diff --git "a/checkpoint-40000/trainer_state.json" "b/checkpoint-40000/trainer_state.json" new file mode 100644--- /dev/null +++ "b/checkpoint-40000/trainer_state.json" @@ -0,0 +1,280077 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.7045121632922653, + "eval_steps": 5867, + "global_step": 40000, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "grad_norm": 3.46875, + "learning_rate": 2.0000000000000003e-06, + "loss": 7.1562, + "step": 1 + }, + { + "epoch": 0.0, + "eval_loss": 7.180609703063965, + "eval_runtime": 80.1197, + "eval_samples_per_second": 18.772, + "eval_steps_per_second": 2.346, + "step": 1 + }, + { + "epoch": 0.0, + "grad_norm": 3.21875, + "learning_rate": 4.000000000000001e-06, + "loss": 7.1001, + "step": 2 + }, + { + "epoch": 0.0, + "grad_norm": 3.359375, + "learning_rate": 6e-06, + "loss": 7.3329, + "step": 3 + }, + { + "epoch": 0.0, + "grad_norm": 3.46875, + "learning_rate": 8.000000000000001e-06, + "loss": 7.2222, + "step": 4 + }, + { + "epoch": 0.0, + "grad_norm": 3.453125, + "learning_rate": 1e-05, + "loss": 7.3313, + "step": 5 + }, + { + "epoch": 0.0, + "grad_norm": 3.234375, + "learning_rate": 1.2e-05, + "loss": 7.0367, + "step": 6 + }, + { + "epoch": 0.0, + "grad_norm": 3.40625, + "learning_rate": 1.4000000000000001e-05, + "loss": 7.1454, + "step": 7 + }, + { + "epoch": 0.0, + "grad_norm": 3.328125, + "learning_rate": 1.6000000000000003e-05, + "loss": 7.1308, + "step": 8 + }, + { + "epoch": 0.0, + "grad_norm": 3.40625, + "learning_rate": 1.8e-05, + "loss": 6.9534, + "step": 9 + }, + { + "epoch": 0.0, + "grad_norm": 3.578125, + "learning_rate": 2e-05, + "loss": 6.9062, + "step": 10 + }, + { + "epoch": 0.0, + "grad_norm": 3.546875, + "learning_rate": 2.2000000000000003e-05, + "loss": 6.6814, + "step": 11 + }, + { + "epoch": 0.0, + "grad_norm": 3.546875, + "learning_rate": 2.4e-05, + "loss": 6.4026, + "step": 12 + }, + { + "epoch": 0.0, + "grad_norm": 3.6875, + "learning_rate": 2.6000000000000002e-05, + "loss": 6.2729, + "step": 13 + }, + { + "epoch": 0.0, + "grad_norm": 3.515625, + "learning_rate": 2.8000000000000003e-05, + "loss": 6.2939, + "step": 14 + }, + { + "epoch": 0.0, + "grad_norm": 3.046875, + "learning_rate": 3e-05, + "loss": 5.6891, + "step": 15 + }, + { + "epoch": 0.0, + "grad_norm": 3.421875, + "learning_rate": 3.2000000000000005e-05, + "loss": 5.6529, + "step": 16 + }, + { + "epoch": 0.0, + "grad_norm": 1.9609375, + "learning_rate": 3.4000000000000007e-05, + "loss": 5.3895, + "step": 17 + }, + { + "epoch": 0.0, + "grad_norm": 2.328125, + "learning_rate": 3.6e-05, + "loss": 5.5974, + "step": 18 + }, + { + "epoch": 0.0, + "grad_norm": 2.96875, + "learning_rate": 3.8e-05, + "loss": 5.6802, + "step": 19 + }, + { + "epoch": 0.0, + "grad_norm": 4.0625, + "learning_rate": 4e-05, + "loss": 5.5493, + "step": 20 + }, + { + "epoch": 0.0, + "grad_norm": 3.65625, + "learning_rate": 4.2e-05, + "loss": 5.3961, + "step": 21 + }, + { + "epoch": 0.0, + "grad_norm": 2.515625, + "learning_rate": 4.4000000000000006e-05, + "loss": 5.1233, + "step": 22 + }, + { + "epoch": 0.0, + "grad_norm": 1.5703125, + "learning_rate": 4.600000000000001e-05, + "loss": 4.8111, + "step": 23 + }, + { + "epoch": 0.0, + "grad_norm": 1.0234375, + "learning_rate": 4.8e-05, + "loss": 4.7203, + "step": 24 + }, + { + "epoch": 0.0, + "grad_norm": 1.609375, + "learning_rate": 5e-05, + "loss": 4.8003, + "step": 25 + }, + { + "epoch": 0.0, + "grad_norm": 1.828125, + "learning_rate": 5.2000000000000004e-05, + "loss": 4.7147, + "step": 26 + }, + { + "epoch": 0.0, + "grad_norm": 1.734375, + "learning_rate": 5.4000000000000005e-05, + "loss": 4.6003, + "step": 27 + }, + { + "epoch": 0.0, + "grad_norm": 1.328125, + "learning_rate": 5.6000000000000006e-05, + "loss": 4.8063, + "step": 28 + }, + { + "epoch": 0.0, + "grad_norm": 1.1640625, + "learning_rate": 5.8e-05, + "loss": 4.6076, + "step": 29 + }, + { + "epoch": 0.0, + "grad_norm": 1.1953125, + "learning_rate": 6e-05, + "loss": 4.568, + "step": 30 + }, + { + "epoch": 0.0, + "grad_norm": 1.703125, + "learning_rate": 6.2e-05, + "loss": 4.4197, + "step": 31 + }, + { + "epoch": 0.0, + "grad_norm": 1.5859375, + "learning_rate": 6.400000000000001e-05, + "loss": 4.2863, + "step": 32 + }, + { + "epoch": 0.0, + "grad_norm": 1.1328125, + "learning_rate": 6.6e-05, + "loss": 4.2801, + "step": 33 + }, + { + "epoch": 0.0, + "grad_norm": 0.578125, + "learning_rate": 6.800000000000001e-05, + "loss": 4.2804, + "step": 34 + }, + { + "epoch": 0.0, + "grad_norm": 0.984375, + "learning_rate": 7e-05, + "loss": 4.2813, + "step": 35 + }, + { + "epoch": 0.0, + "grad_norm": 0.7578125, + "learning_rate": 7.2e-05, + "loss": 4.2116, + "step": 36 + }, + { + "epoch": 0.0, + "grad_norm": 0.6328125, + "learning_rate": 7.4e-05, + "loss": 4.0979, + "step": 37 + }, + { + "epoch": 0.0, + "grad_norm": 0.82421875, + "learning_rate": 7.6e-05, + "loss": 4.0928, + "step": 38 + }, + { + "epoch": 0.0, + "grad_norm": 1.3984375, + "learning_rate": 7.800000000000001e-05, + "loss": 4.052, + "step": 39 + }, + { + "epoch": 0.0, + "grad_norm": 1.1328125, + "learning_rate": 8e-05, + "loss": 4.1638, + "step": 40 + }, + { + "epoch": 0.0, + "grad_norm": 0.80078125, + "learning_rate": 8.2e-05, + "loss": 4.0157, + "step": 41 + }, + { + "epoch": 0.0, + "grad_norm": 0.609375, + "learning_rate": 8.4e-05, + "loss": 4.0236, + "step": 42 + }, + { + "epoch": 0.0, + "grad_norm": 0.54296875, + "learning_rate": 8.6e-05, + "loss": 3.9427, + "step": 43 + }, + { + "epoch": 0.0, + "grad_norm": 0.455078125, + "learning_rate": 8.800000000000001e-05, + "loss": 3.8323, + "step": 44 + }, + { + "epoch": 0.0, + "grad_norm": 0.490234375, + "learning_rate": 9e-05, + "loss": 3.7302, + "step": 45 + }, + { + "epoch": 0.0, + "grad_norm": 0.515625, + "learning_rate": 9.200000000000001e-05, + "loss": 3.8291, + "step": 46 + }, + { + "epoch": 0.0, + "grad_norm": 0.4921875, + "learning_rate": 9.4e-05, + "loss": 3.8055, + "step": 47 + }, + { + "epoch": 0.0, + "grad_norm": 0.494140625, + "learning_rate": 9.6e-05, + "loss": 3.7304, + "step": 48 + }, + { + "epoch": 0.0, + "grad_norm": 0.484375, + "learning_rate": 9.8e-05, + "loss": 3.7091, + "step": 49 + }, + { + "epoch": 0.0, + "grad_norm": 0.57421875, + "learning_rate": 0.0001, + "loss": 3.7711, + "step": 50 + }, + { + "epoch": 0.0, + "grad_norm": 0.58984375, + "learning_rate": 0.00010200000000000001, + "loss": 3.7698, + "step": 51 + }, + { + "epoch": 0.0, + "grad_norm": 0.47265625, + "learning_rate": 0.00010400000000000001, + "loss": 3.7403, + "step": 52 + }, + { + "epoch": 0.0, + "grad_norm": 0.431640625, + "learning_rate": 0.00010600000000000002, + "loss": 3.705, + "step": 53 + }, + { + "epoch": 0.0, + "grad_norm": 0.41796875, + "learning_rate": 0.00010800000000000001, + "loss": 3.6324, + "step": 54 + }, + { + "epoch": 0.0, + "grad_norm": 0.40234375, + "learning_rate": 0.00011000000000000002, + "loss": 3.7317, + "step": 55 + }, + { + "epoch": 0.0, + "grad_norm": 0.4296875, + "learning_rate": 0.00011200000000000001, + "loss": 3.5893, + "step": 56 + }, + { + "epoch": 0.0, + "grad_norm": 0.46875, + "learning_rate": 0.00011399999999999999, + "loss": 3.7702, + "step": 57 + }, + { + "epoch": 0.0, + "grad_norm": 0.462890625, + "learning_rate": 0.000116, + "loss": 3.4647, + "step": 58 + }, + { + "epoch": 0.0, + "grad_norm": 0.4609375, + "learning_rate": 0.000118, + "loss": 3.5091, + "step": 59 + }, + { + "epoch": 0.0, + "grad_norm": 0.6015625, + "learning_rate": 0.00012, + "loss": 3.7022, + "step": 60 + }, + { + "epoch": 0.0, + "grad_norm": 0.404296875, + "learning_rate": 0.000122, + "loss": 3.4509, + "step": 61 + }, + { + "epoch": 0.0, + "grad_norm": 0.50390625, + "learning_rate": 0.000124, + "loss": 3.3392, + "step": 62 + }, + { + "epoch": 0.0, + "grad_norm": 0.48046875, + "learning_rate": 0.000126, + "loss": 3.6135, + "step": 63 + }, + { + "epoch": 0.0, + "grad_norm": 0.51953125, + "learning_rate": 0.00012800000000000002, + "loss": 3.253, + "step": 64 + }, + { + "epoch": 0.0, + "grad_norm": 0.6015625, + "learning_rate": 0.00013000000000000002, + "loss": 3.446, + "step": 65 + }, + { + "epoch": 0.0, + "grad_norm": 0.470703125, + "learning_rate": 0.000132, + "loss": 3.3636, + "step": 66 + }, + { + "epoch": 0.0, + "grad_norm": 0.5625, + "learning_rate": 0.000134, + "loss": 3.5737, + "step": 67 + }, + { + "epoch": 0.0, + "grad_norm": 0.59765625, + "learning_rate": 0.00013600000000000003, + "loss": 3.5012, + "step": 68 + }, + { + "epoch": 0.0, + "grad_norm": 0.462890625, + "learning_rate": 0.000138, + "loss": 3.4315, + "step": 69 + }, + { + "epoch": 0.0, + "grad_norm": 0.80859375, + "learning_rate": 0.00014, + "loss": 3.4633, + "step": 70 + }, + { + "epoch": 0.0, + "grad_norm": 0.70703125, + "learning_rate": 0.000142, + "loss": 3.4118, + "step": 71 + }, + { + "epoch": 0.0, + "grad_norm": 0.42578125, + "learning_rate": 0.000144, + "loss": 3.4363, + "step": 72 + }, + { + "epoch": 0.0, + "grad_norm": 0.482421875, + "learning_rate": 0.000146, + "loss": 3.4456, + "step": 73 + }, + { + "epoch": 0.0, + "grad_norm": 0.419921875, + "learning_rate": 0.000148, + "loss": 3.3302, + "step": 74 + }, + { + "epoch": 0.0, + "grad_norm": 0.53125, + "learning_rate": 0.00015000000000000001, + "loss": 3.4253, + "step": 75 + }, + { + "epoch": 0.0, + "grad_norm": 0.54296875, + "learning_rate": 0.000152, + "loss": 3.2784, + "step": 76 + }, + { + "epoch": 0.0, + "grad_norm": 0.376953125, + "learning_rate": 0.000154, + "loss": 3.2848, + "step": 77 + }, + { + "epoch": 0.0, + "grad_norm": 0.5078125, + "learning_rate": 0.00015600000000000002, + "loss": 3.2216, + "step": 78 + }, + { + "epoch": 0.0, + "grad_norm": 0.458984375, + "learning_rate": 0.00015800000000000002, + "loss": 3.2942, + "step": 79 + }, + { + "epoch": 0.0, + "grad_norm": 0.431640625, + "learning_rate": 0.00016, + "loss": 3.3276, + "step": 80 + }, + { + "epoch": 0.0, + "grad_norm": 0.6875, + "learning_rate": 0.000162, + "loss": 3.2128, + "step": 81 + }, + { + "epoch": 0.0, + "grad_norm": 0.86328125, + "learning_rate": 0.000164, + "loss": 3.2895, + "step": 82 + }, + { + "epoch": 0.0, + "grad_norm": 0.63671875, + "learning_rate": 0.000166, + "loss": 3.1896, + "step": 83 + }, + { + "epoch": 0.0, + "grad_norm": 0.390625, + "learning_rate": 0.000168, + "loss": 3.2439, + "step": 84 + }, + { + "epoch": 0.0, + "grad_norm": 0.65625, + "learning_rate": 0.00017, + "loss": 3.2579, + "step": 85 + }, + { + "epoch": 0.0, + "grad_norm": 0.52734375, + "learning_rate": 0.000172, + "loss": 3.2379, + "step": 86 + }, + { + "epoch": 0.0, + "grad_norm": 0.421875, + "learning_rate": 0.000174, + "loss": 3.1947, + "step": 87 + }, + { + "epoch": 0.0, + "grad_norm": 0.53125, + "learning_rate": 0.00017600000000000002, + "loss": 3.2853, + "step": 88 + }, + { + "epoch": 0.0, + "grad_norm": 0.5546875, + "learning_rate": 0.00017800000000000002, + "loss": 3.1707, + "step": 89 + }, + { + "epoch": 0.0, + "grad_norm": 0.486328125, + "learning_rate": 0.00018, + "loss": 3.2692, + "step": 90 + }, + { + "epoch": 0.0, + "grad_norm": 0.375, + "learning_rate": 0.000182, + "loss": 3.2777, + "step": 91 + }, + { + "epoch": 0.0, + "grad_norm": 0.3515625, + "learning_rate": 0.00018400000000000003, + "loss": 3.2584, + "step": 92 + }, + { + "epoch": 0.0, + "grad_norm": 0.515625, + "learning_rate": 0.00018600000000000002, + "loss": 3.2486, + "step": 93 + }, + { + "epoch": 0.0, + "grad_norm": 0.45703125, + "learning_rate": 0.000188, + "loss": 3.1843, + "step": 94 + }, + { + "epoch": 0.0, + "grad_norm": 0.408203125, + "learning_rate": 0.00019, + "loss": 3.27, + "step": 95 + }, + { + "epoch": 0.0, + "grad_norm": 0.474609375, + "learning_rate": 0.000192, + "loss": 3.213, + "step": 96 + }, + { + "epoch": 0.0, + "grad_norm": 0.40625, + "learning_rate": 0.000194, + "loss": 3.2512, + "step": 97 + }, + { + "epoch": 0.0, + "grad_norm": 0.44921875, + "learning_rate": 0.000196, + "loss": 3.2023, + "step": 98 + }, + { + "epoch": 0.0, + "grad_norm": 0.455078125, + "learning_rate": 0.00019800000000000002, + "loss": 3.2321, + "step": 99 + }, + { + "epoch": 0.0, + "grad_norm": 0.42578125, + "learning_rate": 0.0002, + "loss": 3.3023, + "step": 100 + }, + { + "epoch": 0.0, + "grad_norm": 0.50390625, + "learning_rate": 0.0001999999997750183, + "loss": 3.1008, + "step": 101 + }, + { + "epoch": 0.0, + "grad_norm": 0.67578125, + "learning_rate": 0.0001999999991000731, + "loss": 3.1713, + "step": 102 + }, + { + "epoch": 0.0, + "grad_norm": 0.921875, + "learning_rate": 0.00019999999797516447, + "loss": 3.1691, + "step": 103 + }, + { + "epoch": 0.0, + "grad_norm": 0.90625, + "learning_rate": 0.0001999999964002924, + "loss": 3.1726, + "step": 104 + }, + { + "epoch": 0.0, + "grad_norm": 0.37109375, + "learning_rate": 0.00019999999437545686, + "loss": 3.1777, + "step": 105 + }, + { + "epoch": 0.0, + "grad_norm": 0.76171875, + "learning_rate": 0.00019999999190065796, + "loss": 3.2884, + "step": 106 + }, + { + "epoch": 0.0, + "grad_norm": 0.80078125, + "learning_rate": 0.00019999998897589558, + "loss": 3.2384, + "step": 107 + }, + { + "epoch": 0.0, + "grad_norm": 0.3984375, + "learning_rate": 0.00019999998560116982, + "loss": 3.3567, + "step": 108 + }, + { + "epoch": 0.0, + "grad_norm": 0.64453125, + "learning_rate": 0.00019999998177648064, + "loss": 3.2055, + "step": 109 + }, + { + "epoch": 0.0, + "grad_norm": 0.490234375, + "learning_rate": 0.00019999997750182815, + "loss": 3.1533, + "step": 110 + }, + { + "epoch": 0.0, + "grad_norm": 0.52734375, + "learning_rate": 0.00019999997277721226, + "loss": 3.3401, + "step": 111 + }, + { + "epoch": 0.0, + "grad_norm": 0.625, + "learning_rate": 0.00019999996760263305, + "loss": 3.349, + "step": 112 + }, + { + "epoch": 0.0, + "grad_norm": 0.458984375, + "learning_rate": 0.00019999996197809052, + "loss": 3.2612, + "step": 113 + }, + { + "epoch": 0.0, + "grad_norm": 0.7734375, + "learning_rate": 0.00019999995590358474, + "loss": 3.2335, + "step": 114 + }, + { + "epoch": 0.0, + "grad_norm": 0.4453125, + "learning_rate": 0.00019999994937911568, + "loss": 3.1518, + "step": 115 + }, + { + "epoch": 0.0, + "grad_norm": 0.51953125, + "learning_rate": 0.0001999999424046834, + "loss": 3.1451, + "step": 116 + }, + { + "epoch": 0.0, + "grad_norm": 0.44921875, + "learning_rate": 0.0001999999349802879, + "loss": 3.1221, + "step": 117 + }, + { + "epoch": 0.01, + "grad_norm": 0.4375, + "learning_rate": 0.00019999992710592927, + "loss": 3.1549, + "step": 118 + }, + { + "epoch": 0.01, + "grad_norm": 0.439453125, + "learning_rate": 0.0001999999187816075, + "loss": 3.1547, + "step": 119 + }, + { + "epoch": 0.01, + "grad_norm": 0.39453125, + "learning_rate": 0.00019999991000732264, + "loss": 3.05, + "step": 120 + }, + { + "epoch": 0.01, + "grad_norm": 0.4609375, + "learning_rate": 0.00019999990078307473, + "loss": 3.2157, + "step": 121 + }, + { + "epoch": 0.01, + "grad_norm": 0.419921875, + "learning_rate": 0.0001999998911088638, + "loss": 3.2546, + "step": 122 + }, + { + "epoch": 0.01, + "grad_norm": 0.5390625, + "learning_rate": 0.00019999988098468995, + "loss": 3.1888, + "step": 123 + }, + { + "epoch": 0.01, + "grad_norm": 0.49609375, + "learning_rate": 0.00019999987041055314, + "loss": 3.1858, + "step": 124 + }, + { + "epoch": 0.01, + "grad_norm": 0.455078125, + "learning_rate": 0.00019999985938645346, + "loss": 3.1353, + "step": 125 + }, + { + "epoch": 0.01, + "grad_norm": 0.53125, + "learning_rate": 0.00019999984791239098, + "loss": 3.2962, + "step": 126 + }, + { + "epoch": 0.01, + "grad_norm": 0.49609375, + "learning_rate": 0.00019999983598836574, + "loss": 3.0766, + "step": 127 + }, + { + "epoch": 0.01, + "grad_norm": 0.392578125, + "learning_rate": 0.00019999982361437776, + "loss": 3.0214, + "step": 128 + }, + { + "epoch": 0.01, + "grad_norm": 0.455078125, + "learning_rate": 0.00019999981079042711, + "loss": 3.0065, + "step": 129 + }, + { + "epoch": 0.01, + "grad_norm": 0.375, + "learning_rate": 0.0001999997975165139, + "loss": 2.9795, + "step": 130 + }, + { + "epoch": 0.01, + "grad_norm": 0.470703125, + "learning_rate": 0.0001999997837926381, + "loss": 3.1533, + "step": 131 + }, + { + "epoch": 0.01, + "grad_norm": 0.37890625, + "learning_rate": 0.00019999976961879984, + "loss": 3.1483, + "step": 132 + }, + { + "epoch": 0.01, + "grad_norm": 0.466796875, + "learning_rate": 0.00019999975499499916, + "loss": 3.18, + "step": 133 + }, + { + "epoch": 0.01, + "grad_norm": 0.5078125, + "learning_rate": 0.00019999973992123612, + "loss": 3.1565, + "step": 134 + }, + { + "epoch": 0.01, + "grad_norm": 0.421875, + "learning_rate": 0.00019999972439751083, + "loss": 3.2291, + "step": 135 + }, + { + "epoch": 0.01, + "grad_norm": 0.482421875, + "learning_rate": 0.0001999997084238233, + "loss": 3.2038, + "step": 136 + }, + { + "epoch": 0.01, + "grad_norm": 0.4453125, + "learning_rate": 0.0001999996920001736, + "loss": 3.1005, + "step": 137 + }, + { + "epoch": 0.01, + "grad_norm": 0.3984375, + "learning_rate": 0.00019999967512656185, + "loss": 3.2345, + "step": 138 + }, + { + "epoch": 0.01, + "grad_norm": 0.435546875, + "learning_rate": 0.00019999965780298813, + "loss": 2.9556, + "step": 139 + }, + { + "epoch": 0.01, + "grad_norm": 0.3828125, + "learning_rate": 0.0001999996400294525, + "loss": 3.079, + "step": 140 + }, + { + "epoch": 0.01, + "grad_norm": 0.439453125, + "learning_rate": 0.00019999962180595503, + "loss": 3.0212, + "step": 141 + }, + { + "epoch": 0.01, + "grad_norm": 0.451171875, + "learning_rate": 0.00019999960313249576, + "loss": 3.0247, + "step": 142 + }, + { + "epoch": 0.01, + "grad_norm": 0.41015625, + "learning_rate": 0.00019999958400907487, + "loss": 3.1062, + "step": 143 + }, + { + "epoch": 0.01, + "grad_norm": 0.43359375, + "learning_rate": 0.00019999956443569238, + "loss": 2.9942, + "step": 144 + }, + { + "epoch": 0.01, + "grad_norm": 0.431640625, + "learning_rate": 0.0001999995444123484, + "loss": 3.1007, + "step": 145 + }, + { + "epoch": 0.01, + "grad_norm": 0.439453125, + "learning_rate": 0.00019999952393904305, + "loss": 3.1555, + "step": 146 + }, + { + "epoch": 0.01, + "grad_norm": 0.443359375, + "learning_rate": 0.00019999950301577632, + "loss": 3.1222, + "step": 147 + }, + { + "epoch": 0.01, + "grad_norm": 0.404296875, + "learning_rate": 0.00019999948164254843, + "loss": 3.0358, + "step": 148 + }, + { + "epoch": 0.01, + "grad_norm": 0.5, + "learning_rate": 0.0001999994598193594, + "loss": 3.0916, + "step": 149 + }, + { + "epoch": 0.01, + "grad_norm": 0.404296875, + "learning_rate": 0.00019999943754620933, + "loss": 3.1161, + "step": 150 + }, + { + "epoch": 0.01, + "grad_norm": 0.416015625, + "learning_rate": 0.00019999941482309836, + "loss": 3.0201, + "step": 151 + }, + { + "epoch": 0.01, + "grad_norm": 0.42578125, + "learning_rate": 0.00019999939165002656, + "loss": 3.0233, + "step": 152 + }, + { + "epoch": 0.01, + "grad_norm": 0.380859375, + "learning_rate": 0.00019999936802699405, + "loss": 3.0434, + "step": 153 + }, + { + "epoch": 0.01, + "grad_norm": 0.515625, + "learning_rate": 0.00019999934395400088, + "loss": 3.1672, + "step": 154 + }, + { + "epoch": 0.01, + "grad_norm": 0.42578125, + "learning_rate": 0.00019999931943104725, + "loss": 3.1117, + "step": 155 + }, + { + "epoch": 0.01, + "grad_norm": 0.41796875, + "learning_rate": 0.00019999929445813324, + "loss": 2.9897, + "step": 156 + }, + { + "epoch": 0.01, + "grad_norm": 0.37890625, + "learning_rate": 0.00019999926903525894, + "loss": 2.9579, + "step": 157 + }, + { + "epoch": 0.01, + "grad_norm": 0.58203125, + "learning_rate": 0.00019999924316242446, + "loss": 3.0794, + "step": 158 + }, + { + "epoch": 0.01, + "grad_norm": 0.42578125, + "learning_rate": 0.00019999921683962993, + "loss": 3.1814, + "step": 159 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.0001999991900668755, + "loss": 3.1038, + "step": 160 + }, + { + "epoch": 0.01, + "grad_norm": 0.416015625, + "learning_rate": 0.00019999916284416125, + "loss": 3.2264, + "step": 161 + }, + { + "epoch": 0.01, + "grad_norm": 0.42578125, + "learning_rate": 0.0001999991351714873, + "loss": 2.9753, + "step": 162 + }, + { + "epoch": 0.01, + "grad_norm": 0.396484375, + "learning_rate": 0.0001999991070488538, + "loss": 3.187, + "step": 163 + }, + { + "epoch": 0.01, + "grad_norm": 0.419921875, + "learning_rate": 0.00019999907847626086, + "loss": 3.1198, + "step": 164 + }, + { + "epoch": 0.01, + "grad_norm": 0.515625, + "learning_rate": 0.0001999990494537086, + "loss": 3.1058, + "step": 165 + }, + { + "epoch": 0.01, + "grad_norm": 0.439453125, + "learning_rate": 0.00019999901998119716, + "loss": 3.1592, + "step": 166 + }, + { + "epoch": 0.01, + "grad_norm": 0.392578125, + "learning_rate": 0.00019999899005872668, + "loss": 2.9997, + "step": 167 + }, + { + "epoch": 0.01, + "grad_norm": 0.5, + "learning_rate": 0.00019999895968629731, + "loss": 3.0619, + "step": 168 + }, + { + "epoch": 0.01, + "grad_norm": 0.439453125, + "learning_rate": 0.00019999892886390916, + "loss": 3.0711, + "step": 169 + }, + { + "epoch": 0.01, + "grad_norm": 0.388671875, + "learning_rate": 0.00019999889759156236, + "loss": 2.9735, + "step": 170 + }, + { + "epoch": 0.01, + "grad_norm": 0.498046875, + "learning_rate": 0.00019999886586925708, + "loss": 3.0384, + "step": 171 + }, + { + "epoch": 0.01, + "grad_norm": 0.4296875, + "learning_rate": 0.00019999883369699344, + "loss": 3.0791, + "step": 172 + }, + { + "epoch": 0.01, + "grad_norm": 0.47265625, + "learning_rate": 0.0001999988010747716, + "loss": 3.0633, + "step": 173 + }, + { + "epoch": 0.01, + "grad_norm": 0.54296875, + "learning_rate": 0.0001999987680025917, + "loss": 3.1706, + "step": 174 + }, + { + "epoch": 0.01, + "grad_norm": 0.38671875, + "learning_rate": 0.0001999987344804539, + "loss": 2.9778, + "step": 175 + }, + { + "epoch": 0.01, + "grad_norm": 0.75, + "learning_rate": 0.00019999870050835833, + "loss": 3.1589, + "step": 176 + }, + { + "epoch": 0.01, + "grad_norm": 0.57421875, + "learning_rate": 0.00019999866608630512, + "loss": 2.9964, + "step": 177 + }, + { + "epoch": 0.01, + "grad_norm": 0.5078125, + "learning_rate": 0.00019999863121429452, + "loss": 3.1404, + "step": 178 + }, + { + "epoch": 0.01, + "grad_norm": 0.625, + "learning_rate": 0.0001999985958923266, + "loss": 2.9679, + "step": 179 + }, + { + "epoch": 0.01, + "grad_norm": 0.390625, + "learning_rate": 0.00019999856012040154, + "loss": 2.9801, + "step": 180 + }, + { + "epoch": 0.01, + "grad_norm": 0.69140625, + "learning_rate": 0.0001999985238985195, + "loss": 3.0648, + "step": 181 + }, + { + "epoch": 0.01, + "grad_norm": 0.40625, + "learning_rate": 0.00019999848722668064, + "loss": 2.9979, + "step": 182 + }, + { + "epoch": 0.01, + "grad_norm": 0.6171875, + "learning_rate": 0.00019999845010488514, + "loss": 3.0607, + "step": 183 + }, + { + "epoch": 0.01, + "grad_norm": 0.55859375, + "learning_rate": 0.00019999841253313317, + "loss": 2.9959, + "step": 184 + }, + { + "epoch": 0.01, + "grad_norm": 0.451171875, + "learning_rate": 0.0001999983745114249, + "loss": 3.0744, + "step": 185 + }, + { + "epoch": 0.01, + "grad_norm": 0.5703125, + "learning_rate": 0.00019999833603976045, + "loss": 2.9898, + "step": 186 + }, + { + "epoch": 0.01, + "grad_norm": 0.384765625, + "learning_rate": 0.00019999829711814005, + "loss": 3.0758, + "step": 187 + }, + { + "epoch": 0.01, + "grad_norm": 0.515625, + "learning_rate": 0.00019999825774656387, + "loss": 3.1186, + "step": 188 + }, + { + "epoch": 0.01, + "grad_norm": 0.3828125, + "learning_rate": 0.00019999821792503206, + "loss": 3.0504, + "step": 189 + }, + { + "epoch": 0.01, + "grad_norm": 0.51171875, + "learning_rate": 0.00019999817765354485, + "loss": 3.0273, + "step": 190 + }, + { + "epoch": 0.01, + "grad_norm": 0.3515625, + "learning_rate": 0.00019999813693210235, + "loss": 2.9842, + "step": 191 + }, + { + "epoch": 0.01, + "grad_norm": 0.478515625, + "learning_rate": 0.0001999980957607048, + "loss": 3.0415, + "step": 192 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.00019999805413935232, + "loss": 2.9458, + "step": 193 + }, + { + "epoch": 0.01, + "grad_norm": 0.447265625, + "learning_rate": 0.0001999980120680452, + "loss": 2.9704, + "step": 194 + }, + { + "epoch": 0.01, + "grad_norm": 0.396484375, + "learning_rate": 0.0001999979695467835, + "loss": 3.1304, + "step": 195 + }, + { + "epoch": 0.01, + "grad_norm": 0.390625, + "learning_rate": 0.00019999792657556756, + "loss": 3.0718, + "step": 196 + }, + { + "epoch": 0.01, + "grad_norm": 0.41015625, + "learning_rate": 0.00019999788315439748, + "loss": 2.9656, + "step": 197 + }, + { + "epoch": 0.01, + "grad_norm": 0.443359375, + "learning_rate": 0.00019999783928327343, + "loss": 3.0236, + "step": 198 + }, + { + "epoch": 0.01, + "grad_norm": 0.421875, + "learning_rate": 0.00019999779496219567, + "loss": 3.0188, + "step": 199 + }, + { + "epoch": 0.01, + "grad_norm": 0.392578125, + "learning_rate": 0.0001999977501911644, + "loss": 2.8401, + "step": 200 + }, + { + "epoch": 0.01, + "grad_norm": 0.384765625, + "learning_rate": 0.00019999770497017974, + "loss": 2.9852, + "step": 201 + }, + { + "epoch": 0.01, + "grad_norm": 0.3828125, + "learning_rate": 0.000199997659299242, + "loss": 2.9811, + "step": 202 + }, + { + "epoch": 0.01, + "grad_norm": 0.361328125, + "learning_rate": 0.00019999761317835132, + "loss": 2.9913, + "step": 203 + }, + { + "epoch": 0.01, + "grad_norm": 0.431640625, + "learning_rate": 0.00019999756660750795, + "loss": 3.0011, + "step": 204 + }, + { + "epoch": 0.01, + "grad_norm": 0.3984375, + "learning_rate": 0.00019999751958671205, + "loss": 3.1635, + "step": 205 + }, + { + "epoch": 0.01, + "grad_norm": 0.5625, + "learning_rate": 0.00019999747211596388, + "loss": 2.997, + "step": 206 + }, + { + "epoch": 0.01, + "grad_norm": 0.435546875, + "learning_rate": 0.00019999742419526362, + "loss": 3.1379, + "step": 207 + }, + { + "epoch": 0.01, + "grad_norm": 0.515625, + "learning_rate": 0.0001999973758246115, + "loss": 3.1223, + "step": 208 + }, + { + "epoch": 0.01, + "grad_norm": 0.421875, + "learning_rate": 0.0001999973270040077, + "loss": 3.0984, + "step": 209 + }, + { + "epoch": 0.01, + "grad_norm": 0.478515625, + "learning_rate": 0.00019999727773345252, + "loss": 2.9996, + "step": 210 + }, + { + "epoch": 0.01, + "grad_norm": 0.44140625, + "learning_rate": 0.0001999972280129461, + "loss": 2.9018, + "step": 211 + }, + { + "epoch": 0.01, + "grad_norm": 0.43359375, + "learning_rate": 0.00019999717784248872, + "loss": 3.1169, + "step": 212 + }, + { + "epoch": 0.01, + "grad_norm": 0.44140625, + "learning_rate": 0.00019999712722208057, + "loss": 3.0597, + "step": 213 + }, + { + "epoch": 0.01, + "grad_norm": 0.404296875, + "learning_rate": 0.0001999970761517219, + "loss": 3.0226, + "step": 214 + }, + { + "epoch": 0.01, + "grad_norm": 0.400390625, + "learning_rate": 0.00019999702463141295, + "loss": 2.8395, + "step": 215 + }, + { + "epoch": 0.01, + "grad_norm": 0.396484375, + "learning_rate": 0.0001999969726611539, + "loss": 2.9025, + "step": 216 + }, + { + "epoch": 0.01, + "grad_norm": 0.400390625, + "learning_rate": 0.00019999692024094502, + "loss": 3.0649, + "step": 217 + }, + { + "epoch": 0.01, + "grad_norm": 0.412109375, + "learning_rate": 0.00019999686737078658, + "loss": 3.0628, + "step": 218 + }, + { + "epoch": 0.01, + "grad_norm": 0.4453125, + "learning_rate": 0.00019999681405067875, + "loss": 3.076, + "step": 219 + }, + { + "epoch": 0.01, + "grad_norm": 0.37890625, + "learning_rate": 0.00019999676028062183, + "loss": 2.9837, + "step": 220 + }, + { + "epoch": 0.01, + "grad_norm": 0.4140625, + "learning_rate": 0.000199996706060616, + "loss": 3.0668, + "step": 221 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.00019999665139066154, + "loss": 2.8893, + "step": 222 + }, + { + "epoch": 0.01, + "grad_norm": 0.373046875, + "learning_rate": 0.0001999965962707587, + "loss": 2.8354, + "step": 223 + }, + { + "epoch": 0.01, + "grad_norm": 0.50390625, + "learning_rate": 0.00019999654070090774, + "loss": 3.0428, + "step": 224 + }, + { + "epoch": 0.01, + "grad_norm": 0.3828125, + "learning_rate": 0.0001999964846811089, + "loss": 3.0063, + "step": 225 + }, + { + "epoch": 0.01, + "grad_norm": 0.40625, + "learning_rate": 0.00019999642821136237, + "loss": 2.9854, + "step": 226 + }, + { + "epoch": 0.01, + "grad_norm": 0.333984375, + "learning_rate": 0.0001999963712916685, + "loss": 2.893, + "step": 227 + }, + { + "epoch": 0.01, + "grad_norm": 0.455078125, + "learning_rate": 0.00019999631392202748, + "loss": 3.0147, + "step": 228 + }, + { + "epoch": 0.01, + "grad_norm": 0.375, + "learning_rate": 0.0001999962561024396, + "loss": 2.8574, + "step": 229 + }, + { + "epoch": 0.01, + "grad_norm": 0.39453125, + "learning_rate": 0.00019999619783290514, + "loss": 2.903, + "step": 230 + }, + { + "epoch": 0.01, + "grad_norm": 0.388671875, + "learning_rate": 0.00019999613911342429, + "loss": 3.0274, + "step": 231 + }, + { + "epoch": 0.01, + "grad_norm": 0.427734375, + "learning_rate": 0.00019999607994399739, + "loss": 2.9789, + "step": 232 + }, + { + "epoch": 0.01, + "grad_norm": 0.390625, + "learning_rate": 0.00019999602032462466, + "loss": 3.0969, + "step": 233 + }, + { + "epoch": 0.01, + "grad_norm": 0.447265625, + "learning_rate": 0.00019999596025530638, + "loss": 2.9564, + "step": 234 + }, + { + "epoch": 0.01, + "grad_norm": 0.490234375, + "learning_rate": 0.00019999589973604284, + "loss": 3.0853, + "step": 235 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.00019999583876683427, + "loss": 2.9917, + "step": 236 + }, + { + "epoch": 0.01, + "grad_norm": 0.4296875, + "learning_rate": 0.000199995777347681, + "loss": 3.0586, + "step": 237 + }, + { + "epoch": 0.01, + "grad_norm": 0.435546875, + "learning_rate": 0.00019999571547858324, + "loss": 3.0314, + "step": 238 + }, + { + "epoch": 0.01, + "grad_norm": 0.44140625, + "learning_rate": 0.00019999565315954133, + "loss": 2.8872, + "step": 239 + }, + { + "epoch": 0.01, + "grad_norm": 0.439453125, + "learning_rate": 0.00019999559039055552, + "loss": 2.9112, + "step": 240 + }, + { + "epoch": 0.01, + "grad_norm": 0.416015625, + "learning_rate": 0.0001999955271716261, + "loss": 2.9389, + "step": 241 + }, + { + "epoch": 0.01, + "grad_norm": 0.37109375, + "learning_rate": 0.00019999546350275336, + "loss": 2.8747, + "step": 242 + }, + { + "epoch": 0.01, + "grad_norm": 0.458984375, + "learning_rate": 0.00019999539938393755, + "loss": 3.1738, + "step": 243 + }, + { + "epoch": 0.01, + "grad_norm": 0.36328125, + "learning_rate": 0.000199995334815179, + "loss": 2.9884, + "step": 244 + }, + { + "epoch": 0.01, + "grad_norm": 0.40625, + "learning_rate": 0.00019999526979647803, + "loss": 3.0609, + "step": 245 + }, + { + "epoch": 0.01, + "grad_norm": 0.435546875, + "learning_rate": 0.00019999520432783483, + "loss": 2.8316, + "step": 246 + }, + { + "epoch": 0.01, + "grad_norm": 0.392578125, + "learning_rate": 0.00019999513840924976, + "loss": 2.9614, + "step": 247 + }, + { + "epoch": 0.01, + "grad_norm": 0.53125, + "learning_rate": 0.00019999507204072314, + "loss": 2.967, + "step": 248 + }, + { + "epoch": 0.01, + "grad_norm": 0.4140625, + "learning_rate": 0.00019999500522225523, + "loss": 2.9058, + "step": 249 + }, + { + "epoch": 0.01, + "grad_norm": 0.462890625, + "learning_rate": 0.00019999493795384634, + "loss": 3.0447, + "step": 250 + }, + { + "epoch": 0.01, + "grad_norm": 0.404296875, + "learning_rate": 0.00019999487023549678, + "loss": 2.9499, + "step": 251 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.00019999480206720684, + "loss": 2.9084, + "step": 252 + }, + { + "epoch": 0.01, + "grad_norm": 0.380859375, + "learning_rate": 0.00019999473344897682, + "loss": 2.8387, + "step": 253 + }, + { + "epoch": 0.01, + "grad_norm": 0.494140625, + "learning_rate": 0.00019999466438080704, + "loss": 2.9222, + "step": 254 + }, + { + "epoch": 0.01, + "grad_norm": 0.357421875, + "learning_rate": 0.00019999459486269786, + "loss": 2.8517, + "step": 255 + }, + { + "epoch": 0.01, + "grad_norm": 0.49609375, + "learning_rate": 0.00019999452489464952, + "loss": 2.9447, + "step": 256 + }, + { + "epoch": 0.01, + "grad_norm": 0.3828125, + "learning_rate": 0.00019999445447666235, + "loss": 3.0335, + "step": 257 + }, + { + "epoch": 0.01, + "grad_norm": 0.435546875, + "learning_rate": 0.0001999943836087367, + "loss": 3.0047, + "step": 258 + }, + { + "epoch": 0.01, + "grad_norm": 0.373046875, + "learning_rate": 0.00019999431229087287, + "loss": 2.8986, + "step": 259 + }, + { + "epoch": 0.01, + "grad_norm": 0.44921875, + "learning_rate": 0.00019999424052307113, + "loss": 2.8396, + "step": 260 + }, + { + "epoch": 0.01, + "grad_norm": 0.412109375, + "learning_rate": 0.0001999941683053319, + "loss": 2.9557, + "step": 261 + }, + { + "epoch": 0.01, + "grad_norm": 0.390625, + "learning_rate": 0.00019999409563765545, + "loss": 2.8918, + "step": 262 + }, + { + "epoch": 0.01, + "grad_norm": 0.41015625, + "learning_rate": 0.00019999402252004212, + "loss": 3.0669, + "step": 263 + }, + { + "epoch": 0.01, + "grad_norm": 0.310546875, + "learning_rate": 0.0001999939489524922, + "loss": 2.911, + "step": 264 + }, + { + "epoch": 0.01, + "grad_norm": 0.4453125, + "learning_rate": 0.00019999387493500605, + "loss": 2.8925, + "step": 265 + }, + { + "epoch": 0.01, + "grad_norm": 0.392578125, + "learning_rate": 0.00019999380046758404, + "loss": 2.8713, + "step": 266 + }, + { + "epoch": 0.01, + "grad_norm": 0.490234375, + "learning_rate": 0.00019999372555022646, + "loss": 3.0231, + "step": 267 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.00019999365018293368, + "loss": 3.0208, + "step": 268 + }, + { + "epoch": 0.01, + "grad_norm": 0.50390625, + "learning_rate": 0.000199993574365706, + "loss": 2.9675, + "step": 269 + }, + { + "epoch": 0.01, + "grad_norm": 0.41015625, + "learning_rate": 0.00019999349809854376, + "loss": 2.9086, + "step": 270 + }, + { + "epoch": 0.01, + "grad_norm": 0.48046875, + "learning_rate": 0.00019999342138144735, + "loss": 3.058, + "step": 271 + }, + { + "epoch": 0.01, + "grad_norm": 0.435546875, + "learning_rate": 0.00019999334421441707, + "loss": 2.9038, + "step": 272 + }, + { + "epoch": 0.01, + "grad_norm": 0.4375, + "learning_rate": 0.00019999326659745327, + "loss": 2.9405, + "step": 273 + }, + { + "epoch": 0.01, + "grad_norm": 0.45703125, + "learning_rate": 0.00019999318853055633, + "loss": 2.9583, + "step": 274 + }, + { + "epoch": 0.01, + "grad_norm": 0.359375, + "learning_rate": 0.00019999311001372657, + "loss": 2.8592, + "step": 275 + }, + { + "epoch": 0.01, + "grad_norm": 0.54296875, + "learning_rate": 0.00019999303104696438, + "loss": 2.9743, + "step": 276 + }, + { + "epoch": 0.01, + "grad_norm": 0.37890625, + "learning_rate": 0.0001999929516302701, + "loss": 3.0254, + "step": 277 + }, + { + "epoch": 0.01, + "grad_norm": 0.46875, + "learning_rate": 0.00019999287176364407, + "loss": 2.817, + "step": 278 + }, + { + "epoch": 0.01, + "grad_norm": 0.376953125, + "learning_rate": 0.00019999279144708668, + "loss": 2.9325, + "step": 279 + }, + { + "epoch": 0.01, + "grad_norm": 0.416015625, + "learning_rate": 0.00019999271068059824, + "loss": 2.9373, + "step": 280 + }, + { + "epoch": 0.01, + "grad_norm": 0.36328125, + "learning_rate": 0.00019999262946417915, + "loss": 2.819, + "step": 281 + }, + { + "epoch": 0.01, + "grad_norm": 0.3671875, + "learning_rate": 0.00019999254779782977, + "loss": 2.967, + "step": 282 + }, + { + "epoch": 0.01, + "grad_norm": 0.40234375, + "learning_rate": 0.00019999246568155047, + "loss": 3.104, + "step": 283 + }, + { + "epoch": 0.01, + "grad_norm": 0.34375, + "learning_rate": 0.00019999238311534164, + "loss": 2.8445, + "step": 284 + }, + { + "epoch": 0.01, + "grad_norm": 0.45703125, + "learning_rate": 0.00019999230009920362, + "loss": 3.1596, + "step": 285 + }, + { + "epoch": 0.01, + "grad_norm": 0.359375, + "learning_rate": 0.00019999221663313677, + "loss": 2.9223, + "step": 286 + }, + { + "epoch": 0.01, + "grad_norm": 0.5078125, + "learning_rate": 0.0001999921327171415, + "loss": 2.9239, + "step": 287 + }, + { + "epoch": 0.01, + "grad_norm": 0.4453125, + "learning_rate": 0.00019999204835121818, + "loss": 2.7919, + "step": 288 + }, + { + "epoch": 0.01, + "grad_norm": 0.41796875, + "learning_rate": 0.00019999196353536717, + "loss": 2.9704, + "step": 289 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.0001999918782695889, + "loss": 2.9066, + "step": 290 + }, + { + "epoch": 0.01, + "grad_norm": 0.416015625, + "learning_rate": 0.0001999917925538837, + "loss": 2.9115, + "step": 291 + }, + { + "epoch": 0.01, + "grad_norm": 0.443359375, + "learning_rate": 0.00019999170638825199, + "loss": 2.9575, + "step": 292 + }, + { + "epoch": 0.01, + "grad_norm": 0.4375, + "learning_rate": 0.0001999916197726941, + "loss": 3.0282, + "step": 293 + }, + { + "epoch": 0.01, + "grad_norm": 0.42578125, + "learning_rate": 0.0001999915327072105, + "loss": 3.0804, + "step": 294 + }, + { + "epoch": 0.01, + "grad_norm": 0.365234375, + "learning_rate": 0.00019999144519180156, + "loss": 2.8288, + "step": 295 + }, + { + "epoch": 0.01, + "grad_norm": 0.42578125, + "learning_rate": 0.00019999135722646762, + "loss": 2.9392, + "step": 296 + }, + { + "epoch": 0.01, + "grad_norm": 0.37890625, + "learning_rate": 0.00019999126881120916, + "loss": 3.0594, + "step": 297 + }, + { + "epoch": 0.01, + "grad_norm": 0.4140625, + "learning_rate": 0.00019999117994602648, + "loss": 2.8726, + "step": 298 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.00019999109063092006, + "loss": 3.0382, + "step": 299 + }, + { + "epoch": 0.01, + "grad_norm": 0.3828125, + "learning_rate": 0.0001999910008658903, + "loss": 2.8011, + "step": 300 + }, + { + "epoch": 0.01, + "grad_norm": 0.4140625, + "learning_rate": 0.00019999091065093754, + "loss": 3.0393, + "step": 301 + }, + { + "epoch": 0.01, + "grad_norm": 0.421875, + "learning_rate": 0.00019999081998606225, + "loss": 2.8949, + "step": 302 + }, + { + "epoch": 0.01, + "grad_norm": 0.35546875, + "learning_rate": 0.0001999907288712648, + "loss": 2.8561, + "step": 303 + }, + { + "epoch": 0.01, + "grad_norm": 0.390625, + "learning_rate": 0.0001999906373065456, + "loss": 2.8422, + "step": 304 + }, + { + "epoch": 0.01, + "grad_norm": 0.439453125, + "learning_rate": 0.0001999905452919051, + "loss": 3.1119, + "step": 305 + }, + { + "epoch": 0.01, + "grad_norm": 0.36328125, + "learning_rate": 0.00019999045282734364, + "loss": 2.7697, + "step": 306 + }, + { + "epoch": 0.01, + "grad_norm": 0.388671875, + "learning_rate": 0.00019999035991286174, + "loss": 2.929, + "step": 307 + }, + { + "epoch": 0.01, + "grad_norm": 0.345703125, + "learning_rate": 0.00019999026654845975, + "loss": 2.9842, + "step": 308 + }, + { + "epoch": 0.01, + "grad_norm": 0.4140625, + "learning_rate": 0.0001999901727341381, + "loss": 2.891, + "step": 309 + }, + { + "epoch": 0.01, + "grad_norm": 0.4140625, + "learning_rate": 0.0001999900784698972, + "loss": 2.9519, + "step": 310 + }, + { + "epoch": 0.01, + "grad_norm": 0.447265625, + "learning_rate": 0.00019998998375573751, + "loss": 2.9289, + "step": 311 + }, + { + "epoch": 0.01, + "grad_norm": 0.40625, + "learning_rate": 0.00019998988859165943, + "loss": 3.0018, + "step": 312 + }, + { + "epoch": 0.01, + "grad_norm": 0.5546875, + "learning_rate": 0.00019998979297766338, + "loss": 2.973, + "step": 313 + }, + { + "epoch": 0.01, + "grad_norm": 0.41796875, + "learning_rate": 0.00019998969691374983, + "loss": 3.058, + "step": 314 + }, + { + "epoch": 0.01, + "grad_norm": 0.58203125, + "learning_rate": 0.00019998960039991915, + "loss": 2.9068, + "step": 315 + }, + { + "epoch": 0.01, + "grad_norm": 0.44140625, + "learning_rate": 0.00019998950343617185, + "loss": 2.9679, + "step": 316 + }, + { + "epoch": 0.01, + "grad_norm": 0.490234375, + "learning_rate": 0.0001999894060225083, + "loss": 2.9917, + "step": 317 + }, + { + "epoch": 0.01, + "grad_norm": 0.5234375, + "learning_rate": 0.00019998930815892898, + "loss": 2.8351, + "step": 318 + }, + { + "epoch": 0.01, + "grad_norm": 0.466796875, + "learning_rate": 0.0001999892098454343, + "loss": 3.0294, + "step": 319 + }, + { + "epoch": 0.01, + "grad_norm": 0.515625, + "learning_rate": 0.0001999891110820247, + "loss": 2.9669, + "step": 320 + }, + { + "epoch": 0.01, + "grad_norm": 0.357421875, + "learning_rate": 0.00019998901186870068, + "loss": 3.1359, + "step": 321 + }, + { + "epoch": 0.01, + "grad_norm": 0.408203125, + "learning_rate": 0.00019998891220546265, + "loss": 2.9543, + "step": 322 + }, + { + "epoch": 0.01, + "grad_norm": 0.39453125, + "learning_rate": 0.00019998881209231102, + "loss": 2.8291, + "step": 323 + }, + { + "epoch": 0.01, + "grad_norm": 0.578125, + "learning_rate": 0.00019998871152924629, + "loss": 2.8852, + "step": 324 + }, + { + "epoch": 0.01, + "grad_norm": 0.435546875, + "learning_rate": 0.00019998861051626892, + "loss": 2.9124, + "step": 325 + }, + { + "epoch": 0.01, + "grad_norm": 0.435546875, + "learning_rate": 0.00019998850905337935, + "loss": 3.0296, + "step": 326 + }, + { + "epoch": 0.01, + "grad_norm": 0.400390625, + "learning_rate": 0.00019998840714057802, + "loss": 2.9988, + "step": 327 + }, + { + "epoch": 0.01, + "grad_norm": 0.4296875, + "learning_rate": 0.00019998830477786536, + "loss": 2.9692, + "step": 328 + }, + { + "epoch": 0.01, + "grad_norm": 0.44140625, + "learning_rate": 0.00019998820196524193, + "loss": 2.8354, + "step": 329 + }, + { + "epoch": 0.01, + "grad_norm": 0.388671875, + "learning_rate": 0.0001999880987027081, + "loss": 2.8584, + "step": 330 + }, + { + "epoch": 0.01, + "grad_norm": 0.42578125, + "learning_rate": 0.00019998799499026438, + "loss": 2.8714, + "step": 331 + }, + { + "epoch": 0.01, + "grad_norm": 0.337890625, + "learning_rate": 0.0001999878908279112, + "loss": 2.7848, + "step": 332 + }, + { + "epoch": 0.01, + "grad_norm": 0.41015625, + "learning_rate": 0.00019998778621564906, + "loss": 2.8968, + "step": 333 + }, + { + "epoch": 0.01, + "grad_norm": 0.390625, + "learning_rate": 0.00019998768115347842, + "loss": 2.8181, + "step": 334 + }, + { + "epoch": 0.01, + "grad_norm": 0.40234375, + "learning_rate": 0.0001999875756413998, + "loss": 3.0017, + "step": 335 + }, + { + "epoch": 0.01, + "grad_norm": 0.388671875, + "learning_rate": 0.0001999874696794136, + "loss": 2.9875, + "step": 336 + }, + { + "epoch": 0.01, + "grad_norm": 0.42578125, + "learning_rate": 0.00019998736326752033, + "loss": 2.8357, + "step": 337 + }, + { + "epoch": 0.01, + "grad_norm": 0.37890625, + "learning_rate": 0.00019998725640572044, + "loss": 2.9211, + "step": 338 + }, + { + "epoch": 0.01, + "grad_norm": 0.3984375, + "learning_rate": 0.00019998714909401447, + "loss": 2.9469, + "step": 339 + }, + { + "epoch": 0.01, + "grad_norm": 0.451171875, + "learning_rate": 0.00019998704133240288, + "loss": 2.9184, + "step": 340 + }, + { + "epoch": 0.01, + "grad_norm": 0.3515625, + "learning_rate": 0.00019998693312088612, + "loss": 2.9064, + "step": 341 + }, + { + "epoch": 0.01, + "grad_norm": 0.474609375, + "learning_rate": 0.0001999868244594647, + "loss": 2.9196, + "step": 342 + }, + { + "epoch": 0.01, + "grad_norm": 0.361328125, + "learning_rate": 0.00019998671534813913, + "loss": 2.7997, + "step": 343 + }, + { + "epoch": 0.01, + "grad_norm": 0.392578125, + "learning_rate": 0.00019998660578690987, + "loss": 2.8879, + "step": 344 + }, + { + "epoch": 0.01, + "grad_norm": 0.3671875, + "learning_rate": 0.00019998649577577743, + "loss": 3.0988, + "step": 345 + }, + { + "epoch": 0.01, + "grad_norm": 0.37890625, + "learning_rate": 0.0001999863853147423, + "loss": 2.934, + "step": 346 + }, + { + "epoch": 0.01, + "grad_norm": 0.310546875, + "learning_rate": 0.00019998627440380494, + "loss": 2.8819, + "step": 347 + }, + { + "epoch": 0.01, + "grad_norm": 0.423828125, + "learning_rate": 0.00019998616304296593, + "loss": 2.8644, + "step": 348 + }, + { + "epoch": 0.01, + "grad_norm": 0.376953125, + "learning_rate": 0.00019998605123222572, + "loss": 2.9251, + "step": 349 + }, + { + "epoch": 0.01, + "grad_norm": 0.369140625, + "learning_rate": 0.00019998593897158483, + "loss": 2.9905, + "step": 350 + }, + { + "epoch": 0.01, + "grad_norm": 0.400390625, + "learning_rate": 0.0001999858262610437, + "loss": 2.9717, + "step": 351 + }, + { + "epoch": 0.01, + "grad_norm": 0.388671875, + "learning_rate": 0.00019998571310060297, + "loss": 2.9464, + "step": 352 + }, + { + "epoch": 0.02, + "grad_norm": 0.416015625, + "learning_rate": 0.00019998559949026303, + "loss": 2.8882, + "step": 353 + }, + { + "epoch": 0.02, + "grad_norm": 0.40234375, + "learning_rate": 0.00019998548543002446, + "loss": 2.9792, + "step": 354 + }, + { + "epoch": 0.02, + "grad_norm": 0.380859375, + "learning_rate": 0.0001999853709198877, + "loss": 2.8253, + "step": 355 + }, + { + "epoch": 0.02, + "grad_norm": 0.357421875, + "learning_rate": 0.00019998525595985335, + "loss": 2.7395, + "step": 356 + }, + { + "epoch": 0.02, + "grad_norm": 0.376953125, + "learning_rate": 0.00019998514054992184, + "loss": 2.927, + "step": 357 + }, + { + "epoch": 0.02, + "grad_norm": 0.470703125, + "learning_rate": 0.0001999850246900938, + "loss": 2.9253, + "step": 358 + }, + { + "epoch": 0.02, + "grad_norm": 0.33984375, + "learning_rate": 0.00019998490838036965, + "loss": 2.8672, + "step": 359 + }, + { + "epoch": 0.02, + "grad_norm": 0.412109375, + "learning_rate": 0.00019998479162074996, + "loss": 2.9752, + "step": 360 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.00019998467441123525, + "loss": 2.865, + "step": 361 + }, + { + "epoch": 0.02, + "grad_norm": 0.4140625, + "learning_rate": 0.00019998455675182607, + "loss": 2.6808, + "step": 362 + }, + { + "epoch": 0.02, + "grad_norm": 0.40234375, + "learning_rate": 0.00019998443864252289, + "loss": 2.9794, + "step": 363 + }, + { + "epoch": 0.02, + "grad_norm": 0.3515625, + "learning_rate": 0.00019998432008332627, + "loss": 2.9561, + "step": 364 + }, + { + "epoch": 0.02, + "grad_norm": 0.380859375, + "learning_rate": 0.0001999842010742368, + "loss": 2.9177, + "step": 365 + }, + { + "epoch": 0.02, + "grad_norm": 0.34375, + "learning_rate": 0.00019998408161525492, + "loss": 2.606, + "step": 366 + }, + { + "epoch": 0.02, + "grad_norm": 0.3671875, + "learning_rate": 0.00019998396170638124, + "loss": 2.7952, + "step": 367 + }, + { + "epoch": 0.02, + "grad_norm": 0.3828125, + "learning_rate": 0.00019998384134761626, + "loss": 2.8876, + "step": 368 + }, + { + "epoch": 0.02, + "grad_norm": 0.365234375, + "learning_rate": 0.00019998372053896054, + "loss": 2.7478, + "step": 369 + }, + { + "epoch": 0.02, + "grad_norm": 0.349609375, + "learning_rate": 0.00019998359928041459, + "loss": 2.9535, + "step": 370 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.00019998347757197902, + "loss": 2.9289, + "step": 371 + }, + { + "epoch": 0.02, + "grad_norm": 0.4296875, + "learning_rate": 0.0001999833554136543, + "loss": 2.8045, + "step": 372 + }, + { + "epoch": 0.02, + "grad_norm": 0.396484375, + "learning_rate": 0.00019998323280544103, + "loss": 2.8746, + "step": 373 + }, + { + "epoch": 0.02, + "grad_norm": 0.375, + "learning_rate": 0.00019998310974733978, + "loss": 2.968, + "step": 374 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.00019998298623935106, + "loss": 2.8573, + "step": 375 + }, + { + "epoch": 0.02, + "grad_norm": 0.466796875, + "learning_rate": 0.00019998286228147544, + "loss": 3.0727, + "step": 376 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.0001999827378737135, + "loss": 3.0352, + "step": 377 + }, + { + "epoch": 0.02, + "grad_norm": 0.48046875, + "learning_rate": 0.00019998261301606575, + "loss": 2.9972, + "step": 378 + }, + { + "epoch": 0.02, + "grad_norm": 0.400390625, + "learning_rate": 0.0001999824877085328, + "loss": 2.9786, + "step": 379 + }, + { + "epoch": 0.02, + "grad_norm": 0.419921875, + "learning_rate": 0.00019998236195111518, + "loss": 2.8412, + "step": 380 + }, + { + "epoch": 0.02, + "grad_norm": 0.400390625, + "learning_rate": 0.00019998223574381346, + "loss": 2.8533, + "step": 381 + }, + { + "epoch": 0.02, + "grad_norm": 0.478515625, + "learning_rate": 0.00019998210908662824, + "loss": 2.815, + "step": 382 + }, + { + "epoch": 0.02, + "grad_norm": 0.36328125, + "learning_rate": 0.0001999819819795601, + "loss": 2.8949, + "step": 383 + }, + { + "epoch": 0.02, + "grad_norm": 0.40234375, + "learning_rate": 0.00019998185442260952, + "loss": 2.8564, + "step": 384 + }, + { + "epoch": 0.02, + "grad_norm": 0.416015625, + "learning_rate": 0.00019998172641577718, + "loss": 2.7571, + "step": 385 + }, + { + "epoch": 0.02, + "grad_norm": 0.3984375, + "learning_rate": 0.00019998159795906357, + "loss": 2.8882, + "step": 386 + }, + { + "epoch": 0.02, + "grad_norm": 0.3828125, + "learning_rate": 0.00019998146905246932, + "loss": 2.9497, + "step": 387 + }, + { + "epoch": 0.02, + "grad_norm": 0.375, + "learning_rate": 0.00019998133969599503, + "loss": 2.9013, + "step": 388 + }, + { + "epoch": 0.02, + "grad_norm": 0.36328125, + "learning_rate": 0.0001999812098896412, + "loss": 2.8398, + "step": 389 + }, + { + "epoch": 0.02, + "grad_norm": 0.33203125, + "learning_rate": 0.00019998107963340852, + "loss": 2.9038, + "step": 390 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.0001999809489272975, + "loss": 3.0663, + "step": 391 + }, + { + "epoch": 0.02, + "grad_norm": 0.3828125, + "learning_rate": 0.00019998081777130872, + "loss": 2.9746, + "step": 392 + }, + { + "epoch": 0.02, + "grad_norm": 0.38671875, + "learning_rate": 0.00019998068616544284, + "loss": 2.8285, + "step": 393 + }, + { + "epoch": 0.02, + "grad_norm": 0.34765625, + "learning_rate": 0.00019998055410970036, + "loss": 2.7727, + "step": 394 + }, + { + "epoch": 0.02, + "grad_norm": 0.470703125, + "learning_rate": 0.00019998042160408196, + "loss": 2.9663, + "step": 395 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.0001999802886485882, + "loss": 2.8824, + "step": 396 + }, + { + "epoch": 0.02, + "grad_norm": 0.443359375, + "learning_rate": 0.00019998015524321967, + "loss": 2.8265, + "step": 397 + }, + { + "epoch": 0.02, + "grad_norm": 0.3515625, + "learning_rate": 0.000199980021387977, + "loss": 2.8145, + "step": 398 + }, + { + "epoch": 0.02, + "grad_norm": 0.369140625, + "learning_rate": 0.00019997988708286079, + "loss": 2.8546, + "step": 399 + }, + { + "epoch": 0.02, + "grad_norm": 0.416015625, + "learning_rate": 0.00019997975232787155, + "loss": 2.9629, + "step": 400 + }, + { + "epoch": 0.02, + "grad_norm": 0.447265625, + "learning_rate": 0.00019997961712301003, + "loss": 2.8371, + "step": 401 + }, + { + "epoch": 0.02, + "grad_norm": 0.4140625, + "learning_rate": 0.00019997948146827674, + "loss": 2.8997, + "step": 402 + }, + { + "epoch": 0.02, + "grad_norm": 0.3984375, + "learning_rate": 0.00019997934536367234, + "loss": 2.8762, + "step": 403 + }, + { + "epoch": 0.02, + "grad_norm": 0.427734375, + "learning_rate": 0.0001999792088091974, + "loss": 2.8775, + "step": 404 + }, + { + "epoch": 0.02, + "grad_norm": 0.396484375, + "learning_rate": 0.00019997907180485257, + "loss": 2.9633, + "step": 405 + }, + { + "epoch": 0.02, + "grad_norm": 0.3359375, + "learning_rate": 0.00019997893435063846, + "loss": 2.7826, + "step": 406 + }, + { + "epoch": 0.02, + "grad_norm": 0.42578125, + "learning_rate": 0.0001999787964465557, + "loss": 2.9735, + "step": 407 + }, + { + "epoch": 0.02, + "grad_norm": 0.396484375, + "learning_rate": 0.00019997865809260487, + "loss": 2.8161, + "step": 408 + }, + { + "epoch": 0.02, + "grad_norm": 0.384765625, + "learning_rate": 0.00019997851928878663, + "loss": 2.8986, + "step": 409 + }, + { + "epoch": 0.02, + "grad_norm": 0.376953125, + "learning_rate": 0.00019997838003510158, + "loss": 2.7999, + "step": 410 + }, + { + "epoch": 0.02, + "grad_norm": 0.3984375, + "learning_rate": 0.00019997824033155036, + "loss": 2.9925, + "step": 411 + }, + { + "epoch": 0.02, + "grad_norm": 0.357421875, + "learning_rate": 0.0001999781001781336, + "loss": 2.8399, + "step": 412 + }, + { + "epoch": 0.02, + "grad_norm": 0.39453125, + "learning_rate": 0.00019997795957485192, + "loss": 2.876, + "step": 413 + }, + { + "epoch": 0.02, + "grad_norm": 0.34375, + "learning_rate": 0.00019997781852170595, + "loss": 2.7873, + "step": 414 + }, + { + "epoch": 0.02, + "grad_norm": 0.44140625, + "learning_rate": 0.0001999776770186964, + "loss": 2.8264, + "step": 415 + }, + { + "epoch": 0.02, + "grad_norm": 0.39453125, + "learning_rate": 0.00019997753506582376, + "loss": 2.9551, + "step": 416 + }, + { + "epoch": 0.02, + "grad_norm": 0.396484375, + "learning_rate": 0.00019997739266308878, + "loss": 2.9165, + "step": 417 + }, + { + "epoch": 0.02, + "grad_norm": 0.41015625, + "learning_rate": 0.00019997724981049206, + "loss": 2.7744, + "step": 418 + }, + { + "epoch": 0.02, + "grad_norm": 0.37890625, + "learning_rate": 0.00019997710650803427, + "loss": 2.8312, + "step": 419 + }, + { + "epoch": 0.02, + "grad_norm": 0.48828125, + "learning_rate": 0.00019997696275571604, + "loss": 2.8724, + "step": 420 + }, + { + "epoch": 0.02, + "grad_norm": 0.361328125, + "learning_rate": 0.000199976818553538, + "loss": 2.8697, + "step": 421 + }, + { + "epoch": 0.02, + "grad_norm": 0.45703125, + "learning_rate": 0.00019997667390150083, + "loss": 2.7684, + "step": 422 + }, + { + "epoch": 0.02, + "grad_norm": 0.37890625, + "learning_rate": 0.00019997652879960516, + "loss": 2.9745, + "step": 423 + }, + { + "epoch": 0.02, + "grad_norm": 0.51171875, + "learning_rate": 0.00019997638324785165, + "loss": 2.9649, + "step": 424 + }, + { + "epoch": 0.02, + "grad_norm": 0.462890625, + "learning_rate": 0.00019997623724624095, + "loss": 2.9377, + "step": 425 + }, + { + "epoch": 0.02, + "grad_norm": 0.435546875, + "learning_rate": 0.0001999760907947737, + "loss": 2.9186, + "step": 426 + }, + { + "epoch": 0.02, + "grad_norm": 0.515625, + "learning_rate": 0.0001999759438934506, + "loss": 2.8647, + "step": 427 + }, + { + "epoch": 0.02, + "grad_norm": 0.376953125, + "learning_rate": 0.0001999757965422723, + "loss": 2.8557, + "step": 428 + }, + { + "epoch": 0.02, + "grad_norm": 0.431640625, + "learning_rate": 0.0001999756487412394, + "loss": 2.8014, + "step": 429 + }, + { + "epoch": 0.02, + "grad_norm": 0.41015625, + "learning_rate": 0.00019997550049035266, + "loss": 2.9567, + "step": 430 + }, + { + "epoch": 0.02, + "grad_norm": 0.439453125, + "learning_rate": 0.0001999753517896127, + "loss": 2.8281, + "step": 431 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.0001999752026390202, + "loss": 2.7995, + "step": 432 + }, + { + "epoch": 0.02, + "grad_norm": 0.4609375, + "learning_rate": 0.00019997505303857583, + "loss": 2.8931, + "step": 433 + }, + { + "epoch": 0.02, + "grad_norm": 0.4453125, + "learning_rate": 0.00019997490298828024, + "loss": 2.8787, + "step": 434 + }, + { + "epoch": 0.02, + "grad_norm": 0.33984375, + "learning_rate": 0.00019997475248813412, + "loss": 2.9079, + "step": 435 + }, + { + "epoch": 0.02, + "grad_norm": 0.39453125, + "learning_rate": 0.00019997460153813816, + "loss": 2.8648, + "step": 436 + }, + { + "epoch": 0.02, + "grad_norm": 0.37109375, + "learning_rate": 0.00019997445013829305, + "loss": 2.9262, + "step": 437 + }, + { + "epoch": 0.02, + "grad_norm": 0.42578125, + "learning_rate": 0.00019997429828859942, + "loss": 2.9402, + "step": 438 + }, + { + "epoch": 0.02, + "grad_norm": 0.37109375, + "learning_rate": 0.00019997414598905798, + "loss": 2.7925, + "step": 439 + }, + { + "epoch": 0.02, + "grad_norm": 0.3515625, + "learning_rate": 0.00019997399323966946, + "loss": 2.8178, + "step": 440 + }, + { + "epoch": 0.02, + "grad_norm": 0.40625, + "learning_rate": 0.00019997384004043448, + "loss": 2.8276, + "step": 441 + }, + { + "epoch": 0.02, + "grad_norm": 0.439453125, + "learning_rate": 0.0001999736863913537, + "loss": 3.0277, + "step": 442 + }, + { + "epoch": 0.02, + "grad_norm": 0.462890625, + "learning_rate": 0.00019997353229242798, + "loss": 2.9568, + "step": 443 + }, + { + "epoch": 0.02, + "grad_norm": 0.4609375, + "learning_rate": 0.00019997337774365782, + "loss": 2.8098, + "step": 444 + }, + { + "epoch": 0.02, + "grad_norm": 0.375, + "learning_rate": 0.00019997322274504403, + "loss": 2.8627, + "step": 445 + }, + { + "epoch": 0.02, + "grad_norm": 0.408203125, + "learning_rate": 0.00019997306729658724, + "loss": 2.909, + "step": 446 + }, + { + "epoch": 0.02, + "grad_norm": 0.408203125, + "learning_rate": 0.00019997291139828822, + "loss": 2.8814, + "step": 447 + }, + { + "epoch": 0.02, + "grad_norm": 0.41015625, + "learning_rate": 0.00019997275505014762, + "loss": 2.7775, + "step": 448 + }, + { + "epoch": 0.02, + "grad_norm": 0.44921875, + "learning_rate": 0.00019997259825216618, + "loss": 2.8662, + "step": 449 + }, + { + "epoch": 0.02, + "grad_norm": 0.4453125, + "learning_rate": 0.00019997244100434456, + "loss": 2.8841, + "step": 450 + }, + { + "epoch": 0.02, + "grad_norm": 0.48046875, + "learning_rate": 0.00019997228330668348, + "loss": 2.9179, + "step": 451 + }, + { + "epoch": 0.02, + "grad_norm": 0.48828125, + "learning_rate": 0.0001999721251591837, + "loss": 2.8967, + "step": 452 + }, + { + "epoch": 0.02, + "grad_norm": 0.41796875, + "learning_rate": 0.00019997196656184585, + "loss": 2.8205, + "step": 453 + }, + { + "epoch": 0.02, + "grad_norm": 0.50390625, + "learning_rate": 0.0001999718075146707, + "loss": 2.9586, + "step": 454 + }, + { + "epoch": 0.02, + "grad_norm": 0.376953125, + "learning_rate": 0.00019997164801765894, + "loss": 2.8235, + "step": 455 + }, + { + "epoch": 0.02, + "grad_norm": 0.400390625, + "learning_rate": 0.00019997148807081132, + "loss": 2.7422, + "step": 456 + }, + { + "epoch": 0.02, + "grad_norm": 0.443359375, + "learning_rate": 0.00019997132767412854, + "loss": 2.7851, + "step": 457 + }, + { + "epoch": 0.02, + "grad_norm": 0.423828125, + "learning_rate": 0.0001999711668276113, + "loss": 2.8916, + "step": 458 + }, + { + "epoch": 0.02, + "grad_norm": 0.462890625, + "learning_rate": 0.00019997100553126039, + "loss": 2.8032, + "step": 459 + }, + { + "epoch": 0.02, + "grad_norm": 0.380859375, + "learning_rate": 0.00019997084378507648, + "loss": 2.8547, + "step": 460 + }, + { + "epoch": 0.02, + "grad_norm": 0.451171875, + "learning_rate": 0.00019997068158906025, + "loss": 2.8974, + "step": 461 + }, + { + "epoch": 0.02, + "grad_norm": 0.42578125, + "learning_rate": 0.00019997051894321254, + "loss": 2.8392, + "step": 462 + }, + { + "epoch": 0.02, + "grad_norm": 0.515625, + "learning_rate": 0.000199970355847534, + "loss": 3.0236, + "step": 463 + }, + { + "epoch": 0.02, + "grad_norm": 0.439453125, + "learning_rate": 0.00019997019230202543, + "loss": 2.863, + "step": 464 + }, + { + "epoch": 0.02, + "grad_norm": 0.443359375, + "learning_rate": 0.00019997002830668753, + "loss": 3.0395, + "step": 465 + }, + { + "epoch": 0.02, + "grad_norm": 0.4921875, + "learning_rate": 0.00019996986386152098, + "loss": 2.933, + "step": 466 + }, + { + "epoch": 0.02, + "grad_norm": 0.39453125, + "learning_rate": 0.00019996969896652662, + "loss": 2.8892, + "step": 467 + }, + { + "epoch": 0.02, + "grad_norm": 0.474609375, + "learning_rate": 0.00019996953362170516, + "loss": 2.898, + "step": 468 + }, + { + "epoch": 0.02, + "grad_norm": 0.427734375, + "learning_rate": 0.00019996936782705731, + "loss": 2.9514, + "step": 469 + }, + { + "epoch": 0.02, + "grad_norm": 0.48046875, + "learning_rate": 0.00019996920158258386, + "loss": 2.8762, + "step": 470 + }, + { + "epoch": 0.02, + "grad_norm": 0.439453125, + "learning_rate": 0.0001999690348882855, + "loss": 2.8783, + "step": 471 + }, + { + "epoch": 0.02, + "grad_norm": 0.421875, + "learning_rate": 0.00019996886774416303, + "loss": 2.8473, + "step": 472 + }, + { + "epoch": 0.02, + "grad_norm": 0.435546875, + "learning_rate": 0.0001999687001502172, + "loss": 2.959, + "step": 473 + }, + { + "epoch": 0.02, + "grad_norm": 0.470703125, + "learning_rate": 0.00019996853210644877, + "loss": 3.1498, + "step": 474 + }, + { + "epoch": 0.02, + "grad_norm": 0.37890625, + "learning_rate": 0.00019996836361285846, + "loss": 2.7756, + "step": 475 + }, + { + "epoch": 0.02, + "grad_norm": 0.466796875, + "learning_rate": 0.00019996819466944706, + "loss": 2.95, + "step": 476 + }, + { + "epoch": 0.02, + "grad_norm": 0.376953125, + "learning_rate": 0.00019996802527621533, + "loss": 2.9222, + "step": 477 + }, + { + "epoch": 0.02, + "grad_norm": 0.46875, + "learning_rate": 0.00019996785543316395, + "loss": 2.944, + "step": 478 + }, + { + "epoch": 0.02, + "grad_norm": 0.419921875, + "learning_rate": 0.00019996768514029384, + "loss": 2.8327, + "step": 479 + }, + { + "epoch": 0.02, + "grad_norm": 0.3984375, + "learning_rate": 0.00019996751439760562, + "loss": 2.8951, + "step": 480 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.00019996734320510017, + "loss": 2.8041, + "step": 481 + }, + { + "epoch": 0.02, + "grad_norm": 0.376953125, + "learning_rate": 0.00019996717156277818, + "loss": 2.8659, + "step": 482 + }, + { + "epoch": 0.02, + "grad_norm": 0.365234375, + "learning_rate": 0.00019996699947064046, + "loss": 2.9262, + "step": 483 + }, + { + "epoch": 0.02, + "grad_norm": 0.44140625, + "learning_rate": 0.00019996682692868779, + "loss": 2.9619, + "step": 484 + }, + { + "epoch": 0.02, + "grad_norm": 0.396484375, + "learning_rate": 0.00019996665393692088, + "loss": 2.7081, + "step": 485 + }, + { + "epoch": 0.02, + "grad_norm": 0.404296875, + "learning_rate": 0.0001999664804953406, + "loss": 2.9466, + "step": 486 + }, + { + "epoch": 0.02, + "grad_norm": 0.462890625, + "learning_rate": 0.0001999663066039477, + "loss": 2.9821, + "step": 487 + }, + { + "epoch": 0.02, + "grad_norm": 0.40234375, + "learning_rate": 0.00019996613226274296, + "loss": 2.8766, + "step": 488 + }, + { + "epoch": 0.02, + "grad_norm": 0.400390625, + "learning_rate": 0.00019996595747172712, + "loss": 2.9897, + "step": 489 + }, + { + "epoch": 0.02, + "grad_norm": 0.380859375, + "learning_rate": 0.00019996578223090104, + "loss": 2.8142, + "step": 490 + }, + { + "epoch": 0.02, + "grad_norm": 0.38671875, + "learning_rate": 0.00019996560654026544, + "loss": 2.7896, + "step": 491 + }, + { + "epoch": 0.02, + "grad_norm": 0.373046875, + "learning_rate": 0.00019996543039982117, + "loss": 2.7787, + "step": 492 + }, + { + "epoch": 0.02, + "grad_norm": 0.423828125, + "learning_rate": 0.00019996525380956898, + "loss": 2.8439, + "step": 493 + }, + { + "epoch": 0.02, + "grad_norm": 0.490234375, + "learning_rate": 0.0001999650767695097, + "loss": 2.8281, + "step": 494 + }, + { + "epoch": 0.02, + "grad_norm": 0.349609375, + "learning_rate": 0.0001999648992796441, + "loss": 2.8516, + "step": 495 + }, + { + "epoch": 0.02, + "grad_norm": 0.625, + "learning_rate": 0.00019996472133997298, + "loss": 2.7259, + "step": 496 + }, + { + "epoch": 0.02, + "grad_norm": 0.392578125, + "learning_rate": 0.00019996454295049713, + "loss": 2.8337, + "step": 497 + }, + { + "epoch": 0.02, + "grad_norm": 0.47265625, + "learning_rate": 0.0001999643641112174, + "loss": 2.9552, + "step": 498 + }, + { + "epoch": 0.02, + "grad_norm": 0.373046875, + "learning_rate": 0.00019996418482213458, + "loss": 2.8788, + "step": 499 + }, + { + "epoch": 0.02, + "grad_norm": 0.4765625, + "learning_rate": 0.00019996400508324943, + "loss": 2.8328, + "step": 500 + }, + { + "epoch": 0.02, + "grad_norm": 0.392578125, + "learning_rate": 0.00019996382489456278, + "loss": 2.8079, + "step": 501 + }, + { + "epoch": 0.02, + "grad_norm": 0.3828125, + "learning_rate": 0.00019996364425607546, + "loss": 2.7692, + "step": 502 + }, + { + "epoch": 0.02, + "grad_norm": 0.369140625, + "learning_rate": 0.0001999634631677883, + "loss": 2.9378, + "step": 503 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.00019996328162970203, + "loss": 2.8174, + "step": 504 + }, + { + "epoch": 0.02, + "grad_norm": 0.33984375, + "learning_rate": 0.00019996309964181758, + "loss": 2.7265, + "step": 505 + }, + { + "epoch": 0.02, + "grad_norm": 0.384765625, + "learning_rate": 0.0001999629172041357, + "loss": 2.7853, + "step": 506 + }, + { + "epoch": 0.02, + "grad_norm": 0.384765625, + "learning_rate": 0.0001999627343166572, + "loss": 2.7807, + "step": 507 + }, + { + "epoch": 0.02, + "grad_norm": 0.396484375, + "learning_rate": 0.00019996255097938298, + "loss": 2.8288, + "step": 508 + }, + { + "epoch": 0.02, + "grad_norm": 0.3671875, + "learning_rate": 0.00019996236719231375, + "loss": 2.7173, + "step": 509 + }, + { + "epoch": 0.02, + "grad_norm": 0.40625, + "learning_rate": 0.00019996218295545047, + "loss": 2.8945, + "step": 510 + }, + { + "epoch": 0.02, + "grad_norm": 0.357421875, + "learning_rate": 0.00019996199826879382, + "loss": 2.8969, + "step": 511 + }, + { + "epoch": 0.02, + "grad_norm": 0.353515625, + "learning_rate": 0.00019996181313234476, + "loss": 2.8435, + "step": 512 + }, + { + "epoch": 0.02, + "grad_norm": 0.361328125, + "learning_rate": 0.00019996162754610408, + "loss": 2.8734, + "step": 513 + }, + { + "epoch": 0.02, + "grad_norm": 0.37890625, + "learning_rate": 0.00019996144151007258, + "loss": 2.8554, + "step": 514 + }, + { + "epoch": 0.02, + "grad_norm": 0.43359375, + "learning_rate": 0.0001999612550242511, + "loss": 2.9599, + "step": 515 + }, + { + "epoch": 0.02, + "grad_norm": 0.34765625, + "learning_rate": 0.00019996106808864057, + "loss": 2.7091, + "step": 516 + }, + { + "epoch": 0.02, + "grad_norm": 0.443359375, + "learning_rate": 0.00019996088070324173, + "loss": 2.7764, + "step": 517 + }, + { + "epoch": 0.02, + "grad_norm": 0.375, + "learning_rate": 0.00019996069286805547, + "loss": 2.8091, + "step": 518 + }, + { + "epoch": 0.02, + "grad_norm": 0.404296875, + "learning_rate": 0.00019996050458308262, + "loss": 2.8333, + "step": 519 + }, + { + "epoch": 0.02, + "grad_norm": 0.353515625, + "learning_rate": 0.00019996031584832403, + "loss": 2.8445, + "step": 520 + }, + { + "epoch": 0.02, + "grad_norm": 0.41796875, + "learning_rate": 0.00019996012666378054, + "loss": 2.8067, + "step": 521 + }, + { + "epoch": 0.02, + "grad_norm": 0.369140625, + "learning_rate": 0.000199959937029453, + "loss": 2.7756, + "step": 522 + }, + { + "epoch": 0.02, + "grad_norm": 0.45703125, + "learning_rate": 0.0001999597469453423, + "loss": 2.7676, + "step": 523 + }, + { + "epoch": 0.02, + "grad_norm": 0.423828125, + "learning_rate": 0.00019995955641144926, + "loss": 2.8912, + "step": 524 + }, + { + "epoch": 0.02, + "grad_norm": 0.40234375, + "learning_rate": 0.00019995936542777475, + "loss": 2.7065, + "step": 525 + }, + { + "epoch": 0.02, + "grad_norm": 0.392578125, + "learning_rate": 0.00019995917399431962, + "loss": 2.8163, + "step": 526 + }, + { + "epoch": 0.02, + "grad_norm": 0.423828125, + "learning_rate": 0.00019995898211108476, + "loss": 2.7149, + "step": 527 + }, + { + "epoch": 0.02, + "grad_norm": 0.359375, + "learning_rate": 0.00019995878977807098, + "loss": 2.8849, + "step": 528 + }, + { + "epoch": 0.02, + "grad_norm": 0.431640625, + "learning_rate": 0.00019995859699527922, + "loss": 2.8433, + "step": 529 + }, + { + "epoch": 0.02, + "grad_norm": 0.408203125, + "learning_rate": 0.00019995840376271025, + "loss": 2.778, + "step": 530 + }, + { + "epoch": 0.02, + "grad_norm": 0.39453125, + "learning_rate": 0.00019995821008036502, + "loss": 2.8018, + "step": 531 + }, + { + "epoch": 0.02, + "grad_norm": 0.37890625, + "learning_rate": 0.00019995801594824438, + "loss": 2.9173, + "step": 532 + }, + { + "epoch": 0.02, + "grad_norm": 0.359375, + "learning_rate": 0.00019995782136634919, + "loss": 2.8076, + "step": 533 + }, + { + "epoch": 0.02, + "grad_norm": 0.345703125, + "learning_rate": 0.00019995762633468034, + "loss": 2.7743, + "step": 534 + }, + { + "epoch": 0.02, + "grad_norm": 0.36328125, + "learning_rate": 0.0001999574308532387, + "loss": 2.8404, + "step": 535 + }, + { + "epoch": 0.02, + "grad_norm": 0.365234375, + "learning_rate": 0.00019995723492202515, + "loss": 2.8477, + "step": 536 + }, + { + "epoch": 0.02, + "grad_norm": 0.404296875, + "learning_rate": 0.00019995703854104057, + "loss": 2.9853, + "step": 537 + }, + { + "epoch": 0.02, + "grad_norm": 0.416015625, + "learning_rate": 0.00019995684171028586, + "loss": 2.9137, + "step": 538 + }, + { + "epoch": 0.02, + "grad_norm": 0.3828125, + "learning_rate": 0.0001999566444297619, + "loss": 2.7442, + "step": 539 + }, + { + "epoch": 0.02, + "grad_norm": 0.392578125, + "learning_rate": 0.00019995644669946954, + "loss": 2.6945, + "step": 540 + }, + { + "epoch": 0.02, + "grad_norm": 0.384765625, + "learning_rate": 0.00019995624851940972, + "loss": 2.6755, + "step": 541 + }, + { + "epoch": 0.02, + "grad_norm": 0.423828125, + "learning_rate": 0.0001999560498895833, + "loss": 3.0838, + "step": 542 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.0001999558508099912, + "loss": 2.8937, + "step": 543 + }, + { + "epoch": 0.02, + "grad_norm": 0.322265625, + "learning_rate": 0.0001999556512806343, + "loss": 2.6839, + "step": 544 + }, + { + "epoch": 0.02, + "grad_norm": 0.39453125, + "learning_rate": 0.00019995545130151347, + "loss": 2.9284, + "step": 545 + }, + { + "epoch": 0.02, + "grad_norm": 0.380859375, + "learning_rate": 0.00019995525087262964, + "loss": 2.781, + "step": 546 + }, + { + "epoch": 0.02, + "grad_norm": 0.388671875, + "learning_rate": 0.00019995504999398373, + "loss": 2.7728, + "step": 547 + }, + { + "epoch": 0.02, + "grad_norm": 0.40625, + "learning_rate": 0.00019995484866557663, + "loss": 2.776, + "step": 548 + }, + { + "epoch": 0.02, + "grad_norm": 0.353515625, + "learning_rate": 0.00019995464688740927, + "loss": 2.8316, + "step": 549 + }, + { + "epoch": 0.02, + "grad_norm": 0.390625, + "learning_rate": 0.00019995444465948246, + "loss": 2.9048, + "step": 550 + }, + { + "epoch": 0.02, + "grad_norm": 0.404296875, + "learning_rate": 0.00019995424198179722, + "loss": 2.8826, + "step": 551 + }, + { + "epoch": 0.02, + "grad_norm": 0.419921875, + "learning_rate": 0.00019995403885435443, + "loss": 2.9167, + "step": 552 + }, + { + "epoch": 0.02, + "grad_norm": 0.341796875, + "learning_rate": 0.00019995383527715498, + "loss": 2.7507, + "step": 553 + }, + { + "epoch": 0.02, + "grad_norm": 0.4140625, + "learning_rate": 0.00019995363125019976, + "loss": 2.9202, + "step": 554 + }, + { + "epoch": 0.02, + "grad_norm": 0.359375, + "learning_rate": 0.00019995342677348976, + "loss": 2.8244, + "step": 555 + }, + { + "epoch": 0.02, + "grad_norm": 0.35546875, + "learning_rate": 0.00019995322184702589, + "loss": 2.8051, + "step": 556 + }, + { + "epoch": 0.02, + "grad_norm": 0.4296875, + "learning_rate": 0.000199953016470809, + "loss": 2.959, + "step": 557 + }, + { + "epoch": 0.02, + "grad_norm": 0.39453125, + "learning_rate": 0.0001999528106448401, + "loss": 2.9421, + "step": 558 + }, + { + "epoch": 0.02, + "grad_norm": 0.40625, + "learning_rate": 0.00019995260436912008, + "loss": 2.7995, + "step": 559 + }, + { + "epoch": 0.02, + "grad_norm": 0.37109375, + "learning_rate": 0.00019995239764364987, + "loss": 2.8559, + "step": 560 + }, + { + "epoch": 0.02, + "grad_norm": 0.400390625, + "learning_rate": 0.0001999521904684304, + "loss": 2.7795, + "step": 561 + }, + { + "epoch": 0.02, + "grad_norm": 0.345703125, + "learning_rate": 0.00019995198284346262, + "loss": 2.7468, + "step": 562 + }, + { + "epoch": 0.02, + "grad_norm": 0.373046875, + "learning_rate": 0.0001999517747687474, + "loss": 2.6835, + "step": 563 + }, + { + "epoch": 0.02, + "grad_norm": 0.373046875, + "learning_rate": 0.00019995156624428575, + "loss": 2.6975, + "step": 564 + }, + { + "epoch": 0.02, + "grad_norm": 0.384765625, + "learning_rate": 0.00019995135727007856, + "loss": 2.7875, + "step": 565 + }, + { + "epoch": 0.02, + "grad_norm": 0.3984375, + "learning_rate": 0.0001999511478461268, + "loss": 2.9563, + "step": 566 + }, + { + "epoch": 0.02, + "grad_norm": 0.376953125, + "learning_rate": 0.0001999509379724314, + "loss": 2.624, + "step": 567 + }, + { + "epoch": 0.02, + "grad_norm": 0.375, + "learning_rate": 0.00019995072764899331, + "loss": 2.8615, + "step": 568 + }, + { + "epoch": 0.02, + "grad_norm": 0.3984375, + "learning_rate": 0.0001999505168758135, + "loss": 2.8732, + "step": 569 + }, + { + "epoch": 0.02, + "grad_norm": 0.408203125, + "learning_rate": 0.00019995030565289284, + "loss": 2.9231, + "step": 570 + }, + { + "epoch": 0.02, + "grad_norm": 0.419921875, + "learning_rate": 0.00019995009398023238, + "loss": 2.8584, + "step": 571 + }, + { + "epoch": 0.02, + "grad_norm": 0.4140625, + "learning_rate": 0.000199949881857833, + "loss": 2.8224, + "step": 572 + }, + { + "epoch": 0.02, + "grad_norm": 0.3828125, + "learning_rate": 0.0001999496692856957, + "loss": 2.7789, + "step": 573 + }, + { + "epoch": 0.02, + "grad_norm": 0.39453125, + "learning_rate": 0.0001999494562638214, + "loss": 2.8285, + "step": 574 + }, + { + "epoch": 0.02, + "grad_norm": 0.494140625, + "learning_rate": 0.00019994924279221108, + "loss": 2.965, + "step": 575 + }, + { + "epoch": 0.02, + "grad_norm": 0.4375, + "learning_rate": 0.00019994902887086568, + "loss": 2.7817, + "step": 576 + }, + { + "epoch": 0.02, + "grad_norm": 0.5234375, + "learning_rate": 0.00019994881449978617, + "loss": 2.8276, + "step": 577 + }, + { + "epoch": 0.02, + "grad_norm": 0.5, + "learning_rate": 0.00019994859967897356, + "loss": 2.7642, + "step": 578 + }, + { + "epoch": 0.02, + "grad_norm": 0.361328125, + "learning_rate": 0.00019994838440842876, + "loss": 2.8004, + "step": 579 + }, + { + "epoch": 0.02, + "grad_norm": 0.51953125, + "learning_rate": 0.00019994816868815275, + "loss": 2.9257, + "step": 580 + }, + { + "epoch": 0.02, + "grad_norm": 0.345703125, + "learning_rate": 0.0001999479525181465, + "loss": 2.848, + "step": 581 + }, + { + "epoch": 0.02, + "grad_norm": 0.494140625, + "learning_rate": 0.00019994773589841102, + "loss": 2.9703, + "step": 582 + }, + { + "epoch": 0.02, + "grad_norm": 0.369140625, + "learning_rate": 0.00019994751882894721, + "loss": 2.9679, + "step": 583 + }, + { + "epoch": 0.02, + "grad_norm": 0.50390625, + "learning_rate": 0.00019994730130975613, + "loss": 2.7688, + "step": 584 + }, + { + "epoch": 0.02, + "grad_norm": 0.455078125, + "learning_rate": 0.00019994708334083872, + "loss": 2.811, + "step": 585 + }, + { + "epoch": 0.02, + "grad_norm": 0.498046875, + "learning_rate": 0.00019994686492219593, + "loss": 2.7985, + "step": 586 + }, + { + "epoch": 0.03, + "grad_norm": 0.37109375, + "learning_rate": 0.0001999466460538288, + "loss": 2.7941, + "step": 587 + }, + { + "epoch": 0.03, + "grad_norm": 0.451171875, + "learning_rate": 0.00019994642673573833, + "loss": 2.9673, + "step": 588 + }, + { + "epoch": 0.03, + "grad_norm": 0.44921875, + "learning_rate": 0.00019994620696792538, + "loss": 3.0286, + "step": 589 + }, + { + "epoch": 0.03, + "grad_norm": 0.45703125, + "learning_rate": 0.00019994598675039107, + "loss": 2.8919, + "step": 590 + }, + { + "epoch": 0.03, + "grad_norm": 0.345703125, + "learning_rate": 0.00019994576608313636, + "loss": 2.69, + "step": 591 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.00019994554496616221, + "loss": 2.8918, + "step": 592 + }, + { + "epoch": 0.03, + "grad_norm": 0.458984375, + "learning_rate": 0.0001999453233994696, + "loss": 2.9248, + "step": 593 + }, + { + "epoch": 0.03, + "grad_norm": 0.451171875, + "learning_rate": 0.0001999451013830596, + "loss": 2.8646, + "step": 594 + }, + { + "epoch": 0.03, + "grad_norm": 0.439453125, + "learning_rate": 0.00019994487891693318, + "loss": 2.7815, + "step": 595 + }, + { + "epoch": 0.03, + "grad_norm": 0.482421875, + "learning_rate": 0.0001999446560010913, + "loss": 2.9057, + "step": 596 + }, + { + "epoch": 0.03, + "grad_norm": 0.42578125, + "learning_rate": 0.000199944432635535, + "loss": 2.9106, + "step": 597 + }, + { + "epoch": 0.03, + "grad_norm": 0.515625, + "learning_rate": 0.0001999442088202653, + "loss": 2.9303, + "step": 598 + }, + { + "epoch": 0.03, + "grad_norm": 0.408203125, + "learning_rate": 0.00019994398455528315, + "loss": 2.8191, + "step": 599 + }, + { + "epoch": 0.03, + "grad_norm": 0.439453125, + "learning_rate": 0.00019994375984058965, + "loss": 2.8671, + "step": 600 + }, + { + "epoch": 0.03, + "grad_norm": 0.35546875, + "learning_rate": 0.0001999435346761857, + "loss": 2.7149, + "step": 601 + }, + { + "epoch": 0.03, + "grad_norm": 0.455078125, + "learning_rate": 0.0001999433090620724, + "loss": 2.7532, + "step": 602 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.00019994308299825072, + "loss": 2.8107, + "step": 603 + }, + { + "epoch": 0.03, + "grad_norm": 0.4375, + "learning_rate": 0.0001999428564847217, + "loss": 2.8546, + "step": 604 + }, + { + "epoch": 0.03, + "grad_norm": 0.44140625, + "learning_rate": 0.00019994262952148634, + "loss": 2.9689, + "step": 605 + }, + { + "epoch": 0.03, + "grad_norm": 0.41015625, + "learning_rate": 0.0001999424021085457, + "loss": 2.8501, + "step": 606 + }, + { + "epoch": 0.03, + "grad_norm": 0.416015625, + "learning_rate": 0.00019994217424590077, + "loss": 2.7232, + "step": 607 + }, + { + "epoch": 0.03, + "grad_norm": 0.423828125, + "learning_rate": 0.00019994194593355255, + "loss": 2.9007, + "step": 608 + }, + { + "epoch": 0.03, + "grad_norm": 0.396484375, + "learning_rate": 0.00019994171717150212, + "loss": 2.8821, + "step": 609 + }, + { + "epoch": 0.03, + "grad_norm": 0.39453125, + "learning_rate": 0.00019994148795975046, + "loss": 2.8118, + "step": 610 + }, + { + "epoch": 0.03, + "grad_norm": 0.361328125, + "learning_rate": 0.00019994125829829866, + "loss": 2.9016, + "step": 611 + }, + { + "epoch": 0.03, + "grad_norm": 0.390625, + "learning_rate": 0.00019994102818714771, + "loss": 2.8237, + "step": 612 + }, + { + "epoch": 0.03, + "grad_norm": 0.37109375, + "learning_rate": 0.00019994079762629868, + "loss": 2.8315, + "step": 613 + }, + { + "epoch": 0.03, + "grad_norm": 0.3671875, + "learning_rate": 0.00019994056661575257, + "loss": 2.7998, + "step": 614 + }, + { + "epoch": 0.03, + "grad_norm": 0.361328125, + "learning_rate": 0.0001999403351555104, + "loss": 2.7634, + "step": 615 + }, + { + "epoch": 0.03, + "grad_norm": 0.34375, + "learning_rate": 0.0001999401032455733, + "loss": 2.7547, + "step": 616 + }, + { + "epoch": 0.03, + "grad_norm": 0.41015625, + "learning_rate": 0.00019993987088594222, + "loss": 2.7742, + "step": 617 + }, + { + "epoch": 0.03, + "grad_norm": 0.345703125, + "learning_rate": 0.00019993963807661826, + "loss": 2.7405, + "step": 618 + }, + { + "epoch": 0.03, + "grad_norm": 0.42578125, + "learning_rate": 0.00019993940481760245, + "loss": 2.8408, + "step": 619 + }, + { + "epoch": 0.03, + "grad_norm": 0.3828125, + "learning_rate": 0.00019993917110889585, + "loss": 2.8193, + "step": 620 + }, + { + "epoch": 0.03, + "grad_norm": 0.423828125, + "learning_rate": 0.0001999389369504995, + "loss": 2.8511, + "step": 621 + }, + { + "epoch": 0.03, + "grad_norm": 0.41015625, + "learning_rate": 0.00019993870234241446, + "loss": 2.6963, + "step": 622 + }, + { + "epoch": 0.03, + "grad_norm": 0.47265625, + "learning_rate": 0.00019993846728464174, + "loss": 2.8813, + "step": 623 + }, + { + "epoch": 0.03, + "grad_norm": 0.41796875, + "learning_rate": 0.0001999382317771825, + "loss": 2.8489, + "step": 624 + }, + { + "epoch": 0.03, + "grad_norm": 0.486328125, + "learning_rate": 0.00019993799582003772, + "loss": 2.7785, + "step": 625 + }, + { + "epoch": 0.03, + "grad_norm": 0.36328125, + "learning_rate": 0.00019993775941320847, + "loss": 2.8223, + "step": 626 + }, + { + "epoch": 0.03, + "grad_norm": 0.53125, + "learning_rate": 0.00019993752255669582, + "loss": 2.8157, + "step": 627 + }, + { + "epoch": 0.03, + "grad_norm": 0.3828125, + "learning_rate": 0.00019993728525050087, + "loss": 2.7656, + "step": 628 + }, + { + "epoch": 0.03, + "grad_norm": 0.4765625, + "learning_rate": 0.00019993704749462465, + "loss": 2.8305, + "step": 629 + }, + { + "epoch": 0.03, + "grad_norm": 0.380859375, + "learning_rate": 0.00019993680928906825, + "loss": 2.9123, + "step": 630 + }, + { + "epoch": 0.03, + "grad_norm": 0.515625, + "learning_rate": 0.0001999365706338327, + "loss": 2.8842, + "step": 631 + }, + { + "epoch": 0.03, + "grad_norm": 0.4453125, + "learning_rate": 0.0001999363315289191, + "loss": 2.8535, + "step": 632 + }, + { + "epoch": 0.03, + "grad_norm": 0.53515625, + "learning_rate": 0.00019993609197432858, + "loss": 2.7919, + "step": 633 + }, + { + "epoch": 0.03, + "grad_norm": 0.41796875, + "learning_rate": 0.00019993585197006211, + "loss": 2.8927, + "step": 634 + }, + { + "epoch": 0.03, + "grad_norm": 0.466796875, + "learning_rate": 0.00019993561151612088, + "loss": 2.8391, + "step": 635 + }, + { + "epoch": 0.03, + "grad_norm": 0.435546875, + "learning_rate": 0.00019993537061250592, + "loss": 2.7669, + "step": 636 + }, + { + "epoch": 0.03, + "grad_norm": 0.44140625, + "learning_rate": 0.00019993512925921828, + "loss": 2.7891, + "step": 637 + }, + { + "epoch": 0.03, + "grad_norm": 0.52734375, + "learning_rate": 0.0001999348874562591, + "loss": 2.8889, + "step": 638 + }, + { + "epoch": 0.03, + "grad_norm": 0.384765625, + "learning_rate": 0.00019993464520362943, + "loss": 2.8971, + "step": 639 + }, + { + "epoch": 0.03, + "grad_norm": 0.408203125, + "learning_rate": 0.0001999344025013304, + "loss": 2.877, + "step": 640 + }, + { + "epoch": 0.03, + "grad_norm": 0.380859375, + "learning_rate": 0.00019993415934936305, + "loss": 2.667, + "step": 641 + }, + { + "epoch": 0.03, + "grad_norm": 0.40234375, + "learning_rate": 0.00019993391574772855, + "loss": 2.7317, + "step": 642 + }, + { + "epoch": 0.03, + "grad_norm": 0.3984375, + "learning_rate": 0.00019993367169642793, + "loss": 2.7958, + "step": 643 + }, + { + "epoch": 0.03, + "grad_norm": 0.408203125, + "learning_rate": 0.00019993342719546228, + "loss": 2.9668, + "step": 644 + }, + { + "epoch": 0.03, + "grad_norm": 0.400390625, + "learning_rate": 0.00019993318224483276, + "loss": 2.6714, + "step": 645 + }, + { + "epoch": 0.03, + "grad_norm": 0.373046875, + "learning_rate": 0.00019993293684454044, + "loss": 2.8943, + "step": 646 + }, + { + "epoch": 0.03, + "grad_norm": 0.400390625, + "learning_rate": 0.00019993269099458642, + "loss": 2.8355, + "step": 647 + }, + { + "epoch": 0.03, + "grad_norm": 0.443359375, + "learning_rate": 0.00019993244469497181, + "loss": 2.8691, + "step": 648 + }, + { + "epoch": 0.03, + "grad_norm": 0.3671875, + "learning_rate": 0.00019993219794569773, + "loss": 2.7961, + "step": 649 + }, + { + "epoch": 0.03, + "grad_norm": 0.42578125, + "learning_rate": 0.00019993195074676529, + "loss": 2.7163, + "step": 650 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.00019993170309817555, + "loss": 2.6957, + "step": 651 + }, + { + "epoch": 0.03, + "grad_norm": 0.42578125, + "learning_rate": 0.0001999314549999297, + "loss": 2.791, + "step": 652 + }, + { + "epoch": 0.03, + "grad_norm": 0.4609375, + "learning_rate": 0.00019993120645202882, + "loss": 2.8182, + "step": 653 + }, + { + "epoch": 0.03, + "grad_norm": 0.41015625, + "learning_rate": 0.000199930957454474, + "loss": 2.819, + "step": 654 + }, + { + "epoch": 0.03, + "grad_norm": 0.392578125, + "learning_rate": 0.00019993070800726647, + "loss": 2.8096, + "step": 655 + }, + { + "epoch": 0.03, + "grad_norm": 0.416015625, + "learning_rate": 0.0001999304581104072, + "loss": 2.7438, + "step": 656 + }, + { + "epoch": 0.03, + "grad_norm": 0.42578125, + "learning_rate": 0.00019993020776389741, + "loss": 2.8845, + "step": 657 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.0001999299569677382, + "loss": 2.8444, + "step": 658 + }, + { + "epoch": 0.03, + "grad_norm": 0.34375, + "learning_rate": 0.00019992970572193074, + "loss": 2.7266, + "step": 659 + }, + { + "epoch": 0.03, + "grad_norm": 0.380859375, + "learning_rate": 0.0001999294540264761, + "loss": 2.8351, + "step": 660 + }, + { + "epoch": 0.03, + "grad_norm": 0.35546875, + "learning_rate": 0.00019992920188137542, + "loss": 2.7916, + "step": 661 + }, + { + "epoch": 0.03, + "grad_norm": 0.4296875, + "learning_rate": 0.00019992894928662987, + "loss": 2.8688, + "step": 662 + }, + { + "epoch": 0.03, + "grad_norm": 0.380859375, + "learning_rate": 0.00019992869624224055, + "loss": 2.6982, + "step": 663 + }, + { + "epoch": 0.03, + "grad_norm": 0.46484375, + "learning_rate": 0.00019992844274820863, + "loss": 2.7133, + "step": 664 + }, + { + "epoch": 0.03, + "grad_norm": 0.3671875, + "learning_rate": 0.00019992818880453523, + "loss": 2.6784, + "step": 665 + }, + { + "epoch": 0.03, + "grad_norm": 0.40625, + "learning_rate": 0.00019992793441122152, + "loss": 2.8097, + "step": 666 + }, + { + "epoch": 0.03, + "grad_norm": 0.400390625, + "learning_rate": 0.0001999276795682686, + "loss": 2.8072, + "step": 667 + }, + { + "epoch": 0.03, + "grad_norm": 0.439453125, + "learning_rate": 0.00019992742427567769, + "loss": 2.8068, + "step": 668 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.00019992716853344984, + "loss": 2.8223, + "step": 669 + }, + { + "epoch": 0.03, + "grad_norm": 0.400390625, + "learning_rate": 0.00019992691234158624, + "loss": 2.8741, + "step": 670 + }, + { + "epoch": 0.03, + "grad_norm": 0.408203125, + "learning_rate": 0.0001999266557000881, + "loss": 2.9065, + "step": 671 + }, + { + "epoch": 0.03, + "grad_norm": 0.427734375, + "learning_rate": 0.0001999263986089565, + "loss": 2.8528, + "step": 672 + }, + { + "epoch": 0.03, + "grad_norm": 0.408203125, + "learning_rate": 0.00019992614106819262, + "loss": 2.843, + "step": 673 + }, + { + "epoch": 0.03, + "grad_norm": 0.423828125, + "learning_rate": 0.00019992588307779765, + "loss": 2.9235, + "step": 674 + }, + { + "epoch": 0.03, + "grad_norm": 0.412109375, + "learning_rate": 0.0001999256246377727, + "loss": 2.8917, + "step": 675 + }, + { + "epoch": 0.03, + "grad_norm": 0.34765625, + "learning_rate": 0.000199925365748119, + "loss": 2.7896, + "step": 676 + }, + { + "epoch": 0.03, + "grad_norm": 0.375, + "learning_rate": 0.00019992510640883763, + "loss": 2.7192, + "step": 677 + }, + { + "epoch": 0.03, + "grad_norm": 0.36328125, + "learning_rate": 0.0001999248466199298, + "loss": 2.8067, + "step": 678 + }, + { + "epoch": 0.03, + "grad_norm": 0.41796875, + "learning_rate": 0.00019992458638139668, + "loss": 2.8809, + "step": 679 + }, + { + "epoch": 0.03, + "grad_norm": 0.41796875, + "learning_rate": 0.00019992432569323945, + "loss": 2.8239, + "step": 680 + }, + { + "epoch": 0.03, + "grad_norm": 0.462890625, + "learning_rate": 0.0001999240645554593, + "loss": 2.837, + "step": 681 + }, + { + "epoch": 0.03, + "grad_norm": 0.37890625, + "learning_rate": 0.00019992380296805732, + "loss": 2.7722, + "step": 682 + }, + { + "epoch": 0.03, + "grad_norm": 0.40234375, + "learning_rate": 0.0001999235409310348, + "loss": 2.7932, + "step": 683 + }, + { + "epoch": 0.03, + "grad_norm": 0.392578125, + "learning_rate": 0.00019992327844439283, + "loss": 2.7244, + "step": 684 + }, + { + "epoch": 0.03, + "grad_norm": 0.40234375, + "learning_rate": 0.0001999230155081326, + "loss": 2.8719, + "step": 685 + }, + { + "epoch": 0.03, + "grad_norm": 0.447265625, + "learning_rate": 0.00019992275212225537, + "loss": 2.8205, + "step": 686 + }, + { + "epoch": 0.03, + "grad_norm": 0.365234375, + "learning_rate": 0.00019992248828676226, + "loss": 2.8865, + "step": 687 + }, + { + "epoch": 0.03, + "grad_norm": 0.462890625, + "learning_rate": 0.00019992222400165447, + "loss": 2.7703, + "step": 688 + }, + { + "epoch": 0.03, + "grad_norm": 0.341796875, + "learning_rate": 0.00019992195926693315, + "loss": 2.797, + "step": 689 + }, + { + "epoch": 0.03, + "grad_norm": 0.37109375, + "learning_rate": 0.00019992169408259957, + "loss": 2.8574, + "step": 690 + }, + { + "epoch": 0.03, + "grad_norm": 0.41015625, + "learning_rate": 0.0001999214284486549, + "loss": 2.861, + "step": 691 + }, + { + "epoch": 0.03, + "grad_norm": 0.404296875, + "learning_rate": 0.00019992116236510027, + "loss": 2.9068, + "step": 692 + }, + { + "epoch": 0.03, + "grad_norm": 0.43359375, + "learning_rate": 0.00019992089583193695, + "loss": 2.8729, + "step": 693 + }, + { + "epoch": 0.03, + "grad_norm": 0.400390625, + "learning_rate": 0.00019992062884916612, + "loss": 2.8084, + "step": 694 + }, + { + "epoch": 0.03, + "grad_norm": 0.4375, + "learning_rate": 0.00019992036141678902, + "loss": 2.7717, + "step": 695 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.00019992009353480671, + "loss": 2.9615, + "step": 696 + }, + { + "epoch": 0.03, + "grad_norm": 0.46875, + "learning_rate": 0.00019991982520322057, + "loss": 2.9062, + "step": 697 + }, + { + "epoch": 0.03, + "grad_norm": 0.365234375, + "learning_rate": 0.0001999195564220317, + "loss": 2.8099, + "step": 698 + }, + { + "epoch": 0.03, + "grad_norm": 0.474609375, + "learning_rate": 0.0001999192871912414, + "loss": 2.8646, + "step": 699 + }, + { + "epoch": 0.03, + "grad_norm": 0.40625, + "learning_rate": 0.00019991901751085077, + "loss": 2.8207, + "step": 700 + }, + { + "epoch": 0.03, + "grad_norm": 0.41015625, + "learning_rate": 0.00019991874738086108, + "loss": 2.7678, + "step": 701 + }, + { + "epoch": 0.03, + "grad_norm": 0.412109375, + "learning_rate": 0.00019991847680127353, + "loss": 2.7642, + "step": 702 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.00019991820577208938, + "loss": 2.7838, + "step": 703 + }, + { + "epoch": 0.03, + "grad_norm": 0.4375, + "learning_rate": 0.00019991793429330982, + "loss": 2.7921, + "step": 704 + }, + { + "epoch": 0.03, + "grad_norm": 0.380859375, + "learning_rate": 0.00019991766236493604, + "loss": 2.8979, + "step": 705 + }, + { + "epoch": 0.03, + "grad_norm": 0.408203125, + "learning_rate": 0.00019991738998696934, + "loss": 2.7012, + "step": 706 + }, + { + "epoch": 0.03, + "grad_norm": 0.416015625, + "learning_rate": 0.0001999171171594109, + "loss": 2.8537, + "step": 707 + }, + { + "epoch": 0.03, + "grad_norm": 0.421875, + "learning_rate": 0.0001999168438822619, + "loss": 2.817, + "step": 708 + }, + { + "epoch": 0.03, + "grad_norm": 0.330078125, + "learning_rate": 0.00019991657015552367, + "loss": 2.6327, + "step": 709 + }, + { + "epoch": 0.03, + "grad_norm": 0.419921875, + "learning_rate": 0.00019991629597919736, + "loss": 2.6423, + "step": 710 + }, + { + "epoch": 0.03, + "grad_norm": 0.375, + "learning_rate": 0.00019991602135328423, + "loss": 2.7801, + "step": 711 + }, + { + "epoch": 0.03, + "grad_norm": 0.447265625, + "learning_rate": 0.0001999157462777855, + "loss": 2.8075, + "step": 712 + }, + { + "epoch": 0.03, + "grad_norm": 0.380859375, + "learning_rate": 0.00019991547075270247, + "loss": 2.9507, + "step": 713 + }, + { + "epoch": 0.03, + "grad_norm": 0.37109375, + "learning_rate": 0.0001999151947780363, + "loss": 2.7742, + "step": 714 + }, + { + "epoch": 0.03, + "grad_norm": 0.3515625, + "learning_rate": 0.00019991491835378832, + "loss": 2.7673, + "step": 715 + }, + { + "epoch": 0.03, + "grad_norm": 0.431640625, + "learning_rate": 0.00019991464147995966, + "loss": 2.9137, + "step": 716 + }, + { + "epoch": 0.03, + "grad_norm": 0.34765625, + "learning_rate": 0.00019991436415655169, + "loss": 2.8148, + "step": 717 + }, + { + "epoch": 0.03, + "grad_norm": 0.36328125, + "learning_rate": 0.00019991408638356557, + "loss": 2.7225, + "step": 718 + }, + { + "epoch": 0.03, + "grad_norm": 0.408203125, + "learning_rate": 0.00019991380816100255, + "loss": 2.7583, + "step": 719 + }, + { + "epoch": 0.03, + "grad_norm": 0.359375, + "learning_rate": 0.00019991352948886394, + "loss": 2.8173, + "step": 720 + }, + { + "epoch": 0.03, + "grad_norm": 0.40625, + "learning_rate": 0.00019991325036715094, + "loss": 2.8057, + "step": 721 + }, + { + "epoch": 0.03, + "grad_norm": 0.3671875, + "learning_rate": 0.00019991297079586485, + "loss": 2.717, + "step": 722 + }, + { + "epoch": 0.03, + "grad_norm": 0.3828125, + "learning_rate": 0.0001999126907750069, + "loss": 2.7582, + "step": 723 + }, + { + "epoch": 0.03, + "grad_norm": 0.412109375, + "learning_rate": 0.00019991241030457832, + "loss": 2.8119, + "step": 724 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.00019991212938458045, + "loss": 2.8168, + "step": 725 + }, + { + "epoch": 0.03, + "grad_norm": 0.376953125, + "learning_rate": 0.00019991184801501448, + "loss": 2.8594, + "step": 726 + }, + { + "epoch": 0.03, + "grad_norm": 0.3671875, + "learning_rate": 0.00019991156619588173, + "loss": 2.8035, + "step": 727 + }, + { + "epoch": 0.03, + "grad_norm": 0.443359375, + "learning_rate": 0.00019991128392718346, + "loss": 2.7997, + "step": 728 + }, + { + "epoch": 0.03, + "grad_norm": 0.376953125, + "learning_rate": 0.00019991100120892087, + "loss": 2.7691, + "step": 729 + }, + { + "epoch": 0.03, + "grad_norm": 0.40234375, + "learning_rate": 0.00019991071804109533, + "loss": 2.871, + "step": 730 + }, + { + "epoch": 0.03, + "grad_norm": 0.3984375, + "learning_rate": 0.00019991043442370808, + "loss": 2.7818, + "step": 731 + }, + { + "epoch": 0.03, + "grad_norm": 0.388671875, + "learning_rate": 0.00019991015035676038, + "loss": 2.7875, + "step": 732 + }, + { + "epoch": 0.03, + "grad_norm": 0.337890625, + "learning_rate": 0.00019990986584025348, + "loss": 2.8322, + "step": 733 + }, + { + "epoch": 0.03, + "grad_norm": 0.390625, + "learning_rate": 0.00019990958087418874, + "loss": 2.6878, + "step": 734 + }, + { + "epoch": 0.03, + "grad_norm": 0.388671875, + "learning_rate": 0.00019990929545856737, + "loss": 2.8884, + "step": 735 + }, + { + "epoch": 0.03, + "grad_norm": 0.4375, + "learning_rate": 0.0001999090095933907, + "loss": 2.7926, + "step": 736 + }, + { + "epoch": 0.03, + "grad_norm": 0.3671875, + "learning_rate": 0.00019990872327866002, + "loss": 2.8227, + "step": 737 + }, + { + "epoch": 0.03, + "grad_norm": 0.40234375, + "learning_rate": 0.0001999084365143766, + "loss": 2.8617, + "step": 738 + }, + { + "epoch": 0.03, + "grad_norm": 0.390625, + "learning_rate": 0.00019990814930054168, + "loss": 2.7872, + "step": 739 + }, + { + "epoch": 0.03, + "grad_norm": 0.357421875, + "learning_rate": 0.00019990786163715663, + "loss": 2.8044, + "step": 740 + }, + { + "epoch": 0.03, + "grad_norm": 0.373046875, + "learning_rate": 0.0001999075735242227, + "loss": 2.8684, + "step": 741 + }, + { + "epoch": 0.03, + "grad_norm": 0.380859375, + "learning_rate": 0.0001999072849617412, + "loss": 2.7262, + "step": 742 + }, + { + "epoch": 0.03, + "grad_norm": 0.396484375, + "learning_rate": 0.00019990699594971345, + "loss": 2.7887, + "step": 743 + }, + { + "epoch": 0.03, + "grad_norm": 0.396484375, + "learning_rate": 0.00019990670648814071, + "loss": 2.7464, + "step": 744 + }, + { + "epoch": 0.03, + "grad_norm": 0.43359375, + "learning_rate": 0.0001999064165770243, + "loss": 2.7101, + "step": 745 + }, + { + "epoch": 0.03, + "grad_norm": 0.423828125, + "learning_rate": 0.00019990612621636556, + "loss": 2.6845, + "step": 746 + }, + { + "epoch": 0.03, + "grad_norm": 0.380859375, + "learning_rate": 0.00019990583540616574, + "loss": 2.6884, + "step": 747 + }, + { + "epoch": 0.03, + "grad_norm": 0.388671875, + "learning_rate": 0.00019990554414642616, + "loss": 2.6795, + "step": 748 + }, + { + "epoch": 0.03, + "grad_norm": 0.3671875, + "learning_rate": 0.00019990525243714818, + "loss": 2.6654, + "step": 749 + }, + { + "epoch": 0.03, + "grad_norm": 0.361328125, + "learning_rate": 0.00019990496027833302, + "loss": 2.8463, + "step": 750 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.00019990466766998212, + "loss": 2.8341, + "step": 751 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.00019990437461209668, + "loss": 2.7191, + "step": 752 + }, + { + "epoch": 0.03, + "grad_norm": 0.466796875, + "learning_rate": 0.00019990408110467806, + "loss": 2.9061, + "step": 753 + }, + { + "epoch": 0.03, + "grad_norm": 0.4375, + "learning_rate": 0.00019990378714772761, + "loss": 2.8019, + "step": 754 + }, + { + "epoch": 0.03, + "grad_norm": 0.39453125, + "learning_rate": 0.0001999034927412466, + "loss": 2.7557, + "step": 755 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.0001999031978852364, + "loss": 2.7709, + "step": 756 + }, + { + "epoch": 0.03, + "grad_norm": 0.412109375, + "learning_rate": 0.00019990290257969836, + "loss": 2.8648, + "step": 757 + }, + { + "epoch": 0.03, + "grad_norm": 0.45703125, + "learning_rate": 0.0001999026068246337, + "loss": 2.6924, + "step": 758 + }, + { + "epoch": 0.03, + "grad_norm": 0.3984375, + "learning_rate": 0.00019990231062004386, + "loss": 2.7706, + "step": 759 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.00019990201396593012, + "loss": 2.9446, + "step": 760 + }, + { + "epoch": 0.03, + "grad_norm": 0.40625, + "learning_rate": 0.00019990171686229383, + "loss": 3.0295, + "step": 761 + }, + { + "epoch": 0.03, + "grad_norm": 0.45703125, + "learning_rate": 0.00019990141930913632, + "loss": 2.9046, + "step": 762 + }, + { + "epoch": 0.03, + "grad_norm": 0.431640625, + "learning_rate": 0.00019990112130645896, + "loss": 3.0577, + "step": 763 + }, + { + "epoch": 0.03, + "grad_norm": 0.419921875, + "learning_rate": 0.000199900822854263, + "loss": 2.7211, + "step": 764 + }, + { + "epoch": 0.03, + "grad_norm": 0.392578125, + "learning_rate": 0.0001999005239525499, + "loss": 2.7983, + "step": 765 + }, + { + "epoch": 0.03, + "grad_norm": 0.353515625, + "learning_rate": 0.00019990022460132095, + "loss": 2.7501, + "step": 766 + }, + { + "epoch": 0.03, + "grad_norm": 0.333984375, + "learning_rate": 0.00019989992480057747, + "loss": 2.8149, + "step": 767 + }, + { + "epoch": 0.03, + "grad_norm": 0.37109375, + "learning_rate": 0.00019989962455032085, + "loss": 2.7917, + "step": 768 + }, + { + "epoch": 0.03, + "grad_norm": 0.400390625, + "learning_rate": 0.00019989932385055243, + "loss": 2.7995, + "step": 769 + }, + { + "epoch": 0.03, + "grad_norm": 0.427734375, + "learning_rate": 0.00019989902270127358, + "loss": 2.7623, + "step": 770 + }, + { + "epoch": 0.03, + "grad_norm": 0.4375, + "learning_rate": 0.00019989872110248561, + "loss": 2.9141, + "step": 771 + }, + { + "epoch": 0.03, + "grad_norm": 0.412109375, + "learning_rate": 0.00019989841905418992, + "loss": 2.6409, + "step": 772 + }, + { + "epoch": 0.03, + "grad_norm": 0.37109375, + "learning_rate": 0.00019989811655638786, + "loss": 2.7973, + "step": 773 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.00019989781360908078, + "loss": 2.7895, + "step": 774 + }, + { + "epoch": 0.03, + "grad_norm": 0.4296875, + "learning_rate": 0.00019989751021227004, + "loss": 2.8014, + "step": 775 + }, + { + "epoch": 0.03, + "grad_norm": 0.392578125, + "learning_rate": 0.000199897206365957, + "loss": 2.7797, + "step": 776 + }, + { + "epoch": 0.03, + "grad_norm": 0.416015625, + "learning_rate": 0.00019989690207014306, + "loss": 2.8712, + "step": 777 + }, + { + "epoch": 0.03, + "grad_norm": 0.361328125, + "learning_rate": 0.00019989659732482954, + "loss": 2.7419, + "step": 778 + }, + { + "epoch": 0.03, + "grad_norm": 0.427734375, + "learning_rate": 0.00019989629213001785, + "loss": 2.7999, + "step": 779 + }, + { + "epoch": 0.03, + "grad_norm": 0.375, + "learning_rate": 0.00019989598648570937, + "loss": 2.975, + "step": 780 + }, + { + "epoch": 0.03, + "grad_norm": 0.376953125, + "learning_rate": 0.00019989568039190548, + "loss": 2.8901, + "step": 781 + }, + { + "epoch": 0.03, + "grad_norm": 0.431640625, + "learning_rate": 0.0001998953738486075, + "loss": 2.9208, + "step": 782 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.00019989506685581685, + "loss": 2.7873, + "step": 783 + }, + { + "epoch": 0.03, + "grad_norm": 0.419921875, + "learning_rate": 0.00019989475941353493, + "loss": 2.7873, + "step": 784 + }, + { + "epoch": 0.03, + "grad_norm": 0.4296875, + "learning_rate": 0.00019989445152176306, + "loss": 2.8324, + "step": 785 + }, + { + "epoch": 0.03, + "grad_norm": 0.412109375, + "learning_rate": 0.00019989414318050269, + "loss": 2.7445, + "step": 786 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.00019989383438975518, + "loss": 2.713, + "step": 787 + }, + { + "epoch": 0.03, + "grad_norm": 0.3828125, + "learning_rate": 0.00019989352514952194, + "loss": 2.668, + "step": 788 + }, + { + "epoch": 0.03, + "grad_norm": 0.396484375, + "learning_rate": 0.0001998932154598043, + "loss": 2.7078, + "step": 789 + }, + { + "epoch": 0.03, + "grad_norm": 0.43359375, + "learning_rate": 0.00019989290532060374, + "loss": 2.7804, + "step": 790 + }, + { + "epoch": 0.03, + "grad_norm": 0.388671875, + "learning_rate": 0.00019989259473192158, + "loss": 2.8228, + "step": 791 + }, + { + "epoch": 0.03, + "grad_norm": 0.431640625, + "learning_rate": 0.00019989228369375926, + "loss": 2.9684, + "step": 792 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.00019989197220611818, + "loss": 2.672, + "step": 793 + }, + { + "epoch": 0.03, + "grad_norm": 0.41796875, + "learning_rate": 0.0001998916602689997, + "loss": 2.8269, + "step": 794 + }, + { + "epoch": 0.03, + "grad_norm": 0.3984375, + "learning_rate": 0.0001998913478824053, + "loss": 2.9078, + "step": 795 + }, + { + "epoch": 0.03, + "grad_norm": 0.3984375, + "learning_rate": 0.00019989103504633633, + "loss": 2.8206, + "step": 796 + }, + { + "epoch": 0.03, + "grad_norm": 0.3828125, + "learning_rate": 0.0001998907217607942, + "loss": 2.7561, + "step": 797 + }, + { + "epoch": 0.03, + "grad_norm": 0.373046875, + "learning_rate": 0.00019989040802578033, + "loss": 2.7574, + "step": 798 + }, + { + "epoch": 0.03, + "grad_norm": 0.408203125, + "learning_rate": 0.0001998900938412961, + "loss": 2.6602, + "step": 799 + }, + { + "epoch": 0.03, + "grad_norm": 0.412109375, + "learning_rate": 0.000199889779207343, + "loss": 2.797, + "step": 800 + }, + { + "epoch": 0.03, + "grad_norm": 0.392578125, + "learning_rate": 0.0001998894641239224, + "loss": 2.8302, + "step": 801 + }, + { + "epoch": 0.03, + "grad_norm": 0.38671875, + "learning_rate": 0.00019988914859103569, + "loss": 2.7278, + "step": 802 + }, + { + "epoch": 0.03, + "grad_norm": 0.34765625, + "learning_rate": 0.00019988883260868432, + "loss": 2.6998, + "step": 803 + }, + { + "epoch": 0.03, + "grad_norm": 0.3984375, + "learning_rate": 0.00019988851617686973, + "loss": 2.8717, + "step": 804 + }, + { + "epoch": 0.03, + "grad_norm": 0.4140625, + "learning_rate": 0.00019988819929559327, + "loss": 2.8851, + "step": 805 + }, + { + "epoch": 0.03, + "grad_norm": 0.40625, + "learning_rate": 0.00019988788196485647, + "loss": 2.8661, + "step": 806 + }, + { + "epoch": 0.03, + "grad_norm": 0.4296875, + "learning_rate": 0.00019988756418466068, + "loss": 2.8024, + "step": 807 + }, + { + "epoch": 0.03, + "grad_norm": 0.37109375, + "learning_rate": 0.00019988724595500736, + "loss": 2.8162, + "step": 808 + }, + { + "epoch": 0.03, + "grad_norm": 0.41796875, + "learning_rate": 0.00019988692727589794, + "loss": 2.8202, + "step": 809 + }, + { + "epoch": 0.03, + "grad_norm": 0.423828125, + "learning_rate": 0.00019988660814733387, + "loss": 2.7832, + "step": 810 + }, + { + "epoch": 0.03, + "grad_norm": 0.421875, + "learning_rate": 0.00019988628856931656, + "loss": 2.8333, + "step": 811 + }, + { + "epoch": 0.03, + "grad_norm": 0.376953125, + "learning_rate": 0.00019988596854184743, + "loss": 2.7087, + "step": 812 + }, + { + "epoch": 0.03, + "grad_norm": 0.41796875, + "learning_rate": 0.00019988564806492796, + "loss": 2.8774, + "step": 813 + }, + { + "epoch": 0.03, + "grad_norm": 0.341796875, + "learning_rate": 0.00019988532713855958, + "loss": 2.7935, + "step": 814 + }, + { + "epoch": 0.03, + "grad_norm": 0.37109375, + "learning_rate": 0.00019988500576274373, + "loss": 2.8027, + "step": 815 + }, + { + "epoch": 0.03, + "grad_norm": 0.390625, + "learning_rate": 0.00019988468393748187, + "loss": 2.7285, + "step": 816 + }, + { + "epoch": 0.03, + "grad_norm": 0.39453125, + "learning_rate": 0.0001998843616627754, + "loss": 2.8807, + "step": 817 + }, + { + "epoch": 0.03, + "grad_norm": 0.373046875, + "learning_rate": 0.00019988403893862587, + "loss": 2.6898, + "step": 818 + }, + { + "epoch": 0.03, + "grad_norm": 0.40625, + "learning_rate": 0.00019988371576503463, + "loss": 2.7741, + "step": 819 + }, + { + "epoch": 0.03, + "grad_norm": 0.37890625, + "learning_rate": 0.00019988339214200318, + "loss": 2.8941, + "step": 820 + }, + { + "epoch": 0.03, + "grad_norm": 0.388671875, + "learning_rate": 0.00019988306806953293, + "loss": 2.7323, + "step": 821 + }, + { + "epoch": 0.04, + "grad_norm": 0.359375, + "learning_rate": 0.00019988274354762543, + "loss": 2.8075, + "step": 822 + }, + { + "epoch": 0.04, + "grad_norm": 0.376953125, + "learning_rate": 0.00019988241857628208, + "loss": 2.7923, + "step": 823 + }, + { + "epoch": 0.04, + "grad_norm": 0.390625, + "learning_rate": 0.00019988209315550433, + "loss": 2.7867, + "step": 824 + }, + { + "epoch": 0.04, + "grad_norm": 0.37890625, + "learning_rate": 0.00019988176728529363, + "loss": 2.7578, + "step": 825 + }, + { + "epoch": 0.04, + "grad_norm": 0.39453125, + "learning_rate": 0.00019988144096565153, + "loss": 2.8842, + "step": 826 + }, + { + "epoch": 0.04, + "grad_norm": 0.388671875, + "learning_rate": 0.00019988111419657944, + "loss": 2.8619, + "step": 827 + }, + { + "epoch": 0.04, + "grad_norm": 0.3671875, + "learning_rate": 0.00019988078697807883, + "loss": 2.8537, + "step": 828 + }, + { + "epoch": 0.04, + "grad_norm": 0.384765625, + "learning_rate": 0.0001998804593101512, + "loss": 2.7522, + "step": 829 + }, + { + "epoch": 0.04, + "grad_norm": 0.431640625, + "learning_rate": 0.000199880131192798, + "loss": 2.7277, + "step": 830 + }, + { + "epoch": 0.04, + "grad_norm": 0.373046875, + "learning_rate": 0.0001998798026260207, + "loss": 2.6437, + "step": 831 + }, + { + "epoch": 0.04, + "grad_norm": 0.466796875, + "learning_rate": 0.0001998794736098208, + "loss": 2.7139, + "step": 832 + }, + { + "epoch": 0.04, + "grad_norm": 0.361328125, + "learning_rate": 0.00019987914414419975, + "loss": 2.7994, + "step": 833 + }, + { + "epoch": 0.04, + "grad_norm": 0.48828125, + "learning_rate": 0.00019987881422915907, + "loss": 2.7696, + "step": 834 + }, + { + "epoch": 0.04, + "grad_norm": 0.427734375, + "learning_rate": 0.00019987848386470022, + "loss": 2.7781, + "step": 835 + }, + { + "epoch": 0.04, + "grad_norm": 0.458984375, + "learning_rate": 0.00019987815305082469, + "loss": 2.9178, + "step": 836 + }, + { + "epoch": 0.04, + "grad_norm": 0.4296875, + "learning_rate": 0.000199877821787534, + "loss": 2.862, + "step": 837 + }, + { + "epoch": 0.04, + "grad_norm": 0.40625, + "learning_rate": 0.00019987749007482956, + "loss": 2.9911, + "step": 838 + }, + { + "epoch": 0.04, + "grad_norm": 0.5234375, + "learning_rate": 0.00019987715791271296, + "loss": 2.7864, + "step": 839 + }, + { + "epoch": 0.04, + "grad_norm": 0.58203125, + "learning_rate": 0.00019987682530118566, + "loss": 2.7319, + "step": 840 + }, + { + "epoch": 0.04, + "grad_norm": 0.46484375, + "learning_rate": 0.0001998764922402491, + "loss": 2.7141, + "step": 841 + }, + { + "epoch": 0.04, + "grad_norm": 0.3828125, + "learning_rate": 0.00019987615872990485, + "loss": 2.7154, + "step": 842 + }, + { + "epoch": 0.04, + "grad_norm": 0.408203125, + "learning_rate": 0.00019987582477015437, + "loss": 2.7321, + "step": 843 + }, + { + "epoch": 0.04, + "grad_norm": 0.419921875, + "learning_rate": 0.00019987549036099922, + "loss": 2.6971, + "step": 844 + }, + { + "epoch": 0.04, + "grad_norm": 0.416015625, + "learning_rate": 0.00019987515550244086, + "loss": 2.8206, + "step": 845 + }, + { + "epoch": 0.04, + "grad_norm": 0.4140625, + "learning_rate": 0.00019987482019448075, + "loss": 2.7104, + "step": 846 + }, + { + "epoch": 0.04, + "grad_norm": 0.359375, + "learning_rate": 0.0001998744844371205, + "loss": 2.7685, + "step": 847 + }, + { + "epoch": 0.04, + "grad_norm": 0.45703125, + "learning_rate": 0.00019987414823036156, + "loss": 2.7973, + "step": 848 + }, + { + "epoch": 0.04, + "grad_norm": 0.427734375, + "learning_rate": 0.00019987381157420545, + "loss": 2.7813, + "step": 849 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.00019987347446865367, + "loss": 2.7008, + "step": 850 + }, + { + "epoch": 0.04, + "grad_norm": 0.4453125, + "learning_rate": 0.00019987313691370776, + "loss": 2.9312, + "step": 851 + }, + { + "epoch": 0.04, + "grad_norm": 0.392578125, + "learning_rate": 0.00019987279890936924, + "loss": 2.6993, + "step": 852 + }, + { + "epoch": 0.04, + "grad_norm": 0.42578125, + "learning_rate": 0.00019987246045563963, + "loss": 2.7986, + "step": 853 + }, + { + "epoch": 0.04, + "grad_norm": 0.38671875, + "learning_rate": 0.0001998721215525204, + "loss": 2.6991, + "step": 854 + }, + { + "epoch": 0.04, + "grad_norm": 0.451171875, + "learning_rate": 0.0001998717822000132, + "loss": 2.8508, + "step": 855 + }, + { + "epoch": 0.04, + "grad_norm": 0.359375, + "learning_rate": 0.00019987144239811942, + "loss": 2.5706, + "step": 856 + }, + { + "epoch": 0.04, + "grad_norm": 0.52734375, + "learning_rate": 0.0001998711021468407, + "loss": 2.9212, + "step": 857 + }, + { + "epoch": 0.04, + "grad_norm": 0.39453125, + "learning_rate": 0.00019987076144617846, + "loss": 2.8297, + "step": 858 + }, + { + "epoch": 0.04, + "grad_norm": 0.48046875, + "learning_rate": 0.00019987042029613434, + "loss": 2.9606, + "step": 859 + }, + { + "epoch": 0.04, + "grad_norm": 0.390625, + "learning_rate": 0.0001998700786967098, + "loss": 2.8298, + "step": 860 + }, + { + "epoch": 0.04, + "grad_norm": 0.4609375, + "learning_rate": 0.0001998697366479064, + "loss": 2.766, + "step": 861 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.0001998693941497257, + "loss": 2.8831, + "step": 862 + }, + { + "epoch": 0.04, + "grad_norm": 0.390625, + "learning_rate": 0.00019986905120216922, + "loss": 2.7624, + "step": 863 + }, + { + "epoch": 0.04, + "grad_norm": 0.3671875, + "learning_rate": 0.0001998687078052385, + "loss": 2.756, + "step": 864 + }, + { + "epoch": 0.04, + "grad_norm": 0.400390625, + "learning_rate": 0.0001998683639589351, + "loss": 2.7175, + "step": 865 + }, + { + "epoch": 0.04, + "grad_norm": 0.466796875, + "learning_rate": 0.00019986801966326057, + "loss": 2.9611, + "step": 866 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.00019986767491821645, + "loss": 2.7248, + "step": 867 + }, + { + "epoch": 0.04, + "grad_norm": 0.4375, + "learning_rate": 0.00019986732972380428, + "loss": 2.7931, + "step": 868 + }, + { + "epoch": 0.04, + "grad_norm": 0.35546875, + "learning_rate": 0.00019986698408002563, + "loss": 2.7196, + "step": 869 + }, + { + "epoch": 0.04, + "grad_norm": 0.39453125, + "learning_rate": 0.000199866637986882, + "loss": 2.8162, + "step": 870 + }, + { + "epoch": 0.04, + "grad_norm": 0.412109375, + "learning_rate": 0.00019986629144437505, + "loss": 2.7875, + "step": 871 + }, + { + "epoch": 0.04, + "grad_norm": 0.388671875, + "learning_rate": 0.00019986594445250627, + "loss": 2.7188, + "step": 872 + }, + { + "epoch": 0.04, + "grad_norm": 0.359375, + "learning_rate": 0.00019986559701127725, + "loss": 2.7961, + "step": 873 + }, + { + "epoch": 0.04, + "grad_norm": 0.41796875, + "learning_rate": 0.00019986524912068954, + "loss": 2.6908, + "step": 874 + }, + { + "epoch": 0.04, + "grad_norm": 0.369140625, + "learning_rate": 0.00019986490078074468, + "loss": 2.6783, + "step": 875 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.0001998645519914443, + "loss": 2.8143, + "step": 876 + }, + { + "epoch": 0.04, + "grad_norm": 0.423828125, + "learning_rate": 0.0001998642027527899, + "loss": 2.8287, + "step": 877 + }, + { + "epoch": 0.04, + "grad_norm": 0.38671875, + "learning_rate": 0.00019986385306478309, + "loss": 2.8289, + "step": 878 + }, + { + "epoch": 0.04, + "grad_norm": 0.359375, + "learning_rate": 0.00019986350292742545, + "loss": 2.7163, + "step": 879 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.00019986315234071854, + "loss": 2.8149, + "step": 880 + }, + { + "epoch": 0.04, + "grad_norm": 0.3984375, + "learning_rate": 0.00019986280130466393, + "loss": 2.7327, + "step": 881 + }, + { + "epoch": 0.04, + "grad_norm": 0.404296875, + "learning_rate": 0.0001998624498192632, + "loss": 2.8391, + "step": 882 + }, + { + "epoch": 0.04, + "grad_norm": 0.359375, + "learning_rate": 0.00019986209788451796, + "loss": 2.614, + "step": 883 + }, + { + "epoch": 0.04, + "grad_norm": 0.404296875, + "learning_rate": 0.00019986174550042977, + "loss": 2.9515, + "step": 884 + }, + { + "epoch": 0.04, + "grad_norm": 0.419921875, + "learning_rate": 0.00019986139266700022, + "loss": 2.7625, + "step": 885 + }, + { + "epoch": 0.04, + "grad_norm": 0.380859375, + "learning_rate": 0.00019986103938423092, + "loss": 2.7499, + "step": 886 + }, + { + "epoch": 0.04, + "grad_norm": 0.375, + "learning_rate": 0.0001998606856521234, + "loss": 2.9021, + "step": 887 + }, + { + "epoch": 0.04, + "grad_norm": 0.37890625, + "learning_rate": 0.0001998603314706793, + "loss": 2.642, + "step": 888 + }, + { + "epoch": 0.04, + "grad_norm": 0.404296875, + "learning_rate": 0.00019985997683990024, + "loss": 2.8076, + "step": 889 + }, + { + "epoch": 0.04, + "grad_norm": 0.3984375, + "learning_rate": 0.0001998596217597877, + "loss": 2.9055, + "step": 890 + }, + { + "epoch": 0.04, + "grad_norm": 0.388671875, + "learning_rate": 0.00019985926623034345, + "loss": 2.7496, + "step": 891 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.00019985891025156894, + "loss": 2.8047, + "step": 892 + }, + { + "epoch": 0.04, + "grad_norm": 0.37890625, + "learning_rate": 0.00019985855382346584, + "loss": 2.7576, + "step": 893 + }, + { + "epoch": 0.04, + "grad_norm": 0.375, + "learning_rate": 0.00019985819694603573, + "loss": 2.9777, + "step": 894 + }, + { + "epoch": 0.04, + "grad_norm": 0.333984375, + "learning_rate": 0.00019985783961928025, + "loss": 2.8472, + "step": 895 + }, + { + "epoch": 0.04, + "grad_norm": 0.451171875, + "learning_rate": 0.00019985748184320098, + "loss": 2.8683, + "step": 896 + }, + { + "epoch": 0.04, + "grad_norm": 0.4140625, + "learning_rate": 0.00019985712361779952, + "loss": 3.0099, + "step": 897 + }, + { + "epoch": 0.04, + "grad_norm": 0.390625, + "learning_rate": 0.00019985676494307751, + "loss": 2.7721, + "step": 898 + }, + { + "epoch": 0.04, + "grad_norm": 0.390625, + "learning_rate": 0.00019985640581903654, + "loss": 2.8165, + "step": 899 + }, + { + "epoch": 0.04, + "grad_norm": 0.458984375, + "learning_rate": 0.00019985604624567826, + "loss": 2.7707, + "step": 900 + }, + { + "epoch": 0.04, + "grad_norm": 0.421875, + "learning_rate": 0.00019985568622300426, + "loss": 2.6981, + "step": 901 + }, + { + "epoch": 0.04, + "grad_norm": 0.40234375, + "learning_rate": 0.00019985532575101614, + "loss": 2.7431, + "step": 902 + }, + { + "epoch": 0.04, + "grad_norm": 0.40625, + "learning_rate": 0.00019985496482971557, + "loss": 2.837, + "step": 903 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.00019985460345910413, + "loss": 2.7816, + "step": 904 + }, + { + "epoch": 0.04, + "grad_norm": 0.435546875, + "learning_rate": 0.0001998542416391835, + "loss": 2.7444, + "step": 905 + }, + { + "epoch": 0.04, + "grad_norm": 0.39453125, + "learning_rate": 0.00019985387936995524, + "loss": 2.7379, + "step": 906 + }, + { + "epoch": 0.04, + "grad_norm": 0.423828125, + "learning_rate": 0.00019985351665142105, + "loss": 2.8296, + "step": 907 + }, + { + "epoch": 0.04, + "grad_norm": 0.39453125, + "learning_rate": 0.0001998531534835825, + "loss": 2.7522, + "step": 908 + }, + { + "epoch": 0.04, + "grad_norm": 0.380859375, + "learning_rate": 0.00019985278986644128, + "loss": 2.7179, + "step": 909 + }, + { + "epoch": 0.04, + "grad_norm": 0.361328125, + "learning_rate": 0.00019985242579999898, + "loss": 2.8193, + "step": 910 + }, + { + "epoch": 0.04, + "grad_norm": 0.375, + "learning_rate": 0.00019985206128425725, + "loss": 2.6891, + "step": 911 + }, + { + "epoch": 0.04, + "grad_norm": 0.357421875, + "learning_rate": 0.00019985169631921772, + "loss": 2.8679, + "step": 912 + }, + { + "epoch": 0.04, + "grad_norm": 0.4375, + "learning_rate": 0.00019985133090488208, + "loss": 2.7349, + "step": 913 + }, + { + "epoch": 0.04, + "grad_norm": 0.33203125, + "learning_rate": 0.00019985096504125194, + "loss": 2.6359, + "step": 914 + }, + { + "epoch": 0.04, + "grad_norm": 0.3671875, + "learning_rate": 0.0001998505987283289, + "loss": 2.9283, + "step": 915 + }, + { + "epoch": 0.04, + "grad_norm": 0.373046875, + "learning_rate": 0.00019985023196611473, + "loss": 2.8162, + "step": 916 + }, + { + "epoch": 0.04, + "grad_norm": 0.408203125, + "learning_rate": 0.00019984986475461098, + "loss": 2.7631, + "step": 917 + }, + { + "epoch": 0.04, + "grad_norm": 0.4609375, + "learning_rate": 0.0001998494970938193, + "loss": 2.7019, + "step": 918 + }, + { + "epoch": 0.04, + "grad_norm": 0.376953125, + "learning_rate": 0.00019984912898374143, + "loss": 2.8368, + "step": 919 + }, + { + "epoch": 0.04, + "grad_norm": 0.435546875, + "learning_rate": 0.00019984876042437893, + "loss": 2.8929, + "step": 920 + }, + { + "epoch": 0.04, + "grad_norm": 0.337890625, + "learning_rate": 0.0001998483914157335, + "loss": 2.7709, + "step": 921 + }, + { + "epoch": 0.04, + "grad_norm": 0.474609375, + "learning_rate": 0.0001998480219578068, + "loss": 2.7299, + "step": 922 + }, + { + "epoch": 0.04, + "grad_norm": 0.35546875, + "learning_rate": 0.0001998476520506005, + "loss": 2.7485, + "step": 923 + }, + { + "epoch": 0.04, + "grad_norm": 0.46875, + "learning_rate": 0.00019984728169411623, + "loss": 2.7325, + "step": 924 + }, + { + "epoch": 0.04, + "grad_norm": 0.392578125, + "learning_rate": 0.00019984691088835573, + "loss": 2.7543, + "step": 925 + }, + { + "epoch": 0.04, + "grad_norm": 0.455078125, + "learning_rate": 0.00019984653963332058, + "loss": 2.8602, + "step": 926 + }, + { + "epoch": 0.04, + "grad_norm": 0.439453125, + "learning_rate": 0.00019984616792901252, + "loss": 2.9026, + "step": 927 + }, + { + "epoch": 0.04, + "grad_norm": 0.41015625, + "learning_rate": 0.0001998457957754332, + "loss": 2.7523, + "step": 928 + }, + { + "epoch": 0.04, + "grad_norm": 0.4375, + "learning_rate": 0.00019984542317258425, + "loss": 2.6775, + "step": 929 + }, + { + "epoch": 0.04, + "grad_norm": 0.42578125, + "learning_rate": 0.0001998450501204674, + "loss": 2.8657, + "step": 930 + }, + { + "epoch": 0.04, + "grad_norm": 0.41796875, + "learning_rate": 0.00019984467661908432, + "loss": 2.7326, + "step": 931 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.00019984430266843667, + "loss": 2.8123, + "step": 932 + }, + { + "epoch": 0.04, + "grad_norm": 0.427734375, + "learning_rate": 0.00019984392826852618, + "loss": 2.8626, + "step": 933 + }, + { + "epoch": 0.04, + "grad_norm": 0.41015625, + "learning_rate": 0.00019984355341935448, + "loss": 2.7842, + "step": 934 + }, + { + "epoch": 0.04, + "grad_norm": 0.384765625, + "learning_rate": 0.0001998431781209233, + "loss": 2.8643, + "step": 935 + }, + { + "epoch": 0.04, + "grad_norm": 0.42578125, + "learning_rate": 0.00019984280237323428, + "loss": 2.7311, + "step": 936 + }, + { + "epoch": 0.04, + "grad_norm": 0.337890625, + "learning_rate": 0.00019984242617628915, + "loss": 2.7318, + "step": 937 + }, + { + "epoch": 0.04, + "grad_norm": 0.427734375, + "learning_rate": 0.00019984204953008957, + "loss": 2.7964, + "step": 938 + }, + { + "epoch": 0.04, + "grad_norm": 0.39453125, + "learning_rate": 0.00019984167243463724, + "loss": 2.8075, + "step": 939 + }, + { + "epoch": 0.04, + "grad_norm": 0.44921875, + "learning_rate": 0.00019984129488993392, + "loss": 2.7758, + "step": 940 + }, + { + "epoch": 0.04, + "grad_norm": 0.423828125, + "learning_rate": 0.00019984091689598124, + "loss": 2.8788, + "step": 941 + }, + { + "epoch": 0.04, + "grad_norm": 0.38671875, + "learning_rate": 0.00019984053845278094, + "loss": 2.7883, + "step": 942 + }, + { + "epoch": 0.04, + "grad_norm": 0.353515625, + "learning_rate": 0.00019984015956033468, + "loss": 2.7742, + "step": 943 + }, + { + "epoch": 0.04, + "grad_norm": 0.376953125, + "learning_rate": 0.0001998397802186442, + "loss": 2.8316, + "step": 944 + }, + { + "epoch": 0.04, + "grad_norm": 0.33984375, + "learning_rate": 0.0001998394004277112, + "loss": 2.7438, + "step": 945 + }, + { + "epoch": 0.04, + "grad_norm": 0.41015625, + "learning_rate": 0.00019983902018753737, + "loss": 2.7952, + "step": 946 + }, + { + "epoch": 0.04, + "grad_norm": 0.40625, + "learning_rate": 0.00019983863949812444, + "loss": 2.6653, + "step": 947 + }, + { + "epoch": 0.04, + "grad_norm": 0.328125, + "learning_rate": 0.00019983825835947415, + "loss": 2.7755, + "step": 948 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.00019983787677158813, + "loss": 2.8652, + "step": 949 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.0001998374947344682, + "loss": 2.817, + "step": 950 + }, + { + "epoch": 0.04, + "grad_norm": 0.400390625, + "learning_rate": 0.000199837112248116, + "loss": 2.9617, + "step": 951 + }, + { + "epoch": 0.04, + "grad_norm": 0.353515625, + "learning_rate": 0.0001998367293125333, + "loss": 2.8305, + "step": 952 + }, + { + "epoch": 0.04, + "grad_norm": 0.365234375, + "learning_rate": 0.0001998363459277218, + "loss": 2.834, + "step": 953 + }, + { + "epoch": 0.04, + "grad_norm": 0.39453125, + "learning_rate": 0.00019983596209368322, + "loss": 2.8046, + "step": 954 + }, + { + "epoch": 0.04, + "grad_norm": 0.412109375, + "learning_rate": 0.0001998355778104193, + "loss": 2.7221, + "step": 955 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.00019983519307793178, + "loss": 2.807, + "step": 956 + }, + { + "epoch": 0.04, + "grad_norm": 0.380859375, + "learning_rate": 0.00019983480789622236, + "loss": 2.7552, + "step": 957 + }, + { + "epoch": 0.04, + "grad_norm": 0.41015625, + "learning_rate": 0.0001998344222652928, + "loss": 2.7117, + "step": 958 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.0001998340361851448, + "loss": 2.7054, + "step": 959 + }, + { + "epoch": 0.04, + "grad_norm": 0.369140625, + "learning_rate": 0.00019983364965578014, + "loss": 2.6753, + "step": 960 + }, + { + "epoch": 0.04, + "grad_norm": 0.30859375, + "learning_rate": 0.00019983326267720056, + "loss": 2.6258, + "step": 961 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.00019983287524940775, + "loss": 2.877, + "step": 962 + }, + { + "epoch": 0.04, + "grad_norm": 0.40625, + "learning_rate": 0.0001998324873724035, + "loss": 2.6792, + "step": 963 + }, + { + "epoch": 0.04, + "grad_norm": 0.365234375, + "learning_rate": 0.00019983209904618954, + "loss": 2.5209, + "step": 964 + }, + { + "epoch": 0.04, + "grad_norm": 0.4140625, + "learning_rate": 0.00019983171027076763, + "loss": 2.7313, + "step": 965 + }, + { + "epoch": 0.04, + "grad_norm": 0.39453125, + "learning_rate": 0.00019983132104613952, + "loss": 2.7668, + "step": 966 + }, + { + "epoch": 0.04, + "grad_norm": 0.484375, + "learning_rate": 0.0001998309313723069, + "loss": 2.7689, + "step": 967 + }, + { + "epoch": 0.04, + "grad_norm": 0.36328125, + "learning_rate": 0.0001998305412492716, + "loss": 2.791, + "step": 968 + }, + { + "epoch": 0.04, + "grad_norm": 0.4453125, + "learning_rate": 0.00019983015067703535, + "loss": 2.7689, + "step": 969 + }, + { + "epoch": 0.04, + "grad_norm": 0.36328125, + "learning_rate": 0.0001998297596555999, + "loss": 2.7125, + "step": 970 + }, + { + "epoch": 0.04, + "grad_norm": 0.384765625, + "learning_rate": 0.00019982936818496705, + "loss": 2.7416, + "step": 971 + }, + { + "epoch": 0.04, + "grad_norm": 0.404296875, + "learning_rate": 0.0001998289762651385, + "loss": 2.7738, + "step": 972 + }, + { + "epoch": 0.04, + "grad_norm": 0.37890625, + "learning_rate": 0.00019982858389611604, + "loss": 2.717, + "step": 973 + }, + { + "epoch": 0.04, + "grad_norm": 0.388671875, + "learning_rate": 0.0001998281910779014, + "loss": 2.6783, + "step": 974 + }, + { + "epoch": 0.04, + "grad_norm": 0.376953125, + "learning_rate": 0.00019982779781049643, + "loss": 2.8953, + "step": 975 + }, + { + "epoch": 0.04, + "grad_norm": 0.408203125, + "learning_rate": 0.00019982740409390287, + "loss": 2.7128, + "step": 976 + }, + { + "epoch": 0.04, + "grad_norm": 0.404296875, + "learning_rate": 0.00019982700992812242, + "loss": 2.7434, + "step": 977 + }, + { + "epoch": 0.04, + "grad_norm": 0.38671875, + "learning_rate": 0.00019982661531315693, + "loss": 2.7204, + "step": 978 + }, + { + "epoch": 0.04, + "grad_norm": 0.36328125, + "learning_rate": 0.00019982622024900817, + "loss": 2.8663, + "step": 979 + }, + { + "epoch": 0.04, + "grad_norm": 0.4296875, + "learning_rate": 0.00019982582473567785, + "loss": 2.9168, + "step": 980 + }, + { + "epoch": 0.04, + "grad_norm": 0.40625, + "learning_rate": 0.00019982542877316787, + "loss": 2.8995, + "step": 981 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.00019982503236147992, + "loss": 2.7, + "step": 982 + }, + { + "epoch": 0.04, + "grad_norm": 0.44921875, + "learning_rate": 0.00019982463550061578, + "loss": 2.7957, + "step": 983 + }, + { + "epoch": 0.04, + "grad_norm": 0.392578125, + "learning_rate": 0.00019982423819057726, + "loss": 2.7677, + "step": 984 + }, + { + "epoch": 0.04, + "grad_norm": 0.400390625, + "learning_rate": 0.0001998238404313662, + "loss": 2.8407, + "step": 985 + }, + { + "epoch": 0.04, + "grad_norm": 0.35546875, + "learning_rate": 0.0001998234422229843, + "loss": 2.6961, + "step": 986 + }, + { + "epoch": 0.04, + "grad_norm": 0.376953125, + "learning_rate": 0.0001998230435654334, + "loss": 2.6056, + "step": 987 + }, + { + "epoch": 0.04, + "grad_norm": 0.392578125, + "learning_rate": 0.0001998226444587153, + "loss": 2.8699, + "step": 988 + }, + { + "epoch": 0.04, + "grad_norm": 0.33984375, + "learning_rate": 0.00019982224490283174, + "loss": 2.6993, + "step": 989 + }, + { + "epoch": 0.04, + "grad_norm": 0.3828125, + "learning_rate": 0.00019982184489778456, + "loss": 2.6285, + "step": 990 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.0001998214444435756, + "loss": 2.8632, + "step": 991 + }, + { + "epoch": 0.04, + "grad_norm": 0.46875, + "learning_rate": 0.0001998210435402066, + "loss": 2.8337, + "step": 992 + }, + { + "epoch": 0.04, + "grad_norm": 0.40234375, + "learning_rate": 0.00019982064218767938, + "loss": 2.7682, + "step": 993 + }, + { + "epoch": 0.04, + "grad_norm": 0.3828125, + "learning_rate": 0.00019982024038599577, + "loss": 2.6836, + "step": 994 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.00019981983813515756, + "loss": 2.7163, + "step": 995 + }, + { + "epoch": 0.04, + "grad_norm": 0.380859375, + "learning_rate": 0.00019981943543516653, + "loss": 2.6953, + "step": 996 + }, + { + "epoch": 0.04, + "grad_norm": 0.4140625, + "learning_rate": 0.00019981903228602455, + "loss": 2.6644, + "step": 997 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.00019981862868773337, + "loss": 2.6766, + "step": 998 + }, + { + "epoch": 0.04, + "grad_norm": 0.41796875, + "learning_rate": 0.00019981822464029488, + "loss": 2.769, + "step": 999 + }, + { + "epoch": 0.04, + "grad_norm": 0.384765625, + "learning_rate": 0.0001998178201437108, + "loss": 2.787, + "step": 1000 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.00019981741519798307, + "loss": 2.7641, + "step": 1001 + }, + { + "epoch": 0.04, + "grad_norm": 0.40234375, + "learning_rate": 0.00019981700980311345, + "loss": 2.8225, + "step": 1002 + }, + { + "epoch": 0.04, + "grad_norm": 0.408203125, + "learning_rate": 0.00019981660395910372, + "loss": 2.6549, + "step": 1003 + }, + { + "epoch": 0.04, + "grad_norm": 0.419921875, + "learning_rate": 0.00019981619766595577, + "loss": 2.7811, + "step": 1004 + }, + { + "epoch": 0.04, + "grad_norm": 0.388671875, + "learning_rate": 0.00019981579092367142, + "loss": 2.7035, + "step": 1005 + }, + { + "epoch": 0.04, + "grad_norm": 0.3984375, + "learning_rate": 0.00019981538373225249, + "loss": 2.8386, + "step": 1006 + }, + { + "epoch": 0.04, + "grad_norm": 0.384765625, + "learning_rate": 0.0001998149760917008, + "loss": 2.8294, + "step": 1007 + }, + { + "epoch": 0.04, + "grad_norm": 0.349609375, + "learning_rate": 0.00019981456800201818, + "loss": 2.6131, + "step": 1008 + }, + { + "epoch": 0.04, + "grad_norm": 0.416015625, + "learning_rate": 0.0001998141594632065, + "loss": 2.9039, + "step": 1009 + }, + { + "epoch": 0.04, + "grad_norm": 0.3515625, + "learning_rate": 0.00019981375047526762, + "loss": 2.6994, + "step": 1010 + }, + { + "epoch": 0.04, + "grad_norm": 0.345703125, + "learning_rate": 0.00019981334103820326, + "loss": 2.6721, + "step": 1011 + }, + { + "epoch": 0.04, + "grad_norm": 0.390625, + "learning_rate": 0.00019981293115201541, + "loss": 2.6865, + "step": 1012 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.00019981252081670582, + "loss": 2.894, + "step": 1013 + }, + { + "epoch": 0.04, + "grad_norm": 0.390625, + "learning_rate": 0.00019981211003227634, + "loss": 2.8131, + "step": 1014 + }, + { + "epoch": 0.04, + "grad_norm": 0.380859375, + "learning_rate": 0.00019981169879872889, + "loss": 2.772, + "step": 1015 + }, + { + "epoch": 0.04, + "grad_norm": 0.400390625, + "learning_rate": 0.00019981128711606526, + "loss": 2.6958, + "step": 1016 + }, + { + "epoch": 0.04, + "grad_norm": 0.419921875, + "learning_rate": 0.0001998108749842873, + "loss": 2.6507, + "step": 1017 + }, + { + "epoch": 0.04, + "grad_norm": 0.369140625, + "learning_rate": 0.00019981046240339688, + "loss": 2.8301, + "step": 1018 + }, + { + "epoch": 0.04, + "grad_norm": 0.421875, + "learning_rate": 0.00019981004937339586, + "loss": 2.8455, + "step": 1019 + }, + { + "epoch": 0.04, + "grad_norm": 0.404296875, + "learning_rate": 0.0001998096358942861, + "loss": 2.8067, + "step": 1020 + }, + { + "epoch": 0.04, + "grad_norm": 0.40625, + "learning_rate": 0.00019980922196606945, + "loss": 2.7061, + "step": 1021 + }, + { + "epoch": 0.04, + "grad_norm": 0.3828125, + "learning_rate": 0.00019980880758874777, + "loss": 2.7493, + "step": 1022 + }, + { + "epoch": 0.04, + "grad_norm": 0.46484375, + "learning_rate": 0.00019980839276232293, + "loss": 2.7142, + "step": 1023 + }, + { + "epoch": 0.04, + "grad_norm": 0.384765625, + "learning_rate": 0.00019980797748679683, + "loss": 2.7601, + "step": 1024 + }, + { + "epoch": 0.04, + "grad_norm": 0.40625, + "learning_rate": 0.00019980756176217127, + "loss": 2.7406, + "step": 1025 + }, + { + "epoch": 0.04, + "grad_norm": 0.373046875, + "learning_rate": 0.00019980714558844818, + "loss": 2.6718, + "step": 1026 + }, + { + "epoch": 0.04, + "grad_norm": 0.380859375, + "learning_rate": 0.0001998067289656294, + "loss": 2.7601, + "step": 1027 + }, + { + "epoch": 0.04, + "grad_norm": 0.404296875, + "learning_rate": 0.0001998063118937168, + "loss": 2.7758, + "step": 1028 + }, + { + "epoch": 0.04, + "grad_norm": 0.3828125, + "learning_rate": 0.0001998058943727123, + "loss": 2.8031, + "step": 1029 + }, + { + "epoch": 0.04, + "grad_norm": 0.3828125, + "learning_rate": 0.00019980547640261774, + "loss": 2.7371, + "step": 1030 + }, + { + "epoch": 0.04, + "grad_norm": 0.41796875, + "learning_rate": 0.000199805057983435, + "loss": 2.8315, + "step": 1031 + }, + { + "epoch": 0.04, + "grad_norm": 0.4296875, + "learning_rate": 0.00019980463911516597, + "loss": 2.6994, + "step": 1032 + }, + { + "epoch": 0.04, + "grad_norm": 0.38671875, + "learning_rate": 0.00019980421979781256, + "loss": 2.9324, + "step": 1033 + }, + { + "epoch": 0.04, + "grad_norm": 0.404296875, + "learning_rate": 0.00019980380003137662, + "loss": 2.755, + "step": 1034 + }, + { + "epoch": 0.04, + "grad_norm": 0.361328125, + "learning_rate": 0.00019980337981586005, + "loss": 2.6836, + "step": 1035 + }, + { + "epoch": 0.04, + "grad_norm": 0.451171875, + "learning_rate": 0.00019980295915126477, + "loss": 2.7743, + "step": 1036 + }, + { + "epoch": 0.04, + "grad_norm": 0.365234375, + "learning_rate": 0.00019980253803759262, + "loss": 2.8334, + "step": 1037 + }, + { + "epoch": 0.04, + "grad_norm": 0.421875, + "learning_rate": 0.00019980211647484554, + "loss": 2.8798, + "step": 1038 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.0001998016944630254, + "loss": 2.8147, + "step": 1039 + }, + { + "epoch": 0.04, + "grad_norm": 0.44921875, + "learning_rate": 0.00019980127200213413, + "loss": 2.8461, + "step": 1040 + }, + { + "epoch": 0.04, + "grad_norm": 0.384765625, + "learning_rate": 0.00019980084909217358, + "loss": 2.8476, + "step": 1041 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.0001998004257331457, + "loss": 2.6852, + "step": 1042 + }, + { + "epoch": 0.04, + "grad_norm": 0.37109375, + "learning_rate": 0.00019980000192505238, + "loss": 2.757, + "step": 1043 + }, + { + "epoch": 0.04, + "grad_norm": 0.400390625, + "learning_rate": 0.0001997995776678955, + "loss": 2.8584, + "step": 1044 + }, + { + "epoch": 0.04, + "grad_norm": 0.4375, + "learning_rate": 0.000199799152961677, + "loss": 2.7665, + "step": 1045 + }, + { + "epoch": 0.04, + "grad_norm": 0.375, + "learning_rate": 0.0001997987278063988, + "loss": 2.7804, + "step": 1046 + }, + { + "epoch": 0.04, + "grad_norm": 0.427734375, + "learning_rate": 0.00019979830220206283, + "loss": 2.74, + "step": 1047 + }, + { + "epoch": 0.04, + "grad_norm": 0.34765625, + "learning_rate": 0.0001997978761486709, + "loss": 2.6542, + "step": 1048 + }, + { + "epoch": 0.04, + "grad_norm": 0.40625, + "learning_rate": 0.00019979744964622502, + "loss": 2.5845, + "step": 1049 + }, + { + "epoch": 0.04, + "grad_norm": 0.341796875, + "learning_rate": 0.0001997970226947271, + "loss": 2.6352, + "step": 1050 + }, + { + "epoch": 0.04, + "grad_norm": 0.41015625, + "learning_rate": 0.00019979659529417907, + "loss": 2.7471, + "step": 1051 + }, + { + "epoch": 0.04, + "grad_norm": 0.38671875, + "learning_rate": 0.0001997961674445828, + "loss": 2.8438, + "step": 1052 + }, + { + "epoch": 0.04, + "grad_norm": 0.400390625, + "learning_rate": 0.00019979573914594026, + "loss": 2.7895, + "step": 1053 + }, + { + "epoch": 0.04, + "grad_norm": 0.376953125, + "learning_rate": 0.00019979531039825336, + "loss": 2.9076, + "step": 1054 + }, + { + "epoch": 0.04, + "grad_norm": 0.423828125, + "learning_rate": 0.00019979488120152404, + "loss": 2.8292, + "step": 1055 + }, + { + "epoch": 0.04, + "grad_norm": 0.396484375, + "learning_rate": 0.00019979445155575423, + "loss": 2.6039, + "step": 1056 + }, + { + "epoch": 0.05, + "grad_norm": 0.412109375, + "learning_rate": 0.00019979402146094586, + "loss": 2.7418, + "step": 1057 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.00019979359091710084, + "loss": 2.7091, + "step": 1058 + }, + { + "epoch": 0.05, + "grad_norm": 0.36328125, + "learning_rate": 0.00019979315992422114, + "loss": 2.7368, + "step": 1059 + }, + { + "epoch": 0.05, + "grad_norm": 0.41015625, + "learning_rate": 0.00019979272848230872, + "loss": 2.8388, + "step": 1060 + }, + { + "epoch": 0.05, + "grad_norm": 0.41796875, + "learning_rate": 0.00019979229659136544, + "loss": 2.724, + "step": 1061 + }, + { + "epoch": 0.05, + "grad_norm": 0.365234375, + "learning_rate": 0.00019979186425139335, + "loss": 2.7185, + "step": 1062 + }, + { + "epoch": 0.05, + "grad_norm": 0.384765625, + "learning_rate": 0.00019979143146239432, + "loss": 2.6574, + "step": 1063 + }, + { + "epoch": 0.05, + "grad_norm": 0.322265625, + "learning_rate": 0.0001997909982243703, + "loss": 2.7088, + "step": 1064 + }, + { + "epoch": 0.05, + "grad_norm": 0.41015625, + "learning_rate": 0.00019979056453732328, + "loss": 2.8788, + "step": 1065 + }, + { + "epoch": 0.05, + "grad_norm": 0.373046875, + "learning_rate": 0.00019979013040125518, + "loss": 2.6762, + "step": 1066 + }, + { + "epoch": 0.05, + "grad_norm": 0.375, + "learning_rate": 0.00019978969581616797, + "loss": 2.7654, + "step": 1067 + }, + { + "epoch": 0.05, + "grad_norm": 0.423828125, + "learning_rate": 0.00019978926078206358, + "loss": 2.7751, + "step": 1068 + }, + { + "epoch": 0.05, + "grad_norm": 0.375, + "learning_rate": 0.000199788825298944, + "loss": 2.6884, + "step": 1069 + }, + { + "epoch": 0.05, + "grad_norm": 0.373046875, + "learning_rate": 0.00019978838936681118, + "loss": 2.6724, + "step": 1070 + }, + { + "epoch": 0.05, + "grad_norm": 0.4296875, + "learning_rate": 0.00019978795298566707, + "loss": 2.7573, + "step": 1071 + }, + { + "epoch": 0.05, + "grad_norm": 0.384765625, + "learning_rate": 0.0001997875161555136, + "loss": 2.679, + "step": 1072 + }, + { + "epoch": 0.05, + "grad_norm": 0.494140625, + "learning_rate": 0.00019978707887635284, + "loss": 2.8191, + "step": 1073 + }, + { + "epoch": 0.05, + "grad_norm": 0.392578125, + "learning_rate": 0.00019978664114818665, + "loss": 2.8367, + "step": 1074 + }, + { + "epoch": 0.05, + "grad_norm": 0.392578125, + "learning_rate": 0.00019978620297101706, + "loss": 2.8034, + "step": 1075 + }, + { + "epoch": 0.05, + "grad_norm": 0.419921875, + "learning_rate": 0.00019978576434484603, + "loss": 2.7466, + "step": 1076 + }, + { + "epoch": 0.05, + "grad_norm": 0.3671875, + "learning_rate": 0.00019978532526967552, + "loss": 2.7674, + "step": 1077 + }, + { + "epoch": 0.05, + "grad_norm": 0.396484375, + "learning_rate": 0.0001997848857455075, + "loss": 2.704, + "step": 1078 + }, + { + "epoch": 0.05, + "grad_norm": 0.427734375, + "learning_rate": 0.00019978444577234397, + "loss": 2.8428, + "step": 1079 + }, + { + "epoch": 0.05, + "grad_norm": 0.4140625, + "learning_rate": 0.00019978400535018692, + "loss": 2.8025, + "step": 1080 + }, + { + "epoch": 0.05, + "grad_norm": 0.41796875, + "learning_rate": 0.0001997835644790383, + "loss": 2.883, + "step": 1081 + }, + { + "epoch": 0.05, + "grad_norm": 0.365234375, + "learning_rate": 0.0001997831231589001, + "loss": 2.7366, + "step": 1082 + }, + { + "epoch": 0.05, + "grad_norm": 0.40234375, + "learning_rate": 0.00019978268138977435, + "loss": 2.7959, + "step": 1083 + }, + { + "epoch": 0.05, + "grad_norm": 0.361328125, + "learning_rate": 0.00019978223917166294, + "loss": 2.7287, + "step": 1084 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.00019978179650456795, + "loss": 2.699, + "step": 1085 + }, + { + "epoch": 0.05, + "grad_norm": 0.400390625, + "learning_rate": 0.00019978135338849135, + "loss": 2.779, + "step": 1086 + }, + { + "epoch": 0.05, + "grad_norm": 0.390625, + "learning_rate": 0.0001997809098234351, + "loss": 2.6427, + "step": 1087 + }, + { + "epoch": 0.05, + "grad_norm": 0.37109375, + "learning_rate": 0.00019978046580940127, + "loss": 2.8521, + "step": 1088 + }, + { + "epoch": 0.05, + "grad_norm": 0.41015625, + "learning_rate": 0.00019978002134639176, + "loss": 2.7689, + "step": 1089 + }, + { + "epoch": 0.05, + "grad_norm": 0.47265625, + "learning_rate": 0.00019977957643440864, + "loss": 2.7849, + "step": 1090 + }, + { + "epoch": 0.05, + "grad_norm": 0.435546875, + "learning_rate": 0.0001997791310734539, + "loss": 2.8278, + "step": 1091 + }, + { + "epoch": 0.05, + "grad_norm": 0.373046875, + "learning_rate": 0.0001997786852635295, + "loss": 2.6928, + "step": 1092 + }, + { + "epoch": 0.05, + "grad_norm": 0.44140625, + "learning_rate": 0.0001997782390046375, + "loss": 2.7616, + "step": 1093 + }, + { + "epoch": 0.05, + "grad_norm": 0.388671875, + "learning_rate": 0.00019977779229677988, + "loss": 2.8625, + "step": 1094 + }, + { + "epoch": 0.05, + "grad_norm": 0.396484375, + "learning_rate": 0.00019977734513995868, + "loss": 2.5792, + "step": 1095 + }, + { + "epoch": 0.05, + "grad_norm": 0.439453125, + "learning_rate": 0.0001997768975341759, + "loss": 2.9749, + "step": 1096 + }, + { + "epoch": 0.05, + "grad_norm": 0.3984375, + "learning_rate": 0.00019977644947943352, + "loss": 2.5984, + "step": 1097 + }, + { + "epoch": 0.05, + "grad_norm": 0.49609375, + "learning_rate": 0.00019977600097573358, + "loss": 2.8107, + "step": 1098 + }, + { + "epoch": 0.05, + "grad_norm": 0.40234375, + "learning_rate": 0.00019977555202307815, + "loss": 2.7979, + "step": 1099 + }, + { + "epoch": 0.05, + "grad_norm": 0.498046875, + "learning_rate": 0.00019977510262146915, + "loss": 2.8356, + "step": 1100 + }, + { + "epoch": 0.05, + "grad_norm": 0.453125, + "learning_rate": 0.00019977465277090864, + "loss": 2.8662, + "step": 1101 + }, + { + "epoch": 0.05, + "grad_norm": 0.435546875, + "learning_rate": 0.0001997742024713987, + "loss": 2.6772, + "step": 1102 + }, + { + "epoch": 0.05, + "grad_norm": 0.375, + "learning_rate": 0.00019977375172294126, + "loss": 2.7349, + "step": 1103 + }, + { + "epoch": 0.05, + "grad_norm": 0.416015625, + "learning_rate": 0.00019977330052553844, + "loss": 2.6574, + "step": 1104 + }, + { + "epoch": 0.05, + "grad_norm": 0.390625, + "learning_rate": 0.0001997728488791922, + "loss": 2.7375, + "step": 1105 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.00019977239678390464, + "loss": 2.6151, + "step": 1106 + }, + { + "epoch": 0.05, + "grad_norm": 0.466796875, + "learning_rate": 0.00019977194423967775, + "loss": 2.8165, + "step": 1107 + }, + { + "epoch": 0.05, + "grad_norm": 0.423828125, + "learning_rate": 0.00019977149124651358, + "loss": 2.753, + "step": 1108 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.00019977103780441412, + "loss": 2.7741, + "step": 1109 + }, + { + "epoch": 0.05, + "grad_norm": 0.365234375, + "learning_rate": 0.00019977058391338148, + "loss": 2.5828, + "step": 1110 + }, + { + "epoch": 0.05, + "grad_norm": 0.373046875, + "learning_rate": 0.0001997701295734177, + "loss": 2.7994, + "step": 1111 + }, + { + "epoch": 0.05, + "grad_norm": 0.46484375, + "learning_rate": 0.00019976967478452476, + "loss": 2.848, + "step": 1112 + }, + { + "epoch": 0.05, + "grad_norm": 0.35546875, + "learning_rate": 0.00019976921954670474, + "loss": 2.6064, + "step": 1113 + }, + { + "epoch": 0.05, + "grad_norm": 0.3671875, + "learning_rate": 0.00019976876385995973, + "loss": 2.6031, + "step": 1114 + }, + { + "epoch": 0.05, + "grad_norm": 0.404296875, + "learning_rate": 0.0001997683077242917, + "loss": 2.7886, + "step": 1115 + }, + { + "epoch": 0.05, + "grad_norm": 0.361328125, + "learning_rate": 0.00019976785113970277, + "loss": 2.7608, + "step": 1116 + }, + { + "epoch": 0.05, + "grad_norm": 0.455078125, + "learning_rate": 0.00019976739410619497, + "loss": 2.8269, + "step": 1117 + }, + { + "epoch": 0.05, + "grad_norm": 0.353515625, + "learning_rate": 0.00019976693662377036, + "loss": 2.7227, + "step": 1118 + }, + { + "epoch": 0.05, + "grad_norm": 0.478515625, + "learning_rate": 0.000199766478692431, + "loss": 2.8164, + "step": 1119 + }, + { + "epoch": 0.05, + "grad_norm": 0.388671875, + "learning_rate": 0.00019976602031217895, + "loss": 2.7418, + "step": 1120 + }, + { + "epoch": 0.05, + "grad_norm": 0.4140625, + "learning_rate": 0.00019976556148301625, + "loss": 2.7591, + "step": 1121 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.000199765102204945, + "loss": 2.8311, + "step": 1122 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.00019976464247796724, + "loss": 2.7518, + "step": 1123 + }, + { + "epoch": 0.05, + "grad_norm": 0.40625, + "learning_rate": 0.00019976418230208506, + "loss": 2.7707, + "step": 1124 + }, + { + "epoch": 0.05, + "grad_norm": 0.34765625, + "learning_rate": 0.0001997637216773005, + "loss": 2.7336, + "step": 1125 + }, + { + "epoch": 0.05, + "grad_norm": 0.3828125, + "learning_rate": 0.00019976326060361566, + "loss": 2.7381, + "step": 1126 + }, + { + "epoch": 0.05, + "grad_norm": 0.40625, + "learning_rate": 0.0001997627990810326, + "loss": 2.8329, + "step": 1127 + }, + { + "epoch": 0.05, + "grad_norm": 0.39453125, + "learning_rate": 0.00019976233710955344, + "loss": 2.751, + "step": 1128 + }, + { + "epoch": 0.05, + "grad_norm": 0.38671875, + "learning_rate": 0.0001997618746891802, + "loss": 2.7584, + "step": 1129 + }, + { + "epoch": 0.05, + "grad_norm": 0.38671875, + "learning_rate": 0.00019976141181991496, + "loss": 2.8425, + "step": 1130 + }, + { + "epoch": 0.05, + "grad_norm": 0.353515625, + "learning_rate": 0.00019976094850175983, + "loss": 2.7212, + "step": 1131 + }, + { + "epoch": 0.05, + "grad_norm": 0.4375, + "learning_rate": 0.0001997604847347169, + "loss": 2.7949, + "step": 1132 + }, + { + "epoch": 0.05, + "grad_norm": 0.375, + "learning_rate": 0.00019976002051878827, + "loss": 2.8072, + "step": 1133 + }, + { + "epoch": 0.05, + "grad_norm": 0.353515625, + "learning_rate": 0.00019975955585397598, + "loss": 2.8405, + "step": 1134 + }, + { + "epoch": 0.05, + "grad_norm": 0.3515625, + "learning_rate": 0.00019975909074028216, + "loss": 2.6398, + "step": 1135 + }, + { + "epoch": 0.05, + "grad_norm": 0.3828125, + "learning_rate": 0.00019975862517770886, + "loss": 2.7585, + "step": 1136 + }, + { + "epoch": 0.05, + "grad_norm": 0.37890625, + "learning_rate": 0.00019975815916625824, + "loss": 2.7191, + "step": 1137 + }, + { + "epoch": 0.05, + "grad_norm": 0.369140625, + "learning_rate": 0.00019975769270593233, + "loss": 2.6713, + "step": 1138 + }, + { + "epoch": 0.05, + "grad_norm": 0.3984375, + "learning_rate": 0.00019975722579673329, + "loss": 2.9426, + "step": 1139 + }, + { + "epoch": 0.05, + "grad_norm": 0.392578125, + "learning_rate": 0.00019975675843866318, + "loss": 2.8375, + "step": 1140 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.0001997562906317241, + "loss": 2.947, + "step": 1141 + }, + { + "epoch": 0.05, + "grad_norm": 0.326171875, + "learning_rate": 0.00019975582237591818, + "loss": 2.6514, + "step": 1142 + }, + { + "epoch": 0.05, + "grad_norm": 0.412109375, + "learning_rate": 0.00019975535367124754, + "loss": 2.6778, + "step": 1143 + }, + { + "epoch": 0.05, + "grad_norm": 0.380859375, + "learning_rate": 0.0001997548845177142, + "loss": 2.7624, + "step": 1144 + }, + { + "epoch": 0.05, + "grad_norm": 0.380859375, + "learning_rate": 0.0001997544149153204, + "loss": 2.6612, + "step": 1145 + }, + { + "epoch": 0.05, + "grad_norm": 0.3984375, + "learning_rate": 0.00019975394486406817, + "loss": 2.7137, + "step": 1146 + }, + { + "epoch": 0.05, + "grad_norm": 0.37890625, + "learning_rate": 0.00019975347436395965, + "loss": 2.8383, + "step": 1147 + }, + { + "epoch": 0.05, + "grad_norm": 0.427734375, + "learning_rate": 0.00019975300341499693, + "loss": 2.7738, + "step": 1148 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.00019975253201718216, + "loss": 2.6181, + "step": 1149 + }, + { + "epoch": 0.05, + "grad_norm": 0.40625, + "learning_rate": 0.00019975206017051748, + "loss": 2.8381, + "step": 1150 + }, + { + "epoch": 0.05, + "grad_norm": 0.427734375, + "learning_rate": 0.00019975158787500493, + "loss": 2.8192, + "step": 1151 + }, + { + "epoch": 0.05, + "grad_norm": 0.41015625, + "learning_rate": 0.00019975111513064673, + "loss": 2.7952, + "step": 1152 + }, + { + "epoch": 0.05, + "grad_norm": 0.396484375, + "learning_rate": 0.00019975064193744494, + "loss": 2.6195, + "step": 1153 + }, + { + "epoch": 0.05, + "grad_norm": 0.42578125, + "learning_rate": 0.00019975016829540173, + "loss": 2.7536, + "step": 1154 + }, + { + "epoch": 0.05, + "grad_norm": 0.47265625, + "learning_rate": 0.00019974969420451924, + "loss": 2.943, + "step": 1155 + }, + { + "epoch": 0.05, + "grad_norm": 0.404296875, + "learning_rate": 0.00019974921966479955, + "loss": 2.7297, + "step": 1156 + }, + { + "epoch": 0.05, + "grad_norm": 0.4296875, + "learning_rate": 0.00019974874467624485, + "loss": 2.8096, + "step": 1157 + }, + { + "epoch": 0.05, + "grad_norm": 0.388671875, + "learning_rate": 0.0001997482692388572, + "loss": 2.6425, + "step": 1158 + }, + { + "epoch": 0.05, + "grad_norm": 0.37109375, + "learning_rate": 0.00019974779335263886, + "loss": 2.6923, + "step": 1159 + }, + { + "epoch": 0.05, + "grad_norm": 0.421875, + "learning_rate": 0.00019974731701759185, + "loss": 2.7239, + "step": 1160 + }, + { + "epoch": 0.05, + "grad_norm": 0.38671875, + "learning_rate": 0.00019974684023371843, + "loss": 2.7548, + "step": 1161 + }, + { + "epoch": 0.05, + "grad_norm": 0.423828125, + "learning_rate": 0.00019974636300102062, + "loss": 2.8876, + "step": 1162 + }, + { + "epoch": 0.05, + "grad_norm": 0.4296875, + "learning_rate": 0.00019974588531950063, + "loss": 2.6772, + "step": 1163 + }, + { + "epoch": 0.05, + "grad_norm": 0.4140625, + "learning_rate": 0.00019974540718916065, + "loss": 2.7042, + "step": 1164 + }, + { + "epoch": 0.05, + "grad_norm": 0.447265625, + "learning_rate": 0.00019974492861000277, + "loss": 2.8468, + "step": 1165 + }, + { + "epoch": 0.05, + "grad_norm": 0.373046875, + "learning_rate": 0.00019974444958202915, + "loss": 2.6244, + "step": 1166 + }, + { + "epoch": 0.05, + "grad_norm": 0.392578125, + "learning_rate": 0.00019974397010524196, + "loss": 2.8165, + "step": 1167 + }, + { + "epoch": 0.05, + "grad_norm": 0.416015625, + "learning_rate": 0.00019974349017964338, + "loss": 2.8737, + "step": 1168 + }, + { + "epoch": 0.05, + "grad_norm": 0.384765625, + "learning_rate": 0.00019974300980523554, + "loss": 2.8722, + "step": 1169 + }, + { + "epoch": 0.05, + "grad_norm": 0.396484375, + "learning_rate": 0.00019974252898202058, + "loss": 2.7398, + "step": 1170 + }, + { + "epoch": 0.05, + "grad_norm": 0.392578125, + "learning_rate": 0.00019974204771000076, + "loss": 2.8176, + "step": 1171 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.0001997415659891781, + "loss": 2.7998, + "step": 1172 + }, + { + "epoch": 0.05, + "grad_norm": 0.34375, + "learning_rate": 0.0001997410838195549, + "loss": 2.6974, + "step": 1173 + }, + { + "epoch": 0.05, + "grad_norm": 0.375, + "learning_rate": 0.00019974060120113324, + "loss": 2.7792, + "step": 1174 + }, + { + "epoch": 0.05, + "grad_norm": 0.369140625, + "learning_rate": 0.00019974011813391536, + "loss": 2.7668, + "step": 1175 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.00019973963461790336, + "loss": 2.9063, + "step": 1176 + }, + { + "epoch": 0.05, + "grad_norm": 0.359375, + "learning_rate": 0.00019973915065309947, + "loss": 2.7141, + "step": 1177 + }, + { + "epoch": 0.05, + "grad_norm": 0.404296875, + "learning_rate": 0.00019973866623950584, + "loss": 2.7246, + "step": 1178 + }, + { + "epoch": 0.05, + "grad_norm": 0.375, + "learning_rate": 0.00019973818137712467, + "loss": 2.7191, + "step": 1179 + }, + { + "epoch": 0.05, + "grad_norm": 0.390625, + "learning_rate": 0.00019973769606595817, + "loss": 2.8547, + "step": 1180 + }, + { + "epoch": 0.05, + "grad_norm": 0.36328125, + "learning_rate": 0.00019973721030600844, + "loss": 2.7642, + "step": 1181 + }, + { + "epoch": 0.05, + "grad_norm": 0.4296875, + "learning_rate": 0.00019973672409727772, + "loss": 2.5895, + "step": 1182 + }, + { + "epoch": 0.05, + "grad_norm": 0.38671875, + "learning_rate": 0.0001997362374397682, + "loss": 2.8322, + "step": 1183 + }, + { + "epoch": 0.05, + "grad_norm": 0.4453125, + "learning_rate": 0.00019973575033348207, + "loss": 2.8871, + "step": 1184 + }, + { + "epoch": 0.05, + "grad_norm": 0.388671875, + "learning_rate": 0.0001997352627784215, + "loss": 2.6959, + "step": 1185 + }, + { + "epoch": 0.05, + "grad_norm": 0.380859375, + "learning_rate": 0.00019973477477458872, + "loss": 2.8372, + "step": 1186 + }, + { + "epoch": 0.05, + "grad_norm": 0.416015625, + "learning_rate": 0.00019973428632198586, + "loss": 2.8623, + "step": 1187 + }, + { + "epoch": 0.05, + "grad_norm": 0.34765625, + "learning_rate": 0.0001997337974206152, + "loss": 2.6638, + "step": 1188 + }, + { + "epoch": 0.05, + "grad_norm": 0.453125, + "learning_rate": 0.0001997333080704789, + "loss": 2.7264, + "step": 1189 + }, + { + "epoch": 0.05, + "grad_norm": 0.388671875, + "learning_rate": 0.00019973281827157915, + "loss": 2.7912, + "step": 1190 + }, + { + "epoch": 0.05, + "grad_norm": 0.400390625, + "learning_rate": 0.00019973232802391816, + "loss": 2.6959, + "step": 1191 + }, + { + "epoch": 0.05, + "grad_norm": 0.41015625, + "learning_rate": 0.0001997318373274981, + "loss": 2.7546, + "step": 1192 + }, + { + "epoch": 0.05, + "grad_norm": 0.392578125, + "learning_rate": 0.0001997313461823213, + "loss": 2.7897, + "step": 1193 + }, + { + "epoch": 0.05, + "grad_norm": 0.4375, + "learning_rate": 0.00019973085458838984, + "loss": 2.7252, + "step": 1194 + }, + { + "epoch": 0.05, + "grad_norm": 0.365234375, + "learning_rate": 0.000199730362545706, + "loss": 2.7835, + "step": 1195 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.00019972987005427198, + "loss": 2.7673, + "step": 1196 + }, + { + "epoch": 0.05, + "grad_norm": 0.458984375, + "learning_rate": 0.00019972937711408995, + "loss": 2.7273, + "step": 1197 + }, + { + "epoch": 0.05, + "grad_norm": 0.390625, + "learning_rate": 0.00019972888372516223, + "loss": 2.7443, + "step": 1198 + }, + { + "epoch": 0.05, + "grad_norm": 0.39453125, + "learning_rate": 0.00019972838988749095, + "loss": 2.6028, + "step": 1199 + }, + { + "epoch": 0.05, + "grad_norm": 0.46484375, + "learning_rate": 0.00019972789560107837, + "loss": 2.9105, + "step": 1200 + }, + { + "epoch": 0.05, + "grad_norm": 0.380859375, + "learning_rate": 0.0001997274008659267, + "loss": 2.6712, + "step": 1201 + }, + { + "epoch": 0.05, + "grad_norm": 0.41015625, + "learning_rate": 0.0001997269056820382, + "loss": 2.7476, + "step": 1202 + }, + { + "epoch": 0.05, + "grad_norm": 0.458984375, + "learning_rate": 0.00019972641004941503, + "loss": 2.8514, + "step": 1203 + }, + { + "epoch": 0.05, + "grad_norm": 0.439453125, + "learning_rate": 0.00019972591396805948, + "loss": 2.7163, + "step": 1204 + }, + { + "epoch": 0.05, + "grad_norm": 0.4375, + "learning_rate": 0.00019972541743797374, + "loss": 2.8814, + "step": 1205 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.0001997249204591601, + "loss": 2.73, + "step": 1206 + }, + { + "epoch": 0.05, + "grad_norm": 0.423828125, + "learning_rate": 0.00019972442303162072, + "loss": 2.7668, + "step": 1207 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.00019972392515535793, + "loss": 2.6186, + "step": 1208 + }, + { + "epoch": 0.05, + "grad_norm": 0.3828125, + "learning_rate": 0.0001997234268303739, + "loss": 2.7903, + "step": 1209 + }, + { + "epoch": 0.05, + "grad_norm": 0.34375, + "learning_rate": 0.0001997229280566709, + "loss": 2.7652, + "step": 1210 + }, + { + "epoch": 0.05, + "grad_norm": 0.390625, + "learning_rate": 0.00019972242883425117, + "loss": 2.8083, + "step": 1211 + }, + { + "epoch": 0.05, + "grad_norm": 0.3984375, + "learning_rate": 0.00019972192916311695, + "loss": 2.7778, + "step": 1212 + }, + { + "epoch": 0.05, + "grad_norm": 0.3671875, + "learning_rate": 0.00019972142904327048, + "loss": 2.6382, + "step": 1213 + }, + { + "epoch": 0.05, + "grad_norm": 0.40625, + "learning_rate": 0.00019972092847471404, + "loss": 2.6459, + "step": 1214 + }, + { + "epoch": 0.05, + "grad_norm": 0.42578125, + "learning_rate": 0.00019972042745744985, + "loss": 2.7575, + "step": 1215 + }, + { + "epoch": 0.05, + "grad_norm": 0.3828125, + "learning_rate": 0.0001997199259914802, + "loss": 2.6954, + "step": 1216 + }, + { + "epoch": 0.05, + "grad_norm": 0.388671875, + "learning_rate": 0.0001997194240768073, + "loss": 2.7773, + "step": 1217 + }, + { + "epoch": 0.05, + "grad_norm": 0.375, + "learning_rate": 0.00019971892171343348, + "loss": 2.7311, + "step": 1218 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.0001997184189013609, + "loss": 2.8687, + "step": 1219 + }, + { + "epoch": 0.05, + "grad_norm": 0.365234375, + "learning_rate": 0.0001997179156405919, + "loss": 2.7496, + "step": 1220 + }, + { + "epoch": 0.05, + "grad_norm": 0.396484375, + "learning_rate": 0.00019971741193112875, + "loss": 2.7503, + "step": 1221 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.00019971690777297364, + "loss": 2.6172, + "step": 1222 + }, + { + "epoch": 0.05, + "grad_norm": 0.43359375, + "learning_rate": 0.00019971640316612894, + "loss": 2.7501, + "step": 1223 + }, + { + "epoch": 0.05, + "grad_norm": 0.412109375, + "learning_rate": 0.00019971589811059685, + "loss": 2.7683, + "step": 1224 + }, + { + "epoch": 0.05, + "grad_norm": 0.44140625, + "learning_rate": 0.00019971539260637964, + "loss": 2.7014, + "step": 1225 + }, + { + "epoch": 0.05, + "grad_norm": 0.3984375, + "learning_rate": 0.00019971488665347962, + "loss": 2.7971, + "step": 1226 + }, + { + "epoch": 0.05, + "grad_norm": 0.451171875, + "learning_rate": 0.00019971438025189904, + "loss": 2.7462, + "step": 1227 + }, + { + "epoch": 0.05, + "grad_norm": 0.365234375, + "learning_rate": 0.0001997138734016402, + "loss": 2.7658, + "step": 1228 + }, + { + "epoch": 0.05, + "grad_norm": 0.412109375, + "learning_rate": 0.00019971336610270538, + "loss": 2.7647, + "step": 1229 + }, + { + "epoch": 0.05, + "grad_norm": 0.46875, + "learning_rate": 0.00019971285835509683, + "loss": 2.8276, + "step": 1230 + }, + { + "epoch": 0.05, + "grad_norm": 0.412109375, + "learning_rate": 0.00019971235015881687, + "loss": 2.8129, + "step": 1231 + }, + { + "epoch": 0.05, + "grad_norm": 0.494140625, + "learning_rate": 0.00019971184151386777, + "loss": 2.7639, + "step": 1232 + }, + { + "epoch": 0.05, + "grad_norm": 0.404296875, + "learning_rate": 0.00019971133242025185, + "loss": 2.7703, + "step": 1233 + }, + { + "epoch": 0.05, + "grad_norm": 0.421875, + "learning_rate": 0.00019971082287797133, + "loss": 2.8498, + "step": 1234 + }, + { + "epoch": 0.05, + "grad_norm": 0.412109375, + "learning_rate": 0.0001997103128870286, + "loss": 2.7949, + "step": 1235 + }, + { + "epoch": 0.05, + "grad_norm": 0.421875, + "learning_rate": 0.00019970980244742584, + "loss": 2.7392, + "step": 1236 + }, + { + "epoch": 0.05, + "grad_norm": 0.365234375, + "learning_rate": 0.00019970929155916546, + "loss": 2.6055, + "step": 1237 + }, + { + "epoch": 0.05, + "grad_norm": 0.435546875, + "learning_rate": 0.0001997087802222497, + "loss": 2.785, + "step": 1238 + }, + { + "epoch": 0.05, + "grad_norm": 0.3984375, + "learning_rate": 0.00019970826843668084, + "loss": 2.5358, + "step": 1239 + }, + { + "epoch": 0.05, + "grad_norm": 0.408203125, + "learning_rate": 0.00019970775620246125, + "loss": 2.7725, + "step": 1240 + }, + { + "epoch": 0.05, + "grad_norm": 0.4140625, + "learning_rate": 0.00019970724351959317, + "loss": 2.696, + "step": 1241 + }, + { + "epoch": 0.05, + "grad_norm": 0.404296875, + "learning_rate": 0.00019970673038807894, + "loss": 2.8503, + "step": 1242 + }, + { + "epoch": 0.05, + "grad_norm": 0.369140625, + "learning_rate": 0.00019970621680792086, + "loss": 2.703, + "step": 1243 + }, + { + "epoch": 0.05, + "grad_norm": 0.431640625, + "learning_rate": 0.00019970570277912125, + "loss": 2.6454, + "step": 1244 + }, + { + "epoch": 0.05, + "grad_norm": 0.341796875, + "learning_rate": 0.0001997051883016824, + "loss": 2.5527, + "step": 1245 + }, + { + "epoch": 0.05, + "grad_norm": 0.439453125, + "learning_rate": 0.00019970467337560664, + "loss": 2.7692, + "step": 1246 + }, + { + "epoch": 0.05, + "grad_norm": 0.392578125, + "learning_rate": 0.0001997041580008963, + "loss": 2.9124, + "step": 1247 + }, + { + "epoch": 0.05, + "grad_norm": 0.423828125, + "learning_rate": 0.0001997036421775537, + "loss": 2.947, + "step": 1248 + }, + { + "epoch": 0.05, + "grad_norm": 0.3828125, + "learning_rate": 0.00019970312590558112, + "loss": 2.7362, + "step": 1249 + }, + { + "epoch": 0.05, + "grad_norm": 0.39453125, + "learning_rate": 0.00019970260918498095, + "loss": 2.5994, + "step": 1250 + }, + { + "epoch": 0.05, + "grad_norm": 0.396484375, + "learning_rate": 0.00019970209201575543, + "loss": 2.8055, + "step": 1251 + }, + { + "epoch": 0.05, + "grad_norm": 0.37890625, + "learning_rate": 0.00019970157439790697, + "loss": 2.6261, + "step": 1252 + }, + { + "epoch": 0.05, + "grad_norm": 0.404296875, + "learning_rate": 0.00019970105633143782, + "loss": 2.7717, + "step": 1253 + }, + { + "epoch": 0.05, + "grad_norm": 0.4140625, + "learning_rate": 0.00019970053781635038, + "loss": 2.7628, + "step": 1254 + }, + { + "epoch": 0.05, + "grad_norm": 0.4140625, + "learning_rate": 0.00019970001885264696, + "loss": 2.8353, + "step": 1255 + }, + { + "epoch": 0.05, + "grad_norm": 0.390625, + "learning_rate": 0.00019969949944032987, + "loss": 2.6702, + "step": 1256 + }, + { + "epoch": 0.05, + "grad_norm": 0.384765625, + "learning_rate": 0.00019969897957940148, + "loss": 2.7322, + "step": 1257 + }, + { + "epoch": 0.05, + "grad_norm": 0.412109375, + "learning_rate": 0.00019969845926986413, + "loss": 2.7537, + "step": 1258 + }, + { + "epoch": 0.05, + "grad_norm": 0.419921875, + "learning_rate": 0.00019969793851172014, + "loss": 2.8587, + "step": 1259 + }, + { + "epoch": 0.05, + "grad_norm": 0.3671875, + "learning_rate": 0.0001996974173049719, + "loss": 2.8543, + "step": 1260 + }, + { + "epoch": 0.05, + "grad_norm": 0.357421875, + "learning_rate": 0.00019969689564962163, + "loss": 2.6771, + "step": 1261 + }, + { + "epoch": 0.05, + "grad_norm": 0.388671875, + "learning_rate": 0.00019969637354567185, + "loss": 2.7027, + "step": 1262 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.00019969585099312477, + "loss": 2.6899, + "step": 1263 + }, + { + "epoch": 0.05, + "grad_norm": 0.37890625, + "learning_rate": 0.00019969532799198284, + "loss": 2.8012, + "step": 1264 + }, + { + "epoch": 0.05, + "grad_norm": 0.4140625, + "learning_rate": 0.00019969480454224832, + "loss": 2.7147, + "step": 1265 + }, + { + "epoch": 0.05, + "grad_norm": 0.3671875, + "learning_rate": 0.00019969428064392364, + "loss": 2.6373, + "step": 1266 + }, + { + "epoch": 0.05, + "grad_norm": 0.373046875, + "learning_rate": 0.00019969375629701113, + "loss": 2.8072, + "step": 1267 + }, + { + "epoch": 0.05, + "grad_norm": 0.365234375, + "learning_rate": 0.00019969323150151314, + "loss": 2.6861, + "step": 1268 + }, + { + "epoch": 0.05, + "grad_norm": 0.384765625, + "learning_rate": 0.00019969270625743202, + "loss": 2.777, + "step": 1269 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.0001996921805647702, + "loss": 2.6927, + "step": 1270 + }, + { + "epoch": 0.05, + "grad_norm": 0.396484375, + "learning_rate": 0.00019969165442352998, + "loss": 2.8347, + "step": 1271 + }, + { + "epoch": 0.05, + "grad_norm": 0.35546875, + "learning_rate": 0.00019969112783371377, + "loss": 2.7279, + "step": 1272 + }, + { + "epoch": 0.05, + "grad_norm": 0.443359375, + "learning_rate": 0.00019969060079532393, + "loss": 2.7612, + "step": 1273 + }, + { + "epoch": 0.05, + "grad_norm": 0.376953125, + "learning_rate": 0.00019969007330836275, + "loss": 2.7113, + "step": 1274 + }, + { + "epoch": 0.05, + "grad_norm": 0.373046875, + "learning_rate": 0.00019968954537283274, + "loss": 2.7785, + "step": 1275 + }, + { + "epoch": 0.05, + "grad_norm": 0.404296875, + "learning_rate": 0.00019968901698873617, + "loss": 2.6354, + "step": 1276 + }, + { + "epoch": 0.05, + "grad_norm": 0.4453125, + "learning_rate": 0.00019968848815607546, + "loss": 2.8419, + "step": 1277 + }, + { + "epoch": 0.05, + "grad_norm": 0.4296875, + "learning_rate": 0.000199687958874853, + "loss": 2.8793, + "step": 1278 + }, + { + "epoch": 0.05, + "grad_norm": 0.380859375, + "learning_rate": 0.00019968742914507114, + "loss": 2.8139, + "step": 1279 + }, + { + "epoch": 0.05, + "grad_norm": 0.42578125, + "learning_rate": 0.0001996868989667323, + "loss": 2.7854, + "step": 1280 + }, + { + "epoch": 0.05, + "grad_norm": 0.44140625, + "learning_rate": 0.00019968636833983883, + "loss": 2.89, + "step": 1281 + }, + { + "epoch": 0.05, + "grad_norm": 0.361328125, + "learning_rate": 0.00019968583726439313, + "loss": 2.7496, + "step": 1282 + }, + { + "epoch": 0.05, + "grad_norm": 0.419921875, + "learning_rate": 0.0001996853057403976, + "loss": 2.6643, + "step": 1283 + }, + { + "epoch": 0.05, + "grad_norm": 0.423828125, + "learning_rate": 0.00019968477376785463, + "loss": 2.7586, + "step": 1284 + }, + { + "epoch": 0.05, + "grad_norm": 0.3828125, + "learning_rate": 0.0001996842413467666, + "loss": 2.632, + "step": 1285 + }, + { + "epoch": 0.05, + "grad_norm": 0.404296875, + "learning_rate": 0.00019968370847713593, + "loss": 2.8392, + "step": 1286 + }, + { + "epoch": 0.05, + "grad_norm": 0.380859375, + "learning_rate": 0.000199683175158965, + "loss": 2.7973, + "step": 1287 + }, + { + "epoch": 0.05, + "grad_norm": 0.40234375, + "learning_rate": 0.00019968264139225622, + "loss": 2.7586, + "step": 1288 + }, + { + "epoch": 0.05, + "grad_norm": 0.39453125, + "learning_rate": 0.00019968210717701194, + "loss": 2.7779, + "step": 1289 + }, + { + "epoch": 0.05, + "grad_norm": 0.375, + "learning_rate": 0.00019968157251323466, + "loss": 2.768, + "step": 1290 + }, + { + "epoch": 0.06, + "grad_norm": 0.369140625, + "learning_rate": 0.00019968103740092668, + "loss": 2.6432, + "step": 1291 + }, + { + "epoch": 0.06, + "grad_norm": 0.412109375, + "learning_rate": 0.00019968050184009052, + "loss": 2.8364, + "step": 1292 + }, + { + "epoch": 0.06, + "grad_norm": 0.380859375, + "learning_rate": 0.0001996799658307285, + "loss": 2.7284, + "step": 1293 + }, + { + "epoch": 0.06, + "grad_norm": 0.380859375, + "learning_rate": 0.00019967942937284305, + "loss": 2.7444, + "step": 1294 + }, + { + "epoch": 0.06, + "grad_norm": 0.419921875, + "learning_rate": 0.00019967889246643662, + "loss": 2.6405, + "step": 1295 + }, + { + "epoch": 0.06, + "grad_norm": 0.408203125, + "learning_rate": 0.00019967835511151155, + "loss": 2.8522, + "step": 1296 + }, + { + "epoch": 0.06, + "grad_norm": 0.466796875, + "learning_rate": 0.00019967781730807036, + "loss": 2.8677, + "step": 1297 + }, + { + "epoch": 0.06, + "grad_norm": 0.435546875, + "learning_rate": 0.00019967727905611545, + "loss": 2.8503, + "step": 1298 + }, + { + "epoch": 0.06, + "grad_norm": 0.419921875, + "learning_rate": 0.00019967674035564916, + "loss": 2.6253, + "step": 1299 + }, + { + "epoch": 0.06, + "grad_norm": 0.396484375, + "learning_rate": 0.00019967620120667397, + "loss": 2.7583, + "step": 1300 + }, + { + "epoch": 0.06, + "grad_norm": 0.412109375, + "learning_rate": 0.0001996756616091923, + "loss": 2.6036, + "step": 1301 + }, + { + "epoch": 0.06, + "grad_norm": 0.361328125, + "learning_rate": 0.0001996751215632066, + "loss": 2.7057, + "step": 1302 + }, + { + "epoch": 0.06, + "grad_norm": 0.40234375, + "learning_rate": 0.0001996745810687193, + "loss": 2.6715, + "step": 1303 + }, + { + "epoch": 0.06, + "grad_norm": 0.41015625, + "learning_rate": 0.00019967404012573276, + "loss": 2.6436, + "step": 1304 + }, + { + "epoch": 0.06, + "grad_norm": 0.41796875, + "learning_rate": 0.0001996734987342495, + "loss": 2.7564, + "step": 1305 + }, + { + "epoch": 0.06, + "grad_norm": 0.408203125, + "learning_rate": 0.0001996729568942719, + "loss": 2.7149, + "step": 1306 + }, + { + "epoch": 0.06, + "grad_norm": 0.408203125, + "learning_rate": 0.00019967241460580243, + "loss": 2.7548, + "step": 1307 + }, + { + "epoch": 0.06, + "grad_norm": 0.4453125, + "learning_rate": 0.00019967187186884354, + "loss": 2.711, + "step": 1308 + }, + { + "epoch": 0.06, + "grad_norm": 0.400390625, + "learning_rate": 0.00019967132868339764, + "loss": 2.7586, + "step": 1309 + }, + { + "epoch": 0.06, + "grad_norm": 0.41015625, + "learning_rate": 0.00019967078504946716, + "loss": 2.7755, + "step": 1310 + }, + { + "epoch": 0.06, + "grad_norm": 0.390625, + "learning_rate": 0.0001996702409670546, + "loss": 2.8388, + "step": 1311 + }, + { + "epoch": 0.06, + "grad_norm": 0.408203125, + "learning_rate": 0.00019966969643616238, + "loss": 2.838, + "step": 1312 + }, + { + "epoch": 0.06, + "grad_norm": 0.388671875, + "learning_rate": 0.00019966915145679294, + "loss": 2.7001, + "step": 1313 + }, + { + "epoch": 0.06, + "grad_norm": 0.412109375, + "learning_rate": 0.00019966860602894877, + "loss": 2.7249, + "step": 1314 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019966806015263226, + "loss": 2.6976, + "step": 1315 + }, + { + "epoch": 0.06, + "grad_norm": 0.4296875, + "learning_rate": 0.0001996675138278459, + "loss": 2.7623, + "step": 1316 + }, + { + "epoch": 0.06, + "grad_norm": 0.4765625, + "learning_rate": 0.0001996669670545922, + "loss": 2.8246, + "step": 1317 + }, + { + "epoch": 0.06, + "grad_norm": 0.412109375, + "learning_rate": 0.00019966641983287355, + "loss": 2.6196, + "step": 1318 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.00019966587216269243, + "loss": 2.7351, + "step": 1319 + }, + { + "epoch": 0.06, + "grad_norm": 0.390625, + "learning_rate": 0.00019966532404405132, + "loss": 2.7513, + "step": 1320 + }, + { + "epoch": 0.06, + "grad_norm": 0.43359375, + "learning_rate": 0.00019966477547695265, + "loss": 2.7315, + "step": 1321 + }, + { + "epoch": 0.06, + "grad_norm": 0.396484375, + "learning_rate": 0.00019966422646139895, + "loss": 2.7222, + "step": 1322 + }, + { + "epoch": 0.06, + "grad_norm": 0.396484375, + "learning_rate": 0.00019966367699739262, + "loss": 2.709, + "step": 1323 + }, + { + "epoch": 0.06, + "grad_norm": 0.431640625, + "learning_rate": 0.0001996631270849362, + "loss": 2.7506, + "step": 1324 + }, + { + "epoch": 0.06, + "grad_norm": 0.38671875, + "learning_rate": 0.0001996625767240321, + "loss": 2.704, + "step": 1325 + }, + { + "epoch": 0.06, + "grad_norm": 0.365234375, + "learning_rate": 0.0001996620259146828, + "loss": 2.6927, + "step": 1326 + }, + { + "epoch": 0.06, + "grad_norm": 0.44921875, + "learning_rate": 0.00019966147465689087, + "loss": 2.7284, + "step": 1327 + }, + { + "epoch": 0.06, + "grad_norm": 0.38671875, + "learning_rate": 0.0001996609229506587, + "loss": 2.7541, + "step": 1328 + }, + { + "epoch": 0.06, + "grad_norm": 0.453125, + "learning_rate": 0.0001996603707959888, + "loss": 2.8266, + "step": 1329 + }, + { + "epoch": 0.06, + "grad_norm": 0.435546875, + "learning_rate": 0.00019965981819288365, + "loss": 2.739, + "step": 1330 + }, + { + "epoch": 0.06, + "grad_norm": 0.44921875, + "learning_rate": 0.00019965926514134574, + "loss": 2.7464, + "step": 1331 + }, + { + "epoch": 0.06, + "grad_norm": 0.41015625, + "learning_rate": 0.00019965871164137755, + "loss": 2.7677, + "step": 1332 + }, + { + "epoch": 0.06, + "grad_norm": 0.423828125, + "learning_rate": 0.0001996581576929816, + "loss": 2.8463, + "step": 1333 + }, + { + "epoch": 0.06, + "grad_norm": 0.392578125, + "learning_rate": 0.00019965760329616035, + "loss": 2.7805, + "step": 1334 + }, + { + "epoch": 0.06, + "grad_norm": 0.416015625, + "learning_rate": 0.0001996570484509163, + "loss": 2.7583, + "step": 1335 + }, + { + "epoch": 0.06, + "grad_norm": 0.388671875, + "learning_rate": 0.00019965649315725195, + "loss": 2.6682, + "step": 1336 + }, + { + "epoch": 0.06, + "grad_norm": 0.390625, + "learning_rate": 0.00019965593741516982, + "loss": 2.7368, + "step": 1337 + }, + { + "epoch": 0.06, + "grad_norm": 0.466796875, + "learning_rate": 0.00019965538122467238, + "loss": 2.8559, + "step": 1338 + }, + { + "epoch": 0.06, + "grad_norm": 0.3828125, + "learning_rate": 0.00019965482458576215, + "loss": 2.8215, + "step": 1339 + }, + { + "epoch": 0.06, + "grad_norm": 0.486328125, + "learning_rate": 0.00019965426749844162, + "loss": 2.7535, + "step": 1340 + }, + { + "epoch": 0.06, + "grad_norm": 0.390625, + "learning_rate": 0.00019965370996271331, + "loss": 2.6048, + "step": 1341 + }, + { + "epoch": 0.06, + "grad_norm": 0.423828125, + "learning_rate": 0.00019965315197857973, + "loss": 2.7309, + "step": 1342 + }, + { + "epoch": 0.06, + "grad_norm": 0.431640625, + "learning_rate": 0.00019965259354604337, + "loss": 2.6996, + "step": 1343 + }, + { + "epoch": 0.06, + "grad_norm": 0.41015625, + "learning_rate": 0.00019965203466510678, + "loss": 2.8535, + "step": 1344 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019965147533577241, + "loss": 2.8604, + "step": 1345 + }, + { + "epoch": 0.06, + "grad_norm": 0.38671875, + "learning_rate": 0.00019965091555804287, + "loss": 2.8585, + "step": 1346 + }, + { + "epoch": 0.06, + "grad_norm": 0.365234375, + "learning_rate": 0.0001996503553319206, + "loss": 2.7129, + "step": 1347 + }, + { + "epoch": 0.06, + "grad_norm": 0.3984375, + "learning_rate": 0.00019964979465740817, + "loss": 2.802, + "step": 1348 + }, + { + "epoch": 0.06, + "grad_norm": 0.40625, + "learning_rate": 0.00019964923353450807, + "loss": 2.6555, + "step": 1349 + }, + { + "epoch": 0.06, + "grad_norm": 0.36328125, + "learning_rate": 0.0001996486719632228, + "loss": 2.6409, + "step": 1350 + }, + { + "epoch": 0.06, + "grad_norm": 0.41015625, + "learning_rate": 0.000199648109943555, + "loss": 2.7055, + "step": 1351 + }, + { + "epoch": 0.06, + "grad_norm": 0.3828125, + "learning_rate": 0.00019964754747550708, + "loss": 2.6971, + "step": 1352 + }, + { + "epoch": 0.06, + "grad_norm": 0.455078125, + "learning_rate": 0.0001996469845590816, + "loss": 2.6973, + "step": 1353 + }, + { + "epoch": 0.06, + "grad_norm": 0.3828125, + "learning_rate": 0.00019964642119428113, + "loss": 2.605, + "step": 1354 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.00019964585738110814, + "loss": 2.7263, + "step": 1355 + }, + { + "epoch": 0.06, + "grad_norm": 0.421875, + "learning_rate": 0.00019964529311956524, + "loss": 2.7381, + "step": 1356 + }, + { + "epoch": 0.06, + "grad_norm": 0.37109375, + "learning_rate": 0.00019964472840965496, + "loss": 2.7034, + "step": 1357 + }, + { + "epoch": 0.06, + "grad_norm": 0.453125, + "learning_rate": 0.00019964416325137977, + "loss": 2.665, + "step": 1358 + }, + { + "epoch": 0.06, + "grad_norm": 0.392578125, + "learning_rate": 0.0001996435976447423, + "loss": 2.7413, + "step": 1359 + }, + { + "epoch": 0.06, + "grad_norm": 0.435546875, + "learning_rate": 0.00019964303158974503, + "loss": 2.7922, + "step": 1360 + }, + { + "epoch": 0.06, + "grad_norm": 0.421875, + "learning_rate": 0.00019964246508639058, + "loss": 2.6762, + "step": 1361 + }, + { + "epoch": 0.06, + "grad_norm": 0.3984375, + "learning_rate": 0.0001996418981346814, + "loss": 2.6735, + "step": 1362 + }, + { + "epoch": 0.06, + "grad_norm": 0.412109375, + "learning_rate": 0.0001996413307346201, + "loss": 2.6921, + "step": 1363 + }, + { + "epoch": 0.06, + "grad_norm": 0.404296875, + "learning_rate": 0.00019964076288620926, + "loss": 2.8728, + "step": 1364 + }, + { + "epoch": 0.06, + "grad_norm": 0.404296875, + "learning_rate": 0.00019964019458945136, + "loss": 2.7824, + "step": 1365 + }, + { + "epoch": 0.06, + "grad_norm": 0.431640625, + "learning_rate": 0.00019963962584434902, + "loss": 2.7537, + "step": 1366 + }, + { + "epoch": 0.06, + "grad_norm": 0.3828125, + "learning_rate": 0.00019963905665090477, + "loss": 2.7183, + "step": 1367 + }, + { + "epoch": 0.06, + "grad_norm": 0.3828125, + "learning_rate": 0.0001996384870091212, + "loss": 2.7975, + "step": 1368 + }, + { + "epoch": 0.06, + "grad_norm": 0.359375, + "learning_rate": 0.00019963791691900084, + "loss": 2.7321, + "step": 1369 + }, + { + "epoch": 0.06, + "grad_norm": 0.37890625, + "learning_rate": 0.00019963734638054625, + "loss": 2.8158, + "step": 1370 + }, + { + "epoch": 0.06, + "grad_norm": 0.408203125, + "learning_rate": 0.00019963677539376002, + "loss": 2.8415, + "step": 1371 + }, + { + "epoch": 0.06, + "grad_norm": 0.40234375, + "learning_rate": 0.00019963620395864476, + "loss": 2.8343, + "step": 1372 + }, + { + "epoch": 0.06, + "grad_norm": 0.376953125, + "learning_rate": 0.00019963563207520295, + "loss": 2.8161, + "step": 1373 + }, + { + "epoch": 0.06, + "grad_norm": 0.392578125, + "learning_rate": 0.0001996350597434372, + "loss": 2.7957, + "step": 1374 + }, + { + "epoch": 0.06, + "grad_norm": 0.369140625, + "learning_rate": 0.0001996344869633501, + "loss": 2.7217, + "step": 1375 + }, + { + "epoch": 0.06, + "grad_norm": 0.412109375, + "learning_rate": 0.00019963391373494424, + "loss": 2.717, + "step": 1376 + }, + { + "epoch": 0.06, + "grad_norm": 0.38671875, + "learning_rate": 0.00019963334005822217, + "loss": 2.7599, + "step": 1377 + }, + { + "epoch": 0.06, + "grad_norm": 0.392578125, + "learning_rate": 0.00019963276593318647, + "loss": 2.7437, + "step": 1378 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019963219135983975, + "loss": 2.7578, + "step": 1379 + }, + { + "epoch": 0.06, + "grad_norm": 0.5546875, + "learning_rate": 0.0001996316163381846, + "loss": 2.8257, + "step": 1380 + }, + { + "epoch": 0.06, + "grad_norm": 0.37109375, + "learning_rate": 0.00019963104086822355, + "loss": 2.7035, + "step": 1381 + }, + { + "epoch": 0.06, + "grad_norm": 0.388671875, + "learning_rate": 0.00019963046494995924, + "loss": 2.6125, + "step": 1382 + }, + { + "epoch": 0.06, + "grad_norm": 0.4140625, + "learning_rate": 0.00019962988858339425, + "loss": 2.884, + "step": 1383 + }, + { + "epoch": 0.06, + "grad_norm": 0.4140625, + "learning_rate": 0.00019962931176853118, + "loss": 2.8788, + "step": 1384 + }, + { + "epoch": 0.06, + "grad_norm": 0.388671875, + "learning_rate": 0.00019962873450537263, + "loss": 2.6987, + "step": 1385 + }, + { + "epoch": 0.06, + "grad_norm": 0.4140625, + "learning_rate": 0.00019962815679392114, + "loss": 2.6878, + "step": 1386 + }, + { + "epoch": 0.06, + "grad_norm": 0.400390625, + "learning_rate": 0.00019962757863417938, + "loss": 2.8994, + "step": 1387 + }, + { + "epoch": 0.06, + "grad_norm": 0.380859375, + "learning_rate": 0.00019962700002614994, + "loss": 2.6566, + "step": 1388 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019962642096983538, + "loss": 2.7937, + "step": 1389 + }, + { + "epoch": 0.06, + "grad_norm": 0.388671875, + "learning_rate": 0.00019962584146523835, + "loss": 2.8892, + "step": 1390 + }, + { + "epoch": 0.06, + "grad_norm": 0.419921875, + "learning_rate": 0.00019962526151236145, + "loss": 2.7332, + "step": 1391 + }, + { + "epoch": 0.06, + "grad_norm": 0.373046875, + "learning_rate": 0.00019962468111120725, + "loss": 2.534, + "step": 1392 + }, + { + "epoch": 0.06, + "grad_norm": 0.38671875, + "learning_rate": 0.00019962410026177842, + "loss": 2.6257, + "step": 1393 + }, + { + "epoch": 0.06, + "grad_norm": 0.365234375, + "learning_rate": 0.00019962351896407756, + "loss": 2.7034, + "step": 1394 + }, + { + "epoch": 0.06, + "grad_norm": 0.359375, + "learning_rate": 0.0001996229372181073, + "loss": 2.7464, + "step": 1395 + }, + { + "epoch": 0.06, + "grad_norm": 0.408203125, + "learning_rate": 0.00019962235502387013, + "loss": 2.7143, + "step": 1396 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019962177238136883, + "loss": 2.7298, + "step": 1397 + }, + { + "epoch": 0.06, + "grad_norm": 0.38671875, + "learning_rate": 0.00019962118929060598, + "loss": 2.7054, + "step": 1398 + }, + { + "epoch": 0.06, + "grad_norm": 0.4375, + "learning_rate": 0.00019962060575158416, + "loss": 2.7282, + "step": 1399 + }, + { + "epoch": 0.06, + "grad_norm": 0.40234375, + "learning_rate": 0.00019962002176430603, + "loss": 2.6822, + "step": 1400 + }, + { + "epoch": 0.06, + "grad_norm": 0.40234375, + "learning_rate": 0.00019961943732877422, + "loss": 2.7568, + "step": 1401 + }, + { + "epoch": 0.06, + "grad_norm": 0.40625, + "learning_rate": 0.00019961885244499132, + "loss": 2.8174, + "step": 1402 + }, + { + "epoch": 0.06, + "grad_norm": 0.416015625, + "learning_rate": 0.00019961826711296002, + "loss": 2.7715, + "step": 1403 + }, + { + "epoch": 0.06, + "grad_norm": 0.427734375, + "learning_rate": 0.0001996176813326829, + "loss": 2.8073, + "step": 1404 + }, + { + "epoch": 0.06, + "grad_norm": 0.4140625, + "learning_rate": 0.00019961709510416263, + "loss": 2.8711, + "step": 1405 + }, + { + "epoch": 0.06, + "grad_norm": 0.369140625, + "learning_rate": 0.00019961650842740186, + "loss": 2.7149, + "step": 1406 + }, + { + "epoch": 0.06, + "grad_norm": 0.416015625, + "learning_rate": 0.00019961592130240317, + "loss": 2.7737, + "step": 1407 + }, + { + "epoch": 0.06, + "grad_norm": 0.369140625, + "learning_rate": 0.00019961533372916926, + "loss": 2.7538, + "step": 1408 + }, + { + "epoch": 0.06, + "grad_norm": 0.388671875, + "learning_rate": 0.00019961474570770273, + "loss": 2.6914, + "step": 1409 + }, + { + "epoch": 0.06, + "grad_norm": 0.396484375, + "learning_rate": 0.00019961415723800628, + "loss": 2.6939, + "step": 1410 + }, + { + "epoch": 0.06, + "grad_norm": 0.451171875, + "learning_rate": 0.00019961356832008252, + "loss": 2.719, + "step": 1411 + }, + { + "epoch": 0.06, + "grad_norm": 0.37890625, + "learning_rate": 0.00019961297895393408, + "loss": 2.7437, + "step": 1412 + }, + { + "epoch": 0.06, + "grad_norm": 0.423828125, + "learning_rate": 0.00019961238913956367, + "loss": 2.7766, + "step": 1413 + }, + { + "epoch": 0.06, + "grad_norm": 0.396484375, + "learning_rate": 0.0001996117988769739, + "loss": 2.8816, + "step": 1414 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.00019961120816616745, + "loss": 2.7233, + "step": 1415 + }, + { + "epoch": 0.06, + "grad_norm": 0.390625, + "learning_rate": 0.00019961061700714695, + "loss": 2.9264, + "step": 1416 + }, + { + "epoch": 0.06, + "grad_norm": 0.365234375, + "learning_rate": 0.0001996100253999151, + "loss": 2.7529, + "step": 1417 + }, + { + "epoch": 0.06, + "grad_norm": 0.35546875, + "learning_rate": 0.00019960943334447454, + "loss": 2.601, + "step": 1418 + }, + { + "epoch": 0.06, + "grad_norm": 0.357421875, + "learning_rate": 0.00019960884084082792, + "loss": 2.7824, + "step": 1419 + }, + { + "epoch": 0.06, + "grad_norm": 0.392578125, + "learning_rate": 0.0001996082478889779, + "loss": 2.7704, + "step": 1420 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.0001996076544889272, + "loss": 2.7206, + "step": 1421 + }, + { + "epoch": 0.06, + "grad_norm": 0.421875, + "learning_rate": 0.00019960706064067843, + "loss": 2.6936, + "step": 1422 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.0001996064663442343, + "loss": 2.7901, + "step": 1423 + }, + { + "epoch": 0.06, + "grad_norm": 0.3984375, + "learning_rate": 0.00019960587159959748, + "loss": 2.7629, + "step": 1424 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.00019960527640677062, + "loss": 2.6544, + "step": 1425 + }, + { + "epoch": 0.06, + "grad_norm": 0.380859375, + "learning_rate": 0.00019960468076575644, + "loss": 2.6551, + "step": 1426 + }, + { + "epoch": 0.06, + "grad_norm": 0.400390625, + "learning_rate": 0.0001996040846765576, + "loss": 2.7557, + "step": 1427 + }, + { + "epoch": 0.06, + "grad_norm": 0.40625, + "learning_rate": 0.00019960348813917678, + "loss": 2.7305, + "step": 1428 + }, + { + "epoch": 0.06, + "grad_norm": 0.353515625, + "learning_rate": 0.00019960289115361664, + "loss": 2.809, + "step": 1429 + }, + { + "epoch": 0.06, + "grad_norm": 0.421875, + "learning_rate": 0.0001996022937198799, + "loss": 2.6578, + "step": 1430 + }, + { + "epoch": 0.06, + "grad_norm": 0.37109375, + "learning_rate": 0.00019960169583796926, + "loss": 2.8897, + "step": 1431 + }, + { + "epoch": 0.06, + "grad_norm": 0.419921875, + "learning_rate": 0.0001996010975078874, + "loss": 2.8582, + "step": 1432 + }, + { + "epoch": 0.06, + "grad_norm": 0.40234375, + "learning_rate": 0.00019960049872963693, + "loss": 2.7145, + "step": 1433 + }, + { + "epoch": 0.06, + "grad_norm": 0.408203125, + "learning_rate": 0.00019959989950322067, + "loss": 2.8803, + "step": 1434 + }, + { + "epoch": 0.06, + "grad_norm": 0.361328125, + "learning_rate": 0.00019959929982864126, + "loss": 2.6798, + "step": 1435 + }, + { + "epoch": 0.06, + "grad_norm": 0.34765625, + "learning_rate": 0.0001995986997059014, + "loss": 2.8352, + "step": 1436 + }, + { + "epoch": 0.06, + "grad_norm": 0.44140625, + "learning_rate": 0.00019959809913500378, + "loss": 2.7267, + "step": 1437 + }, + { + "epoch": 0.06, + "grad_norm": 0.359375, + "learning_rate": 0.0001995974981159511, + "loss": 2.713, + "step": 1438 + }, + { + "epoch": 0.06, + "grad_norm": 0.361328125, + "learning_rate": 0.0001995968966487461, + "loss": 2.6793, + "step": 1439 + }, + { + "epoch": 0.06, + "grad_norm": 0.404296875, + "learning_rate": 0.00019959629473339147, + "loss": 2.8629, + "step": 1440 + }, + { + "epoch": 0.06, + "grad_norm": 0.421875, + "learning_rate": 0.00019959569236988988, + "loss": 2.7967, + "step": 1441 + }, + { + "epoch": 0.06, + "grad_norm": 0.37109375, + "learning_rate": 0.0001995950895582441, + "loss": 2.8085, + "step": 1442 + }, + { + "epoch": 0.06, + "grad_norm": 0.3984375, + "learning_rate": 0.0001995944862984568, + "loss": 2.7468, + "step": 1443 + }, + { + "epoch": 0.06, + "grad_norm": 0.375, + "learning_rate": 0.00019959388259053071, + "loss": 2.6254, + "step": 1444 + }, + { + "epoch": 0.06, + "grad_norm": 0.412109375, + "learning_rate": 0.00019959327843446857, + "loss": 2.8061, + "step": 1445 + }, + { + "epoch": 0.06, + "grad_norm": 0.419921875, + "learning_rate": 0.00019959267383027306, + "loss": 2.8766, + "step": 1446 + }, + { + "epoch": 0.06, + "grad_norm": 0.376953125, + "learning_rate": 0.0001995920687779469, + "loss": 2.6783, + "step": 1447 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.00019959146327749283, + "loss": 2.9334, + "step": 1448 + }, + { + "epoch": 0.06, + "grad_norm": 0.361328125, + "learning_rate": 0.0001995908573289136, + "loss": 2.7572, + "step": 1449 + }, + { + "epoch": 0.06, + "grad_norm": 0.4140625, + "learning_rate": 0.00019959025093221186, + "loss": 2.65, + "step": 1450 + }, + { + "epoch": 0.06, + "grad_norm": 0.41015625, + "learning_rate": 0.00019958964408739043, + "loss": 2.7596, + "step": 1451 + }, + { + "epoch": 0.06, + "grad_norm": 0.3828125, + "learning_rate": 0.00019958903679445197, + "loss": 2.7359, + "step": 1452 + }, + { + "epoch": 0.06, + "grad_norm": 0.34765625, + "learning_rate": 0.00019958842905339925, + "loss": 2.6906, + "step": 1453 + }, + { + "epoch": 0.06, + "grad_norm": 0.404296875, + "learning_rate": 0.000199587820864235, + "loss": 2.7852, + "step": 1454 + }, + { + "epoch": 0.06, + "grad_norm": 0.3828125, + "learning_rate": 0.00019958721222696195, + "loss": 2.7444, + "step": 1455 + }, + { + "epoch": 0.06, + "grad_norm": 0.41015625, + "learning_rate": 0.00019958660314158284, + "loss": 2.732, + "step": 1456 + }, + { + "epoch": 0.06, + "grad_norm": 0.46484375, + "learning_rate": 0.00019958599360810043, + "loss": 2.8024, + "step": 1457 + }, + { + "epoch": 0.06, + "grad_norm": 0.435546875, + "learning_rate": 0.0001995853836265174, + "loss": 2.5738, + "step": 1458 + }, + { + "epoch": 0.06, + "grad_norm": 0.35546875, + "learning_rate": 0.00019958477319683655, + "loss": 2.776, + "step": 1459 + }, + { + "epoch": 0.06, + "grad_norm": 0.361328125, + "learning_rate": 0.00019958416231906064, + "loss": 2.6982, + "step": 1460 + }, + { + "epoch": 0.06, + "grad_norm": 0.43359375, + "learning_rate": 0.00019958355099319237, + "loss": 2.6483, + "step": 1461 + }, + { + "epoch": 0.06, + "grad_norm": 0.345703125, + "learning_rate": 0.00019958293921923452, + "loss": 2.7428, + "step": 1462 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.00019958232699718984, + "loss": 2.7559, + "step": 1463 + }, + { + "epoch": 0.06, + "grad_norm": 0.392578125, + "learning_rate": 0.00019958171432706107, + "loss": 2.7689, + "step": 1464 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.000199581101208851, + "loss": 2.8955, + "step": 1465 + }, + { + "epoch": 0.06, + "grad_norm": 0.408203125, + "learning_rate": 0.00019958048764256235, + "loss": 2.7676, + "step": 1466 + }, + { + "epoch": 0.06, + "grad_norm": 0.400390625, + "learning_rate": 0.0001995798736281979, + "loss": 2.8308, + "step": 1467 + }, + { + "epoch": 0.06, + "grad_norm": 0.458984375, + "learning_rate": 0.0001995792591657604, + "loss": 2.8626, + "step": 1468 + }, + { + "epoch": 0.06, + "grad_norm": 0.388671875, + "learning_rate": 0.00019957864425525267, + "loss": 2.8104, + "step": 1469 + }, + { + "epoch": 0.06, + "grad_norm": 0.369140625, + "learning_rate": 0.0001995780288966774, + "loss": 2.6867, + "step": 1470 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.0001995774130900374, + "loss": 2.7549, + "step": 1471 + }, + { + "epoch": 0.06, + "grad_norm": 0.357421875, + "learning_rate": 0.0001995767968353354, + "loss": 2.6305, + "step": 1472 + }, + { + "epoch": 0.06, + "grad_norm": 0.423828125, + "learning_rate": 0.00019957618013257425, + "loss": 2.6969, + "step": 1473 + }, + { + "epoch": 0.06, + "grad_norm": 0.3828125, + "learning_rate": 0.00019957556298175666, + "loss": 2.6943, + "step": 1474 + }, + { + "epoch": 0.06, + "grad_norm": 0.36328125, + "learning_rate": 0.00019957494538288542, + "loss": 2.749, + "step": 1475 + }, + { + "epoch": 0.06, + "grad_norm": 0.671875, + "learning_rate": 0.00019957432733596332, + "loss": 2.7117, + "step": 1476 + }, + { + "epoch": 0.06, + "grad_norm": 0.42578125, + "learning_rate": 0.00019957370884099315, + "loss": 2.6811, + "step": 1477 + }, + { + "epoch": 0.06, + "grad_norm": 0.3671875, + "learning_rate": 0.0001995730898979777, + "loss": 2.6572, + "step": 1478 + }, + { + "epoch": 0.06, + "grad_norm": 0.390625, + "learning_rate": 0.00019957247050691969, + "loss": 2.731, + "step": 1479 + }, + { + "epoch": 0.06, + "grad_norm": 0.36328125, + "learning_rate": 0.00019957185066782197, + "loss": 2.6605, + "step": 1480 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019957123038068732, + "loss": 2.7937, + "step": 1481 + }, + { + "epoch": 0.06, + "grad_norm": 0.4375, + "learning_rate": 0.0001995706096455185, + "loss": 2.7939, + "step": 1482 + }, + { + "epoch": 0.06, + "grad_norm": 0.453125, + "learning_rate": 0.00019956998846231837, + "loss": 2.6326, + "step": 1483 + }, + { + "epoch": 0.06, + "grad_norm": 0.400390625, + "learning_rate": 0.00019956936683108962, + "loss": 2.6174, + "step": 1484 + }, + { + "epoch": 0.06, + "grad_norm": 0.49609375, + "learning_rate": 0.00019956874475183516, + "loss": 2.8021, + "step": 1485 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.0001995681222245577, + "loss": 2.8076, + "step": 1486 + }, + { + "epoch": 0.06, + "grad_norm": 0.482421875, + "learning_rate": 0.0001995674992492601, + "loss": 2.7639, + "step": 1487 + }, + { + "epoch": 0.06, + "grad_norm": 0.46484375, + "learning_rate": 0.00019956687582594517, + "loss": 2.6008, + "step": 1488 + }, + { + "epoch": 0.06, + "grad_norm": 0.416015625, + "learning_rate": 0.00019956625195461562, + "loss": 2.6414, + "step": 1489 + }, + { + "epoch": 0.06, + "grad_norm": 0.51171875, + "learning_rate": 0.00019956562763527436, + "loss": 2.7534, + "step": 1490 + }, + { + "epoch": 0.06, + "grad_norm": 0.359375, + "learning_rate": 0.00019956500286792418, + "loss": 2.6804, + "step": 1491 + }, + { + "epoch": 0.06, + "grad_norm": 0.45703125, + "learning_rate": 0.00019956437765256783, + "loss": 2.7634, + "step": 1492 + }, + { + "epoch": 0.06, + "grad_norm": 0.40234375, + "learning_rate": 0.0001995637519892082, + "loss": 2.6281, + "step": 1493 + }, + { + "epoch": 0.06, + "grad_norm": 0.38671875, + "learning_rate": 0.00019956312587784803, + "loss": 2.7197, + "step": 1494 + }, + { + "epoch": 0.06, + "grad_norm": 0.45703125, + "learning_rate": 0.00019956249931849023, + "loss": 2.6846, + "step": 1495 + }, + { + "epoch": 0.06, + "grad_norm": 0.373046875, + "learning_rate": 0.00019956187231113753, + "loss": 2.7206, + "step": 1496 + }, + { + "epoch": 0.06, + "grad_norm": 0.41015625, + "learning_rate": 0.00019956124485579282, + "loss": 2.8379, + "step": 1497 + }, + { + "epoch": 0.06, + "grad_norm": 0.3984375, + "learning_rate": 0.0001995606169524589, + "loss": 2.8853, + "step": 1498 + }, + { + "epoch": 0.06, + "grad_norm": 0.376953125, + "learning_rate": 0.00019955998860113857, + "loss": 2.7121, + "step": 1499 + }, + { + "epoch": 0.06, + "grad_norm": 0.400390625, + "learning_rate": 0.00019955935980183468, + "loss": 2.7315, + "step": 1500 + }, + { + "epoch": 0.06, + "grad_norm": 0.357421875, + "learning_rate": 0.00019955873055455006, + "loss": 2.6593, + "step": 1501 + }, + { + "epoch": 0.06, + "grad_norm": 0.380859375, + "learning_rate": 0.00019955810085928752, + "loss": 2.639, + "step": 1502 + }, + { + "epoch": 0.06, + "grad_norm": 0.40234375, + "learning_rate": 0.0001995574707160499, + "loss": 2.5535, + "step": 1503 + }, + { + "epoch": 0.06, + "grad_norm": 0.375, + "learning_rate": 0.0001995568401248401, + "loss": 2.6431, + "step": 1504 + }, + { + "epoch": 0.06, + "grad_norm": 0.404296875, + "learning_rate": 0.00019955620908566085, + "loss": 2.7781, + "step": 1505 + }, + { + "epoch": 0.06, + "grad_norm": 0.361328125, + "learning_rate": 0.00019955557759851507, + "loss": 2.8159, + "step": 1506 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.00019955494566340558, + "loss": 2.6577, + "step": 1507 + }, + { + "epoch": 0.06, + "grad_norm": 0.4140625, + "learning_rate": 0.0001995543132803352, + "loss": 2.7526, + "step": 1508 + }, + { + "epoch": 0.06, + "grad_norm": 0.353515625, + "learning_rate": 0.0001995536804493068, + "loss": 2.6957, + "step": 1509 + }, + { + "epoch": 0.06, + "grad_norm": 0.412109375, + "learning_rate": 0.00019955304717032322, + "loss": 2.7323, + "step": 1510 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.0001995524134433873, + "loss": 2.8324, + "step": 1511 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019955177926850193, + "loss": 2.7936, + "step": 1512 + }, + { + "epoch": 0.06, + "grad_norm": 0.380859375, + "learning_rate": 0.0001995511446456699, + "loss": 2.7726, + "step": 1513 + }, + { + "epoch": 0.06, + "grad_norm": 0.451171875, + "learning_rate": 0.00019955050957489413, + "loss": 2.7862, + "step": 1514 + }, + { + "epoch": 0.06, + "grad_norm": 0.439453125, + "learning_rate": 0.00019954987405617743, + "loss": 2.7804, + "step": 1515 + }, + { + "epoch": 0.06, + "grad_norm": 0.384765625, + "learning_rate": 0.00019954923808952272, + "loss": 2.554, + "step": 1516 + }, + { + "epoch": 0.06, + "grad_norm": 0.42578125, + "learning_rate": 0.00019954860167493274, + "loss": 2.77, + "step": 1517 + }, + { + "epoch": 0.06, + "grad_norm": 0.400390625, + "learning_rate": 0.0001995479648124105, + "loss": 2.748, + "step": 1518 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019954732750195877, + "loss": 2.7744, + "step": 1519 + }, + { + "epoch": 0.06, + "grad_norm": 0.39453125, + "learning_rate": 0.00019954668974358044, + "loss": 2.7227, + "step": 1520 + }, + { + "epoch": 0.06, + "grad_norm": 0.353515625, + "learning_rate": 0.00019954605153727842, + "loss": 2.8005, + "step": 1521 + }, + { + "epoch": 0.06, + "grad_norm": 0.400390625, + "learning_rate": 0.00019954541288305552, + "loss": 2.6983, + "step": 1522 + }, + { + "epoch": 0.06, + "grad_norm": 0.365234375, + "learning_rate": 0.00019954477378091465, + "loss": 2.7234, + "step": 1523 + }, + { + "epoch": 0.06, + "grad_norm": 0.35546875, + "learning_rate": 0.0001995441342308587, + "loss": 2.6892, + "step": 1524 + }, + { + "epoch": 0.06, + "grad_norm": 0.3984375, + "learning_rate": 0.00019954349423289048, + "loss": 2.6665, + "step": 1525 + }, + { + "epoch": 0.07, + "grad_norm": 0.4140625, + "learning_rate": 0.00019954285378701294, + "loss": 2.7153, + "step": 1526 + }, + { + "epoch": 0.07, + "grad_norm": 0.423828125, + "learning_rate": 0.00019954221289322893, + "loss": 2.7266, + "step": 1527 + }, + { + "epoch": 0.07, + "grad_norm": 0.390625, + "learning_rate": 0.00019954157155154133, + "loss": 2.7149, + "step": 1528 + }, + { + "epoch": 0.07, + "grad_norm": 0.490234375, + "learning_rate": 0.00019954092976195306, + "loss": 2.7499, + "step": 1529 + }, + { + "epoch": 0.07, + "grad_norm": 0.40234375, + "learning_rate": 0.00019954028752446697, + "loss": 3.035, + "step": 1530 + }, + { + "epoch": 0.07, + "grad_norm": 0.373046875, + "learning_rate": 0.00019953964483908598, + "loss": 2.7525, + "step": 1531 + }, + { + "epoch": 0.07, + "grad_norm": 0.4140625, + "learning_rate": 0.00019953900170581297, + "loss": 2.6496, + "step": 1532 + }, + { + "epoch": 0.07, + "grad_norm": 0.359375, + "learning_rate": 0.00019953835812465078, + "loss": 2.6439, + "step": 1533 + }, + { + "epoch": 0.07, + "grad_norm": 0.41796875, + "learning_rate": 0.0001995377140956024, + "loss": 2.5917, + "step": 1534 + }, + { + "epoch": 0.07, + "grad_norm": 0.37890625, + "learning_rate": 0.0001995370696186707, + "loss": 2.8055, + "step": 1535 + }, + { + "epoch": 0.07, + "grad_norm": 0.35546875, + "learning_rate": 0.00019953642469385853, + "loss": 2.6225, + "step": 1536 + }, + { + "epoch": 0.07, + "grad_norm": 0.431640625, + "learning_rate": 0.00019953577932116883, + "loss": 2.8089, + "step": 1537 + }, + { + "epoch": 0.07, + "grad_norm": 0.4140625, + "learning_rate": 0.00019953513350060448, + "loss": 2.6722, + "step": 1538 + }, + { + "epoch": 0.07, + "grad_norm": 0.373046875, + "learning_rate": 0.00019953448723216842, + "loss": 2.753, + "step": 1539 + }, + { + "epoch": 0.07, + "grad_norm": 0.435546875, + "learning_rate": 0.00019953384051586354, + "loss": 2.7707, + "step": 1540 + }, + { + "epoch": 0.07, + "grad_norm": 0.423828125, + "learning_rate": 0.00019953319335169279, + "loss": 2.6786, + "step": 1541 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.00019953254573965902, + "loss": 2.7, + "step": 1542 + }, + { + "epoch": 0.07, + "grad_norm": 0.37890625, + "learning_rate": 0.00019953189767976515, + "loss": 2.7289, + "step": 1543 + }, + { + "epoch": 0.07, + "grad_norm": 0.36328125, + "learning_rate": 0.00019953124917201413, + "loss": 2.7062, + "step": 1544 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.0001995306002164089, + "loss": 2.7809, + "step": 1545 + }, + { + "epoch": 0.07, + "grad_norm": 0.41796875, + "learning_rate": 0.00019952995081295226, + "loss": 2.6239, + "step": 1546 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.00019952930096164728, + "loss": 2.6631, + "step": 1547 + }, + { + "epoch": 0.07, + "grad_norm": 0.376953125, + "learning_rate": 0.0001995286506624968, + "loss": 2.4992, + "step": 1548 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.0001995279999155038, + "loss": 2.6941, + "step": 1549 + }, + { + "epoch": 0.07, + "grad_norm": 0.40234375, + "learning_rate": 0.00019952734872067112, + "loss": 2.649, + "step": 1550 + }, + { + "epoch": 0.07, + "grad_norm": 0.330078125, + "learning_rate": 0.00019952669707800177, + "loss": 2.5426, + "step": 1551 + }, + { + "epoch": 0.07, + "grad_norm": 0.416015625, + "learning_rate": 0.00019952604498749864, + "loss": 2.7204, + "step": 1552 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.0001995253924491647, + "loss": 2.7156, + "step": 1553 + }, + { + "epoch": 0.07, + "grad_norm": 0.37890625, + "learning_rate": 0.00019952473946300287, + "loss": 2.6008, + "step": 1554 + }, + { + "epoch": 0.07, + "grad_norm": 0.431640625, + "learning_rate": 0.00019952408602901607, + "loss": 2.8707, + "step": 1555 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.00019952343214720727, + "loss": 2.6483, + "step": 1556 + }, + { + "epoch": 0.07, + "grad_norm": 0.396484375, + "learning_rate": 0.00019952277781757935, + "loss": 2.6183, + "step": 1557 + }, + { + "epoch": 0.07, + "grad_norm": 0.37890625, + "learning_rate": 0.00019952212304013536, + "loss": 2.7063, + "step": 1558 + }, + { + "epoch": 0.07, + "grad_norm": 0.5078125, + "learning_rate": 0.00019952146781487815, + "loss": 2.7496, + "step": 1559 + }, + { + "epoch": 0.07, + "grad_norm": 0.40234375, + "learning_rate": 0.0001995208121418107, + "loss": 2.7886, + "step": 1560 + }, + { + "epoch": 0.07, + "grad_norm": 0.482421875, + "learning_rate": 0.000199520156020936, + "loss": 2.6906, + "step": 1561 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.00019951949945225698, + "loss": 2.6827, + "step": 1562 + }, + { + "epoch": 0.07, + "grad_norm": 0.439453125, + "learning_rate": 0.00019951884243577653, + "loss": 2.838, + "step": 1563 + }, + { + "epoch": 0.07, + "grad_norm": 0.4296875, + "learning_rate": 0.0001995181849714977, + "loss": 2.6659, + "step": 1564 + }, + { + "epoch": 0.07, + "grad_norm": 0.427734375, + "learning_rate": 0.0001995175270594234, + "loss": 2.8187, + "step": 1565 + }, + { + "epoch": 0.07, + "grad_norm": 0.416015625, + "learning_rate": 0.0001995168686995566, + "loss": 2.855, + "step": 1566 + }, + { + "epoch": 0.07, + "grad_norm": 0.38671875, + "learning_rate": 0.00019951620989190023, + "loss": 2.6887, + "step": 1567 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.0001995155506364573, + "loss": 2.8318, + "step": 1568 + }, + { + "epoch": 0.07, + "grad_norm": 0.40234375, + "learning_rate": 0.00019951489093323076, + "loss": 2.6905, + "step": 1569 + }, + { + "epoch": 0.07, + "grad_norm": 0.384765625, + "learning_rate": 0.0001995142307822236, + "loss": 2.733, + "step": 1570 + }, + { + "epoch": 0.07, + "grad_norm": 0.408203125, + "learning_rate": 0.00019951357018343875, + "loss": 2.6126, + "step": 1571 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.00019951290913687922, + "loss": 2.7457, + "step": 1572 + }, + { + "epoch": 0.07, + "grad_norm": 0.40234375, + "learning_rate": 0.00019951224764254796, + "loss": 2.7323, + "step": 1573 + }, + { + "epoch": 0.07, + "grad_norm": 0.431640625, + "learning_rate": 0.00019951158570044792, + "loss": 2.725, + "step": 1574 + }, + { + "epoch": 0.07, + "grad_norm": 0.4140625, + "learning_rate": 0.00019951092331058217, + "loss": 2.874, + "step": 1575 + }, + { + "epoch": 0.07, + "grad_norm": 0.40234375, + "learning_rate": 0.00019951026047295358, + "loss": 2.7576, + "step": 1576 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.00019950959718756522, + "loss": 2.8556, + "step": 1577 + }, + { + "epoch": 0.07, + "grad_norm": 0.3671875, + "learning_rate": 0.00019950893345442, + "loss": 2.7155, + "step": 1578 + }, + { + "epoch": 0.07, + "grad_norm": 0.408203125, + "learning_rate": 0.00019950826927352098, + "loss": 2.8298, + "step": 1579 + }, + { + "epoch": 0.07, + "grad_norm": 0.392578125, + "learning_rate": 0.00019950760464487112, + "loss": 2.7577, + "step": 1580 + }, + { + "epoch": 0.07, + "grad_norm": 0.361328125, + "learning_rate": 0.00019950693956847338, + "loss": 2.7318, + "step": 1581 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.0001995062740443308, + "loss": 2.7515, + "step": 1582 + }, + { + "epoch": 0.07, + "grad_norm": 0.423828125, + "learning_rate": 0.0001995056080724463, + "loss": 2.7818, + "step": 1583 + }, + { + "epoch": 0.07, + "grad_norm": 0.359375, + "learning_rate": 0.000199504941652823, + "loss": 2.7378, + "step": 1584 + }, + { + "epoch": 0.07, + "grad_norm": 0.392578125, + "learning_rate": 0.0001995042747854638, + "loss": 2.7702, + "step": 1585 + }, + { + "epoch": 0.07, + "grad_norm": 0.40625, + "learning_rate": 0.00019950360747037174, + "loss": 2.7826, + "step": 1586 + }, + { + "epoch": 0.07, + "grad_norm": 0.396484375, + "learning_rate": 0.00019950293970754976, + "loss": 2.7327, + "step": 1587 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.00019950227149700096, + "loss": 2.6379, + "step": 1588 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.0001995016028387283, + "loss": 2.9073, + "step": 1589 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.0001995009337327348, + "loss": 2.5948, + "step": 1590 + }, + { + "epoch": 0.07, + "grad_norm": 0.384765625, + "learning_rate": 0.00019950026417902344, + "loss": 2.7106, + "step": 1591 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.00019949959417759724, + "loss": 2.6708, + "step": 1592 + }, + { + "epoch": 0.07, + "grad_norm": 0.365234375, + "learning_rate": 0.00019949892372845928, + "loss": 2.6496, + "step": 1593 + }, + { + "epoch": 0.07, + "grad_norm": 0.419921875, + "learning_rate": 0.00019949825283161246, + "loss": 2.7347, + "step": 1594 + }, + { + "epoch": 0.07, + "grad_norm": 0.384765625, + "learning_rate": 0.00019949758148705992, + "loss": 2.7847, + "step": 1595 + }, + { + "epoch": 0.07, + "grad_norm": 0.375, + "learning_rate": 0.00019949690969480458, + "loss": 2.6442, + "step": 1596 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.00019949623745484955, + "loss": 2.7795, + "step": 1597 + }, + { + "epoch": 0.07, + "grad_norm": 0.40625, + "learning_rate": 0.0001994955647671978, + "loss": 2.8488, + "step": 1598 + }, + { + "epoch": 0.07, + "grad_norm": 0.408203125, + "learning_rate": 0.00019949489163185237, + "loss": 2.8831, + "step": 1599 + }, + { + "epoch": 0.07, + "grad_norm": 0.380859375, + "learning_rate": 0.00019949421804881626, + "loss": 2.7142, + "step": 1600 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.0001994935440180926, + "loss": 2.8931, + "step": 1601 + }, + { + "epoch": 0.07, + "grad_norm": 0.388671875, + "learning_rate": 0.00019949286953968427, + "loss": 2.6841, + "step": 1602 + }, + { + "epoch": 0.07, + "grad_norm": 0.33203125, + "learning_rate": 0.00019949219461359445, + "loss": 2.6254, + "step": 1603 + }, + { + "epoch": 0.07, + "grad_norm": 0.376953125, + "learning_rate": 0.00019949151923982607, + "loss": 2.8619, + "step": 1604 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.00019949084341838226, + "loss": 2.7369, + "step": 1605 + }, + { + "epoch": 0.07, + "grad_norm": 0.392578125, + "learning_rate": 0.00019949016714926597, + "loss": 2.7691, + "step": 1606 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.0001994894904324803, + "loss": 2.6433, + "step": 1607 + }, + { + "epoch": 0.07, + "grad_norm": 0.38671875, + "learning_rate": 0.00019948881326802828, + "loss": 2.6675, + "step": 1608 + }, + { + "epoch": 0.07, + "grad_norm": 0.373046875, + "learning_rate": 0.000199488135655913, + "loss": 2.6951, + "step": 1609 + }, + { + "epoch": 0.07, + "grad_norm": 0.388671875, + "learning_rate": 0.0001994874575961374, + "loss": 2.7467, + "step": 1610 + }, + { + "epoch": 0.07, + "grad_norm": 0.373046875, + "learning_rate": 0.00019948677908870465, + "loss": 2.7271, + "step": 1611 + }, + { + "epoch": 0.07, + "grad_norm": 0.369140625, + "learning_rate": 0.00019948610013361773, + "loss": 2.5922, + "step": 1612 + }, + { + "epoch": 0.07, + "grad_norm": 0.375, + "learning_rate": 0.00019948542073087975, + "loss": 2.5943, + "step": 1613 + }, + { + "epoch": 0.07, + "grad_norm": 0.40625, + "learning_rate": 0.0001994847408804937, + "loss": 2.6697, + "step": 1614 + }, + { + "epoch": 0.07, + "grad_norm": 0.357421875, + "learning_rate": 0.00019948406058246267, + "loss": 2.635, + "step": 1615 + }, + { + "epoch": 0.07, + "grad_norm": 0.390625, + "learning_rate": 0.00019948337983678976, + "loss": 2.5233, + "step": 1616 + }, + { + "epoch": 0.07, + "grad_norm": 0.36328125, + "learning_rate": 0.00019948269864347796, + "loss": 2.8317, + "step": 1617 + }, + { + "epoch": 0.07, + "grad_norm": 0.337890625, + "learning_rate": 0.0001994820170025304, + "loss": 2.7131, + "step": 1618 + }, + { + "epoch": 0.07, + "grad_norm": 0.365234375, + "learning_rate": 0.00019948133491395009, + "loss": 2.6211, + "step": 1619 + }, + { + "epoch": 0.07, + "grad_norm": 0.35546875, + "learning_rate": 0.00019948065237774013, + "loss": 2.623, + "step": 1620 + }, + { + "epoch": 0.07, + "grad_norm": 0.408203125, + "learning_rate": 0.0001994799693939036, + "loss": 2.7803, + "step": 1621 + }, + { + "epoch": 0.07, + "grad_norm": 0.36328125, + "learning_rate": 0.0001994792859624436, + "loss": 2.6976, + "step": 1622 + }, + { + "epoch": 0.07, + "grad_norm": 0.4140625, + "learning_rate": 0.00019947860208336312, + "loss": 2.5959, + "step": 1623 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.00019947791775666533, + "loss": 2.7109, + "step": 1624 + }, + { + "epoch": 0.07, + "grad_norm": 0.375, + "learning_rate": 0.00019947723298235323, + "loss": 2.6562, + "step": 1625 + }, + { + "epoch": 0.07, + "grad_norm": 0.388671875, + "learning_rate": 0.00019947654776042995, + "loss": 2.7432, + "step": 1626 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.00019947586209089854, + "loss": 2.5527, + "step": 1627 + }, + { + "epoch": 0.07, + "grad_norm": 0.373046875, + "learning_rate": 0.00019947517597376213, + "loss": 2.7104, + "step": 1628 + }, + { + "epoch": 0.07, + "grad_norm": 0.41015625, + "learning_rate": 0.00019947448940902377, + "loss": 2.7067, + "step": 1629 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.0001994738023966866, + "loss": 2.7143, + "step": 1630 + }, + { + "epoch": 0.07, + "grad_norm": 0.373046875, + "learning_rate": 0.00019947311493675362, + "loss": 2.6682, + "step": 1631 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.000199472427029228, + "loss": 2.6211, + "step": 1632 + }, + { + "epoch": 0.07, + "grad_norm": 0.5, + "learning_rate": 0.00019947173867411276, + "loss": 2.7231, + "step": 1633 + }, + { + "epoch": 0.07, + "grad_norm": 0.36328125, + "learning_rate": 0.00019947104987141112, + "loss": 2.8106, + "step": 1634 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.00019947036062112607, + "loss": 2.5911, + "step": 1635 + }, + { + "epoch": 0.07, + "grad_norm": 0.361328125, + "learning_rate": 0.00019946967092326077, + "loss": 2.6317, + "step": 1636 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.0001994689807778183, + "loss": 2.9378, + "step": 1637 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.00019946829018480176, + "loss": 2.7131, + "step": 1638 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.00019946759914421427, + "loss": 2.7063, + "step": 1639 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.0001994669076560589, + "loss": 2.7415, + "step": 1640 + }, + { + "epoch": 0.07, + "grad_norm": 0.337890625, + "learning_rate": 0.00019946621572033887, + "loss": 2.6733, + "step": 1641 + }, + { + "epoch": 0.07, + "grad_norm": 0.453125, + "learning_rate": 0.00019946552333705716, + "loss": 2.8119, + "step": 1642 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.00019946483050621697, + "loss": 2.7311, + "step": 1643 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.0001994641372278214, + "loss": 2.6169, + "step": 1644 + }, + { + "epoch": 0.07, + "grad_norm": 0.36328125, + "learning_rate": 0.00019946344350187353, + "loss": 2.7564, + "step": 1645 + }, + { + "epoch": 0.07, + "grad_norm": 0.435546875, + "learning_rate": 0.00019946274932837655, + "loss": 2.6551, + "step": 1646 + }, + { + "epoch": 0.07, + "grad_norm": 0.328125, + "learning_rate": 0.00019946205470733352, + "loss": 2.5953, + "step": 1647 + }, + { + "epoch": 0.07, + "grad_norm": 0.376953125, + "learning_rate": 0.0001994613596387476, + "loss": 2.7389, + "step": 1648 + }, + { + "epoch": 0.07, + "grad_norm": 0.408203125, + "learning_rate": 0.00019946066412262193, + "loss": 2.5992, + "step": 1649 + }, + { + "epoch": 0.07, + "grad_norm": 0.40234375, + "learning_rate": 0.0001994599681589596, + "loss": 2.745, + "step": 1650 + }, + { + "epoch": 0.07, + "grad_norm": 0.369140625, + "learning_rate": 0.00019945927174776378, + "loss": 2.7746, + "step": 1651 + }, + { + "epoch": 0.07, + "grad_norm": 0.388671875, + "learning_rate": 0.00019945857488903757, + "loss": 2.6808, + "step": 1652 + }, + { + "epoch": 0.07, + "grad_norm": 0.38671875, + "learning_rate": 0.0001994578775827841, + "loss": 2.5152, + "step": 1653 + }, + { + "epoch": 0.07, + "grad_norm": 0.419921875, + "learning_rate": 0.00019945717982900658, + "loss": 2.7769, + "step": 1654 + }, + { + "epoch": 0.07, + "grad_norm": 0.38671875, + "learning_rate": 0.00019945648162770808, + "loss": 2.6639, + "step": 1655 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.0001994557829788917, + "loss": 2.6401, + "step": 1656 + }, + { + "epoch": 0.07, + "grad_norm": 0.357421875, + "learning_rate": 0.00019945508388256072, + "loss": 2.8244, + "step": 1657 + }, + { + "epoch": 0.07, + "grad_norm": 0.435546875, + "learning_rate": 0.00019945438433871816, + "loss": 2.797, + "step": 1658 + }, + { + "epoch": 0.07, + "grad_norm": 0.361328125, + "learning_rate": 0.00019945368434736726, + "loss": 2.6518, + "step": 1659 + }, + { + "epoch": 0.07, + "grad_norm": 0.408203125, + "learning_rate": 0.00019945298390851112, + "loss": 2.6698, + "step": 1660 + }, + { + "epoch": 0.07, + "grad_norm": 0.455078125, + "learning_rate": 0.0001994522830221529, + "loss": 2.8191, + "step": 1661 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.00019945158168829571, + "loss": 2.5325, + "step": 1662 + }, + { + "epoch": 0.07, + "grad_norm": 0.357421875, + "learning_rate": 0.00019945087990694277, + "loss": 2.5962, + "step": 1663 + }, + { + "epoch": 0.07, + "grad_norm": 0.41796875, + "learning_rate": 0.00019945017767809722, + "loss": 2.7323, + "step": 1664 + }, + { + "epoch": 0.07, + "grad_norm": 0.365234375, + "learning_rate": 0.00019944947500176224, + "loss": 2.5737, + "step": 1665 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.00019944877187794092, + "loss": 2.8904, + "step": 1666 + }, + { + "epoch": 0.07, + "grad_norm": 0.44921875, + "learning_rate": 0.0001994480683066365, + "loss": 2.6667, + "step": 1667 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.0001994473642878521, + "loss": 2.8499, + "step": 1668 + }, + { + "epoch": 0.07, + "grad_norm": 0.419921875, + "learning_rate": 0.0001994466598215909, + "loss": 2.7588, + "step": 1669 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.0001994459549078561, + "loss": 2.6843, + "step": 1670 + }, + { + "epoch": 0.07, + "grad_norm": 0.423828125, + "learning_rate": 0.00019944524954665084, + "loss": 2.6271, + "step": 1671 + }, + { + "epoch": 0.07, + "grad_norm": 0.416015625, + "learning_rate": 0.0001994445437379783, + "loss": 2.6763, + "step": 1672 + }, + { + "epoch": 0.07, + "grad_norm": 0.39453125, + "learning_rate": 0.00019944383748184166, + "loss": 2.7659, + "step": 1673 + }, + { + "epoch": 0.07, + "grad_norm": 0.34765625, + "learning_rate": 0.00019944313077824408, + "loss": 2.7816, + "step": 1674 + }, + { + "epoch": 0.07, + "grad_norm": 0.37109375, + "learning_rate": 0.00019944242362718875, + "loss": 2.6842, + "step": 1675 + }, + { + "epoch": 0.07, + "grad_norm": 0.42578125, + "learning_rate": 0.00019944171602867888, + "loss": 2.7483, + "step": 1676 + }, + { + "epoch": 0.07, + "grad_norm": 0.390625, + "learning_rate": 0.0001994410079827176, + "loss": 2.7386, + "step": 1677 + }, + { + "epoch": 0.07, + "grad_norm": 0.421875, + "learning_rate": 0.00019944029948930814, + "loss": 2.7366, + "step": 1678 + }, + { + "epoch": 0.07, + "grad_norm": 0.396484375, + "learning_rate": 0.00019943959054845366, + "loss": 2.7069, + "step": 1679 + }, + { + "epoch": 0.07, + "grad_norm": 0.37890625, + "learning_rate": 0.0001994388811601574, + "loss": 2.5655, + "step": 1680 + }, + { + "epoch": 0.07, + "grad_norm": 0.451171875, + "learning_rate": 0.00019943817132442247, + "loss": 2.794, + "step": 1681 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.00019943746104125217, + "loss": 2.7044, + "step": 1682 + }, + { + "epoch": 0.07, + "grad_norm": 0.447265625, + "learning_rate": 0.00019943675031064957, + "loss": 2.8494, + "step": 1683 + }, + { + "epoch": 0.07, + "grad_norm": 0.435546875, + "learning_rate": 0.00019943603913261798, + "loss": 2.9019, + "step": 1684 + }, + { + "epoch": 0.07, + "grad_norm": 0.431640625, + "learning_rate": 0.00019943532750716053, + "loss": 2.7235, + "step": 1685 + }, + { + "epoch": 0.07, + "grad_norm": 0.45703125, + "learning_rate": 0.00019943461543428046, + "loss": 2.7638, + "step": 1686 + }, + { + "epoch": 0.07, + "grad_norm": 0.490234375, + "learning_rate": 0.00019943390291398094, + "loss": 2.566, + "step": 1687 + }, + { + "epoch": 0.07, + "grad_norm": 0.451171875, + "learning_rate": 0.0001994331899462652, + "loss": 2.7093, + "step": 1688 + }, + { + "epoch": 0.07, + "grad_norm": 0.439453125, + "learning_rate": 0.00019943247653113645, + "loss": 2.7071, + "step": 1689 + }, + { + "epoch": 0.07, + "grad_norm": 0.388671875, + "learning_rate": 0.0001994317626685979, + "loss": 2.7334, + "step": 1690 + }, + { + "epoch": 0.07, + "grad_norm": 0.458984375, + "learning_rate": 0.00019943104835865274, + "loss": 2.8478, + "step": 1691 + }, + { + "epoch": 0.07, + "grad_norm": 0.41015625, + "learning_rate": 0.00019943033360130422, + "loss": 2.7603, + "step": 1692 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.00019942961839655552, + "loss": 2.5852, + "step": 1693 + }, + { + "epoch": 0.07, + "grad_norm": 0.42578125, + "learning_rate": 0.00019942890274440988, + "loss": 2.7425, + "step": 1694 + }, + { + "epoch": 0.07, + "grad_norm": 0.40234375, + "learning_rate": 0.00019942818664487054, + "loss": 2.8415, + "step": 1695 + }, + { + "epoch": 0.07, + "grad_norm": 0.357421875, + "learning_rate": 0.00019942747009794065, + "loss": 2.6079, + "step": 1696 + }, + { + "epoch": 0.07, + "grad_norm": 0.37890625, + "learning_rate": 0.00019942675310362354, + "loss": 2.7936, + "step": 1697 + }, + { + "epoch": 0.07, + "grad_norm": 0.42578125, + "learning_rate": 0.00019942603566192233, + "loss": 2.7469, + "step": 1698 + }, + { + "epoch": 0.07, + "grad_norm": 0.392578125, + "learning_rate": 0.0001994253177728403, + "loss": 2.6583, + "step": 1699 + }, + { + "epoch": 0.07, + "grad_norm": 0.392578125, + "learning_rate": 0.00019942459943638068, + "loss": 2.6245, + "step": 1700 + }, + { + "epoch": 0.07, + "grad_norm": 0.375, + "learning_rate": 0.00019942388065254671, + "loss": 2.5826, + "step": 1701 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.00019942316142134158, + "loss": 2.5786, + "step": 1702 + }, + { + "epoch": 0.07, + "grad_norm": 0.384765625, + "learning_rate": 0.00019942244174276857, + "loss": 2.6131, + "step": 1703 + }, + { + "epoch": 0.07, + "grad_norm": 0.390625, + "learning_rate": 0.00019942172161683096, + "loss": 2.5075, + "step": 1704 + }, + { + "epoch": 0.07, + "grad_norm": 0.392578125, + "learning_rate": 0.00019942100104353188, + "loss": 2.7218, + "step": 1705 + }, + { + "epoch": 0.07, + "grad_norm": 0.373046875, + "learning_rate": 0.00019942028002287464, + "loss": 2.8233, + "step": 1706 + }, + { + "epoch": 0.07, + "grad_norm": 0.42578125, + "learning_rate": 0.0001994195585548625, + "loss": 2.7734, + "step": 1707 + }, + { + "epoch": 0.07, + "grad_norm": 0.359375, + "learning_rate": 0.00019941883663949865, + "loss": 2.5882, + "step": 1708 + }, + { + "epoch": 0.07, + "grad_norm": 0.353515625, + "learning_rate": 0.0001994181142767864, + "loss": 2.5934, + "step": 1709 + }, + { + "epoch": 0.07, + "grad_norm": 0.439453125, + "learning_rate": 0.00019941739146672892, + "loss": 2.6977, + "step": 1710 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.00019941666820932956, + "loss": 2.7801, + "step": 1711 + }, + { + "epoch": 0.07, + "grad_norm": 0.431640625, + "learning_rate": 0.0001994159445045915, + "loss": 2.6993, + "step": 1712 + }, + { + "epoch": 0.07, + "grad_norm": 0.431640625, + "learning_rate": 0.000199415220352518, + "loss": 2.781, + "step": 1713 + }, + { + "epoch": 0.07, + "grad_norm": 0.431640625, + "learning_rate": 0.00019941449575311236, + "loss": 2.8069, + "step": 1714 + }, + { + "epoch": 0.07, + "grad_norm": 0.392578125, + "learning_rate": 0.00019941377070637785, + "loss": 2.7439, + "step": 1715 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.0001994130452123177, + "loss": 2.7347, + "step": 1716 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.00019941231927093512, + "loss": 2.7719, + "step": 1717 + }, + { + "epoch": 0.07, + "grad_norm": 0.416015625, + "learning_rate": 0.0001994115928822335, + "loss": 2.7706, + "step": 1718 + }, + { + "epoch": 0.07, + "grad_norm": 0.365234375, + "learning_rate": 0.00019941086604621601, + "loss": 2.6648, + "step": 1719 + }, + { + "epoch": 0.07, + "grad_norm": 0.41015625, + "learning_rate": 0.00019941013876288596, + "loss": 2.8035, + "step": 1720 + }, + { + "epoch": 0.07, + "grad_norm": 0.4140625, + "learning_rate": 0.00019940941103224662, + "loss": 2.662, + "step": 1721 + }, + { + "epoch": 0.07, + "grad_norm": 0.37109375, + "learning_rate": 0.00019940868285430125, + "loss": 2.7209, + "step": 1722 + }, + { + "epoch": 0.07, + "grad_norm": 0.396484375, + "learning_rate": 0.00019940795422905314, + "loss": 2.6147, + "step": 1723 + }, + { + "epoch": 0.07, + "grad_norm": 0.392578125, + "learning_rate": 0.00019940722515650555, + "loss": 2.7321, + "step": 1724 + }, + { + "epoch": 0.07, + "grad_norm": 0.375, + "learning_rate": 0.00019940649563666182, + "loss": 2.7636, + "step": 1725 + }, + { + "epoch": 0.07, + "grad_norm": 0.404296875, + "learning_rate": 0.00019940576566952517, + "loss": 2.7237, + "step": 1726 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.0001994050352550989, + "loss": 2.6364, + "step": 1727 + }, + { + "epoch": 0.07, + "grad_norm": 0.376953125, + "learning_rate": 0.0001994043043933863, + "loss": 2.6989, + "step": 1728 + }, + { + "epoch": 0.07, + "grad_norm": 0.396484375, + "learning_rate": 0.00019940357308439064, + "loss": 2.7663, + "step": 1729 + }, + { + "epoch": 0.07, + "grad_norm": 0.3515625, + "learning_rate": 0.00019940284132811525, + "loss": 2.8448, + "step": 1730 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.0001994021091245634, + "loss": 2.7255, + "step": 1731 + }, + { + "epoch": 0.07, + "grad_norm": 0.43359375, + "learning_rate": 0.00019940137647373834, + "loss": 2.6661, + "step": 1732 + }, + { + "epoch": 0.07, + "grad_norm": 0.388671875, + "learning_rate": 0.00019940064337564347, + "loss": 2.8497, + "step": 1733 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.00019939990983028203, + "loss": 2.7327, + "step": 1734 + }, + { + "epoch": 0.07, + "grad_norm": 0.41015625, + "learning_rate": 0.0001993991758376573, + "loss": 2.9028, + "step": 1735 + }, + { + "epoch": 0.07, + "grad_norm": 0.390625, + "learning_rate": 0.00019939844139777263, + "loss": 2.6259, + "step": 1736 + }, + { + "epoch": 0.07, + "grad_norm": 0.44921875, + "learning_rate": 0.00019939770651063128, + "loss": 2.6474, + "step": 1737 + }, + { + "epoch": 0.07, + "grad_norm": 0.44140625, + "learning_rate": 0.00019939697117623656, + "loss": 2.8042, + "step": 1738 + }, + { + "epoch": 0.07, + "grad_norm": 0.376953125, + "learning_rate": 0.00019939623539459183, + "loss": 2.753, + "step": 1739 + }, + { + "epoch": 0.07, + "grad_norm": 0.400390625, + "learning_rate": 0.00019939549916570034, + "loss": 2.6953, + "step": 1740 + }, + { + "epoch": 0.07, + "grad_norm": 0.396484375, + "learning_rate": 0.00019939476248956545, + "loss": 2.7573, + "step": 1741 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.00019939402536619044, + "loss": 2.7115, + "step": 1742 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.00019939328779557865, + "loss": 2.7348, + "step": 1743 + }, + { + "epoch": 0.07, + "grad_norm": 0.42578125, + "learning_rate": 0.00019939254977773335, + "loss": 2.8131, + "step": 1744 + }, + { + "epoch": 0.07, + "grad_norm": 0.3984375, + "learning_rate": 0.00019939181131265797, + "loss": 2.8404, + "step": 1745 + }, + { + "epoch": 0.07, + "grad_norm": 0.416015625, + "learning_rate": 0.0001993910724003557, + "loss": 2.7754, + "step": 1746 + }, + { + "epoch": 0.07, + "grad_norm": 0.37890625, + "learning_rate": 0.00019939033304082994, + "loss": 2.7011, + "step": 1747 + }, + { + "epoch": 0.07, + "grad_norm": 0.412109375, + "learning_rate": 0.00019938959323408405, + "loss": 2.708, + "step": 1748 + }, + { + "epoch": 0.07, + "grad_norm": 0.466796875, + "learning_rate": 0.00019938885298012124, + "loss": 2.8183, + "step": 1749 + }, + { + "epoch": 0.07, + "grad_norm": 0.3828125, + "learning_rate": 0.00019938811227894498, + "loss": 2.571, + "step": 1750 + }, + { + "epoch": 0.07, + "grad_norm": 0.478515625, + "learning_rate": 0.0001993873711305585, + "loss": 2.7053, + "step": 1751 + }, + { + "epoch": 0.07, + "grad_norm": 0.431640625, + "learning_rate": 0.0001993866295349652, + "loss": 2.6686, + "step": 1752 + }, + { + "epoch": 0.07, + "grad_norm": 0.384765625, + "learning_rate": 0.00019938588749216835, + "loss": 2.6261, + "step": 1753 + }, + { + "epoch": 0.07, + "grad_norm": 0.53515625, + "learning_rate": 0.00019938514500217135, + "loss": 2.7209, + "step": 1754 + }, + { + "epoch": 0.07, + "grad_norm": 0.41015625, + "learning_rate": 0.00019938440206497752, + "loss": 2.6887, + "step": 1755 + }, + { + "epoch": 0.07, + "grad_norm": 0.408203125, + "learning_rate": 0.00019938365868059019, + "loss": 2.7166, + "step": 1756 + }, + { + "epoch": 0.07, + "grad_norm": 0.4375, + "learning_rate": 0.00019938291484901272, + "loss": 2.8062, + "step": 1757 + }, + { + "epoch": 0.07, + "grad_norm": 0.427734375, + "learning_rate": 0.00019938217057024847, + "loss": 2.7669, + "step": 1758 + }, + { + "epoch": 0.07, + "grad_norm": 0.4453125, + "learning_rate": 0.00019938142584430076, + "loss": 2.6373, + "step": 1759 + }, + { + "epoch": 0.07, + "grad_norm": 0.42578125, + "learning_rate": 0.00019938068067117294, + "loss": 2.5893, + "step": 1760 + }, + { + "epoch": 0.08, + "grad_norm": 0.455078125, + "learning_rate": 0.0001993799350508684, + "loss": 2.6635, + "step": 1761 + }, + { + "epoch": 0.08, + "grad_norm": 0.408203125, + "learning_rate": 0.00019937918898339048, + "loss": 2.7126, + "step": 1762 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.0001993784424687425, + "loss": 2.9276, + "step": 1763 + }, + { + "epoch": 0.08, + "grad_norm": 0.421875, + "learning_rate": 0.00019937769550692783, + "loss": 2.6714, + "step": 1764 + }, + { + "epoch": 0.08, + "grad_norm": 0.427734375, + "learning_rate": 0.0001993769480979499, + "loss": 2.6648, + "step": 1765 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.00019937620024181198, + "loss": 2.6664, + "step": 1766 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.0001993754519385175, + "loss": 2.6861, + "step": 1767 + }, + { + "epoch": 0.08, + "grad_norm": 0.421875, + "learning_rate": 0.0001993747031880698, + "loss": 2.7579, + "step": 1768 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.00019937395399047226, + "loss": 2.705, + "step": 1769 + }, + { + "epoch": 0.08, + "grad_norm": 0.357421875, + "learning_rate": 0.00019937320434572825, + "loss": 2.7641, + "step": 1770 + }, + { + "epoch": 0.08, + "grad_norm": 0.466796875, + "learning_rate": 0.0001993724542538411, + "loss": 2.7099, + "step": 1771 + }, + { + "epoch": 0.08, + "grad_norm": 0.3828125, + "learning_rate": 0.00019937170371481426, + "loss": 2.6996, + "step": 1772 + }, + { + "epoch": 0.08, + "grad_norm": 0.41796875, + "learning_rate": 0.00019937095272865105, + "loss": 2.7319, + "step": 1773 + }, + { + "epoch": 0.08, + "grad_norm": 0.486328125, + "learning_rate": 0.00019937020129535488, + "loss": 2.7596, + "step": 1774 + }, + { + "epoch": 0.08, + "grad_norm": 0.439453125, + "learning_rate": 0.0001993694494149291, + "loss": 2.7483, + "step": 1775 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.00019936869708737714, + "loss": 2.7209, + "step": 1776 + }, + { + "epoch": 0.08, + "grad_norm": 0.42578125, + "learning_rate": 0.00019936794431270232, + "loss": 2.8446, + "step": 1777 + }, + { + "epoch": 0.08, + "grad_norm": 0.427734375, + "learning_rate": 0.00019936719109090812, + "loss": 2.5644, + "step": 1778 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.00019936643742199782, + "loss": 2.6476, + "step": 1779 + }, + { + "epoch": 0.08, + "grad_norm": 0.41796875, + "learning_rate": 0.00019936568330597489, + "loss": 2.6694, + "step": 1780 + }, + { + "epoch": 0.08, + "grad_norm": 0.3984375, + "learning_rate": 0.00019936492874284272, + "loss": 2.7983, + "step": 1781 + }, + { + "epoch": 0.08, + "grad_norm": 0.376953125, + "learning_rate": 0.00019936417373260466, + "loss": 2.7202, + "step": 1782 + }, + { + "epoch": 0.08, + "grad_norm": 0.404296875, + "learning_rate": 0.00019936341827526412, + "loss": 2.8498, + "step": 1783 + }, + { + "epoch": 0.08, + "grad_norm": 0.38671875, + "learning_rate": 0.00019936266237082455, + "loss": 2.6468, + "step": 1784 + }, + { + "epoch": 0.08, + "grad_norm": 0.3984375, + "learning_rate": 0.00019936190601928927, + "loss": 2.7668, + "step": 1785 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.00019936114922066176, + "loss": 2.7256, + "step": 1786 + }, + { + "epoch": 0.08, + "grad_norm": 0.3828125, + "learning_rate": 0.00019936039197494538, + "loss": 2.7364, + "step": 1787 + }, + { + "epoch": 0.08, + "grad_norm": 0.375, + "learning_rate": 0.00019935963428214352, + "loss": 2.6842, + "step": 1788 + }, + { + "epoch": 0.08, + "grad_norm": 0.392578125, + "learning_rate": 0.00019935887614225967, + "loss": 2.7527, + "step": 1789 + }, + { + "epoch": 0.08, + "grad_norm": 0.353515625, + "learning_rate": 0.00019935811755529713, + "loss": 2.7118, + "step": 1790 + }, + { + "epoch": 0.08, + "grad_norm": 0.359375, + "learning_rate": 0.00019935735852125942, + "loss": 2.7299, + "step": 1791 + }, + { + "epoch": 0.08, + "grad_norm": 0.400390625, + "learning_rate": 0.0001993565990401499, + "loss": 2.6309, + "step": 1792 + }, + { + "epoch": 0.08, + "grad_norm": 0.427734375, + "learning_rate": 0.00019935583911197195, + "loss": 2.6176, + "step": 1793 + }, + { + "epoch": 0.08, + "grad_norm": 0.443359375, + "learning_rate": 0.0001993550787367291, + "loss": 2.7154, + "step": 1794 + }, + { + "epoch": 0.08, + "grad_norm": 0.38671875, + "learning_rate": 0.00019935431791442466, + "loss": 2.7259, + "step": 1795 + }, + { + "epoch": 0.08, + "grad_norm": 0.3828125, + "learning_rate": 0.00019935355664506212, + "loss": 2.6767, + "step": 1796 + }, + { + "epoch": 0.08, + "grad_norm": 0.419921875, + "learning_rate": 0.00019935279492864488, + "loss": 2.8334, + "step": 1797 + }, + { + "epoch": 0.08, + "grad_norm": 0.408203125, + "learning_rate": 0.00019935203276517637, + "loss": 2.7022, + "step": 1798 + }, + { + "epoch": 0.08, + "grad_norm": 0.43359375, + "learning_rate": 0.00019935127015466002, + "loss": 2.7695, + "step": 1799 + }, + { + "epoch": 0.08, + "grad_norm": 0.40234375, + "learning_rate": 0.00019935050709709925, + "loss": 2.6263, + "step": 1800 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.00019934974359249753, + "loss": 2.7151, + "step": 1801 + }, + { + "epoch": 0.08, + "grad_norm": 0.384765625, + "learning_rate": 0.00019934897964085825, + "loss": 2.5945, + "step": 1802 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.0001993482152421849, + "loss": 2.5886, + "step": 1803 + }, + { + "epoch": 0.08, + "grad_norm": 0.41796875, + "learning_rate": 0.00019934745039648085, + "loss": 2.7798, + "step": 1804 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.00019934668510374962, + "loss": 2.6509, + "step": 1805 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.0001993459193639946, + "loss": 2.7171, + "step": 1806 + }, + { + "epoch": 0.08, + "grad_norm": 0.400390625, + "learning_rate": 0.00019934515317721925, + "loss": 2.7615, + "step": 1807 + }, + { + "epoch": 0.08, + "grad_norm": 0.3984375, + "learning_rate": 0.00019934438654342702, + "loss": 2.7678, + "step": 1808 + }, + { + "epoch": 0.08, + "grad_norm": 0.376953125, + "learning_rate": 0.00019934361946262135, + "loss": 2.676, + "step": 1809 + }, + { + "epoch": 0.08, + "grad_norm": 0.39453125, + "learning_rate": 0.0001993428519348057, + "loss": 2.6934, + "step": 1810 + }, + { + "epoch": 0.08, + "grad_norm": 0.44921875, + "learning_rate": 0.0001993420839599835, + "loss": 2.6533, + "step": 1811 + }, + { + "epoch": 0.08, + "grad_norm": 0.38671875, + "learning_rate": 0.00019934131553815828, + "loss": 2.6111, + "step": 1812 + }, + { + "epoch": 0.08, + "grad_norm": 0.443359375, + "learning_rate": 0.0001993405466693334, + "loss": 2.746, + "step": 1813 + }, + { + "epoch": 0.08, + "grad_norm": 0.443359375, + "learning_rate": 0.00019933977735351238, + "loss": 2.7825, + "step": 1814 + }, + { + "epoch": 0.08, + "grad_norm": 0.373046875, + "learning_rate": 0.00019933900759069867, + "loss": 2.6702, + "step": 1815 + }, + { + "epoch": 0.08, + "grad_norm": 0.455078125, + "learning_rate": 0.00019933823738089572, + "loss": 2.7262, + "step": 1816 + }, + { + "epoch": 0.08, + "grad_norm": 0.3515625, + "learning_rate": 0.000199337466724107, + "loss": 2.4778, + "step": 1817 + }, + { + "epoch": 0.08, + "grad_norm": 0.47265625, + "learning_rate": 0.000199336695620336, + "loss": 2.7601, + "step": 1818 + }, + { + "epoch": 0.08, + "grad_norm": 0.404296875, + "learning_rate": 0.00019933592406958618, + "loss": 2.8568, + "step": 1819 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.00019933515207186096, + "loss": 2.7449, + "step": 1820 + }, + { + "epoch": 0.08, + "grad_norm": 0.419921875, + "learning_rate": 0.00019933437962716392, + "loss": 2.7135, + "step": 1821 + }, + { + "epoch": 0.08, + "grad_norm": 0.412109375, + "learning_rate": 0.00019933360673549846, + "loss": 2.6458, + "step": 1822 + }, + { + "epoch": 0.08, + "grad_norm": 0.41015625, + "learning_rate": 0.00019933283339686805, + "loss": 2.6106, + "step": 1823 + }, + { + "epoch": 0.08, + "grad_norm": 0.43359375, + "learning_rate": 0.00019933205961127618, + "loss": 2.5585, + "step": 1824 + }, + { + "epoch": 0.08, + "grad_norm": 0.427734375, + "learning_rate": 0.00019933128537872637, + "loss": 2.7571, + "step": 1825 + }, + { + "epoch": 0.08, + "grad_norm": 0.408203125, + "learning_rate": 0.0001993305106992221, + "loss": 2.6724, + "step": 1826 + }, + { + "epoch": 0.08, + "grad_norm": 0.396484375, + "learning_rate": 0.00019932973557276677, + "loss": 2.984, + "step": 1827 + }, + { + "epoch": 0.08, + "grad_norm": 0.41015625, + "learning_rate": 0.00019932895999936397, + "loss": 2.839, + "step": 1828 + }, + { + "epoch": 0.08, + "grad_norm": 0.470703125, + "learning_rate": 0.00019932818397901713, + "loss": 2.6974, + "step": 1829 + }, + { + "epoch": 0.08, + "grad_norm": 0.435546875, + "learning_rate": 0.00019932740751172978, + "loss": 2.7892, + "step": 1830 + }, + { + "epoch": 0.08, + "grad_norm": 0.46875, + "learning_rate": 0.00019932663059750542, + "loss": 2.7987, + "step": 1831 + }, + { + "epoch": 0.08, + "grad_norm": 0.384765625, + "learning_rate": 0.00019932585323634747, + "loss": 2.7963, + "step": 1832 + }, + { + "epoch": 0.08, + "grad_norm": 0.42578125, + "learning_rate": 0.0001993250754282595, + "loss": 2.7758, + "step": 1833 + }, + { + "epoch": 0.08, + "grad_norm": 0.416015625, + "learning_rate": 0.00019932429717324501, + "loss": 2.7236, + "step": 1834 + }, + { + "epoch": 0.08, + "grad_norm": 0.455078125, + "learning_rate": 0.00019932351847130745, + "loss": 2.765, + "step": 1835 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.0001993227393224504, + "loss": 2.688, + "step": 1836 + }, + { + "epoch": 0.08, + "grad_norm": 0.392578125, + "learning_rate": 0.00019932195972667728, + "loss": 2.6736, + "step": 1837 + }, + { + "epoch": 0.08, + "grad_norm": 0.3984375, + "learning_rate": 0.00019932117968399164, + "loss": 2.5537, + "step": 1838 + }, + { + "epoch": 0.08, + "grad_norm": 0.41015625, + "learning_rate": 0.000199320399194397, + "loss": 2.7277, + "step": 1839 + }, + { + "epoch": 0.08, + "grad_norm": 0.404296875, + "learning_rate": 0.00019931961825789686, + "loss": 2.5632, + "step": 1840 + }, + { + "epoch": 0.08, + "grad_norm": 0.44140625, + "learning_rate": 0.00019931883687449476, + "loss": 2.8835, + "step": 1841 + }, + { + "epoch": 0.08, + "grad_norm": 0.390625, + "learning_rate": 0.00019931805504419417, + "loss": 2.7082, + "step": 1842 + }, + { + "epoch": 0.08, + "grad_norm": 0.38671875, + "learning_rate": 0.00019931727276699862, + "loss": 2.6581, + "step": 1843 + }, + { + "epoch": 0.08, + "grad_norm": 0.42578125, + "learning_rate": 0.00019931649004291163, + "loss": 2.7384, + "step": 1844 + }, + { + "epoch": 0.08, + "grad_norm": 0.37109375, + "learning_rate": 0.00019931570687193676, + "loss": 2.4858, + "step": 1845 + }, + { + "epoch": 0.08, + "grad_norm": 0.3828125, + "learning_rate": 0.0001993149232540775, + "loss": 2.8182, + "step": 1846 + }, + { + "epoch": 0.08, + "grad_norm": 0.4453125, + "learning_rate": 0.00019931413918933735, + "loss": 2.7531, + "step": 1847 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.0001993133546777199, + "loss": 2.6825, + "step": 1848 + }, + { + "epoch": 0.08, + "grad_norm": 0.447265625, + "learning_rate": 0.00019931256971922863, + "loss": 2.8556, + "step": 1849 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.0001993117843138671, + "loss": 2.7795, + "step": 1850 + }, + { + "epoch": 0.08, + "grad_norm": 0.3828125, + "learning_rate": 0.00019931099846163885, + "loss": 2.6545, + "step": 1851 + }, + { + "epoch": 0.08, + "grad_norm": 0.412109375, + "learning_rate": 0.00019931021216254736, + "loss": 2.8389, + "step": 1852 + }, + { + "epoch": 0.08, + "grad_norm": 0.43359375, + "learning_rate": 0.00019930942541659626, + "loss": 2.7473, + "step": 1853 + }, + { + "epoch": 0.08, + "grad_norm": 0.447265625, + "learning_rate": 0.00019930863822378901, + "loss": 2.7911, + "step": 1854 + }, + { + "epoch": 0.08, + "grad_norm": 0.3515625, + "learning_rate": 0.00019930785058412917, + "loss": 2.6233, + "step": 1855 + }, + { + "epoch": 0.08, + "grad_norm": 0.392578125, + "learning_rate": 0.00019930706249762033, + "loss": 2.7159, + "step": 1856 + }, + { + "epoch": 0.08, + "grad_norm": 0.37890625, + "learning_rate": 0.00019930627396426597, + "loss": 2.5892, + "step": 1857 + }, + { + "epoch": 0.08, + "grad_norm": 0.453125, + "learning_rate": 0.00019930548498406969, + "loss": 2.8262, + "step": 1858 + }, + { + "epoch": 0.08, + "grad_norm": 0.404296875, + "learning_rate": 0.000199304695557035, + "loss": 2.6817, + "step": 1859 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.0001993039056831655, + "loss": 2.6654, + "step": 1860 + }, + { + "epoch": 0.08, + "grad_norm": 0.380859375, + "learning_rate": 0.0001993031153624647, + "loss": 2.6533, + "step": 1861 + }, + { + "epoch": 0.08, + "grad_norm": 0.4609375, + "learning_rate": 0.00019930232459493616, + "loss": 2.7617, + "step": 1862 + }, + { + "epoch": 0.08, + "grad_norm": 0.44140625, + "learning_rate": 0.00019930153338058347, + "loss": 2.7998, + "step": 1863 + }, + { + "epoch": 0.08, + "grad_norm": 0.375, + "learning_rate": 0.00019930074171941017, + "loss": 2.5981, + "step": 1864 + }, + { + "epoch": 0.08, + "grad_norm": 0.396484375, + "learning_rate": 0.0001992999496114198, + "loss": 2.7307, + "step": 1865 + }, + { + "epoch": 0.08, + "grad_norm": 0.400390625, + "learning_rate": 0.00019929915705661598, + "loss": 2.7535, + "step": 1866 + }, + { + "epoch": 0.08, + "grad_norm": 0.408203125, + "learning_rate": 0.00019929836405500222, + "loss": 2.8018, + "step": 1867 + }, + { + "epoch": 0.08, + "grad_norm": 0.353515625, + "learning_rate": 0.00019929757060658208, + "loss": 2.7979, + "step": 1868 + }, + { + "epoch": 0.08, + "grad_norm": 0.4453125, + "learning_rate": 0.0001992967767113592, + "loss": 2.6929, + "step": 1869 + }, + { + "epoch": 0.08, + "grad_norm": 0.392578125, + "learning_rate": 0.00019929598236933713, + "loss": 2.7747, + "step": 1870 + }, + { + "epoch": 0.08, + "grad_norm": 0.373046875, + "learning_rate": 0.0001992951875805194, + "loss": 2.7356, + "step": 1871 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.00019929439234490963, + "loss": 2.7242, + "step": 1872 + }, + { + "epoch": 0.08, + "grad_norm": 0.390625, + "learning_rate": 0.00019929359666251138, + "loss": 2.6513, + "step": 1873 + }, + { + "epoch": 0.08, + "grad_norm": 0.4296875, + "learning_rate": 0.00019929280053332821, + "loss": 2.5182, + "step": 1874 + }, + { + "epoch": 0.08, + "grad_norm": 0.396484375, + "learning_rate": 0.00019929200395736376, + "loss": 2.7905, + "step": 1875 + }, + { + "epoch": 0.08, + "grad_norm": 0.42578125, + "learning_rate": 0.00019929120693462155, + "loss": 2.7224, + "step": 1876 + }, + { + "epoch": 0.08, + "grad_norm": 0.443359375, + "learning_rate": 0.0001992904094651052, + "loss": 2.6895, + "step": 1877 + }, + { + "epoch": 0.08, + "grad_norm": 0.3515625, + "learning_rate": 0.00019928961154881832, + "loss": 2.6624, + "step": 1878 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.00019928881318576443, + "loss": 2.7043, + "step": 1879 + }, + { + "epoch": 0.08, + "grad_norm": 0.41015625, + "learning_rate": 0.0001992880143759472, + "loss": 2.6704, + "step": 1880 + }, + { + "epoch": 0.08, + "grad_norm": 0.427734375, + "learning_rate": 0.00019928721511937018, + "loss": 2.6712, + "step": 1881 + }, + { + "epoch": 0.08, + "grad_norm": 0.419921875, + "learning_rate": 0.00019928641541603698, + "loss": 2.7657, + "step": 1882 + }, + { + "epoch": 0.08, + "grad_norm": 0.41796875, + "learning_rate": 0.0001992856152659512, + "loss": 2.5951, + "step": 1883 + }, + { + "epoch": 0.08, + "grad_norm": 0.4296875, + "learning_rate": 0.00019928481466911644, + "loss": 2.8508, + "step": 1884 + }, + { + "epoch": 0.08, + "grad_norm": 0.357421875, + "learning_rate": 0.00019928401362553628, + "loss": 2.6222, + "step": 1885 + }, + { + "epoch": 0.08, + "grad_norm": 0.45703125, + "learning_rate": 0.00019928321213521434, + "loss": 2.8365, + "step": 1886 + }, + { + "epoch": 0.08, + "grad_norm": 0.400390625, + "learning_rate": 0.00019928241019815426, + "loss": 2.7106, + "step": 1887 + }, + { + "epoch": 0.08, + "grad_norm": 0.388671875, + "learning_rate": 0.0001992816078143596, + "loss": 2.7162, + "step": 1888 + }, + { + "epoch": 0.08, + "grad_norm": 0.4453125, + "learning_rate": 0.000199280804983834, + "loss": 2.6705, + "step": 1889 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.00019928000170658103, + "loss": 2.7153, + "step": 1890 + }, + { + "epoch": 0.08, + "grad_norm": 0.421875, + "learning_rate": 0.00019927919798260434, + "loss": 2.5607, + "step": 1891 + }, + { + "epoch": 0.08, + "grad_norm": 0.38671875, + "learning_rate": 0.00019927839381190756, + "loss": 2.5974, + "step": 1892 + }, + { + "epoch": 0.08, + "grad_norm": 0.435546875, + "learning_rate": 0.00019927758919449428, + "loss": 2.7655, + "step": 1893 + }, + { + "epoch": 0.08, + "grad_norm": 0.37890625, + "learning_rate": 0.00019927678413036816, + "loss": 2.6555, + "step": 1894 + }, + { + "epoch": 0.08, + "grad_norm": 0.42578125, + "learning_rate": 0.00019927597861953275, + "loss": 2.6979, + "step": 1895 + }, + { + "epoch": 0.08, + "grad_norm": 0.45703125, + "learning_rate": 0.00019927517266199172, + "loss": 2.7529, + "step": 1896 + }, + { + "epoch": 0.08, + "grad_norm": 0.3984375, + "learning_rate": 0.00019927436625774873, + "loss": 2.7625, + "step": 1897 + }, + { + "epoch": 0.08, + "grad_norm": 0.359375, + "learning_rate": 0.00019927355940680736, + "loss": 2.6893, + "step": 1898 + }, + { + "epoch": 0.08, + "grad_norm": 0.388671875, + "learning_rate": 0.00019927275210917123, + "loss": 2.732, + "step": 1899 + }, + { + "epoch": 0.08, + "grad_norm": 0.396484375, + "learning_rate": 0.00019927194436484403, + "loss": 2.6844, + "step": 1900 + }, + { + "epoch": 0.08, + "grad_norm": 0.36328125, + "learning_rate": 0.00019927113617382934, + "loss": 2.6372, + "step": 1901 + }, + { + "epoch": 0.08, + "grad_norm": 0.412109375, + "learning_rate": 0.00019927032753613082, + "loss": 2.7977, + "step": 1902 + }, + { + "epoch": 0.08, + "grad_norm": 0.462890625, + "learning_rate": 0.0001992695184517521, + "loss": 2.7896, + "step": 1903 + }, + { + "epoch": 0.08, + "grad_norm": 0.39453125, + "learning_rate": 0.00019926870892069683, + "loss": 2.6754, + "step": 1904 + }, + { + "epoch": 0.08, + "grad_norm": 0.369140625, + "learning_rate": 0.00019926789894296864, + "loss": 2.7555, + "step": 1905 + }, + { + "epoch": 0.08, + "grad_norm": 0.4296875, + "learning_rate": 0.00019926708851857118, + "loss": 2.6606, + "step": 1906 + }, + { + "epoch": 0.08, + "grad_norm": 0.3671875, + "learning_rate": 0.00019926627764750812, + "loss": 2.8078, + "step": 1907 + }, + { + "epoch": 0.08, + "grad_norm": 0.427734375, + "learning_rate": 0.00019926546632978307, + "loss": 2.6173, + "step": 1908 + }, + { + "epoch": 0.08, + "grad_norm": 0.380859375, + "learning_rate": 0.0001992646545653997, + "loss": 2.5832, + "step": 1909 + }, + { + "epoch": 0.08, + "grad_norm": 0.427734375, + "learning_rate": 0.00019926384235436167, + "loss": 2.8547, + "step": 1910 + }, + { + "epoch": 0.08, + "grad_norm": 0.3984375, + "learning_rate": 0.00019926302969667264, + "loss": 2.7276, + "step": 1911 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.0001992622165923362, + "loss": 2.6138, + "step": 1912 + }, + { + "epoch": 0.08, + "grad_norm": 0.4296875, + "learning_rate": 0.00019926140304135612, + "loss": 2.7894, + "step": 1913 + }, + { + "epoch": 0.08, + "grad_norm": 0.388671875, + "learning_rate": 0.000199260589043736, + "loss": 2.6956, + "step": 1914 + }, + { + "epoch": 0.08, + "grad_norm": 0.4453125, + "learning_rate": 0.0001992597745994795, + "loss": 2.8069, + "step": 1915 + }, + { + "epoch": 0.08, + "grad_norm": 0.37890625, + "learning_rate": 0.00019925895970859022, + "loss": 2.7307, + "step": 1916 + }, + { + "epoch": 0.08, + "grad_norm": 0.447265625, + "learning_rate": 0.00019925814437107196, + "loss": 2.7447, + "step": 1917 + }, + { + "epoch": 0.08, + "grad_norm": 0.390625, + "learning_rate": 0.00019925732858692833, + "loss": 2.6928, + "step": 1918 + }, + { + "epoch": 0.08, + "grad_norm": 0.435546875, + "learning_rate": 0.00019925651235616298, + "loss": 2.7137, + "step": 1919 + }, + { + "epoch": 0.08, + "grad_norm": 0.42578125, + "learning_rate": 0.00019925569567877963, + "loss": 2.7713, + "step": 1920 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.00019925487855478188, + "loss": 2.6959, + "step": 1921 + }, + { + "epoch": 0.08, + "grad_norm": 0.42578125, + "learning_rate": 0.00019925406098417344, + "loss": 2.7175, + "step": 1922 + }, + { + "epoch": 0.08, + "grad_norm": 0.419921875, + "learning_rate": 0.00019925324296695803, + "loss": 2.6254, + "step": 1923 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.00019925242450313931, + "loss": 2.6113, + "step": 1924 + }, + { + "epoch": 0.08, + "grad_norm": 0.396484375, + "learning_rate": 0.00019925160559272094, + "loss": 2.6411, + "step": 1925 + }, + { + "epoch": 0.08, + "grad_norm": 0.431640625, + "learning_rate": 0.0001992507862357066, + "loss": 2.782, + "step": 1926 + }, + { + "epoch": 0.08, + "grad_norm": 0.44140625, + "learning_rate": 0.0001992499664321, + "loss": 2.62, + "step": 1927 + }, + { + "epoch": 0.08, + "grad_norm": 0.375, + "learning_rate": 0.00019924914618190483, + "loss": 2.7481, + "step": 1928 + }, + { + "epoch": 0.08, + "grad_norm": 0.392578125, + "learning_rate": 0.0001992483254851248, + "loss": 2.7372, + "step": 1929 + }, + { + "epoch": 0.08, + "grad_norm": 0.40234375, + "learning_rate": 0.0001992475043417635, + "loss": 2.7478, + "step": 1930 + }, + { + "epoch": 0.08, + "grad_norm": 0.37109375, + "learning_rate": 0.00019924668275182476, + "loss": 2.6279, + "step": 1931 + }, + { + "epoch": 0.08, + "grad_norm": 0.376953125, + "learning_rate": 0.00019924586071531222, + "loss": 2.6074, + "step": 1932 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.00019924503823222958, + "loss": 2.767, + "step": 1933 + }, + { + "epoch": 0.08, + "grad_norm": 0.419921875, + "learning_rate": 0.0001992442153025805, + "loss": 2.5781, + "step": 1934 + }, + { + "epoch": 0.08, + "grad_norm": 0.384765625, + "learning_rate": 0.00019924339192636876, + "loss": 2.7, + "step": 1935 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.000199242568103598, + "loss": 2.7058, + "step": 1936 + }, + { + "epoch": 0.08, + "grad_norm": 0.392578125, + "learning_rate": 0.00019924174383427197, + "loss": 2.6135, + "step": 1937 + }, + { + "epoch": 0.08, + "grad_norm": 0.369140625, + "learning_rate": 0.00019924091911839435, + "loss": 2.7066, + "step": 1938 + }, + { + "epoch": 0.08, + "grad_norm": 0.384765625, + "learning_rate": 0.00019924009395596887, + "loss": 2.7034, + "step": 1939 + }, + { + "epoch": 0.08, + "grad_norm": 0.353515625, + "learning_rate": 0.00019923926834699925, + "loss": 2.5761, + "step": 1940 + }, + { + "epoch": 0.08, + "grad_norm": 0.390625, + "learning_rate": 0.00019923844229148915, + "loss": 2.6621, + "step": 1941 + }, + { + "epoch": 0.08, + "grad_norm": 0.365234375, + "learning_rate": 0.00019923761578944236, + "loss": 2.6291, + "step": 1942 + }, + { + "epoch": 0.08, + "grad_norm": 0.4296875, + "learning_rate": 0.00019923678884086255, + "loss": 2.6613, + "step": 1943 + }, + { + "epoch": 0.08, + "grad_norm": 0.40625, + "learning_rate": 0.00019923596144575344, + "loss": 2.7033, + "step": 1944 + }, + { + "epoch": 0.08, + "grad_norm": 0.37890625, + "learning_rate": 0.00019923513360411882, + "loss": 2.6215, + "step": 1945 + }, + { + "epoch": 0.08, + "grad_norm": 0.357421875, + "learning_rate": 0.00019923430531596233, + "loss": 2.6991, + "step": 1946 + }, + { + "epoch": 0.08, + "grad_norm": 0.376953125, + "learning_rate": 0.00019923347658128772, + "loss": 2.5654, + "step": 1947 + }, + { + "epoch": 0.08, + "grad_norm": 0.453125, + "learning_rate": 0.00019923264740009876, + "loss": 2.6793, + "step": 1948 + }, + { + "epoch": 0.08, + "grad_norm": 0.3671875, + "learning_rate": 0.00019923181777239913, + "loss": 2.7013, + "step": 1949 + }, + { + "epoch": 0.08, + "grad_norm": 0.447265625, + "learning_rate": 0.0001992309876981926, + "loss": 2.7268, + "step": 1950 + }, + { + "epoch": 0.08, + "grad_norm": 0.416015625, + "learning_rate": 0.0001992301571774829, + "loss": 2.887, + "step": 1951 + }, + { + "epoch": 0.08, + "grad_norm": 0.392578125, + "learning_rate": 0.00019922932621027376, + "loss": 2.7968, + "step": 1952 + }, + { + "epoch": 0.08, + "grad_norm": 0.3671875, + "learning_rate": 0.0001992284947965689, + "loss": 2.625, + "step": 1953 + }, + { + "epoch": 0.08, + "grad_norm": 0.375, + "learning_rate": 0.00019922766293637207, + "loss": 2.7486, + "step": 1954 + }, + { + "epoch": 0.08, + "grad_norm": 0.423828125, + "learning_rate": 0.00019922683062968704, + "loss": 2.7045, + "step": 1955 + }, + { + "epoch": 0.08, + "grad_norm": 0.39453125, + "learning_rate": 0.0001992259978765175, + "loss": 2.6884, + "step": 1956 + }, + { + "epoch": 0.08, + "grad_norm": 0.388671875, + "learning_rate": 0.00019922516467686725, + "loss": 2.5708, + "step": 1957 + }, + { + "epoch": 0.08, + "grad_norm": 0.40234375, + "learning_rate": 0.00019922433103074005, + "loss": 2.6575, + "step": 1958 + }, + { + "epoch": 0.08, + "grad_norm": 0.3671875, + "learning_rate": 0.0001992234969381396, + "loss": 2.6588, + "step": 1959 + }, + { + "epoch": 0.08, + "grad_norm": 0.36328125, + "learning_rate": 0.00019922266239906966, + "loss": 2.6631, + "step": 1960 + }, + { + "epoch": 0.08, + "grad_norm": 0.396484375, + "learning_rate": 0.000199221827413534, + "loss": 2.6366, + "step": 1961 + }, + { + "epoch": 0.08, + "grad_norm": 0.390625, + "learning_rate": 0.00019922099198153644, + "loss": 2.7294, + "step": 1962 + }, + { + "epoch": 0.08, + "grad_norm": 0.478515625, + "learning_rate": 0.0001992201561030806, + "loss": 2.8698, + "step": 1963 + }, + { + "epoch": 0.08, + "grad_norm": 0.400390625, + "learning_rate": 0.00019921931977817038, + "loss": 2.6334, + "step": 1964 + }, + { + "epoch": 0.08, + "grad_norm": 0.4296875, + "learning_rate": 0.00019921848300680944, + "loss": 2.8701, + "step": 1965 + }, + { + "epoch": 0.08, + "grad_norm": 0.3984375, + "learning_rate": 0.0001992176457890016, + "loss": 2.7257, + "step": 1966 + }, + { + "epoch": 0.08, + "grad_norm": 0.341796875, + "learning_rate": 0.0001992168081247506, + "loss": 2.6936, + "step": 1967 + }, + { + "epoch": 0.08, + "grad_norm": 0.390625, + "learning_rate": 0.00019921597001406024, + "loss": 2.7079, + "step": 1968 + }, + { + "epoch": 0.08, + "grad_norm": 0.388671875, + "learning_rate": 0.0001992151314569343, + "loss": 2.7133, + "step": 1969 + }, + { + "epoch": 0.08, + "grad_norm": 0.4140625, + "learning_rate": 0.00019921429245337645, + "loss": 2.6055, + "step": 1970 + }, + { + "epoch": 0.08, + "grad_norm": 0.404296875, + "learning_rate": 0.00019921345300339064, + "loss": 2.5836, + "step": 1971 + }, + { + "epoch": 0.08, + "grad_norm": 0.41015625, + "learning_rate": 0.00019921261310698048, + "loss": 2.7947, + "step": 1972 + }, + { + "epoch": 0.08, + "grad_norm": 0.392578125, + "learning_rate": 0.00019921177276414982, + "loss": 2.7672, + "step": 1973 + }, + { + "epoch": 0.08, + "grad_norm": 0.44921875, + "learning_rate": 0.00019921093197490248, + "loss": 2.7272, + "step": 1974 + }, + { + "epoch": 0.08, + "grad_norm": 0.435546875, + "learning_rate": 0.00019921009073924217, + "loss": 2.7959, + "step": 1975 + }, + { + "epoch": 0.08, + "grad_norm": 0.404296875, + "learning_rate": 0.00019920924905717272, + "loss": 2.8208, + "step": 1976 + }, + { + "epoch": 0.08, + "grad_norm": 0.43359375, + "learning_rate": 0.0001992084069286979, + "loss": 2.7016, + "step": 1977 + }, + { + "epoch": 0.08, + "grad_norm": 0.388671875, + "learning_rate": 0.00019920756435382153, + "loss": 2.7574, + "step": 1978 + }, + { + "epoch": 0.08, + "grad_norm": 0.34375, + "learning_rate": 0.00019920672133254735, + "loss": 2.7549, + "step": 1979 + }, + { + "epoch": 0.08, + "grad_norm": 0.4296875, + "learning_rate": 0.0001992058778648792, + "loss": 2.696, + "step": 1980 + }, + { + "epoch": 0.08, + "grad_norm": 0.3984375, + "learning_rate": 0.00019920503395082085, + "loss": 2.7054, + "step": 1981 + }, + { + "epoch": 0.08, + "grad_norm": 0.40234375, + "learning_rate": 0.0001992041895903761, + "loss": 2.7539, + "step": 1982 + }, + { + "epoch": 0.08, + "grad_norm": 0.365234375, + "learning_rate": 0.00019920334478354873, + "loss": 2.5828, + "step": 1983 + }, + { + "epoch": 0.08, + "grad_norm": 0.419921875, + "learning_rate": 0.0001992024995303426, + "loss": 2.7171, + "step": 1984 + }, + { + "epoch": 0.08, + "grad_norm": 0.427734375, + "learning_rate": 0.00019920165383076146, + "loss": 2.5784, + "step": 1985 + }, + { + "epoch": 0.08, + "grad_norm": 0.44140625, + "learning_rate": 0.00019920080768480913, + "loss": 2.8386, + "step": 1986 + }, + { + "epoch": 0.08, + "grad_norm": 0.41796875, + "learning_rate": 0.00019919996109248942, + "loss": 2.6509, + "step": 1987 + }, + { + "epoch": 0.08, + "grad_norm": 0.435546875, + "learning_rate": 0.00019919911405380614, + "loss": 2.7498, + "step": 1988 + }, + { + "epoch": 0.08, + "grad_norm": 0.482421875, + "learning_rate": 0.00019919826656876312, + "loss": 2.7155, + "step": 1989 + }, + { + "epoch": 0.08, + "grad_norm": 0.416015625, + "learning_rate": 0.00019919741863736408, + "loss": 2.8099, + "step": 1990 + }, + { + "epoch": 0.08, + "grad_norm": 0.484375, + "learning_rate": 0.00019919657025961298, + "loss": 2.6176, + "step": 1991 + }, + { + "epoch": 0.08, + "grad_norm": 0.400390625, + "learning_rate": 0.00019919572143551356, + "loss": 2.639, + "step": 1992 + }, + { + "epoch": 0.08, + "grad_norm": 0.458984375, + "learning_rate": 0.00019919487216506963, + "loss": 2.8329, + "step": 1993 + }, + { + "epoch": 0.08, + "grad_norm": 0.458984375, + "learning_rate": 0.00019919402244828502, + "loss": 2.6371, + "step": 1994 + }, + { + "epoch": 0.09, + "grad_norm": 0.41796875, + "learning_rate": 0.00019919317228516355, + "loss": 2.6138, + "step": 1995 + }, + { + "epoch": 0.09, + "grad_norm": 0.451171875, + "learning_rate": 0.0001991923216757091, + "loss": 2.7474, + "step": 1996 + }, + { + "epoch": 0.09, + "grad_norm": 0.42578125, + "learning_rate": 0.00019919147061992542, + "loss": 2.6903, + "step": 1997 + }, + { + "epoch": 0.09, + "grad_norm": 0.484375, + "learning_rate": 0.0001991906191178164, + "loss": 2.7328, + "step": 1998 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.00019918976716938583, + "loss": 2.7466, + "step": 1999 + }, + { + "epoch": 0.09, + "grad_norm": 0.392578125, + "learning_rate": 0.0001991889147746375, + "loss": 2.735, + "step": 2000 + }, + { + "epoch": 0.09, + "grad_norm": 0.486328125, + "learning_rate": 0.0001991880619335754, + "loss": 2.5764, + "step": 2001 + }, + { + "epoch": 0.09, + "grad_norm": 0.37890625, + "learning_rate": 0.0001991872086462032, + "loss": 2.6172, + "step": 2002 + }, + { + "epoch": 0.09, + "grad_norm": 0.50390625, + "learning_rate": 0.00019918635491252484, + "loss": 2.7728, + "step": 2003 + }, + { + "epoch": 0.09, + "grad_norm": 0.412109375, + "learning_rate": 0.00019918550073254413, + "loss": 2.6772, + "step": 2004 + }, + { + "epoch": 0.09, + "grad_norm": 0.396484375, + "learning_rate": 0.00019918464610626493, + "loss": 2.6641, + "step": 2005 + }, + { + "epoch": 0.09, + "grad_norm": 0.38671875, + "learning_rate": 0.00019918379103369102, + "loss": 2.6655, + "step": 2006 + }, + { + "epoch": 0.09, + "grad_norm": 0.40625, + "learning_rate": 0.00019918293551482634, + "loss": 2.5734, + "step": 2007 + }, + { + "epoch": 0.09, + "grad_norm": 0.412109375, + "learning_rate": 0.0001991820795496747, + "loss": 2.6409, + "step": 2008 + }, + { + "epoch": 0.09, + "grad_norm": 0.34765625, + "learning_rate": 0.00019918122313823992, + "loss": 2.6626, + "step": 2009 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.00019918036628052593, + "loss": 2.6396, + "step": 2010 + }, + { + "epoch": 0.09, + "grad_norm": 0.46484375, + "learning_rate": 0.00019917950897653647, + "loss": 2.8743, + "step": 2011 + }, + { + "epoch": 0.09, + "grad_norm": 0.37890625, + "learning_rate": 0.00019917865122627554, + "loss": 2.6408, + "step": 2012 + }, + { + "epoch": 0.09, + "grad_norm": 0.36328125, + "learning_rate": 0.00019917779302974686, + "loss": 2.6992, + "step": 2013 + }, + { + "epoch": 0.09, + "grad_norm": 0.41796875, + "learning_rate": 0.00019917693438695438, + "loss": 2.5389, + "step": 2014 + }, + { + "epoch": 0.09, + "grad_norm": 0.4453125, + "learning_rate": 0.00019917607529790194, + "loss": 2.8168, + "step": 2015 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.00019917521576259342, + "loss": 2.6312, + "step": 2016 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019917435578103266, + "loss": 2.8025, + "step": 2017 + }, + { + "epoch": 0.09, + "grad_norm": 0.365234375, + "learning_rate": 0.00019917349535322355, + "loss": 2.6314, + "step": 2018 + }, + { + "epoch": 0.09, + "grad_norm": 0.384765625, + "learning_rate": 0.00019917263447916995, + "loss": 2.7022, + "step": 2019 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.00019917177315887573, + "loss": 2.7255, + "step": 2020 + }, + { + "epoch": 0.09, + "grad_norm": 0.373046875, + "learning_rate": 0.0001991709113923448, + "loss": 2.7139, + "step": 2021 + }, + { + "epoch": 0.09, + "grad_norm": 0.5078125, + "learning_rate": 0.00019917004917958099, + "loss": 2.7412, + "step": 2022 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.0001991691865205882, + "loss": 2.675, + "step": 2023 + }, + { + "epoch": 0.09, + "grad_norm": 0.380859375, + "learning_rate": 0.00019916832341537033, + "loss": 2.5722, + "step": 2024 + }, + { + "epoch": 0.09, + "grad_norm": 0.458984375, + "learning_rate": 0.0001991674598639312, + "loss": 2.9125, + "step": 2025 + }, + { + "epoch": 0.09, + "grad_norm": 0.4453125, + "learning_rate": 0.00019916659586627483, + "loss": 2.6586, + "step": 2026 + }, + { + "epoch": 0.09, + "grad_norm": 0.3828125, + "learning_rate": 0.00019916573142240496, + "loss": 2.8212, + "step": 2027 + }, + { + "epoch": 0.09, + "grad_norm": 0.431640625, + "learning_rate": 0.00019916486653232554, + "loss": 2.5967, + "step": 2028 + }, + { + "epoch": 0.09, + "grad_norm": 0.380859375, + "learning_rate": 0.00019916400119604047, + "loss": 2.8164, + "step": 2029 + }, + { + "epoch": 0.09, + "grad_norm": 0.369140625, + "learning_rate": 0.00019916313541355364, + "loss": 2.6021, + "step": 2030 + }, + { + "epoch": 0.09, + "grad_norm": 0.453125, + "learning_rate": 0.0001991622691848689, + "loss": 2.7872, + "step": 2031 + }, + { + "epoch": 0.09, + "grad_norm": 0.421875, + "learning_rate": 0.00019916140250999024, + "loss": 2.6965, + "step": 2032 + }, + { + "epoch": 0.09, + "grad_norm": 0.431640625, + "learning_rate": 0.00019916053538892147, + "loss": 2.6606, + "step": 2033 + }, + { + "epoch": 0.09, + "grad_norm": 0.396484375, + "learning_rate": 0.00019915966782166656, + "loss": 2.7918, + "step": 2034 + }, + { + "epoch": 0.09, + "grad_norm": 0.423828125, + "learning_rate": 0.0001991587998082294, + "loss": 2.7707, + "step": 2035 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019915793134861384, + "loss": 2.6981, + "step": 2036 + }, + { + "epoch": 0.09, + "grad_norm": 0.361328125, + "learning_rate": 0.00019915706244282385, + "loss": 2.5794, + "step": 2037 + }, + { + "epoch": 0.09, + "grad_norm": 0.3828125, + "learning_rate": 0.0001991561930908633, + "loss": 2.7338, + "step": 2038 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.00019915532329273613, + "loss": 2.6532, + "step": 2039 + }, + { + "epoch": 0.09, + "grad_norm": 0.375, + "learning_rate": 0.00019915445304844622, + "loss": 2.7338, + "step": 2040 + }, + { + "epoch": 0.09, + "grad_norm": 0.375, + "learning_rate": 0.00019915358235799753, + "loss": 2.6427, + "step": 2041 + }, + { + "epoch": 0.09, + "grad_norm": 0.375, + "learning_rate": 0.00019915271122139397, + "loss": 2.7838, + "step": 2042 + }, + { + "epoch": 0.09, + "grad_norm": 0.3671875, + "learning_rate": 0.0001991518396386394, + "loss": 2.7031, + "step": 2043 + }, + { + "epoch": 0.09, + "grad_norm": 0.435546875, + "learning_rate": 0.00019915096760973782, + "loss": 2.6825, + "step": 2044 + }, + { + "epoch": 0.09, + "grad_norm": 0.421875, + "learning_rate": 0.0001991500951346931, + "loss": 2.7763, + "step": 2045 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019914922221350923, + "loss": 2.6468, + "step": 2046 + }, + { + "epoch": 0.09, + "grad_norm": 0.375, + "learning_rate": 0.00019914834884619004, + "loss": 2.7377, + "step": 2047 + }, + { + "epoch": 0.09, + "grad_norm": 0.41015625, + "learning_rate": 0.00019914747503273957, + "loss": 2.6688, + "step": 2048 + }, + { + "epoch": 0.09, + "grad_norm": 0.4140625, + "learning_rate": 0.00019914660077316164, + "loss": 2.7499, + "step": 2049 + }, + { + "epoch": 0.09, + "grad_norm": 0.4375, + "learning_rate": 0.0001991457260674603, + "loss": 2.7322, + "step": 2050 + }, + { + "epoch": 0.09, + "grad_norm": 0.4375, + "learning_rate": 0.00019914485091563936, + "loss": 2.7843, + "step": 2051 + }, + { + "epoch": 0.09, + "grad_norm": 0.423828125, + "learning_rate": 0.0001991439753177029, + "loss": 2.896, + "step": 2052 + }, + { + "epoch": 0.09, + "grad_norm": 0.369140625, + "learning_rate": 0.0001991430992736547, + "loss": 2.6288, + "step": 2053 + }, + { + "epoch": 0.09, + "grad_norm": 0.39453125, + "learning_rate": 0.00019914222278349887, + "loss": 2.6674, + "step": 2054 + }, + { + "epoch": 0.09, + "grad_norm": 0.38671875, + "learning_rate": 0.00019914134584723923, + "loss": 2.6484, + "step": 2055 + }, + { + "epoch": 0.09, + "grad_norm": 0.396484375, + "learning_rate": 0.00019914046846487976, + "loss": 2.6589, + "step": 2056 + }, + { + "epoch": 0.09, + "grad_norm": 0.37890625, + "learning_rate": 0.0001991395906364244, + "loss": 2.549, + "step": 2057 + }, + { + "epoch": 0.09, + "grad_norm": 0.412109375, + "learning_rate": 0.00019913871236187713, + "loss": 2.6737, + "step": 2058 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.0001991378336412419, + "loss": 2.7705, + "step": 2059 + }, + { + "epoch": 0.09, + "grad_norm": 0.42578125, + "learning_rate": 0.00019913695447452265, + "loss": 2.6354, + "step": 2060 + }, + { + "epoch": 0.09, + "grad_norm": 0.380859375, + "learning_rate": 0.0001991360748617233, + "loss": 2.6193, + "step": 2061 + }, + { + "epoch": 0.09, + "grad_norm": 0.396484375, + "learning_rate": 0.00019913519480284787, + "loss": 2.6597, + "step": 2062 + }, + { + "epoch": 0.09, + "grad_norm": 0.41015625, + "learning_rate": 0.0001991343142979003, + "loss": 2.6189, + "step": 2063 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.00019913343334688452, + "loss": 2.7605, + "step": 2064 + }, + { + "epoch": 0.09, + "grad_norm": 0.43359375, + "learning_rate": 0.00019913255194980452, + "loss": 2.7743, + "step": 2065 + }, + { + "epoch": 0.09, + "grad_norm": 0.369140625, + "learning_rate": 0.0001991316701066643, + "loss": 2.6179, + "step": 2066 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.00019913078781746776, + "loss": 2.6533, + "step": 2067 + }, + { + "epoch": 0.09, + "grad_norm": 0.40625, + "learning_rate": 0.00019912990508221893, + "loss": 2.8356, + "step": 2068 + }, + { + "epoch": 0.09, + "grad_norm": 0.36328125, + "learning_rate": 0.00019912902190092173, + "loss": 2.6125, + "step": 2069 + }, + { + "epoch": 0.09, + "grad_norm": 0.423828125, + "learning_rate": 0.00019912813827358016, + "loss": 2.6285, + "step": 2070 + }, + { + "epoch": 0.09, + "grad_norm": 0.37109375, + "learning_rate": 0.00019912725420019822, + "loss": 2.669, + "step": 2071 + }, + { + "epoch": 0.09, + "grad_norm": 0.39453125, + "learning_rate": 0.00019912636968077985, + "loss": 2.7873, + "step": 2072 + }, + { + "epoch": 0.09, + "grad_norm": 0.37109375, + "learning_rate": 0.00019912548471532905, + "loss": 2.6717, + "step": 2073 + }, + { + "epoch": 0.09, + "grad_norm": 0.37109375, + "learning_rate": 0.00019912459930384978, + "loss": 2.748, + "step": 2074 + }, + { + "epoch": 0.09, + "grad_norm": 0.4453125, + "learning_rate": 0.00019912371344634603, + "loss": 2.7911, + "step": 2075 + }, + { + "epoch": 0.09, + "grad_norm": 0.396484375, + "learning_rate": 0.00019912282714282183, + "loss": 2.7807, + "step": 2076 + }, + { + "epoch": 0.09, + "grad_norm": 0.365234375, + "learning_rate": 0.0001991219403932811, + "loss": 2.7924, + "step": 2077 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.00019912105319772785, + "loss": 2.757, + "step": 2078 + }, + { + "epoch": 0.09, + "grad_norm": 0.41015625, + "learning_rate": 0.00019912016555616614, + "loss": 2.7873, + "step": 2079 + }, + { + "epoch": 0.09, + "grad_norm": 0.45703125, + "learning_rate": 0.00019911927746859987, + "loss": 2.6729, + "step": 2080 + }, + { + "epoch": 0.09, + "grad_norm": 0.451171875, + "learning_rate": 0.00019911838893503308, + "loss": 2.7622, + "step": 2081 + }, + { + "epoch": 0.09, + "grad_norm": 0.384765625, + "learning_rate": 0.00019911749995546978, + "loss": 2.7533, + "step": 2082 + }, + { + "epoch": 0.09, + "grad_norm": 0.36328125, + "learning_rate": 0.00019911661052991392, + "loss": 2.6807, + "step": 2083 + }, + { + "epoch": 0.09, + "grad_norm": 0.38671875, + "learning_rate": 0.00019911572065836958, + "loss": 2.642, + "step": 2084 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019911483034084067, + "loss": 2.849, + "step": 2085 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019911393957733128, + "loss": 2.6924, + "step": 2086 + }, + { + "epoch": 0.09, + "grad_norm": 0.43359375, + "learning_rate": 0.00019911304836784538, + "loss": 2.8231, + "step": 2087 + }, + { + "epoch": 0.09, + "grad_norm": 0.443359375, + "learning_rate": 0.00019911215671238698, + "loss": 2.662, + "step": 2088 + }, + { + "epoch": 0.09, + "grad_norm": 0.34765625, + "learning_rate": 0.00019911126461096007, + "loss": 2.6717, + "step": 2089 + }, + { + "epoch": 0.09, + "grad_norm": 0.4140625, + "learning_rate": 0.00019911037206356874, + "loss": 2.7453, + "step": 2090 + }, + { + "epoch": 0.09, + "grad_norm": 0.404296875, + "learning_rate": 0.00019910947907021692, + "loss": 2.804, + "step": 2091 + }, + { + "epoch": 0.09, + "grad_norm": 0.357421875, + "learning_rate": 0.00019910858563090868, + "loss": 2.5852, + "step": 2092 + }, + { + "epoch": 0.09, + "grad_norm": 0.400390625, + "learning_rate": 0.00019910769174564803, + "loss": 2.8811, + "step": 2093 + }, + { + "epoch": 0.09, + "grad_norm": 0.400390625, + "learning_rate": 0.00019910679741443898, + "loss": 2.6352, + "step": 2094 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.00019910590263728554, + "loss": 2.7649, + "step": 2095 + }, + { + "epoch": 0.09, + "grad_norm": 0.431640625, + "learning_rate": 0.0001991050074141918, + "loss": 2.816, + "step": 2096 + }, + { + "epoch": 0.09, + "grad_norm": 0.40234375, + "learning_rate": 0.00019910411174516173, + "loss": 2.7343, + "step": 2097 + }, + { + "epoch": 0.09, + "grad_norm": 0.369140625, + "learning_rate": 0.00019910321563019936, + "loss": 2.7597, + "step": 2098 + }, + { + "epoch": 0.09, + "grad_norm": 0.3984375, + "learning_rate": 0.00019910231906930875, + "loss": 2.8012, + "step": 2099 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019910142206249393, + "loss": 2.7279, + "step": 2100 + }, + { + "epoch": 0.09, + "grad_norm": 0.37109375, + "learning_rate": 0.0001991005246097589, + "loss": 2.7667, + "step": 2101 + }, + { + "epoch": 0.09, + "grad_norm": 0.462890625, + "learning_rate": 0.00019909962671110773, + "loss": 2.7446, + "step": 2102 + }, + { + "epoch": 0.09, + "grad_norm": 0.412109375, + "learning_rate": 0.0001990987283665445, + "loss": 2.8122, + "step": 2103 + }, + { + "epoch": 0.09, + "grad_norm": 0.41015625, + "learning_rate": 0.00019909782957607317, + "loss": 2.6787, + "step": 2104 + }, + { + "epoch": 0.09, + "grad_norm": 0.423828125, + "learning_rate": 0.0001990969303396978, + "loss": 2.6596, + "step": 2105 + }, + { + "epoch": 0.09, + "grad_norm": 0.4375, + "learning_rate": 0.00019909603065742252, + "loss": 2.7693, + "step": 2106 + }, + { + "epoch": 0.09, + "grad_norm": 0.4609375, + "learning_rate": 0.00019909513052925127, + "loss": 2.753, + "step": 2107 + }, + { + "epoch": 0.09, + "grad_norm": 0.431640625, + "learning_rate": 0.00019909422995518817, + "loss": 2.7868, + "step": 2108 + }, + { + "epoch": 0.09, + "grad_norm": 0.44921875, + "learning_rate": 0.00019909332893523724, + "loss": 2.7608, + "step": 2109 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.00019909242746940256, + "loss": 2.6859, + "step": 2110 + }, + { + "epoch": 0.09, + "grad_norm": 0.392578125, + "learning_rate": 0.00019909152555768817, + "loss": 2.6296, + "step": 2111 + }, + { + "epoch": 0.09, + "grad_norm": 0.41015625, + "learning_rate": 0.0001990906232000981, + "loss": 2.6686, + "step": 2112 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019908972039663647, + "loss": 2.723, + "step": 2113 + }, + { + "epoch": 0.09, + "grad_norm": 0.439453125, + "learning_rate": 0.0001990888171473073, + "loss": 2.792, + "step": 2114 + }, + { + "epoch": 0.09, + "grad_norm": 0.408203125, + "learning_rate": 0.00019908791345211465, + "loss": 2.591, + "step": 2115 + }, + { + "epoch": 0.09, + "grad_norm": 0.38671875, + "learning_rate": 0.0001990870093110626, + "loss": 2.7267, + "step": 2116 + }, + { + "epoch": 0.09, + "grad_norm": 0.3828125, + "learning_rate": 0.00019908610472415528, + "loss": 2.5635, + "step": 2117 + }, + { + "epoch": 0.09, + "grad_norm": 0.40625, + "learning_rate": 0.00019908519969139663, + "loss": 2.6483, + "step": 2118 + }, + { + "epoch": 0.09, + "grad_norm": 0.36328125, + "learning_rate": 0.00019908429421279086, + "loss": 2.665, + "step": 2119 + }, + { + "epoch": 0.09, + "grad_norm": 0.345703125, + "learning_rate": 0.00019908338828834193, + "loss": 2.7444, + "step": 2120 + }, + { + "epoch": 0.09, + "grad_norm": 0.3828125, + "learning_rate": 0.00019908248191805394, + "loss": 2.7377, + "step": 2121 + }, + { + "epoch": 0.09, + "grad_norm": 0.380859375, + "learning_rate": 0.00019908157510193104, + "loss": 2.6363, + "step": 2122 + }, + { + "epoch": 0.09, + "grad_norm": 0.3984375, + "learning_rate": 0.00019908066783997725, + "loss": 2.7363, + "step": 2123 + }, + { + "epoch": 0.09, + "grad_norm": 0.400390625, + "learning_rate": 0.00019907976013219664, + "loss": 2.7024, + "step": 2124 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019907885197859335, + "loss": 2.8456, + "step": 2125 + }, + { + "epoch": 0.09, + "grad_norm": 0.431640625, + "learning_rate": 0.00019907794337917142, + "loss": 2.6286, + "step": 2126 + }, + { + "epoch": 0.09, + "grad_norm": 0.380859375, + "learning_rate": 0.00019907703433393495, + "loss": 2.6641, + "step": 2127 + }, + { + "epoch": 0.09, + "grad_norm": 0.369140625, + "learning_rate": 0.00019907612484288808, + "loss": 2.7004, + "step": 2128 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.00019907521490603478, + "loss": 2.7048, + "step": 2129 + }, + { + "epoch": 0.09, + "grad_norm": 0.38671875, + "learning_rate": 0.00019907430452337926, + "loss": 2.512, + "step": 2130 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.0001990733936949256, + "loss": 2.6096, + "step": 2131 + }, + { + "epoch": 0.09, + "grad_norm": 0.35546875, + "learning_rate": 0.00019907248242067785, + "loss": 2.8126, + "step": 2132 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.00019907157070064014, + "loss": 2.603, + "step": 2133 + }, + { + "epoch": 0.09, + "grad_norm": 0.375, + "learning_rate": 0.00019907065853481655, + "loss": 2.8088, + "step": 2134 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.00019906974592321123, + "loss": 2.535, + "step": 2135 + }, + { + "epoch": 0.09, + "grad_norm": 0.380859375, + "learning_rate": 0.00019906883286582825, + "loss": 2.5722, + "step": 2136 + }, + { + "epoch": 0.09, + "grad_norm": 0.400390625, + "learning_rate": 0.0001990679193626717, + "loss": 2.76, + "step": 2137 + }, + { + "epoch": 0.09, + "grad_norm": 0.435546875, + "learning_rate": 0.00019906700541374577, + "loss": 2.7037, + "step": 2138 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019906609101905448, + "loss": 2.8087, + "step": 2139 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.000199065176178602, + "loss": 2.6797, + "step": 2140 + }, + { + "epoch": 0.09, + "grad_norm": 0.365234375, + "learning_rate": 0.00019906426089239245, + "loss": 2.8238, + "step": 2141 + }, + { + "epoch": 0.09, + "grad_norm": 0.36328125, + "learning_rate": 0.0001990633451604299, + "loss": 2.5375, + "step": 2142 + }, + { + "epoch": 0.09, + "grad_norm": 0.421875, + "learning_rate": 0.0001990624289827185, + "loss": 2.859, + "step": 2143 + }, + { + "epoch": 0.09, + "grad_norm": 0.3359375, + "learning_rate": 0.00019906151235926238, + "loss": 2.7063, + "step": 2144 + }, + { + "epoch": 0.09, + "grad_norm": 0.3984375, + "learning_rate": 0.00019906059529006568, + "loss": 2.601, + "step": 2145 + }, + { + "epoch": 0.09, + "grad_norm": 0.41796875, + "learning_rate": 0.00019905967777513248, + "loss": 2.6871, + "step": 2146 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.00019905875981446694, + "loss": 2.7524, + "step": 2147 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.00019905784140807317, + "loss": 2.726, + "step": 2148 + }, + { + "epoch": 0.09, + "grad_norm": 0.443359375, + "learning_rate": 0.00019905692255595532, + "loss": 2.6326, + "step": 2149 + }, + { + "epoch": 0.09, + "grad_norm": 0.458984375, + "learning_rate": 0.00019905600325811752, + "loss": 2.7216, + "step": 2150 + }, + { + "epoch": 0.09, + "grad_norm": 0.359375, + "learning_rate": 0.00019905508351456393, + "loss": 2.6053, + "step": 2151 + }, + { + "epoch": 0.09, + "grad_norm": 0.421875, + "learning_rate": 0.00019905416332529865, + "loss": 2.7354, + "step": 2152 + }, + { + "epoch": 0.09, + "grad_norm": 0.359375, + "learning_rate": 0.00019905324269032582, + "loss": 2.5916, + "step": 2153 + }, + { + "epoch": 0.09, + "grad_norm": 0.408203125, + "learning_rate": 0.0001990523216096496, + "loss": 2.6704, + "step": 2154 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019905140008327413, + "loss": 2.7196, + "step": 2155 + }, + { + "epoch": 0.09, + "grad_norm": 0.404296875, + "learning_rate": 0.00019905047811120355, + "loss": 2.6677, + "step": 2156 + }, + { + "epoch": 0.09, + "grad_norm": 0.486328125, + "learning_rate": 0.00019904955569344205, + "loss": 2.7508, + "step": 2157 + }, + { + "epoch": 0.09, + "grad_norm": 0.396484375, + "learning_rate": 0.00019904863282999377, + "loss": 2.6902, + "step": 2158 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.0001990477095208628, + "loss": 2.7799, + "step": 2159 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.0001990467857660533, + "loss": 2.6151, + "step": 2160 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019904586156556953, + "loss": 2.7009, + "step": 2161 + }, + { + "epoch": 0.09, + "grad_norm": 0.40234375, + "learning_rate": 0.00019904493691941556, + "loss": 2.7223, + "step": 2162 + }, + { + "epoch": 0.09, + "grad_norm": 0.400390625, + "learning_rate": 0.00019904401182759554, + "loss": 2.5899, + "step": 2163 + }, + { + "epoch": 0.09, + "grad_norm": 0.400390625, + "learning_rate": 0.00019904308629011368, + "loss": 2.7818, + "step": 2164 + }, + { + "epoch": 0.09, + "grad_norm": 0.408203125, + "learning_rate": 0.00019904216030697413, + "loss": 2.7136, + "step": 2165 + }, + { + "epoch": 0.09, + "grad_norm": 0.365234375, + "learning_rate": 0.00019904123387818107, + "loss": 2.6556, + "step": 2166 + }, + { + "epoch": 0.09, + "grad_norm": 0.345703125, + "learning_rate": 0.00019904030700373864, + "loss": 2.8147, + "step": 2167 + }, + { + "epoch": 0.09, + "grad_norm": 0.40625, + "learning_rate": 0.00019903937968365103, + "loss": 2.7609, + "step": 2168 + }, + { + "epoch": 0.09, + "grad_norm": 0.404296875, + "learning_rate": 0.0001990384519179224, + "loss": 2.7672, + "step": 2169 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019903752370655693, + "loss": 2.7015, + "step": 2170 + }, + { + "epoch": 0.09, + "grad_norm": 0.423828125, + "learning_rate": 0.0001990365950495588, + "loss": 2.7802, + "step": 2171 + }, + { + "epoch": 0.09, + "grad_norm": 0.404296875, + "learning_rate": 0.00019903566594693217, + "loss": 2.7217, + "step": 2172 + }, + { + "epoch": 0.09, + "grad_norm": 0.408203125, + "learning_rate": 0.00019903473639868127, + "loss": 2.6668, + "step": 2173 + }, + { + "epoch": 0.09, + "grad_norm": 0.3984375, + "learning_rate": 0.0001990338064048102, + "loss": 2.6122, + "step": 2174 + }, + { + "epoch": 0.09, + "grad_norm": 0.427734375, + "learning_rate": 0.00019903287596532323, + "loss": 2.738, + "step": 2175 + }, + { + "epoch": 0.09, + "grad_norm": 0.349609375, + "learning_rate": 0.00019903194508022451, + "loss": 2.7403, + "step": 2176 + }, + { + "epoch": 0.09, + "grad_norm": 0.462890625, + "learning_rate": 0.00019903101374951822, + "loss": 2.7957, + "step": 2177 + }, + { + "epoch": 0.09, + "grad_norm": 0.3984375, + "learning_rate": 0.00019903008197320857, + "loss": 2.7463, + "step": 2178 + }, + { + "epoch": 0.09, + "grad_norm": 0.38671875, + "learning_rate": 0.00019902914975129975, + "loss": 2.5923, + "step": 2179 + }, + { + "epoch": 0.09, + "grad_norm": 0.474609375, + "learning_rate": 0.00019902821708379592, + "loss": 2.8195, + "step": 2180 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.00019902728397070131, + "loss": 2.7395, + "step": 2181 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.00019902635041202013, + "loss": 2.6928, + "step": 2182 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.00019902541640775656, + "loss": 2.691, + "step": 2183 + }, + { + "epoch": 0.09, + "grad_norm": 0.384765625, + "learning_rate": 0.00019902448195791478, + "loss": 2.7016, + "step": 2184 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019902354706249903, + "loss": 2.6369, + "step": 2185 + }, + { + "epoch": 0.09, + "grad_norm": 0.3515625, + "learning_rate": 0.00019902261172151354, + "loss": 2.639, + "step": 2186 + }, + { + "epoch": 0.09, + "grad_norm": 0.462890625, + "learning_rate": 0.00019902167593496243, + "loss": 2.7048, + "step": 2187 + }, + { + "epoch": 0.09, + "grad_norm": 0.43359375, + "learning_rate": 0.00019902073970285003, + "loss": 2.768, + "step": 2188 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.00019901980302518046, + "loss": 2.6657, + "step": 2189 + }, + { + "epoch": 0.09, + "grad_norm": 0.4296875, + "learning_rate": 0.00019901886590195796, + "loss": 2.6968, + "step": 2190 + }, + { + "epoch": 0.09, + "grad_norm": 0.447265625, + "learning_rate": 0.00019901792833318673, + "loss": 2.6241, + "step": 2191 + }, + { + "epoch": 0.09, + "grad_norm": 0.375, + "learning_rate": 0.00019901699031887104, + "loss": 2.5816, + "step": 2192 + }, + { + "epoch": 0.09, + "grad_norm": 0.42578125, + "learning_rate": 0.0001990160518590151, + "loss": 2.8102, + "step": 2193 + }, + { + "epoch": 0.09, + "grad_norm": 0.484375, + "learning_rate": 0.00019901511295362308, + "loss": 2.6963, + "step": 2194 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019901417360269924, + "loss": 2.7085, + "step": 2195 + }, + { + "epoch": 0.09, + "grad_norm": 0.38671875, + "learning_rate": 0.0001990132338062478, + "loss": 2.666, + "step": 2196 + }, + { + "epoch": 0.09, + "grad_norm": 0.3984375, + "learning_rate": 0.000199012293564273, + "loss": 2.5781, + "step": 2197 + }, + { + "epoch": 0.09, + "grad_norm": 0.453125, + "learning_rate": 0.00019901135287677906, + "loss": 2.756, + "step": 2198 + }, + { + "epoch": 0.09, + "grad_norm": 0.373046875, + "learning_rate": 0.0001990104117437702, + "loss": 2.6591, + "step": 2199 + }, + { + "epoch": 0.09, + "grad_norm": 0.3828125, + "learning_rate": 0.0001990094701652507, + "loss": 2.5932, + "step": 2200 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019900852814122474, + "loss": 2.8447, + "step": 2201 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.00019900758567169665, + "loss": 2.6973, + "step": 2202 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019900664275667053, + "loss": 2.575, + "step": 2203 + }, + { + "epoch": 0.09, + "grad_norm": 0.453125, + "learning_rate": 0.00019900569939615076, + "loss": 2.7551, + "step": 2204 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.0001990047555901415, + "loss": 2.7544, + "step": 2205 + }, + { + "epoch": 0.09, + "grad_norm": 0.43359375, + "learning_rate": 0.00019900381133864702, + "loss": 2.8995, + "step": 2206 + }, + { + "epoch": 0.09, + "grad_norm": 0.4296875, + "learning_rate": 0.00019900286664167156, + "loss": 2.8178, + "step": 2207 + }, + { + "epoch": 0.09, + "grad_norm": 0.451171875, + "learning_rate": 0.00019900192149921938, + "loss": 2.7801, + "step": 2208 + }, + { + "epoch": 0.09, + "grad_norm": 0.416015625, + "learning_rate": 0.00019900097591129477, + "loss": 2.6908, + "step": 2209 + }, + { + "epoch": 0.09, + "grad_norm": 0.390625, + "learning_rate": 0.00019900002987790192, + "loss": 2.7721, + "step": 2210 + }, + { + "epoch": 0.09, + "grad_norm": 0.4140625, + "learning_rate": 0.00019899908339904513, + "loss": 2.796, + "step": 2211 + }, + { + "epoch": 0.09, + "grad_norm": 0.4296875, + "learning_rate": 0.00019899813647472863, + "loss": 2.8204, + "step": 2212 + }, + { + "epoch": 0.09, + "grad_norm": 0.376953125, + "learning_rate": 0.0001989971891049567, + "loss": 2.7724, + "step": 2213 + }, + { + "epoch": 0.09, + "grad_norm": 0.396484375, + "learning_rate": 0.00019899624128973359, + "loss": 2.6855, + "step": 2214 + }, + { + "epoch": 0.09, + "grad_norm": 0.427734375, + "learning_rate": 0.0001989952930290636, + "loss": 2.7617, + "step": 2215 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019899434432295096, + "loss": 2.6465, + "step": 2216 + }, + { + "epoch": 0.09, + "grad_norm": 0.43359375, + "learning_rate": 0.00019899339517139995, + "loss": 2.8258, + "step": 2217 + }, + { + "epoch": 0.09, + "grad_norm": 0.419921875, + "learning_rate": 0.00019899244557441482, + "loss": 2.7425, + "step": 2218 + }, + { + "epoch": 0.09, + "grad_norm": 0.380859375, + "learning_rate": 0.00019899149553199988, + "loss": 2.5556, + "step": 2219 + }, + { + "epoch": 0.09, + "grad_norm": 0.384765625, + "learning_rate": 0.00019899054504415937, + "loss": 2.6678, + "step": 2220 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.0001989895941108976, + "loss": 2.5645, + "step": 2221 + }, + { + "epoch": 0.09, + "grad_norm": 0.400390625, + "learning_rate": 0.00019898864273221883, + "loss": 2.6759, + "step": 2222 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.00019898769090812737, + "loss": 2.7711, + "step": 2223 + }, + { + "epoch": 0.09, + "grad_norm": 0.408203125, + "learning_rate": 0.00019898673863862746, + "loss": 2.7706, + "step": 2224 + }, + { + "epoch": 0.09, + "grad_norm": 0.439453125, + "learning_rate": 0.00019898578592372337, + "loss": 2.6965, + "step": 2225 + }, + { + "epoch": 0.09, + "grad_norm": 0.40234375, + "learning_rate": 0.00019898483276341943, + "loss": 2.6485, + "step": 2226 + }, + { + "epoch": 0.09, + "grad_norm": 0.359375, + "learning_rate": 0.00019898387915771992, + "loss": 2.7146, + "step": 2227 + }, + { + "epoch": 0.09, + "grad_norm": 0.40625, + "learning_rate": 0.00019898292510662916, + "loss": 2.5707, + "step": 2228 + }, + { + "epoch": 0.09, + "grad_norm": 0.388671875, + "learning_rate": 0.00019898197061015137, + "loss": 2.7306, + "step": 2229 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.00019898101566829092, + "loss": 2.7448, + "step": 2230 + }, + { + "epoch": 0.1, + "grad_norm": 0.373046875, + "learning_rate": 0.00019898006028105204, + "loss": 2.7465, + "step": 2231 + }, + { + "epoch": 0.1, + "grad_norm": 0.369140625, + "learning_rate": 0.00019897910444843908, + "loss": 2.738, + "step": 2232 + }, + { + "epoch": 0.1, + "grad_norm": 0.41015625, + "learning_rate": 0.00019897814817045634, + "loss": 2.7679, + "step": 2233 + }, + { + "epoch": 0.1, + "grad_norm": 0.375, + "learning_rate": 0.0001989771914471081, + "loss": 2.6936, + "step": 2234 + }, + { + "epoch": 0.1, + "grad_norm": 0.435546875, + "learning_rate": 0.00019897623427839864, + "loss": 2.7169, + "step": 2235 + }, + { + "epoch": 0.1, + "grad_norm": 0.38671875, + "learning_rate": 0.0001989752766643323, + "loss": 2.6627, + "step": 2236 + }, + { + "epoch": 0.1, + "grad_norm": 0.44140625, + "learning_rate": 0.00019897431860491343, + "loss": 2.6499, + "step": 2237 + }, + { + "epoch": 0.1, + "grad_norm": 0.380859375, + "learning_rate": 0.00019897336010014626, + "loss": 2.6138, + "step": 2238 + }, + { + "epoch": 0.1, + "grad_norm": 0.439453125, + "learning_rate": 0.00019897240115003512, + "loss": 2.7808, + "step": 2239 + }, + { + "epoch": 0.1, + "grad_norm": 0.4296875, + "learning_rate": 0.0001989714417545844, + "loss": 2.6765, + "step": 2240 + }, + { + "epoch": 0.1, + "grad_norm": 0.427734375, + "learning_rate": 0.0001989704819137983, + "loss": 2.7794, + "step": 2241 + }, + { + "epoch": 0.1, + "grad_norm": 0.451171875, + "learning_rate": 0.00019896952162768127, + "loss": 2.7584, + "step": 2242 + }, + { + "epoch": 0.1, + "grad_norm": 0.42578125, + "learning_rate": 0.00019896856089623752, + "loss": 2.6277, + "step": 2243 + }, + { + "epoch": 0.1, + "grad_norm": 0.3515625, + "learning_rate": 0.00019896759971947144, + "loss": 2.621, + "step": 2244 + }, + { + "epoch": 0.1, + "grad_norm": 0.421875, + "learning_rate": 0.0001989666380973873, + "loss": 2.8129, + "step": 2245 + }, + { + "epoch": 0.1, + "grad_norm": 0.427734375, + "learning_rate": 0.00019896567602998948, + "loss": 2.6986, + "step": 2246 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019896471351728225, + "loss": 2.6095, + "step": 2247 + }, + { + "epoch": 0.1, + "grad_norm": 0.40234375, + "learning_rate": 0.00019896375055927003, + "loss": 2.6918, + "step": 2248 + }, + { + "epoch": 0.1, + "grad_norm": 0.474609375, + "learning_rate": 0.00019896278715595707, + "loss": 2.7115, + "step": 2249 + }, + { + "epoch": 0.1, + "grad_norm": 0.44140625, + "learning_rate": 0.00019896182330734772, + "loss": 2.7786, + "step": 2250 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.00019896085901344637, + "loss": 2.6543, + "step": 2251 + }, + { + "epoch": 0.1, + "grad_norm": 0.4140625, + "learning_rate": 0.00019895989427425728, + "loss": 2.6633, + "step": 2252 + }, + { + "epoch": 0.1, + "grad_norm": 0.4609375, + "learning_rate": 0.00019895892908978487, + "loss": 2.6786, + "step": 2253 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.00019895796346003342, + "loss": 2.7776, + "step": 2254 + }, + { + "epoch": 0.1, + "grad_norm": 0.404296875, + "learning_rate": 0.0001989569973850073, + "loss": 2.7216, + "step": 2255 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.00019895603086471085, + "loss": 2.563, + "step": 2256 + }, + { + "epoch": 0.1, + "grad_norm": 0.39453125, + "learning_rate": 0.00019895506389914843, + "loss": 2.7114, + "step": 2257 + }, + { + "epoch": 0.1, + "grad_norm": 0.427734375, + "learning_rate": 0.0001989540964883244, + "loss": 2.7527, + "step": 2258 + }, + { + "epoch": 0.1, + "grad_norm": 0.431640625, + "learning_rate": 0.0001989531286322431, + "loss": 2.7558, + "step": 2259 + }, + { + "epoch": 0.1, + "grad_norm": 0.4453125, + "learning_rate": 0.00019895216033090886, + "loss": 2.7017, + "step": 2260 + }, + { + "epoch": 0.1, + "grad_norm": 0.412109375, + "learning_rate": 0.00019895119158432607, + "loss": 2.7721, + "step": 2261 + }, + { + "epoch": 0.1, + "grad_norm": 0.3828125, + "learning_rate": 0.0001989502223924991, + "loss": 2.6814, + "step": 2262 + }, + { + "epoch": 0.1, + "grad_norm": 0.42578125, + "learning_rate": 0.00019894925275543227, + "loss": 2.7709, + "step": 2263 + }, + { + "epoch": 0.1, + "grad_norm": 0.34375, + "learning_rate": 0.00019894828267312997, + "loss": 2.7107, + "step": 2264 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019894731214559655, + "loss": 2.6362, + "step": 2265 + }, + { + "epoch": 0.1, + "grad_norm": 0.361328125, + "learning_rate": 0.0001989463411728364, + "loss": 2.6701, + "step": 2266 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.00019894536975485392, + "loss": 2.5993, + "step": 2267 + }, + { + "epoch": 0.1, + "grad_norm": 0.345703125, + "learning_rate": 0.00019894439789165338, + "loss": 2.5608, + "step": 2268 + }, + { + "epoch": 0.1, + "grad_norm": 0.3828125, + "learning_rate": 0.0001989434255832392, + "loss": 2.6923, + "step": 2269 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.0001989424528296158, + "loss": 2.6734, + "step": 2270 + }, + { + "epoch": 0.1, + "grad_norm": 0.423828125, + "learning_rate": 0.0001989414796307875, + "loss": 2.7051, + "step": 2271 + }, + { + "epoch": 0.1, + "grad_norm": 0.421875, + "learning_rate": 0.00019894050598675872, + "loss": 2.6298, + "step": 2272 + }, + { + "epoch": 0.1, + "grad_norm": 0.423828125, + "learning_rate": 0.0001989395318975338, + "loss": 2.7764, + "step": 2273 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.00019893855736311715, + "loss": 2.7141, + "step": 2274 + }, + { + "epoch": 0.1, + "grad_norm": 0.404296875, + "learning_rate": 0.00019893758238351316, + "loss": 2.7845, + "step": 2275 + }, + { + "epoch": 0.1, + "grad_norm": 0.38671875, + "learning_rate": 0.0001989366069587262, + "loss": 2.6359, + "step": 2276 + }, + { + "epoch": 0.1, + "grad_norm": 0.412109375, + "learning_rate": 0.00019893563108876067, + "loss": 2.6872, + "step": 2277 + }, + { + "epoch": 0.1, + "grad_norm": 0.423828125, + "learning_rate": 0.00019893465477362096, + "loss": 2.8187, + "step": 2278 + }, + { + "epoch": 0.1, + "grad_norm": 0.462890625, + "learning_rate": 0.00019893367801331144, + "loss": 2.8102, + "step": 2279 + }, + { + "epoch": 0.1, + "grad_norm": 0.361328125, + "learning_rate": 0.00019893270080783654, + "loss": 2.6265, + "step": 2280 + }, + { + "epoch": 0.1, + "grad_norm": 0.453125, + "learning_rate": 0.00019893172315720064, + "loss": 2.6668, + "step": 2281 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.00019893074506140815, + "loss": 2.5925, + "step": 2282 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.00019892976652046345, + "loss": 2.5085, + "step": 2283 + }, + { + "epoch": 0.1, + "grad_norm": 0.451171875, + "learning_rate": 0.00019892878753437096, + "loss": 2.6991, + "step": 2284 + }, + { + "epoch": 0.1, + "grad_norm": 0.390625, + "learning_rate": 0.00019892780810313508, + "loss": 2.7375, + "step": 2285 + }, + { + "epoch": 0.1, + "grad_norm": 0.447265625, + "learning_rate": 0.00019892682822676022, + "loss": 2.7406, + "step": 2286 + }, + { + "epoch": 0.1, + "grad_norm": 0.365234375, + "learning_rate": 0.00019892584790525078, + "loss": 2.5705, + "step": 2287 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019892486713861117, + "loss": 2.7569, + "step": 2288 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.00019892388592684585, + "loss": 2.6293, + "step": 2289 + }, + { + "epoch": 0.1, + "grad_norm": 0.40625, + "learning_rate": 0.00019892290426995915, + "loss": 2.7556, + "step": 2290 + }, + { + "epoch": 0.1, + "grad_norm": 0.4296875, + "learning_rate": 0.00019892192216795556, + "loss": 2.838, + "step": 2291 + }, + { + "epoch": 0.1, + "grad_norm": 0.375, + "learning_rate": 0.00019892093962083947, + "loss": 2.7534, + "step": 2292 + }, + { + "epoch": 0.1, + "grad_norm": 0.43359375, + "learning_rate": 0.00019891995662861528, + "loss": 2.7034, + "step": 2293 + }, + { + "epoch": 0.1, + "grad_norm": 0.41796875, + "learning_rate": 0.00019891897319128745, + "loss": 2.7062, + "step": 2294 + }, + { + "epoch": 0.1, + "grad_norm": 0.41015625, + "learning_rate": 0.0001989179893088604, + "loss": 2.8704, + "step": 2295 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.00019891700498133856, + "loss": 2.8304, + "step": 2296 + }, + { + "epoch": 0.1, + "grad_norm": 0.421875, + "learning_rate": 0.0001989160202087263, + "loss": 2.7132, + "step": 2297 + }, + { + "epoch": 0.1, + "grad_norm": 0.41796875, + "learning_rate": 0.00019891503499102814, + "loss": 2.7335, + "step": 2298 + }, + { + "epoch": 0.1, + "grad_norm": 0.3828125, + "learning_rate": 0.00019891404932824848, + "loss": 2.722, + "step": 2299 + }, + { + "epoch": 0.1, + "grad_norm": 0.4375, + "learning_rate": 0.00019891306322039173, + "loss": 2.7242, + "step": 2300 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.00019891207666746232, + "loss": 2.8608, + "step": 2301 + }, + { + "epoch": 0.1, + "grad_norm": 0.4375, + "learning_rate": 0.00019891108966946475, + "loss": 2.7995, + "step": 2302 + }, + { + "epoch": 0.1, + "grad_norm": 0.36328125, + "learning_rate": 0.0001989101022264034, + "loss": 2.7988, + "step": 2303 + }, + { + "epoch": 0.1, + "grad_norm": 0.359375, + "learning_rate": 0.00019890911433828275, + "loss": 2.7075, + "step": 2304 + }, + { + "epoch": 0.1, + "grad_norm": 0.359375, + "learning_rate": 0.00019890812600510725, + "loss": 2.6097, + "step": 2305 + }, + { + "epoch": 0.1, + "grad_norm": 0.41015625, + "learning_rate": 0.0001989071372268813, + "loss": 2.7985, + "step": 2306 + }, + { + "epoch": 0.1, + "grad_norm": 0.419921875, + "learning_rate": 0.0001989061480036094, + "loss": 2.6498, + "step": 2307 + }, + { + "epoch": 0.1, + "grad_norm": 1.203125, + "learning_rate": 0.00019890515833529595, + "loss": 2.8615, + "step": 2308 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.00019890416822194545, + "loss": 2.5379, + "step": 2309 + }, + { + "epoch": 0.1, + "grad_norm": 0.40625, + "learning_rate": 0.00019890317766356233, + "loss": 2.5383, + "step": 2310 + }, + { + "epoch": 0.1, + "grad_norm": 0.373046875, + "learning_rate": 0.00019890218666015108, + "loss": 2.7216, + "step": 2311 + }, + { + "epoch": 0.1, + "grad_norm": 0.3515625, + "learning_rate": 0.00019890119521171612, + "loss": 2.589, + "step": 2312 + }, + { + "epoch": 0.1, + "grad_norm": 0.35546875, + "learning_rate": 0.00019890020331826192, + "loss": 2.5498, + "step": 2313 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.00019889921097979296, + "loss": 2.6513, + "step": 2314 + }, + { + "epoch": 0.1, + "grad_norm": 0.3984375, + "learning_rate": 0.00019889821819631367, + "loss": 2.619, + "step": 2315 + }, + { + "epoch": 0.1, + "grad_norm": 0.46484375, + "learning_rate": 0.00019889722496782856, + "loss": 2.7622, + "step": 2316 + }, + { + "epoch": 0.1, + "grad_norm": 0.38671875, + "learning_rate": 0.0001988962312943421, + "loss": 2.6946, + "step": 2317 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.00019889523717585872, + "loss": 2.845, + "step": 2318 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.00019889424261238293, + "loss": 2.6158, + "step": 2319 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.00019889324760391917, + "loss": 2.7806, + "step": 2320 + }, + { + "epoch": 0.1, + "grad_norm": 0.3828125, + "learning_rate": 0.00019889225215047194, + "loss": 2.5861, + "step": 2321 + }, + { + "epoch": 0.1, + "grad_norm": 0.40234375, + "learning_rate": 0.0001988912562520457, + "loss": 2.7272, + "step": 2322 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.000198890259908645, + "loss": 2.5689, + "step": 2323 + }, + { + "epoch": 0.1, + "grad_norm": 0.37890625, + "learning_rate": 0.00019888926312027424, + "loss": 2.6623, + "step": 2324 + }, + { + "epoch": 0.1, + "grad_norm": 0.44140625, + "learning_rate": 0.0001988882658869379, + "loss": 2.7539, + "step": 2325 + }, + { + "epoch": 0.1, + "grad_norm": 0.43359375, + "learning_rate": 0.00019888726820864054, + "loss": 2.6211, + "step": 2326 + }, + { + "epoch": 0.1, + "grad_norm": 0.396484375, + "learning_rate": 0.00019888627008538662, + "loss": 2.6565, + "step": 2327 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.0001988852715171806, + "loss": 2.6793, + "step": 2328 + }, + { + "epoch": 0.1, + "grad_norm": 0.369140625, + "learning_rate": 0.00019888427250402702, + "loss": 2.7492, + "step": 2329 + }, + { + "epoch": 0.1, + "grad_norm": 0.4296875, + "learning_rate": 0.00019888327304593036, + "loss": 2.4699, + "step": 2330 + }, + { + "epoch": 0.1, + "grad_norm": 0.4140625, + "learning_rate": 0.00019888227314289507, + "loss": 2.6757, + "step": 2331 + }, + { + "epoch": 0.1, + "grad_norm": 0.365234375, + "learning_rate": 0.0001988812727949257, + "loss": 2.7099, + "step": 2332 + }, + { + "epoch": 0.1, + "grad_norm": 0.396484375, + "learning_rate": 0.00019888027200202677, + "loss": 2.7002, + "step": 2333 + }, + { + "epoch": 0.1, + "grad_norm": 0.3671875, + "learning_rate": 0.00019887927076420273, + "loss": 2.7419, + "step": 2334 + }, + { + "epoch": 0.1, + "grad_norm": 0.4375, + "learning_rate": 0.0001988782690814581, + "loss": 2.6826, + "step": 2335 + }, + { + "epoch": 0.1, + "grad_norm": 0.37109375, + "learning_rate": 0.0001988772669537974, + "loss": 2.6547, + "step": 2336 + }, + { + "epoch": 0.1, + "grad_norm": 0.427734375, + "learning_rate": 0.00019887626438122515, + "loss": 2.8023, + "step": 2337 + }, + { + "epoch": 0.1, + "grad_norm": 0.39453125, + "learning_rate": 0.00019887526136374583, + "loss": 2.5637, + "step": 2338 + }, + { + "epoch": 0.1, + "grad_norm": 0.40625, + "learning_rate": 0.00019887425790136395, + "loss": 2.7154, + "step": 2339 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.00019887325399408408, + "loss": 2.6879, + "step": 2340 + }, + { + "epoch": 0.1, + "grad_norm": 0.396484375, + "learning_rate": 0.0001988722496419107, + "loss": 2.6473, + "step": 2341 + }, + { + "epoch": 0.1, + "grad_norm": 0.419921875, + "learning_rate": 0.00019887124484484832, + "loss": 2.5116, + "step": 2342 + }, + { + "epoch": 0.1, + "grad_norm": 0.380859375, + "learning_rate": 0.00019887023960290146, + "loss": 2.6078, + "step": 2343 + }, + { + "epoch": 0.1, + "grad_norm": 0.4453125, + "learning_rate": 0.00019886923391607467, + "loss": 2.7034, + "step": 2344 + }, + { + "epoch": 0.1, + "grad_norm": 0.4140625, + "learning_rate": 0.00019886822778437244, + "loss": 2.7569, + "step": 2345 + }, + { + "epoch": 0.1, + "grad_norm": 0.37109375, + "learning_rate": 0.00019886722120779932, + "loss": 2.7067, + "step": 2346 + }, + { + "epoch": 0.1, + "grad_norm": 0.396484375, + "learning_rate": 0.00019886621418635985, + "loss": 2.7191, + "step": 2347 + }, + { + "epoch": 0.1, + "grad_norm": 0.404296875, + "learning_rate": 0.00019886520672005858, + "loss": 2.7166, + "step": 2348 + }, + { + "epoch": 0.1, + "grad_norm": 0.40234375, + "learning_rate": 0.00019886419880889994, + "loss": 2.6077, + "step": 2349 + }, + { + "epoch": 0.1, + "grad_norm": 0.376953125, + "learning_rate": 0.0001988631904528886, + "loss": 2.6114, + "step": 2350 + }, + { + "epoch": 0.1, + "grad_norm": 0.40234375, + "learning_rate": 0.00019886218165202898, + "loss": 2.7485, + "step": 2351 + }, + { + "epoch": 0.1, + "grad_norm": 0.408203125, + "learning_rate": 0.00019886117240632573, + "loss": 2.7598, + "step": 2352 + }, + { + "epoch": 0.1, + "grad_norm": 0.375, + "learning_rate": 0.0001988601627157833, + "loss": 2.5854, + "step": 2353 + }, + { + "epoch": 0.1, + "grad_norm": 0.451171875, + "learning_rate": 0.00019885915258040625, + "loss": 2.7968, + "step": 2354 + }, + { + "epoch": 0.1, + "grad_norm": 0.4296875, + "learning_rate": 0.00019885814200019914, + "loss": 2.7571, + "step": 2355 + }, + { + "epoch": 0.1, + "grad_norm": 0.5, + "learning_rate": 0.00019885713097516657, + "loss": 2.8386, + "step": 2356 + }, + { + "epoch": 0.1, + "grad_norm": 0.439453125, + "learning_rate": 0.000198856119505313, + "loss": 2.7939, + "step": 2357 + }, + { + "epoch": 0.1, + "grad_norm": 0.416015625, + "learning_rate": 0.00019885510759064303, + "loss": 2.7388, + "step": 2358 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.0001988540952311612, + "loss": 2.7481, + "step": 2359 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019885308242687208, + "loss": 2.756, + "step": 2360 + }, + { + "epoch": 0.1, + "grad_norm": 0.419921875, + "learning_rate": 0.00019885206917778022, + "loss": 2.6143, + "step": 2361 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019885105548389013, + "loss": 2.7639, + "step": 2362 + }, + { + "epoch": 0.1, + "grad_norm": 0.376953125, + "learning_rate": 0.00019885004134520643, + "loss": 2.7689, + "step": 2363 + }, + { + "epoch": 0.1, + "grad_norm": 0.380859375, + "learning_rate": 0.00019884902676173372, + "loss": 2.6079, + "step": 2364 + }, + { + "epoch": 0.1, + "grad_norm": 0.3828125, + "learning_rate": 0.00019884801173347647, + "loss": 2.7027, + "step": 2365 + }, + { + "epoch": 0.1, + "grad_norm": 0.419921875, + "learning_rate": 0.0001988469962604393, + "loss": 2.7983, + "step": 2366 + }, + { + "epoch": 0.1, + "grad_norm": 0.3828125, + "learning_rate": 0.00019884598034262678, + "loss": 2.8337, + "step": 2367 + }, + { + "epoch": 0.1, + "grad_norm": 0.41015625, + "learning_rate": 0.00019884496398004346, + "loss": 2.6474, + "step": 2368 + }, + { + "epoch": 0.1, + "grad_norm": 0.41796875, + "learning_rate": 0.00019884394717269394, + "loss": 2.7561, + "step": 2369 + }, + { + "epoch": 0.1, + "grad_norm": 0.3515625, + "learning_rate": 0.00019884292992058278, + "loss": 2.6344, + "step": 2370 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.00019884191222371454, + "loss": 2.7536, + "step": 2371 + }, + { + "epoch": 0.1, + "grad_norm": 0.38671875, + "learning_rate": 0.00019884089408209383, + "loss": 2.6735, + "step": 2372 + }, + { + "epoch": 0.1, + "grad_norm": 0.41015625, + "learning_rate": 0.00019883987549572523, + "loss": 2.6323, + "step": 2373 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.0001988388564646133, + "loss": 2.6236, + "step": 2374 + }, + { + "epoch": 0.1, + "grad_norm": 0.40625, + "learning_rate": 0.00019883783698876263, + "loss": 2.6989, + "step": 2375 + }, + { + "epoch": 0.1, + "grad_norm": 0.453125, + "learning_rate": 0.00019883681706817782, + "loss": 2.6953, + "step": 2376 + }, + { + "epoch": 0.1, + "grad_norm": 0.396484375, + "learning_rate": 0.00019883579670286347, + "loss": 2.6898, + "step": 2377 + }, + { + "epoch": 0.1, + "grad_norm": 0.3828125, + "learning_rate": 0.00019883477589282414, + "loss": 2.6486, + "step": 2378 + }, + { + "epoch": 0.1, + "grad_norm": 0.458984375, + "learning_rate": 0.00019883375463806445, + "loss": 2.8596, + "step": 2379 + }, + { + "epoch": 0.1, + "grad_norm": 0.4453125, + "learning_rate": 0.00019883273293858897, + "loss": 2.7548, + "step": 2380 + }, + { + "epoch": 0.1, + "grad_norm": 0.431640625, + "learning_rate": 0.0001988317107944023, + "loss": 2.6779, + "step": 2381 + }, + { + "epoch": 0.1, + "grad_norm": 0.3984375, + "learning_rate": 0.00019883068820550907, + "loss": 2.7033, + "step": 2382 + }, + { + "epoch": 0.1, + "grad_norm": 0.423828125, + "learning_rate": 0.0001988296651719139, + "loss": 2.6886, + "step": 2383 + }, + { + "epoch": 0.1, + "grad_norm": 0.423828125, + "learning_rate": 0.00019882864169362127, + "loss": 2.7754, + "step": 2384 + }, + { + "epoch": 0.1, + "grad_norm": 0.443359375, + "learning_rate": 0.00019882761777063592, + "loss": 2.7173, + "step": 2385 + }, + { + "epoch": 0.1, + "grad_norm": 0.384765625, + "learning_rate": 0.0001988265934029624, + "loss": 2.6597, + "step": 2386 + }, + { + "epoch": 0.1, + "grad_norm": 0.470703125, + "learning_rate": 0.00019882556859060534, + "loss": 2.6758, + "step": 2387 + }, + { + "epoch": 0.1, + "grad_norm": 0.37109375, + "learning_rate": 0.00019882454333356933, + "loss": 2.5756, + "step": 2388 + }, + { + "epoch": 0.1, + "grad_norm": 0.44140625, + "learning_rate": 0.00019882351763185897, + "loss": 2.7098, + "step": 2389 + }, + { + "epoch": 0.1, + "grad_norm": 0.396484375, + "learning_rate": 0.00019882249148547892, + "loss": 2.6662, + "step": 2390 + }, + { + "epoch": 0.1, + "grad_norm": 0.51171875, + "learning_rate": 0.00019882146489443375, + "loss": 2.7183, + "step": 2391 + }, + { + "epoch": 0.1, + "grad_norm": 0.470703125, + "learning_rate": 0.0001988204378587281, + "loss": 2.6972, + "step": 2392 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.00019881941037836663, + "loss": 2.6549, + "step": 2393 + }, + { + "epoch": 0.1, + "grad_norm": 0.408203125, + "learning_rate": 0.00019881838245335392, + "loss": 2.6766, + "step": 2394 + }, + { + "epoch": 0.1, + "grad_norm": 0.39453125, + "learning_rate": 0.0001988173540836946, + "loss": 2.6141, + "step": 2395 + }, + { + "epoch": 0.1, + "grad_norm": 0.41796875, + "learning_rate": 0.00019881632526939328, + "loss": 2.7191, + "step": 2396 + }, + { + "epoch": 0.1, + "grad_norm": 0.451171875, + "learning_rate": 0.00019881529601045465, + "loss": 2.6084, + "step": 2397 + }, + { + "epoch": 0.1, + "grad_norm": 0.3515625, + "learning_rate": 0.00019881426630688327, + "loss": 2.5593, + "step": 2398 + }, + { + "epoch": 0.1, + "grad_norm": 0.458984375, + "learning_rate": 0.0001988132361586838, + "loss": 2.6459, + "step": 2399 + }, + { + "epoch": 0.1, + "grad_norm": 0.408203125, + "learning_rate": 0.00019881220556586086, + "loss": 2.7439, + "step": 2400 + }, + { + "epoch": 0.1, + "grad_norm": 0.39453125, + "learning_rate": 0.00019881117452841916, + "loss": 2.7342, + "step": 2401 + }, + { + "epoch": 0.1, + "grad_norm": 0.447265625, + "learning_rate": 0.00019881014304636325, + "loss": 2.7095, + "step": 2402 + }, + { + "epoch": 0.1, + "grad_norm": 0.4375, + "learning_rate": 0.00019880911111969782, + "loss": 2.7905, + "step": 2403 + }, + { + "epoch": 0.1, + "grad_norm": 0.412109375, + "learning_rate": 0.0001988080787484275, + "loss": 2.7308, + "step": 2404 + }, + { + "epoch": 0.1, + "grad_norm": 0.396484375, + "learning_rate": 0.00019880704593255692, + "loss": 2.6909, + "step": 2405 + }, + { + "epoch": 0.1, + "grad_norm": 0.404296875, + "learning_rate": 0.00019880601267209077, + "loss": 2.6249, + "step": 2406 + }, + { + "epoch": 0.1, + "grad_norm": 0.396484375, + "learning_rate": 0.00019880497896703365, + "loss": 2.656, + "step": 2407 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.00019880394481739026, + "loss": 2.571, + "step": 2408 + }, + { + "epoch": 0.1, + "grad_norm": 0.41015625, + "learning_rate": 0.00019880291022316522, + "loss": 2.6413, + "step": 2409 + }, + { + "epoch": 0.1, + "grad_norm": 0.40625, + "learning_rate": 0.00019880187518436318, + "loss": 2.5251, + "step": 2410 + }, + { + "epoch": 0.1, + "grad_norm": 0.380859375, + "learning_rate": 0.0001988008397009888, + "loss": 2.6447, + "step": 2411 + }, + { + "epoch": 0.1, + "grad_norm": 0.41015625, + "learning_rate": 0.00019879980377304676, + "loss": 2.7215, + "step": 2412 + }, + { + "epoch": 0.1, + "grad_norm": 0.44140625, + "learning_rate": 0.00019879876740054174, + "loss": 2.7568, + "step": 2413 + }, + { + "epoch": 0.1, + "grad_norm": 0.427734375, + "learning_rate": 0.00019879773058347833, + "loss": 2.7076, + "step": 2414 + }, + { + "epoch": 0.1, + "grad_norm": 0.361328125, + "learning_rate": 0.00019879669332186124, + "loss": 2.4794, + "step": 2415 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.00019879565561569515, + "loss": 2.6013, + "step": 2416 + }, + { + "epoch": 0.1, + "grad_norm": 0.44140625, + "learning_rate": 0.0001987946174649847, + "loss": 2.7349, + "step": 2417 + }, + { + "epoch": 0.1, + "grad_norm": 0.41796875, + "learning_rate": 0.00019879357886973462, + "loss": 2.6624, + "step": 2418 + }, + { + "epoch": 0.1, + "grad_norm": 0.474609375, + "learning_rate": 0.0001987925398299495, + "loss": 2.8332, + "step": 2419 + }, + { + "epoch": 0.1, + "grad_norm": 0.45703125, + "learning_rate": 0.00019879150034563407, + "loss": 2.666, + "step": 2420 + }, + { + "epoch": 0.1, + "grad_norm": 0.40625, + "learning_rate": 0.000198790460416793, + "loss": 2.7636, + "step": 2421 + }, + { + "epoch": 0.1, + "grad_norm": 0.423828125, + "learning_rate": 0.00019878942004343094, + "loss": 2.7943, + "step": 2422 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019878837922555263, + "loss": 2.7567, + "step": 2423 + }, + { + "epoch": 0.1, + "grad_norm": 0.421875, + "learning_rate": 0.00019878733796316266, + "loss": 2.5841, + "step": 2424 + }, + { + "epoch": 0.1, + "grad_norm": 0.375, + "learning_rate": 0.00019878629625626583, + "loss": 2.6013, + "step": 2425 + }, + { + "epoch": 0.1, + "grad_norm": 0.412109375, + "learning_rate": 0.00019878525410486675, + "loss": 2.75, + "step": 2426 + }, + { + "epoch": 0.1, + "grad_norm": 0.373046875, + "learning_rate": 0.00019878421150897011, + "loss": 2.6761, + "step": 2427 + }, + { + "epoch": 0.1, + "grad_norm": 0.384765625, + "learning_rate": 0.00019878316846858062, + "loss": 2.6013, + "step": 2428 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.000198782124983703, + "loss": 2.7106, + "step": 2429 + }, + { + "epoch": 0.1, + "grad_norm": 0.384765625, + "learning_rate": 0.00019878108105434187, + "loss": 2.7353, + "step": 2430 + }, + { + "epoch": 0.1, + "grad_norm": 0.4375, + "learning_rate": 0.00019878003668050202, + "loss": 2.622, + "step": 2431 + }, + { + "epoch": 0.1, + "grad_norm": 0.41796875, + "learning_rate": 0.00019877899186218813, + "loss": 2.7633, + "step": 2432 + }, + { + "epoch": 0.1, + "grad_norm": 0.40234375, + "learning_rate": 0.00019877794659940482, + "loss": 2.5699, + "step": 2433 + }, + { + "epoch": 0.1, + "grad_norm": 0.345703125, + "learning_rate": 0.00019877690089215686, + "loss": 2.7472, + "step": 2434 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019877585474044896, + "loss": 2.682, + "step": 2435 + }, + { + "epoch": 0.1, + "grad_norm": 0.4140625, + "learning_rate": 0.0001987748081442858, + "loss": 2.623, + "step": 2436 + }, + { + "epoch": 0.1, + "grad_norm": 0.359375, + "learning_rate": 0.0001987737611036721, + "loss": 2.8066, + "step": 2437 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019877271361861258, + "loss": 2.7303, + "step": 2438 + }, + { + "epoch": 0.1, + "grad_norm": 0.43359375, + "learning_rate": 0.00019877166568911197, + "loss": 2.6658, + "step": 2439 + }, + { + "epoch": 0.1, + "grad_norm": 0.41015625, + "learning_rate": 0.0001987706173151749, + "loss": 2.6867, + "step": 2440 + }, + { + "epoch": 0.1, + "grad_norm": 0.392578125, + "learning_rate": 0.0001987695684968062, + "loss": 2.7179, + "step": 2441 + }, + { + "epoch": 0.1, + "grad_norm": 0.3984375, + "learning_rate": 0.00019876851923401052, + "loss": 2.7579, + "step": 2442 + }, + { + "epoch": 0.1, + "grad_norm": 0.380859375, + "learning_rate": 0.0001987674695267926, + "loss": 2.7397, + "step": 2443 + }, + { + "epoch": 0.1, + "grad_norm": 0.42578125, + "learning_rate": 0.00019876641937515717, + "loss": 2.6844, + "step": 2444 + }, + { + "epoch": 0.1, + "grad_norm": 0.427734375, + "learning_rate": 0.00019876536877910895, + "loss": 2.6934, + "step": 2445 + }, + { + "epoch": 0.1, + "grad_norm": 0.4453125, + "learning_rate": 0.00019876431773865267, + "loss": 2.6584, + "step": 2446 + }, + { + "epoch": 0.1, + "grad_norm": 0.388671875, + "learning_rate": 0.00019876326625379304, + "loss": 2.6311, + "step": 2447 + }, + { + "epoch": 0.1, + "grad_norm": 0.404296875, + "learning_rate": 0.00019876221432453482, + "loss": 2.657, + "step": 2448 + }, + { + "epoch": 0.1, + "grad_norm": 0.40625, + "learning_rate": 0.0001987611619508827, + "loss": 2.8143, + "step": 2449 + }, + { + "epoch": 0.1, + "grad_norm": 0.3984375, + "learning_rate": 0.00019876010913284144, + "loss": 2.6643, + "step": 2450 + }, + { + "epoch": 0.1, + "grad_norm": 0.400390625, + "learning_rate": 0.0001987590558704158, + "loss": 2.6, + "step": 2451 + }, + { + "epoch": 0.1, + "grad_norm": 0.439453125, + "learning_rate": 0.00019875800216361054, + "loss": 2.6422, + "step": 2452 + }, + { + "epoch": 0.1, + "grad_norm": 0.435546875, + "learning_rate": 0.00019875694801243032, + "loss": 2.6621, + "step": 2453 + }, + { + "epoch": 0.1, + "grad_norm": 0.4453125, + "learning_rate": 0.00019875589341687992, + "loss": 2.5829, + "step": 2454 + }, + { + "epoch": 0.1, + "grad_norm": 0.408203125, + "learning_rate": 0.00019875483837696409, + "loss": 2.5415, + "step": 2455 + }, + { + "epoch": 0.1, + "grad_norm": 0.365234375, + "learning_rate": 0.0001987537828926876, + "loss": 2.5639, + "step": 2456 + }, + { + "epoch": 0.1, + "grad_norm": 0.3828125, + "learning_rate": 0.00019875272696405518, + "loss": 2.6948, + "step": 2457 + }, + { + "epoch": 0.1, + "grad_norm": 0.380859375, + "learning_rate": 0.00019875167059107157, + "loss": 2.6483, + "step": 2458 + }, + { + "epoch": 0.1, + "grad_norm": 0.439453125, + "learning_rate": 0.0001987506137737415, + "loss": 2.6463, + "step": 2459 + }, + { + "epoch": 0.1, + "grad_norm": 0.359375, + "learning_rate": 0.0001987495565120698, + "loss": 2.6362, + "step": 2460 + }, + { + "epoch": 0.1, + "grad_norm": 0.3984375, + "learning_rate": 0.00019874849880606118, + "loss": 2.8176, + "step": 2461 + }, + { + "epoch": 0.1, + "grad_norm": 0.43359375, + "learning_rate": 0.00019874744065572037, + "loss": 2.806, + "step": 2462 + }, + { + "epoch": 0.1, + "grad_norm": 0.435546875, + "learning_rate": 0.0001987463820610522, + "loss": 2.7078, + "step": 2463 + }, + { + "epoch": 0.1, + "grad_norm": 0.421875, + "learning_rate": 0.0001987453230220614, + "loss": 2.7103, + "step": 2464 + }, + { + "epoch": 0.11, + "grad_norm": 0.400390625, + "learning_rate": 0.00019874426353875272, + "loss": 2.5555, + "step": 2465 + }, + { + "epoch": 0.11, + "grad_norm": 0.396484375, + "learning_rate": 0.00019874320361113092, + "loss": 2.6686, + "step": 2466 + }, + { + "epoch": 0.11, + "grad_norm": 0.4296875, + "learning_rate": 0.00019874214323920084, + "loss": 2.7108, + "step": 2467 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.00019874108242296716, + "loss": 2.7913, + "step": 2468 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019874002116243472, + "loss": 2.6619, + "step": 2469 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.00019873895945760826, + "loss": 2.8024, + "step": 2470 + }, + { + "epoch": 0.11, + "grad_norm": 0.44140625, + "learning_rate": 0.00019873789730849258, + "loss": 2.7898, + "step": 2471 + }, + { + "epoch": 0.11, + "grad_norm": 0.396484375, + "learning_rate": 0.0001987368347150924, + "loss": 2.6745, + "step": 2472 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.0001987357716774126, + "loss": 2.7127, + "step": 2473 + }, + { + "epoch": 0.11, + "grad_norm": 0.44921875, + "learning_rate": 0.0001987347081954579, + "loss": 2.6472, + "step": 2474 + }, + { + "epoch": 0.11, + "grad_norm": 0.50390625, + "learning_rate": 0.00019873364426923308, + "loss": 2.8777, + "step": 2475 + }, + { + "epoch": 0.11, + "grad_norm": 0.40625, + "learning_rate": 0.00019873257989874295, + "loss": 2.6904, + "step": 2476 + }, + { + "epoch": 0.11, + "grad_norm": 0.400390625, + "learning_rate": 0.0001987315150839923, + "loss": 2.4623, + "step": 2477 + }, + { + "epoch": 0.11, + "grad_norm": 0.37890625, + "learning_rate": 0.00019873044982498592, + "loss": 2.6593, + "step": 2478 + }, + { + "epoch": 0.11, + "grad_norm": 0.3828125, + "learning_rate": 0.00019872938412172857, + "loss": 2.6812, + "step": 2479 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.00019872831797422511, + "loss": 2.5949, + "step": 2480 + }, + { + "epoch": 0.11, + "grad_norm": 0.384765625, + "learning_rate": 0.00019872725138248029, + "loss": 2.7727, + "step": 2481 + }, + { + "epoch": 0.11, + "grad_norm": 0.447265625, + "learning_rate": 0.00019872618434649888, + "loss": 2.6384, + "step": 2482 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.00019872511686628577, + "loss": 2.6945, + "step": 2483 + }, + { + "epoch": 0.11, + "grad_norm": 0.38671875, + "learning_rate": 0.0001987240489418457, + "loss": 2.6739, + "step": 2484 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019872298057318347, + "loss": 2.7062, + "step": 2485 + }, + { + "epoch": 0.11, + "grad_norm": 0.419921875, + "learning_rate": 0.0001987219117603039, + "loss": 2.6956, + "step": 2486 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.00019872084250321183, + "loss": 2.6723, + "step": 2487 + }, + { + "epoch": 0.11, + "grad_norm": 0.36328125, + "learning_rate": 0.00019871977280191204, + "loss": 2.7041, + "step": 2488 + }, + { + "epoch": 0.11, + "grad_norm": 0.4375, + "learning_rate": 0.00019871870265640937, + "loss": 2.6491, + "step": 2489 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.00019871763206670863, + "loss": 2.6903, + "step": 2490 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.00019871656103281457, + "loss": 2.6155, + "step": 2491 + }, + { + "epoch": 0.11, + "grad_norm": 0.4140625, + "learning_rate": 0.00019871548955473206, + "loss": 2.7648, + "step": 2492 + }, + { + "epoch": 0.11, + "grad_norm": 0.42578125, + "learning_rate": 0.00019871441763246597, + "loss": 2.6031, + "step": 2493 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.00019871334526602103, + "loss": 2.6602, + "step": 2494 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.00019871227245540216, + "loss": 2.7286, + "step": 2495 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.0001987111992006141, + "loss": 2.666, + "step": 2496 + }, + { + "epoch": 0.11, + "grad_norm": 0.326171875, + "learning_rate": 0.00019871012550166174, + "loss": 2.5162, + "step": 2497 + }, + { + "epoch": 0.11, + "grad_norm": 0.365234375, + "learning_rate": 0.00019870905135854988, + "loss": 2.5072, + "step": 2498 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.00019870797677128338, + "loss": 2.7281, + "step": 2499 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.00019870690173986707, + "loss": 2.7208, + "step": 2500 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.00019870582626430573, + "loss": 2.799, + "step": 2501 + }, + { + "epoch": 0.11, + "grad_norm": 0.40625, + "learning_rate": 0.00019870475034460428, + "loss": 2.7017, + "step": 2502 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.0001987036739807675, + "loss": 2.728, + "step": 2503 + }, + { + "epoch": 0.11, + "grad_norm": 0.3984375, + "learning_rate": 0.00019870259717280026, + "loss": 2.6669, + "step": 2504 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.0001987015199207074, + "loss": 2.7544, + "step": 2505 + }, + { + "epoch": 0.11, + "grad_norm": 0.44140625, + "learning_rate": 0.0001987004422244938, + "loss": 2.7237, + "step": 2506 + }, + { + "epoch": 0.11, + "grad_norm": 0.384765625, + "learning_rate": 0.00019869936408416426, + "loss": 2.591, + "step": 2507 + }, + { + "epoch": 0.11, + "grad_norm": 0.40234375, + "learning_rate": 0.00019869828549972366, + "loss": 2.6671, + "step": 2508 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.00019869720647117686, + "loss": 2.7133, + "step": 2509 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.00019869612699852866, + "loss": 2.7591, + "step": 2510 + }, + { + "epoch": 0.11, + "grad_norm": 0.419921875, + "learning_rate": 0.000198695047081784, + "loss": 2.7182, + "step": 2511 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019869396672094765, + "loss": 2.638, + "step": 2512 + }, + { + "epoch": 0.11, + "grad_norm": 0.431640625, + "learning_rate": 0.00019869288591602458, + "loss": 2.6678, + "step": 2513 + }, + { + "epoch": 0.11, + "grad_norm": 0.419921875, + "learning_rate": 0.00019869180466701957, + "loss": 2.6697, + "step": 2514 + }, + { + "epoch": 0.11, + "grad_norm": 0.37109375, + "learning_rate": 0.00019869072297393748, + "loss": 2.568, + "step": 2515 + }, + { + "epoch": 0.11, + "grad_norm": 0.3671875, + "learning_rate": 0.00019868964083678324, + "loss": 2.6391, + "step": 2516 + }, + { + "epoch": 0.11, + "grad_norm": 0.36328125, + "learning_rate": 0.00019868855825556166, + "loss": 2.7151, + "step": 2517 + }, + { + "epoch": 0.11, + "grad_norm": 0.408203125, + "learning_rate": 0.00019868747523027762, + "loss": 2.7086, + "step": 2518 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.00019868639176093603, + "loss": 2.6576, + "step": 2519 + }, + { + "epoch": 0.11, + "grad_norm": 0.37890625, + "learning_rate": 0.00019868530784754174, + "loss": 2.6882, + "step": 2520 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.00019868422349009963, + "loss": 2.6955, + "step": 2521 + }, + { + "epoch": 0.11, + "grad_norm": 0.4765625, + "learning_rate": 0.00019868313868861458, + "loss": 2.6903, + "step": 2522 + }, + { + "epoch": 0.11, + "grad_norm": 0.41796875, + "learning_rate": 0.00019868205344309144, + "loss": 2.7559, + "step": 2523 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.00019868096775353517, + "loss": 2.6972, + "step": 2524 + }, + { + "epoch": 0.11, + "grad_norm": 0.49609375, + "learning_rate": 0.0001986798816199506, + "loss": 2.7424, + "step": 2525 + }, + { + "epoch": 0.11, + "grad_norm": 0.404296875, + "learning_rate": 0.00019867879504234263, + "loss": 2.7598, + "step": 2526 + }, + { + "epoch": 0.11, + "grad_norm": 0.359375, + "learning_rate": 0.0001986777080207161, + "loss": 2.5573, + "step": 2527 + }, + { + "epoch": 0.11, + "grad_norm": 0.4296875, + "learning_rate": 0.000198676620555076, + "loss": 2.5902, + "step": 2528 + }, + { + "epoch": 0.11, + "grad_norm": 0.435546875, + "learning_rate": 0.00019867553264542713, + "loss": 2.6598, + "step": 2529 + }, + { + "epoch": 0.11, + "grad_norm": 0.435546875, + "learning_rate": 0.00019867444429177446, + "loss": 2.7654, + "step": 2530 + }, + { + "epoch": 0.11, + "grad_norm": 0.3984375, + "learning_rate": 0.00019867335549412281, + "loss": 2.7361, + "step": 2531 + }, + { + "epoch": 0.11, + "grad_norm": 0.375, + "learning_rate": 0.00019867226625247718, + "loss": 2.5266, + "step": 2532 + }, + { + "epoch": 0.11, + "grad_norm": 0.404296875, + "learning_rate": 0.00019867117656684235, + "loss": 2.6694, + "step": 2533 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.00019867008643722334, + "loss": 2.677, + "step": 2534 + }, + { + "epoch": 0.11, + "grad_norm": 0.43359375, + "learning_rate": 0.00019866899586362498, + "loss": 2.8155, + "step": 2535 + }, + { + "epoch": 0.11, + "grad_norm": 0.462890625, + "learning_rate": 0.00019866790484605217, + "loss": 2.6736, + "step": 2536 + }, + { + "epoch": 0.11, + "grad_norm": 0.375, + "learning_rate": 0.00019866681338450988, + "loss": 2.5678, + "step": 2537 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.00019866572147900297, + "loss": 2.7559, + "step": 2538 + }, + { + "epoch": 0.11, + "grad_norm": 0.439453125, + "learning_rate": 0.00019866462912953637, + "loss": 2.6822, + "step": 2539 + }, + { + "epoch": 0.11, + "grad_norm": 0.419921875, + "learning_rate": 0.00019866353633611502, + "loss": 2.6452, + "step": 2540 + }, + { + "epoch": 0.11, + "grad_norm": 0.404296875, + "learning_rate": 0.00019866244309874382, + "loss": 2.7778, + "step": 2541 + }, + { + "epoch": 0.11, + "grad_norm": 0.404296875, + "learning_rate": 0.00019866134941742765, + "loss": 2.6448, + "step": 2542 + }, + { + "epoch": 0.11, + "grad_norm": 0.392578125, + "learning_rate": 0.00019866025529217146, + "loss": 2.6128, + "step": 2543 + }, + { + "epoch": 0.11, + "grad_norm": 0.4453125, + "learning_rate": 0.00019865916072298024, + "loss": 2.7295, + "step": 2544 + }, + { + "epoch": 0.11, + "grad_norm": 0.361328125, + "learning_rate": 0.0001986580657098588, + "loss": 2.4968, + "step": 2545 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.0001986569702528121, + "loss": 2.5818, + "step": 2546 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.00019865587435184513, + "loss": 2.7503, + "step": 2547 + }, + { + "epoch": 0.11, + "grad_norm": 0.333984375, + "learning_rate": 0.00019865477800696274, + "loss": 2.6887, + "step": 2548 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019865368121816995, + "loss": 2.7089, + "step": 2549 + }, + { + "epoch": 0.11, + "grad_norm": 0.404296875, + "learning_rate": 0.00019865258398547164, + "loss": 2.6718, + "step": 2550 + }, + { + "epoch": 0.11, + "grad_norm": 0.40234375, + "learning_rate": 0.00019865148630887275, + "loss": 2.7463, + "step": 2551 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.0001986503881883782, + "loss": 2.7328, + "step": 2552 + }, + { + "epoch": 0.11, + "grad_norm": 0.41796875, + "learning_rate": 0.00019864928962399296, + "loss": 2.6267, + "step": 2553 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.000198648190615722, + "loss": 2.7145, + "step": 2554 + }, + { + "epoch": 0.11, + "grad_norm": 0.431640625, + "learning_rate": 0.00019864709116357022, + "loss": 2.7661, + "step": 2555 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.00019864599126754258, + "loss": 2.673, + "step": 2556 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.00019864489092764401, + "loss": 2.6418, + "step": 2557 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019864379014387953, + "loss": 2.5912, + "step": 2558 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.000198642688916254, + "loss": 2.752, + "step": 2559 + }, + { + "epoch": 0.11, + "grad_norm": 0.435546875, + "learning_rate": 0.00019864158724477245, + "loss": 2.6783, + "step": 2560 + }, + { + "epoch": 0.11, + "grad_norm": 0.439453125, + "learning_rate": 0.00019864048512943978, + "loss": 2.612, + "step": 2561 + }, + { + "epoch": 0.11, + "grad_norm": 0.435546875, + "learning_rate": 0.00019863938257026097, + "loss": 2.7496, + "step": 2562 + }, + { + "epoch": 0.11, + "grad_norm": 0.48046875, + "learning_rate": 0.00019863827956724103, + "loss": 2.5761, + "step": 2563 + }, + { + "epoch": 0.11, + "grad_norm": 0.45703125, + "learning_rate": 0.00019863717612038484, + "loss": 2.7349, + "step": 2564 + }, + { + "epoch": 0.11, + "grad_norm": 0.42578125, + "learning_rate": 0.00019863607222969743, + "loss": 2.7253, + "step": 2565 + }, + { + "epoch": 0.11, + "grad_norm": 0.48828125, + "learning_rate": 0.00019863496789518373, + "loss": 2.6955, + "step": 2566 + }, + { + "epoch": 0.11, + "grad_norm": 0.458984375, + "learning_rate": 0.00019863386311684872, + "loss": 2.6859, + "step": 2567 + }, + { + "epoch": 0.11, + "grad_norm": 0.380859375, + "learning_rate": 0.00019863275789469738, + "loss": 2.7287, + "step": 2568 + }, + { + "epoch": 0.11, + "grad_norm": 0.45703125, + "learning_rate": 0.00019863165222873465, + "loss": 2.7705, + "step": 2569 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.00019863054611896558, + "loss": 2.6814, + "step": 2570 + }, + { + "epoch": 0.11, + "grad_norm": 0.42578125, + "learning_rate": 0.00019862943956539507, + "loss": 2.718, + "step": 2571 + }, + { + "epoch": 0.11, + "grad_norm": 0.404296875, + "learning_rate": 0.0001986283325680281, + "loss": 2.5986, + "step": 2572 + }, + { + "epoch": 0.11, + "grad_norm": 0.369140625, + "learning_rate": 0.00019862722512686972, + "loss": 2.6672, + "step": 2573 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.00019862611724192486, + "loss": 2.6861, + "step": 2574 + }, + { + "epoch": 0.11, + "grad_norm": 0.4453125, + "learning_rate": 0.0001986250089131985, + "loss": 2.7433, + "step": 2575 + }, + { + "epoch": 0.11, + "grad_norm": 0.388671875, + "learning_rate": 0.00019862390014069566, + "loss": 2.6606, + "step": 2576 + }, + { + "epoch": 0.11, + "grad_norm": 0.40234375, + "learning_rate": 0.00019862279092442128, + "loss": 2.667, + "step": 2577 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.00019862168126438042, + "loss": 2.8487, + "step": 2578 + }, + { + "epoch": 0.11, + "grad_norm": 0.365234375, + "learning_rate": 0.00019862057116057803, + "loss": 2.6353, + "step": 2579 + }, + { + "epoch": 0.11, + "grad_norm": 0.400390625, + "learning_rate": 0.00019861946061301914, + "loss": 2.6962, + "step": 2580 + }, + { + "epoch": 0.11, + "grad_norm": 0.3984375, + "learning_rate": 0.00019861834962170866, + "loss": 2.7465, + "step": 2581 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.0001986172381866517, + "loss": 2.7061, + "step": 2582 + }, + { + "epoch": 0.11, + "grad_norm": 0.40625, + "learning_rate": 0.00019861612630785316, + "loss": 2.5858, + "step": 2583 + }, + { + "epoch": 0.11, + "grad_norm": 0.375, + "learning_rate": 0.0001986150139853181, + "loss": 2.6278, + "step": 2584 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.00019861390121905155, + "loss": 2.63, + "step": 2585 + }, + { + "epoch": 0.11, + "grad_norm": 0.431640625, + "learning_rate": 0.00019861278800905848, + "loss": 2.6342, + "step": 2586 + }, + { + "epoch": 0.11, + "grad_norm": 0.408203125, + "learning_rate": 0.00019861167435534388, + "loss": 2.6126, + "step": 2587 + }, + { + "epoch": 0.11, + "grad_norm": 0.4296875, + "learning_rate": 0.0001986105602579128, + "loss": 2.7842, + "step": 2588 + }, + { + "epoch": 0.11, + "grad_norm": 0.4609375, + "learning_rate": 0.0001986094457167702, + "loss": 2.5672, + "step": 2589 + }, + { + "epoch": 0.11, + "grad_norm": 0.408203125, + "learning_rate": 0.00019860833073192117, + "loss": 2.6445, + "step": 2590 + }, + { + "epoch": 0.11, + "grad_norm": 0.408203125, + "learning_rate": 0.00019860721530337068, + "loss": 2.7264, + "step": 2591 + }, + { + "epoch": 0.11, + "grad_norm": 0.388671875, + "learning_rate": 0.00019860609943112375, + "loss": 2.55, + "step": 2592 + }, + { + "epoch": 0.11, + "grad_norm": 0.44921875, + "learning_rate": 0.00019860498311518544, + "loss": 2.9818, + "step": 2593 + }, + { + "epoch": 0.11, + "grad_norm": 0.4921875, + "learning_rate": 0.0001986038663555607, + "loss": 2.5515, + "step": 2594 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019860274915225464, + "loss": 2.5904, + "step": 2595 + }, + { + "epoch": 0.11, + "grad_norm": 0.46875, + "learning_rate": 0.0001986016315052722, + "loss": 2.7884, + "step": 2596 + }, + { + "epoch": 0.11, + "grad_norm": 0.41796875, + "learning_rate": 0.00019860051341461847, + "loss": 2.7111, + "step": 2597 + }, + { + "epoch": 0.11, + "grad_norm": 0.427734375, + "learning_rate": 0.00019859939488029845, + "loss": 2.7356, + "step": 2598 + }, + { + "epoch": 0.11, + "grad_norm": 0.439453125, + "learning_rate": 0.00019859827590231724, + "loss": 2.5915, + "step": 2599 + }, + { + "epoch": 0.11, + "grad_norm": 0.375, + "learning_rate": 0.00019859715648067977, + "loss": 2.7452, + "step": 2600 + }, + { + "epoch": 0.11, + "grad_norm": 0.3984375, + "learning_rate": 0.00019859603661539113, + "loss": 2.7467, + "step": 2601 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.0001985949163064564, + "loss": 2.6617, + "step": 2602 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.00019859379555388053, + "loss": 2.7191, + "step": 2603 + }, + { + "epoch": 0.11, + "grad_norm": 0.427734375, + "learning_rate": 0.00019859267435766865, + "loss": 2.7969, + "step": 2604 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.00019859155271782574, + "loss": 2.5134, + "step": 2605 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.0001985904306343569, + "loss": 2.6167, + "step": 2606 + }, + { + "epoch": 0.11, + "grad_norm": 0.3671875, + "learning_rate": 0.00019858930810726712, + "loss": 2.5074, + "step": 2607 + }, + { + "epoch": 0.11, + "grad_norm": 0.40625, + "learning_rate": 0.0001985881851365615, + "loss": 2.6873, + "step": 2608 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.0001985870617222451, + "loss": 2.7078, + "step": 2609 + }, + { + "epoch": 0.11, + "grad_norm": 0.4140625, + "learning_rate": 0.0001985859378643229, + "loss": 2.6953, + "step": 2610 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.00019858481356280002, + "loss": 2.7443, + "step": 2611 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.00019858368881768156, + "loss": 2.7929, + "step": 2612 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.00019858256362897248, + "loss": 2.7975, + "step": 2613 + }, + { + "epoch": 0.11, + "grad_norm": 0.408203125, + "learning_rate": 0.0001985814379966779, + "loss": 2.5942, + "step": 2614 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.0001985803119208029, + "loss": 2.6529, + "step": 2615 + }, + { + "epoch": 0.11, + "grad_norm": 0.396484375, + "learning_rate": 0.0001985791854013525, + "loss": 2.5493, + "step": 2616 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.0001985780584383318, + "loss": 2.6679, + "step": 2617 + }, + { + "epoch": 0.11, + "grad_norm": 0.451171875, + "learning_rate": 0.00019857693103174584, + "loss": 2.6173, + "step": 2618 + }, + { + "epoch": 0.11, + "grad_norm": 0.375, + "learning_rate": 0.00019857580318159975, + "loss": 2.5094, + "step": 2619 + }, + { + "epoch": 0.11, + "grad_norm": 0.408203125, + "learning_rate": 0.00019857467488789855, + "loss": 2.5263, + "step": 2620 + }, + { + "epoch": 0.11, + "grad_norm": 0.431640625, + "learning_rate": 0.00019857354615064737, + "loss": 2.7174, + "step": 2621 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.0001985724169698512, + "loss": 2.6954, + "step": 2622 + }, + { + "epoch": 0.11, + "grad_norm": 0.4375, + "learning_rate": 0.00019857128734551523, + "loss": 2.5815, + "step": 2623 + }, + { + "epoch": 0.11, + "grad_norm": 0.388671875, + "learning_rate": 0.00019857015727764449, + "loss": 2.5138, + "step": 2624 + }, + { + "epoch": 0.11, + "grad_norm": 0.443359375, + "learning_rate": 0.00019856902676624404, + "loss": 2.5502, + "step": 2625 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019856789581131898, + "loss": 2.6817, + "step": 2626 + }, + { + "epoch": 0.11, + "grad_norm": 0.384765625, + "learning_rate": 0.00019856676441287443, + "loss": 2.6496, + "step": 2627 + }, + { + "epoch": 0.11, + "grad_norm": 0.361328125, + "learning_rate": 0.00019856563257091546, + "loss": 2.7308, + "step": 2628 + }, + { + "epoch": 0.11, + "grad_norm": 0.46484375, + "learning_rate": 0.00019856450028544716, + "loss": 2.6156, + "step": 2629 + }, + { + "epoch": 0.11, + "grad_norm": 0.388671875, + "learning_rate": 0.00019856336755647464, + "loss": 2.7754, + "step": 2630 + }, + { + "epoch": 0.11, + "grad_norm": 0.384765625, + "learning_rate": 0.00019856223438400298, + "loss": 2.6671, + "step": 2631 + }, + { + "epoch": 0.11, + "grad_norm": 0.375, + "learning_rate": 0.0001985611007680373, + "loss": 2.7193, + "step": 2632 + }, + { + "epoch": 0.11, + "grad_norm": 0.4375, + "learning_rate": 0.00019855996670858268, + "loss": 2.7804, + "step": 2633 + }, + { + "epoch": 0.11, + "grad_norm": 0.392578125, + "learning_rate": 0.0001985588322056442, + "loss": 2.6877, + "step": 2634 + }, + { + "epoch": 0.11, + "grad_norm": 0.384765625, + "learning_rate": 0.00019855769725922703, + "loss": 2.6616, + "step": 2635 + }, + { + "epoch": 0.11, + "grad_norm": 0.369140625, + "learning_rate": 0.00019855656186933626, + "loss": 2.6561, + "step": 2636 + }, + { + "epoch": 0.11, + "grad_norm": 0.38671875, + "learning_rate": 0.00019855542603597695, + "loss": 2.6908, + "step": 2637 + }, + { + "epoch": 0.11, + "grad_norm": 0.3984375, + "learning_rate": 0.00019855428975915424, + "loss": 2.6442, + "step": 2638 + }, + { + "epoch": 0.11, + "grad_norm": 0.400390625, + "learning_rate": 0.00019855315303887324, + "loss": 2.7411, + "step": 2639 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.00019855201587513905, + "loss": 2.6161, + "step": 2640 + }, + { + "epoch": 0.11, + "grad_norm": 0.4296875, + "learning_rate": 0.00019855087826795684, + "loss": 2.7181, + "step": 2641 + }, + { + "epoch": 0.11, + "grad_norm": 0.408203125, + "learning_rate": 0.00019854974021733167, + "loss": 2.682, + "step": 2642 + }, + { + "epoch": 0.11, + "grad_norm": 0.396484375, + "learning_rate": 0.00019854860172326873, + "loss": 2.757, + "step": 2643 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.00019854746278577304, + "loss": 2.7482, + "step": 2644 + }, + { + "epoch": 0.11, + "grad_norm": 0.392578125, + "learning_rate": 0.00019854632340484982, + "loss": 2.6439, + "step": 2645 + }, + { + "epoch": 0.11, + "grad_norm": 0.427734375, + "learning_rate": 0.00019854518358050413, + "loss": 2.6401, + "step": 2646 + }, + { + "epoch": 0.11, + "grad_norm": 0.4296875, + "learning_rate": 0.00019854404331274116, + "loss": 2.7724, + "step": 2647 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.00019854290260156598, + "loss": 2.6902, + "step": 2648 + }, + { + "epoch": 0.11, + "grad_norm": 0.36328125, + "learning_rate": 0.00019854176144698378, + "loss": 2.7284, + "step": 2649 + }, + { + "epoch": 0.11, + "grad_norm": 0.40625, + "learning_rate": 0.00019854061984899964, + "loss": 2.795, + "step": 2650 + }, + { + "epoch": 0.11, + "grad_norm": 0.41796875, + "learning_rate": 0.00019853947780761874, + "loss": 2.7899, + "step": 2651 + }, + { + "epoch": 0.11, + "grad_norm": 0.400390625, + "learning_rate": 0.00019853833532284618, + "loss": 2.7421, + "step": 2652 + }, + { + "epoch": 0.11, + "grad_norm": 0.44921875, + "learning_rate": 0.00019853719239468716, + "loss": 2.7099, + "step": 2653 + }, + { + "epoch": 0.11, + "grad_norm": 0.357421875, + "learning_rate": 0.00019853604902314676, + "loss": 2.5934, + "step": 2654 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.00019853490520823014, + "loss": 2.6897, + "step": 2655 + }, + { + "epoch": 0.11, + "grad_norm": 0.392578125, + "learning_rate": 0.00019853376094994247, + "loss": 2.6333, + "step": 2656 + }, + { + "epoch": 0.11, + "grad_norm": 0.400390625, + "learning_rate": 0.0001985326162482889, + "loss": 2.6022, + "step": 2657 + }, + { + "epoch": 0.11, + "grad_norm": 0.47265625, + "learning_rate": 0.00019853147110327456, + "loss": 2.6406, + "step": 2658 + }, + { + "epoch": 0.11, + "grad_norm": 0.4453125, + "learning_rate": 0.0001985303255149046, + "loss": 2.7633, + "step": 2659 + }, + { + "epoch": 0.11, + "grad_norm": 0.451171875, + "learning_rate": 0.0001985291794831842, + "loss": 2.6029, + "step": 2660 + }, + { + "epoch": 0.11, + "grad_norm": 0.3671875, + "learning_rate": 0.00019852803300811846, + "loss": 2.6021, + "step": 2661 + }, + { + "epoch": 0.11, + "grad_norm": 0.36328125, + "learning_rate": 0.00019852688608971262, + "loss": 2.646, + "step": 2662 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.0001985257387279718, + "loss": 2.7338, + "step": 2663 + }, + { + "epoch": 0.11, + "grad_norm": 0.416015625, + "learning_rate": 0.00019852459092290113, + "loss": 2.7537, + "step": 2664 + }, + { + "epoch": 0.11, + "grad_norm": 0.412109375, + "learning_rate": 0.0001985234426745058, + "loss": 2.7742, + "step": 2665 + }, + { + "epoch": 0.11, + "grad_norm": 0.453125, + "learning_rate": 0.00019852229398279104, + "loss": 2.7492, + "step": 2666 + }, + { + "epoch": 0.11, + "grad_norm": 0.396484375, + "learning_rate": 0.00019852114484776194, + "loss": 2.6786, + "step": 2667 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019851999526942368, + "loss": 2.6278, + "step": 2668 + }, + { + "epoch": 0.11, + "grad_norm": 0.404296875, + "learning_rate": 0.00019851884524778145, + "loss": 2.7656, + "step": 2669 + }, + { + "epoch": 0.11, + "grad_norm": 0.41015625, + "learning_rate": 0.00019851769478284044, + "loss": 2.767, + "step": 2670 + }, + { + "epoch": 0.11, + "grad_norm": 0.4453125, + "learning_rate": 0.0001985165438746058, + "loss": 2.748, + "step": 2671 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.00019851539252308271, + "loss": 2.5636, + "step": 2672 + }, + { + "epoch": 0.11, + "grad_norm": 0.380859375, + "learning_rate": 0.00019851424072827635, + "loss": 2.6932, + "step": 2673 + }, + { + "epoch": 0.11, + "grad_norm": 0.390625, + "learning_rate": 0.00019851308849019191, + "loss": 2.7656, + "step": 2674 + }, + { + "epoch": 0.11, + "grad_norm": 0.380859375, + "learning_rate": 0.00019851193580883457, + "loss": 2.7387, + "step": 2675 + }, + { + "epoch": 0.11, + "grad_norm": 0.43359375, + "learning_rate": 0.00019851078268420954, + "loss": 2.6572, + "step": 2676 + }, + { + "epoch": 0.11, + "grad_norm": 0.40234375, + "learning_rate": 0.00019850962911632198, + "loss": 2.7404, + "step": 2677 + }, + { + "epoch": 0.11, + "grad_norm": 0.453125, + "learning_rate": 0.0001985084751051771, + "loss": 2.7704, + "step": 2678 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.00019850732065078007, + "loss": 2.7129, + "step": 2679 + }, + { + "epoch": 0.11, + "grad_norm": 0.400390625, + "learning_rate": 0.0001985061657531361, + "loss": 2.5214, + "step": 2680 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.00019850501041225038, + "loss": 2.6446, + "step": 2681 + }, + { + "epoch": 0.11, + "grad_norm": 0.43359375, + "learning_rate": 0.0001985038546281281, + "loss": 2.6538, + "step": 2682 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.0001985026984007745, + "loss": 2.6721, + "step": 2683 + }, + { + "epoch": 0.11, + "grad_norm": 0.458984375, + "learning_rate": 0.00019850154173019476, + "loss": 2.7076, + "step": 2684 + }, + { + "epoch": 0.11, + "grad_norm": 0.404296875, + "learning_rate": 0.00019850038461639407, + "loss": 2.6619, + "step": 2685 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.00019849922705937763, + "loss": 2.5964, + "step": 2686 + }, + { + "epoch": 0.11, + "grad_norm": 0.423828125, + "learning_rate": 0.0001984980690591507, + "loss": 2.6794, + "step": 2687 + }, + { + "epoch": 0.11, + "grad_norm": 0.39453125, + "learning_rate": 0.0001984969106157184, + "loss": 2.6959, + "step": 2688 + }, + { + "epoch": 0.11, + "grad_norm": 0.37890625, + "learning_rate": 0.00019849575172908605, + "loss": 2.7199, + "step": 2689 + }, + { + "epoch": 0.11, + "grad_norm": 0.421875, + "learning_rate": 0.0001984945923992588, + "loss": 2.6525, + "step": 2690 + }, + { + "epoch": 0.11, + "grad_norm": 0.4296875, + "learning_rate": 0.00019849343262624188, + "loss": 2.7203, + "step": 2691 + }, + { + "epoch": 0.11, + "grad_norm": 0.400390625, + "learning_rate": 0.0001984922724100405, + "loss": 2.6779, + "step": 2692 + }, + { + "epoch": 0.11, + "grad_norm": 0.40625, + "learning_rate": 0.00019849111175065988, + "loss": 2.6236, + "step": 2693 + }, + { + "epoch": 0.11, + "grad_norm": 0.40625, + "learning_rate": 0.00019848995064810528, + "loss": 2.5719, + "step": 2694 + }, + { + "epoch": 0.11, + "grad_norm": 0.4140625, + "learning_rate": 0.00019848878910238188, + "loss": 2.753, + "step": 2695 + }, + { + "epoch": 0.11, + "grad_norm": 0.498046875, + "learning_rate": 0.0001984876271134949, + "loss": 2.7484, + "step": 2696 + }, + { + "epoch": 0.11, + "grad_norm": 0.369140625, + "learning_rate": 0.00019848646468144962, + "loss": 2.6397, + "step": 2697 + }, + { + "epoch": 0.11, + "grad_norm": 0.388671875, + "learning_rate": 0.00019848530180625128, + "loss": 2.709, + "step": 2698 + }, + { + "epoch": 0.12, + "grad_norm": 0.380859375, + "learning_rate": 0.000198484138487905, + "loss": 2.6711, + "step": 2699 + }, + { + "epoch": 0.12, + "grad_norm": 0.3984375, + "learning_rate": 0.00019848297472641613, + "loss": 2.6803, + "step": 2700 + }, + { + "epoch": 0.12, + "grad_norm": 0.404296875, + "learning_rate": 0.00019848181052178985, + "loss": 2.7065, + "step": 2701 + }, + { + "epoch": 0.12, + "grad_norm": 0.419921875, + "learning_rate": 0.0001984806458740314, + "loss": 2.6412, + "step": 2702 + }, + { + "epoch": 0.12, + "grad_norm": 0.412109375, + "learning_rate": 0.00019847948078314606, + "loss": 2.6556, + "step": 2703 + }, + { + "epoch": 0.12, + "grad_norm": 0.380859375, + "learning_rate": 0.00019847831524913906, + "loss": 2.5856, + "step": 2704 + }, + { + "epoch": 0.12, + "grad_norm": 0.4375, + "learning_rate": 0.00019847714927201558, + "loss": 2.6902, + "step": 2705 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019847598285178096, + "loss": 2.5951, + "step": 2706 + }, + { + "epoch": 0.12, + "grad_norm": 0.44921875, + "learning_rate": 0.0001984748159884404, + "loss": 2.6771, + "step": 2707 + }, + { + "epoch": 0.12, + "grad_norm": 0.373046875, + "learning_rate": 0.00019847364868199913, + "loss": 2.6739, + "step": 2708 + }, + { + "epoch": 0.12, + "grad_norm": 0.427734375, + "learning_rate": 0.00019847248093246247, + "loss": 2.6659, + "step": 2709 + }, + { + "epoch": 0.12, + "grad_norm": 0.455078125, + "learning_rate": 0.0001984713127398356, + "loss": 2.7224, + "step": 2710 + }, + { + "epoch": 0.12, + "grad_norm": 0.447265625, + "learning_rate": 0.00019847014410412383, + "loss": 2.7249, + "step": 2711 + }, + { + "epoch": 0.12, + "grad_norm": 0.41796875, + "learning_rate": 0.0001984689750253324, + "loss": 2.5554, + "step": 2712 + }, + { + "epoch": 0.12, + "grad_norm": 0.390625, + "learning_rate": 0.00019846780550346655, + "loss": 2.8132, + "step": 2713 + }, + { + "epoch": 0.12, + "grad_norm": 0.392578125, + "learning_rate": 0.00019846663553853158, + "loss": 2.7099, + "step": 2714 + }, + { + "epoch": 0.12, + "grad_norm": 0.41015625, + "learning_rate": 0.00019846546513053271, + "loss": 2.6475, + "step": 2715 + }, + { + "epoch": 0.12, + "grad_norm": 0.361328125, + "learning_rate": 0.00019846429427947525, + "loss": 2.6097, + "step": 2716 + }, + { + "epoch": 0.12, + "grad_norm": 0.37109375, + "learning_rate": 0.00019846312298536445, + "loss": 2.8248, + "step": 2717 + }, + { + "epoch": 0.12, + "grad_norm": 0.404296875, + "learning_rate": 0.0001984619512482056, + "loss": 2.6263, + "step": 2718 + }, + { + "epoch": 0.12, + "grad_norm": 0.373046875, + "learning_rate": 0.00019846077906800392, + "loss": 2.6654, + "step": 2719 + }, + { + "epoch": 0.12, + "grad_norm": 0.427734375, + "learning_rate": 0.00019845960644476475, + "loss": 2.696, + "step": 2720 + }, + { + "epoch": 0.12, + "grad_norm": 0.361328125, + "learning_rate": 0.0001984584333784933, + "loss": 2.6237, + "step": 2721 + }, + { + "epoch": 0.12, + "grad_norm": 0.412109375, + "learning_rate": 0.0001984572598691949, + "loss": 2.6527, + "step": 2722 + }, + { + "epoch": 0.12, + "grad_norm": 0.423828125, + "learning_rate": 0.00019845608591687478, + "loss": 2.6364, + "step": 2723 + }, + { + "epoch": 0.12, + "grad_norm": 0.4765625, + "learning_rate": 0.0001984549115215383, + "loss": 2.7784, + "step": 2724 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.00019845373668319068, + "loss": 2.6645, + "step": 2725 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.00019845256140183724, + "loss": 2.7235, + "step": 2726 + }, + { + "epoch": 0.12, + "grad_norm": 0.39453125, + "learning_rate": 0.00019845138567748324, + "loss": 2.7958, + "step": 2727 + }, + { + "epoch": 0.12, + "grad_norm": 0.443359375, + "learning_rate": 0.00019845020951013399, + "loss": 2.6153, + "step": 2728 + }, + { + "epoch": 0.12, + "grad_norm": 0.390625, + "learning_rate": 0.00019844903289979477, + "loss": 2.609, + "step": 2729 + }, + { + "epoch": 0.12, + "grad_norm": 0.40625, + "learning_rate": 0.00019844785584647087, + "loss": 2.6565, + "step": 2730 + }, + { + "epoch": 0.12, + "grad_norm": 0.43359375, + "learning_rate": 0.0001984466783501676, + "loss": 2.7536, + "step": 2731 + }, + { + "epoch": 0.12, + "grad_norm": 0.3828125, + "learning_rate": 0.00019844550041089027, + "loss": 2.6366, + "step": 2732 + }, + { + "epoch": 0.12, + "grad_norm": 0.384765625, + "learning_rate": 0.0001984443220286442, + "loss": 2.72, + "step": 2733 + }, + { + "epoch": 0.12, + "grad_norm": 0.419921875, + "learning_rate": 0.0001984431432034346, + "loss": 2.5885, + "step": 2734 + }, + { + "epoch": 0.12, + "grad_norm": 0.439453125, + "learning_rate": 0.00019844196393526686, + "loss": 2.6673, + "step": 2735 + }, + { + "epoch": 0.12, + "grad_norm": 0.455078125, + "learning_rate": 0.00019844078422414624, + "loss": 2.8676, + "step": 2736 + }, + { + "epoch": 0.12, + "grad_norm": 0.447265625, + "learning_rate": 0.0001984396040700781, + "loss": 2.6167, + "step": 2737 + }, + { + "epoch": 0.12, + "grad_norm": 0.5078125, + "learning_rate": 0.00019843842347306767, + "loss": 2.7161, + "step": 2738 + }, + { + "epoch": 0.12, + "grad_norm": 0.46484375, + "learning_rate": 0.00019843724243312036, + "loss": 2.8328, + "step": 2739 + }, + { + "epoch": 0.12, + "grad_norm": 0.451171875, + "learning_rate": 0.0001984360609502414, + "loss": 2.7856, + "step": 2740 + }, + { + "epoch": 0.12, + "grad_norm": 0.470703125, + "learning_rate": 0.00019843487902443613, + "loss": 2.828, + "step": 2741 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.00019843369665570993, + "loss": 2.6858, + "step": 2742 + }, + { + "epoch": 0.12, + "grad_norm": 0.39453125, + "learning_rate": 0.00019843251384406803, + "loss": 2.7397, + "step": 2743 + }, + { + "epoch": 0.12, + "grad_norm": 0.4296875, + "learning_rate": 0.0001984313305895158, + "loss": 2.6889, + "step": 2744 + }, + { + "epoch": 0.12, + "grad_norm": 0.396484375, + "learning_rate": 0.00019843014689205853, + "loss": 2.8629, + "step": 2745 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.0001984289627517016, + "loss": 2.5459, + "step": 2746 + }, + { + "epoch": 0.12, + "grad_norm": 0.43359375, + "learning_rate": 0.0001984277781684503, + "loss": 2.7337, + "step": 2747 + }, + { + "epoch": 0.12, + "grad_norm": 0.439453125, + "learning_rate": 0.00019842659314230998, + "loss": 2.6361, + "step": 2748 + }, + { + "epoch": 0.12, + "grad_norm": 0.400390625, + "learning_rate": 0.00019842540767328596, + "loss": 2.7583, + "step": 2749 + }, + { + "epoch": 0.12, + "grad_norm": 0.359375, + "learning_rate": 0.00019842422176138356, + "loss": 2.6049, + "step": 2750 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019842303540660813, + "loss": 2.6912, + "step": 2751 + }, + { + "epoch": 0.12, + "grad_norm": 0.435546875, + "learning_rate": 0.00019842184860896503, + "loss": 2.7153, + "step": 2752 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.00019842066136845955, + "loss": 2.6361, + "step": 2753 + }, + { + "epoch": 0.12, + "grad_norm": 0.466796875, + "learning_rate": 0.0001984194736850971, + "loss": 2.7509, + "step": 2754 + }, + { + "epoch": 0.12, + "grad_norm": 0.455078125, + "learning_rate": 0.00019841828555888297, + "loss": 2.7174, + "step": 2755 + }, + { + "epoch": 0.12, + "grad_norm": 0.41796875, + "learning_rate": 0.00019841709698982253, + "loss": 2.5741, + "step": 2756 + }, + { + "epoch": 0.12, + "grad_norm": 0.390625, + "learning_rate": 0.00019841590797792114, + "loss": 2.5695, + "step": 2757 + }, + { + "epoch": 0.12, + "grad_norm": 0.43359375, + "learning_rate": 0.00019841471852318407, + "loss": 2.6737, + "step": 2758 + }, + { + "epoch": 0.12, + "grad_norm": 0.396484375, + "learning_rate": 0.0001984135286256168, + "loss": 2.7133, + "step": 2759 + }, + { + "epoch": 0.12, + "grad_norm": 0.38671875, + "learning_rate": 0.00019841233828522456, + "loss": 2.73, + "step": 2760 + }, + { + "epoch": 0.12, + "grad_norm": 0.400390625, + "learning_rate": 0.00019841114750201278, + "loss": 2.6414, + "step": 2761 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.0001984099562759868, + "loss": 2.7136, + "step": 2762 + }, + { + "epoch": 0.12, + "grad_norm": 0.4140625, + "learning_rate": 0.00019840876460715202, + "loss": 2.7073, + "step": 2763 + }, + { + "epoch": 0.12, + "grad_norm": 0.455078125, + "learning_rate": 0.00019840757249551375, + "loss": 2.6407, + "step": 2764 + }, + { + "epoch": 0.12, + "grad_norm": 0.447265625, + "learning_rate": 0.00019840637994107734, + "loss": 2.8299, + "step": 2765 + }, + { + "epoch": 0.12, + "grad_norm": 0.451171875, + "learning_rate": 0.0001984051869438482, + "loss": 2.6679, + "step": 2766 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.00019840399350383168, + "loss": 2.6822, + "step": 2767 + }, + { + "epoch": 0.12, + "grad_norm": 0.4140625, + "learning_rate": 0.00019840279962103315, + "loss": 2.5608, + "step": 2768 + }, + { + "epoch": 0.12, + "grad_norm": 0.38671875, + "learning_rate": 0.000198401605295458, + "loss": 2.4816, + "step": 2769 + }, + { + "epoch": 0.12, + "grad_norm": 0.44921875, + "learning_rate": 0.00019840041052711157, + "loss": 2.6391, + "step": 2770 + }, + { + "epoch": 0.12, + "grad_norm": 0.42578125, + "learning_rate": 0.00019839921531599922, + "loss": 2.595, + "step": 2771 + }, + { + "epoch": 0.12, + "grad_norm": 0.427734375, + "learning_rate": 0.0001983980196621264, + "loss": 2.7454, + "step": 2772 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.00019839682356549847, + "loss": 2.688, + "step": 2773 + }, + { + "epoch": 0.12, + "grad_norm": 0.435546875, + "learning_rate": 0.00019839562702612078, + "loss": 2.6541, + "step": 2774 + }, + { + "epoch": 0.12, + "grad_norm": 0.4296875, + "learning_rate": 0.00019839443004399871, + "loss": 2.6937, + "step": 2775 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.00019839323261913768, + "loss": 2.8675, + "step": 2776 + }, + { + "epoch": 0.12, + "grad_norm": 0.392578125, + "learning_rate": 0.00019839203475154307, + "loss": 2.7209, + "step": 2777 + }, + { + "epoch": 0.12, + "grad_norm": 0.412109375, + "learning_rate": 0.00019839083644122022, + "loss": 2.6941, + "step": 2778 + }, + { + "epoch": 0.12, + "grad_norm": 0.400390625, + "learning_rate": 0.0001983896376881746, + "loss": 2.6705, + "step": 2779 + }, + { + "epoch": 0.12, + "grad_norm": 0.4375, + "learning_rate": 0.00019838843849241154, + "loss": 2.6936, + "step": 2780 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.0001983872388539365, + "loss": 2.7673, + "step": 2781 + }, + { + "epoch": 0.12, + "grad_norm": 0.375, + "learning_rate": 0.00019838603877275483, + "loss": 2.599, + "step": 2782 + }, + { + "epoch": 0.12, + "grad_norm": 0.412109375, + "learning_rate": 0.00019838483824887194, + "loss": 2.6663, + "step": 2783 + }, + { + "epoch": 0.12, + "grad_norm": 0.376953125, + "learning_rate": 0.00019838363728229324, + "loss": 2.5332, + "step": 2784 + }, + { + "epoch": 0.12, + "grad_norm": 0.39453125, + "learning_rate": 0.0001983824358730241, + "loss": 2.6745, + "step": 2785 + }, + { + "epoch": 0.12, + "grad_norm": 0.435546875, + "learning_rate": 0.00019838123402107, + "loss": 2.7089, + "step": 2786 + }, + { + "epoch": 0.12, + "grad_norm": 0.4296875, + "learning_rate": 0.00019838003172643624, + "loss": 2.624, + "step": 2787 + }, + { + "epoch": 0.12, + "grad_norm": 0.384765625, + "learning_rate": 0.00019837882898912832, + "loss": 2.7333, + "step": 2788 + }, + { + "epoch": 0.12, + "grad_norm": 0.419921875, + "learning_rate": 0.00019837762580915164, + "loss": 2.6083, + "step": 2789 + }, + { + "epoch": 0.12, + "grad_norm": 0.39453125, + "learning_rate": 0.00019837642218651158, + "loss": 2.6037, + "step": 2790 + }, + { + "epoch": 0.12, + "grad_norm": 0.42578125, + "learning_rate": 0.0001983752181212136, + "loss": 2.7561, + "step": 2791 + }, + { + "epoch": 0.12, + "grad_norm": 0.41796875, + "learning_rate": 0.00019837401361326305, + "loss": 2.6655, + "step": 2792 + }, + { + "epoch": 0.12, + "grad_norm": 0.41015625, + "learning_rate": 0.0001983728086626654, + "loss": 2.6435, + "step": 2793 + }, + { + "epoch": 0.12, + "grad_norm": 0.46484375, + "learning_rate": 0.00019837160326942606, + "loss": 2.777, + "step": 2794 + }, + { + "epoch": 0.12, + "grad_norm": 0.412109375, + "learning_rate": 0.00019837039743355047, + "loss": 2.6155, + "step": 2795 + }, + { + "epoch": 0.12, + "grad_norm": 0.40625, + "learning_rate": 0.00019836919115504405, + "loss": 2.8125, + "step": 2796 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019836798443391218, + "loss": 2.5456, + "step": 2797 + }, + { + "epoch": 0.12, + "grad_norm": 0.412109375, + "learning_rate": 0.00019836677727016038, + "loss": 2.6874, + "step": 2798 + }, + { + "epoch": 0.12, + "grad_norm": 0.390625, + "learning_rate": 0.00019836556966379402, + "loss": 2.547, + "step": 2799 + }, + { + "epoch": 0.12, + "grad_norm": 0.79296875, + "learning_rate": 0.0001983643616148185, + "loss": 2.732, + "step": 2800 + }, + { + "epoch": 0.12, + "grad_norm": 0.369140625, + "learning_rate": 0.00019836315312323933, + "loss": 2.6127, + "step": 2801 + }, + { + "epoch": 0.12, + "grad_norm": 0.42578125, + "learning_rate": 0.00019836194418906193, + "loss": 2.5852, + "step": 2802 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.00019836073481229173, + "loss": 2.7253, + "step": 2803 + }, + { + "epoch": 0.12, + "grad_norm": 0.37890625, + "learning_rate": 0.00019835952499293417, + "loss": 2.6792, + "step": 2804 + }, + { + "epoch": 0.12, + "grad_norm": 0.3984375, + "learning_rate": 0.00019835831473099467, + "loss": 2.6416, + "step": 2805 + }, + { + "epoch": 0.12, + "grad_norm": 0.37109375, + "learning_rate": 0.0001983571040264787, + "loss": 2.647, + "step": 2806 + }, + { + "epoch": 0.12, + "grad_norm": 0.376953125, + "learning_rate": 0.00019835589287939173, + "loss": 2.5961, + "step": 2807 + }, + { + "epoch": 0.12, + "grad_norm": 0.404296875, + "learning_rate": 0.00019835468128973917, + "loss": 2.5769, + "step": 2808 + }, + { + "epoch": 0.12, + "grad_norm": 0.4296875, + "learning_rate": 0.0001983534692575265, + "loss": 2.6791, + "step": 2809 + }, + { + "epoch": 0.12, + "grad_norm": 0.4296875, + "learning_rate": 0.00019835225678275915, + "loss": 2.6563, + "step": 2810 + }, + { + "epoch": 0.12, + "grad_norm": 0.4140625, + "learning_rate": 0.00019835104386544262, + "loss": 2.649, + "step": 2811 + }, + { + "epoch": 0.12, + "grad_norm": 0.447265625, + "learning_rate": 0.00019834983050558228, + "loss": 2.7791, + "step": 2812 + }, + { + "epoch": 0.12, + "grad_norm": 0.37890625, + "learning_rate": 0.0001983486167031837, + "loss": 2.7712, + "step": 2813 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.00019834740245825225, + "loss": 2.7119, + "step": 2814 + }, + { + "epoch": 0.12, + "grad_norm": 0.4140625, + "learning_rate": 0.00019834618777079345, + "loss": 2.7662, + "step": 2815 + }, + { + "epoch": 0.12, + "grad_norm": 0.431640625, + "learning_rate": 0.00019834497264081274, + "loss": 2.5462, + "step": 2816 + }, + { + "epoch": 0.12, + "grad_norm": 0.427734375, + "learning_rate": 0.0001983437570683156, + "loss": 2.8404, + "step": 2817 + }, + { + "epoch": 0.12, + "grad_norm": 0.443359375, + "learning_rate": 0.0001983425410533075, + "loss": 2.6928, + "step": 2818 + }, + { + "epoch": 0.12, + "grad_norm": 0.400390625, + "learning_rate": 0.0001983413245957939, + "loss": 2.6517, + "step": 2819 + }, + { + "epoch": 0.12, + "grad_norm": 0.40625, + "learning_rate": 0.00019834010769578027, + "loss": 2.6667, + "step": 2820 + }, + { + "epoch": 0.12, + "grad_norm": 0.388671875, + "learning_rate": 0.0001983388903532721, + "loss": 2.6248, + "step": 2821 + }, + { + "epoch": 0.12, + "grad_norm": 0.404296875, + "learning_rate": 0.00019833767256827484, + "loss": 2.5949, + "step": 2822 + }, + { + "epoch": 0.12, + "grad_norm": 0.3828125, + "learning_rate": 0.000198336454340794, + "loss": 2.5882, + "step": 2823 + }, + { + "epoch": 0.12, + "grad_norm": 0.42578125, + "learning_rate": 0.0001983352356708351, + "loss": 2.6772, + "step": 2824 + }, + { + "epoch": 0.12, + "grad_norm": 0.404296875, + "learning_rate": 0.00019833401655840354, + "loss": 2.7086, + "step": 2825 + }, + { + "epoch": 0.12, + "grad_norm": 0.4140625, + "learning_rate": 0.00019833279700350484, + "loss": 2.7672, + "step": 2826 + }, + { + "epoch": 0.12, + "grad_norm": 0.44140625, + "learning_rate": 0.00019833157700614452, + "loss": 2.6533, + "step": 2827 + }, + { + "epoch": 0.12, + "grad_norm": 0.486328125, + "learning_rate": 0.00019833035656632798, + "loss": 2.6467, + "step": 2828 + }, + { + "epoch": 0.12, + "grad_norm": 0.41796875, + "learning_rate": 0.00019832913568406082, + "loss": 2.7537, + "step": 2829 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019832791435934848, + "loss": 2.6028, + "step": 2830 + }, + { + "epoch": 0.12, + "grad_norm": 0.42578125, + "learning_rate": 0.00019832669259219646, + "loss": 2.6474, + "step": 2831 + }, + { + "epoch": 0.12, + "grad_norm": 0.3828125, + "learning_rate": 0.00019832547038261023, + "loss": 2.6121, + "step": 2832 + }, + { + "epoch": 0.12, + "grad_norm": 0.392578125, + "learning_rate": 0.00019832424773059534, + "loss": 2.7693, + "step": 2833 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019832302463615724, + "loss": 2.76, + "step": 2834 + }, + { + "epoch": 0.12, + "grad_norm": 0.376953125, + "learning_rate": 0.0001983218010993015, + "loss": 2.7334, + "step": 2835 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019832057712003358, + "loss": 2.6027, + "step": 2836 + }, + { + "epoch": 0.12, + "grad_norm": 0.37890625, + "learning_rate": 0.00019831935269835899, + "loss": 2.6395, + "step": 2837 + }, + { + "epoch": 0.12, + "grad_norm": 0.396484375, + "learning_rate": 0.00019831812783428326, + "loss": 2.6956, + "step": 2838 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.00019831690252781185, + "loss": 2.7472, + "step": 2839 + }, + { + "epoch": 0.12, + "grad_norm": 0.404296875, + "learning_rate": 0.00019831567677895033, + "loss": 2.5872, + "step": 2840 + }, + { + "epoch": 0.12, + "grad_norm": 0.42578125, + "learning_rate": 0.0001983144505877042, + "loss": 2.7684, + "step": 2841 + }, + { + "epoch": 0.12, + "grad_norm": 0.37890625, + "learning_rate": 0.00019831322395407895, + "loss": 2.7188, + "step": 2842 + }, + { + "epoch": 0.12, + "grad_norm": 0.40625, + "learning_rate": 0.00019831199687808014, + "loss": 2.6931, + "step": 2843 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.00019831076935971327, + "loss": 2.6708, + "step": 2844 + }, + { + "epoch": 0.12, + "grad_norm": 0.47265625, + "learning_rate": 0.00019830954139898386, + "loss": 2.8102, + "step": 2845 + }, + { + "epoch": 0.12, + "grad_norm": 0.400390625, + "learning_rate": 0.00019830831299589744, + "loss": 2.6486, + "step": 2846 + }, + { + "epoch": 0.12, + "grad_norm": 0.400390625, + "learning_rate": 0.00019830708415045955, + "loss": 2.6515, + "step": 2847 + }, + { + "epoch": 0.12, + "grad_norm": 0.40234375, + "learning_rate": 0.00019830585486267573, + "loss": 2.7322, + "step": 2848 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019830462513255147, + "loss": 2.6359, + "step": 2849 + }, + { + "epoch": 0.12, + "grad_norm": 0.3984375, + "learning_rate": 0.00019830339496009232, + "loss": 2.7783, + "step": 2850 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.0001983021643453038, + "loss": 2.7482, + "step": 2851 + }, + { + "epoch": 0.12, + "grad_norm": 0.388671875, + "learning_rate": 0.0001983009332881915, + "loss": 2.5529, + "step": 2852 + }, + { + "epoch": 0.12, + "grad_norm": 0.412109375, + "learning_rate": 0.0001982997017887609, + "loss": 2.7033, + "step": 2853 + }, + { + "epoch": 0.12, + "grad_norm": 0.46484375, + "learning_rate": 0.00019829846984701757, + "loss": 2.812, + "step": 2854 + }, + { + "epoch": 0.12, + "grad_norm": 0.390625, + "learning_rate": 0.00019829723746296704, + "loss": 2.6675, + "step": 2855 + }, + { + "epoch": 0.12, + "grad_norm": 0.4375, + "learning_rate": 0.0001982960046366149, + "loss": 2.7457, + "step": 2856 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.00019829477136796663, + "loss": 2.8505, + "step": 2857 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.0001982935376570278, + "loss": 2.6295, + "step": 2858 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.00019829230350380402, + "loss": 2.7559, + "step": 2859 + }, + { + "epoch": 0.12, + "grad_norm": 0.474609375, + "learning_rate": 0.00019829106890830077, + "loss": 2.7548, + "step": 2860 + }, + { + "epoch": 0.12, + "grad_norm": 0.390625, + "learning_rate": 0.00019828983387052364, + "loss": 2.5794, + "step": 2861 + }, + { + "epoch": 0.12, + "grad_norm": 0.41796875, + "learning_rate": 0.00019828859839047817, + "loss": 2.7156, + "step": 2862 + }, + { + "epoch": 0.12, + "grad_norm": 0.462890625, + "learning_rate": 0.0001982873624681699, + "loss": 2.7692, + "step": 2863 + }, + { + "epoch": 0.12, + "grad_norm": 0.439453125, + "learning_rate": 0.00019828612610360447, + "loss": 2.6707, + "step": 2864 + }, + { + "epoch": 0.12, + "grad_norm": 0.3984375, + "learning_rate": 0.00019828488929678737, + "loss": 2.5957, + "step": 2865 + }, + { + "epoch": 0.12, + "grad_norm": 0.45703125, + "learning_rate": 0.00019828365204772416, + "loss": 2.7831, + "step": 2866 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.00019828241435642047, + "loss": 2.7183, + "step": 2867 + }, + { + "epoch": 0.12, + "grad_norm": 0.435546875, + "learning_rate": 0.00019828117622288182, + "loss": 2.6205, + "step": 2868 + }, + { + "epoch": 0.12, + "grad_norm": 0.439453125, + "learning_rate": 0.0001982799376471138, + "loss": 2.6988, + "step": 2869 + }, + { + "epoch": 0.12, + "grad_norm": 0.404296875, + "learning_rate": 0.00019827869862912198, + "loss": 2.5238, + "step": 2870 + }, + { + "epoch": 0.12, + "grad_norm": 0.423828125, + "learning_rate": 0.00019827745916891194, + "loss": 2.6846, + "step": 2871 + }, + { + "epoch": 0.12, + "grad_norm": 0.427734375, + "learning_rate": 0.00019827621926648924, + "loss": 2.6908, + "step": 2872 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019827497892185947, + "loss": 2.6598, + "step": 2873 + }, + { + "epoch": 0.12, + "grad_norm": 0.453125, + "learning_rate": 0.0001982737381350282, + "loss": 2.6848, + "step": 2874 + }, + { + "epoch": 0.12, + "grad_norm": 0.419921875, + "learning_rate": 0.00019827249690600103, + "loss": 2.6456, + "step": 2875 + }, + { + "epoch": 0.12, + "grad_norm": 0.4296875, + "learning_rate": 0.00019827125523478353, + "loss": 2.6387, + "step": 2876 + }, + { + "epoch": 0.12, + "grad_norm": 0.455078125, + "learning_rate": 0.00019827001312138133, + "loss": 2.6878, + "step": 2877 + }, + { + "epoch": 0.12, + "grad_norm": 0.40625, + "learning_rate": 0.00019826877056579995, + "loss": 2.6996, + "step": 2878 + }, + { + "epoch": 0.12, + "grad_norm": 0.443359375, + "learning_rate": 0.00019826752756804504, + "loss": 2.6634, + "step": 2879 + }, + { + "epoch": 0.12, + "grad_norm": 0.462890625, + "learning_rate": 0.00019826628412812216, + "loss": 2.6992, + "step": 2880 + }, + { + "epoch": 0.12, + "grad_norm": 0.41015625, + "learning_rate": 0.0001982650402460369, + "loss": 2.7223, + "step": 2881 + }, + { + "epoch": 0.12, + "grad_norm": 0.4765625, + "learning_rate": 0.00019826379592179487, + "loss": 2.7305, + "step": 2882 + }, + { + "epoch": 0.12, + "grad_norm": 0.396484375, + "learning_rate": 0.0001982625511554017, + "loss": 2.6701, + "step": 2883 + }, + { + "epoch": 0.12, + "grad_norm": 0.3671875, + "learning_rate": 0.00019826130594686293, + "loss": 2.6813, + "step": 2884 + }, + { + "epoch": 0.12, + "grad_norm": 0.490234375, + "learning_rate": 0.0001982600602961842, + "loss": 2.7658, + "step": 2885 + }, + { + "epoch": 0.12, + "grad_norm": 0.369140625, + "learning_rate": 0.0001982588142033711, + "loss": 2.5604, + "step": 2886 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019825756766842927, + "loss": 2.7729, + "step": 2887 + }, + { + "epoch": 0.12, + "grad_norm": 0.390625, + "learning_rate": 0.00019825632069136429, + "loss": 2.5643, + "step": 2888 + }, + { + "epoch": 0.12, + "grad_norm": 0.41015625, + "learning_rate": 0.00019825507327218177, + "loss": 2.6365, + "step": 2889 + }, + { + "epoch": 0.12, + "grad_norm": 0.423828125, + "learning_rate": 0.0001982538254108873, + "loss": 2.734, + "step": 2890 + }, + { + "epoch": 0.12, + "grad_norm": 0.462890625, + "learning_rate": 0.00019825257710748658, + "loss": 2.751, + "step": 2891 + }, + { + "epoch": 0.12, + "grad_norm": 0.41796875, + "learning_rate": 0.00019825132836198513, + "loss": 2.6156, + "step": 2892 + }, + { + "epoch": 0.12, + "grad_norm": 0.357421875, + "learning_rate": 0.0001982500791743886, + "loss": 2.6714, + "step": 2893 + }, + { + "epoch": 0.12, + "grad_norm": 0.427734375, + "learning_rate": 0.00019824882954470264, + "loss": 2.5695, + "step": 2894 + }, + { + "epoch": 0.12, + "grad_norm": 0.400390625, + "learning_rate": 0.00019824757947293286, + "loss": 2.648, + "step": 2895 + }, + { + "epoch": 0.12, + "grad_norm": 0.42578125, + "learning_rate": 0.0001982463289590849, + "loss": 2.7206, + "step": 2896 + }, + { + "epoch": 0.12, + "grad_norm": 0.435546875, + "learning_rate": 0.0001982450780031643, + "loss": 2.7936, + "step": 2897 + }, + { + "epoch": 0.12, + "grad_norm": 0.419921875, + "learning_rate": 0.0001982438266051768, + "loss": 2.6847, + "step": 2898 + }, + { + "epoch": 0.12, + "grad_norm": 0.37890625, + "learning_rate": 0.00019824257476512797, + "loss": 2.7419, + "step": 2899 + }, + { + "epoch": 0.12, + "grad_norm": 0.388671875, + "learning_rate": 0.00019824132248302345, + "loss": 2.6729, + "step": 2900 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.00019824006975886887, + "loss": 2.6464, + "step": 2901 + }, + { + "epoch": 0.12, + "grad_norm": 0.412109375, + "learning_rate": 0.00019823881659266994, + "loss": 2.6158, + "step": 2902 + }, + { + "epoch": 0.12, + "grad_norm": 0.4375, + "learning_rate": 0.0001982375629844322, + "loss": 2.8958, + "step": 2903 + }, + { + "epoch": 0.12, + "grad_norm": 0.376953125, + "learning_rate": 0.0001982363089341613, + "loss": 2.5488, + "step": 2904 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.00019823505444186294, + "loss": 2.6474, + "step": 2905 + }, + { + "epoch": 0.12, + "grad_norm": 0.357421875, + "learning_rate": 0.00019823379950754271, + "loss": 2.6871, + "step": 2906 + }, + { + "epoch": 0.12, + "grad_norm": 0.404296875, + "learning_rate": 0.00019823254413120633, + "loss": 2.7183, + "step": 2907 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.00019823128831285937, + "loss": 2.7571, + "step": 2908 + }, + { + "epoch": 0.12, + "grad_norm": 0.42578125, + "learning_rate": 0.00019823003205250752, + "loss": 2.6491, + "step": 2909 + }, + { + "epoch": 0.12, + "grad_norm": 0.392578125, + "learning_rate": 0.00019822877535015642, + "loss": 2.6472, + "step": 2910 + }, + { + "epoch": 0.12, + "grad_norm": 0.380859375, + "learning_rate": 0.00019822751820581174, + "loss": 2.6799, + "step": 2911 + }, + { + "epoch": 0.12, + "grad_norm": 0.37890625, + "learning_rate": 0.00019822626061947912, + "loss": 2.7032, + "step": 2912 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.00019822500259116424, + "loss": 2.7215, + "step": 2913 + }, + { + "epoch": 0.12, + "grad_norm": 0.353515625, + "learning_rate": 0.0001982237441208727, + "loss": 2.5994, + "step": 2914 + }, + { + "epoch": 0.12, + "grad_norm": 0.341796875, + "learning_rate": 0.00019822248520861023, + "loss": 2.5732, + "step": 2915 + }, + { + "epoch": 0.12, + "grad_norm": 0.41015625, + "learning_rate": 0.0001982212258543825, + "loss": 2.6058, + "step": 2916 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019821996605819513, + "loss": 2.6175, + "step": 2917 + }, + { + "epoch": 0.12, + "grad_norm": 0.44921875, + "learning_rate": 0.0001982187058200538, + "loss": 2.6368, + "step": 2918 + }, + { + "epoch": 0.12, + "grad_norm": 0.40625, + "learning_rate": 0.00019821744513996419, + "loss": 2.7635, + "step": 2919 + }, + { + "epoch": 0.12, + "grad_norm": 0.35546875, + "learning_rate": 0.00019821618401793197, + "loss": 2.6723, + "step": 2920 + }, + { + "epoch": 0.12, + "grad_norm": 0.408203125, + "learning_rate": 0.00019821492245396285, + "loss": 2.6369, + "step": 2921 + }, + { + "epoch": 0.12, + "grad_norm": 0.421875, + "learning_rate": 0.00019821366044806243, + "loss": 2.6284, + "step": 2922 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.00019821239800023644, + "loss": 2.7098, + "step": 2923 + }, + { + "epoch": 0.12, + "grad_norm": 0.37109375, + "learning_rate": 0.00019821113511049056, + "loss": 2.5578, + "step": 2924 + }, + { + "epoch": 0.12, + "grad_norm": 0.443359375, + "learning_rate": 0.00019820987177883047, + "loss": 2.709, + "step": 2925 + }, + { + "epoch": 0.12, + "grad_norm": 0.4453125, + "learning_rate": 0.00019820860800526182, + "loss": 2.8171, + "step": 2926 + }, + { + "epoch": 0.12, + "grad_norm": 0.431640625, + "learning_rate": 0.00019820734378979036, + "loss": 2.7683, + "step": 2927 + }, + { + "epoch": 0.12, + "grad_norm": 0.416015625, + "learning_rate": 0.00019820607913242173, + "loss": 2.686, + "step": 2928 + }, + { + "epoch": 0.12, + "grad_norm": 0.384765625, + "learning_rate": 0.00019820481403316165, + "loss": 2.4675, + "step": 2929 + }, + { + "epoch": 0.12, + "grad_norm": 0.392578125, + "learning_rate": 0.00019820354849201576, + "loss": 2.6209, + "step": 2930 + }, + { + "epoch": 0.12, + "grad_norm": 0.392578125, + "learning_rate": 0.0001982022825089898, + "loss": 2.53, + "step": 2931 + }, + { + "epoch": 0.12, + "grad_norm": 0.423828125, + "learning_rate": 0.00019820101608408947, + "loss": 2.7325, + "step": 2932 + }, + { + "epoch": 0.12, + "grad_norm": 0.470703125, + "learning_rate": 0.00019819974921732049, + "loss": 2.6182, + "step": 2933 + }, + { + "epoch": 0.13, + "grad_norm": 0.40625, + "learning_rate": 0.00019819848190868845, + "loss": 2.6692, + "step": 2934 + }, + { + "epoch": 0.13, + "grad_norm": 0.4140625, + "learning_rate": 0.00019819721415819918, + "loss": 2.6751, + "step": 2935 + }, + { + "epoch": 0.13, + "grad_norm": 0.443359375, + "learning_rate": 0.00019819594596585832, + "loss": 2.7935, + "step": 2936 + }, + { + "epoch": 0.13, + "grad_norm": 0.451171875, + "learning_rate": 0.00019819467733167158, + "loss": 2.8393, + "step": 2937 + }, + { + "epoch": 0.13, + "grad_norm": 0.443359375, + "learning_rate": 0.0001981934082556447, + "loss": 2.5982, + "step": 2938 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.00019819213873778336, + "loss": 2.845, + "step": 2939 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.0001981908687780933, + "loss": 2.6866, + "step": 2940 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.0001981895983765802, + "loss": 2.5513, + "step": 2941 + }, + { + "epoch": 0.13, + "grad_norm": 0.37109375, + "learning_rate": 0.00019818832753324976, + "loss": 2.6009, + "step": 2942 + }, + { + "epoch": 0.13, + "grad_norm": 0.390625, + "learning_rate": 0.00019818705624810777, + "loss": 2.6018, + "step": 2943 + }, + { + "epoch": 0.13, + "grad_norm": 0.404296875, + "learning_rate": 0.00019818578452115987, + "loss": 2.6956, + "step": 2944 + }, + { + "epoch": 0.13, + "grad_norm": 0.423828125, + "learning_rate": 0.00019818451235241187, + "loss": 2.6991, + "step": 2945 + }, + { + "epoch": 0.13, + "grad_norm": 0.416015625, + "learning_rate": 0.00019818323974186942, + "loss": 2.6539, + "step": 2946 + }, + { + "epoch": 0.13, + "grad_norm": 0.439453125, + "learning_rate": 0.00019818196668953828, + "loss": 2.7238, + "step": 2947 + }, + { + "epoch": 0.13, + "grad_norm": 0.365234375, + "learning_rate": 0.00019818069319542416, + "loss": 2.736, + "step": 2948 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.00019817941925953284, + "loss": 2.5983, + "step": 2949 + }, + { + "epoch": 0.13, + "grad_norm": 0.4140625, + "learning_rate": 0.00019817814488186998, + "loss": 2.6191, + "step": 2950 + }, + { + "epoch": 0.13, + "grad_norm": 0.41015625, + "learning_rate": 0.00019817687006244133, + "loss": 2.5294, + "step": 2951 + }, + { + "epoch": 0.13, + "grad_norm": 0.40625, + "learning_rate": 0.0001981755948012527, + "loss": 2.6918, + "step": 2952 + }, + { + "epoch": 0.13, + "grad_norm": 0.412109375, + "learning_rate": 0.00019817431909830972, + "loss": 2.7077, + "step": 2953 + }, + { + "epoch": 0.13, + "grad_norm": 0.412109375, + "learning_rate": 0.0001981730429536182, + "loss": 2.5567, + "step": 2954 + }, + { + "epoch": 0.13, + "grad_norm": 0.41015625, + "learning_rate": 0.00019817176636718388, + "loss": 2.7207, + "step": 2955 + }, + { + "epoch": 0.13, + "grad_norm": 0.474609375, + "learning_rate": 0.0001981704893390125, + "loss": 2.6646, + "step": 2956 + }, + { + "epoch": 0.13, + "grad_norm": 0.412109375, + "learning_rate": 0.00019816921186910973, + "loss": 2.792, + "step": 2957 + }, + { + "epoch": 0.13, + "grad_norm": 0.412109375, + "learning_rate": 0.00019816793395748144, + "loss": 2.5997, + "step": 2958 + }, + { + "epoch": 0.13, + "grad_norm": 0.443359375, + "learning_rate": 0.0001981666556041333, + "loss": 2.7441, + "step": 2959 + }, + { + "epoch": 0.13, + "grad_norm": 0.4296875, + "learning_rate": 0.0001981653768090711, + "loss": 2.6393, + "step": 2960 + }, + { + "epoch": 0.13, + "grad_norm": 0.388671875, + "learning_rate": 0.00019816409757230058, + "loss": 2.6971, + "step": 2961 + }, + { + "epoch": 0.13, + "grad_norm": 0.470703125, + "learning_rate": 0.00019816281789382747, + "loss": 2.7121, + "step": 2962 + }, + { + "epoch": 0.13, + "grad_norm": 0.419921875, + "learning_rate": 0.00019816153777365758, + "loss": 2.6484, + "step": 2963 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.00019816025721179666, + "loss": 2.7068, + "step": 2964 + }, + { + "epoch": 0.13, + "grad_norm": 0.431640625, + "learning_rate": 0.00019815897620825042, + "loss": 2.6623, + "step": 2965 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.0001981576947630247, + "loss": 2.5946, + "step": 2966 + }, + { + "epoch": 0.13, + "grad_norm": 0.482421875, + "learning_rate": 0.00019815641287612517, + "loss": 2.7237, + "step": 2967 + }, + { + "epoch": 0.13, + "grad_norm": 0.455078125, + "learning_rate": 0.00019815513054755767, + "loss": 2.7886, + "step": 2968 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.00019815384777732798, + "loss": 2.7383, + "step": 2969 + }, + { + "epoch": 0.13, + "grad_norm": 0.4375, + "learning_rate": 0.00019815256456544184, + "loss": 2.5985, + "step": 2970 + }, + { + "epoch": 0.13, + "grad_norm": 0.4375, + "learning_rate": 0.000198151280911905, + "loss": 2.8579, + "step": 2971 + }, + { + "epoch": 0.13, + "grad_norm": 0.4765625, + "learning_rate": 0.0001981499968167233, + "loss": 2.7144, + "step": 2972 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.00019814871227990247, + "loss": 2.7078, + "step": 2973 + }, + { + "epoch": 0.13, + "grad_norm": 0.43359375, + "learning_rate": 0.00019814742730144832, + "loss": 2.6529, + "step": 2974 + }, + { + "epoch": 0.13, + "grad_norm": 0.50390625, + "learning_rate": 0.0001981461418813666, + "loss": 2.7635, + "step": 2975 + }, + { + "epoch": 0.13, + "grad_norm": 0.447265625, + "learning_rate": 0.00019814485601966312, + "loss": 2.7199, + "step": 2976 + }, + { + "epoch": 0.13, + "grad_norm": 0.404296875, + "learning_rate": 0.00019814356971634364, + "loss": 2.6883, + "step": 2977 + }, + { + "epoch": 0.13, + "grad_norm": 0.482421875, + "learning_rate": 0.000198142282971414, + "loss": 2.6536, + "step": 2978 + }, + { + "epoch": 0.13, + "grad_norm": 0.38671875, + "learning_rate": 0.00019814099578487992, + "loss": 2.5785, + "step": 2979 + }, + { + "epoch": 0.13, + "grad_norm": 0.40625, + "learning_rate": 0.00019813970815674723, + "loss": 2.6083, + "step": 2980 + }, + { + "epoch": 0.13, + "grad_norm": 0.43359375, + "learning_rate": 0.00019813842008702172, + "loss": 2.6589, + "step": 2981 + }, + { + "epoch": 0.13, + "grad_norm": 0.384765625, + "learning_rate": 0.00019813713157570918, + "loss": 2.667, + "step": 2982 + }, + { + "epoch": 0.13, + "grad_norm": 0.404296875, + "learning_rate": 0.0001981358426228154, + "loss": 2.7676, + "step": 2983 + }, + { + "epoch": 0.13, + "grad_norm": 0.421875, + "learning_rate": 0.00019813455322834626, + "loss": 2.7484, + "step": 2984 + }, + { + "epoch": 0.13, + "grad_norm": 0.38671875, + "learning_rate": 0.00019813326339230742, + "loss": 2.6086, + "step": 2985 + }, + { + "epoch": 0.13, + "grad_norm": 0.498046875, + "learning_rate": 0.0001981319731147048, + "loss": 2.7246, + "step": 2986 + }, + { + "epoch": 0.13, + "grad_norm": 0.375, + "learning_rate": 0.00019813068239554413, + "loss": 2.6364, + "step": 2987 + }, + { + "epoch": 0.13, + "grad_norm": 0.451171875, + "learning_rate": 0.00019812939123483127, + "loss": 2.7095, + "step": 2988 + }, + { + "epoch": 0.13, + "grad_norm": 0.455078125, + "learning_rate": 0.00019812809963257202, + "loss": 2.6697, + "step": 2989 + }, + { + "epoch": 0.13, + "grad_norm": 0.439453125, + "learning_rate": 0.0001981268075887722, + "loss": 2.6538, + "step": 2990 + }, + { + "epoch": 0.13, + "grad_norm": 0.44921875, + "learning_rate": 0.00019812551510343756, + "loss": 2.5972, + "step": 2991 + }, + { + "epoch": 0.13, + "grad_norm": 0.4609375, + "learning_rate": 0.00019812422217657398, + "loss": 2.6722, + "step": 2992 + }, + { + "epoch": 0.13, + "grad_norm": 0.390625, + "learning_rate": 0.00019812292880818725, + "loss": 2.6485, + "step": 2993 + }, + { + "epoch": 0.13, + "grad_norm": 0.421875, + "learning_rate": 0.00019812163499828325, + "loss": 2.7932, + "step": 2994 + }, + { + "epoch": 0.13, + "grad_norm": 0.388671875, + "learning_rate": 0.0001981203407468677, + "loss": 2.5745, + "step": 2995 + }, + { + "epoch": 0.13, + "grad_norm": 0.46875, + "learning_rate": 0.0001981190460539465, + "loss": 2.7674, + "step": 2996 + }, + { + "epoch": 0.13, + "grad_norm": 0.43359375, + "learning_rate": 0.00019811775091952544, + "loss": 2.7508, + "step": 2997 + }, + { + "epoch": 0.13, + "grad_norm": 0.4296875, + "learning_rate": 0.0001981164553436104, + "loss": 2.6983, + "step": 2998 + }, + { + "epoch": 0.13, + "grad_norm": 0.40625, + "learning_rate": 0.0001981151593262071, + "loss": 2.6824, + "step": 2999 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.00019811386286732146, + "loss": 2.6832, + "step": 3000 + }, + { + "epoch": 0.13, + "grad_norm": 0.384765625, + "learning_rate": 0.00019811256596695932, + "loss": 2.5472, + "step": 3001 + }, + { + "epoch": 0.13, + "grad_norm": 0.439453125, + "learning_rate": 0.00019811126862512648, + "loss": 2.8282, + "step": 3002 + }, + { + "epoch": 0.13, + "grad_norm": 0.408203125, + "learning_rate": 0.00019810997084182877, + "loss": 2.7493, + "step": 3003 + }, + { + "epoch": 0.13, + "grad_norm": 0.4296875, + "learning_rate": 0.00019810867261707207, + "loss": 2.7752, + "step": 3004 + }, + { + "epoch": 0.13, + "grad_norm": 0.38671875, + "learning_rate": 0.00019810737395086218, + "loss": 2.6602, + "step": 3005 + }, + { + "epoch": 0.13, + "grad_norm": 0.37890625, + "learning_rate": 0.00019810607484320496, + "loss": 2.5715, + "step": 3006 + }, + { + "epoch": 0.13, + "grad_norm": 0.349609375, + "learning_rate": 0.00019810477529410626, + "loss": 2.7352, + "step": 3007 + }, + { + "epoch": 0.13, + "grad_norm": 0.47265625, + "learning_rate": 0.00019810347530357193, + "loss": 2.8715, + "step": 3008 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.00019810217487160783, + "loss": 2.728, + "step": 3009 + }, + { + "epoch": 0.13, + "grad_norm": 0.404296875, + "learning_rate": 0.00019810087399821976, + "loss": 2.7985, + "step": 3010 + }, + { + "epoch": 0.13, + "grad_norm": 0.455078125, + "learning_rate": 0.00019809957268341366, + "loss": 2.6121, + "step": 3011 + }, + { + "epoch": 0.13, + "grad_norm": 0.462890625, + "learning_rate": 0.0001980982709271953, + "loss": 2.7274, + "step": 3012 + }, + { + "epoch": 0.13, + "grad_norm": 0.396484375, + "learning_rate": 0.00019809696872957056, + "loss": 2.7809, + "step": 3013 + }, + { + "epoch": 0.13, + "grad_norm": 0.443359375, + "learning_rate": 0.00019809566609054532, + "loss": 2.6393, + "step": 3014 + }, + { + "epoch": 0.13, + "grad_norm": 0.443359375, + "learning_rate": 0.00019809436301012544, + "loss": 2.7912, + "step": 3015 + }, + { + "epoch": 0.13, + "grad_norm": 0.40625, + "learning_rate": 0.00019809305948831677, + "loss": 2.6085, + "step": 3016 + }, + { + "epoch": 0.13, + "grad_norm": 0.396484375, + "learning_rate": 0.00019809175552512517, + "loss": 2.6303, + "step": 3017 + }, + { + "epoch": 0.13, + "grad_norm": 0.453125, + "learning_rate": 0.00019809045112055654, + "loss": 2.7616, + "step": 3018 + }, + { + "epoch": 0.13, + "grad_norm": 0.4453125, + "learning_rate": 0.00019808914627461672, + "loss": 2.7107, + "step": 3019 + }, + { + "epoch": 0.13, + "grad_norm": 0.390625, + "learning_rate": 0.00019808784098731159, + "loss": 2.5605, + "step": 3020 + }, + { + "epoch": 0.13, + "grad_norm": 0.369140625, + "learning_rate": 0.00019808653525864702, + "loss": 2.4685, + "step": 3021 + }, + { + "epoch": 0.13, + "grad_norm": 0.4375, + "learning_rate": 0.0001980852290886289, + "loss": 2.726, + "step": 3022 + }, + { + "epoch": 0.13, + "grad_norm": 0.37890625, + "learning_rate": 0.00019808392247726308, + "loss": 2.7032, + "step": 3023 + }, + { + "epoch": 0.13, + "grad_norm": 0.4140625, + "learning_rate": 0.00019808261542455545, + "loss": 2.7103, + "step": 3024 + }, + { + "epoch": 0.13, + "grad_norm": 0.423828125, + "learning_rate": 0.00019808130793051192, + "loss": 2.7178, + "step": 3025 + }, + { + "epoch": 0.13, + "grad_norm": 0.39453125, + "learning_rate": 0.00019807999999513833, + "loss": 2.5797, + "step": 3026 + }, + { + "epoch": 0.13, + "grad_norm": 0.373046875, + "learning_rate": 0.0001980786916184406, + "loss": 2.4603, + "step": 3027 + }, + { + "epoch": 0.13, + "grad_norm": 0.37890625, + "learning_rate": 0.00019807738280042459, + "loss": 2.6536, + "step": 3028 + }, + { + "epoch": 0.13, + "grad_norm": 0.4140625, + "learning_rate": 0.0001980760735410962, + "loss": 2.6754, + "step": 3029 + }, + { + "epoch": 0.13, + "grad_norm": 0.400390625, + "learning_rate": 0.0001980747638404613, + "loss": 2.696, + "step": 3030 + }, + { + "epoch": 0.13, + "grad_norm": 0.396484375, + "learning_rate": 0.00019807345369852582, + "loss": 2.6833, + "step": 3031 + }, + { + "epoch": 0.13, + "grad_norm": 0.439453125, + "learning_rate": 0.00019807214311529564, + "loss": 2.7199, + "step": 3032 + }, + { + "epoch": 0.13, + "grad_norm": 0.384765625, + "learning_rate": 0.00019807083209077669, + "loss": 2.5314, + "step": 3033 + }, + { + "epoch": 0.13, + "grad_norm": 0.40234375, + "learning_rate": 0.0001980695206249748, + "loss": 2.7367, + "step": 3034 + }, + { + "epoch": 0.13, + "grad_norm": 0.447265625, + "learning_rate": 0.00019806820871789591, + "loss": 2.7596, + "step": 3035 + }, + { + "epoch": 0.13, + "grad_norm": 0.400390625, + "learning_rate": 0.00019806689636954594, + "loss": 2.7064, + "step": 3036 + }, + { + "epoch": 0.13, + "grad_norm": 0.408203125, + "learning_rate": 0.00019806558357993078, + "loss": 2.736, + "step": 3037 + }, + { + "epoch": 0.13, + "grad_norm": 0.423828125, + "learning_rate": 0.0001980642703490563, + "loss": 2.7662, + "step": 3038 + }, + { + "epoch": 0.13, + "grad_norm": 0.408203125, + "learning_rate": 0.00019806295667692845, + "loss": 2.5844, + "step": 3039 + }, + { + "epoch": 0.13, + "grad_norm": 0.365234375, + "learning_rate": 0.00019806164256355315, + "loss": 2.7376, + "step": 3040 + }, + { + "epoch": 0.13, + "grad_norm": 0.369140625, + "learning_rate": 0.00019806032800893627, + "loss": 2.6013, + "step": 3041 + }, + { + "epoch": 0.13, + "grad_norm": 0.41015625, + "learning_rate": 0.00019805901301308377, + "loss": 2.6519, + "step": 3042 + }, + { + "epoch": 0.13, + "grad_norm": 0.455078125, + "learning_rate": 0.00019805769757600157, + "loss": 2.7362, + "step": 3043 + }, + { + "epoch": 0.13, + "grad_norm": 0.423828125, + "learning_rate": 0.00019805638169769553, + "loss": 2.6462, + "step": 3044 + }, + { + "epoch": 0.13, + "grad_norm": 0.4296875, + "learning_rate": 0.0001980550653781716, + "loss": 2.6406, + "step": 3045 + }, + { + "epoch": 0.13, + "grad_norm": 0.431640625, + "learning_rate": 0.00019805374861743577, + "loss": 2.5912, + "step": 3046 + }, + { + "epoch": 0.13, + "grad_norm": 0.423828125, + "learning_rate": 0.00019805243141549384, + "loss": 2.7934, + "step": 3047 + }, + { + "epoch": 0.13, + "grad_norm": 0.419921875, + "learning_rate": 0.00019805111377235184, + "loss": 2.8109, + "step": 3048 + }, + { + "epoch": 0.13, + "grad_norm": 0.474609375, + "learning_rate": 0.00019804979568801567, + "loss": 2.748, + "step": 3049 + }, + { + "epoch": 0.13, + "grad_norm": 0.419921875, + "learning_rate": 0.00019804847716249123, + "loss": 2.7447, + "step": 3050 + }, + { + "epoch": 0.13, + "grad_norm": 0.40625, + "learning_rate": 0.00019804715819578448, + "loss": 2.627, + "step": 3051 + }, + { + "epoch": 0.13, + "grad_norm": 0.458984375, + "learning_rate": 0.00019804583878790135, + "loss": 2.6862, + "step": 3052 + }, + { + "epoch": 0.13, + "grad_norm": 0.388671875, + "learning_rate": 0.00019804451893884776, + "loss": 2.6154, + "step": 3053 + }, + { + "epoch": 0.13, + "grad_norm": 0.44921875, + "learning_rate": 0.0001980431986486297, + "loss": 2.6947, + "step": 3054 + }, + { + "epoch": 0.13, + "grad_norm": 0.361328125, + "learning_rate": 0.00019804187791725304, + "loss": 2.6083, + "step": 3055 + }, + { + "epoch": 0.13, + "grad_norm": 0.5, + "learning_rate": 0.0001980405567447238, + "loss": 2.6665, + "step": 3056 + }, + { + "epoch": 0.13, + "grad_norm": 0.396484375, + "learning_rate": 0.00019803923513104786, + "loss": 2.6783, + "step": 3057 + }, + { + "epoch": 0.13, + "grad_norm": 0.390625, + "learning_rate": 0.0001980379130762312, + "loss": 2.6209, + "step": 3058 + }, + { + "epoch": 0.13, + "grad_norm": 0.40625, + "learning_rate": 0.00019803659058027973, + "loss": 2.7908, + "step": 3059 + }, + { + "epoch": 0.13, + "grad_norm": 0.380859375, + "learning_rate": 0.00019803526764319948, + "loss": 2.6055, + "step": 3060 + }, + { + "epoch": 0.13, + "grad_norm": 0.388671875, + "learning_rate": 0.00019803394426499632, + "loss": 2.6694, + "step": 3061 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.00019803262044567626, + "loss": 2.6643, + "step": 3062 + }, + { + "epoch": 0.13, + "grad_norm": 0.4296875, + "learning_rate": 0.0001980312961852452, + "loss": 2.6511, + "step": 3063 + }, + { + "epoch": 0.13, + "grad_norm": 0.404296875, + "learning_rate": 0.00019802997148370916, + "loss": 2.5074, + "step": 3064 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.00019802864634107408, + "loss": 2.7277, + "step": 3065 + }, + { + "epoch": 0.13, + "grad_norm": 0.4375, + "learning_rate": 0.0001980273207573459, + "loss": 2.8275, + "step": 3066 + }, + { + "epoch": 0.13, + "grad_norm": 0.50390625, + "learning_rate": 0.00019802599473253061, + "loss": 2.8086, + "step": 3067 + }, + { + "epoch": 0.13, + "grad_norm": 0.470703125, + "learning_rate": 0.00019802466826663415, + "loss": 2.6828, + "step": 3068 + }, + { + "epoch": 0.13, + "grad_norm": 0.392578125, + "learning_rate": 0.00019802334135966254, + "loss": 2.6329, + "step": 3069 + }, + { + "epoch": 0.13, + "grad_norm": 0.416015625, + "learning_rate": 0.0001980220140116217, + "loss": 2.6804, + "step": 3070 + }, + { + "epoch": 0.13, + "grad_norm": 0.416015625, + "learning_rate": 0.00019802068622251762, + "loss": 2.6143, + "step": 3071 + }, + { + "epoch": 0.13, + "grad_norm": 0.51953125, + "learning_rate": 0.00019801935799235626, + "loss": 2.7302, + "step": 3072 + }, + { + "epoch": 0.13, + "grad_norm": 0.43359375, + "learning_rate": 0.00019801802932114363, + "loss": 2.8454, + "step": 3073 + }, + { + "epoch": 0.13, + "grad_norm": 0.439453125, + "learning_rate": 0.0001980167002088857, + "loss": 2.7245, + "step": 3074 + }, + { + "epoch": 0.13, + "grad_norm": 0.435546875, + "learning_rate": 0.00019801537065558842, + "loss": 2.6673, + "step": 3075 + }, + { + "epoch": 0.13, + "grad_norm": 0.4453125, + "learning_rate": 0.0001980140406612578, + "loss": 2.7201, + "step": 3076 + }, + { + "epoch": 0.13, + "grad_norm": 0.416015625, + "learning_rate": 0.00019801271022589983, + "loss": 2.7024, + "step": 3077 + }, + { + "epoch": 0.13, + "grad_norm": 0.41015625, + "learning_rate": 0.00019801137934952046, + "loss": 2.6887, + "step": 3078 + }, + { + "epoch": 0.13, + "grad_norm": 0.44140625, + "learning_rate": 0.00019801004803212568, + "loss": 2.7091, + "step": 3079 + }, + { + "epoch": 0.13, + "grad_norm": 0.421875, + "learning_rate": 0.00019800871627372155, + "loss": 2.816, + "step": 3080 + }, + { + "epoch": 0.13, + "grad_norm": 0.40234375, + "learning_rate": 0.000198007384074314, + "loss": 2.6309, + "step": 3081 + }, + { + "epoch": 0.13, + "grad_norm": 0.453125, + "learning_rate": 0.00019800605143390905, + "loss": 2.5605, + "step": 3082 + }, + { + "epoch": 0.13, + "grad_norm": 0.408203125, + "learning_rate": 0.00019800471835251267, + "loss": 2.6355, + "step": 3083 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.00019800338483013085, + "loss": 2.6635, + "step": 3084 + }, + { + "epoch": 0.13, + "grad_norm": 0.435546875, + "learning_rate": 0.00019800205086676966, + "loss": 2.6981, + "step": 3085 + }, + { + "epoch": 0.13, + "grad_norm": 0.419921875, + "learning_rate": 0.00019800071646243502, + "loss": 2.7488, + "step": 3086 + }, + { + "epoch": 0.13, + "grad_norm": 0.419921875, + "learning_rate": 0.000197999381617133, + "loss": 2.5554, + "step": 3087 + }, + { + "epoch": 0.13, + "grad_norm": 0.435546875, + "learning_rate": 0.00019799804633086957, + "loss": 2.5722, + "step": 3088 + }, + { + "epoch": 0.13, + "grad_norm": 0.421875, + "learning_rate": 0.00019799671060365074, + "loss": 2.7047, + "step": 3089 + }, + { + "epoch": 0.13, + "grad_norm": 0.3671875, + "learning_rate": 0.00019799537443548253, + "loss": 2.6218, + "step": 3090 + }, + { + "epoch": 0.13, + "grad_norm": 0.431640625, + "learning_rate": 0.00019799403782637094, + "loss": 2.6455, + "step": 3091 + }, + { + "epoch": 0.13, + "grad_norm": 0.4609375, + "learning_rate": 0.000197992700776322, + "loss": 2.7678, + "step": 3092 + }, + { + "epoch": 0.13, + "grad_norm": 0.45703125, + "learning_rate": 0.0001979913632853417, + "loss": 2.6786, + "step": 3093 + }, + { + "epoch": 0.13, + "grad_norm": 0.4296875, + "learning_rate": 0.00019799002535343609, + "loss": 2.7055, + "step": 3094 + }, + { + "epoch": 0.13, + "grad_norm": 0.431640625, + "learning_rate": 0.00019798868698061122, + "loss": 2.7803, + "step": 3095 + }, + { + "epoch": 0.13, + "grad_norm": 0.412109375, + "learning_rate": 0.00019798734816687302, + "loss": 2.6504, + "step": 3096 + }, + { + "epoch": 0.13, + "grad_norm": 0.42578125, + "learning_rate": 0.00019798600891222755, + "loss": 2.5697, + "step": 3097 + }, + { + "epoch": 0.13, + "grad_norm": 0.40234375, + "learning_rate": 0.00019798466921668088, + "loss": 2.6623, + "step": 3098 + }, + { + "epoch": 0.13, + "grad_norm": 0.5078125, + "learning_rate": 0.000197983329080239, + "loss": 2.6992, + "step": 3099 + }, + { + "epoch": 0.13, + "grad_norm": 0.455078125, + "learning_rate": 0.00019798198850290793, + "loss": 2.7103, + "step": 3100 + }, + { + "epoch": 0.13, + "grad_norm": 0.408203125, + "learning_rate": 0.0001979806474846938, + "loss": 2.647, + "step": 3101 + }, + { + "epoch": 0.13, + "grad_norm": 0.48828125, + "learning_rate": 0.00019797930602560247, + "loss": 2.6828, + "step": 3102 + }, + { + "epoch": 0.13, + "grad_norm": 0.39453125, + "learning_rate": 0.00019797796412564011, + "loss": 2.6349, + "step": 3103 + }, + { + "epoch": 0.13, + "grad_norm": 0.380859375, + "learning_rate": 0.0001979766217848127, + "loss": 2.526, + "step": 3104 + }, + { + "epoch": 0.13, + "grad_norm": 0.4453125, + "learning_rate": 0.00019797527900312632, + "loss": 2.6722, + "step": 3105 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.000197973935780587, + "loss": 2.6977, + "step": 3106 + }, + { + "epoch": 0.13, + "grad_norm": 0.4140625, + "learning_rate": 0.00019797259211720074, + "loss": 2.6417, + "step": 3107 + }, + { + "epoch": 0.13, + "grad_norm": 0.4453125, + "learning_rate": 0.00019797124801297362, + "loss": 2.8432, + "step": 3108 + }, + { + "epoch": 0.13, + "grad_norm": 0.400390625, + "learning_rate": 0.00019796990346791172, + "loss": 2.701, + "step": 3109 + }, + { + "epoch": 0.13, + "grad_norm": 0.455078125, + "learning_rate": 0.00019796855848202105, + "loss": 2.7102, + "step": 3110 + }, + { + "epoch": 0.13, + "grad_norm": 0.439453125, + "learning_rate": 0.00019796721305530765, + "loss": 2.7535, + "step": 3111 + }, + { + "epoch": 0.13, + "grad_norm": 0.404296875, + "learning_rate": 0.00019796586718777762, + "loss": 2.7123, + "step": 3112 + }, + { + "epoch": 0.13, + "grad_norm": 0.404296875, + "learning_rate": 0.00019796452087943696, + "loss": 2.6255, + "step": 3113 + }, + { + "epoch": 0.13, + "grad_norm": 0.408203125, + "learning_rate": 0.00019796317413029178, + "loss": 2.7144, + "step": 3114 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.00019796182694034812, + "loss": 2.6522, + "step": 3115 + }, + { + "epoch": 0.13, + "grad_norm": 0.396484375, + "learning_rate": 0.000197960479309612, + "loss": 2.6192, + "step": 3116 + }, + { + "epoch": 0.13, + "grad_norm": 0.458984375, + "learning_rate": 0.0001979591312380896, + "loss": 2.5994, + "step": 3117 + }, + { + "epoch": 0.13, + "grad_norm": 0.435546875, + "learning_rate": 0.00019795778272578684, + "loss": 2.5942, + "step": 3118 + }, + { + "epoch": 0.13, + "grad_norm": 0.431640625, + "learning_rate": 0.0001979564337727099, + "loss": 2.6394, + "step": 3119 + }, + { + "epoch": 0.13, + "grad_norm": 0.451171875, + "learning_rate": 0.00019795508437886477, + "loss": 2.6738, + "step": 3120 + }, + { + "epoch": 0.13, + "grad_norm": 0.45703125, + "learning_rate": 0.0001979537345442576, + "loss": 2.8327, + "step": 3121 + }, + { + "epoch": 0.13, + "grad_norm": 0.455078125, + "learning_rate": 0.0001979523842688944, + "loss": 2.6435, + "step": 3122 + }, + { + "epoch": 0.13, + "grad_norm": 0.443359375, + "learning_rate": 0.00019795103355278127, + "loss": 2.6267, + "step": 3123 + }, + { + "epoch": 0.13, + "grad_norm": 0.484375, + "learning_rate": 0.0001979496823959243, + "loss": 2.7308, + "step": 3124 + }, + { + "epoch": 0.13, + "grad_norm": 0.42578125, + "learning_rate": 0.0001979483307983295, + "loss": 2.7333, + "step": 3125 + }, + { + "epoch": 0.13, + "grad_norm": 0.4296875, + "learning_rate": 0.00019794697876000306, + "loss": 2.6045, + "step": 3126 + }, + { + "epoch": 0.13, + "grad_norm": 0.46875, + "learning_rate": 0.00019794562628095103, + "loss": 2.6905, + "step": 3127 + }, + { + "epoch": 0.13, + "grad_norm": 0.369140625, + "learning_rate": 0.00019794427336117942, + "loss": 2.8275, + "step": 3128 + }, + { + "epoch": 0.13, + "grad_norm": 0.412109375, + "learning_rate": 0.0001979429200006944, + "loss": 2.713, + "step": 3129 + }, + { + "epoch": 0.13, + "grad_norm": 0.423828125, + "learning_rate": 0.00019794156619950203, + "loss": 2.7209, + "step": 3130 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.0001979402119576084, + "loss": 2.8027, + "step": 3131 + }, + { + "epoch": 0.13, + "grad_norm": 0.431640625, + "learning_rate": 0.00019793885727501965, + "loss": 2.755, + "step": 3132 + }, + { + "epoch": 0.13, + "grad_norm": 0.431640625, + "learning_rate": 0.0001979375021517418, + "loss": 2.7504, + "step": 3133 + }, + { + "epoch": 0.13, + "grad_norm": 0.46484375, + "learning_rate": 0.00019793614658778096, + "loss": 2.6011, + "step": 3134 + }, + { + "epoch": 0.13, + "grad_norm": 0.44140625, + "learning_rate": 0.00019793479058314328, + "loss": 2.6934, + "step": 3135 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.00019793343413783486, + "loss": 2.6002, + "step": 3136 + }, + { + "epoch": 0.13, + "grad_norm": 0.41796875, + "learning_rate": 0.00019793207725186173, + "loss": 2.7093, + "step": 3137 + }, + { + "epoch": 0.13, + "grad_norm": 0.466796875, + "learning_rate": 0.00019793071992523008, + "loss": 2.6688, + "step": 3138 + }, + { + "epoch": 0.13, + "grad_norm": 0.427734375, + "learning_rate": 0.00019792936215794595, + "loss": 2.7202, + "step": 3139 + }, + { + "epoch": 0.13, + "grad_norm": 0.419921875, + "learning_rate": 0.0001979280039500155, + "loss": 2.7878, + "step": 3140 + }, + { + "epoch": 0.13, + "grad_norm": 0.453125, + "learning_rate": 0.0001979266453014448, + "loss": 2.6998, + "step": 3141 + }, + { + "epoch": 0.13, + "grad_norm": 0.412109375, + "learning_rate": 0.00019792528621224, + "loss": 2.7195, + "step": 3142 + }, + { + "epoch": 0.13, + "grad_norm": 0.439453125, + "learning_rate": 0.0001979239266824072, + "loss": 2.7637, + "step": 3143 + }, + { + "epoch": 0.13, + "grad_norm": 0.39453125, + "learning_rate": 0.00019792256671195252, + "loss": 2.6543, + "step": 3144 + }, + { + "epoch": 0.13, + "grad_norm": 0.419921875, + "learning_rate": 0.00019792120630088205, + "loss": 2.7923, + "step": 3145 + }, + { + "epoch": 0.13, + "grad_norm": 0.451171875, + "learning_rate": 0.00019791984544920198, + "loss": 2.5437, + "step": 3146 + }, + { + "epoch": 0.13, + "grad_norm": 0.37890625, + "learning_rate": 0.00019791848415691833, + "loss": 2.7326, + "step": 3147 + }, + { + "epoch": 0.13, + "grad_norm": 0.4609375, + "learning_rate": 0.00019791712242403736, + "loss": 2.6403, + "step": 3148 + }, + { + "epoch": 0.13, + "grad_norm": 0.4375, + "learning_rate": 0.00019791576025056506, + "loss": 2.6774, + "step": 3149 + }, + { + "epoch": 0.13, + "grad_norm": 0.380859375, + "learning_rate": 0.00019791439763650764, + "loss": 2.7419, + "step": 3150 + }, + { + "epoch": 0.13, + "grad_norm": 0.35546875, + "learning_rate": 0.00019791303458187125, + "loss": 2.7001, + "step": 3151 + }, + { + "epoch": 0.13, + "grad_norm": 0.421875, + "learning_rate": 0.00019791167108666199, + "loss": 2.7174, + "step": 3152 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.00019791030715088595, + "loss": 2.5743, + "step": 3153 + }, + { + "epoch": 0.13, + "grad_norm": 0.44140625, + "learning_rate": 0.0001979089427745493, + "loss": 2.622, + "step": 3154 + }, + { + "epoch": 0.13, + "grad_norm": 0.412109375, + "learning_rate": 0.00019790757795765823, + "loss": 2.7224, + "step": 3155 + }, + { + "epoch": 0.13, + "grad_norm": 0.392578125, + "learning_rate": 0.00019790621270021882, + "loss": 2.629, + "step": 3156 + }, + { + "epoch": 0.13, + "grad_norm": 0.384765625, + "learning_rate": 0.00019790484700223725, + "loss": 2.6893, + "step": 3157 + }, + { + "epoch": 0.13, + "grad_norm": 0.3828125, + "learning_rate": 0.00019790348086371964, + "loss": 2.6806, + "step": 3158 + }, + { + "epoch": 0.13, + "grad_norm": 0.40234375, + "learning_rate": 0.00019790211428467213, + "loss": 2.7204, + "step": 3159 + }, + { + "epoch": 0.13, + "grad_norm": 0.423828125, + "learning_rate": 0.0001979007472651009, + "loss": 2.6733, + "step": 3160 + }, + { + "epoch": 0.13, + "grad_norm": 0.431640625, + "learning_rate": 0.00019789937980501206, + "loss": 2.7377, + "step": 3161 + }, + { + "epoch": 0.13, + "grad_norm": 0.458984375, + "learning_rate": 0.00019789801190441185, + "loss": 2.7888, + "step": 3162 + }, + { + "epoch": 0.13, + "grad_norm": 0.3984375, + "learning_rate": 0.00019789664356330633, + "loss": 2.6796, + "step": 3163 + }, + { + "epoch": 0.13, + "grad_norm": 0.404296875, + "learning_rate": 0.0001978952747817017, + "loss": 2.5668, + "step": 3164 + }, + { + "epoch": 0.13, + "grad_norm": 0.443359375, + "learning_rate": 0.00019789390555960407, + "loss": 2.5821, + "step": 3165 + }, + { + "epoch": 0.13, + "grad_norm": 0.416015625, + "learning_rate": 0.00019789253589701965, + "loss": 2.6705, + "step": 3166 + }, + { + "epoch": 0.13, + "grad_norm": 0.37890625, + "learning_rate": 0.00019789116579395464, + "loss": 2.6375, + "step": 3167 + }, + { + "epoch": 0.13, + "grad_norm": 0.373046875, + "learning_rate": 0.0001978897952504151, + "loss": 2.6209, + "step": 3168 + }, + { + "epoch": 0.14, + "grad_norm": 0.3671875, + "learning_rate": 0.0001978884242664073, + "loss": 2.544, + "step": 3169 + }, + { + "epoch": 0.14, + "grad_norm": 0.419921875, + "learning_rate": 0.0001978870528419374, + "loss": 2.6992, + "step": 3170 + }, + { + "epoch": 0.14, + "grad_norm": 0.48828125, + "learning_rate": 0.00019788568097701145, + "loss": 2.7018, + "step": 3171 + }, + { + "epoch": 0.14, + "grad_norm": 0.3984375, + "learning_rate": 0.00019788430867163576, + "loss": 2.7871, + "step": 3172 + }, + { + "epoch": 0.14, + "grad_norm": 0.447265625, + "learning_rate": 0.00019788293592581645, + "loss": 2.7832, + "step": 3173 + }, + { + "epoch": 0.14, + "grad_norm": 0.41796875, + "learning_rate": 0.0001978815627395597, + "loss": 2.5834, + "step": 3174 + }, + { + "epoch": 0.14, + "grad_norm": 0.439453125, + "learning_rate": 0.00019788018911287165, + "loss": 2.6979, + "step": 3175 + }, + { + "epoch": 0.14, + "grad_norm": 0.390625, + "learning_rate": 0.00019787881504575855, + "loss": 2.7032, + "step": 3176 + }, + { + "epoch": 0.14, + "grad_norm": 0.40625, + "learning_rate": 0.00019787744053822656, + "loss": 2.6414, + "step": 3177 + }, + { + "epoch": 0.14, + "grad_norm": 0.46484375, + "learning_rate": 0.00019787606559028186, + "loss": 2.5933, + "step": 3178 + }, + { + "epoch": 0.14, + "grad_norm": 0.474609375, + "learning_rate": 0.0001978746902019306, + "loss": 2.7356, + "step": 3179 + }, + { + "epoch": 0.14, + "grad_norm": 0.3984375, + "learning_rate": 0.00019787331437317907, + "loss": 2.8122, + "step": 3180 + }, + { + "epoch": 0.14, + "grad_norm": 0.443359375, + "learning_rate": 0.00019787193810403333, + "loss": 2.6274, + "step": 3181 + }, + { + "epoch": 0.14, + "grad_norm": 0.412109375, + "learning_rate": 0.00019787056139449965, + "loss": 2.6727, + "step": 3182 + }, + { + "epoch": 0.14, + "grad_norm": 0.458984375, + "learning_rate": 0.00019786918424458423, + "loss": 2.7211, + "step": 3183 + }, + { + "epoch": 0.14, + "grad_norm": 0.431640625, + "learning_rate": 0.00019786780665429326, + "loss": 2.6155, + "step": 3184 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.00019786642862363286, + "loss": 2.6632, + "step": 3185 + }, + { + "epoch": 0.14, + "grad_norm": 0.384765625, + "learning_rate": 0.00019786505015260936, + "loss": 2.7063, + "step": 3186 + }, + { + "epoch": 0.14, + "grad_norm": 0.48046875, + "learning_rate": 0.0001978636712412289, + "loss": 2.6877, + "step": 3187 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019786229188949766, + "loss": 2.5791, + "step": 3188 + }, + { + "epoch": 0.14, + "grad_norm": 0.396484375, + "learning_rate": 0.00019786091209742187, + "loss": 2.7383, + "step": 3189 + }, + { + "epoch": 0.14, + "grad_norm": 0.462890625, + "learning_rate": 0.00019785953186500776, + "loss": 2.6805, + "step": 3190 + }, + { + "epoch": 0.14, + "grad_norm": 0.3828125, + "learning_rate": 0.0001978581511922615, + "loss": 2.5919, + "step": 3191 + }, + { + "epoch": 0.14, + "grad_norm": 0.490234375, + "learning_rate": 0.00019785677007918935, + "loss": 2.6406, + "step": 3192 + }, + { + "epoch": 0.14, + "grad_norm": 0.44921875, + "learning_rate": 0.00019785538852579748, + "loss": 2.7774, + "step": 3193 + }, + { + "epoch": 0.14, + "grad_norm": 0.41796875, + "learning_rate": 0.0001978540065320921, + "loss": 2.6878, + "step": 3194 + }, + { + "epoch": 0.14, + "grad_norm": 0.443359375, + "learning_rate": 0.0001978526240980795, + "loss": 2.7994, + "step": 3195 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019785124122376582, + "loss": 2.6944, + "step": 3196 + }, + { + "epoch": 0.14, + "grad_norm": 0.3984375, + "learning_rate": 0.0001978498579091573, + "loss": 2.6796, + "step": 3197 + }, + { + "epoch": 0.14, + "grad_norm": 0.4375, + "learning_rate": 0.0001978484741542602, + "loss": 2.5977, + "step": 3198 + }, + { + "epoch": 0.14, + "grad_norm": 0.451171875, + "learning_rate": 0.00019784708995908076, + "loss": 2.7334, + "step": 3199 + }, + { + "epoch": 0.14, + "grad_norm": 0.423828125, + "learning_rate": 0.00019784570532362512, + "loss": 2.7479, + "step": 3200 + }, + { + "epoch": 0.14, + "grad_norm": 0.44140625, + "learning_rate": 0.0001978443202478996, + "loss": 2.616, + "step": 3201 + }, + { + "epoch": 0.14, + "grad_norm": 0.3828125, + "learning_rate": 0.00019784293473191036, + "loss": 2.712, + "step": 3202 + }, + { + "epoch": 0.14, + "grad_norm": 0.439453125, + "learning_rate": 0.0001978415487756637, + "loss": 2.6435, + "step": 3203 + }, + { + "epoch": 0.14, + "grad_norm": 0.388671875, + "learning_rate": 0.00019784016237916582, + "loss": 2.64, + "step": 3204 + }, + { + "epoch": 0.14, + "grad_norm": 0.427734375, + "learning_rate": 0.00019783877554242296, + "loss": 2.7632, + "step": 3205 + }, + { + "epoch": 0.14, + "grad_norm": 0.466796875, + "learning_rate": 0.0001978373882654414, + "loss": 2.5894, + "step": 3206 + }, + { + "epoch": 0.14, + "grad_norm": 0.419921875, + "learning_rate": 0.0001978360005482273, + "loss": 2.6816, + "step": 3207 + }, + { + "epoch": 0.14, + "grad_norm": 0.453125, + "learning_rate": 0.00019783461239078698, + "loss": 2.5766, + "step": 3208 + }, + { + "epoch": 0.14, + "grad_norm": 0.41796875, + "learning_rate": 0.00019783322379312665, + "loss": 2.6288, + "step": 3209 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.0001978318347552526, + "loss": 2.6505, + "step": 3210 + }, + { + "epoch": 0.14, + "grad_norm": 0.373046875, + "learning_rate": 0.00019783044527717098, + "loss": 2.7752, + "step": 3211 + }, + { + "epoch": 0.14, + "grad_norm": 0.453125, + "learning_rate": 0.00019782905535888813, + "loss": 2.623, + "step": 3212 + }, + { + "epoch": 0.14, + "grad_norm": 0.482421875, + "learning_rate": 0.00019782766500041032, + "loss": 2.8317, + "step": 3213 + }, + { + "epoch": 0.14, + "grad_norm": 0.4296875, + "learning_rate": 0.00019782627420174376, + "loss": 2.6107, + "step": 3214 + }, + { + "epoch": 0.14, + "grad_norm": 0.384765625, + "learning_rate": 0.0001978248829628947, + "loss": 2.6513, + "step": 3215 + }, + { + "epoch": 0.14, + "grad_norm": 0.41796875, + "learning_rate": 0.0001978234912838694, + "loss": 2.6261, + "step": 3216 + }, + { + "epoch": 0.14, + "grad_norm": 0.3984375, + "learning_rate": 0.00019782209916467417, + "loss": 2.6906, + "step": 3217 + }, + { + "epoch": 0.14, + "grad_norm": 0.41796875, + "learning_rate": 0.00019782070660531527, + "loss": 2.6789, + "step": 3218 + }, + { + "epoch": 0.14, + "grad_norm": 0.392578125, + "learning_rate": 0.0001978193136057989, + "loss": 2.6929, + "step": 3219 + }, + { + "epoch": 0.14, + "grad_norm": 0.388671875, + "learning_rate": 0.00019781792016613137, + "loss": 2.649, + "step": 3220 + }, + { + "epoch": 0.14, + "grad_norm": 0.470703125, + "learning_rate": 0.00019781652628631896, + "loss": 2.6841, + "step": 3221 + }, + { + "epoch": 0.14, + "grad_norm": 0.400390625, + "learning_rate": 0.00019781513196636793, + "loss": 2.6093, + "step": 3222 + }, + { + "epoch": 0.14, + "grad_norm": 0.390625, + "learning_rate": 0.00019781373720628454, + "loss": 2.7004, + "step": 3223 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019781234200607508, + "loss": 2.6662, + "step": 3224 + }, + { + "epoch": 0.14, + "grad_norm": 0.46875, + "learning_rate": 0.0001978109463657458, + "loss": 2.7062, + "step": 3225 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019780955028530305, + "loss": 2.7839, + "step": 3226 + }, + { + "epoch": 0.14, + "grad_norm": 0.439453125, + "learning_rate": 0.00019780815376475305, + "loss": 2.7088, + "step": 3227 + }, + { + "epoch": 0.14, + "grad_norm": 0.462890625, + "learning_rate": 0.00019780675680410213, + "loss": 2.6177, + "step": 3228 + }, + { + "epoch": 0.14, + "grad_norm": 0.427734375, + "learning_rate": 0.0001978053594033565, + "loss": 2.7777, + "step": 3229 + }, + { + "epoch": 0.14, + "grad_norm": 0.44140625, + "learning_rate": 0.00019780396156252253, + "loss": 2.6576, + "step": 3230 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019780256328160649, + "loss": 2.5592, + "step": 3231 + }, + { + "epoch": 0.14, + "grad_norm": 0.39453125, + "learning_rate": 0.00019780116456061465, + "loss": 2.6865, + "step": 3232 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.0001977997653995533, + "loss": 2.6478, + "step": 3233 + }, + { + "epoch": 0.14, + "grad_norm": 0.375, + "learning_rate": 0.00019779836579842874, + "loss": 2.6557, + "step": 3234 + }, + { + "epoch": 0.14, + "grad_norm": 0.373046875, + "learning_rate": 0.0001977969657572473, + "loss": 2.6313, + "step": 3235 + }, + { + "epoch": 0.14, + "grad_norm": 0.412109375, + "learning_rate": 0.00019779556527601525, + "loss": 2.7686, + "step": 3236 + }, + { + "epoch": 0.14, + "grad_norm": 0.408203125, + "learning_rate": 0.00019779416435473888, + "loss": 2.7332, + "step": 3237 + }, + { + "epoch": 0.14, + "grad_norm": 0.396484375, + "learning_rate": 0.0001977927629934245, + "loss": 2.6566, + "step": 3238 + }, + { + "epoch": 0.14, + "grad_norm": 0.40625, + "learning_rate": 0.00019779136119207845, + "loss": 2.6728, + "step": 3239 + }, + { + "epoch": 0.14, + "grad_norm": 0.328125, + "learning_rate": 0.00019778995895070698, + "loss": 2.6191, + "step": 3240 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019778855626931648, + "loss": 2.6703, + "step": 3241 + }, + { + "epoch": 0.14, + "grad_norm": 0.4609375, + "learning_rate": 0.00019778715314791318, + "loss": 2.8403, + "step": 3242 + }, + { + "epoch": 0.14, + "grad_norm": 0.345703125, + "learning_rate": 0.00019778574958650345, + "loss": 2.6001, + "step": 3243 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.00019778434558509358, + "loss": 2.6799, + "step": 3244 + }, + { + "epoch": 0.14, + "grad_norm": 0.375, + "learning_rate": 0.00019778294114368986, + "loss": 2.6747, + "step": 3245 + }, + { + "epoch": 0.14, + "grad_norm": 0.435546875, + "learning_rate": 0.00019778153626229868, + "loss": 2.748, + "step": 3246 + }, + { + "epoch": 0.14, + "grad_norm": 0.41015625, + "learning_rate": 0.0001977801309409263, + "loss": 2.7548, + "step": 3247 + }, + { + "epoch": 0.14, + "grad_norm": 0.400390625, + "learning_rate": 0.00019777872517957908, + "loss": 2.6171, + "step": 3248 + }, + { + "epoch": 0.14, + "grad_norm": 0.427734375, + "learning_rate": 0.00019777731897826332, + "loss": 2.669, + "step": 3249 + }, + { + "epoch": 0.14, + "grad_norm": 0.4921875, + "learning_rate": 0.00019777591233698536, + "loss": 2.6067, + "step": 3250 + }, + { + "epoch": 0.14, + "grad_norm": 0.43359375, + "learning_rate": 0.00019777450525575152, + "loss": 2.7202, + "step": 3251 + }, + { + "epoch": 0.14, + "grad_norm": 0.380859375, + "learning_rate": 0.00019777309773456815, + "loss": 2.6687, + "step": 3252 + }, + { + "epoch": 0.14, + "grad_norm": 0.421875, + "learning_rate": 0.00019777168977344159, + "loss": 2.7422, + "step": 3253 + }, + { + "epoch": 0.14, + "grad_norm": 0.45703125, + "learning_rate": 0.00019777028137237811, + "loss": 2.8683, + "step": 3254 + }, + { + "epoch": 0.14, + "grad_norm": 0.38671875, + "learning_rate": 0.0001977688725313841, + "loss": 2.7104, + "step": 3255 + }, + { + "epoch": 0.14, + "grad_norm": 0.365234375, + "learning_rate": 0.00019776746325046593, + "loss": 2.6001, + "step": 3256 + }, + { + "epoch": 0.14, + "grad_norm": 0.365234375, + "learning_rate": 0.0001977660535296299, + "loss": 2.6397, + "step": 3257 + }, + { + "epoch": 0.14, + "grad_norm": 0.423828125, + "learning_rate": 0.00019776464336888235, + "loss": 2.6101, + "step": 3258 + }, + { + "epoch": 0.14, + "grad_norm": 0.474609375, + "learning_rate": 0.00019776323276822963, + "loss": 2.7563, + "step": 3259 + }, + { + "epoch": 0.14, + "grad_norm": 0.43359375, + "learning_rate": 0.0001977618217276781, + "loss": 2.8316, + "step": 3260 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019776041024723408, + "loss": 2.623, + "step": 3261 + }, + { + "epoch": 0.14, + "grad_norm": 0.390625, + "learning_rate": 0.00019775899832690397, + "loss": 2.7635, + "step": 3262 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019775758596669407, + "loss": 2.8598, + "step": 3263 + }, + { + "epoch": 0.14, + "grad_norm": 0.3828125, + "learning_rate": 0.0001977561731666108, + "loss": 2.7696, + "step": 3264 + }, + { + "epoch": 0.14, + "grad_norm": 0.40234375, + "learning_rate": 0.00019775475992666043, + "loss": 2.579, + "step": 3265 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019775334624684936, + "loss": 2.6673, + "step": 3266 + }, + { + "epoch": 0.14, + "grad_norm": 0.39453125, + "learning_rate": 0.000197751932127184, + "loss": 2.7475, + "step": 3267 + }, + { + "epoch": 0.14, + "grad_norm": 0.384765625, + "learning_rate": 0.00019775051756767064, + "loss": 2.5659, + "step": 3268 + }, + { + "epoch": 0.14, + "grad_norm": 0.474609375, + "learning_rate": 0.00019774910256831568, + "loss": 2.7018, + "step": 3269 + }, + { + "epoch": 0.14, + "grad_norm": 0.431640625, + "learning_rate": 0.00019774768712912548, + "loss": 2.7506, + "step": 3270 + }, + { + "epoch": 0.14, + "grad_norm": 0.4765625, + "learning_rate": 0.00019774627125010642, + "loss": 2.6992, + "step": 3271 + }, + { + "epoch": 0.14, + "grad_norm": 0.384765625, + "learning_rate": 0.00019774485493126486, + "loss": 2.6394, + "step": 3272 + }, + { + "epoch": 0.14, + "grad_norm": 0.474609375, + "learning_rate": 0.00019774343817260717, + "loss": 2.6753, + "step": 3273 + }, + { + "epoch": 0.14, + "grad_norm": 0.423828125, + "learning_rate": 0.0001977420209741397, + "loss": 2.6574, + "step": 3274 + }, + { + "epoch": 0.14, + "grad_norm": 0.375, + "learning_rate": 0.0001977406033358689, + "loss": 2.5771, + "step": 3275 + }, + { + "epoch": 0.14, + "grad_norm": 0.443359375, + "learning_rate": 0.0001977391852578011, + "loss": 2.5701, + "step": 3276 + }, + { + "epoch": 0.14, + "grad_norm": 0.37109375, + "learning_rate": 0.00019773776673994266, + "loss": 2.6894, + "step": 3277 + }, + { + "epoch": 0.14, + "grad_norm": 0.373046875, + "learning_rate": 0.00019773634778230003, + "loss": 2.7504, + "step": 3278 + }, + { + "epoch": 0.14, + "grad_norm": 0.466796875, + "learning_rate": 0.00019773492838487953, + "loss": 2.7278, + "step": 3279 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.00019773350854768756, + "loss": 2.6455, + "step": 3280 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.00019773208827073054, + "loss": 2.6451, + "step": 3281 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019773066755401483, + "loss": 2.7157, + "step": 3282 + }, + { + "epoch": 0.14, + "grad_norm": 0.408203125, + "learning_rate": 0.00019772924639754682, + "loss": 2.6643, + "step": 3283 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.00019772782480133292, + "loss": 2.7172, + "step": 3284 + }, + { + "epoch": 0.14, + "grad_norm": 0.34765625, + "learning_rate": 0.00019772640276537954, + "loss": 2.673, + "step": 3285 + }, + { + "epoch": 0.14, + "grad_norm": 0.3671875, + "learning_rate": 0.00019772498028969307, + "loss": 2.6183, + "step": 3286 + }, + { + "epoch": 0.14, + "grad_norm": 0.365234375, + "learning_rate": 0.00019772355737427989, + "loss": 2.573, + "step": 3287 + }, + { + "epoch": 0.14, + "grad_norm": 0.40234375, + "learning_rate": 0.00019772213401914642, + "loss": 2.6768, + "step": 3288 + }, + { + "epoch": 0.14, + "grad_norm": 0.37890625, + "learning_rate": 0.00019772071022429902, + "loss": 2.6861, + "step": 3289 + }, + { + "epoch": 0.14, + "grad_norm": 0.39453125, + "learning_rate": 0.00019771928598974418, + "loss": 2.7339, + "step": 3290 + }, + { + "epoch": 0.14, + "grad_norm": 0.44140625, + "learning_rate": 0.00019771786131548826, + "loss": 2.6468, + "step": 3291 + }, + { + "epoch": 0.14, + "grad_norm": 0.435546875, + "learning_rate": 0.00019771643620153766, + "loss": 2.7964, + "step": 3292 + }, + { + "epoch": 0.14, + "grad_norm": 0.388671875, + "learning_rate": 0.00019771501064789882, + "loss": 2.594, + "step": 3293 + }, + { + "epoch": 0.14, + "grad_norm": 0.427734375, + "learning_rate": 0.00019771358465457816, + "loss": 2.6436, + "step": 3294 + }, + { + "epoch": 0.14, + "grad_norm": 0.40625, + "learning_rate": 0.00019771215822158206, + "loss": 2.7355, + "step": 3295 + }, + { + "epoch": 0.14, + "grad_norm": 0.40234375, + "learning_rate": 0.00019771073134891695, + "loss": 2.6781, + "step": 3296 + }, + { + "epoch": 0.14, + "grad_norm": 0.421875, + "learning_rate": 0.00019770930403658927, + "loss": 2.8342, + "step": 3297 + }, + { + "epoch": 0.14, + "grad_norm": 0.4375, + "learning_rate": 0.00019770787628460543, + "loss": 2.5697, + "step": 3298 + }, + { + "epoch": 0.14, + "grad_norm": 0.431640625, + "learning_rate": 0.00019770644809297185, + "loss": 2.6197, + "step": 3299 + }, + { + "epoch": 0.14, + "grad_norm": 0.36328125, + "learning_rate": 0.00019770501946169496, + "loss": 2.648, + "step": 3300 + }, + { + "epoch": 0.14, + "grad_norm": 0.404296875, + "learning_rate": 0.0001977035903907812, + "loss": 2.4808, + "step": 3301 + }, + { + "epoch": 0.14, + "grad_norm": 0.41796875, + "learning_rate": 0.000197702160880237, + "loss": 2.6508, + "step": 3302 + }, + { + "epoch": 0.14, + "grad_norm": 0.4609375, + "learning_rate": 0.00019770073093006876, + "loss": 2.6036, + "step": 3303 + }, + { + "epoch": 0.14, + "grad_norm": 0.423828125, + "learning_rate": 0.00019769930054028295, + "loss": 2.8816, + "step": 3304 + }, + { + "epoch": 0.14, + "grad_norm": 0.4453125, + "learning_rate": 0.00019769786971088595, + "loss": 2.5932, + "step": 3305 + }, + { + "epoch": 0.14, + "grad_norm": 0.435546875, + "learning_rate": 0.00019769643844188428, + "loss": 2.7031, + "step": 3306 + }, + { + "epoch": 0.14, + "grad_norm": 0.419921875, + "learning_rate": 0.00019769500673328435, + "loss": 2.5832, + "step": 3307 + }, + { + "epoch": 0.14, + "grad_norm": 0.388671875, + "learning_rate": 0.00019769357458509256, + "loss": 2.651, + "step": 3308 + }, + { + "epoch": 0.14, + "grad_norm": 0.453125, + "learning_rate": 0.00019769214199731543, + "loss": 2.6416, + "step": 3309 + }, + { + "epoch": 0.14, + "grad_norm": 0.486328125, + "learning_rate": 0.00019769070896995933, + "loss": 2.7787, + "step": 3310 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019768927550303076, + "loss": 2.5383, + "step": 3311 + }, + { + "epoch": 0.14, + "grad_norm": 0.470703125, + "learning_rate": 0.00019768784159653616, + "loss": 2.6576, + "step": 3312 + }, + { + "epoch": 0.14, + "grad_norm": 0.41796875, + "learning_rate": 0.00019768640725048196, + "loss": 2.6589, + "step": 3313 + }, + { + "epoch": 0.14, + "grad_norm": 0.46875, + "learning_rate": 0.0001976849724648746, + "loss": 2.7573, + "step": 3314 + }, + { + "epoch": 0.14, + "grad_norm": 0.4765625, + "learning_rate": 0.0001976835372397206, + "loss": 2.6041, + "step": 3315 + }, + { + "epoch": 0.14, + "grad_norm": 0.47265625, + "learning_rate": 0.00019768210157502635, + "loss": 2.707, + "step": 3316 + }, + { + "epoch": 0.14, + "grad_norm": 0.453125, + "learning_rate": 0.00019768066547079839, + "loss": 2.7711, + "step": 3317 + }, + { + "epoch": 0.14, + "grad_norm": 0.4375, + "learning_rate": 0.0001976792289270431, + "loss": 2.6564, + "step": 3318 + }, + { + "epoch": 0.14, + "grad_norm": 0.443359375, + "learning_rate": 0.000197677791943767, + "loss": 2.7179, + "step": 3319 + }, + { + "epoch": 0.14, + "grad_norm": 0.447265625, + "learning_rate": 0.0001976763545209765, + "loss": 2.729, + "step": 3320 + }, + { + "epoch": 0.14, + "grad_norm": 0.40625, + "learning_rate": 0.0001976749166586781, + "loss": 2.7061, + "step": 3321 + }, + { + "epoch": 0.14, + "grad_norm": 0.45703125, + "learning_rate": 0.0001976734783568783, + "loss": 2.6517, + "step": 3322 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019767203961558352, + "loss": 2.73, + "step": 3323 + }, + { + "epoch": 0.14, + "grad_norm": 0.4609375, + "learning_rate": 0.0001976706004348003, + "loss": 2.6306, + "step": 3324 + }, + { + "epoch": 0.14, + "grad_norm": 0.5234375, + "learning_rate": 0.000197669160814535, + "loss": 2.725, + "step": 3325 + }, + { + "epoch": 0.14, + "grad_norm": 0.384765625, + "learning_rate": 0.00019766772075479423, + "loss": 2.5864, + "step": 3326 + }, + { + "epoch": 0.14, + "grad_norm": 0.46875, + "learning_rate": 0.00019766628025558438, + "loss": 2.73, + "step": 3327 + }, + { + "epoch": 0.14, + "grad_norm": 0.462890625, + "learning_rate": 0.000197664839316912, + "loss": 2.6147, + "step": 3328 + }, + { + "epoch": 0.14, + "grad_norm": 0.392578125, + "learning_rate": 0.0001976633979387835, + "loss": 2.638, + "step": 3329 + }, + { + "epoch": 0.14, + "grad_norm": 0.4921875, + "learning_rate": 0.00019766195612120542, + "loss": 2.6665, + "step": 3330 + }, + { + "epoch": 0.14, + "grad_norm": 0.365234375, + "learning_rate": 0.0001976605138641842, + "loss": 2.6248, + "step": 3331 + }, + { + "epoch": 0.14, + "grad_norm": 0.466796875, + "learning_rate": 0.00019765907116772637, + "loss": 2.7563, + "step": 3332 + }, + { + "epoch": 0.14, + "grad_norm": 0.435546875, + "learning_rate": 0.00019765762803183843, + "loss": 2.6524, + "step": 3333 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019765618445652681, + "loss": 2.7193, + "step": 3334 + }, + { + "epoch": 0.14, + "grad_norm": 0.466796875, + "learning_rate": 0.00019765474044179811, + "loss": 2.6427, + "step": 3335 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.00019765329598765873, + "loss": 2.7311, + "step": 3336 + }, + { + "epoch": 0.14, + "grad_norm": 0.439453125, + "learning_rate": 0.0001976518510941152, + "loss": 2.6537, + "step": 3337 + }, + { + "epoch": 0.14, + "grad_norm": 0.400390625, + "learning_rate": 0.00019765040576117407, + "loss": 2.7239, + "step": 3338 + }, + { + "epoch": 0.14, + "grad_norm": 0.359375, + "learning_rate": 0.00019764895998884176, + "loss": 2.7298, + "step": 3339 + }, + { + "epoch": 0.14, + "grad_norm": 0.423828125, + "learning_rate": 0.00019764751377712483, + "loss": 2.6804, + "step": 3340 + }, + { + "epoch": 0.14, + "grad_norm": 0.37890625, + "learning_rate": 0.00019764606712602975, + "loss": 2.49, + "step": 3341 + }, + { + "epoch": 0.14, + "grad_norm": 0.427734375, + "learning_rate": 0.0001976446200355631, + "loss": 2.7129, + "step": 3342 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.00019764317250573133, + "loss": 2.8225, + "step": 3343 + }, + { + "epoch": 0.14, + "grad_norm": 0.4375, + "learning_rate": 0.00019764172453654095, + "loss": 2.7927, + "step": 3344 + }, + { + "epoch": 0.14, + "grad_norm": 0.421875, + "learning_rate": 0.00019764027612799852, + "loss": 2.6337, + "step": 3345 + }, + { + "epoch": 0.14, + "grad_norm": 0.44140625, + "learning_rate": 0.0001976388272801105, + "loss": 2.6963, + "step": 3346 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019763737799288343, + "loss": 2.5469, + "step": 3347 + }, + { + "epoch": 0.14, + "grad_norm": 0.470703125, + "learning_rate": 0.00019763592826632388, + "loss": 2.5754, + "step": 3348 + }, + { + "epoch": 0.14, + "grad_norm": 0.451171875, + "learning_rate": 0.0001976344781004383, + "loss": 2.6975, + "step": 3349 + }, + { + "epoch": 0.14, + "grad_norm": 0.384765625, + "learning_rate": 0.0001976330274952333, + "loss": 2.6679, + "step": 3350 + }, + { + "epoch": 0.14, + "grad_norm": 0.466796875, + "learning_rate": 0.00019763157645071532, + "loss": 2.7875, + "step": 3351 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019763012496689094, + "loss": 2.6698, + "step": 3352 + }, + { + "epoch": 0.14, + "grad_norm": 0.416015625, + "learning_rate": 0.00019762867304376667, + "loss": 2.7118, + "step": 3353 + }, + { + "epoch": 0.14, + "grad_norm": 0.4765625, + "learning_rate": 0.00019762722068134903, + "loss": 2.7126, + "step": 3354 + }, + { + "epoch": 0.14, + "grad_norm": 0.40625, + "learning_rate": 0.00019762576787964463, + "loss": 2.6496, + "step": 3355 + }, + { + "epoch": 0.14, + "grad_norm": 0.43359375, + "learning_rate": 0.0001976243146386599, + "loss": 2.6154, + "step": 3356 + }, + { + "epoch": 0.14, + "grad_norm": 0.43359375, + "learning_rate": 0.00019762286095840144, + "loss": 2.5147, + "step": 3357 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.0001976214068388758, + "loss": 2.6699, + "step": 3358 + }, + { + "epoch": 0.14, + "grad_norm": 0.412109375, + "learning_rate": 0.00019761995228008948, + "loss": 2.6157, + "step": 3359 + }, + { + "epoch": 0.14, + "grad_norm": 0.4140625, + "learning_rate": 0.00019761849728204908, + "loss": 2.6635, + "step": 3360 + }, + { + "epoch": 0.14, + "grad_norm": 0.408203125, + "learning_rate": 0.0001976170418447611, + "loss": 2.7606, + "step": 3361 + }, + { + "epoch": 0.14, + "grad_norm": 0.40234375, + "learning_rate": 0.00019761558596823213, + "loss": 2.6786, + "step": 3362 + }, + { + "epoch": 0.14, + "grad_norm": 0.4296875, + "learning_rate": 0.00019761412965246867, + "loss": 2.6584, + "step": 3363 + }, + { + "epoch": 0.14, + "grad_norm": 0.40234375, + "learning_rate": 0.00019761267289747733, + "loss": 2.6786, + "step": 3364 + }, + { + "epoch": 0.14, + "grad_norm": 0.443359375, + "learning_rate": 0.00019761121570326462, + "loss": 2.5767, + "step": 3365 + }, + { + "epoch": 0.14, + "grad_norm": 0.37890625, + "learning_rate": 0.0001976097580698371, + "loss": 2.6652, + "step": 3366 + }, + { + "epoch": 0.14, + "grad_norm": 0.455078125, + "learning_rate": 0.00019760829999720137, + "loss": 2.6483, + "step": 3367 + }, + { + "epoch": 0.14, + "grad_norm": 0.345703125, + "learning_rate": 0.00019760684148536394, + "loss": 2.7043, + "step": 3368 + }, + { + "epoch": 0.14, + "grad_norm": 0.412109375, + "learning_rate": 0.0001976053825343314, + "loss": 2.7537, + "step": 3369 + }, + { + "epoch": 0.14, + "grad_norm": 0.373046875, + "learning_rate": 0.00019760392314411028, + "loss": 2.6212, + "step": 3370 + }, + { + "epoch": 0.14, + "grad_norm": 0.47265625, + "learning_rate": 0.00019760246331470722, + "loss": 2.6737, + "step": 3371 + }, + { + "epoch": 0.14, + "grad_norm": 0.486328125, + "learning_rate": 0.00019760100304612875, + "loss": 2.7118, + "step": 3372 + }, + { + "epoch": 0.14, + "grad_norm": 0.396484375, + "learning_rate": 0.0001975995423383814, + "loss": 2.5491, + "step": 3373 + }, + { + "epoch": 0.14, + "grad_norm": 0.373046875, + "learning_rate": 0.00019759808119147176, + "loss": 2.673, + "step": 3374 + }, + { + "epoch": 0.14, + "grad_norm": 0.447265625, + "learning_rate": 0.00019759661960540648, + "loss": 2.8053, + "step": 3375 + }, + { + "epoch": 0.14, + "grad_norm": 0.431640625, + "learning_rate": 0.00019759515758019204, + "loss": 2.6912, + "step": 3376 + }, + { + "epoch": 0.14, + "grad_norm": 0.3671875, + "learning_rate": 0.00019759369511583507, + "loss": 2.5201, + "step": 3377 + }, + { + "epoch": 0.14, + "grad_norm": 0.408203125, + "learning_rate": 0.00019759223221234214, + "loss": 2.6835, + "step": 3378 + }, + { + "epoch": 0.14, + "grad_norm": 0.47265625, + "learning_rate": 0.00019759076886971983, + "loss": 2.7421, + "step": 3379 + }, + { + "epoch": 0.14, + "grad_norm": 0.39453125, + "learning_rate": 0.00019758930508797474, + "loss": 2.564, + "step": 3380 + }, + { + "epoch": 0.14, + "grad_norm": 0.400390625, + "learning_rate": 0.00019758784086711342, + "loss": 2.486, + "step": 3381 + }, + { + "epoch": 0.14, + "grad_norm": 0.43359375, + "learning_rate": 0.00019758637620714248, + "loss": 2.7166, + "step": 3382 + }, + { + "epoch": 0.14, + "grad_norm": 0.40234375, + "learning_rate": 0.00019758491110806852, + "loss": 2.5332, + "step": 3383 + }, + { + "epoch": 0.14, + "grad_norm": 0.423828125, + "learning_rate": 0.00019758344556989813, + "loss": 2.6589, + "step": 3384 + }, + { + "epoch": 0.14, + "grad_norm": 0.392578125, + "learning_rate": 0.00019758197959263786, + "loss": 2.6937, + "step": 3385 + }, + { + "epoch": 0.14, + "grad_norm": 0.435546875, + "learning_rate": 0.0001975805131762944, + "loss": 2.6114, + "step": 3386 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019757904632087425, + "loss": 2.76, + "step": 3387 + }, + { + "epoch": 0.14, + "grad_norm": 0.43359375, + "learning_rate": 0.00019757757902638407, + "loss": 2.6557, + "step": 3388 + }, + { + "epoch": 0.14, + "grad_norm": 0.423828125, + "learning_rate": 0.00019757611129283045, + "loss": 2.6602, + "step": 3389 + }, + { + "epoch": 0.14, + "grad_norm": 0.42578125, + "learning_rate": 0.00019757464312021998, + "loss": 2.6466, + "step": 3390 + }, + { + "epoch": 0.14, + "grad_norm": 0.369140625, + "learning_rate": 0.00019757317450855928, + "loss": 2.6147, + "step": 3391 + }, + { + "epoch": 0.14, + "grad_norm": 0.40234375, + "learning_rate": 0.00019757170545785496, + "loss": 2.7371, + "step": 3392 + }, + { + "epoch": 0.14, + "grad_norm": 0.431640625, + "learning_rate": 0.00019757023596811361, + "loss": 2.7506, + "step": 3393 + }, + { + "epoch": 0.14, + "grad_norm": 0.408203125, + "learning_rate": 0.00019756876603934184, + "loss": 2.7011, + "step": 3394 + }, + { + "epoch": 0.14, + "grad_norm": 0.458984375, + "learning_rate": 0.00019756729567154633, + "loss": 2.6701, + "step": 3395 + }, + { + "epoch": 0.14, + "grad_norm": 0.482421875, + "learning_rate": 0.00019756582486473362, + "loss": 2.7364, + "step": 3396 + }, + { + "epoch": 0.14, + "grad_norm": 0.453125, + "learning_rate": 0.0001975643536189103, + "loss": 2.8153, + "step": 3397 + }, + { + "epoch": 0.14, + "grad_norm": 0.439453125, + "learning_rate": 0.0001975628819340831, + "loss": 2.7038, + "step": 3398 + }, + { + "epoch": 0.14, + "grad_norm": 0.44921875, + "learning_rate": 0.0001975614098102586, + "loss": 2.7342, + "step": 3399 + }, + { + "epoch": 0.14, + "grad_norm": 0.458984375, + "learning_rate": 0.0001975599372474434, + "loss": 2.6682, + "step": 3400 + }, + { + "epoch": 0.14, + "grad_norm": 0.375, + "learning_rate": 0.00019755846424564415, + "loss": 2.7093, + "step": 3401 + }, + { + "epoch": 0.14, + "grad_norm": 0.48828125, + "learning_rate": 0.00019755699080486742, + "loss": 2.7561, + "step": 3402 + }, + { + "epoch": 0.15, + "grad_norm": 0.44140625, + "learning_rate": 0.00019755551692511993, + "loss": 2.7948, + "step": 3403 + }, + { + "epoch": 0.15, + "grad_norm": 0.466796875, + "learning_rate": 0.00019755404260640824, + "loss": 2.7683, + "step": 3404 + }, + { + "epoch": 0.15, + "grad_norm": 0.41796875, + "learning_rate": 0.000197552567848739, + "loss": 2.7405, + "step": 3405 + }, + { + "epoch": 0.15, + "grad_norm": 0.49609375, + "learning_rate": 0.00019755109265211888, + "loss": 2.8028, + "step": 3406 + }, + { + "epoch": 0.15, + "grad_norm": 0.46484375, + "learning_rate": 0.0001975496170165545, + "loss": 2.9043, + "step": 3407 + }, + { + "epoch": 0.15, + "grad_norm": 0.44921875, + "learning_rate": 0.00019754814094205246, + "loss": 2.6087, + "step": 3408 + }, + { + "epoch": 0.15, + "grad_norm": 0.400390625, + "learning_rate": 0.00019754666442861946, + "loss": 2.4933, + "step": 3409 + }, + { + "epoch": 0.15, + "grad_norm": 0.3828125, + "learning_rate": 0.00019754518747626211, + "loss": 2.6847, + "step": 3410 + }, + { + "epoch": 0.15, + "grad_norm": 0.47265625, + "learning_rate": 0.00019754371008498707, + "loss": 2.7286, + "step": 3411 + }, + { + "epoch": 0.15, + "grad_norm": 0.404296875, + "learning_rate": 0.00019754223225480096, + "loss": 2.6759, + "step": 3412 + }, + { + "epoch": 0.15, + "grad_norm": 0.4453125, + "learning_rate": 0.0001975407539857105, + "loss": 2.5696, + "step": 3413 + }, + { + "epoch": 0.15, + "grad_norm": 0.427734375, + "learning_rate": 0.00019753927527772227, + "loss": 2.6368, + "step": 3414 + }, + { + "epoch": 0.15, + "grad_norm": 0.421875, + "learning_rate": 0.00019753779613084293, + "loss": 2.7095, + "step": 3415 + }, + { + "epoch": 0.15, + "grad_norm": 0.4453125, + "learning_rate": 0.00019753631654507918, + "loss": 2.6479, + "step": 3416 + }, + { + "epoch": 0.15, + "grad_norm": 0.474609375, + "learning_rate": 0.00019753483652043766, + "loss": 2.7674, + "step": 3417 + }, + { + "epoch": 0.15, + "grad_norm": 0.45703125, + "learning_rate": 0.00019753335605692497, + "loss": 2.7635, + "step": 3418 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.00019753187515454785, + "loss": 2.6653, + "step": 3419 + }, + { + "epoch": 0.15, + "grad_norm": 0.451171875, + "learning_rate": 0.00019753039381331295, + "loss": 2.6157, + "step": 3420 + }, + { + "epoch": 0.15, + "grad_norm": 0.46484375, + "learning_rate": 0.00019752891203322692, + "loss": 2.8191, + "step": 3421 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.0001975274298142964, + "loss": 2.7314, + "step": 3422 + }, + { + "epoch": 0.15, + "grad_norm": 0.3828125, + "learning_rate": 0.0001975259471565281, + "loss": 2.6435, + "step": 3423 + }, + { + "epoch": 0.15, + "grad_norm": 0.45703125, + "learning_rate": 0.0001975244640599287, + "loss": 2.6943, + "step": 3424 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.00019752298052450485, + "loss": 2.6303, + "step": 3425 + }, + { + "epoch": 0.15, + "grad_norm": 0.42578125, + "learning_rate": 0.0001975214965502632, + "loss": 2.6426, + "step": 3426 + }, + { + "epoch": 0.15, + "grad_norm": 0.439453125, + "learning_rate": 0.00019752001213721047, + "loss": 2.5363, + "step": 3427 + }, + { + "epoch": 0.15, + "grad_norm": 0.484375, + "learning_rate": 0.00019751852728535333, + "loss": 2.7774, + "step": 3428 + }, + { + "epoch": 0.15, + "grad_norm": 0.4765625, + "learning_rate": 0.00019751704199469846, + "loss": 2.8255, + "step": 3429 + }, + { + "epoch": 0.15, + "grad_norm": 0.486328125, + "learning_rate": 0.0001975155562652525, + "loss": 2.8258, + "step": 3430 + }, + { + "epoch": 0.15, + "grad_norm": 0.41015625, + "learning_rate": 0.00019751407009702224, + "loss": 2.4864, + "step": 3431 + }, + { + "epoch": 0.15, + "grad_norm": 0.42578125, + "learning_rate": 0.00019751258349001426, + "loss": 2.8476, + "step": 3432 + }, + { + "epoch": 0.15, + "grad_norm": 0.435546875, + "learning_rate": 0.00019751109644423532, + "loss": 2.7248, + "step": 3433 + }, + { + "epoch": 0.15, + "grad_norm": 0.435546875, + "learning_rate": 0.00019750960895969203, + "loss": 2.6814, + "step": 3434 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.0001975081210363912, + "loss": 2.7872, + "step": 3435 + }, + { + "epoch": 0.15, + "grad_norm": 0.48046875, + "learning_rate": 0.00019750663267433942, + "loss": 2.8251, + "step": 3436 + }, + { + "epoch": 0.15, + "grad_norm": 0.40234375, + "learning_rate": 0.00019750514387354345, + "loss": 2.7796, + "step": 3437 + }, + { + "epoch": 0.15, + "grad_norm": 0.423828125, + "learning_rate": 0.00019750365463400992, + "loss": 2.6598, + "step": 3438 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.00019750216495574563, + "loss": 2.5772, + "step": 3439 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.0001975006748387572, + "loss": 2.6594, + "step": 3440 + }, + { + "epoch": 0.15, + "grad_norm": 0.4453125, + "learning_rate": 0.00019749918428305137, + "loss": 2.7244, + "step": 3441 + }, + { + "epoch": 0.15, + "grad_norm": 0.419921875, + "learning_rate": 0.00019749769328863484, + "loss": 2.6177, + "step": 3442 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.00019749620185551434, + "loss": 2.8506, + "step": 3443 + }, + { + "epoch": 0.15, + "grad_norm": 0.453125, + "learning_rate": 0.00019749470998369658, + "loss": 2.7696, + "step": 3444 + }, + { + "epoch": 0.15, + "grad_norm": 0.427734375, + "learning_rate": 0.0001974932176731882, + "loss": 2.6933, + "step": 3445 + }, + { + "epoch": 0.15, + "grad_norm": 0.44140625, + "learning_rate": 0.000197491724923996, + "loss": 2.8397, + "step": 3446 + }, + { + "epoch": 0.15, + "grad_norm": 0.447265625, + "learning_rate": 0.0001974902317361267, + "loss": 2.7858, + "step": 3447 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.00019748873810958694, + "loss": 2.6538, + "step": 3448 + }, + { + "epoch": 0.15, + "grad_norm": 0.421875, + "learning_rate": 0.00019748724404438349, + "loss": 2.8236, + "step": 3449 + }, + { + "epoch": 0.15, + "grad_norm": 0.46484375, + "learning_rate": 0.00019748574954052308, + "loss": 2.8557, + "step": 3450 + }, + { + "epoch": 0.15, + "grad_norm": 0.484375, + "learning_rate": 0.0001974842545980124, + "loss": 2.7471, + "step": 3451 + }, + { + "epoch": 0.15, + "grad_norm": 0.4140625, + "learning_rate": 0.0001974827592168582, + "loss": 2.638, + "step": 3452 + }, + { + "epoch": 0.15, + "grad_norm": 0.439453125, + "learning_rate": 0.00019748126339706726, + "loss": 2.6559, + "step": 3453 + }, + { + "epoch": 0.15, + "grad_norm": 0.470703125, + "learning_rate": 0.00019747976713864622, + "loss": 2.7979, + "step": 3454 + }, + { + "epoch": 0.15, + "grad_norm": 0.435546875, + "learning_rate": 0.00019747827044160187, + "loss": 2.7378, + "step": 3455 + }, + { + "epoch": 0.15, + "grad_norm": 0.376953125, + "learning_rate": 0.00019747677330594093, + "loss": 2.6009, + "step": 3456 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.0001974752757316701, + "loss": 2.6378, + "step": 3457 + }, + { + "epoch": 0.15, + "grad_norm": 0.4140625, + "learning_rate": 0.0001974737777187962, + "loss": 2.6485, + "step": 3458 + }, + { + "epoch": 0.15, + "grad_norm": 0.435546875, + "learning_rate": 0.0001974722792673259, + "loss": 2.6864, + "step": 3459 + }, + { + "epoch": 0.15, + "grad_norm": 0.39453125, + "learning_rate": 0.00019747078037726597, + "loss": 2.5888, + "step": 3460 + }, + { + "epoch": 0.15, + "grad_norm": 0.412109375, + "learning_rate": 0.00019746928104862312, + "loss": 2.7077, + "step": 3461 + }, + { + "epoch": 0.15, + "grad_norm": 0.4375, + "learning_rate": 0.00019746778128140418, + "loss": 2.6855, + "step": 3462 + }, + { + "epoch": 0.15, + "grad_norm": 0.423828125, + "learning_rate": 0.00019746628107561584, + "loss": 2.6071, + "step": 3463 + }, + { + "epoch": 0.15, + "grad_norm": 0.41015625, + "learning_rate": 0.00019746478043126478, + "loss": 2.7323, + "step": 3464 + }, + { + "epoch": 0.15, + "grad_norm": 0.451171875, + "learning_rate": 0.0001974632793483579, + "loss": 2.6361, + "step": 3465 + }, + { + "epoch": 0.15, + "grad_norm": 0.390625, + "learning_rate": 0.00019746177782690186, + "loss": 2.4885, + "step": 3466 + }, + { + "epoch": 0.15, + "grad_norm": 0.412109375, + "learning_rate": 0.00019746027586690343, + "loss": 2.5939, + "step": 3467 + }, + { + "epoch": 0.15, + "grad_norm": 0.470703125, + "learning_rate": 0.0001974587734683694, + "loss": 2.6917, + "step": 3468 + }, + { + "epoch": 0.15, + "grad_norm": 0.3984375, + "learning_rate": 0.00019745727063130649, + "loss": 2.7885, + "step": 3469 + }, + { + "epoch": 0.15, + "grad_norm": 0.427734375, + "learning_rate": 0.00019745576735572147, + "loss": 2.5905, + "step": 3470 + }, + { + "epoch": 0.15, + "grad_norm": 0.427734375, + "learning_rate": 0.0001974542636416211, + "loss": 2.6782, + "step": 3471 + }, + { + "epoch": 0.15, + "grad_norm": 0.392578125, + "learning_rate": 0.00019745275948901221, + "loss": 2.7244, + "step": 3472 + }, + { + "epoch": 0.15, + "grad_norm": 0.423828125, + "learning_rate": 0.0001974512548979015, + "loss": 2.5647, + "step": 3473 + }, + { + "epoch": 0.15, + "grad_norm": 0.369140625, + "learning_rate": 0.00019744974986829573, + "loss": 2.6866, + "step": 3474 + }, + { + "epoch": 0.15, + "grad_norm": 0.408203125, + "learning_rate": 0.0001974482444002017, + "loss": 2.7063, + "step": 3475 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.00019744673849362622, + "loss": 2.7433, + "step": 3476 + }, + { + "epoch": 0.15, + "grad_norm": 0.3984375, + "learning_rate": 0.000197445232148576, + "loss": 2.5427, + "step": 3477 + }, + { + "epoch": 0.15, + "grad_norm": 0.439453125, + "learning_rate": 0.00019744372536505787, + "loss": 2.7345, + "step": 3478 + }, + { + "epoch": 0.15, + "grad_norm": 0.390625, + "learning_rate": 0.00019744221814307856, + "loss": 2.663, + "step": 3479 + }, + { + "epoch": 0.15, + "grad_norm": 0.466796875, + "learning_rate": 0.0001974407104826449, + "loss": 2.78, + "step": 3480 + }, + { + "epoch": 0.15, + "grad_norm": 0.380859375, + "learning_rate": 0.00019743920238376363, + "loss": 2.5184, + "step": 3481 + }, + { + "epoch": 0.15, + "grad_norm": 0.388671875, + "learning_rate": 0.0001974376938464416, + "loss": 2.6337, + "step": 3482 + }, + { + "epoch": 0.15, + "grad_norm": 0.40234375, + "learning_rate": 0.00019743618487068552, + "loss": 2.5681, + "step": 3483 + }, + { + "epoch": 0.15, + "grad_norm": 0.4296875, + "learning_rate": 0.00019743467545650222, + "loss": 2.8108, + "step": 3484 + }, + { + "epoch": 0.15, + "grad_norm": 0.4375, + "learning_rate": 0.00019743316560389852, + "loss": 2.7335, + "step": 3485 + }, + { + "epoch": 0.15, + "grad_norm": 0.423828125, + "learning_rate": 0.00019743165531288114, + "loss": 2.5861, + "step": 3486 + }, + { + "epoch": 0.15, + "grad_norm": 0.38671875, + "learning_rate": 0.00019743014458345695, + "loss": 2.558, + "step": 3487 + }, + { + "epoch": 0.15, + "grad_norm": 0.39453125, + "learning_rate": 0.0001974286334156327, + "loss": 2.6204, + "step": 3488 + }, + { + "epoch": 0.15, + "grad_norm": 0.421875, + "learning_rate": 0.0001974271218094152, + "loss": 2.6526, + "step": 3489 + }, + { + "epoch": 0.15, + "grad_norm": 0.44921875, + "learning_rate": 0.00019742560976481128, + "loss": 2.6478, + "step": 3490 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.00019742409728182772, + "loss": 2.6916, + "step": 3491 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.0001974225843604713, + "loss": 2.7561, + "step": 3492 + }, + { + "epoch": 0.15, + "grad_norm": 0.419921875, + "learning_rate": 0.00019742107100074888, + "loss": 2.6252, + "step": 3493 + }, + { + "epoch": 0.15, + "grad_norm": 0.427734375, + "learning_rate": 0.00019741955720266723, + "loss": 2.6561, + "step": 3494 + }, + { + "epoch": 0.15, + "grad_norm": 0.369140625, + "learning_rate": 0.00019741804296623315, + "loss": 2.6883, + "step": 3495 + }, + { + "epoch": 0.15, + "grad_norm": 0.439453125, + "learning_rate": 0.00019741652829145353, + "loss": 2.7993, + "step": 3496 + }, + { + "epoch": 0.15, + "grad_norm": 0.41796875, + "learning_rate": 0.00019741501317833512, + "loss": 2.6193, + "step": 3497 + }, + { + "epoch": 0.15, + "grad_norm": 0.40234375, + "learning_rate": 0.00019741349762688476, + "loss": 2.5918, + "step": 3498 + }, + { + "epoch": 0.15, + "grad_norm": 0.435546875, + "learning_rate": 0.00019741198163710923, + "loss": 2.4737, + "step": 3499 + }, + { + "epoch": 0.15, + "grad_norm": 0.443359375, + "learning_rate": 0.0001974104652090154, + "loss": 2.8569, + "step": 3500 + }, + { + "epoch": 0.15, + "grad_norm": 0.404296875, + "learning_rate": 0.00019740894834261008, + "loss": 2.7281, + "step": 3501 + }, + { + "epoch": 0.15, + "grad_norm": 0.3671875, + "learning_rate": 0.00019740743103790006, + "loss": 2.6928, + "step": 3502 + }, + { + "epoch": 0.15, + "grad_norm": 0.41796875, + "learning_rate": 0.00019740591329489224, + "loss": 2.6012, + "step": 3503 + }, + { + "epoch": 0.15, + "grad_norm": 0.3984375, + "learning_rate": 0.0001974043951135934, + "loss": 2.6456, + "step": 3504 + }, + { + "epoch": 0.15, + "grad_norm": 0.4296875, + "learning_rate": 0.00019740287649401038, + "loss": 2.6194, + "step": 3505 + }, + { + "epoch": 0.15, + "grad_norm": 0.474609375, + "learning_rate": 0.00019740135743615, + "loss": 2.6272, + "step": 3506 + }, + { + "epoch": 0.15, + "grad_norm": 0.421875, + "learning_rate": 0.00019739983794001911, + "loss": 2.6109, + "step": 3507 + }, + { + "epoch": 0.15, + "grad_norm": 0.4453125, + "learning_rate": 0.00019739831800562454, + "loss": 2.7009, + "step": 3508 + }, + { + "epoch": 0.15, + "grad_norm": 0.431640625, + "learning_rate": 0.00019739679763297314, + "loss": 2.6653, + "step": 3509 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.00019739527682207175, + "loss": 2.6026, + "step": 3510 + }, + { + "epoch": 0.15, + "grad_norm": 0.390625, + "learning_rate": 0.0001973937555729272, + "loss": 2.4856, + "step": 3511 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.00019739223388554633, + "loss": 2.7086, + "step": 3512 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.00019739071175993602, + "loss": 2.678, + "step": 3513 + }, + { + "epoch": 0.15, + "grad_norm": 0.458984375, + "learning_rate": 0.0001973891891961031, + "loss": 2.6452, + "step": 3514 + }, + { + "epoch": 0.15, + "grad_norm": 0.431640625, + "learning_rate": 0.00019738766619405442, + "loss": 2.6566, + "step": 3515 + }, + { + "epoch": 0.15, + "grad_norm": 0.453125, + "learning_rate": 0.00019738614275379684, + "loss": 2.7403, + "step": 3516 + }, + { + "epoch": 0.15, + "grad_norm": 0.392578125, + "learning_rate": 0.0001973846188753372, + "loss": 2.6784, + "step": 3517 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.00019738309455868237, + "loss": 2.7583, + "step": 3518 + }, + { + "epoch": 0.15, + "grad_norm": 0.42578125, + "learning_rate": 0.00019738156980383918, + "loss": 2.63, + "step": 3519 + }, + { + "epoch": 0.15, + "grad_norm": 0.400390625, + "learning_rate": 0.00019738004461081454, + "loss": 2.7007, + "step": 3520 + }, + { + "epoch": 0.15, + "grad_norm": 0.376953125, + "learning_rate": 0.00019737851897961528, + "loss": 2.6268, + "step": 3521 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.00019737699291024827, + "loss": 2.5142, + "step": 3522 + }, + { + "epoch": 0.15, + "grad_norm": 0.4609375, + "learning_rate": 0.00019737546640272039, + "loss": 2.6962, + "step": 3523 + }, + { + "epoch": 0.15, + "grad_norm": 0.3984375, + "learning_rate": 0.00019737393945703848, + "loss": 2.6527, + "step": 3524 + }, + { + "epoch": 0.15, + "grad_norm": 0.4453125, + "learning_rate": 0.00019737241207320945, + "loss": 2.632, + "step": 3525 + }, + { + "epoch": 0.15, + "grad_norm": 0.400390625, + "learning_rate": 0.00019737088425124012, + "loss": 2.549, + "step": 3526 + }, + { + "epoch": 0.15, + "grad_norm": 0.400390625, + "learning_rate": 0.00019736935599113742, + "loss": 2.5606, + "step": 3527 + }, + { + "epoch": 0.15, + "grad_norm": 0.4296875, + "learning_rate": 0.0001973678272929082, + "loss": 2.6748, + "step": 3528 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.00019736629815655935, + "loss": 2.6546, + "step": 3529 + }, + { + "epoch": 0.15, + "grad_norm": 0.439453125, + "learning_rate": 0.0001973647685820977, + "loss": 2.6157, + "step": 3530 + }, + { + "epoch": 0.15, + "grad_norm": 0.447265625, + "learning_rate": 0.00019736323856953022, + "loss": 2.6936, + "step": 3531 + }, + { + "epoch": 0.15, + "grad_norm": 0.40234375, + "learning_rate": 0.0001973617081188637, + "loss": 2.7087, + "step": 3532 + }, + { + "epoch": 0.15, + "grad_norm": 0.5, + "learning_rate": 0.00019736017723010512, + "loss": 2.8678, + "step": 3533 + }, + { + "epoch": 0.15, + "grad_norm": 0.376953125, + "learning_rate": 0.00019735864590326134, + "loss": 2.5237, + "step": 3534 + }, + { + "epoch": 0.15, + "grad_norm": 0.453125, + "learning_rate": 0.0001973571141383392, + "loss": 2.537, + "step": 3535 + }, + { + "epoch": 0.15, + "grad_norm": 0.494140625, + "learning_rate": 0.0001973555819353456, + "loss": 2.7085, + "step": 3536 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.00019735404929428752, + "loss": 2.7264, + "step": 3537 + }, + { + "epoch": 0.15, + "grad_norm": 0.4375, + "learning_rate": 0.00019735251621517177, + "loss": 2.7138, + "step": 3538 + }, + { + "epoch": 0.15, + "grad_norm": 0.42578125, + "learning_rate": 0.00019735098269800527, + "loss": 2.6414, + "step": 3539 + }, + { + "epoch": 0.15, + "grad_norm": 0.427734375, + "learning_rate": 0.00019734944874279494, + "loss": 2.6284, + "step": 3540 + }, + { + "epoch": 0.15, + "grad_norm": 0.423828125, + "learning_rate": 0.00019734791434954763, + "loss": 2.647, + "step": 3541 + }, + { + "epoch": 0.15, + "grad_norm": 0.392578125, + "learning_rate": 0.00019734637951827032, + "loss": 2.6675, + "step": 3542 + }, + { + "epoch": 0.15, + "grad_norm": 0.419921875, + "learning_rate": 0.00019734484424896987, + "loss": 2.6996, + "step": 3543 + }, + { + "epoch": 0.15, + "grad_norm": 0.4765625, + "learning_rate": 0.00019734330854165322, + "loss": 2.8638, + "step": 3544 + }, + { + "epoch": 0.15, + "grad_norm": 0.44140625, + "learning_rate": 0.0001973417723963272, + "loss": 2.5598, + "step": 3545 + }, + { + "epoch": 0.15, + "grad_norm": 0.419921875, + "learning_rate": 0.0001973402358129988, + "loss": 2.6556, + "step": 3546 + }, + { + "epoch": 0.15, + "grad_norm": 0.373046875, + "learning_rate": 0.00019733869879167494, + "loss": 2.7026, + "step": 3547 + }, + { + "epoch": 0.15, + "grad_norm": 0.41796875, + "learning_rate": 0.00019733716133236252, + "loss": 2.7829, + "step": 3548 + }, + { + "epoch": 0.15, + "grad_norm": 0.369140625, + "learning_rate": 0.0001973356234350684, + "loss": 2.5064, + "step": 3549 + }, + { + "epoch": 0.15, + "grad_norm": 0.439453125, + "learning_rate": 0.00019733408509979957, + "loss": 2.7058, + "step": 3550 + }, + { + "epoch": 0.15, + "grad_norm": 0.41015625, + "learning_rate": 0.00019733254632656291, + "loss": 2.6327, + "step": 3551 + }, + { + "epoch": 0.15, + "grad_norm": 0.412109375, + "learning_rate": 0.0001973310071153654, + "loss": 2.6734, + "step": 3552 + }, + { + "epoch": 0.15, + "grad_norm": 0.38671875, + "learning_rate": 0.00019732946746621387, + "loss": 2.5298, + "step": 3553 + }, + { + "epoch": 0.15, + "grad_norm": 0.423828125, + "learning_rate": 0.00019732792737911536, + "loss": 2.7212, + "step": 3554 + }, + { + "epoch": 0.15, + "grad_norm": 0.466796875, + "learning_rate": 0.00019732638685407671, + "loss": 2.6423, + "step": 3555 + }, + { + "epoch": 0.15, + "grad_norm": 0.408203125, + "learning_rate": 0.0001973248458911049, + "loss": 2.7224, + "step": 3556 + }, + { + "epoch": 0.15, + "grad_norm": 0.408203125, + "learning_rate": 0.00019732330449020687, + "loss": 2.6899, + "step": 3557 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.0001973217626513895, + "loss": 2.7829, + "step": 3558 + }, + { + "epoch": 0.15, + "grad_norm": 0.412109375, + "learning_rate": 0.0001973202203746598, + "loss": 2.7318, + "step": 3559 + }, + { + "epoch": 0.15, + "grad_norm": 0.4609375, + "learning_rate": 0.00019731867766002465, + "loss": 2.8566, + "step": 3560 + }, + { + "epoch": 0.15, + "grad_norm": 0.3671875, + "learning_rate": 0.00019731713450749102, + "loss": 2.4985, + "step": 3561 + }, + { + "epoch": 0.15, + "grad_norm": 0.412109375, + "learning_rate": 0.0001973155909170659, + "loss": 2.6867, + "step": 3562 + }, + { + "epoch": 0.15, + "grad_norm": 0.419921875, + "learning_rate": 0.00019731404688875612, + "loss": 2.6569, + "step": 3563 + }, + { + "epoch": 0.15, + "grad_norm": 0.44140625, + "learning_rate": 0.00019731250242256874, + "loss": 2.6904, + "step": 3564 + }, + { + "epoch": 0.15, + "grad_norm": 0.515625, + "learning_rate": 0.00019731095751851063, + "loss": 2.8864, + "step": 3565 + }, + { + "epoch": 0.15, + "grad_norm": 0.458984375, + "learning_rate": 0.0001973094121765888, + "loss": 2.7371, + "step": 3566 + }, + { + "epoch": 0.15, + "grad_norm": 0.421875, + "learning_rate": 0.00019730786639681017, + "loss": 2.6434, + "step": 3567 + }, + { + "epoch": 0.15, + "grad_norm": 0.390625, + "learning_rate": 0.00019730632017918168, + "loss": 2.6317, + "step": 3568 + }, + { + "epoch": 0.15, + "grad_norm": 0.35546875, + "learning_rate": 0.00019730477352371032, + "loss": 2.528, + "step": 3569 + }, + { + "epoch": 0.15, + "grad_norm": 0.48046875, + "learning_rate": 0.00019730322643040304, + "loss": 2.7437, + "step": 3570 + }, + { + "epoch": 0.15, + "grad_norm": 0.435546875, + "learning_rate": 0.00019730167889926681, + "loss": 2.6319, + "step": 3571 + }, + { + "epoch": 0.15, + "grad_norm": 0.482421875, + "learning_rate": 0.00019730013093030859, + "loss": 2.7002, + "step": 3572 + }, + { + "epoch": 0.15, + "grad_norm": 0.408203125, + "learning_rate": 0.00019729858252353532, + "loss": 2.6924, + "step": 3573 + }, + { + "epoch": 0.15, + "grad_norm": 0.38671875, + "learning_rate": 0.000197297033678954, + "loss": 2.7202, + "step": 3574 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.0001972954843965716, + "loss": 2.6323, + "step": 3575 + }, + { + "epoch": 0.15, + "grad_norm": 0.390625, + "learning_rate": 0.00019729393467639506, + "loss": 2.4636, + "step": 3576 + }, + { + "epoch": 0.15, + "grad_norm": 0.396484375, + "learning_rate": 0.00019729238451843137, + "loss": 2.6793, + "step": 3577 + }, + { + "epoch": 0.15, + "grad_norm": 0.453125, + "learning_rate": 0.0001972908339226875, + "loss": 2.7027, + "step": 3578 + }, + { + "epoch": 0.15, + "grad_norm": 0.412109375, + "learning_rate": 0.00019728928288917046, + "loss": 2.6465, + "step": 3579 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.00019728773141788718, + "loss": 2.6271, + "step": 3580 + }, + { + "epoch": 0.15, + "grad_norm": 0.376953125, + "learning_rate": 0.00019728617950884466, + "loss": 2.636, + "step": 3581 + }, + { + "epoch": 0.15, + "grad_norm": 0.435546875, + "learning_rate": 0.00019728462716204993, + "loss": 2.6896, + "step": 3582 + }, + { + "epoch": 0.15, + "grad_norm": 0.3828125, + "learning_rate": 0.00019728307437750985, + "loss": 2.6807, + "step": 3583 + }, + { + "epoch": 0.15, + "grad_norm": 0.4765625, + "learning_rate": 0.00019728152115523154, + "loss": 2.7786, + "step": 3584 + }, + { + "epoch": 0.15, + "grad_norm": 0.42578125, + "learning_rate": 0.00019727996749522195, + "loss": 2.6135, + "step": 3585 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.00019727841339748804, + "loss": 2.703, + "step": 3586 + }, + { + "epoch": 0.15, + "grad_norm": 0.48828125, + "learning_rate": 0.00019727685886203683, + "loss": 2.6263, + "step": 3587 + }, + { + "epoch": 0.15, + "grad_norm": 0.404296875, + "learning_rate": 0.00019727530388887528, + "loss": 2.8114, + "step": 3588 + }, + { + "epoch": 0.15, + "grad_norm": 0.396484375, + "learning_rate": 0.00019727374847801043, + "loss": 2.6719, + "step": 3589 + }, + { + "epoch": 0.15, + "grad_norm": 0.451171875, + "learning_rate": 0.00019727219262944927, + "loss": 2.588, + "step": 3590 + }, + { + "epoch": 0.15, + "grad_norm": 0.3984375, + "learning_rate": 0.00019727063634319877, + "loss": 2.613, + "step": 3591 + }, + { + "epoch": 0.15, + "grad_norm": 0.4375, + "learning_rate": 0.00019726907961926599, + "loss": 2.542, + "step": 3592 + }, + { + "epoch": 0.15, + "grad_norm": 0.412109375, + "learning_rate": 0.00019726752245765786, + "loss": 2.6352, + "step": 3593 + }, + { + "epoch": 0.15, + "grad_norm": 0.4140625, + "learning_rate": 0.00019726596485838144, + "loss": 2.6409, + "step": 3594 + }, + { + "epoch": 0.15, + "grad_norm": 0.41796875, + "learning_rate": 0.00019726440682144375, + "loss": 2.7404, + "step": 3595 + }, + { + "epoch": 0.15, + "grad_norm": 0.4140625, + "learning_rate": 0.00019726284834685175, + "loss": 2.658, + "step": 3596 + }, + { + "epoch": 0.15, + "grad_norm": 0.4140625, + "learning_rate": 0.00019726128943461246, + "loss": 2.6893, + "step": 3597 + }, + { + "epoch": 0.15, + "grad_norm": 0.392578125, + "learning_rate": 0.00019725973008473293, + "loss": 2.6221, + "step": 3598 + }, + { + "epoch": 0.15, + "grad_norm": 0.384765625, + "learning_rate": 0.0001972581702972202, + "loss": 2.5286, + "step": 3599 + }, + { + "epoch": 0.15, + "grad_norm": 0.455078125, + "learning_rate": 0.0001972566100720812, + "loss": 2.6813, + "step": 3600 + }, + { + "epoch": 0.15, + "grad_norm": 0.447265625, + "learning_rate": 0.000197255049409323, + "loss": 2.7513, + "step": 3601 + }, + { + "epoch": 0.15, + "grad_norm": 0.421875, + "learning_rate": 0.00019725348830895262, + "loss": 2.7003, + "step": 3602 + }, + { + "epoch": 0.15, + "grad_norm": 0.419921875, + "learning_rate": 0.0001972519267709771, + "loss": 2.7387, + "step": 3603 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.0001972503647954034, + "loss": 2.6564, + "step": 3604 + }, + { + "epoch": 0.15, + "grad_norm": 0.400390625, + "learning_rate": 0.0001972488023822387, + "loss": 2.6638, + "step": 3605 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.00019724723953148983, + "loss": 2.5394, + "step": 3606 + }, + { + "epoch": 0.15, + "grad_norm": 0.38671875, + "learning_rate": 0.00019724567624316398, + "loss": 2.7939, + "step": 3607 + }, + { + "epoch": 0.15, + "grad_norm": 0.45703125, + "learning_rate": 0.0001972441125172681, + "loss": 2.6437, + "step": 3608 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.00019724254835380928, + "loss": 2.5873, + "step": 3609 + }, + { + "epoch": 0.15, + "grad_norm": 0.384765625, + "learning_rate": 0.0001972409837527945, + "loss": 2.6789, + "step": 3610 + }, + { + "epoch": 0.15, + "grad_norm": 0.396484375, + "learning_rate": 0.00019723941871423086, + "loss": 2.6692, + "step": 3611 + }, + { + "epoch": 0.15, + "grad_norm": 0.427734375, + "learning_rate": 0.00019723785323812537, + "loss": 2.8213, + "step": 3612 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.00019723628732448505, + "loss": 2.6955, + "step": 3613 + }, + { + "epoch": 0.15, + "grad_norm": 0.41015625, + "learning_rate": 0.00019723472097331698, + "loss": 2.5948, + "step": 3614 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.00019723315418462823, + "loss": 2.7064, + "step": 3615 + }, + { + "epoch": 0.15, + "grad_norm": 0.43359375, + "learning_rate": 0.0001972315869584258, + "loss": 2.7493, + "step": 3616 + }, + { + "epoch": 0.15, + "grad_norm": 0.451171875, + "learning_rate": 0.00019723001929471674, + "loss": 2.7979, + "step": 3617 + }, + { + "epoch": 0.15, + "grad_norm": 0.37109375, + "learning_rate": 0.00019722845119350818, + "loss": 2.5747, + "step": 3618 + }, + { + "epoch": 0.15, + "grad_norm": 0.376953125, + "learning_rate": 0.00019722688265480707, + "loss": 2.6218, + "step": 3619 + }, + { + "epoch": 0.15, + "grad_norm": 0.37890625, + "learning_rate": 0.00019722531367862055, + "loss": 2.5584, + "step": 3620 + }, + { + "epoch": 0.15, + "grad_norm": 0.42578125, + "learning_rate": 0.00019722374426495563, + "loss": 2.6513, + "step": 3621 + }, + { + "epoch": 0.15, + "grad_norm": 0.40625, + "learning_rate": 0.0001972221744138194, + "loss": 2.7153, + "step": 3622 + }, + { + "epoch": 0.15, + "grad_norm": 0.431640625, + "learning_rate": 0.0001972206041252189, + "loss": 2.7706, + "step": 3623 + }, + { + "epoch": 0.15, + "grad_norm": 0.453125, + "learning_rate": 0.00019721903339916125, + "loss": 2.5704, + "step": 3624 + }, + { + "epoch": 0.15, + "grad_norm": 0.416015625, + "learning_rate": 0.00019721746223565347, + "loss": 2.6786, + "step": 3625 + }, + { + "epoch": 0.15, + "grad_norm": 0.4453125, + "learning_rate": 0.00019721589063470261, + "loss": 2.6619, + "step": 3626 + }, + { + "epoch": 0.15, + "grad_norm": 0.4296875, + "learning_rate": 0.00019721431859631578, + "loss": 2.6548, + "step": 3627 + }, + { + "epoch": 0.15, + "grad_norm": 0.408203125, + "learning_rate": 0.00019721274612050007, + "loss": 2.7756, + "step": 3628 + }, + { + "epoch": 0.15, + "grad_norm": 0.435546875, + "learning_rate": 0.00019721117320726253, + "loss": 2.7275, + "step": 3629 + }, + { + "epoch": 0.15, + "grad_norm": 0.38671875, + "learning_rate": 0.0001972095998566102, + "loss": 2.6052, + "step": 3630 + }, + { + "epoch": 0.15, + "grad_norm": 0.462890625, + "learning_rate": 0.00019720802606855022, + "loss": 2.6394, + "step": 3631 + }, + { + "epoch": 0.15, + "grad_norm": 0.421875, + "learning_rate": 0.00019720645184308965, + "loss": 2.6661, + "step": 3632 + }, + { + "epoch": 0.15, + "grad_norm": 0.4140625, + "learning_rate": 0.00019720487718023554, + "loss": 2.7041, + "step": 3633 + }, + { + "epoch": 0.15, + "grad_norm": 0.37890625, + "learning_rate": 0.00019720330207999504, + "loss": 2.7362, + "step": 3634 + }, + { + "epoch": 0.15, + "grad_norm": 0.4296875, + "learning_rate": 0.0001972017265423752, + "loss": 2.6506, + "step": 3635 + }, + { + "epoch": 0.15, + "grad_norm": 0.412109375, + "learning_rate": 0.00019720015056738313, + "loss": 2.7262, + "step": 3636 + }, + { + "epoch": 0.15, + "grad_norm": 0.37890625, + "learning_rate": 0.00019719857415502587, + "loss": 2.5957, + "step": 3637 + }, + { + "epoch": 0.16, + "grad_norm": 0.427734375, + "learning_rate": 0.0001971969973053106, + "loss": 2.8088, + "step": 3638 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.0001971954200182443, + "loss": 2.6816, + "step": 3639 + }, + { + "epoch": 0.16, + "grad_norm": 0.390625, + "learning_rate": 0.00019719384229383415, + "loss": 2.6294, + "step": 3640 + }, + { + "epoch": 0.16, + "grad_norm": 0.43359375, + "learning_rate": 0.00019719226413208725, + "loss": 2.5725, + "step": 3641 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.00019719068553301065, + "loss": 2.7184, + "step": 3642 + }, + { + "epoch": 0.16, + "grad_norm": 0.4765625, + "learning_rate": 0.0001971891064966115, + "loss": 2.654, + "step": 3643 + }, + { + "epoch": 0.16, + "grad_norm": 0.466796875, + "learning_rate": 0.0001971875270228969, + "loss": 2.7461, + "step": 3644 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.00019718594711187396, + "loss": 2.6381, + "step": 3645 + }, + { + "epoch": 0.16, + "grad_norm": 0.4609375, + "learning_rate": 0.00019718436676354975, + "loss": 2.7949, + "step": 3646 + }, + { + "epoch": 0.16, + "grad_norm": 0.43359375, + "learning_rate": 0.00019718278597793143, + "loss": 2.7229, + "step": 3647 + }, + { + "epoch": 0.16, + "grad_norm": 0.443359375, + "learning_rate": 0.00019718120475502605, + "loss": 2.5654, + "step": 3648 + }, + { + "epoch": 0.16, + "grad_norm": 0.43359375, + "learning_rate": 0.0001971796230948408, + "loss": 2.7446, + "step": 3649 + }, + { + "epoch": 0.16, + "grad_norm": 0.4375, + "learning_rate": 0.00019717804099738274, + "loss": 2.6591, + "step": 3650 + }, + { + "epoch": 0.16, + "grad_norm": 0.458984375, + "learning_rate": 0.00019717645846265903, + "loss": 2.7591, + "step": 3651 + }, + { + "epoch": 0.16, + "grad_norm": 0.45703125, + "learning_rate": 0.00019717487549067675, + "loss": 2.6599, + "step": 3652 + }, + { + "epoch": 0.16, + "grad_norm": 0.3984375, + "learning_rate": 0.00019717329208144305, + "loss": 2.594, + "step": 3653 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.00019717170823496506, + "loss": 2.8944, + "step": 3654 + }, + { + "epoch": 0.16, + "grad_norm": 0.38671875, + "learning_rate": 0.00019717012395124988, + "loss": 2.6411, + "step": 3655 + }, + { + "epoch": 0.16, + "grad_norm": 0.466796875, + "learning_rate": 0.00019716853923030465, + "loss": 2.6812, + "step": 3656 + }, + { + "epoch": 0.16, + "grad_norm": 0.419921875, + "learning_rate": 0.0001971669540721365, + "loss": 2.5674, + "step": 3657 + }, + { + "epoch": 0.16, + "grad_norm": 0.439453125, + "learning_rate": 0.00019716536847675257, + "loss": 2.7164, + "step": 3658 + }, + { + "epoch": 0.16, + "grad_norm": 0.4140625, + "learning_rate": 0.00019716378244416004, + "loss": 2.6326, + "step": 3659 + }, + { + "epoch": 0.16, + "grad_norm": 0.43359375, + "learning_rate": 0.00019716219597436593, + "loss": 2.7636, + "step": 3660 + }, + { + "epoch": 0.16, + "grad_norm": 0.36328125, + "learning_rate": 0.00019716060906737747, + "loss": 2.4836, + "step": 3661 + }, + { + "epoch": 0.16, + "grad_norm": 0.392578125, + "learning_rate": 0.00019715902172320182, + "loss": 2.6544, + "step": 3662 + }, + { + "epoch": 0.16, + "grad_norm": 0.388671875, + "learning_rate": 0.00019715743394184602, + "loss": 2.66, + "step": 3663 + }, + { + "epoch": 0.16, + "grad_norm": 0.435546875, + "learning_rate": 0.00019715584572331732, + "loss": 2.6386, + "step": 3664 + }, + { + "epoch": 0.16, + "grad_norm": 0.400390625, + "learning_rate": 0.00019715425706762278, + "loss": 2.8161, + "step": 3665 + }, + { + "epoch": 0.16, + "grad_norm": 0.435546875, + "learning_rate": 0.0001971526679747696, + "loss": 2.6385, + "step": 3666 + }, + { + "epoch": 0.16, + "grad_norm": 0.439453125, + "learning_rate": 0.00019715107844476494, + "loss": 2.648, + "step": 3667 + }, + { + "epoch": 0.16, + "grad_norm": 0.3984375, + "learning_rate": 0.00019714948847761592, + "loss": 2.7627, + "step": 3668 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.0001971478980733297, + "loss": 2.7818, + "step": 3669 + }, + { + "epoch": 0.16, + "grad_norm": 0.4140625, + "learning_rate": 0.00019714630723191345, + "loss": 2.5952, + "step": 3670 + }, + { + "epoch": 0.16, + "grad_norm": 0.400390625, + "learning_rate": 0.0001971447159533743, + "loss": 2.6597, + "step": 3671 + }, + { + "epoch": 0.16, + "grad_norm": 0.396484375, + "learning_rate": 0.00019714312423771947, + "loss": 2.7325, + "step": 3672 + }, + { + "epoch": 0.16, + "grad_norm": 0.4765625, + "learning_rate": 0.00019714153208495608, + "loss": 2.7072, + "step": 3673 + }, + { + "epoch": 0.16, + "grad_norm": 0.357421875, + "learning_rate": 0.00019713993949509125, + "loss": 2.5371, + "step": 3674 + }, + { + "epoch": 0.16, + "grad_norm": 0.4375, + "learning_rate": 0.00019713834646813223, + "loss": 2.7548, + "step": 3675 + }, + { + "epoch": 0.16, + "grad_norm": 0.39453125, + "learning_rate": 0.00019713675300408613, + "loss": 2.6169, + "step": 3676 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.00019713515910296017, + "loss": 2.6244, + "step": 3677 + }, + { + "epoch": 0.16, + "grad_norm": 0.44140625, + "learning_rate": 0.00019713356476476146, + "loss": 2.8305, + "step": 3678 + }, + { + "epoch": 0.16, + "grad_norm": 0.37890625, + "learning_rate": 0.00019713196998949725, + "loss": 2.5484, + "step": 3679 + }, + { + "epoch": 0.16, + "grad_norm": 0.474609375, + "learning_rate": 0.00019713037477717463, + "loss": 2.7391, + "step": 3680 + }, + { + "epoch": 0.16, + "grad_norm": 0.3984375, + "learning_rate": 0.00019712877912780082, + "loss": 2.6984, + "step": 3681 + }, + { + "epoch": 0.16, + "grad_norm": 0.419921875, + "learning_rate": 0.00019712718304138303, + "loss": 2.5839, + "step": 3682 + }, + { + "epoch": 0.16, + "grad_norm": 0.373046875, + "learning_rate": 0.00019712558651792842, + "loss": 2.7204, + "step": 3683 + }, + { + "epoch": 0.16, + "grad_norm": 0.412109375, + "learning_rate": 0.0001971239895574441, + "loss": 2.5969, + "step": 3684 + }, + { + "epoch": 0.16, + "grad_norm": 0.380859375, + "learning_rate": 0.0001971223921599374, + "loss": 2.5788, + "step": 3685 + }, + { + "epoch": 0.16, + "grad_norm": 0.416015625, + "learning_rate": 0.00019712079432541538, + "loss": 2.7194, + "step": 3686 + }, + { + "epoch": 0.16, + "grad_norm": 0.3984375, + "learning_rate": 0.00019711919605388528, + "loss": 2.6741, + "step": 3687 + }, + { + "epoch": 0.16, + "grad_norm": 0.40234375, + "learning_rate": 0.00019711759734535432, + "loss": 2.6382, + "step": 3688 + }, + { + "epoch": 0.16, + "grad_norm": 0.404296875, + "learning_rate": 0.00019711599819982964, + "loss": 2.6387, + "step": 3689 + }, + { + "epoch": 0.16, + "grad_norm": 0.4296875, + "learning_rate": 0.00019711439861731846, + "loss": 2.6385, + "step": 3690 + }, + { + "epoch": 0.16, + "grad_norm": 0.384765625, + "learning_rate": 0.000197112798597828, + "loss": 2.7, + "step": 3691 + }, + { + "epoch": 0.16, + "grad_norm": 0.40234375, + "learning_rate": 0.00019711119814136543, + "loss": 2.785, + "step": 3692 + }, + { + "epoch": 0.16, + "grad_norm": 0.44921875, + "learning_rate": 0.00019710959724793793, + "loss": 2.6541, + "step": 3693 + }, + { + "epoch": 0.16, + "grad_norm": 0.37890625, + "learning_rate": 0.00019710799591755277, + "loss": 2.6305, + "step": 3694 + }, + { + "epoch": 0.16, + "grad_norm": 0.380859375, + "learning_rate": 0.0001971063941502171, + "loss": 2.5224, + "step": 3695 + }, + { + "epoch": 0.16, + "grad_norm": 0.443359375, + "learning_rate": 0.00019710479194593816, + "loss": 2.6309, + "step": 3696 + }, + { + "epoch": 0.16, + "grad_norm": 0.373046875, + "learning_rate": 0.00019710318930472314, + "loss": 2.6278, + "step": 3697 + }, + { + "epoch": 0.16, + "grad_norm": 0.408203125, + "learning_rate": 0.00019710158622657922, + "loss": 2.652, + "step": 3698 + }, + { + "epoch": 0.16, + "grad_norm": 0.474609375, + "learning_rate": 0.00019709998271151368, + "loss": 2.7636, + "step": 3699 + }, + { + "epoch": 0.16, + "grad_norm": 0.4140625, + "learning_rate": 0.0001970983787595337, + "loss": 2.6088, + "step": 3700 + }, + { + "epoch": 0.16, + "grad_norm": 0.4609375, + "learning_rate": 0.00019709677437064647, + "loss": 2.7, + "step": 3701 + }, + { + "epoch": 0.16, + "grad_norm": 0.4140625, + "learning_rate": 0.0001970951695448593, + "loss": 2.5974, + "step": 3702 + }, + { + "epoch": 0.16, + "grad_norm": 0.404296875, + "learning_rate": 0.00019709356428217932, + "loss": 2.5022, + "step": 3703 + }, + { + "epoch": 0.16, + "grad_norm": 0.451171875, + "learning_rate": 0.00019709195858261377, + "loss": 2.657, + "step": 3704 + }, + { + "epoch": 0.16, + "grad_norm": 0.40234375, + "learning_rate": 0.0001970903524461699, + "loss": 2.6045, + "step": 3705 + }, + { + "epoch": 0.16, + "grad_norm": 0.453125, + "learning_rate": 0.00019708874587285493, + "loss": 2.6493, + "step": 3706 + }, + { + "epoch": 0.16, + "grad_norm": 0.419921875, + "learning_rate": 0.00019708713886267608, + "loss": 2.7714, + "step": 3707 + }, + { + "epoch": 0.16, + "grad_norm": 0.4375, + "learning_rate": 0.0001970855314156406, + "loss": 2.6817, + "step": 3708 + }, + { + "epoch": 0.16, + "grad_norm": 0.53125, + "learning_rate": 0.0001970839235317557, + "loss": 2.733, + "step": 3709 + }, + { + "epoch": 0.16, + "grad_norm": 0.466796875, + "learning_rate": 0.00019708231521102865, + "loss": 2.729, + "step": 3710 + }, + { + "epoch": 0.16, + "grad_norm": 0.4140625, + "learning_rate": 0.00019708070645346664, + "loss": 2.7118, + "step": 3711 + }, + { + "epoch": 0.16, + "grad_norm": 0.5078125, + "learning_rate": 0.00019707909725907692, + "loss": 2.76, + "step": 3712 + }, + { + "epoch": 0.16, + "grad_norm": 0.4296875, + "learning_rate": 0.00019707748762786675, + "loss": 2.75, + "step": 3713 + }, + { + "epoch": 0.16, + "grad_norm": 0.44140625, + "learning_rate": 0.00019707587755984335, + "loss": 2.7053, + "step": 3714 + }, + { + "epoch": 0.16, + "grad_norm": 0.494140625, + "learning_rate": 0.000197074267055014, + "loss": 2.7232, + "step": 3715 + }, + { + "epoch": 0.16, + "grad_norm": 0.44140625, + "learning_rate": 0.0001970726561133859, + "loss": 2.6162, + "step": 3716 + }, + { + "epoch": 0.16, + "grad_norm": 0.427734375, + "learning_rate": 0.00019707104473496635, + "loss": 2.6666, + "step": 3717 + }, + { + "epoch": 0.16, + "grad_norm": 0.46484375, + "learning_rate": 0.00019706943291976256, + "loss": 2.6532, + "step": 3718 + }, + { + "epoch": 0.16, + "grad_norm": 0.447265625, + "learning_rate": 0.0001970678206677818, + "loss": 2.712, + "step": 3719 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.0001970662079790313, + "loss": 2.6163, + "step": 3720 + }, + { + "epoch": 0.16, + "grad_norm": 0.45703125, + "learning_rate": 0.00019706459485351835, + "loss": 2.5418, + "step": 3721 + }, + { + "epoch": 0.16, + "grad_norm": 0.447265625, + "learning_rate": 0.00019706298129125023, + "loss": 2.6176, + "step": 3722 + }, + { + "epoch": 0.16, + "grad_norm": 0.47265625, + "learning_rate": 0.00019706136729223412, + "loss": 2.6497, + "step": 3723 + }, + { + "epoch": 0.16, + "grad_norm": 0.486328125, + "learning_rate": 0.00019705975285647735, + "loss": 2.5942, + "step": 3724 + }, + { + "epoch": 0.16, + "grad_norm": 0.462890625, + "learning_rate": 0.00019705813798398717, + "loss": 2.7656, + "step": 3725 + }, + { + "epoch": 0.16, + "grad_norm": 0.5, + "learning_rate": 0.00019705652267477085, + "loss": 2.6172, + "step": 3726 + }, + { + "epoch": 0.16, + "grad_norm": 0.43359375, + "learning_rate": 0.0001970549069288356, + "loss": 2.6835, + "step": 3727 + }, + { + "epoch": 0.16, + "grad_norm": 0.435546875, + "learning_rate": 0.0001970532907461888, + "loss": 2.7238, + "step": 3728 + }, + { + "epoch": 0.16, + "grad_norm": 0.515625, + "learning_rate": 0.0001970516741268376, + "loss": 2.5696, + "step": 3729 + }, + { + "epoch": 0.16, + "grad_norm": 0.361328125, + "learning_rate": 0.00019705005707078938, + "loss": 2.5044, + "step": 3730 + }, + { + "epoch": 0.16, + "grad_norm": 0.412109375, + "learning_rate": 0.00019704843957805138, + "loss": 2.714, + "step": 3731 + }, + { + "epoch": 0.16, + "grad_norm": 0.484375, + "learning_rate": 0.00019704682164863082, + "loss": 2.8089, + "step": 3732 + }, + { + "epoch": 0.16, + "grad_norm": 0.462890625, + "learning_rate": 0.00019704520328253503, + "loss": 2.7823, + "step": 3733 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.00019704358447977135, + "loss": 2.6358, + "step": 3734 + }, + { + "epoch": 0.16, + "grad_norm": 0.46875, + "learning_rate": 0.00019704196524034695, + "loss": 2.6801, + "step": 3735 + }, + { + "epoch": 0.16, + "grad_norm": 0.38671875, + "learning_rate": 0.00019704034556426917, + "loss": 2.6941, + "step": 3736 + }, + { + "epoch": 0.16, + "grad_norm": 0.412109375, + "learning_rate": 0.00019703872545154535, + "loss": 2.7212, + "step": 3737 + }, + { + "epoch": 0.16, + "grad_norm": 0.44140625, + "learning_rate": 0.00019703710490218267, + "loss": 2.69, + "step": 3738 + }, + { + "epoch": 0.16, + "grad_norm": 0.40234375, + "learning_rate": 0.00019703548391618853, + "loss": 2.6458, + "step": 3739 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.00019703386249357013, + "loss": 2.657, + "step": 3740 + }, + { + "epoch": 0.16, + "grad_norm": 0.4609375, + "learning_rate": 0.00019703224063433487, + "loss": 2.7706, + "step": 3741 + }, + { + "epoch": 0.16, + "grad_norm": 0.453125, + "learning_rate": 0.00019703061833848995, + "loss": 2.6824, + "step": 3742 + }, + { + "epoch": 0.16, + "grad_norm": 0.4453125, + "learning_rate": 0.0001970289956060427, + "loss": 2.6873, + "step": 3743 + }, + { + "epoch": 0.16, + "grad_norm": 0.40625, + "learning_rate": 0.00019702737243700044, + "loss": 2.6364, + "step": 3744 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.00019702574883137048, + "loss": 2.7821, + "step": 3745 + }, + { + "epoch": 0.16, + "grad_norm": 0.416015625, + "learning_rate": 0.0001970241247891601, + "loss": 2.7749, + "step": 3746 + }, + { + "epoch": 0.16, + "grad_norm": 0.447265625, + "learning_rate": 0.00019702250031037658, + "loss": 2.7009, + "step": 3747 + }, + { + "epoch": 0.16, + "grad_norm": 0.451171875, + "learning_rate": 0.0001970208753950273, + "loss": 2.7626, + "step": 3748 + }, + { + "epoch": 0.16, + "grad_norm": 0.4140625, + "learning_rate": 0.00019701925004311956, + "loss": 2.7514, + "step": 3749 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.00019701762425466062, + "loss": 2.6108, + "step": 3750 + }, + { + "epoch": 0.16, + "grad_norm": 0.431640625, + "learning_rate": 0.00019701599802965784, + "loss": 2.6409, + "step": 3751 + }, + { + "epoch": 0.16, + "grad_norm": 0.45703125, + "learning_rate": 0.00019701437136811852, + "loss": 2.7932, + "step": 3752 + }, + { + "epoch": 0.16, + "grad_norm": 0.435546875, + "learning_rate": 0.00019701274427004996, + "loss": 2.6124, + "step": 3753 + }, + { + "epoch": 0.16, + "grad_norm": 0.390625, + "learning_rate": 0.00019701111673545954, + "loss": 2.6947, + "step": 3754 + }, + { + "epoch": 0.16, + "grad_norm": 0.388671875, + "learning_rate": 0.00019700948876435453, + "loss": 2.8343, + "step": 3755 + }, + { + "epoch": 0.16, + "grad_norm": 0.435546875, + "learning_rate": 0.00019700786035674228, + "loss": 2.7132, + "step": 3756 + }, + { + "epoch": 0.16, + "grad_norm": 0.435546875, + "learning_rate": 0.00019700623151263012, + "loss": 2.6061, + "step": 3757 + }, + { + "epoch": 0.16, + "grad_norm": 0.404296875, + "learning_rate": 0.00019700460223202532, + "loss": 2.6995, + "step": 3758 + }, + { + "epoch": 0.16, + "grad_norm": 0.455078125, + "learning_rate": 0.0001970029725149353, + "loss": 2.7194, + "step": 3759 + }, + { + "epoch": 0.16, + "grad_norm": 0.451171875, + "learning_rate": 0.00019700134236136737, + "loss": 2.6448, + "step": 3760 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.00019699971177132884, + "loss": 2.7288, + "step": 3761 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.00019699808074482704, + "loss": 2.6911, + "step": 3762 + }, + { + "epoch": 0.16, + "grad_norm": 0.439453125, + "learning_rate": 0.00019699644928186934, + "loss": 2.7001, + "step": 3763 + }, + { + "epoch": 0.16, + "grad_norm": 0.408203125, + "learning_rate": 0.00019699481738246307, + "loss": 2.7973, + "step": 3764 + }, + { + "epoch": 0.16, + "grad_norm": 0.458984375, + "learning_rate": 0.00019699318504661555, + "loss": 2.554, + "step": 3765 + }, + { + "epoch": 0.16, + "grad_norm": 0.38671875, + "learning_rate": 0.00019699155227433413, + "loss": 2.6321, + "step": 3766 + }, + { + "epoch": 0.16, + "grad_norm": 0.390625, + "learning_rate": 0.00019698991906562624, + "loss": 2.7279, + "step": 3767 + }, + { + "epoch": 0.16, + "grad_norm": 0.470703125, + "learning_rate": 0.0001969882854204991, + "loss": 2.6674, + "step": 3768 + }, + { + "epoch": 0.16, + "grad_norm": 0.455078125, + "learning_rate": 0.00019698665133896013, + "loss": 2.7195, + "step": 3769 + }, + { + "epoch": 0.16, + "grad_norm": 0.384765625, + "learning_rate": 0.0001969850168210167, + "loss": 2.5934, + "step": 3770 + }, + { + "epoch": 0.16, + "grad_norm": 0.392578125, + "learning_rate": 0.0001969833818666761, + "loss": 2.5953, + "step": 3771 + }, + { + "epoch": 0.16, + "grad_norm": 0.439453125, + "learning_rate": 0.00019698174647594574, + "loss": 2.5998, + "step": 3772 + }, + { + "epoch": 0.16, + "grad_norm": 0.455078125, + "learning_rate": 0.00019698011064883298, + "loss": 2.6127, + "step": 3773 + }, + { + "epoch": 0.16, + "grad_norm": 0.419921875, + "learning_rate": 0.00019697847438534513, + "loss": 2.7556, + "step": 3774 + }, + { + "epoch": 0.16, + "grad_norm": 0.416015625, + "learning_rate": 0.0001969768376854896, + "loss": 2.7135, + "step": 3775 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.00019697520054927375, + "loss": 2.7319, + "step": 3776 + }, + { + "epoch": 0.16, + "grad_norm": 0.46484375, + "learning_rate": 0.00019697356297670494, + "loss": 2.8983, + "step": 3777 + }, + { + "epoch": 0.16, + "grad_norm": 0.431640625, + "learning_rate": 0.00019697192496779056, + "loss": 2.7501, + "step": 3778 + }, + { + "epoch": 0.16, + "grad_norm": 0.53125, + "learning_rate": 0.00019697028652253793, + "loss": 2.5292, + "step": 3779 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.00019696864764095443, + "loss": 2.6512, + "step": 3780 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.00019696700832304748, + "loss": 2.778, + "step": 3781 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.00019696536856882442, + "loss": 2.5085, + "step": 3782 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.00019696372837829264, + "loss": 2.5747, + "step": 3783 + }, + { + "epoch": 0.16, + "grad_norm": 0.462890625, + "learning_rate": 0.00019696208775145952, + "loss": 2.6946, + "step": 3784 + }, + { + "epoch": 0.16, + "grad_norm": 0.46484375, + "learning_rate": 0.00019696044668833244, + "loss": 2.6669, + "step": 3785 + }, + { + "epoch": 0.16, + "grad_norm": 0.46875, + "learning_rate": 0.00019695880518891878, + "loss": 2.8692, + "step": 3786 + }, + { + "epoch": 0.16, + "grad_norm": 0.4453125, + "learning_rate": 0.00019695716325322592, + "loss": 2.822, + "step": 3787 + }, + { + "epoch": 0.16, + "grad_norm": 0.4375, + "learning_rate": 0.0001969555208812613, + "loss": 2.6854, + "step": 3788 + }, + { + "epoch": 0.16, + "grad_norm": 0.4296875, + "learning_rate": 0.00019695387807303224, + "loss": 2.6847, + "step": 3789 + }, + { + "epoch": 0.16, + "grad_norm": 0.427734375, + "learning_rate": 0.00019695223482854617, + "loss": 2.6998, + "step": 3790 + }, + { + "epoch": 0.16, + "grad_norm": 0.44921875, + "learning_rate": 0.00019695059114781046, + "loss": 2.6497, + "step": 3791 + }, + { + "epoch": 0.16, + "grad_norm": 0.384765625, + "learning_rate": 0.00019694894703083256, + "loss": 2.6928, + "step": 3792 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.0001969473024776198, + "loss": 2.5817, + "step": 3793 + }, + { + "epoch": 0.16, + "grad_norm": 0.4296875, + "learning_rate": 0.0001969456574881796, + "loss": 2.5379, + "step": 3794 + }, + { + "epoch": 0.16, + "grad_norm": 0.388671875, + "learning_rate": 0.00019694401206251936, + "loss": 2.6666, + "step": 3795 + }, + { + "epoch": 0.16, + "grad_norm": 0.427734375, + "learning_rate": 0.00019694236620064652, + "loss": 2.806, + "step": 3796 + }, + { + "epoch": 0.16, + "grad_norm": 0.375, + "learning_rate": 0.00019694071990256846, + "loss": 2.4978, + "step": 3797 + }, + { + "epoch": 0.16, + "grad_norm": 0.39453125, + "learning_rate": 0.00019693907316829258, + "loss": 2.4374, + "step": 3798 + }, + { + "epoch": 0.16, + "grad_norm": 0.447265625, + "learning_rate": 0.00019693742599782632, + "loss": 2.8227, + "step": 3799 + }, + { + "epoch": 0.16, + "grad_norm": 0.4296875, + "learning_rate": 0.00019693577839117703, + "loss": 2.6673, + "step": 3800 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.00019693413034835218, + "loss": 2.7347, + "step": 3801 + }, + { + "epoch": 0.16, + "grad_norm": 0.451171875, + "learning_rate": 0.0001969324818693592, + "loss": 2.5898, + "step": 3802 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.00019693083295420545, + "loss": 2.609, + "step": 3803 + }, + { + "epoch": 0.16, + "grad_norm": 0.412109375, + "learning_rate": 0.00019692918360289836, + "loss": 2.7148, + "step": 3804 + }, + { + "epoch": 0.16, + "grad_norm": 0.458984375, + "learning_rate": 0.00019692753381544539, + "loss": 2.6272, + "step": 3805 + }, + { + "epoch": 0.16, + "grad_norm": 0.412109375, + "learning_rate": 0.00019692588359185393, + "loss": 2.6704, + "step": 3806 + }, + { + "epoch": 0.16, + "grad_norm": 0.3828125, + "learning_rate": 0.00019692423293213142, + "loss": 2.6471, + "step": 3807 + }, + { + "epoch": 0.16, + "grad_norm": 0.404296875, + "learning_rate": 0.0001969225818362853, + "loss": 2.6712, + "step": 3808 + }, + { + "epoch": 0.16, + "grad_norm": 0.396484375, + "learning_rate": 0.00019692093030432295, + "loss": 2.7191, + "step": 3809 + }, + { + "epoch": 0.16, + "grad_norm": 0.35546875, + "learning_rate": 0.00019691927833625185, + "loss": 2.6377, + "step": 3810 + }, + { + "epoch": 0.16, + "grad_norm": 0.400390625, + "learning_rate": 0.00019691762593207943, + "loss": 2.6696, + "step": 3811 + }, + { + "epoch": 0.16, + "grad_norm": 0.408203125, + "learning_rate": 0.00019691597309181308, + "loss": 2.5119, + "step": 3812 + }, + { + "epoch": 0.16, + "grad_norm": 0.44140625, + "learning_rate": 0.00019691431981546032, + "loss": 2.8102, + "step": 3813 + }, + { + "epoch": 0.16, + "grad_norm": 0.435546875, + "learning_rate": 0.0001969126661030285, + "loss": 2.6573, + "step": 3814 + }, + { + "epoch": 0.16, + "grad_norm": 0.458984375, + "learning_rate": 0.00019691101195452512, + "loss": 2.7688, + "step": 3815 + }, + { + "epoch": 0.16, + "grad_norm": 0.408203125, + "learning_rate": 0.0001969093573699576, + "loss": 2.7236, + "step": 3816 + }, + { + "epoch": 0.16, + "grad_norm": 0.451171875, + "learning_rate": 0.00019690770234933336, + "loss": 2.6469, + "step": 3817 + }, + { + "epoch": 0.16, + "grad_norm": 0.390625, + "learning_rate": 0.0001969060468926599, + "loss": 2.6667, + "step": 3818 + }, + { + "epoch": 0.16, + "grad_norm": 0.4296875, + "learning_rate": 0.00019690439099994463, + "loss": 2.6018, + "step": 3819 + }, + { + "epoch": 0.16, + "grad_norm": 0.4296875, + "learning_rate": 0.00019690273467119505, + "loss": 2.6651, + "step": 3820 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.00019690107790641857, + "loss": 2.6509, + "step": 3821 + }, + { + "epoch": 0.16, + "grad_norm": 0.41015625, + "learning_rate": 0.00019689942070562263, + "loss": 2.4323, + "step": 3822 + }, + { + "epoch": 0.16, + "grad_norm": 0.41015625, + "learning_rate": 0.0001968977630688147, + "loss": 2.6272, + "step": 3823 + }, + { + "epoch": 0.16, + "grad_norm": 0.388671875, + "learning_rate": 0.00019689610499600227, + "loss": 2.6392, + "step": 3824 + }, + { + "epoch": 0.16, + "grad_norm": 0.365234375, + "learning_rate": 0.00019689444648719276, + "loss": 2.7723, + "step": 3825 + }, + { + "epoch": 0.16, + "grad_norm": 0.365234375, + "learning_rate": 0.00019689278754239368, + "loss": 2.5415, + "step": 3826 + }, + { + "epoch": 0.16, + "grad_norm": 0.412109375, + "learning_rate": 0.00019689112816161246, + "loss": 2.6345, + "step": 3827 + }, + { + "epoch": 0.16, + "grad_norm": 0.4140625, + "learning_rate": 0.00019688946834485655, + "loss": 2.5358, + "step": 3828 + }, + { + "epoch": 0.16, + "grad_norm": 0.4140625, + "learning_rate": 0.00019688780809213347, + "loss": 2.6892, + "step": 3829 + }, + { + "epoch": 0.16, + "grad_norm": 0.423828125, + "learning_rate": 0.00019688614740345065, + "loss": 2.6739, + "step": 3830 + }, + { + "epoch": 0.16, + "grad_norm": 0.453125, + "learning_rate": 0.00019688448627881556, + "loss": 2.522, + "step": 3831 + }, + { + "epoch": 0.16, + "grad_norm": 0.4296875, + "learning_rate": 0.00019688282471823568, + "loss": 2.6679, + "step": 3832 + }, + { + "epoch": 0.16, + "grad_norm": 0.44140625, + "learning_rate": 0.00019688116272171852, + "loss": 2.6786, + "step": 3833 + }, + { + "epoch": 0.16, + "grad_norm": 0.443359375, + "learning_rate": 0.00019687950028927155, + "loss": 2.6581, + "step": 3834 + }, + { + "epoch": 0.16, + "grad_norm": 0.404296875, + "learning_rate": 0.0001968778374209022, + "loss": 2.6647, + "step": 3835 + }, + { + "epoch": 0.16, + "grad_norm": 0.41015625, + "learning_rate": 0.000196876174116618, + "loss": 2.6631, + "step": 3836 + }, + { + "epoch": 0.16, + "grad_norm": 0.451171875, + "learning_rate": 0.0001968745103764264, + "loss": 2.754, + "step": 3837 + }, + { + "epoch": 0.16, + "grad_norm": 0.48828125, + "learning_rate": 0.00019687284620033494, + "loss": 2.8151, + "step": 3838 + }, + { + "epoch": 0.16, + "grad_norm": 0.373046875, + "learning_rate": 0.0001968711815883511, + "loss": 2.4865, + "step": 3839 + }, + { + "epoch": 0.16, + "grad_norm": 0.4453125, + "learning_rate": 0.00019686951654048228, + "loss": 2.7346, + "step": 3840 + }, + { + "epoch": 0.16, + "grad_norm": 0.3984375, + "learning_rate": 0.00019686785105673605, + "loss": 2.5654, + "step": 3841 + }, + { + "epoch": 0.16, + "grad_norm": 0.419921875, + "learning_rate": 0.00019686618513711993, + "loss": 2.6755, + "step": 3842 + }, + { + "epoch": 0.16, + "grad_norm": 0.400390625, + "learning_rate": 0.00019686451878164136, + "loss": 2.6855, + "step": 3843 + }, + { + "epoch": 0.16, + "grad_norm": 0.41015625, + "learning_rate": 0.00019686285199030785, + "loss": 2.6399, + "step": 3844 + }, + { + "epoch": 0.16, + "grad_norm": 0.478515625, + "learning_rate": 0.0001968611847631269, + "loss": 2.6353, + "step": 3845 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.00019685951710010603, + "loss": 2.6615, + "step": 3846 + }, + { + "epoch": 0.16, + "grad_norm": 0.4375, + "learning_rate": 0.00019685784900125273, + "loss": 2.7192, + "step": 3847 + }, + { + "epoch": 0.16, + "grad_norm": 0.380859375, + "learning_rate": 0.0001968561804665745, + "loss": 2.5515, + "step": 3848 + }, + { + "epoch": 0.16, + "grad_norm": 0.4453125, + "learning_rate": 0.00019685451149607886, + "loss": 2.7639, + "step": 3849 + }, + { + "epoch": 0.16, + "grad_norm": 0.3984375, + "learning_rate": 0.0001968528420897733, + "loss": 2.6406, + "step": 3850 + }, + { + "epoch": 0.16, + "grad_norm": 0.4375, + "learning_rate": 0.00019685117224766535, + "loss": 2.6907, + "step": 3851 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.00019684950196976252, + "loss": 2.636, + "step": 3852 + }, + { + "epoch": 0.16, + "grad_norm": 0.421875, + "learning_rate": 0.00019684783125607232, + "loss": 2.7762, + "step": 3853 + }, + { + "epoch": 0.16, + "grad_norm": 0.443359375, + "learning_rate": 0.0001968461601066023, + "loss": 2.7364, + "step": 3854 + }, + { + "epoch": 0.16, + "grad_norm": 0.392578125, + "learning_rate": 0.00019684448852135995, + "loss": 2.5305, + "step": 3855 + }, + { + "epoch": 0.16, + "grad_norm": 0.37890625, + "learning_rate": 0.00019684281650035276, + "loss": 2.6308, + "step": 3856 + }, + { + "epoch": 0.16, + "grad_norm": 0.439453125, + "learning_rate": 0.00019684114404358834, + "loss": 2.7379, + "step": 3857 + }, + { + "epoch": 0.16, + "grad_norm": 0.435546875, + "learning_rate": 0.00019683947115107413, + "loss": 2.7509, + "step": 3858 + }, + { + "epoch": 0.16, + "grad_norm": 0.361328125, + "learning_rate": 0.00019683779782281768, + "loss": 2.5102, + "step": 3859 + }, + { + "epoch": 0.16, + "grad_norm": 0.47265625, + "learning_rate": 0.00019683612405882656, + "loss": 2.5712, + "step": 3860 + }, + { + "epoch": 0.16, + "grad_norm": 0.53515625, + "learning_rate": 0.00019683444985910825, + "loss": 2.68, + "step": 3861 + }, + { + "epoch": 0.16, + "grad_norm": 0.4609375, + "learning_rate": 0.0001968327752236703, + "loss": 2.7084, + "step": 3862 + }, + { + "epoch": 0.16, + "grad_norm": 0.4375, + "learning_rate": 0.00019683110015252027, + "loss": 2.7099, + "step": 3863 + }, + { + "epoch": 0.16, + "grad_norm": 0.423828125, + "learning_rate": 0.00019682942464566567, + "loss": 2.7728, + "step": 3864 + }, + { + "epoch": 0.16, + "grad_norm": 0.44921875, + "learning_rate": 0.00019682774870311405, + "loss": 2.7105, + "step": 3865 + }, + { + "epoch": 0.16, + "grad_norm": 0.43359375, + "learning_rate": 0.00019682607232487294, + "loss": 2.6934, + "step": 3866 + }, + { + "epoch": 0.16, + "grad_norm": 0.42578125, + "learning_rate": 0.00019682439551094988, + "loss": 2.7139, + "step": 3867 + }, + { + "epoch": 0.16, + "grad_norm": 0.47265625, + "learning_rate": 0.00019682271826135245, + "loss": 2.7126, + "step": 3868 + }, + { + "epoch": 0.16, + "grad_norm": 0.41796875, + "learning_rate": 0.00019682104057608814, + "loss": 2.6161, + "step": 3869 + }, + { + "epoch": 0.16, + "grad_norm": 0.40234375, + "learning_rate": 0.00019681936245516456, + "loss": 2.7407, + "step": 3870 + }, + { + "epoch": 0.16, + "grad_norm": 0.4453125, + "learning_rate": 0.00019681768389858922, + "loss": 2.6654, + "step": 3871 + }, + { + "epoch": 0.16, + "grad_norm": 0.44140625, + "learning_rate": 0.00019681600490636968, + "loss": 2.7989, + "step": 3872 + }, + { + "epoch": 0.17, + "grad_norm": 0.416015625, + "learning_rate": 0.00019681432547851352, + "loss": 2.6457, + "step": 3873 + }, + { + "epoch": 0.17, + "grad_norm": 0.4609375, + "learning_rate": 0.00019681264561502825, + "loss": 2.5709, + "step": 3874 + }, + { + "epoch": 0.17, + "grad_norm": 0.396484375, + "learning_rate": 0.00019681096531592147, + "loss": 2.634, + "step": 3875 + }, + { + "epoch": 0.17, + "grad_norm": 0.439453125, + "learning_rate": 0.00019680928458120072, + "loss": 2.6184, + "step": 3876 + }, + { + "epoch": 0.17, + "grad_norm": 0.416015625, + "learning_rate": 0.00019680760341087358, + "loss": 2.7135, + "step": 3877 + }, + { + "epoch": 0.17, + "grad_norm": 0.435546875, + "learning_rate": 0.00019680592180494762, + "loss": 2.7098, + "step": 3878 + }, + { + "epoch": 0.17, + "grad_norm": 0.44921875, + "learning_rate": 0.00019680423976343035, + "loss": 2.809, + "step": 3879 + }, + { + "epoch": 0.17, + "grad_norm": 0.44140625, + "learning_rate": 0.0001968025572863294, + "loss": 2.5854, + "step": 3880 + }, + { + "epoch": 0.17, + "grad_norm": 0.419921875, + "learning_rate": 0.00019680087437365233, + "loss": 2.6917, + "step": 3881 + }, + { + "epoch": 0.17, + "grad_norm": 0.3984375, + "learning_rate": 0.0001967991910254067, + "loss": 2.7472, + "step": 3882 + }, + { + "epoch": 0.17, + "grad_norm": 0.384765625, + "learning_rate": 0.00019679750724160006, + "loss": 2.5037, + "step": 3883 + }, + { + "epoch": 0.17, + "grad_norm": 0.427734375, + "learning_rate": 0.00019679582302224003, + "loss": 2.6891, + "step": 3884 + }, + { + "epoch": 0.17, + "grad_norm": 0.396484375, + "learning_rate": 0.0001967941383673342, + "loss": 2.5814, + "step": 3885 + }, + { + "epoch": 0.17, + "grad_norm": 0.41796875, + "learning_rate": 0.00019679245327689006, + "loss": 2.6812, + "step": 3886 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.0001967907677509153, + "loss": 2.6954, + "step": 3887 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.00019678908178941745, + "loss": 2.6294, + "step": 3888 + }, + { + "epoch": 0.17, + "grad_norm": 0.404296875, + "learning_rate": 0.00019678739539240406, + "loss": 2.7811, + "step": 3889 + }, + { + "epoch": 0.17, + "grad_norm": 0.404296875, + "learning_rate": 0.00019678570855988283, + "loss": 2.7085, + "step": 3890 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.00019678402129186124, + "loss": 2.6848, + "step": 3891 + }, + { + "epoch": 0.17, + "grad_norm": 0.458984375, + "learning_rate": 0.00019678233358834694, + "loss": 2.6838, + "step": 3892 + }, + { + "epoch": 0.17, + "grad_norm": 0.431640625, + "learning_rate": 0.00019678064544934753, + "loss": 2.6556, + "step": 3893 + }, + { + "epoch": 0.17, + "grad_norm": 0.50390625, + "learning_rate": 0.00019677895687487054, + "loss": 2.7292, + "step": 3894 + }, + { + "epoch": 0.17, + "grad_norm": 0.39453125, + "learning_rate": 0.00019677726786492366, + "loss": 2.8077, + "step": 3895 + }, + { + "epoch": 0.17, + "grad_norm": 0.478515625, + "learning_rate": 0.0001967755784195144, + "loss": 2.6392, + "step": 3896 + }, + { + "epoch": 0.17, + "grad_norm": 0.419921875, + "learning_rate": 0.00019677388853865042, + "loss": 2.5572, + "step": 3897 + }, + { + "epoch": 0.17, + "grad_norm": 0.474609375, + "learning_rate": 0.0001967721982223393, + "loss": 2.6414, + "step": 3898 + }, + { + "epoch": 0.17, + "grad_norm": 0.3828125, + "learning_rate": 0.0001967705074705887, + "loss": 2.6646, + "step": 3899 + }, + { + "epoch": 0.17, + "grad_norm": 0.431640625, + "learning_rate": 0.00019676881628340613, + "loss": 2.6379, + "step": 3900 + }, + { + "epoch": 0.17, + "grad_norm": 0.50390625, + "learning_rate": 0.00019676712466079924, + "loss": 2.6906, + "step": 3901 + }, + { + "epoch": 0.17, + "grad_norm": 0.423828125, + "learning_rate": 0.00019676543260277568, + "loss": 2.7343, + "step": 3902 + }, + { + "epoch": 0.17, + "grad_norm": 0.4140625, + "learning_rate": 0.00019676374010934306, + "loss": 2.7984, + "step": 3903 + }, + { + "epoch": 0.17, + "grad_norm": 0.48046875, + "learning_rate": 0.00019676204718050894, + "loss": 2.7869, + "step": 3904 + }, + { + "epoch": 0.17, + "grad_norm": 0.435546875, + "learning_rate": 0.000196760353816281, + "loss": 2.5649, + "step": 3905 + }, + { + "epoch": 0.17, + "grad_norm": 0.4140625, + "learning_rate": 0.0001967586600166668, + "loss": 2.6526, + "step": 3906 + }, + { + "epoch": 0.17, + "grad_norm": 0.4453125, + "learning_rate": 0.000196756965781674, + "loss": 2.5881, + "step": 3907 + }, + { + "epoch": 0.17, + "grad_norm": 0.408203125, + "learning_rate": 0.00019675527111131022, + "loss": 2.5268, + "step": 3908 + }, + { + "epoch": 0.17, + "grad_norm": 0.416015625, + "learning_rate": 0.00019675357600558308, + "loss": 2.6414, + "step": 3909 + }, + { + "epoch": 0.17, + "grad_norm": 0.41015625, + "learning_rate": 0.0001967518804645002, + "loss": 2.7231, + "step": 3910 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.00019675018448806923, + "loss": 2.6936, + "step": 3911 + }, + { + "epoch": 0.17, + "grad_norm": 0.369140625, + "learning_rate": 0.00019674848807629778, + "loss": 2.5405, + "step": 3912 + }, + { + "epoch": 0.17, + "grad_norm": 0.421875, + "learning_rate": 0.0001967467912291935, + "loss": 2.6823, + "step": 3913 + }, + { + "epoch": 0.17, + "grad_norm": 0.375, + "learning_rate": 0.00019674509394676398, + "loss": 2.7007, + "step": 3914 + }, + { + "epoch": 0.17, + "grad_norm": 0.421875, + "learning_rate": 0.00019674339622901695, + "loss": 2.6782, + "step": 3915 + }, + { + "epoch": 0.17, + "grad_norm": 0.375, + "learning_rate": 0.00019674169807595994, + "loss": 2.5926, + "step": 3916 + }, + { + "epoch": 0.17, + "grad_norm": 0.421875, + "learning_rate": 0.0001967399994876007, + "loss": 2.5985, + "step": 3917 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.00019673830046394676, + "loss": 2.4709, + "step": 3918 + }, + { + "epoch": 0.17, + "grad_norm": 0.47265625, + "learning_rate": 0.00019673660100500583, + "loss": 2.6918, + "step": 3919 + }, + { + "epoch": 0.17, + "grad_norm": 0.423828125, + "learning_rate": 0.00019673490111078558, + "loss": 2.6309, + "step": 3920 + }, + { + "epoch": 0.17, + "grad_norm": 0.46484375, + "learning_rate": 0.0001967332007812936, + "loss": 2.6231, + "step": 3921 + }, + { + "epoch": 0.17, + "grad_norm": 0.369140625, + "learning_rate": 0.0001967315000165376, + "loss": 2.6508, + "step": 3922 + }, + { + "epoch": 0.17, + "grad_norm": 0.412109375, + "learning_rate": 0.00019672979881652517, + "loss": 2.7128, + "step": 3923 + }, + { + "epoch": 0.17, + "grad_norm": 0.431640625, + "learning_rate": 0.000196728097181264, + "loss": 2.6522, + "step": 3924 + }, + { + "epoch": 0.17, + "grad_norm": 0.412109375, + "learning_rate": 0.00019672639511076178, + "loss": 2.6732, + "step": 3925 + }, + { + "epoch": 0.17, + "grad_norm": 0.44921875, + "learning_rate": 0.00019672469260502607, + "loss": 2.7549, + "step": 3926 + }, + { + "epoch": 0.17, + "grad_norm": 0.443359375, + "learning_rate": 0.00019672298966406463, + "loss": 2.6943, + "step": 3927 + }, + { + "epoch": 0.17, + "grad_norm": 0.443359375, + "learning_rate": 0.00019672128628788507, + "loss": 2.7652, + "step": 3928 + }, + { + "epoch": 0.17, + "grad_norm": 0.447265625, + "learning_rate": 0.00019671958247649508, + "loss": 2.7855, + "step": 3929 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.0001967178782299023, + "loss": 2.6766, + "step": 3930 + }, + { + "epoch": 0.17, + "grad_norm": 0.41796875, + "learning_rate": 0.00019671617354811446, + "loss": 2.5149, + "step": 3931 + }, + { + "epoch": 0.17, + "grad_norm": 0.361328125, + "learning_rate": 0.00019671446843113913, + "loss": 2.6661, + "step": 3932 + }, + { + "epoch": 0.17, + "grad_norm": 0.3984375, + "learning_rate": 0.0001967127628789841, + "loss": 2.6025, + "step": 3933 + }, + { + "epoch": 0.17, + "grad_norm": 0.43359375, + "learning_rate": 0.0001967110568916569, + "loss": 2.7348, + "step": 3934 + }, + { + "epoch": 0.17, + "grad_norm": 0.3828125, + "learning_rate": 0.00019670935046916532, + "loss": 2.5497, + "step": 3935 + }, + { + "epoch": 0.17, + "grad_norm": 0.4375, + "learning_rate": 0.00019670764361151704, + "loss": 2.7258, + "step": 3936 + }, + { + "epoch": 0.17, + "grad_norm": 0.390625, + "learning_rate": 0.00019670593631871967, + "loss": 2.5949, + "step": 3937 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.00019670422859078093, + "loss": 2.4902, + "step": 3938 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.0001967025204277085, + "loss": 2.7457, + "step": 3939 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.0001967008118295101, + "loss": 2.7065, + "step": 3940 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.00019669910279619337, + "loss": 2.6037, + "step": 3941 + }, + { + "epoch": 0.17, + "grad_norm": 0.43359375, + "learning_rate": 0.000196697393327766, + "loss": 2.7678, + "step": 3942 + }, + { + "epoch": 0.17, + "grad_norm": 0.39453125, + "learning_rate": 0.00019669568342423573, + "loss": 2.6655, + "step": 3943 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.0001966939730856102, + "loss": 2.6068, + "step": 3944 + }, + { + "epoch": 0.17, + "grad_norm": 0.36328125, + "learning_rate": 0.00019669226231189716, + "loss": 2.5901, + "step": 3945 + }, + { + "epoch": 0.17, + "grad_norm": 0.462890625, + "learning_rate": 0.00019669055110310425, + "loss": 2.6817, + "step": 3946 + }, + { + "epoch": 0.17, + "grad_norm": 0.4140625, + "learning_rate": 0.00019668883945923917, + "loss": 2.8428, + "step": 3947 + }, + { + "epoch": 0.17, + "grad_norm": 0.416015625, + "learning_rate": 0.0001966871273803097, + "loss": 2.5962, + "step": 3948 + }, + { + "epoch": 0.17, + "grad_norm": 0.4609375, + "learning_rate": 0.00019668541486632347, + "loss": 2.8551, + "step": 3949 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.0001966837019172882, + "loss": 2.6452, + "step": 3950 + }, + { + "epoch": 0.17, + "grad_norm": 0.453125, + "learning_rate": 0.0001966819885332116, + "loss": 2.7143, + "step": 3951 + }, + { + "epoch": 0.17, + "grad_norm": 0.44921875, + "learning_rate": 0.0001966802747141014, + "loss": 2.8189, + "step": 3952 + }, + { + "epoch": 0.17, + "grad_norm": 0.375, + "learning_rate": 0.0001966785604599653, + "loss": 2.6021, + "step": 3953 + }, + { + "epoch": 0.17, + "grad_norm": 0.365234375, + "learning_rate": 0.00019667684577081098, + "loss": 2.5679, + "step": 3954 + }, + { + "epoch": 0.17, + "grad_norm": 0.439453125, + "learning_rate": 0.00019667513064664622, + "loss": 2.5446, + "step": 3955 + }, + { + "epoch": 0.17, + "grad_norm": 0.4375, + "learning_rate": 0.00019667341508747868, + "loss": 2.6515, + "step": 3956 + }, + { + "epoch": 0.17, + "grad_norm": 0.37890625, + "learning_rate": 0.00019667169909331613, + "loss": 2.6116, + "step": 3957 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.00019666998266416625, + "loss": 2.6825, + "step": 3958 + }, + { + "epoch": 0.17, + "grad_norm": 0.400390625, + "learning_rate": 0.00019666826580003676, + "loss": 2.6471, + "step": 3959 + }, + { + "epoch": 0.17, + "grad_norm": 0.396484375, + "learning_rate": 0.00019666654850093542, + "loss": 2.6526, + "step": 3960 + }, + { + "epoch": 0.17, + "grad_norm": 0.4765625, + "learning_rate": 0.00019666483076686992, + "loss": 2.6847, + "step": 3961 + }, + { + "epoch": 0.17, + "grad_norm": 0.478515625, + "learning_rate": 0.000196663112597848, + "loss": 2.6729, + "step": 3962 + }, + { + "epoch": 0.17, + "grad_norm": 0.5234375, + "learning_rate": 0.00019666139399387747, + "loss": 2.8312, + "step": 3963 + }, + { + "epoch": 0.17, + "grad_norm": 0.431640625, + "learning_rate": 0.00019665967495496593, + "loss": 2.6619, + "step": 3964 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.0001966579554811212, + "loss": 2.7127, + "step": 3965 + }, + { + "epoch": 0.17, + "grad_norm": 0.404296875, + "learning_rate": 0.00019665623557235098, + "loss": 2.6437, + "step": 3966 + }, + { + "epoch": 0.17, + "grad_norm": 0.404296875, + "learning_rate": 0.00019665451522866303, + "loss": 2.7177, + "step": 3967 + }, + { + "epoch": 0.17, + "grad_norm": 0.38671875, + "learning_rate": 0.00019665279445006507, + "loss": 2.5962, + "step": 3968 + }, + { + "epoch": 0.17, + "grad_norm": 0.41015625, + "learning_rate": 0.0001966510732365649, + "loss": 2.7386, + "step": 3969 + }, + { + "epoch": 0.17, + "grad_norm": 0.435546875, + "learning_rate": 0.0001966493515881702, + "loss": 2.7523, + "step": 3970 + }, + { + "epoch": 0.17, + "grad_norm": 0.390625, + "learning_rate": 0.00019664762950488874, + "loss": 2.657, + "step": 3971 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.00019664590698672825, + "loss": 2.7775, + "step": 3972 + }, + { + "epoch": 0.17, + "grad_norm": 0.458984375, + "learning_rate": 0.0001966441840336965, + "loss": 2.7752, + "step": 3973 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.00019664246064580128, + "loss": 2.6677, + "step": 3974 + }, + { + "epoch": 0.17, + "grad_norm": 0.373046875, + "learning_rate": 0.00019664073682305026, + "loss": 2.6842, + "step": 3975 + }, + { + "epoch": 0.17, + "grad_norm": 0.48046875, + "learning_rate": 0.00019663901256545128, + "loss": 2.8015, + "step": 3976 + }, + { + "epoch": 0.17, + "grad_norm": 0.458984375, + "learning_rate": 0.00019663728787301202, + "loss": 2.8048, + "step": 3977 + }, + { + "epoch": 0.17, + "grad_norm": 0.46875, + "learning_rate": 0.0001966355627457403, + "loss": 2.8165, + "step": 3978 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.00019663383718364387, + "loss": 2.7326, + "step": 3979 + }, + { + "epoch": 0.17, + "grad_norm": 0.404296875, + "learning_rate": 0.0001966321111867305, + "loss": 2.5904, + "step": 3980 + }, + { + "epoch": 0.17, + "grad_norm": 0.416015625, + "learning_rate": 0.00019663038475500793, + "loss": 2.6928, + "step": 3981 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.00019662865788848394, + "loss": 2.601, + "step": 3982 + }, + { + "epoch": 0.17, + "grad_norm": 0.4140625, + "learning_rate": 0.0001966269305871663, + "loss": 2.7493, + "step": 3983 + }, + { + "epoch": 0.17, + "grad_norm": 0.4375, + "learning_rate": 0.00019662520285106282, + "loss": 2.6966, + "step": 3984 + }, + { + "epoch": 0.17, + "grad_norm": 0.44921875, + "learning_rate": 0.0001966234746801812, + "loss": 2.4813, + "step": 3985 + }, + { + "epoch": 0.17, + "grad_norm": 0.474609375, + "learning_rate": 0.00019662174607452926, + "loss": 2.6807, + "step": 3986 + }, + { + "epoch": 0.17, + "grad_norm": 0.453125, + "learning_rate": 0.00019662001703411476, + "loss": 2.7412, + "step": 3987 + }, + { + "epoch": 0.17, + "grad_norm": 0.43359375, + "learning_rate": 0.00019661828755894553, + "loss": 2.6957, + "step": 3988 + }, + { + "epoch": 0.17, + "grad_norm": 0.484375, + "learning_rate": 0.00019661655764902932, + "loss": 2.7951, + "step": 3989 + }, + { + "epoch": 0.17, + "grad_norm": 0.44140625, + "learning_rate": 0.0001966148273043739, + "loss": 2.6497, + "step": 3990 + }, + { + "epoch": 0.17, + "grad_norm": 0.46484375, + "learning_rate": 0.00019661309652498703, + "loss": 2.763, + "step": 3991 + }, + { + "epoch": 0.17, + "grad_norm": 0.412109375, + "learning_rate": 0.00019661136531087657, + "loss": 2.7413, + "step": 3992 + }, + { + "epoch": 0.17, + "grad_norm": 0.41015625, + "learning_rate": 0.00019660963366205025, + "loss": 2.6931, + "step": 3993 + }, + { + "epoch": 0.17, + "grad_norm": 0.439453125, + "learning_rate": 0.00019660790157851593, + "loss": 2.6197, + "step": 3994 + }, + { + "epoch": 0.17, + "grad_norm": 0.380859375, + "learning_rate": 0.0001966061690602813, + "loss": 2.754, + "step": 3995 + }, + { + "epoch": 0.17, + "grad_norm": 0.439453125, + "learning_rate": 0.00019660443610735424, + "loss": 2.8171, + "step": 3996 + }, + { + "epoch": 0.17, + "grad_norm": 0.431640625, + "learning_rate": 0.00019660270271974252, + "loss": 2.7168, + "step": 3997 + }, + { + "epoch": 0.17, + "grad_norm": 0.4375, + "learning_rate": 0.00019660096889745398, + "loss": 2.5968, + "step": 3998 + }, + { + "epoch": 0.17, + "grad_norm": 0.431640625, + "learning_rate": 0.00019659923464049634, + "loss": 2.561, + "step": 3999 + }, + { + "epoch": 0.17, + "grad_norm": 0.44921875, + "learning_rate": 0.00019659749994887745, + "loss": 2.667, + "step": 4000 + }, + { + "epoch": 0.17, + "grad_norm": 0.4453125, + "learning_rate": 0.0001965957648226051, + "loss": 2.6471, + "step": 4001 + }, + { + "epoch": 0.17, + "grad_norm": 0.380859375, + "learning_rate": 0.00019659402926168714, + "loss": 2.6949, + "step": 4002 + }, + { + "epoch": 0.17, + "grad_norm": 0.45703125, + "learning_rate": 0.00019659229326613137, + "loss": 2.6233, + "step": 4003 + }, + { + "epoch": 0.17, + "grad_norm": 0.3828125, + "learning_rate": 0.00019659055683594553, + "loss": 2.6966, + "step": 4004 + }, + { + "epoch": 0.17, + "grad_norm": 0.443359375, + "learning_rate": 0.00019658881997113752, + "loss": 2.7097, + "step": 4005 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.00019658708267171512, + "loss": 2.5939, + "step": 4006 + }, + { + "epoch": 0.17, + "grad_norm": 0.41015625, + "learning_rate": 0.0001965853449376861, + "loss": 2.672, + "step": 4007 + }, + { + "epoch": 0.17, + "grad_norm": 0.384765625, + "learning_rate": 0.00019658360676905835, + "loss": 2.7448, + "step": 4008 + }, + { + "epoch": 0.17, + "grad_norm": 0.4140625, + "learning_rate": 0.0001965818681658397, + "loss": 2.7194, + "step": 4009 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.0001965801291280379, + "loss": 2.6633, + "step": 4010 + }, + { + "epoch": 0.17, + "grad_norm": 0.478515625, + "learning_rate": 0.00019657838965566083, + "loss": 2.5168, + "step": 4011 + }, + { + "epoch": 0.17, + "grad_norm": 0.419921875, + "learning_rate": 0.0001965766497487163, + "loss": 2.6707, + "step": 4012 + }, + { + "epoch": 0.17, + "grad_norm": 0.41015625, + "learning_rate": 0.00019657490940721214, + "loss": 2.6792, + "step": 4013 + }, + { + "epoch": 0.17, + "grad_norm": 0.416015625, + "learning_rate": 0.00019657316863115617, + "loss": 2.6976, + "step": 4014 + }, + { + "epoch": 0.17, + "grad_norm": 0.48046875, + "learning_rate": 0.00019657142742055626, + "loss": 2.7159, + "step": 4015 + }, + { + "epoch": 0.17, + "grad_norm": 0.39453125, + "learning_rate": 0.0001965696857754202, + "loss": 2.6247, + "step": 4016 + }, + { + "epoch": 0.17, + "grad_norm": 0.412109375, + "learning_rate": 0.00019656794369575587, + "loss": 2.7145, + "step": 4017 + }, + { + "epoch": 0.17, + "grad_norm": 0.412109375, + "learning_rate": 0.00019656620118157106, + "loss": 2.6181, + "step": 4018 + }, + { + "epoch": 0.17, + "grad_norm": 0.419921875, + "learning_rate": 0.00019656445823287364, + "loss": 2.6164, + "step": 4019 + }, + { + "epoch": 0.17, + "grad_norm": 0.41015625, + "learning_rate": 0.00019656271484967146, + "loss": 2.6862, + "step": 4020 + }, + { + "epoch": 0.17, + "grad_norm": 0.453125, + "learning_rate": 0.00019656097103197233, + "loss": 2.7161, + "step": 4021 + }, + { + "epoch": 0.17, + "grad_norm": 0.423828125, + "learning_rate": 0.00019655922677978415, + "loss": 2.6393, + "step": 4022 + }, + { + "epoch": 0.17, + "grad_norm": 0.470703125, + "learning_rate": 0.00019655748209311473, + "loss": 2.8055, + "step": 4023 + }, + { + "epoch": 0.17, + "grad_norm": 0.490234375, + "learning_rate": 0.0001965557369719719, + "loss": 2.6494, + "step": 4024 + }, + { + "epoch": 0.17, + "grad_norm": 0.40625, + "learning_rate": 0.00019655399141636358, + "loss": 2.697, + "step": 4025 + }, + { + "epoch": 0.17, + "grad_norm": 0.462890625, + "learning_rate": 0.00019655224542629754, + "loss": 2.7458, + "step": 4026 + }, + { + "epoch": 0.17, + "grad_norm": 0.408203125, + "learning_rate": 0.00019655049900178174, + "loss": 2.636, + "step": 4027 + }, + { + "epoch": 0.17, + "grad_norm": 0.47265625, + "learning_rate": 0.00019654875214282394, + "loss": 2.8309, + "step": 4028 + }, + { + "epoch": 0.17, + "grad_norm": 0.451171875, + "learning_rate": 0.00019654700484943204, + "loss": 2.5333, + "step": 4029 + }, + { + "epoch": 0.17, + "grad_norm": 0.431640625, + "learning_rate": 0.00019654525712161389, + "loss": 2.6796, + "step": 4030 + }, + { + "epoch": 0.17, + "grad_norm": 0.4921875, + "learning_rate": 0.00019654350895937742, + "loss": 2.6645, + "step": 4031 + }, + { + "epoch": 0.17, + "grad_norm": 0.4375, + "learning_rate": 0.00019654176036273038, + "loss": 2.6683, + "step": 4032 + }, + { + "epoch": 0.17, + "grad_norm": 0.46875, + "learning_rate": 0.00019654001133168075, + "loss": 2.7039, + "step": 4033 + }, + { + "epoch": 0.17, + "grad_norm": 0.376953125, + "learning_rate": 0.0001965382618662363, + "loss": 2.7538, + "step": 4034 + }, + { + "epoch": 0.17, + "grad_norm": 0.443359375, + "learning_rate": 0.00019653651196640498, + "loss": 2.7666, + "step": 4035 + }, + { + "epoch": 0.17, + "grad_norm": 0.40625, + "learning_rate": 0.00019653476163219465, + "loss": 2.672, + "step": 4036 + }, + { + "epoch": 0.17, + "grad_norm": 0.388671875, + "learning_rate": 0.00019653301086361316, + "loss": 2.7187, + "step": 4037 + }, + { + "epoch": 0.17, + "grad_norm": 0.435546875, + "learning_rate": 0.0001965312596606684, + "loss": 2.6228, + "step": 4038 + }, + { + "epoch": 0.17, + "grad_norm": 0.427734375, + "learning_rate": 0.00019652950802336825, + "loss": 2.6193, + "step": 4039 + }, + { + "epoch": 0.17, + "grad_norm": 0.39453125, + "learning_rate": 0.0001965277559517206, + "loss": 2.6483, + "step": 4040 + }, + { + "epoch": 0.17, + "grad_norm": 0.453125, + "learning_rate": 0.0001965260034457333, + "loss": 2.6116, + "step": 4041 + }, + { + "epoch": 0.17, + "grad_norm": 0.380859375, + "learning_rate": 0.00019652425050541427, + "loss": 2.5908, + "step": 4042 + }, + { + "epoch": 0.17, + "grad_norm": 0.5, + "learning_rate": 0.0001965224971307714, + "loss": 2.8443, + "step": 4043 + }, + { + "epoch": 0.17, + "grad_norm": 0.40625, + "learning_rate": 0.00019652074332181257, + "loss": 2.682, + "step": 4044 + }, + { + "epoch": 0.17, + "grad_norm": 0.416015625, + "learning_rate": 0.00019651898907854565, + "loss": 2.6048, + "step": 4045 + }, + { + "epoch": 0.17, + "grad_norm": 0.458984375, + "learning_rate": 0.00019651723440097856, + "loss": 2.8192, + "step": 4046 + }, + { + "epoch": 0.17, + "grad_norm": 0.439453125, + "learning_rate": 0.0001965154792891192, + "loss": 2.6557, + "step": 4047 + }, + { + "epoch": 0.17, + "grad_norm": 0.42578125, + "learning_rate": 0.00019651372374297546, + "loss": 2.5928, + "step": 4048 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.0001965119677625552, + "loss": 2.7272, + "step": 4049 + }, + { + "epoch": 0.17, + "grad_norm": 0.4921875, + "learning_rate": 0.00019651021134786637, + "loss": 2.5929, + "step": 4050 + }, + { + "epoch": 0.17, + "grad_norm": 0.443359375, + "learning_rate": 0.00019650845449891688, + "loss": 2.5803, + "step": 4051 + }, + { + "epoch": 0.17, + "grad_norm": 0.416015625, + "learning_rate": 0.00019650669721571462, + "loss": 2.7044, + "step": 4052 + }, + { + "epoch": 0.17, + "grad_norm": 0.466796875, + "learning_rate": 0.00019650493949826747, + "loss": 2.7435, + "step": 4053 + }, + { + "epoch": 0.17, + "grad_norm": 0.435546875, + "learning_rate": 0.00019650318134658337, + "loss": 2.6543, + "step": 4054 + }, + { + "epoch": 0.17, + "grad_norm": 0.443359375, + "learning_rate": 0.00019650142276067022, + "loss": 2.7046, + "step": 4055 + }, + { + "epoch": 0.17, + "grad_norm": 0.419921875, + "learning_rate": 0.0001964996637405359, + "loss": 2.6477, + "step": 4056 + }, + { + "epoch": 0.17, + "grad_norm": 0.388671875, + "learning_rate": 0.0001964979042861884, + "loss": 2.7195, + "step": 4057 + }, + { + "epoch": 0.17, + "grad_norm": 0.39453125, + "learning_rate": 0.00019649614439763558, + "loss": 2.5424, + "step": 4058 + }, + { + "epoch": 0.17, + "grad_norm": 0.44140625, + "learning_rate": 0.00019649438407488541, + "loss": 2.6208, + "step": 4059 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.00019649262331794572, + "loss": 2.5729, + "step": 4060 + }, + { + "epoch": 0.17, + "grad_norm": 0.43359375, + "learning_rate": 0.00019649086212682453, + "loss": 2.772, + "step": 4061 + }, + { + "epoch": 0.17, + "grad_norm": 0.435546875, + "learning_rate": 0.00019648910050152973, + "loss": 2.6357, + "step": 4062 + }, + { + "epoch": 0.17, + "grad_norm": 0.4609375, + "learning_rate": 0.0001964873384420692, + "loss": 2.8062, + "step": 4063 + }, + { + "epoch": 0.17, + "grad_norm": 0.484375, + "learning_rate": 0.00019648557594845095, + "loss": 2.749, + "step": 4064 + }, + { + "epoch": 0.17, + "grad_norm": 0.375, + "learning_rate": 0.00019648381302068285, + "loss": 2.6119, + "step": 4065 + }, + { + "epoch": 0.17, + "grad_norm": 0.419921875, + "learning_rate": 0.00019648204965877283, + "loss": 2.7951, + "step": 4066 + }, + { + "epoch": 0.17, + "grad_norm": 0.423828125, + "learning_rate": 0.0001964802858627289, + "loss": 2.5619, + "step": 4067 + }, + { + "epoch": 0.17, + "grad_norm": 0.439453125, + "learning_rate": 0.0001964785216325589, + "loss": 2.601, + "step": 4068 + }, + { + "epoch": 0.17, + "grad_norm": 0.484375, + "learning_rate": 0.00019647675696827084, + "loss": 2.6557, + "step": 4069 + }, + { + "epoch": 0.17, + "grad_norm": 0.462890625, + "learning_rate": 0.0001964749918698726, + "loss": 2.765, + "step": 4070 + }, + { + "epoch": 0.17, + "grad_norm": 0.392578125, + "learning_rate": 0.00019647322633737217, + "loss": 2.6886, + "step": 4071 + }, + { + "epoch": 0.17, + "grad_norm": 0.4609375, + "learning_rate": 0.00019647146037077748, + "loss": 2.6462, + "step": 4072 + }, + { + "epoch": 0.17, + "grad_norm": 0.43359375, + "learning_rate": 0.00019646969397009648, + "loss": 2.7368, + "step": 4073 + }, + { + "epoch": 0.17, + "grad_norm": 0.3984375, + "learning_rate": 0.00019646792713533712, + "loss": 2.6483, + "step": 4074 + }, + { + "epoch": 0.17, + "grad_norm": 0.44921875, + "learning_rate": 0.0001964661598665073, + "loss": 2.6305, + "step": 4075 + }, + { + "epoch": 0.17, + "grad_norm": 0.4140625, + "learning_rate": 0.00019646439216361507, + "loss": 2.676, + "step": 4076 + }, + { + "epoch": 0.17, + "grad_norm": 0.458984375, + "learning_rate": 0.00019646262402666831, + "loss": 2.7507, + "step": 4077 + }, + { + "epoch": 0.17, + "grad_norm": 0.396484375, + "learning_rate": 0.000196460855455675, + "loss": 2.6311, + "step": 4078 + }, + { + "epoch": 0.17, + "grad_norm": 0.451171875, + "learning_rate": 0.00019645908645064307, + "loss": 2.8153, + "step": 4079 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.0001964573170115805, + "loss": 2.7092, + "step": 4080 + }, + { + "epoch": 0.17, + "grad_norm": 0.33984375, + "learning_rate": 0.00019645554713849527, + "loss": 2.6135, + "step": 4081 + }, + { + "epoch": 0.17, + "grad_norm": 0.4765625, + "learning_rate": 0.00019645377683139535, + "loss": 2.7213, + "step": 4082 + }, + { + "epoch": 0.17, + "grad_norm": 0.392578125, + "learning_rate": 0.00019645200609028866, + "loss": 2.5568, + "step": 4083 + }, + { + "epoch": 0.17, + "grad_norm": 0.376953125, + "learning_rate": 0.0001964502349151832, + "loss": 2.61, + "step": 4084 + }, + { + "epoch": 0.17, + "grad_norm": 0.40234375, + "learning_rate": 0.0001964484633060869, + "loss": 2.5853, + "step": 4085 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.00019644669126300782, + "loss": 2.6533, + "step": 4086 + }, + { + "epoch": 0.17, + "grad_norm": 0.427734375, + "learning_rate": 0.00019644491878595383, + "loss": 2.5765, + "step": 4087 + }, + { + "epoch": 0.17, + "grad_norm": 0.447265625, + "learning_rate": 0.00019644314587493296, + "loss": 2.5559, + "step": 4088 + }, + { + "epoch": 0.17, + "grad_norm": 0.412109375, + "learning_rate": 0.00019644137252995317, + "loss": 2.6147, + "step": 4089 + }, + { + "epoch": 0.17, + "grad_norm": 0.388671875, + "learning_rate": 0.0001964395987510225, + "loss": 2.5645, + "step": 4090 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.00019643782453814882, + "loss": 2.5208, + "step": 4091 + }, + { + "epoch": 0.17, + "grad_norm": 0.4609375, + "learning_rate": 0.0001964360498913402, + "loss": 2.6948, + "step": 4092 + }, + { + "epoch": 0.17, + "grad_norm": 0.38671875, + "learning_rate": 0.00019643427481060458, + "loss": 2.7115, + "step": 4093 + }, + { + "epoch": 0.17, + "grad_norm": 0.4140625, + "learning_rate": 0.00019643249929595, + "loss": 2.7106, + "step": 4094 + }, + { + "epoch": 0.17, + "grad_norm": 0.40625, + "learning_rate": 0.00019643072334738442, + "loss": 2.6105, + "step": 4095 + }, + { + "epoch": 0.17, + "grad_norm": 0.376953125, + "learning_rate": 0.00019642894696491578, + "loss": 2.5932, + "step": 4096 + }, + { + "epoch": 0.17, + "grad_norm": 0.431640625, + "learning_rate": 0.00019642717014855214, + "loss": 2.52, + "step": 4097 + }, + { + "epoch": 0.17, + "grad_norm": 0.4140625, + "learning_rate": 0.00019642539289830148, + "loss": 2.62, + "step": 4098 + }, + { + "epoch": 0.17, + "grad_norm": 0.447265625, + "learning_rate": 0.0001964236152141718, + "loss": 2.6406, + "step": 4099 + }, + { + "epoch": 0.17, + "grad_norm": 0.453125, + "learning_rate": 0.0001964218370961711, + "loss": 2.7449, + "step": 4100 + }, + { + "epoch": 0.17, + "grad_norm": 0.447265625, + "learning_rate": 0.00019642005854430733, + "loss": 2.661, + "step": 4101 + }, + { + "epoch": 0.17, + "grad_norm": 0.43359375, + "learning_rate": 0.00019641827955858856, + "loss": 2.509, + "step": 4102 + }, + { + "epoch": 0.17, + "grad_norm": 0.380859375, + "learning_rate": 0.00019641650013902276, + "loss": 2.7245, + "step": 4103 + }, + { + "epoch": 0.17, + "grad_norm": 0.447265625, + "learning_rate": 0.00019641472028561797, + "loss": 2.6321, + "step": 4104 + }, + { + "epoch": 0.17, + "grad_norm": 0.4296875, + "learning_rate": 0.00019641293999838214, + "loss": 2.6323, + "step": 4105 + }, + { + "epoch": 0.17, + "grad_norm": 0.43359375, + "learning_rate": 0.00019641115927732333, + "loss": 2.8076, + "step": 4106 + }, + { + "epoch": 0.18, + "grad_norm": 0.416015625, + "learning_rate": 0.00019640937812244954, + "loss": 2.7235, + "step": 4107 + }, + { + "epoch": 0.18, + "grad_norm": 0.42578125, + "learning_rate": 0.00019640759653376878, + "loss": 2.6482, + "step": 4108 + }, + { + "epoch": 0.18, + "grad_norm": 0.388671875, + "learning_rate": 0.00019640581451128906, + "loss": 2.7559, + "step": 4109 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.0001964040320550184, + "loss": 2.6274, + "step": 4110 + }, + { + "epoch": 0.18, + "grad_norm": 0.380859375, + "learning_rate": 0.00019640224916496485, + "loss": 2.5294, + "step": 4111 + }, + { + "epoch": 0.18, + "grad_norm": 0.498046875, + "learning_rate": 0.00019640046584113642, + "loss": 2.8008, + "step": 4112 + }, + { + "epoch": 0.18, + "grad_norm": 0.37109375, + "learning_rate": 0.0001963986820835411, + "loss": 2.6447, + "step": 4113 + }, + { + "epoch": 0.18, + "grad_norm": 0.4140625, + "learning_rate": 0.00019639689789218695, + "loss": 2.7676, + "step": 4114 + }, + { + "epoch": 0.18, + "grad_norm": 0.447265625, + "learning_rate": 0.000196395113267082, + "loss": 2.7812, + "step": 4115 + }, + { + "epoch": 0.18, + "grad_norm": 0.41796875, + "learning_rate": 0.00019639332820823425, + "loss": 2.6728, + "step": 4116 + }, + { + "epoch": 0.18, + "grad_norm": 0.427734375, + "learning_rate": 0.00019639154271565174, + "loss": 2.7805, + "step": 4117 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019638975678934253, + "loss": 2.655, + "step": 4118 + }, + { + "epoch": 0.18, + "grad_norm": 0.44140625, + "learning_rate": 0.00019638797042931464, + "loss": 2.7487, + "step": 4119 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.0001963861836355761, + "loss": 2.6684, + "step": 4120 + }, + { + "epoch": 0.18, + "grad_norm": 0.376953125, + "learning_rate": 0.00019638439640813496, + "loss": 2.6786, + "step": 4121 + }, + { + "epoch": 0.18, + "grad_norm": 0.419921875, + "learning_rate": 0.00019638260874699928, + "loss": 2.6522, + "step": 4122 + }, + { + "epoch": 0.18, + "grad_norm": 0.458984375, + "learning_rate": 0.00019638082065217706, + "loss": 2.6628, + "step": 4123 + }, + { + "epoch": 0.18, + "grad_norm": 0.39453125, + "learning_rate": 0.00019637903212367638, + "loss": 2.8071, + "step": 4124 + }, + { + "epoch": 0.18, + "grad_norm": 0.443359375, + "learning_rate": 0.0001963772431615053, + "loss": 2.6666, + "step": 4125 + }, + { + "epoch": 0.18, + "grad_norm": 0.373046875, + "learning_rate": 0.0001963754537656718, + "loss": 2.5542, + "step": 4126 + }, + { + "epoch": 0.18, + "grad_norm": 0.39453125, + "learning_rate": 0.00019637366393618404, + "loss": 2.5711, + "step": 4127 + }, + { + "epoch": 0.18, + "grad_norm": 0.376953125, + "learning_rate": 0.00019637187367304998, + "loss": 2.7017, + "step": 4128 + }, + { + "epoch": 0.18, + "grad_norm": 0.4140625, + "learning_rate": 0.0001963700829762777, + "loss": 2.7392, + "step": 4129 + }, + { + "epoch": 0.18, + "grad_norm": 0.42578125, + "learning_rate": 0.00019636829184587528, + "loss": 2.6211, + "step": 4130 + }, + { + "epoch": 0.18, + "grad_norm": 0.384765625, + "learning_rate": 0.00019636650028185076, + "loss": 2.6789, + "step": 4131 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.0001963647082842122, + "loss": 2.67, + "step": 4132 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.00019636291585296768, + "loss": 2.6221, + "step": 4133 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019636112298812526, + "loss": 2.6905, + "step": 4134 + }, + { + "epoch": 0.18, + "grad_norm": 0.4921875, + "learning_rate": 0.000196359329689693, + "loss": 2.8263, + "step": 4135 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.00019635753595767896, + "loss": 2.6023, + "step": 4136 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.00019635574179209124, + "loss": 2.5503, + "step": 4137 + }, + { + "epoch": 0.18, + "grad_norm": 0.412109375, + "learning_rate": 0.00019635394719293788, + "loss": 2.6402, + "step": 4138 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019635215216022694, + "loss": 2.6097, + "step": 4139 + }, + { + "epoch": 0.18, + "grad_norm": 0.41796875, + "learning_rate": 0.0001963503566939666, + "loss": 2.696, + "step": 4140 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019634856079416482, + "loss": 2.6809, + "step": 4141 + }, + { + "epoch": 0.18, + "grad_norm": 0.423828125, + "learning_rate": 0.00019634676446082976, + "loss": 2.7076, + "step": 4142 + }, + { + "epoch": 0.18, + "grad_norm": 0.44140625, + "learning_rate": 0.00019634496769396943, + "loss": 2.6087, + "step": 4143 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.00019634317049359198, + "loss": 2.6192, + "step": 4144 + }, + { + "epoch": 0.18, + "grad_norm": 0.40234375, + "learning_rate": 0.00019634137285970545, + "loss": 2.669, + "step": 4145 + }, + { + "epoch": 0.18, + "grad_norm": 0.40234375, + "learning_rate": 0.00019633957479231798, + "loss": 2.6228, + "step": 4146 + }, + { + "epoch": 0.18, + "grad_norm": 0.494140625, + "learning_rate": 0.0001963377762914376, + "loss": 2.8229, + "step": 4147 + }, + { + "epoch": 0.18, + "grad_norm": 0.416015625, + "learning_rate": 0.00019633597735707248, + "loss": 2.5966, + "step": 4148 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.00019633417798923063, + "loss": 2.6532, + "step": 4149 + }, + { + "epoch": 0.18, + "grad_norm": 0.416015625, + "learning_rate": 0.0001963323781879202, + "loss": 2.622, + "step": 4150 + }, + { + "epoch": 0.18, + "grad_norm": 0.369140625, + "learning_rate": 0.00019633057795314927, + "loss": 2.6146, + "step": 4151 + }, + { + "epoch": 0.18, + "grad_norm": 0.439453125, + "learning_rate": 0.00019632877728492597, + "loss": 2.5328, + "step": 4152 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019632697618325834, + "loss": 2.5881, + "step": 4153 + }, + { + "epoch": 0.18, + "grad_norm": 0.39453125, + "learning_rate": 0.00019632517464815451, + "loss": 2.6811, + "step": 4154 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.0001963233726796226, + "loss": 2.6702, + "step": 4155 + }, + { + "epoch": 0.18, + "grad_norm": 0.412109375, + "learning_rate": 0.00019632157027767074, + "loss": 2.5854, + "step": 4156 + }, + { + "epoch": 0.18, + "grad_norm": 0.390625, + "learning_rate": 0.00019631976744230705, + "loss": 2.5285, + "step": 4157 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019631796417353954, + "loss": 2.7523, + "step": 4158 + }, + { + "epoch": 0.18, + "grad_norm": 0.4140625, + "learning_rate": 0.00019631616047137643, + "loss": 2.6273, + "step": 4159 + }, + { + "epoch": 0.18, + "grad_norm": 0.453125, + "learning_rate": 0.0001963143563358258, + "loss": 2.5872, + "step": 4160 + }, + { + "epoch": 0.18, + "grad_norm": 0.44921875, + "learning_rate": 0.00019631255176689574, + "loss": 2.7129, + "step": 4161 + }, + { + "epoch": 0.18, + "grad_norm": 0.423828125, + "learning_rate": 0.0001963107467645944, + "loss": 2.563, + "step": 4162 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.0001963089413289299, + "loss": 2.5863, + "step": 4163 + }, + { + "epoch": 0.18, + "grad_norm": 0.412109375, + "learning_rate": 0.00019630713545991038, + "loss": 2.6808, + "step": 4164 + }, + { + "epoch": 0.18, + "grad_norm": 0.453125, + "learning_rate": 0.0001963053291575439, + "loss": 2.5871, + "step": 4165 + }, + { + "epoch": 0.18, + "grad_norm": 0.427734375, + "learning_rate": 0.00019630352242183867, + "loss": 2.729, + "step": 4166 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.00019630171525280278, + "loss": 2.5753, + "step": 4167 + }, + { + "epoch": 0.18, + "grad_norm": 0.419921875, + "learning_rate": 0.00019629990765044437, + "loss": 2.5648, + "step": 4168 + }, + { + "epoch": 0.18, + "grad_norm": 0.416015625, + "learning_rate": 0.00019629809961477156, + "loss": 2.8764, + "step": 4169 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.0001962962911457925, + "loss": 2.5643, + "step": 4170 + }, + { + "epoch": 0.18, + "grad_norm": 0.408203125, + "learning_rate": 0.00019629448224351532, + "loss": 2.6975, + "step": 4171 + }, + { + "epoch": 0.18, + "grad_norm": 0.3671875, + "learning_rate": 0.00019629267290794818, + "loss": 2.617, + "step": 4172 + }, + { + "epoch": 0.18, + "grad_norm": 0.400390625, + "learning_rate": 0.00019629086313909918, + "loss": 2.7321, + "step": 4173 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.0001962890529369765, + "loss": 2.5993, + "step": 4174 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.00019628724230158823, + "loss": 2.6633, + "step": 4175 + }, + { + "epoch": 0.18, + "grad_norm": 0.427734375, + "learning_rate": 0.00019628543123294262, + "loss": 2.6038, + "step": 4176 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.0001962836197310477, + "loss": 2.7399, + "step": 4177 + }, + { + "epoch": 0.18, + "grad_norm": 0.41796875, + "learning_rate": 0.00019628180779591173, + "loss": 2.6794, + "step": 4178 + }, + { + "epoch": 0.18, + "grad_norm": 0.384765625, + "learning_rate": 0.00019627999542754277, + "loss": 2.6809, + "step": 4179 + }, + { + "epoch": 0.18, + "grad_norm": 0.41796875, + "learning_rate": 0.00019627818262594904, + "loss": 2.5878, + "step": 4180 + }, + { + "epoch": 0.18, + "grad_norm": 0.41796875, + "learning_rate": 0.0001962763693911387, + "loss": 2.5405, + "step": 4181 + }, + { + "epoch": 0.18, + "grad_norm": 0.42578125, + "learning_rate": 0.00019627455572311984, + "loss": 2.748, + "step": 4182 + }, + { + "epoch": 0.18, + "grad_norm": 0.43359375, + "learning_rate": 0.00019627274162190066, + "loss": 2.6512, + "step": 4183 + }, + { + "epoch": 0.18, + "grad_norm": 0.392578125, + "learning_rate": 0.00019627092708748934, + "loss": 2.5744, + "step": 4184 + }, + { + "epoch": 0.18, + "grad_norm": 0.40625, + "learning_rate": 0.00019626911211989404, + "loss": 2.8217, + "step": 4185 + }, + { + "epoch": 0.18, + "grad_norm": 0.37890625, + "learning_rate": 0.00019626729671912293, + "loss": 2.4955, + "step": 4186 + }, + { + "epoch": 0.18, + "grad_norm": 0.498046875, + "learning_rate": 0.00019626548088518412, + "loss": 2.7546, + "step": 4187 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.00019626366461808587, + "loss": 2.6057, + "step": 4188 + }, + { + "epoch": 0.18, + "grad_norm": 0.40625, + "learning_rate": 0.0001962618479178363, + "loss": 2.7432, + "step": 4189 + }, + { + "epoch": 0.18, + "grad_norm": 0.455078125, + "learning_rate": 0.0001962600307844436, + "loss": 2.7806, + "step": 4190 + }, + { + "epoch": 0.18, + "grad_norm": 0.400390625, + "learning_rate": 0.00019625821321791592, + "loss": 2.6203, + "step": 4191 + }, + { + "epoch": 0.18, + "grad_norm": 0.45703125, + "learning_rate": 0.00019625639521826148, + "loss": 2.7271, + "step": 4192 + }, + { + "epoch": 0.18, + "grad_norm": 0.447265625, + "learning_rate": 0.0001962545767854884, + "loss": 2.7698, + "step": 4193 + }, + { + "epoch": 0.18, + "grad_norm": 0.423828125, + "learning_rate": 0.00019625275791960496, + "loss": 2.7859, + "step": 4194 + }, + { + "epoch": 0.18, + "grad_norm": 0.42578125, + "learning_rate": 0.00019625093862061922, + "loss": 2.7084, + "step": 4195 + }, + { + "epoch": 0.18, + "grad_norm": 0.451171875, + "learning_rate": 0.0001962491188885395, + "loss": 2.623, + "step": 4196 + }, + { + "epoch": 0.18, + "grad_norm": 0.447265625, + "learning_rate": 0.0001962472987233739, + "loss": 2.6696, + "step": 4197 + }, + { + "epoch": 0.18, + "grad_norm": 0.466796875, + "learning_rate": 0.00019624547812513058, + "loss": 2.6683, + "step": 4198 + }, + { + "epoch": 0.18, + "grad_norm": 0.419921875, + "learning_rate": 0.00019624365709381783, + "loss": 2.6646, + "step": 4199 + }, + { + "epoch": 0.18, + "grad_norm": 0.384765625, + "learning_rate": 0.00019624183562944378, + "loss": 2.7349, + "step": 4200 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019624001373201667, + "loss": 2.7146, + "step": 4201 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.00019623819140154462, + "loss": 2.7053, + "step": 4202 + }, + { + "epoch": 0.18, + "grad_norm": 0.48046875, + "learning_rate": 0.00019623636863803592, + "loss": 2.7741, + "step": 4203 + }, + { + "epoch": 0.18, + "grad_norm": 0.44140625, + "learning_rate": 0.0001962345454414987, + "loss": 2.6186, + "step": 4204 + }, + { + "epoch": 0.18, + "grad_norm": 0.419921875, + "learning_rate": 0.0001962327218119412, + "loss": 2.6594, + "step": 4205 + }, + { + "epoch": 0.18, + "grad_norm": 0.408203125, + "learning_rate": 0.00019623089774937165, + "loss": 2.6453, + "step": 4206 + }, + { + "epoch": 0.18, + "grad_norm": 0.43359375, + "learning_rate": 0.00019622907325379823, + "loss": 2.6955, + "step": 4207 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019622724832522913, + "loss": 2.6683, + "step": 4208 + }, + { + "epoch": 0.18, + "grad_norm": 0.46484375, + "learning_rate": 0.0001962254229636726, + "loss": 2.5892, + "step": 4209 + }, + { + "epoch": 0.18, + "grad_norm": 0.44140625, + "learning_rate": 0.0001962235971691368, + "loss": 2.7086, + "step": 4210 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019622177094162998, + "loss": 2.687, + "step": 4211 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.00019621994428116034, + "loss": 2.6479, + "step": 4212 + }, + { + "epoch": 0.18, + "grad_norm": 0.392578125, + "learning_rate": 0.00019621811718773617, + "loss": 2.6684, + "step": 4213 + }, + { + "epoch": 0.18, + "grad_norm": 0.458984375, + "learning_rate": 0.0001962162896613656, + "loss": 2.76, + "step": 4214 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.0001962144617020569, + "loss": 2.6331, + "step": 4215 + }, + { + "epoch": 0.18, + "grad_norm": 0.455078125, + "learning_rate": 0.0001962126333098183, + "loss": 2.5096, + "step": 4216 + }, + { + "epoch": 0.18, + "grad_norm": 0.4375, + "learning_rate": 0.00019621080448465797, + "loss": 2.6907, + "step": 4217 + }, + { + "epoch": 0.18, + "grad_norm": 0.416015625, + "learning_rate": 0.00019620897522658424, + "loss": 2.6494, + "step": 4218 + }, + { + "epoch": 0.18, + "grad_norm": 0.408203125, + "learning_rate": 0.00019620714553560524, + "loss": 2.6869, + "step": 4219 + }, + { + "epoch": 0.18, + "grad_norm": 0.42578125, + "learning_rate": 0.00019620531541172923, + "loss": 2.6535, + "step": 4220 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.0001962034848549645, + "loss": 2.7368, + "step": 4221 + }, + { + "epoch": 0.18, + "grad_norm": 0.41796875, + "learning_rate": 0.00019620165386531922, + "loss": 2.617, + "step": 4222 + }, + { + "epoch": 0.18, + "grad_norm": 0.39453125, + "learning_rate": 0.00019619982244280167, + "loss": 2.7286, + "step": 4223 + }, + { + "epoch": 0.18, + "grad_norm": 0.419921875, + "learning_rate": 0.00019619799058742007, + "loss": 2.6361, + "step": 4224 + }, + { + "epoch": 0.18, + "grad_norm": 0.447265625, + "learning_rate": 0.00019619615829918264, + "loss": 2.7211, + "step": 4225 + }, + { + "epoch": 0.18, + "grad_norm": 0.44140625, + "learning_rate": 0.00019619432557809772, + "loss": 2.6037, + "step": 4226 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.00019619249242417344, + "loss": 2.7252, + "step": 4227 + }, + { + "epoch": 0.18, + "grad_norm": 0.412109375, + "learning_rate": 0.00019619065883741808, + "loss": 2.6995, + "step": 4228 + }, + { + "epoch": 0.18, + "grad_norm": 0.427734375, + "learning_rate": 0.00019618882481783998, + "loss": 2.64, + "step": 4229 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.00019618699036544728, + "loss": 2.671, + "step": 4230 + }, + { + "epoch": 0.18, + "grad_norm": 0.443359375, + "learning_rate": 0.00019618515548024828, + "loss": 2.7269, + "step": 4231 + }, + { + "epoch": 0.18, + "grad_norm": 0.462890625, + "learning_rate": 0.0001961833201622512, + "loss": 2.7241, + "step": 4232 + }, + { + "epoch": 0.18, + "grad_norm": 0.451171875, + "learning_rate": 0.0001961814844114644, + "loss": 2.6321, + "step": 4233 + }, + { + "epoch": 0.18, + "grad_norm": 0.400390625, + "learning_rate": 0.000196179648227896, + "loss": 2.5584, + "step": 4234 + }, + { + "epoch": 0.18, + "grad_norm": 0.4609375, + "learning_rate": 0.00019617781161155437, + "loss": 2.6342, + "step": 4235 + }, + { + "epoch": 0.18, + "grad_norm": 0.439453125, + "learning_rate": 0.0001961759745624477, + "loss": 2.499, + "step": 4236 + }, + { + "epoch": 0.18, + "grad_norm": 0.455078125, + "learning_rate": 0.00019617413708058435, + "loss": 2.6131, + "step": 4237 + }, + { + "epoch": 0.18, + "grad_norm": 0.4609375, + "learning_rate": 0.0001961722991659725, + "loss": 2.7253, + "step": 4238 + }, + { + "epoch": 0.18, + "grad_norm": 0.3984375, + "learning_rate": 0.00019617046081862045, + "loss": 2.6246, + "step": 4239 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.00019616862203853647, + "loss": 2.6597, + "step": 4240 + }, + { + "epoch": 0.18, + "grad_norm": 0.390625, + "learning_rate": 0.00019616678282572883, + "loss": 2.605, + "step": 4241 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.00019616494318020584, + "loss": 2.5739, + "step": 4242 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.00019616310310197573, + "loss": 2.7286, + "step": 4243 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019616126259104682, + "loss": 2.5689, + "step": 4244 + }, + { + "epoch": 0.18, + "grad_norm": 0.40234375, + "learning_rate": 0.00019615942164742734, + "loss": 2.7326, + "step": 4245 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.0001961575802711256, + "loss": 2.5435, + "step": 4246 + }, + { + "epoch": 0.18, + "grad_norm": 0.3828125, + "learning_rate": 0.00019615573846214993, + "loss": 2.6487, + "step": 4247 + }, + { + "epoch": 0.18, + "grad_norm": 0.5, + "learning_rate": 0.00019615389622050855, + "loss": 2.8361, + "step": 4248 + }, + { + "epoch": 0.18, + "grad_norm": 0.466796875, + "learning_rate": 0.0001961520535462098, + "loss": 2.6308, + "step": 4249 + }, + { + "epoch": 0.18, + "grad_norm": 0.466796875, + "learning_rate": 0.00019615021043926193, + "loss": 2.7735, + "step": 4250 + }, + { + "epoch": 0.18, + "grad_norm": 0.419921875, + "learning_rate": 0.00019614836689967326, + "loss": 2.6189, + "step": 4251 + }, + { + "epoch": 0.18, + "grad_norm": 0.419921875, + "learning_rate": 0.0001961465229274521, + "loss": 2.7127, + "step": 4252 + }, + { + "epoch": 0.18, + "grad_norm": 0.4140625, + "learning_rate": 0.00019614467852260666, + "loss": 2.6358, + "step": 4253 + }, + { + "epoch": 0.18, + "grad_norm": 0.412109375, + "learning_rate": 0.00019614283368514534, + "loss": 2.687, + "step": 4254 + }, + { + "epoch": 0.18, + "grad_norm": 0.42578125, + "learning_rate": 0.0001961409884150764, + "loss": 2.7395, + "step": 4255 + }, + { + "epoch": 0.18, + "grad_norm": 0.408203125, + "learning_rate": 0.00019613914271240818, + "loss": 2.5929, + "step": 4256 + }, + { + "epoch": 0.18, + "grad_norm": 0.466796875, + "learning_rate": 0.0001961372965771489, + "loss": 2.7251, + "step": 4257 + }, + { + "epoch": 0.18, + "grad_norm": 0.400390625, + "learning_rate": 0.00019613545000930695, + "loss": 2.6395, + "step": 4258 + }, + { + "epoch": 0.18, + "grad_norm": 0.44921875, + "learning_rate": 0.0001961336030088906, + "loss": 2.6734, + "step": 4259 + }, + { + "epoch": 0.18, + "grad_norm": 0.4140625, + "learning_rate": 0.00019613175557590818, + "loss": 2.6838, + "step": 4260 + }, + { + "epoch": 0.18, + "grad_norm": 0.44140625, + "learning_rate": 0.000196129907710368, + "loss": 2.7128, + "step": 4261 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.00019612805941227835, + "loss": 2.6104, + "step": 4262 + }, + { + "epoch": 0.18, + "grad_norm": 0.408203125, + "learning_rate": 0.00019612621068164753, + "loss": 2.5711, + "step": 4263 + }, + { + "epoch": 0.18, + "grad_norm": 0.392578125, + "learning_rate": 0.00019612436151848395, + "loss": 2.5698, + "step": 4264 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.00019612251192279584, + "loss": 2.5841, + "step": 4265 + }, + { + "epoch": 0.18, + "grad_norm": 0.419921875, + "learning_rate": 0.00019612066189459158, + "loss": 2.7421, + "step": 4266 + }, + { + "epoch": 0.18, + "grad_norm": 0.43359375, + "learning_rate": 0.00019611881143387943, + "loss": 2.6125, + "step": 4267 + }, + { + "epoch": 0.18, + "grad_norm": 0.48828125, + "learning_rate": 0.0001961169605406678, + "loss": 2.78, + "step": 4268 + }, + { + "epoch": 0.18, + "grad_norm": 0.427734375, + "learning_rate": 0.00019611510921496494, + "loss": 2.6646, + "step": 4269 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.00019611325745677924, + "loss": 2.6942, + "step": 4270 + }, + { + "epoch": 0.18, + "grad_norm": 0.4609375, + "learning_rate": 0.000196111405266119, + "loss": 2.6565, + "step": 4271 + }, + { + "epoch": 0.18, + "grad_norm": 0.421875, + "learning_rate": 0.00019610955264299257, + "loss": 2.6427, + "step": 4272 + }, + { + "epoch": 0.18, + "grad_norm": 0.498046875, + "learning_rate": 0.00019610769958740824, + "loss": 2.6611, + "step": 4273 + }, + { + "epoch": 0.18, + "grad_norm": 0.455078125, + "learning_rate": 0.00019610584609937442, + "loss": 2.8059, + "step": 4274 + }, + { + "epoch": 0.18, + "grad_norm": 0.41015625, + "learning_rate": 0.0001961039921788994, + "loss": 2.618, + "step": 4275 + }, + { + "epoch": 0.18, + "grad_norm": 0.49609375, + "learning_rate": 0.00019610213782599153, + "loss": 2.7238, + "step": 4276 + }, + { + "epoch": 0.18, + "grad_norm": 0.423828125, + "learning_rate": 0.00019610028304065918, + "loss": 2.582, + "step": 4277 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.0001960984278229107, + "loss": 2.5319, + "step": 4278 + }, + { + "epoch": 0.18, + "grad_norm": 0.5, + "learning_rate": 0.00019609657217275437, + "loss": 2.8098, + "step": 4279 + }, + { + "epoch": 0.18, + "grad_norm": 0.42578125, + "learning_rate": 0.0001960947160901986, + "loss": 2.6761, + "step": 4280 + }, + { + "epoch": 0.18, + "grad_norm": 0.46875, + "learning_rate": 0.00019609285957525172, + "loss": 2.7028, + "step": 4281 + }, + { + "epoch": 0.18, + "grad_norm": 0.4375, + "learning_rate": 0.0001960910026279221, + "loss": 2.5722, + "step": 4282 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.00019608914524821806, + "loss": 2.707, + "step": 4283 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.00019608728743614803, + "loss": 2.7311, + "step": 4284 + }, + { + "epoch": 0.18, + "grad_norm": 0.478515625, + "learning_rate": 0.00019608542919172028, + "loss": 2.7177, + "step": 4285 + }, + { + "epoch": 0.18, + "grad_norm": 0.5078125, + "learning_rate": 0.0001960835705149432, + "loss": 2.7701, + "step": 4286 + }, + { + "epoch": 0.18, + "grad_norm": 0.46875, + "learning_rate": 0.00019608171140582522, + "loss": 2.8001, + "step": 4287 + }, + { + "epoch": 0.18, + "grad_norm": 0.43359375, + "learning_rate": 0.00019607985186437463, + "loss": 2.7242, + "step": 4288 + }, + { + "epoch": 0.18, + "grad_norm": 0.458984375, + "learning_rate": 0.0001960779918905998, + "loss": 2.7761, + "step": 4289 + }, + { + "epoch": 0.18, + "grad_norm": 0.44140625, + "learning_rate": 0.00019607613148450915, + "loss": 2.6242, + "step": 4290 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.00019607427064611104, + "loss": 2.6427, + "step": 4291 + }, + { + "epoch": 0.18, + "grad_norm": 0.44140625, + "learning_rate": 0.0001960724093754138, + "loss": 2.6364, + "step": 4292 + }, + { + "epoch": 0.18, + "grad_norm": 0.443359375, + "learning_rate": 0.0001960705476724258, + "loss": 2.7831, + "step": 4293 + }, + { + "epoch": 0.18, + "grad_norm": 0.447265625, + "learning_rate": 0.00019606868553715545, + "loss": 2.6817, + "step": 4294 + }, + { + "epoch": 0.18, + "grad_norm": 0.388671875, + "learning_rate": 0.00019606682296961114, + "loss": 2.6016, + "step": 4295 + }, + { + "epoch": 0.18, + "grad_norm": 0.4375, + "learning_rate": 0.00019606495996980124, + "loss": 2.5834, + "step": 4296 + }, + { + "epoch": 0.18, + "grad_norm": 0.4609375, + "learning_rate": 0.00019606309653773414, + "loss": 2.6105, + "step": 4297 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.0001960612326734182, + "loss": 2.734, + "step": 4298 + }, + { + "epoch": 0.18, + "grad_norm": 0.359375, + "learning_rate": 0.0001960593683768618, + "loss": 2.6394, + "step": 4299 + }, + { + "epoch": 0.18, + "grad_norm": 0.466796875, + "learning_rate": 0.00019605750364807336, + "loss": 2.7046, + "step": 4300 + }, + { + "epoch": 0.18, + "grad_norm": 0.416015625, + "learning_rate": 0.00019605563848706126, + "loss": 2.7428, + "step": 4301 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.00019605377289383388, + "loss": 2.7764, + "step": 4302 + }, + { + "epoch": 0.18, + "grad_norm": 0.4140625, + "learning_rate": 0.00019605190686839964, + "loss": 2.6361, + "step": 4303 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.00019605004041076693, + "loss": 2.6805, + "step": 4304 + }, + { + "epoch": 0.18, + "grad_norm": 0.431640625, + "learning_rate": 0.0001960481735209441, + "loss": 2.5858, + "step": 4305 + }, + { + "epoch": 0.18, + "grad_norm": 0.4375, + "learning_rate": 0.00019604630619893964, + "loss": 2.6667, + "step": 4306 + }, + { + "epoch": 0.18, + "grad_norm": 0.443359375, + "learning_rate": 0.0001960444384447619, + "loss": 2.8473, + "step": 4307 + }, + { + "epoch": 0.18, + "grad_norm": 0.478515625, + "learning_rate": 0.00019604257025841927, + "loss": 2.7228, + "step": 4308 + }, + { + "epoch": 0.18, + "grad_norm": 0.404296875, + "learning_rate": 0.00019604070163992016, + "loss": 2.6743, + "step": 4309 + }, + { + "epoch": 0.18, + "grad_norm": 0.451171875, + "learning_rate": 0.000196038832589273, + "loss": 2.7419, + "step": 4310 + }, + { + "epoch": 0.18, + "grad_norm": 0.439453125, + "learning_rate": 0.0001960369631064862, + "loss": 2.5791, + "step": 4311 + }, + { + "epoch": 0.18, + "grad_norm": 0.451171875, + "learning_rate": 0.00019603509319156815, + "loss": 2.7874, + "step": 4312 + }, + { + "epoch": 0.18, + "grad_norm": 0.4375, + "learning_rate": 0.0001960332228445273, + "loss": 2.6456, + "step": 4313 + }, + { + "epoch": 0.18, + "grad_norm": 0.455078125, + "learning_rate": 0.00019603135206537205, + "loss": 2.6101, + "step": 4314 + }, + { + "epoch": 0.18, + "grad_norm": 0.40625, + "learning_rate": 0.00019602948085411078, + "loss": 2.6674, + "step": 4315 + }, + { + "epoch": 0.18, + "grad_norm": 0.38671875, + "learning_rate": 0.00019602760921075193, + "loss": 2.5193, + "step": 4316 + }, + { + "epoch": 0.18, + "grad_norm": 0.466796875, + "learning_rate": 0.00019602573713530398, + "loss": 2.7782, + "step": 4317 + }, + { + "epoch": 0.18, + "grad_norm": 0.453125, + "learning_rate": 0.00019602386462777528, + "loss": 2.6758, + "step": 4318 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.0001960219916881743, + "loss": 2.6108, + "step": 4319 + }, + { + "epoch": 0.18, + "grad_norm": 0.4453125, + "learning_rate": 0.0001960201183165094, + "loss": 2.5812, + "step": 4320 + }, + { + "epoch": 0.18, + "grad_norm": 0.486328125, + "learning_rate": 0.00019601824451278912, + "loss": 2.7745, + "step": 4321 + }, + { + "epoch": 0.18, + "grad_norm": 0.439453125, + "learning_rate": 0.0001960163702770218, + "loss": 2.7034, + "step": 4322 + }, + { + "epoch": 0.18, + "grad_norm": 0.484375, + "learning_rate": 0.0001960144956092159, + "loss": 2.6954, + "step": 4323 + }, + { + "epoch": 0.18, + "grad_norm": 0.435546875, + "learning_rate": 0.00019601262050937987, + "loss": 2.7617, + "step": 4324 + }, + { + "epoch": 0.18, + "grad_norm": 0.451171875, + "learning_rate": 0.00019601074497752215, + "loss": 2.7352, + "step": 4325 + }, + { + "epoch": 0.18, + "grad_norm": 0.38671875, + "learning_rate": 0.00019600886901365114, + "loss": 2.5211, + "step": 4326 + }, + { + "epoch": 0.18, + "grad_norm": 0.490234375, + "learning_rate": 0.0001960069926177753, + "loss": 2.8789, + "step": 4327 + }, + { + "epoch": 0.18, + "grad_norm": 0.455078125, + "learning_rate": 0.00019600511578990307, + "loss": 2.6909, + "step": 4328 + }, + { + "epoch": 0.18, + "grad_norm": 0.443359375, + "learning_rate": 0.00019600323853004295, + "loss": 2.6699, + "step": 4329 + }, + { + "epoch": 0.18, + "grad_norm": 0.4375, + "learning_rate": 0.00019600136083820332, + "loss": 2.5968, + "step": 4330 + }, + { + "epoch": 0.18, + "grad_norm": 0.47265625, + "learning_rate": 0.00019599948271439266, + "loss": 2.6275, + "step": 4331 + }, + { + "epoch": 0.18, + "grad_norm": 0.451171875, + "learning_rate": 0.00019599760415861937, + "loss": 2.7298, + "step": 4332 + }, + { + "epoch": 0.18, + "grad_norm": 0.470703125, + "learning_rate": 0.00019599572517089203, + "loss": 2.7327, + "step": 4333 + }, + { + "epoch": 0.18, + "grad_norm": 0.455078125, + "learning_rate": 0.00019599384575121894, + "loss": 2.5106, + "step": 4334 + }, + { + "epoch": 0.18, + "grad_norm": 0.427734375, + "learning_rate": 0.00019599196589960866, + "loss": 2.5638, + "step": 4335 + }, + { + "epoch": 0.18, + "grad_norm": 0.4609375, + "learning_rate": 0.00019599008561606957, + "loss": 2.6333, + "step": 4336 + }, + { + "epoch": 0.18, + "grad_norm": 0.4765625, + "learning_rate": 0.0001959882049006102, + "loss": 2.7021, + "step": 4337 + }, + { + "epoch": 0.18, + "grad_norm": 0.4296875, + "learning_rate": 0.000195986323753239, + "loss": 2.7278, + "step": 4338 + }, + { + "epoch": 0.18, + "grad_norm": 0.462890625, + "learning_rate": 0.00019598444217396442, + "loss": 2.6113, + "step": 4339 + }, + { + "epoch": 0.18, + "grad_norm": 0.416015625, + "learning_rate": 0.0001959825601627949, + "loss": 2.6564, + "step": 4340 + }, + { + "epoch": 0.18, + "grad_norm": 0.388671875, + "learning_rate": 0.00019598067771973895, + "loss": 2.5468, + "step": 4341 + }, + { + "epoch": 0.19, + "grad_norm": 0.455078125, + "learning_rate": 0.00019597879484480505, + "loss": 2.5826, + "step": 4342 + }, + { + "epoch": 0.19, + "grad_norm": 0.498046875, + "learning_rate": 0.00019597691153800165, + "loss": 2.6321, + "step": 4343 + }, + { + "epoch": 0.19, + "grad_norm": 0.453125, + "learning_rate": 0.0001959750277993372, + "loss": 2.604, + "step": 4344 + }, + { + "epoch": 0.19, + "grad_norm": 0.404296875, + "learning_rate": 0.0001959731436288202, + "loss": 2.5077, + "step": 4345 + }, + { + "epoch": 0.19, + "grad_norm": 0.4140625, + "learning_rate": 0.0001959712590264591, + "loss": 2.7033, + "step": 4346 + }, + { + "epoch": 0.19, + "grad_norm": 0.41015625, + "learning_rate": 0.00019596937399226247, + "loss": 2.6041, + "step": 4347 + }, + { + "epoch": 0.19, + "grad_norm": 0.408203125, + "learning_rate": 0.0001959674885262387, + "loss": 2.7505, + "step": 4348 + }, + { + "epoch": 0.19, + "grad_norm": 0.466796875, + "learning_rate": 0.0001959656026283963, + "loss": 2.7383, + "step": 4349 + }, + { + "epoch": 0.19, + "grad_norm": 0.3984375, + "learning_rate": 0.00019596371629874377, + "loss": 2.6644, + "step": 4350 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019596182953728958, + "loss": 2.5867, + "step": 4351 + }, + { + "epoch": 0.19, + "grad_norm": 0.455078125, + "learning_rate": 0.00019595994234404223, + "loss": 2.684, + "step": 4352 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.0001959580547190102, + "loss": 2.6258, + "step": 4353 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.00019595616666220203, + "loss": 2.7678, + "step": 4354 + }, + { + "epoch": 0.19, + "grad_norm": 0.451171875, + "learning_rate": 0.00019595427817362614, + "loss": 2.7138, + "step": 4355 + }, + { + "epoch": 0.19, + "grad_norm": 0.4296875, + "learning_rate": 0.00019595238925329108, + "loss": 2.6834, + "step": 4356 + }, + { + "epoch": 0.19, + "grad_norm": 0.400390625, + "learning_rate": 0.00019595049990120533, + "loss": 2.5722, + "step": 4357 + }, + { + "epoch": 0.19, + "grad_norm": 0.458984375, + "learning_rate": 0.0001959486101173774, + "loss": 2.6953, + "step": 4358 + }, + { + "epoch": 0.19, + "grad_norm": 0.462890625, + "learning_rate": 0.00019594671990181577, + "loss": 2.6533, + "step": 4359 + }, + { + "epoch": 0.19, + "grad_norm": 0.39453125, + "learning_rate": 0.00019594482925452898, + "loss": 2.6678, + "step": 4360 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.00019594293817552555, + "loss": 2.6748, + "step": 4361 + }, + { + "epoch": 0.19, + "grad_norm": 0.349609375, + "learning_rate": 0.00019594104666481393, + "loss": 2.7368, + "step": 4362 + }, + { + "epoch": 0.19, + "grad_norm": 0.42578125, + "learning_rate": 0.00019593915472240268, + "loss": 2.7757, + "step": 4363 + }, + { + "epoch": 0.19, + "grad_norm": 0.421875, + "learning_rate": 0.0001959372623483003, + "loss": 2.7558, + "step": 4364 + }, + { + "epoch": 0.19, + "grad_norm": 0.4140625, + "learning_rate": 0.00019593536954251524, + "loss": 2.7007, + "step": 4365 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.00019593347630505613, + "loss": 2.6787, + "step": 4366 + }, + { + "epoch": 0.19, + "grad_norm": 0.4453125, + "learning_rate": 0.0001959315826359314, + "loss": 2.7211, + "step": 4367 + }, + { + "epoch": 0.19, + "grad_norm": 0.466796875, + "learning_rate": 0.00019592968853514965, + "loss": 2.8207, + "step": 4368 + }, + { + "epoch": 0.19, + "grad_norm": 0.380859375, + "learning_rate": 0.00019592779400271935, + "loss": 2.7043, + "step": 4369 + }, + { + "epoch": 0.19, + "grad_norm": 0.373046875, + "learning_rate": 0.00019592589903864903, + "loss": 2.6211, + "step": 4370 + }, + { + "epoch": 0.19, + "grad_norm": 0.3828125, + "learning_rate": 0.00019592400364294723, + "loss": 2.6438, + "step": 4371 + }, + { + "epoch": 0.19, + "grad_norm": 0.404296875, + "learning_rate": 0.00019592210781562246, + "loss": 2.6324, + "step": 4372 + }, + { + "epoch": 0.19, + "grad_norm": 0.5, + "learning_rate": 0.00019592021155668326, + "loss": 2.7042, + "step": 4373 + }, + { + "epoch": 0.19, + "grad_norm": 0.380859375, + "learning_rate": 0.00019591831486613813, + "loss": 2.5845, + "step": 4374 + }, + { + "epoch": 0.19, + "grad_norm": 0.419921875, + "learning_rate": 0.0001959164177439957, + "loss": 2.6981, + "step": 4375 + }, + { + "epoch": 0.19, + "grad_norm": 0.484375, + "learning_rate": 0.0001959145201902644, + "loss": 2.8167, + "step": 4376 + }, + { + "epoch": 0.19, + "grad_norm": 0.482421875, + "learning_rate": 0.0001959126222049528, + "loss": 2.7274, + "step": 4377 + }, + { + "epoch": 0.19, + "grad_norm": 0.48828125, + "learning_rate": 0.0001959107237880695, + "loss": 2.6871, + "step": 4378 + }, + { + "epoch": 0.19, + "grad_norm": 0.439453125, + "learning_rate": 0.000195908824939623, + "loss": 2.7931, + "step": 4379 + }, + { + "epoch": 0.19, + "grad_norm": 0.44140625, + "learning_rate": 0.0001959069256596218, + "loss": 2.6404, + "step": 4380 + }, + { + "epoch": 0.19, + "grad_norm": 0.39453125, + "learning_rate": 0.0001959050259480745, + "loss": 2.697, + "step": 4381 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.00019590312580498962, + "loss": 2.799, + "step": 4382 + }, + { + "epoch": 0.19, + "grad_norm": 0.458984375, + "learning_rate": 0.0001959012252303757, + "loss": 2.569, + "step": 4383 + }, + { + "epoch": 0.19, + "grad_norm": 0.439453125, + "learning_rate": 0.00019589932422424137, + "loss": 2.7542, + "step": 4384 + }, + { + "epoch": 0.19, + "grad_norm": 0.447265625, + "learning_rate": 0.00019589742278659512, + "loss": 2.5112, + "step": 4385 + }, + { + "epoch": 0.19, + "grad_norm": 0.453125, + "learning_rate": 0.00019589552091744548, + "loss": 2.6077, + "step": 4386 + }, + { + "epoch": 0.19, + "grad_norm": 0.404296875, + "learning_rate": 0.00019589361861680106, + "loss": 2.7604, + "step": 4387 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.0001958917158846704, + "loss": 2.6776, + "step": 4388 + }, + { + "epoch": 0.19, + "grad_norm": 0.42578125, + "learning_rate": 0.00019588981272106203, + "loss": 2.7269, + "step": 4389 + }, + { + "epoch": 0.19, + "grad_norm": 0.42578125, + "learning_rate": 0.0001958879091259846, + "loss": 2.686, + "step": 4390 + }, + { + "epoch": 0.19, + "grad_norm": 0.470703125, + "learning_rate": 0.00019588600509944662, + "loss": 2.7497, + "step": 4391 + }, + { + "epoch": 0.19, + "grad_norm": 0.45703125, + "learning_rate": 0.0001958841006414566, + "loss": 2.687, + "step": 4392 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.0001958821957520232, + "loss": 2.6169, + "step": 4393 + }, + { + "epoch": 0.19, + "grad_norm": 0.408203125, + "learning_rate": 0.000195880290431155, + "loss": 2.602, + "step": 4394 + }, + { + "epoch": 0.19, + "grad_norm": 0.478515625, + "learning_rate": 0.0001958783846788605, + "loss": 2.7488, + "step": 4395 + }, + { + "epoch": 0.19, + "grad_norm": 0.45703125, + "learning_rate": 0.0001958764784951483, + "loss": 2.7229, + "step": 4396 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.000195874571880027, + "loss": 2.6306, + "step": 4397 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019587266483350518, + "loss": 2.5221, + "step": 4398 + }, + { + "epoch": 0.19, + "grad_norm": 0.51953125, + "learning_rate": 0.00019587075735559136, + "loss": 2.7108, + "step": 4399 + }, + { + "epoch": 0.19, + "grad_norm": 0.44140625, + "learning_rate": 0.00019586884944629423, + "loss": 2.7, + "step": 4400 + }, + { + "epoch": 0.19, + "grad_norm": 0.404296875, + "learning_rate": 0.00019586694110562225, + "loss": 2.5819, + "step": 4401 + }, + { + "epoch": 0.19, + "grad_norm": 0.42578125, + "learning_rate": 0.00019586503233358412, + "loss": 2.7001, + "step": 4402 + }, + { + "epoch": 0.19, + "grad_norm": 0.462890625, + "learning_rate": 0.00019586312313018834, + "loss": 2.7912, + "step": 4403 + }, + { + "epoch": 0.19, + "grad_norm": 0.439453125, + "learning_rate": 0.00019586121349544357, + "loss": 2.7285, + "step": 4404 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019585930342935838, + "loss": 2.6601, + "step": 4405 + }, + { + "epoch": 0.19, + "grad_norm": 0.41015625, + "learning_rate": 0.0001958573929319413, + "loss": 2.5753, + "step": 4406 + }, + { + "epoch": 0.19, + "grad_norm": 0.408203125, + "learning_rate": 0.00019585548200320104, + "loss": 2.5796, + "step": 4407 + }, + { + "epoch": 0.19, + "grad_norm": 0.45703125, + "learning_rate": 0.0001958535706431461, + "loss": 2.7106, + "step": 4408 + }, + { + "epoch": 0.19, + "grad_norm": 0.466796875, + "learning_rate": 0.00019585165885178515, + "loss": 2.7693, + "step": 4409 + }, + { + "epoch": 0.19, + "grad_norm": 0.365234375, + "learning_rate": 0.00019584974662912677, + "loss": 2.6754, + "step": 4410 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.00019584783397517957, + "loss": 2.6435, + "step": 4411 + }, + { + "epoch": 0.19, + "grad_norm": 0.41015625, + "learning_rate": 0.0001958459208899521, + "loss": 2.4901, + "step": 4412 + }, + { + "epoch": 0.19, + "grad_norm": 0.416015625, + "learning_rate": 0.00019584400737345306, + "loss": 2.6432, + "step": 4413 + }, + { + "epoch": 0.19, + "grad_norm": 0.46875, + "learning_rate": 0.000195842093425691, + "loss": 2.7468, + "step": 4414 + }, + { + "epoch": 0.19, + "grad_norm": 0.396484375, + "learning_rate": 0.00019584017904667454, + "loss": 2.666, + "step": 4415 + }, + { + "epoch": 0.19, + "grad_norm": 0.4453125, + "learning_rate": 0.00019583826423641228, + "loss": 2.7382, + "step": 4416 + }, + { + "epoch": 0.19, + "grad_norm": 0.416015625, + "learning_rate": 0.00019583634899491287, + "loss": 2.5299, + "step": 4417 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.00019583443332218492, + "loss": 2.7387, + "step": 4418 + }, + { + "epoch": 0.19, + "grad_norm": 0.390625, + "learning_rate": 0.00019583251721823705, + "loss": 2.6528, + "step": 4419 + }, + { + "epoch": 0.19, + "grad_norm": 0.4140625, + "learning_rate": 0.00019583060068307785, + "loss": 2.6347, + "step": 4420 + }, + { + "epoch": 0.19, + "grad_norm": 0.46484375, + "learning_rate": 0.000195828683716716, + "loss": 2.7265, + "step": 4421 + }, + { + "epoch": 0.19, + "grad_norm": 0.4140625, + "learning_rate": 0.00019582676631916008, + "loss": 2.5729, + "step": 4422 + }, + { + "epoch": 0.19, + "grad_norm": 0.427734375, + "learning_rate": 0.00019582484849041873, + "loss": 2.7075, + "step": 4423 + }, + { + "epoch": 0.19, + "grad_norm": 0.443359375, + "learning_rate": 0.00019582293023050058, + "loss": 2.7889, + "step": 4424 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.00019582101153941428, + "loss": 2.7567, + "step": 4425 + }, + { + "epoch": 0.19, + "grad_norm": 0.41015625, + "learning_rate": 0.00019581909241716842, + "loss": 2.6815, + "step": 4426 + }, + { + "epoch": 0.19, + "grad_norm": 0.51171875, + "learning_rate": 0.0001958171728637717, + "loss": 2.8332, + "step": 4427 + }, + { + "epoch": 0.19, + "grad_norm": 0.470703125, + "learning_rate": 0.00019581525287923267, + "loss": 2.7185, + "step": 4428 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.00019581333246356005, + "loss": 2.6146, + "step": 4429 + }, + { + "epoch": 0.19, + "grad_norm": 0.40234375, + "learning_rate": 0.00019581141161676246, + "loss": 2.5972, + "step": 4430 + }, + { + "epoch": 0.19, + "grad_norm": 0.3984375, + "learning_rate": 0.0001958094903388485, + "loss": 2.7083, + "step": 4431 + }, + { + "epoch": 0.19, + "grad_norm": 0.43359375, + "learning_rate": 0.00019580756862982686, + "loss": 2.7653, + "step": 4432 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019580564648970618, + "loss": 2.704, + "step": 4433 + }, + { + "epoch": 0.19, + "grad_norm": 0.39453125, + "learning_rate": 0.00019580372391849512, + "loss": 2.5702, + "step": 4434 + }, + { + "epoch": 0.19, + "grad_norm": 0.3828125, + "learning_rate": 0.00019580180091620231, + "loss": 2.6065, + "step": 4435 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019579987748283637, + "loss": 2.5708, + "step": 4436 + }, + { + "epoch": 0.19, + "grad_norm": 0.455078125, + "learning_rate": 0.00019579795361840603, + "loss": 2.6765, + "step": 4437 + }, + { + "epoch": 0.19, + "grad_norm": 0.427734375, + "learning_rate": 0.0001957960293229199, + "loss": 2.7171, + "step": 4438 + }, + { + "epoch": 0.19, + "grad_norm": 0.419921875, + "learning_rate": 0.00019579410459638665, + "loss": 2.6969, + "step": 4439 + }, + { + "epoch": 0.19, + "grad_norm": 0.400390625, + "learning_rate": 0.00019579217943881493, + "loss": 2.7192, + "step": 4440 + }, + { + "epoch": 0.19, + "grad_norm": 0.41796875, + "learning_rate": 0.0001957902538502134, + "loss": 2.6156, + "step": 4441 + }, + { + "epoch": 0.19, + "grad_norm": 0.451171875, + "learning_rate": 0.00019578832783059076, + "loss": 2.6933, + "step": 4442 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.00019578640137995562, + "loss": 2.6277, + "step": 4443 + }, + { + "epoch": 0.19, + "grad_norm": 0.3984375, + "learning_rate": 0.0001957844744983167, + "loss": 2.5393, + "step": 4444 + }, + { + "epoch": 0.19, + "grad_norm": 0.470703125, + "learning_rate": 0.00019578254718568267, + "loss": 2.7282, + "step": 4445 + }, + { + "epoch": 0.19, + "grad_norm": 0.41796875, + "learning_rate": 0.00019578061944206215, + "loss": 2.6993, + "step": 4446 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.00019577869126746387, + "loss": 2.8201, + "step": 4447 + }, + { + "epoch": 0.19, + "grad_norm": 0.455078125, + "learning_rate": 0.00019577676266189646, + "loss": 2.6634, + "step": 4448 + }, + { + "epoch": 0.19, + "grad_norm": 0.478515625, + "learning_rate": 0.00019577483362536864, + "loss": 2.715, + "step": 4449 + }, + { + "epoch": 0.19, + "grad_norm": 0.41796875, + "learning_rate": 0.00019577290415788906, + "loss": 2.5974, + "step": 4450 + }, + { + "epoch": 0.19, + "grad_norm": 0.388671875, + "learning_rate": 0.0001957709742594664, + "loss": 2.5759, + "step": 4451 + }, + { + "epoch": 0.19, + "grad_norm": 0.49609375, + "learning_rate": 0.00019576904393010938, + "loss": 2.73, + "step": 4452 + }, + { + "epoch": 0.19, + "grad_norm": 0.416015625, + "learning_rate": 0.00019576711316982666, + "loss": 2.7636, + "step": 4453 + }, + { + "epoch": 0.19, + "grad_norm": 0.447265625, + "learning_rate": 0.0001957651819786269, + "loss": 2.6762, + "step": 4454 + }, + { + "epoch": 0.19, + "grad_norm": 0.416015625, + "learning_rate": 0.00019576325035651886, + "loss": 2.5795, + "step": 4455 + }, + { + "epoch": 0.19, + "grad_norm": 0.369140625, + "learning_rate": 0.00019576131830351115, + "loss": 2.508, + "step": 4456 + }, + { + "epoch": 0.19, + "grad_norm": 0.404296875, + "learning_rate": 0.00019575938581961253, + "loss": 2.8543, + "step": 4457 + }, + { + "epoch": 0.19, + "grad_norm": 0.435546875, + "learning_rate": 0.00019575745290483168, + "loss": 2.7495, + "step": 4458 + }, + { + "epoch": 0.19, + "grad_norm": 0.37890625, + "learning_rate": 0.00019575551955917727, + "loss": 2.8836, + "step": 4459 + }, + { + "epoch": 0.19, + "grad_norm": 0.427734375, + "learning_rate": 0.000195753585782658, + "loss": 2.7195, + "step": 4460 + }, + { + "epoch": 0.19, + "grad_norm": 0.439453125, + "learning_rate": 0.0001957516515752826, + "loss": 2.6052, + "step": 4461 + }, + { + "epoch": 0.19, + "grad_norm": 0.451171875, + "learning_rate": 0.00019574971693705977, + "loss": 2.6866, + "step": 4462 + }, + { + "epoch": 0.19, + "grad_norm": 0.443359375, + "learning_rate": 0.0001957477818679982, + "loss": 2.696, + "step": 4463 + }, + { + "epoch": 0.19, + "grad_norm": 0.392578125, + "learning_rate": 0.0001957458463681066, + "loss": 2.5881, + "step": 4464 + }, + { + "epoch": 0.19, + "grad_norm": 0.49609375, + "learning_rate": 0.0001957439104373937, + "loss": 2.6258, + "step": 4465 + }, + { + "epoch": 0.19, + "grad_norm": 0.451171875, + "learning_rate": 0.00019574197407586818, + "loss": 2.6897, + "step": 4466 + }, + { + "epoch": 0.19, + "grad_norm": 0.443359375, + "learning_rate": 0.00019574003728353877, + "loss": 2.6829, + "step": 4467 + }, + { + "epoch": 0.19, + "grad_norm": 0.39453125, + "learning_rate": 0.0001957381000604142, + "loss": 2.6777, + "step": 4468 + }, + { + "epoch": 0.19, + "grad_norm": 0.451171875, + "learning_rate": 0.00019573616240650313, + "loss": 2.7385, + "step": 4469 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019573422432181436, + "loss": 2.7569, + "step": 4470 + }, + { + "epoch": 0.19, + "grad_norm": 0.423828125, + "learning_rate": 0.00019573228580635653, + "loss": 2.7911, + "step": 4471 + }, + { + "epoch": 0.19, + "grad_norm": 0.48046875, + "learning_rate": 0.00019573034686013843, + "loss": 2.7402, + "step": 4472 + }, + { + "epoch": 0.19, + "grad_norm": 0.447265625, + "learning_rate": 0.00019572840748316873, + "loss": 2.6563, + "step": 4473 + }, + { + "epoch": 0.19, + "grad_norm": 0.48828125, + "learning_rate": 0.00019572646767545619, + "loss": 2.8215, + "step": 4474 + }, + { + "epoch": 0.19, + "grad_norm": 0.37890625, + "learning_rate": 0.00019572452743700953, + "loss": 2.7427, + "step": 4475 + }, + { + "epoch": 0.19, + "grad_norm": 0.41015625, + "learning_rate": 0.00019572258676783748, + "loss": 2.7329, + "step": 4476 + }, + { + "epoch": 0.19, + "grad_norm": 0.4140625, + "learning_rate": 0.00019572064566794877, + "loss": 2.7789, + "step": 4477 + }, + { + "epoch": 0.19, + "grad_norm": 0.462890625, + "learning_rate": 0.00019571870413735213, + "loss": 2.6607, + "step": 4478 + }, + { + "epoch": 0.19, + "grad_norm": 0.4453125, + "learning_rate": 0.0001957167621760563, + "loss": 2.6669, + "step": 4479 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019571481978407004, + "loss": 2.6552, + "step": 4480 + }, + { + "epoch": 0.19, + "grad_norm": 0.396484375, + "learning_rate": 0.00019571287696140206, + "loss": 2.7009, + "step": 4481 + }, + { + "epoch": 0.19, + "grad_norm": 0.419921875, + "learning_rate": 0.0001957109337080611, + "loss": 2.6047, + "step": 4482 + }, + { + "epoch": 0.19, + "grad_norm": 0.435546875, + "learning_rate": 0.00019570899002405593, + "loss": 2.6433, + "step": 4483 + }, + { + "epoch": 0.19, + "grad_norm": 0.41015625, + "learning_rate": 0.0001957070459093953, + "loss": 2.527, + "step": 4484 + }, + { + "epoch": 0.19, + "grad_norm": 0.40625, + "learning_rate": 0.0001957051013640879, + "loss": 2.588, + "step": 4485 + }, + { + "epoch": 0.19, + "grad_norm": 0.44921875, + "learning_rate": 0.00019570315638814257, + "loss": 2.7123, + "step": 4486 + }, + { + "epoch": 0.19, + "grad_norm": 0.484375, + "learning_rate": 0.00019570121098156797, + "loss": 2.6866, + "step": 4487 + }, + { + "epoch": 0.19, + "grad_norm": 0.396484375, + "learning_rate": 0.00019569926514437293, + "loss": 2.7352, + "step": 4488 + }, + { + "epoch": 0.19, + "grad_norm": 0.4296875, + "learning_rate": 0.00019569731887656615, + "loss": 2.6336, + "step": 4489 + }, + { + "epoch": 0.19, + "grad_norm": 0.3984375, + "learning_rate": 0.0001956953721781564, + "loss": 2.6286, + "step": 4490 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.00019569342504915244, + "loss": 2.731, + "step": 4491 + }, + { + "epoch": 0.19, + "grad_norm": 0.404296875, + "learning_rate": 0.00019569147748956306, + "loss": 2.7326, + "step": 4492 + }, + { + "epoch": 0.19, + "grad_norm": 0.416015625, + "learning_rate": 0.000195689529499397, + "loss": 2.6354, + "step": 4493 + }, + { + "epoch": 0.19, + "grad_norm": 0.455078125, + "learning_rate": 0.00019568758107866305, + "loss": 2.6501, + "step": 4494 + }, + { + "epoch": 0.19, + "grad_norm": 0.451171875, + "learning_rate": 0.0001956856322273699, + "loss": 2.6585, + "step": 4495 + }, + { + "epoch": 0.19, + "grad_norm": 0.416015625, + "learning_rate": 0.0001956836829455264, + "loss": 2.6226, + "step": 4496 + }, + { + "epoch": 0.19, + "grad_norm": 0.546875, + "learning_rate": 0.00019568173323314132, + "loss": 2.679, + "step": 4497 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.00019567978309022343, + "loss": 2.5973, + "step": 4498 + }, + { + "epoch": 0.19, + "grad_norm": 0.40625, + "learning_rate": 0.0001956778325167814, + "loss": 2.7215, + "step": 4499 + }, + { + "epoch": 0.19, + "grad_norm": 0.39453125, + "learning_rate": 0.00019567588151282413, + "loss": 2.629, + "step": 4500 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.00019567393007836038, + "loss": 2.6649, + "step": 4501 + }, + { + "epoch": 0.19, + "grad_norm": 0.48828125, + "learning_rate": 0.0001956719782133989, + "loss": 2.7511, + "step": 4502 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.00019567002591794845, + "loss": 2.6188, + "step": 4503 + }, + { + "epoch": 0.19, + "grad_norm": 0.42578125, + "learning_rate": 0.00019566807319201787, + "loss": 2.6501, + "step": 4504 + }, + { + "epoch": 0.19, + "grad_norm": 0.451171875, + "learning_rate": 0.00019566612003561593, + "loss": 2.6517, + "step": 4505 + }, + { + "epoch": 0.19, + "grad_norm": 0.462890625, + "learning_rate": 0.00019566416644875142, + "loss": 2.702, + "step": 4506 + }, + { + "epoch": 0.19, + "grad_norm": 0.482421875, + "learning_rate": 0.00019566221243143308, + "loss": 2.7513, + "step": 4507 + }, + { + "epoch": 0.19, + "grad_norm": 0.453125, + "learning_rate": 0.00019566025798366978, + "loss": 2.6073, + "step": 4508 + }, + { + "epoch": 0.19, + "grad_norm": 0.44921875, + "learning_rate": 0.00019565830310547028, + "loss": 2.7083, + "step": 4509 + }, + { + "epoch": 0.19, + "grad_norm": 0.44921875, + "learning_rate": 0.00019565634779684335, + "loss": 2.6102, + "step": 4510 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019565439205779784, + "loss": 2.6434, + "step": 4511 + }, + { + "epoch": 0.19, + "grad_norm": 0.423828125, + "learning_rate": 0.00019565243588834253, + "loss": 2.52, + "step": 4512 + }, + { + "epoch": 0.19, + "grad_norm": 0.419921875, + "learning_rate": 0.00019565047928848616, + "loss": 2.6007, + "step": 4513 + }, + { + "epoch": 0.19, + "grad_norm": 0.439453125, + "learning_rate": 0.00019564852225823764, + "loss": 2.6366, + "step": 4514 + }, + { + "epoch": 0.19, + "grad_norm": 0.458984375, + "learning_rate": 0.00019564656479760571, + "loss": 2.7457, + "step": 4515 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.0001956446069065992, + "loss": 2.6404, + "step": 4516 + }, + { + "epoch": 0.19, + "grad_norm": 0.380859375, + "learning_rate": 0.00019564264858522692, + "loss": 2.6074, + "step": 4517 + }, + { + "epoch": 0.19, + "grad_norm": 0.421875, + "learning_rate": 0.00019564068983349766, + "loss": 2.6291, + "step": 4518 + }, + { + "epoch": 0.19, + "grad_norm": 0.40234375, + "learning_rate": 0.00019563873065142024, + "loss": 2.6632, + "step": 4519 + }, + { + "epoch": 0.19, + "grad_norm": 0.427734375, + "learning_rate": 0.0001956367710390035, + "loss": 2.6371, + "step": 4520 + }, + { + "epoch": 0.19, + "grad_norm": 0.404296875, + "learning_rate": 0.00019563481099625624, + "loss": 2.5696, + "step": 4521 + }, + { + "epoch": 0.19, + "grad_norm": 0.455078125, + "learning_rate": 0.00019563285052318728, + "loss": 2.6769, + "step": 4522 + }, + { + "epoch": 0.19, + "grad_norm": 0.44921875, + "learning_rate": 0.00019563088961980544, + "loss": 2.6695, + "step": 4523 + }, + { + "epoch": 0.19, + "grad_norm": 0.43359375, + "learning_rate": 0.00019562892828611954, + "loss": 2.5514, + "step": 4524 + }, + { + "epoch": 0.19, + "grad_norm": 0.44921875, + "learning_rate": 0.0001956269665221384, + "loss": 2.5899, + "step": 4525 + }, + { + "epoch": 0.19, + "grad_norm": 0.416015625, + "learning_rate": 0.0001956250043278709, + "loss": 2.7522, + "step": 4526 + }, + { + "epoch": 0.19, + "grad_norm": 0.408203125, + "learning_rate": 0.0001956230417033258, + "loss": 2.5485, + "step": 4527 + }, + { + "epoch": 0.19, + "grad_norm": 0.46875, + "learning_rate": 0.00019562107864851195, + "loss": 2.7369, + "step": 4528 + }, + { + "epoch": 0.19, + "grad_norm": 0.421875, + "learning_rate": 0.0001956191151634382, + "loss": 2.7166, + "step": 4529 + }, + { + "epoch": 0.19, + "grad_norm": 0.4375, + "learning_rate": 0.0001956171512481134, + "loss": 2.6095, + "step": 4530 + }, + { + "epoch": 0.19, + "grad_norm": 0.40625, + "learning_rate": 0.00019561518690254634, + "loss": 2.5736, + "step": 4531 + }, + { + "epoch": 0.19, + "grad_norm": 0.478515625, + "learning_rate": 0.0001956132221267459, + "loss": 2.6846, + "step": 4532 + }, + { + "epoch": 0.19, + "grad_norm": 0.408203125, + "learning_rate": 0.00019561125692072086, + "loss": 2.5889, + "step": 4533 + }, + { + "epoch": 0.19, + "grad_norm": 0.50390625, + "learning_rate": 0.00019560929128448017, + "loss": 2.7908, + "step": 4534 + }, + { + "epoch": 0.19, + "grad_norm": 0.4140625, + "learning_rate": 0.00019560732521803256, + "loss": 2.5698, + "step": 4535 + }, + { + "epoch": 0.19, + "grad_norm": 0.421875, + "learning_rate": 0.00019560535872138694, + "loss": 2.6722, + "step": 4536 + }, + { + "epoch": 0.19, + "grad_norm": 0.38671875, + "learning_rate": 0.00019560339179455213, + "loss": 2.6359, + "step": 4537 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.00019560142443753704, + "loss": 2.5995, + "step": 4538 + }, + { + "epoch": 0.19, + "grad_norm": 0.458984375, + "learning_rate": 0.00019559945665035042, + "loss": 2.7661, + "step": 4539 + }, + { + "epoch": 0.19, + "grad_norm": 0.431640625, + "learning_rate": 0.00019559748843300123, + "loss": 2.8401, + "step": 4540 + }, + { + "epoch": 0.19, + "grad_norm": 0.419921875, + "learning_rate": 0.00019559551978549827, + "loss": 2.5239, + "step": 4541 + }, + { + "epoch": 0.19, + "grad_norm": 0.44921875, + "learning_rate": 0.0001955935507078504, + "loss": 2.5598, + "step": 4542 + }, + { + "epoch": 0.19, + "grad_norm": 0.48046875, + "learning_rate": 0.00019559158120006649, + "loss": 2.7531, + "step": 4543 + }, + { + "epoch": 0.19, + "grad_norm": 0.47265625, + "learning_rate": 0.0001955896112621554, + "loss": 2.7892, + "step": 4544 + }, + { + "epoch": 0.19, + "grad_norm": 0.5, + "learning_rate": 0.000195587640894126, + "loss": 2.7655, + "step": 4545 + }, + { + "epoch": 0.19, + "grad_norm": 0.48828125, + "learning_rate": 0.00019558567009598713, + "loss": 2.8229, + "step": 4546 + }, + { + "epoch": 0.19, + "grad_norm": 0.42578125, + "learning_rate": 0.00019558369886774768, + "loss": 2.6675, + "step": 4547 + }, + { + "epoch": 0.19, + "grad_norm": 0.5078125, + "learning_rate": 0.00019558172720941653, + "loss": 2.7513, + "step": 4548 + }, + { + "epoch": 0.19, + "grad_norm": 0.47265625, + "learning_rate": 0.00019557975512100252, + "loss": 2.6265, + "step": 4549 + }, + { + "epoch": 0.19, + "grad_norm": 0.515625, + "learning_rate": 0.00019557778260251456, + "loss": 2.7455, + "step": 4550 + }, + { + "epoch": 0.19, + "grad_norm": 0.380859375, + "learning_rate": 0.0001955758096539615, + "loss": 2.8018, + "step": 4551 + }, + { + "epoch": 0.19, + "grad_norm": 0.484375, + "learning_rate": 0.0001955738362753522, + "loss": 2.6653, + "step": 4552 + }, + { + "epoch": 0.19, + "grad_norm": 0.435546875, + "learning_rate": 0.0001955718624666956, + "loss": 2.7056, + "step": 4553 + }, + { + "epoch": 0.19, + "grad_norm": 0.443359375, + "learning_rate": 0.00019556988822800055, + "loss": 2.6936, + "step": 4554 + }, + { + "epoch": 0.19, + "grad_norm": 0.3984375, + "learning_rate": 0.00019556791355927592, + "loss": 2.57, + "step": 4555 + }, + { + "epoch": 0.19, + "grad_norm": 0.48828125, + "learning_rate": 0.00019556593846053062, + "loss": 2.7094, + "step": 4556 + }, + { + "epoch": 0.19, + "grad_norm": 0.439453125, + "learning_rate": 0.00019556396293177351, + "loss": 2.6888, + "step": 4557 + }, + { + "epoch": 0.19, + "grad_norm": 0.474609375, + "learning_rate": 0.00019556198697301348, + "loss": 2.6521, + "step": 4558 + }, + { + "epoch": 0.19, + "grad_norm": 0.47265625, + "learning_rate": 0.00019556001058425944, + "loss": 2.5907, + "step": 4559 + }, + { + "epoch": 0.19, + "grad_norm": 0.4296875, + "learning_rate": 0.0001955580337655203, + "loss": 2.689, + "step": 4560 + }, + { + "epoch": 0.19, + "grad_norm": 0.490234375, + "learning_rate": 0.0001955560565168049, + "loss": 2.6653, + "step": 4561 + }, + { + "epoch": 0.19, + "grad_norm": 0.427734375, + "learning_rate": 0.00019555407883812217, + "loss": 2.6966, + "step": 4562 + }, + { + "epoch": 0.19, + "grad_norm": 0.443359375, + "learning_rate": 0.00019555210072948103, + "loss": 2.6322, + "step": 4563 + }, + { + "epoch": 0.19, + "grad_norm": 0.494140625, + "learning_rate": 0.00019555012219089034, + "loss": 2.6963, + "step": 4564 + }, + { + "epoch": 0.19, + "grad_norm": 0.466796875, + "learning_rate": 0.000195548143222359, + "loss": 2.5981, + "step": 4565 + }, + { + "epoch": 0.19, + "grad_norm": 0.3984375, + "learning_rate": 0.00019554616382389594, + "loss": 2.6317, + "step": 4566 + }, + { + "epoch": 0.19, + "grad_norm": 0.439453125, + "learning_rate": 0.00019554418399551007, + "loss": 2.7049, + "step": 4567 + }, + { + "epoch": 0.19, + "grad_norm": 0.46875, + "learning_rate": 0.00019554220373721028, + "loss": 2.5326, + "step": 4568 + }, + { + "epoch": 0.19, + "grad_norm": 0.39453125, + "learning_rate": 0.00019554022304900552, + "loss": 2.6657, + "step": 4569 + }, + { + "epoch": 0.19, + "grad_norm": 0.427734375, + "learning_rate": 0.00019553824193090466, + "loss": 2.5163, + "step": 4570 + }, + { + "epoch": 0.19, + "grad_norm": 0.5, + "learning_rate": 0.00019553626038291662, + "loss": 2.6784, + "step": 4571 + }, + { + "epoch": 0.19, + "grad_norm": 0.412109375, + "learning_rate": 0.00019553427840505032, + "loss": 2.5497, + "step": 4572 + }, + { + "epoch": 0.19, + "grad_norm": 0.400390625, + "learning_rate": 0.00019553229599731468, + "loss": 2.5586, + "step": 4573 + }, + { + "epoch": 0.19, + "grad_norm": 0.494140625, + "learning_rate": 0.00019553031315971864, + "loss": 2.7385, + "step": 4574 + }, + { + "epoch": 0.19, + "grad_norm": 0.421875, + "learning_rate": 0.00019552832989227107, + "loss": 2.6132, + "step": 4575 + }, + { + "epoch": 0.19, + "grad_norm": 0.48046875, + "learning_rate": 0.00019552634619498095, + "loss": 2.7273, + "step": 4576 + }, + { + "epoch": 0.2, + "grad_norm": 0.486328125, + "learning_rate": 0.00019552436206785718, + "loss": 2.6965, + "step": 4577 + }, + { + "epoch": 0.2, + "grad_norm": 0.435546875, + "learning_rate": 0.00019552237751090869, + "loss": 2.6012, + "step": 4578 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.0001955203925241444, + "loss": 2.5448, + "step": 4579 + }, + { + "epoch": 0.2, + "grad_norm": 0.455078125, + "learning_rate": 0.0001955184071075733, + "loss": 2.7223, + "step": 4580 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.0001955164212612042, + "loss": 2.7427, + "step": 4581 + }, + { + "epoch": 0.2, + "grad_norm": 0.439453125, + "learning_rate": 0.00019551443498504618, + "loss": 2.6732, + "step": 4582 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.00019551244827910807, + "loss": 2.7018, + "step": 4583 + }, + { + "epoch": 0.2, + "grad_norm": 0.458984375, + "learning_rate": 0.00019551046114339886, + "loss": 2.663, + "step": 4584 + }, + { + "epoch": 0.2, + "grad_norm": 0.423828125, + "learning_rate": 0.00019550847357792747, + "loss": 2.7627, + "step": 4585 + }, + { + "epoch": 0.2, + "grad_norm": 0.466796875, + "learning_rate": 0.00019550648558270287, + "loss": 2.6954, + "step": 4586 + }, + { + "epoch": 0.2, + "grad_norm": 0.39453125, + "learning_rate": 0.00019550449715773397, + "loss": 2.6995, + "step": 4587 + }, + { + "epoch": 0.2, + "grad_norm": 0.44921875, + "learning_rate": 0.00019550250830302975, + "loss": 2.6693, + "step": 4588 + }, + { + "epoch": 0.2, + "grad_norm": 0.431640625, + "learning_rate": 0.00019550051901859913, + "loss": 2.7328, + "step": 4589 + }, + { + "epoch": 0.2, + "grad_norm": 0.423828125, + "learning_rate": 0.0001954985293044511, + "loss": 2.5183, + "step": 4590 + }, + { + "epoch": 0.2, + "grad_norm": 0.431640625, + "learning_rate": 0.00019549653916059456, + "loss": 2.4946, + "step": 4591 + }, + { + "epoch": 0.2, + "grad_norm": 0.412109375, + "learning_rate": 0.00019549454858703853, + "loss": 2.5727, + "step": 4592 + }, + { + "epoch": 0.2, + "grad_norm": 0.4140625, + "learning_rate": 0.0001954925575837919, + "loss": 2.5837, + "step": 4593 + }, + { + "epoch": 0.2, + "grad_norm": 0.466796875, + "learning_rate": 0.00019549056615086364, + "loss": 2.6853, + "step": 4594 + }, + { + "epoch": 0.2, + "grad_norm": 0.482421875, + "learning_rate": 0.0001954885742882628, + "loss": 2.794, + "step": 4595 + }, + { + "epoch": 0.2, + "grad_norm": 0.435546875, + "learning_rate": 0.00019548658199599822, + "loss": 2.7406, + "step": 4596 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.00019548458927407891, + "loss": 2.704, + "step": 4597 + }, + { + "epoch": 0.2, + "grad_norm": 0.470703125, + "learning_rate": 0.00019548259612251387, + "loss": 2.5978, + "step": 4598 + }, + { + "epoch": 0.2, + "grad_norm": 0.478515625, + "learning_rate": 0.00019548060254131203, + "loss": 2.6366, + "step": 4599 + }, + { + "epoch": 0.2, + "grad_norm": 0.447265625, + "learning_rate": 0.0001954786085304824, + "loss": 2.5971, + "step": 4600 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.00019547661409003388, + "loss": 2.6732, + "step": 4601 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.00019547461921997554, + "loss": 2.5626, + "step": 4602 + }, + { + "epoch": 0.2, + "grad_norm": 0.412109375, + "learning_rate": 0.00019547262392031626, + "loss": 2.7524, + "step": 4603 + }, + { + "epoch": 0.2, + "grad_norm": 0.419921875, + "learning_rate": 0.00019547062819106508, + "loss": 2.6246, + "step": 4604 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.000195468632032231, + "loss": 2.6478, + "step": 4605 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.0001954666354438229, + "loss": 2.6312, + "step": 4606 + }, + { + "epoch": 0.2, + "grad_norm": 0.50390625, + "learning_rate": 0.00019546463842584988, + "loss": 2.7226, + "step": 4607 + }, + { + "epoch": 0.2, + "grad_norm": 0.45703125, + "learning_rate": 0.00019546264097832086, + "loss": 2.8185, + "step": 4608 + }, + { + "epoch": 0.2, + "grad_norm": 0.458984375, + "learning_rate": 0.0001954606431012448, + "loss": 2.6561, + "step": 4609 + }, + { + "epoch": 0.2, + "grad_norm": 0.52734375, + "learning_rate": 0.00019545864479463077, + "loss": 2.5595, + "step": 4610 + }, + { + "epoch": 0.2, + "grad_norm": 0.423828125, + "learning_rate": 0.00019545664605848773, + "loss": 2.7581, + "step": 4611 + }, + { + "epoch": 0.2, + "grad_norm": 0.470703125, + "learning_rate": 0.00019545464689282467, + "loss": 2.7857, + "step": 4612 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.0001954526472976506, + "loss": 2.6251, + "step": 4613 + }, + { + "epoch": 0.2, + "grad_norm": 0.392578125, + "learning_rate": 0.00019545064727297446, + "loss": 2.6555, + "step": 4614 + }, + { + "epoch": 0.2, + "grad_norm": 0.44140625, + "learning_rate": 0.0001954486468188053, + "loss": 2.6452, + "step": 4615 + }, + { + "epoch": 0.2, + "grad_norm": 0.44140625, + "learning_rate": 0.00019544664593515215, + "loss": 2.6351, + "step": 4616 + }, + { + "epoch": 0.2, + "grad_norm": 0.41015625, + "learning_rate": 0.00019544464462202394, + "loss": 2.6018, + "step": 4617 + }, + { + "epoch": 0.2, + "grad_norm": 0.396484375, + "learning_rate": 0.00019544264287942972, + "loss": 2.5642, + "step": 4618 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019544064070737847, + "loss": 2.5781, + "step": 4619 + }, + { + "epoch": 0.2, + "grad_norm": 0.451171875, + "learning_rate": 0.00019543863810587923, + "loss": 2.7785, + "step": 4620 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.00019543663507494102, + "loss": 2.6815, + "step": 4621 + }, + { + "epoch": 0.2, + "grad_norm": 0.421875, + "learning_rate": 0.00019543463161457282, + "loss": 2.6459, + "step": 4622 + }, + { + "epoch": 0.2, + "grad_norm": 0.421875, + "learning_rate": 0.00019543262772478363, + "loss": 2.632, + "step": 4623 + }, + { + "epoch": 0.2, + "grad_norm": 0.46875, + "learning_rate": 0.0001954306234055825, + "loss": 2.6554, + "step": 4624 + }, + { + "epoch": 0.2, + "grad_norm": 0.470703125, + "learning_rate": 0.00019542861865697844, + "loss": 2.7261, + "step": 4625 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.0001954266134789805, + "loss": 2.7005, + "step": 4626 + }, + { + "epoch": 0.2, + "grad_norm": 0.5, + "learning_rate": 0.00019542460787159765, + "loss": 2.7066, + "step": 4627 + }, + { + "epoch": 0.2, + "grad_norm": 0.369140625, + "learning_rate": 0.0001954226018348389, + "loss": 2.4788, + "step": 4628 + }, + { + "epoch": 0.2, + "grad_norm": 0.4375, + "learning_rate": 0.00019542059536871336, + "loss": 2.6835, + "step": 4629 + }, + { + "epoch": 0.2, + "grad_norm": 0.4375, + "learning_rate": 0.00019541858847323002, + "loss": 2.6762, + "step": 4630 + }, + { + "epoch": 0.2, + "grad_norm": 0.44921875, + "learning_rate": 0.0001954165811483979, + "loss": 2.5433, + "step": 4631 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.00019541457339422602, + "loss": 2.521, + "step": 4632 + }, + { + "epoch": 0.2, + "grad_norm": 0.46484375, + "learning_rate": 0.00019541256521072344, + "loss": 2.6852, + "step": 4633 + }, + { + "epoch": 0.2, + "grad_norm": 0.45703125, + "learning_rate": 0.00019541055659789917, + "loss": 2.6376, + "step": 4634 + }, + { + "epoch": 0.2, + "grad_norm": 0.458984375, + "learning_rate": 0.00019540854755576228, + "loss": 2.6766, + "step": 4635 + }, + { + "epoch": 0.2, + "grad_norm": 0.47265625, + "learning_rate": 0.0001954065380843218, + "loss": 2.7113, + "step": 4636 + }, + { + "epoch": 0.2, + "grad_norm": 0.466796875, + "learning_rate": 0.00019540452818358675, + "loss": 2.625, + "step": 4637 + }, + { + "epoch": 0.2, + "grad_norm": 0.490234375, + "learning_rate": 0.0001954025178535662, + "loss": 2.7796, + "step": 4638 + }, + { + "epoch": 0.2, + "grad_norm": 0.431640625, + "learning_rate": 0.00019540050709426917, + "loss": 2.5757, + "step": 4639 + }, + { + "epoch": 0.2, + "grad_norm": 0.42578125, + "learning_rate": 0.00019539849590570474, + "loss": 2.6521, + "step": 4640 + }, + { + "epoch": 0.2, + "grad_norm": 0.427734375, + "learning_rate": 0.00019539648428788193, + "loss": 2.7704, + "step": 4641 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.0001953944722408098, + "loss": 2.6239, + "step": 4642 + }, + { + "epoch": 0.2, + "grad_norm": 0.400390625, + "learning_rate": 0.00019539245976449743, + "loss": 2.65, + "step": 4643 + }, + { + "epoch": 0.2, + "grad_norm": 0.388671875, + "learning_rate": 0.00019539044685895385, + "loss": 2.7332, + "step": 4644 + }, + { + "epoch": 0.2, + "grad_norm": 0.4375, + "learning_rate": 0.0001953884335241881, + "loss": 2.5537, + "step": 4645 + }, + { + "epoch": 0.2, + "grad_norm": 0.4140625, + "learning_rate": 0.00019538641976020926, + "loss": 2.5993, + "step": 4646 + }, + { + "epoch": 0.2, + "grad_norm": 0.45703125, + "learning_rate": 0.00019538440556702639, + "loss": 2.6935, + "step": 4647 + }, + { + "epoch": 0.2, + "grad_norm": 0.431640625, + "learning_rate": 0.00019538239094464856, + "loss": 2.5717, + "step": 4648 + }, + { + "epoch": 0.2, + "grad_norm": 0.3984375, + "learning_rate": 0.00019538037589308485, + "loss": 2.6456, + "step": 4649 + }, + { + "epoch": 0.2, + "grad_norm": 0.431640625, + "learning_rate": 0.0001953783604123443, + "loss": 2.4501, + "step": 4650 + }, + { + "epoch": 0.2, + "grad_norm": 0.4765625, + "learning_rate": 0.00019537634450243597, + "loss": 2.6033, + "step": 4651 + }, + { + "epoch": 0.2, + "grad_norm": 0.435546875, + "learning_rate": 0.00019537432816336896, + "loss": 2.6693, + "step": 4652 + }, + { + "epoch": 0.2, + "grad_norm": 0.439453125, + "learning_rate": 0.0001953723113951523, + "loss": 2.7898, + "step": 4653 + }, + { + "epoch": 0.2, + "grad_norm": 0.466796875, + "learning_rate": 0.0001953702941977951, + "loss": 2.6234, + "step": 4654 + }, + { + "epoch": 0.2, + "grad_norm": 0.447265625, + "learning_rate": 0.00019536827657130648, + "loss": 2.59, + "step": 4655 + }, + { + "epoch": 0.2, + "grad_norm": 0.466796875, + "learning_rate": 0.0001953662585156954, + "loss": 2.5491, + "step": 4656 + }, + { + "epoch": 0.2, + "grad_norm": 0.44921875, + "learning_rate": 0.00019536424003097105, + "loss": 2.6768, + "step": 4657 + }, + { + "epoch": 0.2, + "grad_norm": 0.4765625, + "learning_rate": 0.00019536222111714248, + "loss": 2.5892, + "step": 4658 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.00019536020177421875, + "loss": 2.6041, + "step": 4659 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019535818200220893, + "loss": 2.7361, + "step": 4660 + }, + { + "epoch": 0.2, + "grad_norm": 0.439453125, + "learning_rate": 0.0001953561618011222, + "loss": 2.6954, + "step": 4661 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019535414117096753, + "loss": 2.6705, + "step": 4662 + }, + { + "epoch": 0.2, + "grad_norm": 0.443359375, + "learning_rate": 0.00019535212011175413, + "loss": 2.6613, + "step": 4663 + }, + { + "epoch": 0.2, + "grad_norm": 0.451171875, + "learning_rate": 0.000195350098623491, + "loss": 2.7468, + "step": 4664 + }, + { + "epoch": 0.2, + "grad_norm": 0.474609375, + "learning_rate": 0.00019534807670618727, + "loss": 2.7236, + "step": 4665 + }, + { + "epoch": 0.2, + "grad_norm": 0.41015625, + "learning_rate": 0.00019534605435985207, + "loss": 2.5124, + "step": 4666 + }, + { + "epoch": 0.2, + "grad_norm": 0.443359375, + "learning_rate": 0.00019534403158449445, + "loss": 2.7764, + "step": 4667 + }, + { + "epoch": 0.2, + "grad_norm": 0.478515625, + "learning_rate": 0.00019534200838012354, + "loss": 2.8181, + "step": 4668 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.0001953399847467484, + "loss": 2.6172, + "step": 4669 + }, + { + "epoch": 0.2, + "grad_norm": 0.51171875, + "learning_rate": 0.00019533796068437823, + "loss": 2.8633, + "step": 4670 + }, + { + "epoch": 0.2, + "grad_norm": 0.431640625, + "learning_rate": 0.00019533593619302204, + "loss": 2.5737, + "step": 4671 + }, + { + "epoch": 0.2, + "grad_norm": 0.3828125, + "learning_rate": 0.000195333911272689, + "loss": 2.6992, + "step": 4672 + }, + { + "epoch": 0.2, + "grad_norm": 0.427734375, + "learning_rate": 0.00019533188592338817, + "loss": 2.5284, + "step": 4673 + }, + { + "epoch": 0.2, + "grad_norm": 0.455078125, + "learning_rate": 0.00019532986014512873, + "loss": 2.6667, + "step": 4674 + }, + { + "epoch": 0.2, + "grad_norm": 0.392578125, + "learning_rate": 0.00019532783393791975, + "loss": 2.7251, + "step": 4675 + }, + { + "epoch": 0.2, + "grad_norm": 0.4140625, + "learning_rate": 0.00019532580730177035, + "loss": 2.6876, + "step": 4676 + }, + { + "epoch": 0.2, + "grad_norm": 0.484375, + "learning_rate": 0.00019532378023668962, + "loss": 2.6965, + "step": 4677 + }, + { + "epoch": 0.2, + "grad_norm": 0.453125, + "learning_rate": 0.00019532175274268677, + "loss": 2.6455, + "step": 4678 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.00019531972481977086, + "loss": 2.7072, + "step": 4679 + }, + { + "epoch": 0.2, + "grad_norm": 0.419921875, + "learning_rate": 0.000195317696467951, + "loss": 2.5864, + "step": 4680 + }, + { + "epoch": 0.2, + "grad_norm": 0.455078125, + "learning_rate": 0.00019531566768723636, + "loss": 2.6018, + "step": 4681 + }, + { + "epoch": 0.2, + "grad_norm": 0.4609375, + "learning_rate": 0.00019531363847763604, + "loss": 2.7163, + "step": 4682 + }, + { + "epoch": 0.2, + "grad_norm": 0.455078125, + "learning_rate": 0.00019531160883915914, + "loss": 2.6443, + "step": 4683 + }, + { + "epoch": 0.2, + "grad_norm": 0.44921875, + "learning_rate": 0.0001953095787718149, + "loss": 2.7011, + "step": 4684 + }, + { + "epoch": 0.2, + "grad_norm": 0.498046875, + "learning_rate": 0.00019530754827561236, + "loss": 2.7742, + "step": 4685 + }, + { + "epoch": 0.2, + "grad_norm": 0.484375, + "learning_rate": 0.0001953055173505607, + "loss": 2.5501, + "step": 4686 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.000195303485996669, + "loss": 2.6033, + "step": 4687 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019530145421394648, + "loss": 2.5673, + "step": 4688 + }, + { + "epoch": 0.2, + "grad_norm": 0.482421875, + "learning_rate": 0.00019529942200240222, + "loss": 2.5881, + "step": 4689 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.00019529738936204544, + "loss": 2.5778, + "step": 4690 + }, + { + "epoch": 0.2, + "grad_norm": 0.431640625, + "learning_rate": 0.00019529535629288515, + "loss": 2.6744, + "step": 4691 + }, + { + "epoch": 0.2, + "grad_norm": 0.466796875, + "learning_rate": 0.00019529332279493065, + "loss": 2.5878, + "step": 4692 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.000195291288868191, + "loss": 2.6767, + "step": 4693 + }, + { + "epoch": 0.2, + "grad_norm": 0.48046875, + "learning_rate": 0.0001952892545126754, + "loss": 2.8031, + "step": 4694 + }, + { + "epoch": 0.2, + "grad_norm": 0.47265625, + "learning_rate": 0.00019528721972839294, + "loss": 2.7656, + "step": 4695 + }, + { + "epoch": 0.2, + "grad_norm": 0.44140625, + "learning_rate": 0.00019528518451535285, + "loss": 2.6778, + "step": 4696 + }, + { + "epoch": 0.2, + "grad_norm": 0.41015625, + "learning_rate": 0.0001952831488735642, + "loss": 2.5563, + "step": 4697 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.00019528111280303626, + "loss": 2.6095, + "step": 4698 + }, + { + "epoch": 0.2, + "grad_norm": 0.365234375, + "learning_rate": 0.0001952790763037781, + "loss": 2.6589, + "step": 4699 + }, + { + "epoch": 0.2, + "grad_norm": 0.470703125, + "learning_rate": 0.00019527703937579892, + "loss": 2.7626, + "step": 4700 + }, + { + "epoch": 0.2, + "grad_norm": 0.3984375, + "learning_rate": 0.0001952750020191079, + "loss": 2.5979, + "step": 4701 + }, + { + "epoch": 0.2, + "grad_norm": 0.421875, + "learning_rate": 0.0001952729642337142, + "loss": 2.7044, + "step": 4702 + }, + { + "epoch": 0.2, + "grad_norm": 0.3984375, + "learning_rate": 0.00019527092601962697, + "loss": 2.7312, + "step": 4703 + }, + { + "epoch": 0.2, + "grad_norm": 0.421875, + "learning_rate": 0.00019526888737685533, + "loss": 2.6877, + "step": 4704 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.00019526684830540857, + "loss": 2.5812, + "step": 4705 + }, + { + "epoch": 0.2, + "grad_norm": 0.443359375, + "learning_rate": 0.00019526480880529583, + "loss": 2.7572, + "step": 4706 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.00019526276887652622, + "loss": 2.6827, + "step": 4707 + }, + { + "epoch": 0.2, + "grad_norm": 0.42578125, + "learning_rate": 0.000195260728519109, + "loss": 2.5838, + "step": 4708 + }, + { + "epoch": 0.2, + "grad_norm": 0.46484375, + "learning_rate": 0.00019525868773305328, + "loss": 2.6636, + "step": 4709 + }, + { + "epoch": 0.2, + "grad_norm": 0.4765625, + "learning_rate": 0.0001952566465183683, + "loss": 2.7525, + "step": 4710 + }, + { + "epoch": 0.2, + "grad_norm": 0.451171875, + "learning_rate": 0.00019525460487506322, + "loss": 2.6007, + "step": 4711 + }, + { + "epoch": 0.2, + "grad_norm": 0.46875, + "learning_rate": 0.00019525256280314722, + "loss": 2.6589, + "step": 4712 + }, + { + "epoch": 0.2, + "grad_norm": 0.49609375, + "learning_rate": 0.00019525052030262952, + "loss": 2.755, + "step": 4713 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.00019524847737351927, + "loss": 2.5928, + "step": 4714 + }, + { + "epoch": 0.2, + "grad_norm": 0.4140625, + "learning_rate": 0.0001952464340158257, + "loss": 2.4996, + "step": 4715 + }, + { + "epoch": 0.2, + "grad_norm": 0.46484375, + "learning_rate": 0.00019524439022955797, + "loss": 2.7304, + "step": 4716 + }, + { + "epoch": 0.2, + "grad_norm": 0.40625, + "learning_rate": 0.0001952423460147253, + "loss": 2.7094, + "step": 4717 + }, + { + "epoch": 0.2, + "grad_norm": 0.466796875, + "learning_rate": 0.00019524030137133686, + "loss": 2.6834, + "step": 4718 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.00019523825629940187, + "loss": 2.5745, + "step": 4719 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.00019523621079892955, + "loss": 2.7289, + "step": 4720 + }, + { + "epoch": 0.2, + "grad_norm": 0.4140625, + "learning_rate": 0.0001952341648699291, + "loss": 2.5267, + "step": 4721 + }, + { + "epoch": 0.2, + "grad_norm": 0.453125, + "learning_rate": 0.0001952321185124097, + "loss": 2.6891, + "step": 4722 + }, + { + "epoch": 0.2, + "grad_norm": 0.41015625, + "learning_rate": 0.00019523007172638054, + "loss": 2.6379, + "step": 4723 + }, + { + "epoch": 0.2, + "grad_norm": 0.4609375, + "learning_rate": 0.0001952280245118509, + "loss": 2.674, + "step": 4724 + }, + { + "epoch": 0.2, + "grad_norm": 0.396484375, + "learning_rate": 0.00019522597686882993, + "loss": 2.5774, + "step": 4725 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.00019522392879732687, + "loss": 2.6433, + "step": 4726 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019522188029735094, + "loss": 2.6901, + "step": 4727 + }, + { + "epoch": 0.2, + "grad_norm": 0.4921875, + "learning_rate": 0.00019521983136891132, + "loss": 2.5425, + "step": 4728 + }, + { + "epoch": 0.2, + "grad_norm": 0.404296875, + "learning_rate": 0.00019521778201201731, + "loss": 2.613, + "step": 4729 + }, + { + "epoch": 0.2, + "grad_norm": 0.431640625, + "learning_rate": 0.00019521573222667802, + "loss": 2.5569, + "step": 4730 + }, + { + "epoch": 0.2, + "grad_norm": 0.396484375, + "learning_rate": 0.00019521368201290277, + "loss": 2.6105, + "step": 4731 + }, + { + "epoch": 0.2, + "grad_norm": 0.3671875, + "learning_rate": 0.00019521163137070074, + "loss": 2.553, + "step": 4732 + }, + { + "epoch": 0.2, + "grad_norm": 0.404296875, + "learning_rate": 0.00019520958030008116, + "loss": 2.7795, + "step": 4733 + }, + { + "epoch": 0.2, + "grad_norm": 0.447265625, + "learning_rate": 0.00019520752880105327, + "loss": 2.6142, + "step": 4734 + }, + { + "epoch": 0.2, + "grad_norm": 0.4140625, + "learning_rate": 0.00019520547687362628, + "loss": 2.6272, + "step": 4735 + }, + { + "epoch": 0.2, + "grad_norm": 0.404296875, + "learning_rate": 0.00019520342451780946, + "loss": 2.6862, + "step": 4736 + }, + { + "epoch": 0.2, + "grad_norm": 0.4296875, + "learning_rate": 0.00019520137173361198, + "loss": 2.6245, + "step": 4737 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.00019519931852104319, + "loss": 2.6714, + "step": 4738 + }, + { + "epoch": 0.2, + "grad_norm": 0.40234375, + "learning_rate": 0.0001951972648801122, + "loss": 2.7091, + "step": 4739 + }, + { + "epoch": 0.2, + "grad_norm": 0.38671875, + "learning_rate": 0.00019519521081082832, + "loss": 2.5686, + "step": 4740 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.0001951931563132008, + "loss": 2.721, + "step": 4741 + }, + { + "epoch": 0.2, + "grad_norm": 0.474609375, + "learning_rate": 0.00019519110138723885, + "loss": 2.6984, + "step": 4742 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019518904603295175, + "loss": 2.5571, + "step": 4743 + }, + { + "epoch": 0.2, + "grad_norm": 0.427734375, + "learning_rate": 0.0001951869902503487, + "loss": 2.644, + "step": 4744 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.000195184934039439, + "loss": 2.6538, + "step": 4745 + }, + { + "epoch": 0.2, + "grad_norm": 0.41015625, + "learning_rate": 0.00019518287740023189, + "loss": 2.6122, + "step": 4746 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.0001951808203327366, + "loss": 2.6174, + "step": 4747 + }, + { + "epoch": 0.2, + "grad_norm": 0.41015625, + "learning_rate": 0.00019517876283696241, + "loss": 2.6125, + "step": 4748 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019517670491291857, + "loss": 2.6229, + "step": 4749 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.00019517464656061435, + "loss": 2.5993, + "step": 4750 + }, + { + "epoch": 0.2, + "grad_norm": 0.400390625, + "learning_rate": 0.00019517258778005897, + "loss": 2.7004, + "step": 4751 + }, + { + "epoch": 0.2, + "grad_norm": 0.39453125, + "learning_rate": 0.00019517052857126175, + "loss": 2.5678, + "step": 4752 + }, + { + "epoch": 0.2, + "grad_norm": 0.44921875, + "learning_rate": 0.00019516846893423195, + "loss": 2.6781, + "step": 4753 + }, + { + "epoch": 0.2, + "grad_norm": 0.408203125, + "learning_rate": 0.0001951664088689788, + "loss": 2.7011, + "step": 4754 + }, + { + "epoch": 0.2, + "grad_norm": 0.392578125, + "learning_rate": 0.0001951643483755116, + "loss": 2.539, + "step": 4755 + }, + { + "epoch": 0.2, + "grad_norm": 0.427734375, + "learning_rate": 0.0001951622874538396, + "loss": 2.6713, + "step": 4756 + }, + { + "epoch": 0.2, + "grad_norm": 0.4296875, + "learning_rate": 0.0001951602261039721, + "loss": 2.6928, + "step": 4757 + }, + { + "epoch": 0.2, + "grad_norm": 0.4609375, + "learning_rate": 0.00019515816432591834, + "loss": 2.6387, + "step": 4758 + }, + { + "epoch": 0.2, + "grad_norm": 0.40625, + "learning_rate": 0.0001951561021196876, + "loss": 2.64, + "step": 4759 + }, + { + "epoch": 0.2, + "grad_norm": 0.3984375, + "learning_rate": 0.00019515403948528923, + "loss": 2.6224, + "step": 4760 + }, + { + "epoch": 0.2, + "grad_norm": 0.4296875, + "learning_rate": 0.00019515197642273245, + "loss": 2.6418, + "step": 4761 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.00019514991293202648, + "loss": 2.6474, + "step": 4762 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.00019514784901318076, + "loss": 2.6599, + "step": 4763 + }, + { + "epoch": 0.2, + "grad_norm": 0.48046875, + "learning_rate": 0.00019514578466620446, + "loss": 2.808, + "step": 4764 + }, + { + "epoch": 0.2, + "grad_norm": 0.4609375, + "learning_rate": 0.0001951437198911069, + "loss": 2.5951, + "step": 4765 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.00019514165468789734, + "loss": 2.6451, + "step": 4766 + }, + { + "epoch": 0.2, + "grad_norm": 0.427734375, + "learning_rate": 0.00019513958905658513, + "loss": 2.5311, + "step": 4767 + }, + { + "epoch": 0.2, + "grad_norm": 0.4765625, + "learning_rate": 0.00019513752299717957, + "loss": 2.5962, + "step": 4768 + }, + { + "epoch": 0.2, + "grad_norm": 0.4921875, + "learning_rate": 0.00019513545650968987, + "loss": 2.7778, + "step": 4769 + }, + { + "epoch": 0.2, + "grad_norm": 0.458984375, + "learning_rate": 0.0001951333895941254, + "loss": 2.5933, + "step": 4770 + }, + { + "epoch": 0.2, + "grad_norm": 0.421875, + "learning_rate": 0.0001951313222504955, + "loss": 2.6884, + "step": 4771 + }, + { + "epoch": 0.2, + "grad_norm": 0.42578125, + "learning_rate": 0.00019512925447880933, + "loss": 2.6491, + "step": 4772 + }, + { + "epoch": 0.2, + "grad_norm": 0.47265625, + "learning_rate": 0.00019512718627907633, + "loss": 2.7559, + "step": 4773 + }, + { + "epoch": 0.2, + "grad_norm": 0.44921875, + "learning_rate": 0.00019512511765130576, + "loss": 2.6997, + "step": 4774 + }, + { + "epoch": 0.2, + "grad_norm": 0.44140625, + "learning_rate": 0.00019512304859550689, + "loss": 2.8015, + "step": 4775 + }, + { + "epoch": 0.2, + "grad_norm": 0.44140625, + "learning_rate": 0.00019512097911168912, + "loss": 2.7057, + "step": 4776 + }, + { + "epoch": 0.2, + "grad_norm": 0.486328125, + "learning_rate": 0.00019511890919986168, + "loss": 2.6763, + "step": 4777 + }, + { + "epoch": 0.2, + "grad_norm": 0.4375, + "learning_rate": 0.00019511683886003388, + "loss": 2.5873, + "step": 4778 + }, + { + "epoch": 0.2, + "grad_norm": 0.421875, + "learning_rate": 0.0001951147680922151, + "loss": 2.6759, + "step": 4779 + }, + { + "epoch": 0.2, + "grad_norm": 0.427734375, + "learning_rate": 0.0001951126968964146, + "loss": 2.6281, + "step": 4780 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.00019511062527264174, + "loss": 2.6472, + "step": 4781 + }, + { + "epoch": 0.2, + "grad_norm": 0.46484375, + "learning_rate": 0.00019510855322090582, + "loss": 2.5777, + "step": 4782 + }, + { + "epoch": 0.2, + "grad_norm": 0.46484375, + "learning_rate": 0.0001951064807412162, + "loss": 2.7911, + "step": 4783 + }, + { + "epoch": 0.2, + "grad_norm": 0.4296875, + "learning_rate": 0.00019510440783358213, + "loss": 2.6856, + "step": 4784 + }, + { + "epoch": 0.2, + "grad_norm": 0.43359375, + "learning_rate": 0.000195102334498013, + "loss": 2.7182, + "step": 4785 + }, + { + "epoch": 0.2, + "grad_norm": 0.458984375, + "learning_rate": 0.00019510026073451812, + "loss": 2.6705, + "step": 4786 + }, + { + "epoch": 0.2, + "grad_norm": 0.390625, + "learning_rate": 0.00019509818654310683, + "loss": 2.5242, + "step": 4787 + }, + { + "epoch": 0.2, + "grad_norm": 0.443359375, + "learning_rate": 0.00019509611192378842, + "loss": 2.6591, + "step": 4788 + }, + { + "epoch": 0.2, + "grad_norm": 0.412109375, + "learning_rate": 0.0001950940368765723, + "loss": 2.6777, + "step": 4789 + }, + { + "epoch": 0.2, + "grad_norm": 0.373046875, + "learning_rate": 0.00019509196140146776, + "loss": 2.622, + "step": 4790 + }, + { + "epoch": 0.2, + "grad_norm": 0.45703125, + "learning_rate": 0.00019508988549848412, + "loss": 2.6737, + "step": 4791 + }, + { + "epoch": 0.2, + "grad_norm": 0.435546875, + "learning_rate": 0.00019508780916763078, + "loss": 2.5563, + "step": 4792 + }, + { + "epoch": 0.2, + "grad_norm": 0.390625, + "learning_rate": 0.00019508573240891702, + "loss": 2.6465, + "step": 4793 + }, + { + "epoch": 0.2, + "grad_norm": 0.46484375, + "learning_rate": 0.00019508365522235223, + "loss": 2.6618, + "step": 4794 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019508157760794574, + "loss": 2.6927, + "step": 4795 + }, + { + "epoch": 0.2, + "grad_norm": 0.416015625, + "learning_rate": 0.0001950794995657069, + "loss": 2.6253, + "step": 4796 + }, + { + "epoch": 0.2, + "grad_norm": 0.478515625, + "learning_rate": 0.00019507742109564505, + "loss": 2.8874, + "step": 4797 + }, + { + "epoch": 0.2, + "grad_norm": 0.455078125, + "learning_rate": 0.00019507534219776956, + "loss": 2.8059, + "step": 4798 + }, + { + "epoch": 0.2, + "grad_norm": 0.4140625, + "learning_rate": 0.0001950732628720898, + "loss": 2.5635, + "step": 4799 + }, + { + "epoch": 0.2, + "grad_norm": 0.4296875, + "learning_rate": 0.00019507118311861508, + "loss": 2.7517, + "step": 4800 + }, + { + "epoch": 0.2, + "grad_norm": 0.41796875, + "learning_rate": 0.00019506910293735477, + "loss": 2.7121, + "step": 4801 + }, + { + "epoch": 0.2, + "grad_norm": 0.462890625, + "learning_rate": 0.00019506702232831823, + "loss": 2.6406, + "step": 4802 + }, + { + "epoch": 0.2, + "grad_norm": 0.4140625, + "learning_rate": 0.00019506494129151487, + "loss": 2.509, + "step": 4803 + }, + { + "epoch": 0.2, + "grad_norm": 0.44140625, + "learning_rate": 0.000195062859826954, + "loss": 2.7274, + "step": 4804 + }, + { + "epoch": 0.2, + "grad_norm": 0.439453125, + "learning_rate": 0.000195060777934645, + "loss": 2.6579, + "step": 4805 + }, + { + "epoch": 0.2, + "grad_norm": 0.421875, + "learning_rate": 0.00019505869561459724, + "loss": 2.5432, + "step": 4806 + }, + { + "epoch": 0.2, + "grad_norm": 0.419921875, + "learning_rate": 0.00019505661286682008, + "loss": 2.565, + "step": 4807 + }, + { + "epoch": 0.2, + "grad_norm": 0.4453125, + "learning_rate": 0.00019505452969132293, + "loss": 2.7448, + "step": 4808 + }, + { + "epoch": 0.2, + "grad_norm": 0.462890625, + "learning_rate": 0.0001950524460881151, + "loss": 2.6611, + "step": 4809 + }, + { + "epoch": 0.2, + "grad_norm": 0.392578125, + "learning_rate": 0.00019505036205720605, + "loss": 2.697, + "step": 4810 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.00019504827759860504, + "loss": 2.5876, + "step": 4811 + }, + { + "epoch": 0.21, + "grad_norm": 0.4375, + "learning_rate": 0.00019504619271232157, + "loss": 2.6562, + "step": 4812 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.00019504410739836495, + "loss": 2.7361, + "step": 4813 + }, + { + "epoch": 0.21, + "grad_norm": 0.416015625, + "learning_rate": 0.00019504202165674457, + "loss": 2.7172, + "step": 4814 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.00019503993548746984, + "loss": 2.6022, + "step": 4815 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.00019503784889055012, + "loss": 2.6986, + "step": 4816 + }, + { + "epoch": 0.21, + "grad_norm": 0.427734375, + "learning_rate": 0.0001950357618659948, + "loss": 2.5795, + "step": 4817 + }, + { + "epoch": 0.21, + "grad_norm": 0.41796875, + "learning_rate": 0.00019503367441381334, + "loss": 2.6045, + "step": 4818 + }, + { + "epoch": 0.21, + "grad_norm": 0.458984375, + "learning_rate": 0.00019503158653401507, + "loss": 2.8335, + "step": 4819 + }, + { + "epoch": 0.21, + "grad_norm": 0.4375, + "learning_rate": 0.00019502949822660935, + "loss": 2.6664, + "step": 4820 + }, + { + "epoch": 0.21, + "grad_norm": 0.474609375, + "learning_rate": 0.00019502740949160562, + "loss": 2.6507, + "step": 4821 + }, + { + "epoch": 0.21, + "grad_norm": 0.46484375, + "learning_rate": 0.00019502532032901327, + "loss": 2.6071, + "step": 4822 + }, + { + "epoch": 0.21, + "grad_norm": 0.4453125, + "learning_rate": 0.00019502323073884172, + "loss": 2.718, + "step": 4823 + }, + { + "epoch": 0.21, + "grad_norm": 0.427734375, + "learning_rate": 0.00019502114072110036, + "loss": 2.8323, + "step": 4824 + }, + { + "epoch": 0.21, + "grad_norm": 0.462890625, + "learning_rate": 0.0001950190502757986, + "loss": 2.7957, + "step": 4825 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.00019501695940294585, + "loss": 2.587, + "step": 4826 + }, + { + "epoch": 0.21, + "grad_norm": 0.431640625, + "learning_rate": 0.00019501486810255149, + "loss": 2.6121, + "step": 4827 + }, + { + "epoch": 0.21, + "grad_norm": 0.4140625, + "learning_rate": 0.00019501277637462493, + "loss": 2.7355, + "step": 4828 + }, + { + "epoch": 0.21, + "grad_norm": 0.478515625, + "learning_rate": 0.00019501068421917564, + "loss": 2.8273, + "step": 4829 + }, + { + "epoch": 0.21, + "grad_norm": 0.4140625, + "learning_rate": 0.00019500859163621294, + "loss": 2.6759, + "step": 4830 + }, + { + "epoch": 0.21, + "grad_norm": 0.431640625, + "learning_rate": 0.00019500649862574636, + "loss": 2.6742, + "step": 4831 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.0001950044051877852, + "loss": 2.5974, + "step": 4832 + }, + { + "epoch": 0.21, + "grad_norm": 0.443359375, + "learning_rate": 0.00019500231132233898, + "loss": 2.6354, + "step": 4833 + }, + { + "epoch": 0.21, + "grad_norm": 0.4453125, + "learning_rate": 0.00019500021702941706, + "loss": 2.6368, + "step": 4834 + }, + { + "epoch": 0.21, + "grad_norm": 0.447265625, + "learning_rate": 0.00019499812230902886, + "loss": 2.6362, + "step": 4835 + }, + { + "epoch": 0.21, + "grad_norm": 0.458984375, + "learning_rate": 0.00019499602716118388, + "loss": 2.6998, + "step": 4836 + }, + { + "epoch": 0.21, + "grad_norm": 0.48828125, + "learning_rate": 0.00019499393158589148, + "loss": 2.6328, + "step": 4837 + }, + { + "epoch": 0.21, + "grad_norm": 0.38671875, + "learning_rate": 0.00019499183558316108, + "loss": 2.542, + "step": 4838 + }, + { + "epoch": 0.21, + "grad_norm": 0.4375, + "learning_rate": 0.00019498973915300214, + "loss": 2.6664, + "step": 4839 + }, + { + "epoch": 0.21, + "grad_norm": 0.412109375, + "learning_rate": 0.0001949876422954241, + "loss": 2.5647, + "step": 4840 + }, + { + "epoch": 0.21, + "grad_norm": 0.478515625, + "learning_rate": 0.0001949855450104364, + "loss": 2.6425, + "step": 4841 + }, + { + "epoch": 0.21, + "grad_norm": 0.416015625, + "learning_rate": 0.00019498344729804843, + "loss": 2.7175, + "step": 4842 + }, + { + "epoch": 0.21, + "grad_norm": 0.400390625, + "learning_rate": 0.00019498134915826968, + "loss": 2.6207, + "step": 4843 + }, + { + "epoch": 0.21, + "grad_norm": 0.421875, + "learning_rate": 0.00019497925059110954, + "loss": 2.7626, + "step": 4844 + }, + { + "epoch": 0.21, + "grad_norm": 0.50390625, + "learning_rate": 0.0001949771515965775, + "loss": 2.7348, + "step": 4845 + }, + { + "epoch": 0.21, + "grad_norm": 0.431640625, + "learning_rate": 0.00019497505217468299, + "loss": 2.6012, + "step": 4846 + }, + { + "epoch": 0.21, + "grad_norm": 0.46875, + "learning_rate": 0.00019497295232543546, + "loss": 2.6899, + "step": 4847 + }, + { + "epoch": 0.21, + "grad_norm": 0.44921875, + "learning_rate": 0.00019497085204884436, + "loss": 2.682, + "step": 4848 + }, + { + "epoch": 0.21, + "grad_norm": 0.396484375, + "learning_rate": 0.00019496875134491912, + "loss": 2.6901, + "step": 4849 + }, + { + "epoch": 0.21, + "grad_norm": 0.44140625, + "learning_rate": 0.0001949666502136692, + "loss": 2.6244, + "step": 4850 + }, + { + "epoch": 0.21, + "grad_norm": 0.4921875, + "learning_rate": 0.00019496454865510407, + "loss": 2.7886, + "step": 4851 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.00019496244666923317, + "loss": 2.6938, + "step": 4852 + }, + { + "epoch": 0.21, + "grad_norm": 0.423828125, + "learning_rate": 0.00019496034425606597, + "loss": 2.702, + "step": 4853 + }, + { + "epoch": 0.21, + "grad_norm": 0.408203125, + "learning_rate": 0.0001949582414156119, + "loss": 2.5718, + "step": 4854 + }, + { + "epoch": 0.21, + "grad_norm": 0.400390625, + "learning_rate": 0.0001949561381478805, + "loss": 2.6077, + "step": 4855 + }, + { + "epoch": 0.21, + "grad_norm": 0.40625, + "learning_rate": 0.00019495403445288117, + "loss": 2.7026, + "step": 4856 + }, + { + "epoch": 0.21, + "grad_norm": 0.462890625, + "learning_rate": 0.00019495193033062337, + "loss": 2.5711, + "step": 4857 + }, + { + "epoch": 0.21, + "grad_norm": 0.43359375, + "learning_rate": 0.00019494982578111656, + "loss": 2.6491, + "step": 4858 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.00019494772080437027, + "loss": 2.5905, + "step": 4859 + }, + { + "epoch": 0.21, + "grad_norm": 0.41015625, + "learning_rate": 0.00019494561540039392, + "loss": 2.6348, + "step": 4860 + }, + { + "epoch": 0.21, + "grad_norm": 0.41015625, + "learning_rate": 0.000194943509569197, + "loss": 2.6836, + "step": 4861 + }, + { + "epoch": 0.21, + "grad_norm": 0.44921875, + "learning_rate": 0.00019494140331078897, + "loss": 2.6924, + "step": 4862 + }, + { + "epoch": 0.21, + "grad_norm": 0.404296875, + "learning_rate": 0.00019493929662517937, + "loss": 2.5982, + "step": 4863 + }, + { + "epoch": 0.21, + "grad_norm": 0.4140625, + "learning_rate": 0.00019493718951237759, + "loss": 2.7008, + "step": 4864 + }, + { + "epoch": 0.21, + "grad_norm": 0.40234375, + "learning_rate": 0.00019493508197239318, + "loss": 2.6931, + "step": 4865 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.00019493297400523558, + "loss": 2.6235, + "step": 4866 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.0001949308656109143, + "loss": 2.6411, + "step": 4867 + }, + { + "epoch": 0.21, + "grad_norm": 0.392578125, + "learning_rate": 0.00019492875678943882, + "loss": 2.5478, + "step": 4868 + }, + { + "epoch": 0.21, + "grad_norm": 0.427734375, + "learning_rate": 0.0001949266475408186, + "loss": 2.6416, + "step": 4869 + }, + { + "epoch": 0.21, + "grad_norm": 0.423828125, + "learning_rate": 0.00019492453786506316, + "loss": 2.5453, + "step": 4870 + }, + { + "epoch": 0.21, + "grad_norm": 0.4453125, + "learning_rate": 0.000194922427762182, + "loss": 2.7165, + "step": 4871 + }, + { + "epoch": 0.21, + "grad_norm": 0.51953125, + "learning_rate": 0.00019492031723218462, + "loss": 2.8159, + "step": 4872 + }, + { + "epoch": 0.21, + "grad_norm": 0.447265625, + "learning_rate": 0.00019491820627508052, + "loss": 2.6337, + "step": 4873 + }, + { + "epoch": 0.21, + "grad_norm": 0.41796875, + "learning_rate": 0.00019491609489087912, + "loss": 2.5871, + "step": 4874 + }, + { + "epoch": 0.21, + "grad_norm": 0.462890625, + "learning_rate": 0.00019491398307959005, + "loss": 2.4899, + "step": 4875 + }, + { + "epoch": 0.21, + "grad_norm": 0.431640625, + "learning_rate": 0.00019491187084122268, + "loss": 2.6088, + "step": 4876 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.0001949097581757866, + "loss": 2.6153, + "step": 4877 + }, + { + "epoch": 0.21, + "grad_norm": 0.447265625, + "learning_rate": 0.0001949076450832913, + "loss": 2.554, + "step": 4878 + }, + { + "epoch": 0.21, + "grad_norm": 0.41796875, + "learning_rate": 0.0001949055315637463, + "loss": 2.6478, + "step": 4879 + }, + { + "epoch": 0.21, + "grad_norm": 0.466796875, + "learning_rate": 0.00019490341761716105, + "loss": 2.5779, + "step": 4880 + }, + { + "epoch": 0.21, + "grad_norm": 0.423828125, + "learning_rate": 0.0001949013032435451, + "loss": 2.4938, + "step": 4881 + }, + { + "epoch": 0.21, + "grad_norm": 0.4296875, + "learning_rate": 0.000194899188442908, + "loss": 2.7126, + "step": 4882 + }, + { + "epoch": 0.21, + "grad_norm": 0.455078125, + "learning_rate": 0.00019489707321525924, + "loss": 2.6717, + "step": 4883 + }, + { + "epoch": 0.21, + "grad_norm": 0.419921875, + "learning_rate": 0.00019489495756060831, + "loss": 2.7132, + "step": 4884 + }, + { + "epoch": 0.21, + "grad_norm": 0.4453125, + "learning_rate": 0.00019489284147896477, + "loss": 2.6764, + "step": 4885 + }, + { + "epoch": 0.21, + "grad_norm": 0.470703125, + "learning_rate": 0.00019489072497033813, + "loss": 2.6525, + "step": 4886 + }, + { + "epoch": 0.21, + "grad_norm": 0.376953125, + "learning_rate": 0.0001948886080347379, + "loss": 2.6417, + "step": 4887 + }, + { + "epoch": 0.21, + "grad_norm": 0.421875, + "learning_rate": 0.0001948864906721736, + "loss": 2.6878, + "step": 4888 + }, + { + "epoch": 0.21, + "grad_norm": 0.44140625, + "learning_rate": 0.00019488437288265475, + "loss": 2.6349, + "step": 4889 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.00019488225466619092, + "loss": 2.6006, + "step": 4890 + }, + { + "epoch": 0.21, + "grad_norm": 0.466796875, + "learning_rate": 0.0001948801360227916, + "loss": 2.7394, + "step": 4891 + }, + { + "epoch": 0.21, + "grad_norm": 0.359375, + "learning_rate": 0.0001948780169524664, + "loss": 2.4905, + "step": 4892 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.00019487589745522473, + "loss": 2.6535, + "step": 4893 + }, + { + "epoch": 0.21, + "grad_norm": 0.447265625, + "learning_rate": 0.00019487377753107623, + "loss": 2.6246, + "step": 4894 + }, + { + "epoch": 0.21, + "grad_norm": 0.45703125, + "learning_rate": 0.0001948716571800304, + "loss": 2.7161, + "step": 4895 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.00019486953640209678, + "loss": 2.6212, + "step": 4896 + }, + { + "epoch": 0.21, + "grad_norm": 0.4296875, + "learning_rate": 0.0001948674151972849, + "loss": 2.6705, + "step": 4897 + }, + { + "epoch": 0.21, + "grad_norm": 0.361328125, + "learning_rate": 0.00019486529356560435, + "loss": 2.6981, + "step": 4898 + }, + { + "epoch": 0.21, + "grad_norm": 0.392578125, + "learning_rate": 0.00019486317150706463, + "loss": 2.6626, + "step": 4899 + }, + { + "epoch": 0.21, + "grad_norm": 0.427734375, + "learning_rate": 0.0001948610490216753, + "loss": 2.6881, + "step": 4900 + }, + { + "epoch": 0.21, + "grad_norm": 0.458984375, + "learning_rate": 0.00019485892610944595, + "loss": 2.721, + "step": 4901 + }, + { + "epoch": 0.21, + "grad_norm": 0.412109375, + "learning_rate": 0.0001948568027703861, + "loss": 2.5824, + "step": 4902 + }, + { + "epoch": 0.21, + "grad_norm": 0.48046875, + "learning_rate": 0.00019485467900450527, + "loss": 2.7984, + "step": 4903 + }, + { + "epoch": 0.21, + "grad_norm": 0.45703125, + "learning_rate": 0.00019485255481181305, + "loss": 2.9057, + "step": 4904 + }, + { + "epoch": 0.21, + "grad_norm": 0.4296875, + "learning_rate": 0.00019485043019231904, + "loss": 2.7356, + "step": 4905 + }, + { + "epoch": 0.21, + "grad_norm": 0.40625, + "learning_rate": 0.00019484830514603273, + "loss": 2.784, + "step": 4906 + }, + { + "epoch": 0.21, + "grad_norm": 0.46484375, + "learning_rate": 0.0001948461796729637, + "loss": 2.7248, + "step": 4907 + }, + { + "epoch": 0.21, + "grad_norm": 0.396484375, + "learning_rate": 0.00019484405377312154, + "loss": 2.8023, + "step": 4908 + }, + { + "epoch": 0.21, + "grad_norm": 0.47265625, + "learning_rate": 0.00019484192744651576, + "loss": 2.7432, + "step": 4909 + }, + { + "epoch": 0.21, + "grad_norm": 0.427734375, + "learning_rate": 0.00019483980069315602, + "loss": 2.5222, + "step": 4910 + }, + { + "epoch": 0.21, + "grad_norm": 0.451171875, + "learning_rate": 0.0001948376735130518, + "loss": 2.6612, + "step": 4911 + }, + { + "epoch": 0.21, + "grad_norm": 0.416015625, + "learning_rate": 0.00019483554590621273, + "loss": 2.6306, + "step": 4912 + }, + { + "epoch": 0.21, + "grad_norm": 0.44921875, + "learning_rate": 0.00019483341787264837, + "loss": 2.6517, + "step": 4913 + }, + { + "epoch": 0.21, + "grad_norm": 0.4609375, + "learning_rate": 0.00019483128941236828, + "loss": 2.6918, + "step": 4914 + }, + { + "epoch": 0.21, + "grad_norm": 0.41796875, + "learning_rate": 0.00019482916052538204, + "loss": 2.7777, + "step": 4915 + }, + { + "epoch": 0.21, + "grad_norm": 0.466796875, + "learning_rate": 0.00019482703121169922, + "loss": 2.7323, + "step": 4916 + }, + { + "epoch": 0.21, + "grad_norm": 0.466796875, + "learning_rate": 0.00019482490147132943, + "loss": 2.6622, + "step": 4917 + }, + { + "epoch": 0.21, + "grad_norm": 0.41015625, + "learning_rate": 0.00019482277130428223, + "loss": 2.6292, + "step": 4918 + }, + { + "epoch": 0.21, + "grad_norm": 0.408203125, + "learning_rate": 0.00019482064071056724, + "loss": 2.6805, + "step": 4919 + }, + { + "epoch": 0.21, + "grad_norm": 0.421875, + "learning_rate": 0.000194818509690194, + "loss": 2.6877, + "step": 4920 + }, + { + "epoch": 0.21, + "grad_norm": 0.4140625, + "learning_rate": 0.00019481637824317208, + "loss": 2.5619, + "step": 4921 + }, + { + "epoch": 0.21, + "grad_norm": 0.423828125, + "learning_rate": 0.00019481424636951119, + "loss": 2.6966, + "step": 4922 + }, + { + "epoch": 0.21, + "grad_norm": 0.390625, + "learning_rate": 0.00019481211406922078, + "loss": 2.6246, + "step": 4923 + }, + { + "epoch": 0.21, + "grad_norm": 0.451171875, + "learning_rate": 0.00019480998134231055, + "loss": 2.5804, + "step": 4924 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.00019480784818879, + "loss": 2.6146, + "step": 4925 + }, + { + "epoch": 0.21, + "grad_norm": 0.4296875, + "learning_rate": 0.0001948057146086688, + "loss": 2.6504, + "step": 4926 + }, + { + "epoch": 0.21, + "grad_norm": 0.462890625, + "learning_rate": 0.00019480358060195655, + "loss": 2.7523, + "step": 4927 + }, + { + "epoch": 0.21, + "grad_norm": 0.4609375, + "learning_rate": 0.00019480144616866281, + "loss": 2.6264, + "step": 4928 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.00019479931130879724, + "loss": 2.732, + "step": 4929 + }, + { + "epoch": 0.21, + "grad_norm": 0.47265625, + "learning_rate": 0.00019479717602236942, + "loss": 2.6859, + "step": 4930 + }, + { + "epoch": 0.21, + "grad_norm": 0.4609375, + "learning_rate": 0.0001947950403093889, + "loss": 2.7267, + "step": 4931 + }, + { + "epoch": 0.21, + "grad_norm": 0.455078125, + "learning_rate": 0.00019479290416986536, + "loss": 2.5167, + "step": 4932 + }, + { + "epoch": 0.21, + "grad_norm": 0.41796875, + "learning_rate": 0.00019479076760380842, + "loss": 2.6737, + "step": 4933 + }, + { + "epoch": 0.21, + "grad_norm": 0.466796875, + "learning_rate": 0.00019478863061122767, + "loss": 2.6558, + "step": 4934 + }, + { + "epoch": 0.21, + "grad_norm": 0.46875, + "learning_rate": 0.0001947864931921327, + "loss": 2.7371, + "step": 4935 + }, + { + "epoch": 0.21, + "grad_norm": 0.396484375, + "learning_rate": 0.00019478435534653316, + "loss": 2.4761, + "step": 4936 + }, + { + "epoch": 0.21, + "grad_norm": 0.3984375, + "learning_rate": 0.00019478221707443867, + "loss": 2.4708, + "step": 4937 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.00019478007837585884, + "loss": 2.6489, + "step": 4938 + }, + { + "epoch": 0.21, + "grad_norm": 0.486328125, + "learning_rate": 0.00019477793925080328, + "loss": 2.7762, + "step": 4939 + }, + { + "epoch": 0.21, + "grad_norm": 0.4375, + "learning_rate": 0.00019477579969928163, + "loss": 2.753, + "step": 4940 + }, + { + "epoch": 0.21, + "grad_norm": 0.455078125, + "learning_rate": 0.00019477365972130352, + "loss": 2.7235, + "step": 4941 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.0001947715193168786, + "loss": 2.5922, + "step": 4942 + }, + { + "epoch": 0.21, + "grad_norm": 0.408203125, + "learning_rate": 0.00019476937848601647, + "loss": 2.5616, + "step": 4943 + }, + { + "epoch": 0.21, + "grad_norm": 0.443359375, + "learning_rate": 0.00019476723722872678, + "loss": 2.6394, + "step": 4944 + }, + { + "epoch": 0.21, + "grad_norm": 0.44921875, + "learning_rate": 0.0001947650955450191, + "loss": 2.591, + "step": 4945 + }, + { + "epoch": 0.21, + "grad_norm": 0.45703125, + "learning_rate": 0.00019476295343490317, + "loss": 2.7164, + "step": 4946 + }, + { + "epoch": 0.21, + "grad_norm": 0.3984375, + "learning_rate": 0.00019476081089838857, + "loss": 2.6433, + "step": 4947 + }, + { + "epoch": 0.21, + "grad_norm": 0.453125, + "learning_rate": 0.00019475866793548495, + "loss": 2.8123, + "step": 4948 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.00019475652454620197, + "loss": 2.5795, + "step": 4949 + }, + { + "epoch": 0.21, + "grad_norm": 0.4140625, + "learning_rate": 0.00019475438073054925, + "loss": 2.6304, + "step": 4950 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.00019475223648853643, + "loss": 2.6453, + "step": 4951 + }, + { + "epoch": 0.21, + "grad_norm": 0.44921875, + "learning_rate": 0.00019475009182017317, + "loss": 2.5155, + "step": 4952 + }, + { + "epoch": 0.21, + "grad_norm": 0.421875, + "learning_rate": 0.00019474794672546918, + "loss": 2.5006, + "step": 4953 + }, + { + "epoch": 0.21, + "grad_norm": 0.44921875, + "learning_rate": 0.000194745801204434, + "loss": 2.6731, + "step": 4954 + }, + { + "epoch": 0.21, + "grad_norm": 0.46484375, + "learning_rate": 0.00019474365525707737, + "loss": 2.7773, + "step": 4955 + }, + { + "epoch": 0.21, + "grad_norm": 0.421875, + "learning_rate": 0.0001947415088834089, + "loss": 2.5669, + "step": 4956 + }, + { + "epoch": 0.21, + "grad_norm": 0.423828125, + "learning_rate": 0.00019473936208343827, + "loss": 2.5119, + "step": 4957 + }, + { + "epoch": 0.21, + "grad_norm": 0.44140625, + "learning_rate": 0.0001947372148571751, + "loss": 2.5846, + "step": 4958 + }, + { + "epoch": 0.21, + "grad_norm": 0.4375, + "learning_rate": 0.00019473506720462912, + "loss": 2.6628, + "step": 4959 + }, + { + "epoch": 0.21, + "grad_norm": 0.451171875, + "learning_rate": 0.00019473291912580997, + "loss": 2.6055, + "step": 4960 + }, + { + "epoch": 0.21, + "grad_norm": 0.390625, + "learning_rate": 0.0001947307706207273, + "loss": 2.656, + "step": 4961 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.00019472862168939077, + "loss": 2.7861, + "step": 4962 + }, + { + "epoch": 0.21, + "grad_norm": 0.4453125, + "learning_rate": 0.00019472647233181008, + "loss": 2.7328, + "step": 4963 + }, + { + "epoch": 0.21, + "grad_norm": 0.419921875, + "learning_rate": 0.00019472432254799484, + "loss": 2.6388, + "step": 4964 + }, + { + "epoch": 0.21, + "grad_norm": 0.423828125, + "learning_rate": 0.00019472217233795482, + "loss": 2.7249, + "step": 4965 + }, + { + "epoch": 0.21, + "grad_norm": 0.4609375, + "learning_rate": 0.0001947200217016996, + "loss": 2.5633, + "step": 4966 + }, + { + "epoch": 0.21, + "grad_norm": 0.45703125, + "learning_rate": 0.0001947178706392389, + "loss": 2.6954, + "step": 4967 + }, + { + "epoch": 0.21, + "grad_norm": 0.44140625, + "learning_rate": 0.00019471571915058244, + "loss": 2.757, + "step": 4968 + }, + { + "epoch": 0.21, + "grad_norm": 0.421875, + "learning_rate": 0.0001947135672357398, + "loss": 2.6712, + "step": 4969 + }, + { + "epoch": 0.21, + "grad_norm": 0.427734375, + "learning_rate": 0.00019471141489472077, + "loss": 2.4927, + "step": 4970 + }, + { + "epoch": 0.21, + "grad_norm": 0.490234375, + "learning_rate": 0.00019470926212753498, + "loss": 2.6447, + "step": 4971 + }, + { + "epoch": 0.21, + "grad_norm": 0.447265625, + "learning_rate": 0.00019470710893419212, + "loss": 2.619, + "step": 4972 + }, + { + "epoch": 0.21, + "grad_norm": 0.43359375, + "learning_rate": 0.00019470495531470188, + "loss": 2.7919, + "step": 4973 + }, + { + "epoch": 0.21, + "grad_norm": 0.431640625, + "learning_rate": 0.00019470280126907393, + "loss": 2.588, + "step": 4974 + }, + { + "epoch": 0.21, + "grad_norm": 0.419921875, + "learning_rate": 0.00019470064679731798, + "loss": 2.607, + "step": 4975 + }, + { + "epoch": 0.21, + "grad_norm": 0.4453125, + "learning_rate": 0.00019469849189944377, + "loss": 2.716, + "step": 4976 + }, + { + "epoch": 0.21, + "grad_norm": 0.46875, + "learning_rate": 0.00019469633657546092, + "loss": 2.6626, + "step": 4977 + }, + { + "epoch": 0.21, + "grad_norm": 0.466796875, + "learning_rate": 0.00019469418082537917, + "loss": 2.6661, + "step": 4978 + }, + { + "epoch": 0.21, + "grad_norm": 0.427734375, + "learning_rate": 0.00019469202464920824, + "loss": 2.6298, + "step": 4979 + }, + { + "epoch": 0.21, + "grad_norm": 0.42578125, + "learning_rate": 0.00019468986804695778, + "loss": 2.6426, + "step": 4980 + }, + { + "epoch": 0.21, + "grad_norm": 0.443359375, + "learning_rate": 0.0001946877110186375, + "loss": 2.6402, + "step": 4981 + }, + { + "epoch": 0.21, + "grad_norm": 0.443359375, + "learning_rate": 0.00019468555356425715, + "loss": 2.7386, + "step": 4982 + }, + { + "epoch": 0.21, + "grad_norm": 0.419921875, + "learning_rate": 0.0001946833956838264, + "loss": 2.5546, + "step": 4983 + }, + { + "epoch": 0.21, + "grad_norm": 0.44140625, + "learning_rate": 0.00019468123737735498, + "loss": 2.6774, + "step": 4984 + }, + { + "epoch": 0.21, + "grad_norm": 0.486328125, + "learning_rate": 0.00019467907864485263, + "loss": 2.805, + "step": 4985 + }, + { + "epoch": 0.21, + "grad_norm": 0.4453125, + "learning_rate": 0.00019467691948632897, + "loss": 2.769, + "step": 4986 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.0001946747599017938, + "loss": 2.7663, + "step": 4987 + }, + { + "epoch": 0.21, + "grad_norm": 0.423828125, + "learning_rate": 0.00019467259989125682, + "loss": 2.5465, + "step": 4988 + }, + { + "epoch": 0.21, + "grad_norm": 0.46875, + "learning_rate": 0.0001946704394547277, + "loss": 2.7662, + "step": 4989 + }, + { + "epoch": 0.21, + "grad_norm": 0.4453125, + "learning_rate": 0.00019466827859221625, + "loss": 2.5585, + "step": 4990 + }, + { + "epoch": 0.21, + "grad_norm": 0.4765625, + "learning_rate": 0.00019466611730373213, + "loss": 2.6848, + "step": 4991 + }, + { + "epoch": 0.21, + "grad_norm": 0.40625, + "learning_rate": 0.00019466395558928506, + "loss": 2.6892, + "step": 4992 + }, + { + "epoch": 0.21, + "grad_norm": 0.4375, + "learning_rate": 0.00019466179344888477, + "loss": 2.6634, + "step": 4993 + }, + { + "epoch": 0.21, + "grad_norm": 0.4375, + "learning_rate": 0.00019465963088254103, + "loss": 2.6075, + "step": 4994 + }, + { + "epoch": 0.21, + "grad_norm": 0.44140625, + "learning_rate": 0.00019465746789026355, + "loss": 2.5766, + "step": 4995 + }, + { + "epoch": 0.21, + "grad_norm": 0.40234375, + "learning_rate": 0.00019465530447206205, + "loss": 2.7247, + "step": 4996 + }, + { + "epoch": 0.21, + "grad_norm": 0.474609375, + "learning_rate": 0.00019465314062794628, + "loss": 2.8427, + "step": 4997 + }, + { + "epoch": 0.21, + "grad_norm": 0.427734375, + "learning_rate": 0.00019465097635792597, + "loss": 2.6708, + "step": 4998 + }, + { + "epoch": 0.21, + "grad_norm": 0.416015625, + "learning_rate": 0.00019464881166201083, + "loss": 2.6806, + "step": 4999 + }, + { + "epoch": 0.21, + "grad_norm": 0.41015625, + "learning_rate": 0.00019464664654021064, + "loss": 2.5779, + "step": 5000 + }, + { + "epoch": 0.21, + "grad_norm": 0.4765625, + "learning_rate": 0.00019464448099253514, + "loss": 2.63, + "step": 5001 + }, + { + "epoch": 0.21, + "grad_norm": 0.48828125, + "learning_rate": 0.00019464231501899407, + "loss": 2.7894, + "step": 5002 + }, + { + "epoch": 0.21, + "grad_norm": 0.4609375, + "learning_rate": 0.00019464014861959717, + "loss": 2.7639, + "step": 5003 + }, + { + "epoch": 0.21, + "grad_norm": 0.392578125, + "learning_rate": 0.00019463798179435418, + "loss": 2.6882, + "step": 5004 + }, + { + "epoch": 0.21, + "grad_norm": 0.470703125, + "learning_rate": 0.00019463581454327487, + "loss": 2.7131, + "step": 5005 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.00019463364686636895, + "loss": 2.7411, + "step": 5006 + }, + { + "epoch": 0.21, + "grad_norm": 0.484375, + "learning_rate": 0.00019463147876364623, + "loss": 2.6544, + "step": 5007 + }, + { + "epoch": 0.21, + "grad_norm": 0.388671875, + "learning_rate": 0.00019462931023511644, + "loss": 2.6685, + "step": 5008 + }, + { + "epoch": 0.21, + "grad_norm": 0.419921875, + "learning_rate": 0.00019462714128078932, + "loss": 2.7506, + "step": 5009 + }, + { + "epoch": 0.21, + "grad_norm": 0.41796875, + "learning_rate": 0.00019462497190067464, + "loss": 2.6468, + "step": 5010 + }, + { + "epoch": 0.21, + "grad_norm": 0.419921875, + "learning_rate": 0.00019462280209478218, + "loss": 2.7931, + "step": 5011 + }, + { + "epoch": 0.21, + "grad_norm": 0.455078125, + "learning_rate": 0.0001946206318631217, + "loss": 2.7698, + "step": 5012 + }, + { + "epoch": 0.21, + "grad_norm": 0.443359375, + "learning_rate": 0.00019461846120570294, + "loss": 2.6498, + "step": 5013 + }, + { + "epoch": 0.21, + "grad_norm": 0.466796875, + "learning_rate": 0.00019461629012253572, + "loss": 2.6846, + "step": 5014 + }, + { + "epoch": 0.21, + "grad_norm": 0.5703125, + "learning_rate": 0.00019461411861362973, + "loss": 2.6805, + "step": 5015 + }, + { + "epoch": 0.21, + "grad_norm": 0.3984375, + "learning_rate": 0.0001946119466789948, + "loss": 2.6424, + "step": 5016 + }, + { + "epoch": 0.21, + "grad_norm": 0.423828125, + "learning_rate": 0.0001946097743186407, + "loss": 2.6512, + "step": 5017 + }, + { + "epoch": 0.21, + "grad_norm": 0.482421875, + "learning_rate": 0.00019460760153257716, + "loss": 2.6759, + "step": 5018 + }, + { + "epoch": 0.21, + "grad_norm": 0.404296875, + "learning_rate": 0.00019460542832081402, + "loss": 2.5433, + "step": 5019 + }, + { + "epoch": 0.21, + "grad_norm": 0.404296875, + "learning_rate": 0.000194603254683361, + "loss": 2.618, + "step": 5020 + }, + { + "epoch": 0.21, + "grad_norm": 0.447265625, + "learning_rate": 0.00019460108062022792, + "loss": 2.6371, + "step": 5021 + }, + { + "epoch": 0.21, + "grad_norm": 0.458984375, + "learning_rate": 0.00019459890613142457, + "loss": 2.6174, + "step": 5022 + }, + { + "epoch": 0.21, + "grad_norm": 0.49609375, + "learning_rate": 0.00019459673121696066, + "loss": 2.5894, + "step": 5023 + }, + { + "epoch": 0.21, + "grad_norm": 0.482421875, + "learning_rate": 0.0001945945558768461, + "loss": 2.7065, + "step": 5024 + }, + { + "epoch": 0.21, + "grad_norm": 0.408203125, + "learning_rate": 0.00019459238011109058, + "loss": 2.7174, + "step": 5025 + }, + { + "epoch": 0.21, + "grad_norm": 0.419921875, + "learning_rate": 0.0001945902039197039, + "loss": 2.7206, + "step": 5026 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.00019458802730269587, + "loss": 2.5926, + "step": 5027 + }, + { + "epoch": 0.21, + "grad_norm": 0.4296875, + "learning_rate": 0.00019458585026007632, + "loss": 2.728, + "step": 5028 + }, + { + "epoch": 0.21, + "grad_norm": 0.46875, + "learning_rate": 0.000194583672791855, + "loss": 2.8262, + "step": 5029 + }, + { + "epoch": 0.21, + "grad_norm": 0.3671875, + "learning_rate": 0.0001945814948980417, + "loss": 2.6725, + "step": 5030 + }, + { + "epoch": 0.21, + "grad_norm": 0.47265625, + "learning_rate": 0.00019457931657864625, + "loss": 2.6702, + "step": 5031 + }, + { + "epoch": 0.21, + "grad_norm": 0.4375, + "learning_rate": 0.00019457713783367845, + "loss": 2.626, + "step": 5032 + }, + { + "epoch": 0.21, + "grad_norm": 0.443359375, + "learning_rate": 0.00019457495866314806, + "loss": 2.7379, + "step": 5033 + }, + { + "epoch": 0.21, + "grad_norm": 0.494140625, + "learning_rate": 0.00019457277906706494, + "loss": 2.7076, + "step": 5034 + }, + { + "epoch": 0.21, + "grad_norm": 0.44921875, + "learning_rate": 0.00019457059904543888, + "loss": 2.605, + "step": 5035 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.00019456841859827968, + "loss": 2.4861, + "step": 5036 + }, + { + "epoch": 0.21, + "grad_norm": 0.44921875, + "learning_rate": 0.00019456623772559718, + "loss": 2.6351, + "step": 5037 + }, + { + "epoch": 0.21, + "grad_norm": 0.453125, + "learning_rate": 0.00019456405642740115, + "loss": 2.6823, + "step": 5038 + }, + { + "epoch": 0.21, + "grad_norm": 0.435546875, + "learning_rate": 0.00019456187470370142, + "loss": 2.5421, + "step": 5039 + }, + { + "epoch": 0.21, + "grad_norm": 0.439453125, + "learning_rate": 0.00019455969255450782, + "loss": 2.5827, + "step": 5040 + }, + { + "epoch": 0.21, + "grad_norm": 0.455078125, + "learning_rate": 0.00019455750997983016, + "loss": 2.6333, + "step": 5041 + }, + { + "epoch": 0.21, + "grad_norm": 0.4296875, + "learning_rate": 0.00019455532697967826, + "loss": 2.7489, + "step": 5042 + }, + { + "epoch": 0.21, + "grad_norm": 0.4296875, + "learning_rate": 0.00019455314355406195, + "loss": 2.777, + "step": 5043 + }, + { + "epoch": 0.21, + "grad_norm": 0.443359375, + "learning_rate": 0.00019455095970299105, + "loss": 2.6455, + "step": 5044 + }, + { + "epoch": 0.21, + "grad_norm": 0.41015625, + "learning_rate": 0.00019454877542647537, + "loss": 2.6513, + "step": 5045 + }, + { + "epoch": 0.22, + "grad_norm": 0.365234375, + "learning_rate": 0.00019454659072452477, + "loss": 2.4261, + "step": 5046 + }, + { + "epoch": 0.22, + "grad_norm": 0.46484375, + "learning_rate": 0.00019454440559714905, + "loss": 2.5934, + "step": 5047 + }, + { + "epoch": 0.22, + "grad_norm": 0.4375, + "learning_rate": 0.00019454222004435808, + "loss": 2.7233, + "step": 5048 + }, + { + "epoch": 0.22, + "grad_norm": 0.3984375, + "learning_rate": 0.00019454003406616167, + "loss": 2.5012, + "step": 5049 + }, + { + "epoch": 0.22, + "grad_norm": 0.474609375, + "learning_rate": 0.00019453784766256964, + "loss": 2.7569, + "step": 5050 + }, + { + "epoch": 0.22, + "grad_norm": 0.388671875, + "learning_rate": 0.00019453566083359186, + "loss": 2.7172, + "step": 5051 + }, + { + "epoch": 0.22, + "grad_norm": 0.421875, + "learning_rate": 0.00019453347357923812, + "loss": 2.6549, + "step": 5052 + }, + { + "epoch": 0.22, + "grad_norm": 0.498046875, + "learning_rate": 0.00019453128589951835, + "loss": 2.7023, + "step": 5053 + }, + { + "epoch": 0.22, + "grad_norm": 0.419921875, + "learning_rate": 0.00019452909779444228, + "loss": 2.6769, + "step": 5054 + }, + { + "epoch": 0.22, + "grad_norm": 0.408203125, + "learning_rate": 0.00019452690926401986, + "loss": 2.5419, + "step": 5055 + }, + { + "epoch": 0.22, + "grad_norm": 0.4921875, + "learning_rate": 0.00019452472030826088, + "loss": 2.6958, + "step": 5056 + }, + { + "epoch": 0.22, + "grad_norm": 0.384765625, + "learning_rate": 0.0001945225309271752, + "loss": 2.6805, + "step": 5057 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.0001945203411207727, + "loss": 2.6276, + "step": 5058 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.00019451815088906318, + "loss": 2.7596, + "step": 5059 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.00019451596023205655, + "loss": 2.7571, + "step": 5060 + }, + { + "epoch": 0.22, + "grad_norm": 0.4140625, + "learning_rate": 0.00019451376914976262, + "loss": 2.7943, + "step": 5061 + }, + { + "epoch": 0.22, + "grad_norm": 0.421875, + "learning_rate": 0.00019451157764219126, + "loss": 2.4775, + "step": 5062 + }, + { + "epoch": 0.22, + "grad_norm": 0.4140625, + "learning_rate": 0.0001945093857093524, + "loss": 2.5959, + "step": 5063 + }, + { + "epoch": 0.22, + "grad_norm": 0.443359375, + "learning_rate": 0.00019450719335125576, + "loss": 2.7254, + "step": 5064 + }, + { + "epoch": 0.22, + "grad_norm": 0.4296875, + "learning_rate": 0.00019450500056791134, + "loss": 2.6971, + "step": 5065 + }, + { + "epoch": 0.22, + "grad_norm": 0.421875, + "learning_rate": 0.00019450280735932893, + "loss": 2.6971, + "step": 5066 + }, + { + "epoch": 0.22, + "grad_norm": 0.431640625, + "learning_rate": 0.00019450061372551844, + "loss": 2.7266, + "step": 5067 + }, + { + "epoch": 0.22, + "grad_norm": 0.482421875, + "learning_rate": 0.00019449841966648972, + "loss": 2.7338, + "step": 5068 + }, + { + "epoch": 0.22, + "grad_norm": 0.41796875, + "learning_rate": 0.00019449622518225262, + "loss": 2.5963, + "step": 5069 + }, + { + "epoch": 0.22, + "grad_norm": 0.443359375, + "learning_rate": 0.0001944940302728171, + "loss": 2.5792, + "step": 5070 + }, + { + "epoch": 0.22, + "grad_norm": 0.416015625, + "learning_rate": 0.00019449183493819292, + "loss": 2.5125, + "step": 5071 + }, + { + "epoch": 0.22, + "grad_norm": 0.447265625, + "learning_rate": 0.00019448963917839004, + "loss": 2.5187, + "step": 5072 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.00019448744299341834, + "loss": 2.5585, + "step": 5073 + }, + { + "epoch": 0.22, + "grad_norm": 0.4140625, + "learning_rate": 0.00019448524638328765, + "loss": 2.5432, + "step": 5074 + }, + { + "epoch": 0.22, + "grad_norm": 0.435546875, + "learning_rate": 0.00019448304934800787, + "loss": 2.6821, + "step": 5075 + }, + { + "epoch": 0.22, + "grad_norm": 0.4140625, + "learning_rate": 0.00019448085188758892, + "loss": 2.55, + "step": 5076 + }, + { + "epoch": 0.22, + "grad_norm": 0.490234375, + "learning_rate": 0.00019447865400204066, + "loss": 2.6596, + "step": 5077 + }, + { + "epoch": 0.22, + "grad_norm": 0.4609375, + "learning_rate": 0.000194476455691373, + "loss": 2.6015, + "step": 5078 + }, + { + "epoch": 0.22, + "grad_norm": 0.451171875, + "learning_rate": 0.00019447425695559578, + "loss": 2.788, + "step": 5079 + }, + { + "epoch": 0.22, + "grad_norm": 0.453125, + "learning_rate": 0.00019447205779471897, + "loss": 2.5898, + "step": 5080 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.0001944698582087524, + "loss": 2.6411, + "step": 5081 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.00019446765819770598, + "loss": 2.6024, + "step": 5082 + }, + { + "epoch": 0.22, + "grad_norm": 0.412109375, + "learning_rate": 0.00019446545776158965, + "loss": 2.5434, + "step": 5083 + }, + { + "epoch": 0.22, + "grad_norm": 0.470703125, + "learning_rate": 0.00019446325690041327, + "loss": 2.825, + "step": 5084 + }, + { + "epoch": 0.22, + "grad_norm": 0.4296875, + "learning_rate": 0.00019446105561418678, + "loss": 2.5035, + "step": 5085 + }, + { + "epoch": 0.22, + "grad_norm": 0.52734375, + "learning_rate": 0.00019445885390292003, + "loss": 2.8572, + "step": 5086 + }, + { + "epoch": 0.22, + "grad_norm": 0.423828125, + "learning_rate": 0.00019445665176662295, + "loss": 2.6736, + "step": 5087 + }, + { + "epoch": 0.22, + "grad_norm": 0.4140625, + "learning_rate": 0.00019445444920530549, + "loss": 2.6886, + "step": 5088 + }, + { + "epoch": 0.22, + "grad_norm": 0.376953125, + "learning_rate": 0.0001944522462189775, + "loss": 2.5163, + "step": 5089 + }, + { + "epoch": 0.22, + "grad_norm": 0.44921875, + "learning_rate": 0.00019445004280764896, + "loss": 2.6869, + "step": 5090 + }, + { + "epoch": 0.22, + "grad_norm": 0.439453125, + "learning_rate": 0.0001944478389713297, + "loss": 2.6872, + "step": 5091 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.0001944456347100297, + "loss": 2.6107, + "step": 5092 + }, + { + "epoch": 0.22, + "grad_norm": 0.458984375, + "learning_rate": 0.00019444343002375888, + "loss": 2.5653, + "step": 5093 + }, + { + "epoch": 0.22, + "grad_norm": 0.447265625, + "learning_rate": 0.0001944412249125271, + "loss": 2.6076, + "step": 5094 + }, + { + "epoch": 0.22, + "grad_norm": 0.498046875, + "learning_rate": 0.00019443901937634433, + "loss": 2.7119, + "step": 5095 + }, + { + "epoch": 0.22, + "grad_norm": 0.419921875, + "learning_rate": 0.00019443681341522052, + "loss": 2.6215, + "step": 5096 + }, + { + "epoch": 0.22, + "grad_norm": 0.423828125, + "learning_rate": 0.00019443460702916554, + "loss": 2.819, + "step": 5097 + }, + { + "epoch": 0.22, + "grad_norm": 0.462890625, + "learning_rate": 0.00019443240021818934, + "loss": 2.7065, + "step": 5098 + }, + { + "epoch": 0.22, + "grad_norm": 0.412109375, + "learning_rate": 0.00019443019298230184, + "loss": 2.6, + "step": 5099 + }, + { + "epoch": 0.22, + "grad_norm": 0.42578125, + "learning_rate": 0.000194427985321513, + "loss": 2.6313, + "step": 5100 + }, + { + "epoch": 0.22, + "grad_norm": 0.48046875, + "learning_rate": 0.00019442577723583276, + "loss": 2.5685, + "step": 5101 + }, + { + "epoch": 0.22, + "grad_norm": 0.431640625, + "learning_rate": 0.000194423568725271, + "loss": 2.6905, + "step": 5102 + }, + { + "epoch": 0.22, + "grad_norm": 0.451171875, + "learning_rate": 0.00019442135978983767, + "loss": 2.6977, + "step": 5103 + }, + { + "epoch": 0.22, + "grad_norm": 0.4765625, + "learning_rate": 0.00019441915042954272, + "loss": 2.7582, + "step": 5104 + }, + { + "epoch": 0.22, + "grad_norm": 0.462890625, + "learning_rate": 0.00019441694064439616, + "loss": 2.7455, + "step": 5105 + }, + { + "epoch": 0.22, + "grad_norm": 0.50390625, + "learning_rate": 0.00019441473043440782, + "loss": 2.6485, + "step": 5106 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.0001944125197995877, + "loss": 2.638, + "step": 5107 + }, + { + "epoch": 0.22, + "grad_norm": 0.49609375, + "learning_rate": 0.0001944103087399458, + "loss": 2.7193, + "step": 5108 + }, + { + "epoch": 0.22, + "grad_norm": 0.4453125, + "learning_rate": 0.00019440809725549196, + "loss": 2.4965, + "step": 5109 + }, + { + "epoch": 0.22, + "grad_norm": 0.43359375, + "learning_rate": 0.00019440588534623622, + "loss": 2.5264, + "step": 5110 + }, + { + "epoch": 0.22, + "grad_norm": 0.443359375, + "learning_rate": 0.00019440367301218843, + "loss": 2.75, + "step": 5111 + }, + { + "epoch": 0.22, + "grad_norm": 0.5546875, + "learning_rate": 0.00019440146025335867, + "loss": 2.8112, + "step": 5112 + }, + { + "epoch": 0.22, + "grad_norm": 0.40234375, + "learning_rate": 0.00019439924706975682, + "loss": 2.6444, + "step": 5113 + }, + { + "epoch": 0.22, + "grad_norm": 0.46875, + "learning_rate": 0.00019439703346139288, + "loss": 2.7937, + "step": 5114 + }, + { + "epoch": 0.22, + "grad_norm": 0.51171875, + "learning_rate": 0.00019439481942827677, + "loss": 2.8032, + "step": 5115 + }, + { + "epoch": 0.22, + "grad_norm": 0.4296875, + "learning_rate": 0.00019439260497041848, + "loss": 2.6864, + "step": 5116 + }, + { + "epoch": 0.22, + "grad_norm": 0.4453125, + "learning_rate": 0.00019439039008782797, + "loss": 2.5584, + "step": 5117 + }, + { + "epoch": 0.22, + "grad_norm": 0.443359375, + "learning_rate": 0.00019438817478051518, + "loss": 2.7927, + "step": 5118 + }, + { + "epoch": 0.22, + "grad_norm": 0.47265625, + "learning_rate": 0.00019438595904849008, + "loss": 2.7263, + "step": 5119 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.0001943837428917627, + "loss": 2.6925, + "step": 5120 + }, + { + "epoch": 0.22, + "grad_norm": 0.416015625, + "learning_rate": 0.00019438152631034297, + "loss": 2.5302, + "step": 5121 + }, + { + "epoch": 0.22, + "grad_norm": 0.423828125, + "learning_rate": 0.00019437930930424083, + "loss": 2.7004, + "step": 5122 + }, + { + "epoch": 0.22, + "grad_norm": 0.41796875, + "learning_rate": 0.0001943770918734663, + "loss": 2.6264, + "step": 5123 + }, + { + "epoch": 0.22, + "grad_norm": 0.44921875, + "learning_rate": 0.00019437487401802937, + "loss": 2.7862, + "step": 5124 + }, + { + "epoch": 0.22, + "grad_norm": 0.470703125, + "learning_rate": 0.00019437265573794, + "loss": 2.6964, + "step": 5125 + }, + { + "epoch": 0.22, + "grad_norm": 0.435546875, + "learning_rate": 0.00019437043703320815, + "loss": 2.7117, + "step": 5126 + }, + { + "epoch": 0.22, + "grad_norm": 0.474609375, + "learning_rate": 0.0001943682179038438, + "loss": 2.66, + "step": 5127 + }, + { + "epoch": 0.22, + "grad_norm": 0.421875, + "learning_rate": 0.000194365998349857, + "loss": 2.7593, + "step": 5128 + }, + { + "epoch": 0.22, + "grad_norm": 0.4375, + "learning_rate": 0.00019436377837125766, + "loss": 2.8155, + "step": 5129 + }, + { + "epoch": 0.22, + "grad_norm": 0.439453125, + "learning_rate": 0.0001943615579680558, + "loss": 2.5067, + "step": 5130 + }, + { + "epoch": 0.22, + "grad_norm": 0.431640625, + "learning_rate": 0.00019435933714026144, + "loss": 2.6523, + "step": 5131 + }, + { + "epoch": 0.22, + "grad_norm": 0.49609375, + "learning_rate": 0.00019435711588788456, + "loss": 2.7134, + "step": 5132 + }, + { + "epoch": 0.22, + "grad_norm": 0.46875, + "learning_rate": 0.00019435489421093508, + "loss": 2.6462, + "step": 5133 + }, + { + "epoch": 0.22, + "grad_norm": 0.40234375, + "learning_rate": 0.00019435267210942314, + "loss": 2.5634, + "step": 5134 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.0001943504495833586, + "loss": 2.7197, + "step": 5135 + }, + { + "epoch": 0.22, + "grad_norm": 0.44921875, + "learning_rate": 0.00019434822663275155, + "loss": 2.6461, + "step": 5136 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.00019434600325761193, + "loss": 2.6034, + "step": 5137 + }, + { + "epoch": 0.22, + "grad_norm": 0.458984375, + "learning_rate": 0.0001943437794579498, + "loss": 2.7489, + "step": 5138 + }, + { + "epoch": 0.22, + "grad_norm": 0.484375, + "learning_rate": 0.0001943415552337751, + "loss": 2.6101, + "step": 5139 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.00019433933058509792, + "loss": 2.6359, + "step": 5140 + }, + { + "epoch": 0.22, + "grad_norm": 0.40234375, + "learning_rate": 0.00019433710551192823, + "loss": 2.6245, + "step": 5141 + }, + { + "epoch": 0.22, + "grad_norm": 0.41015625, + "learning_rate": 0.00019433488001427605, + "loss": 2.6902, + "step": 5142 + }, + { + "epoch": 0.22, + "grad_norm": 0.431640625, + "learning_rate": 0.00019433265409215137, + "loss": 2.4939, + "step": 5143 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.0001943304277455642, + "loss": 2.7461, + "step": 5144 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.0001943282009745246, + "loss": 2.7603, + "step": 5145 + }, + { + "epoch": 0.22, + "grad_norm": 0.470703125, + "learning_rate": 0.00019432597377904256, + "loss": 2.6377, + "step": 5146 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.00019432374615912813, + "loss": 2.728, + "step": 5147 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.0001943215181147913, + "loss": 2.6105, + "step": 5148 + }, + { + "epoch": 0.22, + "grad_norm": 0.451171875, + "learning_rate": 0.0001943192896460421, + "loss": 2.6897, + "step": 5149 + }, + { + "epoch": 0.22, + "grad_norm": 0.5078125, + "learning_rate": 0.00019431706075289057, + "loss": 2.6673, + "step": 5150 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.00019431483143534674, + "loss": 2.5276, + "step": 5151 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.00019431260169342065, + "loss": 2.7141, + "step": 5152 + }, + { + "epoch": 0.22, + "grad_norm": 0.421875, + "learning_rate": 0.0001943103715271223, + "loss": 2.694, + "step": 5153 + }, + { + "epoch": 0.22, + "grad_norm": 0.3828125, + "learning_rate": 0.00019430814093646177, + "loss": 2.6906, + "step": 5154 + }, + { + "epoch": 0.22, + "grad_norm": 0.416015625, + "learning_rate": 0.00019430590992144904, + "loss": 2.6373, + "step": 5155 + }, + { + "epoch": 0.22, + "grad_norm": 0.419921875, + "learning_rate": 0.00019430367848209418, + "loss": 2.9271, + "step": 5156 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.00019430144661840726, + "loss": 2.6582, + "step": 5157 + }, + { + "epoch": 0.22, + "grad_norm": 0.416015625, + "learning_rate": 0.00019429921433039828, + "loss": 2.7201, + "step": 5158 + }, + { + "epoch": 0.22, + "grad_norm": 0.44921875, + "learning_rate": 0.0001942969816180773, + "loss": 2.61, + "step": 5159 + }, + { + "epoch": 0.22, + "grad_norm": 0.4375, + "learning_rate": 0.00019429474848145435, + "loss": 2.5835, + "step": 5160 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.0001942925149205395, + "loss": 2.8356, + "step": 5161 + }, + { + "epoch": 0.22, + "grad_norm": 0.37890625, + "learning_rate": 0.00019429028093534282, + "loss": 2.6035, + "step": 5162 + }, + { + "epoch": 0.22, + "grad_norm": 0.42578125, + "learning_rate": 0.00019428804652587428, + "loss": 2.6194, + "step": 5163 + }, + { + "epoch": 0.22, + "grad_norm": 0.40234375, + "learning_rate": 0.00019428581169214404, + "loss": 2.639, + "step": 5164 + }, + { + "epoch": 0.22, + "grad_norm": 0.478515625, + "learning_rate": 0.00019428357643416208, + "loss": 2.6996, + "step": 5165 + }, + { + "epoch": 0.22, + "grad_norm": 0.443359375, + "learning_rate": 0.00019428134075193847, + "loss": 2.6323, + "step": 5166 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.0001942791046454833, + "loss": 2.6199, + "step": 5167 + }, + { + "epoch": 0.22, + "grad_norm": 0.462890625, + "learning_rate": 0.0001942768681148066, + "loss": 2.6342, + "step": 5168 + }, + { + "epoch": 0.22, + "grad_norm": 0.451171875, + "learning_rate": 0.0001942746311599185, + "loss": 2.7585, + "step": 5169 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.00019427239378082896, + "loss": 2.5972, + "step": 5170 + }, + { + "epoch": 0.22, + "grad_norm": 0.48046875, + "learning_rate": 0.00019427015597754812, + "loss": 2.7433, + "step": 5171 + }, + { + "epoch": 0.22, + "grad_norm": 0.423828125, + "learning_rate": 0.00019426791775008603, + "loss": 2.5538, + "step": 5172 + }, + { + "epoch": 0.22, + "grad_norm": 0.359375, + "learning_rate": 0.00019426567909845274, + "loss": 2.6204, + "step": 5173 + }, + { + "epoch": 0.22, + "grad_norm": 0.4609375, + "learning_rate": 0.00019426344002265836, + "loss": 2.602, + "step": 5174 + }, + { + "epoch": 0.22, + "grad_norm": 0.4375, + "learning_rate": 0.00019426120052271295, + "loss": 2.5728, + "step": 5175 + }, + { + "epoch": 0.22, + "grad_norm": 0.435546875, + "learning_rate": 0.0001942589605986266, + "loss": 2.6743, + "step": 5176 + }, + { + "epoch": 0.22, + "grad_norm": 0.45703125, + "learning_rate": 0.00019425672025040935, + "loss": 2.7623, + "step": 5177 + }, + { + "epoch": 0.22, + "grad_norm": 0.42578125, + "learning_rate": 0.00019425447947807133, + "loss": 2.6235, + "step": 5178 + }, + { + "epoch": 0.22, + "grad_norm": 0.451171875, + "learning_rate": 0.00019425223828162258, + "loss": 2.6928, + "step": 5179 + }, + { + "epoch": 0.22, + "grad_norm": 0.41015625, + "learning_rate": 0.0001942499966610732, + "loss": 2.6122, + "step": 5180 + }, + { + "epoch": 0.22, + "grad_norm": 0.498046875, + "learning_rate": 0.00019424775461643326, + "loss": 2.6961, + "step": 5181 + }, + { + "epoch": 0.22, + "grad_norm": 0.470703125, + "learning_rate": 0.00019424551214771292, + "loss": 2.8002, + "step": 5182 + }, + { + "epoch": 0.22, + "grad_norm": 0.4296875, + "learning_rate": 0.0001942432692549222, + "loss": 2.6687, + "step": 5183 + }, + { + "epoch": 0.22, + "grad_norm": 0.408203125, + "learning_rate": 0.00019424102593807122, + "loss": 2.6788, + "step": 5184 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.00019423878219717002, + "loss": 2.7567, + "step": 5185 + }, + { + "epoch": 0.22, + "grad_norm": 0.462890625, + "learning_rate": 0.00019423653803222877, + "loss": 2.5719, + "step": 5186 + }, + { + "epoch": 0.22, + "grad_norm": 0.412109375, + "learning_rate": 0.00019423429344325754, + "loss": 2.5126, + "step": 5187 + }, + { + "epoch": 0.22, + "grad_norm": 0.486328125, + "learning_rate": 0.00019423204843026644, + "loss": 2.6519, + "step": 5188 + }, + { + "epoch": 0.22, + "grad_norm": 0.4453125, + "learning_rate": 0.00019422980299326555, + "loss": 2.5308, + "step": 5189 + }, + { + "epoch": 0.22, + "grad_norm": 0.466796875, + "learning_rate": 0.00019422755713226496, + "loss": 2.6649, + "step": 5190 + }, + { + "epoch": 0.22, + "grad_norm": 0.412109375, + "learning_rate": 0.00019422531084727483, + "loss": 2.6068, + "step": 5191 + }, + { + "epoch": 0.22, + "grad_norm": 0.42578125, + "learning_rate": 0.00019422306413830522, + "loss": 2.5612, + "step": 5192 + }, + { + "epoch": 0.22, + "grad_norm": 0.4609375, + "learning_rate": 0.00019422081700536627, + "loss": 2.6026, + "step": 5193 + }, + { + "epoch": 0.22, + "grad_norm": 0.50390625, + "learning_rate": 0.00019421856944846805, + "loss": 2.7309, + "step": 5194 + }, + { + "epoch": 0.22, + "grad_norm": 0.447265625, + "learning_rate": 0.00019421632146762074, + "loss": 2.7001, + "step": 5195 + }, + { + "epoch": 0.22, + "grad_norm": 0.490234375, + "learning_rate": 0.00019421407306283438, + "loss": 2.7024, + "step": 5196 + }, + { + "epoch": 0.22, + "grad_norm": 0.431640625, + "learning_rate": 0.00019421182423411915, + "loss": 2.7043, + "step": 5197 + }, + { + "epoch": 0.22, + "grad_norm": 0.47265625, + "learning_rate": 0.0001942095749814851, + "loss": 2.6564, + "step": 5198 + }, + { + "epoch": 0.22, + "grad_norm": 0.416015625, + "learning_rate": 0.00019420732530494244, + "loss": 2.5834, + "step": 5199 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.0001942050752045012, + "loss": 2.7559, + "step": 5200 + }, + { + "epoch": 0.22, + "grad_norm": 0.474609375, + "learning_rate": 0.0001942028246801716, + "loss": 2.8067, + "step": 5201 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.0001942005737319637, + "loss": 2.6931, + "step": 5202 + }, + { + "epoch": 0.22, + "grad_norm": 0.4375, + "learning_rate": 0.00019419832235988762, + "loss": 2.7538, + "step": 5203 + }, + { + "epoch": 0.22, + "grad_norm": 0.48046875, + "learning_rate": 0.00019419607056395352, + "loss": 2.8171, + "step": 5204 + }, + { + "epoch": 0.22, + "grad_norm": 0.435546875, + "learning_rate": 0.00019419381834417156, + "loss": 2.5326, + "step": 5205 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.0001941915657005518, + "loss": 2.5724, + "step": 5206 + }, + { + "epoch": 0.22, + "grad_norm": 0.40234375, + "learning_rate": 0.00019418931263310446, + "loss": 2.6259, + "step": 5207 + }, + { + "epoch": 0.22, + "grad_norm": 0.419921875, + "learning_rate": 0.00019418705914183958, + "loss": 2.648, + "step": 5208 + }, + { + "epoch": 0.22, + "grad_norm": 0.478515625, + "learning_rate": 0.0001941848052267674, + "loss": 2.6022, + "step": 5209 + }, + { + "epoch": 0.22, + "grad_norm": 0.421875, + "learning_rate": 0.000194182550887898, + "loss": 2.6553, + "step": 5210 + }, + { + "epoch": 0.22, + "grad_norm": 0.43359375, + "learning_rate": 0.00019418029612524156, + "loss": 2.5849, + "step": 5211 + }, + { + "epoch": 0.22, + "grad_norm": 0.53515625, + "learning_rate": 0.00019417804093880818, + "loss": 2.8456, + "step": 5212 + }, + { + "epoch": 0.22, + "grad_norm": 0.412109375, + "learning_rate": 0.00019417578532860804, + "loss": 2.7353, + "step": 5213 + }, + { + "epoch": 0.22, + "grad_norm": 0.494140625, + "learning_rate": 0.00019417352929465128, + "loss": 2.656, + "step": 5214 + }, + { + "epoch": 0.22, + "grad_norm": 0.458984375, + "learning_rate": 0.00019417127283694808, + "loss": 2.6696, + "step": 5215 + }, + { + "epoch": 0.22, + "grad_norm": 0.396484375, + "learning_rate": 0.00019416901595550857, + "loss": 2.6026, + "step": 5216 + }, + { + "epoch": 0.22, + "grad_norm": 0.484375, + "learning_rate": 0.00019416675865034286, + "loss": 2.7877, + "step": 5217 + }, + { + "epoch": 0.22, + "grad_norm": 0.5078125, + "learning_rate": 0.00019416450092146116, + "loss": 2.6268, + "step": 5218 + }, + { + "epoch": 0.22, + "grad_norm": 0.42578125, + "learning_rate": 0.00019416224276887364, + "loss": 2.6556, + "step": 5219 + }, + { + "epoch": 0.22, + "grad_norm": 0.41796875, + "learning_rate": 0.00019415998419259044, + "loss": 2.5424, + "step": 5220 + }, + { + "epoch": 0.22, + "grad_norm": 0.443359375, + "learning_rate": 0.00019415772519262172, + "loss": 2.6487, + "step": 5221 + }, + { + "epoch": 0.22, + "grad_norm": 0.4375, + "learning_rate": 0.00019415546576897764, + "loss": 2.6921, + "step": 5222 + }, + { + "epoch": 0.22, + "grad_norm": 0.462890625, + "learning_rate": 0.00019415320592166836, + "loss": 2.7056, + "step": 5223 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.00019415094565070412, + "loss": 2.5685, + "step": 5224 + }, + { + "epoch": 0.22, + "grad_norm": 0.44921875, + "learning_rate": 0.00019414868495609496, + "loss": 2.5791, + "step": 5225 + }, + { + "epoch": 0.22, + "grad_norm": 0.41796875, + "learning_rate": 0.00019414642383785115, + "loss": 2.7077, + "step": 5226 + }, + { + "epoch": 0.22, + "grad_norm": 0.4453125, + "learning_rate": 0.00019414416229598288, + "loss": 2.7169, + "step": 5227 + }, + { + "epoch": 0.22, + "grad_norm": 0.423828125, + "learning_rate": 0.00019414190033050023, + "loss": 2.6674, + "step": 5228 + }, + { + "epoch": 0.22, + "grad_norm": 0.466796875, + "learning_rate": 0.00019413963794141345, + "loss": 2.6362, + "step": 5229 + }, + { + "epoch": 0.22, + "grad_norm": 0.49609375, + "learning_rate": 0.00019413737512873273, + "loss": 2.8093, + "step": 5230 + }, + { + "epoch": 0.22, + "grad_norm": 0.423828125, + "learning_rate": 0.0001941351118924682, + "loss": 2.5858, + "step": 5231 + }, + { + "epoch": 0.22, + "grad_norm": 0.404296875, + "learning_rate": 0.00019413284823263007, + "loss": 2.5853, + "step": 5232 + }, + { + "epoch": 0.22, + "grad_norm": 0.396484375, + "learning_rate": 0.00019413058414922854, + "loss": 2.5604, + "step": 5233 + }, + { + "epoch": 0.22, + "grad_norm": 0.4453125, + "learning_rate": 0.00019412831964227376, + "loss": 2.488, + "step": 5234 + }, + { + "epoch": 0.22, + "grad_norm": 0.423828125, + "learning_rate": 0.00019412605471177594, + "loss": 2.5573, + "step": 5235 + }, + { + "epoch": 0.22, + "grad_norm": 0.435546875, + "learning_rate": 0.00019412378935774528, + "loss": 2.6213, + "step": 5236 + }, + { + "epoch": 0.22, + "grad_norm": 0.4921875, + "learning_rate": 0.00019412152358019199, + "loss": 2.7658, + "step": 5237 + }, + { + "epoch": 0.22, + "grad_norm": 0.439453125, + "learning_rate": 0.0001941192573791262, + "loss": 2.5215, + "step": 5238 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.00019411699075455818, + "loss": 2.5049, + "step": 5239 + }, + { + "epoch": 0.22, + "grad_norm": 0.421875, + "learning_rate": 0.0001941147237064981, + "loss": 2.5617, + "step": 5240 + }, + { + "epoch": 0.22, + "grad_norm": 0.380859375, + "learning_rate": 0.00019411245623495615, + "loss": 2.6595, + "step": 5241 + }, + { + "epoch": 0.22, + "grad_norm": 0.421875, + "learning_rate": 0.00019411018833994252, + "loss": 2.7642, + "step": 5242 + }, + { + "epoch": 0.22, + "grad_norm": 0.427734375, + "learning_rate": 0.00019410792002146746, + "loss": 2.7037, + "step": 5243 + }, + { + "epoch": 0.22, + "grad_norm": 0.4765625, + "learning_rate": 0.00019410565127954115, + "loss": 2.7146, + "step": 5244 + }, + { + "epoch": 0.22, + "grad_norm": 0.4609375, + "learning_rate": 0.0001941033821141738, + "loss": 2.7477, + "step": 5245 + }, + { + "epoch": 0.22, + "grad_norm": 0.453125, + "learning_rate": 0.0001941011125253756, + "loss": 2.7547, + "step": 5246 + }, + { + "epoch": 0.22, + "grad_norm": 0.474609375, + "learning_rate": 0.0001940988425131568, + "loss": 2.7048, + "step": 5247 + }, + { + "epoch": 0.22, + "grad_norm": 0.443359375, + "learning_rate": 0.0001940965720775276, + "loss": 2.6182, + "step": 5248 + }, + { + "epoch": 0.22, + "grad_norm": 0.458984375, + "learning_rate": 0.00019409430121849824, + "loss": 2.7543, + "step": 5249 + }, + { + "epoch": 0.22, + "grad_norm": 0.423828125, + "learning_rate": 0.00019409202993607885, + "loss": 2.6267, + "step": 5250 + }, + { + "epoch": 0.22, + "grad_norm": 0.45703125, + "learning_rate": 0.00019408975823027977, + "loss": 2.5945, + "step": 5251 + }, + { + "epoch": 0.22, + "grad_norm": 0.44140625, + "learning_rate": 0.00019408748610111115, + "loss": 2.6933, + "step": 5252 + }, + { + "epoch": 0.22, + "grad_norm": 0.4375, + "learning_rate": 0.0001940852135485832, + "loss": 2.5689, + "step": 5253 + }, + { + "epoch": 0.22, + "grad_norm": 0.482421875, + "learning_rate": 0.0001940829405727062, + "loss": 2.7552, + "step": 5254 + }, + { + "epoch": 0.22, + "grad_norm": 0.45703125, + "learning_rate": 0.00019408066717349035, + "loss": 2.6604, + "step": 5255 + }, + { + "epoch": 0.22, + "grad_norm": 0.419921875, + "learning_rate": 0.00019407839335094587, + "loss": 2.6202, + "step": 5256 + }, + { + "epoch": 0.22, + "grad_norm": 0.48828125, + "learning_rate": 0.000194076119105083, + "loss": 2.7188, + "step": 5257 + }, + { + "epoch": 0.22, + "grad_norm": 0.46875, + "learning_rate": 0.000194073844435912, + "loss": 2.705, + "step": 5258 + }, + { + "epoch": 0.22, + "grad_norm": 0.451171875, + "learning_rate": 0.00019407156934344303, + "loss": 2.6769, + "step": 5259 + }, + { + "epoch": 0.22, + "grad_norm": 0.4375, + "learning_rate": 0.0001940692938276864, + "loss": 2.4971, + "step": 5260 + }, + { + "epoch": 0.22, + "grad_norm": 0.38671875, + "learning_rate": 0.00019406701788865233, + "loss": 2.6813, + "step": 5261 + }, + { + "epoch": 0.22, + "grad_norm": 0.455078125, + "learning_rate": 0.00019406474152635107, + "loss": 2.5736, + "step": 5262 + }, + { + "epoch": 0.22, + "grad_norm": 0.451171875, + "learning_rate": 0.00019406246474079283, + "loss": 2.6038, + "step": 5263 + }, + { + "epoch": 0.22, + "grad_norm": 0.3828125, + "learning_rate": 0.0001940601875319879, + "loss": 2.6566, + "step": 5264 + }, + { + "epoch": 0.22, + "grad_norm": 0.416015625, + "learning_rate": 0.0001940579098999465, + "loss": 2.6067, + "step": 5265 + }, + { + "epoch": 0.22, + "grad_norm": 0.46484375, + "learning_rate": 0.00019405563184467884, + "loss": 2.7673, + "step": 5266 + }, + { + "epoch": 0.22, + "grad_norm": 0.439453125, + "learning_rate": 0.00019405335336619524, + "loss": 2.5098, + "step": 5267 + }, + { + "epoch": 0.22, + "grad_norm": 0.453125, + "learning_rate": 0.00019405107446450592, + "loss": 2.6319, + "step": 5268 + }, + { + "epoch": 0.22, + "grad_norm": 0.470703125, + "learning_rate": 0.00019404879513962113, + "loss": 2.7202, + "step": 5269 + }, + { + "epoch": 0.22, + "grad_norm": 0.4453125, + "learning_rate": 0.00019404651539155115, + "loss": 2.6487, + "step": 5270 + }, + { + "epoch": 0.22, + "grad_norm": 0.400390625, + "learning_rate": 0.0001940442352203062, + "loss": 2.4881, + "step": 5271 + }, + { + "epoch": 0.22, + "grad_norm": 0.462890625, + "learning_rate": 0.00019404195462589656, + "loss": 2.6537, + "step": 5272 + }, + { + "epoch": 0.22, + "grad_norm": 0.50390625, + "learning_rate": 0.00019403967360833254, + "loss": 2.6837, + "step": 5273 + }, + { + "epoch": 0.22, + "grad_norm": 0.431640625, + "learning_rate": 0.0001940373921676243, + "loss": 2.6477, + "step": 5274 + }, + { + "epoch": 0.22, + "grad_norm": 0.46875, + "learning_rate": 0.00019403511030378218, + "loss": 2.7707, + "step": 5275 + }, + { + "epoch": 0.22, + "grad_norm": 0.453125, + "learning_rate": 0.00019403282801681642, + "loss": 2.7117, + "step": 5276 + }, + { + "epoch": 0.22, + "grad_norm": 0.498046875, + "learning_rate": 0.0001940305453067373, + "loss": 2.4899, + "step": 5277 + }, + { + "epoch": 0.22, + "grad_norm": 0.4140625, + "learning_rate": 0.0001940282621735551, + "loss": 2.6564, + "step": 5278 + }, + { + "epoch": 0.22, + "grad_norm": 0.38671875, + "learning_rate": 0.00019402597861728008, + "loss": 2.4762, + "step": 5279 + }, + { + "epoch": 0.22, + "grad_norm": 0.416015625, + "learning_rate": 0.00019402369463792255, + "loss": 2.6721, + "step": 5280 + }, + { + "epoch": 0.23, + "grad_norm": 0.421875, + "learning_rate": 0.0001940214102354927, + "loss": 2.7218, + "step": 5281 + }, + { + "epoch": 0.23, + "grad_norm": 0.4453125, + "learning_rate": 0.0001940191254100009, + "loss": 2.6158, + "step": 5282 + }, + { + "epoch": 0.23, + "grad_norm": 0.427734375, + "learning_rate": 0.0001940168401614574, + "loss": 2.6898, + "step": 5283 + }, + { + "epoch": 0.23, + "grad_norm": 0.4140625, + "learning_rate": 0.00019401455448987245, + "loss": 2.5726, + "step": 5284 + }, + { + "epoch": 0.23, + "grad_norm": 0.37890625, + "learning_rate": 0.00019401226839525638, + "loss": 2.5967, + "step": 5285 + }, + { + "epoch": 0.23, + "grad_norm": 0.462890625, + "learning_rate": 0.00019400998187761944, + "loss": 2.6222, + "step": 5286 + }, + { + "epoch": 0.23, + "grad_norm": 0.365234375, + "learning_rate": 0.00019400769493697198, + "loss": 2.6528, + "step": 5287 + }, + { + "epoch": 0.23, + "grad_norm": 0.419921875, + "learning_rate": 0.00019400540757332424, + "loss": 2.7653, + "step": 5288 + }, + { + "epoch": 0.23, + "grad_norm": 0.46484375, + "learning_rate": 0.0001940031197866865, + "loss": 2.8425, + "step": 5289 + }, + { + "epoch": 0.23, + "grad_norm": 0.4296875, + "learning_rate": 0.00019400083157706909, + "loss": 2.6088, + "step": 5290 + }, + { + "epoch": 0.23, + "grad_norm": 0.45703125, + "learning_rate": 0.00019399854294448226, + "loss": 2.6276, + "step": 5291 + }, + { + "epoch": 0.23, + "grad_norm": 0.458984375, + "learning_rate": 0.00019399625388893637, + "loss": 2.7539, + "step": 5292 + }, + { + "epoch": 0.23, + "grad_norm": 0.3984375, + "learning_rate": 0.0001939939644104417, + "loss": 2.6611, + "step": 5293 + }, + { + "epoch": 0.23, + "grad_norm": 0.455078125, + "learning_rate": 0.0001939916745090085, + "loss": 2.6206, + "step": 5294 + }, + { + "epoch": 0.23, + "grad_norm": 0.455078125, + "learning_rate": 0.00019398938418464714, + "loss": 2.686, + "step": 5295 + }, + { + "epoch": 0.23, + "grad_norm": 0.421875, + "learning_rate": 0.0001939870934373679, + "loss": 2.6967, + "step": 5296 + }, + { + "epoch": 0.23, + "grad_norm": 0.45703125, + "learning_rate": 0.00019398480226718106, + "loss": 2.6447, + "step": 5297 + }, + { + "epoch": 0.23, + "grad_norm": 0.41796875, + "learning_rate": 0.00019398251067409702, + "loss": 2.5653, + "step": 5298 + }, + { + "epoch": 0.23, + "grad_norm": 0.4453125, + "learning_rate": 0.00019398021865812596, + "loss": 2.6468, + "step": 5299 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.0001939779262192783, + "loss": 2.6645, + "step": 5300 + }, + { + "epoch": 0.23, + "grad_norm": 0.390625, + "learning_rate": 0.00019397563335756432, + "loss": 2.6456, + "step": 5301 + }, + { + "epoch": 0.23, + "grad_norm": 0.4609375, + "learning_rate": 0.00019397334007299432, + "loss": 2.4858, + "step": 5302 + }, + { + "epoch": 0.23, + "grad_norm": 0.365234375, + "learning_rate": 0.00019397104636557867, + "loss": 2.5467, + "step": 5303 + }, + { + "epoch": 0.23, + "grad_norm": 0.462890625, + "learning_rate": 0.0001939687522353276, + "loss": 2.7517, + "step": 5304 + }, + { + "epoch": 0.23, + "grad_norm": 0.478515625, + "learning_rate": 0.0001939664576822515, + "loss": 2.712, + "step": 5305 + }, + { + "epoch": 0.23, + "grad_norm": 0.494140625, + "learning_rate": 0.00019396416270636067, + "loss": 2.7058, + "step": 5306 + }, + { + "epoch": 0.23, + "grad_norm": 0.421875, + "learning_rate": 0.00019396186730766547, + "loss": 2.7484, + "step": 5307 + }, + { + "epoch": 0.23, + "grad_norm": 0.4140625, + "learning_rate": 0.0001939595714861762, + "loss": 2.5354, + "step": 5308 + }, + { + "epoch": 0.23, + "grad_norm": 0.416015625, + "learning_rate": 0.00019395727524190322, + "loss": 2.6044, + "step": 5309 + }, + { + "epoch": 0.23, + "grad_norm": 0.404296875, + "learning_rate": 0.00019395497857485681, + "loss": 2.4933, + "step": 5310 + }, + { + "epoch": 0.23, + "grad_norm": 0.46484375, + "learning_rate": 0.00019395268148504735, + "loss": 2.6922, + "step": 5311 + }, + { + "epoch": 0.23, + "grad_norm": 0.466796875, + "learning_rate": 0.00019395038397248515, + "loss": 2.6761, + "step": 5312 + }, + { + "epoch": 0.23, + "grad_norm": 0.470703125, + "learning_rate": 0.00019394808603718055, + "loss": 2.7506, + "step": 5313 + }, + { + "epoch": 0.23, + "grad_norm": 0.431640625, + "learning_rate": 0.00019394578767914392, + "loss": 2.6957, + "step": 5314 + }, + { + "epoch": 0.23, + "grad_norm": 0.46484375, + "learning_rate": 0.00019394348889838556, + "loss": 2.7083, + "step": 5315 + }, + { + "epoch": 0.23, + "grad_norm": 0.4375, + "learning_rate": 0.00019394118969491583, + "loss": 2.7547, + "step": 5316 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.0001939388900687451, + "loss": 2.7711, + "step": 5317 + }, + { + "epoch": 0.23, + "grad_norm": 0.5390625, + "learning_rate": 0.00019393659001988367, + "loss": 2.7876, + "step": 5318 + }, + { + "epoch": 0.23, + "grad_norm": 0.435546875, + "learning_rate": 0.00019393428954834192, + "loss": 2.5066, + "step": 5319 + }, + { + "epoch": 0.23, + "grad_norm": 0.419921875, + "learning_rate": 0.0001939319886541302, + "loss": 2.6491, + "step": 5320 + }, + { + "epoch": 0.23, + "grad_norm": 0.453125, + "learning_rate": 0.00019392968733725888, + "loss": 2.5988, + "step": 5321 + }, + { + "epoch": 0.23, + "grad_norm": 0.451171875, + "learning_rate": 0.00019392738559773824, + "loss": 2.5253, + "step": 5322 + }, + { + "epoch": 0.23, + "grad_norm": 0.484375, + "learning_rate": 0.00019392508343557874, + "loss": 2.6687, + "step": 5323 + }, + { + "epoch": 0.23, + "grad_norm": 0.4609375, + "learning_rate": 0.00019392278085079067, + "loss": 2.6303, + "step": 5324 + }, + { + "epoch": 0.23, + "grad_norm": 0.453125, + "learning_rate": 0.00019392047784338442, + "loss": 2.8063, + "step": 5325 + }, + { + "epoch": 0.23, + "grad_norm": 0.44921875, + "learning_rate": 0.00019391817441337034, + "loss": 2.5723, + "step": 5326 + }, + { + "epoch": 0.23, + "grad_norm": 0.453125, + "learning_rate": 0.0001939158705607588, + "loss": 2.6541, + "step": 5327 + }, + { + "epoch": 0.23, + "grad_norm": 0.494140625, + "learning_rate": 0.00019391356628556018, + "loss": 2.6479, + "step": 5328 + }, + { + "epoch": 0.23, + "grad_norm": 0.39453125, + "learning_rate": 0.0001939112615877848, + "loss": 2.5585, + "step": 5329 + }, + { + "epoch": 0.23, + "grad_norm": 0.46484375, + "learning_rate": 0.0001939089564674431, + "loss": 2.6433, + "step": 5330 + }, + { + "epoch": 0.23, + "grad_norm": 0.41015625, + "learning_rate": 0.0001939066509245454, + "loss": 2.7012, + "step": 5331 + }, + { + "epoch": 0.23, + "grad_norm": 0.466796875, + "learning_rate": 0.00019390434495910206, + "loss": 2.7428, + "step": 5332 + }, + { + "epoch": 0.23, + "grad_norm": 0.455078125, + "learning_rate": 0.0001939020385711235, + "loss": 2.7216, + "step": 5333 + }, + { + "epoch": 0.23, + "grad_norm": 0.4453125, + "learning_rate": 0.0001938997317606201, + "loss": 2.7864, + "step": 5334 + }, + { + "epoch": 0.23, + "grad_norm": 0.443359375, + "learning_rate": 0.00019389742452760223, + "loss": 2.6892, + "step": 5335 + }, + { + "epoch": 0.23, + "grad_norm": 0.416015625, + "learning_rate": 0.00019389511687208026, + "loss": 2.6056, + "step": 5336 + }, + { + "epoch": 0.23, + "grad_norm": 0.462890625, + "learning_rate": 0.00019389280879406458, + "loss": 2.7635, + "step": 5337 + }, + { + "epoch": 0.23, + "grad_norm": 0.443359375, + "learning_rate": 0.00019389050029356556, + "loss": 2.6809, + "step": 5338 + }, + { + "epoch": 0.23, + "grad_norm": 0.431640625, + "learning_rate": 0.0001938881913705936, + "loss": 2.6702, + "step": 5339 + }, + { + "epoch": 0.23, + "grad_norm": 0.443359375, + "learning_rate": 0.0001938858820251591, + "loss": 2.6828, + "step": 5340 + }, + { + "epoch": 0.23, + "grad_norm": 0.431640625, + "learning_rate": 0.00019388357225727244, + "loss": 2.6575, + "step": 5341 + }, + { + "epoch": 0.23, + "grad_norm": 0.431640625, + "learning_rate": 0.00019388126206694401, + "loss": 2.7667, + "step": 5342 + }, + { + "epoch": 0.23, + "grad_norm": 0.3984375, + "learning_rate": 0.00019387895145418421, + "loss": 2.7083, + "step": 5343 + }, + { + "epoch": 0.23, + "grad_norm": 0.462890625, + "learning_rate": 0.00019387664041900347, + "loss": 2.66, + "step": 5344 + }, + { + "epoch": 0.23, + "grad_norm": 0.453125, + "learning_rate": 0.00019387432896141214, + "loss": 2.5842, + "step": 5345 + }, + { + "epoch": 0.23, + "grad_norm": 0.4296875, + "learning_rate": 0.0001938720170814206, + "loss": 2.587, + "step": 5346 + }, + { + "epoch": 0.23, + "grad_norm": 0.474609375, + "learning_rate": 0.00019386970477903935, + "loss": 2.542, + "step": 5347 + }, + { + "epoch": 0.23, + "grad_norm": 0.408203125, + "learning_rate": 0.00019386739205427866, + "loss": 2.5783, + "step": 5348 + }, + { + "epoch": 0.23, + "grad_norm": 0.44140625, + "learning_rate": 0.00019386507890714906, + "loss": 2.5947, + "step": 5349 + }, + { + "epoch": 0.23, + "grad_norm": 0.39453125, + "learning_rate": 0.00019386276533766092, + "loss": 2.686, + "step": 5350 + }, + { + "epoch": 0.23, + "grad_norm": 0.490234375, + "learning_rate": 0.0001938604513458246, + "loss": 2.585, + "step": 5351 + }, + { + "epoch": 0.23, + "grad_norm": 0.46484375, + "learning_rate": 0.00019385813693165057, + "loss": 2.6336, + "step": 5352 + }, + { + "epoch": 0.23, + "grad_norm": 0.44140625, + "learning_rate": 0.00019385582209514922, + "loss": 2.6679, + "step": 5353 + }, + { + "epoch": 0.23, + "grad_norm": 0.4296875, + "learning_rate": 0.00019385350683633098, + "loss": 2.73, + "step": 5354 + }, + { + "epoch": 0.23, + "grad_norm": 0.376953125, + "learning_rate": 0.00019385119115520626, + "loss": 2.6778, + "step": 5355 + }, + { + "epoch": 0.23, + "grad_norm": 0.423828125, + "learning_rate": 0.00019384887505178547, + "loss": 2.5917, + "step": 5356 + }, + { + "epoch": 0.23, + "grad_norm": 0.482421875, + "learning_rate": 0.00019384655852607905, + "loss": 2.876, + "step": 5357 + }, + { + "epoch": 0.23, + "grad_norm": 0.42578125, + "learning_rate": 0.0001938442415780974, + "loss": 2.6385, + "step": 5358 + }, + { + "epoch": 0.23, + "grad_norm": 0.439453125, + "learning_rate": 0.000193841924207851, + "loss": 2.5473, + "step": 5359 + }, + { + "epoch": 0.23, + "grad_norm": 0.435546875, + "learning_rate": 0.00019383960641535018, + "loss": 2.5511, + "step": 5360 + }, + { + "epoch": 0.23, + "grad_norm": 0.5, + "learning_rate": 0.00019383728820060545, + "loss": 2.6521, + "step": 5361 + }, + { + "epoch": 0.23, + "grad_norm": 0.423828125, + "learning_rate": 0.0001938349695636272, + "loss": 2.5474, + "step": 5362 + }, + { + "epoch": 0.23, + "grad_norm": 0.42578125, + "learning_rate": 0.0001938326505044259, + "loss": 2.6357, + "step": 5363 + }, + { + "epoch": 0.23, + "grad_norm": 0.427734375, + "learning_rate": 0.00019383033102301196, + "loss": 2.6083, + "step": 5364 + }, + { + "epoch": 0.23, + "grad_norm": 0.375, + "learning_rate": 0.00019382801111939582, + "loss": 2.7104, + "step": 5365 + }, + { + "epoch": 0.23, + "grad_norm": 0.404296875, + "learning_rate": 0.00019382569079358792, + "loss": 2.5812, + "step": 5366 + }, + { + "epoch": 0.23, + "grad_norm": 0.42578125, + "learning_rate": 0.0001938233700455987, + "loss": 2.5321, + "step": 5367 + }, + { + "epoch": 0.23, + "grad_norm": 0.431640625, + "learning_rate": 0.0001938210488754386, + "loss": 2.6043, + "step": 5368 + }, + { + "epoch": 0.23, + "grad_norm": 0.474609375, + "learning_rate": 0.0001938187272831181, + "loss": 2.8453, + "step": 5369 + }, + { + "epoch": 0.23, + "grad_norm": 0.4296875, + "learning_rate": 0.00019381640526864758, + "loss": 2.6836, + "step": 5370 + }, + { + "epoch": 0.23, + "grad_norm": 0.38671875, + "learning_rate": 0.00019381408283203753, + "loss": 2.6362, + "step": 5371 + }, + { + "epoch": 0.23, + "grad_norm": 0.4296875, + "learning_rate": 0.00019381175997329838, + "loss": 2.6281, + "step": 5372 + }, + { + "epoch": 0.23, + "grad_norm": 0.44921875, + "learning_rate": 0.0001938094366924406, + "loss": 2.6557, + "step": 5373 + }, + { + "epoch": 0.23, + "grad_norm": 0.404296875, + "learning_rate": 0.00019380711298947465, + "loss": 2.554, + "step": 5374 + }, + { + "epoch": 0.23, + "grad_norm": 0.41015625, + "learning_rate": 0.00019380478886441094, + "loss": 2.4728, + "step": 5375 + }, + { + "epoch": 0.23, + "grad_norm": 0.447265625, + "learning_rate": 0.00019380246431726001, + "loss": 2.6583, + "step": 5376 + }, + { + "epoch": 0.23, + "grad_norm": 0.46875, + "learning_rate": 0.00019380013934803224, + "loss": 2.6649, + "step": 5377 + }, + { + "epoch": 0.23, + "grad_norm": 0.47265625, + "learning_rate": 0.00019379781395673814, + "loss": 2.7822, + "step": 5378 + }, + { + "epoch": 0.23, + "grad_norm": 0.4921875, + "learning_rate": 0.00019379548814338818, + "loss": 2.6078, + "step": 5379 + }, + { + "epoch": 0.23, + "grad_norm": 0.453125, + "learning_rate": 0.00019379316190799278, + "loss": 2.7305, + "step": 5380 + }, + { + "epoch": 0.23, + "grad_norm": 0.419921875, + "learning_rate": 0.0001937908352505624, + "loss": 2.6104, + "step": 5381 + }, + { + "epoch": 0.23, + "grad_norm": 0.494140625, + "learning_rate": 0.00019378850817110754, + "loss": 2.7933, + "step": 5382 + }, + { + "epoch": 0.23, + "grad_norm": 0.380859375, + "learning_rate": 0.0001937861806696387, + "loss": 2.613, + "step": 5383 + }, + { + "epoch": 0.23, + "grad_norm": 0.44921875, + "learning_rate": 0.0001937838527461663, + "loss": 2.611, + "step": 5384 + }, + { + "epoch": 0.23, + "grad_norm": 0.412109375, + "learning_rate": 0.00019378152440070086, + "loss": 2.6099, + "step": 5385 + }, + { + "epoch": 0.23, + "grad_norm": 0.5078125, + "learning_rate": 0.00019377919563325283, + "loss": 2.6938, + "step": 5386 + }, + { + "epoch": 0.23, + "grad_norm": 0.408203125, + "learning_rate": 0.0001937768664438327, + "loss": 2.7168, + "step": 5387 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.0001937745368324509, + "loss": 2.7112, + "step": 5388 + }, + { + "epoch": 0.23, + "grad_norm": 0.48828125, + "learning_rate": 0.000193772206799118, + "loss": 2.7095, + "step": 5389 + }, + { + "epoch": 0.23, + "grad_norm": 0.466796875, + "learning_rate": 0.0001937698763438444, + "loss": 2.5741, + "step": 5390 + }, + { + "epoch": 0.23, + "grad_norm": 0.39453125, + "learning_rate": 0.00019376754546664065, + "loss": 2.6385, + "step": 5391 + }, + { + "epoch": 0.23, + "grad_norm": 0.40234375, + "learning_rate": 0.0001937652141675172, + "loss": 2.6599, + "step": 5392 + }, + { + "epoch": 0.23, + "grad_norm": 0.439453125, + "learning_rate": 0.00019376288244648456, + "loss": 2.7743, + "step": 5393 + }, + { + "epoch": 0.23, + "grad_norm": 0.435546875, + "learning_rate": 0.00019376055030355323, + "loss": 2.6662, + "step": 5394 + }, + { + "epoch": 0.23, + "grad_norm": 0.4765625, + "learning_rate": 0.00019375821773873368, + "loss": 2.6451, + "step": 5395 + }, + { + "epoch": 0.23, + "grad_norm": 0.46875, + "learning_rate": 0.00019375588475203638, + "loss": 2.5813, + "step": 5396 + }, + { + "epoch": 0.23, + "grad_norm": 0.462890625, + "learning_rate": 0.00019375355134347192, + "loss": 2.686, + "step": 5397 + }, + { + "epoch": 0.23, + "grad_norm": 0.4921875, + "learning_rate": 0.0001937512175130507, + "loss": 2.7769, + "step": 5398 + }, + { + "epoch": 0.23, + "grad_norm": 0.4453125, + "learning_rate": 0.0001937488832607833, + "loss": 2.7035, + "step": 5399 + }, + { + "epoch": 0.23, + "grad_norm": 0.4140625, + "learning_rate": 0.00019374654858668016, + "loss": 2.4409, + "step": 5400 + }, + { + "epoch": 0.23, + "grad_norm": 0.46484375, + "learning_rate": 0.0001937442134907518, + "loss": 2.7389, + "step": 5401 + }, + { + "epoch": 0.23, + "grad_norm": 0.458984375, + "learning_rate": 0.0001937418779730088, + "loss": 2.66, + "step": 5402 + }, + { + "epoch": 0.23, + "grad_norm": 0.396484375, + "learning_rate": 0.00019373954203346153, + "loss": 2.6017, + "step": 5403 + }, + { + "epoch": 0.23, + "grad_norm": 0.46484375, + "learning_rate": 0.00019373720567212062, + "loss": 2.6291, + "step": 5404 + }, + { + "epoch": 0.23, + "grad_norm": 0.44140625, + "learning_rate": 0.00019373486888899652, + "loss": 2.7862, + "step": 5405 + }, + { + "epoch": 0.23, + "grad_norm": 0.4140625, + "learning_rate": 0.0001937325316840998, + "loss": 2.6925, + "step": 5406 + }, + { + "epoch": 0.23, + "grad_norm": 0.447265625, + "learning_rate": 0.0001937301940574409, + "loss": 2.6346, + "step": 5407 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.0001937278560090304, + "loss": 2.691, + "step": 5408 + }, + { + "epoch": 0.23, + "grad_norm": 0.451171875, + "learning_rate": 0.0001937255175388788, + "loss": 2.5507, + "step": 5409 + }, + { + "epoch": 0.23, + "grad_norm": 0.44921875, + "learning_rate": 0.00019372317864699664, + "loss": 2.6825, + "step": 5410 + }, + { + "epoch": 0.23, + "grad_norm": 0.46484375, + "learning_rate": 0.00019372083933339444, + "loss": 2.6318, + "step": 5411 + }, + { + "epoch": 0.23, + "grad_norm": 0.54296875, + "learning_rate": 0.0001937184995980827, + "loss": 2.7577, + "step": 5412 + }, + { + "epoch": 0.23, + "grad_norm": 0.484375, + "learning_rate": 0.00019371615944107194, + "loss": 2.651, + "step": 5413 + }, + { + "epoch": 0.23, + "grad_norm": 0.451171875, + "learning_rate": 0.00019371381886237272, + "loss": 2.6553, + "step": 5414 + }, + { + "epoch": 0.23, + "grad_norm": 0.470703125, + "learning_rate": 0.00019371147786199558, + "loss": 2.5225, + "step": 5415 + }, + { + "epoch": 0.23, + "grad_norm": 0.45703125, + "learning_rate": 0.00019370913643995102, + "loss": 2.6908, + "step": 5416 + }, + { + "epoch": 0.23, + "grad_norm": 0.439453125, + "learning_rate": 0.0001937067945962496, + "loss": 2.7382, + "step": 5417 + }, + { + "epoch": 0.23, + "grad_norm": 0.419921875, + "learning_rate": 0.00019370445233090185, + "loss": 2.7576, + "step": 5418 + }, + { + "epoch": 0.23, + "grad_norm": 0.5, + "learning_rate": 0.00019370210964391832, + "loss": 2.6266, + "step": 5419 + }, + { + "epoch": 0.23, + "grad_norm": 0.41796875, + "learning_rate": 0.00019369976653530951, + "loss": 2.601, + "step": 5420 + }, + { + "epoch": 0.23, + "grad_norm": 0.4140625, + "learning_rate": 0.000193697423005086, + "loss": 2.5326, + "step": 5421 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.00019369507905325837, + "loss": 2.6592, + "step": 5422 + }, + { + "epoch": 0.23, + "grad_norm": 0.45703125, + "learning_rate": 0.0001936927346798371, + "loss": 2.7059, + "step": 5423 + }, + { + "epoch": 0.23, + "grad_norm": 0.421875, + "learning_rate": 0.00019369038988483275, + "loss": 2.6148, + "step": 5424 + }, + { + "epoch": 0.23, + "grad_norm": 0.4609375, + "learning_rate": 0.0001936880446682559, + "loss": 2.6808, + "step": 5425 + }, + { + "epoch": 0.23, + "grad_norm": 0.435546875, + "learning_rate": 0.00019368569903011707, + "loss": 2.7466, + "step": 5426 + }, + { + "epoch": 0.23, + "grad_norm": 0.435546875, + "learning_rate": 0.00019368335297042686, + "loss": 2.7144, + "step": 5427 + }, + { + "epoch": 0.23, + "grad_norm": 0.392578125, + "learning_rate": 0.0001936810064891958, + "loss": 2.5255, + "step": 5428 + }, + { + "epoch": 0.23, + "grad_norm": 0.431640625, + "learning_rate": 0.00019367865958643443, + "loss": 2.6688, + "step": 5429 + }, + { + "epoch": 0.23, + "grad_norm": 0.44921875, + "learning_rate": 0.00019367631226215336, + "loss": 2.5496, + "step": 5430 + }, + { + "epoch": 0.23, + "grad_norm": 0.400390625, + "learning_rate": 0.00019367396451636307, + "loss": 2.62, + "step": 5431 + }, + { + "epoch": 0.23, + "grad_norm": 0.455078125, + "learning_rate": 0.0001936716163490742, + "loss": 2.5964, + "step": 5432 + }, + { + "epoch": 0.23, + "grad_norm": 0.44921875, + "learning_rate": 0.00019366926776029727, + "loss": 2.7265, + "step": 5433 + }, + { + "epoch": 0.23, + "grad_norm": 0.421875, + "learning_rate": 0.0001936669187500429, + "loss": 2.581, + "step": 5434 + }, + { + "epoch": 0.23, + "grad_norm": 0.40234375, + "learning_rate": 0.0001936645693183216, + "loss": 2.6586, + "step": 5435 + }, + { + "epoch": 0.23, + "grad_norm": 0.412109375, + "learning_rate": 0.000193662219465144, + "loss": 2.698, + "step": 5436 + }, + { + "epoch": 0.23, + "grad_norm": 0.359375, + "learning_rate": 0.0001936598691905206, + "loss": 2.5393, + "step": 5437 + }, + { + "epoch": 0.23, + "grad_norm": 0.380859375, + "learning_rate": 0.00019365751849446207, + "loss": 2.4387, + "step": 5438 + }, + { + "epoch": 0.23, + "grad_norm": 0.443359375, + "learning_rate": 0.00019365516737697886, + "loss": 2.6288, + "step": 5439 + }, + { + "epoch": 0.23, + "grad_norm": 0.49609375, + "learning_rate": 0.00019365281583808168, + "loss": 2.7956, + "step": 5440 + }, + { + "epoch": 0.23, + "grad_norm": 0.498046875, + "learning_rate": 0.00019365046387778104, + "loss": 2.7434, + "step": 5441 + }, + { + "epoch": 0.23, + "grad_norm": 0.435546875, + "learning_rate": 0.00019364811149608755, + "loss": 2.6338, + "step": 5442 + }, + { + "epoch": 0.23, + "grad_norm": 0.404296875, + "learning_rate": 0.00019364575869301178, + "loss": 2.6804, + "step": 5443 + }, + { + "epoch": 0.23, + "grad_norm": 0.44140625, + "learning_rate": 0.0001936434054685643, + "loss": 2.6349, + "step": 5444 + }, + { + "epoch": 0.23, + "grad_norm": 0.455078125, + "learning_rate": 0.00019364105182275573, + "loss": 2.6443, + "step": 5445 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.00019363869775559665, + "loss": 2.736, + "step": 5446 + }, + { + "epoch": 0.23, + "grad_norm": 0.419921875, + "learning_rate": 0.00019363634326709765, + "loss": 2.601, + "step": 5447 + }, + { + "epoch": 0.23, + "grad_norm": 0.404296875, + "learning_rate": 0.0001936339883572693, + "loss": 2.5095, + "step": 5448 + }, + { + "epoch": 0.23, + "grad_norm": 0.453125, + "learning_rate": 0.00019363163302612226, + "loss": 2.7115, + "step": 5449 + }, + { + "epoch": 0.23, + "grad_norm": 0.451171875, + "learning_rate": 0.0001936292772736671, + "loss": 2.7373, + "step": 5450 + }, + { + "epoch": 0.23, + "grad_norm": 0.453125, + "learning_rate": 0.00019362692109991438, + "loss": 2.6174, + "step": 5451 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.00019362456450487474, + "loss": 2.7297, + "step": 5452 + }, + { + "epoch": 0.23, + "grad_norm": 0.447265625, + "learning_rate": 0.00019362220748855876, + "loss": 2.6571, + "step": 5453 + }, + { + "epoch": 0.23, + "grad_norm": 0.396484375, + "learning_rate": 0.00019361985005097708, + "loss": 2.5214, + "step": 5454 + }, + { + "epoch": 0.23, + "grad_norm": 0.40625, + "learning_rate": 0.00019361749219214028, + "loss": 2.6017, + "step": 5455 + }, + { + "epoch": 0.23, + "grad_norm": 0.419921875, + "learning_rate": 0.000193615133912059, + "loss": 2.4493, + "step": 5456 + }, + { + "epoch": 0.23, + "grad_norm": 0.404296875, + "learning_rate": 0.00019361277521074379, + "loss": 2.6749, + "step": 5457 + }, + { + "epoch": 0.23, + "grad_norm": 0.3828125, + "learning_rate": 0.00019361041608820532, + "loss": 2.503, + "step": 5458 + }, + { + "epoch": 0.23, + "grad_norm": 0.392578125, + "learning_rate": 0.00019360805654445418, + "loss": 2.6008, + "step": 5459 + }, + { + "epoch": 0.23, + "grad_norm": 0.4375, + "learning_rate": 0.00019360569657950104, + "loss": 2.881, + "step": 5460 + }, + { + "epoch": 0.23, + "grad_norm": 0.404296875, + "learning_rate": 0.00019360333619335642, + "loss": 2.5269, + "step": 5461 + }, + { + "epoch": 0.23, + "grad_norm": 0.439453125, + "learning_rate": 0.00019360097538603105, + "loss": 2.5784, + "step": 5462 + }, + { + "epoch": 0.23, + "grad_norm": 0.416015625, + "learning_rate": 0.00019359861415753544, + "loss": 2.7485, + "step": 5463 + }, + { + "epoch": 0.23, + "grad_norm": 0.474609375, + "learning_rate": 0.0001935962525078803, + "loss": 2.6457, + "step": 5464 + }, + { + "epoch": 0.23, + "grad_norm": 0.40234375, + "learning_rate": 0.00019359389043707625, + "loss": 2.6634, + "step": 5465 + }, + { + "epoch": 0.23, + "grad_norm": 0.412109375, + "learning_rate": 0.00019359152794513386, + "loss": 2.5604, + "step": 5466 + }, + { + "epoch": 0.23, + "grad_norm": 0.478515625, + "learning_rate": 0.0001935891650320638, + "loss": 2.7421, + "step": 5467 + }, + { + "epoch": 0.23, + "grad_norm": 0.380859375, + "learning_rate": 0.00019358680169787674, + "loss": 2.6227, + "step": 5468 + }, + { + "epoch": 0.23, + "grad_norm": 0.396484375, + "learning_rate": 0.00019358443794258325, + "loss": 2.6857, + "step": 5469 + }, + { + "epoch": 0.23, + "grad_norm": 0.435546875, + "learning_rate": 0.00019358207376619398, + "loss": 2.5933, + "step": 5470 + }, + { + "epoch": 0.23, + "grad_norm": 0.478515625, + "learning_rate": 0.00019357970916871958, + "loss": 2.7668, + "step": 5471 + }, + { + "epoch": 0.23, + "grad_norm": 0.44921875, + "learning_rate": 0.00019357734415017072, + "loss": 2.6886, + "step": 5472 + }, + { + "epoch": 0.23, + "grad_norm": 0.4609375, + "learning_rate": 0.00019357497871055798, + "loss": 2.5986, + "step": 5473 + }, + { + "epoch": 0.23, + "grad_norm": 0.4453125, + "learning_rate": 0.00019357261284989202, + "loss": 2.5655, + "step": 5474 + }, + { + "epoch": 0.23, + "grad_norm": 0.4375, + "learning_rate": 0.00019357024656818355, + "loss": 2.6899, + "step": 5475 + }, + { + "epoch": 0.23, + "grad_norm": 0.453125, + "learning_rate": 0.00019356787986544313, + "loss": 2.7082, + "step": 5476 + }, + { + "epoch": 0.23, + "grad_norm": 0.45703125, + "learning_rate": 0.00019356551274168145, + "loss": 2.6057, + "step": 5477 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.00019356314519690913, + "loss": 2.5877, + "step": 5478 + }, + { + "epoch": 0.23, + "grad_norm": 0.4296875, + "learning_rate": 0.0001935607772311369, + "loss": 2.6712, + "step": 5479 + }, + { + "epoch": 0.23, + "grad_norm": 0.4453125, + "learning_rate": 0.00019355840884437533, + "loss": 2.659, + "step": 5480 + }, + { + "epoch": 0.23, + "grad_norm": 0.443359375, + "learning_rate": 0.00019355604003663514, + "loss": 2.6358, + "step": 5481 + }, + { + "epoch": 0.23, + "grad_norm": 0.458984375, + "learning_rate": 0.00019355367080792696, + "loss": 2.7135, + "step": 5482 + }, + { + "epoch": 0.23, + "grad_norm": 0.4453125, + "learning_rate": 0.00019355130115826145, + "loss": 2.6251, + "step": 5483 + }, + { + "epoch": 0.23, + "grad_norm": 0.4453125, + "learning_rate": 0.0001935489310876492, + "loss": 2.6489, + "step": 5484 + }, + { + "epoch": 0.23, + "grad_norm": 0.43359375, + "learning_rate": 0.00019354656059610104, + "loss": 2.7152, + "step": 5485 + }, + { + "epoch": 0.23, + "grad_norm": 0.484375, + "learning_rate": 0.0001935441896836275, + "loss": 2.6768, + "step": 5486 + }, + { + "epoch": 0.23, + "grad_norm": 0.447265625, + "learning_rate": 0.00019354181835023933, + "loss": 2.7, + "step": 5487 + }, + { + "epoch": 0.23, + "grad_norm": 0.427734375, + "learning_rate": 0.0001935394465959471, + "loss": 2.7949, + "step": 5488 + }, + { + "epoch": 0.23, + "grad_norm": 0.396484375, + "learning_rate": 0.0001935370744207616, + "loss": 2.6118, + "step": 5489 + }, + { + "epoch": 0.23, + "grad_norm": 0.443359375, + "learning_rate": 0.00019353470182469345, + "loss": 2.767, + "step": 5490 + }, + { + "epoch": 0.23, + "grad_norm": 0.408203125, + "learning_rate": 0.0001935323288077533, + "loss": 2.6361, + "step": 5491 + }, + { + "epoch": 0.23, + "grad_norm": 0.443359375, + "learning_rate": 0.00019352995536995186, + "loss": 2.7086, + "step": 5492 + }, + { + "epoch": 0.23, + "grad_norm": 0.490234375, + "learning_rate": 0.00019352758151129979, + "loss": 2.8072, + "step": 5493 + }, + { + "epoch": 0.23, + "grad_norm": 0.486328125, + "learning_rate": 0.0001935252072318078, + "loss": 2.6676, + "step": 5494 + }, + { + "epoch": 0.23, + "grad_norm": 0.419921875, + "learning_rate": 0.00019352283253148657, + "loss": 2.5057, + "step": 5495 + }, + { + "epoch": 0.23, + "grad_norm": 0.486328125, + "learning_rate": 0.00019352045741034677, + "loss": 2.7301, + "step": 5496 + }, + { + "epoch": 0.23, + "grad_norm": 0.4296875, + "learning_rate": 0.00019351808186839905, + "loss": 2.7118, + "step": 5497 + }, + { + "epoch": 0.23, + "grad_norm": 0.447265625, + "learning_rate": 0.00019351570590565417, + "loss": 2.6576, + "step": 5498 + }, + { + "epoch": 0.23, + "grad_norm": 0.45703125, + "learning_rate": 0.0001935133295221228, + "loss": 2.723, + "step": 5499 + }, + { + "epoch": 0.23, + "grad_norm": 0.431640625, + "learning_rate": 0.00019351095271781562, + "loss": 2.6902, + "step": 5500 + }, + { + "epoch": 0.23, + "grad_norm": 0.408203125, + "learning_rate": 0.00019350857549274328, + "loss": 2.6443, + "step": 5501 + }, + { + "epoch": 0.23, + "grad_norm": 0.46875, + "learning_rate": 0.00019350619784691656, + "loss": 2.6104, + "step": 5502 + }, + { + "epoch": 0.23, + "grad_norm": 0.44140625, + "learning_rate": 0.00019350381978034613, + "loss": 2.6787, + "step": 5503 + }, + { + "epoch": 0.23, + "grad_norm": 0.4765625, + "learning_rate": 0.00019350144129304267, + "loss": 2.7194, + "step": 5504 + }, + { + "epoch": 0.23, + "grad_norm": 0.41796875, + "learning_rate": 0.00019349906238501688, + "loss": 2.6607, + "step": 5505 + }, + { + "epoch": 0.23, + "grad_norm": 0.3828125, + "learning_rate": 0.0001934966830562795, + "loss": 2.658, + "step": 5506 + }, + { + "epoch": 0.23, + "grad_norm": 0.447265625, + "learning_rate": 0.0001934943033068412, + "loss": 2.7133, + "step": 5507 + }, + { + "epoch": 0.23, + "grad_norm": 0.421875, + "learning_rate": 0.00019349192313671272, + "loss": 2.593, + "step": 5508 + }, + { + "epoch": 0.23, + "grad_norm": 0.47265625, + "learning_rate": 0.00019348954254590477, + "loss": 2.6628, + "step": 5509 + }, + { + "epoch": 0.23, + "grad_norm": 0.4375, + "learning_rate": 0.00019348716153442803, + "loss": 2.7697, + "step": 5510 + }, + { + "epoch": 0.23, + "grad_norm": 0.51171875, + "learning_rate": 0.0001934847801022932, + "loss": 2.699, + "step": 5511 + }, + { + "epoch": 0.23, + "grad_norm": 0.47265625, + "learning_rate": 0.00019348239824951103, + "loss": 2.6937, + "step": 5512 + }, + { + "epoch": 0.23, + "grad_norm": 0.3359375, + "learning_rate": 0.00019348001597609227, + "loss": 2.4839, + "step": 5513 + }, + { + "epoch": 0.23, + "grad_norm": 0.443359375, + "learning_rate": 0.00019347763328204757, + "loss": 2.5778, + "step": 5514 + }, + { + "epoch": 0.24, + "grad_norm": 0.462890625, + "learning_rate": 0.0001934752501673877, + "loss": 2.6687, + "step": 5515 + }, + { + "epoch": 0.24, + "grad_norm": 0.46484375, + "learning_rate": 0.00019347286663212336, + "loss": 2.7201, + "step": 5516 + }, + { + "epoch": 0.24, + "grad_norm": 0.408203125, + "learning_rate": 0.00019347048267626525, + "loss": 2.4258, + "step": 5517 + }, + { + "epoch": 0.24, + "grad_norm": 0.43359375, + "learning_rate": 0.00019346809829982417, + "loss": 2.5976, + "step": 5518 + }, + { + "epoch": 0.24, + "grad_norm": 0.4140625, + "learning_rate": 0.0001934657135028108, + "loss": 2.5623, + "step": 5519 + }, + { + "epoch": 0.24, + "grad_norm": 0.4140625, + "learning_rate": 0.00019346332828523588, + "loss": 2.6097, + "step": 5520 + }, + { + "epoch": 0.24, + "grad_norm": 0.453125, + "learning_rate": 0.00019346094264711012, + "loss": 2.6274, + "step": 5521 + }, + { + "epoch": 0.24, + "grad_norm": 0.427734375, + "learning_rate": 0.00019345855658844428, + "loss": 2.5989, + "step": 5522 + }, + { + "epoch": 0.24, + "grad_norm": 0.41796875, + "learning_rate": 0.0001934561701092491, + "loss": 2.6873, + "step": 5523 + }, + { + "epoch": 0.24, + "grad_norm": 0.390625, + "learning_rate": 0.0001934537832095353, + "loss": 2.6657, + "step": 5524 + }, + { + "epoch": 0.24, + "grad_norm": 0.376953125, + "learning_rate": 0.00019345139588931363, + "loss": 2.6153, + "step": 5525 + }, + { + "epoch": 0.24, + "grad_norm": 0.421875, + "learning_rate": 0.00019344900814859483, + "loss": 2.7198, + "step": 5526 + }, + { + "epoch": 0.24, + "grad_norm": 0.439453125, + "learning_rate": 0.00019344661998738964, + "loss": 2.6893, + "step": 5527 + }, + { + "epoch": 0.24, + "grad_norm": 0.45703125, + "learning_rate": 0.00019344423140570883, + "loss": 2.6037, + "step": 5528 + }, + { + "epoch": 0.24, + "grad_norm": 0.41015625, + "learning_rate": 0.00019344184240356312, + "loss": 2.5987, + "step": 5529 + }, + { + "epoch": 0.24, + "grad_norm": 0.38671875, + "learning_rate": 0.00019343945298096323, + "loss": 2.5737, + "step": 5530 + }, + { + "epoch": 0.24, + "grad_norm": 0.44921875, + "learning_rate": 0.00019343706313792, + "loss": 2.7221, + "step": 5531 + }, + { + "epoch": 0.24, + "grad_norm": 0.4375, + "learning_rate": 0.00019343467287444416, + "loss": 2.6498, + "step": 5532 + }, + { + "epoch": 0.24, + "grad_norm": 0.50390625, + "learning_rate": 0.00019343228219054638, + "loss": 2.6886, + "step": 5533 + }, + { + "epoch": 0.24, + "grad_norm": 0.4765625, + "learning_rate": 0.0001934298910862375, + "loss": 2.5727, + "step": 5534 + }, + { + "epoch": 0.24, + "grad_norm": 0.494140625, + "learning_rate": 0.00019342749956152822, + "loss": 2.6649, + "step": 5535 + }, + { + "epoch": 0.24, + "grad_norm": 0.451171875, + "learning_rate": 0.00019342510761642938, + "loss": 2.6363, + "step": 5536 + }, + { + "epoch": 0.24, + "grad_norm": 0.423828125, + "learning_rate": 0.00019342271525095165, + "loss": 2.7084, + "step": 5537 + }, + { + "epoch": 0.24, + "grad_norm": 0.40625, + "learning_rate": 0.0001934203224651059, + "loss": 2.531, + "step": 5538 + }, + { + "epoch": 0.24, + "grad_norm": 0.482421875, + "learning_rate": 0.0001934179292589028, + "loss": 2.7077, + "step": 5539 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.00019341553563235317, + "loss": 2.7314, + "step": 5540 + }, + { + "epoch": 0.24, + "grad_norm": 0.421875, + "learning_rate": 0.00019341314158546777, + "loss": 2.6529, + "step": 5541 + }, + { + "epoch": 0.24, + "grad_norm": 0.423828125, + "learning_rate": 0.0001934107471182574, + "loss": 2.7574, + "step": 5542 + }, + { + "epoch": 0.24, + "grad_norm": 0.484375, + "learning_rate": 0.00019340835223073278, + "loss": 2.5817, + "step": 5543 + }, + { + "epoch": 0.24, + "grad_norm": 0.427734375, + "learning_rate": 0.0001934059569229047, + "loss": 2.6859, + "step": 5544 + }, + { + "epoch": 0.24, + "grad_norm": 0.474609375, + "learning_rate": 0.00019340356119478398, + "loss": 2.7029, + "step": 5545 + }, + { + "epoch": 0.24, + "grad_norm": 0.42578125, + "learning_rate": 0.00019340116504638135, + "loss": 2.6744, + "step": 5546 + }, + { + "epoch": 0.24, + "grad_norm": 0.419921875, + "learning_rate": 0.00019339876847770758, + "loss": 2.6597, + "step": 5547 + }, + { + "epoch": 0.24, + "grad_norm": 0.451171875, + "learning_rate": 0.00019339637148877353, + "loss": 2.5533, + "step": 5548 + }, + { + "epoch": 0.24, + "grad_norm": 0.40234375, + "learning_rate": 0.00019339397407958995, + "loss": 2.5424, + "step": 5549 + }, + { + "epoch": 0.24, + "grad_norm": 0.3671875, + "learning_rate": 0.0001933915762501676, + "loss": 2.4543, + "step": 5550 + }, + { + "epoch": 0.24, + "grad_norm": 0.46484375, + "learning_rate": 0.00019338917800051727, + "loss": 2.5924, + "step": 5551 + }, + { + "epoch": 0.24, + "grad_norm": 0.46484375, + "learning_rate": 0.00019338677933064977, + "loss": 2.611, + "step": 5552 + }, + { + "epoch": 0.24, + "grad_norm": 0.412109375, + "learning_rate": 0.00019338438024057593, + "loss": 2.7304, + "step": 5553 + }, + { + "epoch": 0.24, + "grad_norm": 0.451171875, + "learning_rate": 0.00019338198073030647, + "loss": 2.791, + "step": 5554 + }, + { + "epoch": 0.24, + "grad_norm": 0.4609375, + "learning_rate": 0.00019337958079985223, + "loss": 2.77, + "step": 5555 + }, + { + "epoch": 0.24, + "grad_norm": 0.4375, + "learning_rate": 0.000193377180449224, + "loss": 2.676, + "step": 5556 + }, + { + "epoch": 0.24, + "grad_norm": 0.39453125, + "learning_rate": 0.0001933747796784326, + "loss": 2.7064, + "step": 5557 + }, + { + "epoch": 0.24, + "grad_norm": 0.443359375, + "learning_rate": 0.0001933723784874888, + "loss": 2.6118, + "step": 5558 + }, + { + "epoch": 0.24, + "grad_norm": 0.361328125, + "learning_rate": 0.00019336997687640346, + "loss": 2.4736, + "step": 5559 + }, + { + "epoch": 0.24, + "grad_norm": 0.47265625, + "learning_rate": 0.0001933675748451873, + "loss": 2.6964, + "step": 5560 + }, + { + "epoch": 0.24, + "grad_norm": 0.404296875, + "learning_rate": 0.00019336517239385118, + "loss": 2.459, + "step": 5561 + }, + { + "epoch": 0.24, + "grad_norm": 0.42578125, + "learning_rate": 0.00019336276952240593, + "loss": 2.697, + "step": 5562 + }, + { + "epoch": 0.24, + "grad_norm": 0.39453125, + "learning_rate": 0.00019336036623086235, + "loss": 2.5948, + "step": 5563 + }, + { + "epoch": 0.24, + "grad_norm": 0.4609375, + "learning_rate": 0.00019335796251923116, + "loss": 2.6314, + "step": 5564 + }, + { + "epoch": 0.24, + "grad_norm": 0.412109375, + "learning_rate": 0.00019335555838752332, + "loss": 2.7541, + "step": 5565 + }, + { + "epoch": 0.24, + "grad_norm": 0.4453125, + "learning_rate": 0.0001933531538357496, + "loss": 2.7228, + "step": 5566 + }, + { + "epoch": 0.24, + "grad_norm": 0.44140625, + "learning_rate": 0.00019335074886392078, + "loss": 2.7152, + "step": 5567 + }, + { + "epoch": 0.24, + "grad_norm": 0.4296875, + "learning_rate": 0.00019334834347204768, + "loss": 2.5657, + "step": 5568 + }, + { + "epoch": 0.24, + "grad_norm": 0.482421875, + "learning_rate": 0.0001933459376601412, + "loss": 2.7242, + "step": 5569 + }, + { + "epoch": 0.24, + "grad_norm": 0.408203125, + "learning_rate": 0.0001933435314282121, + "loss": 2.6589, + "step": 5570 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.0001933411247762712, + "loss": 2.5324, + "step": 5571 + }, + { + "epoch": 0.24, + "grad_norm": 0.4375, + "learning_rate": 0.00019333871770432936, + "loss": 2.689, + "step": 5572 + }, + { + "epoch": 0.24, + "grad_norm": 0.4375, + "learning_rate": 0.00019333631021239742, + "loss": 2.5699, + "step": 5573 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.00019333390230048619, + "loss": 2.7304, + "step": 5574 + }, + { + "epoch": 0.24, + "grad_norm": 0.44921875, + "learning_rate": 0.00019333149396860648, + "loss": 2.6812, + "step": 5575 + }, + { + "epoch": 0.24, + "grad_norm": 0.46875, + "learning_rate": 0.00019332908521676917, + "loss": 2.6099, + "step": 5576 + }, + { + "epoch": 0.24, + "grad_norm": 0.46875, + "learning_rate": 0.0001933266760449851, + "loss": 2.6255, + "step": 5577 + }, + { + "epoch": 0.24, + "grad_norm": 0.419921875, + "learning_rate": 0.00019332426645326506, + "loss": 2.5541, + "step": 5578 + }, + { + "epoch": 0.24, + "grad_norm": 0.400390625, + "learning_rate": 0.00019332185644161998, + "loss": 2.5467, + "step": 5579 + }, + { + "epoch": 0.24, + "grad_norm": 0.431640625, + "learning_rate": 0.0001933194460100606, + "loss": 2.5301, + "step": 5580 + }, + { + "epoch": 0.24, + "grad_norm": 0.474609375, + "learning_rate": 0.00019331703515859785, + "loss": 2.7438, + "step": 5581 + }, + { + "epoch": 0.24, + "grad_norm": 0.40234375, + "learning_rate": 0.00019331462388724253, + "loss": 2.5631, + "step": 5582 + }, + { + "epoch": 0.24, + "grad_norm": 0.46484375, + "learning_rate": 0.0001933122121960055, + "loss": 2.6513, + "step": 5583 + }, + { + "epoch": 0.24, + "grad_norm": 0.46875, + "learning_rate": 0.00019330980008489763, + "loss": 2.7007, + "step": 5584 + }, + { + "epoch": 0.24, + "grad_norm": 0.46484375, + "learning_rate": 0.0001933073875539298, + "loss": 2.6472, + "step": 5585 + }, + { + "epoch": 0.24, + "grad_norm": 0.47265625, + "learning_rate": 0.00019330497460311278, + "loss": 2.7304, + "step": 5586 + }, + { + "epoch": 0.24, + "grad_norm": 0.470703125, + "learning_rate": 0.00019330256123245744, + "loss": 2.6872, + "step": 5587 + }, + { + "epoch": 0.24, + "grad_norm": 0.421875, + "learning_rate": 0.0001933001474419747, + "loss": 2.7255, + "step": 5588 + }, + { + "epoch": 0.24, + "grad_norm": 0.43359375, + "learning_rate": 0.00019329773323167542, + "loss": 2.6952, + "step": 5589 + }, + { + "epoch": 0.24, + "grad_norm": 0.41015625, + "learning_rate": 0.00019329531860157043, + "loss": 2.6141, + "step": 5590 + }, + { + "epoch": 0.24, + "grad_norm": 0.474609375, + "learning_rate": 0.00019329290355167057, + "loss": 2.7503, + "step": 5591 + }, + { + "epoch": 0.24, + "grad_norm": 0.419921875, + "learning_rate": 0.0001932904880819868, + "loss": 2.6455, + "step": 5592 + }, + { + "epoch": 0.24, + "grad_norm": 0.439453125, + "learning_rate": 0.00019328807219252988, + "loss": 2.5281, + "step": 5593 + }, + { + "epoch": 0.24, + "grad_norm": 0.4140625, + "learning_rate": 0.0001932856558833107, + "loss": 2.5388, + "step": 5594 + }, + { + "epoch": 0.24, + "grad_norm": 0.3984375, + "learning_rate": 0.00019328323915434022, + "loss": 2.6109, + "step": 5595 + }, + { + "epoch": 0.24, + "grad_norm": 0.51953125, + "learning_rate": 0.00019328082200562925, + "loss": 2.7986, + "step": 5596 + }, + { + "epoch": 0.24, + "grad_norm": 0.48828125, + "learning_rate": 0.00019327840443718864, + "loss": 2.7244, + "step": 5597 + }, + { + "epoch": 0.24, + "grad_norm": 0.45703125, + "learning_rate": 0.0001932759864490293, + "loss": 2.6521, + "step": 5598 + }, + { + "epoch": 0.24, + "grad_norm": 0.50390625, + "learning_rate": 0.0001932735680411621, + "loss": 2.684, + "step": 5599 + }, + { + "epoch": 0.24, + "grad_norm": 0.427734375, + "learning_rate": 0.00019327114921359798, + "loss": 2.6357, + "step": 5600 + }, + { + "epoch": 0.24, + "grad_norm": 0.50390625, + "learning_rate": 0.00019326872996634775, + "loss": 2.6208, + "step": 5601 + }, + { + "epoch": 0.24, + "grad_norm": 0.49609375, + "learning_rate": 0.00019326631029942232, + "loss": 2.8177, + "step": 5602 + }, + { + "epoch": 0.24, + "grad_norm": 0.40234375, + "learning_rate": 0.0001932638902128326, + "loss": 2.6595, + "step": 5603 + }, + { + "epoch": 0.24, + "grad_norm": 0.470703125, + "learning_rate": 0.0001932614697065894, + "loss": 2.6085, + "step": 5604 + }, + { + "epoch": 0.24, + "grad_norm": 0.443359375, + "learning_rate": 0.00019325904878070372, + "loss": 2.6792, + "step": 5605 + }, + { + "epoch": 0.24, + "grad_norm": 0.44921875, + "learning_rate": 0.0001932566274351864, + "loss": 2.684, + "step": 5606 + }, + { + "epoch": 0.24, + "grad_norm": 0.38671875, + "learning_rate": 0.00019325420567004834, + "loss": 2.5377, + "step": 5607 + }, + { + "epoch": 0.24, + "grad_norm": 0.453125, + "learning_rate": 0.0001932517834853004, + "loss": 2.6378, + "step": 5608 + }, + { + "epoch": 0.24, + "grad_norm": 0.42578125, + "learning_rate": 0.00019324936088095353, + "loss": 2.6616, + "step": 5609 + }, + { + "epoch": 0.24, + "grad_norm": 0.498046875, + "learning_rate": 0.0001932469378570186, + "loss": 2.6775, + "step": 5610 + }, + { + "epoch": 0.24, + "grad_norm": 0.3984375, + "learning_rate": 0.00019324451441350658, + "loss": 2.6454, + "step": 5611 + }, + { + "epoch": 0.24, + "grad_norm": 0.431640625, + "learning_rate": 0.00019324209055042828, + "loss": 2.6999, + "step": 5612 + }, + { + "epoch": 0.24, + "grad_norm": 0.404296875, + "learning_rate": 0.00019323966626779465, + "loss": 2.6028, + "step": 5613 + }, + { + "epoch": 0.24, + "grad_norm": 0.40625, + "learning_rate": 0.0001932372415656166, + "loss": 2.6775, + "step": 5614 + }, + { + "epoch": 0.24, + "grad_norm": 0.376953125, + "learning_rate": 0.00019323481644390508, + "loss": 2.5951, + "step": 5615 + }, + { + "epoch": 0.24, + "grad_norm": 0.4140625, + "learning_rate": 0.00019323239090267088, + "loss": 2.536, + "step": 5616 + }, + { + "epoch": 0.24, + "grad_norm": 0.443359375, + "learning_rate": 0.00019322996494192506, + "loss": 2.6903, + "step": 5617 + }, + { + "epoch": 0.24, + "grad_norm": 0.43359375, + "learning_rate": 0.00019322753856167843, + "loss": 2.687, + "step": 5618 + }, + { + "epoch": 0.24, + "grad_norm": 0.451171875, + "learning_rate": 0.00019322511176194195, + "loss": 2.6131, + "step": 5619 + }, + { + "epoch": 0.24, + "grad_norm": 0.4140625, + "learning_rate": 0.00019322268454272657, + "loss": 2.6426, + "step": 5620 + }, + { + "epoch": 0.24, + "grad_norm": 0.5234375, + "learning_rate": 0.00019322025690404315, + "loss": 2.6201, + "step": 5621 + }, + { + "epoch": 0.24, + "grad_norm": 0.421875, + "learning_rate": 0.00019321782884590264, + "loss": 2.4875, + "step": 5622 + }, + { + "epoch": 0.24, + "grad_norm": 0.43359375, + "learning_rate": 0.00019321540036831597, + "loss": 2.6614, + "step": 5623 + }, + { + "epoch": 0.24, + "grad_norm": 0.4609375, + "learning_rate": 0.0001932129714712941, + "loss": 2.766, + "step": 5624 + }, + { + "epoch": 0.24, + "grad_norm": 0.41015625, + "learning_rate": 0.00019321054215484787, + "loss": 2.6595, + "step": 5625 + }, + { + "epoch": 0.24, + "grad_norm": 0.443359375, + "learning_rate": 0.00019320811241898831, + "loss": 2.6371, + "step": 5626 + }, + { + "epoch": 0.24, + "grad_norm": 0.47265625, + "learning_rate": 0.00019320568226372629, + "loss": 2.6372, + "step": 5627 + }, + { + "epoch": 0.24, + "grad_norm": 0.376953125, + "learning_rate": 0.00019320325168907277, + "loss": 2.6874, + "step": 5628 + }, + { + "epoch": 0.24, + "grad_norm": 0.38671875, + "learning_rate": 0.00019320082069503865, + "loss": 2.6968, + "step": 5629 + }, + { + "epoch": 0.24, + "grad_norm": 0.37890625, + "learning_rate": 0.00019319838928163492, + "loss": 2.6344, + "step": 5630 + }, + { + "epoch": 0.24, + "grad_norm": 0.439453125, + "learning_rate": 0.0001931959574488725, + "loss": 2.6619, + "step": 5631 + }, + { + "epoch": 0.24, + "grad_norm": 0.5078125, + "learning_rate": 0.00019319352519676233, + "loss": 2.7113, + "step": 5632 + }, + { + "epoch": 0.24, + "grad_norm": 0.44921875, + "learning_rate": 0.00019319109252531535, + "loss": 2.6152, + "step": 5633 + }, + { + "epoch": 0.24, + "grad_norm": 0.45703125, + "learning_rate": 0.0001931886594345425, + "loss": 2.6738, + "step": 5634 + }, + { + "epoch": 0.24, + "grad_norm": 0.4453125, + "learning_rate": 0.00019318622592445474, + "loss": 2.64, + "step": 5635 + }, + { + "epoch": 0.24, + "grad_norm": 0.47265625, + "learning_rate": 0.00019318379199506305, + "loss": 2.6414, + "step": 5636 + }, + { + "epoch": 0.24, + "grad_norm": 0.39453125, + "learning_rate": 0.00019318135764637834, + "loss": 2.6514, + "step": 5637 + }, + { + "epoch": 0.24, + "grad_norm": 0.421875, + "learning_rate": 0.00019317892287841156, + "loss": 2.6409, + "step": 5638 + }, + { + "epoch": 0.24, + "grad_norm": 0.48046875, + "learning_rate": 0.0001931764876911737, + "loss": 2.6883, + "step": 5639 + }, + { + "epoch": 0.24, + "grad_norm": 0.412109375, + "learning_rate": 0.00019317405208467566, + "loss": 2.7023, + "step": 5640 + }, + { + "epoch": 0.24, + "grad_norm": 0.453125, + "learning_rate": 0.0001931716160589285, + "loss": 2.5961, + "step": 5641 + }, + { + "epoch": 0.24, + "grad_norm": 0.443359375, + "learning_rate": 0.00019316917961394306, + "loss": 2.6374, + "step": 5642 + }, + { + "epoch": 0.24, + "grad_norm": 0.44921875, + "learning_rate": 0.00019316674274973036, + "loss": 2.7489, + "step": 5643 + }, + { + "epoch": 0.24, + "grad_norm": 0.431640625, + "learning_rate": 0.0001931643054663014, + "loss": 2.6482, + "step": 5644 + }, + { + "epoch": 0.24, + "grad_norm": 0.4453125, + "learning_rate": 0.0001931618677636671, + "loss": 2.5574, + "step": 5645 + }, + { + "epoch": 0.24, + "grad_norm": 0.421875, + "learning_rate": 0.00019315942964183842, + "loss": 2.6837, + "step": 5646 + }, + { + "epoch": 0.24, + "grad_norm": 0.4296875, + "learning_rate": 0.0001931569911008264, + "loss": 2.6272, + "step": 5647 + }, + { + "epoch": 0.24, + "grad_norm": 0.4296875, + "learning_rate": 0.00019315455214064192, + "loss": 2.6264, + "step": 5648 + }, + { + "epoch": 0.24, + "grad_norm": 0.396484375, + "learning_rate": 0.00019315211276129602, + "loss": 2.733, + "step": 5649 + }, + { + "epoch": 0.24, + "grad_norm": 0.380859375, + "learning_rate": 0.00019314967296279965, + "loss": 2.5935, + "step": 5650 + }, + { + "epoch": 0.24, + "grad_norm": 0.369140625, + "learning_rate": 0.0001931472327451638, + "loss": 2.5749, + "step": 5651 + }, + { + "epoch": 0.24, + "grad_norm": 0.443359375, + "learning_rate": 0.00019314479210839945, + "loss": 2.6202, + "step": 5652 + }, + { + "epoch": 0.24, + "grad_norm": 0.443359375, + "learning_rate": 0.00019314235105251756, + "loss": 2.7659, + "step": 5653 + }, + { + "epoch": 0.24, + "grad_norm": 0.447265625, + "learning_rate": 0.00019313990957752915, + "loss": 2.6966, + "step": 5654 + }, + { + "epoch": 0.24, + "grad_norm": 0.41796875, + "learning_rate": 0.00019313746768344516, + "loss": 2.7291, + "step": 5655 + }, + { + "epoch": 0.24, + "grad_norm": 0.3671875, + "learning_rate": 0.0001931350253702766, + "loss": 2.5788, + "step": 5656 + }, + { + "epoch": 0.24, + "grad_norm": 0.4765625, + "learning_rate": 0.0001931325826380345, + "loss": 2.7932, + "step": 5657 + }, + { + "epoch": 0.24, + "grad_norm": 0.4765625, + "learning_rate": 0.0001931301394867298, + "loss": 2.7201, + "step": 5658 + }, + { + "epoch": 0.24, + "grad_norm": 0.439453125, + "learning_rate": 0.0001931276959163735, + "loss": 2.5062, + "step": 5659 + }, + { + "epoch": 0.24, + "grad_norm": 0.44921875, + "learning_rate": 0.0001931252519269766, + "loss": 2.7441, + "step": 5660 + }, + { + "epoch": 0.24, + "grad_norm": 0.453125, + "learning_rate": 0.0001931228075185501, + "loss": 2.5611, + "step": 5661 + }, + { + "epoch": 0.24, + "grad_norm": 0.46484375, + "learning_rate": 0.000193120362691105, + "loss": 2.7476, + "step": 5662 + }, + { + "epoch": 0.24, + "grad_norm": 0.443359375, + "learning_rate": 0.0001931179174446523, + "loss": 2.6896, + "step": 5663 + }, + { + "epoch": 0.24, + "grad_norm": 0.455078125, + "learning_rate": 0.00019311547177920302, + "loss": 2.6254, + "step": 5664 + }, + { + "epoch": 0.24, + "grad_norm": 0.4375, + "learning_rate": 0.00019311302569476814, + "loss": 2.7413, + "step": 5665 + }, + { + "epoch": 0.24, + "grad_norm": 0.44140625, + "learning_rate": 0.00019311057919135864, + "loss": 2.7074, + "step": 5666 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.00019310813226898558, + "loss": 2.6702, + "step": 5667 + }, + { + "epoch": 0.24, + "grad_norm": 0.416015625, + "learning_rate": 0.00019310568492765997, + "loss": 2.6287, + "step": 5668 + }, + { + "epoch": 0.24, + "grad_norm": 0.427734375, + "learning_rate": 0.0001931032371673928, + "loss": 2.557, + "step": 5669 + }, + { + "epoch": 0.24, + "grad_norm": 0.431640625, + "learning_rate": 0.00019310078898819507, + "loss": 2.7214, + "step": 5670 + }, + { + "epoch": 0.24, + "grad_norm": 0.439453125, + "learning_rate": 0.0001930983403900778, + "loss": 2.652, + "step": 5671 + }, + { + "epoch": 0.24, + "grad_norm": 0.453125, + "learning_rate": 0.00019309589137305206, + "loss": 2.5547, + "step": 5672 + }, + { + "epoch": 0.24, + "grad_norm": 0.51171875, + "learning_rate": 0.00019309344193712882, + "loss": 2.5934, + "step": 5673 + }, + { + "epoch": 0.24, + "grad_norm": 0.53515625, + "learning_rate": 0.0001930909920823191, + "loss": 2.6308, + "step": 5674 + }, + { + "epoch": 0.24, + "grad_norm": 0.453125, + "learning_rate": 0.00019308854180863396, + "loss": 2.5754, + "step": 5675 + }, + { + "epoch": 0.24, + "grad_norm": 0.462890625, + "learning_rate": 0.00019308609111608438, + "loss": 2.7058, + "step": 5676 + }, + { + "epoch": 0.24, + "grad_norm": 0.46484375, + "learning_rate": 0.00019308364000468142, + "loss": 2.5692, + "step": 5677 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.0001930811884744361, + "loss": 2.6689, + "step": 5678 + }, + { + "epoch": 0.24, + "grad_norm": 0.3984375, + "learning_rate": 0.00019307873652535946, + "loss": 2.6273, + "step": 5679 + }, + { + "epoch": 0.24, + "grad_norm": 0.46875, + "learning_rate": 0.00019307628415746249, + "loss": 2.7434, + "step": 5680 + }, + { + "epoch": 0.24, + "grad_norm": 0.41796875, + "learning_rate": 0.00019307383137075627, + "loss": 2.5233, + "step": 5681 + }, + { + "epoch": 0.24, + "grad_norm": 0.4296875, + "learning_rate": 0.00019307137816525183, + "loss": 2.6494, + "step": 5682 + }, + { + "epoch": 0.24, + "grad_norm": 0.43359375, + "learning_rate": 0.00019306892454096022, + "loss": 2.6105, + "step": 5683 + }, + { + "epoch": 0.24, + "grad_norm": 0.46875, + "learning_rate": 0.00019306647049789245, + "loss": 2.7544, + "step": 5684 + }, + { + "epoch": 0.24, + "grad_norm": 0.39453125, + "learning_rate": 0.00019306401603605956, + "loss": 2.7016, + "step": 5685 + }, + { + "epoch": 0.24, + "grad_norm": 0.494140625, + "learning_rate": 0.0001930615611554726, + "loss": 2.6196, + "step": 5686 + }, + { + "epoch": 0.24, + "grad_norm": 0.439453125, + "learning_rate": 0.00019305910585614263, + "loss": 2.5854, + "step": 5687 + }, + { + "epoch": 0.24, + "grad_norm": 0.390625, + "learning_rate": 0.0001930566501380807, + "loss": 2.6109, + "step": 5688 + }, + { + "epoch": 0.24, + "grad_norm": 0.44921875, + "learning_rate": 0.00019305419400129784, + "loss": 2.658, + "step": 5689 + }, + { + "epoch": 0.24, + "grad_norm": 0.470703125, + "learning_rate": 0.00019305173744580512, + "loss": 2.5024, + "step": 5690 + }, + { + "epoch": 0.24, + "grad_norm": 0.53125, + "learning_rate": 0.00019304928047161361, + "loss": 2.7205, + "step": 5691 + }, + { + "epoch": 0.24, + "grad_norm": 0.392578125, + "learning_rate": 0.00019304682307873434, + "loss": 2.6134, + "step": 5692 + }, + { + "epoch": 0.24, + "grad_norm": 0.73046875, + "learning_rate": 0.00019304436526717836, + "loss": 2.6135, + "step": 5693 + }, + { + "epoch": 0.24, + "grad_norm": 0.451171875, + "learning_rate": 0.00019304190703695677, + "loss": 2.6841, + "step": 5694 + }, + { + "epoch": 0.24, + "grad_norm": 0.39453125, + "learning_rate": 0.00019303944838808054, + "loss": 2.6616, + "step": 5695 + }, + { + "epoch": 0.24, + "grad_norm": 0.44921875, + "learning_rate": 0.00019303698932056086, + "loss": 2.7832, + "step": 5696 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.00019303452983440868, + "loss": 2.6194, + "step": 5697 + }, + { + "epoch": 0.24, + "grad_norm": 0.455078125, + "learning_rate": 0.00019303206992963515, + "loss": 2.6278, + "step": 5698 + }, + { + "epoch": 0.24, + "grad_norm": 0.431640625, + "learning_rate": 0.00019302960960625127, + "loss": 2.6158, + "step": 5699 + }, + { + "epoch": 0.24, + "grad_norm": 0.412109375, + "learning_rate": 0.00019302714886426818, + "loss": 2.6183, + "step": 5700 + }, + { + "epoch": 0.24, + "grad_norm": 0.4609375, + "learning_rate": 0.0001930246877036969, + "loss": 2.6204, + "step": 5701 + }, + { + "epoch": 0.24, + "grad_norm": 0.400390625, + "learning_rate": 0.0001930222261245485, + "loss": 2.5507, + "step": 5702 + }, + { + "epoch": 0.24, + "grad_norm": 0.4296875, + "learning_rate": 0.0001930197641268341, + "loss": 2.4698, + "step": 5703 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.00019301730171056476, + "loss": 2.5914, + "step": 5704 + }, + { + "epoch": 0.24, + "grad_norm": 0.42578125, + "learning_rate": 0.00019301483887575156, + "loss": 2.5033, + "step": 5705 + }, + { + "epoch": 0.24, + "grad_norm": 0.515625, + "learning_rate": 0.00019301237562240558, + "loss": 2.7889, + "step": 5706 + }, + { + "epoch": 0.24, + "grad_norm": 0.49609375, + "learning_rate": 0.00019300991195053788, + "loss": 2.6862, + "step": 5707 + }, + { + "epoch": 0.24, + "grad_norm": 0.486328125, + "learning_rate": 0.00019300744786015958, + "loss": 2.4383, + "step": 5708 + }, + { + "epoch": 0.24, + "grad_norm": 0.416015625, + "learning_rate": 0.00019300498335128173, + "loss": 2.6373, + "step": 5709 + }, + { + "epoch": 0.24, + "grad_norm": 0.4453125, + "learning_rate": 0.0001930025184239155, + "loss": 2.6935, + "step": 5710 + }, + { + "epoch": 0.24, + "grad_norm": 0.474609375, + "learning_rate": 0.0001930000530780719, + "loss": 2.6465, + "step": 5711 + }, + { + "epoch": 0.24, + "grad_norm": 0.40234375, + "learning_rate": 0.00019299758731376203, + "loss": 2.6288, + "step": 5712 + }, + { + "epoch": 0.24, + "grad_norm": 0.431640625, + "learning_rate": 0.00019299512113099703, + "loss": 2.669, + "step": 5713 + }, + { + "epoch": 0.24, + "grad_norm": 0.4765625, + "learning_rate": 0.00019299265452978794, + "loss": 2.7775, + "step": 5714 + }, + { + "epoch": 0.24, + "grad_norm": 0.5078125, + "learning_rate": 0.00019299018751014593, + "loss": 2.6585, + "step": 5715 + }, + { + "epoch": 0.24, + "grad_norm": 0.474609375, + "learning_rate": 0.00019298772007208203, + "loss": 2.8315, + "step": 5716 + }, + { + "epoch": 0.24, + "grad_norm": 0.458984375, + "learning_rate": 0.0001929852522156074, + "loss": 2.6913, + "step": 5717 + }, + { + "epoch": 0.24, + "grad_norm": 0.37890625, + "learning_rate": 0.0001929827839407331, + "loss": 2.6893, + "step": 5718 + }, + { + "epoch": 0.24, + "grad_norm": 0.39453125, + "learning_rate": 0.00019298031524747026, + "loss": 2.631, + "step": 5719 + }, + { + "epoch": 0.24, + "grad_norm": 0.478515625, + "learning_rate": 0.00019297784613582993, + "loss": 2.6695, + "step": 5720 + }, + { + "epoch": 0.24, + "grad_norm": 0.421875, + "learning_rate": 0.0001929753766058233, + "loss": 2.6352, + "step": 5721 + }, + { + "epoch": 0.24, + "grad_norm": 0.41015625, + "learning_rate": 0.0001929729066574615, + "loss": 2.7005, + "step": 5722 + }, + { + "epoch": 0.24, + "grad_norm": 0.4609375, + "learning_rate": 0.00019297043629075558, + "loss": 2.7289, + "step": 5723 + }, + { + "epoch": 0.24, + "grad_norm": 0.4453125, + "learning_rate": 0.00019296796550571666, + "loss": 2.5792, + "step": 5724 + }, + { + "epoch": 0.24, + "grad_norm": 0.451171875, + "learning_rate": 0.00019296549430235588, + "loss": 2.6556, + "step": 5725 + }, + { + "epoch": 0.24, + "grad_norm": 0.43359375, + "learning_rate": 0.0001929630226806843, + "loss": 2.8465, + "step": 5726 + }, + { + "epoch": 0.24, + "grad_norm": 0.419921875, + "learning_rate": 0.00019296055064071314, + "loss": 2.6371, + "step": 5727 + }, + { + "epoch": 0.24, + "grad_norm": 0.404296875, + "learning_rate": 0.00019295807818245348, + "loss": 2.6212, + "step": 5728 + }, + { + "epoch": 0.24, + "grad_norm": 0.427734375, + "learning_rate": 0.00019295560530591645, + "loss": 2.5482, + "step": 5729 + }, + { + "epoch": 0.24, + "grad_norm": 0.421875, + "learning_rate": 0.00019295313201111314, + "loss": 2.5884, + "step": 5730 + }, + { + "epoch": 0.24, + "grad_norm": 0.46484375, + "learning_rate": 0.00019295065829805469, + "loss": 2.7469, + "step": 5731 + }, + { + "epoch": 0.24, + "grad_norm": 0.439453125, + "learning_rate": 0.0001929481841667523, + "loss": 2.7761, + "step": 5732 + }, + { + "epoch": 0.24, + "grad_norm": 0.4140625, + "learning_rate": 0.00019294570961721701, + "loss": 2.6559, + "step": 5733 + }, + { + "epoch": 0.24, + "grad_norm": 0.51953125, + "learning_rate": 0.00019294323464946, + "loss": 2.614, + "step": 5734 + }, + { + "epoch": 0.24, + "grad_norm": 0.48828125, + "learning_rate": 0.00019294075926349242, + "loss": 2.6719, + "step": 5735 + }, + { + "epoch": 0.24, + "grad_norm": 0.41796875, + "learning_rate": 0.00019293828345932538, + "loss": 2.7605, + "step": 5736 + }, + { + "epoch": 0.24, + "grad_norm": 0.43359375, + "learning_rate": 0.00019293580723697, + "loss": 2.673, + "step": 5737 + }, + { + "epoch": 0.24, + "grad_norm": 0.466796875, + "learning_rate": 0.0001929333305964375, + "loss": 2.751, + "step": 5738 + }, + { + "epoch": 0.24, + "grad_norm": 0.412109375, + "learning_rate": 0.00019293085353773895, + "loss": 2.7189, + "step": 5739 + }, + { + "epoch": 0.24, + "grad_norm": 0.412109375, + "learning_rate": 0.00019292837606088554, + "loss": 2.651, + "step": 5740 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.00019292589816588836, + "loss": 2.6674, + "step": 5741 + }, + { + "epoch": 0.24, + "grad_norm": 0.4140625, + "learning_rate": 0.0001929234198527586, + "loss": 2.6008, + "step": 5742 + }, + { + "epoch": 0.24, + "grad_norm": 0.4375, + "learning_rate": 0.00019292094112150748, + "loss": 2.7537, + "step": 5743 + }, + { + "epoch": 0.24, + "grad_norm": 0.435546875, + "learning_rate": 0.00019291846197214602, + "loss": 2.7037, + "step": 5744 + }, + { + "epoch": 0.24, + "grad_norm": 0.4140625, + "learning_rate": 0.00019291598240468548, + "loss": 2.7786, + "step": 5745 + }, + { + "epoch": 0.24, + "grad_norm": 0.423828125, + "learning_rate": 0.00019291350241913693, + "loss": 2.7565, + "step": 5746 + }, + { + "epoch": 0.24, + "grad_norm": 0.427734375, + "learning_rate": 0.00019291102201551158, + "loss": 2.6205, + "step": 5747 + }, + { + "epoch": 0.24, + "grad_norm": 0.439453125, + "learning_rate": 0.0001929085411938206, + "loss": 2.7471, + "step": 5748 + }, + { + "epoch": 0.24, + "grad_norm": 0.470703125, + "learning_rate": 0.00019290605995407516, + "loss": 2.6836, + "step": 5749 + }, + { + "epoch": 0.25, + "grad_norm": 0.43359375, + "learning_rate": 0.0001929035782962864, + "loss": 2.6242, + "step": 5750 + }, + { + "epoch": 0.25, + "grad_norm": 0.435546875, + "learning_rate": 0.00019290109622046545, + "loss": 2.6905, + "step": 5751 + }, + { + "epoch": 0.25, + "grad_norm": 0.490234375, + "learning_rate": 0.00019289861372662355, + "loss": 2.7887, + "step": 5752 + }, + { + "epoch": 0.25, + "grad_norm": 0.431640625, + "learning_rate": 0.00019289613081477182, + "loss": 2.6926, + "step": 5753 + }, + { + "epoch": 0.25, + "grad_norm": 0.482421875, + "learning_rate": 0.00019289364748492147, + "loss": 2.7067, + "step": 5754 + }, + { + "epoch": 0.25, + "grad_norm": 0.462890625, + "learning_rate": 0.00019289116373708365, + "loss": 2.6626, + "step": 5755 + }, + { + "epoch": 0.25, + "grad_norm": 0.451171875, + "learning_rate": 0.00019288867957126954, + "loss": 2.6814, + "step": 5756 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019288619498749032, + "loss": 2.6542, + "step": 5757 + }, + { + "epoch": 0.25, + "grad_norm": 0.373046875, + "learning_rate": 0.00019288370998575717, + "loss": 2.5711, + "step": 5758 + }, + { + "epoch": 0.25, + "grad_norm": 0.474609375, + "learning_rate": 0.00019288122456608125, + "loss": 2.648, + "step": 5759 + }, + { + "epoch": 0.25, + "grad_norm": 0.42578125, + "learning_rate": 0.0001928787387284738, + "loss": 2.6299, + "step": 5760 + }, + { + "epoch": 0.25, + "grad_norm": 0.44140625, + "learning_rate": 0.00019287625247294595, + "loss": 2.6925, + "step": 5761 + }, + { + "epoch": 0.25, + "grad_norm": 0.43359375, + "learning_rate": 0.0001928737657995089, + "loss": 2.6253, + "step": 5762 + }, + { + "epoch": 0.25, + "grad_norm": 0.466796875, + "learning_rate": 0.00019287127870817384, + "loss": 2.7235, + "step": 5763 + }, + { + "epoch": 0.25, + "grad_norm": 0.47265625, + "learning_rate": 0.000192868791198952, + "loss": 2.6577, + "step": 5764 + }, + { + "epoch": 0.25, + "grad_norm": 0.470703125, + "learning_rate": 0.00019286630327185452, + "loss": 2.6106, + "step": 5765 + }, + { + "epoch": 0.25, + "grad_norm": 0.392578125, + "learning_rate": 0.0001928638149268926, + "loss": 2.6219, + "step": 5766 + }, + { + "epoch": 0.25, + "grad_norm": 0.419921875, + "learning_rate": 0.00019286132616407745, + "loss": 2.6624, + "step": 5767 + }, + { + "epoch": 0.25, + "grad_norm": 0.416015625, + "learning_rate": 0.00019285883698342028, + "loss": 2.7025, + "step": 5768 + }, + { + "epoch": 0.25, + "grad_norm": 0.42578125, + "learning_rate": 0.00019285634738493227, + "loss": 2.7694, + "step": 5769 + }, + { + "epoch": 0.25, + "grad_norm": 0.423828125, + "learning_rate": 0.00019285385736862466, + "loss": 2.6746, + "step": 5770 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019285136693450862, + "loss": 2.6655, + "step": 5771 + }, + { + "epoch": 0.25, + "grad_norm": 0.41015625, + "learning_rate": 0.00019284887608259532, + "loss": 2.616, + "step": 5772 + }, + { + "epoch": 0.25, + "grad_norm": 0.47265625, + "learning_rate": 0.00019284638481289606, + "loss": 2.5732, + "step": 5773 + }, + { + "epoch": 0.25, + "grad_norm": 0.451171875, + "learning_rate": 0.000192843893125422, + "loss": 2.6851, + "step": 5774 + }, + { + "epoch": 0.25, + "grad_norm": 0.43359375, + "learning_rate": 0.0001928414010201843, + "loss": 2.6975, + "step": 5775 + }, + { + "epoch": 0.25, + "grad_norm": 0.423828125, + "learning_rate": 0.00019283890849719425, + "loss": 2.6557, + "step": 5776 + }, + { + "epoch": 0.25, + "grad_norm": 0.451171875, + "learning_rate": 0.00019283641555646307, + "loss": 2.7141, + "step": 5777 + }, + { + "epoch": 0.25, + "grad_norm": 0.40234375, + "learning_rate": 0.0001928339221980019, + "loss": 2.6727, + "step": 5778 + }, + { + "epoch": 0.25, + "grad_norm": 0.5234375, + "learning_rate": 0.00019283142842182205, + "loss": 2.7189, + "step": 5779 + }, + { + "epoch": 0.25, + "grad_norm": 0.4140625, + "learning_rate": 0.00019282893422793465, + "loss": 2.8105, + "step": 5780 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019282643961635103, + "loss": 2.5467, + "step": 5781 + }, + { + "epoch": 0.25, + "grad_norm": 0.421875, + "learning_rate": 0.00019282394458708233, + "loss": 2.4956, + "step": 5782 + }, + { + "epoch": 0.25, + "grad_norm": 0.3984375, + "learning_rate": 0.0001928214491401398, + "loss": 2.592, + "step": 5783 + }, + { + "epoch": 0.25, + "grad_norm": 0.435546875, + "learning_rate": 0.00019281895327553466, + "loss": 2.5535, + "step": 5784 + }, + { + "epoch": 0.25, + "grad_norm": 0.43359375, + "learning_rate": 0.00019281645699327816, + "loss": 2.7005, + "step": 5785 + }, + { + "epoch": 0.25, + "grad_norm": 0.443359375, + "learning_rate": 0.00019281396029338153, + "loss": 2.7233, + "step": 5786 + }, + { + "epoch": 0.25, + "grad_norm": 0.470703125, + "learning_rate": 0.00019281146317585602, + "loss": 2.8026, + "step": 5787 + }, + { + "epoch": 0.25, + "grad_norm": 0.498046875, + "learning_rate": 0.0001928089656407128, + "loss": 2.782, + "step": 5788 + }, + { + "epoch": 0.25, + "grad_norm": 0.45703125, + "learning_rate": 0.00019280646768796316, + "loss": 2.8069, + "step": 5789 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.00019280396931761838, + "loss": 2.7081, + "step": 5790 + }, + { + "epoch": 0.25, + "grad_norm": 0.400390625, + "learning_rate": 0.00019280147052968958, + "loss": 2.6213, + "step": 5791 + }, + { + "epoch": 0.25, + "grad_norm": 0.4140625, + "learning_rate": 0.00019279897132418815, + "loss": 2.7125, + "step": 5792 + }, + { + "epoch": 0.25, + "grad_norm": 0.421875, + "learning_rate": 0.0001927964717011252, + "loss": 2.605, + "step": 5793 + }, + { + "epoch": 0.25, + "grad_norm": 0.455078125, + "learning_rate": 0.00019279397166051208, + "loss": 2.6836, + "step": 5794 + }, + { + "epoch": 0.25, + "grad_norm": 0.44921875, + "learning_rate": 0.00019279147120235997, + "loss": 2.6467, + "step": 5795 + }, + { + "epoch": 0.25, + "grad_norm": 0.466796875, + "learning_rate": 0.00019278897032668017, + "loss": 2.7591, + "step": 5796 + }, + { + "epoch": 0.25, + "grad_norm": 0.431640625, + "learning_rate": 0.00019278646903348392, + "loss": 2.5516, + "step": 5797 + }, + { + "epoch": 0.25, + "grad_norm": 0.431640625, + "learning_rate": 0.00019278396732278248, + "loss": 2.5878, + "step": 5798 + }, + { + "epoch": 0.25, + "grad_norm": 0.423828125, + "learning_rate": 0.00019278146519458708, + "loss": 2.6859, + "step": 5799 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.00019277896264890897, + "loss": 2.4482, + "step": 5800 + }, + { + "epoch": 0.25, + "grad_norm": 0.400390625, + "learning_rate": 0.00019277645968575948, + "loss": 2.724, + "step": 5801 + }, + { + "epoch": 0.25, + "grad_norm": 0.44140625, + "learning_rate": 0.0001927739563051498, + "loss": 2.6818, + "step": 5802 + }, + { + "epoch": 0.25, + "grad_norm": 0.41796875, + "learning_rate": 0.00019277145250709123, + "loss": 2.5601, + "step": 5803 + }, + { + "epoch": 0.25, + "grad_norm": 0.451171875, + "learning_rate": 0.00019276894829159502, + "loss": 2.5387, + "step": 5804 + }, + { + "epoch": 0.25, + "grad_norm": 0.44140625, + "learning_rate": 0.00019276644365867247, + "loss": 2.6449, + "step": 5805 + }, + { + "epoch": 0.25, + "grad_norm": 0.384765625, + "learning_rate": 0.00019276393860833478, + "loss": 2.6031, + "step": 5806 + }, + { + "epoch": 0.25, + "grad_norm": 0.46875, + "learning_rate": 0.0001927614331405933, + "loss": 2.7522, + "step": 5807 + }, + { + "epoch": 0.25, + "grad_norm": 0.443359375, + "learning_rate": 0.00019275892725545928, + "loss": 2.725, + "step": 5808 + }, + { + "epoch": 0.25, + "grad_norm": 0.4140625, + "learning_rate": 0.00019275642095294395, + "loss": 2.7741, + "step": 5809 + }, + { + "epoch": 0.25, + "grad_norm": 0.44140625, + "learning_rate": 0.00019275391423305866, + "loss": 2.5488, + "step": 5810 + }, + { + "epoch": 0.25, + "grad_norm": 0.412109375, + "learning_rate": 0.00019275140709581464, + "loss": 2.4663, + "step": 5811 + }, + { + "epoch": 0.25, + "grad_norm": 0.3984375, + "learning_rate": 0.00019274889954122316, + "loss": 2.5057, + "step": 5812 + }, + { + "epoch": 0.25, + "grad_norm": 0.4375, + "learning_rate": 0.00019274639156929556, + "loss": 2.7483, + "step": 5813 + }, + { + "epoch": 0.25, + "grad_norm": 0.357421875, + "learning_rate": 0.00019274388318004312, + "loss": 2.6726, + "step": 5814 + }, + { + "epoch": 0.25, + "grad_norm": 0.474609375, + "learning_rate": 0.00019274137437347703, + "loss": 2.5307, + "step": 5815 + }, + { + "epoch": 0.25, + "grad_norm": 0.484375, + "learning_rate": 0.00019273886514960872, + "loss": 2.6565, + "step": 5816 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.00019273635550844936, + "loss": 2.6905, + "step": 5817 + }, + { + "epoch": 0.25, + "grad_norm": 0.4140625, + "learning_rate": 0.0001927338454500103, + "loss": 2.6216, + "step": 5818 + }, + { + "epoch": 0.25, + "grad_norm": 0.412109375, + "learning_rate": 0.00019273133497430282, + "loss": 2.5866, + "step": 5819 + }, + { + "epoch": 0.25, + "grad_norm": 0.5, + "learning_rate": 0.00019272882408133825, + "loss": 2.6518, + "step": 5820 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.00019272631277112784, + "loss": 2.5435, + "step": 5821 + }, + { + "epoch": 0.25, + "grad_norm": 0.392578125, + "learning_rate": 0.00019272380104368291, + "loss": 2.6569, + "step": 5822 + }, + { + "epoch": 0.25, + "grad_norm": 0.478515625, + "learning_rate": 0.00019272128889901476, + "loss": 2.6356, + "step": 5823 + }, + { + "epoch": 0.25, + "grad_norm": 0.41015625, + "learning_rate": 0.00019271877633713472, + "loss": 2.6275, + "step": 5824 + }, + { + "epoch": 0.25, + "grad_norm": 0.47265625, + "learning_rate": 0.00019271626335805404, + "loss": 2.8903, + "step": 5825 + }, + { + "epoch": 0.25, + "grad_norm": 0.431640625, + "learning_rate": 0.00019271374996178406, + "loss": 2.7013, + "step": 5826 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.0001927112361483361, + "loss": 2.6166, + "step": 5827 + }, + { + "epoch": 0.25, + "grad_norm": 0.390625, + "learning_rate": 0.00019270872191772144, + "loss": 2.5588, + "step": 5828 + }, + { + "epoch": 0.25, + "grad_norm": 0.546875, + "learning_rate": 0.00019270620726995145, + "loss": 2.8065, + "step": 5829 + }, + { + "epoch": 0.25, + "grad_norm": 0.486328125, + "learning_rate": 0.0001927036922050374, + "loss": 2.704, + "step": 5830 + }, + { + "epoch": 0.25, + "grad_norm": 0.380859375, + "learning_rate": 0.0001927011767229906, + "loss": 2.7658, + "step": 5831 + }, + { + "epoch": 0.25, + "grad_norm": 0.482421875, + "learning_rate": 0.00019269866082382238, + "loss": 2.7246, + "step": 5832 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.00019269614450754407, + "loss": 2.6854, + "step": 5833 + }, + { + "epoch": 0.25, + "grad_norm": 0.466796875, + "learning_rate": 0.00019269362777416698, + "loss": 2.6618, + "step": 5834 + }, + { + "epoch": 0.25, + "grad_norm": 0.5, + "learning_rate": 0.00019269111062370246, + "loss": 2.6464, + "step": 5835 + }, + { + "epoch": 0.25, + "grad_norm": 0.41796875, + "learning_rate": 0.0001926885930561618, + "loss": 2.6389, + "step": 5836 + }, + { + "epoch": 0.25, + "grad_norm": 0.47265625, + "learning_rate": 0.00019268607507155636, + "loss": 2.5906, + "step": 5837 + }, + { + "epoch": 0.25, + "grad_norm": 0.443359375, + "learning_rate": 0.0001926835566698974, + "loss": 2.7143, + "step": 5838 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019268103785119637, + "loss": 2.6568, + "step": 5839 + }, + { + "epoch": 0.25, + "grad_norm": 0.46875, + "learning_rate": 0.00019267851861546454, + "loss": 2.7423, + "step": 5840 + }, + { + "epoch": 0.25, + "grad_norm": 0.447265625, + "learning_rate": 0.00019267599896271323, + "loss": 2.6476, + "step": 5841 + }, + { + "epoch": 0.25, + "grad_norm": 0.404296875, + "learning_rate": 0.00019267347889295378, + "loss": 2.6376, + "step": 5842 + }, + { + "epoch": 0.25, + "grad_norm": 0.41015625, + "learning_rate": 0.00019267095840619757, + "loss": 2.7084, + "step": 5843 + }, + { + "epoch": 0.25, + "grad_norm": 0.412109375, + "learning_rate": 0.0001926684375024559, + "loss": 2.5215, + "step": 5844 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.00019266591618174013, + "loss": 2.6776, + "step": 5845 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.0001926633944440616, + "loss": 2.6607, + "step": 5846 + }, + { + "epoch": 0.25, + "grad_norm": 0.498046875, + "learning_rate": 0.00019266087228943165, + "loss": 2.7519, + "step": 5847 + }, + { + "epoch": 0.25, + "grad_norm": 0.640625, + "learning_rate": 0.00019265834971786164, + "loss": 2.6575, + "step": 5848 + }, + { + "epoch": 0.25, + "grad_norm": 0.4609375, + "learning_rate": 0.0001926558267293629, + "loss": 2.657, + "step": 5849 + }, + { + "epoch": 0.25, + "grad_norm": 0.451171875, + "learning_rate": 0.00019265330332394687, + "loss": 2.6551, + "step": 5850 + }, + { + "epoch": 0.25, + "grad_norm": 0.435546875, + "learning_rate": 0.00019265077950162478, + "loss": 2.6967, + "step": 5851 + }, + { + "epoch": 0.25, + "grad_norm": 0.447265625, + "learning_rate": 0.00019264825526240804, + "loss": 2.6728, + "step": 5852 + }, + { + "epoch": 0.25, + "grad_norm": 0.416015625, + "learning_rate": 0.00019264573060630803, + "loss": 2.6938, + "step": 5853 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.00019264320553333608, + "loss": 2.677, + "step": 5854 + }, + { + "epoch": 0.25, + "grad_norm": 0.423828125, + "learning_rate": 0.00019264068004350358, + "loss": 2.6727, + "step": 5855 + }, + { + "epoch": 0.25, + "grad_norm": 0.458984375, + "learning_rate": 0.00019263815413682187, + "loss": 2.6547, + "step": 5856 + }, + { + "epoch": 0.25, + "grad_norm": 0.50390625, + "learning_rate": 0.00019263562781330232, + "loss": 2.6618, + "step": 5857 + }, + { + "epoch": 0.25, + "grad_norm": 0.4609375, + "learning_rate": 0.0001926331010729563, + "loss": 2.4998, + "step": 5858 + }, + { + "epoch": 0.25, + "grad_norm": 0.50390625, + "learning_rate": 0.00019263057391579515, + "loss": 2.5981, + "step": 5859 + }, + { + "epoch": 0.25, + "grad_norm": 0.392578125, + "learning_rate": 0.0001926280463418303, + "loss": 2.5007, + "step": 5860 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.0001926255183510731, + "loss": 2.7212, + "step": 5861 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.00019262298994353493, + "loss": 2.6612, + "step": 5862 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019262046111922712, + "loss": 2.5674, + "step": 5863 + }, + { + "epoch": 0.25, + "grad_norm": 0.478515625, + "learning_rate": 0.00019261793187816108, + "loss": 2.6482, + "step": 5864 + }, + { + "epoch": 0.25, + "grad_norm": 0.4375, + "learning_rate": 0.0001926154022203482, + "loss": 2.553, + "step": 5865 + }, + { + "epoch": 0.25, + "grad_norm": 0.462890625, + "learning_rate": 0.00019261287214579988, + "loss": 2.6412, + "step": 5866 + }, + { + "epoch": 0.25, + "grad_norm": 0.451171875, + "learning_rate": 0.00019261034165452746, + "loss": 2.7339, + "step": 5867 + }, + { + "epoch": 0.25, + "eval_loss": 2.6266491413116455, + "eval_runtime": 81.8129, + "eval_samples_per_second": 18.383, + "eval_steps_per_second": 2.298, + "step": 5867 + }, + { + "epoch": 0.25, + "grad_norm": 0.462890625, + "learning_rate": 0.00019260781074654238, + "loss": 2.6039, + "step": 5868 + }, + { + "epoch": 0.25, + "grad_norm": 0.44921875, + "learning_rate": 0.00019260527942185595, + "loss": 2.6164, + "step": 5869 + }, + { + "epoch": 0.25, + "grad_norm": 0.423828125, + "learning_rate": 0.00019260274768047965, + "loss": 2.7912, + "step": 5870 + }, + { + "epoch": 0.25, + "grad_norm": 0.41796875, + "learning_rate": 0.0001926002155224248, + "loss": 2.5674, + "step": 5871 + }, + { + "epoch": 0.25, + "grad_norm": 0.458984375, + "learning_rate": 0.0001925976829477028, + "loss": 2.5116, + "step": 5872 + }, + { + "epoch": 0.25, + "grad_norm": 0.466796875, + "learning_rate": 0.00019259514995632513, + "loss": 2.6302, + "step": 5873 + }, + { + "epoch": 0.25, + "grad_norm": 0.423828125, + "learning_rate": 0.00019259261654830308, + "loss": 2.6876, + "step": 5874 + }, + { + "epoch": 0.25, + "grad_norm": 0.390625, + "learning_rate": 0.0001925900827236481, + "loss": 2.6452, + "step": 5875 + }, + { + "epoch": 0.25, + "grad_norm": 0.486328125, + "learning_rate": 0.00019258754848237159, + "loss": 2.5754, + "step": 5876 + }, + { + "epoch": 0.25, + "grad_norm": 0.51171875, + "learning_rate": 0.00019258501382448496, + "loss": 2.6351, + "step": 5877 + }, + { + "epoch": 0.25, + "grad_norm": 0.4609375, + "learning_rate": 0.00019258247874999958, + "loss": 2.7991, + "step": 5878 + }, + { + "epoch": 0.25, + "grad_norm": 0.416015625, + "learning_rate": 0.0001925799432589269, + "loss": 2.6546, + "step": 5879 + }, + { + "epoch": 0.25, + "grad_norm": 0.474609375, + "learning_rate": 0.0001925774073512783, + "loss": 2.6872, + "step": 5880 + }, + { + "epoch": 0.25, + "grad_norm": 0.42578125, + "learning_rate": 0.0001925748710270652, + "loss": 2.7562, + "step": 5881 + }, + { + "epoch": 0.25, + "grad_norm": 0.45703125, + "learning_rate": 0.000192572334286299, + "loss": 2.697, + "step": 5882 + }, + { + "epoch": 0.25, + "grad_norm": 0.498046875, + "learning_rate": 0.00019256979712899117, + "loss": 2.6606, + "step": 5883 + }, + { + "epoch": 0.25, + "grad_norm": 0.46875, + "learning_rate": 0.00019256725955515305, + "loss": 2.7656, + "step": 5884 + }, + { + "epoch": 0.25, + "grad_norm": 0.375, + "learning_rate": 0.0001925647215647961, + "loss": 2.5934, + "step": 5885 + }, + { + "epoch": 0.25, + "grad_norm": 0.416015625, + "learning_rate": 0.0001925621831579317, + "loss": 2.6704, + "step": 5886 + }, + { + "epoch": 0.25, + "grad_norm": 0.4375, + "learning_rate": 0.00019255964433457135, + "loss": 2.6233, + "step": 5887 + }, + { + "epoch": 0.25, + "grad_norm": 0.404296875, + "learning_rate": 0.0001925571050947264, + "loss": 2.5504, + "step": 5888 + }, + { + "epoch": 0.25, + "grad_norm": 0.43359375, + "learning_rate": 0.0001925545654384083, + "loss": 2.6534, + "step": 5889 + }, + { + "epoch": 0.25, + "grad_norm": 0.431640625, + "learning_rate": 0.0001925520253656285, + "loss": 2.5553, + "step": 5890 + }, + { + "epoch": 0.25, + "grad_norm": 0.486328125, + "learning_rate": 0.00019254948487639842, + "loss": 2.6851, + "step": 5891 + }, + { + "epoch": 0.25, + "grad_norm": 0.474609375, + "learning_rate": 0.00019254694397072945, + "loss": 2.8624, + "step": 5892 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.00019254440264863308, + "loss": 2.4548, + "step": 5893 + }, + { + "epoch": 0.25, + "grad_norm": 0.44140625, + "learning_rate": 0.00019254186091012068, + "loss": 2.6843, + "step": 5894 + }, + { + "epoch": 0.25, + "grad_norm": 0.40625, + "learning_rate": 0.00019253931875520377, + "loss": 2.6157, + "step": 5895 + }, + { + "epoch": 0.25, + "grad_norm": 0.54296875, + "learning_rate": 0.00019253677618389373, + "loss": 2.6879, + "step": 5896 + }, + { + "epoch": 0.25, + "grad_norm": 0.45703125, + "learning_rate": 0.000192534233196202, + "loss": 2.5168, + "step": 5897 + }, + { + "epoch": 0.25, + "grad_norm": 0.474609375, + "learning_rate": 0.00019253168979214004, + "loss": 2.6703, + "step": 5898 + }, + { + "epoch": 0.25, + "grad_norm": 0.50390625, + "learning_rate": 0.00019252914597171934, + "loss": 2.6014, + "step": 5899 + }, + { + "epoch": 0.25, + "grad_norm": 0.421875, + "learning_rate": 0.00019252660173495125, + "loss": 2.5535, + "step": 5900 + }, + { + "epoch": 0.25, + "grad_norm": 0.42578125, + "learning_rate": 0.00019252405708184727, + "loss": 2.6391, + "step": 5901 + }, + { + "epoch": 0.25, + "grad_norm": 0.44140625, + "learning_rate": 0.00019252151201241887, + "loss": 2.5948, + "step": 5902 + }, + { + "epoch": 0.25, + "grad_norm": 0.46484375, + "learning_rate": 0.00019251896652667746, + "loss": 2.6326, + "step": 5903 + }, + { + "epoch": 0.25, + "grad_norm": 0.427734375, + "learning_rate": 0.0001925164206246345, + "loss": 2.5934, + "step": 5904 + }, + { + "epoch": 0.25, + "grad_norm": 0.47265625, + "learning_rate": 0.0001925138743063015, + "loss": 2.5557, + "step": 5905 + }, + { + "epoch": 0.25, + "grad_norm": 0.408203125, + "learning_rate": 0.00019251132757168983, + "loss": 2.6122, + "step": 5906 + }, + { + "epoch": 0.25, + "grad_norm": 0.408203125, + "learning_rate": 0.00019250878042081102, + "loss": 2.6725, + "step": 5907 + }, + { + "epoch": 0.25, + "grad_norm": 0.49609375, + "learning_rate": 0.0001925062328536765, + "loss": 2.6493, + "step": 5908 + }, + { + "epoch": 0.25, + "grad_norm": 0.46484375, + "learning_rate": 0.00019250368487029772, + "loss": 2.6748, + "step": 5909 + }, + { + "epoch": 0.25, + "grad_norm": 0.455078125, + "learning_rate": 0.0001925011364706862, + "loss": 2.7247, + "step": 5910 + }, + { + "epoch": 0.25, + "grad_norm": 0.431640625, + "learning_rate": 0.00019249858765485337, + "loss": 2.6015, + "step": 5911 + }, + { + "epoch": 0.25, + "grad_norm": 0.443359375, + "learning_rate": 0.00019249603842281068, + "loss": 2.648, + "step": 5912 + }, + { + "epoch": 0.25, + "grad_norm": 0.40234375, + "learning_rate": 0.0001924934887745696, + "loss": 2.6281, + "step": 5913 + }, + { + "epoch": 0.25, + "grad_norm": 0.38671875, + "learning_rate": 0.00019249093871014168, + "loss": 2.6149, + "step": 5914 + }, + { + "epoch": 0.25, + "grad_norm": 0.421875, + "learning_rate": 0.0001924883882295383, + "loss": 2.5821, + "step": 5915 + }, + { + "epoch": 0.25, + "grad_norm": 0.390625, + "learning_rate": 0.00019248583733277096, + "loss": 2.6521, + "step": 5916 + }, + { + "epoch": 0.25, + "grad_norm": 0.443359375, + "learning_rate": 0.0001924832860198512, + "loss": 2.4991, + "step": 5917 + }, + { + "epoch": 0.25, + "grad_norm": 0.474609375, + "learning_rate": 0.00019248073429079044, + "loss": 2.6336, + "step": 5918 + }, + { + "epoch": 0.25, + "grad_norm": 0.388671875, + "learning_rate": 0.00019247818214560014, + "loss": 2.5557, + "step": 5919 + }, + { + "epoch": 0.25, + "grad_norm": 0.45703125, + "learning_rate": 0.00019247562958429182, + "loss": 2.6727, + "step": 5920 + }, + { + "epoch": 0.25, + "grad_norm": 0.458984375, + "learning_rate": 0.00019247307660687698, + "loss": 2.7007, + "step": 5921 + }, + { + "epoch": 0.25, + "grad_norm": 0.447265625, + "learning_rate": 0.0001924705232133671, + "loss": 2.7634, + "step": 5922 + }, + { + "epoch": 0.25, + "grad_norm": 0.435546875, + "learning_rate": 0.00019246796940377366, + "loss": 2.7539, + "step": 5923 + }, + { + "epoch": 0.25, + "grad_norm": 0.46484375, + "learning_rate": 0.00019246541517810813, + "loss": 2.5918, + "step": 5924 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.00019246286053638202, + "loss": 2.632, + "step": 5925 + }, + { + "epoch": 0.25, + "grad_norm": 0.50390625, + "learning_rate": 0.00019246030547860686, + "loss": 2.6872, + "step": 5926 + }, + { + "epoch": 0.25, + "grad_norm": 0.435546875, + "learning_rate": 0.00019245775000479407, + "loss": 2.7603, + "step": 5927 + }, + { + "epoch": 0.25, + "grad_norm": 0.421875, + "learning_rate": 0.00019245519411495522, + "loss": 2.4261, + "step": 5928 + }, + { + "epoch": 0.25, + "grad_norm": 0.48046875, + "learning_rate": 0.00019245263780910178, + "loss": 2.6907, + "step": 5929 + }, + { + "epoch": 0.25, + "grad_norm": 0.40234375, + "learning_rate": 0.0001924500810872453, + "loss": 2.715, + "step": 5930 + }, + { + "epoch": 0.25, + "grad_norm": 0.388671875, + "learning_rate": 0.0001924475239493972, + "loss": 2.5949, + "step": 5931 + }, + { + "epoch": 0.25, + "grad_norm": 0.458984375, + "learning_rate": 0.000192444966395569, + "loss": 2.6212, + "step": 5932 + }, + { + "epoch": 0.25, + "grad_norm": 0.5234375, + "learning_rate": 0.00019244240842577229, + "loss": 2.5936, + "step": 5933 + }, + { + "epoch": 0.25, + "grad_norm": 0.42578125, + "learning_rate": 0.0001924398500400185, + "loss": 2.5997, + "step": 5934 + }, + { + "epoch": 0.25, + "grad_norm": 0.419921875, + "learning_rate": 0.00019243729123831916, + "loss": 2.6079, + "step": 5935 + }, + { + "epoch": 0.25, + "grad_norm": 0.45703125, + "learning_rate": 0.00019243473202068583, + "loss": 2.8517, + "step": 5936 + }, + { + "epoch": 0.25, + "grad_norm": 0.4609375, + "learning_rate": 0.00019243217238712994, + "loss": 2.7359, + "step": 5937 + }, + { + "epoch": 0.25, + "grad_norm": 0.447265625, + "learning_rate": 0.0001924296123376631, + "loss": 2.6648, + "step": 5938 + }, + { + "epoch": 0.25, + "grad_norm": 0.423828125, + "learning_rate": 0.00019242705187229675, + "loss": 2.737, + "step": 5939 + }, + { + "epoch": 0.25, + "grad_norm": 0.40234375, + "learning_rate": 0.00019242449099104246, + "loss": 2.6063, + "step": 5940 + }, + { + "epoch": 0.25, + "grad_norm": 0.5, + "learning_rate": 0.00019242192969391172, + "loss": 2.7035, + "step": 5941 + }, + { + "epoch": 0.25, + "grad_norm": 0.466796875, + "learning_rate": 0.0001924193679809161, + "loss": 2.7083, + "step": 5942 + }, + { + "epoch": 0.25, + "grad_norm": 0.451171875, + "learning_rate": 0.00019241680585206706, + "loss": 2.5904, + "step": 5943 + }, + { + "epoch": 0.25, + "grad_norm": 0.42578125, + "learning_rate": 0.00019241424330737622, + "loss": 2.619, + "step": 5944 + }, + { + "epoch": 0.25, + "grad_norm": 0.466796875, + "learning_rate": 0.00019241168034685502, + "loss": 2.6851, + "step": 5945 + }, + { + "epoch": 0.25, + "grad_norm": 0.38671875, + "learning_rate": 0.00019240911697051503, + "loss": 2.5116, + "step": 5946 + }, + { + "epoch": 0.25, + "grad_norm": 0.451171875, + "learning_rate": 0.00019240655317836783, + "loss": 2.7853, + "step": 5947 + }, + { + "epoch": 0.25, + "grad_norm": 0.47265625, + "learning_rate": 0.00019240398897042488, + "loss": 2.6875, + "step": 5948 + }, + { + "epoch": 0.25, + "grad_norm": 0.50390625, + "learning_rate": 0.00019240142434669773, + "loss": 2.8356, + "step": 5949 + }, + { + "epoch": 0.25, + "grad_norm": 0.44921875, + "learning_rate": 0.00019239885930719797, + "loss": 2.6662, + "step": 5950 + }, + { + "epoch": 0.25, + "grad_norm": 0.416015625, + "learning_rate": 0.00019239629385193707, + "loss": 2.5649, + "step": 5951 + }, + { + "epoch": 0.25, + "grad_norm": 0.42578125, + "learning_rate": 0.00019239372798092667, + "loss": 2.5599, + "step": 5952 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.0001923911616941782, + "loss": 2.7163, + "step": 5953 + }, + { + "epoch": 0.25, + "grad_norm": 0.482421875, + "learning_rate": 0.00019238859499170333, + "loss": 2.6579, + "step": 5954 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.00019238602787351349, + "loss": 2.6673, + "step": 5955 + }, + { + "epoch": 0.25, + "grad_norm": 0.4140625, + "learning_rate": 0.0001923834603396203, + "loss": 2.6287, + "step": 5956 + }, + { + "epoch": 0.25, + "grad_norm": 0.423828125, + "learning_rate": 0.00019238089239003535, + "loss": 2.5541, + "step": 5957 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019237832402477012, + "loss": 2.7598, + "step": 5958 + }, + { + "epoch": 0.25, + "grad_norm": 0.427734375, + "learning_rate": 0.00019237575524383617, + "loss": 2.7387, + "step": 5959 + }, + { + "epoch": 0.25, + "grad_norm": 0.443359375, + "learning_rate": 0.00019237318604724507, + "loss": 2.701, + "step": 5960 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019237061643500846, + "loss": 2.755, + "step": 5961 + }, + { + "epoch": 0.25, + "grad_norm": 0.4921875, + "learning_rate": 0.00019236804640713776, + "loss": 2.658, + "step": 5962 + }, + { + "epoch": 0.25, + "grad_norm": 0.453125, + "learning_rate": 0.00019236547596364464, + "loss": 2.6489, + "step": 5963 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.00019236290510454062, + "loss": 2.6096, + "step": 5964 + }, + { + "epoch": 0.25, + "grad_norm": 0.396484375, + "learning_rate": 0.00019236033382983732, + "loss": 2.6408, + "step": 5965 + }, + { + "epoch": 0.25, + "grad_norm": 0.484375, + "learning_rate": 0.00019235776213954623, + "loss": 2.6788, + "step": 5966 + }, + { + "epoch": 0.25, + "grad_norm": 0.478515625, + "learning_rate": 0.000192355190033679, + "loss": 2.7158, + "step": 5967 + }, + { + "epoch": 0.25, + "grad_norm": 0.478515625, + "learning_rate": 0.00019235261751224714, + "loss": 2.5735, + "step": 5968 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.00019235004457526225, + "loss": 2.7075, + "step": 5969 + }, + { + "epoch": 0.25, + "grad_norm": 0.416015625, + "learning_rate": 0.0001923474712227359, + "loss": 2.6119, + "step": 5970 + }, + { + "epoch": 0.25, + "grad_norm": 0.44140625, + "learning_rate": 0.00019234489745467972, + "loss": 2.4482, + "step": 5971 + }, + { + "epoch": 0.25, + "grad_norm": 0.470703125, + "learning_rate": 0.00019234232327110522, + "loss": 2.5469, + "step": 5972 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.000192339748672024, + "loss": 2.6064, + "step": 5973 + }, + { + "epoch": 0.25, + "grad_norm": 0.4453125, + "learning_rate": 0.0001923371736574477, + "loss": 2.5766, + "step": 5974 + }, + { + "epoch": 0.25, + "grad_norm": 0.490234375, + "learning_rate": 0.00019233459822738783, + "loss": 2.6773, + "step": 5975 + }, + { + "epoch": 0.25, + "grad_norm": 0.4296875, + "learning_rate": 0.00019233202238185603, + "loss": 2.5201, + "step": 5976 + }, + { + "epoch": 0.25, + "grad_norm": 0.46875, + "learning_rate": 0.00019232944612086386, + "loss": 2.6002, + "step": 5977 + }, + { + "epoch": 0.25, + "grad_norm": 0.4375, + "learning_rate": 0.0001923268694444229, + "loss": 2.6837, + "step": 5978 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019232429235254483, + "loss": 2.687, + "step": 5979 + }, + { + "epoch": 0.25, + "grad_norm": 0.439453125, + "learning_rate": 0.00019232171484524115, + "loss": 2.5587, + "step": 5980 + }, + { + "epoch": 0.25, + "grad_norm": 0.4375, + "learning_rate": 0.0001923191369225235, + "loss": 2.6667, + "step": 5981 + }, + { + "epoch": 0.25, + "grad_norm": 0.416015625, + "learning_rate": 0.00019231655858440347, + "loss": 2.674, + "step": 5982 + }, + { + "epoch": 0.25, + "grad_norm": 0.458984375, + "learning_rate": 0.00019231397983089266, + "loss": 2.6943, + "step": 5983 + }, + { + "epoch": 0.25, + "grad_norm": 0.44140625, + "learning_rate": 0.00019231140066200268, + "loss": 2.6386, + "step": 5984 + }, + { + "epoch": 0.26, + "grad_norm": 0.419921875, + "learning_rate": 0.00019230882107774513, + "loss": 2.6131, + "step": 5985 + }, + { + "epoch": 0.26, + "grad_norm": 0.5078125, + "learning_rate": 0.00019230624107813163, + "loss": 2.7818, + "step": 5986 + }, + { + "epoch": 0.26, + "grad_norm": 0.408203125, + "learning_rate": 0.00019230366066317375, + "loss": 2.7476, + "step": 5987 + }, + { + "epoch": 0.26, + "grad_norm": 0.455078125, + "learning_rate": 0.00019230107983288317, + "loss": 2.5887, + "step": 5988 + }, + { + "epoch": 0.26, + "grad_norm": 0.44921875, + "learning_rate": 0.00019229849858727144, + "loss": 2.8013, + "step": 5989 + }, + { + "epoch": 0.26, + "grad_norm": 0.40625, + "learning_rate": 0.0001922959169263502, + "loss": 2.7593, + "step": 5990 + }, + { + "epoch": 0.26, + "grad_norm": 0.47265625, + "learning_rate": 0.00019229333485013109, + "loss": 2.6485, + "step": 5991 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.00019229075235862567, + "loss": 2.8098, + "step": 5992 + }, + { + "epoch": 0.26, + "grad_norm": 0.435546875, + "learning_rate": 0.0001922881694518456, + "loss": 2.652, + "step": 5993 + }, + { + "epoch": 0.26, + "grad_norm": 0.4609375, + "learning_rate": 0.0001922855861298025, + "loss": 2.635, + "step": 5994 + }, + { + "epoch": 0.26, + "grad_norm": 0.4921875, + "learning_rate": 0.00019228300239250797, + "loss": 2.7425, + "step": 5995 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.00019228041823997368, + "loss": 2.6877, + "step": 5996 + }, + { + "epoch": 0.26, + "grad_norm": 0.478515625, + "learning_rate": 0.0001922778336722112, + "loss": 2.7727, + "step": 5997 + }, + { + "epoch": 0.26, + "grad_norm": 0.4296875, + "learning_rate": 0.00019227524868923226, + "loss": 2.7047, + "step": 5998 + }, + { + "epoch": 0.26, + "grad_norm": 0.423828125, + "learning_rate": 0.00019227266329104837, + "loss": 2.7127, + "step": 5999 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.0001922700774776712, + "loss": 2.7264, + "step": 6000 + }, + { + "epoch": 0.26, + "grad_norm": 0.416015625, + "learning_rate": 0.00019226749124911244, + "loss": 2.5655, + "step": 6001 + }, + { + "epoch": 0.26, + "grad_norm": 0.48828125, + "learning_rate": 0.0001922649046053837, + "loss": 2.7121, + "step": 6002 + }, + { + "epoch": 0.26, + "grad_norm": 0.5078125, + "learning_rate": 0.0001922623175464966, + "loss": 2.6554, + "step": 6003 + }, + { + "epoch": 0.26, + "grad_norm": 0.423828125, + "learning_rate": 0.00019225973007246276, + "loss": 2.7424, + "step": 6004 + }, + { + "epoch": 0.26, + "grad_norm": 0.421875, + "learning_rate": 0.00019225714218329385, + "loss": 2.6268, + "step": 6005 + }, + { + "epoch": 0.26, + "grad_norm": 0.490234375, + "learning_rate": 0.00019225455387900152, + "loss": 2.7254, + "step": 6006 + }, + { + "epoch": 0.26, + "grad_norm": 0.4921875, + "learning_rate": 0.00019225196515959745, + "loss": 2.5963, + "step": 6007 + }, + { + "epoch": 0.26, + "grad_norm": 0.41015625, + "learning_rate": 0.00019224937602509322, + "loss": 2.7016, + "step": 6008 + }, + { + "epoch": 0.26, + "grad_norm": 0.419921875, + "learning_rate": 0.00019224678647550053, + "loss": 2.6573, + "step": 6009 + }, + { + "epoch": 0.26, + "grad_norm": 0.427734375, + "learning_rate": 0.000192244196510831, + "loss": 2.5903, + "step": 6010 + }, + { + "epoch": 0.26, + "grad_norm": 0.43359375, + "learning_rate": 0.0001922416061310963, + "loss": 2.5584, + "step": 6011 + }, + { + "epoch": 0.26, + "grad_norm": 0.421875, + "learning_rate": 0.00019223901533630807, + "loss": 2.6342, + "step": 6012 + }, + { + "epoch": 0.26, + "grad_norm": 0.4765625, + "learning_rate": 0.00019223642412647803, + "loss": 2.6177, + "step": 6013 + }, + { + "epoch": 0.26, + "grad_norm": 0.42578125, + "learning_rate": 0.00019223383250161773, + "loss": 2.7223, + "step": 6014 + }, + { + "epoch": 0.26, + "grad_norm": 0.4296875, + "learning_rate": 0.00019223124046173894, + "loss": 2.7489, + "step": 6015 + }, + { + "epoch": 0.26, + "grad_norm": 0.369140625, + "learning_rate": 0.00019222864800685322, + "loss": 2.4799, + "step": 6016 + }, + { + "epoch": 0.26, + "grad_norm": 0.50390625, + "learning_rate": 0.0001922260551369723, + "loss": 2.7634, + "step": 6017 + }, + { + "epoch": 0.26, + "grad_norm": 0.421875, + "learning_rate": 0.0001922234618521079, + "loss": 2.5853, + "step": 6018 + }, + { + "epoch": 0.26, + "grad_norm": 0.396484375, + "learning_rate": 0.00019222086815227156, + "loss": 2.5536, + "step": 6019 + }, + { + "epoch": 0.26, + "grad_norm": 0.3984375, + "learning_rate": 0.00019221827403747503, + "loss": 2.5812, + "step": 6020 + }, + { + "epoch": 0.26, + "grad_norm": 0.49609375, + "learning_rate": 0.00019221567950772998, + "loss": 2.6426, + "step": 6021 + }, + { + "epoch": 0.26, + "grad_norm": 0.55859375, + "learning_rate": 0.00019221308456304805, + "loss": 2.5626, + "step": 6022 + }, + { + "epoch": 0.26, + "grad_norm": 0.404296875, + "learning_rate": 0.00019221048920344094, + "loss": 2.5677, + "step": 6023 + }, + { + "epoch": 0.26, + "grad_norm": 0.5390625, + "learning_rate": 0.00019220789342892038, + "loss": 2.6068, + "step": 6024 + }, + { + "epoch": 0.26, + "grad_norm": 0.453125, + "learning_rate": 0.00019220529723949797, + "loss": 2.5786, + "step": 6025 + }, + { + "epoch": 0.26, + "grad_norm": 0.392578125, + "learning_rate": 0.0001922027006351854, + "loss": 2.6184, + "step": 6026 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.00019220010361599438, + "loss": 2.6093, + "step": 6027 + }, + { + "epoch": 0.26, + "grad_norm": 0.451171875, + "learning_rate": 0.0001921975061819366, + "loss": 2.6616, + "step": 6028 + }, + { + "epoch": 0.26, + "grad_norm": 0.453125, + "learning_rate": 0.00019219490833302373, + "loss": 2.7256, + "step": 6029 + }, + { + "epoch": 0.26, + "grad_norm": 0.50390625, + "learning_rate": 0.00019219231006926748, + "loss": 2.6908, + "step": 6030 + }, + { + "epoch": 0.26, + "grad_norm": 0.455078125, + "learning_rate": 0.00019218971139067952, + "loss": 2.545, + "step": 6031 + }, + { + "epoch": 0.26, + "grad_norm": 0.51171875, + "learning_rate": 0.00019218711229727152, + "loss": 2.8297, + "step": 6032 + }, + { + "epoch": 0.26, + "grad_norm": 0.46875, + "learning_rate": 0.00019218451278905525, + "loss": 2.6833, + "step": 6033 + }, + { + "epoch": 0.26, + "grad_norm": 0.466796875, + "learning_rate": 0.00019218191286604234, + "loss": 2.6585, + "step": 6034 + }, + { + "epoch": 0.26, + "grad_norm": 0.43359375, + "learning_rate": 0.0001921793125282445, + "loss": 2.7878, + "step": 6035 + }, + { + "epoch": 0.26, + "grad_norm": 0.41015625, + "learning_rate": 0.00019217671177567346, + "loss": 2.6884, + "step": 6036 + }, + { + "epoch": 0.26, + "grad_norm": 0.4765625, + "learning_rate": 0.0001921741106083409, + "loss": 2.7326, + "step": 6037 + }, + { + "epoch": 0.26, + "grad_norm": 0.431640625, + "learning_rate": 0.00019217150902625852, + "loss": 2.6603, + "step": 6038 + }, + { + "epoch": 0.26, + "grad_norm": 0.388671875, + "learning_rate": 0.00019216890702943804, + "loss": 2.6349, + "step": 6039 + }, + { + "epoch": 0.26, + "grad_norm": 0.48046875, + "learning_rate": 0.00019216630461789115, + "loss": 2.5449, + "step": 6040 + }, + { + "epoch": 0.26, + "grad_norm": 0.427734375, + "learning_rate": 0.0001921637017916296, + "loss": 2.7348, + "step": 6041 + }, + { + "epoch": 0.26, + "grad_norm": 0.39453125, + "learning_rate": 0.00019216109855066507, + "loss": 2.6203, + "step": 6042 + }, + { + "epoch": 0.26, + "grad_norm": 0.451171875, + "learning_rate": 0.00019215849489500925, + "loss": 2.7006, + "step": 6043 + }, + { + "epoch": 0.26, + "grad_norm": 0.451171875, + "learning_rate": 0.0001921558908246739, + "loss": 2.5977, + "step": 6044 + }, + { + "epoch": 0.26, + "grad_norm": 0.4375, + "learning_rate": 0.00019215328633967073, + "loss": 2.6445, + "step": 6045 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.00019215068144001144, + "loss": 2.6867, + "step": 6046 + }, + { + "epoch": 0.26, + "grad_norm": 0.42578125, + "learning_rate": 0.00019214807612570777, + "loss": 2.5709, + "step": 6047 + }, + { + "epoch": 0.26, + "grad_norm": 0.4140625, + "learning_rate": 0.00019214547039677143, + "loss": 2.4815, + "step": 6048 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.00019214286425321412, + "loss": 2.6083, + "step": 6049 + }, + { + "epoch": 0.26, + "grad_norm": 0.4765625, + "learning_rate": 0.00019214025769504763, + "loss": 2.7386, + "step": 6050 + }, + { + "epoch": 0.26, + "grad_norm": 0.453125, + "learning_rate": 0.00019213765072228361, + "loss": 2.5946, + "step": 6051 + }, + { + "epoch": 0.26, + "grad_norm": 0.470703125, + "learning_rate": 0.00019213504333493388, + "loss": 2.8192, + "step": 6052 + }, + { + "epoch": 0.26, + "grad_norm": 0.40625, + "learning_rate": 0.00019213243553301013, + "loss": 2.5544, + "step": 6053 + }, + { + "epoch": 0.26, + "grad_norm": 0.45703125, + "learning_rate": 0.00019212982731652407, + "loss": 2.6176, + "step": 6054 + }, + { + "epoch": 0.26, + "grad_norm": 0.466796875, + "learning_rate": 0.00019212721868548744, + "loss": 2.5972, + "step": 6055 + }, + { + "epoch": 0.26, + "grad_norm": 0.392578125, + "learning_rate": 0.00019212460963991204, + "loss": 2.569, + "step": 6056 + }, + { + "epoch": 0.26, + "grad_norm": 0.404296875, + "learning_rate": 0.00019212200017980954, + "loss": 2.6259, + "step": 6057 + }, + { + "epoch": 0.26, + "grad_norm": 0.4609375, + "learning_rate": 0.0001921193903051917, + "loss": 2.5606, + "step": 6058 + }, + { + "epoch": 0.26, + "grad_norm": 0.423828125, + "learning_rate": 0.00019211678001607028, + "loss": 2.6064, + "step": 6059 + }, + { + "epoch": 0.26, + "grad_norm": 0.37109375, + "learning_rate": 0.00019211416931245697, + "loss": 2.532, + "step": 6060 + }, + { + "epoch": 0.26, + "grad_norm": 0.50390625, + "learning_rate": 0.00019211155819436363, + "loss": 2.726, + "step": 6061 + }, + { + "epoch": 0.26, + "grad_norm": 0.455078125, + "learning_rate": 0.0001921089466618019, + "loss": 2.8122, + "step": 6062 + }, + { + "epoch": 0.26, + "grad_norm": 0.490234375, + "learning_rate": 0.00019210633471478358, + "loss": 2.6255, + "step": 6063 + }, + { + "epoch": 0.26, + "grad_norm": 0.392578125, + "learning_rate": 0.00019210372235332042, + "loss": 2.7339, + "step": 6064 + }, + { + "epoch": 0.26, + "grad_norm": 0.40625, + "learning_rate": 0.00019210110957742417, + "loss": 2.69, + "step": 6065 + }, + { + "epoch": 0.26, + "grad_norm": 0.439453125, + "learning_rate": 0.0001920984963871066, + "loss": 2.8813, + "step": 6066 + }, + { + "epoch": 0.26, + "grad_norm": 0.48046875, + "learning_rate": 0.00019209588278237943, + "loss": 2.714, + "step": 6067 + }, + { + "epoch": 0.26, + "grad_norm": 0.447265625, + "learning_rate": 0.00019209326876325444, + "loss": 2.7032, + "step": 6068 + }, + { + "epoch": 0.26, + "grad_norm": 0.4375, + "learning_rate": 0.00019209065432974342, + "loss": 2.4927, + "step": 6069 + }, + { + "epoch": 0.26, + "grad_norm": 0.484375, + "learning_rate": 0.00019208803948185809, + "loss": 2.7097, + "step": 6070 + }, + { + "epoch": 0.26, + "grad_norm": 0.43359375, + "learning_rate": 0.00019208542421961027, + "loss": 2.5907, + "step": 6071 + }, + { + "epoch": 0.26, + "grad_norm": 0.39453125, + "learning_rate": 0.00019208280854301167, + "loss": 2.5352, + "step": 6072 + }, + { + "epoch": 0.26, + "grad_norm": 0.43359375, + "learning_rate": 0.00019208019245207407, + "loss": 2.583, + "step": 6073 + }, + { + "epoch": 0.26, + "grad_norm": 0.42578125, + "learning_rate": 0.00019207757594680929, + "loss": 2.642, + "step": 6074 + }, + { + "epoch": 0.26, + "grad_norm": 0.408203125, + "learning_rate": 0.00019207495902722903, + "loss": 2.6501, + "step": 6075 + }, + { + "epoch": 0.26, + "grad_norm": 0.435546875, + "learning_rate": 0.00019207234169334511, + "loss": 2.5881, + "step": 6076 + }, + { + "epoch": 0.26, + "grad_norm": 0.396484375, + "learning_rate": 0.00019206972394516934, + "loss": 2.692, + "step": 6077 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.00019206710578271343, + "loss": 2.5606, + "step": 6078 + }, + { + "epoch": 0.26, + "grad_norm": 0.4765625, + "learning_rate": 0.00019206448720598919, + "loss": 2.6778, + "step": 6079 + }, + { + "epoch": 0.26, + "grad_norm": 0.51171875, + "learning_rate": 0.0001920618682150084, + "loss": 2.5836, + "step": 6080 + }, + { + "epoch": 0.26, + "grad_norm": 0.458984375, + "learning_rate": 0.00019205924880978285, + "loss": 2.624, + "step": 6081 + }, + { + "epoch": 0.26, + "grad_norm": 0.44921875, + "learning_rate": 0.00019205662899032434, + "loss": 2.656, + "step": 6082 + }, + { + "epoch": 0.26, + "grad_norm": 0.498046875, + "learning_rate": 0.00019205400875664462, + "loss": 2.636, + "step": 6083 + }, + { + "epoch": 0.26, + "grad_norm": 0.439453125, + "learning_rate": 0.00019205138810875552, + "loss": 2.677, + "step": 6084 + }, + { + "epoch": 0.26, + "grad_norm": 0.484375, + "learning_rate": 0.0001920487670466688, + "loss": 2.6194, + "step": 6085 + }, + { + "epoch": 0.26, + "grad_norm": 0.453125, + "learning_rate": 0.0001920461455703963, + "loss": 2.61, + "step": 6086 + }, + { + "epoch": 0.26, + "grad_norm": 0.443359375, + "learning_rate": 0.00019204352367994975, + "loss": 2.821, + "step": 6087 + }, + { + "epoch": 0.26, + "grad_norm": 0.427734375, + "learning_rate": 0.000192040901375341, + "loss": 2.6396, + "step": 6088 + }, + { + "epoch": 0.26, + "grad_norm": 0.470703125, + "learning_rate": 0.00019203827865658184, + "loss": 2.704, + "step": 6089 + }, + { + "epoch": 0.26, + "grad_norm": 0.5234375, + "learning_rate": 0.00019203565552368404, + "loss": 2.6256, + "step": 6090 + }, + { + "epoch": 0.26, + "grad_norm": 0.443359375, + "learning_rate": 0.0001920330319766594, + "loss": 2.604, + "step": 6091 + }, + { + "epoch": 0.26, + "grad_norm": 0.4609375, + "learning_rate": 0.0001920304080155198, + "loss": 2.634, + "step": 6092 + }, + { + "epoch": 0.26, + "grad_norm": 0.404296875, + "learning_rate": 0.000192027783640277, + "loss": 2.7025, + "step": 6093 + }, + { + "epoch": 0.26, + "grad_norm": 0.46875, + "learning_rate": 0.00019202515885094275, + "loss": 2.5877, + "step": 6094 + }, + { + "epoch": 0.26, + "grad_norm": 0.490234375, + "learning_rate": 0.00019202253364752897, + "loss": 2.7166, + "step": 6095 + }, + { + "epoch": 0.26, + "grad_norm": 0.45703125, + "learning_rate": 0.00019201990803004743, + "loss": 2.6819, + "step": 6096 + }, + { + "epoch": 0.26, + "grad_norm": 0.443359375, + "learning_rate": 0.00019201728199850989, + "loss": 2.6537, + "step": 6097 + }, + { + "epoch": 0.26, + "grad_norm": 0.388671875, + "learning_rate": 0.00019201465555292822, + "loss": 2.5347, + "step": 6098 + }, + { + "epoch": 0.26, + "grad_norm": 0.419921875, + "learning_rate": 0.0001920120286933142, + "loss": 2.649, + "step": 6099 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.00019200940141967973, + "loss": 2.5392, + "step": 6100 + }, + { + "epoch": 0.26, + "grad_norm": 0.427734375, + "learning_rate": 0.00019200677373203655, + "loss": 2.6671, + "step": 6101 + }, + { + "epoch": 0.26, + "grad_norm": 0.40625, + "learning_rate": 0.00019200414563039654, + "loss": 2.4805, + "step": 6102 + }, + { + "epoch": 0.26, + "grad_norm": 0.49609375, + "learning_rate": 0.00019200151711477146, + "loss": 2.6151, + "step": 6103 + }, + { + "epoch": 0.26, + "grad_norm": 0.48828125, + "learning_rate": 0.0001919988881851732, + "loss": 2.6353, + "step": 6104 + }, + { + "epoch": 0.26, + "grad_norm": 0.47265625, + "learning_rate": 0.00019199625884161353, + "loss": 2.5943, + "step": 6105 + }, + { + "epoch": 0.26, + "grad_norm": 0.45703125, + "learning_rate": 0.00019199362908410436, + "loss": 2.5778, + "step": 6106 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.0001919909989126575, + "loss": 2.5412, + "step": 6107 + }, + { + "epoch": 0.26, + "grad_norm": 0.38671875, + "learning_rate": 0.0001919883683272847, + "loss": 2.468, + "step": 6108 + }, + { + "epoch": 0.26, + "grad_norm": 0.421875, + "learning_rate": 0.0001919857373279979, + "loss": 2.5967, + "step": 6109 + }, + { + "epoch": 0.26, + "grad_norm": 0.416015625, + "learning_rate": 0.0001919831059148089, + "loss": 2.6489, + "step": 6110 + }, + { + "epoch": 0.26, + "grad_norm": 0.462890625, + "learning_rate": 0.0001919804740877295, + "loss": 2.6422, + "step": 6111 + }, + { + "epoch": 0.26, + "grad_norm": 0.439453125, + "learning_rate": 0.00019197784184677162, + "loss": 2.7482, + "step": 6112 + }, + { + "epoch": 0.26, + "grad_norm": 0.423828125, + "learning_rate": 0.00019197520919194707, + "loss": 2.7365, + "step": 6113 + }, + { + "epoch": 0.26, + "grad_norm": 0.408203125, + "learning_rate": 0.00019197257612326766, + "loss": 2.5644, + "step": 6114 + }, + { + "epoch": 0.26, + "grad_norm": 0.458984375, + "learning_rate": 0.00019196994264074528, + "loss": 2.8088, + "step": 6115 + }, + { + "epoch": 0.26, + "grad_norm": 0.416015625, + "learning_rate": 0.0001919673087443918, + "loss": 2.6248, + "step": 6116 + }, + { + "epoch": 0.26, + "grad_norm": 0.42578125, + "learning_rate": 0.00019196467443421898, + "loss": 2.6797, + "step": 6117 + }, + { + "epoch": 0.26, + "grad_norm": 0.458984375, + "learning_rate": 0.0001919620397102388, + "loss": 2.7137, + "step": 6118 + }, + { + "epoch": 0.26, + "grad_norm": 0.453125, + "learning_rate": 0.000191959404572463, + "loss": 2.6928, + "step": 6119 + }, + { + "epoch": 0.26, + "grad_norm": 0.458984375, + "learning_rate": 0.0001919567690209035, + "loss": 2.7325, + "step": 6120 + }, + { + "epoch": 0.26, + "grad_norm": 0.490234375, + "learning_rate": 0.00019195413305557216, + "loss": 2.8874, + "step": 6121 + }, + { + "epoch": 0.26, + "grad_norm": 0.4765625, + "learning_rate": 0.00019195149667648083, + "loss": 2.743, + "step": 6122 + }, + { + "epoch": 0.26, + "grad_norm": 0.482421875, + "learning_rate": 0.00019194885988364134, + "loss": 2.5625, + "step": 6123 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.0001919462226770656, + "loss": 2.5666, + "step": 6124 + }, + { + "epoch": 0.26, + "grad_norm": 0.486328125, + "learning_rate": 0.00019194358505676546, + "loss": 2.6611, + "step": 6125 + }, + { + "epoch": 0.26, + "grad_norm": 0.39453125, + "learning_rate": 0.00019194094702275282, + "loss": 2.6234, + "step": 6126 + }, + { + "epoch": 0.26, + "grad_norm": 0.416015625, + "learning_rate": 0.00019193830857503947, + "loss": 2.4836, + "step": 6127 + }, + { + "epoch": 0.26, + "grad_norm": 0.41015625, + "learning_rate": 0.00019193566971363737, + "loss": 2.5919, + "step": 6128 + }, + { + "epoch": 0.26, + "grad_norm": 0.443359375, + "learning_rate": 0.00019193303043855834, + "loss": 2.5976, + "step": 6129 + }, + { + "epoch": 0.26, + "grad_norm": 0.3828125, + "learning_rate": 0.00019193039074981425, + "loss": 2.5203, + "step": 6130 + }, + { + "epoch": 0.26, + "grad_norm": 0.427734375, + "learning_rate": 0.00019192775064741702, + "loss": 2.7204, + "step": 6131 + }, + { + "epoch": 0.26, + "grad_norm": 0.47265625, + "learning_rate": 0.0001919251101313785, + "loss": 2.8028, + "step": 6132 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.00019192246920171058, + "loss": 2.5781, + "step": 6133 + }, + { + "epoch": 0.26, + "grad_norm": 0.4296875, + "learning_rate": 0.00019191982785842515, + "loss": 2.5593, + "step": 6134 + }, + { + "epoch": 0.26, + "grad_norm": 0.41015625, + "learning_rate": 0.00019191718610153408, + "loss": 2.6789, + "step": 6135 + }, + { + "epoch": 0.26, + "grad_norm": 0.431640625, + "learning_rate": 0.00019191454393104927, + "loss": 2.5992, + "step": 6136 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.0001919119013469826, + "loss": 2.7194, + "step": 6137 + }, + { + "epoch": 0.26, + "grad_norm": 0.40625, + "learning_rate": 0.00019190925834934594, + "loss": 2.5648, + "step": 6138 + }, + { + "epoch": 0.26, + "grad_norm": 0.42578125, + "learning_rate": 0.00019190661493815125, + "loss": 2.6165, + "step": 6139 + }, + { + "epoch": 0.26, + "grad_norm": 0.5, + "learning_rate": 0.00019190397111341033, + "loss": 2.6613, + "step": 6140 + }, + { + "epoch": 0.26, + "grad_norm": 0.41015625, + "learning_rate": 0.0001919013268751352, + "loss": 2.6028, + "step": 6141 + }, + { + "epoch": 0.26, + "grad_norm": 0.482421875, + "learning_rate": 0.00019189868222333763, + "loss": 2.5816, + "step": 6142 + }, + { + "epoch": 0.26, + "grad_norm": 0.4375, + "learning_rate": 0.00019189603715802958, + "loss": 2.6736, + "step": 6143 + }, + { + "epoch": 0.26, + "grad_norm": 0.390625, + "learning_rate": 0.00019189339167922296, + "loss": 2.5833, + "step": 6144 + }, + { + "epoch": 0.26, + "grad_norm": 0.5390625, + "learning_rate": 0.00019189074578692967, + "loss": 2.6211, + "step": 6145 + }, + { + "epoch": 0.26, + "grad_norm": 0.45703125, + "learning_rate": 0.00019188809948116157, + "loss": 2.7401, + "step": 6146 + }, + { + "epoch": 0.26, + "grad_norm": 0.423828125, + "learning_rate": 0.00019188545276193063, + "loss": 2.7056, + "step": 6147 + }, + { + "epoch": 0.26, + "grad_norm": 0.45703125, + "learning_rate": 0.00019188280562924871, + "loss": 2.5872, + "step": 6148 + }, + { + "epoch": 0.26, + "grad_norm": 0.4921875, + "learning_rate": 0.0001918801580831278, + "loss": 2.7789, + "step": 6149 + }, + { + "epoch": 0.26, + "grad_norm": 0.4296875, + "learning_rate": 0.0001918775101235797, + "loss": 2.7188, + "step": 6150 + }, + { + "epoch": 0.26, + "grad_norm": 0.482421875, + "learning_rate": 0.0001918748617506164, + "loss": 2.6517, + "step": 6151 + }, + { + "epoch": 0.26, + "grad_norm": 0.50390625, + "learning_rate": 0.0001918722129642498, + "loss": 2.783, + "step": 6152 + }, + { + "epoch": 0.26, + "grad_norm": 0.427734375, + "learning_rate": 0.00019186956376449183, + "loss": 2.6134, + "step": 6153 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.00019186691415135437, + "loss": 2.6373, + "step": 6154 + }, + { + "epoch": 0.26, + "grad_norm": 0.490234375, + "learning_rate": 0.00019186426412484938, + "loss": 2.6361, + "step": 6155 + }, + { + "epoch": 0.26, + "grad_norm": 0.455078125, + "learning_rate": 0.00019186161368498877, + "loss": 2.6045, + "step": 6156 + }, + { + "epoch": 0.26, + "grad_norm": 0.455078125, + "learning_rate": 0.00019185896283178448, + "loss": 2.7091, + "step": 6157 + }, + { + "epoch": 0.26, + "grad_norm": 0.48828125, + "learning_rate": 0.00019185631156524844, + "loss": 2.6273, + "step": 6158 + }, + { + "epoch": 0.26, + "grad_norm": 0.421875, + "learning_rate": 0.00019185365988539258, + "loss": 2.5892, + "step": 6159 + }, + { + "epoch": 0.26, + "grad_norm": 0.494140625, + "learning_rate": 0.00019185100779222877, + "loss": 2.8043, + "step": 6160 + }, + { + "epoch": 0.26, + "grad_norm": 0.423828125, + "learning_rate": 0.000191848355285769, + "loss": 2.5115, + "step": 6161 + }, + { + "epoch": 0.26, + "grad_norm": 0.5078125, + "learning_rate": 0.0001918457023660252, + "loss": 2.8154, + "step": 6162 + }, + { + "epoch": 0.26, + "grad_norm": 0.423828125, + "learning_rate": 0.00019184304903300934, + "loss": 2.5306, + "step": 6163 + }, + { + "epoch": 0.26, + "grad_norm": 0.486328125, + "learning_rate": 0.0001918403952867333, + "loss": 2.6233, + "step": 6164 + }, + { + "epoch": 0.26, + "grad_norm": 0.4921875, + "learning_rate": 0.00019183774112720907, + "loss": 2.6892, + "step": 6165 + }, + { + "epoch": 0.26, + "grad_norm": 0.474609375, + "learning_rate": 0.00019183508655444854, + "loss": 2.6662, + "step": 6166 + }, + { + "epoch": 0.26, + "grad_norm": 0.447265625, + "learning_rate": 0.00019183243156846368, + "loss": 2.6636, + "step": 6167 + }, + { + "epoch": 0.26, + "grad_norm": 0.447265625, + "learning_rate": 0.00019182977616926644, + "loss": 2.6334, + "step": 6168 + }, + { + "epoch": 0.26, + "grad_norm": 0.43359375, + "learning_rate": 0.00019182712035686878, + "loss": 2.61, + "step": 6169 + }, + { + "epoch": 0.26, + "grad_norm": 0.466796875, + "learning_rate": 0.00019182446413128264, + "loss": 2.862, + "step": 6170 + }, + { + "epoch": 0.26, + "grad_norm": 0.45703125, + "learning_rate": 0.00019182180749251996, + "loss": 2.7237, + "step": 6171 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.0001918191504405927, + "loss": 2.5983, + "step": 6172 + }, + { + "epoch": 0.26, + "grad_norm": 0.42578125, + "learning_rate": 0.0001918164929755128, + "loss": 2.5456, + "step": 6173 + }, + { + "epoch": 0.26, + "grad_norm": 0.41796875, + "learning_rate": 0.00019181383509729228, + "loss": 2.559, + "step": 6174 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.00019181117680594305, + "loss": 2.5719, + "step": 6175 + }, + { + "epoch": 0.26, + "grad_norm": 0.45703125, + "learning_rate": 0.00019180851810147705, + "loss": 2.5754, + "step": 6176 + }, + { + "epoch": 0.26, + "grad_norm": 0.458984375, + "learning_rate": 0.00019180585898390627, + "loss": 2.665, + "step": 6177 + }, + { + "epoch": 0.26, + "grad_norm": 0.423828125, + "learning_rate": 0.00019180319945324273, + "loss": 2.6752, + "step": 6178 + }, + { + "epoch": 0.26, + "grad_norm": 0.416015625, + "learning_rate": 0.00019180053950949827, + "loss": 2.5297, + "step": 6179 + }, + { + "epoch": 0.26, + "grad_norm": 0.42578125, + "learning_rate": 0.00019179787915268496, + "loss": 2.6109, + "step": 6180 + }, + { + "epoch": 0.26, + "grad_norm": 0.484375, + "learning_rate": 0.00019179521838281476, + "loss": 2.6616, + "step": 6181 + }, + { + "epoch": 0.26, + "grad_norm": 0.453125, + "learning_rate": 0.00019179255719989962, + "loss": 2.5288, + "step": 6182 + }, + { + "epoch": 0.26, + "grad_norm": 0.498046875, + "learning_rate": 0.00019178989560395148, + "loss": 2.6648, + "step": 6183 + }, + { + "epoch": 0.26, + "grad_norm": 0.41796875, + "learning_rate": 0.0001917872335949824, + "loss": 2.4967, + "step": 6184 + }, + { + "epoch": 0.26, + "grad_norm": 0.470703125, + "learning_rate": 0.0001917845711730043, + "loss": 2.6169, + "step": 6185 + }, + { + "epoch": 0.26, + "grad_norm": 0.427734375, + "learning_rate": 0.00019178190833802917, + "loss": 2.6392, + "step": 6186 + }, + { + "epoch": 0.26, + "grad_norm": 0.48046875, + "learning_rate": 0.000191779245090069, + "loss": 2.7102, + "step": 6187 + }, + { + "epoch": 0.26, + "grad_norm": 0.482421875, + "learning_rate": 0.00019177658142913577, + "loss": 2.6111, + "step": 6188 + }, + { + "epoch": 0.26, + "grad_norm": 0.396484375, + "learning_rate": 0.00019177391735524145, + "loss": 2.4941, + "step": 6189 + }, + { + "epoch": 0.26, + "grad_norm": 0.4375, + "learning_rate": 0.00019177125286839805, + "loss": 2.6959, + "step": 6190 + }, + { + "epoch": 0.26, + "grad_norm": 0.427734375, + "learning_rate": 0.00019176858796861757, + "loss": 2.5729, + "step": 6191 + }, + { + "epoch": 0.26, + "grad_norm": 0.421875, + "learning_rate": 0.00019176592265591195, + "loss": 2.5791, + "step": 6192 + }, + { + "epoch": 0.26, + "grad_norm": 0.466796875, + "learning_rate": 0.00019176325693029322, + "loss": 2.5972, + "step": 6193 + }, + { + "epoch": 0.26, + "grad_norm": 0.5, + "learning_rate": 0.0001917605907917734, + "loss": 2.6675, + "step": 6194 + }, + { + "epoch": 0.26, + "grad_norm": 0.43359375, + "learning_rate": 0.00019175792424036442, + "loss": 2.6101, + "step": 6195 + }, + { + "epoch": 0.26, + "grad_norm": 0.453125, + "learning_rate": 0.0001917552572760783, + "loss": 2.701, + "step": 6196 + }, + { + "epoch": 0.26, + "grad_norm": 0.447265625, + "learning_rate": 0.0001917525898989271, + "loss": 2.7535, + "step": 6197 + }, + { + "epoch": 0.26, + "grad_norm": 0.447265625, + "learning_rate": 0.00019174992210892278, + "loss": 2.7163, + "step": 6198 + }, + { + "epoch": 0.26, + "grad_norm": 0.400390625, + "learning_rate": 0.0001917472539060773, + "loss": 2.6596, + "step": 6199 + }, + { + "epoch": 0.26, + "grad_norm": 0.439453125, + "learning_rate": 0.00019174458529040274, + "loss": 2.6286, + "step": 6200 + }, + { + "epoch": 0.26, + "grad_norm": 0.478515625, + "learning_rate": 0.00019174191626191104, + "loss": 2.5553, + "step": 6201 + }, + { + "epoch": 0.26, + "grad_norm": 0.4140625, + "learning_rate": 0.00019173924682061428, + "loss": 2.5521, + "step": 6202 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.00019173657696652444, + "loss": 2.825, + "step": 6203 + }, + { + "epoch": 0.26, + "grad_norm": 0.498046875, + "learning_rate": 0.0001917339066996535, + "loss": 2.7917, + "step": 6204 + }, + { + "epoch": 0.26, + "grad_norm": 0.451171875, + "learning_rate": 0.00019173123602001354, + "loss": 2.6163, + "step": 6205 + }, + { + "epoch": 0.26, + "grad_norm": 0.435546875, + "learning_rate": 0.00019172856492761652, + "loss": 2.7057, + "step": 6206 + }, + { + "epoch": 0.26, + "grad_norm": 0.46484375, + "learning_rate": 0.00019172589342247447, + "loss": 2.7156, + "step": 6207 + }, + { + "epoch": 0.26, + "grad_norm": 0.44140625, + "learning_rate": 0.00019172322150459942, + "loss": 2.6712, + "step": 6208 + }, + { + "epoch": 0.26, + "grad_norm": 0.3984375, + "learning_rate": 0.0001917205491740034, + "loss": 2.6042, + "step": 6209 + }, + { + "epoch": 0.26, + "grad_norm": 0.4453125, + "learning_rate": 0.00019171787643069846, + "loss": 2.6051, + "step": 6210 + }, + { + "epoch": 0.26, + "grad_norm": 0.48046875, + "learning_rate": 0.00019171520327469657, + "loss": 2.6968, + "step": 6211 + }, + { + "epoch": 0.26, + "grad_norm": 0.443359375, + "learning_rate": 0.00019171252970600976, + "loss": 2.6314, + "step": 6212 + }, + { + "epoch": 0.26, + "grad_norm": 0.490234375, + "learning_rate": 0.00019170985572465013, + "loss": 2.6938, + "step": 6213 + }, + { + "epoch": 0.26, + "grad_norm": 0.443359375, + "learning_rate": 0.00019170718133062963, + "loss": 2.5927, + "step": 6214 + }, + { + "epoch": 0.26, + "grad_norm": 0.435546875, + "learning_rate": 0.00019170450652396034, + "loss": 2.5096, + "step": 6215 + }, + { + "epoch": 0.26, + "grad_norm": 0.451171875, + "learning_rate": 0.00019170183130465428, + "loss": 2.7789, + "step": 6216 + }, + { + "epoch": 0.26, + "grad_norm": 0.5, + "learning_rate": 0.0001916991556727235, + "loss": 2.7169, + "step": 6217 + }, + { + "epoch": 0.26, + "grad_norm": 0.3984375, + "learning_rate": 0.00019169647962818002, + "loss": 2.4727, + "step": 6218 + }, + { + "epoch": 0.27, + "grad_norm": 0.412109375, + "learning_rate": 0.00019169380317103594, + "loss": 2.5838, + "step": 6219 + }, + { + "epoch": 0.27, + "grad_norm": 0.4921875, + "learning_rate": 0.0001916911263013032, + "loss": 2.6748, + "step": 6220 + }, + { + "epoch": 0.27, + "grad_norm": 0.470703125, + "learning_rate": 0.00019168844901899391, + "loss": 2.648, + "step": 6221 + }, + { + "epoch": 0.27, + "grad_norm": 0.447265625, + "learning_rate": 0.00019168577132412014, + "loss": 2.6232, + "step": 6222 + }, + { + "epoch": 0.27, + "grad_norm": 0.490234375, + "learning_rate": 0.00019168309321669389, + "loss": 2.7378, + "step": 6223 + }, + { + "epoch": 0.27, + "grad_norm": 0.43359375, + "learning_rate": 0.00019168041469672722, + "loss": 2.6752, + "step": 6224 + }, + { + "epoch": 0.27, + "grad_norm": 0.3828125, + "learning_rate": 0.0001916777357642322, + "loss": 2.4998, + "step": 6225 + }, + { + "epoch": 0.27, + "grad_norm": 0.392578125, + "learning_rate": 0.00019167505641922087, + "loss": 2.5659, + "step": 6226 + }, + { + "epoch": 0.27, + "grad_norm": 0.416015625, + "learning_rate": 0.00019167237666170526, + "loss": 2.5958, + "step": 6227 + }, + { + "epoch": 0.27, + "grad_norm": 0.451171875, + "learning_rate": 0.00019166969649169748, + "loss": 2.6349, + "step": 6228 + }, + { + "epoch": 0.27, + "grad_norm": 0.490234375, + "learning_rate": 0.00019166701590920957, + "loss": 2.7218, + "step": 6229 + }, + { + "epoch": 0.27, + "grad_norm": 0.44921875, + "learning_rate": 0.00019166433491425358, + "loss": 2.7459, + "step": 6230 + }, + { + "epoch": 0.27, + "grad_norm": 0.3984375, + "learning_rate": 0.0001916616535068416, + "loss": 2.6612, + "step": 6231 + }, + { + "epoch": 0.27, + "grad_norm": 0.404296875, + "learning_rate": 0.00019165897168698566, + "loss": 2.5298, + "step": 6232 + }, + { + "epoch": 0.27, + "grad_norm": 0.4453125, + "learning_rate": 0.00019165628945469788, + "loss": 2.6948, + "step": 6233 + }, + { + "epoch": 0.27, + "grad_norm": 0.384765625, + "learning_rate": 0.00019165360680999024, + "loss": 2.4392, + "step": 6234 + }, + { + "epoch": 0.27, + "grad_norm": 0.46875, + "learning_rate": 0.0001916509237528749, + "loss": 2.6129, + "step": 6235 + }, + { + "epoch": 0.27, + "grad_norm": 0.45703125, + "learning_rate": 0.0001916482402833639, + "loss": 2.7346, + "step": 6236 + }, + { + "epoch": 0.27, + "grad_norm": 0.41796875, + "learning_rate": 0.00019164555640146928, + "loss": 2.46, + "step": 6237 + }, + { + "epoch": 0.27, + "grad_norm": 0.451171875, + "learning_rate": 0.00019164287210720316, + "loss": 2.7316, + "step": 6238 + }, + { + "epoch": 0.27, + "grad_norm": 0.41796875, + "learning_rate": 0.00019164018740057764, + "loss": 2.6174, + "step": 6239 + }, + { + "epoch": 0.27, + "grad_norm": 0.458984375, + "learning_rate": 0.00019163750228160472, + "loss": 2.5948, + "step": 6240 + }, + { + "epoch": 0.27, + "grad_norm": 0.427734375, + "learning_rate": 0.00019163481675029656, + "loss": 2.5207, + "step": 6241 + }, + { + "epoch": 0.27, + "grad_norm": 0.439453125, + "learning_rate": 0.00019163213080666523, + "loss": 2.6601, + "step": 6242 + }, + { + "epoch": 0.27, + "grad_norm": 0.42578125, + "learning_rate": 0.00019162944445072277, + "loss": 2.7458, + "step": 6243 + }, + { + "epoch": 0.27, + "grad_norm": 0.40625, + "learning_rate": 0.0001916267576824813, + "loss": 2.5117, + "step": 6244 + }, + { + "epoch": 0.27, + "grad_norm": 0.46484375, + "learning_rate": 0.0001916240705019529, + "loss": 2.6316, + "step": 6245 + }, + { + "epoch": 0.27, + "grad_norm": 0.466796875, + "learning_rate": 0.00019162138290914967, + "loss": 2.6384, + "step": 6246 + }, + { + "epoch": 0.27, + "grad_norm": 0.48046875, + "learning_rate": 0.00019161869490408372, + "loss": 2.5109, + "step": 6247 + }, + { + "epoch": 0.27, + "grad_norm": 0.490234375, + "learning_rate": 0.0001916160064867671, + "loss": 2.7618, + "step": 6248 + }, + { + "epoch": 0.27, + "grad_norm": 0.451171875, + "learning_rate": 0.00019161331765721196, + "loss": 2.7568, + "step": 6249 + }, + { + "epoch": 0.27, + "grad_norm": 0.3984375, + "learning_rate": 0.00019161062841543033, + "loss": 2.612, + "step": 6250 + }, + { + "epoch": 0.27, + "grad_norm": 0.484375, + "learning_rate": 0.00019160793876143442, + "loss": 2.7292, + "step": 6251 + }, + { + "epoch": 0.27, + "grad_norm": 0.45703125, + "learning_rate": 0.00019160524869523622, + "loss": 2.6401, + "step": 6252 + }, + { + "epoch": 0.27, + "grad_norm": 0.4453125, + "learning_rate": 0.00019160255821684788, + "loss": 2.6955, + "step": 6253 + }, + { + "epoch": 0.27, + "grad_norm": 0.5078125, + "learning_rate": 0.0001915998673262815, + "loss": 2.5598, + "step": 6254 + }, + { + "epoch": 0.27, + "grad_norm": 0.44140625, + "learning_rate": 0.00019159717602354917, + "loss": 2.6078, + "step": 6255 + }, + { + "epoch": 0.27, + "grad_norm": 0.396484375, + "learning_rate": 0.0001915944843086631, + "loss": 2.6157, + "step": 6256 + }, + { + "epoch": 0.27, + "grad_norm": 0.48828125, + "learning_rate": 0.00019159179218163525, + "loss": 2.7717, + "step": 6257 + }, + { + "epoch": 0.27, + "grad_norm": 0.431640625, + "learning_rate": 0.00019158909964247785, + "loss": 2.6205, + "step": 6258 + }, + { + "epoch": 0.27, + "grad_norm": 0.4296875, + "learning_rate": 0.00019158640669120292, + "loss": 2.7792, + "step": 6259 + }, + { + "epoch": 0.27, + "grad_norm": 0.447265625, + "learning_rate": 0.0001915837133278227, + "loss": 2.621, + "step": 6260 + }, + { + "epoch": 0.27, + "grad_norm": 0.455078125, + "learning_rate": 0.0001915810195523492, + "loss": 2.6367, + "step": 6261 + }, + { + "epoch": 0.27, + "grad_norm": 0.4609375, + "learning_rate": 0.0001915783253647946, + "loss": 2.6592, + "step": 6262 + }, + { + "epoch": 0.27, + "grad_norm": 0.421875, + "learning_rate": 0.000191575630765171, + "loss": 2.6063, + "step": 6263 + }, + { + "epoch": 0.27, + "grad_norm": 0.431640625, + "learning_rate": 0.00019157293575349052, + "loss": 2.5573, + "step": 6264 + }, + { + "epoch": 0.27, + "grad_norm": 0.4453125, + "learning_rate": 0.00019157024032976532, + "loss": 2.6478, + "step": 6265 + }, + { + "epoch": 0.27, + "grad_norm": 0.40234375, + "learning_rate": 0.00019156754449400746, + "loss": 2.6452, + "step": 6266 + }, + { + "epoch": 0.27, + "grad_norm": 0.447265625, + "learning_rate": 0.00019156484824622914, + "loss": 2.6465, + "step": 6267 + }, + { + "epoch": 0.27, + "grad_norm": 0.44140625, + "learning_rate": 0.00019156215158644247, + "loss": 2.7796, + "step": 6268 + }, + { + "epoch": 0.27, + "grad_norm": 0.451171875, + "learning_rate": 0.00019155945451465955, + "loss": 2.6309, + "step": 6269 + }, + { + "epoch": 0.27, + "grad_norm": 0.44921875, + "learning_rate": 0.00019155675703089257, + "loss": 2.7025, + "step": 6270 + }, + { + "epoch": 0.27, + "grad_norm": 0.412109375, + "learning_rate": 0.00019155405913515365, + "loss": 2.576, + "step": 6271 + }, + { + "epoch": 0.27, + "grad_norm": 0.4453125, + "learning_rate": 0.00019155136082745492, + "loss": 2.7027, + "step": 6272 + }, + { + "epoch": 0.27, + "grad_norm": 0.421875, + "learning_rate": 0.0001915486621078085, + "loss": 2.594, + "step": 6273 + }, + { + "epoch": 0.27, + "grad_norm": 0.453125, + "learning_rate": 0.0001915459629762266, + "loss": 2.8604, + "step": 6274 + }, + { + "epoch": 0.27, + "grad_norm": 0.484375, + "learning_rate": 0.00019154326343272132, + "loss": 2.7795, + "step": 6275 + }, + { + "epoch": 0.27, + "grad_norm": 0.4921875, + "learning_rate": 0.00019154056347730477, + "loss": 2.8041, + "step": 6276 + }, + { + "epoch": 0.27, + "grad_norm": 0.470703125, + "learning_rate": 0.00019153786310998917, + "loss": 2.6046, + "step": 6277 + }, + { + "epoch": 0.27, + "grad_norm": 0.466796875, + "learning_rate": 0.00019153516233078662, + "loss": 2.6983, + "step": 6278 + }, + { + "epoch": 0.27, + "grad_norm": 0.45703125, + "learning_rate": 0.00019153246113970927, + "loss": 2.6714, + "step": 6279 + }, + { + "epoch": 0.27, + "grad_norm": 0.42578125, + "learning_rate": 0.00019152975953676935, + "loss": 2.546, + "step": 6280 + }, + { + "epoch": 0.27, + "grad_norm": 0.416015625, + "learning_rate": 0.00019152705752197892, + "loss": 2.6942, + "step": 6281 + }, + { + "epoch": 0.27, + "grad_norm": 0.427734375, + "learning_rate": 0.00019152435509535022, + "loss": 2.5997, + "step": 6282 + }, + { + "epoch": 0.27, + "grad_norm": 0.43359375, + "learning_rate": 0.00019152165225689536, + "loss": 2.6607, + "step": 6283 + }, + { + "epoch": 0.27, + "grad_norm": 0.435546875, + "learning_rate": 0.0001915189490066265, + "loss": 2.606, + "step": 6284 + }, + { + "epoch": 0.27, + "grad_norm": 0.431640625, + "learning_rate": 0.00019151624534455582, + "loss": 2.6637, + "step": 6285 + }, + { + "epoch": 0.27, + "grad_norm": 0.46484375, + "learning_rate": 0.00019151354127069548, + "loss": 2.7577, + "step": 6286 + }, + { + "epoch": 0.27, + "grad_norm": 0.4375, + "learning_rate": 0.00019151083678505763, + "loss": 2.5731, + "step": 6287 + }, + { + "epoch": 0.27, + "grad_norm": 0.427734375, + "learning_rate": 0.00019150813188765445, + "loss": 2.669, + "step": 6288 + }, + { + "epoch": 0.27, + "grad_norm": 0.423828125, + "learning_rate": 0.00019150542657849818, + "loss": 2.6649, + "step": 6289 + }, + { + "epoch": 0.27, + "grad_norm": 0.42578125, + "learning_rate": 0.00019150272085760088, + "loss": 2.5193, + "step": 6290 + }, + { + "epoch": 0.27, + "grad_norm": 0.416015625, + "learning_rate": 0.00019150001472497479, + "loss": 2.7169, + "step": 6291 + }, + { + "epoch": 0.27, + "grad_norm": 0.4921875, + "learning_rate": 0.00019149730818063208, + "loss": 2.6825, + "step": 6292 + }, + { + "epoch": 0.27, + "grad_norm": 0.458984375, + "learning_rate": 0.0001914946012245849, + "loss": 2.6918, + "step": 6293 + }, + { + "epoch": 0.27, + "grad_norm": 0.498046875, + "learning_rate": 0.00019149189385684546, + "loss": 2.7596, + "step": 6294 + }, + { + "epoch": 0.27, + "grad_norm": 0.455078125, + "learning_rate": 0.00019148918607742596, + "loss": 2.7501, + "step": 6295 + }, + { + "epoch": 0.27, + "grad_norm": 0.44921875, + "learning_rate": 0.00019148647788633856, + "loss": 2.7112, + "step": 6296 + }, + { + "epoch": 0.27, + "grad_norm": 0.443359375, + "learning_rate": 0.0001914837692835954, + "loss": 2.6213, + "step": 6297 + }, + { + "epoch": 0.27, + "grad_norm": 0.390625, + "learning_rate": 0.00019148106026920875, + "loss": 2.5216, + "step": 6298 + }, + { + "epoch": 0.27, + "grad_norm": 0.421875, + "learning_rate": 0.00019147835084319073, + "loss": 2.6337, + "step": 6299 + }, + { + "epoch": 0.27, + "grad_norm": 0.4765625, + "learning_rate": 0.0001914756410055536, + "loss": 2.5972, + "step": 6300 + }, + { + "epoch": 0.27, + "grad_norm": 0.419921875, + "learning_rate": 0.0001914729307563095, + "loss": 2.6207, + "step": 6301 + }, + { + "epoch": 0.27, + "grad_norm": 0.41015625, + "learning_rate": 0.00019147022009547064, + "loss": 2.6456, + "step": 6302 + }, + { + "epoch": 0.27, + "grad_norm": 0.46484375, + "learning_rate": 0.0001914675090230492, + "loss": 2.6922, + "step": 6303 + }, + { + "epoch": 0.27, + "grad_norm": 0.43359375, + "learning_rate": 0.00019146479753905741, + "loss": 2.6809, + "step": 6304 + }, + { + "epoch": 0.27, + "grad_norm": 0.50390625, + "learning_rate": 0.00019146208564350747, + "loss": 2.7136, + "step": 6305 + }, + { + "epoch": 0.27, + "grad_norm": 0.439453125, + "learning_rate": 0.00019145937333641156, + "loss": 2.6032, + "step": 6306 + }, + { + "epoch": 0.27, + "grad_norm": 0.416015625, + "learning_rate": 0.0001914566606177819, + "loss": 2.6979, + "step": 6307 + }, + { + "epoch": 0.27, + "grad_norm": 0.490234375, + "learning_rate": 0.00019145394748763068, + "loss": 2.5939, + "step": 6308 + }, + { + "epoch": 0.27, + "grad_norm": 0.4296875, + "learning_rate": 0.0001914512339459701, + "loss": 2.6428, + "step": 6309 + }, + { + "epoch": 0.27, + "grad_norm": 0.412109375, + "learning_rate": 0.00019144851999281244, + "loss": 2.5766, + "step": 6310 + }, + { + "epoch": 0.27, + "grad_norm": 0.453125, + "learning_rate": 0.00019144580562816983, + "loss": 2.6134, + "step": 6311 + }, + { + "epoch": 0.27, + "grad_norm": 0.49609375, + "learning_rate": 0.00019144309085205452, + "loss": 2.6376, + "step": 6312 + }, + { + "epoch": 0.27, + "grad_norm": 0.498046875, + "learning_rate": 0.00019144037566447871, + "loss": 2.8198, + "step": 6313 + }, + { + "epoch": 0.27, + "grad_norm": 0.466796875, + "learning_rate": 0.00019143766006545464, + "loss": 2.4939, + "step": 6314 + }, + { + "epoch": 0.27, + "grad_norm": 0.447265625, + "learning_rate": 0.0001914349440549945, + "loss": 2.6088, + "step": 6315 + }, + { + "epoch": 0.27, + "grad_norm": 0.515625, + "learning_rate": 0.00019143222763311056, + "loss": 2.5994, + "step": 6316 + }, + { + "epoch": 0.27, + "grad_norm": 0.423828125, + "learning_rate": 0.00019142951079981498, + "loss": 2.7555, + "step": 6317 + }, + { + "epoch": 0.27, + "grad_norm": 0.439453125, + "learning_rate": 0.00019142679355512002, + "loss": 2.5575, + "step": 6318 + }, + { + "epoch": 0.27, + "grad_norm": 0.4921875, + "learning_rate": 0.0001914240758990379, + "loss": 2.7973, + "step": 6319 + }, + { + "epoch": 0.27, + "grad_norm": 0.44140625, + "learning_rate": 0.00019142135783158088, + "loss": 2.7693, + "step": 6320 + }, + { + "epoch": 0.27, + "grad_norm": 0.447265625, + "learning_rate": 0.00019141863935276114, + "loss": 2.7842, + "step": 6321 + }, + { + "epoch": 0.27, + "grad_norm": 0.5078125, + "learning_rate": 0.00019141592046259092, + "loss": 2.7037, + "step": 6322 + }, + { + "epoch": 0.27, + "grad_norm": 0.478515625, + "learning_rate": 0.00019141320116108248, + "loss": 2.6533, + "step": 6323 + }, + { + "epoch": 0.27, + "grad_norm": 0.453125, + "learning_rate": 0.00019141048144824803, + "loss": 2.7045, + "step": 6324 + }, + { + "epoch": 0.27, + "grad_norm": 0.48828125, + "learning_rate": 0.00019140776132409984, + "loss": 2.6551, + "step": 6325 + }, + { + "epoch": 0.27, + "grad_norm": 0.421875, + "learning_rate": 0.00019140504078865011, + "loss": 2.6907, + "step": 6326 + }, + { + "epoch": 0.27, + "grad_norm": 0.38671875, + "learning_rate": 0.00019140231984191112, + "loss": 2.656, + "step": 6327 + }, + { + "epoch": 0.27, + "grad_norm": 0.453125, + "learning_rate": 0.0001913995984838951, + "loss": 2.7039, + "step": 6328 + }, + { + "epoch": 0.27, + "grad_norm": 0.486328125, + "learning_rate": 0.00019139687671461425, + "loss": 2.6198, + "step": 6329 + }, + { + "epoch": 0.27, + "grad_norm": 0.435546875, + "learning_rate": 0.00019139415453408088, + "loss": 2.6438, + "step": 6330 + }, + { + "epoch": 0.27, + "grad_norm": 0.3828125, + "learning_rate": 0.0001913914319423072, + "loss": 2.6251, + "step": 6331 + }, + { + "epoch": 0.27, + "grad_norm": 0.447265625, + "learning_rate": 0.00019138870893930553, + "loss": 2.6314, + "step": 6332 + }, + { + "epoch": 0.27, + "grad_norm": 0.4296875, + "learning_rate": 0.000191385985525088, + "loss": 2.6614, + "step": 6333 + }, + { + "epoch": 0.27, + "grad_norm": 0.39453125, + "learning_rate": 0.00019138326169966695, + "loss": 2.5747, + "step": 6334 + }, + { + "epoch": 0.27, + "grad_norm": 0.423828125, + "learning_rate": 0.00019138053746305467, + "loss": 2.4941, + "step": 6335 + }, + { + "epoch": 0.27, + "grad_norm": 0.451171875, + "learning_rate": 0.00019137781281526332, + "loss": 2.5912, + "step": 6336 + }, + { + "epoch": 0.27, + "grad_norm": 0.431640625, + "learning_rate": 0.00019137508775630518, + "loss": 2.549, + "step": 6337 + }, + { + "epoch": 0.27, + "grad_norm": 0.41796875, + "learning_rate": 0.0001913723622861926, + "loss": 2.6621, + "step": 6338 + }, + { + "epoch": 0.27, + "grad_norm": 0.400390625, + "learning_rate": 0.00019136963640493777, + "loss": 2.6316, + "step": 6339 + }, + { + "epoch": 0.27, + "grad_norm": 0.42578125, + "learning_rate": 0.00019136691011255294, + "loss": 2.5701, + "step": 6340 + }, + { + "epoch": 0.27, + "grad_norm": 0.484375, + "learning_rate": 0.00019136418340905042, + "loss": 2.565, + "step": 6341 + }, + { + "epoch": 0.27, + "grad_norm": 0.47265625, + "learning_rate": 0.00019136145629444247, + "loss": 2.5234, + "step": 6342 + }, + { + "epoch": 0.27, + "grad_norm": 0.384765625, + "learning_rate": 0.00019135872876874135, + "loss": 2.6753, + "step": 6343 + }, + { + "epoch": 0.27, + "grad_norm": 0.40234375, + "learning_rate": 0.00019135600083195934, + "loss": 2.5717, + "step": 6344 + }, + { + "epoch": 0.27, + "grad_norm": 0.443359375, + "learning_rate": 0.00019135327248410872, + "loss": 2.6095, + "step": 6345 + }, + { + "epoch": 0.27, + "grad_norm": 0.44921875, + "learning_rate": 0.00019135054372520176, + "loss": 2.7249, + "step": 6346 + }, + { + "epoch": 0.27, + "grad_norm": 0.41015625, + "learning_rate": 0.0001913478145552507, + "loss": 2.6598, + "step": 6347 + }, + { + "epoch": 0.27, + "grad_norm": 0.490234375, + "learning_rate": 0.0001913450849742679, + "loss": 2.6119, + "step": 6348 + }, + { + "epoch": 0.27, + "grad_norm": 0.44140625, + "learning_rate": 0.00019134235498226559, + "loss": 2.5887, + "step": 6349 + }, + { + "epoch": 0.27, + "grad_norm": 0.40234375, + "learning_rate": 0.00019133962457925608, + "loss": 2.6646, + "step": 6350 + }, + { + "epoch": 0.27, + "grad_norm": 0.40234375, + "learning_rate": 0.0001913368937652516, + "loss": 2.5863, + "step": 6351 + }, + { + "epoch": 0.27, + "grad_norm": 0.392578125, + "learning_rate": 0.00019133416254026454, + "loss": 2.5568, + "step": 6352 + }, + { + "epoch": 0.27, + "grad_norm": 0.451171875, + "learning_rate": 0.00019133143090430704, + "loss": 2.6166, + "step": 6353 + }, + { + "epoch": 0.27, + "grad_norm": 0.50390625, + "learning_rate": 0.00019132869885739155, + "loss": 2.7552, + "step": 6354 + }, + { + "epoch": 0.27, + "grad_norm": 0.451171875, + "learning_rate": 0.00019132596639953025, + "loss": 2.6897, + "step": 6355 + }, + { + "epoch": 0.27, + "grad_norm": 0.474609375, + "learning_rate": 0.00019132323353073552, + "loss": 2.5122, + "step": 6356 + }, + { + "epoch": 0.27, + "grad_norm": 0.412109375, + "learning_rate": 0.00019132050025101957, + "loss": 2.6458, + "step": 6357 + }, + { + "epoch": 0.27, + "grad_norm": 0.396484375, + "learning_rate": 0.00019131776656039476, + "loss": 2.5704, + "step": 6358 + }, + { + "epoch": 0.27, + "grad_norm": 0.390625, + "learning_rate": 0.00019131503245887338, + "loss": 2.4649, + "step": 6359 + }, + { + "epoch": 0.27, + "grad_norm": 0.439453125, + "learning_rate": 0.00019131229794646775, + "loss": 2.6191, + "step": 6360 + }, + { + "epoch": 0.27, + "grad_norm": 0.408203125, + "learning_rate": 0.00019130956302319013, + "loss": 2.6357, + "step": 6361 + }, + { + "epoch": 0.27, + "grad_norm": 0.42578125, + "learning_rate": 0.00019130682768905282, + "loss": 2.733, + "step": 6362 + }, + { + "epoch": 0.27, + "grad_norm": 0.466796875, + "learning_rate": 0.00019130409194406816, + "loss": 2.6133, + "step": 6363 + }, + { + "epoch": 0.27, + "grad_norm": 0.44140625, + "learning_rate": 0.00019130135578824845, + "loss": 2.6008, + "step": 6364 + }, + { + "epoch": 0.27, + "grad_norm": 0.41015625, + "learning_rate": 0.00019129861922160605, + "loss": 2.601, + "step": 6365 + }, + { + "epoch": 0.27, + "grad_norm": 0.39453125, + "learning_rate": 0.0001912958822441532, + "loss": 2.4944, + "step": 6366 + }, + { + "epoch": 0.27, + "grad_norm": 0.41796875, + "learning_rate": 0.00019129314485590224, + "loss": 2.6058, + "step": 6367 + }, + { + "epoch": 0.27, + "grad_norm": 0.419921875, + "learning_rate": 0.00019129040705686553, + "loss": 2.7774, + "step": 6368 + }, + { + "epoch": 0.27, + "grad_norm": 0.451171875, + "learning_rate": 0.0001912876688470553, + "loss": 2.5463, + "step": 6369 + }, + { + "epoch": 0.27, + "grad_norm": 0.431640625, + "learning_rate": 0.00019128493022648396, + "loss": 2.5303, + "step": 6370 + }, + { + "epoch": 0.27, + "grad_norm": 0.453125, + "learning_rate": 0.00019128219119516376, + "loss": 2.5745, + "step": 6371 + }, + { + "epoch": 0.27, + "grad_norm": 0.4921875, + "learning_rate": 0.00019127945175310707, + "loss": 2.6722, + "step": 6372 + }, + { + "epoch": 0.27, + "grad_norm": 0.4296875, + "learning_rate": 0.00019127671190032622, + "loss": 2.665, + "step": 6373 + }, + { + "epoch": 0.27, + "grad_norm": 0.365234375, + "learning_rate": 0.00019127397163683351, + "loss": 2.5176, + "step": 6374 + }, + { + "epoch": 0.27, + "grad_norm": 0.44140625, + "learning_rate": 0.0001912712309626413, + "loss": 2.5947, + "step": 6375 + }, + { + "epoch": 0.27, + "grad_norm": 0.45703125, + "learning_rate": 0.00019126848987776186, + "loss": 2.5771, + "step": 6376 + }, + { + "epoch": 0.27, + "grad_norm": 0.416015625, + "learning_rate": 0.00019126574838220764, + "loss": 2.6529, + "step": 6377 + }, + { + "epoch": 0.27, + "grad_norm": 0.39453125, + "learning_rate": 0.00019126300647599086, + "loss": 2.6708, + "step": 6378 + }, + { + "epoch": 0.27, + "grad_norm": 0.4140625, + "learning_rate": 0.00019126026415912392, + "loss": 2.6327, + "step": 6379 + }, + { + "epoch": 0.27, + "grad_norm": 0.443359375, + "learning_rate": 0.00019125752143161914, + "loss": 2.5715, + "step": 6380 + }, + { + "epoch": 0.27, + "grad_norm": 0.443359375, + "learning_rate": 0.00019125477829348887, + "loss": 2.6959, + "step": 6381 + }, + { + "epoch": 0.27, + "grad_norm": 0.4296875, + "learning_rate": 0.00019125203474474543, + "loss": 2.662, + "step": 6382 + }, + { + "epoch": 0.27, + "grad_norm": 0.3984375, + "learning_rate": 0.00019124929078540118, + "loss": 2.6086, + "step": 6383 + }, + { + "epoch": 0.27, + "grad_norm": 0.47265625, + "learning_rate": 0.00019124654641546851, + "loss": 2.6264, + "step": 6384 + }, + { + "epoch": 0.27, + "grad_norm": 0.400390625, + "learning_rate": 0.0001912438016349597, + "loss": 2.395, + "step": 6385 + }, + { + "epoch": 0.27, + "grad_norm": 0.41015625, + "learning_rate": 0.00019124105644388712, + "loss": 2.4374, + "step": 6386 + }, + { + "epoch": 0.27, + "grad_norm": 0.48046875, + "learning_rate": 0.00019123831084226316, + "loss": 2.6068, + "step": 6387 + }, + { + "epoch": 0.27, + "grad_norm": 0.423828125, + "learning_rate": 0.0001912355648301001, + "loss": 2.7325, + "step": 6388 + }, + { + "epoch": 0.27, + "grad_norm": 0.416015625, + "learning_rate": 0.0001912328184074104, + "loss": 2.6329, + "step": 6389 + }, + { + "epoch": 0.27, + "grad_norm": 0.4375, + "learning_rate": 0.0001912300715742063, + "loss": 2.6219, + "step": 6390 + }, + { + "epoch": 0.27, + "grad_norm": 0.466796875, + "learning_rate": 0.00019122732433050023, + "loss": 2.5715, + "step": 6391 + }, + { + "epoch": 0.27, + "grad_norm": 0.42578125, + "learning_rate": 0.00019122457667630457, + "loss": 2.7551, + "step": 6392 + }, + { + "epoch": 0.27, + "grad_norm": 0.48046875, + "learning_rate": 0.00019122182861163164, + "loss": 2.7108, + "step": 6393 + }, + { + "epoch": 0.27, + "grad_norm": 0.4296875, + "learning_rate": 0.00019121908013649384, + "loss": 2.6845, + "step": 6394 + }, + { + "epoch": 0.27, + "grad_norm": 0.49609375, + "learning_rate": 0.0001912163312509035, + "loss": 2.6635, + "step": 6395 + }, + { + "epoch": 0.27, + "grad_norm": 0.5078125, + "learning_rate": 0.000191213581954873, + "loss": 2.6015, + "step": 6396 + }, + { + "epoch": 0.27, + "grad_norm": 0.49609375, + "learning_rate": 0.0001912108322484147, + "loss": 2.6987, + "step": 6397 + }, + { + "epoch": 0.27, + "grad_norm": 0.453125, + "learning_rate": 0.000191208082131541, + "loss": 2.5384, + "step": 6398 + }, + { + "epoch": 0.27, + "grad_norm": 0.4140625, + "learning_rate": 0.00019120533160426427, + "loss": 2.5575, + "step": 6399 + }, + { + "epoch": 0.27, + "grad_norm": 0.416015625, + "learning_rate": 0.00019120258066659685, + "loss": 2.6186, + "step": 6400 + }, + { + "epoch": 0.27, + "grad_norm": 0.45703125, + "learning_rate": 0.0001911998293185512, + "loss": 2.7293, + "step": 6401 + }, + { + "epoch": 0.27, + "grad_norm": 0.462890625, + "learning_rate": 0.00019119707756013962, + "loss": 2.4981, + "step": 6402 + }, + { + "epoch": 0.27, + "grad_norm": 0.392578125, + "learning_rate": 0.00019119432539137452, + "loss": 2.6901, + "step": 6403 + }, + { + "epoch": 0.27, + "grad_norm": 0.51171875, + "learning_rate": 0.00019119157281226827, + "loss": 2.7409, + "step": 6404 + }, + { + "epoch": 0.27, + "grad_norm": 0.439453125, + "learning_rate": 0.00019118881982283328, + "loss": 2.7215, + "step": 6405 + }, + { + "epoch": 0.27, + "grad_norm": 0.48046875, + "learning_rate": 0.00019118606642308194, + "loss": 2.6719, + "step": 6406 + }, + { + "epoch": 0.27, + "grad_norm": 0.470703125, + "learning_rate": 0.00019118331261302663, + "loss": 2.6712, + "step": 6407 + }, + { + "epoch": 0.27, + "grad_norm": 0.54296875, + "learning_rate": 0.0001911805583926797, + "loss": 2.8085, + "step": 6408 + }, + { + "epoch": 0.27, + "grad_norm": 0.49609375, + "learning_rate": 0.0001911778037620536, + "loss": 2.6329, + "step": 6409 + }, + { + "epoch": 0.27, + "grad_norm": 0.466796875, + "learning_rate": 0.00019117504872116073, + "loss": 2.6622, + "step": 6410 + }, + { + "epoch": 0.27, + "grad_norm": 0.43359375, + "learning_rate": 0.00019117229327001342, + "loss": 2.7563, + "step": 6411 + }, + { + "epoch": 0.27, + "grad_norm": 0.462890625, + "learning_rate": 0.00019116953740862415, + "loss": 2.6118, + "step": 6412 + }, + { + "epoch": 0.27, + "grad_norm": 0.4609375, + "learning_rate": 0.00019116678113700525, + "loss": 2.7256, + "step": 6413 + }, + { + "epoch": 0.27, + "grad_norm": 0.408203125, + "learning_rate": 0.00019116402445516916, + "loss": 2.5462, + "step": 6414 + }, + { + "epoch": 0.27, + "grad_norm": 0.4921875, + "learning_rate": 0.00019116126736312827, + "loss": 2.7483, + "step": 6415 + }, + { + "epoch": 0.27, + "grad_norm": 0.54296875, + "learning_rate": 0.00019115850986089501, + "loss": 2.6827, + "step": 6416 + }, + { + "epoch": 0.27, + "grad_norm": 0.44921875, + "learning_rate": 0.00019115575194848176, + "loss": 2.6682, + "step": 6417 + }, + { + "epoch": 0.27, + "grad_norm": 0.416015625, + "learning_rate": 0.00019115299362590092, + "loss": 2.6674, + "step": 6418 + }, + { + "epoch": 0.27, + "grad_norm": 0.466796875, + "learning_rate": 0.00019115023489316496, + "loss": 2.6786, + "step": 6419 + }, + { + "epoch": 0.27, + "grad_norm": 0.435546875, + "learning_rate": 0.00019114747575028624, + "loss": 2.789, + "step": 6420 + }, + { + "epoch": 0.27, + "grad_norm": 0.4765625, + "learning_rate": 0.00019114471619727718, + "loss": 2.5806, + "step": 6421 + }, + { + "epoch": 0.27, + "grad_norm": 0.447265625, + "learning_rate": 0.00019114195623415024, + "loss": 2.6924, + "step": 6422 + }, + { + "epoch": 0.27, + "grad_norm": 0.39453125, + "learning_rate": 0.00019113919586091776, + "loss": 2.6282, + "step": 6423 + }, + { + "epoch": 0.27, + "grad_norm": 0.43359375, + "learning_rate": 0.00019113643507759223, + "loss": 2.5305, + "step": 6424 + }, + { + "epoch": 0.27, + "grad_norm": 0.421875, + "learning_rate": 0.00019113367388418604, + "loss": 2.455, + "step": 6425 + }, + { + "epoch": 0.27, + "grad_norm": 0.443359375, + "learning_rate": 0.00019113091228071163, + "loss": 2.6649, + "step": 6426 + }, + { + "epoch": 0.27, + "grad_norm": 0.462890625, + "learning_rate": 0.00019112815026718142, + "loss": 2.453, + "step": 6427 + }, + { + "epoch": 0.27, + "grad_norm": 0.39453125, + "learning_rate": 0.00019112538784360784, + "loss": 2.6853, + "step": 6428 + }, + { + "epoch": 0.27, + "grad_norm": 0.42578125, + "learning_rate": 0.00019112262501000328, + "loss": 2.5816, + "step": 6429 + }, + { + "epoch": 0.27, + "grad_norm": 0.423828125, + "learning_rate": 0.00019111986176638027, + "loss": 2.6678, + "step": 6430 + }, + { + "epoch": 0.27, + "grad_norm": 0.4921875, + "learning_rate": 0.00019111709811275114, + "loss": 2.6176, + "step": 6431 + }, + { + "epoch": 0.27, + "grad_norm": 0.47265625, + "learning_rate": 0.00019111433404912838, + "loss": 2.6684, + "step": 6432 + }, + { + "epoch": 0.27, + "grad_norm": 0.427734375, + "learning_rate": 0.00019111156957552443, + "loss": 2.6509, + "step": 6433 + }, + { + "epoch": 0.27, + "grad_norm": 0.4296875, + "learning_rate": 0.00019110880469195166, + "loss": 2.5349, + "step": 6434 + }, + { + "epoch": 0.27, + "grad_norm": 0.40234375, + "learning_rate": 0.00019110603939842262, + "loss": 2.5928, + "step": 6435 + }, + { + "epoch": 0.27, + "grad_norm": 0.44921875, + "learning_rate": 0.00019110327369494968, + "loss": 2.7772, + "step": 6436 + }, + { + "epoch": 0.27, + "grad_norm": 0.45703125, + "learning_rate": 0.00019110050758154533, + "loss": 2.6284, + "step": 6437 + }, + { + "epoch": 0.27, + "grad_norm": 0.423828125, + "learning_rate": 0.00019109774105822197, + "loss": 2.8092, + "step": 6438 + }, + { + "epoch": 0.27, + "grad_norm": 0.46484375, + "learning_rate": 0.00019109497412499204, + "loss": 2.573, + "step": 6439 + }, + { + "epoch": 0.27, + "grad_norm": 0.49609375, + "learning_rate": 0.00019109220678186804, + "loss": 2.6907, + "step": 6440 + }, + { + "epoch": 0.27, + "grad_norm": 0.470703125, + "learning_rate": 0.00019108943902886242, + "loss": 2.6371, + "step": 6441 + }, + { + "epoch": 0.27, + "grad_norm": 0.40625, + "learning_rate": 0.00019108667086598758, + "loss": 2.7024, + "step": 6442 + }, + { + "epoch": 0.27, + "grad_norm": 0.435546875, + "learning_rate": 0.00019108390229325602, + "loss": 2.5566, + "step": 6443 + }, + { + "epoch": 0.27, + "grad_norm": 0.4609375, + "learning_rate": 0.0001910811333106802, + "loss": 2.584, + "step": 6444 + }, + { + "epoch": 0.27, + "grad_norm": 0.46484375, + "learning_rate": 0.00019107836391827257, + "loss": 2.637, + "step": 6445 + }, + { + "epoch": 0.27, + "grad_norm": 0.423828125, + "learning_rate": 0.00019107559411604555, + "loss": 2.6984, + "step": 6446 + }, + { + "epoch": 0.27, + "grad_norm": 0.43359375, + "learning_rate": 0.0001910728239040117, + "loss": 2.7146, + "step": 6447 + }, + { + "epoch": 0.27, + "grad_norm": 0.392578125, + "learning_rate": 0.00019107005328218338, + "loss": 2.6734, + "step": 6448 + }, + { + "epoch": 0.27, + "grad_norm": 0.484375, + "learning_rate": 0.0001910672822505731, + "loss": 2.5767, + "step": 6449 + }, + { + "epoch": 0.27, + "grad_norm": 0.50390625, + "learning_rate": 0.00019106451080919335, + "loss": 2.7756, + "step": 6450 + }, + { + "epoch": 0.27, + "grad_norm": 0.408203125, + "learning_rate": 0.0001910617389580566, + "loss": 2.5896, + "step": 6451 + }, + { + "epoch": 0.27, + "grad_norm": 0.466796875, + "learning_rate": 0.00019105896669717524, + "loss": 2.7068, + "step": 6452 + }, + { + "epoch": 0.27, + "grad_norm": 0.431640625, + "learning_rate": 0.00019105619402656187, + "loss": 2.6445, + "step": 6453 + }, + { + "epoch": 0.28, + "grad_norm": 0.48828125, + "learning_rate": 0.00019105342094622886, + "loss": 2.6064, + "step": 6454 + }, + { + "epoch": 0.28, + "grad_norm": 0.5078125, + "learning_rate": 0.0001910506474561888, + "loss": 2.6333, + "step": 6455 + }, + { + "epoch": 0.28, + "grad_norm": 0.494140625, + "learning_rate": 0.00019104787355645405, + "loss": 2.6246, + "step": 6456 + }, + { + "epoch": 0.28, + "grad_norm": 0.451171875, + "learning_rate": 0.00019104509924703714, + "loss": 2.6007, + "step": 6457 + }, + { + "epoch": 0.28, + "grad_norm": 0.451171875, + "learning_rate": 0.00019104232452795057, + "loss": 2.6627, + "step": 6458 + }, + { + "epoch": 0.28, + "grad_norm": 0.427734375, + "learning_rate": 0.00019103954939920682, + "loss": 2.5752, + "step": 6459 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.00019103677386081834, + "loss": 2.6422, + "step": 6460 + }, + { + "epoch": 0.28, + "grad_norm": 0.43359375, + "learning_rate": 0.00019103399791279766, + "loss": 2.6455, + "step": 6461 + }, + { + "epoch": 0.28, + "grad_norm": 0.419921875, + "learning_rate": 0.00019103122155515726, + "loss": 2.5225, + "step": 6462 + }, + { + "epoch": 0.28, + "grad_norm": 0.515625, + "learning_rate": 0.00019102844478790964, + "loss": 2.5313, + "step": 6463 + }, + { + "epoch": 0.28, + "grad_norm": 0.423828125, + "learning_rate": 0.00019102566761106728, + "loss": 2.7232, + "step": 6464 + }, + { + "epoch": 0.28, + "grad_norm": 0.50390625, + "learning_rate": 0.0001910228900246427, + "loss": 2.6899, + "step": 6465 + }, + { + "epoch": 0.28, + "grad_norm": 0.40625, + "learning_rate": 0.00019102011202864835, + "loss": 2.5492, + "step": 6466 + }, + { + "epoch": 0.28, + "grad_norm": 0.38671875, + "learning_rate": 0.00019101733362309676, + "loss": 2.6249, + "step": 6467 + }, + { + "epoch": 0.28, + "grad_norm": 0.498046875, + "learning_rate": 0.00019101455480800047, + "loss": 2.679, + "step": 6468 + }, + { + "epoch": 0.28, + "grad_norm": 0.470703125, + "learning_rate": 0.00019101177558337188, + "loss": 2.5863, + "step": 6469 + }, + { + "epoch": 0.28, + "grad_norm": 0.43359375, + "learning_rate": 0.00019100899594922362, + "loss": 2.6996, + "step": 6470 + }, + { + "epoch": 0.28, + "grad_norm": 0.447265625, + "learning_rate": 0.0001910062159055681, + "loss": 2.6328, + "step": 6471 + }, + { + "epoch": 0.28, + "grad_norm": 0.423828125, + "learning_rate": 0.0001910034354524179, + "loss": 2.5979, + "step": 6472 + }, + { + "epoch": 0.28, + "grad_norm": 0.4453125, + "learning_rate": 0.0001910006545897855, + "loss": 2.6657, + "step": 6473 + }, + { + "epoch": 0.28, + "grad_norm": 0.427734375, + "learning_rate": 0.00019099787331768338, + "loss": 2.7433, + "step": 6474 + }, + { + "epoch": 0.28, + "grad_norm": 0.41015625, + "learning_rate": 0.0001909950916361241, + "loss": 2.696, + "step": 6475 + }, + { + "epoch": 0.28, + "grad_norm": 0.38671875, + "learning_rate": 0.00019099230954512018, + "loss": 2.4712, + "step": 6476 + }, + { + "epoch": 0.28, + "grad_norm": 0.474609375, + "learning_rate": 0.0001909895270446841, + "loss": 2.6948, + "step": 6477 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.00019098674413482839, + "loss": 2.6698, + "step": 6478 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.0001909839608155656, + "loss": 2.6917, + "step": 6479 + }, + { + "epoch": 0.28, + "grad_norm": 0.421875, + "learning_rate": 0.00019098117708690824, + "loss": 2.6901, + "step": 6480 + }, + { + "epoch": 0.28, + "grad_norm": 0.4375, + "learning_rate": 0.00019097839294886882, + "loss": 2.753, + "step": 6481 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.00019097560840145993, + "loss": 2.4472, + "step": 6482 + }, + { + "epoch": 0.28, + "grad_norm": 0.3984375, + "learning_rate": 0.00019097282344469397, + "loss": 2.5924, + "step": 6483 + }, + { + "epoch": 0.28, + "grad_norm": 0.38671875, + "learning_rate": 0.0001909700380785836, + "loss": 2.739, + "step": 6484 + }, + { + "epoch": 0.28, + "grad_norm": 0.455078125, + "learning_rate": 0.0001909672523031413, + "loss": 2.773, + "step": 6485 + }, + { + "epoch": 0.28, + "grad_norm": 0.427734375, + "learning_rate": 0.00019096446611837958, + "loss": 2.5974, + "step": 6486 + }, + { + "epoch": 0.28, + "grad_norm": 0.400390625, + "learning_rate": 0.00019096167952431105, + "loss": 2.567, + "step": 6487 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.00019095889252094818, + "loss": 2.693, + "step": 6488 + }, + { + "epoch": 0.28, + "grad_norm": 0.49609375, + "learning_rate": 0.00019095610510830356, + "loss": 2.4919, + "step": 6489 + }, + { + "epoch": 0.28, + "grad_norm": 0.412109375, + "learning_rate": 0.00019095331728638968, + "loss": 2.586, + "step": 6490 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.0001909505290552191, + "loss": 2.6379, + "step": 6491 + }, + { + "epoch": 0.28, + "grad_norm": 0.494140625, + "learning_rate": 0.00019094774041480443, + "loss": 2.8073, + "step": 6492 + }, + { + "epoch": 0.28, + "grad_norm": 0.416015625, + "learning_rate": 0.00019094495136515814, + "loss": 2.4884, + "step": 6493 + }, + { + "epoch": 0.28, + "grad_norm": 0.4609375, + "learning_rate": 0.00019094216190629278, + "loss": 2.5954, + "step": 6494 + }, + { + "epoch": 0.28, + "grad_norm": 0.4296875, + "learning_rate": 0.00019093937203822097, + "loss": 2.6033, + "step": 6495 + }, + { + "epoch": 0.28, + "grad_norm": 0.40625, + "learning_rate": 0.0001909365817609552, + "loss": 2.6129, + "step": 6496 + }, + { + "epoch": 0.28, + "grad_norm": 0.439453125, + "learning_rate": 0.00019093379107450802, + "loss": 2.7746, + "step": 6497 + }, + { + "epoch": 0.28, + "grad_norm": 0.48828125, + "learning_rate": 0.00019093099997889203, + "loss": 2.7041, + "step": 6498 + }, + { + "epoch": 0.28, + "grad_norm": 0.392578125, + "learning_rate": 0.0001909282084741198, + "loss": 2.5737, + "step": 6499 + }, + { + "epoch": 0.28, + "grad_norm": 0.5078125, + "learning_rate": 0.00019092541656020383, + "loss": 2.7539, + "step": 6500 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.00019092262423715673, + "loss": 2.5984, + "step": 6501 + }, + { + "epoch": 0.28, + "grad_norm": 0.490234375, + "learning_rate": 0.000190919831504991, + "loss": 2.7689, + "step": 6502 + }, + { + "epoch": 0.28, + "grad_norm": 0.453125, + "learning_rate": 0.00019091703836371932, + "loss": 2.6595, + "step": 6503 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.0001909142448133542, + "loss": 2.6314, + "step": 6504 + }, + { + "epoch": 0.28, + "grad_norm": 0.435546875, + "learning_rate": 0.00019091145085390815, + "loss": 2.5468, + "step": 6505 + }, + { + "epoch": 0.28, + "grad_norm": 0.46875, + "learning_rate": 0.00019090865648539382, + "loss": 2.6934, + "step": 6506 + }, + { + "epoch": 0.28, + "grad_norm": 0.462890625, + "learning_rate": 0.00019090586170782378, + "loss": 2.6371, + "step": 6507 + }, + { + "epoch": 0.28, + "grad_norm": 0.47265625, + "learning_rate": 0.00019090306652121055, + "loss": 2.6902, + "step": 6508 + }, + { + "epoch": 0.28, + "grad_norm": 0.4609375, + "learning_rate": 0.00019090027092556676, + "loss": 2.7554, + "step": 6509 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.00019089747492090497, + "loss": 2.5568, + "step": 6510 + }, + { + "epoch": 0.28, + "grad_norm": 0.43359375, + "learning_rate": 0.00019089467850723777, + "loss": 2.6317, + "step": 6511 + }, + { + "epoch": 0.28, + "grad_norm": 0.42578125, + "learning_rate": 0.00019089188168457772, + "loss": 2.8034, + "step": 6512 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.00019088908445293742, + "loss": 2.6444, + "step": 6513 + }, + { + "epoch": 0.28, + "grad_norm": 0.4140625, + "learning_rate": 0.00019088628681232948, + "loss": 2.7487, + "step": 6514 + }, + { + "epoch": 0.28, + "grad_norm": 0.431640625, + "learning_rate": 0.00019088348876276643, + "loss": 2.6495, + "step": 6515 + }, + { + "epoch": 0.28, + "grad_norm": 0.39453125, + "learning_rate": 0.0001908806903042609, + "loss": 2.6788, + "step": 6516 + }, + { + "epoch": 0.28, + "grad_norm": 0.4453125, + "learning_rate": 0.00019087789143682547, + "loss": 2.7172, + "step": 6517 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.0001908750921604728, + "loss": 2.6586, + "step": 6518 + }, + { + "epoch": 0.28, + "grad_norm": 0.466796875, + "learning_rate": 0.00019087229247521538, + "loss": 2.6169, + "step": 6519 + }, + { + "epoch": 0.28, + "grad_norm": 0.41015625, + "learning_rate": 0.00019086949238106583, + "loss": 2.5505, + "step": 6520 + }, + { + "epoch": 0.28, + "grad_norm": 0.4609375, + "learning_rate": 0.0001908666918780368, + "loss": 2.5469, + "step": 6521 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.0001908638909661409, + "loss": 2.6041, + "step": 6522 + }, + { + "epoch": 0.28, + "grad_norm": 0.3984375, + "learning_rate": 0.00019086108964539065, + "loss": 2.4822, + "step": 6523 + }, + { + "epoch": 0.28, + "grad_norm": 0.50390625, + "learning_rate": 0.00019085828791579873, + "loss": 2.7616, + "step": 6524 + }, + { + "epoch": 0.28, + "grad_norm": 0.4921875, + "learning_rate": 0.00019085548577737772, + "loss": 2.84, + "step": 6525 + }, + { + "epoch": 0.28, + "grad_norm": 0.439453125, + "learning_rate": 0.0001908526832301402, + "loss": 2.6536, + "step": 6526 + }, + { + "epoch": 0.28, + "grad_norm": 0.419921875, + "learning_rate": 0.0001908498802740988, + "loss": 2.5198, + "step": 6527 + }, + { + "epoch": 0.28, + "grad_norm": 0.451171875, + "learning_rate": 0.00019084707690926616, + "loss": 2.5752, + "step": 6528 + }, + { + "epoch": 0.28, + "grad_norm": 0.447265625, + "learning_rate": 0.00019084427313565487, + "loss": 2.4963, + "step": 6529 + }, + { + "epoch": 0.28, + "grad_norm": 0.45703125, + "learning_rate": 0.00019084146895327758, + "loss": 2.6011, + "step": 6530 + }, + { + "epoch": 0.28, + "grad_norm": 0.4375, + "learning_rate": 0.00019083866436214684, + "loss": 2.6247, + "step": 6531 + }, + { + "epoch": 0.28, + "grad_norm": 0.435546875, + "learning_rate": 0.00019083585936227534, + "loss": 2.5974, + "step": 6532 + }, + { + "epoch": 0.28, + "grad_norm": 0.412109375, + "learning_rate": 0.00019083305395367565, + "loss": 2.582, + "step": 6533 + }, + { + "epoch": 0.28, + "grad_norm": 0.4609375, + "learning_rate": 0.0001908302481363604, + "loss": 2.6542, + "step": 6534 + }, + { + "epoch": 0.28, + "grad_norm": 0.421875, + "learning_rate": 0.00019082744191034224, + "loss": 2.5189, + "step": 6535 + }, + { + "epoch": 0.28, + "grad_norm": 0.439453125, + "learning_rate": 0.0001908246352756338, + "loss": 2.6177, + "step": 6536 + }, + { + "epoch": 0.28, + "grad_norm": 0.4765625, + "learning_rate": 0.00019082182823224769, + "loss": 2.5682, + "step": 6537 + }, + { + "epoch": 0.28, + "grad_norm": 0.5, + "learning_rate": 0.00019081902078019652, + "loss": 2.6474, + "step": 6538 + }, + { + "epoch": 0.28, + "grad_norm": 0.498046875, + "learning_rate": 0.00019081621291949297, + "loss": 2.7101, + "step": 6539 + }, + { + "epoch": 0.28, + "grad_norm": 0.3984375, + "learning_rate": 0.00019081340465014964, + "loss": 2.6562, + "step": 6540 + }, + { + "epoch": 0.28, + "grad_norm": 0.451171875, + "learning_rate": 0.0001908105959721792, + "loss": 2.5498, + "step": 6541 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.00019080778688559423, + "loss": 2.4914, + "step": 6542 + }, + { + "epoch": 0.28, + "grad_norm": 0.427734375, + "learning_rate": 0.00019080497739040744, + "loss": 2.5845, + "step": 6543 + }, + { + "epoch": 0.28, + "grad_norm": 0.46484375, + "learning_rate": 0.00019080216748663142, + "loss": 2.6883, + "step": 6544 + }, + { + "epoch": 0.28, + "grad_norm": 0.4921875, + "learning_rate": 0.00019079935717427886, + "loss": 2.6738, + "step": 6545 + }, + { + "epoch": 0.28, + "grad_norm": 0.5, + "learning_rate": 0.00019079654645336234, + "loss": 2.5611, + "step": 6546 + }, + { + "epoch": 0.28, + "grad_norm": 0.484375, + "learning_rate": 0.00019079373532389452, + "loss": 2.5072, + "step": 6547 + }, + { + "epoch": 0.28, + "grad_norm": 0.466796875, + "learning_rate": 0.00019079092378588813, + "loss": 2.5268, + "step": 6548 + }, + { + "epoch": 0.28, + "grad_norm": 0.419921875, + "learning_rate": 0.00019078811183935573, + "loss": 2.6618, + "step": 6549 + }, + { + "epoch": 0.28, + "grad_norm": 0.484375, + "learning_rate": 0.00019078529948431, + "loss": 2.6664, + "step": 6550 + }, + { + "epoch": 0.28, + "grad_norm": 0.50390625, + "learning_rate": 0.00019078248672076362, + "loss": 2.6728, + "step": 6551 + }, + { + "epoch": 0.28, + "grad_norm": 0.4609375, + "learning_rate": 0.00019077967354872924, + "loss": 2.5716, + "step": 6552 + }, + { + "epoch": 0.28, + "grad_norm": 0.423828125, + "learning_rate": 0.00019077685996821947, + "loss": 2.531, + "step": 6553 + }, + { + "epoch": 0.28, + "grad_norm": 0.453125, + "learning_rate": 0.00019077404597924705, + "loss": 2.6707, + "step": 6554 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.00019077123158182455, + "loss": 2.5844, + "step": 6555 + }, + { + "epoch": 0.28, + "grad_norm": 0.451171875, + "learning_rate": 0.00019076841677596472, + "loss": 2.7514, + "step": 6556 + }, + { + "epoch": 0.28, + "grad_norm": 0.458984375, + "learning_rate": 0.00019076560156168016, + "loss": 2.5247, + "step": 6557 + }, + { + "epoch": 0.28, + "grad_norm": 0.50390625, + "learning_rate": 0.00019076278593898355, + "loss": 2.7225, + "step": 6558 + }, + { + "epoch": 0.28, + "grad_norm": 0.447265625, + "learning_rate": 0.0001907599699078876, + "loss": 2.5058, + "step": 6559 + }, + { + "epoch": 0.28, + "grad_norm": 0.51171875, + "learning_rate": 0.00019075715346840495, + "loss": 2.6438, + "step": 6560 + }, + { + "epoch": 0.28, + "grad_norm": 0.4765625, + "learning_rate": 0.00019075433662054829, + "loss": 2.5733, + "step": 6561 + }, + { + "epoch": 0.28, + "grad_norm": 0.4296875, + "learning_rate": 0.00019075151936433023, + "loss": 2.5954, + "step": 6562 + }, + { + "epoch": 0.28, + "grad_norm": 0.455078125, + "learning_rate": 0.00019074870169976356, + "loss": 2.4232, + "step": 6563 + }, + { + "epoch": 0.28, + "grad_norm": 0.3984375, + "learning_rate": 0.00019074588362686087, + "loss": 2.6236, + "step": 6564 + }, + { + "epoch": 0.28, + "grad_norm": 0.45703125, + "learning_rate": 0.00019074306514563487, + "loss": 2.6546, + "step": 6565 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.00019074024625609826, + "loss": 2.6971, + "step": 6566 + }, + { + "epoch": 0.28, + "grad_norm": 0.474609375, + "learning_rate": 0.00019073742695826365, + "loss": 2.6979, + "step": 6567 + }, + { + "epoch": 0.28, + "grad_norm": 0.484375, + "learning_rate": 0.00019073460725214383, + "loss": 2.6817, + "step": 6568 + }, + { + "epoch": 0.28, + "grad_norm": 0.462890625, + "learning_rate": 0.0001907317871377514, + "loss": 2.5909, + "step": 6569 + }, + { + "epoch": 0.28, + "grad_norm": 0.421875, + "learning_rate": 0.00019072896661509914, + "loss": 2.5974, + "step": 6570 + }, + { + "epoch": 0.28, + "grad_norm": 0.431640625, + "learning_rate": 0.00019072614568419965, + "loss": 2.6068, + "step": 6571 + }, + { + "epoch": 0.28, + "grad_norm": 0.478515625, + "learning_rate": 0.00019072332434506568, + "loss": 2.6626, + "step": 6572 + }, + { + "epoch": 0.28, + "grad_norm": 0.50390625, + "learning_rate": 0.0001907205025977099, + "loss": 2.7178, + "step": 6573 + }, + { + "epoch": 0.28, + "grad_norm": 0.42578125, + "learning_rate": 0.00019071768044214498, + "loss": 2.6829, + "step": 6574 + }, + { + "epoch": 0.28, + "grad_norm": 0.51953125, + "learning_rate": 0.00019071485787838368, + "loss": 2.7618, + "step": 6575 + }, + { + "epoch": 0.28, + "grad_norm": 0.46484375, + "learning_rate": 0.0001907120349064387, + "loss": 2.7158, + "step": 6576 + }, + { + "epoch": 0.28, + "grad_norm": 0.39453125, + "learning_rate": 0.00019070921152632267, + "loss": 2.7243, + "step": 6577 + }, + { + "epoch": 0.28, + "grad_norm": 0.423828125, + "learning_rate": 0.00019070638773804833, + "loss": 2.6854, + "step": 6578 + }, + { + "epoch": 0.28, + "grad_norm": 0.451171875, + "learning_rate": 0.00019070356354162842, + "loss": 2.6761, + "step": 6579 + }, + { + "epoch": 0.28, + "grad_norm": 0.408203125, + "learning_rate": 0.00019070073893707562, + "loss": 2.7132, + "step": 6580 + }, + { + "epoch": 0.28, + "grad_norm": 0.435546875, + "learning_rate": 0.00019069791392440264, + "loss": 2.6286, + "step": 6581 + }, + { + "epoch": 0.28, + "grad_norm": 0.455078125, + "learning_rate": 0.00019069508850362219, + "loss": 2.5861, + "step": 6582 + }, + { + "epoch": 0.28, + "grad_norm": 0.447265625, + "learning_rate": 0.000190692262674747, + "loss": 2.5257, + "step": 6583 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.00019068943643778976, + "loss": 2.6204, + "step": 6584 + }, + { + "epoch": 0.28, + "grad_norm": 0.43359375, + "learning_rate": 0.0001906866097927632, + "loss": 2.6482, + "step": 6585 + }, + { + "epoch": 0.28, + "grad_norm": 0.478515625, + "learning_rate": 0.00019068378273968004, + "loss": 2.7272, + "step": 6586 + }, + { + "epoch": 0.28, + "grad_norm": 0.384765625, + "learning_rate": 0.00019068095527855302, + "loss": 2.5623, + "step": 6587 + }, + { + "epoch": 0.28, + "grad_norm": 0.419921875, + "learning_rate": 0.0001906781274093948, + "loss": 2.6793, + "step": 6588 + }, + { + "epoch": 0.28, + "grad_norm": 0.388671875, + "learning_rate": 0.00019067529913221817, + "loss": 2.5751, + "step": 6589 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.0001906724704470358, + "loss": 2.6269, + "step": 6590 + }, + { + "epoch": 0.28, + "grad_norm": 0.44140625, + "learning_rate": 0.0001906696413538605, + "loss": 2.5964, + "step": 6591 + }, + { + "epoch": 0.28, + "grad_norm": 0.380859375, + "learning_rate": 0.00019066681185270494, + "loss": 2.6737, + "step": 6592 + }, + { + "epoch": 0.28, + "grad_norm": 0.416015625, + "learning_rate": 0.00019066398194358185, + "loss": 2.7958, + "step": 6593 + }, + { + "epoch": 0.28, + "grad_norm": 0.419921875, + "learning_rate": 0.000190661151626504, + "loss": 2.649, + "step": 6594 + }, + { + "epoch": 0.28, + "grad_norm": 0.451171875, + "learning_rate": 0.00019065832090148406, + "loss": 2.6905, + "step": 6595 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.00019065548976853483, + "loss": 2.6389, + "step": 6596 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.000190652658227669, + "loss": 2.7011, + "step": 6597 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.00019064982627889937, + "loss": 2.8578, + "step": 6598 + }, + { + "epoch": 0.28, + "grad_norm": 0.412109375, + "learning_rate": 0.00019064699392223867, + "loss": 2.7125, + "step": 6599 + }, + { + "epoch": 0.28, + "grad_norm": 0.390625, + "learning_rate": 0.00019064416115769956, + "loss": 2.6754, + "step": 6600 + }, + { + "epoch": 0.28, + "grad_norm": 0.46875, + "learning_rate": 0.0001906413279852949, + "loss": 2.6753, + "step": 6601 + }, + { + "epoch": 0.28, + "grad_norm": 0.490234375, + "learning_rate": 0.00019063849440503736, + "loss": 2.5896, + "step": 6602 + }, + { + "epoch": 0.28, + "grad_norm": 0.45703125, + "learning_rate": 0.00019063566041693976, + "loss": 2.5892, + "step": 6603 + }, + { + "epoch": 0.28, + "grad_norm": 0.439453125, + "learning_rate": 0.00019063282602101476, + "loss": 2.686, + "step": 6604 + }, + { + "epoch": 0.28, + "grad_norm": 0.44140625, + "learning_rate": 0.00019062999121727522, + "loss": 2.5953, + "step": 6605 + }, + { + "epoch": 0.28, + "grad_norm": 0.421875, + "learning_rate": 0.0001906271560057338, + "loss": 2.6173, + "step": 6606 + }, + { + "epoch": 0.28, + "grad_norm": 0.4453125, + "learning_rate": 0.00019062432038640333, + "loss": 2.602, + "step": 6607 + }, + { + "epoch": 0.28, + "grad_norm": 0.52734375, + "learning_rate": 0.0001906214843592965, + "loss": 2.6589, + "step": 6608 + }, + { + "epoch": 0.28, + "grad_norm": 0.4296875, + "learning_rate": 0.00019061864792442612, + "loss": 2.5721, + "step": 6609 + }, + { + "epoch": 0.28, + "grad_norm": 0.408203125, + "learning_rate": 0.00019061581108180495, + "loss": 2.6238, + "step": 6610 + }, + { + "epoch": 0.28, + "grad_norm": 0.44140625, + "learning_rate": 0.00019061297383144576, + "loss": 2.527, + "step": 6611 + }, + { + "epoch": 0.28, + "grad_norm": 0.4765625, + "learning_rate": 0.00019061013617336126, + "loss": 2.7459, + "step": 6612 + }, + { + "epoch": 0.28, + "grad_norm": 0.380859375, + "learning_rate": 0.0001906072981075643, + "loss": 2.4994, + "step": 6613 + }, + { + "epoch": 0.28, + "grad_norm": 0.416015625, + "learning_rate": 0.0001906044596340676, + "loss": 2.6596, + "step": 6614 + }, + { + "epoch": 0.28, + "grad_norm": 0.453125, + "learning_rate": 0.00019060162075288398, + "loss": 2.6402, + "step": 6615 + }, + { + "epoch": 0.28, + "grad_norm": 0.47265625, + "learning_rate": 0.00019059878146402613, + "loss": 2.5378, + "step": 6616 + }, + { + "epoch": 0.28, + "grad_norm": 0.453125, + "learning_rate": 0.00019059594176750686, + "loss": 2.5416, + "step": 6617 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.000190593101663339, + "loss": 2.4584, + "step": 6618 + }, + { + "epoch": 0.28, + "grad_norm": 0.4296875, + "learning_rate": 0.0001905902611515353, + "loss": 2.6609, + "step": 6619 + }, + { + "epoch": 0.28, + "grad_norm": 0.44140625, + "learning_rate": 0.00019058742023210848, + "loss": 2.5462, + "step": 6620 + }, + { + "epoch": 0.28, + "grad_norm": 0.447265625, + "learning_rate": 0.0001905845789050714, + "loss": 2.7026, + "step": 6621 + }, + { + "epoch": 0.28, + "grad_norm": 0.3984375, + "learning_rate": 0.00019058173717043683, + "loss": 2.5067, + "step": 6622 + }, + { + "epoch": 0.28, + "grad_norm": 0.466796875, + "learning_rate": 0.00019057889502821756, + "loss": 2.6801, + "step": 6623 + }, + { + "epoch": 0.28, + "grad_norm": 0.390625, + "learning_rate": 0.00019057605247842633, + "loss": 2.6689, + "step": 6624 + }, + { + "epoch": 0.28, + "grad_norm": 0.40234375, + "learning_rate": 0.00019057320952107598, + "loss": 2.5725, + "step": 6625 + }, + { + "epoch": 0.28, + "grad_norm": 0.46484375, + "learning_rate": 0.00019057036615617928, + "loss": 2.7035, + "step": 6626 + }, + { + "epoch": 0.28, + "grad_norm": 0.447265625, + "learning_rate": 0.00019056752238374907, + "loss": 2.5427, + "step": 6627 + }, + { + "epoch": 0.28, + "grad_norm": 0.4765625, + "learning_rate": 0.00019056467820379805, + "loss": 2.6853, + "step": 6628 + }, + { + "epoch": 0.28, + "grad_norm": 0.37890625, + "learning_rate": 0.00019056183361633913, + "loss": 2.5745, + "step": 6629 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.00019055898862138505, + "loss": 2.6379, + "step": 6630 + }, + { + "epoch": 0.28, + "grad_norm": 0.484375, + "learning_rate": 0.00019055614321894856, + "loss": 2.5776, + "step": 6631 + }, + { + "epoch": 0.28, + "grad_norm": 0.47265625, + "learning_rate": 0.0001905532974090426, + "loss": 2.7391, + "step": 6632 + }, + { + "epoch": 0.28, + "grad_norm": 0.45703125, + "learning_rate": 0.00019055045119167985, + "loss": 2.7254, + "step": 6633 + }, + { + "epoch": 0.28, + "grad_norm": 0.42578125, + "learning_rate": 0.00019054760456687313, + "loss": 2.5845, + "step": 6634 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.00019054475753463535, + "loss": 2.6539, + "step": 6635 + }, + { + "epoch": 0.28, + "grad_norm": 0.50390625, + "learning_rate": 0.0001905419100949792, + "loss": 2.7713, + "step": 6636 + }, + { + "epoch": 0.28, + "grad_norm": 0.375, + "learning_rate": 0.00019053906224791755, + "loss": 2.4548, + "step": 6637 + }, + { + "epoch": 0.28, + "grad_norm": 0.439453125, + "learning_rate": 0.00019053621399346321, + "loss": 2.5854, + "step": 6638 + }, + { + "epoch": 0.28, + "grad_norm": 0.462890625, + "learning_rate": 0.00019053336533162902, + "loss": 2.6354, + "step": 6639 + }, + { + "epoch": 0.28, + "grad_norm": 0.484375, + "learning_rate": 0.00019053051626242775, + "loss": 2.6186, + "step": 6640 + }, + { + "epoch": 0.28, + "grad_norm": 0.498046875, + "learning_rate": 0.00019052766678587224, + "loss": 2.6094, + "step": 6641 + }, + { + "epoch": 0.28, + "grad_norm": 0.50390625, + "learning_rate": 0.0001905248169019753, + "loss": 2.6664, + "step": 6642 + }, + { + "epoch": 0.28, + "grad_norm": 0.4375, + "learning_rate": 0.00019052196661074983, + "loss": 2.7024, + "step": 6643 + }, + { + "epoch": 0.28, + "grad_norm": 0.4296875, + "learning_rate": 0.00019051911591220854, + "loss": 2.6334, + "step": 6644 + }, + { + "epoch": 0.28, + "grad_norm": 0.416015625, + "learning_rate": 0.0001905162648063643, + "loss": 2.6204, + "step": 6645 + }, + { + "epoch": 0.28, + "grad_norm": 0.4453125, + "learning_rate": 0.00019051341329322997, + "loss": 2.5562, + "step": 6646 + }, + { + "epoch": 0.28, + "grad_norm": 0.4375, + "learning_rate": 0.00019051056137281837, + "loss": 2.58, + "step": 6647 + }, + { + "epoch": 0.28, + "grad_norm": 0.47265625, + "learning_rate": 0.0001905077090451423, + "loss": 2.6581, + "step": 6648 + }, + { + "epoch": 0.28, + "grad_norm": 0.396484375, + "learning_rate": 0.00019050485631021464, + "loss": 2.5494, + "step": 6649 + }, + { + "epoch": 0.28, + "grad_norm": 0.455078125, + "learning_rate": 0.00019050200316804818, + "loss": 2.6725, + "step": 6650 + }, + { + "epoch": 0.28, + "grad_norm": 0.427734375, + "learning_rate": 0.0001904991496186558, + "loss": 2.5998, + "step": 6651 + }, + { + "epoch": 0.28, + "grad_norm": 0.4296875, + "learning_rate": 0.0001904962956620503, + "loss": 2.6465, + "step": 6652 + }, + { + "epoch": 0.28, + "grad_norm": 0.423828125, + "learning_rate": 0.00019049344129824455, + "loss": 2.6797, + "step": 6653 + }, + { + "epoch": 0.28, + "grad_norm": 0.484375, + "learning_rate": 0.0001904905865272514, + "loss": 2.7609, + "step": 6654 + }, + { + "epoch": 0.28, + "grad_norm": 0.439453125, + "learning_rate": 0.00019048773134908365, + "loss": 2.7016, + "step": 6655 + }, + { + "epoch": 0.28, + "grad_norm": 0.43359375, + "learning_rate": 0.0001904848757637542, + "loss": 2.733, + "step": 6656 + }, + { + "epoch": 0.28, + "grad_norm": 0.45703125, + "learning_rate": 0.00019048201977127585, + "loss": 2.4612, + "step": 6657 + }, + { + "epoch": 0.28, + "grad_norm": 0.466796875, + "learning_rate": 0.0001904791633716615, + "loss": 2.7128, + "step": 6658 + }, + { + "epoch": 0.28, + "grad_norm": 0.41796875, + "learning_rate": 0.00019047630656492402, + "loss": 2.5534, + "step": 6659 + }, + { + "epoch": 0.28, + "grad_norm": 0.439453125, + "learning_rate": 0.00019047344935107615, + "loss": 2.6027, + "step": 6660 + }, + { + "epoch": 0.28, + "grad_norm": 0.4765625, + "learning_rate": 0.0001904705917301309, + "loss": 2.7831, + "step": 6661 + }, + { + "epoch": 0.28, + "grad_norm": 0.474609375, + "learning_rate": 0.000190467733702101, + "loss": 2.652, + "step": 6662 + }, + { + "epoch": 0.28, + "grad_norm": 0.46484375, + "learning_rate": 0.00019046487526699937, + "loss": 2.623, + "step": 6663 + }, + { + "epoch": 0.28, + "grad_norm": 0.396484375, + "learning_rate": 0.00019046201642483889, + "loss": 2.5712, + "step": 6664 + }, + { + "epoch": 0.28, + "grad_norm": 0.486328125, + "learning_rate": 0.00019045915717563235, + "loss": 2.6829, + "step": 6665 + }, + { + "epoch": 0.28, + "grad_norm": 0.462890625, + "learning_rate": 0.00019045629751939267, + "loss": 2.6326, + "step": 6666 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.00019045343745613272, + "loss": 2.5801, + "step": 6667 + }, + { + "epoch": 0.28, + "grad_norm": 0.392578125, + "learning_rate": 0.00019045057698586536, + "loss": 2.4998, + "step": 6668 + }, + { + "epoch": 0.28, + "grad_norm": 0.439453125, + "learning_rate": 0.00019044771610860346, + "loss": 2.5978, + "step": 6669 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.0001904448548243599, + "loss": 2.6805, + "step": 6670 + }, + { + "epoch": 0.28, + "grad_norm": 0.4453125, + "learning_rate": 0.00019044199313314754, + "loss": 2.6382, + "step": 6671 + }, + { + "epoch": 0.28, + "grad_norm": 0.408203125, + "learning_rate": 0.00019043913103497926, + "loss": 2.5604, + "step": 6672 + }, + { + "epoch": 0.28, + "grad_norm": 0.421875, + "learning_rate": 0.00019043626852986793, + "loss": 2.5384, + "step": 6673 + }, + { + "epoch": 0.28, + "grad_norm": 0.44921875, + "learning_rate": 0.0001904334056178265, + "loss": 2.6675, + "step": 6674 + }, + { + "epoch": 0.28, + "grad_norm": 0.3984375, + "learning_rate": 0.0001904305422988677, + "loss": 2.5791, + "step": 6675 + }, + { + "epoch": 0.28, + "grad_norm": 0.447265625, + "learning_rate": 0.00019042767857300457, + "loss": 2.607, + "step": 6676 + }, + { + "epoch": 0.28, + "grad_norm": 0.451171875, + "learning_rate": 0.00019042481444024993, + "loss": 2.5418, + "step": 6677 + }, + { + "epoch": 0.28, + "grad_norm": 0.46875, + "learning_rate": 0.00019042194990061666, + "loss": 2.6615, + "step": 6678 + }, + { + "epoch": 0.28, + "grad_norm": 0.443359375, + "learning_rate": 0.00019041908495411766, + "loss": 2.6412, + "step": 6679 + }, + { + "epoch": 0.28, + "grad_norm": 0.4609375, + "learning_rate": 0.00019041621960076582, + "loss": 2.6522, + "step": 6680 + }, + { + "epoch": 0.28, + "grad_norm": 0.458984375, + "learning_rate": 0.00019041335384057404, + "loss": 2.6121, + "step": 6681 + }, + { + "epoch": 0.28, + "grad_norm": 0.5, + "learning_rate": 0.00019041048767355523, + "loss": 2.6672, + "step": 6682 + }, + { + "epoch": 0.28, + "grad_norm": 0.40234375, + "learning_rate": 0.00019040762109972224, + "loss": 2.5679, + "step": 6683 + }, + { + "epoch": 0.28, + "grad_norm": 0.42578125, + "learning_rate": 0.000190404754119088, + "loss": 2.6297, + "step": 6684 + }, + { + "epoch": 0.28, + "grad_norm": 0.408203125, + "learning_rate": 0.00019040188673166543, + "loss": 2.6205, + "step": 6685 + }, + { + "epoch": 0.28, + "grad_norm": 0.48828125, + "learning_rate": 0.00019039901893746738, + "loss": 2.6167, + "step": 6686 + }, + { + "epoch": 0.28, + "grad_norm": 0.46875, + "learning_rate": 0.00019039615073650678, + "loss": 2.6201, + "step": 6687 + }, + { + "epoch": 0.28, + "grad_norm": 0.46875, + "learning_rate": 0.00019039328212879655, + "loss": 2.7091, + "step": 6688 + }, + { + "epoch": 0.29, + "grad_norm": 0.43359375, + "learning_rate": 0.00019039041311434958, + "loss": 2.4926, + "step": 6689 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019038754369317878, + "loss": 2.6792, + "step": 6690 + }, + { + "epoch": 0.29, + "grad_norm": 0.4140625, + "learning_rate": 0.0001903846738652971, + "loss": 2.524, + "step": 6691 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019038180363071738, + "loss": 2.7391, + "step": 6692 + }, + { + "epoch": 0.29, + "grad_norm": 0.4609375, + "learning_rate": 0.00019037893298945257, + "loss": 2.5957, + "step": 6693 + }, + { + "epoch": 0.29, + "grad_norm": 0.4296875, + "learning_rate": 0.00019037606194151563, + "loss": 2.7016, + "step": 6694 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019037319048691943, + "loss": 2.7321, + "step": 6695 + }, + { + "epoch": 0.29, + "grad_norm": 0.431640625, + "learning_rate": 0.0001903703186256769, + "loss": 2.592, + "step": 6696 + }, + { + "epoch": 0.29, + "grad_norm": 0.44921875, + "learning_rate": 0.00019036744635780092, + "loss": 2.6937, + "step": 6697 + }, + { + "epoch": 0.29, + "grad_norm": 0.4296875, + "learning_rate": 0.0001903645736833045, + "loss": 2.5982, + "step": 6698 + }, + { + "epoch": 0.29, + "grad_norm": 0.451171875, + "learning_rate": 0.0001903617006022005, + "loss": 2.6942, + "step": 6699 + }, + { + "epoch": 0.29, + "grad_norm": 0.4765625, + "learning_rate": 0.0001903588271145019, + "loss": 2.7131, + "step": 6700 + }, + { + "epoch": 0.29, + "grad_norm": 0.396484375, + "learning_rate": 0.00019035595322022158, + "loss": 2.3987, + "step": 6701 + }, + { + "epoch": 0.29, + "grad_norm": 0.42578125, + "learning_rate": 0.0001903530789193725, + "loss": 2.5523, + "step": 6702 + }, + { + "epoch": 0.29, + "grad_norm": 0.408203125, + "learning_rate": 0.00019035020421196759, + "loss": 2.5189, + "step": 6703 + }, + { + "epoch": 0.29, + "grad_norm": 0.455078125, + "learning_rate": 0.00019034732909801975, + "loss": 2.6893, + "step": 6704 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019034445357754196, + "loss": 2.8113, + "step": 6705 + }, + { + "epoch": 0.29, + "grad_norm": 0.40234375, + "learning_rate": 0.00019034157765054717, + "loss": 2.7315, + "step": 6706 + }, + { + "epoch": 0.29, + "grad_norm": 0.416015625, + "learning_rate": 0.00019033870131704826, + "loss": 2.5821, + "step": 6707 + }, + { + "epoch": 0.29, + "grad_norm": 0.443359375, + "learning_rate": 0.00019033582457705824, + "loss": 2.7681, + "step": 6708 + }, + { + "epoch": 0.29, + "grad_norm": 0.3984375, + "learning_rate": 0.00019033294743059002, + "loss": 2.6016, + "step": 6709 + }, + { + "epoch": 0.29, + "grad_norm": 0.4296875, + "learning_rate": 0.00019033006987765653, + "loss": 2.6837, + "step": 6710 + }, + { + "epoch": 0.29, + "grad_norm": 0.4765625, + "learning_rate": 0.00019032719191827075, + "loss": 2.6137, + "step": 6711 + }, + { + "epoch": 0.29, + "grad_norm": 0.546875, + "learning_rate": 0.00019032431355244563, + "loss": 2.7751, + "step": 6712 + }, + { + "epoch": 0.29, + "grad_norm": 0.455078125, + "learning_rate": 0.00019032143478019408, + "loss": 2.646, + "step": 6713 + }, + { + "epoch": 0.29, + "grad_norm": 0.4609375, + "learning_rate": 0.00019031855560152912, + "loss": 2.7392, + "step": 6714 + }, + { + "epoch": 0.29, + "grad_norm": 0.466796875, + "learning_rate": 0.00019031567601646363, + "loss": 2.6125, + "step": 6715 + }, + { + "epoch": 0.29, + "grad_norm": 0.404296875, + "learning_rate": 0.00019031279602501063, + "loss": 2.5237, + "step": 6716 + }, + { + "epoch": 0.29, + "grad_norm": 0.451171875, + "learning_rate": 0.00019030991562718304, + "loss": 2.7866, + "step": 6717 + }, + { + "epoch": 0.29, + "grad_norm": 0.419921875, + "learning_rate": 0.00019030703482299385, + "loss": 2.6897, + "step": 6718 + }, + { + "epoch": 0.29, + "grad_norm": 0.4140625, + "learning_rate": 0.000190304153612456, + "loss": 2.6851, + "step": 6719 + }, + { + "epoch": 0.29, + "grad_norm": 0.416015625, + "learning_rate": 0.00019030127199558242, + "loss": 2.5468, + "step": 6720 + }, + { + "epoch": 0.29, + "grad_norm": 0.486328125, + "learning_rate": 0.00019029838997238616, + "loss": 2.695, + "step": 6721 + }, + { + "epoch": 0.29, + "grad_norm": 0.408203125, + "learning_rate": 0.00019029550754288014, + "loss": 2.7276, + "step": 6722 + }, + { + "epoch": 0.29, + "grad_norm": 0.421875, + "learning_rate": 0.00019029262470707733, + "loss": 2.5576, + "step": 6723 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019028974146499073, + "loss": 2.5481, + "step": 6724 + }, + { + "epoch": 0.29, + "grad_norm": 0.44140625, + "learning_rate": 0.00019028685781663325, + "loss": 2.6412, + "step": 6725 + }, + { + "epoch": 0.29, + "grad_norm": 0.423828125, + "learning_rate": 0.00019028397376201795, + "loss": 2.6043, + "step": 6726 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00019028108930115774, + "loss": 2.665, + "step": 6727 + }, + { + "epoch": 0.29, + "grad_norm": 0.447265625, + "learning_rate": 0.00019027820443406562, + "loss": 2.6988, + "step": 6728 + }, + { + "epoch": 0.29, + "grad_norm": 0.455078125, + "learning_rate": 0.00019027531916075455, + "loss": 2.5954, + "step": 6729 + }, + { + "epoch": 0.29, + "grad_norm": 0.466796875, + "learning_rate": 0.00019027243348123757, + "loss": 2.7956, + "step": 6730 + }, + { + "epoch": 0.29, + "grad_norm": 0.44140625, + "learning_rate": 0.00019026954739552763, + "loss": 2.6092, + "step": 6731 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.0001902666609036377, + "loss": 2.6888, + "step": 6732 + }, + { + "epoch": 0.29, + "grad_norm": 0.427734375, + "learning_rate": 0.00019026377400558077, + "loss": 2.5022, + "step": 6733 + }, + { + "epoch": 0.29, + "grad_norm": 0.4765625, + "learning_rate": 0.0001902608867013699, + "loss": 2.6936, + "step": 6734 + }, + { + "epoch": 0.29, + "grad_norm": 0.484375, + "learning_rate": 0.000190257998991018, + "loss": 2.7894, + "step": 6735 + }, + { + "epoch": 0.29, + "grad_norm": 0.4375, + "learning_rate": 0.00019025511087453808, + "loss": 2.6301, + "step": 6736 + }, + { + "epoch": 0.29, + "grad_norm": 0.421875, + "learning_rate": 0.00019025222235194314, + "loss": 2.5067, + "step": 6737 + }, + { + "epoch": 0.29, + "grad_norm": 0.431640625, + "learning_rate": 0.00019024933342324618, + "loss": 2.7066, + "step": 6738 + }, + { + "epoch": 0.29, + "grad_norm": 0.408203125, + "learning_rate": 0.00019024644408846019, + "loss": 2.6295, + "step": 6739 + }, + { + "epoch": 0.29, + "grad_norm": 0.40234375, + "learning_rate": 0.00019024355434759822, + "loss": 2.6321, + "step": 6740 + }, + { + "epoch": 0.29, + "grad_norm": 0.421875, + "learning_rate": 0.00019024066420067324, + "loss": 2.6188, + "step": 6741 + }, + { + "epoch": 0.29, + "grad_norm": 0.41015625, + "learning_rate": 0.0001902377736476982, + "loss": 2.6613, + "step": 6742 + }, + { + "epoch": 0.29, + "grad_norm": 0.451171875, + "learning_rate": 0.0001902348826886862, + "loss": 2.623, + "step": 6743 + }, + { + "epoch": 0.29, + "grad_norm": 0.439453125, + "learning_rate": 0.0001902319913236502, + "loss": 2.521, + "step": 6744 + }, + { + "epoch": 0.29, + "grad_norm": 0.400390625, + "learning_rate": 0.0001902290995526032, + "loss": 2.5906, + "step": 6745 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00019022620737555826, + "loss": 2.5964, + "step": 6746 + }, + { + "epoch": 0.29, + "grad_norm": 0.451171875, + "learning_rate": 0.00019022331479252833, + "loss": 2.6259, + "step": 6747 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019022042180352647, + "loss": 2.594, + "step": 6748 + }, + { + "epoch": 0.29, + "grad_norm": 0.435546875, + "learning_rate": 0.00019021752840856567, + "loss": 2.697, + "step": 6749 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019021463460765895, + "loss": 2.6175, + "step": 6750 + }, + { + "epoch": 0.29, + "grad_norm": 0.388671875, + "learning_rate": 0.00019021174040081938, + "loss": 2.6029, + "step": 6751 + }, + { + "epoch": 0.29, + "grad_norm": 0.421875, + "learning_rate": 0.00019020884578805993, + "loss": 2.5722, + "step": 6752 + }, + { + "epoch": 0.29, + "grad_norm": 0.404296875, + "learning_rate": 0.00019020595076939362, + "loss": 2.6655, + "step": 6753 + }, + { + "epoch": 0.29, + "grad_norm": 0.41796875, + "learning_rate": 0.00019020305534483354, + "loss": 2.5616, + "step": 6754 + }, + { + "epoch": 0.29, + "grad_norm": 0.5, + "learning_rate": 0.00019020015951439265, + "loss": 2.7363, + "step": 6755 + }, + { + "epoch": 0.29, + "grad_norm": 0.451171875, + "learning_rate": 0.000190197263278084, + "loss": 2.7809, + "step": 6756 + }, + { + "epoch": 0.29, + "grad_norm": 0.400390625, + "learning_rate": 0.00019019436663592065, + "loss": 2.598, + "step": 6757 + }, + { + "epoch": 0.29, + "grad_norm": 0.46875, + "learning_rate": 0.00019019146958791563, + "loss": 2.6419, + "step": 6758 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019018857213408193, + "loss": 2.61, + "step": 6759 + }, + { + "epoch": 0.29, + "grad_norm": 0.43359375, + "learning_rate": 0.00019018567427443262, + "loss": 2.6191, + "step": 6760 + }, + { + "epoch": 0.29, + "grad_norm": 0.4765625, + "learning_rate": 0.0001901827760089807, + "loss": 2.6162, + "step": 6761 + }, + { + "epoch": 0.29, + "grad_norm": 0.453125, + "learning_rate": 0.00019017987733773928, + "loss": 2.5223, + "step": 6762 + }, + { + "epoch": 0.29, + "grad_norm": 0.431640625, + "learning_rate": 0.00019017697826072136, + "loss": 2.5037, + "step": 6763 + }, + { + "epoch": 0.29, + "grad_norm": 0.4375, + "learning_rate": 0.00019017407877794002, + "loss": 2.62, + "step": 6764 + }, + { + "epoch": 0.29, + "grad_norm": 0.447265625, + "learning_rate": 0.00019017117888940824, + "loss": 2.6025, + "step": 6765 + }, + { + "epoch": 0.29, + "grad_norm": 0.462890625, + "learning_rate": 0.00019016827859513915, + "loss": 2.6497, + "step": 6766 + }, + { + "epoch": 0.29, + "grad_norm": 0.41015625, + "learning_rate": 0.00019016537789514574, + "loss": 2.5929, + "step": 6767 + }, + { + "epoch": 0.29, + "grad_norm": 0.5, + "learning_rate": 0.00019016247678944109, + "loss": 2.724, + "step": 6768 + }, + { + "epoch": 0.29, + "grad_norm": 0.4140625, + "learning_rate": 0.0001901595752780382, + "loss": 2.4286, + "step": 6769 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00019015667336095025, + "loss": 2.6981, + "step": 6770 + }, + { + "epoch": 0.29, + "grad_norm": 0.43359375, + "learning_rate": 0.00019015377103819017, + "loss": 2.5985, + "step": 6771 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00019015086830977108, + "loss": 2.5755, + "step": 6772 + }, + { + "epoch": 0.29, + "grad_norm": 0.392578125, + "learning_rate": 0.00019014796517570598, + "loss": 2.6336, + "step": 6773 + }, + { + "epoch": 0.29, + "grad_norm": 0.435546875, + "learning_rate": 0.00019014506163600805, + "loss": 2.5443, + "step": 6774 + }, + { + "epoch": 0.29, + "grad_norm": 0.46875, + "learning_rate": 0.00019014215769069023, + "loss": 2.8236, + "step": 6775 + }, + { + "epoch": 0.29, + "grad_norm": 0.40234375, + "learning_rate": 0.00019013925333976567, + "loss": 2.4919, + "step": 6776 + }, + { + "epoch": 0.29, + "grad_norm": 0.404296875, + "learning_rate": 0.00019013634858324743, + "loss": 2.6902, + "step": 6777 + }, + { + "epoch": 0.29, + "grad_norm": 0.4375, + "learning_rate": 0.00019013344342114856, + "loss": 2.5938, + "step": 6778 + }, + { + "epoch": 0.29, + "grad_norm": 0.49609375, + "learning_rate": 0.00019013053785348213, + "loss": 2.6354, + "step": 6779 + }, + { + "epoch": 0.29, + "grad_norm": 0.50390625, + "learning_rate": 0.00019012763188026123, + "loss": 2.8255, + "step": 6780 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00019012472550149887, + "loss": 2.8229, + "step": 6781 + }, + { + "epoch": 0.29, + "grad_norm": 0.44140625, + "learning_rate": 0.00019012181871720822, + "loss": 2.6546, + "step": 6782 + }, + { + "epoch": 0.29, + "grad_norm": 0.453125, + "learning_rate": 0.00019011891152740233, + "loss": 2.5441, + "step": 6783 + }, + { + "epoch": 0.29, + "grad_norm": 0.5, + "learning_rate": 0.00019011600393209425, + "loss": 2.5788, + "step": 6784 + }, + { + "epoch": 0.29, + "grad_norm": 0.474609375, + "learning_rate": 0.00019011309593129711, + "loss": 2.7258, + "step": 6785 + }, + { + "epoch": 0.29, + "grad_norm": 0.435546875, + "learning_rate": 0.00019011018752502397, + "loss": 2.5856, + "step": 6786 + }, + { + "epoch": 0.29, + "grad_norm": 0.470703125, + "learning_rate": 0.0001901072787132879, + "loss": 2.7439, + "step": 6787 + }, + { + "epoch": 0.29, + "grad_norm": 0.474609375, + "learning_rate": 0.000190104369496102, + "loss": 2.4982, + "step": 6788 + }, + { + "epoch": 0.29, + "grad_norm": 0.453125, + "learning_rate": 0.0001901014598734794, + "loss": 2.7329, + "step": 6789 + }, + { + "epoch": 0.29, + "grad_norm": 0.486328125, + "learning_rate": 0.0001900985498454331, + "loss": 2.5183, + "step": 6790 + }, + { + "epoch": 0.29, + "grad_norm": 0.421875, + "learning_rate": 0.0001900956394119763, + "loss": 2.6229, + "step": 6791 + }, + { + "epoch": 0.29, + "grad_norm": 0.478515625, + "learning_rate": 0.00019009272857312205, + "loss": 2.7285, + "step": 6792 + }, + { + "epoch": 0.29, + "grad_norm": 0.4375, + "learning_rate": 0.00019008981732888342, + "loss": 2.624, + "step": 6793 + }, + { + "epoch": 0.29, + "grad_norm": 0.44140625, + "learning_rate": 0.00019008690567927356, + "loss": 2.4679, + "step": 6794 + }, + { + "epoch": 0.29, + "grad_norm": 0.50390625, + "learning_rate": 0.00019008399362430554, + "loss": 2.7478, + "step": 6795 + }, + { + "epoch": 0.29, + "grad_norm": 0.447265625, + "learning_rate": 0.00019008108116399248, + "loss": 2.641, + "step": 6796 + }, + { + "epoch": 0.29, + "grad_norm": 0.443359375, + "learning_rate": 0.00019007816829834747, + "loss": 2.5603, + "step": 6797 + }, + { + "epoch": 0.29, + "grad_norm": 0.4609375, + "learning_rate": 0.00019007525502738362, + "loss": 2.5319, + "step": 6798 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00019007234135111402, + "loss": 2.621, + "step": 6799 + }, + { + "epoch": 0.29, + "grad_norm": 0.380859375, + "learning_rate": 0.00019006942726955184, + "loss": 2.4981, + "step": 6800 + }, + { + "epoch": 0.29, + "grad_norm": 0.46875, + "learning_rate": 0.00019006651278271014, + "loss": 2.7171, + "step": 6801 + }, + { + "epoch": 0.29, + "grad_norm": 0.470703125, + "learning_rate": 0.00019006359789060202, + "loss": 2.5744, + "step": 6802 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00019006068259324064, + "loss": 2.5908, + "step": 6803 + }, + { + "epoch": 0.29, + "grad_norm": 0.37109375, + "learning_rate": 0.00019005776689063914, + "loss": 2.5178, + "step": 6804 + }, + { + "epoch": 0.29, + "grad_norm": 0.4375, + "learning_rate": 0.00019005485078281057, + "loss": 2.8337, + "step": 6805 + }, + { + "epoch": 0.29, + "grad_norm": 0.416015625, + "learning_rate": 0.00019005193426976805, + "loss": 2.6718, + "step": 6806 + }, + { + "epoch": 0.29, + "grad_norm": 0.462890625, + "learning_rate": 0.0001900490173515248, + "loss": 2.7218, + "step": 6807 + }, + { + "epoch": 0.29, + "grad_norm": 0.50390625, + "learning_rate": 0.00019004610002809386, + "loss": 2.7789, + "step": 6808 + }, + { + "epoch": 0.29, + "grad_norm": 0.4609375, + "learning_rate": 0.00019004318229948837, + "loss": 2.7602, + "step": 6809 + }, + { + "epoch": 0.29, + "grad_norm": 0.404296875, + "learning_rate": 0.00019004026416572146, + "loss": 2.5051, + "step": 6810 + }, + { + "epoch": 0.29, + "grad_norm": 0.4609375, + "learning_rate": 0.00019003734562680625, + "loss": 2.584, + "step": 6811 + }, + { + "epoch": 0.29, + "grad_norm": 0.4921875, + "learning_rate": 0.00019003442668275592, + "loss": 2.7479, + "step": 6812 + }, + { + "epoch": 0.29, + "grad_norm": 0.43359375, + "learning_rate": 0.00019003150733358357, + "loss": 2.6146, + "step": 6813 + }, + { + "epoch": 0.29, + "grad_norm": 0.4375, + "learning_rate": 0.00019002858757930234, + "loss": 2.6406, + "step": 6814 + }, + { + "epoch": 0.29, + "grad_norm": 0.44921875, + "learning_rate": 0.00019002566741992537, + "loss": 2.4818, + "step": 6815 + }, + { + "epoch": 0.29, + "grad_norm": 0.462890625, + "learning_rate": 0.00019002274685546577, + "loss": 2.6756, + "step": 6816 + }, + { + "epoch": 0.29, + "grad_norm": 0.435546875, + "learning_rate": 0.00019001982588593675, + "loss": 2.6666, + "step": 6817 + }, + { + "epoch": 0.29, + "grad_norm": 0.392578125, + "learning_rate": 0.00019001690451135136, + "loss": 2.4887, + "step": 6818 + }, + { + "epoch": 0.29, + "grad_norm": 0.41796875, + "learning_rate": 0.00019001398273172285, + "loss": 2.4903, + "step": 6819 + }, + { + "epoch": 0.29, + "grad_norm": 0.4140625, + "learning_rate": 0.0001900110605470643, + "loss": 2.6391, + "step": 6820 + }, + { + "epoch": 0.29, + "grad_norm": 0.4765625, + "learning_rate": 0.00019000813795738887, + "loss": 2.6195, + "step": 6821 + }, + { + "epoch": 0.29, + "grad_norm": 0.486328125, + "learning_rate": 0.0001900052149627097, + "loss": 2.7548, + "step": 6822 + }, + { + "epoch": 0.29, + "grad_norm": 0.470703125, + "learning_rate": 0.00019000229156303995, + "loss": 2.692, + "step": 6823 + }, + { + "epoch": 0.29, + "grad_norm": 0.466796875, + "learning_rate": 0.0001899993677583928, + "loss": 2.6217, + "step": 6824 + }, + { + "epoch": 0.29, + "grad_norm": 0.48828125, + "learning_rate": 0.00018999644354878136, + "loss": 2.6686, + "step": 6825 + }, + { + "epoch": 0.29, + "grad_norm": 0.427734375, + "learning_rate": 0.00018999351893421885, + "loss": 2.6901, + "step": 6826 + }, + { + "epoch": 0.29, + "grad_norm": 0.482421875, + "learning_rate": 0.00018999059391471837, + "loss": 2.8176, + "step": 6827 + }, + { + "epoch": 0.29, + "grad_norm": 0.416015625, + "learning_rate": 0.00018998766849029312, + "loss": 2.688, + "step": 6828 + }, + { + "epoch": 0.29, + "grad_norm": 0.400390625, + "learning_rate": 0.00018998474266095623, + "loss": 2.6926, + "step": 6829 + }, + { + "epoch": 0.29, + "grad_norm": 0.474609375, + "learning_rate": 0.00018998181642672092, + "loss": 2.4697, + "step": 6830 + }, + { + "epoch": 0.29, + "grad_norm": 0.474609375, + "learning_rate": 0.00018997888978760028, + "loss": 2.7417, + "step": 6831 + }, + { + "epoch": 0.29, + "grad_norm": 0.455078125, + "learning_rate": 0.00018997596274360754, + "loss": 2.6802, + "step": 6832 + }, + { + "epoch": 0.29, + "grad_norm": 0.41796875, + "learning_rate": 0.00018997303529475583, + "loss": 2.6256, + "step": 6833 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00018997010744105838, + "loss": 2.7273, + "step": 6834 + }, + { + "epoch": 0.29, + "grad_norm": 0.41796875, + "learning_rate": 0.00018996717918252828, + "loss": 2.6562, + "step": 6835 + }, + { + "epoch": 0.29, + "grad_norm": 0.423828125, + "learning_rate": 0.00018996425051917882, + "loss": 2.5163, + "step": 6836 + }, + { + "epoch": 0.29, + "grad_norm": 0.458984375, + "learning_rate": 0.00018996132145102305, + "loss": 2.5842, + "step": 6837 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.0001899583919780742, + "loss": 2.63, + "step": 6838 + }, + { + "epoch": 0.29, + "grad_norm": 0.43359375, + "learning_rate": 0.00018995546210034552, + "loss": 2.6301, + "step": 6839 + }, + { + "epoch": 0.29, + "grad_norm": 0.400390625, + "learning_rate": 0.0001899525318178501, + "loss": 2.7683, + "step": 6840 + }, + { + "epoch": 0.29, + "grad_norm": 0.453125, + "learning_rate": 0.00018994960113060118, + "loss": 2.6398, + "step": 6841 + }, + { + "epoch": 0.29, + "grad_norm": 0.41796875, + "learning_rate": 0.00018994667003861192, + "loss": 2.476, + "step": 6842 + }, + { + "epoch": 0.29, + "grad_norm": 0.5078125, + "learning_rate": 0.00018994373854189548, + "loss": 2.6888, + "step": 6843 + }, + { + "epoch": 0.29, + "grad_norm": 0.43359375, + "learning_rate": 0.00018994080664046513, + "loss": 2.643, + "step": 6844 + }, + { + "epoch": 0.29, + "grad_norm": 0.462890625, + "learning_rate": 0.000189937874334334, + "loss": 2.658, + "step": 6845 + }, + { + "epoch": 0.29, + "grad_norm": 0.46484375, + "learning_rate": 0.00018993494162351533, + "loss": 2.6123, + "step": 6846 + }, + { + "epoch": 0.29, + "grad_norm": 0.4140625, + "learning_rate": 0.00018993200850802227, + "loss": 2.6773, + "step": 6847 + }, + { + "epoch": 0.29, + "grad_norm": 0.462890625, + "learning_rate": 0.00018992907498786802, + "loss": 2.7565, + "step": 6848 + }, + { + "epoch": 0.29, + "grad_norm": 0.419921875, + "learning_rate": 0.00018992614106306586, + "loss": 2.7342, + "step": 6849 + }, + { + "epoch": 0.29, + "grad_norm": 0.4765625, + "learning_rate": 0.00018992320673362888, + "loss": 2.7012, + "step": 6850 + }, + { + "epoch": 0.29, + "grad_norm": 0.419921875, + "learning_rate": 0.00018992027199957035, + "loss": 2.6573, + "step": 6851 + }, + { + "epoch": 0.29, + "grad_norm": 0.470703125, + "learning_rate": 0.00018991733686090343, + "loss": 2.6519, + "step": 6852 + }, + { + "epoch": 0.29, + "grad_norm": 0.451171875, + "learning_rate": 0.00018991440131764138, + "loss": 2.5921, + "step": 6853 + }, + { + "epoch": 0.29, + "grad_norm": 0.4140625, + "learning_rate": 0.00018991146536979735, + "loss": 2.6012, + "step": 6854 + }, + { + "epoch": 0.29, + "grad_norm": 0.46875, + "learning_rate": 0.00018990852901738464, + "loss": 2.8242, + "step": 6855 + }, + { + "epoch": 0.29, + "grad_norm": 0.421875, + "learning_rate": 0.0001899055922604164, + "loss": 2.6525, + "step": 6856 + }, + { + "epoch": 0.29, + "grad_norm": 0.52734375, + "learning_rate": 0.00018990265509890583, + "loss": 2.7375, + "step": 6857 + }, + { + "epoch": 0.29, + "grad_norm": 0.41796875, + "learning_rate": 0.00018989971753286615, + "loss": 2.5875, + "step": 6858 + }, + { + "epoch": 0.29, + "grad_norm": 0.451171875, + "learning_rate": 0.0001898967795623106, + "loss": 2.5476, + "step": 6859 + }, + { + "epoch": 0.29, + "grad_norm": 0.435546875, + "learning_rate": 0.00018989384118725243, + "loss": 2.5586, + "step": 6860 + }, + { + "epoch": 0.29, + "grad_norm": 0.453125, + "learning_rate": 0.0001898909024077048, + "loss": 2.7632, + "step": 6861 + }, + { + "epoch": 0.29, + "grad_norm": 0.484375, + "learning_rate": 0.00018988796322368096, + "loss": 2.8234, + "step": 6862 + }, + { + "epoch": 0.29, + "grad_norm": 0.51171875, + "learning_rate": 0.00018988502363519416, + "loss": 2.6693, + "step": 6863 + }, + { + "epoch": 0.29, + "grad_norm": 0.44140625, + "learning_rate": 0.00018988208364225759, + "loss": 2.6332, + "step": 6864 + }, + { + "epoch": 0.29, + "grad_norm": 0.3828125, + "learning_rate": 0.0001898791432448845, + "loss": 2.471, + "step": 6865 + }, + { + "epoch": 0.29, + "grad_norm": 0.48828125, + "learning_rate": 0.00018987620244308807, + "loss": 2.6156, + "step": 6866 + }, + { + "epoch": 0.29, + "grad_norm": 0.431640625, + "learning_rate": 0.00018987326123688162, + "loss": 2.6878, + "step": 6867 + }, + { + "epoch": 0.29, + "grad_norm": 0.38671875, + "learning_rate": 0.00018987031962627832, + "loss": 2.6859, + "step": 6868 + }, + { + "epoch": 0.29, + "grad_norm": 0.51171875, + "learning_rate": 0.00018986737761129144, + "loss": 2.6581, + "step": 6869 + }, + { + "epoch": 0.29, + "grad_norm": 0.458984375, + "learning_rate": 0.00018986443519193418, + "loss": 2.6192, + "step": 6870 + }, + { + "epoch": 0.29, + "grad_norm": 0.50390625, + "learning_rate": 0.0001898614923682198, + "loss": 2.4922, + "step": 6871 + }, + { + "epoch": 0.29, + "grad_norm": 0.427734375, + "learning_rate": 0.0001898585491401616, + "loss": 2.5578, + "step": 6872 + }, + { + "epoch": 0.29, + "grad_norm": 0.361328125, + "learning_rate": 0.00018985560550777272, + "loss": 2.5464, + "step": 6873 + }, + { + "epoch": 0.29, + "grad_norm": 0.458984375, + "learning_rate": 0.00018985266147106646, + "loss": 2.6422, + "step": 6874 + }, + { + "epoch": 0.29, + "grad_norm": 0.453125, + "learning_rate": 0.00018984971703005606, + "loss": 2.7389, + "step": 6875 + }, + { + "epoch": 0.29, + "grad_norm": 0.435546875, + "learning_rate": 0.0001898467721847548, + "loss": 2.6534, + "step": 6876 + }, + { + "epoch": 0.29, + "grad_norm": 0.427734375, + "learning_rate": 0.00018984382693517586, + "loss": 2.5673, + "step": 6877 + }, + { + "epoch": 0.29, + "grad_norm": 0.466796875, + "learning_rate": 0.00018984088128133256, + "loss": 2.5956, + "step": 6878 + }, + { + "epoch": 0.29, + "grad_norm": 0.431640625, + "learning_rate": 0.00018983793522323812, + "loss": 2.6035, + "step": 6879 + }, + { + "epoch": 0.29, + "grad_norm": 0.458984375, + "learning_rate": 0.0001898349887609058, + "loss": 2.5254, + "step": 6880 + }, + { + "epoch": 0.29, + "grad_norm": 0.439453125, + "learning_rate": 0.00018983204189434885, + "loss": 2.621, + "step": 6881 + }, + { + "epoch": 0.29, + "grad_norm": 0.46484375, + "learning_rate": 0.00018982909462358057, + "loss": 2.6761, + "step": 6882 + }, + { + "epoch": 0.29, + "grad_norm": 0.4140625, + "learning_rate": 0.00018982614694861415, + "loss": 2.5146, + "step": 6883 + }, + { + "epoch": 0.29, + "grad_norm": 0.37890625, + "learning_rate": 0.00018982319886946293, + "loss": 2.5584, + "step": 6884 + }, + { + "epoch": 0.29, + "grad_norm": 0.439453125, + "learning_rate": 0.00018982025038614015, + "loss": 2.5816, + "step": 6885 + }, + { + "epoch": 0.29, + "grad_norm": 0.484375, + "learning_rate": 0.00018981730149865905, + "loss": 2.6352, + "step": 6886 + }, + { + "epoch": 0.29, + "grad_norm": 0.419921875, + "learning_rate": 0.0001898143522070329, + "loss": 2.5654, + "step": 6887 + }, + { + "epoch": 0.29, + "grad_norm": 0.4296875, + "learning_rate": 0.00018981140251127503, + "loss": 2.636, + "step": 6888 + }, + { + "epoch": 0.29, + "grad_norm": 0.515625, + "learning_rate": 0.00018980845241139865, + "loss": 2.7909, + "step": 6889 + }, + { + "epoch": 0.29, + "grad_norm": 0.443359375, + "learning_rate": 0.00018980550190741705, + "loss": 2.5026, + "step": 6890 + }, + { + "epoch": 0.29, + "grad_norm": 0.39453125, + "learning_rate": 0.00018980255099934352, + "loss": 2.5418, + "step": 6891 + }, + { + "epoch": 0.29, + "grad_norm": 0.404296875, + "learning_rate": 0.00018979959968719133, + "loss": 2.6335, + "step": 6892 + }, + { + "epoch": 0.29, + "grad_norm": 0.462890625, + "learning_rate": 0.00018979664797097375, + "loss": 2.7386, + "step": 6893 + }, + { + "epoch": 0.29, + "grad_norm": 0.435546875, + "learning_rate": 0.00018979369585070405, + "loss": 2.6794, + "step": 6894 + }, + { + "epoch": 0.29, + "grad_norm": 0.43359375, + "learning_rate": 0.0001897907433263956, + "loss": 2.5864, + "step": 6895 + }, + { + "epoch": 0.29, + "grad_norm": 0.44921875, + "learning_rate": 0.00018978779039806155, + "loss": 2.4932, + "step": 6896 + }, + { + "epoch": 0.29, + "grad_norm": 0.466796875, + "learning_rate": 0.0001897848370657153, + "loss": 2.6093, + "step": 6897 + }, + { + "epoch": 0.29, + "grad_norm": 0.46484375, + "learning_rate": 0.00018978188332937008, + "loss": 2.7115, + "step": 6898 + }, + { + "epoch": 0.29, + "grad_norm": 0.486328125, + "learning_rate": 0.0001897789291890392, + "loss": 2.6542, + "step": 6899 + }, + { + "epoch": 0.29, + "grad_norm": 0.44921875, + "learning_rate": 0.00018977597464473595, + "loss": 2.6191, + "step": 6900 + }, + { + "epoch": 0.29, + "grad_norm": 0.494140625, + "learning_rate": 0.00018977301969647363, + "loss": 2.7044, + "step": 6901 + }, + { + "epoch": 0.29, + "grad_norm": 0.466796875, + "learning_rate": 0.00018977006434426553, + "loss": 2.7289, + "step": 6902 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00018976710858812493, + "loss": 2.7194, + "step": 6903 + }, + { + "epoch": 0.29, + "grad_norm": 0.51171875, + "learning_rate": 0.00018976415242806516, + "loss": 2.6087, + "step": 6904 + }, + { + "epoch": 0.29, + "grad_norm": 0.44921875, + "learning_rate": 0.0001897611958640995, + "loss": 2.5181, + "step": 6905 + }, + { + "epoch": 0.29, + "grad_norm": 0.4453125, + "learning_rate": 0.00018975823889624127, + "loss": 2.6465, + "step": 6906 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00018975528152450377, + "loss": 2.5532, + "step": 6907 + }, + { + "epoch": 0.29, + "grad_norm": 0.4609375, + "learning_rate": 0.0001897523237489003, + "loss": 2.6116, + "step": 6908 + }, + { + "epoch": 0.29, + "grad_norm": 0.419921875, + "learning_rate": 0.00018974936556944417, + "loss": 2.5154, + "step": 6909 + }, + { + "epoch": 0.29, + "grad_norm": 0.416015625, + "learning_rate": 0.00018974640698614872, + "loss": 2.572, + "step": 6910 + }, + { + "epoch": 0.29, + "grad_norm": 0.455078125, + "learning_rate": 0.0001897434479990272, + "loss": 2.7254, + "step": 6911 + }, + { + "epoch": 0.29, + "grad_norm": 0.455078125, + "learning_rate": 0.000189740488608093, + "loss": 2.5727, + "step": 6912 + }, + { + "epoch": 0.29, + "grad_norm": 0.4140625, + "learning_rate": 0.00018973752881335938, + "loss": 2.6904, + "step": 6913 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00018973456861483964, + "loss": 2.7145, + "step": 6914 + }, + { + "epoch": 0.29, + "grad_norm": 0.45703125, + "learning_rate": 0.00018973160801254716, + "loss": 2.606, + "step": 6915 + }, + { + "epoch": 0.29, + "grad_norm": 0.48046875, + "learning_rate": 0.00018972864700649527, + "loss": 2.7415, + "step": 6916 + }, + { + "epoch": 0.29, + "grad_norm": 0.443359375, + "learning_rate": 0.00018972568559669724, + "loss": 2.668, + "step": 6917 + }, + { + "epoch": 0.29, + "grad_norm": 0.458984375, + "learning_rate": 0.00018972272378316643, + "loss": 2.5135, + "step": 6918 + }, + { + "epoch": 0.29, + "grad_norm": 0.443359375, + "learning_rate": 0.0001897197615659161, + "loss": 2.5984, + "step": 6919 + }, + { + "epoch": 0.29, + "grad_norm": 0.4609375, + "learning_rate": 0.00018971679894495965, + "loss": 2.8097, + "step": 6920 + }, + { + "epoch": 0.29, + "grad_norm": 0.484375, + "learning_rate": 0.00018971383592031042, + "loss": 2.5962, + "step": 6921 + }, + { + "epoch": 0.29, + "grad_norm": 0.455078125, + "learning_rate": 0.0001897108724919817, + "loss": 2.6985, + "step": 6922 + }, + { + "epoch": 0.3, + "grad_norm": 0.490234375, + "learning_rate": 0.00018970790865998686, + "loss": 2.6704, + "step": 6923 + }, + { + "epoch": 0.3, + "grad_norm": 0.498046875, + "learning_rate": 0.0001897049444243392, + "loss": 2.606, + "step": 6924 + }, + { + "epoch": 0.3, + "grad_norm": 0.447265625, + "learning_rate": 0.00018970197978505207, + "loss": 2.7139, + "step": 6925 + }, + { + "epoch": 0.3, + "grad_norm": 0.443359375, + "learning_rate": 0.0001896990147421388, + "loss": 2.6083, + "step": 6926 + }, + { + "epoch": 0.3, + "grad_norm": 0.3984375, + "learning_rate": 0.00018969604929561278, + "loss": 2.6899, + "step": 6927 + }, + { + "epoch": 0.3, + "grad_norm": 0.498046875, + "learning_rate": 0.0001896930834454873, + "loss": 2.6643, + "step": 6928 + }, + { + "epoch": 0.3, + "grad_norm": 0.451171875, + "learning_rate": 0.00018969011719177571, + "loss": 2.7348, + "step": 6929 + }, + { + "epoch": 0.3, + "grad_norm": 0.435546875, + "learning_rate": 0.0001896871505344914, + "loss": 2.6847, + "step": 6930 + }, + { + "epoch": 0.3, + "grad_norm": 0.431640625, + "learning_rate": 0.00018968418347364767, + "loss": 2.5689, + "step": 6931 + }, + { + "epoch": 0.3, + "grad_norm": 0.478515625, + "learning_rate": 0.0001896812160092579, + "loss": 2.7169, + "step": 6932 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.00018967824814133542, + "loss": 2.5952, + "step": 6933 + }, + { + "epoch": 0.3, + "grad_norm": 0.373046875, + "learning_rate": 0.0001896752798698936, + "loss": 2.5671, + "step": 6934 + }, + { + "epoch": 0.3, + "grad_norm": 0.466796875, + "learning_rate": 0.00018967231119494578, + "loss": 2.5973, + "step": 6935 + }, + { + "epoch": 0.3, + "grad_norm": 0.48046875, + "learning_rate": 0.00018966934211650535, + "loss": 2.5164, + "step": 6936 + }, + { + "epoch": 0.3, + "grad_norm": 0.51953125, + "learning_rate": 0.00018966637263458566, + "loss": 2.6288, + "step": 6937 + }, + { + "epoch": 0.3, + "grad_norm": 0.470703125, + "learning_rate": 0.00018966340274920006, + "loss": 2.7187, + "step": 6938 + }, + { + "epoch": 0.3, + "grad_norm": 0.4296875, + "learning_rate": 0.00018966043246036187, + "loss": 2.5932, + "step": 6939 + }, + { + "epoch": 0.3, + "grad_norm": 0.48828125, + "learning_rate": 0.00018965746176808455, + "loss": 2.6484, + "step": 6940 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.0001896544906723814, + "loss": 2.5865, + "step": 6941 + }, + { + "epoch": 0.3, + "grad_norm": 0.4140625, + "learning_rate": 0.0001896515191732658, + "loss": 2.5766, + "step": 6942 + }, + { + "epoch": 0.3, + "grad_norm": 0.419921875, + "learning_rate": 0.00018964854727075118, + "loss": 2.6239, + "step": 6943 + }, + { + "epoch": 0.3, + "grad_norm": 0.455078125, + "learning_rate": 0.0001896455749648508, + "loss": 2.5922, + "step": 6944 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.0001896426022555781, + "loss": 2.7953, + "step": 6945 + }, + { + "epoch": 0.3, + "grad_norm": 0.400390625, + "learning_rate": 0.00018963962914294647, + "loss": 2.6356, + "step": 6946 + }, + { + "epoch": 0.3, + "grad_norm": 0.380859375, + "learning_rate": 0.00018963665562696927, + "loss": 2.6202, + "step": 6947 + }, + { + "epoch": 0.3, + "grad_norm": 0.443359375, + "learning_rate": 0.00018963368170765986, + "loss": 2.6984, + "step": 6948 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.00018963070738503167, + "loss": 2.7431, + "step": 6949 + }, + { + "epoch": 0.3, + "grad_norm": 0.4765625, + "learning_rate": 0.00018962773265909801, + "loss": 2.5598, + "step": 6950 + }, + { + "epoch": 0.3, + "grad_norm": 0.427734375, + "learning_rate": 0.00018962475752987235, + "loss": 2.6711, + "step": 6951 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.00018962178199736802, + "loss": 2.5861, + "step": 6952 + }, + { + "epoch": 0.3, + "grad_norm": 0.4140625, + "learning_rate": 0.0001896188060615984, + "loss": 2.5768, + "step": 6953 + }, + { + "epoch": 0.3, + "grad_norm": 0.484375, + "learning_rate": 0.00018961582972257693, + "loss": 2.6448, + "step": 6954 + }, + { + "epoch": 0.3, + "grad_norm": 0.470703125, + "learning_rate": 0.000189612852980317, + "loss": 2.5977, + "step": 6955 + }, + { + "epoch": 0.3, + "grad_norm": 0.392578125, + "learning_rate": 0.00018960987583483194, + "loss": 2.5983, + "step": 6956 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.00018960689828613521, + "loss": 2.4649, + "step": 6957 + }, + { + "epoch": 0.3, + "grad_norm": 0.458984375, + "learning_rate": 0.00018960392033424018, + "loss": 2.6772, + "step": 6958 + }, + { + "epoch": 0.3, + "grad_norm": 0.466796875, + "learning_rate": 0.00018960094197916023, + "loss": 2.742, + "step": 6959 + }, + { + "epoch": 0.3, + "grad_norm": 0.478515625, + "learning_rate": 0.0001895979632209088, + "loss": 2.8179, + "step": 6960 + }, + { + "epoch": 0.3, + "grad_norm": 0.416015625, + "learning_rate": 0.00018959498405949932, + "loss": 2.7145, + "step": 6961 + }, + { + "epoch": 0.3, + "grad_norm": 0.52734375, + "learning_rate": 0.00018959200449494512, + "loss": 2.6056, + "step": 6962 + }, + { + "epoch": 0.3, + "grad_norm": 0.3984375, + "learning_rate": 0.0001895890245272596, + "loss": 2.5113, + "step": 6963 + }, + { + "epoch": 0.3, + "grad_norm": 0.41796875, + "learning_rate": 0.0001895860441564563, + "loss": 2.5404, + "step": 6964 + }, + { + "epoch": 0.3, + "grad_norm": 0.46484375, + "learning_rate": 0.00018958306338254848, + "loss": 2.6752, + "step": 6965 + }, + { + "epoch": 0.3, + "grad_norm": 0.427734375, + "learning_rate": 0.00018958008220554964, + "loss": 2.7034, + "step": 6966 + }, + { + "epoch": 0.3, + "grad_norm": 0.4765625, + "learning_rate": 0.00018957710062547314, + "loss": 2.5729, + "step": 6967 + }, + { + "epoch": 0.3, + "grad_norm": 0.490234375, + "learning_rate": 0.00018957411864233245, + "loss": 2.7084, + "step": 6968 + }, + { + "epoch": 0.3, + "grad_norm": 0.490234375, + "learning_rate": 0.00018957113625614095, + "loss": 2.6687, + "step": 6969 + }, + { + "epoch": 0.3, + "grad_norm": 0.451171875, + "learning_rate": 0.00018956815346691207, + "loss": 2.555, + "step": 6970 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.00018956517027465925, + "loss": 2.7156, + "step": 6971 + }, + { + "epoch": 0.3, + "grad_norm": 0.408203125, + "learning_rate": 0.00018956218667939587, + "loss": 2.5949, + "step": 6972 + }, + { + "epoch": 0.3, + "grad_norm": 0.458984375, + "learning_rate": 0.00018955920268113542, + "loss": 2.5559, + "step": 6973 + }, + { + "epoch": 0.3, + "grad_norm": 0.498046875, + "learning_rate": 0.00018955621827989125, + "loss": 2.6185, + "step": 6974 + }, + { + "epoch": 0.3, + "grad_norm": 0.4765625, + "learning_rate": 0.00018955323347567683, + "loss": 2.698, + "step": 6975 + }, + { + "epoch": 0.3, + "grad_norm": 0.4296875, + "learning_rate": 0.0001895502482685056, + "loss": 2.6046, + "step": 6976 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.000189547262658391, + "loss": 2.5916, + "step": 6977 + }, + { + "epoch": 0.3, + "grad_norm": 0.443359375, + "learning_rate": 0.0001895442766453464, + "loss": 2.6153, + "step": 6978 + }, + { + "epoch": 0.3, + "grad_norm": 0.49609375, + "learning_rate": 0.0001895412902293853, + "loss": 2.7673, + "step": 6979 + }, + { + "epoch": 0.3, + "grad_norm": 0.431640625, + "learning_rate": 0.00018953830341052113, + "loss": 2.7178, + "step": 6980 + }, + { + "epoch": 0.3, + "grad_norm": 0.4296875, + "learning_rate": 0.00018953531618876731, + "loss": 2.627, + "step": 6981 + }, + { + "epoch": 0.3, + "grad_norm": 0.453125, + "learning_rate": 0.00018953232856413728, + "loss": 2.6621, + "step": 6982 + }, + { + "epoch": 0.3, + "grad_norm": 0.4140625, + "learning_rate": 0.0001895293405366445, + "loss": 2.6482, + "step": 6983 + }, + { + "epoch": 0.3, + "grad_norm": 0.45703125, + "learning_rate": 0.0001895263521063024, + "loss": 2.6166, + "step": 6984 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.00018952336327312443, + "loss": 2.6242, + "step": 6985 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.00018952037403712404, + "loss": 2.6092, + "step": 6986 + }, + { + "epoch": 0.3, + "grad_norm": 0.48046875, + "learning_rate": 0.00018951738439831471, + "loss": 2.7392, + "step": 6987 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.00018951439435670983, + "loss": 2.7801, + "step": 6988 + }, + { + "epoch": 0.3, + "grad_norm": 0.427734375, + "learning_rate": 0.00018951140391232292, + "loss": 2.585, + "step": 6989 + }, + { + "epoch": 0.3, + "grad_norm": 0.48046875, + "learning_rate": 0.00018950841306516737, + "loss": 2.7151, + "step": 6990 + }, + { + "epoch": 0.3, + "grad_norm": 0.48828125, + "learning_rate": 0.0001895054218152567, + "loss": 2.6304, + "step": 6991 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.00018950243016260435, + "loss": 2.6601, + "step": 6992 + }, + { + "epoch": 0.3, + "grad_norm": 0.427734375, + "learning_rate": 0.00018949943810722374, + "loss": 2.6883, + "step": 6993 + }, + { + "epoch": 0.3, + "grad_norm": 0.4921875, + "learning_rate": 0.00018949644564912838, + "loss": 2.6612, + "step": 6994 + }, + { + "epoch": 0.3, + "grad_norm": 0.427734375, + "learning_rate": 0.00018949345278833175, + "loss": 2.5534, + "step": 6995 + }, + { + "epoch": 0.3, + "grad_norm": 0.443359375, + "learning_rate": 0.00018949045952484727, + "loss": 2.6974, + "step": 6996 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.00018948746585868841, + "loss": 2.4401, + "step": 6997 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.0001894844717898687, + "loss": 2.6337, + "step": 6998 + }, + { + "epoch": 0.3, + "grad_norm": 0.486328125, + "learning_rate": 0.00018948147731840154, + "loss": 2.6007, + "step": 6999 + }, + { + "epoch": 0.3, + "grad_norm": 0.470703125, + "learning_rate": 0.00018947848244430042, + "loss": 2.5964, + "step": 7000 + }, + { + "epoch": 0.3, + "grad_norm": 0.431640625, + "learning_rate": 0.00018947548716757883, + "loss": 2.6321, + "step": 7001 + }, + { + "epoch": 0.3, + "grad_norm": 0.37890625, + "learning_rate": 0.0001894724914882503, + "loss": 2.6771, + "step": 7002 + }, + { + "epoch": 0.3, + "grad_norm": 0.4609375, + "learning_rate": 0.00018946949540632818, + "loss": 2.6534, + "step": 7003 + }, + { + "epoch": 0.3, + "grad_norm": 0.4453125, + "learning_rate": 0.00018946649892182609, + "loss": 2.7232, + "step": 7004 + }, + { + "epoch": 0.3, + "grad_norm": 0.46875, + "learning_rate": 0.0001894635020347574, + "loss": 2.6894, + "step": 7005 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.0001894605047451357, + "loss": 2.6839, + "step": 7006 + }, + { + "epoch": 0.3, + "grad_norm": 0.515625, + "learning_rate": 0.00018945750705297437, + "loss": 2.6166, + "step": 7007 + }, + { + "epoch": 0.3, + "grad_norm": 0.5234375, + "learning_rate": 0.000189454508958287, + "loss": 2.6297, + "step": 7008 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.00018945151046108701, + "loss": 2.6022, + "step": 7009 + }, + { + "epoch": 0.3, + "grad_norm": 0.4296875, + "learning_rate": 0.0001894485115613879, + "loss": 2.5766, + "step": 7010 + }, + { + "epoch": 0.3, + "grad_norm": 0.4765625, + "learning_rate": 0.00018944551225920318, + "loss": 2.6476, + "step": 7011 + }, + { + "epoch": 0.3, + "grad_norm": 0.46484375, + "learning_rate": 0.00018944251255454637, + "loss": 2.7016, + "step": 7012 + }, + { + "epoch": 0.3, + "grad_norm": 0.490234375, + "learning_rate": 0.0001894395124474309, + "loss": 2.6013, + "step": 7013 + }, + { + "epoch": 0.3, + "grad_norm": 0.482421875, + "learning_rate": 0.00018943651193787034, + "loss": 2.6762, + "step": 7014 + }, + { + "epoch": 0.3, + "grad_norm": 0.447265625, + "learning_rate": 0.00018943351102587816, + "loss": 2.6874, + "step": 7015 + }, + { + "epoch": 0.3, + "grad_norm": 0.447265625, + "learning_rate": 0.00018943050971146786, + "loss": 2.4906, + "step": 7016 + }, + { + "epoch": 0.3, + "grad_norm": 0.455078125, + "learning_rate": 0.00018942750799465298, + "loss": 2.6181, + "step": 7017 + }, + { + "epoch": 0.3, + "grad_norm": 0.45703125, + "learning_rate": 0.00018942450587544697, + "loss": 2.712, + "step": 7018 + }, + { + "epoch": 0.3, + "grad_norm": 0.431640625, + "learning_rate": 0.00018942150335386334, + "loss": 2.581, + "step": 7019 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.00018941850042991564, + "loss": 2.6335, + "step": 7020 + }, + { + "epoch": 0.3, + "grad_norm": 0.484375, + "learning_rate": 0.0001894154971036174, + "loss": 2.7112, + "step": 7021 + }, + { + "epoch": 0.3, + "grad_norm": 0.453125, + "learning_rate": 0.00018941249337498206, + "loss": 2.6576, + "step": 7022 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.0001894094892440232, + "loss": 2.8136, + "step": 7023 + }, + { + "epoch": 0.3, + "grad_norm": 0.486328125, + "learning_rate": 0.00018940648471075432, + "loss": 2.7405, + "step": 7024 + }, + { + "epoch": 0.3, + "grad_norm": 0.470703125, + "learning_rate": 0.0001894034797751889, + "loss": 2.6203, + "step": 7025 + }, + { + "epoch": 0.3, + "grad_norm": 0.4453125, + "learning_rate": 0.0001894004744373405, + "loss": 2.705, + "step": 7026 + }, + { + "epoch": 0.3, + "grad_norm": 0.50390625, + "learning_rate": 0.00018939746869722264, + "loss": 2.6402, + "step": 7027 + }, + { + "epoch": 0.3, + "grad_norm": 0.48828125, + "learning_rate": 0.00018939446255484884, + "loss": 2.6821, + "step": 7028 + }, + { + "epoch": 0.3, + "grad_norm": 0.451171875, + "learning_rate": 0.00018939145601023266, + "loss": 2.7838, + "step": 7029 + }, + { + "epoch": 0.3, + "grad_norm": 0.4609375, + "learning_rate": 0.00018938844906338757, + "loss": 2.6663, + "step": 7030 + }, + { + "epoch": 0.3, + "grad_norm": 0.478515625, + "learning_rate": 0.00018938544171432712, + "loss": 2.5784, + "step": 7031 + }, + { + "epoch": 0.3, + "grad_norm": 0.447265625, + "learning_rate": 0.00018938243396306488, + "loss": 2.5613, + "step": 7032 + }, + { + "epoch": 0.3, + "grad_norm": 0.46484375, + "learning_rate": 0.00018937942580961434, + "loss": 2.7079, + "step": 7033 + }, + { + "epoch": 0.3, + "grad_norm": 0.43359375, + "learning_rate": 0.00018937641725398903, + "loss": 2.6499, + "step": 7034 + }, + { + "epoch": 0.3, + "grad_norm": 0.44921875, + "learning_rate": 0.00018937340829620252, + "loss": 2.5951, + "step": 7035 + }, + { + "epoch": 0.3, + "grad_norm": 0.41015625, + "learning_rate": 0.00018937039893626834, + "loss": 2.6314, + "step": 7036 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.0001893673891742, + "loss": 2.6109, + "step": 7037 + }, + { + "epoch": 0.3, + "grad_norm": 0.4453125, + "learning_rate": 0.00018936437901001112, + "loss": 2.7441, + "step": 7038 + }, + { + "epoch": 0.3, + "grad_norm": 0.455078125, + "learning_rate": 0.00018936136844371516, + "loss": 2.6834, + "step": 7039 + }, + { + "epoch": 0.3, + "grad_norm": 0.384765625, + "learning_rate": 0.0001893583574753257, + "loss": 2.6036, + "step": 7040 + }, + { + "epoch": 0.3, + "grad_norm": 0.43359375, + "learning_rate": 0.00018935534610485627, + "loss": 2.7, + "step": 7041 + }, + { + "epoch": 0.3, + "grad_norm": 0.443359375, + "learning_rate": 0.00018935233433232048, + "loss": 2.6493, + "step": 7042 + }, + { + "epoch": 0.3, + "grad_norm": 0.46875, + "learning_rate": 0.0001893493221577318, + "loss": 2.633, + "step": 7043 + }, + { + "epoch": 0.3, + "grad_norm": 0.46875, + "learning_rate": 0.00018934630958110385, + "loss": 2.7744, + "step": 7044 + }, + { + "epoch": 0.3, + "grad_norm": 0.404296875, + "learning_rate": 0.00018934329660245014, + "loss": 2.7244, + "step": 7045 + }, + { + "epoch": 0.3, + "grad_norm": 0.486328125, + "learning_rate": 0.00018934028322178427, + "loss": 2.5891, + "step": 7046 + }, + { + "epoch": 0.3, + "grad_norm": 0.453125, + "learning_rate": 0.00018933726943911973, + "loss": 2.6718, + "step": 7047 + }, + { + "epoch": 0.3, + "grad_norm": 0.423828125, + "learning_rate": 0.00018933425525447016, + "loss": 2.7683, + "step": 7048 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.00018933124066784911, + "loss": 2.4764, + "step": 7049 + }, + { + "epoch": 0.3, + "grad_norm": 0.4921875, + "learning_rate": 0.0001893282256792701, + "loss": 2.5218, + "step": 7050 + }, + { + "epoch": 0.3, + "grad_norm": 0.4609375, + "learning_rate": 0.00018932521028874674, + "loss": 2.788, + "step": 7051 + }, + { + "epoch": 0.3, + "grad_norm": 0.4296875, + "learning_rate": 0.00018932219449629254, + "loss": 2.5783, + "step": 7052 + }, + { + "epoch": 0.3, + "grad_norm": 0.37890625, + "learning_rate": 0.00018931917830192113, + "loss": 2.5087, + "step": 7053 + }, + { + "epoch": 0.3, + "grad_norm": 0.380859375, + "learning_rate": 0.00018931616170564608, + "loss": 2.5861, + "step": 7054 + }, + { + "epoch": 0.3, + "grad_norm": 0.50390625, + "learning_rate": 0.00018931314470748092, + "loss": 2.7285, + "step": 7055 + }, + { + "epoch": 0.3, + "grad_norm": 0.4140625, + "learning_rate": 0.0001893101273074392, + "loss": 2.6773, + "step": 7056 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.00018930710950553463, + "loss": 2.5184, + "step": 7057 + }, + { + "epoch": 0.3, + "grad_norm": 0.42578125, + "learning_rate": 0.00018930409130178067, + "loss": 2.6707, + "step": 7058 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.00018930107269619095, + "loss": 2.5981, + "step": 7059 + }, + { + "epoch": 0.3, + "grad_norm": 0.4296875, + "learning_rate": 0.00018929805368877905, + "loss": 2.6267, + "step": 7060 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.0001892950342795585, + "loss": 2.6181, + "step": 7061 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.00018929201446854296, + "loss": 2.597, + "step": 7062 + }, + { + "epoch": 0.3, + "grad_norm": 0.4609375, + "learning_rate": 0.00018928899425574595, + "loss": 2.5967, + "step": 7063 + }, + { + "epoch": 0.3, + "grad_norm": 0.421875, + "learning_rate": 0.00018928597364118114, + "loss": 2.5129, + "step": 7064 + }, + { + "epoch": 0.3, + "grad_norm": 0.4765625, + "learning_rate": 0.00018928295262486206, + "loss": 2.6915, + "step": 7065 + }, + { + "epoch": 0.3, + "grad_norm": 0.44140625, + "learning_rate": 0.00018927993120680233, + "loss": 2.5703, + "step": 7066 + }, + { + "epoch": 0.3, + "grad_norm": 0.419921875, + "learning_rate": 0.00018927690938701553, + "loss": 2.5153, + "step": 7067 + }, + { + "epoch": 0.3, + "grad_norm": 0.447265625, + "learning_rate": 0.00018927388716551526, + "loss": 2.6581, + "step": 7068 + }, + { + "epoch": 0.3, + "grad_norm": 0.482421875, + "learning_rate": 0.00018927086454231513, + "loss": 2.5918, + "step": 7069 + }, + { + "epoch": 0.3, + "grad_norm": 0.435546875, + "learning_rate": 0.00018926784151742875, + "loss": 2.7105, + "step": 7070 + }, + { + "epoch": 0.3, + "grad_norm": 0.447265625, + "learning_rate": 0.0001892648180908697, + "loss": 2.6959, + "step": 7071 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.00018926179426265158, + "loss": 2.6304, + "step": 7072 + }, + { + "epoch": 0.3, + "grad_norm": 0.3984375, + "learning_rate": 0.000189258770032788, + "loss": 2.5772, + "step": 7073 + }, + { + "epoch": 0.3, + "grad_norm": 0.46875, + "learning_rate": 0.0001892557454012926, + "loss": 2.7684, + "step": 7074 + }, + { + "epoch": 0.3, + "grad_norm": 0.462890625, + "learning_rate": 0.00018925272036817895, + "loss": 2.6733, + "step": 7075 + }, + { + "epoch": 0.3, + "grad_norm": 0.4609375, + "learning_rate": 0.0001892496949334607, + "loss": 2.6742, + "step": 7076 + }, + { + "epoch": 0.3, + "grad_norm": 0.451171875, + "learning_rate": 0.00018924666909715142, + "loss": 2.6915, + "step": 7077 + }, + { + "epoch": 0.3, + "grad_norm": 0.45703125, + "learning_rate": 0.00018924364285926473, + "loss": 2.6414, + "step": 7078 + }, + { + "epoch": 0.3, + "grad_norm": 0.41796875, + "learning_rate": 0.0001892406162198143, + "loss": 2.5579, + "step": 7079 + }, + { + "epoch": 0.3, + "grad_norm": 0.470703125, + "learning_rate": 0.00018923758917881368, + "loss": 2.5571, + "step": 7080 + }, + { + "epoch": 0.3, + "grad_norm": 0.451171875, + "learning_rate": 0.0001892345617362765, + "loss": 2.5966, + "step": 7081 + }, + { + "epoch": 0.3, + "grad_norm": 0.470703125, + "learning_rate": 0.00018923153389221645, + "loss": 2.6557, + "step": 7082 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.00018922850564664708, + "loss": 2.6756, + "step": 7083 + }, + { + "epoch": 0.3, + "grad_norm": 0.46484375, + "learning_rate": 0.00018922547699958204, + "loss": 2.6165, + "step": 7084 + }, + { + "epoch": 0.3, + "grad_norm": 0.53125, + "learning_rate": 0.00018922244795103497, + "loss": 2.8027, + "step": 7085 + }, + { + "epoch": 0.3, + "grad_norm": 0.458984375, + "learning_rate": 0.00018921941850101953, + "loss": 2.7345, + "step": 7086 + }, + { + "epoch": 0.3, + "grad_norm": 0.482421875, + "learning_rate": 0.00018921638864954928, + "loss": 2.6541, + "step": 7087 + }, + { + "epoch": 0.3, + "grad_norm": 0.5, + "learning_rate": 0.00018921335839663787, + "loss": 2.6942, + "step": 7088 + }, + { + "epoch": 0.3, + "grad_norm": 0.431640625, + "learning_rate": 0.00018921032774229897, + "loss": 2.644, + "step": 7089 + }, + { + "epoch": 0.3, + "grad_norm": 0.482421875, + "learning_rate": 0.00018920729668654617, + "loss": 2.5813, + "step": 7090 + }, + { + "epoch": 0.3, + "grad_norm": 0.40625, + "learning_rate": 0.00018920426522939316, + "loss": 2.6625, + "step": 7091 + }, + { + "epoch": 0.3, + "grad_norm": 0.453125, + "learning_rate": 0.00018920123337085355, + "loss": 2.5316, + "step": 7092 + }, + { + "epoch": 0.3, + "grad_norm": 0.416015625, + "learning_rate": 0.000189198201110941, + "loss": 2.6334, + "step": 7093 + }, + { + "epoch": 0.3, + "grad_norm": 0.40625, + "learning_rate": 0.00018919516844966916, + "loss": 2.4844, + "step": 7094 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.00018919213538705164, + "loss": 2.5279, + "step": 7095 + }, + { + "epoch": 0.3, + "grad_norm": 0.4609375, + "learning_rate": 0.00018918910192310207, + "loss": 2.6124, + "step": 7096 + }, + { + "epoch": 0.3, + "grad_norm": 0.427734375, + "learning_rate": 0.0001891860680578342, + "loss": 2.6195, + "step": 7097 + }, + { + "epoch": 0.3, + "grad_norm": 0.48046875, + "learning_rate": 0.00018918303379126159, + "loss": 2.6116, + "step": 7098 + }, + { + "epoch": 0.3, + "grad_norm": 0.4453125, + "learning_rate": 0.00018917999912339793, + "loss": 2.6054, + "step": 7099 + }, + { + "epoch": 0.3, + "grad_norm": 0.41796875, + "learning_rate": 0.00018917696405425685, + "loss": 2.6651, + "step": 7100 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.000189173928583852, + "loss": 2.7378, + "step": 7101 + }, + { + "epoch": 0.3, + "grad_norm": 0.447265625, + "learning_rate": 0.0001891708927121971, + "loss": 2.7668, + "step": 7102 + }, + { + "epoch": 0.3, + "grad_norm": 0.466796875, + "learning_rate": 0.00018916785643930575, + "loss": 2.6294, + "step": 7103 + }, + { + "epoch": 0.3, + "grad_norm": 0.470703125, + "learning_rate": 0.00018916481976519165, + "loss": 2.8321, + "step": 7104 + }, + { + "epoch": 0.3, + "grad_norm": 0.4609375, + "learning_rate": 0.00018916178268986843, + "loss": 2.6536, + "step": 7105 + }, + { + "epoch": 0.3, + "grad_norm": 0.443359375, + "learning_rate": 0.00018915874521334978, + "loss": 2.8009, + "step": 7106 + }, + { + "epoch": 0.3, + "grad_norm": 0.4140625, + "learning_rate": 0.00018915570733564935, + "loss": 2.7392, + "step": 7107 + }, + { + "epoch": 0.3, + "grad_norm": 0.46484375, + "learning_rate": 0.00018915266905678086, + "loss": 2.7501, + "step": 7108 + }, + { + "epoch": 0.3, + "grad_norm": 0.494140625, + "learning_rate": 0.0001891496303767579, + "loss": 2.7524, + "step": 7109 + }, + { + "epoch": 0.3, + "grad_norm": 0.46875, + "learning_rate": 0.0001891465912955942, + "loss": 2.6, + "step": 7110 + }, + { + "epoch": 0.3, + "grad_norm": 0.5, + "learning_rate": 0.00018914355181330343, + "loss": 2.623, + "step": 7111 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.00018914051192989924, + "loss": 2.6529, + "step": 7112 + }, + { + "epoch": 0.3, + "grad_norm": 0.423828125, + "learning_rate": 0.0001891374716453953, + "loss": 2.5487, + "step": 7113 + }, + { + "epoch": 0.3, + "grad_norm": 0.462890625, + "learning_rate": 0.00018913443095980534, + "loss": 2.6185, + "step": 7114 + }, + { + "epoch": 0.3, + "grad_norm": 0.42578125, + "learning_rate": 0.000189131389873143, + "loss": 2.7307, + "step": 7115 + }, + { + "epoch": 0.3, + "grad_norm": 0.4453125, + "learning_rate": 0.000189128348385422, + "loss": 2.5112, + "step": 7116 + }, + { + "epoch": 0.3, + "grad_norm": 0.431640625, + "learning_rate": 0.00018912530649665602, + "loss": 2.5909, + "step": 7117 + }, + { + "epoch": 0.3, + "grad_norm": 0.46875, + "learning_rate": 0.00018912226420685868, + "loss": 2.6417, + "step": 7118 + }, + { + "epoch": 0.3, + "grad_norm": 0.392578125, + "learning_rate": 0.00018911922151604378, + "loss": 2.7983, + "step": 7119 + }, + { + "epoch": 0.3, + "grad_norm": 0.51953125, + "learning_rate": 0.0001891161784242249, + "loss": 2.6961, + "step": 7120 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.00018911313493141582, + "loss": 2.4988, + "step": 7121 + }, + { + "epoch": 0.3, + "grad_norm": 0.466796875, + "learning_rate": 0.0001891100910376302, + "loss": 2.5639, + "step": 7122 + }, + { + "epoch": 0.3, + "grad_norm": 0.388671875, + "learning_rate": 0.00018910704674288172, + "loss": 2.6455, + "step": 7123 + }, + { + "epoch": 0.3, + "grad_norm": 0.45703125, + "learning_rate": 0.00018910400204718412, + "loss": 2.553, + "step": 7124 + }, + { + "epoch": 0.3, + "grad_norm": 0.46484375, + "learning_rate": 0.00018910095695055104, + "loss": 2.6648, + "step": 7125 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.00018909791145299622, + "loss": 2.673, + "step": 7126 + }, + { + "epoch": 0.3, + "grad_norm": 0.392578125, + "learning_rate": 0.00018909486555453335, + "loss": 2.6145, + "step": 7127 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.0001890918192551762, + "loss": 2.7611, + "step": 7128 + }, + { + "epoch": 0.3, + "grad_norm": 0.48046875, + "learning_rate": 0.0001890887725549384, + "loss": 2.5072, + "step": 7129 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.00018908572545383364, + "loss": 2.5749, + "step": 7130 + }, + { + "epoch": 0.3, + "grad_norm": 0.51953125, + "learning_rate": 0.0001890826779518757, + "loss": 2.6841, + "step": 7131 + }, + { + "epoch": 0.3, + "grad_norm": 0.474609375, + "learning_rate": 0.00018907963004907827, + "loss": 2.7055, + "step": 7132 + }, + { + "epoch": 0.3, + "grad_norm": 0.47265625, + "learning_rate": 0.00018907658174545505, + "loss": 2.6288, + "step": 7133 + }, + { + "epoch": 0.3, + "grad_norm": 0.486328125, + "learning_rate": 0.00018907353304101977, + "loss": 2.7729, + "step": 7134 + }, + { + "epoch": 0.3, + "grad_norm": 0.43359375, + "learning_rate": 0.00018907048393578614, + "loss": 2.777, + "step": 7135 + }, + { + "epoch": 0.3, + "grad_norm": 0.431640625, + "learning_rate": 0.00018906743442976791, + "loss": 2.6498, + "step": 7136 + }, + { + "epoch": 0.3, + "grad_norm": 0.44140625, + "learning_rate": 0.00018906438452297874, + "loss": 2.6864, + "step": 7137 + }, + { + "epoch": 0.3, + "grad_norm": 0.447265625, + "learning_rate": 0.00018906133421543243, + "loss": 2.7171, + "step": 7138 + }, + { + "epoch": 0.3, + "grad_norm": 0.47265625, + "learning_rate": 0.00018905828350714263, + "loss": 2.5651, + "step": 7139 + }, + { + "epoch": 0.3, + "grad_norm": 0.5234375, + "learning_rate": 0.00018905523239812314, + "loss": 2.7513, + "step": 7140 + }, + { + "epoch": 0.3, + "grad_norm": 0.396484375, + "learning_rate": 0.00018905218088838762, + "loss": 2.5995, + "step": 7141 + }, + { + "epoch": 0.3, + "grad_norm": 0.408203125, + "learning_rate": 0.00018904912897794982, + "loss": 2.6571, + "step": 7142 + }, + { + "epoch": 0.3, + "grad_norm": 0.478515625, + "learning_rate": 0.00018904607666682348, + "loss": 2.6201, + "step": 7143 + }, + { + "epoch": 0.3, + "grad_norm": 0.39453125, + "learning_rate": 0.00018904302395502236, + "loss": 2.5693, + "step": 7144 + }, + { + "epoch": 0.3, + "grad_norm": 0.431640625, + "learning_rate": 0.00018903997084256018, + "loss": 2.5773, + "step": 7145 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.00018903691732945067, + "loss": 2.6085, + "step": 7146 + }, + { + "epoch": 0.3, + "grad_norm": 0.486328125, + "learning_rate": 0.00018903386341570752, + "loss": 2.7136, + "step": 7147 + }, + { + "epoch": 0.3, + "grad_norm": 0.482421875, + "learning_rate": 0.0001890308091013446, + "loss": 2.717, + "step": 7148 + }, + { + "epoch": 0.3, + "grad_norm": 0.439453125, + "learning_rate": 0.00018902775438637553, + "loss": 2.6244, + "step": 7149 + }, + { + "epoch": 0.3, + "grad_norm": 0.419921875, + "learning_rate": 0.0001890246992708141, + "loss": 2.5853, + "step": 7150 + }, + { + "epoch": 0.3, + "grad_norm": 0.41015625, + "learning_rate": 0.00018902164375467407, + "loss": 2.3998, + "step": 7151 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.00018901858783796916, + "loss": 2.5701, + "step": 7152 + }, + { + "epoch": 0.3, + "grad_norm": 0.4375, + "learning_rate": 0.00018901553152071316, + "loss": 2.5905, + "step": 7153 + }, + { + "epoch": 0.3, + "grad_norm": 0.41015625, + "learning_rate": 0.00018901247480291979, + "loss": 2.5851, + "step": 7154 + }, + { + "epoch": 0.3, + "grad_norm": 0.455078125, + "learning_rate": 0.0001890094176846028, + "loss": 2.6915, + "step": 7155 + }, + { + "epoch": 0.3, + "grad_norm": 0.44921875, + "learning_rate": 0.00018900636016577597, + "loss": 2.6744, + "step": 7156 + }, + { + "epoch": 0.3, + "grad_norm": 0.396484375, + "learning_rate": 0.00018900330224645305, + "loss": 2.6684, + "step": 7157 + }, + { + "epoch": 0.31, + "grad_norm": 0.431640625, + "learning_rate": 0.00018900024392664783, + "loss": 2.7079, + "step": 7158 + }, + { + "epoch": 0.31, + "grad_norm": 0.431640625, + "learning_rate": 0.00018899718520637402, + "loss": 2.5687, + "step": 7159 + }, + { + "epoch": 0.31, + "grad_norm": 0.462890625, + "learning_rate": 0.00018899412608564537, + "loss": 2.6137, + "step": 7160 + }, + { + "epoch": 0.31, + "grad_norm": 0.44140625, + "learning_rate": 0.0001889910665644757, + "loss": 2.71, + "step": 7161 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.00018898800664287875, + "loss": 2.8186, + "step": 7162 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.0001889849463208683, + "loss": 2.6357, + "step": 7163 + }, + { + "epoch": 0.31, + "grad_norm": 0.45703125, + "learning_rate": 0.00018898188559845808, + "loss": 2.6039, + "step": 7164 + }, + { + "epoch": 0.31, + "grad_norm": 0.41015625, + "learning_rate": 0.00018897882447566194, + "loss": 2.6716, + "step": 7165 + }, + { + "epoch": 0.31, + "grad_norm": 0.38671875, + "learning_rate": 0.00018897576295249358, + "loss": 2.5251, + "step": 7166 + }, + { + "epoch": 0.31, + "grad_norm": 0.478515625, + "learning_rate": 0.00018897270102896682, + "loss": 2.6173, + "step": 7167 + }, + { + "epoch": 0.31, + "grad_norm": 0.42578125, + "learning_rate": 0.0001889696387050954, + "loss": 2.5678, + "step": 7168 + }, + { + "epoch": 0.31, + "grad_norm": 0.453125, + "learning_rate": 0.0001889665759808931, + "loss": 2.6928, + "step": 7169 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.00018896351285637377, + "loss": 2.7123, + "step": 7170 + }, + { + "epoch": 0.31, + "grad_norm": 0.443359375, + "learning_rate": 0.0001889604493315511, + "loss": 2.6716, + "step": 7171 + }, + { + "epoch": 0.31, + "grad_norm": 0.435546875, + "learning_rate": 0.00018895738540643895, + "loss": 2.6103, + "step": 7172 + }, + { + "epoch": 0.31, + "grad_norm": 0.423828125, + "learning_rate": 0.00018895432108105106, + "loss": 2.6127, + "step": 7173 + }, + { + "epoch": 0.31, + "grad_norm": 0.447265625, + "learning_rate": 0.00018895125635540126, + "loss": 2.5668, + "step": 7174 + }, + { + "epoch": 0.31, + "grad_norm": 0.427734375, + "learning_rate": 0.0001889481912295033, + "loss": 2.5755, + "step": 7175 + }, + { + "epoch": 0.31, + "grad_norm": 0.359375, + "learning_rate": 0.00018894512570337099, + "loss": 2.6325, + "step": 7176 + }, + { + "epoch": 0.31, + "grad_norm": 0.416015625, + "learning_rate": 0.0001889420597770181, + "loss": 2.6437, + "step": 7177 + }, + { + "epoch": 0.31, + "grad_norm": 0.4375, + "learning_rate": 0.00018893899345045847, + "loss": 2.54, + "step": 7178 + }, + { + "epoch": 0.31, + "grad_norm": 0.419921875, + "learning_rate": 0.00018893592672370585, + "loss": 2.6489, + "step": 7179 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.0001889328595967741, + "loss": 2.5936, + "step": 7180 + }, + { + "epoch": 0.31, + "grad_norm": 0.49609375, + "learning_rate": 0.00018892979206967697, + "loss": 2.6586, + "step": 7181 + }, + { + "epoch": 0.31, + "grad_norm": 0.484375, + "learning_rate": 0.00018892672414242822, + "loss": 2.6191, + "step": 7182 + }, + { + "epoch": 0.31, + "grad_norm": 0.474609375, + "learning_rate": 0.00018892365581504178, + "loss": 2.6682, + "step": 7183 + }, + { + "epoch": 0.31, + "grad_norm": 0.421875, + "learning_rate": 0.00018892058708753134, + "loss": 2.6997, + "step": 7184 + }, + { + "epoch": 0.31, + "grad_norm": 0.44140625, + "learning_rate": 0.00018891751795991078, + "loss": 2.5831, + "step": 7185 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.0001889144484321939, + "loss": 2.6067, + "step": 7186 + }, + { + "epoch": 0.31, + "grad_norm": 0.443359375, + "learning_rate": 0.00018891137850439446, + "loss": 2.682, + "step": 7187 + }, + { + "epoch": 0.31, + "grad_norm": 0.4609375, + "learning_rate": 0.00018890830817652635, + "loss": 2.7545, + "step": 7188 + }, + { + "epoch": 0.31, + "grad_norm": 0.45703125, + "learning_rate": 0.00018890523744860332, + "loss": 2.5917, + "step": 7189 + }, + { + "epoch": 0.31, + "grad_norm": 0.435546875, + "learning_rate": 0.00018890216632063922, + "loss": 2.4931, + "step": 7190 + }, + { + "epoch": 0.31, + "grad_norm": 0.48828125, + "learning_rate": 0.00018889909479264785, + "loss": 2.5862, + "step": 7191 + }, + { + "epoch": 0.31, + "grad_norm": 0.412109375, + "learning_rate": 0.00018889602286464304, + "loss": 2.7071, + "step": 7192 + }, + { + "epoch": 0.31, + "grad_norm": 0.451171875, + "learning_rate": 0.00018889295053663862, + "loss": 2.6234, + "step": 7193 + }, + { + "epoch": 0.31, + "grad_norm": 0.4140625, + "learning_rate": 0.00018888987780864844, + "loss": 2.5777, + "step": 7194 + }, + { + "epoch": 0.31, + "grad_norm": 0.5, + "learning_rate": 0.00018888680468068624, + "loss": 2.6381, + "step": 7195 + }, + { + "epoch": 0.31, + "grad_norm": 0.423828125, + "learning_rate": 0.00018888373115276595, + "loss": 2.631, + "step": 7196 + }, + { + "epoch": 0.31, + "grad_norm": 0.482421875, + "learning_rate": 0.00018888065722490134, + "loss": 2.6387, + "step": 7197 + }, + { + "epoch": 0.31, + "grad_norm": 0.40625, + "learning_rate": 0.00018887758289710623, + "loss": 2.6236, + "step": 7198 + }, + { + "epoch": 0.31, + "grad_norm": 0.474609375, + "learning_rate": 0.0001888745081693945, + "loss": 2.6379, + "step": 7199 + }, + { + "epoch": 0.31, + "grad_norm": 0.482421875, + "learning_rate": 0.00018887143304177997, + "loss": 2.7766, + "step": 7200 + }, + { + "epoch": 0.31, + "grad_norm": 0.4140625, + "learning_rate": 0.00018886835751427648, + "loss": 2.5703, + "step": 7201 + }, + { + "epoch": 0.31, + "grad_norm": 0.453125, + "learning_rate": 0.00018886528158689782, + "loss": 2.6074, + "step": 7202 + }, + { + "epoch": 0.31, + "grad_norm": 0.466796875, + "learning_rate": 0.00018886220525965789, + "loss": 2.6462, + "step": 7203 + }, + { + "epoch": 0.31, + "grad_norm": 0.4453125, + "learning_rate": 0.00018885912853257053, + "loss": 2.611, + "step": 7204 + }, + { + "epoch": 0.31, + "grad_norm": 0.41015625, + "learning_rate": 0.00018885605140564954, + "loss": 2.6793, + "step": 7205 + }, + { + "epoch": 0.31, + "grad_norm": 0.490234375, + "learning_rate": 0.0001888529738789088, + "loss": 2.7177, + "step": 7206 + }, + { + "epoch": 0.31, + "grad_norm": 0.40234375, + "learning_rate": 0.00018884989595236213, + "loss": 2.5444, + "step": 7207 + }, + { + "epoch": 0.31, + "grad_norm": 0.490234375, + "learning_rate": 0.00018884681762602345, + "loss": 2.6261, + "step": 7208 + }, + { + "epoch": 0.31, + "grad_norm": 0.4921875, + "learning_rate": 0.00018884373889990652, + "loss": 2.5921, + "step": 7209 + }, + { + "epoch": 0.31, + "grad_norm": 0.3984375, + "learning_rate": 0.00018884065977402522, + "loss": 2.5711, + "step": 7210 + }, + { + "epoch": 0.31, + "grad_norm": 0.390625, + "learning_rate": 0.00018883758024839346, + "loss": 2.6222, + "step": 7211 + }, + { + "epoch": 0.31, + "grad_norm": 0.453125, + "learning_rate": 0.00018883450032302504, + "loss": 2.5978, + "step": 7212 + }, + { + "epoch": 0.31, + "grad_norm": 0.42578125, + "learning_rate": 0.00018883141999793385, + "loss": 2.6135, + "step": 7213 + }, + { + "epoch": 0.31, + "grad_norm": 0.447265625, + "learning_rate": 0.0001888283392731337, + "loss": 2.5758, + "step": 7214 + }, + { + "epoch": 0.31, + "grad_norm": 0.447265625, + "learning_rate": 0.0001888252581486385, + "loss": 2.6654, + "step": 7215 + }, + { + "epoch": 0.31, + "grad_norm": 0.447265625, + "learning_rate": 0.0001888221766244621, + "loss": 2.5358, + "step": 7216 + }, + { + "epoch": 0.31, + "grad_norm": 0.40625, + "learning_rate": 0.00018881909470061837, + "loss": 2.6943, + "step": 7217 + }, + { + "epoch": 0.31, + "grad_norm": 0.435546875, + "learning_rate": 0.00018881601237712115, + "loss": 2.5717, + "step": 7218 + }, + { + "epoch": 0.31, + "grad_norm": 0.39453125, + "learning_rate": 0.00018881292965398436, + "loss": 2.5676, + "step": 7219 + }, + { + "epoch": 0.31, + "grad_norm": 0.470703125, + "learning_rate": 0.00018880984653122184, + "loss": 2.5596, + "step": 7220 + }, + { + "epoch": 0.31, + "grad_norm": 0.53125, + "learning_rate": 0.00018880676300884748, + "loss": 2.7818, + "step": 7221 + }, + { + "epoch": 0.31, + "grad_norm": 0.408203125, + "learning_rate": 0.0001888036790868751, + "loss": 2.5503, + "step": 7222 + }, + { + "epoch": 0.31, + "grad_norm": 0.41015625, + "learning_rate": 0.00018880059476531863, + "loss": 2.4274, + "step": 7223 + }, + { + "epoch": 0.31, + "grad_norm": 0.416015625, + "learning_rate": 0.00018879751004419198, + "loss": 2.6834, + "step": 7224 + }, + { + "epoch": 0.31, + "grad_norm": 0.40625, + "learning_rate": 0.00018879442492350897, + "loss": 2.5531, + "step": 7225 + }, + { + "epoch": 0.31, + "grad_norm": 0.423828125, + "learning_rate": 0.00018879133940328348, + "loss": 2.7471, + "step": 7226 + }, + { + "epoch": 0.31, + "grad_norm": 0.44921875, + "learning_rate": 0.0001887882534835294, + "loss": 2.546, + "step": 7227 + }, + { + "epoch": 0.31, + "grad_norm": 0.44140625, + "learning_rate": 0.00018878516716426064, + "loss": 2.6213, + "step": 7228 + }, + { + "epoch": 0.31, + "grad_norm": 0.388671875, + "learning_rate": 0.00018878208044549108, + "loss": 2.6084, + "step": 7229 + }, + { + "epoch": 0.31, + "grad_norm": 0.4453125, + "learning_rate": 0.0001887789933272346, + "loss": 2.6626, + "step": 7230 + }, + { + "epoch": 0.31, + "grad_norm": 0.44921875, + "learning_rate": 0.00018877590580950512, + "loss": 2.6236, + "step": 7231 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.00018877281789231648, + "loss": 2.5848, + "step": 7232 + }, + { + "epoch": 0.31, + "grad_norm": 0.427734375, + "learning_rate": 0.0001887697295756826, + "loss": 2.6842, + "step": 7233 + }, + { + "epoch": 0.31, + "grad_norm": 0.421875, + "learning_rate": 0.0001887666408596174, + "loss": 2.6955, + "step": 7234 + }, + { + "epoch": 0.31, + "grad_norm": 0.4765625, + "learning_rate": 0.00018876355174413475, + "loss": 2.5556, + "step": 7235 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.00018876046222924856, + "loss": 2.725, + "step": 7236 + }, + { + "epoch": 0.31, + "grad_norm": 0.4609375, + "learning_rate": 0.0001887573723149727, + "loss": 2.5231, + "step": 7237 + }, + { + "epoch": 0.31, + "grad_norm": 0.4609375, + "learning_rate": 0.00018875428200132113, + "loss": 2.6261, + "step": 7238 + }, + { + "epoch": 0.31, + "grad_norm": 0.423828125, + "learning_rate": 0.00018875119128830773, + "loss": 2.5825, + "step": 7239 + }, + { + "epoch": 0.31, + "grad_norm": 0.494140625, + "learning_rate": 0.00018874810017594637, + "loss": 2.7992, + "step": 7240 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.00018874500866425102, + "loss": 2.7302, + "step": 7241 + }, + { + "epoch": 0.31, + "grad_norm": 0.4375, + "learning_rate": 0.00018874191675323555, + "loss": 2.7197, + "step": 7242 + }, + { + "epoch": 0.31, + "grad_norm": 0.4296875, + "learning_rate": 0.0001887388244429139, + "loss": 2.7981, + "step": 7243 + }, + { + "epoch": 0.31, + "grad_norm": 0.48828125, + "learning_rate": 0.00018873573173329997, + "loss": 2.7179, + "step": 7244 + }, + { + "epoch": 0.31, + "grad_norm": 0.427734375, + "learning_rate": 0.00018873263862440766, + "loss": 2.7316, + "step": 7245 + }, + { + "epoch": 0.31, + "grad_norm": 0.4375, + "learning_rate": 0.00018872954511625091, + "loss": 2.6445, + "step": 7246 + }, + { + "epoch": 0.31, + "grad_norm": 0.48046875, + "learning_rate": 0.00018872645120884368, + "loss": 2.566, + "step": 7247 + }, + { + "epoch": 0.31, + "grad_norm": 0.466796875, + "learning_rate": 0.00018872335690219979, + "loss": 2.7005, + "step": 7248 + }, + { + "epoch": 0.31, + "grad_norm": 0.482421875, + "learning_rate": 0.00018872026219633324, + "loss": 2.7564, + "step": 7249 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.0001887171670912579, + "loss": 2.6444, + "step": 7250 + }, + { + "epoch": 0.31, + "grad_norm": 0.4765625, + "learning_rate": 0.00018871407158698778, + "loss": 2.7398, + "step": 7251 + }, + { + "epoch": 0.31, + "grad_norm": 0.45703125, + "learning_rate": 0.00018871097568353673, + "loss": 2.5946, + "step": 7252 + }, + { + "epoch": 0.31, + "grad_norm": 0.458984375, + "learning_rate": 0.0001887078793809187, + "loss": 2.6006, + "step": 7253 + }, + { + "epoch": 0.31, + "grad_norm": 0.3828125, + "learning_rate": 0.00018870478267914766, + "loss": 2.7015, + "step": 7254 + }, + { + "epoch": 0.31, + "grad_norm": 0.7109375, + "learning_rate": 0.0001887016855782375, + "loss": 2.6848, + "step": 7255 + }, + { + "epoch": 0.31, + "grad_norm": 0.470703125, + "learning_rate": 0.00018869858807820218, + "loss": 2.765, + "step": 7256 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.00018869549017905564, + "loss": 2.6746, + "step": 7257 + }, + { + "epoch": 0.31, + "grad_norm": 0.416015625, + "learning_rate": 0.00018869239188081178, + "loss": 2.5312, + "step": 7258 + }, + { + "epoch": 0.31, + "grad_norm": 0.478515625, + "learning_rate": 0.0001886892931834846, + "loss": 2.5856, + "step": 7259 + }, + { + "epoch": 0.31, + "grad_norm": 0.4140625, + "learning_rate": 0.00018868619408708796, + "loss": 2.5423, + "step": 7260 + }, + { + "epoch": 0.31, + "grad_norm": 0.458984375, + "learning_rate": 0.0001886830945916359, + "loss": 2.7179, + "step": 7261 + }, + { + "epoch": 0.31, + "grad_norm": 0.470703125, + "learning_rate": 0.00018867999469714234, + "loss": 2.7567, + "step": 7262 + }, + { + "epoch": 0.31, + "grad_norm": 0.47265625, + "learning_rate": 0.0001886768944036212, + "loss": 2.6245, + "step": 7263 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.00018867379371108643, + "loss": 2.6273, + "step": 7264 + }, + { + "epoch": 0.31, + "grad_norm": 0.4375, + "learning_rate": 0.00018867069261955198, + "loss": 2.6151, + "step": 7265 + }, + { + "epoch": 0.31, + "grad_norm": 0.51953125, + "learning_rate": 0.00018866759112903182, + "loss": 2.7067, + "step": 7266 + }, + { + "epoch": 0.31, + "grad_norm": 0.412109375, + "learning_rate": 0.00018866448923953995, + "loss": 2.7125, + "step": 7267 + }, + { + "epoch": 0.31, + "grad_norm": 0.4921875, + "learning_rate": 0.00018866138695109024, + "loss": 2.6917, + "step": 7268 + }, + { + "epoch": 0.31, + "grad_norm": 0.453125, + "learning_rate": 0.00018865828426369672, + "loss": 2.5441, + "step": 7269 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.00018865518117737332, + "loss": 2.663, + "step": 7270 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.00018865207769213399, + "loss": 2.7061, + "step": 7271 + }, + { + "epoch": 0.31, + "grad_norm": 0.53125, + "learning_rate": 0.00018864897380799269, + "loss": 2.7037, + "step": 7272 + }, + { + "epoch": 0.31, + "grad_norm": 0.39453125, + "learning_rate": 0.0001886458695249634, + "loss": 2.4932, + "step": 7273 + }, + { + "epoch": 0.31, + "grad_norm": 0.4375, + "learning_rate": 0.00018864276484306013, + "loss": 2.6005, + "step": 7274 + }, + { + "epoch": 0.31, + "grad_norm": 0.431640625, + "learning_rate": 0.00018863965976229682, + "loss": 2.6806, + "step": 7275 + }, + { + "epoch": 0.31, + "grad_norm": 0.4921875, + "learning_rate": 0.0001886365542826874, + "loss": 2.7811, + "step": 7276 + }, + { + "epoch": 0.31, + "grad_norm": 0.455078125, + "learning_rate": 0.0001886334484042459, + "loss": 2.5443, + "step": 7277 + }, + { + "epoch": 0.31, + "grad_norm": 0.4453125, + "learning_rate": 0.00018863034212698626, + "loss": 2.6001, + "step": 7278 + }, + { + "epoch": 0.31, + "grad_norm": 0.47265625, + "learning_rate": 0.0001886272354509225, + "loss": 2.552, + "step": 7279 + }, + { + "epoch": 0.31, + "grad_norm": 0.48828125, + "learning_rate": 0.00018862412837606854, + "loss": 2.7152, + "step": 7280 + }, + { + "epoch": 0.31, + "grad_norm": 0.48046875, + "learning_rate": 0.0001886210209024384, + "loss": 2.7679, + "step": 7281 + }, + { + "epoch": 0.31, + "grad_norm": 0.486328125, + "learning_rate": 0.00018861791303004605, + "loss": 2.7317, + "step": 7282 + }, + { + "epoch": 0.31, + "grad_norm": 0.431640625, + "learning_rate": 0.0001886148047589055, + "loss": 2.6749, + "step": 7283 + }, + { + "epoch": 0.31, + "grad_norm": 0.453125, + "learning_rate": 0.0001886116960890307, + "loss": 2.6684, + "step": 7284 + }, + { + "epoch": 0.31, + "grad_norm": 0.470703125, + "learning_rate": 0.00018860858702043562, + "loss": 2.7863, + "step": 7285 + }, + { + "epoch": 0.31, + "grad_norm": 0.419921875, + "learning_rate": 0.00018860547755313433, + "loss": 2.6031, + "step": 7286 + }, + { + "epoch": 0.31, + "grad_norm": 0.42578125, + "learning_rate": 0.00018860236768714075, + "loss": 2.6016, + "step": 7287 + }, + { + "epoch": 0.31, + "grad_norm": 0.458984375, + "learning_rate": 0.00018859925742246888, + "loss": 2.5378, + "step": 7288 + }, + { + "epoch": 0.31, + "grad_norm": 0.470703125, + "learning_rate": 0.00018859614675913275, + "loss": 2.6504, + "step": 7289 + }, + { + "epoch": 0.31, + "grad_norm": 0.43359375, + "learning_rate": 0.00018859303569714636, + "loss": 2.6477, + "step": 7290 + }, + { + "epoch": 0.31, + "grad_norm": 0.4296875, + "learning_rate": 0.00018858992423652365, + "loss": 2.6655, + "step": 7291 + }, + { + "epoch": 0.31, + "grad_norm": 0.435546875, + "learning_rate": 0.00018858681237727867, + "loss": 2.6129, + "step": 7292 + }, + { + "epoch": 0.31, + "grad_norm": 0.451171875, + "learning_rate": 0.0001885837001194254, + "loss": 2.6042, + "step": 7293 + }, + { + "epoch": 0.31, + "grad_norm": 0.421875, + "learning_rate": 0.00018858058746297785, + "loss": 2.7284, + "step": 7294 + }, + { + "epoch": 0.31, + "grad_norm": 0.482421875, + "learning_rate": 0.00018857747440795007, + "loss": 2.6024, + "step": 7295 + }, + { + "epoch": 0.31, + "grad_norm": 0.49609375, + "learning_rate": 0.00018857436095435598, + "loss": 2.7359, + "step": 7296 + }, + { + "epoch": 0.31, + "grad_norm": 0.466796875, + "learning_rate": 0.00018857124710220966, + "loss": 2.5005, + "step": 7297 + }, + { + "epoch": 0.31, + "grad_norm": 0.44140625, + "learning_rate": 0.00018856813285152507, + "loss": 2.6317, + "step": 7298 + }, + { + "epoch": 0.31, + "grad_norm": 0.52734375, + "learning_rate": 0.0001885650182023163, + "loss": 2.6992, + "step": 7299 + }, + { + "epoch": 0.31, + "grad_norm": 0.421875, + "learning_rate": 0.00018856190315459726, + "loss": 2.6097, + "step": 7300 + }, + { + "epoch": 0.31, + "grad_norm": 0.486328125, + "learning_rate": 0.00018855878770838205, + "loss": 2.6014, + "step": 7301 + }, + { + "epoch": 0.31, + "grad_norm": 0.388671875, + "learning_rate": 0.00018855567186368467, + "loss": 2.5133, + "step": 7302 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.0001885525556205191, + "loss": 2.6609, + "step": 7303 + }, + { + "epoch": 0.31, + "grad_norm": 0.474609375, + "learning_rate": 0.00018854943897889941, + "loss": 2.6161, + "step": 7304 + }, + { + "epoch": 0.31, + "grad_norm": 0.466796875, + "learning_rate": 0.00018854632193883962, + "loss": 2.588, + "step": 7305 + }, + { + "epoch": 0.31, + "grad_norm": 0.43359375, + "learning_rate": 0.00018854320450035372, + "loss": 2.6411, + "step": 7306 + }, + { + "epoch": 0.31, + "grad_norm": 0.490234375, + "learning_rate": 0.00018854008666345574, + "loss": 2.5941, + "step": 7307 + }, + { + "epoch": 0.31, + "grad_norm": 0.439453125, + "learning_rate": 0.00018853696842815977, + "loss": 2.6842, + "step": 7308 + }, + { + "epoch": 0.31, + "grad_norm": 0.45703125, + "learning_rate": 0.0001885338497944798, + "loss": 2.636, + "step": 7309 + }, + { + "epoch": 0.31, + "grad_norm": 0.44140625, + "learning_rate": 0.00018853073076242983, + "loss": 2.4382, + "step": 7310 + }, + { + "epoch": 0.31, + "grad_norm": 0.45703125, + "learning_rate": 0.00018852761133202392, + "loss": 2.5435, + "step": 7311 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.00018852449150327615, + "loss": 2.7148, + "step": 7312 + }, + { + "epoch": 0.31, + "grad_norm": 0.46875, + "learning_rate": 0.0001885213712762005, + "loss": 2.7429, + "step": 7313 + }, + { + "epoch": 0.31, + "grad_norm": 0.42578125, + "learning_rate": 0.00018851825065081102, + "loss": 2.6594, + "step": 7314 + }, + { + "epoch": 0.31, + "grad_norm": 0.48046875, + "learning_rate": 0.00018851512962712177, + "loss": 2.7739, + "step": 7315 + }, + { + "epoch": 0.31, + "grad_norm": 0.421875, + "learning_rate": 0.00018851200820514677, + "loss": 2.4203, + "step": 7316 + }, + { + "epoch": 0.31, + "grad_norm": 0.447265625, + "learning_rate": 0.0001885088863849001, + "loss": 2.6265, + "step": 7317 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.00018850576416639577, + "loss": 2.6427, + "step": 7318 + }, + { + "epoch": 0.31, + "grad_norm": 0.4140625, + "learning_rate": 0.00018850264154964788, + "loss": 2.6091, + "step": 7319 + }, + { + "epoch": 0.31, + "grad_norm": 0.4375, + "learning_rate": 0.00018849951853467043, + "loss": 2.4835, + "step": 7320 + }, + { + "epoch": 0.31, + "grad_norm": 0.478515625, + "learning_rate": 0.0001884963951214775, + "loss": 2.6835, + "step": 7321 + }, + { + "epoch": 0.31, + "grad_norm": 0.486328125, + "learning_rate": 0.0001884932713100831, + "loss": 2.6471, + "step": 7322 + }, + { + "epoch": 0.31, + "grad_norm": 0.4375, + "learning_rate": 0.00018849014710050136, + "loss": 2.5778, + "step": 7323 + }, + { + "epoch": 0.31, + "grad_norm": 0.51171875, + "learning_rate": 0.00018848702249274629, + "loss": 2.744, + "step": 7324 + }, + { + "epoch": 0.31, + "grad_norm": 0.453125, + "learning_rate": 0.00018848389748683193, + "loss": 2.7064, + "step": 7325 + }, + { + "epoch": 0.31, + "grad_norm": 0.388671875, + "learning_rate": 0.00018848077208277237, + "loss": 2.7065, + "step": 7326 + }, + { + "epoch": 0.31, + "grad_norm": 0.458984375, + "learning_rate": 0.0001884776462805817, + "loss": 2.6238, + "step": 7327 + }, + { + "epoch": 0.31, + "grad_norm": 0.408203125, + "learning_rate": 0.00018847452008027396, + "loss": 2.5912, + "step": 7328 + }, + { + "epoch": 0.31, + "grad_norm": 0.41796875, + "learning_rate": 0.00018847139348186315, + "loss": 2.57, + "step": 7329 + }, + { + "epoch": 0.31, + "grad_norm": 0.447265625, + "learning_rate": 0.00018846826648536346, + "loss": 2.7738, + "step": 7330 + }, + { + "epoch": 0.31, + "grad_norm": 0.49609375, + "learning_rate": 0.0001884651390907889, + "loss": 2.7744, + "step": 7331 + }, + { + "epoch": 0.31, + "grad_norm": 0.443359375, + "learning_rate": 0.00018846201129815352, + "loss": 2.6041, + "step": 7332 + }, + { + "epoch": 0.31, + "grad_norm": 0.435546875, + "learning_rate": 0.00018845888310747144, + "loss": 2.57, + "step": 7333 + }, + { + "epoch": 0.31, + "grad_norm": 0.50390625, + "learning_rate": 0.00018845575451875672, + "loss": 2.6966, + "step": 7334 + }, + { + "epoch": 0.31, + "grad_norm": 0.390625, + "learning_rate": 0.0001884526255320234, + "loss": 2.6239, + "step": 7335 + }, + { + "epoch": 0.31, + "grad_norm": 0.46484375, + "learning_rate": 0.00018844949614728562, + "loss": 2.6742, + "step": 7336 + }, + { + "epoch": 0.31, + "grad_norm": 0.484375, + "learning_rate": 0.00018844636636455742, + "loss": 2.7183, + "step": 7337 + }, + { + "epoch": 0.31, + "grad_norm": 0.44921875, + "learning_rate": 0.00018844323618385288, + "loss": 2.6087, + "step": 7338 + }, + { + "epoch": 0.31, + "grad_norm": 0.416015625, + "learning_rate": 0.00018844010560518614, + "loss": 2.5588, + "step": 7339 + }, + { + "epoch": 0.31, + "grad_norm": 0.4765625, + "learning_rate": 0.0001884369746285712, + "loss": 2.581, + "step": 7340 + }, + { + "epoch": 0.31, + "grad_norm": 0.478515625, + "learning_rate": 0.00018843384325402222, + "loss": 2.6628, + "step": 7341 + }, + { + "epoch": 0.31, + "grad_norm": 0.447265625, + "learning_rate": 0.00018843071148155326, + "loss": 2.5813, + "step": 7342 + }, + { + "epoch": 0.31, + "grad_norm": 0.400390625, + "learning_rate": 0.0001884275793111784, + "loss": 2.5564, + "step": 7343 + }, + { + "epoch": 0.31, + "grad_norm": 0.439453125, + "learning_rate": 0.00018842444674291179, + "loss": 2.808, + "step": 7344 + }, + { + "epoch": 0.31, + "grad_norm": 0.478515625, + "learning_rate": 0.00018842131377676746, + "loss": 2.6035, + "step": 7345 + }, + { + "epoch": 0.31, + "grad_norm": 0.43359375, + "learning_rate": 0.0001884181804127595, + "loss": 2.5054, + "step": 7346 + }, + { + "epoch": 0.31, + "grad_norm": 0.498046875, + "learning_rate": 0.0001884150466509021, + "loss": 2.7079, + "step": 7347 + }, + { + "epoch": 0.31, + "grad_norm": 0.421875, + "learning_rate": 0.00018841191249120928, + "loss": 2.7212, + "step": 7348 + }, + { + "epoch": 0.31, + "grad_norm": 0.50390625, + "learning_rate": 0.00018840877793369515, + "loss": 2.654, + "step": 7349 + }, + { + "epoch": 0.31, + "grad_norm": 0.51953125, + "learning_rate": 0.00018840564297837385, + "loss": 2.6697, + "step": 7350 + }, + { + "epoch": 0.31, + "grad_norm": 0.466796875, + "learning_rate": 0.00018840250762525943, + "loss": 2.6072, + "step": 7351 + }, + { + "epoch": 0.31, + "grad_norm": 0.53125, + "learning_rate": 0.00018839937187436608, + "loss": 2.7857, + "step": 7352 + }, + { + "epoch": 0.31, + "grad_norm": 0.46875, + "learning_rate": 0.00018839623572570782, + "loss": 2.7645, + "step": 7353 + }, + { + "epoch": 0.31, + "grad_norm": 0.498046875, + "learning_rate": 0.00018839309917929886, + "loss": 2.7084, + "step": 7354 + }, + { + "epoch": 0.31, + "grad_norm": 0.494140625, + "learning_rate": 0.00018838996223515323, + "loss": 2.6059, + "step": 7355 + }, + { + "epoch": 0.31, + "grad_norm": 0.435546875, + "learning_rate": 0.00018838682489328505, + "loss": 2.5199, + "step": 7356 + }, + { + "epoch": 0.31, + "grad_norm": 0.41796875, + "learning_rate": 0.00018838368715370847, + "loss": 2.5658, + "step": 7357 + }, + { + "epoch": 0.31, + "grad_norm": 0.50390625, + "learning_rate": 0.0001883805490164376, + "loss": 2.5198, + "step": 7358 + }, + { + "epoch": 0.31, + "grad_norm": 0.5, + "learning_rate": 0.00018837741048148657, + "loss": 2.7014, + "step": 7359 + }, + { + "epoch": 0.31, + "grad_norm": 0.41015625, + "learning_rate": 0.0001883742715488695, + "loss": 2.5227, + "step": 7360 + }, + { + "epoch": 0.31, + "grad_norm": 0.49609375, + "learning_rate": 0.00018837113221860046, + "loss": 2.6694, + "step": 7361 + }, + { + "epoch": 0.31, + "grad_norm": 0.4296875, + "learning_rate": 0.00018836799249069365, + "loss": 2.6109, + "step": 7362 + }, + { + "epoch": 0.31, + "grad_norm": 0.439453125, + "learning_rate": 0.00018836485236516318, + "loss": 2.5948, + "step": 7363 + }, + { + "epoch": 0.31, + "grad_norm": 0.462890625, + "learning_rate": 0.00018836171184202316, + "loss": 2.7046, + "step": 7364 + }, + { + "epoch": 0.31, + "grad_norm": 0.48046875, + "learning_rate": 0.0001883585709212877, + "loss": 2.6332, + "step": 7365 + }, + { + "epoch": 0.31, + "grad_norm": 0.470703125, + "learning_rate": 0.00018835542960297098, + "loss": 2.5305, + "step": 7366 + }, + { + "epoch": 0.31, + "grad_norm": 0.482421875, + "learning_rate": 0.00018835228788708714, + "loss": 2.6571, + "step": 7367 + }, + { + "epoch": 0.31, + "grad_norm": 0.439453125, + "learning_rate": 0.0001883491457736503, + "loss": 2.719, + "step": 7368 + }, + { + "epoch": 0.31, + "grad_norm": 0.390625, + "learning_rate": 0.00018834600326267454, + "loss": 2.5091, + "step": 7369 + }, + { + "epoch": 0.31, + "grad_norm": 0.470703125, + "learning_rate": 0.0001883428603541741, + "loss": 2.5486, + "step": 7370 + }, + { + "epoch": 0.31, + "grad_norm": 0.41796875, + "learning_rate": 0.00018833971704816307, + "loss": 2.585, + "step": 7371 + }, + { + "epoch": 0.31, + "grad_norm": 0.421875, + "learning_rate": 0.00018833657334465557, + "loss": 2.619, + "step": 7372 + }, + { + "epoch": 0.31, + "grad_norm": 0.4609375, + "learning_rate": 0.00018833342924366576, + "loss": 2.6438, + "step": 7373 + }, + { + "epoch": 0.31, + "grad_norm": 0.447265625, + "learning_rate": 0.00018833028474520783, + "loss": 2.5955, + "step": 7374 + }, + { + "epoch": 0.31, + "grad_norm": 0.458984375, + "learning_rate": 0.00018832713984929592, + "loss": 2.5944, + "step": 7375 + }, + { + "epoch": 0.31, + "grad_norm": 0.44140625, + "learning_rate": 0.00018832399455594414, + "loss": 2.528, + "step": 7376 + }, + { + "epoch": 0.31, + "grad_norm": 0.43359375, + "learning_rate": 0.00018832084886516665, + "loss": 2.6844, + "step": 7377 + }, + { + "epoch": 0.31, + "grad_norm": 0.40625, + "learning_rate": 0.00018831770277697764, + "loss": 2.5987, + "step": 7378 + }, + { + "epoch": 0.31, + "grad_norm": 0.462890625, + "learning_rate": 0.00018831455629139123, + "loss": 2.6231, + "step": 7379 + }, + { + "epoch": 0.31, + "grad_norm": 0.431640625, + "learning_rate": 0.0001883114094084216, + "loss": 2.6305, + "step": 7380 + }, + { + "epoch": 0.31, + "grad_norm": 0.4140625, + "learning_rate": 0.0001883082621280829, + "loss": 2.6915, + "step": 7381 + }, + { + "epoch": 0.31, + "grad_norm": 0.44140625, + "learning_rate": 0.0001883051144503893, + "loss": 2.6224, + "step": 7382 + }, + { + "epoch": 0.31, + "grad_norm": 0.53515625, + "learning_rate": 0.00018830196637535494, + "loss": 2.5509, + "step": 7383 + }, + { + "epoch": 0.31, + "grad_norm": 0.4453125, + "learning_rate": 0.000188298817902994, + "loss": 2.7241, + "step": 7384 + }, + { + "epoch": 0.31, + "grad_norm": 0.451171875, + "learning_rate": 0.00018829566903332066, + "loss": 2.719, + "step": 7385 + }, + { + "epoch": 0.31, + "grad_norm": 0.439453125, + "learning_rate": 0.00018829251976634908, + "loss": 2.7144, + "step": 7386 + }, + { + "epoch": 0.31, + "grad_norm": 0.423828125, + "learning_rate": 0.0001882893701020934, + "loss": 2.5797, + "step": 7387 + }, + { + "epoch": 0.31, + "grad_norm": 0.474609375, + "learning_rate": 0.00018828622004056787, + "loss": 2.6644, + "step": 7388 + }, + { + "epoch": 0.31, + "grad_norm": 0.412109375, + "learning_rate": 0.0001882830695817866, + "loss": 2.5921, + "step": 7389 + }, + { + "epoch": 0.31, + "grad_norm": 0.41015625, + "learning_rate": 0.00018827991872576375, + "loss": 2.598, + "step": 7390 + }, + { + "epoch": 0.31, + "grad_norm": 0.431640625, + "learning_rate": 0.00018827676747251354, + "loss": 2.6203, + "step": 7391 + }, + { + "epoch": 0.31, + "grad_norm": 0.39453125, + "learning_rate": 0.00018827361582205015, + "loss": 2.5556, + "step": 7392 + }, + { + "epoch": 0.32, + "grad_norm": 0.451171875, + "learning_rate": 0.00018827046377438775, + "loss": 2.48, + "step": 7393 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.0001882673113295405, + "loss": 2.6358, + "step": 7394 + }, + { + "epoch": 0.32, + "grad_norm": 0.392578125, + "learning_rate": 0.00018826415848752266, + "loss": 2.6292, + "step": 7395 + }, + { + "epoch": 0.32, + "grad_norm": 0.431640625, + "learning_rate": 0.0001882610052483483, + "loss": 2.5725, + "step": 7396 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.0001882578516120317, + "loss": 2.6525, + "step": 7397 + }, + { + "epoch": 0.32, + "grad_norm": 0.447265625, + "learning_rate": 0.00018825469757858705, + "loss": 2.7232, + "step": 7398 + }, + { + "epoch": 0.32, + "grad_norm": 0.39453125, + "learning_rate": 0.00018825154314802846, + "loss": 2.6215, + "step": 7399 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.0001882483883203702, + "loss": 2.6861, + "step": 7400 + }, + { + "epoch": 0.32, + "grad_norm": 0.435546875, + "learning_rate": 0.00018824523309562646, + "loss": 2.599, + "step": 7401 + }, + { + "epoch": 0.32, + "grad_norm": 0.4453125, + "learning_rate": 0.00018824207747381141, + "loss": 2.5815, + "step": 7402 + }, + { + "epoch": 0.32, + "grad_norm": 0.44921875, + "learning_rate": 0.00018823892145493926, + "loss": 2.5092, + "step": 7403 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.0001882357650390242, + "loss": 2.5718, + "step": 7404 + }, + { + "epoch": 0.32, + "grad_norm": 0.470703125, + "learning_rate": 0.00018823260822608046, + "loss": 2.6209, + "step": 7405 + }, + { + "epoch": 0.32, + "grad_norm": 0.48828125, + "learning_rate": 0.00018822945101612223, + "loss": 2.7179, + "step": 7406 + }, + { + "epoch": 0.32, + "grad_norm": 0.48046875, + "learning_rate": 0.00018822629340916372, + "loss": 2.6973, + "step": 7407 + }, + { + "epoch": 0.32, + "grad_norm": 0.451171875, + "learning_rate": 0.00018822313540521913, + "loss": 2.6547, + "step": 7408 + }, + { + "epoch": 0.32, + "grad_norm": 0.4609375, + "learning_rate": 0.00018821997700430262, + "loss": 2.4975, + "step": 7409 + }, + { + "epoch": 0.32, + "grad_norm": 0.4609375, + "learning_rate": 0.0001882168182064285, + "loss": 2.686, + "step": 7410 + }, + { + "epoch": 0.32, + "grad_norm": 0.439453125, + "learning_rate": 0.00018821365901161094, + "loss": 2.7609, + "step": 7411 + }, + { + "epoch": 0.32, + "grad_norm": 0.474609375, + "learning_rate": 0.00018821049941986418, + "loss": 2.7323, + "step": 7412 + }, + { + "epoch": 0.32, + "grad_norm": 0.41796875, + "learning_rate": 0.00018820733943120238, + "loss": 2.736, + "step": 7413 + }, + { + "epoch": 0.32, + "grad_norm": 0.48828125, + "learning_rate": 0.00018820417904563975, + "loss": 2.774, + "step": 7414 + }, + { + "epoch": 0.32, + "grad_norm": 0.486328125, + "learning_rate": 0.00018820101826319058, + "loss": 2.6141, + "step": 7415 + }, + { + "epoch": 0.32, + "grad_norm": 0.453125, + "learning_rate": 0.00018819785708386907, + "loss": 2.6487, + "step": 7416 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.00018819469550768945, + "loss": 2.6793, + "step": 7417 + }, + { + "epoch": 0.32, + "grad_norm": 0.48046875, + "learning_rate": 0.00018819153353466593, + "loss": 2.5581, + "step": 7418 + }, + { + "epoch": 0.32, + "grad_norm": 0.490234375, + "learning_rate": 0.0001881883711648127, + "loss": 2.5422, + "step": 7419 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.00018818520839814404, + "loss": 2.4373, + "step": 7420 + }, + { + "epoch": 0.32, + "grad_norm": 0.4765625, + "learning_rate": 0.0001881820452346742, + "loss": 2.54, + "step": 7421 + }, + { + "epoch": 0.32, + "grad_norm": 0.478515625, + "learning_rate": 0.00018817888167441734, + "loss": 2.8225, + "step": 7422 + }, + { + "epoch": 0.32, + "grad_norm": 0.453125, + "learning_rate": 0.00018817571771738778, + "loss": 2.578, + "step": 7423 + }, + { + "epoch": 0.32, + "grad_norm": 0.44140625, + "learning_rate": 0.00018817255336359972, + "loss": 2.6022, + "step": 7424 + }, + { + "epoch": 0.32, + "grad_norm": 0.435546875, + "learning_rate": 0.00018816938861306737, + "loss": 2.5965, + "step": 7425 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.000188166223465805, + "loss": 2.6603, + "step": 7426 + }, + { + "epoch": 0.32, + "grad_norm": 0.470703125, + "learning_rate": 0.00018816305792182682, + "loss": 2.7487, + "step": 7427 + }, + { + "epoch": 0.32, + "grad_norm": 0.4140625, + "learning_rate": 0.00018815989198114715, + "loss": 2.6713, + "step": 7428 + }, + { + "epoch": 0.32, + "grad_norm": 0.455078125, + "learning_rate": 0.00018815672564378012, + "loss": 2.6513, + "step": 7429 + }, + { + "epoch": 0.32, + "grad_norm": 0.44140625, + "learning_rate": 0.0001881535589097401, + "loss": 2.6737, + "step": 7430 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.00018815039177904125, + "loss": 2.5584, + "step": 7431 + }, + { + "epoch": 0.32, + "grad_norm": 0.5234375, + "learning_rate": 0.00018814722425169786, + "loss": 2.7874, + "step": 7432 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.00018814405632772418, + "loss": 2.6761, + "step": 7433 + }, + { + "epoch": 0.32, + "grad_norm": 0.451171875, + "learning_rate": 0.00018814088800713446, + "loss": 2.4884, + "step": 7434 + }, + { + "epoch": 0.32, + "grad_norm": 0.4296875, + "learning_rate": 0.00018813771928994296, + "loss": 2.5925, + "step": 7435 + }, + { + "epoch": 0.32, + "grad_norm": 0.48046875, + "learning_rate": 0.00018813455017616392, + "loss": 2.6504, + "step": 7436 + }, + { + "epoch": 0.32, + "grad_norm": 0.412109375, + "learning_rate": 0.0001881313806658116, + "loss": 2.5974, + "step": 7437 + }, + { + "epoch": 0.32, + "grad_norm": 0.470703125, + "learning_rate": 0.0001881282107589003, + "loss": 2.7064, + "step": 7438 + }, + { + "epoch": 0.32, + "grad_norm": 0.44140625, + "learning_rate": 0.00018812504045544426, + "loss": 2.5874, + "step": 7439 + }, + { + "epoch": 0.32, + "grad_norm": 0.451171875, + "learning_rate": 0.00018812186975545773, + "loss": 2.6975, + "step": 7440 + }, + { + "epoch": 0.32, + "grad_norm": 0.431640625, + "learning_rate": 0.00018811869865895502, + "loss": 2.5536, + "step": 7441 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.00018811552716595033, + "loss": 2.61, + "step": 7442 + }, + { + "epoch": 0.32, + "grad_norm": 0.439453125, + "learning_rate": 0.00018811235527645798, + "loss": 2.6089, + "step": 7443 + }, + { + "epoch": 0.32, + "grad_norm": 0.458984375, + "learning_rate": 0.00018810918299049225, + "loss": 2.6968, + "step": 7444 + }, + { + "epoch": 0.32, + "grad_norm": 0.416015625, + "learning_rate": 0.0001881060103080674, + "loss": 2.5571, + "step": 7445 + }, + { + "epoch": 0.32, + "grad_norm": 0.4453125, + "learning_rate": 0.00018810283722919767, + "loss": 2.7729, + "step": 7446 + }, + { + "epoch": 0.32, + "grad_norm": 0.490234375, + "learning_rate": 0.0001880996637538974, + "loss": 2.7205, + "step": 7447 + }, + { + "epoch": 0.32, + "grad_norm": 0.5078125, + "learning_rate": 0.00018809648988218084, + "loss": 2.6636, + "step": 7448 + }, + { + "epoch": 0.32, + "grad_norm": 0.486328125, + "learning_rate": 0.00018809331561406225, + "loss": 2.6072, + "step": 7449 + }, + { + "epoch": 0.32, + "grad_norm": 0.423828125, + "learning_rate": 0.00018809014094955595, + "loss": 2.5474, + "step": 7450 + }, + { + "epoch": 0.32, + "grad_norm": 0.39453125, + "learning_rate": 0.00018808696588867618, + "loss": 2.6084, + "step": 7451 + }, + { + "epoch": 0.32, + "grad_norm": 0.5078125, + "learning_rate": 0.00018808379043143728, + "loss": 2.6862, + "step": 7452 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.00018808061457785354, + "loss": 2.6286, + "step": 7453 + }, + { + "epoch": 0.32, + "grad_norm": 0.462890625, + "learning_rate": 0.0001880774383279392, + "loss": 2.692, + "step": 7454 + }, + { + "epoch": 0.32, + "grad_norm": 0.44921875, + "learning_rate": 0.00018807426168170857, + "loss": 2.5574, + "step": 7455 + }, + { + "epoch": 0.32, + "grad_norm": 0.4609375, + "learning_rate": 0.00018807108463917595, + "loss": 2.6348, + "step": 7456 + }, + { + "epoch": 0.32, + "grad_norm": 0.52734375, + "learning_rate": 0.00018806790720035562, + "loss": 2.6916, + "step": 7457 + }, + { + "epoch": 0.32, + "grad_norm": 0.435546875, + "learning_rate": 0.00018806472936526195, + "loss": 2.6245, + "step": 7458 + }, + { + "epoch": 0.32, + "grad_norm": 0.423828125, + "learning_rate": 0.00018806155113390913, + "loss": 2.6264, + "step": 7459 + }, + { + "epoch": 0.32, + "grad_norm": 0.515625, + "learning_rate": 0.0001880583725063115, + "loss": 2.5925, + "step": 7460 + }, + { + "epoch": 0.32, + "grad_norm": 0.5078125, + "learning_rate": 0.0001880551934824834, + "loss": 2.6608, + "step": 7461 + }, + { + "epoch": 0.32, + "grad_norm": 0.421875, + "learning_rate": 0.00018805201406243914, + "loss": 2.5728, + "step": 7462 + }, + { + "epoch": 0.32, + "grad_norm": 0.46484375, + "learning_rate": 0.00018804883424619297, + "loss": 2.6533, + "step": 7463 + }, + { + "epoch": 0.32, + "grad_norm": 0.470703125, + "learning_rate": 0.00018804565403375922, + "loss": 2.6434, + "step": 7464 + }, + { + "epoch": 0.32, + "grad_norm": 0.451171875, + "learning_rate": 0.0001880424734251522, + "loss": 2.5073, + "step": 7465 + }, + { + "epoch": 0.32, + "grad_norm": 0.458984375, + "learning_rate": 0.00018803929242038625, + "loss": 2.5414, + "step": 7466 + }, + { + "epoch": 0.32, + "grad_norm": 0.52734375, + "learning_rate": 0.00018803611101947564, + "loss": 2.7794, + "step": 7467 + }, + { + "epoch": 0.32, + "grad_norm": 0.419921875, + "learning_rate": 0.00018803292922243468, + "loss": 2.6403, + "step": 7468 + }, + { + "epoch": 0.32, + "grad_norm": 0.431640625, + "learning_rate": 0.00018802974702927775, + "loss": 2.7725, + "step": 7469 + }, + { + "epoch": 0.32, + "grad_norm": 0.41796875, + "learning_rate": 0.0001880265644400191, + "loss": 2.7209, + "step": 7470 + }, + { + "epoch": 0.32, + "grad_norm": 0.400390625, + "learning_rate": 0.0001880233814546731, + "loss": 2.5849, + "step": 7471 + }, + { + "epoch": 0.32, + "grad_norm": 0.431640625, + "learning_rate": 0.00018802019807325407, + "loss": 2.6627, + "step": 7472 + }, + { + "epoch": 0.32, + "grad_norm": 0.458984375, + "learning_rate": 0.00018801701429577627, + "loss": 2.7056, + "step": 7473 + }, + { + "epoch": 0.32, + "grad_norm": 0.439453125, + "learning_rate": 0.0001880138301222541, + "loss": 2.5657, + "step": 7474 + }, + { + "epoch": 0.32, + "grad_norm": 0.486328125, + "learning_rate": 0.00018801064555270185, + "loss": 2.6094, + "step": 7475 + }, + { + "epoch": 0.32, + "grad_norm": 0.47265625, + "learning_rate": 0.00018800746058713385, + "loss": 2.5824, + "step": 7476 + }, + { + "epoch": 0.32, + "grad_norm": 0.4609375, + "learning_rate": 0.00018800427522556448, + "loss": 2.7332, + "step": 7477 + }, + { + "epoch": 0.32, + "grad_norm": 0.48828125, + "learning_rate": 0.000188001089468008, + "loss": 2.609, + "step": 7478 + }, + { + "epoch": 0.32, + "grad_norm": 0.447265625, + "learning_rate": 0.0001879979033144788, + "loss": 2.6012, + "step": 7479 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.0001879947167649912, + "loss": 2.6, + "step": 7480 + }, + { + "epoch": 0.32, + "grad_norm": 0.515625, + "learning_rate": 0.0001879915298195595, + "loss": 2.7761, + "step": 7481 + }, + { + "epoch": 0.32, + "grad_norm": 0.470703125, + "learning_rate": 0.00018798834247819808, + "loss": 2.5864, + "step": 7482 + }, + { + "epoch": 0.32, + "grad_norm": 0.482421875, + "learning_rate": 0.00018798515474092127, + "loss": 2.6455, + "step": 7483 + }, + { + "epoch": 0.32, + "grad_norm": 0.46875, + "learning_rate": 0.00018798196660774343, + "loss": 2.7493, + "step": 7484 + }, + { + "epoch": 0.32, + "grad_norm": 0.4921875, + "learning_rate": 0.0001879787780786789, + "loss": 2.5809, + "step": 7485 + }, + { + "epoch": 0.32, + "grad_norm": 0.43359375, + "learning_rate": 0.00018797558915374202, + "loss": 2.6322, + "step": 7486 + }, + { + "epoch": 0.32, + "grad_norm": 0.474609375, + "learning_rate": 0.00018797239983294714, + "loss": 2.6003, + "step": 7487 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.0001879692101163086, + "loss": 2.6353, + "step": 7488 + }, + { + "epoch": 0.32, + "grad_norm": 0.4453125, + "learning_rate": 0.00018796602000384076, + "loss": 2.5545, + "step": 7489 + }, + { + "epoch": 0.32, + "grad_norm": 0.46875, + "learning_rate": 0.000187962829495558, + "loss": 2.6595, + "step": 7490 + }, + { + "epoch": 0.32, + "grad_norm": 0.44140625, + "learning_rate": 0.00018795963859147464, + "loss": 2.5943, + "step": 7491 + }, + { + "epoch": 0.32, + "grad_norm": 0.43359375, + "learning_rate": 0.00018795644729160507, + "loss": 2.5906, + "step": 7492 + }, + { + "epoch": 0.32, + "grad_norm": 0.419921875, + "learning_rate": 0.0001879532555959636, + "loss": 2.5652, + "step": 7493 + }, + { + "epoch": 0.32, + "grad_norm": 0.4765625, + "learning_rate": 0.00018795006350456466, + "loss": 2.6812, + "step": 7494 + }, + { + "epoch": 0.32, + "grad_norm": 0.451171875, + "learning_rate": 0.00018794687101742255, + "loss": 2.7314, + "step": 7495 + }, + { + "epoch": 0.32, + "grad_norm": 0.490234375, + "learning_rate": 0.00018794367813455165, + "loss": 2.7551, + "step": 7496 + }, + { + "epoch": 0.32, + "grad_norm": 0.447265625, + "learning_rate": 0.00018794048485596635, + "loss": 2.6809, + "step": 7497 + }, + { + "epoch": 0.32, + "grad_norm": 0.453125, + "learning_rate": 0.000187937291181681, + "loss": 2.5828, + "step": 7498 + }, + { + "epoch": 0.32, + "grad_norm": 0.458984375, + "learning_rate": 0.00018793409711171, + "loss": 2.5995, + "step": 7499 + }, + { + "epoch": 0.32, + "grad_norm": 0.5234375, + "learning_rate": 0.0001879309026460677, + "loss": 2.6611, + "step": 7500 + }, + { + "epoch": 0.32, + "grad_norm": 0.45703125, + "learning_rate": 0.00018792770778476844, + "loss": 2.7069, + "step": 7501 + }, + { + "epoch": 0.32, + "grad_norm": 0.41796875, + "learning_rate": 0.00018792451252782667, + "loss": 2.6362, + "step": 7502 + }, + { + "epoch": 0.32, + "grad_norm": 0.42578125, + "learning_rate": 0.0001879213168752567, + "loss": 2.6042, + "step": 7503 + }, + { + "epoch": 0.32, + "grad_norm": 0.400390625, + "learning_rate": 0.00018791812082707293, + "loss": 2.4734, + "step": 7504 + }, + { + "epoch": 0.32, + "grad_norm": 0.400390625, + "learning_rate": 0.00018791492438328977, + "loss": 2.5843, + "step": 7505 + }, + { + "epoch": 0.32, + "grad_norm": 0.404296875, + "learning_rate": 0.00018791172754392156, + "loss": 2.5915, + "step": 7506 + }, + { + "epoch": 0.32, + "grad_norm": 0.439453125, + "learning_rate": 0.0001879085303089827, + "loss": 2.7088, + "step": 7507 + }, + { + "epoch": 0.32, + "grad_norm": 0.44921875, + "learning_rate": 0.0001879053326784876, + "loss": 2.5194, + "step": 7508 + }, + { + "epoch": 0.32, + "grad_norm": 0.5234375, + "learning_rate": 0.00018790213465245064, + "loss": 2.8223, + "step": 7509 + }, + { + "epoch": 0.32, + "grad_norm": 0.490234375, + "learning_rate": 0.0001878989362308862, + "loss": 2.7366, + "step": 7510 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.0001878957374138086, + "loss": 2.6903, + "step": 7511 + }, + { + "epoch": 0.32, + "grad_norm": 0.400390625, + "learning_rate": 0.00018789253820123238, + "loss": 2.7246, + "step": 7512 + }, + { + "epoch": 0.32, + "grad_norm": 0.4296875, + "learning_rate": 0.00018788933859317184, + "loss": 2.5925, + "step": 7513 + }, + { + "epoch": 0.32, + "grad_norm": 0.421875, + "learning_rate": 0.00018788613858964137, + "loss": 2.5808, + "step": 7514 + }, + { + "epoch": 0.32, + "grad_norm": 0.4609375, + "learning_rate": 0.00018788293819065543, + "loss": 2.5715, + "step": 7515 + }, + { + "epoch": 0.32, + "grad_norm": 0.384765625, + "learning_rate": 0.00018787973739622837, + "loss": 2.6395, + "step": 7516 + }, + { + "epoch": 0.32, + "grad_norm": 0.44140625, + "learning_rate": 0.00018787653620637462, + "loss": 2.7224, + "step": 7517 + }, + { + "epoch": 0.32, + "grad_norm": 0.453125, + "learning_rate": 0.00018787333462110857, + "loss": 2.6602, + "step": 7518 + }, + { + "epoch": 0.32, + "grad_norm": 0.4609375, + "learning_rate": 0.00018787013264044458, + "loss": 2.6215, + "step": 7519 + }, + { + "epoch": 0.32, + "grad_norm": 0.48046875, + "learning_rate": 0.00018786693026439716, + "loss": 2.6827, + "step": 7520 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.00018786372749298065, + "loss": 2.6281, + "step": 7521 + }, + { + "epoch": 0.32, + "grad_norm": 0.47265625, + "learning_rate": 0.0001878605243262095, + "loss": 2.4414, + "step": 7522 + }, + { + "epoch": 0.32, + "grad_norm": 0.412109375, + "learning_rate": 0.00018785732076409805, + "loss": 2.5986, + "step": 7523 + }, + { + "epoch": 0.32, + "grad_norm": 0.44140625, + "learning_rate": 0.0001878541168066608, + "loss": 2.5586, + "step": 7524 + }, + { + "epoch": 0.32, + "grad_norm": 0.435546875, + "learning_rate": 0.0001878509124539121, + "loss": 2.636, + "step": 7525 + }, + { + "epoch": 0.32, + "grad_norm": 0.48828125, + "learning_rate": 0.00018784770770586645, + "loss": 2.7677, + "step": 7526 + }, + { + "epoch": 0.32, + "grad_norm": 0.484375, + "learning_rate": 0.0001878445025625382, + "loss": 2.7038, + "step": 7527 + }, + { + "epoch": 0.32, + "grad_norm": 0.4296875, + "learning_rate": 0.00018784129702394177, + "loss": 2.6893, + "step": 7528 + }, + { + "epoch": 0.32, + "grad_norm": 0.419921875, + "learning_rate": 0.00018783809109009162, + "loss": 2.693, + "step": 7529 + }, + { + "epoch": 0.32, + "grad_norm": 0.453125, + "learning_rate": 0.00018783488476100218, + "loss": 2.8024, + "step": 7530 + }, + { + "epoch": 0.32, + "grad_norm": 0.423828125, + "learning_rate": 0.00018783167803668782, + "loss": 2.6516, + "step": 7531 + }, + { + "epoch": 0.32, + "grad_norm": 0.458984375, + "learning_rate": 0.00018782847091716303, + "loss": 2.7524, + "step": 7532 + }, + { + "epoch": 0.32, + "grad_norm": 0.42578125, + "learning_rate": 0.00018782526340244222, + "loss": 2.4962, + "step": 7533 + }, + { + "epoch": 0.32, + "grad_norm": 0.435546875, + "learning_rate": 0.00018782205549253982, + "loss": 2.7154, + "step": 7534 + }, + { + "epoch": 0.32, + "grad_norm": 0.462890625, + "learning_rate": 0.00018781884718747027, + "loss": 2.6001, + "step": 7535 + }, + { + "epoch": 0.32, + "grad_norm": 0.392578125, + "learning_rate": 0.000187815638487248, + "loss": 2.7159, + "step": 7536 + }, + { + "epoch": 0.32, + "grad_norm": 0.435546875, + "learning_rate": 0.00018781242939188743, + "loss": 2.6996, + "step": 7537 + }, + { + "epoch": 0.32, + "grad_norm": 0.396484375, + "learning_rate": 0.00018780921990140304, + "loss": 2.5475, + "step": 7538 + }, + { + "epoch": 0.32, + "grad_norm": 0.44921875, + "learning_rate": 0.00018780601001580925, + "loss": 2.6419, + "step": 7539 + }, + { + "epoch": 0.32, + "grad_norm": 0.455078125, + "learning_rate": 0.00018780279973512048, + "loss": 2.5493, + "step": 7540 + }, + { + "epoch": 0.32, + "grad_norm": 0.453125, + "learning_rate": 0.00018779958905935123, + "loss": 2.6573, + "step": 7541 + }, + { + "epoch": 0.32, + "grad_norm": 0.4453125, + "learning_rate": 0.0001877963779885159, + "loss": 2.6812, + "step": 7542 + }, + { + "epoch": 0.32, + "grad_norm": 0.439453125, + "learning_rate": 0.00018779316652262895, + "loss": 2.6714, + "step": 7543 + }, + { + "epoch": 0.32, + "grad_norm": 0.419921875, + "learning_rate": 0.00018778995466170486, + "loss": 2.7997, + "step": 7544 + }, + { + "epoch": 0.32, + "grad_norm": 0.45703125, + "learning_rate": 0.00018778674240575805, + "loss": 2.5761, + "step": 7545 + }, + { + "epoch": 0.32, + "grad_norm": 0.46484375, + "learning_rate": 0.00018778352975480295, + "loss": 2.6816, + "step": 7546 + }, + { + "epoch": 0.32, + "grad_norm": 0.419921875, + "learning_rate": 0.0001877803167088541, + "loss": 2.7064, + "step": 7547 + }, + { + "epoch": 0.32, + "grad_norm": 0.4296875, + "learning_rate": 0.00018777710326792587, + "loss": 2.6739, + "step": 7548 + }, + { + "epoch": 0.32, + "grad_norm": 0.46875, + "learning_rate": 0.00018777388943203276, + "loss": 2.6682, + "step": 7549 + }, + { + "epoch": 0.32, + "grad_norm": 0.48828125, + "learning_rate": 0.00018777067520118923, + "loss": 2.6642, + "step": 7550 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.0001877674605754097, + "loss": 2.5965, + "step": 7551 + }, + { + "epoch": 0.32, + "grad_norm": 0.486328125, + "learning_rate": 0.00018776424555470874, + "loss": 2.6037, + "step": 7552 + }, + { + "epoch": 0.32, + "grad_norm": 0.51953125, + "learning_rate": 0.0001877610301391007, + "loss": 2.6263, + "step": 7553 + }, + { + "epoch": 0.32, + "grad_norm": 0.447265625, + "learning_rate": 0.00018775781432860011, + "loss": 2.791, + "step": 7554 + }, + { + "epoch": 0.32, + "grad_norm": 0.458984375, + "learning_rate": 0.00018775459812322143, + "loss": 2.6221, + "step": 7555 + }, + { + "epoch": 0.32, + "grad_norm": 0.51953125, + "learning_rate": 0.00018775138152297914, + "loss": 2.6903, + "step": 7556 + }, + { + "epoch": 0.32, + "grad_norm": 0.41796875, + "learning_rate": 0.0001877481645278877, + "loss": 2.562, + "step": 7557 + }, + { + "epoch": 0.32, + "grad_norm": 0.484375, + "learning_rate": 0.0001877449471379616, + "loss": 2.5638, + "step": 7558 + }, + { + "epoch": 0.32, + "grad_norm": 0.44921875, + "learning_rate": 0.00018774172935321528, + "loss": 2.5708, + "step": 7559 + }, + { + "epoch": 0.32, + "grad_norm": 0.46484375, + "learning_rate": 0.00018773851117366322, + "loss": 2.5971, + "step": 7560 + }, + { + "epoch": 0.32, + "grad_norm": 0.53125, + "learning_rate": 0.00018773529259931998, + "loss": 2.7499, + "step": 7561 + }, + { + "epoch": 0.32, + "grad_norm": 0.447265625, + "learning_rate": 0.00018773207363019994, + "loss": 2.5868, + "step": 7562 + }, + { + "epoch": 0.32, + "grad_norm": 0.482421875, + "learning_rate": 0.00018772885426631767, + "loss": 2.614, + "step": 7563 + }, + { + "epoch": 0.32, + "grad_norm": 0.482421875, + "learning_rate": 0.0001877256345076876, + "loss": 2.6769, + "step": 7564 + }, + { + "epoch": 0.32, + "grad_norm": 0.48046875, + "learning_rate": 0.00018772241435432424, + "loss": 2.69, + "step": 7565 + }, + { + "epoch": 0.32, + "grad_norm": 0.439453125, + "learning_rate": 0.00018771919380624206, + "loss": 2.6875, + "step": 7566 + }, + { + "epoch": 0.32, + "grad_norm": 0.5078125, + "learning_rate": 0.00018771597286345557, + "loss": 2.6788, + "step": 7567 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.00018771275152597926, + "loss": 2.6575, + "step": 7568 + }, + { + "epoch": 0.32, + "grad_norm": 0.41796875, + "learning_rate": 0.00018770952979382764, + "loss": 2.5351, + "step": 7569 + }, + { + "epoch": 0.32, + "grad_norm": 0.41796875, + "learning_rate": 0.00018770630766701518, + "loss": 2.4311, + "step": 7570 + }, + { + "epoch": 0.32, + "grad_norm": 0.470703125, + "learning_rate": 0.00018770308514555638, + "loss": 2.7017, + "step": 7571 + }, + { + "epoch": 0.32, + "grad_norm": 0.46484375, + "learning_rate": 0.00018769986222946577, + "loss": 2.607, + "step": 7572 + }, + { + "epoch": 0.32, + "grad_norm": 0.45703125, + "learning_rate": 0.0001876966389187578, + "loss": 2.7734, + "step": 7573 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.00018769341521344702, + "loss": 2.5086, + "step": 7574 + }, + { + "epoch": 0.32, + "grad_norm": 0.431640625, + "learning_rate": 0.0001876901911135479, + "loss": 2.493, + "step": 7575 + }, + { + "epoch": 0.32, + "grad_norm": 0.48828125, + "learning_rate": 0.000187686966619075, + "loss": 2.6827, + "step": 7576 + }, + { + "epoch": 0.32, + "grad_norm": 0.416015625, + "learning_rate": 0.00018768374173004277, + "loss": 2.6936, + "step": 7577 + }, + { + "epoch": 0.32, + "grad_norm": 0.439453125, + "learning_rate": 0.00018768051644646577, + "loss": 2.5651, + "step": 7578 + }, + { + "epoch": 0.32, + "grad_norm": 0.46875, + "learning_rate": 0.00018767729076835847, + "loss": 2.5756, + "step": 7579 + }, + { + "epoch": 0.32, + "grad_norm": 0.451171875, + "learning_rate": 0.0001876740646957354, + "loss": 2.6044, + "step": 7580 + }, + { + "epoch": 0.32, + "grad_norm": 0.4453125, + "learning_rate": 0.0001876708382286111, + "loss": 2.5828, + "step": 7581 + }, + { + "epoch": 0.32, + "grad_norm": 0.404296875, + "learning_rate": 0.00018766761136700005, + "loss": 2.5246, + "step": 7582 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.00018766438411091683, + "loss": 2.6947, + "step": 7583 + }, + { + "epoch": 0.32, + "grad_norm": 0.3984375, + "learning_rate": 0.00018766115646037584, + "loss": 2.6568, + "step": 7584 + }, + { + "epoch": 0.32, + "grad_norm": 0.41015625, + "learning_rate": 0.00018765792841539175, + "loss": 2.5376, + "step": 7585 + }, + { + "epoch": 0.32, + "grad_norm": 0.490234375, + "learning_rate": 0.000187654699975979, + "loss": 2.6618, + "step": 7586 + }, + { + "epoch": 0.32, + "grad_norm": 0.447265625, + "learning_rate": 0.00018765147114215214, + "loss": 2.6782, + "step": 7587 + }, + { + "epoch": 0.32, + "grad_norm": 0.494140625, + "learning_rate": 0.0001876482419139257, + "loss": 2.6369, + "step": 7588 + }, + { + "epoch": 0.32, + "grad_norm": 0.45703125, + "learning_rate": 0.00018764501229131416, + "loss": 2.6129, + "step": 7589 + }, + { + "epoch": 0.32, + "grad_norm": 0.388671875, + "learning_rate": 0.00018764178227433212, + "loss": 2.6519, + "step": 7590 + }, + { + "epoch": 0.32, + "grad_norm": 0.494140625, + "learning_rate": 0.00018763855186299409, + "loss": 2.593, + "step": 7591 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.0001876353210573146, + "loss": 2.6235, + "step": 7592 + }, + { + "epoch": 0.32, + "grad_norm": 0.46484375, + "learning_rate": 0.0001876320898573082, + "loss": 2.6307, + "step": 7593 + }, + { + "epoch": 0.32, + "grad_norm": 0.470703125, + "learning_rate": 0.00018762885826298945, + "loss": 2.6033, + "step": 7594 + }, + { + "epoch": 0.32, + "grad_norm": 0.458984375, + "learning_rate": 0.00018762562627437282, + "loss": 2.592, + "step": 7595 + }, + { + "epoch": 0.32, + "grad_norm": 0.44140625, + "learning_rate": 0.0001876223938914729, + "loss": 2.6612, + "step": 7596 + }, + { + "epoch": 0.32, + "grad_norm": 0.431640625, + "learning_rate": 0.00018761916111430425, + "loss": 2.6737, + "step": 7597 + }, + { + "epoch": 0.32, + "grad_norm": 0.470703125, + "learning_rate": 0.00018761592794288136, + "loss": 2.7669, + "step": 7598 + }, + { + "epoch": 0.32, + "grad_norm": 0.390625, + "learning_rate": 0.00018761269437721884, + "loss": 2.6045, + "step": 7599 + }, + { + "epoch": 0.32, + "grad_norm": 0.51953125, + "learning_rate": 0.0001876094604173312, + "loss": 2.5781, + "step": 7600 + }, + { + "epoch": 0.32, + "grad_norm": 0.4453125, + "learning_rate": 0.000187606226063233, + "loss": 2.6515, + "step": 7601 + }, + { + "epoch": 0.32, + "grad_norm": 0.419921875, + "learning_rate": 0.00018760299131493887, + "loss": 2.5714, + "step": 7602 + }, + { + "epoch": 0.32, + "grad_norm": 0.447265625, + "learning_rate": 0.0001875997561724632, + "loss": 2.6164, + "step": 7603 + }, + { + "epoch": 0.32, + "grad_norm": 0.419921875, + "learning_rate": 0.0001875965206358207, + "loss": 2.5603, + "step": 7604 + }, + { + "epoch": 0.32, + "grad_norm": 0.46875, + "learning_rate": 0.00018759328470502586, + "loss": 2.5954, + "step": 7605 + }, + { + "epoch": 0.32, + "grad_norm": 0.431640625, + "learning_rate": 0.00018759004838009324, + "loss": 2.6949, + "step": 7606 + }, + { + "epoch": 0.32, + "grad_norm": 0.486328125, + "learning_rate": 0.0001875868116610374, + "loss": 2.5964, + "step": 7607 + }, + { + "epoch": 0.32, + "grad_norm": 0.66015625, + "learning_rate": 0.00018758357454787294, + "loss": 2.6483, + "step": 7608 + }, + { + "epoch": 0.32, + "grad_norm": 0.47265625, + "learning_rate": 0.0001875803370406144, + "loss": 2.7603, + "step": 7609 + }, + { + "epoch": 0.32, + "grad_norm": 0.453125, + "learning_rate": 0.00018757709913927635, + "loss": 2.6608, + "step": 7610 + }, + { + "epoch": 0.32, + "grad_norm": 0.40234375, + "learning_rate": 0.00018757386084387333, + "loss": 2.554, + "step": 7611 + }, + { + "epoch": 0.32, + "grad_norm": 0.462890625, + "learning_rate": 0.00018757062215442, + "loss": 2.6121, + "step": 7612 + }, + { + "epoch": 0.32, + "grad_norm": 0.482421875, + "learning_rate": 0.00018756738307093083, + "loss": 2.7647, + "step": 7613 + }, + { + "epoch": 0.32, + "grad_norm": 0.447265625, + "learning_rate": 0.00018756414359342042, + "loss": 2.5768, + "step": 7614 + }, + { + "epoch": 0.32, + "grad_norm": 0.44140625, + "learning_rate": 0.0001875609037219034, + "loss": 2.6277, + "step": 7615 + }, + { + "epoch": 0.32, + "grad_norm": 0.427734375, + "learning_rate": 0.00018755766345639432, + "loss": 2.718, + "step": 7616 + }, + { + "epoch": 0.32, + "grad_norm": 0.44921875, + "learning_rate": 0.00018755442279690774, + "loss": 2.6575, + "step": 7617 + }, + { + "epoch": 0.32, + "grad_norm": 0.5078125, + "learning_rate": 0.00018755118174345825, + "loss": 2.6318, + "step": 7618 + }, + { + "epoch": 0.32, + "grad_norm": 0.4140625, + "learning_rate": 0.00018754794029606045, + "loss": 2.5763, + "step": 7619 + }, + { + "epoch": 0.32, + "grad_norm": 0.4375, + "learning_rate": 0.00018754469845472891, + "loss": 2.429, + "step": 7620 + }, + { + "epoch": 0.32, + "grad_norm": 0.5390625, + "learning_rate": 0.0001875414562194782, + "loss": 2.7089, + "step": 7621 + }, + { + "epoch": 0.32, + "grad_norm": 0.42578125, + "learning_rate": 0.00018753821359032295, + "loss": 2.6467, + "step": 7622 + }, + { + "epoch": 0.32, + "grad_norm": 0.443359375, + "learning_rate": 0.00018753497056727772, + "loss": 2.6472, + "step": 7623 + }, + { + "epoch": 0.32, + "grad_norm": 0.419921875, + "learning_rate": 0.00018753172715035713, + "loss": 2.5131, + "step": 7624 + }, + { + "epoch": 0.32, + "grad_norm": 0.42578125, + "learning_rate": 0.00018752848333957576, + "loss": 2.6937, + "step": 7625 + }, + { + "epoch": 0.32, + "grad_norm": 0.5078125, + "learning_rate": 0.00018752523913494816, + "loss": 2.8542, + "step": 7626 + }, + { + "epoch": 0.33, + "grad_norm": 0.484375, + "learning_rate": 0.00018752199453648903, + "loss": 2.7618, + "step": 7627 + }, + { + "epoch": 0.33, + "grad_norm": 0.44921875, + "learning_rate": 0.00018751874954421287, + "loss": 2.6258, + "step": 7628 + }, + { + "epoch": 0.33, + "grad_norm": 0.416015625, + "learning_rate": 0.00018751550415813432, + "loss": 2.5558, + "step": 7629 + }, + { + "epoch": 0.33, + "grad_norm": 0.4140625, + "learning_rate": 0.000187512258378268, + "loss": 2.5538, + "step": 7630 + }, + { + "epoch": 0.33, + "grad_norm": 0.435546875, + "learning_rate": 0.00018750901220462849, + "loss": 2.6325, + "step": 7631 + }, + { + "epoch": 0.33, + "grad_norm": 0.396484375, + "learning_rate": 0.0001875057656372304, + "loss": 2.5883, + "step": 7632 + }, + { + "epoch": 0.33, + "grad_norm": 0.416015625, + "learning_rate": 0.00018750251867608836, + "loss": 2.6266, + "step": 7633 + }, + { + "epoch": 0.33, + "grad_norm": 0.373046875, + "learning_rate": 0.00018749927132121697, + "loss": 2.7689, + "step": 7634 + }, + { + "epoch": 0.33, + "grad_norm": 0.50390625, + "learning_rate": 0.00018749602357263085, + "loss": 2.883, + "step": 7635 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.00018749277543034457, + "loss": 2.652, + "step": 7636 + }, + { + "epoch": 0.33, + "grad_norm": 0.458984375, + "learning_rate": 0.00018748952689437277, + "loss": 2.6123, + "step": 7637 + }, + { + "epoch": 0.33, + "grad_norm": 0.39453125, + "learning_rate": 0.0001874862779647301, + "loss": 2.5992, + "step": 7638 + }, + { + "epoch": 0.33, + "grad_norm": 0.484375, + "learning_rate": 0.0001874830286414311, + "loss": 2.5965, + "step": 7639 + }, + { + "epoch": 0.33, + "grad_norm": 0.447265625, + "learning_rate": 0.00018747977892449053, + "loss": 2.6509, + "step": 7640 + }, + { + "epoch": 0.33, + "grad_norm": 0.380859375, + "learning_rate": 0.00018747652881392286, + "loss": 2.6467, + "step": 7641 + }, + { + "epoch": 0.33, + "grad_norm": 0.40625, + "learning_rate": 0.00018747327830974278, + "loss": 2.7453, + "step": 7642 + }, + { + "epoch": 0.33, + "grad_norm": 0.404296875, + "learning_rate": 0.00018747002741196494, + "loss": 2.608, + "step": 7643 + }, + { + "epoch": 0.33, + "grad_norm": 0.4765625, + "learning_rate": 0.00018746677612060394, + "loss": 2.586, + "step": 7644 + }, + { + "epoch": 0.33, + "grad_norm": 0.4609375, + "learning_rate": 0.00018746352443567438, + "loss": 2.6417, + "step": 7645 + }, + { + "epoch": 0.33, + "grad_norm": 0.36328125, + "learning_rate": 0.00018746027235719097, + "loss": 2.6364, + "step": 7646 + }, + { + "epoch": 0.33, + "grad_norm": 0.490234375, + "learning_rate": 0.00018745701988516825, + "loss": 2.6531, + "step": 7647 + }, + { + "epoch": 0.33, + "grad_norm": 0.455078125, + "learning_rate": 0.00018745376701962094, + "loss": 2.6186, + "step": 7648 + }, + { + "epoch": 0.33, + "grad_norm": 0.435546875, + "learning_rate": 0.00018745051376056364, + "loss": 2.5367, + "step": 7649 + }, + { + "epoch": 0.33, + "grad_norm": 0.447265625, + "learning_rate": 0.00018744726010801096, + "loss": 2.6975, + "step": 7650 + }, + { + "epoch": 0.33, + "grad_norm": 0.44921875, + "learning_rate": 0.0001874440060619776, + "loss": 2.6839, + "step": 7651 + }, + { + "epoch": 0.33, + "grad_norm": 0.42578125, + "learning_rate": 0.00018744075162247816, + "loss": 2.7226, + "step": 7652 + }, + { + "epoch": 0.33, + "grad_norm": 0.5078125, + "learning_rate": 0.00018743749678952728, + "loss": 2.7296, + "step": 7653 + }, + { + "epoch": 0.33, + "grad_norm": 0.42578125, + "learning_rate": 0.00018743424156313966, + "loss": 2.5889, + "step": 7654 + }, + { + "epoch": 0.33, + "grad_norm": 0.474609375, + "learning_rate": 0.00018743098594332987, + "loss": 2.5702, + "step": 7655 + }, + { + "epoch": 0.33, + "grad_norm": 0.421875, + "learning_rate": 0.0001874277299301126, + "loss": 2.5114, + "step": 7656 + }, + { + "epoch": 0.33, + "grad_norm": 0.458984375, + "learning_rate": 0.00018742447352350247, + "loss": 2.6663, + "step": 7657 + }, + { + "epoch": 0.33, + "grad_norm": 0.43359375, + "learning_rate": 0.00018742121672351417, + "loss": 2.5826, + "step": 7658 + }, + { + "epoch": 0.33, + "grad_norm": 0.453125, + "learning_rate": 0.00018741795953016238, + "loss": 2.5534, + "step": 7659 + }, + { + "epoch": 0.33, + "grad_norm": 0.443359375, + "learning_rate": 0.0001874147019434617, + "loss": 2.618, + "step": 7660 + }, + { + "epoch": 0.33, + "grad_norm": 0.48828125, + "learning_rate": 0.0001874114439634268, + "loss": 2.6344, + "step": 7661 + }, + { + "epoch": 0.33, + "grad_norm": 0.4609375, + "learning_rate": 0.00018740818559007235, + "loss": 2.6254, + "step": 7662 + }, + { + "epoch": 0.33, + "grad_norm": 0.46875, + "learning_rate": 0.00018740492682341303, + "loss": 2.649, + "step": 7663 + }, + { + "epoch": 0.33, + "grad_norm": 0.5625, + "learning_rate": 0.00018740166766346346, + "loss": 2.7838, + "step": 7664 + }, + { + "epoch": 0.33, + "grad_norm": 0.439453125, + "learning_rate": 0.0001873984081102383, + "loss": 2.5993, + "step": 7665 + }, + { + "epoch": 0.33, + "grad_norm": 0.47265625, + "learning_rate": 0.0001873951481637523, + "loss": 2.5913, + "step": 7666 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.00018739188782402005, + "loss": 2.771, + "step": 7667 + }, + { + "epoch": 0.33, + "grad_norm": 0.46484375, + "learning_rate": 0.00018738862709105622, + "loss": 2.5796, + "step": 7668 + }, + { + "epoch": 0.33, + "grad_norm": 0.423828125, + "learning_rate": 0.00018738536596487552, + "loss": 2.643, + "step": 7669 + }, + { + "epoch": 0.33, + "grad_norm": 0.4296875, + "learning_rate": 0.00018738210444549264, + "loss": 2.6143, + "step": 7670 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.0001873788425329222, + "loss": 2.6682, + "step": 7671 + }, + { + "epoch": 0.33, + "grad_norm": 0.400390625, + "learning_rate": 0.0001873755802271789, + "loss": 2.6154, + "step": 7672 + }, + { + "epoch": 0.33, + "grad_norm": 0.50390625, + "learning_rate": 0.00018737231752827743, + "loss": 2.5828, + "step": 7673 + }, + { + "epoch": 0.33, + "grad_norm": 0.400390625, + "learning_rate": 0.00018736905443623246, + "loss": 2.5863, + "step": 7674 + }, + { + "epoch": 0.33, + "grad_norm": 0.421875, + "learning_rate": 0.00018736579095105866, + "loss": 2.6006, + "step": 7675 + }, + { + "epoch": 0.33, + "grad_norm": 0.408203125, + "learning_rate": 0.00018736252707277074, + "loss": 2.6563, + "step": 7676 + }, + { + "epoch": 0.33, + "grad_norm": 0.4453125, + "learning_rate": 0.00018735926280138336, + "loss": 2.7017, + "step": 7677 + }, + { + "epoch": 0.33, + "grad_norm": 0.392578125, + "learning_rate": 0.00018735599813691126, + "loss": 2.6798, + "step": 7678 + }, + { + "epoch": 0.33, + "grad_norm": 0.42578125, + "learning_rate": 0.00018735273307936905, + "loss": 2.6734, + "step": 7679 + }, + { + "epoch": 0.33, + "grad_norm": 0.44921875, + "learning_rate": 0.0001873494676287715, + "loss": 2.6419, + "step": 7680 + }, + { + "epoch": 0.33, + "grad_norm": 0.421875, + "learning_rate": 0.00018734620178513324, + "loss": 2.7275, + "step": 7681 + }, + { + "epoch": 0.33, + "grad_norm": 0.40625, + "learning_rate": 0.000187342935548469, + "loss": 2.658, + "step": 7682 + }, + { + "epoch": 0.33, + "grad_norm": 0.453125, + "learning_rate": 0.00018733966891879347, + "loss": 2.4848, + "step": 7683 + }, + { + "epoch": 0.33, + "grad_norm": 0.494140625, + "learning_rate": 0.00018733640189612136, + "loss": 2.6536, + "step": 7684 + }, + { + "epoch": 0.33, + "grad_norm": 0.41796875, + "learning_rate": 0.00018733313448046737, + "loss": 2.6055, + "step": 7685 + }, + { + "epoch": 0.33, + "grad_norm": 0.4765625, + "learning_rate": 0.00018732986667184618, + "loss": 2.5585, + "step": 7686 + }, + { + "epoch": 0.33, + "grad_norm": 0.447265625, + "learning_rate": 0.00018732659847027253, + "loss": 2.6863, + "step": 7687 + }, + { + "epoch": 0.33, + "grad_norm": 0.412109375, + "learning_rate": 0.00018732332987576105, + "loss": 2.5678, + "step": 7688 + }, + { + "epoch": 0.33, + "grad_norm": 0.392578125, + "learning_rate": 0.00018732006088832651, + "loss": 2.6466, + "step": 7689 + }, + { + "epoch": 0.33, + "grad_norm": 0.423828125, + "learning_rate": 0.00018731679150798362, + "loss": 2.5986, + "step": 7690 + }, + { + "epoch": 0.33, + "grad_norm": 0.47265625, + "learning_rate": 0.00018731352173474708, + "loss": 2.6658, + "step": 7691 + }, + { + "epoch": 0.33, + "grad_norm": 0.443359375, + "learning_rate": 0.0001873102515686316, + "loss": 2.687, + "step": 7692 + }, + { + "epoch": 0.33, + "grad_norm": 0.470703125, + "learning_rate": 0.0001873069810096519, + "loss": 2.6728, + "step": 7693 + }, + { + "epoch": 0.33, + "grad_norm": 0.408203125, + "learning_rate": 0.00018730371005782267, + "loss": 2.6119, + "step": 7694 + }, + { + "epoch": 0.33, + "grad_norm": 0.423828125, + "learning_rate": 0.00018730043871315868, + "loss": 2.5231, + "step": 7695 + }, + { + "epoch": 0.33, + "grad_norm": 0.486328125, + "learning_rate": 0.00018729716697567464, + "loss": 2.6375, + "step": 7696 + }, + { + "epoch": 0.33, + "grad_norm": 0.482421875, + "learning_rate": 0.0001872938948453852, + "loss": 2.6404, + "step": 7697 + }, + { + "epoch": 0.33, + "grad_norm": 0.419921875, + "learning_rate": 0.00018729062232230516, + "loss": 2.5121, + "step": 7698 + }, + { + "epoch": 0.33, + "grad_norm": 0.421875, + "learning_rate": 0.00018728734940644923, + "loss": 2.5412, + "step": 7699 + }, + { + "epoch": 0.33, + "grad_norm": 0.470703125, + "learning_rate": 0.0001872840760978321, + "loss": 2.6152, + "step": 7700 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018728080239646856, + "loss": 2.6428, + "step": 7701 + }, + { + "epoch": 0.33, + "grad_norm": 0.474609375, + "learning_rate": 0.00018727752830237327, + "loss": 2.4931, + "step": 7702 + }, + { + "epoch": 0.33, + "grad_norm": 0.42578125, + "learning_rate": 0.00018727425381556104, + "loss": 2.6114, + "step": 7703 + }, + { + "epoch": 0.33, + "grad_norm": 0.44140625, + "learning_rate": 0.00018727097893604652, + "loss": 2.7049, + "step": 7704 + }, + { + "epoch": 0.33, + "grad_norm": 0.45703125, + "learning_rate": 0.00018726770366384453, + "loss": 2.5829, + "step": 7705 + }, + { + "epoch": 0.33, + "grad_norm": 0.453125, + "learning_rate": 0.00018726442799896974, + "loss": 2.5643, + "step": 7706 + }, + { + "epoch": 0.33, + "grad_norm": 0.48046875, + "learning_rate": 0.0001872611519414369, + "loss": 2.7313, + "step": 7707 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018725787549126079, + "loss": 2.5027, + "step": 7708 + }, + { + "epoch": 0.33, + "grad_norm": 0.46484375, + "learning_rate": 0.00018725459864845614, + "loss": 2.6337, + "step": 7709 + }, + { + "epoch": 0.33, + "grad_norm": 0.361328125, + "learning_rate": 0.00018725132141303766, + "loss": 2.6199, + "step": 7710 + }, + { + "epoch": 0.33, + "grad_norm": 0.396484375, + "learning_rate": 0.00018724804378502015, + "loss": 2.716, + "step": 7711 + }, + { + "epoch": 0.33, + "grad_norm": 0.4609375, + "learning_rate": 0.0001872447657644183, + "loss": 2.6062, + "step": 7712 + }, + { + "epoch": 0.33, + "grad_norm": 0.498046875, + "learning_rate": 0.0001872414873512469, + "loss": 2.7257, + "step": 7713 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.0001872382085455207, + "loss": 2.6353, + "step": 7714 + }, + { + "epoch": 0.33, + "grad_norm": 0.443359375, + "learning_rate": 0.00018723492934725442, + "loss": 2.6666, + "step": 7715 + }, + { + "epoch": 0.33, + "grad_norm": 0.416015625, + "learning_rate": 0.00018723164975646282, + "loss": 2.6046, + "step": 7716 + }, + { + "epoch": 0.33, + "grad_norm": 0.44921875, + "learning_rate": 0.00018722836977316072, + "loss": 2.6049, + "step": 7717 + }, + { + "epoch": 0.33, + "grad_norm": 0.4453125, + "learning_rate": 0.00018722508939736284, + "loss": 2.701, + "step": 7718 + }, + { + "epoch": 0.33, + "grad_norm": 0.49609375, + "learning_rate": 0.00018722180862908391, + "loss": 2.7104, + "step": 7719 + }, + { + "epoch": 0.33, + "grad_norm": 0.45703125, + "learning_rate": 0.0001872185274683387, + "loss": 2.6069, + "step": 7720 + }, + { + "epoch": 0.33, + "grad_norm": 0.443359375, + "learning_rate": 0.00018721524591514203, + "loss": 2.6765, + "step": 7721 + }, + { + "epoch": 0.33, + "grad_norm": 0.47265625, + "learning_rate": 0.0001872119639695086, + "loss": 2.6503, + "step": 7722 + }, + { + "epoch": 0.33, + "grad_norm": 0.4765625, + "learning_rate": 0.00018720868163145325, + "loss": 2.6238, + "step": 7723 + }, + { + "epoch": 0.33, + "grad_norm": 0.46484375, + "learning_rate": 0.00018720539890099066, + "loss": 2.7072, + "step": 7724 + }, + { + "epoch": 0.33, + "grad_norm": 0.412109375, + "learning_rate": 0.00018720211577813561, + "loss": 2.6423, + "step": 7725 + }, + { + "epoch": 0.33, + "grad_norm": 0.416015625, + "learning_rate": 0.00018719883226290298, + "loss": 2.5513, + "step": 7726 + }, + { + "epoch": 0.33, + "grad_norm": 0.51171875, + "learning_rate": 0.00018719554835530745, + "loss": 2.659, + "step": 7727 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.00018719226405536382, + "loss": 2.6041, + "step": 7728 + }, + { + "epoch": 0.33, + "grad_norm": 0.486328125, + "learning_rate": 0.00018718897936308685, + "loss": 2.66, + "step": 7729 + }, + { + "epoch": 0.33, + "grad_norm": 0.478515625, + "learning_rate": 0.00018718569427849136, + "loss": 2.6376, + "step": 7730 + }, + { + "epoch": 0.33, + "grad_norm": 0.462890625, + "learning_rate": 0.0001871824088015921, + "loss": 2.6959, + "step": 7731 + }, + { + "epoch": 0.33, + "grad_norm": 0.421875, + "learning_rate": 0.00018717912293240383, + "loss": 2.5704, + "step": 7732 + }, + { + "epoch": 0.33, + "grad_norm": 0.482421875, + "learning_rate": 0.0001871758366709414, + "loss": 2.8336, + "step": 7733 + }, + { + "epoch": 0.33, + "grad_norm": 0.455078125, + "learning_rate": 0.00018717255001721953, + "loss": 2.6666, + "step": 7734 + }, + { + "epoch": 0.33, + "grad_norm": 0.421875, + "learning_rate": 0.00018716926297125311, + "loss": 2.5639, + "step": 7735 + }, + { + "epoch": 0.33, + "grad_norm": 0.40234375, + "learning_rate": 0.0001871659755330568, + "loss": 2.7171, + "step": 7736 + }, + { + "epoch": 0.33, + "grad_norm": 0.486328125, + "learning_rate": 0.0001871626877026455, + "loss": 2.6578, + "step": 7737 + }, + { + "epoch": 0.33, + "grad_norm": 0.4609375, + "learning_rate": 0.00018715939948003393, + "loss": 2.5345, + "step": 7738 + }, + { + "epoch": 0.33, + "grad_norm": 0.45703125, + "learning_rate": 0.00018715611086523692, + "loss": 2.5616, + "step": 7739 + }, + { + "epoch": 0.33, + "grad_norm": 0.47265625, + "learning_rate": 0.00018715282185826927, + "loss": 2.6347, + "step": 7740 + }, + { + "epoch": 0.33, + "grad_norm": 0.4765625, + "learning_rate": 0.00018714953245914576, + "loss": 2.5689, + "step": 7741 + }, + { + "epoch": 0.33, + "grad_norm": 0.494140625, + "learning_rate": 0.00018714624266788122, + "loss": 2.5825, + "step": 7742 + }, + { + "epoch": 0.33, + "grad_norm": 0.439453125, + "learning_rate": 0.00018714295248449043, + "loss": 2.7045, + "step": 7743 + }, + { + "epoch": 0.33, + "grad_norm": 0.451171875, + "learning_rate": 0.0001871396619089882, + "loss": 2.6347, + "step": 7744 + }, + { + "epoch": 0.33, + "grad_norm": 0.46875, + "learning_rate": 0.00018713637094138933, + "loss": 2.7369, + "step": 7745 + }, + { + "epoch": 0.33, + "grad_norm": 0.5078125, + "learning_rate": 0.00018713307958170865, + "loss": 2.5681, + "step": 7746 + }, + { + "epoch": 0.33, + "grad_norm": 0.451171875, + "learning_rate": 0.00018712978782996093, + "loss": 2.6936, + "step": 7747 + }, + { + "epoch": 0.33, + "grad_norm": 0.5078125, + "learning_rate": 0.00018712649568616106, + "loss": 2.6685, + "step": 7748 + }, + { + "epoch": 0.33, + "grad_norm": 0.435546875, + "learning_rate": 0.00018712320315032374, + "loss": 2.5908, + "step": 7749 + }, + { + "epoch": 0.33, + "grad_norm": 0.455078125, + "learning_rate": 0.00018711991022246388, + "loss": 2.5806, + "step": 7750 + }, + { + "epoch": 0.33, + "grad_norm": 0.474609375, + "learning_rate": 0.00018711661690259626, + "loss": 2.6169, + "step": 7751 + }, + { + "epoch": 0.33, + "grad_norm": 0.43359375, + "learning_rate": 0.00018711332319073568, + "loss": 2.7223, + "step": 7752 + }, + { + "epoch": 0.33, + "grad_norm": 0.455078125, + "learning_rate": 0.000187110029086897, + "loss": 2.5894, + "step": 7753 + }, + { + "epoch": 0.33, + "grad_norm": 0.421875, + "learning_rate": 0.00018710673459109502, + "loss": 2.6906, + "step": 7754 + }, + { + "epoch": 0.33, + "grad_norm": 0.453125, + "learning_rate": 0.00018710343970334455, + "loss": 2.665, + "step": 7755 + }, + { + "epoch": 0.33, + "grad_norm": 0.462890625, + "learning_rate": 0.00018710014442366047, + "loss": 2.6922, + "step": 7756 + }, + { + "epoch": 0.33, + "grad_norm": 0.46484375, + "learning_rate": 0.00018709684875205758, + "loss": 2.7633, + "step": 7757 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.00018709355268855065, + "loss": 2.6895, + "step": 7758 + }, + { + "epoch": 0.33, + "grad_norm": 0.486328125, + "learning_rate": 0.0001870902562331546, + "loss": 2.5067, + "step": 7759 + }, + { + "epoch": 0.33, + "grad_norm": 0.474609375, + "learning_rate": 0.0001870869593858842, + "loss": 2.6175, + "step": 7760 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018708366214675433, + "loss": 2.6669, + "step": 7761 + }, + { + "epoch": 0.33, + "grad_norm": 0.388671875, + "learning_rate": 0.00018708036451577982, + "loss": 2.565, + "step": 7762 + }, + { + "epoch": 0.33, + "grad_norm": 0.3828125, + "learning_rate": 0.00018707706649297547, + "loss": 2.5788, + "step": 7763 + }, + { + "epoch": 0.33, + "grad_norm": 0.443359375, + "learning_rate": 0.00018707376807835613, + "loss": 2.4519, + "step": 7764 + }, + { + "epoch": 0.33, + "grad_norm": 0.419921875, + "learning_rate": 0.00018707046927193667, + "loss": 2.6387, + "step": 7765 + }, + { + "epoch": 0.33, + "grad_norm": 0.435546875, + "learning_rate": 0.00018706717007373194, + "loss": 2.684, + "step": 7766 + }, + { + "epoch": 0.33, + "grad_norm": 0.462890625, + "learning_rate": 0.00018706387048375673, + "loss": 2.6229, + "step": 7767 + }, + { + "epoch": 0.33, + "grad_norm": 0.408203125, + "learning_rate": 0.00018706057050202595, + "loss": 2.5292, + "step": 7768 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.0001870572701285544, + "loss": 2.7118, + "step": 7769 + }, + { + "epoch": 0.33, + "grad_norm": 0.412109375, + "learning_rate": 0.00018705396936335699, + "loss": 2.6152, + "step": 7770 + }, + { + "epoch": 0.33, + "grad_norm": 0.431640625, + "learning_rate": 0.00018705066820644847, + "loss": 2.7051, + "step": 7771 + }, + { + "epoch": 0.33, + "grad_norm": 0.4765625, + "learning_rate": 0.0001870473666578438, + "loss": 2.7928, + "step": 7772 + }, + { + "epoch": 0.33, + "grad_norm": 0.453125, + "learning_rate": 0.00018704406471755776, + "loss": 2.734, + "step": 7773 + }, + { + "epoch": 0.33, + "grad_norm": 0.486328125, + "learning_rate": 0.00018704076238560525, + "loss": 2.7637, + "step": 7774 + }, + { + "epoch": 0.33, + "grad_norm": 0.375, + "learning_rate": 0.00018703745966200112, + "loss": 2.4813, + "step": 7775 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018703415654676024, + "loss": 2.5654, + "step": 7776 + }, + { + "epoch": 0.33, + "grad_norm": 0.43359375, + "learning_rate": 0.00018703085303989746, + "loss": 2.643, + "step": 7777 + }, + { + "epoch": 0.33, + "grad_norm": 0.474609375, + "learning_rate": 0.00018702754914142763, + "loss": 2.5741, + "step": 7778 + }, + { + "epoch": 0.33, + "grad_norm": 0.484375, + "learning_rate": 0.00018702424485136568, + "loss": 2.7103, + "step": 7779 + }, + { + "epoch": 0.33, + "grad_norm": 0.482421875, + "learning_rate": 0.0001870209401697264, + "loss": 2.6848, + "step": 7780 + }, + { + "epoch": 0.33, + "grad_norm": 0.412109375, + "learning_rate": 0.00018701763509652468, + "loss": 2.5785, + "step": 7781 + }, + { + "epoch": 0.33, + "grad_norm": 0.396484375, + "learning_rate": 0.0001870143296317754, + "loss": 2.6833, + "step": 7782 + }, + { + "epoch": 0.33, + "grad_norm": 0.42578125, + "learning_rate": 0.00018701102377549348, + "loss": 2.6085, + "step": 7783 + }, + { + "epoch": 0.33, + "grad_norm": 0.44140625, + "learning_rate": 0.00018700771752769372, + "loss": 2.5553, + "step": 7784 + }, + { + "epoch": 0.33, + "grad_norm": 0.435546875, + "learning_rate": 0.00018700441088839105, + "loss": 2.7425, + "step": 7785 + }, + { + "epoch": 0.33, + "grad_norm": 0.48828125, + "learning_rate": 0.0001870011038576003, + "loss": 2.7374, + "step": 7786 + }, + { + "epoch": 0.33, + "grad_norm": 0.453125, + "learning_rate": 0.00018699779643533638, + "loss": 2.5538, + "step": 7787 + }, + { + "epoch": 0.33, + "grad_norm": 0.42578125, + "learning_rate": 0.00018699448862161418, + "loss": 2.5744, + "step": 7788 + }, + { + "epoch": 0.33, + "grad_norm": 0.46484375, + "learning_rate": 0.00018699118041644858, + "loss": 2.668, + "step": 7789 + }, + { + "epoch": 0.33, + "grad_norm": 0.439453125, + "learning_rate": 0.00018698787181985446, + "loss": 2.6292, + "step": 7790 + }, + { + "epoch": 0.33, + "grad_norm": 0.48828125, + "learning_rate": 0.00018698456283184668, + "loss": 2.5645, + "step": 7791 + }, + { + "epoch": 0.33, + "grad_norm": 0.447265625, + "learning_rate": 0.00018698125345244018, + "loss": 2.4961, + "step": 7792 + }, + { + "epoch": 0.33, + "grad_norm": 0.466796875, + "learning_rate": 0.00018697794368164984, + "loss": 2.7599, + "step": 7793 + }, + { + "epoch": 0.33, + "grad_norm": 0.458984375, + "learning_rate": 0.00018697463351949055, + "loss": 2.6564, + "step": 7794 + }, + { + "epoch": 0.33, + "grad_norm": 0.48046875, + "learning_rate": 0.00018697132296597712, + "loss": 2.7723, + "step": 7795 + }, + { + "epoch": 0.33, + "grad_norm": 0.451171875, + "learning_rate": 0.0001869680120211246, + "loss": 2.6012, + "step": 7796 + }, + { + "epoch": 0.33, + "grad_norm": 0.435546875, + "learning_rate": 0.00018696470068494775, + "loss": 2.6027, + "step": 7797 + }, + { + "epoch": 0.33, + "grad_norm": 0.4765625, + "learning_rate": 0.0001869613889574616, + "loss": 2.6508, + "step": 7798 + }, + { + "epoch": 0.33, + "grad_norm": 0.384765625, + "learning_rate": 0.0001869580768386809, + "loss": 2.5068, + "step": 7799 + }, + { + "epoch": 0.33, + "grad_norm": 0.458984375, + "learning_rate": 0.00018695476432862068, + "loss": 2.5911, + "step": 7800 + }, + { + "epoch": 0.33, + "grad_norm": 0.474609375, + "learning_rate": 0.00018695145142729583, + "loss": 2.6685, + "step": 7801 + }, + { + "epoch": 0.33, + "grad_norm": 0.408203125, + "learning_rate": 0.00018694813813472115, + "loss": 2.6559, + "step": 7802 + }, + { + "epoch": 0.33, + "grad_norm": 0.474609375, + "learning_rate": 0.00018694482445091168, + "loss": 2.6871, + "step": 7803 + }, + { + "epoch": 0.33, + "grad_norm": 0.455078125, + "learning_rate": 0.00018694151037588225, + "loss": 2.7017, + "step": 7804 + }, + { + "epoch": 0.33, + "grad_norm": 0.43359375, + "learning_rate": 0.00018693819590964782, + "loss": 2.5305, + "step": 7805 + }, + { + "epoch": 0.33, + "grad_norm": 0.4609375, + "learning_rate": 0.00018693488105222327, + "loss": 2.5732, + "step": 7806 + }, + { + "epoch": 0.33, + "grad_norm": 0.48828125, + "learning_rate": 0.0001869315658036235, + "loss": 2.7062, + "step": 7807 + }, + { + "epoch": 0.33, + "grad_norm": 0.470703125, + "learning_rate": 0.00018692825016386347, + "loss": 2.6953, + "step": 7808 + }, + { + "epoch": 0.33, + "grad_norm": 0.5234375, + "learning_rate": 0.0001869249341329581, + "loss": 2.6824, + "step": 7809 + }, + { + "epoch": 0.33, + "grad_norm": 0.416015625, + "learning_rate": 0.00018692161771092226, + "loss": 2.607, + "step": 7810 + }, + { + "epoch": 0.33, + "grad_norm": 0.46484375, + "learning_rate": 0.00018691830089777095, + "loss": 2.6486, + "step": 7811 + }, + { + "epoch": 0.33, + "grad_norm": 0.47265625, + "learning_rate": 0.00018691498369351902, + "loss": 2.7377, + "step": 7812 + }, + { + "epoch": 0.33, + "grad_norm": 0.46484375, + "learning_rate": 0.0001869116660981814, + "loss": 2.674, + "step": 7813 + }, + { + "epoch": 0.33, + "grad_norm": 0.443359375, + "learning_rate": 0.0001869083481117731, + "loss": 2.6016, + "step": 7814 + }, + { + "epoch": 0.33, + "grad_norm": 0.44921875, + "learning_rate": 0.000186905029734309, + "loss": 2.664, + "step": 7815 + }, + { + "epoch": 0.33, + "grad_norm": 0.39453125, + "learning_rate": 0.000186901710965804, + "loss": 2.496, + "step": 7816 + }, + { + "epoch": 0.33, + "grad_norm": 0.44921875, + "learning_rate": 0.00018689839180627304, + "loss": 2.6046, + "step": 7817 + }, + { + "epoch": 0.33, + "grad_norm": 0.45703125, + "learning_rate": 0.0001868950722557311, + "loss": 2.5288, + "step": 7818 + }, + { + "epoch": 0.33, + "grad_norm": 0.40234375, + "learning_rate": 0.0001868917523141931, + "loss": 2.584, + "step": 7819 + }, + { + "epoch": 0.33, + "grad_norm": 0.44921875, + "learning_rate": 0.00018688843198167397, + "loss": 2.6626, + "step": 7820 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018688511125818864, + "loss": 2.6593, + "step": 7821 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018688179014375204, + "loss": 2.5168, + "step": 7822 + }, + { + "epoch": 0.33, + "grad_norm": 0.49609375, + "learning_rate": 0.00018687846863837916, + "loss": 2.7006, + "step": 7823 + }, + { + "epoch": 0.33, + "grad_norm": 0.466796875, + "learning_rate": 0.00018687514674208493, + "loss": 2.6011, + "step": 7824 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018687182445488427, + "loss": 2.8102, + "step": 7825 + }, + { + "epoch": 0.33, + "grad_norm": 0.447265625, + "learning_rate": 0.00018686850177679215, + "loss": 2.5553, + "step": 7826 + }, + { + "epoch": 0.33, + "grad_norm": 0.455078125, + "learning_rate": 0.00018686517870782353, + "loss": 2.6088, + "step": 7827 + }, + { + "epoch": 0.33, + "grad_norm": 0.37890625, + "learning_rate": 0.00018686185524799333, + "loss": 2.5519, + "step": 7828 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.00018685853139731654, + "loss": 2.602, + "step": 7829 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018685520715580812, + "loss": 2.6536, + "step": 7830 + }, + { + "epoch": 0.33, + "grad_norm": 0.45703125, + "learning_rate": 0.00018685188252348298, + "loss": 2.5695, + "step": 7831 + }, + { + "epoch": 0.33, + "grad_norm": 0.458984375, + "learning_rate": 0.0001868485575003561, + "loss": 2.6483, + "step": 7832 + }, + { + "epoch": 0.33, + "grad_norm": 0.408203125, + "learning_rate": 0.0001868452320864425, + "loss": 2.5072, + "step": 7833 + }, + { + "epoch": 0.33, + "grad_norm": 0.443359375, + "learning_rate": 0.00018684190628175706, + "loss": 2.5966, + "step": 7834 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.00018683858008631477, + "loss": 2.5127, + "step": 7835 + }, + { + "epoch": 0.33, + "grad_norm": 0.423828125, + "learning_rate": 0.00018683525350013063, + "loss": 2.5574, + "step": 7836 + }, + { + "epoch": 0.33, + "grad_norm": 0.421875, + "learning_rate": 0.0001868319265232196, + "loss": 2.6307, + "step": 7837 + }, + { + "epoch": 0.33, + "grad_norm": 0.486328125, + "learning_rate": 0.00018682859915559658, + "loss": 2.8323, + "step": 7838 + }, + { + "epoch": 0.33, + "grad_norm": 0.4453125, + "learning_rate": 0.00018682527139727664, + "loss": 2.747, + "step": 7839 + }, + { + "epoch": 0.33, + "grad_norm": 0.4140625, + "learning_rate": 0.0001868219432482747, + "loss": 2.6341, + "step": 7840 + }, + { + "epoch": 0.33, + "grad_norm": 0.4375, + "learning_rate": 0.00018681861470860574, + "loss": 2.5549, + "step": 7841 + }, + { + "epoch": 0.33, + "grad_norm": 0.474609375, + "learning_rate": 0.00018681528577828472, + "loss": 2.6305, + "step": 7842 + }, + { + "epoch": 0.33, + "grad_norm": 0.40625, + "learning_rate": 0.00018681195645732664, + "loss": 2.5941, + "step": 7843 + }, + { + "epoch": 0.33, + "grad_norm": 0.42578125, + "learning_rate": 0.00018680862674574653, + "loss": 2.4912, + "step": 7844 + }, + { + "epoch": 0.33, + "grad_norm": 0.4140625, + "learning_rate": 0.0001868052966435593, + "loss": 2.5648, + "step": 7845 + }, + { + "epoch": 0.33, + "grad_norm": 0.380859375, + "learning_rate": 0.00018680196615077996, + "loss": 2.6804, + "step": 7846 + }, + { + "epoch": 0.33, + "grad_norm": 0.43359375, + "learning_rate": 0.0001867986352674235, + "loss": 2.6264, + "step": 7847 + }, + { + "epoch": 0.33, + "grad_norm": 0.470703125, + "learning_rate": 0.0001867953039935049, + "loss": 2.6645, + "step": 7848 + }, + { + "epoch": 0.33, + "grad_norm": 0.54296875, + "learning_rate": 0.00018679197232903912, + "loss": 2.8695, + "step": 7849 + }, + { + "epoch": 0.33, + "grad_norm": 0.41796875, + "learning_rate": 0.0001867886402740412, + "loss": 2.5314, + "step": 7850 + }, + { + "epoch": 0.33, + "grad_norm": 0.44921875, + "learning_rate": 0.00018678530782852617, + "loss": 2.5668, + "step": 7851 + }, + { + "epoch": 0.33, + "grad_norm": 0.498046875, + "learning_rate": 0.00018678197499250892, + "loss": 2.694, + "step": 7852 + }, + { + "epoch": 0.33, + "grad_norm": 0.451171875, + "learning_rate": 0.0001867786417660045, + "loss": 2.517, + "step": 7853 + }, + { + "epoch": 0.33, + "grad_norm": 0.41015625, + "learning_rate": 0.00018677530814902791, + "loss": 2.6712, + "step": 7854 + }, + { + "epoch": 0.33, + "grad_norm": 0.431640625, + "learning_rate": 0.00018677197414159415, + "loss": 2.5958, + "step": 7855 + }, + { + "epoch": 0.33, + "grad_norm": 0.427734375, + "learning_rate": 0.00018676863974371823, + "loss": 2.6819, + "step": 7856 + }, + { + "epoch": 0.33, + "grad_norm": 0.46484375, + "learning_rate": 0.00018676530495541512, + "loss": 2.7103, + "step": 7857 + }, + { + "epoch": 0.33, + "grad_norm": 0.443359375, + "learning_rate": 0.00018676196977669983, + "loss": 2.6081, + "step": 7858 + }, + { + "epoch": 0.33, + "grad_norm": 0.4140625, + "learning_rate": 0.0001867586342075874, + "loss": 2.6975, + "step": 7859 + }, + { + "epoch": 0.33, + "grad_norm": 0.51171875, + "learning_rate": 0.00018675529824809285, + "loss": 2.6201, + "step": 7860 + }, + { + "epoch": 0.33, + "grad_norm": 0.53515625, + "learning_rate": 0.00018675196189823115, + "loss": 2.6439, + "step": 7861 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018674862515801732, + "loss": 2.6091, + "step": 7862 + }, + { + "epoch": 0.34, + "grad_norm": 0.43359375, + "learning_rate": 0.00018674528802746637, + "loss": 2.529, + "step": 7863 + }, + { + "epoch": 0.34, + "grad_norm": 0.43359375, + "learning_rate": 0.00018674195050659336, + "loss": 2.654, + "step": 7864 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.00018673861259541324, + "loss": 2.5476, + "step": 7865 + }, + { + "epoch": 0.34, + "grad_norm": 0.478515625, + "learning_rate": 0.00018673527429394107, + "loss": 2.7429, + "step": 7866 + }, + { + "epoch": 0.34, + "grad_norm": 0.455078125, + "learning_rate": 0.00018673193560219187, + "loss": 2.626, + "step": 7867 + }, + { + "epoch": 0.34, + "grad_norm": 0.435546875, + "learning_rate": 0.00018672859652018066, + "loss": 2.6394, + "step": 7868 + }, + { + "epoch": 0.34, + "grad_norm": 0.44921875, + "learning_rate": 0.00018672525704792244, + "loss": 2.5161, + "step": 7869 + }, + { + "epoch": 0.34, + "grad_norm": 0.44140625, + "learning_rate": 0.00018672191718543228, + "loss": 2.6209, + "step": 7870 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018671857693272517, + "loss": 2.6254, + "step": 7871 + }, + { + "epoch": 0.34, + "grad_norm": 0.4921875, + "learning_rate": 0.00018671523628981617, + "loss": 2.7121, + "step": 7872 + }, + { + "epoch": 0.34, + "grad_norm": 0.494140625, + "learning_rate": 0.0001867118952567203, + "loss": 2.8168, + "step": 7873 + }, + { + "epoch": 0.34, + "grad_norm": 0.455078125, + "learning_rate": 0.00018670855383345257, + "loss": 2.676, + "step": 7874 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.00018670521202002804, + "loss": 2.5835, + "step": 7875 + }, + { + "epoch": 0.34, + "grad_norm": 0.4296875, + "learning_rate": 0.00018670186981646175, + "loss": 2.7401, + "step": 7876 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.00018669852722276873, + "loss": 2.6434, + "step": 7877 + }, + { + "epoch": 0.34, + "grad_norm": 0.416015625, + "learning_rate": 0.00018669518423896402, + "loss": 2.5503, + "step": 7878 + }, + { + "epoch": 0.34, + "grad_norm": 0.482421875, + "learning_rate": 0.0001866918408650627, + "loss": 2.6764, + "step": 7879 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.0001866884971010797, + "loss": 2.5888, + "step": 7880 + }, + { + "epoch": 0.34, + "grad_norm": 0.486328125, + "learning_rate": 0.00018668515294703022, + "loss": 2.774, + "step": 7881 + }, + { + "epoch": 0.34, + "grad_norm": 0.51953125, + "learning_rate": 0.0001866818084029292, + "loss": 2.6849, + "step": 7882 + }, + { + "epoch": 0.34, + "grad_norm": 0.443359375, + "learning_rate": 0.0001866784634687917, + "loss": 2.7337, + "step": 7883 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.00018667511814463283, + "loss": 2.7349, + "step": 7884 + }, + { + "epoch": 0.34, + "grad_norm": 0.375, + "learning_rate": 0.00018667177243046759, + "loss": 2.6225, + "step": 7885 + }, + { + "epoch": 0.34, + "grad_norm": 0.421875, + "learning_rate": 0.00018666842632631105, + "loss": 2.5134, + "step": 7886 + }, + { + "epoch": 0.34, + "grad_norm": 0.427734375, + "learning_rate": 0.00018666507983217828, + "loss": 2.6732, + "step": 7887 + }, + { + "epoch": 0.34, + "grad_norm": 0.439453125, + "learning_rate": 0.0001866617329480843, + "loss": 2.5367, + "step": 7888 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.00018665838567404422, + "loss": 2.6994, + "step": 7889 + }, + { + "epoch": 0.34, + "grad_norm": 0.47265625, + "learning_rate": 0.00018665503801007302, + "loss": 2.6539, + "step": 7890 + }, + { + "epoch": 0.34, + "grad_norm": 0.478515625, + "learning_rate": 0.00018665168995618588, + "loss": 2.836, + "step": 7891 + }, + { + "epoch": 0.34, + "grad_norm": 0.43359375, + "learning_rate": 0.00018664834151239777, + "loss": 2.6723, + "step": 7892 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.0001866449926787238, + "loss": 2.7971, + "step": 7893 + }, + { + "epoch": 0.34, + "grad_norm": 0.46875, + "learning_rate": 0.000186641643455179, + "loss": 2.6339, + "step": 7894 + }, + { + "epoch": 0.34, + "grad_norm": 0.439453125, + "learning_rate": 0.00018663829384177848, + "loss": 2.5776, + "step": 7895 + }, + { + "epoch": 0.34, + "grad_norm": 0.42578125, + "learning_rate": 0.00018663494383853733, + "loss": 2.5906, + "step": 7896 + }, + { + "epoch": 0.34, + "grad_norm": 0.482421875, + "learning_rate": 0.0001866315934454706, + "loss": 2.5039, + "step": 7897 + }, + { + "epoch": 0.34, + "grad_norm": 0.46875, + "learning_rate": 0.0001866282426625933, + "loss": 2.7113, + "step": 7898 + }, + { + "epoch": 0.34, + "grad_norm": 0.431640625, + "learning_rate": 0.0001866248914899206, + "loss": 2.5776, + "step": 7899 + }, + { + "epoch": 0.34, + "grad_norm": 0.41796875, + "learning_rate": 0.00018662153992746755, + "loss": 2.6162, + "step": 7900 + }, + { + "epoch": 0.34, + "grad_norm": 0.49609375, + "learning_rate": 0.00018661818797524924, + "loss": 2.6514, + "step": 7901 + }, + { + "epoch": 0.34, + "grad_norm": 0.408203125, + "learning_rate": 0.00018661483563328074, + "loss": 2.7085, + "step": 7902 + }, + { + "epoch": 0.34, + "grad_norm": 0.421875, + "learning_rate": 0.0001866114829015771, + "loss": 2.5629, + "step": 7903 + }, + { + "epoch": 0.34, + "grad_norm": 0.412109375, + "learning_rate": 0.00018660812978015345, + "loss": 2.6316, + "step": 7904 + }, + { + "epoch": 0.34, + "grad_norm": 0.470703125, + "learning_rate": 0.00018660477626902488, + "loss": 2.5604, + "step": 7905 + }, + { + "epoch": 0.34, + "grad_norm": 0.50390625, + "learning_rate": 0.00018660142236820646, + "loss": 2.7327, + "step": 7906 + }, + { + "epoch": 0.34, + "grad_norm": 0.47265625, + "learning_rate": 0.00018659806807771328, + "loss": 2.7549, + "step": 7907 + }, + { + "epoch": 0.34, + "grad_norm": 0.43359375, + "learning_rate": 0.00018659471339756048, + "loss": 2.5334, + "step": 7908 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.00018659135832776305, + "loss": 2.5582, + "step": 7909 + }, + { + "epoch": 0.34, + "grad_norm": 0.48046875, + "learning_rate": 0.00018658800286833622, + "loss": 2.5882, + "step": 7910 + }, + { + "epoch": 0.34, + "grad_norm": 0.48046875, + "learning_rate": 0.000186584647019295, + "loss": 2.596, + "step": 7911 + }, + { + "epoch": 0.34, + "grad_norm": 0.470703125, + "learning_rate": 0.0001865812907806545, + "loss": 2.6933, + "step": 7912 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018657793415242983, + "loss": 2.6034, + "step": 7913 + }, + { + "epoch": 0.34, + "grad_norm": 0.474609375, + "learning_rate": 0.0001865745771346361, + "loss": 2.5477, + "step": 7914 + }, + { + "epoch": 0.34, + "grad_norm": 0.470703125, + "learning_rate": 0.0001865712197272884, + "loss": 2.5878, + "step": 7915 + }, + { + "epoch": 0.34, + "grad_norm": 0.462890625, + "learning_rate": 0.0001865678619304019, + "loss": 2.5349, + "step": 7916 + }, + { + "epoch": 0.34, + "grad_norm": 0.46875, + "learning_rate": 0.00018656450374399163, + "loss": 2.5757, + "step": 7917 + }, + { + "epoch": 0.34, + "grad_norm": 0.427734375, + "learning_rate": 0.0001865611451680727, + "loss": 2.6567, + "step": 7918 + }, + { + "epoch": 0.34, + "grad_norm": 0.412109375, + "learning_rate": 0.0001865577862026603, + "loss": 2.6313, + "step": 7919 + }, + { + "epoch": 0.34, + "grad_norm": 0.50390625, + "learning_rate": 0.00018655442684776947, + "loss": 2.6965, + "step": 7920 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.00018655106710341538, + "loss": 2.5516, + "step": 7921 + }, + { + "epoch": 0.34, + "grad_norm": 0.478515625, + "learning_rate": 0.00018654770696961305, + "loss": 2.6163, + "step": 7922 + }, + { + "epoch": 0.34, + "grad_norm": 0.4375, + "learning_rate": 0.00018654434644637774, + "loss": 2.7623, + "step": 7923 + }, + { + "epoch": 0.34, + "grad_norm": 0.44921875, + "learning_rate": 0.00018654098553372444, + "loss": 2.5517, + "step": 7924 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.00018653762423166836, + "loss": 2.6917, + "step": 7925 + }, + { + "epoch": 0.34, + "grad_norm": 0.439453125, + "learning_rate": 0.0001865342625402246, + "loss": 2.7179, + "step": 7926 + }, + { + "epoch": 0.34, + "grad_norm": 0.498046875, + "learning_rate": 0.00018653090045940828, + "loss": 2.7659, + "step": 7927 + }, + { + "epoch": 0.34, + "grad_norm": 0.45703125, + "learning_rate": 0.00018652753798923454, + "loss": 2.6978, + "step": 7928 + }, + { + "epoch": 0.34, + "grad_norm": 0.482421875, + "learning_rate": 0.00018652417512971849, + "loss": 2.7048, + "step": 7929 + }, + { + "epoch": 0.34, + "grad_norm": 0.443359375, + "learning_rate": 0.00018652081188087527, + "loss": 2.7133, + "step": 7930 + }, + { + "epoch": 0.34, + "grad_norm": 0.416015625, + "learning_rate": 0.00018651744824272004, + "loss": 2.6334, + "step": 7931 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.0001865140842152679, + "loss": 2.6972, + "step": 7932 + }, + { + "epoch": 0.34, + "grad_norm": 0.4375, + "learning_rate": 0.00018651071979853398, + "loss": 2.6662, + "step": 7933 + }, + { + "epoch": 0.34, + "grad_norm": 0.494140625, + "learning_rate": 0.00018650735499253345, + "loss": 2.6018, + "step": 7934 + }, + { + "epoch": 0.34, + "grad_norm": 0.466796875, + "learning_rate": 0.00018650398979728144, + "loss": 2.6, + "step": 7935 + }, + { + "epoch": 0.34, + "grad_norm": 0.490234375, + "learning_rate": 0.0001865006242127931, + "loss": 2.5782, + "step": 7936 + }, + { + "epoch": 0.34, + "grad_norm": 0.416015625, + "learning_rate": 0.00018649725823908356, + "loss": 2.46, + "step": 7937 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018649389187616792, + "loss": 2.6521, + "step": 7938 + }, + { + "epoch": 0.34, + "grad_norm": 0.3984375, + "learning_rate": 0.00018649052512406143, + "loss": 2.7271, + "step": 7939 + }, + { + "epoch": 0.34, + "grad_norm": 0.41796875, + "learning_rate": 0.00018648715798277917, + "loss": 2.4819, + "step": 7940 + }, + { + "epoch": 0.34, + "grad_norm": 0.451171875, + "learning_rate": 0.0001864837904523363, + "loss": 2.6301, + "step": 7941 + }, + { + "epoch": 0.34, + "grad_norm": 0.49609375, + "learning_rate": 0.000186480422532748, + "loss": 2.7668, + "step": 7942 + }, + { + "epoch": 0.34, + "grad_norm": 0.466796875, + "learning_rate": 0.0001864770542240294, + "loss": 2.809, + "step": 7943 + }, + { + "epoch": 0.34, + "grad_norm": 0.486328125, + "learning_rate": 0.00018647368552619564, + "loss": 2.6965, + "step": 7944 + }, + { + "epoch": 0.34, + "grad_norm": 0.5390625, + "learning_rate": 0.00018647031643926191, + "loss": 2.714, + "step": 7945 + }, + { + "epoch": 0.34, + "grad_norm": 0.482421875, + "learning_rate": 0.00018646694696324335, + "loss": 2.5664, + "step": 7946 + }, + { + "epoch": 0.34, + "grad_norm": 0.4609375, + "learning_rate": 0.00018646357709815515, + "loss": 2.6303, + "step": 7947 + }, + { + "epoch": 0.34, + "grad_norm": 0.494140625, + "learning_rate": 0.00018646020684401243, + "loss": 2.7309, + "step": 7948 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.00018645683620083036, + "loss": 2.6023, + "step": 7949 + }, + { + "epoch": 0.34, + "grad_norm": 0.470703125, + "learning_rate": 0.00018645346516862413, + "loss": 2.58, + "step": 7950 + }, + { + "epoch": 0.34, + "grad_norm": 0.43359375, + "learning_rate": 0.00018645009374740896, + "loss": 2.7576, + "step": 7951 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.00018644672193719993, + "loss": 2.6984, + "step": 7952 + }, + { + "epoch": 0.34, + "grad_norm": 0.466796875, + "learning_rate": 0.0001864433497380122, + "loss": 2.6163, + "step": 7953 + }, + { + "epoch": 0.34, + "grad_norm": 0.427734375, + "learning_rate": 0.00018643997714986102, + "loss": 2.6962, + "step": 7954 + }, + { + "epoch": 0.34, + "grad_norm": 0.45703125, + "learning_rate": 0.00018643660417276155, + "loss": 2.4977, + "step": 7955 + }, + { + "epoch": 0.34, + "grad_norm": 0.41015625, + "learning_rate": 0.0001864332308067289, + "loss": 2.6559, + "step": 7956 + }, + { + "epoch": 0.34, + "grad_norm": 0.404296875, + "learning_rate": 0.00018642985705177834, + "loss": 2.6293, + "step": 7957 + }, + { + "epoch": 0.34, + "grad_norm": 0.3671875, + "learning_rate": 0.00018642648290792503, + "loss": 2.6511, + "step": 7958 + }, + { + "epoch": 0.34, + "grad_norm": 0.484375, + "learning_rate": 0.00018642310837518409, + "loss": 2.7088, + "step": 7959 + }, + { + "epoch": 0.34, + "grad_norm": 0.486328125, + "learning_rate": 0.00018641973345357078, + "loss": 2.6932, + "step": 7960 + }, + { + "epoch": 0.34, + "grad_norm": 0.439453125, + "learning_rate": 0.0001864163581431002, + "loss": 2.7195, + "step": 7961 + }, + { + "epoch": 0.34, + "grad_norm": 0.478515625, + "learning_rate": 0.00018641298244378765, + "loss": 2.7185, + "step": 7962 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018640960635564824, + "loss": 2.5776, + "step": 7963 + }, + { + "epoch": 0.34, + "grad_norm": 0.4609375, + "learning_rate": 0.00018640622987869718, + "loss": 2.6601, + "step": 7964 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.00018640285301294964, + "loss": 2.6892, + "step": 7965 + }, + { + "epoch": 0.34, + "grad_norm": 0.435546875, + "learning_rate": 0.00018639947575842087, + "loss": 2.6022, + "step": 7966 + }, + { + "epoch": 0.34, + "grad_norm": 0.443359375, + "learning_rate": 0.000186396098115126, + "loss": 2.6946, + "step": 7967 + }, + { + "epoch": 0.34, + "grad_norm": 0.41796875, + "learning_rate": 0.0001863927200830803, + "loss": 2.5953, + "step": 7968 + }, + { + "epoch": 0.34, + "grad_norm": 0.48046875, + "learning_rate": 0.00018638934166229892, + "loss": 2.7424, + "step": 7969 + }, + { + "epoch": 0.34, + "grad_norm": 0.443359375, + "learning_rate": 0.00018638596285279707, + "loss": 2.6105, + "step": 7970 + }, + { + "epoch": 0.34, + "grad_norm": 0.462890625, + "learning_rate": 0.00018638258365458993, + "loss": 2.648, + "step": 7971 + }, + { + "epoch": 0.34, + "grad_norm": 0.4375, + "learning_rate": 0.00018637920406769278, + "loss": 2.7142, + "step": 7972 + }, + { + "epoch": 0.34, + "grad_norm": 0.486328125, + "learning_rate": 0.00018637582409212076, + "loss": 2.7071, + "step": 7973 + }, + { + "epoch": 0.34, + "grad_norm": 0.52734375, + "learning_rate": 0.0001863724437278891, + "loss": 2.7357, + "step": 7974 + }, + { + "epoch": 0.34, + "grad_norm": 0.45703125, + "learning_rate": 0.00018636906297501299, + "loss": 2.6509, + "step": 7975 + }, + { + "epoch": 0.34, + "grad_norm": 0.53515625, + "learning_rate": 0.0001863656818335077, + "loss": 2.6323, + "step": 7976 + }, + { + "epoch": 0.34, + "grad_norm": 0.462890625, + "learning_rate": 0.00018636230030338836, + "loss": 2.6014, + "step": 7977 + }, + { + "epoch": 0.34, + "grad_norm": 0.41015625, + "learning_rate": 0.00018635891838467025, + "loss": 2.5019, + "step": 7978 + }, + { + "epoch": 0.34, + "grad_norm": 0.478515625, + "learning_rate": 0.00018635553607736859, + "loss": 2.6005, + "step": 7979 + }, + { + "epoch": 0.34, + "grad_norm": 0.4609375, + "learning_rate": 0.00018635215338149856, + "loss": 2.5442, + "step": 7980 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.0001863487702970754, + "loss": 2.6151, + "step": 7981 + }, + { + "epoch": 0.34, + "grad_norm": 0.5, + "learning_rate": 0.0001863453868241143, + "loss": 2.6415, + "step": 7982 + }, + { + "epoch": 0.34, + "grad_norm": 0.49609375, + "learning_rate": 0.00018634200296263056, + "loss": 2.6936, + "step": 7983 + }, + { + "epoch": 0.34, + "grad_norm": 0.4140625, + "learning_rate": 0.00018633861871263934, + "loss": 2.6629, + "step": 7984 + }, + { + "epoch": 0.34, + "grad_norm": 0.482421875, + "learning_rate": 0.00018633523407415588, + "loss": 2.6165, + "step": 7985 + }, + { + "epoch": 0.34, + "grad_norm": 0.49609375, + "learning_rate": 0.00018633184904719543, + "loss": 2.7516, + "step": 7986 + }, + { + "epoch": 0.34, + "grad_norm": 0.470703125, + "learning_rate": 0.0001863284636317732, + "loss": 2.6883, + "step": 7987 + }, + { + "epoch": 0.34, + "grad_norm": 0.455078125, + "learning_rate": 0.00018632507782790446, + "loss": 2.5278, + "step": 7988 + }, + { + "epoch": 0.34, + "grad_norm": 0.4609375, + "learning_rate": 0.0001863216916356044, + "loss": 2.5575, + "step": 7989 + }, + { + "epoch": 0.34, + "grad_norm": 0.4609375, + "learning_rate": 0.00018631830505488827, + "loss": 2.6906, + "step": 7990 + }, + { + "epoch": 0.34, + "grad_norm": 0.43359375, + "learning_rate": 0.00018631491808577134, + "loss": 2.5468, + "step": 7991 + }, + { + "epoch": 0.34, + "grad_norm": 0.48046875, + "learning_rate": 0.00018631153072826877, + "loss": 2.5991, + "step": 7992 + }, + { + "epoch": 0.34, + "grad_norm": 0.466796875, + "learning_rate": 0.0001863081429823959, + "loss": 2.5757, + "step": 7993 + }, + { + "epoch": 0.34, + "grad_norm": 0.45703125, + "learning_rate": 0.00018630475484816792, + "loss": 2.6271, + "step": 7994 + }, + { + "epoch": 0.34, + "grad_norm": 0.5078125, + "learning_rate": 0.00018630136632560008, + "loss": 2.7563, + "step": 7995 + }, + { + "epoch": 0.34, + "grad_norm": 0.46484375, + "learning_rate": 0.0001862979774147076, + "loss": 2.6011, + "step": 7996 + }, + { + "epoch": 0.34, + "grad_norm": 0.4921875, + "learning_rate": 0.0001862945881155058, + "loss": 2.5943, + "step": 7997 + }, + { + "epoch": 0.34, + "grad_norm": 0.4921875, + "learning_rate": 0.00018629119842800989, + "loss": 2.794, + "step": 7998 + }, + { + "epoch": 0.34, + "grad_norm": 0.47265625, + "learning_rate": 0.00018628780835223512, + "loss": 2.6321, + "step": 7999 + }, + { + "epoch": 0.34, + "grad_norm": 0.416015625, + "learning_rate": 0.0001862844178881967, + "loss": 2.5801, + "step": 8000 + }, + { + "epoch": 0.34, + "grad_norm": 0.51171875, + "learning_rate": 0.00018628102703591, + "loss": 2.6936, + "step": 8001 + }, + { + "epoch": 0.34, + "grad_norm": 0.482421875, + "learning_rate": 0.0001862776357953902, + "loss": 2.6441, + "step": 8002 + }, + { + "epoch": 0.34, + "grad_norm": 0.40234375, + "learning_rate": 0.00018627424416665253, + "loss": 2.5151, + "step": 8003 + }, + { + "epoch": 0.34, + "grad_norm": 0.439453125, + "learning_rate": 0.00018627085214971233, + "loss": 2.5869, + "step": 8004 + }, + { + "epoch": 0.34, + "grad_norm": 0.44140625, + "learning_rate": 0.0001862674597445848, + "loss": 2.7017, + "step": 8005 + }, + { + "epoch": 0.34, + "grad_norm": 0.470703125, + "learning_rate": 0.00018626406695128523, + "loss": 2.6918, + "step": 8006 + }, + { + "epoch": 0.34, + "grad_norm": 0.470703125, + "learning_rate": 0.0001862606737698289, + "loss": 2.5422, + "step": 8007 + }, + { + "epoch": 0.34, + "grad_norm": 0.482421875, + "learning_rate": 0.00018625728020023105, + "loss": 2.7105, + "step": 8008 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018625388624250695, + "loss": 2.713, + "step": 8009 + }, + { + "epoch": 0.34, + "grad_norm": 0.44140625, + "learning_rate": 0.00018625049189667191, + "loss": 2.6196, + "step": 8010 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018624709716274116, + "loss": 2.7198, + "step": 8011 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018624370204073002, + "loss": 2.7079, + "step": 8012 + }, + { + "epoch": 0.34, + "grad_norm": 0.41796875, + "learning_rate": 0.00018624030653065373, + "loss": 2.6134, + "step": 8013 + }, + { + "epoch": 0.34, + "grad_norm": 0.421875, + "learning_rate": 0.00018623691063252755, + "loss": 2.6223, + "step": 8014 + }, + { + "epoch": 0.34, + "grad_norm": 0.392578125, + "learning_rate": 0.0001862335143463668, + "loss": 2.5425, + "step": 8015 + }, + { + "epoch": 0.34, + "grad_norm": 0.4609375, + "learning_rate": 0.00018623011767218673, + "loss": 2.6116, + "step": 8016 + }, + { + "epoch": 0.34, + "grad_norm": 0.4375, + "learning_rate": 0.00018622672061000272, + "loss": 2.6079, + "step": 8017 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.0001862233231598299, + "loss": 2.5918, + "step": 8018 + }, + { + "epoch": 0.34, + "grad_norm": 0.494140625, + "learning_rate": 0.0001862199253216837, + "loss": 2.5983, + "step": 8019 + }, + { + "epoch": 0.34, + "grad_norm": 0.396484375, + "learning_rate": 0.00018621652709557927, + "loss": 2.5591, + "step": 8020 + }, + { + "epoch": 0.34, + "grad_norm": 0.4140625, + "learning_rate": 0.00018621312848153204, + "loss": 2.596, + "step": 8021 + }, + { + "epoch": 0.34, + "grad_norm": 0.455078125, + "learning_rate": 0.0001862097294795572, + "loss": 2.7309, + "step": 8022 + }, + { + "epoch": 0.34, + "grad_norm": 0.515625, + "learning_rate": 0.0001862063300896701, + "loss": 2.7025, + "step": 8023 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.00018620293031188602, + "loss": 2.6721, + "step": 8024 + }, + { + "epoch": 0.34, + "grad_norm": 0.443359375, + "learning_rate": 0.00018619953014622026, + "loss": 2.7036, + "step": 8025 + }, + { + "epoch": 0.34, + "grad_norm": 0.40625, + "learning_rate": 0.00018619612959268813, + "loss": 2.5352, + "step": 8026 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.0001861927286513049, + "loss": 2.7046, + "step": 8027 + }, + { + "epoch": 0.34, + "grad_norm": 0.43359375, + "learning_rate": 0.00018618932732208587, + "loss": 2.5629, + "step": 8028 + }, + { + "epoch": 0.34, + "grad_norm": 0.498046875, + "learning_rate": 0.0001861859256050464, + "loss": 2.6331, + "step": 8029 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018618252350020175, + "loss": 2.6018, + "step": 8030 + }, + { + "epoch": 0.34, + "grad_norm": 0.435546875, + "learning_rate": 0.00018617912100756721, + "loss": 2.6847, + "step": 8031 + }, + { + "epoch": 0.34, + "grad_norm": 0.4765625, + "learning_rate": 0.00018617571812715817, + "loss": 2.7374, + "step": 8032 + }, + { + "epoch": 0.34, + "grad_norm": 0.455078125, + "learning_rate": 0.00018617231485898987, + "loss": 2.6783, + "step": 8033 + }, + { + "epoch": 0.34, + "grad_norm": 0.4609375, + "learning_rate": 0.00018616891120307766, + "loss": 2.6247, + "step": 8034 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.00018616550715943683, + "loss": 2.5752, + "step": 8035 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.0001861621027280827, + "loss": 2.7265, + "step": 8036 + }, + { + "epoch": 0.34, + "grad_norm": 0.46875, + "learning_rate": 0.00018615869790903063, + "loss": 2.6778, + "step": 8037 + }, + { + "epoch": 0.34, + "grad_norm": 0.431640625, + "learning_rate": 0.00018615529270229585, + "loss": 2.72, + "step": 8038 + }, + { + "epoch": 0.34, + "grad_norm": 0.4296875, + "learning_rate": 0.00018615188710789378, + "loss": 2.5553, + "step": 8039 + }, + { + "epoch": 0.34, + "grad_norm": 0.44140625, + "learning_rate": 0.00018614848112583968, + "loss": 2.5527, + "step": 8040 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.00018614507475614892, + "loss": 2.6363, + "step": 8041 + }, + { + "epoch": 0.34, + "grad_norm": 0.416015625, + "learning_rate": 0.00018614166799883682, + "loss": 2.602, + "step": 8042 + }, + { + "epoch": 0.34, + "grad_norm": 0.5078125, + "learning_rate": 0.00018613826085391865, + "loss": 2.6, + "step": 8043 + }, + { + "epoch": 0.34, + "grad_norm": 0.451171875, + "learning_rate": 0.0001861348533214098, + "loss": 2.6558, + "step": 8044 + }, + { + "epoch": 0.34, + "grad_norm": 0.484375, + "learning_rate": 0.0001861314454013256, + "loss": 2.6402, + "step": 8045 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.00018612803709368138, + "loss": 2.6407, + "step": 8046 + }, + { + "epoch": 0.34, + "grad_norm": 0.4453125, + "learning_rate": 0.00018612462839849244, + "loss": 2.6422, + "step": 8047 + }, + { + "epoch": 0.34, + "grad_norm": 0.435546875, + "learning_rate": 0.0001861212193157742, + "loss": 2.6947, + "step": 8048 + }, + { + "epoch": 0.34, + "grad_norm": 0.515625, + "learning_rate": 0.0001861178098455419, + "loss": 2.7232, + "step": 8049 + }, + { + "epoch": 0.34, + "grad_norm": 0.4375, + "learning_rate": 0.00018611439998781094, + "loss": 2.6401, + "step": 8050 + }, + { + "epoch": 0.34, + "grad_norm": 0.400390625, + "learning_rate": 0.00018611098974259666, + "loss": 2.507, + "step": 8051 + }, + { + "epoch": 0.34, + "grad_norm": 0.41015625, + "learning_rate": 0.0001861075791099144, + "loss": 2.5973, + "step": 8052 + }, + { + "epoch": 0.34, + "grad_norm": 0.45703125, + "learning_rate": 0.0001861041680897795, + "loss": 2.5417, + "step": 8053 + }, + { + "epoch": 0.34, + "grad_norm": 0.43359375, + "learning_rate": 0.0001861007566822073, + "loss": 2.595, + "step": 8054 + }, + { + "epoch": 0.34, + "grad_norm": 0.453125, + "learning_rate": 0.00018609734488721317, + "loss": 2.5936, + "step": 8055 + }, + { + "epoch": 0.34, + "grad_norm": 0.466796875, + "learning_rate": 0.00018609393270481241, + "loss": 2.5991, + "step": 8056 + }, + { + "epoch": 0.34, + "grad_norm": 0.490234375, + "learning_rate": 0.00018609052013502045, + "loss": 2.5303, + "step": 8057 + }, + { + "epoch": 0.34, + "grad_norm": 0.435546875, + "learning_rate": 0.00018608710717785262, + "loss": 2.5352, + "step": 8058 + }, + { + "epoch": 0.34, + "grad_norm": 0.41796875, + "learning_rate": 0.00018608369383332426, + "loss": 2.6142, + "step": 8059 + }, + { + "epoch": 0.34, + "grad_norm": 0.431640625, + "learning_rate": 0.00018608028010145075, + "loss": 2.7121, + "step": 8060 + }, + { + "epoch": 0.34, + "grad_norm": 0.4296875, + "learning_rate": 0.00018607686598224742, + "loss": 2.765, + "step": 8061 + }, + { + "epoch": 0.34, + "grad_norm": 0.55859375, + "learning_rate": 0.00018607345147572966, + "loss": 2.7408, + "step": 8062 + }, + { + "epoch": 0.34, + "grad_norm": 0.421875, + "learning_rate": 0.00018607003658191283, + "loss": 2.58, + "step": 8063 + }, + { + "epoch": 0.34, + "grad_norm": 0.455078125, + "learning_rate": 0.0001860666213008123, + "loss": 2.5222, + "step": 8064 + }, + { + "epoch": 0.34, + "grad_norm": 0.4921875, + "learning_rate": 0.0001860632056324434, + "loss": 2.7455, + "step": 8065 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.00018605978957682154, + "loss": 2.6904, + "step": 8066 + }, + { + "epoch": 0.34, + "grad_norm": 0.443359375, + "learning_rate": 0.0001860563731339621, + "loss": 2.4936, + "step": 8067 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.0001860529563038804, + "loss": 2.5997, + "step": 8068 + }, + { + "epoch": 0.34, + "grad_norm": 0.462890625, + "learning_rate": 0.00018604953908659186, + "loss": 2.5804, + "step": 8069 + }, + { + "epoch": 0.34, + "grad_norm": 0.408203125, + "learning_rate": 0.00018604612148211182, + "loss": 2.536, + "step": 8070 + }, + { + "epoch": 0.34, + "grad_norm": 0.48046875, + "learning_rate": 0.00018604270349045573, + "loss": 2.6162, + "step": 8071 + }, + { + "epoch": 0.34, + "grad_norm": 0.46484375, + "learning_rate": 0.00018603928511163887, + "loss": 2.6173, + "step": 8072 + }, + { + "epoch": 0.34, + "grad_norm": 0.490234375, + "learning_rate": 0.00018603586634567668, + "loss": 2.7208, + "step": 8073 + }, + { + "epoch": 0.34, + "grad_norm": 0.400390625, + "learning_rate": 0.00018603244719258454, + "loss": 2.5837, + "step": 8074 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.00018602902765237783, + "loss": 2.7224, + "step": 8075 + }, + { + "epoch": 0.34, + "grad_norm": 0.427734375, + "learning_rate": 0.00018602560772507193, + "loss": 2.6732, + "step": 8076 + }, + { + "epoch": 0.34, + "grad_norm": 0.39453125, + "learning_rate": 0.00018602218741068223, + "loss": 2.6499, + "step": 8077 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.00018601876670922414, + "loss": 2.5609, + "step": 8078 + }, + { + "epoch": 0.34, + "grad_norm": 0.4921875, + "learning_rate": 0.00018601534562071303, + "loss": 2.7041, + "step": 8079 + }, + { + "epoch": 0.34, + "grad_norm": 0.42578125, + "learning_rate": 0.0001860119241451643, + "loss": 2.6119, + "step": 8080 + }, + { + "epoch": 0.34, + "grad_norm": 0.408203125, + "learning_rate": 0.00018600850228259335, + "loss": 2.5136, + "step": 8081 + }, + { + "epoch": 0.34, + "grad_norm": 0.458984375, + "learning_rate": 0.00018600508003301555, + "loss": 2.7001, + "step": 8082 + }, + { + "epoch": 0.34, + "grad_norm": 0.4765625, + "learning_rate": 0.00018600165739644634, + "loss": 2.6381, + "step": 8083 + }, + { + "epoch": 0.34, + "grad_norm": 0.44921875, + "learning_rate": 0.00018599823437290112, + "loss": 2.6387, + "step": 8084 + }, + { + "epoch": 0.34, + "grad_norm": 0.435546875, + "learning_rate": 0.00018599481096239523, + "loss": 2.672, + "step": 8085 + }, + { + "epoch": 0.34, + "grad_norm": 0.447265625, + "learning_rate": 0.00018599138716494416, + "loss": 2.5758, + "step": 8086 + }, + { + "epoch": 0.34, + "grad_norm": 0.400390625, + "learning_rate": 0.00018598796298056324, + "loss": 2.7789, + "step": 8087 + }, + { + "epoch": 0.34, + "grad_norm": 0.40625, + "learning_rate": 0.00018598453840926792, + "loss": 2.5286, + "step": 8088 + }, + { + "epoch": 0.34, + "grad_norm": 0.55078125, + "learning_rate": 0.00018598111345107362, + "loss": 2.6481, + "step": 8089 + }, + { + "epoch": 0.34, + "grad_norm": 0.439453125, + "learning_rate": 0.00018597768810599572, + "loss": 2.6988, + "step": 8090 + }, + { + "epoch": 0.34, + "grad_norm": 0.55859375, + "learning_rate": 0.00018597426237404964, + "loss": 2.73, + "step": 8091 + }, + { + "epoch": 0.34, + "grad_norm": 0.51953125, + "learning_rate": 0.0001859708362552508, + "loss": 2.6858, + "step": 8092 + }, + { + "epoch": 0.34, + "grad_norm": 0.4609375, + "learning_rate": 0.00018596740974961465, + "loss": 2.6887, + "step": 8093 + }, + { + "epoch": 0.34, + "grad_norm": 0.45703125, + "learning_rate": 0.00018596398285715654, + "loss": 2.549, + "step": 8094 + }, + { + "epoch": 0.34, + "grad_norm": 0.45703125, + "learning_rate": 0.00018596055557789192, + "loss": 2.7304, + "step": 8095 + }, + { + "epoch": 0.34, + "grad_norm": 0.45703125, + "learning_rate": 0.00018595712791183623, + "loss": 2.6044, + "step": 8096 + }, + { + "epoch": 0.35, + "grad_norm": 0.4140625, + "learning_rate": 0.00018595369985900483, + "loss": 2.5916, + "step": 8097 + }, + { + "epoch": 0.35, + "grad_norm": 0.40625, + "learning_rate": 0.00018595027141941322, + "loss": 2.812, + "step": 8098 + }, + { + "epoch": 0.35, + "grad_norm": 0.46875, + "learning_rate": 0.00018594684259307684, + "loss": 2.6248, + "step": 8099 + }, + { + "epoch": 0.35, + "grad_norm": 0.455078125, + "learning_rate": 0.00018594341338001104, + "loss": 2.7876, + "step": 8100 + }, + { + "epoch": 0.35, + "grad_norm": 0.4375, + "learning_rate": 0.0001859399837802313, + "loss": 2.7165, + "step": 8101 + }, + { + "epoch": 0.35, + "grad_norm": 0.453125, + "learning_rate": 0.00018593655379375306, + "loss": 2.5884, + "step": 8102 + }, + { + "epoch": 0.35, + "grad_norm": 0.466796875, + "learning_rate": 0.0001859331234205917, + "loss": 2.7795, + "step": 8103 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018592969266076272, + "loss": 2.4588, + "step": 8104 + }, + { + "epoch": 0.35, + "grad_norm": 0.439453125, + "learning_rate": 0.0001859262615142815, + "loss": 2.5524, + "step": 8105 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.0001859228299811635, + "loss": 2.5439, + "step": 8106 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018591939806142418, + "loss": 2.5419, + "step": 8107 + }, + { + "epoch": 0.35, + "grad_norm": 0.41796875, + "learning_rate": 0.000185915965755079, + "loss": 2.4953, + "step": 8108 + }, + { + "epoch": 0.35, + "grad_norm": 0.453125, + "learning_rate": 0.00018591253306214333, + "loss": 2.7168, + "step": 8109 + }, + { + "epoch": 0.35, + "grad_norm": 0.427734375, + "learning_rate": 0.00018590909998263269, + "loss": 2.683, + "step": 8110 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.00018590566651656247, + "loss": 2.7012, + "step": 8111 + }, + { + "epoch": 0.35, + "grad_norm": 0.451171875, + "learning_rate": 0.00018590223266394814, + "loss": 2.6704, + "step": 8112 + }, + { + "epoch": 0.35, + "grad_norm": 0.404296875, + "learning_rate": 0.00018589879842480518, + "loss": 2.5387, + "step": 8113 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.000185895363799149, + "loss": 2.677, + "step": 8114 + }, + { + "epoch": 0.35, + "grad_norm": 0.46875, + "learning_rate": 0.00018589192878699508, + "loss": 2.7079, + "step": 8115 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.0001858884933883589, + "loss": 2.6051, + "step": 8116 + }, + { + "epoch": 0.35, + "grad_norm": 0.48046875, + "learning_rate": 0.00018588505760325585, + "loss": 2.6514, + "step": 8117 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.00018588162143170144, + "loss": 2.7249, + "step": 8118 + }, + { + "epoch": 0.35, + "grad_norm": 0.4609375, + "learning_rate": 0.00018587818487371112, + "loss": 2.5953, + "step": 8119 + }, + { + "epoch": 0.35, + "grad_norm": 0.40234375, + "learning_rate": 0.00018587474792930035, + "loss": 2.5628, + "step": 8120 + }, + { + "epoch": 0.35, + "grad_norm": 0.443359375, + "learning_rate": 0.00018587131059848462, + "loss": 2.6245, + "step": 8121 + }, + { + "epoch": 0.35, + "grad_norm": 0.45703125, + "learning_rate": 0.00018586787288127937, + "loss": 2.6448, + "step": 8122 + }, + { + "epoch": 0.35, + "grad_norm": 0.40625, + "learning_rate": 0.0001858644347777, + "loss": 2.3905, + "step": 8123 + }, + { + "epoch": 0.35, + "grad_norm": 0.439453125, + "learning_rate": 0.00018586099628776215, + "loss": 2.5861, + "step": 8124 + }, + { + "epoch": 0.35, + "grad_norm": 0.466796875, + "learning_rate": 0.00018585755741148114, + "loss": 2.6456, + "step": 8125 + }, + { + "epoch": 0.35, + "grad_norm": 0.453125, + "learning_rate": 0.00018585411814887252, + "loss": 2.7197, + "step": 8126 + }, + { + "epoch": 0.35, + "grad_norm": 0.484375, + "learning_rate": 0.00018585067849995174, + "loss": 2.573, + "step": 8127 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018584723846473429, + "loss": 2.5901, + "step": 8128 + }, + { + "epoch": 0.35, + "grad_norm": 0.439453125, + "learning_rate": 0.00018584379804323565, + "loss": 2.639, + "step": 8129 + }, + { + "epoch": 0.35, + "grad_norm": 0.455078125, + "learning_rate": 0.00018584035723547126, + "loss": 2.5662, + "step": 8130 + }, + { + "epoch": 0.35, + "grad_norm": 0.52734375, + "learning_rate": 0.00018583691604145665, + "loss": 2.8482, + "step": 8131 + }, + { + "epoch": 0.35, + "grad_norm": 0.4375, + "learning_rate": 0.0001858334744612073, + "loss": 2.5971, + "step": 8132 + }, + { + "epoch": 0.35, + "grad_norm": 0.40625, + "learning_rate": 0.00018583003249473867, + "loss": 2.5419, + "step": 8133 + }, + { + "epoch": 0.35, + "grad_norm": 0.482421875, + "learning_rate": 0.00018582659014206626, + "loss": 2.6162, + "step": 8134 + }, + { + "epoch": 0.35, + "grad_norm": 0.44140625, + "learning_rate": 0.00018582314740320555, + "loss": 2.4176, + "step": 8135 + }, + { + "epoch": 0.35, + "grad_norm": 0.48828125, + "learning_rate": 0.00018581970427817207, + "loss": 2.5883, + "step": 8136 + }, + { + "epoch": 0.35, + "grad_norm": 0.427734375, + "learning_rate": 0.00018581626076698125, + "loss": 2.6221, + "step": 8137 + }, + { + "epoch": 0.35, + "grad_norm": 0.4609375, + "learning_rate": 0.00018581281686964865, + "loss": 2.5564, + "step": 8138 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.0001858093725861897, + "loss": 2.5944, + "step": 8139 + }, + { + "epoch": 0.35, + "grad_norm": 0.45703125, + "learning_rate": 0.00018580592791661994, + "loss": 2.606, + "step": 8140 + }, + { + "epoch": 0.35, + "grad_norm": 0.4140625, + "learning_rate": 0.0001858024828609549, + "loss": 2.6307, + "step": 8141 + }, + { + "epoch": 0.35, + "grad_norm": 0.43359375, + "learning_rate": 0.00018579903741921, + "loss": 2.5448, + "step": 8142 + }, + { + "epoch": 0.35, + "grad_norm": 0.466796875, + "learning_rate": 0.00018579559159140083, + "loss": 2.5835, + "step": 8143 + }, + { + "epoch": 0.35, + "grad_norm": 0.421875, + "learning_rate": 0.00018579214537754282, + "loss": 2.4952, + "step": 8144 + }, + { + "epoch": 0.35, + "grad_norm": 0.4921875, + "learning_rate": 0.00018578869877765154, + "loss": 2.6476, + "step": 8145 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018578525179174246, + "loss": 2.5577, + "step": 8146 + }, + { + "epoch": 0.35, + "grad_norm": 0.5078125, + "learning_rate": 0.00018578180441983106, + "loss": 2.5999, + "step": 8147 + }, + { + "epoch": 0.35, + "grad_norm": 0.439453125, + "learning_rate": 0.00018577835666193292, + "loss": 2.4885, + "step": 8148 + }, + { + "epoch": 0.35, + "grad_norm": 0.478515625, + "learning_rate": 0.00018577490851806353, + "loss": 2.6662, + "step": 8149 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.00018577145998823838, + "loss": 2.6025, + "step": 8150 + }, + { + "epoch": 0.35, + "grad_norm": 0.3828125, + "learning_rate": 0.00018576801107247302, + "loss": 2.5997, + "step": 8151 + }, + { + "epoch": 0.35, + "grad_norm": 0.5078125, + "learning_rate": 0.0001857645617707829, + "loss": 2.6657, + "step": 8152 + }, + { + "epoch": 0.35, + "grad_norm": 0.478515625, + "learning_rate": 0.00018576111208318366, + "loss": 2.6098, + "step": 8153 + }, + { + "epoch": 0.35, + "grad_norm": 0.462890625, + "learning_rate": 0.00018575766200969076, + "loss": 2.6402, + "step": 8154 + }, + { + "epoch": 0.35, + "grad_norm": 0.486328125, + "learning_rate": 0.00018575421155031969, + "loss": 2.5797, + "step": 8155 + }, + { + "epoch": 0.35, + "grad_norm": 0.392578125, + "learning_rate": 0.00018575076070508604, + "loss": 2.543, + "step": 8156 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018574730947400528, + "loss": 2.5946, + "step": 8157 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.00018574385785709297, + "loss": 2.7331, + "step": 8158 + }, + { + "epoch": 0.35, + "grad_norm": 0.55859375, + "learning_rate": 0.00018574040585436462, + "loss": 2.6708, + "step": 8159 + }, + { + "epoch": 0.35, + "grad_norm": 0.52734375, + "learning_rate": 0.0001857369534658358, + "loss": 2.5761, + "step": 8160 + }, + { + "epoch": 0.35, + "grad_norm": 0.44921875, + "learning_rate": 0.00018573350069152203, + "loss": 2.6821, + "step": 8161 + }, + { + "epoch": 0.35, + "grad_norm": 0.484375, + "learning_rate": 0.00018573004753143883, + "loss": 2.8052, + "step": 8162 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018572659398560174, + "loss": 2.6771, + "step": 8163 + }, + { + "epoch": 0.35, + "grad_norm": 0.5, + "learning_rate": 0.00018572314005402632, + "loss": 2.6881, + "step": 8164 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.00018571968573672808, + "loss": 2.4968, + "step": 8165 + }, + { + "epoch": 0.35, + "grad_norm": 0.419921875, + "learning_rate": 0.0001857162310337226, + "loss": 2.6616, + "step": 8166 + }, + { + "epoch": 0.35, + "grad_norm": 0.490234375, + "learning_rate": 0.00018571277594502537, + "loss": 2.6076, + "step": 8167 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.00018570932047065202, + "loss": 2.5448, + "step": 8168 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.000185705864610618, + "loss": 2.694, + "step": 8169 + }, + { + "epoch": 0.35, + "grad_norm": 0.4765625, + "learning_rate": 0.00018570240836493894, + "loss": 2.5659, + "step": 8170 + }, + { + "epoch": 0.35, + "grad_norm": 0.48046875, + "learning_rate": 0.00018569895173363036, + "loss": 2.7984, + "step": 8171 + }, + { + "epoch": 0.35, + "grad_norm": 0.40625, + "learning_rate": 0.00018569549471670783, + "loss": 2.7946, + "step": 8172 + }, + { + "epoch": 0.35, + "grad_norm": 0.42578125, + "learning_rate": 0.0001856920373141869, + "loss": 2.6805, + "step": 8173 + }, + { + "epoch": 0.35, + "grad_norm": 0.431640625, + "learning_rate": 0.00018568857952608306, + "loss": 2.5938, + "step": 8174 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.000185685121352412, + "loss": 2.5824, + "step": 8175 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.00018568166279318916, + "loss": 2.6516, + "step": 8176 + }, + { + "epoch": 0.35, + "grad_norm": 0.45703125, + "learning_rate": 0.00018567820384843018, + "loss": 2.5484, + "step": 8177 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.0001856747445181506, + "loss": 2.5882, + "step": 8178 + }, + { + "epoch": 0.35, + "grad_norm": 0.412109375, + "learning_rate": 0.00018567128480236593, + "loss": 2.6878, + "step": 8179 + }, + { + "epoch": 0.35, + "grad_norm": 0.44921875, + "learning_rate": 0.00018566782470109182, + "loss": 2.6042, + "step": 8180 + }, + { + "epoch": 0.35, + "grad_norm": 0.46875, + "learning_rate": 0.00018566436421434381, + "loss": 2.6758, + "step": 8181 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.00018566090334213746, + "loss": 2.5929, + "step": 8182 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.00018565744208448836, + "loss": 2.6436, + "step": 8183 + }, + { + "epoch": 0.35, + "grad_norm": 0.46875, + "learning_rate": 0.00018565398044141206, + "loss": 2.5876, + "step": 8184 + }, + { + "epoch": 0.35, + "grad_norm": 0.466796875, + "learning_rate": 0.00018565051841292415, + "loss": 2.6494, + "step": 8185 + }, + { + "epoch": 0.35, + "grad_norm": 0.462890625, + "learning_rate": 0.00018564705599904025, + "loss": 2.5518, + "step": 8186 + }, + { + "epoch": 0.35, + "grad_norm": 0.44921875, + "learning_rate": 0.00018564359319977585, + "loss": 2.7445, + "step": 8187 + }, + { + "epoch": 0.35, + "grad_norm": 0.515625, + "learning_rate": 0.0001856401300151466, + "loss": 2.6868, + "step": 8188 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018563666644516805, + "loss": 2.4898, + "step": 8189 + }, + { + "epoch": 0.35, + "grad_norm": 0.4296875, + "learning_rate": 0.00018563320248985582, + "loss": 2.6868, + "step": 8190 + }, + { + "epoch": 0.35, + "grad_norm": 0.431640625, + "learning_rate": 0.00018562973814922545, + "loss": 2.5738, + "step": 8191 + }, + { + "epoch": 0.35, + "grad_norm": 0.5078125, + "learning_rate": 0.00018562627342329257, + "loss": 2.7773, + "step": 8192 + }, + { + "epoch": 0.35, + "grad_norm": 0.4140625, + "learning_rate": 0.00018562280831207274, + "loss": 2.65, + "step": 8193 + }, + { + "epoch": 0.35, + "grad_norm": 0.470703125, + "learning_rate": 0.00018561934281558157, + "loss": 2.5907, + "step": 8194 + }, + { + "epoch": 0.35, + "grad_norm": 0.46484375, + "learning_rate": 0.00018561587693383466, + "loss": 2.6952, + "step": 8195 + }, + { + "epoch": 0.35, + "grad_norm": 0.423828125, + "learning_rate": 0.00018561241066684757, + "loss": 2.724, + "step": 8196 + }, + { + "epoch": 0.35, + "grad_norm": 0.46484375, + "learning_rate": 0.00018560894401463594, + "loss": 2.6065, + "step": 8197 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.00018560547697721533, + "loss": 2.6557, + "step": 8198 + }, + { + "epoch": 0.35, + "grad_norm": 0.47265625, + "learning_rate": 0.00018560200955460138, + "loss": 2.5726, + "step": 8199 + }, + { + "epoch": 0.35, + "grad_norm": 0.4921875, + "learning_rate": 0.00018559854174680967, + "loss": 2.5857, + "step": 8200 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.00018559507355385576, + "loss": 2.7191, + "step": 8201 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018559160497575534, + "loss": 2.77, + "step": 8202 + }, + { + "epoch": 0.35, + "grad_norm": 0.40234375, + "learning_rate": 0.00018558813601252396, + "loss": 2.5787, + "step": 8203 + }, + { + "epoch": 0.35, + "grad_norm": 0.421875, + "learning_rate": 0.00018558466666417726, + "loss": 2.6539, + "step": 8204 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.00018558119693073085, + "loss": 2.5588, + "step": 8205 + }, + { + "epoch": 0.35, + "grad_norm": 0.431640625, + "learning_rate": 0.00018557772681220032, + "loss": 2.693, + "step": 8206 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018557425630860128, + "loss": 2.6129, + "step": 8207 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.00018557078541994938, + "loss": 2.5915, + "step": 8208 + }, + { + "epoch": 0.35, + "grad_norm": 0.44921875, + "learning_rate": 0.0001855673141462602, + "loss": 2.7112, + "step": 8209 + }, + { + "epoch": 0.35, + "grad_norm": 0.427734375, + "learning_rate": 0.0001855638424875494, + "loss": 2.7243, + "step": 8210 + }, + { + "epoch": 0.35, + "grad_norm": 0.40625, + "learning_rate": 0.00018556037044383257, + "loss": 2.586, + "step": 8211 + }, + { + "epoch": 0.35, + "grad_norm": 0.4296875, + "learning_rate": 0.0001855568980151253, + "loss": 2.5465, + "step": 8212 + }, + { + "epoch": 0.35, + "grad_norm": 0.443359375, + "learning_rate": 0.0001855534252014433, + "loss": 2.7048, + "step": 8213 + }, + { + "epoch": 0.35, + "grad_norm": 0.5078125, + "learning_rate": 0.00018554995200280216, + "loss": 2.6639, + "step": 8214 + }, + { + "epoch": 0.35, + "grad_norm": 0.486328125, + "learning_rate": 0.0001855464784192175, + "loss": 2.6465, + "step": 8215 + }, + { + "epoch": 0.35, + "grad_norm": 0.486328125, + "learning_rate": 0.00018554300445070495, + "loss": 2.5136, + "step": 8216 + }, + { + "epoch": 0.35, + "grad_norm": 0.48046875, + "learning_rate": 0.00018553953009728012, + "loss": 2.6849, + "step": 8217 + }, + { + "epoch": 0.35, + "grad_norm": 0.451171875, + "learning_rate": 0.0001855360553589587, + "loss": 2.7126, + "step": 8218 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.00018553258023575625, + "loss": 2.7143, + "step": 8219 + }, + { + "epoch": 0.35, + "grad_norm": 0.4765625, + "learning_rate": 0.00018552910472768844, + "loss": 2.6413, + "step": 8220 + }, + { + "epoch": 0.35, + "grad_norm": 0.4296875, + "learning_rate": 0.00018552562883477095, + "loss": 2.5087, + "step": 8221 + }, + { + "epoch": 0.35, + "grad_norm": 0.490234375, + "learning_rate": 0.0001855221525570194, + "loss": 2.7208, + "step": 8222 + }, + { + "epoch": 0.35, + "grad_norm": 0.439453125, + "learning_rate": 0.00018551867589444937, + "loss": 2.6988, + "step": 8223 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018551519884707658, + "loss": 2.63, + "step": 8224 + }, + { + "epoch": 0.35, + "grad_norm": 0.45703125, + "learning_rate": 0.00018551172141491665, + "loss": 2.6649, + "step": 8225 + }, + { + "epoch": 0.35, + "grad_norm": 0.40234375, + "learning_rate": 0.0001855082435979852, + "loss": 2.5063, + "step": 8226 + }, + { + "epoch": 0.35, + "grad_norm": 0.55078125, + "learning_rate": 0.00018550476539629793, + "loss": 2.5627, + "step": 8227 + }, + { + "epoch": 0.35, + "grad_norm": 0.45703125, + "learning_rate": 0.00018550128680987047, + "loss": 2.5785, + "step": 8228 + }, + { + "epoch": 0.35, + "grad_norm": 0.462890625, + "learning_rate": 0.00018549780783871843, + "loss": 2.6426, + "step": 8229 + }, + { + "epoch": 0.35, + "grad_norm": 0.51171875, + "learning_rate": 0.00018549432848285756, + "loss": 2.5186, + "step": 8230 + }, + { + "epoch": 0.35, + "grad_norm": 0.419921875, + "learning_rate": 0.0001854908487423034, + "loss": 2.4323, + "step": 8231 + }, + { + "epoch": 0.35, + "grad_norm": 0.478515625, + "learning_rate": 0.00018548736861707168, + "loss": 2.7138, + "step": 8232 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.00018548388810717806, + "loss": 2.6774, + "step": 8233 + }, + { + "epoch": 0.35, + "grad_norm": 0.443359375, + "learning_rate": 0.00018548040721263814, + "loss": 2.6603, + "step": 8234 + }, + { + "epoch": 0.35, + "grad_norm": 0.46875, + "learning_rate": 0.0001854769259334677, + "loss": 2.7694, + "step": 8235 + }, + { + "epoch": 0.35, + "grad_norm": 0.404296875, + "learning_rate": 0.00018547344426968229, + "loss": 2.49, + "step": 8236 + }, + { + "epoch": 0.35, + "grad_norm": 0.53515625, + "learning_rate": 0.0001854699622212976, + "loss": 2.6161, + "step": 8237 + }, + { + "epoch": 0.35, + "grad_norm": 0.421875, + "learning_rate": 0.00018546647978832934, + "loss": 2.5933, + "step": 8238 + }, + { + "epoch": 0.35, + "grad_norm": 0.462890625, + "learning_rate": 0.00018546299697079315, + "loss": 2.7261, + "step": 8239 + }, + { + "epoch": 0.35, + "grad_norm": 0.474609375, + "learning_rate": 0.0001854595137687047, + "loss": 2.7036, + "step": 8240 + }, + { + "epoch": 0.35, + "grad_norm": 0.470703125, + "learning_rate": 0.00018545603018207968, + "loss": 2.6547, + "step": 8241 + }, + { + "epoch": 0.35, + "grad_norm": 0.490234375, + "learning_rate": 0.00018545254621093373, + "loss": 2.7255, + "step": 8242 + }, + { + "epoch": 0.35, + "grad_norm": 0.474609375, + "learning_rate": 0.00018544906185528258, + "loss": 2.6544, + "step": 8243 + }, + { + "epoch": 0.35, + "grad_norm": 0.4296875, + "learning_rate": 0.00018544557711514187, + "loss": 2.7091, + "step": 8244 + }, + { + "epoch": 0.35, + "grad_norm": 0.484375, + "learning_rate": 0.0001854420919905273, + "loss": 2.7219, + "step": 8245 + }, + { + "epoch": 0.35, + "grad_norm": 0.49609375, + "learning_rate": 0.00018543860648145453, + "loss": 2.8023, + "step": 8246 + }, + { + "epoch": 0.35, + "grad_norm": 0.39453125, + "learning_rate": 0.00018543512058793925, + "loss": 2.5022, + "step": 8247 + }, + { + "epoch": 0.35, + "grad_norm": 0.462890625, + "learning_rate": 0.00018543163430999717, + "loss": 2.5851, + "step": 8248 + }, + { + "epoch": 0.35, + "grad_norm": 0.408203125, + "learning_rate": 0.00018542814764764396, + "loss": 2.5062, + "step": 8249 + }, + { + "epoch": 0.35, + "grad_norm": 0.478515625, + "learning_rate": 0.00018542466060089528, + "loss": 2.3698, + "step": 8250 + }, + { + "epoch": 0.35, + "grad_norm": 0.462890625, + "learning_rate": 0.00018542117316976687, + "loss": 2.5443, + "step": 8251 + }, + { + "epoch": 0.35, + "grad_norm": 0.50390625, + "learning_rate": 0.00018541768535427437, + "loss": 2.6796, + "step": 8252 + }, + { + "epoch": 0.35, + "grad_norm": 0.515625, + "learning_rate": 0.00018541419715443352, + "loss": 2.6277, + "step": 8253 + }, + { + "epoch": 0.35, + "grad_norm": 0.490234375, + "learning_rate": 0.00018541070857026, + "loss": 2.6053, + "step": 8254 + }, + { + "epoch": 0.35, + "grad_norm": 0.3984375, + "learning_rate": 0.0001854072196017695, + "loss": 2.5279, + "step": 8255 + }, + { + "epoch": 0.35, + "grad_norm": 0.4765625, + "learning_rate": 0.00018540373024897772, + "loss": 2.7677, + "step": 8256 + }, + { + "epoch": 0.35, + "grad_norm": 0.486328125, + "learning_rate": 0.00018540024051190036, + "loss": 2.6586, + "step": 8257 + }, + { + "epoch": 0.35, + "grad_norm": 0.41796875, + "learning_rate": 0.00018539675039055316, + "loss": 2.6833, + "step": 8258 + }, + { + "epoch": 0.35, + "grad_norm": 0.494140625, + "learning_rate": 0.00018539325988495178, + "loss": 2.6742, + "step": 8259 + }, + { + "epoch": 0.35, + "grad_norm": 0.3984375, + "learning_rate": 0.00018538976899511194, + "loss": 2.7231, + "step": 8260 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.00018538627772104934, + "loss": 2.6139, + "step": 8261 + }, + { + "epoch": 0.35, + "grad_norm": 0.4140625, + "learning_rate": 0.00018538278606277967, + "loss": 2.7045, + "step": 8262 + }, + { + "epoch": 0.35, + "grad_norm": 0.498046875, + "learning_rate": 0.00018537929402031867, + "loss": 2.7845, + "step": 8263 + }, + { + "epoch": 0.35, + "grad_norm": 0.4296875, + "learning_rate": 0.00018537580159368208, + "loss": 2.5115, + "step": 8264 + }, + { + "epoch": 0.35, + "grad_norm": 0.51953125, + "learning_rate": 0.00018537230878288557, + "loss": 2.6346, + "step": 8265 + }, + { + "epoch": 0.35, + "grad_norm": 0.41796875, + "learning_rate": 0.0001853688155879449, + "loss": 2.5635, + "step": 8266 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.0001853653220088757, + "loss": 2.609, + "step": 8267 + }, + { + "epoch": 0.35, + "grad_norm": 0.482421875, + "learning_rate": 0.00018536182804569382, + "loss": 2.6903, + "step": 8268 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.00018535833369841487, + "loss": 2.697, + "step": 8269 + }, + { + "epoch": 0.35, + "grad_norm": 0.455078125, + "learning_rate": 0.00018535483896705462, + "loss": 2.6964, + "step": 8270 + }, + { + "epoch": 0.35, + "grad_norm": 0.4296875, + "learning_rate": 0.00018535134385162875, + "loss": 2.6391, + "step": 8271 + }, + { + "epoch": 0.35, + "grad_norm": 0.453125, + "learning_rate": 0.00018534784835215305, + "loss": 2.6252, + "step": 8272 + }, + { + "epoch": 0.35, + "grad_norm": 0.51953125, + "learning_rate": 0.00018534435246864327, + "loss": 2.6245, + "step": 8273 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.00018534085620111504, + "loss": 2.5402, + "step": 8274 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.00018533735954958416, + "loss": 2.6732, + "step": 8275 + }, + { + "epoch": 0.35, + "grad_norm": 0.4609375, + "learning_rate": 0.00018533386251406634, + "loss": 2.6243, + "step": 8276 + }, + { + "epoch": 0.35, + "grad_norm": 0.453125, + "learning_rate": 0.00018533036509457734, + "loss": 2.6228, + "step": 8277 + }, + { + "epoch": 0.35, + "grad_norm": 0.44921875, + "learning_rate": 0.00018532686729113285, + "loss": 2.5837, + "step": 8278 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.0001853233691037487, + "loss": 2.572, + "step": 8279 + }, + { + "epoch": 0.35, + "grad_norm": 0.490234375, + "learning_rate": 0.00018531987053244052, + "loss": 2.6826, + "step": 8280 + }, + { + "epoch": 0.35, + "grad_norm": 0.478515625, + "learning_rate": 0.0001853163715772241, + "loss": 2.684, + "step": 8281 + }, + { + "epoch": 0.35, + "grad_norm": 0.478515625, + "learning_rate": 0.0001853128722381152, + "loss": 2.5856, + "step": 8282 + }, + { + "epoch": 0.35, + "grad_norm": 0.451171875, + "learning_rate": 0.0001853093725151296, + "loss": 2.6562, + "step": 8283 + }, + { + "epoch": 0.35, + "grad_norm": 0.490234375, + "learning_rate": 0.00018530587240828296, + "loss": 2.5883, + "step": 8284 + }, + { + "epoch": 0.35, + "grad_norm": 0.50390625, + "learning_rate": 0.00018530237191759105, + "loss": 2.7402, + "step": 8285 + }, + { + "epoch": 0.35, + "grad_norm": 0.474609375, + "learning_rate": 0.00018529887104306968, + "loss": 2.7587, + "step": 8286 + }, + { + "epoch": 0.35, + "grad_norm": 0.40234375, + "learning_rate": 0.00018529536978473455, + "loss": 2.5801, + "step": 8287 + }, + { + "epoch": 0.35, + "grad_norm": 0.443359375, + "learning_rate": 0.00018529186814260143, + "loss": 2.6496, + "step": 8288 + }, + { + "epoch": 0.35, + "grad_norm": 0.43359375, + "learning_rate": 0.00018528836611668605, + "loss": 2.5539, + "step": 8289 + }, + { + "epoch": 0.35, + "grad_norm": 0.3984375, + "learning_rate": 0.00018528486370700422, + "loss": 2.5603, + "step": 8290 + }, + { + "epoch": 0.35, + "grad_norm": 0.412109375, + "learning_rate": 0.00018528136091357166, + "loss": 2.6501, + "step": 8291 + }, + { + "epoch": 0.35, + "grad_norm": 0.4140625, + "learning_rate": 0.00018527785773640414, + "loss": 2.5052, + "step": 8292 + }, + { + "epoch": 0.35, + "grad_norm": 0.44921875, + "learning_rate": 0.00018527435417551745, + "loss": 2.6427, + "step": 8293 + }, + { + "epoch": 0.35, + "grad_norm": 0.427734375, + "learning_rate": 0.00018527085023092727, + "loss": 2.5327, + "step": 8294 + }, + { + "epoch": 0.35, + "grad_norm": 0.478515625, + "learning_rate": 0.0001852673459026495, + "loss": 2.5776, + "step": 8295 + }, + { + "epoch": 0.35, + "grad_norm": 0.4296875, + "learning_rate": 0.00018526384119069977, + "loss": 2.5074, + "step": 8296 + }, + { + "epoch": 0.35, + "grad_norm": 0.46484375, + "learning_rate": 0.00018526033609509397, + "loss": 2.7573, + "step": 8297 + }, + { + "epoch": 0.35, + "grad_norm": 0.412109375, + "learning_rate": 0.0001852568306158478, + "loss": 2.5238, + "step": 8298 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.00018525332475297703, + "loss": 2.4791, + "step": 8299 + }, + { + "epoch": 0.35, + "grad_norm": 0.427734375, + "learning_rate": 0.00018524981850649748, + "loss": 2.7273, + "step": 8300 + }, + { + "epoch": 0.35, + "grad_norm": 0.419921875, + "learning_rate": 0.00018524631187642488, + "loss": 2.7169, + "step": 8301 + }, + { + "epoch": 0.35, + "grad_norm": 0.458984375, + "learning_rate": 0.00018524280486277505, + "loss": 2.7129, + "step": 8302 + }, + { + "epoch": 0.35, + "grad_norm": 0.423828125, + "learning_rate": 0.00018523929746556373, + "loss": 2.6029, + "step": 8303 + }, + { + "epoch": 0.35, + "grad_norm": 0.4140625, + "learning_rate": 0.00018523578968480673, + "loss": 2.7143, + "step": 8304 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.00018523228152051983, + "loss": 2.4809, + "step": 8305 + }, + { + "epoch": 0.35, + "grad_norm": 0.5078125, + "learning_rate": 0.0001852287729727188, + "loss": 2.6823, + "step": 8306 + }, + { + "epoch": 0.35, + "grad_norm": 0.4375, + "learning_rate": 0.00018522526404141947, + "loss": 2.6287, + "step": 8307 + }, + { + "epoch": 0.35, + "grad_norm": 0.435546875, + "learning_rate": 0.00018522175472663758, + "loss": 2.5803, + "step": 8308 + }, + { + "epoch": 0.35, + "grad_norm": 0.423828125, + "learning_rate": 0.00018521824502838892, + "loss": 2.4739, + "step": 8309 + }, + { + "epoch": 0.35, + "grad_norm": 0.44140625, + "learning_rate": 0.00018521473494668931, + "loss": 2.6361, + "step": 8310 + }, + { + "epoch": 0.35, + "grad_norm": 0.443359375, + "learning_rate": 0.00018521122448155456, + "loss": 2.7943, + "step": 8311 + }, + { + "epoch": 0.35, + "grad_norm": 0.4609375, + "learning_rate": 0.00018520771363300043, + "loss": 2.7959, + "step": 8312 + }, + { + "epoch": 0.35, + "grad_norm": 0.451171875, + "learning_rate": 0.00018520420240104274, + "loss": 2.6889, + "step": 8313 + }, + { + "epoch": 0.35, + "grad_norm": 0.41796875, + "learning_rate": 0.00018520069078569725, + "loss": 2.6618, + "step": 8314 + }, + { + "epoch": 0.35, + "grad_norm": 0.453125, + "learning_rate": 0.00018519717878697983, + "loss": 2.6315, + "step": 8315 + }, + { + "epoch": 0.35, + "grad_norm": 0.421875, + "learning_rate": 0.0001851936664049062, + "loss": 2.6352, + "step": 8316 + }, + { + "epoch": 0.35, + "grad_norm": 0.451171875, + "learning_rate": 0.00018519015363949222, + "loss": 2.5445, + "step": 8317 + }, + { + "epoch": 0.35, + "grad_norm": 0.4453125, + "learning_rate": 0.00018518664049075367, + "loss": 2.6678, + "step": 8318 + }, + { + "epoch": 0.35, + "grad_norm": 0.470703125, + "learning_rate": 0.00018518312695870638, + "loss": 2.5701, + "step": 8319 + }, + { + "epoch": 0.35, + "grad_norm": 0.494140625, + "learning_rate": 0.00018517961304336615, + "loss": 2.7076, + "step": 8320 + }, + { + "epoch": 0.35, + "grad_norm": 0.40625, + "learning_rate": 0.0001851760987447488, + "loss": 2.5753, + "step": 8321 + }, + { + "epoch": 0.35, + "grad_norm": 0.451171875, + "learning_rate": 0.0001851725840628701, + "loss": 2.7067, + "step": 8322 + }, + { + "epoch": 0.35, + "grad_norm": 0.447265625, + "learning_rate": 0.00018516906899774594, + "loss": 2.4466, + "step": 8323 + }, + { + "epoch": 0.35, + "grad_norm": 0.45703125, + "learning_rate": 0.00018516555354939208, + "loss": 2.6073, + "step": 8324 + }, + { + "epoch": 0.35, + "grad_norm": 0.47265625, + "learning_rate": 0.00018516203771782433, + "loss": 2.5845, + "step": 8325 + }, + { + "epoch": 0.35, + "grad_norm": 0.455078125, + "learning_rate": 0.00018515852150305858, + "loss": 2.6684, + "step": 8326 + }, + { + "epoch": 0.35, + "grad_norm": 0.455078125, + "learning_rate": 0.0001851550049051106, + "loss": 2.5935, + "step": 8327 + }, + { + "epoch": 0.35, + "grad_norm": 0.4609375, + "learning_rate": 0.00018515148792399617, + "loss": 2.7405, + "step": 8328 + }, + { + "epoch": 0.35, + "grad_norm": 0.416015625, + "learning_rate": 0.0001851479705597312, + "loss": 2.5654, + "step": 8329 + }, + { + "epoch": 0.35, + "grad_norm": 0.51171875, + "learning_rate": 0.00018514445281233148, + "loss": 2.5656, + "step": 8330 + }, + { + "epoch": 0.36, + "grad_norm": 0.466796875, + "learning_rate": 0.00018514093468181283, + "loss": 2.7023, + "step": 8331 + }, + { + "epoch": 0.36, + "grad_norm": 0.408203125, + "learning_rate": 0.0001851374161681911, + "loss": 2.5977, + "step": 8332 + }, + { + "epoch": 0.36, + "grad_norm": 0.451171875, + "learning_rate": 0.00018513389727148208, + "loss": 2.7112, + "step": 8333 + }, + { + "epoch": 0.36, + "grad_norm": 0.439453125, + "learning_rate": 0.00018513037799170167, + "loss": 2.598, + "step": 8334 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.00018512685832886567, + "loss": 2.6013, + "step": 8335 + }, + { + "epoch": 0.36, + "grad_norm": 0.478515625, + "learning_rate": 0.0001851233382829899, + "loss": 2.4824, + "step": 8336 + }, + { + "epoch": 0.36, + "grad_norm": 0.45703125, + "learning_rate": 0.00018511981785409022, + "loss": 2.629, + "step": 8337 + }, + { + "epoch": 0.36, + "grad_norm": 0.48046875, + "learning_rate": 0.00018511629704218247, + "loss": 2.5955, + "step": 8338 + }, + { + "epoch": 0.36, + "grad_norm": 0.4765625, + "learning_rate": 0.0001851127758472825, + "loss": 2.8321, + "step": 8339 + }, + { + "epoch": 0.36, + "grad_norm": 0.4140625, + "learning_rate": 0.00018510925426940613, + "loss": 2.5109, + "step": 8340 + }, + { + "epoch": 0.36, + "grad_norm": 0.41015625, + "learning_rate": 0.00018510573230856922, + "loss": 2.5415, + "step": 8341 + }, + { + "epoch": 0.36, + "grad_norm": 0.51171875, + "learning_rate": 0.0001851022099647876, + "loss": 2.6603, + "step": 8342 + }, + { + "epoch": 0.36, + "grad_norm": 0.41796875, + "learning_rate": 0.0001850986872380772, + "loss": 2.6256, + "step": 8343 + }, + { + "epoch": 0.36, + "grad_norm": 0.470703125, + "learning_rate": 0.00018509516412845372, + "loss": 2.5777, + "step": 8344 + }, + { + "epoch": 0.36, + "grad_norm": 0.462890625, + "learning_rate": 0.00018509164063593318, + "loss": 2.6453, + "step": 8345 + }, + { + "epoch": 0.36, + "grad_norm": 0.443359375, + "learning_rate": 0.0001850881167605313, + "loss": 2.6197, + "step": 8346 + }, + { + "epoch": 0.36, + "grad_norm": 0.451171875, + "learning_rate": 0.00018508459250226404, + "loss": 2.6517, + "step": 8347 + }, + { + "epoch": 0.36, + "grad_norm": 0.484375, + "learning_rate": 0.00018508106786114717, + "loss": 2.6311, + "step": 8348 + }, + { + "epoch": 0.36, + "grad_norm": 0.482421875, + "learning_rate": 0.0001850775428371966, + "loss": 2.6289, + "step": 8349 + }, + { + "epoch": 0.36, + "grad_norm": 0.43359375, + "learning_rate": 0.0001850740174304282, + "loss": 2.6216, + "step": 8350 + }, + { + "epoch": 0.36, + "grad_norm": 0.470703125, + "learning_rate": 0.00018507049164085776, + "loss": 2.5994, + "step": 8351 + }, + { + "epoch": 0.36, + "grad_norm": 0.421875, + "learning_rate": 0.00018506696546850124, + "loss": 2.5028, + "step": 8352 + }, + { + "epoch": 0.36, + "grad_norm": 0.4296875, + "learning_rate": 0.00018506343891337444, + "loss": 2.7221, + "step": 8353 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.00018505991197549326, + "loss": 2.7081, + "step": 8354 + }, + { + "epoch": 0.36, + "grad_norm": 0.412109375, + "learning_rate": 0.00018505638465487358, + "loss": 2.6417, + "step": 8355 + }, + { + "epoch": 0.36, + "grad_norm": 0.462890625, + "learning_rate": 0.00018505285695153122, + "loss": 2.6456, + "step": 8356 + }, + { + "epoch": 0.36, + "grad_norm": 0.5390625, + "learning_rate": 0.0001850493288654821, + "loss": 2.7534, + "step": 8357 + }, + { + "epoch": 0.36, + "grad_norm": 0.39453125, + "learning_rate": 0.0001850458003967421, + "loss": 2.5162, + "step": 8358 + }, + { + "epoch": 0.36, + "grad_norm": 0.4375, + "learning_rate": 0.00018504227154532709, + "loss": 2.5833, + "step": 8359 + }, + { + "epoch": 0.36, + "grad_norm": 0.423828125, + "learning_rate": 0.0001850387423112529, + "loss": 2.6602, + "step": 8360 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.00018503521269453547, + "loss": 2.678, + "step": 8361 + }, + { + "epoch": 0.36, + "grad_norm": 0.4375, + "learning_rate": 0.00018503168269519063, + "loss": 2.7452, + "step": 8362 + }, + { + "epoch": 0.36, + "grad_norm": 0.49609375, + "learning_rate": 0.00018502815231323432, + "loss": 2.6523, + "step": 8363 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.0001850246215486824, + "loss": 2.5256, + "step": 8364 + }, + { + "epoch": 0.36, + "grad_norm": 0.44140625, + "learning_rate": 0.00018502109040155075, + "loss": 2.6454, + "step": 8365 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.00018501755887185528, + "loss": 2.6489, + "step": 8366 + }, + { + "epoch": 0.36, + "grad_norm": 0.447265625, + "learning_rate": 0.00018501402695961182, + "loss": 2.7403, + "step": 8367 + }, + { + "epoch": 0.36, + "grad_norm": 0.455078125, + "learning_rate": 0.00018501049466483635, + "loss": 2.6153, + "step": 8368 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.0001850069619875447, + "loss": 2.6454, + "step": 8369 + }, + { + "epoch": 0.36, + "grad_norm": 0.48828125, + "learning_rate": 0.00018500342892775282, + "loss": 2.6926, + "step": 8370 + }, + { + "epoch": 0.36, + "grad_norm": 0.46875, + "learning_rate": 0.00018499989548547655, + "loss": 2.7577, + "step": 8371 + }, + { + "epoch": 0.36, + "grad_norm": 0.458984375, + "learning_rate": 0.0001849963616607318, + "loss": 2.6529, + "step": 8372 + }, + { + "epoch": 0.36, + "grad_norm": 0.4765625, + "learning_rate": 0.0001849928274535345, + "loss": 2.6595, + "step": 8373 + }, + { + "epoch": 0.36, + "grad_norm": 0.5, + "learning_rate": 0.00018498929286390054, + "loss": 2.5941, + "step": 8374 + }, + { + "epoch": 0.36, + "grad_norm": 0.466796875, + "learning_rate": 0.00018498575789184583, + "loss": 2.6788, + "step": 8375 + }, + { + "epoch": 0.36, + "grad_norm": 0.44140625, + "learning_rate": 0.00018498222253738624, + "loss": 2.5625, + "step": 8376 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.00018497868680053775, + "loss": 2.668, + "step": 8377 + }, + { + "epoch": 0.36, + "grad_norm": 0.451171875, + "learning_rate": 0.0001849751506813162, + "loss": 2.5662, + "step": 8378 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.0001849716141797375, + "loss": 2.6865, + "step": 8379 + }, + { + "epoch": 0.36, + "grad_norm": 0.404296875, + "learning_rate": 0.00018496807729581765, + "loss": 2.5958, + "step": 8380 + }, + { + "epoch": 0.36, + "grad_norm": 0.498046875, + "learning_rate": 0.00018496454002957246, + "loss": 2.56, + "step": 8381 + }, + { + "epoch": 0.36, + "grad_norm": 0.466796875, + "learning_rate": 0.0001849610023810179, + "loss": 2.4957, + "step": 8382 + }, + { + "epoch": 0.36, + "grad_norm": 0.51953125, + "learning_rate": 0.00018495746435016989, + "loss": 2.709, + "step": 8383 + }, + { + "epoch": 0.36, + "grad_norm": 0.451171875, + "learning_rate": 0.0001849539259370443, + "loss": 2.4956, + "step": 8384 + }, + { + "epoch": 0.36, + "grad_norm": 0.443359375, + "learning_rate": 0.00018495038714165716, + "loss": 2.6346, + "step": 8385 + }, + { + "epoch": 0.36, + "grad_norm": 0.42578125, + "learning_rate": 0.00018494684796402426, + "loss": 2.5012, + "step": 8386 + }, + { + "epoch": 0.36, + "grad_norm": 0.45703125, + "learning_rate": 0.00018494330840416163, + "loss": 2.6274, + "step": 8387 + }, + { + "epoch": 0.36, + "grad_norm": 0.4140625, + "learning_rate": 0.00018493976846208513, + "loss": 2.6273, + "step": 8388 + }, + { + "epoch": 0.36, + "grad_norm": 0.46875, + "learning_rate": 0.00018493622813781072, + "loss": 2.6539, + "step": 8389 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.00018493268743135434, + "loss": 2.7293, + "step": 8390 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.0001849291463427319, + "loss": 2.6934, + "step": 8391 + }, + { + "epoch": 0.36, + "grad_norm": 0.404296875, + "learning_rate": 0.00018492560487195933, + "loss": 2.5995, + "step": 8392 + }, + { + "epoch": 0.36, + "grad_norm": 0.4921875, + "learning_rate": 0.00018492206301905258, + "loss": 2.5573, + "step": 8393 + }, + { + "epoch": 0.36, + "grad_norm": 0.451171875, + "learning_rate": 0.00018491852078402756, + "loss": 2.6613, + "step": 8394 + }, + { + "epoch": 0.36, + "grad_norm": 0.4765625, + "learning_rate": 0.00018491497816690026, + "loss": 2.6512, + "step": 8395 + }, + { + "epoch": 0.36, + "grad_norm": 0.4375, + "learning_rate": 0.00018491143516768656, + "loss": 2.6678, + "step": 8396 + }, + { + "epoch": 0.36, + "grad_norm": 0.423828125, + "learning_rate": 0.00018490789178640248, + "loss": 2.6306, + "step": 8397 + }, + { + "epoch": 0.36, + "grad_norm": 0.439453125, + "learning_rate": 0.00018490434802306388, + "loss": 2.651, + "step": 8398 + }, + { + "epoch": 0.36, + "grad_norm": 0.400390625, + "learning_rate": 0.00018490080387768677, + "loss": 2.5481, + "step": 8399 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.00018489725935028704, + "loss": 2.7109, + "step": 8400 + }, + { + "epoch": 0.36, + "grad_norm": 0.4609375, + "learning_rate": 0.0001848937144408807, + "loss": 2.6697, + "step": 8401 + }, + { + "epoch": 0.36, + "grad_norm": 0.443359375, + "learning_rate": 0.00018489016914948362, + "loss": 2.6357, + "step": 8402 + }, + { + "epoch": 0.36, + "grad_norm": 0.45703125, + "learning_rate": 0.00018488662347611188, + "loss": 2.6635, + "step": 8403 + }, + { + "epoch": 0.36, + "grad_norm": 0.392578125, + "learning_rate": 0.0001848830774207813, + "loss": 2.6075, + "step": 8404 + }, + { + "epoch": 0.36, + "grad_norm": 0.44140625, + "learning_rate": 0.0001848795309835079, + "loss": 2.5759, + "step": 8405 + }, + { + "epoch": 0.36, + "grad_norm": 0.486328125, + "learning_rate": 0.00018487598416430765, + "loss": 2.7486, + "step": 8406 + }, + { + "epoch": 0.36, + "grad_norm": 0.48828125, + "learning_rate": 0.00018487243696319646, + "loss": 2.6028, + "step": 8407 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.00018486888938019032, + "loss": 2.664, + "step": 8408 + }, + { + "epoch": 0.36, + "grad_norm": 0.458984375, + "learning_rate": 0.00018486534141530522, + "loss": 2.6075, + "step": 8409 + }, + { + "epoch": 0.36, + "grad_norm": 0.458984375, + "learning_rate": 0.0001848617930685571, + "loss": 2.5434, + "step": 8410 + }, + { + "epoch": 0.36, + "grad_norm": 0.486328125, + "learning_rate": 0.00018485824433996193, + "loss": 2.6766, + "step": 8411 + }, + { + "epoch": 0.36, + "grad_norm": 0.484375, + "learning_rate": 0.00018485469522953565, + "loss": 2.7099, + "step": 8412 + }, + { + "epoch": 0.36, + "grad_norm": 0.4375, + "learning_rate": 0.00018485114573729426, + "loss": 2.6068, + "step": 8413 + }, + { + "epoch": 0.36, + "grad_norm": 0.5234375, + "learning_rate": 0.0001848475958632537, + "loss": 2.7232, + "step": 8414 + }, + { + "epoch": 0.36, + "grad_norm": 0.43359375, + "learning_rate": 0.00018484404560743, + "loss": 2.6766, + "step": 8415 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.0001848404949698391, + "loss": 2.5299, + "step": 8416 + }, + { + "epoch": 0.36, + "grad_norm": 0.4921875, + "learning_rate": 0.00018483694395049697, + "loss": 2.5722, + "step": 8417 + }, + { + "epoch": 0.36, + "grad_norm": 0.41796875, + "learning_rate": 0.00018483339254941962, + "loss": 2.6278, + "step": 8418 + }, + { + "epoch": 0.36, + "grad_norm": 0.46875, + "learning_rate": 0.00018482984076662295, + "loss": 2.6076, + "step": 8419 + }, + { + "epoch": 0.36, + "grad_norm": 0.48046875, + "learning_rate": 0.00018482628860212304, + "loss": 2.7115, + "step": 8420 + }, + { + "epoch": 0.36, + "grad_norm": 0.435546875, + "learning_rate": 0.00018482273605593583, + "loss": 2.5929, + "step": 8421 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.0001848191831280773, + "loss": 2.6574, + "step": 8422 + }, + { + "epoch": 0.36, + "grad_norm": 0.447265625, + "learning_rate": 0.00018481562981856344, + "loss": 2.6279, + "step": 8423 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.00018481207612741025, + "loss": 2.6121, + "step": 8424 + }, + { + "epoch": 0.36, + "grad_norm": 0.5, + "learning_rate": 0.0001848085220546337, + "loss": 2.6947, + "step": 8425 + }, + { + "epoch": 0.36, + "grad_norm": 0.4140625, + "learning_rate": 0.00018480496760024982, + "loss": 2.6255, + "step": 8426 + }, + { + "epoch": 0.36, + "grad_norm": 0.45703125, + "learning_rate": 0.00018480141276427454, + "loss": 2.561, + "step": 8427 + }, + { + "epoch": 0.36, + "grad_norm": 0.4296875, + "learning_rate": 0.00018479785754672392, + "loss": 2.6641, + "step": 8428 + }, + { + "epoch": 0.36, + "grad_norm": 0.5078125, + "learning_rate": 0.00018479430194761395, + "loss": 2.4961, + "step": 8429 + }, + { + "epoch": 0.36, + "grad_norm": 0.435546875, + "learning_rate": 0.00018479074596696056, + "loss": 2.5565, + "step": 8430 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.00018478718960477983, + "loss": 2.537, + "step": 8431 + }, + { + "epoch": 0.36, + "grad_norm": 0.494140625, + "learning_rate": 0.00018478363286108772, + "loss": 2.7331, + "step": 8432 + }, + { + "epoch": 0.36, + "grad_norm": 0.435546875, + "learning_rate": 0.00018478007573590027, + "loss": 2.6477, + "step": 8433 + }, + { + "epoch": 0.36, + "grad_norm": 0.439453125, + "learning_rate": 0.00018477651822923345, + "loss": 2.5851, + "step": 8434 + }, + { + "epoch": 0.36, + "grad_norm": 0.50390625, + "learning_rate": 0.00018477296034110328, + "loss": 2.6978, + "step": 8435 + }, + { + "epoch": 0.36, + "grad_norm": 0.474609375, + "learning_rate": 0.00018476940207152578, + "loss": 2.7244, + "step": 8436 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.00018476584342051695, + "loss": 2.5812, + "step": 8437 + }, + { + "epoch": 0.36, + "grad_norm": 0.498046875, + "learning_rate": 0.00018476228438809276, + "loss": 2.609, + "step": 8438 + }, + { + "epoch": 0.36, + "grad_norm": 0.44140625, + "learning_rate": 0.00018475872497426932, + "loss": 2.6451, + "step": 8439 + }, + { + "epoch": 0.36, + "grad_norm": 0.46875, + "learning_rate": 0.00018475516517906258, + "loss": 2.696, + "step": 8440 + }, + { + "epoch": 0.36, + "grad_norm": 0.47265625, + "learning_rate": 0.00018475160500248857, + "loss": 2.6917, + "step": 8441 + }, + { + "epoch": 0.36, + "grad_norm": 0.453125, + "learning_rate": 0.0001847480444445633, + "loss": 2.5573, + "step": 8442 + }, + { + "epoch": 0.36, + "grad_norm": 0.462890625, + "learning_rate": 0.0001847444835053028, + "loss": 2.7536, + "step": 8443 + }, + { + "epoch": 0.36, + "grad_norm": 0.49609375, + "learning_rate": 0.0001847409221847231, + "loss": 2.6377, + "step": 8444 + }, + { + "epoch": 0.36, + "grad_norm": 0.427734375, + "learning_rate": 0.00018473736048284022, + "loss": 2.5638, + "step": 8445 + }, + { + "epoch": 0.36, + "grad_norm": 0.447265625, + "learning_rate": 0.0001847337983996702, + "loss": 2.6432, + "step": 8446 + }, + { + "epoch": 0.36, + "grad_norm": 0.453125, + "learning_rate": 0.00018473023593522904, + "loss": 2.6885, + "step": 8447 + }, + { + "epoch": 0.36, + "grad_norm": 0.439453125, + "learning_rate": 0.00018472667308953278, + "loss": 2.7141, + "step": 8448 + }, + { + "epoch": 0.36, + "grad_norm": 0.4921875, + "learning_rate": 0.00018472310986259748, + "loss": 2.6168, + "step": 8449 + }, + { + "epoch": 0.36, + "grad_norm": 0.443359375, + "learning_rate": 0.00018471954625443913, + "loss": 2.5822, + "step": 8450 + }, + { + "epoch": 0.36, + "grad_norm": 0.419921875, + "learning_rate": 0.00018471598226507379, + "loss": 2.5974, + "step": 8451 + }, + { + "epoch": 0.36, + "grad_norm": 0.49609375, + "learning_rate": 0.00018471241789451748, + "loss": 2.7041, + "step": 8452 + }, + { + "epoch": 0.36, + "grad_norm": 0.458984375, + "learning_rate": 0.00018470885314278624, + "loss": 2.5796, + "step": 8453 + }, + { + "epoch": 0.36, + "grad_norm": 0.43359375, + "learning_rate": 0.00018470528800989614, + "loss": 2.4887, + "step": 8454 + }, + { + "epoch": 0.36, + "grad_norm": 0.431640625, + "learning_rate": 0.00018470172249586316, + "loss": 2.5511, + "step": 8455 + }, + { + "epoch": 0.36, + "grad_norm": 0.431640625, + "learning_rate": 0.00018469815660070342, + "loss": 2.6366, + "step": 8456 + }, + { + "epoch": 0.36, + "grad_norm": 0.388671875, + "learning_rate": 0.00018469459032443295, + "loss": 2.5919, + "step": 8457 + }, + { + "epoch": 0.36, + "grad_norm": 0.396484375, + "learning_rate": 0.00018469102366706772, + "loss": 2.5069, + "step": 8458 + }, + { + "epoch": 0.36, + "grad_norm": 0.419921875, + "learning_rate": 0.00018468745662862386, + "loss": 2.8184, + "step": 8459 + }, + { + "epoch": 0.36, + "grad_norm": 0.421875, + "learning_rate": 0.0001846838892091174, + "loss": 2.4903, + "step": 8460 + }, + { + "epoch": 0.36, + "grad_norm": 0.45703125, + "learning_rate": 0.0001846803214085644, + "loss": 2.7088, + "step": 8461 + }, + { + "epoch": 0.36, + "grad_norm": 0.5, + "learning_rate": 0.0001846767532269809, + "loss": 2.6663, + "step": 8462 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.0001846731846643829, + "loss": 2.7268, + "step": 8463 + }, + { + "epoch": 0.36, + "grad_norm": 0.439453125, + "learning_rate": 0.00018466961572078655, + "loss": 2.5858, + "step": 8464 + }, + { + "epoch": 0.36, + "grad_norm": 0.453125, + "learning_rate": 0.0001846660463962079, + "loss": 2.5982, + "step": 8465 + }, + { + "epoch": 0.36, + "grad_norm": 0.44140625, + "learning_rate": 0.00018466247669066298, + "loss": 2.6161, + "step": 8466 + }, + { + "epoch": 0.36, + "grad_norm": 0.427734375, + "learning_rate": 0.00018465890660416783, + "loss": 2.5245, + "step": 8467 + }, + { + "epoch": 0.36, + "grad_norm": 0.45703125, + "learning_rate": 0.00018465533613673852, + "loss": 2.7492, + "step": 8468 + }, + { + "epoch": 0.36, + "grad_norm": 0.478515625, + "learning_rate": 0.00018465176528839117, + "loss": 2.7043, + "step": 8469 + }, + { + "epoch": 0.36, + "grad_norm": 0.41796875, + "learning_rate": 0.0001846481940591418, + "loss": 2.6186, + "step": 8470 + }, + { + "epoch": 0.36, + "grad_norm": 0.4765625, + "learning_rate": 0.0001846446224490065, + "loss": 2.662, + "step": 8471 + }, + { + "epoch": 0.36, + "grad_norm": 0.474609375, + "learning_rate": 0.00018464105045800134, + "loss": 2.7113, + "step": 8472 + }, + { + "epoch": 0.36, + "grad_norm": 0.484375, + "learning_rate": 0.00018463747808614234, + "loss": 2.6551, + "step": 8473 + }, + { + "epoch": 0.36, + "grad_norm": 0.4296875, + "learning_rate": 0.00018463390533344568, + "loss": 2.7104, + "step": 8474 + }, + { + "epoch": 0.36, + "grad_norm": 0.41796875, + "learning_rate": 0.00018463033219992734, + "loss": 2.6544, + "step": 8475 + }, + { + "epoch": 0.36, + "grad_norm": 0.5390625, + "learning_rate": 0.00018462675868560345, + "loss": 2.5995, + "step": 8476 + }, + { + "epoch": 0.36, + "grad_norm": 0.486328125, + "learning_rate": 0.0001846231847904901, + "loss": 2.7777, + "step": 8477 + }, + { + "epoch": 0.36, + "grad_norm": 0.47265625, + "learning_rate": 0.00018461961051460327, + "loss": 2.6143, + "step": 8478 + }, + { + "epoch": 0.36, + "grad_norm": 0.431640625, + "learning_rate": 0.00018461603585795918, + "loss": 2.6352, + "step": 8479 + }, + { + "epoch": 0.36, + "grad_norm": 0.4765625, + "learning_rate": 0.00018461246082057385, + "loss": 2.7596, + "step": 8480 + }, + { + "epoch": 0.36, + "grad_norm": 0.48046875, + "learning_rate": 0.00018460888540246332, + "loss": 2.5582, + "step": 8481 + }, + { + "epoch": 0.36, + "grad_norm": 0.4375, + "learning_rate": 0.00018460530960364376, + "loss": 2.604, + "step": 8482 + }, + { + "epoch": 0.36, + "grad_norm": 0.423828125, + "learning_rate": 0.00018460173342413122, + "loss": 2.6238, + "step": 8483 + }, + { + "epoch": 0.36, + "grad_norm": 0.43359375, + "learning_rate": 0.00018459815686394182, + "loss": 2.4735, + "step": 8484 + }, + { + "epoch": 0.36, + "grad_norm": 0.447265625, + "learning_rate": 0.00018459457992309162, + "loss": 2.5008, + "step": 8485 + }, + { + "epoch": 0.36, + "grad_norm": 0.439453125, + "learning_rate": 0.0001845910026015967, + "loss": 2.5654, + "step": 8486 + }, + { + "epoch": 0.36, + "grad_norm": 0.400390625, + "learning_rate": 0.00018458742489947322, + "loss": 2.6379, + "step": 8487 + }, + { + "epoch": 0.36, + "grad_norm": 0.408203125, + "learning_rate": 0.00018458384681673722, + "loss": 2.6221, + "step": 8488 + }, + { + "epoch": 0.36, + "grad_norm": 0.380859375, + "learning_rate": 0.00018458026835340482, + "loss": 2.4859, + "step": 8489 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.0001845766895094921, + "loss": 2.4806, + "step": 8490 + }, + { + "epoch": 0.36, + "grad_norm": 0.478515625, + "learning_rate": 0.00018457311028501522, + "loss": 2.7016, + "step": 8491 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.00018456953067999025, + "loss": 2.5892, + "step": 8492 + }, + { + "epoch": 0.36, + "grad_norm": 0.41796875, + "learning_rate": 0.0001845659506944333, + "loss": 2.6074, + "step": 8493 + }, + { + "epoch": 0.36, + "grad_norm": 0.427734375, + "learning_rate": 0.0001845623703283605, + "loss": 2.6002, + "step": 8494 + }, + { + "epoch": 0.36, + "grad_norm": 0.443359375, + "learning_rate": 0.0001845587895817879, + "loss": 2.5917, + "step": 8495 + }, + { + "epoch": 0.36, + "grad_norm": 0.39453125, + "learning_rate": 0.00018455520845473165, + "loss": 2.5835, + "step": 8496 + }, + { + "epoch": 0.36, + "grad_norm": 0.466796875, + "learning_rate": 0.00018455162694720788, + "loss": 2.6447, + "step": 8497 + }, + { + "epoch": 0.36, + "grad_norm": 0.453125, + "learning_rate": 0.0001845480450592327, + "loss": 2.64, + "step": 8498 + }, + { + "epoch": 0.36, + "grad_norm": 0.4375, + "learning_rate": 0.00018454446279082216, + "loss": 2.6936, + "step": 8499 + }, + { + "epoch": 0.36, + "grad_norm": 0.484375, + "learning_rate": 0.00018454088014199248, + "loss": 2.6123, + "step": 8500 + }, + { + "epoch": 0.36, + "grad_norm": 0.5, + "learning_rate": 0.00018453729711275974, + "loss": 2.7509, + "step": 8501 + }, + { + "epoch": 0.36, + "grad_norm": 0.474609375, + "learning_rate": 0.00018453371370314003, + "loss": 2.5443, + "step": 8502 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.00018453012991314952, + "loss": 2.7154, + "step": 8503 + }, + { + "epoch": 0.36, + "grad_norm": 0.4765625, + "learning_rate": 0.00018452654574280432, + "loss": 2.676, + "step": 8504 + }, + { + "epoch": 0.36, + "grad_norm": 0.453125, + "learning_rate": 0.00018452296119212056, + "loss": 2.6544, + "step": 8505 + }, + { + "epoch": 0.36, + "grad_norm": 0.443359375, + "learning_rate": 0.00018451937626111432, + "loss": 2.5842, + "step": 8506 + }, + { + "epoch": 0.36, + "grad_norm": 0.4609375, + "learning_rate": 0.00018451579094980183, + "loss": 2.575, + "step": 8507 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.00018451220525819918, + "loss": 2.7073, + "step": 8508 + }, + { + "epoch": 0.36, + "grad_norm": 0.462890625, + "learning_rate": 0.00018450861918632242, + "loss": 2.6708, + "step": 8509 + }, + { + "epoch": 0.36, + "grad_norm": 0.419921875, + "learning_rate": 0.00018450503273418782, + "loss": 2.4329, + "step": 8510 + }, + { + "epoch": 0.36, + "grad_norm": 0.49609375, + "learning_rate": 0.00018450144590181143, + "loss": 2.7141, + "step": 8511 + }, + { + "epoch": 0.36, + "grad_norm": 0.458984375, + "learning_rate": 0.0001844978586892094, + "loss": 2.6623, + "step": 8512 + }, + { + "epoch": 0.36, + "grad_norm": 0.5078125, + "learning_rate": 0.0001844942710963979, + "loss": 2.6408, + "step": 8513 + }, + { + "epoch": 0.36, + "grad_norm": 0.43359375, + "learning_rate": 0.00018449068312339305, + "loss": 2.5202, + "step": 8514 + }, + { + "epoch": 0.36, + "grad_norm": 0.435546875, + "learning_rate": 0.00018448709477021104, + "loss": 2.7056, + "step": 8515 + }, + { + "epoch": 0.36, + "grad_norm": 0.421875, + "learning_rate": 0.00018448350603686797, + "loss": 2.5489, + "step": 8516 + }, + { + "epoch": 0.36, + "grad_norm": 0.50390625, + "learning_rate": 0.00018447991692337998, + "loss": 2.5596, + "step": 8517 + }, + { + "epoch": 0.36, + "grad_norm": 0.408203125, + "learning_rate": 0.00018447632742976325, + "loss": 2.5694, + "step": 8518 + }, + { + "epoch": 0.36, + "grad_norm": 0.50390625, + "learning_rate": 0.0001844727375560339, + "loss": 2.6549, + "step": 8519 + }, + { + "epoch": 0.36, + "grad_norm": 0.455078125, + "learning_rate": 0.00018446914730220812, + "loss": 2.64, + "step": 8520 + }, + { + "epoch": 0.36, + "grad_norm": 0.443359375, + "learning_rate": 0.00018446555666830206, + "loss": 2.641, + "step": 8521 + }, + { + "epoch": 0.36, + "grad_norm": 0.486328125, + "learning_rate": 0.00018446196565433183, + "loss": 2.6149, + "step": 8522 + }, + { + "epoch": 0.36, + "grad_norm": 0.4375, + "learning_rate": 0.00018445837426031363, + "loss": 2.4714, + "step": 8523 + }, + { + "epoch": 0.36, + "grad_norm": 0.435546875, + "learning_rate": 0.00018445478248626363, + "loss": 2.6273, + "step": 8524 + }, + { + "epoch": 0.36, + "grad_norm": 0.42578125, + "learning_rate": 0.00018445119033219795, + "loss": 2.6504, + "step": 8525 + }, + { + "epoch": 0.36, + "grad_norm": 0.478515625, + "learning_rate": 0.0001844475977981328, + "loss": 2.6654, + "step": 8526 + }, + { + "epoch": 0.36, + "grad_norm": 0.3984375, + "learning_rate": 0.00018444400488408432, + "loss": 2.5658, + "step": 8527 + }, + { + "epoch": 0.36, + "grad_norm": 0.49609375, + "learning_rate": 0.00018444041159006863, + "loss": 2.582, + "step": 8528 + }, + { + "epoch": 0.36, + "grad_norm": 0.44921875, + "learning_rate": 0.00018443681791610202, + "loss": 2.7277, + "step": 8529 + }, + { + "epoch": 0.36, + "grad_norm": 0.427734375, + "learning_rate": 0.00018443322386220055, + "loss": 2.5076, + "step": 8530 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.00018442962942838043, + "loss": 2.5602, + "step": 8531 + }, + { + "epoch": 0.36, + "grad_norm": 0.48828125, + "learning_rate": 0.00018442603461465783, + "loss": 2.5311, + "step": 8532 + }, + { + "epoch": 0.36, + "grad_norm": 0.45703125, + "learning_rate": 0.00018442243942104892, + "loss": 2.4662, + "step": 8533 + }, + { + "epoch": 0.36, + "grad_norm": 0.380859375, + "learning_rate": 0.00018441884384756994, + "loss": 2.5683, + "step": 8534 + }, + { + "epoch": 0.36, + "grad_norm": 0.5625, + "learning_rate": 0.00018441524789423696, + "loss": 2.6917, + "step": 8535 + }, + { + "epoch": 0.36, + "grad_norm": 0.50390625, + "learning_rate": 0.00018441165156106624, + "loss": 2.6889, + "step": 8536 + }, + { + "epoch": 0.36, + "grad_norm": 0.478515625, + "learning_rate": 0.00018440805484807392, + "loss": 2.6151, + "step": 8537 + }, + { + "epoch": 0.36, + "grad_norm": 0.48046875, + "learning_rate": 0.0001844044577552762, + "loss": 2.6284, + "step": 8538 + }, + { + "epoch": 0.36, + "grad_norm": 0.48828125, + "learning_rate": 0.00018440086028268928, + "loss": 2.6988, + "step": 8539 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.00018439726243032933, + "loss": 2.5238, + "step": 8540 + }, + { + "epoch": 0.36, + "grad_norm": 0.478515625, + "learning_rate": 0.00018439366419821252, + "loss": 2.6975, + "step": 8541 + }, + { + "epoch": 0.36, + "grad_norm": 0.412109375, + "learning_rate": 0.0001843900655863551, + "loss": 2.5926, + "step": 8542 + }, + { + "epoch": 0.36, + "grad_norm": 0.443359375, + "learning_rate": 0.00018438646659477321, + "loss": 2.5823, + "step": 8543 + }, + { + "epoch": 0.36, + "grad_norm": 0.4296875, + "learning_rate": 0.00018438286722348307, + "loss": 2.6261, + "step": 8544 + }, + { + "epoch": 0.36, + "grad_norm": 0.462890625, + "learning_rate": 0.00018437926747250086, + "loss": 2.4471, + "step": 8545 + }, + { + "epoch": 0.36, + "grad_norm": 0.44140625, + "learning_rate": 0.00018437566734184278, + "loss": 2.7314, + "step": 8546 + }, + { + "epoch": 0.36, + "grad_norm": 0.45703125, + "learning_rate": 0.00018437206683152502, + "loss": 2.7125, + "step": 8547 + }, + { + "epoch": 0.36, + "grad_norm": 0.5078125, + "learning_rate": 0.0001843684659415638, + "loss": 2.6112, + "step": 8548 + }, + { + "epoch": 0.36, + "grad_norm": 0.4609375, + "learning_rate": 0.00018436486467197534, + "loss": 2.6144, + "step": 8549 + }, + { + "epoch": 0.36, + "grad_norm": 0.46484375, + "learning_rate": 0.00018436126302277578, + "loss": 2.7111, + "step": 8550 + }, + { + "epoch": 0.36, + "grad_norm": 0.453125, + "learning_rate": 0.0001843576609939814, + "loss": 2.79, + "step": 8551 + }, + { + "epoch": 0.36, + "grad_norm": 0.43359375, + "learning_rate": 0.00018435405858560836, + "loss": 2.588, + "step": 8552 + }, + { + "epoch": 0.36, + "grad_norm": 0.5, + "learning_rate": 0.0001843504557976729, + "loss": 2.6505, + "step": 8553 + }, + { + "epoch": 0.36, + "grad_norm": 0.494140625, + "learning_rate": 0.0001843468526301912, + "loss": 2.5886, + "step": 8554 + }, + { + "epoch": 0.36, + "grad_norm": 0.49609375, + "learning_rate": 0.0001843432490831795, + "loss": 2.659, + "step": 8555 + }, + { + "epoch": 0.36, + "grad_norm": 0.48046875, + "learning_rate": 0.000184339645156654, + "loss": 2.4737, + "step": 8556 + }, + { + "epoch": 0.36, + "grad_norm": 0.408203125, + "learning_rate": 0.0001843360408506309, + "loss": 2.4737, + "step": 8557 + }, + { + "epoch": 0.36, + "grad_norm": 0.4453125, + "learning_rate": 0.00018433243616512645, + "loss": 2.6648, + "step": 8558 + }, + { + "epoch": 0.36, + "grad_norm": 0.51171875, + "learning_rate": 0.0001843288311001569, + "loss": 2.6226, + "step": 8559 + }, + { + "epoch": 0.36, + "grad_norm": 0.5, + "learning_rate": 0.00018432522565573837, + "loss": 2.7898, + "step": 8560 + }, + { + "epoch": 0.36, + "grad_norm": 0.478515625, + "learning_rate": 0.00018432161983188716, + "loss": 2.7273, + "step": 8561 + }, + { + "epoch": 0.36, + "grad_norm": 0.470703125, + "learning_rate": 0.00018431801362861947, + "loss": 2.6363, + "step": 8562 + }, + { + "epoch": 0.36, + "grad_norm": 0.4296875, + "learning_rate": 0.00018431440704595153, + "loss": 2.4427, + "step": 8563 + }, + { + "epoch": 0.36, + "grad_norm": 0.427734375, + "learning_rate": 0.0001843108000838996, + "loss": 2.4678, + "step": 8564 + }, + { + "epoch": 0.36, + "grad_norm": 0.484375, + "learning_rate": 0.00018430719274247986, + "loss": 2.6441, + "step": 8565 + }, + { + "epoch": 0.37, + "grad_norm": 0.474609375, + "learning_rate": 0.00018430358502170857, + "loss": 2.7085, + "step": 8566 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018429997692160196, + "loss": 2.5666, + "step": 8567 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.00018429636844217626, + "loss": 2.6381, + "step": 8568 + }, + { + "epoch": 0.37, + "grad_norm": 0.404296875, + "learning_rate": 0.0001842927595834477, + "loss": 2.6491, + "step": 8569 + }, + { + "epoch": 0.37, + "grad_norm": 0.470703125, + "learning_rate": 0.00018428915034543253, + "loss": 2.6211, + "step": 8570 + }, + { + "epoch": 0.37, + "grad_norm": 0.4921875, + "learning_rate": 0.00018428554072814698, + "loss": 2.6141, + "step": 8571 + }, + { + "epoch": 0.37, + "grad_norm": 0.515625, + "learning_rate": 0.00018428193073160732, + "loss": 2.681, + "step": 8572 + }, + { + "epoch": 0.37, + "grad_norm": 0.490234375, + "learning_rate": 0.00018427832035582974, + "loss": 2.553, + "step": 8573 + }, + { + "epoch": 0.37, + "grad_norm": 0.486328125, + "learning_rate": 0.00018427470960083055, + "loss": 2.7614, + "step": 8574 + }, + { + "epoch": 0.37, + "grad_norm": 0.474609375, + "learning_rate": 0.00018427109846662595, + "loss": 2.6419, + "step": 8575 + }, + { + "epoch": 0.37, + "grad_norm": 0.49609375, + "learning_rate": 0.00018426748695323224, + "loss": 2.7996, + "step": 8576 + }, + { + "epoch": 0.37, + "grad_norm": 0.5234375, + "learning_rate": 0.0001842638750606656, + "loss": 2.7922, + "step": 8577 + }, + { + "epoch": 0.37, + "grad_norm": 0.455078125, + "learning_rate": 0.0001842602627889423, + "loss": 2.5653, + "step": 8578 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018425665013807865, + "loss": 2.6705, + "step": 8579 + }, + { + "epoch": 0.37, + "grad_norm": 0.412109375, + "learning_rate": 0.00018425303710809085, + "loss": 2.6415, + "step": 8580 + }, + { + "epoch": 0.37, + "grad_norm": 0.458984375, + "learning_rate": 0.00018424942369899515, + "loss": 2.606, + "step": 8581 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.0001842458099108079, + "loss": 2.6106, + "step": 8582 + }, + { + "epoch": 0.37, + "grad_norm": 0.4609375, + "learning_rate": 0.00018424219574354525, + "loss": 2.6823, + "step": 8583 + }, + { + "epoch": 0.37, + "grad_norm": 0.45703125, + "learning_rate": 0.0001842385811972235, + "loss": 2.5849, + "step": 8584 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.0001842349662718589, + "loss": 2.6539, + "step": 8585 + }, + { + "epoch": 0.37, + "grad_norm": 0.455078125, + "learning_rate": 0.0001842313509674678, + "loss": 2.8066, + "step": 8586 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.00018422773528406635, + "loss": 2.4853, + "step": 8587 + }, + { + "epoch": 0.37, + "grad_norm": 0.48046875, + "learning_rate": 0.0001842241192216709, + "loss": 2.6032, + "step": 8588 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018422050278029767, + "loss": 2.6156, + "step": 8589 + }, + { + "epoch": 0.37, + "grad_norm": 0.447265625, + "learning_rate": 0.00018421688595996299, + "loss": 2.6887, + "step": 8590 + }, + { + "epoch": 0.37, + "grad_norm": 0.48046875, + "learning_rate": 0.00018421326876068308, + "loss": 2.668, + "step": 8591 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.00018420965118247425, + "loss": 2.5609, + "step": 8592 + }, + { + "epoch": 0.37, + "grad_norm": 0.484375, + "learning_rate": 0.00018420603322535274, + "loss": 2.6338, + "step": 8593 + }, + { + "epoch": 0.37, + "grad_norm": 0.4296875, + "learning_rate": 0.00018420241488933487, + "loss": 2.4186, + "step": 8594 + }, + { + "epoch": 0.37, + "grad_norm": 0.458984375, + "learning_rate": 0.00018419879617443693, + "loss": 2.7936, + "step": 8595 + }, + { + "epoch": 0.37, + "grad_norm": 0.4921875, + "learning_rate": 0.00018419517708067514, + "loss": 2.7739, + "step": 8596 + }, + { + "epoch": 0.37, + "grad_norm": 0.498046875, + "learning_rate": 0.0001841915576080658, + "loss": 2.5963, + "step": 8597 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.00018418793775662527, + "loss": 2.7344, + "step": 8598 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.00018418431752636978, + "loss": 2.6479, + "step": 8599 + }, + { + "epoch": 0.37, + "grad_norm": 0.412109375, + "learning_rate": 0.00018418069691731562, + "loss": 2.6027, + "step": 8600 + }, + { + "epoch": 0.37, + "grad_norm": 0.4375, + "learning_rate": 0.00018417707592947912, + "loss": 2.6694, + "step": 8601 + }, + { + "epoch": 0.37, + "grad_norm": 0.44921875, + "learning_rate": 0.00018417345456287648, + "loss": 2.7364, + "step": 8602 + }, + { + "epoch": 0.37, + "grad_norm": 0.3828125, + "learning_rate": 0.0001841698328175241, + "loss": 2.5692, + "step": 8603 + }, + { + "epoch": 0.37, + "grad_norm": 0.439453125, + "learning_rate": 0.0001841662106934382, + "loss": 2.5295, + "step": 8604 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.00018416258819063512, + "loss": 2.5849, + "step": 8605 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.00018415896530913118, + "loss": 2.6399, + "step": 8606 + }, + { + "epoch": 0.37, + "grad_norm": 0.4375, + "learning_rate": 0.00018415534204894264, + "loss": 2.6667, + "step": 8607 + }, + { + "epoch": 0.37, + "grad_norm": 0.51953125, + "learning_rate": 0.00018415171841008578, + "loss": 2.7698, + "step": 8608 + }, + { + "epoch": 0.37, + "grad_norm": 0.396484375, + "learning_rate": 0.00018414809439257698, + "loss": 2.6872, + "step": 8609 + }, + { + "epoch": 0.37, + "grad_norm": 0.4921875, + "learning_rate": 0.0001841444699964325, + "loss": 2.6501, + "step": 8610 + }, + { + "epoch": 0.37, + "grad_norm": 0.48828125, + "learning_rate": 0.00018414084522166865, + "loss": 2.6734, + "step": 8611 + }, + { + "epoch": 0.37, + "grad_norm": 0.47265625, + "learning_rate": 0.00018413722006830176, + "loss": 2.6786, + "step": 8612 + }, + { + "epoch": 0.37, + "grad_norm": 0.48828125, + "learning_rate": 0.00018413359453634814, + "loss": 2.6456, + "step": 8613 + }, + { + "epoch": 0.37, + "grad_norm": 0.458984375, + "learning_rate": 0.00018412996862582402, + "loss": 2.6569, + "step": 8614 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.00018412634233674587, + "loss": 2.5444, + "step": 8615 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.0001841227156691299, + "loss": 2.5838, + "step": 8616 + }, + { + "epoch": 0.37, + "grad_norm": 0.47265625, + "learning_rate": 0.00018411908862299244, + "loss": 2.6842, + "step": 8617 + }, + { + "epoch": 0.37, + "grad_norm": 0.4453125, + "learning_rate": 0.00018411546119834982, + "loss": 2.5126, + "step": 8618 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.00018411183339521838, + "loss": 2.6441, + "step": 8619 + }, + { + "epoch": 0.37, + "grad_norm": 0.41015625, + "learning_rate": 0.00018410820521361443, + "loss": 2.7198, + "step": 8620 + }, + { + "epoch": 0.37, + "grad_norm": 0.4921875, + "learning_rate": 0.00018410457665355426, + "loss": 2.7044, + "step": 8621 + }, + { + "epoch": 0.37, + "grad_norm": 0.42578125, + "learning_rate": 0.00018410094771505428, + "loss": 2.4935, + "step": 8622 + }, + { + "epoch": 0.37, + "grad_norm": 0.4375, + "learning_rate": 0.00018409731839813076, + "loss": 2.7117, + "step": 8623 + }, + { + "epoch": 0.37, + "grad_norm": 0.458984375, + "learning_rate": 0.0001840936887028, + "loss": 2.5571, + "step": 8624 + }, + { + "epoch": 0.37, + "grad_norm": 0.47265625, + "learning_rate": 0.00018409005862907843, + "loss": 2.6181, + "step": 8625 + }, + { + "epoch": 0.37, + "grad_norm": 0.474609375, + "learning_rate": 0.00018408642817698229, + "loss": 2.6603, + "step": 8626 + }, + { + "epoch": 0.37, + "grad_norm": 0.51171875, + "learning_rate": 0.00018408279734652796, + "loss": 2.5353, + "step": 8627 + }, + { + "epoch": 0.37, + "grad_norm": 0.455078125, + "learning_rate": 0.0001840791661377318, + "loss": 2.6095, + "step": 8628 + }, + { + "epoch": 0.37, + "grad_norm": 0.423828125, + "learning_rate": 0.0001840755345506101, + "loss": 2.5294, + "step": 8629 + }, + { + "epoch": 0.37, + "grad_norm": 0.455078125, + "learning_rate": 0.00018407190258517918, + "loss": 2.5433, + "step": 8630 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018406827024145547, + "loss": 2.5741, + "step": 8631 + }, + { + "epoch": 0.37, + "grad_norm": 0.37109375, + "learning_rate": 0.00018406463751945528, + "loss": 2.6404, + "step": 8632 + }, + { + "epoch": 0.37, + "grad_norm": 0.408203125, + "learning_rate": 0.00018406100441919492, + "loss": 2.5543, + "step": 8633 + }, + { + "epoch": 0.37, + "grad_norm": 0.427734375, + "learning_rate": 0.00018405737094069076, + "loss": 2.68, + "step": 8634 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.00018405373708395916, + "loss": 2.6736, + "step": 8635 + }, + { + "epoch": 0.37, + "grad_norm": 0.42578125, + "learning_rate": 0.00018405010284901641, + "loss": 2.5269, + "step": 8636 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.000184046468235879, + "loss": 2.604, + "step": 8637 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018404283324456312, + "loss": 2.6692, + "step": 8638 + }, + { + "epoch": 0.37, + "grad_norm": 0.4609375, + "learning_rate": 0.00018403919787508526, + "loss": 2.5518, + "step": 8639 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.0001840355621274617, + "loss": 2.4953, + "step": 8640 + }, + { + "epoch": 0.37, + "grad_norm": 0.53125, + "learning_rate": 0.00018403192600170882, + "loss": 2.7517, + "step": 8641 + }, + { + "epoch": 0.37, + "grad_norm": 0.4296875, + "learning_rate": 0.000184028289497843, + "loss": 2.5967, + "step": 8642 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.00018402465261588055, + "loss": 2.6006, + "step": 8643 + }, + { + "epoch": 0.37, + "grad_norm": 0.4609375, + "learning_rate": 0.00018402101535583792, + "loss": 2.5056, + "step": 8644 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018401737771773138, + "loss": 2.7006, + "step": 8645 + }, + { + "epoch": 0.37, + "grad_norm": 0.5, + "learning_rate": 0.00018401373970157735, + "loss": 2.6208, + "step": 8646 + }, + { + "epoch": 0.37, + "grad_norm": 0.5078125, + "learning_rate": 0.0001840101013073922, + "loss": 2.6014, + "step": 8647 + }, + { + "epoch": 0.37, + "grad_norm": 0.482421875, + "learning_rate": 0.00018400646253519237, + "loss": 2.6274, + "step": 8648 + }, + { + "epoch": 0.37, + "grad_norm": 0.470703125, + "learning_rate": 0.00018400282338499406, + "loss": 2.6286, + "step": 8649 + }, + { + "epoch": 0.37, + "grad_norm": 0.43359375, + "learning_rate": 0.0001839991838568138, + "loss": 2.5519, + "step": 8650 + }, + { + "epoch": 0.37, + "grad_norm": 0.494140625, + "learning_rate": 0.0001839955439506679, + "loss": 2.6923, + "step": 8651 + }, + { + "epoch": 0.37, + "grad_norm": 0.51953125, + "learning_rate": 0.00018399190366657272, + "loss": 2.6878, + "step": 8652 + }, + { + "epoch": 0.37, + "grad_norm": 0.48828125, + "learning_rate": 0.00018398826300454468, + "loss": 2.5218, + "step": 8653 + }, + { + "epoch": 0.37, + "grad_norm": 0.3984375, + "learning_rate": 0.00018398462196460018, + "loss": 2.6301, + "step": 8654 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.00018398098054675556, + "loss": 2.4975, + "step": 8655 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.00018397733875102722, + "loss": 2.6589, + "step": 8656 + }, + { + "epoch": 0.37, + "grad_norm": 0.49609375, + "learning_rate": 0.00018397369657743151, + "loss": 2.6777, + "step": 8657 + }, + { + "epoch": 0.37, + "grad_norm": 0.419921875, + "learning_rate": 0.0001839700540259849, + "loss": 2.5089, + "step": 8658 + }, + { + "epoch": 0.37, + "grad_norm": 0.4375, + "learning_rate": 0.00018396641109670372, + "loss": 2.5906, + "step": 8659 + }, + { + "epoch": 0.37, + "grad_norm": 0.416015625, + "learning_rate": 0.00018396276778960438, + "loss": 2.5862, + "step": 8660 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.00018395912410470327, + "loss": 2.7421, + "step": 8661 + }, + { + "epoch": 0.37, + "grad_norm": 0.47265625, + "learning_rate": 0.00018395548004201677, + "loss": 2.6237, + "step": 8662 + }, + { + "epoch": 0.37, + "grad_norm": 0.455078125, + "learning_rate": 0.0001839518356015613, + "loss": 2.5528, + "step": 8663 + }, + { + "epoch": 0.37, + "grad_norm": 0.40234375, + "learning_rate": 0.00018394819078335325, + "loss": 2.6175, + "step": 8664 + }, + { + "epoch": 0.37, + "grad_norm": 0.470703125, + "learning_rate": 0.000183944545587409, + "loss": 2.5733, + "step": 8665 + }, + { + "epoch": 0.37, + "grad_norm": 0.455078125, + "learning_rate": 0.000183940900013745, + "loss": 2.5041, + "step": 8666 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.00018393725406237765, + "loss": 2.6634, + "step": 8667 + }, + { + "epoch": 0.37, + "grad_norm": 0.45703125, + "learning_rate": 0.0001839336077333233, + "loss": 2.6943, + "step": 8668 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.0001839299610265984, + "loss": 2.6709, + "step": 8669 + }, + { + "epoch": 0.37, + "grad_norm": 0.408203125, + "learning_rate": 0.00018392631394221934, + "loss": 2.6244, + "step": 8670 + }, + { + "epoch": 0.37, + "grad_norm": 0.494140625, + "learning_rate": 0.00018392266648020256, + "loss": 2.8072, + "step": 8671 + }, + { + "epoch": 0.37, + "grad_norm": 0.4296875, + "learning_rate": 0.00018391901864056443, + "loss": 2.6152, + "step": 8672 + }, + { + "epoch": 0.37, + "grad_norm": 0.423828125, + "learning_rate": 0.00018391537042332137, + "loss": 2.5929, + "step": 8673 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.0001839117218284898, + "loss": 2.6902, + "step": 8674 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.0001839080728560862, + "loss": 2.53, + "step": 8675 + }, + { + "epoch": 0.37, + "grad_norm": 0.42578125, + "learning_rate": 0.0001839044235061269, + "loss": 2.6027, + "step": 8676 + }, + { + "epoch": 0.37, + "grad_norm": 0.4296875, + "learning_rate": 0.00018390077377862838, + "loss": 2.6025, + "step": 8677 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.00018389712367360703, + "loss": 2.5738, + "step": 8678 + }, + { + "epoch": 0.37, + "grad_norm": 0.390625, + "learning_rate": 0.0001838934731910793, + "loss": 2.6841, + "step": 8679 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018388982233106158, + "loss": 2.5951, + "step": 8680 + }, + { + "epoch": 0.37, + "grad_norm": 0.451171875, + "learning_rate": 0.00018388617109357035, + "loss": 2.6217, + "step": 8681 + }, + { + "epoch": 0.37, + "grad_norm": 0.427734375, + "learning_rate": 0.00018388251947862197, + "loss": 2.5709, + "step": 8682 + }, + { + "epoch": 0.37, + "grad_norm": 0.423828125, + "learning_rate": 0.00018387886748623292, + "loss": 2.5761, + "step": 8683 + }, + { + "epoch": 0.37, + "grad_norm": 0.42578125, + "learning_rate": 0.00018387521511641962, + "loss": 2.8245, + "step": 8684 + }, + { + "epoch": 0.37, + "grad_norm": 0.46484375, + "learning_rate": 0.0001838715623691985, + "loss": 2.6201, + "step": 8685 + }, + { + "epoch": 0.37, + "grad_norm": 0.5078125, + "learning_rate": 0.000183867909244586, + "loss": 2.7428, + "step": 8686 + }, + { + "epoch": 0.37, + "grad_norm": 0.455078125, + "learning_rate": 0.00018386425574259857, + "loss": 2.7092, + "step": 8687 + }, + { + "epoch": 0.37, + "grad_norm": 0.451171875, + "learning_rate": 0.00018386060186325261, + "loss": 2.7042, + "step": 8688 + }, + { + "epoch": 0.37, + "grad_norm": 0.412109375, + "learning_rate": 0.00018385694760656462, + "loss": 2.6423, + "step": 8689 + }, + { + "epoch": 0.37, + "grad_norm": 0.5234375, + "learning_rate": 0.000183853292972551, + "loss": 2.5432, + "step": 8690 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.00018384963796122823, + "loss": 2.6131, + "step": 8691 + }, + { + "epoch": 0.37, + "grad_norm": 0.5390625, + "learning_rate": 0.0001838459825726127, + "loss": 2.6121, + "step": 8692 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.0001838423268067209, + "loss": 2.6726, + "step": 8693 + }, + { + "epoch": 0.37, + "grad_norm": 0.447265625, + "learning_rate": 0.00018383867066356925, + "loss": 2.449, + "step": 8694 + }, + { + "epoch": 0.37, + "grad_norm": 0.416015625, + "learning_rate": 0.00018383501414317427, + "loss": 2.694, + "step": 8695 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018383135724555234, + "loss": 2.6105, + "step": 8696 + }, + { + "epoch": 0.37, + "grad_norm": 0.439453125, + "learning_rate": 0.00018382769997071994, + "loss": 2.4968, + "step": 8697 + }, + { + "epoch": 0.37, + "grad_norm": 0.50390625, + "learning_rate": 0.00018382404231869352, + "loss": 2.6645, + "step": 8698 + }, + { + "epoch": 0.37, + "grad_norm": 0.41796875, + "learning_rate": 0.00018382038428948954, + "loss": 2.6613, + "step": 8699 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.00018381672588312447, + "loss": 2.7258, + "step": 8700 + }, + { + "epoch": 0.37, + "grad_norm": 0.3984375, + "learning_rate": 0.00018381306709961477, + "loss": 2.4687, + "step": 8701 + }, + { + "epoch": 0.37, + "grad_norm": 0.400390625, + "learning_rate": 0.0001838094079389769, + "loss": 2.5966, + "step": 8702 + }, + { + "epoch": 0.37, + "grad_norm": 0.423828125, + "learning_rate": 0.0001838057484012273, + "loss": 2.4363, + "step": 8703 + }, + { + "epoch": 0.37, + "grad_norm": 0.42578125, + "learning_rate": 0.0001838020884863825, + "loss": 2.6443, + "step": 8704 + }, + { + "epoch": 0.37, + "grad_norm": 0.50390625, + "learning_rate": 0.00018379842819445888, + "loss": 2.5641, + "step": 8705 + }, + { + "epoch": 0.37, + "grad_norm": 0.44921875, + "learning_rate": 0.000183794767525473, + "loss": 2.6054, + "step": 8706 + }, + { + "epoch": 0.37, + "grad_norm": 0.43359375, + "learning_rate": 0.00018379110647944127, + "loss": 2.6924, + "step": 8707 + }, + { + "epoch": 0.37, + "grad_norm": 0.4296875, + "learning_rate": 0.00018378744505638018, + "loss": 2.658, + "step": 8708 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.00018378378325630622, + "loss": 2.5866, + "step": 8709 + }, + { + "epoch": 0.37, + "grad_norm": 0.474609375, + "learning_rate": 0.00018378012107923586, + "loss": 2.6108, + "step": 8710 + }, + { + "epoch": 0.37, + "grad_norm": 0.4140625, + "learning_rate": 0.00018377645852518554, + "loss": 2.6441, + "step": 8711 + }, + { + "epoch": 0.37, + "grad_norm": 0.3984375, + "learning_rate": 0.0001837727955941718, + "loss": 2.5255, + "step": 8712 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.0001837691322862111, + "loss": 2.6445, + "step": 8713 + }, + { + "epoch": 0.37, + "grad_norm": 0.439453125, + "learning_rate": 0.00018376546860131992, + "loss": 2.6627, + "step": 8714 + }, + { + "epoch": 0.37, + "grad_norm": 0.4140625, + "learning_rate": 0.0001837618045395147, + "loss": 2.6014, + "step": 8715 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.00018375814010081204, + "loss": 2.6455, + "step": 8716 + }, + { + "epoch": 0.37, + "grad_norm": 0.42578125, + "learning_rate": 0.0001837544752852283, + "loss": 2.4375, + "step": 8717 + }, + { + "epoch": 0.37, + "grad_norm": 0.486328125, + "learning_rate": 0.00018375081009278007, + "loss": 2.5732, + "step": 8718 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.00018374714452348378, + "loss": 2.5451, + "step": 8719 + }, + { + "epoch": 0.37, + "grad_norm": 0.419921875, + "learning_rate": 0.00018374347857735596, + "loss": 2.5414, + "step": 8720 + }, + { + "epoch": 0.37, + "grad_norm": 0.455078125, + "learning_rate": 0.00018373981225441307, + "loss": 2.5953, + "step": 8721 + }, + { + "epoch": 0.37, + "grad_norm": 0.39453125, + "learning_rate": 0.00018373614555467166, + "loss": 2.4941, + "step": 8722 + }, + { + "epoch": 0.37, + "grad_norm": 0.4140625, + "learning_rate": 0.00018373247847814817, + "loss": 2.6197, + "step": 8723 + }, + { + "epoch": 0.37, + "grad_norm": 0.4296875, + "learning_rate": 0.00018372881102485911, + "loss": 2.6081, + "step": 8724 + }, + { + "epoch": 0.37, + "grad_norm": 0.404296875, + "learning_rate": 0.00018372514319482103, + "loss": 2.5466, + "step": 8725 + }, + { + "epoch": 0.37, + "grad_norm": 0.396484375, + "learning_rate": 0.00018372147498805044, + "loss": 2.6418, + "step": 8726 + }, + { + "epoch": 0.37, + "grad_norm": 0.45703125, + "learning_rate": 0.00018371780640456375, + "loss": 2.5728, + "step": 8727 + }, + { + "epoch": 0.37, + "grad_norm": 0.50390625, + "learning_rate": 0.00018371413744437754, + "loss": 2.5926, + "step": 8728 + }, + { + "epoch": 0.37, + "grad_norm": 0.451171875, + "learning_rate": 0.0001837104681075083, + "loss": 2.576, + "step": 8729 + }, + { + "epoch": 0.37, + "grad_norm": 0.5234375, + "learning_rate": 0.00018370679839397255, + "loss": 2.6686, + "step": 8730 + }, + { + "epoch": 0.37, + "grad_norm": 0.5078125, + "learning_rate": 0.00018370312830378682, + "loss": 2.4948, + "step": 8731 + }, + { + "epoch": 0.37, + "grad_norm": 0.41796875, + "learning_rate": 0.00018369945783696762, + "loss": 2.6896, + "step": 8732 + }, + { + "epoch": 0.37, + "grad_norm": 0.53515625, + "learning_rate": 0.0001836957869935314, + "loss": 2.7454, + "step": 8733 + }, + { + "epoch": 0.37, + "grad_norm": 0.5078125, + "learning_rate": 0.00018369211577349472, + "loss": 2.6932, + "step": 8734 + }, + { + "epoch": 0.37, + "grad_norm": 0.53125, + "learning_rate": 0.00018368844417687415, + "loss": 2.6073, + "step": 8735 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.00018368477220368612, + "loss": 2.6564, + "step": 8736 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.00018368109985394723, + "loss": 2.6127, + "step": 8737 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.00018367742712767398, + "loss": 2.6313, + "step": 8738 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018367375402488287, + "loss": 2.5969, + "step": 8739 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.00018367008054559042, + "loss": 2.5519, + "step": 8740 + }, + { + "epoch": 0.37, + "grad_norm": 0.427734375, + "learning_rate": 0.00018366640668981323, + "loss": 2.6664, + "step": 8741 + }, + { + "epoch": 0.37, + "grad_norm": 0.4765625, + "learning_rate": 0.0001836627324575678, + "loss": 2.6086, + "step": 8742 + }, + { + "epoch": 0.37, + "grad_norm": 0.435546875, + "learning_rate": 0.0001836590578488706, + "loss": 2.6906, + "step": 8743 + }, + { + "epoch": 0.37, + "grad_norm": 0.4375, + "learning_rate": 0.00018365538286373823, + "loss": 2.558, + "step": 8744 + }, + { + "epoch": 0.37, + "grad_norm": 0.419921875, + "learning_rate": 0.00018365170750218721, + "loss": 2.5236, + "step": 8745 + }, + { + "epoch": 0.37, + "grad_norm": 0.40625, + "learning_rate": 0.00018364803176423408, + "loss": 2.6222, + "step": 8746 + }, + { + "epoch": 0.37, + "grad_norm": 0.484375, + "learning_rate": 0.00018364435564989536, + "loss": 2.6989, + "step": 8747 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.00018364067915918763, + "loss": 2.7029, + "step": 8748 + }, + { + "epoch": 0.37, + "grad_norm": 0.466796875, + "learning_rate": 0.00018363700229212743, + "loss": 2.5849, + "step": 8749 + }, + { + "epoch": 0.37, + "grad_norm": 0.4453125, + "learning_rate": 0.00018363332504873125, + "loss": 2.5503, + "step": 8750 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.00018362964742901569, + "loss": 2.5939, + "step": 8751 + }, + { + "epoch": 0.37, + "grad_norm": 0.44921875, + "learning_rate": 0.0001836259694329973, + "loss": 2.6634, + "step": 8752 + }, + { + "epoch": 0.37, + "grad_norm": 0.4609375, + "learning_rate": 0.00018362229106069254, + "loss": 2.4944, + "step": 8753 + }, + { + "epoch": 0.37, + "grad_norm": 0.44140625, + "learning_rate": 0.0001836186123121181, + "loss": 2.6412, + "step": 8754 + }, + { + "epoch": 0.37, + "grad_norm": 0.423828125, + "learning_rate": 0.00018361493318729043, + "loss": 2.5708, + "step": 8755 + }, + { + "epoch": 0.37, + "grad_norm": 0.5234375, + "learning_rate": 0.00018361125368622613, + "loss": 2.8112, + "step": 8756 + }, + { + "epoch": 0.37, + "grad_norm": 0.408203125, + "learning_rate": 0.00018360757380894173, + "loss": 2.6327, + "step": 8757 + }, + { + "epoch": 0.37, + "grad_norm": 0.447265625, + "learning_rate": 0.00018360389355545382, + "loss": 2.5867, + "step": 8758 + }, + { + "epoch": 0.37, + "grad_norm": 0.494140625, + "learning_rate": 0.00018360021292577891, + "loss": 2.6229, + "step": 8759 + }, + { + "epoch": 0.37, + "grad_norm": 0.42578125, + "learning_rate": 0.00018359653191993365, + "loss": 2.5569, + "step": 8760 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.00018359285053793452, + "loss": 2.6474, + "step": 8761 + }, + { + "epoch": 0.37, + "grad_norm": 0.4765625, + "learning_rate": 0.00018358916877979811, + "loss": 2.6482, + "step": 8762 + }, + { + "epoch": 0.37, + "grad_norm": 0.486328125, + "learning_rate": 0.00018358548664554095, + "loss": 2.5255, + "step": 8763 + }, + { + "epoch": 0.37, + "grad_norm": 0.44140625, + "learning_rate": 0.0001835818041351797, + "loss": 2.5278, + "step": 8764 + }, + { + "epoch": 0.37, + "grad_norm": 0.48046875, + "learning_rate": 0.0001835781212487309, + "loss": 2.6396, + "step": 8765 + }, + { + "epoch": 0.37, + "grad_norm": 0.494140625, + "learning_rate": 0.00018357443798621103, + "loss": 2.6845, + "step": 8766 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.00018357075434763677, + "loss": 2.6985, + "step": 8767 + }, + { + "epoch": 0.37, + "grad_norm": 0.458984375, + "learning_rate": 0.00018356707033302465, + "loss": 2.6409, + "step": 8768 + }, + { + "epoch": 0.37, + "grad_norm": 0.451171875, + "learning_rate": 0.00018356338594239128, + "loss": 2.5488, + "step": 8769 + }, + { + "epoch": 0.37, + "grad_norm": 0.5078125, + "learning_rate": 0.00018355970117575317, + "loss": 2.6482, + "step": 8770 + }, + { + "epoch": 0.37, + "grad_norm": 0.45703125, + "learning_rate": 0.00018355601603312695, + "loss": 2.5304, + "step": 8771 + }, + { + "epoch": 0.37, + "grad_norm": 0.4296875, + "learning_rate": 0.0001835523305145292, + "loss": 2.577, + "step": 8772 + }, + { + "epoch": 0.37, + "grad_norm": 0.51953125, + "learning_rate": 0.00018354864461997652, + "loss": 2.8118, + "step": 8773 + }, + { + "epoch": 0.37, + "grad_norm": 0.439453125, + "learning_rate": 0.00018354495834948544, + "loss": 2.667, + "step": 8774 + }, + { + "epoch": 0.37, + "grad_norm": 0.453125, + "learning_rate": 0.0001835412717030726, + "loss": 2.7671, + "step": 8775 + }, + { + "epoch": 0.37, + "grad_norm": 0.45703125, + "learning_rate": 0.00018353758468075455, + "loss": 2.7424, + "step": 8776 + }, + { + "epoch": 0.37, + "grad_norm": 0.439453125, + "learning_rate": 0.0001835338972825479, + "loss": 2.6775, + "step": 8777 + }, + { + "epoch": 0.37, + "grad_norm": 0.45703125, + "learning_rate": 0.00018353020950846928, + "loss": 2.5785, + "step": 8778 + }, + { + "epoch": 0.37, + "grad_norm": 0.408203125, + "learning_rate": 0.00018352652135853522, + "loss": 2.6283, + "step": 8779 + }, + { + "epoch": 0.37, + "grad_norm": 0.48828125, + "learning_rate": 0.00018352283283276232, + "loss": 2.5838, + "step": 8780 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.0001835191439311672, + "loss": 2.7025, + "step": 8781 + }, + { + "epoch": 0.37, + "grad_norm": 0.431640625, + "learning_rate": 0.00018351545465376647, + "loss": 2.6215, + "step": 8782 + }, + { + "epoch": 0.37, + "grad_norm": 0.5234375, + "learning_rate": 0.0001835117650005767, + "loss": 2.6485, + "step": 8783 + }, + { + "epoch": 0.37, + "grad_norm": 0.419921875, + "learning_rate": 0.00018350807497161453, + "loss": 2.5918, + "step": 8784 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.00018350438456689653, + "loss": 2.7347, + "step": 8785 + }, + { + "epoch": 0.37, + "grad_norm": 0.45703125, + "learning_rate": 0.00018350069378643934, + "loss": 2.7733, + "step": 8786 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.0001834970026302595, + "loss": 2.5724, + "step": 8787 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.0001834933110983737, + "loss": 2.6084, + "step": 8788 + }, + { + "epoch": 0.37, + "grad_norm": 0.43359375, + "learning_rate": 0.00018348961919079847, + "loss": 2.6462, + "step": 8789 + }, + { + "epoch": 0.37, + "grad_norm": 0.443359375, + "learning_rate": 0.0001834859269075505, + "loss": 2.5854, + "step": 8790 + }, + { + "epoch": 0.37, + "grad_norm": 0.486328125, + "learning_rate": 0.00018348223424864636, + "loss": 2.7072, + "step": 8791 + }, + { + "epoch": 0.37, + "grad_norm": 0.419921875, + "learning_rate": 0.0001834785412141027, + "loss": 2.5234, + "step": 8792 + }, + { + "epoch": 0.37, + "grad_norm": 0.47265625, + "learning_rate": 0.00018347484780393606, + "loss": 2.6098, + "step": 8793 + }, + { + "epoch": 0.37, + "grad_norm": 0.4296875, + "learning_rate": 0.00018347115401816315, + "loss": 2.6682, + "step": 8794 + }, + { + "epoch": 0.37, + "grad_norm": 0.50390625, + "learning_rate": 0.00018346745985680054, + "loss": 2.7215, + "step": 8795 + }, + { + "epoch": 0.37, + "grad_norm": 0.462890625, + "learning_rate": 0.00018346376531986486, + "loss": 2.5916, + "step": 8796 + }, + { + "epoch": 0.37, + "grad_norm": 0.447265625, + "learning_rate": 0.00018346007040737273, + "loss": 2.5974, + "step": 8797 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.0001834563751193408, + "loss": 2.6513, + "step": 8798 + }, + { + "epoch": 0.37, + "grad_norm": 0.51171875, + "learning_rate": 0.0001834526794557857, + "loss": 2.7561, + "step": 8799 + }, + { + "epoch": 0.37, + "grad_norm": 0.46875, + "learning_rate": 0.00018344898341672404, + "loss": 2.6721, + "step": 8800 + }, + { + "epoch": 0.38, + "grad_norm": 0.47265625, + "learning_rate": 0.00018344528700217242, + "loss": 2.5866, + "step": 8801 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.00018344159021214753, + "loss": 2.6797, + "step": 8802 + }, + { + "epoch": 0.38, + "grad_norm": 0.515625, + "learning_rate": 0.00018343789304666594, + "loss": 2.6891, + "step": 8803 + }, + { + "epoch": 0.38, + "grad_norm": 0.447265625, + "learning_rate": 0.00018343419550574436, + "loss": 2.6324, + "step": 8804 + }, + { + "epoch": 0.38, + "grad_norm": 0.376953125, + "learning_rate": 0.0001834304975893994, + "loss": 2.6538, + "step": 8805 + }, + { + "epoch": 0.38, + "grad_norm": 0.4609375, + "learning_rate": 0.00018342679929764764, + "loss": 2.7309, + "step": 8806 + }, + { + "epoch": 0.38, + "grad_norm": 0.4453125, + "learning_rate": 0.00018342310063050582, + "loss": 2.6314, + "step": 8807 + }, + { + "epoch": 0.38, + "grad_norm": 0.46875, + "learning_rate": 0.00018341940158799052, + "loss": 2.634, + "step": 8808 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.00018341570217011841, + "loss": 2.7445, + "step": 8809 + }, + { + "epoch": 0.38, + "grad_norm": 0.44921875, + "learning_rate": 0.0001834120023769061, + "loss": 2.7545, + "step": 8810 + }, + { + "epoch": 0.38, + "grad_norm": 0.482421875, + "learning_rate": 0.0001834083022083703, + "loss": 2.5928, + "step": 8811 + }, + { + "epoch": 0.38, + "grad_norm": 0.39453125, + "learning_rate": 0.0001834046016645276, + "loss": 2.4779, + "step": 8812 + }, + { + "epoch": 0.38, + "grad_norm": 0.48046875, + "learning_rate": 0.00018340090074539468, + "loss": 2.7082, + "step": 8813 + }, + { + "epoch": 0.38, + "grad_norm": 0.458984375, + "learning_rate": 0.00018339719945098816, + "loss": 2.5201, + "step": 8814 + }, + { + "epoch": 0.38, + "grad_norm": 0.443359375, + "learning_rate": 0.00018339349778132474, + "loss": 2.7343, + "step": 8815 + }, + { + "epoch": 0.38, + "grad_norm": 0.478515625, + "learning_rate": 0.00018338979573642107, + "loss": 2.7933, + "step": 8816 + }, + { + "epoch": 0.38, + "grad_norm": 0.578125, + "learning_rate": 0.00018338609331629377, + "loss": 2.7186, + "step": 8817 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.00018338239052095956, + "loss": 2.6879, + "step": 8818 + }, + { + "epoch": 0.38, + "grad_norm": 0.47265625, + "learning_rate": 0.00018337868735043502, + "loss": 2.6618, + "step": 8819 + }, + { + "epoch": 0.38, + "grad_norm": 0.50390625, + "learning_rate": 0.0001833749838047369, + "loss": 2.743, + "step": 8820 + }, + { + "epoch": 0.38, + "grad_norm": 0.419921875, + "learning_rate": 0.00018337127988388182, + "loss": 2.5318, + "step": 8821 + }, + { + "epoch": 0.38, + "grad_norm": 0.423828125, + "learning_rate": 0.0001833675755878864, + "loss": 2.5936, + "step": 8822 + }, + { + "epoch": 0.38, + "grad_norm": 0.486328125, + "learning_rate": 0.00018336387091676741, + "loss": 2.5495, + "step": 8823 + }, + { + "epoch": 0.38, + "grad_norm": 0.453125, + "learning_rate": 0.00018336016587054146, + "loss": 2.5129, + "step": 8824 + }, + { + "epoch": 0.38, + "grad_norm": 0.439453125, + "learning_rate": 0.0001833564604492252, + "loss": 2.5391, + "step": 8825 + }, + { + "epoch": 0.38, + "grad_norm": 0.416015625, + "learning_rate": 0.00018335275465283535, + "loss": 2.572, + "step": 8826 + }, + { + "epoch": 0.38, + "grad_norm": 0.447265625, + "learning_rate": 0.00018334904848138855, + "loss": 2.6071, + "step": 8827 + }, + { + "epoch": 0.38, + "grad_norm": 0.482421875, + "learning_rate": 0.00018334534193490152, + "loss": 2.6904, + "step": 8828 + }, + { + "epoch": 0.38, + "grad_norm": 0.427734375, + "learning_rate": 0.0001833416350133909, + "loss": 2.57, + "step": 8829 + }, + { + "epoch": 0.38, + "grad_norm": 0.51171875, + "learning_rate": 0.00018333792771687337, + "loss": 2.7687, + "step": 8830 + }, + { + "epoch": 0.38, + "grad_norm": 0.47265625, + "learning_rate": 0.0001833342200453656, + "loss": 2.5883, + "step": 8831 + }, + { + "epoch": 0.38, + "grad_norm": 0.4375, + "learning_rate": 0.00018333051199888436, + "loss": 2.7476, + "step": 8832 + }, + { + "epoch": 0.38, + "grad_norm": 0.44921875, + "learning_rate": 0.00018332680357744622, + "loss": 2.6945, + "step": 8833 + }, + { + "epoch": 0.38, + "grad_norm": 0.390625, + "learning_rate": 0.00018332309478106792, + "loss": 2.6237, + "step": 8834 + }, + { + "epoch": 0.38, + "grad_norm": 0.43359375, + "learning_rate": 0.00018331938560976616, + "loss": 2.6004, + "step": 8835 + }, + { + "epoch": 0.38, + "grad_norm": 0.4296875, + "learning_rate": 0.00018331567606355762, + "loss": 2.7265, + "step": 8836 + }, + { + "epoch": 0.38, + "grad_norm": 0.455078125, + "learning_rate": 0.00018331196614245894, + "loss": 2.6443, + "step": 8837 + }, + { + "epoch": 0.38, + "grad_norm": 0.478515625, + "learning_rate": 0.00018330825584648693, + "loss": 2.6757, + "step": 8838 + }, + { + "epoch": 0.38, + "grad_norm": 0.431640625, + "learning_rate": 0.00018330454517565816, + "loss": 2.4188, + "step": 8839 + }, + { + "epoch": 0.38, + "grad_norm": 0.466796875, + "learning_rate": 0.0001833008341299894, + "loss": 2.6904, + "step": 8840 + }, + { + "epoch": 0.38, + "grad_norm": 0.4375, + "learning_rate": 0.00018329712270949734, + "loss": 2.594, + "step": 8841 + }, + { + "epoch": 0.38, + "grad_norm": 0.474609375, + "learning_rate": 0.00018329341091419864, + "loss": 2.5804, + "step": 8842 + }, + { + "epoch": 0.38, + "grad_norm": 0.390625, + "learning_rate": 0.00018328969874411004, + "loss": 2.4978, + "step": 8843 + }, + { + "epoch": 0.38, + "grad_norm": 0.482421875, + "learning_rate": 0.00018328598619924826, + "loss": 2.7862, + "step": 8844 + }, + { + "epoch": 0.38, + "grad_norm": 0.41015625, + "learning_rate": 0.00018328227327962993, + "loss": 2.4721, + "step": 8845 + }, + { + "epoch": 0.38, + "grad_norm": 0.46484375, + "learning_rate": 0.00018327855998527185, + "loss": 2.5993, + "step": 8846 + }, + { + "epoch": 0.38, + "grad_norm": 0.416015625, + "learning_rate": 0.0001832748463161907, + "loss": 2.7438, + "step": 8847 + }, + { + "epoch": 0.38, + "grad_norm": 0.40234375, + "learning_rate": 0.00018327113227240314, + "loss": 2.521, + "step": 8848 + }, + { + "epoch": 0.38, + "grad_norm": 0.44140625, + "learning_rate": 0.00018326741785392595, + "loss": 2.5815, + "step": 8849 + }, + { + "epoch": 0.38, + "grad_norm": 0.44140625, + "learning_rate": 0.00018326370306077577, + "loss": 2.562, + "step": 8850 + }, + { + "epoch": 0.38, + "grad_norm": 0.466796875, + "learning_rate": 0.00018325998789296938, + "loss": 2.7299, + "step": 8851 + }, + { + "epoch": 0.38, + "grad_norm": 0.458984375, + "learning_rate": 0.00018325627235052345, + "loss": 2.6042, + "step": 8852 + }, + { + "epoch": 0.38, + "grad_norm": 0.494140625, + "learning_rate": 0.00018325255643345475, + "loss": 2.8231, + "step": 8853 + }, + { + "epoch": 0.38, + "grad_norm": 0.4140625, + "learning_rate": 0.00018324884014177993, + "loss": 2.5862, + "step": 8854 + }, + { + "epoch": 0.38, + "grad_norm": 0.49609375, + "learning_rate": 0.0001832451234755158, + "loss": 2.7085, + "step": 8855 + }, + { + "epoch": 0.38, + "grad_norm": 0.462890625, + "learning_rate": 0.00018324140643467902, + "loss": 2.757, + "step": 8856 + }, + { + "epoch": 0.38, + "grad_norm": 0.50390625, + "learning_rate": 0.00018323768901928634, + "loss": 2.5387, + "step": 8857 + }, + { + "epoch": 0.38, + "grad_norm": 0.431640625, + "learning_rate": 0.00018323397122935444, + "loss": 2.7351, + "step": 8858 + }, + { + "epoch": 0.38, + "grad_norm": 0.439453125, + "learning_rate": 0.0001832302530649001, + "loss": 2.5209, + "step": 8859 + }, + { + "epoch": 0.38, + "grad_norm": 0.46484375, + "learning_rate": 0.00018322653452594007, + "loss": 2.5361, + "step": 8860 + }, + { + "epoch": 0.38, + "grad_norm": 0.421875, + "learning_rate": 0.00018322281561249107, + "loss": 2.7453, + "step": 8861 + }, + { + "epoch": 0.38, + "grad_norm": 0.515625, + "learning_rate": 0.00018321909632456978, + "loss": 2.7234, + "step": 8862 + }, + { + "epoch": 0.38, + "grad_norm": 0.42578125, + "learning_rate": 0.00018321537666219295, + "loss": 2.4972, + "step": 8863 + }, + { + "epoch": 0.38, + "grad_norm": 0.5078125, + "learning_rate": 0.00018321165662537735, + "loss": 2.6541, + "step": 8864 + }, + { + "epoch": 0.38, + "grad_norm": 0.44921875, + "learning_rate": 0.00018320793621413972, + "loss": 2.5983, + "step": 8865 + }, + { + "epoch": 0.38, + "grad_norm": 0.431640625, + "learning_rate": 0.0001832042154284968, + "loss": 2.4687, + "step": 8866 + }, + { + "epoch": 0.38, + "grad_norm": 0.412109375, + "learning_rate": 0.00018320049426846525, + "loss": 2.6086, + "step": 8867 + }, + { + "epoch": 0.38, + "grad_norm": 0.427734375, + "learning_rate": 0.00018319677273406196, + "loss": 2.5916, + "step": 8868 + }, + { + "epoch": 0.38, + "grad_norm": 0.4765625, + "learning_rate": 0.00018319305082530355, + "loss": 2.6698, + "step": 8869 + }, + { + "epoch": 0.38, + "grad_norm": 0.4375, + "learning_rate": 0.00018318932854220684, + "loss": 2.6849, + "step": 8870 + }, + { + "epoch": 0.38, + "grad_norm": 0.447265625, + "learning_rate": 0.00018318560588478854, + "loss": 2.5167, + "step": 8871 + }, + { + "epoch": 0.38, + "grad_norm": 0.41015625, + "learning_rate": 0.00018318188285306543, + "loss": 2.5603, + "step": 8872 + }, + { + "epoch": 0.38, + "grad_norm": 0.5, + "learning_rate": 0.00018317815944705426, + "loss": 2.7057, + "step": 8873 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.00018317443566677173, + "loss": 2.5923, + "step": 8874 + }, + { + "epoch": 0.38, + "grad_norm": 0.44921875, + "learning_rate": 0.00018317071151223465, + "loss": 2.6509, + "step": 8875 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.00018316698698345976, + "loss": 2.7371, + "step": 8876 + }, + { + "epoch": 0.38, + "grad_norm": 0.3984375, + "learning_rate": 0.00018316326208046382, + "loss": 2.614, + "step": 8877 + }, + { + "epoch": 0.38, + "grad_norm": 0.4765625, + "learning_rate": 0.0001831595368032636, + "loss": 2.487, + "step": 8878 + }, + { + "epoch": 0.38, + "grad_norm": 0.498046875, + "learning_rate": 0.00018315581115187584, + "loss": 2.7923, + "step": 8879 + }, + { + "epoch": 0.38, + "grad_norm": 0.46484375, + "learning_rate": 0.00018315208512631736, + "loss": 2.5118, + "step": 8880 + }, + { + "epoch": 0.38, + "grad_norm": 0.474609375, + "learning_rate": 0.00018314835872660487, + "loss": 2.6712, + "step": 8881 + }, + { + "epoch": 0.38, + "grad_norm": 0.44921875, + "learning_rate": 0.00018314463195275515, + "loss": 2.5809, + "step": 8882 + }, + { + "epoch": 0.38, + "grad_norm": 0.43359375, + "learning_rate": 0.00018314090480478495, + "loss": 2.6383, + "step": 8883 + }, + { + "epoch": 0.38, + "grad_norm": 0.419921875, + "learning_rate": 0.00018313717728271113, + "loss": 2.6663, + "step": 8884 + }, + { + "epoch": 0.38, + "grad_norm": 0.423828125, + "learning_rate": 0.00018313344938655033, + "loss": 2.6066, + "step": 8885 + }, + { + "epoch": 0.38, + "grad_norm": 0.484375, + "learning_rate": 0.00018312972111631938, + "loss": 2.6187, + "step": 8886 + }, + { + "epoch": 0.38, + "grad_norm": 0.404296875, + "learning_rate": 0.0001831259924720351, + "loss": 2.5381, + "step": 8887 + }, + { + "epoch": 0.38, + "grad_norm": 0.41796875, + "learning_rate": 0.00018312226345371425, + "loss": 2.6567, + "step": 8888 + }, + { + "epoch": 0.38, + "grad_norm": 0.455078125, + "learning_rate": 0.00018311853406137357, + "loss": 2.6642, + "step": 8889 + }, + { + "epoch": 0.38, + "grad_norm": 0.431640625, + "learning_rate": 0.00018311480429502984, + "loss": 2.6419, + "step": 8890 + }, + { + "epoch": 0.38, + "grad_norm": 0.458984375, + "learning_rate": 0.0001831110741546999, + "loss": 2.7176, + "step": 8891 + }, + { + "epoch": 0.38, + "grad_norm": 0.478515625, + "learning_rate": 0.00018310734364040048, + "loss": 2.6787, + "step": 8892 + }, + { + "epoch": 0.38, + "grad_norm": 0.44140625, + "learning_rate": 0.0001831036127521484, + "loss": 2.5514, + "step": 8893 + }, + { + "epoch": 0.38, + "grad_norm": 0.4453125, + "learning_rate": 0.00018309988148996041, + "loss": 2.6686, + "step": 8894 + }, + { + "epoch": 0.38, + "grad_norm": 0.443359375, + "learning_rate": 0.00018309614985385336, + "loss": 2.7324, + "step": 8895 + }, + { + "epoch": 0.38, + "grad_norm": 0.43359375, + "learning_rate": 0.00018309241784384396, + "loss": 2.5831, + "step": 8896 + }, + { + "epoch": 0.38, + "grad_norm": 0.48828125, + "learning_rate": 0.00018308868545994906, + "loss": 2.5714, + "step": 8897 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.00018308495270218545, + "loss": 2.7782, + "step": 8898 + }, + { + "epoch": 0.38, + "grad_norm": 0.408203125, + "learning_rate": 0.00018308121957056993, + "loss": 2.5742, + "step": 8899 + }, + { + "epoch": 0.38, + "grad_norm": 0.4921875, + "learning_rate": 0.00018307748606511925, + "loss": 2.6227, + "step": 8900 + }, + { + "epoch": 0.38, + "grad_norm": 0.427734375, + "learning_rate": 0.00018307375218585027, + "loss": 2.5645, + "step": 8901 + }, + { + "epoch": 0.38, + "grad_norm": 0.390625, + "learning_rate": 0.0001830700179327798, + "loss": 2.5969, + "step": 8902 + }, + { + "epoch": 0.38, + "grad_norm": 0.453125, + "learning_rate": 0.00018306628330592455, + "loss": 2.5175, + "step": 8903 + }, + { + "epoch": 0.38, + "grad_norm": 0.4296875, + "learning_rate": 0.0001830625483053014, + "loss": 2.5021, + "step": 8904 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.00018305881293092713, + "loss": 2.6525, + "step": 8905 + }, + { + "epoch": 0.38, + "grad_norm": 0.51171875, + "learning_rate": 0.00018305507718281857, + "loss": 2.6157, + "step": 8906 + }, + { + "epoch": 0.38, + "grad_norm": 0.46875, + "learning_rate": 0.0001830513410609925, + "loss": 2.6745, + "step": 8907 + }, + { + "epoch": 0.38, + "grad_norm": 0.443359375, + "learning_rate": 0.00018304760456546573, + "loss": 2.5718, + "step": 8908 + }, + { + "epoch": 0.38, + "grad_norm": 0.474609375, + "learning_rate": 0.00018304386769625515, + "loss": 2.6709, + "step": 8909 + }, + { + "epoch": 0.38, + "grad_norm": 0.453125, + "learning_rate": 0.0001830401304533775, + "loss": 2.6176, + "step": 8910 + }, + { + "epoch": 0.38, + "grad_norm": 0.44921875, + "learning_rate": 0.00018303639283684956, + "loss": 2.6166, + "step": 8911 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.00018303265484668826, + "loss": 2.8205, + "step": 8912 + }, + { + "epoch": 0.38, + "grad_norm": 0.455078125, + "learning_rate": 0.0001830289164829103, + "loss": 2.6978, + "step": 8913 + }, + { + "epoch": 0.38, + "grad_norm": 0.48828125, + "learning_rate": 0.0001830251777455326, + "loss": 2.7134, + "step": 8914 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.00018302143863457192, + "loss": 2.6261, + "step": 8915 + }, + { + "epoch": 0.38, + "grad_norm": 0.45703125, + "learning_rate": 0.00018301769915004515, + "loss": 2.5724, + "step": 8916 + }, + { + "epoch": 0.38, + "grad_norm": 0.4609375, + "learning_rate": 0.00018301395929196904, + "loss": 2.6605, + "step": 8917 + }, + { + "epoch": 0.38, + "grad_norm": 0.484375, + "learning_rate": 0.00018301021906036043, + "loss": 2.6468, + "step": 8918 + }, + { + "epoch": 0.38, + "grad_norm": 0.390625, + "learning_rate": 0.0001830064784552362, + "loss": 2.8567, + "step": 8919 + }, + { + "epoch": 0.38, + "grad_norm": 0.41796875, + "learning_rate": 0.00018300273747661315, + "loss": 2.6516, + "step": 8920 + }, + { + "epoch": 0.38, + "grad_norm": 0.443359375, + "learning_rate": 0.0001829989961245081, + "loss": 2.5628, + "step": 8921 + }, + { + "epoch": 0.38, + "grad_norm": 0.53515625, + "learning_rate": 0.00018299525439893793, + "loss": 2.7446, + "step": 8922 + }, + { + "epoch": 0.38, + "grad_norm": 0.41796875, + "learning_rate": 0.00018299151229991942, + "loss": 2.6293, + "step": 8923 + }, + { + "epoch": 0.38, + "grad_norm": 0.51171875, + "learning_rate": 0.00018298776982746945, + "loss": 2.6332, + "step": 8924 + }, + { + "epoch": 0.38, + "grad_norm": 0.458984375, + "learning_rate": 0.00018298402698160483, + "loss": 2.5499, + "step": 8925 + }, + { + "epoch": 0.38, + "grad_norm": 0.470703125, + "learning_rate": 0.0001829802837623424, + "loss": 2.723, + "step": 8926 + }, + { + "epoch": 0.38, + "grad_norm": 0.48046875, + "learning_rate": 0.00018297654016969904, + "loss": 2.7657, + "step": 8927 + }, + { + "epoch": 0.38, + "grad_norm": 0.5, + "learning_rate": 0.0001829727962036916, + "loss": 2.6459, + "step": 8928 + }, + { + "epoch": 0.38, + "grad_norm": 0.478515625, + "learning_rate": 0.00018296905186433686, + "loss": 2.5757, + "step": 8929 + }, + { + "epoch": 0.38, + "grad_norm": 0.46875, + "learning_rate": 0.00018296530715165174, + "loss": 2.5549, + "step": 8930 + }, + { + "epoch": 0.38, + "grad_norm": 0.423828125, + "learning_rate": 0.00018296156206565304, + "loss": 2.6798, + "step": 8931 + }, + { + "epoch": 0.38, + "grad_norm": 0.421875, + "learning_rate": 0.00018295781660635765, + "loss": 2.5562, + "step": 8932 + }, + { + "epoch": 0.38, + "grad_norm": 0.46875, + "learning_rate": 0.00018295407077378238, + "loss": 2.7661, + "step": 8933 + }, + { + "epoch": 0.38, + "grad_norm": 0.408203125, + "learning_rate": 0.0001829503245679441, + "loss": 2.6645, + "step": 8934 + }, + { + "epoch": 0.38, + "grad_norm": 0.484375, + "learning_rate": 0.00018294657798885967, + "loss": 2.7484, + "step": 8935 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.00018294283103654599, + "loss": 2.6878, + "step": 8936 + }, + { + "epoch": 0.38, + "grad_norm": 0.421875, + "learning_rate": 0.00018293908371101988, + "loss": 2.4912, + "step": 8937 + }, + { + "epoch": 0.38, + "grad_norm": 0.4609375, + "learning_rate": 0.00018293533601229815, + "loss": 2.5637, + "step": 8938 + }, + { + "epoch": 0.38, + "grad_norm": 0.4375, + "learning_rate": 0.00018293158794039778, + "loss": 2.5654, + "step": 8939 + }, + { + "epoch": 0.38, + "grad_norm": 0.490234375, + "learning_rate": 0.00018292783949533554, + "loss": 2.7325, + "step": 8940 + }, + { + "epoch": 0.38, + "grad_norm": 0.421875, + "learning_rate": 0.00018292409067712832, + "loss": 2.6212, + "step": 8941 + }, + { + "epoch": 0.38, + "grad_norm": 0.47265625, + "learning_rate": 0.00018292034148579303, + "loss": 2.605, + "step": 8942 + }, + { + "epoch": 0.38, + "grad_norm": 0.5078125, + "learning_rate": 0.00018291659192134647, + "loss": 2.7508, + "step": 8943 + }, + { + "epoch": 0.38, + "grad_norm": 0.4375, + "learning_rate": 0.0001829128419838056, + "loss": 2.7026, + "step": 8944 + }, + { + "epoch": 0.38, + "grad_norm": 0.45703125, + "learning_rate": 0.00018290909167318717, + "loss": 2.5812, + "step": 8945 + }, + { + "epoch": 0.38, + "grad_norm": 0.52734375, + "learning_rate": 0.0001829053409895082, + "loss": 2.6497, + "step": 8946 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.00018290158993278545, + "loss": 2.551, + "step": 8947 + }, + { + "epoch": 0.38, + "grad_norm": 0.439453125, + "learning_rate": 0.00018289783850303587, + "loss": 2.7231, + "step": 8948 + }, + { + "epoch": 0.38, + "grad_norm": 0.416015625, + "learning_rate": 0.00018289408670027629, + "loss": 2.5926, + "step": 8949 + }, + { + "epoch": 0.38, + "grad_norm": 0.443359375, + "learning_rate": 0.0001828903345245236, + "loss": 2.6472, + "step": 8950 + }, + { + "epoch": 0.38, + "grad_norm": 0.4921875, + "learning_rate": 0.00018288658197579473, + "loss": 2.5513, + "step": 8951 + }, + { + "epoch": 0.38, + "grad_norm": 0.43359375, + "learning_rate": 0.00018288282905410652, + "loss": 2.6388, + "step": 8952 + }, + { + "epoch": 0.38, + "grad_norm": 0.458984375, + "learning_rate": 0.00018287907575947586, + "loss": 2.6732, + "step": 8953 + }, + { + "epoch": 0.38, + "grad_norm": 0.5078125, + "learning_rate": 0.00018287532209191966, + "loss": 2.5535, + "step": 8954 + }, + { + "epoch": 0.38, + "grad_norm": 0.466796875, + "learning_rate": 0.0001828715680514548, + "loss": 2.7844, + "step": 8955 + }, + { + "epoch": 0.38, + "grad_norm": 0.4375, + "learning_rate": 0.00018286781363809816, + "loss": 2.638, + "step": 8956 + }, + { + "epoch": 0.38, + "grad_norm": 0.421875, + "learning_rate": 0.00018286405885186666, + "loss": 2.5366, + "step": 8957 + }, + { + "epoch": 0.38, + "grad_norm": 0.470703125, + "learning_rate": 0.0001828603036927772, + "loss": 2.4998, + "step": 8958 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.0001828565481608466, + "loss": 2.4674, + "step": 8959 + }, + { + "epoch": 0.38, + "grad_norm": 0.46484375, + "learning_rate": 0.00018285279225609182, + "loss": 2.6987, + "step": 8960 + }, + { + "epoch": 0.38, + "grad_norm": 0.4921875, + "learning_rate": 0.0001828490359785298, + "loss": 2.6886, + "step": 8961 + }, + { + "epoch": 0.38, + "grad_norm": 0.45703125, + "learning_rate": 0.00018284527932817734, + "loss": 2.6982, + "step": 8962 + }, + { + "epoch": 0.38, + "grad_norm": 0.478515625, + "learning_rate": 0.00018284152230505142, + "loss": 2.5936, + "step": 8963 + }, + { + "epoch": 0.38, + "grad_norm": 0.494140625, + "learning_rate": 0.00018283776490916895, + "loss": 2.5549, + "step": 8964 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.00018283400714054677, + "loss": 2.5065, + "step": 8965 + }, + { + "epoch": 0.38, + "grad_norm": 0.466796875, + "learning_rate": 0.00018283024899920188, + "loss": 2.5712, + "step": 8966 + }, + { + "epoch": 0.38, + "grad_norm": 0.447265625, + "learning_rate": 0.0001828264904851511, + "loss": 2.6198, + "step": 8967 + }, + { + "epoch": 0.38, + "grad_norm": 0.53515625, + "learning_rate": 0.0001828227315984114, + "loss": 2.4676, + "step": 8968 + }, + { + "epoch": 0.38, + "grad_norm": 0.486328125, + "learning_rate": 0.00018281897233899966, + "loss": 2.6749, + "step": 8969 + }, + { + "epoch": 0.38, + "grad_norm": 0.455078125, + "learning_rate": 0.00018281521270693281, + "loss": 2.7306, + "step": 8970 + }, + { + "epoch": 0.38, + "grad_norm": 0.439453125, + "learning_rate": 0.00018281145270222782, + "loss": 2.6509, + "step": 8971 + }, + { + "epoch": 0.38, + "grad_norm": 0.4140625, + "learning_rate": 0.00018280769232490148, + "loss": 2.567, + "step": 8972 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.00018280393157497086, + "loss": 2.623, + "step": 8973 + }, + { + "epoch": 0.38, + "grad_norm": 0.388671875, + "learning_rate": 0.00018280017045245277, + "loss": 2.5984, + "step": 8974 + }, + { + "epoch": 0.38, + "grad_norm": 0.3828125, + "learning_rate": 0.0001827964089573642, + "loss": 2.6257, + "step": 8975 + }, + { + "epoch": 0.38, + "grad_norm": 0.416015625, + "learning_rate": 0.000182792647089722, + "loss": 2.567, + "step": 8976 + }, + { + "epoch": 0.38, + "grad_norm": 0.53515625, + "learning_rate": 0.00018278888484954316, + "loss": 2.6669, + "step": 8977 + }, + { + "epoch": 0.38, + "grad_norm": 0.482421875, + "learning_rate": 0.00018278512223684462, + "loss": 2.6831, + "step": 8978 + }, + { + "epoch": 0.38, + "grad_norm": 0.45703125, + "learning_rate": 0.00018278135925164327, + "loss": 2.6467, + "step": 8979 + }, + { + "epoch": 0.38, + "grad_norm": 0.470703125, + "learning_rate": 0.00018277759589395607, + "loss": 2.6232, + "step": 8980 + }, + { + "epoch": 0.38, + "grad_norm": 0.4296875, + "learning_rate": 0.0001827738321637999, + "loss": 2.7175, + "step": 8981 + }, + { + "epoch": 0.38, + "grad_norm": 0.462890625, + "learning_rate": 0.00018277006806119176, + "loss": 2.6117, + "step": 8982 + }, + { + "epoch": 0.38, + "grad_norm": 0.421875, + "learning_rate": 0.0001827663035861486, + "loss": 2.6025, + "step": 8983 + }, + { + "epoch": 0.38, + "grad_norm": 0.41015625, + "learning_rate": 0.00018276253873868728, + "loss": 2.6901, + "step": 8984 + }, + { + "epoch": 0.38, + "grad_norm": 0.453125, + "learning_rate": 0.00018275877351882477, + "loss": 2.5141, + "step": 8985 + }, + { + "epoch": 0.38, + "grad_norm": 0.43359375, + "learning_rate": 0.00018275500792657806, + "loss": 2.6504, + "step": 8986 + }, + { + "epoch": 0.38, + "grad_norm": 0.515625, + "learning_rate": 0.00018275124196196404, + "loss": 2.5542, + "step": 8987 + }, + { + "epoch": 0.38, + "grad_norm": 0.39453125, + "learning_rate": 0.00018274747562499965, + "loss": 2.5515, + "step": 8988 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.0001827437089157019, + "loss": 2.6909, + "step": 8989 + }, + { + "epoch": 0.38, + "grad_norm": 0.396484375, + "learning_rate": 0.0001827399418340877, + "loss": 2.4966, + "step": 8990 + }, + { + "epoch": 0.38, + "grad_norm": 0.455078125, + "learning_rate": 0.00018273617438017398, + "loss": 2.6292, + "step": 8991 + }, + { + "epoch": 0.38, + "grad_norm": 0.42578125, + "learning_rate": 0.00018273240655397772, + "loss": 2.7472, + "step": 8992 + }, + { + "epoch": 0.38, + "grad_norm": 0.45703125, + "learning_rate": 0.00018272863835551583, + "loss": 2.5862, + "step": 8993 + }, + { + "epoch": 0.38, + "grad_norm": 0.412109375, + "learning_rate": 0.00018272486978480537, + "loss": 2.6659, + "step": 8994 + }, + { + "epoch": 0.38, + "grad_norm": 0.478515625, + "learning_rate": 0.00018272110084186318, + "loss": 2.6188, + "step": 8995 + }, + { + "epoch": 0.38, + "grad_norm": 0.40625, + "learning_rate": 0.00018271733152670628, + "loss": 2.6197, + "step": 8996 + }, + { + "epoch": 0.38, + "grad_norm": 0.4453125, + "learning_rate": 0.0001827135618393516, + "loss": 2.6775, + "step": 8997 + }, + { + "epoch": 0.38, + "grad_norm": 0.455078125, + "learning_rate": 0.00018270979177981614, + "loss": 2.5771, + "step": 8998 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.00018270602134811686, + "loss": 2.5128, + "step": 8999 + }, + { + "epoch": 0.38, + "grad_norm": 0.51953125, + "learning_rate": 0.0001827022505442707, + "loss": 2.6276, + "step": 9000 + }, + { + "epoch": 0.38, + "grad_norm": 0.458984375, + "learning_rate": 0.00018269847936829462, + "loss": 2.4458, + "step": 9001 + }, + { + "epoch": 0.38, + "grad_norm": 0.451171875, + "learning_rate": 0.0001826947078202056, + "loss": 2.5562, + "step": 9002 + }, + { + "epoch": 0.38, + "grad_norm": 0.431640625, + "learning_rate": 0.00018269093590002062, + "loss": 2.5922, + "step": 9003 + }, + { + "epoch": 0.38, + "grad_norm": 0.515625, + "learning_rate": 0.00018268716360775666, + "loss": 2.6934, + "step": 9004 + }, + { + "epoch": 0.38, + "grad_norm": 0.423828125, + "learning_rate": 0.00018268339094343068, + "loss": 2.465, + "step": 9005 + }, + { + "epoch": 0.38, + "grad_norm": 0.46875, + "learning_rate": 0.00018267961790705966, + "loss": 2.6694, + "step": 9006 + }, + { + "epoch": 0.38, + "grad_norm": 0.439453125, + "learning_rate": 0.00018267584449866057, + "loss": 2.593, + "step": 9007 + }, + { + "epoch": 0.38, + "grad_norm": 0.498046875, + "learning_rate": 0.00018267207071825042, + "loss": 2.6043, + "step": 9008 + }, + { + "epoch": 0.38, + "grad_norm": 0.439453125, + "learning_rate": 0.00018266829656584613, + "loss": 2.7667, + "step": 9009 + }, + { + "epoch": 0.38, + "grad_norm": 0.482421875, + "learning_rate": 0.00018266452204146472, + "loss": 2.574, + "step": 9010 + }, + { + "epoch": 0.38, + "grad_norm": 0.404296875, + "learning_rate": 0.00018266074714512318, + "loss": 2.6932, + "step": 9011 + }, + { + "epoch": 0.38, + "grad_norm": 0.421875, + "learning_rate": 0.0001826569718768385, + "loss": 2.4029, + "step": 9012 + }, + { + "epoch": 0.38, + "grad_norm": 0.419921875, + "learning_rate": 0.00018265319623662763, + "loss": 2.5039, + "step": 9013 + }, + { + "epoch": 0.38, + "grad_norm": 0.44921875, + "learning_rate": 0.0001826494202245076, + "loss": 2.5187, + "step": 9014 + }, + { + "epoch": 0.38, + "grad_norm": 0.40234375, + "learning_rate": 0.00018264564384049537, + "loss": 2.6169, + "step": 9015 + }, + { + "epoch": 0.38, + "grad_norm": 0.5234375, + "learning_rate": 0.00018264186708460796, + "loss": 2.6712, + "step": 9016 + }, + { + "epoch": 0.38, + "grad_norm": 0.443359375, + "learning_rate": 0.00018263808995686235, + "loss": 2.5158, + "step": 9017 + }, + { + "epoch": 0.38, + "grad_norm": 0.44140625, + "learning_rate": 0.00018263431245727553, + "loss": 2.5876, + "step": 9018 + }, + { + "epoch": 0.38, + "grad_norm": 0.40625, + "learning_rate": 0.00018263053458586449, + "loss": 2.5349, + "step": 9019 + }, + { + "epoch": 0.38, + "grad_norm": 0.423828125, + "learning_rate": 0.00018262675634264628, + "loss": 2.549, + "step": 9020 + }, + { + "epoch": 0.38, + "grad_norm": 0.384765625, + "learning_rate": 0.00018262297772763783, + "loss": 2.5892, + "step": 9021 + }, + { + "epoch": 0.38, + "grad_norm": 0.458984375, + "learning_rate": 0.00018261919874085618, + "loss": 2.6431, + "step": 9022 + }, + { + "epoch": 0.38, + "grad_norm": 0.453125, + "learning_rate": 0.00018261541938231835, + "loss": 2.67, + "step": 9023 + }, + { + "epoch": 0.38, + "grad_norm": 0.41015625, + "learning_rate": 0.0001826116396520413, + "loss": 2.6814, + "step": 9024 + }, + { + "epoch": 0.38, + "grad_norm": 0.47265625, + "learning_rate": 0.00018260785955004209, + "loss": 2.6351, + "step": 9025 + }, + { + "epoch": 0.38, + "grad_norm": 0.478515625, + "learning_rate": 0.00018260407907633768, + "loss": 2.8127, + "step": 9026 + }, + { + "epoch": 0.38, + "grad_norm": 0.419921875, + "learning_rate": 0.00018260029823094511, + "loss": 2.6913, + "step": 9027 + }, + { + "epoch": 0.38, + "grad_norm": 0.416015625, + "learning_rate": 0.0001825965170138814, + "loss": 2.6321, + "step": 9028 + }, + { + "epoch": 0.38, + "grad_norm": 0.482421875, + "learning_rate": 0.0001825927354251635, + "loss": 2.6755, + "step": 9029 + }, + { + "epoch": 0.38, + "grad_norm": 0.447265625, + "learning_rate": 0.0001825889534648085, + "loss": 2.5925, + "step": 9030 + }, + { + "epoch": 0.38, + "grad_norm": 0.46484375, + "learning_rate": 0.0001825851711328334, + "loss": 2.6104, + "step": 9031 + }, + { + "epoch": 0.38, + "grad_norm": 0.4609375, + "learning_rate": 0.00018258138842925522, + "loss": 2.6659, + "step": 9032 + }, + { + "epoch": 0.38, + "grad_norm": 0.482421875, + "learning_rate": 0.00018257760535409098, + "loss": 2.5621, + "step": 9033 + }, + { + "epoch": 0.38, + "grad_norm": 0.435546875, + "learning_rate": 0.00018257382190735766, + "loss": 2.647, + "step": 9034 + }, + { + "epoch": 0.39, + "grad_norm": 0.42578125, + "learning_rate": 0.00018257003808907232, + "loss": 2.6658, + "step": 9035 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.000182566253899252, + "loss": 2.6521, + "step": 9036 + }, + { + "epoch": 0.39, + "grad_norm": 0.490234375, + "learning_rate": 0.0001825624693379137, + "loss": 2.6063, + "step": 9037 + }, + { + "epoch": 0.39, + "grad_norm": 0.4140625, + "learning_rate": 0.00018255868440507448, + "loss": 2.5477, + "step": 9038 + }, + { + "epoch": 0.39, + "grad_norm": 0.48046875, + "learning_rate": 0.00018255489910075134, + "loss": 2.6009, + "step": 9039 + }, + { + "epoch": 0.39, + "grad_norm": 0.419921875, + "learning_rate": 0.00018255111342496135, + "loss": 2.4381, + "step": 9040 + }, + { + "epoch": 0.39, + "grad_norm": 0.4375, + "learning_rate": 0.0001825473273777215, + "loss": 2.6344, + "step": 9041 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.00018254354095904884, + "loss": 2.6375, + "step": 9042 + }, + { + "epoch": 0.39, + "grad_norm": 0.4921875, + "learning_rate": 0.00018253975416896042, + "loss": 2.6259, + "step": 9043 + }, + { + "epoch": 0.39, + "grad_norm": 0.404296875, + "learning_rate": 0.00018253596700747323, + "loss": 2.5681, + "step": 9044 + }, + { + "epoch": 0.39, + "grad_norm": 0.447265625, + "learning_rate": 0.0001825321794746044, + "loss": 2.6388, + "step": 9045 + }, + { + "epoch": 0.39, + "grad_norm": 0.486328125, + "learning_rate": 0.0001825283915703709, + "loss": 2.7344, + "step": 9046 + }, + { + "epoch": 0.39, + "grad_norm": 0.44921875, + "learning_rate": 0.00018252460329478982, + "loss": 2.4853, + "step": 9047 + }, + { + "epoch": 0.39, + "grad_norm": 0.458984375, + "learning_rate": 0.0001825208146478782, + "loss": 2.5688, + "step": 9048 + }, + { + "epoch": 0.39, + "grad_norm": 0.5546875, + "learning_rate": 0.00018251702562965304, + "loss": 2.5556, + "step": 9049 + }, + { + "epoch": 0.39, + "grad_norm": 0.453125, + "learning_rate": 0.00018251323624013144, + "loss": 2.6561, + "step": 9050 + }, + { + "epoch": 0.39, + "grad_norm": 0.439453125, + "learning_rate": 0.0001825094464793304, + "loss": 2.5632, + "step": 9051 + }, + { + "epoch": 0.39, + "grad_norm": 0.458984375, + "learning_rate": 0.00018250565634726705, + "loss": 2.7269, + "step": 9052 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.00018250186584395833, + "loss": 2.5264, + "step": 9053 + }, + { + "epoch": 0.39, + "grad_norm": 0.416015625, + "learning_rate": 0.00018249807496942144, + "loss": 2.7377, + "step": 9054 + }, + { + "epoch": 0.39, + "grad_norm": 0.458984375, + "learning_rate": 0.0001824942837236733, + "loss": 2.6754, + "step": 9055 + }, + { + "epoch": 0.39, + "grad_norm": 0.408203125, + "learning_rate": 0.0001824904921067311, + "loss": 2.6808, + "step": 9056 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018248670011861178, + "loss": 2.6225, + "step": 9057 + }, + { + "epoch": 0.39, + "grad_norm": 0.486328125, + "learning_rate": 0.00018248290775933244, + "loss": 2.6795, + "step": 9058 + }, + { + "epoch": 0.39, + "grad_norm": 0.4609375, + "learning_rate": 0.0001824791150289102, + "loss": 2.6296, + "step": 9059 + }, + { + "epoch": 0.39, + "grad_norm": 0.4375, + "learning_rate": 0.00018247532192736208, + "loss": 2.5914, + "step": 9060 + }, + { + "epoch": 0.39, + "grad_norm": 0.435546875, + "learning_rate": 0.00018247152845470515, + "loss": 2.6102, + "step": 9061 + }, + { + "epoch": 0.39, + "grad_norm": 0.486328125, + "learning_rate": 0.00018246773461095647, + "loss": 2.5834, + "step": 9062 + }, + { + "epoch": 0.39, + "grad_norm": 0.462890625, + "learning_rate": 0.00018246394039613312, + "loss": 2.6379, + "step": 9063 + }, + { + "epoch": 0.39, + "grad_norm": 0.466796875, + "learning_rate": 0.0001824601458102522, + "loss": 2.672, + "step": 9064 + }, + { + "epoch": 0.39, + "grad_norm": 0.474609375, + "learning_rate": 0.00018245635085333074, + "loss": 2.4986, + "step": 9065 + }, + { + "epoch": 0.39, + "grad_norm": 0.478515625, + "learning_rate": 0.00018245255552538586, + "loss": 2.6426, + "step": 9066 + }, + { + "epoch": 0.39, + "grad_norm": 0.458984375, + "learning_rate": 0.00018244875982643456, + "loss": 2.6149, + "step": 9067 + }, + { + "epoch": 0.39, + "grad_norm": 0.4921875, + "learning_rate": 0.00018244496375649402, + "loss": 2.8838, + "step": 9068 + }, + { + "epoch": 0.39, + "grad_norm": 0.47265625, + "learning_rate": 0.00018244116731558124, + "loss": 2.7112, + "step": 9069 + }, + { + "epoch": 0.39, + "grad_norm": 0.478515625, + "learning_rate": 0.00018243737050371338, + "loss": 2.5222, + "step": 9070 + }, + { + "epoch": 0.39, + "grad_norm": 0.470703125, + "learning_rate": 0.00018243357332090745, + "loss": 2.644, + "step": 9071 + }, + { + "epoch": 0.39, + "grad_norm": 0.470703125, + "learning_rate": 0.00018242977576718057, + "loss": 2.8159, + "step": 9072 + }, + { + "epoch": 0.39, + "grad_norm": 0.44921875, + "learning_rate": 0.00018242597784254985, + "loss": 2.638, + "step": 9073 + }, + { + "epoch": 0.39, + "grad_norm": 0.4453125, + "learning_rate": 0.0001824221795470323, + "loss": 2.6482, + "step": 9074 + }, + { + "epoch": 0.39, + "grad_norm": 0.423828125, + "learning_rate": 0.00018241838088064509, + "loss": 2.5643, + "step": 9075 + }, + { + "epoch": 0.39, + "grad_norm": 0.43359375, + "learning_rate": 0.0001824145818434053, + "loss": 2.6073, + "step": 9076 + }, + { + "epoch": 0.39, + "grad_norm": 0.447265625, + "learning_rate": 0.00018241078243533, + "loss": 2.5669, + "step": 9077 + }, + { + "epoch": 0.39, + "grad_norm": 1.265625, + "learning_rate": 0.00018240698265643628, + "loss": 2.5359, + "step": 9078 + }, + { + "epoch": 0.39, + "grad_norm": 0.47265625, + "learning_rate": 0.00018240318250674128, + "loss": 2.7382, + "step": 9079 + }, + { + "epoch": 0.39, + "grad_norm": 0.427734375, + "learning_rate": 0.00018239938198626207, + "loss": 2.5404, + "step": 9080 + }, + { + "epoch": 0.39, + "grad_norm": 0.443359375, + "learning_rate": 0.00018239558109501577, + "loss": 2.6546, + "step": 9081 + }, + { + "epoch": 0.39, + "grad_norm": 0.47265625, + "learning_rate": 0.00018239177983301943, + "loss": 2.6663, + "step": 9082 + }, + { + "epoch": 0.39, + "grad_norm": 0.46484375, + "learning_rate": 0.0001823879782002902, + "loss": 2.6598, + "step": 9083 + }, + { + "epoch": 0.39, + "grad_norm": 0.48046875, + "learning_rate": 0.00018238417619684518, + "loss": 2.373, + "step": 9084 + }, + { + "epoch": 0.39, + "grad_norm": 0.419921875, + "learning_rate": 0.00018238037382270147, + "loss": 2.5499, + "step": 9085 + }, + { + "epoch": 0.39, + "grad_norm": 0.462890625, + "learning_rate": 0.0001823765710778762, + "loss": 2.702, + "step": 9086 + }, + { + "epoch": 0.39, + "grad_norm": 0.46875, + "learning_rate": 0.00018237276796238648, + "loss": 2.6831, + "step": 9087 + }, + { + "epoch": 0.39, + "grad_norm": 0.453125, + "learning_rate": 0.00018236896447624935, + "loss": 2.6774, + "step": 9088 + }, + { + "epoch": 0.39, + "grad_norm": 0.4453125, + "learning_rate": 0.00018236516061948205, + "loss": 2.5716, + "step": 9089 + }, + { + "epoch": 0.39, + "grad_norm": 0.4296875, + "learning_rate": 0.0001823613563921016, + "loss": 2.7018, + "step": 9090 + }, + { + "epoch": 0.39, + "grad_norm": 0.412109375, + "learning_rate": 0.00018235755179412514, + "loss": 2.4147, + "step": 9091 + }, + { + "epoch": 0.39, + "grad_norm": 0.443359375, + "learning_rate": 0.0001823537468255698, + "loss": 2.6443, + "step": 9092 + }, + { + "epoch": 0.39, + "grad_norm": 0.48046875, + "learning_rate": 0.00018234994148645268, + "loss": 2.7282, + "step": 9093 + }, + { + "epoch": 0.39, + "grad_norm": 0.4296875, + "learning_rate": 0.00018234613577679094, + "loss": 2.5458, + "step": 9094 + }, + { + "epoch": 0.39, + "grad_norm": 0.46484375, + "learning_rate": 0.00018234232969660167, + "loss": 2.5888, + "step": 9095 + }, + { + "epoch": 0.39, + "grad_norm": 0.423828125, + "learning_rate": 0.00018233852324590203, + "loss": 2.5534, + "step": 9096 + }, + { + "epoch": 0.39, + "grad_norm": 0.435546875, + "learning_rate": 0.00018233471642470911, + "loss": 2.546, + "step": 9097 + }, + { + "epoch": 0.39, + "grad_norm": 0.466796875, + "learning_rate": 0.00018233090923304007, + "loss": 2.5686, + "step": 9098 + }, + { + "epoch": 0.39, + "grad_norm": 0.4375, + "learning_rate": 0.000182327101670912, + "loss": 2.6056, + "step": 9099 + }, + { + "epoch": 0.39, + "grad_norm": 0.447265625, + "learning_rate": 0.00018232329373834208, + "loss": 2.6157, + "step": 9100 + }, + { + "epoch": 0.39, + "grad_norm": 0.4921875, + "learning_rate": 0.00018231948543534742, + "loss": 2.7092, + "step": 9101 + }, + { + "epoch": 0.39, + "grad_norm": 0.50390625, + "learning_rate": 0.00018231567676194515, + "loss": 2.6078, + "step": 9102 + }, + { + "epoch": 0.39, + "grad_norm": 0.494140625, + "learning_rate": 0.00018231186771815243, + "loss": 2.7167, + "step": 9103 + }, + { + "epoch": 0.39, + "grad_norm": 0.5078125, + "learning_rate": 0.00018230805830398637, + "loss": 2.9472, + "step": 9104 + }, + { + "epoch": 0.39, + "grad_norm": 0.43359375, + "learning_rate": 0.00018230424851946414, + "loss": 2.4603, + "step": 9105 + }, + { + "epoch": 0.39, + "grad_norm": 0.48046875, + "learning_rate": 0.00018230043836460288, + "loss": 2.5542, + "step": 9106 + }, + { + "epoch": 0.39, + "grad_norm": 0.4921875, + "learning_rate": 0.00018229662783941973, + "loss": 2.6024, + "step": 9107 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.00018229281694393182, + "loss": 2.7159, + "step": 9108 + }, + { + "epoch": 0.39, + "grad_norm": 0.470703125, + "learning_rate": 0.00018228900567815627, + "loss": 2.5587, + "step": 9109 + }, + { + "epoch": 0.39, + "grad_norm": 0.5390625, + "learning_rate": 0.00018228519404211034, + "loss": 2.4993, + "step": 9110 + }, + { + "epoch": 0.39, + "grad_norm": 0.423828125, + "learning_rate": 0.00018228138203581104, + "loss": 2.5143, + "step": 9111 + }, + { + "epoch": 0.39, + "grad_norm": 0.490234375, + "learning_rate": 0.00018227756965927565, + "loss": 2.6689, + "step": 9112 + }, + { + "epoch": 0.39, + "grad_norm": 0.4765625, + "learning_rate": 0.00018227375691252123, + "loss": 2.7272, + "step": 9113 + }, + { + "epoch": 0.39, + "grad_norm": 0.431640625, + "learning_rate": 0.00018226994379556496, + "loss": 2.6559, + "step": 9114 + }, + { + "epoch": 0.39, + "grad_norm": 0.4453125, + "learning_rate": 0.00018226613030842406, + "loss": 2.5589, + "step": 9115 + }, + { + "epoch": 0.39, + "grad_norm": 0.46484375, + "learning_rate": 0.0001822623164511156, + "loss": 2.4954, + "step": 9116 + }, + { + "epoch": 0.39, + "grad_norm": 0.431640625, + "learning_rate": 0.0001822585022236568, + "loss": 2.5146, + "step": 9117 + }, + { + "epoch": 0.39, + "grad_norm": 0.51171875, + "learning_rate": 0.00018225468762606478, + "loss": 2.5256, + "step": 9118 + }, + { + "epoch": 0.39, + "grad_norm": 0.431640625, + "learning_rate": 0.00018225087265835672, + "loss": 2.5707, + "step": 9119 + }, + { + "epoch": 0.39, + "grad_norm": 0.490234375, + "learning_rate": 0.0001822470573205498, + "loss": 2.7074, + "step": 9120 + }, + { + "epoch": 0.39, + "grad_norm": 0.41015625, + "learning_rate": 0.0001822432416126612, + "loss": 2.4995, + "step": 9121 + }, + { + "epoch": 0.39, + "grad_norm": 0.52734375, + "learning_rate": 0.00018223942553470806, + "loss": 2.6119, + "step": 9122 + }, + { + "epoch": 0.39, + "grad_norm": 0.474609375, + "learning_rate": 0.00018223560908670755, + "loss": 2.567, + "step": 9123 + }, + { + "epoch": 0.39, + "grad_norm": 0.46875, + "learning_rate": 0.00018223179226867686, + "loss": 2.6142, + "step": 9124 + }, + { + "epoch": 0.39, + "grad_norm": 0.455078125, + "learning_rate": 0.00018222797508063315, + "loss": 2.497, + "step": 9125 + }, + { + "epoch": 0.39, + "grad_norm": 0.41796875, + "learning_rate": 0.00018222415752259364, + "loss": 2.6441, + "step": 9126 + }, + { + "epoch": 0.39, + "grad_norm": 0.5, + "learning_rate": 0.00018222033959457542, + "loss": 2.7272, + "step": 9127 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018221652129659575, + "loss": 2.5464, + "step": 9128 + }, + { + "epoch": 0.39, + "grad_norm": 0.453125, + "learning_rate": 0.00018221270262867175, + "loss": 2.6752, + "step": 9129 + }, + { + "epoch": 0.39, + "grad_norm": 0.4296875, + "learning_rate": 0.00018220888359082066, + "loss": 2.5457, + "step": 9130 + }, + { + "epoch": 0.39, + "grad_norm": 0.5234375, + "learning_rate": 0.00018220506418305962, + "loss": 2.6531, + "step": 9131 + }, + { + "epoch": 0.39, + "grad_norm": 0.45703125, + "learning_rate": 0.00018220124440540585, + "loss": 2.6232, + "step": 9132 + }, + { + "epoch": 0.39, + "grad_norm": 0.435546875, + "learning_rate": 0.00018219742425787651, + "loss": 2.6249, + "step": 9133 + }, + { + "epoch": 0.39, + "grad_norm": 0.482421875, + "learning_rate": 0.00018219360374048882, + "loss": 2.6974, + "step": 9134 + }, + { + "epoch": 0.39, + "grad_norm": 0.443359375, + "learning_rate": 0.00018218978285325995, + "loss": 2.5933, + "step": 9135 + }, + { + "epoch": 0.39, + "grad_norm": 0.443359375, + "learning_rate": 0.00018218596159620706, + "loss": 2.6115, + "step": 9136 + }, + { + "epoch": 0.39, + "grad_norm": 0.478515625, + "learning_rate": 0.00018218213996934744, + "loss": 2.659, + "step": 9137 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018217831797269818, + "loss": 2.5383, + "step": 9138 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018217449560627654, + "loss": 2.659, + "step": 9139 + }, + { + "epoch": 0.39, + "grad_norm": 0.455078125, + "learning_rate": 0.00018217067287009967, + "loss": 2.5407, + "step": 9140 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018216684976418482, + "loss": 2.6333, + "step": 9141 + }, + { + "epoch": 0.39, + "grad_norm": 0.474609375, + "learning_rate": 0.0001821630262885492, + "loss": 2.7218, + "step": 9142 + }, + { + "epoch": 0.39, + "grad_norm": 0.48046875, + "learning_rate": 0.00018215920244320994, + "loss": 2.7689, + "step": 9143 + }, + { + "epoch": 0.39, + "grad_norm": 0.47265625, + "learning_rate": 0.00018215537822818432, + "loss": 2.6483, + "step": 9144 + }, + { + "epoch": 0.39, + "grad_norm": 0.45703125, + "learning_rate": 0.00018215155364348953, + "loss": 2.5131, + "step": 9145 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.00018214772868914273, + "loss": 2.6228, + "step": 9146 + }, + { + "epoch": 0.39, + "grad_norm": 0.462890625, + "learning_rate": 0.00018214390336516122, + "loss": 2.6159, + "step": 9147 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018214007767156215, + "loss": 2.5983, + "step": 9148 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.00018213625160836276, + "loss": 2.668, + "step": 9149 + }, + { + "epoch": 0.39, + "grad_norm": 0.400390625, + "learning_rate": 0.00018213242517558023, + "loss": 2.6315, + "step": 9150 + }, + { + "epoch": 0.39, + "grad_norm": 0.47265625, + "learning_rate": 0.00018212859837323184, + "loss": 2.6026, + "step": 9151 + }, + { + "epoch": 0.39, + "grad_norm": 0.47265625, + "learning_rate": 0.00018212477120133476, + "loss": 2.4984, + "step": 9152 + }, + { + "epoch": 0.39, + "grad_norm": 0.447265625, + "learning_rate": 0.0001821209436599062, + "loss": 2.5731, + "step": 9153 + }, + { + "epoch": 0.39, + "grad_norm": 0.408203125, + "learning_rate": 0.0001821171157489634, + "loss": 2.533, + "step": 9154 + }, + { + "epoch": 0.39, + "grad_norm": 0.4609375, + "learning_rate": 0.00018211328746852362, + "loss": 2.6697, + "step": 9155 + }, + { + "epoch": 0.39, + "grad_norm": 0.470703125, + "learning_rate": 0.00018210945881860403, + "loss": 2.5885, + "step": 9156 + }, + { + "epoch": 0.39, + "grad_norm": 0.478515625, + "learning_rate": 0.0001821056297992219, + "loss": 2.7187, + "step": 9157 + }, + { + "epoch": 0.39, + "grad_norm": 0.40625, + "learning_rate": 0.00018210180041039443, + "loss": 2.6419, + "step": 9158 + }, + { + "epoch": 0.39, + "grad_norm": 0.423828125, + "learning_rate": 0.0001820979706521389, + "loss": 2.622, + "step": 9159 + }, + { + "epoch": 0.39, + "grad_norm": 0.455078125, + "learning_rate": 0.00018209414052447245, + "loss": 2.6879, + "step": 9160 + }, + { + "epoch": 0.39, + "grad_norm": 0.50390625, + "learning_rate": 0.00018209031002741242, + "loss": 2.5804, + "step": 9161 + }, + { + "epoch": 0.39, + "grad_norm": 0.45703125, + "learning_rate": 0.00018208647916097593, + "loss": 2.6286, + "step": 9162 + }, + { + "epoch": 0.39, + "grad_norm": 0.439453125, + "learning_rate": 0.00018208264792518036, + "loss": 2.5593, + "step": 9163 + }, + { + "epoch": 0.39, + "grad_norm": 0.466796875, + "learning_rate": 0.00018207881632004283, + "loss": 2.5839, + "step": 9164 + }, + { + "epoch": 0.39, + "grad_norm": 0.52734375, + "learning_rate": 0.00018207498434558064, + "loss": 2.7104, + "step": 9165 + }, + { + "epoch": 0.39, + "grad_norm": 0.435546875, + "learning_rate": 0.00018207115200181097, + "loss": 2.7034, + "step": 9166 + }, + { + "epoch": 0.39, + "grad_norm": 0.470703125, + "learning_rate": 0.00018206731928875115, + "loss": 2.6919, + "step": 9167 + }, + { + "epoch": 0.39, + "grad_norm": 0.43359375, + "learning_rate": 0.0001820634862064184, + "loss": 2.6305, + "step": 9168 + }, + { + "epoch": 0.39, + "grad_norm": 0.40625, + "learning_rate": 0.0001820596527548299, + "loss": 2.6897, + "step": 9169 + }, + { + "epoch": 0.39, + "grad_norm": 0.4921875, + "learning_rate": 0.000182055818934003, + "loss": 2.5958, + "step": 9170 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018205198474395489, + "loss": 2.5844, + "step": 9171 + }, + { + "epoch": 0.39, + "grad_norm": 0.4453125, + "learning_rate": 0.00018204815018470284, + "loss": 2.6556, + "step": 9172 + }, + { + "epoch": 0.39, + "grad_norm": 0.515625, + "learning_rate": 0.0001820443152562641, + "loss": 2.6923, + "step": 9173 + }, + { + "epoch": 0.39, + "grad_norm": 0.41015625, + "learning_rate": 0.00018204047995865592, + "loss": 2.5814, + "step": 9174 + }, + { + "epoch": 0.39, + "grad_norm": 0.53515625, + "learning_rate": 0.00018203664429189553, + "loss": 2.5018, + "step": 9175 + }, + { + "epoch": 0.39, + "grad_norm": 0.486328125, + "learning_rate": 0.00018203280825600026, + "loss": 2.7729, + "step": 9176 + }, + { + "epoch": 0.39, + "grad_norm": 0.453125, + "learning_rate": 0.00018202897185098734, + "loss": 2.6419, + "step": 9177 + }, + { + "epoch": 0.39, + "grad_norm": 0.466796875, + "learning_rate": 0.000182025135076874, + "loss": 2.6783, + "step": 9178 + }, + { + "epoch": 0.39, + "grad_norm": 0.43359375, + "learning_rate": 0.00018202129793367755, + "loss": 2.5279, + "step": 9179 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.00018201746042141522, + "loss": 2.564, + "step": 9180 + }, + { + "epoch": 0.39, + "grad_norm": 0.4453125, + "learning_rate": 0.00018201362254010429, + "loss": 2.7307, + "step": 9181 + }, + { + "epoch": 0.39, + "grad_norm": 0.41015625, + "learning_rate": 0.00018200978428976203, + "loss": 2.7163, + "step": 9182 + }, + { + "epoch": 0.39, + "grad_norm": 0.474609375, + "learning_rate": 0.0001820059456704057, + "loss": 2.5696, + "step": 9183 + }, + { + "epoch": 0.39, + "grad_norm": 0.5, + "learning_rate": 0.00018200210668205263, + "loss": 2.61, + "step": 9184 + }, + { + "epoch": 0.39, + "grad_norm": 0.470703125, + "learning_rate": 0.00018199826732471998, + "loss": 2.4709, + "step": 9185 + }, + { + "epoch": 0.39, + "grad_norm": 0.49609375, + "learning_rate": 0.00018199442759842517, + "loss": 2.6373, + "step": 9186 + }, + { + "epoch": 0.39, + "grad_norm": 0.53515625, + "learning_rate": 0.00018199058750318535, + "loss": 2.6728, + "step": 9187 + }, + { + "epoch": 0.39, + "grad_norm": 0.47265625, + "learning_rate": 0.0001819867470390179, + "loss": 2.7036, + "step": 9188 + }, + { + "epoch": 0.39, + "grad_norm": 0.427734375, + "learning_rate": 0.00018198290620593998, + "loss": 2.66, + "step": 9189 + }, + { + "epoch": 0.39, + "grad_norm": 0.45703125, + "learning_rate": 0.00018197906500396898, + "loss": 2.6305, + "step": 9190 + }, + { + "epoch": 0.39, + "grad_norm": 0.419921875, + "learning_rate": 0.00018197522343312215, + "loss": 2.5945, + "step": 9191 + }, + { + "epoch": 0.39, + "grad_norm": 0.44921875, + "learning_rate": 0.00018197138149341675, + "loss": 2.551, + "step": 9192 + }, + { + "epoch": 0.39, + "grad_norm": 0.478515625, + "learning_rate": 0.00018196753918487012, + "loss": 2.6305, + "step": 9193 + }, + { + "epoch": 0.39, + "grad_norm": 0.4375, + "learning_rate": 0.0001819636965074995, + "loss": 2.6625, + "step": 9194 + }, + { + "epoch": 0.39, + "grad_norm": 0.53125, + "learning_rate": 0.00018195985346132221, + "loss": 2.698, + "step": 9195 + }, + { + "epoch": 0.39, + "grad_norm": 0.5, + "learning_rate": 0.00018195601004635553, + "loss": 2.7821, + "step": 9196 + }, + { + "epoch": 0.39, + "grad_norm": 0.494140625, + "learning_rate": 0.00018195216626261677, + "loss": 2.6109, + "step": 9197 + }, + { + "epoch": 0.39, + "grad_norm": 0.494140625, + "learning_rate": 0.00018194832211012322, + "loss": 2.6437, + "step": 9198 + }, + { + "epoch": 0.39, + "grad_norm": 0.466796875, + "learning_rate": 0.00018194447758889212, + "loss": 2.6116, + "step": 9199 + }, + { + "epoch": 0.39, + "grad_norm": 0.46875, + "learning_rate": 0.00018194063269894086, + "loss": 2.5551, + "step": 9200 + }, + { + "epoch": 0.39, + "grad_norm": 0.482421875, + "learning_rate": 0.00018193678744028667, + "loss": 2.7235, + "step": 9201 + }, + { + "epoch": 0.39, + "grad_norm": 0.412109375, + "learning_rate": 0.0001819329418129469, + "loss": 2.6325, + "step": 9202 + }, + { + "epoch": 0.39, + "grad_norm": 0.447265625, + "learning_rate": 0.00018192909581693885, + "loss": 2.6389, + "step": 9203 + }, + { + "epoch": 0.39, + "grad_norm": 0.421875, + "learning_rate": 0.0001819252494522798, + "loss": 2.546, + "step": 9204 + }, + { + "epoch": 0.39, + "grad_norm": 0.427734375, + "learning_rate": 0.00018192140271898705, + "loss": 2.5255, + "step": 9205 + }, + { + "epoch": 0.39, + "grad_norm": 0.42578125, + "learning_rate": 0.00018191755561707793, + "loss": 2.5607, + "step": 9206 + }, + { + "epoch": 0.39, + "grad_norm": 0.46484375, + "learning_rate": 0.00018191370814656975, + "loss": 2.6258, + "step": 9207 + }, + { + "epoch": 0.39, + "grad_norm": 0.404296875, + "learning_rate": 0.00018190986030747983, + "loss": 2.4877, + "step": 9208 + }, + { + "epoch": 0.39, + "grad_norm": 0.443359375, + "learning_rate": 0.00018190601209982544, + "loss": 2.5808, + "step": 9209 + }, + { + "epoch": 0.39, + "grad_norm": 0.419921875, + "learning_rate": 0.00018190216352362398, + "loss": 2.5044, + "step": 9210 + }, + { + "epoch": 0.39, + "grad_norm": 0.478515625, + "learning_rate": 0.0001818983145788927, + "loss": 2.5941, + "step": 9211 + }, + { + "epoch": 0.39, + "grad_norm": 0.55078125, + "learning_rate": 0.00018189446526564891, + "loss": 2.6266, + "step": 9212 + }, + { + "epoch": 0.39, + "grad_norm": 0.423828125, + "learning_rate": 0.00018189061558390997, + "loss": 2.6241, + "step": 9213 + }, + { + "epoch": 0.39, + "grad_norm": 0.447265625, + "learning_rate": 0.00018188676553369319, + "loss": 2.6791, + "step": 9214 + }, + { + "epoch": 0.39, + "grad_norm": 0.4453125, + "learning_rate": 0.0001818829151150159, + "loss": 2.5226, + "step": 9215 + }, + { + "epoch": 0.39, + "grad_norm": 0.462890625, + "learning_rate": 0.0001818790643278954, + "loss": 2.602, + "step": 9216 + }, + { + "epoch": 0.39, + "grad_norm": 0.5, + "learning_rate": 0.00018187521317234904, + "loss": 2.6947, + "step": 9217 + }, + { + "epoch": 0.39, + "grad_norm": 0.482421875, + "learning_rate": 0.00018187136164839415, + "loss": 2.6518, + "step": 9218 + }, + { + "epoch": 0.39, + "grad_norm": 0.390625, + "learning_rate": 0.00018186750975604805, + "loss": 2.5804, + "step": 9219 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018186365749532808, + "loss": 2.4672, + "step": 9220 + }, + { + "epoch": 0.39, + "grad_norm": 0.455078125, + "learning_rate": 0.00018185980486625156, + "loss": 2.6473, + "step": 9221 + }, + { + "epoch": 0.39, + "grad_norm": 0.44921875, + "learning_rate": 0.0001818559518688358, + "loss": 2.6097, + "step": 9222 + }, + { + "epoch": 0.39, + "grad_norm": 0.462890625, + "learning_rate": 0.00018185209850309824, + "loss": 2.627, + "step": 9223 + }, + { + "epoch": 0.39, + "grad_norm": 0.44921875, + "learning_rate": 0.0001818482447690561, + "loss": 2.6562, + "step": 9224 + }, + { + "epoch": 0.39, + "grad_norm": 0.458984375, + "learning_rate": 0.0001818443906667268, + "loss": 2.7237, + "step": 9225 + }, + { + "epoch": 0.39, + "grad_norm": 0.484375, + "learning_rate": 0.00018184053619612765, + "loss": 2.6377, + "step": 9226 + }, + { + "epoch": 0.39, + "grad_norm": 0.494140625, + "learning_rate": 0.00018183668135727596, + "loss": 2.6955, + "step": 9227 + }, + { + "epoch": 0.39, + "grad_norm": 0.416015625, + "learning_rate": 0.00018183282615018916, + "loss": 2.6429, + "step": 9228 + }, + { + "epoch": 0.39, + "grad_norm": 0.45703125, + "learning_rate": 0.00018182897057488452, + "loss": 2.5199, + "step": 9229 + }, + { + "epoch": 0.39, + "grad_norm": 0.486328125, + "learning_rate": 0.0001818251146313794, + "loss": 2.6375, + "step": 9230 + }, + { + "epoch": 0.39, + "grad_norm": 0.470703125, + "learning_rate": 0.00018182125831969122, + "loss": 2.6077, + "step": 9231 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.00018181740163983724, + "loss": 2.6171, + "step": 9232 + }, + { + "epoch": 0.39, + "grad_norm": 0.421875, + "learning_rate": 0.0001818135445918349, + "loss": 2.5825, + "step": 9233 + }, + { + "epoch": 0.39, + "grad_norm": 0.439453125, + "learning_rate": 0.00018180968717570144, + "loss": 2.6706, + "step": 9234 + }, + { + "epoch": 0.39, + "grad_norm": 0.48828125, + "learning_rate": 0.00018180582939145433, + "loss": 2.6358, + "step": 9235 + }, + { + "epoch": 0.39, + "grad_norm": 0.48828125, + "learning_rate": 0.00018180197123911088, + "loss": 2.6657, + "step": 9236 + }, + { + "epoch": 0.39, + "grad_norm": 0.486328125, + "learning_rate": 0.00018179811271868844, + "loss": 2.5638, + "step": 9237 + }, + { + "epoch": 0.39, + "grad_norm": 0.439453125, + "learning_rate": 0.00018179425383020437, + "loss": 2.5682, + "step": 9238 + }, + { + "epoch": 0.39, + "grad_norm": 0.408203125, + "learning_rate": 0.0001817903945736761, + "loss": 2.6816, + "step": 9239 + }, + { + "epoch": 0.39, + "grad_norm": 0.45703125, + "learning_rate": 0.00018178653494912091, + "loss": 2.6166, + "step": 9240 + }, + { + "epoch": 0.39, + "grad_norm": 0.439453125, + "learning_rate": 0.00018178267495655624, + "loss": 2.7231, + "step": 9241 + }, + { + "epoch": 0.39, + "grad_norm": 0.455078125, + "learning_rate": 0.00018177881459599939, + "loss": 2.6718, + "step": 9242 + }, + { + "epoch": 0.39, + "grad_norm": 0.5, + "learning_rate": 0.00018177495386746776, + "loss": 2.6075, + "step": 9243 + }, + { + "epoch": 0.39, + "grad_norm": 0.443359375, + "learning_rate": 0.00018177109277097873, + "loss": 2.669, + "step": 9244 + }, + { + "epoch": 0.39, + "grad_norm": 0.421875, + "learning_rate": 0.00018176723130654968, + "loss": 2.7127, + "step": 9245 + }, + { + "epoch": 0.39, + "grad_norm": 0.478515625, + "learning_rate": 0.00018176336947419794, + "loss": 2.6532, + "step": 9246 + }, + { + "epoch": 0.39, + "grad_norm": 0.44921875, + "learning_rate": 0.00018175950727394096, + "loss": 2.5968, + "step": 9247 + }, + { + "epoch": 0.39, + "grad_norm": 0.435546875, + "learning_rate": 0.00018175564470579602, + "loss": 2.6423, + "step": 9248 + }, + { + "epoch": 0.39, + "grad_norm": 0.435546875, + "learning_rate": 0.0001817517817697806, + "loss": 2.743, + "step": 9249 + }, + { + "epoch": 0.39, + "grad_norm": 0.408203125, + "learning_rate": 0.000181747918465912, + "loss": 2.5753, + "step": 9250 + }, + { + "epoch": 0.39, + "grad_norm": 0.447265625, + "learning_rate": 0.0001817440547942077, + "loss": 2.7169, + "step": 9251 + }, + { + "epoch": 0.39, + "grad_norm": 0.7421875, + "learning_rate": 0.000181740190754685, + "loss": 2.7196, + "step": 9252 + }, + { + "epoch": 0.39, + "grad_norm": 0.427734375, + "learning_rate": 0.0001817363263473613, + "loss": 2.7122, + "step": 9253 + }, + { + "epoch": 0.39, + "grad_norm": 0.451171875, + "learning_rate": 0.00018173246157225404, + "loss": 2.6821, + "step": 9254 + }, + { + "epoch": 0.39, + "grad_norm": 0.4921875, + "learning_rate": 0.00018172859642938053, + "loss": 2.6731, + "step": 9255 + }, + { + "epoch": 0.39, + "grad_norm": 0.373046875, + "learning_rate": 0.00018172473091875824, + "loss": 2.6174, + "step": 9256 + }, + { + "epoch": 0.39, + "grad_norm": 0.455078125, + "learning_rate": 0.00018172086504040448, + "loss": 2.6158, + "step": 9257 + }, + { + "epoch": 0.39, + "grad_norm": 0.5078125, + "learning_rate": 0.00018171699879433674, + "loss": 2.7032, + "step": 9258 + }, + { + "epoch": 0.39, + "grad_norm": 0.44140625, + "learning_rate": 0.00018171313218057235, + "loss": 2.7388, + "step": 9259 + }, + { + "epoch": 0.39, + "grad_norm": 0.47265625, + "learning_rate": 0.00018170926519912875, + "loss": 2.6387, + "step": 9260 + }, + { + "epoch": 0.39, + "grad_norm": 0.46484375, + "learning_rate": 0.0001817053978500233, + "loss": 2.7109, + "step": 9261 + }, + { + "epoch": 0.39, + "grad_norm": 0.4765625, + "learning_rate": 0.0001817015301332734, + "loss": 2.609, + "step": 9262 + }, + { + "epoch": 0.39, + "grad_norm": 0.41015625, + "learning_rate": 0.00018169766204889648, + "loss": 2.6247, + "step": 9263 + }, + { + "epoch": 0.39, + "grad_norm": 0.43359375, + "learning_rate": 0.00018169379359690994, + "loss": 2.5833, + "step": 9264 + }, + { + "epoch": 0.39, + "grad_norm": 0.3984375, + "learning_rate": 0.0001816899247773312, + "loss": 2.6421, + "step": 9265 + }, + { + "epoch": 0.39, + "grad_norm": 0.52734375, + "learning_rate": 0.00018168605559017767, + "loss": 2.7556, + "step": 9266 + }, + { + "epoch": 0.39, + "grad_norm": 0.55078125, + "learning_rate": 0.00018168218603546672, + "loss": 2.7428, + "step": 9267 + }, + { + "epoch": 0.39, + "grad_norm": 0.4453125, + "learning_rate": 0.00018167831611321574, + "loss": 2.5596, + "step": 9268 + }, + { + "epoch": 0.39, + "grad_norm": 0.44921875, + "learning_rate": 0.00018167444582344226, + "loss": 2.6081, + "step": 9269 + }, + { + "epoch": 0.4, + "grad_norm": 0.46875, + "learning_rate": 0.0001816705751661636, + "loss": 2.6586, + "step": 9270 + }, + { + "epoch": 0.4, + "grad_norm": 0.49609375, + "learning_rate": 0.00018166670414139716, + "loss": 2.6584, + "step": 9271 + }, + { + "epoch": 0.4, + "grad_norm": 0.4375, + "learning_rate": 0.00018166283274916045, + "loss": 2.6984, + "step": 9272 + }, + { + "epoch": 0.4, + "grad_norm": 0.43359375, + "learning_rate": 0.00018165896098947082, + "loss": 2.5499, + "step": 9273 + }, + { + "epoch": 0.4, + "grad_norm": 0.474609375, + "learning_rate": 0.0001816550888623457, + "loss": 2.7561, + "step": 9274 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018165121636780254, + "loss": 2.6569, + "step": 9275 + }, + { + "epoch": 0.4, + "grad_norm": 0.400390625, + "learning_rate": 0.00018164734350585872, + "loss": 2.627, + "step": 9276 + }, + { + "epoch": 0.4, + "grad_norm": 0.5, + "learning_rate": 0.00018164347027653171, + "loss": 2.6784, + "step": 9277 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.0001816395966798389, + "loss": 2.5621, + "step": 9278 + }, + { + "epoch": 0.4, + "grad_norm": 0.466796875, + "learning_rate": 0.00018163572271579778, + "loss": 2.6409, + "step": 9279 + }, + { + "epoch": 0.4, + "grad_norm": 0.43359375, + "learning_rate": 0.0001816318483844257, + "loss": 2.6262, + "step": 9280 + }, + { + "epoch": 0.4, + "grad_norm": 0.451171875, + "learning_rate": 0.00018162797368574015, + "loss": 2.4231, + "step": 9281 + }, + { + "epoch": 0.4, + "grad_norm": 0.427734375, + "learning_rate": 0.00018162409861975857, + "loss": 2.6395, + "step": 9282 + }, + { + "epoch": 0.4, + "grad_norm": 0.51953125, + "learning_rate": 0.00018162022318649835, + "loss": 2.7127, + "step": 9283 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.00018161634738597698, + "loss": 2.5433, + "step": 9284 + }, + { + "epoch": 0.4, + "grad_norm": 0.494140625, + "learning_rate": 0.00018161247121821185, + "loss": 2.6215, + "step": 9285 + }, + { + "epoch": 0.4, + "grad_norm": 0.47265625, + "learning_rate": 0.00018160859468322043, + "loss": 2.6788, + "step": 9286 + }, + { + "epoch": 0.4, + "grad_norm": 0.53515625, + "learning_rate": 0.00018160471778102017, + "loss": 2.7681, + "step": 9287 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.0001816008405116285, + "loss": 2.5924, + "step": 9288 + }, + { + "epoch": 0.4, + "grad_norm": 0.41796875, + "learning_rate": 0.00018159696287506285, + "loss": 2.615, + "step": 9289 + }, + { + "epoch": 0.4, + "grad_norm": 0.48046875, + "learning_rate": 0.0001815930848713407, + "loss": 2.614, + "step": 9290 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018158920650047952, + "loss": 2.5367, + "step": 9291 + }, + { + "epoch": 0.4, + "grad_norm": 0.4453125, + "learning_rate": 0.00018158532776249667, + "loss": 2.4684, + "step": 9292 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018158144865740972, + "loss": 2.7041, + "step": 9293 + }, + { + "epoch": 0.4, + "grad_norm": 0.435546875, + "learning_rate": 0.00018157756918523604, + "loss": 2.5286, + "step": 9294 + }, + { + "epoch": 0.4, + "grad_norm": 0.50390625, + "learning_rate": 0.00018157368934599308, + "loss": 2.8053, + "step": 9295 + }, + { + "epoch": 0.4, + "grad_norm": 0.44140625, + "learning_rate": 0.00018156980913969835, + "loss": 2.5608, + "step": 9296 + }, + { + "epoch": 0.4, + "grad_norm": 0.494140625, + "learning_rate": 0.0001815659285663693, + "loss": 2.6146, + "step": 9297 + }, + { + "epoch": 0.4, + "grad_norm": 0.50390625, + "learning_rate": 0.00018156204762602337, + "loss": 2.5515, + "step": 9298 + }, + { + "epoch": 0.4, + "grad_norm": 0.5234375, + "learning_rate": 0.00018155816631867805, + "loss": 2.7694, + "step": 9299 + }, + { + "epoch": 0.4, + "grad_norm": 0.41015625, + "learning_rate": 0.0001815542846443508, + "loss": 2.687, + "step": 9300 + }, + { + "epoch": 0.4, + "grad_norm": 0.44921875, + "learning_rate": 0.00018155040260305904, + "loss": 2.715, + "step": 9301 + }, + { + "epoch": 0.4, + "grad_norm": 0.3828125, + "learning_rate": 0.00018154652019482028, + "loss": 2.656, + "step": 9302 + }, + { + "epoch": 0.4, + "grad_norm": 0.494140625, + "learning_rate": 0.00018154263741965195, + "loss": 2.6085, + "step": 9303 + }, + { + "epoch": 0.4, + "grad_norm": 0.462890625, + "learning_rate": 0.0001815387542775716, + "loss": 2.6925, + "step": 9304 + }, + { + "epoch": 0.4, + "grad_norm": 0.486328125, + "learning_rate": 0.00018153487076859665, + "loss": 2.6421, + "step": 9305 + }, + { + "epoch": 0.4, + "grad_norm": 0.40625, + "learning_rate": 0.00018153098689274456, + "loss": 2.697, + "step": 9306 + }, + { + "epoch": 0.4, + "grad_norm": 0.453125, + "learning_rate": 0.00018152710265003283, + "loss": 2.7565, + "step": 9307 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.00018152321804047893, + "loss": 2.6199, + "step": 9308 + }, + { + "epoch": 0.4, + "grad_norm": 0.412109375, + "learning_rate": 0.00018151933306410038, + "loss": 2.6586, + "step": 9309 + }, + { + "epoch": 0.4, + "grad_norm": 0.412109375, + "learning_rate": 0.00018151544772091458, + "loss": 2.6214, + "step": 9310 + }, + { + "epoch": 0.4, + "grad_norm": 0.421875, + "learning_rate": 0.00018151156201093907, + "loss": 2.6162, + "step": 9311 + }, + { + "epoch": 0.4, + "grad_norm": 0.494140625, + "learning_rate": 0.00018150767593419135, + "loss": 2.8508, + "step": 9312 + }, + { + "epoch": 0.4, + "grad_norm": 0.439453125, + "learning_rate": 0.0001815037894906888, + "loss": 2.6723, + "step": 9313 + }, + { + "epoch": 0.4, + "grad_norm": 0.44140625, + "learning_rate": 0.00018149990268044906, + "loss": 2.6011, + "step": 9314 + }, + { + "epoch": 0.4, + "grad_norm": 0.408203125, + "learning_rate": 0.00018149601550348948, + "loss": 2.6216, + "step": 9315 + }, + { + "epoch": 0.4, + "grad_norm": 0.45703125, + "learning_rate": 0.00018149212795982766, + "loss": 2.7897, + "step": 9316 + }, + { + "epoch": 0.4, + "grad_norm": 0.455078125, + "learning_rate": 0.00018148824004948102, + "loss": 2.5841, + "step": 9317 + }, + { + "epoch": 0.4, + "grad_norm": 0.41796875, + "learning_rate": 0.0001814843517724671, + "loss": 2.4831, + "step": 9318 + }, + { + "epoch": 0.4, + "grad_norm": 0.490234375, + "learning_rate": 0.00018148046312880335, + "loss": 2.8301, + "step": 9319 + }, + { + "epoch": 0.4, + "grad_norm": 0.65234375, + "learning_rate": 0.00018147657411850734, + "loss": 2.6607, + "step": 9320 + }, + { + "epoch": 0.4, + "grad_norm": 0.453125, + "learning_rate": 0.0001814726847415965, + "loss": 2.7327, + "step": 9321 + }, + { + "epoch": 0.4, + "grad_norm": 0.451171875, + "learning_rate": 0.00018146879499808833, + "loss": 2.5687, + "step": 9322 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.00018146490488800038, + "loss": 2.5266, + "step": 9323 + }, + { + "epoch": 0.4, + "grad_norm": 0.484375, + "learning_rate": 0.0001814610144113501, + "loss": 2.5998, + "step": 9324 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.00018145712356815508, + "loss": 2.603, + "step": 9325 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018145323235843272, + "loss": 2.6057, + "step": 9326 + }, + { + "epoch": 0.4, + "grad_norm": 0.48828125, + "learning_rate": 0.0001814493407822006, + "loss": 2.5152, + "step": 9327 + }, + { + "epoch": 0.4, + "grad_norm": 0.4375, + "learning_rate": 0.00018144544883947621, + "loss": 2.6389, + "step": 9328 + }, + { + "epoch": 0.4, + "grad_norm": 0.4921875, + "learning_rate": 0.0001814415565302771, + "loss": 2.604, + "step": 9329 + }, + { + "epoch": 0.4, + "grad_norm": 0.44140625, + "learning_rate": 0.00018143766385462072, + "loss": 2.5366, + "step": 9330 + }, + { + "epoch": 0.4, + "grad_norm": 0.435546875, + "learning_rate": 0.00018143377081252462, + "loss": 2.5779, + "step": 9331 + }, + { + "epoch": 0.4, + "grad_norm": 0.52734375, + "learning_rate": 0.0001814298774040063, + "loss": 2.6486, + "step": 9332 + }, + { + "epoch": 0.4, + "grad_norm": 0.478515625, + "learning_rate": 0.0001814259836290833, + "loss": 2.545, + "step": 9333 + }, + { + "epoch": 0.4, + "grad_norm": 0.5546875, + "learning_rate": 0.00018142208948777313, + "loss": 2.5555, + "step": 9334 + }, + { + "epoch": 0.4, + "grad_norm": 0.4765625, + "learning_rate": 0.00018141819498009333, + "loss": 2.5359, + "step": 9335 + }, + { + "epoch": 0.4, + "grad_norm": 0.486328125, + "learning_rate": 0.00018141430010606136, + "loss": 2.636, + "step": 9336 + }, + { + "epoch": 0.4, + "grad_norm": 0.451171875, + "learning_rate": 0.00018141040486569483, + "loss": 2.6821, + "step": 9337 + }, + { + "epoch": 0.4, + "grad_norm": 0.39453125, + "learning_rate": 0.0001814065092590112, + "loss": 2.5141, + "step": 9338 + }, + { + "epoch": 0.4, + "grad_norm": 0.478515625, + "learning_rate": 0.00018140261328602808, + "loss": 2.6266, + "step": 9339 + }, + { + "epoch": 0.4, + "grad_norm": 0.490234375, + "learning_rate": 0.0001813987169467629, + "loss": 2.5953, + "step": 9340 + }, + { + "epoch": 0.4, + "grad_norm": 0.45703125, + "learning_rate": 0.00018139482024123326, + "loss": 2.6517, + "step": 9341 + }, + { + "epoch": 0.4, + "grad_norm": 0.44140625, + "learning_rate": 0.0001813909231694567, + "loss": 2.6907, + "step": 9342 + }, + { + "epoch": 0.4, + "grad_norm": 0.439453125, + "learning_rate": 0.00018138702573145072, + "loss": 2.596, + "step": 9343 + }, + { + "epoch": 0.4, + "grad_norm": 0.4453125, + "learning_rate": 0.00018138312792723285, + "loss": 2.5859, + "step": 9344 + }, + { + "epoch": 0.4, + "grad_norm": 0.486328125, + "learning_rate": 0.00018137922975682065, + "loss": 2.6309, + "step": 9345 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.00018137533122023167, + "loss": 2.4724, + "step": 9346 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.0001813714323174834, + "loss": 2.8202, + "step": 9347 + }, + { + "epoch": 0.4, + "grad_norm": 0.44140625, + "learning_rate": 0.00018136753304859345, + "loss": 2.6802, + "step": 9348 + }, + { + "epoch": 0.4, + "grad_norm": 0.46875, + "learning_rate": 0.00018136363341357935, + "loss": 2.5812, + "step": 9349 + }, + { + "epoch": 0.4, + "grad_norm": 0.5546875, + "learning_rate": 0.00018135973341245864, + "loss": 2.8395, + "step": 9350 + }, + { + "epoch": 0.4, + "grad_norm": 0.47265625, + "learning_rate": 0.00018135583304524885, + "loss": 2.5765, + "step": 9351 + }, + { + "epoch": 0.4, + "grad_norm": 0.41015625, + "learning_rate": 0.00018135193231196754, + "loss": 2.6044, + "step": 9352 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.00018134803121263226, + "loss": 2.7106, + "step": 9353 + }, + { + "epoch": 0.4, + "grad_norm": 0.466796875, + "learning_rate": 0.00018134412974726057, + "loss": 2.5746, + "step": 9354 + }, + { + "epoch": 0.4, + "grad_norm": 0.51953125, + "learning_rate": 0.00018134022791587004, + "loss": 2.648, + "step": 9355 + }, + { + "epoch": 0.4, + "grad_norm": 0.482421875, + "learning_rate": 0.0001813363257184782, + "loss": 2.7416, + "step": 9356 + }, + { + "epoch": 0.4, + "grad_norm": 0.486328125, + "learning_rate": 0.00018133242315510262, + "loss": 2.6404, + "step": 9357 + }, + { + "epoch": 0.4, + "grad_norm": 0.451171875, + "learning_rate": 0.00018132852022576085, + "loss": 2.6239, + "step": 9358 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.00018132461693047048, + "loss": 2.6574, + "step": 9359 + }, + { + "epoch": 0.4, + "grad_norm": 0.453125, + "learning_rate": 0.00018132071326924904, + "loss": 2.5709, + "step": 9360 + }, + { + "epoch": 0.4, + "grad_norm": 0.42578125, + "learning_rate": 0.00018131680924211413, + "loss": 2.6936, + "step": 9361 + }, + { + "epoch": 0.4, + "grad_norm": 0.4921875, + "learning_rate": 0.00018131290484908326, + "loss": 2.6819, + "step": 9362 + }, + { + "epoch": 0.4, + "grad_norm": 0.451171875, + "learning_rate": 0.0001813090000901741, + "loss": 2.6589, + "step": 9363 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.0001813050949654041, + "loss": 2.7336, + "step": 9364 + }, + { + "epoch": 0.4, + "grad_norm": 0.484375, + "learning_rate": 0.00018130118947479087, + "loss": 2.6572, + "step": 9365 + }, + { + "epoch": 0.4, + "grad_norm": 0.5, + "learning_rate": 0.00018129728361835203, + "loss": 2.6805, + "step": 9366 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018129337739610513, + "loss": 2.7222, + "step": 9367 + }, + { + "epoch": 0.4, + "grad_norm": 0.4375, + "learning_rate": 0.00018128947080806772, + "loss": 2.6173, + "step": 9368 + }, + { + "epoch": 0.4, + "grad_norm": 0.455078125, + "learning_rate": 0.00018128556385425742, + "loss": 2.6969, + "step": 9369 + }, + { + "epoch": 0.4, + "grad_norm": 0.44921875, + "learning_rate": 0.00018128165653469175, + "loss": 2.5528, + "step": 9370 + }, + { + "epoch": 0.4, + "grad_norm": 0.48828125, + "learning_rate": 0.00018127774884938838, + "loss": 2.5894, + "step": 9371 + }, + { + "epoch": 0.4, + "grad_norm": 0.453125, + "learning_rate": 0.00018127384079836482, + "loss": 2.5822, + "step": 9372 + }, + { + "epoch": 0.4, + "grad_norm": 0.412109375, + "learning_rate": 0.0001812699323816387, + "loss": 2.5871, + "step": 9373 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018126602359922758, + "loss": 2.6283, + "step": 9374 + }, + { + "epoch": 0.4, + "grad_norm": 0.421875, + "learning_rate": 0.000181262114451149, + "loss": 2.6745, + "step": 9375 + }, + { + "epoch": 0.4, + "grad_norm": 0.455078125, + "learning_rate": 0.00018125820493742066, + "loss": 2.5307, + "step": 9376 + }, + { + "epoch": 0.4, + "grad_norm": 0.435546875, + "learning_rate": 0.0001812542950580601, + "loss": 2.5649, + "step": 9377 + }, + { + "epoch": 0.4, + "grad_norm": 0.400390625, + "learning_rate": 0.00018125038481308487, + "loss": 2.4859, + "step": 9378 + }, + { + "epoch": 0.4, + "grad_norm": 0.431640625, + "learning_rate": 0.00018124647420251263, + "loss": 2.5587, + "step": 9379 + }, + { + "epoch": 0.4, + "grad_norm": 0.44921875, + "learning_rate": 0.00018124256322636092, + "loss": 2.6817, + "step": 9380 + }, + { + "epoch": 0.4, + "grad_norm": 0.447265625, + "learning_rate": 0.00018123865188464738, + "loss": 2.6179, + "step": 9381 + }, + { + "epoch": 0.4, + "grad_norm": 0.490234375, + "learning_rate": 0.0001812347401773896, + "loss": 2.6035, + "step": 9382 + }, + { + "epoch": 0.4, + "grad_norm": 0.51953125, + "learning_rate": 0.00018123082810460518, + "loss": 2.704, + "step": 9383 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.00018122691566631172, + "loss": 2.6641, + "step": 9384 + }, + { + "epoch": 0.4, + "grad_norm": 0.4453125, + "learning_rate": 0.0001812230028625268, + "loss": 2.7061, + "step": 9385 + }, + { + "epoch": 0.4, + "grad_norm": 0.466796875, + "learning_rate": 0.00018121908969326807, + "loss": 2.8118, + "step": 9386 + }, + { + "epoch": 0.4, + "grad_norm": 0.408203125, + "learning_rate": 0.00018121517615855314, + "loss": 2.6048, + "step": 9387 + }, + { + "epoch": 0.4, + "grad_norm": 0.42578125, + "learning_rate": 0.00018121126225839956, + "loss": 2.6207, + "step": 9388 + }, + { + "epoch": 0.4, + "grad_norm": 0.50390625, + "learning_rate": 0.000181207347992825, + "loss": 2.6038, + "step": 9389 + }, + { + "epoch": 0.4, + "grad_norm": 0.4140625, + "learning_rate": 0.00018120343336184707, + "loss": 2.6401, + "step": 9390 + }, + { + "epoch": 0.4, + "grad_norm": 0.45703125, + "learning_rate": 0.00018119951836548332, + "loss": 2.6355, + "step": 9391 + }, + { + "epoch": 0.4, + "grad_norm": 0.4375, + "learning_rate": 0.00018119560300375144, + "loss": 2.5424, + "step": 9392 + }, + { + "epoch": 0.4, + "grad_norm": 0.478515625, + "learning_rate": 0.00018119168727666901, + "loss": 2.5644, + "step": 9393 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.0001811877711842537, + "loss": 2.4244, + "step": 9394 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.00018118385472652303, + "loss": 2.5026, + "step": 9395 + }, + { + "epoch": 0.4, + "grad_norm": 0.51953125, + "learning_rate": 0.00018117993790349473, + "loss": 2.5486, + "step": 9396 + }, + { + "epoch": 0.4, + "grad_norm": 0.4375, + "learning_rate": 0.00018117602071518634, + "loss": 2.6645, + "step": 9397 + }, + { + "epoch": 0.4, + "grad_norm": 0.408203125, + "learning_rate": 0.00018117210316161554, + "loss": 2.5544, + "step": 9398 + }, + { + "epoch": 0.4, + "grad_norm": 0.466796875, + "learning_rate": 0.00018116818524279994, + "loss": 2.705, + "step": 9399 + }, + { + "epoch": 0.4, + "grad_norm": 0.451171875, + "learning_rate": 0.00018116426695875717, + "loss": 2.6219, + "step": 9400 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.0001811603483095049, + "loss": 2.6537, + "step": 9401 + }, + { + "epoch": 0.4, + "grad_norm": 0.416015625, + "learning_rate": 0.00018115642929506068, + "loss": 2.7084, + "step": 9402 + }, + { + "epoch": 0.4, + "grad_norm": 0.49609375, + "learning_rate": 0.00018115250991544216, + "loss": 2.6624, + "step": 9403 + }, + { + "epoch": 0.4, + "grad_norm": 0.435546875, + "learning_rate": 0.00018114859017066706, + "loss": 2.5558, + "step": 9404 + }, + { + "epoch": 0.4, + "grad_norm": 0.4921875, + "learning_rate": 0.0001811446700607529, + "loss": 2.6525, + "step": 9405 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018114074958571746, + "loss": 2.5393, + "step": 9406 + }, + { + "epoch": 0.4, + "grad_norm": 0.392578125, + "learning_rate": 0.00018113682874557822, + "loss": 2.6289, + "step": 9407 + }, + { + "epoch": 0.4, + "grad_norm": 0.396484375, + "learning_rate": 0.00018113290754035295, + "loss": 2.6948, + "step": 9408 + }, + { + "epoch": 0.4, + "grad_norm": 0.44921875, + "learning_rate": 0.00018112898597005925, + "loss": 2.6647, + "step": 9409 + }, + { + "epoch": 0.4, + "grad_norm": 0.435546875, + "learning_rate": 0.00018112506403471473, + "loss": 2.5971, + "step": 9410 + }, + { + "epoch": 0.4, + "grad_norm": 0.453125, + "learning_rate": 0.00018112114173433707, + "loss": 2.5985, + "step": 9411 + }, + { + "epoch": 0.4, + "grad_norm": 0.40234375, + "learning_rate": 0.00018111721906894396, + "loss": 2.4571, + "step": 9412 + }, + { + "epoch": 0.4, + "grad_norm": 0.435546875, + "learning_rate": 0.000181113296038553, + "loss": 2.5577, + "step": 9413 + }, + { + "epoch": 0.4, + "grad_norm": 0.4296875, + "learning_rate": 0.00018110937264318182, + "loss": 2.6112, + "step": 9414 + }, + { + "epoch": 0.4, + "grad_norm": 0.474609375, + "learning_rate": 0.00018110544888284814, + "loss": 2.7488, + "step": 9415 + }, + { + "epoch": 0.4, + "grad_norm": 0.384765625, + "learning_rate": 0.00018110152475756954, + "loss": 2.6165, + "step": 9416 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018109760026736375, + "loss": 2.725, + "step": 9417 + }, + { + "epoch": 0.4, + "grad_norm": 0.46875, + "learning_rate": 0.0001810936754122484, + "loss": 2.7315, + "step": 9418 + }, + { + "epoch": 0.4, + "grad_norm": 0.44140625, + "learning_rate": 0.0001810897501922411, + "loss": 2.6135, + "step": 9419 + }, + { + "epoch": 0.4, + "grad_norm": 0.4921875, + "learning_rate": 0.0001810858246073596, + "loss": 2.6091, + "step": 9420 + }, + { + "epoch": 0.4, + "grad_norm": 0.48046875, + "learning_rate": 0.00018108189865762153, + "loss": 2.6391, + "step": 9421 + }, + { + "epoch": 0.4, + "grad_norm": 0.453125, + "learning_rate": 0.00018107797234304454, + "loss": 2.4897, + "step": 9422 + }, + { + "epoch": 0.4, + "grad_norm": 0.50390625, + "learning_rate": 0.0001810740456636463, + "loss": 2.6881, + "step": 9423 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.00018107011861944448, + "loss": 2.6812, + "step": 9424 + }, + { + "epoch": 0.4, + "grad_norm": 0.4296875, + "learning_rate": 0.00018106619121045675, + "loss": 2.5854, + "step": 9425 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.0001810622634367008, + "loss": 2.6745, + "step": 9426 + }, + { + "epoch": 0.4, + "grad_norm": 0.455078125, + "learning_rate": 0.00018105833529819427, + "loss": 2.5979, + "step": 9427 + }, + { + "epoch": 0.4, + "grad_norm": 0.455078125, + "learning_rate": 0.0001810544067949549, + "loss": 2.5401, + "step": 9428 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018105047792700029, + "loss": 2.6725, + "step": 9429 + }, + { + "epoch": 0.4, + "grad_norm": 0.52734375, + "learning_rate": 0.00018104654869434812, + "loss": 2.6644, + "step": 9430 + }, + { + "epoch": 0.4, + "grad_norm": 0.45703125, + "learning_rate": 0.00018104261909701613, + "loss": 2.5858, + "step": 9431 + }, + { + "epoch": 0.4, + "grad_norm": 0.50390625, + "learning_rate": 0.00018103868913502198, + "loss": 2.7166, + "step": 9432 + }, + { + "epoch": 0.4, + "grad_norm": 0.423828125, + "learning_rate": 0.00018103475880838332, + "loss": 2.4584, + "step": 9433 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.0001810308281171179, + "loss": 2.6841, + "step": 9434 + }, + { + "epoch": 0.4, + "grad_norm": 0.46875, + "learning_rate": 0.0001810268970612433, + "loss": 2.7329, + "step": 9435 + }, + { + "epoch": 0.4, + "grad_norm": 0.41015625, + "learning_rate": 0.00018102296564077733, + "loss": 2.5769, + "step": 9436 + }, + { + "epoch": 0.4, + "grad_norm": 0.416015625, + "learning_rate": 0.0001810190338557376, + "loss": 2.5324, + "step": 9437 + }, + { + "epoch": 0.4, + "grad_norm": 0.486328125, + "learning_rate": 0.00018101510170614185, + "loss": 2.7782, + "step": 9438 + }, + { + "epoch": 0.4, + "grad_norm": 0.470703125, + "learning_rate": 0.00018101116919200773, + "loss": 2.6271, + "step": 9439 + }, + { + "epoch": 0.4, + "grad_norm": 0.470703125, + "learning_rate": 0.00018100723631335298, + "loss": 2.6487, + "step": 9440 + }, + { + "epoch": 0.4, + "grad_norm": 0.5, + "learning_rate": 0.00018100330307019527, + "loss": 2.4901, + "step": 9441 + }, + { + "epoch": 0.4, + "grad_norm": 0.470703125, + "learning_rate": 0.00018099936946255226, + "loss": 2.6193, + "step": 9442 + }, + { + "epoch": 0.4, + "grad_norm": 0.50390625, + "learning_rate": 0.00018099543549044173, + "loss": 2.6727, + "step": 9443 + }, + { + "epoch": 0.4, + "grad_norm": 0.4296875, + "learning_rate": 0.00018099150115388134, + "loss": 2.6704, + "step": 9444 + }, + { + "epoch": 0.4, + "grad_norm": 0.4921875, + "learning_rate": 0.00018098756645288874, + "loss": 2.6408, + "step": 9445 + }, + { + "epoch": 0.4, + "grad_norm": 0.44921875, + "learning_rate": 0.00018098363138748175, + "loss": 2.6399, + "step": 9446 + }, + { + "epoch": 0.4, + "grad_norm": 0.54296875, + "learning_rate": 0.000180979695957678, + "loss": 2.6418, + "step": 9447 + }, + { + "epoch": 0.4, + "grad_norm": 0.4453125, + "learning_rate": 0.0001809757601634952, + "loss": 2.6401, + "step": 9448 + }, + { + "epoch": 0.4, + "grad_norm": 0.5, + "learning_rate": 0.00018097182400495106, + "loss": 2.7145, + "step": 9449 + }, + { + "epoch": 0.4, + "grad_norm": 0.38671875, + "learning_rate": 0.00018096788748206332, + "loss": 2.5618, + "step": 9450 + }, + { + "epoch": 0.4, + "grad_norm": 0.40625, + "learning_rate": 0.00018096395059484968, + "loss": 2.6305, + "step": 9451 + }, + { + "epoch": 0.4, + "grad_norm": 0.455078125, + "learning_rate": 0.00018096001334332786, + "loss": 2.6279, + "step": 9452 + }, + { + "epoch": 0.4, + "grad_norm": 0.419921875, + "learning_rate": 0.00018095607572751556, + "loss": 2.7389, + "step": 9453 + }, + { + "epoch": 0.4, + "grad_norm": 0.48828125, + "learning_rate": 0.0001809521377474305, + "loss": 2.7053, + "step": 9454 + }, + { + "epoch": 0.4, + "grad_norm": 0.453125, + "learning_rate": 0.0001809481994030904, + "loss": 2.6544, + "step": 9455 + }, + { + "epoch": 0.4, + "grad_norm": 0.470703125, + "learning_rate": 0.000180944260694513, + "loss": 2.6121, + "step": 9456 + }, + { + "epoch": 0.4, + "grad_norm": 0.466796875, + "learning_rate": 0.00018094032162171598, + "loss": 2.6937, + "step": 9457 + }, + { + "epoch": 0.4, + "grad_norm": 0.458984375, + "learning_rate": 0.00018093638218471715, + "loss": 2.7448, + "step": 9458 + }, + { + "epoch": 0.4, + "grad_norm": 0.4140625, + "learning_rate": 0.00018093244238353412, + "loss": 2.609, + "step": 9459 + }, + { + "epoch": 0.4, + "grad_norm": 0.4609375, + "learning_rate": 0.00018092850221818472, + "loss": 2.6749, + "step": 9460 + }, + { + "epoch": 0.4, + "grad_norm": 0.478515625, + "learning_rate": 0.0001809245616886866, + "loss": 2.6084, + "step": 9461 + }, + { + "epoch": 0.4, + "grad_norm": 0.474609375, + "learning_rate": 0.00018092062079505753, + "loss": 2.698, + "step": 9462 + }, + { + "epoch": 0.4, + "grad_norm": 0.48046875, + "learning_rate": 0.00018091667953731527, + "loss": 2.4675, + "step": 9463 + }, + { + "epoch": 0.4, + "grad_norm": 0.49609375, + "learning_rate": 0.00018091273791547748, + "loss": 2.6247, + "step": 9464 + }, + { + "epoch": 0.4, + "grad_norm": 0.484375, + "learning_rate": 0.00018090879592956197, + "loss": 2.6886, + "step": 9465 + }, + { + "epoch": 0.4, + "grad_norm": 0.4453125, + "learning_rate": 0.00018090485357958648, + "loss": 2.5328, + "step": 9466 + }, + { + "epoch": 0.4, + "grad_norm": 0.46875, + "learning_rate": 0.00018090091086556866, + "loss": 2.7586, + "step": 9467 + }, + { + "epoch": 0.4, + "grad_norm": 0.45703125, + "learning_rate": 0.00018089696778752633, + "loss": 2.3828, + "step": 9468 + }, + { + "epoch": 0.4, + "grad_norm": 0.43359375, + "learning_rate": 0.00018089302434547722, + "loss": 2.5462, + "step": 9469 + }, + { + "epoch": 0.4, + "grad_norm": 0.421875, + "learning_rate": 0.00018088908053943902, + "loss": 2.5728, + "step": 9470 + }, + { + "epoch": 0.4, + "grad_norm": 0.439453125, + "learning_rate": 0.00018088513636942954, + "loss": 2.5885, + "step": 9471 + }, + { + "epoch": 0.4, + "grad_norm": 0.421875, + "learning_rate": 0.00018088119183546652, + "loss": 2.5324, + "step": 9472 + }, + { + "epoch": 0.4, + "grad_norm": 0.52734375, + "learning_rate": 0.0001808772469375677, + "loss": 2.6656, + "step": 9473 + }, + { + "epoch": 0.4, + "grad_norm": 0.421875, + "learning_rate": 0.0001808733016757508, + "loss": 2.5663, + "step": 9474 + }, + { + "epoch": 0.4, + "grad_norm": 0.42578125, + "learning_rate": 0.00018086935605003365, + "loss": 2.6657, + "step": 9475 + }, + { + "epoch": 0.4, + "grad_norm": 0.453125, + "learning_rate": 0.0001808654100604339, + "loss": 2.6715, + "step": 9476 + }, + { + "epoch": 0.4, + "grad_norm": 0.412109375, + "learning_rate": 0.0001808614637069694, + "loss": 2.5461, + "step": 9477 + }, + { + "epoch": 0.4, + "grad_norm": 0.392578125, + "learning_rate": 0.0001808575169896578, + "loss": 2.5225, + "step": 9478 + }, + { + "epoch": 0.4, + "grad_norm": 0.470703125, + "learning_rate": 0.000180853569908517, + "loss": 2.6997, + "step": 9479 + }, + { + "epoch": 0.4, + "grad_norm": 0.484375, + "learning_rate": 0.00018084962246356462, + "loss": 2.6763, + "step": 9480 + }, + { + "epoch": 0.4, + "grad_norm": 0.43359375, + "learning_rate": 0.0001808456746548185, + "loss": 2.5231, + "step": 9481 + }, + { + "epoch": 0.4, + "grad_norm": 0.435546875, + "learning_rate": 0.0001808417264822964, + "loss": 2.6589, + "step": 9482 + }, + { + "epoch": 0.4, + "grad_norm": 0.482421875, + "learning_rate": 0.0001808377779460161, + "loss": 2.4651, + "step": 9483 + }, + { + "epoch": 0.4, + "grad_norm": 0.4375, + "learning_rate": 0.0001808338290459953, + "loss": 2.6053, + "step": 9484 + }, + { + "epoch": 0.4, + "grad_norm": 0.431640625, + "learning_rate": 0.00018082987978225185, + "loss": 2.6386, + "step": 9485 + }, + { + "epoch": 0.4, + "grad_norm": 0.431640625, + "learning_rate": 0.00018082593015480346, + "loss": 2.7121, + "step": 9486 + }, + { + "epoch": 0.4, + "grad_norm": 0.47265625, + "learning_rate": 0.0001808219801636679, + "loss": 2.4945, + "step": 9487 + }, + { + "epoch": 0.4, + "grad_norm": 0.431640625, + "learning_rate": 0.000180818029808863, + "loss": 2.5287, + "step": 9488 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.0001808140790904065, + "loss": 2.6322, + "step": 9489 + }, + { + "epoch": 0.4, + "grad_norm": 0.46484375, + "learning_rate": 0.00018081012800831619, + "loss": 2.6809, + "step": 9490 + }, + { + "epoch": 0.4, + "grad_norm": 0.443359375, + "learning_rate": 0.00018080617656260982, + "loss": 2.6035, + "step": 9491 + }, + { + "epoch": 0.4, + "grad_norm": 0.455078125, + "learning_rate": 0.00018080222475330517, + "loss": 2.5779, + "step": 9492 + }, + { + "epoch": 0.4, + "grad_norm": 0.484375, + "learning_rate": 0.00018079827258042004, + "loss": 2.5729, + "step": 9493 + }, + { + "epoch": 0.4, + "grad_norm": 0.484375, + "learning_rate": 0.00018079432004397222, + "loss": 2.7107, + "step": 9494 + }, + { + "epoch": 0.4, + "grad_norm": 0.498046875, + "learning_rate": 0.0001807903671439795, + "loss": 2.541, + "step": 9495 + }, + { + "epoch": 0.4, + "grad_norm": 0.474609375, + "learning_rate": 0.00018078641388045964, + "loss": 2.5923, + "step": 9496 + }, + { + "epoch": 0.4, + "grad_norm": 0.40234375, + "learning_rate": 0.00018078246025343044, + "loss": 2.7285, + "step": 9497 + }, + { + "epoch": 0.4, + "grad_norm": 0.46875, + "learning_rate": 0.0001807785062629097, + "loss": 2.69, + "step": 9498 + }, + { + "epoch": 0.4, + "grad_norm": 0.435546875, + "learning_rate": 0.0001807745519089152, + "loss": 2.5222, + "step": 9499 + }, + { + "epoch": 0.4, + "grad_norm": 0.416015625, + "learning_rate": 0.00018077059719146473, + "loss": 2.6455, + "step": 9500 + }, + { + "epoch": 0.4, + "grad_norm": 0.46875, + "learning_rate": 0.00018076664211057612, + "loss": 2.6787, + "step": 9501 + }, + { + "epoch": 0.4, + "grad_norm": 0.451171875, + "learning_rate": 0.0001807626866662671, + "loss": 2.7201, + "step": 9502 + }, + { + "epoch": 0.4, + "grad_norm": 0.470703125, + "learning_rate": 0.00018075873085855553, + "loss": 2.5801, + "step": 9503 + }, + { + "epoch": 0.4, + "grad_norm": 0.41796875, + "learning_rate": 0.00018075477468745916, + "loss": 2.6252, + "step": 9504 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.00018075081815299585, + "loss": 2.6402, + "step": 9505 + }, + { + "epoch": 0.41, + "grad_norm": 0.44921875, + "learning_rate": 0.00018074686125518335, + "loss": 2.6332, + "step": 9506 + }, + { + "epoch": 0.41, + "grad_norm": 0.42578125, + "learning_rate": 0.00018074290399403947, + "loss": 2.5254, + "step": 9507 + }, + { + "epoch": 0.41, + "grad_norm": 0.388671875, + "learning_rate": 0.00018073894636958207, + "loss": 2.5945, + "step": 9508 + }, + { + "epoch": 0.41, + "grad_norm": 0.453125, + "learning_rate": 0.00018073498838182888, + "loss": 2.7186, + "step": 9509 + }, + { + "epoch": 0.41, + "grad_norm": 0.4921875, + "learning_rate": 0.00018073103003079776, + "loss": 2.6507, + "step": 9510 + }, + { + "epoch": 0.41, + "grad_norm": 0.431640625, + "learning_rate": 0.0001807270713165065, + "loss": 2.5797, + "step": 9511 + }, + { + "epoch": 0.41, + "grad_norm": 0.4375, + "learning_rate": 0.00018072311223897293, + "loss": 2.6856, + "step": 9512 + }, + { + "epoch": 0.41, + "grad_norm": 0.474609375, + "learning_rate": 0.00018071915279821487, + "loss": 2.5553, + "step": 9513 + }, + { + "epoch": 0.41, + "grad_norm": 0.439453125, + "learning_rate": 0.0001807151929942501, + "loss": 2.5193, + "step": 9514 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018071123282709648, + "loss": 2.4935, + "step": 9515 + }, + { + "epoch": 0.41, + "grad_norm": 0.53125, + "learning_rate": 0.00018070727229677178, + "loss": 2.8536, + "step": 9516 + }, + { + "epoch": 0.41, + "grad_norm": 0.4765625, + "learning_rate": 0.00018070331140329385, + "loss": 2.5398, + "step": 9517 + }, + { + "epoch": 0.41, + "grad_norm": 0.47265625, + "learning_rate": 0.00018069935014668054, + "loss": 2.7448, + "step": 9518 + }, + { + "epoch": 0.41, + "grad_norm": 0.48046875, + "learning_rate": 0.00018069538852694966, + "loss": 2.724, + "step": 9519 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.000180691426544119, + "loss": 2.7242, + "step": 9520 + }, + { + "epoch": 0.41, + "grad_norm": 0.419921875, + "learning_rate": 0.00018068746419820636, + "loss": 2.5292, + "step": 9521 + }, + { + "epoch": 0.41, + "grad_norm": 0.466796875, + "learning_rate": 0.00018068350148922968, + "loss": 2.4978, + "step": 9522 + }, + { + "epoch": 0.41, + "grad_norm": 0.45703125, + "learning_rate": 0.00018067953841720673, + "loss": 2.6358, + "step": 9523 + }, + { + "epoch": 0.41, + "grad_norm": 0.490234375, + "learning_rate": 0.00018067557498215533, + "loss": 2.5878, + "step": 9524 + }, + { + "epoch": 0.41, + "grad_norm": 0.423828125, + "learning_rate": 0.00018067161118409333, + "loss": 2.6449, + "step": 9525 + }, + { + "epoch": 0.41, + "grad_norm": 0.44921875, + "learning_rate": 0.00018066764702303852, + "loss": 2.5175, + "step": 9526 + }, + { + "epoch": 0.41, + "grad_norm": 0.482421875, + "learning_rate": 0.00018066368249900885, + "loss": 2.5573, + "step": 9527 + }, + { + "epoch": 0.41, + "grad_norm": 0.427734375, + "learning_rate": 0.00018065971761202206, + "loss": 2.6788, + "step": 9528 + }, + { + "epoch": 0.41, + "grad_norm": 0.423828125, + "learning_rate": 0.000180655752362096, + "loss": 2.5536, + "step": 9529 + }, + { + "epoch": 0.41, + "grad_norm": 0.435546875, + "learning_rate": 0.00018065178674924853, + "loss": 2.601, + "step": 9530 + }, + { + "epoch": 0.41, + "grad_norm": 0.53125, + "learning_rate": 0.00018064782077349754, + "loss": 2.6807, + "step": 9531 + }, + { + "epoch": 0.41, + "grad_norm": 0.490234375, + "learning_rate": 0.00018064385443486075, + "loss": 2.7944, + "step": 9532 + }, + { + "epoch": 0.41, + "grad_norm": 0.490234375, + "learning_rate": 0.00018063988773335616, + "loss": 2.6481, + "step": 9533 + }, + { + "epoch": 0.41, + "grad_norm": 0.474609375, + "learning_rate": 0.0001806359206690015, + "loss": 2.5935, + "step": 9534 + }, + { + "epoch": 0.41, + "grad_norm": 0.4453125, + "learning_rate": 0.0001806319532418147, + "loss": 2.749, + "step": 9535 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.00018062798545181357, + "loss": 2.6073, + "step": 9536 + }, + { + "epoch": 0.41, + "grad_norm": 0.3984375, + "learning_rate": 0.00018062401729901595, + "loss": 2.6032, + "step": 9537 + }, + { + "epoch": 0.41, + "grad_norm": 0.48046875, + "learning_rate": 0.00018062004878343973, + "loss": 2.6017, + "step": 9538 + }, + { + "epoch": 0.41, + "grad_norm": 0.4375, + "learning_rate": 0.00018061607990510278, + "loss": 2.678, + "step": 9539 + }, + { + "epoch": 0.41, + "grad_norm": 0.48046875, + "learning_rate": 0.0001806121106640229, + "loss": 2.7335, + "step": 9540 + }, + { + "epoch": 0.41, + "grad_norm": 0.5, + "learning_rate": 0.00018060814106021797, + "loss": 2.833, + "step": 9541 + }, + { + "epoch": 0.41, + "grad_norm": 0.408203125, + "learning_rate": 0.00018060417109370592, + "loss": 2.7581, + "step": 9542 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.0001806002007645045, + "loss": 2.5914, + "step": 9543 + }, + { + "epoch": 0.41, + "grad_norm": 0.431640625, + "learning_rate": 0.00018059623007263164, + "loss": 2.6354, + "step": 9544 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.00018059225901810523, + "loss": 2.4857, + "step": 9545 + }, + { + "epoch": 0.41, + "grad_norm": 0.427734375, + "learning_rate": 0.0001805882876009431, + "loss": 2.5482, + "step": 9546 + }, + { + "epoch": 0.41, + "grad_norm": 0.423828125, + "learning_rate": 0.00018058431582116313, + "loss": 2.6185, + "step": 9547 + }, + { + "epoch": 0.41, + "grad_norm": 0.443359375, + "learning_rate": 0.00018058034367878317, + "loss": 2.596, + "step": 9548 + }, + { + "epoch": 0.41, + "grad_norm": 0.5, + "learning_rate": 0.00018057637117382112, + "loss": 2.6921, + "step": 9549 + }, + { + "epoch": 0.41, + "grad_norm": 0.5, + "learning_rate": 0.00018057239830629484, + "loss": 2.6762, + "step": 9550 + }, + { + "epoch": 0.41, + "grad_norm": 0.470703125, + "learning_rate": 0.0001805684250762222, + "loss": 2.7446, + "step": 9551 + }, + { + "epoch": 0.41, + "grad_norm": 0.423828125, + "learning_rate": 0.00018056445148362113, + "loss": 2.5607, + "step": 9552 + }, + { + "epoch": 0.41, + "grad_norm": 0.46484375, + "learning_rate": 0.00018056047752850946, + "loss": 2.5446, + "step": 9553 + }, + { + "epoch": 0.41, + "grad_norm": 0.42578125, + "learning_rate": 0.00018055650321090506, + "loss": 2.5816, + "step": 9554 + }, + { + "epoch": 0.41, + "grad_norm": 0.466796875, + "learning_rate": 0.00018055252853082588, + "loss": 2.5897, + "step": 9555 + }, + { + "epoch": 0.41, + "grad_norm": 0.40625, + "learning_rate": 0.00018054855348828973, + "loss": 2.4356, + "step": 9556 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.00018054457808331452, + "loss": 2.6816, + "step": 9557 + }, + { + "epoch": 0.41, + "grad_norm": 0.47265625, + "learning_rate": 0.00018054060231591815, + "loss": 2.6905, + "step": 9558 + }, + { + "epoch": 0.41, + "grad_norm": 0.41796875, + "learning_rate": 0.0001805366261861185, + "loss": 2.6093, + "step": 9559 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018053264969393346, + "loss": 2.5144, + "step": 9560 + }, + { + "epoch": 0.41, + "grad_norm": 0.51953125, + "learning_rate": 0.00018052867283938095, + "loss": 2.4845, + "step": 9561 + }, + { + "epoch": 0.41, + "grad_norm": 0.49609375, + "learning_rate": 0.00018052469562247883, + "loss": 2.6605, + "step": 9562 + }, + { + "epoch": 0.41, + "grad_norm": 0.48828125, + "learning_rate": 0.00018052071804324502, + "loss": 2.6866, + "step": 9563 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018051674010169738, + "loss": 2.6586, + "step": 9564 + }, + { + "epoch": 0.41, + "grad_norm": 0.486328125, + "learning_rate": 0.00018051276179785385, + "loss": 2.6678, + "step": 9565 + }, + { + "epoch": 0.41, + "grad_norm": 0.412109375, + "learning_rate": 0.00018050878313173233, + "loss": 2.5776, + "step": 9566 + }, + { + "epoch": 0.41, + "grad_norm": 0.490234375, + "learning_rate": 0.0001805048041033507, + "loss": 2.5813, + "step": 9567 + }, + { + "epoch": 0.41, + "grad_norm": 0.421875, + "learning_rate": 0.00018050082471272684, + "loss": 2.6835, + "step": 9568 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.00018049684495987872, + "loss": 2.5615, + "step": 9569 + }, + { + "epoch": 0.41, + "grad_norm": 0.482421875, + "learning_rate": 0.0001804928648448242, + "loss": 2.6818, + "step": 9570 + }, + { + "epoch": 0.41, + "grad_norm": 0.392578125, + "learning_rate": 0.00018048888436758125, + "loss": 2.587, + "step": 9571 + }, + { + "epoch": 0.41, + "grad_norm": 0.48828125, + "learning_rate": 0.0001804849035281677, + "loss": 2.5284, + "step": 9572 + }, + { + "epoch": 0.41, + "grad_norm": 0.443359375, + "learning_rate": 0.00018048092232660147, + "loss": 2.6316, + "step": 9573 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.00018047694076290055, + "loss": 2.5992, + "step": 9574 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.0001804729588370828, + "loss": 2.7021, + "step": 9575 + }, + { + "epoch": 0.41, + "grad_norm": 0.443359375, + "learning_rate": 0.00018046897654916614, + "loss": 2.5837, + "step": 9576 + }, + { + "epoch": 0.41, + "grad_norm": 0.447265625, + "learning_rate": 0.00018046499389916848, + "loss": 2.5937, + "step": 9577 + }, + { + "epoch": 0.41, + "grad_norm": 0.431640625, + "learning_rate": 0.00018046101088710774, + "loss": 2.7175, + "step": 9578 + }, + { + "epoch": 0.41, + "grad_norm": 0.462890625, + "learning_rate": 0.00018045702751300188, + "loss": 2.5093, + "step": 9579 + }, + { + "epoch": 0.41, + "grad_norm": 0.408203125, + "learning_rate": 0.00018045304377686884, + "loss": 2.5291, + "step": 9580 + }, + { + "epoch": 0.41, + "grad_norm": 0.4453125, + "learning_rate": 0.00018044905967872643, + "loss": 2.597, + "step": 9581 + }, + { + "epoch": 0.41, + "grad_norm": 0.494140625, + "learning_rate": 0.0001804450752185927, + "loss": 2.5793, + "step": 9582 + }, + { + "epoch": 0.41, + "grad_norm": 0.4765625, + "learning_rate": 0.00018044109039648548, + "loss": 2.6098, + "step": 9583 + }, + { + "epoch": 0.41, + "grad_norm": 0.47265625, + "learning_rate": 0.00018043710521242282, + "loss": 2.5488, + "step": 9584 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.00018043311966642252, + "loss": 2.5697, + "step": 9585 + }, + { + "epoch": 0.41, + "grad_norm": 0.470703125, + "learning_rate": 0.00018042913375850264, + "loss": 2.649, + "step": 9586 + }, + { + "epoch": 0.41, + "grad_norm": 0.51171875, + "learning_rate": 0.000180425147488681, + "loss": 2.7459, + "step": 9587 + }, + { + "epoch": 0.41, + "grad_norm": 0.453125, + "learning_rate": 0.0001804211608569756, + "loss": 2.6401, + "step": 9588 + }, + { + "epoch": 0.41, + "grad_norm": 0.44921875, + "learning_rate": 0.0001804171738634044, + "loss": 2.5877, + "step": 9589 + }, + { + "epoch": 0.41, + "grad_norm": 0.4375, + "learning_rate": 0.00018041318650798525, + "loss": 2.5835, + "step": 9590 + }, + { + "epoch": 0.41, + "grad_norm": 0.447265625, + "learning_rate": 0.00018040919879073618, + "loss": 2.6189, + "step": 9591 + }, + { + "epoch": 0.41, + "grad_norm": 0.421875, + "learning_rate": 0.00018040521071167512, + "loss": 2.5923, + "step": 9592 + }, + { + "epoch": 0.41, + "grad_norm": 0.443359375, + "learning_rate": 0.00018040122227081998, + "loss": 2.521, + "step": 9593 + }, + { + "epoch": 0.41, + "grad_norm": 0.474609375, + "learning_rate": 0.0001803972334681887, + "loss": 2.7257, + "step": 9594 + }, + { + "epoch": 0.41, + "grad_norm": 0.431640625, + "learning_rate": 0.00018039324430379927, + "loss": 2.5472, + "step": 9595 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.00018038925477766965, + "loss": 2.6521, + "step": 9596 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.0001803852648898177, + "loss": 2.6188, + "step": 9597 + }, + { + "epoch": 0.41, + "grad_norm": 0.466796875, + "learning_rate": 0.00018038127464026148, + "loss": 2.5719, + "step": 9598 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.0001803772840290189, + "loss": 2.7162, + "step": 9599 + }, + { + "epoch": 0.41, + "grad_norm": 0.486328125, + "learning_rate": 0.0001803732930561079, + "loss": 2.6142, + "step": 9600 + }, + { + "epoch": 0.41, + "grad_norm": 0.5, + "learning_rate": 0.00018036930172154648, + "loss": 2.6364, + "step": 9601 + }, + { + "epoch": 0.41, + "grad_norm": 0.4453125, + "learning_rate": 0.00018036531002535254, + "loss": 2.7042, + "step": 9602 + }, + { + "epoch": 0.41, + "grad_norm": 0.490234375, + "learning_rate": 0.00018036131796754406, + "loss": 2.7549, + "step": 9603 + }, + { + "epoch": 0.41, + "grad_norm": 0.458984375, + "learning_rate": 0.00018035732554813909, + "loss": 2.5675, + "step": 9604 + }, + { + "epoch": 0.41, + "grad_norm": 0.42578125, + "learning_rate": 0.00018035333276715548, + "loss": 2.5963, + "step": 9605 + }, + { + "epoch": 0.41, + "grad_norm": 0.54296875, + "learning_rate": 0.00018034933962461123, + "loss": 2.7917, + "step": 9606 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018034534612052433, + "loss": 2.5643, + "step": 9607 + }, + { + "epoch": 0.41, + "grad_norm": 0.44921875, + "learning_rate": 0.0001803413522549127, + "loss": 2.602, + "step": 9608 + }, + { + "epoch": 0.41, + "grad_norm": 0.447265625, + "learning_rate": 0.0001803373580277944, + "loss": 2.6128, + "step": 9609 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018033336343918732, + "loss": 2.5512, + "step": 9610 + }, + { + "epoch": 0.41, + "grad_norm": 0.3984375, + "learning_rate": 0.00018032936848910947, + "loss": 2.7152, + "step": 9611 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.0001803253731775788, + "loss": 2.6139, + "step": 9612 + }, + { + "epoch": 0.41, + "grad_norm": 0.51171875, + "learning_rate": 0.00018032137750461334, + "loss": 2.6505, + "step": 9613 + }, + { + "epoch": 0.41, + "grad_norm": 0.4375, + "learning_rate": 0.00018031738147023102, + "loss": 2.7329, + "step": 9614 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018031338507444985, + "loss": 2.5317, + "step": 9615 + }, + { + "epoch": 0.41, + "grad_norm": 0.443359375, + "learning_rate": 0.00018030938831728777, + "loss": 2.5333, + "step": 9616 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.0001803053911987628, + "loss": 2.8287, + "step": 9617 + }, + { + "epoch": 0.41, + "grad_norm": 0.47265625, + "learning_rate": 0.00018030139371889292, + "loss": 2.5392, + "step": 9618 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018029739587769613, + "loss": 2.6923, + "step": 9619 + }, + { + "epoch": 0.41, + "grad_norm": 0.4609375, + "learning_rate": 0.00018029339767519037, + "loss": 2.6062, + "step": 9620 + }, + { + "epoch": 0.41, + "grad_norm": 0.4609375, + "learning_rate": 0.00018028939911139368, + "loss": 2.7256, + "step": 9621 + }, + { + "epoch": 0.41, + "grad_norm": 0.48828125, + "learning_rate": 0.00018028540018632405, + "loss": 2.6956, + "step": 9622 + }, + { + "epoch": 0.41, + "grad_norm": 0.45703125, + "learning_rate": 0.00018028140089999944, + "loss": 2.5659, + "step": 9623 + }, + { + "epoch": 0.41, + "grad_norm": 0.421875, + "learning_rate": 0.0001802774012524379, + "loss": 2.5242, + "step": 9624 + }, + { + "epoch": 0.41, + "grad_norm": 0.462890625, + "learning_rate": 0.00018027340124365736, + "loss": 2.5597, + "step": 9625 + }, + { + "epoch": 0.41, + "grad_norm": 0.40625, + "learning_rate": 0.00018026940087367588, + "loss": 2.6788, + "step": 9626 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.0001802654001425114, + "loss": 2.5627, + "step": 9627 + }, + { + "epoch": 0.41, + "grad_norm": 0.474609375, + "learning_rate": 0.00018026139905018195, + "loss": 2.6164, + "step": 9628 + }, + { + "epoch": 0.41, + "grad_norm": 0.458984375, + "learning_rate": 0.00018025739759670555, + "loss": 2.7629, + "step": 9629 + }, + { + "epoch": 0.41, + "grad_norm": 0.44140625, + "learning_rate": 0.00018025339578210022, + "loss": 2.5719, + "step": 9630 + }, + { + "epoch": 0.41, + "grad_norm": 0.48046875, + "learning_rate": 0.0001802493936063839, + "loss": 2.6538, + "step": 9631 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.00018024539106957464, + "loss": 2.5844, + "step": 9632 + }, + { + "epoch": 0.41, + "grad_norm": 0.50390625, + "learning_rate": 0.00018024138817169044, + "loss": 2.6193, + "step": 9633 + }, + { + "epoch": 0.41, + "grad_norm": 0.44140625, + "learning_rate": 0.00018023738491274935, + "loss": 2.5093, + "step": 9634 + }, + { + "epoch": 0.41, + "grad_norm": 0.51953125, + "learning_rate": 0.00018023338129276932, + "loss": 2.6978, + "step": 9635 + }, + { + "epoch": 0.41, + "grad_norm": 0.412109375, + "learning_rate": 0.00018022937731176843, + "loss": 2.4464, + "step": 9636 + }, + { + "epoch": 0.41, + "grad_norm": 0.431640625, + "learning_rate": 0.00018022537296976464, + "loss": 2.6267, + "step": 9637 + }, + { + "epoch": 0.41, + "grad_norm": 0.484375, + "learning_rate": 0.00018022136826677598, + "loss": 2.5357, + "step": 9638 + }, + { + "epoch": 0.41, + "grad_norm": 0.41015625, + "learning_rate": 0.00018021736320282053, + "loss": 2.6248, + "step": 9639 + }, + { + "epoch": 0.41, + "grad_norm": 0.439453125, + "learning_rate": 0.00018021335777791618, + "loss": 2.5697, + "step": 9640 + }, + { + "epoch": 0.41, + "grad_norm": 0.427734375, + "learning_rate": 0.00018020935199208112, + "loss": 2.6229, + "step": 9641 + }, + { + "epoch": 0.41, + "grad_norm": 0.419921875, + "learning_rate": 0.0001802053458453332, + "loss": 2.6604, + "step": 9642 + }, + { + "epoch": 0.41, + "grad_norm": 0.47265625, + "learning_rate": 0.0001802013393376906, + "loss": 2.6405, + "step": 9643 + }, + { + "epoch": 0.41, + "grad_norm": 0.4453125, + "learning_rate": 0.00018019733246917128, + "loss": 2.6618, + "step": 9644 + }, + { + "epoch": 0.41, + "grad_norm": 0.48828125, + "learning_rate": 0.00018019332523979326, + "loss": 2.7231, + "step": 9645 + }, + { + "epoch": 0.41, + "grad_norm": 0.40234375, + "learning_rate": 0.0001801893176495746, + "loss": 2.5961, + "step": 9646 + }, + { + "epoch": 0.41, + "grad_norm": 0.435546875, + "learning_rate": 0.00018018530969853328, + "loss": 2.6083, + "step": 9647 + }, + { + "epoch": 0.41, + "grad_norm": 0.4296875, + "learning_rate": 0.0001801813013866874, + "loss": 2.5603, + "step": 9648 + }, + { + "epoch": 0.41, + "grad_norm": 0.51171875, + "learning_rate": 0.00018017729271405493, + "loss": 2.5896, + "step": 9649 + }, + { + "epoch": 0.41, + "grad_norm": 0.435546875, + "learning_rate": 0.000180173283680654, + "loss": 2.5883, + "step": 9650 + }, + { + "epoch": 0.41, + "grad_norm": 0.46484375, + "learning_rate": 0.00018016927428650257, + "loss": 2.6562, + "step": 9651 + }, + { + "epoch": 0.41, + "grad_norm": 0.453125, + "learning_rate": 0.0001801652645316187, + "loss": 2.6752, + "step": 9652 + }, + { + "epoch": 0.41, + "grad_norm": 0.4609375, + "learning_rate": 0.00018016125441602042, + "loss": 2.6949, + "step": 9653 + }, + { + "epoch": 0.41, + "grad_norm": 0.453125, + "learning_rate": 0.00018015724393972583, + "loss": 2.7159, + "step": 9654 + }, + { + "epoch": 0.41, + "grad_norm": 0.421875, + "learning_rate": 0.0001801532331027529, + "loss": 2.5509, + "step": 9655 + }, + { + "epoch": 0.41, + "grad_norm": 0.431640625, + "learning_rate": 0.00018014922190511974, + "loss": 2.7333, + "step": 9656 + }, + { + "epoch": 0.41, + "grad_norm": 0.50390625, + "learning_rate": 0.00018014521034684434, + "loss": 2.6085, + "step": 9657 + }, + { + "epoch": 0.41, + "grad_norm": 0.51171875, + "learning_rate": 0.00018014119842794482, + "loss": 2.7821, + "step": 9658 + }, + { + "epoch": 0.41, + "grad_norm": 0.40625, + "learning_rate": 0.00018013718614843917, + "loss": 2.473, + "step": 9659 + }, + { + "epoch": 0.41, + "grad_norm": 0.419921875, + "learning_rate": 0.00018013317350834549, + "loss": 2.601, + "step": 9660 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018012916050768178, + "loss": 2.5783, + "step": 9661 + }, + { + "epoch": 0.41, + "grad_norm": 0.5, + "learning_rate": 0.00018012514714646618, + "loss": 2.5633, + "step": 9662 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018012113342471666, + "loss": 2.5717, + "step": 9663 + }, + { + "epoch": 0.41, + "grad_norm": 0.44140625, + "learning_rate": 0.00018011711934245132, + "loss": 2.4683, + "step": 9664 + }, + { + "epoch": 0.41, + "grad_norm": 0.43359375, + "learning_rate": 0.00018011310489968825, + "loss": 2.467, + "step": 9665 + }, + { + "epoch": 0.41, + "grad_norm": 0.44921875, + "learning_rate": 0.00018010909009644546, + "loss": 2.6466, + "step": 9666 + }, + { + "epoch": 0.41, + "grad_norm": 0.416015625, + "learning_rate": 0.00018010507493274104, + "loss": 2.5477, + "step": 9667 + }, + { + "epoch": 0.41, + "grad_norm": 0.4765625, + "learning_rate": 0.0001801010594085931, + "loss": 2.5555, + "step": 9668 + }, + { + "epoch": 0.41, + "grad_norm": 0.3984375, + "learning_rate": 0.00018009704352401964, + "loss": 2.4975, + "step": 9669 + }, + { + "epoch": 0.41, + "grad_norm": 0.416015625, + "learning_rate": 0.00018009302727903874, + "loss": 2.6974, + "step": 9670 + }, + { + "epoch": 0.41, + "grad_norm": 0.439453125, + "learning_rate": 0.0001800890106736685, + "loss": 2.6717, + "step": 9671 + }, + { + "epoch": 0.41, + "grad_norm": 0.38671875, + "learning_rate": 0.00018008499370792698, + "loss": 2.5687, + "step": 9672 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018008097638183226, + "loss": 2.6473, + "step": 9673 + }, + { + "epoch": 0.41, + "grad_norm": 0.486328125, + "learning_rate": 0.00018007695869540243, + "loss": 2.5858, + "step": 9674 + }, + { + "epoch": 0.41, + "grad_norm": 0.453125, + "learning_rate": 0.00018007294064865552, + "loss": 2.6104, + "step": 9675 + }, + { + "epoch": 0.41, + "grad_norm": 0.447265625, + "learning_rate": 0.00018006892224160966, + "loss": 2.6564, + "step": 9676 + }, + { + "epoch": 0.41, + "grad_norm": 0.42578125, + "learning_rate": 0.0001800649034742829, + "loss": 2.42, + "step": 9677 + }, + { + "epoch": 0.41, + "grad_norm": 0.466796875, + "learning_rate": 0.00018006088434669335, + "loss": 2.6918, + "step": 9678 + }, + { + "epoch": 0.41, + "grad_norm": 0.453125, + "learning_rate": 0.00018005686485885906, + "loss": 2.6543, + "step": 9679 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.00018005284501079817, + "loss": 2.7196, + "step": 9680 + }, + { + "epoch": 0.41, + "grad_norm": 0.439453125, + "learning_rate": 0.00018004882480252867, + "loss": 2.5064, + "step": 9681 + }, + { + "epoch": 0.41, + "grad_norm": 0.458984375, + "learning_rate": 0.00018004480423406877, + "loss": 2.7156, + "step": 9682 + }, + { + "epoch": 0.41, + "grad_norm": 0.474609375, + "learning_rate": 0.00018004078330543648, + "loss": 2.662, + "step": 9683 + }, + { + "epoch": 0.41, + "grad_norm": 0.486328125, + "learning_rate": 0.00018003676201664993, + "loss": 2.6418, + "step": 9684 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.0001800327403677272, + "loss": 2.5458, + "step": 9685 + }, + { + "epoch": 0.41, + "grad_norm": 0.462890625, + "learning_rate": 0.00018002871835868636, + "loss": 2.4491, + "step": 9686 + }, + { + "epoch": 0.41, + "grad_norm": 0.498046875, + "learning_rate": 0.00018002469598954555, + "loss": 2.7387, + "step": 9687 + }, + { + "epoch": 0.41, + "grad_norm": 0.4765625, + "learning_rate": 0.00018002067326032285, + "loss": 2.5618, + "step": 9688 + }, + { + "epoch": 0.41, + "grad_norm": 0.494140625, + "learning_rate": 0.00018001665017103639, + "loss": 2.6322, + "step": 9689 + }, + { + "epoch": 0.41, + "grad_norm": 0.4765625, + "learning_rate": 0.00018001262672170422, + "loss": 2.6377, + "step": 9690 + }, + { + "epoch": 0.41, + "grad_norm": 0.54296875, + "learning_rate": 0.00018000860291234448, + "loss": 2.7203, + "step": 9691 + }, + { + "epoch": 0.41, + "grad_norm": 0.494140625, + "learning_rate": 0.00018000457874297528, + "loss": 2.7506, + "step": 9692 + }, + { + "epoch": 0.41, + "grad_norm": 0.4453125, + "learning_rate": 0.00018000055421361468, + "loss": 2.6398, + "step": 9693 + }, + { + "epoch": 0.41, + "grad_norm": 0.44921875, + "learning_rate": 0.00017999652932428086, + "loss": 2.6034, + "step": 9694 + }, + { + "epoch": 0.41, + "grad_norm": 0.47265625, + "learning_rate": 0.00017999250407499186, + "loss": 2.5704, + "step": 9695 + }, + { + "epoch": 0.41, + "grad_norm": 0.462890625, + "learning_rate": 0.00017998847846576585, + "loss": 2.4758, + "step": 9696 + }, + { + "epoch": 0.41, + "grad_norm": 0.46875, + "learning_rate": 0.0001799844524966209, + "loss": 2.6194, + "step": 9697 + }, + { + "epoch": 0.41, + "grad_norm": 0.435546875, + "learning_rate": 0.00017998042616757518, + "loss": 2.6022, + "step": 9698 + }, + { + "epoch": 0.41, + "grad_norm": 0.44140625, + "learning_rate": 0.00017997639947864676, + "loss": 2.6425, + "step": 9699 + }, + { + "epoch": 0.41, + "grad_norm": 0.416015625, + "learning_rate": 0.00017997237242985375, + "loss": 2.6048, + "step": 9700 + }, + { + "epoch": 0.41, + "grad_norm": 0.486328125, + "learning_rate": 0.00017996834502121428, + "loss": 2.6223, + "step": 9701 + }, + { + "epoch": 0.41, + "grad_norm": 0.4609375, + "learning_rate": 0.0001799643172527465, + "loss": 2.5954, + "step": 9702 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.00017996028912446854, + "loss": 2.5895, + "step": 9703 + }, + { + "epoch": 0.41, + "grad_norm": 0.462890625, + "learning_rate": 0.00017995626063639847, + "loss": 2.7397, + "step": 9704 + }, + { + "epoch": 0.41, + "grad_norm": 0.4375, + "learning_rate": 0.00017995223178855446, + "loss": 2.6301, + "step": 9705 + }, + { + "epoch": 0.41, + "grad_norm": 0.4140625, + "learning_rate": 0.00017994820258095462, + "loss": 2.6486, + "step": 9706 + }, + { + "epoch": 0.41, + "grad_norm": 0.419921875, + "learning_rate": 0.00017994417301361707, + "loss": 2.6444, + "step": 9707 + }, + { + "epoch": 0.41, + "grad_norm": 0.365234375, + "learning_rate": 0.00017994014308655998, + "loss": 2.6374, + "step": 9708 + }, + { + "epoch": 0.41, + "grad_norm": 0.474609375, + "learning_rate": 0.00017993611279980146, + "loss": 2.6475, + "step": 9709 + }, + { + "epoch": 0.41, + "grad_norm": 0.494140625, + "learning_rate": 0.00017993208215335966, + "loss": 2.6904, + "step": 9710 + }, + { + "epoch": 0.41, + "grad_norm": 0.431640625, + "learning_rate": 0.00017992805114725265, + "loss": 2.5558, + "step": 9711 + }, + { + "epoch": 0.41, + "grad_norm": 0.408203125, + "learning_rate": 0.00017992401978149868, + "loss": 2.4212, + "step": 9712 + }, + { + "epoch": 0.41, + "grad_norm": 0.451171875, + "learning_rate": 0.0001799199880561158, + "loss": 2.6071, + "step": 9713 + }, + { + "epoch": 0.41, + "grad_norm": 0.423828125, + "learning_rate": 0.0001799159559711222, + "loss": 2.577, + "step": 9714 + }, + { + "epoch": 0.41, + "grad_norm": 0.419921875, + "learning_rate": 0.000179911923526536, + "loss": 2.6134, + "step": 9715 + }, + { + "epoch": 0.41, + "grad_norm": 0.443359375, + "learning_rate": 0.00017990789072237532, + "loss": 2.5454, + "step": 9716 + }, + { + "epoch": 0.41, + "grad_norm": 0.5390625, + "learning_rate": 0.00017990385755865836, + "loss": 2.5561, + "step": 9717 + }, + { + "epoch": 0.41, + "grad_norm": 0.462890625, + "learning_rate": 0.00017989982403540325, + "loss": 2.6987, + "step": 9718 + }, + { + "epoch": 0.41, + "grad_norm": 0.474609375, + "learning_rate": 0.0001798957901526281, + "loss": 2.563, + "step": 9719 + }, + { + "epoch": 0.41, + "grad_norm": 0.482421875, + "learning_rate": 0.0001798917559103511, + "loss": 2.5955, + "step": 9720 + }, + { + "epoch": 0.41, + "grad_norm": 0.4609375, + "learning_rate": 0.0001798877213085904, + "loss": 2.5054, + "step": 9721 + }, + { + "epoch": 0.41, + "grad_norm": 0.51171875, + "learning_rate": 0.00017988368634736415, + "loss": 2.6454, + "step": 9722 + }, + { + "epoch": 0.41, + "grad_norm": 0.4609375, + "learning_rate": 0.00017987965102669054, + "loss": 2.5104, + "step": 9723 + }, + { + "epoch": 0.41, + "grad_norm": 0.404296875, + "learning_rate": 0.00017987561534658764, + "loss": 2.5312, + "step": 9724 + }, + { + "epoch": 0.41, + "grad_norm": 0.46484375, + "learning_rate": 0.00017987157930707373, + "loss": 2.6868, + "step": 9725 + }, + { + "epoch": 0.41, + "grad_norm": 0.44140625, + "learning_rate": 0.00017986754290816686, + "loss": 2.5277, + "step": 9726 + }, + { + "epoch": 0.41, + "grad_norm": 0.515625, + "learning_rate": 0.00017986350614988524, + "loss": 2.3591, + "step": 9727 + }, + { + "epoch": 0.41, + "grad_norm": 0.435546875, + "learning_rate": 0.000179859469032247, + "loss": 2.5952, + "step": 9728 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.00017985543155527037, + "loss": 2.6248, + "step": 9729 + }, + { + "epoch": 0.41, + "grad_norm": 0.439453125, + "learning_rate": 0.00017985139371897346, + "loss": 2.5991, + "step": 9730 + }, + { + "epoch": 0.41, + "grad_norm": 0.455078125, + "learning_rate": 0.00017984735552337448, + "loss": 2.6267, + "step": 9731 + }, + { + "epoch": 0.41, + "grad_norm": 0.392578125, + "learning_rate": 0.00017984331696849158, + "loss": 2.6271, + "step": 9732 + }, + { + "epoch": 0.41, + "grad_norm": 0.484375, + "learning_rate": 0.00017983927805434292, + "loss": 2.5307, + "step": 9733 + }, + { + "epoch": 0.41, + "grad_norm": 0.462890625, + "learning_rate": 0.0001798352387809467, + "loss": 2.6475, + "step": 9734 + }, + { + "epoch": 0.41, + "grad_norm": 0.4765625, + "learning_rate": 0.00017983119914832108, + "loss": 2.5827, + "step": 9735 + }, + { + "epoch": 0.41, + "grad_norm": 0.44140625, + "learning_rate": 0.00017982715915648421, + "loss": 2.5424, + "step": 9736 + }, + { + "epoch": 0.41, + "grad_norm": 0.40234375, + "learning_rate": 0.00017982311880545431, + "loss": 2.6383, + "step": 9737 + }, + { + "epoch": 0.41, + "grad_norm": 0.439453125, + "learning_rate": 0.00017981907809524957, + "loss": 2.5869, + "step": 9738 + }, + { + "epoch": 0.42, + "grad_norm": 0.46484375, + "learning_rate": 0.00017981503702588813, + "loss": 2.6427, + "step": 9739 + }, + { + "epoch": 0.42, + "grad_norm": 0.455078125, + "learning_rate": 0.0001798109955973882, + "loss": 2.7417, + "step": 9740 + }, + { + "epoch": 0.42, + "grad_norm": 0.46484375, + "learning_rate": 0.00017980695380976798, + "loss": 2.5415, + "step": 9741 + }, + { + "epoch": 0.42, + "grad_norm": 0.435546875, + "learning_rate": 0.0001798029116630456, + "loss": 2.5474, + "step": 9742 + }, + { + "epoch": 0.42, + "grad_norm": 0.451171875, + "learning_rate": 0.0001797988691572393, + "loss": 2.6605, + "step": 9743 + }, + { + "epoch": 0.42, + "grad_norm": 0.421875, + "learning_rate": 0.00017979482629236723, + "loss": 2.7111, + "step": 9744 + }, + { + "epoch": 0.42, + "grad_norm": 0.4765625, + "learning_rate": 0.00017979078306844763, + "loss": 2.5856, + "step": 9745 + }, + { + "epoch": 0.42, + "grad_norm": 0.44140625, + "learning_rate": 0.00017978673948549864, + "loss": 2.4576, + "step": 9746 + }, + { + "epoch": 0.42, + "grad_norm": 0.470703125, + "learning_rate": 0.0001797826955435385, + "loss": 2.6277, + "step": 9747 + }, + { + "epoch": 0.42, + "grad_norm": 0.4375, + "learning_rate": 0.0001797786512425854, + "loss": 2.5169, + "step": 9748 + }, + { + "epoch": 0.42, + "grad_norm": 0.455078125, + "learning_rate": 0.0001797746065826575, + "loss": 2.6233, + "step": 9749 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.00017977056156377304, + "loss": 2.5358, + "step": 9750 + }, + { + "epoch": 0.42, + "grad_norm": 0.45703125, + "learning_rate": 0.00017976651618595017, + "loss": 2.4689, + "step": 9751 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.00017976247044920718, + "loss": 2.752, + "step": 9752 + }, + { + "epoch": 0.42, + "grad_norm": 0.462890625, + "learning_rate": 0.00017975842435356221, + "loss": 2.622, + "step": 9753 + }, + { + "epoch": 0.42, + "grad_norm": 0.5234375, + "learning_rate": 0.00017975437789903348, + "loss": 2.5866, + "step": 9754 + }, + { + "epoch": 0.42, + "grad_norm": 0.478515625, + "learning_rate": 0.00017975033108563916, + "loss": 2.5356, + "step": 9755 + }, + { + "epoch": 0.42, + "grad_norm": 0.55078125, + "learning_rate": 0.0001797462839133975, + "loss": 2.6861, + "step": 9756 + }, + { + "epoch": 0.42, + "grad_norm": 0.4296875, + "learning_rate": 0.00017974223638232673, + "loss": 2.5897, + "step": 9757 + }, + { + "epoch": 0.42, + "grad_norm": 0.494140625, + "learning_rate": 0.00017973818849244505, + "loss": 2.6601, + "step": 9758 + }, + { + "epoch": 0.42, + "grad_norm": 0.48828125, + "learning_rate": 0.00017973414024377064, + "loss": 2.7242, + "step": 9759 + }, + { + "epoch": 0.42, + "grad_norm": 0.48046875, + "learning_rate": 0.00017973009163632176, + "loss": 2.5873, + "step": 9760 + }, + { + "epoch": 0.42, + "grad_norm": 0.484375, + "learning_rate": 0.00017972604267011658, + "loss": 2.7279, + "step": 9761 + }, + { + "epoch": 0.42, + "grad_norm": 0.38671875, + "learning_rate": 0.00017972199334517335, + "loss": 2.5444, + "step": 9762 + }, + { + "epoch": 0.42, + "grad_norm": 0.490234375, + "learning_rate": 0.00017971794366151032, + "loss": 2.6511, + "step": 9763 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.00017971389361914563, + "loss": 2.6149, + "step": 9764 + }, + { + "epoch": 0.42, + "grad_norm": 0.466796875, + "learning_rate": 0.00017970984321809757, + "loss": 2.694, + "step": 9765 + }, + { + "epoch": 0.42, + "grad_norm": 0.4765625, + "learning_rate": 0.00017970579245838434, + "loss": 2.5915, + "step": 9766 + }, + { + "epoch": 0.42, + "grad_norm": 0.451171875, + "learning_rate": 0.0001797017413400242, + "loss": 2.5367, + "step": 9767 + }, + { + "epoch": 0.42, + "grad_norm": 0.46484375, + "learning_rate": 0.00017969768986303533, + "loss": 2.7894, + "step": 9768 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.00017969363802743598, + "loss": 2.6514, + "step": 9769 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.00017968958583324438, + "loss": 2.5615, + "step": 9770 + }, + { + "epoch": 0.42, + "grad_norm": 0.447265625, + "learning_rate": 0.00017968553328047876, + "loss": 2.5451, + "step": 9771 + }, + { + "epoch": 0.42, + "grad_norm": 0.421875, + "learning_rate": 0.00017968148036915738, + "loss": 2.5269, + "step": 9772 + }, + { + "epoch": 0.42, + "grad_norm": 0.4140625, + "learning_rate": 0.00017967742709929847, + "loss": 2.6762, + "step": 9773 + }, + { + "epoch": 0.42, + "grad_norm": 0.404296875, + "learning_rate": 0.00017967337347092022, + "loss": 2.6596, + "step": 9774 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.00017966931948404092, + "loss": 2.6058, + "step": 9775 + }, + { + "epoch": 0.42, + "grad_norm": 0.48046875, + "learning_rate": 0.0001796652651386788, + "loss": 2.6807, + "step": 9776 + }, + { + "epoch": 0.42, + "grad_norm": 0.48046875, + "learning_rate": 0.00017966121043485215, + "loss": 2.6191, + "step": 9777 + }, + { + "epoch": 0.42, + "grad_norm": 0.44921875, + "learning_rate": 0.0001796571553725791, + "loss": 2.5551, + "step": 9778 + }, + { + "epoch": 0.42, + "grad_norm": 0.4375, + "learning_rate": 0.000179653099951878, + "loss": 2.6052, + "step": 9779 + }, + { + "epoch": 0.42, + "grad_norm": 0.498046875, + "learning_rate": 0.00017964904417276705, + "loss": 2.6348, + "step": 9780 + }, + { + "epoch": 0.42, + "grad_norm": 0.3984375, + "learning_rate": 0.0001796449880352645, + "loss": 2.4634, + "step": 9781 + }, + { + "epoch": 0.42, + "grad_norm": 0.462890625, + "learning_rate": 0.0001796409315393886, + "loss": 2.5992, + "step": 9782 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.0001796368746851576, + "loss": 2.5917, + "step": 9783 + }, + { + "epoch": 0.42, + "grad_norm": 0.462890625, + "learning_rate": 0.00017963281747258983, + "loss": 2.7422, + "step": 9784 + }, + { + "epoch": 0.42, + "grad_norm": 0.474609375, + "learning_rate": 0.00017962875990170345, + "loss": 2.6559, + "step": 9785 + }, + { + "epoch": 0.42, + "grad_norm": 0.474609375, + "learning_rate": 0.00017962470197251676, + "loss": 2.5698, + "step": 9786 + }, + { + "epoch": 0.42, + "grad_norm": 0.41796875, + "learning_rate": 0.000179620643685048, + "loss": 2.5796, + "step": 9787 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.00017961658503931545, + "loss": 2.5891, + "step": 9788 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.00017961252603533733, + "loss": 2.6009, + "step": 9789 + }, + { + "epoch": 0.42, + "grad_norm": 0.40234375, + "learning_rate": 0.00017960846667313196, + "loss": 2.6322, + "step": 9790 + }, + { + "epoch": 0.42, + "grad_norm": 0.4453125, + "learning_rate": 0.00017960440695271757, + "loss": 2.6585, + "step": 9791 + }, + { + "epoch": 0.42, + "grad_norm": 0.400390625, + "learning_rate": 0.0001796003468741125, + "loss": 2.732, + "step": 9792 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.00017959628643733492, + "loss": 2.6373, + "step": 9793 + }, + { + "epoch": 0.42, + "grad_norm": 0.453125, + "learning_rate": 0.00017959222564240313, + "loss": 2.7396, + "step": 9794 + }, + { + "epoch": 0.42, + "grad_norm": 0.421875, + "learning_rate": 0.00017958816448933543, + "loss": 2.5958, + "step": 9795 + }, + { + "epoch": 0.42, + "grad_norm": 0.515625, + "learning_rate": 0.00017958410297815006, + "loss": 2.7182, + "step": 9796 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.00017958004110886532, + "loss": 2.5606, + "step": 9797 + }, + { + "epoch": 0.42, + "grad_norm": 0.48828125, + "learning_rate": 0.00017957597888149947, + "loss": 2.7052, + "step": 9798 + }, + { + "epoch": 0.42, + "grad_norm": 0.5, + "learning_rate": 0.00017957191629607082, + "loss": 2.6437, + "step": 9799 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.0001795678533525976, + "loss": 2.6966, + "step": 9800 + }, + { + "epoch": 0.42, + "grad_norm": 0.486328125, + "learning_rate": 0.00017956379005109813, + "loss": 2.5898, + "step": 9801 + }, + { + "epoch": 0.42, + "grad_norm": 0.4609375, + "learning_rate": 0.00017955972639159068, + "loss": 2.6074, + "step": 9802 + }, + { + "epoch": 0.42, + "grad_norm": 0.4609375, + "learning_rate": 0.00017955566237409353, + "loss": 2.7325, + "step": 9803 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.000179551597998625, + "loss": 2.5159, + "step": 9804 + }, + { + "epoch": 0.42, + "grad_norm": 0.43359375, + "learning_rate": 0.00017954753326520332, + "loss": 2.4667, + "step": 9805 + }, + { + "epoch": 0.42, + "grad_norm": 0.421875, + "learning_rate": 0.00017954346817384683, + "loss": 2.6282, + "step": 9806 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.00017953940272457379, + "loss": 2.606, + "step": 9807 + }, + { + "epoch": 0.42, + "grad_norm": 0.490234375, + "learning_rate": 0.00017953533691740251, + "loss": 2.6612, + "step": 9808 + }, + { + "epoch": 0.42, + "grad_norm": 0.392578125, + "learning_rate": 0.00017953127075235128, + "loss": 2.3886, + "step": 9809 + }, + { + "epoch": 0.42, + "grad_norm": 0.376953125, + "learning_rate": 0.00017952720422943842, + "loss": 2.7082, + "step": 9810 + }, + { + "epoch": 0.42, + "grad_norm": 0.447265625, + "learning_rate": 0.0001795231373486822, + "loss": 2.697, + "step": 9811 + }, + { + "epoch": 0.42, + "grad_norm": 0.55859375, + "learning_rate": 0.00017951907011010088, + "loss": 2.7286, + "step": 9812 + }, + { + "epoch": 0.42, + "grad_norm": 0.484375, + "learning_rate": 0.00017951500251371286, + "loss": 2.7909, + "step": 9813 + }, + { + "epoch": 0.42, + "grad_norm": 0.462890625, + "learning_rate": 0.00017951093455953636, + "loss": 2.4771, + "step": 9814 + }, + { + "epoch": 0.42, + "grad_norm": 0.44140625, + "learning_rate": 0.0001795068662475897, + "loss": 2.6751, + "step": 9815 + }, + { + "epoch": 0.42, + "grad_norm": 0.50390625, + "learning_rate": 0.00017950279757789121, + "loss": 2.6009, + "step": 9816 + }, + { + "epoch": 0.42, + "grad_norm": 0.408203125, + "learning_rate": 0.0001794987285504592, + "loss": 2.5885, + "step": 9817 + }, + { + "epoch": 0.42, + "grad_norm": 0.42578125, + "learning_rate": 0.00017949465916531197, + "loss": 2.6168, + "step": 9818 + }, + { + "epoch": 0.42, + "grad_norm": 0.490234375, + "learning_rate": 0.0001794905894224678, + "loss": 2.6635, + "step": 9819 + }, + { + "epoch": 0.42, + "grad_norm": 0.486328125, + "learning_rate": 0.00017948651932194503, + "loss": 2.5748, + "step": 9820 + }, + { + "epoch": 0.42, + "grad_norm": 0.462890625, + "learning_rate": 0.000179482448863762, + "loss": 2.5704, + "step": 9821 + }, + { + "epoch": 0.42, + "grad_norm": 0.45703125, + "learning_rate": 0.00017947837804793695, + "loss": 2.7549, + "step": 9822 + }, + { + "epoch": 0.42, + "grad_norm": 0.5078125, + "learning_rate": 0.0001794743068744883, + "loss": 2.6152, + "step": 9823 + }, + { + "epoch": 0.42, + "grad_norm": 0.48828125, + "learning_rate": 0.0001794702353434343, + "loss": 2.6187, + "step": 9824 + }, + { + "epoch": 0.42, + "grad_norm": 0.466796875, + "learning_rate": 0.00017946616345479324, + "loss": 2.6123, + "step": 9825 + }, + { + "epoch": 0.42, + "grad_norm": 0.50390625, + "learning_rate": 0.00017946209120858354, + "loss": 2.6739, + "step": 9826 + }, + { + "epoch": 0.42, + "grad_norm": 0.453125, + "learning_rate": 0.00017945801860482345, + "loss": 2.5927, + "step": 9827 + }, + { + "epoch": 0.42, + "grad_norm": 0.474609375, + "learning_rate": 0.0001794539456435313, + "loss": 2.5895, + "step": 9828 + }, + { + "epoch": 0.42, + "grad_norm": 0.42578125, + "learning_rate": 0.00017944987232472546, + "loss": 2.5274, + "step": 9829 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.0001794457986484242, + "loss": 2.5789, + "step": 9830 + }, + { + "epoch": 0.42, + "grad_norm": 0.439453125, + "learning_rate": 0.0001794417246146459, + "loss": 2.5349, + "step": 9831 + }, + { + "epoch": 0.42, + "grad_norm": 0.45703125, + "learning_rate": 0.00017943765022340887, + "loss": 2.6638, + "step": 9832 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.00017943357547473143, + "loss": 2.5212, + "step": 9833 + }, + { + "epoch": 0.42, + "grad_norm": 0.4609375, + "learning_rate": 0.00017942950036863196, + "loss": 2.7309, + "step": 9834 + }, + { + "epoch": 0.42, + "grad_norm": 0.46484375, + "learning_rate": 0.00017942542490512873, + "loss": 2.5886, + "step": 9835 + }, + { + "epoch": 0.42, + "grad_norm": 0.5078125, + "learning_rate": 0.00017942134908424014, + "loss": 2.7292, + "step": 9836 + }, + { + "epoch": 0.42, + "grad_norm": 0.484375, + "learning_rate": 0.0001794172729059845, + "loss": 2.6627, + "step": 9837 + }, + { + "epoch": 0.42, + "grad_norm": 0.4765625, + "learning_rate": 0.00017941319637038015, + "loss": 2.4421, + "step": 9838 + }, + { + "epoch": 0.42, + "grad_norm": 0.453125, + "learning_rate": 0.00017940911947744544, + "loss": 2.6413, + "step": 9839 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.00017940504222719873, + "loss": 2.5506, + "step": 9840 + }, + { + "epoch": 0.42, + "grad_norm": 0.435546875, + "learning_rate": 0.00017940096461965832, + "loss": 2.5444, + "step": 9841 + }, + { + "epoch": 0.42, + "grad_norm": 0.453125, + "learning_rate": 0.0001793968866548426, + "loss": 2.5508, + "step": 9842 + }, + { + "epoch": 0.42, + "grad_norm": 0.4375, + "learning_rate": 0.0001793928083327699, + "loss": 2.6359, + "step": 9843 + }, + { + "epoch": 0.42, + "grad_norm": 0.46875, + "learning_rate": 0.00017938872965345858, + "loss": 2.6397, + "step": 9844 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.000179384650616927, + "loss": 2.5831, + "step": 9845 + }, + { + "epoch": 0.42, + "grad_norm": 0.44140625, + "learning_rate": 0.0001793805712231935, + "loss": 2.5653, + "step": 9846 + }, + { + "epoch": 0.42, + "grad_norm": 0.470703125, + "learning_rate": 0.00017937649147227645, + "loss": 2.5742, + "step": 9847 + }, + { + "epoch": 0.42, + "grad_norm": 0.408203125, + "learning_rate": 0.0001793724113641942, + "loss": 2.8227, + "step": 9848 + }, + { + "epoch": 0.42, + "grad_norm": 0.423828125, + "learning_rate": 0.00017936833089896506, + "loss": 2.733, + "step": 9849 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.00017936425007660745, + "loss": 2.5557, + "step": 9850 + }, + { + "epoch": 0.42, + "grad_norm": 0.44921875, + "learning_rate": 0.00017936016889713975, + "loss": 2.6253, + "step": 9851 + }, + { + "epoch": 0.42, + "grad_norm": 0.474609375, + "learning_rate": 0.00017935608736058029, + "loss": 2.6079, + "step": 9852 + }, + { + "epoch": 0.42, + "grad_norm": 0.4453125, + "learning_rate": 0.00017935200546694742, + "loss": 2.566, + "step": 9853 + }, + { + "epoch": 0.42, + "grad_norm": 0.39453125, + "learning_rate": 0.00017934792321625955, + "loss": 2.5476, + "step": 9854 + }, + { + "epoch": 0.42, + "grad_norm": 0.4765625, + "learning_rate": 0.000179343840608535, + "loss": 2.5668, + "step": 9855 + }, + { + "epoch": 0.42, + "grad_norm": 0.423828125, + "learning_rate": 0.00017933975764379217, + "loss": 2.5653, + "step": 9856 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.00017933567432204943, + "loss": 2.7211, + "step": 9857 + }, + { + "epoch": 0.42, + "grad_norm": 0.4765625, + "learning_rate": 0.00017933159064332515, + "loss": 2.6026, + "step": 9858 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.0001793275066076377, + "loss": 2.57, + "step": 9859 + }, + { + "epoch": 0.42, + "grad_norm": 0.439453125, + "learning_rate": 0.00017932342221500545, + "loss": 2.5426, + "step": 9860 + }, + { + "epoch": 0.42, + "grad_norm": 0.490234375, + "learning_rate": 0.0001793193374654468, + "loss": 2.6128, + "step": 9861 + }, + { + "epoch": 0.42, + "grad_norm": 0.435546875, + "learning_rate": 0.00017931525235898013, + "loss": 2.7401, + "step": 9862 + }, + { + "epoch": 0.42, + "grad_norm": 0.498046875, + "learning_rate": 0.0001793111668956238, + "loss": 2.6419, + "step": 9863 + }, + { + "epoch": 0.42, + "grad_norm": 0.44140625, + "learning_rate": 0.0001793070810753962, + "loss": 2.6214, + "step": 9864 + }, + { + "epoch": 0.42, + "grad_norm": 0.50390625, + "learning_rate": 0.00017930299489831572, + "loss": 2.5245, + "step": 9865 + }, + { + "epoch": 0.42, + "grad_norm": 0.478515625, + "learning_rate": 0.00017929890836440075, + "loss": 2.5052, + "step": 9866 + }, + { + "epoch": 0.42, + "grad_norm": 0.439453125, + "learning_rate": 0.00017929482147366967, + "loss": 2.5369, + "step": 9867 + }, + { + "epoch": 0.42, + "grad_norm": 0.48046875, + "learning_rate": 0.00017929073422614088, + "loss": 2.7793, + "step": 9868 + }, + { + "epoch": 0.42, + "grad_norm": 0.4140625, + "learning_rate": 0.00017928664662183276, + "loss": 2.6028, + "step": 9869 + }, + { + "epoch": 0.42, + "grad_norm": 0.43359375, + "learning_rate": 0.00017928255866076368, + "loss": 2.553, + "step": 9870 + }, + { + "epoch": 0.42, + "grad_norm": 0.4921875, + "learning_rate": 0.0001792784703429521, + "loss": 2.6904, + "step": 9871 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.00017927438166841637, + "loss": 2.6117, + "step": 9872 + }, + { + "epoch": 0.42, + "grad_norm": 0.419921875, + "learning_rate": 0.0001792702926371749, + "loss": 2.5263, + "step": 9873 + }, + { + "epoch": 0.42, + "grad_norm": 0.462890625, + "learning_rate": 0.00017926620324924602, + "loss": 2.6124, + "step": 9874 + }, + { + "epoch": 0.42, + "grad_norm": 0.44921875, + "learning_rate": 0.00017926211350464827, + "loss": 2.7676, + "step": 9875 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.00017925802340339994, + "loss": 2.5195, + "step": 9876 + }, + { + "epoch": 0.42, + "grad_norm": 0.462890625, + "learning_rate": 0.0001792539329455195, + "loss": 2.4998, + "step": 9877 + }, + { + "epoch": 0.42, + "grad_norm": 0.470703125, + "learning_rate": 0.0001792498421310253, + "loss": 2.6048, + "step": 9878 + }, + { + "epoch": 0.42, + "grad_norm": 0.484375, + "learning_rate": 0.00017924575095993578, + "loss": 2.7054, + "step": 9879 + }, + { + "epoch": 0.42, + "grad_norm": 0.447265625, + "learning_rate": 0.0001792416594322693, + "loss": 2.7048, + "step": 9880 + }, + { + "epoch": 0.42, + "grad_norm": 0.48828125, + "learning_rate": 0.00017923756754804437, + "loss": 2.6162, + "step": 9881 + }, + { + "epoch": 0.42, + "grad_norm": 0.44921875, + "learning_rate": 0.00017923347530727934, + "loss": 2.5503, + "step": 9882 + }, + { + "epoch": 0.42, + "grad_norm": 0.44921875, + "learning_rate": 0.0001792293827099926, + "loss": 2.601, + "step": 9883 + }, + { + "epoch": 0.42, + "grad_norm": 0.4921875, + "learning_rate": 0.0001792252897562026, + "loss": 2.7637, + "step": 9884 + }, + { + "epoch": 0.42, + "grad_norm": 0.44140625, + "learning_rate": 0.00017922119644592775, + "loss": 2.5901, + "step": 9885 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.0001792171027791865, + "loss": 2.7529, + "step": 9886 + }, + { + "epoch": 0.42, + "grad_norm": 0.494140625, + "learning_rate": 0.0001792130087559972, + "loss": 2.5421, + "step": 9887 + }, + { + "epoch": 0.42, + "grad_norm": 0.482421875, + "learning_rate": 0.0001792089143763783, + "loss": 2.5781, + "step": 9888 + }, + { + "epoch": 0.42, + "grad_norm": 0.46484375, + "learning_rate": 0.00017920481964034827, + "loss": 2.4903, + "step": 9889 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.00017920072454792547, + "loss": 2.7546, + "step": 9890 + }, + { + "epoch": 0.42, + "grad_norm": 0.474609375, + "learning_rate": 0.00017919662909912835, + "loss": 2.5918, + "step": 9891 + }, + { + "epoch": 0.42, + "grad_norm": 0.5078125, + "learning_rate": 0.0001791925332939754, + "loss": 2.6699, + "step": 9892 + }, + { + "epoch": 0.42, + "grad_norm": 0.470703125, + "learning_rate": 0.00017918843713248493, + "loss": 2.5254, + "step": 9893 + }, + { + "epoch": 0.42, + "grad_norm": 0.443359375, + "learning_rate": 0.00017918434061467547, + "loss": 2.5094, + "step": 9894 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.00017918024374056536, + "loss": 2.5459, + "step": 9895 + }, + { + "epoch": 0.42, + "grad_norm": 0.44140625, + "learning_rate": 0.00017917614651017315, + "loss": 2.6219, + "step": 9896 + }, + { + "epoch": 0.42, + "grad_norm": 0.431640625, + "learning_rate": 0.00017917204892351719, + "loss": 2.6574, + "step": 9897 + }, + { + "epoch": 0.42, + "grad_norm": 0.486328125, + "learning_rate": 0.00017916795098061594, + "loss": 2.5641, + "step": 9898 + }, + { + "epoch": 0.42, + "grad_norm": 0.404296875, + "learning_rate": 0.00017916385268148784, + "loss": 2.5797, + "step": 9899 + }, + { + "epoch": 0.42, + "grad_norm": 0.431640625, + "learning_rate": 0.00017915975402615138, + "loss": 2.5835, + "step": 9900 + }, + { + "epoch": 0.42, + "grad_norm": 0.4921875, + "learning_rate": 0.0001791556550146249, + "loss": 2.5743, + "step": 9901 + }, + { + "epoch": 0.42, + "grad_norm": 0.419921875, + "learning_rate": 0.00017915155564692691, + "loss": 2.6536, + "step": 9902 + }, + { + "epoch": 0.42, + "grad_norm": 0.4453125, + "learning_rate": 0.00017914745592307587, + "loss": 2.5741, + "step": 9903 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.0001791433558430902, + "loss": 2.5438, + "step": 9904 + }, + { + "epoch": 0.42, + "grad_norm": 0.4375, + "learning_rate": 0.00017913925540698832, + "loss": 2.5897, + "step": 9905 + }, + { + "epoch": 0.42, + "grad_norm": 0.44921875, + "learning_rate": 0.00017913515461478877, + "loss": 2.5819, + "step": 9906 + }, + { + "epoch": 0.42, + "grad_norm": 0.53125, + "learning_rate": 0.00017913105346650988, + "loss": 2.5926, + "step": 9907 + }, + { + "epoch": 0.42, + "grad_norm": 0.45703125, + "learning_rate": 0.0001791269519621702, + "loss": 2.5469, + "step": 9908 + }, + { + "epoch": 0.42, + "grad_norm": 0.51953125, + "learning_rate": 0.00017912285010178818, + "loss": 2.8572, + "step": 9909 + }, + { + "epoch": 0.42, + "grad_norm": 0.466796875, + "learning_rate": 0.0001791187478853822, + "loss": 2.5697, + "step": 9910 + }, + { + "epoch": 0.42, + "grad_norm": 0.455078125, + "learning_rate": 0.0001791146453129708, + "loss": 2.551, + "step": 9911 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.00017911054238457238, + "loss": 2.6613, + "step": 9912 + }, + { + "epoch": 0.42, + "grad_norm": 0.44921875, + "learning_rate": 0.00017910643910020547, + "loss": 2.701, + "step": 9913 + }, + { + "epoch": 0.42, + "grad_norm": 0.5703125, + "learning_rate": 0.00017910233545988849, + "loss": 2.6036, + "step": 9914 + }, + { + "epoch": 0.42, + "grad_norm": 0.455078125, + "learning_rate": 0.00017909823146363988, + "loss": 2.5611, + "step": 9915 + }, + { + "epoch": 0.42, + "grad_norm": 0.431640625, + "learning_rate": 0.00017909412711147814, + "loss": 2.6363, + "step": 9916 + }, + { + "epoch": 0.42, + "grad_norm": 0.4296875, + "learning_rate": 0.00017909002240342178, + "loss": 2.6791, + "step": 9917 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.00017908591733948917, + "loss": 2.5659, + "step": 9918 + }, + { + "epoch": 0.42, + "grad_norm": 0.42578125, + "learning_rate": 0.00017908181191969886, + "loss": 2.5851, + "step": 9919 + }, + { + "epoch": 0.42, + "grad_norm": 0.494140625, + "learning_rate": 0.00017907770614406928, + "loss": 2.7115, + "step": 9920 + }, + { + "epoch": 0.42, + "grad_norm": 0.478515625, + "learning_rate": 0.00017907360001261896, + "loss": 2.4381, + "step": 9921 + }, + { + "epoch": 0.42, + "grad_norm": 0.50390625, + "learning_rate": 0.0001790694935253663, + "loss": 2.5112, + "step": 9922 + }, + { + "epoch": 0.42, + "grad_norm": 0.4140625, + "learning_rate": 0.00017906538668232983, + "loss": 2.5349, + "step": 9923 + }, + { + "epoch": 0.42, + "grad_norm": 0.419921875, + "learning_rate": 0.000179061279483528, + "loss": 2.6663, + "step": 9924 + }, + { + "epoch": 0.42, + "grad_norm": 0.435546875, + "learning_rate": 0.00017905717192897932, + "loss": 2.5369, + "step": 9925 + }, + { + "epoch": 0.42, + "grad_norm": 0.46875, + "learning_rate": 0.00017905306401870227, + "loss": 2.5793, + "step": 9926 + }, + { + "epoch": 0.42, + "grad_norm": 0.4609375, + "learning_rate": 0.0001790489557527153, + "loss": 2.6444, + "step": 9927 + }, + { + "epoch": 0.42, + "grad_norm": 0.4453125, + "learning_rate": 0.0001790448471310369, + "loss": 2.5976, + "step": 9928 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.0001790407381536856, + "loss": 2.5451, + "step": 9929 + }, + { + "epoch": 0.42, + "grad_norm": 0.53515625, + "learning_rate": 0.0001790366288206799, + "loss": 2.7317, + "step": 9930 + }, + { + "epoch": 0.42, + "grad_norm": 0.486328125, + "learning_rate": 0.0001790325191320382, + "loss": 2.6628, + "step": 9931 + }, + { + "epoch": 0.42, + "grad_norm": 0.498046875, + "learning_rate": 0.00017902840908777908, + "loss": 2.5303, + "step": 9932 + }, + { + "epoch": 0.42, + "grad_norm": 0.40625, + "learning_rate": 0.00017902429868792098, + "loss": 2.5853, + "step": 9933 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.00017902018793248243, + "loss": 2.5136, + "step": 9934 + }, + { + "epoch": 0.42, + "grad_norm": 0.419921875, + "learning_rate": 0.00017901607682148192, + "loss": 2.6801, + "step": 9935 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.0001790119653549379, + "loss": 2.7469, + "step": 9936 + }, + { + "epoch": 0.42, + "grad_norm": 0.4453125, + "learning_rate": 0.00017900785353286896, + "loss": 2.4737, + "step": 9937 + }, + { + "epoch": 0.42, + "grad_norm": 0.4453125, + "learning_rate": 0.00017900374135529354, + "loss": 2.6337, + "step": 9938 + }, + { + "epoch": 0.42, + "grad_norm": 0.5, + "learning_rate": 0.00017899962882223015, + "loss": 2.6122, + "step": 9939 + }, + { + "epoch": 0.42, + "grad_norm": 0.41015625, + "learning_rate": 0.0001789955159336973, + "loss": 2.5622, + "step": 9940 + }, + { + "epoch": 0.42, + "grad_norm": 0.44140625, + "learning_rate": 0.00017899140268971348, + "loss": 2.66, + "step": 9941 + }, + { + "epoch": 0.42, + "grad_norm": 0.466796875, + "learning_rate": 0.00017898728909029725, + "loss": 2.6618, + "step": 9942 + }, + { + "epoch": 0.42, + "grad_norm": 0.431640625, + "learning_rate": 0.00017898317513546707, + "loss": 2.5451, + "step": 9943 + }, + { + "epoch": 0.42, + "grad_norm": 0.482421875, + "learning_rate": 0.00017897906082524147, + "loss": 2.5959, + "step": 9944 + }, + { + "epoch": 0.42, + "grad_norm": 0.4609375, + "learning_rate": 0.00017897494615963896, + "loss": 2.5165, + "step": 9945 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.00017897083113867804, + "loss": 2.6831, + "step": 9946 + }, + { + "epoch": 0.42, + "grad_norm": 0.470703125, + "learning_rate": 0.00017896671576237726, + "loss": 2.5981, + "step": 9947 + }, + { + "epoch": 0.42, + "grad_norm": 0.462890625, + "learning_rate": 0.0001789626000307551, + "loss": 2.7394, + "step": 9948 + }, + { + "epoch": 0.42, + "grad_norm": 0.435546875, + "learning_rate": 0.00017895848394383014, + "loss": 2.4726, + "step": 9949 + }, + { + "epoch": 0.42, + "grad_norm": 0.51171875, + "learning_rate": 0.0001789543675016208, + "loss": 2.6838, + "step": 9950 + }, + { + "epoch": 0.42, + "grad_norm": 0.453125, + "learning_rate": 0.00017895025070414572, + "loss": 2.5341, + "step": 9951 + }, + { + "epoch": 0.42, + "grad_norm": 0.453125, + "learning_rate": 0.00017894613355142332, + "loss": 2.5501, + "step": 9952 + }, + { + "epoch": 0.42, + "grad_norm": 0.384765625, + "learning_rate": 0.00017894201604347218, + "loss": 2.6318, + "step": 9953 + }, + { + "epoch": 0.42, + "grad_norm": 0.451171875, + "learning_rate": 0.0001789378981803108, + "loss": 2.6241, + "step": 9954 + }, + { + "epoch": 0.42, + "grad_norm": 0.48828125, + "learning_rate": 0.00017893377996195775, + "loss": 2.5573, + "step": 9955 + }, + { + "epoch": 0.42, + "grad_norm": 0.43359375, + "learning_rate": 0.00017892966138843155, + "loss": 2.5952, + "step": 9956 + }, + { + "epoch": 0.42, + "grad_norm": 0.515625, + "learning_rate": 0.00017892554245975067, + "loss": 2.7561, + "step": 9957 + }, + { + "epoch": 0.42, + "grad_norm": 0.46875, + "learning_rate": 0.00017892142317593374, + "loss": 2.6204, + "step": 9958 + }, + { + "epoch": 0.42, + "grad_norm": 0.50390625, + "learning_rate": 0.00017891730353699925, + "loss": 2.5995, + "step": 9959 + }, + { + "epoch": 0.42, + "grad_norm": 0.447265625, + "learning_rate": 0.0001789131835429657, + "loss": 2.5633, + "step": 9960 + }, + { + "epoch": 0.42, + "grad_norm": 0.43359375, + "learning_rate": 0.0001789090631938517, + "loss": 2.7011, + "step": 9961 + }, + { + "epoch": 0.42, + "grad_norm": 0.4765625, + "learning_rate": 0.00017890494248967573, + "loss": 2.5992, + "step": 9962 + }, + { + "epoch": 0.42, + "grad_norm": 0.458984375, + "learning_rate": 0.00017890082143045638, + "loss": 2.6622, + "step": 9963 + }, + { + "epoch": 0.42, + "grad_norm": 0.45703125, + "learning_rate": 0.00017889670001621215, + "loss": 2.7099, + "step": 9964 + }, + { + "epoch": 0.42, + "grad_norm": 0.482421875, + "learning_rate": 0.0001788925782469616, + "loss": 2.7598, + "step": 9965 + }, + { + "epoch": 0.42, + "grad_norm": 0.427734375, + "learning_rate": 0.00017888845612272326, + "loss": 2.5879, + "step": 9966 + }, + { + "epoch": 0.42, + "grad_norm": 0.5234375, + "learning_rate": 0.00017888433364351576, + "loss": 2.7392, + "step": 9967 + }, + { + "epoch": 0.42, + "grad_norm": 0.47265625, + "learning_rate": 0.00017888021080935754, + "loss": 2.7145, + "step": 9968 + }, + { + "epoch": 0.42, + "grad_norm": 0.439453125, + "learning_rate": 0.0001788760876202672, + "loss": 2.6397, + "step": 9969 + }, + { + "epoch": 0.42, + "grad_norm": 0.4609375, + "learning_rate": 0.0001788719640762633, + "loss": 2.5163, + "step": 9970 + }, + { + "epoch": 0.42, + "grad_norm": 0.4375, + "learning_rate": 0.00017886784017736437, + "loss": 2.5932, + "step": 9971 + }, + { + "epoch": 0.42, + "grad_norm": 0.41796875, + "learning_rate": 0.00017886371592358902, + "loss": 2.7317, + "step": 9972 + }, + { + "epoch": 0.42, + "grad_norm": 0.451171875, + "learning_rate": 0.00017885959131495573, + "loss": 2.6574, + "step": 9973 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.00017885546635148313, + "loss": 2.7011, + "step": 9974 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.00017885134103318975, + "loss": 2.6786, + "step": 9975 + }, + { + "epoch": 0.43, + "grad_norm": 0.439453125, + "learning_rate": 0.00017884721536009411, + "loss": 2.6859, + "step": 9976 + }, + { + "epoch": 0.43, + "grad_norm": 0.421875, + "learning_rate": 0.00017884308933221487, + "loss": 2.6695, + "step": 9977 + }, + { + "epoch": 0.43, + "grad_norm": 0.5234375, + "learning_rate": 0.0001788389629495705, + "loss": 2.645, + "step": 9978 + }, + { + "epoch": 0.43, + "grad_norm": 0.462890625, + "learning_rate": 0.00017883483621217964, + "loss": 2.5532, + "step": 9979 + }, + { + "epoch": 0.43, + "grad_norm": 0.51171875, + "learning_rate": 0.00017883070912006081, + "loss": 2.7992, + "step": 9980 + }, + { + "epoch": 0.43, + "grad_norm": 0.5234375, + "learning_rate": 0.0001788265816732326, + "loss": 2.6128, + "step": 9981 + }, + { + "epoch": 0.43, + "grad_norm": 0.5234375, + "learning_rate": 0.0001788224538717136, + "loss": 2.5784, + "step": 9982 + }, + { + "epoch": 0.43, + "grad_norm": 0.498046875, + "learning_rate": 0.00017881832571552234, + "loss": 2.6578, + "step": 9983 + }, + { + "epoch": 0.43, + "grad_norm": 0.44140625, + "learning_rate": 0.0001788141972046774, + "loss": 2.5214, + "step": 9984 + }, + { + "epoch": 0.43, + "grad_norm": 0.5078125, + "learning_rate": 0.00017881006833919742, + "loss": 2.6839, + "step": 9985 + }, + { + "epoch": 0.43, + "grad_norm": 0.470703125, + "learning_rate": 0.0001788059391191009, + "loss": 2.691, + "step": 9986 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.0001788018095444065, + "loss": 2.7066, + "step": 9987 + }, + { + "epoch": 0.43, + "grad_norm": 0.435546875, + "learning_rate": 0.0001787976796151327, + "loss": 2.6393, + "step": 9988 + }, + { + "epoch": 0.43, + "grad_norm": 0.4453125, + "learning_rate": 0.00017879354933129818, + "loss": 2.6703, + "step": 9989 + }, + { + "epoch": 0.43, + "grad_norm": 0.515625, + "learning_rate": 0.00017878941869292145, + "loss": 2.7324, + "step": 9990 + }, + { + "epoch": 0.43, + "grad_norm": 0.4453125, + "learning_rate": 0.00017878528770002116, + "loss": 2.6573, + "step": 9991 + }, + { + "epoch": 0.43, + "grad_norm": 0.515625, + "learning_rate": 0.00017878115635261584, + "loss": 2.7012, + "step": 9992 + }, + { + "epoch": 0.43, + "grad_norm": 0.4609375, + "learning_rate": 0.00017877702465072414, + "loss": 2.5205, + "step": 9993 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.00017877289259436458, + "loss": 2.6961, + "step": 9994 + }, + { + "epoch": 0.43, + "grad_norm": 0.478515625, + "learning_rate": 0.00017876876018355583, + "loss": 2.7039, + "step": 9995 + }, + { + "epoch": 0.43, + "grad_norm": 0.392578125, + "learning_rate": 0.00017876462741831642, + "loss": 2.7253, + "step": 9996 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.000178760494298665, + "loss": 2.7244, + "step": 9997 + }, + { + "epoch": 0.43, + "grad_norm": 0.404296875, + "learning_rate": 0.0001787563608246201, + "loss": 2.6083, + "step": 9998 + }, + { + "epoch": 0.43, + "grad_norm": 0.458984375, + "learning_rate": 0.0001787522269962004, + "loss": 2.6635, + "step": 9999 + }, + { + "epoch": 0.43, + "grad_norm": 0.44921875, + "learning_rate": 0.00017874809281342443, + "loss": 2.4847, + "step": 10000 + }, + { + "epoch": 0.43, + "grad_norm": 0.47265625, + "learning_rate": 0.00017874395827631082, + "loss": 2.6449, + "step": 10001 + }, + { + "epoch": 0.43, + "grad_norm": 0.419921875, + "learning_rate": 0.0001787398233848782, + "loss": 2.5821, + "step": 10002 + }, + { + "epoch": 0.43, + "grad_norm": 0.51171875, + "learning_rate": 0.00017873568813914513, + "loss": 2.601, + "step": 10003 + }, + { + "epoch": 0.43, + "grad_norm": 0.48046875, + "learning_rate": 0.00017873155253913026, + "loss": 2.5768, + "step": 10004 + }, + { + "epoch": 0.43, + "grad_norm": 0.51171875, + "learning_rate": 0.00017872741658485214, + "loss": 2.6257, + "step": 10005 + }, + { + "epoch": 0.43, + "grad_norm": 0.419921875, + "learning_rate": 0.00017872328027632945, + "loss": 2.6593, + "step": 10006 + }, + { + "epoch": 0.43, + "grad_norm": 0.427734375, + "learning_rate": 0.00017871914361358075, + "loss": 2.6411, + "step": 10007 + }, + { + "epoch": 0.43, + "grad_norm": 0.43359375, + "learning_rate": 0.00017871500659662467, + "loss": 2.597, + "step": 10008 + }, + { + "epoch": 0.43, + "grad_norm": 0.490234375, + "learning_rate": 0.00017871086922547984, + "loss": 2.5721, + "step": 10009 + }, + { + "epoch": 0.43, + "grad_norm": 0.443359375, + "learning_rate": 0.00017870673150016488, + "loss": 2.5446, + "step": 10010 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017870259342069836, + "loss": 2.644, + "step": 10011 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017869845498709897, + "loss": 2.6214, + "step": 10012 + }, + { + "epoch": 0.43, + "grad_norm": 0.431640625, + "learning_rate": 0.00017869431619938525, + "loss": 2.6975, + "step": 10013 + }, + { + "epoch": 0.43, + "grad_norm": 0.462890625, + "learning_rate": 0.00017869017705757586, + "loss": 2.5697, + "step": 10014 + }, + { + "epoch": 0.43, + "grad_norm": 0.4375, + "learning_rate": 0.0001786860375616895, + "loss": 2.6592, + "step": 10015 + }, + { + "epoch": 0.43, + "grad_norm": 0.52734375, + "learning_rate": 0.00017868189771174465, + "loss": 2.6357, + "step": 10016 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017867775750776005, + "loss": 2.5527, + "step": 10017 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.00017867361694975432, + "loss": 2.6488, + "step": 10018 + }, + { + "epoch": 0.43, + "grad_norm": 0.4140625, + "learning_rate": 0.000178669476037746, + "loss": 2.6004, + "step": 10019 + }, + { + "epoch": 0.43, + "grad_norm": 0.486328125, + "learning_rate": 0.00017866533477175385, + "loss": 2.6622, + "step": 10020 + }, + { + "epoch": 0.43, + "grad_norm": 0.439453125, + "learning_rate": 0.00017866119315179644, + "loss": 2.6207, + "step": 10021 + }, + { + "epoch": 0.43, + "grad_norm": 0.4375, + "learning_rate": 0.0001786570511778924, + "loss": 2.7305, + "step": 10022 + }, + { + "epoch": 0.43, + "grad_norm": 0.431640625, + "learning_rate": 0.00017865290885006032, + "loss": 2.6904, + "step": 10023 + }, + { + "epoch": 0.43, + "grad_norm": 0.42578125, + "learning_rate": 0.00017864876616831895, + "loss": 2.5057, + "step": 10024 + }, + { + "epoch": 0.43, + "grad_norm": 0.50390625, + "learning_rate": 0.00017864462313268685, + "loss": 2.6015, + "step": 10025 + }, + { + "epoch": 0.43, + "grad_norm": 0.3984375, + "learning_rate": 0.00017864047974318272, + "loss": 2.5665, + "step": 10026 + }, + { + "epoch": 0.43, + "grad_norm": 0.41796875, + "learning_rate": 0.00017863633599982513, + "loss": 2.6754, + "step": 10027 + }, + { + "epoch": 0.43, + "grad_norm": 0.4609375, + "learning_rate": 0.0001786321919026328, + "loss": 2.5835, + "step": 10028 + }, + { + "epoch": 0.43, + "grad_norm": 0.4296875, + "learning_rate": 0.00017862804745162435, + "loss": 2.5026, + "step": 10029 + }, + { + "epoch": 0.43, + "grad_norm": 0.462890625, + "learning_rate": 0.0001786239026468184, + "loss": 2.644, + "step": 10030 + }, + { + "epoch": 0.43, + "grad_norm": 0.44140625, + "learning_rate": 0.0001786197574882336, + "loss": 2.4584, + "step": 10031 + }, + { + "epoch": 0.43, + "grad_norm": 0.443359375, + "learning_rate": 0.00017861561197588866, + "loss": 2.4838, + "step": 10032 + }, + { + "epoch": 0.43, + "grad_norm": 0.44140625, + "learning_rate": 0.00017861146610980216, + "loss": 2.5457, + "step": 10033 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017860731988999286, + "loss": 2.6971, + "step": 10034 + }, + { + "epoch": 0.43, + "grad_norm": 0.40625, + "learning_rate": 0.00017860317331647928, + "loss": 2.5634, + "step": 10035 + }, + { + "epoch": 0.43, + "grad_norm": 0.416015625, + "learning_rate": 0.00017859902638928018, + "loss": 2.5196, + "step": 10036 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.00017859487910841418, + "loss": 2.6946, + "step": 10037 + }, + { + "epoch": 0.43, + "grad_norm": 0.5, + "learning_rate": 0.00017859073147389993, + "loss": 2.721, + "step": 10038 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.00017858658348575613, + "loss": 2.619, + "step": 10039 + }, + { + "epoch": 0.43, + "grad_norm": 0.421875, + "learning_rate": 0.00017858243514400144, + "loss": 2.5569, + "step": 10040 + }, + { + "epoch": 0.43, + "grad_norm": 0.49609375, + "learning_rate": 0.0001785782864486545, + "loss": 2.6112, + "step": 10041 + }, + { + "epoch": 0.43, + "grad_norm": 0.46875, + "learning_rate": 0.00017857413739973395, + "loss": 2.6572, + "step": 10042 + }, + { + "epoch": 0.43, + "grad_norm": 0.4765625, + "learning_rate": 0.00017856998799725853, + "loss": 2.6461, + "step": 10043 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.00017856583824124686, + "loss": 2.5788, + "step": 10044 + }, + { + "epoch": 0.43, + "grad_norm": 0.515625, + "learning_rate": 0.00017856168813171766, + "loss": 2.6009, + "step": 10045 + }, + { + "epoch": 0.43, + "grad_norm": 0.421875, + "learning_rate": 0.00017855753766868955, + "loss": 2.5681, + "step": 10046 + }, + { + "epoch": 0.43, + "grad_norm": 0.482421875, + "learning_rate": 0.00017855338685218123, + "loss": 2.5399, + "step": 10047 + }, + { + "epoch": 0.43, + "grad_norm": 0.482421875, + "learning_rate": 0.0001785492356822114, + "loss": 2.6512, + "step": 10048 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.0001785450841587987, + "loss": 2.6313, + "step": 10049 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.00017854093228196182, + "loss": 2.5, + "step": 10050 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017853678005171943, + "loss": 2.7121, + "step": 10051 + }, + { + "epoch": 0.43, + "grad_norm": 0.44921875, + "learning_rate": 0.00017853262746809028, + "loss": 2.5737, + "step": 10052 + }, + { + "epoch": 0.43, + "grad_norm": 0.515625, + "learning_rate": 0.00017852847453109298, + "loss": 2.7432, + "step": 10053 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017852432124074624, + "loss": 2.5948, + "step": 10054 + }, + { + "epoch": 0.43, + "grad_norm": 0.458984375, + "learning_rate": 0.00017852016759706878, + "loss": 2.4888, + "step": 10055 + }, + { + "epoch": 0.43, + "grad_norm": 0.439453125, + "learning_rate": 0.00017851601360007924, + "loss": 2.5338, + "step": 10056 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.00017851185924979633, + "loss": 2.7174, + "step": 10057 + }, + { + "epoch": 0.43, + "grad_norm": 0.455078125, + "learning_rate": 0.00017850770454623876, + "loss": 2.5686, + "step": 10058 + }, + { + "epoch": 0.43, + "grad_norm": 0.44921875, + "learning_rate": 0.00017850354948942515, + "loss": 2.7469, + "step": 10059 + }, + { + "epoch": 0.43, + "grad_norm": 0.451171875, + "learning_rate": 0.00017849939407937434, + "loss": 2.5798, + "step": 10060 + }, + { + "epoch": 0.43, + "grad_norm": 0.400390625, + "learning_rate": 0.00017849523831610492, + "loss": 2.468, + "step": 10061 + }, + { + "epoch": 0.43, + "grad_norm": 0.44140625, + "learning_rate": 0.0001784910821996356, + "loss": 2.6638, + "step": 10062 + }, + { + "epoch": 0.43, + "grad_norm": 0.443359375, + "learning_rate": 0.0001784869257299851, + "loss": 2.5525, + "step": 10063 + }, + { + "epoch": 0.43, + "grad_norm": 0.474609375, + "learning_rate": 0.0001784827689071721, + "loss": 2.5635, + "step": 10064 + }, + { + "epoch": 0.43, + "grad_norm": 0.400390625, + "learning_rate": 0.00017847861173121536, + "loss": 2.3436, + "step": 10065 + }, + { + "epoch": 0.43, + "grad_norm": 0.439453125, + "learning_rate": 0.00017847445420213353, + "loss": 2.5949, + "step": 10066 + }, + { + "epoch": 0.43, + "grad_norm": 0.52734375, + "learning_rate": 0.00017847029631994535, + "loss": 2.6112, + "step": 10067 + }, + { + "epoch": 0.43, + "grad_norm": 0.43359375, + "learning_rate": 0.00017846613808466947, + "loss": 2.5264, + "step": 10068 + }, + { + "epoch": 0.43, + "grad_norm": 0.412109375, + "learning_rate": 0.0001784619794963247, + "loss": 2.7276, + "step": 10069 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017845782055492964, + "loss": 2.6617, + "step": 10070 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017845366126050308, + "loss": 2.5913, + "step": 10071 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017844950161306376, + "loss": 2.7996, + "step": 10072 + }, + { + "epoch": 0.43, + "grad_norm": 0.435546875, + "learning_rate": 0.0001784453416126303, + "loss": 2.6586, + "step": 10073 + }, + { + "epoch": 0.43, + "grad_norm": 0.451171875, + "learning_rate": 0.0001784411812592215, + "loss": 2.7085, + "step": 10074 + }, + { + "epoch": 0.43, + "grad_norm": 0.451171875, + "learning_rate": 0.00017843702055285606, + "loss": 2.6288, + "step": 10075 + }, + { + "epoch": 0.43, + "grad_norm": 0.462890625, + "learning_rate": 0.00017843285949355265, + "loss": 2.656, + "step": 10076 + }, + { + "epoch": 0.43, + "grad_norm": 0.490234375, + "learning_rate": 0.00017842869808133008, + "loss": 2.6409, + "step": 10077 + }, + { + "epoch": 0.43, + "grad_norm": 0.515625, + "learning_rate": 0.00017842453631620702, + "loss": 2.6845, + "step": 10078 + }, + { + "epoch": 0.43, + "grad_norm": 0.5078125, + "learning_rate": 0.0001784203741982022, + "loss": 2.6571, + "step": 10079 + }, + { + "epoch": 0.43, + "grad_norm": 0.416015625, + "learning_rate": 0.00017841621172733435, + "loss": 2.578, + "step": 10080 + }, + { + "epoch": 0.43, + "grad_norm": 0.4921875, + "learning_rate": 0.0001784120489036222, + "loss": 2.6568, + "step": 10081 + }, + { + "epoch": 0.43, + "grad_norm": 0.4453125, + "learning_rate": 0.0001784078857270845, + "loss": 2.5187, + "step": 10082 + }, + { + "epoch": 0.43, + "grad_norm": 0.50390625, + "learning_rate": 0.00017840372219773997, + "loss": 2.6648, + "step": 10083 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017839955831560733, + "loss": 2.6793, + "step": 10084 + }, + { + "epoch": 0.43, + "grad_norm": 0.48046875, + "learning_rate": 0.00017839539408070535, + "loss": 2.6359, + "step": 10085 + }, + { + "epoch": 0.43, + "grad_norm": 0.427734375, + "learning_rate": 0.0001783912294930527, + "loss": 2.611, + "step": 10086 + }, + { + "epoch": 0.43, + "grad_norm": 0.41015625, + "learning_rate": 0.00017838706455266824, + "loss": 2.5824, + "step": 10087 + }, + { + "epoch": 0.43, + "grad_norm": 0.474609375, + "learning_rate": 0.0001783828992595706, + "loss": 2.78, + "step": 10088 + }, + { + "epoch": 0.43, + "grad_norm": 0.42578125, + "learning_rate": 0.00017837873361377854, + "loss": 2.4436, + "step": 10089 + }, + { + "epoch": 0.43, + "grad_norm": 0.470703125, + "learning_rate": 0.00017837456761531084, + "loss": 2.7481, + "step": 10090 + }, + { + "epoch": 0.43, + "grad_norm": 0.431640625, + "learning_rate": 0.00017837040126418623, + "loss": 2.5592, + "step": 10091 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.00017836623456042345, + "loss": 2.5483, + "step": 10092 + }, + { + "epoch": 0.43, + "grad_norm": 0.44140625, + "learning_rate": 0.00017836206750404125, + "loss": 2.7957, + "step": 10093 + }, + { + "epoch": 0.43, + "grad_norm": 0.44140625, + "learning_rate": 0.0001783579000950584, + "loss": 2.7251, + "step": 10094 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017835373233349362, + "loss": 2.6322, + "step": 10095 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.0001783495642193657, + "loss": 2.6095, + "step": 10096 + }, + { + "epoch": 0.43, + "grad_norm": 0.427734375, + "learning_rate": 0.00017834539575269337, + "loss": 2.6524, + "step": 10097 + }, + { + "epoch": 0.43, + "grad_norm": 0.5078125, + "learning_rate": 0.00017834122693349537, + "loss": 2.6133, + "step": 10098 + }, + { + "epoch": 0.43, + "grad_norm": 0.46875, + "learning_rate": 0.00017833705776179048, + "loss": 2.5897, + "step": 10099 + }, + { + "epoch": 0.43, + "grad_norm": 0.48828125, + "learning_rate": 0.0001783328882375975, + "loss": 2.6487, + "step": 10100 + }, + { + "epoch": 0.43, + "grad_norm": 0.4609375, + "learning_rate": 0.00017832871836093509, + "loss": 2.583, + "step": 10101 + }, + { + "epoch": 0.43, + "grad_norm": 0.423828125, + "learning_rate": 0.00017832454813182213, + "loss": 2.5859, + "step": 10102 + }, + { + "epoch": 0.43, + "grad_norm": 0.474609375, + "learning_rate": 0.0001783203775502773, + "loss": 2.6707, + "step": 10103 + }, + { + "epoch": 0.43, + "grad_norm": 0.451171875, + "learning_rate": 0.0001783162066163194, + "loss": 2.5926, + "step": 10104 + }, + { + "epoch": 0.43, + "grad_norm": 0.423828125, + "learning_rate": 0.0001783120353299672, + "loss": 2.5322, + "step": 10105 + }, + { + "epoch": 0.43, + "grad_norm": 0.498046875, + "learning_rate": 0.00017830786369123943, + "loss": 2.6092, + "step": 10106 + }, + { + "epoch": 0.43, + "grad_norm": 0.431640625, + "learning_rate": 0.0001783036917001549, + "loss": 2.5782, + "step": 10107 + }, + { + "epoch": 0.43, + "grad_norm": 0.427734375, + "learning_rate": 0.0001782995193567324, + "loss": 2.5008, + "step": 10108 + }, + { + "epoch": 0.43, + "grad_norm": 0.51171875, + "learning_rate": 0.00017829534666099066, + "loss": 2.6067, + "step": 10109 + }, + { + "epoch": 0.43, + "grad_norm": 0.458984375, + "learning_rate": 0.0001782911736129485, + "loss": 2.5846, + "step": 10110 + }, + { + "epoch": 0.43, + "grad_norm": 0.4765625, + "learning_rate": 0.00017828700021262463, + "loss": 2.6403, + "step": 10111 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.0001782828264600379, + "loss": 2.5531, + "step": 10112 + }, + { + "epoch": 0.43, + "grad_norm": 0.4765625, + "learning_rate": 0.00017827865235520705, + "loss": 2.7315, + "step": 10113 + }, + { + "epoch": 0.43, + "grad_norm": 0.470703125, + "learning_rate": 0.00017827447789815084, + "loss": 2.7018, + "step": 10114 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017827030308888812, + "loss": 2.6918, + "step": 10115 + }, + { + "epoch": 0.43, + "grad_norm": 0.48046875, + "learning_rate": 0.00017826612792743764, + "loss": 2.6317, + "step": 10116 + }, + { + "epoch": 0.43, + "grad_norm": 0.458984375, + "learning_rate": 0.00017826195241381817, + "loss": 2.6655, + "step": 10117 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017825777654804857, + "loss": 2.6273, + "step": 10118 + }, + { + "epoch": 0.43, + "grad_norm": 0.400390625, + "learning_rate": 0.0001782536003301475, + "loss": 2.6465, + "step": 10119 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.00017824942376013387, + "loss": 2.6424, + "step": 10120 + }, + { + "epoch": 0.43, + "grad_norm": 0.458984375, + "learning_rate": 0.00017824524683802642, + "loss": 2.58, + "step": 10121 + }, + { + "epoch": 0.43, + "grad_norm": 0.4609375, + "learning_rate": 0.00017824106956384393, + "loss": 2.7267, + "step": 10122 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.00017823689193760526, + "loss": 2.6623, + "step": 10123 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017823271395932917, + "loss": 2.4949, + "step": 10124 + }, + { + "epoch": 0.43, + "grad_norm": 0.451171875, + "learning_rate": 0.00017822853562903443, + "loss": 2.636, + "step": 10125 + }, + { + "epoch": 0.43, + "grad_norm": 0.46875, + "learning_rate": 0.0001782243569467399, + "loss": 2.6611, + "step": 10126 + }, + { + "epoch": 0.43, + "grad_norm": 0.43359375, + "learning_rate": 0.0001782201779124643, + "loss": 2.5265, + "step": 10127 + }, + { + "epoch": 0.43, + "grad_norm": 0.462890625, + "learning_rate": 0.00017821599852622653, + "loss": 2.7087, + "step": 10128 + }, + { + "epoch": 0.43, + "grad_norm": 0.474609375, + "learning_rate": 0.00017821181878804537, + "loss": 2.6397, + "step": 10129 + }, + { + "epoch": 0.43, + "grad_norm": 0.47265625, + "learning_rate": 0.00017820763869793959, + "loss": 2.626, + "step": 10130 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017820345825592798, + "loss": 2.6619, + "step": 10131 + }, + { + "epoch": 0.43, + "grad_norm": 0.4765625, + "learning_rate": 0.0001781992774620294, + "loss": 2.773, + "step": 10132 + }, + { + "epoch": 0.43, + "grad_norm": 0.462890625, + "learning_rate": 0.0001781950963162627, + "loss": 2.7054, + "step": 10133 + }, + { + "epoch": 0.43, + "grad_norm": 0.47265625, + "learning_rate": 0.0001781909148186466, + "loss": 2.7783, + "step": 10134 + }, + { + "epoch": 0.43, + "grad_norm": 0.423828125, + "learning_rate": 0.00017818673296919995, + "loss": 2.5419, + "step": 10135 + }, + { + "epoch": 0.43, + "grad_norm": 0.412109375, + "learning_rate": 0.0001781825507679416, + "loss": 2.5371, + "step": 10136 + }, + { + "epoch": 0.43, + "grad_norm": 0.421875, + "learning_rate": 0.00017817836821489033, + "loss": 2.5076, + "step": 10137 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017817418531006497, + "loss": 2.8742, + "step": 10138 + }, + { + "epoch": 0.43, + "grad_norm": 0.4921875, + "learning_rate": 0.00017817000205348435, + "loss": 2.6307, + "step": 10139 + }, + { + "epoch": 0.43, + "grad_norm": 0.474609375, + "learning_rate": 0.0001781658184451673, + "loss": 2.5472, + "step": 10140 + }, + { + "epoch": 0.43, + "grad_norm": 0.458984375, + "learning_rate": 0.00017816163448513262, + "loss": 2.676, + "step": 10141 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017815745017339918, + "loss": 2.7049, + "step": 10142 + }, + { + "epoch": 0.43, + "grad_norm": 0.439453125, + "learning_rate": 0.00017815326550998576, + "loss": 2.6628, + "step": 10143 + }, + { + "epoch": 0.43, + "grad_norm": 0.4765625, + "learning_rate": 0.0001781490804949112, + "loss": 2.6886, + "step": 10144 + }, + { + "epoch": 0.43, + "grad_norm": 0.41015625, + "learning_rate": 0.0001781448951281943, + "loss": 2.5437, + "step": 10145 + }, + { + "epoch": 0.43, + "grad_norm": 0.48046875, + "learning_rate": 0.000178140709409854, + "loss": 2.59, + "step": 10146 + }, + { + "epoch": 0.43, + "grad_norm": 0.4453125, + "learning_rate": 0.00017813652333990905, + "loss": 2.6563, + "step": 10147 + }, + { + "epoch": 0.43, + "grad_norm": 0.486328125, + "learning_rate": 0.00017813233691837827, + "loss": 2.6479, + "step": 10148 + }, + { + "epoch": 0.43, + "grad_norm": 0.47265625, + "learning_rate": 0.00017812815014528054, + "loss": 2.6993, + "step": 10149 + }, + { + "epoch": 0.43, + "grad_norm": 0.427734375, + "learning_rate": 0.0001781239630206347, + "loss": 2.5744, + "step": 10150 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.00017811977554445956, + "loss": 2.6161, + "step": 10151 + }, + { + "epoch": 0.43, + "grad_norm": 0.482421875, + "learning_rate": 0.000178115587716774, + "loss": 2.6422, + "step": 10152 + }, + { + "epoch": 0.43, + "grad_norm": 0.40625, + "learning_rate": 0.0001781113995375968, + "loss": 2.4282, + "step": 10153 + }, + { + "epoch": 0.43, + "grad_norm": 0.478515625, + "learning_rate": 0.0001781072110069469, + "loss": 2.61, + "step": 10154 + }, + { + "epoch": 0.43, + "grad_norm": 0.51953125, + "learning_rate": 0.00017810302212484307, + "loss": 2.5683, + "step": 10155 + }, + { + "epoch": 0.43, + "grad_norm": 0.49609375, + "learning_rate": 0.0001780988328913042, + "loss": 2.5563, + "step": 10156 + }, + { + "epoch": 0.43, + "grad_norm": 0.48046875, + "learning_rate": 0.00017809464330634911, + "loss": 2.6405, + "step": 10157 + }, + { + "epoch": 0.43, + "grad_norm": 0.50390625, + "learning_rate": 0.00017809045336999665, + "loss": 2.6437, + "step": 10158 + }, + { + "epoch": 0.43, + "grad_norm": 0.44921875, + "learning_rate": 0.00017808626308226573, + "loss": 2.6427, + "step": 10159 + }, + { + "epoch": 0.43, + "grad_norm": 0.4765625, + "learning_rate": 0.00017808207244317514, + "loss": 2.5355, + "step": 10160 + }, + { + "epoch": 0.43, + "grad_norm": 0.51953125, + "learning_rate": 0.00017807788145274373, + "loss": 2.579, + "step": 10161 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.0001780736901109904, + "loss": 2.7328, + "step": 10162 + }, + { + "epoch": 0.43, + "grad_norm": 0.51953125, + "learning_rate": 0.00017806949841793403, + "loss": 2.5795, + "step": 10163 + }, + { + "epoch": 0.43, + "grad_norm": 0.435546875, + "learning_rate": 0.00017806530637359344, + "loss": 2.5366, + "step": 10164 + }, + { + "epoch": 0.43, + "grad_norm": 0.439453125, + "learning_rate": 0.00017806111397798748, + "loss": 2.5455, + "step": 10165 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.00017805692123113505, + "loss": 2.666, + "step": 10166 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.000178052728133055, + "loss": 2.6883, + "step": 10167 + }, + { + "epoch": 0.43, + "grad_norm": 0.447265625, + "learning_rate": 0.00017804853468376624, + "loss": 2.6044, + "step": 10168 + }, + { + "epoch": 0.43, + "grad_norm": 0.4765625, + "learning_rate": 0.00017804434088328753, + "loss": 2.6683, + "step": 10169 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.00017804014673163785, + "loss": 2.6678, + "step": 10170 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.000178035952228836, + "loss": 2.5599, + "step": 10171 + }, + { + "epoch": 0.43, + "grad_norm": 0.48828125, + "learning_rate": 0.0001780317573749009, + "loss": 2.7574, + "step": 10172 + }, + { + "epoch": 0.43, + "grad_norm": 0.515625, + "learning_rate": 0.00017802756216985142, + "loss": 2.7186, + "step": 10173 + }, + { + "epoch": 0.43, + "grad_norm": 0.412109375, + "learning_rate": 0.00017802336661370638, + "loss": 2.6164, + "step": 10174 + }, + { + "epoch": 0.43, + "grad_norm": 0.5234375, + "learning_rate": 0.00017801917070648475, + "loss": 2.5906, + "step": 10175 + }, + { + "epoch": 0.43, + "grad_norm": 0.466796875, + "learning_rate": 0.00017801497444820536, + "loss": 2.578, + "step": 10176 + }, + { + "epoch": 0.43, + "grad_norm": 0.53125, + "learning_rate": 0.0001780107778388871, + "loss": 2.5584, + "step": 10177 + }, + { + "epoch": 0.43, + "grad_norm": 0.42578125, + "learning_rate": 0.0001780065808785488, + "loss": 2.6081, + "step": 10178 + }, + { + "epoch": 0.43, + "grad_norm": 0.435546875, + "learning_rate": 0.00017800238356720946, + "loss": 2.4538, + "step": 10179 + }, + { + "epoch": 0.43, + "grad_norm": 0.5, + "learning_rate": 0.00017799818590488785, + "loss": 2.772, + "step": 10180 + }, + { + "epoch": 0.43, + "grad_norm": 0.451171875, + "learning_rate": 0.00017799398789160296, + "loss": 2.602, + "step": 10181 + }, + { + "epoch": 0.43, + "grad_norm": 0.51171875, + "learning_rate": 0.00017798978952737358, + "loss": 2.7263, + "step": 10182 + }, + { + "epoch": 0.43, + "grad_norm": 0.435546875, + "learning_rate": 0.00017798559081221867, + "loss": 2.5458, + "step": 10183 + }, + { + "epoch": 0.43, + "grad_norm": 0.52734375, + "learning_rate": 0.00017798139174615712, + "loss": 2.7387, + "step": 10184 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.00017797719232920777, + "loss": 2.5668, + "step": 10185 + }, + { + "epoch": 0.43, + "grad_norm": 0.41015625, + "learning_rate": 0.00017797299256138956, + "loss": 2.5062, + "step": 10186 + }, + { + "epoch": 0.43, + "grad_norm": 0.48046875, + "learning_rate": 0.0001779687924427214, + "loss": 2.6615, + "step": 10187 + }, + { + "epoch": 0.43, + "grad_norm": 0.54296875, + "learning_rate": 0.00017796459197322214, + "loss": 2.682, + "step": 10188 + }, + { + "epoch": 0.43, + "grad_norm": 0.44921875, + "learning_rate": 0.00017796039115291072, + "loss": 2.7061, + "step": 10189 + }, + { + "epoch": 0.43, + "grad_norm": 0.4453125, + "learning_rate": 0.00017795618998180603, + "loss": 2.5209, + "step": 10190 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.000177951988459927, + "loss": 2.6528, + "step": 10191 + }, + { + "epoch": 0.43, + "grad_norm": 0.47265625, + "learning_rate": 0.0001779477865872925, + "loss": 2.6912, + "step": 10192 + }, + { + "epoch": 0.43, + "grad_norm": 0.43359375, + "learning_rate": 0.0001779435843639214, + "loss": 2.5744, + "step": 10193 + }, + { + "epoch": 0.43, + "grad_norm": 0.4296875, + "learning_rate": 0.00017793938178983272, + "loss": 2.5161, + "step": 10194 + }, + { + "epoch": 0.43, + "grad_norm": 0.435546875, + "learning_rate": 0.00017793517886504528, + "loss": 2.5606, + "step": 10195 + }, + { + "epoch": 0.43, + "grad_norm": 0.453125, + "learning_rate": 0.000177930975589578, + "loss": 2.5444, + "step": 10196 + }, + { + "epoch": 0.43, + "grad_norm": 0.45703125, + "learning_rate": 0.00017792677196344983, + "loss": 2.6377, + "step": 10197 + }, + { + "epoch": 0.43, + "grad_norm": 0.455078125, + "learning_rate": 0.00017792256798667964, + "loss": 2.5365, + "step": 10198 + }, + { + "epoch": 0.43, + "grad_norm": 0.50390625, + "learning_rate": 0.0001779183636592864, + "loss": 2.6044, + "step": 10199 + }, + { + "epoch": 0.43, + "grad_norm": 0.34375, + "learning_rate": 0.00017791415898128903, + "loss": 2.6547, + "step": 10200 + }, + { + "epoch": 0.43, + "grad_norm": 0.451171875, + "learning_rate": 0.00017790995395270635, + "loss": 2.71, + "step": 10201 + }, + { + "epoch": 0.43, + "grad_norm": 0.427734375, + "learning_rate": 0.0001779057485735574, + "loss": 2.5994, + "step": 10202 + }, + { + "epoch": 0.43, + "grad_norm": 0.470703125, + "learning_rate": 0.000177901542843861, + "loss": 2.5882, + "step": 10203 + }, + { + "epoch": 0.43, + "grad_norm": 0.4140625, + "learning_rate": 0.00017789733676363618, + "loss": 2.5066, + "step": 10204 + }, + { + "epoch": 0.43, + "grad_norm": 0.42578125, + "learning_rate": 0.0001778931303329018, + "loss": 2.664, + "step": 10205 + }, + { + "epoch": 0.43, + "grad_norm": 0.46484375, + "learning_rate": 0.0001778889235516768, + "loss": 2.7076, + "step": 10206 + }, + { + "epoch": 0.43, + "grad_norm": 0.478515625, + "learning_rate": 0.00017788471641998007, + "loss": 2.5695, + "step": 10207 + }, + { + "epoch": 0.43, + "grad_norm": 0.474609375, + "learning_rate": 0.00017788050893783062, + "loss": 2.4975, + "step": 10208 + }, + { + "epoch": 0.44, + "grad_norm": 0.4609375, + "learning_rate": 0.00017787630110524733, + "loss": 2.6428, + "step": 10209 + }, + { + "epoch": 0.44, + "grad_norm": 0.447265625, + "learning_rate": 0.00017787209292224917, + "loss": 2.4825, + "step": 10210 + }, + { + "epoch": 0.44, + "grad_norm": 0.435546875, + "learning_rate": 0.00017786788438885505, + "loss": 2.6143, + "step": 10211 + }, + { + "epoch": 0.44, + "grad_norm": 0.435546875, + "learning_rate": 0.00017786367550508386, + "loss": 2.706, + "step": 10212 + }, + { + "epoch": 0.44, + "grad_norm": 0.55859375, + "learning_rate": 0.00017785946627095462, + "loss": 2.7193, + "step": 10213 + }, + { + "epoch": 0.44, + "grad_norm": 0.4296875, + "learning_rate": 0.0001778552566864862, + "loss": 2.4809, + "step": 10214 + }, + { + "epoch": 0.44, + "grad_norm": 0.484375, + "learning_rate": 0.00017785104675169765, + "loss": 2.663, + "step": 10215 + }, + { + "epoch": 0.44, + "grad_norm": 0.431640625, + "learning_rate": 0.00017784683646660782, + "loss": 2.628, + "step": 10216 + }, + { + "epoch": 0.44, + "grad_norm": 0.46484375, + "learning_rate": 0.0001778426258312356, + "loss": 2.6738, + "step": 10217 + }, + { + "epoch": 0.44, + "grad_norm": 0.419921875, + "learning_rate": 0.00017783841484560008, + "loss": 2.6137, + "step": 10218 + }, + { + "epoch": 0.44, + "grad_norm": 0.484375, + "learning_rate": 0.00017783420350972012, + "loss": 2.6568, + "step": 10219 + }, + { + "epoch": 0.44, + "grad_norm": 0.439453125, + "learning_rate": 0.00017782999182361473, + "loss": 2.5344, + "step": 10220 + }, + { + "epoch": 0.44, + "grad_norm": 0.42578125, + "learning_rate": 0.00017782577978730276, + "loss": 2.6811, + "step": 10221 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017782156740080325, + "loss": 2.6678, + "step": 10222 + }, + { + "epoch": 0.44, + "grad_norm": 0.404296875, + "learning_rate": 0.00017781735466413515, + "loss": 2.613, + "step": 10223 + }, + { + "epoch": 0.44, + "grad_norm": 0.4453125, + "learning_rate": 0.00017781314157731733, + "loss": 2.5851, + "step": 10224 + }, + { + "epoch": 0.44, + "grad_norm": 0.48828125, + "learning_rate": 0.00017780892814036887, + "loss": 2.7613, + "step": 10225 + }, + { + "epoch": 0.44, + "grad_norm": 0.421875, + "learning_rate": 0.00017780471435330865, + "loss": 2.7542, + "step": 10226 + }, + { + "epoch": 0.44, + "grad_norm": 0.498046875, + "learning_rate": 0.00017780050021615567, + "loss": 2.7632, + "step": 10227 + }, + { + "epoch": 0.44, + "grad_norm": 0.48828125, + "learning_rate": 0.00017779628572892883, + "loss": 2.7021, + "step": 10228 + }, + { + "epoch": 0.44, + "grad_norm": 0.451171875, + "learning_rate": 0.00017779207089164715, + "loss": 2.6247, + "step": 10229 + }, + { + "epoch": 0.44, + "grad_norm": 0.4609375, + "learning_rate": 0.00017778785570432958, + "loss": 2.6172, + "step": 10230 + }, + { + "epoch": 0.44, + "grad_norm": 0.470703125, + "learning_rate": 0.0001777836401669951, + "loss": 2.6124, + "step": 10231 + }, + { + "epoch": 0.44, + "grad_norm": 0.453125, + "learning_rate": 0.00017777942427966268, + "loss": 2.5715, + "step": 10232 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.00017777520804235125, + "loss": 2.6392, + "step": 10233 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.0001777709914550798, + "loss": 2.6426, + "step": 10234 + }, + { + "epoch": 0.44, + "grad_norm": 0.44140625, + "learning_rate": 0.00017776677451786737, + "loss": 2.6358, + "step": 10235 + }, + { + "epoch": 0.44, + "grad_norm": 0.49609375, + "learning_rate": 0.00017776255723073278, + "loss": 2.6139, + "step": 10236 + }, + { + "epoch": 0.44, + "grad_norm": 0.486328125, + "learning_rate": 0.00017775833959369518, + "loss": 2.7216, + "step": 10237 + }, + { + "epoch": 0.44, + "grad_norm": 0.458984375, + "learning_rate": 0.00017775412160677342, + "loss": 2.789, + "step": 10238 + }, + { + "epoch": 0.44, + "grad_norm": 0.4375, + "learning_rate": 0.00017774990326998658, + "loss": 2.5012, + "step": 10239 + }, + { + "epoch": 0.44, + "grad_norm": 0.4453125, + "learning_rate": 0.00017774568458335353, + "loss": 2.7182, + "step": 10240 + }, + { + "epoch": 0.44, + "grad_norm": 0.4609375, + "learning_rate": 0.00017774146554689336, + "loss": 2.638, + "step": 10241 + }, + { + "epoch": 0.44, + "grad_norm": 0.478515625, + "learning_rate": 0.000177737246160625, + "loss": 2.6514, + "step": 10242 + }, + { + "epoch": 0.44, + "grad_norm": 0.431640625, + "learning_rate": 0.0001777330264245674, + "loss": 2.6403, + "step": 10243 + }, + { + "epoch": 0.44, + "grad_norm": 0.45703125, + "learning_rate": 0.00017772880633873964, + "loss": 2.5708, + "step": 10244 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.0001777245859031606, + "loss": 2.5832, + "step": 10245 + }, + { + "epoch": 0.44, + "grad_norm": 0.4453125, + "learning_rate": 0.0001777203651178494, + "loss": 2.6078, + "step": 10246 + }, + { + "epoch": 0.44, + "grad_norm": 0.439453125, + "learning_rate": 0.0001777161439828249, + "loss": 2.6247, + "step": 10247 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.00017771192249810616, + "loss": 2.7099, + "step": 10248 + }, + { + "epoch": 0.44, + "grad_norm": 0.447265625, + "learning_rate": 0.00017770770066371223, + "loss": 2.6517, + "step": 10249 + }, + { + "epoch": 0.44, + "grad_norm": 0.474609375, + "learning_rate": 0.000177703478479662, + "loss": 2.4231, + "step": 10250 + }, + { + "epoch": 0.44, + "grad_norm": 0.48046875, + "learning_rate": 0.00017769925594597453, + "loss": 2.7317, + "step": 10251 + }, + { + "epoch": 0.44, + "grad_norm": 0.42578125, + "learning_rate": 0.00017769503306266877, + "loss": 2.4526, + "step": 10252 + }, + { + "epoch": 0.44, + "grad_norm": 0.455078125, + "learning_rate": 0.00017769080982976378, + "loss": 2.5465, + "step": 10253 + }, + { + "epoch": 0.44, + "grad_norm": 0.443359375, + "learning_rate": 0.00017768658624727854, + "loss": 2.6511, + "step": 10254 + }, + { + "epoch": 0.44, + "grad_norm": 0.494140625, + "learning_rate": 0.00017768236231523207, + "loss": 2.6799, + "step": 10255 + }, + { + "epoch": 0.44, + "grad_norm": 0.443359375, + "learning_rate": 0.00017767813803364331, + "loss": 2.5936, + "step": 10256 + }, + { + "epoch": 0.44, + "grad_norm": 0.43359375, + "learning_rate": 0.00017767391340253136, + "loss": 2.6572, + "step": 10257 + }, + { + "epoch": 0.44, + "grad_norm": 0.431640625, + "learning_rate": 0.00017766968842191516, + "loss": 2.5832, + "step": 10258 + }, + { + "epoch": 0.44, + "grad_norm": 0.443359375, + "learning_rate": 0.00017766546309181378, + "loss": 2.6389, + "step": 10259 + }, + { + "epoch": 0.44, + "grad_norm": 0.435546875, + "learning_rate": 0.00017766123741224616, + "loss": 2.6261, + "step": 10260 + }, + { + "epoch": 0.44, + "grad_norm": 0.380859375, + "learning_rate": 0.0001776570113832314, + "loss": 2.5423, + "step": 10261 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.00017765278500478843, + "loss": 2.4778, + "step": 10262 + }, + { + "epoch": 0.44, + "grad_norm": 0.48046875, + "learning_rate": 0.00017764855827693632, + "loss": 2.6108, + "step": 10263 + }, + { + "epoch": 0.44, + "grad_norm": 0.41015625, + "learning_rate": 0.00017764433119969408, + "loss": 2.6127, + "step": 10264 + }, + { + "epoch": 0.44, + "grad_norm": 0.431640625, + "learning_rate": 0.0001776401037730807, + "loss": 2.6332, + "step": 10265 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017763587599711525, + "loss": 2.5144, + "step": 10266 + }, + { + "epoch": 0.44, + "grad_norm": 0.421875, + "learning_rate": 0.00017763164787181675, + "loss": 2.6203, + "step": 10267 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017762741939720417, + "loss": 2.5488, + "step": 10268 + }, + { + "epoch": 0.44, + "grad_norm": 0.44140625, + "learning_rate": 0.00017762319057329658, + "loss": 2.549, + "step": 10269 + }, + { + "epoch": 0.44, + "grad_norm": 0.453125, + "learning_rate": 0.000177618961400113, + "loss": 2.5475, + "step": 10270 + }, + { + "epoch": 0.44, + "grad_norm": 0.4609375, + "learning_rate": 0.00017761473187767246, + "loss": 2.6263, + "step": 10271 + }, + { + "epoch": 0.44, + "grad_norm": 0.435546875, + "learning_rate": 0.00017761050200599401, + "loss": 2.5939, + "step": 10272 + }, + { + "epoch": 0.44, + "grad_norm": 0.41796875, + "learning_rate": 0.00017760627178509665, + "loss": 2.5818, + "step": 10273 + }, + { + "epoch": 0.44, + "grad_norm": 0.458984375, + "learning_rate": 0.00017760204121499942, + "loss": 2.6253, + "step": 10274 + }, + { + "epoch": 0.44, + "grad_norm": 0.5234375, + "learning_rate": 0.00017759781029572138, + "loss": 2.6582, + "step": 10275 + }, + { + "epoch": 0.44, + "grad_norm": 0.4921875, + "learning_rate": 0.00017759357902728153, + "loss": 2.5137, + "step": 10276 + }, + { + "epoch": 0.44, + "grad_norm": 0.5234375, + "learning_rate": 0.00017758934740969896, + "loss": 2.6445, + "step": 10277 + }, + { + "epoch": 0.44, + "grad_norm": 0.490234375, + "learning_rate": 0.0001775851154429927, + "loss": 2.6409, + "step": 10278 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017758088312718173, + "loss": 2.6413, + "step": 10279 + }, + { + "epoch": 0.44, + "grad_norm": 0.453125, + "learning_rate": 0.00017757665046228514, + "loss": 2.6212, + "step": 10280 + }, + { + "epoch": 0.44, + "grad_norm": 0.447265625, + "learning_rate": 0.00017757241744832198, + "loss": 2.6903, + "step": 10281 + }, + { + "epoch": 0.44, + "grad_norm": 0.42578125, + "learning_rate": 0.0001775681840853113, + "loss": 2.5442, + "step": 10282 + }, + { + "epoch": 0.44, + "grad_norm": 0.498046875, + "learning_rate": 0.00017756395037327215, + "loss": 2.8234, + "step": 10283 + }, + { + "epoch": 0.44, + "grad_norm": 0.43359375, + "learning_rate": 0.00017755971631222358, + "loss": 2.5723, + "step": 10284 + }, + { + "epoch": 0.44, + "grad_norm": 0.45703125, + "learning_rate": 0.0001775554819021846, + "loss": 2.5364, + "step": 10285 + }, + { + "epoch": 0.44, + "grad_norm": 0.431640625, + "learning_rate": 0.0001775512471431743, + "loss": 2.5378, + "step": 10286 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.00017754701203521173, + "loss": 2.5344, + "step": 10287 + }, + { + "epoch": 0.44, + "grad_norm": 0.51953125, + "learning_rate": 0.00017754277657831594, + "loss": 2.7815, + "step": 10288 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.000177538540772506, + "loss": 2.5285, + "step": 10289 + }, + { + "epoch": 0.44, + "grad_norm": 0.470703125, + "learning_rate": 0.00017753430461780098, + "loss": 2.5699, + "step": 10290 + }, + { + "epoch": 0.44, + "grad_norm": 0.439453125, + "learning_rate": 0.00017753006811421993, + "loss": 2.5452, + "step": 10291 + }, + { + "epoch": 0.44, + "grad_norm": 0.498046875, + "learning_rate": 0.0001775258312617819, + "loss": 2.6906, + "step": 10292 + }, + { + "epoch": 0.44, + "grad_norm": 0.46484375, + "learning_rate": 0.00017752159406050593, + "loss": 2.6464, + "step": 10293 + }, + { + "epoch": 0.44, + "grad_norm": 0.478515625, + "learning_rate": 0.00017751735651041112, + "loss": 2.5956, + "step": 10294 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.00017751311861151657, + "loss": 2.6159, + "step": 10295 + }, + { + "epoch": 0.44, + "grad_norm": 0.490234375, + "learning_rate": 0.0001775088803638413, + "loss": 2.6801, + "step": 10296 + }, + { + "epoch": 0.44, + "grad_norm": 0.5234375, + "learning_rate": 0.0001775046417674044, + "loss": 2.8263, + "step": 10297 + }, + { + "epoch": 0.44, + "grad_norm": 0.494140625, + "learning_rate": 0.0001775004028222249, + "loss": 2.6393, + "step": 10298 + }, + { + "epoch": 0.44, + "grad_norm": 0.470703125, + "learning_rate": 0.00017749616352832196, + "loss": 2.6522, + "step": 10299 + }, + { + "epoch": 0.44, + "grad_norm": 0.462890625, + "learning_rate": 0.0001774919238857146, + "loss": 2.6676, + "step": 10300 + }, + { + "epoch": 0.44, + "grad_norm": 0.482421875, + "learning_rate": 0.00017748768389442188, + "loss": 2.6764, + "step": 10301 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.00017748344355446292, + "loss": 2.7229, + "step": 10302 + }, + { + "epoch": 0.44, + "grad_norm": 0.482421875, + "learning_rate": 0.00017747920286585676, + "loss": 2.6202, + "step": 10303 + }, + { + "epoch": 0.44, + "grad_norm": 0.447265625, + "learning_rate": 0.0001774749618286225, + "loss": 2.7641, + "step": 10304 + }, + { + "epoch": 0.44, + "grad_norm": 0.458984375, + "learning_rate": 0.00017747072044277924, + "loss": 2.6104, + "step": 10305 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.0001774664787083461, + "loss": 2.5064, + "step": 10306 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.00017746223662534202, + "loss": 2.5066, + "step": 10307 + }, + { + "epoch": 0.44, + "grad_norm": 0.50390625, + "learning_rate": 0.00017745799419378625, + "loss": 2.5969, + "step": 10308 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.00017745375141369778, + "loss": 2.6307, + "step": 10309 + }, + { + "epoch": 0.44, + "grad_norm": 0.54296875, + "learning_rate": 0.00017744950828509575, + "loss": 2.646, + "step": 10310 + }, + { + "epoch": 0.44, + "grad_norm": 0.416015625, + "learning_rate": 0.00017744526480799924, + "loss": 2.5485, + "step": 10311 + }, + { + "epoch": 0.44, + "grad_norm": 0.439453125, + "learning_rate": 0.00017744102098242733, + "loss": 2.6178, + "step": 10312 + }, + { + "epoch": 0.44, + "grad_norm": 0.388671875, + "learning_rate": 0.00017743677680839915, + "loss": 2.5271, + "step": 10313 + }, + { + "epoch": 0.44, + "grad_norm": 0.439453125, + "learning_rate": 0.00017743253228593372, + "loss": 2.6483, + "step": 10314 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017742828741505027, + "loss": 2.6238, + "step": 10315 + }, + { + "epoch": 0.44, + "grad_norm": 0.427734375, + "learning_rate": 0.00017742404219576777, + "loss": 2.6533, + "step": 10316 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.0001774197966281054, + "loss": 2.7319, + "step": 10317 + }, + { + "epoch": 0.44, + "grad_norm": 0.470703125, + "learning_rate": 0.00017741555071208221, + "loss": 2.6808, + "step": 10318 + }, + { + "epoch": 0.44, + "grad_norm": 0.45703125, + "learning_rate": 0.00017741130444771737, + "loss": 2.649, + "step": 10319 + }, + { + "epoch": 0.44, + "grad_norm": 0.494140625, + "learning_rate": 0.00017740705783502993, + "loss": 2.7049, + "step": 10320 + }, + { + "epoch": 0.44, + "grad_norm": 0.4609375, + "learning_rate": 0.000177402810874039, + "loss": 2.7126, + "step": 10321 + }, + { + "epoch": 0.44, + "grad_norm": 0.474609375, + "learning_rate": 0.00017739856356476376, + "loss": 2.5839, + "step": 10322 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017739431590722325, + "loss": 2.6067, + "step": 10323 + }, + { + "epoch": 0.44, + "grad_norm": 0.48828125, + "learning_rate": 0.00017739006790143657, + "loss": 2.6198, + "step": 10324 + }, + { + "epoch": 0.44, + "grad_norm": 0.44140625, + "learning_rate": 0.0001773858195474229, + "loss": 2.5647, + "step": 10325 + }, + { + "epoch": 0.44, + "grad_norm": 0.482421875, + "learning_rate": 0.0001773815708452013, + "loss": 2.5761, + "step": 10326 + }, + { + "epoch": 0.44, + "grad_norm": 0.455078125, + "learning_rate": 0.00017737732179479095, + "loss": 2.596, + "step": 10327 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017737307239621092, + "loss": 2.602, + "step": 10328 + }, + { + "epoch": 0.44, + "grad_norm": 0.52734375, + "learning_rate": 0.0001773688226494803, + "loss": 2.6653, + "step": 10329 + }, + { + "epoch": 0.44, + "grad_norm": 0.51171875, + "learning_rate": 0.0001773645725546183, + "loss": 2.746, + "step": 10330 + }, + { + "epoch": 0.44, + "grad_norm": 0.50390625, + "learning_rate": 0.00017736032211164394, + "loss": 2.6597, + "step": 10331 + }, + { + "epoch": 0.44, + "grad_norm": 0.41796875, + "learning_rate": 0.00017735607132057647, + "loss": 2.5484, + "step": 10332 + }, + { + "epoch": 0.44, + "grad_norm": 0.4609375, + "learning_rate": 0.0001773518201814349, + "loss": 2.6244, + "step": 10333 + }, + { + "epoch": 0.44, + "grad_norm": 0.443359375, + "learning_rate": 0.00017734756869423844, + "loss": 2.5494, + "step": 10334 + }, + { + "epoch": 0.44, + "grad_norm": 0.4453125, + "learning_rate": 0.00017734331685900615, + "loss": 2.8086, + "step": 10335 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.00017733906467575725, + "loss": 2.5705, + "step": 10336 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.0001773348121445108, + "loss": 2.6024, + "step": 10337 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017733055926528595, + "loss": 2.6164, + "step": 10338 + }, + { + "epoch": 0.44, + "grad_norm": 0.45703125, + "learning_rate": 0.00017732630603810185, + "loss": 2.752, + "step": 10339 + }, + { + "epoch": 0.44, + "grad_norm": 0.42578125, + "learning_rate": 0.00017732205246297761, + "loss": 2.5118, + "step": 10340 + }, + { + "epoch": 0.44, + "grad_norm": 0.498046875, + "learning_rate": 0.00017731779853993242, + "loss": 2.6669, + "step": 10341 + }, + { + "epoch": 0.44, + "grad_norm": 0.470703125, + "learning_rate": 0.00017731354426898542, + "loss": 2.628, + "step": 10342 + }, + { + "epoch": 0.44, + "grad_norm": 0.419921875, + "learning_rate": 0.00017730928965015567, + "loss": 2.4218, + "step": 10343 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.00017730503468346237, + "loss": 2.5618, + "step": 10344 + }, + { + "epoch": 0.44, + "grad_norm": 0.4921875, + "learning_rate": 0.0001773007793689247, + "loss": 2.4429, + "step": 10345 + }, + { + "epoch": 0.44, + "grad_norm": 0.478515625, + "learning_rate": 0.00017729652370656175, + "loss": 2.5815, + "step": 10346 + }, + { + "epoch": 0.44, + "grad_norm": 0.51171875, + "learning_rate": 0.0001772922676963927, + "loss": 2.6498, + "step": 10347 + }, + { + "epoch": 0.44, + "grad_norm": 0.419921875, + "learning_rate": 0.0001772880113384367, + "loss": 2.6661, + "step": 10348 + }, + { + "epoch": 0.44, + "grad_norm": 0.427734375, + "learning_rate": 0.0001772837546327129, + "loss": 2.5138, + "step": 10349 + }, + { + "epoch": 0.44, + "grad_norm": 0.408203125, + "learning_rate": 0.00017727949757924045, + "loss": 2.6564, + "step": 10350 + }, + { + "epoch": 0.44, + "grad_norm": 0.458984375, + "learning_rate": 0.0001772752401780385, + "loss": 2.7459, + "step": 10351 + }, + { + "epoch": 0.44, + "grad_norm": 0.5, + "learning_rate": 0.00017727098242912618, + "loss": 2.6659, + "step": 10352 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.0001772667243325227, + "loss": 2.6055, + "step": 10353 + }, + { + "epoch": 0.44, + "grad_norm": 0.51171875, + "learning_rate": 0.00017726246588824722, + "loss": 2.6085, + "step": 10354 + }, + { + "epoch": 0.44, + "grad_norm": 0.478515625, + "learning_rate": 0.00017725820709631887, + "loss": 2.6692, + "step": 10355 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.0001772539479567568, + "loss": 2.5697, + "step": 10356 + }, + { + "epoch": 0.44, + "grad_norm": 0.45703125, + "learning_rate": 0.0001772496884695802, + "loss": 2.6115, + "step": 10357 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.00017724542863480823, + "loss": 2.6973, + "step": 10358 + }, + { + "epoch": 0.44, + "grad_norm": 0.4921875, + "learning_rate": 0.0001772411684524601, + "loss": 2.7518, + "step": 10359 + }, + { + "epoch": 0.44, + "grad_norm": 0.404296875, + "learning_rate": 0.00017723690792255492, + "loss": 2.704, + "step": 10360 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.00017723264704511187, + "loss": 2.6899, + "step": 10361 + }, + { + "epoch": 0.44, + "grad_norm": 0.38671875, + "learning_rate": 0.00017722838582015014, + "loss": 2.5306, + "step": 10362 + }, + { + "epoch": 0.44, + "grad_norm": 0.443359375, + "learning_rate": 0.00017722412424768887, + "loss": 2.6786, + "step": 10363 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.0001772198623277473, + "loss": 2.6312, + "step": 10364 + }, + { + "epoch": 0.44, + "grad_norm": 0.46484375, + "learning_rate": 0.00017721560006034456, + "loss": 2.6388, + "step": 10365 + }, + { + "epoch": 0.44, + "grad_norm": 0.447265625, + "learning_rate": 0.00017721133744549983, + "loss": 2.482, + "step": 10366 + }, + { + "epoch": 0.44, + "grad_norm": 0.45703125, + "learning_rate": 0.0001772070744832323, + "loss": 2.5908, + "step": 10367 + }, + { + "epoch": 0.44, + "grad_norm": 0.443359375, + "learning_rate": 0.00017720281117356116, + "loss": 2.64, + "step": 10368 + }, + { + "epoch": 0.44, + "grad_norm": 0.42578125, + "learning_rate": 0.0001771985475165056, + "loss": 2.6461, + "step": 10369 + }, + { + "epoch": 0.44, + "grad_norm": 0.41015625, + "learning_rate": 0.00017719428351208478, + "loss": 2.5866, + "step": 10370 + }, + { + "epoch": 0.44, + "grad_norm": 0.447265625, + "learning_rate": 0.00017719001916031787, + "loss": 2.4881, + "step": 10371 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.0001771857544612241, + "loss": 2.5495, + "step": 10372 + }, + { + "epoch": 0.44, + "grad_norm": 0.494140625, + "learning_rate": 0.00017718148941482263, + "loss": 2.7069, + "step": 10373 + }, + { + "epoch": 0.44, + "grad_norm": 0.478515625, + "learning_rate": 0.0001771772240211327, + "loss": 2.778, + "step": 10374 + }, + { + "epoch": 0.44, + "grad_norm": 0.4453125, + "learning_rate": 0.0001771729582801734, + "loss": 2.5985, + "step": 10375 + }, + { + "epoch": 0.44, + "grad_norm": 0.39453125, + "learning_rate": 0.00017716869219196405, + "loss": 2.5967, + "step": 10376 + }, + { + "epoch": 0.44, + "grad_norm": 0.396484375, + "learning_rate": 0.00017716442575652379, + "loss": 2.6158, + "step": 10377 + }, + { + "epoch": 0.44, + "grad_norm": 0.46484375, + "learning_rate": 0.0001771601589738718, + "loss": 2.6993, + "step": 10378 + }, + { + "epoch": 0.44, + "grad_norm": 0.431640625, + "learning_rate": 0.0001771558918440273, + "loss": 2.5459, + "step": 10379 + }, + { + "epoch": 0.44, + "grad_norm": 0.447265625, + "learning_rate": 0.00017715162436700944, + "loss": 2.5053, + "step": 10380 + }, + { + "epoch": 0.44, + "grad_norm": 0.484375, + "learning_rate": 0.00017714735654283752, + "loss": 2.6179, + "step": 10381 + }, + { + "epoch": 0.44, + "grad_norm": 0.5, + "learning_rate": 0.00017714308837153068, + "loss": 2.6422, + "step": 10382 + }, + { + "epoch": 0.44, + "grad_norm": 0.458984375, + "learning_rate": 0.0001771388198531081, + "loss": 2.6243, + "step": 10383 + }, + { + "epoch": 0.44, + "grad_norm": 0.478515625, + "learning_rate": 0.00017713455098758907, + "loss": 2.6389, + "step": 10384 + }, + { + "epoch": 0.44, + "grad_norm": 0.451171875, + "learning_rate": 0.00017713028177499272, + "loss": 2.7386, + "step": 10385 + }, + { + "epoch": 0.44, + "grad_norm": 0.48828125, + "learning_rate": 0.00017712601221533828, + "loss": 2.611, + "step": 10386 + }, + { + "epoch": 0.44, + "grad_norm": 0.5, + "learning_rate": 0.000177121742308645, + "loss": 2.6868, + "step": 10387 + }, + { + "epoch": 0.44, + "grad_norm": 0.43359375, + "learning_rate": 0.00017711747205493205, + "loss": 2.4496, + "step": 10388 + }, + { + "epoch": 0.44, + "grad_norm": 0.439453125, + "learning_rate": 0.0001771132014542187, + "loss": 2.6165, + "step": 10389 + }, + { + "epoch": 0.44, + "grad_norm": 0.439453125, + "learning_rate": 0.00017710893050652406, + "loss": 2.6548, + "step": 10390 + }, + { + "epoch": 0.44, + "grad_norm": 0.484375, + "learning_rate": 0.00017710465921186744, + "loss": 2.6377, + "step": 10391 + }, + { + "epoch": 0.44, + "grad_norm": 0.443359375, + "learning_rate": 0.000177100387570268, + "loss": 2.5327, + "step": 10392 + }, + { + "epoch": 0.44, + "grad_norm": 0.431640625, + "learning_rate": 0.00017709611558174503, + "loss": 2.5468, + "step": 10393 + }, + { + "epoch": 0.44, + "grad_norm": 0.5078125, + "learning_rate": 0.00017709184324631773, + "loss": 2.5695, + "step": 10394 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.00017708757056400528, + "loss": 2.7567, + "step": 10395 + }, + { + "epoch": 0.44, + "grad_norm": 0.4609375, + "learning_rate": 0.00017708329753482693, + "loss": 2.6221, + "step": 10396 + }, + { + "epoch": 0.44, + "grad_norm": 0.48828125, + "learning_rate": 0.00017707902415880193, + "loss": 2.557, + "step": 10397 + }, + { + "epoch": 0.44, + "grad_norm": 0.43359375, + "learning_rate": 0.00017707475043594948, + "loss": 2.6222, + "step": 10398 + }, + { + "epoch": 0.44, + "grad_norm": 0.462890625, + "learning_rate": 0.00017707047636628885, + "loss": 2.5695, + "step": 10399 + }, + { + "epoch": 0.44, + "grad_norm": 0.427734375, + "learning_rate": 0.0001770662019498392, + "loss": 2.6258, + "step": 10400 + }, + { + "epoch": 0.44, + "grad_norm": 0.419921875, + "learning_rate": 0.00017706192718661986, + "loss": 2.6282, + "step": 10401 + }, + { + "epoch": 0.44, + "grad_norm": 0.39453125, + "learning_rate": 0.00017705765207664994, + "loss": 2.6008, + "step": 10402 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.00017705337661994878, + "loss": 2.4811, + "step": 10403 + }, + { + "epoch": 0.44, + "grad_norm": 0.490234375, + "learning_rate": 0.00017704910081653557, + "loss": 2.7203, + "step": 10404 + }, + { + "epoch": 0.44, + "grad_norm": 0.462890625, + "learning_rate": 0.0001770448246664296, + "loss": 2.5442, + "step": 10405 + }, + { + "epoch": 0.44, + "grad_norm": 0.443359375, + "learning_rate": 0.00017704054816965005, + "loss": 2.7089, + "step": 10406 + }, + { + "epoch": 0.44, + "grad_norm": 0.451171875, + "learning_rate": 0.0001770362713262162, + "loss": 2.6536, + "step": 10407 + }, + { + "epoch": 0.44, + "grad_norm": 0.45703125, + "learning_rate": 0.00017703199413614728, + "loss": 2.6182, + "step": 10408 + }, + { + "epoch": 0.44, + "grad_norm": 0.52734375, + "learning_rate": 0.00017702771659946253, + "loss": 2.7571, + "step": 10409 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.00017702343871618122, + "loss": 2.5056, + "step": 10410 + }, + { + "epoch": 0.44, + "grad_norm": 0.431640625, + "learning_rate": 0.00017701916048632258, + "loss": 2.5504, + "step": 10411 + }, + { + "epoch": 0.44, + "grad_norm": 0.54296875, + "learning_rate": 0.00017701488190990589, + "loss": 2.5591, + "step": 10412 + }, + { + "epoch": 0.44, + "grad_norm": 0.498046875, + "learning_rate": 0.00017701060298695032, + "loss": 2.7633, + "step": 10413 + }, + { + "epoch": 0.44, + "grad_norm": 0.490234375, + "learning_rate": 0.00017700632371747522, + "loss": 2.6742, + "step": 10414 + }, + { + "epoch": 0.44, + "grad_norm": 0.51171875, + "learning_rate": 0.0001770020441014998, + "loss": 2.7476, + "step": 10415 + }, + { + "epoch": 0.44, + "grad_norm": 0.470703125, + "learning_rate": 0.00017699776413904333, + "loss": 2.529, + "step": 10416 + }, + { + "epoch": 0.44, + "grad_norm": 0.44140625, + "learning_rate": 0.00017699348383012504, + "loss": 2.7258, + "step": 10417 + }, + { + "epoch": 0.44, + "grad_norm": 0.462890625, + "learning_rate": 0.00017698920317476419, + "loss": 2.549, + "step": 10418 + }, + { + "epoch": 0.44, + "grad_norm": 0.46484375, + "learning_rate": 0.0001769849221729801, + "loss": 2.6441, + "step": 10419 + }, + { + "epoch": 0.44, + "grad_norm": 0.486328125, + "learning_rate": 0.000176980640824792, + "loss": 2.7052, + "step": 10420 + }, + { + "epoch": 0.44, + "grad_norm": 0.43359375, + "learning_rate": 0.00017697635913021913, + "loss": 2.6058, + "step": 10421 + }, + { + "epoch": 0.44, + "grad_norm": 0.458984375, + "learning_rate": 0.00017697207708928077, + "loss": 2.7241, + "step": 10422 + }, + { + "epoch": 0.44, + "grad_norm": 0.404296875, + "learning_rate": 0.00017696779470199623, + "loss": 2.5182, + "step": 10423 + }, + { + "epoch": 0.44, + "grad_norm": 0.44140625, + "learning_rate": 0.0001769635119683847, + "loss": 2.6976, + "step": 10424 + }, + { + "epoch": 0.44, + "grad_norm": 0.44921875, + "learning_rate": 0.00017695922888846552, + "loss": 2.6387, + "step": 10425 + }, + { + "epoch": 0.44, + "grad_norm": 0.421875, + "learning_rate": 0.00017695494546225794, + "loss": 2.4687, + "step": 10426 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.0001769506616897812, + "loss": 2.6818, + "step": 10427 + }, + { + "epoch": 0.44, + "grad_norm": 0.439453125, + "learning_rate": 0.00017694637757105462, + "loss": 2.6451, + "step": 10428 + }, + { + "epoch": 0.44, + "grad_norm": 0.4765625, + "learning_rate": 0.00017694209310609743, + "loss": 2.6169, + "step": 10429 + }, + { + "epoch": 0.44, + "grad_norm": 0.46484375, + "learning_rate": 0.00017693780829492899, + "loss": 2.6421, + "step": 10430 + }, + { + "epoch": 0.44, + "grad_norm": 0.484375, + "learning_rate": 0.00017693352313756852, + "loss": 2.7194, + "step": 10431 + }, + { + "epoch": 0.44, + "grad_norm": 0.458984375, + "learning_rate": 0.0001769292376340353, + "loss": 2.6582, + "step": 10432 + }, + { + "epoch": 0.44, + "grad_norm": 0.53125, + "learning_rate": 0.00017692495178434862, + "loss": 2.691, + "step": 10433 + }, + { + "epoch": 0.44, + "grad_norm": 0.46875, + "learning_rate": 0.00017692066558852777, + "loss": 2.5656, + "step": 10434 + }, + { + "epoch": 0.44, + "grad_norm": 0.41796875, + "learning_rate": 0.00017691637904659207, + "loss": 2.6989, + "step": 10435 + }, + { + "epoch": 0.44, + "grad_norm": 0.466796875, + "learning_rate": 0.00017691209215856075, + "loss": 2.6627, + "step": 10436 + }, + { + "epoch": 0.44, + "grad_norm": 0.4453125, + "learning_rate": 0.00017690780492445312, + "loss": 2.6574, + "step": 10437 + }, + { + "epoch": 0.44, + "grad_norm": 0.44140625, + "learning_rate": 0.00017690351734428846, + "loss": 2.6226, + "step": 10438 + }, + { + "epoch": 0.44, + "grad_norm": 0.453125, + "learning_rate": 0.0001768992294180861, + "loss": 2.5659, + "step": 10439 + }, + { + "epoch": 0.44, + "grad_norm": 0.45703125, + "learning_rate": 0.00017689494114586532, + "loss": 2.5683, + "step": 10440 + }, + { + "epoch": 0.44, + "grad_norm": 0.458984375, + "learning_rate": 0.00017689065252764538, + "loss": 2.5767, + "step": 10441 + }, + { + "epoch": 0.44, + "grad_norm": 0.419921875, + "learning_rate": 0.00017688636356344558, + "loss": 2.5813, + "step": 10442 + }, + { + "epoch": 0.45, + "grad_norm": 0.4296875, + "learning_rate": 0.0001768820742532853, + "loss": 2.4657, + "step": 10443 + }, + { + "epoch": 0.45, + "grad_norm": 0.44921875, + "learning_rate": 0.00017687778459718375, + "loss": 2.5623, + "step": 10444 + }, + { + "epoch": 0.45, + "grad_norm": 0.462890625, + "learning_rate": 0.0001768734945951603, + "loss": 2.6069, + "step": 10445 + }, + { + "epoch": 0.45, + "grad_norm": 0.486328125, + "learning_rate": 0.0001768692042472342, + "loss": 2.6949, + "step": 10446 + }, + { + "epoch": 0.45, + "grad_norm": 0.45703125, + "learning_rate": 0.00017686491355342474, + "loss": 2.5779, + "step": 10447 + }, + { + "epoch": 0.45, + "grad_norm": 0.46484375, + "learning_rate": 0.00017686062251375128, + "loss": 2.5792, + "step": 10448 + }, + { + "epoch": 0.45, + "grad_norm": 0.447265625, + "learning_rate": 0.00017685633112823315, + "loss": 2.5392, + "step": 10449 + }, + { + "epoch": 0.45, + "grad_norm": 0.447265625, + "learning_rate": 0.00017685203939688958, + "loss": 2.4868, + "step": 10450 + }, + { + "epoch": 0.45, + "grad_norm": 0.43359375, + "learning_rate": 0.0001768477473197399, + "loss": 2.5588, + "step": 10451 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.00017684345489680348, + "loss": 2.6181, + "step": 10452 + }, + { + "epoch": 0.45, + "grad_norm": 0.462890625, + "learning_rate": 0.00017683916212809958, + "loss": 2.5712, + "step": 10453 + }, + { + "epoch": 0.45, + "grad_norm": 0.51171875, + "learning_rate": 0.0001768348690136475, + "loss": 2.7989, + "step": 10454 + }, + { + "epoch": 0.45, + "grad_norm": 0.48828125, + "learning_rate": 0.00017683057555346662, + "loss": 2.6088, + "step": 10455 + }, + { + "epoch": 0.45, + "grad_norm": 0.396484375, + "learning_rate": 0.00017682628174757624, + "loss": 2.5931, + "step": 10456 + }, + { + "epoch": 0.45, + "grad_norm": 0.53125, + "learning_rate": 0.0001768219875959957, + "loss": 2.6539, + "step": 10457 + }, + { + "epoch": 0.45, + "grad_norm": 0.419921875, + "learning_rate": 0.00017681769309874423, + "loss": 2.6145, + "step": 10458 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.00017681339825584123, + "loss": 2.7589, + "step": 10459 + }, + { + "epoch": 0.45, + "grad_norm": 0.462890625, + "learning_rate": 0.000176809103067306, + "loss": 2.65, + "step": 10460 + }, + { + "epoch": 0.45, + "grad_norm": 0.51171875, + "learning_rate": 0.0001768048075331579, + "loss": 2.5358, + "step": 10461 + }, + { + "epoch": 0.45, + "grad_norm": 0.443359375, + "learning_rate": 0.0001768005116534162, + "loss": 2.5677, + "step": 10462 + }, + { + "epoch": 0.45, + "grad_norm": 0.46875, + "learning_rate": 0.0001767962154281003, + "loss": 2.6226, + "step": 10463 + }, + { + "epoch": 0.45, + "grad_norm": 0.515625, + "learning_rate": 0.0001767919188572295, + "loss": 2.6187, + "step": 10464 + }, + { + "epoch": 0.45, + "grad_norm": 0.51171875, + "learning_rate": 0.00017678762194082312, + "loss": 2.6566, + "step": 10465 + }, + { + "epoch": 0.45, + "grad_norm": 0.482421875, + "learning_rate": 0.00017678332467890048, + "loss": 2.6406, + "step": 10466 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017677902707148097, + "loss": 2.5261, + "step": 10467 + }, + { + "epoch": 0.45, + "grad_norm": 0.51171875, + "learning_rate": 0.0001767747291185839, + "loss": 2.6738, + "step": 10468 + }, + { + "epoch": 0.45, + "grad_norm": 0.486328125, + "learning_rate": 0.0001767704308202286, + "loss": 2.6888, + "step": 10469 + }, + { + "epoch": 0.45, + "grad_norm": 0.50390625, + "learning_rate": 0.0001767661321764344, + "loss": 2.618, + "step": 10470 + }, + { + "epoch": 0.45, + "grad_norm": 0.50390625, + "learning_rate": 0.00017676183318722068, + "loss": 2.7314, + "step": 10471 + }, + { + "epoch": 0.45, + "grad_norm": 0.447265625, + "learning_rate": 0.00017675753385260675, + "loss": 2.6502, + "step": 10472 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.00017675323417261197, + "loss": 2.6833, + "step": 10473 + }, + { + "epoch": 0.45, + "grad_norm": 0.431640625, + "learning_rate": 0.0001767489341472557, + "loss": 2.5984, + "step": 10474 + }, + { + "epoch": 0.45, + "grad_norm": 0.494140625, + "learning_rate": 0.00017674463377655728, + "loss": 2.6785, + "step": 10475 + }, + { + "epoch": 0.45, + "grad_norm": 0.466796875, + "learning_rate": 0.00017674033306053604, + "loss": 2.5733, + "step": 10476 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.00017673603199921134, + "loss": 2.6327, + "step": 10477 + }, + { + "epoch": 0.45, + "grad_norm": 0.41796875, + "learning_rate": 0.00017673173059260254, + "loss": 2.6702, + "step": 10478 + }, + { + "epoch": 0.45, + "grad_norm": 0.4140625, + "learning_rate": 0.00017672742884072902, + "loss": 2.5964, + "step": 10479 + }, + { + "epoch": 0.45, + "grad_norm": 0.44140625, + "learning_rate": 0.0001767231267436101, + "loss": 2.5568, + "step": 10480 + }, + { + "epoch": 0.45, + "grad_norm": 0.435546875, + "learning_rate": 0.00017671882430126513, + "loss": 2.5324, + "step": 10481 + }, + { + "epoch": 0.45, + "grad_norm": 0.462890625, + "learning_rate": 0.0001767145215137135, + "loss": 2.6446, + "step": 10482 + }, + { + "epoch": 0.45, + "grad_norm": 0.458984375, + "learning_rate": 0.00017671021838097458, + "loss": 2.6932, + "step": 10483 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.0001767059149030677, + "loss": 2.4953, + "step": 10484 + }, + { + "epoch": 0.45, + "grad_norm": 0.404296875, + "learning_rate": 0.0001767016110800122, + "loss": 2.5068, + "step": 10485 + }, + { + "epoch": 0.45, + "grad_norm": 0.515625, + "learning_rate": 0.00017669730691182751, + "loss": 2.6267, + "step": 10486 + }, + { + "epoch": 0.45, + "grad_norm": 0.4453125, + "learning_rate": 0.00017669300239853296, + "loss": 2.508, + "step": 10487 + }, + { + "epoch": 0.45, + "grad_norm": 0.5, + "learning_rate": 0.00017668869754014794, + "loss": 2.5531, + "step": 10488 + }, + { + "epoch": 0.45, + "grad_norm": 0.470703125, + "learning_rate": 0.0001766843923366918, + "loss": 2.5739, + "step": 10489 + }, + { + "epoch": 0.45, + "grad_norm": 0.52734375, + "learning_rate": 0.00017668008678818392, + "loss": 2.6752, + "step": 10490 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.00017667578089464368, + "loss": 2.5989, + "step": 10491 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.00017667147465609044, + "loss": 2.6526, + "step": 10492 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.00017666716807254357, + "loss": 2.553, + "step": 10493 + }, + { + "epoch": 0.45, + "grad_norm": 0.52734375, + "learning_rate": 0.0001766628611440225, + "loss": 2.5949, + "step": 10494 + }, + { + "epoch": 0.45, + "grad_norm": 0.47265625, + "learning_rate": 0.00017665855387054654, + "loss": 2.5539, + "step": 10495 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.00017665424625213513, + "loss": 2.5779, + "step": 10496 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.0001766499382888076, + "loss": 2.5329, + "step": 10497 + }, + { + "epoch": 0.45, + "grad_norm": 0.48828125, + "learning_rate": 0.0001766456299805834, + "loss": 2.6741, + "step": 10498 + }, + { + "epoch": 0.45, + "grad_norm": 0.43359375, + "learning_rate": 0.00017664132132748183, + "loss": 2.7038, + "step": 10499 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.00017663701232952233, + "loss": 2.7428, + "step": 10500 + }, + { + "epoch": 0.45, + "grad_norm": 0.5, + "learning_rate": 0.0001766327029867243, + "loss": 2.6992, + "step": 10501 + }, + { + "epoch": 0.45, + "grad_norm": 0.466796875, + "learning_rate": 0.00017662839329910709, + "loss": 2.6196, + "step": 10502 + }, + { + "epoch": 0.45, + "grad_norm": 0.48828125, + "learning_rate": 0.00017662408326669013, + "loss": 2.6897, + "step": 10503 + }, + { + "epoch": 0.45, + "grad_norm": 0.490234375, + "learning_rate": 0.00017661977288949281, + "loss": 2.599, + "step": 10504 + }, + { + "epoch": 0.45, + "grad_norm": 0.462890625, + "learning_rate": 0.0001766154621675345, + "loss": 2.6761, + "step": 10505 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017661115110083458, + "loss": 2.5875, + "step": 10506 + }, + { + "epoch": 0.45, + "grad_norm": 0.5234375, + "learning_rate": 0.0001766068396894125, + "loss": 2.7302, + "step": 10507 + }, + { + "epoch": 0.45, + "grad_norm": 0.458984375, + "learning_rate": 0.00017660252793328768, + "loss": 2.6641, + "step": 10508 + }, + { + "epoch": 0.45, + "grad_norm": 0.41796875, + "learning_rate": 0.00017659821583247944, + "loss": 2.5953, + "step": 10509 + }, + { + "epoch": 0.45, + "grad_norm": 0.49609375, + "learning_rate": 0.00017659390338700722, + "loss": 2.7724, + "step": 10510 + }, + { + "epoch": 0.45, + "grad_norm": 0.4453125, + "learning_rate": 0.00017658959059689043, + "loss": 2.7458, + "step": 10511 + }, + { + "epoch": 0.45, + "grad_norm": 0.451171875, + "learning_rate": 0.0001765852774621485, + "loss": 2.5223, + "step": 10512 + }, + { + "epoch": 0.45, + "grad_norm": 0.427734375, + "learning_rate": 0.00017658096398280075, + "loss": 2.4939, + "step": 10513 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.0001765766501588667, + "loss": 2.5609, + "step": 10514 + }, + { + "epoch": 0.45, + "grad_norm": 0.43359375, + "learning_rate": 0.00017657233599036571, + "loss": 2.5244, + "step": 10515 + }, + { + "epoch": 0.45, + "grad_norm": 0.5078125, + "learning_rate": 0.00017656802147731718, + "loss": 2.683, + "step": 10516 + }, + { + "epoch": 0.45, + "grad_norm": 0.45703125, + "learning_rate": 0.00017656370661974053, + "loss": 2.6682, + "step": 10517 + }, + { + "epoch": 0.45, + "grad_norm": 0.5234375, + "learning_rate": 0.0001765593914176552, + "loss": 2.6256, + "step": 10518 + }, + { + "epoch": 0.45, + "grad_norm": 0.37890625, + "learning_rate": 0.00017655507587108056, + "loss": 2.628, + "step": 10519 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.00017655075998003605, + "loss": 2.6985, + "step": 10520 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.0001765464437445411, + "loss": 2.6655, + "step": 10521 + }, + { + "epoch": 0.45, + "grad_norm": 0.3984375, + "learning_rate": 0.00017654212716461517, + "loss": 2.4609, + "step": 10522 + }, + { + "epoch": 0.45, + "grad_norm": 0.4296875, + "learning_rate": 0.00017653781024027762, + "loss": 2.6843, + "step": 10523 + }, + { + "epoch": 0.45, + "grad_norm": 0.44140625, + "learning_rate": 0.00017653349297154787, + "loss": 2.4988, + "step": 10524 + }, + { + "epoch": 0.45, + "grad_norm": 0.48828125, + "learning_rate": 0.0001765291753584454, + "loss": 2.6226, + "step": 10525 + }, + { + "epoch": 0.45, + "grad_norm": 0.486328125, + "learning_rate": 0.0001765248574009896, + "loss": 2.729, + "step": 10526 + }, + { + "epoch": 0.45, + "grad_norm": 0.49609375, + "learning_rate": 0.0001765205390991999, + "loss": 2.5492, + "step": 10527 + }, + { + "epoch": 0.45, + "grad_norm": 0.412109375, + "learning_rate": 0.00017651622045309574, + "loss": 2.5319, + "step": 10528 + }, + { + "epoch": 0.45, + "grad_norm": 0.51171875, + "learning_rate": 0.00017651190146269655, + "loss": 2.6085, + "step": 10529 + }, + { + "epoch": 0.45, + "grad_norm": 0.41796875, + "learning_rate": 0.0001765075821280218, + "loss": 2.6087, + "step": 10530 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.00017650326244909086, + "loss": 2.5959, + "step": 10531 + }, + { + "epoch": 0.45, + "grad_norm": 0.47265625, + "learning_rate": 0.0001764989424259232, + "loss": 2.5912, + "step": 10532 + }, + { + "epoch": 0.45, + "grad_norm": 0.490234375, + "learning_rate": 0.00017649462205853827, + "loss": 2.6065, + "step": 10533 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.00017649030134695549, + "loss": 2.399, + "step": 10534 + }, + { + "epoch": 0.45, + "grad_norm": 0.48046875, + "learning_rate": 0.00017648598029119426, + "loss": 2.6365, + "step": 10535 + }, + { + "epoch": 0.45, + "grad_norm": 0.435546875, + "learning_rate": 0.00017648165889127415, + "loss": 2.6729, + "step": 10536 + }, + { + "epoch": 0.45, + "grad_norm": 0.416015625, + "learning_rate": 0.00017647733714721447, + "loss": 2.6474, + "step": 10537 + }, + { + "epoch": 0.45, + "grad_norm": 0.4375, + "learning_rate": 0.00017647301505903476, + "loss": 2.5129, + "step": 10538 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.0001764686926267544, + "loss": 2.5587, + "step": 10539 + }, + { + "epoch": 0.45, + "grad_norm": 0.416015625, + "learning_rate": 0.00017646436985039286, + "loss": 2.638, + "step": 10540 + }, + { + "epoch": 0.45, + "grad_norm": 0.458984375, + "learning_rate": 0.00017646004672996964, + "loss": 2.5117, + "step": 10541 + }, + { + "epoch": 0.45, + "grad_norm": 0.470703125, + "learning_rate": 0.00017645572326550413, + "loss": 2.6479, + "step": 10542 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.0001764513994570158, + "loss": 2.593, + "step": 10543 + }, + { + "epoch": 0.45, + "grad_norm": 0.439453125, + "learning_rate": 0.00017644707530452413, + "loss": 2.5177, + "step": 10544 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.00017644275080804856, + "loss": 2.5323, + "step": 10545 + }, + { + "epoch": 0.45, + "grad_norm": 0.404296875, + "learning_rate": 0.00017643842596760852, + "loss": 2.6056, + "step": 10546 + }, + { + "epoch": 0.45, + "grad_norm": 0.462890625, + "learning_rate": 0.00017643410078322353, + "loss": 2.6541, + "step": 10547 + }, + { + "epoch": 0.45, + "grad_norm": 0.408203125, + "learning_rate": 0.00017642977525491303, + "loss": 2.6664, + "step": 10548 + }, + { + "epoch": 0.45, + "grad_norm": 0.494140625, + "learning_rate": 0.00017642544938269644, + "loss": 2.6439, + "step": 10549 + }, + { + "epoch": 0.45, + "grad_norm": 0.50390625, + "learning_rate": 0.00017642112316659328, + "loss": 2.5069, + "step": 10550 + }, + { + "epoch": 0.45, + "grad_norm": 0.45703125, + "learning_rate": 0.000176416796606623, + "loss": 2.6058, + "step": 10551 + }, + { + "epoch": 0.45, + "grad_norm": 0.462890625, + "learning_rate": 0.00017641246970280503, + "loss": 2.5052, + "step": 10552 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.00017640814245515886, + "loss": 2.599, + "step": 10553 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017640381486370402, + "loss": 2.6458, + "step": 10554 + }, + { + "epoch": 0.45, + "grad_norm": 0.451171875, + "learning_rate": 0.0001763994869284599, + "loss": 2.6065, + "step": 10555 + }, + { + "epoch": 0.45, + "grad_norm": 0.51171875, + "learning_rate": 0.00017639515864944602, + "loss": 2.7822, + "step": 10556 + }, + { + "epoch": 0.45, + "grad_norm": 0.43359375, + "learning_rate": 0.00017639083002668186, + "loss": 2.7022, + "step": 10557 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.0001763865010601869, + "loss": 2.6878, + "step": 10558 + }, + { + "epoch": 0.45, + "grad_norm": 0.458984375, + "learning_rate": 0.00017638217174998055, + "loss": 2.5299, + "step": 10559 + }, + { + "epoch": 0.45, + "grad_norm": 0.482421875, + "learning_rate": 0.00017637784209608237, + "loss": 2.6903, + "step": 10560 + }, + { + "epoch": 0.45, + "grad_norm": 0.455078125, + "learning_rate": 0.0001763735120985118, + "loss": 2.6402, + "step": 10561 + }, + { + "epoch": 0.45, + "grad_norm": 0.45703125, + "learning_rate": 0.00017636918175728837, + "loss": 2.6574, + "step": 10562 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.00017636485107243148, + "loss": 2.7254, + "step": 10563 + }, + { + "epoch": 0.45, + "grad_norm": 0.44140625, + "learning_rate": 0.00017636052004396072, + "loss": 2.6039, + "step": 10564 + }, + { + "epoch": 0.45, + "grad_norm": 0.388671875, + "learning_rate": 0.0001763561886718955, + "loss": 2.6993, + "step": 10565 + }, + { + "epoch": 0.45, + "grad_norm": 0.466796875, + "learning_rate": 0.00017635185695625533, + "loss": 2.7106, + "step": 10566 + }, + { + "epoch": 0.45, + "grad_norm": 0.45703125, + "learning_rate": 0.00017634752489705972, + "loss": 2.6255, + "step": 10567 + }, + { + "epoch": 0.45, + "grad_norm": 0.41796875, + "learning_rate": 0.00017634319249432812, + "loss": 2.5718, + "step": 10568 + }, + { + "epoch": 0.45, + "grad_norm": 0.421875, + "learning_rate": 0.0001763388597480801, + "loss": 2.6023, + "step": 10569 + }, + { + "epoch": 0.45, + "grad_norm": 0.5625, + "learning_rate": 0.0001763345266583351, + "loss": 2.7661, + "step": 10570 + }, + { + "epoch": 0.45, + "grad_norm": 0.4453125, + "learning_rate": 0.00017633019322511263, + "loss": 2.6242, + "step": 10571 + }, + { + "epoch": 0.45, + "grad_norm": 0.447265625, + "learning_rate": 0.00017632585944843217, + "loss": 2.5327, + "step": 10572 + }, + { + "epoch": 0.45, + "grad_norm": 0.466796875, + "learning_rate": 0.00017632152532831322, + "loss": 2.6973, + "step": 10573 + }, + { + "epoch": 0.45, + "grad_norm": 0.455078125, + "learning_rate": 0.00017631719086477535, + "loss": 2.6744, + "step": 10574 + }, + { + "epoch": 0.45, + "grad_norm": 0.53515625, + "learning_rate": 0.00017631285605783795, + "loss": 2.7322, + "step": 10575 + }, + { + "epoch": 0.45, + "grad_norm": 0.50390625, + "learning_rate": 0.00017630852090752064, + "loss": 2.611, + "step": 10576 + }, + { + "epoch": 0.45, + "grad_norm": 0.46875, + "learning_rate": 0.00017630418541384285, + "loss": 2.5355, + "step": 10577 + }, + { + "epoch": 0.45, + "grad_norm": 0.400390625, + "learning_rate": 0.00017629984957682408, + "loss": 2.5957, + "step": 10578 + }, + { + "epoch": 0.45, + "grad_norm": 0.46875, + "learning_rate": 0.00017629551339648391, + "loss": 2.6078, + "step": 10579 + }, + { + "epoch": 0.45, + "grad_norm": 0.44921875, + "learning_rate": 0.00017629117687284184, + "loss": 2.5935, + "step": 10580 + }, + { + "epoch": 0.45, + "grad_norm": 0.478515625, + "learning_rate": 0.0001762868400059173, + "loss": 2.7253, + "step": 10581 + }, + { + "epoch": 0.45, + "grad_norm": 0.44921875, + "learning_rate": 0.00017628250279572987, + "loss": 2.5846, + "step": 10582 + }, + { + "epoch": 0.45, + "grad_norm": 0.44921875, + "learning_rate": 0.00017627816524229907, + "loss": 2.5933, + "step": 10583 + }, + { + "epoch": 0.45, + "grad_norm": 0.48828125, + "learning_rate": 0.0001762738273456444, + "loss": 2.6939, + "step": 10584 + }, + { + "epoch": 0.45, + "grad_norm": 0.478515625, + "learning_rate": 0.0001762694891057854, + "loss": 2.5728, + "step": 10585 + }, + { + "epoch": 0.45, + "grad_norm": 0.53125, + "learning_rate": 0.00017626515052274156, + "loss": 2.6908, + "step": 10586 + }, + { + "epoch": 0.45, + "grad_norm": 0.53515625, + "learning_rate": 0.00017626081159653238, + "loss": 2.6628, + "step": 10587 + }, + { + "epoch": 0.45, + "grad_norm": 0.482421875, + "learning_rate": 0.00017625647232717747, + "loss": 2.6729, + "step": 10588 + }, + { + "epoch": 0.45, + "grad_norm": 0.470703125, + "learning_rate": 0.00017625213271469625, + "loss": 2.6085, + "step": 10589 + }, + { + "epoch": 0.45, + "grad_norm": 0.462890625, + "learning_rate": 0.00017624779275910834, + "loss": 2.7234, + "step": 10590 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.00017624345246043324, + "loss": 2.5668, + "step": 10591 + }, + { + "epoch": 0.45, + "grad_norm": 0.5078125, + "learning_rate": 0.00017623911181869045, + "loss": 2.5578, + "step": 10592 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.00017623477083389947, + "loss": 2.6314, + "step": 10593 + }, + { + "epoch": 0.45, + "grad_norm": 0.490234375, + "learning_rate": 0.00017623042950607994, + "loss": 2.7019, + "step": 10594 + }, + { + "epoch": 0.45, + "grad_norm": 0.443359375, + "learning_rate": 0.0001762260878352513, + "loss": 2.594, + "step": 10595 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017622174582143316, + "loss": 2.5504, + "step": 10596 + }, + { + "epoch": 0.45, + "grad_norm": 0.50390625, + "learning_rate": 0.00017621740346464502, + "loss": 2.6869, + "step": 10597 + }, + { + "epoch": 0.45, + "grad_norm": 0.494140625, + "learning_rate": 0.00017621306076490641, + "loss": 2.6386, + "step": 10598 + }, + { + "epoch": 0.45, + "grad_norm": 0.41015625, + "learning_rate": 0.00017620871772223686, + "loss": 2.5857, + "step": 10599 + }, + { + "epoch": 0.45, + "grad_norm": 0.41796875, + "learning_rate": 0.00017620437433665592, + "loss": 2.4826, + "step": 10600 + }, + { + "epoch": 0.45, + "grad_norm": 0.408203125, + "learning_rate": 0.0001762000306081832, + "loss": 2.5501, + "step": 10601 + }, + { + "epoch": 0.45, + "grad_norm": 0.447265625, + "learning_rate": 0.00017619568653683815, + "loss": 2.6237, + "step": 10602 + }, + { + "epoch": 0.45, + "grad_norm": 0.46875, + "learning_rate": 0.00017619134212264035, + "loss": 2.7252, + "step": 10603 + }, + { + "epoch": 0.45, + "grad_norm": 0.4921875, + "learning_rate": 0.00017618699736560937, + "loss": 2.7636, + "step": 10604 + }, + { + "epoch": 0.45, + "grad_norm": 0.4765625, + "learning_rate": 0.00017618265226576477, + "loss": 2.5864, + "step": 10605 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017617830682312605, + "loss": 2.7172, + "step": 10606 + }, + { + "epoch": 0.45, + "grad_norm": 0.451171875, + "learning_rate": 0.00017617396103771277, + "loss": 2.5735, + "step": 10607 + }, + { + "epoch": 0.45, + "grad_norm": 0.451171875, + "learning_rate": 0.0001761696149095445, + "loss": 2.6309, + "step": 10608 + }, + { + "epoch": 0.45, + "grad_norm": 0.498046875, + "learning_rate": 0.00017616526843864085, + "loss": 2.6392, + "step": 10609 + }, + { + "epoch": 0.45, + "grad_norm": 0.490234375, + "learning_rate": 0.0001761609216250213, + "loss": 2.6452, + "step": 10610 + }, + { + "epoch": 0.45, + "grad_norm": 0.466796875, + "learning_rate": 0.00017615657446870545, + "loss": 2.5184, + "step": 10611 + }, + { + "epoch": 0.45, + "grad_norm": 0.458984375, + "learning_rate": 0.00017615222696971283, + "loss": 2.4807, + "step": 10612 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.00017614787912806302, + "loss": 2.6155, + "step": 10613 + }, + { + "epoch": 0.45, + "grad_norm": 0.5, + "learning_rate": 0.0001761435309437756, + "loss": 2.7749, + "step": 10614 + }, + { + "epoch": 0.45, + "grad_norm": 0.4375, + "learning_rate": 0.00017613918241687012, + "loss": 2.5805, + "step": 10615 + }, + { + "epoch": 0.45, + "grad_norm": 0.453125, + "learning_rate": 0.00017613483354736608, + "loss": 2.7459, + "step": 10616 + }, + { + "epoch": 0.45, + "grad_norm": 0.5078125, + "learning_rate": 0.00017613048433528318, + "loss": 2.7792, + "step": 10617 + }, + { + "epoch": 0.45, + "grad_norm": 0.447265625, + "learning_rate": 0.0001761261347806409, + "loss": 2.4891, + "step": 10618 + }, + { + "epoch": 0.45, + "grad_norm": 0.486328125, + "learning_rate": 0.00017612178488345885, + "loss": 2.6059, + "step": 10619 + }, + { + "epoch": 0.45, + "grad_norm": 0.419921875, + "learning_rate": 0.00017611743464375658, + "loss": 2.5365, + "step": 10620 + }, + { + "epoch": 0.45, + "grad_norm": 0.4375, + "learning_rate": 0.00017611308406155366, + "loss": 2.5382, + "step": 10621 + }, + { + "epoch": 0.45, + "grad_norm": 0.5078125, + "learning_rate": 0.0001761087331368697, + "loss": 2.6678, + "step": 10622 + }, + { + "epoch": 0.45, + "grad_norm": 0.50390625, + "learning_rate": 0.00017610438186972424, + "loss": 2.6034, + "step": 10623 + }, + { + "epoch": 0.45, + "grad_norm": 0.455078125, + "learning_rate": 0.00017610003026013687, + "loss": 2.5259, + "step": 10624 + }, + { + "epoch": 0.45, + "grad_norm": 0.49609375, + "learning_rate": 0.00017609567830812717, + "loss": 2.7007, + "step": 10625 + }, + { + "epoch": 0.45, + "grad_norm": 0.458984375, + "learning_rate": 0.00017609132601371474, + "loss": 2.5342, + "step": 10626 + }, + { + "epoch": 0.45, + "grad_norm": 0.49609375, + "learning_rate": 0.00017608697337691918, + "loss": 2.6851, + "step": 10627 + }, + { + "epoch": 0.45, + "grad_norm": 0.47265625, + "learning_rate": 0.00017608262039776, + "loss": 2.7326, + "step": 10628 + }, + { + "epoch": 0.45, + "grad_norm": 0.466796875, + "learning_rate": 0.00017607826707625687, + "loss": 2.6561, + "step": 10629 + }, + { + "epoch": 0.45, + "grad_norm": 0.48046875, + "learning_rate": 0.0001760739134124293, + "loss": 2.6922, + "step": 10630 + }, + { + "epoch": 0.45, + "grad_norm": 0.5703125, + "learning_rate": 0.00017606955940629692, + "loss": 2.6324, + "step": 10631 + }, + { + "epoch": 0.45, + "grad_norm": 0.47265625, + "learning_rate": 0.00017606520505787936, + "loss": 2.4371, + "step": 10632 + }, + { + "epoch": 0.45, + "grad_norm": 0.49609375, + "learning_rate": 0.00017606085036719617, + "loss": 2.7158, + "step": 10633 + }, + { + "epoch": 0.45, + "grad_norm": 0.41015625, + "learning_rate": 0.00017605649533426694, + "loss": 2.5487, + "step": 10634 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.0001760521399591113, + "loss": 2.5417, + "step": 10635 + }, + { + "epoch": 0.45, + "grad_norm": 0.400390625, + "learning_rate": 0.00017604778424174877, + "loss": 2.4878, + "step": 10636 + }, + { + "epoch": 0.45, + "grad_norm": 0.50390625, + "learning_rate": 0.00017604342818219906, + "loss": 2.6099, + "step": 10637 + }, + { + "epoch": 0.45, + "grad_norm": 0.396484375, + "learning_rate": 0.0001760390717804817, + "loss": 2.6323, + "step": 10638 + }, + { + "epoch": 0.45, + "grad_norm": 0.42578125, + "learning_rate": 0.00017603471503661625, + "loss": 2.4724, + "step": 10639 + }, + { + "epoch": 0.45, + "grad_norm": 0.490234375, + "learning_rate": 0.00017603035795062244, + "loss": 2.6814, + "step": 10640 + }, + { + "epoch": 0.45, + "grad_norm": 0.44921875, + "learning_rate": 0.0001760260005225198, + "loss": 2.535, + "step": 10641 + }, + { + "epoch": 0.45, + "grad_norm": 0.458984375, + "learning_rate": 0.0001760216427523279, + "loss": 2.7214, + "step": 10642 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017601728464006642, + "loss": 2.7258, + "step": 10643 + }, + { + "epoch": 0.45, + "grad_norm": 0.423828125, + "learning_rate": 0.00017601292618575493, + "loss": 2.6782, + "step": 10644 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017600856738941305, + "loss": 2.5547, + "step": 10645 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017600420825106042, + "loss": 2.5255, + "step": 10646 + }, + { + "epoch": 0.45, + "grad_norm": 0.50390625, + "learning_rate": 0.0001759998487707166, + "loss": 2.6505, + "step": 10647 + }, + { + "epoch": 0.45, + "grad_norm": 0.486328125, + "learning_rate": 0.00017599548894840124, + "loss": 2.7311, + "step": 10648 + }, + { + "epoch": 0.45, + "grad_norm": 0.423828125, + "learning_rate": 0.00017599112878413395, + "loss": 2.5111, + "step": 10649 + }, + { + "epoch": 0.45, + "grad_norm": 0.447265625, + "learning_rate": 0.00017598676827793437, + "loss": 2.6207, + "step": 10650 + }, + { + "epoch": 0.45, + "grad_norm": 0.458984375, + "learning_rate": 0.0001759824074298221, + "loss": 2.6297, + "step": 10651 + }, + { + "epoch": 0.45, + "grad_norm": 0.486328125, + "learning_rate": 0.00017597804623981675, + "loss": 2.5905, + "step": 10652 + }, + { + "epoch": 0.45, + "grad_norm": 0.49609375, + "learning_rate": 0.00017597368470793798, + "loss": 2.6346, + "step": 10653 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017596932283420538, + "loss": 2.7215, + "step": 10654 + }, + { + "epoch": 0.45, + "grad_norm": 0.5, + "learning_rate": 0.00017596496061863856, + "loss": 2.5029, + "step": 10655 + }, + { + "epoch": 0.45, + "grad_norm": 0.439453125, + "learning_rate": 0.00017596059806125722, + "loss": 2.5451, + "step": 10656 + }, + { + "epoch": 0.45, + "grad_norm": 0.455078125, + "learning_rate": 0.00017595623516208095, + "loss": 2.6851, + "step": 10657 + }, + { + "epoch": 0.45, + "grad_norm": 0.45703125, + "learning_rate": 0.00017595187192112936, + "loss": 2.5586, + "step": 10658 + }, + { + "epoch": 0.45, + "grad_norm": 0.447265625, + "learning_rate": 0.0001759475083384221, + "loss": 2.5962, + "step": 10659 + }, + { + "epoch": 0.45, + "grad_norm": 0.4140625, + "learning_rate": 0.00017594314441397884, + "loss": 2.548, + "step": 10660 + }, + { + "epoch": 0.45, + "grad_norm": 0.41015625, + "learning_rate": 0.00017593878014781917, + "loss": 2.5303, + "step": 10661 + }, + { + "epoch": 0.45, + "grad_norm": 0.466796875, + "learning_rate": 0.00017593441553996275, + "loss": 2.6919, + "step": 10662 + }, + { + "epoch": 0.45, + "grad_norm": 0.451171875, + "learning_rate": 0.0001759300505904292, + "loss": 2.6541, + "step": 10663 + }, + { + "epoch": 0.45, + "grad_norm": 0.5078125, + "learning_rate": 0.00017592568529923815, + "loss": 2.4866, + "step": 10664 + }, + { + "epoch": 0.45, + "grad_norm": 0.486328125, + "learning_rate": 0.0001759213196664093, + "loss": 2.5628, + "step": 10665 + }, + { + "epoch": 0.45, + "grad_norm": 0.48046875, + "learning_rate": 0.00017591695369196224, + "loss": 2.656, + "step": 10666 + }, + { + "epoch": 0.45, + "grad_norm": 0.419921875, + "learning_rate": 0.00017591258737591661, + "loss": 2.5506, + "step": 10667 + }, + { + "epoch": 0.45, + "grad_norm": 0.478515625, + "learning_rate": 0.00017590822071829212, + "loss": 2.5487, + "step": 10668 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.00017590385371910836, + "loss": 2.5703, + "step": 10669 + }, + { + "epoch": 0.45, + "grad_norm": 0.431640625, + "learning_rate": 0.000175899486378385, + "loss": 2.6826, + "step": 10670 + }, + { + "epoch": 0.45, + "grad_norm": 0.4609375, + "learning_rate": 0.00017589511869614167, + "loss": 2.5482, + "step": 10671 + }, + { + "epoch": 0.45, + "grad_norm": 0.515625, + "learning_rate": 0.00017589075067239807, + "loss": 2.7016, + "step": 10672 + }, + { + "epoch": 0.45, + "grad_norm": 0.474609375, + "learning_rate": 0.00017588638230717384, + "loss": 2.6009, + "step": 10673 + }, + { + "epoch": 0.45, + "grad_norm": 0.46484375, + "learning_rate": 0.00017588201360048858, + "loss": 2.6181, + "step": 10674 + }, + { + "epoch": 0.45, + "grad_norm": 0.380859375, + "learning_rate": 0.00017587764455236201, + "loss": 2.6075, + "step": 10675 + }, + { + "epoch": 0.45, + "grad_norm": 0.419921875, + "learning_rate": 0.0001758732751628138, + "loss": 2.6809, + "step": 10676 + }, + { + "epoch": 0.45, + "grad_norm": 0.40625, + "learning_rate": 0.00017586890543186355, + "loss": 2.5034, + "step": 10677 + }, + { + "epoch": 0.46, + "grad_norm": 0.458984375, + "learning_rate": 0.00017586453535953095, + "loss": 2.6754, + "step": 10678 + }, + { + "epoch": 0.46, + "grad_norm": 0.4453125, + "learning_rate": 0.00017586016494583565, + "loss": 2.5066, + "step": 10679 + }, + { + "epoch": 0.46, + "grad_norm": 0.412109375, + "learning_rate": 0.00017585579419079737, + "loss": 2.6082, + "step": 10680 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.00017585142309443571, + "loss": 2.6072, + "step": 10681 + }, + { + "epoch": 0.46, + "grad_norm": 0.4296875, + "learning_rate": 0.00017584705165677039, + "loss": 2.5778, + "step": 10682 + }, + { + "epoch": 0.46, + "grad_norm": 0.427734375, + "learning_rate": 0.00017584267987782104, + "loss": 2.6207, + "step": 10683 + }, + { + "epoch": 0.46, + "grad_norm": 0.45703125, + "learning_rate": 0.00017583830775760734, + "loss": 2.5358, + "step": 10684 + }, + { + "epoch": 0.46, + "grad_norm": 0.435546875, + "learning_rate": 0.00017583393529614898, + "loss": 2.5367, + "step": 10685 + }, + { + "epoch": 0.46, + "grad_norm": 0.486328125, + "learning_rate": 0.00017582956249346564, + "loss": 2.6036, + "step": 10686 + }, + { + "epoch": 0.46, + "grad_norm": 0.443359375, + "learning_rate": 0.00017582518934957696, + "loss": 2.7602, + "step": 10687 + }, + { + "epoch": 0.46, + "grad_norm": 0.46875, + "learning_rate": 0.00017582081586450265, + "loss": 2.5875, + "step": 10688 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.00017581644203826238, + "loss": 2.5793, + "step": 10689 + }, + { + "epoch": 0.46, + "grad_norm": 0.41796875, + "learning_rate": 0.00017581206787087582, + "loss": 2.4395, + "step": 10690 + }, + { + "epoch": 0.46, + "grad_norm": 0.47265625, + "learning_rate": 0.00017580769336236266, + "loss": 2.6011, + "step": 10691 + }, + { + "epoch": 0.46, + "grad_norm": 0.44921875, + "learning_rate": 0.0001758033185127426, + "loss": 2.5566, + "step": 10692 + }, + { + "epoch": 0.46, + "grad_norm": 0.447265625, + "learning_rate": 0.0001757989433220353, + "loss": 2.5799, + "step": 10693 + }, + { + "epoch": 0.46, + "grad_norm": 0.443359375, + "learning_rate": 0.00017579456779026048, + "loss": 2.6373, + "step": 10694 + }, + { + "epoch": 0.46, + "grad_norm": 0.451171875, + "learning_rate": 0.00017579019191743775, + "loss": 2.6291, + "step": 10695 + }, + { + "epoch": 0.46, + "grad_norm": 0.46484375, + "learning_rate": 0.0001757858157035869, + "loss": 2.7749, + "step": 10696 + }, + { + "epoch": 0.46, + "grad_norm": 0.4765625, + "learning_rate": 0.00017578143914872754, + "loss": 2.6516, + "step": 10697 + }, + { + "epoch": 0.46, + "grad_norm": 0.3828125, + "learning_rate": 0.00017577706225287943, + "loss": 2.5471, + "step": 10698 + }, + { + "epoch": 0.46, + "grad_norm": 0.423828125, + "learning_rate": 0.00017577268501606222, + "loss": 2.6356, + "step": 10699 + }, + { + "epoch": 0.46, + "grad_norm": 0.49609375, + "learning_rate": 0.0001757683074382956, + "loss": 2.71, + "step": 10700 + }, + { + "epoch": 0.46, + "grad_norm": 0.462890625, + "learning_rate": 0.00017576392951959932, + "loss": 2.6142, + "step": 10701 + }, + { + "epoch": 0.46, + "grad_norm": 0.48046875, + "learning_rate": 0.00017575955125999303, + "loss": 2.6555, + "step": 10702 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017575517265949647, + "loss": 2.6091, + "step": 10703 + }, + { + "epoch": 0.46, + "grad_norm": 0.5, + "learning_rate": 0.0001757507937181293, + "loss": 2.7655, + "step": 10704 + }, + { + "epoch": 0.46, + "grad_norm": 0.515625, + "learning_rate": 0.00017574641443591124, + "loss": 2.7431, + "step": 10705 + }, + { + "epoch": 0.46, + "grad_norm": 0.435546875, + "learning_rate": 0.000175742034812862, + "loss": 2.5132, + "step": 10706 + }, + { + "epoch": 0.46, + "grad_norm": 0.427734375, + "learning_rate": 0.00017573765484900125, + "loss": 2.6308, + "step": 10707 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017573327454434878, + "loss": 2.6717, + "step": 10708 + }, + { + "epoch": 0.46, + "grad_norm": 0.474609375, + "learning_rate": 0.00017572889389892425, + "loss": 2.5399, + "step": 10709 + }, + { + "epoch": 0.46, + "grad_norm": 0.453125, + "learning_rate": 0.00017572451291274734, + "loss": 2.6068, + "step": 10710 + }, + { + "epoch": 0.46, + "grad_norm": 0.5234375, + "learning_rate": 0.00017572013158583783, + "loss": 2.5467, + "step": 10711 + }, + { + "epoch": 0.46, + "grad_norm": 0.427734375, + "learning_rate": 0.00017571574991821537, + "loss": 2.6818, + "step": 10712 + }, + { + "epoch": 0.46, + "grad_norm": 0.447265625, + "learning_rate": 0.00017571136790989976, + "loss": 2.5356, + "step": 10713 + }, + { + "epoch": 0.46, + "grad_norm": 0.416015625, + "learning_rate": 0.00017570698556091058, + "loss": 2.6694, + "step": 10714 + }, + { + "epoch": 0.46, + "grad_norm": 0.396484375, + "learning_rate": 0.00017570260287126768, + "loss": 2.4201, + "step": 10715 + }, + { + "epoch": 0.46, + "grad_norm": 0.46484375, + "learning_rate": 0.00017569821984099075, + "loss": 2.5602, + "step": 10716 + }, + { + "epoch": 0.46, + "grad_norm": 0.4140625, + "learning_rate": 0.00017569383647009947, + "loss": 2.5961, + "step": 10717 + }, + { + "epoch": 0.46, + "grad_norm": 0.408203125, + "learning_rate": 0.00017568945275861362, + "loss": 2.4906, + "step": 10718 + }, + { + "epoch": 0.46, + "grad_norm": 0.44921875, + "learning_rate": 0.00017568506870655287, + "loss": 2.6616, + "step": 10719 + }, + { + "epoch": 0.46, + "grad_norm": 0.478515625, + "learning_rate": 0.00017568068431393695, + "loss": 2.435, + "step": 10720 + }, + { + "epoch": 0.46, + "grad_norm": 0.451171875, + "learning_rate": 0.00017567629958078565, + "loss": 2.6455, + "step": 10721 + }, + { + "epoch": 0.46, + "grad_norm": 0.458984375, + "learning_rate": 0.00017567191450711862, + "loss": 2.6634, + "step": 10722 + }, + { + "epoch": 0.46, + "grad_norm": 0.400390625, + "learning_rate": 0.00017566752909295566, + "loss": 2.4455, + "step": 10723 + }, + { + "epoch": 0.46, + "grad_norm": 0.423828125, + "learning_rate": 0.00017566314333831645, + "loss": 2.6161, + "step": 10724 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.00017565875724322077, + "loss": 2.6468, + "step": 10725 + }, + { + "epoch": 0.46, + "grad_norm": 0.486328125, + "learning_rate": 0.0001756543708076883, + "loss": 2.6811, + "step": 10726 + }, + { + "epoch": 0.46, + "grad_norm": 0.4140625, + "learning_rate": 0.00017564998403173882, + "loss": 2.6131, + "step": 10727 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.00017564559691539205, + "loss": 2.7308, + "step": 10728 + }, + { + "epoch": 0.46, + "grad_norm": 0.412109375, + "learning_rate": 0.00017564120945866778, + "loss": 2.5599, + "step": 10729 + }, + { + "epoch": 0.46, + "grad_norm": 0.498046875, + "learning_rate": 0.00017563682166158568, + "loss": 2.618, + "step": 10730 + }, + { + "epoch": 0.46, + "grad_norm": 0.43359375, + "learning_rate": 0.00017563243352416554, + "loss": 2.6295, + "step": 10731 + }, + { + "epoch": 0.46, + "grad_norm": 0.458984375, + "learning_rate": 0.00017562804504642706, + "loss": 2.4936, + "step": 10732 + }, + { + "epoch": 0.46, + "grad_norm": 0.44921875, + "learning_rate": 0.00017562365622839006, + "loss": 2.672, + "step": 10733 + }, + { + "epoch": 0.46, + "grad_norm": 0.439453125, + "learning_rate": 0.0001756192670700742, + "loss": 2.4783, + "step": 10734 + }, + { + "epoch": 0.46, + "grad_norm": 0.45703125, + "learning_rate": 0.00017561487757149932, + "loss": 2.5961, + "step": 10735 + }, + { + "epoch": 0.46, + "grad_norm": 0.44921875, + "learning_rate": 0.00017561048773268508, + "loss": 2.595, + "step": 10736 + }, + { + "epoch": 0.46, + "grad_norm": 0.4765625, + "learning_rate": 0.00017560609755365133, + "loss": 2.5421, + "step": 10737 + }, + { + "epoch": 0.46, + "grad_norm": 0.443359375, + "learning_rate": 0.00017560170703441772, + "loss": 2.5685, + "step": 10738 + }, + { + "epoch": 0.46, + "grad_norm": 0.451171875, + "learning_rate": 0.00017559731617500409, + "loss": 2.565, + "step": 10739 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017559292497543015, + "loss": 2.5684, + "step": 10740 + }, + { + "epoch": 0.46, + "grad_norm": 0.455078125, + "learning_rate": 0.0001755885334357157, + "loss": 2.6126, + "step": 10741 + }, + { + "epoch": 0.46, + "grad_norm": 0.455078125, + "learning_rate": 0.0001755841415558804, + "loss": 2.6989, + "step": 10742 + }, + { + "epoch": 0.46, + "grad_norm": 0.43359375, + "learning_rate": 0.00017557974933594416, + "loss": 2.5962, + "step": 10743 + }, + { + "epoch": 0.46, + "grad_norm": 0.474609375, + "learning_rate": 0.00017557535677592664, + "loss": 2.7529, + "step": 10744 + }, + { + "epoch": 0.46, + "grad_norm": 0.48046875, + "learning_rate": 0.00017557096387584763, + "loss": 2.6891, + "step": 10745 + }, + { + "epoch": 0.46, + "grad_norm": 0.419921875, + "learning_rate": 0.00017556657063572692, + "loss": 2.58, + "step": 10746 + }, + { + "epoch": 0.46, + "grad_norm": 0.421875, + "learning_rate": 0.00017556217705558423, + "loss": 2.6569, + "step": 10747 + }, + { + "epoch": 0.46, + "grad_norm": 0.421875, + "learning_rate": 0.0001755577831354394, + "loss": 2.4966, + "step": 10748 + }, + { + "epoch": 0.46, + "grad_norm": 0.439453125, + "learning_rate": 0.00017555338887531214, + "loss": 2.6447, + "step": 10749 + }, + { + "epoch": 0.46, + "grad_norm": 0.50390625, + "learning_rate": 0.00017554899427522225, + "loss": 2.5527, + "step": 10750 + }, + { + "epoch": 0.46, + "grad_norm": 0.421875, + "learning_rate": 0.00017554459933518944, + "loss": 2.5474, + "step": 10751 + }, + { + "epoch": 0.46, + "grad_norm": 0.46875, + "learning_rate": 0.0001755402040552336, + "loss": 2.6696, + "step": 10752 + }, + { + "epoch": 0.46, + "grad_norm": 0.49609375, + "learning_rate": 0.00017553580843537444, + "loss": 2.6003, + "step": 10753 + }, + { + "epoch": 0.46, + "grad_norm": 0.41015625, + "learning_rate": 0.00017553141247563173, + "loss": 2.4595, + "step": 10754 + }, + { + "epoch": 0.46, + "grad_norm": 0.4140625, + "learning_rate": 0.00017552701617602528, + "loss": 2.6157, + "step": 10755 + }, + { + "epoch": 0.46, + "grad_norm": 0.421875, + "learning_rate": 0.00017552261953657488, + "loss": 2.5397, + "step": 10756 + }, + { + "epoch": 0.46, + "grad_norm": 0.462890625, + "learning_rate": 0.00017551822255730025, + "loss": 2.6941, + "step": 10757 + }, + { + "epoch": 0.46, + "grad_norm": 0.4453125, + "learning_rate": 0.00017551382523822124, + "loss": 2.6264, + "step": 10758 + }, + { + "epoch": 0.46, + "grad_norm": 0.455078125, + "learning_rate": 0.00017550942757935762, + "loss": 2.4514, + "step": 10759 + }, + { + "epoch": 0.46, + "grad_norm": 0.5234375, + "learning_rate": 0.00017550502958072915, + "loss": 2.696, + "step": 10760 + }, + { + "epoch": 0.46, + "grad_norm": 0.44140625, + "learning_rate": 0.00017550063124235568, + "loss": 2.4916, + "step": 10761 + }, + { + "epoch": 0.46, + "grad_norm": 0.4453125, + "learning_rate": 0.00017549623256425692, + "loss": 2.6052, + "step": 10762 + }, + { + "epoch": 0.46, + "grad_norm": 0.490234375, + "learning_rate": 0.0001754918335464527, + "loss": 2.5667, + "step": 10763 + }, + { + "epoch": 0.46, + "grad_norm": 0.431640625, + "learning_rate": 0.00017548743418896285, + "loss": 2.5653, + "step": 10764 + }, + { + "epoch": 0.46, + "grad_norm": 0.462890625, + "learning_rate": 0.00017548303449180713, + "loss": 2.634, + "step": 10765 + }, + { + "epoch": 0.46, + "grad_norm": 0.4140625, + "learning_rate": 0.0001754786344550053, + "loss": 2.6677, + "step": 10766 + }, + { + "epoch": 0.46, + "grad_norm": 0.486328125, + "learning_rate": 0.00017547423407857725, + "loss": 2.5766, + "step": 10767 + }, + { + "epoch": 0.46, + "grad_norm": 0.42578125, + "learning_rate": 0.00017546983336254272, + "loss": 2.5483, + "step": 10768 + }, + { + "epoch": 0.46, + "grad_norm": 0.4921875, + "learning_rate": 0.0001754654323069215, + "loss": 2.6741, + "step": 10769 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017546103091173342, + "loss": 2.7404, + "step": 10770 + }, + { + "epoch": 0.46, + "grad_norm": 0.443359375, + "learning_rate": 0.0001754566291769983, + "loss": 2.5874, + "step": 10771 + }, + { + "epoch": 0.46, + "grad_norm": 0.427734375, + "learning_rate": 0.0001754522271027359, + "loss": 2.6304, + "step": 10772 + }, + { + "epoch": 0.46, + "grad_norm": 0.46484375, + "learning_rate": 0.00017544782468896606, + "loss": 2.5099, + "step": 10773 + }, + { + "epoch": 0.46, + "grad_norm": 0.42578125, + "learning_rate": 0.00017544342193570857, + "loss": 2.5732, + "step": 10774 + }, + { + "epoch": 0.46, + "grad_norm": 0.451171875, + "learning_rate": 0.00017543901884298327, + "loss": 2.7349, + "step": 10775 + }, + { + "epoch": 0.46, + "grad_norm": 0.494140625, + "learning_rate": 0.00017543461541080996, + "loss": 2.657, + "step": 10776 + }, + { + "epoch": 0.46, + "grad_norm": 0.4296875, + "learning_rate": 0.00017543021163920846, + "loss": 2.5042, + "step": 10777 + }, + { + "epoch": 0.46, + "grad_norm": 0.478515625, + "learning_rate": 0.00017542580752819852, + "loss": 2.7036, + "step": 10778 + }, + { + "epoch": 0.46, + "grad_norm": 0.50390625, + "learning_rate": 0.00017542140307780007, + "loss": 2.7004, + "step": 10779 + }, + { + "epoch": 0.46, + "grad_norm": 0.431640625, + "learning_rate": 0.00017541699828803285, + "loss": 2.5745, + "step": 10780 + }, + { + "epoch": 0.46, + "grad_norm": 0.5, + "learning_rate": 0.00017541259315891669, + "loss": 2.5963, + "step": 10781 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.0001754081876904714, + "loss": 2.5647, + "step": 10782 + }, + { + "epoch": 0.46, + "grad_norm": 0.439453125, + "learning_rate": 0.00017540378188271686, + "loss": 2.6127, + "step": 10783 + }, + { + "epoch": 0.46, + "grad_norm": 0.37890625, + "learning_rate": 0.00017539937573567285, + "loss": 2.6029, + "step": 10784 + }, + { + "epoch": 0.46, + "grad_norm": 0.46875, + "learning_rate": 0.0001753949692493592, + "loss": 2.652, + "step": 10785 + }, + { + "epoch": 0.46, + "grad_norm": 0.515625, + "learning_rate": 0.00017539056242379575, + "loss": 2.5971, + "step": 10786 + }, + { + "epoch": 0.46, + "grad_norm": 0.470703125, + "learning_rate": 0.0001753861552590023, + "loss": 2.7507, + "step": 10787 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017538174775499874, + "loss": 2.7081, + "step": 10788 + }, + { + "epoch": 0.46, + "grad_norm": 0.486328125, + "learning_rate": 0.00017537733991180484, + "loss": 2.5616, + "step": 10789 + }, + { + "epoch": 0.46, + "grad_norm": 0.474609375, + "learning_rate": 0.00017537293172944046, + "loss": 2.6579, + "step": 10790 + }, + { + "epoch": 0.46, + "grad_norm": 0.4140625, + "learning_rate": 0.0001753685232079254, + "loss": 2.7191, + "step": 10791 + }, + { + "epoch": 0.46, + "grad_norm": 0.4765625, + "learning_rate": 0.00017536411434727957, + "loss": 2.5925, + "step": 10792 + }, + { + "epoch": 0.46, + "grad_norm": 0.486328125, + "learning_rate": 0.00017535970514752277, + "loss": 2.6336, + "step": 10793 + }, + { + "epoch": 0.46, + "grad_norm": 0.494140625, + "learning_rate": 0.0001753552956086748, + "loss": 2.7215, + "step": 10794 + }, + { + "epoch": 0.46, + "grad_norm": 0.423828125, + "learning_rate": 0.00017535088573075557, + "loss": 2.5456, + "step": 10795 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.0001753464755137849, + "loss": 2.5786, + "step": 10796 + }, + { + "epoch": 0.46, + "grad_norm": 0.453125, + "learning_rate": 0.00017534206495778263, + "loss": 2.678, + "step": 10797 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.00017533765406276856, + "loss": 2.6379, + "step": 10798 + }, + { + "epoch": 0.46, + "grad_norm": 0.47265625, + "learning_rate": 0.00017533324282876261, + "loss": 2.4801, + "step": 10799 + }, + { + "epoch": 0.46, + "grad_norm": 0.48046875, + "learning_rate": 0.0001753288312557846, + "loss": 2.5725, + "step": 10800 + }, + { + "epoch": 0.46, + "grad_norm": 0.439453125, + "learning_rate": 0.00017532441934385438, + "loss": 2.6703, + "step": 10801 + }, + { + "epoch": 0.46, + "grad_norm": 0.458984375, + "learning_rate": 0.0001753200070929918, + "loss": 2.3831, + "step": 10802 + }, + { + "epoch": 0.46, + "grad_norm": 0.443359375, + "learning_rate": 0.00017531559450321667, + "loss": 2.498, + "step": 10803 + }, + { + "epoch": 0.46, + "grad_norm": 0.470703125, + "learning_rate": 0.00017531118157454895, + "loss": 2.6756, + "step": 10804 + }, + { + "epoch": 0.46, + "grad_norm": 0.50390625, + "learning_rate": 0.00017530676830700842, + "loss": 2.6196, + "step": 10805 + }, + { + "epoch": 0.46, + "grad_norm": 0.40234375, + "learning_rate": 0.00017530235470061495, + "loss": 2.6301, + "step": 10806 + }, + { + "epoch": 0.46, + "grad_norm": 0.46875, + "learning_rate": 0.0001752979407553884, + "loss": 2.6091, + "step": 10807 + }, + { + "epoch": 0.46, + "grad_norm": 0.44921875, + "learning_rate": 0.00017529352647134862, + "loss": 2.6597, + "step": 10808 + }, + { + "epoch": 0.46, + "grad_norm": 0.458984375, + "learning_rate": 0.0001752891118485155, + "loss": 2.7093, + "step": 10809 + }, + { + "epoch": 0.46, + "grad_norm": 0.4765625, + "learning_rate": 0.00017528469688690888, + "loss": 2.6554, + "step": 10810 + }, + { + "epoch": 0.46, + "grad_norm": 0.470703125, + "learning_rate": 0.0001752802815865487, + "loss": 2.6656, + "step": 10811 + }, + { + "epoch": 0.46, + "grad_norm": 0.51171875, + "learning_rate": 0.0001752758659474547, + "loss": 2.6222, + "step": 10812 + }, + { + "epoch": 0.46, + "grad_norm": 0.490234375, + "learning_rate": 0.00017527144996964684, + "loss": 2.7097, + "step": 10813 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.00017526703365314494, + "loss": 2.538, + "step": 10814 + }, + { + "epoch": 0.46, + "grad_norm": 0.39453125, + "learning_rate": 0.00017526261699796891, + "loss": 2.6114, + "step": 10815 + }, + { + "epoch": 0.46, + "grad_norm": 0.5078125, + "learning_rate": 0.0001752582000041386, + "loss": 2.6061, + "step": 10816 + }, + { + "epoch": 0.46, + "grad_norm": 0.5, + "learning_rate": 0.00017525378267167391, + "loss": 2.6376, + "step": 10817 + }, + { + "epoch": 0.46, + "grad_norm": 0.4453125, + "learning_rate": 0.00017524936500059468, + "loss": 2.5776, + "step": 10818 + }, + { + "epoch": 0.46, + "grad_norm": 0.50390625, + "learning_rate": 0.00017524494699092084, + "loss": 2.6993, + "step": 10819 + }, + { + "epoch": 0.46, + "grad_norm": 0.466796875, + "learning_rate": 0.0001752405286426722, + "loss": 2.6475, + "step": 10820 + }, + { + "epoch": 0.46, + "grad_norm": 0.46484375, + "learning_rate": 0.0001752361099558687, + "loss": 2.5699, + "step": 10821 + }, + { + "epoch": 0.46, + "grad_norm": 0.421875, + "learning_rate": 0.0001752316909305302, + "loss": 2.6811, + "step": 10822 + }, + { + "epoch": 0.46, + "grad_norm": 0.48828125, + "learning_rate": 0.0001752272715666766, + "loss": 2.7754, + "step": 10823 + }, + { + "epoch": 0.46, + "grad_norm": 0.455078125, + "learning_rate": 0.00017522285186432775, + "loss": 2.6359, + "step": 10824 + }, + { + "epoch": 0.46, + "grad_norm": 0.451171875, + "learning_rate": 0.00017521843182350354, + "loss": 2.58, + "step": 10825 + }, + { + "epoch": 0.46, + "grad_norm": 0.451171875, + "learning_rate": 0.0001752140114442239, + "loss": 2.6015, + "step": 10826 + }, + { + "epoch": 0.46, + "grad_norm": 0.45703125, + "learning_rate": 0.00017520959072650868, + "loss": 2.6057, + "step": 10827 + }, + { + "epoch": 0.46, + "grad_norm": 0.466796875, + "learning_rate": 0.00017520516967037782, + "loss": 2.7184, + "step": 10828 + }, + { + "epoch": 0.46, + "grad_norm": 0.45703125, + "learning_rate": 0.00017520074827585117, + "loss": 2.659, + "step": 10829 + }, + { + "epoch": 0.46, + "grad_norm": 0.41796875, + "learning_rate": 0.0001751963265429486, + "loss": 2.6134, + "step": 10830 + }, + { + "epoch": 0.46, + "grad_norm": 0.40234375, + "learning_rate": 0.00017519190447169006, + "loss": 2.636, + "step": 10831 + }, + { + "epoch": 0.46, + "grad_norm": 0.5, + "learning_rate": 0.00017518748206209544, + "loss": 2.5978, + "step": 10832 + }, + { + "epoch": 0.46, + "grad_norm": 0.392578125, + "learning_rate": 0.00017518305931418465, + "loss": 2.6059, + "step": 10833 + }, + { + "epoch": 0.46, + "grad_norm": 0.451171875, + "learning_rate": 0.0001751786362279775, + "loss": 2.6835, + "step": 10834 + }, + { + "epoch": 0.46, + "grad_norm": 0.51171875, + "learning_rate": 0.000175174212803494, + "loss": 2.6752, + "step": 10835 + }, + { + "epoch": 0.46, + "grad_norm": 0.4140625, + "learning_rate": 0.00017516978904075407, + "loss": 2.5516, + "step": 10836 + }, + { + "epoch": 0.46, + "grad_norm": 0.5078125, + "learning_rate": 0.00017516536493977747, + "loss": 2.7428, + "step": 10837 + }, + { + "epoch": 0.46, + "grad_norm": 0.4765625, + "learning_rate": 0.00017516094050058428, + "loss": 2.6951, + "step": 10838 + }, + { + "epoch": 0.46, + "grad_norm": 0.43359375, + "learning_rate": 0.00017515651572319428, + "loss": 2.6168, + "step": 10839 + }, + { + "epoch": 0.46, + "grad_norm": 0.484375, + "learning_rate": 0.00017515209060762742, + "loss": 2.6088, + "step": 10840 + }, + { + "epoch": 0.46, + "grad_norm": 0.453125, + "learning_rate": 0.00017514766515390364, + "loss": 2.4952, + "step": 10841 + }, + { + "epoch": 0.46, + "grad_norm": 0.451171875, + "learning_rate": 0.00017514323936204283, + "loss": 2.6103, + "step": 10842 + }, + { + "epoch": 0.46, + "grad_norm": 0.435546875, + "learning_rate": 0.0001751388132320649, + "loss": 2.763, + "step": 10843 + }, + { + "epoch": 0.46, + "grad_norm": 0.421875, + "learning_rate": 0.00017513438676398978, + "loss": 2.5857, + "step": 10844 + }, + { + "epoch": 0.46, + "grad_norm": 0.51953125, + "learning_rate": 0.0001751299599578373, + "loss": 2.5092, + "step": 10845 + }, + { + "epoch": 0.46, + "grad_norm": 0.48046875, + "learning_rate": 0.00017512553281362757, + "loss": 2.6712, + "step": 10846 + }, + { + "epoch": 0.46, + "grad_norm": 0.5234375, + "learning_rate": 0.00017512110533138034, + "loss": 2.7541, + "step": 10847 + }, + { + "epoch": 0.46, + "grad_norm": 0.443359375, + "learning_rate": 0.00017511667751111558, + "loss": 2.6834, + "step": 10848 + }, + { + "epoch": 0.46, + "grad_norm": 0.443359375, + "learning_rate": 0.00017511224935285325, + "loss": 2.6007, + "step": 10849 + }, + { + "epoch": 0.46, + "grad_norm": 0.5234375, + "learning_rate": 0.0001751078208566132, + "loss": 2.6536, + "step": 10850 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017510339202241547, + "loss": 2.6975, + "step": 10851 + }, + { + "epoch": 0.46, + "grad_norm": 0.404296875, + "learning_rate": 0.0001750989628502799, + "loss": 2.5884, + "step": 10852 + }, + { + "epoch": 0.46, + "grad_norm": 0.470703125, + "learning_rate": 0.00017509453334022644, + "loss": 2.6931, + "step": 10853 + }, + { + "epoch": 0.46, + "grad_norm": 0.4453125, + "learning_rate": 0.000175090103492275, + "loss": 2.6446, + "step": 10854 + }, + { + "epoch": 0.46, + "grad_norm": 0.494140625, + "learning_rate": 0.00017508567330644557, + "loss": 2.6043, + "step": 10855 + }, + { + "epoch": 0.46, + "grad_norm": 0.384765625, + "learning_rate": 0.00017508124278275804, + "loss": 2.4962, + "step": 10856 + }, + { + "epoch": 0.46, + "grad_norm": 0.43359375, + "learning_rate": 0.00017507681192123234, + "loss": 2.4759, + "step": 10857 + }, + { + "epoch": 0.46, + "grad_norm": 0.490234375, + "learning_rate": 0.00017507238072188844, + "loss": 2.7109, + "step": 10858 + }, + { + "epoch": 0.46, + "grad_norm": 0.5, + "learning_rate": 0.0001750679491847463, + "loss": 2.5684, + "step": 10859 + }, + { + "epoch": 0.46, + "grad_norm": 0.45703125, + "learning_rate": 0.00017506351730982577, + "loss": 2.4999, + "step": 10860 + }, + { + "epoch": 0.46, + "grad_norm": 0.4375, + "learning_rate": 0.00017505908509714686, + "loss": 2.5611, + "step": 10861 + }, + { + "epoch": 0.46, + "grad_norm": 0.474609375, + "learning_rate": 0.0001750546525467295, + "loss": 2.6764, + "step": 10862 + }, + { + "epoch": 0.46, + "grad_norm": 0.4453125, + "learning_rate": 0.00017505021965859364, + "loss": 2.4188, + "step": 10863 + }, + { + "epoch": 0.46, + "grad_norm": 0.447265625, + "learning_rate": 0.0001750457864327592, + "loss": 2.5708, + "step": 10864 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017504135286924613, + "loss": 2.6418, + "step": 10865 + }, + { + "epoch": 0.46, + "grad_norm": 0.515625, + "learning_rate": 0.00017503691896807446, + "loss": 2.6601, + "step": 10866 + }, + { + "epoch": 0.46, + "grad_norm": 0.49609375, + "learning_rate": 0.00017503248472926404, + "loss": 2.5837, + "step": 10867 + }, + { + "epoch": 0.46, + "grad_norm": 0.4140625, + "learning_rate": 0.00017502805015283485, + "loss": 2.5689, + "step": 10868 + }, + { + "epoch": 0.46, + "grad_norm": 0.439453125, + "learning_rate": 0.00017502361523880685, + "loss": 2.592, + "step": 10869 + }, + { + "epoch": 0.46, + "grad_norm": 0.439453125, + "learning_rate": 0.00017501917998720003, + "loss": 2.6126, + "step": 10870 + }, + { + "epoch": 0.46, + "grad_norm": 0.435546875, + "learning_rate": 0.00017501474439803432, + "loss": 2.387, + "step": 10871 + }, + { + "epoch": 0.46, + "grad_norm": 0.48046875, + "learning_rate": 0.00017501030847132964, + "loss": 2.5726, + "step": 10872 + }, + { + "epoch": 0.46, + "grad_norm": 0.453125, + "learning_rate": 0.00017500587220710601, + "loss": 2.5666, + "step": 10873 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017500143560538335, + "loss": 2.6436, + "step": 10874 + }, + { + "epoch": 0.46, + "grad_norm": 0.494140625, + "learning_rate": 0.00017499699866618164, + "loss": 2.9126, + "step": 10875 + }, + { + "epoch": 0.46, + "grad_norm": 0.484375, + "learning_rate": 0.00017499256138952087, + "loss": 2.7521, + "step": 10876 + }, + { + "epoch": 0.46, + "grad_norm": 0.453125, + "learning_rate": 0.00017498812377542096, + "loss": 2.5734, + "step": 10877 + }, + { + "epoch": 0.46, + "grad_norm": 0.4140625, + "learning_rate": 0.0001749836858239019, + "loss": 2.6487, + "step": 10878 + }, + { + "epoch": 0.46, + "grad_norm": 0.5078125, + "learning_rate": 0.00017497924753498366, + "loss": 2.652, + "step": 10879 + }, + { + "epoch": 0.46, + "grad_norm": 0.396484375, + "learning_rate": 0.00017497480890868618, + "loss": 2.4917, + "step": 10880 + }, + { + "epoch": 0.46, + "grad_norm": 0.421875, + "learning_rate": 0.00017497036994502949, + "loss": 2.5626, + "step": 10881 + }, + { + "epoch": 0.46, + "grad_norm": 0.4765625, + "learning_rate": 0.00017496593064403356, + "loss": 2.6503, + "step": 10882 + }, + { + "epoch": 0.46, + "grad_norm": 0.474609375, + "learning_rate": 0.00017496149100571827, + "loss": 2.658, + "step": 10883 + }, + { + "epoch": 0.46, + "grad_norm": 0.470703125, + "learning_rate": 0.00017495705103010373, + "loss": 2.6948, + "step": 10884 + }, + { + "epoch": 0.46, + "grad_norm": 0.435546875, + "learning_rate": 0.00017495261071720982, + "loss": 2.6501, + "step": 10885 + }, + { + "epoch": 0.46, + "grad_norm": 0.419921875, + "learning_rate": 0.00017494817006705657, + "loss": 2.5509, + "step": 10886 + }, + { + "epoch": 0.46, + "grad_norm": 0.5, + "learning_rate": 0.00017494372907966393, + "loss": 2.7043, + "step": 10887 + }, + { + "epoch": 0.46, + "grad_norm": 0.486328125, + "learning_rate": 0.0001749392877550519, + "loss": 2.587, + "step": 10888 + }, + { + "epoch": 0.46, + "grad_norm": 0.404296875, + "learning_rate": 0.00017493484609324046, + "loss": 2.5528, + "step": 10889 + }, + { + "epoch": 0.46, + "grad_norm": 0.498046875, + "learning_rate": 0.0001749304040942496, + "loss": 2.7052, + "step": 10890 + }, + { + "epoch": 0.46, + "grad_norm": 0.447265625, + "learning_rate": 0.00017492596175809933, + "loss": 2.5661, + "step": 10891 + }, + { + "epoch": 0.46, + "grad_norm": 0.458984375, + "learning_rate": 0.00017492151908480962, + "loss": 2.7126, + "step": 10892 + }, + { + "epoch": 0.46, + "grad_norm": 0.4609375, + "learning_rate": 0.00017491707607440043, + "loss": 2.7694, + "step": 10893 + }, + { + "epoch": 0.46, + "grad_norm": 0.48828125, + "learning_rate": 0.00017491263272689176, + "loss": 2.567, + "step": 10894 + }, + { + "epoch": 0.46, + "grad_norm": 0.482421875, + "learning_rate": 0.00017490818904230363, + "loss": 2.5961, + "step": 10895 + }, + { + "epoch": 0.46, + "grad_norm": 0.427734375, + "learning_rate": 0.00017490374502065602, + "loss": 2.6004, + "step": 10896 + }, + { + "epoch": 0.46, + "grad_norm": 0.455078125, + "learning_rate": 0.00017489930066196896, + "loss": 2.6137, + "step": 10897 + }, + { + "epoch": 0.46, + "grad_norm": 0.48828125, + "learning_rate": 0.0001748948559662624, + "loss": 2.6879, + "step": 10898 + }, + { + "epoch": 0.46, + "grad_norm": 0.423828125, + "learning_rate": 0.00017489041093355636, + "loss": 2.7208, + "step": 10899 + }, + { + "epoch": 0.46, + "grad_norm": 0.462890625, + "learning_rate": 0.00017488596556387086, + "loss": 2.5775, + "step": 10900 + }, + { + "epoch": 0.46, + "grad_norm": 0.4453125, + "learning_rate": 0.00017488151985722584, + "loss": 2.6946, + "step": 10901 + }, + { + "epoch": 0.46, + "grad_norm": 0.51953125, + "learning_rate": 0.00017487707381364138, + "loss": 2.6969, + "step": 10902 + }, + { + "epoch": 0.46, + "grad_norm": 0.47265625, + "learning_rate": 0.00017487262743313747, + "loss": 2.6879, + "step": 10903 + }, + { + "epoch": 0.46, + "grad_norm": 0.40625, + "learning_rate": 0.00017486818071573405, + "loss": 2.4854, + "step": 10904 + }, + { + "epoch": 0.46, + "grad_norm": 0.46484375, + "learning_rate": 0.00017486373366145118, + "loss": 2.6438, + "step": 10905 + }, + { + "epoch": 0.46, + "grad_norm": 0.462890625, + "learning_rate": 0.0001748592862703089, + "loss": 2.6348, + "step": 10906 + }, + { + "epoch": 0.46, + "grad_norm": 0.41796875, + "learning_rate": 0.0001748548385423272, + "loss": 2.5624, + "step": 10907 + }, + { + "epoch": 0.46, + "grad_norm": 0.4765625, + "learning_rate": 0.00017485039047752604, + "loss": 2.6721, + "step": 10908 + }, + { + "epoch": 0.46, + "grad_norm": 0.458984375, + "learning_rate": 0.0001748459420759255, + "loss": 2.5654, + "step": 10909 + }, + { + "epoch": 0.46, + "grad_norm": 0.5234375, + "learning_rate": 0.00017484149333754558, + "loss": 2.7091, + "step": 10910 + }, + { + "epoch": 0.46, + "grad_norm": 0.470703125, + "learning_rate": 0.00017483704426240627, + "loss": 2.6011, + "step": 10911 + }, + { + "epoch": 0.46, + "grad_norm": 0.447265625, + "learning_rate": 0.00017483259485052765, + "loss": 2.6894, + "step": 10912 + }, + { + "epoch": 0.47, + "grad_norm": 0.458984375, + "learning_rate": 0.00017482814510192966, + "loss": 2.5464, + "step": 10913 + }, + { + "epoch": 0.47, + "grad_norm": 0.466796875, + "learning_rate": 0.00017482369501663238, + "loss": 2.6011, + "step": 10914 + }, + { + "epoch": 0.47, + "grad_norm": 0.41796875, + "learning_rate": 0.00017481924459465583, + "loss": 2.4083, + "step": 10915 + }, + { + "epoch": 0.47, + "grad_norm": 0.40234375, + "learning_rate": 0.00017481479383602003, + "loss": 2.5639, + "step": 10916 + }, + { + "epoch": 0.47, + "grad_norm": 0.44921875, + "learning_rate": 0.00017481034274074495, + "loss": 2.6831, + "step": 10917 + }, + { + "epoch": 0.47, + "grad_norm": 0.52734375, + "learning_rate": 0.0001748058913088507, + "loss": 2.6331, + "step": 10918 + }, + { + "epoch": 0.47, + "grad_norm": 0.4375, + "learning_rate": 0.00017480143954035727, + "loss": 2.6009, + "step": 10919 + }, + { + "epoch": 0.47, + "grad_norm": 0.4375, + "learning_rate": 0.0001747969874352847, + "loss": 2.6911, + "step": 10920 + }, + { + "epoch": 0.47, + "grad_norm": 0.439453125, + "learning_rate": 0.00017479253499365303, + "loss": 2.6651, + "step": 10921 + }, + { + "epoch": 0.47, + "grad_norm": 0.455078125, + "learning_rate": 0.00017478808221548227, + "loss": 2.6728, + "step": 10922 + }, + { + "epoch": 0.47, + "grad_norm": 0.455078125, + "learning_rate": 0.0001747836291007925, + "loss": 2.4759, + "step": 10923 + }, + { + "epoch": 0.47, + "grad_norm": 0.44140625, + "learning_rate": 0.0001747791756496037, + "loss": 2.6386, + "step": 10924 + }, + { + "epoch": 0.47, + "grad_norm": 0.451171875, + "learning_rate": 0.00017477472186193594, + "loss": 2.6452, + "step": 10925 + }, + { + "epoch": 0.47, + "grad_norm": 0.470703125, + "learning_rate": 0.0001747702677378093, + "loss": 2.7712, + "step": 10926 + }, + { + "epoch": 0.47, + "grad_norm": 0.4140625, + "learning_rate": 0.00017476581327724374, + "loss": 2.6776, + "step": 10927 + }, + { + "epoch": 0.47, + "grad_norm": 0.421875, + "learning_rate": 0.00017476135848025937, + "loss": 2.7535, + "step": 10928 + }, + { + "epoch": 0.47, + "grad_norm": 0.431640625, + "learning_rate": 0.0001747569033468762, + "loss": 2.6568, + "step": 10929 + }, + { + "epoch": 0.47, + "grad_norm": 0.52734375, + "learning_rate": 0.00017475244787711427, + "loss": 2.5444, + "step": 10930 + }, + { + "epoch": 0.47, + "grad_norm": 0.4375, + "learning_rate": 0.00017474799207099366, + "loss": 2.6512, + "step": 10931 + }, + { + "epoch": 0.47, + "grad_norm": 0.4375, + "learning_rate": 0.0001747435359285344, + "loss": 2.6137, + "step": 10932 + }, + { + "epoch": 0.47, + "grad_norm": 0.416015625, + "learning_rate": 0.00017473907944975654, + "loss": 2.4598, + "step": 10933 + }, + { + "epoch": 0.47, + "grad_norm": 0.484375, + "learning_rate": 0.00017473462263468013, + "loss": 2.5303, + "step": 10934 + }, + { + "epoch": 0.47, + "grad_norm": 0.50390625, + "learning_rate": 0.00017473016548332523, + "loss": 2.6808, + "step": 10935 + }, + { + "epoch": 0.47, + "grad_norm": 0.431640625, + "learning_rate": 0.0001747257079957119, + "loss": 2.6147, + "step": 10936 + }, + { + "epoch": 0.47, + "grad_norm": 0.5078125, + "learning_rate": 0.0001747212501718602, + "loss": 2.5581, + "step": 10937 + }, + { + "epoch": 0.47, + "grad_norm": 0.484375, + "learning_rate": 0.00017471679201179016, + "loss": 2.6573, + "step": 10938 + }, + { + "epoch": 0.47, + "grad_norm": 0.4375, + "learning_rate": 0.00017471233351552188, + "loss": 2.7923, + "step": 10939 + }, + { + "epoch": 0.47, + "grad_norm": 0.408203125, + "learning_rate": 0.00017470787468307542, + "loss": 2.599, + "step": 10940 + }, + { + "epoch": 0.47, + "grad_norm": 0.451171875, + "learning_rate": 0.0001747034155144708, + "loss": 2.6343, + "step": 10941 + }, + { + "epoch": 0.47, + "grad_norm": 0.431640625, + "learning_rate": 0.00017469895600972813, + "loss": 2.6043, + "step": 10942 + }, + { + "epoch": 0.47, + "grad_norm": 0.46875, + "learning_rate": 0.00017469449616886744, + "loss": 2.5816, + "step": 10943 + }, + { + "epoch": 0.47, + "grad_norm": 0.4609375, + "learning_rate": 0.0001746900359919088, + "loss": 2.512, + "step": 10944 + }, + { + "epoch": 0.47, + "grad_norm": 0.462890625, + "learning_rate": 0.00017468557547887232, + "loss": 2.5404, + "step": 10945 + }, + { + "epoch": 0.47, + "grad_norm": 0.44921875, + "learning_rate": 0.00017468111462977804, + "loss": 2.8093, + "step": 10946 + }, + { + "epoch": 0.47, + "grad_norm": 0.43359375, + "learning_rate": 0.00017467665344464604, + "loss": 2.6773, + "step": 10947 + }, + { + "epoch": 0.47, + "grad_norm": 0.46875, + "learning_rate": 0.0001746721919234964, + "loss": 2.6997, + "step": 10948 + }, + { + "epoch": 0.47, + "grad_norm": 0.466796875, + "learning_rate": 0.00017466773006634919, + "loss": 2.6339, + "step": 10949 + }, + { + "epoch": 0.47, + "grad_norm": 0.4453125, + "learning_rate": 0.00017466326787322447, + "loss": 2.5933, + "step": 10950 + }, + { + "epoch": 0.47, + "grad_norm": 0.4140625, + "learning_rate": 0.00017465880534414232, + "loss": 2.6496, + "step": 10951 + }, + { + "epoch": 0.47, + "grad_norm": 0.423828125, + "learning_rate": 0.00017465434247912283, + "loss": 2.4627, + "step": 10952 + }, + { + "epoch": 0.47, + "grad_norm": 0.51171875, + "learning_rate": 0.0001746498792781861, + "loss": 2.5756, + "step": 10953 + }, + { + "epoch": 0.47, + "grad_norm": 0.447265625, + "learning_rate": 0.00017464541574135217, + "loss": 2.7057, + "step": 10954 + }, + { + "epoch": 0.47, + "grad_norm": 0.455078125, + "learning_rate": 0.00017464095186864117, + "loss": 2.6174, + "step": 10955 + }, + { + "epoch": 0.47, + "grad_norm": 0.4453125, + "learning_rate": 0.00017463648766007317, + "loss": 2.5114, + "step": 10956 + }, + { + "epoch": 0.47, + "grad_norm": 0.462890625, + "learning_rate": 0.0001746320231156682, + "loss": 2.4549, + "step": 10957 + }, + { + "epoch": 0.47, + "grad_norm": 0.4765625, + "learning_rate": 0.00017462755823544648, + "loss": 2.5171, + "step": 10958 + }, + { + "epoch": 0.47, + "grad_norm": 0.48828125, + "learning_rate": 0.000174623093019428, + "loss": 2.5309, + "step": 10959 + }, + { + "epoch": 0.47, + "grad_norm": 0.4296875, + "learning_rate": 0.00017461862746763284, + "loss": 2.5993, + "step": 10960 + }, + { + "epoch": 0.47, + "grad_norm": 0.439453125, + "learning_rate": 0.00017461416158008115, + "loss": 2.5828, + "step": 10961 + }, + { + "epoch": 0.47, + "grad_norm": 0.421875, + "learning_rate": 0.000174609695356793, + "loss": 2.6484, + "step": 10962 + }, + { + "epoch": 0.47, + "grad_norm": 0.470703125, + "learning_rate": 0.00017460522879778845, + "loss": 2.6634, + "step": 10963 + }, + { + "epoch": 0.47, + "grad_norm": 0.42578125, + "learning_rate": 0.0001746007619030877, + "loss": 2.498, + "step": 10964 + }, + { + "epoch": 0.47, + "grad_norm": 0.44140625, + "learning_rate": 0.00017459629467271075, + "loss": 2.6254, + "step": 10965 + }, + { + "epoch": 0.47, + "grad_norm": 0.4375, + "learning_rate": 0.00017459182710667773, + "loss": 2.5248, + "step": 10966 + }, + { + "epoch": 0.47, + "grad_norm": 0.41796875, + "learning_rate": 0.00017458735920500876, + "loss": 2.6353, + "step": 10967 + }, + { + "epoch": 0.47, + "grad_norm": 0.490234375, + "learning_rate": 0.00017458289096772395, + "loss": 2.4913, + "step": 10968 + }, + { + "epoch": 0.47, + "grad_norm": 0.47265625, + "learning_rate": 0.00017457842239484335, + "loss": 2.4582, + "step": 10969 + }, + { + "epoch": 0.47, + "grad_norm": 0.447265625, + "learning_rate": 0.0001745739534863871, + "loss": 2.5966, + "step": 10970 + }, + { + "epoch": 0.47, + "grad_norm": 0.400390625, + "learning_rate": 0.00017456948424237533, + "loss": 2.5334, + "step": 10971 + }, + { + "epoch": 0.47, + "grad_norm": 0.455078125, + "learning_rate": 0.00017456501466282814, + "loss": 2.6346, + "step": 10972 + }, + { + "epoch": 0.47, + "grad_norm": 0.4296875, + "learning_rate": 0.00017456054474776565, + "loss": 2.6255, + "step": 10973 + }, + { + "epoch": 0.47, + "grad_norm": 0.423828125, + "learning_rate": 0.00017455607449720794, + "loss": 2.4312, + "step": 10974 + }, + { + "epoch": 0.47, + "grad_norm": 0.42578125, + "learning_rate": 0.0001745516039111751, + "loss": 2.5425, + "step": 10975 + }, + { + "epoch": 0.47, + "grad_norm": 0.427734375, + "learning_rate": 0.00017454713298968737, + "loss": 2.4535, + "step": 10976 + }, + { + "epoch": 0.47, + "grad_norm": 0.45703125, + "learning_rate": 0.00017454266173276475, + "loss": 2.792, + "step": 10977 + }, + { + "epoch": 0.47, + "grad_norm": 0.412109375, + "learning_rate": 0.00017453819014042737, + "loss": 2.5198, + "step": 10978 + }, + { + "epoch": 0.47, + "grad_norm": 0.50390625, + "learning_rate": 0.00017453371821269536, + "loss": 2.673, + "step": 10979 + }, + { + "epoch": 0.47, + "grad_norm": 0.443359375, + "learning_rate": 0.0001745292459495889, + "loss": 2.4851, + "step": 10980 + }, + { + "epoch": 0.47, + "grad_norm": 0.4296875, + "learning_rate": 0.00017452477335112807, + "loss": 2.5122, + "step": 10981 + }, + { + "epoch": 0.47, + "grad_norm": 0.408203125, + "learning_rate": 0.000174520300417333, + "loss": 2.5414, + "step": 10982 + }, + { + "epoch": 0.47, + "grad_norm": 0.478515625, + "learning_rate": 0.00017451582714822378, + "loss": 2.6067, + "step": 10983 + }, + { + "epoch": 0.47, + "grad_norm": 0.4375, + "learning_rate": 0.0001745113535438206, + "loss": 2.6486, + "step": 10984 + }, + { + "epoch": 0.47, + "grad_norm": 0.462890625, + "learning_rate": 0.00017450687960414354, + "loss": 2.5957, + "step": 10985 + }, + { + "epoch": 0.47, + "grad_norm": 0.431640625, + "learning_rate": 0.00017450240532921276, + "loss": 2.5864, + "step": 10986 + }, + { + "epoch": 0.47, + "grad_norm": 0.47265625, + "learning_rate": 0.00017449793071904837, + "loss": 2.5317, + "step": 10987 + }, + { + "epoch": 0.47, + "grad_norm": 0.443359375, + "learning_rate": 0.00017449345577367055, + "loss": 2.5557, + "step": 10988 + }, + { + "epoch": 0.47, + "grad_norm": 0.435546875, + "learning_rate": 0.00017448898049309938, + "loss": 2.7005, + "step": 10989 + }, + { + "epoch": 0.47, + "grad_norm": 0.455078125, + "learning_rate": 0.00017448450487735504, + "loss": 2.6523, + "step": 10990 + }, + { + "epoch": 0.47, + "grad_norm": 0.45703125, + "learning_rate": 0.0001744800289264576, + "loss": 2.5944, + "step": 10991 + }, + { + "epoch": 0.47, + "grad_norm": 0.49609375, + "learning_rate": 0.0001744755526404273, + "loss": 2.6744, + "step": 10992 + }, + { + "epoch": 0.47, + "grad_norm": 0.443359375, + "learning_rate": 0.0001744710760192842, + "loss": 2.63, + "step": 10993 + }, + { + "epoch": 0.47, + "grad_norm": 0.431640625, + "learning_rate": 0.0001744665990630485, + "loss": 2.6562, + "step": 10994 + }, + { + "epoch": 0.47, + "grad_norm": 0.5, + "learning_rate": 0.0001744621217717403, + "loss": 2.6605, + "step": 10995 + }, + { + "epoch": 0.47, + "grad_norm": 0.404296875, + "learning_rate": 0.00017445764414537976, + "loss": 2.5287, + "step": 10996 + }, + { + "epoch": 0.47, + "grad_norm": 0.451171875, + "learning_rate": 0.00017445316618398707, + "loss": 2.5071, + "step": 10997 + }, + { + "epoch": 0.47, + "grad_norm": 0.50390625, + "learning_rate": 0.0001744486878875823, + "loss": 2.5814, + "step": 10998 + }, + { + "epoch": 0.47, + "grad_norm": 0.42578125, + "learning_rate": 0.00017444420925618565, + "loss": 2.4896, + "step": 10999 + }, + { + "epoch": 0.47, + "grad_norm": 0.4453125, + "learning_rate": 0.00017443973028981728, + "loss": 2.639, + "step": 11000 + }, + { + "epoch": 0.47, + "grad_norm": 0.412109375, + "learning_rate": 0.0001744352509884973, + "loss": 2.5358, + "step": 11001 + }, + { + "epoch": 0.47, + "grad_norm": 0.546875, + "learning_rate": 0.00017443077135224596, + "loss": 2.6758, + "step": 11002 + }, + { + "epoch": 0.47, + "grad_norm": 0.490234375, + "learning_rate": 0.00017442629138108328, + "loss": 2.5649, + "step": 11003 + }, + { + "epoch": 0.47, + "grad_norm": 0.435546875, + "learning_rate": 0.0001744218110750295, + "loss": 2.5765, + "step": 11004 + }, + { + "epoch": 0.47, + "grad_norm": 0.474609375, + "learning_rate": 0.0001744173304341048, + "loss": 2.5982, + "step": 11005 + }, + { + "epoch": 0.47, + "grad_norm": 0.4296875, + "learning_rate": 0.0001744128494583293, + "loss": 2.6377, + "step": 11006 + }, + { + "epoch": 0.47, + "grad_norm": 0.47265625, + "learning_rate": 0.00017440836814772314, + "loss": 2.624, + "step": 11007 + }, + { + "epoch": 0.47, + "grad_norm": 0.431640625, + "learning_rate": 0.00017440388650230654, + "loss": 2.6968, + "step": 11008 + }, + { + "epoch": 0.47, + "grad_norm": 0.478515625, + "learning_rate": 0.00017439940452209965, + "loss": 2.7164, + "step": 11009 + }, + { + "epoch": 0.47, + "grad_norm": 0.4375, + "learning_rate": 0.00017439492220712262, + "loss": 2.6933, + "step": 11010 + }, + { + "epoch": 0.47, + "grad_norm": 0.45703125, + "learning_rate": 0.0001743904395573956, + "loss": 2.5841, + "step": 11011 + }, + { + "epoch": 0.47, + "grad_norm": 0.45703125, + "learning_rate": 0.00017438595657293883, + "loss": 2.4536, + "step": 11012 + }, + { + "epoch": 0.47, + "grad_norm": 0.5, + "learning_rate": 0.00017438147325377244, + "loss": 2.6469, + "step": 11013 + }, + { + "epoch": 0.47, + "grad_norm": 0.451171875, + "learning_rate": 0.0001743769895999166, + "loss": 2.6049, + "step": 11014 + }, + { + "epoch": 0.47, + "grad_norm": 0.439453125, + "learning_rate": 0.00017437250561139144, + "loss": 2.6726, + "step": 11015 + }, + { + "epoch": 0.47, + "grad_norm": 0.44140625, + "learning_rate": 0.00017436802128821722, + "loss": 2.4939, + "step": 11016 + }, + { + "epoch": 0.47, + "grad_norm": 0.46875, + "learning_rate": 0.00017436353663041412, + "loss": 2.6106, + "step": 11017 + }, + { + "epoch": 0.47, + "grad_norm": 0.443359375, + "learning_rate": 0.00017435905163800227, + "loss": 2.4894, + "step": 11018 + }, + { + "epoch": 0.47, + "grad_norm": 0.47265625, + "learning_rate": 0.00017435456631100183, + "loss": 2.5977, + "step": 11019 + }, + { + "epoch": 0.47, + "grad_norm": 0.49609375, + "learning_rate": 0.00017435008064943302, + "loss": 2.7274, + "step": 11020 + }, + { + "epoch": 0.47, + "grad_norm": 0.51953125, + "learning_rate": 0.00017434559465331605, + "loss": 2.6494, + "step": 11021 + }, + { + "epoch": 0.47, + "grad_norm": 0.44921875, + "learning_rate": 0.000174341108322671, + "loss": 2.7519, + "step": 11022 + }, + { + "epoch": 0.47, + "grad_norm": 0.421875, + "learning_rate": 0.0001743366216575182, + "loss": 2.4804, + "step": 11023 + }, + { + "epoch": 0.47, + "grad_norm": 0.4453125, + "learning_rate": 0.00017433213465787775, + "loss": 2.5045, + "step": 11024 + }, + { + "epoch": 0.47, + "grad_norm": 0.43359375, + "learning_rate": 0.00017432764732376986, + "loss": 2.5461, + "step": 11025 + }, + { + "epoch": 0.47, + "grad_norm": 0.5234375, + "learning_rate": 0.0001743231596552147, + "loss": 2.6135, + "step": 11026 + }, + { + "epoch": 0.47, + "grad_norm": 0.47265625, + "learning_rate": 0.00017431867165223252, + "loss": 2.7271, + "step": 11027 + }, + { + "epoch": 0.47, + "grad_norm": 0.490234375, + "learning_rate": 0.00017431418331484346, + "loss": 2.6681, + "step": 11028 + }, + { + "epoch": 0.47, + "grad_norm": 0.46875, + "learning_rate": 0.00017430969464306774, + "loss": 2.7689, + "step": 11029 + }, + { + "epoch": 0.47, + "grad_norm": 0.39453125, + "learning_rate": 0.00017430520563692557, + "loss": 2.6038, + "step": 11030 + }, + { + "epoch": 0.47, + "grad_norm": 0.439453125, + "learning_rate": 0.0001743007162964371, + "loss": 2.5827, + "step": 11031 + }, + { + "epoch": 0.47, + "grad_norm": 0.5546875, + "learning_rate": 0.00017429622662162254, + "loss": 2.6048, + "step": 11032 + }, + { + "epoch": 0.47, + "grad_norm": 0.392578125, + "learning_rate": 0.00017429173661250214, + "loss": 2.6366, + "step": 11033 + }, + { + "epoch": 0.47, + "grad_norm": 0.453125, + "learning_rate": 0.0001742872462690961, + "loss": 2.6271, + "step": 11034 + }, + { + "epoch": 0.47, + "grad_norm": 0.5078125, + "learning_rate": 0.00017428275559142455, + "loss": 2.6159, + "step": 11035 + }, + { + "epoch": 0.47, + "grad_norm": 0.466796875, + "learning_rate": 0.0001742782645795078, + "loss": 2.7083, + "step": 11036 + }, + { + "epoch": 0.47, + "grad_norm": 0.41015625, + "learning_rate": 0.00017427377323336594, + "loss": 2.4883, + "step": 11037 + }, + { + "epoch": 0.47, + "grad_norm": 0.51171875, + "learning_rate": 0.00017426928155301932, + "loss": 2.6808, + "step": 11038 + }, + { + "epoch": 0.47, + "grad_norm": 0.4453125, + "learning_rate": 0.00017426478953848804, + "loss": 2.6559, + "step": 11039 + }, + { + "epoch": 0.47, + "grad_norm": 0.455078125, + "learning_rate": 0.00017426029718979234, + "loss": 2.6414, + "step": 11040 + }, + { + "epoch": 0.47, + "grad_norm": 0.4765625, + "learning_rate": 0.00017425580450695243, + "loss": 2.6063, + "step": 11041 + }, + { + "epoch": 0.47, + "grad_norm": 0.474609375, + "learning_rate": 0.00017425131148998856, + "loss": 2.5825, + "step": 11042 + }, + { + "epoch": 0.47, + "grad_norm": 0.41015625, + "learning_rate": 0.00017424681813892092, + "loss": 2.6976, + "step": 11043 + }, + { + "epoch": 0.47, + "grad_norm": 0.42578125, + "learning_rate": 0.00017424232445376972, + "loss": 2.5189, + "step": 11044 + }, + { + "epoch": 0.47, + "grad_norm": 0.51953125, + "learning_rate": 0.0001742378304345552, + "loss": 2.6121, + "step": 11045 + }, + { + "epoch": 0.47, + "grad_norm": 0.4609375, + "learning_rate": 0.0001742333360812976, + "loss": 2.7035, + "step": 11046 + }, + { + "epoch": 0.47, + "grad_norm": 0.51171875, + "learning_rate": 0.0001742288413940171, + "loss": 2.5896, + "step": 11047 + }, + { + "epoch": 0.47, + "grad_norm": 0.48046875, + "learning_rate": 0.00017422434637273394, + "loss": 2.6784, + "step": 11048 + }, + { + "epoch": 0.47, + "grad_norm": 0.4921875, + "learning_rate": 0.0001742198510174683, + "loss": 2.6744, + "step": 11049 + }, + { + "epoch": 0.47, + "grad_norm": 0.4296875, + "learning_rate": 0.00017421535532824054, + "loss": 2.6095, + "step": 11050 + }, + { + "epoch": 0.47, + "grad_norm": 0.423828125, + "learning_rate": 0.00017421085930507076, + "loss": 2.5565, + "step": 11051 + }, + { + "epoch": 0.47, + "grad_norm": 0.447265625, + "learning_rate": 0.00017420636294797923, + "loss": 2.6236, + "step": 11052 + }, + { + "epoch": 0.47, + "grad_norm": 0.44140625, + "learning_rate": 0.0001742018662569862, + "loss": 2.5264, + "step": 11053 + }, + { + "epoch": 0.47, + "grad_norm": 0.45703125, + "learning_rate": 0.00017419736923211188, + "loss": 2.5831, + "step": 11054 + }, + { + "epoch": 0.47, + "grad_norm": 0.51953125, + "learning_rate": 0.0001741928718733765, + "loss": 2.5966, + "step": 11055 + }, + { + "epoch": 0.47, + "grad_norm": 0.5, + "learning_rate": 0.0001741883741808003, + "loss": 2.6064, + "step": 11056 + }, + { + "epoch": 0.47, + "grad_norm": 0.4453125, + "learning_rate": 0.0001741838761544036, + "loss": 2.6647, + "step": 11057 + }, + { + "epoch": 0.47, + "grad_norm": 0.4921875, + "learning_rate": 0.0001741793777942065, + "loss": 2.6335, + "step": 11058 + }, + { + "epoch": 0.47, + "grad_norm": 0.431640625, + "learning_rate": 0.00017417487910022933, + "loss": 2.5642, + "step": 11059 + }, + { + "epoch": 0.47, + "grad_norm": 0.458984375, + "learning_rate": 0.0001741703800724923, + "loss": 2.5486, + "step": 11060 + }, + { + "epoch": 0.47, + "grad_norm": 0.404296875, + "learning_rate": 0.00017416588071101572, + "loss": 2.5173, + "step": 11061 + }, + { + "epoch": 0.47, + "grad_norm": 0.46484375, + "learning_rate": 0.0001741613810158197, + "loss": 2.6584, + "step": 11062 + }, + { + "epoch": 0.47, + "grad_norm": 0.46875, + "learning_rate": 0.00017415688098692465, + "loss": 2.6471, + "step": 11063 + }, + { + "epoch": 0.47, + "grad_norm": 0.4296875, + "learning_rate": 0.00017415238062435068, + "loss": 2.6088, + "step": 11064 + }, + { + "epoch": 0.47, + "grad_norm": 0.421875, + "learning_rate": 0.00017414787992811812, + "loss": 2.6024, + "step": 11065 + }, + { + "epoch": 0.47, + "grad_norm": 0.435546875, + "learning_rate": 0.0001741433788982472, + "loss": 2.5761, + "step": 11066 + }, + { + "epoch": 0.47, + "grad_norm": 0.474609375, + "learning_rate": 0.00017413887753475816, + "loss": 2.7108, + "step": 11067 + }, + { + "epoch": 0.47, + "grad_norm": 0.4609375, + "learning_rate": 0.00017413437583767128, + "loss": 2.6333, + "step": 11068 + }, + { + "epoch": 0.47, + "grad_norm": 0.41796875, + "learning_rate": 0.00017412987380700679, + "loss": 2.5049, + "step": 11069 + }, + { + "epoch": 0.47, + "grad_norm": 0.48828125, + "learning_rate": 0.00017412537144278494, + "loss": 2.5596, + "step": 11070 + }, + { + "epoch": 0.47, + "grad_norm": 0.443359375, + "learning_rate": 0.00017412086874502602, + "loss": 2.5281, + "step": 11071 + }, + { + "epoch": 0.47, + "grad_norm": 0.46875, + "learning_rate": 0.00017411636571375028, + "loss": 2.6676, + "step": 11072 + }, + { + "epoch": 0.47, + "grad_norm": 0.451171875, + "learning_rate": 0.000174111862348978, + "loss": 2.57, + "step": 11073 + }, + { + "epoch": 0.47, + "grad_norm": 0.466796875, + "learning_rate": 0.00017410735865072943, + "loss": 2.5141, + "step": 11074 + }, + { + "epoch": 0.47, + "grad_norm": 0.49609375, + "learning_rate": 0.0001741028546190248, + "loss": 2.5362, + "step": 11075 + }, + { + "epoch": 0.47, + "grad_norm": 0.515625, + "learning_rate": 0.0001740983502538844, + "loss": 2.4511, + "step": 11076 + }, + { + "epoch": 0.47, + "grad_norm": 0.453125, + "learning_rate": 0.00017409384555532853, + "loss": 2.5348, + "step": 11077 + }, + { + "epoch": 0.47, + "grad_norm": 0.42578125, + "learning_rate": 0.00017408934052337745, + "loss": 2.5274, + "step": 11078 + }, + { + "epoch": 0.47, + "grad_norm": 0.42578125, + "learning_rate": 0.00017408483515805138, + "loss": 2.6105, + "step": 11079 + }, + { + "epoch": 0.47, + "grad_norm": 0.447265625, + "learning_rate": 0.00017408032945937063, + "loss": 2.6359, + "step": 11080 + }, + { + "epoch": 0.47, + "grad_norm": 0.53515625, + "learning_rate": 0.00017407582342735548, + "loss": 2.5246, + "step": 11081 + }, + { + "epoch": 0.47, + "grad_norm": 0.478515625, + "learning_rate": 0.0001740713170620262, + "loss": 2.4902, + "step": 11082 + }, + { + "epoch": 0.47, + "grad_norm": 0.44140625, + "learning_rate": 0.00017406681036340305, + "loss": 2.6575, + "step": 11083 + }, + { + "epoch": 0.47, + "grad_norm": 0.51171875, + "learning_rate": 0.00017406230333150632, + "loss": 2.6763, + "step": 11084 + }, + { + "epoch": 0.47, + "grad_norm": 0.474609375, + "learning_rate": 0.0001740577959663563, + "loss": 2.6512, + "step": 11085 + }, + { + "epoch": 0.47, + "grad_norm": 0.48828125, + "learning_rate": 0.00017405328826797323, + "loss": 2.6285, + "step": 11086 + }, + { + "epoch": 0.47, + "grad_norm": 0.458984375, + "learning_rate": 0.00017404878023637748, + "loss": 2.6642, + "step": 11087 + }, + { + "epoch": 0.47, + "grad_norm": 0.466796875, + "learning_rate": 0.00017404427187158928, + "loss": 2.5796, + "step": 11088 + }, + { + "epoch": 0.47, + "grad_norm": 0.45703125, + "learning_rate": 0.00017403976317362889, + "loss": 2.5454, + "step": 11089 + }, + { + "epoch": 0.47, + "grad_norm": 0.443359375, + "learning_rate": 0.0001740352541425166, + "loss": 2.6565, + "step": 11090 + }, + { + "epoch": 0.47, + "grad_norm": 0.494140625, + "learning_rate": 0.00017403074477827276, + "loss": 2.6968, + "step": 11091 + }, + { + "epoch": 0.47, + "grad_norm": 0.494140625, + "learning_rate": 0.00017402623508091765, + "loss": 2.6333, + "step": 11092 + }, + { + "epoch": 0.47, + "grad_norm": 0.44921875, + "learning_rate": 0.00017402172505047146, + "loss": 2.7273, + "step": 11093 + }, + { + "epoch": 0.47, + "grad_norm": 0.48046875, + "learning_rate": 0.00017401721468695463, + "loss": 2.845, + "step": 11094 + }, + { + "epoch": 0.47, + "grad_norm": 0.4765625, + "learning_rate": 0.00017401270399038733, + "loss": 2.6204, + "step": 11095 + }, + { + "epoch": 0.47, + "grad_norm": 0.48828125, + "learning_rate": 0.00017400819296078997, + "loss": 2.6135, + "step": 11096 + }, + { + "epoch": 0.47, + "grad_norm": 0.470703125, + "learning_rate": 0.00017400368159818276, + "loss": 2.5241, + "step": 11097 + }, + { + "epoch": 0.47, + "grad_norm": 0.42578125, + "learning_rate": 0.000173999169902586, + "loss": 2.4748, + "step": 11098 + }, + { + "epoch": 0.47, + "grad_norm": 0.458984375, + "learning_rate": 0.00017399465787402006, + "loss": 2.5005, + "step": 11099 + }, + { + "epoch": 0.47, + "grad_norm": 0.458984375, + "learning_rate": 0.00017399014551250515, + "loss": 2.6444, + "step": 11100 + }, + { + "epoch": 0.47, + "grad_norm": 0.48828125, + "learning_rate": 0.00017398563281806166, + "loss": 2.6298, + "step": 11101 + }, + { + "epoch": 0.47, + "grad_norm": 0.455078125, + "learning_rate": 0.00017398111979070988, + "loss": 2.6218, + "step": 11102 + }, + { + "epoch": 0.47, + "grad_norm": 0.447265625, + "learning_rate": 0.00017397660643047008, + "loss": 2.4241, + "step": 11103 + }, + { + "epoch": 0.47, + "grad_norm": 0.44921875, + "learning_rate": 0.00017397209273736257, + "loss": 2.8667, + "step": 11104 + }, + { + "epoch": 0.47, + "grad_norm": 0.46484375, + "learning_rate": 0.00017396757871140768, + "loss": 2.544, + "step": 11105 + }, + { + "epoch": 0.47, + "grad_norm": 0.490234375, + "learning_rate": 0.00017396306435262573, + "loss": 2.6305, + "step": 11106 + }, + { + "epoch": 0.47, + "grad_norm": 0.51953125, + "learning_rate": 0.00017395854966103703, + "loss": 2.7357, + "step": 11107 + }, + { + "epoch": 0.47, + "grad_norm": 0.48046875, + "learning_rate": 0.00017395403463666187, + "loss": 2.6185, + "step": 11108 + }, + { + "epoch": 0.47, + "grad_norm": 0.462890625, + "learning_rate": 0.0001739495192795206, + "loss": 2.5447, + "step": 11109 + }, + { + "epoch": 0.47, + "grad_norm": 0.416015625, + "learning_rate": 0.00017394500358963348, + "loss": 2.4936, + "step": 11110 + }, + { + "epoch": 0.47, + "grad_norm": 0.41015625, + "learning_rate": 0.0001739404875670209, + "loss": 2.6811, + "step": 11111 + }, + { + "epoch": 0.47, + "grad_norm": 0.49609375, + "learning_rate": 0.00017393597121170316, + "loss": 2.5508, + "step": 11112 + }, + { + "epoch": 0.47, + "grad_norm": 0.4609375, + "learning_rate": 0.00017393145452370052, + "loss": 2.4452, + "step": 11113 + }, + { + "epoch": 0.47, + "grad_norm": 0.5078125, + "learning_rate": 0.00017392693750303338, + "loss": 2.6913, + "step": 11114 + }, + { + "epoch": 0.47, + "grad_norm": 0.451171875, + "learning_rate": 0.00017392242014972205, + "loss": 2.6259, + "step": 11115 + }, + { + "epoch": 0.47, + "grad_norm": 0.4140625, + "learning_rate": 0.00017391790246378683, + "loss": 2.6735, + "step": 11116 + }, + { + "epoch": 0.47, + "grad_norm": 0.4609375, + "learning_rate": 0.00017391338444524808, + "loss": 2.5392, + "step": 11117 + }, + { + "epoch": 0.47, + "grad_norm": 0.44921875, + "learning_rate": 0.00017390886609412612, + "loss": 2.5236, + "step": 11118 + }, + { + "epoch": 0.47, + "grad_norm": 0.5078125, + "learning_rate": 0.00017390434741044124, + "loss": 2.7724, + "step": 11119 + }, + { + "epoch": 0.47, + "grad_norm": 0.466796875, + "learning_rate": 0.0001738998283942138, + "loss": 2.6179, + "step": 11120 + }, + { + "epoch": 0.47, + "grad_norm": 0.412109375, + "learning_rate": 0.00017389530904546416, + "loss": 2.6931, + "step": 11121 + }, + { + "epoch": 0.47, + "grad_norm": 0.494140625, + "learning_rate": 0.00017389078936421264, + "loss": 2.6486, + "step": 11122 + }, + { + "epoch": 0.47, + "grad_norm": 0.40234375, + "learning_rate": 0.00017388626935047955, + "loss": 2.6522, + "step": 11123 + }, + { + "epoch": 0.47, + "grad_norm": 0.5078125, + "learning_rate": 0.0001738817490042853, + "loss": 2.5427, + "step": 11124 + }, + { + "epoch": 0.47, + "grad_norm": 0.5078125, + "learning_rate": 0.00017387722832565015, + "loss": 2.6811, + "step": 11125 + }, + { + "epoch": 0.47, + "grad_norm": 0.52734375, + "learning_rate": 0.00017387270731459445, + "loss": 2.5509, + "step": 11126 + }, + { + "epoch": 0.47, + "grad_norm": 0.375, + "learning_rate": 0.0001738681859711386, + "loss": 2.5207, + "step": 11127 + }, + { + "epoch": 0.47, + "grad_norm": 0.431640625, + "learning_rate": 0.0001738636642953029, + "loss": 2.589, + "step": 11128 + }, + { + "epoch": 0.47, + "grad_norm": 0.478515625, + "learning_rate": 0.00017385914228710766, + "loss": 2.6117, + "step": 11129 + }, + { + "epoch": 0.47, + "grad_norm": 0.52734375, + "learning_rate": 0.00017385461994657333, + "loss": 2.6814, + "step": 11130 + }, + { + "epoch": 0.47, + "grad_norm": 0.423828125, + "learning_rate": 0.00017385009727372016, + "loss": 2.6623, + "step": 11131 + }, + { + "epoch": 0.47, + "grad_norm": 0.3984375, + "learning_rate": 0.00017384557426856857, + "loss": 2.5572, + "step": 11132 + }, + { + "epoch": 0.47, + "grad_norm": 0.419921875, + "learning_rate": 0.00017384105093113888, + "loss": 2.5218, + "step": 11133 + }, + { + "epoch": 0.47, + "grad_norm": 0.40625, + "learning_rate": 0.00017383652726145145, + "loss": 2.652, + "step": 11134 + }, + { + "epoch": 0.47, + "grad_norm": 0.37109375, + "learning_rate": 0.0001738320032595266, + "loss": 2.6156, + "step": 11135 + }, + { + "epoch": 0.47, + "grad_norm": 0.453125, + "learning_rate": 0.00017382747892538474, + "loss": 2.7005, + "step": 11136 + }, + { + "epoch": 0.47, + "grad_norm": 0.42578125, + "learning_rate": 0.00017382295425904623, + "loss": 2.5189, + "step": 11137 + }, + { + "epoch": 0.47, + "grad_norm": 0.40234375, + "learning_rate": 0.00017381842926053137, + "loss": 2.5462, + "step": 11138 + }, + { + "epoch": 0.47, + "grad_norm": 0.453125, + "learning_rate": 0.00017381390392986055, + "loss": 2.528, + "step": 11139 + }, + { + "epoch": 0.47, + "grad_norm": 0.4453125, + "learning_rate": 0.00017380937826705418, + "loss": 2.5993, + "step": 11140 + }, + { + "epoch": 0.47, + "grad_norm": 0.470703125, + "learning_rate": 0.00017380485227213254, + "loss": 2.6362, + "step": 11141 + }, + { + "epoch": 0.47, + "grad_norm": 0.466796875, + "learning_rate": 0.00017380032594511605, + "loss": 2.6098, + "step": 11142 + }, + { + "epoch": 0.47, + "grad_norm": 0.455078125, + "learning_rate": 0.0001737957992860251, + "loss": 2.607, + "step": 11143 + }, + { + "epoch": 0.47, + "grad_norm": 0.4453125, + "learning_rate": 0.00017379127229487997, + "loss": 2.6173, + "step": 11144 + }, + { + "epoch": 0.47, + "grad_norm": 0.45703125, + "learning_rate": 0.0001737867449717011, + "loss": 2.5685, + "step": 11145 + }, + { + "epoch": 0.47, + "grad_norm": 0.416015625, + "learning_rate": 0.00017378221731650885, + "loss": 2.664, + "step": 11146 + }, + { + "epoch": 0.48, + "grad_norm": 0.427734375, + "learning_rate": 0.0001737776893293236, + "loss": 2.4693, + "step": 11147 + }, + { + "epoch": 0.48, + "grad_norm": 0.451171875, + "learning_rate": 0.0001737731610101657, + "loss": 2.6951, + "step": 11148 + }, + { + "epoch": 0.48, + "grad_norm": 0.443359375, + "learning_rate": 0.0001737686323590555, + "loss": 2.6945, + "step": 11149 + }, + { + "epoch": 0.48, + "grad_norm": 0.4140625, + "learning_rate": 0.00017376410337601345, + "loss": 2.5207, + "step": 11150 + }, + { + "epoch": 0.48, + "grad_norm": 0.5, + "learning_rate": 0.0001737595740610599, + "loss": 2.7229, + "step": 11151 + }, + { + "epoch": 0.48, + "grad_norm": 0.474609375, + "learning_rate": 0.00017375504441421522, + "loss": 2.6617, + "step": 11152 + }, + { + "epoch": 0.48, + "grad_norm": 0.52734375, + "learning_rate": 0.00017375051443549975, + "loss": 2.6666, + "step": 11153 + }, + { + "epoch": 0.48, + "grad_norm": 0.48046875, + "learning_rate": 0.00017374598412493397, + "loss": 2.5903, + "step": 11154 + }, + { + "epoch": 0.48, + "grad_norm": 0.49609375, + "learning_rate": 0.0001737414534825382, + "loss": 2.671, + "step": 11155 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.00017373692250833282, + "loss": 2.653, + "step": 11156 + }, + { + "epoch": 0.48, + "grad_norm": 0.4375, + "learning_rate": 0.00017373239120233827, + "loss": 2.5973, + "step": 11157 + }, + { + "epoch": 0.48, + "grad_norm": 0.392578125, + "learning_rate": 0.00017372785956457486, + "loss": 2.644, + "step": 11158 + }, + { + "epoch": 0.48, + "grad_norm": 0.46484375, + "learning_rate": 0.00017372332759506306, + "loss": 2.5877, + "step": 11159 + }, + { + "epoch": 0.48, + "grad_norm": 0.46484375, + "learning_rate": 0.00017371879529382323, + "loss": 2.6082, + "step": 11160 + }, + { + "epoch": 0.48, + "grad_norm": 0.4453125, + "learning_rate": 0.00017371426266087574, + "loss": 2.4923, + "step": 11161 + }, + { + "epoch": 0.48, + "grad_norm": 0.48046875, + "learning_rate": 0.000173709729696241, + "loss": 2.7246, + "step": 11162 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.00017370519639993946, + "loss": 2.6346, + "step": 11163 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.00017370066277199144, + "loss": 2.524, + "step": 11164 + }, + { + "epoch": 0.48, + "grad_norm": 0.4296875, + "learning_rate": 0.0001736961288124174, + "loss": 2.6872, + "step": 11165 + }, + { + "epoch": 0.48, + "grad_norm": 0.458984375, + "learning_rate": 0.00017369159452123767, + "loss": 2.6427, + "step": 11166 + }, + { + "epoch": 0.48, + "grad_norm": 0.46875, + "learning_rate": 0.00017368705989847274, + "loss": 2.5377, + "step": 11167 + }, + { + "epoch": 0.48, + "grad_norm": 0.455078125, + "learning_rate": 0.00017368252494414293, + "loss": 2.4853, + "step": 11168 + }, + { + "epoch": 0.48, + "grad_norm": 0.478515625, + "learning_rate": 0.0001736779896582687, + "loss": 2.642, + "step": 11169 + }, + { + "epoch": 0.48, + "grad_norm": 0.431640625, + "learning_rate": 0.00017367345404087044, + "loss": 2.6182, + "step": 11170 + }, + { + "epoch": 0.48, + "grad_norm": 0.4765625, + "learning_rate": 0.00017366891809196858, + "loss": 2.56, + "step": 11171 + }, + { + "epoch": 0.48, + "grad_norm": 0.42578125, + "learning_rate": 0.0001736643818115835, + "loss": 2.5418, + "step": 11172 + }, + { + "epoch": 0.48, + "grad_norm": 0.4375, + "learning_rate": 0.00017365984519973564, + "loss": 2.6107, + "step": 11173 + }, + { + "epoch": 0.48, + "grad_norm": 0.478515625, + "learning_rate": 0.00017365530825644537, + "loss": 2.5897, + "step": 11174 + }, + { + "epoch": 0.48, + "grad_norm": 0.43359375, + "learning_rate": 0.00017365077098173312, + "loss": 2.3952, + "step": 11175 + }, + { + "epoch": 0.48, + "grad_norm": 0.4296875, + "learning_rate": 0.00017364623337561935, + "loss": 2.6108, + "step": 11176 + }, + { + "epoch": 0.48, + "grad_norm": 0.46875, + "learning_rate": 0.00017364169543812442, + "loss": 2.6869, + "step": 11177 + }, + { + "epoch": 0.48, + "grad_norm": 0.4765625, + "learning_rate": 0.00017363715716926882, + "loss": 2.6011, + "step": 11178 + }, + { + "epoch": 0.48, + "grad_norm": 0.45703125, + "learning_rate": 0.00017363261856907287, + "loss": 2.6932, + "step": 11179 + }, + { + "epoch": 0.48, + "grad_norm": 0.462890625, + "learning_rate": 0.00017362807963755707, + "loss": 2.6475, + "step": 11180 + }, + { + "epoch": 0.48, + "grad_norm": 0.466796875, + "learning_rate": 0.0001736235403747418, + "loss": 2.665, + "step": 11181 + }, + { + "epoch": 0.48, + "grad_norm": 0.44921875, + "learning_rate": 0.0001736190007806475, + "loss": 2.5934, + "step": 11182 + }, + { + "epoch": 0.48, + "grad_norm": 0.462890625, + "learning_rate": 0.00017361446085529462, + "loss": 2.6024, + "step": 11183 + }, + { + "epoch": 0.48, + "grad_norm": 0.50390625, + "learning_rate": 0.00017360992059870358, + "loss": 2.6, + "step": 11184 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.00017360538001089478, + "loss": 2.605, + "step": 11185 + }, + { + "epoch": 0.48, + "grad_norm": 0.478515625, + "learning_rate": 0.00017360083909188868, + "loss": 2.6003, + "step": 11186 + }, + { + "epoch": 0.48, + "grad_norm": 0.41015625, + "learning_rate": 0.0001735962978417057, + "loss": 2.5199, + "step": 11187 + }, + { + "epoch": 0.48, + "grad_norm": 0.4296875, + "learning_rate": 0.0001735917562603663, + "loss": 2.6196, + "step": 11188 + }, + { + "epoch": 0.48, + "grad_norm": 0.5, + "learning_rate": 0.00017358721434789084, + "loss": 2.5389, + "step": 11189 + }, + { + "epoch": 0.48, + "grad_norm": 0.5234375, + "learning_rate": 0.00017358267210429985, + "loss": 2.5457, + "step": 11190 + }, + { + "epoch": 0.48, + "grad_norm": 0.41796875, + "learning_rate": 0.00017357812952961373, + "loss": 2.7212, + "step": 11191 + }, + { + "epoch": 0.48, + "grad_norm": 0.427734375, + "learning_rate": 0.00017357358662385287, + "loss": 2.5686, + "step": 11192 + }, + { + "epoch": 0.48, + "grad_norm": 0.421875, + "learning_rate": 0.0001735690433870378, + "loss": 2.553, + "step": 11193 + }, + { + "epoch": 0.48, + "grad_norm": 0.458984375, + "learning_rate": 0.0001735644998191889, + "loss": 2.6158, + "step": 11194 + }, + { + "epoch": 0.48, + "grad_norm": 0.486328125, + "learning_rate": 0.00017355995592032664, + "loss": 2.6453, + "step": 11195 + }, + { + "epoch": 0.48, + "grad_norm": 0.470703125, + "learning_rate": 0.0001735554116904715, + "loss": 2.5929, + "step": 11196 + }, + { + "epoch": 0.48, + "grad_norm": 0.416015625, + "learning_rate": 0.00017355086712964382, + "loss": 2.7547, + "step": 11197 + }, + { + "epoch": 0.48, + "grad_norm": 0.455078125, + "learning_rate": 0.00017354632223786417, + "loss": 2.4959, + "step": 11198 + }, + { + "epoch": 0.48, + "grad_norm": 0.455078125, + "learning_rate": 0.00017354177701515294, + "loss": 2.5018, + "step": 11199 + }, + { + "epoch": 0.48, + "grad_norm": 0.45703125, + "learning_rate": 0.00017353723146153058, + "loss": 2.6762, + "step": 11200 + }, + { + "epoch": 0.48, + "grad_norm": 0.41015625, + "learning_rate": 0.00017353268557701755, + "loss": 2.4403, + "step": 11201 + }, + { + "epoch": 0.48, + "grad_norm": 0.451171875, + "learning_rate": 0.0001735281393616343, + "loss": 2.5292, + "step": 11202 + }, + { + "epoch": 0.48, + "grad_norm": 0.52734375, + "learning_rate": 0.0001735235928154013, + "loss": 2.6719, + "step": 11203 + }, + { + "epoch": 0.48, + "grad_norm": 0.478515625, + "learning_rate": 0.000173519045938339, + "loss": 2.6023, + "step": 11204 + }, + { + "epoch": 0.48, + "grad_norm": 0.419921875, + "learning_rate": 0.00017351449873046789, + "loss": 2.6031, + "step": 11205 + }, + { + "epoch": 0.48, + "grad_norm": 0.421875, + "learning_rate": 0.00017350995119180836, + "loss": 2.4984, + "step": 11206 + }, + { + "epoch": 0.48, + "grad_norm": 0.451171875, + "learning_rate": 0.00017350540332238093, + "loss": 2.6997, + "step": 11207 + }, + { + "epoch": 0.48, + "grad_norm": 0.451171875, + "learning_rate": 0.00017350085512220605, + "loss": 2.591, + "step": 11208 + }, + { + "epoch": 0.48, + "grad_norm": 0.400390625, + "learning_rate": 0.00017349630659130422, + "loss": 2.5742, + "step": 11209 + }, + { + "epoch": 0.48, + "grad_norm": 0.48046875, + "learning_rate": 0.0001734917577296958, + "loss": 2.5893, + "step": 11210 + }, + { + "epoch": 0.48, + "grad_norm": 0.47265625, + "learning_rate": 0.00017348720853740138, + "loss": 2.6029, + "step": 11211 + }, + { + "epoch": 0.48, + "grad_norm": 0.46875, + "learning_rate": 0.00017348265901444138, + "loss": 2.663, + "step": 11212 + }, + { + "epoch": 0.48, + "grad_norm": 0.419921875, + "learning_rate": 0.00017347810916083626, + "loss": 2.6081, + "step": 11213 + }, + { + "epoch": 0.48, + "grad_norm": 0.470703125, + "learning_rate": 0.00017347355897660648, + "loss": 2.6192, + "step": 11214 + }, + { + "epoch": 0.48, + "grad_norm": 0.474609375, + "learning_rate": 0.00017346900846177256, + "loss": 2.769, + "step": 11215 + }, + { + "epoch": 0.48, + "grad_norm": 0.4375, + "learning_rate": 0.00017346445761635493, + "loss": 2.5592, + "step": 11216 + }, + { + "epoch": 0.48, + "grad_norm": 0.435546875, + "learning_rate": 0.00017345990644037412, + "loss": 2.5519, + "step": 11217 + }, + { + "epoch": 0.48, + "grad_norm": 0.431640625, + "learning_rate": 0.00017345535493385055, + "loss": 2.6124, + "step": 11218 + }, + { + "epoch": 0.48, + "grad_norm": 0.46875, + "learning_rate": 0.00017345080309680473, + "loss": 2.8132, + "step": 11219 + }, + { + "epoch": 0.48, + "grad_norm": 0.466796875, + "learning_rate": 0.00017344625092925714, + "loss": 2.6931, + "step": 11220 + }, + { + "epoch": 0.48, + "grad_norm": 0.431640625, + "learning_rate": 0.00017344169843122828, + "loss": 2.5591, + "step": 11221 + }, + { + "epoch": 0.48, + "grad_norm": 0.474609375, + "learning_rate": 0.0001734371456027386, + "loss": 2.5319, + "step": 11222 + }, + { + "epoch": 0.48, + "grad_norm": 0.4609375, + "learning_rate": 0.00017343259244380864, + "loss": 2.7418, + "step": 11223 + }, + { + "epoch": 0.48, + "grad_norm": 0.5, + "learning_rate": 0.0001734280389544588, + "loss": 2.6955, + "step": 11224 + }, + { + "epoch": 0.48, + "grad_norm": 0.546875, + "learning_rate": 0.00017342348513470964, + "loss": 2.6087, + "step": 11225 + }, + { + "epoch": 0.48, + "grad_norm": 0.439453125, + "learning_rate": 0.00017341893098458163, + "loss": 2.6428, + "step": 11226 + }, + { + "epoch": 0.48, + "grad_norm": 0.439453125, + "learning_rate": 0.00017341437650409525, + "loss": 2.6247, + "step": 11227 + }, + { + "epoch": 0.48, + "grad_norm": 0.458984375, + "learning_rate": 0.00017340982169327103, + "loss": 2.5218, + "step": 11228 + }, + { + "epoch": 0.48, + "grad_norm": 0.51171875, + "learning_rate": 0.00017340526655212942, + "loss": 2.7098, + "step": 11229 + }, + { + "epoch": 0.48, + "grad_norm": 0.462890625, + "learning_rate": 0.00017340071108069095, + "loss": 2.59, + "step": 11230 + }, + { + "epoch": 0.48, + "grad_norm": 0.43359375, + "learning_rate": 0.0001733961552789761, + "loss": 2.4948, + "step": 11231 + }, + { + "epoch": 0.48, + "grad_norm": 0.412109375, + "learning_rate": 0.00017339159914700537, + "loss": 2.6248, + "step": 11232 + }, + { + "epoch": 0.48, + "grad_norm": 0.53125, + "learning_rate": 0.00017338704268479927, + "loss": 2.7378, + "step": 11233 + }, + { + "epoch": 0.48, + "grad_norm": 0.47265625, + "learning_rate": 0.0001733824858923783, + "loss": 2.6319, + "step": 11234 + }, + { + "epoch": 0.48, + "grad_norm": 0.42578125, + "learning_rate": 0.00017337792876976295, + "loss": 2.467, + "step": 11235 + }, + { + "epoch": 0.48, + "grad_norm": 0.486328125, + "learning_rate": 0.00017337337131697375, + "loss": 2.6371, + "step": 11236 + }, + { + "epoch": 0.48, + "grad_norm": 0.419921875, + "learning_rate": 0.00017336881353403118, + "loss": 2.6675, + "step": 11237 + }, + { + "epoch": 0.48, + "grad_norm": 0.44140625, + "learning_rate": 0.00017336425542095578, + "loss": 2.4218, + "step": 11238 + }, + { + "epoch": 0.48, + "grad_norm": 0.431640625, + "learning_rate": 0.00017335969697776803, + "loss": 2.6016, + "step": 11239 + }, + { + "epoch": 0.48, + "grad_norm": 0.4609375, + "learning_rate": 0.00017335513820448844, + "loss": 2.5222, + "step": 11240 + }, + { + "epoch": 0.48, + "grad_norm": 0.4296875, + "learning_rate": 0.00017335057910113758, + "loss": 2.6307, + "step": 11241 + }, + { + "epoch": 0.48, + "grad_norm": 0.44140625, + "learning_rate": 0.0001733460196677359, + "loss": 2.6977, + "step": 11242 + }, + { + "epoch": 0.48, + "grad_norm": 0.447265625, + "learning_rate": 0.00017334145990430394, + "loss": 2.6121, + "step": 11243 + }, + { + "epoch": 0.48, + "grad_norm": 0.45703125, + "learning_rate": 0.0001733368998108622, + "loss": 2.6487, + "step": 11244 + }, + { + "epoch": 0.48, + "grad_norm": 0.458984375, + "learning_rate": 0.00017333233938743122, + "loss": 2.4706, + "step": 11245 + }, + { + "epoch": 0.48, + "grad_norm": 0.46875, + "learning_rate": 0.0001733277786340315, + "loss": 2.5905, + "step": 11246 + }, + { + "epoch": 0.48, + "grad_norm": 0.462890625, + "learning_rate": 0.0001733232175506836, + "loss": 2.5615, + "step": 11247 + }, + { + "epoch": 0.48, + "grad_norm": 0.490234375, + "learning_rate": 0.00017331865613740797, + "loss": 2.5934, + "step": 11248 + }, + { + "epoch": 0.48, + "grad_norm": 0.4296875, + "learning_rate": 0.00017331409439422523, + "loss": 2.5862, + "step": 11249 + }, + { + "epoch": 0.48, + "grad_norm": 0.4453125, + "learning_rate": 0.00017330953232115584, + "loss": 2.5845, + "step": 11250 + }, + { + "epoch": 0.48, + "grad_norm": 0.4765625, + "learning_rate": 0.00017330496991822036, + "loss": 2.6769, + "step": 11251 + }, + { + "epoch": 0.48, + "grad_norm": 0.462890625, + "learning_rate": 0.0001733004071854393, + "loss": 2.4725, + "step": 11252 + }, + { + "epoch": 0.48, + "grad_norm": 0.4765625, + "learning_rate": 0.00017329584412283317, + "loss": 2.6763, + "step": 11253 + }, + { + "epoch": 0.48, + "grad_norm": 0.431640625, + "learning_rate": 0.00017329128073042256, + "loss": 2.5183, + "step": 11254 + }, + { + "epoch": 0.48, + "grad_norm": 0.400390625, + "learning_rate": 0.00017328671700822797, + "loss": 2.5417, + "step": 11255 + }, + { + "epoch": 0.48, + "grad_norm": 0.48046875, + "learning_rate": 0.0001732821529562699, + "loss": 2.5509, + "step": 11256 + }, + { + "epoch": 0.48, + "grad_norm": 0.37890625, + "learning_rate": 0.00017327758857456894, + "loss": 2.5242, + "step": 11257 + }, + { + "epoch": 0.48, + "grad_norm": 0.53125, + "learning_rate": 0.00017327302386314564, + "loss": 2.6636, + "step": 11258 + }, + { + "epoch": 0.48, + "grad_norm": 0.470703125, + "learning_rate": 0.0001732684588220205, + "loss": 2.6382, + "step": 11259 + }, + { + "epoch": 0.48, + "grad_norm": 0.400390625, + "learning_rate": 0.00017326389345121402, + "loss": 2.4944, + "step": 11260 + }, + { + "epoch": 0.48, + "grad_norm": 0.54296875, + "learning_rate": 0.00017325932775074686, + "loss": 2.6198, + "step": 11261 + }, + { + "epoch": 0.48, + "grad_norm": 0.40234375, + "learning_rate": 0.0001732547617206395, + "loss": 2.5649, + "step": 11262 + }, + { + "epoch": 0.48, + "grad_norm": 0.451171875, + "learning_rate": 0.00017325019536091246, + "loss": 2.6515, + "step": 11263 + }, + { + "epoch": 0.48, + "grad_norm": 0.48828125, + "learning_rate": 0.0001732456286715863, + "loss": 2.6292, + "step": 11264 + }, + { + "epoch": 0.48, + "grad_norm": 0.451171875, + "learning_rate": 0.00017324106165268158, + "loss": 2.6477, + "step": 11265 + }, + { + "epoch": 0.48, + "grad_norm": 0.419921875, + "learning_rate": 0.00017323649430421887, + "loss": 2.5184, + "step": 11266 + }, + { + "epoch": 0.48, + "grad_norm": 0.5390625, + "learning_rate": 0.0001732319266262187, + "loss": 2.7713, + "step": 11267 + }, + { + "epoch": 0.48, + "grad_norm": 0.443359375, + "learning_rate": 0.0001732273586187016, + "loss": 2.4516, + "step": 11268 + }, + { + "epoch": 0.48, + "grad_norm": 0.427734375, + "learning_rate": 0.00017322279028168816, + "loss": 2.6363, + "step": 11269 + }, + { + "epoch": 0.48, + "grad_norm": 0.458984375, + "learning_rate": 0.00017321822161519892, + "loss": 2.5089, + "step": 11270 + }, + { + "epoch": 0.48, + "grad_norm": 0.404296875, + "learning_rate": 0.00017321365261925445, + "loss": 2.622, + "step": 11271 + }, + { + "epoch": 0.48, + "grad_norm": 0.376953125, + "learning_rate": 0.00017320908329387528, + "loss": 2.605, + "step": 11272 + }, + { + "epoch": 0.48, + "grad_norm": 0.47265625, + "learning_rate": 0.00017320451363908202, + "loss": 2.5932, + "step": 11273 + }, + { + "epoch": 0.48, + "grad_norm": 0.44921875, + "learning_rate": 0.00017319994365489518, + "loss": 2.717, + "step": 11274 + }, + { + "epoch": 0.48, + "grad_norm": 0.466796875, + "learning_rate": 0.00017319537334133534, + "loss": 2.6169, + "step": 11275 + }, + { + "epoch": 0.48, + "grad_norm": 0.466796875, + "learning_rate": 0.0001731908026984231, + "loss": 2.5348, + "step": 11276 + }, + { + "epoch": 0.48, + "grad_norm": 0.451171875, + "learning_rate": 0.000173186231726179, + "loss": 2.6074, + "step": 11277 + }, + { + "epoch": 0.48, + "grad_norm": 0.400390625, + "learning_rate": 0.00017318166042462355, + "loss": 2.4947, + "step": 11278 + }, + { + "epoch": 0.48, + "grad_norm": 0.49609375, + "learning_rate": 0.0001731770887937774, + "loss": 2.7093, + "step": 11279 + }, + { + "epoch": 0.48, + "grad_norm": 0.4140625, + "learning_rate": 0.0001731725168336611, + "loss": 2.5837, + "step": 11280 + }, + { + "epoch": 0.48, + "grad_norm": 0.478515625, + "learning_rate": 0.0001731679445442952, + "loss": 2.7216, + "step": 11281 + }, + { + "epoch": 0.48, + "grad_norm": 0.384765625, + "learning_rate": 0.00017316337192570033, + "loss": 2.5501, + "step": 11282 + }, + { + "epoch": 0.48, + "grad_norm": 0.50390625, + "learning_rate": 0.000173158798977897, + "loss": 2.613, + "step": 11283 + }, + { + "epoch": 0.48, + "grad_norm": 0.41796875, + "learning_rate": 0.00017315422570090582, + "loss": 2.5798, + "step": 11284 + }, + { + "epoch": 0.48, + "grad_norm": 0.46484375, + "learning_rate": 0.0001731496520947473, + "loss": 2.6462, + "step": 11285 + }, + { + "epoch": 0.48, + "grad_norm": 0.455078125, + "learning_rate": 0.00017314507815944217, + "loss": 2.6091, + "step": 11286 + }, + { + "epoch": 0.48, + "grad_norm": 0.466796875, + "learning_rate": 0.00017314050389501085, + "loss": 2.6803, + "step": 11287 + }, + { + "epoch": 0.48, + "grad_norm": 0.466796875, + "learning_rate": 0.00017313592930147403, + "loss": 2.5541, + "step": 11288 + }, + { + "epoch": 0.48, + "grad_norm": 0.392578125, + "learning_rate": 0.00017313135437885225, + "loss": 2.5812, + "step": 11289 + }, + { + "epoch": 0.48, + "grad_norm": 0.4609375, + "learning_rate": 0.0001731267791271661, + "loss": 2.6157, + "step": 11290 + }, + { + "epoch": 0.48, + "grad_norm": 0.50390625, + "learning_rate": 0.00017312220354643616, + "loss": 2.6567, + "step": 11291 + }, + { + "epoch": 0.48, + "grad_norm": 0.439453125, + "learning_rate": 0.00017311762763668305, + "loss": 2.7543, + "step": 11292 + }, + { + "epoch": 0.48, + "grad_norm": 0.4375, + "learning_rate": 0.0001731130513979273, + "loss": 2.4532, + "step": 11293 + }, + { + "epoch": 0.48, + "grad_norm": 0.47265625, + "learning_rate": 0.00017310847483018957, + "loss": 2.6232, + "step": 11294 + }, + { + "epoch": 0.48, + "grad_norm": 0.4609375, + "learning_rate": 0.0001731038979334904, + "loss": 2.6445, + "step": 11295 + }, + { + "epoch": 0.48, + "grad_norm": 0.423828125, + "learning_rate": 0.0001730993207078504, + "loss": 2.6576, + "step": 11296 + }, + { + "epoch": 0.48, + "grad_norm": 0.46484375, + "learning_rate": 0.0001730947431532902, + "loss": 2.664, + "step": 11297 + }, + { + "epoch": 0.48, + "grad_norm": 0.47265625, + "learning_rate": 0.00017309016526983038, + "loss": 2.6635, + "step": 11298 + }, + { + "epoch": 0.48, + "grad_norm": 0.46484375, + "learning_rate": 0.0001730855870574915, + "loss": 2.751, + "step": 11299 + }, + { + "epoch": 0.48, + "grad_norm": 0.45703125, + "learning_rate": 0.00017308100851629417, + "loss": 2.6686, + "step": 11300 + }, + { + "epoch": 0.48, + "grad_norm": 0.5, + "learning_rate": 0.00017307642964625903, + "loss": 2.603, + "step": 11301 + }, + { + "epoch": 0.48, + "grad_norm": 0.435546875, + "learning_rate": 0.00017307185044740665, + "loss": 2.5452, + "step": 11302 + }, + { + "epoch": 0.48, + "grad_norm": 0.494140625, + "learning_rate": 0.00017306727091975769, + "loss": 2.784, + "step": 11303 + }, + { + "epoch": 0.48, + "grad_norm": 0.4609375, + "learning_rate": 0.00017306269106333265, + "loss": 2.6236, + "step": 11304 + }, + { + "epoch": 0.48, + "grad_norm": 0.4296875, + "learning_rate": 0.00017305811087815225, + "loss": 2.527, + "step": 11305 + }, + { + "epoch": 0.48, + "grad_norm": 0.48828125, + "learning_rate": 0.00017305353036423706, + "loss": 2.6806, + "step": 11306 + }, + { + "epoch": 0.48, + "grad_norm": 0.48046875, + "learning_rate": 0.00017304894952160766, + "loss": 2.5, + "step": 11307 + }, + { + "epoch": 0.48, + "grad_norm": 0.458984375, + "learning_rate": 0.00017304436835028465, + "loss": 2.7488, + "step": 11308 + }, + { + "epoch": 0.48, + "grad_norm": 0.455078125, + "learning_rate": 0.00017303978685028875, + "loss": 2.7088, + "step": 11309 + }, + { + "epoch": 0.48, + "grad_norm": 0.546875, + "learning_rate": 0.00017303520502164044, + "loss": 2.7139, + "step": 11310 + }, + { + "epoch": 0.48, + "grad_norm": 0.4921875, + "learning_rate": 0.00017303062286436042, + "loss": 2.4739, + "step": 11311 + }, + { + "epoch": 0.48, + "grad_norm": 0.494140625, + "learning_rate": 0.00017302604037846927, + "loss": 2.5855, + "step": 11312 + }, + { + "epoch": 0.48, + "grad_norm": 0.54296875, + "learning_rate": 0.00017302145756398766, + "loss": 2.6621, + "step": 11313 + }, + { + "epoch": 0.48, + "grad_norm": 0.427734375, + "learning_rate": 0.00017301687442093617, + "loss": 2.6183, + "step": 11314 + }, + { + "epoch": 0.48, + "grad_norm": 0.5, + "learning_rate": 0.00017301229094933543, + "loss": 2.4273, + "step": 11315 + }, + { + "epoch": 0.48, + "grad_norm": 0.40625, + "learning_rate": 0.00017300770714920603, + "loss": 2.4934, + "step": 11316 + }, + { + "epoch": 0.48, + "grad_norm": 0.51953125, + "learning_rate": 0.00017300312302056864, + "loss": 2.7343, + "step": 11317 + }, + { + "epoch": 0.48, + "grad_norm": 0.46875, + "learning_rate": 0.00017299853856344392, + "loss": 2.6034, + "step": 11318 + }, + { + "epoch": 0.48, + "grad_norm": 0.44921875, + "learning_rate": 0.0001729939537778524, + "loss": 2.5463, + "step": 11319 + }, + { + "epoch": 0.48, + "grad_norm": 0.47265625, + "learning_rate": 0.0001729893686638148, + "loss": 2.5852, + "step": 11320 + }, + { + "epoch": 0.48, + "grad_norm": 0.4609375, + "learning_rate": 0.00017298478322135173, + "loss": 2.6353, + "step": 11321 + }, + { + "epoch": 0.48, + "grad_norm": 0.484375, + "learning_rate": 0.00017298019745048374, + "loss": 2.6982, + "step": 11322 + }, + { + "epoch": 0.48, + "grad_norm": 0.478515625, + "learning_rate": 0.0001729756113512316, + "loss": 2.5651, + "step": 11323 + }, + { + "epoch": 0.48, + "grad_norm": 0.578125, + "learning_rate": 0.00017297102492361586, + "loss": 2.6601, + "step": 11324 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.00017296643816765717, + "loss": 2.4849, + "step": 11325 + }, + { + "epoch": 0.48, + "grad_norm": 0.470703125, + "learning_rate": 0.00017296185108337618, + "loss": 2.6064, + "step": 11326 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.0001729572636707935, + "loss": 2.6267, + "step": 11327 + }, + { + "epoch": 0.48, + "grad_norm": 0.48828125, + "learning_rate": 0.00017295267592992982, + "loss": 2.6011, + "step": 11328 + }, + { + "epoch": 0.48, + "grad_norm": 0.443359375, + "learning_rate": 0.00017294808786080576, + "loss": 2.4468, + "step": 11329 + }, + { + "epoch": 0.48, + "grad_norm": 0.43359375, + "learning_rate": 0.00017294349946344196, + "loss": 2.7871, + "step": 11330 + }, + { + "epoch": 0.48, + "grad_norm": 0.486328125, + "learning_rate": 0.00017293891073785905, + "loss": 2.6906, + "step": 11331 + }, + { + "epoch": 0.48, + "grad_norm": 0.43359375, + "learning_rate": 0.00017293432168407773, + "loss": 2.5679, + "step": 11332 + }, + { + "epoch": 0.48, + "grad_norm": 0.45703125, + "learning_rate": 0.0001729297323021186, + "loss": 2.6864, + "step": 11333 + }, + { + "epoch": 0.48, + "grad_norm": 0.5, + "learning_rate": 0.00017292514259200235, + "loss": 2.6417, + "step": 11334 + }, + { + "epoch": 0.48, + "grad_norm": 0.40625, + "learning_rate": 0.00017292055255374957, + "loss": 2.5823, + "step": 11335 + }, + { + "epoch": 0.48, + "grad_norm": 0.4375, + "learning_rate": 0.00017291596218738095, + "loss": 2.6073, + "step": 11336 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.0001729113714929172, + "loss": 2.7131, + "step": 11337 + }, + { + "epoch": 0.48, + "grad_norm": 0.431640625, + "learning_rate": 0.00017290678047037888, + "loss": 2.6915, + "step": 11338 + }, + { + "epoch": 0.48, + "grad_norm": 0.46875, + "learning_rate": 0.0001729021891197867, + "loss": 2.4937, + "step": 11339 + }, + { + "epoch": 0.48, + "grad_norm": 0.53515625, + "learning_rate": 0.00017289759744116133, + "loss": 2.6015, + "step": 11340 + }, + { + "epoch": 0.48, + "grad_norm": 0.45703125, + "learning_rate": 0.0001728930054345234, + "loss": 2.5468, + "step": 11341 + }, + { + "epoch": 0.48, + "grad_norm": 0.4609375, + "learning_rate": 0.0001728884130998936, + "loss": 2.6436, + "step": 11342 + }, + { + "epoch": 0.48, + "grad_norm": 0.4453125, + "learning_rate": 0.00017288382043729254, + "loss": 2.6046, + "step": 11343 + }, + { + "epoch": 0.48, + "grad_norm": 0.470703125, + "learning_rate": 0.00017287922744674096, + "loss": 2.6896, + "step": 11344 + }, + { + "epoch": 0.48, + "grad_norm": 0.44921875, + "learning_rate": 0.00017287463412825948, + "loss": 2.6148, + "step": 11345 + }, + { + "epoch": 0.48, + "grad_norm": 0.458984375, + "learning_rate": 0.00017287004048186877, + "loss": 2.5012, + "step": 11346 + }, + { + "epoch": 0.48, + "grad_norm": 0.4453125, + "learning_rate": 0.00017286544650758955, + "loss": 2.5216, + "step": 11347 + }, + { + "epoch": 0.48, + "grad_norm": 0.439453125, + "learning_rate": 0.00017286085220544239, + "loss": 2.6208, + "step": 11348 + }, + { + "epoch": 0.48, + "grad_norm": 0.43359375, + "learning_rate": 0.00017285625757544804, + "loss": 2.5417, + "step": 11349 + }, + { + "epoch": 0.48, + "grad_norm": 0.46484375, + "learning_rate": 0.00017285166261762718, + "loss": 2.5415, + "step": 11350 + }, + { + "epoch": 0.48, + "grad_norm": 0.45703125, + "learning_rate": 0.00017284706733200044, + "loss": 2.5592, + "step": 11351 + }, + { + "epoch": 0.48, + "grad_norm": 0.4921875, + "learning_rate": 0.00017284247171858854, + "loss": 2.6053, + "step": 11352 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.00017283787577741214, + "loss": 2.6222, + "step": 11353 + }, + { + "epoch": 0.48, + "grad_norm": 0.478515625, + "learning_rate": 0.00017283327950849188, + "loss": 2.5892, + "step": 11354 + }, + { + "epoch": 0.48, + "grad_norm": 0.5, + "learning_rate": 0.00017282868291184852, + "loss": 2.549, + "step": 11355 + }, + { + "epoch": 0.48, + "grad_norm": 0.44921875, + "learning_rate": 0.00017282408598750268, + "loss": 2.5137, + "step": 11356 + }, + { + "epoch": 0.48, + "grad_norm": 0.380859375, + "learning_rate": 0.00017281948873547508, + "loss": 2.6895, + "step": 11357 + }, + { + "epoch": 0.48, + "grad_norm": 0.4921875, + "learning_rate": 0.0001728148911557864, + "loss": 2.5657, + "step": 11358 + }, + { + "epoch": 0.48, + "grad_norm": 0.421875, + "learning_rate": 0.0001728102932484573, + "loss": 2.5344, + "step": 11359 + }, + { + "epoch": 0.48, + "grad_norm": 0.4609375, + "learning_rate": 0.0001728056950135085, + "loss": 2.6503, + "step": 11360 + }, + { + "epoch": 0.48, + "grad_norm": 0.4921875, + "learning_rate": 0.00017280109645096068, + "loss": 2.5652, + "step": 11361 + }, + { + "epoch": 0.48, + "grad_norm": 0.388671875, + "learning_rate": 0.0001727964975608345, + "loss": 2.6098, + "step": 11362 + }, + { + "epoch": 0.48, + "grad_norm": 0.50390625, + "learning_rate": 0.00017279189834315073, + "loss": 2.5295, + "step": 11363 + }, + { + "epoch": 0.48, + "grad_norm": 0.4921875, + "learning_rate": 0.00017278729879793, + "loss": 2.6361, + "step": 11364 + }, + { + "epoch": 0.48, + "grad_norm": 0.458984375, + "learning_rate": 0.00017278269892519304, + "loss": 2.7065, + "step": 11365 + }, + { + "epoch": 0.48, + "grad_norm": 0.482421875, + "learning_rate": 0.00017277809872496053, + "loss": 2.642, + "step": 11366 + }, + { + "epoch": 0.48, + "grad_norm": 0.4921875, + "learning_rate": 0.00017277349819725318, + "loss": 2.5839, + "step": 11367 + }, + { + "epoch": 0.48, + "grad_norm": 0.4375, + "learning_rate": 0.00017276889734209165, + "loss": 2.664, + "step": 11368 + }, + { + "epoch": 0.48, + "grad_norm": 0.4921875, + "learning_rate": 0.0001727642961594967, + "loss": 2.5053, + "step": 11369 + }, + { + "epoch": 0.48, + "grad_norm": 0.451171875, + "learning_rate": 0.000172759694649489, + "loss": 2.5543, + "step": 11370 + }, + { + "epoch": 0.48, + "grad_norm": 0.4453125, + "learning_rate": 0.00017275509281208928, + "loss": 2.5569, + "step": 11371 + }, + { + "epoch": 0.48, + "grad_norm": 0.5234375, + "learning_rate": 0.0001727504906473182, + "loss": 2.6825, + "step": 11372 + }, + { + "epoch": 0.48, + "grad_norm": 0.5, + "learning_rate": 0.00017274588815519655, + "loss": 2.6841, + "step": 11373 + }, + { + "epoch": 0.48, + "grad_norm": 0.51171875, + "learning_rate": 0.00017274128533574493, + "loss": 2.5117, + "step": 11374 + }, + { + "epoch": 0.48, + "grad_norm": 0.494140625, + "learning_rate": 0.00017273668218898414, + "loss": 2.5703, + "step": 11375 + }, + { + "epoch": 0.48, + "grad_norm": 0.486328125, + "learning_rate": 0.0001727320787149349, + "loss": 2.7217, + "step": 11376 + }, + { + "epoch": 0.48, + "grad_norm": 0.453125, + "learning_rate": 0.00017272747491361786, + "loss": 2.6994, + "step": 11377 + }, + { + "epoch": 0.48, + "grad_norm": 0.4140625, + "learning_rate": 0.00017272287078505373, + "loss": 2.6179, + "step": 11378 + }, + { + "epoch": 0.48, + "grad_norm": 0.47265625, + "learning_rate": 0.00017271826632926329, + "loss": 2.665, + "step": 11379 + }, + { + "epoch": 0.48, + "grad_norm": 0.48828125, + "learning_rate": 0.00017271366154626721, + "loss": 2.5511, + "step": 11380 + }, + { + "epoch": 0.48, + "grad_norm": 0.443359375, + "learning_rate": 0.00017270905643608625, + "loss": 2.6144, + "step": 11381 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.0001727044509987411, + "loss": 2.7241, + "step": 11382 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.00017269984523425244, + "loss": 2.4577, + "step": 11383 + }, + { + "epoch": 0.49, + "grad_norm": 0.41796875, + "learning_rate": 0.00017269523914264113, + "loss": 2.5866, + "step": 11384 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.00017269063272392775, + "loss": 2.6651, + "step": 11385 + }, + { + "epoch": 0.49, + "grad_norm": 0.5, + "learning_rate": 0.0001726860259781331, + "loss": 2.6179, + "step": 11386 + }, + { + "epoch": 0.49, + "grad_norm": 0.462890625, + "learning_rate": 0.0001726814189052779, + "loss": 2.6034, + "step": 11387 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.00017267681150538288, + "loss": 2.4891, + "step": 11388 + }, + { + "epoch": 0.49, + "grad_norm": 0.435546875, + "learning_rate": 0.00017267220377846878, + "loss": 2.4817, + "step": 11389 + }, + { + "epoch": 0.49, + "grad_norm": 0.43359375, + "learning_rate": 0.00017266759572455632, + "loss": 2.4828, + "step": 11390 + }, + { + "epoch": 0.49, + "grad_norm": 0.5078125, + "learning_rate": 0.0001726629873436662, + "loss": 2.7067, + "step": 11391 + }, + { + "epoch": 0.49, + "grad_norm": 0.458984375, + "learning_rate": 0.00017265837863581924, + "loss": 2.5271, + "step": 11392 + }, + { + "epoch": 0.49, + "grad_norm": 0.4609375, + "learning_rate": 0.0001726537696010361, + "loss": 2.624, + "step": 11393 + }, + { + "epoch": 0.49, + "grad_norm": 0.39453125, + "learning_rate": 0.00017264916023933754, + "loss": 2.5999, + "step": 11394 + }, + { + "epoch": 0.49, + "grad_norm": 0.486328125, + "learning_rate": 0.00017264455055074435, + "loss": 2.6792, + "step": 11395 + }, + { + "epoch": 0.49, + "grad_norm": 0.5, + "learning_rate": 0.00017263994053527722, + "loss": 2.7531, + "step": 11396 + }, + { + "epoch": 0.49, + "grad_norm": 0.453125, + "learning_rate": 0.0001726353301929569, + "loss": 2.5689, + "step": 11397 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.00017263071952380412, + "loss": 2.6215, + "step": 11398 + }, + { + "epoch": 0.49, + "grad_norm": 0.419921875, + "learning_rate": 0.00017262610852783965, + "loss": 2.5906, + "step": 11399 + }, + { + "epoch": 0.49, + "grad_norm": 0.42578125, + "learning_rate": 0.00017262149720508426, + "loss": 2.574, + "step": 11400 + }, + { + "epoch": 0.49, + "grad_norm": 0.490234375, + "learning_rate": 0.00017261688555555863, + "loss": 2.7139, + "step": 11401 + }, + { + "epoch": 0.49, + "grad_norm": 0.5078125, + "learning_rate": 0.00017261227357928357, + "loss": 2.6519, + "step": 11402 + }, + { + "epoch": 0.49, + "grad_norm": 0.482421875, + "learning_rate": 0.00017260766127627982, + "loss": 2.6253, + "step": 11403 + }, + { + "epoch": 0.49, + "grad_norm": 0.5234375, + "learning_rate": 0.00017260304864656815, + "loss": 2.5685, + "step": 11404 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.00017259843569016926, + "loss": 2.6245, + "step": 11405 + }, + { + "epoch": 0.49, + "grad_norm": 0.5078125, + "learning_rate": 0.00017259382240710397, + "loss": 2.7998, + "step": 11406 + }, + { + "epoch": 0.49, + "grad_norm": 0.451171875, + "learning_rate": 0.000172589208797393, + "loss": 2.7123, + "step": 11407 + }, + { + "epoch": 0.49, + "grad_norm": 0.474609375, + "learning_rate": 0.00017258459486105713, + "loss": 2.6581, + "step": 11408 + }, + { + "epoch": 0.49, + "grad_norm": 0.455078125, + "learning_rate": 0.0001725799805981171, + "loss": 2.664, + "step": 11409 + }, + { + "epoch": 0.49, + "grad_norm": 0.52734375, + "learning_rate": 0.00017257536600859367, + "loss": 2.6349, + "step": 11410 + }, + { + "epoch": 0.49, + "grad_norm": 0.4453125, + "learning_rate": 0.00017257075109250762, + "loss": 2.6916, + "step": 11411 + }, + { + "epoch": 0.49, + "grad_norm": 0.3984375, + "learning_rate": 0.0001725661358498797, + "loss": 2.6001, + "step": 11412 + }, + { + "epoch": 0.49, + "grad_norm": 0.4453125, + "learning_rate": 0.00017256152028073075, + "loss": 2.5352, + "step": 11413 + }, + { + "epoch": 0.49, + "grad_norm": 0.4921875, + "learning_rate": 0.0001725569043850814, + "loss": 2.5721, + "step": 11414 + }, + { + "epoch": 0.49, + "grad_norm": 0.466796875, + "learning_rate": 0.00017255228816295254, + "loss": 2.6426, + "step": 11415 + }, + { + "epoch": 0.49, + "grad_norm": 0.390625, + "learning_rate": 0.0001725476716143649, + "loss": 2.6344, + "step": 11416 + }, + { + "epoch": 0.49, + "grad_norm": 0.54296875, + "learning_rate": 0.00017254305473933922, + "loss": 2.7014, + "step": 11417 + }, + { + "epoch": 0.49, + "grad_norm": 0.453125, + "learning_rate": 0.00017253843753789632, + "loss": 2.6013, + "step": 11418 + }, + { + "epoch": 0.49, + "grad_norm": 0.44140625, + "learning_rate": 0.00017253382001005695, + "loss": 2.5249, + "step": 11419 + }, + { + "epoch": 0.49, + "grad_norm": 0.421875, + "learning_rate": 0.0001725292021558419, + "loss": 2.5591, + "step": 11420 + }, + { + "epoch": 0.49, + "grad_norm": 0.478515625, + "learning_rate": 0.000172524583975272, + "loss": 2.5955, + "step": 11421 + }, + { + "epoch": 0.49, + "grad_norm": 0.42578125, + "learning_rate": 0.0001725199654683679, + "loss": 2.583, + "step": 11422 + }, + { + "epoch": 0.49, + "grad_norm": 0.53515625, + "learning_rate": 0.0001725153466351505, + "loss": 2.7631, + "step": 11423 + }, + { + "epoch": 0.49, + "grad_norm": 0.408203125, + "learning_rate": 0.00017251072747564053, + "loss": 2.5251, + "step": 11424 + }, + { + "epoch": 0.49, + "grad_norm": 0.484375, + "learning_rate": 0.0001725061079898588, + "loss": 2.7072, + "step": 11425 + }, + { + "epoch": 0.49, + "grad_norm": 0.482421875, + "learning_rate": 0.00017250148817782605, + "loss": 2.5408, + "step": 11426 + }, + { + "epoch": 0.49, + "grad_norm": 0.416015625, + "learning_rate": 0.00017249686803956314, + "loss": 2.5743, + "step": 11427 + }, + { + "epoch": 0.49, + "grad_norm": 0.470703125, + "learning_rate": 0.0001724922475750908, + "loss": 2.4998, + "step": 11428 + }, + { + "epoch": 0.49, + "grad_norm": 0.46484375, + "learning_rate": 0.00017248762678442984, + "loss": 2.7044, + "step": 11429 + }, + { + "epoch": 0.49, + "grad_norm": 0.462890625, + "learning_rate": 0.00017248300566760104, + "loss": 2.5378, + "step": 11430 + }, + { + "epoch": 0.49, + "grad_norm": 0.435546875, + "learning_rate": 0.00017247838422462523, + "loss": 2.5577, + "step": 11431 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.00017247376245552317, + "loss": 2.5262, + "step": 11432 + }, + { + "epoch": 0.49, + "grad_norm": 0.49609375, + "learning_rate": 0.00017246914036031565, + "loss": 2.665, + "step": 11433 + }, + { + "epoch": 0.49, + "grad_norm": 0.462890625, + "learning_rate": 0.0001724645179390235, + "loss": 2.6557, + "step": 11434 + }, + { + "epoch": 0.49, + "grad_norm": 0.484375, + "learning_rate": 0.00017245989519166748, + "loss": 2.6713, + "step": 11435 + }, + { + "epoch": 0.49, + "grad_norm": 0.443359375, + "learning_rate": 0.00017245527211826844, + "loss": 2.6504, + "step": 11436 + }, + { + "epoch": 0.49, + "grad_norm": 0.478515625, + "learning_rate": 0.00017245064871884713, + "loss": 2.6567, + "step": 11437 + }, + { + "epoch": 0.49, + "grad_norm": 0.421875, + "learning_rate": 0.00017244602499342436, + "loss": 2.6574, + "step": 11438 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.000172441400942021, + "loss": 2.7026, + "step": 11439 + }, + { + "epoch": 0.49, + "grad_norm": 0.458984375, + "learning_rate": 0.0001724367765646578, + "loss": 2.6424, + "step": 11440 + }, + { + "epoch": 0.49, + "grad_norm": 0.44921875, + "learning_rate": 0.00017243215186135558, + "loss": 2.6577, + "step": 11441 + }, + { + "epoch": 0.49, + "grad_norm": 0.4375, + "learning_rate": 0.00017242752683213514, + "loss": 2.5382, + "step": 11442 + }, + { + "epoch": 0.49, + "grad_norm": 0.48046875, + "learning_rate": 0.00017242290147701729, + "loss": 2.5912, + "step": 11443 + }, + { + "epoch": 0.49, + "grad_norm": 0.49609375, + "learning_rate": 0.00017241827579602285, + "loss": 2.6355, + "step": 11444 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.00017241364978917264, + "loss": 2.5992, + "step": 11445 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.00017240902345648748, + "loss": 2.5791, + "step": 11446 + }, + { + "epoch": 0.49, + "grad_norm": 0.451171875, + "learning_rate": 0.00017240439679798816, + "loss": 2.5227, + "step": 11447 + }, + { + "epoch": 0.49, + "grad_norm": 0.46484375, + "learning_rate": 0.0001723997698136955, + "loss": 2.5928, + "step": 11448 + }, + { + "epoch": 0.49, + "grad_norm": 0.427734375, + "learning_rate": 0.00017239514250363042, + "loss": 2.7425, + "step": 11449 + }, + { + "epoch": 0.49, + "grad_norm": 0.484375, + "learning_rate": 0.00017239051486781358, + "loss": 2.592, + "step": 11450 + }, + { + "epoch": 0.49, + "grad_norm": 0.435546875, + "learning_rate": 0.00017238588690626587, + "loss": 2.6613, + "step": 11451 + }, + { + "epoch": 0.49, + "grad_norm": 0.4296875, + "learning_rate": 0.00017238125861900817, + "loss": 2.5605, + "step": 11452 + }, + { + "epoch": 0.49, + "grad_norm": 0.408203125, + "learning_rate": 0.00017237663000606123, + "loss": 2.6229, + "step": 11453 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.0001723720010674459, + "loss": 2.6272, + "step": 11454 + }, + { + "epoch": 0.49, + "grad_norm": 0.490234375, + "learning_rate": 0.00017236737180318303, + "loss": 2.672, + "step": 11455 + }, + { + "epoch": 0.49, + "grad_norm": 0.53125, + "learning_rate": 0.00017236274221329342, + "loss": 2.7442, + "step": 11456 + }, + { + "epoch": 0.49, + "grad_norm": 0.486328125, + "learning_rate": 0.00017235811229779794, + "loss": 2.7764, + "step": 11457 + }, + { + "epoch": 0.49, + "grad_norm": 0.474609375, + "learning_rate": 0.0001723534820567174, + "loss": 2.7127, + "step": 11458 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.00017234885149007258, + "loss": 2.5476, + "step": 11459 + }, + { + "epoch": 0.49, + "grad_norm": 0.408203125, + "learning_rate": 0.00017234422059788442, + "loss": 2.517, + "step": 11460 + }, + { + "epoch": 0.49, + "grad_norm": 0.5234375, + "learning_rate": 0.00017233958938017368, + "loss": 2.6887, + "step": 11461 + }, + { + "epoch": 0.49, + "grad_norm": 0.455078125, + "learning_rate": 0.00017233495783696126, + "loss": 2.6522, + "step": 11462 + }, + { + "epoch": 0.49, + "grad_norm": 0.51171875, + "learning_rate": 0.00017233032596826792, + "loss": 2.7242, + "step": 11463 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.00017232569377411457, + "loss": 2.4498, + "step": 11464 + }, + { + "epoch": 0.49, + "grad_norm": 0.435546875, + "learning_rate": 0.00017232106125452204, + "loss": 2.6159, + "step": 11465 + }, + { + "epoch": 0.49, + "grad_norm": 0.5078125, + "learning_rate": 0.00017231642840951115, + "loss": 2.5514, + "step": 11466 + }, + { + "epoch": 0.49, + "grad_norm": 0.423828125, + "learning_rate": 0.00017231179523910276, + "loss": 2.5572, + "step": 11467 + }, + { + "epoch": 0.49, + "grad_norm": 0.458984375, + "learning_rate": 0.0001723071617433177, + "loss": 2.5898, + "step": 11468 + }, + { + "epoch": 0.49, + "grad_norm": 0.46484375, + "learning_rate": 0.00017230252792217686, + "loss": 2.6654, + "step": 11469 + }, + { + "epoch": 0.49, + "grad_norm": 0.490234375, + "learning_rate": 0.0001722978937757011, + "loss": 2.7207, + "step": 11470 + }, + { + "epoch": 0.49, + "grad_norm": 0.451171875, + "learning_rate": 0.00017229325930391117, + "loss": 2.6689, + "step": 11471 + }, + { + "epoch": 0.49, + "grad_norm": 0.4453125, + "learning_rate": 0.00017228862450682805, + "loss": 2.5125, + "step": 11472 + }, + { + "epoch": 0.49, + "grad_norm": 0.427734375, + "learning_rate": 0.00017228398938447253, + "loss": 2.608, + "step": 11473 + }, + { + "epoch": 0.49, + "grad_norm": 0.48828125, + "learning_rate": 0.00017227935393686545, + "loss": 2.5866, + "step": 11474 + }, + { + "epoch": 0.49, + "grad_norm": 0.466796875, + "learning_rate": 0.00017227471816402772, + "loss": 2.6284, + "step": 11475 + }, + { + "epoch": 0.49, + "grad_norm": 0.41015625, + "learning_rate": 0.00017227008206598015, + "loss": 2.6164, + "step": 11476 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.00017226544564274363, + "loss": 2.535, + "step": 11477 + }, + { + "epoch": 0.49, + "grad_norm": 0.427734375, + "learning_rate": 0.00017226080889433902, + "loss": 2.3851, + "step": 11478 + }, + { + "epoch": 0.49, + "grad_norm": 0.44140625, + "learning_rate": 0.00017225617182078717, + "loss": 2.6348, + "step": 11479 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.00017225153442210894, + "loss": 2.5094, + "step": 11480 + }, + { + "epoch": 0.49, + "grad_norm": 0.41796875, + "learning_rate": 0.00017224689669832524, + "loss": 2.7245, + "step": 11481 + }, + { + "epoch": 0.49, + "grad_norm": 0.48828125, + "learning_rate": 0.00017224225864945688, + "loss": 2.6509, + "step": 11482 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.0001722376202755248, + "loss": 2.6185, + "step": 11483 + }, + { + "epoch": 0.49, + "grad_norm": 0.5078125, + "learning_rate": 0.00017223298157654978, + "loss": 2.6667, + "step": 11484 + }, + { + "epoch": 0.49, + "grad_norm": 0.458984375, + "learning_rate": 0.00017222834255255278, + "loss": 2.5065, + "step": 11485 + }, + { + "epoch": 0.49, + "grad_norm": 0.45703125, + "learning_rate": 0.0001722237032035546, + "loss": 2.611, + "step": 11486 + }, + { + "epoch": 0.49, + "grad_norm": 0.470703125, + "learning_rate": 0.00017221906352957617, + "loss": 2.6496, + "step": 11487 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.00017221442353063834, + "loss": 2.4803, + "step": 11488 + }, + { + "epoch": 0.49, + "grad_norm": 0.486328125, + "learning_rate": 0.000172209783206762, + "loss": 2.5912, + "step": 11489 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.00017220514255796802, + "loss": 2.6127, + "step": 11490 + }, + { + "epoch": 0.49, + "grad_norm": 0.45703125, + "learning_rate": 0.0001722005015842773, + "loss": 2.5566, + "step": 11491 + }, + { + "epoch": 0.49, + "grad_norm": 0.43359375, + "learning_rate": 0.00017219586028571068, + "loss": 2.565, + "step": 11492 + }, + { + "epoch": 0.49, + "grad_norm": 0.53515625, + "learning_rate": 0.0001721912186622891, + "loss": 2.6677, + "step": 11493 + }, + { + "epoch": 0.49, + "grad_norm": 0.478515625, + "learning_rate": 0.00017218657671403342, + "loss": 2.4312, + "step": 11494 + }, + { + "epoch": 0.49, + "grad_norm": 0.453125, + "learning_rate": 0.00017218193444096453, + "loss": 2.6369, + "step": 11495 + }, + { + "epoch": 0.49, + "grad_norm": 0.51171875, + "learning_rate": 0.0001721772918431033, + "loss": 2.562, + "step": 11496 + }, + { + "epoch": 0.49, + "grad_norm": 0.5234375, + "learning_rate": 0.0001721726489204706, + "loss": 2.5706, + "step": 11497 + }, + { + "epoch": 0.49, + "grad_norm": 0.419921875, + "learning_rate": 0.00017216800567308743, + "loss": 2.5235, + "step": 11498 + }, + { + "epoch": 0.49, + "grad_norm": 0.443359375, + "learning_rate": 0.00017216336210097456, + "loss": 2.6543, + "step": 11499 + }, + { + "epoch": 0.49, + "grad_norm": 0.486328125, + "learning_rate": 0.00017215871820415294, + "loss": 2.5175, + "step": 11500 + }, + { + "epoch": 0.49, + "grad_norm": 0.4453125, + "learning_rate": 0.00017215407398264346, + "loss": 2.7747, + "step": 11501 + }, + { + "epoch": 0.49, + "grad_norm": 0.421875, + "learning_rate": 0.00017214942943646704, + "loss": 2.5507, + "step": 11502 + }, + { + "epoch": 0.49, + "grad_norm": 0.48828125, + "learning_rate": 0.00017214478456564455, + "loss": 2.6003, + "step": 11503 + }, + { + "epoch": 0.49, + "grad_norm": 0.421875, + "learning_rate": 0.00017214013937019685, + "loss": 2.5902, + "step": 11504 + }, + { + "epoch": 0.49, + "grad_norm": 0.45703125, + "learning_rate": 0.00017213549385014493, + "loss": 2.556, + "step": 11505 + }, + { + "epoch": 0.49, + "grad_norm": 0.421875, + "learning_rate": 0.00017213084800550964, + "loss": 2.5569, + "step": 11506 + }, + { + "epoch": 0.49, + "grad_norm": 0.455078125, + "learning_rate": 0.0001721262018363119, + "loss": 2.6931, + "step": 11507 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.0001721215553425726, + "loss": 2.6435, + "step": 11508 + }, + { + "epoch": 0.49, + "grad_norm": 0.4453125, + "learning_rate": 0.00017211690852431266, + "loss": 2.5009, + "step": 11509 + }, + { + "epoch": 0.49, + "grad_norm": 0.515625, + "learning_rate": 0.000172112261381553, + "loss": 2.6655, + "step": 11510 + }, + { + "epoch": 0.49, + "grad_norm": 0.443359375, + "learning_rate": 0.0001721076139143145, + "loss": 2.4927, + "step": 11511 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.0001721029661226181, + "loss": 2.6316, + "step": 11512 + }, + { + "epoch": 0.49, + "grad_norm": 0.48828125, + "learning_rate": 0.0001720983180064847, + "loss": 2.7441, + "step": 11513 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.0001720936695659352, + "loss": 2.5308, + "step": 11514 + }, + { + "epoch": 0.49, + "grad_norm": 0.42578125, + "learning_rate": 0.00017208902080099057, + "loss": 2.552, + "step": 11515 + }, + { + "epoch": 0.49, + "grad_norm": 0.455078125, + "learning_rate": 0.00017208437171167167, + "loss": 2.4988, + "step": 11516 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.00017207972229799944, + "loss": 2.7108, + "step": 11517 + }, + { + "epoch": 0.49, + "grad_norm": 0.41796875, + "learning_rate": 0.0001720750725599948, + "loss": 2.6421, + "step": 11518 + }, + { + "epoch": 0.49, + "grad_norm": 0.44140625, + "learning_rate": 0.0001720704224976787, + "loss": 2.6049, + "step": 11519 + }, + { + "epoch": 0.49, + "grad_norm": 0.427734375, + "learning_rate": 0.00017206577211107202, + "loss": 2.5673, + "step": 11520 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.00017206112140019571, + "loss": 2.635, + "step": 11521 + }, + { + "epoch": 0.49, + "grad_norm": 0.470703125, + "learning_rate": 0.00017205647036507067, + "loss": 2.5607, + "step": 11522 + }, + { + "epoch": 0.49, + "grad_norm": 0.484375, + "learning_rate": 0.00017205181900571784, + "loss": 2.608, + "step": 11523 + }, + { + "epoch": 0.49, + "grad_norm": 0.44921875, + "learning_rate": 0.0001720471673221582, + "loss": 2.5881, + "step": 11524 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.0001720425153144126, + "loss": 2.5594, + "step": 11525 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.000172037862982502, + "loss": 2.5714, + "step": 11526 + }, + { + "epoch": 0.49, + "grad_norm": 0.419921875, + "learning_rate": 0.0001720332103264474, + "loss": 2.4825, + "step": 11527 + }, + { + "epoch": 0.49, + "grad_norm": 0.451171875, + "learning_rate": 0.00017202855734626963, + "loss": 2.6521, + "step": 11528 + }, + { + "epoch": 0.49, + "grad_norm": 0.52734375, + "learning_rate": 0.00017202390404198967, + "loss": 2.6612, + "step": 11529 + }, + { + "epoch": 0.49, + "grad_norm": 0.4453125, + "learning_rate": 0.00017201925041362848, + "loss": 2.5552, + "step": 11530 + }, + { + "epoch": 0.49, + "grad_norm": 0.443359375, + "learning_rate": 0.000172014596461207, + "loss": 2.613, + "step": 11531 + }, + { + "epoch": 0.49, + "grad_norm": 0.453125, + "learning_rate": 0.00017200994218474614, + "loss": 2.6602, + "step": 11532 + }, + { + "epoch": 0.49, + "grad_norm": 0.45703125, + "learning_rate": 0.00017200528758426684, + "loss": 2.5679, + "step": 11533 + }, + { + "epoch": 0.49, + "grad_norm": 0.4609375, + "learning_rate": 0.0001720006326597901, + "loss": 2.5439, + "step": 11534 + }, + { + "epoch": 0.49, + "grad_norm": 0.451171875, + "learning_rate": 0.0001719959774113368, + "loss": 2.5599, + "step": 11535 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.0001719913218389279, + "loss": 2.615, + "step": 11536 + }, + { + "epoch": 0.49, + "grad_norm": 0.443359375, + "learning_rate": 0.00017198666594258436, + "loss": 2.6658, + "step": 11537 + }, + { + "epoch": 0.49, + "grad_norm": 0.46484375, + "learning_rate": 0.00017198200972232716, + "loss": 2.6032, + "step": 11538 + }, + { + "epoch": 0.49, + "grad_norm": 0.4453125, + "learning_rate": 0.0001719773531781772, + "loss": 2.7139, + "step": 11539 + }, + { + "epoch": 0.49, + "grad_norm": 0.46484375, + "learning_rate": 0.0001719726963101555, + "loss": 2.4639, + "step": 11540 + }, + { + "epoch": 0.49, + "grad_norm": 0.396484375, + "learning_rate": 0.00017196803911828293, + "loss": 2.5191, + "step": 11541 + }, + { + "epoch": 0.49, + "grad_norm": 0.462890625, + "learning_rate": 0.00017196338160258047, + "loss": 2.5589, + "step": 11542 + }, + { + "epoch": 0.49, + "grad_norm": 0.4609375, + "learning_rate": 0.00017195872376306914, + "loss": 2.5355, + "step": 11543 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.0001719540655997698, + "loss": 2.6242, + "step": 11544 + }, + { + "epoch": 0.49, + "grad_norm": 0.44921875, + "learning_rate": 0.0001719494071127035, + "loss": 2.5517, + "step": 11545 + }, + { + "epoch": 0.49, + "grad_norm": 0.4375, + "learning_rate": 0.00017194474830189115, + "loss": 2.5697, + "step": 11546 + }, + { + "epoch": 0.49, + "grad_norm": 0.5390625, + "learning_rate": 0.00017194008916735371, + "loss": 2.6573, + "step": 11547 + }, + { + "epoch": 0.49, + "grad_norm": 0.48828125, + "learning_rate": 0.0001719354297091122, + "loss": 2.7088, + "step": 11548 + }, + { + "epoch": 0.49, + "grad_norm": 0.47265625, + "learning_rate": 0.00017193076992718748, + "loss": 2.6344, + "step": 11549 + }, + { + "epoch": 0.49, + "grad_norm": 0.421875, + "learning_rate": 0.00017192610982160063, + "loss": 2.4772, + "step": 11550 + }, + { + "epoch": 0.49, + "grad_norm": 0.4375, + "learning_rate": 0.00017192144939237258, + "loss": 2.6079, + "step": 11551 + }, + { + "epoch": 0.49, + "grad_norm": 0.458984375, + "learning_rate": 0.00017191678863952428, + "loss": 2.5874, + "step": 11552 + }, + { + "epoch": 0.49, + "grad_norm": 0.412109375, + "learning_rate": 0.0001719121275630767, + "loss": 2.5767, + "step": 11553 + }, + { + "epoch": 0.49, + "grad_norm": 0.46484375, + "learning_rate": 0.00017190746616305085, + "loss": 2.5582, + "step": 11554 + }, + { + "epoch": 0.49, + "grad_norm": 0.466796875, + "learning_rate": 0.00017190280443946766, + "loss": 2.6828, + "step": 11555 + }, + { + "epoch": 0.49, + "grad_norm": 0.498046875, + "learning_rate": 0.00017189814239234813, + "loss": 2.5705, + "step": 11556 + }, + { + "epoch": 0.49, + "grad_norm": 0.416015625, + "learning_rate": 0.00017189348002171324, + "loss": 2.5663, + "step": 11557 + }, + { + "epoch": 0.49, + "grad_norm": 0.42578125, + "learning_rate": 0.00017188881732758398, + "loss": 2.6321, + "step": 11558 + }, + { + "epoch": 0.49, + "grad_norm": 0.490234375, + "learning_rate": 0.0001718841543099813, + "loss": 2.6534, + "step": 11559 + }, + { + "epoch": 0.49, + "grad_norm": 0.453125, + "learning_rate": 0.00017187949096892616, + "loss": 2.5652, + "step": 11560 + }, + { + "epoch": 0.49, + "grad_norm": 0.447265625, + "learning_rate": 0.0001718748273044396, + "loss": 2.7158, + "step": 11561 + }, + { + "epoch": 0.49, + "grad_norm": 0.5078125, + "learning_rate": 0.00017187016331654266, + "loss": 2.6156, + "step": 11562 + }, + { + "epoch": 0.49, + "grad_norm": 0.44921875, + "learning_rate": 0.00017186549900525616, + "loss": 2.6182, + "step": 11563 + }, + { + "epoch": 0.49, + "grad_norm": 0.44140625, + "learning_rate": 0.0001718608343706012, + "loss": 2.5483, + "step": 11564 + }, + { + "epoch": 0.49, + "grad_norm": 0.44140625, + "learning_rate": 0.00017185616941259877, + "loss": 2.5569, + "step": 11565 + }, + { + "epoch": 0.49, + "grad_norm": 0.4921875, + "learning_rate": 0.00017185150413126982, + "loss": 2.7071, + "step": 11566 + }, + { + "epoch": 0.49, + "grad_norm": 0.427734375, + "learning_rate": 0.00017184683852663536, + "loss": 2.493, + "step": 11567 + }, + { + "epoch": 0.49, + "grad_norm": 0.416015625, + "learning_rate": 0.0001718421725987164, + "loss": 2.5049, + "step": 11568 + }, + { + "epoch": 0.49, + "grad_norm": 0.47265625, + "learning_rate": 0.0001718375063475339, + "loss": 2.5582, + "step": 11569 + }, + { + "epoch": 0.49, + "grad_norm": 0.4765625, + "learning_rate": 0.0001718328397731089, + "loss": 2.6559, + "step": 11570 + }, + { + "epoch": 0.49, + "grad_norm": 0.478515625, + "learning_rate": 0.00017182817287546235, + "loss": 2.6241, + "step": 11571 + }, + { + "epoch": 0.49, + "grad_norm": 0.490234375, + "learning_rate": 0.00017182350565461528, + "loss": 2.5316, + "step": 11572 + }, + { + "epoch": 0.49, + "grad_norm": 0.4296875, + "learning_rate": 0.00017181883811058868, + "loss": 2.4547, + "step": 11573 + }, + { + "epoch": 0.49, + "grad_norm": 0.498046875, + "learning_rate": 0.00017181417024340355, + "loss": 2.5873, + "step": 11574 + }, + { + "epoch": 0.49, + "grad_norm": 0.46484375, + "learning_rate": 0.0001718095020530809, + "loss": 2.4986, + "step": 11575 + }, + { + "epoch": 0.49, + "grad_norm": 0.48046875, + "learning_rate": 0.00017180483353964175, + "loss": 2.5683, + "step": 11576 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.00017180016470310706, + "loss": 2.685, + "step": 11577 + }, + { + "epoch": 0.49, + "grad_norm": 0.5, + "learning_rate": 0.0001717954955434979, + "loss": 2.5639, + "step": 11578 + }, + { + "epoch": 0.49, + "grad_norm": 0.45703125, + "learning_rate": 0.00017179082606083522, + "loss": 2.6351, + "step": 11579 + }, + { + "epoch": 0.49, + "grad_norm": 0.466796875, + "learning_rate": 0.00017178615625514006, + "loss": 2.5288, + "step": 11580 + }, + { + "epoch": 0.49, + "grad_norm": 0.396484375, + "learning_rate": 0.0001717814861264335, + "loss": 2.5197, + "step": 11581 + }, + { + "epoch": 0.49, + "grad_norm": 0.455078125, + "learning_rate": 0.0001717768156747364, + "loss": 2.6096, + "step": 11582 + }, + { + "epoch": 0.49, + "grad_norm": 0.44140625, + "learning_rate": 0.0001717721449000699, + "loss": 2.6093, + "step": 11583 + }, + { + "epoch": 0.49, + "grad_norm": 0.44921875, + "learning_rate": 0.00017176747380245498, + "loss": 2.5745, + "step": 11584 + }, + { + "epoch": 0.49, + "grad_norm": 0.498046875, + "learning_rate": 0.00017176280238191262, + "loss": 2.5953, + "step": 11585 + }, + { + "epoch": 0.49, + "grad_norm": 0.48828125, + "learning_rate": 0.0001717581306384639, + "loss": 2.79, + "step": 11586 + }, + { + "epoch": 0.49, + "grad_norm": 0.439453125, + "learning_rate": 0.00017175345857212983, + "loss": 2.5339, + "step": 11587 + }, + { + "epoch": 0.49, + "grad_norm": 0.462890625, + "learning_rate": 0.0001717487861829314, + "loss": 2.7207, + "step": 11588 + }, + { + "epoch": 0.49, + "grad_norm": 0.48828125, + "learning_rate": 0.00017174411347088968, + "loss": 2.595, + "step": 11589 + }, + { + "epoch": 0.49, + "grad_norm": 0.5234375, + "learning_rate": 0.00017173944043602567, + "loss": 2.7864, + "step": 11590 + }, + { + "epoch": 0.49, + "grad_norm": 0.49609375, + "learning_rate": 0.00017173476707836035, + "loss": 2.7444, + "step": 11591 + }, + { + "epoch": 0.49, + "grad_norm": 0.47265625, + "learning_rate": 0.00017173009339791482, + "loss": 2.6101, + "step": 11592 + }, + { + "epoch": 0.49, + "grad_norm": 0.490234375, + "learning_rate": 0.00017172541939471008, + "loss": 2.5246, + "step": 11593 + }, + { + "epoch": 0.49, + "grad_norm": 0.498046875, + "learning_rate": 0.0001717207450687672, + "loss": 2.4609, + "step": 11594 + }, + { + "epoch": 0.49, + "grad_norm": 0.43359375, + "learning_rate": 0.00017171607042010712, + "loss": 2.5473, + "step": 11595 + }, + { + "epoch": 0.49, + "grad_norm": 0.462890625, + "learning_rate": 0.00017171139544875096, + "loss": 2.5844, + "step": 11596 + }, + { + "epoch": 0.49, + "grad_norm": 0.421875, + "learning_rate": 0.00017170672015471972, + "loss": 2.6289, + "step": 11597 + }, + { + "epoch": 0.49, + "grad_norm": 0.416015625, + "learning_rate": 0.00017170204453803448, + "loss": 2.5544, + "step": 11598 + }, + { + "epoch": 0.49, + "grad_norm": 0.51171875, + "learning_rate": 0.00017169736859871623, + "loss": 2.6862, + "step": 11599 + }, + { + "epoch": 0.49, + "grad_norm": 0.486328125, + "learning_rate": 0.000171692692336786, + "loss": 2.5738, + "step": 11600 + }, + { + "epoch": 0.49, + "grad_norm": 0.4453125, + "learning_rate": 0.00017168801575226486, + "loss": 2.6076, + "step": 11601 + }, + { + "epoch": 0.49, + "grad_norm": 0.466796875, + "learning_rate": 0.00017168333884517385, + "loss": 2.515, + "step": 11602 + }, + { + "epoch": 0.49, + "grad_norm": 0.51953125, + "learning_rate": 0.000171678661615534, + "loss": 2.6538, + "step": 11603 + }, + { + "epoch": 0.49, + "grad_norm": 0.44140625, + "learning_rate": 0.0001716739840633664, + "loss": 2.7247, + "step": 11604 + }, + { + "epoch": 0.49, + "grad_norm": 0.482421875, + "learning_rate": 0.00017166930618869205, + "loss": 2.5513, + "step": 11605 + }, + { + "epoch": 0.49, + "grad_norm": 0.4765625, + "learning_rate": 0.00017166462799153201, + "loss": 2.604, + "step": 11606 + }, + { + "epoch": 0.49, + "grad_norm": 0.494140625, + "learning_rate": 0.00017165994947190732, + "loss": 2.5991, + "step": 11607 + }, + { + "epoch": 0.49, + "grad_norm": 0.45703125, + "learning_rate": 0.0001716552706298391, + "loss": 2.5574, + "step": 11608 + }, + { + "epoch": 0.49, + "grad_norm": 0.396484375, + "learning_rate": 0.0001716505914653483, + "loss": 2.4441, + "step": 11609 + }, + { + "epoch": 0.49, + "grad_norm": 0.482421875, + "learning_rate": 0.00017164591197845605, + "loss": 2.7068, + "step": 11610 + }, + { + "epoch": 0.49, + "grad_norm": 0.51953125, + "learning_rate": 0.00017164123216918337, + "loss": 2.6571, + "step": 11611 + }, + { + "epoch": 0.49, + "grad_norm": 0.44140625, + "learning_rate": 0.0001716365520375513, + "loss": 2.6327, + "step": 11612 + }, + { + "epoch": 0.49, + "grad_norm": 0.455078125, + "learning_rate": 0.00017163187158358097, + "loss": 2.5425, + "step": 11613 + }, + { + "epoch": 0.49, + "grad_norm": 0.484375, + "learning_rate": 0.00017162719080729337, + "loss": 2.8211, + "step": 11614 + }, + { + "epoch": 0.49, + "grad_norm": 0.431640625, + "learning_rate": 0.00017162250970870958, + "loss": 2.5618, + "step": 11615 + }, + { + "epoch": 0.49, + "grad_norm": 0.4921875, + "learning_rate": 0.0001716178282878507, + "loss": 2.6536, + "step": 11616 + }, + { + "epoch": 0.5, + "grad_norm": 0.447265625, + "learning_rate": 0.00017161314654473776, + "loss": 2.6819, + "step": 11617 + }, + { + "epoch": 0.5, + "grad_norm": 0.48046875, + "learning_rate": 0.00017160846447939184, + "loss": 2.5802, + "step": 11618 + }, + { + "epoch": 0.5, + "grad_norm": 0.455078125, + "learning_rate": 0.00017160378209183397, + "loss": 2.6046, + "step": 11619 + }, + { + "epoch": 0.5, + "grad_norm": 0.421875, + "learning_rate": 0.00017159909938208525, + "loss": 2.6388, + "step": 11620 + }, + { + "epoch": 0.5, + "grad_norm": 0.44140625, + "learning_rate": 0.00017159441635016678, + "loss": 2.617, + "step": 11621 + }, + { + "epoch": 0.5, + "grad_norm": 0.4609375, + "learning_rate": 0.00017158973299609957, + "loss": 2.5903, + "step": 11622 + }, + { + "epoch": 0.5, + "grad_norm": 0.48046875, + "learning_rate": 0.00017158504931990475, + "loss": 2.6836, + "step": 11623 + }, + { + "epoch": 0.5, + "grad_norm": 0.4609375, + "learning_rate": 0.00017158036532160335, + "loss": 2.6764, + "step": 11624 + }, + { + "epoch": 0.5, + "grad_norm": 0.486328125, + "learning_rate": 0.00017157568100121647, + "loss": 2.618, + "step": 11625 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.0001715709963587652, + "loss": 2.6455, + "step": 11626 + }, + { + "epoch": 0.5, + "grad_norm": 0.447265625, + "learning_rate": 0.00017156631139427056, + "loss": 2.6284, + "step": 11627 + }, + { + "epoch": 0.5, + "grad_norm": 0.46484375, + "learning_rate": 0.00017156162610775373, + "loss": 2.7635, + "step": 11628 + }, + { + "epoch": 0.5, + "grad_norm": 0.51171875, + "learning_rate": 0.0001715569404992357, + "loss": 2.7209, + "step": 11629 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.0001715522545687376, + "loss": 2.5993, + "step": 11630 + }, + { + "epoch": 0.5, + "grad_norm": 0.453125, + "learning_rate": 0.00017154756831628049, + "loss": 2.5848, + "step": 11631 + }, + { + "epoch": 0.5, + "grad_norm": 0.484375, + "learning_rate": 0.00017154288174188548, + "loss": 2.5497, + "step": 11632 + }, + { + "epoch": 0.5, + "grad_norm": 0.4296875, + "learning_rate": 0.00017153819484557365, + "loss": 2.5743, + "step": 11633 + }, + { + "epoch": 0.5, + "grad_norm": 0.482421875, + "learning_rate": 0.00017153350762736609, + "loss": 2.7301, + "step": 11634 + }, + { + "epoch": 0.5, + "grad_norm": 0.443359375, + "learning_rate": 0.0001715288200872839, + "loss": 2.5967, + "step": 11635 + }, + { + "epoch": 0.5, + "grad_norm": 0.46484375, + "learning_rate": 0.00017152413222534812, + "loss": 2.8073, + "step": 11636 + }, + { + "epoch": 0.5, + "grad_norm": 0.48046875, + "learning_rate": 0.00017151944404157992, + "loss": 2.5871, + "step": 11637 + }, + { + "epoch": 0.5, + "grad_norm": 0.435546875, + "learning_rate": 0.00017151475553600032, + "loss": 2.6138, + "step": 11638 + }, + { + "epoch": 0.5, + "grad_norm": 0.478515625, + "learning_rate": 0.0001715100667086305, + "loss": 2.6243, + "step": 11639 + }, + { + "epoch": 0.5, + "grad_norm": 0.470703125, + "learning_rate": 0.00017150537755949153, + "loss": 2.6107, + "step": 11640 + }, + { + "epoch": 0.5, + "grad_norm": 0.470703125, + "learning_rate": 0.00017150068808860445, + "loss": 2.6368, + "step": 11641 + }, + { + "epoch": 0.5, + "grad_norm": 0.5, + "learning_rate": 0.0001714959982959904, + "loss": 2.5862, + "step": 11642 + }, + { + "epoch": 0.5, + "grad_norm": 0.404296875, + "learning_rate": 0.00017149130818167052, + "loss": 2.5689, + "step": 11643 + }, + { + "epoch": 0.5, + "grad_norm": 0.396484375, + "learning_rate": 0.00017148661774566586, + "loss": 2.4626, + "step": 11644 + }, + { + "epoch": 0.5, + "grad_norm": 0.421875, + "learning_rate": 0.00017148192698799753, + "loss": 2.5258, + "step": 11645 + }, + { + "epoch": 0.5, + "grad_norm": 0.5234375, + "learning_rate": 0.0001714772359086867, + "loss": 2.6091, + "step": 11646 + }, + { + "epoch": 0.5, + "grad_norm": 0.478515625, + "learning_rate": 0.00017147254450775438, + "loss": 2.4757, + "step": 11647 + }, + { + "epoch": 0.5, + "grad_norm": 0.439453125, + "learning_rate": 0.00017146785278522174, + "loss": 2.4923, + "step": 11648 + }, + { + "epoch": 0.5, + "grad_norm": 0.453125, + "learning_rate": 0.0001714631607411099, + "loss": 2.6275, + "step": 11649 + }, + { + "epoch": 0.5, + "grad_norm": 0.43359375, + "learning_rate": 0.00017145846837543995, + "loss": 2.6341, + "step": 11650 + }, + { + "epoch": 0.5, + "grad_norm": 0.478515625, + "learning_rate": 0.00017145377568823298, + "loss": 2.6476, + "step": 11651 + }, + { + "epoch": 0.5, + "grad_norm": 0.443359375, + "learning_rate": 0.00017144908267951016, + "loss": 2.5303, + "step": 11652 + }, + { + "epoch": 0.5, + "grad_norm": 0.494140625, + "learning_rate": 0.0001714443893492926, + "loss": 2.5598, + "step": 11653 + }, + { + "epoch": 0.5, + "grad_norm": 0.458984375, + "learning_rate": 0.00017143969569760134, + "loss": 2.4722, + "step": 11654 + }, + { + "epoch": 0.5, + "grad_norm": 0.4453125, + "learning_rate": 0.0001714350017244576, + "loss": 2.6593, + "step": 11655 + }, + { + "epoch": 0.5, + "grad_norm": 0.48046875, + "learning_rate": 0.00017143030742988247, + "loss": 2.6136, + "step": 11656 + }, + { + "epoch": 0.5, + "grad_norm": 0.48828125, + "learning_rate": 0.00017142561281389702, + "loss": 2.5724, + "step": 11657 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.00017142091787652244, + "loss": 2.628, + "step": 11658 + }, + { + "epoch": 0.5, + "grad_norm": 0.474609375, + "learning_rate": 0.00017141622261777984, + "loss": 2.529, + "step": 11659 + }, + { + "epoch": 0.5, + "grad_norm": 0.51953125, + "learning_rate": 0.0001714115270376903, + "loss": 2.7984, + "step": 11660 + }, + { + "epoch": 0.5, + "grad_norm": 0.486328125, + "learning_rate": 0.00017140683113627502, + "loss": 2.5252, + "step": 11661 + }, + { + "epoch": 0.5, + "grad_norm": 0.470703125, + "learning_rate": 0.0001714021349135551, + "loss": 2.6, + "step": 11662 + }, + { + "epoch": 0.5, + "grad_norm": 0.5390625, + "learning_rate": 0.00017139743836955162, + "loss": 2.6124, + "step": 11663 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017139274150428582, + "loss": 2.7195, + "step": 11664 + }, + { + "epoch": 0.5, + "grad_norm": 0.51171875, + "learning_rate": 0.00017138804431777874, + "loss": 2.7775, + "step": 11665 + }, + { + "epoch": 0.5, + "grad_norm": 0.53515625, + "learning_rate": 0.00017138334681005154, + "loss": 2.679, + "step": 11666 + }, + { + "epoch": 0.5, + "grad_norm": 0.458984375, + "learning_rate": 0.0001713786489811254, + "loss": 2.5829, + "step": 11667 + }, + { + "epoch": 0.5, + "grad_norm": 0.380859375, + "learning_rate": 0.00017137395083102142, + "loss": 2.5934, + "step": 11668 + }, + { + "epoch": 0.5, + "grad_norm": 0.466796875, + "learning_rate": 0.0001713692523597607, + "loss": 2.6875, + "step": 11669 + }, + { + "epoch": 0.5, + "grad_norm": 0.5234375, + "learning_rate": 0.0001713645535673645, + "loss": 2.6438, + "step": 11670 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017135985445385386, + "loss": 2.6195, + "step": 11671 + }, + { + "epoch": 0.5, + "grad_norm": 0.51953125, + "learning_rate": 0.00017135515501924992, + "loss": 2.7133, + "step": 11672 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.0001713504552635739, + "loss": 2.4744, + "step": 11673 + }, + { + "epoch": 0.5, + "grad_norm": 0.5, + "learning_rate": 0.0001713457551868469, + "loss": 2.6371, + "step": 11674 + }, + { + "epoch": 0.5, + "grad_norm": 0.44921875, + "learning_rate": 0.00017134105478909005, + "loss": 2.5341, + "step": 11675 + }, + { + "epoch": 0.5, + "grad_norm": 0.43359375, + "learning_rate": 0.00017133635407032456, + "loss": 2.5854, + "step": 11676 + }, + { + "epoch": 0.5, + "grad_norm": 0.408203125, + "learning_rate": 0.00017133165303057153, + "loss": 2.5779, + "step": 11677 + }, + { + "epoch": 0.5, + "grad_norm": 0.46875, + "learning_rate": 0.00017132695166985213, + "loss": 2.6718, + "step": 11678 + }, + { + "epoch": 0.5, + "grad_norm": 0.427734375, + "learning_rate": 0.00017132224998818752, + "loss": 2.5467, + "step": 11679 + }, + { + "epoch": 0.5, + "grad_norm": 0.451171875, + "learning_rate": 0.00017131754798559886, + "loss": 2.6122, + "step": 11680 + }, + { + "epoch": 0.5, + "grad_norm": 0.46484375, + "learning_rate": 0.00017131284566210725, + "loss": 2.5649, + "step": 11681 + }, + { + "epoch": 0.5, + "grad_norm": 0.46484375, + "learning_rate": 0.00017130814301773395, + "loss": 2.6908, + "step": 11682 + }, + { + "epoch": 0.5, + "grad_norm": 0.484375, + "learning_rate": 0.00017130344005250004, + "loss": 2.6554, + "step": 11683 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.00017129873676642672, + "loss": 2.5281, + "step": 11684 + }, + { + "epoch": 0.5, + "grad_norm": 0.40234375, + "learning_rate": 0.00017129403315953513, + "loss": 2.5148, + "step": 11685 + }, + { + "epoch": 0.5, + "grad_norm": 0.4921875, + "learning_rate": 0.00017128932923184644, + "loss": 2.5763, + "step": 11686 + }, + { + "epoch": 0.5, + "grad_norm": 0.466796875, + "learning_rate": 0.00017128462498338184, + "loss": 2.6581, + "step": 11687 + }, + { + "epoch": 0.5, + "grad_norm": 0.423828125, + "learning_rate": 0.0001712799204141625, + "loss": 2.5045, + "step": 11688 + }, + { + "epoch": 0.5, + "grad_norm": 0.421875, + "learning_rate": 0.00017127521552420954, + "loss": 2.5268, + "step": 11689 + }, + { + "epoch": 0.5, + "grad_norm": 0.451171875, + "learning_rate": 0.00017127051031354414, + "loss": 2.5799, + "step": 11690 + }, + { + "epoch": 0.5, + "grad_norm": 0.494140625, + "learning_rate": 0.0001712658047821875, + "loss": 2.6254, + "step": 11691 + }, + { + "epoch": 0.5, + "grad_norm": 0.44921875, + "learning_rate": 0.0001712610989301608, + "loss": 2.7682, + "step": 11692 + }, + { + "epoch": 0.5, + "grad_norm": 0.423828125, + "learning_rate": 0.0001712563927574852, + "loss": 2.5533, + "step": 11693 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017125168626418185, + "loss": 2.5839, + "step": 11694 + }, + { + "epoch": 0.5, + "grad_norm": 0.5078125, + "learning_rate": 0.00017124697945027198, + "loss": 2.5799, + "step": 11695 + }, + { + "epoch": 0.5, + "grad_norm": 0.51171875, + "learning_rate": 0.00017124227231577672, + "loss": 2.7246, + "step": 11696 + }, + { + "epoch": 0.5, + "grad_norm": 0.439453125, + "learning_rate": 0.00017123756486071728, + "loss": 2.6473, + "step": 11697 + }, + { + "epoch": 0.5, + "grad_norm": 0.515625, + "learning_rate": 0.0001712328570851148, + "loss": 2.6111, + "step": 11698 + }, + { + "epoch": 0.5, + "grad_norm": 0.44921875, + "learning_rate": 0.00017122814898899055, + "loss": 2.4931, + "step": 11699 + }, + { + "epoch": 0.5, + "grad_norm": 0.53125, + "learning_rate": 0.00017122344057236564, + "loss": 2.6737, + "step": 11700 + }, + { + "epoch": 0.5, + "grad_norm": 0.470703125, + "learning_rate": 0.00017121873183526128, + "loss": 2.6471, + "step": 11701 + }, + { + "epoch": 0.5, + "grad_norm": 0.5078125, + "learning_rate": 0.00017121402277769864, + "loss": 2.677, + "step": 11702 + }, + { + "epoch": 0.5, + "grad_norm": 0.4921875, + "learning_rate": 0.0001712093133996989, + "loss": 2.8107, + "step": 11703 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.00017120460370128332, + "loss": 2.6269, + "step": 11704 + }, + { + "epoch": 0.5, + "grad_norm": 0.455078125, + "learning_rate": 0.00017119989368247302, + "loss": 2.5798, + "step": 11705 + }, + { + "epoch": 0.5, + "grad_norm": 0.490234375, + "learning_rate": 0.00017119518334328924, + "loss": 2.6195, + "step": 11706 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017119047268375313, + "loss": 2.7729, + "step": 11707 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017118576170388594, + "loss": 2.6969, + "step": 11708 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.0001711810504037088, + "loss": 2.637, + "step": 11709 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.00017117633878324298, + "loss": 2.6973, + "step": 11710 + }, + { + "epoch": 0.5, + "grad_norm": 0.431640625, + "learning_rate": 0.0001711716268425096, + "loss": 2.7428, + "step": 11711 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017116691458152997, + "loss": 2.5706, + "step": 11712 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.00017116220200032522, + "loss": 2.7524, + "step": 11713 + }, + { + "epoch": 0.5, + "grad_norm": 0.4140625, + "learning_rate": 0.00017115748909891655, + "loss": 2.4855, + "step": 11714 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.00017115277587732516, + "loss": 2.5138, + "step": 11715 + }, + { + "epoch": 0.5, + "grad_norm": 0.421875, + "learning_rate": 0.00017114806233557231, + "loss": 2.6987, + "step": 11716 + }, + { + "epoch": 0.5, + "grad_norm": 0.4375, + "learning_rate": 0.0001711433484736792, + "loss": 2.6577, + "step": 11717 + }, + { + "epoch": 0.5, + "grad_norm": 0.515625, + "learning_rate": 0.000171138634291667, + "loss": 2.583, + "step": 11718 + }, + { + "epoch": 0.5, + "grad_norm": 0.458984375, + "learning_rate": 0.00017113391978955695, + "loss": 2.6991, + "step": 11719 + }, + { + "epoch": 0.5, + "grad_norm": 0.44921875, + "learning_rate": 0.00017112920496737023, + "loss": 2.5921, + "step": 11720 + }, + { + "epoch": 0.5, + "grad_norm": 0.447265625, + "learning_rate": 0.00017112448982512811, + "loss": 2.6055, + "step": 11721 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.0001711197743628518, + "loss": 2.5873, + "step": 11722 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017111505858056244, + "loss": 2.7862, + "step": 11723 + }, + { + "epoch": 0.5, + "grad_norm": 0.48046875, + "learning_rate": 0.00017111034247828134, + "loss": 2.5051, + "step": 11724 + }, + { + "epoch": 0.5, + "grad_norm": 0.466796875, + "learning_rate": 0.0001711056260560297, + "loss": 2.5819, + "step": 11725 + }, + { + "epoch": 0.5, + "grad_norm": 0.458984375, + "learning_rate": 0.00017110090931382873, + "loss": 2.724, + "step": 11726 + }, + { + "epoch": 0.5, + "grad_norm": 0.42578125, + "learning_rate": 0.00017109619225169962, + "loss": 2.5622, + "step": 11727 + }, + { + "epoch": 0.5, + "grad_norm": 0.453125, + "learning_rate": 0.00017109147486966367, + "loss": 2.6951, + "step": 11728 + }, + { + "epoch": 0.5, + "grad_norm": 0.474609375, + "learning_rate": 0.000171086757167742, + "loss": 2.7709, + "step": 11729 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017108203914595595, + "loss": 2.5798, + "step": 11730 + }, + { + "epoch": 0.5, + "grad_norm": 0.4765625, + "learning_rate": 0.00017107732080432668, + "loss": 2.6382, + "step": 11731 + }, + { + "epoch": 0.5, + "grad_norm": 0.435546875, + "learning_rate": 0.00017107260214287547, + "loss": 2.7053, + "step": 11732 + }, + { + "epoch": 0.5, + "grad_norm": 0.486328125, + "learning_rate": 0.0001710678831616235, + "loss": 2.6468, + "step": 11733 + }, + { + "epoch": 0.5, + "grad_norm": 0.486328125, + "learning_rate": 0.00017106316386059206, + "loss": 2.6905, + "step": 11734 + }, + { + "epoch": 0.5, + "eval_loss": 2.587218999862671, + "eval_runtime": 82.828, + "eval_samples_per_second": 18.158, + "eval_steps_per_second": 2.27, + "step": 11734 + }, + { + "epoch": 0.5, + "grad_norm": 0.46484375, + "learning_rate": 0.0001710584442398023, + "loss": 2.6853, + "step": 11735 + }, + { + "epoch": 0.5, + "grad_norm": 0.451171875, + "learning_rate": 0.00017105372429927554, + "loss": 2.6542, + "step": 11736 + }, + { + "epoch": 0.5, + "grad_norm": 0.435546875, + "learning_rate": 0.000171049004039033, + "loss": 2.6363, + "step": 11737 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.00017104428345909587, + "loss": 2.5518, + "step": 11738 + }, + { + "epoch": 0.5, + "grad_norm": 0.419921875, + "learning_rate": 0.00017103956255948549, + "loss": 2.5329, + "step": 11739 + }, + { + "epoch": 0.5, + "grad_norm": 0.46484375, + "learning_rate": 0.00017103484134022297, + "loss": 2.5656, + "step": 11740 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.00017103011980132968, + "loss": 2.6847, + "step": 11741 + }, + { + "epoch": 0.5, + "grad_norm": 0.443359375, + "learning_rate": 0.00017102539794282677, + "loss": 2.5679, + "step": 11742 + }, + { + "epoch": 0.5, + "grad_norm": 0.53515625, + "learning_rate": 0.00017102067576473554, + "loss": 2.5757, + "step": 11743 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017101595326707723, + "loss": 2.6696, + "step": 11744 + }, + { + "epoch": 0.5, + "grad_norm": 0.4375, + "learning_rate": 0.00017101123044987307, + "loss": 2.4313, + "step": 11745 + }, + { + "epoch": 0.5, + "grad_norm": 0.427734375, + "learning_rate": 0.00017100650731314434, + "loss": 2.6669, + "step": 11746 + }, + { + "epoch": 0.5, + "grad_norm": 0.40234375, + "learning_rate": 0.0001710017838569123, + "loss": 2.5715, + "step": 11747 + }, + { + "epoch": 0.5, + "grad_norm": 0.55078125, + "learning_rate": 0.00017099706008119814, + "loss": 2.7513, + "step": 11748 + }, + { + "epoch": 0.5, + "grad_norm": 0.455078125, + "learning_rate": 0.00017099233598602315, + "loss": 2.6526, + "step": 11749 + }, + { + "epoch": 0.5, + "grad_norm": 0.458984375, + "learning_rate": 0.0001709876115714086, + "loss": 2.5083, + "step": 11750 + }, + { + "epoch": 0.5, + "grad_norm": 0.42578125, + "learning_rate": 0.00017098288683737578, + "loss": 2.5883, + "step": 11751 + }, + { + "epoch": 0.5, + "grad_norm": 0.44140625, + "learning_rate": 0.00017097816178394587, + "loss": 2.5423, + "step": 11752 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.00017097343641114018, + "loss": 2.6219, + "step": 11753 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017096871071897995, + "loss": 2.6056, + "step": 11754 + }, + { + "epoch": 0.5, + "grad_norm": 0.482421875, + "learning_rate": 0.0001709639847074865, + "loss": 2.5679, + "step": 11755 + }, + { + "epoch": 0.5, + "grad_norm": 0.439453125, + "learning_rate": 0.00017095925837668104, + "loss": 2.6299, + "step": 11756 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.00017095453172658482, + "loss": 2.696, + "step": 11757 + }, + { + "epoch": 0.5, + "grad_norm": 0.423828125, + "learning_rate": 0.00017094980475721916, + "loss": 2.5219, + "step": 11758 + }, + { + "epoch": 0.5, + "grad_norm": 0.470703125, + "learning_rate": 0.0001709450774686053, + "loss": 2.5758, + "step": 11759 + }, + { + "epoch": 0.5, + "grad_norm": 0.478515625, + "learning_rate": 0.00017094034986076453, + "loss": 2.7292, + "step": 11760 + }, + { + "epoch": 0.5, + "grad_norm": 0.447265625, + "learning_rate": 0.0001709356219337181, + "loss": 2.5671, + "step": 11761 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017093089368748729, + "loss": 2.4671, + "step": 11762 + }, + { + "epoch": 0.5, + "grad_norm": 0.3671875, + "learning_rate": 0.0001709261651220934, + "loss": 2.5497, + "step": 11763 + }, + { + "epoch": 0.5, + "grad_norm": 0.439453125, + "learning_rate": 0.0001709214362375577, + "loss": 2.6276, + "step": 11764 + }, + { + "epoch": 0.5, + "grad_norm": 0.44921875, + "learning_rate": 0.0001709167070339014, + "loss": 2.6263, + "step": 11765 + }, + { + "epoch": 0.5, + "grad_norm": 0.486328125, + "learning_rate": 0.00017091197751114588, + "loss": 2.653, + "step": 11766 + }, + { + "epoch": 0.5, + "grad_norm": 0.4921875, + "learning_rate": 0.00017090724766931237, + "loss": 2.681, + "step": 11767 + }, + { + "epoch": 0.5, + "grad_norm": 0.46484375, + "learning_rate": 0.00017090251750842214, + "loss": 2.6306, + "step": 11768 + }, + { + "epoch": 0.5, + "grad_norm": 0.435546875, + "learning_rate": 0.00017089778702849653, + "loss": 2.6446, + "step": 11769 + }, + { + "epoch": 0.5, + "grad_norm": 0.453125, + "learning_rate": 0.00017089305622955677, + "loss": 2.6756, + "step": 11770 + }, + { + "epoch": 0.5, + "grad_norm": 0.455078125, + "learning_rate": 0.00017088832511162413, + "loss": 2.6266, + "step": 11771 + }, + { + "epoch": 0.5, + "grad_norm": 0.439453125, + "learning_rate": 0.00017088359367472, + "loss": 2.6343, + "step": 11772 + }, + { + "epoch": 0.5, + "grad_norm": 0.427734375, + "learning_rate": 0.00017087886191886558, + "loss": 2.5178, + "step": 11773 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017087412984408216, + "loss": 2.5969, + "step": 11774 + }, + { + "epoch": 0.5, + "grad_norm": 0.447265625, + "learning_rate": 0.00017086939745039107, + "loss": 2.54, + "step": 11775 + }, + { + "epoch": 0.5, + "grad_norm": 0.53125, + "learning_rate": 0.0001708646647378136, + "loss": 2.5842, + "step": 11776 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017085993170637103, + "loss": 2.6451, + "step": 11777 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.00017085519835608465, + "loss": 2.6344, + "step": 11778 + }, + { + "epoch": 0.5, + "grad_norm": 0.439453125, + "learning_rate": 0.00017085046468697578, + "loss": 2.7586, + "step": 11779 + }, + { + "epoch": 0.5, + "grad_norm": 0.4375, + "learning_rate": 0.00017084573069906572, + "loss": 2.7007, + "step": 11780 + }, + { + "epoch": 0.5, + "grad_norm": 0.44140625, + "learning_rate": 0.00017084099639237576, + "loss": 2.628, + "step": 11781 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.0001708362617669272, + "loss": 2.6875, + "step": 11782 + }, + { + "epoch": 0.5, + "grad_norm": 0.421875, + "learning_rate": 0.00017083152682274136, + "loss": 2.6312, + "step": 11783 + }, + { + "epoch": 0.5, + "grad_norm": 0.46484375, + "learning_rate": 0.00017082679155983955, + "loss": 2.6181, + "step": 11784 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017082205597824304, + "loss": 2.529, + "step": 11785 + }, + { + "epoch": 0.5, + "grad_norm": 0.494140625, + "learning_rate": 0.00017081732007797316, + "loss": 2.7501, + "step": 11786 + }, + { + "epoch": 0.5, + "grad_norm": 0.515625, + "learning_rate": 0.00017081258385905123, + "loss": 2.7622, + "step": 11787 + }, + { + "epoch": 0.5, + "grad_norm": 0.455078125, + "learning_rate": 0.00017080784732149853, + "loss": 2.7891, + "step": 11788 + }, + { + "epoch": 0.5, + "grad_norm": 0.4453125, + "learning_rate": 0.00017080311046533642, + "loss": 2.6933, + "step": 11789 + }, + { + "epoch": 0.5, + "grad_norm": 0.48046875, + "learning_rate": 0.00017079837329058616, + "loss": 2.6013, + "step": 11790 + }, + { + "epoch": 0.5, + "grad_norm": 0.5, + "learning_rate": 0.00017079363579726914, + "loss": 2.7099, + "step": 11791 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.00017078889798540658, + "loss": 2.6379, + "step": 11792 + }, + { + "epoch": 0.5, + "grad_norm": 0.435546875, + "learning_rate": 0.00017078415985501987, + "loss": 2.6192, + "step": 11793 + }, + { + "epoch": 0.5, + "grad_norm": 0.45703125, + "learning_rate": 0.0001707794214061303, + "loss": 2.6474, + "step": 11794 + }, + { + "epoch": 0.5, + "grad_norm": 0.51953125, + "learning_rate": 0.00017077468263875924, + "loss": 2.6258, + "step": 11795 + }, + { + "epoch": 0.5, + "grad_norm": 0.5, + "learning_rate": 0.00017076994355292793, + "loss": 2.5608, + "step": 11796 + }, + { + "epoch": 0.5, + "grad_norm": 0.43359375, + "learning_rate": 0.00017076520414865775, + "loss": 2.6879, + "step": 11797 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017076046442597, + "loss": 2.725, + "step": 11798 + }, + { + "epoch": 0.5, + "grad_norm": 0.482421875, + "learning_rate": 0.00017075572438488606, + "loss": 2.7283, + "step": 11799 + }, + { + "epoch": 0.5, + "grad_norm": 0.42578125, + "learning_rate": 0.0001707509840254272, + "loss": 2.7169, + "step": 11800 + }, + { + "epoch": 0.5, + "grad_norm": 0.486328125, + "learning_rate": 0.00017074624334761476, + "loss": 2.6634, + "step": 11801 + }, + { + "epoch": 0.5, + "grad_norm": 0.40234375, + "learning_rate": 0.00017074150235147007, + "loss": 2.5188, + "step": 11802 + }, + { + "epoch": 0.5, + "grad_norm": 0.458984375, + "learning_rate": 0.0001707367610370145, + "loss": 2.454, + "step": 11803 + }, + { + "epoch": 0.5, + "grad_norm": 0.494140625, + "learning_rate": 0.00017073201940426936, + "loss": 2.5797, + "step": 11804 + }, + { + "epoch": 0.5, + "grad_norm": 0.44140625, + "learning_rate": 0.00017072727745325595, + "loss": 2.5661, + "step": 11805 + }, + { + "epoch": 0.5, + "grad_norm": 0.421875, + "learning_rate": 0.00017072253518399567, + "loss": 2.5722, + "step": 11806 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.0001707177925965098, + "loss": 2.6582, + "step": 11807 + }, + { + "epoch": 0.5, + "grad_norm": 0.41796875, + "learning_rate": 0.00017071304969081973, + "loss": 2.6244, + "step": 11808 + }, + { + "epoch": 0.5, + "grad_norm": 0.451171875, + "learning_rate": 0.00017070830646694676, + "loss": 2.5738, + "step": 11809 + }, + { + "epoch": 0.5, + "grad_norm": 0.419921875, + "learning_rate": 0.00017070356292491225, + "loss": 2.5925, + "step": 11810 + }, + { + "epoch": 0.5, + "grad_norm": 0.44140625, + "learning_rate": 0.00017069881906473758, + "loss": 2.5014, + "step": 11811 + }, + { + "epoch": 0.5, + "grad_norm": 0.51953125, + "learning_rate": 0.000170694074886444, + "loss": 2.7104, + "step": 11812 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017068933039005298, + "loss": 2.6463, + "step": 11813 + }, + { + "epoch": 0.5, + "grad_norm": 0.5, + "learning_rate": 0.00017068458557558576, + "loss": 2.6861, + "step": 11814 + }, + { + "epoch": 0.5, + "grad_norm": 0.455078125, + "learning_rate": 0.0001706798404430638, + "loss": 2.6527, + "step": 11815 + }, + { + "epoch": 0.5, + "grad_norm": 0.404296875, + "learning_rate": 0.00017067509499250834, + "loss": 2.5852, + "step": 11816 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017067034922394077, + "loss": 2.5345, + "step": 11817 + }, + { + "epoch": 0.5, + "grad_norm": 0.470703125, + "learning_rate": 0.00017066560313738248, + "loss": 2.6099, + "step": 11818 + }, + { + "epoch": 0.5, + "grad_norm": 0.44140625, + "learning_rate": 0.00017066085673285482, + "loss": 2.6157, + "step": 11819 + }, + { + "epoch": 0.5, + "grad_norm": 0.44140625, + "learning_rate": 0.00017065611001037908, + "loss": 2.5471, + "step": 11820 + }, + { + "epoch": 0.5, + "grad_norm": 0.46875, + "learning_rate": 0.00017065136296997674, + "loss": 2.7715, + "step": 11821 + }, + { + "epoch": 0.5, + "grad_norm": 0.458984375, + "learning_rate": 0.000170646615611669, + "loss": 2.5959, + "step": 11822 + }, + { + "epoch": 0.5, + "grad_norm": 0.515625, + "learning_rate": 0.00017064186793547736, + "loss": 2.7954, + "step": 11823 + }, + { + "epoch": 0.5, + "grad_norm": 0.50390625, + "learning_rate": 0.00017063711994142313, + "loss": 2.6154, + "step": 11824 + }, + { + "epoch": 0.5, + "grad_norm": 0.455078125, + "learning_rate": 0.00017063237162952765, + "loss": 2.584, + "step": 11825 + }, + { + "epoch": 0.5, + "grad_norm": 0.451171875, + "learning_rate": 0.00017062762299981232, + "loss": 2.6036, + "step": 11826 + }, + { + "epoch": 0.5, + "grad_norm": 0.44140625, + "learning_rate": 0.00017062287405229852, + "loss": 2.5273, + "step": 11827 + }, + { + "epoch": 0.5, + "grad_norm": 0.462890625, + "learning_rate": 0.00017061812478700756, + "loss": 2.4985, + "step": 11828 + }, + { + "epoch": 0.5, + "grad_norm": 0.474609375, + "learning_rate": 0.00017061337520396088, + "loss": 2.6236, + "step": 11829 + }, + { + "epoch": 0.5, + "grad_norm": 0.392578125, + "learning_rate": 0.00017060862530317978, + "loss": 2.6627, + "step": 11830 + }, + { + "epoch": 0.5, + "grad_norm": 0.474609375, + "learning_rate": 0.00017060387508468572, + "loss": 2.6891, + "step": 11831 + }, + { + "epoch": 0.5, + "grad_norm": 0.4921875, + "learning_rate": 0.0001705991245485, + "loss": 2.7875, + "step": 11832 + }, + { + "epoch": 0.5, + "grad_norm": 0.447265625, + "learning_rate": 0.00017059437369464398, + "loss": 2.7089, + "step": 11833 + }, + { + "epoch": 0.5, + "grad_norm": 0.478515625, + "learning_rate": 0.00017058962252313915, + "loss": 2.6843, + "step": 11834 + }, + { + "epoch": 0.5, + "grad_norm": 0.4765625, + "learning_rate": 0.00017058487103400677, + "loss": 2.6578, + "step": 11835 + }, + { + "epoch": 0.5, + "grad_norm": 0.470703125, + "learning_rate": 0.00017058011922726829, + "loss": 2.5369, + "step": 11836 + }, + { + "epoch": 0.5, + "grad_norm": 0.443359375, + "learning_rate": 0.00017057536710294506, + "loss": 2.58, + "step": 11837 + }, + { + "epoch": 0.5, + "grad_norm": 0.4140625, + "learning_rate": 0.00017057061466105845, + "loss": 2.6634, + "step": 11838 + }, + { + "epoch": 0.5, + "grad_norm": 0.5, + "learning_rate": 0.0001705658619016299, + "loss": 2.6313, + "step": 11839 + }, + { + "epoch": 0.5, + "grad_norm": 0.6875, + "learning_rate": 0.00017056110882468075, + "loss": 2.649, + "step": 11840 + }, + { + "epoch": 0.5, + "grad_norm": 0.455078125, + "learning_rate": 0.00017055635543023238, + "loss": 2.7946, + "step": 11841 + }, + { + "epoch": 0.5, + "grad_norm": 0.4609375, + "learning_rate": 0.0001705516017183062, + "loss": 2.6705, + "step": 11842 + }, + { + "epoch": 0.5, + "grad_norm": 0.5234375, + "learning_rate": 0.00017054684768892363, + "loss": 2.6636, + "step": 11843 + }, + { + "epoch": 0.5, + "grad_norm": 0.451171875, + "learning_rate": 0.00017054209334210597, + "loss": 2.689, + "step": 11844 + }, + { + "epoch": 0.5, + "grad_norm": 0.4921875, + "learning_rate": 0.0001705373386778747, + "loss": 2.7766, + "step": 11845 + }, + { + "epoch": 0.5, + "grad_norm": 0.427734375, + "learning_rate": 0.00017053258369625118, + "loss": 2.4639, + "step": 11846 + }, + { + "epoch": 0.5, + "grad_norm": 0.47265625, + "learning_rate": 0.00017052782839725684, + "loss": 2.5949, + "step": 11847 + }, + { + "epoch": 0.5, + "grad_norm": 0.431640625, + "learning_rate": 0.00017052307278091303, + "loss": 2.6002, + "step": 11848 + }, + { + "epoch": 0.5, + "grad_norm": 0.439453125, + "learning_rate": 0.00017051831684724113, + "loss": 2.5032, + "step": 11849 + }, + { + "epoch": 0.5, + "grad_norm": 0.4375, + "learning_rate": 0.0001705135605962626, + "loss": 2.4905, + "step": 11850 + }, + { + "epoch": 0.51, + "grad_norm": 0.46484375, + "learning_rate": 0.00017050880402799882, + "loss": 2.6959, + "step": 11851 + }, + { + "epoch": 0.51, + "grad_norm": 0.486328125, + "learning_rate": 0.0001705040471424712, + "loss": 2.7549, + "step": 11852 + }, + { + "epoch": 0.51, + "grad_norm": 0.50390625, + "learning_rate": 0.0001704992899397011, + "loss": 2.7542, + "step": 11853 + }, + { + "epoch": 0.51, + "grad_norm": 0.455078125, + "learning_rate": 0.00017049453241970997, + "loss": 2.546, + "step": 11854 + }, + { + "epoch": 0.51, + "grad_norm": 0.46875, + "learning_rate": 0.00017048977458251926, + "loss": 2.6628, + "step": 11855 + }, + { + "epoch": 0.51, + "grad_norm": 0.52734375, + "learning_rate": 0.00017048501642815027, + "loss": 2.9487, + "step": 11856 + }, + { + "epoch": 0.51, + "grad_norm": 0.46875, + "learning_rate": 0.00017048025795662447, + "loss": 2.7473, + "step": 11857 + }, + { + "epoch": 0.51, + "grad_norm": 0.44921875, + "learning_rate": 0.0001704754991679633, + "loss": 2.5736, + "step": 11858 + }, + { + "epoch": 0.51, + "grad_norm": 0.42578125, + "learning_rate": 0.0001704707400621881, + "loss": 2.6507, + "step": 11859 + }, + { + "epoch": 0.51, + "grad_norm": 0.4609375, + "learning_rate": 0.00017046598063932036, + "loss": 2.5293, + "step": 11860 + }, + { + "epoch": 0.51, + "grad_norm": 0.453125, + "learning_rate": 0.00017046122089938145, + "loss": 2.6223, + "step": 11861 + }, + { + "epoch": 0.51, + "grad_norm": 0.48828125, + "learning_rate": 0.00017045646084239277, + "loss": 2.7091, + "step": 11862 + }, + { + "epoch": 0.51, + "grad_norm": 0.453125, + "learning_rate": 0.0001704517004683758, + "loss": 2.4918, + "step": 11863 + }, + { + "epoch": 0.51, + "grad_norm": 0.458984375, + "learning_rate": 0.0001704469397773519, + "loss": 2.6091, + "step": 11864 + }, + { + "epoch": 0.51, + "grad_norm": 0.45703125, + "learning_rate": 0.00017044217876934252, + "loss": 2.561, + "step": 11865 + }, + { + "epoch": 0.51, + "grad_norm": 0.4921875, + "learning_rate": 0.00017043741744436906, + "loss": 2.5374, + "step": 11866 + }, + { + "epoch": 0.51, + "grad_norm": 0.494140625, + "learning_rate": 0.00017043265580245302, + "loss": 2.5751, + "step": 11867 + }, + { + "epoch": 0.51, + "grad_norm": 0.4453125, + "learning_rate": 0.00017042789384361573, + "loss": 2.6456, + "step": 11868 + }, + { + "epoch": 0.51, + "grad_norm": 0.46875, + "learning_rate": 0.00017042313156787867, + "loss": 2.682, + "step": 11869 + }, + { + "epoch": 0.51, + "grad_norm": 0.412109375, + "learning_rate": 0.00017041836897526325, + "loss": 2.3376, + "step": 11870 + }, + { + "epoch": 0.51, + "grad_norm": 0.41015625, + "learning_rate": 0.00017041360606579088, + "loss": 2.4598, + "step": 11871 + }, + { + "epoch": 0.51, + "grad_norm": 0.41796875, + "learning_rate": 0.00017040884283948303, + "loss": 2.531, + "step": 11872 + }, + { + "epoch": 0.51, + "grad_norm": 0.462890625, + "learning_rate": 0.00017040407929636112, + "loss": 2.6616, + "step": 11873 + }, + { + "epoch": 0.51, + "grad_norm": 0.46875, + "learning_rate": 0.00017039931543644664, + "loss": 2.6654, + "step": 11874 + }, + { + "epoch": 0.51, + "grad_norm": 0.5, + "learning_rate": 0.00017039455125976088, + "loss": 2.628, + "step": 11875 + }, + { + "epoch": 0.51, + "grad_norm": 0.44921875, + "learning_rate": 0.0001703897867663254, + "loss": 2.5801, + "step": 11876 + }, + { + "epoch": 0.51, + "grad_norm": 0.515625, + "learning_rate": 0.0001703850219561616, + "loss": 2.5695, + "step": 11877 + }, + { + "epoch": 0.51, + "grad_norm": 0.47265625, + "learning_rate": 0.00017038025682929094, + "loss": 2.7531, + "step": 11878 + }, + { + "epoch": 0.51, + "grad_norm": 0.435546875, + "learning_rate": 0.00017037549138573484, + "loss": 2.481, + "step": 11879 + }, + { + "epoch": 0.51, + "grad_norm": 0.49609375, + "learning_rate": 0.00017037072562551474, + "loss": 2.6314, + "step": 11880 + }, + { + "epoch": 0.51, + "grad_norm": 0.462890625, + "learning_rate": 0.0001703659595486521, + "loss": 2.6941, + "step": 11881 + }, + { + "epoch": 0.51, + "grad_norm": 0.419921875, + "learning_rate": 0.00017036119315516835, + "loss": 2.602, + "step": 11882 + }, + { + "epoch": 0.51, + "grad_norm": 0.5, + "learning_rate": 0.00017035642644508493, + "loss": 2.6959, + "step": 11883 + }, + { + "epoch": 0.51, + "grad_norm": 0.478515625, + "learning_rate": 0.00017035165941842336, + "loss": 2.6584, + "step": 11884 + }, + { + "epoch": 0.51, + "grad_norm": 0.494140625, + "learning_rate": 0.00017034689207520498, + "loss": 2.7058, + "step": 11885 + }, + { + "epoch": 0.51, + "grad_norm": 0.453125, + "learning_rate": 0.00017034212441545133, + "loss": 2.5113, + "step": 11886 + }, + { + "epoch": 0.51, + "grad_norm": 0.37109375, + "learning_rate": 0.00017033735643918378, + "loss": 2.5295, + "step": 11887 + }, + { + "epoch": 0.51, + "grad_norm": 0.474609375, + "learning_rate": 0.00017033258814642385, + "loss": 2.6446, + "step": 11888 + }, + { + "epoch": 0.51, + "grad_norm": 0.515625, + "learning_rate": 0.000170327819537193, + "loss": 2.5712, + "step": 11889 + }, + { + "epoch": 0.51, + "grad_norm": 0.4609375, + "learning_rate": 0.00017032305061151265, + "loss": 2.5217, + "step": 11890 + }, + { + "epoch": 0.51, + "grad_norm": 0.462890625, + "learning_rate": 0.0001703182813694043, + "loss": 2.5374, + "step": 11891 + }, + { + "epoch": 0.51, + "grad_norm": 0.453125, + "learning_rate": 0.00017031351181088936, + "loss": 2.5769, + "step": 11892 + }, + { + "epoch": 0.51, + "grad_norm": 0.48046875, + "learning_rate": 0.00017030874193598927, + "loss": 2.7046, + "step": 11893 + }, + { + "epoch": 0.51, + "grad_norm": 0.46484375, + "learning_rate": 0.00017030397174472558, + "loss": 2.596, + "step": 11894 + }, + { + "epoch": 0.51, + "grad_norm": 0.52734375, + "learning_rate": 0.00017029920123711972, + "loss": 2.7434, + "step": 11895 + }, + { + "epoch": 0.51, + "grad_norm": 0.51171875, + "learning_rate": 0.00017029443041319313, + "loss": 2.6868, + "step": 11896 + }, + { + "epoch": 0.51, + "grad_norm": 0.44921875, + "learning_rate": 0.00017028965927296728, + "loss": 2.6085, + "step": 11897 + }, + { + "epoch": 0.51, + "grad_norm": 0.427734375, + "learning_rate": 0.00017028488781646367, + "loss": 2.4948, + "step": 11898 + }, + { + "epoch": 0.51, + "grad_norm": 0.466796875, + "learning_rate": 0.00017028011604370376, + "loss": 2.6969, + "step": 11899 + }, + { + "epoch": 0.51, + "grad_norm": 0.494140625, + "learning_rate": 0.00017027534395470899, + "loss": 2.534, + "step": 11900 + }, + { + "epoch": 0.51, + "grad_norm": 0.490234375, + "learning_rate": 0.00017027057154950085, + "loss": 2.8128, + "step": 11901 + }, + { + "epoch": 0.51, + "grad_norm": 0.50390625, + "learning_rate": 0.00017026579882810085, + "loss": 2.6398, + "step": 11902 + }, + { + "epoch": 0.51, + "grad_norm": 0.5, + "learning_rate": 0.00017026102579053038, + "loss": 2.6086, + "step": 11903 + }, + { + "epoch": 0.51, + "grad_norm": 0.53125, + "learning_rate": 0.00017025625243681102, + "loss": 2.4393, + "step": 11904 + }, + { + "epoch": 0.51, + "grad_norm": 0.49609375, + "learning_rate": 0.00017025147876696418, + "loss": 2.5774, + "step": 11905 + }, + { + "epoch": 0.51, + "grad_norm": 0.46875, + "learning_rate": 0.00017024670478101137, + "loss": 2.6164, + "step": 11906 + }, + { + "epoch": 0.51, + "grad_norm": 0.47265625, + "learning_rate": 0.00017024193047897405, + "loss": 2.72, + "step": 11907 + }, + { + "epoch": 0.51, + "grad_norm": 0.421875, + "learning_rate": 0.00017023715586087374, + "loss": 2.6454, + "step": 11908 + }, + { + "epoch": 0.51, + "grad_norm": 0.44921875, + "learning_rate": 0.0001702323809267319, + "loss": 2.6344, + "step": 11909 + }, + { + "epoch": 0.51, + "grad_norm": 0.486328125, + "learning_rate": 0.00017022760567656997, + "loss": 2.6288, + "step": 11910 + }, + { + "epoch": 0.51, + "grad_norm": 0.54296875, + "learning_rate": 0.00017022283011040949, + "loss": 2.662, + "step": 11911 + }, + { + "epoch": 0.51, + "grad_norm": 0.515625, + "learning_rate": 0.00017021805422827194, + "loss": 2.8014, + "step": 11912 + }, + { + "epoch": 0.51, + "grad_norm": 0.42578125, + "learning_rate": 0.00017021327803017883, + "loss": 2.546, + "step": 11913 + }, + { + "epoch": 0.51, + "grad_norm": 0.490234375, + "learning_rate": 0.0001702085015161516, + "loss": 2.6563, + "step": 11914 + }, + { + "epoch": 0.51, + "grad_norm": 0.4765625, + "learning_rate": 0.0001702037246862118, + "loss": 2.6042, + "step": 11915 + }, + { + "epoch": 0.51, + "grad_norm": 0.53125, + "learning_rate": 0.0001701989475403809, + "loss": 2.7885, + "step": 11916 + }, + { + "epoch": 0.51, + "grad_norm": 0.447265625, + "learning_rate": 0.00017019417007868034, + "loss": 2.6407, + "step": 11917 + }, + { + "epoch": 0.51, + "grad_norm": 0.4296875, + "learning_rate": 0.00017018939230113174, + "loss": 2.5132, + "step": 11918 + }, + { + "epoch": 0.51, + "grad_norm": 0.478515625, + "learning_rate": 0.00017018461420775647, + "loss": 2.7599, + "step": 11919 + }, + { + "epoch": 0.51, + "grad_norm": 0.44140625, + "learning_rate": 0.00017017983579857609, + "loss": 2.5451, + "step": 11920 + }, + { + "epoch": 0.51, + "grad_norm": 0.515625, + "learning_rate": 0.00017017505707361212, + "loss": 2.7068, + "step": 11921 + }, + { + "epoch": 0.51, + "grad_norm": 0.462890625, + "learning_rate": 0.00017017027803288603, + "loss": 2.5606, + "step": 11922 + }, + { + "epoch": 0.51, + "grad_norm": 0.447265625, + "learning_rate": 0.0001701654986764193, + "loss": 2.4711, + "step": 11923 + }, + { + "epoch": 0.51, + "grad_norm": 0.466796875, + "learning_rate": 0.00017016071900423355, + "loss": 2.6182, + "step": 11924 + }, + { + "epoch": 0.51, + "grad_norm": 0.458984375, + "learning_rate": 0.00017015593901635012, + "loss": 2.5672, + "step": 11925 + }, + { + "epoch": 0.51, + "grad_norm": 0.439453125, + "learning_rate": 0.00017015115871279064, + "loss": 2.7139, + "step": 11926 + }, + { + "epoch": 0.51, + "grad_norm": 0.47265625, + "learning_rate": 0.00017014637809357658, + "loss": 2.6104, + "step": 11927 + }, + { + "epoch": 0.51, + "grad_norm": 0.478515625, + "learning_rate": 0.00017014159715872946, + "loss": 2.5127, + "step": 11928 + }, + { + "epoch": 0.51, + "grad_norm": 0.43359375, + "learning_rate": 0.0001701368159082708, + "loss": 2.567, + "step": 11929 + }, + { + "epoch": 0.51, + "grad_norm": 0.474609375, + "learning_rate": 0.0001701320343422221, + "loss": 2.5507, + "step": 11930 + }, + { + "epoch": 0.51, + "grad_norm": 0.45703125, + "learning_rate": 0.00017012725246060484, + "loss": 2.5352, + "step": 11931 + }, + { + "epoch": 0.51, + "grad_norm": 0.451171875, + "learning_rate": 0.00017012247026344058, + "loss": 2.628, + "step": 11932 + }, + { + "epoch": 0.51, + "grad_norm": 0.478515625, + "learning_rate": 0.00017011768775075084, + "loss": 2.5787, + "step": 11933 + }, + { + "epoch": 0.51, + "grad_norm": 0.482421875, + "learning_rate": 0.00017011290492255715, + "loss": 2.7061, + "step": 11934 + }, + { + "epoch": 0.51, + "grad_norm": 0.4453125, + "learning_rate": 0.00017010812177888098, + "loss": 2.5528, + "step": 11935 + }, + { + "epoch": 0.51, + "grad_norm": 0.419921875, + "learning_rate": 0.0001701033383197439, + "loss": 2.5887, + "step": 11936 + }, + { + "epoch": 0.51, + "grad_norm": 0.4765625, + "learning_rate": 0.0001700985545451674, + "loss": 2.671, + "step": 11937 + }, + { + "epoch": 0.51, + "grad_norm": 0.427734375, + "learning_rate": 0.00017009377045517306, + "loss": 2.5384, + "step": 11938 + }, + { + "epoch": 0.51, + "grad_norm": 0.431640625, + "learning_rate": 0.00017008898604978234, + "loss": 2.5239, + "step": 11939 + }, + { + "epoch": 0.51, + "grad_norm": 0.5546875, + "learning_rate": 0.00017008420132901678, + "loss": 2.6664, + "step": 11940 + }, + { + "epoch": 0.51, + "grad_norm": 0.4375, + "learning_rate": 0.00017007941629289798, + "loss": 2.571, + "step": 11941 + }, + { + "epoch": 0.51, + "grad_norm": 0.42578125, + "learning_rate": 0.00017007463094144737, + "loss": 2.5696, + "step": 11942 + }, + { + "epoch": 0.51, + "grad_norm": 0.419921875, + "learning_rate": 0.00017006984527468655, + "loss": 2.6063, + "step": 11943 + }, + { + "epoch": 0.51, + "grad_norm": 0.41796875, + "learning_rate": 0.00017006505929263702, + "loss": 2.5569, + "step": 11944 + }, + { + "epoch": 0.51, + "grad_norm": 0.431640625, + "learning_rate": 0.00017006027299532034, + "loss": 2.65, + "step": 11945 + }, + { + "epoch": 0.51, + "grad_norm": 0.4375, + "learning_rate": 0.00017005548638275805, + "loss": 2.6657, + "step": 11946 + }, + { + "epoch": 0.51, + "grad_norm": 0.37890625, + "learning_rate": 0.00017005069945497168, + "loss": 2.6032, + "step": 11947 + }, + { + "epoch": 0.51, + "grad_norm": 0.470703125, + "learning_rate": 0.00017004591221198272, + "loss": 2.634, + "step": 11948 + }, + { + "epoch": 0.51, + "grad_norm": 0.466796875, + "learning_rate": 0.00017004112465381278, + "loss": 2.7375, + "step": 11949 + }, + { + "epoch": 0.51, + "grad_norm": 0.447265625, + "learning_rate": 0.0001700363367804834, + "loss": 2.5955, + "step": 11950 + }, + { + "epoch": 0.51, + "grad_norm": 0.40625, + "learning_rate": 0.00017003154859201604, + "loss": 2.6231, + "step": 11951 + }, + { + "epoch": 0.51, + "grad_norm": 0.498046875, + "learning_rate": 0.00017002676008843234, + "loss": 2.5913, + "step": 11952 + }, + { + "epoch": 0.51, + "grad_norm": 0.42578125, + "learning_rate": 0.0001700219712697538, + "loss": 2.7109, + "step": 11953 + }, + { + "epoch": 0.51, + "grad_norm": 0.466796875, + "learning_rate": 0.000170017182136002, + "loss": 2.6018, + "step": 11954 + }, + { + "epoch": 0.51, + "grad_norm": 0.408203125, + "learning_rate": 0.00017001239268719847, + "loss": 2.5619, + "step": 11955 + }, + { + "epoch": 0.51, + "grad_norm": 0.43359375, + "learning_rate": 0.00017000760292336473, + "loss": 2.6017, + "step": 11956 + }, + { + "epoch": 0.51, + "grad_norm": 0.46875, + "learning_rate": 0.0001700028128445224, + "loss": 2.6435, + "step": 11957 + }, + { + "epoch": 0.51, + "grad_norm": 0.51171875, + "learning_rate": 0.00016999802245069295, + "loss": 2.617, + "step": 11958 + }, + { + "epoch": 0.51, + "grad_norm": 0.423828125, + "learning_rate": 0.000169993231741898, + "loss": 2.6784, + "step": 11959 + }, + { + "epoch": 0.51, + "grad_norm": 0.439453125, + "learning_rate": 0.0001699884407181591, + "loss": 2.8181, + "step": 11960 + }, + { + "epoch": 0.51, + "grad_norm": 0.46484375, + "learning_rate": 0.00016998364937949778, + "loss": 2.6686, + "step": 11961 + }, + { + "epoch": 0.51, + "grad_norm": 0.421875, + "learning_rate": 0.00016997885772593562, + "loss": 2.7602, + "step": 11962 + }, + { + "epoch": 0.51, + "grad_norm": 0.412109375, + "learning_rate": 0.00016997406575749419, + "loss": 2.4619, + "step": 11963 + }, + { + "epoch": 0.51, + "grad_norm": 0.43359375, + "learning_rate": 0.00016996927347419502, + "loss": 2.6131, + "step": 11964 + }, + { + "epoch": 0.51, + "grad_norm": 0.50390625, + "learning_rate": 0.0001699644808760597, + "loss": 2.5435, + "step": 11965 + }, + { + "epoch": 0.51, + "grad_norm": 0.4765625, + "learning_rate": 0.00016995968796310978, + "loss": 2.704, + "step": 11966 + }, + { + "epoch": 0.51, + "grad_norm": 0.451171875, + "learning_rate": 0.00016995489473536682, + "loss": 2.6996, + "step": 11967 + }, + { + "epoch": 0.51, + "grad_norm": 0.40625, + "learning_rate": 0.00016995010119285242, + "loss": 2.6216, + "step": 11968 + }, + { + "epoch": 0.51, + "grad_norm": 0.453125, + "learning_rate": 0.00016994530733558812, + "loss": 2.6209, + "step": 11969 + }, + { + "epoch": 0.51, + "grad_norm": 0.43359375, + "learning_rate": 0.00016994051316359548, + "loss": 2.5619, + "step": 11970 + }, + { + "epoch": 0.51, + "grad_norm": 0.4375, + "learning_rate": 0.00016993571867689612, + "loss": 2.5748, + "step": 11971 + }, + { + "epoch": 0.51, + "grad_norm": 0.490234375, + "learning_rate": 0.00016993092387551154, + "loss": 2.6739, + "step": 11972 + }, + { + "epoch": 0.51, + "grad_norm": 0.423828125, + "learning_rate": 0.0001699261287594634, + "loss": 2.6037, + "step": 11973 + }, + { + "epoch": 0.51, + "grad_norm": 0.447265625, + "learning_rate": 0.00016992133332877324, + "loss": 2.4192, + "step": 11974 + }, + { + "epoch": 0.51, + "grad_norm": 0.458984375, + "learning_rate": 0.0001699165375834626, + "loss": 2.5593, + "step": 11975 + }, + { + "epoch": 0.51, + "grad_norm": 0.474609375, + "learning_rate": 0.0001699117415235531, + "loss": 2.6043, + "step": 11976 + }, + { + "epoch": 0.51, + "grad_norm": 0.435546875, + "learning_rate": 0.00016990694514906632, + "loss": 2.4894, + "step": 11977 + }, + { + "epoch": 0.51, + "grad_norm": 0.5, + "learning_rate": 0.00016990214846002388, + "loss": 2.7323, + "step": 11978 + }, + { + "epoch": 0.51, + "grad_norm": 0.439453125, + "learning_rate": 0.00016989735145644724, + "loss": 2.6789, + "step": 11979 + }, + { + "epoch": 0.51, + "grad_norm": 0.408203125, + "learning_rate": 0.00016989255413835813, + "loss": 2.5912, + "step": 11980 + }, + { + "epoch": 0.51, + "grad_norm": 0.4921875, + "learning_rate": 0.00016988775650577802, + "loss": 2.748, + "step": 11981 + }, + { + "epoch": 0.51, + "grad_norm": 0.54296875, + "learning_rate": 0.00016988295855872856, + "loss": 2.6544, + "step": 11982 + }, + { + "epoch": 0.51, + "grad_norm": 0.47265625, + "learning_rate": 0.00016987816029723134, + "loss": 2.6025, + "step": 11983 + }, + { + "epoch": 0.51, + "grad_norm": 0.48828125, + "learning_rate": 0.00016987336172130793, + "loss": 2.558, + "step": 11984 + }, + { + "epoch": 0.51, + "grad_norm": 0.51171875, + "learning_rate": 0.00016986856283097992, + "loss": 2.7432, + "step": 11985 + }, + { + "epoch": 0.51, + "grad_norm": 0.4921875, + "learning_rate": 0.00016986376362626896, + "loss": 2.6933, + "step": 11986 + }, + { + "epoch": 0.51, + "grad_norm": 0.5, + "learning_rate": 0.00016985896410719653, + "loss": 2.6603, + "step": 11987 + }, + { + "epoch": 0.51, + "grad_norm": 0.4375, + "learning_rate": 0.00016985416427378432, + "loss": 2.5533, + "step": 11988 + }, + { + "epoch": 0.51, + "grad_norm": 0.435546875, + "learning_rate": 0.00016984936412605387, + "loss": 2.8173, + "step": 11989 + }, + { + "epoch": 0.51, + "grad_norm": 0.5390625, + "learning_rate": 0.00016984456366402685, + "loss": 2.6299, + "step": 11990 + }, + { + "epoch": 0.51, + "grad_norm": 0.458984375, + "learning_rate": 0.0001698397628877248, + "loss": 2.657, + "step": 11991 + }, + { + "epoch": 0.51, + "grad_norm": 0.4609375, + "learning_rate": 0.00016983496179716933, + "loss": 2.5454, + "step": 11992 + }, + { + "epoch": 0.51, + "grad_norm": 0.4765625, + "learning_rate": 0.00016983016039238212, + "loss": 2.6499, + "step": 11993 + }, + { + "epoch": 0.51, + "grad_norm": 0.458984375, + "learning_rate": 0.00016982535867338463, + "loss": 2.5411, + "step": 11994 + }, + { + "epoch": 0.51, + "grad_norm": 0.52734375, + "learning_rate": 0.00016982055664019856, + "loss": 2.6424, + "step": 11995 + }, + { + "epoch": 0.51, + "grad_norm": 0.447265625, + "learning_rate": 0.0001698157542928455, + "loss": 2.568, + "step": 11996 + }, + { + "epoch": 0.51, + "grad_norm": 0.435546875, + "learning_rate": 0.00016981095163134708, + "loss": 2.6734, + "step": 11997 + }, + { + "epoch": 0.51, + "grad_norm": 0.453125, + "learning_rate": 0.00016980614865572488, + "loss": 2.607, + "step": 11998 + }, + { + "epoch": 0.51, + "grad_norm": 0.435546875, + "learning_rate": 0.00016980134536600052, + "loss": 2.6477, + "step": 11999 + }, + { + "epoch": 0.51, + "grad_norm": 0.408203125, + "learning_rate": 0.00016979654176219562, + "loss": 2.6038, + "step": 12000 + }, + { + "epoch": 0.51, + "grad_norm": 0.52734375, + "learning_rate": 0.0001697917378443318, + "loss": 2.7406, + "step": 12001 + }, + { + "epoch": 0.51, + "grad_norm": 0.462890625, + "learning_rate": 0.00016978693361243064, + "loss": 2.6664, + "step": 12002 + }, + { + "epoch": 0.51, + "grad_norm": 0.49609375, + "learning_rate": 0.00016978212906651382, + "loss": 2.6077, + "step": 12003 + }, + { + "epoch": 0.51, + "grad_norm": 0.47265625, + "learning_rate": 0.0001697773242066029, + "loss": 2.6764, + "step": 12004 + }, + { + "epoch": 0.51, + "grad_norm": 0.412109375, + "learning_rate": 0.0001697725190327195, + "loss": 2.6639, + "step": 12005 + }, + { + "epoch": 0.51, + "grad_norm": 0.427734375, + "learning_rate": 0.0001697677135448853, + "loss": 2.6941, + "step": 12006 + }, + { + "epoch": 0.51, + "grad_norm": 0.4921875, + "learning_rate": 0.00016976290774312186, + "loss": 2.6376, + "step": 12007 + }, + { + "epoch": 0.51, + "grad_norm": 0.484375, + "learning_rate": 0.00016975810162745084, + "loss": 2.6737, + "step": 12008 + }, + { + "epoch": 0.51, + "grad_norm": 0.455078125, + "learning_rate": 0.00016975329519789388, + "loss": 2.6918, + "step": 12009 + }, + { + "epoch": 0.51, + "grad_norm": 0.482421875, + "learning_rate": 0.00016974848845447256, + "loss": 2.413, + "step": 12010 + }, + { + "epoch": 0.51, + "grad_norm": 0.45703125, + "learning_rate": 0.00016974368139720852, + "loss": 2.5149, + "step": 12011 + }, + { + "epoch": 0.51, + "grad_norm": 0.478515625, + "learning_rate": 0.00016973887402612343, + "loss": 2.5197, + "step": 12012 + }, + { + "epoch": 0.51, + "grad_norm": 0.51953125, + "learning_rate": 0.00016973406634123887, + "loss": 2.6999, + "step": 12013 + }, + { + "epoch": 0.51, + "grad_norm": 0.45703125, + "learning_rate": 0.00016972925834257653, + "loss": 2.4933, + "step": 12014 + }, + { + "epoch": 0.51, + "grad_norm": 0.431640625, + "learning_rate": 0.00016972445003015794, + "loss": 2.6034, + "step": 12015 + }, + { + "epoch": 0.51, + "grad_norm": 0.443359375, + "learning_rate": 0.00016971964140400487, + "loss": 2.5547, + "step": 12016 + }, + { + "epoch": 0.51, + "grad_norm": 0.59375, + "learning_rate": 0.0001697148324641389, + "loss": 2.6767, + "step": 12017 + }, + { + "epoch": 0.51, + "grad_norm": 0.45703125, + "learning_rate": 0.00016971002321058162, + "loss": 2.4038, + "step": 12018 + }, + { + "epoch": 0.51, + "grad_norm": 0.46875, + "learning_rate": 0.00016970521364335472, + "loss": 2.6289, + "step": 12019 + }, + { + "epoch": 0.51, + "grad_norm": 0.4765625, + "learning_rate": 0.00016970040376247984, + "loss": 2.6449, + "step": 12020 + }, + { + "epoch": 0.51, + "grad_norm": 0.50390625, + "learning_rate": 0.00016969559356797863, + "loss": 2.7341, + "step": 12021 + }, + { + "epoch": 0.51, + "grad_norm": 0.51171875, + "learning_rate": 0.0001696907830598727, + "loss": 2.573, + "step": 12022 + }, + { + "epoch": 0.51, + "grad_norm": 0.46484375, + "learning_rate": 0.00016968597223818372, + "loss": 2.659, + "step": 12023 + }, + { + "epoch": 0.51, + "grad_norm": 0.4765625, + "learning_rate": 0.00016968116110293333, + "loss": 2.5713, + "step": 12024 + }, + { + "epoch": 0.51, + "grad_norm": 0.46484375, + "learning_rate": 0.00016967634965414318, + "loss": 2.6224, + "step": 12025 + }, + { + "epoch": 0.51, + "grad_norm": 0.443359375, + "learning_rate": 0.00016967153789183495, + "loss": 2.5644, + "step": 12026 + }, + { + "epoch": 0.51, + "grad_norm": 0.40625, + "learning_rate": 0.00016966672581603023, + "loss": 2.5174, + "step": 12027 + }, + { + "epoch": 0.51, + "grad_norm": 0.466796875, + "learning_rate": 0.0001696619134267507, + "loss": 2.7108, + "step": 12028 + }, + { + "epoch": 0.51, + "grad_norm": 0.421875, + "learning_rate": 0.00016965710072401803, + "loss": 2.6792, + "step": 12029 + }, + { + "epoch": 0.51, + "grad_norm": 0.455078125, + "learning_rate": 0.00016965228770785386, + "loss": 2.5636, + "step": 12030 + }, + { + "epoch": 0.51, + "grad_norm": 0.490234375, + "learning_rate": 0.00016964747437827986, + "loss": 2.5529, + "step": 12031 + }, + { + "epoch": 0.51, + "grad_norm": 0.50390625, + "learning_rate": 0.00016964266073531768, + "loss": 2.6804, + "step": 12032 + }, + { + "epoch": 0.51, + "grad_norm": 0.431640625, + "learning_rate": 0.00016963784677898898, + "loss": 2.6588, + "step": 12033 + }, + { + "epoch": 0.51, + "grad_norm": 0.466796875, + "learning_rate": 0.0001696330325093154, + "loss": 2.6403, + "step": 12034 + }, + { + "epoch": 0.51, + "grad_norm": 0.486328125, + "learning_rate": 0.00016962821792631863, + "loss": 2.5129, + "step": 12035 + }, + { + "epoch": 0.51, + "grad_norm": 0.439453125, + "learning_rate": 0.0001696234030300203, + "loss": 2.5167, + "step": 12036 + }, + { + "epoch": 0.51, + "grad_norm": 0.439453125, + "learning_rate": 0.00016961858782044216, + "loss": 2.6662, + "step": 12037 + }, + { + "epoch": 0.51, + "grad_norm": 0.484375, + "learning_rate": 0.00016961377229760577, + "loss": 2.5361, + "step": 12038 + }, + { + "epoch": 0.51, + "grad_norm": 0.451171875, + "learning_rate": 0.00016960895646153285, + "loss": 2.6677, + "step": 12039 + }, + { + "epoch": 0.51, + "grad_norm": 0.494140625, + "learning_rate": 0.00016960414031224507, + "loss": 2.6036, + "step": 12040 + }, + { + "epoch": 0.51, + "grad_norm": 0.50390625, + "learning_rate": 0.00016959932384976413, + "loss": 2.7685, + "step": 12041 + }, + { + "epoch": 0.51, + "grad_norm": 0.4453125, + "learning_rate": 0.0001695945070741116, + "loss": 2.6237, + "step": 12042 + }, + { + "epoch": 0.51, + "grad_norm": 0.5234375, + "learning_rate": 0.00016958968998530928, + "loss": 2.6111, + "step": 12043 + }, + { + "epoch": 0.51, + "grad_norm": 0.48828125, + "learning_rate": 0.00016958487258337876, + "loss": 2.5916, + "step": 12044 + }, + { + "epoch": 0.51, + "grad_norm": 0.498046875, + "learning_rate": 0.00016958005486834176, + "loss": 2.5972, + "step": 12045 + }, + { + "epoch": 0.51, + "grad_norm": 0.494140625, + "learning_rate": 0.00016957523684021995, + "loss": 2.6112, + "step": 12046 + }, + { + "epoch": 0.51, + "grad_norm": 0.4765625, + "learning_rate": 0.00016957041849903498, + "loss": 2.5479, + "step": 12047 + }, + { + "epoch": 0.51, + "grad_norm": 0.453125, + "learning_rate": 0.00016956559984480856, + "loss": 2.6973, + "step": 12048 + }, + { + "epoch": 0.51, + "grad_norm": 0.4609375, + "learning_rate": 0.00016956078087756237, + "loss": 2.5438, + "step": 12049 + }, + { + "epoch": 0.51, + "grad_norm": 0.443359375, + "learning_rate": 0.00016955596159731805, + "loss": 2.6236, + "step": 12050 + }, + { + "epoch": 0.51, + "grad_norm": 0.44140625, + "learning_rate": 0.00016955114200409738, + "loss": 2.538, + "step": 12051 + }, + { + "epoch": 0.51, + "grad_norm": 0.466796875, + "learning_rate": 0.00016954632209792195, + "loss": 2.5018, + "step": 12052 + }, + { + "epoch": 0.51, + "grad_norm": 0.490234375, + "learning_rate": 0.00016954150187881351, + "loss": 2.6867, + "step": 12053 + }, + { + "epoch": 0.51, + "grad_norm": 0.458984375, + "learning_rate": 0.0001695366813467937, + "loss": 2.5209, + "step": 12054 + }, + { + "epoch": 0.51, + "grad_norm": 0.435546875, + "learning_rate": 0.00016953186050188427, + "loss": 2.7453, + "step": 12055 + }, + { + "epoch": 0.51, + "grad_norm": 0.443359375, + "learning_rate": 0.00016952703934410685, + "loss": 2.6706, + "step": 12056 + }, + { + "epoch": 0.51, + "grad_norm": 0.55078125, + "learning_rate": 0.00016952221787348316, + "loss": 2.5564, + "step": 12057 + }, + { + "epoch": 0.51, + "grad_norm": 0.41796875, + "learning_rate": 0.00016951739609003493, + "loss": 2.5732, + "step": 12058 + }, + { + "epoch": 0.51, + "grad_norm": 0.466796875, + "learning_rate": 0.00016951257399378376, + "loss": 2.645, + "step": 12059 + }, + { + "epoch": 0.51, + "grad_norm": 0.451171875, + "learning_rate": 0.00016950775158475147, + "loss": 2.5567, + "step": 12060 + }, + { + "epoch": 0.51, + "grad_norm": 0.46484375, + "learning_rate": 0.0001695029288629597, + "loss": 2.7283, + "step": 12061 + }, + { + "epoch": 0.51, + "grad_norm": 0.4765625, + "learning_rate": 0.0001694981058284301, + "loss": 2.4129, + "step": 12062 + }, + { + "epoch": 0.51, + "grad_norm": 0.5234375, + "learning_rate": 0.0001694932824811845, + "loss": 2.6348, + "step": 12063 + }, + { + "epoch": 0.51, + "grad_norm": 0.46875, + "learning_rate": 0.00016948845882124445, + "loss": 2.6255, + "step": 12064 + }, + { + "epoch": 0.51, + "grad_norm": 0.47265625, + "learning_rate": 0.00016948363484863176, + "loss": 2.644, + "step": 12065 + }, + { + "epoch": 0.51, + "grad_norm": 0.5078125, + "learning_rate": 0.00016947881056336809, + "loss": 2.5618, + "step": 12066 + }, + { + "epoch": 0.51, + "grad_norm": 0.427734375, + "learning_rate": 0.0001694739859654752, + "loss": 2.6449, + "step": 12067 + }, + { + "epoch": 0.51, + "grad_norm": 0.4453125, + "learning_rate": 0.00016946916105497475, + "loss": 2.6082, + "step": 12068 + }, + { + "epoch": 0.51, + "grad_norm": 0.4609375, + "learning_rate": 0.00016946433583188845, + "loss": 2.4933, + "step": 12069 + }, + { + "epoch": 0.51, + "grad_norm": 0.4453125, + "learning_rate": 0.00016945951029623802, + "loss": 2.6975, + "step": 12070 + }, + { + "epoch": 0.51, + "grad_norm": 0.421875, + "learning_rate": 0.00016945468444804523, + "loss": 2.5854, + "step": 12071 + }, + { + "epoch": 0.51, + "grad_norm": 0.412109375, + "learning_rate": 0.00016944985828733168, + "loss": 2.5238, + "step": 12072 + }, + { + "epoch": 0.51, + "grad_norm": 0.416015625, + "learning_rate": 0.00016944503181411918, + "loss": 2.5027, + "step": 12073 + }, + { + "epoch": 0.51, + "grad_norm": 0.435546875, + "learning_rate": 0.0001694402050284294, + "loss": 2.6776, + "step": 12074 + }, + { + "epoch": 0.51, + "grad_norm": 0.45703125, + "learning_rate": 0.0001694353779302841, + "loss": 2.5348, + "step": 12075 + }, + { + "epoch": 0.51, + "grad_norm": 0.4375, + "learning_rate": 0.00016943055051970495, + "loss": 2.5822, + "step": 12076 + }, + { + "epoch": 0.51, + "grad_norm": 0.455078125, + "learning_rate": 0.00016942572279671371, + "loss": 2.4709, + "step": 12077 + }, + { + "epoch": 0.51, + "grad_norm": 0.455078125, + "learning_rate": 0.0001694208947613321, + "loss": 2.5963, + "step": 12078 + }, + { + "epoch": 0.51, + "grad_norm": 0.515625, + "learning_rate": 0.0001694160664135818, + "loss": 2.6347, + "step": 12079 + }, + { + "epoch": 0.51, + "grad_norm": 0.455078125, + "learning_rate": 0.0001694112377534846, + "loss": 2.5615, + "step": 12080 + }, + { + "epoch": 0.51, + "grad_norm": 0.53125, + "learning_rate": 0.00016940640878106218, + "loss": 2.64, + "step": 12081 + }, + { + "epoch": 0.51, + "grad_norm": 0.451171875, + "learning_rate": 0.0001694015794963363, + "loss": 2.5401, + "step": 12082 + }, + { + "epoch": 0.51, + "grad_norm": 0.439453125, + "learning_rate": 0.00016939674989932866, + "loss": 2.7094, + "step": 12083 + }, + { + "epoch": 0.51, + "grad_norm": 0.494140625, + "learning_rate": 0.000169391919990061, + "loss": 2.6435, + "step": 12084 + }, + { + "epoch": 0.51, + "grad_norm": 0.443359375, + "learning_rate": 0.0001693870897685551, + "loss": 2.6248, + "step": 12085 + }, + { + "epoch": 0.52, + "grad_norm": 0.388671875, + "learning_rate": 0.0001693822592348326, + "loss": 2.5011, + "step": 12086 + }, + { + "epoch": 0.52, + "grad_norm": 0.4609375, + "learning_rate": 0.00016937742838891534, + "loss": 2.5727, + "step": 12087 + }, + { + "epoch": 0.52, + "grad_norm": 0.515625, + "learning_rate": 0.00016937259723082498, + "loss": 2.6456, + "step": 12088 + }, + { + "epoch": 0.52, + "grad_norm": 0.435546875, + "learning_rate": 0.0001693677657605833, + "loss": 2.561, + "step": 12089 + }, + { + "epoch": 0.52, + "grad_norm": 0.41796875, + "learning_rate": 0.00016936293397821199, + "loss": 2.538, + "step": 12090 + }, + { + "epoch": 0.52, + "grad_norm": 0.484375, + "learning_rate": 0.00016935810188373286, + "loss": 2.7405, + "step": 12091 + }, + { + "epoch": 0.52, + "grad_norm": 0.4453125, + "learning_rate": 0.00016935326947716762, + "loss": 2.546, + "step": 12092 + }, + { + "epoch": 0.52, + "grad_norm": 0.45703125, + "learning_rate": 0.00016934843675853802, + "loss": 2.54, + "step": 12093 + }, + { + "epoch": 0.52, + "grad_norm": 0.447265625, + "learning_rate": 0.00016934360372786576, + "loss": 2.597, + "step": 12094 + }, + { + "epoch": 0.52, + "grad_norm": 0.50390625, + "learning_rate": 0.00016933877038517267, + "loss": 2.6015, + "step": 12095 + }, + { + "epoch": 0.52, + "grad_norm": 0.4453125, + "learning_rate": 0.0001693339367304804, + "loss": 2.5323, + "step": 12096 + }, + { + "epoch": 0.52, + "grad_norm": 0.486328125, + "learning_rate": 0.0001693291027638108, + "loss": 2.788, + "step": 12097 + }, + { + "epoch": 0.52, + "grad_norm": 0.435546875, + "learning_rate": 0.00016932426848518555, + "loss": 2.6327, + "step": 12098 + }, + { + "epoch": 0.52, + "grad_norm": 0.46484375, + "learning_rate": 0.00016931943389462647, + "loss": 2.7117, + "step": 12099 + }, + { + "epoch": 0.52, + "grad_norm": 0.46484375, + "learning_rate": 0.00016931459899215522, + "loss": 2.4708, + "step": 12100 + }, + { + "epoch": 0.52, + "grad_norm": 0.44921875, + "learning_rate": 0.00016930976377779362, + "loss": 2.4355, + "step": 12101 + }, + { + "epoch": 0.52, + "grad_norm": 0.42578125, + "learning_rate": 0.0001693049282515634, + "loss": 2.826, + "step": 12102 + }, + { + "epoch": 0.52, + "grad_norm": 0.421875, + "learning_rate": 0.00016930009241348636, + "loss": 2.6698, + "step": 12103 + }, + { + "epoch": 0.52, + "grad_norm": 0.439453125, + "learning_rate": 0.0001692952562635842, + "loss": 2.4441, + "step": 12104 + }, + { + "epoch": 0.52, + "grad_norm": 0.447265625, + "learning_rate": 0.00016929041980187875, + "loss": 2.5304, + "step": 12105 + }, + { + "epoch": 0.52, + "grad_norm": 0.498046875, + "learning_rate": 0.0001692855830283917, + "loss": 2.5871, + "step": 12106 + }, + { + "epoch": 0.52, + "grad_norm": 0.486328125, + "learning_rate": 0.00016928074594314488, + "loss": 2.5636, + "step": 12107 + }, + { + "epoch": 0.52, + "grad_norm": 0.4765625, + "learning_rate": 0.00016927590854615997, + "loss": 2.5364, + "step": 12108 + }, + { + "epoch": 0.52, + "grad_norm": 0.515625, + "learning_rate": 0.00016927107083745884, + "loss": 2.7513, + "step": 12109 + }, + { + "epoch": 0.52, + "grad_norm": 0.466796875, + "learning_rate": 0.0001692662328170632, + "loss": 2.6998, + "step": 12110 + }, + { + "epoch": 0.52, + "grad_norm": 0.4453125, + "learning_rate": 0.0001692613944849948, + "loss": 2.6138, + "step": 12111 + }, + { + "epoch": 0.52, + "grad_norm": 0.52734375, + "learning_rate": 0.00016925655584127547, + "loss": 2.7622, + "step": 12112 + }, + { + "epoch": 0.52, + "grad_norm": 0.51171875, + "learning_rate": 0.0001692517168859269, + "loss": 2.6597, + "step": 12113 + }, + { + "epoch": 0.52, + "grad_norm": 0.419921875, + "learning_rate": 0.00016924687761897097, + "loss": 2.6259, + "step": 12114 + }, + { + "epoch": 0.52, + "grad_norm": 0.52734375, + "learning_rate": 0.00016924203804042937, + "loss": 2.6303, + "step": 12115 + }, + { + "epoch": 0.52, + "grad_norm": 0.455078125, + "learning_rate": 0.0001692371981503239, + "loss": 2.6267, + "step": 12116 + }, + { + "epoch": 0.52, + "grad_norm": 0.408203125, + "learning_rate": 0.00016923235794867635, + "loss": 2.6119, + "step": 12117 + }, + { + "epoch": 0.52, + "grad_norm": 0.484375, + "learning_rate": 0.0001692275174355085, + "loss": 2.6965, + "step": 12118 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.0001692226766108421, + "loss": 2.5869, + "step": 12119 + }, + { + "epoch": 0.52, + "grad_norm": 0.447265625, + "learning_rate": 0.00016921783547469893, + "loss": 2.6063, + "step": 12120 + }, + { + "epoch": 0.52, + "grad_norm": 0.431640625, + "learning_rate": 0.00016921299402710086, + "loss": 2.6079, + "step": 12121 + }, + { + "epoch": 0.52, + "grad_norm": 0.4140625, + "learning_rate": 0.00016920815226806956, + "loss": 2.6494, + "step": 12122 + }, + { + "epoch": 0.52, + "grad_norm": 0.5625, + "learning_rate": 0.0001692033101976269, + "loss": 2.5618, + "step": 12123 + }, + { + "epoch": 0.52, + "grad_norm": 0.515625, + "learning_rate": 0.0001691984678157946, + "loss": 2.6432, + "step": 12124 + }, + { + "epoch": 0.52, + "grad_norm": 0.44140625, + "learning_rate": 0.00016919362512259452, + "loss": 2.6373, + "step": 12125 + }, + { + "epoch": 0.52, + "grad_norm": 0.470703125, + "learning_rate": 0.0001691887821180484, + "loss": 2.6298, + "step": 12126 + }, + { + "epoch": 0.52, + "grad_norm": 0.455078125, + "learning_rate": 0.00016918393880217806, + "loss": 2.6593, + "step": 12127 + }, + { + "epoch": 0.52, + "grad_norm": 0.486328125, + "learning_rate": 0.00016917909517500526, + "loss": 2.6335, + "step": 12128 + }, + { + "epoch": 0.52, + "grad_norm": 0.42578125, + "learning_rate": 0.00016917425123655182, + "loss": 2.5418, + "step": 12129 + }, + { + "epoch": 0.52, + "grad_norm": 0.4375, + "learning_rate": 0.0001691694069868395, + "loss": 2.608, + "step": 12130 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.00016916456242589017, + "loss": 2.7699, + "step": 12131 + }, + { + "epoch": 0.52, + "grad_norm": 0.45703125, + "learning_rate": 0.00016915971755372558, + "loss": 2.5451, + "step": 12132 + }, + { + "epoch": 0.52, + "grad_norm": 0.4921875, + "learning_rate": 0.00016915487237036754, + "loss": 2.6737, + "step": 12133 + }, + { + "epoch": 0.52, + "grad_norm": 0.392578125, + "learning_rate": 0.00016915002687583784, + "loss": 2.4895, + "step": 12134 + }, + { + "epoch": 0.52, + "grad_norm": 0.51171875, + "learning_rate": 0.00016914518107015826, + "loss": 2.5858, + "step": 12135 + }, + { + "epoch": 0.52, + "grad_norm": 0.5, + "learning_rate": 0.0001691403349533507, + "loss": 2.6615, + "step": 12136 + }, + { + "epoch": 0.52, + "grad_norm": 0.484375, + "learning_rate": 0.00016913548852543682, + "loss": 2.7143, + "step": 12137 + }, + { + "epoch": 0.52, + "grad_norm": 0.4609375, + "learning_rate": 0.00016913064178643858, + "loss": 2.6708, + "step": 12138 + }, + { + "epoch": 0.52, + "grad_norm": 0.462890625, + "learning_rate": 0.00016912579473637765, + "loss": 2.7346, + "step": 12139 + }, + { + "epoch": 0.52, + "grad_norm": 0.453125, + "learning_rate": 0.00016912094737527595, + "loss": 2.4523, + "step": 12140 + }, + { + "epoch": 0.52, + "grad_norm": 0.484375, + "learning_rate": 0.00016911609970315526, + "loss": 2.5152, + "step": 12141 + }, + { + "epoch": 0.52, + "grad_norm": 0.435546875, + "learning_rate": 0.00016911125172003737, + "loss": 2.6957, + "step": 12142 + }, + { + "epoch": 0.52, + "grad_norm": 0.439453125, + "learning_rate": 0.00016910640342594405, + "loss": 2.6067, + "step": 12143 + }, + { + "epoch": 0.52, + "grad_norm": 0.46875, + "learning_rate": 0.00016910155482089724, + "loss": 2.5114, + "step": 12144 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.00016909670590491867, + "loss": 2.5268, + "step": 12145 + }, + { + "epoch": 0.52, + "grad_norm": 0.412109375, + "learning_rate": 0.00016909185667803015, + "loss": 2.4887, + "step": 12146 + }, + { + "epoch": 0.52, + "grad_norm": 0.42578125, + "learning_rate": 0.00016908700714025353, + "loss": 2.5728, + "step": 12147 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.00016908215729161065, + "loss": 2.5398, + "step": 12148 + }, + { + "epoch": 0.52, + "grad_norm": 0.50390625, + "learning_rate": 0.00016907730713212332, + "loss": 2.6035, + "step": 12149 + }, + { + "epoch": 0.52, + "grad_norm": 0.431640625, + "learning_rate": 0.00016907245666181333, + "loss": 2.6178, + "step": 12150 + }, + { + "epoch": 0.52, + "grad_norm": 0.5078125, + "learning_rate": 0.0001690676058807025, + "loss": 2.6616, + "step": 12151 + }, + { + "epoch": 0.52, + "grad_norm": 0.455078125, + "learning_rate": 0.00016906275478881274, + "loss": 2.5543, + "step": 12152 + }, + { + "epoch": 0.52, + "grad_norm": 0.43359375, + "learning_rate": 0.00016905790338616575, + "loss": 2.6635, + "step": 12153 + }, + { + "epoch": 0.52, + "grad_norm": 0.4765625, + "learning_rate": 0.00016905305167278352, + "loss": 2.7407, + "step": 12154 + }, + { + "epoch": 0.52, + "grad_norm": 0.4375, + "learning_rate": 0.00016904819964868773, + "loss": 2.5187, + "step": 12155 + }, + { + "epoch": 0.52, + "grad_norm": 0.462890625, + "learning_rate": 0.00016904334731390034, + "loss": 2.6014, + "step": 12156 + }, + { + "epoch": 0.52, + "grad_norm": 0.5078125, + "learning_rate": 0.00016903849466844302, + "loss": 2.6307, + "step": 12157 + }, + { + "epoch": 0.52, + "grad_norm": 0.53125, + "learning_rate": 0.0001690336417123378, + "loss": 2.7849, + "step": 12158 + }, + { + "epoch": 0.52, + "grad_norm": 0.490234375, + "learning_rate": 0.00016902878844560638, + "loss": 2.616, + "step": 12159 + }, + { + "epoch": 0.52, + "grad_norm": 0.48046875, + "learning_rate": 0.00016902393486827063, + "loss": 2.5967, + "step": 12160 + }, + { + "epoch": 0.52, + "grad_norm": 0.478515625, + "learning_rate": 0.00016901908098035243, + "loss": 2.4656, + "step": 12161 + }, + { + "epoch": 0.52, + "grad_norm": 0.474609375, + "learning_rate": 0.00016901422678187355, + "loss": 2.5897, + "step": 12162 + }, + { + "epoch": 0.52, + "grad_norm": 0.443359375, + "learning_rate": 0.0001690093722728559, + "loss": 2.495, + "step": 12163 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.0001690045174533213, + "loss": 2.5881, + "step": 12164 + }, + { + "epoch": 0.52, + "grad_norm": 0.416015625, + "learning_rate": 0.00016899966232329153, + "loss": 2.5157, + "step": 12165 + }, + { + "epoch": 0.52, + "grad_norm": 0.4453125, + "learning_rate": 0.00016899480688278856, + "loss": 2.4859, + "step": 12166 + }, + { + "epoch": 0.52, + "grad_norm": 0.40625, + "learning_rate": 0.00016898995113183416, + "loss": 2.5563, + "step": 12167 + }, + { + "epoch": 0.52, + "grad_norm": 0.421875, + "learning_rate": 0.0001689850950704502, + "loss": 2.5263, + "step": 12168 + }, + { + "epoch": 0.52, + "grad_norm": 0.48046875, + "learning_rate": 0.0001689802386986585, + "loss": 2.6583, + "step": 12169 + }, + { + "epoch": 0.52, + "grad_norm": 0.48046875, + "learning_rate": 0.00016897538201648095, + "loss": 2.6095, + "step": 12170 + }, + { + "epoch": 0.52, + "grad_norm": 0.447265625, + "learning_rate": 0.00016897052502393937, + "loss": 2.6134, + "step": 12171 + }, + { + "epoch": 0.52, + "grad_norm": 0.466796875, + "learning_rate": 0.00016896566772105567, + "loss": 2.6781, + "step": 12172 + }, + { + "epoch": 0.52, + "grad_norm": 0.466796875, + "learning_rate": 0.00016896081010785166, + "loss": 2.5569, + "step": 12173 + }, + { + "epoch": 0.52, + "grad_norm": 0.392578125, + "learning_rate": 0.00016895595218434917, + "loss": 2.5817, + "step": 12174 + }, + { + "epoch": 0.52, + "grad_norm": 0.40234375, + "learning_rate": 0.00016895109395057012, + "loss": 2.4395, + "step": 12175 + }, + { + "epoch": 0.52, + "grad_norm": 0.453125, + "learning_rate": 0.00016894623540653636, + "loss": 2.5482, + "step": 12176 + }, + { + "epoch": 0.52, + "grad_norm": 0.431640625, + "learning_rate": 0.0001689413765522697, + "loss": 2.6505, + "step": 12177 + }, + { + "epoch": 0.52, + "grad_norm": 0.421875, + "learning_rate": 0.0001689365173877921, + "loss": 2.6681, + "step": 12178 + }, + { + "epoch": 0.52, + "grad_norm": 0.51953125, + "learning_rate": 0.0001689316579131253, + "loss": 2.6384, + "step": 12179 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.00016892679812829128, + "loss": 2.5996, + "step": 12180 + }, + { + "epoch": 0.52, + "grad_norm": 0.43359375, + "learning_rate": 0.00016892193803331185, + "loss": 2.6755, + "step": 12181 + }, + { + "epoch": 0.52, + "grad_norm": 0.4375, + "learning_rate": 0.00016891707762820888, + "loss": 2.584, + "step": 12182 + }, + { + "epoch": 0.52, + "grad_norm": 0.484375, + "learning_rate": 0.00016891221691300423, + "loss": 2.5898, + "step": 12183 + }, + { + "epoch": 0.52, + "grad_norm": 0.470703125, + "learning_rate": 0.0001689073558877198, + "loss": 2.7093, + "step": 12184 + }, + { + "epoch": 0.52, + "grad_norm": 0.55078125, + "learning_rate": 0.00016890249455237742, + "loss": 2.769, + "step": 12185 + }, + { + "epoch": 0.52, + "grad_norm": 0.451171875, + "learning_rate": 0.00016889763290699907, + "loss": 2.5438, + "step": 12186 + }, + { + "epoch": 0.52, + "grad_norm": 0.53125, + "learning_rate": 0.0001688927709516065, + "loss": 2.6856, + "step": 12187 + }, + { + "epoch": 0.52, + "grad_norm": 0.44140625, + "learning_rate": 0.00016888790868622164, + "loss": 2.5407, + "step": 12188 + }, + { + "epoch": 0.52, + "grad_norm": 0.4375, + "learning_rate": 0.00016888304611086636, + "loss": 2.6205, + "step": 12189 + }, + { + "epoch": 0.52, + "grad_norm": 0.421875, + "learning_rate": 0.0001688781832255626, + "loss": 2.4539, + "step": 12190 + }, + { + "epoch": 0.52, + "grad_norm": 0.42578125, + "learning_rate": 0.00016887332003033212, + "loss": 2.6337, + "step": 12191 + }, + { + "epoch": 0.52, + "grad_norm": 0.46875, + "learning_rate": 0.00016886845652519693, + "loss": 2.6885, + "step": 12192 + }, + { + "epoch": 0.52, + "grad_norm": 0.5, + "learning_rate": 0.0001688635927101788, + "loss": 2.6698, + "step": 12193 + }, + { + "epoch": 0.52, + "grad_norm": 0.4453125, + "learning_rate": 0.0001688587285852997, + "loss": 2.6245, + "step": 12194 + }, + { + "epoch": 0.52, + "grad_norm": 0.46875, + "learning_rate": 0.00016885386415058147, + "loss": 2.5187, + "step": 12195 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.00016884899940604605, + "loss": 2.6887, + "step": 12196 + }, + { + "epoch": 0.52, + "grad_norm": 0.439453125, + "learning_rate": 0.00016884413435171522, + "loss": 2.5591, + "step": 12197 + }, + { + "epoch": 0.52, + "grad_norm": 0.4453125, + "learning_rate": 0.00016883926898761102, + "loss": 2.7184, + "step": 12198 + }, + { + "epoch": 0.52, + "grad_norm": 0.478515625, + "learning_rate": 0.00016883440331375523, + "loss": 2.535, + "step": 12199 + }, + { + "epoch": 0.52, + "grad_norm": 0.478515625, + "learning_rate": 0.00016882953733016979, + "loss": 2.7015, + "step": 12200 + }, + { + "epoch": 0.52, + "grad_norm": 0.53125, + "learning_rate": 0.00016882467103687657, + "loss": 2.5382, + "step": 12201 + }, + { + "epoch": 0.52, + "grad_norm": 0.4609375, + "learning_rate": 0.00016881980443389752, + "loss": 2.5705, + "step": 12202 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.0001688149375212545, + "loss": 2.6382, + "step": 12203 + }, + { + "epoch": 0.52, + "grad_norm": 0.5078125, + "learning_rate": 0.00016881007029896937, + "loss": 2.5697, + "step": 12204 + }, + { + "epoch": 0.52, + "grad_norm": 0.384765625, + "learning_rate": 0.00016880520276706408, + "loss": 2.5233, + "step": 12205 + }, + { + "epoch": 0.52, + "grad_norm": 0.43359375, + "learning_rate": 0.00016880033492556055, + "loss": 2.7351, + "step": 12206 + }, + { + "epoch": 0.52, + "grad_norm": 0.490234375, + "learning_rate": 0.00016879546677448063, + "loss": 2.4513, + "step": 12207 + }, + { + "epoch": 0.52, + "grad_norm": 0.45703125, + "learning_rate": 0.00016879059831384626, + "loss": 2.7283, + "step": 12208 + }, + { + "epoch": 0.52, + "grad_norm": 0.435546875, + "learning_rate": 0.00016878572954367934, + "loss": 2.5588, + "step": 12209 + }, + { + "epoch": 0.52, + "grad_norm": 0.474609375, + "learning_rate": 0.0001687808604640018, + "loss": 2.5811, + "step": 12210 + }, + { + "epoch": 0.52, + "grad_norm": 0.453125, + "learning_rate": 0.00016877599107483548, + "loss": 2.5188, + "step": 12211 + }, + { + "epoch": 0.52, + "grad_norm": 0.4921875, + "learning_rate": 0.00016877112137620237, + "loss": 2.599, + "step": 12212 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.00016876625136812433, + "loss": 2.6862, + "step": 12213 + }, + { + "epoch": 0.52, + "grad_norm": 0.466796875, + "learning_rate": 0.00016876138105062328, + "loss": 2.5928, + "step": 12214 + }, + { + "epoch": 0.52, + "grad_norm": 0.474609375, + "learning_rate": 0.00016875651042372116, + "loss": 2.5631, + "step": 12215 + }, + { + "epoch": 0.52, + "grad_norm": 0.404296875, + "learning_rate": 0.00016875163948743986, + "loss": 2.5705, + "step": 12216 + }, + { + "epoch": 0.52, + "grad_norm": 0.466796875, + "learning_rate": 0.0001687467682418013, + "loss": 2.5865, + "step": 12217 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.00016874189668682744, + "loss": 2.6773, + "step": 12218 + }, + { + "epoch": 0.52, + "grad_norm": 0.4140625, + "learning_rate": 0.00016873702482254018, + "loss": 2.675, + "step": 12219 + }, + { + "epoch": 0.52, + "grad_norm": 0.451171875, + "learning_rate": 0.00016873215264896137, + "loss": 2.4332, + "step": 12220 + }, + { + "epoch": 0.52, + "grad_norm": 0.52734375, + "learning_rate": 0.00016872728016611302, + "loss": 2.6221, + "step": 12221 + }, + { + "epoch": 0.52, + "grad_norm": 0.4921875, + "learning_rate": 0.00016872240737401704, + "loss": 2.5842, + "step": 12222 + }, + { + "epoch": 0.52, + "grad_norm": 0.439453125, + "learning_rate": 0.00016871753427269532, + "loss": 2.5853, + "step": 12223 + }, + { + "epoch": 0.52, + "grad_norm": 0.490234375, + "learning_rate": 0.00016871266086216982, + "loss": 2.6861, + "step": 12224 + }, + { + "epoch": 0.52, + "grad_norm": 0.4921875, + "learning_rate": 0.00016870778714246244, + "loss": 2.6153, + "step": 12225 + }, + { + "epoch": 0.52, + "grad_norm": 0.40625, + "learning_rate": 0.00016870291311359516, + "loss": 2.5253, + "step": 12226 + }, + { + "epoch": 0.52, + "grad_norm": 0.44140625, + "learning_rate": 0.00016869803877558987, + "loss": 2.6421, + "step": 12227 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.0001686931641284685, + "loss": 2.5259, + "step": 12228 + }, + { + "epoch": 0.52, + "grad_norm": 0.439453125, + "learning_rate": 0.000168688289172253, + "loss": 2.5167, + "step": 12229 + }, + { + "epoch": 0.52, + "grad_norm": 0.447265625, + "learning_rate": 0.0001686834139069653, + "loss": 2.7297, + "step": 12230 + }, + { + "epoch": 0.52, + "grad_norm": 0.404296875, + "learning_rate": 0.00016867853833262735, + "loss": 2.5933, + "step": 12231 + }, + { + "epoch": 0.52, + "grad_norm": 0.4453125, + "learning_rate": 0.00016867366244926104, + "loss": 2.5946, + "step": 12232 + }, + { + "epoch": 0.52, + "grad_norm": 0.4765625, + "learning_rate": 0.0001686687862568884, + "loss": 2.4226, + "step": 12233 + }, + { + "epoch": 0.52, + "grad_norm": 0.4609375, + "learning_rate": 0.0001686639097555313, + "loss": 2.7114, + "step": 12234 + }, + { + "epoch": 0.52, + "grad_norm": 0.50390625, + "learning_rate": 0.00016865903294521165, + "loss": 2.5304, + "step": 12235 + }, + { + "epoch": 0.52, + "grad_norm": 0.5234375, + "learning_rate": 0.0001686541558259515, + "loss": 2.7103, + "step": 12236 + }, + { + "epoch": 0.52, + "grad_norm": 0.53125, + "learning_rate": 0.0001686492783977727, + "loss": 2.6211, + "step": 12237 + }, + { + "epoch": 0.52, + "grad_norm": 0.4921875, + "learning_rate": 0.00016864440066069727, + "loss": 2.4983, + "step": 12238 + }, + { + "epoch": 0.52, + "grad_norm": 0.49609375, + "learning_rate": 0.0001686395226147471, + "loss": 2.5664, + "step": 12239 + }, + { + "epoch": 0.52, + "grad_norm": 0.470703125, + "learning_rate": 0.0001686346442599442, + "loss": 2.4359, + "step": 12240 + }, + { + "epoch": 0.52, + "grad_norm": 0.4609375, + "learning_rate": 0.00016862976559631044, + "loss": 2.5152, + "step": 12241 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.00016862488662386786, + "loss": 2.5369, + "step": 12242 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.00016862000734263835, + "loss": 2.6235, + "step": 12243 + }, + { + "epoch": 0.52, + "grad_norm": 0.484375, + "learning_rate": 0.00016861512775264386, + "loss": 2.5514, + "step": 12244 + }, + { + "epoch": 0.52, + "grad_norm": 0.490234375, + "learning_rate": 0.0001686102478539064, + "loss": 2.7067, + "step": 12245 + }, + { + "epoch": 0.52, + "grad_norm": 0.453125, + "learning_rate": 0.0001686053676464479, + "loss": 2.5866, + "step": 12246 + }, + { + "epoch": 0.52, + "grad_norm": 0.431640625, + "learning_rate": 0.0001686004871302903, + "loss": 2.7043, + "step": 12247 + }, + { + "epoch": 0.52, + "grad_norm": 0.4375, + "learning_rate": 0.00016859560630545558, + "loss": 2.5711, + "step": 12248 + }, + { + "epoch": 0.52, + "grad_norm": 0.43359375, + "learning_rate": 0.0001685907251719657, + "loss": 2.6544, + "step": 12249 + }, + { + "epoch": 0.52, + "grad_norm": 0.416015625, + "learning_rate": 0.00016858584372984262, + "loss": 2.6114, + "step": 12250 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.00016858096197910834, + "loss": 2.6194, + "step": 12251 + }, + { + "epoch": 0.52, + "grad_norm": 0.4765625, + "learning_rate": 0.00016857607991978477, + "loss": 2.5545, + "step": 12252 + }, + { + "epoch": 0.52, + "grad_norm": 0.5390625, + "learning_rate": 0.0001685711975518939, + "loss": 2.6858, + "step": 12253 + }, + { + "epoch": 0.52, + "grad_norm": 0.52734375, + "learning_rate": 0.0001685663148754577, + "loss": 2.6129, + "step": 12254 + }, + { + "epoch": 0.52, + "grad_norm": 0.453125, + "learning_rate": 0.00016856143189049815, + "loss": 2.6096, + "step": 12255 + }, + { + "epoch": 0.52, + "grad_norm": 0.431640625, + "learning_rate": 0.0001685565485970372, + "loss": 2.6466, + "step": 12256 + }, + { + "epoch": 0.52, + "grad_norm": 0.48828125, + "learning_rate": 0.00016855166499509688, + "loss": 2.6605, + "step": 12257 + }, + { + "epoch": 0.52, + "grad_norm": 0.44921875, + "learning_rate": 0.00016854678108469906, + "loss": 2.6329, + "step": 12258 + }, + { + "epoch": 0.52, + "grad_norm": 0.546875, + "learning_rate": 0.0001685418968658658, + "loss": 2.6346, + "step": 12259 + }, + { + "epoch": 0.52, + "grad_norm": 0.421875, + "learning_rate": 0.00016853701233861905, + "loss": 2.5277, + "step": 12260 + }, + { + "epoch": 0.52, + "grad_norm": 0.478515625, + "learning_rate": 0.00016853212750298082, + "loss": 2.5538, + "step": 12261 + }, + { + "epoch": 0.52, + "grad_norm": 0.51171875, + "learning_rate": 0.00016852724235897303, + "loss": 2.5196, + "step": 12262 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.0001685223569066177, + "loss": 2.5886, + "step": 12263 + }, + { + "epoch": 0.52, + "grad_norm": 0.423828125, + "learning_rate": 0.0001685174711459368, + "loss": 2.6359, + "step": 12264 + }, + { + "epoch": 0.52, + "grad_norm": 0.416015625, + "learning_rate": 0.00016851258507695236, + "loss": 2.4731, + "step": 12265 + }, + { + "epoch": 0.52, + "grad_norm": 0.4296875, + "learning_rate": 0.00016850769869968627, + "loss": 2.5502, + "step": 12266 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.00016850281201416062, + "loss": 2.676, + "step": 12267 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.00016849792502039729, + "loss": 2.5742, + "step": 12268 + }, + { + "epoch": 0.52, + "grad_norm": 0.515625, + "learning_rate": 0.0001684930377184184, + "loss": 2.7905, + "step": 12269 + }, + { + "epoch": 0.52, + "grad_norm": 0.466796875, + "learning_rate": 0.00016848815010824582, + "loss": 2.5178, + "step": 12270 + }, + { + "epoch": 0.52, + "grad_norm": 0.462890625, + "learning_rate": 0.00016848326218990163, + "loss": 2.5618, + "step": 12271 + }, + { + "epoch": 0.52, + "grad_norm": 0.44921875, + "learning_rate": 0.00016847837396340775, + "loss": 2.5958, + "step": 12272 + }, + { + "epoch": 0.52, + "grad_norm": 0.5390625, + "learning_rate": 0.00016847348542878624, + "loss": 2.6724, + "step": 12273 + }, + { + "epoch": 0.52, + "grad_norm": 0.486328125, + "learning_rate": 0.00016846859658605908, + "loss": 2.5607, + "step": 12274 + }, + { + "epoch": 0.52, + "grad_norm": 0.51171875, + "learning_rate": 0.00016846370743524823, + "loss": 2.5772, + "step": 12275 + }, + { + "epoch": 0.52, + "grad_norm": 0.470703125, + "learning_rate": 0.00016845881797637573, + "loss": 2.6246, + "step": 12276 + }, + { + "epoch": 0.52, + "grad_norm": 0.408203125, + "learning_rate": 0.00016845392820946358, + "loss": 2.5284, + "step": 12277 + }, + { + "epoch": 0.52, + "grad_norm": 0.51953125, + "learning_rate": 0.00016844903813453374, + "loss": 2.6791, + "step": 12278 + }, + { + "epoch": 0.52, + "grad_norm": 0.46875, + "learning_rate": 0.00016844414775160827, + "loss": 2.6204, + "step": 12279 + }, + { + "epoch": 0.52, + "grad_norm": 0.4765625, + "learning_rate": 0.00016843925706070914, + "loss": 2.6416, + "step": 12280 + }, + { + "epoch": 0.52, + "grad_norm": 0.439453125, + "learning_rate": 0.00016843436606185836, + "loss": 2.606, + "step": 12281 + }, + { + "epoch": 0.52, + "grad_norm": 0.423828125, + "learning_rate": 0.00016842947475507797, + "loss": 2.6682, + "step": 12282 + }, + { + "epoch": 0.52, + "grad_norm": 0.470703125, + "learning_rate": 0.00016842458314038992, + "loss": 2.6646, + "step": 12283 + }, + { + "epoch": 0.52, + "grad_norm": 0.3984375, + "learning_rate": 0.00016841969121781625, + "loss": 2.635, + "step": 12284 + }, + { + "epoch": 0.52, + "grad_norm": 0.48828125, + "learning_rate": 0.000168414798987379, + "loss": 2.7519, + "step": 12285 + }, + { + "epoch": 0.52, + "grad_norm": 0.46875, + "learning_rate": 0.00016840990644910014, + "loss": 2.5953, + "step": 12286 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.00016840501360300173, + "loss": 2.673, + "step": 12287 + }, + { + "epoch": 0.52, + "grad_norm": 0.466796875, + "learning_rate": 0.0001684001204491057, + "loss": 2.5734, + "step": 12288 + }, + { + "epoch": 0.52, + "grad_norm": 0.4296875, + "learning_rate": 0.00016839522698743417, + "loss": 2.4269, + "step": 12289 + }, + { + "epoch": 0.52, + "grad_norm": 0.5078125, + "learning_rate": 0.0001683903332180091, + "loss": 2.6367, + "step": 12290 + }, + { + "epoch": 0.52, + "grad_norm": 0.44140625, + "learning_rate": 0.00016838543914085251, + "loss": 2.6489, + "step": 12291 + }, + { + "epoch": 0.52, + "grad_norm": 0.40625, + "learning_rate": 0.00016838054475598645, + "loss": 2.5253, + "step": 12292 + }, + { + "epoch": 0.52, + "grad_norm": 0.48046875, + "learning_rate": 0.00016837565006343295, + "loss": 2.7188, + "step": 12293 + }, + { + "epoch": 0.52, + "grad_norm": 0.5078125, + "learning_rate": 0.00016837075506321398, + "loss": 2.6136, + "step": 12294 + }, + { + "epoch": 0.52, + "grad_norm": 0.458984375, + "learning_rate": 0.00016836585975535162, + "loss": 2.5962, + "step": 12295 + }, + { + "epoch": 0.52, + "grad_norm": 0.482421875, + "learning_rate": 0.00016836096413986787, + "loss": 2.6287, + "step": 12296 + }, + { + "epoch": 0.52, + "grad_norm": 0.443359375, + "learning_rate": 0.00016835606821678475, + "loss": 2.5594, + "step": 12297 + }, + { + "epoch": 0.52, + "grad_norm": 0.470703125, + "learning_rate": 0.00016835117198612433, + "loss": 2.6342, + "step": 12298 + }, + { + "epoch": 0.52, + "grad_norm": 0.451171875, + "learning_rate": 0.00016834627544790859, + "loss": 2.6025, + "step": 12299 + }, + { + "epoch": 0.52, + "grad_norm": 0.455078125, + "learning_rate": 0.00016834137860215963, + "loss": 2.6527, + "step": 12300 + }, + { + "epoch": 0.52, + "grad_norm": 0.455078125, + "learning_rate": 0.0001683364814488994, + "loss": 2.6616, + "step": 12301 + }, + { + "epoch": 0.52, + "grad_norm": 0.51171875, + "learning_rate": 0.00016833158398815, + "loss": 2.6345, + "step": 12302 + }, + { + "epoch": 0.52, + "grad_norm": 0.482421875, + "learning_rate": 0.00016832668621993344, + "loss": 2.5962, + "step": 12303 + }, + { + "epoch": 0.52, + "grad_norm": 0.47265625, + "learning_rate": 0.00016832178814427175, + "loss": 2.7088, + "step": 12304 + }, + { + "epoch": 0.52, + "grad_norm": 0.48046875, + "learning_rate": 0.000168316889761187, + "loss": 2.676, + "step": 12305 + }, + { + "epoch": 0.52, + "grad_norm": 0.4296875, + "learning_rate": 0.00016831199107070118, + "loss": 2.5708, + "step": 12306 + }, + { + "epoch": 0.52, + "grad_norm": 0.48046875, + "learning_rate": 0.00016830709207283642, + "loss": 2.6731, + "step": 12307 + }, + { + "epoch": 0.52, + "grad_norm": 0.451171875, + "learning_rate": 0.00016830219276761464, + "loss": 2.6933, + "step": 12308 + }, + { + "epoch": 0.52, + "grad_norm": 0.462890625, + "learning_rate": 0.00016829729315505801, + "loss": 2.6426, + "step": 12309 + }, + { + "epoch": 0.52, + "grad_norm": 0.41796875, + "learning_rate": 0.0001682923932351885, + "loss": 2.6668, + "step": 12310 + }, + { + "epoch": 0.52, + "grad_norm": 0.416015625, + "learning_rate": 0.00016828749300802818, + "loss": 2.5685, + "step": 12311 + }, + { + "epoch": 0.52, + "grad_norm": 0.427734375, + "learning_rate": 0.0001682825924735991, + "loss": 2.6355, + "step": 12312 + }, + { + "epoch": 0.52, + "grad_norm": 0.470703125, + "learning_rate": 0.0001682776916319233, + "loss": 2.6199, + "step": 12313 + }, + { + "epoch": 0.52, + "grad_norm": 0.41015625, + "learning_rate": 0.00016827279048302283, + "loss": 2.6612, + "step": 12314 + }, + { + "epoch": 0.52, + "grad_norm": 0.435546875, + "learning_rate": 0.0001682678890269198, + "loss": 2.608, + "step": 12315 + }, + { + "epoch": 0.52, + "grad_norm": 0.482421875, + "learning_rate": 0.00016826298726363618, + "loss": 2.6986, + "step": 12316 + }, + { + "epoch": 0.52, + "grad_norm": 0.44921875, + "learning_rate": 0.0001682580851931941, + "loss": 2.594, + "step": 12317 + }, + { + "epoch": 0.52, + "grad_norm": 0.50390625, + "learning_rate": 0.00016825318281561553, + "loss": 2.5675, + "step": 12318 + }, + { + "epoch": 0.52, + "grad_norm": 0.453125, + "learning_rate": 0.00016824828013092263, + "loss": 2.5929, + "step": 12319 + }, + { + "epoch": 0.52, + "grad_norm": 0.443359375, + "learning_rate": 0.00016824337713913735, + "loss": 2.528, + "step": 12320 + }, + { + "epoch": 0.53, + "grad_norm": 0.4140625, + "learning_rate": 0.0001682384738402819, + "loss": 2.528, + "step": 12321 + }, + { + "epoch": 0.53, + "grad_norm": 0.49609375, + "learning_rate": 0.0001682335702343782, + "loss": 2.5972, + "step": 12322 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.00016822866632144838, + "loss": 2.5719, + "step": 12323 + }, + { + "epoch": 0.53, + "grad_norm": 0.478515625, + "learning_rate": 0.00016822376210151447, + "loss": 2.5673, + "step": 12324 + }, + { + "epoch": 0.53, + "grad_norm": 0.40234375, + "learning_rate": 0.0001682188575745986, + "loss": 2.681, + "step": 12325 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.00016821395274072278, + "loss": 2.568, + "step": 12326 + }, + { + "epoch": 0.53, + "grad_norm": 0.494140625, + "learning_rate": 0.00016820904759990912, + "loss": 2.6328, + "step": 12327 + }, + { + "epoch": 0.53, + "grad_norm": 0.43359375, + "learning_rate": 0.00016820414215217965, + "loss": 2.545, + "step": 12328 + }, + { + "epoch": 0.53, + "grad_norm": 0.41796875, + "learning_rate": 0.0001681992363975565, + "loss": 2.6564, + "step": 12329 + }, + { + "epoch": 0.53, + "grad_norm": 0.53515625, + "learning_rate": 0.00016819433033606168, + "loss": 2.5744, + "step": 12330 + }, + { + "epoch": 0.53, + "grad_norm": 0.443359375, + "learning_rate": 0.0001681894239677173, + "loss": 2.5042, + "step": 12331 + }, + { + "epoch": 0.53, + "grad_norm": 0.45703125, + "learning_rate": 0.00016818451729254546, + "loss": 2.6919, + "step": 12332 + }, + { + "epoch": 0.53, + "grad_norm": 0.4375, + "learning_rate": 0.0001681796103105682, + "loss": 2.7818, + "step": 12333 + }, + { + "epoch": 0.53, + "grad_norm": 0.466796875, + "learning_rate": 0.0001681747030218076, + "loss": 2.7006, + "step": 12334 + }, + { + "epoch": 0.53, + "grad_norm": 0.486328125, + "learning_rate": 0.00016816979542628576, + "loss": 2.8275, + "step": 12335 + }, + { + "epoch": 0.53, + "grad_norm": 0.46875, + "learning_rate": 0.0001681648875240247, + "loss": 2.6294, + "step": 12336 + }, + { + "epoch": 0.53, + "grad_norm": 0.470703125, + "learning_rate": 0.00016815997931504665, + "loss": 2.6699, + "step": 12337 + }, + { + "epoch": 0.53, + "grad_norm": 0.4296875, + "learning_rate": 0.00016815507079937354, + "loss": 2.5681, + "step": 12338 + }, + { + "epoch": 0.53, + "grad_norm": 0.431640625, + "learning_rate": 0.00016815016197702754, + "loss": 2.3392, + "step": 12339 + }, + { + "epoch": 0.53, + "grad_norm": 0.4765625, + "learning_rate": 0.00016814525284803068, + "loss": 2.5607, + "step": 12340 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.00016814034341240515, + "loss": 2.5418, + "step": 12341 + }, + { + "epoch": 0.53, + "grad_norm": 0.482421875, + "learning_rate": 0.00016813543367017294, + "loss": 2.727, + "step": 12342 + }, + { + "epoch": 0.53, + "grad_norm": 0.404296875, + "learning_rate": 0.00016813052362135619, + "loss": 2.6349, + "step": 12343 + }, + { + "epoch": 0.53, + "grad_norm": 0.453125, + "learning_rate": 0.000168125613265977, + "loss": 2.5097, + "step": 12344 + }, + { + "epoch": 0.53, + "grad_norm": 0.466796875, + "learning_rate": 0.00016812070260405738, + "loss": 2.5934, + "step": 12345 + }, + { + "epoch": 0.53, + "grad_norm": 0.44140625, + "learning_rate": 0.00016811579163561956, + "loss": 2.5557, + "step": 12346 + }, + { + "epoch": 0.53, + "grad_norm": 0.48828125, + "learning_rate": 0.00016811088036068554, + "loss": 2.5827, + "step": 12347 + }, + { + "epoch": 0.53, + "grad_norm": 0.421875, + "learning_rate": 0.00016810596877927747, + "loss": 2.5557, + "step": 12348 + }, + { + "epoch": 0.53, + "grad_norm": 0.419921875, + "learning_rate": 0.00016810105689141742, + "loss": 2.6198, + "step": 12349 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.0001680961446971275, + "loss": 2.6563, + "step": 12350 + }, + { + "epoch": 0.53, + "grad_norm": 0.4921875, + "learning_rate": 0.00016809123219642982, + "loss": 2.5819, + "step": 12351 + }, + { + "epoch": 0.53, + "grad_norm": 0.484375, + "learning_rate": 0.00016808631938934648, + "loss": 2.6663, + "step": 12352 + }, + { + "epoch": 0.53, + "grad_norm": 0.455078125, + "learning_rate": 0.00016808140627589958, + "loss": 2.5725, + "step": 12353 + }, + { + "epoch": 0.53, + "grad_norm": 0.4609375, + "learning_rate": 0.00016807649285611125, + "loss": 2.5098, + "step": 12354 + }, + { + "epoch": 0.53, + "grad_norm": 0.439453125, + "learning_rate": 0.00016807157913000355, + "loss": 2.6391, + "step": 12355 + }, + { + "epoch": 0.53, + "grad_norm": 0.423828125, + "learning_rate": 0.00016806666509759866, + "loss": 2.5598, + "step": 12356 + }, + { + "epoch": 0.53, + "grad_norm": 0.50390625, + "learning_rate": 0.0001680617507589186, + "loss": 2.6022, + "step": 12357 + }, + { + "epoch": 0.53, + "grad_norm": 0.5, + "learning_rate": 0.0001680568361139856, + "loss": 2.6409, + "step": 12358 + }, + { + "epoch": 0.53, + "grad_norm": 0.5234375, + "learning_rate": 0.00016805192116282165, + "loss": 2.6124, + "step": 12359 + }, + { + "epoch": 0.53, + "grad_norm": 0.40625, + "learning_rate": 0.00016804700590544895, + "loss": 2.5372, + "step": 12360 + }, + { + "epoch": 0.53, + "grad_norm": 0.416015625, + "learning_rate": 0.0001680420903418896, + "loss": 2.5678, + "step": 12361 + }, + { + "epoch": 0.53, + "grad_norm": 0.412109375, + "learning_rate": 0.0001680371744721657, + "loss": 2.587, + "step": 12362 + }, + { + "epoch": 0.53, + "grad_norm": 0.453125, + "learning_rate": 0.0001680322582962994, + "loss": 2.5071, + "step": 12363 + }, + { + "epoch": 0.53, + "grad_norm": 0.48046875, + "learning_rate": 0.00016802734181431276, + "loss": 2.6071, + "step": 12364 + }, + { + "epoch": 0.53, + "grad_norm": 0.486328125, + "learning_rate": 0.00016802242502622798, + "loss": 2.569, + "step": 12365 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.0001680175079320671, + "loss": 2.6231, + "step": 12366 + }, + { + "epoch": 0.53, + "grad_norm": 0.439453125, + "learning_rate": 0.00016801259053185233, + "loss": 2.605, + "step": 12367 + }, + { + "epoch": 0.53, + "grad_norm": 0.439453125, + "learning_rate": 0.00016800767282560577, + "loss": 2.5671, + "step": 12368 + }, + { + "epoch": 0.53, + "grad_norm": 0.447265625, + "learning_rate": 0.00016800275481334951, + "loss": 2.6887, + "step": 12369 + }, + { + "epoch": 0.53, + "grad_norm": 0.46484375, + "learning_rate": 0.0001679978364951057, + "loss": 2.672, + "step": 12370 + }, + { + "epoch": 0.53, + "grad_norm": 0.490234375, + "learning_rate": 0.0001679929178708965, + "loss": 2.6466, + "step": 12371 + }, + { + "epoch": 0.53, + "grad_norm": 0.466796875, + "learning_rate": 0.000167987998940744, + "loss": 2.712, + "step": 12372 + }, + { + "epoch": 0.53, + "grad_norm": 0.50390625, + "learning_rate": 0.00016798307970467038, + "loss": 2.6539, + "step": 12373 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.00016797816016269772, + "loss": 2.6165, + "step": 12374 + }, + { + "epoch": 0.53, + "grad_norm": 0.45703125, + "learning_rate": 0.00016797324031484818, + "loss": 2.6618, + "step": 12375 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.0001679683201611439, + "loss": 2.7019, + "step": 12376 + }, + { + "epoch": 0.53, + "grad_norm": 0.462890625, + "learning_rate": 0.00016796339970160705, + "loss": 2.6681, + "step": 12377 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.00016795847893625972, + "loss": 2.7126, + "step": 12378 + }, + { + "epoch": 0.53, + "grad_norm": 0.43359375, + "learning_rate": 0.00016795355786512406, + "loss": 2.5204, + "step": 12379 + }, + { + "epoch": 0.53, + "grad_norm": 0.4765625, + "learning_rate": 0.00016794863648822222, + "loss": 2.5948, + "step": 12380 + }, + { + "epoch": 0.53, + "grad_norm": 0.4140625, + "learning_rate": 0.00016794371480557637, + "loss": 2.51, + "step": 12381 + }, + { + "epoch": 0.53, + "grad_norm": 0.482421875, + "learning_rate": 0.0001679387928172086, + "loss": 2.5248, + "step": 12382 + }, + { + "epoch": 0.53, + "grad_norm": 0.53125, + "learning_rate": 0.00016793387052314114, + "loss": 2.7147, + "step": 12383 + }, + { + "epoch": 0.53, + "grad_norm": 0.44140625, + "learning_rate": 0.00016792894792339605, + "loss": 2.7047, + "step": 12384 + }, + { + "epoch": 0.53, + "grad_norm": 0.44140625, + "learning_rate": 0.00016792402501799556, + "loss": 2.6337, + "step": 12385 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.00016791910180696172, + "loss": 2.5607, + "step": 12386 + }, + { + "epoch": 0.53, + "grad_norm": 0.462890625, + "learning_rate": 0.0001679141782903168, + "loss": 2.5896, + "step": 12387 + }, + { + "epoch": 0.53, + "grad_norm": 0.4375, + "learning_rate": 0.00016790925446808288, + "loss": 2.495, + "step": 12388 + }, + { + "epoch": 0.53, + "grad_norm": 0.490234375, + "learning_rate": 0.00016790433034028212, + "loss": 2.7631, + "step": 12389 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.0001678994059069367, + "loss": 2.5939, + "step": 12390 + }, + { + "epoch": 0.53, + "grad_norm": 0.49609375, + "learning_rate": 0.00016789448116806877, + "loss": 2.6291, + "step": 12391 + }, + { + "epoch": 0.53, + "grad_norm": 0.478515625, + "learning_rate": 0.00016788955612370048, + "loss": 2.6601, + "step": 12392 + }, + { + "epoch": 0.53, + "grad_norm": 0.50390625, + "learning_rate": 0.000167884630773854, + "loss": 2.6354, + "step": 12393 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.0001678797051185515, + "loss": 2.6825, + "step": 12394 + }, + { + "epoch": 0.53, + "grad_norm": 0.546875, + "learning_rate": 0.0001678747791578151, + "loss": 2.6208, + "step": 12395 + }, + { + "epoch": 0.53, + "grad_norm": 0.494140625, + "learning_rate": 0.00016786985289166703, + "loss": 2.6563, + "step": 12396 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.00016786492632012942, + "loss": 2.5587, + "step": 12397 + }, + { + "epoch": 0.53, + "grad_norm": 0.46484375, + "learning_rate": 0.00016785999944322444, + "loss": 2.6068, + "step": 12398 + }, + { + "epoch": 0.53, + "grad_norm": 0.431640625, + "learning_rate": 0.0001678550722609743, + "loss": 2.6161, + "step": 12399 + }, + { + "epoch": 0.53, + "grad_norm": 0.443359375, + "learning_rate": 0.00016785014477340107, + "loss": 2.6398, + "step": 12400 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.00016784521698052701, + "loss": 2.698, + "step": 12401 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.00016784028888237429, + "loss": 2.5552, + "step": 12402 + }, + { + "epoch": 0.53, + "grad_norm": 0.490234375, + "learning_rate": 0.00016783536047896503, + "loss": 2.631, + "step": 12403 + }, + { + "epoch": 0.53, + "grad_norm": 0.53125, + "learning_rate": 0.00016783043177032148, + "loss": 2.6648, + "step": 12404 + }, + { + "epoch": 0.53, + "grad_norm": 0.4296875, + "learning_rate": 0.00016782550275646574, + "loss": 2.633, + "step": 12405 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.00016782057343742003, + "loss": 2.6501, + "step": 12406 + }, + { + "epoch": 0.53, + "grad_norm": 0.44140625, + "learning_rate": 0.00016781564381320653, + "loss": 2.6144, + "step": 12407 + }, + { + "epoch": 0.53, + "grad_norm": 0.427734375, + "learning_rate": 0.00016781071388384744, + "loss": 2.5795, + "step": 12408 + }, + { + "epoch": 0.53, + "grad_norm": 0.486328125, + "learning_rate": 0.00016780578364936487, + "loss": 2.6491, + "step": 12409 + }, + { + "epoch": 0.53, + "grad_norm": 0.421875, + "learning_rate": 0.00016780085310978107, + "loss": 2.6242, + "step": 12410 + }, + { + "epoch": 0.53, + "grad_norm": 0.455078125, + "learning_rate": 0.0001677959222651182, + "loss": 2.7117, + "step": 12411 + }, + { + "epoch": 0.53, + "grad_norm": 0.52734375, + "learning_rate": 0.0001677909911153985, + "loss": 2.6914, + "step": 12412 + }, + { + "epoch": 0.53, + "grad_norm": 0.40625, + "learning_rate": 0.0001677860596606441, + "loss": 2.4926, + "step": 12413 + }, + { + "epoch": 0.53, + "grad_norm": 0.50390625, + "learning_rate": 0.00016778112790087718, + "loss": 2.5718, + "step": 12414 + }, + { + "epoch": 0.53, + "grad_norm": 0.423828125, + "learning_rate": 0.00016777619583611996, + "loss": 2.5007, + "step": 12415 + }, + { + "epoch": 0.53, + "grad_norm": 0.4609375, + "learning_rate": 0.00016777126346639466, + "loss": 2.6637, + "step": 12416 + }, + { + "epoch": 0.53, + "grad_norm": 0.498046875, + "learning_rate": 0.00016776633079172343, + "loss": 2.5278, + "step": 12417 + }, + { + "epoch": 0.53, + "grad_norm": 0.48046875, + "learning_rate": 0.00016776139781212845, + "loss": 2.511, + "step": 12418 + }, + { + "epoch": 0.53, + "grad_norm": 0.4375, + "learning_rate": 0.000167756464527632, + "loss": 2.681, + "step": 12419 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.00016775153093825614, + "loss": 2.6207, + "step": 12420 + }, + { + "epoch": 0.53, + "grad_norm": 0.4375, + "learning_rate": 0.0001677465970440232, + "loss": 2.5751, + "step": 12421 + }, + { + "epoch": 0.53, + "grad_norm": 0.4609375, + "learning_rate": 0.00016774166284495533, + "loss": 2.6226, + "step": 12422 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.00016773672834107472, + "loss": 2.6758, + "step": 12423 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.00016773179353240362, + "loss": 2.5527, + "step": 12424 + }, + { + "epoch": 0.53, + "grad_norm": 0.4140625, + "learning_rate": 0.00016772685841896418, + "loss": 2.4347, + "step": 12425 + }, + { + "epoch": 0.53, + "grad_norm": 0.4140625, + "learning_rate": 0.00016772192300077863, + "loss": 2.6337, + "step": 12426 + }, + { + "epoch": 0.53, + "grad_norm": 0.439453125, + "learning_rate": 0.00016771698727786918, + "loss": 2.6048, + "step": 12427 + }, + { + "epoch": 0.53, + "grad_norm": 0.484375, + "learning_rate": 0.00016771205125025803, + "loss": 2.6801, + "step": 12428 + }, + { + "epoch": 0.53, + "grad_norm": 0.416015625, + "learning_rate": 0.00016770711491796738, + "loss": 2.5065, + "step": 12429 + }, + { + "epoch": 0.53, + "grad_norm": 0.48046875, + "learning_rate": 0.0001677021782810195, + "loss": 2.7157, + "step": 12430 + }, + { + "epoch": 0.53, + "grad_norm": 0.51171875, + "learning_rate": 0.00016769724133943652, + "loss": 2.6013, + "step": 12431 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.0001676923040932407, + "loss": 2.5945, + "step": 12432 + }, + { + "epoch": 0.53, + "grad_norm": 0.4765625, + "learning_rate": 0.00016768736654245426, + "loss": 2.5452, + "step": 12433 + }, + { + "epoch": 0.53, + "grad_norm": 0.466796875, + "learning_rate": 0.0001676824286870994, + "loss": 2.5953, + "step": 12434 + }, + { + "epoch": 0.53, + "grad_norm": 0.4765625, + "learning_rate": 0.00016767749052719834, + "loss": 2.7058, + "step": 12435 + }, + { + "epoch": 0.53, + "grad_norm": 0.416015625, + "learning_rate": 0.00016767255206277332, + "loss": 2.5407, + "step": 12436 + }, + { + "epoch": 0.53, + "grad_norm": 0.494140625, + "learning_rate": 0.00016766761329384654, + "loss": 2.4733, + "step": 12437 + }, + { + "epoch": 0.53, + "grad_norm": 0.46484375, + "learning_rate": 0.00016766267422044023, + "loss": 2.5794, + "step": 12438 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.00016765773484257664, + "loss": 2.5829, + "step": 12439 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.00016765279516027794, + "loss": 2.5701, + "step": 12440 + }, + { + "epoch": 0.53, + "grad_norm": 0.455078125, + "learning_rate": 0.0001676478551735664, + "loss": 2.5487, + "step": 12441 + }, + { + "epoch": 0.53, + "grad_norm": 0.421875, + "learning_rate": 0.0001676429148824642, + "loss": 2.5229, + "step": 12442 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.00016763797428699363, + "loss": 2.5951, + "step": 12443 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.0001676330333871769, + "loss": 2.5955, + "step": 12444 + }, + { + "epoch": 0.53, + "grad_norm": 0.5234375, + "learning_rate": 0.00016762809218303626, + "loss": 2.6868, + "step": 12445 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.00016762315067459386, + "loss": 2.7166, + "step": 12446 + }, + { + "epoch": 0.53, + "grad_norm": 0.447265625, + "learning_rate": 0.00016761820886187202, + "loss": 2.6889, + "step": 12447 + }, + { + "epoch": 0.53, + "grad_norm": 0.486328125, + "learning_rate": 0.000167613266744893, + "loss": 2.6273, + "step": 12448 + }, + { + "epoch": 0.53, + "grad_norm": 0.396484375, + "learning_rate": 0.00016760832432367891, + "loss": 2.4542, + "step": 12449 + }, + { + "epoch": 0.53, + "grad_norm": 0.439453125, + "learning_rate": 0.00016760338159825213, + "loss": 2.5795, + "step": 12450 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.0001675984385686348, + "loss": 2.6169, + "step": 12451 + }, + { + "epoch": 0.53, + "grad_norm": 0.462890625, + "learning_rate": 0.00016759349523484923, + "loss": 2.5281, + "step": 12452 + }, + { + "epoch": 0.53, + "grad_norm": 0.431640625, + "learning_rate": 0.00016758855159691762, + "loss": 2.5306, + "step": 12453 + }, + { + "epoch": 0.53, + "grad_norm": 0.50390625, + "learning_rate": 0.00016758360765486226, + "loss": 2.6516, + "step": 12454 + }, + { + "epoch": 0.53, + "grad_norm": 0.453125, + "learning_rate": 0.00016757866340870532, + "loss": 2.5512, + "step": 12455 + }, + { + "epoch": 0.53, + "grad_norm": 0.43359375, + "learning_rate": 0.00016757371885846908, + "loss": 2.6429, + "step": 12456 + }, + { + "epoch": 0.53, + "grad_norm": 0.51171875, + "learning_rate": 0.00016756877400417585, + "loss": 2.6012, + "step": 12457 + }, + { + "epoch": 0.53, + "grad_norm": 0.482421875, + "learning_rate": 0.0001675638288458478, + "loss": 2.5956, + "step": 12458 + }, + { + "epoch": 0.53, + "grad_norm": 0.421875, + "learning_rate": 0.0001675588833835072, + "loss": 2.7055, + "step": 12459 + }, + { + "epoch": 0.53, + "grad_norm": 0.435546875, + "learning_rate": 0.00016755393761717636, + "loss": 2.5439, + "step": 12460 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.00016754899154687743, + "loss": 2.6629, + "step": 12461 + }, + { + "epoch": 0.53, + "grad_norm": 0.470703125, + "learning_rate": 0.00016754404517263277, + "loss": 2.6736, + "step": 12462 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.0001675390984944646, + "loss": 2.6946, + "step": 12463 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.00016753415151239516, + "loss": 2.4746, + "step": 12464 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.00016752920422644668, + "loss": 2.6875, + "step": 12465 + }, + { + "epoch": 0.53, + "grad_norm": 0.453125, + "learning_rate": 0.00016752425663664152, + "loss": 2.7536, + "step": 12466 + }, + { + "epoch": 0.53, + "grad_norm": 0.4296875, + "learning_rate": 0.00016751930874300182, + "loss": 2.5321, + "step": 12467 + }, + { + "epoch": 0.53, + "grad_norm": 0.4609375, + "learning_rate": 0.00016751436054554994, + "loss": 2.6296, + "step": 12468 + }, + { + "epoch": 0.53, + "grad_norm": 0.4921875, + "learning_rate": 0.0001675094120443081, + "loss": 2.6817, + "step": 12469 + }, + { + "epoch": 0.53, + "grad_norm": 0.451171875, + "learning_rate": 0.00016750446323929858, + "loss": 2.5462, + "step": 12470 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.00016749951413054366, + "loss": 2.5218, + "step": 12471 + }, + { + "epoch": 0.53, + "grad_norm": 0.453125, + "learning_rate": 0.0001674945647180656, + "loss": 2.7371, + "step": 12472 + }, + { + "epoch": 0.53, + "grad_norm": 0.421875, + "learning_rate": 0.00016748961500188662, + "loss": 2.5626, + "step": 12473 + }, + { + "epoch": 0.53, + "grad_norm": 0.45703125, + "learning_rate": 0.00016748466498202907, + "loss": 2.5926, + "step": 12474 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.0001674797146585152, + "loss": 2.5577, + "step": 12475 + }, + { + "epoch": 0.53, + "grad_norm": 0.4296875, + "learning_rate": 0.00016747476403136728, + "loss": 2.5721, + "step": 12476 + }, + { + "epoch": 0.53, + "grad_norm": 0.4140625, + "learning_rate": 0.00016746981310060757, + "loss": 2.6779, + "step": 12477 + }, + { + "epoch": 0.53, + "grad_norm": 0.435546875, + "learning_rate": 0.00016746486186625833, + "loss": 2.5776, + "step": 12478 + }, + { + "epoch": 0.53, + "grad_norm": 0.478515625, + "learning_rate": 0.0001674599103283419, + "loss": 2.5993, + "step": 12479 + }, + { + "epoch": 0.53, + "grad_norm": 0.51171875, + "learning_rate": 0.0001674549584868805, + "loss": 2.5094, + "step": 12480 + }, + { + "epoch": 0.53, + "grad_norm": 0.431640625, + "learning_rate": 0.00016745000634189647, + "loss": 2.5854, + "step": 12481 + }, + { + "epoch": 0.53, + "grad_norm": 0.416015625, + "learning_rate": 0.00016744505389341204, + "loss": 2.6617, + "step": 12482 + }, + { + "epoch": 0.53, + "grad_norm": 0.443359375, + "learning_rate": 0.00016744010114144953, + "loss": 2.4772, + "step": 12483 + }, + { + "epoch": 0.53, + "grad_norm": 0.451171875, + "learning_rate": 0.0001674351480860312, + "loss": 2.5518, + "step": 12484 + }, + { + "epoch": 0.53, + "grad_norm": 0.392578125, + "learning_rate": 0.00016743019472717936, + "loss": 2.4892, + "step": 12485 + }, + { + "epoch": 0.53, + "grad_norm": 0.478515625, + "learning_rate": 0.00016742524106491628, + "loss": 2.5486, + "step": 12486 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.00016742028709926425, + "loss": 2.5159, + "step": 12487 + }, + { + "epoch": 0.53, + "grad_norm": 0.486328125, + "learning_rate": 0.00016741533283024556, + "loss": 2.481, + "step": 12488 + }, + { + "epoch": 0.53, + "grad_norm": 0.4140625, + "learning_rate": 0.00016741037825788256, + "loss": 2.6163, + "step": 12489 + }, + { + "epoch": 0.53, + "grad_norm": 0.451171875, + "learning_rate": 0.0001674054233821974, + "loss": 2.666, + "step": 12490 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.00016740046820321256, + "loss": 2.4878, + "step": 12491 + }, + { + "epoch": 0.53, + "grad_norm": 0.490234375, + "learning_rate": 0.0001673955127209502, + "loss": 2.7645, + "step": 12492 + }, + { + "epoch": 0.53, + "grad_norm": 0.412109375, + "learning_rate": 0.00016739055693543268, + "loss": 2.5358, + "step": 12493 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.00016738560084668226, + "loss": 2.5323, + "step": 12494 + }, + { + "epoch": 0.53, + "grad_norm": 0.41796875, + "learning_rate": 0.0001673806444547213, + "loss": 2.6514, + "step": 12495 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.00016737568775957202, + "loss": 2.4596, + "step": 12496 + }, + { + "epoch": 0.53, + "grad_norm": 0.451171875, + "learning_rate": 0.0001673707307612568, + "loss": 2.452, + "step": 12497 + }, + { + "epoch": 0.53, + "grad_norm": 0.39453125, + "learning_rate": 0.00016736577345979789, + "loss": 2.4486, + "step": 12498 + }, + { + "epoch": 0.53, + "grad_norm": 0.3984375, + "learning_rate": 0.00016736081585521766, + "loss": 2.4931, + "step": 12499 + }, + { + "epoch": 0.53, + "grad_norm": 0.46484375, + "learning_rate": 0.00016735585794753834, + "loss": 2.6347, + "step": 12500 + }, + { + "epoch": 0.53, + "grad_norm": 0.46875, + "learning_rate": 0.00016735089973678229, + "loss": 2.6663, + "step": 12501 + }, + { + "epoch": 0.53, + "grad_norm": 0.45703125, + "learning_rate": 0.00016734594122297177, + "loss": 2.4906, + "step": 12502 + }, + { + "epoch": 0.53, + "grad_norm": 0.44140625, + "learning_rate": 0.00016734098240612916, + "loss": 2.6378, + "step": 12503 + }, + { + "epoch": 0.53, + "grad_norm": 0.455078125, + "learning_rate": 0.00016733602328627677, + "loss": 2.6644, + "step": 12504 + }, + { + "epoch": 0.53, + "grad_norm": 0.5, + "learning_rate": 0.00016733106386343684, + "loss": 2.6236, + "step": 12505 + }, + { + "epoch": 0.53, + "grad_norm": 0.455078125, + "learning_rate": 0.00016732610413763174, + "loss": 2.465, + "step": 12506 + }, + { + "epoch": 0.53, + "grad_norm": 0.4375, + "learning_rate": 0.00016732114410888377, + "loss": 2.6663, + "step": 12507 + }, + { + "epoch": 0.53, + "grad_norm": 0.48828125, + "learning_rate": 0.00016731618377721524, + "loss": 2.6179, + "step": 12508 + }, + { + "epoch": 0.53, + "grad_norm": 0.421875, + "learning_rate": 0.00016731122314264852, + "loss": 2.6513, + "step": 12509 + }, + { + "epoch": 0.53, + "grad_norm": 0.455078125, + "learning_rate": 0.00016730626220520586, + "loss": 2.5781, + "step": 12510 + }, + { + "epoch": 0.53, + "grad_norm": 0.4375, + "learning_rate": 0.00016730130096490964, + "loss": 2.5753, + "step": 12511 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.00016729633942178215, + "loss": 2.707, + "step": 12512 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.00016729137757584575, + "loss": 2.5601, + "step": 12513 + }, + { + "epoch": 0.53, + "grad_norm": 0.451171875, + "learning_rate": 0.0001672864154271227, + "loss": 2.587, + "step": 12514 + }, + { + "epoch": 0.53, + "grad_norm": 0.41796875, + "learning_rate": 0.00016728145297563544, + "loss": 2.6151, + "step": 12515 + }, + { + "epoch": 0.53, + "grad_norm": 0.48828125, + "learning_rate": 0.00016727649022140617, + "loss": 2.5663, + "step": 12516 + }, + { + "epoch": 0.53, + "grad_norm": 0.37890625, + "learning_rate": 0.00016727152716445727, + "loss": 2.6228, + "step": 12517 + }, + { + "epoch": 0.53, + "grad_norm": 0.431640625, + "learning_rate": 0.00016726656380481114, + "loss": 2.6559, + "step": 12518 + }, + { + "epoch": 0.53, + "grad_norm": 0.4921875, + "learning_rate": 0.00016726160014249005, + "loss": 2.6352, + "step": 12519 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.0001672566361775163, + "loss": 2.6703, + "step": 12520 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.0001672516719099123, + "loss": 2.6624, + "step": 12521 + }, + { + "epoch": 0.53, + "grad_norm": 0.482421875, + "learning_rate": 0.0001672467073397003, + "loss": 2.7324, + "step": 12522 + }, + { + "epoch": 0.53, + "grad_norm": 0.48828125, + "learning_rate": 0.00016724174246690273, + "loss": 2.5203, + "step": 12523 + }, + { + "epoch": 0.53, + "grad_norm": 0.470703125, + "learning_rate": 0.0001672367772915419, + "loss": 2.5517, + "step": 12524 + }, + { + "epoch": 0.53, + "grad_norm": 0.474609375, + "learning_rate": 0.00016723181181364015, + "loss": 2.7056, + "step": 12525 + }, + { + "epoch": 0.53, + "grad_norm": 0.43359375, + "learning_rate": 0.0001672268460332198, + "loss": 2.5524, + "step": 12526 + }, + { + "epoch": 0.53, + "grad_norm": 0.44140625, + "learning_rate": 0.00016722187995030322, + "loss": 2.4425, + "step": 12527 + }, + { + "epoch": 0.53, + "grad_norm": 0.423828125, + "learning_rate": 0.00016721691356491274, + "loss": 2.4626, + "step": 12528 + }, + { + "epoch": 0.53, + "grad_norm": 0.458984375, + "learning_rate": 0.00016721194687707074, + "loss": 2.5883, + "step": 12529 + }, + { + "epoch": 0.53, + "grad_norm": 0.53515625, + "learning_rate": 0.0001672069798867995, + "loss": 2.6292, + "step": 12530 + }, + { + "epoch": 0.53, + "grad_norm": 0.439453125, + "learning_rate": 0.00016720201259412144, + "loss": 2.4764, + "step": 12531 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.00016719704499905887, + "loss": 2.6297, + "step": 12532 + }, + { + "epoch": 0.53, + "grad_norm": 0.462890625, + "learning_rate": 0.00016719207710163417, + "loss": 2.5181, + "step": 12533 + }, + { + "epoch": 0.53, + "grad_norm": 0.421875, + "learning_rate": 0.00016718710890186968, + "loss": 2.587, + "step": 12534 + }, + { + "epoch": 0.53, + "grad_norm": 0.470703125, + "learning_rate": 0.00016718214039978776, + "loss": 2.6689, + "step": 12535 + }, + { + "epoch": 0.53, + "grad_norm": 0.4765625, + "learning_rate": 0.00016717717159541072, + "loss": 2.6028, + "step": 12536 + }, + { + "epoch": 0.53, + "grad_norm": 0.515625, + "learning_rate": 0.00016717220248876097, + "loss": 2.6619, + "step": 12537 + }, + { + "epoch": 0.53, + "grad_norm": 0.51171875, + "learning_rate": 0.00016716723307986085, + "loss": 2.5798, + "step": 12538 + }, + { + "epoch": 0.53, + "grad_norm": 0.5, + "learning_rate": 0.00016716226336873277, + "loss": 2.6571, + "step": 12539 + }, + { + "epoch": 0.53, + "grad_norm": 0.62890625, + "learning_rate": 0.00016715729335539905, + "loss": 2.6053, + "step": 12540 + }, + { + "epoch": 0.53, + "grad_norm": 0.486328125, + "learning_rate": 0.000167152323039882, + "loss": 2.688, + "step": 12541 + }, + { + "epoch": 0.53, + "grad_norm": 0.439453125, + "learning_rate": 0.00016714735242220408, + "loss": 2.5137, + "step": 12542 + }, + { + "epoch": 0.53, + "grad_norm": 0.41796875, + "learning_rate": 0.0001671423815023876, + "loss": 2.5424, + "step": 12543 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.00016713741028045496, + "loss": 2.6788, + "step": 12544 + }, + { + "epoch": 0.53, + "grad_norm": 0.466796875, + "learning_rate": 0.0001671324387564285, + "loss": 2.6006, + "step": 12545 + }, + { + "epoch": 0.53, + "grad_norm": 0.45703125, + "learning_rate": 0.00016712746693033057, + "loss": 2.5598, + "step": 12546 + }, + { + "epoch": 0.53, + "grad_norm": 0.45703125, + "learning_rate": 0.00016712249480218362, + "loss": 2.5054, + "step": 12547 + }, + { + "epoch": 0.53, + "grad_norm": 0.40234375, + "learning_rate": 0.00016711752237200993, + "loss": 2.4676, + "step": 12548 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.00016711254963983193, + "loss": 2.5627, + "step": 12549 + }, + { + "epoch": 0.53, + "grad_norm": 0.462890625, + "learning_rate": 0.000167107576605672, + "loss": 2.5848, + "step": 12550 + }, + { + "epoch": 0.53, + "grad_norm": 0.42578125, + "learning_rate": 0.0001671026032695525, + "loss": 2.6333, + "step": 12551 + }, + { + "epoch": 0.53, + "grad_norm": 0.44921875, + "learning_rate": 0.0001670976296314958, + "loss": 2.4335, + "step": 12552 + }, + { + "epoch": 0.53, + "grad_norm": 0.4453125, + "learning_rate": 0.00016709265569152425, + "loss": 2.5497, + "step": 12553 + }, + { + "epoch": 0.53, + "grad_norm": 0.494140625, + "learning_rate": 0.00016708768144966033, + "loss": 2.714, + "step": 12554 + }, + { + "epoch": 0.54, + "grad_norm": 0.46484375, + "learning_rate": 0.00016708270690592633, + "loss": 2.6763, + "step": 12555 + }, + { + "epoch": 0.54, + "grad_norm": 0.43359375, + "learning_rate": 0.00016707773206034468, + "loss": 2.4809, + "step": 12556 + }, + { + "epoch": 0.54, + "grad_norm": 0.466796875, + "learning_rate": 0.00016707275691293772, + "loss": 2.6758, + "step": 12557 + }, + { + "epoch": 0.54, + "grad_norm": 0.404296875, + "learning_rate": 0.0001670677814637279, + "loss": 2.5802, + "step": 12558 + }, + { + "epoch": 0.54, + "grad_norm": 0.455078125, + "learning_rate": 0.00016706280571273757, + "loss": 2.65, + "step": 12559 + }, + { + "epoch": 0.54, + "grad_norm": 0.482421875, + "learning_rate": 0.0001670578296599891, + "loss": 2.5688, + "step": 12560 + }, + { + "epoch": 0.54, + "grad_norm": 0.48828125, + "learning_rate": 0.0001670528533055049, + "loss": 2.6147, + "step": 12561 + }, + { + "epoch": 0.54, + "grad_norm": 0.451171875, + "learning_rate": 0.0001670478766493074, + "loss": 2.5605, + "step": 12562 + }, + { + "epoch": 0.54, + "grad_norm": 0.443359375, + "learning_rate": 0.00016704289969141892, + "loss": 2.6911, + "step": 12563 + }, + { + "epoch": 0.54, + "grad_norm": 0.48046875, + "learning_rate": 0.00016703792243186194, + "loss": 2.6725, + "step": 12564 + }, + { + "epoch": 0.54, + "grad_norm": 0.4296875, + "learning_rate": 0.00016703294487065877, + "loss": 2.6044, + "step": 12565 + }, + { + "epoch": 0.54, + "grad_norm": 0.44921875, + "learning_rate": 0.00016702796700783187, + "loss": 2.6558, + "step": 12566 + }, + { + "epoch": 0.54, + "grad_norm": 0.455078125, + "learning_rate": 0.0001670229888434036, + "loss": 2.5857, + "step": 12567 + }, + { + "epoch": 0.54, + "grad_norm": 0.466796875, + "learning_rate": 0.00016701801037739635, + "loss": 2.5847, + "step": 12568 + }, + { + "epoch": 0.54, + "grad_norm": 0.45703125, + "learning_rate": 0.0001670130316098326, + "loss": 2.6013, + "step": 12569 + }, + { + "epoch": 0.54, + "grad_norm": 0.455078125, + "learning_rate": 0.00016700805254073466, + "loss": 2.5943, + "step": 12570 + }, + { + "epoch": 0.54, + "grad_norm": 0.490234375, + "learning_rate": 0.000167003073170125, + "loss": 2.4778, + "step": 12571 + }, + { + "epoch": 0.54, + "grad_norm": 0.4296875, + "learning_rate": 0.000166998093498026, + "loss": 2.5971, + "step": 12572 + }, + { + "epoch": 0.54, + "grad_norm": 0.412109375, + "learning_rate": 0.00016699311352446, + "loss": 2.4654, + "step": 12573 + }, + { + "epoch": 0.54, + "grad_norm": 0.3828125, + "learning_rate": 0.00016698813324944953, + "loss": 2.5635, + "step": 12574 + }, + { + "epoch": 0.54, + "grad_norm": 0.5, + "learning_rate": 0.0001669831526730169, + "loss": 2.5934, + "step": 12575 + }, + { + "epoch": 0.54, + "grad_norm": 0.44921875, + "learning_rate": 0.0001669781717951846, + "loss": 2.5957, + "step": 12576 + }, + { + "epoch": 0.54, + "grad_norm": 0.416015625, + "learning_rate": 0.00016697319061597502, + "loss": 2.6718, + "step": 12577 + }, + { + "epoch": 0.54, + "grad_norm": 0.453125, + "learning_rate": 0.00016696820913541052, + "loss": 2.6562, + "step": 12578 + }, + { + "epoch": 0.54, + "grad_norm": 0.4375, + "learning_rate": 0.00016696322735351357, + "loss": 2.5488, + "step": 12579 + }, + { + "epoch": 0.54, + "grad_norm": 0.51171875, + "learning_rate": 0.00016695824527030654, + "loss": 2.56, + "step": 12580 + }, + { + "epoch": 0.54, + "grad_norm": 0.427734375, + "learning_rate": 0.00016695326288581193, + "loss": 2.5576, + "step": 12581 + }, + { + "epoch": 0.54, + "grad_norm": 0.451171875, + "learning_rate": 0.00016694828020005206, + "loss": 2.6691, + "step": 12582 + }, + { + "epoch": 0.54, + "grad_norm": 0.484375, + "learning_rate": 0.00016694329721304944, + "loss": 2.6907, + "step": 12583 + }, + { + "epoch": 0.54, + "grad_norm": 0.41796875, + "learning_rate": 0.0001669383139248264, + "loss": 2.4987, + "step": 12584 + }, + { + "epoch": 0.54, + "grad_norm": 0.443359375, + "learning_rate": 0.00016693333033540543, + "loss": 2.4892, + "step": 12585 + }, + { + "epoch": 0.54, + "grad_norm": 0.4296875, + "learning_rate": 0.00016692834644480896, + "loss": 2.6971, + "step": 12586 + }, + { + "epoch": 0.54, + "grad_norm": 0.498046875, + "learning_rate": 0.00016692336225305936, + "loss": 2.6915, + "step": 12587 + }, + { + "epoch": 0.54, + "grad_norm": 0.408203125, + "learning_rate": 0.0001669183777601791, + "loss": 2.4564, + "step": 12588 + }, + { + "epoch": 0.54, + "grad_norm": 0.48046875, + "learning_rate": 0.00016691339296619057, + "loss": 2.5282, + "step": 12589 + }, + { + "epoch": 0.54, + "grad_norm": 0.46875, + "learning_rate": 0.00016690840787111626, + "loss": 2.6561, + "step": 12590 + }, + { + "epoch": 0.54, + "grad_norm": 0.5078125, + "learning_rate": 0.00016690342247497854, + "loss": 2.7371, + "step": 12591 + }, + { + "epoch": 0.54, + "grad_norm": 0.474609375, + "learning_rate": 0.0001668984367777999, + "loss": 2.6799, + "step": 12592 + }, + { + "epoch": 0.54, + "grad_norm": 0.46875, + "learning_rate": 0.00016689345077960274, + "loss": 2.5864, + "step": 12593 + }, + { + "epoch": 0.54, + "grad_norm": 0.46484375, + "learning_rate": 0.00016688846448040948, + "loss": 2.7108, + "step": 12594 + }, + { + "epoch": 0.54, + "grad_norm": 0.49609375, + "learning_rate": 0.00016688347788024256, + "loss": 2.4611, + "step": 12595 + }, + { + "epoch": 0.54, + "grad_norm": 0.48828125, + "learning_rate": 0.00016687849097912446, + "loss": 2.6223, + "step": 12596 + }, + { + "epoch": 0.54, + "grad_norm": 0.466796875, + "learning_rate": 0.0001668735037770776, + "loss": 2.7407, + "step": 12597 + }, + { + "epoch": 0.54, + "grad_norm": 0.439453125, + "learning_rate": 0.00016686851627412437, + "loss": 2.568, + "step": 12598 + }, + { + "epoch": 0.54, + "grad_norm": 0.408203125, + "learning_rate": 0.0001668635284702873, + "loss": 2.5785, + "step": 12599 + }, + { + "epoch": 0.54, + "grad_norm": 0.439453125, + "learning_rate": 0.00016685854036558876, + "loss": 2.5316, + "step": 12600 + }, + { + "epoch": 0.54, + "grad_norm": 0.4140625, + "learning_rate": 0.00016685355196005124, + "loss": 2.7818, + "step": 12601 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.00016684856325369714, + "loss": 2.6712, + "step": 12602 + }, + { + "epoch": 0.54, + "grad_norm": 0.427734375, + "learning_rate": 0.00016684357424654895, + "loss": 2.5963, + "step": 12603 + }, + { + "epoch": 0.54, + "grad_norm": 0.5, + "learning_rate": 0.00016683858493862913, + "loss": 2.541, + "step": 12604 + }, + { + "epoch": 0.54, + "grad_norm": 0.47265625, + "learning_rate": 0.00016683359532996005, + "loss": 2.4512, + "step": 12605 + }, + { + "epoch": 0.54, + "grad_norm": 0.44921875, + "learning_rate": 0.0001668286054205643, + "loss": 2.5071, + "step": 12606 + }, + { + "epoch": 0.54, + "grad_norm": 0.48828125, + "learning_rate": 0.00016682361521046417, + "loss": 2.5425, + "step": 12607 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.0001668186246996822, + "loss": 2.5612, + "step": 12608 + }, + { + "epoch": 0.54, + "grad_norm": 0.45703125, + "learning_rate": 0.00016681363388824086, + "loss": 2.6161, + "step": 12609 + }, + { + "epoch": 0.54, + "grad_norm": 0.486328125, + "learning_rate": 0.0001668086427761626, + "loss": 2.6138, + "step": 12610 + }, + { + "epoch": 0.54, + "grad_norm": 0.45703125, + "learning_rate": 0.00016680365136346984, + "loss": 2.6316, + "step": 12611 + }, + { + "epoch": 0.54, + "grad_norm": 0.439453125, + "learning_rate": 0.00016679865965018507, + "loss": 2.5811, + "step": 12612 + }, + { + "epoch": 0.54, + "grad_norm": 0.421875, + "learning_rate": 0.00016679366763633076, + "loss": 2.6203, + "step": 12613 + }, + { + "epoch": 0.54, + "grad_norm": 0.45703125, + "learning_rate": 0.00016678867532192933, + "loss": 2.6342, + "step": 12614 + }, + { + "epoch": 0.54, + "grad_norm": 0.486328125, + "learning_rate": 0.00016678368270700328, + "loss": 2.6056, + "step": 12615 + }, + { + "epoch": 0.54, + "grad_norm": 0.47265625, + "learning_rate": 0.00016677868979157506, + "loss": 2.8154, + "step": 12616 + }, + { + "epoch": 0.54, + "grad_norm": 0.435546875, + "learning_rate": 0.00016677369657566715, + "loss": 2.7203, + "step": 12617 + }, + { + "epoch": 0.54, + "grad_norm": 0.47265625, + "learning_rate": 0.00016676870305930202, + "loss": 2.6418, + "step": 12618 + }, + { + "epoch": 0.54, + "grad_norm": 0.4921875, + "learning_rate": 0.00016676370924250208, + "loss": 2.5801, + "step": 12619 + }, + { + "epoch": 0.54, + "grad_norm": 0.5, + "learning_rate": 0.00016675871512528988, + "loss": 2.7088, + "step": 12620 + }, + { + "epoch": 0.54, + "grad_norm": 0.4296875, + "learning_rate": 0.00016675372070768787, + "loss": 2.562, + "step": 12621 + }, + { + "epoch": 0.54, + "grad_norm": 0.4453125, + "learning_rate": 0.00016674872598971852, + "loss": 2.5332, + "step": 12622 + }, + { + "epoch": 0.54, + "grad_norm": 0.42578125, + "learning_rate": 0.00016674373097140427, + "loss": 2.6774, + "step": 12623 + }, + { + "epoch": 0.54, + "grad_norm": 0.44140625, + "learning_rate": 0.0001667387356527676, + "loss": 2.6174, + "step": 12624 + }, + { + "epoch": 0.54, + "grad_norm": 0.4453125, + "learning_rate": 0.00016673374003383105, + "loss": 2.5976, + "step": 12625 + }, + { + "epoch": 0.54, + "grad_norm": 0.47265625, + "learning_rate": 0.00016672874411461706, + "loss": 2.7325, + "step": 12626 + }, + { + "epoch": 0.54, + "grad_norm": 0.5234375, + "learning_rate": 0.0001667237478951481, + "loss": 2.6071, + "step": 12627 + }, + { + "epoch": 0.54, + "grad_norm": 0.5234375, + "learning_rate": 0.00016671875137544664, + "loss": 2.7963, + "step": 12628 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.0001667137545555352, + "loss": 2.5944, + "step": 12629 + }, + { + "epoch": 0.54, + "grad_norm": 0.494140625, + "learning_rate": 0.00016670875743543623, + "loss": 2.733, + "step": 12630 + }, + { + "epoch": 0.54, + "grad_norm": 0.478515625, + "learning_rate": 0.00016670376001517225, + "loss": 2.6334, + "step": 12631 + }, + { + "epoch": 0.54, + "grad_norm": 0.421875, + "learning_rate": 0.00016669876229476573, + "loss": 2.583, + "step": 12632 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.00016669376427423912, + "loss": 2.5722, + "step": 12633 + }, + { + "epoch": 0.54, + "grad_norm": 0.4609375, + "learning_rate": 0.00016668876595361497, + "loss": 2.6244, + "step": 12634 + }, + { + "epoch": 0.54, + "grad_norm": 0.50390625, + "learning_rate": 0.00016668376733291576, + "loss": 2.6138, + "step": 12635 + }, + { + "epoch": 0.54, + "grad_norm": 0.50390625, + "learning_rate": 0.00016667876841216397, + "loss": 2.5706, + "step": 12636 + }, + { + "epoch": 0.54, + "grad_norm": 0.484375, + "learning_rate": 0.00016667376919138205, + "loss": 2.675, + "step": 12637 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.00016666876967059256, + "loss": 2.6904, + "step": 12638 + }, + { + "epoch": 0.54, + "grad_norm": 0.494140625, + "learning_rate": 0.00016666376984981795, + "loss": 2.3895, + "step": 12639 + }, + { + "epoch": 0.54, + "grad_norm": 0.5078125, + "learning_rate": 0.0001666587697290808, + "loss": 2.5693, + "step": 12640 + }, + { + "epoch": 0.54, + "grad_norm": 0.51171875, + "learning_rate": 0.00016665376930840347, + "loss": 2.6696, + "step": 12641 + }, + { + "epoch": 0.54, + "grad_norm": 0.443359375, + "learning_rate": 0.0001666487685878086, + "loss": 2.6336, + "step": 12642 + }, + { + "epoch": 0.54, + "grad_norm": 0.52734375, + "learning_rate": 0.00016664376756731858, + "loss": 2.6597, + "step": 12643 + }, + { + "epoch": 0.54, + "grad_norm": 0.439453125, + "learning_rate": 0.000166638766246956, + "loss": 2.5575, + "step": 12644 + }, + { + "epoch": 0.54, + "grad_norm": 0.447265625, + "learning_rate": 0.00016663376462674328, + "loss": 2.598, + "step": 12645 + }, + { + "epoch": 0.54, + "grad_norm": 0.455078125, + "learning_rate": 0.000166628762706703, + "loss": 2.5526, + "step": 12646 + }, + { + "epoch": 0.54, + "grad_norm": 0.4453125, + "learning_rate": 0.00016662376048685764, + "loss": 2.5746, + "step": 12647 + }, + { + "epoch": 0.54, + "grad_norm": 0.435546875, + "learning_rate": 0.00016661875796722972, + "loss": 2.5084, + "step": 12648 + }, + { + "epoch": 0.54, + "grad_norm": 0.44921875, + "learning_rate": 0.00016661375514784172, + "loss": 2.6987, + "step": 12649 + }, + { + "epoch": 0.54, + "grad_norm": 0.51171875, + "learning_rate": 0.00016660875202871616, + "loss": 2.6472, + "step": 12650 + }, + { + "epoch": 0.54, + "grad_norm": 0.4296875, + "learning_rate": 0.00016660374860987556, + "loss": 2.5464, + "step": 12651 + }, + { + "epoch": 0.54, + "grad_norm": 0.5, + "learning_rate": 0.00016659874489134243, + "loss": 2.6787, + "step": 12652 + }, + { + "epoch": 0.54, + "grad_norm": 0.466796875, + "learning_rate": 0.00016659374087313927, + "loss": 2.7225, + "step": 12653 + }, + { + "epoch": 0.54, + "grad_norm": 0.482421875, + "learning_rate": 0.00016658873655528867, + "loss": 2.6391, + "step": 12654 + }, + { + "epoch": 0.54, + "grad_norm": 0.490234375, + "learning_rate": 0.00016658373193781305, + "loss": 2.7541, + "step": 12655 + }, + { + "epoch": 0.54, + "grad_norm": 0.482421875, + "learning_rate": 0.00016657872702073497, + "loss": 2.7257, + "step": 12656 + }, + { + "epoch": 0.54, + "grad_norm": 0.439453125, + "learning_rate": 0.00016657372180407695, + "loss": 2.7403, + "step": 12657 + }, + { + "epoch": 0.54, + "grad_norm": 0.4375, + "learning_rate": 0.00016656871628786154, + "loss": 2.6081, + "step": 12658 + }, + { + "epoch": 0.54, + "grad_norm": 0.51953125, + "learning_rate": 0.0001665637104721112, + "loss": 2.7963, + "step": 12659 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.00016655870435684854, + "loss": 2.4951, + "step": 12660 + }, + { + "epoch": 0.54, + "grad_norm": 0.423828125, + "learning_rate": 0.00016655369794209597, + "loss": 2.583, + "step": 12661 + }, + { + "epoch": 0.54, + "grad_norm": 0.431640625, + "learning_rate": 0.0001665486912278761, + "loss": 2.6811, + "step": 12662 + }, + { + "epoch": 0.54, + "grad_norm": 0.5078125, + "learning_rate": 0.00016654368421421145, + "loss": 2.7195, + "step": 12663 + }, + { + "epoch": 0.54, + "grad_norm": 0.470703125, + "learning_rate": 0.00016653867690112455, + "loss": 2.5212, + "step": 12664 + }, + { + "epoch": 0.54, + "grad_norm": 0.48046875, + "learning_rate": 0.0001665336692886379, + "loss": 2.681, + "step": 12665 + }, + { + "epoch": 0.54, + "grad_norm": 0.49609375, + "learning_rate": 0.00016652866137677408, + "loss": 2.6798, + "step": 12666 + }, + { + "epoch": 0.54, + "grad_norm": 0.400390625, + "learning_rate": 0.00016652365316555557, + "loss": 2.6082, + "step": 12667 + }, + { + "epoch": 0.54, + "grad_norm": 0.494140625, + "learning_rate": 0.00016651864465500494, + "loss": 2.5911, + "step": 12668 + }, + { + "epoch": 0.54, + "grad_norm": 0.453125, + "learning_rate": 0.00016651363584514473, + "loss": 2.5892, + "step": 12669 + }, + { + "epoch": 0.54, + "grad_norm": 0.49609375, + "learning_rate": 0.00016650862673599746, + "loss": 2.5155, + "step": 12670 + }, + { + "epoch": 0.54, + "grad_norm": 0.50390625, + "learning_rate": 0.00016650361732758567, + "loss": 2.6911, + "step": 12671 + }, + { + "epoch": 0.54, + "grad_norm": 0.408203125, + "learning_rate": 0.0001664986076199319, + "loss": 2.5486, + "step": 12672 + }, + { + "epoch": 0.54, + "grad_norm": 0.453125, + "learning_rate": 0.0001664935976130587, + "loss": 2.4465, + "step": 12673 + }, + { + "epoch": 0.54, + "grad_norm": 0.447265625, + "learning_rate": 0.00016648858730698861, + "loss": 2.6788, + "step": 12674 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.0001664835767017442, + "loss": 2.6159, + "step": 12675 + }, + { + "epoch": 0.54, + "grad_norm": 0.50390625, + "learning_rate": 0.000166478565797348, + "loss": 2.5268, + "step": 12676 + }, + { + "epoch": 0.54, + "grad_norm": 0.498046875, + "learning_rate": 0.00016647355459382252, + "loss": 2.5025, + "step": 12677 + }, + { + "epoch": 0.54, + "grad_norm": 0.439453125, + "learning_rate": 0.00016646854309119037, + "loss": 2.6436, + "step": 12678 + }, + { + "epoch": 0.54, + "grad_norm": 0.484375, + "learning_rate": 0.000166463531289474, + "loss": 2.6955, + "step": 12679 + }, + { + "epoch": 0.54, + "grad_norm": 0.453125, + "learning_rate": 0.00016645851918869607, + "loss": 2.6436, + "step": 12680 + }, + { + "epoch": 0.54, + "grad_norm": 0.484375, + "learning_rate": 0.0001664535067888791, + "loss": 2.6529, + "step": 12681 + }, + { + "epoch": 0.54, + "grad_norm": 0.46484375, + "learning_rate": 0.00016644849409004563, + "loss": 2.5931, + "step": 12682 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.00016644348109221823, + "loss": 2.6962, + "step": 12683 + }, + { + "epoch": 0.54, + "grad_norm": 0.40625, + "learning_rate": 0.00016643846779541943, + "loss": 2.4569, + "step": 12684 + }, + { + "epoch": 0.54, + "grad_norm": 0.4140625, + "learning_rate": 0.0001664334541996718, + "loss": 2.5349, + "step": 12685 + }, + { + "epoch": 0.54, + "grad_norm": 0.478515625, + "learning_rate": 0.0001664284403049979, + "loss": 2.5961, + "step": 12686 + }, + { + "epoch": 0.54, + "grad_norm": 0.55859375, + "learning_rate": 0.00016642342611142033, + "loss": 2.6601, + "step": 12687 + }, + { + "epoch": 0.54, + "grad_norm": 0.474609375, + "learning_rate": 0.00016641841161896162, + "loss": 2.5441, + "step": 12688 + }, + { + "epoch": 0.54, + "grad_norm": 0.478515625, + "learning_rate": 0.0001664133968276443, + "loss": 2.5915, + "step": 12689 + }, + { + "epoch": 0.54, + "grad_norm": 0.443359375, + "learning_rate": 0.00016640838173749098, + "loss": 2.568, + "step": 12690 + }, + { + "epoch": 0.54, + "grad_norm": 0.4296875, + "learning_rate": 0.0001664033663485242, + "loss": 2.448, + "step": 12691 + }, + { + "epoch": 0.54, + "grad_norm": 0.451171875, + "learning_rate": 0.00016639835066076657, + "loss": 2.6017, + "step": 12692 + }, + { + "epoch": 0.54, + "grad_norm": 0.45703125, + "learning_rate": 0.00016639333467424058, + "loss": 2.6, + "step": 12693 + }, + { + "epoch": 0.54, + "grad_norm": 0.38671875, + "learning_rate": 0.00016638831838896887, + "loss": 2.5472, + "step": 12694 + }, + { + "epoch": 0.54, + "grad_norm": 0.4140625, + "learning_rate": 0.00016638330180497398, + "loss": 2.5126, + "step": 12695 + }, + { + "epoch": 0.54, + "grad_norm": 0.427734375, + "learning_rate": 0.00016637828492227855, + "loss": 2.5172, + "step": 12696 + }, + { + "epoch": 0.54, + "grad_norm": 0.39453125, + "learning_rate": 0.00016637326774090504, + "loss": 2.4292, + "step": 12697 + }, + { + "epoch": 0.54, + "grad_norm": 0.431640625, + "learning_rate": 0.00016636825026087606, + "loss": 2.6803, + "step": 12698 + }, + { + "epoch": 0.54, + "grad_norm": 0.482421875, + "learning_rate": 0.00016636323248221424, + "loss": 2.6183, + "step": 12699 + }, + { + "epoch": 0.54, + "grad_norm": 0.42578125, + "learning_rate": 0.00016635821440494215, + "loss": 2.7086, + "step": 12700 + }, + { + "epoch": 0.54, + "grad_norm": 0.48828125, + "learning_rate": 0.00016635319602908231, + "loss": 2.5724, + "step": 12701 + }, + { + "epoch": 0.54, + "grad_norm": 0.50390625, + "learning_rate": 0.00016634817735465735, + "loss": 2.7172, + "step": 12702 + }, + { + "epoch": 0.54, + "grad_norm": 0.482421875, + "learning_rate": 0.00016634315838168985, + "loss": 2.6579, + "step": 12703 + }, + { + "epoch": 0.54, + "grad_norm": 0.51953125, + "learning_rate": 0.00016633813911020237, + "loss": 2.6017, + "step": 12704 + }, + { + "epoch": 0.54, + "grad_norm": 0.50390625, + "learning_rate": 0.0001663331195402175, + "loss": 2.6591, + "step": 12705 + }, + { + "epoch": 0.54, + "grad_norm": 0.4609375, + "learning_rate": 0.00016632809967175786, + "loss": 2.6536, + "step": 12706 + }, + { + "epoch": 0.54, + "grad_norm": 0.388671875, + "learning_rate": 0.000166323079504846, + "loss": 2.5608, + "step": 12707 + }, + { + "epoch": 0.54, + "grad_norm": 0.48046875, + "learning_rate": 0.0001663180590395045, + "loss": 2.5467, + "step": 12708 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.00016631303827575598, + "loss": 2.5089, + "step": 12709 + }, + { + "epoch": 0.54, + "grad_norm": 0.474609375, + "learning_rate": 0.00016630801721362304, + "loss": 2.5604, + "step": 12710 + }, + { + "epoch": 0.54, + "grad_norm": 0.48046875, + "learning_rate": 0.00016630299585312824, + "loss": 2.4692, + "step": 12711 + }, + { + "epoch": 0.54, + "grad_norm": 0.451171875, + "learning_rate": 0.00016629797419429423, + "loss": 2.6823, + "step": 12712 + }, + { + "epoch": 0.54, + "grad_norm": 0.4921875, + "learning_rate": 0.0001662929522371435, + "loss": 2.6945, + "step": 12713 + }, + { + "epoch": 0.54, + "grad_norm": 0.46484375, + "learning_rate": 0.00016628792998169876, + "loss": 2.4823, + "step": 12714 + }, + { + "epoch": 0.54, + "grad_norm": 0.4921875, + "learning_rate": 0.00016628290742798255, + "loss": 2.6621, + "step": 12715 + }, + { + "epoch": 0.54, + "grad_norm": 0.443359375, + "learning_rate": 0.00016627788457601746, + "loss": 2.6835, + "step": 12716 + }, + { + "epoch": 0.54, + "grad_norm": 0.416015625, + "learning_rate": 0.00016627286142582612, + "loss": 2.5484, + "step": 12717 + }, + { + "epoch": 0.54, + "grad_norm": 0.48046875, + "learning_rate": 0.00016626783797743112, + "loss": 2.5996, + "step": 12718 + }, + { + "epoch": 0.54, + "grad_norm": 0.45703125, + "learning_rate": 0.0001662628142308551, + "loss": 2.6354, + "step": 12719 + }, + { + "epoch": 0.54, + "grad_norm": 0.4765625, + "learning_rate": 0.0001662577901861206, + "loss": 2.6477, + "step": 12720 + }, + { + "epoch": 0.54, + "grad_norm": 0.431640625, + "learning_rate": 0.00016625276584325026, + "loss": 2.7031, + "step": 12721 + }, + { + "epoch": 0.54, + "grad_norm": 0.470703125, + "learning_rate": 0.00016624774120226673, + "loss": 2.5187, + "step": 12722 + }, + { + "epoch": 0.54, + "grad_norm": 0.4921875, + "learning_rate": 0.00016624271626319251, + "loss": 2.6916, + "step": 12723 + }, + { + "epoch": 0.54, + "grad_norm": 0.453125, + "learning_rate": 0.0001662376910260503, + "loss": 2.5921, + "step": 12724 + }, + { + "epoch": 0.54, + "grad_norm": 0.50390625, + "learning_rate": 0.0001662326654908627, + "loss": 2.6966, + "step": 12725 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.0001662276396576523, + "loss": 2.5535, + "step": 12726 + }, + { + "epoch": 0.54, + "grad_norm": 0.44921875, + "learning_rate": 0.00016622261352644172, + "loss": 2.7103, + "step": 12727 + }, + { + "epoch": 0.54, + "grad_norm": 0.451171875, + "learning_rate": 0.0001662175870972536, + "loss": 2.6884, + "step": 12728 + }, + { + "epoch": 0.54, + "grad_norm": 0.5, + "learning_rate": 0.00016621256037011053, + "loss": 2.6351, + "step": 12729 + }, + { + "epoch": 0.54, + "grad_norm": 0.439453125, + "learning_rate": 0.0001662075333450351, + "loss": 2.5426, + "step": 12730 + }, + { + "epoch": 0.54, + "grad_norm": 0.486328125, + "learning_rate": 0.00016620250602205003, + "loss": 2.6157, + "step": 12731 + }, + { + "epoch": 0.54, + "grad_norm": 0.5, + "learning_rate": 0.00016619747840117784, + "loss": 2.552, + "step": 12732 + }, + { + "epoch": 0.54, + "grad_norm": 0.5078125, + "learning_rate": 0.0001661924504824412, + "loss": 2.8696, + "step": 12733 + }, + { + "epoch": 0.54, + "grad_norm": 0.48828125, + "learning_rate": 0.0001661874222658627, + "loss": 2.6398, + "step": 12734 + }, + { + "epoch": 0.54, + "grad_norm": 0.41796875, + "learning_rate": 0.00016618239375146502, + "loss": 2.5401, + "step": 12735 + }, + { + "epoch": 0.54, + "grad_norm": 0.4375, + "learning_rate": 0.00016617736493927074, + "loss": 2.4961, + "step": 12736 + }, + { + "epoch": 0.54, + "grad_norm": 0.5078125, + "learning_rate": 0.00016617233582930248, + "loss": 2.6223, + "step": 12737 + }, + { + "epoch": 0.54, + "grad_norm": 0.44140625, + "learning_rate": 0.00016616730642158293, + "loss": 2.617, + "step": 12738 + }, + { + "epoch": 0.54, + "grad_norm": 0.458984375, + "learning_rate": 0.00016616227671613468, + "loss": 2.7305, + "step": 12739 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.00016615724671298034, + "loss": 2.5632, + "step": 12740 + }, + { + "epoch": 0.54, + "grad_norm": 0.515625, + "learning_rate": 0.0001661522164121426, + "loss": 2.6145, + "step": 12741 + }, + { + "epoch": 0.54, + "grad_norm": 0.439453125, + "learning_rate": 0.00016614718581364405, + "loss": 2.6083, + "step": 12742 + }, + { + "epoch": 0.54, + "grad_norm": 0.546875, + "learning_rate": 0.00016614215491750733, + "loss": 2.6242, + "step": 12743 + }, + { + "epoch": 0.54, + "grad_norm": 0.451171875, + "learning_rate": 0.00016613712372375508, + "loss": 2.529, + "step": 12744 + }, + { + "epoch": 0.54, + "grad_norm": 0.51171875, + "learning_rate": 0.00016613209223240993, + "loss": 2.6625, + "step": 12745 + }, + { + "epoch": 0.54, + "grad_norm": 0.486328125, + "learning_rate": 0.00016612706044349455, + "loss": 2.7239, + "step": 12746 + }, + { + "epoch": 0.54, + "grad_norm": 0.453125, + "learning_rate": 0.0001661220283570316, + "loss": 2.5033, + "step": 12747 + }, + { + "epoch": 0.54, + "grad_norm": 0.466796875, + "learning_rate": 0.00016611699597304363, + "loss": 2.5204, + "step": 12748 + }, + { + "epoch": 0.54, + "grad_norm": 0.482421875, + "learning_rate": 0.00016611196329155336, + "loss": 2.6192, + "step": 12749 + }, + { + "epoch": 0.54, + "grad_norm": 0.435546875, + "learning_rate": 0.0001661069303125834, + "loss": 2.5327, + "step": 12750 + }, + { + "epoch": 0.54, + "grad_norm": 0.43359375, + "learning_rate": 0.0001661018970361565, + "loss": 2.427, + "step": 12751 + }, + { + "epoch": 0.54, + "grad_norm": 0.466796875, + "learning_rate": 0.00016609686346229512, + "loss": 2.6131, + "step": 12752 + }, + { + "epoch": 0.54, + "grad_norm": 0.486328125, + "learning_rate": 0.00016609182959102206, + "loss": 2.5561, + "step": 12753 + }, + { + "epoch": 0.54, + "grad_norm": 0.51953125, + "learning_rate": 0.0001660867954223599, + "loss": 2.607, + "step": 12754 + }, + { + "epoch": 0.54, + "grad_norm": 0.427734375, + "learning_rate": 0.0001660817609563313, + "loss": 2.5599, + "step": 12755 + }, + { + "epoch": 0.54, + "grad_norm": 0.453125, + "learning_rate": 0.00016607672619295896, + "loss": 2.5295, + "step": 12756 + }, + { + "epoch": 0.54, + "grad_norm": 0.515625, + "learning_rate": 0.0001660716911322655, + "loss": 2.5671, + "step": 12757 + }, + { + "epoch": 0.54, + "grad_norm": 0.46875, + "learning_rate": 0.00016606665577427354, + "loss": 2.4958, + "step": 12758 + }, + { + "epoch": 0.54, + "grad_norm": 0.466796875, + "learning_rate": 0.00016606162011900582, + "loss": 2.5052, + "step": 12759 + }, + { + "epoch": 0.54, + "grad_norm": 0.443359375, + "learning_rate": 0.00016605658416648492, + "loss": 2.6515, + "step": 12760 + }, + { + "epoch": 0.54, + "grad_norm": 0.423828125, + "learning_rate": 0.00016605154791673354, + "loss": 2.5511, + "step": 12761 + }, + { + "epoch": 0.54, + "grad_norm": 0.484375, + "learning_rate": 0.0001660465113697743, + "loss": 2.6976, + "step": 12762 + }, + { + "epoch": 0.54, + "grad_norm": 0.484375, + "learning_rate": 0.00016604147452562996, + "loss": 2.6208, + "step": 12763 + }, + { + "epoch": 0.54, + "grad_norm": 0.5, + "learning_rate": 0.0001660364373843231, + "loss": 2.6083, + "step": 12764 + }, + { + "epoch": 0.54, + "grad_norm": 0.427734375, + "learning_rate": 0.00016603139994587638, + "loss": 2.6705, + "step": 12765 + }, + { + "epoch": 0.54, + "grad_norm": 0.5078125, + "learning_rate": 0.00016602636221031252, + "loss": 2.5469, + "step": 12766 + }, + { + "epoch": 0.54, + "grad_norm": 0.51171875, + "learning_rate": 0.00016602132417765412, + "loss": 2.6369, + "step": 12767 + }, + { + "epoch": 0.54, + "grad_norm": 0.45703125, + "learning_rate": 0.00016601628584792392, + "loss": 2.6049, + "step": 12768 + }, + { + "epoch": 0.54, + "grad_norm": 0.48828125, + "learning_rate": 0.00016601124722114455, + "loss": 2.6753, + "step": 12769 + }, + { + "epoch": 0.54, + "grad_norm": 0.494140625, + "learning_rate": 0.0001660062082973387, + "loss": 2.532, + "step": 12770 + }, + { + "epoch": 0.54, + "grad_norm": 0.40625, + "learning_rate": 0.00016600116907652905, + "loss": 2.5564, + "step": 12771 + }, + { + "epoch": 0.54, + "grad_norm": 0.478515625, + "learning_rate": 0.0001659961295587382, + "loss": 2.6167, + "step": 12772 + }, + { + "epoch": 0.54, + "grad_norm": 0.51953125, + "learning_rate": 0.00016599108974398895, + "loss": 2.7053, + "step": 12773 + }, + { + "epoch": 0.54, + "grad_norm": 0.453125, + "learning_rate": 0.00016598604963230387, + "loss": 2.5793, + "step": 12774 + }, + { + "epoch": 0.54, + "grad_norm": 0.46484375, + "learning_rate": 0.00016598100922370575, + "loss": 2.5388, + "step": 12775 + }, + { + "epoch": 0.54, + "grad_norm": 0.447265625, + "learning_rate": 0.00016597596851821713, + "loss": 2.7638, + "step": 12776 + }, + { + "epoch": 0.54, + "grad_norm": 0.4921875, + "learning_rate": 0.0001659709275158608, + "loss": 2.6603, + "step": 12777 + }, + { + "epoch": 0.54, + "grad_norm": 0.482421875, + "learning_rate": 0.0001659658862166594, + "loss": 2.6274, + "step": 12778 + }, + { + "epoch": 0.54, + "grad_norm": 0.42578125, + "learning_rate": 0.0001659608446206356, + "loss": 2.5558, + "step": 12779 + }, + { + "epoch": 0.54, + "grad_norm": 0.458984375, + "learning_rate": 0.00016595580272781214, + "loss": 2.5778, + "step": 12780 + }, + { + "epoch": 0.54, + "grad_norm": 0.455078125, + "learning_rate": 0.00016595076053821164, + "loss": 2.6965, + "step": 12781 + }, + { + "epoch": 0.54, + "grad_norm": 0.4609375, + "learning_rate": 0.00016594571805185682, + "loss": 2.6643, + "step": 12782 + }, + { + "epoch": 0.54, + "grad_norm": 0.447265625, + "learning_rate": 0.0001659406752687704, + "loss": 2.6181, + "step": 12783 + }, + { + "epoch": 0.54, + "grad_norm": 0.388671875, + "learning_rate": 0.000165935632188975, + "loss": 2.5471, + "step": 12784 + }, + { + "epoch": 0.54, + "grad_norm": 0.427734375, + "learning_rate": 0.0001659305888124934, + "loss": 2.5004, + "step": 12785 + }, + { + "epoch": 0.54, + "grad_norm": 0.462890625, + "learning_rate": 0.0001659255451393482, + "loss": 2.6084, + "step": 12786 + }, + { + "epoch": 0.54, + "grad_norm": 0.3984375, + "learning_rate": 0.00016592050116956216, + "loss": 2.4831, + "step": 12787 + }, + { + "epoch": 0.54, + "grad_norm": 0.4765625, + "learning_rate": 0.00016591545690315795, + "loss": 2.5461, + "step": 12788 + }, + { + "epoch": 0.54, + "grad_norm": 0.47265625, + "learning_rate": 0.00016591041234015826, + "loss": 2.5919, + "step": 12789 + }, + { + "epoch": 0.55, + "grad_norm": 0.443359375, + "learning_rate": 0.00016590536748058583, + "loss": 2.5615, + "step": 12790 + }, + { + "epoch": 0.55, + "grad_norm": 0.482421875, + "learning_rate": 0.00016590032232446334, + "loss": 2.5849, + "step": 12791 + }, + { + "epoch": 0.55, + "grad_norm": 0.4453125, + "learning_rate": 0.00016589527687181342, + "loss": 2.5577, + "step": 12792 + }, + { + "epoch": 0.55, + "grad_norm": 0.474609375, + "learning_rate": 0.00016589023112265888, + "loss": 2.8095, + "step": 12793 + }, + { + "epoch": 0.55, + "grad_norm": 0.4921875, + "learning_rate": 0.0001658851850770224, + "loss": 2.6826, + "step": 12794 + }, + { + "epoch": 0.55, + "grad_norm": 0.53515625, + "learning_rate": 0.00016588013873492665, + "loss": 2.7081, + "step": 12795 + }, + { + "epoch": 0.55, + "grad_norm": 0.435546875, + "learning_rate": 0.00016587509209639435, + "loss": 2.6399, + "step": 12796 + }, + { + "epoch": 0.55, + "grad_norm": 0.5078125, + "learning_rate": 0.00016587004516144819, + "loss": 2.6765, + "step": 12797 + }, + { + "epoch": 0.55, + "grad_norm": 0.4609375, + "learning_rate": 0.0001658649979301109, + "loss": 2.6678, + "step": 12798 + }, + { + "epoch": 0.55, + "grad_norm": 0.453125, + "learning_rate": 0.00016585995040240524, + "loss": 2.6307, + "step": 12799 + }, + { + "epoch": 0.55, + "grad_norm": 0.458984375, + "learning_rate": 0.00016585490257835383, + "loss": 2.6039, + "step": 12800 + }, + { + "epoch": 0.55, + "grad_norm": 0.52734375, + "learning_rate": 0.00016584985445797945, + "loss": 2.6354, + "step": 12801 + }, + { + "epoch": 0.55, + "grad_norm": 0.51953125, + "learning_rate": 0.00016584480604130475, + "loss": 2.5815, + "step": 12802 + }, + { + "epoch": 0.55, + "grad_norm": 0.431640625, + "learning_rate": 0.00016583975732835252, + "loss": 2.5897, + "step": 12803 + }, + { + "epoch": 0.55, + "grad_norm": 0.482421875, + "learning_rate": 0.00016583470831914543, + "loss": 2.6594, + "step": 12804 + }, + { + "epoch": 0.55, + "grad_norm": 0.44140625, + "learning_rate": 0.00016582965901370625, + "loss": 2.6059, + "step": 12805 + }, + { + "epoch": 0.55, + "grad_norm": 0.474609375, + "learning_rate": 0.0001658246094120576, + "loss": 2.672, + "step": 12806 + }, + { + "epoch": 0.55, + "grad_norm": 0.435546875, + "learning_rate": 0.0001658195595142223, + "loss": 2.5968, + "step": 12807 + }, + { + "epoch": 0.55, + "grad_norm": 0.474609375, + "learning_rate": 0.000165814509320223, + "loss": 2.5877, + "step": 12808 + }, + { + "epoch": 0.55, + "grad_norm": 0.490234375, + "learning_rate": 0.00016580945883008251, + "loss": 2.5768, + "step": 12809 + }, + { + "epoch": 0.55, + "grad_norm": 0.455078125, + "learning_rate": 0.0001658044080438235, + "loss": 2.5976, + "step": 12810 + }, + { + "epoch": 0.55, + "grad_norm": 0.51953125, + "learning_rate": 0.0001657993569614687, + "loss": 2.9346, + "step": 12811 + }, + { + "epoch": 0.55, + "grad_norm": 0.408203125, + "learning_rate": 0.0001657943055830408, + "loss": 2.5055, + "step": 12812 + }, + { + "epoch": 0.55, + "grad_norm": 0.490234375, + "learning_rate": 0.0001657892539085626, + "loss": 2.6196, + "step": 12813 + }, + { + "epoch": 0.55, + "grad_norm": 0.50390625, + "learning_rate": 0.00016578420193805678, + "loss": 2.6549, + "step": 12814 + }, + { + "epoch": 0.55, + "grad_norm": 0.45703125, + "learning_rate": 0.00016577914967154614, + "loss": 2.6449, + "step": 12815 + }, + { + "epoch": 0.55, + "grad_norm": 0.421875, + "learning_rate": 0.00016577409710905334, + "loss": 2.5497, + "step": 12816 + }, + { + "epoch": 0.55, + "grad_norm": 0.38671875, + "learning_rate": 0.00016576904425060117, + "loss": 2.6025, + "step": 12817 + }, + { + "epoch": 0.55, + "grad_norm": 0.388671875, + "learning_rate": 0.00016576399109621227, + "loss": 2.5185, + "step": 12818 + }, + { + "epoch": 0.55, + "grad_norm": 0.42578125, + "learning_rate": 0.0001657589376459095, + "loss": 2.5295, + "step": 12819 + }, + { + "epoch": 0.55, + "grad_norm": 0.45703125, + "learning_rate": 0.0001657538838997155, + "loss": 2.7438, + "step": 12820 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.0001657488298576531, + "loss": 2.5169, + "step": 12821 + }, + { + "epoch": 0.55, + "grad_norm": 0.486328125, + "learning_rate": 0.00016574377551974493, + "loss": 2.7091, + "step": 12822 + }, + { + "epoch": 0.55, + "grad_norm": 0.4375, + "learning_rate": 0.00016573872088601385, + "loss": 2.5824, + "step": 12823 + }, + { + "epoch": 0.55, + "grad_norm": 0.51171875, + "learning_rate": 0.00016573366595648255, + "loss": 2.6797, + "step": 12824 + }, + { + "epoch": 0.55, + "grad_norm": 0.439453125, + "learning_rate": 0.00016572861073117377, + "loss": 2.6167, + "step": 12825 + }, + { + "epoch": 0.55, + "grad_norm": 0.419921875, + "learning_rate": 0.00016572355521011025, + "loss": 2.529, + "step": 12826 + }, + { + "epoch": 0.55, + "grad_norm": 0.458984375, + "learning_rate": 0.00016571849939331476, + "loss": 2.6047, + "step": 12827 + }, + { + "epoch": 0.55, + "grad_norm": 0.474609375, + "learning_rate": 0.00016571344328081, + "loss": 2.6442, + "step": 12828 + }, + { + "epoch": 0.55, + "grad_norm": 0.462890625, + "learning_rate": 0.00016570838687261885, + "loss": 2.5903, + "step": 12829 + }, + { + "epoch": 0.55, + "grad_norm": 0.447265625, + "learning_rate": 0.00016570333016876392, + "loss": 2.504, + "step": 12830 + }, + { + "epoch": 0.55, + "grad_norm": 0.45703125, + "learning_rate": 0.000165698273169268, + "loss": 2.6103, + "step": 12831 + }, + { + "epoch": 0.55, + "grad_norm": 0.42578125, + "learning_rate": 0.00016569321587415388, + "loss": 2.5646, + "step": 12832 + }, + { + "epoch": 0.55, + "grad_norm": 0.431640625, + "learning_rate": 0.00016568815828344428, + "loss": 2.5071, + "step": 12833 + }, + { + "epoch": 0.55, + "grad_norm": 0.455078125, + "learning_rate": 0.00016568310039716198, + "loss": 2.3889, + "step": 12834 + }, + { + "epoch": 0.55, + "grad_norm": 0.462890625, + "learning_rate": 0.00016567804221532978, + "loss": 2.6072, + "step": 12835 + }, + { + "epoch": 0.55, + "grad_norm": 0.431640625, + "learning_rate": 0.00016567298373797032, + "loss": 2.5849, + "step": 12836 + }, + { + "epoch": 0.55, + "grad_norm": 0.44140625, + "learning_rate": 0.00016566792496510647, + "loss": 2.5079, + "step": 12837 + }, + { + "epoch": 0.55, + "grad_norm": 0.466796875, + "learning_rate": 0.00016566286589676094, + "loss": 2.5599, + "step": 12838 + }, + { + "epoch": 0.55, + "grad_norm": 0.45703125, + "learning_rate": 0.00016565780653295654, + "loss": 2.6314, + "step": 12839 + }, + { + "epoch": 0.55, + "grad_norm": 0.484375, + "learning_rate": 0.000165652746873716, + "loss": 2.6489, + "step": 12840 + }, + { + "epoch": 0.55, + "grad_norm": 0.43359375, + "learning_rate": 0.0001656476869190621, + "loss": 2.5875, + "step": 12841 + }, + { + "epoch": 0.55, + "grad_norm": 0.484375, + "learning_rate": 0.00016564262666901757, + "loss": 2.6223, + "step": 12842 + }, + { + "epoch": 0.55, + "grad_norm": 0.419921875, + "learning_rate": 0.00016563756612360524, + "loss": 2.6005, + "step": 12843 + }, + { + "epoch": 0.55, + "grad_norm": 0.4609375, + "learning_rate": 0.00016563250528284782, + "loss": 2.5704, + "step": 12844 + }, + { + "epoch": 0.55, + "grad_norm": 0.494140625, + "learning_rate": 0.00016562744414676812, + "loss": 2.6466, + "step": 12845 + }, + { + "epoch": 0.55, + "grad_norm": 0.48828125, + "learning_rate": 0.00016562238271538893, + "loss": 2.6019, + "step": 12846 + }, + { + "epoch": 0.55, + "grad_norm": 0.46875, + "learning_rate": 0.00016561732098873297, + "loss": 2.4345, + "step": 12847 + }, + { + "epoch": 0.55, + "grad_norm": 0.474609375, + "learning_rate": 0.00016561225896682307, + "loss": 2.5111, + "step": 12848 + }, + { + "epoch": 0.55, + "grad_norm": 0.451171875, + "learning_rate": 0.00016560719664968193, + "loss": 2.4028, + "step": 12849 + }, + { + "epoch": 0.55, + "grad_norm": 0.46484375, + "learning_rate": 0.00016560213403733244, + "loss": 2.6362, + "step": 12850 + }, + { + "epoch": 0.55, + "grad_norm": 0.4765625, + "learning_rate": 0.0001655970711297973, + "loss": 2.668, + "step": 12851 + }, + { + "epoch": 0.55, + "grad_norm": 0.462890625, + "learning_rate": 0.0001655920079270993, + "loss": 2.592, + "step": 12852 + }, + { + "epoch": 0.55, + "grad_norm": 0.51953125, + "learning_rate": 0.00016558694442926126, + "loss": 2.4869, + "step": 12853 + }, + { + "epoch": 0.55, + "grad_norm": 0.478515625, + "learning_rate": 0.00016558188063630594, + "loss": 2.589, + "step": 12854 + }, + { + "epoch": 0.55, + "grad_norm": 0.5078125, + "learning_rate": 0.0001655768165482561, + "loss": 2.6235, + "step": 12855 + }, + { + "epoch": 0.55, + "grad_norm": 0.48046875, + "learning_rate": 0.00016557175216513458, + "loss": 2.5589, + "step": 12856 + }, + { + "epoch": 0.55, + "grad_norm": 0.49609375, + "learning_rate": 0.00016556668748696413, + "loss": 2.5877, + "step": 12857 + }, + { + "epoch": 0.55, + "grad_norm": 0.55859375, + "learning_rate": 0.00016556162251376755, + "loss": 2.6925, + "step": 12858 + }, + { + "epoch": 0.55, + "grad_norm": 0.412109375, + "learning_rate": 0.0001655565572455676, + "loss": 2.6122, + "step": 12859 + }, + { + "epoch": 0.55, + "grad_norm": 0.51171875, + "learning_rate": 0.00016555149168238713, + "loss": 2.7141, + "step": 12860 + }, + { + "epoch": 0.55, + "grad_norm": 0.435546875, + "learning_rate": 0.0001655464258242489, + "loss": 2.5615, + "step": 12861 + }, + { + "epoch": 0.55, + "grad_norm": 0.51953125, + "learning_rate": 0.00016554135967117572, + "loss": 2.7982, + "step": 12862 + }, + { + "epoch": 0.55, + "grad_norm": 0.4609375, + "learning_rate": 0.00016553629322319037, + "loss": 2.5491, + "step": 12863 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.00016553122648031563, + "loss": 2.5757, + "step": 12864 + }, + { + "epoch": 0.55, + "grad_norm": 0.50390625, + "learning_rate": 0.0001655261594425743, + "loss": 2.534, + "step": 12865 + }, + { + "epoch": 0.55, + "grad_norm": 0.478515625, + "learning_rate": 0.00016552109210998925, + "loss": 2.6796, + "step": 12866 + }, + { + "epoch": 0.55, + "grad_norm": 0.486328125, + "learning_rate": 0.00016551602448258322, + "loss": 2.7129, + "step": 12867 + }, + { + "epoch": 0.55, + "grad_norm": 0.50390625, + "learning_rate": 0.000165510956560379, + "loss": 2.7273, + "step": 12868 + }, + { + "epoch": 0.55, + "grad_norm": 0.43359375, + "learning_rate": 0.00016550588834339942, + "loss": 2.5712, + "step": 12869 + }, + { + "epoch": 0.55, + "grad_norm": 0.48046875, + "learning_rate": 0.0001655008198316673, + "loss": 2.5954, + "step": 12870 + }, + { + "epoch": 0.55, + "grad_norm": 0.4453125, + "learning_rate": 0.0001654957510252054, + "loss": 2.6212, + "step": 12871 + }, + { + "epoch": 0.55, + "grad_norm": 0.4609375, + "learning_rate": 0.0001654906819240366, + "loss": 2.4592, + "step": 12872 + }, + { + "epoch": 0.55, + "grad_norm": 0.44921875, + "learning_rate": 0.00016548561252818363, + "loss": 2.7046, + "step": 12873 + }, + { + "epoch": 0.55, + "grad_norm": 0.421875, + "learning_rate": 0.00016548054283766937, + "loss": 2.6388, + "step": 12874 + }, + { + "epoch": 0.55, + "grad_norm": 0.5, + "learning_rate": 0.00016547547285251658, + "loss": 2.6368, + "step": 12875 + }, + { + "epoch": 0.55, + "grad_norm": 0.46875, + "learning_rate": 0.00016547040257274809, + "loss": 2.531, + "step": 12876 + }, + { + "epoch": 0.55, + "grad_norm": 0.4453125, + "learning_rate": 0.00016546533199838668, + "loss": 2.759, + "step": 12877 + }, + { + "epoch": 0.55, + "grad_norm": 0.47265625, + "learning_rate": 0.00016546026112945525, + "loss": 2.7063, + "step": 12878 + }, + { + "epoch": 0.55, + "grad_norm": 0.46875, + "learning_rate": 0.00016545518996597655, + "loss": 2.689, + "step": 12879 + }, + { + "epoch": 0.55, + "grad_norm": 0.451171875, + "learning_rate": 0.0001654501185079734, + "loss": 2.7053, + "step": 12880 + }, + { + "epoch": 0.55, + "grad_norm": 0.4296875, + "learning_rate": 0.00016544504675546865, + "loss": 2.7306, + "step": 12881 + }, + { + "epoch": 0.55, + "grad_norm": 0.462890625, + "learning_rate": 0.00016543997470848515, + "loss": 2.6915, + "step": 12882 + }, + { + "epoch": 0.55, + "grad_norm": 0.447265625, + "learning_rate": 0.00016543490236704562, + "loss": 2.6869, + "step": 12883 + }, + { + "epoch": 0.55, + "grad_norm": 0.46875, + "learning_rate": 0.00016542982973117296, + "loss": 2.5503, + "step": 12884 + }, + { + "epoch": 0.55, + "grad_norm": 0.466796875, + "learning_rate": 0.00016542475680088998, + "loss": 2.6065, + "step": 12885 + }, + { + "epoch": 0.55, + "grad_norm": 0.4375, + "learning_rate": 0.0001654196835762195, + "loss": 2.5361, + "step": 12886 + }, + { + "epoch": 0.55, + "grad_norm": 0.5, + "learning_rate": 0.00016541461005718436, + "loss": 2.599, + "step": 12887 + }, + { + "epoch": 0.55, + "grad_norm": 0.5, + "learning_rate": 0.00016540953624380738, + "loss": 2.6265, + "step": 12888 + }, + { + "epoch": 0.55, + "grad_norm": 0.46875, + "learning_rate": 0.00016540446213611138, + "loss": 2.6325, + "step": 12889 + }, + { + "epoch": 0.55, + "grad_norm": 0.431640625, + "learning_rate": 0.00016539938773411922, + "loss": 2.6193, + "step": 12890 + }, + { + "epoch": 0.55, + "grad_norm": 0.50390625, + "learning_rate": 0.00016539431303785372, + "loss": 2.6269, + "step": 12891 + }, + { + "epoch": 0.55, + "grad_norm": 0.46875, + "learning_rate": 0.00016538923804733767, + "loss": 2.3282, + "step": 12892 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.00016538416276259399, + "loss": 2.7096, + "step": 12893 + }, + { + "epoch": 0.55, + "grad_norm": 0.439453125, + "learning_rate": 0.00016537908718364545, + "loss": 2.676, + "step": 12894 + }, + { + "epoch": 0.55, + "grad_norm": 0.41796875, + "learning_rate": 0.00016537401131051492, + "loss": 2.5499, + "step": 12895 + }, + { + "epoch": 0.55, + "grad_norm": 0.4609375, + "learning_rate": 0.00016536893514322522, + "loss": 2.6218, + "step": 12896 + }, + { + "epoch": 0.55, + "grad_norm": 0.490234375, + "learning_rate": 0.00016536385868179922, + "loss": 2.6917, + "step": 12897 + }, + { + "epoch": 0.55, + "grad_norm": 0.46484375, + "learning_rate": 0.00016535878192625974, + "loss": 2.6, + "step": 12898 + }, + { + "epoch": 0.55, + "grad_norm": 0.478515625, + "learning_rate": 0.0001653537048766296, + "loss": 2.6069, + "step": 12899 + }, + { + "epoch": 0.55, + "grad_norm": 0.50390625, + "learning_rate": 0.0001653486275329317, + "loss": 2.6786, + "step": 12900 + }, + { + "epoch": 0.55, + "grad_norm": 0.416015625, + "learning_rate": 0.00016534354989518888, + "loss": 2.4519, + "step": 12901 + }, + { + "epoch": 0.55, + "grad_norm": 0.455078125, + "learning_rate": 0.00016533847196342394, + "loss": 2.6621, + "step": 12902 + }, + { + "epoch": 0.55, + "grad_norm": 0.451171875, + "learning_rate": 0.00016533339373765973, + "loss": 2.5185, + "step": 12903 + }, + { + "epoch": 0.55, + "grad_norm": 0.44140625, + "learning_rate": 0.00016532831521791916, + "loss": 2.4954, + "step": 12904 + }, + { + "epoch": 0.55, + "grad_norm": 0.447265625, + "learning_rate": 0.00016532323640422502, + "loss": 2.6323, + "step": 12905 + }, + { + "epoch": 0.55, + "grad_norm": 0.447265625, + "learning_rate": 0.00016531815729660022, + "loss": 2.6988, + "step": 12906 + }, + { + "epoch": 0.55, + "grad_norm": 0.494140625, + "learning_rate": 0.00016531307789506756, + "loss": 2.7105, + "step": 12907 + }, + { + "epoch": 0.55, + "grad_norm": 0.412109375, + "learning_rate": 0.00016530799819964993, + "loss": 2.5798, + "step": 12908 + }, + { + "epoch": 0.55, + "grad_norm": 0.458984375, + "learning_rate": 0.00016530291821037016, + "loss": 2.618, + "step": 12909 + }, + { + "epoch": 0.55, + "grad_norm": 0.515625, + "learning_rate": 0.00016529783792725114, + "loss": 2.6282, + "step": 12910 + }, + { + "epoch": 0.55, + "grad_norm": 0.455078125, + "learning_rate": 0.0001652927573503157, + "loss": 2.4501, + "step": 12911 + }, + { + "epoch": 0.55, + "grad_norm": 0.47265625, + "learning_rate": 0.00016528767647958672, + "loss": 2.5057, + "step": 12912 + }, + { + "epoch": 0.55, + "grad_norm": 0.50390625, + "learning_rate": 0.00016528259531508704, + "loss": 2.6964, + "step": 12913 + }, + { + "epoch": 0.55, + "grad_norm": 0.458984375, + "learning_rate": 0.00016527751385683956, + "loss": 2.615, + "step": 12914 + }, + { + "epoch": 0.55, + "grad_norm": 0.41015625, + "learning_rate": 0.00016527243210486713, + "loss": 2.5765, + "step": 12915 + }, + { + "epoch": 0.55, + "grad_norm": 0.4921875, + "learning_rate": 0.00016526735005919257, + "loss": 2.7158, + "step": 12916 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.00016526226771983884, + "loss": 2.7092, + "step": 12917 + }, + { + "epoch": 0.55, + "grad_norm": 0.51171875, + "learning_rate": 0.00016525718508682874, + "loss": 2.5574, + "step": 12918 + }, + { + "epoch": 0.55, + "grad_norm": 0.419921875, + "learning_rate": 0.00016525210216018513, + "loss": 2.5929, + "step": 12919 + }, + { + "epoch": 0.55, + "grad_norm": 0.5, + "learning_rate": 0.00016524701893993093, + "loss": 2.7707, + "step": 12920 + }, + { + "epoch": 0.55, + "grad_norm": 0.462890625, + "learning_rate": 0.000165241935426089, + "loss": 2.7121, + "step": 12921 + }, + { + "epoch": 0.55, + "grad_norm": 0.455078125, + "learning_rate": 0.00016523685161868218, + "loss": 2.6424, + "step": 12922 + }, + { + "epoch": 0.55, + "grad_norm": 0.486328125, + "learning_rate": 0.00016523176751773337, + "loss": 2.5788, + "step": 12923 + }, + { + "epoch": 0.55, + "grad_norm": 0.490234375, + "learning_rate": 0.00016522668312326545, + "loss": 2.7511, + "step": 12924 + }, + { + "epoch": 0.55, + "grad_norm": 0.412109375, + "learning_rate": 0.0001652215984353013, + "loss": 2.5312, + "step": 12925 + }, + { + "epoch": 0.55, + "grad_norm": 0.41796875, + "learning_rate": 0.0001652165134538638, + "loss": 2.5878, + "step": 12926 + }, + { + "epoch": 0.55, + "grad_norm": 0.404296875, + "learning_rate": 0.0001652114281789758, + "loss": 2.547, + "step": 12927 + }, + { + "epoch": 0.55, + "grad_norm": 0.400390625, + "learning_rate": 0.0001652063426106602, + "loss": 2.5384, + "step": 12928 + }, + { + "epoch": 0.55, + "grad_norm": 0.408203125, + "learning_rate": 0.00016520125674893992, + "loss": 2.5625, + "step": 12929 + }, + { + "epoch": 0.55, + "grad_norm": 0.419921875, + "learning_rate": 0.0001651961705938378, + "loss": 2.5134, + "step": 12930 + }, + { + "epoch": 0.55, + "grad_norm": 0.42578125, + "learning_rate": 0.00016519108414537675, + "loss": 2.6819, + "step": 12931 + }, + { + "epoch": 0.55, + "grad_norm": 0.37109375, + "learning_rate": 0.00016518599740357966, + "loss": 2.5617, + "step": 12932 + }, + { + "epoch": 0.55, + "grad_norm": 0.45703125, + "learning_rate": 0.00016518091036846934, + "loss": 2.5682, + "step": 12933 + }, + { + "epoch": 0.55, + "grad_norm": 0.484375, + "learning_rate": 0.0001651758230400688, + "loss": 2.5667, + "step": 12934 + }, + { + "epoch": 0.55, + "grad_norm": 0.44140625, + "learning_rate": 0.00016517073541840084, + "loss": 2.4996, + "step": 12935 + }, + { + "epoch": 0.55, + "grad_norm": 0.40234375, + "learning_rate": 0.00016516564750348843, + "loss": 2.6843, + "step": 12936 + }, + { + "epoch": 0.55, + "grad_norm": 0.431640625, + "learning_rate": 0.00016516055929535436, + "loss": 2.5433, + "step": 12937 + }, + { + "epoch": 0.55, + "grad_norm": 0.484375, + "learning_rate": 0.00016515547079402166, + "loss": 2.679, + "step": 12938 + }, + { + "epoch": 0.55, + "grad_norm": 0.498046875, + "learning_rate": 0.0001651503819995131, + "loss": 2.7753, + "step": 12939 + }, + { + "epoch": 0.55, + "grad_norm": 0.482421875, + "learning_rate": 0.00016514529291185165, + "loss": 2.5315, + "step": 12940 + }, + { + "epoch": 0.55, + "grad_norm": 0.486328125, + "learning_rate": 0.00016514020353106018, + "loss": 2.5674, + "step": 12941 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.00016513511385716164, + "loss": 2.5736, + "step": 12942 + }, + { + "epoch": 0.55, + "grad_norm": 0.484375, + "learning_rate": 0.00016513002389017883, + "loss": 2.7019, + "step": 12943 + }, + { + "epoch": 0.55, + "grad_norm": 0.4453125, + "learning_rate": 0.00016512493363013476, + "loss": 2.521, + "step": 12944 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.00016511984307705226, + "loss": 2.6488, + "step": 12945 + }, + { + "epoch": 0.55, + "grad_norm": 0.451171875, + "learning_rate": 0.0001651147522309543, + "loss": 2.7949, + "step": 12946 + }, + { + "epoch": 0.55, + "grad_norm": 0.48828125, + "learning_rate": 0.00016510966109186371, + "loss": 2.7145, + "step": 12947 + }, + { + "epoch": 0.55, + "grad_norm": 0.455078125, + "learning_rate": 0.00016510456965980347, + "loss": 2.6275, + "step": 12948 + }, + { + "epoch": 0.55, + "grad_norm": 0.39453125, + "learning_rate": 0.00016509947793479643, + "loss": 2.562, + "step": 12949 + }, + { + "epoch": 0.55, + "grad_norm": 0.458984375, + "learning_rate": 0.00016509438591686556, + "loss": 2.6319, + "step": 12950 + }, + { + "epoch": 0.55, + "grad_norm": 0.494140625, + "learning_rate": 0.00016508929360603374, + "loss": 2.6429, + "step": 12951 + }, + { + "epoch": 0.55, + "grad_norm": 0.453125, + "learning_rate": 0.00016508420100232386, + "loss": 2.7074, + "step": 12952 + }, + { + "epoch": 0.55, + "grad_norm": 0.431640625, + "learning_rate": 0.0001650791081057589, + "loss": 2.6843, + "step": 12953 + }, + { + "epoch": 0.55, + "grad_norm": 0.474609375, + "learning_rate": 0.0001650740149163617, + "loss": 2.6172, + "step": 12954 + }, + { + "epoch": 0.55, + "grad_norm": 0.439453125, + "learning_rate": 0.00016506892143415525, + "loss": 2.4346, + "step": 12955 + }, + { + "epoch": 0.55, + "grad_norm": 0.4609375, + "learning_rate": 0.0001650638276591624, + "loss": 2.439, + "step": 12956 + }, + { + "epoch": 0.55, + "grad_norm": 0.484375, + "learning_rate": 0.00016505873359140614, + "loss": 2.6628, + "step": 12957 + }, + { + "epoch": 0.55, + "grad_norm": 0.490234375, + "learning_rate": 0.00016505363923090933, + "loss": 2.6096, + "step": 12958 + }, + { + "epoch": 0.55, + "grad_norm": 0.515625, + "learning_rate": 0.00016504854457769492, + "loss": 2.715, + "step": 12959 + }, + { + "epoch": 0.55, + "grad_norm": 0.408203125, + "learning_rate": 0.00016504344963178582, + "loss": 2.6398, + "step": 12960 + }, + { + "epoch": 0.55, + "grad_norm": 0.453125, + "learning_rate": 0.000165038354393205, + "loss": 2.4378, + "step": 12961 + }, + { + "epoch": 0.55, + "grad_norm": 0.494140625, + "learning_rate": 0.00016503325886197532, + "loss": 2.6618, + "step": 12962 + }, + { + "epoch": 0.55, + "grad_norm": 0.51953125, + "learning_rate": 0.00016502816303811977, + "loss": 2.6325, + "step": 12963 + }, + { + "epoch": 0.55, + "grad_norm": 0.458984375, + "learning_rate": 0.00016502306692166123, + "loss": 2.6149, + "step": 12964 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.00016501797051262268, + "loss": 2.652, + "step": 12965 + }, + { + "epoch": 0.55, + "grad_norm": 0.484375, + "learning_rate": 0.00016501287381102698, + "loss": 2.7277, + "step": 12966 + }, + { + "epoch": 0.55, + "grad_norm": 0.51953125, + "learning_rate": 0.00016500777681689713, + "loss": 2.6246, + "step": 12967 + }, + { + "epoch": 0.55, + "grad_norm": 0.408203125, + "learning_rate": 0.00016500267953025604, + "loss": 2.5359, + "step": 12968 + }, + { + "epoch": 0.55, + "grad_norm": 0.462890625, + "learning_rate": 0.0001649975819511267, + "loss": 2.5093, + "step": 12969 + }, + { + "epoch": 0.55, + "grad_norm": 0.462890625, + "learning_rate": 0.00016499248407953192, + "loss": 2.5958, + "step": 12970 + }, + { + "epoch": 0.55, + "grad_norm": 0.482421875, + "learning_rate": 0.00016498738591549475, + "loss": 2.4994, + "step": 12971 + }, + { + "epoch": 0.55, + "grad_norm": 0.490234375, + "learning_rate": 0.00016498228745903806, + "loss": 2.641, + "step": 12972 + }, + { + "epoch": 0.55, + "grad_norm": 0.45703125, + "learning_rate": 0.0001649771887101849, + "loss": 2.6012, + "step": 12973 + }, + { + "epoch": 0.55, + "grad_norm": 0.478515625, + "learning_rate": 0.00016497208966895804, + "loss": 2.6144, + "step": 12974 + }, + { + "epoch": 0.55, + "grad_norm": 0.45703125, + "learning_rate": 0.0001649669903353806, + "loss": 2.4603, + "step": 12975 + }, + { + "epoch": 0.55, + "grad_norm": 0.4453125, + "learning_rate": 0.0001649618907094754, + "loss": 2.508, + "step": 12976 + }, + { + "epoch": 0.55, + "grad_norm": 0.482421875, + "learning_rate": 0.00016495679079126543, + "loss": 2.6249, + "step": 12977 + }, + { + "epoch": 0.55, + "grad_norm": 0.39453125, + "learning_rate": 0.00016495169058077368, + "loss": 2.6494, + "step": 12978 + }, + { + "epoch": 0.55, + "grad_norm": 0.484375, + "learning_rate": 0.00016494659007802303, + "loss": 2.5198, + "step": 12979 + }, + { + "epoch": 0.55, + "grad_norm": 0.49609375, + "learning_rate": 0.00016494148928303644, + "loss": 2.7027, + "step": 12980 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.00016493638819583693, + "loss": 2.6045, + "step": 12981 + }, + { + "epoch": 0.55, + "grad_norm": 0.40625, + "learning_rate": 0.00016493128681644738, + "loss": 2.4949, + "step": 12982 + }, + { + "epoch": 0.55, + "grad_norm": 0.41796875, + "learning_rate": 0.00016492618514489076, + "loss": 2.6009, + "step": 12983 + }, + { + "epoch": 0.55, + "grad_norm": 0.482421875, + "learning_rate": 0.00016492108318119005, + "loss": 2.4781, + "step": 12984 + }, + { + "epoch": 0.55, + "grad_norm": 0.439453125, + "learning_rate": 0.00016491598092536818, + "loss": 2.5306, + "step": 12985 + }, + { + "epoch": 0.55, + "grad_norm": 0.46484375, + "learning_rate": 0.00016491087837744813, + "loss": 2.6656, + "step": 12986 + }, + { + "epoch": 0.55, + "grad_norm": 0.52734375, + "learning_rate": 0.00016490577553745285, + "loss": 2.5644, + "step": 12987 + }, + { + "epoch": 0.55, + "grad_norm": 0.478515625, + "learning_rate": 0.00016490067240540528, + "loss": 2.6179, + "step": 12988 + }, + { + "epoch": 0.55, + "grad_norm": 0.4609375, + "learning_rate": 0.00016489556898132844, + "loss": 2.7003, + "step": 12989 + }, + { + "epoch": 0.55, + "grad_norm": 0.490234375, + "learning_rate": 0.00016489046526524522, + "loss": 2.6653, + "step": 12990 + }, + { + "epoch": 0.55, + "grad_norm": 0.494140625, + "learning_rate": 0.00016488536125717866, + "loss": 2.8001, + "step": 12991 + }, + { + "epoch": 0.55, + "grad_norm": 0.51171875, + "learning_rate": 0.00016488025695715166, + "loss": 2.6247, + "step": 12992 + }, + { + "epoch": 0.55, + "grad_norm": 0.54296875, + "learning_rate": 0.00016487515236518724, + "loss": 2.7268, + "step": 12993 + }, + { + "epoch": 0.55, + "grad_norm": 0.462890625, + "learning_rate": 0.00016487004748130829, + "loss": 2.46, + "step": 12994 + }, + { + "epoch": 0.55, + "grad_norm": 0.470703125, + "learning_rate": 0.00016486494230553785, + "loss": 2.6507, + "step": 12995 + }, + { + "epoch": 0.55, + "grad_norm": 0.44140625, + "learning_rate": 0.0001648598368378989, + "loss": 2.5086, + "step": 12996 + }, + { + "epoch": 0.55, + "grad_norm": 0.45703125, + "learning_rate": 0.0001648547310784144, + "loss": 2.7338, + "step": 12997 + }, + { + "epoch": 0.55, + "grad_norm": 0.5, + "learning_rate": 0.00016484962502710728, + "loss": 2.6353, + "step": 12998 + }, + { + "epoch": 0.55, + "grad_norm": 0.46875, + "learning_rate": 0.00016484451868400057, + "loss": 2.5507, + "step": 12999 + }, + { + "epoch": 0.55, + "grad_norm": 0.52734375, + "learning_rate": 0.0001648394120491172, + "loss": 2.6771, + "step": 13000 + }, + { + "epoch": 0.55, + "grad_norm": 0.447265625, + "learning_rate": 0.0001648343051224802, + "loss": 2.6218, + "step": 13001 + }, + { + "epoch": 0.55, + "grad_norm": 0.5078125, + "learning_rate": 0.0001648291979041125, + "loss": 2.6401, + "step": 13002 + }, + { + "epoch": 0.55, + "grad_norm": 0.48046875, + "learning_rate": 0.0001648240903940371, + "loss": 2.6705, + "step": 13003 + }, + { + "epoch": 0.55, + "grad_norm": 0.4296875, + "learning_rate": 0.00016481898259227697, + "loss": 2.4999, + "step": 13004 + }, + { + "epoch": 0.55, + "grad_norm": 0.44140625, + "learning_rate": 0.00016481387449885518, + "loss": 2.5524, + "step": 13005 + }, + { + "epoch": 0.55, + "grad_norm": 0.447265625, + "learning_rate": 0.00016480876611379457, + "loss": 2.5136, + "step": 13006 + }, + { + "epoch": 0.55, + "grad_norm": 0.49609375, + "learning_rate": 0.00016480365743711821, + "loss": 2.6013, + "step": 13007 + }, + { + "epoch": 0.55, + "grad_norm": 0.4921875, + "learning_rate": 0.00016479854846884905, + "loss": 2.6391, + "step": 13008 + }, + { + "epoch": 0.55, + "grad_norm": 0.435546875, + "learning_rate": 0.00016479343920901016, + "loss": 2.5415, + "step": 13009 + }, + { + "epoch": 0.55, + "grad_norm": 0.57421875, + "learning_rate": 0.00016478832965762444, + "loss": 2.6679, + "step": 13010 + }, + { + "epoch": 0.55, + "grad_norm": 0.515625, + "learning_rate": 0.00016478321981471493, + "loss": 2.6005, + "step": 13011 + }, + { + "epoch": 0.55, + "grad_norm": 0.494140625, + "learning_rate": 0.00016477810968030458, + "loss": 2.4617, + "step": 13012 + }, + { + "epoch": 0.55, + "grad_norm": 0.44140625, + "learning_rate": 0.00016477299925441645, + "loss": 2.5544, + "step": 13013 + }, + { + "epoch": 0.55, + "grad_norm": 0.435546875, + "learning_rate": 0.00016476788853707344, + "loss": 2.5555, + "step": 13014 + }, + { + "epoch": 0.55, + "grad_norm": 0.515625, + "learning_rate": 0.00016476277752829868, + "loss": 2.6672, + "step": 13015 + }, + { + "epoch": 0.55, + "grad_norm": 0.52734375, + "learning_rate": 0.00016475766622811502, + "loss": 2.7713, + "step": 13016 + }, + { + "epoch": 0.55, + "grad_norm": 0.4765625, + "learning_rate": 0.00016475255463654556, + "loss": 2.5518, + "step": 13017 + }, + { + "epoch": 0.55, + "grad_norm": 0.451171875, + "learning_rate": 0.00016474744275361326, + "loss": 2.7086, + "step": 13018 + }, + { + "epoch": 0.55, + "grad_norm": 0.4921875, + "learning_rate": 0.00016474233057934114, + "loss": 2.6094, + "step": 13019 + }, + { + "epoch": 0.55, + "grad_norm": 0.4453125, + "learning_rate": 0.0001647372181137522, + "loss": 2.5838, + "step": 13020 + }, + { + "epoch": 0.55, + "grad_norm": 0.458984375, + "learning_rate": 0.00016473210535686937, + "loss": 2.5845, + "step": 13021 + }, + { + "epoch": 0.55, + "grad_norm": 0.486328125, + "learning_rate": 0.0001647269923087158, + "loss": 2.765, + "step": 13022 + }, + { + "epoch": 0.55, + "grad_norm": 0.4453125, + "learning_rate": 0.0001647218789693144, + "loss": 2.827, + "step": 13023 + }, + { + "epoch": 0.55, + "grad_norm": 0.451171875, + "learning_rate": 0.00016471676533868815, + "loss": 2.5771, + "step": 13024 + }, + { + "epoch": 0.56, + "grad_norm": 0.45703125, + "learning_rate": 0.00016471165141686015, + "loss": 2.5518, + "step": 13025 + }, + { + "epoch": 0.56, + "grad_norm": 0.5, + "learning_rate": 0.00016470653720385337, + "loss": 2.6709, + "step": 13026 + }, + { + "epoch": 0.56, + "grad_norm": 0.435546875, + "learning_rate": 0.00016470142269969085, + "loss": 2.66, + "step": 13027 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.0001646963079043955, + "loss": 2.4685, + "step": 13028 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.00016469119281799048, + "loss": 2.6345, + "step": 13029 + }, + { + "epoch": 0.56, + "grad_norm": 0.427734375, + "learning_rate": 0.0001646860774404987, + "loss": 2.5881, + "step": 13030 + }, + { + "epoch": 0.56, + "grad_norm": 0.427734375, + "learning_rate": 0.0001646809617719432, + "loss": 2.5581, + "step": 13031 + }, + { + "epoch": 0.56, + "grad_norm": 0.416015625, + "learning_rate": 0.00016467584581234703, + "loss": 2.5774, + "step": 13032 + }, + { + "epoch": 0.56, + "grad_norm": 0.4921875, + "learning_rate": 0.00016467072956173316, + "loss": 2.5037, + "step": 13033 + }, + { + "epoch": 0.56, + "grad_norm": 0.447265625, + "learning_rate": 0.0001646656130201247, + "loss": 2.7346, + "step": 13034 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.00016466049618754457, + "loss": 2.5259, + "step": 13035 + }, + { + "epoch": 0.56, + "grad_norm": 0.484375, + "learning_rate": 0.00016465537906401584, + "loss": 2.5676, + "step": 13036 + }, + { + "epoch": 0.56, + "grad_norm": 0.48046875, + "learning_rate": 0.00016465026164956155, + "loss": 2.5971, + "step": 13037 + }, + { + "epoch": 0.56, + "grad_norm": 0.443359375, + "learning_rate": 0.0001646451439442047, + "loss": 2.6342, + "step": 13038 + }, + { + "epoch": 0.56, + "grad_norm": 0.470703125, + "learning_rate": 0.00016464002594796832, + "loss": 2.6633, + "step": 13039 + }, + { + "epoch": 0.56, + "grad_norm": 0.494140625, + "learning_rate": 0.00016463490766087542, + "loss": 2.6588, + "step": 13040 + }, + { + "epoch": 0.56, + "grad_norm": 0.4375, + "learning_rate": 0.00016462978908294911, + "loss": 2.4185, + "step": 13041 + }, + { + "epoch": 0.56, + "grad_norm": 0.474609375, + "learning_rate": 0.00016462467021421238, + "loss": 2.622, + "step": 13042 + }, + { + "epoch": 0.56, + "grad_norm": 0.390625, + "learning_rate": 0.00016461955105468823, + "loss": 2.4965, + "step": 13043 + }, + { + "epoch": 0.56, + "grad_norm": 0.443359375, + "learning_rate": 0.0001646144316043997, + "loss": 2.5632, + "step": 13044 + }, + { + "epoch": 0.56, + "grad_norm": 0.474609375, + "learning_rate": 0.00016460931186336984, + "loss": 2.5663, + "step": 13045 + }, + { + "epoch": 0.56, + "grad_norm": 0.42578125, + "learning_rate": 0.00016460419183162168, + "loss": 2.5554, + "step": 13046 + }, + { + "epoch": 0.56, + "grad_norm": 0.453125, + "learning_rate": 0.0001645990715091783, + "loss": 2.6282, + "step": 13047 + }, + { + "epoch": 0.56, + "grad_norm": 0.458984375, + "learning_rate": 0.0001645939508960627, + "loss": 2.5586, + "step": 13048 + }, + { + "epoch": 0.56, + "grad_norm": 0.44921875, + "learning_rate": 0.0001645888299922979, + "loss": 2.6989, + "step": 13049 + }, + { + "epoch": 0.56, + "grad_norm": 0.419921875, + "learning_rate": 0.00016458370879790697, + "loss": 2.6201, + "step": 13050 + }, + { + "epoch": 0.56, + "grad_norm": 0.462890625, + "learning_rate": 0.000164578587312913, + "loss": 2.5774, + "step": 13051 + }, + { + "epoch": 0.56, + "grad_norm": 0.4140625, + "learning_rate": 0.00016457346553733896, + "loss": 2.701, + "step": 13052 + }, + { + "epoch": 0.56, + "grad_norm": 0.466796875, + "learning_rate": 0.0001645683434712079, + "loss": 2.5676, + "step": 13053 + }, + { + "epoch": 0.56, + "grad_norm": 0.435546875, + "learning_rate": 0.0001645632211145429, + "loss": 2.5184, + "step": 13054 + }, + { + "epoch": 0.56, + "grad_norm": 0.44921875, + "learning_rate": 0.00016455809846736705, + "loss": 2.6077, + "step": 13055 + }, + { + "epoch": 0.56, + "grad_norm": 0.51953125, + "learning_rate": 0.0001645529755297033, + "loss": 2.6975, + "step": 13056 + }, + { + "epoch": 0.56, + "grad_norm": 0.48046875, + "learning_rate": 0.00016454785230157477, + "loss": 2.5692, + "step": 13057 + }, + { + "epoch": 0.56, + "grad_norm": 0.36328125, + "learning_rate": 0.00016454272878300448, + "loss": 2.6073, + "step": 13058 + }, + { + "epoch": 0.56, + "grad_norm": 0.421875, + "learning_rate": 0.00016453760497401552, + "loss": 2.4898, + "step": 13059 + }, + { + "epoch": 0.56, + "grad_norm": 0.52734375, + "learning_rate": 0.00016453248087463092, + "loss": 2.4464, + "step": 13060 + }, + { + "epoch": 0.56, + "grad_norm": 0.439453125, + "learning_rate": 0.00016452735648487372, + "loss": 2.6701, + "step": 13061 + }, + { + "epoch": 0.56, + "grad_norm": 0.482421875, + "learning_rate": 0.000164522231804767, + "loss": 2.5467, + "step": 13062 + }, + { + "epoch": 0.56, + "grad_norm": 0.4765625, + "learning_rate": 0.00016451710683433384, + "loss": 2.5674, + "step": 13063 + }, + { + "epoch": 0.56, + "grad_norm": 0.41796875, + "learning_rate": 0.00016451198157359725, + "loss": 2.5922, + "step": 13064 + }, + { + "epoch": 0.56, + "grad_norm": 0.53125, + "learning_rate": 0.00016450685602258035, + "loss": 2.7049, + "step": 13065 + }, + { + "epoch": 0.56, + "grad_norm": 0.431640625, + "learning_rate": 0.00016450173018130616, + "loss": 2.52, + "step": 13066 + }, + { + "epoch": 0.56, + "grad_norm": 0.40234375, + "learning_rate": 0.00016449660404979774, + "loss": 2.592, + "step": 13067 + }, + { + "epoch": 0.56, + "grad_norm": 0.431640625, + "learning_rate": 0.00016449147762807818, + "loss": 2.5858, + "step": 13068 + }, + { + "epoch": 0.56, + "grad_norm": 0.466796875, + "learning_rate": 0.00016448635091617054, + "loss": 2.6168, + "step": 13069 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.00016448122391409792, + "loss": 2.5206, + "step": 13070 + }, + { + "epoch": 0.56, + "grad_norm": 0.435546875, + "learning_rate": 0.00016447609662188333, + "loss": 2.4889, + "step": 13071 + }, + { + "epoch": 0.56, + "grad_norm": 0.46875, + "learning_rate": 0.00016447096903954983, + "loss": 2.4157, + "step": 13072 + }, + { + "epoch": 0.56, + "grad_norm": 0.5078125, + "learning_rate": 0.00016446584116712058, + "loss": 2.5701, + "step": 13073 + }, + { + "epoch": 0.56, + "grad_norm": 0.5, + "learning_rate": 0.0001644607130046186, + "loss": 2.5285, + "step": 13074 + }, + { + "epoch": 0.56, + "grad_norm": 0.447265625, + "learning_rate": 0.00016445558455206694, + "loss": 2.5448, + "step": 13075 + }, + { + "epoch": 0.56, + "grad_norm": 0.4609375, + "learning_rate": 0.00016445045580948874, + "loss": 2.6043, + "step": 13076 + }, + { + "epoch": 0.56, + "grad_norm": 0.4921875, + "learning_rate": 0.00016444532677690706, + "loss": 2.5294, + "step": 13077 + }, + { + "epoch": 0.56, + "grad_norm": 0.451171875, + "learning_rate": 0.00016444019745434488, + "loss": 2.6247, + "step": 13078 + }, + { + "epoch": 0.56, + "grad_norm": 0.490234375, + "learning_rate": 0.00016443506784182544, + "loss": 2.7049, + "step": 13079 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.0001644299379393717, + "loss": 2.7643, + "step": 13080 + }, + { + "epoch": 0.56, + "grad_norm": 0.421875, + "learning_rate": 0.00016442480774700682, + "loss": 2.6138, + "step": 13081 + }, + { + "epoch": 0.56, + "grad_norm": 0.453125, + "learning_rate": 0.00016441967726475384, + "loss": 2.4497, + "step": 13082 + }, + { + "epoch": 0.56, + "grad_norm": 0.443359375, + "learning_rate": 0.00016441454649263583, + "loss": 2.5829, + "step": 13083 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.00016440941543067593, + "loss": 2.6103, + "step": 13084 + }, + { + "epoch": 0.56, + "grad_norm": 0.44140625, + "learning_rate": 0.0001644042840788972, + "loss": 2.6271, + "step": 13085 + }, + { + "epoch": 0.56, + "grad_norm": 0.474609375, + "learning_rate": 0.00016439915243732274, + "loss": 2.7123, + "step": 13086 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.00016439402050597558, + "loss": 2.8116, + "step": 13087 + }, + { + "epoch": 0.56, + "grad_norm": 0.48046875, + "learning_rate": 0.00016438888828487888, + "loss": 2.6366, + "step": 13088 + }, + { + "epoch": 0.56, + "grad_norm": 0.53125, + "learning_rate": 0.00016438375577405574, + "loss": 2.695, + "step": 13089 + }, + { + "epoch": 0.56, + "grad_norm": 0.47265625, + "learning_rate": 0.00016437862297352923, + "loss": 2.6936, + "step": 13090 + }, + { + "epoch": 0.56, + "grad_norm": 0.5546875, + "learning_rate": 0.0001643734898833224, + "loss": 2.7465, + "step": 13091 + }, + { + "epoch": 0.56, + "grad_norm": 0.44921875, + "learning_rate": 0.00016436835650345842, + "loss": 2.4455, + "step": 13092 + }, + { + "epoch": 0.56, + "grad_norm": 0.4296875, + "learning_rate": 0.00016436322283396035, + "loss": 2.5856, + "step": 13093 + }, + { + "epoch": 0.56, + "grad_norm": 0.392578125, + "learning_rate": 0.0001643580888748513, + "loss": 2.4661, + "step": 13094 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.00016435295462615437, + "loss": 2.575, + "step": 13095 + }, + { + "epoch": 0.56, + "grad_norm": 0.498046875, + "learning_rate": 0.00016434782008789269, + "loss": 2.5674, + "step": 13096 + }, + { + "epoch": 0.56, + "grad_norm": 0.4609375, + "learning_rate": 0.00016434268526008928, + "loss": 2.4285, + "step": 13097 + }, + { + "epoch": 0.56, + "grad_norm": 0.4140625, + "learning_rate": 0.00016433755014276736, + "loss": 2.5439, + "step": 13098 + }, + { + "epoch": 0.56, + "grad_norm": 0.478515625, + "learning_rate": 0.00016433241473594995, + "loss": 2.6714, + "step": 13099 + }, + { + "epoch": 0.56, + "grad_norm": 0.47265625, + "learning_rate": 0.0001643272790396602, + "loss": 2.575, + "step": 13100 + }, + { + "epoch": 0.56, + "grad_norm": 0.439453125, + "learning_rate": 0.00016432214305392117, + "loss": 2.5265, + "step": 13101 + }, + { + "epoch": 0.56, + "grad_norm": 0.44921875, + "learning_rate": 0.00016431700677875603, + "loss": 2.6493, + "step": 13102 + }, + { + "epoch": 0.56, + "grad_norm": 0.455078125, + "learning_rate": 0.00016431187021418787, + "loss": 2.7157, + "step": 13103 + }, + { + "epoch": 0.56, + "grad_norm": 0.39453125, + "learning_rate": 0.00016430673336023977, + "loss": 2.6124, + "step": 13104 + }, + { + "epoch": 0.56, + "grad_norm": 0.4296875, + "learning_rate": 0.00016430159621693487, + "loss": 2.5809, + "step": 13105 + }, + { + "epoch": 0.56, + "grad_norm": 0.435546875, + "learning_rate": 0.00016429645878429633, + "loss": 2.6107, + "step": 13106 + }, + { + "epoch": 0.56, + "grad_norm": 0.49609375, + "learning_rate": 0.0001642913210623472, + "loss": 2.6543, + "step": 13107 + }, + { + "epoch": 0.56, + "grad_norm": 0.49609375, + "learning_rate": 0.0001642861830511106, + "loss": 2.5852, + "step": 13108 + }, + { + "epoch": 0.56, + "grad_norm": 0.47265625, + "learning_rate": 0.0001642810447506097, + "loss": 2.6514, + "step": 13109 + }, + { + "epoch": 0.56, + "grad_norm": 0.48046875, + "learning_rate": 0.00016427590616086758, + "loss": 2.5603, + "step": 13110 + }, + { + "epoch": 0.56, + "grad_norm": 0.482421875, + "learning_rate": 0.00016427076728190736, + "loss": 2.574, + "step": 13111 + }, + { + "epoch": 0.56, + "grad_norm": 0.42578125, + "learning_rate": 0.0001642656281137522, + "loss": 2.4604, + "step": 13112 + }, + { + "epoch": 0.56, + "grad_norm": 0.48046875, + "learning_rate": 0.0001642604886564252, + "loss": 2.6957, + "step": 13113 + }, + { + "epoch": 0.56, + "grad_norm": 0.458984375, + "learning_rate": 0.00016425534890994944, + "loss": 2.5949, + "step": 13114 + }, + { + "epoch": 0.56, + "grad_norm": 0.439453125, + "learning_rate": 0.0001642502088743481, + "loss": 2.6448, + "step": 13115 + }, + { + "epoch": 0.56, + "grad_norm": 0.42578125, + "learning_rate": 0.00016424506854964435, + "loss": 2.6008, + "step": 13116 + }, + { + "epoch": 0.56, + "grad_norm": 0.421875, + "learning_rate": 0.00016423992793586126, + "loss": 2.6884, + "step": 13117 + }, + { + "epoch": 0.56, + "grad_norm": 0.478515625, + "learning_rate": 0.00016423478703302195, + "loss": 2.4347, + "step": 13118 + }, + { + "epoch": 0.56, + "grad_norm": 0.515625, + "learning_rate": 0.00016422964584114958, + "loss": 2.574, + "step": 13119 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.00016422450436026726, + "loss": 2.548, + "step": 13120 + }, + { + "epoch": 0.56, + "grad_norm": 0.455078125, + "learning_rate": 0.00016421936259039817, + "loss": 2.5145, + "step": 13121 + }, + { + "epoch": 0.56, + "grad_norm": 0.46484375, + "learning_rate": 0.0001642142205315654, + "loss": 2.566, + "step": 13122 + }, + { + "epoch": 0.56, + "grad_norm": 0.44921875, + "learning_rate": 0.0001642090781837921, + "loss": 2.662, + "step": 13123 + }, + { + "epoch": 0.56, + "grad_norm": 0.3984375, + "learning_rate": 0.00016420393554710146, + "loss": 2.5897, + "step": 13124 + }, + { + "epoch": 0.56, + "grad_norm": 0.44140625, + "learning_rate": 0.00016419879262151652, + "loss": 2.5958, + "step": 13125 + }, + { + "epoch": 0.56, + "grad_norm": 0.466796875, + "learning_rate": 0.0001641936494070605, + "loss": 2.5373, + "step": 13126 + }, + { + "epoch": 0.56, + "grad_norm": 0.44921875, + "learning_rate": 0.0001641885059037565, + "loss": 2.6454, + "step": 13127 + }, + { + "epoch": 0.56, + "grad_norm": 0.474609375, + "learning_rate": 0.0001641833621116277, + "loss": 2.601, + "step": 13128 + }, + { + "epoch": 0.56, + "grad_norm": 0.47265625, + "learning_rate": 0.00016417821803069721, + "loss": 2.6356, + "step": 13129 + }, + { + "epoch": 0.56, + "grad_norm": 0.44140625, + "learning_rate": 0.0001641730736609882, + "loss": 2.5957, + "step": 13130 + }, + { + "epoch": 0.56, + "grad_norm": 0.482421875, + "learning_rate": 0.00016416792900252384, + "loss": 2.6152, + "step": 13131 + }, + { + "epoch": 0.56, + "grad_norm": 0.4609375, + "learning_rate": 0.0001641627840553272, + "loss": 2.5643, + "step": 13132 + }, + { + "epoch": 0.56, + "grad_norm": 0.4140625, + "learning_rate": 0.00016415763881942152, + "loss": 2.5778, + "step": 13133 + }, + { + "epoch": 0.56, + "grad_norm": 0.474609375, + "learning_rate": 0.0001641524932948299, + "loss": 2.6062, + "step": 13134 + }, + { + "epoch": 0.56, + "grad_norm": 0.49609375, + "learning_rate": 0.0001641473474815755, + "loss": 2.5331, + "step": 13135 + }, + { + "epoch": 0.56, + "grad_norm": 0.4765625, + "learning_rate": 0.00016414220137968147, + "loss": 2.5942, + "step": 13136 + }, + { + "epoch": 0.56, + "grad_norm": 0.453125, + "learning_rate": 0.000164137054989171, + "loss": 2.389, + "step": 13137 + }, + { + "epoch": 0.56, + "grad_norm": 0.498046875, + "learning_rate": 0.00016413190831006722, + "loss": 2.5521, + "step": 13138 + }, + { + "epoch": 0.56, + "grad_norm": 0.419921875, + "learning_rate": 0.00016412676134239327, + "loss": 2.6656, + "step": 13139 + }, + { + "epoch": 0.56, + "grad_norm": 0.5234375, + "learning_rate": 0.00016412161408617236, + "loss": 2.7284, + "step": 13140 + }, + { + "epoch": 0.56, + "grad_norm": 0.41796875, + "learning_rate": 0.00016411646654142757, + "loss": 2.5192, + "step": 13141 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.00016411131870818215, + "loss": 2.6874, + "step": 13142 + }, + { + "epoch": 0.56, + "grad_norm": 0.44140625, + "learning_rate": 0.00016410617058645923, + "loss": 2.5742, + "step": 13143 + }, + { + "epoch": 0.56, + "grad_norm": 0.46484375, + "learning_rate": 0.00016410102217628196, + "loss": 2.5512, + "step": 13144 + }, + { + "epoch": 0.56, + "grad_norm": 0.392578125, + "learning_rate": 0.0001640958734776735, + "loss": 2.476, + "step": 13145 + }, + { + "epoch": 0.56, + "grad_norm": 0.48828125, + "learning_rate": 0.00016409072449065707, + "loss": 2.632, + "step": 13146 + }, + { + "epoch": 0.56, + "grad_norm": 0.462890625, + "learning_rate": 0.00016408557521525578, + "loss": 2.7472, + "step": 13147 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.00016408042565149283, + "loss": 2.587, + "step": 13148 + }, + { + "epoch": 0.56, + "grad_norm": 0.408203125, + "learning_rate": 0.00016407527579939134, + "loss": 2.6612, + "step": 13149 + }, + { + "epoch": 0.56, + "grad_norm": 0.5, + "learning_rate": 0.0001640701256589746, + "loss": 2.6326, + "step": 13150 + }, + { + "epoch": 0.56, + "grad_norm": 0.48828125, + "learning_rate": 0.00016406497523026564, + "loss": 2.6313, + "step": 13151 + }, + { + "epoch": 0.56, + "grad_norm": 0.4296875, + "learning_rate": 0.00016405982451328773, + "loss": 2.4557, + "step": 13152 + }, + { + "epoch": 0.56, + "grad_norm": 0.447265625, + "learning_rate": 0.00016405467350806406, + "loss": 2.584, + "step": 13153 + }, + { + "epoch": 0.56, + "grad_norm": 0.458984375, + "learning_rate": 0.00016404952221461773, + "loss": 2.5769, + "step": 13154 + }, + { + "epoch": 0.56, + "grad_norm": 0.435546875, + "learning_rate": 0.00016404437063297195, + "loss": 2.5747, + "step": 13155 + }, + { + "epoch": 0.56, + "grad_norm": 0.443359375, + "learning_rate": 0.00016403921876314992, + "loss": 2.607, + "step": 13156 + }, + { + "epoch": 0.56, + "grad_norm": 0.439453125, + "learning_rate": 0.00016403406660517482, + "loss": 2.5309, + "step": 13157 + }, + { + "epoch": 0.56, + "grad_norm": 0.48046875, + "learning_rate": 0.0001640289141590698, + "loss": 2.5057, + "step": 13158 + }, + { + "epoch": 0.56, + "grad_norm": 0.50390625, + "learning_rate": 0.0001640237614248581, + "loss": 2.5752, + "step": 13159 + }, + { + "epoch": 0.56, + "grad_norm": 0.423828125, + "learning_rate": 0.00016401860840256285, + "loss": 2.6774, + "step": 13160 + }, + { + "epoch": 0.56, + "grad_norm": 0.451171875, + "learning_rate": 0.00016401345509220724, + "loss": 2.5447, + "step": 13161 + }, + { + "epoch": 0.56, + "grad_norm": 0.419921875, + "learning_rate": 0.00016400830149381453, + "loss": 2.5005, + "step": 13162 + }, + { + "epoch": 0.56, + "grad_norm": 0.435546875, + "learning_rate": 0.0001640031476074078, + "loss": 2.6455, + "step": 13163 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.0001639979934330103, + "loss": 2.5809, + "step": 13164 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.00016399283897064525, + "loss": 2.5866, + "step": 13165 + }, + { + "epoch": 0.56, + "grad_norm": 0.474609375, + "learning_rate": 0.00016398768422033579, + "loss": 2.7012, + "step": 13166 + }, + { + "epoch": 0.56, + "grad_norm": 0.412109375, + "learning_rate": 0.00016398252918210516, + "loss": 2.5174, + "step": 13167 + }, + { + "epoch": 0.56, + "grad_norm": 0.455078125, + "learning_rate": 0.0001639773738559765, + "loss": 2.5334, + "step": 13168 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.00016397221824197304, + "loss": 2.5341, + "step": 13169 + }, + { + "epoch": 0.56, + "grad_norm": 0.453125, + "learning_rate": 0.000163967062340118, + "loss": 2.6337, + "step": 13170 + }, + { + "epoch": 0.56, + "grad_norm": 0.46875, + "learning_rate": 0.00016396190615043454, + "loss": 2.576, + "step": 13171 + }, + { + "epoch": 0.56, + "grad_norm": 0.423828125, + "learning_rate": 0.0001639567496729459, + "loss": 2.5633, + "step": 13172 + }, + { + "epoch": 0.56, + "grad_norm": 0.498046875, + "learning_rate": 0.00016395159290767523, + "loss": 2.6146, + "step": 13173 + }, + { + "epoch": 0.56, + "grad_norm": 0.484375, + "learning_rate": 0.00016394643585464577, + "loss": 2.661, + "step": 13174 + }, + { + "epoch": 0.56, + "grad_norm": 0.42578125, + "learning_rate": 0.00016394127851388074, + "loss": 2.6111, + "step": 13175 + }, + { + "epoch": 0.56, + "grad_norm": 0.4921875, + "learning_rate": 0.0001639361208854033, + "loss": 2.7305, + "step": 13176 + }, + { + "epoch": 0.56, + "grad_norm": 0.46484375, + "learning_rate": 0.0001639309629692367, + "loss": 2.6387, + "step": 13177 + }, + { + "epoch": 0.56, + "grad_norm": 0.4921875, + "learning_rate": 0.00016392580476540412, + "loss": 2.4143, + "step": 13178 + }, + { + "epoch": 0.56, + "grad_norm": 0.47265625, + "learning_rate": 0.00016392064627392878, + "loss": 2.4746, + "step": 13179 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.0001639154874948339, + "loss": 2.8132, + "step": 13180 + }, + { + "epoch": 0.56, + "grad_norm": 0.453125, + "learning_rate": 0.0001639103284281427, + "loss": 2.6045, + "step": 13181 + }, + { + "epoch": 0.56, + "grad_norm": 0.5078125, + "learning_rate": 0.00016390516907387835, + "loss": 2.6127, + "step": 13182 + }, + { + "epoch": 0.56, + "grad_norm": 0.458984375, + "learning_rate": 0.0001639000094320641, + "loss": 2.486, + "step": 13183 + }, + { + "epoch": 0.56, + "grad_norm": 0.482421875, + "learning_rate": 0.00016389484950272317, + "loss": 2.5916, + "step": 13184 + }, + { + "epoch": 0.56, + "grad_norm": 0.494140625, + "learning_rate": 0.00016388968928587874, + "loss": 2.5732, + "step": 13185 + }, + { + "epoch": 0.56, + "grad_norm": 0.482421875, + "learning_rate": 0.0001638845287815541, + "loss": 2.4659, + "step": 13186 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.00016387936798977238, + "loss": 2.5577, + "step": 13187 + }, + { + "epoch": 0.56, + "grad_norm": 0.484375, + "learning_rate": 0.0001638742069105569, + "loss": 2.6894, + "step": 13188 + }, + { + "epoch": 0.56, + "grad_norm": 0.42578125, + "learning_rate": 0.0001638690455439308, + "loss": 2.4814, + "step": 13189 + }, + { + "epoch": 0.56, + "grad_norm": 0.5078125, + "learning_rate": 0.00016386388388991735, + "loss": 2.4286, + "step": 13190 + }, + { + "epoch": 0.56, + "grad_norm": 0.447265625, + "learning_rate": 0.00016385872194853976, + "loss": 2.5542, + "step": 13191 + }, + { + "epoch": 0.56, + "grad_norm": 0.51953125, + "learning_rate": 0.00016385355971982124, + "loss": 2.6106, + "step": 13192 + }, + { + "epoch": 0.56, + "grad_norm": 0.4140625, + "learning_rate": 0.00016384839720378507, + "loss": 2.6682, + "step": 13193 + }, + { + "epoch": 0.56, + "grad_norm": 0.5, + "learning_rate": 0.00016384323440045445, + "loss": 2.612, + "step": 13194 + }, + { + "epoch": 0.56, + "grad_norm": 0.474609375, + "learning_rate": 0.00016383807130985253, + "loss": 2.5951, + "step": 13195 + }, + { + "epoch": 0.56, + "grad_norm": 0.462890625, + "learning_rate": 0.0001638329079320027, + "loss": 2.6517, + "step": 13196 + }, + { + "epoch": 0.56, + "grad_norm": 0.412109375, + "learning_rate": 0.00016382774426692808, + "loss": 2.5672, + "step": 13197 + }, + { + "epoch": 0.56, + "grad_norm": 0.498046875, + "learning_rate": 0.00016382258031465193, + "loss": 2.5954, + "step": 13198 + }, + { + "epoch": 0.56, + "grad_norm": 0.4765625, + "learning_rate": 0.0001638174160751975, + "loss": 2.5469, + "step": 13199 + }, + { + "epoch": 0.56, + "grad_norm": 0.466796875, + "learning_rate": 0.00016381225154858805, + "loss": 2.5343, + "step": 13200 + }, + { + "epoch": 0.56, + "grad_norm": 0.5078125, + "learning_rate": 0.00016380708673484676, + "loss": 2.5587, + "step": 13201 + }, + { + "epoch": 0.56, + "grad_norm": 0.490234375, + "learning_rate": 0.00016380192163399692, + "loss": 2.7263, + "step": 13202 + }, + { + "epoch": 0.56, + "grad_norm": 0.3984375, + "learning_rate": 0.00016379675624606175, + "loss": 2.5377, + "step": 13203 + }, + { + "epoch": 0.56, + "grad_norm": 0.50390625, + "learning_rate": 0.00016379159057106448, + "loss": 2.631, + "step": 13204 + }, + { + "epoch": 0.56, + "grad_norm": 0.4296875, + "learning_rate": 0.00016378642460902835, + "loss": 2.4396, + "step": 13205 + }, + { + "epoch": 0.56, + "grad_norm": 0.45703125, + "learning_rate": 0.0001637812583599766, + "loss": 2.6625, + "step": 13206 + }, + { + "epoch": 0.56, + "grad_norm": 0.40625, + "learning_rate": 0.00016377609182393257, + "loss": 2.49, + "step": 13207 + }, + { + "epoch": 0.56, + "grad_norm": 0.427734375, + "learning_rate": 0.0001637709250009194, + "loss": 2.639, + "step": 13208 + }, + { + "epoch": 0.56, + "grad_norm": 0.412109375, + "learning_rate": 0.00016376575789096037, + "loss": 2.5727, + "step": 13209 + }, + { + "epoch": 0.56, + "grad_norm": 0.431640625, + "learning_rate": 0.0001637605904940787, + "loss": 2.6163, + "step": 13210 + }, + { + "epoch": 0.56, + "grad_norm": 0.4375, + "learning_rate": 0.00016375542281029773, + "loss": 2.7366, + "step": 13211 + }, + { + "epoch": 0.56, + "grad_norm": 0.4140625, + "learning_rate": 0.00016375025483964062, + "loss": 2.5676, + "step": 13212 + }, + { + "epoch": 0.56, + "grad_norm": 0.392578125, + "learning_rate": 0.0001637450865821307, + "loss": 2.5185, + "step": 13213 + }, + { + "epoch": 0.56, + "grad_norm": 0.4375, + "learning_rate": 0.00016373991803779114, + "loss": 2.5654, + "step": 13214 + }, + { + "epoch": 0.56, + "grad_norm": 0.41796875, + "learning_rate": 0.00016373474920664528, + "loss": 2.6471, + "step": 13215 + }, + { + "epoch": 0.56, + "grad_norm": 0.443359375, + "learning_rate": 0.0001637295800887163, + "loss": 2.4975, + "step": 13216 + }, + { + "epoch": 0.56, + "grad_norm": 0.5078125, + "learning_rate": 0.00016372441068402752, + "loss": 2.7623, + "step": 13217 + }, + { + "epoch": 0.56, + "grad_norm": 0.482421875, + "learning_rate": 0.0001637192409926022, + "loss": 2.7546, + "step": 13218 + }, + { + "epoch": 0.56, + "grad_norm": 0.43359375, + "learning_rate": 0.00016371407101446357, + "loss": 2.5289, + "step": 13219 + }, + { + "epoch": 0.56, + "grad_norm": 0.4921875, + "learning_rate": 0.00016370890074963489, + "loss": 2.6419, + "step": 13220 + }, + { + "epoch": 0.56, + "grad_norm": 0.4765625, + "learning_rate": 0.00016370373019813946, + "loss": 2.6054, + "step": 13221 + }, + { + "epoch": 0.56, + "grad_norm": 0.458984375, + "learning_rate": 0.0001636985593600005, + "loss": 2.4808, + "step": 13222 + }, + { + "epoch": 0.56, + "grad_norm": 0.44140625, + "learning_rate": 0.00016369338823524134, + "loss": 2.6417, + "step": 13223 + }, + { + "epoch": 0.56, + "grad_norm": 0.57421875, + "learning_rate": 0.00016368821682388518, + "loss": 2.589, + "step": 13224 + }, + { + "epoch": 0.56, + "grad_norm": 0.45703125, + "learning_rate": 0.00016368304512595535, + "loss": 2.7627, + "step": 13225 + }, + { + "epoch": 0.56, + "grad_norm": 0.4296875, + "learning_rate": 0.00016367787314147507, + "loss": 2.5403, + "step": 13226 + }, + { + "epoch": 0.56, + "grad_norm": 0.53515625, + "learning_rate": 0.00016367270087046763, + "loss": 2.6786, + "step": 13227 + }, + { + "epoch": 0.56, + "grad_norm": 0.412109375, + "learning_rate": 0.00016366752831295628, + "loss": 2.6035, + "step": 13228 + }, + { + "epoch": 0.56, + "grad_norm": 0.46875, + "learning_rate": 0.00016366235546896437, + "loss": 2.6022, + "step": 13229 + }, + { + "epoch": 0.56, + "grad_norm": 0.435546875, + "learning_rate": 0.0001636571823385151, + "loss": 2.6238, + "step": 13230 + }, + { + "epoch": 0.56, + "grad_norm": 0.458984375, + "learning_rate": 0.0001636520089216318, + "loss": 2.7208, + "step": 13231 + }, + { + "epoch": 0.56, + "grad_norm": 0.431640625, + "learning_rate": 0.00016364683521833767, + "loss": 2.5798, + "step": 13232 + }, + { + "epoch": 0.56, + "grad_norm": 0.42578125, + "learning_rate": 0.00016364166122865607, + "loss": 2.5116, + "step": 13233 + }, + { + "epoch": 0.56, + "grad_norm": 0.4453125, + "learning_rate": 0.00016363648695261028, + "loss": 2.5623, + "step": 13234 + }, + { + "epoch": 0.56, + "grad_norm": 0.384765625, + "learning_rate": 0.00016363131239022352, + "loss": 2.5787, + "step": 13235 + }, + { + "epoch": 0.56, + "grad_norm": 0.470703125, + "learning_rate": 0.00016362613754151913, + "loss": 2.5756, + "step": 13236 + }, + { + "epoch": 0.56, + "grad_norm": 0.431640625, + "learning_rate": 0.00016362096240652037, + "loss": 2.641, + "step": 13237 + }, + { + "epoch": 0.56, + "grad_norm": 0.44140625, + "learning_rate": 0.0001636157869852505, + "loss": 2.4786, + "step": 13238 + }, + { + "epoch": 0.56, + "grad_norm": 0.41015625, + "learning_rate": 0.00016361061127773287, + "loss": 2.437, + "step": 13239 + }, + { + "epoch": 0.56, + "grad_norm": 0.48828125, + "learning_rate": 0.00016360543528399073, + "loss": 2.6034, + "step": 13240 + }, + { + "epoch": 0.56, + "grad_norm": 0.419921875, + "learning_rate": 0.00016360025900404737, + "loss": 2.5736, + "step": 13241 + }, + { + "epoch": 0.56, + "grad_norm": 0.494140625, + "learning_rate": 0.00016359508243792608, + "loss": 2.5797, + "step": 13242 + }, + { + "epoch": 0.56, + "grad_norm": 0.462890625, + "learning_rate": 0.00016358990558565018, + "loss": 2.6523, + "step": 13243 + }, + { + "epoch": 0.56, + "grad_norm": 0.466796875, + "learning_rate": 0.00016358472844724294, + "loss": 2.6993, + "step": 13244 + }, + { + "epoch": 0.56, + "grad_norm": 0.48828125, + "learning_rate": 0.00016357955102272765, + "loss": 2.6006, + "step": 13245 + }, + { + "epoch": 0.56, + "grad_norm": 0.4765625, + "learning_rate": 0.00016357437331212762, + "loss": 2.6682, + "step": 13246 + }, + { + "epoch": 0.56, + "grad_norm": 0.48828125, + "learning_rate": 0.00016356919531546616, + "loss": 2.7119, + "step": 13247 + }, + { + "epoch": 0.56, + "grad_norm": 0.4375, + "learning_rate": 0.00016356401703276652, + "loss": 2.567, + "step": 13248 + }, + { + "epoch": 0.56, + "grad_norm": 0.52734375, + "learning_rate": 0.00016355883846405208, + "loss": 2.6445, + "step": 13249 + }, + { + "epoch": 0.56, + "grad_norm": 0.44140625, + "learning_rate": 0.00016355365960934602, + "loss": 2.468, + "step": 13250 + }, + { + "epoch": 0.56, + "grad_norm": 0.462890625, + "learning_rate": 0.00016354848046867178, + "loss": 2.6221, + "step": 13251 + }, + { + "epoch": 0.56, + "grad_norm": 0.5078125, + "learning_rate": 0.00016354330104205258, + "loss": 2.6399, + "step": 13252 + }, + { + "epoch": 0.56, + "grad_norm": 0.466796875, + "learning_rate": 0.00016353812132951175, + "loss": 2.5871, + "step": 13253 + }, + { + "epoch": 0.56, + "grad_norm": 0.396484375, + "learning_rate": 0.0001635329413310726, + "loss": 2.6443, + "step": 13254 + }, + { + "epoch": 0.56, + "grad_norm": 0.46875, + "learning_rate": 0.00016352776104675844, + "loss": 2.7425, + "step": 13255 + }, + { + "epoch": 0.56, + "grad_norm": 0.5, + "learning_rate": 0.00016352258047659256, + "loss": 2.659, + "step": 13256 + }, + { + "epoch": 0.56, + "grad_norm": 0.408203125, + "learning_rate": 0.0001635173996205983, + "loss": 2.5687, + "step": 13257 + }, + { + "epoch": 0.56, + "grad_norm": 0.45703125, + "learning_rate": 0.00016351221847879895, + "loss": 2.6014, + "step": 13258 + }, + { + "epoch": 0.57, + "grad_norm": 0.458984375, + "learning_rate": 0.0001635070370512178, + "loss": 2.4897, + "step": 13259 + }, + { + "epoch": 0.57, + "grad_norm": 0.51953125, + "learning_rate": 0.0001635018553378782, + "loss": 2.6227, + "step": 13260 + }, + { + "epoch": 0.57, + "grad_norm": 0.484375, + "learning_rate": 0.00016349667333880348, + "loss": 2.6655, + "step": 13261 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.00016349149105401693, + "loss": 2.6205, + "step": 13262 + }, + { + "epoch": 0.57, + "grad_norm": 0.443359375, + "learning_rate": 0.00016348630848354184, + "loss": 2.5104, + "step": 13263 + }, + { + "epoch": 0.57, + "grad_norm": 0.486328125, + "learning_rate": 0.0001634811256274016, + "loss": 2.6078, + "step": 13264 + }, + { + "epoch": 0.57, + "grad_norm": 0.484375, + "learning_rate": 0.0001634759424856195, + "loss": 2.5227, + "step": 13265 + }, + { + "epoch": 0.57, + "grad_norm": 0.515625, + "learning_rate": 0.00016347075905821882, + "loss": 2.6419, + "step": 13266 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.00016346557534522294, + "loss": 2.5983, + "step": 13267 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.00016346039134665518, + "loss": 2.6739, + "step": 13268 + }, + { + "epoch": 0.57, + "grad_norm": 0.50390625, + "learning_rate": 0.00016345520706253884, + "loss": 2.5054, + "step": 13269 + }, + { + "epoch": 0.57, + "grad_norm": 0.435546875, + "learning_rate": 0.00016345002249289724, + "loss": 2.6356, + "step": 13270 + }, + { + "epoch": 0.57, + "grad_norm": 0.52734375, + "learning_rate": 0.00016344483763775376, + "loss": 2.696, + "step": 13271 + }, + { + "epoch": 0.57, + "grad_norm": 0.474609375, + "learning_rate": 0.0001634396524971317, + "loss": 2.7501, + "step": 13272 + }, + { + "epoch": 0.57, + "grad_norm": 0.439453125, + "learning_rate": 0.00016343446707105437, + "loss": 2.5323, + "step": 13273 + }, + { + "epoch": 0.57, + "grad_norm": 0.43359375, + "learning_rate": 0.0001634292813595451, + "loss": 2.5209, + "step": 13274 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.00016342409536262729, + "loss": 2.545, + "step": 13275 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.00016341890908032417, + "loss": 2.4381, + "step": 13276 + }, + { + "epoch": 0.57, + "grad_norm": 0.490234375, + "learning_rate": 0.00016341372251265921, + "loss": 2.7091, + "step": 13277 + }, + { + "epoch": 0.57, + "grad_norm": 0.50390625, + "learning_rate": 0.0001634085356596556, + "loss": 2.6974, + "step": 13278 + }, + { + "epoch": 0.57, + "grad_norm": 0.53125, + "learning_rate": 0.0001634033485213368, + "loss": 2.5102, + "step": 13279 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.00016339816109772607, + "loss": 2.5262, + "step": 13280 + }, + { + "epoch": 0.57, + "grad_norm": 0.5234375, + "learning_rate": 0.00016339297338884684, + "loss": 2.6596, + "step": 13281 + }, + { + "epoch": 0.57, + "grad_norm": 0.474609375, + "learning_rate": 0.00016338778539472235, + "loss": 2.5774, + "step": 13282 + }, + { + "epoch": 0.57, + "grad_norm": 0.4453125, + "learning_rate": 0.000163382597115376, + "loss": 2.6967, + "step": 13283 + }, + { + "epoch": 0.57, + "grad_norm": 0.447265625, + "learning_rate": 0.00016337740855083114, + "loss": 2.5863, + "step": 13284 + }, + { + "epoch": 0.57, + "grad_norm": 0.53125, + "learning_rate": 0.0001633722197011111, + "loss": 2.4658, + "step": 13285 + }, + { + "epoch": 0.57, + "grad_norm": 0.53125, + "learning_rate": 0.00016336703056623922, + "loss": 2.6461, + "step": 13286 + }, + { + "epoch": 0.57, + "grad_norm": 0.46484375, + "learning_rate": 0.00016336184114623884, + "loss": 2.5167, + "step": 13287 + }, + { + "epoch": 0.57, + "grad_norm": 0.431640625, + "learning_rate": 0.00016335665144113338, + "loss": 2.6131, + "step": 13288 + }, + { + "epoch": 0.57, + "grad_norm": 0.45703125, + "learning_rate": 0.00016335146145094611, + "loss": 2.7329, + "step": 13289 + }, + { + "epoch": 0.57, + "grad_norm": 0.490234375, + "learning_rate": 0.00016334627117570042, + "loss": 2.6816, + "step": 13290 + }, + { + "epoch": 0.57, + "grad_norm": 0.474609375, + "learning_rate": 0.00016334108061541967, + "loss": 2.6079, + "step": 13291 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.0001633358897701272, + "loss": 2.5721, + "step": 13292 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.00016333069863984638, + "loss": 2.5484, + "step": 13293 + }, + { + "epoch": 0.57, + "grad_norm": 0.515625, + "learning_rate": 0.00016332550722460054, + "loss": 2.6649, + "step": 13294 + }, + { + "epoch": 0.57, + "grad_norm": 0.453125, + "learning_rate": 0.0001633203155244131, + "loss": 2.5451, + "step": 13295 + }, + { + "epoch": 0.57, + "grad_norm": 0.458984375, + "learning_rate": 0.00016331512353930734, + "loss": 2.5295, + "step": 13296 + }, + { + "epoch": 0.57, + "grad_norm": 0.462890625, + "learning_rate": 0.00016330993126930667, + "loss": 2.5706, + "step": 13297 + }, + { + "epoch": 0.57, + "grad_norm": 0.5, + "learning_rate": 0.00016330473871443445, + "loss": 2.6217, + "step": 13298 + }, + { + "epoch": 0.57, + "grad_norm": 0.474609375, + "learning_rate": 0.00016329954587471405, + "loss": 2.6061, + "step": 13299 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.0001632943527501688, + "loss": 2.5807, + "step": 13300 + }, + { + "epoch": 0.57, + "grad_norm": 0.494140625, + "learning_rate": 0.00016328915934082217, + "loss": 2.7103, + "step": 13301 + }, + { + "epoch": 0.57, + "grad_norm": 0.4765625, + "learning_rate": 0.00016328396564669736, + "loss": 2.7329, + "step": 13302 + }, + { + "epoch": 0.57, + "grad_norm": 0.462890625, + "learning_rate": 0.00016327877166781788, + "loss": 2.5661, + "step": 13303 + }, + { + "epoch": 0.57, + "grad_norm": 0.55078125, + "learning_rate": 0.00016327357740420704, + "loss": 2.7578, + "step": 13304 + }, + { + "epoch": 0.57, + "grad_norm": 0.421875, + "learning_rate": 0.00016326838285588826, + "loss": 2.5207, + "step": 13305 + }, + { + "epoch": 0.57, + "grad_norm": 0.494140625, + "learning_rate": 0.00016326318802288483, + "loss": 2.607, + "step": 13306 + }, + { + "epoch": 0.57, + "grad_norm": 0.478515625, + "learning_rate": 0.0001632579929052202, + "loss": 2.5909, + "step": 13307 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.0001632527975029177, + "loss": 2.6007, + "step": 13308 + }, + { + "epoch": 0.57, + "grad_norm": 0.56640625, + "learning_rate": 0.00016324760181600075, + "loss": 2.6496, + "step": 13309 + }, + { + "epoch": 0.57, + "grad_norm": 0.45703125, + "learning_rate": 0.00016324240584449272, + "loss": 2.5732, + "step": 13310 + }, + { + "epoch": 0.57, + "grad_norm": 0.498046875, + "learning_rate": 0.00016323720958841693, + "loss": 2.5677, + "step": 13311 + }, + { + "epoch": 0.57, + "grad_norm": 0.427734375, + "learning_rate": 0.0001632320130477968, + "loss": 2.5164, + "step": 13312 + }, + { + "epoch": 0.57, + "grad_norm": 0.48828125, + "learning_rate": 0.00016322681622265577, + "loss": 2.5945, + "step": 13313 + }, + { + "epoch": 0.57, + "grad_norm": 0.478515625, + "learning_rate": 0.00016322161911301713, + "loss": 2.6835, + "step": 13314 + }, + { + "epoch": 0.57, + "grad_norm": 0.470703125, + "learning_rate": 0.00016321642171890436, + "loss": 2.6521, + "step": 13315 + }, + { + "epoch": 0.57, + "grad_norm": 0.45703125, + "learning_rate": 0.00016321122404034077, + "loss": 2.6639, + "step": 13316 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.00016320602607734977, + "loss": 2.546, + "step": 13317 + }, + { + "epoch": 0.57, + "grad_norm": 0.52734375, + "learning_rate": 0.00016320082782995474, + "loss": 2.5859, + "step": 13318 + }, + { + "epoch": 0.57, + "grad_norm": 0.48828125, + "learning_rate": 0.00016319562929817912, + "loss": 2.7352, + "step": 13319 + }, + { + "epoch": 0.57, + "grad_norm": 0.4609375, + "learning_rate": 0.00016319043048204624, + "loss": 2.7185, + "step": 13320 + }, + { + "epoch": 0.57, + "grad_norm": 0.43359375, + "learning_rate": 0.0001631852313815795, + "loss": 2.4976, + "step": 13321 + }, + { + "epoch": 0.57, + "grad_norm": 0.4453125, + "learning_rate": 0.00016318003199680234, + "loss": 2.5431, + "step": 13322 + }, + { + "epoch": 0.57, + "grad_norm": 0.439453125, + "learning_rate": 0.0001631748323277381, + "loss": 2.6435, + "step": 13323 + }, + { + "epoch": 0.57, + "grad_norm": 0.4296875, + "learning_rate": 0.0001631696323744102, + "loss": 2.5193, + "step": 13324 + }, + { + "epoch": 0.57, + "grad_norm": 0.50390625, + "learning_rate": 0.00016316443213684208, + "loss": 2.5613, + "step": 13325 + }, + { + "epoch": 0.57, + "grad_norm": 0.453125, + "learning_rate": 0.00016315923161505706, + "loss": 2.4499, + "step": 13326 + }, + { + "epoch": 0.57, + "grad_norm": 0.494140625, + "learning_rate": 0.0001631540308090786, + "loss": 2.6544, + "step": 13327 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.0001631488297189301, + "loss": 2.548, + "step": 13328 + }, + { + "epoch": 0.57, + "grad_norm": 0.43359375, + "learning_rate": 0.0001631436283446349, + "loss": 2.5903, + "step": 13329 + }, + { + "epoch": 0.57, + "grad_norm": 0.38671875, + "learning_rate": 0.00016313842668621645, + "loss": 2.46, + "step": 13330 + }, + { + "epoch": 0.57, + "grad_norm": 0.46484375, + "learning_rate": 0.00016313322474369816, + "loss": 2.615, + "step": 13331 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.00016312802251710346, + "loss": 2.6459, + "step": 13332 + }, + { + "epoch": 0.57, + "grad_norm": 0.439453125, + "learning_rate": 0.00016312282000645572, + "loss": 2.5836, + "step": 13333 + }, + { + "epoch": 0.57, + "grad_norm": 0.419921875, + "learning_rate": 0.00016311761721177835, + "loss": 2.5944, + "step": 13334 + }, + { + "epoch": 0.57, + "grad_norm": 0.447265625, + "learning_rate": 0.00016311241413309477, + "loss": 2.6779, + "step": 13335 + }, + { + "epoch": 0.57, + "grad_norm": 0.41796875, + "learning_rate": 0.00016310721077042838, + "loss": 2.4575, + "step": 13336 + }, + { + "epoch": 0.57, + "grad_norm": 0.50390625, + "learning_rate": 0.00016310200712380262, + "loss": 2.5881, + "step": 13337 + }, + { + "epoch": 0.57, + "grad_norm": 0.44921875, + "learning_rate": 0.0001630968031932409, + "loss": 2.602, + "step": 13338 + }, + { + "epoch": 0.57, + "grad_norm": 0.4609375, + "learning_rate": 0.0001630915989787666, + "loss": 2.6196, + "step": 13339 + }, + { + "epoch": 0.57, + "grad_norm": 0.4375, + "learning_rate": 0.00016308639448040318, + "loss": 2.6008, + "step": 13340 + }, + { + "epoch": 0.57, + "grad_norm": 0.490234375, + "learning_rate": 0.000163081189698174, + "loss": 2.5537, + "step": 13341 + }, + { + "epoch": 0.57, + "grad_norm": 0.474609375, + "learning_rate": 0.00016307598463210257, + "loss": 2.6241, + "step": 13342 + }, + { + "epoch": 0.57, + "grad_norm": 0.484375, + "learning_rate": 0.00016307077928221224, + "loss": 2.6304, + "step": 13343 + }, + { + "epoch": 0.57, + "grad_norm": 0.478515625, + "learning_rate": 0.0001630655736485265, + "loss": 2.6928, + "step": 13344 + }, + { + "epoch": 0.57, + "grad_norm": 0.40234375, + "learning_rate": 0.00016306036773106866, + "loss": 2.6259, + "step": 13345 + }, + { + "epoch": 0.57, + "grad_norm": 0.470703125, + "learning_rate": 0.00016305516152986222, + "loss": 2.5424, + "step": 13346 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.0001630499550449306, + "loss": 2.69, + "step": 13347 + }, + { + "epoch": 0.57, + "grad_norm": 0.392578125, + "learning_rate": 0.00016304474827629725, + "loss": 2.5678, + "step": 13348 + }, + { + "epoch": 0.57, + "grad_norm": 0.498046875, + "learning_rate": 0.00016303954122398557, + "loss": 2.5816, + "step": 13349 + }, + { + "epoch": 0.57, + "grad_norm": 0.462890625, + "learning_rate": 0.00016303433388801896, + "loss": 2.566, + "step": 13350 + }, + { + "epoch": 0.57, + "grad_norm": 0.4375, + "learning_rate": 0.0001630291262684209, + "loss": 2.5776, + "step": 13351 + }, + { + "epoch": 0.57, + "grad_norm": 0.43359375, + "learning_rate": 0.00016302391836521484, + "loss": 2.689, + "step": 13352 + }, + { + "epoch": 0.57, + "grad_norm": 0.427734375, + "learning_rate": 0.00016301871017842414, + "loss": 2.5413, + "step": 13353 + }, + { + "epoch": 0.57, + "grad_norm": 0.43359375, + "learning_rate": 0.00016301350170807227, + "loss": 2.4869, + "step": 13354 + }, + { + "epoch": 0.57, + "grad_norm": 0.419921875, + "learning_rate": 0.0001630082929541827, + "loss": 2.5822, + "step": 13355 + }, + { + "epoch": 0.57, + "grad_norm": 0.4765625, + "learning_rate": 0.00016300308391677882, + "loss": 2.7108, + "step": 13356 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.0001629978745958841, + "loss": 2.5753, + "step": 13357 + }, + { + "epoch": 0.57, + "grad_norm": 0.443359375, + "learning_rate": 0.00016299266499152196, + "loss": 2.5174, + "step": 13358 + }, + { + "epoch": 0.57, + "grad_norm": 0.400390625, + "learning_rate": 0.00016298745510371585, + "loss": 2.4669, + "step": 13359 + }, + { + "epoch": 0.57, + "grad_norm": 0.474609375, + "learning_rate": 0.0001629822449324892, + "loss": 2.6245, + "step": 13360 + }, + { + "epoch": 0.57, + "grad_norm": 0.51953125, + "learning_rate": 0.00016297703447786546, + "loss": 2.6707, + "step": 13361 + }, + { + "epoch": 0.57, + "grad_norm": 0.4296875, + "learning_rate": 0.00016297182373986813, + "loss": 2.588, + "step": 13362 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.00016296661271852057, + "loss": 2.6189, + "step": 13363 + }, + { + "epoch": 0.57, + "grad_norm": 0.49609375, + "learning_rate": 0.00016296140141384627, + "loss": 2.5896, + "step": 13364 + }, + { + "epoch": 0.57, + "grad_norm": 0.484375, + "learning_rate": 0.00016295618982586864, + "loss": 2.7094, + "step": 13365 + }, + { + "epoch": 0.57, + "grad_norm": 0.478515625, + "learning_rate": 0.00016295097795461123, + "loss": 2.6647, + "step": 13366 + }, + { + "epoch": 0.57, + "grad_norm": 0.419921875, + "learning_rate": 0.00016294576580009736, + "loss": 2.5578, + "step": 13367 + }, + { + "epoch": 0.57, + "grad_norm": 0.470703125, + "learning_rate": 0.0001629405533623506, + "loss": 2.5725, + "step": 13368 + }, + { + "epoch": 0.57, + "grad_norm": 0.40234375, + "learning_rate": 0.0001629353406413943, + "loss": 2.5484, + "step": 13369 + }, + { + "epoch": 0.57, + "grad_norm": 0.48828125, + "learning_rate": 0.00016293012763725202, + "loss": 2.5779, + "step": 13370 + }, + { + "epoch": 0.57, + "grad_norm": 0.44921875, + "learning_rate": 0.00016292491434994713, + "loss": 2.6715, + "step": 13371 + }, + { + "epoch": 0.57, + "grad_norm": 0.498046875, + "learning_rate": 0.00016291970077950314, + "loss": 2.6685, + "step": 13372 + }, + { + "epoch": 0.57, + "grad_norm": 0.4453125, + "learning_rate": 0.0001629144869259435, + "loss": 2.5764, + "step": 13373 + }, + { + "epoch": 0.57, + "grad_norm": 0.51953125, + "learning_rate": 0.00016290927278929163, + "loss": 2.5627, + "step": 13374 + }, + { + "epoch": 0.57, + "grad_norm": 0.3984375, + "learning_rate": 0.00016290405836957104, + "loss": 2.507, + "step": 13375 + }, + { + "epoch": 0.57, + "grad_norm": 0.5078125, + "learning_rate": 0.00016289884366680517, + "loss": 2.6209, + "step": 13376 + }, + { + "epoch": 0.57, + "grad_norm": 0.478515625, + "learning_rate": 0.00016289362868101748, + "loss": 2.6296, + "step": 13377 + }, + { + "epoch": 0.57, + "grad_norm": 0.51953125, + "learning_rate": 0.00016288841341223148, + "loss": 2.7534, + "step": 13378 + }, + { + "epoch": 0.57, + "grad_norm": 0.50390625, + "learning_rate": 0.00016288319786047058, + "loss": 2.6479, + "step": 13379 + }, + { + "epoch": 0.57, + "grad_norm": 0.439453125, + "learning_rate": 0.00016287798202575828, + "loss": 2.6394, + "step": 13380 + }, + { + "epoch": 0.57, + "grad_norm": 0.451171875, + "learning_rate": 0.00016287276590811804, + "loss": 2.5956, + "step": 13381 + }, + { + "epoch": 0.57, + "grad_norm": 0.46875, + "learning_rate": 0.0001628675495075733, + "loss": 2.6619, + "step": 13382 + }, + { + "epoch": 0.57, + "grad_norm": 0.4609375, + "learning_rate": 0.0001628623328241476, + "loss": 2.6928, + "step": 13383 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.00016285711585786438, + "loss": 2.6255, + "step": 13384 + }, + { + "epoch": 0.57, + "grad_norm": 0.462890625, + "learning_rate": 0.00016285189860874708, + "loss": 2.5411, + "step": 13385 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.00016284668107681923, + "loss": 2.6666, + "step": 13386 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.00016284146326210426, + "loss": 2.6074, + "step": 13387 + }, + { + "epoch": 0.57, + "grad_norm": 0.52734375, + "learning_rate": 0.0001628362451646257, + "loss": 2.7287, + "step": 13388 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.00016283102678440697, + "loss": 2.7208, + "step": 13389 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.0001628258081214716, + "loss": 2.4857, + "step": 13390 + }, + { + "epoch": 0.57, + "grad_norm": 0.4609375, + "learning_rate": 0.00016282058917584303, + "loss": 2.6339, + "step": 13391 + }, + { + "epoch": 0.57, + "grad_norm": 0.51171875, + "learning_rate": 0.0001628153699475448, + "loss": 2.6545, + "step": 13392 + }, + { + "epoch": 0.57, + "grad_norm": 0.458984375, + "learning_rate": 0.00016281015043660035, + "loss": 2.597, + "step": 13393 + }, + { + "epoch": 0.57, + "grad_norm": 0.431640625, + "learning_rate": 0.00016280493064303316, + "loss": 2.4708, + "step": 13394 + }, + { + "epoch": 0.57, + "grad_norm": 0.498046875, + "learning_rate": 0.00016279971056686678, + "loss": 2.6788, + "step": 13395 + }, + { + "epoch": 0.57, + "grad_norm": 0.4296875, + "learning_rate": 0.0001627944902081246, + "loss": 2.5777, + "step": 13396 + }, + { + "epoch": 0.57, + "grad_norm": 0.451171875, + "learning_rate": 0.0001627892695668302, + "loss": 2.4945, + "step": 13397 + }, + { + "epoch": 0.57, + "grad_norm": 0.421875, + "learning_rate": 0.000162784048643007, + "loss": 2.6254, + "step": 13398 + }, + { + "epoch": 0.57, + "grad_norm": 0.4765625, + "learning_rate": 0.00016277882743667856, + "loss": 2.5447, + "step": 13399 + }, + { + "epoch": 0.57, + "grad_norm": 0.48046875, + "learning_rate": 0.00016277360594786833, + "loss": 2.5973, + "step": 13400 + }, + { + "epoch": 0.57, + "grad_norm": 0.5, + "learning_rate": 0.0001627683841765998, + "loss": 2.6286, + "step": 13401 + }, + { + "epoch": 0.57, + "grad_norm": 0.45703125, + "learning_rate": 0.0001627631621228965, + "loss": 2.608, + "step": 13402 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.0001627579397867819, + "loss": 2.7324, + "step": 13403 + }, + { + "epoch": 0.57, + "grad_norm": 0.439453125, + "learning_rate": 0.00016275271716827955, + "loss": 2.5207, + "step": 13404 + }, + { + "epoch": 0.57, + "grad_norm": 0.423828125, + "learning_rate": 0.00016274749426741284, + "loss": 2.636, + "step": 13405 + }, + { + "epoch": 0.57, + "grad_norm": 0.455078125, + "learning_rate": 0.00016274227108420536, + "loss": 2.6633, + "step": 13406 + }, + { + "epoch": 0.57, + "grad_norm": 0.4765625, + "learning_rate": 0.00016273704761868055, + "loss": 2.6522, + "step": 13407 + }, + { + "epoch": 0.57, + "grad_norm": 0.490234375, + "learning_rate": 0.00016273182387086202, + "loss": 2.5862, + "step": 13408 + }, + { + "epoch": 0.57, + "grad_norm": 0.51953125, + "learning_rate": 0.00016272659984077316, + "loss": 2.6992, + "step": 13409 + }, + { + "epoch": 0.57, + "grad_norm": 0.419921875, + "learning_rate": 0.00016272137552843756, + "loss": 2.5942, + "step": 13410 + }, + { + "epoch": 0.57, + "grad_norm": 0.45703125, + "learning_rate": 0.00016271615093387868, + "loss": 2.5055, + "step": 13411 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.00016271092605712003, + "loss": 2.6331, + "step": 13412 + }, + { + "epoch": 0.57, + "grad_norm": 0.498046875, + "learning_rate": 0.00016270570089818516, + "loss": 2.7235, + "step": 13413 + }, + { + "epoch": 0.57, + "grad_norm": 0.458984375, + "learning_rate": 0.00016270047545709752, + "loss": 2.5961, + "step": 13414 + }, + { + "epoch": 0.57, + "grad_norm": 0.45703125, + "learning_rate": 0.00016269524973388066, + "loss": 2.7358, + "step": 13415 + }, + { + "epoch": 0.57, + "grad_norm": 0.494140625, + "learning_rate": 0.0001626900237285581, + "loss": 2.6602, + "step": 13416 + }, + { + "epoch": 0.57, + "grad_norm": 0.49609375, + "learning_rate": 0.00016268479744115336, + "loss": 2.589, + "step": 13417 + }, + { + "epoch": 0.57, + "grad_norm": 0.396484375, + "learning_rate": 0.0001626795708716899, + "loss": 2.5309, + "step": 13418 + }, + { + "epoch": 0.57, + "grad_norm": 0.462890625, + "learning_rate": 0.00016267434402019132, + "loss": 2.5237, + "step": 13419 + }, + { + "epoch": 0.57, + "grad_norm": 0.400390625, + "learning_rate": 0.00016266911688668105, + "loss": 2.6085, + "step": 13420 + }, + { + "epoch": 0.57, + "grad_norm": 0.48046875, + "learning_rate": 0.00016266388947118267, + "loss": 2.656, + "step": 13421 + }, + { + "epoch": 0.57, + "grad_norm": 0.48046875, + "learning_rate": 0.0001626586617737197, + "loss": 2.464, + "step": 13422 + }, + { + "epoch": 0.57, + "grad_norm": 0.453125, + "learning_rate": 0.00016265343379431567, + "loss": 2.6037, + "step": 13423 + }, + { + "epoch": 0.57, + "grad_norm": 0.52734375, + "learning_rate": 0.00016264820553299404, + "loss": 2.7008, + "step": 13424 + }, + { + "epoch": 0.57, + "grad_norm": 0.478515625, + "learning_rate": 0.00016264297698977842, + "loss": 2.7159, + "step": 13425 + }, + { + "epoch": 0.57, + "grad_norm": 0.455078125, + "learning_rate": 0.0001626377481646923, + "loss": 2.5582, + "step": 13426 + }, + { + "epoch": 0.57, + "grad_norm": 0.46484375, + "learning_rate": 0.00016263251905775916, + "loss": 2.643, + "step": 13427 + }, + { + "epoch": 0.57, + "grad_norm": 0.42578125, + "learning_rate": 0.0001626272896690026, + "loss": 2.5818, + "step": 13428 + }, + { + "epoch": 0.57, + "grad_norm": 0.42578125, + "learning_rate": 0.00016262205999844614, + "loss": 2.6055, + "step": 13429 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.00016261683004611328, + "loss": 2.6311, + "step": 13430 + }, + { + "epoch": 0.57, + "grad_norm": 0.51953125, + "learning_rate": 0.00016261159981202755, + "loss": 2.681, + "step": 13431 + }, + { + "epoch": 0.57, + "grad_norm": 0.4140625, + "learning_rate": 0.00016260636929621254, + "loss": 2.7377, + "step": 13432 + }, + { + "epoch": 0.57, + "grad_norm": 0.439453125, + "learning_rate": 0.00016260113849869172, + "loss": 2.6229, + "step": 13433 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.00016259590741948867, + "loss": 2.5703, + "step": 13434 + }, + { + "epoch": 0.57, + "grad_norm": 0.498046875, + "learning_rate": 0.00016259067605862691, + "loss": 2.607, + "step": 13435 + }, + { + "epoch": 0.57, + "grad_norm": 0.515625, + "learning_rate": 0.00016258544441612998, + "loss": 2.6249, + "step": 13436 + }, + { + "epoch": 0.57, + "grad_norm": 0.5, + "learning_rate": 0.00016258021249202143, + "loss": 2.6409, + "step": 13437 + }, + { + "epoch": 0.57, + "grad_norm": 0.435546875, + "learning_rate": 0.0001625749802863248, + "loss": 2.4318, + "step": 13438 + }, + { + "epoch": 0.57, + "grad_norm": 0.4921875, + "learning_rate": 0.0001625697477990636, + "loss": 2.6475, + "step": 13439 + }, + { + "epoch": 0.57, + "grad_norm": 0.435546875, + "learning_rate": 0.0001625645150302614, + "loss": 2.5226, + "step": 13440 + }, + { + "epoch": 0.57, + "grad_norm": 0.41796875, + "learning_rate": 0.00016255928197994178, + "loss": 2.6864, + "step": 13441 + }, + { + "epoch": 0.57, + "grad_norm": 0.435546875, + "learning_rate": 0.00016255404864812825, + "loss": 2.6605, + "step": 13442 + }, + { + "epoch": 0.57, + "grad_norm": 0.46875, + "learning_rate": 0.00016254881503484435, + "loss": 2.6151, + "step": 13443 + }, + { + "epoch": 0.57, + "grad_norm": 0.4765625, + "learning_rate": 0.00016254358114011364, + "loss": 2.5905, + "step": 13444 + }, + { + "epoch": 0.57, + "grad_norm": 0.46484375, + "learning_rate": 0.00016253834696395965, + "loss": 2.6141, + "step": 13445 + }, + { + "epoch": 0.57, + "grad_norm": 0.50390625, + "learning_rate": 0.00016253311250640596, + "loss": 2.5622, + "step": 13446 + }, + { + "epoch": 0.57, + "grad_norm": 0.42578125, + "learning_rate": 0.00016252787776747617, + "loss": 2.5584, + "step": 13447 + }, + { + "epoch": 0.57, + "grad_norm": 0.462890625, + "learning_rate": 0.00016252264274719375, + "loss": 2.5053, + "step": 13448 + }, + { + "epoch": 0.57, + "grad_norm": 0.427734375, + "learning_rate": 0.00016251740744558228, + "loss": 2.5147, + "step": 13449 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.00016251217186266529, + "loss": 2.658, + "step": 13450 + }, + { + "epoch": 0.57, + "grad_norm": 0.486328125, + "learning_rate": 0.00016250693599846643, + "loss": 2.5773, + "step": 13451 + }, + { + "epoch": 0.57, + "grad_norm": 0.458984375, + "learning_rate": 0.00016250169985300919, + "loss": 2.6327, + "step": 13452 + }, + { + "epoch": 0.57, + "grad_norm": 0.427734375, + "learning_rate": 0.00016249646342631713, + "loss": 2.3957, + "step": 13453 + }, + { + "epoch": 0.57, + "grad_norm": 0.48046875, + "learning_rate": 0.0001624912267184138, + "loss": 2.5763, + "step": 13454 + }, + { + "epoch": 0.57, + "grad_norm": 0.443359375, + "learning_rate": 0.00016248598972932286, + "loss": 2.5675, + "step": 13455 + }, + { + "epoch": 0.57, + "grad_norm": 0.4296875, + "learning_rate": 0.00016248075245906774, + "loss": 2.582, + "step": 13456 + }, + { + "epoch": 0.57, + "grad_norm": 0.4296875, + "learning_rate": 0.0001624755149076721, + "loss": 2.5706, + "step": 13457 + }, + { + "epoch": 0.57, + "grad_norm": 0.4609375, + "learning_rate": 0.00016247027707515943, + "loss": 2.6471, + "step": 13458 + }, + { + "epoch": 0.57, + "grad_norm": 0.396484375, + "learning_rate": 0.0001624650389615534, + "loss": 2.4923, + "step": 13459 + }, + { + "epoch": 0.57, + "grad_norm": 0.484375, + "learning_rate": 0.00016245980056687746, + "loss": 2.5608, + "step": 13460 + }, + { + "epoch": 0.57, + "grad_norm": 0.486328125, + "learning_rate": 0.0001624545618911553, + "loss": 2.7216, + "step": 13461 + }, + { + "epoch": 0.57, + "grad_norm": 0.478515625, + "learning_rate": 0.0001624493229344104, + "loss": 2.5798, + "step": 13462 + }, + { + "epoch": 0.57, + "grad_norm": 0.4921875, + "learning_rate": 0.00016244408369666638, + "loss": 2.4928, + "step": 13463 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.00016243884417794683, + "loss": 2.5102, + "step": 13464 + }, + { + "epoch": 0.57, + "grad_norm": 0.404296875, + "learning_rate": 0.00016243360437827526, + "loss": 2.5241, + "step": 13465 + }, + { + "epoch": 0.57, + "grad_norm": 0.46484375, + "learning_rate": 0.0001624283642976753, + "loss": 2.522, + "step": 13466 + }, + { + "epoch": 0.57, + "grad_norm": 0.5078125, + "learning_rate": 0.00016242312393617054, + "loss": 2.6594, + "step": 13467 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.0001624178832937845, + "loss": 2.58, + "step": 13468 + }, + { + "epoch": 0.57, + "grad_norm": 0.43359375, + "learning_rate": 0.00016241264237054084, + "loss": 2.4239, + "step": 13469 + }, + { + "epoch": 0.57, + "grad_norm": 0.447265625, + "learning_rate": 0.00016240740116646305, + "loss": 2.5208, + "step": 13470 + }, + { + "epoch": 0.57, + "grad_norm": 0.466796875, + "learning_rate": 0.0001624021596815748, + "loss": 2.6693, + "step": 13471 + }, + { + "epoch": 0.57, + "grad_norm": 0.4609375, + "learning_rate": 0.0001623969179158996, + "loss": 2.6507, + "step": 13472 + }, + { + "epoch": 0.57, + "grad_norm": 0.470703125, + "learning_rate": 0.0001623916758694611, + "loss": 2.5994, + "step": 13473 + }, + { + "epoch": 0.57, + "grad_norm": 0.47265625, + "learning_rate": 0.00016238643354228284, + "loss": 2.783, + "step": 13474 + }, + { + "epoch": 0.57, + "grad_norm": 0.435546875, + "learning_rate": 0.00016238119093438846, + "loss": 2.5633, + "step": 13475 + }, + { + "epoch": 0.57, + "grad_norm": 0.43359375, + "learning_rate": 0.0001623759480458015, + "loss": 2.5878, + "step": 13476 + }, + { + "epoch": 0.57, + "grad_norm": 0.451171875, + "learning_rate": 0.00016237070487654557, + "loss": 2.5521, + "step": 13477 + }, + { + "epoch": 0.57, + "grad_norm": 0.423828125, + "learning_rate": 0.00016236546142664427, + "loss": 2.6208, + "step": 13478 + }, + { + "epoch": 0.57, + "grad_norm": 0.458984375, + "learning_rate": 0.0001623602176961212, + "loss": 2.6852, + "step": 13479 + }, + { + "epoch": 0.57, + "grad_norm": 0.419921875, + "learning_rate": 0.0001623549736849999, + "loss": 2.5793, + "step": 13480 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.00016234972939330403, + "loss": 2.6282, + "step": 13481 + }, + { + "epoch": 0.57, + "grad_norm": 0.451171875, + "learning_rate": 0.00016234448482105714, + "loss": 2.679, + "step": 13482 + }, + { + "epoch": 0.57, + "grad_norm": 0.453125, + "learning_rate": 0.00016233923996828288, + "loss": 2.58, + "step": 13483 + }, + { + "epoch": 0.57, + "grad_norm": 0.4296875, + "learning_rate": 0.00016233399483500479, + "loss": 2.5826, + "step": 13484 + }, + { + "epoch": 0.57, + "grad_norm": 0.48046875, + "learning_rate": 0.00016232874942124657, + "loss": 2.5499, + "step": 13485 + }, + { + "epoch": 0.57, + "grad_norm": 0.44140625, + "learning_rate": 0.00016232350372703168, + "loss": 2.5121, + "step": 13486 + }, + { + "epoch": 0.57, + "grad_norm": 0.4375, + "learning_rate": 0.00016231825775238386, + "loss": 2.5558, + "step": 13487 + }, + { + "epoch": 0.57, + "grad_norm": 0.443359375, + "learning_rate": 0.00016231301149732664, + "loss": 2.5665, + "step": 13488 + }, + { + "epoch": 0.57, + "grad_norm": 0.4453125, + "learning_rate": 0.0001623077649618836, + "loss": 2.6328, + "step": 13489 + }, + { + "epoch": 0.57, + "grad_norm": 0.3984375, + "learning_rate": 0.00016230251814607842, + "loss": 2.5464, + "step": 13490 + }, + { + "epoch": 0.57, + "grad_norm": 0.46875, + "learning_rate": 0.0001622972710499347, + "loss": 2.6099, + "step": 13491 + }, + { + "epoch": 0.57, + "grad_norm": 0.390625, + "learning_rate": 0.000162292023673476, + "loss": 2.4962, + "step": 13492 + }, + { + "epoch": 0.57, + "grad_norm": 0.48046875, + "learning_rate": 0.00016228677601672596, + "loss": 2.4409, + "step": 13493 + }, + { + "epoch": 0.58, + "grad_norm": 0.431640625, + "learning_rate": 0.00016228152807970818, + "loss": 2.5195, + "step": 13494 + }, + { + "epoch": 0.58, + "grad_norm": 0.451171875, + "learning_rate": 0.00016227627986244628, + "loss": 2.7124, + "step": 13495 + }, + { + "epoch": 0.58, + "grad_norm": 0.447265625, + "learning_rate": 0.00016227103136496392, + "loss": 2.48, + "step": 13496 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016226578258728464, + "loss": 2.5355, + "step": 13497 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016226053352943212, + "loss": 2.6416, + "step": 13498 + }, + { + "epoch": 0.58, + "grad_norm": 0.498046875, + "learning_rate": 0.00016225528419142995, + "loss": 2.6787, + "step": 13499 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016225003457330173, + "loss": 2.6238, + "step": 13500 + }, + { + "epoch": 0.58, + "grad_norm": 0.365234375, + "learning_rate": 0.00016224478467507112, + "loss": 2.4805, + "step": 13501 + }, + { + "epoch": 0.58, + "grad_norm": 0.51171875, + "learning_rate": 0.0001622395344967617, + "loss": 2.5191, + "step": 13502 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016223428403839716, + "loss": 2.5813, + "step": 13503 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016222903330000108, + "loss": 2.722, + "step": 13504 + }, + { + "epoch": 0.58, + "grad_norm": 0.482421875, + "learning_rate": 0.00016222378228159708, + "loss": 2.6919, + "step": 13505 + }, + { + "epoch": 0.58, + "grad_norm": 0.453125, + "learning_rate": 0.00016221853098320882, + "loss": 2.5998, + "step": 13506 + }, + { + "epoch": 0.58, + "grad_norm": 0.46484375, + "learning_rate": 0.00016221327940485987, + "loss": 2.5867, + "step": 13507 + }, + { + "epoch": 0.58, + "grad_norm": 0.466796875, + "learning_rate": 0.0001622080275465739, + "loss": 2.7191, + "step": 13508 + }, + { + "epoch": 0.58, + "grad_norm": 0.458984375, + "learning_rate": 0.0001622027754083746, + "loss": 2.6994, + "step": 13509 + }, + { + "epoch": 0.58, + "grad_norm": 0.494140625, + "learning_rate": 0.0001621975229902855, + "loss": 2.5307, + "step": 13510 + }, + { + "epoch": 0.58, + "grad_norm": 0.490234375, + "learning_rate": 0.0001621922702923303, + "loss": 2.6901, + "step": 13511 + }, + { + "epoch": 0.58, + "grad_norm": 0.451171875, + "learning_rate": 0.00016218701731453256, + "loss": 2.6876, + "step": 13512 + }, + { + "epoch": 0.58, + "grad_norm": 0.412109375, + "learning_rate": 0.000162181764056916, + "loss": 2.6247, + "step": 13513 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016217651051950423, + "loss": 2.5691, + "step": 13514 + }, + { + "epoch": 0.58, + "grad_norm": 0.46484375, + "learning_rate": 0.0001621712567023209, + "loss": 2.5956, + "step": 13515 + }, + { + "epoch": 0.58, + "grad_norm": 0.453125, + "learning_rate": 0.0001621660026053896, + "loss": 2.4562, + "step": 13516 + }, + { + "epoch": 0.58, + "grad_norm": 0.50390625, + "learning_rate": 0.00016216074822873405, + "loss": 2.5863, + "step": 13517 + }, + { + "epoch": 0.58, + "grad_norm": 0.4375, + "learning_rate": 0.00016215549357237783, + "loss": 2.6636, + "step": 13518 + }, + { + "epoch": 0.58, + "grad_norm": 0.52734375, + "learning_rate": 0.0001621502386363446, + "loss": 2.8345, + "step": 13519 + }, + { + "epoch": 0.58, + "grad_norm": 0.5, + "learning_rate": 0.000162144983420658, + "loss": 2.6793, + "step": 13520 + }, + { + "epoch": 0.58, + "grad_norm": 0.5390625, + "learning_rate": 0.00016213972792534172, + "loss": 2.5652, + "step": 13521 + }, + { + "epoch": 0.58, + "grad_norm": 0.4609375, + "learning_rate": 0.00016213447215041935, + "loss": 2.5822, + "step": 13522 + }, + { + "epoch": 0.58, + "grad_norm": 0.447265625, + "learning_rate": 0.00016212921609591455, + "loss": 2.744, + "step": 13523 + }, + { + "epoch": 0.58, + "grad_norm": 0.4140625, + "learning_rate": 0.000162123959761851, + "loss": 2.5428, + "step": 13524 + }, + { + "epoch": 0.58, + "grad_norm": 0.3984375, + "learning_rate": 0.00016211870314825232, + "loss": 2.6618, + "step": 13525 + }, + { + "epoch": 0.58, + "grad_norm": 0.451171875, + "learning_rate": 0.0001621134462551422, + "loss": 2.5852, + "step": 13526 + }, + { + "epoch": 0.58, + "grad_norm": 0.451171875, + "learning_rate": 0.00016210818908254427, + "loss": 2.6148, + "step": 13527 + }, + { + "epoch": 0.58, + "grad_norm": 0.439453125, + "learning_rate": 0.00016210293163048216, + "loss": 2.563, + "step": 13528 + }, + { + "epoch": 0.58, + "grad_norm": 0.4921875, + "learning_rate": 0.00016209767389897958, + "loss": 2.4538, + "step": 13529 + }, + { + "epoch": 0.58, + "grad_norm": 0.490234375, + "learning_rate": 0.00016209241588806017, + "loss": 2.7606, + "step": 13530 + }, + { + "epoch": 0.58, + "grad_norm": 0.431640625, + "learning_rate": 0.00016208715759774756, + "loss": 2.517, + "step": 13531 + }, + { + "epoch": 0.58, + "grad_norm": 0.494140625, + "learning_rate": 0.00016208189902806544, + "loss": 2.6307, + "step": 13532 + }, + { + "epoch": 0.58, + "grad_norm": 0.44921875, + "learning_rate": 0.00016207664017903745, + "loss": 2.7303, + "step": 13533 + }, + { + "epoch": 0.58, + "grad_norm": 0.470703125, + "learning_rate": 0.00016207138105068727, + "loss": 2.6052, + "step": 13534 + }, + { + "epoch": 0.58, + "grad_norm": 0.443359375, + "learning_rate": 0.0001620661216430386, + "loss": 2.6917, + "step": 13535 + }, + { + "epoch": 0.58, + "grad_norm": 0.455078125, + "learning_rate": 0.000162060861956115, + "loss": 2.6081, + "step": 13536 + }, + { + "epoch": 0.58, + "grad_norm": 0.41796875, + "learning_rate": 0.00016205560198994024, + "loss": 2.5565, + "step": 13537 + }, + { + "epoch": 0.58, + "grad_norm": 0.396484375, + "learning_rate": 0.00016205034174453795, + "loss": 2.5526, + "step": 13538 + }, + { + "epoch": 0.58, + "grad_norm": 0.51171875, + "learning_rate": 0.00016204508121993178, + "loss": 2.6386, + "step": 13539 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016203982041614542, + "loss": 2.6627, + "step": 13540 + }, + { + "epoch": 0.58, + "grad_norm": 0.451171875, + "learning_rate": 0.00016203455933320258, + "loss": 2.583, + "step": 13541 + }, + { + "epoch": 0.58, + "grad_norm": 0.45703125, + "learning_rate": 0.00016202929797112683, + "loss": 2.5522, + "step": 13542 + }, + { + "epoch": 0.58, + "grad_norm": 0.451171875, + "learning_rate": 0.00016202403632994194, + "loss": 2.5644, + "step": 13543 + }, + { + "epoch": 0.58, + "grad_norm": 0.4609375, + "learning_rate": 0.00016201877440967158, + "loss": 2.5678, + "step": 13544 + }, + { + "epoch": 0.58, + "grad_norm": 0.484375, + "learning_rate": 0.00016201351221033938, + "loss": 2.6354, + "step": 13545 + }, + { + "epoch": 0.58, + "grad_norm": 0.39453125, + "learning_rate": 0.00016200824973196902, + "loss": 2.5884, + "step": 13546 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.0001620029869745842, + "loss": 2.4645, + "step": 13547 + }, + { + "epoch": 0.58, + "grad_norm": 0.486328125, + "learning_rate": 0.00016199772393820863, + "loss": 2.5302, + "step": 13548 + }, + { + "epoch": 0.58, + "grad_norm": 0.453125, + "learning_rate": 0.00016199246062286596, + "loss": 2.6254, + "step": 13549 + }, + { + "epoch": 0.58, + "grad_norm": 0.48828125, + "learning_rate": 0.00016198719702857983, + "loss": 2.6031, + "step": 13550 + }, + { + "epoch": 0.58, + "grad_norm": 0.474609375, + "learning_rate": 0.000161981933155374, + "loss": 2.5509, + "step": 13551 + }, + { + "epoch": 0.58, + "grad_norm": 0.49609375, + "learning_rate": 0.00016197666900327211, + "loss": 2.5514, + "step": 13552 + }, + { + "epoch": 0.58, + "grad_norm": 0.451171875, + "learning_rate": 0.00016197140457229788, + "loss": 2.6248, + "step": 13553 + }, + { + "epoch": 0.58, + "grad_norm": 0.482421875, + "learning_rate": 0.00016196613986247494, + "loss": 2.7078, + "step": 13554 + }, + { + "epoch": 0.58, + "grad_norm": 0.53515625, + "learning_rate": 0.00016196087487382706, + "loss": 2.5569, + "step": 13555 + }, + { + "epoch": 0.58, + "grad_norm": 0.423828125, + "learning_rate": 0.00016195560960637783, + "loss": 2.4814, + "step": 13556 + }, + { + "epoch": 0.58, + "grad_norm": 0.396484375, + "learning_rate": 0.00016195034406015104, + "loss": 2.5606, + "step": 13557 + }, + { + "epoch": 0.58, + "grad_norm": 0.44140625, + "learning_rate": 0.00016194507823517036, + "loss": 2.5118, + "step": 13558 + }, + { + "epoch": 0.58, + "grad_norm": 0.498046875, + "learning_rate": 0.00016193981213145944, + "loss": 2.6691, + "step": 13559 + }, + { + "epoch": 0.58, + "grad_norm": 0.49609375, + "learning_rate": 0.00016193454574904198, + "loss": 2.5708, + "step": 13560 + }, + { + "epoch": 0.58, + "grad_norm": 0.4609375, + "learning_rate": 0.00016192927908794172, + "loss": 2.6019, + "step": 13561 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016192401214818234, + "loss": 2.6504, + "step": 13562 + }, + { + "epoch": 0.58, + "grad_norm": 0.482421875, + "learning_rate": 0.00016191874492978752, + "loss": 2.462, + "step": 13563 + }, + { + "epoch": 0.58, + "grad_norm": 0.482421875, + "learning_rate": 0.000161913477432781, + "loss": 2.5481, + "step": 13564 + }, + { + "epoch": 0.58, + "grad_norm": 0.443359375, + "learning_rate": 0.00016190820965718644, + "loss": 2.5997, + "step": 13565 + }, + { + "epoch": 0.58, + "grad_norm": 0.44140625, + "learning_rate": 0.0001619029416030276, + "loss": 2.568, + "step": 13566 + }, + { + "epoch": 0.58, + "grad_norm": 0.46875, + "learning_rate": 0.00016189767327032812, + "loss": 2.6423, + "step": 13567 + }, + { + "epoch": 0.58, + "grad_norm": 0.45703125, + "learning_rate": 0.00016189240465911173, + "loss": 2.5914, + "step": 13568 + }, + { + "epoch": 0.58, + "grad_norm": 0.46484375, + "learning_rate": 0.00016188713576940214, + "loss": 2.7243, + "step": 13569 + }, + { + "epoch": 0.58, + "grad_norm": 0.44921875, + "learning_rate": 0.00016188186660122305, + "loss": 2.5808, + "step": 13570 + }, + { + "epoch": 0.58, + "grad_norm": 0.45703125, + "learning_rate": 0.00016187659715459818, + "loss": 2.6253, + "step": 13571 + }, + { + "epoch": 0.58, + "grad_norm": 0.470703125, + "learning_rate": 0.00016187132742955123, + "loss": 2.5749, + "step": 13572 + }, + { + "epoch": 0.58, + "grad_norm": 0.453125, + "learning_rate": 0.0001618660574261059, + "loss": 2.6756, + "step": 13573 + }, + { + "epoch": 0.58, + "grad_norm": 0.4453125, + "learning_rate": 0.00016186078714428598, + "loss": 2.6569, + "step": 13574 + }, + { + "epoch": 0.58, + "grad_norm": 0.51171875, + "learning_rate": 0.00016185551658411506, + "loss": 2.7789, + "step": 13575 + }, + { + "epoch": 0.58, + "grad_norm": 0.466796875, + "learning_rate": 0.00016185024574561695, + "loss": 2.7008, + "step": 13576 + }, + { + "epoch": 0.58, + "grad_norm": 0.46484375, + "learning_rate": 0.00016184497462881533, + "loss": 2.5634, + "step": 13577 + }, + { + "epoch": 0.58, + "grad_norm": 0.458984375, + "learning_rate": 0.00016183970323373393, + "loss": 2.5507, + "step": 13578 + }, + { + "epoch": 0.58, + "grad_norm": 0.490234375, + "learning_rate": 0.0001618344315603965, + "loss": 2.5951, + "step": 13579 + }, + { + "epoch": 0.58, + "grad_norm": 0.478515625, + "learning_rate": 0.00016182915960882667, + "loss": 2.591, + "step": 13580 + }, + { + "epoch": 0.58, + "grad_norm": 0.484375, + "learning_rate": 0.00016182388737904827, + "loss": 2.6253, + "step": 13581 + }, + { + "epoch": 0.58, + "grad_norm": 0.439453125, + "learning_rate": 0.00016181861487108492, + "loss": 2.5904, + "step": 13582 + }, + { + "epoch": 0.58, + "grad_norm": 0.482421875, + "learning_rate": 0.0001618133420849604, + "loss": 2.5922, + "step": 13583 + }, + { + "epoch": 0.58, + "grad_norm": 0.45703125, + "learning_rate": 0.00016180806902069846, + "loss": 2.6887, + "step": 13584 + }, + { + "epoch": 0.58, + "grad_norm": 0.43359375, + "learning_rate": 0.0001618027956783228, + "loss": 2.4539, + "step": 13585 + }, + { + "epoch": 0.58, + "grad_norm": 0.439453125, + "learning_rate": 0.00016179752205785713, + "loss": 2.6999, + "step": 13586 + }, + { + "epoch": 0.58, + "grad_norm": 0.50390625, + "learning_rate": 0.0001617922481593252, + "loss": 2.695, + "step": 13587 + }, + { + "epoch": 0.58, + "grad_norm": 0.447265625, + "learning_rate": 0.00016178697398275073, + "loss": 2.6087, + "step": 13588 + }, + { + "epoch": 0.58, + "grad_norm": 0.443359375, + "learning_rate": 0.00016178169952815746, + "loss": 2.5561, + "step": 13589 + }, + { + "epoch": 0.58, + "grad_norm": 0.447265625, + "learning_rate": 0.0001617764247955691, + "loss": 2.5968, + "step": 13590 + }, + { + "epoch": 0.58, + "grad_norm": 0.46484375, + "learning_rate": 0.00016177114978500943, + "loss": 2.5385, + "step": 13591 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016176587449650217, + "loss": 2.5605, + "step": 13592 + }, + { + "epoch": 0.58, + "grad_norm": 0.47265625, + "learning_rate": 0.00016176059893007103, + "loss": 2.6879, + "step": 13593 + }, + { + "epoch": 0.58, + "grad_norm": 0.458984375, + "learning_rate": 0.00016175532308573977, + "loss": 2.6759, + "step": 13594 + }, + { + "epoch": 0.58, + "grad_norm": 0.44921875, + "learning_rate": 0.0001617500469635321, + "loss": 2.7035, + "step": 13595 + }, + { + "epoch": 0.58, + "grad_norm": 0.5078125, + "learning_rate": 0.00016174477056347185, + "loss": 2.6448, + "step": 13596 + }, + { + "epoch": 0.58, + "grad_norm": 0.478515625, + "learning_rate": 0.00016173949388558265, + "loss": 2.4888, + "step": 13597 + }, + { + "epoch": 0.58, + "grad_norm": 0.453125, + "learning_rate": 0.0001617342169298883, + "loss": 2.6085, + "step": 13598 + }, + { + "epoch": 0.58, + "grad_norm": 0.4140625, + "learning_rate": 0.00016172893969641251, + "loss": 2.5239, + "step": 13599 + }, + { + "epoch": 0.58, + "grad_norm": 0.412109375, + "learning_rate": 0.00016172366218517912, + "loss": 2.5087, + "step": 13600 + }, + { + "epoch": 0.58, + "grad_norm": 0.431640625, + "learning_rate": 0.00016171838439621173, + "loss": 2.4712, + "step": 13601 + }, + { + "epoch": 0.58, + "grad_norm": 0.494140625, + "learning_rate": 0.0001617131063295342, + "loss": 2.5844, + "step": 13602 + }, + { + "epoch": 0.58, + "grad_norm": 0.470703125, + "learning_rate": 0.00016170782798517026, + "loss": 2.6307, + "step": 13603 + }, + { + "epoch": 0.58, + "grad_norm": 0.43359375, + "learning_rate": 0.00016170254936314363, + "loss": 2.4016, + "step": 13604 + }, + { + "epoch": 0.58, + "grad_norm": 0.427734375, + "learning_rate": 0.00016169727046347806, + "loss": 2.5022, + "step": 13605 + }, + { + "epoch": 0.58, + "grad_norm": 0.443359375, + "learning_rate": 0.00016169199128619735, + "loss": 2.6728, + "step": 13606 + }, + { + "epoch": 0.58, + "grad_norm": 0.4140625, + "learning_rate": 0.00016168671183132522, + "loss": 2.5545, + "step": 13607 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016168143209888543, + "loss": 2.5323, + "step": 13608 + }, + { + "epoch": 0.58, + "grad_norm": 0.447265625, + "learning_rate": 0.00016167615208890172, + "loss": 2.4794, + "step": 13609 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016167087180139785, + "loss": 2.3546, + "step": 13610 + }, + { + "epoch": 0.58, + "grad_norm": 0.466796875, + "learning_rate": 0.00016166559123639764, + "loss": 2.5548, + "step": 13611 + }, + { + "epoch": 0.58, + "grad_norm": 0.50390625, + "learning_rate": 0.0001616603103939248, + "loss": 2.5595, + "step": 13612 + }, + { + "epoch": 0.58, + "grad_norm": 0.423828125, + "learning_rate": 0.00016165502927400306, + "loss": 2.6359, + "step": 13613 + }, + { + "epoch": 0.58, + "grad_norm": 0.50390625, + "learning_rate": 0.00016164974787665628, + "loss": 2.6527, + "step": 13614 + }, + { + "epoch": 0.58, + "grad_norm": 0.51953125, + "learning_rate": 0.0001616444662019081, + "loss": 2.6, + "step": 13615 + }, + { + "epoch": 0.58, + "grad_norm": 0.5, + "learning_rate": 0.00016163918424978238, + "loss": 2.5243, + "step": 13616 + }, + { + "epoch": 0.58, + "grad_norm": 0.54296875, + "learning_rate": 0.00016163390202030285, + "loss": 2.566, + "step": 13617 + }, + { + "epoch": 0.58, + "grad_norm": 0.4609375, + "learning_rate": 0.00016162861951349333, + "loss": 2.5797, + "step": 13618 + }, + { + "epoch": 0.58, + "grad_norm": 0.474609375, + "learning_rate": 0.0001616233367293775, + "loss": 2.7122, + "step": 13619 + }, + { + "epoch": 0.58, + "grad_norm": 0.498046875, + "learning_rate": 0.00016161805366797917, + "loss": 2.667, + "step": 13620 + }, + { + "epoch": 0.58, + "grad_norm": 0.4375, + "learning_rate": 0.00016161277032932214, + "loss": 2.374, + "step": 13621 + }, + { + "epoch": 0.58, + "grad_norm": 0.48828125, + "learning_rate": 0.00016160748671343017, + "loss": 2.6137, + "step": 13622 + }, + { + "epoch": 0.58, + "grad_norm": 0.427734375, + "learning_rate": 0.000161602202820327, + "loss": 2.52, + "step": 13623 + }, + { + "epoch": 0.58, + "grad_norm": 0.482421875, + "learning_rate": 0.00016159691865003643, + "loss": 2.4791, + "step": 13624 + }, + { + "epoch": 0.58, + "grad_norm": 0.51171875, + "learning_rate": 0.00016159163420258223, + "loss": 2.6272, + "step": 13625 + }, + { + "epoch": 0.58, + "grad_norm": 0.4453125, + "learning_rate": 0.0001615863494779882, + "loss": 2.6382, + "step": 13626 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016158106447627813, + "loss": 2.6061, + "step": 13627 + }, + { + "epoch": 0.58, + "grad_norm": 0.51171875, + "learning_rate": 0.00016157577919747573, + "loss": 2.6279, + "step": 13628 + }, + { + "epoch": 0.58, + "grad_norm": 0.455078125, + "learning_rate": 0.00016157049364160485, + "loss": 2.6791, + "step": 13629 + }, + { + "epoch": 0.58, + "grad_norm": 0.48828125, + "learning_rate": 0.00016156520780868924, + "loss": 2.6218, + "step": 13630 + }, + { + "epoch": 0.58, + "grad_norm": 0.48046875, + "learning_rate": 0.00016155992169875272, + "loss": 2.5944, + "step": 13631 + }, + { + "epoch": 0.58, + "grad_norm": 0.466796875, + "learning_rate": 0.00016155463531181903, + "loss": 2.5965, + "step": 13632 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016154934864791195, + "loss": 2.6387, + "step": 13633 + }, + { + "epoch": 0.58, + "grad_norm": 0.46875, + "learning_rate": 0.00016154406170705533, + "loss": 2.5931, + "step": 13634 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.0001615387744892729, + "loss": 2.6198, + "step": 13635 + }, + { + "epoch": 0.58, + "grad_norm": 0.4609375, + "learning_rate": 0.00016153348699458854, + "loss": 2.6191, + "step": 13636 + }, + { + "epoch": 0.58, + "grad_norm": 0.46875, + "learning_rate": 0.00016152819922302588, + "loss": 2.6536, + "step": 13637 + }, + { + "epoch": 0.58, + "grad_norm": 0.51171875, + "learning_rate": 0.00016152291117460887, + "loss": 2.425, + "step": 13638 + }, + { + "epoch": 0.58, + "grad_norm": 0.453125, + "learning_rate": 0.00016151762284936122, + "loss": 2.5936, + "step": 13639 + }, + { + "epoch": 0.58, + "grad_norm": 0.478515625, + "learning_rate": 0.00016151233424730677, + "loss": 2.5784, + "step": 13640 + }, + { + "epoch": 0.58, + "grad_norm": 0.49609375, + "learning_rate": 0.0001615070453684693, + "loss": 2.688, + "step": 13641 + }, + { + "epoch": 0.58, + "grad_norm": 0.4296875, + "learning_rate": 0.00016150175621287258, + "loss": 2.64, + "step": 13642 + }, + { + "epoch": 0.58, + "grad_norm": 0.44140625, + "learning_rate": 0.00016149646678054045, + "loss": 2.6573, + "step": 13643 + }, + { + "epoch": 0.58, + "grad_norm": 0.5625, + "learning_rate": 0.00016149117707149667, + "loss": 2.5922, + "step": 13644 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.0001614858870857651, + "loss": 2.5845, + "step": 13645 + }, + { + "epoch": 0.58, + "grad_norm": 0.466796875, + "learning_rate": 0.00016148059682336949, + "loss": 2.5804, + "step": 13646 + }, + { + "epoch": 0.58, + "grad_norm": 0.478515625, + "learning_rate": 0.00016147530628433368, + "loss": 2.7147, + "step": 13647 + }, + { + "epoch": 0.58, + "grad_norm": 0.474609375, + "learning_rate": 0.00016147001546868144, + "loss": 2.6499, + "step": 13648 + }, + { + "epoch": 0.58, + "grad_norm": 0.490234375, + "learning_rate": 0.00016146472437643657, + "loss": 2.5883, + "step": 13649 + }, + { + "epoch": 0.58, + "grad_norm": 0.515625, + "learning_rate": 0.00016145943300762295, + "loss": 2.6115, + "step": 13650 + }, + { + "epoch": 0.58, + "grad_norm": 0.5390625, + "learning_rate": 0.00016145414136226432, + "loss": 2.6517, + "step": 13651 + }, + { + "epoch": 0.58, + "grad_norm": 0.412109375, + "learning_rate": 0.00016144884944038453, + "loss": 2.5863, + "step": 13652 + }, + { + "epoch": 0.58, + "grad_norm": 0.41796875, + "learning_rate": 0.00016144355724200736, + "loss": 2.5385, + "step": 13653 + }, + { + "epoch": 0.58, + "grad_norm": 0.4609375, + "learning_rate": 0.00016143826476715663, + "loss": 2.6859, + "step": 13654 + }, + { + "epoch": 0.58, + "grad_norm": 0.453125, + "learning_rate": 0.00016143297201585613, + "loss": 2.6421, + "step": 13655 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016142767898812978, + "loss": 2.6985, + "step": 13656 + }, + { + "epoch": 0.58, + "grad_norm": 0.435546875, + "learning_rate": 0.00016142238568400127, + "loss": 2.5718, + "step": 13657 + }, + { + "epoch": 0.58, + "grad_norm": 0.48046875, + "learning_rate": 0.00016141709210349453, + "loss": 2.6497, + "step": 13658 + }, + { + "epoch": 0.58, + "grad_norm": 0.4375, + "learning_rate": 0.00016141179824663327, + "loss": 2.4855, + "step": 13659 + }, + { + "epoch": 0.58, + "grad_norm": 0.4453125, + "learning_rate": 0.00016140650411344137, + "loss": 2.5193, + "step": 13660 + }, + { + "epoch": 0.58, + "grad_norm": 0.390625, + "learning_rate": 0.00016140120970394267, + "loss": 2.5678, + "step": 13661 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016139591501816093, + "loss": 2.6714, + "step": 13662 + }, + { + "epoch": 0.58, + "grad_norm": 0.44140625, + "learning_rate": 0.00016139062005612003, + "loss": 2.5639, + "step": 13663 + }, + { + "epoch": 0.58, + "grad_norm": 0.427734375, + "learning_rate": 0.00016138532481784377, + "loss": 2.677, + "step": 13664 + }, + { + "epoch": 0.58, + "grad_norm": 0.431640625, + "learning_rate": 0.00016138002930335597, + "loss": 2.6095, + "step": 13665 + }, + { + "epoch": 0.58, + "grad_norm": 0.48046875, + "learning_rate": 0.00016137473351268046, + "loss": 2.6782, + "step": 13666 + }, + { + "epoch": 0.58, + "grad_norm": 0.458984375, + "learning_rate": 0.00016136943744584113, + "loss": 2.551, + "step": 13667 + }, + { + "epoch": 0.58, + "grad_norm": 0.443359375, + "learning_rate": 0.00016136414110286173, + "loss": 2.6806, + "step": 13668 + }, + { + "epoch": 0.58, + "grad_norm": 0.447265625, + "learning_rate": 0.0001613588444837661, + "loss": 2.5689, + "step": 13669 + }, + { + "epoch": 0.58, + "grad_norm": 0.478515625, + "learning_rate": 0.00016135354758857812, + "loss": 2.611, + "step": 13670 + }, + { + "epoch": 0.58, + "grad_norm": 0.470703125, + "learning_rate": 0.0001613482504173216, + "loss": 2.5609, + "step": 13671 + }, + { + "epoch": 0.58, + "grad_norm": 0.455078125, + "learning_rate": 0.0001613429529700204, + "loss": 2.6336, + "step": 13672 + }, + { + "epoch": 0.58, + "grad_norm": 0.5078125, + "learning_rate": 0.00016133765524669827, + "loss": 2.6651, + "step": 13673 + }, + { + "epoch": 0.58, + "grad_norm": 0.51171875, + "learning_rate": 0.00016133235724737912, + "loss": 2.6033, + "step": 13674 + }, + { + "epoch": 0.58, + "grad_norm": 0.48046875, + "learning_rate": 0.00016132705897208677, + "loss": 2.6559, + "step": 13675 + }, + { + "epoch": 0.58, + "grad_norm": 0.50390625, + "learning_rate": 0.0001613217604208451, + "loss": 2.5424, + "step": 13676 + }, + { + "epoch": 0.58, + "grad_norm": 0.494140625, + "learning_rate": 0.00016131646159367794, + "loss": 2.6442, + "step": 13677 + }, + { + "epoch": 0.58, + "grad_norm": 0.484375, + "learning_rate": 0.00016131116249060905, + "loss": 2.5452, + "step": 13678 + }, + { + "epoch": 0.58, + "grad_norm": 0.5390625, + "learning_rate": 0.00016130586311166237, + "loss": 2.6923, + "step": 13679 + }, + { + "epoch": 0.58, + "grad_norm": 0.51953125, + "learning_rate": 0.0001613005634568617, + "loss": 2.8029, + "step": 13680 + }, + { + "epoch": 0.58, + "grad_norm": 0.404296875, + "learning_rate": 0.0001612952635262309, + "loss": 2.4418, + "step": 13681 + }, + { + "epoch": 0.58, + "grad_norm": 0.51953125, + "learning_rate": 0.00016128996331979381, + "loss": 2.6599, + "step": 13682 + }, + { + "epoch": 0.58, + "grad_norm": 0.42578125, + "learning_rate": 0.00016128466283757428, + "loss": 2.5089, + "step": 13683 + }, + { + "epoch": 0.58, + "grad_norm": 0.4296875, + "learning_rate": 0.0001612793620795962, + "loss": 2.5649, + "step": 13684 + }, + { + "epoch": 0.58, + "grad_norm": 0.515625, + "learning_rate": 0.00016127406104588334, + "loss": 2.5754, + "step": 13685 + }, + { + "epoch": 0.58, + "grad_norm": 0.46875, + "learning_rate": 0.00016126875973645962, + "loss": 2.6642, + "step": 13686 + }, + { + "epoch": 0.58, + "grad_norm": 0.43359375, + "learning_rate": 0.0001612634581513489, + "loss": 2.6149, + "step": 13687 + }, + { + "epoch": 0.58, + "grad_norm": 0.44921875, + "learning_rate": 0.00016125815629057497, + "loss": 2.5614, + "step": 13688 + }, + { + "epoch": 0.58, + "grad_norm": 0.458984375, + "learning_rate": 0.00016125285415416173, + "loss": 2.4891, + "step": 13689 + }, + { + "epoch": 0.58, + "grad_norm": 0.447265625, + "learning_rate": 0.00016124755174213302, + "loss": 2.5881, + "step": 13690 + }, + { + "epoch": 0.58, + "grad_norm": 0.427734375, + "learning_rate": 0.00016124224905451272, + "loss": 2.445, + "step": 13691 + }, + { + "epoch": 0.58, + "grad_norm": 0.486328125, + "learning_rate": 0.00016123694609132468, + "loss": 2.6323, + "step": 13692 + }, + { + "epoch": 0.58, + "grad_norm": 0.41015625, + "learning_rate": 0.0001612316428525928, + "loss": 2.4637, + "step": 13693 + }, + { + "epoch": 0.58, + "grad_norm": 0.470703125, + "learning_rate": 0.00016122633933834086, + "loss": 2.5995, + "step": 13694 + }, + { + "epoch": 0.58, + "grad_norm": 0.484375, + "learning_rate": 0.0001612210355485928, + "loss": 2.6755, + "step": 13695 + }, + { + "epoch": 0.58, + "grad_norm": 0.416015625, + "learning_rate": 0.00016121573148337242, + "loss": 2.5446, + "step": 13696 + }, + { + "epoch": 0.58, + "grad_norm": 0.486328125, + "learning_rate": 0.00016121042714270365, + "loss": 2.6554, + "step": 13697 + }, + { + "epoch": 0.58, + "grad_norm": 0.423828125, + "learning_rate": 0.0001612051225266103, + "loss": 2.6438, + "step": 13698 + }, + { + "epoch": 0.58, + "grad_norm": 0.498046875, + "learning_rate": 0.0001611998176351163, + "loss": 2.6454, + "step": 13699 + }, + { + "epoch": 0.58, + "grad_norm": 0.484375, + "learning_rate": 0.00016119451246824544, + "loss": 2.525, + "step": 13700 + }, + { + "epoch": 0.58, + "grad_norm": 0.40625, + "learning_rate": 0.00016118920702602166, + "loss": 2.5479, + "step": 13701 + }, + { + "epoch": 0.58, + "grad_norm": 0.42578125, + "learning_rate": 0.00016118390130846882, + "loss": 2.6896, + "step": 13702 + }, + { + "epoch": 0.58, + "grad_norm": 0.462890625, + "learning_rate": 0.00016117859531561077, + "loss": 2.6455, + "step": 13703 + }, + { + "epoch": 0.58, + "grad_norm": 0.443359375, + "learning_rate": 0.0001611732890474714, + "loss": 2.6223, + "step": 13704 + }, + { + "epoch": 0.58, + "grad_norm": 0.44140625, + "learning_rate": 0.0001611679825040746, + "loss": 2.5645, + "step": 13705 + }, + { + "epoch": 0.58, + "grad_norm": 0.4609375, + "learning_rate": 0.00016116267568544421, + "loss": 2.6906, + "step": 13706 + }, + { + "epoch": 0.58, + "grad_norm": 0.431640625, + "learning_rate": 0.00016115736859160417, + "loss": 2.5339, + "step": 13707 + }, + { + "epoch": 0.58, + "grad_norm": 0.447265625, + "learning_rate": 0.00016115206122257832, + "loss": 2.4761, + "step": 13708 + }, + { + "epoch": 0.58, + "grad_norm": 0.46875, + "learning_rate": 0.0001611467535783905, + "loss": 2.5084, + "step": 13709 + }, + { + "epoch": 0.58, + "grad_norm": 0.46484375, + "learning_rate": 0.00016114144565906467, + "loss": 2.6304, + "step": 13710 + }, + { + "epoch": 0.58, + "grad_norm": 0.47265625, + "learning_rate": 0.00016113613746462465, + "loss": 2.5629, + "step": 13711 + }, + { + "epoch": 0.58, + "grad_norm": 0.47265625, + "learning_rate": 0.00016113082899509438, + "loss": 2.4704, + "step": 13712 + }, + { + "epoch": 0.58, + "grad_norm": 0.4453125, + "learning_rate": 0.0001611255202504977, + "loss": 2.5943, + "step": 13713 + }, + { + "epoch": 0.58, + "grad_norm": 0.43359375, + "learning_rate": 0.00016112021123085852, + "loss": 2.642, + "step": 13714 + }, + { + "epoch": 0.58, + "grad_norm": 0.470703125, + "learning_rate": 0.00016111490193620076, + "loss": 2.6324, + "step": 13715 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016110959236654825, + "loss": 2.6755, + "step": 13716 + }, + { + "epoch": 0.58, + "grad_norm": 0.4453125, + "learning_rate": 0.0001611042825219249, + "loss": 2.4962, + "step": 13717 + }, + { + "epoch": 0.58, + "grad_norm": 0.45703125, + "learning_rate": 0.0001610989724023546, + "loss": 2.5101, + "step": 13718 + }, + { + "epoch": 0.58, + "grad_norm": 0.451171875, + "learning_rate": 0.0001610936620078613, + "loss": 2.4741, + "step": 13719 + }, + { + "epoch": 0.58, + "grad_norm": 0.4453125, + "learning_rate": 0.0001610883513384688, + "loss": 2.6072, + "step": 13720 + }, + { + "epoch": 0.58, + "grad_norm": 0.52734375, + "learning_rate": 0.00016108304039420108, + "loss": 2.6856, + "step": 13721 + }, + { + "epoch": 0.58, + "grad_norm": 0.466796875, + "learning_rate": 0.00016107772917508197, + "loss": 2.5532, + "step": 13722 + }, + { + "epoch": 0.58, + "grad_norm": 0.4765625, + "learning_rate": 0.00016107241768113543, + "loss": 2.6372, + "step": 13723 + }, + { + "epoch": 0.58, + "grad_norm": 0.51171875, + "learning_rate": 0.00016106710591238532, + "loss": 2.5891, + "step": 13724 + }, + { + "epoch": 0.58, + "grad_norm": 0.404296875, + "learning_rate": 0.00016106179386885555, + "loss": 2.4743, + "step": 13725 + }, + { + "epoch": 0.58, + "grad_norm": 0.494140625, + "learning_rate": 0.00016105648155057, + "loss": 2.7237, + "step": 13726 + }, + { + "epoch": 0.58, + "grad_norm": 0.466796875, + "learning_rate": 0.00016105116895755265, + "loss": 2.7364, + "step": 13727 + }, + { + "epoch": 0.58, + "grad_norm": 0.443359375, + "learning_rate": 0.00016104585608982732, + "loss": 2.45, + "step": 13728 + }, + { + "epoch": 0.59, + "grad_norm": 0.515625, + "learning_rate": 0.00016104054294741792, + "loss": 2.5652, + "step": 13729 + }, + { + "epoch": 0.59, + "grad_norm": 0.5, + "learning_rate": 0.00016103522953034841, + "loss": 2.5606, + "step": 13730 + }, + { + "epoch": 0.59, + "grad_norm": 0.462890625, + "learning_rate": 0.0001610299158386427, + "loss": 2.5171, + "step": 13731 + }, + { + "epoch": 0.59, + "grad_norm": 0.453125, + "learning_rate": 0.0001610246018723246, + "loss": 2.5313, + "step": 13732 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.00016101928763141818, + "loss": 2.4, + "step": 13733 + }, + { + "epoch": 0.59, + "grad_norm": 0.466796875, + "learning_rate": 0.0001610139731159472, + "loss": 2.6408, + "step": 13734 + }, + { + "epoch": 0.59, + "grad_norm": 0.4921875, + "learning_rate": 0.00016100865832593568, + "loss": 2.7435, + "step": 13735 + }, + { + "epoch": 0.59, + "grad_norm": 0.44921875, + "learning_rate": 0.00016100334326140747, + "loss": 2.6347, + "step": 13736 + }, + { + "epoch": 0.59, + "grad_norm": 0.453125, + "learning_rate": 0.00016099802792238652, + "loss": 2.6566, + "step": 13737 + }, + { + "epoch": 0.59, + "grad_norm": 0.498046875, + "learning_rate": 0.00016099271230889676, + "loss": 2.6961, + "step": 13738 + }, + { + "epoch": 0.59, + "grad_norm": 0.43359375, + "learning_rate": 0.00016098739642096202, + "loss": 2.597, + "step": 13739 + }, + { + "epoch": 0.59, + "grad_norm": 0.43359375, + "learning_rate": 0.00016098208025860632, + "loss": 2.5514, + "step": 13740 + }, + { + "epoch": 0.59, + "grad_norm": 0.431640625, + "learning_rate": 0.00016097676382185357, + "loss": 2.517, + "step": 13741 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.00016097144711072763, + "loss": 2.5843, + "step": 13742 + }, + { + "epoch": 0.59, + "grad_norm": 0.45703125, + "learning_rate": 0.00016096613012525247, + "loss": 2.6895, + "step": 13743 + }, + { + "epoch": 0.59, + "grad_norm": 0.474609375, + "learning_rate": 0.000160960812865452, + "loss": 2.5678, + "step": 13744 + }, + { + "epoch": 0.59, + "grad_norm": 0.5, + "learning_rate": 0.00016095549533135016, + "loss": 2.7487, + "step": 13745 + }, + { + "epoch": 0.59, + "grad_norm": 0.482421875, + "learning_rate": 0.0001609501775229709, + "loss": 2.7502, + "step": 13746 + }, + { + "epoch": 0.59, + "grad_norm": 0.447265625, + "learning_rate": 0.0001609448594403381, + "loss": 2.6045, + "step": 13747 + }, + { + "epoch": 0.59, + "grad_norm": 0.484375, + "learning_rate": 0.00016093954108347566, + "loss": 2.5467, + "step": 13748 + }, + { + "epoch": 0.59, + "grad_norm": 0.48046875, + "learning_rate": 0.00016093422245240762, + "loss": 2.623, + "step": 13749 + }, + { + "epoch": 0.59, + "grad_norm": 0.5078125, + "learning_rate": 0.00016092890354715776, + "loss": 2.7305, + "step": 13750 + }, + { + "epoch": 0.59, + "grad_norm": 0.453125, + "learning_rate": 0.0001609235843677502, + "loss": 2.5981, + "step": 13751 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.00016091826491420873, + "loss": 2.6805, + "step": 13752 + }, + { + "epoch": 0.59, + "grad_norm": 0.4453125, + "learning_rate": 0.00016091294518655734, + "loss": 2.4692, + "step": 13753 + }, + { + "epoch": 0.59, + "grad_norm": 0.447265625, + "learning_rate": 0.0001609076251848199, + "loss": 2.4656, + "step": 13754 + }, + { + "epoch": 0.59, + "grad_norm": 0.53125, + "learning_rate": 0.0001609023049090205, + "loss": 2.7411, + "step": 13755 + }, + { + "epoch": 0.59, + "grad_norm": 0.453125, + "learning_rate": 0.00016089698435918292, + "loss": 2.6726, + "step": 13756 + }, + { + "epoch": 0.59, + "grad_norm": 0.5703125, + "learning_rate": 0.0001608916635353312, + "loss": 2.6913, + "step": 13757 + }, + { + "epoch": 0.59, + "grad_norm": 0.462890625, + "learning_rate": 0.0001608863424374892, + "loss": 2.6501, + "step": 13758 + }, + { + "epoch": 0.59, + "grad_norm": 0.46484375, + "learning_rate": 0.00016088102106568097, + "loss": 2.6458, + "step": 13759 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016087569941993036, + "loss": 2.6448, + "step": 13760 + }, + { + "epoch": 0.59, + "grad_norm": 0.466796875, + "learning_rate": 0.00016087037750026138, + "loss": 2.5125, + "step": 13761 + }, + { + "epoch": 0.59, + "grad_norm": 0.439453125, + "learning_rate": 0.00016086505530669791, + "loss": 2.6855, + "step": 13762 + }, + { + "epoch": 0.59, + "grad_norm": 0.5078125, + "learning_rate": 0.00016085973283926395, + "loss": 2.4866, + "step": 13763 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016085441009798342, + "loss": 2.5896, + "step": 13764 + }, + { + "epoch": 0.59, + "grad_norm": 0.419921875, + "learning_rate": 0.0001608490870828803, + "loss": 2.5021, + "step": 13765 + }, + { + "epoch": 0.59, + "grad_norm": 0.4765625, + "learning_rate": 0.00016084376379397854, + "loss": 2.6954, + "step": 13766 + }, + { + "epoch": 0.59, + "grad_norm": 0.484375, + "learning_rate": 0.00016083844023130208, + "loss": 2.5733, + "step": 13767 + }, + { + "epoch": 0.59, + "grad_norm": 0.439453125, + "learning_rate": 0.00016083311639487484, + "loss": 2.6085, + "step": 13768 + }, + { + "epoch": 0.59, + "grad_norm": 0.55078125, + "learning_rate": 0.00016082779228472082, + "loss": 2.6467, + "step": 13769 + }, + { + "epoch": 0.59, + "grad_norm": 0.474609375, + "learning_rate": 0.00016082246790086397, + "loss": 2.6002, + "step": 13770 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016081714324332825, + "loss": 2.6995, + "step": 13771 + }, + { + "epoch": 0.59, + "grad_norm": 0.44921875, + "learning_rate": 0.0001608118183121376, + "loss": 2.5635, + "step": 13772 + }, + { + "epoch": 0.59, + "grad_norm": 0.5390625, + "learning_rate": 0.00016080649310731598, + "loss": 2.6287, + "step": 13773 + }, + { + "epoch": 0.59, + "grad_norm": 0.48046875, + "learning_rate": 0.00016080116762888738, + "loss": 2.723, + "step": 13774 + }, + { + "epoch": 0.59, + "grad_norm": 0.482421875, + "learning_rate": 0.00016079584187687576, + "loss": 2.6141, + "step": 13775 + }, + { + "epoch": 0.59, + "grad_norm": 0.451171875, + "learning_rate": 0.00016079051585130504, + "loss": 2.4842, + "step": 13776 + }, + { + "epoch": 0.59, + "grad_norm": 0.546875, + "learning_rate": 0.00016078518955219923, + "loss": 2.8243, + "step": 13777 + }, + { + "epoch": 0.59, + "grad_norm": 0.4921875, + "learning_rate": 0.0001607798629795823, + "loss": 2.7981, + "step": 13778 + }, + { + "epoch": 0.59, + "grad_norm": 0.494140625, + "learning_rate": 0.00016077453613347816, + "loss": 2.5889, + "step": 13779 + }, + { + "epoch": 0.59, + "grad_norm": 0.4765625, + "learning_rate": 0.00016076920901391084, + "loss": 2.5733, + "step": 13780 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.00016076388162090428, + "loss": 2.6013, + "step": 13781 + }, + { + "epoch": 0.59, + "grad_norm": 0.4609375, + "learning_rate": 0.00016075855395448246, + "loss": 2.6588, + "step": 13782 + }, + { + "epoch": 0.59, + "grad_norm": 0.50390625, + "learning_rate": 0.00016075322601466938, + "loss": 2.6042, + "step": 13783 + }, + { + "epoch": 0.59, + "grad_norm": 0.474609375, + "learning_rate": 0.00016074789780148895, + "loss": 2.4749, + "step": 13784 + }, + { + "epoch": 0.59, + "grad_norm": 0.40234375, + "learning_rate": 0.00016074256931496522, + "loss": 2.5401, + "step": 13785 + }, + { + "epoch": 0.59, + "grad_norm": 0.466796875, + "learning_rate": 0.00016073724055512206, + "loss": 2.4546, + "step": 13786 + }, + { + "epoch": 0.59, + "grad_norm": 0.423828125, + "learning_rate": 0.0001607319115219836, + "loss": 2.6142, + "step": 13787 + }, + { + "epoch": 0.59, + "grad_norm": 0.40625, + "learning_rate": 0.00016072658221557366, + "loss": 2.5923, + "step": 13788 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016072125263591632, + "loss": 2.5978, + "step": 13789 + }, + { + "epoch": 0.59, + "grad_norm": 0.4765625, + "learning_rate": 0.00016071592278303553, + "loss": 2.5529, + "step": 13790 + }, + { + "epoch": 0.59, + "grad_norm": 0.453125, + "learning_rate": 0.0001607105926569553, + "loss": 2.6456, + "step": 13791 + }, + { + "epoch": 0.59, + "grad_norm": 0.416015625, + "learning_rate": 0.00016070526225769958, + "loss": 2.5237, + "step": 13792 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.00016069993158529235, + "loss": 2.5424, + "step": 13793 + }, + { + "epoch": 0.59, + "grad_norm": 0.421875, + "learning_rate": 0.00016069460063975762, + "loss": 2.5176, + "step": 13794 + }, + { + "epoch": 0.59, + "grad_norm": 0.486328125, + "learning_rate": 0.00016068926942111936, + "loss": 2.6062, + "step": 13795 + }, + { + "epoch": 0.59, + "grad_norm": 0.427734375, + "learning_rate": 0.00016068393792940158, + "loss": 2.619, + "step": 13796 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016067860616462824, + "loss": 2.5333, + "step": 13797 + }, + { + "epoch": 0.59, + "grad_norm": 0.44921875, + "learning_rate": 0.00016067327412682336, + "loss": 2.5328, + "step": 13798 + }, + { + "epoch": 0.59, + "grad_norm": 0.453125, + "learning_rate": 0.0001606679418160109, + "loss": 2.5415, + "step": 13799 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.0001606626092322149, + "loss": 2.5829, + "step": 13800 + }, + { + "epoch": 0.59, + "grad_norm": 0.478515625, + "learning_rate": 0.00016065727637545933, + "loss": 2.6295, + "step": 13801 + }, + { + "epoch": 0.59, + "grad_norm": 0.5625, + "learning_rate": 0.00016065194324576819, + "loss": 2.6828, + "step": 13802 + }, + { + "epoch": 0.59, + "grad_norm": 0.5078125, + "learning_rate": 0.00016064660984316543, + "loss": 2.7036, + "step": 13803 + }, + { + "epoch": 0.59, + "grad_norm": 0.47265625, + "learning_rate": 0.00016064127616767512, + "loss": 2.6205, + "step": 13804 + }, + { + "epoch": 0.59, + "grad_norm": 0.486328125, + "learning_rate": 0.00016063594221932124, + "loss": 2.5679, + "step": 13805 + }, + { + "epoch": 0.59, + "grad_norm": 0.443359375, + "learning_rate": 0.0001606306079981277, + "loss": 2.5588, + "step": 13806 + }, + { + "epoch": 0.59, + "grad_norm": 0.51171875, + "learning_rate": 0.0001606252735041187, + "loss": 2.6887, + "step": 13807 + }, + { + "epoch": 0.59, + "grad_norm": 0.54296875, + "learning_rate": 0.00016061993873731805, + "loss": 2.6925, + "step": 13808 + }, + { + "epoch": 0.59, + "grad_norm": 0.443359375, + "learning_rate": 0.0001606146036977499, + "loss": 2.5876, + "step": 13809 + }, + { + "epoch": 0.59, + "grad_norm": 0.466796875, + "learning_rate": 0.0001606092683854381, + "loss": 2.5469, + "step": 13810 + }, + { + "epoch": 0.59, + "grad_norm": 0.4296875, + "learning_rate": 0.0001606039328004068, + "loss": 2.5032, + "step": 13811 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.0001605985969426799, + "loss": 2.6856, + "step": 13812 + }, + { + "epoch": 0.59, + "grad_norm": 0.478515625, + "learning_rate": 0.0001605932608122815, + "loss": 2.6691, + "step": 13813 + }, + { + "epoch": 0.59, + "grad_norm": 0.462890625, + "learning_rate": 0.00016058792440923556, + "loss": 2.5739, + "step": 13814 + }, + { + "epoch": 0.59, + "grad_norm": 0.494140625, + "learning_rate": 0.0001605825877335661, + "loss": 2.6421, + "step": 13815 + }, + { + "epoch": 0.59, + "grad_norm": 0.484375, + "learning_rate": 0.00016057725078529714, + "loss": 2.496, + "step": 13816 + }, + { + "epoch": 0.59, + "grad_norm": 0.416015625, + "learning_rate": 0.0001605719135644527, + "loss": 2.5738, + "step": 13817 + }, + { + "epoch": 0.59, + "grad_norm": 0.404296875, + "learning_rate": 0.00016056657607105675, + "loss": 2.5391, + "step": 13818 + }, + { + "epoch": 0.59, + "grad_norm": 0.490234375, + "learning_rate": 0.00016056123830513337, + "loss": 2.6913, + "step": 13819 + }, + { + "epoch": 0.59, + "grad_norm": 0.5234375, + "learning_rate": 0.00016055590026670656, + "loss": 2.5855, + "step": 13820 + }, + { + "epoch": 0.59, + "grad_norm": 0.443359375, + "learning_rate": 0.0001605505619558003, + "loss": 2.5596, + "step": 13821 + }, + { + "epoch": 0.59, + "grad_norm": 0.41796875, + "learning_rate": 0.00016054522337243866, + "loss": 2.4593, + "step": 13822 + }, + { + "epoch": 0.59, + "grad_norm": 0.48046875, + "learning_rate": 0.00016053988451664563, + "loss": 2.7157, + "step": 13823 + }, + { + "epoch": 0.59, + "grad_norm": 0.404296875, + "learning_rate": 0.00016053454538844525, + "loss": 2.5605, + "step": 13824 + }, + { + "epoch": 0.59, + "grad_norm": 0.44921875, + "learning_rate": 0.00016052920598786155, + "loss": 2.6043, + "step": 13825 + }, + { + "epoch": 0.59, + "grad_norm": 0.482421875, + "learning_rate": 0.00016052386631491855, + "loss": 2.6676, + "step": 13826 + }, + { + "epoch": 0.59, + "grad_norm": 0.44140625, + "learning_rate": 0.00016051852636964021, + "loss": 2.6167, + "step": 13827 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.00016051318615205068, + "loss": 2.6177, + "step": 13828 + }, + { + "epoch": 0.59, + "grad_norm": 0.44921875, + "learning_rate": 0.00016050784566217389, + "loss": 2.6554, + "step": 13829 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016050250490003395, + "loss": 2.6615, + "step": 13830 + }, + { + "epoch": 0.59, + "grad_norm": 0.453125, + "learning_rate": 0.0001604971638656548, + "loss": 2.5389, + "step": 13831 + }, + { + "epoch": 0.59, + "grad_norm": 0.5, + "learning_rate": 0.0001604918225590606, + "loss": 2.587, + "step": 13832 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.00016048648098027522, + "loss": 2.6394, + "step": 13833 + }, + { + "epoch": 0.59, + "grad_norm": 0.486328125, + "learning_rate": 0.00016048113912932283, + "loss": 2.6172, + "step": 13834 + }, + { + "epoch": 0.59, + "grad_norm": 0.51171875, + "learning_rate": 0.00016047579700622738, + "loss": 2.6682, + "step": 13835 + }, + { + "epoch": 0.59, + "grad_norm": 0.416015625, + "learning_rate": 0.00016047045461101296, + "loss": 2.6149, + "step": 13836 + }, + { + "epoch": 0.59, + "grad_norm": 0.53515625, + "learning_rate": 0.00016046511194370362, + "loss": 2.4972, + "step": 13837 + }, + { + "epoch": 0.59, + "grad_norm": 0.439453125, + "learning_rate": 0.00016045976900432336, + "loss": 2.6111, + "step": 13838 + }, + { + "epoch": 0.59, + "grad_norm": 0.5234375, + "learning_rate": 0.00016045442579289623, + "loss": 2.6951, + "step": 13839 + }, + { + "epoch": 0.59, + "grad_norm": 0.451171875, + "learning_rate": 0.00016044908230944627, + "loss": 2.6912, + "step": 13840 + }, + { + "epoch": 0.59, + "grad_norm": 0.51171875, + "learning_rate": 0.00016044373855399755, + "loss": 2.6165, + "step": 13841 + }, + { + "epoch": 0.59, + "grad_norm": 0.462890625, + "learning_rate": 0.0001604383945265741, + "loss": 2.575, + "step": 13842 + }, + { + "epoch": 0.59, + "grad_norm": 0.4609375, + "learning_rate": 0.00016043305022719992, + "loss": 2.6421, + "step": 13843 + }, + { + "epoch": 0.59, + "grad_norm": 0.3984375, + "learning_rate": 0.00016042770565589913, + "loss": 2.5465, + "step": 13844 + }, + { + "epoch": 0.59, + "grad_norm": 0.431640625, + "learning_rate": 0.00016042236081269574, + "loss": 2.611, + "step": 13845 + }, + { + "epoch": 0.59, + "grad_norm": 0.484375, + "learning_rate": 0.00016041701569761383, + "loss": 2.5248, + "step": 13846 + }, + { + "epoch": 0.59, + "grad_norm": 0.478515625, + "learning_rate": 0.0001604116703106774, + "loss": 2.6629, + "step": 13847 + }, + { + "epoch": 0.59, + "grad_norm": 0.447265625, + "learning_rate": 0.00016040632465191053, + "loss": 2.6041, + "step": 13848 + }, + { + "epoch": 0.59, + "grad_norm": 0.439453125, + "learning_rate": 0.0001604009787213373, + "loss": 2.5353, + "step": 13849 + }, + { + "epoch": 0.59, + "grad_norm": 0.4609375, + "learning_rate": 0.00016039563251898174, + "loss": 2.5069, + "step": 13850 + }, + { + "epoch": 0.59, + "grad_norm": 0.51953125, + "learning_rate": 0.0001603902860448679, + "loss": 2.6675, + "step": 13851 + }, + { + "epoch": 0.59, + "grad_norm": 0.447265625, + "learning_rate": 0.00016038493929901982, + "loss": 2.6419, + "step": 13852 + }, + { + "epoch": 0.59, + "grad_norm": 0.47265625, + "learning_rate": 0.00016037959228146158, + "loss": 2.6087, + "step": 13853 + }, + { + "epoch": 0.59, + "grad_norm": 0.44140625, + "learning_rate": 0.0001603742449922173, + "loss": 2.6531, + "step": 13854 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.0001603688974313109, + "loss": 2.5321, + "step": 13855 + }, + { + "epoch": 0.59, + "grad_norm": 0.48046875, + "learning_rate": 0.00016036354959876657, + "loss": 2.7056, + "step": 13856 + }, + { + "epoch": 0.59, + "grad_norm": 0.419921875, + "learning_rate": 0.00016035820149460834, + "loss": 2.518, + "step": 13857 + }, + { + "epoch": 0.59, + "grad_norm": 0.447265625, + "learning_rate": 0.00016035285311886022, + "loss": 2.5551, + "step": 13858 + }, + { + "epoch": 0.59, + "grad_norm": 0.50390625, + "learning_rate": 0.00016034750447154635, + "loss": 2.5546, + "step": 13859 + }, + { + "epoch": 0.59, + "grad_norm": 0.458984375, + "learning_rate": 0.00016034215555269077, + "loss": 2.6771, + "step": 13860 + }, + { + "epoch": 0.59, + "grad_norm": 0.4609375, + "learning_rate": 0.0001603368063623175, + "loss": 2.6366, + "step": 13861 + }, + { + "epoch": 0.59, + "grad_norm": 0.4609375, + "learning_rate": 0.0001603314569004507, + "loss": 2.6502, + "step": 13862 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016032610716711437, + "loss": 2.6006, + "step": 13863 + }, + { + "epoch": 0.59, + "grad_norm": 0.498046875, + "learning_rate": 0.0001603207571623326, + "loss": 2.6038, + "step": 13864 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.00016031540688612948, + "loss": 2.6826, + "step": 13865 + }, + { + "epoch": 0.59, + "grad_norm": 0.42578125, + "learning_rate": 0.00016031005633852907, + "loss": 2.5343, + "step": 13866 + }, + { + "epoch": 0.59, + "grad_norm": 0.427734375, + "learning_rate": 0.00016030470551955546, + "loss": 2.6827, + "step": 13867 + }, + { + "epoch": 0.59, + "grad_norm": 0.408203125, + "learning_rate": 0.00016029935442923269, + "loss": 2.6153, + "step": 13868 + }, + { + "epoch": 0.59, + "grad_norm": 0.482421875, + "learning_rate": 0.00016029400306758485, + "loss": 2.6494, + "step": 13869 + }, + { + "epoch": 0.59, + "grad_norm": 0.474609375, + "learning_rate": 0.00016028865143463605, + "loss": 2.6014, + "step": 13870 + }, + { + "epoch": 0.59, + "grad_norm": 0.43359375, + "learning_rate": 0.00016028329953041035, + "loss": 2.5102, + "step": 13871 + }, + { + "epoch": 0.59, + "grad_norm": 0.451171875, + "learning_rate": 0.00016027794735493184, + "loss": 2.5023, + "step": 13872 + }, + { + "epoch": 0.59, + "grad_norm": 0.51953125, + "learning_rate": 0.00016027259490822458, + "loss": 2.6398, + "step": 13873 + }, + { + "epoch": 0.59, + "grad_norm": 0.46875, + "learning_rate": 0.00016026724219031272, + "loss": 2.7251, + "step": 13874 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.00016026188920122023, + "loss": 2.5922, + "step": 13875 + }, + { + "epoch": 0.59, + "grad_norm": 0.41796875, + "learning_rate": 0.0001602565359409713, + "loss": 2.5067, + "step": 13876 + }, + { + "epoch": 0.59, + "grad_norm": 0.458984375, + "learning_rate": 0.00016025118240958994, + "loss": 2.4526, + "step": 13877 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016024582860710033, + "loss": 2.674, + "step": 13878 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.00016024047453352646, + "loss": 2.4937, + "step": 13879 + }, + { + "epoch": 0.59, + "grad_norm": 0.4453125, + "learning_rate": 0.00016023512018889253, + "loss": 2.6083, + "step": 13880 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.0001602297655732225, + "loss": 2.5387, + "step": 13881 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.00016022441068654062, + "loss": 2.6592, + "step": 13882 + }, + { + "epoch": 0.59, + "grad_norm": 0.54296875, + "learning_rate": 0.00016021905552887083, + "loss": 2.6831, + "step": 13883 + }, + { + "epoch": 0.59, + "grad_norm": 0.427734375, + "learning_rate": 0.00016021370010023734, + "loss": 2.6737, + "step": 13884 + }, + { + "epoch": 0.59, + "grad_norm": 0.462890625, + "learning_rate": 0.00016020834440066416, + "loss": 2.5934, + "step": 13885 + }, + { + "epoch": 0.59, + "grad_norm": 0.44140625, + "learning_rate": 0.00016020298843017549, + "loss": 2.6438, + "step": 13886 + }, + { + "epoch": 0.59, + "grad_norm": 0.5234375, + "learning_rate": 0.00016019763218879534, + "loss": 2.7538, + "step": 13887 + }, + { + "epoch": 0.59, + "grad_norm": 0.478515625, + "learning_rate": 0.00016019227567654783, + "loss": 2.6691, + "step": 13888 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.0001601869188934571, + "loss": 2.5864, + "step": 13889 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.00016018156183954723, + "loss": 2.7065, + "step": 13890 + }, + { + "epoch": 0.59, + "grad_norm": 0.45703125, + "learning_rate": 0.0001601762045148423, + "loss": 2.6785, + "step": 13891 + }, + { + "epoch": 0.59, + "grad_norm": 0.40234375, + "learning_rate": 0.0001601708469193665, + "loss": 2.636, + "step": 13892 + }, + { + "epoch": 0.59, + "grad_norm": 0.4453125, + "learning_rate": 0.00016016548905314383, + "loss": 2.5229, + "step": 13893 + }, + { + "epoch": 0.59, + "grad_norm": 0.50390625, + "learning_rate": 0.0001601601309161985, + "loss": 2.479, + "step": 13894 + }, + { + "epoch": 0.59, + "grad_norm": 0.427734375, + "learning_rate": 0.00016015477250855452, + "loss": 2.5786, + "step": 13895 + }, + { + "epoch": 0.59, + "grad_norm": 0.51953125, + "learning_rate": 0.00016014941383023606, + "loss": 2.6967, + "step": 13896 + }, + { + "epoch": 0.59, + "grad_norm": 0.52734375, + "learning_rate": 0.0001601440548812672, + "loss": 2.6908, + "step": 13897 + }, + { + "epoch": 0.59, + "grad_norm": 0.46484375, + "learning_rate": 0.00016013869566167208, + "loss": 2.5997, + "step": 13898 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.00016013333617147485, + "loss": 2.5783, + "step": 13899 + }, + { + "epoch": 0.59, + "grad_norm": 0.494140625, + "learning_rate": 0.00016012797641069953, + "loss": 2.6891, + "step": 13900 + }, + { + "epoch": 0.59, + "grad_norm": 0.47265625, + "learning_rate": 0.0001601226163793703, + "loss": 2.6817, + "step": 13901 + }, + { + "epoch": 0.59, + "grad_norm": 0.5234375, + "learning_rate": 0.00016011725607751129, + "loss": 2.659, + "step": 13902 + }, + { + "epoch": 0.59, + "grad_norm": 0.490234375, + "learning_rate": 0.00016011189550514657, + "loss": 2.6803, + "step": 13903 + }, + { + "epoch": 0.59, + "grad_norm": 0.50390625, + "learning_rate": 0.0001601065346623003, + "loss": 2.6541, + "step": 13904 + }, + { + "epoch": 0.59, + "grad_norm": 0.43359375, + "learning_rate": 0.0001601011735489966, + "loss": 2.6115, + "step": 13905 + }, + { + "epoch": 0.59, + "grad_norm": 0.443359375, + "learning_rate": 0.0001600958121652596, + "loss": 2.6639, + "step": 13906 + }, + { + "epoch": 0.59, + "grad_norm": 0.4140625, + "learning_rate": 0.00016009045051111338, + "loss": 2.6896, + "step": 13907 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.0001600850885865821, + "loss": 2.5729, + "step": 13908 + }, + { + "epoch": 0.59, + "grad_norm": 0.453125, + "learning_rate": 0.00016007972639168986, + "loss": 2.558, + "step": 13909 + }, + { + "epoch": 0.59, + "grad_norm": 0.44140625, + "learning_rate": 0.00016007436392646084, + "loss": 2.6753, + "step": 13910 + }, + { + "epoch": 0.59, + "grad_norm": 0.419921875, + "learning_rate": 0.00016006900119091912, + "loss": 2.5995, + "step": 13911 + }, + { + "epoch": 0.59, + "grad_norm": 0.46875, + "learning_rate": 0.00016006363818508887, + "loss": 2.5641, + "step": 13912 + }, + { + "epoch": 0.59, + "grad_norm": 0.447265625, + "learning_rate": 0.0001600582749089942, + "loss": 2.5225, + "step": 13913 + }, + { + "epoch": 0.59, + "grad_norm": 0.423828125, + "learning_rate": 0.0001600529113626592, + "loss": 2.5889, + "step": 13914 + }, + { + "epoch": 0.59, + "grad_norm": 0.466796875, + "learning_rate": 0.00016004754754610806, + "loss": 2.6216, + "step": 13915 + }, + { + "epoch": 0.59, + "grad_norm": 0.458984375, + "learning_rate": 0.00016004218345936494, + "loss": 2.6094, + "step": 13916 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.0001600368191024539, + "loss": 2.5683, + "step": 13917 + }, + { + "epoch": 0.59, + "grad_norm": 0.466796875, + "learning_rate": 0.00016003145447539914, + "loss": 2.5805, + "step": 13918 + }, + { + "epoch": 0.59, + "grad_norm": 0.486328125, + "learning_rate": 0.0001600260895782247, + "loss": 2.6288, + "step": 13919 + }, + { + "epoch": 0.59, + "grad_norm": 0.44921875, + "learning_rate": 0.0001600207244109549, + "loss": 2.6193, + "step": 13920 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.0001600153589736137, + "loss": 2.6899, + "step": 13921 + }, + { + "epoch": 0.59, + "grad_norm": 0.462890625, + "learning_rate": 0.00016000999326622534, + "loss": 2.6132, + "step": 13922 + }, + { + "epoch": 0.59, + "grad_norm": 0.4453125, + "learning_rate": 0.00016000462728881396, + "loss": 2.4913, + "step": 13923 + }, + { + "epoch": 0.59, + "grad_norm": 0.470703125, + "learning_rate": 0.00015999926104140366, + "loss": 2.7057, + "step": 13924 + }, + { + "epoch": 0.59, + "grad_norm": 0.4609375, + "learning_rate": 0.00015999389452401862, + "loss": 2.5988, + "step": 13925 + }, + { + "epoch": 0.59, + "grad_norm": 0.42578125, + "learning_rate": 0.000159988527736683, + "loss": 2.4504, + "step": 13926 + }, + { + "epoch": 0.59, + "grad_norm": 0.451171875, + "learning_rate": 0.0001599831606794209, + "loss": 2.5412, + "step": 13927 + }, + { + "epoch": 0.59, + "grad_norm": 0.41015625, + "learning_rate": 0.0001599777933522565, + "loss": 2.5639, + "step": 13928 + }, + { + "epoch": 0.59, + "grad_norm": 0.494140625, + "learning_rate": 0.000159972425755214, + "loss": 2.682, + "step": 13929 + }, + { + "epoch": 0.59, + "grad_norm": 0.447265625, + "learning_rate": 0.00015996705788831744, + "loss": 2.5025, + "step": 13930 + }, + { + "epoch": 0.59, + "grad_norm": 0.48828125, + "learning_rate": 0.00015996168975159104, + "loss": 2.476, + "step": 13931 + }, + { + "epoch": 0.59, + "grad_norm": 0.5078125, + "learning_rate": 0.00015995632134505899, + "loss": 2.6744, + "step": 13932 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.0001599509526687454, + "loss": 2.5407, + "step": 13933 + }, + { + "epoch": 0.59, + "grad_norm": 0.482421875, + "learning_rate": 0.00015994558372267438, + "loss": 2.6121, + "step": 13934 + }, + { + "epoch": 0.59, + "grad_norm": 0.515625, + "learning_rate": 0.0001599402145068702, + "loss": 2.7192, + "step": 13935 + }, + { + "epoch": 0.59, + "grad_norm": 0.484375, + "learning_rate": 0.00015993484502135693, + "loss": 2.6575, + "step": 13936 + }, + { + "epoch": 0.59, + "grad_norm": 0.419921875, + "learning_rate": 0.0001599294752661588, + "loss": 2.5837, + "step": 13937 + }, + { + "epoch": 0.59, + "grad_norm": 0.494140625, + "learning_rate": 0.00015992410524129987, + "loss": 2.5778, + "step": 13938 + }, + { + "epoch": 0.59, + "grad_norm": 0.48046875, + "learning_rate": 0.0001599187349468044, + "loss": 2.7065, + "step": 13939 + }, + { + "epoch": 0.59, + "grad_norm": 0.46484375, + "learning_rate": 0.00015991336438269649, + "loss": 2.6532, + "step": 13940 + }, + { + "epoch": 0.59, + "grad_norm": 0.51171875, + "learning_rate": 0.00015990799354900038, + "loss": 2.6044, + "step": 13941 + }, + { + "epoch": 0.59, + "grad_norm": 0.431640625, + "learning_rate": 0.00015990262244574017, + "loss": 2.6052, + "step": 13942 + }, + { + "epoch": 0.59, + "grad_norm": 0.515625, + "learning_rate": 0.00015989725107294004, + "loss": 2.686, + "step": 13943 + }, + { + "epoch": 0.59, + "grad_norm": 0.4296875, + "learning_rate": 0.0001598918794306242, + "loss": 2.6113, + "step": 13944 + }, + { + "epoch": 0.59, + "grad_norm": 0.482421875, + "learning_rate": 0.00015988650751881677, + "loss": 2.6782, + "step": 13945 + }, + { + "epoch": 0.59, + "grad_norm": 0.466796875, + "learning_rate": 0.00015988113533754196, + "loss": 2.6179, + "step": 13946 + }, + { + "epoch": 0.59, + "grad_norm": 0.44140625, + "learning_rate": 0.0001598757628868239, + "loss": 2.6146, + "step": 13947 + }, + { + "epoch": 0.59, + "grad_norm": 0.50390625, + "learning_rate": 0.0001598703901666868, + "loss": 2.6872, + "step": 13948 + }, + { + "epoch": 0.59, + "grad_norm": 0.515625, + "learning_rate": 0.00015986501717715483, + "loss": 2.6226, + "step": 13949 + }, + { + "epoch": 0.59, + "grad_norm": 0.486328125, + "learning_rate": 0.00015985964391825218, + "loss": 2.5556, + "step": 13950 + }, + { + "epoch": 0.59, + "grad_norm": 0.46875, + "learning_rate": 0.00015985427039000297, + "loss": 2.5872, + "step": 13951 + }, + { + "epoch": 0.59, + "grad_norm": 0.47265625, + "learning_rate": 0.00015984889659243143, + "loss": 2.6138, + "step": 13952 + }, + { + "epoch": 0.59, + "grad_norm": 0.482421875, + "learning_rate": 0.00015984352252556173, + "loss": 2.7265, + "step": 13953 + }, + { + "epoch": 0.59, + "grad_norm": 0.5078125, + "learning_rate": 0.00015983814818941805, + "loss": 2.6462, + "step": 13954 + }, + { + "epoch": 0.59, + "grad_norm": 0.51953125, + "learning_rate": 0.00015983277358402458, + "loss": 2.605, + "step": 13955 + }, + { + "epoch": 0.59, + "grad_norm": 0.455078125, + "learning_rate": 0.0001598273987094055, + "loss": 2.5416, + "step": 13956 + }, + { + "epoch": 0.59, + "grad_norm": 0.5, + "learning_rate": 0.00015982202356558495, + "loss": 2.5759, + "step": 13957 + }, + { + "epoch": 0.59, + "grad_norm": 0.451171875, + "learning_rate": 0.0001598166481525872, + "loss": 2.6669, + "step": 13958 + }, + { + "epoch": 0.59, + "grad_norm": 0.478515625, + "learning_rate": 0.00015981127247043637, + "loss": 2.5251, + "step": 13959 + }, + { + "epoch": 0.59, + "grad_norm": 0.50390625, + "learning_rate": 0.00015980589651915666, + "loss": 2.6084, + "step": 13960 + }, + { + "epoch": 0.59, + "grad_norm": 0.4375, + "learning_rate": 0.00015980052029877234, + "loss": 2.6675, + "step": 13961 + }, + { + "epoch": 0.59, + "grad_norm": 0.4453125, + "learning_rate": 0.00015979514380930744, + "loss": 2.6873, + "step": 13962 + }, + { + "epoch": 0.6, + "grad_norm": 0.515625, + "learning_rate": 0.00015978976705078633, + "loss": 2.5408, + "step": 13963 + }, + { + "epoch": 0.6, + "grad_norm": 0.49609375, + "learning_rate": 0.00015978439002323306, + "loss": 2.5991, + "step": 13964 + }, + { + "epoch": 0.6, + "grad_norm": 0.46875, + "learning_rate": 0.00015977901272667193, + "loss": 2.5801, + "step": 13965 + }, + { + "epoch": 0.6, + "grad_norm": 0.490234375, + "learning_rate": 0.00015977363516112703, + "loss": 2.5889, + "step": 13966 + }, + { + "epoch": 0.6, + "grad_norm": 0.515625, + "learning_rate": 0.00015976825732662265, + "loss": 2.7526, + "step": 13967 + }, + { + "epoch": 0.6, + "grad_norm": 0.439453125, + "learning_rate": 0.00015976287922318295, + "loss": 2.4597, + "step": 13968 + }, + { + "epoch": 0.6, + "grad_norm": 0.43359375, + "learning_rate": 0.00015975750085083217, + "loss": 2.5461, + "step": 13969 + }, + { + "epoch": 0.6, + "grad_norm": 0.474609375, + "learning_rate": 0.00015975212220959443, + "loss": 2.4787, + "step": 13970 + }, + { + "epoch": 0.6, + "grad_norm": 0.455078125, + "learning_rate": 0.000159746743299494, + "loss": 2.7829, + "step": 13971 + }, + { + "epoch": 0.6, + "grad_norm": 0.4921875, + "learning_rate": 0.00015974136412055505, + "loss": 2.7179, + "step": 13972 + }, + { + "epoch": 0.6, + "grad_norm": 0.484375, + "learning_rate": 0.0001597359846728018, + "loss": 2.5605, + "step": 13973 + }, + { + "epoch": 0.6, + "grad_norm": 0.51953125, + "learning_rate": 0.00015973060495625842, + "loss": 2.5594, + "step": 13974 + }, + { + "epoch": 0.6, + "grad_norm": 0.455078125, + "learning_rate": 0.0001597252249709492, + "loss": 2.6832, + "step": 13975 + }, + { + "epoch": 0.6, + "grad_norm": 0.455078125, + "learning_rate": 0.00015971984471689824, + "loss": 2.5156, + "step": 13976 + }, + { + "epoch": 0.6, + "grad_norm": 0.4765625, + "learning_rate": 0.00015971446419412987, + "loss": 2.555, + "step": 13977 + }, + { + "epoch": 0.6, + "grad_norm": 0.44140625, + "learning_rate": 0.00015970908340266817, + "loss": 2.6377, + "step": 13978 + }, + { + "epoch": 0.6, + "grad_norm": 0.40625, + "learning_rate": 0.00015970370234253745, + "loss": 2.5637, + "step": 13979 + }, + { + "epoch": 0.6, + "grad_norm": 0.470703125, + "learning_rate": 0.00015969832101376185, + "loss": 2.6287, + "step": 13980 + }, + { + "epoch": 0.6, + "grad_norm": 0.412109375, + "learning_rate": 0.00015969293941636568, + "loss": 2.5638, + "step": 13981 + }, + { + "epoch": 0.6, + "grad_norm": 0.4921875, + "learning_rate": 0.00015968755755037302, + "loss": 2.5021, + "step": 13982 + }, + { + "epoch": 0.6, + "grad_norm": 0.458984375, + "learning_rate": 0.00015968217541580826, + "loss": 2.573, + "step": 13983 + }, + { + "epoch": 0.6, + "grad_norm": 0.490234375, + "learning_rate": 0.00015967679301269545, + "loss": 2.5363, + "step": 13984 + }, + { + "epoch": 0.6, + "grad_norm": 0.48828125, + "learning_rate": 0.0001596714103410589, + "loss": 2.6895, + "step": 13985 + }, + { + "epoch": 0.6, + "grad_norm": 0.51171875, + "learning_rate": 0.00015966602740092282, + "loss": 2.7375, + "step": 13986 + }, + { + "epoch": 0.6, + "grad_norm": 0.5390625, + "learning_rate": 0.00015966064419231145, + "loss": 2.7426, + "step": 13987 + }, + { + "epoch": 0.6, + "grad_norm": 0.470703125, + "learning_rate": 0.00015965526071524895, + "loss": 2.4593, + "step": 13988 + }, + { + "epoch": 0.6, + "grad_norm": 0.388671875, + "learning_rate": 0.00015964987696975958, + "loss": 2.5879, + "step": 13989 + }, + { + "epoch": 0.6, + "grad_norm": 0.427734375, + "learning_rate": 0.00015964449295586755, + "loss": 2.6285, + "step": 13990 + }, + { + "epoch": 0.6, + "grad_norm": 0.435546875, + "learning_rate": 0.00015963910867359713, + "loss": 2.6345, + "step": 13991 + }, + { + "epoch": 0.6, + "grad_norm": 0.515625, + "learning_rate": 0.00015963372412297248, + "loss": 2.6025, + "step": 13992 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.00015962833930401792, + "loss": 2.5588, + "step": 13993 + }, + { + "epoch": 0.6, + "grad_norm": 0.5, + "learning_rate": 0.00015962295421675758, + "loss": 2.6154, + "step": 13994 + }, + { + "epoch": 0.6, + "grad_norm": 0.45703125, + "learning_rate": 0.00015961756886121576, + "loss": 2.7608, + "step": 13995 + }, + { + "epoch": 0.6, + "grad_norm": 0.46484375, + "learning_rate": 0.00015961218323741662, + "loss": 2.6606, + "step": 13996 + }, + { + "epoch": 0.6, + "grad_norm": 0.47265625, + "learning_rate": 0.00015960679734538448, + "loss": 2.5911, + "step": 13997 + }, + { + "epoch": 0.6, + "grad_norm": 0.4296875, + "learning_rate": 0.00015960141118514354, + "loss": 2.6568, + "step": 13998 + }, + { + "epoch": 0.6, + "grad_norm": 0.443359375, + "learning_rate": 0.00015959602475671805, + "loss": 2.6408, + "step": 13999 + }, + { + "epoch": 0.6, + "grad_norm": 0.423828125, + "learning_rate": 0.0001595906380601322, + "loss": 2.4831, + "step": 14000 + }, + { + "epoch": 0.6, + "grad_norm": 0.416015625, + "learning_rate": 0.00015958525109541024, + "loss": 2.7405, + "step": 14001 + }, + { + "epoch": 0.6, + "grad_norm": 0.41015625, + "learning_rate": 0.00015957986386257644, + "loss": 2.6951, + "step": 14002 + }, + { + "epoch": 0.6, + "grad_norm": 0.404296875, + "learning_rate": 0.00015957447636165504, + "loss": 2.688, + "step": 14003 + }, + { + "epoch": 0.6, + "grad_norm": 0.435546875, + "learning_rate": 0.00015956908859267026, + "loss": 2.5753, + "step": 14004 + }, + { + "epoch": 0.6, + "grad_norm": 0.49609375, + "learning_rate": 0.00015956370055564636, + "loss": 2.7487, + "step": 14005 + }, + { + "epoch": 0.6, + "grad_norm": 0.369140625, + "learning_rate": 0.00015955831225060755, + "loss": 2.4954, + "step": 14006 + }, + { + "epoch": 0.6, + "grad_norm": 0.458984375, + "learning_rate": 0.00015955292367757812, + "loss": 2.6052, + "step": 14007 + }, + { + "epoch": 0.6, + "grad_norm": 0.4609375, + "learning_rate": 0.00015954753483658228, + "loss": 2.565, + "step": 14008 + }, + { + "epoch": 0.6, + "grad_norm": 0.419921875, + "learning_rate": 0.00015954214572764435, + "loss": 2.5075, + "step": 14009 + }, + { + "epoch": 0.6, + "grad_norm": 0.451171875, + "learning_rate": 0.00015953675635078848, + "loss": 2.5879, + "step": 14010 + }, + { + "epoch": 0.6, + "grad_norm": 0.408203125, + "learning_rate": 0.00015953136670603896, + "loss": 2.499, + "step": 14011 + }, + { + "epoch": 0.6, + "grad_norm": 0.435546875, + "learning_rate": 0.00015952597679342006, + "loss": 2.5441, + "step": 14012 + }, + { + "epoch": 0.6, + "grad_norm": 0.4375, + "learning_rate": 0.000159520586612956, + "loss": 2.711, + "step": 14013 + }, + { + "epoch": 0.6, + "grad_norm": 0.421875, + "learning_rate": 0.0001595151961646711, + "loss": 2.5802, + "step": 14014 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.0001595098054485895, + "loss": 2.6317, + "step": 14015 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.00015950441446473557, + "loss": 2.5743, + "step": 14016 + }, + { + "epoch": 0.6, + "grad_norm": 0.427734375, + "learning_rate": 0.0001594990232131335, + "loss": 2.5652, + "step": 14017 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.00015949363169380759, + "loss": 2.5133, + "step": 14018 + }, + { + "epoch": 0.6, + "grad_norm": 0.421875, + "learning_rate": 0.00015948823990678208, + "loss": 2.6075, + "step": 14019 + }, + { + "epoch": 0.6, + "grad_norm": 0.44921875, + "learning_rate": 0.0001594828478520812, + "loss": 2.6522, + "step": 14020 + }, + { + "epoch": 0.6, + "grad_norm": 0.43359375, + "learning_rate": 0.00015947745552972927, + "loss": 2.5372, + "step": 14021 + }, + { + "epoch": 0.6, + "grad_norm": 0.419921875, + "learning_rate": 0.0001594720629397505, + "loss": 2.4551, + "step": 14022 + }, + { + "epoch": 0.6, + "grad_norm": 0.46484375, + "learning_rate": 0.00015946667008216923, + "loss": 2.74, + "step": 14023 + }, + { + "epoch": 0.6, + "grad_norm": 0.486328125, + "learning_rate": 0.00015946127695700964, + "loss": 2.4822, + "step": 14024 + }, + { + "epoch": 0.6, + "grad_norm": 0.443359375, + "learning_rate": 0.00015945588356429603, + "loss": 2.6237, + "step": 14025 + }, + { + "epoch": 0.6, + "grad_norm": 0.439453125, + "learning_rate": 0.00015945048990405268, + "loss": 2.5378, + "step": 14026 + }, + { + "epoch": 0.6, + "grad_norm": 0.349609375, + "learning_rate": 0.0001594450959763039, + "loss": 2.5635, + "step": 14027 + }, + { + "epoch": 0.6, + "grad_norm": 0.474609375, + "learning_rate": 0.00015943970178107386, + "loss": 2.6656, + "step": 14028 + }, + { + "epoch": 0.6, + "grad_norm": 0.494140625, + "learning_rate": 0.00015943430731838688, + "loss": 2.4329, + "step": 14029 + }, + { + "epoch": 0.6, + "grad_norm": 0.466796875, + "learning_rate": 0.00015942891258826725, + "loss": 2.7523, + "step": 14030 + }, + { + "epoch": 0.6, + "grad_norm": 0.451171875, + "learning_rate": 0.00015942351759073927, + "loss": 2.6625, + "step": 14031 + }, + { + "epoch": 0.6, + "grad_norm": 0.427734375, + "learning_rate": 0.0001594181223258271, + "loss": 2.5075, + "step": 14032 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.00015941272679355513, + "loss": 2.6958, + "step": 14033 + }, + { + "epoch": 0.6, + "grad_norm": 0.486328125, + "learning_rate": 0.0001594073309939476, + "loss": 2.5701, + "step": 14034 + }, + { + "epoch": 0.6, + "grad_norm": 0.58203125, + "learning_rate": 0.00015940193492702882, + "loss": 2.6808, + "step": 14035 + }, + { + "epoch": 0.6, + "grad_norm": 0.435546875, + "learning_rate": 0.00015939653859282303, + "loss": 2.7102, + "step": 14036 + }, + { + "epoch": 0.6, + "grad_norm": 0.46875, + "learning_rate": 0.0001593911419913545, + "loss": 2.6567, + "step": 14037 + }, + { + "epoch": 0.6, + "grad_norm": 0.455078125, + "learning_rate": 0.00015938574512264756, + "loss": 2.5801, + "step": 14038 + }, + { + "epoch": 0.6, + "grad_norm": 0.5390625, + "learning_rate": 0.00015938034798672647, + "loss": 2.5176, + "step": 14039 + }, + { + "epoch": 0.6, + "grad_norm": 0.466796875, + "learning_rate": 0.0001593749505836155, + "loss": 2.4531, + "step": 14040 + }, + { + "epoch": 0.6, + "grad_norm": 0.462890625, + "learning_rate": 0.00015936955291333895, + "loss": 2.6865, + "step": 14041 + }, + { + "epoch": 0.6, + "grad_norm": 0.5078125, + "learning_rate": 0.00015936415497592112, + "loss": 2.6303, + "step": 14042 + }, + { + "epoch": 0.6, + "grad_norm": 0.447265625, + "learning_rate": 0.0001593587567713863, + "loss": 2.5959, + "step": 14043 + }, + { + "epoch": 0.6, + "grad_norm": 0.486328125, + "learning_rate": 0.00015935335829975875, + "loss": 2.5428, + "step": 14044 + }, + { + "epoch": 0.6, + "grad_norm": 0.421875, + "learning_rate": 0.0001593479595610628, + "loss": 2.7893, + "step": 14045 + }, + { + "epoch": 0.6, + "grad_norm": 0.453125, + "learning_rate": 0.00015934256055532265, + "loss": 2.5623, + "step": 14046 + }, + { + "epoch": 0.6, + "grad_norm": 0.451171875, + "learning_rate": 0.00015933716128256275, + "loss": 2.6853, + "step": 14047 + }, + { + "epoch": 0.6, + "grad_norm": 0.4375, + "learning_rate": 0.00015933176174280726, + "loss": 2.6589, + "step": 14048 + }, + { + "epoch": 0.6, + "grad_norm": 0.4296875, + "learning_rate": 0.00015932636193608058, + "loss": 2.4319, + "step": 14049 + }, + { + "epoch": 0.6, + "grad_norm": 0.5078125, + "learning_rate": 0.00015932096186240687, + "loss": 2.6445, + "step": 14050 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.00015931556152181056, + "loss": 2.5496, + "step": 14051 + }, + { + "epoch": 0.6, + "grad_norm": 0.5078125, + "learning_rate": 0.0001593101609143159, + "loss": 2.4719, + "step": 14052 + }, + { + "epoch": 0.6, + "grad_norm": 0.40625, + "learning_rate": 0.0001593047600399472, + "loss": 2.4938, + "step": 14053 + }, + { + "epoch": 0.6, + "grad_norm": 0.482421875, + "learning_rate": 0.00015929935889872875, + "loss": 2.5402, + "step": 14054 + }, + { + "epoch": 0.6, + "grad_norm": 0.419921875, + "learning_rate": 0.00015929395749068482, + "loss": 2.5733, + "step": 14055 + }, + { + "epoch": 0.6, + "grad_norm": 0.447265625, + "learning_rate": 0.0001592885558158398, + "loss": 2.5551, + "step": 14056 + }, + { + "epoch": 0.6, + "grad_norm": 0.474609375, + "learning_rate": 0.00015928315387421792, + "loss": 2.6706, + "step": 14057 + }, + { + "epoch": 0.6, + "grad_norm": 0.470703125, + "learning_rate": 0.00015927775166584354, + "loss": 2.7366, + "step": 14058 + }, + { + "epoch": 0.6, + "grad_norm": 0.4140625, + "learning_rate": 0.0001592723491907409, + "loss": 2.5953, + "step": 14059 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.0001592669464489344, + "loss": 2.5947, + "step": 14060 + }, + { + "epoch": 0.6, + "grad_norm": 0.4140625, + "learning_rate": 0.00015926154344044825, + "loss": 2.6585, + "step": 14061 + }, + { + "epoch": 0.6, + "grad_norm": 0.48828125, + "learning_rate": 0.00015925614016530683, + "loss": 2.5847, + "step": 14062 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.00015925073662353444, + "loss": 2.4984, + "step": 14063 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.0001592453328151554, + "loss": 2.6156, + "step": 14064 + }, + { + "epoch": 0.6, + "grad_norm": 0.466796875, + "learning_rate": 0.00015923992874019395, + "loss": 2.5777, + "step": 14065 + }, + { + "epoch": 0.6, + "grad_norm": 0.49609375, + "learning_rate": 0.00015923452439867454, + "loss": 2.752, + "step": 14066 + }, + { + "epoch": 0.6, + "grad_norm": 0.44921875, + "learning_rate": 0.0001592291197906214, + "loss": 2.6627, + "step": 14067 + }, + { + "epoch": 0.6, + "grad_norm": 0.484375, + "learning_rate": 0.00015922371491605886, + "loss": 2.6002, + "step": 14068 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.00015921830977501123, + "loss": 2.4481, + "step": 14069 + }, + { + "epoch": 0.6, + "grad_norm": 0.44921875, + "learning_rate": 0.0001592129043675029, + "loss": 2.6363, + "step": 14070 + }, + { + "epoch": 0.6, + "grad_norm": 0.46875, + "learning_rate": 0.00015920749869355812, + "loss": 2.5352, + "step": 14071 + }, + { + "epoch": 0.6, + "grad_norm": 0.421875, + "learning_rate": 0.0001592020927532012, + "loss": 2.6228, + "step": 14072 + }, + { + "epoch": 0.6, + "grad_norm": 0.439453125, + "learning_rate": 0.00015919668654645652, + "loss": 2.3991, + "step": 14073 + }, + { + "epoch": 0.6, + "grad_norm": 0.5, + "learning_rate": 0.0001591912800733484, + "loss": 2.5976, + "step": 14074 + }, + { + "epoch": 0.6, + "grad_norm": 0.439453125, + "learning_rate": 0.0001591858733339011, + "loss": 2.4908, + "step": 14075 + }, + { + "epoch": 0.6, + "grad_norm": 0.4375, + "learning_rate": 0.00015918046632813906, + "loss": 2.4562, + "step": 14076 + }, + { + "epoch": 0.6, + "grad_norm": 0.44921875, + "learning_rate": 0.00015917505905608655, + "loss": 2.5595, + "step": 14077 + }, + { + "epoch": 0.6, + "grad_norm": 0.458984375, + "learning_rate": 0.00015916965151776787, + "loss": 2.507, + "step": 14078 + }, + { + "epoch": 0.6, + "grad_norm": 0.5078125, + "learning_rate": 0.0001591642437132074, + "loss": 2.7014, + "step": 14079 + }, + { + "epoch": 0.6, + "grad_norm": 0.427734375, + "learning_rate": 0.00015915883564242945, + "loss": 2.5172, + "step": 14080 + }, + { + "epoch": 0.6, + "grad_norm": 0.4765625, + "learning_rate": 0.00015915342730545834, + "loss": 2.6002, + "step": 14081 + }, + { + "epoch": 0.6, + "grad_norm": 0.51953125, + "learning_rate": 0.00015914801870231844, + "loss": 2.6248, + "step": 14082 + }, + { + "epoch": 0.6, + "grad_norm": 0.482421875, + "learning_rate": 0.00015914260983303406, + "loss": 2.6362, + "step": 14083 + }, + { + "epoch": 0.6, + "grad_norm": 0.482421875, + "learning_rate": 0.00015913720069762955, + "loss": 2.6682, + "step": 14084 + }, + { + "epoch": 0.6, + "grad_norm": 0.43359375, + "learning_rate": 0.00015913179129612925, + "loss": 2.5339, + "step": 14085 + }, + { + "epoch": 0.6, + "grad_norm": 0.43359375, + "learning_rate": 0.00015912638162855753, + "loss": 2.6063, + "step": 14086 + }, + { + "epoch": 0.6, + "grad_norm": 0.51171875, + "learning_rate": 0.00015912097169493866, + "loss": 2.5919, + "step": 14087 + }, + { + "epoch": 0.6, + "grad_norm": 0.447265625, + "learning_rate": 0.00015911556149529707, + "loss": 2.5319, + "step": 14088 + }, + { + "epoch": 0.6, + "grad_norm": 0.47265625, + "learning_rate": 0.000159110151029657, + "loss": 2.6427, + "step": 14089 + }, + { + "epoch": 0.6, + "grad_norm": 0.40625, + "learning_rate": 0.00015910474029804292, + "loss": 2.4153, + "step": 14090 + }, + { + "epoch": 0.6, + "grad_norm": 0.443359375, + "learning_rate": 0.00015909932930047906, + "loss": 2.6788, + "step": 14091 + }, + { + "epoch": 0.6, + "grad_norm": 0.408203125, + "learning_rate": 0.00015909391803698984, + "loss": 2.726, + "step": 14092 + }, + { + "epoch": 0.6, + "grad_norm": 0.39453125, + "learning_rate": 0.00015908850650759955, + "loss": 2.3539, + "step": 14093 + }, + { + "epoch": 0.6, + "grad_norm": 0.46875, + "learning_rate": 0.00015908309471233262, + "loss": 2.6351, + "step": 14094 + }, + { + "epoch": 0.6, + "grad_norm": 0.47265625, + "learning_rate": 0.00015907768265121334, + "loss": 2.7158, + "step": 14095 + }, + { + "epoch": 0.6, + "grad_norm": 0.47265625, + "learning_rate": 0.00015907227032426605, + "loss": 2.7362, + "step": 14096 + }, + { + "epoch": 0.6, + "grad_norm": 0.474609375, + "learning_rate": 0.00015906685773151518, + "loss": 2.5405, + "step": 14097 + }, + { + "epoch": 0.6, + "grad_norm": 0.498046875, + "learning_rate": 0.00015906144487298502, + "loss": 2.8124, + "step": 14098 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.00015905603174869994, + "loss": 2.4649, + "step": 14099 + }, + { + "epoch": 0.6, + "grad_norm": 0.451171875, + "learning_rate": 0.0001590506183586843, + "loss": 2.5666, + "step": 14100 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.00015904520470296242, + "loss": 2.7476, + "step": 14101 + }, + { + "epoch": 0.6, + "grad_norm": 0.423828125, + "learning_rate": 0.00015903979078155876, + "loss": 2.5551, + "step": 14102 + }, + { + "epoch": 0.6, + "grad_norm": 0.490234375, + "learning_rate": 0.0001590343765944976, + "loss": 2.489, + "step": 14103 + }, + { + "epoch": 0.6, + "grad_norm": 0.4921875, + "learning_rate": 0.0001590289621418033, + "loss": 2.5245, + "step": 14104 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.00015902354742350022, + "loss": 2.6531, + "step": 14105 + }, + { + "epoch": 0.6, + "grad_norm": 0.423828125, + "learning_rate": 0.0001590181324396128, + "loss": 2.5055, + "step": 14106 + }, + { + "epoch": 0.6, + "grad_norm": 0.435546875, + "learning_rate": 0.00015901271719016532, + "loss": 2.6515, + "step": 14107 + }, + { + "epoch": 0.6, + "grad_norm": 0.48828125, + "learning_rate": 0.0001590073016751822, + "loss": 2.6512, + "step": 14108 + }, + { + "epoch": 0.6, + "grad_norm": 0.4453125, + "learning_rate": 0.0001590018858946878, + "loss": 2.5017, + "step": 14109 + }, + { + "epoch": 0.6, + "grad_norm": 0.423828125, + "learning_rate": 0.00015899646984870644, + "loss": 2.5079, + "step": 14110 + }, + { + "epoch": 0.6, + "grad_norm": 0.453125, + "learning_rate": 0.00015899105353726252, + "loss": 2.6091, + "step": 14111 + }, + { + "epoch": 0.6, + "grad_norm": 0.458984375, + "learning_rate": 0.00015898563696038046, + "loss": 2.5054, + "step": 14112 + }, + { + "epoch": 0.6, + "grad_norm": 0.515625, + "learning_rate": 0.00015898022011808457, + "loss": 2.653, + "step": 14113 + }, + { + "epoch": 0.6, + "grad_norm": 0.49609375, + "learning_rate": 0.00015897480301039924, + "loss": 2.7542, + "step": 14114 + }, + { + "epoch": 0.6, + "grad_norm": 0.419921875, + "learning_rate": 0.00015896938563734884, + "loss": 2.3862, + "step": 14115 + }, + { + "epoch": 0.6, + "grad_norm": 0.490234375, + "learning_rate": 0.00015896396799895777, + "loss": 2.6199, + "step": 14116 + }, + { + "epoch": 0.6, + "grad_norm": 0.466796875, + "learning_rate": 0.0001589585500952504, + "loss": 2.646, + "step": 14117 + }, + { + "epoch": 0.6, + "grad_norm": 0.4609375, + "learning_rate": 0.0001589531319262511, + "loss": 2.5675, + "step": 14118 + }, + { + "epoch": 0.6, + "grad_norm": 0.462890625, + "learning_rate": 0.00015894771349198424, + "loss": 2.614, + "step": 14119 + }, + { + "epoch": 0.6, + "grad_norm": 0.484375, + "learning_rate": 0.00015894229479247423, + "loss": 2.5286, + "step": 14120 + }, + { + "epoch": 0.6, + "grad_norm": 0.478515625, + "learning_rate": 0.00015893687582774538, + "loss": 2.6544, + "step": 14121 + }, + { + "epoch": 0.6, + "grad_norm": 0.453125, + "learning_rate": 0.0001589314565978222, + "loss": 2.5366, + "step": 14122 + }, + { + "epoch": 0.6, + "grad_norm": 0.5, + "learning_rate": 0.00015892603710272895, + "loss": 2.6518, + "step": 14123 + }, + { + "epoch": 0.6, + "grad_norm": 0.439453125, + "learning_rate": 0.00015892061734249012, + "loss": 2.5836, + "step": 14124 + }, + { + "epoch": 0.6, + "grad_norm": 0.423828125, + "learning_rate": 0.00015891519731712998, + "loss": 2.546, + "step": 14125 + }, + { + "epoch": 0.6, + "grad_norm": 0.44140625, + "learning_rate": 0.00015890977702667303, + "loss": 2.496, + "step": 14126 + }, + { + "epoch": 0.6, + "grad_norm": 0.48828125, + "learning_rate": 0.00015890435647114361, + "loss": 2.6204, + "step": 14127 + }, + { + "epoch": 0.6, + "grad_norm": 0.455078125, + "learning_rate": 0.0001588989356505661, + "loss": 2.596, + "step": 14128 + }, + { + "epoch": 0.6, + "grad_norm": 0.486328125, + "learning_rate": 0.0001588935145649649, + "loss": 2.6218, + "step": 14129 + }, + { + "epoch": 0.6, + "grad_norm": 0.431640625, + "learning_rate": 0.0001588880932143644, + "loss": 2.7241, + "step": 14130 + }, + { + "epoch": 0.6, + "grad_norm": 0.43359375, + "learning_rate": 0.00015888267159878903, + "loss": 2.7736, + "step": 14131 + }, + { + "epoch": 0.6, + "grad_norm": 0.46484375, + "learning_rate": 0.00015887724971826313, + "loss": 2.5954, + "step": 14132 + }, + { + "epoch": 0.6, + "grad_norm": 0.52734375, + "learning_rate": 0.00015887182757281113, + "loss": 2.7343, + "step": 14133 + }, + { + "epoch": 0.6, + "grad_norm": 0.45703125, + "learning_rate": 0.00015886640516245745, + "loss": 2.6083, + "step": 14134 + }, + { + "epoch": 0.6, + "grad_norm": 0.44921875, + "learning_rate": 0.00015886098248722643, + "loss": 2.5317, + "step": 14135 + }, + { + "epoch": 0.6, + "grad_norm": 0.49609375, + "learning_rate": 0.00015885555954714252, + "loss": 2.6258, + "step": 14136 + }, + { + "epoch": 0.6, + "grad_norm": 0.41015625, + "learning_rate": 0.00015885013634223008, + "loss": 2.4829, + "step": 14137 + }, + { + "epoch": 0.6, + "grad_norm": 0.42578125, + "learning_rate": 0.00015884471287251354, + "loss": 2.5693, + "step": 14138 + }, + { + "epoch": 0.6, + "grad_norm": 0.44921875, + "learning_rate": 0.00015883928913801732, + "loss": 2.5671, + "step": 14139 + }, + { + "epoch": 0.6, + "grad_norm": 0.4765625, + "learning_rate": 0.00015883386513876574, + "loss": 2.5457, + "step": 14140 + }, + { + "epoch": 0.6, + "grad_norm": 0.51953125, + "learning_rate": 0.0001588284408747833, + "loss": 2.7235, + "step": 14141 + }, + { + "epoch": 0.6, + "grad_norm": 0.45703125, + "learning_rate": 0.00015882301634609442, + "loss": 2.6587, + "step": 14142 + }, + { + "epoch": 0.6, + "grad_norm": 0.4765625, + "learning_rate": 0.00015881759155272343, + "loss": 2.5758, + "step": 14143 + }, + { + "epoch": 0.6, + "grad_norm": 0.427734375, + "learning_rate": 0.00015881216649469478, + "loss": 2.4713, + "step": 14144 + }, + { + "epoch": 0.6, + "grad_norm": 0.462890625, + "learning_rate": 0.00015880674117203284, + "loss": 2.6368, + "step": 14145 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.0001588013155847621, + "loss": 2.5922, + "step": 14146 + }, + { + "epoch": 0.6, + "grad_norm": 0.4609375, + "learning_rate": 0.0001587958897329069, + "loss": 2.7588, + "step": 14147 + }, + { + "epoch": 0.6, + "grad_norm": 0.50390625, + "learning_rate": 0.0001587904636164917, + "loss": 2.5906, + "step": 14148 + }, + { + "epoch": 0.6, + "grad_norm": 0.498046875, + "learning_rate": 0.00015878503723554087, + "loss": 2.6836, + "step": 14149 + }, + { + "epoch": 0.6, + "grad_norm": 0.462890625, + "learning_rate": 0.0001587796105900789, + "loss": 2.5984, + "step": 14150 + }, + { + "epoch": 0.6, + "grad_norm": 0.458984375, + "learning_rate": 0.00015877418368013014, + "loss": 2.6312, + "step": 14151 + }, + { + "epoch": 0.6, + "grad_norm": 0.455078125, + "learning_rate": 0.00015876875650571904, + "loss": 2.5513, + "step": 14152 + }, + { + "epoch": 0.6, + "grad_norm": 0.478515625, + "learning_rate": 0.00015876332906687002, + "loss": 2.6183, + "step": 14153 + }, + { + "epoch": 0.6, + "grad_norm": 0.40625, + "learning_rate": 0.0001587579013636075, + "loss": 2.5831, + "step": 14154 + }, + { + "epoch": 0.6, + "grad_norm": 0.47265625, + "learning_rate": 0.00015875247339595586, + "loss": 2.6389, + "step": 14155 + }, + { + "epoch": 0.6, + "grad_norm": 0.431640625, + "learning_rate": 0.00015874704516393958, + "loss": 2.6, + "step": 14156 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.00015874161666758308, + "loss": 2.7459, + "step": 14157 + }, + { + "epoch": 0.6, + "grad_norm": 0.46875, + "learning_rate": 0.0001587361879069108, + "loss": 2.608, + "step": 14158 + }, + { + "epoch": 0.6, + "grad_norm": 0.42578125, + "learning_rate": 0.00015873075888194705, + "loss": 2.4069, + "step": 14159 + }, + { + "epoch": 0.6, + "grad_norm": 0.45703125, + "learning_rate": 0.00015872532959271642, + "loss": 2.5845, + "step": 14160 + }, + { + "epoch": 0.6, + "grad_norm": 0.447265625, + "learning_rate": 0.00015871990003924323, + "loss": 2.507, + "step": 14161 + }, + { + "epoch": 0.6, + "grad_norm": 0.490234375, + "learning_rate": 0.000158714470221552, + "loss": 2.5509, + "step": 14162 + }, + { + "epoch": 0.6, + "grad_norm": 0.44140625, + "learning_rate": 0.0001587090401396671, + "loss": 2.6104, + "step": 14163 + }, + { + "epoch": 0.6, + "grad_norm": 0.466796875, + "learning_rate": 0.00015870360979361293, + "loss": 2.5082, + "step": 14164 + }, + { + "epoch": 0.6, + "grad_norm": 0.494140625, + "learning_rate": 0.000158698179183414, + "loss": 2.5508, + "step": 14165 + }, + { + "epoch": 0.6, + "grad_norm": 0.53515625, + "learning_rate": 0.00015869274830909472, + "loss": 2.6842, + "step": 14166 + }, + { + "epoch": 0.6, + "grad_norm": 0.474609375, + "learning_rate": 0.0001586873171706795, + "loss": 2.5485, + "step": 14167 + }, + { + "epoch": 0.6, + "grad_norm": 0.39453125, + "learning_rate": 0.00015868188576819282, + "loss": 2.4225, + "step": 14168 + }, + { + "epoch": 0.6, + "grad_norm": 0.439453125, + "learning_rate": 0.0001586764541016591, + "loss": 2.6266, + "step": 14169 + }, + { + "epoch": 0.6, + "grad_norm": 0.478515625, + "learning_rate": 0.0001586710221711028, + "loss": 2.6505, + "step": 14170 + }, + { + "epoch": 0.6, + "grad_norm": 0.4296875, + "learning_rate": 0.0001586655899765483, + "loss": 2.5531, + "step": 14171 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.00015866015751802011, + "loss": 2.6628, + "step": 14172 + }, + { + "epoch": 0.6, + "grad_norm": 0.515625, + "learning_rate": 0.0001586547247955426, + "loss": 2.6915, + "step": 14173 + }, + { + "epoch": 0.6, + "grad_norm": 0.412109375, + "learning_rate": 0.00015864929180914032, + "loss": 2.5374, + "step": 14174 + }, + { + "epoch": 0.6, + "grad_norm": 0.48828125, + "learning_rate": 0.00015864385855883766, + "loss": 2.5813, + "step": 14175 + }, + { + "epoch": 0.6, + "grad_norm": 0.4765625, + "learning_rate": 0.00015863842504465904, + "loss": 2.707, + "step": 14176 + }, + { + "epoch": 0.6, + "grad_norm": 0.4609375, + "learning_rate": 0.000158632991266629, + "loss": 2.6737, + "step": 14177 + }, + { + "epoch": 0.6, + "grad_norm": 0.451171875, + "learning_rate": 0.00015862755722477186, + "loss": 2.5255, + "step": 14178 + }, + { + "epoch": 0.6, + "grad_norm": 0.48046875, + "learning_rate": 0.00015862212291911214, + "loss": 2.5036, + "step": 14179 + }, + { + "epoch": 0.6, + "grad_norm": 0.453125, + "learning_rate": 0.0001586166883496743, + "loss": 2.6235, + "step": 14180 + }, + { + "epoch": 0.6, + "grad_norm": 0.50390625, + "learning_rate": 0.0001586112535164828, + "loss": 2.5981, + "step": 14181 + }, + { + "epoch": 0.6, + "grad_norm": 0.4140625, + "learning_rate": 0.00015860581841956208, + "loss": 2.4385, + "step": 14182 + }, + { + "epoch": 0.6, + "grad_norm": 0.4375, + "learning_rate": 0.00015860038305893657, + "loss": 2.5455, + "step": 14183 + }, + { + "epoch": 0.6, + "grad_norm": 0.421875, + "learning_rate": 0.0001585949474346308, + "loss": 2.5795, + "step": 14184 + }, + { + "epoch": 0.6, + "grad_norm": 0.431640625, + "learning_rate": 0.00015858951154666911, + "loss": 2.5296, + "step": 14185 + }, + { + "epoch": 0.6, + "grad_norm": 0.490234375, + "learning_rate": 0.00015858407539507609, + "loss": 2.6004, + "step": 14186 + }, + { + "epoch": 0.6, + "grad_norm": 0.427734375, + "learning_rate": 0.00015857863897987613, + "loss": 2.4318, + "step": 14187 + }, + { + "epoch": 0.6, + "grad_norm": 0.42578125, + "learning_rate": 0.0001585732023010937, + "loss": 2.5284, + "step": 14188 + }, + { + "epoch": 0.6, + "grad_norm": 0.43359375, + "learning_rate": 0.00015856776535875326, + "loss": 2.4432, + "step": 14189 + }, + { + "epoch": 0.6, + "grad_norm": 0.4609375, + "learning_rate": 0.00015856232815287925, + "loss": 2.6828, + "step": 14190 + }, + { + "epoch": 0.6, + "grad_norm": 0.51171875, + "learning_rate": 0.00015855689068349621, + "loss": 2.481, + "step": 14191 + }, + { + "epoch": 0.6, + "grad_norm": 0.421875, + "learning_rate": 0.00015855145295062855, + "loss": 2.524, + "step": 14192 + }, + { + "epoch": 0.6, + "grad_norm": 0.4921875, + "learning_rate": 0.00015854601495430076, + "loss": 2.6496, + "step": 14193 + }, + { + "epoch": 0.6, + "grad_norm": 0.458984375, + "learning_rate": 0.00015854057669453727, + "loss": 2.5476, + "step": 14194 + }, + { + "epoch": 0.6, + "grad_norm": 0.4140625, + "learning_rate": 0.00015853513817136257, + "loss": 2.5825, + "step": 14195 + }, + { + "epoch": 0.6, + "grad_norm": 0.490234375, + "learning_rate": 0.00015852969938480118, + "loss": 2.5248, + "step": 14196 + }, + { + "epoch": 0.6, + "grad_norm": 0.49609375, + "learning_rate": 0.00015852426033487752, + "loss": 2.6467, + "step": 14197 + }, + { + "epoch": 0.61, + "grad_norm": 0.4296875, + "learning_rate": 0.00015851882102161605, + "loss": 2.5652, + "step": 14198 + }, + { + "epoch": 0.61, + "grad_norm": 0.458984375, + "learning_rate": 0.00015851338144504134, + "loss": 2.4905, + "step": 14199 + }, + { + "epoch": 0.61, + "grad_norm": 0.486328125, + "learning_rate": 0.00015850794160517773, + "loss": 2.6945, + "step": 14200 + }, + { + "epoch": 0.61, + "grad_norm": 0.478515625, + "learning_rate": 0.0001585025015020498, + "loss": 2.6419, + "step": 14201 + }, + { + "epoch": 0.61, + "grad_norm": 0.41796875, + "learning_rate": 0.00015849706113568198, + "loss": 2.5537, + "step": 14202 + }, + { + "epoch": 0.61, + "grad_norm": 0.390625, + "learning_rate": 0.00015849162050609876, + "loss": 2.4116, + "step": 14203 + }, + { + "epoch": 0.61, + "grad_norm": 0.51171875, + "learning_rate": 0.00015848617961332462, + "loss": 2.5303, + "step": 14204 + }, + { + "epoch": 0.61, + "grad_norm": 0.486328125, + "learning_rate": 0.00015848073845738404, + "loss": 2.4759, + "step": 14205 + }, + { + "epoch": 0.61, + "grad_norm": 0.490234375, + "learning_rate": 0.00015847529703830152, + "loss": 2.6096, + "step": 14206 + }, + { + "epoch": 0.61, + "grad_norm": 0.466796875, + "learning_rate": 0.00015846985535610157, + "loss": 2.6117, + "step": 14207 + }, + { + "epoch": 0.61, + "grad_norm": 0.423828125, + "learning_rate": 0.0001584644134108086, + "loss": 2.6129, + "step": 14208 + }, + { + "epoch": 0.61, + "grad_norm": 0.4296875, + "learning_rate": 0.00015845897120244712, + "loss": 2.7372, + "step": 14209 + }, + { + "epoch": 0.61, + "grad_norm": 0.419921875, + "learning_rate": 0.00015845352873104166, + "loss": 2.51, + "step": 14210 + }, + { + "epoch": 0.61, + "grad_norm": 0.494140625, + "learning_rate": 0.0001584480859966167, + "loss": 2.6481, + "step": 14211 + }, + { + "epoch": 0.61, + "grad_norm": 0.4375, + "learning_rate": 0.00015844264299919668, + "loss": 2.4899, + "step": 14212 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015843719973880616, + "loss": 2.5545, + "step": 14213 + }, + { + "epoch": 0.61, + "grad_norm": 0.4765625, + "learning_rate": 0.00015843175621546956, + "loss": 2.696, + "step": 14214 + }, + { + "epoch": 0.61, + "grad_norm": 0.4375, + "learning_rate": 0.0001584263124292114, + "loss": 2.5483, + "step": 14215 + }, + { + "epoch": 0.61, + "grad_norm": 0.50390625, + "learning_rate": 0.00015842086838005622, + "loss": 2.6262, + "step": 14216 + }, + { + "epoch": 0.61, + "grad_norm": 0.443359375, + "learning_rate": 0.0001584154240680285, + "loss": 2.6829, + "step": 14217 + }, + { + "epoch": 0.61, + "grad_norm": 0.45703125, + "learning_rate": 0.0001584099794931527, + "loss": 2.4394, + "step": 14218 + }, + { + "epoch": 0.61, + "grad_norm": 0.451171875, + "learning_rate": 0.00015840453465545333, + "loss": 2.557, + "step": 14219 + }, + { + "epoch": 0.61, + "grad_norm": 0.453125, + "learning_rate": 0.00015839908955495488, + "loss": 2.7423, + "step": 14220 + }, + { + "epoch": 0.61, + "grad_norm": 0.455078125, + "learning_rate": 0.00015839364419168193, + "loss": 2.6211, + "step": 14221 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015838819856565884, + "loss": 2.6175, + "step": 14222 + }, + { + "epoch": 0.61, + "grad_norm": 0.462890625, + "learning_rate": 0.00015838275267691026, + "loss": 2.6207, + "step": 14223 + }, + { + "epoch": 0.61, + "grad_norm": 0.484375, + "learning_rate": 0.00015837730652546058, + "loss": 2.7515, + "step": 14224 + }, + { + "epoch": 0.61, + "grad_norm": 0.46484375, + "learning_rate": 0.00015837186011133443, + "loss": 2.595, + "step": 14225 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015836641343455616, + "loss": 2.6347, + "step": 14226 + }, + { + "epoch": 0.61, + "grad_norm": 0.498046875, + "learning_rate": 0.00015836096649515043, + "loss": 2.7366, + "step": 14227 + }, + { + "epoch": 0.61, + "grad_norm": 0.4296875, + "learning_rate": 0.0001583555192931416, + "loss": 2.4557, + "step": 14228 + }, + { + "epoch": 0.61, + "grad_norm": 0.45703125, + "learning_rate": 0.00015835007182855433, + "loss": 2.7016, + "step": 14229 + }, + { + "epoch": 0.61, + "grad_norm": 0.447265625, + "learning_rate": 0.00015834462410141303, + "loss": 2.4858, + "step": 14230 + }, + { + "epoch": 0.61, + "grad_norm": 0.47265625, + "learning_rate": 0.00015833917611174226, + "loss": 2.5922, + "step": 14231 + }, + { + "epoch": 0.61, + "grad_norm": 0.4765625, + "learning_rate": 0.00015833372785956649, + "loss": 2.709, + "step": 14232 + }, + { + "epoch": 0.61, + "grad_norm": 0.484375, + "learning_rate": 0.0001583282793449103, + "loss": 2.5609, + "step": 14233 + }, + { + "epoch": 0.61, + "grad_norm": 0.453125, + "learning_rate": 0.00015832283056779812, + "loss": 2.5766, + "step": 14234 + }, + { + "epoch": 0.61, + "grad_norm": 0.435546875, + "learning_rate": 0.00015831738152825454, + "loss": 2.5056, + "step": 14235 + }, + { + "epoch": 0.61, + "grad_norm": 0.5, + "learning_rate": 0.00015831193222630408, + "loss": 2.6464, + "step": 14236 + }, + { + "epoch": 0.61, + "grad_norm": 0.498046875, + "learning_rate": 0.00015830648266197122, + "loss": 2.4566, + "step": 14237 + }, + { + "epoch": 0.61, + "grad_norm": 0.4375, + "learning_rate": 0.0001583010328352805, + "loss": 2.6193, + "step": 14238 + }, + { + "epoch": 0.61, + "grad_norm": 0.451171875, + "learning_rate": 0.00015829558274625644, + "loss": 2.5746, + "step": 14239 + }, + { + "epoch": 0.61, + "grad_norm": 0.4453125, + "learning_rate": 0.00015829013239492353, + "loss": 2.6118, + "step": 14240 + }, + { + "epoch": 0.61, + "grad_norm": 0.4765625, + "learning_rate": 0.00015828468178130637, + "loss": 2.6559, + "step": 14241 + }, + { + "epoch": 0.61, + "grad_norm": 0.498046875, + "learning_rate": 0.00015827923090542943, + "loss": 2.6042, + "step": 14242 + }, + { + "epoch": 0.61, + "grad_norm": 0.427734375, + "learning_rate": 0.00015827377976731724, + "loss": 2.4686, + "step": 14243 + }, + { + "epoch": 0.61, + "grad_norm": 0.40625, + "learning_rate": 0.00015826832836699436, + "loss": 2.6081, + "step": 14244 + }, + { + "epoch": 0.61, + "grad_norm": 0.44921875, + "learning_rate": 0.00015826287670448526, + "loss": 2.5817, + "step": 14245 + }, + { + "epoch": 0.61, + "grad_norm": 0.50390625, + "learning_rate": 0.00015825742477981454, + "loss": 2.7377, + "step": 14246 + }, + { + "epoch": 0.61, + "grad_norm": 0.451171875, + "learning_rate": 0.0001582519725930067, + "loss": 2.5231, + "step": 14247 + }, + { + "epoch": 0.61, + "grad_norm": 0.4375, + "learning_rate": 0.00015824652014408624, + "loss": 2.6298, + "step": 14248 + }, + { + "epoch": 0.61, + "grad_norm": 0.45703125, + "learning_rate": 0.00015824106743307775, + "loss": 2.5777, + "step": 14249 + }, + { + "epoch": 0.61, + "grad_norm": 0.46484375, + "learning_rate": 0.00015823561446000572, + "loss": 2.5872, + "step": 14250 + }, + { + "epoch": 0.61, + "grad_norm": 0.4375, + "learning_rate": 0.00015823016122489474, + "loss": 2.5563, + "step": 14251 + }, + { + "epoch": 0.61, + "grad_norm": 0.48828125, + "learning_rate": 0.0001582247077277693, + "loss": 2.4978, + "step": 14252 + }, + { + "epoch": 0.61, + "grad_norm": 0.423828125, + "learning_rate": 0.00015821925396865397, + "loss": 2.4369, + "step": 14253 + }, + { + "epoch": 0.61, + "grad_norm": 0.412109375, + "learning_rate": 0.00015821379994757323, + "loss": 2.5015, + "step": 14254 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015820834566455168, + "loss": 2.5687, + "step": 14255 + }, + { + "epoch": 0.61, + "grad_norm": 0.46875, + "learning_rate": 0.00015820289111961386, + "loss": 2.5321, + "step": 14256 + }, + { + "epoch": 0.61, + "grad_norm": 0.462890625, + "learning_rate": 0.00015819743631278433, + "loss": 2.5948, + "step": 14257 + }, + { + "epoch": 0.61, + "grad_norm": 0.43359375, + "learning_rate": 0.00015819198124408755, + "loss": 2.6017, + "step": 14258 + }, + { + "epoch": 0.61, + "grad_norm": 0.515625, + "learning_rate": 0.00015818652591354814, + "loss": 2.4466, + "step": 14259 + }, + { + "epoch": 0.61, + "grad_norm": 0.486328125, + "learning_rate": 0.00015818107032119062, + "loss": 2.6256, + "step": 14260 + }, + { + "epoch": 0.61, + "grad_norm": 0.48828125, + "learning_rate": 0.00015817561446703957, + "loss": 2.5225, + "step": 14261 + }, + { + "epoch": 0.61, + "grad_norm": 0.46875, + "learning_rate": 0.0001581701583511195, + "loss": 2.5816, + "step": 14262 + }, + { + "epoch": 0.61, + "grad_norm": 0.4453125, + "learning_rate": 0.00015816470197345498, + "loss": 2.4957, + "step": 14263 + }, + { + "epoch": 0.61, + "grad_norm": 0.44140625, + "learning_rate": 0.00015815924533407055, + "loss": 2.6181, + "step": 14264 + }, + { + "epoch": 0.61, + "grad_norm": 0.48828125, + "learning_rate": 0.0001581537884329908, + "loss": 2.5979, + "step": 14265 + }, + { + "epoch": 0.61, + "grad_norm": 0.46484375, + "learning_rate": 0.00015814833127024023, + "loss": 2.6956, + "step": 14266 + }, + { + "epoch": 0.61, + "grad_norm": 0.466796875, + "learning_rate": 0.00015814287384584343, + "loss": 2.551, + "step": 14267 + }, + { + "epoch": 0.61, + "grad_norm": 0.458984375, + "learning_rate": 0.00015813741615982492, + "loss": 2.4532, + "step": 14268 + }, + { + "epoch": 0.61, + "grad_norm": 0.515625, + "learning_rate": 0.00015813195821220933, + "loss": 2.5192, + "step": 14269 + }, + { + "epoch": 0.61, + "grad_norm": 0.439453125, + "learning_rate": 0.00015812650000302116, + "loss": 2.6026, + "step": 14270 + }, + { + "epoch": 0.61, + "grad_norm": 0.48046875, + "learning_rate": 0.00015812104153228496, + "loss": 2.556, + "step": 14271 + }, + { + "epoch": 0.61, + "grad_norm": 0.44140625, + "learning_rate": 0.00015811558280002532, + "loss": 2.4485, + "step": 14272 + }, + { + "epoch": 0.61, + "grad_norm": 0.41015625, + "learning_rate": 0.00015811012380626682, + "loss": 2.4436, + "step": 14273 + }, + { + "epoch": 0.61, + "grad_norm": 0.5, + "learning_rate": 0.00015810466455103397, + "loss": 2.6484, + "step": 14274 + }, + { + "epoch": 0.61, + "grad_norm": 0.46484375, + "learning_rate": 0.00015809920503435137, + "loss": 2.7036, + "step": 14275 + }, + { + "epoch": 0.61, + "grad_norm": 0.4609375, + "learning_rate": 0.00015809374525624356, + "loss": 2.5075, + "step": 14276 + }, + { + "epoch": 0.61, + "grad_norm": 0.443359375, + "learning_rate": 0.00015808828521673518, + "loss": 2.6827, + "step": 14277 + }, + { + "epoch": 0.61, + "grad_norm": 0.42578125, + "learning_rate": 0.00015808282491585072, + "loss": 2.4667, + "step": 14278 + }, + { + "epoch": 0.61, + "grad_norm": 0.404296875, + "learning_rate": 0.00015807736435361478, + "loss": 2.6041, + "step": 14279 + }, + { + "epoch": 0.61, + "grad_norm": 0.41015625, + "learning_rate": 0.0001580719035300519, + "loss": 2.5226, + "step": 14280 + }, + { + "epoch": 0.61, + "grad_norm": 0.419921875, + "learning_rate": 0.00015806644244518672, + "loss": 2.6968, + "step": 14281 + }, + { + "epoch": 0.61, + "grad_norm": 0.5078125, + "learning_rate": 0.00015806098109904374, + "loss": 2.5784, + "step": 14282 + }, + { + "epoch": 0.61, + "grad_norm": 0.4140625, + "learning_rate": 0.00015805551949164755, + "loss": 2.5272, + "step": 14283 + }, + { + "epoch": 0.61, + "grad_norm": 0.46875, + "learning_rate": 0.00015805005762302277, + "loss": 2.6653, + "step": 14284 + }, + { + "epoch": 0.61, + "grad_norm": 0.462890625, + "learning_rate": 0.00015804459549319396, + "loss": 2.6666, + "step": 14285 + }, + { + "epoch": 0.61, + "grad_norm": 0.43359375, + "learning_rate": 0.00015803913310218564, + "loss": 2.6147, + "step": 14286 + }, + { + "epoch": 0.61, + "grad_norm": 0.443359375, + "learning_rate": 0.00015803367045002245, + "loss": 2.5262, + "step": 14287 + }, + { + "epoch": 0.61, + "grad_norm": 0.470703125, + "learning_rate": 0.00015802820753672893, + "loss": 2.6586, + "step": 14288 + }, + { + "epoch": 0.61, + "grad_norm": 0.478515625, + "learning_rate": 0.0001580227443623297, + "loss": 2.5677, + "step": 14289 + }, + { + "epoch": 0.61, + "grad_norm": 0.48046875, + "learning_rate": 0.00015801728092684934, + "loss": 2.586, + "step": 14290 + }, + { + "epoch": 0.61, + "grad_norm": 0.466796875, + "learning_rate": 0.0001580118172303124, + "loss": 2.6022, + "step": 14291 + }, + { + "epoch": 0.61, + "grad_norm": 0.45703125, + "learning_rate": 0.0001580063532727435, + "loss": 2.6758, + "step": 14292 + }, + { + "epoch": 0.61, + "grad_norm": 0.369140625, + "learning_rate": 0.0001580008890541672, + "loss": 2.5191, + "step": 14293 + }, + { + "epoch": 0.61, + "grad_norm": 0.478515625, + "learning_rate": 0.00015799542457460808, + "loss": 2.6988, + "step": 14294 + }, + { + "epoch": 0.61, + "grad_norm": 0.423828125, + "learning_rate": 0.00015798995983409077, + "loss": 2.6678, + "step": 14295 + }, + { + "epoch": 0.61, + "grad_norm": 0.4609375, + "learning_rate": 0.00015798449483263983, + "loss": 2.5087, + "step": 14296 + }, + { + "epoch": 0.61, + "grad_norm": 0.498046875, + "learning_rate": 0.00015797902957027983, + "loss": 2.5876, + "step": 14297 + }, + { + "epoch": 0.61, + "grad_norm": 0.39453125, + "learning_rate": 0.00015797356404703542, + "loss": 2.5058, + "step": 14298 + }, + { + "epoch": 0.61, + "grad_norm": 0.3828125, + "learning_rate": 0.00015796809826293116, + "loss": 2.613, + "step": 14299 + }, + { + "epoch": 0.61, + "grad_norm": 0.490234375, + "learning_rate": 0.00015796263221799163, + "loss": 2.5281, + "step": 14300 + }, + { + "epoch": 0.61, + "grad_norm": 0.43359375, + "learning_rate": 0.0001579571659122414, + "loss": 2.6267, + "step": 14301 + }, + { + "epoch": 0.61, + "grad_norm": 0.45703125, + "learning_rate": 0.00015795169934570519, + "loss": 2.5932, + "step": 14302 + }, + { + "epoch": 0.61, + "grad_norm": 0.50390625, + "learning_rate": 0.00015794623251840742, + "loss": 2.6158, + "step": 14303 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015794076543037283, + "loss": 2.5367, + "step": 14304 + }, + { + "epoch": 0.61, + "grad_norm": 0.51953125, + "learning_rate": 0.00015793529808162595, + "loss": 2.6284, + "step": 14305 + }, + { + "epoch": 0.61, + "grad_norm": 0.482421875, + "learning_rate": 0.0001579298304721914, + "loss": 2.5909, + "step": 14306 + }, + { + "epoch": 0.61, + "grad_norm": 0.484375, + "learning_rate": 0.00015792436260209382, + "loss": 2.4882, + "step": 14307 + }, + { + "epoch": 0.61, + "grad_norm": 0.515625, + "learning_rate": 0.00015791889447135774, + "loss": 2.5621, + "step": 14308 + }, + { + "epoch": 0.61, + "grad_norm": 0.44140625, + "learning_rate": 0.00015791342608000784, + "loss": 2.5089, + "step": 14309 + }, + { + "epoch": 0.61, + "grad_norm": 0.3984375, + "learning_rate": 0.00015790795742806862, + "loss": 2.5501, + "step": 14310 + }, + { + "epoch": 0.61, + "grad_norm": 0.5078125, + "learning_rate": 0.00015790248851556483, + "loss": 2.7447, + "step": 14311 + }, + { + "epoch": 0.61, + "grad_norm": 0.498046875, + "learning_rate": 0.00015789701934252096, + "loss": 2.6376, + "step": 14312 + }, + { + "epoch": 0.61, + "grad_norm": 0.48828125, + "learning_rate": 0.00015789154990896167, + "loss": 2.5942, + "step": 14313 + }, + { + "epoch": 0.61, + "grad_norm": 0.4296875, + "learning_rate": 0.00015788608021491158, + "loss": 2.4067, + "step": 14314 + }, + { + "epoch": 0.61, + "grad_norm": 0.458984375, + "learning_rate": 0.00015788061026039528, + "loss": 2.5203, + "step": 14315 + }, + { + "epoch": 0.61, + "grad_norm": 0.46875, + "learning_rate": 0.00015787514004543738, + "loss": 2.5931, + "step": 14316 + }, + { + "epoch": 0.61, + "grad_norm": 0.43359375, + "learning_rate": 0.00015786966957006247, + "loss": 2.5613, + "step": 14317 + }, + { + "epoch": 0.61, + "grad_norm": 0.4609375, + "learning_rate": 0.0001578641988342952, + "loss": 2.5761, + "step": 14318 + }, + { + "epoch": 0.61, + "grad_norm": 0.4453125, + "learning_rate": 0.0001578587278381602, + "loss": 2.5479, + "step": 14319 + }, + { + "epoch": 0.61, + "grad_norm": 0.478515625, + "learning_rate": 0.0001578532565816821, + "loss": 2.5445, + "step": 14320 + }, + { + "epoch": 0.61, + "grad_norm": 0.478515625, + "learning_rate": 0.00015784778506488544, + "loss": 2.5643, + "step": 14321 + }, + { + "epoch": 0.61, + "grad_norm": 0.458984375, + "learning_rate": 0.00015784231328779493, + "loss": 2.4998, + "step": 14322 + }, + { + "epoch": 0.61, + "grad_norm": 0.453125, + "learning_rate": 0.00015783684125043513, + "loss": 2.487, + "step": 14323 + }, + { + "epoch": 0.61, + "grad_norm": 0.47265625, + "learning_rate": 0.00015783136895283068, + "loss": 2.5325, + "step": 14324 + }, + { + "epoch": 0.61, + "grad_norm": 0.5, + "learning_rate": 0.0001578258963950062, + "loss": 2.676, + "step": 14325 + }, + { + "epoch": 0.61, + "grad_norm": 0.4375, + "learning_rate": 0.00015782042357698631, + "loss": 2.5942, + "step": 14326 + }, + { + "epoch": 0.61, + "grad_norm": 0.41796875, + "learning_rate": 0.00015781495049879566, + "loss": 2.5191, + "step": 14327 + }, + { + "epoch": 0.61, + "grad_norm": 0.443359375, + "learning_rate": 0.0001578094771604589, + "loss": 2.6219, + "step": 14328 + }, + { + "epoch": 0.61, + "grad_norm": 0.443359375, + "learning_rate": 0.00015780400356200057, + "loss": 2.6569, + "step": 14329 + }, + { + "epoch": 0.61, + "grad_norm": 0.39453125, + "learning_rate": 0.00015779852970344534, + "loss": 2.4888, + "step": 14330 + }, + { + "epoch": 0.61, + "grad_norm": 0.498046875, + "learning_rate": 0.00015779305558481787, + "loss": 2.6384, + "step": 14331 + }, + { + "epoch": 0.61, + "grad_norm": 0.392578125, + "learning_rate": 0.00015778758120614276, + "loss": 2.5031, + "step": 14332 + }, + { + "epoch": 0.61, + "grad_norm": 0.486328125, + "learning_rate": 0.00015778210656744468, + "loss": 2.6284, + "step": 14333 + }, + { + "epoch": 0.61, + "grad_norm": 0.44921875, + "learning_rate": 0.00015777663166874822, + "loss": 2.5818, + "step": 14334 + }, + { + "epoch": 0.61, + "grad_norm": 0.431640625, + "learning_rate": 0.00015777115651007803, + "loss": 2.3992, + "step": 14335 + }, + { + "epoch": 0.61, + "grad_norm": 0.443359375, + "learning_rate": 0.00015776568109145876, + "loss": 2.5815, + "step": 14336 + }, + { + "epoch": 0.61, + "grad_norm": 0.46875, + "learning_rate": 0.00015776020541291504, + "loss": 2.5086, + "step": 14337 + }, + { + "epoch": 0.61, + "grad_norm": 0.48828125, + "learning_rate": 0.00015775472947447146, + "loss": 2.6329, + "step": 14338 + }, + { + "epoch": 0.61, + "grad_norm": 0.52734375, + "learning_rate": 0.00015774925327615273, + "loss": 2.5818, + "step": 14339 + }, + { + "epoch": 0.61, + "grad_norm": 0.490234375, + "learning_rate": 0.0001577437768179835, + "loss": 2.7442, + "step": 14340 + }, + { + "epoch": 0.61, + "grad_norm": 0.51953125, + "learning_rate": 0.00015773830009998833, + "loss": 2.6396, + "step": 14341 + }, + { + "epoch": 0.61, + "grad_norm": 0.431640625, + "learning_rate": 0.00015773282312219196, + "loss": 2.5615, + "step": 14342 + }, + { + "epoch": 0.61, + "grad_norm": 0.466796875, + "learning_rate": 0.00015772734588461894, + "loss": 2.5203, + "step": 14343 + }, + { + "epoch": 0.61, + "grad_norm": 0.451171875, + "learning_rate": 0.00015772186838729397, + "loss": 2.5692, + "step": 14344 + }, + { + "epoch": 0.61, + "grad_norm": 0.455078125, + "learning_rate": 0.0001577163906302417, + "loss": 2.6137, + "step": 14345 + }, + { + "epoch": 0.61, + "grad_norm": 0.51171875, + "learning_rate": 0.00015771091261348677, + "loss": 2.6852, + "step": 14346 + }, + { + "epoch": 0.61, + "grad_norm": 0.458984375, + "learning_rate": 0.00015770543433705377, + "loss": 2.5575, + "step": 14347 + }, + { + "epoch": 0.61, + "grad_norm": 0.455078125, + "learning_rate": 0.00015769995580096748, + "loss": 2.5944, + "step": 14348 + }, + { + "epoch": 0.61, + "grad_norm": 0.4140625, + "learning_rate": 0.0001576944770052524, + "loss": 2.5922, + "step": 14349 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015768899794993331, + "loss": 2.6387, + "step": 14350 + }, + { + "epoch": 0.61, + "grad_norm": 0.462890625, + "learning_rate": 0.0001576835186350348, + "loss": 2.5613, + "step": 14351 + }, + { + "epoch": 0.61, + "grad_norm": 0.4453125, + "learning_rate": 0.00015767803906058156, + "loss": 2.4593, + "step": 14352 + }, + { + "epoch": 0.61, + "grad_norm": 0.466796875, + "learning_rate": 0.0001576725592265982, + "loss": 2.5894, + "step": 14353 + }, + { + "epoch": 0.61, + "grad_norm": 0.52734375, + "learning_rate": 0.00015766707913310941, + "loss": 2.5268, + "step": 14354 + }, + { + "epoch": 0.61, + "grad_norm": 0.5078125, + "learning_rate": 0.00015766159878013983, + "loss": 2.6313, + "step": 14355 + }, + { + "epoch": 0.61, + "grad_norm": 0.431640625, + "learning_rate": 0.00015765611816771414, + "loss": 2.47, + "step": 14356 + }, + { + "epoch": 0.61, + "grad_norm": 0.423828125, + "learning_rate": 0.00015765063729585697, + "loss": 2.5793, + "step": 14357 + }, + { + "epoch": 0.61, + "grad_norm": 0.458984375, + "learning_rate": 0.000157645156164593, + "loss": 2.6161, + "step": 14358 + }, + { + "epoch": 0.61, + "grad_norm": 0.53125, + "learning_rate": 0.0001576396747739469, + "loss": 2.5693, + "step": 14359 + }, + { + "epoch": 0.61, + "grad_norm": 0.416015625, + "learning_rate": 0.00015763419312394335, + "loss": 2.5582, + "step": 14360 + }, + { + "epoch": 0.61, + "grad_norm": 0.419921875, + "learning_rate": 0.00015762871121460697, + "loss": 2.6953, + "step": 14361 + }, + { + "epoch": 0.61, + "grad_norm": 0.44140625, + "learning_rate": 0.00015762322904596245, + "loss": 2.5799, + "step": 14362 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015761774661803447, + "loss": 2.5946, + "step": 14363 + }, + { + "epoch": 0.61, + "grad_norm": 0.4453125, + "learning_rate": 0.00015761226393084767, + "loss": 2.5432, + "step": 14364 + }, + { + "epoch": 0.61, + "grad_norm": 0.453125, + "learning_rate": 0.00015760678098442673, + "loss": 2.4436, + "step": 14365 + }, + { + "epoch": 0.61, + "grad_norm": 0.478515625, + "learning_rate": 0.00015760129777879638, + "loss": 2.5697, + "step": 14366 + }, + { + "epoch": 0.61, + "grad_norm": 0.486328125, + "learning_rate": 0.0001575958143139812, + "loss": 2.7094, + "step": 14367 + }, + { + "epoch": 0.61, + "grad_norm": 0.490234375, + "learning_rate": 0.00015759033059000593, + "loss": 2.525, + "step": 14368 + }, + { + "epoch": 0.61, + "grad_norm": 0.416015625, + "learning_rate": 0.00015758484660689519, + "loss": 2.6728, + "step": 14369 + }, + { + "epoch": 0.61, + "grad_norm": 0.447265625, + "learning_rate": 0.0001575793623646737, + "loss": 2.5902, + "step": 14370 + }, + { + "epoch": 0.61, + "grad_norm": 0.427734375, + "learning_rate": 0.0001575738778633661, + "loss": 2.5704, + "step": 14371 + }, + { + "epoch": 0.61, + "grad_norm": 0.44140625, + "learning_rate": 0.00015756839310299711, + "loss": 2.5832, + "step": 14372 + }, + { + "epoch": 0.61, + "grad_norm": 0.4765625, + "learning_rate": 0.0001575629080835914, + "loss": 2.6536, + "step": 14373 + }, + { + "epoch": 0.61, + "grad_norm": 0.50390625, + "learning_rate": 0.00015755742280517364, + "loss": 2.6532, + "step": 14374 + }, + { + "epoch": 0.61, + "grad_norm": 0.46875, + "learning_rate": 0.0001575519372677685, + "loss": 2.5658, + "step": 14375 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015754645147140068, + "loss": 2.5364, + "step": 14376 + }, + { + "epoch": 0.61, + "grad_norm": 0.40625, + "learning_rate": 0.00015754096541609484, + "loss": 2.5819, + "step": 14377 + }, + { + "epoch": 0.61, + "grad_norm": 0.4921875, + "learning_rate": 0.00015753547910187573, + "loss": 2.6331, + "step": 14378 + }, + { + "epoch": 0.61, + "grad_norm": 0.46484375, + "learning_rate": 0.00015752999252876792, + "loss": 2.5128, + "step": 14379 + }, + { + "epoch": 0.61, + "grad_norm": 0.42578125, + "learning_rate": 0.00015752450569679624, + "loss": 2.5761, + "step": 14380 + }, + { + "epoch": 0.61, + "grad_norm": 0.48828125, + "learning_rate": 0.00015751901860598525, + "loss": 2.5676, + "step": 14381 + }, + { + "epoch": 0.61, + "grad_norm": 0.42578125, + "learning_rate": 0.00015751353125635975, + "loss": 2.518, + "step": 14382 + }, + { + "epoch": 0.61, + "grad_norm": 0.427734375, + "learning_rate": 0.00015750804364794436, + "loss": 2.4935, + "step": 14383 + }, + { + "epoch": 0.61, + "grad_norm": 0.447265625, + "learning_rate": 0.00015750255578076376, + "loss": 2.6034, + "step": 14384 + }, + { + "epoch": 0.61, + "grad_norm": 0.427734375, + "learning_rate": 0.0001574970676548427, + "loss": 2.5277, + "step": 14385 + }, + { + "epoch": 0.61, + "grad_norm": 0.470703125, + "learning_rate": 0.00015749157927020586, + "loss": 2.6233, + "step": 14386 + }, + { + "epoch": 0.61, + "grad_norm": 0.458984375, + "learning_rate": 0.00015748609062687789, + "loss": 2.6752, + "step": 14387 + }, + { + "epoch": 0.61, + "grad_norm": 0.49609375, + "learning_rate": 0.00015748060172488356, + "loss": 2.5246, + "step": 14388 + }, + { + "epoch": 0.61, + "grad_norm": 0.484375, + "learning_rate": 0.00015747511256424752, + "loss": 2.55, + "step": 14389 + }, + { + "epoch": 0.61, + "grad_norm": 0.40625, + "learning_rate": 0.00015746962314499446, + "loss": 2.5533, + "step": 14390 + }, + { + "epoch": 0.61, + "grad_norm": 0.490234375, + "learning_rate": 0.00015746413346714907, + "loss": 2.5775, + "step": 14391 + }, + { + "epoch": 0.61, + "grad_norm": 0.4765625, + "learning_rate": 0.00015745864353073613, + "loss": 2.6995, + "step": 14392 + }, + { + "epoch": 0.61, + "grad_norm": 0.45703125, + "learning_rate": 0.00015745315333578027, + "loss": 2.5987, + "step": 14393 + }, + { + "epoch": 0.61, + "grad_norm": 0.474609375, + "learning_rate": 0.00015744766288230624, + "loss": 2.6485, + "step": 14394 + }, + { + "epoch": 0.61, + "grad_norm": 0.458984375, + "learning_rate": 0.0001574421721703387, + "loss": 2.5857, + "step": 14395 + }, + { + "epoch": 0.61, + "grad_norm": 0.462890625, + "learning_rate": 0.0001574366811999024, + "loss": 2.5618, + "step": 14396 + }, + { + "epoch": 0.61, + "grad_norm": 0.515625, + "learning_rate": 0.000157431189971022, + "loss": 2.6493, + "step": 14397 + }, + { + "epoch": 0.61, + "grad_norm": 0.53515625, + "learning_rate": 0.00015742569848372227, + "loss": 2.6908, + "step": 14398 + }, + { + "epoch": 0.61, + "grad_norm": 0.4140625, + "learning_rate": 0.00015742020673802784, + "loss": 2.5004, + "step": 14399 + }, + { + "epoch": 0.61, + "grad_norm": 0.482421875, + "learning_rate": 0.0001574147147339635, + "loss": 2.5849, + "step": 14400 + }, + { + "epoch": 0.61, + "grad_norm": 0.435546875, + "learning_rate": 0.00015740922247155388, + "loss": 2.709, + "step": 14401 + }, + { + "epoch": 0.61, + "grad_norm": 0.4609375, + "learning_rate": 0.0001574037299508238, + "loss": 2.4552, + "step": 14402 + }, + { + "epoch": 0.61, + "grad_norm": 0.4765625, + "learning_rate": 0.00015739823717179788, + "loss": 2.6398, + "step": 14403 + }, + { + "epoch": 0.61, + "grad_norm": 0.462890625, + "learning_rate": 0.00015739274413450086, + "loss": 2.5625, + "step": 14404 + }, + { + "epoch": 0.61, + "grad_norm": 0.44140625, + "learning_rate": 0.00015738725083895748, + "loss": 2.626, + "step": 14405 + }, + { + "epoch": 0.61, + "grad_norm": 0.44140625, + "learning_rate": 0.00015738175728519246, + "loss": 2.5593, + "step": 14406 + }, + { + "epoch": 0.61, + "grad_norm": 0.486328125, + "learning_rate": 0.00015737626347323043, + "loss": 2.6575, + "step": 14407 + }, + { + "epoch": 0.61, + "grad_norm": 0.43359375, + "learning_rate": 0.00015737076940309627, + "loss": 2.6034, + "step": 14408 + }, + { + "epoch": 0.61, + "grad_norm": 0.4609375, + "learning_rate": 0.00015736527507481457, + "loss": 2.4889, + "step": 14409 + }, + { + "epoch": 0.61, + "grad_norm": 0.42578125, + "learning_rate": 0.0001573597804884101, + "loss": 2.6435, + "step": 14410 + }, + { + "epoch": 0.61, + "grad_norm": 0.455078125, + "learning_rate": 0.0001573542856439076, + "loss": 2.5655, + "step": 14411 + }, + { + "epoch": 0.61, + "grad_norm": 0.44921875, + "learning_rate": 0.00015734879054133178, + "loss": 2.558, + "step": 14412 + }, + { + "epoch": 0.61, + "grad_norm": 0.41015625, + "learning_rate": 0.00015734329518070732, + "loss": 2.4519, + "step": 14413 + }, + { + "epoch": 0.61, + "grad_norm": 0.404296875, + "learning_rate": 0.00015733779956205902, + "loss": 2.5931, + "step": 14414 + }, + { + "epoch": 0.61, + "grad_norm": 0.466796875, + "learning_rate": 0.00015733230368541155, + "loss": 2.6582, + "step": 14415 + }, + { + "epoch": 0.61, + "grad_norm": 0.408203125, + "learning_rate": 0.00015732680755078973, + "loss": 2.4647, + "step": 14416 + }, + { + "epoch": 0.61, + "grad_norm": 0.41796875, + "learning_rate": 0.00015732131115821818, + "loss": 2.5953, + "step": 14417 + }, + { + "epoch": 0.61, + "grad_norm": 0.455078125, + "learning_rate": 0.00015731581450772168, + "loss": 2.594, + "step": 14418 + }, + { + "epoch": 0.61, + "grad_norm": 0.423828125, + "learning_rate": 0.00015731031759932496, + "loss": 2.4054, + "step": 14419 + }, + { + "epoch": 0.61, + "grad_norm": 0.45703125, + "learning_rate": 0.00015730482043305278, + "loss": 2.7157, + "step": 14420 + }, + { + "epoch": 0.61, + "grad_norm": 0.42578125, + "learning_rate": 0.0001572993230089298, + "loss": 2.5716, + "step": 14421 + }, + { + "epoch": 0.61, + "grad_norm": 0.482421875, + "learning_rate": 0.00015729382532698086, + "loss": 2.5657, + "step": 14422 + }, + { + "epoch": 0.61, + "grad_norm": 0.423828125, + "learning_rate": 0.0001572883273872306, + "loss": 2.7384, + "step": 14423 + }, + { + "epoch": 0.61, + "grad_norm": 0.453125, + "learning_rate": 0.0001572828291897038, + "loss": 2.4954, + "step": 14424 + }, + { + "epoch": 0.61, + "grad_norm": 0.427734375, + "learning_rate": 0.00015727733073442524, + "loss": 2.6333, + "step": 14425 + }, + { + "epoch": 0.61, + "grad_norm": 0.431640625, + "learning_rate": 0.00015727183202141962, + "loss": 2.6646, + "step": 14426 + }, + { + "epoch": 0.61, + "grad_norm": 0.40625, + "learning_rate": 0.00015726633305071167, + "loss": 2.609, + "step": 14427 + }, + { + "epoch": 0.61, + "grad_norm": 0.484375, + "learning_rate": 0.00015726083382232618, + "loss": 2.5187, + "step": 14428 + }, + { + "epoch": 0.61, + "grad_norm": 0.470703125, + "learning_rate": 0.00015725533433628782, + "loss": 2.7027, + "step": 14429 + }, + { + "epoch": 0.61, + "grad_norm": 0.392578125, + "learning_rate": 0.0001572498345926214, + "loss": 2.5179, + "step": 14430 + }, + { + "epoch": 0.61, + "grad_norm": 0.490234375, + "learning_rate": 0.00015724433459135166, + "loss": 2.7215, + "step": 14431 + }, + { + "epoch": 0.61, + "grad_norm": 0.400390625, + "learning_rate": 0.00015723883433250331, + "loss": 2.5276, + "step": 14432 + }, + { + "epoch": 0.62, + "grad_norm": 0.4765625, + "learning_rate": 0.00015723333381610112, + "loss": 2.6435, + "step": 14433 + }, + { + "epoch": 0.62, + "grad_norm": 0.44921875, + "learning_rate": 0.00015722783304216988, + "loss": 2.5019, + "step": 14434 + }, + { + "epoch": 0.62, + "grad_norm": 0.474609375, + "learning_rate": 0.00015722233201073427, + "loss": 2.7012, + "step": 14435 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.00015721683072181909, + "loss": 2.639, + "step": 14436 + }, + { + "epoch": 0.62, + "grad_norm": 0.478515625, + "learning_rate": 0.00015721132917544912, + "loss": 2.6161, + "step": 14437 + }, + { + "epoch": 0.62, + "grad_norm": 0.498046875, + "learning_rate": 0.00015720582737164903, + "loss": 2.5489, + "step": 14438 + }, + { + "epoch": 0.62, + "grad_norm": 0.42578125, + "learning_rate": 0.00015720032531044365, + "loss": 2.6584, + "step": 14439 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.00015719482299185767, + "loss": 2.5822, + "step": 14440 + }, + { + "epoch": 0.62, + "grad_norm": 0.44921875, + "learning_rate": 0.00015718932041591593, + "loss": 2.6669, + "step": 14441 + }, + { + "epoch": 0.62, + "grad_norm": 0.46875, + "learning_rate": 0.00015718381758264314, + "loss": 2.6436, + "step": 14442 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.00015717831449206407, + "loss": 2.5495, + "step": 14443 + }, + { + "epoch": 0.62, + "grad_norm": 0.41796875, + "learning_rate": 0.00015717281114420345, + "loss": 2.4777, + "step": 14444 + }, + { + "epoch": 0.62, + "grad_norm": 0.5078125, + "learning_rate": 0.0001571673075390861, + "loss": 2.6478, + "step": 14445 + }, + { + "epoch": 0.62, + "grad_norm": 0.4609375, + "learning_rate": 0.00015716180367673677, + "loss": 2.503, + "step": 14446 + }, + { + "epoch": 0.62, + "grad_norm": 0.5, + "learning_rate": 0.0001571562995571802, + "loss": 2.6184, + "step": 14447 + }, + { + "epoch": 0.62, + "grad_norm": 0.484375, + "learning_rate": 0.00015715079518044116, + "loss": 2.5527, + "step": 14448 + }, + { + "epoch": 0.62, + "grad_norm": 0.4921875, + "learning_rate": 0.00015714529054654443, + "loss": 2.6054, + "step": 14449 + }, + { + "epoch": 0.62, + "grad_norm": 0.421875, + "learning_rate": 0.00015713978565551478, + "loss": 2.6598, + "step": 14450 + }, + { + "epoch": 0.62, + "grad_norm": 0.48046875, + "learning_rate": 0.000157134280507377, + "loss": 2.6128, + "step": 14451 + }, + { + "epoch": 0.62, + "grad_norm": 0.5, + "learning_rate": 0.00015712877510215583, + "loss": 2.5384, + "step": 14452 + }, + { + "epoch": 0.62, + "grad_norm": 0.482421875, + "learning_rate": 0.00015712326943987604, + "loss": 2.5633, + "step": 14453 + }, + { + "epoch": 0.62, + "grad_norm": 0.5390625, + "learning_rate": 0.0001571177635205624, + "loss": 2.5837, + "step": 14454 + }, + { + "epoch": 0.62, + "grad_norm": 0.486328125, + "learning_rate": 0.00015711225734423972, + "loss": 2.6815, + "step": 14455 + }, + { + "epoch": 0.62, + "grad_norm": 0.43359375, + "learning_rate": 0.00015710675091093276, + "loss": 2.5685, + "step": 14456 + }, + { + "epoch": 0.62, + "grad_norm": 0.44140625, + "learning_rate": 0.00015710124422066624, + "loss": 2.5153, + "step": 14457 + }, + { + "epoch": 0.62, + "grad_norm": 0.486328125, + "learning_rate": 0.00015709573727346502, + "loss": 2.641, + "step": 14458 + }, + { + "epoch": 0.62, + "grad_norm": 0.447265625, + "learning_rate": 0.00015709023006935387, + "loss": 2.6548, + "step": 14459 + }, + { + "epoch": 0.62, + "grad_norm": 0.4609375, + "learning_rate": 0.00015708472260835753, + "loss": 2.5252, + "step": 14460 + }, + { + "epoch": 0.62, + "grad_norm": 0.46875, + "learning_rate": 0.00015707921489050084, + "loss": 2.5926, + "step": 14461 + }, + { + "epoch": 0.62, + "grad_norm": 0.54296875, + "learning_rate": 0.0001570737069158085, + "loss": 2.6325, + "step": 14462 + }, + { + "epoch": 0.62, + "grad_norm": 0.431640625, + "learning_rate": 0.00015706819868430532, + "loss": 2.504, + "step": 14463 + }, + { + "epoch": 0.62, + "grad_norm": 0.431640625, + "learning_rate": 0.0001570626901960161, + "loss": 2.5075, + "step": 14464 + }, + { + "epoch": 0.62, + "grad_norm": 0.46484375, + "learning_rate": 0.0001570571814509657, + "loss": 2.5458, + "step": 14465 + }, + { + "epoch": 0.62, + "grad_norm": 0.4375, + "learning_rate": 0.0001570516724491788, + "loss": 2.6276, + "step": 14466 + }, + { + "epoch": 0.62, + "grad_norm": 0.41796875, + "learning_rate": 0.0001570461631906802, + "loss": 2.5217, + "step": 14467 + }, + { + "epoch": 0.62, + "grad_norm": 0.49609375, + "learning_rate": 0.0001570406536754947, + "loss": 2.6025, + "step": 14468 + }, + { + "epoch": 0.62, + "grad_norm": 0.46484375, + "learning_rate": 0.00015703514390364714, + "loss": 2.4728, + "step": 14469 + }, + { + "epoch": 0.62, + "grad_norm": 0.423828125, + "learning_rate": 0.00015702963387516228, + "loss": 2.5995, + "step": 14470 + }, + { + "epoch": 0.62, + "grad_norm": 0.44921875, + "learning_rate": 0.0001570241235900649, + "loss": 2.6978, + "step": 14471 + }, + { + "epoch": 0.62, + "grad_norm": 0.48046875, + "learning_rate": 0.00015701861304837978, + "loss": 2.6871, + "step": 14472 + }, + { + "epoch": 0.62, + "grad_norm": 0.427734375, + "learning_rate": 0.00015701310225013177, + "loss": 2.5401, + "step": 14473 + }, + { + "epoch": 0.62, + "grad_norm": 0.4453125, + "learning_rate": 0.00015700759119534563, + "loss": 2.545, + "step": 14474 + }, + { + "epoch": 0.62, + "grad_norm": 0.423828125, + "learning_rate": 0.00015700207988404615, + "loss": 2.6328, + "step": 14475 + }, + { + "epoch": 0.62, + "grad_norm": 0.455078125, + "learning_rate": 0.00015699656831625816, + "loss": 2.4815, + "step": 14476 + }, + { + "epoch": 0.62, + "grad_norm": 0.447265625, + "learning_rate": 0.00015699105649200645, + "loss": 2.6524, + "step": 14477 + }, + { + "epoch": 0.62, + "grad_norm": 0.4765625, + "learning_rate": 0.00015698554441131578, + "loss": 2.7073, + "step": 14478 + }, + { + "epoch": 0.62, + "grad_norm": 0.443359375, + "learning_rate": 0.00015698003207421102, + "loss": 2.556, + "step": 14479 + }, + { + "epoch": 0.62, + "grad_norm": 0.435546875, + "learning_rate": 0.0001569745194807169, + "loss": 2.4882, + "step": 14480 + }, + { + "epoch": 0.62, + "grad_norm": 0.46484375, + "learning_rate": 0.00015696900663085832, + "loss": 2.6582, + "step": 14481 + }, + { + "epoch": 0.62, + "grad_norm": 0.46484375, + "learning_rate": 0.00015696349352465998, + "loss": 2.6843, + "step": 14482 + }, + { + "epoch": 0.62, + "grad_norm": 0.412109375, + "learning_rate": 0.00015695798016214675, + "loss": 2.5815, + "step": 14483 + }, + { + "epoch": 0.62, + "grad_norm": 0.470703125, + "learning_rate": 0.00015695246654334343, + "loss": 2.4945, + "step": 14484 + }, + { + "epoch": 0.62, + "grad_norm": 0.482421875, + "learning_rate": 0.00015694695266827485, + "loss": 2.6002, + "step": 14485 + }, + { + "epoch": 0.62, + "grad_norm": 0.490234375, + "learning_rate": 0.00015694143853696573, + "loss": 2.6444, + "step": 14486 + }, + { + "epoch": 0.62, + "grad_norm": 0.408203125, + "learning_rate": 0.000156935924149441, + "loss": 2.6161, + "step": 14487 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.0001569304095057254, + "loss": 2.6197, + "step": 14488 + }, + { + "epoch": 0.62, + "grad_norm": 0.43359375, + "learning_rate": 0.0001569248946058438, + "loss": 2.5808, + "step": 14489 + }, + { + "epoch": 0.62, + "grad_norm": 0.4609375, + "learning_rate": 0.00015691937944982092, + "loss": 2.5482, + "step": 14490 + }, + { + "epoch": 0.62, + "grad_norm": 0.435546875, + "learning_rate": 0.00015691386403768166, + "loss": 2.6295, + "step": 14491 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.0001569083483694508, + "loss": 2.5757, + "step": 14492 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.00015690283244515317, + "loss": 2.5728, + "step": 14493 + }, + { + "epoch": 0.62, + "grad_norm": 0.46484375, + "learning_rate": 0.0001568973162648136, + "loss": 2.5411, + "step": 14494 + }, + { + "epoch": 0.62, + "grad_norm": 0.51171875, + "learning_rate": 0.00015689179982845692, + "loss": 2.6376, + "step": 14495 + }, + { + "epoch": 0.62, + "grad_norm": 0.4921875, + "learning_rate": 0.0001568862831361079, + "loss": 2.6781, + "step": 14496 + }, + { + "epoch": 0.62, + "grad_norm": 0.49609375, + "learning_rate": 0.00015688076618779144, + "loss": 2.596, + "step": 14497 + }, + { + "epoch": 0.62, + "grad_norm": 0.435546875, + "learning_rate": 0.00015687524898353228, + "loss": 2.6546, + "step": 14498 + }, + { + "epoch": 0.62, + "grad_norm": 0.50390625, + "learning_rate": 0.00015686973152335528, + "loss": 2.7229, + "step": 14499 + }, + { + "epoch": 0.62, + "grad_norm": 0.44921875, + "learning_rate": 0.0001568642138072853, + "loss": 2.5556, + "step": 14500 + }, + { + "epoch": 0.62, + "grad_norm": 0.474609375, + "learning_rate": 0.00015685869583534711, + "loss": 2.572, + "step": 14501 + }, + { + "epoch": 0.62, + "grad_norm": 0.46875, + "learning_rate": 0.0001568531776075656, + "loss": 2.6084, + "step": 14502 + }, + { + "epoch": 0.62, + "grad_norm": 0.404296875, + "learning_rate": 0.00015684765912396556, + "loss": 2.5446, + "step": 14503 + }, + { + "epoch": 0.62, + "grad_norm": 0.451171875, + "learning_rate": 0.00015684214038457178, + "loss": 2.6757, + "step": 14504 + }, + { + "epoch": 0.62, + "grad_norm": 0.48046875, + "learning_rate": 0.0001568366213894092, + "loss": 2.553, + "step": 14505 + }, + { + "epoch": 0.62, + "grad_norm": 0.46875, + "learning_rate": 0.0001568311021385026, + "loss": 2.5878, + "step": 14506 + }, + { + "epoch": 0.62, + "grad_norm": 0.4296875, + "learning_rate": 0.00015682558263187678, + "loss": 2.4035, + "step": 14507 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.0001568200628695566, + "loss": 2.5999, + "step": 14508 + }, + { + "epoch": 0.62, + "grad_norm": 0.416015625, + "learning_rate": 0.00015681454285156691, + "loss": 2.5809, + "step": 14509 + }, + { + "epoch": 0.62, + "grad_norm": 0.4140625, + "learning_rate": 0.00015680902257793252, + "loss": 2.5582, + "step": 14510 + }, + { + "epoch": 0.62, + "grad_norm": 0.462890625, + "learning_rate": 0.00015680350204867836, + "loss": 2.562, + "step": 14511 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.0001567979812638291, + "loss": 2.5694, + "step": 14512 + }, + { + "epoch": 0.62, + "grad_norm": 0.412109375, + "learning_rate": 0.00015679246022340978, + "loss": 2.6626, + "step": 14513 + }, + { + "epoch": 0.62, + "grad_norm": 0.412109375, + "learning_rate": 0.00015678693892744508, + "loss": 2.4934, + "step": 14514 + }, + { + "epoch": 0.62, + "grad_norm": 0.443359375, + "learning_rate": 0.00015678141737595995, + "loss": 2.6029, + "step": 14515 + }, + { + "epoch": 0.62, + "grad_norm": 0.490234375, + "learning_rate": 0.00015677589556897917, + "loss": 2.6614, + "step": 14516 + }, + { + "epoch": 0.62, + "grad_norm": 0.53125, + "learning_rate": 0.0001567703735065276, + "loss": 2.6065, + "step": 14517 + }, + { + "epoch": 0.62, + "grad_norm": 0.455078125, + "learning_rate": 0.00015676485118863012, + "loss": 2.6887, + "step": 14518 + }, + { + "epoch": 0.62, + "grad_norm": 0.439453125, + "learning_rate": 0.00015675932861531152, + "loss": 2.5978, + "step": 14519 + }, + { + "epoch": 0.62, + "grad_norm": 0.51171875, + "learning_rate": 0.00015675380578659672, + "loss": 2.6737, + "step": 14520 + }, + { + "epoch": 0.62, + "grad_norm": 0.39453125, + "learning_rate": 0.00015674828270251053, + "loss": 2.5065, + "step": 14521 + }, + { + "epoch": 0.62, + "grad_norm": 0.404296875, + "learning_rate": 0.0001567427593630778, + "loss": 2.5513, + "step": 14522 + }, + { + "epoch": 0.62, + "grad_norm": 0.48828125, + "learning_rate": 0.00015673723576832338, + "loss": 2.5378, + "step": 14523 + }, + { + "epoch": 0.62, + "grad_norm": 0.478515625, + "learning_rate": 0.00015673171191827215, + "loss": 2.5744, + "step": 14524 + }, + { + "epoch": 0.62, + "grad_norm": 0.455078125, + "learning_rate": 0.00015672618781294894, + "loss": 2.5572, + "step": 14525 + }, + { + "epoch": 0.62, + "grad_norm": 0.4921875, + "learning_rate": 0.00015672066345237862, + "loss": 2.4903, + "step": 14526 + }, + { + "epoch": 0.62, + "grad_norm": 0.458984375, + "learning_rate": 0.00015671513883658603, + "loss": 2.536, + "step": 14527 + }, + { + "epoch": 0.62, + "grad_norm": 0.44140625, + "learning_rate": 0.00015670961396559605, + "loss": 2.6739, + "step": 14528 + }, + { + "epoch": 0.62, + "grad_norm": 0.40625, + "learning_rate": 0.00015670408883943355, + "loss": 2.5177, + "step": 14529 + }, + { + "epoch": 0.62, + "grad_norm": 0.41015625, + "learning_rate": 0.00015669856345812335, + "loss": 2.6583, + "step": 14530 + }, + { + "epoch": 0.62, + "grad_norm": 0.4765625, + "learning_rate": 0.00015669303782169034, + "loss": 2.5955, + "step": 14531 + }, + { + "epoch": 0.62, + "grad_norm": 0.435546875, + "learning_rate": 0.00015668751193015937, + "loss": 2.5128, + "step": 14532 + }, + { + "epoch": 0.62, + "grad_norm": 0.4375, + "learning_rate": 0.00015668198578355536, + "loss": 2.4875, + "step": 14533 + }, + { + "epoch": 0.62, + "grad_norm": 0.419921875, + "learning_rate": 0.00015667645938190308, + "loss": 2.6886, + "step": 14534 + }, + { + "epoch": 0.62, + "grad_norm": 0.478515625, + "learning_rate": 0.00015667093272522746, + "loss": 2.5451, + "step": 14535 + }, + { + "epoch": 0.62, + "grad_norm": 0.482421875, + "learning_rate": 0.00015666540581355337, + "loss": 2.6112, + "step": 14536 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.00015665987864690566, + "loss": 2.544, + "step": 14537 + }, + { + "epoch": 0.62, + "grad_norm": 0.462890625, + "learning_rate": 0.00015665435122530917, + "loss": 2.5402, + "step": 14538 + }, + { + "epoch": 0.62, + "grad_norm": 0.51171875, + "learning_rate": 0.00015664882354878882, + "loss": 2.7282, + "step": 14539 + }, + { + "epoch": 0.62, + "grad_norm": 0.34765625, + "learning_rate": 0.0001566432956173695, + "loss": 2.3937, + "step": 14540 + }, + { + "epoch": 0.62, + "grad_norm": 0.40625, + "learning_rate": 0.000156637767431076, + "loss": 2.5271, + "step": 14541 + }, + { + "epoch": 0.62, + "grad_norm": 0.4453125, + "learning_rate": 0.00015663223898993326, + "loss": 2.7401, + "step": 14542 + }, + { + "epoch": 0.62, + "grad_norm": 0.490234375, + "learning_rate": 0.00015662671029396616, + "loss": 2.4511, + "step": 14543 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.00015662118134319956, + "loss": 2.5607, + "step": 14544 + }, + { + "epoch": 0.62, + "grad_norm": 0.48046875, + "learning_rate": 0.00015661565213765833, + "loss": 2.5768, + "step": 14545 + }, + { + "epoch": 0.62, + "grad_norm": 0.52734375, + "learning_rate": 0.00015661012267736735, + "loss": 2.6332, + "step": 14546 + }, + { + "epoch": 0.62, + "grad_norm": 0.40625, + "learning_rate": 0.0001566045929623515, + "loss": 2.5537, + "step": 14547 + }, + { + "epoch": 0.62, + "grad_norm": 0.5, + "learning_rate": 0.00015659906299263568, + "loss": 2.7679, + "step": 14548 + }, + { + "epoch": 0.62, + "grad_norm": 0.447265625, + "learning_rate": 0.00015659353276824473, + "loss": 2.6, + "step": 14549 + }, + { + "epoch": 0.62, + "grad_norm": 0.453125, + "learning_rate": 0.0001565880022892036, + "loss": 2.732, + "step": 14550 + }, + { + "epoch": 0.62, + "grad_norm": 0.458984375, + "learning_rate": 0.0001565824715555371, + "loss": 2.7237, + "step": 14551 + }, + { + "epoch": 0.62, + "grad_norm": 0.419921875, + "learning_rate": 0.00015657694056727021, + "loss": 2.6266, + "step": 14552 + }, + { + "epoch": 0.62, + "grad_norm": 0.466796875, + "learning_rate": 0.00015657140932442775, + "loss": 2.6381, + "step": 14553 + }, + { + "epoch": 0.62, + "grad_norm": 0.462890625, + "learning_rate": 0.0001565658778270346, + "loss": 2.5969, + "step": 14554 + }, + { + "epoch": 0.62, + "grad_norm": 0.388671875, + "learning_rate": 0.00015656034607511567, + "loss": 2.3749, + "step": 14555 + }, + { + "epoch": 0.62, + "grad_norm": 0.5, + "learning_rate": 0.0001565548140686959, + "loss": 2.5978, + "step": 14556 + }, + { + "epoch": 0.62, + "grad_norm": 0.5078125, + "learning_rate": 0.00015654928180780007, + "loss": 2.5949, + "step": 14557 + }, + { + "epoch": 0.62, + "grad_norm": 0.439453125, + "learning_rate": 0.0001565437492924532, + "loss": 2.6348, + "step": 14558 + }, + { + "epoch": 0.62, + "grad_norm": 0.416015625, + "learning_rate": 0.0001565382165226801, + "loss": 2.4631, + "step": 14559 + }, + { + "epoch": 0.62, + "grad_norm": 0.4296875, + "learning_rate": 0.0001565326834985057, + "loss": 2.6478, + "step": 14560 + }, + { + "epoch": 0.62, + "grad_norm": 0.43359375, + "learning_rate": 0.00015652715021995483, + "loss": 2.5912, + "step": 14561 + }, + { + "epoch": 0.62, + "grad_norm": 0.494140625, + "learning_rate": 0.00015652161668705249, + "loss": 2.5983, + "step": 14562 + }, + { + "epoch": 0.62, + "grad_norm": 0.48828125, + "learning_rate": 0.00015651608289982352, + "loss": 2.6559, + "step": 14563 + }, + { + "epoch": 0.62, + "grad_norm": 0.41796875, + "learning_rate": 0.00015651054885829283, + "loss": 2.6307, + "step": 14564 + }, + { + "epoch": 0.62, + "grad_norm": 0.369140625, + "learning_rate": 0.00015650501456248534, + "loss": 2.4905, + "step": 14565 + }, + { + "epoch": 0.62, + "grad_norm": 0.4375, + "learning_rate": 0.0001564994800124259, + "loss": 2.4909, + "step": 14566 + }, + { + "epoch": 0.62, + "grad_norm": 0.453125, + "learning_rate": 0.00015649394520813947, + "loss": 2.581, + "step": 14567 + }, + { + "epoch": 0.62, + "grad_norm": 0.453125, + "learning_rate": 0.00015648841014965093, + "loss": 2.6351, + "step": 14568 + }, + { + "epoch": 0.62, + "grad_norm": 0.482421875, + "learning_rate": 0.00015648287483698519, + "loss": 2.4988, + "step": 14569 + }, + { + "epoch": 0.62, + "grad_norm": 0.4609375, + "learning_rate": 0.00015647733927016716, + "loss": 2.6345, + "step": 14570 + }, + { + "epoch": 0.62, + "grad_norm": 0.3828125, + "learning_rate": 0.00015647180344922172, + "loss": 2.4928, + "step": 14571 + }, + { + "epoch": 0.62, + "grad_norm": 0.5078125, + "learning_rate": 0.00015646626737417384, + "loss": 2.6315, + "step": 14572 + }, + { + "epoch": 0.62, + "grad_norm": 0.458984375, + "learning_rate": 0.00015646073104504837, + "loss": 2.6966, + "step": 14573 + }, + { + "epoch": 0.62, + "grad_norm": 0.5078125, + "learning_rate": 0.00015645519446187028, + "loss": 2.742, + "step": 14574 + }, + { + "epoch": 0.62, + "grad_norm": 0.478515625, + "learning_rate": 0.00015644965762466438, + "loss": 2.5257, + "step": 14575 + }, + { + "epoch": 0.62, + "grad_norm": 0.4453125, + "learning_rate": 0.0001564441205334557, + "loss": 2.6092, + "step": 14576 + }, + { + "epoch": 0.62, + "grad_norm": 0.494140625, + "learning_rate": 0.0001564385831882691, + "loss": 2.6714, + "step": 14577 + }, + { + "epoch": 0.62, + "grad_norm": 0.486328125, + "learning_rate": 0.0001564330455891295, + "loss": 2.623, + "step": 14578 + }, + { + "epoch": 0.62, + "grad_norm": 0.412109375, + "learning_rate": 0.0001564275077360618, + "loss": 2.498, + "step": 14579 + }, + { + "epoch": 0.62, + "grad_norm": 0.50390625, + "learning_rate": 0.00015642196962909095, + "loss": 2.7182, + "step": 14580 + }, + { + "epoch": 0.62, + "grad_norm": 0.51171875, + "learning_rate": 0.00015641643126824184, + "loss": 2.6338, + "step": 14581 + }, + { + "epoch": 0.62, + "grad_norm": 0.5078125, + "learning_rate": 0.00015641089265353944, + "loss": 2.5653, + "step": 14582 + }, + { + "epoch": 0.62, + "grad_norm": 0.474609375, + "learning_rate": 0.0001564053537850086, + "loss": 2.6591, + "step": 14583 + }, + { + "epoch": 0.62, + "grad_norm": 0.416015625, + "learning_rate": 0.0001563998146626743, + "loss": 2.5987, + "step": 14584 + }, + { + "epoch": 0.62, + "grad_norm": 0.486328125, + "learning_rate": 0.00015639427528656144, + "loss": 2.6271, + "step": 14585 + }, + { + "epoch": 0.62, + "grad_norm": 0.412109375, + "learning_rate": 0.00015638873565669497, + "loss": 2.5063, + "step": 14586 + }, + { + "epoch": 0.62, + "grad_norm": 0.482421875, + "learning_rate": 0.00015638319577309975, + "loss": 2.6879, + "step": 14587 + }, + { + "epoch": 0.62, + "grad_norm": 0.53515625, + "learning_rate": 0.00015637765563580086, + "loss": 2.6808, + "step": 14588 + }, + { + "epoch": 0.62, + "grad_norm": 0.447265625, + "learning_rate": 0.00015637211524482302, + "loss": 2.6251, + "step": 14589 + }, + { + "epoch": 0.62, + "grad_norm": 0.46875, + "learning_rate": 0.0001563665746001913, + "loss": 2.5772, + "step": 14590 + }, + { + "epoch": 0.62, + "grad_norm": 0.392578125, + "learning_rate": 0.00015636103370193059, + "loss": 2.4406, + "step": 14591 + }, + { + "epoch": 0.62, + "grad_norm": 0.439453125, + "learning_rate": 0.00015635549255006584, + "loss": 2.6072, + "step": 14592 + }, + { + "epoch": 0.62, + "grad_norm": 0.51953125, + "learning_rate": 0.00015634995114462193, + "loss": 2.5391, + "step": 14593 + }, + { + "epoch": 0.62, + "grad_norm": 0.498046875, + "learning_rate": 0.00015634440948562387, + "loss": 2.6591, + "step": 14594 + }, + { + "epoch": 0.62, + "grad_norm": 0.490234375, + "learning_rate": 0.00015633886757309654, + "loss": 2.5393, + "step": 14595 + }, + { + "epoch": 0.62, + "grad_norm": 0.435546875, + "learning_rate": 0.0001563333254070649, + "loss": 2.5624, + "step": 14596 + }, + { + "epoch": 0.62, + "grad_norm": 0.416015625, + "learning_rate": 0.0001563277829875539, + "loss": 2.5234, + "step": 14597 + }, + { + "epoch": 0.62, + "grad_norm": 0.48046875, + "learning_rate": 0.00015632224031458843, + "loss": 2.673, + "step": 14598 + }, + { + "epoch": 0.62, + "grad_norm": 0.443359375, + "learning_rate": 0.0001563166973881935, + "loss": 2.6095, + "step": 14599 + }, + { + "epoch": 0.62, + "grad_norm": 0.421875, + "learning_rate": 0.000156311154208394, + "loss": 2.5086, + "step": 14600 + }, + { + "epoch": 0.62, + "grad_norm": 0.4453125, + "learning_rate": 0.00015630561077521488, + "loss": 2.5801, + "step": 14601 + }, + { + "epoch": 0.62, + "grad_norm": 0.51171875, + "learning_rate": 0.00015630006708868108, + "loss": 2.6424, + "step": 14602 + }, + { + "epoch": 0.62, + "grad_norm": 0.48828125, + "learning_rate": 0.00015629452314881755, + "loss": 2.5766, + "step": 14603 + }, + { + "epoch": 0.62, + "grad_norm": 0.44140625, + "learning_rate": 0.0001562889789556493, + "loss": 2.5955, + "step": 14604 + }, + { + "epoch": 0.62, + "grad_norm": 0.431640625, + "learning_rate": 0.00015628343450920114, + "loss": 2.5298, + "step": 14605 + }, + { + "epoch": 0.62, + "grad_norm": 0.470703125, + "learning_rate": 0.00015627788980949813, + "loss": 2.7691, + "step": 14606 + }, + { + "epoch": 0.62, + "grad_norm": 0.43359375, + "learning_rate": 0.00015627234485656515, + "loss": 2.4899, + "step": 14607 + }, + { + "epoch": 0.62, + "grad_norm": 0.46875, + "learning_rate": 0.00015626679965042724, + "loss": 2.5234, + "step": 14608 + }, + { + "epoch": 0.62, + "grad_norm": 0.42578125, + "learning_rate": 0.00015626125419110923, + "loss": 2.5028, + "step": 14609 + }, + { + "epoch": 0.62, + "grad_norm": 0.46875, + "learning_rate": 0.0001562557084786362, + "loss": 2.669, + "step": 14610 + }, + { + "epoch": 0.62, + "grad_norm": 0.443359375, + "learning_rate": 0.00015625016251303298, + "loss": 2.6344, + "step": 14611 + }, + { + "epoch": 0.62, + "grad_norm": 0.447265625, + "learning_rate": 0.00015624461629432465, + "loss": 2.4959, + "step": 14612 + }, + { + "epoch": 0.62, + "grad_norm": 0.44140625, + "learning_rate": 0.00015623906982253604, + "loss": 2.7077, + "step": 14613 + }, + { + "epoch": 0.62, + "grad_norm": 0.4609375, + "learning_rate": 0.0001562335230976922, + "loss": 2.5757, + "step": 14614 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.00015622797611981805, + "loss": 2.6565, + "step": 14615 + }, + { + "epoch": 0.62, + "grad_norm": 0.390625, + "learning_rate": 0.00015622242888893857, + "loss": 2.5624, + "step": 14616 + }, + { + "epoch": 0.62, + "grad_norm": 0.408203125, + "learning_rate": 0.0001562168814050787, + "loss": 2.6274, + "step": 14617 + }, + { + "epoch": 0.62, + "grad_norm": 0.482421875, + "learning_rate": 0.0001562113336682634, + "loss": 2.5189, + "step": 14618 + }, + { + "epoch": 0.62, + "grad_norm": 0.46875, + "learning_rate": 0.00015620578567851763, + "loss": 2.5447, + "step": 14619 + }, + { + "epoch": 0.62, + "grad_norm": 0.453125, + "learning_rate": 0.00015620023743586643, + "loss": 2.5728, + "step": 14620 + }, + { + "epoch": 0.62, + "grad_norm": 0.458984375, + "learning_rate": 0.00015619468894033462, + "loss": 2.7345, + "step": 14621 + }, + { + "epoch": 0.62, + "grad_norm": 0.44921875, + "learning_rate": 0.00015618914019194732, + "loss": 2.6136, + "step": 14622 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.00015618359119072935, + "loss": 2.7251, + "step": 14623 + }, + { + "epoch": 0.62, + "grad_norm": 0.4765625, + "learning_rate": 0.00015617804193670577, + "loss": 2.6653, + "step": 14624 + }, + { + "epoch": 0.62, + "grad_norm": 0.4296875, + "learning_rate": 0.00015617249242990156, + "loss": 2.5312, + "step": 14625 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.00015616694267034164, + "loss": 2.529, + "step": 14626 + }, + { + "epoch": 0.62, + "grad_norm": 0.376953125, + "learning_rate": 0.000156161392658051, + "loss": 2.506, + "step": 14627 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.00015615584239305467, + "loss": 2.568, + "step": 14628 + }, + { + "epoch": 0.62, + "grad_norm": 0.453125, + "learning_rate": 0.00015615029187537752, + "loss": 2.6309, + "step": 14629 + }, + { + "epoch": 0.62, + "grad_norm": 0.44921875, + "learning_rate": 0.0001561447411050446, + "loss": 2.5948, + "step": 14630 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.00015613919008208087, + "loss": 2.5579, + "step": 14631 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.00015613363880651127, + "loss": 2.4978, + "step": 14632 + }, + { + "epoch": 0.62, + "grad_norm": 0.53125, + "learning_rate": 0.00015612808727836081, + "loss": 2.6458, + "step": 14633 + }, + { + "epoch": 0.62, + "grad_norm": 0.44921875, + "learning_rate": 0.00015612253549765449, + "loss": 2.5475, + "step": 14634 + }, + { + "epoch": 0.62, + "grad_norm": 0.44921875, + "learning_rate": 0.00015611698346441728, + "loss": 2.521, + "step": 14635 + }, + { + "epoch": 0.62, + "grad_norm": 0.4375, + "learning_rate": 0.00015611143117867416, + "loss": 2.4905, + "step": 14636 + }, + { + "epoch": 0.62, + "grad_norm": 0.400390625, + "learning_rate": 0.00015610587864045008, + "loss": 2.6563, + "step": 14637 + }, + { + "epoch": 0.62, + "grad_norm": 0.49609375, + "learning_rate": 0.00015610032584977004, + "loss": 2.6462, + "step": 14638 + }, + { + "epoch": 0.62, + "grad_norm": 0.435546875, + "learning_rate": 0.00015609477280665904, + "loss": 2.5909, + "step": 14639 + }, + { + "epoch": 0.62, + "grad_norm": 0.49609375, + "learning_rate": 0.00015608921951114208, + "loss": 2.6231, + "step": 14640 + }, + { + "epoch": 0.62, + "grad_norm": 0.423828125, + "learning_rate": 0.00015608366596324413, + "loss": 2.5198, + "step": 14641 + }, + { + "epoch": 0.62, + "grad_norm": 0.546875, + "learning_rate": 0.00015607811216299016, + "loss": 2.7226, + "step": 14642 + }, + { + "epoch": 0.62, + "grad_norm": 0.49609375, + "learning_rate": 0.00015607255811040521, + "loss": 2.593, + "step": 14643 + }, + { + "epoch": 0.62, + "grad_norm": 0.4609375, + "learning_rate": 0.00015606700380551422, + "loss": 2.5199, + "step": 14644 + }, + { + "epoch": 0.62, + "grad_norm": 0.42578125, + "learning_rate": 0.00015606144924834223, + "loss": 2.6326, + "step": 14645 + }, + { + "epoch": 0.62, + "grad_norm": 0.447265625, + "learning_rate": 0.00015605589443891418, + "loss": 2.5885, + "step": 14646 + }, + { + "epoch": 0.62, + "grad_norm": 0.4609375, + "learning_rate": 0.00015605033937725508, + "loss": 2.6186, + "step": 14647 + }, + { + "epoch": 0.62, + "grad_norm": 0.41015625, + "learning_rate": 0.00015604478406339, + "loss": 2.6028, + "step": 14648 + }, + { + "epoch": 0.62, + "grad_norm": 0.466796875, + "learning_rate": 0.0001560392284973438, + "loss": 2.5975, + "step": 14649 + }, + { + "epoch": 0.62, + "grad_norm": 0.51953125, + "learning_rate": 0.00015603367267914158, + "loss": 2.5637, + "step": 14650 + }, + { + "epoch": 0.62, + "grad_norm": 0.427734375, + "learning_rate": 0.0001560281166088083, + "loss": 2.4393, + "step": 14651 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.00015602256028636897, + "loss": 2.6623, + "step": 14652 + }, + { + "epoch": 0.62, + "grad_norm": 0.4296875, + "learning_rate": 0.00015601700371184858, + "loss": 2.4373, + "step": 14653 + }, + { + "epoch": 0.62, + "grad_norm": 0.416015625, + "learning_rate": 0.0001560114468852722, + "loss": 2.4843, + "step": 14654 + }, + { + "epoch": 0.62, + "grad_norm": 0.466796875, + "learning_rate": 0.00015600588980666476, + "loss": 2.4547, + "step": 14655 + }, + { + "epoch": 0.62, + "grad_norm": 0.494140625, + "learning_rate": 0.00015600033247605127, + "loss": 2.6201, + "step": 14656 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.00015599477489345673, + "loss": 2.5865, + "step": 14657 + }, + { + "epoch": 0.62, + "grad_norm": 0.439453125, + "learning_rate": 0.00015598921705890618, + "loss": 2.5102, + "step": 14658 + }, + { + "epoch": 0.62, + "grad_norm": 0.45703125, + "learning_rate": 0.00015598365897242462, + "loss": 2.6396, + "step": 14659 + }, + { + "epoch": 0.62, + "grad_norm": 0.51953125, + "learning_rate": 0.00015597810063403705, + "loss": 2.6481, + "step": 14660 + }, + { + "epoch": 0.62, + "grad_norm": 0.470703125, + "learning_rate": 0.00015597254204376853, + "loss": 2.6118, + "step": 14661 + }, + { + "epoch": 0.62, + "grad_norm": 0.5, + "learning_rate": 0.000155966983201644, + "loss": 2.6196, + "step": 14662 + }, + { + "epoch": 0.62, + "grad_norm": 0.484375, + "learning_rate": 0.00015596142410768845, + "loss": 2.594, + "step": 14663 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.000155955864761927, + "loss": 2.6423, + "step": 14664 + }, + { + "epoch": 0.62, + "grad_norm": 0.47265625, + "learning_rate": 0.0001559503051643846, + "loss": 2.5861, + "step": 14665 + }, + { + "epoch": 0.62, + "grad_norm": 0.4765625, + "learning_rate": 0.00015594474531508621, + "loss": 2.6044, + "step": 14666 + }, + { + "epoch": 0.63, + "grad_norm": 0.421875, + "learning_rate": 0.000155939185214057, + "loss": 2.5567, + "step": 14667 + }, + { + "epoch": 0.63, + "grad_norm": 0.49609375, + "learning_rate": 0.00015593362486132186, + "loss": 2.6401, + "step": 14668 + }, + { + "epoch": 0.63, + "grad_norm": 0.416015625, + "learning_rate": 0.00015592806425690584, + "loss": 2.5482, + "step": 14669 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.00015592250340083397, + "loss": 2.6224, + "step": 14670 + }, + { + "epoch": 0.63, + "grad_norm": 0.51171875, + "learning_rate": 0.00015591694229313133, + "loss": 2.4742, + "step": 14671 + }, + { + "epoch": 0.63, + "grad_norm": 0.462890625, + "learning_rate": 0.0001559113809338228, + "loss": 2.6533, + "step": 14672 + }, + { + "epoch": 0.63, + "grad_norm": 0.4296875, + "learning_rate": 0.00015590581932293355, + "loss": 2.6443, + "step": 14673 + }, + { + "epoch": 0.63, + "grad_norm": 0.53515625, + "learning_rate": 0.0001559002574604885, + "loss": 2.7951, + "step": 14674 + }, + { + "epoch": 0.63, + "grad_norm": 0.486328125, + "learning_rate": 0.00015589469534651277, + "loss": 2.6511, + "step": 14675 + }, + { + "epoch": 0.63, + "grad_norm": 0.5390625, + "learning_rate": 0.00015588913298103128, + "loss": 2.5758, + "step": 14676 + }, + { + "epoch": 0.63, + "grad_norm": 0.486328125, + "learning_rate": 0.00015588357036406918, + "loss": 2.5858, + "step": 14677 + }, + { + "epoch": 0.63, + "grad_norm": 0.39453125, + "learning_rate": 0.0001558780074956514, + "loss": 2.4954, + "step": 14678 + }, + { + "epoch": 0.63, + "grad_norm": 0.50390625, + "learning_rate": 0.00015587244437580302, + "loss": 2.5124, + "step": 14679 + }, + { + "epoch": 0.63, + "grad_norm": 0.419921875, + "learning_rate": 0.00015586688100454905, + "loss": 2.6448, + "step": 14680 + }, + { + "epoch": 0.63, + "grad_norm": 0.453125, + "learning_rate": 0.00015586131738191454, + "loss": 2.5461, + "step": 14681 + }, + { + "epoch": 0.63, + "grad_norm": 0.50390625, + "learning_rate": 0.0001558557535079245, + "loss": 2.6452, + "step": 14682 + }, + { + "epoch": 0.63, + "grad_norm": 0.4765625, + "learning_rate": 0.000155850189382604, + "loss": 2.4808, + "step": 14683 + }, + { + "epoch": 0.63, + "grad_norm": 0.419921875, + "learning_rate": 0.00015584462500597805, + "loss": 2.5826, + "step": 14684 + }, + { + "epoch": 0.63, + "grad_norm": 0.55078125, + "learning_rate": 0.0001558390603780717, + "loss": 2.7325, + "step": 14685 + }, + { + "epoch": 0.63, + "grad_norm": 0.5, + "learning_rate": 0.00015583349549890997, + "loss": 2.7776, + "step": 14686 + }, + { + "epoch": 0.63, + "grad_norm": 0.455078125, + "learning_rate": 0.00015582793036851794, + "loss": 2.6402, + "step": 14687 + }, + { + "epoch": 0.63, + "grad_norm": 0.5234375, + "learning_rate": 0.0001558223649869206, + "loss": 2.589, + "step": 14688 + }, + { + "epoch": 0.63, + "grad_norm": 0.46875, + "learning_rate": 0.00015581679935414307, + "loss": 2.6737, + "step": 14689 + }, + { + "epoch": 0.63, + "grad_norm": 0.455078125, + "learning_rate": 0.0001558112334702103, + "loss": 2.4463, + "step": 14690 + }, + { + "epoch": 0.63, + "grad_norm": 0.515625, + "learning_rate": 0.0001558056673351474, + "loss": 2.8319, + "step": 14691 + }, + { + "epoch": 0.63, + "grad_norm": 0.45703125, + "learning_rate": 0.00015580010094897934, + "loss": 2.5811, + "step": 14692 + }, + { + "epoch": 0.63, + "grad_norm": 0.462890625, + "learning_rate": 0.00015579453431173126, + "loss": 2.5306, + "step": 14693 + }, + { + "epoch": 0.63, + "grad_norm": 0.45703125, + "learning_rate": 0.00015578896742342814, + "loss": 2.5097, + "step": 14694 + }, + { + "epoch": 0.63, + "grad_norm": 0.470703125, + "learning_rate": 0.0001557834002840951, + "loss": 2.5502, + "step": 14695 + }, + { + "epoch": 0.63, + "grad_norm": 0.439453125, + "learning_rate": 0.0001557778328937571, + "loss": 2.3708, + "step": 14696 + }, + { + "epoch": 0.63, + "grad_norm": 0.490234375, + "learning_rate": 0.00015577226525243925, + "loss": 2.5817, + "step": 14697 + }, + { + "epoch": 0.63, + "grad_norm": 0.458984375, + "learning_rate": 0.00015576669736016657, + "loss": 2.58, + "step": 14698 + }, + { + "epoch": 0.63, + "grad_norm": 0.51953125, + "learning_rate": 0.0001557611292169642, + "loss": 2.5868, + "step": 14699 + }, + { + "epoch": 0.63, + "grad_norm": 0.439453125, + "learning_rate": 0.00015575556082285706, + "loss": 2.6382, + "step": 14700 + }, + { + "epoch": 0.63, + "grad_norm": 0.51171875, + "learning_rate": 0.0001557499921778703, + "loss": 2.6984, + "step": 14701 + }, + { + "epoch": 0.63, + "grad_norm": 0.423828125, + "learning_rate": 0.00015574442328202893, + "loss": 2.5155, + "step": 14702 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.00015573885413535808, + "loss": 2.5847, + "step": 14703 + }, + { + "epoch": 0.63, + "grad_norm": 0.486328125, + "learning_rate": 0.00015573328473788268, + "loss": 2.5906, + "step": 14704 + }, + { + "epoch": 0.63, + "grad_norm": 0.451171875, + "learning_rate": 0.0001557277150896279, + "loss": 2.4769, + "step": 14705 + }, + { + "epoch": 0.63, + "grad_norm": 0.453125, + "learning_rate": 0.0001557221451906188, + "loss": 2.6226, + "step": 14706 + }, + { + "epoch": 0.63, + "grad_norm": 0.46484375, + "learning_rate": 0.00015571657504088036, + "loss": 2.6212, + "step": 14707 + }, + { + "epoch": 0.63, + "grad_norm": 0.455078125, + "learning_rate": 0.00015571100464043772, + "loss": 2.5892, + "step": 14708 + }, + { + "epoch": 0.63, + "grad_norm": 0.439453125, + "learning_rate": 0.00015570543398931593, + "loss": 2.5493, + "step": 14709 + }, + { + "epoch": 0.63, + "grad_norm": 0.44921875, + "learning_rate": 0.00015569986308754, + "loss": 2.542, + "step": 14710 + }, + { + "epoch": 0.63, + "grad_norm": 0.443359375, + "learning_rate": 0.0001556942919351351, + "loss": 2.6463, + "step": 14711 + }, + { + "epoch": 0.63, + "grad_norm": 0.41796875, + "learning_rate": 0.00015568872053212622, + "loss": 2.5128, + "step": 14712 + }, + { + "epoch": 0.63, + "grad_norm": 0.466796875, + "learning_rate": 0.00015568314887853845, + "loss": 2.4638, + "step": 14713 + }, + { + "epoch": 0.63, + "grad_norm": 0.453125, + "learning_rate": 0.00015567757697439684, + "loss": 2.6852, + "step": 14714 + }, + { + "epoch": 0.63, + "grad_norm": 0.470703125, + "learning_rate": 0.00015567200481972652, + "loss": 2.5556, + "step": 14715 + }, + { + "epoch": 0.63, + "grad_norm": 0.431640625, + "learning_rate": 0.00015566643241455248, + "loss": 2.4841, + "step": 14716 + }, + { + "epoch": 0.63, + "grad_norm": 0.482421875, + "learning_rate": 0.00015566085975889985, + "loss": 2.5909, + "step": 14717 + }, + { + "epoch": 0.63, + "grad_norm": 0.400390625, + "learning_rate": 0.0001556552868527937, + "loss": 2.607, + "step": 14718 + }, + { + "epoch": 0.63, + "grad_norm": 0.451171875, + "learning_rate": 0.0001556497136962591, + "loss": 2.5165, + "step": 14719 + }, + { + "epoch": 0.63, + "grad_norm": 0.5, + "learning_rate": 0.00015564414028932112, + "loss": 2.5463, + "step": 14720 + }, + { + "epoch": 0.63, + "grad_norm": 0.453125, + "learning_rate": 0.00015563856663200487, + "loss": 2.6544, + "step": 14721 + }, + { + "epoch": 0.63, + "grad_norm": 0.455078125, + "learning_rate": 0.00015563299272433536, + "loss": 2.741, + "step": 14722 + }, + { + "epoch": 0.63, + "grad_norm": 0.4296875, + "learning_rate": 0.00015562741856633776, + "loss": 2.7091, + "step": 14723 + }, + { + "epoch": 0.63, + "grad_norm": 0.431640625, + "learning_rate": 0.00015562184415803708, + "loss": 2.5649, + "step": 14724 + }, + { + "epoch": 0.63, + "grad_norm": 0.451171875, + "learning_rate": 0.00015561626949945842, + "loss": 2.531, + "step": 14725 + }, + { + "epoch": 0.63, + "grad_norm": 0.484375, + "learning_rate": 0.0001556106945906269, + "loss": 2.5862, + "step": 14726 + }, + { + "epoch": 0.63, + "grad_norm": 0.39453125, + "learning_rate": 0.00015560511943156757, + "loss": 2.5833, + "step": 14727 + }, + { + "epoch": 0.63, + "grad_norm": 0.43359375, + "learning_rate": 0.00015559954402230553, + "loss": 2.5983, + "step": 14728 + }, + { + "epoch": 0.63, + "grad_norm": 0.4921875, + "learning_rate": 0.00015559396836286582, + "loss": 2.7127, + "step": 14729 + }, + { + "epoch": 0.63, + "grad_norm": 0.45703125, + "learning_rate": 0.0001555883924532736, + "loss": 2.6501, + "step": 14730 + }, + { + "epoch": 0.63, + "grad_norm": 0.47265625, + "learning_rate": 0.00015558281629355394, + "loss": 2.6114, + "step": 14731 + }, + { + "epoch": 0.63, + "grad_norm": 0.43359375, + "learning_rate": 0.0001555772398837319, + "loss": 2.6729, + "step": 14732 + }, + { + "epoch": 0.63, + "grad_norm": 0.47265625, + "learning_rate": 0.00015557166322383262, + "loss": 2.6472, + "step": 14733 + }, + { + "epoch": 0.63, + "grad_norm": 0.5078125, + "learning_rate": 0.0001555660863138811, + "loss": 2.6952, + "step": 14734 + }, + { + "epoch": 0.63, + "grad_norm": 0.435546875, + "learning_rate": 0.00015556050915390256, + "loss": 2.5517, + "step": 14735 + }, + { + "epoch": 0.63, + "grad_norm": 0.455078125, + "learning_rate": 0.00015555493174392203, + "loss": 2.4884, + "step": 14736 + }, + { + "epoch": 0.63, + "grad_norm": 0.5390625, + "learning_rate": 0.0001555493540839646, + "loss": 2.4389, + "step": 14737 + }, + { + "epoch": 0.63, + "grad_norm": 0.404296875, + "learning_rate": 0.00015554377617405532, + "loss": 2.5417, + "step": 14738 + }, + { + "epoch": 0.63, + "grad_norm": 0.45703125, + "learning_rate": 0.00015553819801421943, + "loss": 2.6588, + "step": 14739 + }, + { + "epoch": 0.63, + "grad_norm": 0.416015625, + "learning_rate": 0.0001555326196044819, + "loss": 2.5767, + "step": 14740 + }, + { + "epoch": 0.63, + "grad_norm": 0.3984375, + "learning_rate": 0.00015552704094486792, + "loss": 2.6328, + "step": 14741 + }, + { + "epoch": 0.63, + "grad_norm": 0.43359375, + "learning_rate": 0.0001555214620354025, + "loss": 2.4434, + "step": 14742 + }, + { + "epoch": 0.63, + "grad_norm": 0.423828125, + "learning_rate": 0.00015551588287611082, + "loss": 2.3359, + "step": 14743 + }, + { + "epoch": 0.63, + "grad_norm": 0.4375, + "learning_rate": 0.00015551030346701796, + "loss": 2.5703, + "step": 14744 + }, + { + "epoch": 0.63, + "grad_norm": 0.4765625, + "learning_rate": 0.000155504723808149, + "loss": 2.7588, + "step": 14745 + }, + { + "epoch": 0.63, + "grad_norm": 0.48046875, + "learning_rate": 0.00015549914389952906, + "loss": 2.6624, + "step": 14746 + }, + { + "epoch": 0.63, + "grad_norm": 0.46484375, + "learning_rate": 0.0001554935637411833, + "loss": 2.6937, + "step": 14747 + }, + { + "epoch": 0.63, + "grad_norm": 0.439453125, + "learning_rate": 0.00015548798333313676, + "loss": 2.4828, + "step": 14748 + }, + { + "epoch": 0.63, + "grad_norm": 0.421875, + "learning_rate": 0.00015548240267541455, + "loss": 2.5108, + "step": 14749 + }, + { + "epoch": 0.63, + "grad_norm": 0.435546875, + "learning_rate": 0.0001554768217680418, + "loss": 2.7085, + "step": 14750 + }, + { + "epoch": 0.63, + "grad_norm": 0.4609375, + "learning_rate": 0.00015547124061104367, + "loss": 2.5289, + "step": 14751 + }, + { + "epoch": 0.63, + "grad_norm": 0.41796875, + "learning_rate": 0.0001554656592044452, + "loss": 2.5692, + "step": 14752 + }, + { + "epoch": 0.63, + "grad_norm": 0.4765625, + "learning_rate": 0.00015546007754827158, + "loss": 2.6135, + "step": 14753 + }, + { + "epoch": 0.63, + "grad_norm": 0.443359375, + "learning_rate": 0.0001554544956425478, + "loss": 2.5587, + "step": 14754 + }, + { + "epoch": 0.63, + "grad_norm": 0.462890625, + "learning_rate": 0.00015544891348729913, + "loss": 2.6309, + "step": 14755 + }, + { + "epoch": 0.63, + "grad_norm": 0.478515625, + "learning_rate": 0.00015544333108255057, + "loss": 2.6583, + "step": 14756 + }, + { + "epoch": 0.63, + "grad_norm": 0.47265625, + "learning_rate": 0.00015543774842832728, + "loss": 2.6423, + "step": 14757 + }, + { + "epoch": 0.63, + "grad_norm": 0.482421875, + "learning_rate": 0.0001554321655246544, + "loss": 2.5525, + "step": 14758 + }, + { + "epoch": 0.63, + "grad_norm": 0.453125, + "learning_rate": 0.00015542658237155705, + "loss": 2.5527, + "step": 14759 + }, + { + "epoch": 0.63, + "grad_norm": 0.41796875, + "learning_rate": 0.0001554209989690603, + "loss": 2.5223, + "step": 14760 + }, + { + "epoch": 0.63, + "grad_norm": 0.474609375, + "learning_rate": 0.00015541541531718933, + "loss": 2.6303, + "step": 14761 + }, + { + "epoch": 0.63, + "grad_norm": 0.44140625, + "learning_rate": 0.00015540983141596924, + "loss": 2.7142, + "step": 14762 + }, + { + "epoch": 0.63, + "grad_norm": 0.44921875, + "learning_rate": 0.00015540424726542517, + "loss": 2.5813, + "step": 14763 + }, + { + "epoch": 0.63, + "grad_norm": 0.404296875, + "learning_rate": 0.0001553986628655822, + "loss": 2.6316, + "step": 14764 + }, + { + "epoch": 0.63, + "grad_norm": 0.47265625, + "learning_rate": 0.0001553930782164655, + "loss": 2.5491, + "step": 14765 + }, + { + "epoch": 0.63, + "grad_norm": 0.48046875, + "learning_rate": 0.00015538749331810023, + "loss": 2.7005, + "step": 14766 + }, + { + "epoch": 0.63, + "grad_norm": 0.453125, + "learning_rate": 0.00015538190817051146, + "loss": 2.6907, + "step": 14767 + }, + { + "epoch": 0.63, + "grad_norm": 0.51171875, + "learning_rate": 0.00015537632277372434, + "loss": 2.6869, + "step": 14768 + }, + { + "epoch": 0.63, + "grad_norm": 0.546875, + "learning_rate": 0.000155370737127764, + "loss": 2.6523, + "step": 14769 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.00015536515123265558, + "loss": 2.5799, + "step": 14770 + }, + { + "epoch": 0.63, + "grad_norm": 0.451171875, + "learning_rate": 0.00015535956508842423, + "loss": 2.6302, + "step": 14771 + }, + { + "epoch": 0.63, + "grad_norm": 0.4296875, + "learning_rate": 0.00015535397869509505, + "loss": 2.5381, + "step": 14772 + }, + { + "epoch": 0.63, + "grad_norm": 0.47265625, + "learning_rate": 0.0001553483920526932, + "loss": 2.6562, + "step": 14773 + }, + { + "epoch": 0.63, + "grad_norm": 0.515625, + "learning_rate": 0.00015534280516124382, + "loss": 2.6064, + "step": 14774 + }, + { + "epoch": 0.63, + "grad_norm": 0.4765625, + "learning_rate": 0.00015533721802077203, + "loss": 2.5906, + "step": 14775 + }, + { + "epoch": 0.63, + "grad_norm": 0.455078125, + "learning_rate": 0.000155331630631303, + "loss": 2.6771, + "step": 14776 + }, + { + "epoch": 0.63, + "grad_norm": 0.49609375, + "learning_rate": 0.00015532604299286183, + "loss": 2.6992, + "step": 14777 + }, + { + "epoch": 0.63, + "grad_norm": 0.470703125, + "learning_rate": 0.0001553204551054737, + "loss": 2.7061, + "step": 14778 + }, + { + "epoch": 0.63, + "grad_norm": 0.53515625, + "learning_rate": 0.00015531486696916376, + "loss": 2.615, + "step": 14779 + }, + { + "epoch": 0.63, + "grad_norm": 0.404296875, + "learning_rate": 0.0001553092785839571, + "loss": 2.5818, + "step": 14780 + }, + { + "epoch": 0.63, + "grad_norm": 0.51171875, + "learning_rate": 0.0001553036899498789, + "loss": 2.6205, + "step": 14781 + }, + { + "epoch": 0.63, + "grad_norm": 0.4453125, + "learning_rate": 0.00015529810106695434, + "loss": 2.5429, + "step": 14782 + }, + { + "epoch": 0.63, + "grad_norm": 0.470703125, + "learning_rate": 0.00015529251193520852, + "loss": 2.4923, + "step": 14783 + }, + { + "epoch": 0.63, + "grad_norm": 0.546875, + "learning_rate": 0.00015528692255466656, + "loss": 2.7091, + "step": 14784 + }, + { + "epoch": 0.63, + "grad_norm": 0.435546875, + "learning_rate": 0.0001552813329253537, + "loss": 2.5753, + "step": 14785 + }, + { + "epoch": 0.63, + "grad_norm": 0.458984375, + "learning_rate": 0.00015527574304729505, + "loss": 2.5124, + "step": 14786 + }, + { + "epoch": 0.63, + "grad_norm": 0.46484375, + "learning_rate": 0.00015527015292051574, + "loss": 2.6162, + "step": 14787 + }, + { + "epoch": 0.63, + "grad_norm": 0.46875, + "learning_rate": 0.00015526456254504094, + "loss": 2.6792, + "step": 14788 + }, + { + "epoch": 0.63, + "grad_norm": 0.5234375, + "learning_rate": 0.0001552589719208958, + "loss": 2.6208, + "step": 14789 + }, + { + "epoch": 0.63, + "grad_norm": 0.52734375, + "learning_rate": 0.00015525338104810553, + "loss": 2.7595, + "step": 14790 + }, + { + "epoch": 0.63, + "grad_norm": 0.43359375, + "learning_rate": 0.00015524778992669517, + "loss": 2.6434, + "step": 14791 + }, + { + "epoch": 0.63, + "grad_norm": 0.431640625, + "learning_rate": 0.00015524219855668996, + "loss": 2.611, + "step": 14792 + }, + { + "epoch": 0.63, + "grad_norm": 0.486328125, + "learning_rate": 0.00015523660693811506, + "loss": 2.6745, + "step": 14793 + }, + { + "epoch": 0.63, + "grad_norm": 0.470703125, + "learning_rate": 0.00015523101507099562, + "loss": 2.683, + "step": 14794 + }, + { + "epoch": 0.63, + "grad_norm": 0.5234375, + "learning_rate": 0.0001552254229553568, + "loss": 2.6968, + "step": 14795 + }, + { + "epoch": 0.63, + "grad_norm": 0.388671875, + "learning_rate": 0.00015521983059122376, + "loss": 2.6376, + "step": 14796 + }, + { + "epoch": 0.63, + "grad_norm": 0.50390625, + "learning_rate": 0.00015521423797862163, + "loss": 2.7591, + "step": 14797 + }, + { + "epoch": 0.63, + "grad_norm": 0.4609375, + "learning_rate": 0.00015520864511757565, + "loss": 2.6296, + "step": 14798 + }, + { + "epoch": 0.63, + "grad_norm": 0.431640625, + "learning_rate": 0.00015520305200811092, + "loss": 2.5942, + "step": 14799 + }, + { + "epoch": 0.63, + "grad_norm": 0.546875, + "learning_rate": 0.0001551974586502526, + "loss": 2.6319, + "step": 14800 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.00015519186504402594, + "loss": 2.5919, + "step": 14801 + }, + { + "epoch": 0.63, + "grad_norm": 0.416015625, + "learning_rate": 0.00015518627118945607, + "loss": 2.6003, + "step": 14802 + }, + { + "epoch": 0.63, + "grad_norm": 0.55078125, + "learning_rate": 0.0001551806770865681, + "loss": 2.6901, + "step": 14803 + }, + { + "epoch": 0.63, + "grad_norm": 0.435546875, + "learning_rate": 0.00015517508273538727, + "loss": 2.6819, + "step": 14804 + }, + { + "epoch": 0.63, + "grad_norm": 0.412109375, + "learning_rate": 0.0001551694881359387, + "loss": 2.5126, + "step": 14805 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.00015516389328824764, + "loss": 2.5613, + "step": 14806 + }, + { + "epoch": 0.63, + "grad_norm": 0.43359375, + "learning_rate": 0.00015515829819233922, + "loss": 2.5579, + "step": 14807 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.0001551527028482386, + "loss": 2.6485, + "step": 14808 + }, + { + "epoch": 0.63, + "grad_norm": 0.46875, + "learning_rate": 0.00015514710725597096, + "loss": 2.6304, + "step": 14809 + }, + { + "epoch": 0.63, + "grad_norm": 0.43359375, + "learning_rate": 0.00015514151141556154, + "loss": 2.5198, + "step": 14810 + }, + { + "epoch": 0.63, + "grad_norm": 0.45703125, + "learning_rate": 0.00015513591532703542, + "loss": 2.5987, + "step": 14811 + }, + { + "epoch": 0.63, + "grad_norm": 0.48046875, + "learning_rate": 0.00015513031899041786, + "loss": 2.4795, + "step": 14812 + }, + { + "epoch": 0.63, + "grad_norm": 0.421875, + "learning_rate": 0.000155124722405734, + "loss": 2.4558, + "step": 14813 + }, + { + "epoch": 0.63, + "grad_norm": 0.4296875, + "learning_rate": 0.00015511912557300903, + "loss": 2.5621, + "step": 14814 + }, + { + "epoch": 0.63, + "grad_norm": 0.47265625, + "learning_rate": 0.00015511352849226814, + "loss": 2.4705, + "step": 14815 + }, + { + "epoch": 0.63, + "grad_norm": 0.49609375, + "learning_rate": 0.00015510793116353653, + "loss": 2.7696, + "step": 14816 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.00015510233358683934, + "loss": 2.6187, + "step": 14817 + }, + { + "epoch": 0.63, + "grad_norm": 0.462890625, + "learning_rate": 0.0001550967357622018, + "loss": 2.6462, + "step": 14818 + }, + { + "epoch": 0.63, + "grad_norm": 0.4609375, + "learning_rate": 0.0001550911376896491, + "loss": 2.6519, + "step": 14819 + }, + { + "epoch": 0.63, + "grad_norm": 0.486328125, + "learning_rate": 0.00015508553936920637, + "loss": 2.7017, + "step": 14820 + }, + { + "epoch": 0.63, + "grad_norm": 0.515625, + "learning_rate": 0.00015507994080089888, + "loss": 2.5157, + "step": 14821 + }, + { + "epoch": 0.63, + "grad_norm": 0.50390625, + "learning_rate": 0.0001550743419847518, + "loss": 2.6201, + "step": 14822 + }, + { + "epoch": 0.63, + "grad_norm": 0.392578125, + "learning_rate": 0.0001550687429207903, + "loss": 2.4918, + "step": 14823 + }, + { + "epoch": 0.63, + "grad_norm": 0.50390625, + "learning_rate": 0.00015506314360903955, + "loss": 2.683, + "step": 14824 + }, + { + "epoch": 0.63, + "grad_norm": 0.46484375, + "learning_rate": 0.0001550575440495248, + "loss": 2.5682, + "step": 14825 + }, + { + "epoch": 0.63, + "grad_norm": 0.458984375, + "learning_rate": 0.00015505194424227122, + "loss": 2.5913, + "step": 14826 + }, + { + "epoch": 0.63, + "grad_norm": 0.44921875, + "learning_rate": 0.000155046344187304, + "loss": 2.6453, + "step": 14827 + }, + { + "epoch": 0.63, + "grad_norm": 0.48828125, + "learning_rate": 0.00015504074388464838, + "loss": 2.6948, + "step": 14828 + }, + { + "epoch": 0.63, + "grad_norm": 0.4609375, + "learning_rate": 0.00015503514333432952, + "loss": 2.7977, + "step": 14829 + }, + { + "epoch": 0.63, + "grad_norm": 0.3828125, + "learning_rate": 0.00015502954253637262, + "loss": 2.4942, + "step": 14830 + }, + { + "epoch": 0.63, + "grad_norm": 0.47265625, + "learning_rate": 0.0001550239414908029, + "loss": 2.5107, + "step": 14831 + }, + { + "epoch": 0.63, + "grad_norm": 0.458984375, + "learning_rate": 0.0001550183401976455, + "loss": 2.4075, + "step": 14832 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.00015501273865692572, + "loss": 2.6238, + "step": 14833 + }, + { + "epoch": 0.63, + "grad_norm": 0.42578125, + "learning_rate": 0.00015500713686866872, + "loss": 2.5777, + "step": 14834 + }, + { + "epoch": 0.63, + "grad_norm": 0.46875, + "learning_rate": 0.0001550015348328997, + "loss": 2.5519, + "step": 14835 + }, + { + "epoch": 0.63, + "grad_norm": 0.447265625, + "learning_rate": 0.00015499593254964387, + "loss": 2.6386, + "step": 14836 + }, + { + "epoch": 0.63, + "grad_norm": 0.44921875, + "learning_rate": 0.00015499033001892647, + "loss": 2.6042, + "step": 14837 + }, + { + "epoch": 0.63, + "grad_norm": 0.494140625, + "learning_rate": 0.00015498472724077264, + "loss": 2.5785, + "step": 14838 + }, + { + "epoch": 0.63, + "grad_norm": 0.412109375, + "learning_rate": 0.00015497912421520767, + "loss": 2.6441, + "step": 14839 + }, + { + "epoch": 0.63, + "grad_norm": 0.48828125, + "learning_rate": 0.0001549735209422567, + "loss": 2.6012, + "step": 14840 + }, + { + "epoch": 0.63, + "grad_norm": 0.470703125, + "learning_rate": 0.000154967917421945, + "loss": 2.4744, + "step": 14841 + }, + { + "epoch": 0.63, + "grad_norm": 0.4453125, + "learning_rate": 0.00015496231365429774, + "loss": 2.6338, + "step": 14842 + }, + { + "epoch": 0.63, + "grad_norm": 0.41015625, + "learning_rate": 0.00015495670963934017, + "loss": 2.5106, + "step": 14843 + }, + { + "epoch": 0.63, + "grad_norm": 0.486328125, + "learning_rate": 0.0001549511053770975, + "loss": 2.648, + "step": 14844 + }, + { + "epoch": 0.63, + "grad_norm": 0.435546875, + "learning_rate": 0.0001549455008675949, + "loss": 2.6948, + "step": 14845 + }, + { + "epoch": 0.63, + "grad_norm": 0.458984375, + "learning_rate": 0.00015493989611085768, + "loss": 2.6367, + "step": 14846 + }, + { + "epoch": 0.63, + "grad_norm": 0.4921875, + "learning_rate": 0.00015493429110691097, + "loss": 2.5476, + "step": 14847 + }, + { + "epoch": 0.63, + "grad_norm": 0.494140625, + "learning_rate": 0.00015492868585578005, + "loss": 2.6477, + "step": 14848 + }, + { + "epoch": 0.63, + "grad_norm": 0.51953125, + "learning_rate": 0.0001549230803574901, + "loss": 2.5862, + "step": 14849 + }, + { + "epoch": 0.63, + "grad_norm": 0.451171875, + "learning_rate": 0.00015491747461206636, + "loss": 2.5511, + "step": 14850 + }, + { + "epoch": 0.63, + "grad_norm": 0.443359375, + "learning_rate": 0.00015491186861953403, + "loss": 2.4434, + "step": 14851 + }, + { + "epoch": 0.63, + "grad_norm": 0.484375, + "learning_rate": 0.00015490626237991842, + "loss": 2.5418, + "step": 14852 + }, + { + "epoch": 0.63, + "grad_norm": 0.494140625, + "learning_rate": 0.00015490065589324466, + "loss": 2.5871, + "step": 14853 + }, + { + "epoch": 0.63, + "grad_norm": 0.453125, + "learning_rate": 0.000154895049159538, + "loss": 2.5115, + "step": 14854 + }, + { + "epoch": 0.63, + "grad_norm": 0.392578125, + "learning_rate": 0.00015488944217882368, + "loss": 2.5892, + "step": 14855 + }, + { + "epoch": 0.63, + "grad_norm": 0.443359375, + "learning_rate": 0.00015488383495112698, + "loss": 2.4425, + "step": 14856 + }, + { + "epoch": 0.63, + "grad_norm": 0.498046875, + "learning_rate": 0.00015487822747647302, + "loss": 2.6703, + "step": 14857 + }, + { + "epoch": 0.63, + "grad_norm": 0.4453125, + "learning_rate": 0.00015487261975488713, + "loss": 2.5744, + "step": 14858 + }, + { + "epoch": 0.63, + "grad_norm": 0.470703125, + "learning_rate": 0.0001548670117863945, + "loss": 2.5698, + "step": 14859 + }, + { + "epoch": 0.63, + "grad_norm": 0.44140625, + "learning_rate": 0.00015486140357102036, + "loss": 2.5259, + "step": 14860 + }, + { + "epoch": 0.63, + "grad_norm": 0.427734375, + "learning_rate": 0.00015485579510878995, + "loss": 2.5545, + "step": 14861 + }, + { + "epoch": 0.63, + "grad_norm": 0.4609375, + "learning_rate": 0.0001548501863997285, + "loss": 2.6315, + "step": 14862 + }, + { + "epoch": 0.63, + "grad_norm": 0.474609375, + "learning_rate": 0.00015484457744386124, + "loss": 2.5565, + "step": 14863 + }, + { + "epoch": 0.63, + "grad_norm": 0.458984375, + "learning_rate": 0.00015483896824121347, + "loss": 2.7894, + "step": 14864 + }, + { + "epoch": 0.63, + "grad_norm": 0.4921875, + "learning_rate": 0.00015483335879181036, + "loss": 2.5866, + "step": 14865 + }, + { + "epoch": 0.63, + "grad_norm": 0.47265625, + "learning_rate": 0.0001548277490956772, + "loss": 2.542, + "step": 14866 + }, + { + "epoch": 0.63, + "grad_norm": 0.494140625, + "learning_rate": 0.0001548221391528392, + "loss": 2.5893, + "step": 14867 + }, + { + "epoch": 0.63, + "grad_norm": 0.45703125, + "learning_rate": 0.0001548165289633216, + "loss": 2.6958, + "step": 14868 + }, + { + "epoch": 0.63, + "grad_norm": 0.4921875, + "learning_rate": 0.00015481091852714965, + "loss": 2.6153, + "step": 14869 + }, + { + "epoch": 0.63, + "grad_norm": 0.45703125, + "learning_rate": 0.00015480530784434858, + "loss": 2.5413, + "step": 14870 + }, + { + "epoch": 0.63, + "grad_norm": 0.439453125, + "learning_rate": 0.00015479969691494368, + "loss": 2.4973, + "step": 14871 + }, + { + "epoch": 0.63, + "grad_norm": 0.419921875, + "learning_rate": 0.00015479408573896016, + "loss": 2.546, + "step": 14872 + }, + { + "epoch": 0.63, + "grad_norm": 0.451171875, + "learning_rate": 0.0001547884743164233, + "loss": 2.6515, + "step": 14873 + }, + { + "epoch": 0.63, + "grad_norm": 0.470703125, + "learning_rate": 0.00015478286264735831, + "loss": 2.6388, + "step": 14874 + }, + { + "epoch": 0.63, + "grad_norm": 0.435546875, + "learning_rate": 0.00015477725073179046, + "loss": 2.6236, + "step": 14875 + }, + { + "epoch": 0.63, + "grad_norm": 0.400390625, + "learning_rate": 0.000154771638569745, + "loss": 2.5611, + "step": 14876 + }, + { + "epoch": 0.63, + "grad_norm": 0.486328125, + "learning_rate": 0.0001547660261612472, + "loss": 2.5948, + "step": 14877 + }, + { + "epoch": 0.63, + "grad_norm": 0.453125, + "learning_rate": 0.00015476041350632227, + "loss": 2.5148, + "step": 14878 + }, + { + "epoch": 0.63, + "grad_norm": 0.57421875, + "learning_rate": 0.0001547548006049955, + "loss": 2.8326, + "step": 14879 + }, + { + "epoch": 0.63, + "grad_norm": 0.52734375, + "learning_rate": 0.00015474918745729215, + "loss": 2.743, + "step": 14880 + }, + { + "epoch": 0.63, + "grad_norm": 0.439453125, + "learning_rate": 0.00015474357406323742, + "loss": 2.4261, + "step": 14881 + }, + { + "epoch": 0.63, + "grad_norm": 0.46875, + "learning_rate": 0.00015473796042285667, + "loss": 2.5634, + "step": 14882 + }, + { + "epoch": 0.63, + "grad_norm": 0.44140625, + "learning_rate": 0.0001547323465361751, + "loss": 2.5344, + "step": 14883 + }, + { + "epoch": 0.63, + "grad_norm": 0.416015625, + "learning_rate": 0.00015472673240321796, + "loss": 2.4265, + "step": 14884 + }, + { + "epoch": 0.63, + "grad_norm": 0.486328125, + "learning_rate": 0.0001547211180240105, + "loss": 2.7228, + "step": 14885 + }, + { + "epoch": 0.63, + "grad_norm": 0.392578125, + "learning_rate": 0.00015471550339857804, + "loss": 2.5134, + "step": 14886 + }, + { + "epoch": 0.63, + "grad_norm": 0.42578125, + "learning_rate": 0.00015470988852694576, + "loss": 2.6115, + "step": 14887 + }, + { + "epoch": 0.63, + "grad_norm": 0.45703125, + "learning_rate": 0.00015470427340913903, + "loss": 2.4309, + "step": 14888 + }, + { + "epoch": 0.63, + "grad_norm": 0.40625, + "learning_rate": 0.00015469865804518303, + "loss": 2.5716, + "step": 14889 + }, + { + "epoch": 0.63, + "grad_norm": 0.482421875, + "learning_rate": 0.00015469304243510308, + "loss": 2.7309, + "step": 14890 + }, + { + "epoch": 0.63, + "grad_norm": 0.455078125, + "learning_rate": 0.00015468742657892442, + "loss": 2.5631, + "step": 14891 + }, + { + "epoch": 0.63, + "grad_norm": 0.4453125, + "learning_rate": 0.00015468181047667234, + "loss": 2.6272, + "step": 14892 + }, + { + "epoch": 0.63, + "grad_norm": 0.421875, + "learning_rate": 0.0001546761941283721, + "loss": 2.4444, + "step": 14893 + }, + { + "epoch": 0.63, + "grad_norm": 0.4375, + "learning_rate": 0.00015467057753404894, + "loss": 2.4758, + "step": 14894 + }, + { + "epoch": 0.63, + "grad_norm": 0.419921875, + "learning_rate": 0.00015466496069372818, + "loss": 2.6297, + "step": 14895 + }, + { + "epoch": 0.63, + "grad_norm": 0.4921875, + "learning_rate": 0.00015465934360743506, + "loss": 2.4839, + "step": 14896 + }, + { + "epoch": 0.63, + "grad_norm": 0.451171875, + "learning_rate": 0.0001546537262751949, + "loss": 2.6178, + "step": 14897 + }, + { + "epoch": 0.63, + "grad_norm": 0.474609375, + "learning_rate": 0.0001546481086970329, + "loss": 2.5485, + "step": 14898 + }, + { + "epoch": 0.63, + "grad_norm": 0.5234375, + "learning_rate": 0.0001546424908729744, + "loss": 2.7212, + "step": 14899 + }, + { + "epoch": 0.63, + "grad_norm": 0.46875, + "learning_rate": 0.0001546368728030447, + "loss": 2.6076, + "step": 14900 + }, + { + "epoch": 0.63, + "grad_norm": 0.46484375, + "learning_rate": 0.00015463125448726895, + "loss": 2.4759, + "step": 14901 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.0001546256359256726, + "loss": 2.5287, + "step": 14902 + }, + { + "epoch": 0.64, + "grad_norm": 0.494140625, + "learning_rate": 0.0001546200171182808, + "loss": 2.6141, + "step": 14903 + }, + { + "epoch": 0.64, + "grad_norm": 0.44921875, + "learning_rate": 0.00015461439806511895, + "loss": 2.5161, + "step": 14904 + }, + { + "epoch": 0.64, + "grad_norm": 0.4765625, + "learning_rate": 0.00015460877876621222, + "loss": 2.7213, + "step": 14905 + }, + { + "epoch": 0.64, + "grad_norm": 0.48046875, + "learning_rate": 0.00015460315922158596, + "loss": 2.6281, + "step": 14906 + }, + { + "epoch": 0.64, + "grad_norm": 0.4296875, + "learning_rate": 0.00015459753943126544, + "loss": 2.6213, + "step": 14907 + }, + { + "epoch": 0.64, + "grad_norm": 0.4375, + "learning_rate": 0.00015459191939527593, + "loss": 2.5098, + "step": 14908 + }, + { + "epoch": 0.64, + "grad_norm": 0.43359375, + "learning_rate": 0.00015458629911364275, + "loss": 2.5966, + "step": 14909 + }, + { + "epoch": 0.64, + "grad_norm": 0.51171875, + "learning_rate": 0.00015458067858639115, + "loss": 2.6651, + "step": 14910 + }, + { + "epoch": 0.64, + "grad_norm": 0.470703125, + "learning_rate": 0.00015457505781354646, + "loss": 2.5991, + "step": 14911 + }, + { + "epoch": 0.64, + "grad_norm": 0.3984375, + "learning_rate": 0.00015456943679513393, + "loss": 2.6404, + "step": 14912 + }, + { + "epoch": 0.64, + "grad_norm": 0.51953125, + "learning_rate": 0.00015456381553117894, + "loss": 2.5601, + "step": 14913 + }, + { + "epoch": 0.64, + "grad_norm": 0.4921875, + "learning_rate": 0.00015455819402170667, + "loss": 2.5994, + "step": 14914 + }, + { + "epoch": 0.64, + "grad_norm": 0.4765625, + "learning_rate": 0.00015455257226674248, + "loss": 2.5342, + "step": 14915 + }, + { + "epoch": 0.64, + "grad_norm": 0.498046875, + "learning_rate": 0.00015454695026631164, + "loss": 2.622, + "step": 14916 + }, + { + "epoch": 0.64, + "grad_norm": 0.40625, + "learning_rate": 0.0001545413280204395, + "loss": 2.6419, + "step": 14917 + }, + { + "epoch": 0.64, + "grad_norm": 0.4296875, + "learning_rate": 0.00015453570552915124, + "loss": 2.5985, + "step": 14918 + }, + { + "epoch": 0.64, + "grad_norm": 0.455078125, + "learning_rate": 0.00015453008279247229, + "loss": 2.655, + "step": 14919 + }, + { + "epoch": 0.64, + "grad_norm": 0.44921875, + "learning_rate": 0.00015452445981042786, + "loss": 2.6459, + "step": 14920 + }, + { + "epoch": 0.64, + "grad_norm": 0.51953125, + "learning_rate": 0.00015451883658304335, + "loss": 2.6148, + "step": 14921 + }, + { + "epoch": 0.64, + "grad_norm": 0.421875, + "learning_rate": 0.00015451321311034394, + "loss": 2.5484, + "step": 14922 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.000154507589392355, + "loss": 2.5866, + "step": 14923 + }, + { + "epoch": 0.64, + "grad_norm": 0.416015625, + "learning_rate": 0.00015450196542910185, + "loss": 2.5692, + "step": 14924 + }, + { + "epoch": 0.64, + "grad_norm": 0.44921875, + "learning_rate": 0.0001544963412206098, + "loss": 2.6657, + "step": 14925 + }, + { + "epoch": 0.64, + "grad_norm": 0.44140625, + "learning_rate": 0.00015449071676690404, + "loss": 2.5041, + "step": 14926 + }, + { + "epoch": 0.64, + "grad_norm": 0.384765625, + "learning_rate": 0.00015448509206801006, + "loss": 2.6122, + "step": 14927 + }, + { + "epoch": 0.64, + "grad_norm": 0.50390625, + "learning_rate": 0.00015447946712395303, + "loss": 2.5374, + "step": 14928 + }, + { + "epoch": 0.64, + "grad_norm": 0.474609375, + "learning_rate": 0.00015447384193475834, + "loss": 2.6274, + "step": 14929 + }, + { + "epoch": 0.64, + "grad_norm": 0.443359375, + "learning_rate": 0.00015446821650045122, + "loss": 2.5943, + "step": 14930 + }, + { + "epoch": 0.64, + "grad_norm": 0.458984375, + "learning_rate": 0.00015446259082105708, + "loss": 2.5775, + "step": 14931 + }, + { + "epoch": 0.64, + "grad_norm": 0.44921875, + "learning_rate": 0.00015445696489660115, + "loss": 2.593, + "step": 14932 + }, + { + "epoch": 0.64, + "grad_norm": 0.482421875, + "learning_rate": 0.0001544513387271088, + "loss": 2.5434, + "step": 14933 + }, + { + "epoch": 0.64, + "grad_norm": 0.466796875, + "learning_rate": 0.00015444571231260532, + "loss": 2.5146, + "step": 14934 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.00015444008565311605, + "loss": 2.4936, + "step": 14935 + }, + { + "epoch": 0.64, + "grad_norm": 0.515625, + "learning_rate": 0.00015443445874866626, + "loss": 2.6455, + "step": 14936 + }, + { + "epoch": 0.64, + "grad_norm": 0.443359375, + "learning_rate": 0.00015442883159928132, + "loss": 2.6696, + "step": 14937 + }, + { + "epoch": 0.64, + "grad_norm": 0.412109375, + "learning_rate": 0.0001544232042049865, + "loss": 2.495, + "step": 14938 + }, + { + "epoch": 0.64, + "grad_norm": 0.4296875, + "learning_rate": 0.0001544175765658072, + "loss": 2.6615, + "step": 14939 + }, + { + "epoch": 0.64, + "grad_norm": 0.486328125, + "learning_rate": 0.00015441194868176864, + "loss": 2.6227, + "step": 14940 + }, + { + "epoch": 0.64, + "grad_norm": 0.4765625, + "learning_rate": 0.00015440632055289624, + "loss": 2.601, + "step": 14941 + }, + { + "epoch": 0.64, + "grad_norm": 0.55859375, + "learning_rate": 0.00015440069217921524, + "loss": 2.4833, + "step": 14942 + }, + { + "epoch": 0.64, + "grad_norm": 0.466796875, + "learning_rate": 0.00015439506356075105, + "loss": 2.6724, + "step": 14943 + }, + { + "epoch": 0.64, + "grad_norm": 0.41015625, + "learning_rate": 0.00015438943469752892, + "loss": 2.5131, + "step": 14944 + }, + { + "epoch": 0.64, + "grad_norm": 0.431640625, + "learning_rate": 0.00015438380558957423, + "loss": 2.635, + "step": 14945 + }, + { + "epoch": 0.64, + "grad_norm": 0.50390625, + "learning_rate": 0.0001543781762369123, + "loss": 2.6638, + "step": 14946 + }, + { + "epoch": 0.64, + "grad_norm": 0.50390625, + "learning_rate": 0.00015437254663956841, + "loss": 2.736, + "step": 14947 + }, + { + "epoch": 0.64, + "grad_norm": 0.451171875, + "learning_rate": 0.00015436691679756795, + "loss": 2.6226, + "step": 14948 + }, + { + "epoch": 0.64, + "grad_norm": 0.42578125, + "learning_rate": 0.00015436128671093626, + "loss": 2.5526, + "step": 14949 + }, + { + "epoch": 0.64, + "grad_norm": 0.5078125, + "learning_rate": 0.00015435565637969862, + "loss": 2.6378, + "step": 14950 + }, + { + "epoch": 0.64, + "grad_norm": 0.390625, + "learning_rate": 0.0001543500258038804, + "loss": 2.5004, + "step": 14951 + }, + { + "epoch": 0.64, + "grad_norm": 0.453125, + "learning_rate": 0.00015434439498350692, + "loss": 2.498, + "step": 14952 + }, + { + "epoch": 0.64, + "grad_norm": 0.470703125, + "learning_rate": 0.00015433876391860354, + "loss": 2.6117, + "step": 14953 + }, + { + "epoch": 0.64, + "grad_norm": 0.4140625, + "learning_rate": 0.00015433313260919556, + "loss": 2.4363, + "step": 14954 + }, + { + "epoch": 0.64, + "grad_norm": 0.423828125, + "learning_rate": 0.00015432750105530833, + "loss": 2.435, + "step": 14955 + }, + { + "epoch": 0.64, + "grad_norm": 0.431640625, + "learning_rate": 0.0001543218692569672, + "loss": 2.5149, + "step": 14956 + }, + { + "epoch": 0.64, + "grad_norm": 0.5234375, + "learning_rate": 0.00015431623721419755, + "loss": 2.6579, + "step": 14957 + }, + { + "epoch": 0.64, + "grad_norm": 0.515625, + "learning_rate": 0.00015431060492702466, + "loss": 2.6134, + "step": 14958 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.00015430497239547387, + "loss": 2.5789, + "step": 14959 + }, + { + "epoch": 0.64, + "grad_norm": 0.51171875, + "learning_rate": 0.00015429933961957058, + "loss": 2.6577, + "step": 14960 + }, + { + "epoch": 0.64, + "grad_norm": 0.48046875, + "learning_rate": 0.0001542937065993401, + "loss": 2.515, + "step": 14961 + }, + { + "epoch": 0.64, + "grad_norm": 0.47265625, + "learning_rate": 0.00015428807333480777, + "loss": 2.5317, + "step": 14962 + }, + { + "epoch": 0.64, + "grad_norm": 0.4140625, + "learning_rate": 0.000154282439825999, + "loss": 2.6259, + "step": 14963 + }, + { + "epoch": 0.64, + "grad_norm": 0.416015625, + "learning_rate": 0.000154276806072939, + "loss": 2.505, + "step": 14964 + }, + { + "epoch": 0.64, + "grad_norm": 0.4765625, + "learning_rate": 0.0001542711720756533, + "loss": 2.6061, + "step": 14965 + }, + { + "epoch": 0.64, + "grad_norm": 0.384765625, + "learning_rate": 0.0001542655378341671, + "loss": 2.522, + "step": 14966 + }, + { + "epoch": 0.64, + "grad_norm": 0.443359375, + "learning_rate": 0.00015425990334850585, + "loss": 2.5544, + "step": 14967 + }, + { + "epoch": 0.64, + "grad_norm": 0.455078125, + "learning_rate": 0.00015425426861869482, + "loss": 2.6286, + "step": 14968 + }, + { + "epoch": 0.64, + "grad_norm": 0.490234375, + "learning_rate": 0.00015424863364475944, + "loss": 2.6661, + "step": 14969 + }, + { + "epoch": 0.64, + "grad_norm": 0.455078125, + "learning_rate": 0.00015424299842672502, + "loss": 2.4956, + "step": 14970 + }, + { + "epoch": 0.64, + "grad_norm": 0.45703125, + "learning_rate": 0.00015423736296461696, + "loss": 2.8197, + "step": 14971 + }, + { + "epoch": 0.64, + "grad_norm": 0.45703125, + "learning_rate": 0.00015423172725846057, + "loss": 2.5315, + "step": 14972 + }, + { + "epoch": 0.64, + "grad_norm": 0.451171875, + "learning_rate": 0.00015422609130828123, + "loss": 2.5531, + "step": 14973 + }, + { + "epoch": 0.64, + "grad_norm": 0.51171875, + "learning_rate": 0.0001542204551141043, + "loss": 2.665, + "step": 14974 + }, + { + "epoch": 0.64, + "grad_norm": 0.466796875, + "learning_rate": 0.00015421481867595514, + "loss": 2.6159, + "step": 14975 + }, + { + "epoch": 0.64, + "grad_norm": 0.45703125, + "learning_rate": 0.00015420918199385907, + "loss": 2.6923, + "step": 14976 + }, + { + "epoch": 0.64, + "grad_norm": 0.41015625, + "learning_rate": 0.00015420354506784153, + "loss": 2.4996, + "step": 14977 + }, + { + "epoch": 0.64, + "grad_norm": 0.4296875, + "learning_rate": 0.00015419790789792783, + "loss": 2.5387, + "step": 14978 + }, + { + "epoch": 0.64, + "grad_norm": 0.41796875, + "learning_rate": 0.00015419227048414337, + "loss": 2.6213, + "step": 14979 + }, + { + "epoch": 0.64, + "grad_norm": 0.5234375, + "learning_rate": 0.0001541866328265135, + "loss": 2.7219, + "step": 14980 + }, + { + "epoch": 0.64, + "grad_norm": 0.478515625, + "learning_rate": 0.00015418099492506355, + "loss": 2.5633, + "step": 14981 + }, + { + "epoch": 0.64, + "grad_norm": 0.455078125, + "learning_rate": 0.00015417535677981893, + "loss": 2.6373, + "step": 14982 + }, + { + "epoch": 0.64, + "grad_norm": 0.51171875, + "learning_rate": 0.00015416971839080502, + "loss": 2.618, + "step": 14983 + }, + { + "epoch": 0.64, + "grad_norm": 0.41796875, + "learning_rate": 0.00015416407975804715, + "loss": 2.5074, + "step": 14984 + }, + { + "epoch": 0.64, + "grad_norm": 0.44140625, + "learning_rate": 0.00015415844088157078, + "loss": 2.5376, + "step": 14985 + }, + { + "epoch": 0.64, + "grad_norm": 0.455078125, + "learning_rate": 0.00015415280176140118, + "loss": 2.6836, + "step": 14986 + }, + { + "epoch": 0.64, + "grad_norm": 0.46875, + "learning_rate": 0.00015414716239756375, + "loss": 2.5502, + "step": 14987 + }, + { + "epoch": 0.64, + "grad_norm": 0.5078125, + "learning_rate": 0.0001541415227900839, + "loss": 2.5262, + "step": 14988 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.00015413588293898697, + "loss": 2.5886, + "step": 14989 + }, + { + "epoch": 0.64, + "grad_norm": 0.419921875, + "learning_rate": 0.00015413024284429834, + "loss": 2.5763, + "step": 14990 + }, + { + "epoch": 0.64, + "grad_norm": 0.494140625, + "learning_rate": 0.0001541246025060434, + "loss": 2.6962, + "step": 14991 + }, + { + "epoch": 0.64, + "grad_norm": 0.46484375, + "learning_rate": 0.00015411896192424752, + "loss": 2.5417, + "step": 14992 + }, + { + "epoch": 0.64, + "grad_norm": 0.439453125, + "learning_rate": 0.00015411332109893614, + "loss": 2.5588, + "step": 14993 + }, + { + "epoch": 0.64, + "grad_norm": 0.53515625, + "learning_rate": 0.00015410768003013456, + "loss": 2.6446, + "step": 14994 + }, + { + "epoch": 0.64, + "grad_norm": 0.494140625, + "learning_rate": 0.00015410203871786822, + "loss": 2.53, + "step": 14995 + }, + { + "epoch": 0.64, + "grad_norm": 0.421875, + "learning_rate": 0.00015409639716216242, + "loss": 2.4573, + "step": 14996 + }, + { + "epoch": 0.64, + "grad_norm": 0.50390625, + "learning_rate": 0.00015409075536304263, + "loss": 2.761, + "step": 14997 + }, + { + "epoch": 0.64, + "grad_norm": 0.439453125, + "learning_rate": 0.00015408511332053423, + "loss": 2.5474, + "step": 14998 + }, + { + "epoch": 0.64, + "grad_norm": 0.41796875, + "learning_rate": 0.00015407947103466255, + "loss": 2.5331, + "step": 14999 + }, + { + "epoch": 0.64, + "grad_norm": 0.4921875, + "learning_rate": 0.00015407382850545303, + "loss": 2.5261, + "step": 15000 + }, + { + "epoch": 0.64, + "grad_norm": 0.44140625, + "learning_rate": 0.00015406818573293106, + "loss": 2.5876, + "step": 15001 + }, + { + "epoch": 0.64, + "grad_norm": 0.47265625, + "learning_rate": 0.000154062542717122, + "loss": 2.5947, + "step": 15002 + }, + { + "epoch": 0.64, + "grad_norm": 0.5234375, + "learning_rate": 0.00015405689945805127, + "loss": 2.6251, + "step": 15003 + }, + { + "epoch": 0.64, + "grad_norm": 0.451171875, + "learning_rate": 0.00015405125595574422, + "loss": 2.6951, + "step": 15004 + }, + { + "epoch": 0.64, + "grad_norm": 0.470703125, + "learning_rate": 0.0001540456122102263, + "loss": 2.6913, + "step": 15005 + }, + { + "epoch": 0.64, + "grad_norm": 0.5859375, + "learning_rate": 0.00015403996822152286, + "loss": 2.5571, + "step": 15006 + }, + { + "epoch": 0.64, + "grad_norm": 0.515625, + "learning_rate": 0.00015403432398965934, + "loss": 2.5931, + "step": 15007 + }, + { + "epoch": 0.64, + "grad_norm": 0.45703125, + "learning_rate": 0.00015402867951466105, + "loss": 2.6983, + "step": 15008 + }, + { + "epoch": 0.64, + "grad_norm": 0.51171875, + "learning_rate": 0.00015402303479655348, + "loss": 2.5575, + "step": 15009 + }, + { + "epoch": 0.64, + "grad_norm": 0.51171875, + "learning_rate": 0.00015401738983536202, + "loss": 2.5735, + "step": 15010 + }, + { + "epoch": 0.64, + "grad_norm": 0.484375, + "learning_rate": 0.00015401174463111202, + "loss": 2.5471, + "step": 15011 + }, + { + "epoch": 0.64, + "grad_norm": 0.3984375, + "learning_rate": 0.00015400609918382892, + "loss": 2.4325, + "step": 15012 + }, + { + "epoch": 0.64, + "grad_norm": 0.46484375, + "learning_rate": 0.0001540004534935381, + "loss": 2.7103, + "step": 15013 + }, + { + "epoch": 0.64, + "grad_norm": 0.462890625, + "learning_rate": 0.000153994807560265, + "loss": 2.4372, + "step": 15014 + }, + { + "epoch": 0.64, + "grad_norm": 0.4765625, + "learning_rate": 0.00015398916138403495, + "loss": 2.5524, + "step": 15015 + }, + { + "epoch": 0.64, + "grad_norm": 0.478515625, + "learning_rate": 0.00015398351496487343, + "loss": 2.6578, + "step": 15016 + }, + { + "epoch": 0.64, + "grad_norm": 0.5078125, + "learning_rate": 0.00015397786830280582, + "loss": 2.6522, + "step": 15017 + }, + { + "epoch": 0.64, + "grad_norm": 0.44921875, + "learning_rate": 0.00015397222139785754, + "loss": 2.5937, + "step": 15018 + }, + { + "epoch": 0.64, + "grad_norm": 0.431640625, + "learning_rate": 0.00015396657425005399, + "loss": 2.5269, + "step": 15019 + }, + { + "epoch": 0.64, + "grad_norm": 0.453125, + "learning_rate": 0.00015396092685942056, + "loss": 2.5911, + "step": 15020 + }, + { + "epoch": 0.64, + "grad_norm": 0.435546875, + "learning_rate": 0.0001539552792259827, + "loss": 2.6432, + "step": 15021 + }, + { + "epoch": 0.64, + "grad_norm": 0.4453125, + "learning_rate": 0.0001539496313497658, + "loss": 2.517, + "step": 15022 + }, + { + "epoch": 0.64, + "grad_norm": 0.458984375, + "learning_rate": 0.00015394398323079526, + "loss": 2.4622, + "step": 15023 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.0001539383348690965, + "loss": 2.594, + "step": 15024 + }, + { + "epoch": 0.64, + "grad_norm": 0.484375, + "learning_rate": 0.000153932686264695, + "loss": 2.6032, + "step": 15025 + }, + { + "epoch": 0.64, + "grad_norm": 0.482421875, + "learning_rate": 0.00015392703741761606, + "loss": 2.742, + "step": 15026 + }, + { + "epoch": 0.64, + "grad_norm": 0.46875, + "learning_rate": 0.00015392138832788515, + "loss": 2.5003, + "step": 15027 + }, + { + "epoch": 0.64, + "grad_norm": 0.50390625, + "learning_rate": 0.00015391573899552777, + "loss": 2.7216, + "step": 15028 + }, + { + "epoch": 0.64, + "grad_norm": 0.474609375, + "learning_rate": 0.00015391008942056922, + "loss": 2.5814, + "step": 15029 + }, + { + "epoch": 0.64, + "grad_norm": 0.47265625, + "learning_rate": 0.00015390443960303497, + "loss": 2.6598, + "step": 15030 + }, + { + "epoch": 0.64, + "grad_norm": 0.431640625, + "learning_rate": 0.00015389878954295044, + "loss": 2.5664, + "step": 15031 + }, + { + "epoch": 0.64, + "grad_norm": 0.466796875, + "learning_rate": 0.0001538931392403411, + "loss": 2.5705, + "step": 15032 + }, + { + "epoch": 0.64, + "grad_norm": 0.484375, + "learning_rate": 0.0001538874886952323, + "loss": 2.6159, + "step": 15033 + }, + { + "epoch": 0.64, + "grad_norm": 0.47265625, + "learning_rate": 0.00015388183790764946, + "loss": 2.7524, + "step": 15034 + }, + { + "epoch": 0.64, + "grad_norm": 0.453125, + "learning_rate": 0.0001538761868776181, + "loss": 2.6813, + "step": 15035 + }, + { + "epoch": 0.64, + "grad_norm": 0.478515625, + "learning_rate": 0.00015387053560516356, + "loss": 2.5539, + "step": 15036 + }, + { + "epoch": 0.64, + "grad_norm": 0.46484375, + "learning_rate": 0.00015386488409031127, + "loss": 2.4619, + "step": 15037 + }, + { + "epoch": 0.64, + "grad_norm": 0.462890625, + "learning_rate": 0.00015385923233308672, + "loss": 2.4273, + "step": 15038 + }, + { + "epoch": 0.64, + "grad_norm": 0.5078125, + "learning_rate": 0.0001538535803335153, + "loss": 2.7743, + "step": 15039 + }, + { + "epoch": 0.64, + "grad_norm": 0.412109375, + "learning_rate": 0.00015384792809162244, + "loss": 2.5199, + "step": 15040 + }, + { + "epoch": 0.64, + "grad_norm": 0.478515625, + "learning_rate": 0.0001538422756074336, + "loss": 2.7453, + "step": 15041 + }, + { + "epoch": 0.64, + "grad_norm": 0.365234375, + "learning_rate": 0.0001538366228809742, + "loss": 2.4689, + "step": 15042 + }, + { + "epoch": 0.64, + "grad_norm": 0.45703125, + "learning_rate": 0.00015383096991226965, + "loss": 2.5728, + "step": 15043 + }, + { + "epoch": 0.64, + "grad_norm": 0.443359375, + "learning_rate": 0.00015382531670134544, + "loss": 2.703, + "step": 15044 + }, + { + "epoch": 0.64, + "grad_norm": 0.421875, + "learning_rate": 0.00015381966324822695, + "loss": 2.3994, + "step": 15045 + }, + { + "epoch": 0.64, + "grad_norm": 0.53515625, + "learning_rate": 0.00015381400955293963, + "loss": 2.6297, + "step": 15046 + }, + { + "epoch": 0.64, + "grad_norm": 0.5234375, + "learning_rate": 0.00015380835561550898, + "loss": 2.7472, + "step": 15047 + }, + { + "epoch": 0.64, + "grad_norm": 0.458984375, + "learning_rate": 0.00015380270143596037, + "loss": 2.5238, + "step": 15048 + }, + { + "epoch": 0.64, + "grad_norm": 0.44921875, + "learning_rate": 0.00015379704701431925, + "loss": 2.6043, + "step": 15049 + }, + { + "epoch": 0.64, + "grad_norm": 0.404296875, + "learning_rate": 0.0001537913923506111, + "loss": 2.487, + "step": 15050 + }, + { + "epoch": 0.64, + "grad_norm": 0.41796875, + "learning_rate": 0.00015378573744486134, + "loss": 2.4881, + "step": 15051 + }, + { + "epoch": 0.64, + "grad_norm": 0.53125, + "learning_rate": 0.00015378008229709544, + "loss": 2.6944, + "step": 15052 + }, + { + "epoch": 0.64, + "grad_norm": 0.490234375, + "learning_rate": 0.00015377442690733877, + "loss": 2.6864, + "step": 15053 + }, + { + "epoch": 0.64, + "grad_norm": 0.451171875, + "learning_rate": 0.00015376877127561687, + "loss": 2.5406, + "step": 15054 + }, + { + "epoch": 0.64, + "grad_norm": 0.498046875, + "learning_rate": 0.00015376311540195512, + "loss": 2.5764, + "step": 15055 + }, + { + "epoch": 0.64, + "grad_norm": 0.490234375, + "learning_rate": 0.00015375745928637903, + "loss": 2.7343, + "step": 15056 + }, + { + "epoch": 0.64, + "grad_norm": 0.408203125, + "learning_rate": 0.000153751802928914, + "loss": 2.5636, + "step": 15057 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.00015374614632958554, + "loss": 2.5784, + "step": 15058 + }, + { + "epoch": 0.64, + "grad_norm": 0.490234375, + "learning_rate": 0.00015374048948841898, + "loss": 2.5305, + "step": 15059 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.00015373483240543995, + "loss": 2.7131, + "step": 15060 + }, + { + "epoch": 0.64, + "grad_norm": 0.47265625, + "learning_rate": 0.00015372917508067375, + "loss": 2.4854, + "step": 15061 + }, + { + "epoch": 0.64, + "grad_norm": 0.439453125, + "learning_rate": 0.00015372351751414594, + "loss": 2.5272, + "step": 15062 + }, + { + "epoch": 0.64, + "grad_norm": 0.53515625, + "learning_rate": 0.00015371785970588192, + "loss": 2.7455, + "step": 15063 + }, + { + "epoch": 0.64, + "grad_norm": 0.498046875, + "learning_rate": 0.00015371220165590714, + "loss": 2.566, + "step": 15064 + }, + { + "epoch": 0.64, + "grad_norm": 0.466796875, + "learning_rate": 0.0001537065433642471, + "loss": 2.5381, + "step": 15065 + }, + { + "epoch": 0.64, + "grad_norm": 0.46875, + "learning_rate": 0.00015370088483092725, + "loss": 2.5844, + "step": 15066 + }, + { + "epoch": 0.64, + "grad_norm": 0.484375, + "learning_rate": 0.000153695226055973, + "loss": 2.6276, + "step": 15067 + }, + { + "epoch": 0.64, + "grad_norm": 0.482421875, + "learning_rate": 0.00015368956703940993, + "loss": 2.6234, + "step": 15068 + }, + { + "epoch": 0.64, + "grad_norm": 0.43359375, + "learning_rate": 0.00015368390778126338, + "loss": 2.5197, + "step": 15069 + }, + { + "epoch": 0.64, + "grad_norm": 0.43359375, + "learning_rate": 0.00015367824828155886, + "loss": 2.5324, + "step": 15070 + }, + { + "epoch": 0.64, + "grad_norm": 0.52734375, + "learning_rate": 0.00015367258854032188, + "loss": 2.5402, + "step": 15071 + }, + { + "epoch": 0.64, + "grad_norm": 0.4375, + "learning_rate": 0.00015366692855757786, + "loss": 2.6205, + "step": 15072 + }, + { + "epoch": 0.64, + "grad_norm": 0.49609375, + "learning_rate": 0.00015366126833335223, + "loss": 2.644, + "step": 15073 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.00015365560786767052, + "loss": 2.6816, + "step": 15074 + }, + { + "epoch": 0.64, + "grad_norm": 0.48046875, + "learning_rate": 0.0001536499471605582, + "loss": 2.6569, + "step": 15075 + }, + { + "epoch": 0.64, + "grad_norm": 0.447265625, + "learning_rate": 0.00015364428621204073, + "loss": 2.5944, + "step": 15076 + }, + { + "epoch": 0.64, + "grad_norm": 0.38671875, + "learning_rate": 0.00015363862502214355, + "loss": 2.51, + "step": 15077 + }, + { + "epoch": 0.64, + "grad_norm": 0.462890625, + "learning_rate": 0.0001536329635908922, + "loss": 2.6017, + "step": 15078 + }, + { + "epoch": 0.64, + "grad_norm": 0.44140625, + "learning_rate": 0.00015362730191831206, + "loss": 2.6219, + "step": 15079 + }, + { + "epoch": 0.64, + "grad_norm": 0.41796875, + "learning_rate": 0.0001536216400044287, + "loss": 2.5461, + "step": 15080 + }, + { + "epoch": 0.64, + "grad_norm": 0.44921875, + "learning_rate": 0.00015361597784926753, + "loss": 2.5707, + "step": 15081 + }, + { + "epoch": 0.64, + "grad_norm": 0.435546875, + "learning_rate": 0.0001536103154528541, + "loss": 2.554, + "step": 15082 + }, + { + "epoch": 0.64, + "grad_norm": 0.44921875, + "learning_rate": 0.00015360465281521378, + "loss": 2.5217, + "step": 15083 + }, + { + "epoch": 0.64, + "grad_norm": 0.490234375, + "learning_rate": 0.00015359898993637215, + "loss": 2.5762, + "step": 15084 + }, + { + "epoch": 0.64, + "grad_norm": 0.45703125, + "learning_rate": 0.00015359332681635465, + "loss": 2.6382, + "step": 15085 + }, + { + "epoch": 0.64, + "grad_norm": 0.44140625, + "learning_rate": 0.00015358766345518676, + "loss": 2.527, + "step": 15086 + }, + { + "epoch": 0.64, + "grad_norm": 0.48046875, + "learning_rate": 0.000153581999852894, + "loss": 2.6248, + "step": 15087 + }, + { + "epoch": 0.64, + "grad_norm": 0.5078125, + "learning_rate": 0.0001535763360095018, + "loss": 2.5206, + "step": 15088 + }, + { + "epoch": 0.64, + "grad_norm": 0.455078125, + "learning_rate": 0.00015357067192503564, + "loss": 2.616, + "step": 15089 + }, + { + "epoch": 0.64, + "grad_norm": 0.455078125, + "learning_rate": 0.00015356500759952107, + "loss": 2.6024, + "step": 15090 + }, + { + "epoch": 0.64, + "grad_norm": 0.43359375, + "learning_rate": 0.0001535593430329835, + "loss": 2.585, + "step": 15091 + }, + { + "epoch": 0.64, + "grad_norm": 0.4453125, + "learning_rate": 0.00015355367822544853, + "loss": 2.6243, + "step": 15092 + }, + { + "epoch": 0.64, + "grad_norm": 0.4296875, + "learning_rate": 0.0001535480131769415, + "loss": 2.3513, + "step": 15093 + }, + { + "epoch": 0.64, + "grad_norm": 0.447265625, + "learning_rate": 0.00015354234788748805, + "loss": 2.5646, + "step": 15094 + }, + { + "epoch": 0.64, + "grad_norm": 0.466796875, + "learning_rate": 0.00015353668235711356, + "loss": 2.7195, + "step": 15095 + }, + { + "epoch": 0.64, + "grad_norm": 0.5, + "learning_rate": 0.00015353101658584356, + "loss": 2.7079, + "step": 15096 + }, + { + "epoch": 0.64, + "grad_norm": 0.447265625, + "learning_rate": 0.0001535253505737036, + "loss": 2.4884, + "step": 15097 + }, + { + "epoch": 0.64, + "grad_norm": 0.431640625, + "learning_rate": 0.0001535196843207191, + "loss": 2.5878, + "step": 15098 + }, + { + "epoch": 0.64, + "grad_norm": 0.5, + "learning_rate": 0.00015351401782691553, + "loss": 2.6075, + "step": 15099 + }, + { + "epoch": 0.64, + "grad_norm": 0.423828125, + "learning_rate": 0.0001535083510923185, + "loss": 2.4925, + "step": 15100 + }, + { + "epoch": 0.64, + "grad_norm": 0.46875, + "learning_rate": 0.00015350268411695338, + "loss": 2.6295, + "step": 15101 + }, + { + "epoch": 0.64, + "grad_norm": 0.54296875, + "learning_rate": 0.0001534970169008458, + "loss": 2.7051, + "step": 15102 + }, + { + "epoch": 0.64, + "grad_norm": 0.421875, + "learning_rate": 0.00015349134944402115, + "loss": 2.4034, + "step": 15103 + }, + { + "epoch": 0.64, + "grad_norm": 0.46875, + "learning_rate": 0.000153485681746505, + "loss": 2.5785, + "step": 15104 + }, + { + "epoch": 0.64, + "grad_norm": 0.4609375, + "learning_rate": 0.0001534800138083228, + "loss": 2.5919, + "step": 15105 + }, + { + "epoch": 0.64, + "grad_norm": 0.494140625, + "learning_rate": 0.00015347434562950013, + "loss": 2.625, + "step": 15106 + }, + { + "epoch": 0.64, + "grad_norm": 0.47265625, + "learning_rate": 0.00015346867721006242, + "loss": 2.6066, + "step": 15107 + }, + { + "epoch": 0.64, + "grad_norm": 0.41796875, + "learning_rate": 0.00015346300855003522, + "loss": 2.6686, + "step": 15108 + }, + { + "epoch": 0.64, + "grad_norm": 0.435546875, + "learning_rate": 0.000153457339649444, + "loss": 2.6016, + "step": 15109 + }, + { + "epoch": 0.64, + "grad_norm": 0.5078125, + "learning_rate": 0.0001534516705083143, + "loss": 2.6104, + "step": 15110 + }, + { + "epoch": 0.64, + "grad_norm": 0.40234375, + "learning_rate": 0.00015344600112667163, + "loss": 2.5239, + "step": 15111 + }, + { + "epoch": 0.64, + "grad_norm": 0.46875, + "learning_rate": 0.00015344033150454145, + "loss": 2.5535, + "step": 15112 + }, + { + "epoch": 0.64, + "grad_norm": 0.4375, + "learning_rate": 0.00015343466164194933, + "loss": 2.5888, + "step": 15113 + }, + { + "epoch": 0.64, + "grad_norm": 0.5078125, + "learning_rate": 0.00015342899153892076, + "loss": 2.7777, + "step": 15114 + }, + { + "epoch": 0.64, + "grad_norm": 0.484375, + "learning_rate": 0.00015342332119548123, + "loss": 2.5096, + "step": 15115 + }, + { + "epoch": 0.64, + "grad_norm": 0.435546875, + "learning_rate": 0.00015341765061165633, + "loss": 2.5601, + "step": 15116 + }, + { + "epoch": 0.64, + "grad_norm": 0.4296875, + "learning_rate": 0.00015341197978747148, + "loss": 2.6686, + "step": 15117 + }, + { + "epoch": 0.64, + "grad_norm": 0.48828125, + "learning_rate": 0.00015340630872295226, + "loss": 2.6588, + "step": 15118 + }, + { + "epoch": 0.64, + "grad_norm": 0.46875, + "learning_rate": 0.00015340063741812418, + "loss": 2.487, + "step": 15119 + }, + { + "epoch": 0.64, + "grad_norm": 0.474609375, + "learning_rate": 0.00015339496587301273, + "loss": 2.7737, + "step": 15120 + }, + { + "epoch": 0.64, + "grad_norm": 0.427734375, + "learning_rate": 0.00015338929408764343, + "loss": 2.5494, + "step": 15121 + }, + { + "epoch": 0.64, + "grad_norm": 0.455078125, + "learning_rate": 0.00015338362206204184, + "loss": 2.5583, + "step": 15122 + }, + { + "epoch": 0.64, + "grad_norm": 0.453125, + "learning_rate": 0.00015337794979623345, + "loss": 2.6002, + "step": 15123 + }, + { + "epoch": 0.64, + "grad_norm": 0.46484375, + "learning_rate": 0.0001533722772902438, + "loss": 2.6707, + "step": 15124 + }, + { + "epoch": 0.64, + "grad_norm": 0.439453125, + "learning_rate": 0.00015336660454409842, + "loss": 2.6504, + "step": 15125 + }, + { + "epoch": 0.64, + "grad_norm": 0.447265625, + "learning_rate": 0.00015336093155782283, + "loss": 2.5285, + "step": 15126 + }, + { + "epoch": 0.64, + "grad_norm": 0.435546875, + "learning_rate": 0.0001533552583314425, + "loss": 2.5588, + "step": 15127 + }, + { + "epoch": 0.64, + "grad_norm": 0.494140625, + "learning_rate": 0.00015334958486498305, + "loss": 2.4389, + "step": 15128 + }, + { + "epoch": 0.64, + "grad_norm": 0.47265625, + "learning_rate": 0.00015334391115846998, + "loss": 2.6685, + "step": 15129 + }, + { + "epoch": 0.64, + "grad_norm": 0.42578125, + "learning_rate": 0.00015333823721192874, + "loss": 2.5798, + "step": 15130 + }, + { + "epoch": 0.64, + "grad_norm": 0.4921875, + "learning_rate": 0.00015333256302538497, + "loss": 2.6437, + "step": 15131 + }, + { + "epoch": 0.64, + "grad_norm": 0.470703125, + "learning_rate": 0.00015332688859886418, + "loss": 2.5966, + "step": 15132 + }, + { + "epoch": 0.64, + "grad_norm": 0.494140625, + "learning_rate": 0.00015332121393239184, + "loss": 2.6055, + "step": 15133 + }, + { + "epoch": 0.64, + "grad_norm": 0.4765625, + "learning_rate": 0.00015331553902599352, + "loss": 2.5749, + "step": 15134 + }, + { + "epoch": 0.64, + "grad_norm": 0.421875, + "learning_rate": 0.00015330986387969477, + "loss": 2.4535, + "step": 15135 + }, + { + "epoch": 0.64, + "grad_norm": 0.470703125, + "learning_rate": 0.00015330418849352113, + "loss": 2.5644, + "step": 15136 + }, + { + "epoch": 0.65, + "grad_norm": 0.490234375, + "learning_rate": 0.0001532985128674981, + "loss": 2.5332, + "step": 15137 + }, + { + "epoch": 0.65, + "grad_norm": 0.4765625, + "learning_rate": 0.00015329283700165128, + "loss": 2.6647, + "step": 15138 + }, + { + "epoch": 0.65, + "grad_norm": 0.53125, + "learning_rate": 0.0001532871608960061, + "loss": 2.6118, + "step": 15139 + }, + { + "epoch": 0.65, + "grad_norm": 0.4375, + "learning_rate": 0.00015328148455058825, + "loss": 2.6252, + "step": 15140 + }, + { + "epoch": 0.65, + "grad_norm": 0.4140625, + "learning_rate": 0.0001532758079654231, + "loss": 2.575, + "step": 15141 + }, + { + "epoch": 0.65, + "grad_norm": 0.486328125, + "learning_rate": 0.00015327013114053637, + "loss": 2.5406, + "step": 15142 + }, + { + "epoch": 0.65, + "grad_norm": 0.466796875, + "learning_rate": 0.0001532644540759535, + "loss": 2.5851, + "step": 15143 + }, + { + "epoch": 0.65, + "grad_norm": 0.46875, + "learning_rate": 0.00015325877677170002, + "loss": 2.5198, + "step": 15144 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 0.00015325309922780154, + "loss": 2.6259, + "step": 15145 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015324742144428354, + "loss": 2.6049, + "step": 15146 + }, + { + "epoch": 0.65, + "grad_norm": 0.474609375, + "learning_rate": 0.00015324174342117163, + "loss": 2.6324, + "step": 15147 + }, + { + "epoch": 0.65, + "grad_norm": 0.470703125, + "learning_rate": 0.00015323606515849135, + "loss": 2.7519, + "step": 15148 + }, + { + "epoch": 0.65, + "grad_norm": 0.4609375, + "learning_rate": 0.0001532303866562682, + "loss": 2.7124, + "step": 15149 + }, + { + "epoch": 0.65, + "grad_norm": 0.42578125, + "learning_rate": 0.00015322470791452773, + "loss": 2.5608, + "step": 15150 + }, + { + "epoch": 0.65, + "grad_norm": 0.4765625, + "learning_rate": 0.0001532190289332956, + "loss": 2.5953, + "step": 15151 + }, + { + "epoch": 0.65, + "grad_norm": 0.462890625, + "learning_rate": 0.00015321334971259723, + "loss": 2.5787, + "step": 15152 + }, + { + "epoch": 0.65, + "grad_norm": 0.4453125, + "learning_rate": 0.00015320767025245828, + "loss": 2.4423, + "step": 15153 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015320199055290424, + "loss": 2.4601, + "step": 15154 + }, + { + "epoch": 0.65, + "grad_norm": 0.3984375, + "learning_rate": 0.0001531963106139607, + "loss": 2.4082, + "step": 15155 + }, + { + "epoch": 0.65, + "grad_norm": 0.451171875, + "learning_rate": 0.0001531906304356532, + "loss": 2.5009, + "step": 15156 + }, + { + "epoch": 0.65, + "grad_norm": 0.41796875, + "learning_rate": 0.00015318495001800727, + "loss": 2.5395, + "step": 15157 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.0001531792693610485, + "loss": 2.5579, + "step": 15158 + }, + { + "epoch": 0.65, + "grad_norm": 0.4375, + "learning_rate": 0.00015317358846480249, + "loss": 2.5765, + "step": 15159 + }, + { + "epoch": 0.65, + "grad_norm": 0.46875, + "learning_rate": 0.00015316790732929474, + "loss": 2.5764, + "step": 15160 + }, + { + "epoch": 0.65, + "grad_norm": 0.5625, + "learning_rate": 0.00015316222595455087, + "loss": 2.6372, + "step": 15161 + }, + { + "epoch": 0.65, + "grad_norm": 0.43359375, + "learning_rate": 0.00015315654434059636, + "loss": 2.5455, + "step": 15162 + }, + { + "epoch": 0.65, + "grad_norm": 0.423828125, + "learning_rate": 0.00015315086248745687, + "loss": 2.6302, + "step": 15163 + }, + { + "epoch": 0.65, + "grad_norm": 0.4453125, + "learning_rate": 0.00015314518039515792, + "loss": 2.5925, + "step": 15164 + }, + { + "epoch": 0.65, + "grad_norm": 0.451171875, + "learning_rate": 0.00015313949806372508, + "loss": 2.5515, + "step": 15165 + }, + { + "epoch": 0.65, + "grad_norm": 0.49609375, + "learning_rate": 0.0001531338154931839, + "loss": 2.5995, + "step": 15166 + }, + { + "epoch": 0.65, + "grad_norm": 0.4609375, + "learning_rate": 0.00015312813268355998, + "loss": 2.5744, + "step": 15167 + }, + { + "epoch": 0.65, + "grad_norm": 0.416015625, + "learning_rate": 0.00015312244963487887, + "loss": 2.5156, + "step": 15168 + }, + { + "epoch": 0.65, + "grad_norm": 0.439453125, + "learning_rate": 0.00015311676634716616, + "loss": 2.6636, + "step": 15169 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.0001531110828204474, + "loss": 2.6336, + "step": 15170 + }, + { + "epoch": 0.65, + "grad_norm": 0.447265625, + "learning_rate": 0.00015310539905474818, + "loss": 2.5506, + "step": 15171 + }, + { + "epoch": 0.65, + "grad_norm": 0.48828125, + "learning_rate": 0.00015309971505009408, + "loss": 2.5433, + "step": 15172 + }, + { + "epoch": 0.65, + "grad_norm": 0.470703125, + "learning_rate": 0.0001530940308065107, + "loss": 2.6687, + "step": 15173 + }, + { + "epoch": 0.65, + "grad_norm": 0.490234375, + "learning_rate": 0.00015308834632402353, + "loss": 2.7342, + "step": 15174 + }, + { + "epoch": 0.65, + "grad_norm": 0.3828125, + "learning_rate": 0.00015308266160265826, + "loss": 2.3729, + "step": 15175 + }, + { + "epoch": 0.65, + "grad_norm": 0.462890625, + "learning_rate": 0.00015307697664244035, + "loss": 2.4739, + "step": 15176 + }, + { + "epoch": 0.65, + "grad_norm": 0.47265625, + "learning_rate": 0.00015307129144339547, + "loss": 2.587, + "step": 15177 + }, + { + "epoch": 0.65, + "grad_norm": 0.451171875, + "learning_rate": 0.00015306560600554918, + "loss": 2.5928, + "step": 15178 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015305992032892704, + "loss": 2.6272, + "step": 15179 + }, + { + "epoch": 0.65, + "grad_norm": 0.44140625, + "learning_rate": 0.00015305423441355466, + "loss": 2.606, + "step": 15180 + }, + { + "epoch": 0.65, + "grad_norm": 0.53125, + "learning_rate": 0.00015304854825945764, + "loss": 2.6172, + "step": 15181 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 0.00015304286186666147, + "loss": 2.5995, + "step": 15182 + }, + { + "epoch": 0.65, + "grad_norm": 0.51171875, + "learning_rate": 0.00015303717523519188, + "loss": 2.6135, + "step": 15183 + }, + { + "epoch": 0.65, + "grad_norm": 0.48828125, + "learning_rate": 0.00015303148836507435, + "loss": 2.7258, + "step": 15184 + }, + { + "epoch": 0.65, + "grad_norm": 0.46875, + "learning_rate": 0.0001530258012563345, + "loss": 2.6059, + "step": 15185 + }, + { + "epoch": 0.65, + "grad_norm": 0.48828125, + "learning_rate": 0.00015302011390899792, + "loss": 2.6261, + "step": 15186 + }, + { + "epoch": 0.65, + "grad_norm": 0.451171875, + "learning_rate": 0.00015301442632309026, + "loss": 2.5871, + "step": 15187 + }, + { + "epoch": 0.65, + "grad_norm": 0.443359375, + "learning_rate": 0.000153008738498637, + "loss": 2.5524, + "step": 15188 + }, + { + "epoch": 0.65, + "grad_norm": 0.4140625, + "learning_rate": 0.00015300305043566378, + "loss": 2.3969, + "step": 15189 + }, + { + "epoch": 0.65, + "grad_norm": 0.4296875, + "learning_rate": 0.00015299736213419623, + "loss": 2.5594, + "step": 15190 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.00015299167359425993, + "loss": 2.5775, + "step": 15191 + }, + { + "epoch": 0.65, + "grad_norm": 0.49609375, + "learning_rate": 0.00015298598481588042, + "loss": 2.5959, + "step": 15192 + }, + { + "epoch": 0.65, + "grad_norm": 0.466796875, + "learning_rate": 0.00015298029579908338, + "loss": 2.619, + "step": 15193 + }, + { + "epoch": 0.65, + "grad_norm": 0.458984375, + "learning_rate": 0.00015297460654389435, + "loss": 2.4065, + "step": 15194 + }, + { + "epoch": 0.65, + "grad_norm": 0.39453125, + "learning_rate": 0.00015296891705033898, + "loss": 2.519, + "step": 15195 + }, + { + "epoch": 0.65, + "grad_norm": 0.4765625, + "learning_rate": 0.00015296322731844282, + "loss": 2.6464, + "step": 15196 + }, + { + "epoch": 0.65, + "grad_norm": 0.451171875, + "learning_rate": 0.00015295753734823148, + "loss": 2.6256, + "step": 15197 + }, + { + "epoch": 0.65, + "grad_norm": 0.51171875, + "learning_rate": 0.00015295184713973062, + "loss": 2.5861, + "step": 15198 + }, + { + "epoch": 0.65, + "grad_norm": 0.494140625, + "learning_rate": 0.00015294615669296578, + "loss": 2.6309, + "step": 15199 + }, + { + "epoch": 0.65, + "grad_norm": 0.470703125, + "learning_rate": 0.00015294046600796253, + "loss": 2.6308, + "step": 15200 + }, + { + "epoch": 0.65, + "grad_norm": 0.466796875, + "learning_rate": 0.0001529347750847466, + "loss": 2.6376, + "step": 15201 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 0.00015292908392334348, + "loss": 2.4929, + "step": 15202 + }, + { + "epoch": 0.65, + "grad_norm": 0.48046875, + "learning_rate": 0.00015292339252377887, + "loss": 2.6677, + "step": 15203 + }, + { + "epoch": 0.65, + "grad_norm": 0.43359375, + "learning_rate": 0.00015291770088607831, + "loss": 2.575, + "step": 15204 + }, + { + "epoch": 0.65, + "grad_norm": 0.41796875, + "learning_rate": 0.00015291200901026744, + "loss": 2.4898, + "step": 15205 + }, + { + "epoch": 0.65, + "grad_norm": 0.474609375, + "learning_rate": 0.00015290631689637184, + "loss": 2.5906, + "step": 15206 + }, + { + "epoch": 0.65, + "grad_norm": 0.5546875, + "learning_rate": 0.00015290062454441722, + "loss": 2.705, + "step": 15207 + }, + { + "epoch": 0.65, + "grad_norm": 0.44921875, + "learning_rate": 0.00015289493195442906, + "loss": 2.55, + "step": 15208 + }, + { + "epoch": 0.65, + "grad_norm": 0.416015625, + "learning_rate": 0.00015288923912643306, + "loss": 2.7027, + "step": 15209 + }, + { + "epoch": 0.65, + "grad_norm": 0.455078125, + "learning_rate": 0.0001528835460604548, + "loss": 2.6137, + "step": 15210 + }, + { + "epoch": 0.65, + "grad_norm": 0.5390625, + "learning_rate": 0.0001528778527565199, + "loss": 2.6176, + "step": 15211 + }, + { + "epoch": 0.65, + "grad_norm": 0.4140625, + "learning_rate": 0.00015287215921465404, + "loss": 2.3992, + "step": 15212 + }, + { + "epoch": 0.65, + "grad_norm": 0.474609375, + "learning_rate": 0.00015286646543488276, + "loss": 2.5626, + "step": 15213 + }, + { + "epoch": 0.65, + "grad_norm": 0.4375, + "learning_rate": 0.00015286077141723166, + "loss": 2.545, + "step": 15214 + }, + { + "epoch": 0.65, + "grad_norm": 0.416015625, + "learning_rate": 0.00015285507716172647, + "loss": 2.6467, + "step": 15215 + }, + { + "epoch": 0.65, + "grad_norm": 0.4453125, + "learning_rate": 0.0001528493826683927, + "loss": 2.6181, + "step": 15216 + }, + { + "epoch": 0.65, + "grad_norm": 0.390625, + "learning_rate": 0.00015284368793725604, + "loss": 2.5445, + "step": 15217 + }, + { + "epoch": 0.65, + "grad_norm": 0.44140625, + "learning_rate": 0.0001528379929683421, + "loss": 2.6042, + "step": 15218 + }, + { + "epoch": 0.65, + "grad_norm": 0.466796875, + "learning_rate": 0.00015283229776167652, + "loss": 2.5819, + "step": 15219 + }, + { + "epoch": 0.65, + "grad_norm": 0.455078125, + "learning_rate": 0.0001528266023172849, + "loss": 2.5639, + "step": 15220 + }, + { + "epoch": 0.65, + "grad_norm": 0.431640625, + "learning_rate": 0.00015282090663519286, + "loss": 2.5257, + "step": 15221 + }, + { + "epoch": 0.65, + "grad_norm": 0.4296875, + "learning_rate": 0.00015281521071542603, + "loss": 2.5954, + "step": 15222 + }, + { + "epoch": 0.65, + "grad_norm": 0.439453125, + "learning_rate": 0.00015280951455801009, + "loss": 2.5506, + "step": 15223 + }, + { + "epoch": 0.65, + "grad_norm": 0.490234375, + "learning_rate": 0.0001528038181629706, + "loss": 2.5313, + "step": 15224 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015279812153033327, + "loss": 2.6384, + "step": 15225 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015279242466012365, + "loss": 2.4517, + "step": 15226 + }, + { + "epoch": 0.65, + "grad_norm": 0.474609375, + "learning_rate": 0.00015278672755236742, + "loss": 2.5509, + "step": 15227 + }, + { + "epoch": 0.65, + "grad_norm": 0.5078125, + "learning_rate": 0.00015278103020709022, + "loss": 2.6667, + "step": 15228 + }, + { + "epoch": 0.65, + "grad_norm": 0.431640625, + "learning_rate": 0.00015277533262431767, + "loss": 2.6067, + "step": 15229 + }, + { + "epoch": 0.65, + "grad_norm": 0.4921875, + "learning_rate": 0.0001527696348040754, + "loss": 2.561, + "step": 15230 + }, + { + "epoch": 0.65, + "grad_norm": 0.40625, + "learning_rate": 0.0001527639367463891, + "loss": 2.6698, + "step": 15231 + }, + { + "epoch": 0.65, + "grad_norm": 0.443359375, + "learning_rate": 0.0001527582384512843, + "loss": 2.6699, + "step": 15232 + }, + { + "epoch": 0.65, + "grad_norm": 0.447265625, + "learning_rate": 0.00015275253991878674, + "loss": 2.49, + "step": 15233 + }, + { + "epoch": 0.65, + "grad_norm": 0.44921875, + "learning_rate": 0.00015274684114892204, + "loss": 2.5558, + "step": 15234 + }, + { + "epoch": 0.65, + "grad_norm": 0.470703125, + "learning_rate": 0.00015274114214171582, + "loss": 2.573, + "step": 15235 + }, + { + "epoch": 0.65, + "grad_norm": 0.43359375, + "learning_rate": 0.00015273544289719375, + "loss": 2.6126, + "step": 15236 + }, + { + "epoch": 0.65, + "grad_norm": 0.40625, + "learning_rate": 0.0001527297434153814, + "loss": 2.4239, + "step": 15237 + }, + { + "epoch": 0.65, + "grad_norm": 0.419921875, + "learning_rate": 0.00015272404369630452, + "loss": 2.452, + "step": 15238 + }, + { + "epoch": 0.65, + "grad_norm": 0.49609375, + "learning_rate": 0.00015271834373998874, + "loss": 2.5009, + "step": 15239 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 0.00015271264354645964, + "loss": 2.5146, + "step": 15240 + }, + { + "epoch": 0.65, + "grad_norm": 0.431640625, + "learning_rate": 0.00015270694311574293, + "loss": 2.625, + "step": 15241 + }, + { + "epoch": 0.65, + "grad_norm": 0.419921875, + "learning_rate": 0.00015270124244786423, + "loss": 2.5694, + "step": 15242 + }, + { + "epoch": 0.65, + "grad_norm": 0.515625, + "learning_rate": 0.00015269554154284922, + "loss": 2.7886, + "step": 15243 + }, + { + "epoch": 0.65, + "grad_norm": 0.49609375, + "learning_rate": 0.0001526898404007235, + "loss": 2.585, + "step": 15244 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 0.00015268413902151274, + "loss": 2.6391, + "step": 15245 + }, + { + "epoch": 0.65, + "grad_norm": 0.421875, + "learning_rate": 0.00015267843740524265, + "loss": 2.5773, + "step": 15246 + }, + { + "epoch": 0.65, + "grad_norm": 0.439453125, + "learning_rate": 0.00015267273555193885, + "loss": 2.7206, + "step": 15247 + }, + { + "epoch": 0.65, + "grad_norm": 0.46484375, + "learning_rate": 0.0001526670334616269, + "loss": 2.5134, + "step": 15248 + }, + { + "epoch": 0.65, + "grad_norm": 0.462890625, + "learning_rate": 0.00015266133113433266, + "loss": 2.6131, + "step": 15249 + }, + { + "epoch": 0.65, + "grad_norm": 0.5, + "learning_rate": 0.0001526556285700816, + "loss": 2.6981, + "step": 15250 + }, + { + "epoch": 0.65, + "grad_norm": 0.412109375, + "learning_rate": 0.00015264992576889948, + "loss": 2.4616, + "step": 15251 + }, + { + "epoch": 0.65, + "grad_norm": 0.5390625, + "learning_rate": 0.00015264422273081197, + "loss": 2.4762, + "step": 15252 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 0.00015263851945584466, + "loss": 2.5767, + "step": 15253 + }, + { + "epoch": 0.65, + "grad_norm": 0.4296875, + "learning_rate": 0.00015263281594402324, + "loss": 2.4838, + "step": 15254 + }, + { + "epoch": 0.65, + "grad_norm": 0.462890625, + "learning_rate": 0.00015262711219537342, + "loss": 2.631, + "step": 15255 + }, + { + "epoch": 0.65, + "grad_norm": 0.423828125, + "learning_rate": 0.00015262140820992078, + "loss": 2.5612, + "step": 15256 + }, + { + "epoch": 0.65, + "grad_norm": 0.5078125, + "learning_rate": 0.00015261570398769107, + "loss": 2.6392, + "step": 15257 + }, + { + "epoch": 0.65, + "grad_norm": 0.44140625, + "learning_rate": 0.00015260999952870988, + "loss": 2.4455, + "step": 15258 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 0.00015260429483300296, + "loss": 2.6254, + "step": 15259 + }, + { + "epoch": 0.65, + "grad_norm": 0.4453125, + "learning_rate": 0.00015259858990059595, + "loss": 2.5618, + "step": 15260 + }, + { + "epoch": 0.65, + "grad_norm": 0.486328125, + "learning_rate": 0.00015259288473151448, + "loss": 2.6244, + "step": 15261 + }, + { + "epoch": 0.65, + "grad_norm": 0.466796875, + "learning_rate": 0.00015258717932578424, + "loss": 2.5499, + "step": 15262 + }, + { + "epoch": 0.65, + "grad_norm": 0.455078125, + "learning_rate": 0.00015258147368343093, + "loss": 2.4342, + "step": 15263 + }, + { + "epoch": 0.65, + "grad_norm": 0.43359375, + "learning_rate": 0.0001525757678044802, + "loss": 2.5931, + "step": 15264 + }, + { + "epoch": 0.65, + "grad_norm": 0.462890625, + "learning_rate": 0.00015257006168895775, + "loss": 2.5343, + "step": 15265 + }, + { + "epoch": 0.65, + "grad_norm": 0.53125, + "learning_rate": 0.0001525643553368892, + "loss": 2.5792, + "step": 15266 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015255864874830026, + "loss": 2.4958, + "step": 15267 + }, + { + "epoch": 0.65, + "grad_norm": 0.453125, + "learning_rate": 0.00015255294192321664, + "loss": 2.6014, + "step": 15268 + }, + { + "epoch": 0.65, + "grad_norm": 0.4453125, + "learning_rate": 0.00015254723486166397, + "loss": 2.5755, + "step": 15269 + }, + { + "epoch": 0.65, + "grad_norm": 0.435546875, + "learning_rate": 0.00015254152756366795, + "loss": 2.6281, + "step": 15270 + }, + { + "epoch": 0.65, + "grad_norm": 0.5, + "learning_rate": 0.00015253582002925427, + "loss": 2.7433, + "step": 15271 + }, + { + "epoch": 0.65, + "grad_norm": 0.431640625, + "learning_rate": 0.00015253011225844858, + "loss": 2.4895, + "step": 15272 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.00015252440425127657, + "loss": 2.5863, + "step": 15273 + }, + { + "epoch": 0.65, + "grad_norm": 0.443359375, + "learning_rate": 0.000152518696007764, + "loss": 2.6718, + "step": 15274 + }, + { + "epoch": 0.65, + "grad_norm": 0.6171875, + "learning_rate": 0.0001525129875279364, + "loss": 2.6825, + "step": 15275 + }, + { + "epoch": 0.65, + "grad_norm": 0.423828125, + "learning_rate": 0.00015250727881181957, + "loss": 2.4976, + "step": 15276 + }, + { + "epoch": 0.65, + "grad_norm": 0.40625, + "learning_rate": 0.00015250156985943921, + "loss": 2.4502, + "step": 15277 + }, + { + "epoch": 0.65, + "grad_norm": 0.44921875, + "learning_rate": 0.00015249586067082099, + "loss": 2.4979, + "step": 15278 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.00015249015124599056, + "loss": 2.6041, + "step": 15279 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.0001524844415849736, + "loss": 2.464, + "step": 15280 + }, + { + "epoch": 0.65, + "grad_norm": 0.498046875, + "learning_rate": 0.00015247873168779585, + "loss": 2.56, + "step": 15281 + }, + { + "epoch": 0.65, + "grad_norm": 0.52734375, + "learning_rate": 0.000152473021554483, + "loss": 2.6876, + "step": 15282 + }, + { + "epoch": 0.65, + "grad_norm": 0.421875, + "learning_rate": 0.00015246731118506073, + "loss": 2.6319, + "step": 15283 + }, + { + "epoch": 0.65, + "grad_norm": 0.443359375, + "learning_rate": 0.00015246160057955473, + "loss": 2.6396, + "step": 15284 + }, + { + "epoch": 0.65, + "grad_norm": 0.455078125, + "learning_rate": 0.00015245588973799068, + "loss": 2.6168, + "step": 15285 + }, + { + "epoch": 0.65, + "grad_norm": 0.49609375, + "learning_rate": 0.00015245017866039433, + "loss": 2.6739, + "step": 15286 + }, + { + "epoch": 0.65, + "grad_norm": 0.48828125, + "learning_rate": 0.00015244446734679135, + "loss": 2.5473, + "step": 15287 + }, + { + "epoch": 0.65, + "grad_norm": 0.41796875, + "learning_rate": 0.0001524387557972074, + "loss": 2.5717, + "step": 15288 + }, + { + "epoch": 0.65, + "grad_norm": 0.41015625, + "learning_rate": 0.0001524330440116682, + "loss": 2.5174, + "step": 15289 + }, + { + "epoch": 0.65, + "grad_norm": 0.443359375, + "learning_rate": 0.00015242733199019953, + "loss": 2.4546, + "step": 15290 + }, + { + "epoch": 0.65, + "grad_norm": 0.5078125, + "learning_rate": 0.00015242161973282693, + "loss": 2.6782, + "step": 15291 + }, + { + "epoch": 0.65, + "grad_norm": 0.46484375, + "learning_rate": 0.00015241590723957628, + "loss": 2.6244, + "step": 15292 + }, + { + "epoch": 0.65, + "grad_norm": 0.435546875, + "learning_rate": 0.00015241019451047317, + "loss": 2.4426, + "step": 15293 + }, + { + "epoch": 0.65, + "grad_norm": 0.416015625, + "learning_rate": 0.00015240448154554335, + "loss": 2.6256, + "step": 15294 + }, + { + "epoch": 0.65, + "grad_norm": 0.498046875, + "learning_rate": 0.00015239876834481254, + "loss": 2.6982, + "step": 15295 + }, + { + "epoch": 0.65, + "grad_norm": 0.484375, + "learning_rate": 0.00015239305490830638, + "loss": 2.6335, + "step": 15296 + }, + { + "epoch": 0.65, + "grad_norm": 0.48046875, + "learning_rate": 0.00015238734123605062, + "loss": 2.5624, + "step": 15297 + }, + { + "epoch": 0.65, + "grad_norm": 0.451171875, + "learning_rate": 0.000152381627328071, + "loss": 2.7161, + "step": 15298 + }, + { + "epoch": 0.65, + "grad_norm": 0.4765625, + "learning_rate": 0.00015237591318439315, + "loss": 2.5662, + "step": 15299 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.00015237019880504288, + "loss": 2.5143, + "step": 15300 + }, + { + "epoch": 0.65, + "grad_norm": 0.4921875, + "learning_rate": 0.00015236448419004583, + "loss": 2.6965, + "step": 15301 + }, + { + "epoch": 0.65, + "grad_norm": 0.4453125, + "learning_rate": 0.00015235876933942776, + "loss": 2.6034, + "step": 15302 + }, + { + "epoch": 0.65, + "grad_norm": 0.462890625, + "learning_rate": 0.00015235305425321434, + "loss": 2.5777, + "step": 15303 + }, + { + "epoch": 0.65, + "grad_norm": 0.51171875, + "learning_rate": 0.00015234733893143134, + "loss": 2.5349, + "step": 15304 + }, + { + "epoch": 0.65, + "grad_norm": 0.4375, + "learning_rate": 0.0001523416233741044, + "loss": 2.5285, + "step": 15305 + }, + { + "epoch": 0.65, + "grad_norm": 0.4296875, + "learning_rate": 0.00015233590758125933, + "loss": 2.5516, + "step": 15306 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.0001523301915529218, + "loss": 2.5068, + "step": 15307 + }, + { + "epoch": 0.65, + "grad_norm": 0.4609375, + "learning_rate": 0.00015232447528911748, + "loss": 2.6649, + "step": 15308 + }, + { + "epoch": 0.65, + "grad_norm": 0.4375, + "learning_rate": 0.0001523187587898722, + "loss": 2.5637, + "step": 15309 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015231304205521158, + "loss": 2.6162, + "step": 15310 + }, + { + "epoch": 0.65, + "grad_norm": 0.435546875, + "learning_rate": 0.00015230732508516144, + "loss": 2.5832, + "step": 15311 + }, + { + "epoch": 0.65, + "grad_norm": 0.42578125, + "learning_rate": 0.0001523016078797474, + "loss": 2.5363, + "step": 15312 + }, + { + "epoch": 0.65, + "grad_norm": 0.458984375, + "learning_rate": 0.00015229589043899526, + "loss": 2.6177, + "step": 15313 + }, + { + "epoch": 0.65, + "grad_norm": 0.431640625, + "learning_rate": 0.00015229017276293077, + "loss": 2.6553, + "step": 15314 + }, + { + "epoch": 0.65, + "grad_norm": 0.484375, + "learning_rate": 0.00015228445485157953, + "loss": 2.5743, + "step": 15315 + }, + { + "epoch": 0.65, + "grad_norm": 0.462890625, + "learning_rate": 0.0001522787367049674, + "loss": 2.5127, + "step": 15316 + }, + { + "epoch": 0.65, + "grad_norm": 0.4609375, + "learning_rate": 0.00015227301832312006, + "loss": 2.4647, + "step": 15317 + }, + { + "epoch": 0.65, + "grad_norm": 0.44140625, + "learning_rate": 0.00015226729970606323, + "loss": 2.6034, + "step": 15318 + }, + { + "epoch": 0.65, + "grad_norm": 0.515625, + "learning_rate": 0.00015226158085382267, + "loss": 2.5817, + "step": 15319 + }, + { + "epoch": 0.65, + "grad_norm": 0.5, + "learning_rate": 0.00015225586176642407, + "loss": 2.621, + "step": 15320 + }, + { + "epoch": 0.65, + "grad_norm": 0.50390625, + "learning_rate": 0.0001522501424438932, + "loss": 2.6254, + "step": 15321 + }, + { + "epoch": 0.65, + "grad_norm": 0.5078125, + "learning_rate": 0.0001522444228862558, + "loss": 2.6127, + "step": 15322 + }, + { + "epoch": 0.65, + "grad_norm": 0.4375, + "learning_rate": 0.00015223870309353756, + "loss": 2.4427, + "step": 15323 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015223298306576427, + "loss": 2.5373, + "step": 15324 + }, + { + "epoch": 0.65, + "grad_norm": 0.48046875, + "learning_rate": 0.00015222726280296163, + "loss": 2.581, + "step": 15325 + }, + { + "epoch": 0.65, + "grad_norm": 0.4375, + "learning_rate": 0.0001522215423051554, + "loss": 2.541, + "step": 15326 + }, + { + "epoch": 0.65, + "grad_norm": 0.470703125, + "learning_rate": 0.00015221582157237134, + "loss": 2.6728, + "step": 15327 + }, + { + "epoch": 0.65, + "grad_norm": 0.546875, + "learning_rate": 0.00015221010060463512, + "loss": 2.5972, + "step": 15328 + }, + { + "epoch": 0.65, + "grad_norm": 0.490234375, + "learning_rate": 0.00015220437940197256, + "loss": 2.6795, + "step": 15329 + }, + { + "epoch": 0.65, + "grad_norm": 0.5078125, + "learning_rate": 0.00015219865796440936, + "loss": 2.5793, + "step": 15330 + }, + { + "epoch": 0.65, + "grad_norm": 0.5078125, + "learning_rate": 0.00015219293629197128, + "loss": 2.7481, + "step": 15331 + }, + { + "epoch": 0.65, + "grad_norm": 0.48046875, + "learning_rate": 0.00015218721438468405, + "loss": 2.6036, + "step": 15332 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.00015218149224257346, + "loss": 2.5005, + "step": 15333 + }, + { + "epoch": 0.65, + "grad_norm": 0.484375, + "learning_rate": 0.00015217576986566517, + "loss": 2.6987, + "step": 15334 + }, + { + "epoch": 0.65, + "grad_norm": 0.5234375, + "learning_rate": 0.00015217004725398504, + "loss": 2.5807, + "step": 15335 + }, + { + "epoch": 0.65, + "grad_norm": 0.458984375, + "learning_rate": 0.00015216432440755875, + "loss": 2.6436, + "step": 15336 + }, + { + "epoch": 0.65, + "grad_norm": 0.49609375, + "learning_rate": 0.00015215860132641203, + "loss": 2.7065, + "step": 15337 + }, + { + "epoch": 0.65, + "grad_norm": 0.45703125, + "learning_rate": 0.0001521528780105707, + "loss": 2.5409, + "step": 15338 + }, + { + "epoch": 0.65, + "grad_norm": 0.4375, + "learning_rate": 0.00015214715446006046, + "loss": 2.4916, + "step": 15339 + }, + { + "epoch": 0.65, + "grad_norm": 0.365234375, + "learning_rate": 0.00015214143067490707, + "loss": 2.5288, + "step": 15340 + }, + { + "epoch": 0.65, + "grad_norm": 0.486328125, + "learning_rate": 0.00015213570665513634, + "loss": 2.5375, + "step": 15341 + }, + { + "epoch": 0.65, + "grad_norm": 0.478515625, + "learning_rate": 0.00015212998240077394, + "loss": 2.5942, + "step": 15342 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.0001521242579118457, + "loss": 2.545, + "step": 15343 + }, + { + "epoch": 0.65, + "grad_norm": 0.5078125, + "learning_rate": 0.00015211853318837734, + "loss": 2.5799, + "step": 15344 + }, + { + "epoch": 0.65, + "grad_norm": 0.443359375, + "learning_rate": 0.0001521128082303946, + "loss": 2.5648, + "step": 15345 + }, + { + "epoch": 0.65, + "grad_norm": 0.4609375, + "learning_rate": 0.00015210708303792334, + "loss": 2.5793, + "step": 15346 + }, + { + "epoch": 0.65, + "grad_norm": 0.423828125, + "learning_rate": 0.0001521013576109892, + "loss": 2.6586, + "step": 15347 + }, + { + "epoch": 0.65, + "grad_norm": 0.494140625, + "learning_rate": 0.00015209563194961797, + "loss": 2.4898, + "step": 15348 + }, + { + "epoch": 0.65, + "grad_norm": 0.5, + "learning_rate": 0.00015208990605383547, + "loss": 2.6744, + "step": 15349 + }, + { + "epoch": 0.65, + "grad_norm": 0.474609375, + "learning_rate": 0.0001520841799236674, + "loss": 2.6381, + "step": 15350 + }, + { + "epoch": 0.65, + "grad_norm": 0.4921875, + "learning_rate": 0.00015207845355913962, + "loss": 2.5936, + "step": 15351 + }, + { + "epoch": 0.65, + "grad_norm": 0.419921875, + "learning_rate": 0.00015207272696027776, + "loss": 2.5092, + "step": 15352 + }, + { + "epoch": 0.65, + "grad_norm": 0.484375, + "learning_rate": 0.0001520670001271077, + "loss": 2.6018, + "step": 15353 + }, + { + "epoch": 0.65, + "grad_norm": 0.490234375, + "learning_rate": 0.00015206127305965518, + "loss": 2.6745, + "step": 15354 + }, + { + "epoch": 0.65, + "grad_norm": 0.419921875, + "learning_rate": 0.00015205554575794587, + "loss": 2.5375, + "step": 15355 + }, + { + "epoch": 0.65, + "grad_norm": 0.4765625, + "learning_rate": 0.00015204981822200571, + "loss": 2.713, + "step": 15356 + }, + { + "epoch": 0.65, + "grad_norm": 0.49609375, + "learning_rate": 0.00015204409045186035, + "loss": 2.7074, + "step": 15357 + }, + { + "epoch": 0.65, + "grad_norm": 0.458984375, + "learning_rate": 0.00015203836244753565, + "loss": 2.6544, + "step": 15358 + }, + { + "epoch": 0.65, + "grad_norm": 0.453125, + "learning_rate": 0.0001520326342090573, + "loss": 2.5841, + "step": 15359 + }, + { + "epoch": 0.65, + "grad_norm": 0.451171875, + "learning_rate": 0.00015202690573645113, + "loss": 2.5912, + "step": 15360 + }, + { + "epoch": 0.65, + "grad_norm": 0.451171875, + "learning_rate": 0.00015202117702974288, + "loss": 2.4968, + "step": 15361 + }, + { + "epoch": 0.65, + "grad_norm": 0.427734375, + "learning_rate": 0.00015201544808895835, + "loss": 2.5896, + "step": 15362 + }, + { + "epoch": 0.65, + "grad_norm": 0.462890625, + "learning_rate": 0.00015200971891412332, + "loss": 2.5078, + "step": 15363 + }, + { + "epoch": 0.65, + "grad_norm": 0.41796875, + "learning_rate": 0.00015200398950526356, + "loss": 2.6468, + "step": 15364 + }, + { + "epoch": 0.65, + "grad_norm": 0.43359375, + "learning_rate": 0.00015199825986240485, + "loss": 2.4653, + "step": 15365 + }, + { + "epoch": 0.65, + "grad_norm": 0.431640625, + "learning_rate": 0.000151992529985573, + "loss": 2.6113, + "step": 15366 + }, + { + "epoch": 0.65, + "grad_norm": 0.43359375, + "learning_rate": 0.00015198679987479368, + "loss": 2.5252, + "step": 15367 + }, + { + "epoch": 0.65, + "grad_norm": 0.51953125, + "learning_rate": 0.00015198106953009285, + "loss": 2.6591, + "step": 15368 + }, + { + "epoch": 0.65, + "grad_norm": 0.466796875, + "learning_rate": 0.00015197533895149618, + "loss": 2.5523, + "step": 15369 + }, + { + "epoch": 0.65, + "grad_norm": 0.42578125, + "learning_rate": 0.00015196960813902949, + "loss": 2.5069, + "step": 15370 + }, + { + "epoch": 0.66, + "grad_norm": 0.490234375, + "learning_rate": 0.00015196387709271854, + "loss": 2.6215, + "step": 15371 + }, + { + "epoch": 0.66, + "grad_norm": 0.431640625, + "learning_rate": 0.0001519581458125891, + "loss": 2.5839, + "step": 15372 + }, + { + "epoch": 0.66, + "grad_norm": 0.41015625, + "learning_rate": 0.00015195241429866708, + "loss": 2.4761, + "step": 15373 + }, + { + "epoch": 0.66, + "grad_norm": 0.5, + "learning_rate": 0.0001519466825509781, + "loss": 2.7033, + "step": 15374 + }, + { + "epoch": 0.66, + "grad_norm": 0.50390625, + "learning_rate": 0.00015194095056954805, + "loss": 2.6437, + "step": 15375 + }, + { + "epoch": 0.66, + "grad_norm": 0.423828125, + "learning_rate": 0.00015193521835440272, + "loss": 2.536, + "step": 15376 + }, + { + "epoch": 0.66, + "grad_norm": 0.515625, + "learning_rate": 0.00015192948590556787, + "loss": 2.5973, + "step": 15377 + }, + { + "epoch": 0.66, + "grad_norm": 0.4765625, + "learning_rate": 0.00015192375322306935, + "loss": 2.5923, + "step": 15378 + }, + { + "epoch": 0.66, + "grad_norm": 0.44921875, + "learning_rate": 0.00015191802030693288, + "loss": 2.6249, + "step": 15379 + }, + { + "epoch": 0.66, + "grad_norm": 0.439453125, + "learning_rate": 0.00015191228715718432, + "loss": 2.5127, + "step": 15380 + }, + { + "epoch": 0.66, + "grad_norm": 0.458984375, + "learning_rate": 0.00015190655377384941, + "loss": 2.595, + "step": 15381 + }, + { + "epoch": 0.66, + "grad_norm": 0.458984375, + "learning_rate": 0.00015190082015695402, + "loss": 2.5236, + "step": 15382 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.00015189508630652384, + "loss": 2.6119, + "step": 15383 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.0001518893522225848, + "loss": 2.6864, + "step": 15384 + }, + { + "epoch": 0.66, + "grad_norm": 0.4375, + "learning_rate": 0.0001518836179051626, + "loss": 2.5443, + "step": 15385 + }, + { + "epoch": 0.66, + "grad_norm": 0.474609375, + "learning_rate": 0.0001518778833542831, + "loss": 2.6947, + "step": 15386 + }, + { + "epoch": 0.66, + "grad_norm": 0.51171875, + "learning_rate": 0.00015187214856997208, + "loss": 2.6009, + "step": 15387 + }, + { + "epoch": 0.66, + "grad_norm": 0.451171875, + "learning_rate": 0.00015186641355225536, + "loss": 2.5537, + "step": 15388 + }, + { + "epoch": 0.66, + "grad_norm": 0.455078125, + "learning_rate": 0.00015186067830115873, + "loss": 2.5622, + "step": 15389 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.000151854942816708, + "loss": 2.6191, + "step": 15390 + }, + { + "epoch": 0.66, + "grad_norm": 0.490234375, + "learning_rate": 0.00015184920709892892, + "loss": 2.6178, + "step": 15391 + }, + { + "epoch": 0.66, + "grad_norm": 0.451171875, + "learning_rate": 0.00015184347114784744, + "loss": 2.4314, + "step": 15392 + }, + { + "epoch": 0.66, + "grad_norm": 0.412109375, + "learning_rate": 0.00015183773496348926, + "loss": 2.6676, + "step": 15393 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.00015183199854588018, + "loss": 2.5653, + "step": 15394 + }, + { + "epoch": 0.66, + "grad_norm": 0.47265625, + "learning_rate": 0.00015182626189504613, + "loss": 2.5686, + "step": 15395 + }, + { + "epoch": 0.66, + "grad_norm": 0.42578125, + "learning_rate": 0.00015182052501101274, + "loss": 2.461, + "step": 15396 + }, + { + "epoch": 0.66, + "grad_norm": 0.4609375, + "learning_rate": 0.000151814787893806, + "loss": 2.6403, + "step": 15397 + }, + { + "epoch": 0.66, + "grad_norm": 0.5, + "learning_rate": 0.0001518090505434516, + "loss": 2.4956, + "step": 15398 + }, + { + "epoch": 0.66, + "grad_norm": 0.40625, + "learning_rate": 0.00015180331295997545, + "loss": 2.5678, + "step": 15399 + }, + { + "epoch": 0.66, + "grad_norm": 0.421875, + "learning_rate": 0.00015179757514340332, + "loss": 2.6283, + "step": 15400 + }, + { + "epoch": 0.66, + "grad_norm": 0.5390625, + "learning_rate": 0.000151791837093761, + "loss": 2.5143, + "step": 15401 + }, + { + "epoch": 0.66, + "grad_norm": 0.50390625, + "learning_rate": 0.00015178609881107437, + "loss": 2.4696, + "step": 15402 + }, + { + "epoch": 0.66, + "grad_norm": 0.453125, + "learning_rate": 0.0001517803602953692, + "loss": 2.5763, + "step": 15403 + }, + { + "epoch": 0.66, + "grad_norm": 0.45703125, + "learning_rate": 0.00015177462154667134, + "loss": 2.4635, + "step": 15404 + }, + { + "epoch": 0.66, + "grad_norm": 0.458984375, + "learning_rate": 0.00015176888256500663, + "loss": 2.5825, + "step": 15405 + }, + { + "epoch": 0.66, + "grad_norm": 0.494140625, + "learning_rate": 0.00015176314335040085, + "loss": 2.5722, + "step": 15406 + }, + { + "epoch": 0.66, + "grad_norm": 0.453125, + "learning_rate": 0.00015175740390287981, + "loss": 2.6399, + "step": 15407 + }, + { + "epoch": 0.66, + "grad_norm": 0.439453125, + "learning_rate": 0.00015175166422246942, + "loss": 2.5509, + "step": 15408 + }, + { + "epoch": 0.66, + "grad_norm": 0.50390625, + "learning_rate": 0.00015174592430919543, + "loss": 2.6099, + "step": 15409 + }, + { + "epoch": 0.66, + "grad_norm": 0.478515625, + "learning_rate": 0.00015174018416308366, + "loss": 2.5324, + "step": 15410 + }, + { + "epoch": 0.66, + "grad_norm": 0.490234375, + "learning_rate": 0.00015173444378416005, + "loss": 2.5678, + "step": 15411 + }, + { + "epoch": 0.66, + "grad_norm": 0.48046875, + "learning_rate": 0.00015172870317245026, + "loss": 2.6794, + "step": 15412 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.0001517229623279803, + "loss": 2.6911, + "step": 15413 + }, + { + "epoch": 0.66, + "grad_norm": 0.435546875, + "learning_rate": 0.00015171722125077584, + "loss": 2.5797, + "step": 15414 + }, + { + "epoch": 0.66, + "grad_norm": 0.4453125, + "learning_rate": 0.0001517114799408628, + "loss": 2.6527, + "step": 15415 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.000151705738398267, + "loss": 2.5082, + "step": 15416 + }, + { + "epoch": 0.66, + "grad_norm": 0.45703125, + "learning_rate": 0.00015169999662301427, + "loss": 2.7041, + "step": 15417 + }, + { + "epoch": 0.66, + "grad_norm": 0.47265625, + "learning_rate": 0.00015169425461513048, + "loss": 2.6835, + "step": 15418 + }, + { + "epoch": 0.66, + "grad_norm": 0.494140625, + "learning_rate": 0.00015168851237464144, + "loss": 2.592, + "step": 15419 + }, + { + "epoch": 0.66, + "grad_norm": 0.447265625, + "learning_rate": 0.00015168276990157293, + "loss": 2.5721, + "step": 15420 + }, + { + "epoch": 0.66, + "grad_norm": 0.478515625, + "learning_rate": 0.00015167702719595086, + "loss": 2.612, + "step": 15421 + }, + { + "epoch": 0.66, + "grad_norm": 0.44921875, + "learning_rate": 0.00015167128425780109, + "loss": 2.6225, + "step": 15422 + }, + { + "epoch": 0.66, + "grad_norm": 0.46875, + "learning_rate": 0.0001516655410871494, + "loss": 2.5224, + "step": 15423 + }, + { + "epoch": 0.66, + "grad_norm": 0.466796875, + "learning_rate": 0.00015165979768402165, + "loss": 2.683, + "step": 15424 + }, + { + "epoch": 0.66, + "grad_norm": 0.462890625, + "learning_rate": 0.0001516540540484437, + "loss": 2.6328, + "step": 15425 + }, + { + "epoch": 0.66, + "grad_norm": 0.53515625, + "learning_rate": 0.00015164831018044139, + "loss": 2.6653, + "step": 15426 + }, + { + "epoch": 0.66, + "grad_norm": 0.474609375, + "learning_rate": 0.00015164256608004055, + "loss": 2.6558, + "step": 15427 + }, + { + "epoch": 0.66, + "grad_norm": 0.41796875, + "learning_rate": 0.00015163682174726704, + "loss": 2.419, + "step": 15428 + }, + { + "epoch": 0.66, + "grad_norm": 0.4453125, + "learning_rate": 0.0001516310771821467, + "loss": 2.5989, + "step": 15429 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.00015162533238470538, + "loss": 2.5293, + "step": 15430 + }, + { + "epoch": 0.66, + "grad_norm": 0.447265625, + "learning_rate": 0.00015161958735496891, + "loss": 2.487, + "step": 15431 + }, + { + "epoch": 0.66, + "grad_norm": 0.4140625, + "learning_rate": 0.00015161384209296322, + "loss": 2.5418, + "step": 15432 + }, + { + "epoch": 0.66, + "grad_norm": 0.490234375, + "learning_rate": 0.00015160809659871407, + "loss": 2.6914, + "step": 15433 + }, + { + "epoch": 0.66, + "grad_norm": 0.408203125, + "learning_rate": 0.00015160235087224733, + "loss": 2.6336, + "step": 15434 + }, + { + "epoch": 0.66, + "grad_norm": 0.40234375, + "learning_rate": 0.00015159660491358888, + "loss": 2.5379, + "step": 15435 + }, + { + "epoch": 0.66, + "grad_norm": 0.490234375, + "learning_rate": 0.00015159085872276459, + "loss": 2.5073, + "step": 15436 + }, + { + "epoch": 0.66, + "grad_norm": 0.4609375, + "learning_rate": 0.00015158511229980027, + "loss": 2.6401, + "step": 15437 + }, + { + "epoch": 0.66, + "grad_norm": 0.43359375, + "learning_rate": 0.0001515793656447218, + "loss": 2.5191, + "step": 15438 + }, + { + "epoch": 0.66, + "grad_norm": 0.42578125, + "learning_rate": 0.00015157361875755502, + "loss": 2.632, + "step": 15439 + }, + { + "epoch": 0.66, + "grad_norm": 0.435546875, + "learning_rate": 0.0001515678716383258, + "loss": 2.6826, + "step": 15440 + }, + { + "epoch": 0.66, + "grad_norm": 0.4453125, + "learning_rate": 0.00015156212428706005, + "loss": 2.5483, + "step": 15441 + }, + { + "epoch": 0.66, + "grad_norm": 0.435546875, + "learning_rate": 0.00015155637670378352, + "loss": 2.4933, + "step": 15442 + }, + { + "epoch": 0.66, + "grad_norm": 0.50390625, + "learning_rate": 0.00015155062888852217, + "loss": 2.4752, + "step": 15443 + }, + { + "epoch": 0.66, + "grad_norm": 0.482421875, + "learning_rate": 0.00015154488084130182, + "loss": 2.6295, + "step": 15444 + }, + { + "epoch": 0.66, + "grad_norm": 0.4453125, + "learning_rate": 0.00015153913256214837, + "loss": 2.6668, + "step": 15445 + }, + { + "epoch": 0.66, + "grad_norm": 0.421875, + "learning_rate": 0.00015153338405108764, + "loss": 2.4085, + "step": 15446 + }, + { + "epoch": 0.66, + "grad_norm": 0.455078125, + "learning_rate": 0.00015152763530814552, + "loss": 2.6414, + "step": 15447 + }, + { + "epoch": 0.66, + "grad_norm": 0.458984375, + "learning_rate": 0.00015152188633334786, + "loss": 2.5385, + "step": 15448 + }, + { + "epoch": 0.66, + "grad_norm": 0.44921875, + "learning_rate": 0.00015151613712672055, + "loss": 2.5995, + "step": 15449 + }, + { + "epoch": 0.66, + "grad_norm": 0.462890625, + "learning_rate": 0.00015151038768828942, + "loss": 2.6718, + "step": 15450 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.00015150463801808043, + "loss": 2.4744, + "step": 15451 + }, + { + "epoch": 0.66, + "grad_norm": 0.51953125, + "learning_rate": 0.00015149888811611938, + "loss": 2.4558, + "step": 15452 + }, + { + "epoch": 0.66, + "grad_norm": 0.416015625, + "learning_rate": 0.00015149313798243212, + "loss": 2.6322, + "step": 15453 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.00015148738761704456, + "loss": 2.6722, + "step": 15454 + }, + { + "epoch": 0.66, + "grad_norm": 0.4375, + "learning_rate": 0.0001514816370199826, + "loss": 2.7442, + "step": 15455 + }, + { + "epoch": 0.66, + "grad_norm": 0.55859375, + "learning_rate": 0.00015147588619127207, + "loss": 2.7301, + "step": 15456 + }, + { + "epoch": 0.66, + "grad_norm": 0.4765625, + "learning_rate": 0.00015147013513093887, + "loss": 2.7124, + "step": 15457 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.00015146438383900886, + "loss": 2.6178, + "step": 15458 + }, + { + "epoch": 0.66, + "grad_norm": 0.42578125, + "learning_rate": 0.00015145863231550796, + "loss": 2.5373, + "step": 15459 + }, + { + "epoch": 0.66, + "grad_norm": 0.3984375, + "learning_rate": 0.00015145288056046198, + "loss": 2.5333, + "step": 15460 + }, + { + "epoch": 0.66, + "grad_norm": 0.546875, + "learning_rate": 0.00015144712857389684, + "loss": 2.6727, + "step": 15461 + }, + { + "epoch": 0.66, + "grad_norm": 0.447265625, + "learning_rate": 0.0001514413763558385, + "loss": 2.5688, + "step": 15462 + }, + { + "epoch": 0.66, + "grad_norm": 0.423828125, + "learning_rate": 0.00015143562390631268, + "loss": 2.5512, + "step": 15463 + }, + { + "epoch": 0.66, + "grad_norm": 0.4609375, + "learning_rate": 0.00015142987122534538, + "loss": 2.4712, + "step": 15464 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.00015142411831296245, + "loss": 2.5599, + "step": 15465 + }, + { + "epoch": 0.66, + "grad_norm": 0.431640625, + "learning_rate": 0.0001514183651691898, + "loss": 2.6231, + "step": 15466 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.0001514126117940533, + "loss": 2.6359, + "step": 15467 + }, + { + "epoch": 0.66, + "grad_norm": 0.443359375, + "learning_rate": 0.00015140685818757881, + "loss": 2.5661, + "step": 15468 + }, + { + "epoch": 0.66, + "grad_norm": 0.486328125, + "learning_rate": 0.00015140110434979224, + "loss": 2.6466, + "step": 15469 + }, + { + "epoch": 0.66, + "grad_norm": 0.453125, + "learning_rate": 0.0001513953502807195, + "loss": 2.5708, + "step": 15470 + }, + { + "epoch": 0.66, + "grad_norm": 0.498046875, + "learning_rate": 0.00015138959598038647, + "loss": 2.6746, + "step": 15471 + }, + { + "epoch": 0.66, + "grad_norm": 0.51171875, + "learning_rate": 0.00015138384144881904, + "loss": 2.6877, + "step": 15472 + }, + { + "epoch": 0.66, + "grad_norm": 0.5234375, + "learning_rate": 0.0001513780866860431, + "loss": 2.6008, + "step": 15473 + }, + { + "epoch": 0.66, + "grad_norm": 0.515625, + "learning_rate": 0.00015137233169208454, + "loss": 2.6139, + "step": 15474 + }, + { + "epoch": 0.66, + "grad_norm": 0.48828125, + "learning_rate": 0.0001513665764669693, + "loss": 2.5343, + "step": 15475 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.00015136082101072318, + "loss": 2.6254, + "step": 15476 + }, + { + "epoch": 0.66, + "grad_norm": 0.4375, + "learning_rate": 0.00015135506532337216, + "loss": 2.6858, + "step": 15477 + }, + { + "epoch": 0.66, + "grad_norm": 0.462890625, + "learning_rate": 0.00015134930940494213, + "loss": 2.5105, + "step": 15478 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.00015134355325545896, + "loss": 2.4973, + "step": 15479 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.00015133779687494855, + "loss": 2.7602, + "step": 15480 + }, + { + "epoch": 0.66, + "grad_norm": 0.421875, + "learning_rate": 0.00015133204026343682, + "loss": 2.4793, + "step": 15481 + }, + { + "epoch": 0.66, + "grad_norm": 0.419921875, + "learning_rate": 0.00015132628342094968, + "loss": 2.4648, + "step": 15482 + }, + { + "epoch": 0.66, + "grad_norm": 0.455078125, + "learning_rate": 0.00015132052634751303, + "loss": 2.493, + "step": 15483 + }, + { + "epoch": 0.66, + "grad_norm": 0.486328125, + "learning_rate": 0.00015131476904315275, + "loss": 2.5607, + "step": 15484 + }, + { + "epoch": 0.66, + "grad_norm": 0.44140625, + "learning_rate": 0.0001513090115078948, + "loss": 2.5295, + "step": 15485 + }, + { + "epoch": 0.66, + "grad_norm": 0.4296875, + "learning_rate": 0.000151303253741765, + "loss": 2.5694, + "step": 15486 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.00015129749574478932, + "loss": 2.5747, + "step": 15487 + }, + { + "epoch": 0.66, + "grad_norm": 0.40234375, + "learning_rate": 0.00015129173751699367, + "loss": 2.4678, + "step": 15488 + }, + { + "epoch": 0.66, + "grad_norm": 0.486328125, + "learning_rate": 0.00015128597905840393, + "loss": 2.4536, + "step": 15489 + }, + { + "epoch": 0.66, + "grad_norm": 0.419921875, + "learning_rate": 0.00015128022036904604, + "loss": 2.5839, + "step": 15490 + }, + { + "epoch": 0.66, + "grad_norm": 0.443359375, + "learning_rate": 0.0001512744614489459, + "loss": 2.4734, + "step": 15491 + }, + { + "epoch": 0.66, + "grad_norm": 0.466796875, + "learning_rate": 0.0001512687022981294, + "loss": 2.7022, + "step": 15492 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.0001512629429166225, + "loss": 2.4915, + "step": 15493 + }, + { + "epoch": 0.66, + "grad_norm": 0.42578125, + "learning_rate": 0.0001512571833044511, + "loss": 2.5077, + "step": 15494 + }, + { + "epoch": 0.66, + "grad_norm": 0.44921875, + "learning_rate": 0.00015125142346164108, + "loss": 2.6201, + "step": 15495 + }, + { + "epoch": 0.66, + "grad_norm": 0.474609375, + "learning_rate": 0.0001512456633882184, + "loss": 2.6839, + "step": 15496 + }, + { + "epoch": 0.66, + "grad_norm": 0.498046875, + "learning_rate": 0.00015123990308420896, + "loss": 2.485, + "step": 15497 + }, + { + "epoch": 0.66, + "grad_norm": 0.447265625, + "learning_rate": 0.0001512341425496387, + "loss": 2.5361, + "step": 15498 + }, + { + "epoch": 0.66, + "grad_norm": 0.443359375, + "learning_rate": 0.0001512283817845335, + "loss": 2.5363, + "step": 15499 + }, + { + "epoch": 0.66, + "grad_norm": 0.46875, + "learning_rate": 0.00015122262078891928, + "loss": 2.5409, + "step": 15500 + }, + { + "epoch": 0.66, + "grad_norm": 0.482421875, + "learning_rate": 0.00015121685956282204, + "loss": 2.6126, + "step": 15501 + }, + { + "epoch": 0.66, + "grad_norm": 0.4375, + "learning_rate": 0.0001512110981062676, + "loss": 2.5606, + "step": 15502 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.00015120533641928199, + "loss": 2.5284, + "step": 15503 + }, + { + "epoch": 0.66, + "grad_norm": 0.451171875, + "learning_rate": 0.00015119957450189105, + "loss": 2.5607, + "step": 15504 + }, + { + "epoch": 0.66, + "grad_norm": 0.51171875, + "learning_rate": 0.00015119381235412073, + "loss": 2.6827, + "step": 15505 + }, + { + "epoch": 0.66, + "grad_norm": 0.43359375, + "learning_rate": 0.00015118804997599694, + "loss": 2.5419, + "step": 15506 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.0001511822873675457, + "loss": 2.5729, + "step": 15507 + }, + { + "epoch": 0.66, + "grad_norm": 0.4453125, + "learning_rate": 0.0001511765245287928, + "loss": 2.6732, + "step": 15508 + }, + { + "epoch": 0.66, + "grad_norm": 0.53125, + "learning_rate": 0.00015117076145976428, + "loss": 2.613, + "step": 15509 + }, + { + "epoch": 0.66, + "grad_norm": 0.482421875, + "learning_rate": 0.00015116499816048603, + "loss": 2.5779, + "step": 15510 + }, + { + "epoch": 0.66, + "grad_norm": 0.4453125, + "learning_rate": 0.000151159234630984, + "loss": 2.4456, + "step": 15511 + }, + { + "epoch": 0.66, + "grad_norm": 0.498046875, + "learning_rate": 0.00015115347087128408, + "loss": 2.6719, + "step": 15512 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.00015114770688141226, + "loss": 2.642, + "step": 15513 + }, + { + "epoch": 0.66, + "grad_norm": 0.4140625, + "learning_rate": 0.00015114194266139443, + "loss": 2.5181, + "step": 15514 + }, + { + "epoch": 0.66, + "grad_norm": 0.48828125, + "learning_rate": 0.00015113617821125658, + "loss": 2.6265, + "step": 15515 + }, + { + "epoch": 0.66, + "grad_norm": 0.439453125, + "learning_rate": 0.0001511304135310246, + "loss": 2.5654, + "step": 15516 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.00015112464862072442, + "loss": 2.6109, + "step": 15517 + }, + { + "epoch": 0.66, + "grad_norm": 0.458984375, + "learning_rate": 0.00015111888348038206, + "loss": 2.5752, + "step": 15518 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.00015111311811002334, + "loss": 2.6509, + "step": 15519 + }, + { + "epoch": 0.66, + "grad_norm": 0.435546875, + "learning_rate": 0.00015110735250967433, + "loss": 2.5403, + "step": 15520 + }, + { + "epoch": 0.66, + "grad_norm": 0.4296875, + "learning_rate": 0.00015110158667936086, + "loss": 2.6761, + "step": 15521 + }, + { + "epoch": 0.66, + "grad_norm": 0.42578125, + "learning_rate": 0.00015109582061910895, + "loss": 2.4423, + "step": 15522 + }, + { + "epoch": 0.66, + "grad_norm": 0.48828125, + "learning_rate": 0.00015109005432894455, + "loss": 2.6235, + "step": 15523 + }, + { + "epoch": 0.66, + "grad_norm": 0.427734375, + "learning_rate": 0.00015108428780889353, + "loss": 2.6598, + "step": 15524 + }, + { + "epoch": 0.66, + "grad_norm": 0.42578125, + "learning_rate": 0.00015107852105898187, + "loss": 2.5894, + "step": 15525 + }, + { + "epoch": 0.66, + "grad_norm": 0.431640625, + "learning_rate": 0.0001510727540792356, + "loss": 2.5833, + "step": 15526 + }, + { + "epoch": 0.66, + "grad_norm": 0.474609375, + "learning_rate": 0.00015106698686968053, + "loss": 2.5585, + "step": 15527 + }, + { + "epoch": 0.66, + "grad_norm": 0.50390625, + "learning_rate": 0.00015106121943034273, + "loss": 2.4528, + "step": 15528 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.0001510554517612481, + "loss": 2.5135, + "step": 15529 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.00015104968386242258, + "loss": 2.6375, + "step": 15530 + }, + { + "epoch": 0.66, + "grad_norm": 0.427734375, + "learning_rate": 0.00015104391573389213, + "loss": 2.4902, + "step": 15531 + }, + { + "epoch": 0.66, + "grad_norm": 0.50390625, + "learning_rate": 0.00015103814737568274, + "loss": 2.5304, + "step": 15532 + }, + { + "epoch": 0.66, + "grad_norm": 0.50390625, + "learning_rate": 0.00015103237878782034, + "loss": 2.4552, + "step": 15533 + }, + { + "epoch": 0.66, + "grad_norm": 0.45703125, + "learning_rate": 0.00015102660997033088, + "loss": 2.6984, + "step": 15534 + }, + { + "epoch": 0.66, + "grad_norm": 0.423828125, + "learning_rate": 0.0001510208409232403, + "loss": 2.6217, + "step": 15535 + }, + { + "epoch": 0.66, + "grad_norm": 0.4609375, + "learning_rate": 0.00015101507164657458, + "loss": 2.5562, + "step": 15536 + }, + { + "epoch": 0.66, + "grad_norm": 0.490234375, + "learning_rate": 0.0001510093021403597, + "loss": 2.6774, + "step": 15537 + }, + { + "epoch": 0.66, + "grad_norm": 0.462890625, + "learning_rate": 0.00015100353240462161, + "loss": 2.5223, + "step": 15538 + }, + { + "epoch": 0.66, + "grad_norm": 0.48046875, + "learning_rate": 0.00015099776243938625, + "loss": 2.58, + "step": 15539 + }, + { + "epoch": 0.66, + "grad_norm": 0.42578125, + "learning_rate": 0.00015099199224467963, + "loss": 2.7476, + "step": 15540 + }, + { + "epoch": 0.66, + "grad_norm": 0.498046875, + "learning_rate": 0.0001509862218205276, + "loss": 2.6727, + "step": 15541 + }, + { + "epoch": 0.66, + "grad_norm": 0.447265625, + "learning_rate": 0.0001509804511669563, + "loss": 2.6026, + "step": 15542 + }, + { + "epoch": 0.66, + "grad_norm": 0.443359375, + "learning_rate": 0.00015097468028399153, + "loss": 2.5761, + "step": 15543 + }, + { + "epoch": 0.66, + "grad_norm": 0.443359375, + "learning_rate": 0.00015096890917165937, + "loss": 2.5934, + "step": 15544 + }, + { + "epoch": 0.66, + "grad_norm": 0.470703125, + "learning_rate": 0.00015096313782998576, + "loss": 2.5523, + "step": 15545 + }, + { + "epoch": 0.66, + "grad_norm": 0.498046875, + "learning_rate": 0.0001509573662589966, + "loss": 2.653, + "step": 15546 + }, + { + "epoch": 0.66, + "grad_norm": 0.427734375, + "learning_rate": 0.00015095159445871795, + "loss": 2.527, + "step": 15547 + }, + { + "epoch": 0.66, + "grad_norm": 0.4453125, + "learning_rate": 0.00015094582242917572, + "loss": 2.6781, + "step": 15548 + }, + { + "epoch": 0.66, + "grad_norm": 0.486328125, + "learning_rate": 0.00015094005017039593, + "loss": 2.6714, + "step": 15549 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.00015093427768240452, + "loss": 2.6925, + "step": 15550 + }, + { + "epoch": 0.66, + "grad_norm": 0.486328125, + "learning_rate": 0.00015092850496522747, + "loss": 2.6948, + "step": 15551 + }, + { + "epoch": 0.66, + "grad_norm": 0.423828125, + "learning_rate": 0.00015092273201889076, + "loss": 2.4948, + "step": 15552 + }, + { + "epoch": 0.66, + "grad_norm": 0.4140625, + "learning_rate": 0.0001509169588434204, + "loss": 2.4757, + "step": 15553 + }, + { + "epoch": 0.66, + "grad_norm": 0.404296875, + "learning_rate": 0.0001509111854388423, + "loss": 2.5359, + "step": 15554 + }, + { + "epoch": 0.66, + "grad_norm": 0.416015625, + "learning_rate": 0.0001509054118051825, + "loss": 2.5566, + "step": 15555 + }, + { + "epoch": 0.66, + "grad_norm": 0.482421875, + "learning_rate": 0.00015089963794246694, + "loss": 2.5181, + "step": 15556 + }, + { + "epoch": 0.66, + "grad_norm": 0.4921875, + "learning_rate": 0.00015089386385072157, + "loss": 2.6829, + "step": 15557 + }, + { + "epoch": 0.66, + "grad_norm": 0.427734375, + "learning_rate": 0.00015088808952997248, + "loss": 2.5844, + "step": 15558 + }, + { + "epoch": 0.66, + "grad_norm": 0.47265625, + "learning_rate": 0.00015088231498024552, + "loss": 2.4985, + "step": 15559 + }, + { + "epoch": 0.66, + "grad_norm": 0.4765625, + "learning_rate": 0.00015087654020156676, + "loss": 2.6285, + "step": 15560 + }, + { + "epoch": 0.66, + "grad_norm": 0.455078125, + "learning_rate": 0.00015087076519396217, + "loss": 2.5973, + "step": 15561 + }, + { + "epoch": 0.66, + "grad_norm": 0.5, + "learning_rate": 0.00015086498995745774, + "loss": 2.6126, + "step": 15562 + }, + { + "epoch": 0.66, + "grad_norm": 0.447265625, + "learning_rate": 0.00015085921449207944, + "loss": 2.7452, + "step": 15563 + }, + { + "epoch": 0.66, + "grad_norm": 0.49609375, + "learning_rate": 0.00015085343879785325, + "loss": 2.502, + "step": 15564 + }, + { + "epoch": 0.66, + "grad_norm": 0.44921875, + "learning_rate": 0.00015084766287480514, + "loss": 2.6984, + "step": 15565 + }, + { + "epoch": 0.66, + "grad_norm": 0.490234375, + "learning_rate": 0.0001508418867229612, + "loss": 2.4821, + "step": 15566 + }, + { + "epoch": 0.66, + "grad_norm": 0.458984375, + "learning_rate": 0.00015083611034234727, + "loss": 2.6012, + "step": 15567 + }, + { + "epoch": 0.66, + "grad_norm": 0.421875, + "learning_rate": 0.0001508303337329895, + "loss": 2.5333, + "step": 15568 + }, + { + "epoch": 0.66, + "grad_norm": 0.50390625, + "learning_rate": 0.00015082455689491376, + "loss": 2.5094, + "step": 15569 + }, + { + "epoch": 0.66, + "grad_norm": 0.49609375, + "learning_rate": 0.0001508187798281461, + "loss": 2.5803, + "step": 15570 + }, + { + "epoch": 0.66, + "grad_norm": 0.40625, + "learning_rate": 0.00015081300253271248, + "loss": 2.7362, + "step": 15571 + }, + { + "epoch": 0.66, + "grad_norm": 0.443359375, + "learning_rate": 0.00015080722500863898, + "loss": 2.6098, + "step": 15572 + }, + { + "epoch": 0.66, + "grad_norm": 0.421875, + "learning_rate": 0.00015080144725595145, + "loss": 2.6102, + "step": 15573 + }, + { + "epoch": 0.66, + "grad_norm": 0.490234375, + "learning_rate": 0.00015079566927467605, + "loss": 2.6328, + "step": 15574 + }, + { + "epoch": 0.66, + "grad_norm": 0.455078125, + "learning_rate": 0.00015078989106483866, + "loss": 2.6276, + "step": 15575 + }, + { + "epoch": 0.66, + "grad_norm": 0.435546875, + "learning_rate": 0.00015078411262646534, + "loss": 2.5199, + "step": 15576 + }, + { + "epoch": 0.66, + "grad_norm": 0.4609375, + "learning_rate": 0.00015077833395958206, + "loss": 2.6496, + "step": 15577 + }, + { + "epoch": 0.66, + "grad_norm": 0.4140625, + "learning_rate": 0.00015077255506421483, + "loss": 2.7286, + "step": 15578 + }, + { + "epoch": 0.66, + "grad_norm": 0.48046875, + "learning_rate": 0.00015076677594038967, + "loss": 2.5372, + "step": 15579 + }, + { + "epoch": 0.66, + "grad_norm": 0.462890625, + "learning_rate": 0.00015076099658813257, + "loss": 2.5249, + "step": 15580 + }, + { + "epoch": 0.66, + "grad_norm": 0.427734375, + "learning_rate": 0.00015075521700746953, + "loss": 2.6349, + "step": 15581 + }, + { + "epoch": 0.66, + "grad_norm": 0.474609375, + "learning_rate": 0.00015074943719842657, + "loss": 2.642, + "step": 15582 + }, + { + "epoch": 0.66, + "grad_norm": 0.4609375, + "learning_rate": 0.00015074365716102967, + "loss": 2.539, + "step": 15583 + }, + { + "epoch": 0.66, + "grad_norm": 0.52734375, + "learning_rate": 0.0001507378768953049, + "loss": 2.6149, + "step": 15584 + }, + { + "epoch": 0.66, + "grad_norm": 0.431640625, + "learning_rate": 0.0001507320964012782, + "loss": 2.596, + "step": 15585 + }, + { + "epoch": 0.66, + "grad_norm": 0.44140625, + "learning_rate": 0.0001507263156789756, + "loss": 2.6786, + "step": 15586 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.0001507205347284231, + "loss": 2.5714, + "step": 15587 + }, + { + "epoch": 0.66, + "grad_norm": 0.4453125, + "learning_rate": 0.0001507147535496468, + "loss": 2.5198, + "step": 15588 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.00015070897214267256, + "loss": 2.68, + "step": 15589 + }, + { + "epoch": 0.66, + "grad_norm": 0.486328125, + "learning_rate": 0.00015070319050752655, + "loss": 2.613, + "step": 15590 + }, + { + "epoch": 0.66, + "grad_norm": 0.453125, + "learning_rate": 0.00015069740864423465, + "loss": 2.7315, + "step": 15591 + }, + { + "epoch": 0.66, + "grad_norm": 0.478515625, + "learning_rate": 0.00015069162655282297, + "loss": 2.6463, + "step": 15592 + }, + { + "epoch": 0.66, + "grad_norm": 0.4609375, + "learning_rate": 0.0001506858442333175, + "loss": 2.6016, + "step": 15593 + }, + { + "epoch": 0.66, + "grad_norm": 0.4609375, + "learning_rate": 0.00015068006168574423, + "loss": 2.6573, + "step": 15594 + }, + { + "epoch": 0.66, + "grad_norm": 0.423828125, + "learning_rate": 0.00015067427891012923, + "loss": 2.5357, + "step": 15595 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.00015066849590649847, + "loss": 2.592, + "step": 15596 + }, + { + "epoch": 0.66, + "grad_norm": 0.46875, + "learning_rate": 0.00015066271267487802, + "loss": 2.5926, + "step": 15597 + }, + { + "epoch": 0.66, + "grad_norm": 0.455078125, + "learning_rate": 0.0001506569292152939, + "loss": 2.5968, + "step": 15598 + }, + { + "epoch": 0.66, + "grad_norm": 0.482421875, + "learning_rate": 0.00015065114552777203, + "loss": 2.769, + "step": 15599 + }, + { + "epoch": 0.66, + "grad_norm": 0.486328125, + "learning_rate": 0.00015064536161233858, + "loss": 2.7032, + "step": 15600 + }, + { + "epoch": 0.66, + "grad_norm": 0.46484375, + "learning_rate": 0.0001506395774690195, + "loss": 2.5484, + "step": 15601 + }, + { + "epoch": 0.66, + "grad_norm": 0.5078125, + "learning_rate": 0.00015063379309784085, + "loss": 2.6285, + "step": 15602 + }, + { + "epoch": 0.66, + "grad_norm": 0.48828125, + "learning_rate": 0.0001506280084988286, + "loss": 2.6594, + "step": 15603 + }, + { + "epoch": 0.66, + "grad_norm": 0.361328125, + "learning_rate": 0.00015062222367200883, + "loss": 2.5325, + "step": 15604 + }, + { + "epoch": 0.66, + "grad_norm": 0.453125, + "learning_rate": 0.00015061643861740754, + "loss": 2.4871, + "step": 15605 + }, + { + "epoch": 0.67, + "grad_norm": 0.408203125, + "learning_rate": 0.00015061065333505077, + "loss": 2.6058, + "step": 15606 + }, + { + "epoch": 0.67, + "grad_norm": 0.484375, + "learning_rate": 0.00015060486782496457, + "loss": 2.6674, + "step": 15607 + }, + { + "epoch": 0.67, + "grad_norm": 0.462890625, + "learning_rate": 0.00015059908208717496, + "loss": 2.6412, + "step": 15608 + }, + { + "epoch": 0.67, + "grad_norm": 0.44140625, + "learning_rate": 0.000150593296121708, + "loss": 2.5569, + "step": 15609 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.00015058750992858966, + "loss": 2.4813, + "step": 15610 + }, + { + "epoch": 0.67, + "grad_norm": 0.4921875, + "learning_rate": 0.00015058172350784603, + "loss": 2.5193, + "step": 15611 + }, + { + "epoch": 0.67, + "grad_norm": 0.40234375, + "learning_rate": 0.0001505759368595031, + "loss": 2.5917, + "step": 15612 + }, + { + "epoch": 0.67, + "grad_norm": 0.51953125, + "learning_rate": 0.00015057014998358693, + "loss": 2.6235, + "step": 15613 + }, + { + "epoch": 0.67, + "grad_norm": 0.47265625, + "learning_rate": 0.00015056436288012363, + "loss": 2.6202, + "step": 15614 + }, + { + "epoch": 0.67, + "grad_norm": 0.46484375, + "learning_rate": 0.00015055857554913913, + "loss": 2.6315, + "step": 15615 + }, + { + "epoch": 0.67, + "grad_norm": 0.47265625, + "learning_rate": 0.00015055278799065952, + "loss": 2.6906, + "step": 15616 + }, + { + "epoch": 0.67, + "grad_norm": 0.4921875, + "learning_rate": 0.00015054700020471088, + "loss": 2.4289, + "step": 15617 + }, + { + "epoch": 0.67, + "grad_norm": 0.4921875, + "learning_rate": 0.00015054121219131917, + "loss": 2.5845, + "step": 15618 + }, + { + "epoch": 0.67, + "grad_norm": 0.45703125, + "learning_rate": 0.00015053542395051047, + "loss": 2.6555, + "step": 15619 + }, + { + "epoch": 0.67, + "grad_norm": 0.4453125, + "learning_rate": 0.00015052963548231084, + "loss": 2.5138, + "step": 15620 + }, + { + "epoch": 0.67, + "grad_norm": 0.423828125, + "learning_rate": 0.0001505238467867463, + "loss": 2.6626, + "step": 15621 + }, + { + "epoch": 0.67, + "grad_norm": 0.44140625, + "learning_rate": 0.000150518057863843, + "loss": 2.6217, + "step": 15622 + }, + { + "epoch": 0.67, + "grad_norm": 0.462890625, + "learning_rate": 0.00015051226871362681, + "loss": 2.6265, + "step": 15623 + }, + { + "epoch": 0.67, + "grad_norm": 0.45703125, + "learning_rate": 0.0001505064793361239, + "loss": 2.5257, + "step": 15624 + }, + { + "epoch": 0.67, + "grad_norm": 0.431640625, + "learning_rate": 0.0001505006897313603, + "loss": 2.6068, + "step": 15625 + }, + { + "epoch": 0.67, + "grad_norm": 0.3828125, + "learning_rate": 0.00015049489989936202, + "loss": 2.4978, + "step": 15626 + }, + { + "epoch": 0.67, + "grad_norm": 0.439453125, + "learning_rate": 0.00015048910984015516, + "loss": 2.5226, + "step": 15627 + }, + { + "epoch": 0.67, + "grad_norm": 0.44921875, + "learning_rate": 0.00015048331955376578, + "loss": 2.5418, + "step": 15628 + }, + { + "epoch": 0.67, + "grad_norm": 0.490234375, + "learning_rate": 0.00015047752904021988, + "loss": 2.6474, + "step": 15629 + }, + { + "epoch": 0.67, + "grad_norm": 0.4296875, + "learning_rate": 0.00015047173829954357, + "loss": 2.6456, + "step": 15630 + }, + { + "epoch": 0.67, + "grad_norm": 0.474609375, + "learning_rate": 0.00015046594733176288, + "loss": 2.6413, + "step": 15631 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.00015046015613690385, + "loss": 2.7432, + "step": 15632 + }, + { + "epoch": 0.67, + "grad_norm": 0.421875, + "learning_rate": 0.00015045436471499262, + "loss": 2.5012, + "step": 15633 + }, + { + "epoch": 0.67, + "grad_norm": 0.48828125, + "learning_rate": 0.0001504485730660551, + "loss": 2.6255, + "step": 15634 + }, + { + "epoch": 0.67, + "grad_norm": 0.4609375, + "learning_rate": 0.0001504427811901175, + "loss": 2.7498, + "step": 15635 + }, + { + "epoch": 0.67, + "grad_norm": 0.443359375, + "learning_rate": 0.0001504369890872058, + "loss": 2.5613, + "step": 15636 + }, + { + "epoch": 0.67, + "grad_norm": 0.46875, + "learning_rate": 0.00015043119675734606, + "loss": 2.54, + "step": 15637 + }, + { + "epoch": 0.67, + "grad_norm": 0.416015625, + "learning_rate": 0.0001504254042005644, + "loss": 2.4562, + "step": 15638 + }, + { + "epoch": 0.67, + "grad_norm": 0.443359375, + "learning_rate": 0.00015041961141688684, + "loss": 2.4698, + "step": 15639 + }, + { + "epoch": 0.67, + "grad_norm": 0.484375, + "learning_rate": 0.0001504138184063394, + "loss": 2.6713, + "step": 15640 + }, + { + "epoch": 0.67, + "grad_norm": 0.515625, + "learning_rate": 0.0001504080251689483, + "loss": 2.6147, + "step": 15641 + }, + { + "epoch": 0.67, + "grad_norm": 0.447265625, + "learning_rate": 0.00015040223170473946, + "loss": 2.5775, + "step": 15642 + }, + { + "epoch": 0.67, + "grad_norm": 0.494140625, + "learning_rate": 0.00015039643801373898, + "loss": 2.6472, + "step": 15643 + }, + { + "epoch": 0.67, + "grad_norm": 0.4609375, + "learning_rate": 0.00015039064409597296, + "loss": 2.6208, + "step": 15644 + }, + { + "epoch": 0.67, + "grad_norm": 0.462890625, + "learning_rate": 0.00015038484995146746, + "loss": 2.5423, + "step": 15645 + }, + { + "epoch": 0.67, + "grad_norm": 0.490234375, + "learning_rate": 0.00015037905558024857, + "loss": 2.5528, + "step": 15646 + }, + { + "epoch": 0.67, + "grad_norm": 0.494140625, + "learning_rate": 0.00015037326098234232, + "loss": 2.5996, + "step": 15647 + }, + { + "epoch": 0.67, + "grad_norm": 0.462890625, + "learning_rate": 0.0001503674661577748, + "loss": 2.4844, + "step": 15648 + }, + { + "epoch": 0.67, + "grad_norm": 0.408203125, + "learning_rate": 0.00015036167110657212, + "loss": 2.6257, + "step": 15649 + }, + { + "epoch": 0.67, + "grad_norm": 0.42578125, + "learning_rate": 0.00015035587582876028, + "loss": 2.6562, + "step": 15650 + }, + { + "epoch": 0.67, + "grad_norm": 0.48046875, + "learning_rate": 0.00015035008032436543, + "loss": 2.542, + "step": 15651 + }, + { + "epoch": 0.67, + "grad_norm": 0.5390625, + "learning_rate": 0.0001503442845934136, + "loss": 2.5704, + "step": 15652 + }, + { + "epoch": 0.67, + "grad_norm": 0.435546875, + "learning_rate": 0.0001503384886359309, + "loss": 2.5434, + "step": 15653 + }, + { + "epoch": 0.67, + "grad_norm": 0.470703125, + "learning_rate": 0.0001503326924519434, + "loss": 2.5546, + "step": 15654 + }, + { + "epoch": 0.67, + "grad_norm": 0.466796875, + "learning_rate": 0.00015032689604147718, + "loss": 2.6456, + "step": 15655 + }, + { + "epoch": 0.67, + "grad_norm": 0.447265625, + "learning_rate": 0.0001503210994045583, + "loss": 2.609, + "step": 15656 + }, + { + "epoch": 0.67, + "grad_norm": 0.498046875, + "learning_rate": 0.0001503153025412129, + "loss": 2.546, + "step": 15657 + }, + { + "epoch": 0.67, + "grad_norm": 0.427734375, + "learning_rate": 0.00015030950545146698, + "loss": 2.5646, + "step": 15658 + }, + { + "epoch": 0.67, + "grad_norm": 0.48828125, + "learning_rate": 0.0001503037081353467, + "loss": 2.5821, + "step": 15659 + }, + { + "epoch": 0.67, + "grad_norm": 0.4375, + "learning_rate": 0.00015029791059287812, + "loss": 2.5957, + "step": 15660 + }, + { + "epoch": 0.67, + "grad_norm": 0.494140625, + "learning_rate": 0.00015029211282408732, + "loss": 2.5433, + "step": 15661 + }, + { + "epoch": 0.67, + "grad_norm": 0.439453125, + "learning_rate": 0.0001502863148290004, + "loss": 2.5073, + "step": 15662 + }, + { + "epoch": 0.67, + "grad_norm": 0.4296875, + "learning_rate": 0.00015028051660764344, + "loss": 2.7236, + "step": 15663 + }, + { + "epoch": 0.67, + "grad_norm": 0.49609375, + "learning_rate": 0.0001502747181600425, + "loss": 2.4741, + "step": 15664 + }, + { + "epoch": 0.67, + "grad_norm": 0.43359375, + "learning_rate": 0.00015026891948622374, + "loss": 2.6135, + "step": 15665 + }, + { + "epoch": 0.67, + "grad_norm": 0.51171875, + "learning_rate": 0.0001502631205862132, + "loss": 2.6763, + "step": 15666 + }, + { + "epoch": 0.67, + "grad_norm": 0.4296875, + "learning_rate": 0.00015025732146003697, + "loss": 2.4146, + "step": 15667 + }, + { + "epoch": 0.67, + "grad_norm": 0.470703125, + "learning_rate": 0.0001502515221077212, + "loss": 2.618, + "step": 15668 + }, + { + "epoch": 0.67, + "grad_norm": 0.435546875, + "learning_rate": 0.0001502457225292919, + "loss": 2.5216, + "step": 15669 + }, + { + "epoch": 0.67, + "grad_norm": 0.404296875, + "learning_rate": 0.00015023992272477523, + "loss": 2.5702, + "step": 15670 + }, + { + "epoch": 0.67, + "grad_norm": 0.53515625, + "learning_rate": 0.0001502341226941973, + "loss": 2.7462, + "step": 15671 + }, + { + "epoch": 0.67, + "grad_norm": 0.40234375, + "learning_rate": 0.0001502283224375841, + "loss": 2.552, + "step": 15672 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.00015022252195496188, + "loss": 2.5786, + "step": 15673 + }, + { + "epoch": 0.67, + "grad_norm": 0.421875, + "learning_rate": 0.0001502167212463566, + "loss": 2.4554, + "step": 15674 + }, + { + "epoch": 0.67, + "grad_norm": 0.44140625, + "learning_rate": 0.00015021092031179448, + "loss": 2.5963, + "step": 15675 + }, + { + "epoch": 0.67, + "grad_norm": 0.48828125, + "learning_rate": 0.00015020511915130155, + "loss": 2.6646, + "step": 15676 + }, + { + "epoch": 0.67, + "grad_norm": 0.443359375, + "learning_rate": 0.0001501993177649039, + "loss": 2.5992, + "step": 15677 + }, + { + "epoch": 0.67, + "grad_norm": 0.40234375, + "learning_rate": 0.00015019351615262774, + "loss": 2.4482, + "step": 15678 + }, + { + "epoch": 0.67, + "grad_norm": 0.482421875, + "learning_rate": 0.00015018771431449905, + "loss": 2.4056, + "step": 15679 + }, + { + "epoch": 0.67, + "grad_norm": 0.41015625, + "learning_rate": 0.00015018191225054398, + "loss": 2.504, + "step": 15680 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.0001501761099607887, + "loss": 2.593, + "step": 15681 + }, + { + "epoch": 0.67, + "grad_norm": 0.4375, + "learning_rate": 0.00015017030744525923, + "loss": 2.5606, + "step": 15682 + }, + { + "epoch": 0.67, + "grad_norm": 0.46484375, + "learning_rate": 0.0001501645047039817, + "loss": 2.5467, + "step": 15683 + }, + { + "epoch": 0.67, + "grad_norm": 0.482421875, + "learning_rate": 0.00015015870173698222, + "loss": 2.6238, + "step": 15684 + }, + { + "epoch": 0.67, + "grad_norm": 0.439453125, + "learning_rate": 0.00015015289854428693, + "loss": 2.6799, + "step": 15685 + }, + { + "epoch": 0.67, + "grad_norm": 0.466796875, + "learning_rate": 0.00015014709512592192, + "loss": 2.5334, + "step": 15686 + }, + { + "epoch": 0.67, + "grad_norm": 0.41015625, + "learning_rate": 0.00015014129148191332, + "loss": 2.5524, + "step": 15687 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.00015013548761228722, + "loss": 2.6565, + "step": 15688 + }, + { + "epoch": 0.67, + "grad_norm": 0.478515625, + "learning_rate": 0.00015012968351706978, + "loss": 2.6512, + "step": 15689 + }, + { + "epoch": 0.67, + "grad_norm": 0.46484375, + "learning_rate": 0.00015012387919628703, + "loss": 2.4944, + "step": 15690 + }, + { + "epoch": 0.67, + "grad_norm": 0.42578125, + "learning_rate": 0.00015011807464996517, + "loss": 2.5876, + "step": 15691 + }, + { + "epoch": 0.67, + "grad_norm": 0.439453125, + "learning_rate": 0.00015011226987813033, + "loss": 2.6682, + "step": 15692 + }, + { + "epoch": 0.67, + "grad_norm": 0.52734375, + "learning_rate": 0.00015010646488080852, + "loss": 2.6236, + "step": 15693 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.00015010065965802597, + "loss": 2.6283, + "step": 15694 + }, + { + "epoch": 0.67, + "grad_norm": 0.490234375, + "learning_rate": 0.00015009485420980877, + "loss": 2.6334, + "step": 15695 + }, + { + "epoch": 0.67, + "grad_norm": 0.50390625, + "learning_rate": 0.000150089048536183, + "loss": 2.5787, + "step": 15696 + }, + { + "epoch": 0.67, + "grad_norm": 0.578125, + "learning_rate": 0.00015008324263717483, + "loss": 2.7065, + "step": 15697 + }, + { + "epoch": 0.67, + "grad_norm": 0.466796875, + "learning_rate": 0.00015007743651281036, + "loss": 2.5285, + "step": 15698 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.00015007163016311577, + "loss": 2.6338, + "step": 15699 + }, + { + "epoch": 0.67, + "grad_norm": 0.5078125, + "learning_rate": 0.0001500658235881171, + "loss": 2.7366, + "step": 15700 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.00015006001678784055, + "loss": 2.5492, + "step": 15701 + }, + { + "epoch": 0.67, + "grad_norm": 0.46875, + "learning_rate": 0.0001500542097623122, + "loss": 2.6327, + "step": 15702 + }, + { + "epoch": 0.67, + "grad_norm": 0.5, + "learning_rate": 0.0001500484025115582, + "loss": 2.6149, + "step": 15703 + }, + { + "epoch": 0.67, + "grad_norm": 0.5078125, + "learning_rate": 0.00015004259503560466, + "loss": 2.5174, + "step": 15704 + }, + { + "epoch": 0.67, + "grad_norm": 0.4453125, + "learning_rate": 0.00015003678733447776, + "loss": 2.5667, + "step": 15705 + }, + { + "epoch": 0.67, + "grad_norm": 0.4765625, + "learning_rate": 0.0001500309794082036, + "loss": 2.505, + "step": 15706 + }, + { + "epoch": 0.67, + "grad_norm": 0.427734375, + "learning_rate": 0.00015002517125680832, + "loss": 2.5426, + "step": 15707 + }, + { + "epoch": 0.67, + "grad_norm": 0.474609375, + "learning_rate": 0.00015001936288031807, + "loss": 2.7343, + "step": 15708 + }, + { + "epoch": 0.67, + "grad_norm": 0.41015625, + "learning_rate": 0.0001500135542787589, + "loss": 2.6284, + "step": 15709 + }, + { + "epoch": 0.67, + "grad_norm": 0.4140625, + "learning_rate": 0.00015000774545215708, + "loss": 2.5374, + "step": 15710 + }, + { + "epoch": 0.67, + "grad_norm": 0.404296875, + "learning_rate": 0.00015000193640053865, + "loss": 2.4158, + "step": 15711 + }, + { + "epoch": 0.67, + "grad_norm": 0.478515625, + "learning_rate": 0.00014999612712392977, + "loss": 2.5504, + "step": 15712 + }, + { + "epoch": 0.67, + "grad_norm": 0.4453125, + "learning_rate": 0.00014999031762235663, + "loss": 2.6156, + "step": 15713 + }, + { + "epoch": 0.67, + "grad_norm": 0.412109375, + "learning_rate": 0.0001499845078958453, + "loss": 2.5997, + "step": 15714 + }, + { + "epoch": 0.67, + "grad_norm": 0.466796875, + "learning_rate": 0.00014997869794442195, + "loss": 2.5859, + "step": 15715 + }, + { + "epoch": 0.67, + "grad_norm": 0.40625, + "learning_rate": 0.00014997288776811275, + "loss": 2.6218, + "step": 15716 + }, + { + "epoch": 0.67, + "grad_norm": 0.431640625, + "learning_rate": 0.0001499670773669438, + "loss": 2.5572, + "step": 15717 + }, + { + "epoch": 0.67, + "grad_norm": 0.41015625, + "learning_rate": 0.00014996126674094126, + "loss": 2.6766, + "step": 15718 + }, + { + "epoch": 0.67, + "grad_norm": 0.482421875, + "learning_rate": 0.0001499554558901313, + "loss": 2.6458, + "step": 15719 + }, + { + "epoch": 0.67, + "grad_norm": 0.455078125, + "learning_rate": 0.00014994964481454, + "loss": 2.721, + "step": 15720 + }, + { + "epoch": 0.67, + "grad_norm": 0.4375, + "learning_rate": 0.00014994383351419358, + "loss": 2.6149, + "step": 15721 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.00014993802198911817, + "loss": 2.6453, + "step": 15722 + }, + { + "epoch": 0.67, + "grad_norm": 0.431640625, + "learning_rate": 0.0001499322102393399, + "loss": 2.5297, + "step": 15723 + }, + { + "epoch": 0.67, + "grad_norm": 0.494140625, + "learning_rate": 0.00014992639826488493, + "loss": 2.6242, + "step": 15724 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.00014992058606577943, + "loss": 2.5526, + "step": 15725 + }, + { + "epoch": 0.67, + "grad_norm": 0.5234375, + "learning_rate": 0.0001499147736420495, + "loss": 2.6016, + "step": 15726 + }, + { + "epoch": 0.67, + "grad_norm": 0.45703125, + "learning_rate": 0.00014990896099372137, + "loss": 2.6536, + "step": 15727 + }, + { + "epoch": 0.67, + "grad_norm": 0.439453125, + "learning_rate": 0.00014990314812082114, + "loss": 2.6995, + "step": 15728 + }, + { + "epoch": 0.67, + "grad_norm": 0.466796875, + "learning_rate": 0.000149897335023375, + "loss": 2.5923, + "step": 15729 + }, + { + "epoch": 0.67, + "grad_norm": 0.419921875, + "learning_rate": 0.00014989152170140905, + "loss": 2.5441, + "step": 15730 + }, + { + "epoch": 0.67, + "grad_norm": 0.498046875, + "learning_rate": 0.00014988570815494952, + "loss": 2.739, + "step": 15731 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.0001498798943840225, + "loss": 2.5762, + "step": 15732 + }, + { + "epoch": 0.67, + "grad_norm": 0.419921875, + "learning_rate": 0.0001498740803886542, + "loss": 2.645, + "step": 15733 + }, + { + "epoch": 0.67, + "grad_norm": 0.498046875, + "learning_rate": 0.0001498682661688708, + "loss": 2.6686, + "step": 15734 + }, + { + "epoch": 0.67, + "grad_norm": 0.478515625, + "learning_rate": 0.0001498624517246984, + "loss": 2.554, + "step": 15735 + }, + { + "epoch": 0.67, + "grad_norm": 0.478515625, + "learning_rate": 0.00014985663705616316, + "loss": 2.6325, + "step": 15736 + }, + { + "epoch": 0.67, + "grad_norm": 0.384765625, + "learning_rate": 0.00014985082216329132, + "loss": 2.4888, + "step": 15737 + }, + { + "epoch": 0.67, + "grad_norm": 0.416015625, + "learning_rate": 0.00014984500704610896, + "loss": 2.4547, + "step": 15738 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.0001498391917046423, + "loss": 2.5555, + "step": 15739 + }, + { + "epoch": 0.67, + "grad_norm": 0.50390625, + "learning_rate": 0.0001498333761389175, + "loss": 2.6615, + "step": 15740 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.0001498275603489607, + "loss": 2.5525, + "step": 15741 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.00014982174433479813, + "loss": 2.6423, + "step": 15742 + }, + { + "epoch": 0.67, + "grad_norm": 0.408203125, + "learning_rate": 0.0001498159280964559, + "loss": 2.4929, + "step": 15743 + }, + { + "epoch": 0.67, + "grad_norm": 0.4609375, + "learning_rate": 0.00014981011163396014, + "loss": 2.5335, + "step": 15744 + }, + { + "epoch": 0.67, + "grad_norm": 0.44921875, + "learning_rate": 0.00014980429494733716, + "loss": 2.4662, + "step": 15745 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.000149798478036613, + "loss": 2.5831, + "step": 15746 + }, + { + "epoch": 0.67, + "grad_norm": 0.47265625, + "learning_rate": 0.00014979266090181393, + "loss": 2.5313, + "step": 15747 + }, + { + "epoch": 0.67, + "grad_norm": 0.5, + "learning_rate": 0.00014978684354296606, + "loss": 2.6374, + "step": 15748 + }, + { + "epoch": 0.67, + "grad_norm": 0.423828125, + "learning_rate": 0.0001497810259600956, + "loss": 2.4787, + "step": 15749 + }, + { + "epoch": 0.67, + "grad_norm": 0.50390625, + "learning_rate": 0.00014977520815322873, + "loss": 2.5586, + "step": 15750 + }, + { + "epoch": 0.67, + "grad_norm": 0.51953125, + "learning_rate": 0.0001497693901223916, + "loss": 2.603, + "step": 15751 + }, + { + "epoch": 0.67, + "grad_norm": 0.4375, + "learning_rate": 0.0001497635718676104, + "loss": 2.546, + "step": 15752 + }, + { + "epoch": 0.67, + "grad_norm": 0.5390625, + "learning_rate": 0.0001497577533889113, + "loss": 2.6528, + "step": 15753 + }, + { + "epoch": 0.67, + "grad_norm": 0.44921875, + "learning_rate": 0.0001497519346863205, + "loss": 2.6522, + "step": 15754 + }, + { + "epoch": 0.67, + "grad_norm": 0.412109375, + "learning_rate": 0.00014974611575986418, + "loss": 2.5171, + "step": 15755 + }, + { + "epoch": 0.67, + "grad_norm": 0.5, + "learning_rate": 0.00014974029660956854, + "loss": 2.716, + "step": 15756 + }, + { + "epoch": 0.67, + "grad_norm": 0.46484375, + "learning_rate": 0.0001497344772354597, + "loss": 2.4351, + "step": 15757 + }, + { + "epoch": 0.67, + "grad_norm": 0.474609375, + "learning_rate": 0.0001497286576375639, + "loss": 2.5509, + "step": 15758 + }, + { + "epoch": 0.67, + "grad_norm": 0.46875, + "learning_rate": 0.00014972283781590734, + "loss": 2.5728, + "step": 15759 + }, + { + "epoch": 0.67, + "grad_norm": 0.45703125, + "learning_rate": 0.00014971701777051615, + "loss": 2.5832, + "step": 15760 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.0001497111975014166, + "loss": 2.5993, + "step": 15761 + }, + { + "epoch": 0.67, + "grad_norm": 0.50390625, + "learning_rate": 0.0001497053770086348, + "loss": 2.54, + "step": 15762 + }, + { + "epoch": 0.67, + "grad_norm": 0.455078125, + "learning_rate": 0.00014969955629219695, + "loss": 2.5975, + "step": 15763 + }, + { + "epoch": 0.67, + "grad_norm": 0.435546875, + "learning_rate": 0.00014969373535212926, + "loss": 2.637, + "step": 15764 + }, + { + "epoch": 0.67, + "grad_norm": 0.486328125, + "learning_rate": 0.00014968791418845794, + "loss": 2.7961, + "step": 15765 + }, + { + "epoch": 0.67, + "grad_norm": 0.50390625, + "learning_rate": 0.00014968209280120918, + "loss": 2.6609, + "step": 15766 + }, + { + "epoch": 0.67, + "grad_norm": 0.43359375, + "learning_rate": 0.0001496762711904091, + "loss": 2.7135, + "step": 15767 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.00014967044935608402, + "loss": 2.5423, + "step": 15768 + }, + { + "epoch": 0.67, + "grad_norm": 0.439453125, + "learning_rate": 0.00014966462729826006, + "loss": 2.5058, + "step": 15769 + }, + { + "epoch": 0.67, + "grad_norm": 0.443359375, + "learning_rate": 0.0001496588050169634, + "loss": 2.6441, + "step": 15770 + }, + { + "epoch": 0.67, + "grad_norm": 0.423828125, + "learning_rate": 0.00014965298251222027, + "loss": 2.5957, + "step": 15771 + }, + { + "epoch": 0.67, + "grad_norm": 0.498046875, + "learning_rate": 0.0001496471597840569, + "loss": 2.4808, + "step": 15772 + }, + { + "epoch": 0.67, + "grad_norm": 0.50390625, + "learning_rate": 0.00014964133683249942, + "loss": 2.531, + "step": 15773 + }, + { + "epoch": 0.67, + "grad_norm": 0.4375, + "learning_rate": 0.00014963551365757406, + "loss": 2.5588, + "step": 15774 + }, + { + "epoch": 0.67, + "grad_norm": 0.43359375, + "learning_rate": 0.00014962969025930707, + "loss": 2.5222, + "step": 15775 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.00014962386663772457, + "loss": 2.5492, + "step": 15776 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.0001496180427928528, + "loss": 2.5808, + "step": 15777 + }, + { + "epoch": 0.67, + "grad_norm": 0.5, + "learning_rate": 0.00014961221872471803, + "loss": 2.5043, + "step": 15778 + }, + { + "epoch": 0.67, + "grad_norm": 0.4453125, + "learning_rate": 0.00014960639443334635, + "loss": 2.5389, + "step": 15779 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.00014960056991876407, + "loss": 2.7129, + "step": 15780 + }, + { + "epoch": 0.67, + "grad_norm": 0.4609375, + "learning_rate": 0.00014959474518099728, + "loss": 2.5798, + "step": 15781 + }, + { + "epoch": 0.67, + "grad_norm": 0.392578125, + "learning_rate": 0.00014958892022007232, + "loss": 2.5462, + "step": 15782 + }, + { + "epoch": 0.67, + "grad_norm": 0.4921875, + "learning_rate": 0.0001495830950360153, + "loss": 2.578, + "step": 15783 + }, + { + "epoch": 0.67, + "grad_norm": 0.46484375, + "learning_rate": 0.00014957726962885252, + "loss": 2.6424, + "step": 15784 + }, + { + "epoch": 0.67, + "grad_norm": 0.447265625, + "learning_rate": 0.00014957144399861013, + "loss": 2.5291, + "step": 15785 + }, + { + "epoch": 0.67, + "grad_norm": 0.466796875, + "learning_rate": 0.00014956561814531436, + "loss": 2.6241, + "step": 15786 + }, + { + "epoch": 0.67, + "grad_norm": 0.46484375, + "learning_rate": 0.0001495597920689914, + "loss": 2.5748, + "step": 15787 + }, + { + "epoch": 0.67, + "grad_norm": 0.455078125, + "learning_rate": 0.00014955396576966752, + "loss": 2.548, + "step": 15788 + }, + { + "epoch": 0.67, + "grad_norm": 0.45703125, + "learning_rate": 0.00014954813924736887, + "loss": 2.5337, + "step": 15789 + }, + { + "epoch": 0.67, + "grad_norm": 0.4375, + "learning_rate": 0.0001495423125021217, + "loss": 2.536, + "step": 15790 + }, + { + "epoch": 0.67, + "grad_norm": 0.490234375, + "learning_rate": 0.00014953648553395226, + "loss": 2.5459, + "step": 15791 + }, + { + "epoch": 0.67, + "grad_norm": 0.47265625, + "learning_rate": 0.00014953065834288673, + "loss": 2.5351, + "step": 15792 + }, + { + "epoch": 0.67, + "grad_norm": 0.431640625, + "learning_rate": 0.00014952483092895135, + "loss": 2.4645, + "step": 15793 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.0001495190032921723, + "loss": 2.5861, + "step": 15794 + }, + { + "epoch": 0.67, + "grad_norm": 0.54296875, + "learning_rate": 0.00014951317543257585, + "loss": 2.7213, + "step": 15795 + }, + { + "epoch": 0.67, + "grad_norm": 0.455078125, + "learning_rate": 0.00014950734735018822, + "loss": 2.463, + "step": 15796 + }, + { + "epoch": 0.67, + "grad_norm": 0.45703125, + "learning_rate": 0.0001495015190450356, + "loss": 2.6478, + "step": 15797 + }, + { + "epoch": 0.67, + "grad_norm": 0.40625, + "learning_rate": 0.00014949569051714423, + "loss": 2.5447, + "step": 15798 + }, + { + "epoch": 0.67, + "grad_norm": 0.455078125, + "learning_rate": 0.00014948986176654033, + "loss": 2.5466, + "step": 15799 + }, + { + "epoch": 0.67, + "grad_norm": 0.470703125, + "learning_rate": 0.00014948403279325018, + "loss": 2.5519, + "step": 15800 + }, + { + "epoch": 0.67, + "grad_norm": 0.486328125, + "learning_rate": 0.00014947820359729997, + "loss": 2.733, + "step": 15801 + }, + { + "epoch": 0.67, + "grad_norm": 0.455078125, + "learning_rate": 0.0001494723741787159, + "loss": 2.5973, + "step": 15802 + }, + { + "epoch": 0.67, + "grad_norm": 0.50390625, + "learning_rate": 0.0001494665445375242, + "loss": 2.5241, + "step": 15803 + }, + { + "epoch": 0.67, + "grad_norm": 0.421875, + "learning_rate": 0.0001494607146737512, + "loss": 2.6045, + "step": 15804 + }, + { + "epoch": 0.67, + "grad_norm": 0.396484375, + "learning_rate": 0.00014945488458742296, + "loss": 2.4365, + "step": 15805 + }, + { + "epoch": 0.67, + "grad_norm": 0.400390625, + "learning_rate": 0.00014944905427856587, + "loss": 2.5642, + "step": 15806 + }, + { + "epoch": 0.67, + "grad_norm": 0.44140625, + "learning_rate": 0.0001494432237472061, + "loss": 2.5688, + "step": 15807 + }, + { + "epoch": 0.67, + "grad_norm": 0.48046875, + "learning_rate": 0.0001494373929933699, + "loss": 2.7569, + "step": 15808 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.0001494315620170835, + "loss": 2.6199, + "step": 15809 + }, + { + "epoch": 0.67, + "grad_norm": 0.51171875, + "learning_rate": 0.00014942573081837314, + "loss": 2.533, + "step": 15810 + }, + { + "epoch": 0.67, + "grad_norm": 0.490234375, + "learning_rate": 0.00014941989939726505, + "loss": 2.5351, + "step": 15811 + }, + { + "epoch": 0.67, + "grad_norm": 0.470703125, + "learning_rate": 0.00014941406775378547, + "loss": 2.6987, + "step": 15812 + }, + { + "epoch": 0.67, + "grad_norm": 0.408203125, + "learning_rate": 0.00014940823588796063, + "loss": 2.6489, + "step": 15813 + }, + { + "epoch": 0.67, + "grad_norm": 0.474609375, + "learning_rate": 0.0001494024037998168, + "loss": 2.5928, + "step": 15814 + }, + { + "epoch": 0.67, + "grad_norm": 0.462890625, + "learning_rate": 0.00014939657148938022, + "loss": 2.6015, + "step": 15815 + }, + { + "epoch": 0.67, + "grad_norm": 0.439453125, + "learning_rate": 0.0001493907389566771, + "loss": 2.7448, + "step": 15816 + }, + { + "epoch": 0.67, + "grad_norm": 0.421875, + "learning_rate": 0.0001493849062017337, + "loss": 2.555, + "step": 15817 + }, + { + "epoch": 0.67, + "grad_norm": 0.431640625, + "learning_rate": 0.0001493790732245763, + "loss": 2.5754, + "step": 15818 + }, + { + "epoch": 0.67, + "grad_norm": 0.431640625, + "learning_rate": 0.00014937324002523106, + "loss": 2.5517, + "step": 15819 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.00014936740660372437, + "loss": 2.7479, + "step": 15820 + }, + { + "epoch": 0.67, + "grad_norm": 0.451171875, + "learning_rate": 0.00014936157296008233, + "loss": 2.6021, + "step": 15821 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.00014935573909433125, + "loss": 2.5004, + "step": 15822 + }, + { + "epoch": 0.67, + "grad_norm": 0.396484375, + "learning_rate": 0.0001493499050064974, + "loss": 2.5976, + "step": 15823 + }, + { + "epoch": 0.67, + "grad_norm": 0.478515625, + "learning_rate": 0.00014934407069660703, + "loss": 2.7684, + "step": 15824 + }, + { + "epoch": 0.67, + "grad_norm": 0.462890625, + "learning_rate": 0.00014933823616468636, + "loss": 2.6421, + "step": 15825 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.00014933240141076163, + "loss": 2.4953, + "step": 15826 + }, + { + "epoch": 0.67, + "grad_norm": 0.458984375, + "learning_rate": 0.00014932656643485914, + "loss": 2.5731, + "step": 15827 + }, + { + "epoch": 0.67, + "grad_norm": 0.421875, + "learning_rate": 0.00014932073123700515, + "loss": 2.5852, + "step": 15828 + }, + { + "epoch": 0.67, + "grad_norm": 0.53515625, + "learning_rate": 0.00014931489581722587, + "loss": 2.5619, + "step": 15829 + }, + { + "epoch": 0.67, + "grad_norm": 0.435546875, + "learning_rate": 0.0001493090601755476, + "loss": 2.4689, + "step": 15830 + }, + { + "epoch": 0.67, + "grad_norm": 0.44140625, + "learning_rate": 0.00014930322431199657, + "loss": 2.536, + "step": 15831 + }, + { + "epoch": 0.67, + "grad_norm": 0.453125, + "learning_rate": 0.00014929738822659904, + "loss": 2.7606, + "step": 15832 + }, + { + "epoch": 0.67, + "grad_norm": 0.478515625, + "learning_rate": 0.00014929155191938128, + "loss": 2.6633, + "step": 15833 + }, + { + "epoch": 0.67, + "grad_norm": 0.474609375, + "learning_rate": 0.00014928571539036956, + "loss": 2.5156, + "step": 15834 + }, + { + "epoch": 0.67, + "grad_norm": 0.51171875, + "learning_rate": 0.00014927987863959014, + "loss": 2.6368, + "step": 15835 + }, + { + "epoch": 0.67, + "grad_norm": 0.404296875, + "learning_rate": 0.00014927404166706926, + "loss": 2.3095, + "step": 15836 + }, + { + "epoch": 0.67, + "grad_norm": 0.48046875, + "learning_rate": 0.0001492682044728332, + "loss": 2.5363, + "step": 15837 + }, + { + "epoch": 0.67, + "grad_norm": 0.4453125, + "learning_rate": 0.00014926236705690825, + "loss": 2.6653, + "step": 15838 + }, + { + "epoch": 0.67, + "grad_norm": 0.484375, + "learning_rate": 0.00014925652941932062, + "loss": 2.538, + "step": 15839 + }, + { + "epoch": 0.67, + "grad_norm": 0.427734375, + "learning_rate": 0.00014925069156009665, + "loss": 2.6634, + "step": 15840 + }, + { + "epoch": 0.68, + "grad_norm": 0.458984375, + "learning_rate": 0.0001492448534792625, + "loss": 2.6573, + "step": 15841 + }, + { + "epoch": 0.68, + "grad_norm": 0.48046875, + "learning_rate": 0.00014923901517684458, + "loss": 2.5697, + "step": 15842 + }, + { + "epoch": 0.68, + "grad_norm": 0.423828125, + "learning_rate": 0.00014923317665286904, + "loss": 2.4845, + "step": 15843 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.0001492273379073622, + "loss": 2.6818, + "step": 15844 + }, + { + "epoch": 0.68, + "grad_norm": 0.45703125, + "learning_rate": 0.00014922149894035033, + "loss": 2.4476, + "step": 15845 + }, + { + "epoch": 0.68, + "grad_norm": 0.392578125, + "learning_rate": 0.0001492156597518597, + "loss": 2.5163, + "step": 15846 + }, + { + "epoch": 0.68, + "grad_norm": 0.455078125, + "learning_rate": 0.00014920982034191663, + "loss": 2.55, + "step": 15847 + }, + { + "epoch": 0.68, + "grad_norm": 0.41796875, + "learning_rate": 0.00014920398071054732, + "loss": 2.6788, + "step": 15848 + }, + { + "epoch": 0.68, + "grad_norm": 0.474609375, + "learning_rate": 0.00014919814085777807, + "loss": 2.5124, + "step": 15849 + }, + { + "epoch": 0.68, + "grad_norm": 0.46875, + "learning_rate": 0.0001491923007836352, + "loss": 2.5403, + "step": 15850 + }, + { + "epoch": 0.68, + "grad_norm": 0.453125, + "learning_rate": 0.00014918646048814492, + "loss": 2.6267, + "step": 15851 + }, + { + "epoch": 0.68, + "grad_norm": 0.49609375, + "learning_rate": 0.00014918061997133358, + "loss": 2.4842, + "step": 15852 + }, + { + "epoch": 0.68, + "grad_norm": 0.466796875, + "learning_rate": 0.00014917477923322736, + "loss": 2.6196, + "step": 15853 + }, + { + "epoch": 0.68, + "grad_norm": 0.47265625, + "learning_rate": 0.00014916893827385266, + "loss": 2.683, + "step": 15854 + }, + { + "epoch": 0.68, + "grad_norm": 0.44140625, + "learning_rate": 0.0001491630970932357, + "loss": 2.5004, + "step": 15855 + }, + { + "epoch": 0.68, + "grad_norm": 0.494140625, + "learning_rate": 0.00014915725569140275, + "loss": 2.6001, + "step": 15856 + }, + { + "epoch": 0.68, + "grad_norm": 0.498046875, + "learning_rate": 0.00014915141406838012, + "loss": 2.6587, + "step": 15857 + }, + { + "epoch": 0.68, + "grad_norm": 0.48828125, + "learning_rate": 0.0001491455722241941, + "loss": 2.7278, + "step": 15858 + }, + { + "epoch": 0.68, + "grad_norm": 0.478515625, + "learning_rate": 0.00014913973015887093, + "loss": 2.6741, + "step": 15859 + }, + { + "epoch": 0.68, + "grad_norm": 0.48046875, + "learning_rate": 0.000149133887872437, + "loss": 2.5119, + "step": 15860 + }, + { + "epoch": 0.68, + "grad_norm": 0.447265625, + "learning_rate": 0.00014912804536491847, + "loss": 2.5948, + "step": 15861 + }, + { + "epoch": 0.68, + "grad_norm": 0.41796875, + "learning_rate": 0.00014912220263634172, + "loss": 2.5475, + "step": 15862 + }, + { + "epoch": 0.68, + "grad_norm": 0.4375, + "learning_rate": 0.000149116359686733, + "loss": 2.4977, + "step": 15863 + }, + { + "epoch": 0.68, + "grad_norm": 0.470703125, + "learning_rate": 0.00014911051651611863, + "loss": 2.6044, + "step": 15864 + }, + { + "epoch": 0.68, + "grad_norm": 0.443359375, + "learning_rate": 0.00014910467312452484, + "loss": 2.5498, + "step": 15865 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.00014909882951197803, + "loss": 2.542, + "step": 15866 + }, + { + "epoch": 0.68, + "grad_norm": 0.421875, + "learning_rate": 0.00014909298567850437, + "loss": 2.5778, + "step": 15867 + }, + { + "epoch": 0.68, + "grad_norm": 0.4453125, + "learning_rate": 0.00014908714162413026, + "loss": 2.6046, + "step": 15868 + }, + { + "epoch": 0.68, + "grad_norm": 0.494140625, + "learning_rate": 0.00014908129734888192, + "loss": 2.6876, + "step": 15869 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.0001490754528527857, + "loss": 2.3262, + "step": 15870 + }, + { + "epoch": 0.68, + "grad_norm": 0.4140625, + "learning_rate": 0.00014906960813586788, + "loss": 2.7478, + "step": 15871 + }, + { + "epoch": 0.68, + "grad_norm": 0.4375, + "learning_rate": 0.00014906376319815474, + "loss": 2.6481, + "step": 15872 + }, + { + "epoch": 0.68, + "grad_norm": 0.462890625, + "learning_rate": 0.0001490579180396726, + "loss": 2.5257, + "step": 15873 + }, + { + "epoch": 0.68, + "grad_norm": 0.380859375, + "learning_rate": 0.00014905207266044775, + "loss": 2.4571, + "step": 15874 + }, + { + "epoch": 0.68, + "grad_norm": 0.4296875, + "learning_rate": 0.00014904622706050653, + "loss": 2.5665, + "step": 15875 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.00014904038123987517, + "loss": 2.5646, + "step": 15876 + }, + { + "epoch": 0.68, + "grad_norm": 0.41796875, + "learning_rate": 0.00014903453519858007, + "loss": 2.5778, + "step": 15877 + }, + { + "epoch": 0.68, + "grad_norm": 0.46484375, + "learning_rate": 0.00014902868893664742, + "loss": 2.4879, + "step": 15878 + }, + { + "epoch": 0.68, + "grad_norm": 0.5078125, + "learning_rate": 0.00014902284245410364, + "loss": 2.6757, + "step": 15879 + }, + { + "epoch": 0.68, + "grad_norm": 0.404296875, + "learning_rate": 0.00014901699575097497, + "loss": 2.47, + "step": 15880 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.0001490111488272877, + "loss": 2.6191, + "step": 15881 + }, + { + "epoch": 0.68, + "grad_norm": 0.47265625, + "learning_rate": 0.00014900530168306822, + "loss": 2.6346, + "step": 15882 + }, + { + "epoch": 0.68, + "grad_norm": 0.44140625, + "learning_rate": 0.00014899945431834277, + "loss": 2.5844, + "step": 15883 + }, + { + "epoch": 0.68, + "grad_norm": 0.4140625, + "learning_rate": 0.00014899360673313768, + "loss": 2.4963, + "step": 15884 + }, + { + "epoch": 0.68, + "grad_norm": 0.52734375, + "learning_rate": 0.00014898775892747927, + "loss": 2.5478, + "step": 15885 + }, + { + "epoch": 0.68, + "grad_norm": 0.4921875, + "learning_rate": 0.00014898191090139383, + "loss": 2.5707, + "step": 15886 + }, + { + "epoch": 0.68, + "grad_norm": 0.466796875, + "learning_rate": 0.0001489760626549077, + "loss": 2.7549, + "step": 15887 + }, + { + "epoch": 0.68, + "grad_norm": 0.43359375, + "learning_rate": 0.00014897021418804719, + "loss": 2.6429, + "step": 15888 + }, + { + "epoch": 0.68, + "grad_norm": 0.46484375, + "learning_rate": 0.00014896436550083863, + "loss": 2.751, + "step": 15889 + }, + { + "epoch": 0.68, + "grad_norm": 0.447265625, + "learning_rate": 0.0001489585165933083, + "loss": 2.6884, + "step": 15890 + }, + { + "epoch": 0.68, + "grad_norm": 0.42578125, + "learning_rate": 0.00014895266746548256, + "loss": 2.4994, + "step": 15891 + }, + { + "epoch": 0.68, + "grad_norm": 0.466796875, + "learning_rate": 0.00014894681811738767, + "loss": 2.6643, + "step": 15892 + }, + { + "epoch": 0.68, + "grad_norm": 0.447265625, + "learning_rate": 0.00014894096854905, + "loss": 2.4473, + "step": 15893 + }, + { + "epoch": 0.68, + "grad_norm": 0.435546875, + "learning_rate": 0.00014893511876049587, + "loss": 2.5955, + "step": 15894 + }, + { + "epoch": 0.68, + "grad_norm": 0.404296875, + "learning_rate": 0.0001489292687517516, + "loss": 2.4137, + "step": 15895 + }, + { + "epoch": 0.68, + "grad_norm": 0.412109375, + "learning_rate": 0.0001489234185228435, + "loss": 2.6585, + "step": 15896 + }, + { + "epoch": 0.68, + "grad_norm": 0.421875, + "learning_rate": 0.00014891756807379787, + "loss": 2.4986, + "step": 15897 + }, + { + "epoch": 0.68, + "grad_norm": 0.4453125, + "learning_rate": 0.00014891171740464106, + "loss": 2.5104, + "step": 15898 + }, + { + "epoch": 0.68, + "grad_norm": 0.4140625, + "learning_rate": 0.00014890586651539943, + "loss": 2.449, + "step": 15899 + }, + { + "epoch": 0.68, + "grad_norm": 0.47265625, + "learning_rate": 0.00014890001540609924, + "loss": 2.6266, + "step": 15900 + }, + { + "epoch": 0.68, + "grad_norm": 0.4609375, + "learning_rate": 0.00014889416407676685, + "loss": 2.5878, + "step": 15901 + }, + { + "epoch": 0.68, + "grad_norm": 0.47265625, + "learning_rate": 0.00014888831252742862, + "loss": 2.6351, + "step": 15902 + }, + { + "epoch": 0.68, + "grad_norm": 0.4609375, + "learning_rate": 0.00014888246075811084, + "loss": 2.5656, + "step": 15903 + }, + { + "epoch": 0.68, + "grad_norm": 0.50390625, + "learning_rate": 0.00014887660876883985, + "loss": 2.5938, + "step": 15904 + }, + { + "epoch": 0.68, + "grad_norm": 0.498046875, + "learning_rate": 0.00014887075655964198, + "loss": 2.5742, + "step": 15905 + }, + { + "epoch": 0.68, + "grad_norm": 0.482421875, + "learning_rate": 0.00014886490413054353, + "loss": 2.6569, + "step": 15906 + }, + { + "epoch": 0.68, + "grad_norm": 0.484375, + "learning_rate": 0.00014885905148157092, + "loss": 2.6333, + "step": 15907 + }, + { + "epoch": 0.68, + "grad_norm": 0.46875, + "learning_rate": 0.0001488531986127504, + "loss": 2.4542, + "step": 15908 + }, + { + "epoch": 0.68, + "grad_norm": 0.494140625, + "learning_rate": 0.0001488473455241084, + "loss": 2.5689, + "step": 15909 + }, + { + "epoch": 0.68, + "grad_norm": 0.486328125, + "learning_rate": 0.00014884149221567114, + "loss": 2.614, + "step": 15910 + }, + { + "epoch": 0.68, + "grad_norm": 0.4375, + "learning_rate": 0.00014883563868746502, + "loss": 2.5034, + "step": 15911 + }, + { + "epoch": 0.68, + "grad_norm": 0.44921875, + "learning_rate": 0.0001488297849395164, + "loss": 2.5771, + "step": 15912 + }, + { + "epoch": 0.68, + "grad_norm": 0.46875, + "learning_rate": 0.00014882393097185158, + "loss": 2.5967, + "step": 15913 + }, + { + "epoch": 0.68, + "grad_norm": 0.44921875, + "learning_rate": 0.00014881807678449692, + "loss": 2.4818, + "step": 15914 + }, + { + "epoch": 0.68, + "grad_norm": 0.466796875, + "learning_rate": 0.00014881222237747875, + "loss": 2.5641, + "step": 15915 + }, + { + "epoch": 0.68, + "grad_norm": 0.45703125, + "learning_rate": 0.00014880636775082336, + "loss": 2.5753, + "step": 15916 + }, + { + "epoch": 0.68, + "grad_norm": 0.47265625, + "learning_rate": 0.00014880051290455724, + "loss": 2.6724, + "step": 15917 + }, + { + "epoch": 0.68, + "grad_norm": 0.443359375, + "learning_rate": 0.0001487946578387066, + "loss": 2.5783, + "step": 15918 + }, + { + "epoch": 0.68, + "grad_norm": 0.431640625, + "learning_rate": 0.00014878880255329787, + "loss": 2.5765, + "step": 15919 + }, + { + "epoch": 0.68, + "grad_norm": 0.484375, + "learning_rate": 0.00014878294704835732, + "loss": 2.6104, + "step": 15920 + }, + { + "epoch": 0.68, + "grad_norm": 0.50390625, + "learning_rate": 0.00014877709132391135, + "loss": 2.647, + "step": 15921 + }, + { + "epoch": 0.68, + "grad_norm": 0.5, + "learning_rate": 0.0001487712353799863, + "loss": 2.6519, + "step": 15922 + }, + { + "epoch": 0.68, + "grad_norm": 0.423828125, + "learning_rate": 0.0001487653792166085, + "loss": 2.4995, + "step": 15923 + }, + { + "epoch": 0.68, + "grad_norm": 0.458984375, + "learning_rate": 0.00014875952283380432, + "loss": 2.5693, + "step": 15924 + }, + { + "epoch": 0.68, + "grad_norm": 0.435546875, + "learning_rate": 0.0001487536662316001, + "loss": 2.4991, + "step": 15925 + }, + { + "epoch": 0.68, + "grad_norm": 0.478515625, + "learning_rate": 0.00014874780941002223, + "loss": 2.5848, + "step": 15926 + }, + { + "epoch": 0.68, + "grad_norm": 0.498046875, + "learning_rate": 0.000148741952369097, + "loss": 2.6224, + "step": 15927 + }, + { + "epoch": 0.68, + "grad_norm": 0.5078125, + "learning_rate": 0.00014873609510885082, + "loss": 2.7142, + "step": 15928 + }, + { + "epoch": 0.68, + "grad_norm": 0.4453125, + "learning_rate": 0.00014873023762931002, + "loss": 2.5057, + "step": 15929 + }, + { + "epoch": 0.68, + "grad_norm": 0.48046875, + "learning_rate": 0.00014872437993050094, + "loss": 2.6759, + "step": 15930 + }, + { + "epoch": 0.68, + "grad_norm": 0.455078125, + "learning_rate": 0.00014871852201244997, + "loss": 2.5223, + "step": 15931 + }, + { + "epoch": 0.68, + "grad_norm": 0.49609375, + "learning_rate": 0.00014871266387518346, + "loss": 2.6058, + "step": 15932 + }, + { + "epoch": 0.68, + "grad_norm": 0.53515625, + "learning_rate": 0.00014870680551872778, + "loss": 2.5754, + "step": 15933 + }, + { + "epoch": 0.68, + "grad_norm": 0.51953125, + "learning_rate": 0.00014870094694310922, + "loss": 2.6414, + "step": 15934 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.00014869508814835424, + "loss": 2.5851, + "step": 15935 + }, + { + "epoch": 0.68, + "grad_norm": 0.5, + "learning_rate": 0.00014868922913448916, + "loss": 2.5706, + "step": 15936 + }, + { + "epoch": 0.68, + "grad_norm": 0.50390625, + "learning_rate": 0.00014868336990154035, + "loss": 2.5808, + "step": 15937 + }, + { + "epoch": 0.68, + "grad_norm": 0.41015625, + "learning_rate": 0.00014867751044953415, + "loss": 2.4833, + "step": 15938 + }, + { + "epoch": 0.68, + "grad_norm": 0.4453125, + "learning_rate": 0.00014867165077849697, + "loss": 2.6339, + "step": 15939 + }, + { + "epoch": 0.68, + "grad_norm": 0.431640625, + "learning_rate": 0.0001486657908884551, + "loss": 2.6057, + "step": 15940 + }, + { + "epoch": 0.68, + "grad_norm": 0.5078125, + "learning_rate": 0.00014865993077943499, + "loss": 2.6219, + "step": 15941 + }, + { + "epoch": 0.68, + "grad_norm": 0.5234375, + "learning_rate": 0.00014865407045146296, + "loss": 2.5089, + "step": 15942 + }, + { + "epoch": 0.68, + "grad_norm": 0.443359375, + "learning_rate": 0.00014864820990456539, + "loss": 2.5123, + "step": 15943 + }, + { + "epoch": 0.68, + "grad_norm": 0.5, + "learning_rate": 0.00014864234913876867, + "loss": 2.6957, + "step": 15944 + }, + { + "epoch": 0.68, + "grad_norm": 0.46484375, + "learning_rate": 0.00014863648815409915, + "loss": 2.5577, + "step": 15945 + }, + { + "epoch": 0.68, + "grad_norm": 0.4375, + "learning_rate": 0.0001486306269505832, + "loss": 2.496, + "step": 15946 + }, + { + "epoch": 0.68, + "grad_norm": 0.4296875, + "learning_rate": 0.00014862476552824722, + "loss": 2.4713, + "step": 15947 + }, + { + "epoch": 0.68, + "grad_norm": 0.49609375, + "learning_rate": 0.00014861890388711754, + "loss": 2.531, + "step": 15948 + }, + { + "epoch": 0.68, + "grad_norm": 0.43359375, + "learning_rate": 0.0001486130420272206, + "loss": 2.512, + "step": 15949 + }, + { + "epoch": 0.68, + "grad_norm": 0.44140625, + "learning_rate": 0.00014860717994858268, + "loss": 2.6554, + "step": 15950 + }, + { + "epoch": 0.68, + "grad_norm": 0.458984375, + "learning_rate": 0.00014860131765123024, + "loss": 2.5272, + "step": 15951 + }, + { + "epoch": 0.68, + "grad_norm": 0.453125, + "learning_rate": 0.00014859545513518966, + "loss": 2.575, + "step": 15952 + }, + { + "epoch": 0.68, + "grad_norm": 0.48828125, + "learning_rate": 0.00014858959240048728, + "loss": 2.6293, + "step": 15953 + }, + { + "epoch": 0.68, + "grad_norm": 0.484375, + "learning_rate": 0.00014858372944714948, + "loss": 2.5734, + "step": 15954 + }, + { + "epoch": 0.68, + "grad_norm": 0.453125, + "learning_rate": 0.00014857786627520264, + "loss": 2.4813, + "step": 15955 + }, + { + "epoch": 0.68, + "grad_norm": 0.443359375, + "learning_rate": 0.00014857200288467316, + "loss": 2.6273, + "step": 15956 + }, + { + "epoch": 0.68, + "grad_norm": 0.4921875, + "learning_rate": 0.00014856613927558744, + "loss": 2.812, + "step": 15957 + }, + { + "epoch": 0.68, + "grad_norm": 0.41796875, + "learning_rate": 0.00014856027544797183, + "loss": 2.5276, + "step": 15958 + }, + { + "epoch": 0.68, + "grad_norm": 0.427734375, + "learning_rate": 0.0001485544114018527, + "loss": 2.4732, + "step": 15959 + }, + { + "epoch": 0.68, + "grad_norm": 0.5078125, + "learning_rate": 0.0001485485471372565, + "loss": 2.7403, + "step": 15960 + }, + { + "epoch": 0.68, + "grad_norm": 0.51171875, + "learning_rate": 0.0001485426826542096, + "loss": 2.7123, + "step": 15961 + }, + { + "epoch": 0.68, + "grad_norm": 0.42578125, + "learning_rate": 0.00014853681795273834, + "loss": 2.5396, + "step": 15962 + }, + { + "epoch": 0.68, + "grad_norm": 0.4296875, + "learning_rate": 0.00014853095303286914, + "loss": 2.6078, + "step": 15963 + }, + { + "epoch": 0.68, + "grad_norm": 0.49609375, + "learning_rate": 0.00014852508789462836, + "loss": 2.588, + "step": 15964 + }, + { + "epoch": 0.68, + "grad_norm": 0.515625, + "learning_rate": 0.00014851922253804244, + "loss": 2.6752, + "step": 15965 + }, + { + "epoch": 0.68, + "grad_norm": 0.53515625, + "learning_rate": 0.00014851335696313775, + "loss": 2.6051, + "step": 15966 + }, + { + "epoch": 0.68, + "grad_norm": 0.44140625, + "learning_rate": 0.00014850749116994066, + "loss": 2.4557, + "step": 15967 + }, + { + "epoch": 0.68, + "grad_norm": 0.400390625, + "learning_rate": 0.00014850162515847763, + "loss": 2.6617, + "step": 15968 + }, + { + "epoch": 0.68, + "grad_norm": 0.44921875, + "learning_rate": 0.000148495758928775, + "loss": 2.4369, + "step": 15969 + }, + { + "epoch": 0.68, + "grad_norm": 0.44921875, + "learning_rate": 0.00014848989248085918, + "loss": 2.5499, + "step": 15970 + }, + { + "epoch": 0.68, + "grad_norm": 0.48828125, + "learning_rate": 0.00014848402581475655, + "loss": 2.6162, + "step": 15971 + }, + { + "epoch": 0.68, + "grad_norm": 0.45703125, + "learning_rate": 0.00014847815893049353, + "loss": 2.679, + "step": 15972 + }, + { + "epoch": 0.68, + "grad_norm": 0.435546875, + "learning_rate": 0.00014847229182809652, + "loss": 2.5735, + "step": 15973 + }, + { + "epoch": 0.68, + "grad_norm": 0.458984375, + "learning_rate": 0.0001484664245075919, + "loss": 2.5447, + "step": 15974 + }, + { + "epoch": 0.68, + "grad_norm": 0.4921875, + "learning_rate": 0.00014846055696900607, + "loss": 2.5119, + "step": 15975 + }, + { + "epoch": 0.68, + "grad_norm": 0.44140625, + "learning_rate": 0.00014845468921236547, + "loss": 2.4943, + "step": 15976 + }, + { + "epoch": 0.68, + "grad_norm": 0.51953125, + "learning_rate": 0.00014844882123769647, + "loss": 2.6526, + "step": 15977 + }, + { + "epoch": 0.68, + "grad_norm": 0.470703125, + "learning_rate": 0.00014844295304502548, + "loss": 2.661, + "step": 15978 + }, + { + "epoch": 0.68, + "grad_norm": 0.43359375, + "learning_rate": 0.00014843708463437892, + "loss": 2.5941, + "step": 15979 + }, + { + "epoch": 0.68, + "grad_norm": 0.515625, + "learning_rate": 0.00014843121600578311, + "loss": 2.6785, + "step": 15980 + }, + { + "epoch": 0.68, + "grad_norm": 0.4765625, + "learning_rate": 0.00014842534715926462, + "loss": 2.6884, + "step": 15981 + }, + { + "epoch": 0.68, + "grad_norm": 0.46875, + "learning_rate": 0.0001484194780948497, + "loss": 2.6363, + "step": 15982 + }, + { + "epoch": 0.68, + "grad_norm": 0.4609375, + "learning_rate": 0.00014841360881256486, + "loss": 2.6362, + "step": 15983 + }, + { + "epoch": 0.68, + "grad_norm": 0.5078125, + "learning_rate": 0.00014840773931243644, + "loss": 2.5488, + "step": 15984 + }, + { + "epoch": 0.68, + "grad_norm": 0.44921875, + "learning_rate": 0.00014840186959449092, + "loss": 2.6242, + "step": 15985 + }, + { + "epoch": 0.68, + "grad_norm": 0.384765625, + "learning_rate": 0.00014839599965875468, + "loss": 2.5636, + "step": 15986 + }, + { + "epoch": 0.68, + "grad_norm": 0.44921875, + "learning_rate": 0.0001483901295052541, + "loss": 2.5231, + "step": 15987 + }, + { + "epoch": 0.68, + "grad_norm": 0.478515625, + "learning_rate": 0.00014838425913401562, + "loss": 2.5684, + "step": 15988 + }, + { + "epoch": 0.68, + "grad_norm": 0.4296875, + "learning_rate": 0.0001483783885450657, + "loss": 2.4508, + "step": 15989 + }, + { + "epoch": 0.68, + "grad_norm": 0.458984375, + "learning_rate": 0.00014837251773843065, + "loss": 2.4412, + "step": 15990 + }, + { + "epoch": 0.68, + "grad_norm": 0.51171875, + "learning_rate": 0.00014836664671413698, + "loss": 2.5705, + "step": 15991 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.00014836077547221107, + "loss": 2.548, + "step": 15992 + }, + { + "epoch": 0.68, + "grad_norm": 0.490234375, + "learning_rate": 0.00014835490401267934, + "loss": 2.6039, + "step": 15993 + }, + { + "epoch": 0.68, + "grad_norm": 0.47265625, + "learning_rate": 0.00014834903233556826, + "loss": 2.5973, + "step": 15994 + }, + { + "epoch": 0.68, + "grad_norm": 0.50390625, + "learning_rate": 0.00014834316044090416, + "loss": 2.7238, + "step": 15995 + }, + { + "epoch": 0.68, + "grad_norm": 0.427734375, + "learning_rate": 0.0001483372883287135, + "loss": 2.6565, + "step": 15996 + }, + { + "epoch": 0.68, + "grad_norm": 0.46484375, + "learning_rate": 0.00014833141599902275, + "loss": 2.5741, + "step": 15997 + }, + { + "epoch": 0.68, + "grad_norm": 0.43359375, + "learning_rate": 0.00014832554345185826, + "loss": 2.3925, + "step": 15998 + }, + { + "epoch": 0.68, + "grad_norm": 0.53515625, + "learning_rate": 0.00014831967068724646, + "loss": 2.6156, + "step": 15999 + }, + { + "epoch": 0.68, + "grad_norm": 0.53125, + "learning_rate": 0.00014831379770521386, + "loss": 2.6527, + "step": 16000 + }, + { + "epoch": 0.68, + "grad_norm": 0.404296875, + "learning_rate": 0.00014830792450578676, + "loss": 2.4382, + "step": 16001 + }, + { + "epoch": 0.68, + "grad_norm": 0.419921875, + "learning_rate": 0.00014830205108899168, + "loss": 2.6585, + "step": 16002 + }, + { + "epoch": 0.68, + "grad_norm": 0.4921875, + "learning_rate": 0.00014829617745485505, + "loss": 2.528, + "step": 16003 + }, + { + "epoch": 0.68, + "grad_norm": 0.458984375, + "learning_rate": 0.00014829030360340325, + "loss": 2.5234, + "step": 16004 + }, + { + "epoch": 0.68, + "grad_norm": 0.43359375, + "learning_rate": 0.00014828442953466274, + "loss": 2.6344, + "step": 16005 + }, + { + "epoch": 0.68, + "grad_norm": 0.50390625, + "learning_rate": 0.00014827855524865994, + "loss": 2.5602, + "step": 16006 + }, + { + "epoch": 0.68, + "grad_norm": 0.5078125, + "learning_rate": 0.00014827268074542126, + "loss": 2.5919, + "step": 16007 + }, + { + "epoch": 0.68, + "grad_norm": 0.453125, + "learning_rate": 0.0001482668060249732, + "loss": 2.6739, + "step": 16008 + }, + { + "epoch": 0.68, + "grad_norm": 0.453125, + "learning_rate": 0.00014826093108734212, + "loss": 2.5517, + "step": 16009 + }, + { + "epoch": 0.68, + "grad_norm": 0.435546875, + "learning_rate": 0.0001482550559325545, + "loss": 2.5288, + "step": 16010 + }, + { + "epoch": 0.68, + "grad_norm": 0.408203125, + "learning_rate": 0.00014824918056063676, + "loss": 2.4707, + "step": 16011 + }, + { + "epoch": 0.68, + "grad_norm": 0.4609375, + "learning_rate": 0.00014824330497161537, + "loss": 2.6023, + "step": 16012 + }, + { + "epoch": 0.68, + "grad_norm": 0.41015625, + "learning_rate": 0.0001482374291655167, + "loss": 2.5231, + "step": 16013 + }, + { + "epoch": 0.68, + "grad_norm": 0.470703125, + "learning_rate": 0.00014823155314236725, + "loss": 2.5751, + "step": 16014 + }, + { + "epoch": 0.68, + "grad_norm": 0.369140625, + "learning_rate": 0.0001482256769021934, + "loss": 2.7254, + "step": 16015 + }, + { + "epoch": 0.68, + "grad_norm": 0.423828125, + "learning_rate": 0.00014821980044502168, + "loss": 2.5252, + "step": 16016 + }, + { + "epoch": 0.68, + "grad_norm": 0.435546875, + "learning_rate": 0.00014821392377087843, + "loss": 2.5778, + "step": 16017 + }, + { + "epoch": 0.68, + "grad_norm": 0.48046875, + "learning_rate": 0.00014820804687979019, + "loss": 2.6168, + "step": 16018 + }, + { + "epoch": 0.68, + "grad_norm": 0.4609375, + "learning_rate": 0.00014820216977178333, + "loss": 2.4903, + "step": 16019 + }, + { + "epoch": 0.68, + "grad_norm": 0.4375, + "learning_rate": 0.00014819629244688432, + "loss": 2.6868, + "step": 16020 + }, + { + "epoch": 0.68, + "grad_norm": 0.443359375, + "learning_rate": 0.00014819041490511965, + "loss": 2.5816, + "step": 16021 + }, + { + "epoch": 0.68, + "grad_norm": 0.453125, + "learning_rate": 0.00014818453714651568, + "loss": 2.7032, + "step": 16022 + }, + { + "epoch": 0.68, + "grad_norm": 0.46484375, + "learning_rate": 0.0001481786591710989, + "loss": 2.6645, + "step": 16023 + }, + { + "epoch": 0.68, + "grad_norm": 0.47265625, + "learning_rate": 0.0001481727809788958, + "loss": 2.6875, + "step": 16024 + }, + { + "epoch": 0.68, + "grad_norm": 0.482421875, + "learning_rate": 0.00014816690256993272, + "loss": 2.5928, + "step": 16025 + }, + { + "epoch": 0.68, + "grad_norm": 0.470703125, + "learning_rate": 0.00014816102394423625, + "loss": 2.464, + "step": 16026 + }, + { + "epoch": 0.68, + "grad_norm": 0.455078125, + "learning_rate": 0.00014815514510183275, + "loss": 2.6022, + "step": 16027 + }, + { + "epoch": 0.68, + "grad_norm": 0.49609375, + "learning_rate": 0.0001481492660427487, + "loss": 2.5384, + "step": 16028 + }, + { + "epoch": 0.68, + "grad_norm": 0.5390625, + "learning_rate": 0.00014814338676701053, + "loss": 2.7225, + "step": 16029 + }, + { + "epoch": 0.68, + "grad_norm": 0.439453125, + "learning_rate": 0.00014813750727464474, + "loss": 2.5483, + "step": 16030 + }, + { + "epoch": 0.68, + "grad_norm": 0.52734375, + "learning_rate": 0.00014813162756567772, + "loss": 2.5052, + "step": 16031 + }, + { + "epoch": 0.68, + "grad_norm": 0.546875, + "learning_rate": 0.00014812574764013603, + "loss": 2.7071, + "step": 16032 + }, + { + "epoch": 0.68, + "grad_norm": 0.50390625, + "learning_rate": 0.000148119867498046, + "loss": 2.6283, + "step": 16033 + }, + { + "epoch": 0.68, + "grad_norm": 0.486328125, + "learning_rate": 0.0001481139871394342, + "loss": 2.6757, + "step": 16034 + }, + { + "epoch": 0.68, + "grad_norm": 0.423828125, + "learning_rate": 0.000148108106564327, + "loss": 2.5121, + "step": 16035 + }, + { + "epoch": 0.68, + "grad_norm": 0.4375, + "learning_rate": 0.00014810222577275094, + "loss": 2.493, + "step": 16036 + }, + { + "epoch": 0.68, + "grad_norm": 0.48046875, + "learning_rate": 0.00014809634476473246, + "loss": 2.6789, + "step": 16037 + }, + { + "epoch": 0.68, + "grad_norm": 0.5, + "learning_rate": 0.00014809046354029794, + "loss": 2.5799, + "step": 16038 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.00014808458209947394, + "loss": 2.7728, + "step": 16039 + }, + { + "epoch": 0.68, + "grad_norm": 0.412109375, + "learning_rate": 0.0001480787004422869, + "loss": 2.5802, + "step": 16040 + }, + { + "epoch": 0.68, + "grad_norm": 0.447265625, + "learning_rate": 0.00014807281856876327, + "loss": 2.5588, + "step": 16041 + }, + { + "epoch": 0.68, + "grad_norm": 0.435546875, + "learning_rate": 0.00014806693647892953, + "loss": 2.6646, + "step": 16042 + }, + { + "epoch": 0.68, + "grad_norm": 0.45703125, + "learning_rate": 0.00014806105417281216, + "loss": 2.6304, + "step": 16043 + }, + { + "epoch": 0.68, + "grad_norm": 0.43359375, + "learning_rate": 0.00014805517165043756, + "loss": 2.6077, + "step": 16044 + }, + { + "epoch": 0.68, + "grad_norm": 0.482421875, + "learning_rate": 0.0001480492889118323, + "loss": 2.5088, + "step": 16045 + }, + { + "epoch": 0.68, + "grad_norm": 0.458984375, + "learning_rate": 0.00014804340595702276, + "loss": 2.5641, + "step": 16046 + }, + { + "epoch": 0.68, + "grad_norm": 0.46484375, + "learning_rate": 0.0001480375227860355, + "loss": 2.5501, + "step": 16047 + }, + { + "epoch": 0.68, + "grad_norm": 0.44921875, + "learning_rate": 0.0001480316393988969, + "loss": 2.528, + "step": 16048 + }, + { + "epoch": 0.68, + "grad_norm": 0.447265625, + "learning_rate": 0.0001480257557956335, + "loss": 2.601, + "step": 16049 + }, + { + "epoch": 0.68, + "grad_norm": 0.484375, + "learning_rate": 0.00014801987197627175, + "loss": 2.6587, + "step": 16050 + }, + { + "epoch": 0.68, + "grad_norm": 0.3984375, + "learning_rate": 0.0001480139879408381, + "loss": 2.4856, + "step": 16051 + }, + { + "epoch": 0.68, + "grad_norm": 0.486328125, + "learning_rate": 0.00014800810368935906, + "loss": 2.606, + "step": 16052 + }, + { + "epoch": 0.68, + "grad_norm": 0.439453125, + "learning_rate": 0.00014800221922186112, + "loss": 2.5617, + "step": 16053 + }, + { + "epoch": 0.68, + "grad_norm": 0.484375, + "learning_rate": 0.00014799633453837073, + "loss": 2.6632, + "step": 16054 + }, + { + "epoch": 0.68, + "grad_norm": 0.5, + "learning_rate": 0.00014799044963891434, + "loss": 2.7397, + "step": 16055 + }, + { + "epoch": 0.68, + "grad_norm": 0.44921875, + "learning_rate": 0.00014798456452351855, + "loss": 2.5911, + "step": 16056 + }, + { + "epoch": 0.68, + "grad_norm": 0.40234375, + "learning_rate": 0.00014797867919220966, + "loss": 2.4085, + "step": 16057 + }, + { + "epoch": 0.68, + "grad_norm": 0.396484375, + "learning_rate": 0.00014797279364501428, + "loss": 2.6366, + "step": 16058 + }, + { + "epoch": 0.68, + "grad_norm": 0.53515625, + "learning_rate": 0.00014796690788195887, + "loss": 2.645, + "step": 16059 + }, + { + "epoch": 0.68, + "grad_norm": 0.439453125, + "learning_rate": 0.0001479610219030699, + "loss": 2.6419, + "step": 16060 + }, + { + "epoch": 0.68, + "grad_norm": 0.470703125, + "learning_rate": 0.0001479551357083739, + "loss": 2.7582, + "step": 16061 + }, + { + "epoch": 0.68, + "grad_norm": 0.51953125, + "learning_rate": 0.00014794924929789727, + "loss": 2.641, + "step": 16062 + }, + { + "epoch": 0.68, + "grad_norm": 0.53125, + "learning_rate": 0.00014794336267166653, + "loss": 2.5782, + "step": 16063 + }, + { + "epoch": 0.68, + "grad_norm": 0.46875, + "learning_rate": 0.0001479374758297082, + "loss": 2.6332, + "step": 16064 + }, + { + "epoch": 0.68, + "grad_norm": 0.447265625, + "learning_rate": 0.00014793158877204874, + "loss": 2.6001, + "step": 16065 + }, + { + "epoch": 0.68, + "grad_norm": 0.400390625, + "learning_rate": 0.00014792570149871464, + "loss": 2.5113, + "step": 16066 + }, + { + "epoch": 0.68, + "grad_norm": 0.48046875, + "learning_rate": 0.00014791981400973244, + "loss": 2.6086, + "step": 16067 + }, + { + "epoch": 0.68, + "grad_norm": 0.451171875, + "learning_rate": 0.00014791392630512855, + "loss": 2.5046, + "step": 16068 + }, + { + "epoch": 0.68, + "grad_norm": 0.48046875, + "learning_rate": 0.00014790803838492952, + "loss": 2.6417, + "step": 16069 + }, + { + "epoch": 0.68, + "grad_norm": 0.435546875, + "learning_rate": 0.00014790215024916182, + "loss": 2.6244, + "step": 16070 + }, + { + "epoch": 0.68, + "grad_norm": 0.4453125, + "learning_rate": 0.00014789626189785194, + "loss": 2.5157, + "step": 16071 + }, + { + "epoch": 0.68, + "grad_norm": 0.44140625, + "learning_rate": 0.0001478903733310264, + "loss": 2.6085, + "step": 16072 + }, + { + "epoch": 0.68, + "grad_norm": 0.46875, + "learning_rate": 0.00014788448454871166, + "loss": 2.6172, + "step": 16073 + }, + { + "epoch": 0.68, + "grad_norm": 0.4609375, + "learning_rate": 0.00014787859555093427, + "loss": 2.6016, + "step": 16074 + }, + { + "epoch": 0.69, + "grad_norm": 0.44140625, + "learning_rate": 0.0001478727063377207, + "loss": 2.4711, + "step": 16075 + }, + { + "epoch": 0.69, + "grad_norm": 0.45703125, + "learning_rate": 0.0001478668169090974, + "loss": 2.5513, + "step": 16076 + }, + { + "epoch": 0.69, + "grad_norm": 0.435546875, + "learning_rate": 0.00014786092726509095, + "loss": 2.596, + "step": 16077 + }, + { + "epoch": 0.69, + "grad_norm": 0.42578125, + "learning_rate": 0.00014785503740572783, + "loss": 2.568, + "step": 16078 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.0001478491473310345, + "loss": 2.5066, + "step": 16079 + }, + { + "epoch": 0.69, + "grad_norm": 0.408203125, + "learning_rate": 0.00014784325704103753, + "loss": 2.4627, + "step": 16080 + }, + { + "epoch": 0.69, + "grad_norm": 0.51171875, + "learning_rate": 0.00014783736653576336, + "loss": 2.6755, + "step": 16081 + }, + { + "epoch": 0.69, + "grad_norm": 0.404296875, + "learning_rate": 0.00014783147581523852, + "loss": 2.4477, + "step": 16082 + }, + { + "epoch": 0.69, + "grad_norm": 0.408203125, + "learning_rate": 0.00014782558487948954, + "loss": 2.5669, + "step": 16083 + }, + { + "epoch": 0.69, + "grad_norm": 0.44921875, + "learning_rate": 0.0001478196937285429, + "loss": 2.7618, + "step": 16084 + }, + { + "epoch": 0.69, + "grad_norm": 0.52734375, + "learning_rate": 0.00014781380236242512, + "loss": 2.7122, + "step": 16085 + }, + { + "epoch": 0.69, + "grad_norm": 0.408203125, + "learning_rate": 0.0001478079107811627, + "loss": 2.5952, + "step": 16086 + }, + { + "epoch": 0.69, + "grad_norm": 0.5234375, + "learning_rate": 0.0001478020189847821, + "loss": 2.5751, + "step": 16087 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.00014779612697331, + "loss": 2.5569, + "step": 16088 + }, + { + "epoch": 0.69, + "grad_norm": 0.474609375, + "learning_rate": 0.00014779023474677269, + "loss": 2.5306, + "step": 16089 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.0001477843423051968, + "loss": 2.5538, + "step": 16090 + }, + { + "epoch": 0.69, + "grad_norm": 0.466796875, + "learning_rate": 0.0001477784496486089, + "loss": 2.5408, + "step": 16091 + }, + { + "epoch": 0.69, + "grad_norm": 0.40234375, + "learning_rate": 0.0001477725567770354, + "loss": 2.625, + "step": 16092 + }, + { + "epoch": 0.69, + "grad_norm": 0.47265625, + "learning_rate": 0.00014776666369050286, + "loss": 2.6643, + "step": 16093 + }, + { + "epoch": 0.69, + "grad_norm": 0.412109375, + "learning_rate": 0.00014776077038903778, + "loss": 2.5216, + "step": 16094 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.00014775487687266668, + "loss": 2.5114, + "step": 16095 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.0001477489831414161, + "loss": 2.6874, + "step": 16096 + }, + { + "epoch": 0.69, + "grad_norm": 0.478515625, + "learning_rate": 0.00014774308919531252, + "loss": 2.6287, + "step": 16097 + }, + { + "epoch": 0.69, + "grad_norm": 0.466796875, + "learning_rate": 0.00014773719503438252, + "loss": 2.6141, + "step": 16098 + }, + { + "epoch": 0.69, + "grad_norm": 0.3984375, + "learning_rate": 0.00014773130065865258, + "loss": 2.4818, + "step": 16099 + }, + { + "epoch": 0.69, + "grad_norm": 0.48828125, + "learning_rate": 0.0001477254060681492, + "loss": 2.6151, + "step": 16100 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.00014771951126289897, + "loss": 2.6103, + "step": 16101 + }, + { + "epoch": 0.69, + "grad_norm": 0.47265625, + "learning_rate": 0.00014771361624292838, + "loss": 2.5515, + "step": 16102 + }, + { + "epoch": 0.69, + "grad_norm": 0.46484375, + "learning_rate": 0.00014770772100826393, + "loss": 2.5912, + "step": 16103 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.00014770182555893222, + "loss": 2.629, + "step": 16104 + }, + { + "epoch": 0.69, + "grad_norm": 0.45703125, + "learning_rate": 0.0001476959298949597, + "loss": 2.6655, + "step": 16105 + }, + { + "epoch": 0.69, + "grad_norm": 0.482421875, + "learning_rate": 0.00014769003401637288, + "loss": 2.6627, + "step": 16106 + }, + { + "epoch": 0.69, + "grad_norm": 0.56640625, + "learning_rate": 0.0001476841379231984, + "loss": 2.7138, + "step": 16107 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.0001476782416154627, + "loss": 2.6416, + "step": 16108 + }, + { + "epoch": 0.69, + "grad_norm": 0.455078125, + "learning_rate": 0.00014767234509319234, + "loss": 2.5789, + "step": 16109 + }, + { + "epoch": 0.69, + "grad_norm": 0.4375, + "learning_rate": 0.00014766644835641386, + "loss": 2.6002, + "step": 16110 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.00014766055140515375, + "loss": 2.6029, + "step": 16111 + }, + { + "epoch": 0.69, + "grad_norm": 0.44921875, + "learning_rate": 0.00014765465423943863, + "loss": 2.5971, + "step": 16112 + }, + { + "epoch": 0.69, + "grad_norm": 0.51953125, + "learning_rate": 0.00014764875685929492, + "loss": 2.5323, + "step": 16113 + }, + { + "epoch": 0.69, + "grad_norm": 0.48828125, + "learning_rate": 0.00014764285926474924, + "loss": 2.6731, + "step": 16114 + }, + { + "epoch": 0.69, + "grad_norm": 0.486328125, + "learning_rate": 0.00014763696145582813, + "loss": 2.5479, + "step": 16115 + }, + { + "epoch": 0.69, + "grad_norm": 0.40234375, + "learning_rate": 0.00014763106343255807, + "loss": 2.5712, + "step": 16116 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.00014762516519496564, + "loss": 2.611, + "step": 16117 + }, + { + "epoch": 0.69, + "grad_norm": 0.458984375, + "learning_rate": 0.0001476192667430774, + "loss": 2.669, + "step": 16118 + }, + { + "epoch": 0.69, + "grad_norm": 0.43359375, + "learning_rate": 0.00014761336807691983, + "loss": 2.6117, + "step": 16119 + }, + { + "epoch": 0.69, + "grad_norm": 0.482421875, + "learning_rate": 0.0001476074691965195, + "loss": 2.6759, + "step": 16120 + }, + { + "epoch": 0.69, + "grad_norm": 0.46484375, + "learning_rate": 0.000147601570101903, + "loss": 2.6084, + "step": 16121 + }, + { + "epoch": 0.69, + "grad_norm": 0.47265625, + "learning_rate": 0.00014759567079309678, + "loss": 2.5838, + "step": 16122 + }, + { + "epoch": 0.69, + "grad_norm": 0.455078125, + "learning_rate": 0.00014758977127012746, + "loss": 2.5948, + "step": 16123 + }, + { + "epoch": 0.69, + "grad_norm": 0.45703125, + "learning_rate": 0.00014758387153302155, + "loss": 2.5707, + "step": 16124 + }, + { + "epoch": 0.69, + "grad_norm": 0.50390625, + "learning_rate": 0.00014757797158180565, + "loss": 2.5818, + "step": 16125 + }, + { + "epoch": 0.69, + "grad_norm": 0.484375, + "learning_rate": 0.00014757207141650623, + "loss": 2.4294, + "step": 16126 + }, + { + "epoch": 0.69, + "grad_norm": 0.490234375, + "learning_rate": 0.00014756617103714987, + "loss": 2.6895, + "step": 16127 + }, + { + "epoch": 0.69, + "grad_norm": 0.515625, + "learning_rate": 0.00014756027044376312, + "loss": 2.6231, + "step": 16128 + }, + { + "epoch": 0.69, + "grad_norm": 0.451171875, + "learning_rate": 0.00014755436963637255, + "loss": 2.3812, + "step": 16129 + }, + { + "epoch": 0.69, + "grad_norm": 0.515625, + "learning_rate": 0.0001475484686150047, + "loss": 2.4938, + "step": 16130 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.0001475425673796861, + "loss": 2.6166, + "step": 16131 + }, + { + "epoch": 0.69, + "grad_norm": 0.45703125, + "learning_rate": 0.00014753666593044334, + "loss": 2.6235, + "step": 16132 + }, + { + "epoch": 0.69, + "grad_norm": 0.427734375, + "learning_rate": 0.00014753076426730295, + "loss": 2.6848, + "step": 16133 + }, + { + "epoch": 0.69, + "grad_norm": 0.478515625, + "learning_rate": 0.00014752486239029148, + "loss": 2.5336, + "step": 16134 + }, + { + "epoch": 0.69, + "grad_norm": 0.458984375, + "learning_rate": 0.0001475189602994355, + "loss": 2.5641, + "step": 16135 + }, + { + "epoch": 0.69, + "grad_norm": 0.515625, + "learning_rate": 0.00014751305799476158, + "loss": 2.6483, + "step": 16136 + }, + { + "epoch": 0.69, + "grad_norm": 0.466796875, + "learning_rate": 0.00014750715547629626, + "loss": 2.6226, + "step": 16137 + }, + { + "epoch": 0.69, + "grad_norm": 0.380859375, + "learning_rate": 0.00014750125274406607, + "loss": 2.6047, + "step": 16138 + }, + { + "epoch": 0.69, + "grad_norm": 0.4453125, + "learning_rate": 0.00014749534979809764, + "loss": 2.4656, + "step": 16139 + }, + { + "epoch": 0.69, + "grad_norm": 0.5390625, + "learning_rate": 0.0001474894466384175, + "loss": 2.6021, + "step": 16140 + }, + { + "epoch": 0.69, + "grad_norm": 0.4375, + "learning_rate": 0.00014748354326505218, + "loss": 2.5295, + "step": 16141 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.0001474776396780283, + "loss": 2.5129, + "step": 16142 + }, + { + "epoch": 0.69, + "grad_norm": 0.53515625, + "learning_rate": 0.00014747173587737234, + "loss": 2.4966, + "step": 16143 + }, + { + "epoch": 0.69, + "grad_norm": 0.4921875, + "learning_rate": 0.00014746583186311095, + "loss": 2.6285, + "step": 16144 + }, + { + "epoch": 0.69, + "grad_norm": 0.5, + "learning_rate": 0.00014745992763527067, + "loss": 2.5629, + "step": 16145 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.000147454023193878, + "loss": 2.5316, + "step": 16146 + }, + { + "epoch": 0.69, + "grad_norm": 0.439453125, + "learning_rate": 0.00014744811853895966, + "loss": 2.4452, + "step": 16147 + }, + { + "epoch": 0.69, + "grad_norm": 0.455078125, + "learning_rate": 0.00014744221367054208, + "loss": 2.6168, + "step": 16148 + }, + { + "epoch": 0.69, + "grad_norm": 0.486328125, + "learning_rate": 0.00014743630858865187, + "loss": 2.7809, + "step": 16149 + }, + { + "epoch": 0.69, + "grad_norm": 0.474609375, + "learning_rate": 0.00014743040329331564, + "loss": 2.723, + "step": 16150 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.00014742449778455989, + "loss": 2.4152, + "step": 16151 + }, + { + "epoch": 0.69, + "grad_norm": 0.43359375, + "learning_rate": 0.00014741859206241124, + "loss": 2.5793, + "step": 16152 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.0001474126861268963, + "loss": 2.4604, + "step": 16153 + }, + { + "epoch": 0.69, + "grad_norm": 0.431640625, + "learning_rate": 0.00014740677997804154, + "loss": 2.5204, + "step": 16154 + }, + { + "epoch": 0.69, + "grad_norm": 0.57421875, + "learning_rate": 0.00014740087361587363, + "loss": 2.6243, + "step": 16155 + }, + { + "epoch": 0.69, + "grad_norm": 0.435546875, + "learning_rate": 0.0001473949670404191, + "loss": 2.5604, + "step": 16156 + }, + { + "epoch": 0.69, + "grad_norm": 0.490234375, + "learning_rate": 0.00014738906025170453, + "loss": 2.5445, + "step": 16157 + }, + { + "epoch": 0.69, + "grad_norm": 0.4453125, + "learning_rate": 0.0001473831532497565, + "loss": 2.6016, + "step": 16158 + }, + { + "epoch": 0.69, + "grad_norm": 0.498046875, + "learning_rate": 0.00014737724603460162, + "loss": 2.5343, + "step": 16159 + }, + { + "epoch": 0.69, + "grad_norm": 0.4609375, + "learning_rate": 0.00014737133860626646, + "loss": 2.6669, + "step": 16160 + }, + { + "epoch": 0.69, + "grad_norm": 0.4609375, + "learning_rate": 0.0001473654309647776, + "loss": 2.4715, + "step": 16161 + }, + { + "epoch": 0.69, + "grad_norm": 0.400390625, + "learning_rate": 0.00014735952311016153, + "loss": 2.4503, + "step": 16162 + }, + { + "epoch": 0.69, + "grad_norm": 0.5234375, + "learning_rate": 0.000147353615042445, + "loss": 2.6423, + "step": 16163 + }, + { + "epoch": 0.69, + "grad_norm": 0.48828125, + "learning_rate": 0.00014734770676165445, + "loss": 2.6082, + "step": 16164 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.00014734179826781657, + "loss": 2.4434, + "step": 16165 + }, + { + "epoch": 0.69, + "grad_norm": 0.4296875, + "learning_rate": 0.00014733588956095784, + "loss": 2.6399, + "step": 16166 + }, + { + "epoch": 0.69, + "grad_norm": 0.50390625, + "learning_rate": 0.00014732998064110497, + "loss": 2.7311, + "step": 16167 + }, + { + "epoch": 0.69, + "grad_norm": 0.42578125, + "learning_rate": 0.00014732407150828446, + "loss": 2.5911, + "step": 16168 + }, + { + "epoch": 0.69, + "grad_norm": 0.42578125, + "learning_rate": 0.00014731816216252293, + "loss": 2.6522, + "step": 16169 + }, + { + "epoch": 0.69, + "grad_norm": 0.48828125, + "learning_rate": 0.00014731225260384694, + "loss": 2.5062, + "step": 16170 + }, + { + "epoch": 0.69, + "grad_norm": 0.4140625, + "learning_rate": 0.00014730634283228314, + "loss": 2.5154, + "step": 16171 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.000147300432847858, + "loss": 2.498, + "step": 16172 + }, + { + "epoch": 0.69, + "grad_norm": 0.453125, + "learning_rate": 0.0001472945226505983, + "loss": 2.6692, + "step": 16173 + }, + { + "epoch": 0.69, + "grad_norm": 0.431640625, + "learning_rate": 0.00014728861224053047, + "loss": 2.5665, + "step": 16174 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.0001472827016176812, + "loss": 2.4821, + "step": 16175 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.00014727679078207706, + "loss": 2.7066, + "step": 16176 + }, + { + "epoch": 0.69, + "grad_norm": 0.45703125, + "learning_rate": 0.00014727087973374464, + "loss": 2.496, + "step": 16177 + }, + { + "epoch": 0.69, + "grad_norm": 0.46875, + "learning_rate": 0.00014726496847271048, + "loss": 2.4942, + "step": 16178 + }, + { + "epoch": 0.69, + "grad_norm": 0.4921875, + "learning_rate": 0.00014725905699900128, + "loss": 2.6958, + "step": 16179 + }, + { + "epoch": 0.69, + "grad_norm": 0.42578125, + "learning_rate": 0.00014725314531264359, + "loss": 2.5492, + "step": 16180 + }, + { + "epoch": 0.69, + "grad_norm": 0.4609375, + "learning_rate": 0.000147247233413664, + "loss": 2.6086, + "step": 16181 + }, + { + "epoch": 0.69, + "grad_norm": 0.43359375, + "learning_rate": 0.00014724132130208915, + "loss": 2.6563, + "step": 16182 + }, + { + "epoch": 0.69, + "grad_norm": 0.46484375, + "learning_rate": 0.00014723540897794562, + "loss": 2.5457, + "step": 16183 + }, + { + "epoch": 0.69, + "grad_norm": 0.427734375, + "learning_rate": 0.00014722949644125998, + "loss": 2.4716, + "step": 16184 + }, + { + "epoch": 0.69, + "grad_norm": 0.447265625, + "learning_rate": 0.00014722358369205892, + "loss": 2.7141, + "step": 16185 + }, + { + "epoch": 0.69, + "grad_norm": 0.494140625, + "learning_rate": 0.00014721767073036892, + "loss": 2.7024, + "step": 16186 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.00014721175755621672, + "loss": 2.6838, + "step": 16187 + }, + { + "epoch": 0.69, + "grad_norm": 0.4453125, + "learning_rate": 0.00014720584416962884, + "loss": 2.6569, + "step": 16188 + }, + { + "epoch": 0.69, + "grad_norm": 0.44140625, + "learning_rate": 0.0001471999305706319, + "loss": 2.6654, + "step": 16189 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.00014719401675925256, + "loss": 2.7271, + "step": 16190 + }, + { + "epoch": 0.69, + "grad_norm": 0.4296875, + "learning_rate": 0.00014718810273551736, + "loss": 2.6319, + "step": 16191 + }, + { + "epoch": 0.69, + "grad_norm": 0.42578125, + "learning_rate": 0.00014718218849945296, + "loss": 2.5621, + "step": 16192 + }, + { + "epoch": 0.69, + "grad_norm": 0.5078125, + "learning_rate": 0.00014717627405108594, + "loss": 2.5532, + "step": 16193 + }, + { + "epoch": 0.69, + "grad_norm": 0.45703125, + "learning_rate": 0.00014717035939044293, + "loss": 2.52, + "step": 16194 + }, + { + "epoch": 0.69, + "grad_norm": 0.447265625, + "learning_rate": 0.00014716444451755057, + "loss": 2.6174, + "step": 16195 + }, + { + "epoch": 0.69, + "grad_norm": 0.44921875, + "learning_rate": 0.0001471585294324354, + "loss": 2.6199, + "step": 16196 + }, + { + "epoch": 0.69, + "grad_norm": 0.435546875, + "learning_rate": 0.00014715261413512408, + "loss": 2.5157, + "step": 16197 + }, + { + "epoch": 0.69, + "grad_norm": 0.45703125, + "learning_rate": 0.00014714669862564327, + "loss": 2.6856, + "step": 16198 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.00014714078290401953, + "loss": 2.7225, + "step": 16199 + }, + { + "epoch": 0.69, + "grad_norm": 0.478515625, + "learning_rate": 0.0001471348669702795, + "loss": 2.6806, + "step": 16200 + }, + { + "epoch": 0.69, + "grad_norm": 0.453125, + "learning_rate": 0.0001471289508244498, + "loss": 2.5081, + "step": 16201 + }, + { + "epoch": 0.69, + "grad_norm": 0.47265625, + "learning_rate": 0.00014712303446655703, + "loss": 2.6457, + "step": 16202 + }, + { + "epoch": 0.69, + "grad_norm": 0.48046875, + "learning_rate": 0.00014711711789662784, + "loss": 2.591, + "step": 16203 + }, + { + "epoch": 0.69, + "grad_norm": 0.447265625, + "learning_rate": 0.00014711120111468882, + "loss": 2.4126, + "step": 16204 + }, + { + "epoch": 0.69, + "grad_norm": 0.447265625, + "learning_rate": 0.0001471052841207666, + "loss": 2.5444, + "step": 16205 + }, + { + "epoch": 0.69, + "grad_norm": 0.427734375, + "learning_rate": 0.00014709936691488786, + "loss": 2.5653, + "step": 16206 + }, + { + "epoch": 0.69, + "grad_norm": 0.470703125, + "learning_rate": 0.00014709344949707915, + "loss": 2.4941, + "step": 16207 + }, + { + "epoch": 0.69, + "grad_norm": 0.46875, + "learning_rate": 0.00014708753186736713, + "loss": 2.7066, + "step": 16208 + }, + { + "epoch": 0.69, + "grad_norm": 0.44140625, + "learning_rate": 0.00014708161402577843, + "loss": 2.5285, + "step": 16209 + }, + { + "epoch": 0.69, + "grad_norm": 0.50390625, + "learning_rate": 0.00014707569597233967, + "loss": 2.5819, + "step": 16210 + }, + { + "epoch": 0.69, + "grad_norm": 0.39453125, + "learning_rate": 0.00014706977770707753, + "loss": 2.5248, + "step": 16211 + }, + { + "epoch": 0.69, + "grad_norm": 0.4296875, + "learning_rate": 0.0001470638592300185, + "loss": 2.6134, + "step": 16212 + }, + { + "epoch": 0.69, + "grad_norm": 0.4921875, + "learning_rate": 0.00014705794054118936, + "loss": 2.7689, + "step": 16213 + }, + { + "epoch": 0.69, + "grad_norm": 0.396484375, + "learning_rate": 0.00014705202164061668, + "loss": 2.5543, + "step": 16214 + }, + { + "epoch": 0.69, + "grad_norm": 0.44921875, + "learning_rate": 0.00014704610252832712, + "loss": 2.5396, + "step": 16215 + }, + { + "epoch": 0.69, + "grad_norm": 0.466796875, + "learning_rate": 0.00014704018320434728, + "loss": 2.6468, + "step": 16216 + }, + { + "epoch": 0.69, + "grad_norm": 0.44921875, + "learning_rate": 0.00014703426366870378, + "loss": 2.5469, + "step": 16217 + }, + { + "epoch": 0.69, + "grad_norm": 0.435546875, + "learning_rate": 0.00014702834392142327, + "loss": 2.5163, + "step": 16218 + }, + { + "epoch": 0.69, + "grad_norm": 0.423828125, + "learning_rate": 0.00014702242396253247, + "loss": 2.5972, + "step": 16219 + }, + { + "epoch": 0.69, + "grad_norm": 0.451171875, + "learning_rate": 0.00014701650379205788, + "loss": 2.6658, + "step": 16220 + }, + { + "epoch": 0.69, + "grad_norm": 0.462890625, + "learning_rate": 0.00014701058341002625, + "loss": 2.6532, + "step": 16221 + }, + { + "epoch": 0.69, + "grad_norm": 0.376953125, + "learning_rate": 0.00014700466281646415, + "loss": 2.501, + "step": 16222 + }, + { + "epoch": 0.69, + "grad_norm": 0.41015625, + "learning_rate": 0.00014699874201139825, + "loss": 2.491, + "step": 16223 + }, + { + "epoch": 0.69, + "grad_norm": 0.408203125, + "learning_rate": 0.0001469928209948552, + "loss": 2.5862, + "step": 16224 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.00014698689976686162, + "loss": 2.6112, + "step": 16225 + }, + { + "epoch": 0.69, + "grad_norm": 0.474609375, + "learning_rate": 0.00014698097832744413, + "loss": 2.747, + "step": 16226 + }, + { + "epoch": 0.69, + "grad_norm": 0.458984375, + "learning_rate": 0.00014697505667662946, + "loss": 2.547, + "step": 16227 + }, + { + "epoch": 0.69, + "grad_norm": 0.4609375, + "learning_rate": 0.00014696913481444416, + "loss": 2.4911, + "step": 16228 + }, + { + "epoch": 0.69, + "grad_norm": 0.447265625, + "learning_rate": 0.00014696321274091493, + "loss": 2.5267, + "step": 16229 + }, + { + "epoch": 0.69, + "grad_norm": 0.43359375, + "learning_rate": 0.00014695729045606843, + "loss": 2.5406, + "step": 16230 + }, + { + "epoch": 0.69, + "grad_norm": 0.490234375, + "learning_rate": 0.00014695136795993128, + "loss": 2.5074, + "step": 16231 + }, + { + "epoch": 0.69, + "grad_norm": 0.478515625, + "learning_rate": 0.0001469454452525301, + "loss": 2.5711, + "step": 16232 + }, + { + "epoch": 0.69, + "grad_norm": 0.44140625, + "learning_rate": 0.00014693952233389162, + "loss": 2.6712, + "step": 16233 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.0001469335992040424, + "loss": 2.6186, + "step": 16234 + }, + { + "epoch": 0.69, + "grad_norm": 0.43359375, + "learning_rate": 0.00014692767586300918, + "loss": 2.5064, + "step": 16235 + }, + { + "epoch": 0.69, + "grad_norm": 0.46875, + "learning_rate": 0.0001469217523108185, + "loss": 2.4695, + "step": 16236 + }, + { + "epoch": 0.69, + "grad_norm": 0.453125, + "learning_rate": 0.00014691582854749714, + "loss": 2.5087, + "step": 16237 + }, + { + "epoch": 0.69, + "grad_norm": 0.5234375, + "learning_rate": 0.0001469099045730717, + "loss": 2.4812, + "step": 16238 + }, + { + "epoch": 0.69, + "grad_norm": 0.474609375, + "learning_rate": 0.0001469039803875688, + "loss": 2.6493, + "step": 16239 + }, + { + "epoch": 0.69, + "grad_norm": 0.435546875, + "learning_rate": 0.00014689805599101514, + "loss": 2.5006, + "step": 16240 + }, + { + "epoch": 0.69, + "grad_norm": 0.5078125, + "learning_rate": 0.0001468921313834374, + "loss": 2.7442, + "step": 16241 + }, + { + "epoch": 0.69, + "grad_norm": 0.46484375, + "learning_rate": 0.00014688620656486216, + "loss": 2.5843, + "step": 16242 + }, + { + "epoch": 0.69, + "grad_norm": 0.416015625, + "learning_rate": 0.00014688028153531614, + "loss": 2.5444, + "step": 16243 + }, + { + "epoch": 0.69, + "grad_norm": 0.419921875, + "learning_rate": 0.000146874356294826, + "loss": 2.594, + "step": 16244 + }, + { + "epoch": 0.69, + "grad_norm": 0.423828125, + "learning_rate": 0.00014686843084341837, + "loss": 2.4915, + "step": 16245 + }, + { + "epoch": 0.69, + "grad_norm": 0.396484375, + "learning_rate": 0.00014686250518111994, + "loss": 2.5666, + "step": 16246 + }, + { + "epoch": 0.69, + "grad_norm": 0.43359375, + "learning_rate": 0.00014685657930795735, + "loss": 2.5765, + "step": 16247 + }, + { + "epoch": 0.69, + "grad_norm": 0.5, + "learning_rate": 0.00014685065322395728, + "loss": 2.6271, + "step": 16248 + }, + { + "epoch": 0.69, + "grad_norm": 0.5, + "learning_rate": 0.0001468447269291464, + "loss": 2.5734, + "step": 16249 + }, + { + "epoch": 0.69, + "grad_norm": 0.4765625, + "learning_rate": 0.00014683880042355138, + "loss": 2.6588, + "step": 16250 + }, + { + "epoch": 0.69, + "grad_norm": 0.42578125, + "learning_rate": 0.00014683287370719888, + "loss": 2.7792, + "step": 16251 + }, + { + "epoch": 0.69, + "grad_norm": 0.4453125, + "learning_rate": 0.00014682694678011554, + "loss": 2.4688, + "step": 16252 + }, + { + "epoch": 0.69, + "grad_norm": 0.46484375, + "learning_rate": 0.00014682101964232804, + "loss": 2.4729, + "step": 16253 + }, + { + "epoch": 0.69, + "grad_norm": 0.490234375, + "learning_rate": 0.0001468150922938631, + "loss": 2.5511, + "step": 16254 + }, + { + "epoch": 0.69, + "grad_norm": 0.404296875, + "learning_rate": 0.00014680916473474732, + "loss": 2.5643, + "step": 16255 + }, + { + "epoch": 0.69, + "grad_norm": 0.427734375, + "learning_rate": 0.00014680323696500741, + "loss": 2.5506, + "step": 16256 + }, + { + "epoch": 0.69, + "grad_norm": 0.482421875, + "learning_rate": 0.00014679730898467007, + "loss": 2.6209, + "step": 16257 + }, + { + "epoch": 0.69, + "grad_norm": 0.408203125, + "learning_rate": 0.0001467913807937619, + "loss": 2.5549, + "step": 16258 + }, + { + "epoch": 0.69, + "grad_norm": 0.4453125, + "learning_rate": 0.00014678545239230965, + "loss": 2.3727, + "step": 16259 + }, + { + "epoch": 0.69, + "grad_norm": 0.51171875, + "learning_rate": 0.00014677952378033997, + "loss": 2.7613, + "step": 16260 + }, + { + "epoch": 0.69, + "grad_norm": 0.427734375, + "learning_rate": 0.00014677359495787948, + "loss": 2.3253, + "step": 16261 + }, + { + "epoch": 0.69, + "grad_norm": 0.486328125, + "learning_rate": 0.00014676766592495498, + "loss": 2.6999, + "step": 16262 + }, + { + "epoch": 0.69, + "grad_norm": 0.443359375, + "learning_rate": 0.00014676173668159298, + "loss": 2.4838, + "step": 16263 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.00014675580722782033, + "loss": 2.6645, + "step": 16264 + }, + { + "epoch": 0.69, + "grad_norm": 0.470703125, + "learning_rate": 0.0001467498775636636, + "loss": 2.49, + "step": 16265 + }, + { + "epoch": 0.69, + "grad_norm": 0.455078125, + "learning_rate": 0.00014674394768914955, + "loss": 2.7052, + "step": 16266 + }, + { + "epoch": 0.69, + "grad_norm": 0.474609375, + "learning_rate": 0.00014673801760430478, + "loss": 2.5969, + "step": 16267 + }, + { + "epoch": 0.69, + "grad_norm": 0.51171875, + "learning_rate": 0.000146732087309156, + "loss": 2.6256, + "step": 16268 + }, + { + "epoch": 0.69, + "grad_norm": 0.4375, + "learning_rate": 0.00014672615680372993, + "loss": 2.5347, + "step": 16269 + }, + { + "epoch": 0.69, + "grad_norm": 0.46484375, + "learning_rate": 0.00014672022608805323, + "loss": 2.5905, + "step": 16270 + }, + { + "epoch": 0.69, + "grad_norm": 0.421875, + "learning_rate": 0.00014671429516215255, + "loss": 2.5125, + "step": 16271 + }, + { + "epoch": 0.69, + "grad_norm": 0.51171875, + "learning_rate": 0.00014670836402605467, + "loss": 2.6916, + "step": 16272 + }, + { + "epoch": 0.69, + "grad_norm": 0.455078125, + "learning_rate": 0.00014670243267978617, + "loss": 2.5989, + "step": 16273 + }, + { + "epoch": 0.69, + "grad_norm": 0.44921875, + "learning_rate": 0.00014669650112337384, + "loss": 2.6748, + "step": 16274 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.0001466905693568443, + "loss": 2.6953, + "step": 16275 + }, + { + "epoch": 0.69, + "grad_norm": 0.43359375, + "learning_rate": 0.00014668463738022424, + "loss": 2.4895, + "step": 16276 + }, + { + "epoch": 0.69, + "grad_norm": 0.431640625, + "learning_rate": 0.00014667870519354038, + "loss": 2.588, + "step": 16277 + }, + { + "epoch": 0.69, + "grad_norm": 0.427734375, + "learning_rate": 0.00014667277279681944, + "loss": 2.4387, + "step": 16278 + }, + { + "epoch": 0.69, + "grad_norm": 0.447265625, + "learning_rate": 0.00014666684019008804, + "loss": 2.5323, + "step": 16279 + }, + { + "epoch": 0.69, + "grad_norm": 0.44140625, + "learning_rate": 0.00014666090737337295, + "loss": 2.6648, + "step": 16280 + }, + { + "epoch": 0.69, + "grad_norm": 0.4453125, + "learning_rate": 0.00014665497434670078, + "loss": 2.5498, + "step": 16281 + }, + { + "epoch": 0.69, + "grad_norm": 0.453125, + "learning_rate": 0.00014664904111009834, + "loss": 2.5037, + "step": 16282 + }, + { + "epoch": 0.69, + "grad_norm": 0.431640625, + "learning_rate": 0.00014664310766359222, + "loss": 2.4668, + "step": 16283 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.00014663717400720917, + "loss": 2.6115, + "step": 16284 + }, + { + "epoch": 0.69, + "grad_norm": 0.5234375, + "learning_rate": 0.00014663124014097586, + "loss": 2.663, + "step": 16285 + }, + { + "epoch": 0.69, + "grad_norm": 0.451171875, + "learning_rate": 0.00014662530606491907, + "loss": 2.6694, + "step": 16286 + }, + { + "epoch": 0.69, + "grad_norm": 0.380859375, + "learning_rate": 0.00014661937177906537, + "loss": 2.5343, + "step": 16287 + }, + { + "epoch": 0.69, + "grad_norm": 0.51171875, + "learning_rate": 0.00014661343728344158, + "loss": 2.4947, + "step": 16288 + }, + { + "epoch": 0.69, + "grad_norm": 0.408203125, + "learning_rate": 0.00014660750257807433, + "loss": 2.5781, + "step": 16289 + }, + { + "epoch": 0.69, + "grad_norm": 0.39453125, + "learning_rate": 0.00014660156766299037, + "loss": 2.5603, + "step": 16290 + }, + { + "epoch": 0.69, + "grad_norm": 0.46484375, + "learning_rate": 0.00014659563253821643, + "loss": 2.6451, + "step": 16291 + }, + { + "epoch": 0.69, + "grad_norm": 0.48046875, + "learning_rate": 0.00014658969720377911, + "loss": 2.593, + "step": 16292 + }, + { + "epoch": 0.69, + "grad_norm": 0.54296875, + "learning_rate": 0.00014658376165970518, + "loss": 2.6261, + "step": 16293 + }, + { + "epoch": 0.69, + "grad_norm": 0.474609375, + "learning_rate": 0.0001465778259060214, + "loss": 2.5803, + "step": 16294 + }, + { + "epoch": 0.69, + "grad_norm": 0.44140625, + "learning_rate": 0.00014657188994275435, + "loss": 2.565, + "step": 16295 + }, + { + "epoch": 0.69, + "grad_norm": 0.490234375, + "learning_rate": 0.00014656595376993085, + "loss": 2.5843, + "step": 16296 + }, + { + "epoch": 0.69, + "grad_norm": 0.4921875, + "learning_rate": 0.00014656001738757758, + "loss": 2.6285, + "step": 16297 + }, + { + "epoch": 0.69, + "grad_norm": 0.53515625, + "learning_rate": 0.00014655408079572126, + "loss": 2.5565, + "step": 16298 + }, + { + "epoch": 0.69, + "grad_norm": 0.4609375, + "learning_rate": 0.00014654814399438858, + "loss": 2.5874, + "step": 16299 + }, + { + "epoch": 0.69, + "grad_norm": 0.49609375, + "learning_rate": 0.00014654220698360625, + "loss": 2.6484, + "step": 16300 + }, + { + "epoch": 0.69, + "grad_norm": 0.423828125, + "learning_rate": 0.000146536269763401, + "loss": 2.5447, + "step": 16301 + }, + { + "epoch": 0.69, + "grad_norm": 0.490234375, + "learning_rate": 0.00014653033233379958, + "loss": 2.561, + "step": 16302 + }, + { + "epoch": 0.69, + "grad_norm": 0.443359375, + "learning_rate": 0.00014652439469482861, + "loss": 2.5768, + "step": 16303 + }, + { + "epoch": 0.69, + "grad_norm": 0.5, + "learning_rate": 0.0001465184568465149, + "loss": 2.5461, + "step": 16304 + }, + { + "epoch": 0.69, + "grad_norm": 0.40625, + "learning_rate": 0.00014651251878888515, + "loss": 2.5053, + "step": 16305 + }, + { + "epoch": 0.69, + "grad_norm": 0.431640625, + "learning_rate": 0.00014650658052196605, + "loss": 2.5869, + "step": 16306 + }, + { + "epoch": 0.69, + "grad_norm": 0.408203125, + "learning_rate": 0.00014650064204578433, + "loss": 2.5997, + "step": 16307 + }, + { + "epoch": 0.69, + "grad_norm": 0.44921875, + "learning_rate": 0.00014649470336036672, + "loss": 2.5891, + "step": 16308 + }, + { + "epoch": 0.69, + "grad_norm": 0.48046875, + "learning_rate": 0.0001464887644657399, + "loss": 2.5563, + "step": 16309 + }, + { + "epoch": 0.7, + "grad_norm": 0.45703125, + "learning_rate": 0.0001464828253619307, + "loss": 2.5661, + "step": 16310 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.0001464768860489657, + "loss": 2.5826, + "step": 16311 + }, + { + "epoch": 0.7, + "grad_norm": 0.4140625, + "learning_rate": 0.00014647094652687175, + "loss": 2.4699, + "step": 16312 + }, + { + "epoch": 0.7, + "grad_norm": 0.4296875, + "learning_rate": 0.00014646500679567551, + "loss": 2.6955, + "step": 16313 + }, + { + "epoch": 0.7, + "grad_norm": 0.41015625, + "learning_rate": 0.00014645906685540374, + "loss": 2.5428, + "step": 16314 + }, + { + "epoch": 0.7, + "grad_norm": 0.4296875, + "learning_rate": 0.0001464531267060831, + "loss": 2.6513, + "step": 16315 + }, + { + "epoch": 0.7, + "grad_norm": 0.46484375, + "learning_rate": 0.0001464471863477404, + "loss": 2.5445, + "step": 16316 + }, + { + "epoch": 0.7, + "grad_norm": 0.4765625, + "learning_rate": 0.00014644124578040233, + "loss": 2.6174, + "step": 16317 + }, + { + "epoch": 0.7, + "grad_norm": 0.474609375, + "learning_rate": 0.00014643530500409565, + "loss": 2.5966, + "step": 16318 + }, + { + "epoch": 0.7, + "grad_norm": 0.419921875, + "learning_rate": 0.00014642936401884702, + "loss": 2.6591, + "step": 16319 + }, + { + "epoch": 0.7, + "grad_norm": 0.423828125, + "learning_rate": 0.00014642342282468324, + "loss": 2.5984, + "step": 16320 + }, + { + "epoch": 0.7, + "grad_norm": 0.41015625, + "learning_rate": 0.00014641748142163103, + "loss": 2.7035, + "step": 16321 + }, + { + "epoch": 0.7, + "grad_norm": 0.439453125, + "learning_rate": 0.0001464115398097171, + "loss": 2.5004, + "step": 16322 + }, + { + "epoch": 0.7, + "grad_norm": 0.435546875, + "learning_rate": 0.00014640559798896822, + "loss": 2.5894, + "step": 16323 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.0001463996559594111, + "loss": 2.5833, + "step": 16324 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.00014639371372107246, + "loss": 2.3887, + "step": 16325 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.00014638777127397914, + "loss": 2.5904, + "step": 16326 + }, + { + "epoch": 0.7, + "grad_norm": 0.431640625, + "learning_rate": 0.0001463818286181577, + "loss": 2.6245, + "step": 16327 + }, + { + "epoch": 0.7, + "grad_norm": 0.4609375, + "learning_rate": 0.0001463758857536351, + "loss": 2.6509, + "step": 16328 + }, + { + "epoch": 0.7, + "grad_norm": 0.5, + "learning_rate": 0.00014636994268043787, + "loss": 2.5567, + "step": 16329 + }, + { + "epoch": 0.7, + "grad_norm": 0.490234375, + "learning_rate": 0.00014636399939859288, + "loss": 2.5635, + "step": 16330 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.00014635805590812679, + "loss": 2.5303, + "step": 16331 + }, + { + "epoch": 0.7, + "grad_norm": 0.484375, + "learning_rate": 0.00014635211220906646, + "loss": 2.6151, + "step": 16332 + }, + { + "epoch": 0.7, + "grad_norm": 0.49609375, + "learning_rate": 0.0001463461683014385, + "loss": 2.6819, + "step": 16333 + }, + { + "epoch": 0.7, + "grad_norm": 0.466796875, + "learning_rate": 0.0001463402241852698, + "loss": 2.6262, + "step": 16334 + }, + { + "epoch": 0.7, + "grad_norm": 0.486328125, + "learning_rate": 0.00014633427986058696, + "loss": 2.6617, + "step": 16335 + }, + { + "epoch": 0.7, + "grad_norm": 0.51171875, + "learning_rate": 0.0001463283353274168, + "loss": 2.6937, + "step": 16336 + }, + { + "epoch": 0.7, + "grad_norm": 0.48046875, + "learning_rate": 0.0001463223905857861, + "loss": 2.6128, + "step": 16337 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.00014631644563572156, + "loss": 2.5237, + "step": 16338 + }, + { + "epoch": 0.7, + "grad_norm": 0.54296875, + "learning_rate": 0.0001463105004772499, + "loss": 2.8655, + "step": 16339 + }, + { + "epoch": 0.7, + "grad_norm": 0.435546875, + "learning_rate": 0.00014630455511039796, + "loss": 2.4752, + "step": 16340 + }, + { + "epoch": 0.7, + "grad_norm": 0.4375, + "learning_rate": 0.00014629860953519242, + "loss": 2.4062, + "step": 16341 + }, + { + "epoch": 0.7, + "grad_norm": 0.48046875, + "learning_rate": 0.0001462926637516601, + "loss": 2.5631, + "step": 16342 + }, + { + "epoch": 0.7, + "grad_norm": 0.412109375, + "learning_rate": 0.00014628671775982762, + "loss": 2.5673, + "step": 16343 + }, + { + "epoch": 0.7, + "grad_norm": 0.404296875, + "learning_rate": 0.00014628077155972192, + "loss": 2.4241, + "step": 16344 + }, + { + "epoch": 0.7, + "grad_norm": 0.4921875, + "learning_rate": 0.0001462748251513696, + "loss": 2.6671, + "step": 16345 + }, + { + "epoch": 0.7, + "grad_norm": 0.47265625, + "learning_rate": 0.00014626887853479753, + "loss": 2.5342, + "step": 16346 + }, + { + "epoch": 0.7, + "grad_norm": 0.455078125, + "learning_rate": 0.00014626293171003234, + "loss": 2.635, + "step": 16347 + }, + { + "epoch": 0.7, + "grad_norm": 0.4453125, + "learning_rate": 0.00014625698467710094, + "loss": 2.5768, + "step": 16348 + }, + { + "epoch": 0.7, + "grad_norm": 0.51171875, + "learning_rate": 0.00014625103743602996, + "loss": 2.6549, + "step": 16349 + }, + { + "epoch": 0.7, + "grad_norm": 0.46484375, + "learning_rate": 0.00014624508998684627, + "loss": 2.4974, + "step": 16350 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014623914232957653, + "loss": 2.4609, + "step": 16351 + }, + { + "epoch": 0.7, + "grad_norm": 0.5078125, + "learning_rate": 0.00014623319446424758, + "loss": 2.4482, + "step": 16352 + }, + { + "epoch": 0.7, + "grad_norm": 0.41015625, + "learning_rate": 0.00014622724639088612, + "loss": 2.5487, + "step": 16353 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.00014622129810951896, + "loss": 2.4575, + "step": 16354 + }, + { + "epoch": 0.7, + "grad_norm": 0.462890625, + "learning_rate": 0.00014621534962017281, + "loss": 2.5293, + "step": 16355 + }, + { + "epoch": 0.7, + "grad_norm": 0.48046875, + "learning_rate": 0.00014620940092287455, + "loss": 2.6445, + "step": 16356 + }, + { + "epoch": 0.7, + "grad_norm": 0.470703125, + "learning_rate": 0.00014620345201765082, + "loss": 2.6394, + "step": 16357 + }, + { + "epoch": 0.7, + "grad_norm": 0.50390625, + "learning_rate": 0.00014619750290452847, + "loss": 2.4256, + "step": 16358 + }, + { + "epoch": 0.7, + "grad_norm": 0.423828125, + "learning_rate": 0.00014619155358353426, + "loss": 2.6071, + "step": 16359 + }, + { + "epoch": 0.7, + "grad_norm": 0.462890625, + "learning_rate": 0.0001461856040546949, + "loss": 2.5962, + "step": 16360 + }, + { + "epoch": 0.7, + "grad_norm": 0.4453125, + "learning_rate": 0.00014617965431803725, + "loss": 2.6403, + "step": 16361 + }, + { + "epoch": 0.7, + "grad_norm": 0.4765625, + "learning_rate": 0.00014617370437358795, + "loss": 2.7152, + "step": 16362 + }, + { + "epoch": 0.7, + "grad_norm": 0.451171875, + "learning_rate": 0.0001461677542213739, + "loss": 2.5761, + "step": 16363 + }, + { + "epoch": 0.7, + "grad_norm": 0.4921875, + "learning_rate": 0.00014616180386142183, + "loss": 2.6641, + "step": 16364 + }, + { + "epoch": 0.7, + "grad_norm": 0.40625, + "learning_rate": 0.0001461558532937585, + "loss": 2.7205, + "step": 16365 + }, + { + "epoch": 0.7, + "grad_norm": 0.474609375, + "learning_rate": 0.0001461499025184107, + "loss": 2.5558, + "step": 16366 + }, + { + "epoch": 0.7, + "grad_norm": 0.42578125, + "learning_rate": 0.00014614395153540527, + "loss": 2.6228, + "step": 16367 + }, + { + "epoch": 0.7, + "grad_norm": 0.390625, + "learning_rate": 0.00014613800034476884, + "loss": 2.6293, + "step": 16368 + }, + { + "epoch": 0.7, + "grad_norm": 0.50390625, + "learning_rate": 0.0001461320489465283, + "loss": 2.6321, + "step": 16369 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.00014612609734071037, + "loss": 2.4862, + "step": 16370 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.0001461201455273419, + "loss": 2.4381, + "step": 16371 + }, + { + "epoch": 0.7, + "grad_norm": 0.515625, + "learning_rate": 0.0001461141935064496, + "loss": 2.5334, + "step": 16372 + }, + { + "epoch": 0.7, + "grad_norm": 0.470703125, + "learning_rate": 0.00014610824127806032, + "loss": 2.5801, + "step": 16373 + }, + { + "epoch": 0.7, + "grad_norm": 0.466796875, + "learning_rate": 0.00014610228884220073, + "loss": 2.5296, + "step": 16374 + }, + { + "epoch": 0.7, + "grad_norm": 0.49609375, + "learning_rate": 0.0001460963361988978, + "loss": 2.6874, + "step": 16375 + }, + { + "epoch": 0.7, + "grad_norm": 0.4453125, + "learning_rate": 0.0001460903833481781, + "loss": 2.602, + "step": 16376 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.0001460844302900686, + "loss": 2.4683, + "step": 16377 + }, + { + "epoch": 0.7, + "grad_norm": 0.474609375, + "learning_rate": 0.00014607847702459594, + "loss": 2.6732, + "step": 16378 + }, + { + "epoch": 0.7, + "grad_norm": 0.44140625, + "learning_rate": 0.000146072523551787, + "loss": 2.4641, + "step": 16379 + }, + { + "epoch": 0.7, + "grad_norm": 0.5078125, + "learning_rate": 0.00014606656987166855, + "loss": 2.6557, + "step": 16380 + }, + { + "epoch": 0.7, + "grad_norm": 0.46484375, + "learning_rate": 0.00014606061598426736, + "loss": 2.48, + "step": 16381 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.0001460546618896102, + "loss": 2.5149, + "step": 16382 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.00014604870758772398, + "loss": 2.616, + "step": 16383 + }, + { + "epoch": 0.7, + "grad_norm": 0.5234375, + "learning_rate": 0.00014604275307863532, + "loss": 2.6396, + "step": 16384 + }, + { + "epoch": 0.7, + "grad_norm": 0.4765625, + "learning_rate": 0.00014603679836237114, + "loss": 2.5378, + "step": 16385 + }, + { + "epoch": 0.7, + "grad_norm": 0.51953125, + "learning_rate": 0.00014603084343895817, + "loss": 2.5133, + "step": 16386 + }, + { + "epoch": 0.7, + "grad_norm": 0.4453125, + "learning_rate": 0.00014602488830842323, + "loss": 2.5628, + "step": 16387 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.00014601893297079313, + "loss": 2.6417, + "step": 16388 + }, + { + "epoch": 0.7, + "grad_norm": 0.515625, + "learning_rate": 0.00014601297742609465, + "loss": 2.6674, + "step": 16389 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.00014600702167435456, + "loss": 2.6089, + "step": 16390 + }, + { + "epoch": 0.7, + "grad_norm": 0.41796875, + "learning_rate": 0.00014600106571559972, + "loss": 2.4649, + "step": 16391 + }, + { + "epoch": 0.7, + "grad_norm": 0.494140625, + "learning_rate": 0.00014599510954985687, + "loss": 2.5664, + "step": 16392 + }, + { + "epoch": 0.7, + "grad_norm": 0.466796875, + "learning_rate": 0.00014598915317715286, + "loss": 2.6299, + "step": 16393 + }, + { + "epoch": 0.7, + "grad_norm": 0.42578125, + "learning_rate": 0.00014598319659751445, + "loss": 2.6306, + "step": 16394 + }, + { + "epoch": 0.7, + "grad_norm": 0.5, + "learning_rate": 0.00014597723981096846, + "loss": 2.6129, + "step": 16395 + }, + { + "epoch": 0.7, + "grad_norm": 0.41796875, + "learning_rate": 0.0001459712828175417, + "loss": 2.6328, + "step": 16396 + }, + { + "epoch": 0.7, + "grad_norm": 0.4921875, + "learning_rate": 0.00014596532561726098, + "loss": 2.5779, + "step": 16397 + }, + { + "epoch": 0.7, + "grad_norm": 0.412109375, + "learning_rate": 0.00014595936821015304, + "loss": 2.5698, + "step": 16398 + }, + { + "epoch": 0.7, + "grad_norm": 0.439453125, + "learning_rate": 0.00014595341059624482, + "loss": 2.5818, + "step": 16399 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.00014594745277556298, + "loss": 2.4253, + "step": 16400 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.00014594149474813443, + "loss": 2.5334, + "step": 16401 + }, + { + "epoch": 0.7, + "grad_norm": 0.4453125, + "learning_rate": 0.00014593553651398594, + "loss": 2.4661, + "step": 16402 + }, + { + "epoch": 0.7, + "grad_norm": 0.435546875, + "learning_rate": 0.0001459295780731443, + "loss": 2.5329, + "step": 16403 + }, + { + "epoch": 0.7, + "grad_norm": 0.435546875, + "learning_rate": 0.00014592361942563636, + "loss": 2.5984, + "step": 16404 + }, + { + "epoch": 0.7, + "grad_norm": 0.421875, + "learning_rate": 0.0001459176605714889, + "loss": 2.5037, + "step": 16405 + }, + { + "epoch": 0.7, + "grad_norm": 0.4921875, + "learning_rate": 0.00014591170151072877, + "loss": 2.5312, + "step": 16406 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.00014590574224338277, + "loss": 2.4392, + "step": 16407 + }, + { + "epoch": 0.7, + "grad_norm": 0.478515625, + "learning_rate": 0.0001458997827694777, + "loss": 2.5786, + "step": 16408 + }, + { + "epoch": 0.7, + "grad_norm": 0.4921875, + "learning_rate": 0.00014589382308904036, + "loss": 2.6871, + "step": 16409 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.0001458878632020976, + "loss": 2.526, + "step": 16410 + }, + { + "epoch": 0.7, + "grad_norm": 0.435546875, + "learning_rate": 0.00014588190310867622, + "loss": 2.5065, + "step": 16411 + }, + { + "epoch": 0.7, + "grad_norm": 0.54296875, + "learning_rate": 0.00014587594280880307, + "loss": 2.6371, + "step": 16412 + }, + { + "epoch": 0.7, + "grad_norm": 0.494140625, + "learning_rate": 0.00014586998230250493, + "loss": 2.6485, + "step": 16413 + }, + { + "epoch": 0.7, + "grad_norm": 0.515625, + "learning_rate": 0.0001458640215898086, + "loss": 2.5957, + "step": 16414 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.00014585806067074099, + "loss": 2.6962, + "step": 16415 + }, + { + "epoch": 0.7, + "grad_norm": 0.490234375, + "learning_rate": 0.00014585209954532882, + "loss": 2.4923, + "step": 16416 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.00014584613821359897, + "loss": 2.5041, + "step": 16417 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.00014584017667557827, + "loss": 2.6025, + "step": 16418 + }, + { + "epoch": 0.7, + "grad_norm": 0.443359375, + "learning_rate": 0.00014583421493129353, + "loss": 2.6371, + "step": 16419 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014582825298077156, + "loss": 2.5324, + "step": 16420 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.0001458222908240392, + "loss": 2.7403, + "step": 16421 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.00014581632846112326, + "loss": 2.5453, + "step": 16422 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.00014581036589205061, + "loss": 2.6965, + "step": 16423 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014580440311684804, + "loss": 2.6738, + "step": 16424 + }, + { + "epoch": 0.7, + "grad_norm": 0.4609375, + "learning_rate": 0.00014579844013554242, + "loss": 2.5921, + "step": 16425 + }, + { + "epoch": 0.7, + "grad_norm": 0.4921875, + "learning_rate": 0.00014579247694816055, + "loss": 2.7053, + "step": 16426 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.00014578651355472925, + "loss": 2.5755, + "step": 16427 + }, + { + "epoch": 0.7, + "grad_norm": 0.478515625, + "learning_rate": 0.0001457805499552754, + "loss": 2.6254, + "step": 16428 + }, + { + "epoch": 0.7, + "grad_norm": 0.470703125, + "learning_rate": 0.00014577458614982577, + "loss": 2.6403, + "step": 16429 + }, + { + "epoch": 0.7, + "grad_norm": 0.419921875, + "learning_rate": 0.00014576862213840726, + "loss": 2.5037, + "step": 16430 + }, + { + "epoch": 0.7, + "grad_norm": 0.5, + "learning_rate": 0.00014576265792104665, + "loss": 2.5115, + "step": 16431 + }, + { + "epoch": 0.7, + "grad_norm": 0.42578125, + "learning_rate": 0.00014575669349777082, + "loss": 2.5725, + "step": 16432 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.0001457507288686066, + "loss": 2.4928, + "step": 16433 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.00014574476403358082, + "loss": 2.6247, + "step": 16434 + }, + { + "epoch": 0.7, + "grad_norm": 0.470703125, + "learning_rate": 0.0001457387989927203, + "loss": 2.6607, + "step": 16435 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014573283374605194, + "loss": 2.6067, + "step": 16436 + }, + { + "epoch": 0.7, + "grad_norm": 0.4296875, + "learning_rate": 0.0001457268682936025, + "loss": 2.5304, + "step": 16437 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014572090263539886, + "loss": 2.5646, + "step": 16438 + }, + { + "epoch": 0.7, + "grad_norm": 0.455078125, + "learning_rate": 0.0001457149367714679, + "loss": 2.5177, + "step": 16439 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.0001457089707018364, + "loss": 2.5719, + "step": 16440 + }, + { + "epoch": 0.7, + "grad_norm": 0.4375, + "learning_rate": 0.00014570300442653126, + "loss": 2.5944, + "step": 16441 + }, + { + "epoch": 0.7, + "grad_norm": 0.435546875, + "learning_rate": 0.00014569703794557932, + "loss": 2.5466, + "step": 16442 + }, + { + "epoch": 0.7, + "grad_norm": 0.427734375, + "learning_rate": 0.0001456910712590074, + "loss": 2.58, + "step": 16443 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.00014568510436684232, + "loss": 2.5887, + "step": 16444 + }, + { + "epoch": 0.7, + "grad_norm": 0.478515625, + "learning_rate": 0.000145679137269111, + "loss": 2.7071, + "step": 16445 + }, + { + "epoch": 0.7, + "grad_norm": 0.400390625, + "learning_rate": 0.00014567316996584023, + "loss": 2.659, + "step": 16446 + }, + { + "epoch": 0.7, + "grad_norm": 0.50390625, + "learning_rate": 0.0001456672024570569, + "loss": 2.5883, + "step": 16447 + }, + { + "epoch": 0.7, + "grad_norm": 0.431640625, + "learning_rate": 0.00014566123474278784, + "loss": 2.5375, + "step": 16448 + }, + { + "epoch": 0.7, + "grad_norm": 0.42578125, + "learning_rate": 0.00014565526682305992, + "loss": 2.6155, + "step": 16449 + }, + { + "epoch": 0.7, + "grad_norm": 0.52734375, + "learning_rate": 0.0001456492986979, + "loss": 2.5446, + "step": 16450 + }, + { + "epoch": 0.7, + "grad_norm": 0.52734375, + "learning_rate": 0.0001456433303673349, + "loss": 2.6368, + "step": 16451 + }, + { + "epoch": 0.7, + "grad_norm": 0.451171875, + "learning_rate": 0.00014563736183139148, + "loss": 2.611, + "step": 16452 + }, + { + "epoch": 0.7, + "grad_norm": 0.484375, + "learning_rate": 0.00014563139309009665, + "loss": 2.6116, + "step": 16453 + }, + { + "epoch": 0.7, + "grad_norm": 0.439453125, + "learning_rate": 0.0001456254241434772, + "loss": 2.5684, + "step": 16454 + }, + { + "epoch": 0.7, + "grad_norm": 0.423828125, + "learning_rate": 0.00014561945499156004, + "loss": 2.5344, + "step": 16455 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.00014561348563437197, + "loss": 2.5479, + "step": 16456 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.0001456075160719399, + "loss": 2.4648, + "step": 16457 + }, + { + "epoch": 0.7, + "grad_norm": 0.4375, + "learning_rate": 0.00014560154630429068, + "loss": 2.4762, + "step": 16458 + }, + { + "epoch": 0.7, + "grad_norm": 0.54296875, + "learning_rate": 0.00014559557633145117, + "loss": 2.5126, + "step": 16459 + }, + { + "epoch": 0.7, + "grad_norm": 0.423828125, + "learning_rate": 0.0001455896061534482, + "loss": 2.545, + "step": 16460 + }, + { + "epoch": 0.7, + "grad_norm": 0.490234375, + "learning_rate": 0.0001455836357703087, + "loss": 2.5324, + "step": 16461 + }, + { + "epoch": 0.7, + "grad_norm": 0.5234375, + "learning_rate": 0.00014557766518205948, + "loss": 2.6603, + "step": 16462 + }, + { + "epoch": 0.7, + "grad_norm": 0.47265625, + "learning_rate": 0.00014557169438872745, + "loss": 2.6219, + "step": 16463 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.00014556572339033942, + "loss": 2.4606, + "step": 16464 + }, + { + "epoch": 0.7, + "grad_norm": 0.470703125, + "learning_rate": 0.00014555975218692228, + "loss": 2.6069, + "step": 16465 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014555378077850292, + "loss": 2.6428, + "step": 16466 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.00014554780916510823, + "loss": 2.5308, + "step": 16467 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.00014554183734676499, + "loss": 2.5193, + "step": 16468 + }, + { + "epoch": 0.7, + "grad_norm": 0.435546875, + "learning_rate": 0.00014553586532350013, + "loss": 2.5494, + "step": 16469 + }, + { + "epoch": 0.7, + "grad_norm": 0.462890625, + "learning_rate": 0.00014552989309534053, + "loss": 2.6747, + "step": 16470 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.0001455239206623131, + "loss": 2.4155, + "step": 16471 + }, + { + "epoch": 0.7, + "grad_norm": 0.5078125, + "learning_rate": 0.00014551794802444457, + "loss": 2.5194, + "step": 16472 + }, + { + "epoch": 0.7, + "grad_norm": 0.515625, + "learning_rate": 0.00014551197518176197, + "loss": 2.5129, + "step": 16473 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.0001455060021342921, + "loss": 2.6619, + "step": 16474 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.00014550002888206185, + "loss": 2.6346, + "step": 16475 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.00014549405542509808, + "loss": 2.6969, + "step": 16476 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.00014548808176342772, + "loss": 2.5605, + "step": 16477 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.00014548210789707759, + "loss": 2.5221, + "step": 16478 + }, + { + "epoch": 0.7, + "grad_norm": 0.435546875, + "learning_rate": 0.00014547613382607464, + "loss": 2.6203, + "step": 16479 + }, + { + "epoch": 0.7, + "grad_norm": 0.400390625, + "learning_rate": 0.00014547015955044566, + "loss": 2.6192, + "step": 16480 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.0001454641850702176, + "loss": 2.6256, + "step": 16481 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.00014545821038541731, + "loss": 2.4575, + "step": 16482 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.0001454522354960717, + "loss": 2.5551, + "step": 16483 + }, + { + "epoch": 0.7, + "grad_norm": 0.46484375, + "learning_rate": 0.00014544626040220764, + "loss": 2.5556, + "step": 16484 + }, + { + "epoch": 0.7, + "grad_norm": 0.515625, + "learning_rate": 0.000145440285103852, + "loss": 2.4953, + "step": 16485 + }, + { + "epoch": 0.7, + "grad_norm": 0.478515625, + "learning_rate": 0.0001454343096010317, + "loss": 2.6742, + "step": 16486 + }, + { + "epoch": 0.7, + "grad_norm": 0.44140625, + "learning_rate": 0.00014542833389377362, + "loss": 2.5453, + "step": 16487 + }, + { + "epoch": 0.7, + "grad_norm": 0.474609375, + "learning_rate": 0.00014542235798210463, + "loss": 2.678, + "step": 16488 + }, + { + "epoch": 0.7, + "grad_norm": 0.470703125, + "learning_rate": 0.0001454163818660516, + "loss": 2.6999, + "step": 16489 + }, + { + "epoch": 0.7, + "grad_norm": 0.50390625, + "learning_rate": 0.00014541040554564148, + "loss": 2.5731, + "step": 16490 + }, + { + "epoch": 0.7, + "grad_norm": 0.41015625, + "learning_rate": 0.00014540442902090113, + "loss": 2.4677, + "step": 16491 + }, + { + "epoch": 0.7, + "grad_norm": 0.46484375, + "learning_rate": 0.00014539845229185746, + "loss": 2.5542, + "step": 16492 + }, + { + "epoch": 0.7, + "grad_norm": 0.45703125, + "learning_rate": 0.0001453924753585373, + "loss": 2.5612, + "step": 16493 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.0001453864982209676, + "loss": 2.6442, + "step": 16494 + }, + { + "epoch": 0.7, + "grad_norm": 0.443359375, + "learning_rate": 0.0001453805208791753, + "loss": 2.5173, + "step": 16495 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.00014537454333318718, + "loss": 2.5784, + "step": 16496 + }, + { + "epoch": 0.7, + "grad_norm": 0.38671875, + "learning_rate": 0.0001453685655830302, + "loss": 2.6008, + "step": 16497 + }, + { + "epoch": 0.7, + "grad_norm": 0.474609375, + "learning_rate": 0.0001453625876287313, + "loss": 2.6327, + "step": 16498 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.00014535660947031728, + "loss": 2.587, + "step": 16499 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.00014535063110781513, + "loss": 2.4159, + "step": 16500 + }, + { + "epoch": 0.7, + "grad_norm": 0.427734375, + "learning_rate": 0.00014534465254125172, + "loss": 2.5601, + "step": 16501 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014533867377065392, + "loss": 2.6334, + "step": 16502 + }, + { + "epoch": 0.7, + "grad_norm": 0.47265625, + "learning_rate": 0.00014533269479604868, + "loss": 2.5936, + "step": 16503 + }, + { + "epoch": 0.7, + "grad_norm": 0.466796875, + "learning_rate": 0.00014532671561746284, + "loss": 2.5208, + "step": 16504 + }, + { + "epoch": 0.7, + "grad_norm": 0.52734375, + "learning_rate": 0.00014532073623492337, + "loss": 2.695, + "step": 16505 + }, + { + "epoch": 0.7, + "grad_norm": 0.48046875, + "learning_rate": 0.00014531475664845717, + "loss": 2.5667, + "step": 16506 + }, + { + "epoch": 0.7, + "grad_norm": 0.55859375, + "learning_rate": 0.0001453087768580911, + "loss": 2.6663, + "step": 16507 + }, + { + "epoch": 0.7, + "grad_norm": 0.51171875, + "learning_rate": 0.0001453027968638521, + "loss": 2.6469, + "step": 16508 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.00014529681666576707, + "loss": 2.5285, + "step": 16509 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.0001452908362638629, + "loss": 2.5864, + "step": 16510 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014528485565816655, + "loss": 2.5907, + "step": 16511 + }, + { + "epoch": 0.7, + "grad_norm": 0.4921875, + "learning_rate": 0.0001452788748487049, + "loss": 2.514, + "step": 16512 + }, + { + "epoch": 0.7, + "grad_norm": 0.400390625, + "learning_rate": 0.00014527289383550483, + "loss": 2.6297, + "step": 16513 + }, + { + "epoch": 0.7, + "grad_norm": 0.470703125, + "learning_rate": 0.00014526691261859331, + "loss": 2.5632, + "step": 16514 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014526093119799719, + "loss": 2.4678, + "step": 16515 + }, + { + "epoch": 0.7, + "grad_norm": 0.423828125, + "learning_rate": 0.00014525494957374346, + "loss": 2.7011, + "step": 16516 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.00014524896774585898, + "loss": 2.5597, + "step": 16517 + }, + { + "epoch": 0.7, + "grad_norm": 0.48828125, + "learning_rate": 0.00014524298571437069, + "loss": 2.7338, + "step": 16518 + }, + { + "epoch": 0.7, + "grad_norm": 0.51171875, + "learning_rate": 0.0001452370034793055, + "loss": 2.563, + "step": 16519 + }, + { + "epoch": 0.7, + "grad_norm": 0.44921875, + "learning_rate": 0.00014523102104069033, + "loss": 2.5392, + "step": 16520 + }, + { + "epoch": 0.7, + "grad_norm": 0.51171875, + "learning_rate": 0.00014522503839855207, + "loss": 2.5045, + "step": 16521 + }, + { + "epoch": 0.7, + "grad_norm": 0.4296875, + "learning_rate": 0.00014521905555291771, + "loss": 2.6958, + "step": 16522 + }, + { + "epoch": 0.7, + "grad_norm": 0.51953125, + "learning_rate": 0.00014521307250381406, + "loss": 2.6027, + "step": 16523 + }, + { + "epoch": 0.7, + "grad_norm": 0.466796875, + "learning_rate": 0.00014520708925126817, + "loss": 2.5777, + "step": 16524 + }, + { + "epoch": 0.7, + "grad_norm": 0.453125, + "learning_rate": 0.00014520110579530686, + "loss": 2.6359, + "step": 16525 + }, + { + "epoch": 0.7, + "grad_norm": 0.423828125, + "learning_rate": 0.0001451951221359571, + "loss": 2.5034, + "step": 16526 + }, + { + "epoch": 0.7, + "grad_norm": 0.458984375, + "learning_rate": 0.00014518913827324582, + "loss": 2.5836, + "step": 16527 + }, + { + "epoch": 0.7, + "grad_norm": 0.431640625, + "learning_rate": 0.00014518315420719992, + "loss": 2.395, + "step": 16528 + }, + { + "epoch": 0.7, + "grad_norm": 0.423828125, + "learning_rate": 0.00014517716993784634, + "loss": 2.5653, + "step": 16529 + }, + { + "epoch": 0.7, + "grad_norm": 0.5078125, + "learning_rate": 0.00014517118546521202, + "loss": 2.5373, + "step": 16530 + }, + { + "epoch": 0.7, + "grad_norm": 0.46484375, + "learning_rate": 0.00014516520078932386, + "loss": 2.4328, + "step": 16531 + }, + { + "epoch": 0.7, + "grad_norm": 0.494140625, + "learning_rate": 0.0001451592159102088, + "loss": 2.6411, + "step": 16532 + }, + { + "epoch": 0.7, + "grad_norm": 0.43359375, + "learning_rate": 0.00014515323082789377, + "loss": 2.5793, + "step": 16533 + }, + { + "epoch": 0.7, + "grad_norm": 0.462890625, + "learning_rate": 0.00014514724554240573, + "loss": 2.7192, + "step": 16534 + }, + { + "epoch": 0.7, + "grad_norm": 0.47265625, + "learning_rate": 0.00014514126005377157, + "loss": 2.6618, + "step": 16535 + }, + { + "epoch": 0.7, + "grad_norm": 0.490234375, + "learning_rate": 0.00014513527436201822, + "loss": 2.6298, + "step": 16536 + }, + { + "epoch": 0.7, + "grad_norm": 0.474609375, + "learning_rate": 0.00014512928846717265, + "loss": 2.5861, + "step": 16537 + }, + { + "epoch": 0.7, + "grad_norm": 0.4296875, + "learning_rate": 0.0001451233023692618, + "loss": 2.5818, + "step": 16538 + }, + { + "epoch": 0.7, + "grad_norm": 0.48046875, + "learning_rate": 0.00014511731606831253, + "loss": 2.5697, + "step": 16539 + }, + { + "epoch": 0.7, + "grad_norm": 0.404296875, + "learning_rate": 0.00014511132956435182, + "loss": 2.4358, + "step": 16540 + }, + { + "epoch": 0.7, + "grad_norm": 0.38671875, + "learning_rate": 0.00014510534285740664, + "loss": 2.4445, + "step": 16541 + }, + { + "epoch": 0.7, + "grad_norm": 0.447265625, + "learning_rate": 0.00014509935594750394, + "loss": 2.5706, + "step": 16542 + }, + { + "epoch": 0.7, + "grad_norm": 0.484375, + "learning_rate": 0.00014509336883467054, + "loss": 2.5443, + "step": 16543 + }, + { + "epoch": 0.7, + "grad_norm": 0.46875, + "learning_rate": 0.00014508738151893354, + "loss": 2.5822, + "step": 16544 + }, + { + "epoch": 0.71, + "grad_norm": 0.47265625, + "learning_rate": 0.00014508139400031973, + "loss": 2.5312, + "step": 16545 + }, + { + "epoch": 0.71, + "grad_norm": 0.423828125, + "learning_rate": 0.00014507540627885617, + "loss": 2.5274, + "step": 16546 + }, + { + "epoch": 0.71, + "grad_norm": 0.427734375, + "learning_rate": 0.0001450694183545697, + "loss": 2.6018, + "step": 16547 + }, + { + "epoch": 0.71, + "grad_norm": 0.419921875, + "learning_rate": 0.00014506343022748736, + "loss": 2.6156, + "step": 16548 + }, + { + "epoch": 0.71, + "grad_norm": 0.46484375, + "learning_rate": 0.00014505744189763605, + "loss": 2.6527, + "step": 16549 + }, + { + "epoch": 0.71, + "grad_norm": 0.5078125, + "learning_rate": 0.00014505145336504274, + "loss": 2.7289, + "step": 16550 + }, + { + "epoch": 0.71, + "grad_norm": 0.44140625, + "learning_rate": 0.00014504546462973432, + "loss": 2.6249, + "step": 16551 + }, + { + "epoch": 0.71, + "grad_norm": 0.5390625, + "learning_rate": 0.00014503947569173778, + "loss": 2.5877, + "step": 16552 + }, + { + "epoch": 0.71, + "grad_norm": 0.416015625, + "learning_rate": 0.00014503348655108006, + "loss": 2.639, + "step": 16553 + }, + { + "epoch": 0.71, + "grad_norm": 0.478515625, + "learning_rate": 0.00014502749720778812, + "loss": 2.5464, + "step": 16554 + }, + { + "epoch": 0.71, + "grad_norm": 0.5234375, + "learning_rate": 0.00014502150766188888, + "loss": 2.8086, + "step": 16555 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.0001450155179134093, + "loss": 2.6135, + "step": 16556 + }, + { + "epoch": 0.71, + "grad_norm": 0.47265625, + "learning_rate": 0.00014500952796237638, + "loss": 2.6115, + "step": 16557 + }, + { + "epoch": 0.71, + "grad_norm": 0.46484375, + "learning_rate": 0.000145003537808817, + "loss": 2.5368, + "step": 16558 + }, + { + "epoch": 0.71, + "grad_norm": 0.482421875, + "learning_rate": 0.00014499754745275818, + "loss": 2.6269, + "step": 16559 + }, + { + "epoch": 0.71, + "grad_norm": 0.45703125, + "learning_rate": 0.0001449915568942268, + "loss": 2.5656, + "step": 16560 + }, + { + "epoch": 0.71, + "grad_norm": 0.470703125, + "learning_rate": 0.00014498556613324988, + "loss": 2.5193, + "step": 16561 + }, + { + "epoch": 0.71, + "grad_norm": 0.470703125, + "learning_rate": 0.00014497957516985437, + "loss": 2.6083, + "step": 16562 + }, + { + "epoch": 0.71, + "grad_norm": 0.51953125, + "learning_rate": 0.00014497358400406715, + "loss": 2.604, + "step": 16563 + }, + { + "epoch": 0.71, + "grad_norm": 0.474609375, + "learning_rate": 0.00014496759263591527, + "loss": 2.5313, + "step": 16564 + }, + { + "epoch": 0.71, + "grad_norm": 0.43359375, + "learning_rate": 0.00014496160106542569, + "loss": 2.5959, + "step": 16565 + }, + { + "epoch": 0.71, + "grad_norm": 0.46484375, + "learning_rate": 0.0001449556092926253, + "loss": 2.6219, + "step": 16566 + }, + { + "epoch": 0.71, + "grad_norm": 0.419921875, + "learning_rate": 0.00014494961731754114, + "loss": 2.4272, + "step": 16567 + }, + { + "epoch": 0.71, + "grad_norm": 0.443359375, + "learning_rate": 0.00014494362514020007, + "loss": 2.5539, + "step": 16568 + }, + { + "epoch": 0.71, + "grad_norm": 0.4921875, + "learning_rate": 0.00014493763276062912, + "loss": 2.6438, + "step": 16569 + }, + { + "epoch": 0.71, + "grad_norm": 0.498046875, + "learning_rate": 0.0001449316401788553, + "loss": 2.5899, + "step": 16570 + }, + { + "epoch": 0.71, + "grad_norm": 0.443359375, + "learning_rate": 0.00014492564739490546, + "loss": 2.714, + "step": 16571 + }, + { + "epoch": 0.71, + "grad_norm": 0.50390625, + "learning_rate": 0.00014491965440880666, + "loss": 2.5298, + "step": 16572 + }, + { + "epoch": 0.71, + "grad_norm": 0.439453125, + "learning_rate": 0.00014491366122058584, + "loss": 2.5851, + "step": 16573 + }, + { + "epoch": 0.71, + "grad_norm": 0.439453125, + "learning_rate": 0.00014490766783026993, + "loss": 2.5126, + "step": 16574 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.00014490167423788591, + "loss": 2.5654, + "step": 16575 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.0001448956804434608, + "loss": 2.5689, + "step": 16576 + }, + { + "epoch": 0.71, + "grad_norm": 0.5, + "learning_rate": 0.00014488968644702153, + "loss": 2.581, + "step": 16577 + }, + { + "epoch": 0.71, + "grad_norm": 0.44140625, + "learning_rate": 0.0001448836922485951, + "loss": 2.4898, + "step": 16578 + }, + { + "epoch": 0.71, + "grad_norm": 0.45703125, + "learning_rate": 0.00014487769784820838, + "loss": 2.3859, + "step": 16579 + }, + { + "epoch": 0.71, + "grad_norm": 0.45703125, + "learning_rate": 0.00014487170324588847, + "loss": 2.5015, + "step": 16580 + }, + { + "epoch": 0.71, + "grad_norm": 0.3828125, + "learning_rate": 0.0001448657084416623, + "loss": 2.4689, + "step": 16581 + }, + { + "epoch": 0.71, + "grad_norm": 0.494140625, + "learning_rate": 0.00014485971343555684, + "loss": 2.7397, + "step": 16582 + }, + { + "epoch": 0.71, + "grad_norm": 0.416015625, + "learning_rate": 0.00014485371822759908, + "loss": 2.5012, + "step": 16583 + }, + { + "epoch": 0.71, + "grad_norm": 0.408203125, + "learning_rate": 0.00014484772281781597, + "loss": 2.6401, + "step": 16584 + }, + { + "epoch": 0.71, + "grad_norm": 0.412109375, + "learning_rate": 0.00014484172720623446, + "loss": 2.5164, + "step": 16585 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.0001448357313928816, + "loss": 2.6727, + "step": 16586 + }, + { + "epoch": 0.71, + "grad_norm": 0.49609375, + "learning_rate": 0.00014482973537778432, + "loss": 2.5605, + "step": 16587 + }, + { + "epoch": 0.71, + "grad_norm": 0.482421875, + "learning_rate": 0.0001448237391609696, + "loss": 2.5439, + "step": 16588 + }, + { + "epoch": 0.71, + "grad_norm": 0.470703125, + "learning_rate": 0.00014481774274246448, + "loss": 2.649, + "step": 16589 + }, + { + "epoch": 0.71, + "grad_norm": 0.439453125, + "learning_rate": 0.00014481174612229587, + "loss": 2.6315, + "step": 16590 + }, + { + "epoch": 0.71, + "grad_norm": 0.478515625, + "learning_rate": 0.0001448057493004908, + "loss": 2.743, + "step": 16591 + }, + { + "epoch": 0.71, + "grad_norm": 0.44140625, + "learning_rate": 0.00014479975227707617, + "loss": 2.5288, + "step": 16592 + }, + { + "epoch": 0.71, + "grad_norm": 0.435546875, + "learning_rate": 0.0001447937550520791, + "loss": 2.5984, + "step": 16593 + }, + { + "epoch": 0.71, + "grad_norm": 0.48828125, + "learning_rate": 0.00014478775762552648, + "loss": 2.7158, + "step": 16594 + }, + { + "epoch": 0.71, + "grad_norm": 0.5078125, + "learning_rate": 0.0001447817599974453, + "loss": 2.5251, + "step": 16595 + }, + { + "epoch": 0.71, + "grad_norm": 0.515625, + "learning_rate": 0.00014477576216786258, + "loss": 2.5956, + "step": 16596 + }, + { + "epoch": 0.71, + "grad_norm": 0.474609375, + "learning_rate": 0.0001447697641368053, + "loss": 2.4662, + "step": 16597 + }, + { + "epoch": 0.71, + "grad_norm": 0.5, + "learning_rate": 0.0001447637659043004, + "loss": 2.7316, + "step": 16598 + }, + { + "epoch": 0.71, + "grad_norm": 0.48828125, + "learning_rate": 0.00014475776747037498, + "loss": 2.5776, + "step": 16599 + }, + { + "epoch": 0.71, + "grad_norm": 0.443359375, + "learning_rate": 0.00014475176883505593, + "loss": 2.6441, + "step": 16600 + }, + { + "epoch": 0.71, + "grad_norm": 0.53125, + "learning_rate": 0.00014474576999837026, + "loss": 2.6931, + "step": 16601 + }, + { + "epoch": 0.71, + "grad_norm": 0.443359375, + "learning_rate": 0.000144739770960345, + "loss": 2.4867, + "step": 16602 + }, + { + "epoch": 0.71, + "grad_norm": 0.419921875, + "learning_rate": 0.00014473377172100714, + "loss": 2.6096, + "step": 16603 + }, + { + "epoch": 0.71, + "grad_norm": 0.5234375, + "learning_rate": 0.0001447277722803836, + "loss": 2.6572, + "step": 16604 + }, + { + "epoch": 0.71, + "grad_norm": 0.412109375, + "learning_rate": 0.0001447217726385015, + "loss": 2.4017, + "step": 16605 + }, + { + "epoch": 0.71, + "grad_norm": 0.46875, + "learning_rate": 0.00014471577279538767, + "loss": 2.5806, + "step": 16606 + }, + { + "epoch": 0.71, + "grad_norm": 0.51171875, + "learning_rate": 0.0001447097727510693, + "loss": 2.7495, + "step": 16607 + }, + { + "epoch": 0.71, + "grad_norm": 0.47265625, + "learning_rate": 0.00014470377250557322, + "loss": 2.6522, + "step": 16608 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.00014469777205892656, + "loss": 2.5568, + "step": 16609 + }, + { + "epoch": 0.71, + "grad_norm": 0.455078125, + "learning_rate": 0.00014469177141115624, + "loss": 2.6167, + "step": 16610 + }, + { + "epoch": 0.71, + "grad_norm": 0.49609375, + "learning_rate": 0.00014468577056228927, + "loss": 2.6211, + "step": 16611 + }, + { + "epoch": 0.71, + "grad_norm": 0.4375, + "learning_rate": 0.00014467976951235264, + "loss": 2.5382, + "step": 16612 + }, + { + "epoch": 0.71, + "grad_norm": 0.453125, + "learning_rate": 0.00014467376826137344, + "loss": 2.631, + "step": 16613 + }, + { + "epoch": 0.71, + "grad_norm": 0.515625, + "learning_rate": 0.00014466776680937857, + "loss": 2.6794, + "step": 16614 + }, + { + "epoch": 0.71, + "grad_norm": 0.44921875, + "learning_rate": 0.00014466176515639506, + "loss": 2.5822, + "step": 16615 + }, + { + "epoch": 0.71, + "grad_norm": 0.439453125, + "learning_rate": 0.00014465576330244996, + "loss": 2.6579, + "step": 16616 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.00014464976124757025, + "loss": 2.5866, + "step": 16617 + }, + { + "epoch": 0.71, + "grad_norm": 0.458984375, + "learning_rate": 0.00014464375899178293, + "loss": 2.458, + "step": 16618 + }, + { + "epoch": 0.71, + "grad_norm": 0.453125, + "learning_rate": 0.00014463775653511497, + "loss": 2.6136, + "step": 16619 + }, + { + "epoch": 0.71, + "grad_norm": 0.482421875, + "learning_rate": 0.00014463175387759345, + "loss": 2.6001, + "step": 16620 + }, + { + "epoch": 0.71, + "grad_norm": 0.439453125, + "learning_rate": 0.00014462575101924536, + "loss": 2.5944, + "step": 16621 + }, + { + "epoch": 0.71, + "grad_norm": 0.5, + "learning_rate": 0.00014461974796009766, + "loss": 2.503, + "step": 16622 + }, + { + "epoch": 0.71, + "grad_norm": 0.4765625, + "learning_rate": 0.00014461374470017743, + "loss": 2.5585, + "step": 16623 + }, + { + "epoch": 0.71, + "grad_norm": 0.439453125, + "learning_rate": 0.00014460774123951163, + "loss": 2.5547, + "step": 16624 + }, + { + "epoch": 0.71, + "grad_norm": 0.5078125, + "learning_rate": 0.00014460173757812735, + "loss": 2.6738, + "step": 16625 + }, + { + "epoch": 0.71, + "grad_norm": 0.484375, + "learning_rate": 0.0001445957337160515, + "loss": 2.6641, + "step": 16626 + }, + { + "epoch": 0.71, + "grad_norm": 0.4140625, + "learning_rate": 0.00014458972965331114, + "loss": 2.694, + "step": 16627 + }, + { + "epoch": 0.71, + "grad_norm": 0.4609375, + "learning_rate": 0.0001445837253899333, + "loss": 2.5617, + "step": 16628 + }, + { + "epoch": 0.71, + "grad_norm": 0.447265625, + "learning_rate": 0.00014457772092594503, + "loss": 2.4356, + "step": 16629 + }, + { + "epoch": 0.71, + "grad_norm": 0.51953125, + "learning_rate": 0.00014457171626137326, + "loss": 2.6409, + "step": 16630 + }, + { + "epoch": 0.71, + "grad_norm": 0.462890625, + "learning_rate": 0.0001445657113962451, + "loss": 2.6921, + "step": 16631 + }, + { + "epoch": 0.71, + "grad_norm": 0.4140625, + "learning_rate": 0.0001445597063305875, + "loss": 2.5059, + "step": 16632 + }, + { + "epoch": 0.71, + "grad_norm": 0.490234375, + "learning_rate": 0.0001445537010644275, + "loss": 2.5108, + "step": 16633 + }, + { + "epoch": 0.71, + "grad_norm": 0.4609375, + "learning_rate": 0.00014454769559779215, + "loss": 2.6392, + "step": 16634 + }, + { + "epoch": 0.71, + "grad_norm": 0.462890625, + "learning_rate": 0.00014454168993070841, + "loss": 2.7333, + "step": 16635 + }, + { + "epoch": 0.71, + "grad_norm": 0.54296875, + "learning_rate": 0.00014453568406320338, + "loss": 2.6534, + "step": 16636 + }, + { + "epoch": 0.71, + "grad_norm": 0.4453125, + "learning_rate": 0.00014452967799530404, + "loss": 2.6065, + "step": 16637 + }, + { + "epoch": 0.71, + "grad_norm": 0.46875, + "learning_rate": 0.0001445236717270374, + "loss": 2.5818, + "step": 16638 + }, + { + "epoch": 0.71, + "grad_norm": 0.431640625, + "learning_rate": 0.00014451766525843052, + "loss": 2.5368, + "step": 16639 + }, + { + "epoch": 0.71, + "grad_norm": 0.458984375, + "learning_rate": 0.00014451165858951043, + "loss": 2.586, + "step": 16640 + }, + { + "epoch": 0.71, + "grad_norm": 0.494140625, + "learning_rate": 0.00014450565172030415, + "loss": 2.5902, + "step": 16641 + }, + { + "epoch": 0.71, + "grad_norm": 0.46484375, + "learning_rate": 0.0001444996446508387, + "loss": 2.6086, + "step": 16642 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.0001444936373811411, + "loss": 2.6466, + "step": 16643 + }, + { + "epoch": 0.71, + "grad_norm": 0.51171875, + "learning_rate": 0.00014448762991123838, + "loss": 2.5269, + "step": 16644 + }, + { + "epoch": 0.71, + "grad_norm": 0.484375, + "learning_rate": 0.0001444816222411576, + "loss": 2.59, + "step": 16645 + }, + { + "epoch": 0.71, + "grad_norm": 0.490234375, + "learning_rate": 0.00014447561437092577, + "loss": 2.6197, + "step": 16646 + }, + { + "epoch": 0.71, + "grad_norm": 0.474609375, + "learning_rate": 0.0001444696063005699, + "loss": 2.5585, + "step": 16647 + }, + { + "epoch": 0.71, + "grad_norm": 0.50390625, + "learning_rate": 0.0001444635980301171, + "loss": 2.6321, + "step": 16648 + }, + { + "epoch": 0.71, + "grad_norm": 0.431640625, + "learning_rate": 0.00014445758955959436, + "loss": 2.6081, + "step": 16649 + }, + { + "epoch": 0.71, + "grad_norm": 0.42578125, + "learning_rate": 0.0001444515808890287, + "loss": 2.4819, + "step": 16650 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.00014444557201844717, + "loss": 2.5039, + "step": 16651 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.0001444395629478768, + "loss": 2.603, + "step": 16652 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.00014443355367734466, + "loss": 2.6394, + "step": 16653 + }, + { + "epoch": 0.71, + "grad_norm": 0.490234375, + "learning_rate": 0.00014442754420687773, + "loss": 2.6148, + "step": 16654 + }, + { + "epoch": 0.71, + "grad_norm": 0.45703125, + "learning_rate": 0.0001444215345365031, + "loss": 2.4723, + "step": 16655 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.00014441552466624784, + "loss": 2.4439, + "step": 16656 + }, + { + "epoch": 0.71, + "grad_norm": 0.4609375, + "learning_rate": 0.00014440951459613891, + "loss": 2.5969, + "step": 16657 + }, + { + "epoch": 0.71, + "grad_norm": 0.50390625, + "learning_rate": 0.0001444035043262034, + "loss": 2.7079, + "step": 16658 + }, + { + "epoch": 0.71, + "grad_norm": 0.490234375, + "learning_rate": 0.00014439749385646838, + "loss": 2.6907, + "step": 16659 + }, + { + "epoch": 0.71, + "grad_norm": 0.43359375, + "learning_rate": 0.00014439148318696077, + "loss": 2.6537, + "step": 16660 + }, + { + "epoch": 0.71, + "grad_norm": 0.42578125, + "learning_rate": 0.00014438547231770783, + "loss": 2.5019, + "step": 16661 + }, + { + "epoch": 0.71, + "grad_norm": 0.4765625, + "learning_rate": 0.0001443794612487364, + "loss": 2.6119, + "step": 16662 + }, + { + "epoch": 0.71, + "grad_norm": 0.466796875, + "learning_rate": 0.00014437344998007365, + "loss": 2.5437, + "step": 16663 + }, + { + "epoch": 0.71, + "grad_norm": 0.41796875, + "learning_rate": 0.0001443674385117466, + "loss": 2.4837, + "step": 16664 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.00014436142684378227, + "loss": 2.5649, + "step": 16665 + }, + { + "epoch": 0.71, + "grad_norm": 0.416015625, + "learning_rate": 0.00014435541497620773, + "loss": 2.5885, + "step": 16666 + }, + { + "epoch": 0.71, + "grad_norm": 0.4375, + "learning_rate": 0.00014434940290905006, + "loss": 2.5491, + "step": 16667 + }, + { + "epoch": 0.71, + "grad_norm": 0.42578125, + "learning_rate": 0.00014434339064233625, + "loss": 2.5462, + "step": 16668 + }, + { + "epoch": 0.71, + "grad_norm": 0.466796875, + "learning_rate": 0.0001443373781760934, + "loss": 2.4949, + "step": 16669 + }, + { + "epoch": 0.71, + "grad_norm": 0.423828125, + "learning_rate": 0.00014433136551034853, + "loss": 2.5738, + "step": 16670 + }, + { + "epoch": 0.71, + "grad_norm": 0.4609375, + "learning_rate": 0.00014432535264512875, + "loss": 2.7433, + "step": 16671 + }, + { + "epoch": 0.71, + "grad_norm": 0.4453125, + "learning_rate": 0.0001443193395804611, + "loss": 2.5473, + "step": 16672 + }, + { + "epoch": 0.71, + "grad_norm": 0.455078125, + "learning_rate": 0.00014431332631637258, + "loss": 2.5912, + "step": 16673 + }, + { + "epoch": 0.71, + "grad_norm": 0.515625, + "learning_rate": 0.0001443073128528903, + "loss": 2.4839, + "step": 16674 + }, + { + "epoch": 0.71, + "grad_norm": 0.388671875, + "learning_rate": 0.00014430129919004129, + "loss": 2.5664, + "step": 16675 + }, + { + "epoch": 0.71, + "grad_norm": 0.51171875, + "learning_rate": 0.00014429528532785264, + "loss": 2.6967, + "step": 16676 + }, + { + "epoch": 0.71, + "grad_norm": 0.458984375, + "learning_rate": 0.00014428927126635142, + "loss": 2.5298, + "step": 16677 + }, + { + "epoch": 0.71, + "grad_norm": 0.443359375, + "learning_rate": 0.0001442832570055646, + "loss": 2.7581, + "step": 16678 + }, + { + "epoch": 0.71, + "grad_norm": 0.50390625, + "learning_rate": 0.00014427724254551934, + "loss": 2.6247, + "step": 16679 + }, + { + "epoch": 0.71, + "grad_norm": 0.490234375, + "learning_rate": 0.0001442712278862427, + "loss": 2.6121, + "step": 16680 + }, + { + "epoch": 0.71, + "grad_norm": 0.41796875, + "learning_rate": 0.0001442652130277617, + "loss": 2.5775, + "step": 16681 + }, + { + "epoch": 0.71, + "grad_norm": 0.466796875, + "learning_rate": 0.00014425919797010338, + "loss": 2.6568, + "step": 16682 + }, + { + "epoch": 0.71, + "grad_norm": 0.41796875, + "learning_rate": 0.00014425318271329485, + "loss": 2.583, + "step": 16683 + }, + { + "epoch": 0.71, + "grad_norm": 0.44140625, + "learning_rate": 0.00014424716725736317, + "loss": 2.5402, + "step": 16684 + }, + { + "epoch": 0.71, + "grad_norm": 0.455078125, + "learning_rate": 0.00014424115160233548, + "loss": 2.5794, + "step": 16685 + }, + { + "epoch": 0.71, + "grad_norm": 0.46875, + "learning_rate": 0.0001442351357482387, + "loss": 2.6693, + "step": 16686 + }, + { + "epoch": 0.71, + "grad_norm": 0.470703125, + "learning_rate": 0.0001442291196951, + "loss": 2.7029, + "step": 16687 + }, + { + "epoch": 0.71, + "grad_norm": 0.453125, + "learning_rate": 0.00014422310344294642, + "loss": 2.7527, + "step": 16688 + }, + { + "epoch": 0.71, + "grad_norm": 0.458984375, + "learning_rate": 0.00014421708699180504, + "loss": 2.5536, + "step": 16689 + }, + { + "epoch": 0.71, + "grad_norm": 0.48046875, + "learning_rate": 0.00014421107034170294, + "loss": 2.6318, + "step": 16690 + }, + { + "epoch": 0.71, + "grad_norm": 0.478515625, + "learning_rate": 0.00014420505349266717, + "loss": 2.6166, + "step": 16691 + }, + { + "epoch": 0.71, + "grad_norm": 0.455078125, + "learning_rate": 0.0001441990364447248, + "loss": 2.5696, + "step": 16692 + }, + { + "epoch": 0.71, + "grad_norm": 0.54296875, + "learning_rate": 0.00014419301919790296, + "loss": 2.6031, + "step": 16693 + }, + { + "epoch": 0.71, + "grad_norm": 0.43359375, + "learning_rate": 0.00014418700175222865, + "loss": 2.5143, + "step": 16694 + }, + { + "epoch": 0.71, + "grad_norm": 0.408203125, + "learning_rate": 0.00014418098410772897, + "loss": 2.7289, + "step": 16695 + }, + { + "epoch": 0.71, + "grad_norm": 0.470703125, + "learning_rate": 0.00014417496626443102, + "loss": 2.5642, + "step": 16696 + }, + { + "epoch": 0.71, + "grad_norm": 0.490234375, + "learning_rate": 0.0001441689482223619, + "loss": 2.6158, + "step": 16697 + }, + { + "epoch": 0.71, + "grad_norm": 0.439453125, + "learning_rate": 0.00014416292998154862, + "loss": 2.5889, + "step": 16698 + }, + { + "epoch": 0.71, + "grad_norm": 0.4375, + "learning_rate": 0.0001441569115420183, + "loss": 2.5054, + "step": 16699 + }, + { + "epoch": 0.71, + "grad_norm": 0.408203125, + "learning_rate": 0.000144150892903798, + "loss": 2.5497, + "step": 16700 + }, + { + "epoch": 0.71, + "grad_norm": 0.53125, + "learning_rate": 0.00014414487406691487, + "loss": 2.5243, + "step": 16701 + }, + { + "epoch": 0.71, + "grad_norm": 0.388671875, + "learning_rate": 0.00014413885503139585, + "loss": 2.6018, + "step": 16702 + }, + { + "epoch": 0.71, + "grad_norm": 0.52734375, + "learning_rate": 0.00014413283579726817, + "loss": 2.5646, + "step": 16703 + }, + { + "epoch": 0.71, + "grad_norm": 0.435546875, + "learning_rate": 0.00014412681636455886, + "loss": 2.515, + "step": 16704 + }, + { + "epoch": 0.71, + "grad_norm": 0.4140625, + "learning_rate": 0.000144120796733295, + "loss": 2.5167, + "step": 16705 + }, + { + "epoch": 0.71, + "grad_norm": 0.5078125, + "learning_rate": 0.00014411477690350367, + "loss": 2.6507, + "step": 16706 + }, + { + "epoch": 0.71, + "grad_norm": 0.44140625, + "learning_rate": 0.000144108756875212, + "loss": 2.5886, + "step": 16707 + }, + { + "epoch": 0.71, + "grad_norm": 0.48046875, + "learning_rate": 0.000144102736648447, + "loss": 2.5475, + "step": 16708 + }, + { + "epoch": 0.71, + "grad_norm": 0.470703125, + "learning_rate": 0.00014409671622323583, + "loss": 2.4203, + "step": 16709 + }, + { + "epoch": 0.71, + "grad_norm": 0.453125, + "learning_rate": 0.00014409069559960554, + "loss": 2.5776, + "step": 16710 + }, + { + "epoch": 0.71, + "grad_norm": 0.435546875, + "learning_rate": 0.00014408467477758323, + "loss": 2.5855, + "step": 16711 + }, + { + "epoch": 0.71, + "grad_norm": 0.470703125, + "learning_rate": 0.00014407865375719601, + "loss": 2.689, + "step": 16712 + }, + { + "epoch": 0.71, + "grad_norm": 0.427734375, + "learning_rate": 0.00014407263253847093, + "loss": 2.5023, + "step": 16713 + }, + { + "epoch": 0.71, + "grad_norm": 0.447265625, + "learning_rate": 0.00014406661112143513, + "loss": 2.5505, + "step": 16714 + }, + { + "epoch": 0.71, + "grad_norm": 0.47265625, + "learning_rate": 0.0001440605895061157, + "loss": 2.6317, + "step": 16715 + }, + { + "epoch": 0.71, + "grad_norm": 0.3984375, + "learning_rate": 0.00014405456769253966, + "loss": 2.595, + "step": 16716 + }, + { + "epoch": 0.71, + "grad_norm": 0.4453125, + "learning_rate": 0.00014404854568073424, + "loss": 2.7008, + "step": 16717 + }, + { + "epoch": 0.71, + "grad_norm": 0.490234375, + "learning_rate": 0.0001440425234707264, + "loss": 2.5688, + "step": 16718 + }, + { + "epoch": 0.71, + "grad_norm": 0.439453125, + "learning_rate": 0.00014403650106254335, + "loss": 2.699, + "step": 16719 + }, + { + "epoch": 0.71, + "grad_norm": 0.462890625, + "learning_rate": 0.00014403047845621208, + "loss": 2.59, + "step": 16720 + }, + { + "epoch": 0.71, + "grad_norm": 0.482421875, + "learning_rate": 0.00014402445565175981, + "loss": 2.6193, + "step": 16721 + }, + { + "epoch": 0.71, + "grad_norm": 0.453125, + "learning_rate": 0.00014401843264921356, + "loss": 2.6554, + "step": 16722 + }, + { + "epoch": 0.71, + "grad_norm": 0.443359375, + "learning_rate": 0.00014401240944860043, + "loss": 2.4666, + "step": 16723 + }, + { + "epoch": 0.71, + "grad_norm": 0.482421875, + "learning_rate": 0.00014400638604994756, + "loss": 2.5683, + "step": 16724 + }, + { + "epoch": 0.71, + "grad_norm": 0.470703125, + "learning_rate": 0.00014400036245328205, + "loss": 2.5608, + "step": 16725 + }, + { + "epoch": 0.71, + "grad_norm": 0.478515625, + "learning_rate": 0.00014399433865863097, + "loss": 2.5979, + "step": 16726 + }, + { + "epoch": 0.71, + "grad_norm": 0.4609375, + "learning_rate": 0.00014398831466602143, + "loss": 2.5556, + "step": 16727 + }, + { + "epoch": 0.71, + "grad_norm": 0.462890625, + "learning_rate": 0.00014398229047548057, + "loss": 2.6712, + "step": 16728 + }, + { + "epoch": 0.71, + "grad_norm": 0.43359375, + "learning_rate": 0.00014397626608703547, + "loss": 2.6102, + "step": 16729 + }, + { + "epoch": 0.71, + "grad_norm": 0.421875, + "learning_rate": 0.00014397024150071325, + "loss": 2.5016, + "step": 16730 + }, + { + "epoch": 0.71, + "grad_norm": 0.458984375, + "learning_rate": 0.000143964216716541, + "loss": 2.6498, + "step": 16731 + }, + { + "epoch": 0.71, + "grad_norm": 0.51171875, + "learning_rate": 0.00014395819173454585, + "loss": 2.5635, + "step": 16732 + }, + { + "epoch": 0.71, + "grad_norm": 0.4296875, + "learning_rate": 0.0001439521665547549, + "loss": 2.6892, + "step": 16733 + }, + { + "epoch": 0.71, + "grad_norm": 0.416015625, + "learning_rate": 0.00014394614117719528, + "loss": 2.4831, + "step": 16734 + }, + { + "epoch": 0.71, + "grad_norm": 0.443359375, + "learning_rate": 0.00014394011560189403, + "loss": 2.5037, + "step": 16735 + }, + { + "epoch": 0.71, + "grad_norm": 0.48046875, + "learning_rate": 0.0001439340898288784, + "loss": 2.5952, + "step": 16736 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.00014392806385817537, + "loss": 2.5226, + "step": 16737 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.00014392203768981212, + "loss": 2.5386, + "step": 16738 + }, + { + "epoch": 0.71, + "grad_norm": 0.45703125, + "learning_rate": 0.00014391601132381573, + "loss": 2.6152, + "step": 16739 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.00014390998476021338, + "loss": 2.685, + "step": 16740 + }, + { + "epoch": 0.71, + "grad_norm": 0.51953125, + "learning_rate": 0.00014390395799903212, + "loss": 2.7297, + "step": 16741 + }, + { + "epoch": 0.71, + "grad_norm": 0.4140625, + "learning_rate": 0.0001438979310402991, + "loss": 2.4904, + "step": 16742 + }, + { + "epoch": 0.71, + "grad_norm": 0.419921875, + "learning_rate": 0.00014389190388404142, + "loss": 2.4812, + "step": 16743 + }, + { + "epoch": 0.71, + "grad_norm": 0.421875, + "learning_rate": 0.00014388587653028622, + "loss": 2.6593, + "step": 16744 + }, + { + "epoch": 0.71, + "grad_norm": 0.490234375, + "learning_rate": 0.0001438798489790606, + "loss": 2.6667, + "step": 16745 + }, + { + "epoch": 0.71, + "grad_norm": 0.4921875, + "learning_rate": 0.00014387382123039172, + "loss": 2.6548, + "step": 16746 + }, + { + "epoch": 0.71, + "grad_norm": 0.419921875, + "learning_rate": 0.00014386779328430668, + "loss": 2.587, + "step": 16747 + }, + { + "epoch": 0.71, + "grad_norm": 0.423828125, + "learning_rate": 0.0001438617651408326, + "loss": 2.5353, + "step": 16748 + }, + { + "epoch": 0.71, + "grad_norm": 0.4453125, + "learning_rate": 0.0001438557367999966, + "loss": 2.5779, + "step": 16749 + }, + { + "epoch": 0.71, + "grad_norm": 0.453125, + "learning_rate": 0.00014384970826182583, + "loss": 2.5503, + "step": 16750 + }, + { + "epoch": 0.71, + "grad_norm": 0.4765625, + "learning_rate": 0.00014384367952634735, + "loss": 2.5902, + "step": 16751 + }, + { + "epoch": 0.71, + "grad_norm": 0.412109375, + "learning_rate": 0.00014383765059358837, + "loss": 2.512, + "step": 16752 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.000143831621463576, + "loss": 2.5004, + "step": 16753 + }, + { + "epoch": 0.71, + "grad_norm": 0.5, + "learning_rate": 0.00014382559213633732, + "loss": 2.4274, + "step": 16754 + }, + { + "epoch": 0.71, + "grad_norm": 0.486328125, + "learning_rate": 0.0001438195626118995, + "loss": 2.6303, + "step": 16755 + }, + { + "epoch": 0.71, + "grad_norm": 0.453125, + "learning_rate": 0.00014381353289028966, + "loss": 2.5254, + "step": 16756 + }, + { + "epoch": 0.71, + "grad_norm": 0.451171875, + "learning_rate": 0.00014380750297153495, + "loss": 2.685, + "step": 16757 + }, + { + "epoch": 0.71, + "grad_norm": 0.478515625, + "learning_rate": 0.0001438014728556625, + "loss": 2.4903, + "step": 16758 + }, + { + "epoch": 0.71, + "grad_norm": 0.443359375, + "learning_rate": 0.00014379544254269937, + "loss": 2.619, + "step": 16759 + }, + { + "epoch": 0.71, + "grad_norm": 0.474609375, + "learning_rate": 0.0001437894120326728, + "loss": 2.498, + "step": 16760 + }, + { + "epoch": 0.71, + "grad_norm": 0.427734375, + "learning_rate": 0.00014378338132560984, + "loss": 2.5075, + "step": 16761 + }, + { + "epoch": 0.71, + "grad_norm": 0.50390625, + "learning_rate": 0.0001437773504215377, + "loss": 2.6646, + "step": 16762 + }, + { + "epoch": 0.71, + "grad_norm": 0.4375, + "learning_rate": 0.0001437713193204835, + "loss": 2.545, + "step": 16763 + }, + { + "epoch": 0.71, + "grad_norm": 0.44921875, + "learning_rate": 0.00014376528802247432, + "loss": 2.7003, + "step": 16764 + }, + { + "epoch": 0.71, + "grad_norm": 0.435546875, + "learning_rate": 0.00014375925652753736, + "loss": 2.6238, + "step": 16765 + }, + { + "epoch": 0.71, + "grad_norm": 0.46875, + "learning_rate": 0.00014375322483569972, + "loss": 2.5224, + "step": 16766 + }, + { + "epoch": 0.71, + "grad_norm": 0.55078125, + "learning_rate": 0.00014374719294698855, + "loss": 2.5568, + "step": 16767 + }, + { + "epoch": 0.71, + "grad_norm": 0.42578125, + "learning_rate": 0.00014374116086143105, + "loss": 2.5421, + "step": 16768 + }, + { + "epoch": 0.71, + "grad_norm": 0.515625, + "learning_rate": 0.00014373512857905426, + "loss": 2.7273, + "step": 16769 + }, + { + "epoch": 0.71, + "grad_norm": 0.5078125, + "learning_rate": 0.00014372909609988542, + "loss": 2.8304, + "step": 16770 + }, + { + "epoch": 0.71, + "grad_norm": 0.4609375, + "learning_rate": 0.0001437230634239516, + "loss": 2.6699, + "step": 16771 + }, + { + "epoch": 0.71, + "grad_norm": 0.4453125, + "learning_rate": 0.00014371703055127995, + "loss": 2.5232, + "step": 16772 + }, + { + "epoch": 0.71, + "grad_norm": 0.466796875, + "learning_rate": 0.0001437109974818977, + "loss": 2.4762, + "step": 16773 + }, + { + "epoch": 0.71, + "grad_norm": 0.4375, + "learning_rate": 0.0001437049642158319, + "loss": 2.5808, + "step": 16774 + }, + { + "epoch": 0.71, + "grad_norm": 0.431640625, + "learning_rate": 0.00014369893075310972, + "loss": 2.5156, + "step": 16775 + }, + { + "epoch": 0.71, + "grad_norm": 0.4921875, + "learning_rate": 0.00014369289709375836, + "loss": 2.549, + "step": 16776 + }, + { + "epoch": 0.71, + "grad_norm": 0.423828125, + "learning_rate": 0.00014368686323780488, + "loss": 2.5509, + "step": 16777 + }, + { + "epoch": 0.71, + "grad_norm": 0.494140625, + "learning_rate": 0.00014368082918527652, + "loss": 2.5365, + "step": 16778 + }, + { + "epoch": 0.72, + "grad_norm": 0.5625, + "learning_rate": 0.00014367479493620038, + "loss": 2.7208, + "step": 16779 + }, + { + "epoch": 0.72, + "grad_norm": 0.4453125, + "learning_rate": 0.00014366876049060363, + "loss": 2.6124, + "step": 16780 + }, + { + "epoch": 0.72, + "grad_norm": 0.51171875, + "learning_rate": 0.00014366272584851342, + "loss": 2.5225, + "step": 16781 + }, + { + "epoch": 0.72, + "grad_norm": 0.466796875, + "learning_rate": 0.0001436566910099569, + "loss": 2.6214, + "step": 16782 + }, + { + "epoch": 0.72, + "grad_norm": 0.482421875, + "learning_rate": 0.0001436506559749612, + "loss": 2.5747, + "step": 16783 + }, + { + "epoch": 0.72, + "grad_norm": 0.53515625, + "learning_rate": 0.00014364462074355352, + "loss": 2.5908, + "step": 16784 + }, + { + "epoch": 0.72, + "grad_norm": 0.4765625, + "learning_rate": 0.000143638585315761, + "loss": 2.5411, + "step": 16785 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.00014363254969161076, + "loss": 2.5069, + "step": 16786 + }, + { + "epoch": 0.72, + "grad_norm": 0.53515625, + "learning_rate": 0.00014362651387113005, + "loss": 2.6779, + "step": 16787 + }, + { + "epoch": 0.72, + "grad_norm": 0.43359375, + "learning_rate": 0.00014362047785434595, + "loss": 2.5504, + "step": 16788 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.00014361444164128563, + "loss": 2.5949, + "step": 16789 + }, + { + "epoch": 0.72, + "grad_norm": 0.48046875, + "learning_rate": 0.00014360840523197627, + "loss": 2.5161, + "step": 16790 + }, + { + "epoch": 0.72, + "grad_norm": 0.5, + "learning_rate": 0.00014360236862644501, + "loss": 2.6414, + "step": 16791 + }, + { + "epoch": 0.72, + "grad_norm": 0.41796875, + "learning_rate": 0.00014359633182471905, + "loss": 2.5191, + "step": 16792 + }, + { + "epoch": 0.72, + "grad_norm": 0.482421875, + "learning_rate": 0.0001435902948268255, + "loss": 2.6715, + "step": 16793 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.00014358425763279158, + "loss": 2.5466, + "step": 16794 + }, + { + "epoch": 0.72, + "grad_norm": 0.4140625, + "learning_rate": 0.0001435782202426444, + "loss": 2.3087, + "step": 16795 + }, + { + "epoch": 0.72, + "grad_norm": 0.466796875, + "learning_rate": 0.0001435721826564112, + "loss": 2.4155, + "step": 16796 + }, + { + "epoch": 0.72, + "grad_norm": 0.4453125, + "learning_rate": 0.00014356614487411905, + "loss": 2.4916, + "step": 16797 + }, + { + "epoch": 0.72, + "grad_norm": 0.470703125, + "learning_rate": 0.0001435601068957952, + "loss": 2.7378, + "step": 16798 + }, + { + "epoch": 0.72, + "grad_norm": 0.515625, + "learning_rate": 0.00014355406872146678, + "loss": 2.668, + "step": 16799 + }, + { + "epoch": 0.72, + "grad_norm": 0.439453125, + "learning_rate": 0.00014354803035116097, + "loss": 2.4669, + "step": 16800 + }, + { + "epoch": 0.72, + "grad_norm": 0.484375, + "learning_rate": 0.00014354199178490492, + "loss": 2.6537, + "step": 16801 + }, + { + "epoch": 0.72, + "grad_norm": 0.42578125, + "learning_rate": 0.00014353595302272587, + "loss": 2.5771, + "step": 16802 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.00014352991406465092, + "loss": 2.5645, + "step": 16803 + }, + { + "epoch": 0.72, + "grad_norm": 0.45703125, + "learning_rate": 0.00014352387491070725, + "loss": 2.4381, + "step": 16804 + }, + { + "epoch": 0.72, + "grad_norm": 0.453125, + "learning_rate": 0.000143517835560922, + "loss": 2.5045, + "step": 16805 + }, + { + "epoch": 0.72, + "grad_norm": 0.5078125, + "learning_rate": 0.0001435117960153225, + "loss": 2.6201, + "step": 16806 + }, + { + "epoch": 0.72, + "grad_norm": 0.3984375, + "learning_rate": 0.00014350575627393574, + "loss": 2.569, + "step": 16807 + }, + { + "epoch": 0.72, + "grad_norm": 0.466796875, + "learning_rate": 0.00014349971633678904, + "loss": 2.6354, + "step": 16808 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.00014349367620390945, + "loss": 2.5231, + "step": 16809 + }, + { + "epoch": 0.72, + "grad_norm": 0.494140625, + "learning_rate": 0.00014348763587532427, + "loss": 2.6657, + "step": 16810 + }, + { + "epoch": 0.72, + "grad_norm": 0.4453125, + "learning_rate": 0.0001434815953510606, + "loss": 2.5118, + "step": 16811 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.00014347555463114562, + "loss": 2.5927, + "step": 16812 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.00014346951371560653, + "loss": 2.5795, + "step": 16813 + }, + { + "epoch": 0.72, + "grad_norm": 0.4453125, + "learning_rate": 0.00014346347260447054, + "loss": 2.4817, + "step": 16814 + }, + { + "epoch": 0.72, + "grad_norm": 0.419921875, + "learning_rate": 0.0001434574312977648, + "loss": 2.6537, + "step": 16815 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.0001434513897955165, + "loss": 2.5864, + "step": 16816 + }, + { + "epoch": 0.72, + "grad_norm": 0.51953125, + "learning_rate": 0.00014344534809775282, + "loss": 2.6505, + "step": 16817 + }, + { + "epoch": 0.72, + "grad_norm": 0.46484375, + "learning_rate": 0.00014343930620450097, + "loss": 2.5744, + "step": 16818 + }, + { + "epoch": 0.72, + "grad_norm": 0.482421875, + "learning_rate": 0.0001434332641157881, + "loss": 2.5458, + "step": 16819 + }, + { + "epoch": 0.72, + "grad_norm": 0.44921875, + "learning_rate": 0.00014342722183164142, + "loss": 2.5305, + "step": 16820 + }, + { + "epoch": 0.72, + "grad_norm": 0.50390625, + "learning_rate": 0.0001434211793520881, + "loss": 2.6891, + "step": 16821 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.00014341513667715536, + "loss": 2.5933, + "step": 16822 + }, + { + "epoch": 0.72, + "grad_norm": 0.48046875, + "learning_rate": 0.00014340909380687036, + "loss": 2.6055, + "step": 16823 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.0001434030507412603, + "loss": 2.7223, + "step": 16824 + }, + { + "epoch": 0.72, + "grad_norm": 0.435546875, + "learning_rate": 0.00014339700748035238, + "loss": 2.5936, + "step": 16825 + }, + { + "epoch": 0.72, + "grad_norm": 0.466796875, + "learning_rate": 0.00014339096402417377, + "loss": 2.5743, + "step": 16826 + }, + { + "epoch": 0.72, + "grad_norm": 0.4453125, + "learning_rate": 0.0001433849203727517, + "loss": 2.5688, + "step": 16827 + }, + { + "epoch": 0.72, + "grad_norm": 0.431640625, + "learning_rate": 0.00014337887652611335, + "loss": 2.5966, + "step": 16828 + }, + { + "epoch": 0.72, + "grad_norm": 0.384765625, + "learning_rate": 0.0001433728324842859, + "loss": 2.5414, + "step": 16829 + }, + { + "epoch": 0.72, + "grad_norm": 0.4765625, + "learning_rate": 0.00014336678824729656, + "loss": 2.574, + "step": 16830 + }, + { + "epoch": 0.72, + "grad_norm": 0.515625, + "learning_rate": 0.0001433607438151725, + "loss": 2.5383, + "step": 16831 + }, + { + "epoch": 0.72, + "grad_norm": 0.4375, + "learning_rate": 0.00014335469918794097, + "loss": 2.4958, + "step": 16832 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.0001433486543656291, + "loss": 2.5524, + "step": 16833 + }, + { + "epoch": 0.72, + "grad_norm": 0.462890625, + "learning_rate": 0.00014334260934826415, + "loss": 2.5679, + "step": 16834 + }, + { + "epoch": 0.72, + "grad_norm": 0.47265625, + "learning_rate": 0.0001433365641358733, + "loss": 2.582, + "step": 16835 + }, + { + "epoch": 0.72, + "grad_norm": 0.451171875, + "learning_rate": 0.00014333051872848375, + "loss": 2.5323, + "step": 16836 + }, + { + "epoch": 0.72, + "grad_norm": 0.482421875, + "learning_rate": 0.00014332447312612268, + "loss": 2.5465, + "step": 16837 + }, + { + "epoch": 0.72, + "grad_norm": 0.490234375, + "learning_rate": 0.00014331842732881735, + "loss": 2.5603, + "step": 16838 + }, + { + "epoch": 0.72, + "grad_norm": 0.4140625, + "learning_rate": 0.0001433123813365949, + "loss": 2.567, + "step": 16839 + }, + { + "epoch": 0.72, + "grad_norm": 0.4765625, + "learning_rate": 0.00014330633514948257, + "loss": 2.7193, + "step": 16840 + }, + { + "epoch": 0.72, + "grad_norm": 0.443359375, + "learning_rate": 0.00014330028876750756, + "loss": 2.6384, + "step": 16841 + }, + { + "epoch": 0.72, + "grad_norm": 0.4609375, + "learning_rate": 0.0001432942421906971, + "loss": 2.5447, + "step": 16842 + }, + { + "epoch": 0.72, + "grad_norm": 0.412109375, + "learning_rate": 0.00014328819541907834, + "loss": 2.507, + "step": 16843 + }, + { + "epoch": 0.72, + "grad_norm": 0.50390625, + "learning_rate": 0.00014328214845267854, + "loss": 2.6852, + "step": 16844 + }, + { + "epoch": 0.72, + "grad_norm": 0.51171875, + "learning_rate": 0.00014327610129152488, + "loss": 2.622, + "step": 16845 + }, + { + "epoch": 0.72, + "grad_norm": 0.484375, + "learning_rate": 0.00014327005393564455, + "loss": 2.6733, + "step": 16846 + }, + { + "epoch": 0.72, + "grad_norm": 0.44140625, + "learning_rate": 0.00014326400638506483, + "loss": 2.5122, + "step": 16847 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.0001432579586398129, + "loss": 2.5276, + "step": 16848 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.00014325191069991595, + "loss": 2.6374, + "step": 16849 + }, + { + "epoch": 0.72, + "grad_norm": 0.4140625, + "learning_rate": 0.00014324586256540117, + "loss": 2.404, + "step": 16850 + }, + { + "epoch": 0.72, + "grad_norm": 0.40625, + "learning_rate": 0.00014323981423629588, + "loss": 2.6232, + "step": 16851 + }, + { + "epoch": 0.72, + "grad_norm": 0.47265625, + "learning_rate": 0.00014323376571262716, + "loss": 2.6794, + "step": 16852 + }, + { + "epoch": 0.72, + "grad_norm": 0.447265625, + "learning_rate": 0.00014322771699442233, + "loss": 2.6073, + "step": 16853 + }, + { + "epoch": 0.72, + "grad_norm": 0.4140625, + "learning_rate": 0.00014322166808170857, + "loss": 2.4318, + "step": 16854 + }, + { + "epoch": 0.72, + "grad_norm": 0.5234375, + "learning_rate": 0.00014321561897451311, + "loss": 2.7411, + "step": 16855 + }, + { + "epoch": 0.72, + "grad_norm": 0.419921875, + "learning_rate": 0.00014320956967286315, + "loss": 2.5969, + "step": 16856 + }, + { + "epoch": 0.72, + "grad_norm": 0.56640625, + "learning_rate": 0.00014320352017678593, + "loss": 2.6802, + "step": 16857 + }, + { + "epoch": 0.72, + "grad_norm": 0.421875, + "learning_rate": 0.00014319747048630865, + "loss": 2.4695, + "step": 16858 + }, + { + "epoch": 0.72, + "grad_norm": 0.490234375, + "learning_rate": 0.00014319142060145853, + "loss": 2.5967, + "step": 16859 + }, + { + "epoch": 0.72, + "grad_norm": 0.462890625, + "learning_rate": 0.00014318537052226282, + "loss": 2.6511, + "step": 16860 + }, + { + "epoch": 0.72, + "grad_norm": 0.51953125, + "learning_rate": 0.00014317932024874872, + "loss": 2.3584, + "step": 16861 + }, + { + "epoch": 0.72, + "grad_norm": 0.5, + "learning_rate": 0.00014317326978094343, + "loss": 2.5728, + "step": 16862 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.00014316721911887423, + "loss": 2.5618, + "step": 16863 + }, + { + "epoch": 0.72, + "grad_norm": 0.494140625, + "learning_rate": 0.0001431611682625683, + "loss": 2.634, + "step": 16864 + }, + { + "epoch": 0.72, + "grad_norm": 0.49609375, + "learning_rate": 0.00014315511721205292, + "loss": 2.6097, + "step": 16865 + }, + { + "epoch": 0.72, + "grad_norm": 0.48046875, + "learning_rate": 0.00014314906596735524, + "loss": 2.5468, + "step": 16866 + }, + { + "epoch": 0.72, + "grad_norm": 0.44140625, + "learning_rate": 0.00014314301452850257, + "loss": 2.4915, + "step": 16867 + }, + { + "epoch": 0.72, + "grad_norm": 0.470703125, + "learning_rate": 0.00014313696289552208, + "loss": 2.4321, + "step": 16868 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.00014313091106844106, + "loss": 2.5887, + "step": 16869 + }, + { + "epoch": 0.72, + "grad_norm": 0.453125, + "learning_rate": 0.00014312485904728668, + "loss": 2.5999, + "step": 16870 + }, + { + "epoch": 0.72, + "grad_norm": 0.486328125, + "learning_rate": 0.0001431188068320862, + "loss": 2.6433, + "step": 16871 + }, + { + "epoch": 0.72, + "grad_norm": 0.43359375, + "learning_rate": 0.0001431127544228668, + "loss": 2.5618, + "step": 16872 + }, + { + "epoch": 0.72, + "grad_norm": 0.435546875, + "learning_rate": 0.00014310670181965583, + "loss": 2.3979, + "step": 16873 + }, + { + "epoch": 0.72, + "grad_norm": 0.44921875, + "learning_rate": 0.00014310064902248038, + "loss": 2.5861, + "step": 16874 + }, + { + "epoch": 0.72, + "grad_norm": 0.43359375, + "learning_rate": 0.0001430945960313678, + "loss": 2.6508, + "step": 16875 + }, + { + "epoch": 0.72, + "grad_norm": 0.484375, + "learning_rate": 0.00014308854284634533, + "loss": 2.5378, + "step": 16876 + }, + { + "epoch": 0.72, + "grad_norm": 0.419921875, + "learning_rate": 0.00014308248946744013, + "loss": 2.6114, + "step": 16877 + }, + { + "epoch": 0.72, + "grad_norm": 0.51171875, + "learning_rate": 0.00014307643589467944, + "loss": 2.6538, + "step": 16878 + }, + { + "epoch": 0.72, + "grad_norm": 0.48828125, + "learning_rate": 0.00014307038212809056, + "loss": 2.5602, + "step": 16879 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.0001430643281677007, + "loss": 2.5813, + "step": 16880 + }, + { + "epoch": 0.72, + "grad_norm": 0.494140625, + "learning_rate": 0.00014305827401353712, + "loss": 2.6414, + "step": 16881 + }, + { + "epoch": 0.72, + "grad_norm": 0.484375, + "learning_rate": 0.00014305221966562698, + "loss": 2.6215, + "step": 16882 + }, + { + "epoch": 0.72, + "grad_norm": 0.435546875, + "learning_rate": 0.00014304616512399764, + "loss": 2.6572, + "step": 16883 + }, + { + "epoch": 0.72, + "grad_norm": 0.44921875, + "learning_rate": 0.0001430401103886763, + "loss": 2.5509, + "step": 16884 + }, + { + "epoch": 0.72, + "grad_norm": 0.43359375, + "learning_rate": 0.00014303405545969014, + "loss": 2.6276, + "step": 16885 + }, + { + "epoch": 0.72, + "grad_norm": 0.443359375, + "learning_rate": 0.00014302800033706652, + "loss": 2.3493, + "step": 16886 + }, + { + "epoch": 0.72, + "grad_norm": 0.48828125, + "learning_rate": 0.0001430219450208326, + "loss": 2.5087, + "step": 16887 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.0001430158895110156, + "loss": 2.6399, + "step": 16888 + }, + { + "epoch": 0.72, + "grad_norm": 0.51171875, + "learning_rate": 0.00014300983380764287, + "loss": 2.6041, + "step": 16889 + }, + { + "epoch": 0.72, + "grad_norm": 0.41796875, + "learning_rate": 0.0001430037779107416, + "loss": 2.5672, + "step": 16890 + }, + { + "epoch": 0.72, + "grad_norm": 0.494140625, + "learning_rate": 0.00014299772182033903, + "loss": 2.7035, + "step": 16891 + }, + { + "epoch": 0.72, + "grad_norm": 0.45703125, + "learning_rate": 0.00014299166553646245, + "loss": 2.5016, + "step": 16892 + }, + { + "epoch": 0.72, + "grad_norm": 0.4765625, + "learning_rate": 0.00014298560905913907, + "loss": 2.5327, + "step": 16893 + }, + { + "epoch": 0.72, + "grad_norm": 0.47265625, + "learning_rate": 0.00014297955238839615, + "loss": 2.4565, + "step": 16894 + }, + { + "epoch": 0.72, + "grad_norm": 0.4140625, + "learning_rate": 0.000142973495524261, + "loss": 2.6305, + "step": 16895 + }, + { + "epoch": 0.72, + "grad_norm": 0.43359375, + "learning_rate": 0.00014296743846676076, + "loss": 2.5399, + "step": 16896 + }, + { + "epoch": 0.72, + "grad_norm": 0.44140625, + "learning_rate": 0.0001429613812159228, + "loss": 2.5114, + "step": 16897 + }, + { + "epoch": 0.72, + "grad_norm": 0.515625, + "learning_rate": 0.00014295532377177429, + "loss": 2.8972, + "step": 16898 + }, + { + "epoch": 0.72, + "grad_norm": 0.4765625, + "learning_rate": 0.00014294926613434253, + "loss": 2.7178, + "step": 16899 + }, + { + "epoch": 0.72, + "grad_norm": 0.48828125, + "learning_rate": 0.00014294320830365476, + "loss": 2.6063, + "step": 16900 + }, + { + "epoch": 0.72, + "grad_norm": 0.416015625, + "learning_rate": 0.00014293715027973827, + "loss": 2.655, + "step": 16901 + }, + { + "epoch": 0.72, + "grad_norm": 0.439453125, + "learning_rate": 0.0001429310920626203, + "loss": 2.5512, + "step": 16902 + }, + { + "epoch": 0.72, + "grad_norm": 0.45703125, + "learning_rate": 0.00014292503365232807, + "loss": 2.4458, + "step": 16903 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.0001429189750488889, + "loss": 2.3778, + "step": 16904 + }, + { + "epoch": 0.72, + "grad_norm": 0.48828125, + "learning_rate": 0.00014291291625233003, + "loss": 2.5927, + "step": 16905 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.0001429068572626787, + "loss": 2.6725, + "step": 16906 + }, + { + "epoch": 0.72, + "grad_norm": 0.466796875, + "learning_rate": 0.00014290079807996223, + "loss": 2.5228, + "step": 16907 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.0001428947387042078, + "loss": 2.6528, + "step": 16908 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.00014288867913544274, + "loss": 2.5316, + "step": 16909 + }, + { + "epoch": 0.72, + "grad_norm": 0.484375, + "learning_rate": 0.0001428826193736943, + "loss": 2.8032, + "step": 16910 + }, + { + "epoch": 0.72, + "grad_norm": 0.4453125, + "learning_rate": 0.00014287655941898976, + "loss": 2.5257, + "step": 16911 + }, + { + "epoch": 0.72, + "grad_norm": 0.40625, + "learning_rate": 0.00014287049927135637, + "loss": 2.3008, + "step": 16912 + }, + { + "epoch": 0.72, + "grad_norm": 0.451171875, + "learning_rate": 0.0001428644389308214, + "loss": 2.7149, + "step": 16913 + }, + { + "epoch": 0.72, + "grad_norm": 0.48046875, + "learning_rate": 0.00014285837839741208, + "loss": 2.6344, + "step": 16914 + }, + { + "epoch": 0.72, + "grad_norm": 0.384765625, + "learning_rate": 0.00014285231767115577, + "loss": 2.6454, + "step": 16915 + }, + { + "epoch": 0.72, + "grad_norm": 0.412109375, + "learning_rate": 0.0001428462567520797, + "loss": 2.505, + "step": 16916 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.00014284019564021107, + "loss": 2.6029, + "step": 16917 + }, + { + "epoch": 0.72, + "grad_norm": 0.466796875, + "learning_rate": 0.00014283413433557727, + "loss": 2.7041, + "step": 16918 + }, + { + "epoch": 0.72, + "grad_norm": 0.4609375, + "learning_rate": 0.0001428280728382055, + "loss": 2.5719, + "step": 16919 + }, + { + "epoch": 0.72, + "grad_norm": 0.392578125, + "learning_rate": 0.00014282201114812303, + "loss": 2.5527, + "step": 16920 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.00014281594926535723, + "loss": 2.61, + "step": 16921 + }, + { + "epoch": 0.72, + "grad_norm": 0.498046875, + "learning_rate": 0.0001428098871899352, + "loss": 2.6462, + "step": 16922 + }, + { + "epoch": 0.72, + "grad_norm": 0.48828125, + "learning_rate": 0.0001428038249218844, + "loss": 2.5983, + "step": 16923 + }, + { + "epoch": 0.72, + "grad_norm": 0.42578125, + "learning_rate": 0.00014279776246123202, + "loss": 2.6084, + "step": 16924 + }, + { + "epoch": 0.72, + "grad_norm": 0.453125, + "learning_rate": 0.00014279169980800535, + "loss": 2.5415, + "step": 16925 + }, + { + "epoch": 0.72, + "grad_norm": 0.462890625, + "learning_rate": 0.00014278563696223167, + "loss": 2.6102, + "step": 16926 + }, + { + "epoch": 0.72, + "grad_norm": 0.478515625, + "learning_rate": 0.00014277957392393824, + "loss": 2.5458, + "step": 16927 + }, + { + "epoch": 0.72, + "grad_norm": 0.416015625, + "learning_rate": 0.00014277351069315235, + "loss": 2.6136, + "step": 16928 + }, + { + "epoch": 0.72, + "grad_norm": 0.44921875, + "learning_rate": 0.00014276744726990134, + "loss": 2.5098, + "step": 16929 + }, + { + "epoch": 0.72, + "grad_norm": 0.48046875, + "learning_rate": 0.00014276138365421241, + "loss": 2.7084, + "step": 16930 + }, + { + "epoch": 0.72, + "grad_norm": 0.453125, + "learning_rate": 0.0001427553198461129, + "loss": 2.5838, + "step": 16931 + }, + { + "epoch": 0.72, + "grad_norm": 0.447265625, + "learning_rate": 0.0001427492558456301, + "loss": 2.5546, + "step": 16932 + }, + { + "epoch": 0.72, + "grad_norm": 0.453125, + "learning_rate": 0.0001427431916527912, + "loss": 2.6198, + "step": 16933 + }, + { + "epoch": 0.72, + "grad_norm": 0.51171875, + "learning_rate": 0.00014273712726762362, + "loss": 2.6129, + "step": 16934 + }, + { + "epoch": 0.72, + "grad_norm": 0.49609375, + "learning_rate": 0.00014273106269015458, + "loss": 2.6757, + "step": 16935 + }, + { + "epoch": 0.72, + "grad_norm": 0.421875, + "learning_rate": 0.00014272499792041137, + "loss": 2.4661, + "step": 16936 + }, + { + "epoch": 0.72, + "grad_norm": 0.470703125, + "learning_rate": 0.00014271893295842128, + "loss": 2.6761, + "step": 16937 + }, + { + "epoch": 0.72, + "grad_norm": 0.43359375, + "learning_rate": 0.00014271286780421157, + "loss": 2.5844, + "step": 16938 + }, + { + "epoch": 0.72, + "grad_norm": 0.455078125, + "learning_rate": 0.0001427068024578096, + "loss": 2.6373, + "step": 16939 + }, + { + "epoch": 0.72, + "grad_norm": 0.439453125, + "learning_rate": 0.00014270073691924264, + "loss": 2.5541, + "step": 16940 + }, + { + "epoch": 0.72, + "grad_norm": 0.455078125, + "learning_rate": 0.00014269467118853795, + "loss": 2.537, + "step": 16941 + }, + { + "epoch": 0.72, + "grad_norm": 0.4296875, + "learning_rate": 0.00014268860526572285, + "loss": 2.4892, + "step": 16942 + }, + { + "epoch": 0.72, + "grad_norm": 0.439453125, + "learning_rate": 0.0001426825391508246, + "loss": 2.5369, + "step": 16943 + }, + { + "epoch": 0.72, + "grad_norm": 0.41015625, + "learning_rate": 0.00014267647284387053, + "loss": 2.6568, + "step": 16944 + }, + { + "epoch": 0.72, + "grad_norm": 0.4609375, + "learning_rate": 0.00014267040634488798, + "loss": 2.6643, + "step": 16945 + }, + { + "epoch": 0.72, + "grad_norm": 0.455078125, + "learning_rate": 0.00014266433965390412, + "loss": 2.551, + "step": 16946 + }, + { + "epoch": 0.72, + "grad_norm": 0.49609375, + "learning_rate": 0.00014265827277094636, + "loss": 2.6286, + "step": 16947 + }, + { + "epoch": 0.72, + "grad_norm": 0.4375, + "learning_rate": 0.000142652205696042, + "loss": 2.6365, + "step": 16948 + }, + { + "epoch": 0.72, + "grad_norm": 0.5078125, + "learning_rate": 0.00014264613842921825, + "loss": 2.6492, + "step": 16949 + }, + { + "epoch": 0.72, + "grad_norm": 0.478515625, + "learning_rate": 0.00014264007097050248, + "loss": 2.5674, + "step": 16950 + }, + { + "epoch": 0.72, + "grad_norm": 0.4375, + "learning_rate": 0.00014263400331992196, + "loss": 2.6548, + "step": 16951 + }, + { + "epoch": 0.72, + "grad_norm": 0.421875, + "learning_rate": 0.00014262793547750402, + "loss": 2.3628, + "step": 16952 + }, + { + "epoch": 0.72, + "grad_norm": 0.4453125, + "learning_rate": 0.00014262186744327596, + "loss": 2.6463, + "step": 16953 + }, + { + "epoch": 0.72, + "grad_norm": 0.44140625, + "learning_rate": 0.00014261579921726506, + "loss": 2.581, + "step": 16954 + }, + { + "epoch": 0.72, + "grad_norm": 0.42578125, + "learning_rate": 0.0001426097307994986, + "loss": 2.4111, + "step": 16955 + }, + { + "epoch": 0.72, + "grad_norm": 0.44140625, + "learning_rate": 0.000142603662190004, + "loss": 2.4994, + "step": 16956 + }, + { + "epoch": 0.72, + "grad_norm": 0.53515625, + "learning_rate": 0.00014259759338880844, + "loss": 2.6093, + "step": 16957 + }, + { + "epoch": 0.72, + "grad_norm": 0.482421875, + "learning_rate": 0.0001425915243959393, + "loss": 2.6819, + "step": 16958 + }, + { + "epoch": 0.72, + "grad_norm": 0.42578125, + "learning_rate": 0.00014258545521142384, + "loss": 2.5774, + "step": 16959 + }, + { + "epoch": 0.72, + "grad_norm": 0.53515625, + "learning_rate": 0.00014257938583528938, + "loss": 2.6069, + "step": 16960 + }, + { + "epoch": 0.72, + "grad_norm": 0.447265625, + "learning_rate": 0.0001425733162675633, + "loss": 2.5798, + "step": 16961 + }, + { + "epoch": 0.72, + "grad_norm": 0.4921875, + "learning_rate": 0.00014256724650827282, + "loss": 2.7844, + "step": 16962 + }, + { + "epoch": 0.72, + "grad_norm": 0.44921875, + "learning_rate": 0.0001425611765574453, + "loss": 2.5591, + "step": 16963 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.00014255510641510803, + "loss": 2.6082, + "step": 16964 + }, + { + "epoch": 0.72, + "grad_norm": 0.443359375, + "learning_rate": 0.00014254903608128835, + "loss": 2.5782, + "step": 16965 + }, + { + "epoch": 0.72, + "grad_norm": 0.46875, + "learning_rate": 0.00014254296555601354, + "loss": 2.5775, + "step": 16966 + }, + { + "epoch": 0.72, + "grad_norm": 0.423828125, + "learning_rate": 0.00014253689483931093, + "loss": 2.6698, + "step": 16967 + }, + { + "epoch": 0.72, + "grad_norm": 0.412109375, + "learning_rate": 0.00014253082393120788, + "loss": 2.5391, + "step": 16968 + }, + { + "epoch": 0.72, + "grad_norm": 0.52734375, + "learning_rate": 0.00014252475283173163, + "loss": 2.6446, + "step": 16969 + }, + { + "epoch": 0.72, + "grad_norm": 0.419921875, + "learning_rate": 0.00014251868154090952, + "loss": 2.5164, + "step": 16970 + }, + { + "epoch": 0.72, + "grad_norm": 0.482421875, + "learning_rate": 0.0001425126100587689, + "loss": 2.4573, + "step": 16971 + }, + { + "epoch": 0.72, + "grad_norm": 0.515625, + "learning_rate": 0.0001425065383853371, + "loss": 2.6431, + "step": 16972 + }, + { + "epoch": 0.72, + "grad_norm": 0.5, + "learning_rate": 0.00014250046652064134, + "loss": 2.5153, + "step": 16973 + }, + { + "epoch": 0.72, + "grad_norm": 0.431640625, + "learning_rate": 0.00014249439446470906, + "loss": 2.526, + "step": 16974 + }, + { + "epoch": 0.72, + "grad_norm": 0.431640625, + "learning_rate": 0.00014248832221756754, + "loss": 2.6225, + "step": 16975 + }, + { + "epoch": 0.72, + "grad_norm": 0.462890625, + "learning_rate": 0.0001424822497792441, + "loss": 2.5807, + "step": 16976 + }, + { + "epoch": 0.72, + "grad_norm": 0.51171875, + "learning_rate": 0.00014247617714976604, + "loss": 2.5989, + "step": 16977 + }, + { + "epoch": 0.72, + "grad_norm": 0.45703125, + "learning_rate": 0.00014247010432916072, + "loss": 2.5627, + "step": 16978 + }, + { + "epoch": 0.72, + "grad_norm": 0.48046875, + "learning_rate": 0.00014246403131745543, + "loss": 2.7693, + "step": 16979 + }, + { + "epoch": 0.72, + "grad_norm": 0.5078125, + "learning_rate": 0.00014245795811467754, + "loss": 2.667, + "step": 16980 + }, + { + "epoch": 0.72, + "grad_norm": 0.41015625, + "learning_rate": 0.00014245188472085434, + "loss": 2.5182, + "step": 16981 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.0001424458111360132, + "loss": 2.4705, + "step": 16982 + }, + { + "epoch": 0.72, + "grad_norm": 0.478515625, + "learning_rate": 0.00014243973736018139, + "loss": 2.5922, + "step": 16983 + }, + { + "epoch": 0.72, + "grad_norm": 0.41015625, + "learning_rate": 0.0001424336633933863, + "loss": 2.4756, + "step": 16984 + }, + { + "epoch": 0.72, + "grad_norm": 0.43359375, + "learning_rate": 0.0001424275892356552, + "loss": 2.6158, + "step": 16985 + }, + { + "epoch": 0.72, + "grad_norm": 0.427734375, + "learning_rate": 0.0001424215148870155, + "loss": 2.4378, + "step": 16986 + }, + { + "epoch": 0.72, + "grad_norm": 0.486328125, + "learning_rate": 0.00014241544034749443, + "loss": 2.6336, + "step": 16987 + }, + { + "epoch": 0.72, + "grad_norm": 0.52734375, + "learning_rate": 0.00014240936561711945, + "loss": 2.6222, + "step": 16988 + }, + { + "epoch": 0.72, + "grad_norm": 0.52734375, + "learning_rate": 0.00014240329069591776, + "loss": 2.5195, + "step": 16989 + }, + { + "epoch": 0.72, + "grad_norm": 0.412109375, + "learning_rate": 0.00014239721558391677, + "loss": 2.6313, + "step": 16990 + }, + { + "epoch": 0.72, + "grad_norm": 0.451171875, + "learning_rate": 0.0001423911402811438, + "loss": 2.6231, + "step": 16991 + }, + { + "epoch": 0.72, + "grad_norm": 0.51171875, + "learning_rate": 0.00014238506478762625, + "loss": 2.5269, + "step": 16992 + }, + { + "epoch": 0.72, + "grad_norm": 0.470703125, + "learning_rate": 0.00014237898910339134, + "loss": 2.4508, + "step": 16993 + }, + { + "epoch": 0.72, + "grad_norm": 0.41796875, + "learning_rate": 0.00014237291322846647, + "loss": 2.5907, + "step": 16994 + }, + { + "epoch": 0.72, + "grad_norm": 0.484375, + "learning_rate": 0.00014236683716287898, + "loss": 2.483, + "step": 16995 + }, + { + "epoch": 0.72, + "grad_norm": 0.4765625, + "learning_rate": 0.00014236076090665622, + "loss": 2.7852, + "step": 16996 + }, + { + "epoch": 0.72, + "grad_norm": 0.50390625, + "learning_rate": 0.0001423546844598255, + "loss": 2.5186, + "step": 16997 + }, + { + "epoch": 0.72, + "grad_norm": 0.486328125, + "learning_rate": 0.00014234860782241419, + "loss": 2.5146, + "step": 16998 + }, + { + "epoch": 0.72, + "grad_norm": 0.451171875, + "learning_rate": 0.00014234253099444964, + "loss": 2.6723, + "step": 16999 + }, + { + "epoch": 0.72, + "grad_norm": 0.4765625, + "learning_rate": 0.00014233645397595913, + "loss": 2.4416, + "step": 17000 + }, + { + "epoch": 0.72, + "grad_norm": 0.451171875, + "learning_rate": 0.0001423303767669701, + "loss": 2.5805, + "step": 17001 + }, + { + "epoch": 0.72, + "grad_norm": 0.458984375, + "learning_rate": 0.00014232429936750983, + "loss": 2.6594, + "step": 17002 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.00014231822177760564, + "loss": 2.5517, + "step": 17003 + }, + { + "epoch": 0.72, + "grad_norm": 0.474609375, + "learning_rate": 0.00014231214399728495, + "loss": 2.4661, + "step": 17004 + }, + { + "epoch": 0.72, + "grad_norm": 0.431640625, + "learning_rate": 0.00014230606602657505, + "loss": 2.4986, + "step": 17005 + }, + { + "epoch": 0.72, + "grad_norm": 0.45703125, + "learning_rate": 0.00014229998786550335, + "loss": 2.4049, + "step": 17006 + }, + { + "epoch": 0.72, + "grad_norm": 0.466796875, + "learning_rate": 0.00014229390951409716, + "loss": 2.5679, + "step": 17007 + }, + { + "epoch": 0.72, + "grad_norm": 0.48828125, + "learning_rate": 0.00014228783097238384, + "loss": 2.5412, + "step": 17008 + }, + { + "epoch": 0.72, + "grad_norm": 0.470703125, + "learning_rate": 0.0001422817522403907, + "loss": 2.6024, + "step": 17009 + }, + { + "epoch": 0.72, + "grad_norm": 0.451171875, + "learning_rate": 0.00014227567331814517, + "loss": 2.5633, + "step": 17010 + }, + { + "epoch": 0.72, + "grad_norm": 0.4453125, + "learning_rate": 0.00014226959420567448, + "loss": 2.4957, + "step": 17011 + }, + { + "epoch": 0.72, + "grad_norm": 0.44921875, + "learning_rate": 0.00014226351490300617, + "loss": 2.6015, + "step": 17012 + }, + { + "epoch": 0.72, + "grad_norm": 0.455078125, + "learning_rate": 0.0001422574354101674, + "loss": 2.5829, + "step": 17013 + }, + { + "epoch": 0.73, + "grad_norm": 0.41796875, + "learning_rate": 0.00014225135572718567, + "loss": 2.465, + "step": 17014 + }, + { + "epoch": 0.73, + "grad_norm": 0.498046875, + "learning_rate": 0.00014224527585408826, + "loss": 2.596, + "step": 17015 + }, + { + "epoch": 0.73, + "grad_norm": 0.455078125, + "learning_rate": 0.00014223919579090254, + "loss": 2.5165, + "step": 17016 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.0001422331155376559, + "loss": 2.6751, + "step": 17017 + }, + { + "epoch": 0.73, + "grad_norm": 0.5078125, + "learning_rate": 0.00014222703509437568, + "loss": 2.6677, + "step": 17018 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.0001422209544610892, + "loss": 2.5586, + "step": 17019 + }, + { + "epoch": 0.73, + "grad_norm": 0.486328125, + "learning_rate": 0.00014221487363782389, + "loss": 2.5918, + "step": 17020 + }, + { + "epoch": 0.73, + "grad_norm": 0.423828125, + "learning_rate": 0.000142208792624607, + "loss": 2.5946, + "step": 17021 + }, + { + "epoch": 0.73, + "grad_norm": 0.423828125, + "learning_rate": 0.00014220271142146603, + "loss": 2.3512, + "step": 17022 + }, + { + "epoch": 0.73, + "grad_norm": 0.4765625, + "learning_rate": 0.00014219663002842828, + "loss": 2.6684, + "step": 17023 + }, + { + "epoch": 0.73, + "grad_norm": 0.50390625, + "learning_rate": 0.00014219054844552114, + "loss": 2.5988, + "step": 17024 + }, + { + "epoch": 0.73, + "grad_norm": 0.443359375, + "learning_rate": 0.0001421844666727719, + "loss": 2.6756, + "step": 17025 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.00014217838471020802, + "loss": 2.5084, + "step": 17026 + }, + { + "epoch": 0.73, + "grad_norm": 0.53125, + "learning_rate": 0.00014217230255785679, + "loss": 2.7076, + "step": 17027 + }, + { + "epoch": 0.73, + "grad_norm": 0.3828125, + "learning_rate": 0.00014216622021574564, + "loss": 2.6263, + "step": 17028 + }, + { + "epoch": 0.73, + "grad_norm": 0.443359375, + "learning_rate": 0.00014216013768390188, + "loss": 2.5896, + "step": 17029 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014215405496235291, + "loss": 2.4999, + "step": 17030 + }, + { + "epoch": 0.73, + "grad_norm": 0.51171875, + "learning_rate": 0.0001421479720511261, + "loss": 2.654, + "step": 17031 + }, + { + "epoch": 0.73, + "grad_norm": 0.443359375, + "learning_rate": 0.00014214188895024884, + "loss": 2.6214, + "step": 17032 + }, + { + "epoch": 0.73, + "grad_norm": 0.431640625, + "learning_rate": 0.0001421358056597485, + "loss": 2.5197, + "step": 17033 + }, + { + "epoch": 0.73, + "grad_norm": 0.46484375, + "learning_rate": 0.0001421297221796524, + "loss": 2.4986, + "step": 17034 + }, + { + "epoch": 0.73, + "grad_norm": 0.46875, + "learning_rate": 0.0001421236385099879, + "loss": 2.6319, + "step": 17035 + }, + { + "epoch": 0.73, + "grad_norm": 0.47265625, + "learning_rate": 0.0001421175546507825, + "loss": 2.6445, + "step": 17036 + }, + { + "epoch": 0.73, + "grad_norm": 0.50390625, + "learning_rate": 0.00014211147060206347, + "loss": 2.6741, + "step": 17037 + }, + { + "epoch": 0.73, + "grad_norm": 0.5234375, + "learning_rate": 0.0001421053863638582, + "loss": 2.5837, + "step": 17038 + }, + { + "epoch": 0.73, + "grad_norm": 0.439453125, + "learning_rate": 0.0001420993019361941, + "loss": 2.6808, + "step": 17039 + }, + { + "epoch": 0.73, + "grad_norm": 0.447265625, + "learning_rate": 0.00014209321731909852, + "loss": 2.6933, + "step": 17040 + }, + { + "epoch": 0.73, + "grad_norm": 0.490234375, + "learning_rate": 0.00014208713251259885, + "loss": 2.5933, + "step": 17041 + }, + { + "epoch": 0.73, + "grad_norm": 0.44921875, + "learning_rate": 0.00014208104751672245, + "loss": 2.4676, + "step": 17042 + }, + { + "epoch": 0.73, + "grad_norm": 0.5234375, + "learning_rate": 0.0001420749623314967, + "loss": 2.7193, + "step": 17043 + }, + { + "epoch": 0.73, + "grad_norm": 0.421875, + "learning_rate": 0.00014206887695694903, + "loss": 2.5202, + "step": 17044 + }, + { + "epoch": 0.73, + "grad_norm": 0.46875, + "learning_rate": 0.00014206279139310676, + "loss": 2.5023, + "step": 17045 + }, + { + "epoch": 0.73, + "grad_norm": 0.439453125, + "learning_rate": 0.00014205670563999733, + "loss": 2.4403, + "step": 17046 + }, + { + "epoch": 0.73, + "grad_norm": 0.5390625, + "learning_rate": 0.00014205061969764808, + "loss": 2.5623, + "step": 17047 + }, + { + "epoch": 0.73, + "grad_norm": 0.4375, + "learning_rate": 0.0001420445335660864, + "loss": 2.5368, + "step": 17048 + }, + { + "epoch": 0.73, + "grad_norm": 0.41796875, + "learning_rate": 0.0001420384472453397, + "loss": 2.5994, + "step": 17049 + }, + { + "epoch": 0.73, + "grad_norm": 0.423828125, + "learning_rate": 0.00014203236073543536, + "loss": 2.5859, + "step": 17050 + }, + { + "epoch": 0.73, + "grad_norm": 0.4296875, + "learning_rate": 0.0001420262740364007, + "loss": 2.5856, + "step": 17051 + }, + { + "epoch": 0.73, + "grad_norm": 0.419921875, + "learning_rate": 0.00014202018714826325, + "loss": 2.496, + "step": 17052 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.00014201410007105025, + "loss": 2.6482, + "step": 17053 + }, + { + "epoch": 0.73, + "grad_norm": 0.4609375, + "learning_rate": 0.00014200801280478916, + "loss": 2.5846, + "step": 17054 + }, + { + "epoch": 0.73, + "grad_norm": 0.4453125, + "learning_rate": 0.0001420019253495074, + "loss": 2.4552, + "step": 17055 + }, + { + "epoch": 0.73, + "grad_norm": 0.48046875, + "learning_rate": 0.00014199583770523231, + "loss": 2.5737, + "step": 17056 + }, + { + "epoch": 0.73, + "grad_norm": 0.4296875, + "learning_rate": 0.0001419897498719913, + "loss": 2.4542, + "step": 17057 + }, + { + "epoch": 0.73, + "grad_norm": 0.443359375, + "learning_rate": 0.00014198366184981176, + "loss": 2.5195, + "step": 17058 + }, + { + "epoch": 0.73, + "grad_norm": 0.46484375, + "learning_rate": 0.00014197757363872105, + "loss": 2.5568, + "step": 17059 + }, + { + "epoch": 0.73, + "grad_norm": 0.4765625, + "learning_rate": 0.00014197148523874666, + "loss": 2.4963, + "step": 17060 + }, + { + "epoch": 0.73, + "grad_norm": 0.470703125, + "learning_rate": 0.00014196539664991588, + "loss": 2.5455, + "step": 17061 + }, + { + "epoch": 0.73, + "grad_norm": 0.44921875, + "learning_rate": 0.00014195930787225619, + "loss": 2.564, + "step": 17062 + }, + { + "epoch": 0.73, + "grad_norm": 0.439453125, + "learning_rate": 0.0001419532189057949, + "loss": 2.4372, + "step": 17063 + }, + { + "epoch": 0.73, + "grad_norm": 0.4765625, + "learning_rate": 0.00014194712975055951, + "loss": 2.7304, + "step": 17064 + }, + { + "epoch": 0.73, + "grad_norm": 0.41796875, + "learning_rate": 0.0001419410404065773, + "loss": 2.6201, + "step": 17065 + }, + { + "epoch": 0.73, + "grad_norm": 0.439453125, + "learning_rate": 0.0001419349508738758, + "loss": 2.5111, + "step": 17066 + }, + { + "epoch": 0.73, + "grad_norm": 0.48046875, + "learning_rate": 0.00014192886115248232, + "loss": 2.6545, + "step": 17067 + }, + { + "epoch": 0.73, + "grad_norm": 0.431640625, + "learning_rate": 0.0001419227712424243, + "loss": 2.5228, + "step": 17068 + }, + { + "epoch": 0.73, + "grad_norm": 0.453125, + "learning_rate": 0.00014191668114372908, + "loss": 2.5013, + "step": 17069 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.00014191059085642417, + "loss": 2.5805, + "step": 17070 + }, + { + "epoch": 0.73, + "grad_norm": 0.482421875, + "learning_rate": 0.0001419045003805369, + "loss": 2.6298, + "step": 17071 + }, + { + "epoch": 0.73, + "grad_norm": 0.455078125, + "learning_rate": 0.0001418984097160947, + "loss": 2.6163, + "step": 17072 + }, + { + "epoch": 0.73, + "grad_norm": 0.4609375, + "learning_rate": 0.00014189231886312497, + "loss": 2.5105, + "step": 17073 + }, + { + "epoch": 0.73, + "grad_norm": 0.484375, + "learning_rate": 0.0001418862278216551, + "loss": 2.6745, + "step": 17074 + }, + { + "epoch": 0.73, + "grad_norm": 0.51953125, + "learning_rate": 0.00014188013659171254, + "loss": 2.6538, + "step": 17075 + }, + { + "epoch": 0.73, + "grad_norm": 0.404296875, + "learning_rate": 0.00014187404517332467, + "loss": 2.4902, + "step": 17076 + }, + { + "epoch": 0.73, + "grad_norm": 0.392578125, + "learning_rate": 0.00014186795356651885, + "loss": 2.4925, + "step": 17077 + }, + { + "epoch": 0.73, + "grad_norm": 0.443359375, + "learning_rate": 0.00014186186177132256, + "loss": 2.6144, + "step": 17078 + }, + { + "epoch": 0.73, + "grad_norm": 0.45703125, + "learning_rate": 0.0001418557697877632, + "loss": 2.5382, + "step": 17079 + }, + { + "epoch": 0.73, + "grad_norm": 0.470703125, + "learning_rate": 0.0001418496776158682, + "loss": 2.7014, + "step": 17080 + }, + { + "epoch": 0.73, + "grad_norm": 0.482421875, + "learning_rate": 0.0001418435852556649, + "loss": 2.57, + "step": 17081 + }, + { + "epoch": 0.73, + "grad_norm": 0.43359375, + "learning_rate": 0.0001418374927071808, + "loss": 2.6705, + "step": 17082 + }, + { + "epoch": 0.73, + "grad_norm": 0.42578125, + "learning_rate": 0.00014183139997044322, + "loss": 2.5654, + "step": 17083 + }, + { + "epoch": 0.73, + "grad_norm": 0.5, + "learning_rate": 0.0001418253070454797, + "loss": 2.6953, + "step": 17084 + }, + { + "epoch": 0.73, + "grad_norm": 0.41796875, + "learning_rate": 0.00014181921393231753, + "loss": 2.5291, + "step": 17085 + }, + { + "epoch": 0.73, + "grad_norm": 0.427734375, + "learning_rate": 0.00014181312063098417, + "loss": 2.4612, + "step": 17086 + }, + { + "epoch": 0.73, + "grad_norm": 0.453125, + "learning_rate": 0.00014180702714150712, + "loss": 2.5172, + "step": 17087 + }, + { + "epoch": 0.73, + "grad_norm": 0.41796875, + "learning_rate": 0.00014180093346391362, + "loss": 2.392, + "step": 17088 + }, + { + "epoch": 0.73, + "grad_norm": 0.453125, + "learning_rate": 0.0001417948395982313, + "loss": 2.6357, + "step": 17089 + }, + { + "epoch": 0.73, + "grad_norm": 0.484375, + "learning_rate": 0.0001417887455444874, + "loss": 2.5813, + "step": 17090 + }, + { + "epoch": 0.73, + "grad_norm": 0.45703125, + "learning_rate": 0.00014178265130270946, + "loss": 2.5142, + "step": 17091 + }, + { + "epoch": 0.73, + "grad_norm": 0.5078125, + "learning_rate": 0.00014177655687292484, + "loss": 2.7011, + "step": 17092 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.000141770462255161, + "loss": 2.5165, + "step": 17093 + }, + { + "epoch": 0.73, + "grad_norm": 0.470703125, + "learning_rate": 0.0001417643674494453, + "loss": 2.5056, + "step": 17094 + }, + { + "epoch": 0.73, + "grad_norm": 0.435546875, + "learning_rate": 0.00014175827245580525, + "loss": 2.5949, + "step": 17095 + }, + { + "epoch": 0.73, + "grad_norm": 0.55078125, + "learning_rate": 0.00014175217727426822, + "loss": 2.6029, + "step": 17096 + }, + { + "epoch": 0.73, + "grad_norm": 0.4453125, + "learning_rate": 0.00014174608190486169, + "loss": 2.5223, + "step": 17097 + }, + { + "epoch": 0.73, + "grad_norm": 0.421875, + "learning_rate": 0.00014173998634761298, + "loss": 2.5497, + "step": 17098 + }, + { + "epoch": 0.73, + "grad_norm": 0.4375, + "learning_rate": 0.00014173389060254964, + "loss": 2.4177, + "step": 17099 + }, + { + "epoch": 0.73, + "grad_norm": 0.42578125, + "learning_rate": 0.00014172779466969902, + "loss": 2.6838, + "step": 17100 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.0001417216985490886, + "loss": 2.5724, + "step": 17101 + }, + { + "epoch": 0.73, + "grad_norm": 0.53125, + "learning_rate": 0.00014171560224074574, + "loss": 2.6575, + "step": 17102 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014170950574469795, + "loss": 2.5588, + "step": 17103 + }, + { + "epoch": 0.73, + "grad_norm": 0.455078125, + "learning_rate": 0.0001417034090609726, + "loss": 2.6197, + "step": 17104 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.00014169731218959715, + "loss": 2.5386, + "step": 17105 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014169121513059907, + "loss": 2.5663, + "step": 17106 + }, + { + "epoch": 0.73, + "grad_norm": 0.455078125, + "learning_rate": 0.00014168511788400574, + "loss": 2.5534, + "step": 17107 + }, + { + "epoch": 0.73, + "grad_norm": 0.453125, + "learning_rate": 0.00014167902044984461, + "loss": 2.5846, + "step": 17108 + }, + { + "epoch": 0.73, + "grad_norm": 0.490234375, + "learning_rate": 0.00014167292282814313, + "loss": 2.6171, + "step": 17109 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.0001416668250189287, + "loss": 2.723, + "step": 17110 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.00014166072702222882, + "loss": 2.6008, + "step": 17111 + }, + { + "epoch": 0.73, + "grad_norm": 0.42578125, + "learning_rate": 0.00014165462883807084, + "loss": 2.5617, + "step": 17112 + }, + { + "epoch": 0.73, + "grad_norm": 0.388671875, + "learning_rate": 0.0001416485304664823, + "loss": 2.5728, + "step": 17113 + }, + { + "epoch": 0.73, + "grad_norm": 0.515625, + "learning_rate": 0.0001416424319074906, + "loss": 2.5339, + "step": 17114 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.00014163633316112314, + "loss": 2.5438, + "step": 17115 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014163023422740745, + "loss": 2.5914, + "step": 17116 + }, + { + "epoch": 0.73, + "grad_norm": 0.419921875, + "learning_rate": 0.00014162413510637086, + "loss": 2.5237, + "step": 17117 + }, + { + "epoch": 0.73, + "grad_norm": 0.40234375, + "learning_rate": 0.0001416180357980409, + "loss": 2.5108, + "step": 17118 + }, + { + "epoch": 0.73, + "grad_norm": 0.419921875, + "learning_rate": 0.000141611936302445, + "loss": 2.596, + "step": 17119 + }, + { + "epoch": 0.73, + "grad_norm": 0.490234375, + "learning_rate": 0.00014160583661961058, + "loss": 2.5213, + "step": 17120 + }, + { + "epoch": 0.73, + "grad_norm": 0.490234375, + "learning_rate": 0.0001415997367495651, + "loss": 2.5432, + "step": 17121 + }, + { + "epoch": 0.73, + "grad_norm": 0.4375, + "learning_rate": 0.000141593636692336, + "loss": 2.3936, + "step": 17122 + }, + { + "epoch": 0.73, + "grad_norm": 0.439453125, + "learning_rate": 0.00014158753644795074, + "loss": 2.6425, + "step": 17123 + }, + { + "epoch": 0.73, + "grad_norm": 0.46484375, + "learning_rate": 0.0001415814360164368, + "loss": 2.7238, + "step": 17124 + }, + { + "epoch": 0.73, + "grad_norm": 0.4296875, + "learning_rate": 0.00014157533539782156, + "loss": 2.556, + "step": 17125 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.0001415692345921325, + "loss": 2.5322, + "step": 17126 + }, + { + "epoch": 0.73, + "grad_norm": 0.47265625, + "learning_rate": 0.0001415631335993971, + "loss": 2.5681, + "step": 17127 + }, + { + "epoch": 0.73, + "grad_norm": 0.45703125, + "learning_rate": 0.00014155703241964276, + "loss": 2.5694, + "step": 17128 + }, + { + "epoch": 0.73, + "grad_norm": 0.4765625, + "learning_rate": 0.000141550931052897, + "loss": 2.5207, + "step": 17129 + }, + { + "epoch": 0.73, + "grad_norm": 0.50390625, + "learning_rate": 0.0001415448294991872, + "loss": 2.5808, + "step": 17130 + }, + { + "epoch": 0.73, + "grad_norm": 0.50390625, + "learning_rate": 0.00014153872775854085, + "loss": 2.509, + "step": 17131 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.0001415326258309854, + "loss": 2.5366, + "step": 17132 + }, + { + "epoch": 0.73, + "grad_norm": 0.439453125, + "learning_rate": 0.00014152652371654835, + "loss": 2.6103, + "step": 17133 + }, + { + "epoch": 0.73, + "grad_norm": 0.5, + "learning_rate": 0.0001415204214152571, + "loss": 2.546, + "step": 17134 + }, + { + "epoch": 0.73, + "grad_norm": 0.46484375, + "learning_rate": 0.00014151431892713912, + "loss": 2.616, + "step": 17135 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014150821625222187, + "loss": 2.4699, + "step": 17136 + }, + { + "epoch": 0.73, + "grad_norm": 0.451171875, + "learning_rate": 0.00014150211339053283, + "loss": 2.6988, + "step": 17137 + }, + { + "epoch": 0.73, + "grad_norm": 0.39453125, + "learning_rate": 0.00014149601034209945, + "loss": 2.5514, + "step": 17138 + }, + { + "epoch": 0.73, + "grad_norm": 0.48828125, + "learning_rate": 0.0001414899071069492, + "loss": 2.6786, + "step": 17139 + }, + { + "epoch": 0.73, + "grad_norm": 0.404296875, + "learning_rate": 0.0001414838036851095, + "loss": 2.538, + "step": 17140 + }, + { + "epoch": 0.73, + "grad_norm": 0.50390625, + "learning_rate": 0.00014147770007660786, + "loss": 2.5095, + "step": 17141 + }, + { + "epoch": 0.73, + "grad_norm": 0.435546875, + "learning_rate": 0.0001414715962814717, + "loss": 2.4513, + "step": 17142 + }, + { + "epoch": 0.73, + "grad_norm": 0.46484375, + "learning_rate": 0.00014146549229972859, + "loss": 2.5774, + "step": 17143 + }, + { + "epoch": 0.73, + "grad_norm": 0.46875, + "learning_rate": 0.00014145938813140584, + "loss": 2.6675, + "step": 17144 + }, + { + "epoch": 0.73, + "grad_norm": 0.484375, + "learning_rate": 0.00014145328377653104, + "loss": 2.4996, + "step": 17145 + }, + { + "epoch": 0.73, + "grad_norm": 0.443359375, + "learning_rate": 0.0001414471792351316, + "loss": 2.5817, + "step": 17146 + }, + { + "epoch": 0.73, + "grad_norm": 0.5390625, + "learning_rate": 0.00014144107450723498, + "loss": 2.6121, + "step": 17147 + }, + { + "epoch": 0.73, + "grad_norm": 0.478515625, + "learning_rate": 0.00014143496959286868, + "loss": 2.5022, + "step": 17148 + }, + { + "epoch": 0.73, + "grad_norm": 0.470703125, + "learning_rate": 0.0001414288644920602, + "loss": 2.5261, + "step": 17149 + }, + { + "epoch": 0.73, + "grad_norm": 0.455078125, + "learning_rate": 0.0001414227592048369, + "loss": 2.6258, + "step": 17150 + }, + { + "epoch": 0.73, + "grad_norm": 0.486328125, + "learning_rate": 0.00014141665373122638, + "loss": 2.6196, + "step": 17151 + }, + { + "epoch": 0.73, + "grad_norm": 0.50390625, + "learning_rate": 0.000141410548071256, + "loss": 2.5921, + "step": 17152 + }, + { + "epoch": 0.73, + "grad_norm": 0.435546875, + "learning_rate": 0.00014140444222495336, + "loss": 2.5917, + "step": 17153 + }, + { + "epoch": 0.73, + "grad_norm": 0.431640625, + "learning_rate": 0.0001413983361923458, + "loss": 2.6769, + "step": 17154 + }, + { + "epoch": 0.73, + "grad_norm": 0.421875, + "learning_rate": 0.00014139222997346088, + "loss": 2.5022, + "step": 17155 + }, + { + "epoch": 0.73, + "grad_norm": 0.46875, + "learning_rate": 0.00014138612356832605, + "loss": 2.5847, + "step": 17156 + }, + { + "epoch": 0.73, + "grad_norm": 0.494140625, + "learning_rate": 0.0001413800169769688, + "loss": 2.7356, + "step": 17157 + }, + { + "epoch": 0.73, + "grad_norm": 0.423828125, + "learning_rate": 0.00014137391019941657, + "loss": 2.5151, + "step": 17158 + }, + { + "epoch": 0.73, + "grad_norm": 0.5, + "learning_rate": 0.00014136780323569689, + "loss": 2.7907, + "step": 17159 + }, + { + "epoch": 0.73, + "grad_norm": 0.515625, + "learning_rate": 0.0001413616960858372, + "loss": 2.5975, + "step": 17160 + }, + { + "epoch": 0.73, + "grad_norm": 0.484375, + "learning_rate": 0.00014135558874986499, + "loss": 2.7301, + "step": 17161 + }, + { + "epoch": 0.73, + "grad_norm": 0.416015625, + "learning_rate": 0.00014134948122780774, + "loss": 2.5611, + "step": 17162 + }, + { + "epoch": 0.73, + "grad_norm": 0.490234375, + "learning_rate": 0.00014134337351969298, + "loss": 2.5997, + "step": 17163 + }, + { + "epoch": 0.73, + "grad_norm": 0.52734375, + "learning_rate": 0.0001413372656255481, + "loss": 2.6045, + "step": 17164 + }, + { + "epoch": 0.73, + "grad_norm": 0.51953125, + "learning_rate": 0.00014133115754540065, + "loss": 2.4866, + "step": 17165 + }, + { + "epoch": 0.73, + "grad_norm": 0.4609375, + "learning_rate": 0.0001413250492792781, + "loss": 2.638, + "step": 17166 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014131894082720795, + "loss": 2.5842, + "step": 17167 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.0001413128321892176, + "loss": 2.6765, + "step": 17168 + }, + { + "epoch": 0.73, + "grad_norm": 0.474609375, + "learning_rate": 0.00014130672336533465, + "loss": 2.6533, + "step": 17169 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014130061435558655, + "loss": 2.5865, + "step": 17170 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014129450516000075, + "loss": 2.587, + "step": 17171 + }, + { + "epoch": 0.73, + "grad_norm": 0.51171875, + "learning_rate": 0.0001412883957786048, + "loss": 2.4924, + "step": 17172 + }, + { + "epoch": 0.73, + "grad_norm": 0.4375, + "learning_rate": 0.00014128228621142615, + "loss": 2.6337, + "step": 17173 + }, + { + "epoch": 0.73, + "grad_norm": 0.447265625, + "learning_rate": 0.00014127617645849226, + "loss": 2.6242, + "step": 17174 + }, + { + "epoch": 0.73, + "grad_norm": 0.56640625, + "learning_rate": 0.00014127006651983072, + "loss": 2.5985, + "step": 17175 + }, + { + "epoch": 0.73, + "grad_norm": 0.474609375, + "learning_rate": 0.0001412639563954689, + "loss": 2.6148, + "step": 17176 + }, + { + "epoch": 0.73, + "grad_norm": 0.44921875, + "learning_rate": 0.0001412578460854344, + "loss": 2.5128, + "step": 17177 + }, + { + "epoch": 0.73, + "grad_norm": 0.470703125, + "learning_rate": 0.00014125173558975467, + "loss": 2.5802, + "step": 17178 + }, + { + "epoch": 0.73, + "grad_norm": 0.451171875, + "learning_rate": 0.0001412456249084572, + "loss": 2.5717, + "step": 17179 + }, + { + "epoch": 0.73, + "grad_norm": 0.451171875, + "learning_rate": 0.00014123951404156946, + "loss": 2.7053, + "step": 17180 + }, + { + "epoch": 0.73, + "grad_norm": 0.419921875, + "learning_rate": 0.000141233402989119, + "loss": 2.4736, + "step": 17181 + }, + { + "epoch": 0.73, + "grad_norm": 0.494140625, + "learning_rate": 0.00014122729175113332, + "loss": 2.5542, + "step": 17182 + }, + { + "epoch": 0.73, + "grad_norm": 0.48046875, + "learning_rate": 0.0001412211803276399, + "loss": 2.5818, + "step": 17183 + }, + { + "epoch": 0.73, + "grad_norm": 0.48828125, + "learning_rate": 0.00014121506871866618, + "loss": 2.5604, + "step": 17184 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.00014120895692423975, + "loss": 2.503, + "step": 17185 + }, + { + "epoch": 0.73, + "grad_norm": 0.4765625, + "learning_rate": 0.00014120284494438808, + "loss": 2.6591, + "step": 17186 + }, + { + "epoch": 0.73, + "grad_norm": 0.55859375, + "learning_rate": 0.00014119673277913862, + "loss": 2.6302, + "step": 17187 + }, + { + "epoch": 0.73, + "grad_norm": 0.46875, + "learning_rate": 0.000141190620428519, + "loss": 2.6129, + "step": 17188 + }, + { + "epoch": 0.73, + "grad_norm": 0.4765625, + "learning_rate": 0.00014118450789255658, + "loss": 2.5132, + "step": 17189 + }, + { + "epoch": 0.73, + "grad_norm": 0.455078125, + "learning_rate": 0.0001411783951712789, + "loss": 2.5753, + "step": 17190 + }, + { + "epoch": 0.73, + "grad_norm": 0.470703125, + "learning_rate": 0.00014117228226471354, + "loss": 2.5094, + "step": 17191 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.00014116616917288793, + "loss": 2.5542, + "step": 17192 + }, + { + "epoch": 0.73, + "grad_norm": 0.490234375, + "learning_rate": 0.00014116005589582964, + "loss": 2.5721, + "step": 17193 + }, + { + "epoch": 0.73, + "grad_norm": 0.478515625, + "learning_rate": 0.0001411539424335661, + "loss": 2.6324, + "step": 17194 + }, + { + "epoch": 0.73, + "grad_norm": 0.48828125, + "learning_rate": 0.0001411478287861249, + "loss": 2.6344, + "step": 17195 + }, + { + "epoch": 0.73, + "grad_norm": 0.412109375, + "learning_rate": 0.00014114171495353348, + "loss": 2.5504, + "step": 17196 + }, + { + "epoch": 0.73, + "grad_norm": 0.462890625, + "learning_rate": 0.00014113560093581939, + "loss": 2.632, + "step": 17197 + }, + { + "epoch": 0.73, + "grad_norm": 0.404296875, + "learning_rate": 0.0001411294867330101, + "loss": 2.5082, + "step": 17198 + }, + { + "epoch": 0.73, + "grad_norm": 0.486328125, + "learning_rate": 0.00014112337234513318, + "loss": 2.6538, + "step": 17199 + }, + { + "epoch": 0.73, + "grad_norm": 0.482421875, + "learning_rate": 0.0001411172577722161, + "loss": 2.6399, + "step": 17200 + }, + { + "epoch": 0.73, + "grad_norm": 0.4921875, + "learning_rate": 0.00014111114301428642, + "loss": 2.7756, + "step": 17201 + }, + { + "epoch": 0.73, + "grad_norm": 0.48828125, + "learning_rate": 0.00014110502807137157, + "loss": 2.6058, + "step": 17202 + }, + { + "epoch": 0.73, + "grad_norm": 0.4375, + "learning_rate": 0.00014109891294349916, + "loss": 2.6298, + "step": 17203 + }, + { + "epoch": 0.73, + "grad_norm": 0.53515625, + "learning_rate": 0.00014109279763069662, + "loss": 2.5656, + "step": 17204 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.00014108668213299154, + "loss": 2.7671, + "step": 17205 + }, + { + "epoch": 0.73, + "grad_norm": 0.46484375, + "learning_rate": 0.0001410805664504114, + "loss": 2.6093, + "step": 17206 + }, + { + "epoch": 0.73, + "grad_norm": 0.4375, + "learning_rate": 0.00014107445058298372, + "loss": 2.5586, + "step": 17207 + }, + { + "epoch": 0.73, + "grad_norm": 0.4375, + "learning_rate": 0.000141068334530736, + "loss": 2.6455, + "step": 17208 + }, + { + "epoch": 0.73, + "grad_norm": 0.486328125, + "learning_rate": 0.00014106221829369582, + "loss": 2.6913, + "step": 17209 + }, + { + "epoch": 0.73, + "grad_norm": 0.5234375, + "learning_rate": 0.00014105610187189067, + "loss": 2.6792, + "step": 17210 + }, + { + "epoch": 0.73, + "grad_norm": 0.55859375, + "learning_rate": 0.00014104998526534804, + "loss": 2.7959, + "step": 17211 + }, + { + "epoch": 0.73, + "grad_norm": 0.419921875, + "learning_rate": 0.00014104386847409547, + "loss": 2.4598, + "step": 17212 + }, + { + "epoch": 0.73, + "grad_norm": 0.41015625, + "learning_rate": 0.00014103775149816052, + "loss": 2.4344, + "step": 17213 + }, + { + "epoch": 0.73, + "grad_norm": 0.451171875, + "learning_rate": 0.00014103163433757068, + "loss": 2.5634, + "step": 17214 + }, + { + "epoch": 0.73, + "grad_norm": 0.41796875, + "learning_rate": 0.00014102551699235347, + "loss": 2.5073, + "step": 17215 + }, + { + "epoch": 0.73, + "grad_norm": 0.43359375, + "learning_rate": 0.00014101939946253644, + "loss": 2.5657, + "step": 17216 + }, + { + "epoch": 0.73, + "grad_norm": 0.48046875, + "learning_rate": 0.00014101328174814708, + "loss": 2.4828, + "step": 17217 + }, + { + "epoch": 0.73, + "grad_norm": 0.51171875, + "learning_rate": 0.00014100716384921297, + "loss": 2.7048, + "step": 17218 + }, + { + "epoch": 0.73, + "grad_norm": 0.427734375, + "learning_rate": 0.00014100104576576157, + "loss": 2.5118, + "step": 17219 + }, + { + "epoch": 0.73, + "grad_norm": 0.5078125, + "learning_rate": 0.0001409949274978205, + "loss": 2.4434, + "step": 17220 + }, + { + "epoch": 0.73, + "grad_norm": 0.47265625, + "learning_rate": 0.00014098880904541718, + "loss": 2.7191, + "step": 17221 + }, + { + "epoch": 0.73, + "grad_norm": 0.5, + "learning_rate": 0.00014098269040857924, + "loss": 2.5415, + "step": 17222 + }, + { + "epoch": 0.73, + "grad_norm": 0.44140625, + "learning_rate": 0.00014097657158733417, + "loss": 2.6528, + "step": 17223 + }, + { + "epoch": 0.73, + "grad_norm": 0.5, + "learning_rate": 0.0001409704525817095, + "loss": 2.6127, + "step": 17224 + }, + { + "epoch": 0.73, + "grad_norm": 0.44921875, + "learning_rate": 0.00014096433339173275, + "loss": 2.4921, + "step": 17225 + }, + { + "epoch": 0.73, + "grad_norm": 0.494140625, + "learning_rate": 0.0001409582140174315, + "loss": 2.6543, + "step": 17226 + }, + { + "epoch": 0.73, + "grad_norm": 0.486328125, + "learning_rate": 0.00014095209445883319, + "loss": 2.5853, + "step": 17227 + }, + { + "epoch": 0.73, + "grad_norm": 0.451171875, + "learning_rate": 0.00014094597471596548, + "loss": 2.4124, + "step": 17228 + }, + { + "epoch": 0.73, + "grad_norm": 0.482421875, + "learning_rate": 0.00014093985478885583, + "loss": 2.7161, + "step": 17229 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014093373467753182, + "loss": 2.6425, + "step": 17230 + }, + { + "epoch": 0.73, + "grad_norm": 0.439453125, + "learning_rate": 0.00014092761438202093, + "loss": 2.6794, + "step": 17231 + }, + { + "epoch": 0.73, + "grad_norm": 0.482421875, + "learning_rate": 0.00014092149390235075, + "loss": 2.6784, + "step": 17232 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014091537323854877, + "loss": 2.5674, + "step": 17233 + }, + { + "epoch": 0.73, + "grad_norm": 0.466796875, + "learning_rate": 0.00014090925239064264, + "loss": 2.6242, + "step": 17234 + }, + { + "epoch": 0.73, + "grad_norm": 0.484375, + "learning_rate": 0.00014090313135865974, + "loss": 2.6912, + "step": 17235 + }, + { + "epoch": 0.73, + "grad_norm": 0.427734375, + "learning_rate": 0.00014089701014262772, + "loss": 2.6132, + "step": 17236 + }, + { + "epoch": 0.73, + "grad_norm": 0.451171875, + "learning_rate": 0.00014089088874257412, + "loss": 2.5315, + "step": 17237 + }, + { + "epoch": 0.73, + "grad_norm": 0.458984375, + "learning_rate": 0.00014088476715852648, + "loss": 2.5586, + "step": 17238 + }, + { + "epoch": 0.73, + "grad_norm": 0.46484375, + "learning_rate": 0.0001408786453905123, + "loss": 2.5196, + "step": 17239 + }, + { + "epoch": 0.73, + "grad_norm": 0.5234375, + "learning_rate": 0.00014087252343855916, + "loss": 2.6779, + "step": 17240 + }, + { + "epoch": 0.73, + "grad_norm": 0.486328125, + "learning_rate": 0.0001408664013026946, + "loss": 2.5533, + "step": 17241 + }, + { + "epoch": 0.73, + "grad_norm": 0.443359375, + "learning_rate": 0.0001408602789829462, + "loss": 2.6174, + "step": 17242 + }, + { + "epoch": 0.73, + "grad_norm": 0.447265625, + "learning_rate": 0.00014085415647934144, + "loss": 2.604, + "step": 17243 + }, + { + "epoch": 0.73, + "grad_norm": 0.4375, + "learning_rate": 0.00014084803379190792, + "loss": 2.5564, + "step": 17244 + }, + { + "epoch": 0.73, + "grad_norm": 0.41796875, + "learning_rate": 0.00014084191092067315, + "loss": 2.477, + "step": 17245 + }, + { + "epoch": 0.73, + "grad_norm": 0.451171875, + "learning_rate": 0.00014083578786566476, + "loss": 2.5896, + "step": 17246 + }, + { + "epoch": 0.73, + "grad_norm": 0.45703125, + "learning_rate": 0.0001408296646269102, + "loss": 2.5483, + "step": 17247 + }, + { + "epoch": 0.73, + "grad_norm": 0.4453125, + "learning_rate": 0.0001408235412044371, + "loss": 2.5953, + "step": 17248 + }, + { + "epoch": 0.74, + "grad_norm": 0.515625, + "learning_rate": 0.00014081741759827297, + "loss": 2.642, + "step": 17249 + }, + { + "epoch": 0.74, + "grad_norm": 0.482421875, + "learning_rate": 0.0001408112938084454, + "loss": 2.6495, + "step": 17250 + }, + { + "epoch": 0.74, + "grad_norm": 0.4609375, + "learning_rate": 0.00014080516983498188, + "loss": 2.5857, + "step": 17251 + }, + { + "epoch": 0.74, + "grad_norm": 0.46484375, + "learning_rate": 0.00014079904567791001, + "loss": 2.563, + "step": 17252 + }, + { + "epoch": 0.74, + "grad_norm": 0.490234375, + "learning_rate": 0.0001407929213372574, + "loss": 2.7336, + "step": 17253 + }, + { + "epoch": 0.74, + "grad_norm": 0.427734375, + "learning_rate": 0.00014078679681305152, + "loss": 2.4337, + "step": 17254 + }, + { + "epoch": 0.74, + "grad_norm": 0.55078125, + "learning_rate": 0.00014078067210531996, + "loss": 2.5918, + "step": 17255 + }, + { + "epoch": 0.74, + "grad_norm": 0.43359375, + "learning_rate": 0.00014077454721409026, + "loss": 2.5899, + "step": 17256 + }, + { + "epoch": 0.74, + "grad_norm": 0.52734375, + "learning_rate": 0.00014076842213939003, + "loss": 2.6025, + "step": 17257 + }, + { + "epoch": 0.74, + "grad_norm": 0.44921875, + "learning_rate": 0.0001407622968812468, + "loss": 2.5203, + "step": 17258 + }, + { + "epoch": 0.74, + "grad_norm": 0.462890625, + "learning_rate": 0.00014075617143968812, + "loss": 2.6309, + "step": 17259 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.00014075004581474155, + "loss": 2.6345, + "step": 17260 + }, + { + "epoch": 0.74, + "grad_norm": 0.53515625, + "learning_rate": 0.00014074392000643467, + "loss": 2.4356, + "step": 17261 + }, + { + "epoch": 0.74, + "grad_norm": 0.453125, + "learning_rate": 0.00014073779401479507, + "loss": 2.6452, + "step": 17262 + }, + { + "epoch": 0.74, + "grad_norm": 0.478515625, + "learning_rate": 0.00014073166783985027, + "loss": 2.4548, + "step": 17263 + }, + { + "epoch": 0.74, + "grad_norm": 0.474609375, + "learning_rate": 0.0001407255414816278, + "loss": 2.5017, + "step": 17264 + }, + { + "epoch": 0.74, + "grad_norm": 0.404296875, + "learning_rate": 0.00014071941494015532, + "loss": 2.5619, + "step": 17265 + }, + { + "epoch": 0.74, + "grad_norm": 0.451171875, + "learning_rate": 0.0001407132882154604, + "loss": 2.5549, + "step": 17266 + }, + { + "epoch": 0.74, + "grad_norm": 0.48828125, + "learning_rate": 0.00014070716130757047, + "loss": 2.5616, + "step": 17267 + }, + { + "epoch": 0.74, + "grad_norm": 0.458984375, + "learning_rate": 0.00014070103421651324, + "loss": 2.6225, + "step": 17268 + }, + { + "epoch": 0.74, + "grad_norm": 0.453125, + "learning_rate": 0.00014069490694231624, + "loss": 2.5913, + "step": 17269 + }, + { + "epoch": 0.74, + "grad_norm": 0.478515625, + "learning_rate": 0.00014068877948500704, + "loss": 2.6667, + "step": 17270 + }, + { + "epoch": 0.74, + "grad_norm": 0.48828125, + "learning_rate": 0.00014068265184461316, + "loss": 2.721, + "step": 17271 + }, + { + "epoch": 0.74, + "grad_norm": 0.42578125, + "learning_rate": 0.00014067652402116224, + "loss": 2.4201, + "step": 17272 + }, + { + "epoch": 0.74, + "grad_norm": 0.388671875, + "learning_rate": 0.0001406703960146818, + "loss": 2.5262, + "step": 17273 + }, + { + "epoch": 0.74, + "grad_norm": 0.40234375, + "learning_rate": 0.00014066426782519947, + "loss": 2.4319, + "step": 17274 + }, + { + "epoch": 0.74, + "grad_norm": 0.47265625, + "learning_rate": 0.0001406581394527428, + "loss": 2.5885, + "step": 17275 + }, + { + "epoch": 0.74, + "grad_norm": 0.4375, + "learning_rate": 0.00014065201089733935, + "loss": 2.6355, + "step": 17276 + }, + { + "epoch": 0.74, + "grad_norm": 0.5, + "learning_rate": 0.0001406458821590167, + "loss": 2.7397, + "step": 17277 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.00014063975323780245, + "loss": 2.5119, + "step": 17278 + }, + { + "epoch": 0.74, + "grad_norm": 0.4921875, + "learning_rate": 0.00014063362413372418, + "loss": 2.5637, + "step": 17279 + }, + { + "epoch": 0.74, + "grad_norm": 0.447265625, + "learning_rate": 0.0001406274948468094, + "loss": 2.4892, + "step": 17280 + }, + { + "epoch": 0.74, + "grad_norm": 0.45703125, + "learning_rate": 0.00014062136537708577, + "loss": 2.5578, + "step": 17281 + }, + { + "epoch": 0.74, + "grad_norm": 0.4765625, + "learning_rate": 0.00014061523572458085, + "loss": 2.6179, + "step": 17282 + }, + { + "epoch": 0.74, + "grad_norm": 0.46875, + "learning_rate": 0.0001406091058893222, + "loss": 2.6194, + "step": 17283 + }, + { + "epoch": 0.74, + "grad_norm": 0.46875, + "learning_rate": 0.0001406029758713374, + "loss": 2.6069, + "step": 17284 + }, + { + "epoch": 0.74, + "grad_norm": 0.4140625, + "learning_rate": 0.00014059684567065408, + "loss": 2.4546, + "step": 17285 + }, + { + "epoch": 0.74, + "grad_norm": 0.494140625, + "learning_rate": 0.00014059071528729978, + "loss": 2.6114, + "step": 17286 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.00014058458472130208, + "loss": 2.6571, + "step": 17287 + }, + { + "epoch": 0.74, + "grad_norm": 0.439453125, + "learning_rate": 0.0001405784539726886, + "loss": 2.5436, + "step": 17288 + }, + { + "epoch": 0.74, + "grad_norm": 0.484375, + "learning_rate": 0.00014057232304148687, + "loss": 2.5276, + "step": 17289 + }, + { + "epoch": 0.74, + "grad_norm": 0.458984375, + "learning_rate": 0.0001405661919277246, + "loss": 2.6076, + "step": 17290 + }, + { + "epoch": 0.74, + "grad_norm": 0.48046875, + "learning_rate": 0.00014056006063142918, + "loss": 2.7504, + "step": 17291 + }, + { + "epoch": 0.74, + "grad_norm": 0.5078125, + "learning_rate": 0.00014055392915262837, + "loss": 2.4501, + "step": 17292 + }, + { + "epoch": 0.74, + "grad_norm": 0.390625, + "learning_rate": 0.00014054779749134969, + "loss": 2.4015, + "step": 17293 + }, + { + "epoch": 0.74, + "grad_norm": 0.423828125, + "learning_rate": 0.00014054166564762076, + "loss": 2.583, + "step": 17294 + }, + { + "epoch": 0.74, + "grad_norm": 0.421875, + "learning_rate": 0.0001405355336214691, + "loss": 2.5193, + "step": 17295 + }, + { + "epoch": 0.74, + "grad_norm": 0.455078125, + "learning_rate": 0.0001405294014129224, + "loss": 2.5421, + "step": 17296 + }, + { + "epoch": 0.74, + "grad_norm": 0.435546875, + "learning_rate": 0.00014052326902200818, + "loss": 2.4923, + "step": 17297 + }, + { + "epoch": 0.74, + "grad_norm": 0.47265625, + "learning_rate": 0.0001405171364487541, + "loss": 2.6445, + "step": 17298 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.0001405110036931877, + "loss": 2.5077, + "step": 17299 + }, + { + "epoch": 0.74, + "grad_norm": 0.458984375, + "learning_rate": 0.00014050487075533655, + "loss": 2.5938, + "step": 17300 + }, + { + "epoch": 0.74, + "grad_norm": 0.439453125, + "learning_rate": 0.00014049873763522828, + "loss": 2.5688, + "step": 17301 + }, + { + "epoch": 0.74, + "grad_norm": 0.435546875, + "learning_rate": 0.00014049260433289054, + "loss": 2.5213, + "step": 17302 + }, + { + "epoch": 0.74, + "grad_norm": 0.4609375, + "learning_rate": 0.00014048647084835084, + "loss": 2.5856, + "step": 17303 + }, + { + "epoch": 0.74, + "grad_norm": 0.48828125, + "learning_rate": 0.00014048033718163687, + "loss": 2.6291, + "step": 17304 + }, + { + "epoch": 0.74, + "grad_norm": 0.51171875, + "learning_rate": 0.00014047420333277612, + "loss": 2.717, + "step": 17305 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.00014046806930179627, + "loss": 2.7063, + "step": 17306 + }, + { + "epoch": 0.74, + "grad_norm": 0.4765625, + "learning_rate": 0.00014046193508872489, + "loss": 2.7537, + "step": 17307 + }, + { + "epoch": 0.74, + "grad_norm": 0.458984375, + "learning_rate": 0.00014045580069358958, + "loss": 2.6612, + "step": 17308 + }, + { + "epoch": 0.74, + "grad_norm": 0.482421875, + "learning_rate": 0.00014044966611641798, + "loss": 2.6445, + "step": 17309 + }, + { + "epoch": 0.74, + "grad_norm": 0.5390625, + "learning_rate": 0.00014044353135723764, + "loss": 2.5376, + "step": 17310 + }, + { + "epoch": 0.74, + "grad_norm": 0.4609375, + "learning_rate": 0.00014043739641607622, + "loss": 2.517, + "step": 17311 + }, + { + "epoch": 0.74, + "grad_norm": 0.55078125, + "learning_rate": 0.00014043126129296128, + "loss": 2.6304, + "step": 17312 + }, + { + "epoch": 0.74, + "grad_norm": 0.412109375, + "learning_rate": 0.0001404251259879204, + "loss": 2.5213, + "step": 17313 + }, + { + "epoch": 0.74, + "grad_norm": 0.40625, + "learning_rate": 0.00014041899050098127, + "loss": 2.5485, + "step": 17314 + }, + { + "epoch": 0.74, + "grad_norm": 0.478515625, + "learning_rate": 0.00014041285483217146, + "loss": 2.5358, + "step": 17315 + }, + { + "epoch": 0.74, + "grad_norm": 0.431640625, + "learning_rate": 0.00014040671898151857, + "loss": 2.5405, + "step": 17316 + }, + { + "epoch": 0.74, + "grad_norm": 0.41796875, + "learning_rate": 0.0001404005829490502, + "loss": 2.5812, + "step": 17317 + }, + { + "epoch": 0.74, + "grad_norm": 0.46875, + "learning_rate": 0.00014039444673479397, + "loss": 2.5148, + "step": 17318 + }, + { + "epoch": 0.74, + "grad_norm": 0.498046875, + "learning_rate": 0.00014038831033877746, + "loss": 2.5895, + "step": 17319 + }, + { + "epoch": 0.74, + "grad_norm": 0.400390625, + "learning_rate": 0.0001403821737610284, + "loss": 2.5885, + "step": 17320 + }, + { + "epoch": 0.74, + "grad_norm": 0.4140625, + "learning_rate": 0.00014037603700157422, + "loss": 2.5093, + "step": 17321 + }, + { + "epoch": 0.74, + "grad_norm": 0.4453125, + "learning_rate": 0.00014036990006044268, + "loss": 2.625, + "step": 17322 + }, + { + "epoch": 0.74, + "grad_norm": 0.451171875, + "learning_rate": 0.00014036376293766133, + "loss": 2.6726, + "step": 17323 + }, + { + "epoch": 0.74, + "grad_norm": 0.419921875, + "learning_rate": 0.00014035762563325782, + "loss": 2.6854, + "step": 17324 + }, + { + "epoch": 0.74, + "grad_norm": 0.5078125, + "learning_rate": 0.00014035148814725971, + "loss": 2.6306, + "step": 17325 + }, + { + "epoch": 0.74, + "grad_norm": 0.44921875, + "learning_rate": 0.0001403453504796947, + "loss": 2.639, + "step": 17326 + }, + { + "epoch": 0.74, + "grad_norm": 0.486328125, + "learning_rate": 0.00014033921263059029, + "loss": 2.6851, + "step": 17327 + }, + { + "epoch": 0.74, + "grad_norm": 0.4453125, + "learning_rate": 0.00014033307459997423, + "loss": 2.5138, + "step": 17328 + }, + { + "epoch": 0.74, + "grad_norm": 0.4375, + "learning_rate": 0.00014032693638787403, + "loss": 2.5013, + "step": 17329 + }, + { + "epoch": 0.74, + "grad_norm": 0.4453125, + "learning_rate": 0.00014032079799431734, + "loss": 2.5618, + "step": 17330 + }, + { + "epoch": 0.74, + "grad_norm": 0.3984375, + "learning_rate": 0.00014031465941933185, + "loss": 2.5035, + "step": 17331 + }, + { + "epoch": 0.74, + "grad_norm": 0.478515625, + "learning_rate": 0.0001403085206629451, + "loss": 2.4819, + "step": 17332 + }, + { + "epoch": 0.74, + "grad_norm": 0.431640625, + "learning_rate": 0.00014030238172518475, + "loss": 2.4444, + "step": 17333 + }, + { + "epoch": 0.74, + "grad_norm": 0.46875, + "learning_rate": 0.0001402962426060784, + "loss": 2.5789, + "step": 17334 + }, + { + "epoch": 0.74, + "grad_norm": 0.44921875, + "learning_rate": 0.00014029010330565366, + "loss": 2.6129, + "step": 17335 + }, + { + "epoch": 0.74, + "grad_norm": 0.42578125, + "learning_rate": 0.0001402839638239382, + "loss": 2.3706, + "step": 17336 + }, + { + "epoch": 0.74, + "grad_norm": 0.482421875, + "learning_rate": 0.00014027782416095965, + "loss": 2.6872, + "step": 17337 + }, + { + "epoch": 0.74, + "grad_norm": 0.451171875, + "learning_rate": 0.00014027168431674555, + "loss": 2.5986, + "step": 17338 + }, + { + "epoch": 0.74, + "grad_norm": 0.4453125, + "learning_rate": 0.00014026554429132367, + "loss": 2.4972, + "step": 17339 + }, + { + "epoch": 0.74, + "grad_norm": 0.494140625, + "learning_rate": 0.0001402594040847215, + "loss": 2.6189, + "step": 17340 + }, + { + "epoch": 0.74, + "grad_norm": 0.439453125, + "learning_rate": 0.00014025326369696673, + "loss": 2.6802, + "step": 17341 + }, + { + "epoch": 0.74, + "grad_norm": 0.451171875, + "learning_rate": 0.000140247123128087, + "loss": 2.5795, + "step": 17342 + }, + { + "epoch": 0.74, + "grad_norm": 0.4921875, + "learning_rate": 0.0001402409823781099, + "loss": 2.7754, + "step": 17343 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.0001402348414470631, + "loss": 2.5865, + "step": 17344 + }, + { + "epoch": 0.74, + "grad_norm": 0.40625, + "learning_rate": 0.00014022870033497423, + "loss": 2.4384, + "step": 17345 + }, + { + "epoch": 0.74, + "grad_norm": 0.49609375, + "learning_rate": 0.0001402225590418709, + "loss": 2.6725, + "step": 17346 + }, + { + "epoch": 0.74, + "grad_norm": 0.482421875, + "learning_rate": 0.00014021641756778075, + "loss": 2.6173, + "step": 17347 + }, + { + "epoch": 0.74, + "grad_norm": 0.408203125, + "learning_rate": 0.00014021027591273142, + "loss": 2.5858, + "step": 17348 + }, + { + "epoch": 0.74, + "grad_norm": 0.48828125, + "learning_rate": 0.00014020413407675054, + "loss": 2.6977, + "step": 17349 + }, + { + "epoch": 0.74, + "grad_norm": 0.412109375, + "learning_rate": 0.00014019799205986576, + "loss": 2.5299, + "step": 17350 + }, + { + "epoch": 0.74, + "grad_norm": 0.443359375, + "learning_rate": 0.00014019184986210472, + "loss": 2.5741, + "step": 17351 + }, + { + "epoch": 0.74, + "grad_norm": 0.478515625, + "learning_rate": 0.00014018570748349502, + "loss": 2.6967, + "step": 17352 + }, + { + "epoch": 0.74, + "grad_norm": 0.48046875, + "learning_rate": 0.00014017956492406436, + "loss": 2.6028, + "step": 17353 + }, + { + "epoch": 0.74, + "grad_norm": 0.4609375, + "learning_rate": 0.00014017342218384027, + "loss": 2.4504, + "step": 17354 + }, + { + "epoch": 0.74, + "grad_norm": 0.400390625, + "learning_rate": 0.00014016727926285052, + "loss": 2.4357, + "step": 17355 + }, + { + "epoch": 0.74, + "grad_norm": 0.486328125, + "learning_rate": 0.00014016113616112266, + "loss": 2.6188, + "step": 17356 + }, + { + "epoch": 0.74, + "grad_norm": 0.482421875, + "learning_rate": 0.00014015499287868435, + "loss": 2.6679, + "step": 17357 + }, + { + "epoch": 0.74, + "grad_norm": 0.498046875, + "learning_rate": 0.0001401488494155633, + "loss": 2.6537, + "step": 17358 + }, + { + "epoch": 0.74, + "grad_norm": 0.494140625, + "learning_rate": 0.0001401427057717871, + "loss": 2.6346, + "step": 17359 + }, + { + "epoch": 0.74, + "grad_norm": 0.46875, + "learning_rate": 0.00014013656194738336, + "loss": 2.7168, + "step": 17360 + }, + { + "epoch": 0.74, + "grad_norm": 0.451171875, + "learning_rate": 0.0001401304179423798, + "loss": 2.5297, + "step": 17361 + }, + { + "epoch": 0.74, + "grad_norm": 0.423828125, + "learning_rate": 0.00014012427375680394, + "loss": 2.4329, + "step": 17362 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.0001401181293906836, + "loss": 2.4639, + "step": 17363 + }, + { + "epoch": 0.74, + "grad_norm": 0.5, + "learning_rate": 0.0001401119848440463, + "loss": 2.5718, + "step": 17364 + }, + { + "epoch": 0.74, + "grad_norm": 0.47265625, + "learning_rate": 0.00014010584011691973, + "loss": 2.5933, + "step": 17365 + }, + { + "epoch": 0.74, + "grad_norm": 0.47265625, + "learning_rate": 0.00014009969520933154, + "loss": 2.5306, + "step": 17366 + }, + { + "epoch": 0.74, + "grad_norm": 0.55859375, + "learning_rate": 0.0001400935501213094, + "loss": 2.6925, + "step": 17367 + }, + { + "epoch": 0.74, + "grad_norm": 0.419921875, + "learning_rate": 0.0001400874048528809, + "loss": 2.4492, + "step": 17368 + }, + { + "epoch": 0.74, + "grad_norm": 0.408203125, + "learning_rate": 0.00014008125940407376, + "loss": 2.4863, + "step": 17369 + }, + { + "epoch": 0.74, + "grad_norm": 0.470703125, + "learning_rate": 0.00014007511377491554, + "loss": 2.5208, + "step": 17370 + }, + { + "epoch": 0.74, + "grad_norm": 0.5, + "learning_rate": 0.00014006896796543402, + "loss": 2.6425, + "step": 17371 + }, + { + "epoch": 0.74, + "grad_norm": 0.484375, + "learning_rate": 0.00014006282197565677, + "loss": 2.4965, + "step": 17372 + }, + { + "epoch": 0.74, + "grad_norm": 0.427734375, + "learning_rate": 0.00014005667580561143, + "loss": 2.5284, + "step": 17373 + }, + { + "epoch": 0.74, + "grad_norm": 0.408203125, + "learning_rate": 0.00014005052945532574, + "loss": 2.5647, + "step": 17374 + }, + { + "epoch": 0.74, + "grad_norm": 0.416015625, + "learning_rate": 0.0001400443829248273, + "loss": 2.4559, + "step": 17375 + }, + { + "epoch": 0.74, + "grad_norm": 0.5078125, + "learning_rate": 0.00014003823621414375, + "loss": 2.6843, + "step": 17376 + }, + { + "epoch": 0.74, + "grad_norm": 0.400390625, + "learning_rate": 0.00014003208932330277, + "loss": 2.458, + "step": 17377 + }, + { + "epoch": 0.74, + "grad_norm": 0.435546875, + "learning_rate": 0.000140025942252332, + "loss": 2.4703, + "step": 17378 + }, + { + "epoch": 0.74, + "grad_norm": 0.4609375, + "learning_rate": 0.00014001979500125915, + "loss": 2.5993, + "step": 17379 + }, + { + "epoch": 0.74, + "grad_norm": 0.4765625, + "learning_rate": 0.00014001364757011182, + "loss": 2.598, + "step": 17380 + }, + { + "epoch": 0.74, + "grad_norm": 0.50390625, + "learning_rate": 0.0001400074999589177, + "loss": 2.4261, + "step": 17381 + }, + { + "epoch": 0.74, + "grad_norm": 0.515625, + "learning_rate": 0.00014000135216770448, + "loss": 2.7579, + "step": 17382 + }, + { + "epoch": 0.74, + "grad_norm": 0.416015625, + "learning_rate": 0.00013999520419649976, + "loss": 2.4519, + "step": 17383 + }, + { + "epoch": 0.74, + "grad_norm": 0.431640625, + "learning_rate": 0.00013998905604533128, + "loss": 2.5536, + "step": 17384 + }, + { + "epoch": 0.74, + "grad_norm": 0.431640625, + "learning_rate": 0.00013998290771422665, + "loss": 2.6109, + "step": 17385 + }, + { + "epoch": 0.74, + "grad_norm": 0.4921875, + "learning_rate": 0.00013997675920321352, + "loss": 2.6142, + "step": 17386 + }, + { + "epoch": 0.74, + "grad_norm": 0.443359375, + "learning_rate": 0.0001399706105123196, + "loss": 2.5462, + "step": 17387 + }, + { + "epoch": 0.74, + "grad_norm": 0.455078125, + "learning_rate": 0.00013996446164157253, + "loss": 2.6163, + "step": 17388 + }, + { + "epoch": 0.74, + "grad_norm": 0.451171875, + "learning_rate": 0.00013995831259100002, + "loss": 2.4136, + "step": 17389 + }, + { + "epoch": 0.74, + "grad_norm": 0.51171875, + "learning_rate": 0.00013995216336062967, + "loss": 2.6658, + "step": 17390 + }, + { + "epoch": 0.74, + "grad_norm": 0.4921875, + "learning_rate": 0.0001399460139504892, + "loss": 2.5707, + "step": 17391 + }, + { + "epoch": 0.74, + "grad_norm": 0.49609375, + "learning_rate": 0.00013993986436060626, + "loss": 2.5934, + "step": 17392 + }, + { + "epoch": 0.74, + "grad_norm": 0.451171875, + "learning_rate": 0.00013993371459100853, + "loss": 2.6238, + "step": 17393 + }, + { + "epoch": 0.74, + "grad_norm": 0.419921875, + "learning_rate": 0.00013992756464172368, + "loss": 2.5964, + "step": 17394 + }, + { + "epoch": 0.74, + "grad_norm": 0.47265625, + "learning_rate": 0.0001399214145127794, + "loss": 2.6284, + "step": 17395 + }, + { + "epoch": 0.74, + "grad_norm": 0.470703125, + "learning_rate": 0.0001399152642042033, + "loss": 2.6872, + "step": 17396 + }, + { + "epoch": 0.74, + "grad_norm": 0.478515625, + "learning_rate": 0.00013990911371602313, + "loss": 2.4859, + "step": 17397 + }, + { + "epoch": 0.74, + "grad_norm": 0.3984375, + "learning_rate": 0.00013990296304826653, + "loss": 2.6084, + "step": 17398 + }, + { + "epoch": 0.74, + "grad_norm": 0.486328125, + "learning_rate": 0.00013989681220096117, + "loss": 2.5761, + "step": 17399 + }, + { + "epoch": 0.74, + "grad_norm": 0.390625, + "learning_rate": 0.00013989066117413477, + "loss": 2.5609, + "step": 17400 + }, + { + "epoch": 0.74, + "grad_norm": 0.453125, + "learning_rate": 0.00013988450996781494, + "loss": 2.6862, + "step": 17401 + }, + { + "epoch": 0.74, + "grad_norm": 0.4921875, + "learning_rate": 0.00013987835858202938, + "loss": 2.5655, + "step": 17402 + }, + { + "epoch": 0.74, + "grad_norm": 0.41796875, + "learning_rate": 0.0001398722070168058, + "loss": 2.5916, + "step": 17403 + }, + { + "epoch": 0.74, + "grad_norm": 0.458984375, + "learning_rate": 0.00013986605527217184, + "loss": 2.5062, + "step": 17404 + }, + { + "epoch": 0.74, + "grad_norm": 0.48046875, + "learning_rate": 0.00013985990334815525, + "loss": 2.5348, + "step": 17405 + }, + { + "epoch": 0.74, + "grad_norm": 0.4296875, + "learning_rate": 0.00013985375124478362, + "loss": 2.6637, + "step": 17406 + }, + { + "epoch": 0.74, + "grad_norm": 0.474609375, + "learning_rate": 0.00013984759896208467, + "loss": 2.6695, + "step": 17407 + }, + { + "epoch": 0.74, + "grad_norm": 0.451171875, + "learning_rate": 0.00013984144650008612, + "loss": 2.4841, + "step": 17408 + }, + { + "epoch": 0.74, + "grad_norm": 0.47265625, + "learning_rate": 0.0001398352938588156, + "loss": 2.482, + "step": 17409 + }, + { + "epoch": 0.74, + "grad_norm": 0.453125, + "learning_rate": 0.0001398291410383008, + "loss": 2.5558, + "step": 17410 + }, + { + "epoch": 0.74, + "grad_norm": 0.48828125, + "learning_rate": 0.00013982298803856948, + "loss": 2.669, + "step": 17411 + }, + { + "epoch": 0.74, + "grad_norm": 0.40234375, + "learning_rate": 0.0001398168348596492, + "loss": 2.4324, + "step": 17412 + }, + { + "epoch": 0.74, + "grad_norm": 0.474609375, + "learning_rate": 0.00013981068150156775, + "loss": 2.4493, + "step": 17413 + }, + { + "epoch": 0.74, + "grad_norm": 0.462890625, + "learning_rate": 0.00013980452796435278, + "loss": 2.4125, + "step": 17414 + }, + { + "epoch": 0.74, + "grad_norm": 0.416015625, + "learning_rate": 0.00013979837424803198, + "loss": 2.5841, + "step": 17415 + }, + { + "epoch": 0.74, + "grad_norm": 0.439453125, + "learning_rate": 0.00013979222035263302, + "loss": 2.4631, + "step": 17416 + }, + { + "epoch": 0.74, + "grad_norm": 0.44921875, + "learning_rate": 0.00013978606627818363, + "loss": 2.5456, + "step": 17417 + }, + { + "epoch": 0.74, + "grad_norm": 0.46484375, + "learning_rate": 0.00013977991202471145, + "loss": 2.5483, + "step": 17418 + }, + { + "epoch": 0.74, + "grad_norm": 0.439453125, + "learning_rate": 0.00013977375759224427, + "loss": 2.4384, + "step": 17419 + }, + { + "epoch": 0.74, + "grad_norm": 0.431640625, + "learning_rate": 0.00013976760298080966, + "loss": 2.7159, + "step": 17420 + }, + { + "epoch": 0.74, + "grad_norm": 0.4609375, + "learning_rate": 0.00013976144819043539, + "loss": 2.564, + "step": 17421 + }, + { + "epoch": 0.74, + "grad_norm": 0.51171875, + "learning_rate": 0.00013975529322114914, + "loss": 2.6054, + "step": 17422 + }, + { + "epoch": 0.74, + "grad_norm": 0.462890625, + "learning_rate": 0.0001397491380729786, + "loss": 2.4403, + "step": 17423 + }, + { + "epoch": 0.74, + "grad_norm": 0.5, + "learning_rate": 0.00013974298274595145, + "loss": 2.5032, + "step": 17424 + }, + { + "epoch": 0.74, + "grad_norm": 0.486328125, + "learning_rate": 0.0001397368272400954, + "loss": 2.7076, + "step": 17425 + }, + { + "epoch": 0.74, + "grad_norm": 0.4609375, + "learning_rate": 0.00013973067155543815, + "loss": 2.5682, + "step": 17426 + }, + { + "epoch": 0.74, + "grad_norm": 0.4609375, + "learning_rate": 0.00013972451569200744, + "loss": 2.6553, + "step": 17427 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.00013971835964983088, + "loss": 2.5679, + "step": 17428 + }, + { + "epoch": 0.74, + "grad_norm": 0.435546875, + "learning_rate": 0.00013971220342893622, + "loss": 2.5427, + "step": 17429 + }, + { + "epoch": 0.74, + "grad_norm": 0.44921875, + "learning_rate": 0.00013970604702935116, + "loss": 2.7631, + "step": 17430 + }, + { + "epoch": 0.74, + "grad_norm": 0.453125, + "learning_rate": 0.00013969989045110344, + "loss": 2.7188, + "step": 17431 + }, + { + "epoch": 0.74, + "grad_norm": 0.416015625, + "learning_rate": 0.00013969373369422067, + "loss": 2.5885, + "step": 17432 + }, + { + "epoch": 0.74, + "grad_norm": 0.453125, + "learning_rate": 0.00013968757675873064, + "loss": 2.5214, + "step": 17433 + }, + { + "epoch": 0.74, + "grad_norm": 0.44921875, + "learning_rate": 0.00013968141964466096, + "loss": 2.5107, + "step": 17434 + }, + { + "epoch": 0.74, + "grad_norm": 0.42578125, + "learning_rate": 0.00013967526235203947, + "loss": 2.5689, + "step": 17435 + }, + { + "epoch": 0.74, + "grad_norm": 0.474609375, + "learning_rate": 0.00013966910488089374, + "loss": 2.6106, + "step": 17436 + }, + { + "epoch": 0.74, + "grad_norm": 0.490234375, + "learning_rate": 0.00013966294723125158, + "loss": 2.543, + "step": 17437 + }, + { + "epoch": 0.74, + "grad_norm": 0.416015625, + "learning_rate": 0.00013965678940314064, + "loss": 2.6047, + "step": 17438 + }, + { + "epoch": 0.74, + "grad_norm": 0.421875, + "learning_rate": 0.00013965063139658864, + "loss": 2.4535, + "step": 17439 + }, + { + "epoch": 0.74, + "grad_norm": 0.48828125, + "learning_rate": 0.00013964447321162328, + "loss": 2.6723, + "step": 17440 + }, + { + "epoch": 0.74, + "grad_norm": 0.5859375, + "learning_rate": 0.00013963831484827227, + "loss": 2.5731, + "step": 17441 + }, + { + "epoch": 0.74, + "grad_norm": 0.4296875, + "learning_rate": 0.00013963215630656334, + "loss": 2.565, + "step": 17442 + }, + { + "epoch": 0.74, + "grad_norm": 0.453125, + "learning_rate": 0.0001396259975865242, + "loss": 2.5093, + "step": 17443 + }, + { + "epoch": 0.74, + "grad_norm": 0.5, + "learning_rate": 0.00013961983868818255, + "loss": 2.7076, + "step": 17444 + }, + { + "epoch": 0.74, + "grad_norm": 0.470703125, + "learning_rate": 0.00013961367961156607, + "loss": 2.6154, + "step": 17445 + }, + { + "epoch": 0.74, + "grad_norm": 0.4375, + "learning_rate": 0.00013960752035670256, + "loss": 2.568, + "step": 17446 + }, + { + "epoch": 0.74, + "grad_norm": 0.515625, + "learning_rate": 0.00013960136092361965, + "loss": 2.6219, + "step": 17447 + }, + { + "epoch": 0.74, + "grad_norm": 0.48828125, + "learning_rate": 0.00013959520131234514, + "loss": 2.443, + "step": 17448 + }, + { + "epoch": 0.74, + "grad_norm": 0.44140625, + "learning_rate": 0.00013958904152290664, + "loss": 2.5904, + "step": 17449 + }, + { + "epoch": 0.74, + "grad_norm": 0.48828125, + "learning_rate": 0.00013958288155533193, + "loss": 2.7456, + "step": 17450 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.00013957672140964874, + "loss": 2.631, + "step": 17451 + }, + { + "epoch": 0.74, + "grad_norm": 0.486328125, + "learning_rate": 0.00013957056108588475, + "loss": 2.5663, + "step": 17452 + }, + { + "epoch": 0.74, + "grad_norm": 0.43359375, + "learning_rate": 0.0001395644005840677, + "loss": 2.5811, + "step": 17453 + }, + { + "epoch": 0.74, + "grad_norm": 0.5234375, + "learning_rate": 0.00013955823990422533, + "loss": 2.5594, + "step": 17454 + }, + { + "epoch": 0.74, + "grad_norm": 0.458984375, + "learning_rate": 0.0001395520790463853, + "loss": 2.6882, + "step": 17455 + }, + { + "epoch": 0.74, + "grad_norm": 0.47265625, + "learning_rate": 0.00013954591801057536, + "loss": 2.6224, + "step": 17456 + }, + { + "epoch": 0.74, + "grad_norm": 0.443359375, + "learning_rate": 0.00013953975679682328, + "loss": 2.5804, + "step": 17457 + }, + { + "epoch": 0.74, + "grad_norm": 0.435546875, + "learning_rate": 0.00013953359540515674, + "loss": 2.4776, + "step": 17458 + }, + { + "epoch": 0.74, + "grad_norm": 0.484375, + "learning_rate": 0.00013952743383560345, + "loss": 2.6144, + "step": 17459 + }, + { + "epoch": 0.74, + "grad_norm": 0.498046875, + "learning_rate": 0.00013952127208819115, + "loss": 2.7415, + "step": 17460 + }, + { + "epoch": 0.74, + "grad_norm": 0.427734375, + "learning_rate": 0.00013951511016294758, + "loss": 2.6409, + "step": 17461 + }, + { + "epoch": 0.74, + "grad_norm": 0.4453125, + "learning_rate": 0.00013950894805990046, + "loss": 2.6418, + "step": 17462 + }, + { + "epoch": 0.74, + "grad_norm": 0.439453125, + "learning_rate": 0.0001395027857790775, + "loss": 2.4535, + "step": 17463 + }, + { + "epoch": 0.74, + "grad_norm": 0.46484375, + "learning_rate": 0.00013949662332050642, + "loss": 2.6474, + "step": 17464 + }, + { + "epoch": 0.74, + "grad_norm": 0.46484375, + "learning_rate": 0.00013949046068421504, + "loss": 2.7112, + "step": 17465 + }, + { + "epoch": 0.74, + "grad_norm": 0.47265625, + "learning_rate": 0.00013948429787023097, + "loss": 2.6445, + "step": 17466 + }, + { + "epoch": 0.74, + "grad_norm": 0.48046875, + "learning_rate": 0.000139478134878582, + "loss": 2.69, + "step": 17467 + }, + { + "epoch": 0.74, + "grad_norm": 0.466796875, + "learning_rate": 0.00013947197170929585, + "loss": 2.7307, + "step": 17468 + }, + { + "epoch": 0.74, + "grad_norm": 0.51953125, + "learning_rate": 0.00013946580836240023, + "loss": 2.6658, + "step": 17469 + }, + { + "epoch": 0.74, + "grad_norm": 0.44921875, + "learning_rate": 0.00013945964483792292, + "loss": 2.5321, + "step": 17470 + }, + { + "epoch": 0.74, + "grad_norm": 0.421875, + "learning_rate": 0.00013945348113589162, + "loss": 2.5245, + "step": 17471 + }, + { + "epoch": 0.74, + "grad_norm": 0.458984375, + "learning_rate": 0.00013944731725633406, + "loss": 2.5742, + "step": 17472 + }, + { + "epoch": 0.74, + "grad_norm": 0.5, + "learning_rate": 0.00013944115319927802, + "loss": 2.6373, + "step": 17473 + }, + { + "epoch": 0.74, + "grad_norm": 0.431640625, + "learning_rate": 0.0001394349889647512, + "loss": 2.643, + "step": 17474 + }, + { + "epoch": 0.74, + "grad_norm": 0.46875, + "learning_rate": 0.0001394288245527813, + "loss": 2.4945, + "step": 17475 + }, + { + "epoch": 0.74, + "grad_norm": 0.443359375, + "learning_rate": 0.00013942265996339616, + "loss": 2.5342, + "step": 17476 + }, + { + "epoch": 0.74, + "grad_norm": 0.486328125, + "learning_rate": 0.0001394164951966234, + "loss": 2.5827, + "step": 17477 + }, + { + "epoch": 0.74, + "grad_norm": 0.51171875, + "learning_rate": 0.00013941033025249088, + "loss": 2.5943, + "step": 17478 + }, + { + "epoch": 0.74, + "grad_norm": 0.435546875, + "learning_rate": 0.00013940416513102624, + "loss": 2.5419, + "step": 17479 + }, + { + "epoch": 0.74, + "grad_norm": 0.46875, + "learning_rate": 0.00013939799983225725, + "loss": 2.4091, + "step": 17480 + }, + { + "epoch": 0.74, + "grad_norm": 0.4140625, + "learning_rate": 0.0001393918343562117, + "loss": 2.5276, + "step": 17481 + }, + { + "epoch": 0.74, + "grad_norm": 0.443359375, + "learning_rate": 0.00013938566870291724, + "loss": 2.565, + "step": 17482 + }, + { + "epoch": 0.74, + "grad_norm": 0.455078125, + "learning_rate": 0.0001393795028724017, + "loss": 2.5426, + "step": 17483 + }, + { + "epoch": 0.75, + "grad_norm": 0.5234375, + "learning_rate": 0.00013937333686469275, + "loss": 2.6034, + "step": 17484 + }, + { + "epoch": 0.75, + "grad_norm": 0.42578125, + "learning_rate": 0.00013936717067981817, + "loss": 2.5261, + "step": 17485 + }, + { + "epoch": 0.75, + "grad_norm": 0.484375, + "learning_rate": 0.00013936100431780575, + "loss": 2.5052, + "step": 17486 + }, + { + "epoch": 0.75, + "grad_norm": 0.51953125, + "learning_rate": 0.00013935483777868319, + "loss": 2.5591, + "step": 17487 + }, + { + "epoch": 0.75, + "grad_norm": 0.498046875, + "learning_rate": 0.00013934867106247823, + "loss": 2.6197, + "step": 17488 + }, + { + "epoch": 0.75, + "grad_norm": 0.490234375, + "learning_rate": 0.00013934250416921864, + "loss": 2.5269, + "step": 17489 + }, + { + "epoch": 0.75, + "grad_norm": 0.462890625, + "learning_rate": 0.00013933633709893215, + "loss": 2.5854, + "step": 17490 + }, + { + "epoch": 0.75, + "grad_norm": 0.4296875, + "learning_rate": 0.00013933016985164654, + "loss": 2.6104, + "step": 17491 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.0001393240024273895, + "loss": 2.445, + "step": 17492 + }, + { + "epoch": 0.75, + "grad_norm": 0.53125, + "learning_rate": 0.00013931783482618883, + "loss": 2.5928, + "step": 17493 + }, + { + "epoch": 0.75, + "grad_norm": 0.46484375, + "learning_rate": 0.0001393116670480723, + "loss": 2.6221, + "step": 17494 + }, + { + "epoch": 0.75, + "grad_norm": 0.478515625, + "learning_rate": 0.0001393054990930676, + "loss": 2.5992, + "step": 17495 + }, + { + "epoch": 0.75, + "grad_norm": 0.404296875, + "learning_rate": 0.00013929933096120254, + "loss": 2.5887, + "step": 17496 + }, + { + "epoch": 0.75, + "grad_norm": 0.494140625, + "learning_rate": 0.00013929316265250488, + "loss": 2.6237, + "step": 17497 + }, + { + "epoch": 0.75, + "grad_norm": 0.49609375, + "learning_rate": 0.0001392869941670023, + "loss": 2.551, + "step": 17498 + }, + { + "epoch": 0.75, + "grad_norm": 0.4609375, + "learning_rate": 0.00013928082550472264, + "loss": 2.6074, + "step": 17499 + }, + { + "epoch": 0.75, + "grad_norm": 0.400390625, + "learning_rate": 0.0001392746566656936, + "loss": 2.522, + "step": 17500 + }, + { + "epoch": 0.75, + "grad_norm": 0.466796875, + "learning_rate": 0.00013926848764994292, + "loss": 2.5827, + "step": 17501 + }, + { + "epoch": 0.75, + "grad_norm": 0.5234375, + "learning_rate": 0.00013926231845749849, + "loss": 2.5175, + "step": 17502 + }, + { + "epoch": 0.75, + "grad_norm": 0.423828125, + "learning_rate": 0.0001392561490883879, + "loss": 2.4811, + "step": 17503 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.000139249979542639, + "loss": 2.5881, + "step": 17504 + }, + { + "epoch": 0.75, + "grad_norm": 0.474609375, + "learning_rate": 0.00013924380982027957, + "loss": 2.5893, + "step": 17505 + }, + { + "epoch": 0.75, + "grad_norm": 0.50390625, + "learning_rate": 0.0001392376399213373, + "loss": 2.7496, + "step": 17506 + }, + { + "epoch": 0.75, + "grad_norm": 0.486328125, + "learning_rate": 0.00013923146984584002, + "loss": 2.491, + "step": 17507 + }, + { + "epoch": 0.75, + "grad_norm": 0.4375, + "learning_rate": 0.00013922529959381542, + "loss": 2.5756, + "step": 17508 + }, + { + "epoch": 0.75, + "grad_norm": 0.4296875, + "learning_rate": 0.00013921912916529133, + "loss": 2.442, + "step": 17509 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.00013921295856029553, + "loss": 2.6518, + "step": 17510 + }, + { + "epoch": 0.75, + "grad_norm": 0.43359375, + "learning_rate": 0.0001392067877788557, + "loss": 2.4657, + "step": 17511 + }, + { + "epoch": 0.75, + "grad_norm": 0.5234375, + "learning_rate": 0.00013920061682099965, + "loss": 2.5709, + "step": 17512 + }, + { + "epoch": 0.75, + "grad_norm": 0.482421875, + "learning_rate": 0.00013919444568675517, + "loss": 2.5967, + "step": 17513 + }, + { + "epoch": 0.75, + "grad_norm": 0.4765625, + "learning_rate": 0.00013918827437615, + "loss": 2.6684, + "step": 17514 + }, + { + "epoch": 0.75, + "grad_norm": 0.41015625, + "learning_rate": 0.00013918210288921194, + "loss": 2.4282, + "step": 17515 + }, + { + "epoch": 0.75, + "grad_norm": 0.447265625, + "learning_rate": 0.00013917593122596871, + "loss": 2.4941, + "step": 17516 + }, + { + "epoch": 0.75, + "grad_norm": 0.455078125, + "learning_rate": 0.0001391697593864481, + "loss": 2.6433, + "step": 17517 + }, + { + "epoch": 0.75, + "grad_norm": 0.453125, + "learning_rate": 0.00013916358737067793, + "loss": 2.623, + "step": 17518 + }, + { + "epoch": 0.75, + "grad_norm": 0.51171875, + "learning_rate": 0.0001391574151786859, + "loss": 2.6354, + "step": 17519 + }, + { + "epoch": 0.75, + "grad_norm": 0.466796875, + "learning_rate": 0.0001391512428104998, + "loss": 2.6701, + "step": 17520 + }, + { + "epoch": 0.75, + "grad_norm": 0.427734375, + "learning_rate": 0.00013914507026614742, + "loss": 2.5415, + "step": 17521 + }, + { + "epoch": 0.75, + "grad_norm": 0.5234375, + "learning_rate": 0.00013913889754565655, + "loss": 2.6329, + "step": 17522 + }, + { + "epoch": 0.75, + "grad_norm": 0.478515625, + "learning_rate": 0.00013913272464905494, + "loss": 2.6732, + "step": 17523 + }, + { + "epoch": 0.75, + "grad_norm": 0.41796875, + "learning_rate": 0.00013912655157637038, + "loss": 2.4335, + "step": 17524 + }, + { + "epoch": 0.75, + "grad_norm": 0.43359375, + "learning_rate": 0.0001391203783276306, + "loss": 2.5835, + "step": 17525 + }, + { + "epoch": 0.75, + "grad_norm": 0.462890625, + "learning_rate": 0.00013911420490286344, + "loss": 2.5378, + "step": 17526 + }, + { + "epoch": 0.75, + "grad_norm": 0.494140625, + "learning_rate": 0.00013910803130209664, + "loss": 2.4972, + "step": 17527 + }, + { + "epoch": 0.75, + "grad_norm": 0.53125, + "learning_rate": 0.000139101857525358, + "loss": 2.7743, + "step": 17528 + }, + { + "epoch": 0.75, + "grad_norm": 0.400390625, + "learning_rate": 0.0001390956835726753, + "loss": 2.4631, + "step": 17529 + }, + { + "epoch": 0.75, + "grad_norm": 0.47265625, + "learning_rate": 0.00013908950944407632, + "loss": 2.5139, + "step": 17530 + }, + { + "epoch": 0.75, + "grad_norm": 0.46875, + "learning_rate": 0.0001390833351395888, + "loss": 2.6043, + "step": 17531 + }, + { + "epoch": 0.75, + "grad_norm": 0.43359375, + "learning_rate": 0.00013907716065924056, + "loss": 2.344, + "step": 17532 + }, + { + "epoch": 0.75, + "grad_norm": 0.466796875, + "learning_rate": 0.0001390709860030594, + "loss": 2.5818, + "step": 17533 + }, + { + "epoch": 0.75, + "grad_norm": 0.458984375, + "learning_rate": 0.0001390648111710731, + "loss": 2.4962, + "step": 17534 + }, + { + "epoch": 0.75, + "grad_norm": 0.427734375, + "learning_rate": 0.00013905863616330935, + "loss": 2.5247, + "step": 17535 + }, + { + "epoch": 0.75, + "grad_norm": 0.435546875, + "learning_rate": 0.00013905246097979608, + "loss": 2.5198, + "step": 17536 + }, + { + "epoch": 0.75, + "grad_norm": 0.474609375, + "learning_rate": 0.00013904628562056099, + "loss": 2.5845, + "step": 17537 + }, + { + "epoch": 0.75, + "grad_norm": 0.484375, + "learning_rate": 0.00013904011008563188, + "loss": 2.5566, + "step": 17538 + }, + { + "epoch": 0.75, + "grad_norm": 0.486328125, + "learning_rate": 0.00013903393437503657, + "loss": 2.7026, + "step": 17539 + }, + { + "epoch": 0.75, + "grad_norm": 0.44140625, + "learning_rate": 0.00013902775848880275, + "loss": 2.4971, + "step": 17540 + }, + { + "epoch": 0.75, + "grad_norm": 0.44921875, + "learning_rate": 0.00013902158242695833, + "loss": 2.5728, + "step": 17541 + }, + { + "epoch": 0.75, + "grad_norm": 0.498046875, + "learning_rate": 0.00013901540618953105, + "loss": 2.6957, + "step": 17542 + }, + { + "epoch": 0.75, + "grad_norm": 0.5078125, + "learning_rate": 0.0001390092297765487, + "loss": 2.7023, + "step": 17543 + }, + { + "epoch": 0.75, + "grad_norm": 0.44921875, + "learning_rate": 0.00013900305318803905, + "loss": 2.5531, + "step": 17544 + }, + { + "epoch": 0.75, + "grad_norm": 0.44921875, + "learning_rate": 0.00013899687642402993, + "loss": 2.6778, + "step": 17545 + }, + { + "epoch": 0.75, + "grad_norm": 0.447265625, + "learning_rate": 0.00013899069948454914, + "loss": 2.6403, + "step": 17546 + }, + { + "epoch": 0.75, + "grad_norm": 0.439453125, + "learning_rate": 0.00013898452236962443, + "loss": 2.5118, + "step": 17547 + }, + { + "epoch": 0.75, + "grad_norm": 0.48046875, + "learning_rate": 0.0001389783450792836, + "loss": 2.6916, + "step": 17548 + }, + { + "epoch": 0.75, + "grad_norm": 0.451171875, + "learning_rate": 0.00013897216761355448, + "loss": 2.5123, + "step": 17549 + }, + { + "epoch": 0.75, + "grad_norm": 0.4609375, + "learning_rate": 0.00013896598997246485, + "loss": 2.6411, + "step": 17550 + }, + { + "epoch": 0.75, + "grad_norm": 0.43359375, + "learning_rate": 0.0001389598121560425, + "loss": 2.5306, + "step": 17551 + }, + { + "epoch": 0.75, + "grad_norm": 0.40234375, + "learning_rate": 0.00013895363416431523, + "loss": 2.5881, + "step": 17552 + }, + { + "epoch": 0.75, + "grad_norm": 0.50390625, + "learning_rate": 0.00013894745599731087, + "loss": 2.5517, + "step": 17553 + }, + { + "epoch": 0.75, + "grad_norm": 0.455078125, + "learning_rate": 0.00013894127765505716, + "loss": 2.5561, + "step": 17554 + }, + { + "epoch": 0.75, + "grad_norm": 0.44140625, + "learning_rate": 0.00013893509913758196, + "loss": 2.4236, + "step": 17555 + }, + { + "epoch": 0.75, + "grad_norm": 0.431640625, + "learning_rate": 0.00013892892044491302, + "loss": 2.6494, + "step": 17556 + }, + { + "epoch": 0.75, + "grad_norm": 0.44921875, + "learning_rate": 0.00013892274157707813, + "loss": 2.5677, + "step": 17557 + }, + { + "epoch": 0.75, + "grad_norm": 0.3984375, + "learning_rate": 0.0001389165625341052, + "loss": 2.6465, + "step": 17558 + }, + { + "epoch": 0.75, + "grad_norm": 0.482421875, + "learning_rate": 0.00013891038331602192, + "loss": 2.4887, + "step": 17559 + }, + { + "epoch": 0.75, + "grad_norm": 0.458984375, + "learning_rate": 0.00013890420392285614, + "loss": 2.6709, + "step": 17560 + }, + { + "epoch": 0.75, + "grad_norm": 0.470703125, + "learning_rate": 0.00013889802435463567, + "loss": 2.5463, + "step": 17561 + }, + { + "epoch": 0.75, + "grad_norm": 0.53515625, + "learning_rate": 0.0001388918446113883, + "loss": 2.6174, + "step": 17562 + }, + { + "epoch": 0.75, + "grad_norm": 0.494140625, + "learning_rate": 0.00013888566469314186, + "loss": 2.5033, + "step": 17563 + }, + { + "epoch": 0.75, + "grad_norm": 0.466796875, + "learning_rate": 0.00013887948459992414, + "loss": 2.6301, + "step": 17564 + }, + { + "epoch": 0.75, + "grad_norm": 0.435546875, + "learning_rate": 0.00013887330433176288, + "loss": 2.6236, + "step": 17565 + }, + { + "epoch": 0.75, + "grad_norm": 0.53515625, + "learning_rate": 0.00013886712388868605, + "loss": 2.6858, + "step": 17566 + }, + { + "epoch": 0.75, + "grad_norm": 0.427734375, + "learning_rate": 0.0001388609432707213, + "loss": 2.6126, + "step": 17567 + }, + { + "epoch": 0.75, + "grad_norm": 0.453125, + "learning_rate": 0.00013885476247789654, + "loss": 2.7003, + "step": 17568 + }, + { + "epoch": 0.75, + "grad_norm": 0.4375, + "learning_rate": 0.00013884858151023955, + "loss": 2.4601, + "step": 17569 + }, + { + "epoch": 0.75, + "grad_norm": 0.439453125, + "learning_rate": 0.00013884240036777813, + "loss": 2.5289, + "step": 17570 + }, + { + "epoch": 0.75, + "grad_norm": 0.462890625, + "learning_rate": 0.0001388362190505401, + "loss": 2.5543, + "step": 17571 + }, + { + "epoch": 0.75, + "grad_norm": 0.50390625, + "learning_rate": 0.00013883003755855327, + "loss": 2.6114, + "step": 17572 + }, + { + "epoch": 0.75, + "grad_norm": 0.43359375, + "learning_rate": 0.0001388238558918455, + "loss": 2.6011, + "step": 17573 + }, + { + "epoch": 0.75, + "grad_norm": 0.451171875, + "learning_rate": 0.00013881767405044454, + "loss": 2.6484, + "step": 17574 + }, + { + "epoch": 0.75, + "grad_norm": 0.5703125, + "learning_rate": 0.00013881149203437824, + "loss": 2.6882, + "step": 17575 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.00013880530984367438, + "loss": 2.4739, + "step": 17576 + }, + { + "epoch": 0.75, + "grad_norm": 0.44140625, + "learning_rate": 0.00013879912747836085, + "loss": 2.578, + "step": 17577 + }, + { + "epoch": 0.75, + "grad_norm": 0.455078125, + "learning_rate": 0.0001387929449384654, + "loss": 2.7336, + "step": 17578 + }, + { + "epoch": 0.75, + "grad_norm": 0.515625, + "learning_rate": 0.00013878676222401587, + "loss": 2.4778, + "step": 17579 + }, + { + "epoch": 0.75, + "grad_norm": 0.494140625, + "learning_rate": 0.00013878057933504011, + "loss": 2.5276, + "step": 17580 + }, + { + "epoch": 0.75, + "grad_norm": 0.47265625, + "learning_rate": 0.0001387743962715659, + "loss": 2.5671, + "step": 17581 + }, + { + "epoch": 0.75, + "grad_norm": 0.46484375, + "learning_rate": 0.0001387682130336211, + "loss": 2.6662, + "step": 17582 + }, + { + "epoch": 0.75, + "grad_norm": 0.482421875, + "learning_rate": 0.00013876202962123348, + "loss": 2.4636, + "step": 17583 + }, + { + "epoch": 0.75, + "grad_norm": 0.447265625, + "learning_rate": 0.0001387558460344309, + "loss": 2.5091, + "step": 17584 + }, + { + "epoch": 0.75, + "grad_norm": 0.453125, + "learning_rate": 0.0001387496622732412, + "loss": 2.6076, + "step": 17585 + }, + { + "epoch": 0.75, + "grad_norm": 0.412109375, + "learning_rate": 0.0001387434783376921, + "loss": 2.5427, + "step": 17586 + }, + { + "epoch": 0.75, + "grad_norm": 0.45703125, + "learning_rate": 0.0001387372942278116, + "loss": 2.5926, + "step": 17587 + }, + { + "epoch": 0.75, + "grad_norm": 0.4375, + "learning_rate": 0.0001387311099436274, + "loss": 2.5588, + "step": 17588 + }, + { + "epoch": 0.75, + "grad_norm": 0.53515625, + "learning_rate": 0.00013872492548516735, + "loss": 2.7948, + "step": 17589 + }, + { + "epoch": 0.75, + "grad_norm": 0.46875, + "learning_rate": 0.00013871874085245932, + "loss": 2.6519, + "step": 17590 + }, + { + "epoch": 0.75, + "grad_norm": 0.48046875, + "learning_rate": 0.00013871255604553108, + "loss": 2.6361, + "step": 17591 + }, + { + "epoch": 0.75, + "grad_norm": 0.51171875, + "learning_rate": 0.00013870637106441048, + "loss": 2.6139, + "step": 17592 + }, + { + "epoch": 0.75, + "grad_norm": 0.470703125, + "learning_rate": 0.00013870018590912537, + "loss": 2.5827, + "step": 17593 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.00013869400057970355, + "loss": 2.5352, + "step": 17594 + }, + { + "epoch": 0.75, + "grad_norm": 0.48828125, + "learning_rate": 0.00013868781507617286, + "loss": 2.5784, + "step": 17595 + }, + { + "epoch": 0.75, + "grad_norm": 0.44140625, + "learning_rate": 0.0001386816293985612, + "loss": 2.491, + "step": 17596 + }, + { + "epoch": 0.75, + "grad_norm": 0.478515625, + "learning_rate": 0.0001386754435468963, + "loss": 2.7932, + "step": 17597 + }, + { + "epoch": 0.75, + "grad_norm": 0.50390625, + "learning_rate": 0.00013866925752120603, + "loss": 2.6717, + "step": 17598 + }, + { + "epoch": 0.75, + "grad_norm": 0.4609375, + "learning_rate": 0.00013866307132151827, + "loss": 2.6469, + "step": 17599 + }, + { + "epoch": 0.75, + "grad_norm": 0.53125, + "learning_rate": 0.00013865688494786075, + "loss": 2.6251, + "step": 17600 + }, + { + "epoch": 0.75, + "grad_norm": 0.447265625, + "learning_rate": 0.00013865069840026146, + "loss": 2.6134, + "step": 17601 + }, + { + "epoch": 0.75, + "eval_loss": 2.554856061935425, + "eval_runtime": 82.0988, + "eval_samples_per_second": 18.319, + "eval_steps_per_second": 2.29, + "step": 17601 + }, + { + "epoch": 0.75, + "grad_norm": 0.470703125, + "learning_rate": 0.0001386445116787481, + "loss": 2.5945, + "step": 17602 + }, + { + "epoch": 0.75, + "grad_norm": 0.484375, + "learning_rate": 0.00013863832478334858, + "loss": 2.6658, + "step": 17603 + }, + { + "epoch": 0.75, + "grad_norm": 0.412109375, + "learning_rate": 0.0001386321377140907, + "loss": 2.5969, + "step": 17604 + }, + { + "epoch": 0.75, + "grad_norm": 0.447265625, + "learning_rate": 0.00013862595047100233, + "loss": 2.4698, + "step": 17605 + }, + { + "epoch": 0.75, + "grad_norm": 0.48046875, + "learning_rate": 0.0001386197630541113, + "loss": 2.5194, + "step": 17606 + }, + { + "epoch": 0.75, + "grad_norm": 0.4453125, + "learning_rate": 0.00013861357546344547, + "loss": 2.6841, + "step": 17607 + }, + { + "epoch": 0.75, + "grad_norm": 0.48046875, + "learning_rate": 0.0001386073876990326, + "loss": 2.6194, + "step": 17608 + }, + { + "epoch": 0.75, + "grad_norm": 0.49609375, + "learning_rate": 0.00013860119976090066, + "loss": 2.6846, + "step": 17609 + }, + { + "epoch": 0.75, + "grad_norm": 0.44921875, + "learning_rate": 0.0001385950116490774, + "loss": 2.5381, + "step": 17610 + }, + { + "epoch": 0.75, + "grad_norm": 0.44921875, + "learning_rate": 0.00013858882336359072, + "loss": 2.5322, + "step": 17611 + }, + { + "epoch": 0.75, + "grad_norm": 0.419921875, + "learning_rate": 0.0001385826349044684, + "loss": 2.5563, + "step": 17612 + }, + { + "epoch": 0.75, + "grad_norm": 0.42578125, + "learning_rate": 0.00013857644627173832, + "loss": 2.5594, + "step": 17613 + }, + { + "epoch": 0.75, + "grad_norm": 0.490234375, + "learning_rate": 0.00013857025746542837, + "loss": 2.6984, + "step": 17614 + }, + { + "epoch": 0.75, + "grad_norm": 0.439453125, + "learning_rate": 0.00013856406848556632, + "loss": 2.6716, + "step": 17615 + }, + { + "epoch": 0.75, + "grad_norm": 0.455078125, + "learning_rate": 0.00013855787933218008, + "loss": 2.7275, + "step": 17616 + }, + { + "epoch": 0.75, + "grad_norm": 0.408203125, + "learning_rate": 0.0001385516900052975, + "loss": 2.5395, + "step": 17617 + }, + { + "epoch": 0.75, + "grad_norm": 0.47265625, + "learning_rate": 0.00013854550050494634, + "loss": 2.679, + "step": 17618 + }, + { + "epoch": 0.75, + "grad_norm": 0.5, + "learning_rate": 0.00013853931083115454, + "loss": 2.6661, + "step": 17619 + }, + { + "epoch": 0.75, + "grad_norm": 0.46484375, + "learning_rate": 0.00013853312098394992, + "loss": 2.5868, + "step": 17620 + }, + { + "epoch": 0.75, + "grad_norm": 0.390625, + "learning_rate": 0.00013852693096336036, + "loss": 2.6502, + "step": 17621 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.0001385207407694137, + "loss": 2.495, + "step": 17622 + }, + { + "epoch": 0.75, + "grad_norm": 0.419921875, + "learning_rate": 0.00013851455040213775, + "loss": 2.5959, + "step": 17623 + }, + { + "epoch": 0.75, + "grad_norm": 0.458984375, + "learning_rate": 0.0001385083598615604, + "loss": 2.4454, + "step": 17624 + }, + { + "epoch": 0.75, + "grad_norm": 0.462890625, + "learning_rate": 0.00013850216914770952, + "loss": 2.6943, + "step": 17625 + }, + { + "epoch": 0.75, + "grad_norm": 0.5, + "learning_rate": 0.00013849597826061293, + "loss": 2.5717, + "step": 17626 + }, + { + "epoch": 0.75, + "grad_norm": 0.498046875, + "learning_rate": 0.00013848978720029854, + "loss": 2.586, + "step": 17627 + }, + { + "epoch": 0.75, + "grad_norm": 0.3984375, + "learning_rate": 0.00013848359596679415, + "loss": 2.5136, + "step": 17628 + }, + { + "epoch": 0.75, + "grad_norm": 0.4453125, + "learning_rate": 0.00013847740456012764, + "loss": 2.4221, + "step": 17629 + }, + { + "epoch": 0.75, + "grad_norm": 0.490234375, + "learning_rate": 0.00013847121298032688, + "loss": 2.5027, + "step": 17630 + }, + { + "epoch": 0.75, + "grad_norm": 0.490234375, + "learning_rate": 0.00013846502122741972, + "loss": 2.6635, + "step": 17631 + }, + { + "epoch": 0.75, + "grad_norm": 0.388671875, + "learning_rate": 0.00013845882930143402, + "loss": 2.5736, + "step": 17632 + }, + { + "epoch": 0.75, + "grad_norm": 0.43359375, + "learning_rate": 0.00013845263720239765, + "loss": 2.6256, + "step": 17633 + }, + { + "epoch": 0.75, + "grad_norm": 0.44140625, + "learning_rate": 0.00013844644493033844, + "loss": 2.5121, + "step": 17634 + }, + { + "epoch": 0.75, + "grad_norm": 0.41015625, + "learning_rate": 0.00013844025248528432, + "loss": 2.5774, + "step": 17635 + }, + { + "epoch": 0.75, + "grad_norm": 0.45703125, + "learning_rate": 0.00013843405986726308, + "loss": 2.6571, + "step": 17636 + }, + { + "epoch": 0.75, + "grad_norm": 0.439453125, + "learning_rate": 0.00013842786707630262, + "loss": 2.6164, + "step": 17637 + }, + { + "epoch": 0.75, + "grad_norm": 0.4375, + "learning_rate": 0.00013842167411243082, + "loss": 2.589, + "step": 17638 + }, + { + "epoch": 0.75, + "grad_norm": 0.455078125, + "learning_rate": 0.00013841548097567552, + "loss": 2.486, + "step": 17639 + }, + { + "epoch": 0.75, + "grad_norm": 0.455078125, + "learning_rate": 0.0001384092876660646, + "loss": 2.6087, + "step": 17640 + }, + { + "epoch": 0.75, + "grad_norm": 0.453125, + "learning_rate": 0.0001384030941836259, + "loss": 2.6382, + "step": 17641 + }, + { + "epoch": 0.75, + "grad_norm": 0.4140625, + "learning_rate": 0.00013839690052838731, + "loss": 2.5791, + "step": 17642 + }, + { + "epoch": 0.75, + "grad_norm": 0.48046875, + "learning_rate": 0.00013839070670037672, + "loss": 2.6663, + "step": 17643 + }, + { + "epoch": 0.75, + "grad_norm": 0.5078125, + "learning_rate": 0.00013838451269962198, + "loss": 2.4488, + "step": 17644 + }, + { + "epoch": 0.75, + "grad_norm": 0.423828125, + "learning_rate": 0.00013837831852615097, + "loss": 2.5465, + "step": 17645 + }, + { + "epoch": 0.75, + "grad_norm": 0.474609375, + "learning_rate": 0.00013837212417999154, + "loss": 2.5548, + "step": 17646 + }, + { + "epoch": 0.75, + "grad_norm": 0.51171875, + "learning_rate": 0.00013836592966117158, + "loss": 2.7318, + "step": 17647 + }, + { + "epoch": 0.75, + "grad_norm": 0.44140625, + "learning_rate": 0.00013835973496971894, + "loss": 2.4773, + "step": 17648 + }, + { + "epoch": 0.75, + "grad_norm": 0.427734375, + "learning_rate": 0.00013835354010566152, + "loss": 2.5865, + "step": 17649 + }, + { + "epoch": 0.75, + "grad_norm": 0.41015625, + "learning_rate": 0.00013834734506902718, + "loss": 2.3931, + "step": 17650 + }, + { + "epoch": 0.75, + "grad_norm": 0.494140625, + "learning_rate": 0.00013834114985984382, + "loss": 2.6214, + "step": 17651 + }, + { + "epoch": 0.75, + "grad_norm": 0.455078125, + "learning_rate": 0.0001383349544781393, + "loss": 2.4686, + "step": 17652 + }, + { + "epoch": 0.75, + "grad_norm": 0.447265625, + "learning_rate": 0.0001383287589239415, + "loss": 2.52, + "step": 17653 + }, + { + "epoch": 0.75, + "grad_norm": 0.52734375, + "learning_rate": 0.00013832256319727825, + "loss": 2.7626, + "step": 17654 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.0001383163672981775, + "loss": 2.5821, + "step": 17655 + }, + { + "epoch": 0.75, + "grad_norm": 0.423828125, + "learning_rate": 0.00013831017122666706, + "loss": 2.5117, + "step": 17656 + }, + { + "epoch": 0.75, + "grad_norm": 0.4921875, + "learning_rate": 0.00013830397498277493, + "loss": 2.5866, + "step": 17657 + }, + { + "epoch": 0.75, + "grad_norm": 0.478515625, + "learning_rate": 0.00013829777856652883, + "loss": 2.6115, + "step": 17658 + }, + { + "epoch": 0.75, + "grad_norm": 0.498046875, + "learning_rate": 0.0001382915819779568, + "loss": 2.6667, + "step": 17659 + }, + { + "epoch": 0.75, + "grad_norm": 0.43359375, + "learning_rate": 0.0001382853852170866, + "loss": 2.6581, + "step": 17660 + }, + { + "epoch": 0.75, + "grad_norm": 0.490234375, + "learning_rate": 0.00013827918828394616, + "loss": 2.6148, + "step": 17661 + }, + { + "epoch": 0.75, + "grad_norm": 0.478515625, + "learning_rate": 0.00013827299117856337, + "loss": 2.5798, + "step": 17662 + }, + { + "epoch": 0.75, + "grad_norm": 0.388671875, + "learning_rate": 0.00013826679390096608, + "loss": 2.6165, + "step": 17663 + }, + { + "epoch": 0.75, + "grad_norm": 0.44140625, + "learning_rate": 0.00013826059645118222, + "loss": 2.5863, + "step": 17664 + }, + { + "epoch": 0.75, + "grad_norm": 0.447265625, + "learning_rate": 0.0001382543988292397, + "loss": 2.6297, + "step": 17665 + }, + { + "epoch": 0.75, + "grad_norm": 0.458984375, + "learning_rate": 0.0001382482010351663, + "loss": 2.5022, + "step": 17666 + }, + { + "epoch": 0.75, + "grad_norm": 0.490234375, + "learning_rate": 0.00013824200306899, + "loss": 2.717, + "step": 17667 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.00013823580493073866, + "loss": 2.6548, + "step": 17668 + }, + { + "epoch": 0.75, + "grad_norm": 0.44140625, + "learning_rate": 0.00013822960662044018, + "loss": 2.4955, + "step": 17669 + }, + { + "epoch": 0.75, + "grad_norm": 0.486328125, + "learning_rate": 0.00013822340813812245, + "loss": 2.618, + "step": 17670 + }, + { + "epoch": 0.75, + "grad_norm": 0.458984375, + "learning_rate": 0.00013821720948381335, + "loss": 2.4289, + "step": 17671 + }, + { + "epoch": 0.75, + "grad_norm": 0.4921875, + "learning_rate": 0.00013821101065754075, + "loss": 2.6799, + "step": 17672 + }, + { + "epoch": 0.75, + "grad_norm": 0.4609375, + "learning_rate": 0.00013820481165933258, + "loss": 2.5854, + "step": 17673 + }, + { + "epoch": 0.75, + "grad_norm": 0.462890625, + "learning_rate": 0.00013819861248921673, + "loss": 2.6479, + "step": 17674 + }, + { + "epoch": 0.75, + "grad_norm": 0.419921875, + "learning_rate": 0.00013819241314722106, + "loss": 2.5637, + "step": 17675 + }, + { + "epoch": 0.75, + "grad_norm": 0.50390625, + "learning_rate": 0.0001381862136333735, + "loss": 2.5748, + "step": 17676 + }, + { + "epoch": 0.75, + "grad_norm": 0.478515625, + "learning_rate": 0.00013818001394770194, + "loss": 2.5335, + "step": 17677 + }, + { + "epoch": 0.75, + "grad_norm": 0.494140625, + "learning_rate": 0.00013817381409023428, + "loss": 2.5616, + "step": 17678 + }, + { + "epoch": 0.75, + "grad_norm": 0.451171875, + "learning_rate": 0.0001381676140609984, + "loss": 2.5763, + "step": 17679 + }, + { + "epoch": 0.75, + "grad_norm": 0.4296875, + "learning_rate": 0.00013816141386002216, + "loss": 2.5713, + "step": 17680 + }, + { + "epoch": 0.75, + "grad_norm": 0.404296875, + "learning_rate": 0.00013815521348733357, + "loss": 2.6101, + "step": 17681 + }, + { + "epoch": 0.75, + "grad_norm": 0.423828125, + "learning_rate": 0.00013814901294296043, + "loss": 2.4964, + "step": 17682 + }, + { + "epoch": 0.75, + "grad_norm": 0.458984375, + "learning_rate": 0.00013814281222693068, + "loss": 2.6077, + "step": 17683 + }, + { + "epoch": 0.75, + "grad_norm": 0.4609375, + "learning_rate": 0.00013813661133927222, + "loss": 2.5026, + "step": 17684 + }, + { + "epoch": 0.75, + "grad_norm": 0.42578125, + "learning_rate": 0.00013813041028001293, + "loss": 2.6903, + "step": 17685 + }, + { + "epoch": 0.75, + "grad_norm": 0.49609375, + "learning_rate": 0.00013812420904918077, + "loss": 2.5475, + "step": 17686 + }, + { + "epoch": 0.75, + "grad_norm": 0.47265625, + "learning_rate": 0.00013811800764680355, + "loss": 2.6676, + "step": 17687 + }, + { + "epoch": 0.75, + "grad_norm": 0.51953125, + "learning_rate": 0.00013811180607290925, + "loss": 2.6469, + "step": 17688 + }, + { + "epoch": 0.75, + "grad_norm": 0.50390625, + "learning_rate": 0.00013810560432752579, + "loss": 2.5942, + "step": 17689 + }, + { + "epoch": 0.75, + "grad_norm": 0.447265625, + "learning_rate": 0.00013809940241068096, + "loss": 2.5236, + "step": 17690 + }, + { + "epoch": 0.75, + "grad_norm": 0.51953125, + "learning_rate": 0.00013809320032240278, + "loss": 2.6045, + "step": 17691 + }, + { + "epoch": 0.75, + "grad_norm": 0.4296875, + "learning_rate": 0.00013808699806271913, + "loss": 2.5565, + "step": 17692 + }, + { + "epoch": 0.75, + "grad_norm": 0.4921875, + "learning_rate": 0.0001380807956316579, + "loss": 2.5585, + "step": 17693 + }, + { + "epoch": 0.75, + "grad_norm": 0.474609375, + "learning_rate": 0.00013807459302924699, + "loss": 2.6195, + "step": 17694 + }, + { + "epoch": 0.75, + "grad_norm": 0.390625, + "learning_rate": 0.00013806839025551435, + "loss": 2.4941, + "step": 17695 + }, + { + "epoch": 0.75, + "grad_norm": 0.462890625, + "learning_rate": 0.0001380621873104879, + "loss": 2.6048, + "step": 17696 + }, + { + "epoch": 0.75, + "grad_norm": 0.4453125, + "learning_rate": 0.00013805598419419543, + "loss": 2.5554, + "step": 17697 + }, + { + "epoch": 0.75, + "grad_norm": 0.4921875, + "learning_rate": 0.000138049780906665, + "loss": 2.4993, + "step": 17698 + }, + { + "epoch": 0.75, + "grad_norm": 0.4296875, + "learning_rate": 0.00013804357744792443, + "loss": 2.5563, + "step": 17699 + }, + { + "epoch": 0.75, + "grad_norm": 0.404296875, + "learning_rate": 0.0001380373738180017, + "loss": 2.5749, + "step": 17700 + }, + { + "epoch": 0.75, + "grad_norm": 0.47265625, + "learning_rate": 0.00013803117001692466, + "loss": 2.708, + "step": 17701 + }, + { + "epoch": 0.75, + "grad_norm": 0.453125, + "learning_rate": 0.00013802496604472126, + "loss": 2.6128, + "step": 17702 + }, + { + "epoch": 0.75, + "grad_norm": 0.48828125, + "learning_rate": 0.0001380187619014194, + "loss": 2.6245, + "step": 17703 + }, + { + "epoch": 0.75, + "grad_norm": 0.435546875, + "learning_rate": 0.000138012557587047, + "loss": 2.6321, + "step": 17704 + }, + { + "epoch": 0.75, + "grad_norm": 0.494140625, + "learning_rate": 0.000138006353101632, + "loss": 2.5798, + "step": 17705 + }, + { + "epoch": 0.75, + "grad_norm": 0.5078125, + "learning_rate": 0.00013800014844520233, + "loss": 2.6418, + "step": 17706 + }, + { + "epoch": 0.75, + "grad_norm": 0.443359375, + "learning_rate": 0.00013799394361778584, + "loss": 2.6098, + "step": 17707 + }, + { + "epoch": 0.75, + "grad_norm": 0.4921875, + "learning_rate": 0.0001379877386194105, + "loss": 2.5213, + "step": 17708 + }, + { + "epoch": 0.75, + "grad_norm": 0.412109375, + "learning_rate": 0.0001379815334501042, + "loss": 2.4413, + "step": 17709 + }, + { + "epoch": 0.75, + "grad_norm": 0.4375, + "learning_rate": 0.00013797532810989492, + "loss": 2.4794, + "step": 17710 + }, + { + "epoch": 0.75, + "grad_norm": 0.451171875, + "learning_rate": 0.00013796912259881052, + "loss": 2.542, + "step": 17711 + }, + { + "epoch": 0.75, + "grad_norm": 0.50390625, + "learning_rate": 0.00013796291691687895, + "loss": 2.5227, + "step": 17712 + }, + { + "epoch": 0.75, + "grad_norm": 0.431640625, + "learning_rate": 0.0001379567110641281, + "loss": 2.6337, + "step": 17713 + }, + { + "epoch": 0.75, + "grad_norm": 0.45703125, + "learning_rate": 0.00013795050504058595, + "loss": 2.5752, + "step": 17714 + }, + { + "epoch": 0.75, + "grad_norm": 0.48046875, + "learning_rate": 0.00013794429884628038, + "loss": 2.6499, + "step": 17715 + }, + { + "epoch": 0.75, + "grad_norm": 0.453125, + "learning_rate": 0.00013793809248123937, + "loss": 2.5187, + "step": 17716 + }, + { + "epoch": 0.75, + "grad_norm": 0.498046875, + "learning_rate": 0.00013793188594549076, + "loss": 2.5246, + "step": 17717 + }, + { + "epoch": 0.76, + "grad_norm": 0.455078125, + "learning_rate": 0.00013792567923906254, + "loss": 2.447, + "step": 17718 + }, + { + "epoch": 0.76, + "grad_norm": 0.44140625, + "learning_rate": 0.00013791947236198264, + "loss": 2.6718, + "step": 17719 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.00013791326531427896, + "loss": 2.6397, + "step": 17720 + }, + { + "epoch": 0.76, + "grad_norm": 0.447265625, + "learning_rate": 0.00013790705809597946, + "loss": 2.5892, + "step": 17721 + }, + { + "epoch": 0.76, + "grad_norm": 0.453125, + "learning_rate": 0.00013790085070711206, + "loss": 2.5602, + "step": 17722 + }, + { + "epoch": 0.76, + "grad_norm": 0.42578125, + "learning_rate": 0.0001378946431477046, + "loss": 2.62, + "step": 17723 + }, + { + "epoch": 0.76, + "grad_norm": 0.4375, + "learning_rate": 0.0001378884354177852, + "loss": 2.5595, + "step": 17724 + }, + { + "epoch": 0.76, + "grad_norm": 0.48828125, + "learning_rate": 0.00013788222751738162, + "loss": 2.5415, + "step": 17725 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.00013787601944652189, + "loss": 2.479, + "step": 17726 + }, + { + "epoch": 0.76, + "grad_norm": 0.498046875, + "learning_rate": 0.0001378698112052339, + "loss": 2.6269, + "step": 17727 + }, + { + "epoch": 0.76, + "grad_norm": 0.435546875, + "learning_rate": 0.00013786360279354563, + "loss": 2.5452, + "step": 17728 + }, + { + "epoch": 0.76, + "grad_norm": 0.53515625, + "learning_rate": 0.00013785739421148495, + "loss": 2.6338, + "step": 17729 + }, + { + "epoch": 0.76, + "grad_norm": 0.490234375, + "learning_rate": 0.00013785118545907985, + "loss": 2.7022, + "step": 17730 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.0001378449765363582, + "loss": 2.7218, + "step": 17731 + }, + { + "epoch": 0.76, + "grad_norm": 0.4765625, + "learning_rate": 0.00013783876744334806, + "loss": 2.5046, + "step": 17732 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.00013783255818007725, + "loss": 2.5935, + "step": 17733 + }, + { + "epoch": 0.76, + "grad_norm": 0.419921875, + "learning_rate": 0.00013782634874657377, + "loss": 2.5347, + "step": 17734 + }, + { + "epoch": 0.76, + "grad_norm": 0.515625, + "learning_rate": 0.00013782013914286556, + "loss": 2.7862, + "step": 17735 + }, + { + "epoch": 0.76, + "grad_norm": 0.4765625, + "learning_rate": 0.0001378139293689805, + "loss": 2.3534, + "step": 17736 + }, + { + "epoch": 0.76, + "grad_norm": 0.455078125, + "learning_rate": 0.0001378077194249466, + "loss": 2.5269, + "step": 17737 + }, + { + "epoch": 0.76, + "grad_norm": 0.470703125, + "learning_rate": 0.00013780150931079181, + "loss": 2.5356, + "step": 17738 + }, + { + "epoch": 0.76, + "grad_norm": 0.419921875, + "learning_rate": 0.000137795299026544, + "loss": 2.526, + "step": 17739 + }, + { + "epoch": 0.76, + "grad_norm": 0.48046875, + "learning_rate": 0.0001377890885722312, + "loss": 2.571, + "step": 17740 + }, + { + "epoch": 0.76, + "grad_norm": 0.43359375, + "learning_rate": 0.00013778287794788128, + "loss": 2.5171, + "step": 17741 + }, + { + "epoch": 0.76, + "grad_norm": 0.43359375, + "learning_rate": 0.00013777666715352222, + "loss": 2.6921, + "step": 17742 + }, + { + "epoch": 0.76, + "grad_norm": 0.39453125, + "learning_rate": 0.00013777045618918197, + "loss": 2.6541, + "step": 17743 + }, + { + "epoch": 0.76, + "grad_norm": 0.48828125, + "learning_rate": 0.00013776424505488848, + "loss": 2.5904, + "step": 17744 + }, + { + "epoch": 0.76, + "grad_norm": 0.46875, + "learning_rate": 0.00013775803375066966, + "loss": 2.7507, + "step": 17745 + }, + { + "epoch": 0.76, + "grad_norm": 0.482421875, + "learning_rate": 0.0001377518222765535, + "loss": 2.6145, + "step": 17746 + }, + { + "epoch": 0.76, + "grad_norm": 0.427734375, + "learning_rate": 0.0001377456106325679, + "loss": 2.6279, + "step": 17747 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.00013773939881874095, + "loss": 2.578, + "step": 17748 + }, + { + "epoch": 0.76, + "grad_norm": 0.48046875, + "learning_rate": 0.0001377331868351004, + "loss": 2.7211, + "step": 17749 + }, + { + "epoch": 0.76, + "grad_norm": 0.419921875, + "learning_rate": 0.00013772697468167434, + "loss": 2.3866, + "step": 17750 + }, + { + "epoch": 0.76, + "grad_norm": 0.443359375, + "learning_rate": 0.00013772076235849068, + "loss": 2.5648, + "step": 17751 + }, + { + "epoch": 0.76, + "grad_norm": 0.435546875, + "learning_rate": 0.00013771454986557738, + "loss": 2.606, + "step": 17752 + }, + { + "epoch": 0.76, + "grad_norm": 0.439453125, + "learning_rate": 0.00013770833720296236, + "loss": 2.6762, + "step": 17753 + }, + { + "epoch": 0.76, + "grad_norm": 0.443359375, + "learning_rate": 0.00013770212437067364, + "loss": 2.5043, + "step": 17754 + }, + { + "epoch": 0.76, + "grad_norm": 0.4140625, + "learning_rate": 0.0001376959113687391, + "loss": 2.3678, + "step": 17755 + }, + { + "epoch": 0.76, + "grad_norm": 0.4609375, + "learning_rate": 0.00013768969819718678, + "loss": 2.4895, + "step": 17756 + }, + { + "epoch": 0.76, + "grad_norm": 0.421875, + "learning_rate": 0.00013768348485604454, + "loss": 2.5477, + "step": 17757 + }, + { + "epoch": 0.76, + "grad_norm": 0.421875, + "learning_rate": 0.00013767727134534043, + "loss": 2.4403, + "step": 17758 + }, + { + "epoch": 0.76, + "grad_norm": 0.453125, + "learning_rate": 0.00013767105766510237, + "loss": 2.6147, + "step": 17759 + }, + { + "epoch": 0.76, + "grad_norm": 0.453125, + "learning_rate": 0.0001376648438153583, + "loss": 2.5987, + "step": 17760 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.0001376586297961362, + "loss": 2.5361, + "step": 17761 + }, + { + "epoch": 0.76, + "grad_norm": 0.51171875, + "learning_rate": 0.00013765241560746401, + "loss": 2.6248, + "step": 17762 + }, + { + "epoch": 0.76, + "grad_norm": 0.435546875, + "learning_rate": 0.00013764620124936974, + "loss": 2.4707, + "step": 17763 + }, + { + "epoch": 0.76, + "grad_norm": 0.412109375, + "learning_rate": 0.00013763998672188131, + "loss": 2.5368, + "step": 17764 + }, + { + "epoch": 0.76, + "grad_norm": 0.466796875, + "learning_rate": 0.00013763377202502669, + "loss": 2.6525, + "step": 17765 + }, + { + "epoch": 0.76, + "grad_norm": 0.482421875, + "learning_rate": 0.00013762755715883386, + "loss": 2.459, + "step": 17766 + }, + { + "epoch": 0.76, + "grad_norm": 0.453125, + "learning_rate": 0.00013762134212333077, + "loss": 2.4436, + "step": 17767 + }, + { + "epoch": 0.76, + "grad_norm": 0.486328125, + "learning_rate": 0.0001376151269185454, + "loss": 2.5772, + "step": 17768 + }, + { + "epoch": 0.76, + "grad_norm": 0.490234375, + "learning_rate": 0.00013760891154450565, + "loss": 2.5436, + "step": 17769 + }, + { + "epoch": 0.76, + "grad_norm": 0.515625, + "learning_rate": 0.00013760269600123962, + "loss": 2.632, + "step": 17770 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.00013759648028877517, + "loss": 2.7243, + "step": 17771 + }, + { + "epoch": 0.76, + "grad_norm": 0.416015625, + "learning_rate": 0.00013759026440714032, + "loss": 2.5558, + "step": 17772 + }, + { + "epoch": 0.76, + "grad_norm": 0.47265625, + "learning_rate": 0.00013758404835636297, + "loss": 2.5809, + "step": 17773 + }, + { + "epoch": 0.76, + "grad_norm": 0.40625, + "learning_rate": 0.00013757783213647117, + "loss": 2.5455, + "step": 17774 + }, + { + "epoch": 0.76, + "grad_norm": 0.451171875, + "learning_rate": 0.00013757161574749283, + "loss": 2.4549, + "step": 17775 + }, + { + "epoch": 0.76, + "grad_norm": 0.44921875, + "learning_rate": 0.00013756539918945601, + "loss": 2.5441, + "step": 17776 + }, + { + "epoch": 0.76, + "grad_norm": 0.431640625, + "learning_rate": 0.00013755918246238856, + "loss": 2.5383, + "step": 17777 + }, + { + "epoch": 0.76, + "grad_norm": 0.38671875, + "learning_rate": 0.00013755296556631856, + "loss": 2.5388, + "step": 17778 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.0001375467485012739, + "loss": 2.559, + "step": 17779 + }, + { + "epoch": 0.76, + "grad_norm": 0.431640625, + "learning_rate": 0.00013754053126728262, + "loss": 2.5527, + "step": 17780 + }, + { + "epoch": 0.76, + "grad_norm": 0.50390625, + "learning_rate": 0.00013753431386437264, + "loss": 2.7386, + "step": 17781 + }, + { + "epoch": 0.76, + "grad_norm": 0.4296875, + "learning_rate": 0.00013752809629257202, + "loss": 2.6127, + "step": 17782 + }, + { + "epoch": 0.76, + "grad_norm": 0.4453125, + "learning_rate": 0.00013752187855190865, + "loss": 2.5123, + "step": 17783 + }, + { + "epoch": 0.76, + "grad_norm": 0.46875, + "learning_rate": 0.00013751566064241052, + "loss": 2.5876, + "step": 17784 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.00013750944256410563, + "loss": 2.4872, + "step": 17785 + }, + { + "epoch": 0.76, + "grad_norm": 0.490234375, + "learning_rate": 0.00013750322431702202, + "loss": 2.5176, + "step": 17786 + }, + { + "epoch": 0.76, + "grad_norm": 0.484375, + "learning_rate": 0.0001374970059011875, + "loss": 2.6746, + "step": 17787 + }, + { + "epoch": 0.76, + "grad_norm": 0.419921875, + "learning_rate": 0.00013749078731663025, + "loss": 2.5917, + "step": 17788 + }, + { + "epoch": 0.76, + "grad_norm": 0.46484375, + "learning_rate": 0.0001374845685633781, + "loss": 2.5849, + "step": 17789 + }, + { + "epoch": 0.76, + "grad_norm": 0.43359375, + "learning_rate": 0.0001374783496414591, + "loss": 2.5145, + "step": 17790 + }, + { + "epoch": 0.76, + "grad_norm": 0.435546875, + "learning_rate": 0.00013747213055090125, + "loss": 2.4655, + "step": 17791 + }, + { + "epoch": 0.76, + "grad_norm": 0.439453125, + "learning_rate": 0.0001374659112917325, + "loss": 2.629, + "step": 17792 + }, + { + "epoch": 0.76, + "grad_norm": 0.4765625, + "learning_rate": 0.00013745969186398083, + "loss": 2.4592, + "step": 17793 + }, + { + "epoch": 0.76, + "grad_norm": 0.453125, + "learning_rate": 0.00013745347226767427, + "loss": 2.5568, + "step": 17794 + }, + { + "epoch": 0.76, + "grad_norm": 0.482421875, + "learning_rate": 0.00013744725250284073, + "loss": 2.6224, + "step": 17795 + }, + { + "epoch": 0.76, + "grad_norm": 0.5, + "learning_rate": 0.00013744103256950823, + "loss": 2.586, + "step": 17796 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.0001374348124677048, + "loss": 2.4545, + "step": 17797 + }, + { + "epoch": 0.76, + "grad_norm": 0.51953125, + "learning_rate": 0.00013742859219745837, + "loss": 2.6705, + "step": 17798 + }, + { + "epoch": 0.76, + "grad_norm": 0.490234375, + "learning_rate": 0.00013742237175879698, + "loss": 2.4722, + "step": 17799 + }, + { + "epoch": 0.76, + "grad_norm": 0.5078125, + "learning_rate": 0.0001374161511517486, + "loss": 2.5742, + "step": 17800 + }, + { + "epoch": 0.76, + "grad_norm": 0.466796875, + "learning_rate": 0.00013740993037634117, + "loss": 2.5555, + "step": 17801 + }, + { + "epoch": 0.76, + "grad_norm": 0.439453125, + "learning_rate": 0.00013740370943260278, + "loss": 2.5372, + "step": 17802 + }, + { + "epoch": 0.76, + "grad_norm": 0.451171875, + "learning_rate": 0.00013739748832056133, + "loss": 2.555, + "step": 17803 + }, + { + "epoch": 0.76, + "grad_norm": 0.455078125, + "learning_rate": 0.00013739126704024484, + "loss": 2.5344, + "step": 17804 + }, + { + "epoch": 0.76, + "grad_norm": 0.46875, + "learning_rate": 0.00013738504559168132, + "loss": 2.5842, + "step": 17805 + }, + { + "epoch": 0.76, + "grad_norm": 0.546875, + "learning_rate": 0.0001373788239748988, + "loss": 2.6449, + "step": 17806 + }, + { + "epoch": 0.76, + "grad_norm": 0.4375, + "learning_rate": 0.00013737260218992518, + "loss": 2.583, + "step": 17807 + }, + { + "epoch": 0.76, + "grad_norm": 0.5, + "learning_rate": 0.00013736638023678855, + "loss": 2.8025, + "step": 17808 + }, + { + "epoch": 0.76, + "grad_norm": 0.5, + "learning_rate": 0.0001373601581155168, + "loss": 2.5843, + "step": 17809 + }, + { + "epoch": 0.76, + "grad_norm": 0.41015625, + "learning_rate": 0.00013735393582613807, + "loss": 2.5965, + "step": 17810 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.00013734771336868025, + "loss": 2.4972, + "step": 17811 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.00013734149074317136, + "loss": 2.588, + "step": 17812 + }, + { + "epoch": 0.76, + "grad_norm": 0.5078125, + "learning_rate": 0.0001373352679496394, + "loss": 2.5632, + "step": 17813 + }, + { + "epoch": 0.76, + "grad_norm": 0.4921875, + "learning_rate": 0.00013732904498811238, + "loss": 2.5653, + "step": 17814 + }, + { + "epoch": 0.76, + "grad_norm": 0.50390625, + "learning_rate": 0.00013732282185861835, + "loss": 2.6699, + "step": 17815 + }, + { + "epoch": 0.76, + "grad_norm": 0.46484375, + "learning_rate": 0.0001373165985611852, + "loss": 2.4853, + "step": 17816 + }, + { + "epoch": 0.76, + "grad_norm": 0.474609375, + "learning_rate": 0.00013731037509584102, + "loss": 2.6671, + "step": 17817 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.00013730415146261378, + "loss": 2.4922, + "step": 17818 + }, + { + "epoch": 0.76, + "grad_norm": 0.41796875, + "learning_rate": 0.0001372979276615315, + "loss": 2.578, + "step": 17819 + }, + { + "epoch": 0.76, + "grad_norm": 0.51171875, + "learning_rate": 0.00013729170369262215, + "loss": 2.7045, + "step": 17820 + }, + { + "epoch": 0.76, + "grad_norm": 0.53125, + "learning_rate": 0.00013728547955591382, + "loss": 2.6276, + "step": 17821 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.0001372792552514344, + "loss": 2.551, + "step": 17822 + }, + { + "epoch": 0.76, + "grad_norm": 0.44921875, + "learning_rate": 0.00013727303077921198, + "loss": 2.5076, + "step": 17823 + }, + { + "epoch": 0.76, + "grad_norm": 0.47265625, + "learning_rate": 0.00013726680613927457, + "loss": 2.4529, + "step": 17824 + }, + { + "epoch": 0.76, + "grad_norm": 0.45703125, + "learning_rate": 0.0001372605813316501, + "loss": 2.5566, + "step": 17825 + }, + { + "epoch": 0.76, + "grad_norm": 0.4609375, + "learning_rate": 0.00013725435635636668, + "loss": 2.6356, + "step": 17826 + }, + { + "epoch": 0.76, + "grad_norm": 0.470703125, + "learning_rate": 0.00013724813121345224, + "loss": 2.4674, + "step": 17827 + }, + { + "epoch": 0.76, + "grad_norm": 0.37890625, + "learning_rate": 0.0001372419059029348, + "loss": 2.5229, + "step": 17828 + }, + { + "epoch": 0.76, + "grad_norm": 0.4765625, + "learning_rate": 0.00013723568042484246, + "loss": 2.5498, + "step": 17829 + }, + { + "epoch": 0.76, + "grad_norm": 0.439453125, + "learning_rate": 0.00013722945477920312, + "loss": 2.4674, + "step": 17830 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.00013722322896604484, + "loss": 2.404, + "step": 17831 + }, + { + "epoch": 0.76, + "grad_norm": 0.412109375, + "learning_rate": 0.00013721700298539565, + "loss": 2.4737, + "step": 17832 + }, + { + "epoch": 0.76, + "grad_norm": 0.44921875, + "learning_rate": 0.00013721077683728352, + "loss": 2.5299, + "step": 17833 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.0001372045505217365, + "loss": 2.6015, + "step": 17834 + }, + { + "epoch": 0.76, + "grad_norm": 0.453125, + "learning_rate": 0.00013719832403878262, + "loss": 2.5112, + "step": 17835 + }, + { + "epoch": 0.76, + "grad_norm": 0.490234375, + "learning_rate": 0.00013719209738844985, + "loss": 2.5571, + "step": 17836 + }, + { + "epoch": 0.76, + "grad_norm": 0.4921875, + "learning_rate": 0.00013718587057076628, + "loss": 2.6367, + "step": 17837 + }, + { + "epoch": 0.76, + "grad_norm": 0.5078125, + "learning_rate": 0.00013717964358575982, + "loss": 2.6575, + "step": 17838 + }, + { + "epoch": 0.76, + "grad_norm": 0.41796875, + "learning_rate": 0.0001371734164334586, + "loss": 2.4891, + "step": 17839 + }, + { + "epoch": 0.76, + "grad_norm": 0.427734375, + "learning_rate": 0.00013716718911389056, + "loss": 2.4217, + "step": 17840 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.00013716096162708378, + "loss": 2.5212, + "step": 17841 + }, + { + "epoch": 0.76, + "grad_norm": 0.431640625, + "learning_rate": 0.00013715473397306625, + "loss": 2.5964, + "step": 17842 + }, + { + "epoch": 0.76, + "grad_norm": 0.5078125, + "learning_rate": 0.000137148506151866, + "loss": 2.5375, + "step": 17843 + }, + { + "epoch": 0.76, + "grad_norm": 0.5078125, + "learning_rate": 0.000137142278163511, + "loss": 2.6678, + "step": 17844 + }, + { + "epoch": 0.76, + "grad_norm": 0.451171875, + "learning_rate": 0.00013713605000802937, + "loss": 2.4896, + "step": 17845 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.0001371298216854491, + "loss": 2.7429, + "step": 17846 + }, + { + "epoch": 0.76, + "grad_norm": 0.466796875, + "learning_rate": 0.00013712359319579815, + "loss": 2.4686, + "step": 17847 + }, + { + "epoch": 0.76, + "grad_norm": 0.45703125, + "learning_rate": 0.00013711736453910463, + "loss": 2.5832, + "step": 17848 + }, + { + "epoch": 0.76, + "grad_norm": 0.419921875, + "learning_rate": 0.00013711113571539653, + "loss": 2.5685, + "step": 17849 + }, + { + "epoch": 0.76, + "grad_norm": 0.470703125, + "learning_rate": 0.0001371049067247019, + "loss": 2.5878, + "step": 17850 + }, + { + "epoch": 0.76, + "grad_norm": 0.439453125, + "learning_rate": 0.00013709867756704874, + "loss": 2.5626, + "step": 17851 + }, + { + "epoch": 0.76, + "grad_norm": 0.44921875, + "learning_rate": 0.00013709244824246504, + "loss": 2.5323, + "step": 17852 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.00013708621875097894, + "loss": 2.5014, + "step": 17853 + }, + { + "epoch": 0.76, + "grad_norm": 0.51171875, + "learning_rate": 0.00013707998909261838, + "loss": 2.5704, + "step": 17854 + }, + { + "epoch": 0.76, + "grad_norm": 0.42578125, + "learning_rate": 0.00013707375926741143, + "loss": 2.6397, + "step": 17855 + }, + { + "epoch": 0.76, + "grad_norm": 0.50390625, + "learning_rate": 0.0001370675292753861, + "loss": 2.5798, + "step": 17856 + }, + { + "epoch": 0.76, + "grad_norm": 0.474609375, + "learning_rate": 0.00013706129911657044, + "loss": 2.6113, + "step": 17857 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.00013705506879099248, + "loss": 2.4963, + "step": 17858 + }, + { + "epoch": 0.76, + "grad_norm": 0.443359375, + "learning_rate": 0.00013704883829868028, + "loss": 2.5201, + "step": 17859 + }, + { + "epoch": 0.76, + "grad_norm": 0.484375, + "learning_rate": 0.0001370426076396618, + "loss": 2.6013, + "step": 17860 + }, + { + "epoch": 0.76, + "grad_norm": 0.4375, + "learning_rate": 0.00013703637681396517, + "loss": 2.5247, + "step": 17861 + }, + { + "epoch": 0.76, + "grad_norm": 0.49609375, + "learning_rate": 0.00013703014582161832, + "loss": 2.6471, + "step": 17862 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.0001370239146626494, + "loss": 2.563, + "step": 17863 + }, + { + "epoch": 0.76, + "grad_norm": 0.39453125, + "learning_rate": 0.00013701768333708637, + "loss": 2.3281, + "step": 17864 + }, + { + "epoch": 0.76, + "grad_norm": 0.44921875, + "learning_rate": 0.00013701145184495732, + "loss": 2.5935, + "step": 17865 + }, + { + "epoch": 0.76, + "grad_norm": 0.478515625, + "learning_rate": 0.00013700522018629023, + "loss": 2.5072, + "step": 17866 + }, + { + "epoch": 0.76, + "grad_norm": 0.470703125, + "learning_rate": 0.00013699898836111319, + "loss": 2.6473, + "step": 17867 + }, + { + "epoch": 0.76, + "grad_norm": 0.53125, + "learning_rate": 0.0001369927563694542, + "loss": 2.6514, + "step": 17868 + }, + { + "epoch": 0.76, + "grad_norm": 0.5, + "learning_rate": 0.00013698652421134137, + "loss": 2.5326, + "step": 17869 + }, + { + "epoch": 0.76, + "grad_norm": 0.439453125, + "learning_rate": 0.00013698029188680266, + "loss": 2.5742, + "step": 17870 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.00013697405939586615, + "loss": 2.5341, + "step": 17871 + }, + { + "epoch": 0.76, + "grad_norm": 0.435546875, + "learning_rate": 0.00013696782673855993, + "loss": 2.4187, + "step": 17872 + }, + { + "epoch": 0.76, + "grad_norm": 0.474609375, + "learning_rate": 0.00013696159391491197, + "loss": 2.5144, + "step": 17873 + }, + { + "epoch": 0.76, + "grad_norm": 0.431640625, + "learning_rate": 0.00013695536092495035, + "loss": 2.6213, + "step": 17874 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.0001369491277687031, + "loss": 2.555, + "step": 17875 + }, + { + "epoch": 0.76, + "grad_norm": 0.5, + "learning_rate": 0.00013694289444619826, + "loss": 2.4921, + "step": 17876 + }, + { + "epoch": 0.76, + "grad_norm": 0.41015625, + "learning_rate": 0.00013693666095746396, + "loss": 2.4142, + "step": 17877 + }, + { + "epoch": 0.76, + "grad_norm": 0.42578125, + "learning_rate": 0.0001369304273025281, + "loss": 2.5287, + "step": 17878 + }, + { + "epoch": 0.76, + "grad_norm": 0.48828125, + "learning_rate": 0.00013692419348141887, + "loss": 2.6309, + "step": 17879 + }, + { + "epoch": 0.76, + "grad_norm": 0.4296875, + "learning_rate": 0.00013691795949416426, + "loss": 2.5876, + "step": 17880 + }, + { + "epoch": 0.76, + "grad_norm": 0.43359375, + "learning_rate": 0.00013691172534079232, + "loss": 2.5999, + "step": 17881 + }, + { + "epoch": 0.76, + "grad_norm": 0.5390625, + "learning_rate": 0.00013690549102133112, + "loss": 2.5957, + "step": 17882 + }, + { + "epoch": 0.76, + "grad_norm": 0.44921875, + "learning_rate": 0.00013689925653580866, + "loss": 2.5635, + "step": 17883 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.00013689302188425306, + "loss": 2.5868, + "step": 17884 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.00013688678706669233, + "loss": 2.5104, + "step": 17885 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.00013688055208315453, + "loss": 2.6673, + "step": 17886 + }, + { + "epoch": 0.76, + "grad_norm": 0.451171875, + "learning_rate": 0.00013687431693366775, + "loss": 2.5103, + "step": 17887 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.00013686808161826, + "loss": 2.5235, + "step": 17888 + }, + { + "epoch": 0.76, + "grad_norm": 0.470703125, + "learning_rate": 0.00013686184613695938, + "loss": 2.6065, + "step": 17889 + }, + { + "epoch": 0.76, + "grad_norm": 0.41015625, + "learning_rate": 0.0001368556104897939, + "loss": 2.5311, + "step": 17890 + }, + { + "epoch": 0.76, + "grad_norm": 0.5078125, + "learning_rate": 0.00013684937467679166, + "loss": 2.6282, + "step": 17891 + }, + { + "epoch": 0.76, + "grad_norm": 0.419921875, + "learning_rate": 0.00013684313869798066, + "loss": 2.5395, + "step": 17892 + }, + { + "epoch": 0.76, + "grad_norm": 0.447265625, + "learning_rate": 0.00013683690255338908, + "loss": 2.4987, + "step": 17893 + }, + { + "epoch": 0.76, + "grad_norm": 0.4140625, + "learning_rate": 0.00013683066624304484, + "loss": 2.4905, + "step": 17894 + }, + { + "epoch": 0.76, + "grad_norm": 0.435546875, + "learning_rate": 0.00013682442976697607, + "loss": 2.5631, + "step": 17895 + }, + { + "epoch": 0.76, + "grad_norm": 0.44921875, + "learning_rate": 0.00013681819312521081, + "loss": 2.5297, + "step": 17896 + }, + { + "epoch": 0.76, + "grad_norm": 0.4609375, + "learning_rate": 0.00013681195631777717, + "loss": 2.4454, + "step": 17897 + }, + { + "epoch": 0.76, + "grad_norm": 0.470703125, + "learning_rate": 0.00013680571934470318, + "loss": 2.5283, + "step": 17898 + }, + { + "epoch": 0.76, + "grad_norm": 0.427734375, + "learning_rate": 0.00013679948220601687, + "loss": 2.6086, + "step": 17899 + }, + { + "epoch": 0.76, + "grad_norm": 0.482421875, + "learning_rate": 0.00013679324490174633, + "loss": 2.5652, + "step": 17900 + }, + { + "epoch": 0.76, + "grad_norm": 0.4765625, + "learning_rate": 0.0001367870074319197, + "loss": 2.6058, + "step": 17901 + }, + { + "epoch": 0.76, + "grad_norm": 0.44921875, + "learning_rate": 0.00013678076979656493, + "loss": 2.5616, + "step": 17902 + }, + { + "epoch": 0.76, + "grad_norm": 0.453125, + "learning_rate": 0.00013677453199571014, + "loss": 2.5678, + "step": 17903 + }, + { + "epoch": 0.76, + "grad_norm": 0.447265625, + "learning_rate": 0.00013676829402938342, + "loss": 2.5996, + "step": 17904 + }, + { + "epoch": 0.76, + "grad_norm": 0.400390625, + "learning_rate": 0.0001367620558976128, + "loss": 2.5468, + "step": 17905 + }, + { + "epoch": 0.76, + "grad_norm": 0.462890625, + "learning_rate": 0.00013675581760042637, + "loss": 2.5479, + "step": 17906 + }, + { + "epoch": 0.76, + "grad_norm": 0.4453125, + "learning_rate": 0.0001367495791378522, + "loss": 2.636, + "step": 17907 + }, + { + "epoch": 0.76, + "grad_norm": 0.478515625, + "learning_rate": 0.00013674334050991832, + "loss": 2.5018, + "step": 17908 + }, + { + "epoch": 0.76, + "grad_norm": 0.48046875, + "learning_rate": 0.00013673710171665287, + "loss": 2.5749, + "step": 17909 + }, + { + "epoch": 0.76, + "grad_norm": 0.498046875, + "learning_rate": 0.00013673086275808386, + "loss": 2.5931, + "step": 17910 + }, + { + "epoch": 0.76, + "grad_norm": 0.44140625, + "learning_rate": 0.00013672462363423942, + "loss": 2.6358, + "step": 17911 + }, + { + "epoch": 0.76, + "grad_norm": 0.47265625, + "learning_rate": 0.00013671838434514756, + "loss": 2.5157, + "step": 17912 + }, + { + "epoch": 0.76, + "grad_norm": 0.4375, + "learning_rate": 0.00013671214489083643, + "loss": 2.6712, + "step": 17913 + }, + { + "epoch": 0.76, + "grad_norm": 0.5078125, + "learning_rate": 0.00013670590527133404, + "loss": 2.6941, + "step": 17914 + }, + { + "epoch": 0.76, + "grad_norm": 0.50390625, + "learning_rate": 0.0001366996654866685, + "loss": 2.6229, + "step": 17915 + }, + { + "epoch": 0.76, + "grad_norm": 0.447265625, + "learning_rate": 0.00013669342553686786, + "loss": 2.5324, + "step": 17916 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.00013668718542196026, + "loss": 2.5236, + "step": 17917 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.00013668094514197366, + "loss": 2.7351, + "step": 17918 + }, + { + "epoch": 0.76, + "grad_norm": 0.482421875, + "learning_rate": 0.0001366747046969363, + "loss": 2.7682, + "step": 17919 + }, + { + "epoch": 0.76, + "grad_norm": 0.486328125, + "learning_rate": 0.0001366684640868761, + "loss": 2.5308, + "step": 17920 + }, + { + "epoch": 0.76, + "grad_norm": 0.458984375, + "learning_rate": 0.00013666222331182122, + "loss": 2.5788, + "step": 17921 + }, + { + "epoch": 0.76, + "grad_norm": 0.5, + "learning_rate": 0.0001366559823717998, + "loss": 2.3838, + "step": 17922 + }, + { + "epoch": 0.76, + "grad_norm": 0.369140625, + "learning_rate": 0.0001366497412668398, + "loss": 2.5247, + "step": 17923 + }, + { + "epoch": 0.76, + "grad_norm": 0.40625, + "learning_rate": 0.00013664349999696935, + "loss": 2.5922, + "step": 17924 + }, + { + "epoch": 0.76, + "grad_norm": 0.486328125, + "learning_rate": 0.00013663725856221656, + "loss": 2.5522, + "step": 17925 + }, + { + "epoch": 0.76, + "grad_norm": 0.486328125, + "learning_rate": 0.0001366310169626095, + "loss": 2.6033, + "step": 17926 + }, + { + "epoch": 0.76, + "grad_norm": 0.404296875, + "learning_rate": 0.00013662477519817622, + "loss": 2.5477, + "step": 17927 + }, + { + "epoch": 0.76, + "grad_norm": 0.44140625, + "learning_rate": 0.00013661853326894487, + "loss": 2.5043, + "step": 17928 + }, + { + "epoch": 0.76, + "grad_norm": 0.48046875, + "learning_rate": 0.0001366122911749435, + "loss": 2.6393, + "step": 17929 + }, + { + "epoch": 0.76, + "grad_norm": 0.470703125, + "learning_rate": 0.00013660604891620022, + "loss": 2.4455, + "step": 17930 + }, + { + "epoch": 0.76, + "grad_norm": 0.4609375, + "learning_rate": 0.00013659980649274306, + "loss": 2.5359, + "step": 17931 + }, + { + "epoch": 0.76, + "grad_norm": 0.423828125, + "learning_rate": 0.00013659356390460014, + "loss": 2.665, + "step": 17932 + }, + { + "epoch": 0.76, + "grad_norm": 0.419921875, + "learning_rate": 0.0001365873211517996, + "loss": 2.5157, + "step": 17933 + }, + { + "epoch": 0.76, + "grad_norm": 0.474609375, + "learning_rate": 0.00013658107823436943, + "loss": 2.575, + "step": 17934 + }, + { + "epoch": 0.76, + "grad_norm": 0.46875, + "learning_rate": 0.0001365748351523378, + "loss": 2.5785, + "step": 17935 + }, + { + "epoch": 0.76, + "grad_norm": 0.51171875, + "learning_rate": 0.0001365685919057328, + "loss": 2.6136, + "step": 17936 + }, + { + "epoch": 0.76, + "grad_norm": 0.46875, + "learning_rate": 0.0001365623484945825, + "loss": 2.5342, + "step": 17937 + }, + { + "epoch": 0.76, + "grad_norm": 0.466796875, + "learning_rate": 0.00013655610491891497, + "loss": 2.61, + "step": 17938 + }, + { + "epoch": 0.76, + "grad_norm": 0.435546875, + "learning_rate": 0.00013654986117875835, + "loss": 2.4284, + "step": 17939 + }, + { + "epoch": 0.76, + "grad_norm": 0.443359375, + "learning_rate": 0.0001365436172741407, + "loss": 2.4683, + "step": 17940 + }, + { + "epoch": 0.76, + "grad_norm": 0.443359375, + "learning_rate": 0.00013653737320509012, + "loss": 2.691, + "step": 17941 + }, + { + "epoch": 0.76, + "grad_norm": 0.443359375, + "learning_rate": 0.0001365311289716347, + "loss": 2.5496, + "step": 17942 + }, + { + "epoch": 0.76, + "grad_norm": 0.48828125, + "learning_rate": 0.00013652488457380256, + "loss": 2.5283, + "step": 17943 + }, + { + "epoch": 0.76, + "grad_norm": 0.4921875, + "learning_rate": 0.00013651864001162185, + "loss": 2.7159, + "step": 17944 + }, + { + "epoch": 0.76, + "grad_norm": 0.482421875, + "learning_rate": 0.0001365123952851205, + "loss": 2.6591, + "step": 17945 + }, + { + "epoch": 0.76, + "grad_norm": 0.443359375, + "learning_rate": 0.0001365061503943268, + "loss": 2.4912, + "step": 17946 + }, + { + "epoch": 0.76, + "grad_norm": 0.435546875, + "learning_rate": 0.00013649990533926872, + "loss": 2.5195, + "step": 17947 + }, + { + "epoch": 0.76, + "grad_norm": 0.47265625, + "learning_rate": 0.00013649366011997445, + "loss": 2.7101, + "step": 17948 + }, + { + "epoch": 0.76, + "grad_norm": 0.51171875, + "learning_rate": 0.000136487414736472, + "loss": 2.5079, + "step": 17949 + }, + { + "epoch": 0.76, + "grad_norm": 0.50390625, + "learning_rate": 0.00013648116918878956, + "loss": 2.5778, + "step": 17950 + }, + { + "epoch": 0.76, + "grad_norm": 0.51171875, + "learning_rate": 0.00013647492347695512, + "loss": 2.6285, + "step": 17951 + }, + { + "epoch": 0.76, + "grad_norm": 0.478515625, + "learning_rate": 0.00013646867760099693, + "loss": 2.5687, + "step": 17952 + }, + { + "epoch": 0.77, + "grad_norm": 0.46484375, + "learning_rate": 0.00013646243156094297, + "loss": 2.4799, + "step": 17953 + }, + { + "epoch": 0.77, + "grad_norm": 0.439453125, + "learning_rate": 0.0001364561853568214, + "loss": 2.5074, + "step": 17954 + }, + { + "epoch": 0.77, + "grad_norm": 0.4140625, + "learning_rate": 0.00013644993898866034, + "loss": 2.5504, + "step": 17955 + }, + { + "epoch": 0.77, + "grad_norm": 0.48046875, + "learning_rate": 0.00013644369245648788, + "loss": 2.7166, + "step": 17956 + }, + { + "epoch": 0.77, + "grad_norm": 0.41015625, + "learning_rate": 0.0001364374457603321, + "loss": 2.719, + "step": 17957 + }, + { + "epoch": 0.77, + "grad_norm": 0.4296875, + "learning_rate": 0.00013643119890022114, + "loss": 2.5458, + "step": 17958 + }, + { + "epoch": 0.77, + "grad_norm": 0.671875, + "learning_rate": 0.0001364249518761831, + "loss": 2.4168, + "step": 17959 + }, + { + "epoch": 0.77, + "grad_norm": 0.44140625, + "learning_rate": 0.0001364187046882461, + "loss": 2.4424, + "step": 17960 + }, + { + "epoch": 0.77, + "grad_norm": 0.435546875, + "learning_rate": 0.0001364124573364382, + "loss": 2.5516, + "step": 17961 + }, + { + "epoch": 0.77, + "grad_norm": 0.48828125, + "learning_rate": 0.00013640620982078758, + "loss": 2.5327, + "step": 17962 + }, + { + "epoch": 0.77, + "grad_norm": 0.44921875, + "learning_rate": 0.0001363999621413223, + "loss": 2.6784, + "step": 17963 + }, + { + "epoch": 0.77, + "grad_norm": 0.439453125, + "learning_rate": 0.00013639371429807054, + "loss": 2.6245, + "step": 17964 + }, + { + "epoch": 0.77, + "grad_norm": 0.48828125, + "learning_rate": 0.0001363874662910603, + "loss": 2.6238, + "step": 17965 + }, + { + "epoch": 0.77, + "grad_norm": 0.431640625, + "learning_rate": 0.0001363812181203198, + "loss": 2.6821, + "step": 17966 + }, + { + "epoch": 0.77, + "grad_norm": 0.484375, + "learning_rate": 0.00013637496978587712, + "loss": 2.6228, + "step": 17967 + }, + { + "epoch": 0.77, + "grad_norm": 0.40625, + "learning_rate": 0.00013636872128776036, + "loss": 2.6476, + "step": 17968 + }, + { + "epoch": 0.77, + "grad_norm": 0.484375, + "learning_rate": 0.00013636247262599764, + "loss": 2.6445, + "step": 17969 + }, + { + "epoch": 0.77, + "grad_norm": 0.52734375, + "learning_rate": 0.00013635622380061708, + "loss": 2.5607, + "step": 17970 + }, + { + "epoch": 0.77, + "grad_norm": 0.423828125, + "learning_rate": 0.0001363499748116468, + "loss": 2.4345, + "step": 17971 + }, + { + "epoch": 0.77, + "grad_norm": 0.4375, + "learning_rate": 0.00013634372565911494, + "loss": 2.5255, + "step": 17972 + }, + { + "epoch": 0.77, + "grad_norm": 0.404296875, + "learning_rate": 0.00013633747634304955, + "loss": 2.6353, + "step": 17973 + }, + { + "epoch": 0.77, + "grad_norm": 0.39453125, + "learning_rate": 0.00013633122686347886, + "loss": 2.4602, + "step": 17974 + }, + { + "epoch": 0.77, + "grad_norm": 0.44140625, + "learning_rate": 0.00013632497722043088, + "loss": 2.6998, + "step": 17975 + }, + { + "epoch": 0.77, + "grad_norm": 0.44921875, + "learning_rate": 0.0001363187274139338, + "loss": 2.4923, + "step": 17976 + }, + { + "epoch": 0.77, + "grad_norm": 0.4609375, + "learning_rate": 0.0001363124774440157, + "loss": 2.5034, + "step": 17977 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.00013630622731070473, + "loss": 2.5244, + "step": 17978 + }, + { + "epoch": 0.77, + "grad_norm": 0.447265625, + "learning_rate": 0.000136299977014029, + "loss": 2.568, + "step": 17979 + }, + { + "epoch": 0.77, + "grad_norm": 0.439453125, + "learning_rate": 0.00013629372655401663, + "loss": 2.6527, + "step": 17980 + }, + { + "epoch": 0.77, + "grad_norm": 0.48046875, + "learning_rate": 0.0001362874759306958, + "loss": 2.6356, + "step": 17981 + }, + { + "epoch": 0.77, + "grad_norm": 0.423828125, + "learning_rate": 0.00013628122514409456, + "loss": 2.5308, + "step": 17982 + }, + { + "epoch": 0.77, + "grad_norm": 0.435546875, + "learning_rate": 0.00013627497419424104, + "loss": 2.5057, + "step": 17983 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.00013626872308116346, + "loss": 2.6484, + "step": 17984 + }, + { + "epoch": 0.77, + "grad_norm": 0.447265625, + "learning_rate": 0.0001362624718048898, + "loss": 2.5927, + "step": 17985 + }, + { + "epoch": 0.77, + "grad_norm": 0.47265625, + "learning_rate": 0.00013625622036544832, + "loss": 2.6331, + "step": 17986 + }, + { + "epoch": 0.77, + "grad_norm": 0.435546875, + "learning_rate": 0.0001362499687628671, + "loss": 2.7084, + "step": 17987 + }, + { + "epoch": 0.77, + "grad_norm": 0.421875, + "learning_rate": 0.00013624371699717426, + "loss": 2.534, + "step": 17988 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.00013623746506839793, + "loss": 2.4843, + "step": 17989 + }, + { + "epoch": 0.77, + "grad_norm": 0.45703125, + "learning_rate": 0.00013623121297656624, + "loss": 2.5997, + "step": 17990 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.00013622496072170735, + "loss": 2.445, + "step": 17991 + }, + { + "epoch": 0.77, + "grad_norm": 0.49609375, + "learning_rate": 0.0001362187083038494, + "loss": 2.7625, + "step": 17992 + }, + { + "epoch": 0.77, + "grad_norm": 0.427734375, + "learning_rate": 0.00013621245572302044, + "loss": 2.4852, + "step": 17993 + }, + { + "epoch": 0.77, + "grad_norm": 0.484375, + "learning_rate": 0.00013620620297924871, + "loss": 2.6083, + "step": 17994 + }, + { + "epoch": 0.77, + "grad_norm": 0.4765625, + "learning_rate": 0.00013619995007256228, + "loss": 2.6076, + "step": 17995 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.00013619369700298932, + "loss": 2.6154, + "step": 17996 + }, + { + "epoch": 0.77, + "grad_norm": 0.419921875, + "learning_rate": 0.00013618744377055794, + "loss": 2.4968, + "step": 17997 + }, + { + "epoch": 0.77, + "grad_norm": 0.4375, + "learning_rate": 0.00013618119037529629, + "loss": 2.4502, + "step": 17998 + }, + { + "epoch": 0.77, + "grad_norm": 0.4921875, + "learning_rate": 0.00013617493681723249, + "loss": 2.4884, + "step": 17999 + }, + { + "epoch": 0.77, + "grad_norm": 0.55859375, + "learning_rate": 0.00013616868309639473, + "loss": 2.5312, + "step": 18000 + }, + { + "epoch": 0.77, + "grad_norm": 0.416015625, + "learning_rate": 0.00013616242921281107, + "loss": 2.4535, + "step": 18001 + }, + { + "epoch": 0.77, + "grad_norm": 0.5234375, + "learning_rate": 0.00013615617516650973, + "loss": 2.6231, + "step": 18002 + }, + { + "epoch": 0.77, + "grad_norm": 0.447265625, + "learning_rate": 0.00013614992095751878, + "loss": 2.411, + "step": 18003 + }, + { + "epoch": 0.77, + "grad_norm": 0.5, + "learning_rate": 0.00013614366658586645, + "loss": 2.5266, + "step": 18004 + }, + { + "epoch": 0.77, + "grad_norm": 0.421875, + "learning_rate": 0.0001361374120515808, + "loss": 2.6288, + "step": 18005 + }, + { + "epoch": 0.77, + "grad_norm": 0.5078125, + "learning_rate": 0.00013613115735469, + "loss": 2.5851, + "step": 18006 + }, + { + "epoch": 0.77, + "grad_norm": 0.4609375, + "learning_rate": 0.0001361249024952222, + "loss": 2.5218, + "step": 18007 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013611864747320554, + "loss": 2.4975, + "step": 18008 + }, + { + "epoch": 0.77, + "grad_norm": 0.5234375, + "learning_rate": 0.00013611239228866816, + "loss": 2.572, + "step": 18009 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013610613694163822, + "loss": 2.6258, + "step": 18010 + }, + { + "epoch": 0.77, + "grad_norm": 0.482421875, + "learning_rate": 0.00013609988143214384, + "loss": 2.71, + "step": 18011 + }, + { + "epoch": 0.77, + "grad_norm": 0.451171875, + "learning_rate": 0.00013609362576021323, + "loss": 2.4902, + "step": 18012 + }, + { + "epoch": 0.77, + "grad_norm": 0.52734375, + "learning_rate": 0.00013608736992587445, + "loss": 2.729, + "step": 18013 + }, + { + "epoch": 0.77, + "grad_norm": 0.443359375, + "learning_rate": 0.00013608111392915571, + "loss": 2.5598, + "step": 18014 + }, + { + "epoch": 0.77, + "grad_norm": 0.494140625, + "learning_rate": 0.00013607485777008513, + "loss": 2.3894, + "step": 18015 + }, + { + "epoch": 0.77, + "grad_norm": 0.490234375, + "learning_rate": 0.0001360686014486909, + "loss": 2.5512, + "step": 18016 + }, + { + "epoch": 0.77, + "grad_norm": 0.447265625, + "learning_rate": 0.00013606234496500114, + "loss": 2.4995, + "step": 18017 + }, + { + "epoch": 0.77, + "grad_norm": 0.3984375, + "learning_rate": 0.00013605608831904399, + "loss": 2.4719, + "step": 18018 + }, + { + "epoch": 0.77, + "grad_norm": 0.416015625, + "learning_rate": 0.00013604983151084766, + "loss": 2.6005, + "step": 18019 + }, + { + "epoch": 0.77, + "grad_norm": 0.41796875, + "learning_rate": 0.0001360435745404402, + "loss": 2.5795, + "step": 18020 + }, + { + "epoch": 0.77, + "grad_norm": 0.431640625, + "learning_rate": 0.00013603731740784988, + "loss": 2.59, + "step": 18021 + }, + { + "epoch": 0.77, + "grad_norm": 1.34375, + "learning_rate": 0.0001360310601131048, + "loss": 2.5771, + "step": 18022 + }, + { + "epoch": 0.77, + "grad_norm": 0.46484375, + "learning_rate": 0.00013602480265623306, + "loss": 2.5576, + "step": 18023 + }, + { + "epoch": 0.77, + "grad_norm": 0.458984375, + "learning_rate": 0.00013601854503726295, + "loss": 2.6059, + "step": 18024 + }, + { + "epoch": 0.77, + "grad_norm": 0.412109375, + "learning_rate": 0.00013601228725622247, + "loss": 2.4683, + "step": 18025 + }, + { + "epoch": 0.77, + "grad_norm": 0.494140625, + "learning_rate": 0.00013600602931313993, + "loss": 2.5687, + "step": 18026 + }, + { + "epoch": 0.77, + "grad_norm": 0.47265625, + "learning_rate": 0.0001359997712080434, + "loss": 2.4701, + "step": 18027 + }, + { + "epoch": 0.77, + "grad_norm": 0.48046875, + "learning_rate": 0.00013599351294096102, + "loss": 2.6674, + "step": 18028 + }, + { + "epoch": 0.77, + "grad_norm": 0.396484375, + "learning_rate": 0.00013598725451192102, + "loss": 2.5984, + "step": 18029 + }, + { + "epoch": 0.77, + "grad_norm": 0.52734375, + "learning_rate": 0.00013598099592095154, + "loss": 2.6855, + "step": 18030 + }, + { + "epoch": 0.77, + "grad_norm": 0.45703125, + "learning_rate": 0.00013597473716808067, + "loss": 2.5763, + "step": 18031 + }, + { + "epoch": 0.77, + "grad_norm": 0.435546875, + "learning_rate": 0.00013596847825333668, + "loss": 2.571, + "step": 18032 + }, + { + "epoch": 0.77, + "grad_norm": 0.447265625, + "learning_rate": 0.00013596221917674766, + "loss": 2.6344, + "step": 18033 + }, + { + "epoch": 0.77, + "grad_norm": 0.484375, + "learning_rate": 0.0001359559599383418, + "loss": 2.6241, + "step": 18034 + }, + { + "epoch": 0.77, + "grad_norm": 0.4609375, + "learning_rate": 0.00013594970053814728, + "loss": 2.6443, + "step": 18035 + }, + { + "epoch": 0.77, + "grad_norm": 0.421875, + "learning_rate": 0.00013594344097619222, + "loss": 2.4599, + "step": 18036 + }, + { + "epoch": 0.77, + "grad_norm": 0.49609375, + "learning_rate": 0.00013593718125250482, + "loss": 2.6726, + "step": 18037 + }, + { + "epoch": 0.77, + "grad_norm": 0.46875, + "learning_rate": 0.00013593092136711326, + "loss": 2.6933, + "step": 18038 + }, + { + "epoch": 0.77, + "grad_norm": 0.5234375, + "learning_rate": 0.00013592466132004565, + "loss": 2.7767, + "step": 18039 + }, + { + "epoch": 0.77, + "grad_norm": 0.46875, + "learning_rate": 0.00013591840111133022, + "loss": 2.5982, + "step": 18040 + }, + { + "epoch": 0.77, + "grad_norm": 0.408203125, + "learning_rate": 0.0001359121407409951, + "loss": 2.5159, + "step": 18041 + }, + { + "epoch": 0.77, + "grad_norm": 0.4140625, + "learning_rate": 0.00013590588020906847, + "loss": 2.4841, + "step": 18042 + }, + { + "epoch": 0.77, + "grad_norm": 0.46875, + "learning_rate": 0.0001358996195155785, + "loss": 2.7358, + "step": 18043 + }, + { + "epoch": 0.77, + "grad_norm": 0.458984375, + "learning_rate": 0.00013589335866055337, + "loss": 2.5418, + "step": 18044 + }, + { + "epoch": 0.77, + "grad_norm": 0.5234375, + "learning_rate": 0.00013588709764402124, + "loss": 2.5339, + "step": 18045 + }, + { + "epoch": 0.77, + "grad_norm": 0.412109375, + "learning_rate": 0.0001358808364660103, + "loss": 2.5933, + "step": 18046 + }, + { + "epoch": 0.77, + "grad_norm": 0.427734375, + "learning_rate": 0.00013587457512654867, + "loss": 2.6478, + "step": 18047 + }, + { + "epoch": 0.77, + "grad_norm": 0.453125, + "learning_rate": 0.0001358683136256646, + "loss": 2.5529, + "step": 18048 + }, + { + "epoch": 0.77, + "grad_norm": 0.5078125, + "learning_rate": 0.0001358620519633862, + "loss": 2.576, + "step": 18049 + }, + { + "epoch": 0.77, + "grad_norm": 0.482421875, + "learning_rate": 0.0001358557901397417, + "loss": 2.502, + "step": 18050 + }, + { + "epoch": 0.77, + "grad_norm": 0.484375, + "learning_rate": 0.0001358495281547592, + "loss": 2.4733, + "step": 18051 + }, + { + "epoch": 0.77, + "grad_norm": 0.49609375, + "learning_rate": 0.00013584326600846698, + "loss": 2.515, + "step": 18052 + }, + { + "epoch": 0.77, + "grad_norm": 0.42578125, + "learning_rate": 0.00013583700370089314, + "loss": 2.5921, + "step": 18053 + }, + { + "epoch": 0.77, + "grad_norm": 0.458984375, + "learning_rate": 0.00013583074123206588, + "loss": 2.5848, + "step": 18054 + }, + { + "epoch": 0.77, + "grad_norm": 0.439453125, + "learning_rate": 0.00013582447860201335, + "loss": 2.4944, + "step": 18055 + }, + { + "epoch": 0.77, + "grad_norm": 0.48828125, + "learning_rate": 0.00013581821581076382, + "loss": 2.6018, + "step": 18056 + }, + { + "epoch": 0.77, + "grad_norm": 0.4375, + "learning_rate": 0.00013581195285834537, + "loss": 2.5336, + "step": 18057 + }, + { + "epoch": 0.77, + "grad_norm": 0.490234375, + "learning_rate": 0.00013580568974478623, + "loss": 2.5635, + "step": 18058 + }, + { + "epoch": 0.77, + "grad_norm": 0.466796875, + "learning_rate": 0.00013579942647011457, + "loss": 2.5106, + "step": 18059 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.00013579316303435855, + "loss": 2.6019, + "step": 18060 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.0001357868994375464, + "loss": 2.4252, + "step": 18061 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.00013578063567970628, + "loss": 2.588, + "step": 18062 + }, + { + "epoch": 0.77, + "grad_norm": 0.4375, + "learning_rate": 0.00013577437176086639, + "loss": 2.5025, + "step": 18063 + }, + { + "epoch": 0.77, + "grad_norm": 0.490234375, + "learning_rate": 0.00013576810768105491, + "loss": 2.5286, + "step": 18064 + }, + { + "epoch": 0.77, + "grad_norm": 0.4921875, + "learning_rate": 0.00013576184344029997, + "loss": 2.5772, + "step": 18065 + }, + { + "epoch": 0.77, + "grad_norm": 0.56640625, + "learning_rate": 0.00013575557903862983, + "loss": 2.6447, + "step": 18066 + }, + { + "epoch": 0.77, + "grad_norm": 0.5390625, + "learning_rate": 0.00013574931447607266, + "loss": 2.5603, + "step": 18067 + }, + { + "epoch": 0.77, + "grad_norm": 0.5078125, + "learning_rate": 0.0001357430497526566, + "loss": 2.6242, + "step": 18068 + }, + { + "epoch": 0.77, + "grad_norm": 0.443359375, + "learning_rate": 0.00013573678486840992, + "loss": 2.4404, + "step": 18069 + }, + { + "epoch": 0.77, + "grad_norm": 0.50390625, + "learning_rate": 0.00013573051982336076, + "loss": 2.6279, + "step": 18070 + }, + { + "epoch": 0.77, + "grad_norm": 0.4140625, + "learning_rate": 0.00013572425461753733, + "loss": 2.5641, + "step": 18071 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.00013571798925096778, + "loss": 2.6414, + "step": 18072 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013571172372368038, + "loss": 2.5619, + "step": 18073 + }, + { + "epoch": 0.77, + "grad_norm": 0.458984375, + "learning_rate": 0.0001357054580357032, + "loss": 2.5319, + "step": 18074 + }, + { + "epoch": 0.77, + "grad_norm": 0.423828125, + "learning_rate": 0.00013569919218706458, + "loss": 2.7564, + "step": 18075 + }, + { + "epoch": 0.77, + "grad_norm": 0.4140625, + "learning_rate": 0.00013569292617779258, + "loss": 2.4341, + "step": 18076 + }, + { + "epoch": 0.77, + "grad_norm": 0.4765625, + "learning_rate": 0.0001356866600079155, + "loss": 2.5198, + "step": 18077 + }, + { + "epoch": 0.77, + "grad_norm": 0.47265625, + "learning_rate": 0.00013568039367746148, + "loss": 2.6879, + "step": 18078 + }, + { + "epoch": 0.77, + "grad_norm": 0.466796875, + "learning_rate": 0.00013567412718645873, + "loss": 2.51, + "step": 18079 + }, + { + "epoch": 0.77, + "grad_norm": 0.41796875, + "learning_rate": 0.00013566786053493545, + "loss": 2.5282, + "step": 18080 + }, + { + "epoch": 0.77, + "grad_norm": 0.462890625, + "learning_rate": 0.00013566159372291981, + "loss": 2.6689, + "step": 18081 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013565532675044003, + "loss": 2.7017, + "step": 18082 + }, + { + "epoch": 0.77, + "grad_norm": 0.498046875, + "learning_rate": 0.00013564905961752435, + "loss": 2.6295, + "step": 18083 + }, + { + "epoch": 0.77, + "grad_norm": 0.486328125, + "learning_rate": 0.0001356427923242009, + "loss": 2.6059, + "step": 18084 + }, + { + "epoch": 0.77, + "grad_norm": 0.453125, + "learning_rate": 0.0001356365248704979, + "loss": 2.6134, + "step": 18085 + }, + { + "epoch": 0.77, + "grad_norm": 0.515625, + "learning_rate": 0.0001356302572564436, + "loss": 2.5966, + "step": 18086 + }, + { + "epoch": 0.77, + "grad_norm": 0.474609375, + "learning_rate": 0.00013562398948206616, + "loss": 2.5478, + "step": 18087 + }, + { + "epoch": 0.77, + "grad_norm": 0.484375, + "learning_rate": 0.00013561772154739376, + "loss": 2.4844, + "step": 18088 + }, + { + "epoch": 0.77, + "grad_norm": 0.45703125, + "learning_rate": 0.00013561145345245462, + "loss": 2.5745, + "step": 18089 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.00013560518519727696, + "loss": 2.6971, + "step": 18090 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.00013559891678188902, + "loss": 2.5493, + "step": 18091 + }, + { + "epoch": 0.77, + "grad_norm": 0.44921875, + "learning_rate": 0.00013559264820631891, + "loss": 2.5468, + "step": 18092 + }, + { + "epoch": 0.77, + "grad_norm": 0.451171875, + "learning_rate": 0.00013558637947059493, + "loss": 2.4543, + "step": 18093 + }, + { + "epoch": 0.77, + "grad_norm": 0.5, + "learning_rate": 0.00013558011057474524, + "loss": 2.72, + "step": 18094 + }, + { + "epoch": 0.77, + "grad_norm": 0.427734375, + "learning_rate": 0.00013557384151879806, + "loss": 2.5636, + "step": 18095 + }, + { + "epoch": 0.77, + "grad_norm": 0.48828125, + "learning_rate": 0.0001355675723027816, + "loss": 2.6135, + "step": 18096 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013556130292672401, + "loss": 2.4588, + "step": 18097 + }, + { + "epoch": 0.77, + "grad_norm": 0.4296875, + "learning_rate": 0.0001355550333906536, + "loss": 2.589, + "step": 18098 + }, + { + "epoch": 0.77, + "grad_norm": 0.5078125, + "learning_rate": 0.00013554876369459854, + "loss": 2.6074, + "step": 18099 + }, + { + "epoch": 0.77, + "grad_norm": 0.451171875, + "learning_rate": 0.00013554249383858702, + "loss": 2.5034, + "step": 18100 + }, + { + "epoch": 0.77, + "grad_norm": 0.46484375, + "learning_rate": 0.00013553622382264725, + "loss": 2.5781, + "step": 18101 + }, + { + "epoch": 0.77, + "grad_norm": 0.47265625, + "learning_rate": 0.00013552995364680747, + "loss": 2.5689, + "step": 18102 + }, + { + "epoch": 0.77, + "grad_norm": 0.45703125, + "learning_rate": 0.0001355236833110959, + "loss": 2.4809, + "step": 18103 + }, + { + "epoch": 0.77, + "grad_norm": 0.46875, + "learning_rate": 0.00013551741281554074, + "loss": 2.5962, + "step": 18104 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.0001355111421601702, + "loss": 2.6741, + "step": 18105 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.00013550487134501244, + "loss": 2.5207, + "step": 18106 + }, + { + "epoch": 0.77, + "grad_norm": 0.41015625, + "learning_rate": 0.0001354986003700958, + "loss": 2.4678, + "step": 18107 + }, + { + "epoch": 0.77, + "grad_norm": 0.462890625, + "learning_rate": 0.00013549232923544842, + "loss": 2.6311, + "step": 18108 + }, + { + "epoch": 0.77, + "grad_norm": 0.484375, + "learning_rate": 0.0001354860579410985, + "loss": 2.5946, + "step": 18109 + }, + { + "epoch": 0.77, + "grad_norm": 0.435546875, + "learning_rate": 0.00013547978648707431, + "loss": 2.5362, + "step": 18110 + }, + { + "epoch": 0.77, + "grad_norm": 0.416015625, + "learning_rate": 0.00013547351487340405, + "loss": 2.6316, + "step": 18111 + }, + { + "epoch": 0.77, + "grad_norm": 0.50390625, + "learning_rate": 0.00013546724310011595, + "loss": 2.6607, + "step": 18112 + }, + { + "epoch": 0.77, + "grad_norm": 0.451171875, + "learning_rate": 0.0001354609711672382, + "loss": 2.6924, + "step": 18113 + }, + { + "epoch": 0.77, + "grad_norm": 0.4140625, + "learning_rate": 0.00013545469907479905, + "loss": 2.5285, + "step": 18114 + }, + { + "epoch": 0.77, + "grad_norm": 0.484375, + "learning_rate": 0.0001354484268228267, + "loss": 2.5754, + "step": 18115 + }, + { + "epoch": 0.77, + "grad_norm": 0.384765625, + "learning_rate": 0.00013544215441134937, + "loss": 2.4331, + "step": 18116 + }, + { + "epoch": 0.77, + "grad_norm": 0.48046875, + "learning_rate": 0.00013543588184039532, + "loss": 2.5481, + "step": 18117 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.00013542960910999273, + "loss": 2.6081, + "step": 18118 + }, + { + "epoch": 0.77, + "grad_norm": 0.451171875, + "learning_rate": 0.0001354233362201699, + "loss": 2.5619, + "step": 18119 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013541706317095494, + "loss": 2.6291, + "step": 18120 + }, + { + "epoch": 0.77, + "grad_norm": 0.55859375, + "learning_rate": 0.00013541078996237617, + "loss": 2.5651, + "step": 18121 + }, + { + "epoch": 0.77, + "grad_norm": 0.4296875, + "learning_rate": 0.00013540451659446178, + "loss": 2.4918, + "step": 18122 + }, + { + "epoch": 0.77, + "grad_norm": 0.46875, + "learning_rate": 0.00013539824306724004, + "loss": 2.6218, + "step": 18123 + }, + { + "epoch": 0.77, + "grad_norm": 0.5234375, + "learning_rate": 0.00013539196938073907, + "loss": 2.5585, + "step": 18124 + }, + { + "epoch": 0.77, + "grad_norm": 0.439453125, + "learning_rate": 0.0001353856955349872, + "loss": 2.5547, + "step": 18125 + }, + { + "epoch": 0.77, + "grad_norm": 0.453125, + "learning_rate": 0.00013537942153001267, + "loss": 2.7013, + "step": 18126 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013537314736584367, + "loss": 2.4415, + "step": 18127 + }, + { + "epoch": 0.77, + "grad_norm": 0.4609375, + "learning_rate": 0.0001353668730425084, + "loss": 2.55, + "step": 18128 + }, + { + "epoch": 0.77, + "grad_norm": 0.435546875, + "learning_rate": 0.00013536059856003515, + "loss": 2.4478, + "step": 18129 + }, + { + "epoch": 0.77, + "grad_norm": 0.466796875, + "learning_rate": 0.00013535432391845208, + "loss": 2.5445, + "step": 18130 + }, + { + "epoch": 0.77, + "grad_norm": 0.462890625, + "learning_rate": 0.00013534804911778753, + "loss": 2.6335, + "step": 18131 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.00013534177415806963, + "loss": 2.5485, + "step": 18132 + }, + { + "epoch": 0.77, + "grad_norm": 0.443359375, + "learning_rate": 0.0001353354990393267, + "loss": 2.557, + "step": 18133 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013532922376158696, + "loss": 2.6005, + "step": 18134 + }, + { + "epoch": 0.77, + "grad_norm": 0.462890625, + "learning_rate": 0.00013532294832487856, + "loss": 2.5328, + "step": 18135 + }, + { + "epoch": 0.77, + "grad_norm": 0.46875, + "learning_rate": 0.00013531667272922984, + "loss": 2.5388, + "step": 18136 + }, + { + "epoch": 0.77, + "grad_norm": 0.44140625, + "learning_rate": 0.000135310396974669, + "loss": 2.4688, + "step": 18137 + }, + { + "epoch": 0.77, + "grad_norm": 0.5, + "learning_rate": 0.00013530412106122424, + "loss": 2.5121, + "step": 18138 + }, + { + "epoch": 0.77, + "grad_norm": 0.4375, + "learning_rate": 0.0001352978449889239, + "loss": 2.5631, + "step": 18139 + }, + { + "epoch": 0.77, + "grad_norm": 0.482421875, + "learning_rate": 0.0001352915687577961, + "loss": 2.6315, + "step": 18140 + }, + { + "epoch": 0.77, + "grad_norm": 0.421875, + "learning_rate": 0.0001352852923678692, + "loss": 2.5067, + "step": 18141 + }, + { + "epoch": 0.77, + "grad_norm": 0.447265625, + "learning_rate": 0.0001352790158191713, + "loss": 2.5952, + "step": 18142 + }, + { + "epoch": 0.77, + "grad_norm": 0.4765625, + "learning_rate": 0.00013527273911173078, + "loss": 2.6391, + "step": 18143 + }, + { + "epoch": 0.77, + "grad_norm": 0.435546875, + "learning_rate": 0.00013526646224557582, + "loss": 2.6209, + "step": 18144 + }, + { + "epoch": 0.77, + "grad_norm": 0.46875, + "learning_rate": 0.00013526018522073464, + "loss": 2.6349, + "step": 18145 + }, + { + "epoch": 0.77, + "grad_norm": 0.474609375, + "learning_rate": 0.00013525390803723552, + "loss": 2.5726, + "step": 18146 + }, + { + "epoch": 0.77, + "grad_norm": 0.435546875, + "learning_rate": 0.00013524763069510672, + "loss": 2.5361, + "step": 18147 + }, + { + "epoch": 0.77, + "grad_norm": 0.443359375, + "learning_rate": 0.00013524135319437644, + "loss": 2.6351, + "step": 18148 + }, + { + "epoch": 0.77, + "grad_norm": 0.46484375, + "learning_rate": 0.00013523507553507298, + "loss": 2.4771, + "step": 18149 + }, + { + "epoch": 0.77, + "grad_norm": 0.388671875, + "learning_rate": 0.00013522879771722454, + "loss": 2.4657, + "step": 18150 + }, + { + "epoch": 0.77, + "grad_norm": 0.37890625, + "learning_rate": 0.00013522251974085937, + "loss": 2.469, + "step": 18151 + }, + { + "epoch": 0.77, + "grad_norm": 0.46484375, + "learning_rate": 0.00013521624160600575, + "loss": 2.4673, + "step": 18152 + }, + { + "epoch": 0.77, + "grad_norm": 0.474609375, + "learning_rate": 0.00013520996331269194, + "loss": 2.579, + "step": 18153 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.0001352036848609461, + "loss": 2.5725, + "step": 18154 + }, + { + "epoch": 0.77, + "grad_norm": 0.474609375, + "learning_rate": 0.0001351974062507966, + "loss": 2.6523, + "step": 18155 + }, + { + "epoch": 0.77, + "grad_norm": 0.451171875, + "learning_rate": 0.0001351911274822716, + "loss": 2.6478, + "step": 18156 + }, + { + "epoch": 0.77, + "grad_norm": 0.474609375, + "learning_rate": 0.0001351848485553994, + "loss": 2.4935, + "step": 18157 + }, + { + "epoch": 0.77, + "grad_norm": 0.365234375, + "learning_rate": 0.00013517856947020823, + "loss": 2.5291, + "step": 18158 + }, + { + "epoch": 0.77, + "grad_norm": 0.4765625, + "learning_rate": 0.00013517229022672639, + "loss": 2.4679, + "step": 18159 + }, + { + "epoch": 0.77, + "grad_norm": 0.490234375, + "learning_rate": 0.00013516601082498206, + "loss": 2.6172, + "step": 18160 + }, + { + "epoch": 0.77, + "grad_norm": 0.46484375, + "learning_rate": 0.00013515973126500353, + "loss": 2.6126, + "step": 18161 + }, + { + "epoch": 0.77, + "grad_norm": 0.5, + "learning_rate": 0.00013515345154681903, + "loss": 2.6072, + "step": 18162 + }, + { + "epoch": 0.77, + "grad_norm": 0.455078125, + "learning_rate": 0.00013514717167045694, + "loss": 2.546, + "step": 18163 + }, + { + "epoch": 0.77, + "grad_norm": 0.388671875, + "learning_rate": 0.00013514089163594534, + "loss": 2.5975, + "step": 18164 + }, + { + "epoch": 0.77, + "grad_norm": 0.44140625, + "learning_rate": 0.00013513461144331262, + "loss": 2.6472, + "step": 18165 + }, + { + "epoch": 0.77, + "grad_norm": 0.50390625, + "learning_rate": 0.00013512833109258695, + "loss": 2.6575, + "step": 18166 + }, + { + "epoch": 0.77, + "grad_norm": 0.466796875, + "learning_rate": 0.00013512205058379664, + "loss": 2.599, + "step": 18167 + }, + { + "epoch": 0.77, + "grad_norm": 0.44921875, + "learning_rate": 0.00013511576991696996, + "loss": 2.565, + "step": 18168 + }, + { + "epoch": 0.77, + "grad_norm": 0.45703125, + "learning_rate": 0.00013510948909213516, + "loss": 2.516, + "step": 18169 + }, + { + "epoch": 0.77, + "grad_norm": 0.388671875, + "learning_rate": 0.0001351032081093204, + "loss": 2.469, + "step": 18170 + }, + { + "epoch": 0.77, + "grad_norm": 0.451171875, + "learning_rate": 0.00013509692696855413, + "loss": 2.5475, + "step": 18171 + }, + { + "epoch": 0.77, + "grad_norm": 0.439453125, + "learning_rate": 0.00013509064566986448, + "loss": 2.6138, + "step": 18172 + }, + { + "epoch": 0.77, + "grad_norm": 0.5234375, + "learning_rate": 0.00013508436421327976, + "loss": 2.6757, + "step": 18173 + }, + { + "epoch": 0.77, + "grad_norm": 0.51953125, + "learning_rate": 0.0001350780825988282, + "loss": 2.5155, + "step": 18174 + }, + { + "epoch": 0.77, + "grad_norm": 0.470703125, + "learning_rate": 0.0001350718008265381, + "loss": 2.4836, + "step": 18175 + }, + { + "epoch": 0.77, + "grad_norm": 0.453125, + "learning_rate": 0.00013506551889643772, + "loss": 2.5991, + "step": 18176 + }, + { + "epoch": 0.77, + "grad_norm": 0.51171875, + "learning_rate": 0.00013505923680855533, + "loss": 2.5871, + "step": 18177 + }, + { + "epoch": 0.77, + "grad_norm": 0.41796875, + "learning_rate": 0.00013505295456291914, + "loss": 2.482, + "step": 18178 + }, + { + "epoch": 0.77, + "grad_norm": 0.5078125, + "learning_rate": 0.0001350466721595575, + "loss": 2.5699, + "step": 18179 + }, + { + "epoch": 0.77, + "grad_norm": 0.4453125, + "learning_rate": 0.00013504038959849865, + "loss": 2.5966, + "step": 18180 + }, + { + "epoch": 0.77, + "grad_norm": 0.478515625, + "learning_rate": 0.00013503410687977082, + "loss": 2.7527, + "step": 18181 + }, + { + "epoch": 0.77, + "grad_norm": 0.46875, + "learning_rate": 0.00013502782400340237, + "loss": 2.678, + "step": 18182 + }, + { + "epoch": 0.77, + "grad_norm": 0.447265625, + "learning_rate": 0.00013502154096942145, + "loss": 2.615, + "step": 18183 + }, + { + "epoch": 0.77, + "grad_norm": 0.5, + "learning_rate": 0.00013501525777785643, + "loss": 2.678, + "step": 18184 + }, + { + "epoch": 0.77, + "grad_norm": 0.46484375, + "learning_rate": 0.00013500897442873556, + "loss": 2.5082, + "step": 18185 + }, + { + "epoch": 0.77, + "grad_norm": 0.478515625, + "learning_rate": 0.00013500269092208707, + "loss": 2.5395, + "step": 18186 + }, + { + "epoch": 0.77, + "grad_norm": 0.42578125, + "learning_rate": 0.00013499640725793928, + "loss": 2.4535, + "step": 18187 + }, + { + "epoch": 0.78, + "grad_norm": 0.52734375, + "learning_rate": 0.00013499012343632042, + "loss": 2.7079, + "step": 18188 + }, + { + "epoch": 0.78, + "grad_norm": 0.484375, + "learning_rate": 0.00013498383945725878, + "loss": 2.6482, + "step": 18189 + }, + { + "epoch": 0.78, + "grad_norm": 0.46875, + "learning_rate": 0.0001349775553207827, + "loss": 2.5207, + "step": 18190 + }, + { + "epoch": 0.78, + "grad_norm": 0.458984375, + "learning_rate": 0.00013497127102692037, + "loss": 2.5867, + "step": 18191 + }, + { + "epoch": 0.78, + "grad_norm": 0.46484375, + "learning_rate": 0.00013496498657570013, + "loss": 2.537, + "step": 18192 + }, + { + "epoch": 0.78, + "grad_norm": 0.4375, + "learning_rate": 0.0001349587019671502, + "loss": 2.6279, + "step": 18193 + }, + { + "epoch": 0.78, + "grad_norm": 0.462890625, + "learning_rate": 0.0001349524172012989, + "loss": 2.6128, + "step": 18194 + }, + { + "epoch": 0.78, + "grad_norm": 0.447265625, + "learning_rate": 0.0001349461322781745, + "loss": 2.5856, + "step": 18195 + }, + { + "epoch": 0.78, + "grad_norm": 0.52734375, + "learning_rate": 0.00013493984719780528, + "loss": 2.7483, + "step": 18196 + }, + { + "epoch": 0.78, + "grad_norm": 0.400390625, + "learning_rate": 0.0001349335619602195, + "loss": 2.5456, + "step": 18197 + }, + { + "epoch": 0.78, + "grad_norm": 0.404296875, + "learning_rate": 0.00013492727656544545, + "loss": 2.5763, + "step": 18198 + }, + { + "epoch": 0.78, + "grad_norm": 0.435546875, + "learning_rate": 0.00013492099101351142, + "loss": 2.4204, + "step": 18199 + }, + { + "epoch": 0.78, + "grad_norm": 0.470703125, + "learning_rate": 0.0001349147053044457, + "loss": 2.6267, + "step": 18200 + }, + { + "epoch": 0.78, + "grad_norm": 0.4765625, + "learning_rate": 0.00013490841943827658, + "loss": 2.5615, + "step": 18201 + }, + { + "epoch": 0.78, + "grad_norm": 0.400390625, + "learning_rate": 0.00013490213341503236, + "loss": 2.551, + "step": 18202 + }, + { + "epoch": 0.78, + "grad_norm": 0.41015625, + "learning_rate": 0.00013489584723474124, + "loss": 2.5939, + "step": 18203 + }, + { + "epoch": 0.78, + "grad_norm": 0.455078125, + "learning_rate": 0.0001348895608974316, + "loss": 2.6287, + "step": 18204 + }, + { + "epoch": 0.78, + "grad_norm": 0.5, + "learning_rate": 0.00013488327440313165, + "loss": 2.4997, + "step": 18205 + }, + { + "epoch": 0.78, + "grad_norm": 0.46484375, + "learning_rate": 0.00013487698775186972, + "loss": 2.544, + "step": 18206 + }, + { + "epoch": 0.78, + "grad_norm": 0.462890625, + "learning_rate": 0.00013487070094367408, + "loss": 2.5995, + "step": 18207 + }, + { + "epoch": 0.78, + "grad_norm": 0.431640625, + "learning_rate": 0.00013486441397857308, + "loss": 2.7022, + "step": 18208 + }, + { + "epoch": 0.78, + "grad_norm": 0.4453125, + "learning_rate": 0.0001348581268565949, + "loss": 2.391, + "step": 18209 + }, + { + "epoch": 0.78, + "grad_norm": 0.51171875, + "learning_rate": 0.00013485183957776794, + "loss": 2.5474, + "step": 18210 + }, + { + "epoch": 0.78, + "grad_norm": 0.447265625, + "learning_rate": 0.0001348455521421204, + "loss": 2.6824, + "step": 18211 + }, + { + "epoch": 0.78, + "grad_norm": 0.470703125, + "learning_rate": 0.00013483926454968062, + "loss": 2.5255, + "step": 18212 + }, + { + "epoch": 0.78, + "grad_norm": 0.5390625, + "learning_rate": 0.0001348329768004769, + "loss": 2.5408, + "step": 18213 + }, + { + "epoch": 0.78, + "grad_norm": 0.478515625, + "learning_rate": 0.00013482668889453752, + "loss": 2.5513, + "step": 18214 + }, + { + "epoch": 0.78, + "grad_norm": 0.474609375, + "learning_rate": 0.00013482040083189072, + "loss": 2.7212, + "step": 18215 + }, + { + "epoch": 0.78, + "grad_norm": 0.50390625, + "learning_rate": 0.00013481411261256486, + "loss": 2.5903, + "step": 18216 + }, + { + "epoch": 0.78, + "grad_norm": 0.431640625, + "learning_rate": 0.00013480782423658824, + "loss": 2.6805, + "step": 18217 + }, + { + "epoch": 0.78, + "grad_norm": 0.431640625, + "learning_rate": 0.00013480153570398913, + "loss": 2.4999, + "step": 18218 + }, + { + "epoch": 0.78, + "grad_norm": 0.400390625, + "learning_rate": 0.00013479524701479582, + "loss": 2.4668, + "step": 18219 + }, + { + "epoch": 0.78, + "grad_norm": 0.52734375, + "learning_rate": 0.0001347889581690366, + "loss": 2.6055, + "step": 18220 + }, + { + "epoch": 0.78, + "grad_norm": 0.474609375, + "learning_rate": 0.00013478266916673977, + "loss": 2.5864, + "step": 18221 + }, + { + "epoch": 0.78, + "grad_norm": 0.392578125, + "learning_rate": 0.0001347763800079337, + "loss": 2.4403, + "step": 18222 + }, + { + "epoch": 0.78, + "grad_norm": 0.5, + "learning_rate": 0.00013477009069264657, + "loss": 2.6514, + "step": 18223 + }, + { + "epoch": 0.78, + "grad_norm": 0.4921875, + "learning_rate": 0.00013476380122090677, + "loss": 2.6438, + "step": 18224 + }, + { + "epoch": 0.78, + "grad_norm": 0.447265625, + "learning_rate": 0.00013475751159274256, + "loss": 2.559, + "step": 18225 + }, + { + "epoch": 0.78, + "grad_norm": 0.455078125, + "learning_rate": 0.00013475122180818224, + "loss": 2.6018, + "step": 18226 + }, + { + "epoch": 0.78, + "grad_norm": 0.427734375, + "learning_rate": 0.00013474493186725413, + "loss": 2.6323, + "step": 18227 + }, + { + "epoch": 0.78, + "grad_norm": 0.421875, + "learning_rate": 0.00013473864176998657, + "loss": 2.5678, + "step": 18228 + }, + { + "epoch": 0.78, + "grad_norm": 0.431640625, + "learning_rate": 0.00013473235151640776, + "loss": 2.6796, + "step": 18229 + }, + { + "epoch": 0.78, + "grad_norm": 0.458984375, + "learning_rate": 0.0001347260611065461, + "loss": 2.6944, + "step": 18230 + }, + { + "epoch": 0.78, + "grad_norm": 0.462890625, + "learning_rate": 0.0001347197705404298, + "loss": 2.6513, + "step": 18231 + }, + { + "epoch": 0.78, + "grad_norm": 0.4921875, + "learning_rate": 0.00013471347981808726, + "loss": 2.672, + "step": 18232 + }, + { + "epoch": 0.78, + "grad_norm": 0.45703125, + "learning_rate": 0.00013470718893954673, + "loss": 2.6543, + "step": 18233 + }, + { + "epoch": 0.78, + "grad_norm": 0.44921875, + "learning_rate": 0.00013470089790483655, + "loss": 2.4585, + "step": 18234 + }, + { + "epoch": 0.78, + "grad_norm": 0.4375, + "learning_rate": 0.00013469460671398502, + "loss": 2.5386, + "step": 18235 + }, + { + "epoch": 0.78, + "grad_norm": 0.4609375, + "learning_rate": 0.00013468831536702042, + "loss": 2.5945, + "step": 18236 + }, + { + "epoch": 0.78, + "grad_norm": 0.494140625, + "learning_rate": 0.00013468202386397106, + "loss": 2.5906, + "step": 18237 + }, + { + "epoch": 0.78, + "grad_norm": 0.486328125, + "learning_rate": 0.00013467573220486532, + "loss": 2.6549, + "step": 18238 + }, + { + "epoch": 0.78, + "grad_norm": 0.453125, + "learning_rate": 0.0001346694403897314, + "loss": 2.5129, + "step": 18239 + }, + { + "epoch": 0.78, + "grad_norm": 0.515625, + "learning_rate": 0.0001346631484185977, + "loss": 2.7428, + "step": 18240 + }, + { + "epoch": 0.78, + "grad_norm": 0.494140625, + "learning_rate": 0.0001346568562914925, + "loss": 2.4974, + "step": 18241 + }, + { + "epoch": 0.78, + "grad_norm": 0.52734375, + "learning_rate": 0.0001346505640084441, + "loss": 2.5687, + "step": 18242 + }, + { + "epoch": 0.78, + "grad_norm": 0.5, + "learning_rate": 0.00013464427156948083, + "loss": 2.764, + "step": 18243 + }, + { + "epoch": 0.78, + "grad_norm": 0.48828125, + "learning_rate": 0.00013463797897463105, + "loss": 2.5651, + "step": 18244 + }, + { + "epoch": 0.78, + "grad_norm": 0.47265625, + "learning_rate": 0.00013463168622392292, + "loss": 2.4991, + "step": 18245 + }, + { + "epoch": 0.78, + "grad_norm": 0.5, + "learning_rate": 0.00013462539331738496, + "loss": 2.6571, + "step": 18246 + }, + { + "epoch": 0.78, + "grad_norm": 0.47265625, + "learning_rate": 0.0001346191002550453, + "loss": 2.5881, + "step": 18247 + }, + { + "epoch": 0.78, + "grad_norm": 0.52734375, + "learning_rate": 0.0001346128070369324, + "loss": 2.6245, + "step": 18248 + }, + { + "epoch": 0.78, + "grad_norm": 0.470703125, + "learning_rate": 0.0001346065136630745, + "loss": 2.6257, + "step": 18249 + }, + { + "epoch": 0.78, + "grad_norm": 0.400390625, + "learning_rate": 0.00013460022013349994, + "loss": 2.5458, + "step": 18250 + }, + { + "epoch": 0.78, + "grad_norm": 0.48046875, + "learning_rate": 0.00013459392644823702, + "loss": 2.5414, + "step": 18251 + }, + { + "epoch": 0.78, + "grad_norm": 0.439453125, + "learning_rate": 0.00013458763260731408, + "loss": 2.5401, + "step": 18252 + }, + { + "epoch": 0.78, + "grad_norm": 0.4609375, + "learning_rate": 0.00013458133861075943, + "loss": 2.4905, + "step": 18253 + }, + { + "epoch": 0.78, + "grad_norm": 0.494140625, + "learning_rate": 0.0001345750444586014, + "loss": 2.5442, + "step": 18254 + }, + { + "epoch": 0.78, + "grad_norm": 0.44140625, + "learning_rate": 0.0001345687501508683, + "loss": 2.5373, + "step": 18255 + }, + { + "epoch": 0.78, + "grad_norm": 0.4765625, + "learning_rate": 0.00013456245568758848, + "loss": 2.6846, + "step": 18256 + }, + { + "epoch": 0.78, + "grad_norm": 0.439453125, + "learning_rate": 0.0001345561610687902, + "loss": 2.5404, + "step": 18257 + }, + { + "epoch": 0.78, + "grad_norm": 0.46875, + "learning_rate": 0.00013454986629450185, + "loss": 2.5012, + "step": 18258 + }, + { + "epoch": 0.78, + "grad_norm": 0.44140625, + "learning_rate": 0.00013454357136475172, + "loss": 2.6617, + "step": 18259 + }, + { + "epoch": 0.78, + "grad_norm": 0.421875, + "learning_rate": 0.00013453727627956818, + "loss": 2.4888, + "step": 18260 + }, + { + "epoch": 0.78, + "grad_norm": 0.5078125, + "learning_rate": 0.00013453098103897946, + "loss": 2.6804, + "step": 18261 + }, + { + "epoch": 0.78, + "grad_norm": 0.4609375, + "learning_rate": 0.000134524685643014, + "loss": 2.5568, + "step": 18262 + }, + { + "epoch": 0.78, + "grad_norm": 0.435546875, + "learning_rate": 0.00013451839009170002, + "loss": 2.5785, + "step": 18263 + }, + { + "epoch": 0.78, + "grad_norm": 0.419921875, + "learning_rate": 0.00013451209438506594, + "loss": 2.5351, + "step": 18264 + }, + { + "epoch": 0.78, + "grad_norm": 0.40234375, + "learning_rate": 0.00013450579852314, + "loss": 2.6068, + "step": 18265 + }, + { + "epoch": 0.78, + "grad_norm": 0.46875, + "learning_rate": 0.00013449950250595064, + "loss": 2.7263, + "step": 18266 + }, + { + "epoch": 0.78, + "grad_norm": 0.48828125, + "learning_rate": 0.00013449320633352607, + "loss": 2.6164, + "step": 18267 + }, + { + "epoch": 0.78, + "grad_norm": 0.486328125, + "learning_rate": 0.00013448691000589473, + "loss": 2.7105, + "step": 18268 + }, + { + "epoch": 0.78, + "grad_norm": 0.392578125, + "learning_rate": 0.00013448061352308485, + "loss": 2.5909, + "step": 18269 + }, + { + "epoch": 0.78, + "grad_norm": 0.4140625, + "learning_rate": 0.00013447431688512484, + "loss": 2.4616, + "step": 18270 + }, + { + "epoch": 0.78, + "grad_norm": 0.408203125, + "learning_rate": 0.000134468020092043, + "loss": 2.5328, + "step": 18271 + }, + { + "epoch": 0.78, + "grad_norm": 0.4921875, + "learning_rate": 0.00013446172314386768, + "loss": 2.6576, + "step": 18272 + }, + { + "epoch": 0.78, + "grad_norm": 0.373046875, + "learning_rate": 0.0001344554260406272, + "loss": 2.5959, + "step": 18273 + }, + { + "epoch": 0.78, + "grad_norm": 0.443359375, + "learning_rate": 0.00013444912878234986, + "loss": 2.738, + "step": 18274 + }, + { + "epoch": 0.78, + "grad_norm": 0.5, + "learning_rate": 0.000134442831369064, + "loss": 2.5561, + "step": 18275 + }, + { + "epoch": 0.78, + "grad_norm": 0.52734375, + "learning_rate": 0.00013443653380079806, + "loss": 2.5814, + "step": 18276 + }, + { + "epoch": 0.78, + "grad_norm": 0.494140625, + "learning_rate": 0.00013443023607758025, + "loss": 2.5788, + "step": 18277 + }, + { + "epoch": 0.78, + "grad_norm": 0.48046875, + "learning_rate": 0.00013442393819943898, + "loss": 2.5822, + "step": 18278 + }, + { + "epoch": 0.78, + "grad_norm": 0.46875, + "learning_rate": 0.0001344176401664026, + "loss": 2.5421, + "step": 18279 + }, + { + "epoch": 0.78, + "grad_norm": 0.4453125, + "learning_rate": 0.00013441134197849942, + "loss": 2.5483, + "step": 18280 + }, + { + "epoch": 0.78, + "grad_norm": 0.4296875, + "learning_rate": 0.00013440504363575772, + "loss": 2.5135, + "step": 18281 + }, + { + "epoch": 0.78, + "grad_norm": 0.490234375, + "learning_rate": 0.00013439874513820595, + "loss": 2.6674, + "step": 18282 + }, + { + "epoch": 0.78, + "grad_norm": 0.47265625, + "learning_rate": 0.00013439244648587236, + "loss": 2.6468, + "step": 18283 + }, + { + "epoch": 0.78, + "grad_norm": 0.40234375, + "learning_rate": 0.00013438614767878534, + "loss": 2.6049, + "step": 18284 + }, + { + "epoch": 0.78, + "grad_norm": 0.423828125, + "learning_rate": 0.00013437984871697323, + "loss": 2.4977, + "step": 18285 + }, + { + "epoch": 0.78, + "grad_norm": 0.3984375, + "learning_rate": 0.00013437354960046436, + "loss": 2.5618, + "step": 18286 + }, + { + "epoch": 0.78, + "grad_norm": 0.42578125, + "learning_rate": 0.00013436725032928707, + "loss": 2.6171, + "step": 18287 + }, + { + "epoch": 0.78, + "grad_norm": 0.470703125, + "learning_rate": 0.00013436095090346974, + "loss": 2.5632, + "step": 18288 + }, + { + "epoch": 0.78, + "grad_norm": 0.439453125, + "learning_rate": 0.00013435465132304066, + "loss": 2.4699, + "step": 18289 + }, + { + "epoch": 0.78, + "grad_norm": 0.47265625, + "learning_rate": 0.0001343483515880282, + "loss": 2.5452, + "step": 18290 + }, + { + "epoch": 0.78, + "grad_norm": 0.431640625, + "learning_rate": 0.00013434205169846072, + "loss": 2.4845, + "step": 18291 + }, + { + "epoch": 0.78, + "grad_norm": 0.4609375, + "learning_rate": 0.0001343357516543666, + "loss": 2.6349, + "step": 18292 + }, + { + "epoch": 0.78, + "grad_norm": 0.439453125, + "learning_rate": 0.00013432945145577408, + "loss": 2.7043, + "step": 18293 + }, + { + "epoch": 0.78, + "grad_norm": 0.50390625, + "learning_rate": 0.0001343231511027116, + "loss": 2.5828, + "step": 18294 + }, + { + "epoch": 0.78, + "grad_norm": 0.51171875, + "learning_rate": 0.00013431685059520752, + "loss": 2.654, + "step": 18295 + }, + { + "epoch": 0.78, + "grad_norm": 0.484375, + "learning_rate": 0.00013431054993329013, + "loss": 2.6355, + "step": 18296 + }, + { + "epoch": 0.78, + "grad_norm": 0.451171875, + "learning_rate": 0.00013430424911698777, + "loss": 2.5094, + "step": 18297 + }, + { + "epoch": 0.78, + "grad_norm": 0.4375, + "learning_rate": 0.00013429794814632888, + "loss": 2.56, + "step": 18298 + }, + { + "epoch": 0.78, + "grad_norm": 0.5390625, + "learning_rate": 0.0001342916470213417, + "loss": 2.611, + "step": 18299 + }, + { + "epoch": 0.78, + "grad_norm": 0.359375, + "learning_rate": 0.0001342853457420547, + "loss": 2.4955, + "step": 18300 + }, + { + "epoch": 0.78, + "grad_norm": 0.416015625, + "learning_rate": 0.00013427904430849617, + "loss": 2.5164, + "step": 18301 + }, + { + "epoch": 0.78, + "grad_norm": 0.4296875, + "learning_rate": 0.0001342727427206944, + "loss": 2.5154, + "step": 18302 + }, + { + "epoch": 0.78, + "grad_norm": 0.482421875, + "learning_rate": 0.00013426644097867788, + "loss": 2.6243, + "step": 18303 + }, + { + "epoch": 0.78, + "grad_norm": 0.5, + "learning_rate": 0.00013426013908247484, + "loss": 2.6478, + "step": 18304 + }, + { + "epoch": 0.78, + "grad_norm": 0.486328125, + "learning_rate": 0.00013425383703211376, + "loss": 2.5967, + "step": 18305 + }, + { + "epoch": 0.78, + "grad_norm": 0.4375, + "learning_rate": 0.00013424753482762288, + "loss": 2.6325, + "step": 18306 + }, + { + "epoch": 0.78, + "grad_norm": 0.421875, + "learning_rate": 0.00013424123246903063, + "loss": 2.5683, + "step": 18307 + }, + { + "epoch": 0.78, + "grad_norm": 0.4609375, + "learning_rate": 0.00013423492995636536, + "loss": 2.5966, + "step": 18308 + }, + { + "epoch": 0.78, + "grad_norm": 0.515625, + "learning_rate": 0.00013422862728965543, + "loss": 2.625, + "step": 18309 + }, + { + "epoch": 0.78, + "grad_norm": 0.484375, + "learning_rate": 0.00013422232446892913, + "loss": 2.6342, + "step": 18310 + }, + { + "epoch": 0.78, + "grad_norm": 0.46484375, + "learning_rate": 0.00013421602149421493, + "loss": 2.5754, + "step": 18311 + }, + { + "epoch": 0.78, + "grad_norm": 0.435546875, + "learning_rate": 0.0001342097183655411, + "loss": 2.5326, + "step": 18312 + }, + { + "epoch": 0.78, + "grad_norm": 0.455078125, + "learning_rate": 0.00013420341508293607, + "loss": 2.459, + "step": 18313 + }, + { + "epoch": 0.78, + "grad_norm": 0.462890625, + "learning_rate": 0.00013419711164642815, + "loss": 2.5567, + "step": 18314 + }, + { + "epoch": 0.78, + "grad_norm": 0.412109375, + "learning_rate": 0.00013419080805604573, + "loss": 2.5161, + "step": 18315 + }, + { + "epoch": 0.78, + "grad_norm": 0.474609375, + "learning_rate": 0.00013418450431181716, + "loss": 2.5683, + "step": 18316 + }, + { + "epoch": 0.78, + "grad_norm": 0.490234375, + "learning_rate": 0.00013417820041377085, + "loss": 2.6163, + "step": 18317 + }, + { + "epoch": 0.78, + "grad_norm": 0.44921875, + "learning_rate": 0.00013417189636193508, + "loss": 2.6176, + "step": 18318 + }, + { + "epoch": 0.78, + "grad_norm": 0.474609375, + "learning_rate": 0.00013416559215633831, + "loss": 2.5191, + "step": 18319 + }, + { + "epoch": 0.78, + "grad_norm": 0.53125, + "learning_rate": 0.00013415928779700881, + "loss": 2.6508, + "step": 18320 + }, + { + "epoch": 0.78, + "grad_norm": 0.498046875, + "learning_rate": 0.00013415298328397506, + "loss": 2.6925, + "step": 18321 + }, + { + "epoch": 0.78, + "grad_norm": 0.5, + "learning_rate": 0.00013414667861726533, + "loss": 2.6498, + "step": 18322 + }, + { + "epoch": 0.78, + "grad_norm": 0.41796875, + "learning_rate": 0.00013414037379690802, + "loss": 2.6253, + "step": 18323 + }, + { + "epoch": 0.78, + "grad_norm": 0.490234375, + "learning_rate": 0.00013413406882293154, + "loss": 2.4614, + "step": 18324 + }, + { + "epoch": 0.78, + "grad_norm": 0.4296875, + "learning_rate": 0.0001341277636953642, + "loss": 2.5505, + "step": 18325 + }, + { + "epoch": 0.78, + "grad_norm": 0.447265625, + "learning_rate": 0.00013412145841423437, + "loss": 2.5736, + "step": 18326 + }, + { + "epoch": 0.78, + "grad_norm": 0.462890625, + "learning_rate": 0.0001341151529795705, + "loss": 2.5816, + "step": 18327 + }, + { + "epoch": 0.78, + "grad_norm": 0.443359375, + "learning_rate": 0.00013410884739140086, + "loss": 2.5604, + "step": 18328 + }, + { + "epoch": 0.78, + "grad_norm": 0.5, + "learning_rate": 0.0001341025416497539, + "loss": 2.651, + "step": 18329 + }, + { + "epoch": 0.78, + "grad_norm": 0.474609375, + "learning_rate": 0.00013409623575465798, + "loss": 2.6156, + "step": 18330 + }, + { + "epoch": 0.78, + "grad_norm": 0.435546875, + "learning_rate": 0.00013408992970614141, + "loss": 2.4797, + "step": 18331 + }, + { + "epoch": 0.78, + "grad_norm": 0.474609375, + "learning_rate": 0.00013408362350423266, + "loss": 2.6282, + "step": 18332 + }, + { + "epoch": 0.78, + "grad_norm": 0.478515625, + "learning_rate": 0.00013407731714896006, + "loss": 2.5643, + "step": 18333 + }, + { + "epoch": 0.78, + "grad_norm": 0.376953125, + "learning_rate": 0.00013407101064035195, + "loss": 2.4557, + "step": 18334 + }, + { + "epoch": 0.78, + "grad_norm": 0.53125, + "learning_rate": 0.00013406470397843678, + "loss": 2.5654, + "step": 18335 + }, + { + "epoch": 0.78, + "grad_norm": 0.5234375, + "learning_rate": 0.00013405839716324285, + "loss": 2.4914, + "step": 18336 + }, + { + "epoch": 0.78, + "grad_norm": 0.42578125, + "learning_rate": 0.00013405209019479863, + "loss": 2.5253, + "step": 18337 + }, + { + "epoch": 0.78, + "grad_norm": 0.474609375, + "learning_rate": 0.00013404578307313241, + "loss": 2.6241, + "step": 18338 + }, + { + "epoch": 0.78, + "grad_norm": 0.408203125, + "learning_rate": 0.00013403947579827263, + "loss": 2.7089, + "step": 18339 + }, + { + "epoch": 0.78, + "grad_norm": 0.423828125, + "learning_rate": 0.00013403316837024763, + "loss": 2.5445, + "step": 18340 + }, + { + "epoch": 0.78, + "grad_norm": 0.4375, + "learning_rate": 0.00013402686078908585, + "loss": 2.5725, + "step": 18341 + }, + { + "epoch": 0.78, + "grad_norm": 0.51171875, + "learning_rate": 0.0001340205530548156, + "loss": 2.6677, + "step": 18342 + }, + { + "epoch": 0.78, + "grad_norm": 0.484375, + "learning_rate": 0.0001340142451674653, + "loss": 2.6602, + "step": 18343 + }, + { + "epoch": 0.78, + "grad_norm": 0.498046875, + "learning_rate": 0.00013400793712706332, + "loss": 2.6248, + "step": 18344 + }, + { + "epoch": 0.78, + "grad_norm": 0.42578125, + "learning_rate": 0.00013400162893363805, + "loss": 2.3996, + "step": 18345 + }, + { + "epoch": 0.78, + "grad_norm": 0.427734375, + "learning_rate": 0.00013399532058721789, + "loss": 2.4778, + "step": 18346 + }, + { + "epoch": 0.78, + "grad_norm": 0.4765625, + "learning_rate": 0.00013398901208783122, + "loss": 2.5824, + "step": 18347 + }, + { + "epoch": 0.78, + "grad_norm": 0.486328125, + "learning_rate": 0.0001339827034355064, + "loss": 2.5968, + "step": 18348 + }, + { + "epoch": 0.78, + "grad_norm": 0.47265625, + "learning_rate": 0.00013397639463027183, + "loss": 2.5281, + "step": 18349 + }, + { + "epoch": 0.78, + "grad_norm": 0.474609375, + "learning_rate": 0.00013397008567215589, + "loss": 2.4818, + "step": 18350 + }, + { + "epoch": 0.78, + "grad_norm": 0.494140625, + "learning_rate": 0.000133963776561187, + "loss": 2.6434, + "step": 18351 + }, + { + "epoch": 0.78, + "grad_norm": 0.462890625, + "learning_rate": 0.00013395746729739355, + "loss": 2.5578, + "step": 18352 + }, + { + "epoch": 0.78, + "grad_norm": 0.47265625, + "learning_rate": 0.00013395115788080385, + "loss": 2.6303, + "step": 18353 + }, + { + "epoch": 0.78, + "grad_norm": 0.458984375, + "learning_rate": 0.00013394484831144642, + "loss": 2.6141, + "step": 18354 + }, + { + "epoch": 0.78, + "grad_norm": 0.5234375, + "learning_rate": 0.0001339385385893495, + "loss": 2.58, + "step": 18355 + }, + { + "epoch": 0.78, + "grad_norm": 0.462890625, + "learning_rate": 0.0001339322287145416, + "loss": 2.5665, + "step": 18356 + }, + { + "epoch": 0.78, + "grad_norm": 0.4609375, + "learning_rate": 0.0001339259186870511, + "loss": 2.4986, + "step": 18357 + }, + { + "epoch": 0.78, + "grad_norm": 0.49609375, + "learning_rate": 0.0001339196085069063, + "loss": 2.6415, + "step": 18358 + }, + { + "epoch": 0.78, + "grad_norm": 0.41796875, + "learning_rate": 0.0001339132981741357, + "loss": 2.5777, + "step": 18359 + }, + { + "epoch": 0.78, + "grad_norm": 0.482421875, + "learning_rate": 0.00013390698768876765, + "loss": 2.4699, + "step": 18360 + }, + { + "epoch": 0.78, + "grad_norm": 0.455078125, + "learning_rate": 0.00013390067705083055, + "loss": 2.6832, + "step": 18361 + }, + { + "epoch": 0.78, + "grad_norm": 0.498046875, + "learning_rate": 0.00013389436626035276, + "loss": 2.6975, + "step": 18362 + }, + { + "epoch": 0.78, + "grad_norm": 0.4609375, + "learning_rate": 0.00013388805531736275, + "loss": 2.6234, + "step": 18363 + }, + { + "epoch": 0.78, + "grad_norm": 0.38671875, + "learning_rate": 0.00013388174422188884, + "loss": 2.5844, + "step": 18364 + }, + { + "epoch": 0.78, + "grad_norm": 0.51953125, + "learning_rate": 0.0001338754329739595, + "loss": 2.7423, + "step": 18365 + }, + { + "epoch": 0.78, + "grad_norm": 0.447265625, + "learning_rate": 0.00013386912157360302, + "loss": 2.6652, + "step": 18366 + }, + { + "epoch": 0.78, + "grad_norm": 0.52734375, + "learning_rate": 0.00013386281002084798, + "loss": 2.791, + "step": 18367 + }, + { + "epoch": 0.78, + "grad_norm": 0.50390625, + "learning_rate": 0.00013385649831572258, + "loss": 2.6613, + "step": 18368 + }, + { + "epoch": 0.78, + "grad_norm": 0.4921875, + "learning_rate": 0.00013385018645825536, + "loss": 2.5083, + "step": 18369 + }, + { + "epoch": 0.78, + "grad_norm": 0.50390625, + "learning_rate": 0.00013384387444847465, + "loss": 2.4539, + "step": 18370 + }, + { + "epoch": 0.78, + "grad_norm": 0.4453125, + "learning_rate": 0.00013383756228640888, + "loss": 2.5419, + "step": 18371 + }, + { + "epoch": 0.78, + "grad_norm": 0.46484375, + "learning_rate": 0.00013383124997208646, + "loss": 2.4672, + "step": 18372 + }, + { + "epoch": 0.78, + "grad_norm": 0.46875, + "learning_rate": 0.00013382493750553574, + "loss": 2.6111, + "step": 18373 + }, + { + "epoch": 0.78, + "grad_norm": 0.421875, + "learning_rate": 0.00013381862488678518, + "loss": 2.4586, + "step": 18374 + }, + { + "epoch": 0.78, + "grad_norm": 0.45703125, + "learning_rate": 0.0001338123121158632, + "loss": 2.6714, + "step": 18375 + }, + { + "epoch": 0.78, + "grad_norm": 0.498046875, + "learning_rate": 0.00013380599919279813, + "loss": 2.7268, + "step": 18376 + }, + { + "epoch": 0.78, + "grad_norm": 0.498046875, + "learning_rate": 0.0001337996861176184, + "loss": 2.5999, + "step": 18377 + }, + { + "epoch": 0.78, + "grad_norm": 0.435546875, + "learning_rate": 0.0001337933728903525, + "loss": 2.6363, + "step": 18378 + }, + { + "epoch": 0.78, + "grad_norm": 0.470703125, + "learning_rate": 0.00013378705951102874, + "loss": 2.5383, + "step": 18379 + }, + { + "epoch": 0.78, + "grad_norm": 0.455078125, + "learning_rate": 0.00013378074597967556, + "loss": 2.65, + "step": 18380 + }, + { + "epoch": 0.78, + "grad_norm": 0.419921875, + "learning_rate": 0.00013377443229632138, + "loss": 2.4495, + "step": 18381 + }, + { + "epoch": 0.78, + "grad_norm": 0.50390625, + "learning_rate": 0.00013376811846099458, + "loss": 2.675, + "step": 18382 + }, + { + "epoch": 0.78, + "grad_norm": 0.451171875, + "learning_rate": 0.00013376180447372362, + "loss": 2.5465, + "step": 18383 + }, + { + "epoch": 0.78, + "grad_norm": 0.419921875, + "learning_rate": 0.00013375549033453683, + "loss": 2.544, + "step": 18384 + }, + { + "epoch": 0.78, + "grad_norm": 0.455078125, + "learning_rate": 0.0001337491760434627, + "loss": 2.6055, + "step": 18385 + }, + { + "epoch": 0.78, + "grad_norm": 0.46875, + "learning_rate": 0.0001337428616005296, + "loss": 2.5817, + "step": 18386 + }, + { + "epoch": 0.78, + "grad_norm": 0.44140625, + "learning_rate": 0.00013373654700576595, + "loss": 2.5456, + "step": 18387 + }, + { + "epoch": 0.78, + "grad_norm": 0.439453125, + "learning_rate": 0.0001337302322592002, + "loss": 2.6133, + "step": 18388 + }, + { + "epoch": 0.78, + "grad_norm": 0.45703125, + "learning_rate": 0.0001337239173608607, + "loss": 2.6485, + "step": 18389 + }, + { + "epoch": 0.78, + "grad_norm": 0.423828125, + "learning_rate": 0.00013371760231077587, + "loss": 2.527, + "step": 18390 + }, + { + "epoch": 0.78, + "grad_norm": 0.419921875, + "learning_rate": 0.00013371128710897422, + "loss": 2.5054, + "step": 18391 + }, + { + "epoch": 0.78, + "grad_norm": 0.451171875, + "learning_rate": 0.00013370497175548406, + "loss": 2.6, + "step": 18392 + }, + { + "epoch": 0.78, + "grad_norm": 0.421875, + "learning_rate": 0.00013369865625033385, + "loss": 2.5497, + "step": 18393 + }, + { + "epoch": 0.78, + "grad_norm": 0.416015625, + "learning_rate": 0.000133692340593552, + "loss": 2.5803, + "step": 18394 + }, + { + "epoch": 0.78, + "grad_norm": 0.515625, + "learning_rate": 0.00013368602478516694, + "loss": 2.5926, + "step": 18395 + }, + { + "epoch": 0.78, + "grad_norm": 0.423828125, + "learning_rate": 0.00013367970882520707, + "loss": 2.5083, + "step": 18396 + }, + { + "epoch": 0.78, + "grad_norm": 0.44921875, + "learning_rate": 0.00013367339271370084, + "loss": 2.4934, + "step": 18397 + }, + { + "epoch": 0.78, + "grad_norm": 0.416015625, + "learning_rate": 0.0001336670764506766, + "loss": 2.596, + "step": 18398 + }, + { + "epoch": 0.78, + "grad_norm": 0.451171875, + "learning_rate": 0.00013366076003616288, + "loss": 2.5858, + "step": 18399 + }, + { + "epoch": 0.78, + "grad_norm": 0.494140625, + "learning_rate": 0.00013365444347018799, + "loss": 2.5811, + "step": 18400 + }, + { + "epoch": 0.78, + "grad_norm": 0.45703125, + "learning_rate": 0.0001336481267527804, + "loss": 2.595, + "step": 18401 + }, + { + "epoch": 0.78, + "grad_norm": 0.48828125, + "learning_rate": 0.00013364180988396858, + "loss": 2.603, + "step": 18402 + }, + { + "epoch": 0.78, + "grad_norm": 0.4375, + "learning_rate": 0.0001336354928637809, + "loss": 2.5821, + "step": 18403 + }, + { + "epoch": 0.78, + "grad_norm": 0.45703125, + "learning_rate": 0.0001336291756922458, + "loss": 2.5662, + "step": 18404 + }, + { + "epoch": 0.78, + "grad_norm": 0.453125, + "learning_rate": 0.0001336228583693917, + "loss": 2.5255, + "step": 18405 + }, + { + "epoch": 0.78, + "grad_norm": 0.443359375, + "learning_rate": 0.00013361654089524697, + "loss": 2.5833, + "step": 18406 + }, + { + "epoch": 0.78, + "grad_norm": 0.4296875, + "learning_rate": 0.00013361022326984016, + "loss": 2.4543, + "step": 18407 + }, + { + "epoch": 0.78, + "grad_norm": 0.451171875, + "learning_rate": 0.0001336039054931996, + "loss": 2.428, + "step": 18408 + }, + { + "epoch": 0.78, + "grad_norm": 0.46875, + "learning_rate": 0.00013359758756535374, + "loss": 2.5119, + "step": 18409 + }, + { + "epoch": 0.78, + "grad_norm": 0.443359375, + "learning_rate": 0.00013359126948633103, + "loss": 2.5666, + "step": 18410 + }, + { + "epoch": 0.78, + "grad_norm": 0.515625, + "learning_rate": 0.00013358495125615988, + "loss": 2.5874, + "step": 18411 + }, + { + "epoch": 0.78, + "grad_norm": 0.4765625, + "learning_rate": 0.00013357863287486872, + "loss": 2.5182, + "step": 18412 + }, + { + "epoch": 0.78, + "grad_norm": 0.427734375, + "learning_rate": 0.00013357231434248597, + "loss": 2.4504, + "step": 18413 + }, + { + "epoch": 0.78, + "grad_norm": 0.443359375, + "learning_rate": 0.0001335659956590401, + "loss": 2.4699, + "step": 18414 + }, + { + "epoch": 0.78, + "grad_norm": 0.5078125, + "learning_rate": 0.0001335596768245595, + "loss": 2.6857, + "step": 18415 + }, + { + "epoch": 0.78, + "grad_norm": 0.46875, + "learning_rate": 0.00013355335783907262, + "loss": 2.5542, + "step": 18416 + }, + { + "epoch": 0.78, + "grad_norm": 0.470703125, + "learning_rate": 0.00013354703870260788, + "loss": 2.7087, + "step": 18417 + }, + { + "epoch": 0.78, + "grad_norm": 0.443359375, + "learning_rate": 0.00013354071941519376, + "loss": 2.5661, + "step": 18418 + }, + { + "epoch": 0.78, + "grad_norm": 0.484375, + "learning_rate": 0.00013353439997685864, + "loss": 2.6477, + "step": 18419 + }, + { + "epoch": 0.78, + "grad_norm": 0.57421875, + "learning_rate": 0.00013352808038763096, + "loss": 2.6191, + "step": 18420 + }, + { + "epoch": 0.78, + "grad_norm": 0.462890625, + "learning_rate": 0.0001335217606475392, + "loss": 2.6265, + "step": 18421 + }, + { + "epoch": 0.79, + "grad_norm": 0.484375, + "learning_rate": 0.00013351544075661175, + "loss": 2.6625, + "step": 18422 + }, + { + "epoch": 0.79, + "grad_norm": 0.400390625, + "learning_rate": 0.00013350912071487708, + "loss": 2.5884, + "step": 18423 + }, + { + "epoch": 0.79, + "grad_norm": 0.40625, + "learning_rate": 0.00013350280052236362, + "loss": 2.4856, + "step": 18424 + }, + { + "epoch": 0.79, + "grad_norm": 0.466796875, + "learning_rate": 0.00013349648017909975, + "loss": 2.4801, + "step": 18425 + }, + { + "epoch": 0.79, + "grad_norm": 0.44921875, + "learning_rate": 0.00013349015968511403, + "loss": 2.7463, + "step": 18426 + }, + { + "epoch": 0.79, + "grad_norm": 0.46484375, + "learning_rate": 0.00013348383904043478, + "loss": 2.4336, + "step": 18427 + }, + { + "epoch": 0.79, + "grad_norm": 0.494140625, + "learning_rate": 0.0001334775182450905, + "loss": 2.5812, + "step": 18428 + }, + { + "epoch": 0.79, + "grad_norm": 0.49609375, + "learning_rate": 0.00013347119729910964, + "loss": 2.5795, + "step": 18429 + }, + { + "epoch": 0.79, + "grad_norm": 0.443359375, + "learning_rate": 0.0001334648762025206, + "loss": 2.5299, + "step": 18430 + }, + { + "epoch": 0.79, + "grad_norm": 0.48046875, + "learning_rate": 0.0001334585549553519, + "loss": 2.5367, + "step": 18431 + }, + { + "epoch": 0.79, + "grad_norm": 0.427734375, + "learning_rate": 0.00013345223355763186, + "loss": 2.51, + "step": 18432 + }, + { + "epoch": 0.79, + "grad_norm": 0.4609375, + "learning_rate": 0.00013344591200938904, + "loss": 2.6783, + "step": 18433 + }, + { + "epoch": 0.79, + "grad_norm": 0.494140625, + "learning_rate": 0.00013343959031065182, + "loss": 2.5846, + "step": 18434 + }, + { + "epoch": 0.79, + "grad_norm": 0.51171875, + "learning_rate": 0.00013343326846144865, + "loss": 2.5539, + "step": 18435 + }, + { + "epoch": 0.79, + "grad_norm": 0.458984375, + "learning_rate": 0.00013342694646180804, + "loss": 2.5189, + "step": 18436 + }, + { + "epoch": 0.79, + "grad_norm": 0.49609375, + "learning_rate": 0.00013342062431175836, + "loss": 2.6118, + "step": 18437 + }, + { + "epoch": 0.79, + "grad_norm": 0.4140625, + "learning_rate": 0.00013341430201132808, + "loss": 2.551, + "step": 18438 + }, + { + "epoch": 0.79, + "grad_norm": 0.431640625, + "learning_rate": 0.00013340797956054565, + "loss": 2.5367, + "step": 18439 + }, + { + "epoch": 0.79, + "grad_norm": 0.52734375, + "learning_rate": 0.00013340165695943952, + "loss": 2.6005, + "step": 18440 + }, + { + "epoch": 0.79, + "grad_norm": 0.46484375, + "learning_rate": 0.00013339533420803814, + "loss": 2.5992, + "step": 18441 + }, + { + "epoch": 0.79, + "grad_norm": 0.41796875, + "learning_rate": 0.00013338901130637, + "loss": 2.5341, + "step": 18442 + }, + { + "epoch": 0.79, + "grad_norm": 0.48046875, + "learning_rate": 0.00013338268825446346, + "loss": 2.5765, + "step": 18443 + }, + { + "epoch": 0.79, + "grad_norm": 0.427734375, + "learning_rate": 0.00013337636505234704, + "loss": 2.5372, + "step": 18444 + }, + { + "epoch": 0.79, + "grad_norm": 0.5, + "learning_rate": 0.00013337004170004915, + "loss": 2.6473, + "step": 18445 + }, + { + "epoch": 0.79, + "grad_norm": 0.462890625, + "learning_rate": 0.00013336371819759832, + "loss": 2.5987, + "step": 18446 + }, + { + "epoch": 0.79, + "grad_norm": 0.431640625, + "learning_rate": 0.00013335739454502292, + "loss": 2.5076, + "step": 18447 + }, + { + "epoch": 0.79, + "grad_norm": 0.49609375, + "learning_rate": 0.00013335107074235143, + "loss": 2.5921, + "step": 18448 + }, + { + "epoch": 0.79, + "grad_norm": 0.51953125, + "learning_rate": 0.0001333447467896123, + "loss": 2.6017, + "step": 18449 + }, + { + "epoch": 0.79, + "grad_norm": 0.443359375, + "learning_rate": 0.00013333842268683406, + "loss": 2.6074, + "step": 18450 + }, + { + "epoch": 0.79, + "grad_norm": 0.435546875, + "learning_rate": 0.00013333209843404502, + "loss": 2.5453, + "step": 18451 + }, + { + "epoch": 0.79, + "grad_norm": 0.41796875, + "learning_rate": 0.00013332577403127378, + "loss": 2.5249, + "step": 18452 + }, + { + "epoch": 0.79, + "grad_norm": 0.474609375, + "learning_rate": 0.00013331944947854872, + "loss": 2.5907, + "step": 18453 + }, + { + "epoch": 0.79, + "grad_norm": 0.412109375, + "learning_rate": 0.0001333131247758983, + "loss": 2.5218, + "step": 18454 + }, + { + "epoch": 0.79, + "grad_norm": 0.44921875, + "learning_rate": 0.000133306799923351, + "loss": 2.5914, + "step": 18455 + }, + { + "epoch": 0.79, + "grad_norm": 0.41015625, + "learning_rate": 0.00013330047492093528, + "loss": 2.4328, + "step": 18456 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.00013329414976867956, + "loss": 2.5138, + "step": 18457 + }, + { + "epoch": 0.79, + "grad_norm": 0.44921875, + "learning_rate": 0.00013328782446661239, + "loss": 2.5865, + "step": 18458 + }, + { + "epoch": 0.79, + "grad_norm": 0.45703125, + "learning_rate": 0.00013328149901476212, + "loss": 2.6048, + "step": 18459 + }, + { + "epoch": 0.79, + "grad_norm": 0.435546875, + "learning_rate": 0.0001332751734131573, + "loss": 2.5665, + "step": 18460 + }, + { + "epoch": 0.79, + "grad_norm": 0.435546875, + "learning_rate": 0.00013326884766182634, + "loss": 2.5883, + "step": 18461 + }, + { + "epoch": 0.79, + "grad_norm": 0.458984375, + "learning_rate": 0.00013326252176079775, + "loss": 2.5795, + "step": 18462 + }, + { + "epoch": 0.79, + "grad_norm": 0.4609375, + "learning_rate": 0.00013325619571009995, + "loss": 2.7187, + "step": 18463 + }, + { + "epoch": 0.79, + "grad_norm": 0.4609375, + "learning_rate": 0.00013324986950976142, + "loss": 2.5253, + "step": 18464 + }, + { + "epoch": 0.79, + "grad_norm": 0.4609375, + "learning_rate": 0.00013324354315981062, + "loss": 2.4905, + "step": 18465 + }, + { + "epoch": 0.79, + "grad_norm": 0.43359375, + "learning_rate": 0.00013323721666027606, + "loss": 2.6865, + "step": 18466 + }, + { + "epoch": 0.79, + "grad_norm": 0.470703125, + "learning_rate": 0.00013323089001118614, + "loss": 2.5345, + "step": 18467 + }, + { + "epoch": 0.79, + "grad_norm": 0.455078125, + "learning_rate": 0.00013322456321256937, + "loss": 2.6224, + "step": 18468 + }, + { + "epoch": 0.79, + "grad_norm": 0.40625, + "learning_rate": 0.0001332182362644542, + "loss": 2.5108, + "step": 18469 + }, + { + "epoch": 0.79, + "grad_norm": 0.423828125, + "learning_rate": 0.00013321190916686915, + "loss": 2.4978, + "step": 18470 + }, + { + "epoch": 0.79, + "grad_norm": 0.416015625, + "learning_rate": 0.0001332055819198426, + "loss": 2.5468, + "step": 18471 + }, + { + "epoch": 0.79, + "grad_norm": 0.431640625, + "learning_rate": 0.00013319925452340308, + "loss": 2.6724, + "step": 18472 + }, + { + "epoch": 0.79, + "grad_norm": 0.498046875, + "learning_rate": 0.00013319292697757903, + "loss": 2.5906, + "step": 18473 + }, + { + "epoch": 0.79, + "grad_norm": 0.419921875, + "learning_rate": 0.000133186599282399, + "loss": 2.522, + "step": 18474 + }, + { + "epoch": 0.79, + "grad_norm": 0.45703125, + "learning_rate": 0.0001331802714378913, + "loss": 2.5436, + "step": 18475 + }, + { + "epoch": 0.79, + "grad_norm": 0.4375, + "learning_rate": 0.00013317394344408457, + "loss": 2.5852, + "step": 18476 + }, + { + "epoch": 0.79, + "grad_norm": 0.494140625, + "learning_rate": 0.0001331676153010072, + "loss": 2.6057, + "step": 18477 + }, + { + "epoch": 0.79, + "grad_norm": 0.5, + "learning_rate": 0.0001331612870086877, + "loss": 2.599, + "step": 18478 + }, + { + "epoch": 0.79, + "grad_norm": 0.515625, + "learning_rate": 0.00013315495856715455, + "loss": 2.7782, + "step": 18479 + }, + { + "epoch": 0.79, + "grad_norm": 0.48046875, + "learning_rate": 0.00013314862997643615, + "loss": 2.668, + "step": 18480 + }, + { + "epoch": 0.79, + "grad_norm": 0.451171875, + "learning_rate": 0.00013314230123656105, + "loss": 2.639, + "step": 18481 + }, + { + "epoch": 0.79, + "grad_norm": 0.5078125, + "learning_rate": 0.00013313597234755771, + "loss": 2.6631, + "step": 18482 + }, + { + "epoch": 0.79, + "grad_norm": 0.5078125, + "learning_rate": 0.00013312964330945462, + "loss": 2.7133, + "step": 18483 + }, + { + "epoch": 0.79, + "grad_norm": 0.44921875, + "learning_rate": 0.0001331233141222802, + "loss": 2.6282, + "step": 18484 + }, + { + "epoch": 0.79, + "grad_norm": 0.44140625, + "learning_rate": 0.00013311698478606302, + "loss": 2.5164, + "step": 18485 + }, + { + "epoch": 0.79, + "grad_norm": 0.5, + "learning_rate": 0.00013311065530083148, + "loss": 2.6151, + "step": 18486 + }, + { + "epoch": 0.79, + "grad_norm": 0.484375, + "learning_rate": 0.0001331043256666141, + "loss": 2.485, + "step": 18487 + }, + { + "epoch": 0.79, + "grad_norm": 0.447265625, + "learning_rate": 0.00013309799588343937, + "loss": 2.5481, + "step": 18488 + }, + { + "epoch": 0.79, + "grad_norm": 0.48828125, + "learning_rate": 0.00013309166595133572, + "loss": 2.5943, + "step": 18489 + }, + { + "epoch": 0.79, + "grad_norm": 0.47265625, + "learning_rate": 0.0001330853358703317, + "loss": 2.6458, + "step": 18490 + }, + { + "epoch": 0.79, + "grad_norm": 0.5078125, + "learning_rate": 0.00013307900564045572, + "loss": 2.6542, + "step": 18491 + }, + { + "epoch": 0.79, + "grad_norm": 0.478515625, + "learning_rate": 0.00013307267526173634, + "loss": 2.5825, + "step": 18492 + }, + { + "epoch": 0.79, + "grad_norm": 0.4296875, + "learning_rate": 0.00013306634473420198, + "loss": 2.5208, + "step": 18493 + }, + { + "epoch": 0.79, + "grad_norm": 0.46484375, + "learning_rate": 0.00013306001405788116, + "loss": 2.5508, + "step": 18494 + }, + { + "epoch": 0.79, + "grad_norm": 0.4375, + "learning_rate": 0.0001330536832328024, + "loss": 2.3978, + "step": 18495 + }, + { + "epoch": 0.79, + "grad_norm": 0.4609375, + "learning_rate": 0.00013304735225899412, + "loss": 2.5605, + "step": 18496 + }, + { + "epoch": 0.79, + "grad_norm": 0.41015625, + "learning_rate": 0.0001330410211364848, + "loss": 2.4348, + "step": 18497 + }, + { + "epoch": 0.79, + "grad_norm": 0.48046875, + "learning_rate": 0.000133034689865303, + "loss": 2.5831, + "step": 18498 + }, + { + "epoch": 0.79, + "grad_norm": 0.4921875, + "learning_rate": 0.00013302835844547714, + "loss": 2.5778, + "step": 18499 + }, + { + "epoch": 0.79, + "grad_norm": 0.470703125, + "learning_rate": 0.00013302202687703575, + "loss": 2.47, + "step": 18500 + }, + { + "epoch": 0.79, + "grad_norm": 0.466796875, + "learning_rate": 0.0001330156951600073, + "loss": 2.4447, + "step": 18501 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.0001330093632944203, + "loss": 2.7254, + "step": 18502 + }, + { + "epoch": 0.79, + "grad_norm": 0.42578125, + "learning_rate": 0.00013300303128030324, + "loss": 2.5554, + "step": 18503 + }, + { + "epoch": 0.79, + "grad_norm": 0.478515625, + "learning_rate": 0.0001329966991176846, + "loss": 2.4838, + "step": 18504 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.00013299036680659282, + "loss": 2.5598, + "step": 18505 + }, + { + "epoch": 0.79, + "grad_norm": 0.51171875, + "learning_rate": 0.00013298403434705653, + "loss": 2.5546, + "step": 18506 + }, + { + "epoch": 0.79, + "grad_norm": 0.455078125, + "learning_rate": 0.00013297770173910408, + "loss": 2.6767, + "step": 18507 + }, + { + "epoch": 0.79, + "grad_norm": 0.419921875, + "learning_rate": 0.00013297136898276403, + "loss": 2.493, + "step": 18508 + }, + { + "epoch": 0.79, + "grad_norm": 0.5078125, + "learning_rate": 0.00013296503607806488, + "loss": 2.5508, + "step": 18509 + }, + { + "epoch": 0.79, + "grad_norm": 0.474609375, + "learning_rate": 0.0001329587030250351, + "loss": 2.4697, + "step": 18510 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.00013295236982370322, + "loss": 2.6856, + "step": 18511 + }, + { + "epoch": 0.79, + "grad_norm": 0.404296875, + "learning_rate": 0.00013294603647409772, + "loss": 2.602, + "step": 18512 + }, + { + "epoch": 0.79, + "grad_norm": 0.439453125, + "learning_rate": 0.00013293970297624706, + "loss": 2.5208, + "step": 18513 + }, + { + "epoch": 0.79, + "grad_norm": 0.486328125, + "learning_rate": 0.00013293336933017984, + "loss": 2.6122, + "step": 18514 + }, + { + "epoch": 0.79, + "grad_norm": 0.427734375, + "learning_rate": 0.00013292703553592446, + "loss": 2.7269, + "step": 18515 + }, + { + "epoch": 0.79, + "grad_norm": 0.453125, + "learning_rate": 0.0001329207015935094, + "loss": 2.5825, + "step": 18516 + }, + { + "epoch": 0.79, + "grad_norm": 0.490234375, + "learning_rate": 0.0001329143675029633, + "loss": 2.7054, + "step": 18517 + }, + { + "epoch": 0.79, + "grad_norm": 0.53515625, + "learning_rate": 0.00013290803326431453, + "loss": 2.6354, + "step": 18518 + }, + { + "epoch": 0.79, + "grad_norm": 0.466796875, + "learning_rate": 0.00013290169887759164, + "loss": 2.6883, + "step": 18519 + }, + { + "epoch": 0.79, + "grad_norm": 0.45703125, + "learning_rate": 0.00013289536434282314, + "loss": 2.5687, + "step": 18520 + }, + { + "epoch": 0.79, + "grad_norm": 0.48828125, + "learning_rate": 0.00013288902966003752, + "loss": 2.5783, + "step": 18521 + }, + { + "epoch": 0.79, + "grad_norm": 0.51953125, + "learning_rate": 0.0001328826948292633, + "loss": 2.717, + "step": 18522 + }, + { + "epoch": 0.79, + "grad_norm": 0.42578125, + "learning_rate": 0.00013287635985052894, + "loss": 2.657, + "step": 18523 + }, + { + "epoch": 0.79, + "grad_norm": 0.44140625, + "learning_rate": 0.000132870024723863, + "loss": 2.591, + "step": 18524 + }, + { + "epoch": 0.79, + "grad_norm": 0.447265625, + "learning_rate": 0.00013286368944929395, + "loss": 2.5152, + "step": 18525 + }, + { + "epoch": 0.79, + "grad_norm": 0.435546875, + "learning_rate": 0.00013285735402685032, + "loss": 2.5212, + "step": 18526 + }, + { + "epoch": 0.79, + "grad_norm": 0.4921875, + "learning_rate": 0.00013285101845656057, + "loss": 2.5726, + "step": 18527 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.0001328446827384533, + "loss": 2.5095, + "step": 18528 + }, + { + "epoch": 0.79, + "grad_norm": 0.48046875, + "learning_rate": 0.0001328383468725569, + "loss": 2.6564, + "step": 18529 + }, + { + "epoch": 0.79, + "grad_norm": 0.4765625, + "learning_rate": 0.00013283201085889998, + "loss": 2.5644, + "step": 18530 + }, + { + "epoch": 0.79, + "grad_norm": 0.41015625, + "learning_rate": 0.00013282567469751098, + "loss": 2.4799, + "step": 18531 + }, + { + "epoch": 0.79, + "grad_norm": 0.44140625, + "learning_rate": 0.00013281933838841847, + "loss": 2.4486, + "step": 18532 + }, + { + "epoch": 0.79, + "grad_norm": 0.46484375, + "learning_rate": 0.00013281300193165093, + "loss": 2.6811, + "step": 18533 + }, + { + "epoch": 0.79, + "grad_norm": 0.470703125, + "learning_rate": 0.00013280666532723684, + "loss": 2.6006, + "step": 18534 + }, + { + "epoch": 0.79, + "grad_norm": 0.4765625, + "learning_rate": 0.00013280032857520478, + "loss": 2.6581, + "step": 18535 + }, + { + "epoch": 0.79, + "grad_norm": 0.43359375, + "learning_rate": 0.00013279399167558323, + "loss": 2.6498, + "step": 18536 + }, + { + "epoch": 0.79, + "grad_norm": 0.470703125, + "learning_rate": 0.00013278765462840066, + "loss": 2.6174, + "step": 18537 + }, + { + "epoch": 0.79, + "grad_norm": 0.439453125, + "learning_rate": 0.0001327813174336857, + "loss": 2.5729, + "step": 18538 + }, + { + "epoch": 0.79, + "grad_norm": 0.484375, + "learning_rate": 0.00013277498009146672, + "loss": 2.5257, + "step": 18539 + }, + { + "epoch": 0.79, + "grad_norm": 0.462890625, + "learning_rate": 0.00013276864260177232, + "loss": 2.566, + "step": 18540 + }, + { + "epoch": 0.79, + "grad_norm": 0.455078125, + "learning_rate": 0.00013276230496463102, + "loss": 2.6375, + "step": 18541 + }, + { + "epoch": 0.79, + "grad_norm": 0.40234375, + "learning_rate": 0.00013275596718007133, + "loss": 2.5899, + "step": 18542 + }, + { + "epoch": 0.79, + "grad_norm": 0.4765625, + "learning_rate": 0.00013274962924812175, + "loss": 2.4955, + "step": 18543 + }, + { + "epoch": 0.79, + "grad_norm": 0.47265625, + "learning_rate": 0.00013274329116881082, + "loss": 2.4858, + "step": 18544 + }, + { + "epoch": 0.79, + "grad_norm": 0.435546875, + "learning_rate": 0.00013273695294216701, + "loss": 2.4787, + "step": 18545 + }, + { + "epoch": 0.79, + "grad_norm": 0.423828125, + "learning_rate": 0.00013273061456821892, + "loss": 2.5031, + "step": 18546 + }, + { + "epoch": 0.79, + "grad_norm": 0.421875, + "learning_rate": 0.000132724276046995, + "loss": 2.6962, + "step": 18547 + }, + { + "epoch": 0.79, + "grad_norm": 0.39453125, + "learning_rate": 0.00013271793737852378, + "loss": 2.51, + "step": 18548 + }, + { + "epoch": 0.79, + "grad_norm": 0.439453125, + "learning_rate": 0.00013271159856283383, + "loss": 2.4365, + "step": 18549 + }, + { + "epoch": 0.79, + "grad_norm": 0.44921875, + "learning_rate": 0.00013270525959995365, + "loss": 2.6004, + "step": 18550 + }, + { + "epoch": 0.79, + "grad_norm": 0.4453125, + "learning_rate": 0.0001326989204899117, + "loss": 2.782, + "step": 18551 + }, + { + "epoch": 0.79, + "grad_norm": 0.53125, + "learning_rate": 0.0001326925812327366, + "loss": 2.6852, + "step": 18552 + }, + { + "epoch": 0.79, + "grad_norm": 0.462890625, + "learning_rate": 0.00013268624182845683, + "loss": 2.6543, + "step": 18553 + }, + { + "epoch": 0.79, + "grad_norm": 0.431640625, + "learning_rate": 0.00013267990227710093, + "loss": 2.5329, + "step": 18554 + }, + { + "epoch": 0.79, + "grad_norm": 0.44140625, + "learning_rate": 0.0001326735625786974, + "loss": 2.5124, + "step": 18555 + }, + { + "epoch": 0.79, + "grad_norm": 0.423828125, + "learning_rate": 0.00013266722273327474, + "loss": 2.4297, + "step": 18556 + }, + { + "epoch": 0.79, + "grad_norm": 0.52734375, + "learning_rate": 0.00013266088274086158, + "loss": 2.5875, + "step": 18557 + }, + { + "epoch": 0.79, + "grad_norm": 0.43359375, + "learning_rate": 0.00013265454260148632, + "loss": 2.5121, + "step": 18558 + }, + { + "epoch": 0.79, + "grad_norm": 0.46484375, + "learning_rate": 0.0001326482023151776, + "loss": 2.5666, + "step": 18559 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.0001326418618819639, + "loss": 2.4369, + "step": 18560 + }, + { + "epoch": 0.79, + "grad_norm": 0.50390625, + "learning_rate": 0.00013263552130187375, + "loss": 2.7498, + "step": 18561 + }, + { + "epoch": 0.79, + "grad_norm": 0.40234375, + "learning_rate": 0.00013262918057493564, + "loss": 2.5455, + "step": 18562 + }, + { + "epoch": 0.79, + "grad_norm": 0.50390625, + "learning_rate": 0.00013262283970117818, + "loss": 2.6362, + "step": 18563 + }, + { + "epoch": 0.79, + "grad_norm": 0.486328125, + "learning_rate": 0.00013261649868062984, + "loss": 2.4598, + "step": 18564 + }, + { + "epoch": 0.79, + "grad_norm": 0.443359375, + "learning_rate": 0.00013261015751331922, + "loss": 2.4639, + "step": 18565 + }, + { + "epoch": 0.79, + "grad_norm": 0.423828125, + "learning_rate": 0.00013260381619927474, + "loss": 2.6211, + "step": 18566 + }, + { + "epoch": 0.79, + "grad_norm": 0.4765625, + "learning_rate": 0.00013259747473852505, + "loss": 2.627, + "step": 18567 + }, + { + "epoch": 0.79, + "grad_norm": 0.41015625, + "learning_rate": 0.00013259113313109864, + "loss": 2.5357, + "step": 18568 + }, + { + "epoch": 0.79, + "grad_norm": 0.44921875, + "learning_rate": 0.00013258479137702404, + "loss": 2.4752, + "step": 18569 + }, + { + "epoch": 0.79, + "grad_norm": 0.4921875, + "learning_rate": 0.00013257844947632978, + "loss": 2.5919, + "step": 18570 + }, + { + "epoch": 0.79, + "grad_norm": 0.462890625, + "learning_rate": 0.0001325721074290444, + "loss": 2.5967, + "step": 18571 + }, + { + "epoch": 0.79, + "grad_norm": 0.52734375, + "learning_rate": 0.0001325657652351964, + "loss": 2.6337, + "step": 18572 + }, + { + "epoch": 0.79, + "grad_norm": 0.41015625, + "learning_rate": 0.00013255942289481444, + "loss": 2.6474, + "step": 18573 + }, + { + "epoch": 0.79, + "grad_norm": 0.4765625, + "learning_rate": 0.00013255308040792692, + "loss": 2.6349, + "step": 18574 + }, + { + "epoch": 0.79, + "grad_norm": 0.48828125, + "learning_rate": 0.00013254673777456246, + "loss": 2.6485, + "step": 18575 + }, + { + "epoch": 0.79, + "grad_norm": 0.455078125, + "learning_rate": 0.00013254039499474956, + "loss": 2.5813, + "step": 18576 + }, + { + "epoch": 0.79, + "grad_norm": 0.470703125, + "learning_rate": 0.00013253405206851678, + "loss": 2.4823, + "step": 18577 + }, + { + "epoch": 0.79, + "grad_norm": 0.455078125, + "learning_rate": 0.00013252770899589266, + "loss": 2.6174, + "step": 18578 + }, + { + "epoch": 0.79, + "grad_norm": 0.44140625, + "learning_rate": 0.00013252136577690573, + "loss": 2.4198, + "step": 18579 + }, + { + "epoch": 0.79, + "grad_norm": 0.458984375, + "learning_rate": 0.00013251502241158454, + "loss": 2.5808, + "step": 18580 + }, + { + "epoch": 0.79, + "grad_norm": 0.455078125, + "learning_rate": 0.00013250867889995764, + "loss": 2.5493, + "step": 18581 + }, + { + "epoch": 0.79, + "grad_norm": 0.453125, + "learning_rate": 0.00013250233524205355, + "loss": 2.6652, + "step": 18582 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.00013249599143790082, + "loss": 2.6903, + "step": 18583 + }, + { + "epoch": 0.79, + "grad_norm": 0.482421875, + "learning_rate": 0.00013248964748752804, + "loss": 2.5462, + "step": 18584 + }, + { + "epoch": 0.79, + "grad_norm": 0.40625, + "learning_rate": 0.00013248330339096372, + "loss": 2.5885, + "step": 18585 + }, + { + "epoch": 0.79, + "grad_norm": 0.43359375, + "learning_rate": 0.0001324769591482364, + "loss": 2.5487, + "step": 18586 + }, + { + "epoch": 0.79, + "grad_norm": 0.462890625, + "learning_rate": 0.0001324706147593746, + "loss": 2.6305, + "step": 18587 + }, + { + "epoch": 0.79, + "grad_norm": 0.412109375, + "learning_rate": 0.0001324642702244069, + "loss": 2.5476, + "step": 18588 + }, + { + "epoch": 0.79, + "grad_norm": 0.53125, + "learning_rate": 0.0001324579255433619, + "loss": 2.6131, + "step": 18589 + }, + { + "epoch": 0.79, + "grad_norm": 0.5625, + "learning_rate": 0.00013245158071626805, + "loss": 2.7624, + "step": 18590 + }, + { + "epoch": 0.79, + "grad_norm": 0.54296875, + "learning_rate": 0.00013244523574315394, + "loss": 2.7904, + "step": 18591 + }, + { + "epoch": 0.79, + "grad_norm": 0.478515625, + "learning_rate": 0.00013243889062404817, + "loss": 2.5514, + "step": 18592 + }, + { + "epoch": 0.79, + "grad_norm": 0.42578125, + "learning_rate": 0.00013243254535897923, + "loss": 2.4665, + "step": 18593 + }, + { + "epoch": 0.79, + "grad_norm": 0.443359375, + "learning_rate": 0.0001324261999479757, + "loss": 2.5365, + "step": 18594 + }, + { + "epoch": 0.79, + "grad_norm": 0.478515625, + "learning_rate": 0.0001324198543910661, + "loss": 2.4748, + "step": 18595 + }, + { + "epoch": 0.79, + "grad_norm": 0.47265625, + "learning_rate": 0.000132413508688279, + "loss": 2.6115, + "step": 18596 + }, + { + "epoch": 0.79, + "grad_norm": 0.470703125, + "learning_rate": 0.000132407162839643, + "loss": 2.6569, + "step": 18597 + }, + { + "epoch": 0.79, + "grad_norm": 0.41796875, + "learning_rate": 0.00013240081684518654, + "loss": 2.4832, + "step": 18598 + }, + { + "epoch": 0.79, + "grad_norm": 0.478515625, + "learning_rate": 0.00013239447070493828, + "loss": 2.621, + "step": 18599 + }, + { + "epoch": 0.79, + "grad_norm": 0.419921875, + "learning_rate": 0.00013238812441892676, + "loss": 2.4693, + "step": 18600 + }, + { + "epoch": 0.79, + "grad_norm": 0.453125, + "learning_rate": 0.0001323817779871805, + "loss": 2.4892, + "step": 18601 + }, + { + "epoch": 0.79, + "grad_norm": 0.4140625, + "learning_rate": 0.00013237543140972807, + "loss": 2.574, + "step": 18602 + }, + { + "epoch": 0.79, + "grad_norm": 0.41796875, + "learning_rate": 0.00013236908468659803, + "loss": 2.4171, + "step": 18603 + }, + { + "epoch": 0.79, + "grad_norm": 0.470703125, + "learning_rate": 0.00013236273781781893, + "loss": 2.6159, + "step": 18604 + }, + { + "epoch": 0.79, + "grad_norm": 0.453125, + "learning_rate": 0.00013235639080341936, + "loss": 2.4836, + "step": 18605 + }, + { + "epoch": 0.79, + "grad_norm": 0.45703125, + "learning_rate": 0.00013235004364342784, + "loss": 2.7196, + "step": 18606 + }, + { + "epoch": 0.79, + "grad_norm": 0.48046875, + "learning_rate": 0.00013234369633787296, + "loss": 2.5977, + "step": 18607 + }, + { + "epoch": 0.79, + "grad_norm": 0.443359375, + "learning_rate": 0.00013233734888678326, + "loss": 2.5829, + "step": 18608 + }, + { + "epoch": 0.79, + "grad_norm": 0.447265625, + "learning_rate": 0.00013233100129018732, + "loss": 2.5702, + "step": 18609 + }, + { + "epoch": 0.79, + "grad_norm": 0.427734375, + "learning_rate": 0.00013232465354811365, + "loss": 2.4439, + "step": 18610 + }, + { + "epoch": 0.79, + "grad_norm": 0.482421875, + "learning_rate": 0.0001323183056605909, + "loss": 2.5744, + "step": 18611 + }, + { + "epoch": 0.79, + "grad_norm": 0.419921875, + "learning_rate": 0.00013231195762764754, + "loss": 2.5042, + "step": 18612 + }, + { + "epoch": 0.79, + "grad_norm": 0.44140625, + "learning_rate": 0.00013230560944931224, + "loss": 2.581, + "step": 18613 + }, + { + "epoch": 0.79, + "grad_norm": 0.431640625, + "learning_rate": 0.00013229926112561343, + "loss": 2.4904, + "step": 18614 + }, + { + "epoch": 0.79, + "grad_norm": 0.412109375, + "learning_rate": 0.00013229291265657979, + "loss": 2.5391, + "step": 18615 + }, + { + "epoch": 0.79, + "grad_norm": 0.484375, + "learning_rate": 0.00013228656404223985, + "loss": 2.609, + "step": 18616 + }, + { + "epoch": 0.79, + "grad_norm": 0.4609375, + "learning_rate": 0.0001322802152826222, + "loss": 2.63, + "step": 18617 + }, + { + "epoch": 0.79, + "grad_norm": 0.50390625, + "learning_rate": 0.00013227386637775532, + "loss": 2.5062, + "step": 18618 + }, + { + "epoch": 0.79, + "grad_norm": 0.486328125, + "learning_rate": 0.00013226751732766783, + "loss": 2.5321, + "step": 18619 + }, + { + "epoch": 0.79, + "grad_norm": 0.466796875, + "learning_rate": 0.00013226116813238833, + "loss": 2.721, + "step": 18620 + }, + { + "epoch": 0.79, + "grad_norm": 0.48046875, + "learning_rate": 0.0001322548187919454, + "loss": 2.4994, + "step": 18621 + }, + { + "epoch": 0.79, + "grad_norm": 0.43359375, + "learning_rate": 0.00013224846930636752, + "loss": 2.4909, + "step": 18622 + }, + { + "epoch": 0.79, + "grad_norm": 0.51953125, + "learning_rate": 0.00013224211967568332, + "loss": 2.6644, + "step": 18623 + }, + { + "epoch": 0.79, + "grad_norm": 0.412109375, + "learning_rate": 0.0001322357698999214, + "loss": 2.5663, + "step": 18624 + }, + { + "epoch": 0.79, + "grad_norm": 0.4609375, + "learning_rate": 0.00013222941997911026, + "loss": 2.6064, + "step": 18625 + }, + { + "epoch": 0.79, + "grad_norm": 0.484375, + "learning_rate": 0.00013222306991327853, + "loss": 2.6975, + "step": 18626 + }, + { + "epoch": 0.79, + "grad_norm": 0.421875, + "learning_rate": 0.00013221671970245472, + "loss": 2.6026, + "step": 18627 + }, + { + "epoch": 0.79, + "grad_norm": 0.45703125, + "learning_rate": 0.0001322103693466675, + "loss": 2.483, + "step": 18628 + }, + { + "epoch": 0.79, + "grad_norm": 0.53125, + "learning_rate": 0.0001322040188459454, + "loss": 2.6399, + "step": 18629 + }, + { + "epoch": 0.79, + "grad_norm": 0.50390625, + "learning_rate": 0.0001321976682003169, + "loss": 2.5338, + "step": 18630 + }, + { + "epoch": 0.79, + "grad_norm": 0.44921875, + "learning_rate": 0.0001321913174098107, + "loss": 2.6096, + "step": 18631 + }, + { + "epoch": 0.79, + "grad_norm": 0.55859375, + "learning_rate": 0.00013218496647445538, + "loss": 2.7017, + "step": 18632 + }, + { + "epoch": 0.79, + "grad_norm": 0.46484375, + "learning_rate": 0.00013217861539427942, + "loss": 2.5816, + "step": 18633 + }, + { + "epoch": 0.79, + "grad_norm": 0.470703125, + "learning_rate": 0.0001321722641693115, + "loss": 2.5896, + "step": 18634 + }, + { + "epoch": 0.79, + "grad_norm": 0.40625, + "learning_rate": 0.0001321659127995801, + "loss": 2.4727, + "step": 18635 + }, + { + "epoch": 0.79, + "grad_norm": 0.50390625, + "learning_rate": 0.00013215956128511387, + "loss": 2.6339, + "step": 18636 + }, + { + "epoch": 0.79, + "grad_norm": 0.5234375, + "learning_rate": 0.00013215320962594137, + "loss": 2.6025, + "step": 18637 + }, + { + "epoch": 0.79, + "grad_norm": 0.45703125, + "learning_rate": 0.00013214685782209114, + "loss": 2.5112, + "step": 18638 + }, + { + "epoch": 0.79, + "grad_norm": 0.427734375, + "learning_rate": 0.00013214050587359187, + "loss": 2.6056, + "step": 18639 + }, + { + "epoch": 0.79, + "grad_norm": 0.4140625, + "learning_rate": 0.00013213415378047203, + "loss": 2.5757, + "step": 18640 + }, + { + "epoch": 0.79, + "grad_norm": 0.439453125, + "learning_rate": 0.00013212780154276023, + "loss": 2.7075, + "step": 18641 + }, + { + "epoch": 0.79, + "grad_norm": 0.4609375, + "learning_rate": 0.00013212144916048508, + "loss": 2.58, + "step": 18642 + }, + { + "epoch": 0.79, + "grad_norm": 0.4140625, + "learning_rate": 0.00013211509663367513, + "loss": 2.3969, + "step": 18643 + }, + { + "epoch": 0.79, + "grad_norm": 0.416015625, + "learning_rate": 0.000132108743962359, + "loss": 2.4294, + "step": 18644 + }, + { + "epoch": 0.79, + "grad_norm": 0.484375, + "learning_rate": 0.00013210239114656527, + "loss": 2.556, + "step": 18645 + }, + { + "epoch": 0.79, + "grad_norm": 0.431640625, + "learning_rate": 0.0001320960381863225, + "loss": 2.5708, + "step": 18646 + }, + { + "epoch": 0.79, + "grad_norm": 0.435546875, + "learning_rate": 0.00013208968508165926, + "loss": 2.6077, + "step": 18647 + }, + { + "epoch": 0.79, + "grad_norm": 0.4296875, + "learning_rate": 0.0001320833318326042, + "loss": 2.4453, + "step": 18648 + }, + { + "epoch": 0.79, + "grad_norm": 0.51171875, + "learning_rate": 0.00013207697843918586, + "loss": 2.5331, + "step": 18649 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.00013207062490143282, + "loss": 2.5969, + "step": 18650 + }, + { + "epoch": 0.79, + "grad_norm": 0.46484375, + "learning_rate": 0.00013206427121937372, + "loss": 2.5908, + "step": 18651 + }, + { + "epoch": 0.79, + "grad_norm": 0.46875, + "learning_rate": 0.0001320579173930371, + "loss": 2.7261, + "step": 18652 + }, + { + "epoch": 0.79, + "grad_norm": 0.435546875, + "learning_rate": 0.0001320515634224516, + "loss": 2.4936, + "step": 18653 + }, + { + "epoch": 0.79, + "grad_norm": 0.484375, + "learning_rate": 0.00013204520930764572, + "loss": 2.6436, + "step": 18654 + }, + { + "epoch": 0.79, + "grad_norm": 0.45703125, + "learning_rate": 0.00013203885504864814, + "loss": 2.4712, + "step": 18655 + }, + { + "epoch": 0.79, + "grad_norm": 0.416015625, + "learning_rate": 0.00013203250064548741, + "loss": 2.4728, + "step": 18656 + }, + { + "epoch": 0.8, + "grad_norm": 0.53125, + "learning_rate": 0.00013202614609819215, + "loss": 2.5909, + "step": 18657 + }, + { + "epoch": 0.8, + "grad_norm": 0.5, + "learning_rate": 0.00013201979140679094, + "loss": 2.7363, + "step": 18658 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.00013201343657131238, + "loss": 2.6019, + "step": 18659 + }, + { + "epoch": 0.8, + "grad_norm": 0.490234375, + "learning_rate": 0.00013200708159178502, + "loss": 2.5269, + "step": 18660 + }, + { + "epoch": 0.8, + "grad_norm": 0.4609375, + "learning_rate": 0.00013200072646823752, + "loss": 2.4221, + "step": 18661 + }, + { + "epoch": 0.8, + "grad_norm": 0.49609375, + "learning_rate": 0.00013199437120069843, + "loss": 2.6391, + "step": 18662 + }, + { + "epoch": 0.8, + "grad_norm": 0.494140625, + "learning_rate": 0.00013198801578919635, + "loss": 2.5824, + "step": 18663 + }, + { + "epoch": 0.8, + "grad_norm": 0.46484375, + "learning_rate": 0.00013198166023375993, + "loss": 2.5836, + "step": 18664 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.0001319753045344177, + "loss": 2.5467, + "step": 18665 + }, + { + "epoch": 0.8, + "grad_norm": 0.47265625, + "learning_rate": 0.00013196894869119827, + "loss": 2.588, + "step": 18666 + }, + { + "epoch": 0.8, + "grad_norm": 0.578125, + "learning_rate": 0.00013196259270413028, + "loss": 2.729, + "step": 18667 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.0001319562365732423, + "loss": 2.4688, + "step": 18668 + }, + { + "epoch": 0.8, + "grad_norm": 0.484375, + "learning_rate": 0.00013194988029856293, + "loss": 2.5885, + "step": 18669 + }, + { + "epoch": 0.8, + "grad_norm": 0.5546875, + "learning_rate": 0.00013194352388012077, + "loss": 2.6397, + "step": 18670 + }, + { + "epoch": 0.8, + "grad_norm": 0.4765625, + "learning_rate": 0.0001319371673179444, + "loss": 2.6027, + "step": 18671 + }, + { + "epoch": 0.8, + "grad_norm": 0.462890625, + "learning_rate": 0.00013193081061206252, + "loss": 2.5089, + "step": 18672 + }, + { + "epoch": 0.8, + "grad_norm": 0.44921875, + "learning_rate": 0.0001319244537625036, + "loss": 2.5966, + "step": 18673 + }, + { + "epoch": 0.8, + "grad_norm": 0.494140625, + "learning_rate": 0.00013191809676929632, + "loss": 2.4766, + "step": 18674 + }, + { + "epoch": 0.8, + "grad_norm": 0.423828125, + "learning_rate": 0.00013191173963246928, + "loss": 2.4616, + "step": 18675 + }, + { + "epoch": 0.8, + "grad_norm": 0.375, + "learning_rate": 0.00013190538235205107, + "loss": 2.4453, + "step": 18676 + }, + { + "epoch": 0.8, + "grad_norm": 0.46875, + "learning_rate": 0.00013189902492807031, + "loss": 2.6136, + "step": 18677 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.00013189266736055556, + "loss": 2.6409, + "step": 18678 + }, + { + "epoch": 0.8, + "grad_norm": 0.4453125, + "learning_rate": 0.00013188630964953548, + "loss": 2.5034, + "step": 18679 + }, + { + "epoch": 0.8, + "grad_norm": 0.4453125, + "learning_rate": 0.0001318799517950387, + "loss": 2.6179, + "step": 18680 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.00013187359379709373, + "loss": 2.5473, + "step": 18681 + }, + { + "epoch": 0.8, + "grad_norm": 0.53125, + "learning_rate": 0.00013186723565572928, + "loss": 2.4444, + "step": 18682 + }, + { + "epoch": 0.8, + "grad_norm": 0.4765625, + "learning_rate": 0.0001318608773709739, + "loss": 2.4393, + "step": 18683 + }, + { + "epoch": 0.8, + "grad_norm": 0.44140625, + "learning_rate": 0.00013185451894285618, + "loss": 2.5885, + "step": 18684 + }, + { + "epoch": 0.8, + "grad_norm": 0.49609375, + "learning_rate": 0.00013184816037140482, + "loss": 2.3629, + "step": 18685 + }, + { + "epoch": 0.8, + "grad_norm": 0.58203125, + "learning_rate": 0.00013184180165664834, + "loss": 2.6284, + "step": 18686 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.0001318354427986154, + "loss": 2.6667, + "step": 18687 + }, + { + "epoch": 0.8, + "grad_norm": 0.4375, + "learning_rate": 0.00013182908379733463, + "loss": 2.4365, + "step": 18688 + }, + { + "epoch": 0.8, + "grad_norm": 0.490234375, + "learning_rate": 0.00013182272465283456, + "loss": 2.6018, + "step": 18689 + }, + { + "epoch": 0.8, + "grad_norm": 0.50390625, + "learning_rate": 0.00013181636536514391, + "loss": 2.7172, + "step": 18690 + }, + { + "epoch": 0.8, + "grad_norm": 0.427734375, + "learning_rate": 0.00013181000593429123, + "loss": 2.6643, + "step": 18691 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.00013180364636030512, + "loss": 2.4881, + "step": 18692 + }, + { + "epoch": 0.8, + "grad_norm": 0.43359375, + "learning_rate": 0.00013179728664321428, + "loss": 2.6356, + "step": 18693 + }, + { + "epoch": 0.8, + "grad_norm": 0.462890625, + "learning_rate": 0.00013179092678304723, + "loss": 2.6159, + "step": 18694 + }, + { + "epoch": 0.8, + "grad_norm": 0.45703125, + "learning_rate": 0.00013178456677983262, + "loss": 2.6066, + "step": 18695 + }, + { + "epoch": 0.8, + "grad_norm": 0.46484375, + "learning_rate": 0.0001317782066335991, + "loss": 2.6971, + "step": 18696 + }, + { + "epoch": 0.8, + "grad_norm": 0.484375, + "learning_rate": 0.00013177184634437526, + "loss": 2.431, + "step": 18697 + }, + { + "epoch": 0.8, + "grad_norm": 0.5234375, + "learning_rate": 0.00013176548591218969, + "loss": 2.5667, + "step": 18698 + }, + { + "epoch": 0.8, + "grad_norm": 0.5, + "learning_rate": 0.00013175912533707108, + "loss": 2.4907, + "step": 18699 + }, + { + "epoch": 0.8, + "grad_norm": 0.443359375, + "learning_rate": 0.000131752764619048, + "loss": 2.5561, + "step": 18700 + }, + { + "epoch": 0.8, + "grad_norm": 0.44921875, + "learning_rate": 0.00013174640375814912, + "loss": 2.6052, + "step": 18701 + }, + { + "epoch": 0.8, + "grad_norm": 0.5078125, + "learning_rate": 0.00013174004275440297, + "loss": 2.6527, + "step": 18702 + }, + { + "epoch": 0.8, + "grad_norm": 0.44140625, + "learning_rate": 0.00013173368160783824, + "loss": 2.7049, + "step": 18703 + }, + { + "epoch": 0.8, + "grad_norm": 0.49609375, + "learning_rate": 0.00013172732031848356, + "loss": 2.5705, + "step": 18704 + }, + { + "epoch": 0.8, + "grad_norm": 0.462890625, + "learning_rate": 0.00013172095888636752, + "loss": 2.4918, + "step": 18705 + }, + { + "epoch": 0.8, + "grad_norm": 0.470703125, + "learning_rate": 0.00013171459731151875, + "loss": 2.7866, + "step": 18706 + }, + { + "epoch": 0.8, + "grad_norm": 0.4296875, + "learning_rate": 0.00013170823559396588, + "loss": 2.5616, + "step": 18707 + }, + { + "epoch": 0.8, + "grad_norm": 0.6328125, + "learning_rate": 0.00013170187373373756, + "loss": 2.6194, + "step": 18708 + }, + { + "epoch": 0.8, + "grad_norm": 0.490234375, + "learning_rate": 0.0001316955117308624, + "loss": 2.5514, + "step": 18709 + }, + { + "epoch": 0.8, + "grad_norm": 0.466796875, + "learning_rate": 0.00013168914958536899, + "loss": 2.647, + "step": 18710 + }, + { + "epoch": 0.8, + "grad_norm": 0.423828125, + "learning_rate": 0.00013168278729728598, + "loss": 2.594, + "step": 18711 + }, + { + "epoch": 0.8, + "grad_norm": 0.455078125, + "learning_rate": 0.00013167642486664206, + "loss": 2.6022, + "step": 18712 + }, + { + "epoch": 0.8, + "grad_norm": 0.447265625, + "learning_rate": 0.00013167006229346576, + "loss": 2.6477, + "step": 18713 + }, + { + "epoch": 0.8, + "grad_norm": 0.515625, + "learning_rate": 0.00013166369957778577, + "loss": 2.535, + "step": 18714 + }, + { + "epoch": 0.8, + "grad_norm": 0.466796875, + "learning_rate": 0.0001316573367196307, + "loss": 2.5127, + "step": 18715 + }, + { + "epoch": 0.8, + "grad_norm": 0.486328125, + "learning_rate": 0.00013165097371902918, + "loss": 2.5576, + "step": 18716 + }, + { + "epoch": 0.8, + "grad_norm": 0.470703125, + "learning_rate": 0.00013164461057600985, + "loss": 2.6057, + "step": 18717 + }, + { + "epoch": 0.8, + "grad_norm": 0.4609375, + "learning_rate": 0.00013163824729060136, + "loss": 2.579, + "step": 18718 + }, + { + "epoch": 0.8, + "grad_norm": 0.462890625, + "learning_rate": 0.00013163188386283226, + "loss": 2.582, + "step": 18719 + }, + { + "epoch": 0.8, + "grad_norm": 0.498046875, + "learning_rate": 0.0001316255202927313, + "loss": 2.6571, + "step": 18720 + }, + { + "epoch": 0.8, + "grad_norm": 0.470703125, + "learning_rate": 0.00013161915658032702, + "loss": 2.5706, + "step": 18721 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.0001316127927256481, + "loss": 2.6417, + "step": 18722 + }, + { + "epoch": 0.8, + "grad_norm": 0.439453125, + "learning_rate": 0.00013160642872872316, + "loss": 2.5018, + "step": 18723 + }, + { + "epoch": 0.8, + "grad_norm": 0.4609375, + "learning_rate": 0.00013160006458958086, + "loss": 2.4992, + "step": 18724 + }, + { + "epoch": 0.8, + "grad_norm": 0.404296875, + "learning_rate": 0.0001315937003082498, + "loss": 2.5836, + "step": 18725 + }, + { + "epoch": 0.8, + "grad_norm": 0.51171875, + "learning_rate": 0.00013158733588475869, + "loss": 2.633, + "step": 18726 + }, + { + "epoch": 0.8, + "grad_norm": 0.423828125, + "learning_rate": 0.00013158097131913605, + "loss": 2.628, + "step": 18727 + }, + { + "epoch": 0.8, + "grad_norm": 0.486328125, + "learning_rate": 0.00013157460661141062, + "loss": 2.551, + "step": 18728 + }, + { + "epoch": 0.8, + "grad_norm": 0.44921875, + "learning_rate": 0.00013156824176161096, + "loss": 2.5524, + "step": 18729 + }, + { + "epoch": 0.8, + "grad_norm": 0.451171875, + "learning_rate": 0.00013156187676976577, + "loss": 2.6859, + "step": 18730 + }, + { + "epoch": 0.8, + "grad_norm": 0.46875, + "learning_rate": 0.00013155551163590367, + "loss": 2.5546, + "step": 18731 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.0001315491463600533, + "loss": 2.6104, + "step": 18732 + }, + { + "epoch": 0.8, + "grad_norm": 0.447265625, + "learning_rate": 0.00013154278094224328, + "loss": 2.4663, + "step": 18733 + }, + { + "epoch": 0.8, + "grad_norm": 0.44921875, + "learning_rate": 0.00013153641538250232, + "loss": 2.429, + "step": 18734 + }, + { + "epoch": 0.8, + "grad_norm": 0.53125, + "learning_rate": 0.00013153004968085896, + "loss": 2.6825, + "step": 18735 + }, + { + "epoch": 0.8, + "grad_norm": 0.408203125, + "learning_rate": 0.00013152368383734196, + "loss": 2.535, + "step": 18736 + }, + { + "epoch": 0.8, + "grad_norm": 0.478515625, + "learning_rate": 0.00013151731785197984, + "loss": 2.7106, + "step": 18737 + }, + { + "epoch": 0.8, + "grad_norm": 0.462890625, + "learning_rate": 0.00013151095172480134, + "loss": 2.7166, + "step": 18738 + }, + { + "epoch": 0.8, + "grad_norm": 0.44921875, + "learning_rate": 0.00013150458545583508, + "loss": 2.5269, + "step": 18739 + }, + { + "epoch": 0.8, + "grad_norm": 0.46875, + "learning_rate": 0.00013149821904510968, + "loss": 2.6948, + "step": 18740 + }, + { + "epoch": 0.8, + "grad_norm": 0.48828125, + "learning_rate": 0.0001314918524926538, + "loss": 2.5789, + "step": 18741 + }, + { + "epoch": 0.8, + "grad_norm": 0.443359375, + "learning_rate": 0.00013148548579849614, + "loss": 2.4955, + "step": 18742 + }, + { + "epoch": 0.8, + "grad_norm": 0.46875, + "learning_rate": 0.00013147911896266526, + "loss": 2.658, + "step": 18743 + }, + { + "epoch": 0.8, + "grad_norm": 0.515625, + "learning_rate": 0.00013147275198518987, + "loss": 2.6057, + "step": 18744 + }, + { + "epoch": 0.8, + "grad_norm": 0.421875, + "learning_rate": 0.00013146638486609858, + "loss": 2.4832, + "step": 18745 + }, + { + "epoch": 0.8, + "grad_norm": 0.421875, + "learning_rate": 0.00013146001760542007, + "loss": 2.5813, + "step": 18746 + }, + { + "epoch": 0.8, + "grad_norm": 0.50390625, + "learning_rate": 0.00013145365020318298, + "loss": 2.5896, + "step": 18747 + }, + { + "epoch": 0.8, + "grad_norm": 0.431640625, + "learning_rate": 0.00013144728265941592, + "loss": 2.4182, + "step": 18748 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.0001314409149741476, + "loss": 2.4698, + "step": 18749 + }, + { + "epoch": 0.8, + "grad_norm": 0.498046875, + "learning_rate": 0.0001314345471474067, + "loss": 2.6008, + "step": 18750 + }, + { + "epoch": 0.8, + "grad_norm": 0.42578125, + "learning_rate": 0.00013142817917922176, + "loss": 2.6069, + "step": 18751 + }, + { + "epoch": 0.8, + "grad_norm": 0.5625, + "learning_rate": 0.0001314218110696215, + "loss": 2.5984, + "step": 18752 + }, + { + "epoch": 0.8, + "grad_norm": 0.50390625, + "learning_rate": 0.00013141544281863464, + "loss": 2.5598, + "step": 18753 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.00013140907442628973, + "loss": 2.5349, + "step": 18754 + }, + { + "epoch": 0.8, + "grad_norm": 0.458984375, + "learning_rate": 0.00013140270589261546, + "loss": 2.4429, + "step": 18755 + }, + { + "epoch": 0.8, + "grad_norm": 0.42578125, + "learning_rate": 0.00013139633721764052, + "loss": 2.5011, + "step": 18756 + }, + { + "epoch": 0.8, + "grad_norm": 0.52734375, + "learning_rate": 0.00013138996840139348, + "loss": 2.6401, + "step": 18757 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.00013138359944390312, + "loss": 2.54, + "step": 18758 + }, + { + "epoch": 0.8, + "grad_norm": 0.466796875, + "learning_rate": 0.00013137723034519798, + "loss": 2.5665, + "step": 18759 + }, + { + "epoch": 0.8, + "grad_norm": 0.4296875, + "learning_rate": 0.00013137086110530678, + "loss": 2.4467, + "step": 18760 + }, + { + "epoch": 0.8, + "grad_norm": 0.41015625, + "learning_rate": 0.0001313644917242582, + "loss": 2.4473, + "step": 18761 + }, + { + "epoch": 0.8, + "grad_norm": 0.4921875, + "learning_rate": 0.00013135812220208082, + "loss": 2.6006, + "step": 18762 + }, + { + "epoch": 0.8, + "grad_norm": 0.416015625, + "learning_rate": 0.00013135175253880342, + "loss": 2.479, + "step": 18763 + }, + { + "epoch": 0.8, + "grad_norm": 0.4296875, + "learning_rate": 0.00013134538273445453, + "loss": 2.5766, + "step": 18764 + }, + { + "epoch": 0.8, + "grad_norm": 0.44921875, + "learning_rate": 0.0001313390127890629, + "loss": 2.503, + "step": 18765 + }, + { + "epoch": 0.8, + "grad_norm": 0.48828125, + "learning_rate": 0.00013133264270265716, + "loss": 2.5326, + "step": 18766 + }, + { + "epoch": 0.8, + "grad_norm": 0.4375, + "learning_rate": 0.00013132627247526598, + "loss": 2.6589, + "step": 18767 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.000131319902106918, + "loss": 2.5932, + "step": 18768 + }, + { + "epoch": 0.8, + "grad_norm": 0.4453125, + "learning_rate": 0.00013131353159764192, + "loss": 2.6334, + "step": 18769 + }, + { + "epoch": 0.8, + "grad_norm": 0.482421875, + "learning_rate": 0.0001313071609474664, + "loss": 2.5588, + "step": 18770 + }, + { + "epoch": 0.8, + "grad_norm": 0.5078125, + "learning_rate": 0.0001313007901564201, + "loss": 2.672, + "step": 18771 + }, + { + "epoch": 0.8, + "grad_norm": 0.435546875, + "learning_rate": 0.00013129441922453167, + "loss": 2.4047, + "step": 18772 + }, + { + "epoch": 0.8, + "grad_norm": 0.490234375, + "learning_rate": 0.0001312880481518298, + "loss": 2.6339, + "step": 18773 + }, + { + "epoch": 0.8, + "grad_norm": 0.52734375, + "learning_rate": 0.00013128167693834315, + "loss": 2.6096, + "step": 18774 + }, + { + "epoch": 0.8, + "grad_norm": 0.48828125, + "learning_rate": 0.00013127530558410037, + "loss": 2.486, + "step": 18775 + }, + { + "epoch": 0.8, + "grad_norm": 0.4375, + "learning_rate": 0.00013126893408913016, + "loss": 2.5382, + "step": 18776 + }, + { + "epoch": 0.8, + "grad_norm": 0.50390625, + "learning_rate": 0.00013126256245346118, + "loss": 2.5407, + "step": 18777 + }, + { + "epoch": 0.8, + "grad_norm": 0.42578125, + "learning_rate": 0.0001312561906771221, + "loss": 2.4911, + "step": 18778 + }, + { + "epoch": 0.8, + "grad_norm": 0.45703125, + "learning_rate": 0.00013124981876014156, + "loss": 2.5076, + "step": 18779 + }, + { + "epoch": 0.8, + "grad_norm": 0.48828125, + "learning_rate": 0.00013124344670254826, + "loss": 2.6921, + "step": 18780 + }, + { + "epoch": 0.8, + "grad_norm": 0.447265625, + "learning_rate": 0.00013123707450437086, + "loss": 2.5486, + "step": 18781 + }, + { + "epoch": 0.8, + "grad_norm": 0.40234375, + "learning_rate": 0.00013123070216563808, + "loss": 2.4754, + "step": 18782 + }, + { + "epoch": 0.8, + "grad_norm": 0.42578125, + "learning_rate": 0.0001312243296863785, + "loss": 2.4991, + "step": 18783 + }, + { + "epoch": 0.8, + "grad_norm": 0.443359375, + "learning_rate": 0.00013121795706662087, + "loss": 2.6551, + "step": 18784 + }, + { + "epoch": 0.8, + "grad_norm": 0.4453125, + "learning_rate": 0.00013121158430639385, + "loss": 2.3987, + "step": 18785 + }, + { + "epoch": 0.8, + "grad_norm": 0.478515625, + "learning_rate": 0.00013120521140572614, + "loss": 2.555, + "step": 18786 + }, + { + "epoch": 0.8, + "grad_norm": 0.43359375, + "learning_rate": 0.00013119883836464633, + "loss": 2.4836, + "step": 18787 + }, + { + "epoch": 0.8, + "grad_norm": 0.443359375, + "learning_rate": 0.00013119246518318315, + "loss": 2.5291, + "step": 18788 + }, + { + "epoch": 0.8, + "grad_norm": 0.50390625, + "learning_rate": 0.0001311860918613653, + "loss": 2.6688, + "step": 18789 + }, + { + "epoch": 0.8, + "grad_norm": 0.51953125, + "learning_rate": 0.00013117971839922142, + "loss": 2.5401, + "step": 18790 + }, + { + "epoch": 0.8, + "grad_norm": 0.494140625, + "learning_rate": 0.00013117334479678024, + "loss": 2.6662, + "step": 18791 + }, + { + "epoch": 0.8, + "grad_norm": 0.470703125, + "learning_rate": 0.00013116697105407035, + "loss": 2.5785, + "step": 18792 + }, + { + "epoch": 0.8, + "grad_norm": 0.45703125, + "learning_rate": 0.00013116059717112053, + "loss": 2.6032, + "step": 18793 + }, + { + "epoch": 0.8, + "grad_norm": 0.423828125, + "learning_rate": 0.00013115422314795936, + "loss": 2.5938, + "step": 18794 + }, + { + "epoch": 0.8, + "grad_norm": 0.443359375, + "learning_rate": 0.0001311478489846156, + "loss": 2.4186, + "step": 18795 + }, + { + "epoch": 0.8, + "grad_norm": 0.43359375, + "learning_rate": 0.00013114147468111788, + "loss": 2.5087, + "step": 18796 + }, + { + "epoch": 0.8, + "grad_norm": 0.4921875, + "learning_rate": 0.00013113510023749492, + "loss": 2.5343, + "step": 18797 + }, + { + "epoch": 0.8, + "grad_norm": 0.455078125, + "learning_rate": 0.0001311287256537754, + "loss": 2.5365, + "step": 18798 + }, + { + "epoch": 0.8, + "grad_norm": 0.390625, + "learning_rate": 0.00013112235092998796, + "loss": 2.4598, + "step": 18799 + }, + { + "epoch": 0.8, + "grad_norm": 0.451171875, + "learning_rate": 0.00013111597606616133, + "loss": 2.5949, + "step": 18800 + }, + { + "epoch": 0.8, + "grad_norm": 0.4609375, + "learning_rate": 0.0001311096010623242, + "loss": 2.5585, + "step": 18801 + }, + { + "epoch": 0.8, + "grad_norm": 0.44140625, + "learning_rate": 0.0001311032259185052, + "loss": 2.5573, + "step": 18802 + }, + { + "epoch": 0.8, + "grad_norm": 0.48046875, + "learning_rate": 0.00013109685063473307, + "loss": 2.6517, + "step": 18803 + }, + { + "epoch": 0.8, + "grad_norm": 0.443359375, + "learning_rate": 0.00013109047521103648, + "loss": 2.581, + "step": 18804 + }, + { + "epoch": 0.8, + "grad_norm": 0.451171875, + "learning_rate": 0.0001310840996474441, + "loss": 2.5545, + "step": 18805 + }, + { + "epoch": 0.8, + "grad_norm": 0.49609375, + "learning_rate": 0.00013107772394398465, + "loss": 2.6142, + "step": 18806 + }, + { + "epoch": 0.8, + "grad_norm": 0.47265625, + "learning_rate": 0.0001310713481006868, + "loss": 2.6809, + "step": 18807 + }, + { + "epoch": 0.8, + "grad_norm": 0.44921875, + "learning_rate": 0.0001310649721175792, + "loss": 2.438, + "step": 18808 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.00013105859599469064, + "loss": 2.6349, + "step": 18809 + }, + { + "epoch": 0.8, + "grad_norm": 0.423828125, + "learning_rate": 0.00013105221973204969, + "loss": 2.4825, + "step": 18810 + }, + { + "epoch": 0.8, + "grad_norm": 0.47265625, + "learning_rate": 0.00013104584332968513, + "loss": 2.5154, + "step": 18811 + }, + { + "epoch": 0.8, + "grad_norm": 0.5078125, + "learning_rate": 0.00013103946678762562, + "loss": 2.6165, + "step": 18812 + }, + { + "epoch": 0.8, + "grad_norm": 0.400390625, + "learning_rate": 0.00013103309010589986, + "loss": 2.3482, + "step": 18813 + }, + { + "epoch": 0.8, + "grad_norm": 0.43359375, + "learning_rate": 0.0001310267132845365, + "loss": 2.6751, + "step": 18814 + }, + { + "epoch": 0.8, + "grad_norm": 0.4375, + "learning_rate": 0.0001310203363235643, + "loss": 2.5423, + "step": 18815 + }, + { + "epoch": 0.8, + "grad_norm": 0.46875, + "learning_rate": 0.00013101395922301192, + "loss": 2.5426, + "step": 18816 + }, + { + "epoch": 0.8, + "grad_norm": 0.515625, + "learning_rate": 0.00013100758198290806, + "loss": 2.6158, + "step": 18817 + }, + { + "epoch": 0.8, + "grad_norm": 0.4453125, + "learning_rate": 0.0001310012046032814, + "loss": 2.6465, + "step": 18818 + }, + { + "epoch": 0.8, + "grad_norm": 0.416015625, + "learning_rate": 0.00013099482708416065, + "loss": 2.5579, + "step": 18819 + }, + { + "epoch": 0.8, + "grad_norm": 0.46484375, + "learning_rate": 0.00013098844942557451, + "loss": 2.585, + "step": 18820 + }, + { + "epoch": 0.8, + "grad_norm": 0.515625, + "learning_rate": 0.00013098207162755168, + "loss": 2.6941, + "step": 18821 + }, + { + "epoch": 0.8, + "grad_norm": 0.41796875, + "learning_rate": 0.00013097569369012083, + "loss": 2.5152, + "step": 18822 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.00013096931561331073, + "loss": 2.6664, + "step": 18823 + }, + { + "epoch": 0.8, + "grad_norm": 0.462890625, + "learning_rate": 0.00013096293739714996, + "loss": 2.5331, + "step": 18824 + }, + { + "epoch": 0.8, + "grad_norm": 0.42578125, + "learning_rate": 0.00013095655904166734, + "loss": 2.5347, + "step": 18825 + }, + { + "epoch": 0.8, + "grad_norm": 0.45703125, + "learning_rate": 0.00013095018054689147, + "loss": 2.7439, + "step": 18826 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.00013094380191285113, + "loss": 2.6112, + "step": 18827 + }, + { + "epoch": 0.8, + "grad_norm": 0.466796875, + "learning_rate": 0.000130937423139575, + "loss": 2.5521, + "step": 18828 + }, + { + "epoch": 0.8, + "grad_norm": 0.5234375, + "learning_rate": 0.00013093104422709175, + "loss": 2.6675, + "step": 18829 + }, + { + "epoch": 0.8, + "grad_norm": 0.4140625, + "learning_rate": 0.00013092466517543013, + "loss": 2.5678, + "step": 18830 + }, + { + "epoch": 0.8, + "grad_norm": 0.400390625, + "learning_rate": 0.0001309182859846188, + "loss": 2.6153, + "step": 18831 + }, + { + "epoch": 0.8, + "grad_norm": 0.490234375, + "learning_rate": 0.00013091190665468643, + "loss": 2.5511, + "step": 18832 + }, + { + "epoch": 0.8, + "grad_norm": 0.4921875, + "learning_rate": 0.00013090552718566189, + "loss": 2.4344, + "step": 18833 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.0001308991475775737, + "loss": 2.5172, + "step": 18834 + }, + { + "epoch": 0.8, + "grad_norm": 0.451171875, + "learning_rate": 0.00013089276783045064, + "loss": 2.6317, + "step": 18835 + }, + { + "epoch": 0.8, + "grad_norm": 0.494140625, + "learning_rate": 0.00013088638794432142, + "loss": 2.5752, + "step": 18836 + }, + { + "epoch": 0.8, + "grad_norm": 0.4375, + "learning_rate": 0.00013088000791921476, + "loss": 2.5073, + "step": 18837 + }, + { + "epoch": 0.8, + "grad_norm": 0.453125, + "learning_rate": 0.00013087362775515934, + "loss": 2.6613, + "step": 18838 + }, + { + "epoch": 0.8, + "grad_norm": 0.46484375, + "learning_rate": 0.00013086724745218389, + "loss": 2.4978, + "step": 18839 + }, + { + "epoch": 0.8, + "grad_norm": 0.490234375, + "learning_rate": 0.00013086086701031706, + "loss": 2.6714, + "step": 18840 + }, + { + "epoch": 0.8, + "grad_norm": 0.490234375, + "learning_rate": 0.00013085448642958765, + "loss": 2.7169, + "step": 18841 + }, + { + "epoch": 0.8, + "grad_norm": 0.455078125, + "learning_rate": 0.00013084810571002432, + "loss": 2.7619, + "step": 18842 + }, + { + "epoch": 0.8, + "grad_norm": 0.5234375, + "learning_rate": 0.00013084172485165578, + "loss": 2.5951, + "step": 18843 + }, + { + "epoch": 0.8, + "grad_norm": 0.470703125, + "learning_rate": 0.00013083534385451075, + "loss": 2.5637, + "step": 18844 + }, + { + "epoch": 0.8, + "grad_norm": 0.427734375, + "learning_rate": 0.00013082896271861792, + "loss": 2.4745, + "step": 18845 + }, + { + "epoch": 0.8, + "grad_norm": 0.46875, + "learning_rate": 0.00013082258144400606, + "loss": 2.5541, + "step": 18846 + }, + { + "epoch": 0.8, + "grad_norm": 0.447265625, + "learning_rate": 0.00013081620003070385, + "loss": 2.5851, + "step": 18847 + }, + { + "epoch": 0.8, + "grad_norm": 0.4375, + "learning_rate": 0.00013080981847873998, + "loss": 2.6239, + "step": 18848 + }, + { + "epoch": 0.8, + "grad_norm": 0.4453125, + "learning_rate": 0.00013080343678814323, + "loss": 2.4513, + "step": 18849 + }, + { + "epoch": 0.8, + "grad_norm": 0.52734375, + "learning_rate": 0.00013079705495894222, + "loss": 2.6843, + "step": 18850 + }, + { + "epoch": 0.8, + "grad_norm": 0.484375, + "learning_rate": 0.0001307906729911657, + "loss": 2.6415, + "step": 18851 + }, + { + "epoch": 0.8, + "grad_norm": 0.4375, + "learning_rate": 0.00013078429088484247, + "loss": 2.4967, + "step": 18852 + }, + { + "epoch": 0.8, + "grad_norm": 0.439453125, + "learning_rate": 0.00013077790864000116, + "loss": 2.5363, + "step": 18853 + }, + { + "epoch": 0.8, + "grad_norm": 0.458984375, + "learning_rate": 0.0001307715262566705, + "loss": 2.5187, + "step": 18854 + }, + { + "epoch": 0.8, + "grad_norm": 0.478515625, + "learning_rate": 0.00013076514373487922, + "loss": 2.6333, + "step": 18855 + }, + { + "epoch": 0.8, + "grad_norm": 0.4609375, + "learning_rate": 0.000130758761074656, + "loss": 2.7553, + "step": 18856 + }, + { + "epoch": 0.8, + "grad_norm": 0.42578125, + "learning_rate": 0.00013075237827602965, + "loss": 2.5138, + "step": 18857 + }, + { + "epoch": 0.8, + "grad_norm": 0.46875, + "learning_rate": 0.00013074599533902882, + "loss": 2.5491, + "step": 18858 + }, + { + "epoch": 0.8, + "grad_norm": 0.486328125, + "learning_rate": 0.00013073961226368224, + "loss": 2.5671, + "step": 18859 + }, + { + "epoch": 0.8, + "grad_norm": 0.4375, + "learning_rate": 0.00013073322905001863, + "loss": 2.569, + "step": 18860 + }, + { + "epoch": 0.8, + "grad_norm": 0.484375, + "learning_rate": 0.00013072684569806676, + "loss": 2.6586, + "step": 18861 + }, + { + "epoch": 0.8, + "grad_norm": 0.408203125, + "learning_rate": 0.00013072046220785526, + "loss": 2.6123, + "step": 18862 + }, + { + "epoch": 0.8, + "grad_norm": 0.4921875, + "learning_rate": 0.00013071407857941295, + "loss": 2.6369, + "step": 18863 + }, + { + "epoch": 0.8, + "grad_norm": 0.486328125, + "learning_rate": 0.00013070769481276847, + "loss": 2.7141, + "step": 18864 + }, + { + "epoch": 0.8, + "grad_norm": 0.46484375, + "learning_rate": 0.00013070131090795066, + "loss": 2.6529, + "step": 18865 + }, + { + "epoch": 0.8, + "grad_norm": 0.474609375, + "learning_rate": 0.0001306949268649881, + "loss": 2.5764, + "step": 18866 + }, + { + "epoch": 0.8, + "grad_norm": 0.44140625, + "learning_rate": 0.0001306885426839096, + "loss": 2.6648, + "step": 18867 + }, + { + "epoch": 0.8, + "grad_norm": 0.443359375, + "learning_rate": 0.0001306821583647439, + "loss": 2.536, + "step": 18868 + }, + { + "epoch": 0.8, + "grad_norm": 0.4140625, + "learning_rate": 0.00013067577390751968, + "loss": 2.555, + "step": 18869 + }, + { + "epoch": 0.8, + "grad_norm": 0.435546875, + "learning_rate": 0.0001306693893122657, + "loss": 2.6489, + "step": 18870 + }, + { + "epoch": 0.8, + "grad_norm": 0.50390625, + "learning_rate": 0.00013066300457901067, + "loss": 2.5411, + "step": 18871 + }, + { + "epoch": 0.8, + "grad_norm": 0.5078125, + "learning_rate": 0.0001306566197077833, + "loss": 2.556, + "step": 18872 + }, + { + "epoch": 0.8, + "grad_norm": 0.4453125, + "learning_rate": 0.00013065023469861238, + "loss": 2.4902, + "step": 18873 + }, + { + "epoch": 0.8, + "grad_norm": 0.478515625, + "learning_rate": 0.0001306438495515266, + "loss": 2.5835, + "step": 18874 + }, + { + "epoch": 0.8, + "grad_norm": 0.52734375, + "learning_rate": 0.0001306374642665547, + "loss": 2.4988, + "step": 18875 + }, + { + "epoch": 0.8, + "grad_norm": 0.42578125, + "learning_rate": 0.0001306310788437254, + "loss": 2.5538, + "step": 18876 + }, + { + "epoch": 0.8, + "grad_norm": 0.49609375, + "learning_rate": 0.00013062469328306743, + "loss": 2.7711, + "step": 18877 + }, + { + "epoch": 0.8, + "grad_norm": 0.455078125, + "learning_rate": 0.00013061830758460954, + "loss": 2.5467, + "step": 18878 + }, + { + "epoch": 0.8, + "grad_norm": 0.4765625, + "learning_rate": 0.00013061192174838046, + "loss": 2.4844, + "step": 18879 + }, + { + "epoch": 0.8, + "grad_norm": 0.484375, + "learning_rate": 0.0001306055357744089, + "loss": 2.5842, + "step": 18880 + }, + { + "epoch": 0.8, + "grad_norm": 0.494140625, + "learning_rate": 0.00013059914966272363, + "loss": 2.5671, + "step": 18881 + }, + { + "epoch": 0.8, + "grad_norm": 0.515625, + "learning_rate": 0.00013059276341335335, + "loss": 2.6431, + "step": 18882 + }, + { + "epoch": 0.8, + "grad_norm": 0.443359375, + "learning_rate": 0.00013058637702632683, + "loss": 2.5328, + "step": 18883 + }, + { + "epoch": 0.8, + "grad_norm": 0.478515625, + "learning_rate": 0.0001305799905016728, + "loss": 2.714, + "step": 18884 + }, + { + "epoch": 0.8, + "grad_norm": 0.421875, + "learning_rate": 0.00013057360383942, + "loss": 2.4831, + "step": 18885 + }, + { + "epoch": 0.8, + "grad_norm": 0.416015625, + "learning_rate": 0.00013056721703959712, + "loss": 2.4201, + "step": 18886 + }, + { + "epoch": 0.8, + "grad_norm": 0.43359375, + "learning_rate": 0.00013056083010223296, + "loss": 2.6644, + "step": 18887 + }, + { + "epoch": 0.8, + "grad_norm": 0.46875, + "learning_rate": 0.0001305544430273562, + "loss": 2.6038, + "step": 18888 + }, + { + "epoch": 0.8, + "grad_norm": 0.455078125, + "learning_rate": 0.00013054805581499566, + "loss": 2.5522, + "step": 18889 + }, + { + "epoch": 0.8, + "grad_norm": 0.412109375, + "learning_rate": 0.00013054166846518001, + "loss": 2.6143, + "step": 18890 + }, + { + "epoch": 0.8, + "grad_norm": 0.4296875, + "learning_rate": 0.000130535280977938, + "loss": 2.4805, + "step": 18891 + }, + { + "epoch": 0.81, + "grad_norm": 0.47265625, + "learning_rate": 0.00013052889335329844, + "loss": 2.6176, + "step": 18892 + }, + { + "epoch": 0.81, + "grad_norm": 0.51953125, + "learning_rate": 0.00013052250559128997, + "loss": 2.5833, + "step": 18893 + }, + { + "epoch": 0.81, + "grad_norm": 0.46484375, + "learning_rate": 0.00013051611769194138, + "loss": 2.7856, + "step": 18894 + }, + { + "epoch": 0.81, + "grad_norm": 0.435546875, + "learning_rate": 0.00013050972965528143, + "loss": 2.5275, + "step": 18895 + }, + { + "epoch": 0.81, + "grad_norm": 0.486328125, + "learning_rate": 0.0001305033414813388, + "loss": 2.6061, + "step": 18896 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.00013049695317014236, + "loss": 2.6054, + "step": 18897 + }, + { + "epoch": 0.81, + "grad_norm": 0.4765625, + "learning_rate": 0.00013049056472172073, + "loss": 2.4968, + "step": 18898 + }, + { + "epoch": 0.81, + "grad_norm": 0.453125, + "learning_rate": 0.00013048417613610268, + "loss": 2.5797, + "step": 18899 + }, + { + "epoch": 0.81, + "grad_norm": 0.41015625, + "learning_rate": 0.00013047778741331702, + "loss": 2.5128, + "step": 18900 + }, + { + "epoch": 0.81, + "grad_norm": 0.466796875, + "learning_rate": 0.00013047139855339244, + "loss": 2.582, + "step": 18901 + }, + { + "epoch": 0.81, + "grad_norm": 0.46875, + "learning_rate": 0.0001304650095563577, + "loss": 2.5751, + "step": 18902 + }, + { + "epoch": 0.81, + "grad_norm": 0.439453125, + "learning_rate": 0.00013045862042224154, + "loss": 2.6472, + "step": 18903 + }, + { + "epoch": 0.81, + "grad_norm": 0.48046875, + "learning_rate": 0.00013045223115107273, + "loss": 2.5187, + "step": 18904 + }, + { + "epoch": 0.81, + "grad_norm": 0.451171875, + "learning_rate": 0.00013044584174288004, + "loss": 2.5675, + "step": 18905 + }, + { + "epoch": 0.81, + "grad_norm": 0.443359375, + "learning_rate": 0.00013043945219769215, + "loss": 2.5789, + "step": 18906 + }, + { + "epoch": 0.81, + "grad_norm": 0.53125, + "learning_rate": 0.00013043306251553783, + "loss": 2.5654, + "step": 18907 + }, + { + "epoch": 0.81, + "grad_norm": 0.46875, + "learning_rate": 0.0001304266726964459, + "loss": 2.5239, + "step": 18908 + }, + { + "epoch": 0.81, + "grad_norm": 0.427734375, + "learning_rate": 0.00013042028274044503, + "loss": 2.5433, + "step": 18909 + }, + { + "epoch": 0.81, + "grad_norm": 0.46484375, + "learning_rate": 0.00013041389264756403, + "loss": 2.6646, + "step": 18910 + }, + { + "epoch": 0.81, + "grad_norm": 0.51171875, + "learning_rate": 0.00013040750241783162, + "loss": 2.5903, + "step": 18911 + }, + { + "epoch": 0.81, + "grad_norm": 0.3984375, + "learning_rate": 0.00013040111205127653, + "loss": 2.5697, + "step": 18912 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.0001303947215479276, + "loss": 2.4441, + "step": 18913 + }, + { + "epoch": 0.81, + "grad_norm": 0.39453125, + "learning_rate": 0.0001303883309078135, + "loss": 2.4309, + "step": 18914 + }, + { + "epoch": 0.81, + "grad_norm": 0.44140625, + "learning_rate": 0.00013038194013096298, + "loss": 2.5782, + "step": 18915 + }, + { + "epoch": 0.81, + "grad_norm": 0.5, + "learning_rate": 0.0001303755492174049, + "loss": 2.6384, + "step": 18916 + }, + { + "epoch": 0.81, + "grad_norm": 0.43359375, + "learning_rate": 0.00013036915816716792, + "loss": 2.632, + "step": 18917 + }, + { + "epoch": 0.81, + "grad_norm": 0.515625, + "learning_rate": 0.00013036276698028082, + "loss": 2.6231, + "step": 18918 + }, + { + "epoch": 0.81, + "grad_norm": 0.435546875, + "learning_rate": 0.00013035637565677238, + "loss": 2.6654, + "step": 18919 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00013034998419667135, + "loss": 2.514, + "step": 18920 + }, + { + "epoch": 0.81, + "grad_norm": 0.416015625, + "learning_rate": 0.00013034359260000647, + "loss": 2.6196, + "step": 18921 + }, + { + "epoch": 0.81, + "grad_norm": 0.41796875, + "learning_rate": 0.00013033720086680651, + "loss": 2.4987, + "step": 18922 + }, + { + "epoch": 0.81, + "grad_norm": 0.490234375, + "learning_rate": 0.00013033080899710022, + "loss": 2.5508, + "step": 18923 + }, + { + "epoch": 0.81, + "grad_norm": 0.419921875, + "learning_rate": 0.00013032441699091643, + "loss": 2.5066, + "step": 18924 + }, + { + "epoch": 0.81, + "grad_norm": 0.40625, + "learning_rate": 0.00013031802484828376, + "loss": 2.6686, + "step": 18925 + }, + { + "epoch": 0.81, + "grad_norm": 0.51171875, + "learning_rate": 0.0001303116325692311, + "loss": 2.5674, + "step": 18926 + }, + { + "epoch": 0.81, + "grad_norm": 0.494140625, + "learning_rate": 0.0001303052401537872, + "loss": 2.4229, + "step": 18927 + }, + { + "epoch": 0.81, + "grad_norm": 0.48828125, + "learning_rate": 0.00013029884760198075, + "loss": 2.6728, + "step": 18928 + }, + { + "epoch": 0.81, + "grad_norm": 0.431640625, + "learning_rate": 0.0001302924549138406, + "loss": 2.5032, + "step": 18929 + }, + { + "epoch": 0.81, + "grad_norm": 0.453125, + "learning_rate": 0.00013028606208939544, + "loss": 2.6097, + "step": 18930 + }, + { + "epoch": 0.81, + "grad_norm": 0.515625, + "learning_rate": 0.0001302796691286741, + "loss": 2.4455, + "step": 18931 + }, + { + "epoch": 0.81, + "grad_norm": 0.490234375, + "learning_rate": 0.00013027327603170528, + "loss": 2.6139, + "step": 18932 + }, + { + "epoch": 0.81, + "grad_norm": 0.498046875, + "learning_rate": 0.0001302668827985178, + "loss": 2.7009, + "step": 18933 + }, + { + "epoch": 0.81, + "grad_norm": 0.42578125, + "learning_rate": 0.0001302604894291404, + "loss": 2.6179, + "step": 18934 + }, + { + "epoch": 0.81, + "grad_norm": 0.431640625, + "learning_rate": 0.00013025409592360188, + "loss": 2.5796, + "step": 18935 + }, + { + "epoch": 0.81, + "grad_norm": 0.458984375, + "learning_rate": 0.00013024770228193096, + "loss": 2.3736, + "step": 18936 + }, + { + "epoch": 0.81, + "grad_norm": 0.5, + "learning_rate": 0.00013024130850415645, + "loss": 2.5526, + "step": 18937 + }, + { + "epoch": 0.81, + "grad_norm": 0.474609375, + "learning_rate": 0.0001302349145903071, + "loss": 2.6027, + "step": 18938 + }, + { + "epoch": 0.81, + "grad_norm": 0.462890625, + "learning_rate": 0.00013022852054041167, + "loss": 2.5079, + "step": 18939 + }, + { + "epoch": 0.81, + "grad_norm": 0.48828125, + "learning_rate": 0.00013022212635449898, + "loss": 2.5611, + "step": 18940 + }, + { + "epoch": 0.81, + "grad_norm": 0.439453125, + "learning_rate": 0.00013021573203259772, + "loss": 2.5019, + "step": 18941 + }, + { + "epoch": 0.81, + "grad_norm": 0.427734375, + "learning_rate": 0.00013020933757473672, + "loss": 2.5307, + "step": 18942 + }, + { + "epoch": 0.81, + "grad_norm": 0.455078125, + "learning_rate": 0.00013020294298094476, + "loss": 2.686, + "step": 18943 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.00013019654825125056, + "loss": 2.6294, + "step": 18944 + }, + { + "epoch": 0.81, + "grad_norm": 0.490234375, + "learning_rate": 0.00013019015338568298, + "loss": 2.5783, + "step": 18945 + }, + { + "epoch": 0.81, + "grad_norm": 0.51171875, + "learning_rate": 0.00013018375838427072, + "loss": 2.6604, + "step": 18946 + }, + { + "epoch": 0.81, + "grad_norm": 0.50390625, + "learning_rate": 0.00013017736324704254, + "loss": 2.5566, + "step": 18947 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00013017096797402733, + "loss": 2.4489, + "step": 18948 + }, + { + "epoch": 0.81, + "grad_norm": 0.4609375, + "learning_rate": 0.00013016457256525372, + "loss": 2.6617, + "step": 18949 + }, + { + "epoch": 0.81, + "grad_norm": 0.435546875, + "learning_rate": 0.0001301581770207506, + "loss": 2.5136, + "step": 18950 + }, + { + "epoch": 0.81, + "grad_norm": 0.470703125, + "learning_rate": 0.00013015178134054666, + "loss": 2.5998, + "step": 18951 + }, + { + "epoch": 0.81, + "grad_norm": 0.40234375, + "learning_rate": 0.00013014538552467078, + "loss": 2.5363, + "step": 18952 + }, + { + "epoch": 0.81, + "grad_norm": 0.423828125, + "learning_rate": 0.00013013898957315166, + "loss": 2.4813, + "step": 18953 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.0001301325934860181, + "loss": 2.5828, + "step": 18954 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00013012619726329884, + "loss": 2.649, + "step": 18955 + }, + { + "epoch": 0.81, + "grad_norm": 0.4375, + "learning_rate": 0.00013011980090502277, + "loss": 2.4751, + "step": 18956 + }, + { + "epoch": 0.81, + "grad_norm": 0.498046875, + "learning_rate": 0.00013011340441121855, + "loss": 2.6795, + "step": 18957 + }, + { + "epoch": 0.81, + "grad_norm": 0.5234375, + "learning_rate": 0.00013010700778191505, + "loss": 2.6465, + "step": 18958 + }, + { + "epoch": 0.81, + "grad_norm": 0.490234375, + "learning_rate": 0.000130100611017141, + "loss": 2.692, + "step": 18959 + }, + { + "epoch": 0.81, + "grad_norm": 0.431640625, + "learning_rate": 0.0001300942141169252, + "loss": 2.51, + "step": 18960 + }, + { + "epoch": 0.81, + "grad_norm": 0.439453125, + "learning_rate": 0.00013008781708129648, + "loss": 2.6401, + "step": 18961 + }, + { + "epoch": 0.81, + "grad_norm": 0.4765625, + "learning_rate": 0.00013008141991028354, + "loss": 2.6195, + "step": 18962 + }, + { + "epoch": 0.81, + "grad_norm": 0.4375, + "learning_rate": 0.00013007502260391516, + "loss": 2.5755, + "step": 18963 + }, + { + "epoch": 0.81, + "grad_norm": 0.47265625, + "learning_rate": 0.00013006862516222025, + "loss": 2.6136, + "step": 18964 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.00013006222758522745, + "loss": 2.6055, + "step": 18965 + }, + { + "epoch": 0.81, + "grad_norm": 0.466796875, + "learning_rate": 0.00013005582987296566, + "loss": 2.6366, + "step": 18966 + }, + { + "epoch": 0.81, + "grad_norm": 0.494140625, + "learning_rate": 0.00013004943202546364, + "loss": 2.6048, + "step": 18967 + }, + { + "epoch": 0.81, + "grad_norm": 0.482421875, + "learning_rate": 0.00013004303404275008, + "loss": 2.5511, + "step": 18968 + }, + { + "epoch": 0.81, + "grad_norm": 0.478515625, + "learning_rate": 0.0001300366359248539, + "loss": 2.6663, + "step": 18969 + }, + { + "epoch": 0.81, + "grad_norm": 0.46484375, + "learning_rate": 0.00013003023767180381, + "loss": 2.3942, + "step": 18970 + }, + { + "epoch": 0.81, + "grad_norm": 0.45703125, + "learning_rate": 0.00013002383928362865, + "loss": 2.6471, + "step": 18971 + }, + { + "epoch": 0.81, + "grad_norm": 0.48046875, + "learning_rate": 0.00013001744076035718, + "loss": 2.6632, + "step": 18972 + }, + { + "epoch": 0.81, + "grad_norm": 0.494140625, + "learning_rate": 0.0001300110421020182, + "loss": 2.6054, + "step": 18973 + }, + { + "epoch": 0.81, + "grad_norm": 0.482421875, + "learning_rate": 0.00013000464330864046, + "loss": 2.6328, + "step": 18974 + }, + { + "epoch": 0.81, + "grad_norm": 0.470703125, + "learning_rate": 0.00012999824438025283, + "loss": 2.5406, + "step": 18975 + }, + { + "epoch": 0.81, + "grad_norm": 0.486328125, + "learning_rate": 0.00012999184531688407, + "loss": 2.4895, + "step": 18976 + }, + { + "epoch": 0.81, + "grad_norm": 0.4765625, + "learning_rate": 0.00012998544611856293, + "loss": 2.569, + "step": 18977 + }, + { + "epoch": 0.81, + "grad_norm": 0.392578125, + "learning_rate": 0.00012997904678531828, + "loss": 2.6115, + "step": 18978 + }, + { + "epoch": 0.81, + "grad_norm": 0.4609375, + "learning_rate": 0.00012997264731717882, + "loss": 2.5666, + "step": 18979 + }, + { + "epoch": 0.81, + "grad_norm": 0.453125, + "learning_rate": 0.00012996624771417347, + "loss": 2.5544, + "step": 18980 + }, + { + "epoch": 0.81, + "grad_norm": 0.47265625, + "learning_rate": 0.00012995984797633088, + "loss": 2.5325, + "step": 18981 + }, + { + "epoch": 0.81, + "grad_norm": 0.447265625, + "learning_rate": 0.00012995344810368, + "loss": 2.5483, + "step": 18982 + }, + { + "epoch": 0.81, + "grad_norm": 0.435546875, + "learning_rate": 0.00012994704809624953, + "loss": 2.5941, + "step": 18983 + }, + { + "epoch": 0.81, + "grad_norm": 0.50390625, + "learning_rate": 0.00012994064795406827, + "loss": 2.5509, + "step": 18984 + }, + { + "epoch": 0.81, + "grad_norm": 0.447265625, + "learning_rate": 0.00012993424767716505, + "loss": 2.4867, + "step": 18985 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.00012992784726556864, + "loss": 2.5053, + "step": 18986 + }, + { + "epoch": 0.81, + "grad_norm": 0.4765625, + "learning_rate": 0.00012992144671930783, + "loss": 2.5073, + "step": 18987 + }, + { + "epoch": 0.81, + "grad_norm": 0.462890625, + "learning_rate": 0.0001299150460384115, + "loss": 2.5694, + "step": 18988 + }, + { + "epoch": 0.81, + "grad_norm": 0.427734375, + "learning_rate": 0.00012990864522290836, + "loss": 2.5262, + "step": 18989 + }, + { + "epoch": 0.81, + "grad_norm": 0.439453125, + "learning_rate": 0.00012990224427282727, + "loss": 2.5889, + "step": 18990 + }, + { + "epoch": 0.81, + "grad_norm": 0.416015625, + "learning_rate": 0.000129895843188197, + "loss": 2.5604, + "step": 18991 + }, + { + "epoch": 0.81, + "grad_norm": 0.470703125, + "learning_rate": 0.00012988944196904635, + "loss": 2.4273, + "step": 18992 + }, + { + "epoch": 0.81, + "grad_norm": 0.412109375, + "learning_rate": 0.00012988304061540412, + "loss": 2.6059, + "step": 18993 + }, + { + "epoch": 0.81, + "grad_norm": 0.46875, + "learning_rate": 0.00012987663912729917, + "loss": 2.6145, + "step": 18994 + }, + { + "epoch": 0.81, + "grad_norm": 0.484375, + "learning_rate": 0.00012987023750476025, + "loss": 2.5435, + "step": 18995 + }, + { + "epoch": 0.81, + "grad_norm": 0.482421875, + "learning_rate": 0.00012986383574781617, + "loss": 2.5603, + "step": 18996 + }, + { + "epoch": 0.81, + "grad_norm": 0.443359375, + "learning_rate": 0.00012985743385649575, + "loss": 2.4947, + "step": 18997 + }, + { + "epoch": 0.81, + "grad_norm": 0.5, + "learning_rate": 0.00012985103183082777, + "loss": 2.62, + "step": 18998 + }, + { + "epoch": 0.81, + "grad_norm": 0.453125, + "learning_rate": 0.00012984462967084105, + "loss": 2.5877, + "step": 18999 + }, + { + "epoch": 0.81, + "grad_norm": 0.451171875, + "learning_rate": 0.00012983822737656445, + "loss": 2.6121, + "step": 19000 + }, + { + "epoch": 0.81, + "grad_norm": 0.380859375, + "learning_rate": 0.0001298318249480267, + "loss": 2.4426, + "step": 19001 + }, + { + "epoch": 0.81, + "grad_norm": 0.42578125, + "learning_rate": 0.00012982542238525662, + "loss": 2.5864, + "step": 19002 + }, + { + "epoch": 0.81, + "grad_norm": 0.5078125, + "learning_rate": 0.00012981901968828308, + "loss": 2.6113, + "step": 19003 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.00012981261685713483, + "loss": 2.487, + "step": 19004 + }, + { + "epoch": 0.81, + "grad_norm": 0.4765625, + "learning_rate": 0.00012980621389184068, + "loss": 2.6495, + "step": 19005 + }, + { + "epoch": 0.81, + "grad_norm": 0.49609375, + "learning_rate": 0.0001297998107924295, + "loss": 2.608, + "step": 19006 + }, + { + "epoch": 0.81, + "grad_norm": 0.47265625, + "learning_rate": 0.00012979340755893005, + "loss": 2.6952, + "step": 19007 + }, + { + "epoch": 0.81, + "grad_norm": 0.443359375, + "learning_rate": 0.00012978700419137115, + "loss": 2.497, + "step": 19008 + }, + { + "epoch": 0.81, + "grad_norm": 0.431640625, + "learning_rate": 0.00012978060068978163, + "loss": 2.52, + "step": 19009 + }, + { + "epoch": 0.81, + "grad_norm": 0.42578125, + "learning_rate": 0.00012977419705419027, + "loss": 2.5773, + "step": 19010 + }, + { + "epoch": 0.81, + "grad_norm": 0.490234375, + "learning_rate": 0.0001297677932846259, + "loss": 2.6374, + "step": 19011 + }, + { + "epoch": 0.81, + "grad_norm": 0.45703125, + "learning_rate": 0.00012976138938111738, + "loss": 2.6242, + "step": 19012 + }, + { + "epoch": 0.81, + "grad_norm": 0.484375, + "learning_rate": 0.0001297549853436934, + "loss": 2.4139, + "step": 19013 + }, + { + "epoch": 0.81, + "grad_norm": 0.51953125, + "learning_rate": 0.00012974858117238293, + "loss": 2.6444, + "step": 19014 + }, + { + "epoch": 0.81, + "grad_norm": 0.484375, + "learning_rate": 0.00012974217686721471, + "loss": 2.6892, + "step": 19015 + }, + { + "epoch": 0.81, + "grad_norm": 0.419921875, + "learning_rate": 0.00012973577242821754, + "loss": 2.5835, + "step": 19016 + }, + { + "epoch": 0.81, + "grad_norm": 0.4609375, + "learning_rate": 0.00012972936785542028, + "loss": 2.5538, + "step": 19017 + }, + { + "epoch": 0.81, + "grad_norm": 0.490234375, + "learning_rate": 0.00012972296314885175, + "loss": 2.5917, + "step": 19018 + }, + { + "epoch": 0.81, + "grad_norm": 0.4375, + "learning_rate": 0.0001297165583085407, + "loss": 2.4626, + "step": 19019 + }, + { + "epoch": 0.81, + "grad_norm": 0.46484375, + "learning_rate": 0.00012971015333451604, + "loss": 2.5739, + "step": 19020 + }, + { + "epoch": 0.81, + "grad_norm": 0.45703125, + "learning_rate": 0.0001297037482268065, + "loss": 2.5809, + "step": 19021 + }, + { + "epoch": 0.81, + "grad_norm": 0.486328125, + "learning_rate": 0.00012969734298544098, + "loss": 2.5139, + "step": 19022 + }, + { + "epoch": 0.81, + "grad_norm": 0.50390625, + "learning_rate": 0.00012969093761044828, + "loss": 2.6318, + "step": 19023 + }, + { + "epoch": 0.81, + "grad_norm": 0.453125, + "learning_rate": 0.0001296845321018572, + "loss": 2.2998, + "step": 19024 + }, + { + "epoch": 0.81, + "grad_norm": 0.478515625, + "learning_rate": 0.0001296781264596966, + "loss": 2.5879, + "step": 19025 + }, + { + "epoch": 0.81, + "grad_norm": 0.484375, + "learning_rate": 0.00012967172068399521, + "loss": 2.6956, + "step": 19026 + }, + { + "epoch": 0.81, + "grad_norm": 0.4140625, + "learning_rate": 0.00012966531477478198, + "loss": 2.5263, + "step": 19027 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00012965890873208567, + "loss": 2.5622, + "step": 19028 + }, + { + "epoch": 0.81, + "grad_norm": 0.53125, + "learning_rate": 0.00012965250255593508, + "loss": 2.5675, + "step": 19029 + }, + { + "epoch": 0.81, + "grad_norm": 0.4140625, + "learning_rate": 0.00012964609624635908, + "loss": 2.5892, + "step": 19030 + }, + { + "epoch": 0.81, + "grad_norm": 0.4296875, + "learning_rate": 0.0001296396898033865, + "loss": 2.6195, + "step": 19031 + }, + { + "epoch": 0.81, + "grad_norm": 0.46875, + "learning_rate": 0.0001296332832270461, + "loss": 2.5407, + "step": 19032 + }, + { + "epoch": 0.81, + "grad_norm": 0.50390625, + "learning_rate": 0.0001296268765173668, + "loss": 2.5701, + "step": 19033 + }, + { + "epoch": 0.81, + "grad_norm": 0.40625, + "learning_rate": 0.00012962046967437738, + "loss": 2.5805, + "step": 19034 + }, + { + "epoch": 0.81, + "grad_norm": 0.427734375, + "learning_rate": 0.00012961406269810664, + "loss": 2.6644, + "step": 19035 + }, + { + "epoch": 0.81, + "grad_norm": 0.455078125, + "learning_rate": 0.00012960765558858348, + "loss": 2.7026, + "step": 19036 + }, + { + "epoch": 0.81, + "grad_norm": 0.40234375, + "learning_rate": 0.00012960124834583667, + "loss": 2.5507, + "step": 19037 + }, + { + "epoch": 0.81, + "grad_norm": 0.421875, + "learning_rate": 0.00012959484096989506, + "loss": 2.5657, + "step": 19038 + }, + { + "epoch": 0.81, + "grad_norm": 0.443359375, + "learning_rate": 0.00012958843346078748, + "loss": 2.4644, + "step": 19039 + }, + { + "epoch": 0.81, + "grad_norm": 0.451171875, + "learning_rate": 0.00012958202581854275, + "loss": 2.6558, + "step": 19040 + }, + { + "epoch": 0.81, + "grad_norm": 0.41015625, + "learning_rate": 0.00012957561804318975, + "loss": 2.6198, + "step": 19041 + }, + { + "epoch": 0.81, + "grad_norm": 0.4375, + "learning_rate": 0.00012956921013475726, + "loss": 2.6447, + "step": 19042 + }, + { + "epoch": 0.81, + "grad_norm": 0.482421875, + "learning_rate": 0.00012956280209327413, + "loss": 2.5049, + "step": 19043 + }, + { + "epoch": 0.81, + "grad_norm": 0.427734375, + "learning_rate": 0.00012955639391876918, + "loss": 2.5463, + "step": 19044 + }, + { + "epoch": 0.81, + "grad_norm": 0.43359375, + "learning_rate": 0.00012954998561127127, + "loss": 2.4846, + "step": 19045 + }, + { + "epoch": 0.81, + "grad_norm": 0.498046875, + "learning_rate": 0.0001295435771708092, + "loss": 2.5014, + "step": 19046 + }, + { + "epoch": 0.81, + "grad_norm": 0.44140625, + "learning_rate": 0.00012953716859741187, + "loss": 2.6443, + "step": 19047 + }, + { + "epoch": 0.81, + "grad_norm": 0.451171875, + "learning_rate": 0.00012953075989110806, + "loss": 2.5391, + "step": 19048 + }, + { + "epoch": 0.81, + "grad_norm": 0.470703125, + "learning_rate": 0.00012952435105192663, + "loss": 2.658, + "step": 19049 + }, + { + "epoch": 0.81, + "grad_norm": 0.47265625, + "learning_rate": 0.0001295179420798964, + "loss": 2.5526, + "step": 19050 + }, + { + "epoch": 0.81, + "grad_norm": 0.458984375, + "learning_rate": 0.00012951153297504622, + "loss": 2.7075, + "step": 19051 + }, + { + "epoch": 0.81, + "grad_norm": 0.453125, + "learning_rate": 0.00012950512373740493, + "loss": 2.5433, + "step": 19052 + }, + { + "epoch": 0.81, + "grad_norm": 0.4296875, + "learning_rate": 0.00012949871436700137, + "loss": 2.5314, + "step": 19053 + }, + { + "epoch": 0.81, + "grad_norm": 0.5078125, + "learning_rate": 0.00012949230486386436, + "loss": 2.7609, + "step": 19054 + }, + { + "epoch": 0.81, + "grad_norm": 0.455078125, + "learning_rate": 0.0001294858952280228, + "loss": 2.6591, + "step": 19055 + }, + { + "epoch": 0.81, + "grad_norm": 0.423828125, + "learning_rate": 0.00012947948545950544, + "loss": 2.505, + "step": 19056 + }, + { + "epoch": 0.81, + "grad_norm": 0.447265625, + "learning_rate": 0.00012947307555834117, + "loss": 2.3893, + "step": 19057 + }, + { + "epoch": 0.81, + "grad_norm": 0.466796875, + "learning_rate": 0.00012946666552455884, + "loss": 2.5861, + "step": 19058 + }, + { + "epoch": 0.81, + "grad_norm": 0.45703125, + "learning_rate": 0.0001294602553581873, + "loss": 2.5973, + "step": 19059 + }, + { + "epoch": 0.81, + "grad_norm": 0.4375, + "learning_rate": 0.00012945384505925537, + "loss": 2.4753, + "step": 19060 + }, + { + "epoch": 0.81, + "grad_norm": 0.45703125, + "learning_rate": 0.0001294474346277919, + "loss": 2.5509, + "step": 19061 + }, + { + "epoch": 0.81, + "grad_norm": 0.4140625, + "learning_rate": 0.0001294410240638257, + "loss": 2.6306, + "step": 19062 + }, + { + "epoch": 0.81, + "grad_norm": 0.455078125, + "learning_rate": 0.0001294346133673857, + "loss": 2.5902, + "step": 19063 + }, + { + "epoch": 0.81, + "grad_norm": 0.486328125, + "learning_rate": 0.00012942820253850068, + "loss": 2.5368, + "step": 19064 + }, + { + "epoch": 0.81, + "grad_norm": 0.51171875, + "learning_rate": 0.00012942179157719952, + "loss": 2.5969, + "step": 19065 + }, + { + "epoch": 0.81, + "grad_norm": 0.431640625, + "learning_rate": 0.00012941538048351104, + "loss": 2.6638, + "step": 19066 + }, + { + "epoch": 0.81, + "grad_norm": 0.451171875, + "learning_rate": 0.00012940896925746408, + "loss": 2.6902, + "step": 19067 + }, + { + "epoch": 0.81, + "grad_norm": 0.423828125, + "learning_rate": 0.00012940255789908753, + "loss": 2.5935, + "step": 19068 + }, + { + "epoch": 0.81, + "grad_norm": 0.474609375, + "learning_rate": 0.0001293961464084102, + "loss": 2.6657, + "step": 19069 + }, + { + "epoch": 0.81, + "grad_norm": 0.478515625, + "learning_rate": 0.00012938973478546096, + "loss": 2.494, + "step": 19070 + }, + { + "epoch": 0.81, + "grad_norm": 0.482421875, + "learning_rate": 0.00012938332303026866, + "loss": 2.5198, + "step": 19071 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00012937691114286212, + "loss": 2.5719, + "step": 19072 + }, + { + "epoch": 0.81, + "grad_norm": 0.451171875, + "learning_rate": 0.00012937049912327024, + "loss": 2.4992, + "step": 19073 + }, + { + "epoch": 0.81, + "grad_norm": 0.515625, + "learning_rate": 0.00012936408697152184, + "loss": 2.6125, + "step": 19074 + }, + { + "epoch": 0.81, + "grad_norm": 0.53515625, + "learning_rate": 0.00012935767468764576, + "loss": 2.6021, + "step": 19075 + }, + { + "epoch": 0.81, + "grad_norm": 0.46875, + "learning_rate": 0.00012935126227167092, + "loss": 2.5203, + "step": 19076 + }, + { + "epoch": 0.81, + "grad_norm": 0.416015625, + "learning_rate": 0.00012934484972362608, + "loss": 2.607, + "step": 19077 + }, + { + "epoch": 0.81, + "grad_norm": 0.478515625, + "learning_rate": 0.00012933843704354017, + "loss": 2.6568, + "step": 19078 + }, + { + "epoch": 0.81, + "grad_norm": 0.41796875, + "learning_rate": 0.000129332024231442, + "loss": 2.4716, + "step": 19079 + }, + { + "epoch": 0.81, + "grad_norm": 0.419921875, + "learning_rate": 0.00012932561128736045, + "loss": 2.4739, + "step": 19080 + }, + { + "epoch": 0.81, + "grad_norm": 0.4375, + "learning_rate": 0.00012931919821132436, + "loss": 2.6094, + "step": 19081 + }, + { + "epoch": 0.81, + "grad_norm": 0.51171875, + "learning_rate": 0.0001293127850033626, + "loss": 2.4947, + "step": 19082 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00012930637166350402, + "loss": 2.5382, + "step": 19083 + }, + { + "epoch": 0.81, + "grad_norm": 0.53515625, + "learning_rate": 0.00012929995819177748, + "loss": 2.5872, + "step": 19084 + }, + { + "epoch": 0.81, + "grad_norm": 0.443359375, + "learning_rate": 0.00012929354458821184, + "loss": 2.5468, + "step": 19085 + }, + { + "epoch": 0.81, + "grad_norm": 0.419921875, + "learning_rate": 0.00012928713085283593, + "loss": 2.5514, + "step": 19086 + }, + { + "epoch": 0.81, + "grad_norm": 0.5390625, + "learning_rate": 0.00012928071698567866, + "loss": 2.6393, + "step": 19087 + }, + { + "epoch": 0.81, + "grad_norm": 0.451171875, + "learning_rate": 0.00012927430298676884, + "loss": 2.6498, + "step": 19088 + }, + { + "epoch": 0.81, + "grad_norm": 0.48046875, + "learning_rate": 0.00012926788885613536, + "loss": 2.5417, + "step": 19089 + }, + { + "epoch": 0.81, + "grad_norm": 0.484375, + "learning_rate": 0.00012926147459380708, + "loss": 2.6013, + "step": 19090 + }, + { + "epoch": 0.81, + "grad_norm": 0.466796875, + "learning_rate": 0.00012925506019981288, + "loss": 2.5867, + "step": 19091 + }, + { + "epoch": 0.81, + "grad_norm": 0.41015625, + "learning_rate": 0.00012924864567418157, + "loss": 2.5078, + "step": 19092 + }, + { + "epoch": 0.81, + "grad_norm": 0.466796875, + "learning_rate": 0.00012924223101694207, + "loss": 2.5975, + "step": 19093 + }, + { + "epoch": 0.81, + "grad_norm": 0.443359375, + "learning_rate": 0.0001292358162281232, + "loss": 2.4297, + "step": 19094 + }, + { + "epoch": 0.81, + "grad_norm": 0.43359375, + "learning_rate": 0.00012922940130775384, + "loss": 2.4227, + "step": 19095 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.00012922298625586286, + "loss": 2.6263, + "step": 19096 + }, + { + "epoch": 0.81, + "grad_norm": 0.50390625, + "learning_rate": 0.00012921657107247913, + "loss": 2.6534, + "step": 19097 + }, + { + "epoch": 0.81, + "grad_norm": 0.482421875, + "learning_rate": 0.0001292101557576315, + "loss": 2.6005, + "step": 19098 + }, + { + "epoch": 0.81, + "grad_norm": 0.486328125, + "learning_rate": 0.00012920374031134886, + "loss": 2.6033, + "step": 19099 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00012919732473366005, + "loss": 2.552, + "step": 19100 + }, + { + "epoch": 0.81, + "grad_norm": 0.44140625, + "learning_rate": 0.00012919090902459394, + "loss": 2.6097, + "step": 19101 + }, + { + "epoch": 0.81, + "grad_norm": 0.48828125, + "learning_rate": 0.0001291844931841794, + "loss": 2.7392, + "step": 19102 + }, + { + "epoch": 0.81, + "grad_norm": 0.48046875, + "learning_rate": 0.00012917807721244534, + "loss": 2.6502, + "step": 19103 + }, + { + "epoch": 0.81, + "grad_norm": 0.498046875, + "learning_rate": 0.00012917166110942057, + "loss": 2.5832, + "step": 19104 + }, + { + "epoch": 0.81, + "grad_norm": 0.453125, + "learning_rate": 0.00012916524487513396, + "loss": 2.5728, + "step": 19105 + }, + { + "epoch": 0.81, + "grad_norm": 0.5390625, + "learning_rate": 0.00012915882850961444, + "loss": 2.6028, + "step": 19106 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.00012915241201289087, + "loss": 2.5258, + "step": 19107 + }, + { + "epoch": 0.81, + "grad_norm": 0.462890625, + "learning_rate": 0.0001291459953849921, + "loss": 2.5096, + "step": 19108 + }, + { + "epoch": 0.81, + "grad_norm": 0.44921875, + "learning_rate": 0.00012913957862594697, + "loss": 2.5359, + "step": 19109 + }, + { + "epoch": 0.81, + "grad_norm": 0.4609375, + "learning_rate": 0.00012913316173578435, + "loss": 2.6945, + "step": 19110 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00012912674471453323, + "loss": 2.6751, + "step": 19111 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.00012912032756222237, + "loss": 2.5448, + "step": 19112 + }, + { + "epoch": 0.81, + "grad_norm": 0.4609375, + "learning_rate": 0.00012911391027888066, + "loss": 2.633, + "step": 19113 + }, + { + "epoch": 0.81, + "grad_norm": 0.431640625, + "learning_rate": 0.000129107492864537, + "loss": 2.5812, + "step": 19114 + }, + { + "epoch": 0.81, + "grad_norm": 0.453125, + "learning_rate": 0.0001291010753192203, + "loss": 2.6489, + "step": 19115 + }, + { + "epoch": 0.81, + "grad_norm": 0.46875, + "learning_rate": 0.00012909465764295936, + "loss": 2.6727, + "step": 19116 + }, + { + "epoch": 0.81, + "grad_norm": 0.439453125, + "learning_rate": 0.0001290882398357831, + "loss": 2.5723, + "step": 19117 + }, + { + "epoch": 0.81, + "grad_norm": 0.443359375, + "learning_rate": 0.00012908182189772037, + "loss": 2.5021, + "step": 19118 + }, + { + "epoch": 0.81, + "grad_norm": 0.4453125, + "learning_rate": 0.0001290754038288001, + "loss": 2.5669, + "step": 19119 + }, + { + "epoch": 0.81, + "grad_norm": 0.4609375, + "learning_rate": 0.00012906898562905112, + "loss": 2.5638, + "step": 19120 + }, + { + "epoch": 0.81, + "grad_norm": 0.462890625, + "learning_rate": 0.00012906256729850233, + "loss": 2.5305, + "step": 19121 + }, + { + "epoch": 0.81, + "grad_norm": 0.5, + "learning_rate": 0.00012905614883718263, + "loss": 2.6428, + "step": 19122 + }, + { + "epoch": 0.81, + "grad_norm": 0.470703125, + "learning_rate": 0.00012904973024512086, + "loss": 2.5739, + "step": 19123 + }, + { + "epoch": 0.81, + "grad_norm": 0.447265625, + "learning_rate": 0.00012904331152234591, + "loss": 2.5579, + "step": 19124 + }, + { + "epoch": 0.81, + "grad_norm": 0.5078125, + "learning_rate": 0.0001290368926688867, + "loss": 2.5619, + "step": 19125 + }, + { + "epoch": 0.82, + "grad_norm": 0.447265625, + "learning_rate": 0.00012903047368477205, + "loss": 2.5436, + "step": 19126 + }, + { + "epoch": 0.82, + "grad_norm": 0.435546875, + "learning_rate": 0.00012902405457003094, + "loss": 2.5567, + "step": 19127 + }, + { + "epoch": 0.82, + "grad_norm": 0.39453125, + "learning_rate": 0.00012901763532469212, + "loss": 2.5608, + "step": 19128 + }, + { + "epoch": 0.82, + "grad_norm": 0.46484375, + "learning_rate": 0.00012901121594878457, + "loss": 2.5304, + "step": 19129 + }, + { + "epoch": 0.82, + "grad_norm": 0.482421875, + "learning_rate": 0.00012900479644233716, + "loss": 2.6023, + "step": 19130 + }, + { + "epoch": 0.82, + "grad_norm": 0.427734375, + "learning_rate": 0.00012899837680537877, + "loss": 2.6543, + "step": 19131 + }, + { + "epoch": 0.82, + "grad_norm": 0.484375, + "learning_rate": 0.00012899195703793828, + "loss": 2.5358, + "step": 19132 + }, + { + "epoch": 0.82, + "grad_norm": 0.46484375, + "learning_rate": 0.00012898553714004457, + "loss": 2.5595, + "step": 19133 + }, + { + "epoch": 0.82, + "grad_norm": 0.515625, + "learning_rate": 0.00012897911711172653, + "loss": 2.631, + "step": 19134 + }, + { + "epoch": 0.82, + "grad_norm": 0.46875, + "learning_rate": 0.00012897269695301308, + "loss": 2.658, + "step": 19135 + }, + { + "epoch": 0.82, + "grad_norm": 0.470703125, + "learning_rate": 0.00012896627666393305, + "loss": 2.5124, + "step": 19136 + }, + { + "epoch": 0.82, + "grad_norm": 0.490234375, + "learning_rate": 0.0001289598562445154, + "loss": 2.6493, + "step": 19137 + }, + { + "epoch": 0.82, + "grad_norm": 0.453125, + "learning_rate": 0.00012895343569478893, + "loss": 2.5576, + "step": 19138 + }, + { + "epoch": 0.82, + "grad_norm": 0.435546875, + "learning_rate": 0.0001289470150147826, + "loss": 2.6183, + "step": 19139 + }, + { + "epoch": 0.82, + "grad_norm": 0.55078125, + "learning_rate": 0.00012894059420452531, + "loss": 2.4642, + "step": 19140 + }, + { + "epoch": 0.82, + "grad_norm": 0.5, + "learning_rate": 0.00012893417326404592, + "loss": 2.5403, + "step": 19141 + }, + { + "epoch": 0.82, + "grad_norm": 0.447265625, + "learning_rate": 0.00012892775219337327, + "loss": 2.4828, + "step": 19142 + }, + { + "epoch": 0.82, + "grad_norm": 0.466796875, + "learning_rate": 0.00012892133099253635, + "loss": 2.6287, + "step": 19143 + }, + { + "epoch": 0.82, + "grad_norm": 0.4609375, + "learning_rate": 0.000128914909661564, + "loss": 2.5668, + "step": 19144 + }, + { + "epoch": 0.82, + "grad_norm": 0.6015625, + "learning_rate": 0.00012890848820048514, + "loss": 2.6369, + "step": 19145 + }, + { + "epoch": 0.82, + "grad_norm": 0.439453125, + "learning_rate": 0.00012890206660932861, + "loss": 2.5115, + "step": 19146 + }, + { + "epoch": 0.82, + "grad_norm": 0.47265625, + "learning_rate": 0.0001288956448881234, + "loss": 2.5646, + "step": 19147 + }, + { + "epoch": 0.82, + "grad_norm": 0.474609375, + "learning_rate": 0.0001288892230368983, + "loss": 2.4663, + "step": 19148 + }, + { + "epoch": 0.82, + "grad_norm": 0.50390625, + "learning_rate": 0.00012888280105568227, + "loss": 2.5325, + "step": 19149 + }, + { + "epoch": 0.82, + "grad_norm": 0.431640625, + "learning_rate": 0.00012887637894450418, + "loss": 2.4349, + "step": 19150 + }, + { + "epoch": 0.82, + "grad_norm": 0.419921875, + "learning_rate": 0.00012886995670339298, + "loss": 2.5394, + "step": 19151 + }, + { + "epoch": 0.82, + "grad_norm": 0.423828125, + "learning_rate": 0.00012886353433237747, + "loss": 2.4596, + "step": 19152 + }, + { + "epoch": 0.82, + "grad_norm": 0.5234375, + "learning_rate": 0.00012885711183148663, + "loss": 2.6774, + "step": 19153 + }, + { + "epoch": 0.82, + "grad_norm": 0.494140625, + "learning_rate": 0.00012885068920074933, + "loss": 2.4055, + "step": 19154 + }, + { + "epoch": 0.82, + "grad_norm": 0.458984375, + "learning_rate": 0.00012884426644019448, + "loss": 2.6132, + "step": 19155 + }, + { + "epoch": 0.82, + "grad_norm": 0.447265625, + "learning_rate": 0.00012883784354985096, + "loss": 2.5873, + "step": 19156 + }, + { + "epoch": 0.82, + "grad_norm": 0.458984375, + "learning_rate": 0.0001288314205297477, + "loss": 2.3584, + "step": 19157 + }, + { + "epoch": 0.82, + "grad_norm": 0.47265625, + "learning_rate": 0.00012882499737991356, + "loss": 2.5674, + "step": 19158 + }, + { + "epoch": 0.82, + "grad_norm": 0.4765625, + "learning_rate": 0.00012881857410037748, + "loss": 2.4675, + "step": 19159 + }, + { + "epoch": 0.82, + "grad_norm": 0.451171875, + "learning_rate": 0.00012881215069116838, + "loss": 2.7235, + "step": 19160 + }, + { + "epoch": 0.82, + "grad_norm": 0.494140625, + "learning_rate": 0.00012880572715231504, + "loss": 2.5196, + "step": 19161 + }, + { + "epoch": 0.82, + "grad_norm": 0.44921875, + "learning_rate": 0.00012879930348384656, + "loss": 2.5697, + "step": 19162 + }, + { + "epoch": 0.82, + "grad_norm": 0.447265625, + "learning_rate": 0.00012879287968579165, + "loss": 2.5897, + "step": 19163 + }, + { + "epoch": 0.82, + "grad_norm": 0.453125, + "learning_rate": 0.00012878645575817936, + "loss": 2.6238, + "step": 19164 + }, + { + "epoch": 0.82, + "grad_norm": 0.451171875, + "learning_rate": 0.00012878003170103855, + "loss": 2.5087, + "step": 19165 + }, + { + "epoch": 0.82, + "grad_norm": 0.46875, + "learning_rate": 0.00012877360751439808, + "loss": 2.5767, + "step": 19166 + }, + { + "epoch": 0.82, + "grad_norm": 0.48046875, + "learning_rate": 0.0001287671831982869, + "loss": 2.5584, + "step": 19167 + }, + { + "epoch": 0.82, + "grad_norm": 0.427734375, + "learning_rate": 0.00012876075875273394, + "loss": 2.6018, + "step": 19168 + }, + { + "epoch": 0.82, + "grad_norm": 0.51953125, + "learning_rate": 0.000128754334177768, + "loss": 2.69, + "step": 19169 + }, + { + "epoch": 0.82, + "grad_norm": 0.4375, + "learning_rate": 0.00012874790947341814, + "loss": 2.5896, + "step": 19170 + }, + { + "epoch": 0.82, + "grad_norm": 0.46875, + "learning_rate": 0.00012874148463971316, + "loss": 2.6286, + "step": 19171 + }, + { + "epoch": 0.82, + "grad_norm": 0.45703125, + "learning_rate": 0.000128735059676682, + "loss": 2.4673, + "step": 19172 + }, + { + "epoch": 0.82, + "grad_norm": 0.470703125, + "learning_rate": 0.0001287286345843536, + "loss": 2.5854, + "step": 19173 + }, + { + "epoch": 0.82, + "grad_norm": 0.416015625, + "learning_rate": 0.00012872220936275682, + "loss": 2.4315, + "step": 19174 + }, + { + "epoch": 0.82, + "grad_norm": 0.515625, + "learning_rate": 0.00012871578401192063, + "loss": 2.7219, + "step": 19175 + }, + { + "epoch": 0.82, + "grad_norm": 0.462890625, + "learning_rate": 0.00012870935853187386, + "loss": 2.4913, + "step": 19176 + }, + { + "epoch": 0.82, + "grad_norm": 0.455078125, + "learning_rate": 0.00012870293292264546, + "loss": 2.5943, + "step": 19177 + }, + { + "epoch": 0.82, + "grad_norm": 0.4921875, + "learning_rate": 0.0001286965071842644, + "loss": 2.5677, + "step": 19178 + }, + { + "epoch": 0.82, + "grad_norm": 0.44921875, + "learning_rate": 0.0001286900813167595, + "loss": 2.6025, + "step": 19179 + }, + { + "epoch": 0.82, + "grad_norm": 0.462890625, + "learning_rate": 0.00012868365532015975, + "loss": 2.6013, + "step": 19180 + }, + { + "epoch": 0.82, + "grad_norm": 0.41796875, + "learning_rate": 0.00012867722919449403, + "loss": 2.64, + "step": 19181 + }, + { + "epoch": 0.82, + "grad_norm": 0.490234375, + "learning_rate": 0.00012867080293979126, + "loss": 2.5784, + "step": 19182 + }, + { + "epoch": 0.82, + "grad_norm": 0.494140625, + "learning_rate": 0.00012866437655608036, + "loss": 2.4785, + "step": 19183 + }, + { + "epoch": 0.82, + "grad_norm": 0.4296875, + "learning_rate": 0.00012865795004339024, + "loss": 2.5978, + "step": 19184 + }, + { + "epoch": 0.82, + "grad_norm": 0.455078125, + "learning_rate": 0.00012865152340174978, + "loss": 2.5387, + "step": 19185 + }, + { + "epoch": 0.82, + "grad_norm": 0.40625, + "learning_rate": 0.00012864509663118798, + "loss": 2.5542, + "step": 19186 + }, + { + "epoch": 0.82, + "grad_norm": 0.47265625, + "learning_rate": 0.00012863866973173368, + "loss": 2.698, + "step": 19187 + }, + { + "epoch": 0.82, + "grad_norm": 0.427734375, + "learning_rate": 0.00012863224270341585, + "loss": 2.6124, + "step": 19188 + }, + { + "epoch": 0.82, + "grad_norm": 0.421875, + "learning_rate": 0.0001286258155462634, + "loss": 2.5538, + "step": 19189 + }, + { + "epoch": 0.82, + "grad_norm": 0.498046875, + "learning_rate": 0.00012861938826030524, + "loss": 2.5585, + "step": 19190 + }, + { + "epoch": 0.82, + "grad_norm": 0.427734375, + "learning_rate": 0.0001286129608455703, + "loss": 2.5345, + "step": 19191 + }, + { + "epoch": 0.82, + "grad_norm": 0.451171875, + "learning_rate": 0.00012860653330208746, + "loss": 2.535, + "step": 19192 + }, + { + "epoch": 0.82, + "grad_norm": 0.455078125, + "learning_rate": 0.00012860010562988568, + "loss": 2.4888, + "step": 19193 + }, + { + "epoch": 0.82, + "grad_norm": 0.4453125, + "learning_rate": 0.00012859367782899392, + "loss": 2.5688, + "step": 19194 + }, + { + "epoch": 0.82, + "grad_norm": 0.490234375, + "learning_rate": 0.000128587249899441, + "loss": 2.6109, + "step": 19195 + }, + { + "epoch": 0.82, + "grad_norm": 0.408203125, + "learning_rate": 0.00012858082184125593, + "loss": 2.5254, + "step": 19196 + }, + { + "epoch": 0.82, + "grad_norm": 0.546875, + "learning_rate": 0.00012857439365446762, + "loss": 2.6192, + "step": 19197 + }, + { + "epoch": 0.82, + "grad_norm": 0.51953125, + "learning_rate": 0.000128567965339105, + "loss": 2.5248, + "step": 19198 + }, + { + "epoch": 0.82, + "grad_norm": 0.44140625, + "learning_rate": 0.0001285615368951969, + "loss": 2.6125, + "step": 19199 + }, + { + "epoch": 0.82, + "grad_norm": 0.482421875, + "learning_rate": 0.0001285551083227724, + "loss": 2.5576, + "step": 19200 + }, + { + "epoch": 0.82, + "grad_norm": 0.455078125, + "learning_rate": 0.00012854867962186028, + "loss": 2.6516, + "step": 19201 + }, + { + "epoch": 0.82, + "grad_norm": 0.482421875, + "learning_rate": 0.0001285422507924896, + "loss": 2.5618, + "step": 19202 + }, + { + "epoch": 0.82, + "grad_norm": 0.435546875, + "learning_rate": 0.00012853582183468918, + "loss": 2.5122, + "step": 19203 + }, + { + "epoch": 0.82, + "grad_norm": 0.42578125, + "learning_rate": 0.000128529392748488, + "loss": 2.5892, + "step": 19204 + }, + { + "epoch": 0.82, + "grad_norm": 0.49609375, + "learning_rate": 0.000128522963533915, + "loss": 2.65, + "step": 19205 + }, + { + "epoch": 0.82, + "grad_norm": 0.453125, + "learning_rate": 0.00012851653419099907, + "loss": 2.5652, + "step": 19206 + }, + { + "epoch": 0.82, + "grad_norm": 0.53125, + "learning_rate": 0.00012851010471976915, + "loss": 2.6277, + "step": 19207 + }, + { + "epoch": 0.82, + "grad_norm": 0.482421875, + "learning_rate": 0.00012850367512025422, + "loss": 2.6109, + "step": 19208 + }, + { + "epoch": 0.82, + "grad_norm": 0.44921875, + "learning_rate": 0.0001284972453924831, + "loss": 2.5711, + "step": 19209 + }, + { + "epoch": 0.82, + "grad_norm": 0.53515625, + "learning_rate": 0.00012849081553648485, + "loss": 2.513, + "step": 19210 + }, + { + "epoch": 0.82, + "grad_norm": 0.5859375, + "learning_rate": 0.0001284843855522883, + "loss": 2.6251, + "step": 19211 + }, + { + "epoch": 0.82, + "grad_norm": 0.498046875, + "learning_rate": 0.00012847795543992246, + "loss": 2.6768, + "step": 19212 + }, + { + "epoch": 0.82, + "grad_norm": 0.453125, + "learning_rate": 0.0001284715251994162, + "loss": 2.4123, + "step": 19213 + }, + { + "epoch": 0.82, + "grad_norm": 0.453125, + "learning_rate": 0.00012846509483079848, + "loss": 2.6657, + "step": 19214 + }, + { + "epoch": 0.82, + "grad_norm": 0.466796875, + "learning_rate": 0.00012845866433409825, + "loss": 2.5685, + "step": 19215 + }, + { + "epoch": 0.82, + "grad_norm": 0.40625, + "learning_rate": 0.00012845223370934445, + "loss": 2.5714, + "step": 19216 + }, + { + "epoch": 0.82, + "grad_norm": 0.40625, + "learning_rate": 0.00012844580295656596, + "loss": 2.5965, + "step": 19217 + }, + { + "epoch": 0.82, + "grad_norm": 0.41015625, + "learning_rate": 0.0001284393720757918, + "loss": 2.5572, + "step": 19218 + }, + { + "epoch": 0.82, + "grad_norm": 0.486328125, + "learning_rate": 0.0001284329410670508, + "loss": 2.5991, + "step": 19219 + }, + { + "epoch": 0.82, + "grad_norm": 0.455078125, + "learning_rate": 0.000128426509930372, + "loss": 2.5022, + "step": 19220 + }, + { + "epoch": 0.82, + "grad_norm": 0.494140625, + "learning_rate": 0.00012842007866578428, + "loss": 2.664, + "step": 19221 + }, + { + "epoch": 0.82, + "grad_norm": 0.44140625, + "learning_rate": 0.0001284136472733166, + "loss": 2.5761, + "step": 19222 + }, + { + "epoch": 0.82, + "grad_norm": 0.44140625, + "learning_rate": 0.00012840721575299786, + "loss": 2.5454, + "step": 19223 + }, + { + "epoch": 0.82, + "grad_norm": 0.40234375, + "learning_rate": 0.00012840078410485708, + "loss": 2.6105, + "step": 19224 + }, + { + "epoch": 0.82, + "grad_norm": 0.5078125, + "learning_rate": 0.0001283943523289231, + "loss": 2.5589, + "step": 19225 + }, + { + "epoch": 0.82, + "grad_norm": 0.458984375, + "learning_rate": 0.00012838792042522496, + "loss": 2.5641, + "step": 19226 + }, + { + "epoch": 0.82, + "grad_norm": 0.484375, + "learning_rate": 0.0001283814883937915, + "loss": 2.6654, + "step": 19227 + }, + { + "epoch": 0.82, + "grad_norm": 0.482421875, + "learning_rate": 0.00012837505623465174, + "loss": 2.5287, + "step": 19228 + }, + { + "epoch": 0.82, + "grad_norm": 0.396484375, + "learning_rate": 0.0001283686239478346, + "loss": 2.4815, + "step": 19229 + }, + { + "epoch": 0.82, + "grad_norm": 0.419921875, + "learning_rate": 0.00012836219153336903, + "loss": 2.5722, + "step": 19230 + }, + { + "epoch": 0.82, + "grad_norm": 0.400390625, + "learning_rate": 0.00012835575899128394, + "loss": 2.5084, + "step": 19231 + }, + { + "epoch": 0.82, + "grad_norm": 0.447265625, + "learning_rate": 0.0001283493263216083, + "loss": 2.5176, + "step": 19232 + }, + { + "epoch": 0.82, + "grad_norm": 0.4921875, + "learning_rate": 0.00012834289352437107, + "loss": 2.5916, + "step": 19233 + }, + { + "epoch": 0.82, + "grad_norm": 0.462890625, + "learning_rate": 0.00012833646059960118, + "loss": 2.579, + "step": 19234 + }, + { + "epoch": 0.82, + "grad_norm": 0.484375, + "learning_rate": 0.00012833002754732755, + "loss": 2.615, + "step": 19235 + }, + { + "epoch": 0.82, + "grad_norm": 0.478515625, + "learning_rate": 0.00012832359436757917, + "loss": 2.6234, + "step": 19236 + }, + { + "epoch": 0.82, + "grad_norm": 0.44140625, + "learning_rate": 0.00012831716106038497, + "loss": 2.582, + "step": 19237 + }, + { + "epoch": 0.82, + "grad_norm": 0.43359375, + "learning_rate": 0.00012831072762577387, + "loss": 2.4609, + "step": 19238 + }, + { + "epoch": 0.82, + "grad_norm": 0.466796875, + "learning_rate": 0.00012830429406377483, + "loss": 2.5527, + "step": 19239 + }, + { + "epoch": 0.82, + "grad_norm": 0.46484375, + "learning_rate": 0.0001282978603744169, + "loss": 2.6152, + "step": 19240 + }, + { + "epoch": 0.82, + "grad_norm": 0.43359375, + "learning_rate": 0.00012829142655772883, + "loss": 2.5399, + "step": 19241 + }, + { + "epoch": 0.82, + "grad_norm": 0.50390625, + "learning_rate": 0.00012828499261373973, + "loss": 2.5807, + "step": 19242 + }, + { + "epoch": 0.82, + "grad_norm": 0.498046875, + "learning_rate": 0.0001282785585424785, + "loss": 2.6636, + "step": 19243 + }, + { + "epoch": 0.82, + "grad_norm": 0.470703125, + "learning_rate": 0.00012827212434397408, + "loss": 2.5488, + "step": 19244 + }, + { + "epoch": 0.82, + "grad_norm": 0.48046875, + "learning_rate": 0.00012826569001825547, + "loss": 2.4337, + "step": 19245 + }, + { + "epoch": 0.82, + "grad_norm": 0.384765625, + "learning_rate": 0.00012825925556535155, + "loss": 2.3629, + "step": 19246 + }, + { + "epoch": 0.82, + "grad_norm": 0.462890625, + "learning_rate": 0.00012825282098529128, + "loss": 2.6385, + "step": 19247 + }, + { + "epoch": 0.82, + "grad_norm": 0.4921875, + "learning_rate": 0.00012824638627810368, + "loss": 2.5433, + "step": 19248 + }, + { + "epoch": 0.82, + "grad_norm": 0.470703125, + "learning_rate": 0.00012823995144381763, + "loss": 2.7009, + "step": 19249 + }, + { + "epoch": 0.82, + "grad_norm": 0.486328125, + "learning_rate": 0.0001282335164824622, + "loss": 2.6315, + "step": 19250 + }, + { + "epoch": 0.82, + "grad_norm": 0.46875, + "learning_rate": 0.00012822708139406617, + "loss": 2.6605, + "step": 19251 + }, + { + "epoch": 0.82, + "grad_norm": 0.48828125, + "learning_rate": 0.00012822064617865864, + "loss": 2.7454, + "step": 19252 + }, + { + "epoch": 0.82, + "grad_norm": 0.462890625, + "learning_rate": 0.0001282142108362685, + "loss": 2.6266, + "step": 19253 + }, + { + "epoch": 0.82, + "grad_norm": 0.412109375, + "learning_rate": 0.00012820777536692475, + "loss": 2.4848, + "step": 19254 + }, + { + "epoch": 0.82, + "grad_norm": 0.494140625, + "learning_rate": 0.00012820133977065626, + "loss": 2.5508, + "step": 19255 + }, + { + "epoch": 0.82, + "grad_norm": 0.431640625, + "learning_rate": 0.00012819490404749213, + "loss": 2.4586, + "step": 19256 + }, + { + "epoch": 0.82, + "grad_norm": 0.453125, + "learning_rate": 0.0001281884681974612, + "loss": 2.5167, + "step": 19257 + }, + { + "epoch": 0.82, + "grad_norm": 0.447265625, + "learning_rate": 0.00012818203222059242, + "loss": 2.6193, + "step": 19258 + }, + { + "epoch": 0.82, + "grad_norm": 0.4921875, + "learning_rate": 0.00012817559611691482, + "loss": 2.6188, + "step": 19259 + }, + { + "epoch": 0.82, + "grad_norm": 0.474609375, + "learning_rate": 0.00012816915988645734, + "loss": 2.6169, + "step": 19260 + }, + { + "epoch": 0.82, + "grad_norm": 0.47265625, + "learning_rate": 0.00012816272352924896, + "loss": 2.5968, + "step": 19261 + }, + { + "epoch": 0.82, + "grad_norm": 0.44140625, + "learning_rate": 0.00012815628704531862, + "loss": 2.5674, + "step": 19262 + }, + { + "epoch": 0.82, + "grad_norm": 0.400390625, + "learning_rate": 0.0001281498504346952, + "loss": 2.5187, + "step": 19263 + }, + { + "epoch": 0.82, + "grad_norm": 0.43359375, + "learning_rate": 0.00012814341369740784, + "loss": 2.5953, + "step": 19264 + }, + { + "epoch": 0.82, + "grad_norm": 0.490234375, + "learning_rate": 0.00012813697683348532, + "loss": 2.4426, + "step": 19265 + }, + { + "epoch": 0.82, + "grad_norm": 0.412109375, + "learning_rate": 0.00012813053984295672, + "loss": 2.6239, + "step": 19266 + }, + { + "epoch": 0.82, + "grad_norm": 0.435546875, + "learning_rate": 0.00012812410272585101, + "loss": 2.4422, + "step": 19267 + }, + { + "epoch": 0.82, + "grad_norm": 0.44921875, + "learning_rate": 0.0001281176654821971, + "loss": 2.3976, + "step": 19268 + }, + { + "epoch": 0.82, + "grad_norm": 0.498046875, + "learning_rate": 0.00012811122811202393, + "loss": 2.5888, + "step": 19269 + }, + { + "epoch": 0.82, + "grad_norm": 0.50390625, + "learning_rate": 0.00012810479061536054, + "loss": 2.665, + "step": 19270 + }, + { + "epoch": 0.82, + "grad_norm": 0.4375, + "learning_rate": 0.00012809835299223586, + "loss": 2.6127, + "step": 19271 + }, + { + "epoch": 0.82, + "grad_norm": 0.52734375, + "learning_rate": 0.00012809191524267886, + "loss": 2.5804, + "step": 19272 + }, + { + "epoch": 0.82, + "grad_norm": 0.4375, + "learning_rate": 0.0001280854773667185, + "loss": 2.4889, + "step": 19273 + }, + { + "epoch": 0.82, + "grad_norm": 0.46484375, + "learning_rate": 0.00012807903936438373, + "loss": 2.6022, + "step": 19274 + }, + { + "epoch": 0.82, + "grad_norm": 0.37109375, + "learning_rate": 0.0001280726012357036, + "loss": 2.542, + "step": 19275 + }, + { + "epoch": 0.82, + "grad_norm": 0.44921875, + "learning_rate": 0.00012806616298070698, + "loss": 2.3793, + "step": 19276 + }, + { + "epoch": 0.82, + "grad_norm": 0.47265625, + "learning_rate": 0.0001280597245994229, + "loss": 2.526, + "step": 19277 + }, + { + "epoch": 0.82, + "grad_norm": 0.50390625, + "learning_rate": 0.00012805328609188032, + "loss": 2.5896, + "step": 19278 + }, + { + "epoch": 0.82, + "grad_norm": 0.4609375, + "learning_rate": 0.0001280468474581082, + "loss": 2.5934, + "step": 19279 + }, + { + "epoch": 0.82, + "grad_norm": 0.412109375, + "learning_rate": 0.00012804040869813553, + "loss": 2.4593, + "step": 19280 + }, + { + "epoch": 0.82, + "grad_norm": 0.443359375, + "learning_rate": 0.00012803396981199124, + "loss": 2.5923, + "step": 19281 + }, + { + "epoch": 0.82, + "grad_norm": 0.42578125, + "learning_rate": 0.00012802753079970433, + "loss": 2.6533, + "step": 19282 + }, + { + "epoch": 0.82, + "grad_norm": 0.46875, + "learning_rate": 0.00012802109166130382, + "loss": 2.5362, + "step": 19283 + }, + { + "epoch": 0.82, + "grad_norm": 0.515625, + "learning_rate": 0.00012801465239681859, + "loss": 2.5942, + "step": 19284 + }, + { + "epoch": 0.82, + "grad_norm": 0.4921875, + "learning_rate": 0.00012800821300627767, + "loss": 2.6847, + "step": 19285 + }, + { + "epoch": 0.82, + "grad_norm": 0.44921875, + "learning_rate": 0.00012800177348971004, + "loss": 2.619, + "step": 19286 + }, + { + "epoch": 0.82, + "grad_norm": 0.474609375, + "learning_rate": 0.00012799533384714467, + "loss": 2.4929, + "step": 19287 + }, + { + "epoch": 0.82, + "grad_norm": 0.421875, + "learning_rate": 0.0001279888940786105, + "loss": 2.5966, + "step": 19288 + }, + { + "epoch": 0.82, + "grad_norm": 0.453125, + "learning_rate": 0.00012798245418413657, + "loss": 2.5295, + "step": 19289 + }, + { + "epoch": 0.82, + "grad_norm": 0.5, + "learning_rate": 0.0001279760141637518, + "loss": 2.6315, + "step": 19290 + }, + { + "epoch": 0.82, + "grad_norm": 0.447265625, + "learning_rate": 0.0001279695740174852, + "loss": 2.5933, + "step": 19291 + }, + { + "epoch": 0.82, + "grad_norm": 0.46875, + "learning_rate": 0.00012796313374536573, + "loss": 2.6058, + "step": 19292 + }, + { + "epoch": 0.82, + "grad_norm": 0.458984375, + "learning_rate": 0.00012795669334742237, + "loss": 2.4798, + "step": 19293 + }, + { + "epoch": 0.82, + "grad_norm": 0.4453125, + "learning_rate": 0.00012795025282368413, + "loss": 2.5625, + "step": 19294 + }, + { + "epoch": 0.82, + "grad_norm": 0.46484375, + "learning_rate": 0.00012794381217417998, + "loss": 2.5224, + "step": 19295 + }, + { + "epoch": 0.82, + "grad_norm": 0.5078125, + "learning_rate": 0.00012793737139893884, + "loss": 2.5623, + "step": 19296 + }, + { + "epoch": 0.82, + "grad_norm": 0.443359375, + "learning_rate": 0.00012793093049798975, + "loss": 2.6113, + "step": 19297 + }, + { + "epoch": 0.82, + "grad_norm": 0.50390625, + "learning_rate": 0.00012792448947136168, + "loss": 2.6709, + "step": 19298 + }, + { + "epoch": 0.82, + "grad_norm": 0.484375, + "learning_rate": 0.00012791804831908368, + "loss": 2.5239, + "step": 19299 + }, + { + "epoch": 0.82, + "grad_norm": 0.451171875, + "learning_rate": 0.0001279116070411846, + "loss": 2.5231, + "step": 19300 + }, + { + "epoch": 0.82, + "grad_norm": 0.478515625, + "learning_rate": 0.0001279051656376935, + "loss": 2.672, + "step": 19301 + }, + { + "epoch": 0.82, + "grad_norm": 0.443359375, + "learning_rate": 0.00012789872410863936, + "loss": 2.6496, + "step": 19302 + }, + { + "epoch": 0.82, + "grad_norm": 0.421875, + "learning_rate": 0.00012789228245405114, + "loss": 2.6067, + "step": 19303 + }, + { + "epoch": 0.82, + "grad_norm": 0.439453125, + "learning_rate": 0.00012788584067395786, + "loss": 2.4638, + "step": 19304 + }, + { + "epoch": 0.82, + "grad_norm": 0.443359375, + "learning_rate": 0.00012787939876838847, + "loss": 2.5667, + "step": 19305 + }, + { + "epoch": 0.82, + "grad_norm": 0.453125, + "learning_rate": 0.00012787295673737196, + "loss": 2.5106, + "step": 19306 + }, + { + "epoch": 0.82, + "grad_norm": 0.462890625, + "learning_rate": 0.0001278665145809374, + "loss": 2.6145, + "step": 19307 + }, + { + "epoch": 0.82, + "grad_norm": 0.416015625, + "learning_rate": 0.00012786007229911362, + "loss": 2.5657, + "step": 19308 + }, + { + "epoch": 0.82, + "grad_norm": 0.427734375, + "learning_rate": 0.00012785362989192975, + "loss": 2.4466, + "step": 19309 + }, + { + "epoch": 0.82, + "grad_norm": 0.47265625, + "learning_rate": 0.0001278471873594147, + "loss": 2.5471, + "step": 19310 + }, + { + "epoch": 0.82, + "grad_norm": 0.4140625, + "learning_rate": 0.00012784074470159753, + "loss": 2.5396, + "step": 19311 + }, + { + "epoch": 0.82, + "grad_norm": 0.4921875, + "learning_rate": 0.00012783430191850713, + "loss": 2.676, + "step": 19312 + }, + { + "epoch": 0.82, + "grad_norm": 0.431640625, + "learning_rate": 0.0001278278590101726, + "loss": 2.5338, + "step": 19313 + }, + { + "epoch": 0.82, + "grad_norm": 0.5078125, + "learning_rate": 0.0001278214159766228, + "loss": 2.6821, + "step": 19314 + }, + { + "epoch": 0.82, + "grad_norm": 0.47265625, + "learning_rate": 0.00012781497281788687, + "loss": 2.6508, + "step": 19315 + }, + { + "epoch": 0.82, + "grad_norm": 0.51171875, + "learning_rate": 0.00012780852953399367, + "loss": 2.6771, + "step": 19316 + }, + { + "epoch": 0.82, + "grad_norm": 0.5390625, + "learning_rate": 0.00012780208612497227, + "loss": 2.6527, + "step": 19317 + }, + { + "epoch": 0.82, + "grad_norm": 0.435546875, + "learning_rate": 0.00012779564259085167, + "loss": 2.5112, + "step": 19318 + }, + { + "epoch": 0.82, + "grad_norm": 0.50390625, + "learning_rate": 0.0001277891989316608, + "loss": 2.586, + "step": 19319 + }, + { + "epoch": 0.82, + "grad_norm": 0.50390625, + "learning_rate": 0.00012778275514742872, + "loss": 2.5846, + "step": 19320 + }, + { + "epoch": 0.82, + "grad_norm": 0.423828125, + "learning_rate": 0.00012777631123818439, + "loss": 2.5008, + "step": 19321 + }, + { + "epoch": 0.82, + "grad_norm": 0.47265625, + "learning_rate": 0.00012776986720395677, + "loss": 2.5063, + "step": 19322 + }, + { + "epoch": 0.82, + "grad_norm": 0.439453125, + "learning_rate": 0.00012776342304477497, + "loss": 2.6551, + "step": 19323 + }, + { + "epoch": 0.82, + "grad_norm": 0.51953125, + "learning_rate": 0.00012775697876066784, + "loss": 2.5656, + "step": 19324 + }, + { + "epoch": 0.82, + "grad_norm": 0.53515625, + "learning_rate": 0.00012775053435166447, + "loss": 2.662, + "step": 19325 + }, + { + "epoch": 0.82, + "grad_norm": 0.45703125, + "learning_rate": 0.00012774408981779386, + "loss": 2.5873, + "step": 19326 + }, + { + "epoch": 0.82, + "grad_norm": 0.451171875, + "learning_rate": 0.00012773764515908496, + "loss": 2.5184, + "step": 19327 + }, + { + "epoch": 0.82, + "grad_norm": 0.51171875, + "learning_rate": 0.0001277312003755668, + "loss": 2.6157, + "step": 19328 + }, + { + "epoch": 0.82, + "grad_norm": 0.53125, + "learning_rate": 0.0001277247554672684, + "loss": 2.5273, + "step": 19329 + }, + { + "epoch": 0.82, + "grad_norm": 0.447265625, + "learning_rate": 0.0001277183104342187, + "loss": 2.4996, + "step": 19330 + }, + { + "epoch": 0.82, + "grad_norm": 0.486328125, + "learning_rate": 0.00012771186527644676, + "loss": 2.6525, + "step": 19331 + }, + { + "epoch": 0.82, + "grad_norm": 0.466796875, + "learning_rate": 0.0001277054199939815, + "loss": 2.504, + "step": 19332 + }, + { + "epoch": 0.82, + "grad_norm": 0.5078125, + "learning_rate": 0.00012769897458685204, + "loss": 2.6658, + "step": 19333 + }, + { + "epoch": 0.82, + "grad_norm": 0.451171875, + "learning_rate": 0.00012769252905508725, + "loss": 2.3939, + "step": 19334 + }, + { + "epoch": 0.82, + "grad_norm": 0.474609375, + "learning_rate": 0.00012768608339871625, + "loss": 2.4532, + "step": 19335 + }, + { + "epoch": 0.82, + "grad_norm": 0.51171875, + "learning_rate": 0.000127679637617768, + "loss": 2.5083, + "step": 19336 + }, + { + "epoch": 0.82, + "grad_norm": 0.46484375, + "learning_rate": 0.0001276731917122715, + "loss": 2.5351, + "step": 19337 + }, + { + "epoch": 0.82, + "grad_norm": 0.396484375, + "learning_rate": 0.00012766674568225568, + "loss": 2.5683, + "step": 19338 + }, + { + "epoch": 0.82, + "grad_norm": 0.44921875, + "learning_rate": 0.0001276602995277497, + "loss": 2.63, + "step": 19339 + }, + { + "epoch": 0.82, + "grad_norm": 0.44140625, + "learning_rate": 0.00012765385324878243, + "loss": 2.6148, + "step": 19340 + }, + { + "epoch": 0.82, + "grad_norm": 0.470703125, + "learning_rate": 0.00012764740684538295, + "loss": 2.6438, + "step": 19341 + }, + { + "epoch": 0.82, + "grad_norm": 0.4453125, + "learning_rate": 0.00012764096031758025, + "loss": 2.5147, + "step": 19342 + }, + { + "epoch": 0.82, + "grad_norm": 0.49609375, + "learning_rate": 0.0001276345136654033, + "loss": 2.6629, + "step": 19343 + }, + { + "epoch": 0.82, + "grad_norm": 0.41015625, + "learning_rate": 0.00012762806688888117, + "loss": 2.4556, + "step": 19344 + }, + { + "epoch": 0.82, + "grad_norm": 0.5, + "learning_rate": 0.0001276216199880428, + "loss": 2.5012, + "step": 19345 + }, + { + "epoch": 0.82, + "grad_norm": 0.455078125, + "learning_rate": 0.00012761517296291724, + "loss": 2.4858, + "step": 19346 + }, + { + "epoch": 0.82, + "grad_norm": 0.45703125, + "learning_rate": 0.00012760872581353354, + "loss": 2.5835, + "step": 19347 + }, + { + "epoch": 0.82, + "grad_norm": 0.443359375, + "learning_rate": 0.00012760227853992058, + "loss": 2.4397, + "step": 19348 + }, + { + "epoch": 0.82, + "grad_norm": 0.46875, + "learning_rate": 0.00012759583114210754, + "loss": 2.6036, + "step": 19349 + }, + { + "epoch": 0.82, + "grad_norm": 0.43359375, + "learning_rate": 0.00012758938362012333, + "loss": 2.5027, + "step": 19350 + }, + { + "epoch": 0.82, + "grad_norm": 0.44140625, + "learning_rate": 0.00012758293597399693, + "loss": 2.6004, + "step": 19351 + }, + { + "epoch": 0.82, + "grad_norm": 0.419921875, + "learning_rate": 0.00012757648820375744, + "loss": 2.6429, + "step": 19352 + }, + { + "epoch": 0.82, + "grad_norm": 0.42578125, + "learning_rate": 0.0001275700403094338, + "loss": 2.6633, + "step": 19353 + }, + { + "epoch": 0.82, + "grad_norm": 0.478515625, + "learning_rate": 0.00012756359229105507, + "loss": 2.6426, + "step": 19354 + }, + { + "epoch": 0.82, + "grad_norm": 0.46484375, + "learning_rate": 0.00012755714414865025, + "loss": 2.5983, + "step": 19355 + }, + { + "epoch": 0.82, + "grad_norm": 0.474609375, + "learning_rate": 0.00012755069588224835, + "loss": 2.5905, + "step": 19356 + }, + { + "epoch": 0.82, + "grad_norm": 0.4765625, + "learning_rate": 0.0001275442474918784, + "loss": 2.5576, + "step": 19357 + }, + { + "epoch": 0.82, + "grad_norm": 0.486328125, + "learning_rate": 0.00012753779897756938, + "loss": 2.6551, + "step": 19358 + }, + { + "epoch": 0.82, + "grad_norm": 0.408203125, + "learning_rate": 0.00012753135033935034, + "loss": 2.5458, + "step": 19359 + }, + { + "epoch": 0.82, + "grad_norm": 0.49609375, + "learning_rate": 0.00012752490157725028, + "loss": 2.5916, + "step": 19360 + }, + { + "epoch": 0.83, + "grad_norm": 0.44921875, + "learning_rate": 0.00012751845269129823, + "loss": 2.6015, + "step": 19361 + }, + { + "epoch": 0.83, + "grad_norm": 0.50390625, + "learning_rate": 0.0001275120036815232, + "loss": 2.4877, + "step": 19362 + }, + { + "epoch": 0.83, + "grad_norm": 0.41796875, + "learning_rate": 0.00012750555454795422, + "loss": 2.566, + "step": 19363 + }, + { + "epoch": 0.83, + "grad_norm": 0.46875, + "learning_rate": 0.00012749910529062026, + "loss": 2.4769, + "step": 19364 + }, + { + "epoch": 0.83, + "grad_norm": 0.447265625, + "learning_rate": 0.0001274926559095504, + "loss": 2.4548, + "step": 19365 + }, + { + "epoch": 0.83, + "grad_norm": 0.49609375, + "learning_rate": 0.00012748620640477362, + "loss": 2.5702, + "step": 19366 + }, + { + "epoch": 0.83, + "grad_norm": 0.546875, + "learning_rate": 0.00012747975677631897, + "loss": 2.6061, + "step": 19367 + }, + { + "epoch": 0.83, + "grad_norm": 0.470703125, + "learning_rate": 0.00012747330702421548, + "loss": 2.5634, + "step": 19368 + }, + { + "epoch": 0.83, + "grad_norm": 0.43359375, + "learning_rate": 0.00012746685714849212, + "loss": 2.5723, + "step": 19369 + }, + { + "epoch": 0.83, + "grad_norm": 0.5, + "learning_rate": 0.00012746040714917793, + "loss": 2.609, + "step": 19370 + }, + { + "epoch": 0.83, + "grad_norm": 0.421875, + "learning_rate": 0.00012745395702630196, + "loss": 2.6304, + "step": 19371 + }, + { + "epoch": 0.83, + "grad_norm": 0.478515625, + "learning_rate": 0.0001274475067798932, + "loss": 2.5845, + "step": 19372 + }, + { + "epoch": 0.83, + "grad_norm": 0.50390625, + "learning_rate": 0.00012744105640998072, + "loss": 2.4855, + "step": 19373 + }, + { + "epoch": 0.83, + "grad_norm": 0.42578125, + "learning_rate": 0.00012743460591659348, + "loss": 2.6009, + "step": 19374 + }, + { + "epoch": 0.83, + "grad_norm": 0.474609375, + "learning_rate": 0.0001274281552997606, + "loss": 2.4501, + "step": 19375 + }, + { + "epoch": 0.83, + "grad_norm": 0.4453125, + "learning_rate": 0.00012742170455951098, + "loss": 2.4754, + "step": 19376 + }, + { + "epoch": 0.83, + "grad_norm": 0.400390625, + "learning_rate": 0.00012741525369587374, + "loss": 2.6272, + "step": 19377 + }, + { + "epoch": 0.83, + "grad_norm": 0.423828125, + "learning_rate": 0.00012740880270887788, + "loss": 2.4422, + "step": 19378 + }, + { + "epoch": 0.83, + "grad_norm": 0.47265625, + "learning_rate": 0.00012740235159855242, + "loss": 2.5591, + "step": 19379 + }, + { + "epoch": 0.83, + "grad_norm": 0.43359375, + "learning_rate": 0.00012739590036492637, + "loss": 2.5256, + "step": 19380 + }, + { + "epoch": 0.83, + "grad_norm": 0.4765625, + "learning_rate": 0.0001273894490080288, + "loss": 2.6543, + "step": 19381 + }, + { + "epoch": 0.83, + "grad_norm": 0.45703125, + "learning_rate": 0.00012738299752788874, + "loss": 2.4986, + "step": 19382 + }, + { + "epoch": 0.83, + "grad_norm": 0.4765625, + "learning_rate": 0.00012737654592453516, + "loss": 2.4996, + "step": 19383 + }, + { + "epoch": 0.83, + "grad_norm": 0.46875, + "learning_rate": 0.00012737009419799716, + "loss": 2.5285, + "step": 19384 + }, + { + "epoch": 0.83, + "grad_norm": 0.474609375, + "learning_rate": 0.00012736364234830375, + "loss": 2.7264, + "step": 19385 + }, + { + "epoch": 0.83, + "grad_norm": 0.419921875, + "learning_rate": 0.00012735719037548388, + "loss": 2.5443, + "step": 19386 + }, + { + "epoch": 0.83, + "grad_norm": 0.46875, + "learning_rate": 0.00012735073827956672, + "loss": 2.5835, + "step": 19387 + }, + { + "epoch": 0.83, + "grad_norm": 0.470703125, + "learning_rate": 0.0001273442860605812, + "loss": 2.4186, + "step": 19388 + }, + { + "epoch": 0.83, + "grad_norm": 0.478515625, + "learning_rate": 0.00012733783371855642, + "loss": 2.6809, + "step": 19389 + }, + { + "epoch": 0.83, + "grad_norm": 0.455078125, + "learning_rate": 0.00012733138125352133, + "loss": 2.3977, + "step": 19390 + }, + { + "epoch": 0.83, + "grad_norm": 0.42578125, + "learning_rate": 0.00012732492866550504, + "loss": 2.5698, + "step": 19391 + }, + { + "epoch": 0.83, + "grad_norm": 0.439453125, + "learning_rate": 0.00012731847595453657, + "loss": 2.418, + "step": 19392 + }, + { + "epoch": 0.83, + "grad_norm": 0.41015625, + "learning_rate": 0.00012731202312064491, + "loss": 2.5711, + "step": 19393 + }, + { + "epoch": 0.83, + "grad_norm": 0.44140625, + "learning_rate": 0.00012730557016385916, + "loss": 2.6185, + "step": 19394 + }, + { + "epoch": 0.83, + "grad_norm": 0.47265625, + "learning_rate": 0.00012729911708420828, + "loss": 2.5306, + "step": 19395 + }, + { + "epoch": 0.83, + "grad_norm": 0.47265625, + "learning_rate": 0.00012729266388172142, + "loss": 2.5762, + "step": 19396 + }, + { + "epoch": 0.83, + "grad_norm": 0.42578125, + "learning_rate": 0.00012728621055642746, + "loss": 2.5137, + "step": 19397 + }, + { + "epoch": 0.83, + "grad_norm": 0.4921875, + "learning_rate": 0.0001272797571083556, + "loss": 2.6637, + "step": 19398 + }, + { + "epoch": 0.83, + "grad_norm": 0.46484375, + "learning_rate": 0.00012727330353753475, + "loss": 2.6079, + "step": 19399 + }, + { + "epoch": 0.83, + "grad_norm": 0.427734375, + "learning_rate": 0.00012726684984399404, + "loss": 2.4593, + "step": 19400 + }, + { + "epoch": 0.83, + "grad_norm": 0.443359375, + "learning_rate": 0.00012726039602776245, + "loss": 2.4973, + "step": 19401 + }, + { + "epoch": 0.83, + "grad_norm": 0.466796875, + "learning_rate": 0.00012725394208886904, + "loss": 2.5667, + "step": 19402 + }, + { + "epoch": 0.83, + "grad_norm": 0.46484375, + "learning_rate": 0.00012724748802734286, + "loss": 2.4897, + "step": 19403 + }, + { + "epoch": 0.83, + "grad_norm": 0.48828125, + "learning_rate": 0.00012724103384321292, + "loss": 2.6243, + "step": 19404 + }, + { + "epoch": 0.83, + "grad_norm": 0.36328125, + "learning_rate": 0.0001272345795365083, + "loss": 2.4519, + "step": 19405 + }, + { + "epoch": 0.83, + "grad_norm": 0.423828125, + "learning_rate": 0.00012722812510725806, + "loss": 2.5164, + "step": 19406 + }, + { + "epoch": 0.83, + "grad_norm": 0.4765625, + "learning_rate": 0.00012722167055549116, + "loss": 2.5985, + "step": 19407 + }, + { + "epoch": 0.83, + "grad_norm": 0.455078125, + "learning_rate": 0.00012721521588123668, + "loss": 2.57, + "step": 19408 + }, + { + "epoch": 0.83, + "grad_norm": 0.474609375, + "learning_rate": 0.0001272087610845237, + "loss": 2.6191, + "step": 19409 + }, + { + "epoch": 0.83, + "grad_norm": 0.51953125, + "learning_rate": 0.00012720230616538125, + "loss": 2.5446, + "step": 19410 + }, + { + "epoch": 0.83, + "grad_norm": 0.51171875, + "learning_rate": 0.00012719585112383835, + "loss": 2.5437, + "step": 19411 + }, + { + "epoch": 0.83, + "grad_norm": 0.55859375, + "learning_rate": 0.00012718939595992405, + "loss": 2.5935, + "step": 19412 + }, + { + "epoch": 0.83, + "grad_norm": 0.4140625, + "learning_rate": 0.0001271829406736674, + "loss": 2.578, + "step": 19413 + }, + { + "epoch": 0.83, + "grad_norm": 0.494140625, + "learning_rate": 0.00012717648526509748, + "loss": 2.5712, + "step": 19414 + }, + { + "epoch": 0.83, + "grad_norm": 0.44921875, + "learning_rate": 0.00012717002973424327, + "loss": 2.5437, + "step": 19415 + }, + { + "epoch": 0.83, + "grad_norm": 0.44140625, + "learning_rate": 0.00012716357408113388, + "loss": 2.4415, + "step": 19416 + }, + { + "epoch": 0.83, + "grad_norm": 0.4609375, + "learning_rate": 0.00012715711830579835, + "loss": 2.5274, + "step": 19417 + }, + { + "epoch": 0.83, + "grad_norm": 0.41015625, + "learning_rate": 0.0001271506624082657, + "loss": 2.6025, + "step": 19418 + }, + { + "epoch": 0.83, + "grad_norm": 0.5, + "learning_rate": 0.000127144206388565, + "loss": 2.6373, + "step": 19419 + }, + { + "epoch": 0.83, + "grad_norm": 0.482421875, + "learning_rate": 0.00012713775024672527, + "loss": 2.704, + "step": 19420 + }, + { + "epoch": 0.83, + "grad_norm": 0.478515625, + "learning_rate": 0.00012713129398277557, + "loss": 2.5417, + "step": 19421 + }, + { + "epoch": 0.83, + "grad_norm": 0.47265625, + "learning_rate": 0.000127124837596745, + "loss": 2.647, + "step": 19422 + }, + { + "epoch": 0.83, + "grad_norm": 0.5, + "learning_rate": 0.00012711838108866254, + "loss": 2.5423, + "step": 19423 + }, + { + "epoch": 0.83, + "grad_norm": 0.52734375, + "learning_rate": 0.0001271119244585573, + "loss": 2.5548, + "step": 19424 + }, + { + "epoch": 0.83, + "grad_norm": 0.43359375, + "learning_rate": 0.0001271054677064583, + "loss": 2.6302, + "step": 19425 + }, + { + "epoch": 0.83, + "grad_norm": 0.404296875, + "learning_rate": 0.0001270990108323946, + "loss": 2.5442, + "step": 19426 + }, + { + "epoch": 0.83, + "grad_norm": 0.447265625, + "learning_rate": 0.00012709255383639527, + "loss": 2.57, + "step": 19427 + }, + { + "epoch": 0.83, + "grad_norm": 0.44140625, + "learning_rate": 0.00012708609671848933, + "loss": 2.5457, + "step": 19428 + }, + { + "epoch": 0.83, + "grad_norm": 0.4453125, + "learning_rate": 0.00012707963947870582, + "loss": 2.6496, + "step": 19429 + }, + { + "epoch": 0.83, + "grad_norm": 0.4375, + "learning_rate": 0.0001270731821170739, + "loss": 2.5644, + "step": 19430 + }, + { + "epoch": 0.83, + "grad_norm": 0.5390625, + "learning_rate": 0.0001270667246336225, + "loss": 2.5875, + "step": 19431 + }, + { + "epoch": 0.83, + "grad_norm": 0.462890625, + "learning_rate": 0.00012706026702838076, + "loss": 2.5092, + "step": 19432 + }, + { + "epoch": 0.83, + "grad_norm": 0.4140625, + "learning_rate": 0.0001270538093013777, + "loss": 2.4596, + "step": 19433 + }, + { + "epoch": 0.83, + "grad_norm": 0.416015625, + "learning_rate": 0.00012704735145264238, + "loss": 2.5614, + "step": 19434 + }, + { + "epoch": 0.83, + "grad_norm": 0.443359375, + "learning_rate": 0.00012704089348220386, + "loss": 2.5223, + "step": 19435 + }, + { + "epoch": 0.83, + "grad_norm": 0.46484375, + "learning_rate": 0.00012703443539009124, + "loss": 2.62, + "step": 19436 + }, + { + "epoch": 0.83, + "grad_norm": 0.4765625, + "learning_rate": 0.0001270279771763335, + "loss": 2.5779, + "step": 19437 + }, + { + "epoch": 0.83, + "grad_norm": 0.451171875, + "learning_rate": 0.00012702151884095975, + "loss": 2.4828, + "step": 19438 + }, + { + "epoch": 0.83, + "grad_norm": 0.43359375, + "learning_rate": 0.00012701506038399903, + "loss": 2.3983, + "step": 19439 + }, + { + "epoch": 0.83, + "grad_norm": 0.44140625, + "learning_rate": 0.00012700860180548044, + "loss": 2.6097, + "step": 19440 + }, + { + "epoch": 0.83, + "grad_norm": 0.5234375, + "learning_rate": 0.000127002143105433, + "loss": 2.5868, + "step": 19441 + }, + { + "epoch": 0.83, + "grad_norm": 0.470703125, + "learning_rate": 0.00012699568428388577, + "loss": 2.626, + "step": 19442 + }, + { + "epoch": 0.83, + "grad_norm": 0.390625, + "learning_rate": 0.00012698922534086783, + "loss": 2.5085, + "step": 19443 + }, + { + "epoch": 0.83, + "grad_norm": 0.46484375, + "learning_rate": 0.00012698276627640825, + "loss": 2.5608, + "step": 19444 + }, + { + "epoch": 0.83, + "grad_norm": 0.48046875, + "learning_rate": 0.00012697630709053608, + "loss": 2.5151, + "step": 19445 + }, + { + "epoch": 0.83, + "grad_norm": 0.482421875, + "learning_rate": 0.0001269698477832804, + "loss": 2.5024, + "step": 19446 + }, + { + "epoch": 0.83, + "grad_norm": 0.427734375, + "learning_rate": 0.00012696338835467024, + "loss": 2.5304, + "step": 19447 + }, + { + "epoch": 0.83, + "grad_norm": 0.43359375, + "learning_rate": 0.00012695692880473466, + "loss": 2.544, + "step": 19448 + }, + { + "epoch": 0.83, + "grad_norm": 0.419921875, + "learning_rate": 0.0001269504691335028, + "loss": 2.4777, + "step": 19449 + }, + { + "epoch": 0.83, + "grad_norm": 0.423828125, + "learning_rate": 0.00012694400934100365, + "loss": 2.5063, + "step": 19450 + }, + { + "epoch": 0.83, + "grad_norm": 0.462890625, + "learning_rate": 0.00012693754942726633, + "loss": 2.6825, + "step": 19451 + }, + { + "epoch": 0.83, + "grad_norm": 0.455078125, + "learning_rate": 0.00012693108939231988, + "loss": 2.6917, + "step": 19452 + }, + { + "epoch": 0.83, + "grad_norm": 0.5234375, + "learning_rate": 0.00012692462923619334, + "loss": 2.6879, + "step": 19453 + }, + { + "epoch": 0.83, + "grad_norm": 0.482421875, + "learning_rate": 0.00012691816895891585, + "loss": 2.6274, + "step": 19454 + }, + { + "epoch": 0.83, + "grad_norm": 0.50390625, + "learning_rate": 0.0001269117085605164, + "loss": 2.5153, + "step": 19455 + }, + { + "epoch": 0.83, + "grad_norm": 0.46875, + "learning_rate": 0.00012690524804102413, + "loss": 2.5213, + "step": 19456 + }, + { + "epoch": 0.83, + "grad_norm": 0.490234375, + "learning_rate": 0.00012689878740046803, + "loss": 2.5259, + "step": 19457 + }, + { + "epoch": 0.83, + "grad_norm": 0.494140625, + "learning_rate": 0.00012689232663887727, + "loss": 2.6945, + "step": 19458 + }, + { + "epoch": 0.83, + "grad_norm": 0.41796875, + "learning_rate": 0.00012688586575628082, + "loss": 2.5676, + "step": 19459 + }, + { + "epoch": 0.83, + "grad_norm": 0.5, + "learning_rate": 0.00012687940475270785, + "loss": 2.5255, + "step": 19460 + }, + { + "epoch": 0.83, + "grad_norm": 0.458984375, + "learning_rate": 0.00012687294362818732, + "loss": 2.5369, + "step": 19461 + }, + { + "epoch": 0.83, + "grad_norm": 0.49609375, + "learning_rate": 0.00012686648238274844, + "loss": 2.515, + "step": 19462 + }, + { + "epoch": 0.83, + "grad_norm": 0.435546875, + "learning_rate": 0.00012686002101642015, + "loss": 2.6986, + "step": 19463 + }, + { + "epoch": 0.83, + "grad_norm": 0.42578125, + "learning_rate": 0.00012685355952923158, + "loss": 2.4673, + "step": 19464 + }, + { + "epoch": 0.83, + "grad_norm": 0.4765625, + "learning_rate": 0.00012684709792121183, + "loss": 2.5487, + "step": 19465 + }, + { + "epoch": 0.83, + "grad_norm": 0.44921875, + "learning_rate": 0.00012684063619238995, + "loss": 2.4449, + "step": 19466 + }, + { + "epoch": 0.83, + "grad_norm": 0.4921875, + "learning_rate": 0.00012683417434279504, + "loss": 2.5765, + "step": 19467 + }, + { + "epoch": 0.83, + "grad_norm": 0.423828125, + "learning_rate": 0.00012682771237245612, + "loss": 2.5285, + "step": 19468 + }, + { + "epoch": 0.83, + "grad_norm": 0.4609375, + "learning_rate": 0.0001268212502814023, + "loss": 2.5196, + "step": 19469 + }, + { + "epoch": 0.83, + "grad_norm": 0.48046875, + "learning_rate": 0.00012681478806966266, + "loss": 2.5417, + "step": 19470 + }, + { + "epoch": 0.83, + "grad_norm": 0.462890625, + "learning_rate": 0.00012680832573726627, + "loss": 2.6072, + "step": 19471 + }, + { + "epoch": 0.83, + "grad_norm": 0.5390625, + "learning_rate": 0.0001268018632842422, + "loss": 2.6451, + "step": 19472 + }, + { + "epoch": 0.83, + "grad_norm": 0.400390625, + "learning_rate": 0.00012679540071061955, + "loss": 2.5878, + "step": 19473 + }, + { + "epoch": 0.83, + "grad_norm": 0.44921875, + "learning_rate": 0.0001267889380164274, + "loss": 2.6681, + "step": 19474 + }, + { + "epoch": 0.83, + "grad_norm": 0.486328125, + "learning_rate": 0.00012678247520169481, + "loss": 2.5719, + "step": 19475 + }, + { + "epoch": 0.83, + "grad_norm": 0.48828125, + "learning_rate": 0.00012677601226645087, + "loss": 2.5665, + "step": 19476 + }, + { + "epoch": 0.83, + "grad_norm": 0.490234375, + "learning_rate": 0.00012676954921072466, + "loss": 2.4784, + "step": 19477 + }, + { + "epoch": 0.83, + "grad_norm": 0.458984375, + "learning_rate": 0.00012676308603454527, + "loss": 2.6362, + "step": 19478 + }, + { + "epoch": 0.83, + "grad_norm": 0.52734375, + "learning_rate": 0.00012675662273794176, + "loss": 2.5965, + "step": 19479 + }, + { + "epoch": 0.83, + "grad_norm": 0.439453125, + "learning_rate": 0.00012675015932094326, + "loss": 2.6351, + "step": 19480 + }, + { + "epoch": 0.83, + "grad_norm": 0.4296875, + "learning_rate": 0.00012674369578357877, + "loss": 2.4674, + "step": 19481 + }, + { + "epoch": 0.83, + "grad_norm": 0.427734375, + "learning_rate": 0.00012673723212587744, + "loss": 2.6221, + "step": 19482 + }, + { + "epoch": 0.83, + "grad_norm": 0.447265625, + "learning_rate": 0.00012673076834786838, + "loss": 2.6761, + "step": 19483 + }, + { + "epoch": 0.83, + "grad_norm": 0.48046875, + "learning_rate": 0.00012672430444958058, + "loss": 2.6078, + "step": 19484 + }, + { + "epoch": 0.83, + "grad_norm": 0.44921875, + "learning_rate": 0.00012671784043104319, + "loss": 2.5248, + "step": 19485 + }, + { + "epoch": 0.83, + "grad_norm": 0.486328125, + "learning_rate": 0.00012671137629228528, + "loss": 2.6405, + "step": 19486 + }, + { + "epoch": 0.83, + "grad_norm": 0.412109375, + "learning_rate": 0.00012670491203333592, + "loss": 2.4239, + "step": 19487 + }, + { + "epoch": 0.83, + "grad_norm": 0.390625, + "learning_rate": 0.00012669844765422425, + "loss": 2.5516, + "step": 19488 + }, + { + "epoch": 0.83, + "grad_norm": 0.515625, + "learning_rate": 0.0001266919831549793, + "loss": 2.6469, + "step": 19489 + }, + { + "epoch": 0.83, + "grad_norm": 0.490234375, + "learning_rate": 0.0001266855185356302, + "loss": 2.5192, + "step": 19490 + }, + { + "epoch": 0.83, + "grad_norm": 0.46875, + "learning_rate": 0.00012667905379620603, + "loss": 2.5879, + "step": 19491 + }, + { + "epoch": 0.83, + "grad_norm": 0.5234375, + "learning_rate": 0.00012667258893673584, + "loss": 2.5778, + "step": 19492 + }, + { + "epoch": 0.83, + "grad_norm": 0.44921875, + "learning_rate": 0.00012666612395724873, + "loss": 2.5234, + "step": 19493 + }, + { + "epoch": 0.83, + "grad_norm": 0.48046875, + "learning_rate": 0.00012665965885777383, + "loss": 2.544, + "step": 19494 + }, + { + "epoch": 0.83, + "grad_norm": 0.451171875, + "learning_rate": 0.0001266531936383402, + "loss": 2.637, + "step": 19495 + }, + { + "epoch": 0.83, + "grad_norm": 0.515625, + "learning_rate": 0.00012664672829897696, + "loss": 2.6127, + "step": 19496 + }, + { + "epoch": 0.83, + "grad_norm": 0.486328125, + "learning_rate": 0.00012664026283971318, + "loss": 2.5055, + "step": 19497 + }, + { + "epoch": 0.83, + "grad_norm": 0.423828125, + "learning_rate": 0.00012663379726057797, + "loss": 2.3709, + "step": 19498 + }, + { + "epoch": 0.83, + "grad_norm": 0.462890625, + "learning_rate": 0.00012662733156160037, + "loss": 2.4535, + "step": 19499 + }, + { + "epoch": 0.83, + "grad_norm": 0.435546875, + "learning_rate": 0.0001266208657428095, + "loss": 2.5329, + "step": 19500 + }, + { + "epoch": 0.83, + "grad_norm": 0.486328125, + "learning_rate": 0.00012661439980423447, + "loss": 2.4946, + "step": 19501 + }, + { + "epoch": 0.83, + "grad_norm": 0.52734375, + "learning_rate": 0.0001266079337459044, + "loss": 2.5475, + "step": 19502 + }, + { + "epoch": 0.83, + "grad_norm": 0.47265625, + "learning_rate": 0.00012660146756784832, + "loss": 2.5305, + "step": 19503 + }, + { + "epoch": 0.83, + "grad_norm": 0.423828125, + "learning_rate": 0.00012659500127009535, + "loss": 2.6378, + "step": 19504 + }, + { + "epoch": 0.83, + "grad_norm": 0.427734375, + "learning_rate": 0.00012658853485267464, + "loss": 2.5184, + "step": 19505 + }, + { + "epoch": 0.83, + "grad_norm": 0.478515625, + "learning_rate": 0.0001265820683156152, + "loss": 2.5877, + "step": 19506 + }, + { + "epoch": 0.83, + "grad_norm": 0.46875, + "learning_rate": 0.0001265756016589462, + "loss": 2.5667, + "step": 19507 + }, + { + "epoch": 0.83, + "grad_norm": 0.431640625, + "learning_rate": 0.0001265691348826967, + "loss": 2.4891, + "step": 19508 + }, + { + "epoch": 0.83, + "grad_norm": 0.474609375, + "learning_rate": 0.00012656266798689574, + "loss": 2.5086, + "step": 19509 + }, + { + "epoch": 0.83, + "grad_norm": 0.43359375, + "learning_rate": 0.00012655620097157258, + "loss": 2.5413, + "step": 19510 + }, + { + "epoch": 0.83, + "grad_norm": 0.47265625, + "learning_rate": 0.00012654973383675612, + "loss": 2.6476, + "step": 19511 + }, + { + "epoch": 0.83, + "grad_norm": 0.486328125, + "learning_rate": 0.00012654326658247562, + "loss": 2.785, + "step": 19512 + }, + { + "epoch": 0.83, + "grad_norm": 0.419921875, + "learning_rate": 0.0001265367992087601, + "loss": 2.4276, + "step": 19513 + }, + { + "epoch": 0.83, + "grad_norm": 0.482421875, + "learning_rate": 0.0001265303317156387, + "loss": 2.55, + "step": 19514 + }, + { + "epoch": 0.83, + "grad_norm": 0.484375, + "learning_rate": 0.0001265238641031405, + "loss": 2.585, + "step": 19515 + }, + { + "epoch": 0.83, + "grad_norm": 0.474609375, + "learning_rate": 0.0001265173963712946, + "loss": 2.6192, + "step": 19516 + }, + { + "epoch": 0.83, + "grad_norm": 0.40625, + "learning_rate": 0.00012651092852013012, + "loss": 2.4898, + "step": 19517 + }, + { + "epoch": 0.83, + "grad_norm": 0.416015625, + "learning_rate": 0.00012650446054967616, + "loss": 2.4294, + "step": 19518 + }, + { + "epoch": 0.83, + "grad_norm": 0.423828125, + "learning_rate": 0.00012649799245996178, + "loss": 2.5984, + "step": 19519 + }, + { + "epoch": 0.83, + "grad_norm": 0.462890625, + "learning_rate": 0.0001264915242510161, + "loss": 2.5763, + "step": 19520 + }, + { + "epoch": 0.83, + "grad_norm": 0.458984375, + "learning_rate": 0.0001264850559228683, + "loss": 2.6985, + "step": 19521 + }, + { + "epoch": 0.83, + "grad_norm": 0.375, + "learning_rate": 0.0001264785874755474, + "loss": 2.5293, + "step": 19522 + }, + { + "epoch": 0.83, + "grad_norm": 0.4921875, + "learning_rate": 0.00012647211890908254, + "loss": 2.5039, + "step": 19523 + }, + { + "epoch": 0.83, + "grad_norm": 0.42578125, + "learning_rate": 0.0001264656502235028, + "loss": 2.5286, + "step": 19524 + }, + { + "epoch": 0.83, + "grad_norm": 0.447265625, + "learning_rate": 0.00012645918141883733, + "loss": 2.5106, + "step": 19525 + }, + { + "epoch": 0.83, + "grad_norm": 0.42578125, + "learning_rate": 0.0001264527124951152, + "loss": 2.5418, + "step": 19526 + }, + { + "epoch": 0.83, + "grad_norm": 0.400390625, + "learning_rate": 0.00012644624345236553, + "loss": 2.4463, + "step": 19527 + }, + { + "epoch": 0.83, + "grad_norm": 0.490234375, + "learning_rate": 0.0001264397742906174, + "loss": 2.6477, + "step": 19528 + }, + { + "epoch": 0.83, + "grad_norm": 0.39453125, + "learning_rate": 0.00012643330500990003, + "loss": 2.5481, + "step": 19529 + }, + { + "epoch": 0.83, + "grad_norm": 0.4375, + "learning_rate": 0.0001264268356102424, + "loss": 2.5934, + "step": 19530 + }, + { + "epoch": 0.83, + "grad_norm": 0.455078125, + "learning_rate": 0.00012642036609167363, + "loss": 2.5085, + "step": 19531 + }, + { + "epoch": 0.83, + "grad_norm": 0.45703125, + "learning_rate": 0.00012641389645422292, + "loss": 2.6495, + "step": 19532 + }, + { + "epoch": 0.83, + "grad_norm": 0.392578125, + "learning_rate": 0.0001264074266979193, + "loss": 2.5473, + "step": 19533 + }, + { + "epoch": 0.83, + "grad_norm": 0.51171875, + "learning_rate": 0.00012640095682279193, + "loss": 2.6012, + "step": 19534 + }, + { + "epoch": 0.83, + "grad_norm": 0.5, + "learning_rate": 0.0001263944868288699, + "loss": 2.6676, + "step": 19535 + }, + { + "epoch": 0.83, + "grad_norm": 0.48828125, + "learning_rate": 0.0001263880167161823, + "loss": 2.4943, + "step": 19536 + }, + { + "epoch": 0.83, + "grad_norm": 0.4453125, + "learning_rate": 0.00012638154648475833, + "loss": 2.4426, + "step": 19537 + }, + { + "epoch": 0.83, + "grad_norm": 0.486328125, + "learning_rate": 0.00012637507613462697, + "loss": 2.642, + "step": 19538 + }, + { + "epoch": 0.83, + "grad_norm": 0.4453125, + "learning_rate": 0.00012636860566581746, + "loss": 2.7174, + "step": 19539 + }, + { + "epoch": 0.83, + "grad_norm": 0.43359375, + "learning_rate": 0.00012636213507835882, + "loss": 2.5354, + "step": 19540 + }, + { + "epoch": 0.83, + "grad_norm": 0.45703125, + "learning_rate": 0.00012635566437228026, + "loss": 2.5897, + "step": 19541 + }, + { + "epoch": 0.83, + "grad_norm": 0.458984375, + "learning_rate": 0.0001263491935476108, + "loss": 2.5192, + "step": 19542 + }, + { + "epoch": 0.83, + "grad_norm": 0.455078125, + "learning_rate": 0.00012634272260437962, + "loss": 2.5256, + "step": 19543 + }, + { + "epoch": 0.83, + "grad_norm": 0.45703125, + "learning_rate": 0.00012633625154261576, + "loss": 2.5628, + "step": 19544 + }, + { + "epoch": 0.83, + "grad_norm": 0.396484375, + "learning_rate": 0.00012632978036234847, + "loss": 2.5008, + "step": 19545 + }, + { + "epoch": 0.83, + "grad_norm": 0.443359375, + "learning_rate": 0.00012632330906360676, + "loss": 2.5888, + "step": 19546 + }, + { + "epoch": 0.83, + "grad_norm": 0.373046875, + "learning_rate": 0.00012631683764641974, + "loss": 2.3573, + "step": 19547 + }, + { + "epoch": 0.83, + "grad_norm": 0.46875, + "learning_rate": 0.0001263103661108166, + "loss": 2.5036, + "step": 19548 + }, + { + "epoch": 0.83, + "grad_norm": 0.400390625, + "learning_rate": 0.00012630389445682648, + "loss": 2.4527, + "step": 19549 + }, + { + "epoch": 0.83, + "grad_norm": 0.416015625, + "learning_rate": 0.00012629742268447836, + "loss": 2.5722, + "step": 19550 + }, + { + "epoch": 0.83, + "grad_norm": 0.49609375, + "learning_rate": 0.0001262909507938015, + "loss": 2.7627, + "step": 19551 + }, + { + "epoch": 0.83, + "grad_norm": 0.53125, + "learning_rate": 0.00012628447878482496, + "loss": 2.6731, + "step": 19552 + }, + { + "epoch": 0.83, + "grad_norm": 0.45703125, + "learning_rate": 0.0001262780066575779, + "loss": 2.423, + "step": 19553 + }, + { + "epoch": 0.83, + "grad_norm": 0.47265625, + "learning_rate": 0.00012627153441208935, + "loss": 2.6886, + "step": 19554 + }, + { + "epoch": 0.83, + "grad_norm": 0.5234375, + "learning_rate": 0.00012626506204838853, + "loss": 2.5386, + "step": 19555 + }, + { + "epoch": 0.83, + "grad_norm": 0.462890625, + "learning_rate": 0.00012625858956650453, + "loss": 2.7205, + "step": 19556 + }, + { + "epoch": 0.83, + "grad_norm": 0.484375, + "learning_rate": 0.00012625211696646647, + "loss": 2.4659, + "step": 19557 + }, + { + "epoch": 0.83, + "grad_norm": 0.5, + "learning_rate": 0.0001262456442483035, + "loss": 2.4759, + "step": 19558 + }, + { + "epoch": 0.83, + "grad_norm": 0.474609375, + "learning_rate": 0.00012623917141204468, + "loss": 2.5252, + "step": 19559 + }, + { + "epoch": 0.83, + "grad_norm": 0.42578125, + "learning_rate": 0.0001262326984577192, + "loss": 2.4511, + "step": 19560 + }, + { + "epoch": 0.83, + "grad_norm": 0.4765625, + "learning_rate": 0.0001262262253853562, + "loss": 2.578, + "step": 19561 + }, + { + "epoch": 0.83, + "grad_norm": 0.5, + "learning_rate": 0.0001262197521949847, + "loss": 2.6805, + "step": 19562 + }, + { + "epoch": 0.83, + "grad_norm": 0.51953125, + "learning_rate": 0.00012621327888663394, + "loss": 2.4884, + "step": 19563 + }, + { + "epoch": 0.83, + "grad_norm": 0.3984375, + "learning_rate": 0.000126206805460333, + "loss": 2.5043, + "step": 19564 + }, + { + "epoch": 0.83, + "grad_norm": 0.52734375, + "learning_rate": 0.000126200331916111, + "loss": 2.5724, + "step": 19565 + }, + { + "epoch": 0.83, + "grad_norm": 0.51953125, + "learning_rate": 0.0001261938582539971, + "loss": 2.7046, + "step": 19566 + }, + { + "epoch": 0.83, + "grad_norm": 0.51171875, + "learning_rate": 0.0001261873844740204, + "loss": 2.6188, + "step": 19567 + }, + { + "epoch": 0.83, + "grad_norm": 0.4453125, + "learning_rate": 0.00012618091057621004, + "loss": 2.4486, + "step": 19568 + }, + { + "epoch": 0.83, + "grad_norm": 0.435546875, + "learning_rate": 0.00012617443656059518, + "loss": 2.5925, + "step": 19569 + }, + { + "epoch": 0.83, + "grad_norm": 0.498046875, + "learning_rate": 0.00012616796242720488, + "loss": 2.56, + "step": 19570 + }, + { + "epoch": 0.83, + "grad_norm": 0.4375, + "learning_rate": 0.00012616148817606832, + "loss": 2.5401, + "step": 19571 + }, + { + "epoch": 0.83, + "grad_norm": 0.451171875, + "learning_rate": 0.00012615501380721465, + "loss": 2.5567, + "step": 19572 + }, + { + "epoch": 0.83, + "grad_norm": 0.4921875, + "learning_rate": 0.00012614853932067295, + "loss": 2.6614, + "step": 19573 + }, + { + "epoch": 0.83, + "grad_norm": 0.46484375, + "learning_rate": 0.00012614206471647237, + "loss": 2.6582, + "step": 19574 + }, + { + "epoch": 0.83, + "grad_norm": 0.419921875, + "learning_rate": 0.00012613558999464207, + "loss": 2.6196, + "step": 19575 + }, + { + "epoch": 0.83, + "grad_norm": 0.462890625, + "learning_rate": 0.00012612911515521116, + "loss": 2.7096, + "step": 19576 + }, + { + "epoch": 0.83, + "grad_norm": 0.44140625, + "learning_rate": 0.00012612264019820881, + "loss": 2.5653, + "step": 19577 + }, + { + "epoch": 0.83, + "grad_norm": 0.455078125, + "learning_rate": 0.00012611616512366407, + "loss": 2.5491, + "step": 19578 + }, + { + "epoch": 0.83, + "grad_norm": 0.5, + "learning_rate": 0.00012610968993160616, + "loss": 2.5968, + "step": 19579 + }, + { + "epoch": 0.83, + "grad_norm": 0.4609375, + "learning_rate": 0.0001261032146220642, + "loss": 2.5115, + "step": 19580 + }, + { + "epoch": 0.83, + "grad_norm": 0.392578125, + "learning_rate": 0.00012609673919506732, + "loss": 2.4718, + "step": 19581 + }, + { + "epoch": 0.83, + "grad_norm": 0.427734375, + "learning_rate": 0.00012609026365064465, + "loss": 2.4705, + "step": 19582 + }, + { + "epoch": 0.83, + "grad_norm": 0.435546875, + "learning_rate": 0.0001260837879888253, + "loss": 2.5601, + "step": 19583 + }, + { + "epoch": 0.83, + "grad_norm": 0.49609375, + "learning_rate": 0.00012607731220963842, + "loss": 2.5361, + "step": 19584 + }, + { + "epoch": 0.83, + "grad_norm": 0.435546875, + "learning_rate": 0.00012607083631311323, + "loss": 2.5172, + "step": 19585 + }, + { + "epoch": 0.83, + "grad_norm": 0.45703125, + "learning_rate": 0.00012606436029927875, + "loss": 2.5653, + "step": 19586 + }, + { + "epoch": 0.83, + "grad_norm": 0.44921875, + "learning_rate": 0.0001260578841681642, + "loss": 2.576, + "step": 19587 + }, + { + "epoch": 0.83, + "grad_norm": 0.416015625, + "learning_rate": 0.0001260514079197987, + "loss": 2.6089, + "step": 19588 + }, + { + "epoch": 0.83, + "grad_norm": 0.453125, + "learning_rate": 0.0001260449315542114, + "loss": 2.6664, + "step": 19589 + }, + { + "epoch": 0.83, + "grad_norm": 0.47265625, + "learning_rate": 0.00012603845507143136, + "loss": 2.4337, + "step": 19590 + }, + { + "epoch": 0.83, + "grad_norm": 0.43359375, + "learning_rate": 0.00012603197847148783, + "loss": 2.5603, + "step": 19591 + }, + { + "epoch": 0.83, + "grad_norm": 0.4765625, + "learning_rate": 0.0001260255017544099, + "loss": 2.552, + "step": 19592 + }, + { + "epoch": 0.83, + "grad_norm": 0.482421875, + "learning_rate": 0.00012601902492022674, + "loss": 2.7159, + "step": 19593 + }, + { + "epoch": 0.83, + "grad_norm": 0.478515625, + "learning_rate": 0.00012601254796896746, + "loss": 2.5771, + "step": 19594 + }, + { + "epoch": 0.83, + "grad_norm": 0.443359375, + "learning_rate": 0.00012600607090066122, + "loss": 2.4706, + "step": 19595 + }, + { + "epoch": 0.84, + "grad_norm": 0.490234375, + "learning_rate": 0.00012599959371533719, + "loss": 2.562, + "step": 19596 + }, + { + "epoch": 0.84, + "grad_norm": 0.4140625, + "learning_rate": 0.00012599311641302448, + "loss": 2.5386, + "step": 19597 + }, + { + "epoch": 0.84, + "grad_norm": 0.5234375, + "learning_rate": 0.00012598663899375225, + "loss": 2.6529, + "step": 19598 + }, + { + "epoch": 0.84, + "grad_norm": 0.46875, + "learning_rate": 0.00012598016145754962, + "loss": 2.6569, + "step": 19599 + }, + { + "epoch": 0.84, + "grad_norm": 0.482421875, + "learning_rate": 0.00012597368380444577, + "loss": 2.6762, + "step": 19600 + }, + { + "epoch": 0.84, + "grad_norm": 0.451171875, + "learning_rate": 0.00012596720603446986, + "loss": 2.607, + "step": 19601 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012596072814765097, + "loss": 2.4723, + "step": 19602 + }, + { + "epoch": 0.84, + "grad_norm": 0.52734375, + "learning_rate": 0.0001259542501440183, + "loss": 2.5997, + "step": 19603 + }, + { + "epoch": 0.84, + "grad_norm": 0.5, + "learning_rate": 0.000125947772023601, + "loss": 2.6197, + "step": 19604 + }, + { + "epoch": 0.84, + "grad_norm": 0.427734375, + "learning_rate": 0.0001259412937864282, + "loss": 2.5409, + "step": 19605 + }, + { + "epoch": 0.84, + "grad_norm": 0.470703125, + "learning_rate": 0.00012593481543252908, + "loss": 2.6158, + "step": 19606 + }, + { + "epoch": 0.84, + "grad_norm": 0.494140625, + "learning_rate": 0.00012592833696193275, + "loss": 2.5263, + "step": 19607 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012592185837466836, + "loss": 2.7767, + "step": 19608 + }, + { + "epoch": 0.84, + "grad_norm": 0.421875, + "learning_rate": 0.00012591537967076514, + "loss": 2.5343, + "step": 19609 + }, + { + "epoch": 0.84, + "grad_norm": 0.4296875, + "learning_rate": 0.00012590890085025213, + "loss": 2.5374, + "step": 19610 + }, + { + "epoch": 0.84, + "grad_norm": 0.490234375, + "learning_rate": 0.00012590242191315855, + "loss": 2.6459, + "step": 19611 + }, + { + "epoch": 0.84, + "grad_norm": 0.5078125, + "learning_rate": 0.00012589594285951353, + "loss": 2.5739, + "step": 19612 + }, + { + "epoch": 0.84, + "grad_norm": 0.435546875, + "learning_rate": 0.00012588946368934624, + "loss": 2.6399, + "step": 19613 + }, + { + "epoch": 0.84, + "grad_norm": 0.474609375, + "learning_rate": 0.00012588298440268582, + "loss": 2.5247, + "step": 19614 + }, + { + "epoch": 0.84, + "grad_norm": 0.42578125, + "learning_rate": 0.00012587650499956145, + "loss": 2.577, + "step": 19615 + }, + { + "epoch": 0.84, + "grad_norm": 0.4453125, + "learning_rate": 0.0001258700254800022, + "loss": 2.5882, + "step": 19616 + }, + { + "epoch": 0.84, + "grad_norm": 0.45703125, + "learning_rate": 0.00012586354584403738, + "loss": 2.5988, + "step": 19617 + }, + { + "epoch": 0.84, + "grad_norm": 0.59375, + "learning_rate": 0.000125857066091696, + "loss": 2.5637, + "step": 19618 + }, + { + "epoch": 0.84, + "grad_norm": 0.515625, + "learning_rate": 0.00012585058622300728, + "loss": 2.6806, + "step": 19619 + }, + { + "epoch": 0.84, + "grad_norm": 0.4765625, + "learning_rate": 0.00012584410623800037, + "loss": 2.5992, + "step": 19620 + }, + { + "epoch": 0.84, + "grad_norm": 0.48046875, + "learning_rate": 0.00012583762613670448, + "loss": 2.5792, + "step": 19621 + }, + { + "epoch": 0.84, + "grad_norm": 0.44921875, + "learning_rate": 0.00012583114591914864, + "loss": 2.633, + "step": 19622 + }, + { + "epoch": 0.84, + "grad_norm": 0.455078125, + "learning_rate": 0.00012582466558536213, + "loss": 2.5243, + "step": 19623 + }, + { + "epoch": 0.84, + "grad_norm": 0.419921875, + "learning_rate": 0.00012581818513537403, + "loss": 2.5836, + "step": 19624 + }, + { + "epoch": 0.84, + "grad_norm": 0.4921875, + "learning_rate": 0.00012581170456921355, + "loss": 2.6143, + "step": 19625 + }, + { + "epoch": 0.84, + "grad_norm": 0.4375, + "learning_rate": 0.00012580522388690982, + "loss": 2.5335, + "step": 19626 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.00012579874308849202, + "loss": 2.5784, + "step": 19627 + }, + { + "epoch": 0.84, + "grad_norm": 0.5, + "learning_rate": 0.0001257922621739893, + "loss": 2.5792, + "step": 19628 + }, + { + "epoch": 0.84, + "grad_norm": 0.455078125, + "learning_rate": 0.00012578578114343084, + "loss": 2.5442, + "step": 19629 + }, + { + "epoch": 0.84, + "grad_norm": 0.470703125, + "learning_rate": 0.0001257792999968458, + "loss": 2.4346, + "step": 19630 + }, + { + "epoch": 0.84, + "grad_norm": 0.486328125, + "learning_rate": 0.0001257728187342633, + "loss": 2.4329, + "step": 19631 + }, + { + "epoch": 0.84, + "grad_norm": 0.44140625, + "learning_rate": 0.00012576633735571254, + "loss": 2.6965, + "step": 19632 + }, + { + "epoch": 0.84, + "grad_norm": 0.54296875, + "learning_rate": 0.00012575985586122269, + "loss": 2.4622, + "step": 19633 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012575337425082286, + "loss": 2.5209, + "step": 19634 + }, + { + "epoch": 0.84, + "grad_norm": 0.431640625, + "learning_rate": 0.00012574689252454232, + "loss": 2.4695, + "step": 19635 + }, + { + "epoch": 0.84, + "grad_norm": 0.455078125, + "learning_rate": 0.00012574041068241017, + "loss": 2.5487, + "step": 19636 + }, + { + "epoch": 0.84, + "grad_norm": 0.490234375, + "learning_rate": 0.00012573392872445552, + "loss": 2.5775, + "step": 19637 + }, + { + "epoch": 0.84, + "grad_norm": 0.470703125, + "learning_rate": 0.00012572744665070767, + "loss": 2.5265, + "step": 19638 + }, + { + "epoch": 0.84, + "grad_norm": 0.455078125, + "learning_rate": 0.00012572096446119568, + "loss": 2.6057, + "step": 19639 + }, + { + "epoch": 0.84, + "grad_norm": 0.412109375, + "learning_rate": 0.00012571448215594873, + "loss": 2.4798, + "step": 19640 + }, + { + "epoch": 0.84, + "grad_norm": 0.4453125, + "learning_rate": 0.00012570799973499604, + "loss": 2.6379, + "step": 19641 + }, + { + "epoch": 0.84, + "grad_norm": 0.447265625, + "learning_rate": 0.00012570151719836672, + "loss": 2.5535, + "step": 19642 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012569503454608994, + "loss": 2.5129, + "step": 19643 + }, + { + "epoch": 0.84, + "grad_norm": 0.439453125, + "learning_rate": 0.000125688551778195, + "loss": 2.6659, + "step": 19644 + }, + { + "epoch": 0.84, + "grad_norm": 0.451171875, + "learning_rate": 0.00012568206889471085, + "loss": 2.5518, + "step": 19645 + }, + { + "epoch": 0.84, + "grad_norm": 0.484375, + "learning_rate": 0.0001256755858956668, + "loss": 2.5534, + "step": 19646 + }, + { + "epoch": 0.84, + "grad_norm": 0.4609375, + "learning_rate": 0.00012566910278109203, + "loss": 2.6522, + "step": 19647 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.00012566261955101568, + "loss": 2.6838, + "step": 19648 + }, + { + "epoch": 0.84, + "grad_norm": 0.466796875, + "learning_rate": 0.0001256561362054669, + "loss": 2.5912, + "step": 19649 + }, + { + "epoch": 0.84, + "grad_norm": 0.427734375, + "learning_rate": 0.0001256496527444749, + "loss": 2.5184, + "step": 19650 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.00012564316916806878, + "loss": 2.5662, + "step": 19651 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.00012563668547627786, + "loss": 2.5537, + "step": 19652 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.00012563020166913118, + "loss": 2.6434, + "step": 19653 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012562371774665795, + "loss": 2.4422, + "step": 19654 + }, + { + "epoch": 0.84, + "grad_norm": 0.427734375, + "learning_rate": 0.00012561723370888735, + "loss": 2.4457, + "step": 19655 + }, + { + "epoch": 0.84, + "grad_norm": 0.490234375, + "learning_rate": 0.0001256107495558486, + "loss": 2.6169, + "step": 19656 + }, + { + "epoch": 0.84, + "grad_norm": 0.47265625, + "learning_rate": 0.0001256042652875708, + "loss": 2.5419, + "step": 19657 + }, + { + "epoch": 0.84, + "grad_norm": 0.451171875, + "learning_rate": 0.00012559778090408318, + "loss": 2.5442, + "step": 19658 + }, + { + "epoch": 0.84, + "grad_norm": 0.482421875, + "learning_rate": 0.0001255912964054149, + "loss": 2.5717, + "step": 19659 + }, + { + "epoch": 0.84, + "grad_norm": 0.4375, + "learning_rate": 0.00012558481179159516, + "loss": 2.609, + "step": 19660 + }, + { + "epoch": 0.84, + "grad_norm": 0.494140625, + "learning_rate": 0.00012557832706265308, + "loss": 2.5399, + "step": 19661 + }, + { + "epoch": 0.84, + "grad_norm": 0.43359375, + "learning_rate": 0.00012557184221861787, + "loss": 2.591, + "step": 19662 + }, + { + "epoch": 0.84, + "grad_norm": 0.490234375, + "learning_rate": 0.00012556535725951874, + "loss": 2.6116, + "step": 19663 + }, + { + "epoch": 0.84, + "grad_norm": 0.455078125, + "learning_rate": 0.00012555887218538485, + "loss": 2.5075, + "step": 19664 + }, + { + "epoch": 0.84, + "grad_norm": 0.447265625, + "learning_rate": 0.00012555238699624537, + "loss": 2.3967, + "step": 19665 + }, + { + "epoch": 0.84, + "grad_norm": 0.43359375, + "learning_rate": 0.00012554590169212948, + "loss": 2.6384, + "step": 19666 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012553941627306634, + "loss": 2.4818, + "step": 19667 + }, + { + "epoch": 0.84, + "grad_norm": 0.42578125, + "learning_rate": 0.0001255329307390852, + "loss": 2.6273, + "step": 19668 + }, + { + "epoch": 0.84, + "grad_norm": 0.44921875, + "learning_rate": 0.0001255264450902152, + "loss": 2.4538, + "step": 19669 + }, + { + "epoch": 0.84, + "grad_norm": 0.498046875, + "learning_rate": 0.00012551995932648553, + "loss": 2.8373, + "step": 19670 + }, + { + "epoch": 0.84, + "grad_norm": 0.447265625, + "learning_rate": 0.00012551347344792534, + "loss": 2.5655, + "step": 19671 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.00012550698745456386, + "loss": 2.5498, + "step": 19672 + }, + { + "epoch": 0.84, + "grad_norm": 0.474609375, + "learning_rate": 0.0001255005013464303, + "loss": 2.4982, + "step": 19673 + }, + { + "epoch": 0.84, + "grad_norm": 0.4296875, + "learning_rate": 0.00012549401512355375, + "loss": 2.5959, + "step": 19674 + }, + { + "epoch": 0.84, + "grad_norm": 0.515625, + "learning_rate": 0.00012548752878596345, + "loss": 2.5637, + "step": 19675 + }, + { + "epoch": 0.84, + "grad_norm": 0.48046875, + "learning_rate": 0.0001254810423336886, + "loss": 2.6542, + "step": 19676 + }, + { + "epoch": 0.84, + "grad_norm": 0.47265625, + "learning_rate": 0.00012547455576675834, + "loss": 2.5678, + "step": 19677 + }, + { + "epoch": 0.84, + "grad_norm": 0.50390625, + "learning_rate": 0.0001254680690852019, + "loss": 2.4953, + "step": 19678 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.00012546158228904852, + "loss": 2.5258, + "step": 19679 + }, + { + "epoch": 0.84, + "grad_norm": 0.490234375, + "learning_rate": 0.00012545509537832726, + "loss": 2.5828, + "step": 19680 + }, + { + "epoch": 0.84, + "grad_norm": 0.5, + "learning_rate": 0.0001254486083530674, + "loss": 2.6703, + "step": 19681 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.0001254421212132981, + "loss": 2.4783, + "step": 19682 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012543563395904853, + "loss": 2.5341, + "step": 19683 + }, + { + "epoch": 0.84, + "grad_norm": 0.5078125, + "learning_rate": 0.00012542914659034797, + "loss": 2.5877, + "step": 19684 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012542265910722546, + "loss": 2.4679, + "step": 19685 + }, + { + "epoch": 0.84, + "grad_norm": 0.4609375, + "learning_rate": 0.00012541617150971032, + "loss": 2.572, + "step": 19686 + }, + { + "epoch": 0.84, + "grad_norm": 0.435546875, + "learning_rate": 0.0001254096837978317, + "loss": 2.5667, + "step": 19687 + }, + { + "epoch": 0.84, + "grad_norm": 0.482421875, + "learning_rate": 0.00012540319597161877, + "loss": 2.6477, + "step": 19688 + }, + { + "epoch": 0.84, + "grad_norm": 0.408203125, + "learning_rate": 0.00012539670803110074, + "loss": 2.4785, + "step": 19689 + }, + { + "epoch": 0.84, + "grad_norm": 0.478515625, + "learning_rate": 0.00012539021997630683, + "loss": 2.5094, + "step": 19690 + }, + { + "epoch": 0.84, + "grad_norm": 0.431640625, + "learning_rate": 0.00012538373180726617, + "loss": 2.5579, + "step": 19691 + }, + { + "epoch": 0.84, + "grad_norm": 0.47265625, + "learning_rate": 0.00012537724352400804, + "loss": 2.6083, + "step": 19692 + }, + { + "epoch": 0.84, + "grad_norm": 0.48046875, + "learning_rate": 0.00012537075512656154, + "loss": 2.5437, + "step": 19693 + }, + { + "epoch": 0.84, + "grad_norm": 0.52734375, + "learning_rate": 0.00012536426661495596, + "loss": 2.5551, + "step": 19694 + }, + { + "epoch": 0.84, + "grad_norm": 0.447265625, + "learning_rate": 0.00012535777798922042, + "loss": 2.5298, + "step": 19695 + }, + { + "epoch": 0.84, + "grad_norm": 0.51171875, + "learning_rate": 0.00012535128924938416, + "loss": 2.6366, + "step": 19696 + }, + { + "epoch": 0.84, + "grad_norm": 0.486328125, + "learning_rate": 0.00012534480039547635, + "loss": 2.5817, + "step": 19697 + }, + { + "epoch": 0.84, + "grad_norm": 0.4375, + "learning_rate": 0.00012533831142752622, + "loss": 2.5608, + "step": 19698 + }, + { + "epoch": 0.84, + "grad_norm": 0.392578125, + "learning_rate": 0.0001253318223455629, + "loss": 2.5669, + "step": 19699 + }, + { + "epoch": 0.84, + "grad_norm": 0.388671875, + "learning_rate": 0.00012532533314961572, + "loss": 2.4442, + "step": 19700 + }, + { + "epoch": 0.84, + "grad_norm": 0.51953125, + "learning_rate": 0.0001253188438397137, + "loss": 2.6849, + "step": 19701 + }, + { + "epoch": 0.84, + "grad_norm": 0.48828125, + "learning_rate": 0.00012531235441588622, + "loss": 2.5132, + "step": 19702 + }, + { + "epoch": 0.84, + "grad_norm": 0.43359375, + "learning_rate": 0.00012530586487816237, + "loss": 2.641, + "step": 19703 + }, + { + "epoch": 0.84, + "grad_norm": 0.474609375, + "learning_rate": 0.00012529937522657135, + "loss": 2.6374, + "step": 19704 + }, + { + "epoch": 0.84, + "grad_norm": 0.40625, + "learning_rate": 0.0001252928854611424, + "loss": 2.5247, + "step": 19705 + }, + { + "epoch": 0.84, + "grad_norm": 0.451171875, + "learning_rate": 0.00012528639558190472, + "loss": 2.4528, + "step": 19706 + }, + { + "epoch": 0.84, + "grad_norm": 0.466796875, + "learning_rate": 0.0001252799055888875, + "loss": 2.5162, + "step": 19707 + }, + { + "epoch": 0.84, + "grad_norm": 0.4921875, + "learning_rate": 0.00012527341548211992, + "loss": 2.4494, + "step": 19708 + }, + { + "epoch": 0.84, + "grad_norm": 0.44921875, + "learning_rate": 0.00012526692526163125, + "loss": 2.486, + "step": 19709 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.0001252604349274506, + "loss": 2.5479, + "step": 19710 + }, + { + "epoch": 0.84, + "grad_norm": 0.423828125, + "learning_rate": 0.00012525394447960726, + "loss": 2.5634, + "step": 19711 + }, + { + "epoch": 0.84, + "grad_norm": 0.43359375, + "learning_rate": 0.0001252474539181304, + "loss": 2.5517, + "step": 19712 + }, + { + "epoch": 0.84, + "grad_norm": 0.46484375, + "learning_rate": 0.0001252409632430492, + "loss": 2.4304, + "step": 19713 + }, + { + "epoch": 0.84, + "grad_norm": 0.3984375, + "learning_rate": 0.00012523447245439294, + "loss": 2.5376, + "step": 19714 + }, + { + "epoch": 0.84, + "grad_norm": 0.43359375, + "learning_rate": 0.0001252279815521907, + "loss": 2.4905, + "step": 19715 + }, + { + "epoch": 0.84, + "grad_norm": 0.486328125, + "learning_rate": 0.0001252214905364718, + "loss": 2.7928, + "step": 19716 + }, + { + "epoch": 0.84, + "grad_norm": 0.48828125, + "learning_rate": 0.00012521499940726543, + "loss": 2.6019, + "step": 19717 + }, + { + "epoch": 0.84, + "grad_norm": 0.5234375, + "learning_rate": 0.00012520850816460076, + "loss": 2.5652, + "step": 19718 + }, + { + "epoch": 0.84, + "grad_norm": 0.455078125, + "learning_rate": 0.00012520201680850702, + "loss": 2.5455, + "step": 19719 + }, + { + "epoch": 0.84, + "grad_norm": 0.4609375, + "learning_rate": 0.0001251955253390134, + "loss": 2.6, + "step": 19720 + }, + { + "epoch": 0.84, + "grad_norm": 0.4609375, + "learning_rate": 0.00012518903375614913, + "loss": 2.6558, + "step": 19721 + }, + { + "epoch": 0.84, + "grad_norm": 0.421875, + "learning_rate": 0.00012518254205994345, + "loss": 2.625, + "step": 19722 + }, + { + "epoch": 0.84, + "grad_norm": 0.451171875, + "learning_rate": 0.00012517605025042547, + "loss": 2.6404, + "step": 19723 + }, + { + "epoch": 0.84, + "grad_norm": 0.447265625, + "learning_rate": 0.00012516955832762448, + "loss": 2.5495, + "step": 19724 + }, + { + "epoch": 0.84, + "grad_norm": 0.50390625, + "learning_rate": 0.00012516306629156972, + "loss": 2.7225, + "step": 19725 + }, + { + "epoch": 0.84, + "grad_norm": 0.42578125, + "learning_rate": 0.00012515657414229034, + "loss": 2.5146, + "step": 19726 + }, + { + "epoch": 0.84, + "grad_norm": 0.52734375, + "learning_rate": 0.00012515008187981557, + "loss": 2.5755, + "step": 19727 + }, + { + "epoch": 0.84, + "grad_norm": 0.439453125, + "learning_rate": 0.0001251435895041746, + "loss": 2.606, + "step": 19728 + }, + { + "epoch": 0.84, + "grad_norm": 0.404296875, + "learning_rate": 0.00012513709701539666, + "loss": 2.4413, + "step": 19729 + }, + { + "epoch": 0.84, + "grad_norm": 0.46875, + "learning_rate": 0.000125130604413511, + "loss": 2.6039, + "step": 19730 + }, + { + "epoch": 0.84, + "grad_norm": 0.404296875, + "learning_rate": 0.0001251241116985468, + "loss": 2.5022, + "step": 19731 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.00012511761887053327, + "loss": 2.5319, + "step": 19732 + }, + { + "epoch": 0.84, + "grad_norm": 0.435546875, + "learning_rate": 0.00012511112592949962, + "loss": 2.5194, + "step": 19733 + }, + { + "epoch": 0.84, + "grad_norm": 0.4140625, + "learning_rate": 0.0001251046328754751, + "loss": 2.5717, + "step": 19734 + }, + { + "epoch": 0.84, + "grad_norm": 0.46875, + "learning_rate": 0.00012509813970848892, + "loss": 2.5667, + "step": 19735 + }, + { + "epoch": 0.84, + "grad_norm": 0.439453125, + "learning_rate": 0.00012509164642857028, + "loss": 2.5445, + "step": 19736 + }, + { + "epoch": 0.84, + "grad_norm": 0.44921875, + "learning_rate": 0.00012508515303574837, + "loss": 2.5742, + "step": 19737 + }, + { + "epoch": 0.84, + "grad_norm": 0.4375, + "learning_rate": 0.00012507865953005246, + "loss": 2.5535, + "step": 19738 + }, + { + "epoch": 0.84, + "grad_norm": 0.46875, + "learning_rate": 0.00012507216591151174, + "loss": 2.533, + "step": 19739 + }, + { + "epoch": 0.84, + "grad_norm": 0.427734375, + "learning_rate": 0.00012506567218015545, + "loss": 2.6375, + "step": 19740 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.0001250591783360128, + "loss": 2.5415, + "step": 19741 + }, + { + "epoch": 0.84, + "grad_norm": 0.45703125, + "learning_rate": 0.00012505268437911296, + "loss": 2.6115, + "step": 19742 + }, + { + "epoch": 0.84, + "grad_norm": 0.4140625, + "learning_rate": 0.00012504619030948526, + "loss": 2.5333, + "step": 19743 + }, + { + "epoch": 0.84, + "grad_norm": 0.4140625, + "learning_rate": 0.0001250396961271588, + "loss": 2.4725, + "step": 19744 + }, + { + "epoch": 0.84, + "grad_norm": 0.4296875, + "learning_rate": 0.0001250332018321629, + "loss": 2.6023, + "step": 19745 + }, + { + "epoch": 0.84, + "grad_norm": 0.443359375, + "learning_rate": 0.00012502670742452673, + "loss": 2.5259, + "step": 19746 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.0001250202129042795, + "loss": 2.5705, + "step": 19747 + }, + { + "epoch": 0.84, + "grad_norm": 0.470703125, + "learning_rate": 0.0001250137182714505, + "loss": 2.5604, + "step": 19748 + }, + { + "epoch": 0.84, + "grad_norm": 0.50390625, + "learning_rate": 0.00012500722352606886, + "loss": 2.7062, + "step": 19749 + }, + { + "epoch": 0.84, + "grad_norm": 0.435546875, + "learning_rate": 0.0001250007286681639, + "loss": 2.4601, + "step": 19750 + }, + { + "epoch": 0.84, + "grad_norm": 0.4296875, + "learning_rate": 0.00012499423369776477, + "loss": 2.6376, + "step": 19751 + }, + { + "epoch": 0.84, + "grad_norm": 0.42578125, + "learning_rate": 0.00012498773861490073, + "loss": 2.5775, + "step": 19752 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.000124981243419601, + "loss": 2.6408, + "step": 19753 + }, + { + "epoch": 0.84, + "grad_norm": 0.43359375, + "learning_rate": 0.0001249747481118948, + "loss": 2.4288, + "step": 19754 + }, + { + "epoch": 0.84, + "grad_norm": 0.462890625, + "learning_rate": 0.00012496825269181134, + "loss": 2.5683, + "step": 19755 + }, + { + "epoch": 0.84, + "grad_norm": 0.46484375, + "learning_rate": 0.00012496175715937988, + "loss": 2.4754, + "step": 19756 + }, + { + "epoch": 0.84, + "grad_norm": 0.46875, + "learning_rate": 0.00012495526151462967, + "loss": 2.5687, + "step": 19757 + }, + { + "epoch": 0.84, + "grad_norm": 0.451171875, + "learning_rate": 0.00012494876575758985, + "loss": 2.5782, + "step": 19758 + }, + { + "epoch": 0.84, + "grad_norm": 0.49609375, + "learning_rate": 0.00012494226988828973, + "loss": 2.5048, + "step": 19759 + }, + { + "epoch": 0.84, + "grad_norm": 0.4609375, + "learning_rate": 0.0001249357739067585, + "loss": 2.5761, + "step": 19760 + }, + { + "epoch": 0.84, + "grad_norm": 0.4921875, + "learning_rate": 0.00012492927781302543, + "loss": 2.5974, + "step": 19761 + }, + { + "epoch": 0.84, + "grad_norm": 0.5234375, + "learning_rate": 0.0001249227816071197, + "loss": 2.4657, + "step": 19762 + }, + { + "epoch": 0.84, + "grad_norm": 0.50390625, + "learning_rate": 0.00012491628528907056, + "loss": 2.5898, + "step": 19763 + }, + { + "epoch": 0.84, + "grad_norm": 0.4609375, + "learning_rate": 0.00012490978885890724, + "loss": 2.5292, + "step": 19764 + }, + { + "epoch": 0.84, + "grad_norm": 0.5, + "learning_rate": 0.000124903292316659, + "loss": 2.5582, + "step": 19765 + }, + { + "epoch": 0.84, + "grad_norm": 0.416015625, + "learning_rate": 0.000124896795662355, + "loss": 2.4569, + "step": 19766 + }, + { + "epoch": 0.84, + "grad_norm": 0.455078125, + "learning_rate": 0.00012489029889602453, + "loss": 2.5114, + "step": 19767 + }, + { + "epoch": 0.84, + "grad_norm": 0.486328125, + "learning_rate": 0.00012488380201769683, + "loss": 2.5713, + "step": 19768 + }, + { + "epoch": 0.84, + "grad_norm": 0.466796875, + "learning_rate": 0.00012487730502740111, + "loss": 2.6726, + "step": 19769 + }, + { + "epoch": 0.84, + "grad_norm": 0.466796875, + "learning_rate": 0.00012487080792516662, + "loss": 2.5463, + "step": 19770 + }, + { + "epoch": 0.84, + "grad_norm": 0.45703125, + "learning_rate": 0.00012486431071102256, + "loss": 2.6665, + "step": 19771 + }, + { + "epoch": 0.84, + "grad_norm": 0.44921875, + "learning_rate": 0.0001248578133849982, + "loss": 2.4562, + "step": 19772 + }, + { + "epoch": 0.84, + "grad_norm": 0.48828125, + "learning_rate": 0.00012485131594712275, + "loss": 2.7411, + "step": 19773 + }, + { + "epoch": 0.84, + "grad_norm": 0.4375, + "learning_rate": 0.00012484481839742544, + "loss": 2.5575, + "step": 19774 + }, + { + "epoch": 0.84, + "grad_norm": 0.451171875, + "learning_rate": 0.00012483832073593559, + "loss": 2.4381, + "step": 19775 + }, + { + "epoch": 0.84, + "grad_norm": 0.421875, + "learning_rate": 0.00012483182296268235, + "loss": 2.5054, + "step": 19776 + }, + { + "epoch": 0.84, + "grad_norm": 0.431640625, + "learning_rate": 0.00012482532507769497, + "loss": 2.5016, + "step": 19777 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.0001248188270810027, + "loss": 2.6206, + "step": 19778 + }, + { + "epoch": 0.84, + "grad_norm": 0.400390625, + "learning_rate": 0.0001248123289726348, + "loss": 2.561, + "step": 19779 + }, + { + "epoch": 0.84, + "grad_norm": 0.40234375, + "learning_rate": 0.00012480583075262044, + "loss": 2.426, + "step": 19780 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.00012479933242098897, + "loss": 2.581, + "step": 19781 + }, + { + "epoch": 0.84, + "grad_norm": 0.51953125, + "learning_rate": 0.0001247928339777695, + "loss": 2.6206, + "step": 19782 + }, + { + "epoch": 0.84, + "grad_norm": 0.46484375, + "learning_rate": 0.00012478633542299138, + "loss": 2.6571, + "step": 19783 + }, + { + "epoch": 0.84, + "grad_norm": 0.435546875, + "learning_rate": 0.00012477983675668382, + "loss": 2.5718, + "step": 19784 + }, + { + "epoch": 0.84, + "grad_norm": 0.45703125, + "learning_rate": 0.00012477333797887604, + "loss": 2.6341, + "step": 19785 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.0001247668390895973, + "loss": 2.4803, + "step": 19786 + }, + { + "epoch": 0.84, + "grad_norm": 0.494140625, + "learning_rate": 0.00012476034008887678, + "loss": 2.6508, + "step": 19787 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.0001247538409767438, + "loss": 2.4211, + "step": 19788 + }, + { + "epoch": 0.84, + "grad_norm": 0.462890625, + "learning_rate": 0.00012474734175322764, + "loss": 2.6118, + "step": 19789 + }, + { + "epoch": 0.84, + "grad_norm": 0.439453125, + "learning_rate": 0.0001247408424183574, + "loss": 2.5105, + "step": 19790 + }, + { + "epoch": 0.84, + "grad_norm": 0.4453125, + "learning_rate": 0.00012473434297216249, + "loss": 2.4468, + "step": 19791 + }, + { + "epoch": 0.84, + "grad_norm": 0.431640625, + "learning_rate": 0.00012472784341467205, + "loss": 2.5026, + "step": 19792 + }, + { + "epoch": 0.84, + "grad_norm": 0.49609375, + "learning_rate": 0.00012472134374591533, + "loss": 2.6492, + "step": 19793 + }, + { + "epoch": 0.84, + "grad_norm": 0.435546875, + "learning_rate": 0.00012471484396592164, + "loss": 2.563, + "step": 19794 + }, + { + "epoch": 0.84, + "grad_norm": 0.4609375, + "learning_rate": 0.00012470834407472014, + "loss": 2.6486, + "step": 19795 + }, + { + "epoch": 0.84, + "grad_norm": 0.421875, + "learning_rate": 0.00012470184407234014, + "loss": 2.4899, + "step": 19796 + }, + { + "epoch": 0.84, + "grad_norm": 0.431640625, + "learning_rate": 0.0001246953439588109, + "loss": 2.5413, + "step": 19797 + }, + { + "epoch": 0.84, + "grad_norm": 0.466796875, + "learning_rate": 0.00012468884373416155, + "loss": 2.4696, + "step": 19798 + }, + { + "epoch": 0.84, + "grad_norm": 0.55078125, + "learning_rate": 0.0001246823433984215, + "loss": 2.5094, + "step": 19799 + }, + { + "epoch": 0.84, + "grad_norm": 0.482421875, + "learning_rate": 0.00012467584295161992, + "loss": 2.5566, + "step": 19800 + }, + { + "epoch": 0.84, + "grad_norm": 0.412109375, + "learning_rate": 0.00012466934239378606, + "loss": 2.5312, + "step": 19801 + }, + { + "epoch": 0.84, + "grad_norm": 0.515625, + "learning_rate": 0.00012466284172494915, + "loss": 2.4663, + "step": 19802 + }, + { + "epoch": 0.84, + "grad_norm": 0.51171875, + "learning_rate": 0.0001246563409451385, + "loss": 2.5752, + "step": 19803 + }, + { + "epoch": 0.84, + "grad_norm": 0.44921875, + "learning_rate": 0.00012464984005438327, + "loss": 2.46, + "step": 19804 + }, + { + "epoch": 0.84, + "grad_norm": 0.439453125, + "learning_rate": 0.00012464333905271285, + "loss": 2.5304, + "step": 19805 + }, + { + "epoch": 0.84, + "grad_norm": 0.435546875, + "learning_rate": 0.00012463683794015633, + "loss": 2.587, + "step": 19806 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.0001246303367167431, + "loss": 2.5134, + "step": 19807 + }, + { + "epoch": 0.84, + "grad_norm": 0.52734375, + "learning_rate": 0.00012462383538250233, + "loss": 2.6605, + "step": 19808 + }, + { + "epoch": 0.84, + "grad_norm": 0.44140625, + "learning_rate": 0.0001246173339374633, + "loss": 2.4653, + "step": 19809 + }, + { + "epoch": 0.84, + "grad_norm": 0.42578125, + "learning_rate": 0.00012461083238165526, + "loss": 2.5048, + "step": 19810 + }, + { + "epoch": 0.84, + "grad_norm": 0.5, + "learning_rate": 0.0001246043307151075, + "loss": 2.5592, + "step": 19811 + }, + { + "epoch": 0.84, + "grad_norm": 0.482421875, + "learning_rate": 0.0001245978289378492, + "loss": 2.5757, + "step": 19812 + }, + { + "epoch": 0.84, + "grad_norm": 0.474609375, + "learning_rate": 0.0001245913270499097, + "loss": 2.5734, + "step": 19813 + }, + { + "epoch": 0.84, + "grad_norm": 0.4453125, + "learning_rate": 0.00012458482505131818, + "loss": 2.5366, + "step": 19814 + }, + { + "epoch": 0.84, + "grad_norm": 0.453125, + "learning_rate": 0.00012457832294210394, + "loss": 2.4858, + "step": 19815 + }, + { + "epoch": 0.84, + "grad_norm": 0.4140625, + "learning_rate": 0.00012457182072229625, + "loss": 2.4954, + "step": 19816 + }, + { + "epoch": 0.84, + "grad_norm": 0.47265625, + "learning_rate": 0.00012456531839192433, + "loss": 2.617, + "step": 19817 + }, + { + "epoch": 0.84, + "grad_norm": 0.4453125, + "learning_rate": 0.00012455881595101745, + "loss": 2.4678, + "step": 19818 + }, + { + "epoch": 0.84, + "grad_norm": 0.494140625, + "learning_rate": 0.0001245523133996049, + "loss": 2.6579, + "step": 19819 + }, + { + "epoch": 0.84, + "grad_norm": 0.46875, + "learning_rate": 0.0001245458107377159, + "loss": 2.5341, + "step": 19820 + }, + { + "epoch": 0.84, + "grad_norm": 0.49609375, + "learning_rate": 0.00012453930796537974, + "loss": 2.596, + "step": 19821 + }, + { + "epoch": 0.84, + "grad_norm": 0.447265625, + "learning_rate": 0.00012453280508262563, + "loss": 2.5159, + "step": 19822 + }, + { + "epoch": 0.84, + "grad_norm": 0.482421875, + "learning_rate": 0.00012452630208948287, + "loss": 2.7034, + "step": 19823 + }, + { + "epoch": 0.84, + "grad_norm": 0.458984375, + "learning_rate": 0.0001245197989859807, + "loss": 2.4046, + "step": 19824 + }, + { + "epoch": 0.84, + "grad_norm": 0.462890625, + "learning_rate": 0.0001245132957721484, + "loss": 2.6054, + "step": 19825 + }, + { + "epoch": 0.84, + "grad_norm": 0.46484375, + "learning_rate": 0.00012450679244801525, + "loss": 2.5093, + "step": 19826 + }, + { + "epoch": 0.84, + "grad_norm": 0.412109375, + "learning_rate": 0.00012450028901361048, + "loss": 2.5159, + "step": 19827 + }, + { + "epoch": 0.84, + "grad_norm": 0.419921875, + "learning_rate": 0.00012449378546896335, + "loss": 2.6254, + "step": 19828 + }, + { + "epoch": 0.84, + "grad_norm": 0.478515625, + "learning_rate": 0.0001244872818141032, + "loss": 2.5817, + "step": 19829 + }, + { + "epoch": 0.85, + "grad_norm": 0.470703125, + "learning_rate": 0.00012448077804905913, + "loss": 2.5422, + "step": 19830 + }, + { + "epoch": 0.85, + "grad_norm": 0.416015625, + "learning_rate": 0.00012447427417386058, + "loss": 2.6552, + "step": 19831 + }, + { + "epoch": 0.85, + "grad_norm": 0.466796875, + "learning_rate": 0.0001244677701885367, + "loss": 2.507, + "step": 19832 + }, + { + "epoch": 0.85, + "grad_norm": 0.4609375, + "learning_rate": 0.0001244612660931168, + "loss": 2.4181, + "step": 19833 + }, + { + "epoch": 0.85, + "grad_norm": 0.46484375, + "learning_rate": 0.00012445476188763016, + "loss": 2.5916, + "step": 19834 + }, + { + "epoch": 0.85, + "grad_norm": 0.51171875, + "learning_rate": 0.00012444825757210604, + "loss": 2.6062, + "step": 19835 + }, + { + "epoch": 0.85, + "grad_norm": 0.455078125, + "learning_rate": 0.00012444175314657368, + "loss": 2.5635, + "step": 19836 + }, + { + "epoch": 0.85, + "grad_norm": 0.474609375, + "learning_rate": 0.00012443524861106237, + "loss": 2.5032, + "step": 19837 + }, + { + "epoch": 0.85, + "grad_norm": 0.451171875, + "learning_rate": 0.00012442874396560133, + "loss": 2.5029, + "step": 19838 + }, + { + "epoch": 0.85, + "grad_norm": 0.4296875, + "learning_rate": 0.00012442223921021992, + "loss": 2.6109, + "step": 19839 + }, + { + "epoch": 0.85, + "grad_norm": 0.46875, + "learning_rate": 0.00012441573434494735, + "loss": 2.5263, + "step": 19840 + }, + { + "epoch": 0.85, + "grad_norm": 0.41796875, + "learning_rate": 0.0001244092293698129, + "loss": 2.5329, + "step": 19841 + }, + { + "epoch": 0.85, + "grad_norm": 0.46875, + "learning_rate": 0.00012440272428484582, + "loss": 2.5773, + "step": 19842 + }, + { + "epoch": 0.85, + "grad_norm": 0.451171875, + "learning_rate": 0.0001243962190900754, + "loss": 2.5869, + "step": 19843 + }, + { + "epoch": 0.85, + "grad_norm": 0.431640625, + "learning_rate": 0.0001243897137855309, + "loss": 2.4838, + "step": 19844 + }, + { + "epoch": 0.85, + "grad_norm": 0.453125, + "learning_rate": 0.00012438320837124167, + "loss": 2.555, + "step": 19845 + }, + { + "epoch": 0.85, + "grad_norm": 0.5234375, + "learning_rate": 0.00012437670284723683, + "loss": 2.5751, + "step": 19846 + }, + { + "epoch": 0.85, + "grad_norm": 0.423828125, + "learning_rate": 0.00012437019721354576, + "loss": 2.3531, + "step": 19847 + }, + { + "epoch": 0.85, + "grad_norm": 0.44140625, + "learning_rate": 0.00012436369147019773, + "loss": 2.5138, + "step": 19848 + }, + { + "epoch": 0.85, + "grad_norm": 0.4453125, + "learning_rate": 0.000124357185617222, + "loss": 2.5648, + "step": 19849 + }, + { + "epoch": 0.85, + "grad_norm": 0.49609375, + "learning_rate": 0.0001243506796546478, + "loss": 2.6195, + "step": 19850 + }, + { + "epoch": 0.85, + "grad_norm": 0.515625, + "learning_rate": 0.00012434417358250446, + "loss": 2.5694, + "step": 19851 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.0001243376674008212, + "loss": 2.4815, + "step": 19852 + }, + { + "epoch": 0.85, + "grad_norm": 0.4765625, + "learning_rate": 0.0001243311611096274, + "loss": 2.6235, + "step": 19853 + }, + { + "epoch": 0.85, + "grad_norm": 0.4609375, + "learning_rate": 0.00012432465470895224, + "loss": 2.7321, + "step": 19854 + }, + { + "epoch": 0.85, + "grad_norm": 0.404296875, + "learning_rate": 0.00012431814819882498, + "loss": 2.4823, + "step": 19855 + }, + { + "epoch": 0.85, + "grad_norm": 0.470703125, + "learning_rate": 0.000124311641579275, + "loss": 2.5643, + "step": 19856 + }, + { + "epoch": 0.85, + "grad_norm": 0.453125, + "learning_rate": 0.0001243051348503315, + "loss": 2.4207, + "step": 19857 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.00012429862801202374, + "loss": 2.6017, + "step": 19858 + }, + { + "epoch": 0.85, + "grad_norm": 0.51953125, + "learning_rate": 0.0001242921210643811, + "loss": 2.6685, + "step": 19859 + }, + { + "epoch": 0.85, + "grad_norm": 0.478515625, + "learning_rate": 0.0001242856140074327, + "loss": 2.6778, + "step": 19860 + }, + { + "epoch": 0.85, + "grad_norm": 0.490234375, + "learning_rate": 0.000124279106841208, + "loss": 2.6048, + "step": 19861 + }, + { + "epoch": 0.85, + "grad_norm": 0.443359375, + "learning_rate": 0.00012427259956573615, + "loss": 2.5552, + "step": 19862 + }, + { + "epoch": 0.85, + "grad_norm": 0.455078125, + "learning_rate": 0.00012426609218104648, + "loss": 2.6507, + "step": 19863 + }, + { + "epoch": 0.85, + "grad_norm": 0.47265625, + "learning_rate": 0.00012425958468716827, + "loss": 2.5879, + "step": 19864 + }, + { + "epoch": 0.85, + "grad_norm": 0.490234375, + "learning_rate": 0.00012425307708413078, + "loss": 2.6076, + "step": 19865 + }, + { + "epoch": 0.85, + "grad_norm": 0.4296875, + "learning_rate": 0.00012424656937196332, + "loss": 2.4736, + "step": 19866 + }, + { + "epoch": 0.85, + "grad_norm": 0.4296875, + "learning_rate": 0.00012424006155069517, + "loss": 2.4789, + "step": 19867 + }, + { + "epoch": 0.85, + "grad_norm": 0.49609375, + "learning_rate": 0.00012423355362035555, + "loss": 2.5265, + "step": 19868 + }, + { + "epoch": 0.85, + "grad_norm": 0.427734375, + "learning_rate": 0.0001242270455809739, + "loss": 2.5072, + "step": 19869 + }, + { + "epoch": 0.85, + "grad_norm": 0.408203125, + "learning_rate": 0.0001242205374325793, + "loss": 2.5149, + "step": 19870 + }, + { + "epoch": 0.85, + "grad_norm": 0.5703125, + "learning_rate": 0.00012421402917520114, + "loss": 2.5841, + "step": 19871 + }, + { + "epoch": 0.85, + "grad_norm": 0.474609375, + "learning_rate": 0.00012420752080886875, + "loss": 2.5835, + "step": 19872 + }, + { + "epoch": 0.85, + "grad_norm": 0.466796875, + "learning_rate": 0.00012420101233361132, + "loss": 2.6052, + "step": 19873 + }, + { + "epoch": 0.85, + "grad_norm": 0.453125, + "learning_rate": 0.00012419450374945821, + "loss": 2.5734, + "step": 19874 + }, + { + "epoch": 0.85, + "grad_norm": 0.431640625, + "learning_rate": 0.00012418799505643866, + "loss": 2.5101, + "step": 19875 + }, + { + "epoch": 0.85, + "grad_norm": 0.486328125, + "learning_rate": 0.00012418148625458194, + "loss": 2.6031, + "step": 19876 + }, + { + "epoch": 0.85, + "grad_norm": 0.470703125, + "learning_rate": 0.00012417497734391742, + "loss": 2.5205, + "step": 19877 + }, + { + "epoch": 0.85, + "grad_norm": 0.466796875, + "learning_rate": 0.00012416846832447433, + "loss": 2.5427, + "step": 19878 + }, + { + "epoch": 0.85, + "grad_norm": 0.419921875, + "learning_rate": 0.00012416195919628196, + "loss": 2.4841, + "step": 19879 + }, + { + "epoch": 0.85, + "grad_norm": 0.435546875, + "learning_rate": 0.0001241554499593696, + "loss": 2.6231, + "step": 19880 + }, + { + "epoch": 0.85, + "grad_norm": 0.470703125, + "learning_rate": 0.00012414894061376651, + "loss": 2.6255, + "step": 19881 + }, + { + "epoch": 0.85, + "grad_norm": 0.42578125, + "learning_rate": 0.00012414243115950205, + "loss": 2.6454, + "step": 19882 + }, + { + "epoch": 0.85, + "grad_norm": 0.439453125, + "learning_rate": 0.00012413592159660547, + "loss": 2.4656, + "step": 19883 + }, + { + "epoch": 0.85, + "grad_norm": 0.52734375, + "learning_rate": 0.0001241294119251061, + "loss": 2.5748, + "step": 19884 + }, + { + "epoch": 0.85, + "grad_norm": 0.4453125, + "learning_rate": 0.00012412290214503313, + "loss": 2.5083, + "step": 19885 + }, + { + "epoch": 0.85, + "grad_norm": 0.45703125, + "learning_rate": 0.00012411639225641595, + "loss": 2.5492, + "step": 19886 + }, + { + "epoch": 0.85, + "grad_norm": 0.482421875, + "learning_rate": 0.00012410988225928376, + "loss": 2.6622, + "step": 19887 + }, + { + "epoch": 0.85, + "grad_norm": 0.498046875, + "learning_rate": 0.000124103372153666, + "loss": 2.4734, + "step": 19888 + }, + { + "epoch": 0.85, + "grad_norm": 0.4609375, + "learning_rate": 0.00012409686193959183, + "loss": 2.517, + "step": 19889 + }, + { + "epoch": 0.85, + "grad_norm": 0.5625, + "learning_rate": 0.00012409035161709057, + "loss": 2.5281, + "step": 19890 + }, + { + "epoch": 0.85, + "grad_norm": 0.478515625, + "learning_rate": 0.00012408384118619157, + "loss": 2.676, + "step": 19891 + }, + { + "epoch": 0.85, + "grad_norm": 0.443359375, + "learning_rate": 0.00012407733064692406, + "loss": 2.4539, + "step": 19892 + }, + { + "epoch": 0.85, + "grad_norm": 0.48828125, + "learning_rate": 0.00012407081999931739, + "loss": 2.6631, + "step": 19893 + }, + { + "epoch": 0.85, + "grad_norm": 0.447265625, + "learning_rate": 0.0001240643092434008, + "loss": 2.7541, + "step": 19894 + }, + { + "epoch": 0.85, + "grad_norm": 0.470703125, + "learning_rate": 0.0001240577983792036, + "loss": 2.6794, + "step": 19895 + }, + { + "epoch": 0.85, + "grad_norm": 0.46875, + "learning_rate": 0.00012405128740675513, + "loss": 2.4458, + "step": 19896 + }, + { + "epoch": 0.85, + "grad_norm": 0.470703125, + "learning_rate": 0.00012404477632608466, + "loss": 2.6798, + "step": 19897 + }, + { + "epoch": 0.85, + "grad_norm": 0.431640625, + "learning_rate": 0.00012403826513722146, + "loss": 2.6057, + "step": 19898 + }, + { + "epoch": 0.85, + "grad_norm": 0.419921875, + "learning_rate": 0.00012403175384019486, + "loss": 2.5825, + "step": 19899 + }, + { + "epoch": 0.85, + "grad_norm": 0.46875, + "learning_rate": 0.00012402524243503414, + "loss": 2.537, + "step": 19900 + }, + { + "epoch": 0.85, + "grad_norm": 0.435546875, + "learning_rate": 0.00012401873092176863, + "loss": 2.3986, + "step": 19901 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.00012401221930042762, + "loss": 2.5377, + "step": 19902 + }, + { + "epoch": 0.85, + "grad_norm": 0.439453125, + "learning_rate": 0.00012400570757104036, + "loss": 2.4553, + "step": 19903 + }, + { + "epoch": 0.85, + "grad_norm": 0.455078125, + "learning_rate": 0.0001239991957336362, + "loss": 2.5497, + "step": 19904 + }, + { + "epoch": 0.85, + "grad_norm": 0.408203125, + "learning_rate": 0.00012399268378824444, + "loss": 2.5732, + "step": 19905 + }, + { + "epoch": 0.85, + "grad_norm": 0.53515625, + "learning_rate": 0.00012398617173489437, + "loss": 2.661, + "step": 19906 + }, + { + "epoch": 0.85, + "grad_norm": 0.494140625, + "learning_rate": 0.00012397965957361532, + "loss": 2.4729, + "step": 19907 + }, + { + "epoch": 0.85, + "grad_norm": 0.42578125, + "learning_rate": 0.0001239731473044365, + "loss": 2.4354, + "step": 19908 + }, + { + "epoch": 0.85, + "grad_norm": 0.50390625, + "learning_rate": 0.00012396663492738734, + "loss": 2.7252, + "step": 19909 + }, + { + "epoch": 0.85, + "grad_norm": 0.52734375, + "learning_rate": 0.00012396012244249706, + "loss": 2.6536, + "step": 19910 + }, + { + "epoch": 0.85, + "grad_norm": 0.482421875, + "learning_rate": 0.00012395360984979496, + "loss": 2.667, + "step": 19911 + }, + { + "epoch": 0.85, + "grad_norm": 0.45703125, + "learning_rate": 0.00012394709714931042, + "loss": 2.6533, + "step": 19912 + }, + { + "epoch": 0.85, + "grad_norm": 0.51953125, + "learning_rate": 0.00012394058434107264, + "loss": 2.5188, + "step": 19913 + }, + { + "epoch": 0.85, + "grad_norm": 0.484375, + "learning_rate": 0.000123934071425111, + "loss": 2.5195, + "step": 19914 + }, + { + "epoch": 0.85, + "grad_norm": 0.466796875, + "learning_rate": 0.0001239275584014548, + "loss": 2.5853, + "step": 19915 + }, + { + "epoch": 0.85, + "grad_norm": 0.447265625, + "learning_rate": 0.00012392104527013333, + "loss": 2.535, + "step": 19916 + }, + { + "epoch": 0.85, + "grad_norm": 0.4921875, + "learning_rate": 0.00012391453203117588, + "loss": 2.3741, + "step": 19917 + }, + { + "epoch": 0.85, + "grad_norm": 0.478515625, + "learning_rate": 0.00012390801868461182, + "loss": 2.4765, + "step": 19918 + }, + { + "epoch": 0.85, + "grad_norm": 0.396484375, + "learning_rate": 0.00012390150523047035, + "loss": 2.6742, + "step": 19919 + }, + { + "epoch": 0.85, + "grad_norm": 0.455078125, + "learning_rate": 0.00012389499166878088, + "loss": 2.5572, + "step": 19920 + }, + { + "epoch": 0.85, + "grad_norm": 0.515625, + "learning_rate": 0.00012388847799957265, + "loss": 2.668, + "step": 19921 + }, + { + "epoch": 0.85, + "grad_norm": 0.451171875, + "learning_rate": 0.00012388196422287505, + "loss": 2.4228, + "step": 19922 + }, + { + "epoch": 0.85, + "grad_norm": 0.49609375, + "learning_rate": 0.00012387545033871732, + "loss": 2.7029, + "step": 19923 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.00012386893634712878, + "loss": 2.4733, + "step": 19924 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.00012386242224813874, + "loss": 2.71, + "step": 19925 + }, + { + "epoch": 0.85, + "grad_norm": 0.4921875, + "learning_rate": 0.00012385590804177654, + "loss": 2.6092, + "step": 19926 + }, + { + "epoch": 0.85, + "grad_norm": 0.4453125, + "learning_rate": 0.00012384939372807143, + "loss": 2.6055, + "step": 19927 + }, + { + "epoch": 0.85, + "grad_norm": 0.423828125, + "learning_rate": 0.00012384287930705283, + "loss": 2.5981, + "step": 19928 + }, + { + "epoch": 0.85, + "grad_norm": 0.453125, + "learning_rate": 0.00012383636477874993, + "loss": 2.5856, + "step": 19929 + }, + { + "epoch": 0.85, + "grad_norm": 0.44140625, + "learning_rate": 0.00012382985014319214, + "loss": 2.5004, + "step": 19930 + }, + { + "epoch": 0.85, + "grad_norm": 0.5, + "learning_rate": 0.0001238233354004087, + "loss": 2.5496, + "step": 19931 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.000123816820550429, + "loss": 2.619, + "step": 19932 + }, + { + "epoch": 0.85, + "grad_norm": 0.455078125, + "learning_rate": 0.00012381030559328226, + "loss": 2.6679, + "step": 19933 + }, + { + "epoch": 0.85, + "grad_norm": 0.3984375, + "learning_rate": 0.0001238037905289979, + "loss": 2.3928, + "step": 19934 + }, + { + "epoch": 0.85, + "grad_norm": 0.45703125, + "learning_rate": 0.0001237972753576051, + "loss": 2.4069, + "step": 19935 + }, + { + "epoch": 0.85, + "grad_norm": 0.474609375, + "learning_rate": 0.00012379076007913335, + "loss": 2.5506, + "step": 19936 + }, + { + "epoch": 0.85, + "grad_norm": 0.4765625, + "learning_rate": 0.0001237842446936118, + "loss": 2.6291, + "step": 19937 + }, + { + "epoch": 0.85, + "grad_norm": 0.400390625, + "learning_rate": 0.00012377772920106988, + "loss": 2.6032, + "step": 19938 + }, + { + "epoch": 0.85, + "grad_norm": 0.458984375, + "learning_rate": 0.00012377121360153688, + "loss": 2.3441, + "step": 19939 + }, + { + "epoch": 0.85, + "grad_norm": 0.392578125, + "learning_rate": 0.0001237646978950421, + "loss": 2.5328, + "step": 19940 + }, + { + "epoch": 0.85, + "grad_norm": 0.423828125, + "learning_rate": 0.00012375818208161486, + "loss": 2.5278, + "step": 19941 + }, + { + "epoch": 0.85, + "grad_norm": 0.443359375, + "learning_rate": 0.00012375166616128446, + "loss": 2.4425, + "step": 19942 + }, + { + "epoch": 0.85, + "grad_norm": 0.447265625, + "learning_rate": 0.00012374515013408025, + "loss": 2.5656, + "step": 19943 + }, + { + "epoch": 0.85, + "grad_norm": 0.4140625, + "learning_rate": 0.0001237386340000316, + "loss": 2.486, + "step": 19944 + }, + { + "epoch": 0.85, + "grad_norm": 0.408203125, + "learning_rate": 0.0001237321177591677, + "loss": 2.5912, + "step": 19945 + }, + { + "epoch": 0.85, + "grad_norm": 0.431640625, + "learning_rate": 0.00012372560141151796, + "loss": 2.4471, + "step": 19946 + }, + { + "epoch": 0.85, + "grad_norm": 0.44140625, + "learning_rate": 0.00012371908495711168, + "loss": 2.6341, + "step": 19947 + }, + { + "epoch": 0.85, + "grad_norm": 0.4609375, + "learning_rate": 0.00012371256839597823, + "loss": 2.6469, + "step": 19948 + }, + { + "epoch": 0.85, + "grad_norm": 0.490234375, + "learning_rate": 0.00012370605172814684, + "loss": 2.4667, + "step": 19949 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.0001236995349536469, + "loss": 2.7334, + "step": 19950 + }, + { + "epoch": 0.85, + "grad_norm": 0.419921875, + "learning_rate": 0.0001236930180725077, + "loss": 2.4725, + "step": 19951 + }, + { + "epoch": 0.85, + "grad_norm": 0.45703125, + "learning_rate": 0.00012368650108475861, + "loss": 2.5511, + "step": 19952 + }, + { + "epoch": 0.85, + "grad_norm": 0.4921875, + "learning_rate": 0.00012367998399042887, + "loss": 2.6029, + "step": 19953 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.00012367346678954789, + "loss": 2.6725, + "step": 19954 + }, + { + "epoch": 0.85, + "grad_norm": 0.435546875, + "learning_rate": 0.00012366694948214496, + "loss": 2.5774, + "step": 19955 + }, + { + "epoch": 0.85, + "grad_norm": 0.48046875, + "learning_rate": 0.00012366043206824942, + "loss": 2.5713, + "step": 19956 + }, + { + "epoch": 0.85, + "grad_norm": 0.46484375, + "learning_rate": 0.00012365391454789056, + "loss": 2.6095, + "step": 19957 + }, + { + "epoch": 0.85, + "grad_norm": 0.51171875, + "learning_rate": 0.00012364739692109775, + "loss": 2.4922, + "step": 19958 + }, + { + "epoch": 0.85, + "grad_norm": 0.451171875, + "learning_rate": 0.00012364087918790027, + "loss": 2.6716, + "step": 19959 + }, + { + "epoch": 0.85, + "grad_norm": 0.439453125, + "learning_rate": 0.0001236343613483275, + "loss": 2.6842, + "step": 19960 + }, + { + "epoch": 0.85, + "grad_norm": 0.421875, + "learning_rate": 0.00012362784340240872, + "loss": 2.5527, + "step": 19961 + }, + { + "epoch": 0.85, + "grad_norm": 0.55078125, + "learning_rate": 0.0001236213253501733, + "loss": 2.6159, + "step": 19962 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.00012361480719165052, + "loss": 2.6151, + "step": 19963 + }, + { + "epoch": 0.85, + "grad_norm": 0.46875, + "learning_rate": 0.00012360828892686975, + "loss": 2.5531, + "step": 19964 + }, + { + "epoch": 0.85, + "grad_norm": 0.41015625, + "learning_rate": 0.00012360177055586031, + "loss": 2.4605, + "step": 19965 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.00012359525207865154, + "loss": 2.6096, + "step": 19966 + }, + { + "epoch": 0.85, + "grad_norm": 0.408203125, + "learning_rate": 0.00012358873349527274, + "loss": 2.5069, + "step": 19967 + }, + { + "epoch": 0.85, + "grad_norm": 0.4609375, + "learning_rate": 0.00012358221480575327, + "loss": 2.616, + "step": 19968 + }, + { + "epoch": 0.85, + "grad_norm": 0.4921875, + "learning_rate": 0.00012357569601012246, + "loss": 2.7246, + "step": 19969 + }, + { + "epoch": 0.85, + "grad_norm": 0.5390625, + "learning_rate": 0.0001235691771084096, + "loss": 2.6558, + "step": 19970 + }, + { + "epoch": 0.85, + "grad_norm": 0.47265625, + "learning_rate": 0.0001235626581006441, + "loss": 2.5095, + "step": 19971 + }, + { + "epoch": 0.85, + "grad_norm": 0.4765625, + "learning_rate": 0.00012355613898685524, + "loss": 2.5863, + "step": 19972 + }, + { + "epoch": 0.85, + "grad_norm": 0.435546875, + "learning_rate": 0.00012354961976707233, + "loss": 2.5006, + "step": 19973 + }, + { + "epoch": 0.85, + "grad_norm": 0.451171875, + "learning_rate": 0.00012354310044132477, + "loss": 2.5491, + "step": 19974 + }, + { + "epoch": 0.85, + "grad_norm": 0.431640625, + "learning_rate": 0.00012353658100964184, + "loss": 2.6902, + "step": 19975 + }, + { + "epoch": 0.85, + "grad_norm": 0.486328125, + "learning_rate": 0.00012353006147205293, + "loss": 2.6471, + "step": 19976 + }, + { + "epoch": 0.85, + "grad_norm": 0.42578125, + "learning_rate": 0.0001235235418285873, + "loss": 2.5213, + "step": 19977 + }, + { + "epoch": 0.85, + "grad_norm": 0.396484375, + "learning_rate": 0.00012351702207927435, + "loss": 2.5631, + "step": 19978 + }, + { + "epoch": 0.85, + "grad_norm": 0.466796875, + "learning_rate": 0.0001235105022241434, + "loss": 2.5028, + "step": 19979 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.00012350398226322376, + "loss": 2.6208, + "step": 19980 + }, + { + "epoch": 0.85, + "grad_norm": 0.404296875, + "learning_rate": 0.00012349746219654484, + "loss": 2.5226, + "step": 19981 + }, + { + "epoch": 0.85, + "grad_norm": 0.50390625, + "learning_rate": 0.0001234909420241359, + "loss": 2.5542, + "step": 19982 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.00012348442174602624, + "loss": 2.6842, + "step": 19983 + }, + { + "epoch": 0.85, + "grad_norm": 0.51171875, + "learning_rate": 0.00012347790136224538, + "loss": 2.5437, + "step": 19984 + }, + { + "epoch": 0.85, + "grad_norm": 0.44921875, + "learning_rate": 0.00012347138087282245, + "loss": 2.548, + "step": 19985 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.00012346486027778694, + "loss": 2.5584, + "step": 19986 + }, + { + "epoch": 0.85, + "grad_norm": 0.435546875, + "learning_rate": 0.00012345833957716812, + "loss": 2.514, + "step": 19987 + }, + { + "epoch": 0.85, + "grad_norm": 0.4765625, + "learning_rate": 0.00012345181877099535, + "loss": 2.6379, + "step": 19988 + }, + { + "epoch": 0.85, + "grad_norm": 0.48046875, + "learning_rate": 0.00012344529785929796, + "loss": 2.5858, + "step": 19989 + }, + { + "epoch": 0.85, + "grad_norm": 0.44921875, + "learning_rate": 0.00012343877684210528, + "loss": 2.5453, + "step": 19990 + }, + { + "epoch": 0.85, + "grad_norm": 0.443359375, + "learning_rate": 0.00012343225571944668, + "loss": 2.551, + "step": 19991 + }, + { + "epoch": 0.85, + "grad_norm": 0.4453125, + "learning_rate": 0.00012342573449135153, + "loss": 2.6544, + "step": 19992 + }, + { + "epoch": 0.85, + "grad_norm": 0.455078125, + "learning_rate": 0.00012341921315784908, + "loss": 2.6084, + "step": 19993 + }, + { + "epoch": 0.85, + "grad_norm": 0.421875, + "learning_rate": 0.00012341269171896877, + "loss": 2.5464, + "step": 19994 + }, + { + "epoch": 0.85, + "grad_norm": 0.45703125, + "learning_rate": 0.00012340617017473986, + "loss": 2.6075, + "step": 19995 + }, + { + "epoch": 0.85, + "grad_norm": 0.453125, + "learning_rate": 0.0001233996485251918, + "loss": 2.4933, + "step": 19996 + }, + { + "epoch": 0.85, + "grad_norm": 0.447265625, + "learning_rate": 0.00012339312677035383, + "loss": 2.5222, + "step": 19997 + }, + { + "epoch": 0.85, + "grad_norm": 0.447265625, + "learning_rate": 0.00012338660491025536, + "loss": 2.5442, + "step": 19998 + }, + { + "epoch": 0.85, + "grad_norm": 0.4921875, + "learning_rate": 0.00012338008294492572, + "loss": 2.4976, + "step": 19999 + }, + { + "epoch": 0.85, + "grad_norm": 0.4765625, + "learning_rate": 0.0001233735608743942, + "loss": 2.4765, + "step": 20000 + }, + { + "epoch": 0.85, + "grad_norm": 0.42578125, + "learning_rate": 0.00012336703869869026, + "loss": 2.71, + "step": 20001 + }, + { + "epoch": 0.85, + "grad_norm": 0.431640625, + "learning_rate": 0.00012336051641784315, + "loss": 2.5529, + "step": 20002 + }, + { + "epoch": 0.85, + "grad_norm": 0.53125, + "learning_rate": 0.0001233539940318823, + "loss": 2.6175, + "step": 20003 + }, + { + "epoch": 0.85, + "grad_norm": 0.4453125, + "learning_rate": 0.00012334747154083692, + "loss": 2.618, + "step": 20004 + }, + { + "epoch": 0.85, + "grad_norm": 0.482421875, + "learning_rate": 0.0001233409489447365, + "loss": 2.595, + "step": 20005 + }, + { + "epoch": 0.85, + "grad_norm": 0.478515625, + "learning_rate": 0.00012333442624361035, + "loss": 2.6032, + "step": 20006 + }, + { + "epoch": 0.85, + "grad_norm": 0.4453125, + "learning_rate": 0.0001233279034374878, + "loss": 2.5093, + "step": 20007 + }, + { + "epoch": 0.85, + "grad_norm": 0.482421875, + "learning_rate": 0.00012332138052639823, + "loss": 2.5121, + "step": 20008 + }, + { + "epoch": 0.85, + "grad_norm": 0.435546875, + "learning_rate": 0.00012331485751037094, + "loss": 2.5293, + "step": 20009 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.0001233083343894353, + "loss": 2.6633, + "step": 20010 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.00012330181116362073, + "loss": 2.6074, + "step": 20011 + }, + { + "epoch": 0.85, + "grad_norm": 0.443359375, + "learning_rate": 0.0001232952878329565, + "loss": 2.617, + "step": 20012 + }, + { + "epoch": 0.85, + "grad_norm": 0.451171875, + "learning_rate": 0.00012328876439747196, + "loss": 2.4361, + "step": 20013 + }, + { + "epoch": 0.85, + "grad_norm": 0.46484375, + "learning_rate": 0.00012328224085719652, + "loss": 2.6391, + "step": 20014 + }, + { + "epoch": 0.85, + "grad_norm": 0.453125, + "learning_rate": 0.0001232757172121595, + "loss": 2.4754, + "step": 20015 + }, + { + "epoch": 0.85, + "grad_norm": 0.470703125, + "learning_rate": 0.00012326919346239027, + "loss": 2.6515, + "step": 20016 + }, + { + "epoch": 0.85, + "grad_norm": 0.470703125, + "learning_rate": 0.00012326266960791817, + "loss": 2.5419, + "step": 20017 + }, + { + "epoch": 0.85, + "grad_norm": 0.416015625, + "learning_rate": 0.00012325614564877253, + "loss": 2.5814, + "step": 20018 + }, + { + "epoch": 0.85, + "grad_norm": 0.412109375, + "learning_rate": 0.00012324962158498278, + "loss": 2.369, + "step": 20019 + }, + { + "epoch": 0.85, + "grad_norm": 0.5078125, + "learning_rate": 0.00012324309741657818, + "loss": 2.6167, + "step": 20020 + }, + { + "epoch": 0.85, + "grad_norm": 0.490234375, + "learning_rate": 0.00012323657314358816, + "loss": 2.4687, + "step": 20021 + }, + { + "epoch": 0.85, + "grad_norm": 0.41015625, + "learning_rate": 0.00012323004876604204, + "loss": 2.4423, + "step": 20022 + }, + { + "epoch": 0.85, + "grad_norm": 0.376953125, + "learning_rate": 0.0001232235242839692, + "loss": 2.5403, + "step": 20023 + }, + { + "epoch": 0.85, + "grad_norm": 0.40625, + "learning_rate": 0.00012321699969739902, + "loss": 2.5776, + "step": 20024 + }, + { + "epoch": 0.85, + "grad_norm": 0.484375, + "learning_rate": 0.0001232104750063608, + "loss": 2.6113, + "step": 20025 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.0001232039502108839, + "loss": 2.6198, + "step": 20026 + }, + { + "epoch": 0.85, + "grad_norm": 0.427734375, + "learning_rate": 0.00012319742531099776, + "loss": 2.5938, + "step": 20027 + }, + { + "epoch": 0.85, + "grad_norm": 0.458984375, + "learning_rate": 0.0001231909003067316, + "loss": 2.5892, + "step": 20028 + }, + { + "epoch": 0.85, + "grad_norm": 0.419921875, + "learning_rate": 0.00012318437519811493, + "loss": 2.3903, + "step": 20029 + }, + { + "epoch": 0.85, + "grad_norm": 0.5, + "learning_rate": 0.00012317784998517703, + "loss": 2.5699, + "step": 20030 + }, + { + "epoch": 0.85, + "grad_norm": 0.482421875, + "learning_rate": 0.00012317132466794727, + "loss": 2.4428, + "step": 20031 + }, + { + "epoch": 0.85, + "grad_norm": 0.447265625, + "learning_rate": 0.00012316479924645503, + "loss": 2.5325, + "step": 20032 + }, + { + "epoch": 0.85, + "grad_norm": 0.486328125, + "learning_rate": 0.00012315827372072965, + "loss": 2.6342, + "step": 20033 + }, + { + "epoch": 0.85, + "grad_norm": 0.478515625, + "learning_rate": 0.0001231517480908005, + "loss": 2.4896, + "step": 20034 + }, + { + "epoch": 0.85, + "grad_norm": 0.466796875, + "learning_rate": 0.00012314522235669695, + "loss": 2.5832, + "step": 20035 + }, + { + "epoch": 0.85, + "grad_norm": 0.484375, + "learning_rate": 0.00012313869651844834, + "loss": 2.4962, + "step": 20036 + }, + { + "epoch": 0.85, + "grad_norm": 0.4375, + "learning_rate": 0.00012313217057608407, + "loss": 2.4782, + "step": 20037 + }, + { + "epoch": 0.85, + "grad_norm": 0.466796875, + "learning_rate": 0.0001231256445296335, + "loss": 2.6125, + "step": 20038 + }, + { + "epoch": 0.85, + "grad_norm": 0.431640625, + "learning_rate": 0.00012311911837912596, + "loss": 2.5472, + "step": 20039 + }, + { + "epoch": 0.85, + "grad_norm": 0.421875, + "learning_rate": 0.00012311259212459082, + "loss": 2.5894, + "step": 20040 + }, + { + "epoch": 0.85, + "grad_norm": 0.458984375, + "learning_rate": 0.0001231060657660575, + "loss": 2.4963, + "step": 20041 + }, + { + "epoch": 0.85, + "grad_norm": 0.447265625, + "learning_rate": 0.00012309953930355529, + "loss": 2.5809, + "step": 20042 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.00012309301273711365, + "loss": 2.5481, + "step": 20043 + }, + { + "epoch": 0.85, + "grad_norm": 0.44140625, + "learning_rate": 0.00012308648606676183, + "loss": 2.5406, + "step": 20044 + }, + { + "epoch": 0.85, + "grad_norm": 0.4296875, + "learning_rate": 0.0001230799592925293, + "loss": 2.4004, + "step": 20045 + }, + { + "epoch": 0.85, + "grad_norm": 0.46875, + "learning_rate": 0.00012307343241444536, + "loss": 2.625, + "step": 20046 + }, + { + "epoch": 0.85, + "grad_norm": 0.50390625, + "learning_rate": 0.00012306690543253943, + "loss": 2.6462, + "step": 20047 + }, + { + "epoch": 0.85, + "grad_norm": 0.443359375, + "learning_rate": 0.00012306037834684084, + "loss": 2.5016, + "step": 20048 + }, + { + "epoch": 0.85, + "grad_norm": 0.41015625, + "learning_rate": 0.000123053851157379, + "loss": 2.528, + "step": 20049 + }, + { + "epoch": 0.85, + "grad_norm": 0.4140625, + "learning_rate": 0.0001230473238641832, + "loss": 2.673, + "step": 20050 + }, + { + "epoch": 0.85, + "grad_norm": 0.484375, + "learning_rate": 0.00012304079646728293, + "loss": 2.5289, + "step": 20051 + }, + { + "epoch": 0.85, + "grad_norm": 0.494140625, + "learning_rate": 0.00012303426896670744, + "loss": 2.5273, + "step": 20052 + }, + { + "epoch": 0.85, + "grad_norm": 0.46875, + "learning_rate": 0.0001230277413624862, + "loss": 2.592, + "step": 20053 + }, + { + "epoch": 0.85, + "grad_norm": 0.462890625, + "learning_rate": 0.00012302121365464854, + "loss": 2.6814, + "step": 20054 + }, + { + "epoch": 0.85, + "grad_norm": 0.45703125, + "learning_rate": 0.00012301468584322383, + "loss": 2.4931, + "step": 20055 + }, + { + "epoch": 0.85, + "grad_norm": 0.439453125, + "learning_rate": 0.00012300815792824142, + "loss": 2.6123, + "step": 20056 + }, + { + "epoch": 0.85, + "grad_norm": 0.443359375, + "learning_rate": 0.0001230016299097307, + "loss": 2.7099, + "step": 20057 + }, + { + "epoch": 0.85, + "grad_norm": 0.46875, + "learning_rate": 0.00012299510178772106, + "loss": 2.7322, + "step": 20058 + }, + { + "epoch": 0.85, + "grad_norm": 0.408203125, + "learning_rate": 0.00012298857356224193, + "loss": 2.6285, + "step": 20059 + }, + { + "epoch": 0.85, + "grad_norm": 0.408203125, + "learning_rate": 0.00012298204523332254, + "loss": 2.5963, + "step": 20060 + }, + { + "epoch": 0.85, + "grad_norm": 0.44921875, + "learning_rate": 0.00012297551680099238, + "loss": 2.5334, + "step": 20061 + }, + { + "epoch": 0.85, + "grad_norm": 0.435546875, + "learning_rate": 0.00012296898826528077, + "loss": 2.4759, + "step": 20062 + }, + { + "epoch": 0.85, + "grad_norm": 0.447265625, + "learning_rate": 0.00012296245962621714, + "loss": 2.4808, + "step": 20063 + }, + { + "epoch": 0.85, + "grad_norm": 0.416015625, + "learning_rate": 0.0001229559308838308, + "loss": 2.5436, + "step": 20064 + }, + { + "epoch": 0.86, + "grad_norm": 0.4609375, + "learning_rate": 0.00012294940203815116, + "loss": 2.5468, + "step": 20065 + }, + { + "epoch": 0.86, + "grad_norm": 0.45703125, + "learning_rate": 0.0001229428730892076, + "loss": 2.5426, + "step": 20066 + }, + { + "epoch": 0.86, + "grad_norm": 0.470703125, + "learning_rate": 0.00012293634403702955, + "loss": 2.5539, + "step": 20067 + }, + { + "epoch": 0.86, + "grad_norm": 0.5234375, + "learning_rate": 0.00012292981488164627, + "loss": 2.4703, + "step": 20068 + }, + { + "epoch": 0.86, + "grad_norm": 0.49609375, + "learning_rate": 0.00012292328562308724, + "loss": 2.4294, + "step": 20069 + }, + { + "epoch": 0.86, + "grad_norm": 0.4765625, + "learning_rate": 0.00012291675626138177, + "loss": 2.6838, + "step": 20070 + }, + { + "epoch": 0.86, + "grad_norm": 0.47265625, + "learning_rate": 0.0001229102267965593, + "loss": 2.4993, + "step": 20071 + }, + { + "epoch": 0.86, + "grad_norm": 0.5078125, + "learning_rate": 0.0001229036972286492, + "loss": 2.593, + "step": 20072 + }, + { + "epoch": 0.86, + "grad_norm": 0.49609375, + "learning_rate": 0.0001228971675576808, + "loss": 2.5829, + "step": 20073 + }, + { + "epoch": 0.86, + "grad_norm": 0.49609375, + "learning_rate": 0.0001228906377836835, + "loss": 2.5137, + "step": 20074 + }, + { + "epoch": 0.86, + "grad_norm": 0.44140625, + "learning_rate": 0.00012288410790668673, + "loss": 2.5134, + "step": 20075 + }, + { + "epoch": 0.86, + "grad_norm": 0.470703125, + "learning_rate": 0.0001228775779267198, + "loss": 2.5672, + "step": 20076 + }, + { + "epoch": 0.86, + "grad_norm": 0.439453125, + "learning_rate": 0.00012287104784381215, + "loss": 2.5218, + "step": 20077 + }, + { + "epoch": 0.86, + "grad_norm": 0.4921875, + "learning_rate": 0.00012286451765799314, + "loss": 2.5422, + "step": 20078 + }, + { + "epoch": 0.86, + "grad_norm": 0.41796875, + "learning_rate": 0.00012285798736929218, + "loss": 2.5256, + "step": 20079 + }, + { + "epoch": 0.86, + "grad_norm": 0.42578125, + "learning_rate": 0.0001228514569777386, + "loss": 2.512, + "step": 20080 + }, + { + "epoch": 0.86, + "grad_norm": 0.404296875, + "learning_rate": 0.00012284492648336184, + "loss": 2.528, + "step": 20081 + }, + { + "epoch": 0.86, + "grad_norm": 0.466796875, + "learning_rate": 0.0001228383958861912, + "loss": 2.5055, + "step": 20082 + }, + { + "epoch": 0.86, + "grad_norm": 0.408203125, + "learning_rate": 0.00012283186518625622, + "loss": 2.5085, + "step": 20083 + }, + { + "epoch": 0.86, + "grad_norm": 0.40625, + "learning_rate": 0.00012282533438358613, + "loss": 2.4073, + "step": 20084 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.00012281880347821039, + "loss": 2.7528, + "step": 20085 + }, + { + "epoch": 0.86, + "grad_norm": 0.4765625, + "learning_rate": 0.0001228122724701584, + "loss": 2.7183, + "step": 20086 + }, + { + "epoch": 0.86, + "grad_norm": 0.392578125, + "learning_rate": 0.00012280574135945947, + "loss": 2.5858, + "step": 20087 + }, + { + "epoch": 0.86, + "grad_norm": 0.41015625, + "learning_rate": 0.00012279921014614307, + "loss": 2.6197, + "step": 20088 + }, + { + "epoch": 0.86, + "grad_norm": 0.427734375, + "learning_rate": 0.00012279267883023856, + "loss": 2.5426, + "step": 20089 + }, + { + "epoch": 0.86, + "grad_norm": 0.5234375, + "learning_rate": 0.00012278614741177528, + "loss": 2.6316, + "step": 20090 + }, + { + "epoch": 0.86, + "grad_norm": 0.408203125, + "learning_rate": 0.00012277961589078273, + "loss": 2.5533, + "step": 20091 + }, + { + "epoch": 0.86, + "grad_norm": 0.5, + "learning_rate": 0.00012277308426729018, + "loss": 2.479, + "step": 20092 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.0001227665525413271, + "loss": 2.5183, + "step": 20093 + }, + { + "epoch": 0.86, + "grad_norm": 0.42578125, + "learning_rate": 0.00012276002071292286, + "loss": 2.6206, + "step": 20094 + }, + { + "epoch": 0.86, + "grad_norm": 0.423828125, + "learning_rate": 0.00012275348878210683, + "loss": 2.4815, + "step": 20095 + }, + { + "epoch": 0.86, + "grad_norm": 0.4921875, + "learning_rate": 0.00012274695674890842, + "loss": 2.5652, + "step": 20096 + }, + { + "epoch": 0.86, + "grad_norm": 0.5, + "learning_rate": 0.00012274042461335702, + "loss": 2.6301, + "step": 20097 + }, + { + "epoch": 0.86, + "grad_norm": 0.439453125, + "learning_rate": 0.00012273389237548197, + "loss": 2.458, + "step": 20098 + }, + { + "epoch": 0.86, + "grad_norm": 0.50390625, + "learning_rate": 0.0001227273600353128, + "loss": 2.6854, + "step": 20099 + }, + { + "epoch": 0.86, + "grad_norm": 0.48828125, + "learning_rate": 0.00012272082759287874, + "loss": 2.5769, + "step": 20100 + }, + { + "epoch": 0.86, + "grad_norm": 0.44921875, + "learning_rate": 0.00012271429504820934, + "loss": 2.4682, + "step": 20101 + }, + { + "epoch": 0.86, + "grad_norm": 0.400390625, + "learning_rate": 0.00012270776240133385, + "loss": 2.5531, + "step": 20102 + }, + { + "epoch": 0.86, + "grad_norm": 0.515625, + "learning_rate": 0.00012270122965228175, + "loss": 2.5966, + "step": 20103 + }, + { + "epoch": 0.86, + "grad_norm": 0.4296875, + "learning_rate": 0.0001226946968010824, + "loss": 2.4988, + "step": 20104 + }, + { + "epoch": 0.86, + "grad_norm": 0.47265625, + "learning_rate": 0.00012268816384776522, + "loss": 2.5466, + "step": 20105 + }, + { + "epoch": 0.86, + "grad_norm": 0.47265625, + "learning_rate": 0.0001226816307923596, + "loss": 2.583, + "step": 20106 + }, + { + "epoch": 0.86, + "grad_norm": 0.42578125, + "learning_rate": 0.00012267509763489492, + "loss": 2.6381, + "step": 20107 + }, + { + "epoch": 0.86, + "grad_norm": 0.466796875, + "learning_rate": 0.00012266856437540057, + "loss": 2.5619, + "step": 20108 + }, + { + "epoch": 0.86, + "grad_norm": 0.42578125, + "learning_rate": 0.000122662031013906, + "loss": 2.6298, + "step": 20109 + }, + { + "epoch": 0.86, + "grad_norm": 0.486328125, + "learning_rate": 0.00012265549755044054, + "loss": 2.5675, + "step": 20110 + }, + { + "epoch": 0.86, + "grad_norm": 0.458984375, + "learning_rate": 0.00012264896398503364, + "loss": 2.5945, + "step": 20111 + }, + { + "epoch": 0.86, + "grad_norm": 0.455078125, + "learning_rate": 0.00012264243031771468, + "loss": 2.5193, + "step": 20112 + }, + { + "epoch": 0.86, + "grad_norm": 0.4609375, + "learning_rate": 0.0001226358965485131, + "loss": 2.4127, + "step": 20113 + }, + { + "epoch": 0.86, + "grad_norm": 0.44140625, + "learning_rate": 0.00012262936267745818, + "loss": 2.6345, + "step": 20114 + }, + { + "epoch": 0.86, + "grad_norm": 0.44921875, + "learning_rate": 0.00012262282870457945, + "loss": 2.507, + "step": 20115 + }, + { + "epoch": 0.86, + "grad_norm": 0.486328125, + "learning_rate": 0.00012261629462990623, + "loss": 2.493, + "step": 20116 + }, + { + "epoch": 0.86, + "grad_norm": 0.482421875, + "learning_rate": 0.00012260976045346794, + "loss": 2.5805, + "step": 20117 + }, + { + "epoch": 0.86, + "grad_norm": 0.474609375, + "learning_rate": 0.00012260322617529403, + "loss": 2.6339, + "step": 20118 + }, + { + "epoch": 0.86, + "grad_norm": 0.44921875, + "learning_rate": 0.00012259669179541383, + "loss": 2.575, + "step": 20119 + }, + { + "epoch": 0.86, + "grad_norm": 0.4140625, + "learning_rate": 0.0001225901573138568, + "loss": 2.4535, + "step": 20120 + }, + { + "epoch": 0.86, + "grad_norm": 0.453125, + "learning_rate": 0.00012258362273065228, + "loss": 2.5056, + "step": 20121 + }, + { + "epoch": 0.86, + "grad_norm": 0.458984375, + "learning_rate": 0.00012257708804582976, + "loss": 2.5422, + "step": 20122 + }, + { + "epoch": 0.86, + "grad_norm": 0.49609375, + "learning_rate": 0.00012257055325941857, + "loss": 2.6579, + "step": 20123 + }, + { + "epoch": 0.86, + "grad_norm": 0.453125, + "learning_rate": 0.00012256401837144816, + "loss": 2.5386, + "step": 20124 + }, + { + "epoch": 0.86, + "grad_norm": 0.4921875, + "learning_rate": 0.00012255748338194786, + "loss": 2.4801, + "step": 20125 + }, + { + "epoch": 0.86, + "grad_norm": 0.49609375, + "learning_rate": 0.0001225509482909472, + "loss": 2.5154, + "step": 20126 + }, + { + "epoch": 0.86, + "grad_norm": 0.49609375, + "learning_rate": 0.00012254441309847544, + "loss": 2.5511, + "step": 20127 + }, + { + "epoch": 0.86, + "grad_norm": 0.421875, + "learning_rate": 0.0001225378778045621, + "loss": 2.5812, + "step": 20128 + }, + { + "epoch": 0.86, + "grad_norm": 0.46875, + "learning_rate": 0.00012253134240923657, + "loss": 2.5404, + "step": 20129 + }, + { + "epoch": 0.86, + "grad_norm": 0.55859375, + "learning_rate": 0.0001225248069125282, + "loss": 2.5382, + "step": 20130 + }, + { + "epoch": 0.86, + "grad_norm": 0.48046875, + "learning_rate": 0.00012251827131446643, + "loss": 2.504, + "step": 20131 + }, + { + "epoch": 0.86, + "grad_norm": 0.474609375, + "learning_rate": 0.00012251173561508066, + "loss": 2.5835, + "step": 20132 + }, + { + "epoch": 0.86, + "grad_norm": 0.48046875, + "learning_rate": 0.00012250519981440032, + "loss": 2.6047, + "step": 20133 + }, + { + "epoch": 0.86, + "grad_norm": 0.478515625, + "learning_rate": 0.00012249866391245484, + "loss": 2.5604, + "step": 20134 + }, + { + "epoch": 0.86, + "grad_norm": 0.42578125, + "learning_rate": 0.00012249212790927354, + "loss": 2.4257, + "step": 20135 + }, + { + "epoch": 0.86, + "grad_norm": 0.45703125, + "learning_rate": 0.0001224855918048859, + "loss": 2.5388, + "step": 20136 + }, + { + "epoch": 0.86, + "grad_norm": 0.439453125, + "learning_rate": 0.00012247905559932133, + "loss": 2.5638, + "step": 20137 + }, + { + "epoch": 0.86, + "grad_norm": 0.43359375, + "learning_rate": 0.00012247251929260923, + "loss": 2.402, + "step": 20138 + }, + { + "epoch": 0.86, + "grad_norm": 0.53125, + "learning_rate": 0.000122465982884779, + "loss": 2.6001, + "step": 20139 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.00012245944637586005, + "loss": 2.5022, + "step": 20140 + }, + { + "epoch": 0.86, + "grad_norm": 0.46875, + "learning_rate": 0.00012245290976588177, + "loss": 2.6165, + "step": 20141 + }, + { + "epoch": 0.86, + "grad_norm": 0.4375, + "learning_rate": 0.00012244637305487365, + "loss": 2.4633, + "step": 20142 + }, + { + "epoch": 0.86, + "grad_norm": 0.453125, + "learning_rate": 0.00012243983624286503, + "loss": 2.3018, + "step": 20143 + }, + { + "epoch": 0.86, + "grad_norm": 0.546875, + "learning_rate": 0.00012243329932988534, + "loss": 2.6812, + "step": 20144 + }, + { + "epoch": 0.86, + "grad_norm": 0.455078125, + "learning_rate": 0.00012242676231596404, + "loss": 2.5289, + "step": 20145 + }, + { + "epoch": 0.86, + "grad_norm": 0.46484375, + "learning_rate": 0.00012242022520113049, + "loss": 2.4784, + "step": 20146 + }, + { + "epoch": 0.86, + "grad_norm": 0.5, + "learning_rate": 0.00012241368798541412, + "loss": 2.6601, + "step": 20147 + }, + { + "epoch": 0.86, + "grad_norm": 0.451171875, + "learning_rate": 0.0001224071506688443, + "loss": 2.5653, + "step": 20148 + }, + { + "epoch": 0.86, + "grad_norm": 0.482421875, + "learning_rate": 0.00012240061325145054, + "loss": 2.6391, + "step": 20149 + }, + { + "epoch": 0.86, + "grad_norm": 0.46484375, + "learning_rate": 0.0001223940757332622, + "loss": 2.5394, + "step": 20150 + }, + { + "epoch": 0.86, + "grad_norm": 0.427734375, + "learning_rate": 0.00012238753811430868, + "loss": 2.5929, + "step": 20151 + }, + { + "epoch": 0.86, + "grad_norm": 0.462890625, + "learning_rate": 0.00012238100039461944, + "loss": 2.5127, + "step": 20152 + }, + { + "epoch": 0.86, + "grad_norm": 0.421875, + "learning_rate": 0.00012237446257422387, + "loss": 2.4722, + "step": 20153 + }, + { + "epoch": 0.86, + "grad_norm": 0.4296875, + "learning_rate": 0.0001223679246531514, + "loss": 2.3815, + "step": 20154 + }, + { + "epoch": 0.86, + "grad_norm": 0.462890625, + "learning_rate": 0.00012236138663143146, + "loss": 2.3787, + "step": 20155 + }, + { + "epoch": 0.86, + "grad_norm": 0.41796875, + "learning_rate": 0.00012235484850909345, + "loss": 2.5085, + "step": 20156 + }, + { + "epoch": 0.86, + "grad_norm": 0.431640625, + "learning_rate": 0.00012234831028616676, + "loss": 2.593, + "step": 20157 + }, + { + "epoch": 0.86, + "grad_norm": 0.4375, + "learning_rate": 0.00012234177196268086, + "loss": 2.4507, + "step": 20158 + }, + { + "epoch": 0.86, + "grad_norm": 0.453125, + "learning_rate": 0.00012233523353866513, + "loss": 2.5443, + "step": 20159 + }, + { + "epoch": 0.86, + "grad_norm": 0.447265625, + "learning_rate": 0.00012232869501414904, + "loss": 2.6063, + "step": 20160 + }, + { + "epoch": 0.86, + "grad_norm": 0.4765625, + "learning_rate": 0.00012232215638916197, + "loss": 2.5916, + "step": 20161 + }, + { + "epoch": 0.86, + "grad_norm": 0.41015625, + "learning_rate": 0.00012231561766373335, + "loss": 2.5351, + "step": 20162 + }, + { + "epoch": 0.86, + "grad_norm": 0.38671875, + "learning_rate": 0.00012230907883789262, + "loss": 2.4854, + "step": 20163 + }, + { + "epoch": 0.86, + "grad_norm": 0.4453125, + "learning_rate": 0.0001223025399116692, + "loss": 2.4805, + "step": 20164 + }, + { + "epoch": 0.86, + "grad_norm": 0.50390625, + "learning_rate": 0.00012229600088509246, + "loss": 2.585, + "step": 20165 + }, + { + "epoch": 0.86, + "grad_norm": 0.498046875, + "learning_rate": 0.0001222894617581919, + "loss": 2.6108, + "step": 20166 + }, + { + "epoch": 0.86, + "grad_norm": 0.51171875, + "learning_rate": 0.00012228292253099689, + "loss": 2.6338, + "step": 20167 + }, + { + "epoch": 0.86, + "grad_norm": 0.51953125, + "learning_rate": 0.0001222763832035369, + "loss": 2.7483, + "step": 20168 + }, + { + "epoch": 0.86, + "grad_norm": 0.453125, + "learning_rate": 0.0001222698437758413, + "loss": 2.6601, + "step": 20169 + }, + { + "epoch": 0.86, + "grad_norm": 0.482421875, + "learning_rate": 0.00012226330424793954, + "loss": 2.5957, + "step": 20170 + }, + { + "epoch": 0.86, + "grad_norm": 0.43359375, + "learning_rate": 0.00012225676461986106, + "loss": 2.6024, + "step": 20171 + }, + { + "epoch": 0.86, + "grad_norm": 0.458984375, + "learning_rate": 0.00012225022489163528, + "loss": 2.572, + "step": 20172 + }, + { + "epoch": 0.86, + "grad_norm": 0.48046875, + "learning_rate": 0.0001222436850632916, + "loss": 2.6945, + "step": 20173 + }, + { + "epoch": 0.86, + "grad_norm": 0.45703125, + "learning_rate": 0.00012223714513485945, + "loss": 2.3944, + "step": 20174 + }, + { + "epoch": 0.86, + "grad_norm": 0.43359375, + "learning_rate": 0.0001222306051063683, + "loss": 2.56, + "step": 20175 + }, + { + "epoch": 0.86, + "grad_norm": 0.408203125, + "learning_rate": 0.00012222406497784755, + "loss": 2.4706, + "step": 20176 + }, + { + "epoch": 0.86, + "grad_norm": 0.466796875, + "learning_rate": 0.0001222175247493266, + "loss": 2.5093, + "step": 20177 + }, + { + "epoch": 0.86, + "grad_norm": 0.458984375, + "learning_rate": 0.00012221098442083495, + "loss": 2.4356, + "step": 20178 + }, + { + "epoch": 0.86, + "grad_norm": 0.515625, + "learning_rate": 0.00012220444399240195, + "loss": 2.7216, + "step": 20179 + }, + { + "epoch": 0.86, + "grad_norm": 0.52734375, + "learning_rate": 0.00012219790346405712, + "loss": 2.6865, + "step": 20180 + }, + { + "epoch": 0.86, + "grad_norm": 0.490234375, + "learning_rate": 0.0001221913628358298, + "loss": 2.5504, + "step": 20181 + }, + { + "epoch": 0.86, + "grad_norm": 0.4609375, + "learning_rate": 0.00012218482210774948, + "loss": 2.5348, + "step": 20182 + }, + { + "epoch": 0.86, + "grad_norm": 0.51953125, + "learning_rate": 0.00012217828127984555, + "loss": 2.5749, + "step": 20183 + }, + { + "epoch": 0.86, + "grad_norm": 0.51953125, + "learning_rate": 0.00012217174035214745, + "loss": 2.6811, + "step": 20184 + }, + { + "epoch": 0.86, + "grad_norm": 0.435546875, + "learning_rate": 0.00012216519932468464, + "loss": 2.5481, + "step": 20185 + }, + { + "epoch": 0.86, + "grad_norm": 0.470703125, + "learning_rate": 0.00012215865819748654, + "loss": 2.6469, + "step": 20186 + }, + { + "epoch": 0.86, + "grad_norm": 0.4765625, + "learning_rate": 0.00012215211697058252, + "loss": 2.5004, + "step": 20187 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.00012214557564400215, + "loss": 2.5651, + "step": 20188 + }, + { + "epoch": 0.86, + "grad_norm": 0.484375, + "learning_rate": 0.0001221390342177747, + "loss": 2.7531, + "step": 20189 + }, + { + "epoch": 0.86, + "grad_norm": 0.4609375, + "learning_rate": 0.00012213249269192974, + "loss": 2.5132, + "step": 20190 + }, + { + "epoch": 0.86, + "grad_norm": 0.462890625, + "learning_rate": 0.00012212595106649664, + "loss": 2.5737, + "step": 20191 + }, + { + "epoch": 0.86, + "grad_norm": 0.447265625, + "learning_rate": 0.00012211940934150488, + "loss": 2.5011, + "step": 20192 + }, + { + "epoch": 0.86, + "grad_norm": 0.44921875, + "learning_rate": 0.0001221128675169838, + "loss": 2.5471, + "step": 20193 + }, + { + "epoch": 0.86, + "grad_norm": 0.462890625, + "learning_rate": 0.00012210632559296296, + "loss": 2.537, + "step": 20194 + }, + { + "epoch": 0.86, + "grad_norm": 0.46484375, + "learning_rate": 0.00012209978356947167, + "loss": 2.5305, + "step": 20195 + }, + { + "epoch": 0.86, + "grad_norm": 0.451171875, + "learning_rate": 0.00012209324144653948, + "loss": 2.6225, + "step": 20196 + }, + { + "epoch": 0.86, + "grad_norm": 0.451171875, + "learning_rate": 0.00012208669922419575, + "loss": 2.5342, + "step": 20197 + }, + { + "epoch": 0.86, + "grad_norm": 0.41015625, + "learning_rate": 0.00012208015690246996, + "loss": 2.5733, + "step": 20198 + }, + { + "epoch": 0.86, + "grad_norm": 0.490234375, + "learning_rate": 0.00012207361448139152, + "loss": 2.6923, + "step": 20199 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.0001220670719609899, + "loss": 2.6557, + "step": 20200 + }, + { + "epoch": 0.86, + "grad_norm": 0.486328125, + "learning_rate": 0.0001220605293412945, + "loss": 2.6271, + "step": 20201 + }, + { + "epoch": 0.86, + "grad_norm": 0.455078125, + "learning_rate": 0.00012205398662233481, + "loss": 2.5368, + "step": 20202 + }, + { + "epoch": 0.86, + "grad_norm": 0.4375, + "learning_rate": 0.0001220474438041402, + "loss": 2.6853, + "step": 20203 + }, + { + "epoch": 0.86, + "grad_norm": 0.46484375, + "learning_rate": 0.00012204090088674021, + "loss": 2.5751, + "step": 20204 + }, + { + "epoch": 0.86, + "grad_norm": 0.421875, + "learning_rate": 0.00012203435787016416, + "loss": 2.3583, + "step": 20205 + }, + { + "epoch": 0.86, + "grad_norm": 0.44140625, + "learning_rate": 0.00012202781475444156, + "loss": 2.4886, + "step": 20206 + }, + { + "epoch": 0.86, + "grad_norm": 0.46875, + "learning_rate": 0.00012202127153960186, + "loss": 2.6015, + "step": 20207 + }, + { + "epoch": 0.86, + "grad_norm": 0.447265625, + "learning_rate": 0.0001220147282256745, + "loss": 2.4666, + "step": 20208 + }, + { + "epoch": 0.86, + "grad_norm": 0.490234375, + "learning_rate": 0.0001220081848126889, + "loss": 2.6743, + "step": 20209 + }, + { + "epoch": 0.86, + "grad_norm": 0.4375, + "learning_rate": 0.00012200164130067451, + "loss": 2.6542, + "step": 20210 + }, + { + "epoch": 0.86, + "grad_norm": 0.484375, + "learning_rate": 0.00012199509768966074, + "loss": 2.548, + "step": 20211 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.00012198855397967713, + "loss": 2.5819, + "step": 20212 + }, + { + "epoch": 0.86, + "grad_norm": 0.462890625, + "learning_rate": 0.00012198201017075299, + "loss": 2.4082, + "step": 20213 + }, + { + "epoch": 0.86, + "grad_norm": 0.5546875, + "learning_rate": 0.0001219754662629179, + "loss": 2.6515, + "step": 20214 + }, + { + "epoch": 0.86, + "grad_norm": 0.4375, + "learning_rate": 0.00012196892225620123, + "loss": 2.4885, + "step": 20215 + }, + { + "epoch": 0.86, + "grad_norm": 0.5, + "learning_rate": 0.00012196237815063243, + "loss": 2.5408, + "step": 20216 + }, + { + "epoch": 0.86, + "grad_norm": 0.451171875, + "learning_rate": 0.00012195583394624096, + "loss": 2.6231, + "step": 20217 + }, + { + "epoch": 0.86, + "grad_norm": 0.44140625, + "learning_rate": 0.00012194928964305625, + "loss": 2.5741, + "step": 20218 + }, + { + "epoch": 0.86, + "grad_norm": 0.4453125, + "learning_rate": 0.00012194274524110774, + "loss": 2.4959, + "step": 20219 + }, + { + "epoch": 0.86, + "grad_norm": 0.46484375, + "learning_rate": 0.00012193620074042495, + "loss": 2.5799, + "step": 20220 + }, + { + "epoch": 0.86, + "grad_norm": 0.5078125, + "learning_rate": 0.00012192965614103722, + "loss": 2.9021, + "step": 20221 + }, + { + "epoch": 0.86, + "grad_norm": 0.47265625, + "learning_rate": 0.00012192311144297408, + "loss": 2.3983, + "step": 20222 + }, + { + "epoch": 0.86, + "grad_norm": 0.4296875, + "learning_rate": 0.00012191656664626493, + "loss": 2.4684, + "step": 20223 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.00012191002175093927, + "loss": 2.5016, + "step": 20224 + }, + { + "epoch": 0.86, + "grad_norm": 0.57421875, + "learning_rate": 0.00012190347675702652, + "loss": 2.4834, + "step": 20225 + }, + { + "epoch": 0.86, + "grad_norm": 0.515625, + "learning_rate": 0.00012189693166455609, + "loss": 2.5082, + "step": 20226 + }, + { + "epoch": 0.86, + "grad_norm": 0.42578125, + "learning_rate": 0.00012189038647355749, + "loss": 2.579, + "step": 20227 + }, + { + "epoch": 0.86, + "grad_norm": 0.431640625, + "learning_rate": 0.00012188384118406018, + "loss": 2.458, + "step": 20228 + }, + { + "epoch": 0.86, + "grad_norm": 0.50390625, + "learning_rate": 0.00012187729579609352, + "loss": 2.5815, + "step": 20229 + }, + { + "epoch": 0.86, + "grad_norm": 0.52734375, + "learning_rate": 0.00012187075030968706, + "loss": 2.5624, + "step": 20230 + }, + { + "epoch": 0.86, + "grad_norm": 0.435546875, + "learning_rate": 0.00012186420472487023, + "loss": 2.6546, + "step": 20231 + }, + { + "epoch": 0.86, + "grad_norm": 0.46484375, + "learning_rate": 0.00012185765904167245, + "loss": 2.5919, + "step": 20232 + }, + { + "epoch": 0.86, + "grad_norm": 0.44140625, + "learning_rate": 0.00012185111326012321, + "loss": 2.5386, + "step": 20233 + }, + { + "epoch": 0.86, + "grad_norm": 0.486328125, + "learning_rate": 0.00012184456738025193, + "loss": 2.5841, + "step": 20234 + }, + { + "epoch": 0.86, + "grad_norm": 0.494140625, + "learning_rate": 0.00012183802140208809, + "loss": 2.5275, + "step": 20235 + }, + { + "epoch": 0.86, + "grad_norm": 0.4765625, + "learning_rate": 0.00012183147532566115, + "loss": 2.5334, + "step": 20236 + }, + { + "epoch": 0.86, + "grad_norm": 0.470703125, + "learning_rate": 0.0001218249291510005, + "loss": 2.5284, + "step": 20237 + }, + { + "epoch": 0.86, + "grad_norm": 0.427734375, + "learning_rate": 0.00012181838287813567, + "loss": 2.6219, + "step": 20238 + }, + { + "epoch": 0.86, + "grad_norm": 0.453125, + "learning_rate": 0.00012181183650709611, + "loss": 2.594, + "step": 20239 + }, + { + "epoch": 0.86, + "grad_norm": 0.40625, + "learning_rate": 0.00012180529003791124, + "loss": 2.5527, + "step": 20240 + }, + { + "epoch": 0.86, + "grad_norm": 0.4609375, + "learning_rate": 0.00012179874347061053, + "loss": 2.4943, + "step": 20241 + }, + { + "epoch": 0.86, + "grad_norm": 0.51953125, + "learning_rate": 0.00012179219680522347, + "loss": 2.5304, + "step": 20242 + }, + { + "epoch": 0.86, + "grad_norm": 0.490234375, + "learning_rate": 0.00012178565004177946, + "loss": 2.6799, + "step": 20243 + }, + { + "epoch": 0.86, + "grad_norm": 0.439453125, + "learning_rate": 0.00012177910318030798, + "loss": 2.5537, + "step": 20244 + }, + { + "epoch": 0.86, + "grad_norm": 0.515625, + "learning_rate": 0.00012177255622083852, + "loss": 2.5732, + "step": 20245 + }, + { + "epoch": 0.86, + "grad_norm": 0.50390625, + "learning_rate": 0.00012176600916340048, + "loss": 2.5712, + "step": 20246 + }, + { + "epoch": 0.86, + "grad_norm": 0.4375, + "learning_rate": 0.00012175946200802339, + "loss": 2.7622, + "step": 20247 + }, + { + "epoch": 0.86, + "grad_norm": 0.466796875, + "learning_rate": 0.00012175291475473665, + "loss": 2.5694, + "step": 20248 + }, + { + "epoch": 0.86, + "grad_norm": 0.44921875, + "learning_rate": 0.00012174636740356975, + "loss": 2.4778, + "step": 20249 + }, + { + "epoch": 0.86, + "grad_norm": 0.490234375, + "learning_rate": 0.00012173981995455215, + "loss": 2.7167, + "step": 20250 + }, + { + "epoch": 0.86, + "grad_norm": 0.40234375, + "learning_rate": 0.00012173327240771331, + "loss": 2.592, + "step": 20251 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.00012172672476308267, + "loss": 2.6382, + "step": 20252 + }, + { + "epoch": 0.86, + "grad_norm": 0.47265625, + "learning_rate": 0.00012172017702068971, + "loss": 2.6856, + "step": 20253 + }, + { + "epoch": 0.86, + "grad_norm": 0.515625, + "learning_rate": 0.00012171362918056388, + "loss": 2.5845, + "step": 20254 + }, + { + "epoch": 0.86, + "grad_norm": 0.54296875, + "learning_rate": 0.00012170708124273469, + "loss": 2.5331, + "step": 20255 + }, + { + "epoch": 0.86, + "grad_norm": 0.44140625, + "learning_rate": 0.00012170053320723153, + "loss": 2.6014, + "step": 20256 + }, + { + "epoch": 0.86, + "grad_norm": 0.455078125, + "learning_rate": 0.00012169398507408394, + "loss": 2.6112, + "step": 20257 + }, + { + "epoch": 0.86, + "grad_norm": 0.453125, + "learning_rate": 0.00012168743684332131, + "loss": 2.4898, + "step": 20258 + }, + { + "epoch": 0.86, + "grad_norm": 0.42578125, + "learning_rate": 0.00012168088851497317, + "loss": 2.4888, + "step": 20259 + }, + { + "epoch": 0.86, + "grad_norm": 0.494140625, + "learning_rate": 0.00012167434008906893, + "loss": 2.5312, + "step": 20260 + }, + { + "epoch": 0.86, + "grad_norm": 0.515625, + "learning_rate": 0.0001216677915656381, + "loss": 2.5978, + "step": 20261 + }, + { + "epoch": 0.86, + "grad_norm": 0.478515625, + "learning_rate": 0.00012166124294471011, + "loss": 2.5593, + "step": 20262 + }, + { + "epoch": 0.86, + "grad_norm": 0.51171875, + "learning_rate": 0.00012165469422631445, + "loss": 2.6445, + "step": 20263 + }, + { + "epoch": 0.86, + "grad_norm": 0.41015625, + "learning_rate": 0.00012164814541048057, + "loss": 2.488, + "step": 20264 + }, + { + "epoch": 0.86, + "grad_norm": 0.5078125, + "learning_rate": 0.00012164159649723797, + "loss": 2.6421, + "step": 20265 + }, + { + "epoch": 0.86, + "grad_norm": 0.455078125, + "learning_rate": 0.00012163504748661608, + "loss": 2.5024, + "step": 20266 + }, + { + "epoch": 0.86, + "grad_norm": 0.4140625, + "learning_rate": 0.00012162849837864438, + "loss": 2.5581, + "step": 20267 + }, + { + "epoch": 0.86, + "grad_norm": 0.5, + "learning_rate": 0.00012162194917335235, + "loss": 2.5663, + "step": 20268 + }, + { + "epoch": 0.86, + "grad_norm": 0.478515625, + "learning_rate": 0.00012161539987076947, + "loss": 2.4811, + "step": 20269 + }, + { + "epoch": 0.86, + "grad_norm": 0.474609375, + "learning_rate": 0.00012160885047092514, + "loss": 2.6967, + "step": 20270 + }, + { + "epoch": 0.86, + "grad_norm": 0.498046875, + "learning_rate": 0.00012160230097384893, + "loss": 2.5445, + "step": 20271 + }, + { + "epoch": 0.86, + "grad_norm": 0.435546875, + "learning_rate": 0.00012159575137957022, + "loss": 2.4144, + "step": 20272 + }, + { + "epoch": 0.86, + "grad_norm": 0.41015625, + "learning_rate": 0.00012158920168811854, + "loss": 2.4204, + "step": 20273 + }, + { + "epoch": 0.86, + "grad_norm": 0.439453125, + "learning_rate": 0.00012158265189952334, + "loss": 2.6092, + "step": 20274 + }, + { + "epoch": 0.86, + "grad_norm": 0.421875, + "learning_rate": 0.00012157610201381409, + "loss": 2.4474, + "step": 20275 + }, + { + "epoch": 0.86, + "grad_norm": 0.416015625, + "learning_rate": 0.00012156955203102028, + "loss": 2.5833, + "step": 20276 + }, + { + "epoch": 0.86, + "grad_norm": 0.5078125, + "learning_rate": 0.00012156300195117136, + "loss": 2.4925, + "step": 20277 + }, + { + "epoch": 0.86, + "grad_norm": 0.462890625, + "learning_rate": 0.00012155645177429679, + "loss": 2.5181, + "step": 20278 + }, + { + "epoch": 0.86, + "grad_norm": 0.48046875, + "learning_rate": 0.0001215499015004261, + "loss": 2.6063, + "step": 20279 + }, + { + "epoch": 0.86, + "grad_norm": 0.439453125, + "learning_rate": 0.0001215433511295887, + "loss": 2.5387, + "step": 20280 + }, + { + "epoch": 0.86, + "grad_norm": 0.443359375, + "learning_rate": 0.00012153680066181412, + "loss": 2.574, + "step": 20281 + }, + { + "epoch": 0.86, + "grad_norm": 0.427734375, + "learning_rate": 0.00012153025009713179, + "loss": 2.6047, + "step": 20282 + }, + { + "epoch": 0.86, + "grad_norm": 0.47265625, + "learning_rate": 0.00012152369943557122, + "loss": 2.7034, + "step": 20283 + }, + { + "epoch": 0.86, + "grad_norm": 0.45703125, + "learning_rate": 0.00012151714867716186, + "loss": 2.5726, + "step": 20284 + }, + { + "epoch": 0.86, + "grad_norm": 0.46484375, + "learning_rate": 0.00012151059782193319, + "loss": 2.4814, + "step": 20285 + }, + { + "epoch": 0.86, + "grad_norm": 0.44921875, + "learning_rate": 0.00012150404686991468, + "loss": 2.5056, + "step": 20286 + }, + { + "epoch": 0.86, + "grad_norm": 0.4765625, + "learning_rate": 0.00012149749582113585, + "loss": 2.5689, + "step": 20287 + }, + { + "epoch": 0.86, + "grad_norm": 0.5234375, + "learning_rate": 0.00012149094467562611, + "loss": 2.6987, + "step": 20288 + }, + { + "epoch": 0.86, + "grad_norm": 0.4765625, + "learning_rate": 0.000121484393433415, + "loss": 2.577, + "step": 20289 + }, + { + "epoch": 0.86, + "grad_norm": 0.447265625, + "learning_rate": 0.00012147784209453197, + "loss": 2.4794, + "step": 20290 + }, + { + "epoch": 0.86, + "grad_norm": 0.47265625, + "learning_rate": 0.00012147129065900648, + "loss": 2.448, + "step": 20291 + }, + { + "epoch": 0.86, + "grad_norm": 0.435546875, + "learning_rate": 0.00012146473912686804, + "loss": 2.4892, + "step": 20292 + }, + { + "epoch": 0.86, + "grad_norm": 0.466796875, + "learning_rate": 0.00012145818749814612, + "loss": 2.6829, + "step": 20293 + }, + { + "epoch": 0.86, + "grad_norm": 0.44140625, + "learning_rate": 0.00012145163577287018, + "loss": 2.6739, + "step": 20294 + }, + { + "epoch": 0.86, + "grad_norm": 0.455078125, + "learning_rate": 0.00012144508395106977, + "loss": 2.5034, + "step": 20295 + }, + { + "epoch": 0.86, + "grad_norm": 0.4453125, + "learning_rate": 0.00012143853203277427, + "loss": 2.597, + "step": 20296 + }, + { + "epoch": 0.86, + "grad_norm": 0.52734375, + "learning_rate": 0.00012143198001801322, + "loss": 2.418, + "step": 20297 + }, + { + "epoch": 0.86, + "grad_norm": 0.46875, + "learning_rate": 0.0001214254279068161, + "loss": 2.6404, + "step": 20298 + }, + { + "epoch": 0.86, + "grad_norm": 0.50390625, + "learning_rate": 0.00012141887569921237, + "loss": 2.6344, + "step": 20299 + }, + { + "epoch": 0.87, + "grad_norm": 0.482421875, + "learning_rate": 0.00012141232339523157, + "loss": 2.5997, + "step": 20300 + }, + { + "epoch": 0.87, + "grad_norm": 0.453125, + "learning_rate": 0.00012140577099490309, + "loss": 2.5173, + "step": 20301 + }, + { + "epoch": 0.87, + "grad_norm": 0.44921875, + "learning_rate": 0.00012139921849825646, + "loss": 2.3712, + "step": 20302 + }, + { + "epoch": 0.87, + "grad_norm": 0.4921875, + "learning_rate": 0.00012139266590532123, + "loss": 2.6058, + "step": 20303 + }, + { + "epoch": 0.87, + "grad_norm": 0.453125, + "learning_rate": 0.00012138611321612676, + "loss": 2.4797, + "step": 20304 + }, + { + "epoch": 0.87, + "grad_norm": 0.515625, + "learning_rate": 0.00012137956043070262, + "loss": 2.6025, + "step": 20305 + }, + { + "epoch": 0.87, + "grad_norm": 0.44140625, + "learning_rate": 0.00012137300754907826, + "loss": 2.4629, + "step": 20306 + }, + { + "epoch": 0.87, + "grad_norm": 0.44921875, + "learning_rate": 0.00012136645457128319, + "loss": 2.6266, + "step": 20307 + }, + { + "epoch": 0.87, + "grad_norm": 0.482421875, + "learning_rate": 0.00012135990149734688, + "loss": 2.6148, + "step": 20308 + }, + { + "epoch": 0.87, + "grad_norm": 0.453125, + "learning_rate": 0.00012135334832729884, + "loss": 2.5576, + "step": 20309 + }, + { + "epoch": 0.87, + "grad_norm": 0.4765625, + "learning_rate": 0.00012134679506116849, + "loss": 2.7294, + "step": 20310 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012134024169898541, + "loss": 2.5217, + "step": 20311 + }, + { + "epoch": 0.87, + "grad_norm": 0.490234375, + "learning_rate": 0.000121333688240779, + "loss": 2.6417, + "step": 20312 + }, + { + "epoch": 0.87, + "grad_norm": 0.46484375, + "learning_rate": 0.00012132713468657881, + "loss": 2.5476, + "step": 20313 + }, + { + "epoch": 0.87, + "grad_norm": 0.50390625, + "learning_rate": 0.0001213205810364143, + "loss": 2.6022, + "step": 20314 + }, + { + "epoch": 0.87, + "grad_norm": 0.431640625, + "learning_rate": 0.00012131402729031498, + "loss": 2.4989, + "step": 20315 + }, + { + "epoch": 0.87, + "grad_norm": 0.494140625, + "learning_rate": 0.00012130747344831032, + "loss": 2.6175, + "step": 20316 + }, + { + "epoch": 0.87, + "grad_norm": 0.455078125, + "learning_rate": 0.00012130091951042983, + "loss": 2.6007, + "step": 20317 + }, + { + "epoch": 0.87, + "grad_norm": 0.48828125, + "learning_rate": 0.00012129436547670296, + "loss": 2.7476, + "step": 20318 + }, + { + "epoch": 0.87, + "grad_norm": 0.443359375, + "learning_rate": 0.00012128781134715925, + "loss": 2.56, + "step": 20319 + }, + { + "epoch": 0.87, + "grad_norm": 0.484375, + "learning_rate": 0.00012128125712182816, + "loss": 2.5317, + "step": 20320 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012127470280073919, + "loss": 2.4994, + "step": 20321 + }, + { + "epoch": 0.87, + "grad_norm": 0.462890625, + "learning_rate": 0.00012126814838392184, + "loss": 2.5909, + "step": 20322 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012126159387140559, + "loss": 2.6592, + "step": 20323 + }, + { + "epoch": 0.87, + "grad_norm": 0.40625, + "learning_rate": 0.00012125503926321993, + "loss": 2.4864, + "step": 20324 + }, + { + "epoch": 0.87, + "grad_norm": 0.53125, + "learning_rate": 0.00012124848455939437, + "loss": 2.6838, + "step": 20325 + }, + { + "epoch": 0.87, + "grad_norm": 0.484375, + "learning_rate": 0.00012124192975995838, + "loss": 2.6719, + "step": 20326 + }, + { + "epoch": 0.87, + "grad_norm": 0.41796875, + "learning_rate": 0.0001212353748649415, + "loss": 2.3981, + "step": 20327 + }, + { + "epoch": 0.87, + "grad_norm": 0.51953125, + "learning_rate": 0.00012122881987437316, + "loss": 2.6403, + "step": 20328 + }, + { + "epoch": 0.87, + "grad_norm": 0.4921875, + "learning_rate": 0.00012122226478828289, + "loss": 2.5601, + "step": 20329 + }, + { + "epoch": 0.87, + "grad_norm": 0.4453125, + "learning_rate": 0.0001212157096067002, + "loss": 2.4133, + "step": 20330 + }, + { + "epoch": 0.87, + "grad_norm": 0.462890625, + "learning_rate": 0.00012120915432965459, + "loss": 2.513, + "step": 20331 + }, + { + "epoch": 0.87, + "grad_norm": 0.455078125, + "learning_rate": 0.00012120259895717552, + "loss": 2.5342, + "step": 20332 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.0001211960434892925, + "loss": 2.529, + "step": 20333 + }, + { + "epoch": 0.87, + "grad_norm": 0.44921875, + "learning_rate": 0.00012118948792603499, + "loss": 2.4561, + "step": 20334 + }, + { + "epoch": 0.87, + "grad_norm": 0.498046875, + "learning_rate": 0.0001211829322674326, + "loss": 2.6572, + "step": 20335 + }, + { + "epoch": 0.87, + "grad_norm": 0.447265625, + "learning_rate": 0.0001211763765135147, + "loss": 2.5947, + "step": 20336 + }, + { + "epoch": 0.87, + "grad_norm": 0.41796875, + "learning_rate": 0.00012116982066431089, + "loss": 2.5337, + "step": 20337 + }, + { + "epoch": 0.87, + "grad_norm": 0.51171875, + "learning_rate": 0.00012116326471985058, + "loss": 2.473, + "step": 20338 + }, + { + "epoch": 0.87, + "grad_norm": 0.515625, + "learning_rate": 0.00012115670868016336, + "loss": 2.4668, + "step": 20339 + }, + { + "epoch": 0.87, + "grad_norm": 0.458984375, + "learning_rate": 0.00012115015254527865, + "loss": 2.4886, + "step": 20340 + }, + { + "epoch": 0.87, + "grad_norm": 0.373046875, + "learning_rate": 0.00012114359631522599, + "loss": 2.466, + "step": 20341 + }, + { + "epoch": 0.87, + "grad_norm": 0.458984375, + "learning_rate": 0.00012113703999003485, + "loss": 2.6817, + "step": 20342 + }, + { + "epoch": 0.87, + "grad_norm": 0.443359375, + "learning_rate": 0.00012113048356973479, + "loss": 2.4911, + "step": 20343 + }, + { + "epoch": 0.87, + "grad_norm": 0.5390625, + "learning_rate": 0.00012112392705435526, + "loss": 2.665, + "step": 20344 + }, + { + "epoch": 0.87, + "grad_norm": 0.44921875, + "learning_rate": 0.00012111737044392579, + "loss": 2.5227, + "step": 20345 + }, + { + "epoch": 0.87, + "grad_norm": 0.46484375, + "learning_rate": 0.00012111081373847586, + "loss": 2.6085, + "step": 20346 + }, + { + "epoch": 0.87, + "grad_norm": 0.443359375, + "learning_rate": 0.000121104256938035, + "loss": 2.6876, + "step": 20347 + }, + { + "epoch": 0.87, + "grad_norm": 0.455078125, + "learning_rate": 0.00012109770004263268, + "loss": 2.602, + "step": 20348 + }, + { + "epoch": 0.87, + "grad_norm": 0.40625, + "learning_rate": 0.00012109114305229842, + "loss": 2.6465, + "step": 20349 + }, + { + "epoch": 0.87, + "grad_norm": 0.546875, + "learning_rate": 0.00012108458596706172, + "loss": 2.3951, + "step": 20350 + }, + { + "epoch": 0.87, + "grad_norm": 0.451171875, + "learning_rate": 0.00012107802878695212, + "loss": 2.4894, + "step": 20351 + }, + { + "epoch": 0.87, + "grad_norm": 0.5546875, + "learning_rate": 0.00012107147151199905, + "loss": 2.715, + "step": 20352 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012106491414223209, + "loss": 2.5764, + "step": 20353 + }, + { + "epoch": 0.87, + "grad_norm": 0.5, + "learning_rate": 0.00012105835667768071, + "loss": 2.5396, + "step": 20354 + }, + { + "epoch": 0.87, + "grad_norm": 0.49609375, + "learning_rate": 0.00012105179911837441, + "loss": 2.51, + "step": 20355 + }, + { + "epoch": 0.87, + "grad_norm": 0.451171875, + "learning_rate": 0.0001210452414643427, + "loss": 2.5197, + "step": 20356 + }, + { + "epoch": 0.87, + "grad_norm": 0.4375, + "learning_rate": 0.00012103868371561513, + "loss": 2.6314, + "step": 20357 + }, + { + "epoch": 0.87, + "grad_norm": 0.4453125, + "learning_rate": 0.00012103212587222112, + "loss": 2.5858, + "step": 20358 + }, + { + "epoch": 0.87, + "grad_norm": 0.4140625, + "learning_rate": 0.00012102556793419029, + "loss": 2.4735, + "step": 20359 + }, + { + "epoch": 0.87, + "grad_norm": 0.482421875, + "learning_rate": 0.00012101900990155203, + "loss": 2.5795, + "step": 20360 + }, + { + "epoch": 0.87, + "grad_norm": 0.427734375, + "learning_rate": 0.00012101245177433593, + "loss": 2.5054, + "step": 20361 + }, + { + "epoch": 0.87, + "grad_norm": 0.466796875, + "learning_rate": 0.00012100589355257148, + "loss": 2.5475, + "step": 20362 + }, + { + "epoch": 0.87, + "grad_norm": 0.41015625, + "learning_rate": 0.00012099933523628819, + "loss": 2.5205, + "step": 20363 + }, + { + "epoch": 0.87, + "grad_norm": 0.404296875, + "learning_rate": 0.00012099277682551557, + "loss": 2.5734, + "step": 20364 + }, + { + "epoch": 0.87, + "grad_norm": 0.49609375, + "learning_rate": 0.00012098621832028311, + "loss": 2.4382, + "step": 20365 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012097965972062036, + "loss": 2.4485, + "step": 20366 + }, + { + "epoch": 0.87, + "grad_norm": 0.451171875, + "learning_rate": 0.00012097310102655681, + "loss": 2.5632, + "step": 20367 + }, + { + "epoch": 0.87, + "grad_norm": 0.4765625, + "learning_rate": 0.00012096654223812196, + "loss": 2.4451, + "step": 20368 + }, + { + "epoch": 0.87, + "grad_norm": 0.53125, + "learning_rate": 0.00012095998335534529, + "loss": 2.6467, + "step": 20369 + }, + { + "epoch": 0.87, + "grad_norm": 0.416015625, + "learning_rate": 0.00012095342437825643, + "loss": 2.4522, + "step": 20370 + }, + { + "epoch": 0.87, + "grad_norm": 0.53125, + "learning_rate": 0.00012094686530688474, + "loss": 2.5348, + "step": 20371 + }, + { + "epoch": 0.87, + "grad_norm": 0.412109375, + "learning_rate": 0.00012094030614125986, + "loss": 2.6785, + "step": 20372 + }, + { + "epoch": 0.87, + "grad_norm": 0.4296875, + "learning_rate": 0.00012093374688141125, + "loss": 2.4966, + "step": 20373 + }, + { + "epoch": 0.87, + "grad_norm": 0.45703125, + "learning_rate": 0.00012092718752736842, + "loss": 2.5094, + "step": 20374 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012092062807916092, + "loss": 2.5837, + "step": 20375 + }, + { + "epoch": 0.87, + "grad_norm": 0.46875, + "learning_rate": 0.00012091406853681822, + "loss": 2.5092, + "step": 20376 + }, + { + "epoch": 0.87, + "grad_norm": 0.458984375, + "learning_rate": 0.00012090750890036984, + "loss": 2.5464, + "step": 20377 + }, + { + "epoch": 0.87, + "grad_norm": 0.384765625, + "learning_rate": 0.00012090094916984534, + "loss": 2.5058, + "step": 20378 + }, + { + "epoch": 0.87, + "grad_norm": 0.443359375, + "learning_rate": 0.00012089438934527417, + "loss": 2.5502, + "step": 20379 + }, + { + "epoch": 0.87, + "grad_norm": 0.482421875, + "learning_rate": 0.0001208878294266859, + "loss": 2.5612, + "step": 20380 + }, + { + "epoch": 0.87, + "grad_norm": 0.478515625, + "learning_rate": 0.00012088126941411004, + "loss": 2.6398, + "step": 20381 + }, + { + "epoch": 0.87, + "grad_norm": 0.4453125, + "learning_rate": 0.00012087470930757609, + "loss": 2.5372, + "step": 20382 + }, + { + "epoch": 0.87, + "grad_norm": 0.4140625, + "learning_rate": 0.00012086814910711359, + "loss": 2.6564, + "step": 20383 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012086158881275205, + "loss": 2.5116, + "step": 20384 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012085502842452095, + "loss": 2.6627, + "step": 20385 + }, + { + "epoch": 0.87, + "grad_norm": 0.482421875, + "learning_rate": 0.00012084846794244988, + "loss": 2.6233, + "step": 20386 + }, + { + "epoch": 0.87, + "grad_norm": 0.423828125, + "learning_rate": 0.0001208419073665683, + "loss": 2.4676, + "step": 20387 + }, + { + "epoch": 0.87, + "grad_norm": 0.45703125, + "learning_rate": 0.00012083534669690577, + "loss": 2.6417, + "step": 20388 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012082878593349178, + "loss": 2.5199, + "step": 20389 + }, + { + "epoch": 0.87, + "grad_norm": 0.486328125, + "learning_rate": 0.00012082222507635587, + "loss": 2.7353, + "step": 20390 + }, + { + "epoch": 0.87, + "grad_norm": 0.42578125, + "learning_rate": 0.00012081566412552759, + "loss": 2.6244, + "step": 20391 + }, + { + "epoch": 0.87, + "grad_norm": 0.4609375, + "learning_rate": 0.00012080910308103638, + "loss": 2.5448, + "step": 20392 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012080254194291181, + "loss": 2.6646, + "step": 20393 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012079598071118346, + "loss": 2.7235, + "step": 20394 + }, + { + "epoch": 0.87, + "grad_norm": 0.4296875, + "learning_rate": 0.00012078941938588073, + "loss": 2.6576, + "step": 20395 + }, + { + "epoch": 0.87, + "grad_norm": 0.494140625, + "learning_rate": 0.00012078285796703323, + "loss": 2.4654, + "step": 20396 + }, + { + "epoch": 0.87, + "grad_norm": 0.466796875, + "learning_rate": 0.0001207762964546705, + "loss": 2.4944, + "step": 20397 + }, + { + "epoch": 0.87, + "grad_norm": 0.39453125, + "learning_rate": 0.00012076973484882199, + "loss": 2.3433, + "step": 20398 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012076317314951729, + "loss": 2.4904, + "step": 20399 + }, + { + "epoch": 0.87, + "grad_norm": 0.5390625, + "learning_rate": 0.00012075661135678588, + "loss": 2.7296, + "step": 20400 + }, + { + "epoch": 0.87, + "grad_norm": 0.4921875, + "learning_rate": 0.00012075004947065729, + "loss": 2.5072, + "step": 20401 + }, + { + "epoch": 0.87, + "grad_norm": 0.458984375, + "learning_rate": 0.0001207434874911611, + "loss": 2.5204, + "step": 20402 + }, + { + "epoch": 0.87, + "grad_norm": 0.51953125, + "learning_rate": 0.00012073692541832677, + "loss": 2.576, + "step": 20403 + }, + { + "epoch": 0.87, + "grad_norm": 0.498046875, + "learning_rate": 0.00012073036325218387, + "loss": 2.5036, + "step": 20404 + }, + { + "epoch": 0.87, + "grad_norm": 0.51171875, + "learning_rate": 0.00012072380099276191, + "loss": 2.5338, + "step": 20405 + }, + { + "epoch": 0.87, + "grad_norm": 0.404296875, + "learning_rate": 0.00012071723864009041, + "loss": 2.4494, + "step": 20406 + }, + { + "epoch": 0.87, + "grad_norm": 0.46484375, + "learning_rate": 0.00012071067619419892, + "loss": 2.4997, + "step": 20407 + }, + { + "epoch": 0.87, + "grad_norm": 0.498046875, + "learning_rate": 0.00012070411365511694, + "loss": 2.6068, + "step": 20408 + }, + { + "epoch": 0.87, + "grad_norm": 0.462890625, + "learning_rate": 0.00012069755102287402, + "loss": 2.5333, + "step": 20409 + }, + { + "epoch": 0.87, + "grad_norm": 0.423828125, + "learning_rate": 0.00012069098829749969, + "loss": 2.4197, + "step": 20410 + }, + { + "epoch": 0.87, + "grad_norm": 0.4609375, + "learning_rate": 0.00012068442547902346, + "loss": 2.4208, + "step": 20411 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012067786256747488, + "loss": 2.4307, + "step": 20412 + }, + { + "epoch": 0.87, + "grad_norm": 0.45703125, + "learning_rate": 0.00012067129956288348, + "loss": 2.5631, + "step": 20413 + }, + { + "epoch": 0.87, + "grad_norm": 0.515625, + "learning_rate": 0.0001206647364652788, + "loss": 2.6592, + "step": 20414 + }, + { + "epoch": 0.87, + "grad_norm": 0.49609375, + "learning_rate": 0.00012065817327469035, + "loss": 2.6628, + "step": 20415 + }, + { + "epoch": 0.87, + "grad_norm": 0.48046875, + "learning_rate": 0.00012065160999114766, + "loss": 2.5674, + "step": 20416 + }, + { + "epoch": 0.87, + "grad_norm": 0.44921875, + "learning_rate": 0.00012064504661468026, + "loss": 2.4892, + "step": 20417 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012063848314531773, + "loss": 2.4788, + "step": 20418 + }, + { + "epoch": 0.87, + "grad_norm": 0.443359375, + "learning_rate": 0.00012063191958308953, + "loss": 2.4435, + "step": 20419 + }, + { + "epoch": 0.87, + "grad_norm": 0.431640625, + "learning_rate": 0.00012062535592802526, + "loss": 2.4871, + "step": 20420 + }, + { + "epoch": 0.87, + "grad_norm": 0.458984375, + "learning_rate": 0.00012061879218015442, + "loss": 2.6153, + "step": 20421 + }, + { + "epoch": 0.87, + "grad_norm": 0.4296875, + "learning_rate": 0.00012061222833950657, + "loss": 2.5893, + "step": 20422 + }, + { + "epoch": 0.87, + "grad_norm": 0.46875, + "learning_rate": 0.00012060566440611118, + "loss": 2.6498, + "step": 20423 + }, + { + "epoch": 0.87, + "grad_norm": 0.447265625, + "learning_rate": 0.00012059910037999787, + "loss": 2.4946, + "step": 20424 + }, + { + "epoch": 0.87, + "grad_norm": 0.4453125, + "learning_rate": 0.00012059253626119611, + "loss": 2.5119, + "step": 20425 + }, + { + "epoch": 0.87, + "grad_norm": 0.466796875, + "learning_rate": 0.0001205859720497355, + "loss": 2.541, + "step": 20426 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012057940774564549, + "loss": 2.5996, + "step": 20427 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012057284334895569, + "loss": 2.5841, + "step": 20428 + }, + { + "epoch": 0.87, + "grad_norm": 0.48828125, + "learning_rate": 0.0001205662788596956, + "loss": 2.6264, + "step": 20429 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012055971427789479, + "loss": 2.4879, + "step": 20430 + }, + { + "epoch": 0.87, + "grad_norm": 0.412109375, + "learning_rate": 0.0001205531496035828, + "loss": 2.5781, + "step": 20431 + }, + { + "epoch": 0.87, + "grad_norm": 0.4609375, + "learning_rate": 0.00012054658483678911, + "loss": 2.6515, + "step": 20432 + }, + { + "epoch": 0.87, + "grad_norm": 0.416015625, + "learning_rate": 0.0001205400199775433, + "loss": 2.3654, + "step": 20433 + }, + { + "epoch": 0.87, + "grad_norm": 0.41796875, + "learning_rate": 0.00012053345502587492, + "loss": 2.4613, + "step": 20434 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012052688998181347, + "loss": 2.4023, + "step": 20435 + }, + { + "epoch": 0.87, + "grad_norm": 0.44140625, + "learning_rate": 0.00012052032484538855, + "loss": 2.6894, + "step": 20436 + }, + { + "epoch": 0.87, + "grad_norm": 0.462890625, + "learning_rate": 0.00012051375961662965, + "loss": 2.5423, + "step": 20437 + }, + { + "epoch": 0.87, + "grad_norm": 0.46875, + "learning_rate": 0.00012050719429556633, + "loss": 2.6117, + "step": 20438 + }, + { + "epoch": 0.87, + "grad_norm": 0.52734375, + "learning_rate": 0.00012050062888222816, + "loss": 2.6353, + "step": 20439 + }, + { + "epoch": 0.87, + "grad_norm": 0.4609375, + "learning_rate": 0.00012049406337664462, + "loss": 2.5888, + "step": 20440 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012048749777884527, + "loss": 2.6207, + "step": 20441 + }, + { + "epoch": 0.87, + "grad_norm": 0.51171875, + "learning_rate": 0.00012048093208885971, + "loss": 2.5266, + "step": 20442 + }, + { + "epoch": 0.87, + "grad_norm": 0.490234375, + "learning_rate": 0.00012047436630671742, + "loss": 2.67, + "step": 20443 + }, + { + "epoch": 0.87, + "grad_norm": 0.50390625, + "learning_rate": 0.00012046780043244797, + "loss": 2.5862, + "step": 20444 + }, + { + "epoch": 0.87, + "grad_norm": 0.44921875, + "learning_rate": 0.00012046123446608088, + "loss": 2.5467, + "step": 20445 + }, + { + "epoch": 0.87, + "grad_norm": 0.47265625, + "learning_rate": 0.00012045466840764572, + "loss": 2.6318, + "step": 20446 + }, + { + "epoch": 0.87, + "grad_norm": 0.51953125, + "learning_rate": 0.00012044810225717205, + "loss": 2.5307, + "step": 20447 + }, + { + "epoch": 0.87, + "grad_norm": 0.46484375, + "learning_rate": 0.00012044153601468937, + "loss": 2.6045, + "step": 20448 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012043496968022723, + "loss": 2.6627, + "step": 20449 + }, + { + "epoch": 0.87, + "grad_norm": 0.427734375, + "learning_rate": 0.00012042840325381526, + "loss": 2.4486, + "step": 20450 + }, + { + "epoch": 0.87, + "grad_norm": 0.48828125, + "learning_rate": 0.00012042183673548287, + "loss": 2.5899, + "step": 20451 + }, + { + "epoch": 0.87, + "grad_norm": 0.455078125, + "learning_rate": 0.0001204152701252597, + "loss": 2.5401, + "step": 20452 + }, + { + "epoch": 0.87, + "grad_norm": 0.447265625, + "learning_rate": 0.0001204087034231753, + "loss": 2.7636, + "step": 20453 + }, + { + "epoch": 0.87, + "grad_norm": 0.484375, + "learning_rate": 0.00012040213662925917, + "loss": 2.6202, + "step": 20454 + }, + { + "epoch": 0.87, + "grad_norm": 0.515625, + "learning_rate": 0.0001203955697435409, + "loss": 2.5488, + "step": 20455 + }, + { + "epoch": 0.87, + "grad_norm": 0.46875, + "learning_rate": 0.00012038900276605, + "loss": 2.5903, + "step": 20456 + }, + { + "epoch": 0.87, + "grad_norm": 0.515625, + "learning_rate": 0.00012038243569681601, + "loss": 2.6197, + "step": 20457 + }, + { + "epoch": 0.87, + "grad_norm": 0.453125, + "learning_rate": 0.00012037586853586858, + "loss": 2.4851, + "step": 20458 + }, + { + "epoch": 0.87, + "grad_norm": 0.431640625, + "learning_rate": 0.00012036930128323714, + "loss": 2.4263, + "step": 20459 + }, + { + "epoch": 0.87, + "grad_norm": 0.40234375, + "learning_rate": 0.0001203627339389513, + "loss": 2.4239, + "step": 20460 + }, + { + "epoch": 0.87, + "grad_norm": 0.490234375, + "learning_rate": 0.0001203561665030406, + "loss": 2.5307, + "step": 20461 + }, + { + "epoch": 0.87, + "grad_norm": 0.470703125, + "learning_rate": 0.0001203495989755346, + "loss": 2.4913, + "step": 20462 + }, + { + "epoch": 0.87, + "grad_norm": 0.458984375, + "learning_rate": 0.00012034303135646283, + "loss": 2.7468, + "step": 20463 + }, + { + "epoch": 0.87, + "grad_norm": 0.447265625, + "learning_rate": 0.00012033646364585484, + "loss": 2.5744, + "step": 20464 + }, + { + "epoch": 0.87, + "grad_norm": 0.478515625, + "learning_rate": 0.00012032989584374019, + "loss": 2.6138, + "step": 20465 + }, + { + "epoch": 0.87, + "grad_norm": 0.42578125, + "learning_rate": 0.00012032332795014849, + "loss": 2.5682, + "step": 20466 + }, + { + "epoch": 0.87, + "grad_norm": 0.4609375, + "learning_rate": 0.0001203167599651092, + "loss": 2.6641, + "step": 20467 + }, + { + "epoch": 0.87, + "grad_norm": 0.458984375, + "learning_rate": 0.00012031019188865192, + "loss": 2.4427, + "step": 20468 + }, + { + "epoch": 0.87, + "grad_norm": 0.494140625, + "learning_rate": 0.00012030362372080621, + "loss": 2.5451, + "step": 20469 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012029705546160161, + "loss": 2.5437, + "step": 20470 + }, + { + "epoch": 0.87, + "grad_norm": 0.51171875, + "learning_rate": 0.00012029048711106767, + "loss": 2.5163, + "step": 20471 + }, + { + "epoch": 0.87, + "grad_norm": 0.46875, + "learning_rate": 0.00012028391866923396, + "loss": 2.5572, + "step": 20472 + }, + { + "epoch": 0.87, + "grad_norm": 0.46484375, + "learning_rate": 0.00012027735013613002, + "loss": 2.5348, + "step": 20473 + }, + { + "epoch": 0.87, + "grad_norm": 0.42578125, + "learning_rate": 0.00012027078151178543, + "loss": 2.3175, + "step": 20474 + }, + { + "epoch": 0.87, + "grad_norm": 0.43359375, + "learning_rate": 0.00012026421279622972, + "loss": 2.5934, + "step": 20475 + }, + { + "epoch": 0.87, + "grad_norm": 0.44921875, + "learning_rate": 0.00012025764398949247, + "loss": 2.5835, + "step": 20476 + }, + { + "epoch": 0.87, + "grad_norm": 0.44921875, + "learning_rate": 0.00012025107509160321, + "loss": 2.5613, + "step": 20477 + }, + { + "epoch": 0.87, + "grad_norm": 0.478515625, + "learning_rate": 0.00012024450610259153, + "loss": 2.7106, + "step": 20478 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012023793702248695, + "loss": 2.6457, + "step": 20479 + }, + { + "epoch": 0.87, + "grad_norm": 0.490234375, + "learning_rate": 0.00012023136785131907, + "loss": 2.6172, + "step": 20480 + }, + { + "epoch": 0.87, + "grad_norm": 0.4765625, + "learning_rate": 0.0001202247985891174, + "loss": 2.4866, + "step": 20481 + }, + { + "epoch": 0.87, + "grad_norm": 0.4296875, + "learning_rate": 0.00012021822923591158, + "loss": 2.6141, + "step": 20482 + }, + { + "epoch": 0.87, + "grad_norm": 0.46484375, + "learning_rate": 0.00012021165979173106, + "loss": 2.6775, + "step": 20483 + }, + { + "epoch": 0.87, + "grad_norm": 0.47265625, + "learning_rate": 0.00012020509025660549, + "loss": 2.4883, + "step": 20484 + }, + { + "epoch": 0.87, + "grad_norm": 0.42578125, + "learning_rate": 0.00012019852063056441, + "loss": 2.5118, + "step": 20485 + }, + { + "epoch": 0.87, + "grad_norm": 0.447265625, + "learning_rate": 0.00012019195091363733, + "loss": 2.6185, + "step": 20486 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012018538110585387, + "loss": 2.4681, + "step": 20487 + }, + { + "epoch": 0.87, + "grad_norm": 0.431640625, + "learning_rate": 0.00012017881120724356, + "loss": 2.388, + "step": 20488 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012017224121783602, + "loss": 2.6496, + "step": 20489 + }, + { + "epoch": 0.87, + "grad_norm": 0.447265625, + "learning_rate": 0.0001201656711376607, + "loss": 2.5084, + "step": 20490 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00012015910096674728, + "loss": 2.5949, + "step": 20491 + }, + { + "epoch": 0.87, + "grad_norm": 0.51953125, + "learning_rate": 0.00012015253070512524, + "loss": 2.4992, + "step": 20492 + }, + { + "epoch": 0.87, + "grad_norm": 0.490234375, + "learning_rate": 0.00012014596035282421, + "loss": 2.5412, + "step": 20493 + }, + { + "epoch": 0.87, + "grad_norm": 0.484375, + "learning_rate": 0.00012013938990987368, + "loss": 2.4618, + "step": 20494 + }, + { + "epoch": 0.87, + "grad_norm": 0.494140625, + "learning_rate": 0.00012013281937630328, + "loss": 2.5703, + "step": 20495 + }, + { + "epoch": 0.87, + "grad_norm": 0.4453125, + "learning_rate": 0.00012012624875214255, + "loss": 2.5772, + "step": 20496 + }, + { + "epoch": 0.87, + "grad_norm": 0.451171875, + "learning_rate": 0.00012011967803742106, + "loss": 2.5888, + "step": 20497 + }, + { + "epoch": 0.87, + "grad_norm": 0.462890625, + "learning_rate": 0.00012011310723216837, + "loss": 2.6312, + "step": 20498 + }, + { + "epoch": 0.87, + "grad_norm": 0.4453125, + "learning_rate": 0.00012010653633641405, + "loss": 2.4984, + "step": 20499 + }, + { + "epoch": 0.87, + "grad_norm": 0.48828125, + "learning_rate": 0.00012009996535018764, + "loss": 2.5145, + "step": 20500 + }, + { + "epoch": 0.87, + "grad_norm": 0.435546875, + "learning_rate": 0.00012009339427351877, + "loss": 2.6046, + "step": 20501 + }, + { + "epoch": 0.87, + "grad_norm": 0.48828125, + "learning_rate": 0.00012008682310643693, + "loss": 2.6549, + "step": 20502 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012008025184897173, + "loss": 2.5898, + "step": 20503 + }, + { + "epoch": 0.87, + "grad_norm": 0.55078125, + "learning_rate": 0.00012007368050115277, + "loss": 2.4799, + "step": 20504 + }, + { + "epoch": 0.87, + "grad_norm": 0.46875, + "learning_rate": 0.00012006710906300957, + "loss": 2.5448, + "step": 20505 + }, + { + "epoch": 0.87, + "grad_norm": 0.4609375, + "learning_rate": 0.00012006053753457172, + "loss": 2.3763, + "step": 20506 + }, + { + "epoch": 0.87, + "grad_norm": 0.5234375, + "learning_rate": 0.00012005396591586874, + "loss": 2.7722, + "step": 20507 + }, + { + "epoch": 0.87, + "grad_norm": 0.4765625, + "learning_rate": 0.00012004739420693028, + "loss": 2.567, + "step": 20508 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.0001200408224077859, + "loss": 2.5465, + "step": 20509 + }, + { + "epoch": 0.87, + "grad_norm": 0.451171875, + "learning_rate": 0.00012003425051846506, + "loss": 2.6117, + "step": 20510 + }, + { + "epoch": 0.87, + "grad_norm": 0.453125, + "learning_rate": 0.00012002767853899749, + "loss": 2.4475, + "step": 20511 + }, + { + "epoch": 0.87, + "grad_norm": 0.4296875, + "learning_rate": 0.00012002110646941265, + "loss": 2.5051, + "step": 20512 + }, + { + "epoch": 0.87, + "grad_norm": 0.453125, + "learning_rate": 0.00012001453430974015, + "loss": 2.5636, + "step": 20513 + }, + { + "epoch": 0.87, + "grad_norm": 0.443359375, + "learning_rate": 0.00012000796206000957, + "loss": 2.6175, + "step": 20514 + }, + { + "epoch": 0.87, + "grad_norm": 0.439453125, + "learning_rate": 0.00012000138972025047, + "loss": 2.4931, + "step": 20515 + }, + { + "epoch": 0.87, + "grad_norm": 0.50390625, + "learning_rate": 0.00011999481729049241, + "loss": 2.5132, + "step": 20516 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.00011998824477076502, + "loss": 2.6646, + "step": 20517 + }, + { + "epoch": 0.87, + "grad_norm": 0.44140625, + "learning_rate": 0.00011998167216109779, + "loss": 2.5505, + "step": 20518 + }, + { + "epoch": 0.87, + "grad_norm": 0.58203125, + "learning_rate": 0.00011997509946152035, + "loss": 2.6547, + "step": 20519 + }, + { + "epoch": 0.87, + "grad_norm": 0.451171875, + "learning_rate": 0.00011996852667206226, + "loss": 2.6392, + "step": 20520 + }, + { + "epoch": 0.87, + "grad_norm": 0.45703125, + "learning_rate": 0.00011996195379275312, + "loss": 2.5153, + "step": 20521 + }, + { + "epoch": 0.87, + "grad_norm": 0.40234375, + "learning_rate": 0.00011995538082362247, + "loss": 2.4609, + "step": 20522 + }, + { + "epoch": 0.87, + "grad_norm": 0.4375, + "learning_rate": 0.0001199488077646999, + "loss": 2.6684, + "step": 20523 + }, + { + "epoch": 0.87, + "grad_norm": 0.46875, + "learning_rate": 0.00011994223461601497, + "loss": 2.4885, + "step": 20524 + }, + { + "epoch": 0.87, + "grad_norm": 0.51953125, + "learning_rate": 0.00011993566137759731, + "loss": 2.5978, + "step": 20525 + }, + { + "epoch": 0.87, + "grad_norm": 0.4375, + "learning_rate": 0.00011992908804947642, + "loss": 2.6761, + "step": 20526 + }, + { + "epoch": 0.87, + "grad_norm": 0.447265625, + "learning_rate": 0.00011992251463168194, + "loss": 2.6407, + "step": 20527 + }, + { + "epoch": 0.87, + "grad_norm": 0.474609375, + "learning_rate": 0.0001199159411242434, + "loss": 2.5583, + "step": 20528 + }, + { + "epoch": 0.87, + "grad_norm": 0.490234375, + "learning_rate": 0.00011990936752719044, + "loss": 2.4935, + "step": 20529 + }, + { + "epoch": 0.87, + "grad_norm": 0.49609375, + "learning_rate": 0.00011990279384055258, + "loss": 2.5259, + "step": 20530 + }, + { + "epoch": 0.87, + "grad_norm": 0.4140625, + "learning_rate": 0.00011989622006435943, + "loss": 2.3873, + "step": 20531 + }, + { + "epoch": 0.87, + "grad_norm": 0.48046875, + "learning_rate": 0.00011988964619864054, + "loss": 2.6829, + "step": 20532 + }, + { + "epoch": 0.87, + "grad_norm": 0.453125, + "learning_rate": 0.00011988307224342555, + "loss": 2.6343, + "step": 20533 + }, + { + "epoch": 0.88, + "grad_norm": 0.451171875, + "learning_rate": 0.00011987649819874395, + "loss": 2.5519, + "step": 20534 + }, + { + "epoch": 0.88, + "grad_norm": 0.416015625, + "learning_rate": 0.00011986992406462541, + "loss": 2.4796, + "step": 20535 + }, + { + "epoch": 0.88, + "grad_norm": 0.474609375, + "learning_rate": 0.00011986334984109946, + "loss": 2.4999, + "step": 20536 + }, + { + "epoch": 0.88, + "grad_norm": 0.458984375, + "learning_rate": 0.0001198567755281957, + "loss": 2.4879, + "step": 20537 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.0001198502011259437, + "loss": 2.6932, + "step": 20538 + }, + { + "epoch": 0.88, + "grad_norm": 0.49609375, + "learning_rate": 0.00011984362663437305, + "loss": 2.6224, + "step": 20539 + }, + { + "epoch": 0.88, + "grad_norm": 0.443359375, + "learning_rate": 0.00011983705205351333, + "loss": 2.6358, + "step": 20540 + }, + { + "epoch": 0.88, + "grad_norm": 0.5234375, + "learning_rate": 0.00011983047738339415, + "loss": 2.6495, + "step": 20541 + }, + { + "epoch": 0.88, + "grad_norm": 0.419921875, + "learning_rate": 0.00011982390262404504, + "loss": 2.4958, + "step": 20542 + }, + { + "epoch": 0.88, + "grad_norm": 0.4375, + "learning_rate": 0.00011981732777549561, + "loss": 2.4082, + "step": 20543 + }, + { + "epoch": 0.88, + "grad_norm": 0.470703125, + "learning_rate": 0.00011981075283777547, + "loss": 2.517, + "step": 20544 + }, + { + "epoch": 0.88, + "grad_norm": 0.498046875, + "learning_rate": 0.00011980417781091418, + "loss": 2.7644, + "step": 20545 + }, + { + "epoch": 0.88, + "grad_norm": 0.427734375, + "learning_rate": 0.00011979760269494131, + "loss": 2.5939, + "step": 20546 + }, + { + "epoch": 0.88, + "grad_norm": 0.439453125, + "learning_rate": 0.00011979102748988648, + "loss": 2.5639, + "step": 20547 + }, + { + "epoch": 0.88, + "grad_norm": 0.41796875, + "learning_rate": 0.00011978445219577925, + "loss": 2.4768, + "step": 20548 + }, + { + "epoch": 0.88, + "grad_norm": 0.50390625, + "learning_rate": 0.00011977787681264922, + "loss": 2.5329, + "step": 20549 + }, + { + "epoch": 0.88, + "grad_norm": 0.46875, + "learning_rate": 0.00011977130134052597, + "loss": 2.4895, + "step": 20550 + }, + { + "epoch": 0.88, + "grad_norm": 0.5, + "learning_rate": 0.00011976472577943908, + "loss": 2.6228, + "step": 20551 + }, + { + "epoch": 0.88, + "grad_norm": 0.52734375, + "learning_rate": 0.00011975815012941818, + "loss": 2.4636, + "step": 20552 + }, + { + "epoch": 0.88, + "grad_norm": 0.474609375, + "learning_rate": 0.00011975157439049281, + "loss": 2.5359, + "step": 20553 + }, + { + "epoch": 0.88, + "grad_norm": 0.470703125, + "learning_rate": 0.00011974499856269255, + "loss": 2.6127, + "step": 20554 + }, + { + "epoch": 0.88, + "grad_norm": 0.451171875, + "learning_rate": 0.00011973842264604704, + "loss": 2.4704, + "step": 20555 + }, + { + "epoch": 0.88, + "grad_norm": 0.43359375, + "learning_rate": 0.00011973184664058581, + "loss": 2.6107, + "step": 20556 + }, + { + "epoch": 0.88, + "grad_norm": 0.431640625, + "learning_rate": 0.00011972527054633853, + "loss": 2.5825, + "step": 20557 + }, + { + "epoch": 0.88, + "grad_norm": 0.44140625, + "learning_rate": 0.00011971869436333469, + "loss": 2.5738, + "step": 20558 + }, + { + "epoch": 0.88, + "grad_norm": 0.458984375, + "learning_rate": 0.00011971211809160397, + "loss": 2.689, + "step": 20559 + }, + { + "epoch": 0.88, + "grad_norm": 0.4609375, + "learning_rate": 0.0001197055417311759, + "loss": 2.652, + "step": 20560 + }, + { + "epoch": 0.88, + "grad_norm": 0.408203125, + "learning_rate": 0.00011969896528208009, + "loss": 2.583, + "step": 20561 + }, + { + "epoch": 0.88, + "grad_norm": 0.419921875, + "learning_rate": 0.00011969238874434617, + "loss": 2.4542, + "step": 20562 + }, + { + "epoch": 0.88, + "grad_norm": 0.44921875, + "learning_rate": 0.00011968581211800365, + "loss": 2.5682, + "step": 20563 + }, + { + "epoch": 0.88, + "grad_norm": 0.427734375, + "learning_rate": 0.00011967923540308219, + "loss": 2.5947, + "step": 20564 + }, + { + "epoch": 0.88, + "grad_norm": 0.51953125, + "learning_rate": 0.00011967265859961139, + "loss": 2.6079, + "step": 20565 + }, + { + "epoch": 0.88, + "grad_norm": 0.43359375, + "learning_rate": 0.00011966608170762077, + "loss": 2.5684, + "step": 20566 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.00011965950472714001, + "loss": 2.5397, + "step": 20567 + }, + { + "epoch": 0.88, + "grad_norm": 0.435546875, + "learning_rate": 0.00011965292765819862, + "loss": 2.5342, + "step": 20568 + }, + { + "epoch": 0.88, + "grad_norm": 0.376953125, + "learning_rate": 0.00011964635050082627, + "loss": 2.451, + "step": 20569 + }, + { + "epoch": 0.88, + "grad_norm": 0.427734375, + "learning_rate": 0.00011963977325505253, + "loss": 2.544, + "step": 20570 + }, + { + "epoch": 0.88, + "grad_norm": 0.48828125, + "learning_rate": 0.00011963319592090696, + "loss": 2.6051, + "step": 20571 + }, + { + "epoch": 0.88, + "grad_norm": 0.443359375, + "learning_rate": 0.00011962661849841918, + "loss": 2.6072, + "step": 20572 + }, + { + "epoch": 0.88, + "grad_norm": 0.427734375, + "learning_rate": 0.00011962004098761883, + "loss": 2.5772, + "step": 20573 + }, + { + "epoch": 0.88, + "grad_norm": 0.4375, + "learning_rate": 0.00011961346338853542, + "loss": 2.616, + "step": 20574 + }, + { + "epoch": 0.88, + "grad_norm": 0.47265625, + "learning_rate": 0.00011960688570119863, + "loss": 2.5089, + "step": 20575 + }, + { + "epoch": 0.88, + "grad_norm": 0.50390625, + "learning_rate": 0.00011960030792563801, + "loss": 2.7041, + "step": 20576 + }, + { + "epoch": 0.88, + "grad_norm": 0.435546875, + "learning_rate": 0.00011959373006188314, + "loss": 2.5034, + "step": 20577 + }, + { + "epoch": 0.88, + "grad_norm": 0.55859375, + "learning_rate": 0.0001195871521099637, + "loss": 2.6786, + "step": 20578 + }, + { + "epoch": 0.88, + "grad_norm": 0.453125, + "learning_rate": 0.0001195805740699092, + "loss": 2.5704, + "step": 20579 + }, + { + "epoch": 0.88, + "grad_norm": 0.5, + "learning_rate": 0.00011957399594174925, + "loss": 2.5141, + "step": 20580 + }, + { + "epoch": 0.88, + "grad_norm": 0.478515625, + "learning_rate": 0.00011956741772551352, + "loss": 2.4915, + "step": 20581 + }, + { + "epoch": 0.88, + "grad_norm": 0.5078125, + "learning_rate": 0.00011956083942123151, + "loss": 2.5836, + "step": 20582 + }, + { + "epoch": 0.88, + "grad_norm": 0.46875, + "learning_rate": 0.0001195542610289329, + "loss": 2.6103, + "step": 20583 + }, + { + "epoch": 0.88, + "grad_norm": 0.453125, + "learning_rate": 0.00011954768254864728, + "loss": 2.5383, + "step": 20584 + }, + { + "epoch": 0.88, + "grad_norm": 0.474609375, + "learning_rate": 0.00011954110398040419, + "loss": 2.4164, + "step": 20585 + }, + { + "epoch": 0.88, + "grad_norm": 0.458984375, + "learning_rate": 0.0001195345253242333, + "loss": 2.5409, + "step": 20586 + }, + { + "epoch": 0.88, + "grad_norm": 0.4296875, + "learning_rate": 0.0001195279465801642, + "loss": 2.4523, + "step": 20587 + }, + { + "epoch": 0.88, + "grad_norm": 0.474609375, + "learning_rate": 0.00011952136774822643, + "loss": 2.6016, + "step": 20588 + }, + { + "epoch": 0.88, + "grad_norm": 0.5078125, + "learning_rate": 0.00011951478882844968, + "loss": 2.5873, + "step": 20589 + }, + { + "epoch": 0.88, + "grad_norm": 0.419921875, + "learning_rate": 0.00011950820982086348, + "loss": 2.5535, + "step": 20590 + }, + { + "epoch": 0.88, + "grad_norm": 0.453125, + "learning_rate": 0.0001195016307254975, + "loss": 2.6496, + "step": 20591 + }, + { + "epoch": 0.88, + "grad_norm": 0.41796875, + "learning_rate": 0.00011949505154238129, + "loss": 2.5143, + "step": 20592 + }, + { + "epoch": 0.88, + "grad_norm": 0.4609375, + "learning_rate": 0.0001194884722715445, + "loss": 2.706, + "step": 20593 + }, + { + "epoch": 0.88, + "grad_norm": 0.42578125, + "learning_rate": 0.00011948189291301668, + "loss": 2.4807, + "step": 20594 + }, + { + "epoch": 0.88, + "grad_norm": 0.5234375, + "learning_rate": 0.00011947531346682745, + "loss": 2.5104, + "step": 20595 + }, + { + "epoch": 0.88, + "grad_norm": 0.435546875, + "learning_rate": 0.00011946873393300641, + "loss": 2.5946, + "step": 20596 + }, + { + "epoch": 0.88, + "grad_norm": 0.404296875, + "learning_rate": 0.00011946215431158325, + "loss": 2.5908, + "step": 20597 + }, + { + "epoch": 0.88, + "grad_norm": 0.453125, + "learning_rate": 0.00011945557460258745, + "loss": 2.6257, + "step": 20598 + }, + { + "epoch": 0.88, + "grad_norm": 0.5, + "learning_rate": 0.0001194489948060487, + "loss": 2.5345, + "step": 20599 + }, + { + "epoch": 0.88, + "grad_norm": 0.431640625, + "learning_rate": 0.0001194424149219966, + "loss": 2.4716, + "step": 20600 + }, + { + "epoch": 0.88, + "grad_norm": 0.486328125, + "learning_rate": 0.00011943583495046071, + "loss": 2.5601, + "step": 20601 + }, + { + "epoch": 0.88, + "grad_norm": 0.384765625, + "learning_rate": 0.00011942925489147068, + "loss": 2.6507, + "step": 20602 + }, + { + "epoch": 0.88, + "grad_norm": 0.4765625, + "learning_rate": 0.00011942267474505609, + "loss": 2.5868, + "step": 20603 + }, + { + "epoch": 0.88, + "grad_norm": 0.482421875, + "learning_rate": 0.00011941609451124654, + "loss": 2.5799, + "step": 20604 + }, + { + "epoch": 0.88, + "grad_norm": 0.43359375, + "learning_rate": 0.00011940951419007171, + "loss": 2.5386, + "step": 20605 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.00011940293378156115, + "loss": 2.5451, + "step": 20606 + }, + { + "epoch": 0.88, + "grad_norm": 0.47265625, + "learning_rate": 0.00011939635328574446, + "loss": 2.592, + "step": 20607 + }, + { + "epoch": 0.88, + "grad_norm": 0.46484375, + "learning_rate": 0.00011938977270265129, + "loss": 2.594, + "step": 20608 + }, + { + "epoch": 0.88, + "grad_norm": 0.42578125, + "learning_rate": 0.00011938319203231119, + "loss": 2.5072, + "step": 20609 + }, + { + "epoch": 0.88, + "grad_norm": 0.42578125, + "learning_rate": 0.00011937661127475384, + "loss": 2.5851, + "step": 20610 + }, + { + "epoch": 0.88, + "grad_norm": 0.515625, + "learning_rate": 0.0001193700304300088, + "loss": 2.7023, + "step": 20611 + }, + { + "epoch": 0.88, + "grad_norm": 0.45703125, + "learning_rate": 0.00011936344949810574, + "loss": 2.5969, + "step": 20612 + }, + { + "epoch": 0.88, + "grad_norm": 0.423828125, + "learning_rate": 0.00011935686847907419, + "loss": 2.6195, + "step": 20613 + }, + { + "epoch": 0.88, + "grad_norm": 0.50390625, + "learning_rate": 0.00011935028737294383, + "loss": 2.652, + "step": 20614 + }, + { + "epoch": 0.88, + "grad_norm": 0.51953125, + "learning_rate": 0.00011934370617974424, + "loss": 2.4009, + "step": 20615 + }, + { + "epoch": 0.88, + "grad_norm": 0.42578125, + "learning_rate": 0.00011933712489950506, + "loss": 2.5492, + "step": 20616 + }, + { + "epoch": 0.88, + "grad_norm": 0.462890625, + "learning_rate": 0.00011933054353225586, + "loss": 2.4504, + "step": 20617 + }, + { + "epoch": 0.88, + "grad_norm": 0.4453125, + "learning_rate": 0.00011932396207802629, + "loss": 2.5356, + "step": 20618 + }, + { + "epoch": 0.88, + "grad_norm": 0.443359375, + "learning_rate": 0.00011931738053684595, + "loss": 2.55, + "step": 20619 + }, + { + "epoch": 0.88, + "grad_norm": 0.421875, + "learning_rate": 0.00011931079890874447, + "loss": 2.5785, + "step": 20620 + }, + { + "epoch": 0.88, + "grad_norm": 0.427734375, + "learning_rate": 0.00011930421719375145, + "loss": 2.5866, + "step": 20621 + }, + { + "epoch": 0.88, + "grad_norm": 0.451171875, + "learning_rate": 0.0001192976353918965, + "loss": 2.4704, + "step": 20622 + }, + { + "epoch": 0.88, + "grad_norm": 0.474609375, + "learning_rate": 0.00011929105350320922, + "loss": 2.5452, + "step": 20623 + }, + { + "epoch": 0.88, + "grad_norm": 0.52734375, + "learning_rate": 0.00011928447152771929, + "loss": 2.7482, + "step": 20624 + }, + { + "epoch": 0.88, + "grad_norm": 0.4609375, + "learning_rate": 0.00011927788946545623, + "loss": 2.4851, + "step": 20625 + }, + { + "epoch": 0.88, + "grad_norm": 0.431640625, + "learning_rate": 0.00011927130731644976, + "loss": 2.6606, + "step": 20626 + }, + { + "epoch": 0.88, + "grad_norm": 0.5234375, + "learning_rate": 0.00011926472508072946, + "loss": 2.5706, + "step": 20627 + }, + { + "epoch": 0.88, + "grad_norm": 0.486328125, + "learning_rate": 0.0001192581427583249, + "loss": 2.6805, + "step": 20628 + }, + { + "epoch": 0.88, + "grad_norm": 0.486328125, + "learning_rate": 0.00011925156034926577, + "loss": 2.6491, + "step": 20629 + }, + { + "epoch": 0.88, + "grad_norm": 0.49609375, + "learning_rate": 0.00011924497785358163, + "loss": 2.5952, + "step": 20630 + }, + { + "epoch": 0.88, + "grad_norm": 0.4765625, + "learning_rate": 0.00011923839527130211, + "loss": 2.5733, + "step": 20631 + }, + { + "epoch": 0.88, + "grad_norm": 0.404296875, + "learning_rate": 0.00011923181260245688, + "loss": 2.5182, + "step": 20632 + }, + { + "epoch": 0.88, + "grad_norm": 0.421875, + "learning_rate": 0.00011922522984707549, + "loss": 2.6109, + "step": 20633 + }, + { + "epoch": 0.88, + "grad_norm": 0.44921875, + "learning_rate": 0.00011921864700518761, + "loss": 2.5199, + "step": 20634 + }, + { + "epoch": 0.88, + "grad_norm": 0.421875, + "learning_rate": 0.00011921206407682284, + "loss": 2.5722, + "step": 20635 + }, + { + "epoch": 0.88, + "grad_norm": 0.482421875, + "learning_rate": 0.00011920548106201082, + "loss": 2.4645, + "step": 20636 + }, + { + "epoch": 0.88, + "grad_norm": 0.458984375, + "learning_rate": 0.00011919889796078112, + "loss": 2.497, + "step": 20637 + }, + { + "epoch": 0.88, + "grad_norm": 0.46875, + "learning_rate": 0.00011919231477316342, + "loss": 2.4448, + "step": 20638 + }, + { + "epoch": 0.88, + "grad_norm": 0.515625, + "learning_rate": 0.00011918573149918728, + "loss": 2.6724, + "step": 20639 + }, + { + "epoch": 0.88, + "grad_norm": 0.43359375, + "learning_rate": 0.00011917914813888242, + "loss": 2.669, + "step": 20640 + }, + { + "epoch": 0.88, + "grad_norm": 0.3671875, + "learning_rate": 0.00011917256469227836, + "loss": 2.5355, + "step": 20641 + }, + { + "epoch": 0.88, + "grad_norm": 0.462890625, + "learning_rate": 0.00011916598115940476, + "loss": 2.5482, + "step": 20642 + }, + { + "epoch": 0.88, + "grad_norm": 0.416015625, + "learning_rate": 0.00011915939754029128, + "loss": 2.5433, + "step": 20643 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.00011915281383496748, + "loss": 2.6479, + "step": 20644 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.00011914623004346304, + "loss": 2.5208, + "step": 20645 + }, + { + "epoch": 0.88, + "grad_norm": 0.470703125, + "learning_rate": 0.00011913964616580757, + "loss": 2.5318, + "step": 20646 + }, + { + "epoch": 0.88, + "grad_norm": 0.4765625, + "learning_rate": 0.00011913306220203066, + "loss": 2.5603, + "step": 20647 + }, + { + "epoch": 0.88, + "grad_norm": 0.470703125, + "learning_rate": 0.000119126478152162, + "loss": 2.5291, + "step": 20648 + }, + { + "epoch": 0.88, + "grad_norm": 0.46484375, + "learning_rate": 0.00011911989401623113, + "loss": 2.5401, + "step": 20649 + }, + { + "epoch": 0.88, + "grad_norm": 0.455078125, + "learning_rate": 0.00011911330979426776, + "loss": 2.668, + "step": 20650 + }, + { + "epoch": 0.88, + "grad_norm": 0.4609375, + "learning_rate": 0.00011910672548630147, + "loss": 2.4721, + "step": 20651 + }, + { + "epoch": 0.88, + "grad_norm": 0.44921875, + "learning_rate": 0.0001191001410923619, + "loss": 2.5587, + "step": 20652 + }, + { + "epoch": 0.88, + "grad_norm": 0.49609375, + "learning_rate": 0.00011909355661247865, + "loss": 2.4926, + "step": 20653 + }, + { + "epoch": 0.88, + "grad_norm": 0.439453125, + "learning_rate": 0.00011908697204668142, + "loss": 2.6026, + "step": 20654 + }, + { + "epoch": 0.88, + "grad_norm": 0.470703125, + "learning_rate": 0.00011908038739499976, + "loss": 2.5328, + "step": 20655 + }, + { + "epoch": 0.88, + "grad_norm": 0.48828125, + "learning_rate": 0.00011907380265746335, + "loss": 2.5673, + "step": 20656 + }, + { + "epoch": 0.88, + "grad_norm": 0.453125, + "learning_rate": 0.00011906721783410178, + "loss": 2.4402, + "step": 20657 + }, + { + "epoch": 0.88, + "grad_norm": 0.48046875, + "learning_rate": 0.00011906063292494469, + "loss": 2.4675, + "step": 20658 + }, + { + "epoch": 0.88, + "grad_norm": 0.4609375, + "learning_rate": 0.00011905404793002176, + "loss": 2.533, + "step": 20659 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.00011904746284936253, + "loss": 2.5435, + "step": 20660 + }, + { + "epoch": 0.88, + "grad_norm": 0.41796875, + "learning_rate": 0.00011904087768299667, + "loss": 2.6722, + "step": 20661 + }, + { + "epoch": 0.88, + "grad_norm": 0.451171875, + "learning_rate": 0.00011903429243095388, + "loss": 2.505, + "step": 20662 + }, + { + "epoch": 0.88, + "grad_norm": 0.4296875, + "learning_rate": 0.00011902770709326367, + "loss": 2.6203, + "step": 20663 + }, + { + "epoch": 0.88, + "grad_norm": 0.5390625, + "learning_rate": 0.0001190211216699558, + "loss": 2.4473, + "step": 20664 + }, + { + "epoch": 0.88, + "grad_norm": 0.419921875, + "learning_rate": 0.00011901453616105977, + "loss": 2.5263, + "step": 20665 + }, + { + "epoch": 0.88, + "grad_norm": 0.498046875, + "learning_rate": 0.0001190079505666053, + "loss": 2.6333, + "step": 20666 + }, + { + "epoch": 0.88, + "grad_norm": 0.443359375, + "learning_rate": 0.00011900136488662201, + "loss": 2.5625, + "step": 20667 + }, + { + "epoch": 0.88, + "grad_norm": 0.451171875, + "learning_rate": 0.0001189947791211395, + "loss": 2.6557, + "step": 20668 + }, + { + "epoch": 0.88, + "grad_norm": 0.451171875, + "learning_rate": 0.00011898819327018742, + "loss": 2.5648, + "step": 20669 + }, + { + "epoch": 0.88, + "grad_norm": 0.419921875, + "learning_rate": 0.00011898160733379547, + "loss": 2.5244, + "step": 20670 + }, + { + "epoch": 0.88, + "grad_norm": 0.44140625, + "learning_rate": 0.00011897502131199314, + "loss": 2.4399, + "step": 20671 + }, + { + "epoch": 0.88, + "grad_norm": 0.474609375, + "learning_rate": 0.00011896843520481021, + "loss": 2.5524, + "step": 20672 + }, + { + "epoch": 0.88, + "grad_norm": 0.4609375, + "learning_rate": 0.00011896184901227623, + "loss": 2.5349, + "step": 20673 + }, + { + "epoch": 0.88, + "grad_norm": 0.423828125, + "learning_rate": 0.00011895526273442086, + "loss": 2.6253, + "step": 20674 + }, + { + "epoch": 0.88, + "grad_norm": 0.5078125, + "learning_rate": 0.00011894867637127374, + "loss": 2.6402, + "step": 20675 + }, + { + "epoch": 0.88, + "grad_norm": 0.44140625, + "learning_rate": 0.00011894208992286451, + "loss": 2.5311, + "step": 20676 + }, + { + "epoch": 0.88, + "grad_norm": 0.412109375, + "learning_rate": 0.00011893550338922277, + "loss": 2.5761, + "step": 20677 + }, + { + "epoch": 0.88, + "grad_norm": 0.5078125, + "learning_rate": 0.00011892891677037821, + "loss": 2.4253, + "step": 20678 + }, + { + "epoch": 0.88, + "grad_norm": 0.5, + "learning_rate": 0.00011892233006636042, + "loss": 2.6846, + "step": 20679 + }, + { + "epoch": 0.88, + "grad_norm": 0.447265625, + "learning_rate": 0.00011891574327719908, + "loss": 2.5316, + "step": 20680 + }, + { + "epoch": 0.88, + "grad_norm": 0.451171875, + "learning_rate": 0.0001189091564029238, + "loss": 2.7216, + "step": 20681 + }, + { + "epoch": 0.88, + "grad_norm": 0.484375, + "learning_rate": 0.00011890256944356424, + "loss": 2.6759, + "step": 20682 + }, + { + "epoch": 0.88, + "grad_norm": 0.41015625, + "learning_rate": 0.00011889598239915, + "loss": 2.6044, + "step": 20683 + }, + { + "epoch": 0.88, + "grad_norm": 0.51953125, + "learning_rate": 0.00011888939526971077, + "loss": 2.6666, + "step": 20684 + }, + { + "epoch": 0.88, + "grad_norm": 0.498046875, + "learning_rate": 0.00011888280805527614, + "loss": 2.6104, + "step": 20685 + }, + { + "epoch": 0.88, + "grad_norm": 0.453125, + "learning_rate": 0.00011887622075587578, + "loss": 2.4618, + "step": 20686 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.00011886963337153934, + "loss": 2.5012, + "step": 20687 + }, + { + "epoch": 0.88, + "grad_norm": 0.484375, + "learning_rate": 0.00011886304590229644, + "loss": 2.6379, + "step": 20688 + }, + { + "epoch": 0.88, + "grad_norm": 0.490234375, + "learning_rate": 0.00011885645834817673, + "loss": 2.6071, + "step": 20689 + }, + { + "epoch": 0.88, + "grad_norm": 0.453125, + "learning_rate": 0.00011884987070920984, + "loss": 2.5097, + "step": 20690 + }, + { + "epoch": 0.88, + "grad_norm": 0.44921875, + "learning_rate": 0.00011884328298542544, + "loss": 2.498, + "step": 20691 + }, + { + "epoch": 0.88, + "grad_norm": 0.5078125, + "learning_rate": 0.00011883669517685315, + "loss": 2.5196, + "step": 20692 + }, + { + "epoch": 0.88, + "grad_norm": 0.435546875, + "learning_rate": 0.00011883010728352257, + "loss": 2.4187, + "step": 20693 + }, + { + "epoch": 0.88, + "grad_norm": 0.53125, + "learning_rate": 0.00011882351930546344, + "loss": 2.8156, + "step": 20694 + }, + { + "epoch": 0.88, + "grad_norm": 0.50390625, + "learning_rate": 0.00011881693124270533, + "loss": 2.6866, + "step": 20695 + }, + { + "epoch": 0.88, + "grad_norm": 0.46875, + "learning_rate": 0.00011881034309527789, + "loss": 2.5044, + "step": 20696 + }, + { + "epoch": 0.88, + "grad_norm": 0.388671875, + "learning_rate": 0.00011880375486321082, + "loss": 2.4718, + "step": 20697 + }, + { + "epoch": 0.88, + "grad_norm": 0.435546875, + "learning_rate": 0.0001187971665465337, + "loss": 2.4131, + "step": 20698 + }, + { + "epoch": 0.88, + "grad_norm": 0.435546875, + "learning_rate": 0.00011879057814527621, + "loss": 2.5285, + "step": 20699 + }, + { + "epoch": 0.88, + "grad_norm": 0.42578125, + "learning_rate": 0.000118783989659468, + "loss": 2.508, + "step": 20700 + }, + { + "epoch": 0.88, + "grad_norm": 0.416015625, + "learning_rate": 0.00011877740108913866, + "loss": 2.4766, + "step": 20701 + }, + { + "epoch": 0.88, + "grad_norm": 0.515625, + "learning_rate": 0.00011877081243431792, + "loss": 2.6069, + "step": 20702 + }, + { + "epoch": 0.88, + "grad_norm": 0.421875, + "learning_rate": 0.00011876422369503534, + "loss": 2.6188, + "step": 20703 + }, + { + "epoch": 0.88, + "grad_norm": 0.48046875, + "learning_rate": 0.00011875763487132065, + "loss": 2.5208, + "step": 20704 + }, + { + "epoch": 0.88, + "grad_norm": 0.412109375, + "learning_rate": 0.00011875104596320343, + "loss": 2.4594, + "step": 20705 + }, + { + "epoch": 0.88, + "grad_norm": 0.48828125, + "learning_rate": 0.00011874445697071339, + "loss": 2.6251, + "step": 20706 + }, + { + "epoch": 0.88, + "grad_norm": 0.4453125, + "learning_rate": 0.00011873786789388013, + "loss": 2.5774, + "step": 20707 + }, + { + "epoch": 0.88, + "grad_norm": 0.47265625, + "learning_rate": 0.0001187312787327333, + "loss": 2.4832, + "step": 20708 + }, + { + "epoch": 0.88, + "grad_norm": 0.47265625, + "learning_rate": 0.00011872468948730255, + "loss": 2.5627, + "step": 20709 + }, + { + "epoch": 0.88, + "grad_norm": 0.447265625, + "learning_rate": 0.00011871810015761759, + "loss": 2.5885, + "step": 20710 + }, + { + "epoch": 0.88, + "grad_norm": 0.48046875, + "learning_rate": 0.00011871151074370797, + "loss": 2.6218, + "step": 20711 + }, + { + "epoch": 0.88, + "grad_norm": 0.4375, + "learning_rate": 0.00011870492124560342, + "loss": 2.505, + "step": 20712 + }, + { + "epoch": 0.88, + "grad_norm": 0.435546875, + "learning_rate": 0.00011869833166333352, + "loss": 2.594, + "step": 20713 + }, + { + "epoch": 0.88, + "grad_norm": 0.490234375, + "learning_rate": 0.000118691741996928, + "loss": 2.5668, + "step": 20714 + }, + { + "epoch": 0.88, + "grad_norm": 0.4609375, + "learning_rate": 0.00011868515224641648, + "loss": 2.6726, + "step": 20715 + }, + { + "epoch": 0.88, + "grad_norm": 0.44140625, + "learning_rate": 0.00011867856241182858, + "loss": 2.474, + "step": 20716 + }, + { + "epoch": 0.88, + "grad_norm": 0.46875, + "learning_rate": 0.00011867197249319395, + "loss": 2.5714, + "step": 20717 + }, + { + "epoch": 0.88, + "grad_norm": 0.51171875, + "learning_rate": 0.00011866538249054233, + "loss": 2.6185, + "step": 20718 + }, + { + "epoch": 0.88, + "grad_norm": 0.408203125, + "learning_rate": 0.00011865879240390328, + "loss": 2.5917, + "step": 20719 + }, + { + "epoch": 0.88, + "grad_norm": 0.486328125, + "learning_rate": 0.00011865220223330646, + "loss": 2.6576, + "step": 20720 + }, + { + "epoch": 0.88, + "grad_norm": 0.40234375, + "learning_rate": 0.00011864561197878159, + "loss": 2.4302, + "step": 20721 + }, + { + "epoch": 0.88, + "grad_norm": 0.515625, + "learning_rate": 0.00011863902164035825, + "loss": 2.6236, + "step": 20722 + }, + { + "epoch": 0.88, + "grad_norm": 0.50390625, + "learning_rate": 0.00011863243121806614, + "loss": 2.5734, + "step": 20723 + }, + { + "epoch": 0.88, + "grad_norm": 0.419921875, + "learning_rate": 0.0001186258407119349, + "loss": 2.4307, + "step": 20724 + }, + { + "epoch": 0.88, + "grad_norm": 0.52734375, + "learning_rate": 0.00011861925012199416, + "loss": 2.5583, + "step": 20725 + }, + { + "epoch": 0.88, + "grad_norm": 0.482421875, + "learning_rate": 0.00011861265944827363, + "loss": 2.6866, + "step": 20726 + }, + { + "epoch": 0.88, + "grad_norm": 0.447265625, + "learning_rate": 0.0001186060686908029, + "loss": 2.5756, + "step": 20727 + }, + { + "epoch": 0.88, + "grad_norm": 0.49609375, + "learning_rate": 0.00011859947784961167, + "loss": 2.5868, + "step": 20728 + }, + { + "epoch": 0.88, + "grad_norm": 0.419921875, + "learning_rate": 0.00011859288692472963, + "loss": 2.4164, + "step": 20729 + }, + { + "epoch": 0.88, + "grad_norm": 0.451171875, + "learning_rate": 0.00011858629591618634, + "loss": 2.6539, + "step": 20730 + }, + { + "epoch": 0.88, + "grad_norm": 0.42578125, + "learning_rate": 0.00011857970482401152, + "loss": 2.4863, + "step": 20731 + }, + { + "epoch": 0.88, + "grad_norm": 0.44921875, + "learning_rate": 0.00011857311364823484, + "loss": 2.5521, + "step": 20732 + }, + { + "epoch": 0.88, + "grad_norm": 0.50390625, + "learning_rate": 0.00011856652238888592, + "loss": 2.5653, + "step": 20733 + }, + { + "epoch": 0.88, + "grad_norm": 0.515625, + "learning_rate": 0.00011855993104599445, + "loss": 2.6649, + "step": 20734 + }, + { + "epoch": 0.88, + "grad_norm": 0.423828125, + "learning_rate": 0.00011855333961959007, + "loss": 2.548, + "step": 20735 + }, + { + "epoch": 0.88, + "grad_norm": 0.453125, + "learning_rate": 0.00011854674810970241, + "loss": 2.6777, + "step": 20736 + }, + { + "epoch": 0.88, + "grad_norm": 0.484375, + "learning_rate": 0.0001185401565163612, + "loss": 2.6233, + "step": 20737 + }, + { + "epoch": 0.88, + "grad_norm": 0.48828125, + "learning_rate": 0.00011853356483959606, + "loss": 2.7408, + "step": 20738 + }, + { + "epoch": 0.88, + "grad_norm": 0.49609375, + "learning_rate": 0.00011852697307943663, + "loss": 2.6308, + "step": 20739 + }, + { + "epoch": 0.88, + "grad_norm": 0.4765625, + "learning_rate": 0.0001185203812359126, + "loss": 2.5162, + "step": 20740 + }, + { + "epoch": 0.88, + "grad_norm": 0.45703125, + "learning_rate": 0.00011851378930905363, + "loss": 2.5538, + "step": 20741 + }, + { + "epoch": 0.88, + "grad_norm": 0.439453125, + "learning_rate": 0.00011850719729888937, + "loss": 2.6118, + "step": 20742 + }, + { + "epoch": 0.88, + "grad_norm": 0.4453125, + "learning_rate": 0.0001185006052054495, + "loss": 2.6662, + "step": 20743 + }, + { + "epoch": 0.88, + "grad_norm": 0.423828125, + "learning_rate": 0.00011849401302876364, + "loss": 2.5642, + "step": 20744 + }, + { + "epoch": 0.88, + "grad_norm": 0.4453125, + "learning_rate": 0.0001184874207688615, + "loss": 2.5586, + "step": 20745 + }, + { + "epoch": 0.88, + "grad_norm": 0.4296875, + "learning_rate": 0.0001184808284257727, + "loss": 2.5113, + "step": 20746 + }, + { + "epoch": 0.88, + "grad_norm": 0.486328125, + "learning_rate": 0.00011847423599952695, + "loss": 2.5145, + "step": 20747 + }, + { + "epoch": 0.88, + "grad_norm": 0.498046875, + "learning_rate": 0.00011846764349015389, + "loss": 2.6431, + "step": 20748 + }, + { + "epoch": 0.88, + "grad_norm": 0.447265625, + "learning_rate": 0.00011846105089768315, + "loss": 2.565, + "step": 20749 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.00011845445822214447, + "loss": 2.5928, + "step": 20750 + }, + { + "epoch": 0.88, + "grad_norm": 0.443359375, + "learning_rate": 0.00011844786546356746, + "loss": 2.6635, + "step": 20751 + }, + { + "epoch": 0.88, + "grad_norm": 0.482421875, + "learning_rate": 0.00011844127262198177, + "loss": 2.6624, + "step": 20752 + }, + { + "epoch": 0.88, + "grad_norm": 0.4609375, + "learning_rate": 0.00011843467969741715, + "loss": 2.4937, + "step": 20753 + }, + { + "epoch": 0.88, + "grad_norm": 0.51171875, + "learning_rate": 0.00011842808668990315, + "loss": 2.6078, + "step": 20754 + }, + { + "epoch": 0.88, + "grad_norm": 0.490234375, + "learning_rate": 0.00011842149359946952, + "loss": 2.616, + "step": 20755 + }, + { + "epoch": 0.88, + "grad_norm": 0.421875, + "learning_rate": 0.0001184149004261459, + "loss": 2.5481, + "step": 20756 + }, + { + "epoch": 0.88, + "grad_norm": 0.5390625, + "learning_rate": 0.00011840830716996196, + "loss": 2.5953, + "step": 20757 + }, + { + "epoch": 0.88, + "grad_norm": 0.41796875, + "learning_rate": 0.00011840171383094736, + "loss": 2.5651, + "step": 20758 + }, + { + "epoch": 0.88, + "grad_norm": 0.466796875, + "learning_rate": 0.00011839512040913178, + "loss": 2.4834, + "step": 20759 + }, + { + "epoch": 0.88, + "grad_norm": 0.46875, + "learning_rate": 0.00011838852690454485, + "loss": 2.6032, + "step": 20760 + }, + { + "epoch": 0.88, + "grad_norm": 0.49609375, + "learning_rate": 0.00011838193331721632, + "loss": 2.7746, + "step": 20761 + }, + { + "epoch": 0.88, + "grad_norm": 0.404296875, + "learning_rate": 0.00011837533964717577, + "loss": 2.4608, + "step": 20762 + }, + { + "epoch": 0.88, + "grad_norm": 0.47265625, + "learning_rate": 0.00011836874589445292, + "loss": 2.5697, + "step": 20763 + }, + { + "epoch": 0.88, + "grad_norm": 0.455078125, + "learning_rate": 0.00011836215205907744, + "loss": 2.4895, + "step": 20764 + }, + { + "epoch": 0.88, + "grad_norm": 0.486328125, + "learning_rate": 0.00011835555814107897, + "loss": 2.5493, + "step": 20765 + }, + { + "epoch": 0.88, + "grad_norm": 0.44140625, + "learning_rate": 0.0001183489641404872, + "loss": 2.5635, + "step": 20766 + }, + { + "epoch": 0.88, + "grad_norm": 0.48828125, + "learning_rate": 0.0001183423700573318, + "loss": 2.5219, + "step": 20767 + }, + { + "epoch": 0.88, + "grad_norm": 0.41796875, + "learning_rate": 0.00011833577589164242, + "loss": 2.5306, + "step": 20768 + }, + { + "epoch": 0.89, + "grad_norm": 0.439453125, + "learning_rate": 0.00011832918164344877, + "loss": 2.5046, + "step": 20769 + }, + { + "epoch": 0.89, + "grad_norm": 0.490234375, + "learning_rate": 0.00011832258731278049, + "loss": 2.5205, + "step": 20770 + }, + { + "epoch": 0.89, + "grad_norm": 0.451171875, + "learning_rate": 0.00011831599289966726, + "loss": 2.622, + "step": 20771 + }, + { + "epoch": 0.89, + "grad_norm": 0.4375, + "learning_rate": 0.00011830939840413878, + "loss": 2.5131, + "step": 20772 + }, + { + "epoch": 0.89, + "grad_norm": 0.51171875, + "learning_rate": 0.00011830280382622468, + "loss": 2.5607, + "step": 20773 + }, + { + "epoch": 0.89, + "grad_norm": 0.49609375, + "learning_rate": 0.00011829620916595465, + "loss": 2.5662, + "step": 20774 + }, + { + "epoch": 0.89, + "grad_norm": 0.4140625, + "learning_rate": 0.00011828961442335837, + "loss": 2.4251, + "step": 20775 + }, + { + "epoch": 0.89, + "grad_norm": 0.41796875, + "learning_rate": 0.0001182830195984655, + "loss": 2.5749, + "step": 20776 + }, + { + "epoch": 0.89, + "grad_norm": 0.431640625, + "learning_rate": 0.00011827642469130575, + "loss": 2.4302, + "step": 20777 + }, + { + "epoch": 0.89, + "grad_norm": 0.51171875, + "learning_rate": 0.00011826982970190874, + "loss": 2.5658, + "step": 20778 + }, + { + "epoch": 0.89, + "grad_norm": 0.44140625, + "learning_rate": 0.00011826323463030417, + "loss": 2.5546, + "step": 20779 + }, + { + "epoch": 0.89, + "grad_norm": 0.384765625, + "learning_rate": 0.00011825663947652174, + "loss": 2.5376, + "step": 20780 + }, + { + "epoch": 0.89, + "grad_norm": 0.421875, + "learning_rate": 0.00011825004424059112, + "loss": 2.5787, + "step": 20781 + }, + { + "epoch": 0.89, + "grad_norm": 0.4375, + "learning_rate": 0.00011824344892254193, + "loss": 2.6211, + "step": 20782 + }, + { + "epoch": 0.89, + "grad_norm": 0.49609375, + "learning_rate": 0.00011823685352240393, + "loss": 2.7341, + "step": 20783 + }, + { + "epoch": 0.89, + "grad_norm": 0.40234375, + "learning_rate": 0.00011823025804020669, + "loss": 2.4596, + "step": 20784 + }, + { + "epoch": 0.89, + "grad_norm": 0.455078125, + "learning_rate": 0.00011822366247598002, + "loss": 2.4335, + "step": 20785 + }, + { + "epoch": 0.89, + "grad_norm": 0.5, + "learning_rate": 0.00011821706682975349, + "loss": 2.5538, + "step": 20786 + }, + { + "epoch": 0.89, + "grad_norm": 0.458984375, + "learning_rate": 0.00011821047110155681, + "loss": 2.6557, + "step": 20787 + }, + { + "epoch": 0.89, + "grad_norm": 0.435546875, + "learning_rate": 0.00011820387529141967, + "loss": 2.6426, + "step": 20788 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011819727939937177, + "loss": 2.5838, + "step": 20789 + }, + { + "epoch": 0.89, + "grad_norm": 0.453125, + "learning_rate": 0.00011819068342544275, + "loss": 2.6359, + "step": 20790 + }, + { + "epoch": 0.89, + "grad_norm": 0.408203125, + "learning_rate": 0.0001181840873696623, + "loss": 2.3835, + "step": 20791 + }, + { + "epoch": 0.89, + "grad_norm": 0.416015625, + "learning_rate": 0.00011817749123206008, + "loss": 2.5796, + "step": 20792 + }, + { + "epoch": 0.89, + "grad_norm": 0.474609375, + "learning_rate": 0.00011817089501266584, + "loss": 2.53, + "step": 20793 + }, + { + "epoch": 0.89, + "grad_norm": 0.44140625, + "learning_rate": 0.00011816429871150918, + "loss": 2.5483, + "step": 20794 + }, + { + "epoch": 0.89, + "grad_norm": 0.4609375, + "learning_rate": 0.00011815770232861982, + "loss": 2.5343, + "step": 20795 + }, + { + "epoch": 0.89, + "grad_norm": 0.46484375, + "learning_rate": 0.00011815110586402744, + "loss": 2.6735, + "step": 20796 + }, + { + "epoch": 0.89, + "grad_norm": 0.494140625, + "learning_rate": 0.00011814450931776171, + "loss": 2.7361, + "step": 20797 + }, + { + "epoch": 0.89, + "grad_norm": 0.4921875, + "learning_rate": 0.00011813791268985232, + "loss": 2.4845, + "step": 20798 + }, + { + "epoch": 0.89, + "grad_norm": 0.462890625, + "learning_rate": 0.00011813131598032896, + "loss": 2.6282, + "step": 20799 + }, + { + "epoch": 0.89, + "grad_norm": 0.48828125, + "learning_rate": 0.00011812471918922127, + "loss": 2.6691, + "step": 20800 + }, + { + "epoch": 0.89, + "grad_norm": 0.439453125, + "learning_rate": 0.00011811812231655903, + "loss": 2.5889, + "step": 20801 + }, + { + "epoch": 0.89, + "grad_norm": 0.435546875, + "learning_rate": 0.00011811152536237182, + "loss": 2.4489, + "step": 20802 + }, + { + "epoch": 0.89, + "grad_norm": 0.384765625, + "learning_rate": 0.00011810492832668938, + "loss": 2.4592, + "step": 20803 + }, + { + "epoch": 0.89, + "grad_norm": 0.47265625, + "learning_rate": 0.00011809833120954137, + "loss": 2.602, + "step": 20804 + }, + { + "epoch": 0.89, + "grad_norm": 0.466796875, + "learning_rate": 0.00011809173401095748, + "loss": 2.5373, + "step": 20805 + }, + { + "epoch": 0.89, + "grad_norm": 0.41015625, + "learning_rate": 0.00011808513673096741, + "loss": 2.4281, + "step": 20806 + }, + { + "epoch": 0.89, + "grad_norm": 0.4296875, + "learning_rate": 0.00011807853936960083, + "loss": 2.5135, + "step": 20807 + }, + { + "epoch": 0.89, + "grad_norm": 0.48046875, + "learning_rate": 0.00011807194192688741, + "loss": 2.5718, + "step": 20808 + }, + { + "epoch": 0.89, + "grad_norm": 0.412109375, + "learning_rate": 0.0001180653444028569, + "loss": 2.3148, + "step": 20809 + }, + { + "epoch": 0.89, + "grad_norm": 0.494140625, + "learning_rate": 0.0001180587467975389, + "loss": 2.576, + "step": 20810 + }, + { + "epoch": 0.89, + "grad_norm": 0.455078125, + "learning_rate": 0.00011805214911096316, + "loss": 2.5849, + "step": 20811 + }, + { + "epoch": 0.89, + "grad_norm": 0.4375, + "learning_rate": 0.00011804555134315936, + "loss": 2.4125, + "step": 20812 + }, + { + "epoch": 0.89, + "grad_norm": 0.412109375, + "learning_rate": 0.00011803895349415715, + "loss": 2.5898, + "step": 20813 + }, + { + "epoch": 0.89, + "grad_norm": 0.44140625, + "learning_rate": 0.00011803235556398624, + "loss": 2.6291, + "step": 20814 + }, + { + "epoch": 0.89, + "grad_norm": 0.408203125, + "learning_rate": 0.00011802575755267633, + "loss": 2.5722, + "step": 20815 + }, + { + "epoch": 0.89, + "grad_norm": 0.486328125, + "learning_rate": 0.00011801915946025707, + "loss": 2.7184, + "step": 20816 + }, + { + "epoch": 0.89, + "grad_norm": 0.44921875, + "learning_rate": 0.00011801256128675824, + "loss": 2.4351, + "step": 20817 + }, + { + "epoch": 0.89, + "grad_norm": 0.470703125, + "learning_rate": 0.0001180059630322094, + "loss": 2.5923, + "step": 20818 + }, + { + "epoch": 0.89, + "grad_norm": 0.5078125, + "learning_rate": 0.00011799936469664035, + "loss": 2.6517, + "step": 20819 + }, + { + "epoch": 0.89, + "grad_norm": 0.50390625, + "learning_rate": 0.00011799276628008073, + "loss": 2.6779, + "step": 20820 + }, + { + "epoch": 0.89, + "grad_norm": 0.474609375, + "learning_rate": 0.00011798616778256022, + "loss": 2.609, + "step": 20821 + }, + { + "epoch": 0.89, + "grad_norm": 0.443359375, + "learning_rate": 0.00011797956920410855, + "loss": 2.5894, + "step": 20822 + }, + { + "epoch": 0.89, + "grad_norm": 0.498046875, + "learning_rate": 0.00011797297054475538, + "loss": 2.4796, + "step": 20823 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011796637180453039, + "loss": 2.541, + "step": 20824 + }, + { + "epoch": 0.89, + "grad_norm": 0.41015625, + "learning_rate": 0.00011795977298346334, + "loss": 2.5133, + "step": 20825 + }, + { + "epoch": 0.89, + "grad_norm": 0.482421875, + "learning_rate": 0.00011795317408158384, + "loss": 2.6084, + "step": 20826 + }, + { + "epoch": 0.89, + "grad_norm": 0.45703125, + "learning_rate": 0.00011794657509892162, + "loss": 2.5138, + "step": 20827 + }, + { + "epoch": 0.89, + "grad_norm": 0.486328125, + "learning_rate": 0.00011793997603550637, + "loss": 2.5844, + "step": 20828 + }, + { + "epoch": 0.89, + "grad_norm": 0.41796875, + "learning_rate": 0.0001179333768913678, + "loss": 2.3711, + "step": 20829 + }, + { + "epoch": 0.89, + "grad_norm": 0.451171875, + "learning_rate": 0.00011792677766653559, + "loss": 2.4906, + "step": 20830 + }, + { + "epoch": 0.89, + "grad_norm": 0.50390625, + "learning_rate": 0.00011792017836103941, + "loss": 2.5151, + "step": 20831 + }, + { + "epoch": 0.89, + "grad_norm": 0.47265625, + "learning_rate": 0.00011791357897490898, + "loss": 2.536, + "step": 20832 + }, + { + "epoch": 0.89, + "grad_norm": 0.51953125, + "learning_rate": 0.00011790697950817403, + "loss": 2.6736, + "step": 20833 + }, + { + "epoch": 0.89, + "grad_norm": 0.4765625, + "learning_rate": 0.00011790037996086416, + "loss": 2.4062, + "step": 20834 + }, + { + "epoch": 0.89, + "grad_norm": 0.490234375, + "learning_rate": 0.00011789378033300915, + "loss": 2.5383, + "step": 20835 + }, + { + "epoch": 0.89, + "grad_norm": 0.4609375, + "learning_rate": 0.00011788718062463866, + "loss": 2.4901, + "step": 20836 + }, + { + "epoch": 0.89, + "grad_norm": 0.435546875, + "learning_rate": 0.0001178805808357824, + "loss": 2.5213, + "step": 20837 + }, + { + "epoch": 0.89, + "grad_norm": 0.494140625, + "learning_rate": 0.00011787398096647007, + "loss": 2.6225, + "step": 20838 + }, + { + "epoch": 0.89, + "grad_norm": 0.423828125, + "learning_rate": 0.00011786738101673136, + "loss": 2.5728, + "step": 20839 + }, + { + "epoch": 0.89, + "grad_norm": 0.4375, + "learning_rate": 0.00011786078098659592, + "loss": 2.5346, + "step": 20840 + }, + { + "epoch": 0.89, + "grad_norm": 0.4921875, + "learning_rate": 0.00011785418087609355, + "loss": 2.5351, + "step": 20841 + }, + { + "epoch": 0.89, + "grad_norm": 0.455078125, + "learning_rate": 0.00011784758068525385, + "loss": 2.5514, + "step": 20842 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011784098041410659, + "loss": 2.6962, + "step": 20843 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011783438006268141, + "loss": 2.6075, + "step": 20844 + }, + { + "epoch": 0.89, + "grad_norm": 0.4609375, + "learning_rate": 0.00011782777963100806, + "loss": 2.5878, + "step": 20845 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011782117911911621, + "loss": 2.5007, + "step": 20846 + }, + { + "epoch": 0.89, + "grad_norm": 0.4296875, + "learning_rate": 0.00011781457852703557, + "loss": 2.5495, + "step": 20847 + }, + { + "epoch": 0.89, + "grad_norm": 0.486328125, + "learning_rate": 0.00011780797785479585, + "loss": 2.5663, + "step": 20848 + }, + { + "epoch": 0.89, + "grad_norm": 0.498046875, + "learning_rate": 0.00011780137710242671, + "loss": 2.4952, + "step": 20849 + }, + { + "epoch": 0.89, + "grad_norm": 0.515625, + "learning_rate": 0.00011779477626995789, + "loss": 2.6054, + "step": 20850 + }, + { + "epoch": 0.89, + "grad_norm": 0.470703125, + "learning_rate": 0.00011778817535741906, + "loss": 2.5534, + "step": 20851 + }, + { + "epoch": 0.89, + "grad_norm": 0.474609375, + "learning_rate": 0.00011778157436483998, + "loss": 2.567, + "step": 20852 + }, + { + "epoch": 0.89, + "grad_norm": 0.416015625, + "learning_rate": 0.00011777497329225029, + "loss": 2.4887, + "step": 20853 + }, + { + "epoch": 0.89, + "grad_norm": 0.435546875, + "learning_rate": 0.00011776837213967972, + "loss": 2.5468, + "step": 20854 + }, + { + "epoch": 0.89, + "grad_norm": 0.50390625, + "learning_rate": 0.00011776177090715796, + "loss": 2.7042, + "step": 20855 + }, + { + "epoch": 0.89, + "grad_norm": 0.4921875, + "learning_rate": 0.00011775516959471475, + "loss": 2.5872, + "step": 20856 + }, + { + "epoch": 0.89, + "grad_norm": 0.451171875, + "learning_rate": 0.00011774856820237974, + "loss": 2.585, + "step": 20857 + }, + { + "epoch": 0.89, + "grad_norm": 0.48828125, + "learning_rate": 0.00011774196673018268, + "loss": 2.4643, + "step": 20858 + }, + { + "epoch": 0.89, + "grad_norm": 0.40234375, + "learning_rate": 0.00011773536517815322, + "loss": 2.4609, + "step": 20859 + }, + { + "epoch": 0.89, + "grad_norm": 0.4375, + "learning_rate": 0.00011772876354632114, + "loss": 2.6046, + "step": 20860 + }, + { + "epoch": 0.89, + "grad_norm": 0.486328125, + "learning_rate": 0.00011772216183471607, + "loss": 2.591, + "step": 20861 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011771556004336777, + "loss": 2.5535, + "step": 20862 + }, + { + "epoch": 0.89, + "grad_norm": 0.50390625, + "learning_rate": 0.00011770895817230592, + "loss": 2.5216, + "step": 20863 + }, + { + "epoch": 0.89, + "grad_norm": 0.451171875, + "learning_rate": 0.0001177023562215602, + "loss": 2.5582, + "step": 20864 + }, + { + "epoch": 0.89, + "grad_norm": 0.416015625, + "learning_rate": 0.00011769575419116039, + "loss": 2.569, + "step": 20865 + }, + { + "epoch": 0.89, + "grad_norm": 0.498046875, + "learning_rate": 0.00011768915208113613, + "loss": 2.5645, + "step": 20866 + }, + { + "epoch": 0.89, + "grad_norm": 0.40234375, + "learning_rate": 0.00011768254989151714, + "loss": 2.4952, + "step": 20867 + }, + { + "epoch": 0.89, + "grad_norm": 0.44921875, + "learning_rate": 0.00011767594762233318, + "loss": 2.6198, + "step": 20868 + }, + { + "epoch": 0.89, + "grad_norm": 0.439453125, + "learning_rate": 0.00011766934527361386, + "loss": 2.5476, + "step": 20869 + }, + { + "epoch": 0.89, + "grad_norm": 0.5, + "learning_rate": 0.00011766274284538899, + "loss": 2.5783, + "step": 20870 + }, + { + "epoch": 0.89, + "grad_norm": 0.474609375, + "learning_rate": 0.00011765614033768822, + "loss": 2.5287, + "step": 20871 + }, + { + "epoch": 0.89, + "grad_norm": 0.427734375, + "learning_rate": 0.00011764953775054126, + "loss": 2.5074, + "step": 20872 + }, + { + "epoch": 0.89, + "grad_norm": 0.458984375, + "learning_rate": 0.00011764293508397783, + "loss": 2.6346, + "step": 20873 + }, + { + "epoch": 0.89, + "grad_norm": 0.462890625, + "learning_rate": 0.00011763633233802767, + "loss": 2.5297, + "step": 20874 + }, + { + "epoch": 0.89, + "grad_norm": 0.4921875, + "learning_rate": 0.0001176297295127204, + "loss": 2.612, + "step": 20875 + }, + { + "epoch": 0.89, + "grad_norm": 0.53515625, + "learning_rate": 0.00011762312660808585, + "loss": 2.5361, + "step": 20876 + }, + { + "epoch": 0.89, + "grad_norm": 0.42578125, + "learning_rate": 0.00011761652362415363, + "loss": 2.5827, + "step": 20877 + }, + { + "epoch": 0.89, + "grad_norm": 0.48046875, + "learning_rate": 0.00011760992056095351, + "loss": 2.6217, + "step": 20878 + }, + { + "epoch": 0.89, + "grad_norm": 0.458984375, + "learning_rate": 0.00011760331741851518, + "loss": 2.7091, + "step": 20879 + }, + { + "epoch": 0.89, + "grad_norm": 0.4609375, + "learning_rate": 0.00011759671419686836, + "loss": 2.5553, + "step": 20880 + }, + { + "epoch": 0.89, + "grad_norm": 0.466796875, + "learning_rate": 0.00011759011089604275, + "loss": 2.4039, + "step": 20881 + }, + { + "epoch": 0.89, + "grad_norm": 0.435546875, + "learning_rate": 0.00011758350751606809, + "loss": 2.5076, + "step": 20882 + }, + { + "epoch": 0.89, + "grad_norm": 0.443359375, + "learning_rate": 0.00011757690405697403, + "loss": 2.4813, + "step": 20883 + }, + { + "epoch": 0.89, + "grad_norm": 0.419921875, + "learning_rate": 0.00011757030051879037, + "loss": 2.4814, + "step": 20884 + }, + { + "epoch": 0.89, + "grad_norm": 0.3984375, + "learning_rate": 0.00011756369690154674, + "loss": 2.508, + "step": 20885 + }, + { + "epoch": 0.89, + "grad_norm": 0.5, + "learning_rate": 0.0001175570932052729, + "loss": 2.5483, + "step": 20886 + }, + { + "epoch": 0.89, + "grad_norm": 0.392578125, + "learning_rate": 0.00011755048942999857, + "loss": 2.4238, + "step": 20887 + }, + { + "epoch": 0.89, + "grad_norm": 0.419921875, + "learning_rate": 0.00011754388557575347, + "loss": 2.535, + "step": 20888 + }, + { + "epoch": 0.89, + "grad_norm": 0.53515625, + "learning_rate": 0.00011753728164256728, + "loss": 2.6527, + "step": 20889 + }, + { + "epoch": 0.89, + "grad_norm": 0.443359375, + "learning_rate": 0.00011753067763046972, + "loss": 2.5714, + "step": 20890 + }, + { + "epoch": 0.89, + "grad_norm": 0.421875, + "learning_rate": 0.0001175240735394905, + "loss": 2.6058, + "step": 20891 + }, + { + "epoch": 0.89, + "grad_norm": 0.4609375, + "learning_rate": 0.00011751746936965941, + "loss": 2.6047, + "step": 20892 + }, + { + "epoch": 0.89, + "grad_norm": 0.404296875, + "learning_rate": 0.00011751086512100606, + "loss": 2.6006, + "step": 20893 + }, + { + "epoch": 0.89, + "grad_norm": 0.466796875, + "learning_rate": 0.00011750426079356026, + "loss": 2.5067, + "step": 20894 + }, + { + "epoch": 0.89, + "grad_norm": 0.431640625, + "learning_rate": 0.00011749765638735163, + "loss": 2.5034, + "step": 20895 + }, + { + "epoch": 0.89, + "grad_norm": 0.474609375, + "learning_rate": 0.00011749105190240999, + "loss": 2.5804, + "step": 20896 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011748444733876498, + "loss": 2.54, + "step": 20897 + }, + { + "epoch": 0.89, + "grad_norm": 0.53515625, + "learning_rate": 0.00011747784269644637, + "loss": 2.5461, + "step": 20898 + }, + { + "epoch": 0.89, + "grad_norm": 0.474609375, + "learning_rate": 0.00011747123797548383, + "loss": 2.6066, + "step": 20899 + }, + { + "epoch": 0.89, + "grad_norm": 0.478515625, + "learning_rate": 0.00011746463317590712, + "loss": 2.5343, + "step": 20900 + }, + { + "epoch": 0.89, + "grad_norm": 0.447265625, + "learning_rate": 0.00011745802829774593, + "loss": 2.4146, + "step": 20901 + }, + { + "epoch": 0.89, + "grad_norm": 0.45703125, + "learning_rate": 0.00011745142334103001, + "loss": 2.5087, + "step": 20902 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011744481830578904, + "loss": 2.6674, + "step": 20903 + }, + { + "epoch": 0.89, + "grad_norm": 0.458984375, + "learning_rate": 0.0001174382131920528, + "loss": 2.6542, + "step": 20904 + }, + { + "epoch": 0.89, + "grad_norm": 0.4921875, + "learning_rate": 0.00011743160799985094, + "loss": 2.6235, + "step": 20905 + }, + { + "epoch": 0.89, + "grad_norm": 0.455078125, + "learning_rate": 0.00011742500272921321, + "loss": 2.4539, + "step": 20906 + }, + { + "epoch": 0.89, + "grad_norm": 0.47265625, + "learning_rate": 0.00011741839738016936, + "loss": 2.5898, + "step": 20907 + }, + { + "epoch": 0.89, + "grad_norm": 0.423828125, + "learning_rate": 0.0001174117919527491, + "loss": 2.4291, + "step": 20908 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011740518644698209, + "loss": 2.5483, + "step": 20909 + }, + { + "epoch": 0.89, + "grad_norm": 0.458984375, + "learning_rate": 0.00011739858086289812, + "loss": 2.7064, + "step": 20910 + }, + { + "epoch": 0.89, + "grad_norm": 0.423828125, + "learning_rate": 0.00011739197520052691, + "loss": 2.5815, + "step": 20911 + }, + { + "epoch": 0.89, + "grad_norm": 0.45703125, + "learning_rate": 0.00011738536945989817, + "loss": 2.6562, + "step": 20912 + }, + { + "epoch": 0.89, + "grad_norm": 0.482421875, + "learning_rate": 0.00011737876364104161, + "loss": 2.4457, + "step": 20913 + }, + { + "epoch": 0.89, + "grad_norm": 0.412109375, + "learning_rate": 0.00011737215774398698, + "loss": 2.489, + "step": 20914 + }, + { + "epoch": 0.89, + "grad_norm": 0.412109375, + "learning_rate": 0.00011736555176876396, + "loss": 2.4453, + "step": 20915 + }, + { + "epoch": 0.89, + "grad_norm": 0.447265625, + "learning_rate": 0.00011735894571540234, + "loss": 2.5529, + "step": 20916 + }, + { + "epoch": 0.89, + "grad_norm": 0.53515625, + "learning_rate": 0.00011735233958393177, + "loss": 2.6794, + "step": 20917 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011734573337438201, + "loss": 2.5301, + "step": 20918 + }, + { + "epoch": 0.89, + "grad_norm": 0.44921875, + "learning_rate": 0.00011733912708678282, + "loss": 2.436, + "step": 20919 + }, + { + "epoch": 0.89, + "grad_norm": 0.51171875, + "learning_rate": 0.0001173325207211639, + "loss": 2.4734, + "step": 20920 + }, + { + "epoch": 0.89, + "grad_norm": 0.4140625, + "learning_rate": 0.00011732591427755496, + "loss": 2.4315, + "step": 20921 + }, + { + "epoch": 0.89, + "grad_norm": 0.45703125, + "learning_rate": 0.00011731930775598574, + "loss": 2.6565, + "step": 20922 + }, + { + "epoch": 0.89, + "grad_norm": 0.462890625, + "learning_rate": 0.00011731270115648594, + "loss": 2.5141, + "step": 20923 + }, + { + "epoch": 0.89, + "grad_norm": 0.4375, + "learning_rate": 0.00011730609447908534, + "loss": 2.5245, + "step": 20924 + }, + { + "epoch": 0.89, + "grad_norm": 0.46484375, + "learning_rate": 0.00011729948772381361, + "loss": 2.5509, + "step": 20925 + }, + { + "epoch": 0.89, + "grad_norm": 0.427734375, + "learning_rate": 0.00011729288089070054, + "loss": 2.6636, + "step": 20926 + }, + { + "epoch": 0.89, + "grad_norm": 0.470703125, + "learning_rate": 0.00011728627397977581, + "loss": 2.533, + "step": 20927 + }, + { + "epoch": 0.89, + "grad_norm": 0.5, + "learning_rate": 0.00011727966699106919, + "loss": 2.6616, + "step": 20928 + }, + { + "epoch": 0.89, + "grad_norm": 0.431640625, + "learning_rate": 0.00011727305992461035, + "loss": 2.5324, + "step": 20929 + }, + { + "epoch": 0.89, + "grad_norm": 0.376953125, + "learning_rate": 0.00011726645278042907, + "loss": 2.5336, + "step": 20930 + }, + { + "epoch": 0.89, + "grad_norm": 0.40625, + "learning_rate": 0.00011725984555855503, + "loss": 2.5977, + "step": 20931 + }, + { + "epoch": 0.89, + "grad_norm": 0.451171875, + "learning_rate": 0.00011725323825901803, + "loss": 2.5458, + "step": 20932 + }, + { + "epoch": 0.89, + "grad_norm": 0.443359375, + "learning_rate": 0.00011724663088184775, + "loss": 2.4983, + "step": 20933 + }, + { + "epoch": 0.89, + "grad_norm": 0.431640625, + "learning_rate": 0.00011724002342707394, + "loss": 2.5413, + "step": 20934 + }, + { + "epoch": 0.89, + "grad_norm": 0.423828125, + "learning_rate": 0.00011723341589472631, + "loss": 2.5523, + "step": 20935 + }, + { + "epoch": 0.89, + "grad_norm": 0.455078125, + "learning_rate": 0.00011722680828483463, + "loss": 2.7721, + "step": 20936 + }, + { + "epoch": 0.89, + "grad_norm": 0.486328125, + "learning_rate": 0.0001172202005974286, + "loss": 2.6325, + "step": 20937 + }, + { + "epoch": 0.89, + "grad_norm": 0.443359375, + "learning_rate": 0.00011721359283253796, + "loss": 2.4805, + "step": 20938 + }, + { + "epoch": 0.89, + "grad_norm": 0.53125, + "learning_rate": 0.00011720698499019243, + "loss": 2.6943, + "step": 20939 + }, + { + "epoch": 0.89, + "grad_norm": 0.50390625, + "learning_rate": 0.00011720037707042178, + "loss": 2.5982, + "step": 20940 + }, + { + "epoch": 0.89, + "grad_norm": 0.458984375, + "learning_rate": 0.00011719376907325568, + "loss": 2.6092, + "step": 20941 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011718716099872393, + "loss": 2.4513, + "step": 20942 + }, + { + "epoch": 0.89, + "grad_norm": 0.490234375, + "learning_rate": 0.00011718055284685625, + "loss": 2.5138, + "step": 20943 + }, + { + "epoch": 0.89, + "grad_norm": 0.48046875, + "learning_rate": 0.00011717394461768234, + "loss": 2.5665, + "step": 20944 + }, + { + "epoch": 0.89, + "grad_norm": 0.451171875, + "learning_rate": 0.00011716733631123198, + "loss": 2.6355, + "step": 20945 + }, + { + "epoch": 0.89, + "grad_norm": 0.486328125, + "learning_rate": 0.00011716072792753486, + "loss": 2.4643, + "step": 20946 + }, + { + "epoch": 0.89, + "grad_norm": 0.4609375, + "learning_rate": 0.00011715411946662072, + "loss": 2.4827, + "step": 20947 + }, + { + "epoch": 0.89, + "grad_norm": 0.431640625, + "learning_rate": 0.00011714751092851935, + "loss": 2.5518, + "step": 20948 + }, + { + "epoch": 0.89, + "grad_norm": 0.484375, + "learning_rate": 0.0001171409023132604, + "loss": 2.5238, + "step": 20949 + }, + { + "epoch": 0.89, + "grad_norm": 0.443359375, + "learning_rate": 0.00011713429362087369, + "loss": 2.6497, + "step": 20950 + }, + { + "epoch": 0.89, + "grad_norm": 0.51171875, + "learning_rate": 0.00011712768485138892, + "loss": 2.636, + "step": 20951 + }, + { + "epoch": 0.89, + "grad_norm": 0.4921875, + "learning_rate": 0.00011712107600483581, + "loss": 2.5839, + "step": 20952 + }, + { + "epoch": 0.89, + "grad_norm": 0.494140625, + "learning_rate": 0.00011711446708124414, + "loss": 2.5166, + "step": 20953 + }, + { + "epoch": 0.89, + "grad_norm": 0.48046875, + "learning_rate": 0.0001171078580806436, + "loss": 2.4743, + "step": 20954 + }, + { + "epoch": 0.89, + "grad_norm": 0.4375, + "learning_rate": 0.00011710124900306394, + "loss": 2.567, + "step": 20955 + }, + { + "epoch": 0.89, + "grad_norm": 0.41015625, + "learning_rate": 0.00011709463984853494, + "loss": 2.5502, + "step": 20956 + }, + { + "epoch": 0.89, + "grad_norm": 0.421875, + "learning_rate": 0.00011708803061708629, + "loss": 2.4997, + "step": 20957 + }, + { + "epoch": 0.89, + "grad_norm": 0.455078125, + "learning_rate": 0.00011708142130874775, + "loss": 2.4942, + "step": 20958 + }, + { + "epoch": 0.89, + "grad_norm": 0.46484375, + "learning_rate": 0.00011707481192354906, + "loss": 2.6314, + "step": 20959 + }, + { + "epoch": 0.89, + "grad_norm": 0.439453125, + "learning_rate": 0.00011706820246151995, + "loss": 2.7194, + "step": 20960 + }, + { + "epoch": 0.89, + "grad_norm": 0.482421875, + "learning_rate": 0.00011706159292269019, + "loss": 2.5586, + "step": 20961 + }, + { + "epoch": 0.89, + "grad_norm": 0.421875, + "learning_rate": 0.00011705498330708949, + "loss": 2.4287, + "step": 20962 + }, + { + "epoch": 0.89, + "grad_norm": 0.4375, + "learning_rate": 0.00011704837361474755, + "loss": 2.4675, + "step": 20963 + }, + { + "epoch": 0.89, + "grad_norm": 0.498046875, + "learning_rate": 0.00011704176384569424, + "loss": 2.5624, + "step": 20964 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011703515399995914, + "loss": 2.6875, + "step": 20965 + }, + { + "epoch": 0.89, + "grad_norm": 0.458984375, + "learning_rate": 0.00011702854407757212, + "loss": 2.5748, + "step": 20966 + }, + { + "epoch": 0.89, + "grad_norm": 0.5234375, + "learning_rate": 0.00011702193407856286, + "loss": 2.6987, + "step": 20967 + }, + { + "epoch": 0.89, + "grad_norm": 0.369140625, + "learning_rate": 0.00011701532400296113, + "loss": 2.5005, + "step": 20968 + }, + { + "epoch": 0.89, + "grad_norm": 0.466796875, + "learning_rate": 0.00011700871385079666, + "loss": 2.5334, + "step": 20969 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011700210362209918, + "loss": 2.6909, + "step": 20970 + }, + { + "epoch": 0.89, + "grad_norm": 0.427734375, + "learning_rate": 0.00011699549331689845, + "loss": 2.7655, + "step": 20971 + }, + { + "epoch": 0.89, + "grad_norm": 0.474609375, + "learning_rate": 0.00011698888293522423, + "loss": 2.6414, + "step": 20972 + }, + { + "epoch": 0.89, + "grad_norm": 0.431640625, + "learning_rate": 0.00011698227247710623, + "loss": 2.5199, + "step": 20973 + }, + { + "epoch": 0.89, + "grad_norm": 0.474609375, + "learning_rate": 0.00011697566194257418, + "loss": 2.475, + "step": 20974 + }, + { + "epoch": 0.89, + "grad_norm": 0.462890625, + "learning_rate": 0.00011696905133165792, + "loss": 2.5343, + "step": 20975 + }, + { + "epoch": 0.89, + "grad_norm": 0.478515625, + "learning_rate": 0.00011696244064438708, + "loss": 2.4446, + "step": 20976 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011695582988079145, + "loss": 2.5982, + "step": 20977 + }, + { + "epoch": 0.89, + "grad_norm": 0.453125, + "learning_rate": 0.00011694921904090081, + "loss": 2.5679, + "step": 20978 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011694260812474489, + "loss": 2.5203, + "step": 20979 + }, + { + "epoch": 0.89, + "grad_norm": 0.49609375, + "learning_rate": 0.00011693599713235342, + "loss": 2.464, + "step": 20980 + }, + { + "epoch": 0.89, + "grad_norm": 0.44921875, + "learning_rate": 0.00011692938606375614, + "loss": 2.6453, + "step": 20981 + }, + { + "epoch": 0.89, + "grad_norm": 0.44921875, + "learning_rate": 0.00011692277491898279, + "loss": 2.4322, + "step": 20982 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011691616369806318, + "loss": 2.5125, + "step": 20983 + }, + { + "epoch": 0.89, + "grad_norm": 0.43359375, + "learning_rate": 0.00011690955240102697, + "loss": 2.6023, + "step": 20984 + }, + { + "epoch": 0.89, + "grad_norm": 0.453125, + "learning_rate": 0.00011690294102790398, + "loss": 2.5699, + "step": 20985 + }, + { + "epoch": 0.89, + "grad_norm": 0.396484375, + "learning_rate": 0.00011689632957872395, + "loss": 2.4083, + "step": 20986 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011688971805351659, + "loss": 2.4722, + "step": 20987 + }, + { + "epoch": 0.89, + "grad_norm": 0.5, + "learning_rate": 0.00011688310645231166, + "loss": 2.6589, + "step": 20988 + }, + { + "epoch": 0.89, + "grad_norm": 0.443359375, + "learning_rate": 0.00011687649477513893, + "loss": 2.5017, + "step": 20989 + }, + { + "epoch": 0.89, + "grad_norm": 0.423828125, + "learning_rate": 0.00011686988302202813, + "loss": 2.5822, + "step": 20990 + }, + { + "epoch": 0.89, + "grad_norm": 0.4453125, + "learning_rate": 0.00011686327119300906, + "loss": 2.6245, + "step": 20991 + }, + { + "epoch": 0.89, + "grad_norm": 0.48828125, + "learning_rate": 0.00011685665928811138, + "loss": 2.5735, + "step": 20992 + }, + { + "epoch": 0.89, + "grad_norm": 0.43359375, + "learning_rate": 0.00011685004730736493, + "loss": 2.5259, + "step": 20993 + }, + { + "epoch": 0.89, + "grad_norm": 0.435546875, + "learning_rate": 0.00011684343525079941, + "loss": 2.6108, + "step": 20994 + }, + { + "epoch": 0.89, + "grad_norm": 0.447265625, + "learning_rate": 0.00011683682311844461, + "loss": 2.6312, + "step": 20995 + }, + { + "epoch": 0.89, + "grad_norm": 0.44140625, + "learning_rate": 0.00011683021091033023, + "loss": 2.4101, + "step": 20996 + }, + { + "epoch": 0.89, + "grad_norm": 0.41015625, + "learning_rate": 0.00011682359862648608, + "loss": 2.529, + "step": 20997 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011681698626694187, + "loss": 2.5838, + "step": 20998 + }, + { + "epoch": 0.89, + "grad_norm": 0.46484375, + "learning_rate": 0.00011681037383172738, + "loss": 2.5775, + "step": 20999 + }, + { + "epoch": 0.89, + "grad_norm": 0.46875, + "learning_rate": 0.00011680376132087232, + "loss": 2.5107, + "step": 21000 + }, + { + "epoch": 0.89, + "grad_norm": 0.453125, + "learning_rate": 0.0001167971487344065, + "loss": 2.5528, + "step": 21001 + }, + { + "epoch": 0.89, + "grad_norm": 0.478515625, + "learning_rate": 0.00011679053607235964, + "loss": 2.5394, + "step": 21002 + }, + { + "epoch": 0.89, + "grad_norm": 0.435546875, + "learning_rate": 0.00011678392333476151, + "loss": 2.6535, + "step": 21003 + }, + { + "epoch": 0.9, + "grad_norm": 0.5234375, + "learning_rate": 0.00011677731052164187, + "loss": 2.6317, + "step": 21004 + }, + { + "epoch": 0.9, + "grad_norm": 0.43359375, + "learning_rate": 0.00011677069763303044, + "loss": 2.4976, + "step": 21005 + }, + { + "epoch": 0.9, + "grad_norm": 0.435546875, + "learning_rate": 0.00011676408466895698, + "loss": 2.5584, + "step": 21006 + }, + { + "epoch": 0.9, + "grad_norm": 0.4375, + "learning_rate": 0.00011675747162945132, + "loss": 2.7556, + "step": 21007 + }, + { + "epoch": 0.9, + "grad_norm": 0.50390625, + "learning_rate": 0.00011675085851454313, + "loss": 2.4503, + "step": 21008 + }, + { + "epoch": 0.9, + "grad_norm": 0.490234375, + "learning_rate": 0.0001167442453242622, + "loss": 2.5079, + "step": 21009 + }, + { + "epoch": 0.9, + "grad_norm": 0.427734375, + "learning_rate": 0.0001167376320586383, + "loss": 2.6594, + "step": 21010 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011673101871770115, + "loss": 2.499, + "step": 21011 + }, + { + "epoch": 0.9, + "grad_norm": 0.4375, + "learning_rate": 0.00011672440530148057, + "loss": 2.4712, + "step": 21012 + }, + { + "epoch": 0.9, + "grad_norm": 0.46484375, + "learning_rate": 0.00011671779181000623, + "loss": 2.5419, + "step": 21013 + }, + { + "epoch": 0.9, + "grad_norm": 0.4296875, + "learning_rate": 0.00011671117824330795, + "loss": 2.5105, + "step": 21014 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011670456460141552, + "loss": 2.5799, + "step": 21015 + }, + { + "epoch": 0.9, + "grad_norm": 0.458984375, + "learning_rate": 0.0001166979508843586, + "loss": 2.5843, + "step": 21016 + }, + { + "epoch": 0.9, + "grad_norm": 0.482421875, + "learning_rate": 0.00011669133709216702, + "loss": 2.6132, + "step": 21017 + }, + { + "epoch": 0.9, + "grad_norm": 0.412109375, + "learning_rate": 0.00011668472322487052, + "loss": 2.456, + "step": 21018 + }, + { + "epoch": 0.9, + "grad_norm": 0.416015625, + "learning_rate": 0.00011667810928249887, + "loss": 2.576, + "step": 21019 + }, + { + "epoch": 0.9, + "grad_norm": 0.470703125, + "learning_rate": 0.00011667149526508183, + "loss": 2.5893, + "step": 21020 + }, + { + "epoch": 0.9, + "grad_norm": 0.494140625, + "learning_rate": 0.00011666488117264914, + "loss": 2.5929, + "step": 21021 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011665826700523057, + "loss": 2.5798, + "step": 21022 + }, + { + "epoch": 0.9, + "grad_norm": 0.458984375, + "learning_rate": 0.00011665165276285593, + "loss": 2.5594, + "step": 21023 + }, + { + "epoch": 0.9, + "grad_norm": 0.435546875, + "learning_rate": 0.00011664503844555488, + "loss": 2.4115, + "step": 21024 + }, + { + "epoch": 0.9, + "grad_norm": 0.447265625, + "learning_rate": 0.00011663842405335725, + "loss": 2.5431, + "step": 21025 + }, + { + "epoch": 0.9, + "grad_norm": 0.40625, + "learning_rate": 0.00011663180958629281, + "loss": 2.526, + "step": 21026 + }, + { + "epoch": 0.9, + "grad_norm": 0.46484375, + "learning_rate": 0.00011662519504439131, + "loss": 2.5644, + "step": 21027 + }, + { + "epoch": 0.9, + "grad_norm": 0.41015625, + "learning_rate": 0.0001166185804276825, + "loss": 2.526, + "step": 21028 + }, + { + "epoch": 0.9, + "grad_norm": 0.4765625, + "learning_rate": 0.00011661196573619614, + "loss": 2.5062, + "step": 21029 + }, + { + "epoch": 0.9, + "grad_norm": 0.47265625, + "learning_rate": 0.00011660535096996201, + "loss": 2.7427, + "step": 21030 + }, + { + "epoch": 0.9, + "grad_norm": 0.439453125, + "learning_rate": 0.0001165987361290099, + "loss": 2.5304, + "step": 21031 + }, + { + "epoch": 0.9, + "grad_norm": 0.5, + "learning_rate": 0.00011659212121336951, + "loss": 2.6284, + "step": 21032 + }, + { + "epoch": 0.9, + "grad_norm": 0.48046875, + "learning_rate": 0.00011658550622307065, + "loss": 2.5925, + "step": 21033 + }, + { + "epoch": 0.9, + "grad_norm": 0.466796875, + "learning_rate": 0.00011657889115814307, + "loss": 2.6423, + "step": 21034 + }, + { + "epoch": 0.9, + "grad_norm": 0.412109375, + "learning_rate": 0.00011657227601861653, + "loss": 2.5021, + "step": 21035 + }, + { + "epoch": 0.9, + "grad_norm": 0.478515625, + "learning_rate": 0.00011656566080452082, + "loss": 2.6036, + "step": 21036 + }, + { + "epoch": 0.9, + "grad_norm": 0.42578125, + "learning_rate": 0.00011655904551588568, + "loss": 2.5318, + "step": 21037 + }, + { + "epoch": 0.9, + "grad_norm": 0.43359375, + "learning_rate": 0.00011655243015274089, + "loss": 2.6067, + "step": 21038 + }, + { + "epoch": 0.9, + "grad_norm": 0.478515625, + "learning_rate": 0.00011654581471511623, + "loss": 2.5748, + "step": 21039 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011653919920304142, + "loss": 2.5011, + "step": 21040 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.00011653258361654626, + "loss": 2.4805, + "step": 21041 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011652596795566056, + "loss": 2.5259, + "step": 21042 + }, + { + "epoch": 0.9, + "grad_norm": 0.423828125, + "learning_rate": 0.000116519352220414, + "loss": 2.5037, + "step": 21043 + }, + { + "epoch": 0.9, + "grad_norm": 0.44921875, + "learning_rate": 0.0001165127364108364, + "loss": 2.547, + "step": 21044 + }, + { + "epoch": 0.9, + "grad_norm": 0.482421875, + "learning_rate": 0.00011650612052695753, + "loss": 2.5751, + "step": 21045 + }, + { + "epoch": 0.9, + "grad_norm": 0.4609375, + "learning_rate": 0.00011649950456880714, + "loss": 2.5056, + "step": 21046 + }, + { + "epoch": 0.9, + "grad_norm": 0.490234375, + "learning_rate": 0.00011649288853641503, + "loss": 2.6456, + "step": 21047 + }, + { + "epoch": 0.9, + "grad_norm": 0.4140625, + "learning_rate": 0.00011648627242981091, + "loss": 2.6613, + "step": 21048 + }, + { + "epoch": 0.9, + "grad_norm": 0.466796875, + "learning_rate": 0.00011647965624902461, + "loss": 2.5503, + "step": 21049 + }, + { + "epoch": 0.9, + "grad_norm": 0.416015625, + "learning_rate": 0.00011647303999408589, + "loss": 2.6147, + "step": 21050 + }, + { + "epoch": 0.9, + "grad_norm": 0.470703125, + "learning_rate": 0.00011646642366502451, + "loss": 2.5248, + "step": 21051 + }, + { + "epoch": 0.9, + "grad_norm": 0.435546875, + "learning_rate": 0.00011645980726187024, + "loss": 2.487, + "step": 21052 + }, + { + "epoch": 0.9, + "grad_norm": 0.427734375, + "learning_rate": 0.00011645319078465284, + "loss": 2.4847, + "step": 21053 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.00011644657423340208, + "loss": 2.43, + "step": 21054 + }, + { + "epoch": 0.9, + "grad_norm": 0.5, + "learning_rate": 0.00011643995760814781, + "loss": 2.5814, + "step": 21055 + }, + { + "epoch": 0.9, + "grad_norm": 0.466796875, + "learning_rate": 0.00011643334090891967, + "loss": 2.4163, + "step": 21056 + }, + { + "epoch": 0.9, + "grad_norm": 0.478515625, + "learning_rate": 0.00011642672413574754, + "loss": 2.5144, + "step": 21057 + }, + { + "epoch": 0.9, + "grad_norm": 0.412109375, + "learning_rate": 0.00011642010728866112, + "loss": 2.55, + "step": 21058 + }, + { + "epoch": 0.9, + "grad_norm": 0.439453125, + "learning_rate": 0.00011641349036769025, + "loss": 2.4434, + "step": 21059 + }, + { + "epoch": 0.9, + "grad_norm": 0.435546875, + "learning_rate": 0.00011640687337286467, + "loss": 2.503, + "step": 21060 + }, + { + "epoch": 0.9, + "grad_norm": 0.419921875, + "learning_rate": 0.00011640025630421416, + "loss": 2.5632, + "step": 21061 + }, + { + "epoch": 0.9, + "grad_norm": 0.5078125, + "learning_rate": 0.00011639363916176847, + "loss": 2.5621, + "step": 21062 + }, + { + "epoch": 0.9, + "grad_norm": 0.435546875, + "learning_rate": 0.00011638702194555741, + "loss": 2.5273, + "step": 21063 + }, + { + "epoch": 0.9, + "grad_norm": 0.443359375, + "learning_rate": 0.0001163804046556107, + "loss": 2.5718, + "step": 21064 + }, + { + "epoch": 0.9, + "grad_norm": 0.412109375, + "learning_rate": 0.00011637378729195821, + "loss": 2.5013, + "step": 21065 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.00011636716985462963, + "loss": 2.6582, + "step": 21066 + }, + { + "epoch": 0.9, + "grad_norm": 0.412109375, + "learning_rate": 0.0001163605523436548, + "loss": 2.629, + "step": 21067 + }, + { + "epoch": 0.9, + "grad_norm": 0.455078125, + "learning_rate": 0.00011635393475906345, + "loss": 2.5332, + "step": 21068 + }, + { + "epoch": 0.9, + "grad_norm": 0.4921875, + "learning_rate": 0.00011634731710088536, + "loss": 2.6586, + "step": 21069 + }, + { + "epoch": 0.9, + "grad_norm": 0.44921875, + "learning_rate": 0.00011634069936915028, + "loss": 2.4252, + "step": 21070 + }, + { + "epoch": 0.9, + "grad_norm": 0.53515625, + "learning_rate": 0.00011633408156388809, + "loss": 2.6062, + "step": 21071 + }, + { + "epoch": 0.9, + "grad_norm": 0.47265625, + "learning_rate": 0.00011632746368512846, + "loss": 2.4854, + "step": 21072 + }, + { + "epoch": 0.9, + "grad_norm": 0.455078125, + "learning_rate": 0.00011632084573290123, + "loss": 2.5852, + "step": 21073 + }, + { + "epoch": 0.9, + "grad_norm": 0.466796875, + "learning_rate": 0.00011631422770723616, + "loss": 2.5968, + "step": 21074 + }, + { + "epoch": 0.9, + "grad_norm": 0.4296875, + "learning_rate": 0.00011630760960816303, + "loss": 2.5793, + "step": 21075 + }, + { + "epoch": 0.9, + "grad_norm": 0.408203125, + "learning_rate": 0.0001163009914357116, + "loss": 2.5835, + "step": 21076 + }, + { + "epoch": 0.9, + "grad_norm": 0.451171875, + "learning_rate": 0.00011629437318991168, + "loss": 2.5544, + "step": 21077 + }, + { + "epoch": 0.9, + "grad_norm": 0.5234375, + "learning_rate": 0.00011628775487079302, + "loss": 2.5357, + "step": 21078 + }, + { + "epoch": 0.9, + "grad_norm": 0.4609375, + "learning_rate": 0.00011628113647838543, + "loss": 2.5383, + "step": 21079 + }, + { + "epoch": 0.9, + "grad_norm": 0.5234375, + "learning_rate": 0.00011627451801271868, + "loss": 2.5507, + "step": 21080 + }, + { + "epoch": 0.9, + "grad_norm": 0.4296875, + "learning_rate": 0.00011626789947382252, + "loss": 2.465, + "step": 21081 + }, + { + "epoch": 0.9, + "grad_norm": 0.451171875, + "learning_rate": 0.00011626128086172678, + "loss": 2.5832, + "step": 21082 + }, + { + "epoch": 0.9, + "grad_norm": 0.431640625, + "learning_rate": 0.00011625466217646122, + "loss": 2.5269, + "step": 21083 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011624804341805563, + "loss": 2.4924, + "step": 21084 + }, + { + "epoch": 0.9, + "grad_norm": 0.466796875, + "learning_rate": 0.00011624142458653976, + "loss": 2.6384, + "step": 21085 + }, + { + "epoch": 0.9, + "grad_norm": 0.443359375, + "learning_rate": 0.0001162348056819434, + "loss": 2.5061, + "step": 21086 + }, + { + "epoch": 0.9, + "grad_norm": 0.515625, + "learning_rate": 0.0001162281867042964, + "loss": 2.6172, + "step": 21087 + }, + { + "epoch": 0.9, + "grad_norm": 0.427734375, + "learning_rate": 0.00011622156765362843, + "loss": 2.3924, + "step": 21088 + }, + { + "epoch": 0.9, + "grad_norm": 0.419921875, + "learning_rate": 0.00011621494852996937, + "loss": 2.5211, + "step": 21089 + }, + { + "epoch": 0.9, + "grad_norm": 0.412109375, + "learning_rate": 0.00011620832933334897, + "loss": 2.5093, + "step": 21090 + }, + { + "epoch": 0.9, + "grad_norm": 0.498046875, + "learning_rate": 0.00011620171006379701, + "loss": 2.5289, + "step": 21091 + }, + { + "epoch": 0.9, + "grad_norm": 0.4453125, + "learning_rate": 0.00011619509072134327, + "loss": 2.5352, + "step": 21092 + }, + { + "epoch": 0.9, + "grad_norm": 0.396484375, + "learning_rate": 0.00011618847130601755, + "loss": 2.5759, + "step": 21093 + }, + { + "epoch": 0.9, + "grad_norm": 0.419921875, + "learning_rate": 0.00011618185181784962, + "loss": 2.4964, + "step": 21094 + }, + { + "epoch": 0.9, + "grad_norm": 0.47265625, + "learning_rate": 0.00011617523225686926, + "loss": 2.6399, + "step": 21095 + }, + { + "epoch": 0.9, + "grad_norm": 0.419921875, + "learning_rate": 0.00011616861262310627, + "loss": 2.5234, + "step": 21096 + }, + { + "epoch": 0.9, + "grad_norm": 0.462890625, + "learning_rate": 0.00011616199291659042, + "loss": 2.6529, + "step": 21097 + }, + { + "epoch": 0.9, + "grad_norm": 0.427734375, + "learning_rate": 0.00011615537313735154, + "loss": 2.689, + "step": 21098 + }, + { + "epoch": 0.9, + "grad_norm": 0.470703125, + "learning_rate": 0.00011614875328541937, + "loss": 2.63, + "step": 21099 + }, + { + "epoch": 0.9, + "grad_norm": 0.4453125, + "learning_rate": 0.00011614213336082369, + "loss": 2.4781, + "step": 21100 + }, + { + "epoch": 0.9, + "grad_norm": 0.48046875, + "learning_rate": 0.00011613551336359434, + "loss": 2.6606, + "step": 21101 + }, + { + "epoch": 0.9, + "grad_norm": 0.4375, + "learning_rate": 0.00011612889329376106, + "loss": 2.6082, + "step": 21102 + }, + { + "epoch": 0.9, + "grad_norm": 0.462890625, + "learning_rate": 0.00011612227315135367, + "loss": 2.5475, + "step": 21103 + }, + { + "epoch": 0.9, + "grad_norm": 0.431640625, + "learning_rate": 0.00011611565293640194, + "loss": 2.5665, + "step": 21104 + }, + { + "epoch": 0.9, + "grad_norm": 0.443359375, + "learning_rate": 0.00011610903264893562, + "loss": 2.6195, + "step": 21105 + }, + { + "epoch": 0.9, + "grad_norm": 0.42578125, + "learning_rate": 0.00011610241228898462, + "loss": 2.4673, + "step": 21106 + }, + { + "epoch": 0.9, + "grad_norm": 0.4921875, + "learning_rate": 0.00011609579185657858, + "loss": 2.5871, + "step": 21107 + }, + { + "epoch": 0.9, + "grad_norm": 0.486328125, + "learning_rate": 0.0001160891713517474, + "loss": 2.5034, + "step": 21108 + }, + { + "epoch": 0.9, + "grad_norm": 0.462890625, + "learning_rate": 0.00011608255077452079, + "loss": 2.538, + "step": 21109 + }, + { + "epoch": 0.9, + "grad_norm": 0.40234375, + "learning_rate": 0.0001160759301249286, + "loss": 2.5241, + "step": 21110 + }, + { + "epoch": 0.9, + "grad_norm": 0.458984375, + "learning_rate": 0.00011606930940300063, + "loss": 2.5092, + "step": 21111 + }, + { + "epoch": 0.9, + "grad_norm": 0.423828125, + "learning_rate": 0.0001160626886087666, + "loss": 2.4965, + "step": 21112 + }, + { + "epoch": 0.9, + "grad_norm": 0.462890625, + "learning_rate": 0.00011605606774225632, + "loss": 2.4324, + "step": 21113 + }, + { + "epoch": 0.9, + "grad_norm": 0.458984375, + "learning_rate": 0.00011604944680349968, + "loss": 2.4905, + "step": 21114 + }, + { + "epoch": 0.9, + "grad_norm": 0.4609375, + "learning_rate": 0.00011604282579252632, + "loss": 2.4502, + "step": 21115 + }, + { + "epoch": 0.9, + "grad_norm": 0.498046875, + "learning_rate": 0.00011603620470936614, + "loss": 2.6791, + "step": 21116 + }, + { + "epoch": 0.9, + "grad_norm": 0.419921875, + "learning_rate": 0.0001160295835540489, + "loss": 2.558, + "step": 21117 + }, + { + "epoch": 0.9, + "grad_norm": 0.4453125, + "learning_rate": 0.00011602296232660438, + "loss": 2.6186, + "step": 21118 + }, + { + "epoch": 0.9, + "grad_norm": 0.4453125, + "learning_rate": 0.0001160163410270624, + "loss": 2.6235, + "step": 21119 + }, + { + "epoch": 0.9, + "grad_norm": 0.443359375, + "learning_rate": 0.00011600971965545274, + "loss": 2.5387, + "step": 21120 + }, + { + "epoch": 0.9, + "grad_norm": 0.447265625, + "learning_rate": 0.00011600309821180518, + "loss": 2.5488, + "step": 21121 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011599647669614955, + "loss": 2.5161, + "step": 21122 + }, + { + "epoch": 0.9, + "grad_norm": 0.451171875, + "learning_rate": 0.00011598985510851559, + "loss": 2.4582, + "step": 21123 + }, + { + "epoch": 0.9, + "grad_norm": 0.453125, + "learning_rate": 0.00011598323344893312, + "loss": 2.7141, + "step": 21124 + }, + { + "epoch": 0.9, + "grad_norm": 0.423828125, + "learning_rate": 0.00011597661171743197, + "loss": 2.6313, + "step": 21125 + }, + { + "epoch": 0.9, + "grad_norm": 0.427734375, + "learning_rate": 0.00011596998991404189, + "loss": 2.5489, + "step": 21126 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011596336803879268, + "loss": 2.4838, + "step": 21127 + }, + { + "epoch": 0.9, + "grad_norm": 0.46875, + "learning_rate": 0.00011595674609171418, + "loss": 2.4724, + "step": 21128 + }, + { + "epoch": 0.9, + "grad_norm": 0.4453125, + "learning_rate": 0.00011595012407283614, + "loss": 2.508, + "step": 21129 + }, + { + "epoch": 0.9, + "grad_norm": 0.453125, + "learning_rate": 0.0001159435019821884, + "loss": 2.7537, + "step": 21130 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.00011593687981980066, + "loss": 2.5633, + "step": 21131 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.00011593025758570282, + "loss": 2.5618, + "step": 21132 + }, + { + "epoch": 0.9, + "grad_norm": 0.458984375, + "learning_rate": 0.00011592363527992465, + "loss": 2.4586, + "step": 21133 + }, + { + "epoch": 0.9, + "grad_norm": 0.4296875, + "learning_rate": 0.00011591701290249595, + "loss": 2.5805, + "step": 21134 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011591039045344648, + "loss": 2.6463, + "step": 21135 + }, + { + "epoch": 0.9, + "grad_norm": 0.466796875, + "learning_rate": 0.00011590376793280608, + "loss": 2.5615, + "step": 21136 + }, + { + "epoch": 0.9, + "grad_norm": 0.490234375, + "learning_rate": 0.00011589714534060455, + "loss": 2.5235, + "step": 21137 + }, + { + "epoch": 0.9, + "grad_norm": 0.478515625, + "learning_rate": 0.00011589052267687166, + "loss": 2.5684, + "step": 21138 + }, + { + "epoch": 0.9, + "grad_norm": 0.419921875, + "learning_rate": 0.00011588389994163722, + "loss": 2.4147, + "step": 21139 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011587727713493104, + "loss": 2.449, + "step": 21140 + }, + { + "epoch": 0.9, + "grad_norm": 0.439453125, + "learning_rate": 0.00011587065425678295, + "loss": 2.5397, + "step": 21141 + }, + { + "epoch": 0.9, + "grad_norm": 0.50390625, + "learning_rate": 0.00011586403130722267, + "loss": 2.4958, + "step": 21142 + }, + { + "epoch": 0.9, + "grad_norm": 0.478515625, + "learning_rate": 0.00011585740828628004, + "loss": 2.5385, + "step": 21143 + }, + { + "epoch": 0.9, + "grad_norm": 0.45703125, + "learning_rate": 0.00011585078519398492, + "loss": 2.5601, + "step": 21144 + }, + { + "epoch": 0.9, + "grad_norm": 0.53125, + "learning_rate": 0.00011584416203036701, + "loss": 2.7844, + "step": 21145 + }, + { + "epoch": 0.9, + "grad_norm": 0.45703125, + "learning_rate": 0.00011583753879545618, + "loss": 2.6378, + "step": 21146 + }, + { + "epoch": 0.9, + "grad_norm": 0.48046875, + "learning_rate": 0.00011583091548928222, + "loss": 2.5979, + "step": 21147 + }, + { + "epoch": 0.9, + "grad_norm": 0.455078125, + "learning_rate": 0.00011582429211187491, + "loss": 2.549, + "step": 21148 + }, + { + "epoch": 0.9, + "grad_norm": 0.42578125, + "learning_rate": 0.00011581766866326409, + "loss": 2.5808, + "step": 21149 + }, + { + "epoch": 0.9, + "grad_norm": 0.423828125, + "learning_rate": 0.00011581104514347954, + "loss": 2.4073, + "step": 21150 + }, + { + "epoch": 0.9, + "grad_norm": 0.439453125, + "learning_rate": 0.00011580442155255102, + "loss": 2.55, + "step": 21151 + }, + { + "epoch": 0.9, + "grad_norm": 0.4296875, + "learning_rate": 0.00011579779789050844, + "loss": 2.4894, + "step": 21152 + }, + { + "epoch": 0.9, + "grad_norm": 0.46484375, + "learning_rate": 0.0001157911741573815, + "loss": 2.465, + "step": 21153 + }, + { + "epoch": 0.9, + "grad_norm": 0.49609375, + "learning_rate": 0.00011578455035320008, + "loss": 2.5884, + "step": 21154 + }, + { + "epoch": 0.9, + "grad_norm": 0.5, + "learning_rate": 0.00011577792647799392, + "loss": 2.6301, + "step": 21155 + }, + { + "epoch": 0.9, + "grad_norm": 0.4375, + "learning_rate": 0.00011577130253179288, + "loss": 2.4172, + "step": 21156 + }, + { + "epoch": 0.9, + "grad_norm": 0.482421875, + "learning_rate": 0.00011576467851462675, + "loss": 2.5832, + "step": 21157 + }, + { + "epoch": 0.9, + "grad_norm": 0.44921875, + "learning_rate": 0.00011575805442652532, + "loss": 2.5471, + "step": 21158 + }, + { + "epoch": 0.9, + "grad_norm": 0.44921875, + "learning_rate": 0.00011575143026751837, + "loss": 2.5289, + "step": 21159 + }, + { + "epoch": 0.9, + "grad_norm": 0.443359375, + "learning_rate": 0.0001157448060376358, + "loss": 2.5468, + "step": 21160 + }, + { + "epoch": 0.9, + "grad_norm": 0.427734375, + "learning_rate": 0.00011573818173690731, + "loss": 2.5618, + "step": 21161 + }, + { + "epoch": 0.9, + "grad_norm": 0.45703125, + "learning_rate": 0.00011573155736536276, + "loss": 2.6901, + "step": 21162 + }, + { + "epoch": 0.9, + "grad_norm": 0.45703125, + "learning_rate": 0.00011572493292303196, + "loss": 2.4892, + "step": 21163 + }, + { + "epoch": 0.9, + "grad_norm": 0.51171875, + "learning_rate": 0.00011571830840994474, + "loss": 2.6975, + "step": 21164 + }, + { + "epoch": 0.9, + "grad_norm": 0.431640625, + "learning_rate": 0.00011571168382613083, + "loss": 2.5196, + "step": 21165 + }, + { + "epoch": 0.9, + "grad_norm": 0.427734375, + "learning_rate": 0.00011570505917162011, + "loss": 2.5936, + "step": 21166 + }, + { + "epoch": 0.9, + "grad_norm": 0.431640625, + "learning_rate": 0.00011569843444644232, + "loss": 2.4748, + "step": 21167 + }, + { + "epoch": 0.9, + "grad_norm": 0.44921875, + "learning_rate": 0.00011569180965062739, + "loss": 2.6708, + "step": 21168 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011568518478420497, + "loss": 2.5839, + "step": 21169 + }, + { + "epoch": 0.9, + "grad_norm": 0.478515625, + "learning_rate": 0.000115678559847205, + "loss": 2.5285, + "step": 21170 + }, + { + "epoch": 0.9, + "grad_norm": 0.41796875, + "learning_rate": 0.00011567193483965723, + "loss": 2.5523, + "step": 21171 + }, + { + "epoch": 0.9, + "grad_norm": 0.498046875, + "learning_rate": 0.00011566530976159148, + "loss": 2.5322, + "step": 21172 + }, + { + "epoch": 0.9, + "grad_norm": 0.53515625, + "learning_rate": 0.00011565868461303755, + "loss": 2.4974, + "step": 21173 + }, + { + "epoch": 0.9, + "grad_norm": 0.51953125, + "learning_rate": 0.00011565205939402528, + "loss": 2.6734, + "step": 21174 + }, + { + "epoch": 0.9, + "grad_norm": 0.53125, + "learning_rate": 0.00011564543410458444, + "loss": 2.5698, + "step": 21175 + }, + { + "epoch": 0.9, + "grad_norm": 0.498046875, + "learning_rate": 0.00011563880874474491, + "loss": 2.5254, + "step": 21176 + }, + { + "epoch": 0.9, + "grad_norm": 0.4375, + "learning_rate": 0.0001156321833145364, + "loss": 2.4033, + "step": 21177 + }, + { + "epoch": 0.9, + "grad_norm": 0.478515625, + "learning_rate": 0.00011562555781398881, + "loss": 2.6179, + "step": 21178 + }, + { + "epoch": 0.9, + "grad_norm": 0.427734375, + "learning_rate": 0.00011561893224313192, + "loss": 2.4877, + "step": 21179 + }, + { + "epoch": 0.9, + "grad_norm": 0.38671875, + "learning_rate": 0.00011561230660199553, + "loss": 2.4829, + "step": 21180 + }, + { + "epoch": 0.9, + "grad_norm": 0.46875, + "learning_rate": 0.00011560568089060948, + "loss": 2.6257, + "step": 21181 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011559905510900354, + "loss": 2.5978, + "step": 21182 + }, + { + "epoch": 0.9, + "grad_norm": 0.455078125, + "learning_rate": 0.00011559242925720757, + "loss": 2.6184, + "step": 21183 + }, + { + "epoch": 0.9, + "grad_norm": 0.4296875, + "learning_rate": 0.0001155858033352514, + "loss": 2.5842, + "step": 21184 + }, + { + "epoch": 0.9, + "grad_norm": 0.50390625, + "learning_rate": 0.00011557917734316477, + "loss": 2.5134, + "step": 21185 + }, + { + "epoch": 0.9, + "grad_norm": 0.4296875, + "learning_rate": 0.00011557255128097754, + "loss": 2.4174, + "step": 21186 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011556592514871954, + "loss": 2.5591, + "step": 21187 + }, + { + "epoch": 0.9, + "grad_norm": 0.478515625, + "learning_rate": 0.00011555929894642055, + "loss": 2.4916, + "step": 21188 + }, + { + "epoch": 0.9, + "grad_norm": 0.55078125, + "learning_rate": 0.0001155526726741104, + "loss": 2.5203, + "step": 21189 + }, + { + "epoch": 0.9, + "grad_norm": 0.44921875, + "learning_rate": 0.00011554604633181891, + "loss": 2.6294, + "step": 21190 + }, + { + "epoch": 0.9, + "grad_norm": 0.431640625, + "learning_rate": 0.00011553941991957588, + "loss": 2.5322, + "step": 21191 + }, + { + "epoch": 0.9, + "grad_norm": 0.61328125, + "learning_rate": 0.00011553279343741117, + "loss": 2.5878, + "step": 21192 + }, + { + "epoch": 0.9, + "grad_norm": 0.494140625, + "learning_rate": 0.00011552616688535454, + "loss": 2.5981, + "step": 21193 + }, + { + "epoch": 0.9, + "grad_norm": 0.4609375, + "learning_rate": 0.00011551954026343585, + "loss": 2.7392, + "step": 21194 + }, + { + "epoch": 0.9, + "grad_norm": 0.462890625, + "learning_rate": 0.0001155129135716849, + "loss": 2.3227, + "step": 21195 + }, + { + "epoch": 0.9, + "grad_norm": 0.482421875, + "learning_rate": 0.00011550628681013149, + "loss": 2.6805, + "step": 21196 + }, + { + "epoch": 0.9, + "grad_norm": 0.453125, + "learning_rate": 0.00011549965997880548, + "loss": 2.5196, + "step": 21197 + }, + { + "epoch": 0.9, + "grad_norm": 0.447265625, + "learning_rate": 0.00011549303307773665, + "loss": 2.6397, + "step": 21198 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011548640610695483, + "loss": 2.4798, + "step": 21199 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011547977906648985, + "loss": 2.6195, + "step": 21200 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011547315195637151, + "loss": 2.5242, + "step": 21201 + }, + { + "epoch": 0.9, + "grad_norm": 0.44921875, + "learning_rate": 0.00011546652477662964, + "loss": 2.6106, + "step": 21202 + }, + { + "epoch": 0.9, + "grad_norm": 0.5078125, + "learning_rate": 0.0001154598975272941, + "loss": 2.5784, + "step": 21203 + }, + { + "epoch": 0.9, + "grad_norm": 0.41015625, + "learning_rate": 0.00011545327020839462, + "loss": 2.4709, + "step": 21204 + }, + { + "epoch": 0.9, + "grad_norm": 0.4765625, + "learning_rate": 0.00011544664281996111, + "loss": 2.5706, + "step": 21205 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.00011544001536202333, + "loss": 2.6345, + "step": 21206 + }, + { + "epoch": 0.9, + "grad_norm": 0.431640625, + "learning_rate": 0.00011543338783461111, + "loss": 2.4284, + "step": 21207 + }, + { + "epoch": 0.9, + "grad_norm": 0.51953125, + "learning_rate": 0.00011542676023775434, + "loss": 2.5462, + "step": 21208 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.00011542013257148273, + "loss": 2.6738, + "step": 21209 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.00011541350483582618, + "loss": 2.4763, + "step": 21210 + }, + { + "epoch": 0.9, + "grad_norm": 0.5078125, + "learning_rate": 0.0001154068770308145, + "loss": 2.5721, + "step": 21211 + }, + { + "epoch": 0.9, + "grad_norm": 0.4453125, + "learning_rate": 0.00011540024915647749, + "loss": 2.5078, + "step": 21212 + }, + { + "epoch": 0.9, + "grad_norm": 0.41015625, + "learning_rate": 0.00011539362121284501, + "loss": 2.5539, + "step": 21213 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011538699319994683, + "loss": 2.639, + "step": 21214 + }, + { + "epoch": 0.9, + "grad_norm": 0.470703125, + "learning_rate": 0.00011538036511781282, + "loss": 2.6564, + "step": 21215 + }, + { + "epoch": 0.9, + "grad_norm": 0.466796875, + "learning_rate": 0.00011537373696647277, + "loss": 2.6164, + "step": 21216 + }, + { + "epoch": 0.9, + "grad_norm": 0.4453125, + "learning_rate": 0.00011536710874595656, + "loss": 2.4462, + "step": 21217 + }, + { + "epoch": 0.9, + "grad_norm": 0.419921875, + "learning_rate": 0.00011536048045629392, + "loss": 2.5875, + "step": 21218 + }, + { + "epoch": 0.9, + "grad_norm": 0.48046875, + "learning_rate": 0.00011535385209751479, + "loss": 2.5828, + "step": 21219 + }, + { + "epoch": 0.9, + "grad_norm": 0.431640625, + "learning_rate": 0.0001153472236696489, + "loss": 2.4637, + "step": 21220 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011534059517272613, + "loss": 2.5367, + "step": 21221 + }, + { + "epoch": 0.9, + "grad_norm": 0.4765625, + "learning_rate": 0.00011533396660677628, + "loss": 2.5588, + "step": 21222 + }, + { + "epoch": 0.9, + "grad_norm": 0.4921875, + "learning_rate": 0.0001153273379718292, + "loss": 2.5701, + "step": 21223 + }, + { + "epoch": 0.9, + "grad_norm": 0.5, + "learning_rate": 0.00011532070926791468, + "loss": 2.4871, + "step": 21224 + }, + { + "epoch": 0.9, + "grad_norm": 0.443359375, + "learning_rate": 0.00011531408049506259, + "loss": 2.5526, + "step": 21225 + }, + { + "epoch": 0.9, + "grad_norm": 0.447265625, + "learning_rate": 0.00011530745165330267, + "loss": 2.4625, + "step": 21226 + }, + { + "epoch": 0.9, + "grad_norm": 0.443359375, + "learning_rate": 0.00011530082274266489, + "loss": 2.5672, + "step": 21227 + }, + { + "epoch": 0.9, + "grad_norm": 0.49609375, + "learning_rate": 0.00011529419376317896, + "loss": 2.6686, + "step": 21228 + }, + { + "epoch": 0.9, + "grad_norm": 0.431640625, + "learning_rate": 0.00011528756471487477, + "loss": 2.3903, + "step": 21229 + }, + { + "epoch": 0.9, + "grad_norm": 0.48828125, + "learning_rate": 0.0001152809355977821, + "loss": 2.4744, + "step": 21230 + }, + { + "epoch": 0.9, + "grad_norm": 0.484375, + "learning_rate": 0.00011527430641193082, + "loss": 2.5542, + "step": 21231 + }, + { + "epoch": 0.9, + "grad_norm": 0.48046875, + "learning_rate": 0.00011526767715735073, + "loss": 2.4829, + "step": 21232 + }, + { + "epoch": 0.9, + "grad_norm": 0.447265625, + "learning_rate": 0.0001152610478340717, + "loss": 2.5484, + "step": 21233 + }, + { + "epoch": 0.9, + "grad_norm": 0.474609375, + "learning_rate": 0.0001152544184421235, + "loss": 2.6385, + "step": 21234 + }, + { + "epoch": 0.9, + "grad_norm": 0.490234375, + "learning_rate": 0.00011524778898153603, + "loss": 2.5626, + "step": 21235 + }, + { + "epoch": 0.9, + "grad_norm": 0.44140625, + "learning_rate": 0.00011524115945233904, + "loss": 2.5301, + "step": 21236 + }, + { + "epoch": 0.9, + "grad_norm": 0.5546875, + "learning_rate": 0.00011523452985456242, + "loss": 2.7464, + "step": 21237 + }, + { + "epoch": 0.91, + "grad_norm": 0.462890625, + "learning_rate": 0.00011522790018823599, + "loss": 2.5543, + "step": 21238 + }, + { + "epoch": 0.91, + "grad_norm": 0.421875, + "learning_rate": 0.00011522127045338957, + "loss": 2.391, + "step": 21239 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011521464065005301, + "loss": 2.5019, + "step": 21240 + }, + { + "epoch": 0.91, + "grad_norm": 0.453125, + "learning_rate": 0.0001152080107782561, + "loss": 2.5966, + "step": 21241 + }, + { + "epoch": 0.91, + "grad_norm": 0.51953125, + "learning_rate": 0.0001152013808380287, + "loss": 2.6668, + "step": 21242 + }, + { + "epoch": 0.91, + "grad_norm": 0.50390625, + "learning_rate": 0.00011519475082940068, + "loss": 2.6178, + "step": 21243 + }, + { + "epoch": 0.91, + "grad_norm": 0.42578125, + "learning_rate": 0.00011518812075240178, + "loss": 2.4728, + "step": 21244 + }, + { + "epoch": 0.91, + "grad_norm": 0.412109375, + "learning_rate": 0.00011518149060706191, + "loss": 2.5728, + "step": 21245 + }, + { + "epoch": 0.91, + "grad_norm": 0.4296875, + "learning_rate": 0.0001151748603934109, + "loss": 2.4376, + "step": 21246 + }, + { + "epoch": 0.91, + "grad_norm": 0.458984375, + "learning_rate": 0.00011516823011147854, + "loss": 2.4203, + "step": 21247 + }, + { + "epoch": 0.91, + "grad_norm": 0.478515625, + "learning_rate": 0.0001151615997612947, + "loss": 2.3885, + "step": 21248 + }, + { + "epoch": 0.91, + "grad_norm": 0.43359375, + "learning_rate": 0.0001151549693428892, + "loss": 2.5388, + "step": 21249 + }, + { + "epoch": 0.91, + "grad_norm": 0.51171875, + "learning_rate": 0.00011514833885629187, + "loss": 2.646, + "step": 21250 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011514170830153257, + "loss": 2.4624, + "step": 21251 + }, + { + "epoch": 0.91, + "grad_norm": 0.44921875, + "learning_rate": 0.0001151350776786411, + "loss": 2.593, + "step": 21252 + }, + { + "epoch": 0.91, + "grad_norm": 0.462890625, + "learning_rate": 0.00011512844698764731, + "loss": 2.5284, + "step": 21253 + }, + { + "epoch": 0.91, + "grad_norm": 0.482421875, + "learning_rate": 0.00011512181622858105, + "loss": 2.6784, + "step": 21254 + }, + { + "epoch": 0.91, + "grad_norm": 0.455078125, + "learning_rate": 0.00011511518540147213, + "loss": 2.7331, + "step": 21255 + }, + { + "epoch": 0.91, + "grad_norm": 0.482421875, + "learning_rate": 0.0001151085545063504, + "loss": 2.5418, + "step": 21256 + }, + { + "epoch": 0.91, + "grad_norm": 0.47265625, + "learning_rate": 0.00011510192354324571, + "loss": 2.5453, + "step": 21257 + }, + { + "epoch": 0.91, + "grad_norm": 0.404296875, + "learning_rate": 0.00011509529251218785, + "loss": 2.4214, + "step": 21258 + }, + { + "epoch": 0.91, + "grad_norm": 0.4765625, + "learning_rate": 0.00011508866141320674, + "loss": 2.5983, + "step": 21259 + }, + { + "epoch": 0.91, + "grad_norm": 0.41796875, + "learning_rate": 0.00011508203024633213, + "loss": 2.6321, + "step": 21260 + }, + { + "epoch": 0.91, + "grad_norm": 0.458984375, + "learning_rate": 0.0001150753990115939, + "loss": 2.4766, + "step": 21261 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.00011506876770902191, + "loss": 2.5469, + "step": 21262 + }, + { + "epoch": 0.91, + "grad_norm": 0.478515625, + "learning_rate": 0.00011506213633864596, + "loss": 2.5339, + "step": 21263 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.0001150555049004959, + "loss": 2.5455, + "step": 21264 + }, + { + "epoch": 0.91, + "grad_norm": 0.54296875, + "learning_rate": 0.00011504887339460156, + "loss": 2.6198, + "step": 21265 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011504224182099277, + "loss": 2.5072, + "step": 21266 + }, + { + "epoch": 0.91, + "grad_norm": 0.470703125, + "learning_rate": 0.00011503561017969944, + "loss": 2.5068, + "step": 21267 + }, + { + "epoch": 0.91, + "grad_norm": 0.4765625, + "learning_rate": 0.00011502897847075132, + "loss": 2.7044, + "step": 21268 + }, + { + "epoch": 0.91, + "grad_norm": 0.427734375, + "learning_rate": 0.0001150223466941783, + "loss": 2.5232, + "step": 21269 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011501571485001021, + "loss": 2.5156, + "step": 21270 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.0001150090829382769, + "loss": 2.4483, + "step": 21271 + }, + { + "epoch": 0.91, + "grad_norm": 0.4375, + "learning_rate": 0.00011500245095900821, + "loss": 2.5115, + "step": 21272 + }, + { + "epoch": 0.91, + "grad_norm": 0.404296875, + "learning_rate": 0.00011499581891223394, + "loss": 2.5148, + "step": 21273 + }, + { + "epoch": 0.91, + "grad_norm": 0.439453125, + "learning_rate": 0.00011498918679798399, + "loss": 2.5564, + "step": 21274 + }, + { + "epoch": 0.91, + "grad_norm": 0.478515625, + "learning_rate": 0.0001149825546162882, + "loss": 2.5382, + "step": 21275 + }, + { + "epoch": 0.91, + "grad_norm": 0.453125, + "learning_rate": 0.00011497592236717633, + "loss": 2.412, + "step": 21276 + }, + { + "epoch": 0.91, + "grad_norm": 0.4609375, + "learning_rate": 0.00011496929005067832, + "loss": 2.4175, + "step": 21277 + }, + { + "epoch": 0.91, + "grad_norm": 0.47265625, + "learning_rate": 0.00011496265766682397, + "loss": 2.5584, + "step": 21278 + }, + { + "epoch": 0.91, + "grad_norm": 0.408203125, + "learning_rate": 0.00011495602521564312, + "loss": 2.5657, + "step": 21279 + }, + { + "epoch": 0.91, + "grad_norm": 0.40234375, + "learning_rate": 0.00011494939269716565, + "loss": 2.4004, + "step": 21280 + }, + { + "epoch": 0.91, + "grad_norm": 0.40625, + "learning_rate": 0.00011494276011142135, + "loss": 2.6626, + "step": 21281 + }, + { + "epoch": 0.91, + "grad_norm": 0.486328125, + "learning_rate": 0.00011493612745844007, + "loss": 2.6111, + "step": 21282 + }, + { + "epoch": 0.91, + "grad_norm": 0.51953125, + "learning_rate": 0.00011492949473825172, + "loss": 2.4809, + "step": 21283 + }, + { + "epoch": 0.91, + "grad_norm": 0.4375, + "learning_rate": 0.00011492286195088607, + "loss": 2.5607, + "step": 21284 + }, + { + "epoch": 0.91, + "grad_norm": 0.453125, + "learning_rate": 0.00011491622909637301, + "loss": 2.5527, + "step": 21285 + }, + { + "epoch": 0.91, + "grad_norm": 0.451171875, + "learning_rate": 0.00011490959617474238, + "loss": 2.6538, + "step": 21286 + }, + { + "epoch": 0.91, + "grad_norm": 0.5078125, + "learning_rate": 0.000114902963186024, + "loss": 2.6802, + "step": 21287 + }, + { + "epoch": 0.91, + "grad_norm": 0.421875, + "learning_rate": 0.00011489633013024774, + "loss": 2.6156, + "step": 21288 + }, + { + "epoch": 0.91, + "grad_norm": 0.3984375, + "learning_rate": 0.00011488969700744342, + "loss": 2.4227, + "step": 21289 + }, + { + "epoch": 0.91, + "grad_norm": 0.47265625, + "learning_rate": 0.00011488306381764091, + "loss": 2.5215, + "step": 21290 + }, + { + "epoch": 0.91, + "grad_norm": 0.466796875, + "learning_rate": 0.0001148764305608701, + "loss": 2.6564, + "step": 21291 + }, + { + "epoch": 0.91, + "grad_norm": 0.5, + "learning_rate": 0.00011486979723716072, + "loss": 2.4888, + "step": 21292 + }, + { + "epoch": 0.91, + "grad_norm": 0.412109375, + "learning_rate": 0.00011486316384654273, + "loss": 2.6244, + "step": 21293 + }, + { + "epoch": 0.91, + "grad_norm": 0.431640625, + "learning_rate": 0.00011485653038904591, + "loss": 2.5197, + "step": 21294 + }, + { + "epoch": 0.91, + "grad_norm": 0.5, + "learning_rate": 0.00011484989686470014, + "loss": 2.6164, + "step": 21295 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.00011484326327353527, + "loss": 2.5845, + "step": 21296 + }, + { + "epoch": 0.91, + "grad_norm": 0.44921875, + "learning_rate": 0.00011483662961558113, + "loss": 2.4361, + "step": 21297 + }, + { + "epoch": 0.91, + "grad_norm": 0.451171875, + "learning_rate": 0.00011482999589086758, + "loss": 2.5784, + "step": 21298 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011482336209942448, + "loss": 2.6351, + "step": 21299 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011481672824128164, + "loss": 2.4271, + "step": 21300 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011481009431646896, + "loss": 2.4315, + "step": 21301 + }, + { + "epoch": 0.91, + "grad_norm": 0.43359375, + "learning_rate": 0.00011480346032501628, + "loss": 2.5081, + "step": 21302 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011479682626695343, + "loss": 2.4867, + "step": 21303 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011479019214231026, + "loss": 2.4127, + "step": 21304 + }, + { + "epoch": 0.91, + "grad_norm": 0.451171875, + "learning_rate": 0.00011478355795111664, + "loss": 2.5852, + "step": 21305 + }, + { + "epoch": 0.91, + "grad_norm": 0.45703125, + "learning_rate": 0.00011477692369340238, + "loss": 2.5453, + "step": 21306 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.00011477028936919743, + "loss": 2.474, + "step": 21307 + }, + { + "epoch": 0.91, + "grad_norm": 0.490234375, + "learning_rate": 0.00011476365497853149, + "loss": 2.5532, + "step": 21308 + }, + { + "epoch": 0.91, + "grad_norm": 0.419921875, + "learning_rate": 0.00011475702052143455, + "loss": 2.5021, + "step": 21309 + }, + { + "epoch": 0.91, + "grad_norm": 0.5078125, + "learning_rate": 0.00011475038599793641, + "loss": 2.4957, + "step": 21310 + }, + { + "epoch": 0.91, + "grad_norm": 0.55078125, + "learning_rate": 0.00011474375140806692, + "loss": 2.7068, + "step": 21311 + }, + { + "epoch": 0.91, + "grad_norm": 0.46875, + "learning_rate": 0.00011473711675185594, + "loss": 2.5447, + "step": 21312 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.00011473048202933331, + "loss": 2.5539, + "step": 21313 + }, + { + "epoch": 0.91, + "grad_norm": 0.482421875, + "learning_rate": 0.00011472384724052889, + "loss": 2.5708, + "step": 21314 + }, + { + "epoch": 0.91, + "grad_norm": 0.470703125, + "learning_rate": 0.00011471721238547258, + "loss": 2.7357, + "step": 21315 + }, + { + "epoch": 0.91, + "grad_norm": 0.46875, + "learning_rate": 0.00011471057746419413, + "loss": 2.5829, + "step": 21316 + }, + { + "epoch": 0.91, + "grad_norm": 0.49609375, + "learning_rate": 0.0001147039424767235, + "loss": 2.5351, + "step": 21317 + }, + { + "epoch": 0.91, + "grad_norm": 0.462890625, + "learning_rate": 0.00011469730742309049, + "loss": 2.571, + "step": 21318 + }, + { + "epoch": 0.91, + "grad_norm": 0.40625, + "learning_rate": 0.00011469067230332498, + "loss": 2.4621, + "step": 21319 + }, + { + "epoch": 0.91, + "grad_norm": 0.4765625, + "learning_rate": 0.00011468403711745679, + "loss": 2.5842, + "step": 21320 + }, + { + "epoch": 0.91, + "grad_norm": 0.48828125, + "learning_rate": 0.00011467740186551581, + "loss": 2.5853, + "step": 21321 + }, + { + "epoch": 0.91, + "grad_norm": 0.41796875, + "learning_rate": 0.00011467076654753188, + "loss": 2.313, + "step": 21322 + }, + { + "epoch": 0.91, + "grad_norm": 0.5078125, + "learning_rate": 0.00011466413116353489, + "loss": 2.5431, + "step": 21323 + }, + { + "epoch": 0.91, + "grad_norm": 0.466796875, + "learning_rate": 0.00011465749571355462, + "loss": 2.649, + "step": 21324 + }, + { + "epoch": 0.91, + "grad_norm": 0.435546875, + "learning_rate": 0.00011465086019762101, + "loss": 2.4609, + "step": 21325 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011464422461576388, + "loss": 2.5764, + "step": 21326 + }, + { + "epoch": 0.91, + "grad_norm": 0.478515625, + "learning_rate": 0.00011463758896801309, + "loss": 2.5481, + "step": 21327 + }, + { + "epoch": 0.91, + "grad_norm": 0.470703125, + "learning_rate": 0.0001146309532543985, + "loss": 2.5273, + "step": 21328 + }, + { + "epoch": 0.91, + "grad_norm": 0.5, + "learning_rate": 0.00011462431747494997, + "loss": 2.5521, + "step": 21329 + }, + { + "epoch": 0.91, + "grad_norm": 0.4609375, + "learning_rate": 0.00011461768162969732, + "loss": 2.4844, + "step": 21330 + }, + { + "epoch": 0.91, + "grad_norm": 0.474609375, + "learning_rate": 0.00011461104571867051, + "loss": 2.4534, + "step": 21331 + }, + { + "epoch": 0.91, + "grad_norm": 0.435546875, + "learning_rate": 0.00011460440974189928, + "loss": 2.3956, + "step": 21332 + }, + { + "epoch": 0.91, + "grad_norm": 0.3984375, + "learning_rate": 0.00011459777369941358, + "loss": 2.575, + "step": 21333 + }, + { + "epoch": 0.91, + "grad_norm": 0.439453125, + "learning_rate": 0.00011459113759124322, + "loss": 2.6552, + "step": 21334 + }, + { + "epoch": 0.91, + "grad_norm": 0.484375, + "learning_rate": 0.00011458450141741806, + "loss": 2.6185, + "step": 21335 + }, + { + "epoch": 0.91, + "grad_norm": 0.427734375, + "learning_rate": 0.00011457786517796798, + "loss": 2.5941, + "step": 21336 + }, + { + "epoch": 0.91, + "grad_norm": 0.455078125, + "learning_rate": 0.00011457122887292284, + "loss": 2.6415, + "step": 21337 + }, + { + "epoch": 0.91, + "grad_norm": 0.453125, + "learning_rate": 0.00011456459250231251, + "loss": 2.5008, + "step": 21338 + }, + { + "epoch": 0.91, + "grad_norm": 0.447265625, + "learning_rate": 0.00011455795606616682, + "loss": 2.6014, + "step": 21339 + }, + { + "epoch": 0.91, + "grad_norm": 0.408203125, + "learning_rate": 0.00011455131956451563, + "loss": 2.4963, + "step": 21340 + }, + { + "epoch": 0.91, + "grad_norm": 0.51953125, + "learning_rate": 0.00011454468299738884, + "loss": 2.5788, + "step": 21341 + }, + { + "epoch": 0.91, + "grad_norm": 0.423828125, + "learning_rate": 0.0001145380463648163, + "loss": 2.4723, + "step": 21342 + }, + { + "epoch": 0.91, + "grad_norm": 0.46875, + "learning_rate": 0.00011453140966682782, + "loss": 2.5567, + "step": 21343 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.00011452477290345336, + "loss": 2.5495, + "step": 21344 + }, + { + "epoch": 0.91, + "grad_norm": 0.400390625, + "learning_rate": 0.00011451813607472273, + "loss": 2.4984, + "step": 21345 + }, + { + "epoch": 0.91, + "grad_norm": 0.46875, + "learning_rate": 0.00011451149918066577, + "loss": 2.4982, + "step": 21346 + }, + { + "epoch": 0.91, + "grad_norm": 0.46484375, + "learning_rate": 0.00011450486222131236, + "loss": 2.6065, + "step": 21347 + }, + { + "epoch": 0.91, + "grad_norm": 0.478515625, + "learning_rate": 0.0001144982251966924, + "loss": 2.4825, + "step": 21348 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011449158810683566, + "loss": 2.5638, + "step": 21349 + }, + { + "epoch": 0.91, + "grad_norm": 0.431640625, + "learning_rate": 0.00011448495095177216, + "loss": 2.4707, + "step": 21350 + }, + { + "epoch": 0.91, + "grad_norm": 0.46484375, + "learning_rate": 0.00011447831373153162, + "loss": 2.6545, + "step": 21351 + }, + { + "epoch": 0.91, + "grad_norm": 0.490234375, + "learning_rate": 0.00011447167644614395, + "loss": 2.5393, + "step": 21352 + }, + { + "epoch": 0.91, + "grad_norm": 0.451171875, + "learning_rate": 0.00011446503909563906, + "loss": 2.401, + "step": 21353 + }, + { + "epoch": 0.91, + "grad_norm": 0.455078125, + "learning_rate": 0.00011445840168004678, + "loss": 2.4373, + "step": 21354 + }, + { + "epoch": 0.91, + "grad_norm": 0.4375, + "learning_rate": 0.00011445176419939697, + "loss": 2.4877, + "step": 21355 + }, + { + "epoch": 0.91, + "grad_norm": 0.48046875, + "learning_rate": 0.00011444512665371949, + "loss": 2.4866, + "step": 21356 + }, + { + "epoch": 0.91, + "grad_norm": 0.458984375, + "learning_rate": 0.0001144384890430442, + "loss": 2.4767, + "step": 21357 + }, + { + "epoch": 0.91, + "grad_norm": 0.458984375, + "learning_rate": 0.00011443185136740102, + "loss": 2.45, + "step": 21358 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011442521362681977, + "loss": 2.6094, + "step": 21359 + }, + { + "epoch": 0.91, + "grad_norm": 0.44921875, + "learning_rate": 0.00011441857582133033, + "loss": 2.5003, + "step": 21360 + }, + { + "epoch": 0.91, + "grad_norm": 0.47265625, + "learning_rate": 0.00011441193795096257, + "loss": 2.4633, + "step": 21361 + }, + { + "epoch": 0.91, + "grad_norm": 0.494140625, + "learning_rate": 0.00011440530001574636, + "loss": 2.538, + "step": 21362 + }, + { + "epoch": 0.91, + "grad_norm": 0.484375, + "learning_rate": 0.00011439866201571156, + "loss": 2.4874, + "step": 21363 + }, + { + "epoch": 0.91, + "grad_norm": 0.50390625, + "learning_rate": 0.00011439202395088806, + "loss": 2.6507, + "step": 21364 + }, + { + "epoch": 0.91, + "grad_norm": 0.46484375, + "learning_rate": 0.00011438538582130567, + "loss": 2.5418, + "step": 21365 + }, + { + "epoch": 0.91, + "grad_norm": 0.41015625, + "learning_rate": 0.00011437874762699437, + "loss": 2.4327, + "step": 21366 + }, + { + "epoch": 0.91, + "grad_norm": 0.40234375, + "learning_rate": 0.00011437210936798388, + "loss": 2.5238, + "step": 21367 + }, + { + "epoch": 0.91, + "grad_norm": 0.48828125, + "learning_rate": 0.0001143654710443042, + "loss": 2.5387, + "step": 21368 + }, + { + "epoch": 0.91, + "grad_norm": 0.46875, + "learning_rate": 0.00011435883265598515, + "loss": 2.5293, + "step": 21369 + }, + { + "epoch": 0.91, + "grad_norm": 0.416015625, + "learning_rate": 0.00011435219420305658, + "loss": 2.5509, + "step": 21370 + }, + { + "epoch": 0.91, + "grad_norm": 0.52734375, + "learning_rate": 0.00011434555568554839, + "loss": 2.528, + "step": 21371 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011433891710349046, + "loss": 2.5185, + "step": 21372 + }, + { + "epoch": 0.91, + "grad_norm": 0.49609375, + "learning_rate": 0.00011433227845691261, + "loss": 2.4626, + "step": 21373 + }, + { + "epoch": 0.91, + "grad_norm": 0.48828125, + "learning_rate": 0.00011432563974584479, + "loss": 2.4636, + "step": 21374 + }, + { + "epoch": 0.91, + "grad_norm": 0.4375, + "learning_rate": 0.00011431900097031676, + "loss": 2.5119, + "step": 21375 + }, + { + "epoch": 0.91, + "grad_norm": 0.5078125, + "learning_rate": 0.00011431236213035852, + "loss": 2.591, + "step": 21376 + }, + { + "epoch": 0.91, + "grad_norm": 0.44921875, + "learning_rate": 0.00011430572322599984, + "loss": 2.4703, + "step": 21377 + }, + { + "epoch": 0.91, + "grad_norm": 0.4140625, + "learning_rate": 0.00011429908425727065, + "loss": 2.5962, + "step": 21378 + }, + { + "epoch": 0.91, + "grad_norm": 0.43359375, + "learning_rate": 0.00011429244522420082, + "loss": 2.4892, + "step": 21379 + }, + { + "epoch": 0.91, + "grad_norm": 0.439453125, + "learning_rate": 0.00011428580612682021, + "loss": 2.5603, + "step": 21380 + }, + { + "epoch": 0.91, + "grad_norm": 0.41796875, + "learning_rate": 0.00011427916696515865, + "loss": 2.4478, + "step": 21381 + }, + { + "epoch": 0.91, + "grad_norm": 0.49609375, + "learning_rate": 0.00011427252773924611, + "loss": 2.5529, + "step": 21382 + }, + { + "epoch": 0.91, + "grad_norm": 0.494140625, + "learning_rate": 0.00011426588844911237, + "loss": 2.63, + "step": 21383 + }, + { + "epoch": 0.91, + "grad_norm": 0.439453125, + "learning_rate": 0.00011425924909478738, + "loss": 2.7722, + "step": 21384 + }, + { + "epoch": 0.91, + "grad_norm": 0.462890625, + "learning_rate": 0.00011425260967630096, + "loss": 2.4682, + "step": 21385 + }, + { + "epoch": 0.91, + "grad_norm": 0.453125, + "learning_rate": 0.00011424597019368302, + "loss": 2.545, + "step": 21386 + }, + { + "epoch": 0.91, + "grad_norm": 0.41796875, + "learning_rate": 0.0001142393306469634, + "loss": 2.4471, + "step": 21387 + }, + { + "epoch": 0.91, + "grad_norm": 0.4375, + "learning_rate": 0.00011423269103617202, + "loss": 2.5433, + "step": 21388 + }, + { + "epoch": 0.91, + "grad_norm": 0.435546875, + "learning_rate": 0.00011422605136133869, + "loss": 2.5606, + "step": 21389 + }, + { + "epoch": 0.91, + "grad_norm": 0.431640625, + "learning_rate": 0.00011421941162249339, + "loss": 2.6007, + "step": 21390 + }, + { + "epoch": 0.91, + "grad_norm": 0.51171875, + "learning_rate": 0.00011421277181966587, + "loss": 2.5541, + "step": 21391 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011420613195288611, + "loss": 2.4008, + "step": 21392 + }, + { + "epoch": 0.91, + "grad_norm": 0.384765625, + "learning_rate": 0.00011419949202218394, + "loss": 2.5032, + "step": 21393 + }, + { + "epoch": 0.91, + "grad_norm": 0.4296875, + "learning_rate": 0.00011419285202758925, + "loss": 2.6284, + "step": 21394 + }, + { + "epoch": 0.91, + "grad_norm": 0.42578125, + "learning_rate": 0.00011418621196913191, + "loss": 2.5188, + "step": 21395 + }, + { + "epoch": 0.91, + "grad_norm": 0.484375, + "learning_rate": 0.0001141795718468418, + "loss": 2.5666, + "step": 21396 + }, + { + "epoch": 0.91, + "grad_norm": 0.431640625, + "learning_rate": 0.00011417293166074878, + "loss": 2.5517, + "step": 21397 + }, + { + "epoch": 0.91, + "grad_norm": 0.53125, + "learning_rate": 0.0001141662914108828, + "loss": 2.4914, + "step": 21398 + }, + { + "epoch": 0.91, + "grad_norm": 0.51171875, + "learning_rate": 0.00011415965109727362, + "loss": 2.5342, + "step": 21399 + }, + { + "epoch": 0.91, + "grad_norm": 0.451171875, + "learning_rate": 0.00011415301071995122, + "loss": 2.6641, + "step": 21400 + }, + { + "epoch": 0.91, + "grad_norm": 0.478515625, + "learning_rate": 0.00011414637027894543, + "loss": 2.5774, + "step": 21401 + }, + { + "epoch": 0.91, + "grad_norm": 0.51171875, + "learning_rate": 0.00011413972977428615, + "loss": 2.546, + "step": 21402 + }, + { + "epoch": 0.91, + "grad_norm": 0.52734375, + "learning_rate": 0.00011413308920600327, + "loss": 2.6243, + "step": 21403 + }, + { + "epoch": 0.91, + "grad_norm": 0.46875, + "learning_rate": 0.00011412644857412664, + "loss": 2.6505, + "step": 21404 + }, + { + "epoch": 0.91, + "grad_norm": 0.39453125, + "learning_rate": 0.00011411980787868612, + "loss": 2.49, + "step": 21405 + }, + { + "epoch": 0.91, + "grad_norm": 0.421875, + "learning_rate": 0.00011411316711971169, + "loss": 2.5067, + "step": 21406 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011410652629723312, + "loss": 2.5587, + "step": 21407 + }, + { + "epoch": 0.91, + "grad_norm": 0.404296875, + "learning_rate": 0.00011409988541128037, + "loss": 2.5106, + "step": 21408 + }, + { + "epoch": 0.91, + "grad_norm": 0.412109375, + "learning_rate": 0.00011409324446188325, + "loss": 2.4951, + "step": 21409 + }, + { + "epoch": 0.91, + "grad_norm": 0.3984375, + "learning_rate": 0.00011408660344907171, + "loss": 2.4335, + "step": 21410 + }, + { + "epoch": 0.91, + "grad_norm": 0.435546875, + "learning_rate": 0.0001140799623728756, + "loss": 2.5558, + "step": 21411 + }, + { + "epoch": 0.91, + "grad_norm": 0.474609375, + "learning_rate": 0.00011407332123332479, + "loss": 2.6061, + "step": 21412 + }, + { + "epoch": 0.91, + "grad_norm": 0.48828125, + "learning_rate": 0.00011406668003044918, + "loss": 2.5808, + "step": 21413 + }, + { + "epoch": 0.91, + "grad_norm": 0.53125, + "learning_rate": 0.00011406003876427867, + "loss": 2.6428, + "step": 21414 + }, + { + "epoch": 0.91, + "grad_norm": 0.4765625, + "learning_rate": 0.00011405339743484309, + "loss": 2.5272, + "step": 21415 + }, + { + "epoch": 0.91, + "grad_norm": 0.474609375, + "learning_rate": 0.00011404675604217239, + "loss": 2.3936, + "step": 21416 + }, + { + "epoch": 0.91, + "grad_norm": 0.447265625, + "learning_rate": 0.0001140401145862964, + "loss": 2.5862, + "step": 21417 + }, + { + "epoch": 0.91, + "grad_norm": 0.44921875, + "learning_rate": 0.00011403347306724503, + "loss": 2.5399, + "step": 21418 + }, + { + "epoch": 0.91, + "grad_norm": 0.40625, + "learning_rate": 0.00011402683148504818, + "loss": 2.61, + "step": 21419 + }, + { + "epoch": 0.91, + "grad_norm": 0.466796875, + "learning_rate": 0.0001140201898397357, + "loss": 2.6094, + "step": 21420 + }, + { + "epoch": 0.91, + "grad_norm": 0.48046875, + "learning_rate": 0.00011401354813133747, + "loss": 2.5385, + "step": 21421 + }, + { + "epoch": 0.91, + "grad_norm": 0.47265625, + "learning_rate": 0.00011400690635988342, + "loss": 2.5748, + "step": 21422 + }, + { + "epoch": 0.91, + "grad_norm": 0.453125, + "learning_rate": 0.0001140002645254034, + "loss": 2.6869, + "step": 21423 + }, + { + "epoch": 0.91, + "grad_norm": 0.455078125, + "learning_rate": 0.0001139936226279273, + "loss": 2.6055, + "step": 21424 + }, + { + "epoch": 0.91, + "grad_norm": 0.498046875, + "learning_rate": 0.00011398698066748503, + "loss": 2.6519, + "step": 21425 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011398033864410647, + "loss": 2.5081, + "step": 21426 + }, + { + "epoch": 0.91, + "grad_norm": 0.494140625, + "learning_rate": 0.00011397369655782146, + "loss": 2.5776, + "step": 21427 + }, + { + "epoch": 0.91, + "grad_norm": 0.462890625, + "learning_rate": 0.00011396705440865994, + "loss": 2.6012, + "step": 21428 + }, + { + "epoch": 0.91, + "grad_norm": 0.451171875, + "learning_rate": 0.00011396041219665177, + "loss": 2.5697, + "step": 21429 + }, + { + "epoch": 0.91, + "grad_norm": 0.44921875, + "learning_rate": 0.00011395376992182687, + "loss": 2.6017, + "step": 21430 + }, + { + "epoch": 0.91, + "grad_norm": 0.478515625, + "learning_rate": 0.00011394712758421507, + "loss": 2.5825, + "step": 21431 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.00011394048518384633, + "loss": 2.5395, + "step": 21432 + }, + { + "epoch": 0.91, + "grad_norm": 0.5234375, + "learning_rate": 0.00011393384272075048, + "loss": 2.6178, + "step": 21433 + }, + { + "epoch": 0.91, + "grad_norm": 0.4921875, + "learning_rate": 0.00011392720019495742, + "loss": 2.5399, + "step": 21434 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.00011392055760649707, + "loss": 2.5549, + "step": 21435 + }, + { + "epoch": 0.91, + "grad_norm": 0.470703125, + "learning_rate": 0.00011391391495539929, + "loss": 2.5426, + "step": 21436 + }, + { + "epoch": 0.91, + "grad_norm": 0.4609375, + "learning_rate": 0.00011390727224169396, + "loss": 2.458, + "step": 21437 + }, + { + "epoch": 0.91, + "grad_norm": 0.431640625, + "learning_rate": 0.00011390062946541102, + "loss": 2.5986, + "step": 21438 + }, + { + "epoch": 0.91, + "grad_norm": 0.484375, + "learning_rate": 0.00011389398662658031, + "loss": 2.4767, + "step": 21439 + }, + { + "epoch": 0.91, + "grad_norm": 0.4375, + "learning_rate": 0.00011388734372523174, + "loss": 2.5158, + "step": 21440 + }, + { + "epoch": 0.91, + "grad_norm": 0.404296875, + "learning_rate": 0.00011388070076139521, + "loss": 2.3887, + "step": 21441 + }, + { + "epoch": 0.91, + "grad_norm": 0.5390625, + "learning_rate": 0.00011387405773510058, + "loss": 2.5912, + "step": 21442 + }, + { + "epoch": 0.91, + "grad_norm": 0.421875, + "learning_rate": 0.00011386741464637778, + "loss": 2.5055, + "step": 21443 + }, + { + "epoch": 0.91, + "grad_norm": 0.4609375, + "learning_rate": 0.00011386077149525667, + "loss": 2.6495, + "step": 21444 + }, + { + "epoch": 0.91, + "grad_norm": 0.45703125, + "learning_rate": 0.00011385412828176714, + "loss": 2.5331, + "step": 21445 + }, + { + "epoch": 0.91, + "grad_norm": 0.421875, + "learning_rate": 0.00011384748500593913, + "loss": 2.5495, + "step": 21446 + }, + { + "epoch": 0.91, + "grad_norm": 0.458984375, + "learning_rate": 0.00011384084166780246, + "loss": 2.617, + "step": 21447 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.00011383419826738708, + "loss": 2.5369, + "step": 21448 + }, + { + "epoch": 0.91, + "grad_norm": 0.421875, + "learning_rate": 0.00011382755480472287, + "loss": 2.5293, + "step": 21449 + }, + { + "epoch": 0.91, + "grad_norm": 0.427734375, + "learning_rate": 0.00011382091127983972, + "loss": 2.4882, + "step": 21450 + }, + { + "epoch": 0.91, + "grad_norm": 0.6015625, + "learning_rate": 0.00011381426769276751, + "loss": 2.423, + "step": 21451 + }, + { + "epoch": 0.91, + "grad_norm": 0.470703125, + "learning_rate": 0.00011380762404353613, + "loss": 2.6173, + "step": 21452 + }, + { + "epoch": 0.91, + "grad_norm": 0.42578125, + "learning_rate": 0.0001138009803321755, + "loss": 2.6969, + "step": 21453 + }, + { + "epoch": 0.91, + "grad_norm": 0.453125, + "learning_rate": 0.00011379433655871552, + "loss": 2.5549, + "step": 21454 + }, + { + "epoch": 0.91, + "grad_norm": 0.54296875, + "learning_rate": 0.00011378769272318605, + "loss": 2.6665, + "step": 21455 + }, + { + "epoch": 0.91, + "grad_norm": 0.5390625, + "learning_rate": 0.000113781048825617, + "loss": 2.4727, + "step": 21456 + }, + { + "epoch": 0.91, + "grad_norm": 0.423828125, + "learning_rate": 0.00011377440486603828, + "loss": 2.5819, + "step": 21457 + }, + { + "epoch": 0.91, + "grad_norm": 0.51953125, + "learning_rate": 0.00011376776084447976, + "loss": 2.5667, + "step": 21458 + }, + { + "epoch": 0.91, + "grad_norm": 0.41796875, + "learning_rate": 0.00011376111676097134, + "loss": 2.5782, + "step": 21459 + }, + { + "epoch": 0.91, + "grad_norm": 0.453125, + "learning_rate": 0.00011375447261554293, + "loss": 2.4825, + "step": 21460 + }, + { + "epoch": 0.91, + "grad_norm": 0.427734375, + "learning_rate": 0.00011374782840822442, + "loss": 2.6634, + "step": 21461 + }, + { + "epoch": 0.91, + "grad_norm": 0.44140625, + "learning_rate": 0.00011374118413904573, + "loss": 2.5732, + "step": 21462 + }, + { + "epoch": 0.91, + "grad_norm": 0.443359375, + "learning_rate": 0.0001137345398080367, + "loss": 2.5821, + "step": 21463 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011372789541522727, + "loss": 2.4862, + "step": 21464 + }, + { + "epoch": 0.91, + "grad_norm": 0.50390625, + "learning_rate": 0.00011372125096064734, + "loss": 2.5386, + "step": 21465 + }, + { + "epoch": 0.91, + "grad_norm": 0.4453125, + "learning_rate": 0.00011371460644432677, + "loss": 2.6157, + "step": 21466 + }, + { + "epoch": 0.91, + "grad_norm": 0.404296875, + "learning_rate": 0.0001137079618662955, + "loss": 2.5901, + "step": 21467 + }, + { + "epoch": 0.91, + "grad_norm": 0.421875, + "learning_rate": 0.00011370131722658341, + "loss": 2.577, + "step": 21468 + }, + { + "epoch": 0.91, + "grad_norm": 0.439453125, + "learning_rate": 0.00011369467252522041, + "loss": 2.4482, + "step": 21469 + }, + { + "epoch": 0.91, + "grad_norm": 0.3828125, + "learning_rate": 0.00011368802776223637, + "loss": 2.4411, + "step": 21470 + }, + { + "epoch": 0.91, + "grad_norm": 0.416015625, + "learning_rate": 0.00011368138293766122, + "loss": 2.5056, + "step": 21471 + }, + { + "epoch": 0.91, + "grad_norm": 0.482421875, + "learning_rate": 0.00011367473805152484, + "loss": 2.5234, + "step": 21472 + }, + { + "epoch": 0.92, + "grad_norm": 0.41796875, + "learning_rate": 0.00011366809310385715, + "loss": 2.5757, + "step": 21473 + }, + { + "epoch": 0.92, + "grad_norm": 0.443359375, + "learning_rate": 0.00011366144809468802, + "loss": 2.5014, + "step": 21474 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.00011365480302404736, + "loss": 2.5472, + "step": 21475 + }, + { + "epoch": 0.92, + "grad_norm": 0.478515625, + "learning_rate": 0.00011364815789196511, + "loss": 2.5085, + "step": 21476 + }, + { + "epoch": 0.92, + "grad_norm": 0.5078125, + "learning_rate": 0.0001136415126984711, + "loss": 2.4845, + "step": 21477 + }, + { + "epoch": 0.92, + "grad_norm": 0.5390625, + "learning_rate": 0.00011363486744359528, + "loss": 2.6259, + "step": 21478 + }, + { + "epoch": 0.92, + "grad_norm": 0.46484375, + "learning_rate": 0.00011362822212736754, + "loss": 2.5237, + "step": 21479 + }, + { + "epoch": 0.92, + "grad_norm": 0.5390625, + "learning_rate": 0.00011362157674981777, + "loss": 2.484, + "step": 21480 + }, + { + "epoch": 0.92, + "grad_norm": 0.44140625, + "learning_rate": 0.00011361493131097594, + "loss": 2.5298, + "step": 21481 + }, + { + "epoch": 0.92, + "grad_norm": 0.46484375, + "learning_rate": 0.00011360828581087182, + "loss": 2.7774, + "step": 21482 + }, + { + "epoch": 0.92, + "grad_norm": 0.453125, + "learning_rate": 0.00011360164024953541, + "loss": 2.4995, + "step": 21483 + }, + { + "epoch": 0.92, + "grad_norm": 0.494140625, + "learning_rate": 0.0001135949946269966, + "loss": 2.542, + "step": 21484 + }, + { + "epoch": 0.92, + "grad_norm": 0.46875, + "learning_rate": 0.00011358834894328529, + "loss": 2.6124, + "step": 21485 + }, + { + "epoch": 0.92, + "grad_norm": 0.447265625, + "learning_rate": 0.00011358170319843138, + "loss": 2.6676, + "step": 21486 + }, + { + "epoch": 0.92, + "grad_norm": 0.515625, + "learning_rate": 0.00011357505739246475, + "loss": 2.5706, + "step": 21487 + }, + { + "epoch": 0.92, + "grad_norm": 0.4375, + "learning_rate": 0.0001135684115254153, + "loss": 2.556, + "step": 21488 + }, + { + "epoch": 0.92, + "grad_norm": 0.43359375, + "learning_rate": 0.00011356176559731302, + "loss": 2.5534, + "step": 21489 + }, + { + "epoch": 0.92, + "grad_norm": 0.482421875, + "learning_rate": 0.0001135551196081877, + "loss": 2.6439, + "step": 21490 + }, + { + "epoch": 0.92, + "grad_norm": 0.4765625, + "learning_rate": 0.00011354847355806932, + "loss": 2.6738, + "step": 21491 + }, + { + "epoch": 0.92, + "grad_norm": 0.5078125, + "learning_rate": 0.00011354182744698776, + "loss": 2.6938, + "step": 21492 + }, + { + "epoch": 0.92, + "grad_norm": 0.470703125, + "learning_rate": 0.00011353518127497291, + "loss": 2.5994, + "step": 21493 + }, + { + "epoch": 0.92, + "grad_norm": 0.40625, + "learning_rate": 0.00011352853504205469, + "loss": 2.6246, + "step": 21494 + }, + { + "epoch": 0.92, + "grad_norm": 0.470703125, + "learning_rate": 0.00011352188874826304, + "loss": 2.5329, + "step": 21495 + }, + { + "epoch": 0.92, + "grad_norm": 0.453125, + "learning_rate": 0.00011351524239362778, + "loss": 2.4801, + "step": 21496 + }, + { + "epoch": 0.92, + "grad_norm": 0.50390625, + "learning_rate": 0.00011350859597817892, + "loss": 2.6325, + "step": 21497 + }, + { + "epoch": 0.92, + "grad_norm": 0.5, + "learning_rate": 0.00011350194950194626, + "loss": 2.5959, + "step": 21498 + }, + { + "epoch": 0.92, + "grad_norm": 0.423828125, + "learning_rate": 0.0001134953029649598, + "loss": 2.5925, + "step": 21499 + }, + { + "epoch": 0.92, + "grad_norm": 0.4375, + "learning_rate": 0.00011348865636724942, + "loss": 2.5625, + "step": 21500 + }, + { + "epoch": 0.92, + "grad_norm": 0.40625, + "learning_rate": 0.000113482009708845, + "loss": 2.4023, + "step": 21501 + }, + { + "epoch": 0.92, + "grad_norm": 0.47265625, + "learning_rate": 0.00011347536298977645, + "loss": 2.6186, + "step": 21502 + }, + { + "epoch": 0.92, + "grad_norm": 0.44140625, + "learning_rate": 0.00011346871621007372, + "loss": 2.5432, + "step": 21503 + }, + { + "epoch": 0.92, + "grad_norm": 0.4296875, + "learning_rate": 0.00011346206936976664, + "loss": 2.5433, + "step": 21504 + }, + { + "epoch": 0.92, + "grad_norm": 0.4453125, + "learning_rate": 0.00011345542246888521, + "loss": 2.5904, + "step": 21505 + }, + { + "epoch": 0.92, + "grad_norm": 0.423828125, + "learning_rate": 0.00011344877550745928, + "loss": 2.5316, + "step": 21506 + }, + { + "epoch": 0.92, + "grad_norm": 0.427734375, + "learning_rate": 0.00011344212848551879, + "loss": 2.4309, + "step": 21507 + }, + { + "epoch": 0.92, + "grad_norm": 0.447265625, + "learning_rate": 0.00011343548140309362, + "loss": 2.6882, + "step": 21508 + }, + { + "epoch": 0.92, + "grad_norm": 0.439453125, + "learning_rate": 0.00011342883426021369, + "loss": 2.6274, + "step": 21509 + }, + { + "epoch": 0.92, + "grad_norm": 0.53125, + "learning_rate": 0.00011342218705690893, + "loss": 2.6711, + "step": 21510 + }, + { + "epoch": 0.92, + "grad_norm": 0.482421875, + "learning_rate": 0.00011341553979320922, + "loss": 2.572, + "step": 21511 + }, + { + "epoch": 0.92, + "grad_norm": 0.44921875, + "learning_rate": 0.00011340889246914448, + "loss": 2.5106, + "step": 21512 + }, + { + "epoch": 0.92, + "grad_norm": 0.46875, + "learning_rate": 0.00011340224508474466, + "loss": 2.6624, + "step": 21513 + }, + { + "epoch": 0.92, + "grad_norm": 0.466796875, + "learning_rate": 0.0001133955976400396, + "loss": 2.567, + "step": 21514 + }, + { + "epoch": 0.92, + "grad_norm": 0.44140625, + "learning_rate": 0.00011338895013505923, + "loss": 2.5444, + "step": 21515 + }, + { + "epoch": 0.92, + "grad_norm": 0.466796875, + "learning_rate": 0.0001133823025698335, + "loss": 2.5934, + "step": 21516 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.00011337565494439233, + "loss": 2.5895, + "step": 21517 + }, + { + "epoch": 0.92, + "grad_norm": 0.447265625, + "learning_rate": 0.00011336900725876556, + "loss": 2.669, + "step": 21518 + }, + { + "epoch": 0.92, + "grad_norm": 0.482421875, + "learning_rate": 0.00011336235951298318, + "loss": 2.5336, + "step": 21519 + }, + { + "epoch": 0.92, + "grad_norm": 0.470703125, + "learning_rate": 0.000113355711707075, + "loss": 2.4977, + "step": 21520 + }, + { + "epoch": 0.92, + "grad_norm": 0.47265625, + "learning_rate": 0.00011334906384107106, + "loss": 2.5963, + "step": 21521 + }, + { + "epoch": 0.92, + "grad_norm": 0.42578125, + "learning_rate": 0.00011334241591500118, + "loss": 2.4399, + "step": 21522 + }, + { + "epoch": 0.92, + "grad_norm": 0.455078125, + "learning_rate": 0.0001133357679288953, + "loss": 2.6308, + "step": 21523 + }, + { + "epoch": 0.92, + "grad_norm": 0.443359375, + "learning_rate": 0.00011332911988278336, + "loss": 2.5499, + "step": 21524 + }, + { + "epoch": 0.92, + "grad_norm": 0.45703125, + "learning_rate": 0.00011332247177669526, + "loss": 2.5113, + "step": 21525 + }, + { + "epoch": 0.92, + "grad_norm": 0.4296875, + "learning_rate": 0.00011331582361066088, + "loss": 2.536, + "step": 21526 + }, + { + "epoch": 0.92, + "grad_norm": 0.4765625, + "learning_rate": 0.00011330917538471018, + "loss": 2.5845, + "step": 21527 + }, + { + "epoch": 0.92, + "grad_norm": 0.42578125, + "learning_rate": 0.00011330252709887303, + "loss": 2.5232, + "step": 21528 + }, + { + "epoch": 0.92, + "grad_norm": 0.46484375, + "learning_rate": 0.00011329587875317941, + "loss": 2.4555, + "step": 21529 + }, + { + "epoch": 0.92, + "grad_norm": 0.474609375, + "learning_rate": 0.00011328923034765915, + "loss": 2.5739, + "step": 21530 + }, + { + "epoch": 0.92, + "grad_norm": 0.423828125, + "learning_rate": 0.00011328258188234224, + "loss": 2.4884, + "step": 21531 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.00011327593335725855, + "loss": 2.5684, + "step": 21532 + }, + { + "epoch": 0.92, + "grad_norm": 0.455078125, + "learning_rate": 0.00011326928477243801, + "loss": 2.6294, + "step": 21533 + }, + { + "epoch": 0.92, + "grad_norm": 0.48046875, + "learning_rate": 0.00011326263612791055, + "loss": 2.4983, + "step": 21534 + }, + { + "epoch": 0.92, + "grad_norm": 0.474609375, + "learning_rate": 0.00011325598742370607, + "loss": 2.6408, + "step": 21535 + }, + { + "epoch": 0.92, + "grad_norm": 0.484375, + "learning_rate": 0.00011324933865985447, + "loss": 2.6224, + "step": 21536 + }, + { + "epoch": 0.92, + "grad_norm": 0.5, + "learning_rate": 0.00011324268983638575, + "loss": 2.5576, + "step": 21537 + }, + { + "epoch": 0.92, + "grad_norm": 0.484375, + "learning_rate": 0.00011323604095332969, + "loss": 2.6787, + "step": 21538 + }, + { + "epoch": 0.92, + "grad_norm": 0.45703125, + "learning_rate": 0.00011322939201071632, + "loss": 2.545, + "step": 21539 + }, + { + "epoch": 0.92, + "grad_norm": 0.400390625, + "learning_rate": 0.00011322274300857554, + "loss": 2.6239, + "step": 21540 + }, + { + "epoch": 0.92, + "grad_norm": 0.5078125, + "learning_rate": 0.00011321609394693722, + "loss": 2.7152, + "step": 21541 + }, + { + "epoch": 0.92, + "grad_norm": 0.373046875, + "learning_rate": 0.00011320944482583131, + "loss": 2.383, + "step": 21542 + }, + { + "epoch": 0.92, + "grad_norm": 0.44140625, + "learning_rate": 0.00011320279564528772, + "loss": 2.6757, + "step": 21543 + }, + { + "epoch": 0.92, + "grad_norm": 0.443359375, + "learning_rate": 0.00011319614640533637, + "loss": 2.3852, + "step": 21544 + }, + { + "epoch": 0.92, + "grad_norm": 0.376953125, + "learning_rate": 0.00011318949710600724, + "loss": 2.6204, + "step": 21545 + }, + { + "epoch": 0.92, + "grad_norm": 0.435546875, + "learning_rate": 0.00011318284774733013, + "loss": 2.6234, + "step": 21546 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.00011317619832933504, + "loss": 2.669, + "step": 21547 + }, + { + "epoch": 0.92, + "grad_norm": 0.435546875, + "learning_rate": 0.00011316954885205189, + "loss": 2.6402, + "step": 21548 + }, + { + "epoch": 0.92, + "grad_norm": 0.453125, + "learning_rate": 0.00011316289931551057, + "loss": 2.6186, + "step": 21549 + }, + { + "epoch": 0.92, + "grad_norm": 0.453125, + "learning_rate": 0.00011315624971974102, + "loss": 2.4872, + "step": 21550 + }, + { + "epoch": 0.92, + "grad_norm": 0.40234375, + "learning_rate": 0.00011314960006477314, + "loss": 2.5885, + "step": 21551 + }, + { + "epoch": 0.92, + "grad_norm": 0.453125, + "learning_rate": 0.00011314295035063684, + "loss": 2.7223, + "step": 21552 + }, + { + "epoch": 0.92, + "grad_norm": 0.439453125, + "learning_rate": 0.00011313630057736214, + "loss": 2.591, + "step": 21553 + }, + { + "epoch": 0.92, + "grad_norm": 0.5390625, + "learning_rate": 0.00011312965074497884, + "loss": 2.6265, + "step": 21554 + }, + { + "epoch": 0.92, + "grad_norm": 0.419921875, + "learning_rate": 0.0001131230008535169, + "loss": 2.4705, + "step": 21555 + }, + { + "epoch": 0.92, + "grad_norm": 0.4375, + "learning_rate": 0.00011311635090300627, + "loss": 2.5115, + "step": 21556 + }, + { + "epoch": 0.92, + "grad_norm": 0.4609375, + "learning_rate": 0.00011310970089347686, + "loss": 2.5252, + "step": 21557 + }, + { + "epoch": 0.92, + "grad_norm": 0.490234375, + "learning_rate": 0.0001131030508249586, + "loss": 2.5428, + "step": 21558 + }, + { + "epoch": 0.92, + "grad_norm": 0.443359375, + "learning_rate": 0.00011309640069748137, + "loss": 2.6127, + "step": 21559 + }, + { + "epoch": 0.92, + "grad_norm": 0.43359375, + "learning_rate": 0.00011308975051107511, + "loss": 2.4545, + "step": 21560 + }, + { + "epoch": 0.92, + "grad_norm": 0.416015625, + "learning_rate": 0.00011308310026576981, + "loss": 2.5631, + "step": 21561 + }, + { + "epoch": 0.92, + "grad_norm": 0.4765625, + "learning_rate": 0.0001130764499615953, + "loss": 2.5411, + "step": 21562 + }, + { + "epoch": 0.92, + "grad_norm": 0.443359375, + "learning_rate": 0.00011306979959858154, + "loss": 2.5729, + "step": 21563 + }, + { + "epoch": 0.92, + "grad_norm": 0.427734375, + "learning_rate": 0.00011306314917675848, + "loss": 2.5269, + "step": 21564 + }, + { + "epoch": 0.92, + "grad_norm": 0.44921875, + "learning_rate": 0.00011305649869615602, + "loss": 2.6457, + "step": 21565 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.00011304984815680409, + "loss": 2.4793, + "step": 21566 + }, + { + "epoch": 0.92, + "grad_norm": 0.392578125, + "learning_rate": 0.0001130431975587326, + "loss": 2.4802, + "step": 21567 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.0001130365469019715, + "loss": 2.5785, + "step": 21568 + }, + { + "epoch": 0.92, + "grad_norm": 0.443359375, + "learning_rate": 0.00011302989618655072, + "loss": 2.5679, + "step": 21569 + }, + { + "epoch": 0.92, + "grad_norm": 0.47265625, + "learning_rate": 0.00011302324541250011, + "loss": 2.5401, + "step": 21570 + }, + { + "epoch": 0.92, + "grad_norm": 0.484375, + "learning_rate": 0.0001130165945798497, + "loss": 2.4652, + "step": 21571 + }, + { + "epoch": 0.92, + "grad_norm": 0.4921875, + "learning_rate": 0.00011300994368862936, + "loss": 2.6386, + "step": 21572 + }, + { + "epoch": 0.92, + "grad_norm": 0.474609375, + "learning_rate": 0.00011300329273886902, + "loss": 2.5112, + "step": 21573 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.00011299664173059863, + "loss": 2.6099, + "step": 21574 + }, + { + "epoch": 0.92, + "grad_norm": 0.466796875, + "learning_rate": 0.00011298999066384813, + "loss": 2.4461, + "step": 21575 + }, + { + "epoch": 0.92, + "grad_norm": 0.498046875, + "learning_rate": 0.00011298333953864737, + "loss": 2.581, + "step": 21576 + }, + { + "epoch": 0.92, + "grad_norm": 0.498046875, + "learning_rate": 0.00011297668835502635, + "loss": 2.5829, + "step": 21577 + }, + { + "epoch": 0.92, + "grad_norm": 0.46875, + "learning_rate": 0.00011297003711301497, + "loss": 2.5157, + "step": 21578 + }, + { + "epoch": 0.92, + "grad_norm": 0.498046875, + "learning_rate": 0.00011296338581264314, + "loss": 2.5814, + "step": 21579 + }, + { + "epoch": 0.92, + "grad_norm": 0.50390625, + "learning_rate": 0.00011295673445394086, + "loss": 2.479, + "step": 21580 + }, + { + "epoch": 0.92, + "grad_norm": 0.46484375, + "learning_rate": 0.00011295008303693797, + "loss": 2.6368, + "step": 21581 + }, + { + "epoch": 0.92, + "grad_norm": 0.447265625, + "learning_rate": 0.00011294343156166445, + "loss": 2.4242, + "step": 21582 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.00011293678002815023, + "loss": 2.8008, + "step": 21583 + }, + { + "epoch": 0.92, + "grad_norm": 0.4375, + "learning_rate": 0.00011293012843642521, + "loss": 2.5763, + "step": 21584 + }, + { + "epoch": 0.92, + "grad_norm": 0.41015625, + "learning_rate": 0.00011292347678651935, + "loss": 2.5096, + "step": 21585 + }, + { + "epoch": 0.92, + "grad_norm": 0.3984375, + "learning_rate": 0.00011291682507846257, + "loss": 2.5972, + "step": 21586 + }, + { + "epoch": 0.92, + "grad_norm": 0.46875, + "learning_rate": 0.00011291017331228477, + "loss": 2.6103, + "step": 21587 + }, + { + "epoch": 0.92, + "grad_norm": 0.5546875, + "learning_rate": 0.00011290352148801595, + "loss": 2.682, + "step": 21588 + }, + { + "epoch": 0.92, + "grad_norm": 0.490234375, + "learning_rate": 0.00011289686960568595, + "loss": 2.5493, + "step": 21589 + }, + { + "epoch": 0.92, + "grad_norm": 0.50390625, + "learning_rate": 0.0001128902176653248, + "loss": 2.6173, + "step": 21590 + }, + { + "epoch": 0.92, + "grad_norm": 0.515625, + "learning_rate": 0.00011288356566696234, + "loss": 2.6485, + "step": 21591 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.00011287691361062858, + "loss": 2.6023, + "step": 21592 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.0001128702614963534, + "loss": 2.5661, + "step": 21593 + }, + { + "epoch": 0.92, + "grad_norm": 0.427734375, + "learning_rate": 0.00011286360932416673, + "loss": 2.5813, + "step": 21594 + }, + { + "epoch": 0.92, + "grad_norm": 0.421875, + "learning_rate": 0.00011285695709409852, + "loss": 2.4945, + "step": 21595 + }, + { + "epoch": 0.92, + "grad_norm": 0.44140625, + "learning_rate": 0.00011285030480617873, + "loss": 2.7485, + "step": 21596 + }, + { + "epoch": 0.92, + "grad_norm": 0.4609375, + "learning_rate": 0.00011284365246043724, + "loss": 2.5648, + "step": 21597 + }, + { + "epoch": 0.92, + "grad_norm": 0.44140625, + "learning_rate": 0.00011283700005690403, + "loss": 2.4834, + "step": 21598 + }, + { + "epoch": 0.92, + "grad_norm": 0.400390625, + "learning_rate": 0.000112830347595609, + "loss": 2.4284, + "step": 21599 + }, + { + "epoch": 0.92, + "grad_norm": 0.486328125, + "learning_rate": 0.00011282369507658208, + "loss": 2.6588, + "step": 21600 + }, + { + "epoch": 0.92, + "grad_norm": 0.458984375, + "learning_rate": 0.0001128170424998532, + "loss": 2.4594, + "step": 21601 + }, + { + "epoch": 0.92, + "grad_norm": 0.48828125, + "learning_rate": 0.00011281038986545238, + "loss": 2.3656, + "step": 21602 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.00011280373717340944, + "loss": 2.5581, + "step": 21603 + }, + { + "epoch": 0.92, + "grad_norm": 0.484375, + "learning_rate": 0.0001127970844237544, + "loss": 2.509, + "step": 21604 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.0001127904316165171, + "loss": 2.4949, + "step": 21605 + }, + { + "epoch": 0.92, + "grad_norm": 0.4921875, + "learning_rate": 0.00011278377875172757, + "loss": 2.5201, + "step": 21606 + }, + { + "epoch": 0.92, + "grad_norm": 0.423828125, + "learning_rate": 0.0001127771258294157, + "loss": 2.5536, + "step": 21607 + }, + { + "epoch": 0.92, + "grad_norm": 0.53515625, + "learning_rate": 0.00011277047284961144, + "loss": 2.6427, + "step": 21608 + }, + { + "epoch": 0.92, + "grad_norm": 0.53125, + "learning_rate": 0.0001127638198123447, + "loss": 2.4951, + "step": 21609 + }, + { + "epoch": 0.92, + "grad_norm": 0.416015625, + "learning_rate": 0.00011275716671764547, + "loss": 2.55, + "step": 21610 + }, + { + "epoch": 0.92, + "grad_norm": 0.443359375, + "learning_rate": 0.0001127505135655436, + "loss": 2.5337, + "step": 21611 + }, + { + "epoch": 0.92, + "grad_norm": 0.4453125, + "learning_rate": 0.00011274386035606914, + "loss": 2.5393, + "step": 21612 + }, + { + "epoch": 0.92, + "grad_norm": 0.458984375, + "learning_rate": 0.00011273720708925193, + "loss": 2.5868, + "step": 21613 + }, + { + "epoch": 0.92, + "grad_norm": 0.41796875, + "learning_rate": 0.00011273055376512193, + "loss": 2.443, + "step": 21614 + }, + { + "epoch": 0.92, + "grad_norm": 0.453125, + "learning_rate": 0.00011272390038370913, + "loss": 2.5219, + "step": 21615 + }, + { + "epoch": 0.92, + "grad_norm": 0.451171875, + "learning_rate": 0.00011271724694504339, + "loss": 2.5235, + "step": 21616 + }, + { + "epoch": 0.92, + "grad_norm": 0.412109375, + "learning_rate": 0.00011271059344915468, + "loss": 2.4453, + "step": 21617 + }, + { + "epoch": 0.92, + "grad_norm": 0.48046875, + "learning_rate": 0.000112703939896073, + "loss": 2.6274, + "step": 21618 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.00011269728628582818, + "loss": 2.6505, + "step": 21619 + }, + { + "epoch": 0.92, + "grad_norm": 0.48046875, + "learning_rate": 0.00011269063261845025, + "loss": 2.5922, + "step": 21620 + }, + { + "epoch": 0.92, + "grad_norm": 0.474609375, + "learning_rate": 0.00011268397889396909, + "loss": 2.4964, + "step": 21621 + }, + { + "epoch": 0.92, + "grad_norm": 0.5, + "learning_rate": 0.00011267732511241467, + "loss": 2.6706, + "step": 21622 + }, + { + "epoch": 0.92, + "grad_norm": 0.427734375, + "learning_rate": 0.00011267067127381692, + "loss": 2.5438, + "step": 21623 + }, + { + "epoch": 0.92, + "grad_norm": 0.455078125, + "learning_rate": 0.00011266401737820577, + "loss": 2.6288, + "step": 21624 + }, + { + "epoch": 0.92, + "grad_norm": 0.4921875, + "learning_rate": 0.00011265736342561115, + "loss": 2.5585, + "step": 21625 + }, + { + "epoch": 0.92, + "grad_norm": 0.45703125, + "learning_rate": 0.00011265070941606306, + "loss": 2.4971, + "step": 21626 + }, + { + "epoch": 0.92, + "grad_norm": 0.470703125, + "learning_rate": 0.00011264405534959137, + "loss": 2.4194, + "step": 21627 + }, + { + "epoch": 0.92, + "grad_norm": 0.416015625, + "learning_rate": 0.00011263740122622608, + "loss": 2.4448, + "step": 21628 + }, + { + "epoch": 0.92, + "grad_norm": 0.447265625, + "learning_rate": 0.0001126307470459971, + "loss": 2.5081, + "step": 21629 + }, + { + "epoch": 0.92, + "grad_norm": 0.4609375, + "learning_rate": 0.00011262409280893435, + "loss": 2.6221, + "step": 21630 + }, + { + "epoch": 0.92, + "grad_norm": 0.48046875, + "learning_rate": 0.00011261743851506783, + "loss": 2.6371, + "step": 21631 + }, + { + "epoch": 0.92, + "grad_norm": 0.478515625, + "learning_rate": 0.00011261078416442741, + "loss": 2.4714, + "step": 21632 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.00011260412975704309, + "loss": 2.4955, + "step": 21633 + }, + { + "epoch": 0.92, + "grad_norm": 0.478515625, + "learning_rate": 0.00011259747529294479, + "loss": 2.5072, + "step": 21634 + }, + { + "epoch": 0.92, + "grad_norm": 0.427734375, + "learning_rate": 0.00011259082077216245, + "loss": 2.4807, + "step": 21635 + }, + { + "epoch": 0.92, + "grad_norm": 0.478515625, + "learning_rate": 0.00011258416619472602, + "loss": 2.4701, + "step": 21636 + }, + { + "epoch": 0.92, + "grad_norm": 0.373046875, + "learning_rate": 0.00011257751156066545, + "loss": 2.4829, + "step": 21637 + }, + { + "epoch": 0.92, + "grad_norm": 0.46484375, + "learning_rate": 0.00011257085687001067, + "loss": 2.5757, + "step": 21638 + }, + { + "epoch": 0.92, + "grad_norm": 0.43359375, + "learning_rate": 0.00011256420212279164, + "loss": 2.5555, + "step": 21639 + }, + { + "epoch": 0.92, + "grad_norm": 0.423828125, + "learning_rate": 0.00011255754731903828, + "loss": 2.3692, + "step": 21640 + }, + { + "epoch": 0.92, + "grad_norm": 0.423828125, + "learning_rate": 0.00011255089245878053, + "loss": 2.4774, + "step": 21641 + }, + { + "epoch": 0.92, + "grad_norm": 0.5, + "learning_rate": 0.0001125442375420484, + "loss": 2.6696, + "step": 21642 + }, + { + "epoch": 0.92, + "grad_norm": 0.43359375, + "learning_rate": 0.00011253758256887172, + "loss": 2.5928, + "step": 21643 + }, + { + "epoch": 0.92, + "grad_norm": 0.546875, + "learning_rate": 0.00011253092753928054, + "loss": 2.5516, + "step": 21644 + }, + { + "epoch": 0.92, + "grad_norm": 0.427734375, + "learning_rate": 0.00011252427245330474, + "loss": 2.4809, + "step": 21645 + }, + { + "epoch": 0.92, + "grad_norm": 0.435546875, + "learning_rate": 0.00011251761731097432, + "loss": 2.5677, + "step": 21646 + }, + { + "epoch": 0.92, + "grad_norm": 0.439453125, + "learning_rate": 0.0001125109621123192, + "loss": 2.7159, + "step": 21647 + }, + { + "epoch": 0.92, + "grad_norm": 0.453125, + "learning_rate": 0.0001125043068573693, + "loss": 2.6169, + "step": 21648 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.0001124976515461546, + "loss": 2.6205, + "step": 21649 + }, + { + "epoch": 0.92, + "grad_norm": 0.4375, + "learning_rate": 0.00011249099617870504, + "loss": 2.5279, + "step": 21650 + }, + { + "epoch": 0.92, + "grad_norm": 0.4765625, + "learning_rate": 0.00011248434075505055, + "loss": 2.74, + "step": 21651 + }, + { + "epoch": 0.92, + "grad_norm": 0.484375, + "learning_rate": 0.00011247768527522109, + "loss": 2.4741, + "step": 21652 + }, + { + "epoch": 0.92, + "grad_norm": 0.4921875, + "learning_rate": 0.00011247102973924662, + "loss": 2.5986, + "step": 21653 + }, + { + "epoch": 0.92, + "grad_norm": 0.4453125, + "learning_rate": 0.00011246437414715705, + "loss": 2.4284, + "step": 21654 + }, + { + "epoch": 0.92, + "grad_norm": 0.4140625, + "learning_rate": 0.00011245771849898237, + "loss": 2.4336, + "step": 21655 + }, + { + "epoch": 0.92, + "grad_norm": 0.462890625, + "learning_rate": 0.00011245106279475251, + "loss": 2.509, + "step": 21656 + }, + { + "epoch": 0.92, + "grad_norm": 0.46484375, + "learning_rate": 0.00011244440703449738, + "loss": 2.6165, + "step": 21657 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.00011243775121824704, + "loss": 2.7093, + "step": 21658 + }, + { + "epoch": 0.92, + "grad_norm": 0.44921875, + "learning_rate": 0.00011243109534603129, + "loss": 2.5012, + "step": 21659 + }, + { + "epoch": 0.92, + "grad_norm": 0.46484375, + "learning_rate": 0.00011242443941788021, + "loss": 2.5728, + "step": 21660 + }, + { + "epoch": 0.92, + "grad_norm": 0.44921875, + "learning_rate": 0.00011241778343382368, + "loss": 2.527, + "step": 21661 + }, + { + "epoch": 0.92, + "grad_norm": 0.423828125, + "learning_rate": 0.00011241112739389164, + "loss": 2.6302, + "step": 21662 + }, + { + "epoch": 0.92, + "grad_norm": 0.3828125, + "learning_rate": 0.00011240447129811409, + "loss": 2.6008, + "step": 21663 + }, + { + "epoch": 0.92, + "grad_norm": 0.46875, + "learning_rate": 0.00011239781514652095, + "loss": 2.5416, + "step": 21664 + }, + { + "epoch": 0.92, + "grad_norm": 0.39453125, + "learning_rate": 0.00011239115893914211, + "loss": 2.4106, + "step": 21665 + }, + { + "epoch": 0.92, + "grad_norm": 0.435546875, + "learning_rate": 0.00011238450267600768, + "loss": 2.4064, + "step": 21666 + }, + { + "epoch": 0.92, + "grad_norm": 0.50390625, + "learning_rate": 0.00011237784635714745, + "loss": 2.6091, + "step": 21667 + }, + { + "epoch": 0.92, + "grad_norm": 0.4453125, + "learning_rate": 0.00011237118998259147, + "loss": 2.5813, + "step": 21668 + }, + { + "epoch": 0.92, + "grad_norm": 0.470703125, + "learning_rate": 0.00011236453355236963, + "loss": 2.4723, + "step": 21669 + }, + { + "epoch": 0.92, + "grad_norm": 0.44921875, + "learning_rate": 0.0001123578770665119, + "loss": 2.571, + "step": 21670 + }, + { + "epoch": 0.92, + "grad_norm": 0.4140625, + "learning_rate": 0.00011235122052504828, + "loss": 2.619, + "step": 21671 + }, + { + "epoch": 0.92, + "grad_norm": 0.482421875, + "learning_rate": 0.00011234456392800865, + "loss": 2.5111, + "step": 21672 + }, + { + "epoch": 0.92, + "grad_norm": 0.447265625, + "learning_rate": 0.000112337907275423, + "loss": 2.4457, + "step": 21673 + }, + { + "epoch": 0.92, + "grad_norm": 0.455078125, + "learning_rate": 0.0001123312505673213, + "loss": 2.6732, + "step": 21674 + }, + { + "epoch": 0.92, + "grad_norm": 0.44140625, + "learning_rate": 0.00011232459380373343, + "loss": 2.4676, + "step": 21675 + }, + { + "epoch": 0.92, + "grad_norm": 0.5078125, + "learning_rate": 0.00011231793698468942, + "loss": 2.5093, + "step": 21676 + }, + { + "epoch": 0.92, + "grad_norm": 0.455078125, + "learning_rate": 0.0001123112801102192, + "loss": 2.6557, + "step": 21677 + }, + { + "epoch": 0.92, + "grad_norm": 0.470703125, + "learning_rate": 0.00011230462318035273, + "loss": 2.7262, + "step": 21678 + }, + { + "epoch": 0.92, + "grad_norm": 0.408203125, + "learning_rate": 0.00011229796619511994, + "loss": 2.5289, + "step": 21679 + }, + { + "epoch": 0.92, + "grad_norm": 0.439453125, + "learning_rate": 0.0001122913091545508, + "loss": 2.4938, + "step": 21680 + }, + { + "epoch": 0.92, + "grad_norm": 0.447265625, + "learning_rate": 0.00011228465205867523, + "loss": 2.5059, + "step": 21681 + }, + { + "epoch": 0.92, + "grad_norm": 0.41015625, + "learning_rate": 0.00011227799490752326, + "loss": 2.4677, + "step": 21682 + }, + { + "epoch": 0.92, + "grad_norm": 0.423828125, + "learning_rate": 0.00011227133770112476, + "loss": 2.5643, + "step": 21683 + }, + { + "epoch": 0.92, + "grad_norm": 0.484375, + "learning_rate": 0.00011226468043950974, + "loss": 2.5328, + "step": 21684 + }, + { + "epoch": 0.92, + "grad_norm": 0.498046875, + "learning_rate": 0.00011225802312270814, + "loss": 2.6449, + "step": 21685 + }, + { + "epoch": 0.92, + "grad_norm": 0.44921875, + "learning_rate": 0.00011225136575074992, + "loss": 2.4606, + "step": 21686 + }, + { + "epoch": 0.92, + "grad_norm": 0.4375, + "learning_rate": 0.00011224470832366505, + "loss": 2.5731, + "step": 21687 + }, + { + "epoch": 0.92, + "grad_norm": 0.43359375, + "learning_rate": 0.00011223805084148344, + "loss": 2.5489, + "step": 21688 + }, + { + "epoch": 0.92, + "grad_norm": 0.54296875, + "learning_rate": 0.00011223139330423506, + "loss": 2.724, + "step": 21689 + }, + { + "epoch": 0.92, + "grad_norm": 0.458984375, + "learning_rate": 0.00011222473571194994, + "loss": 2.5252, + "step": 21690 + }, + { + "epoch": 0.92, + "grad_norm": 0.5078125, + "learning_rate": 0.00011221807806465792, + "loss": 2.7556, + "step": 21691 + }, + { + "epoch": 0.92, + "grad_norm": 0.46484375, + "learning_rate": 0.00011221142036238904, + "loss": 2.5113, + "step": 21692 + }, + { + "epoch": 0.92, + "grad_norm": 0.48046875, + "learning_rate": 0.00011220476260517324, + "loss": 2.5891, + "step": 21693 + }, + { + "epoch": 0.92, + "grad_norm": 0.5, + "learning_rate": 0.00011219810479304045, + "loss": 2.5901, + "step": 21694 + }, + { + "epoch": 0.92, + "grad_norm": 0.4765625, + "learning_rate": 0.00011219144692602064, + "loss": 2.5822, + "step": 21695 + }, + { + "epoch": 0.92, + "grad_norm": 0.451171875, + "learning_rate": 0.0001121847890041438, + "loss": 2.5206, + "step": 21696 + }, + { + "epoch": 0.92, + "grad_norm": 0.51171875, + "learning_rate": 0.00011217813102743982, + "loss": 2.6398, + "step": 21697 + }, + { + "epoch": 0.92, + "grad_norm": 0.494140625, + "learning_rate": 0.00011217147299593876, + "loss": 2.6002, + "step": 21698 + }, + { + "epoch": 0.92, + "grad_norm": 0.376953125, + "learning_rate": 0.00011216481490967048, + "loss": 2.5101, + "step": 21699 + }, + { + "epoch": 0.92, + "grad_norm": 0.546875, + "learning_rate": 0.00011215815676866496, + "loss": 2.6287, + "step": 21700 + }, + { + "epoch": 0.92, + "grad_norm": 0.466796875, + "learning_rate": 0.00011215149857295224, + "loss": 2.6683, + "step": 21701 + }, + { + "epoch": 0.92, + "grad_norm": 0.45703125, + "learning_rate": 0.00011214484032256216, + "loss": 2.5931, + "step": 21702 + }, + { + "epoch": 0.92, + "grad_norm": 0.47265625, + "learning_rate": 0.00011213818201752475, + "loss": 2.5556, + "step": 21703 + }, + { + "epoch": 0.92, + "grad_norm": 0.42578125, + "learning_rate": 0.00011213152365786998, + "loss": 2.4847, + "step": 21704 + }, + { + "epoch": 0.92, + "grad_norm": 0.45703125, + "learning_rate": 0.00011212486524362778, + "loss": 2.5046, + "step": 21705 + }, + { + "epoch": 0.92, + "grad_norm": 0.416015625, + "learning_rate": 0.0001121182067748281, + "loss": 2.4191, + "step": 21706 + }, + { + "epoch": 0.92, + "grad_norm": 0.4609375, + "learning_rate": 0.00011211154825150094, + "loss": 2.5332, + "step": 21707 + }, + { + "epoch": 0.93, + "grad_norm": 0.44140625, + "learning_rate": 0.00011210488967367621, + "loss": 2.4592, + "step": 21708 + }, + { + "epoch": 0.93, + "grad_norm": 0.40625, + "learning_rate": 0.00011209823104138392, + "loss": 2.3978, + "step": 21709 + }, + { + "epoch": 0.93, + "grad_norm": 0.453125, + "learning_rate": 0.000112091572354654, + "loss": 2.5836, + "step": 21710 + }, + { + "epoch": 0.93, + "grad_norm": 0.396484375, + "learning_rate": 0.00011208491361351643, + "loss": 2.4199, + "step": 21711 + }, + { + "epoch": 0.93, + "grad_norm": 0.49609375, + "learning_rate": 0.00011207825481800118, + "loss": 2.6032, + "step": 21712 + }, + { + "epoch": 0.93, + "grad_norm": 0.439453125, + "learning_rate": 0.0001120715959681382, + "loss": 2.6116, + "step": 21713 + }, + { + "epoch": 0.93, + "grad_norm": 0.458984375, + "learning_rate": 0.00011206493706395742, + "loss": 2.6617, + "step": 21714 + }, + { + "epoch": 0.93, + "grad_norm": 0.5078125, + "learning_rate": 0.00011205827810548885, + "loss": 2.6714, + "step": 21715 + }, + { + "epoch": 0.93, + "grad_norm": 0.439453125, + "learning_rate": 0.00011205161909276242, + "loss": 2.4635, + "step": 21716 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.00011204496002580816, + "loss": 2.5987, + "step": 21717 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.0001120383009046559, + "loss": 2.5098, + "step": 21718 + }, + { + "epoch": 0.93, + "grad_norm": 0.404296875, + "learning_rate": 0.00011203164172933574, + "loss": 2.5812, + "step": 21719 + }, + { + "epoch": 0.93, + "grad_norm": 0.44140625, + "learning_rate": 0.00011202498249987758, + "loss": 2.5492, + "step": 21720 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.0001120183232163114, + "loss": 2.5574, + "step": 21721 + }, + { + "epoch": 0.93, + "grad_norm": 0.466796875, + "learning_rate": 0.00011201166387866715, + "loss": 2.5157, + "step": 21722 + }, + { + "epoch": 0.93, + "grad_norm": 0.404296875, + "learning_rate": 0.00011200500448697479, + "loss": 2.5083, + "step": 21723 + }, + { + "epoch": 0.93, + "grad_norm": 0.416015625, + "learning_rate": 0.0001119983450412643, + "loss": 2.4851, + "step": 21724 + }, + { + "epoch": 0.93, + "grad_norm": 0.46484375, + "learning_rate": 0.00011199168554156567, + "loss": 2.5344, + "step": 21725 + }, + { + "epoch": 0.93, + "grad_norm": 0.419921875, + "learning_rate": 0.00011198502598790881, + "loss": 2.6602, + "step": 21726 + }, + { + "epoch": 0.93, + "grad_norm": 0.482421875, + "learning_rate": 0.00011197836638032371, + "loss": 2.5346, + "step": 21727 + }, + { + "epoch": 0.93, + "grad_norm": 0.380859375, + "learning_rate": 0.00011197170671884037, + "loss": 2.4373, + "step": 21728 + }, + { + "epoch": 0.93, + "grad_norm": 0.421875, + "learning_rate": 0.00011196504700348868, + "loss": 2.6194, + "step": 21729 + }, + { + "epoch": 0.93, + "grad_norm": 0.421875, + "learning_rate": 0.0001119583872342987, + "loss": 2.5563, + "step": 21730 + }, + { + "epoch": 0.93, + "grad_norm": 0.5078125, + "learning_rate": 0.00011195172741130032, + "loss": 2.6708, + "step": 21731 + }, + { + "epoch": 0.93, + "grad_norm": 0.41796875, + "learning_rate": 0.00011194506753452352, + "loss": 2.5173, + "step": 21732 + }, + { + "epoch": 0.93, + "grad_norm": 0.443359375, + "learning_rate": 0.00011193840760399832, + "loss": 2.5067, + "step": 21733 + }, + { + "epoch": 0.93, + "grad_norm": 0.4765625, + "learning_rate": 0.00011193174761975462, + "loss": 2.5061, + "step": 21734 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011192508758182243, + "loss": 2.5307, + "step": 21735 + }, + { + "epoch": 0.93, + "grad_norm": 0.4453125, + "learning_rate": 0.0001119184274902317, + "loss": 2.4517, + "step": 21736 + }, + { + "epoch": 0.93, + "grad_norm": 0.4296875, + "learning_rate": 0.00011191176734501238, + "loss": 2.5818, + "step": 21737 + }, + { + "epoch": 0.93, + "grad_norm": 0.47265625, + "learning_rate": 0.00011190510714619451, + "loss": 2.4296, + "step": 21738 + }, + { + "epoch": 0.93, + "grad_norm": 0.439453125, + "learning_rate": 0.00011189844689380798, + "loss": 2.6417, + "step": 21739 + }, + { + "epoch": 0.93, + "grad_norm": 0.458984375, + "learning_rate": 0.00011189178658788275, + "loss": 2.62, + "step": 21740 + }, + { + "epoch": 0.93, + "grad_norm": 0.462890625, + "learning_rate": 0.0001118851262284489, + "loss": 2.532, + "step": 21741 + }, + { + "epoch": 0.93, + "grad_norm": 0.4921875, + "learning_rate": 0.00011187846581553626, + "loss": 2.5989, + "step": 21742 + }, + { + "epoch": 0.93, + "grad_norm": 0.4375, + "learning_rate": 0.0001118718053491749, + "loss": 2.4204, + "step": 21743 + }, + { + "epoch": 0.93, + "grad_norm": 0.427734375, + "learning_rate": 0.00011186514482939476, + "loss": 2.5857, + "step": 21744 + }, + { + "epoch": 0.93, + "grad_norm": 0.466796875, + "learning_rate": 0.00011185848425622578, + "loss": 2.6339, + "step": 21745 + }, + { + "epoch": 0.93, + "grad_norm": 0.443359375, + "learning_rate": 0.00011185182362969797, + "loss": 2.4898, + "step": 21746 + }, + { + "epoch": 0.93, + "grad_norm": 0.4609375, + "learning_rate": 0.00011184516294984128, + "loss": 2.5376, + "step": 21747 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.00011183850221668569, + "loss": 2.5277, + "step": 21748 + }, + { + "epoch": 0.93, + "grad_norm": 0.490234375, + "learning_rate": 0.00011183184143026118, + "loss": 2.5453, + "step": 21749 + }, + { + "epoch": 0.93, + "grad_norm": 0.4609375, + "learning_rate": 0.00011182518059059767, + "loss": 2.4855, + "step": 21750 + }, + { + "epoch": 0.93, + "grad_norm": 0.408203125, + "learning_rate": 0.00011181851969772521, + "loss": 2.5876, + "step": 21751 + }, + { + "epoch": 0.93, + "grad_norm": 0.47265625, + "learning_rate": 0.0001118118587516737, + "loss": 2.5835, + "step": 21752 + }, + { + "epoch": 0.93, + "grad_norm": 0.4453125, + "learning_rate": 0.00011180519775247316, + "loss": 2.4025, + "step": 21753 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011179853670015355, + "loss": 2.5706, + "step": 21754 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011179187559474482, + "loss": 2.5183, + "step": 21755 + }, + { + "epoch": 0.93, + "grad_norm": 0.466796875, + "learning_rate": 0.00011178521443627694, + "loss": 2.4604, + "step": 21756 + }, + { + "epoch": 0.93, + "grad_norm": 0.4453125, + "learning_rate": 0.00011177855322477995, + "loss": 2.5949, + "step": 21757 + }, + { + "epoch": 0.93, + "grad_norm": 0.419921875, + "learning_rate": 0.00011177189196028375, + "loss": 2.6346, + "step": 21758 + }, + { + "epoch": 0.93, + "grad_norm": 0.46875, + "learning_rate": 0.00011176523064281835, + "loss": 2.6248, + "step": 21759 + }, + { + "epoch": 0.93, + "grad_norm": 0.46484375, + "learning_rate": 0.00011175856927241368, + "loss": 2.6471, + "step": 21760 + }, + { + "epoch": 0.93, + "grad_norm": 0.4140625, + "learning_rate": 0.00011175190784909976, + "loss": 2.5201, + "step": 21761 + }, + { + "epoch": 0.93, + "grad_norm": 0.42578125, + "learning_rate": 0.00011174524637290658, + "loss": 2.5349, + "step": 21762 + }, + { + "epoch": 0.93, + "grad_norm": 0.478515625, + "learning_rate": 0.00011173858484386406, + "loss": 2.4803, + "step": 21763 + }, + { + "epoch": 0.93, + "grad_norm": 0.48828125, + "learning_rate": 0.00011173192326200217, + "loss": 2.5412, + "step": 21764 + }, + { + "epoch": 0.93, + "grad_norm": 0.46875, + "learning_rate": 0.00011172526162735097, + "loss": 2.667, + "step": 21765 + }, + { + "epoch": 0.93, + "grad_norm": 0.40234375, + "learning_rate": 0.00011171859993994034, + "loss": 2.4489, + "step": 21766 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.00011171193819980031, + "loss": 2.3795, + "step": 21767 + }, + { + "epoch": 0.93, + "grad_norm": 0.4375, + "learning_rate": 0.00011170527640696083, + "loss": 2.6025, + "step": 21768 + }, + { + "epoch": 0.93, + "grad_norm": 0.494140625, + "learning_rate": 0.00011169861456145187, + "loss": 2.5441, + "step": 21769 + }, + { + "epoch": 0.93, + "grad_norm": 0.470703125, + "learning_rate": 0.00011169195266330344, + "loss": 2.6178, + "step": 21770 + }, + { + "epoch": 0.93, + "grad_norm": 0.48046875, + "learning_rate": 0.00011168529071254549, + "loss": 2.6043, + "step": 21771 + }, + { + "epoch": 0.93, + "grad_norm": 0.451171875, + "learning_rate": 0.000111678628709208, + "loss": 2.5271, + "step": 21772 + }, + { + "epoch": 0.93, + "grad_norm": 0.42578125, + "learning_rate": 0.00011167196665332096, + "loss": 2.5245, + "step": 21773 + }, + { + "epoch": 0.93, + "grad_norm": 0.46875, + "learning_rate": 0.00011166530454491431, + "loss": 2.563, + "step": 21774 + }, + { + "epoch": 0.93, + "grad_norm": 0.484375, + "learning_rate": 0.00011165864238401808, + "loss": 2.4648, + "step": 21775 + }, + { + "epoch": 0.93, + "grad_norm": 0.53125, + "learning_rate": 0.00011165198017066221, + "loss": 2.4241, + "step": 21776 + }, + { + "epoch": 0.93, + "grad_norm": 0.50390625, + "learning_rate": 0.00011164531790487669, + "loss": 2.4973, + "step": 21777 + }, + { + "epoch": 0.93, + "grad_norm": 0.44921875, + "learning_rate": 0.00011163865558669149, + "loss": 2.3942, + "step": 21778 + }, + { + "epoch": 0.93, + "grad_norm": 0.53125, + "learning_rate": 0.0001116319932161366, + "loss": 2.5761, + "step": 21779 + }, + { + "epoch": 0.93, + "grad_norm": 0.4921875, + "learning_rate": 0.00011162533079324198, + "loss": 2.6176, + "step": 21780 + }, + { + "epoch": 0.93, + "grad_norm": 0.4453125, + "learning_rate": 0.00011161866831803764, + "loss": 2.5176, + "step": 21781 + }, + { + "epoch": 0.93, + "grad_norm": 0.42578125, + "learning_rate": 0.0001116120057905535, + "loss": 2.4974, + "step": 21782 + }, + { + "epoch": 0.93, + "grad_norm": 0.52734375, + "learning_rate": 0.00011160534321081961, + "loss": 2.6083, + "step": 21783 + }, + { + "epoch": 0.93, + "grad_norm": 0.474609375, + "learning_rate": 0.00011159868057886591, + "loss": 2.5749, + "step": 21784 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.0001115920178947224, + "loss": 2.6288, + "step": 21785 + }, + { + "epoch": 0.93, + "grad_norm": 0.490234375, + "learning_rate": 0.00011158535515841904, + "loss": 2.5325, + "step": 21786 + }, + { + "epoch": 0.93, + "grad_norm": 0.49609375, + "learning_rate": 0.00011157869236998582, + "loss": 2.6756, + "step": 21787 + }, + { + "epoch": 0.93, + "grad_norm": 0.40234375, + "learning_rate": 0.00011157202952945268, + "loss": 2.4755, + "step": 21788 + }, + { + "epoch": 0.93, + "grad_norm": 0.466796875, + "learning_rate": 0.00011156536663684967, + "loss": 2.554, + "step": 21789 + }, + { + "epoch": 0.93, + "grad_norm": 0.447265625, + "learning_rate": 0.00011155870369220671, + "loss": 2.5538, + "step": 21790 + }, + { + "epoch": 0.93, + "grad_norm": 0.451171875, + "learning_rate": 0.00011155204069555383, + "loss": 2.4342, + "step": 21791 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.000111545377646921, + "loss": 2.6053, + "step": 21792 + }, + { + "epoch": 0.93, + "grad_norm": 0.451171875, + "learning_rate": 0.00011153871454633817, + "loss": 2.4993, + "step": 21793 + }, + { + "epoch": 0.93, + "grad_norm": 0.47265625, + "learning_rate": 0.00011153205139383533, + "loss": 2.5537, + "step": 21794 + }, + { + "epoch": 0.93, + "grad_norm": 0.462890625, + "learning_rate": 0.00011152538818944252, + "loss": 2.5452, + "step": 21795 + }, + { + "epoch": 0.93, + "grad_norm": 0.41015625, + "learning_rate": 0.0001115187249331896, + "loss": 2.4883, + "step": 21796 + }, + { + "epoch": 0.93, + "grad_norm": 0.4921875, + "learning_rate": 0.0001115120616251067, + "loss": 2.6424, + "step": 21797 + }, + { + "epoch": 0.93, + "grad_norm": 0.38671875, + "learning_rate": 0.00011150539826522368, + "loss": 2.5554, + "step": 21798 + }, + { + "epoch": 0.93, + "grad_norm": 0.38671875, + "learning_rate": 0.00011149873485357061, + "loss": 2.4757, + "step": 21799 + }, + { + "epoch": 0.93, + "grad_norm": 0.455078125, + "learning_rate": 0.00011149207139017742, + "loss": 2.5015, + "step": 21800 + }, + { + "epoch": 0.93, + "grad_norm": 0.515625, + "learning_rate": 0.00011148540787507411, + "loss": 2.5619, + "step": 21801 + }, + { + "epoch": 0.93, + "grad_norm": 0.470703125, + "learning_rate": 0.00011147874430829067, + "loss": 2.4373, + "step": 21802 + }, + { + "epoch": 0.93, + "grad_norm": 0.490234375, + "learning_rate": 0.00011147208068985704, + "loss": 2.6767, + "step": 21803 + }, + { + "epoch": 0.93, + "grad_norm": 0.4921875, + "learning_rate": 0.00011146541701980327, + "loss": 2.524, + "step": 21804 + }, + { + "epoch": 0.93, + "grad_norm": 0.423828125, + "learning_rate": 0.00011145875329815931, + "loss": 2.5437, + "step": 21805 + }, + { + "epoch": 0.93, + "grad_norm": 0.453125, + "learning_rate": 0.00011145208952495513, + "loss": 2.6491, + "step": 21806 + }, + { + "epoch": 0.93, + "grad_norm": 0.48828125, + "learning_rate": 0.00011144542570022076, + "loss": 2.6351, + "step": 21807 + }, + { + "epoch": 0.93, + "grad_norm": 0.46484375, + "learning_rate": 0.00011143876182398614, + "loss": 2.5339, + "step": 21808 + }, + { + "epoch": 0.93, + "grad_norm": 0.412109375, + "learning_rate": 0.00011143209789628129, + "loss": 2.5077, + "step": 21809 + }, + { + "epoch": 0.93, + "grad_norm": 0.52734375, + "learning_rate": 0.00011142543391713616, + "loss": 2.5541, + "step": 21810 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.00011141876988658075, + "loss": 2.555, + "step": 21811 + }, + { + "epoch": 0.93, + "grad_norm": 0.50390625, + "learning_rate": 0.00011141210580464504, + "loss": 2.6554, + "step": 21812 + }, + { + "epoch": 0.93, + "grad_norm": 0.3828125, + "learning_rate": 0.00011140544167135907, + "loss": 2.3707, + "step": 21813 + }, + { + "epoch": 0.93, + "grad_norm": 0.4921875, + "learning_rate": 0.00011139877748675273, + "loss": 2.6629, + "step": 21814 + }, + { + "epoch": 0.93, + "grad_norm": 0.423828125, + "learning_rate": 0.00011139211325085607, + "loss": 2.5862, + "step": 21815 + }, + { + "epoch": 0.93, + "grad_norm": 0.423828125, + "learning_rate": 0.0001113854489636991, + "loss": 2.6362, + "step": 21816 + }, + { + "epoch": 0.93, + "grad_norm": 0.40625, + "learning_rate": 0.00011137878462531173, + "loss": 2.4114, + "step": 21817 + }, + { + "epoch": 0.93, + "grad_norm": 0.439453125, + "learning_rate": 0.00011137212023572401, + "loss": 2.6607, + "step": 21818 + }, + { + "epoch": 0.93, + "grad_norm": 0.435546875, + "learning_rate": 0.00011136545579496589, + "loss": 2.6311, + "step": 21819 + }, + { + "epoch": 0.93, + "grad_norm": 0.498046875, + "learning_rate": 0.00011135879130306738, + "loss": 2.5146, + "step": 21820 + }, + { + "epoch": 0.93, + "grad_norm": 0.41796875, + "learning_rate": 0.00011135212676005848, + "loss": 2.5658, + "step": 21821 + }, + { + "epoch": 0.93, + "grad_norm": 0.466796875, + "learning_rate": 0.00011134546216596912, + "loss": 2.5556, + "step": 21822 + }, + { + "epoch": 0.93, + "grad_norm": 0.54296875, + "learning_rate": 0.00011133879752082933, + "loss": 2.4725, + "step": 21823 + }, + { + "epoch": 0.93, + "grad_norm": 0.48828125, + "learning_rate": 0.00011133213282466915, + "loss": 2.6028, + "step": 21824 + }, + { + "epoch": 0.93, + "grad_norm": 0.412109375, + "learning_rate": 0.00011132546807751845, + "loss": 2.5635, + "step": 21825 + }, + { + "epoch": 0.93, + "grad_norm": 0.50390625, + "learning_rate": 0.00011131880327940729, + "loss": 2.6253, + "step": 21826 + }, + { + "epoch": 0.93, + "grad_norm": 0.484375, + "learning_rate": 0.00011131213843036568, + "loss": 2.4324, + "step": 21827 + }, + { + "epoch": 0.93, + "grad_norm": 0.447265625, + "learning_rate": 0.00011130547353042358, + "loss": 2.669, + "step": 21828 + }, + { + "epoch": 0.93, + "grad_norm": 0.435546875, + "learning_rate": 0.00011129880857961097, + "loss": 2.4668, + "step": 21829 + }, + { + "epoch": 0.93, + "grad_norm": 0.41796875, + "learning_rate": 0.00011129214357795784, + "loss": 2.4878, + "step": 21830 + }, + { + "epoch": 0.93, + "grad_norm": 0.48828125, + "learning_rate": 0.00011128547852549418, + "loss": 2.5566, + "step": 21831 + }, + { + "epoch": 0.93, + "grad_norm": 0.451171875, + "learning_rate": 0.00011127881342225004, + "loss": 2.4391, + "step": 21832 + }, + { + "epoch": 0.93, + "grad_norm": 0.41796875, + "learning_rate": 0.00011127214826825531, + "loss": 2.5528, + "step": 21833 + }, + { + "epoch": 0.93, + "grad_norm": 0.423828125, + "learning_rate": 0.00011126548306354006, + "loss": 2.5324, + "step": 21834 + }, + { + "epoch": 0.93, + "grad_norm": 0.494140625, + "learning_rate": 0.00011125881780813425, + "loss": 2.4712, + "step": 21835 + }, + { + "epoch": 0.93, + "grad_norm": 0.46484375, + "learning_rate": 0.00011125215250206786, + "loss": 2.5176, + "step": 21836 + }, + { + "epoch": 0.93, + "grad_norm": 0.44921875, + "learning_rate": 0.00011124548714537093, + "loss": 2.5323, + "step": 21837 + }, + { + "epoch": 0.93, + "grad_norm": 0.48046875, + "learning_rate": 0.00011123882173807339, + "loss": 2.5637, + "step": 21838 + }, + { + "epoch": 0.93, + "grad_norm": 0.490234375, + "learning_rate": 0.00011123215628020524, + "loss": 2.6713, + "step": 21839 + }, + { + "epoch": 0.93, + "grad_norm": 0.470703125, + "learning_rate": 0.00011122549077179655, + "loss": 2.5564, + "step": 21840 + }, + { + "epoch": 0.93, + "grad_norm": 0.453125, + "learning_rate": 0.0001112188252128772, + "loss": 2.5503, + "step": 21841 + }, + { + "epoch": 0.93, + "grad_norm": 0.4296875, + "learning_rate": 0.00011121215960347726, + "loss": 2.6072, + "step": 21842 + }, + { + "epoch": 0.93, + "grad_norm": 0.498046875, + "learning_rate": 0.0001112054939436267, + "loss": 2.5439, + "step": 21843 + }, + { + "epoch": 0.93, + "grad_norm": 0.478515625, + "learning_rate": 0.00011119882823335552, + "loss": 2.4409, + "step": 21844 + }, + { + "epoch": 0.93, + "grad_norm": 0.396484375, + "learning_rate": 0.00011119216247269368, + "loss": 2.4283, + "step": 21845 + }, + { + "epoch": 0.93, + "grad_norm": 0.51953125, + "learning_rate": 0.00011118549666167122, + "loss": 2.6686, + "step": 21846 + }, + { + "epoch": 0.93, + "grad_norm": 0.478515625, + "learning_rate": 0.0001111788308003181, + "loss": 2.5225, + "step": 21847 + }, + { + "epoch": 0.93, + "grad_norm": 0.458984375, + "learning_rate": 0.00011117216488866435, + "loss": 2.5386, + "step": 21848 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011116549892673991, + "loss": 2.5205, + "step": 21849 + }, + { + "epoch": 0.93, + "grad_norm": 0.451171875, + "learning_rate": 0.00011115883291457483, + "loss": 2.6353, + "step": 21850 + }, + { + "epoch": 0.93, + "grad_norm": 0.4453125, + "learning_rate": 0.00011115216685219907, + "loss": 2.5985, + "step": 21851 + }, + { + "epoch": 0.93, + "grad_norm": 0.4296875, + "learning_rate": 0.00011114550073964263, + "loss": 2.447, + "step": 21852 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011113883457693552, + "loss": 2.5189, + "step": 21853 + }, + { + "epoch": 0.93, + "grad_norm": 0.41796875, + "learning_rate": 0.00011113216836410771, + "loss": 2.5622, + "step": 21854 + }, + { + "epoch": 0.93, + "grad_norm": 0.484375, + "learning_rate": 0.00011112550210118922, + "loss": 2.5994, + "step": 21855 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011111883578821005, + "loss": 2.368, + "step": 21856 + }, + { + "epoch": 0.93, + "grad_norm": 0.4921875, + "learning_rate": 0.00011111216942520015, + "loss": 2.4428, + "step": 21857 + }, + { + "epoch": 0.93, + "grad_norm": 0.46875, + "learning_rate": 0.00011110550301218959, + "loss": 2.596, + "step": 21858 + }, + { + "epoch": 0.93, + "grad_norm": 0.466796875, + "learning_rate": 0.0001110988365492083, + "loss": 2.5963, + "step": 21859 + }, + { + "epoch": 0.93, + "grad_norm": 0.44140625, + "learning_rate": 0.00011109217003628629, + "loss": 2.6253, + "step": 21860 + }, + { + "epoch": 0.93, + "grad_norm": 0.48828125, + "learning_rate": 0.00011108550347345359, + "loss": 2.5399, + "step": 21861 + }, + { + "epoch": 0.93, + "grad_norm": 0.470703125, + "learning_rate": 0.00011107883686074019, + "loss": 2.5022, + "step": 21862 + }, + { + "epoch": 0.93, + "grad_norm": 0.392578125, + "learning_rate": 0.00011107217019817602, + "loss": 2.6532, + "step": 21863 + }, + { + "epoch": 0.93, + "grad_norm": 0.478515625, + "learning_rate": 0.0001110655034857912, + "loss": 2.6701, + "step": 21864 + }, + { + "epoch": 0.93, + "grad_norm": 0.404296875, + "learning_rate": 0.00011105883672361558, + "loss": 2.5451, + "step": 21865 + }, + { + "epoch": 0.93, + "grad_norm": 0.412109375, + "learning_rate": 0.00011105216991167928, + "loss": 2.607, + "step": 21866 + }, + { + "epoch": 0.93, + "grad_norm": 0.4921875, + "learning_rate": 0.00011104550305001227, + "loss": 2.6732, + "step": 21867 + }, + { + "epoch": 0.93, + "grad_norm": 0.44140625, + "learning_rate": 0.00011103883613864452, + "loss": 2.6372, + "step": 21868 + }, + { + "epoch": 0.93, + "grad_norm": 0.421875, + "learning_rate": 0.00011103216917760604, + "loss": 2.499, + "step": 21869 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.00011102550216692682, + "loss": 2.4308, + "step": 21870 + }, + { + "epoch": 0.93, + "grad_norm": 0.451171875, + "learning_rate": 0.00011101883510663686, + "loss": 2.5785, + "step": 21871 + }, + { + "epoch": 0.93, + "grad_norm": 0.46484375, + "learning_rate": 0.00011101216799676621, + "loss": 2.5083, + "step": 21872 + }, + { + "epoch": 0.93, + "grad_norm": 0.4375, + "learning_rate": 0.00011100550083734478, + "loss": 2.5184, + "step": 21873 + }, + { + "epoch": 0.93, + "grad_norm": 0.5078125, + "learning_rate": 0.00011099883362840265, + "loss": 2.6139, + "step": 21874 + }, + { + "epoch": 0.93, + "grad_norm": 0.42578125, + "learning_rate": 0.00011099216636996979, + "loss": 2.5336, + "step": 21875 + }, + { + "epoch": 0.93, + "grad_norm": 0.5, + "learning_rate": 0.00011098549906207618, + "loss": 2.5874, + "step": 21876 + }, + { + "epoch": 0.93, + "grad_norm": 0.427734375, + "learning_rate": 0.00011097883170475186, + "loss": 2.5089, + "step": 21877 + }, + { + "epoch": 0.93, + "grad_norm": 0.474609375, + "learning_rate": 0.0001109721642980268, + "loss": 2.5456, + "step": 21878 + }, + { + "epoch": 0.93, + "grad_norm": 0.439453125, + "learning_rate": 0.00011096549684193098, + "loss": 2.602, + "step": 21879 + }, + { + "epoch": 0.93, + "grad_norm": 0.4375, + "learning_rate": 0.00011095882933649449, + "loss": 2.664, + "step": 21880 + }, + { + "epoch": 0.93, + "grad_norm": 0.51953125, + "learning_rate": 0.00011095216178174721, + "loss": 2.6082, + "step": 21881 + }, + { + "epoch": 0.93, + "grad_norm": 0.3984375, + "learning_rate": 0.00011094549417771924, + "loss": 2.3396, + "step": 21882 + }, + { + "epoch": 0.93, + "grad_norm": 0.4609375, + "learning_rate": 0.00011093882652444054, + "loss": 2.5634, + "step": 21883 + }, + { + "epoch": 0.93, + "grad_norm": 0.4921875, + "learning_rate": 0.00011093215882194114, + "loss": 2.605, + "step": 21884 + }, + { + "epoch": 0.93, + "grad_norm": 0.431640625, + "learning_rate": 0.000110925491070251, + "loss": 2.5281, + "step": 21885 + }, + { + "epoch": 0.93, + "grad_norm": 0.421875, + "learning_rate": 0.00011091882326940015, + "loss": 2.4735, + "step": 21886 + }, + { + "epoch": 0.93, + "grad_norm": 0.447265625, + "learning_rate": 0.00011091215541941856, + "loss": 2.5517, + "step": 21887 + }, + { + "epoch": 0.93, + "grad_norm": 0.474609375, + "learning_rate": 0.00011090548752033631, + "loss": 2.6444, + "step": 21888 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.0001108988195721833, + "loss": 2.5068, + "step": 21889 + }, + { + "epoch": 0.93, + "grad_norm": 0.40625, + "learning_rate": 0.0001108921515749896, + "loss": 2.5265, + "step": 21890 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.0001108854835287852, + "loss": 2.6656, + "step": 21891 + }, + { + "epoch": 0.93, + "grad_norm": 0.408203125, + "learning_rate": 0.00011087881543360012, + "loss": 2.6456, + "step": 21892 + }, + { + "epoch": 0.93, + "grad_norm": 0.466796875, + "learning_rate": 0.00011087214728946434, + "loss": 2.6717, + "step": 21893 + }, + { + "epoch": 0.93, + "grad_norm": 0.4765625, + "learning_rate": 0.00011086547909640788, + "loss": 2.6241, + "step": 21894 + }, + { + "epoch": 0.93, + "grad_norm": 0.447265625, + "learning_rate": 0.00011085881085446068, + "loss": 2.5754, + "step": 21895 + }, + { + "epoch": 0.93, + "grad_norm": 0.41796875, + "learning_rate": 0.00011085214256365286, + "loss": 2.5772, + "step": 21896 + }, + { + "epoch": 0.93, + "grad_norm": 0.47265625, + "learning_rate": 0.00011084547422401432, + "loss": 2.5919, + "step": 21897 + }, + { + "epoch": 0.93, + "grad_norm": 0.455078125, + "learning_rate": 0.00011083880583557516, + "loss": 2.5901, + "step": 21898 + }, + { + "epoch": 0.93, + "grad_norm": 0.482421875, + "learning_rate": 0.0001108321373983653, + "loss": 2.5601, + "step": 21899 + }, + { + "epoch": 0.93, + "grad_norm": 0.388671875, + "learning_rate": 0.00011082546891241478, + "loss": 2.456, + "step": 21900 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011081880037775362, + "loss": 2.55, + "step": 21901 + }, + { + "epoch": 0.93, + "grad_norm": 0.46484375, + "learning_rate": 0.0001108121317944118, + "loss": 2.4897, + "step": 21902 + }, + { + "epoch": 0.93, + "grad_norm": 0.458984375, + "learning_rate": 0.00011080546316241933, + "loss": 2.641, + "step": 21903 + }, + { + "epoch": 0.93, + "grad_norm": 0.5078125, + "learning_rate": 0.00011079879448180628, + "loss": 2.5599, + "step": 21904 + }, + { + "epoch": 0.93, + "grad_norm": 0.427734375, + "learning_rate": 0.00011079212575260254, + "loss": 2.518, + "step": 21905 + }, + { + "epoch": 0.93, + "grad_norm": 0.451171875, + "learning_rate": 0.0001107854569748382, + "loss": 2.6794, + "step": 21906 + }, + { + "epoch": 0.93, + "grad_norm": 0.43359375, + "learning_rate": 0.00011077878814854323, + "loss": 2.5884, + "step": 21907 + }, + { + "epoch": 0.93, + "grad_norm": 0.478515625, + "learning_rate": 0.00011077211927374769, + "loss": 2.5427, + "step": 21908 + }, + { + "epoch": 0.93, + "grad_norm": 0.439453125, + "learning_rate": 0.00011076545035048152, + "loss": 2.4184, + "step": 21909 + }, + { + "epoch": 0.93, + "grad_norm": 0.451171875, + "learning_rate": 0.00011075878137877475, + "loss": 2.4579, + "step": 21910 + }, + { + "epoch": 0.93, + "grad_norm": 0.458984375, + "learning_rate": 0.00011075211235865741, + "loss": 2.5335, + "step": 21911 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011074544329015953, + "loss": 2.5092, + "step": 21912 + }, + { + "epoch": 0.93, + "grad_norm": 0.4765625, + "learning_rate": 0.000110738774173311, + "loss": 2.5806, + "step": 21913 + }, + { + "epoch": 0.93, + "grad_norm": 0.5390625, + "learning_rate": 0.00011073210500814198, + "loss": 2.6776, + "step": 21914 + }, + { + "epoch": 0.93, + "grad_norm": 0.447265625, + "learning_rate": 0.00011072543579468237, + "loss": 2.6405, + "step": 21915 + }, + { + "epoch": 0.93, + "grad_norm": 0.45703125, + "learning_rate": 0.00011071876653296225, + "loss": 2.653, + "step": 21916 + }, + { + "epoch": 0.93, + "grad_norm": 0.455078125, + "learning_rate": 0.00011071209722301156, + "loss": 2.4607, + "step": 21917 + }, + { + "epoch": 0.93, + "grad_norm": 0.5078125, + "learning_rate": 0.0001107054278648604, + "loss": 2.64, + "step": 21918 + }, + { + "epoch": 0.93, + "grad_norm": 0.462890625, + "learning_rate": 0.00011069875845853868, + "loss": 2.5653, + "step": 21919 + }, + { + "epoch": 0.93, + "grad_norm": 0.498046875, + "learning_rate": 0.00011069208900407648, + "loss": 2.5999, + "step": 21920 + }, + { + "epoch": 0.93, + "grad_norm": 0.5234375, + "learning_rate": 0.00011068541950150377, + "loss": 2.6034, + "step": 21921 + }, + { + "epoch": 0.93, + "grad_norm": 0.51171875, + "learning_rate": 0.00011067874995085059, + "loss": 2.5504, + "step": 21922 + }, + { + "epoch": 0.93, + "grad_norm": 0.47265625, + "learning_rate": 0.00011067208035214693, + "loss": 2.5326, + "step": 21923 + }, + { + "epoch": 0.93, + "grad_norm": 0.453125, + "learning_rate": 0.00011066541070542282, + "loss": 2.6262, + "step": 21924 + }, + { + "epoch": 0.93, + "grad_norm": 0.44921875, + "learning_rate": 0.00011065874101070826, + "loss": 2.5616, + "step": 21925 + }, + { + "epoch": 0.93, + "grad_norm": 0.48828125, + "learning_rate": 0.00011065207126803325, + "loss": 2.5935, + "step": 21926 + }, + { + "epoch": 0.93, + "grad_norm": 0.490234375, + "learning_rate": 0.00011064540147742781, + "loss": 2.5467, + "step": 21927 + }, + { + "epoch": 0.93, + "grad_norm": 0.41796875, + "learning_rate": 0.00011063873163892198, + "loss": 2.3882, + "step": 21928 + }, + { + "epoch": 0.93, + "grad_norm": 0.455078125, + "learning_rate": 0.0001106320617525457, + "loss": 2.5902, + "step": 21929 + }, + { + "epoch": 0.93, + "grad_norm": 0.458984375, + "learning_rate": 0.00011062539181832908, + "loss": 2.4213, + "step": 21930 + }, + { + "epoch": 0.93, + "grad_norm": 0.48046875, + "learning_rate": 0.00011061872183630204, + "loss": 2.5327, + "step": 21931 + }, + { + "epoch": 0.93, + "grad_norm": 0.486328125, + "learning_rate": 0.00011061205180649466, + "loss": 2.5701, + "step": 21932 + }, + { + "epoch": 0.93, + "grad_norm": 0.515625, + "learning_rate": 0.0001106053817289369, + "loss": 2.5835, + "step": 21933 + }, + { + "epoch": 0.93, + "grad_norm": 0.50390625, + "learning_rate": 0.00011059871160365883, + "loss": 2.6125, + "step": 21934 + }, + { + "epoch": 0.93, + "grad_norm": 0.52734375, + "learning_rate": 0.00011059204143069037, + "loss": 2.5643, + "step": 21935 + }, + { + "epoch": 0.93, + "grad_norm": 0.439453125, + "learning_rate": 0.00011058537121006167, + "loss": 2.6364, + "step": 21936 + }, + { + "epoch": 0.93, + "grad_norm": 0.4609375, + "learning_rate": 0.00011057870094180263, + "loss": 2.5454, + "step": 21937 + }, + { + "epoch": 0.93, + "grad_norm": 0.48046875, + "learning_rate": 0.00011057203062594331, + "loss": 2.536, + "step": 21938 + }, + { + "epoch": 0.93, + "grad_norm": 0.478515625, + "learning_rate": 0.00011056536026251369, + "loss": 2.5924, + "step": 21939 + }, + { + "epoch": 0.93, + "grad_norm": 0.490234375, + "learning_rate": 0.00011055868985154386, + "loss": 2.5558, + "step": 21940 + }, + { + "epoch": 0.93, + "grad_norm": 0.474609375, + "learning_rate": 0.00011055201939306374, + "loss": 2.6378, + "step": 21941 + }, + { + "epoch": 0.94, + "grad_norm": 0.427734375, + "learning_rate": 0.00011054534888710342, + "loss": 2.4936, + "step": 21942 + }, + { + "epoch": 0.94, + "grad_norm": 0.380859375, + "learning_rate": 0.00011053867833369286, + "loss": 2.5188, + "step": 21943 + }, + { + "epoch": 0.94, + "grad_norm": 0.439453125, + "learning_rate": 0.00011053200773286213, + "loss": 2.5279, + "step": 21944 + }, + { + "epoch": 0.94, + "grad_norm": 0.48046875, + "learning_rate": 0.00011052533708464117, + "loss": 2.4829, + "step": 21945 + }, + { + "epoch": 0.94, + "grad_norm": 0.46875, + "learning_rate": 0.00011051866638906005, + "loss": 2.6505, + "step": 21946 + }, + { + "epoch": 0.94, + "grad_norm": 0.458984375, + "learning_rate": 0.0001105119956461488, + "loss": 2.5396, + "step": 21947 + }, + { + "epoch": 0.94, + "grad_norm": 0.41796875, + "learning_rate": 0.00011050532485593738, + "loss": 2.5419, + "step": 21948 + }, + { + "epoch": 0.94, + "grad_norm": 0.416015625, + "learning_rate": 0.00011049865401845587, + "loss": 2.6304, + "step": 21949 + }, + { + "epoch": 0.94, + "grad_norm": 0.462890625, + "learning_rate": 0.00011049198313373424, + "loss": 2.5805, + "step": 21950 + }, + { + "epoch": 0.94, + "grad_norm": 0.470703125, + "learning_rate": 0.00011048531220180252, + "loss": 2.7024, + "step": 21951 + }, + { + "epoch": 0.94, + "grad_norm": 0.490234375, + "learning_rate": 0.00011047864122269074, + "loss": 2.5311, + "step": 21952 + }, + { + "epoch": 0.94, + "grad_norm": 0.50390625, + "learning_rate": 0.00011047197019642887, + "loss": 2.6183, + "step": 21953 + }, + { + "epoch": 0.94, + "grad_norm": 0.38671875, + "learning_rate": 0.00011046529912304698, + "loss": 2.3189, + "step": 21954 + }, + { + "epoch": 0.94, + "grad_norm": 0.466796875, + "learning_rate": 0.00011045862800257509, + "loss": 2.44, + "step": 21955 + }, + { + "epoch": 0.94, + "grad_norm": 0.4765625, + "learning_rate": 0.00011045195683504314, + "loss": 2.5285, + "step": 21956 + }, + { + "epoch": 0.94, + "grad_norm": 0.42578125, + "learning_rate": 0.00011044528562048126, + "loss": 2.7313, + "step": 21957 + }, + { + "epoch": 0.94, + "grad_norm": 0.427734375, + "learning_rate": 0.0001104386143589194, + "loss": 2.5943, + "step": 21958 + }, + { + "epoch": 0.94, + "grad_norm": 0.470703125, + "learning_rate": 0.00011043194305038758, + "loss": 2.4287, + "step": 21959 + }, + { + "epoch": 0.94, + "grad_norm": 0.4375, + "learning_rate": 0.00011042527169491583, + "loss": 2.495, + "step": 21960 + }, + { + "epoch": 0.94, + "grad_norm": 0.46875, + "learning_rate": 0.00011041860029253417, + "loss": 2.4662, + "step": 21961 + }, + { + "epoch": 0.94, + "grad_norm": 0.4453125, + "learning_rate": 0.00011041192884327261, + "loss": 2.508, + "step": 21962 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.00011040525734716121, + "loss": 2.5753, + "step": 21963 + }, + { + "epoch": 0.94, + "grad_norm": 0.4609375, + "learning_rate": 0.00011039858580422991, + "loss": 2.4416, + "step": 21964 + }, + { + "epoch": 0.94, + "grad_norm": 0.4296875, + "learning_rate": 0.00011039191421450879, + "loss": 2.4166, + "step": 21965 + }, + { + "epoch": 0.94, + "grad_norm": 0.51953125, + "learning_rate": 0.00011038524257802785, + "loss": 2.5683, + "step": 21966 + }, + { + "epoch": 0.94, + "grad_norm": 0.412109375, + "learning_rate": 0.00011037857089481712, + "loss": 2.5296, + "step": 21967 + }, + { + "epoch": 0.94, + "grad_norm": 0.408203125, + "learning_rate": 0.0001103718991649066, + "loss": 2.4121, + "step": 21968 + }, + { + "epoch": 0.94, + "grad_norm": 0.51171875, + "learning_rate": 0.00011036522738832635, + "loss": 2.5955, + "step": 21969 + }, + { + "epoch": 0.94, + "grad_norm": 0.40625, + "learning_rate": 0.00011035855556510632, + "loss": 2.4578, + "step": 21970 + }, + { + "epoch": 0.94, + "grad_norm": 0.431640625, + "learning_rate": 0.00011035188369527662, + "loss": 2.53, + "step": 21971 + }, + { + "epoch": 0.94, + "grad_norm": 0.474609375, + "learning_rate": 0.0001103452117788672, + "loss": 2.7472, + "step": 21972 + }, + { + "epoch": 0.94, + "grad_norm": 0.45703125, + "learning_rate": 0.00011033853981590813, + "loss": 2.5442, + "step": 21973 + }, + { + "epoch": 0.94, + "grad_norm": 0.47265625, + "learning_rate": 0.00011033186780642939, + "loss": 2.4183, + "step": 21974 + }, + { + "epoch": 0.94, + "grad_norm": 0.455078125, + "learning_rate": 0.00011032519575046105, + "loss": 2.5174, + "step": 21975 + }, + { + "epoch": 0.94, + "grad_norm": 0.5, + "learning_rate": 0.00011031852364803308, + "loss": 2.5893, + "step": 21976 + }, + { + "epoch": 0.94, + "grad_norm": 0.47265625, + "learning_rate": 0.00011031185149917554, + "loss": 2.529, + "step": 21977 + }, + { + "epoch": 0.94, + "grad_norm": 0.4453125, + "learning_rate": 0.00011030517930391839, + "loss": 2.4494, + "step": 21978 + }, + { + "epoch": 0.94, + "grad_norm": 0.45703125, + "learning_rate": 0.00011029850706229177, + "loss": 2.628, + "step": 21979 + }, + { + "epoch": 0.94, + "grad_norm": 0.42578125, + "learning_rate": 0.00011029183477432557, + "loss": 2.3699, + "step": 21980 + }, + { + "epoch": 0.94, + "grad_norm": 0.458984375, + "learning_rate": 0.00011028516244004989, + "loss": 2.6351, + "step": 21981 + }, + { + "epoch": 0.94, + "grad_norm": 0.50390625, + "learning_rate": 0.00011027849005949475, + "loss": 2.5535, + "step": 21982 + }, + { + "epoch": 0.94, + "grad_norm": 0.46484375, + "learning_rate": 0.00011027181763269017, + "loss": 2.5908, + "step": 21983 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00011026514515966615, + "loss": 2.4656, + "step": 21984 + }, + { + "epoch": 0.94, + "grad_norm": 0.5546875, + "learning_rate": 0.00011025847264045272, + "loss": 2.6411, + "step": 21985 + }, + { + "epoch": 0.94, + "grad_norm": 0.46875, + "learning_rate": 0.00011025180007507992, + "loss": 2.5959, + "step": 21986 + }, + { + "epoch": 0.94, + "grad_norm": 0.498046875, + "learning_rate": 0.0001102451274635778, + "loss": 2.5887, + "step": 21987 + }, + { + "epoch": 0.94, + "grad_norm": 0.4765625, + "learning_rate": 0.0001102384548059763, + "loss": 2.7159, + "step": 21988 + }, + { + "epoch": 0.94, + "grad_norm": 0.4609375, + "learning_rate": 0.00011023178210230551, + "loss": 2.5605, + "step": 21989 + }, + { + "epoch": 0.94, + "grad_norm": 0.49609375, + "learning_rate": 0.00011022510935259544, + "loss": 2.567, + "step": 21990 + }, + { + "epoch": 0.94, + "grad_norm": 0.474609375, + "learning_rate": 0.00011021843655687613, + "loss": 2.6531, + "step": 21991 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00011021176371517758, + "loss": 2.5312, + "step": 21992 + }, + { + "epoch": 0.94, + "grad_norm": 0.466796875, + "learning_rate": 0.00011020509082752983, + "loss": 2.5491, + "step": 21993 + }, + { + "epoch": 0.94, + "grad_norm": 0.46484375, + "learning_rate": 0.00011019841789396287, + "loss": 2.5697, + "step": 21994 + }, + { + "epoch": 0.94, + "grad_norm": 0.478515625, + "learning_rate": 0.00011019174491450682, + "loss": 2.5902, + "step": 21995 + }, + { + "epoch": 0.94, + "grad_norm": 0.484375, + "learning_rate": 0.00011018507188919158, + "loss": 2.5088, + "step": 21996 + }, + { + "epoch": 0.94, + "grad_norm": 0.427734375, + "learning_rate": 0.00011017839881804726, + "loss": 2.5086, + "step": 21997 + }, + { + "epoch": 0.94, + "grad_norm": 0.46875, + "learning_rate": 0.00011017172570110388, + "loss": 2.6646, + "step": 21998 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00011016505253839144, + "loss": 2.558, + "step": 21999 + }, + { + "epoch": 0.94, + "grad_norm": 0.5859375, + "learning_rate": 0.00011015837932993997, + "loss": 2.6576, + "step": 22000 + }, + { + "epoch": 0.94, + "grad_norm": 0.419921875, + "learning_rate": 0.00011015170607577953, + "loss": 2.3966, + "step": 22001 + }, + { + "epoch": 0.94, + "grad_norm": 0.474609375, + "learning_rate": 0.00011014503277594008, + "loss": 2.6159, + "step": 22002 + }, + { + "epoch": 0.94, + "grad_norm": 0.48828125, + "learning_rate": 0.00011013835943045175, + "loss": 2.5573, + "step": 22003 + }, + { + "epoch": 0.94, + "grad_norm": 0.419921875, + "learning_rate": 0.00011013168603934445, + "loss": 2.5274, + "step": 22004 + }, + { + "epoch": 0.94, + "grad_norm": 0.48046875, + "learning_rate": 0.00011012501260264829, + "loss": 2.605, + "step": 22005 + }, + { + "epoch": 0.94, + "grad_norm": 0.478515625, + "learning_rate": 0.00011011833912039328, + "loss": 2.717, + "step": 22006 + }, + { + "epoch": 0.94, + "grad_norm": 0.494140625, + "learning_rate": 0.00011011166559260943, + "loss": 2.4512, + "step": 22007 + }, + { + "epoch": 0.94, + "grad_norm": 0.4375, + "learning_rate": 0.00011010499201932681, + "loss": 2.5798, + "step": 22008 + }, + { + "epoch": 0.94, + "grad_norm": 0.4609375, + "learning_rate": 0.00011009831840057538, + "loss": 2.599, + "step": 22009 + }, + { + "epoch": 0.94, + "grad_norm": 0.458984375, + "learning_rate": 0.0001100916447363852, + "loss": 2.5062, + "step": 22010 + }, + { + "epoch": 0.94, + "grad_norm": 0.419921875, + "learning_rate": 0.00011008497102678636, + "loss": 2.5559, + "step": 22011 + }, + { + "epoch": 0.94, + "grad_norm": 0.4140625, + "learning_rate": 0.00011007829727180878, + "loss": 2.5248, + "step": 22012 + }, + { + "epoch": 0.94, + "grad_norm": 0.50390625, + "learning_rate": 0.00011007162347148258, + "loss": 2.5898, + "step": 22013 + }, + { + "epoch": 0.94, + "grad_norm": 0.494140625, + "learning_rate": 0.00011006494962583776, + "loss": 2.645, + "step": 22014 + }, + { + "epoch": 0.94, + "grad_norm": 0.4609375, + "learning_rate": 0.00011005827573490432, + "loss": 2.5264, + "step": 22015 + }, + { + "epoch": 0.94, + "grad_norm": 0.38671875, + "learning_rate": 0.00011005160179871235, + "loss": 2.603, + "step": 22016 + }, + { + "epoch": 0.94, + "grad_norm": 0.4609375, + "learning_rate": 0.0001100449278172918, + "loss": 2.4899, + "step": 22017 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00011003825379067277, + "loss": 2.5358, + "step": 22018 + }, + { + "epoch": 0.94, + "grad_norm": 0.419921875, + "learning_rate": 0.00011003157971888527, + "loss": 2.5922, + "step": 22019 + }, + { + "epoch": 0.94, + "grad_norm": 0.46484375, + "learning_rate": 0.00011002490560195931, + "loss": 2.6708, + "step": 22020 + }, + { + "epoch": 0.94, + "grad_norm": 0.412109375, + "learning_rate": 0.00011001823143992495, + "loss": 2.529, + "step": 22021 + }, + { + "epoch": 0.94, + "grad_norm": 0.443359375, + "learning_rate": 0.00011001155723281221, + "loss": 2.6639, + "step": 22022 + }, + { + "epoch": 0.94, + "grad_norm": 0.427734375, + "learning_rate": 0.00011000488298065112, + "loss": 2.4189, + "step": 22023 + }, + { + "epoch": 0.94, + "grad_norm": 0.451171875, + "learning_rate": 0.00010999820868347172, + "loss": 2.5261, + "step": 22024 + }, + { + "epoch": 0.94, + "grad_norm": 0.48828125, + "learning_rate": 0.00010999153434130403, + "loss": 2.5162, + "step": 22025 + }, + { + "epoch": 0.94, + "grad_norm": 0.51171875, + "learning_rate": 0.00010998485995417807, + "loss": 2.6417, + "step": 22026 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00010997818552212393, + "loss": 2.6981, + "step": 22027 + }, + { + "epoch": 0.94, + "grad_norm": 0.484375, + "learning_rate": 0.00010997151104517156, + "loss": 2.5562, + "step": 22028 + }, + { + "epoch": 0.94, + "grad_norm": 0.470703125, + "learning_rate": 0.00010996483652335106, + "loss": 2.6519, + "step": 22029 + }, + { + "epoch": 0.94, + "grad_norm": 0.48828125, + "learning_rate": 0.00010995816195669242, + "loss": 2.5364, + "step": 22030 + }, + { + "epoch": 0.94, + "grad_norm": 0.451171875, + "learning_rate": 0.00010995148734522569, + "loss": 2.7032, + "step": 22031 + }, + { + "epoch": 0.94, + "grad_norm": 0.419921875, + "learning_rate": 0.00010994481268898092, + "loss": 2.4495, + "step": 22032 + }, + { + "epoch": 0.94, + "grad_norm": 0.392578125, + "learning_rate": 0.00010993813798798813, + "loss": 2.6168, + "step": 22033 + }, + { + "epoch": 0.94, + "grad_norm": 0.470703125, + "learning_rate": 0.00010993146324227732, + "loss": 2.6483, + "step": 22034 + }, + { + "epoch": 0.94, + "grad_norm": 0.40625, + "learning_rate": 0.00010992478845187859, + "loss": 2.4916, + "step": 22035 + }, + { + "epoch": 0.94, + "grad_norm": 0.48046875, + "learning_rate": 0.00010991811361682191, + "loss": 2.662, + "step": 22036 + }, + { + "epoch": 0.94, + "grad_norm": 0.46484375, + "learning_rate": 0.00010991143873713736, + "loss": 2.4726, + "step": 22037 + }, + { + "epoch": 0.94, + "grad_norm": 0.421875, + "learning_rate": 0.00010990476381285498, + "loss": 2.4784, + "step": 22038 + }, + { + "epoch": 0.94, + "grad_norm": 0.466796875, + "learning_rate": 0.00010989808884400474, + "loss": 2.5302, + "step": 22039 + }, + { + "epoch": 0.94, + "grad_norm": 0.462890625, + "learning_rate": 0.00010989141383061674, + "loss": 2.5829, + "step": 22040 + }, + { + "epoch": 0.94, + "grad_norm": 0.42578125, + "learning_rate": 0.00010988473877272101, + "loss": 2.517, + "step": 22041 + }, + { + "epoch": 0.94, + "grad_norm": 0.486328125, + "learning_rate": 0.00010987806367034753, + "loss": 2.5603, + "step": 22042 + }, + { + "epoch": 0.94, + "grad_norm": 0.427734375, + "learning_rate": 0.00010987138852352641, + "loss": 2.5441, + "step": 22043 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00010986471333228761, + "loss": 2.579, + "step": 22044 + }, + { + "epoch": 0.94, + "grad_norm": 0.412109375, + "learning_rate": 0.00010985803809666122, + "loss": 2.5973, + "step": 22045 + }, + { + "epoch": 0.94, + "grad_norm": 0.44921875, + "learning_rate": 0.00010985136281667727, + "loss": 2.4494, + "step": 22046 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.00010984468749236578, + "loss": 2.4796, + "step": 22047 + }, + { + "epoch": 0.94, + "grad_norm": 0.48046875, + "learning_rate": 0.00010983801212375679, + "loss": 2.5164, + "step": 22048 + }, + { + "epoch": 0.94, + "grad_norm": 0.470703125, + "learning_rate": 0.00010983133671088037, + "loss": 2.6463, + "step": 22049 + }, + { + "epoch": 0.94, + "grad_norm": 0.396484375, + "learning_rate": 0.00010982466125376648, + "loss": 2.316, + "step": 22050 + }, + { + "epoch": 0.94, + "grad_norm": 0.50390625, + "learning_rate": 0.00010981798575244526, + "loss": 2.5428, + "step": 22051 + }, + { + "epoch": 0.94, + "grad_norm": 0.494140625, + "learning_rate": 0.00010981131020694663, + "loss": 2.3776, + "step": 22052 + }, + { + "epoch": 0.94, + "grad_norm": 0.447265625, + "learning_rate": 0.00010980463461730073, + "loss": 2.5416, + "step": 22053 + }, + { + "epoch": 0.94, + "grad_norm": 0.4765625, + "learning_rate": 0.00010979795898353755, + "loss": 2.6719, + "step": 22054 + }, + { + "epoch": 0.94, + "grad_norm": 0.43359375, + "learning_rate": 0.00010979128330568716, + "loss": 2.5422, + "step": 22055 + }, + { + "epoch": 0.94, + "grad_norm": 0.5, + "learning_rate": 0.0001097846075837795, + "loss": 2.7206, + "step": 22056 + }, + { + "epoch": 0.94, + "grad_norm": 0.5234375, + "learning_rate": 0.00010977793181784477, + "loss": 2.7873, + "step": 22057 + }, + { + "epoch": 0.94, + "grad_norm": 0.48828125, + "learning_rate": 0.00010977125600791287, + "loss": 2.5896, + "step": 22058 + }, + { + "epoch": 0.94, + "grad_norm": 0.458984375, + "learning_rate": 0.00010976458015401392, + "loss": 2.4807, + "step": 22059 + }, + { + "epoch": 0.94, + "grad_norm": 0.470703125, + "learning_rate": 0.00010975790425617788, + "loss": 2.5648, + "step": 22060 + }, + { + "epoch": 0.94, + "grad_norm": 0.423828125, + "learning_rate": 0.00010975122831443488, + "loss": 2.5354, + "step": 22061 + }, + { + "epoch": 0.94, + "grad_norm": 0.44921875, + "learning_rate": 0.00010974455232881493, + "loss": 2.5251, + "step": 22062 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.000109737876299348, + "loss": 2.6467, + "step": 22063 + }, + { + "epoch": 0.94, + "grad_norm": 0.494140625, + "learning_rate": 0.00010973120022606423, + "loss": 2.632, + "step": 22064 + }, + { + "epoch": 0.94, + "grad_norm": 0.423828125, + "learning_rate": 0.00010972452410899359, + "loss": 2.5386, + "step": 22065 + }, + { + "epoch": 0.94, + "grad_norm": 0.416015625, + "learning_rate": 0.00010971784794816616, + "loss": 2.4498, + "step": 22066 + }, + { + "epoch": 0.94, + "grad_norm": 0.46875, + "learning_rate": 0.00010971117174361193, + "loss": 2.6374, + "step": 22067 + }, + { + "epoch": 0.94, + "grad_norm": 0.486328125, + "learning_rate": 0.00010970449549536104, + "loss": 2.4835, + "step": 22068 + }, + { + "epoch": 0.94, + "grad_norm": 0.443359375, + "learning_rate": 0.00010969781920344339, + "loss": 2.6525, + "step": 22069 + }, + { + "epoch": 0.94, + "grad_norm": 0.40234375, + "learning_rate": 0.00010969114286788917, + "loss": 2.5747, + "step": 22070 + }, + { + "epoch": 0.94, + "grad_norm": 0.5, + "learning_rate": 0.0001096844664887283, + "loss": 2.5382, + "step": 22071 + }, + { + "epoch": 0.94, + "grad_norm": 0.443359375, + "learning_rate": 0.0001096777900659909, + "loss": 2.5521, + "step": 22072 + }, + { + "epoch": 0.94, + "grad_norm": 0.416015625, + "learning_rate": 0.00010967111359970697, + "loss": 2.5572, + "step": 22073 + }, + { + "epoch": 0.94, + "grad_norm": 0.419921875, + "learning_rate": 0.00010966443708990657, + "loss": 2.5349, + "step": 22074 + }, + { + "epoch": 0.94, + "grad_norm": 0.44921875, + "learning_rate": 0.00010965776053661971, + "loss": 2.5309, + "step": 22075 + }, + { + "epoch": 0.94, + "grad_norm": 0.4453125, + "learning_rate": 0.00010965108393987649, + "loss": 2.5505, + "step": 22076 + }, + { + "epoch": 0.94, + "grad_norm": 0.435546875, + "learning_rate": 0.00010964440729970688, + "loss": 2.5845, + "step": 22077 + }, + { + "epoch": 0.94, + "grad_norm": 0.474609375, + "learning_rate": 0.00010963773061614103, + "loss": 2.5254, + "step": 22078 + }, + { + "epoch": 0.94, + "grad_norm": 0.45703125, + "learning_rate": 0.00010963105388920884, + "loss": 2.5085, + "step": 22079 + }, + { + "epoch": 0.94, + "grad_norm": 0.46875, + "learning_rate": 0.00010962437711894047, + "loss": 2.581, + "step": 22080 + }, + { + "epoch": 0.94, + "grad_norm": 0.5546875, + "learning_rate": 0.00010961770030536591, + "loss": 2.7433, + "step": 22081 + }, + { + "epoch": 0.94, + "grad_norm": 0.412109375, + "learning_rate": 0.00010961102344851522, + "loss": 2.5903, + "step": 22082 + }, + { + "epoch": 0.94, + "grad_norm": 0.423828125, + "learning_rate": 0.0001096043465484184, + "loss": 2.5458, + "step": 22083 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.00010959766960510558, + "loss": 2.487, + "step": 22084 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00010959099261860672, + "loss": 2.4814, + "step": 22085 + }, + { + "epoch": 0.94, + "grad_norm": 0.458984375, + "learning_rate": 0.00010958431558895193, + "loss": 2.4988, + "step": 22086 + }, + { + "epoch": 0.94, + "grad_norm": 0.50390625, + "learning_rate": 0.00010957763851617118, + "loss": 2.5509, + "step": 22087 + }, + { + "epoch": 0.94, + "grad_norm": 0.455078125, + "learning_rate": 0.00010957096140029459, + "loss": 2.4406, + "step": 22088 + }, + { + "epoch": 0.94, + "grad_norm": 0.51171875, + "learning_rate": 0.00010956428424135217, + "loss": 2.4752, + "step": 22089 + }, + { + "epoch": 0.94, + "grad_norm": 0.5, + "learning_rate": 0.00010955760703937396, + "loss": 2.4917, + "step": 22090 + }, + { + "epoch": 0.94, + "grad_norm": 0.45703125, + "learning_rate": 0.00010955092979438999, + "loss": 2.4455, + "step": 22091 + }, + { + "epoch": 0.94, + "grad_norm": 0.47265625, + "learning_rate": 0.00010954425250643036, + "loss": 2.5289, + "step": 22092 + }, + { + "epoch": 0.94, + "grad_norm": 0.416015625, + "learning_rate": 0.00010953757517552505, + "loss": 2.5178, + "step": 22093 + }, + { + "epoch": 0.94, + "grad_norm": 0.423828125, + "learning_rate": 0.00010953089780170419, + "loss": 2.546, + "step": 22094 + }, + { + "epoch": 0.94, + "grad_norm": 0.4609375, + "learning_rate": 0.00010952422038499772, + "loss": 2.5909, + "step": 22095 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.00010951754292543577, + "loss": 2.4891, + "step": 22096 + }, + { + "epoch": 0.94, + "grad_norm": 0.52734375, + "learning_rate": 0.00010951086542304834, + "loss": 2.5187, + "step": 22097 + }, + { + "epoch": 0.94, + "grad_norm": 0.41796875, + "learning_rate": 0.00010950418787786549, + "loss": 2.5982, + "step": 22098 + }, + { + "epoch": 0.94, + "grad_norm": 0.515625, + "learning_rate": 0.00010949751028991725, + "loss": 2.6484, + "step": 22099 + }, + { + "epoch": 0.94, + "grad_norm": 0.478515625, + "learning_rate": 0.00010949083265923374, + "loss": 2.5078, + "step": 22100 + }, + { + "epoch": 0.94, + "grad_norm": 0.45703125, + "learning_rate": 0.0001094841549858449, + "loss": 2.5698, + "step": 22101 + }, + { + "epoch": 0.94, + "grad_norm": 0.486328125, + "learning_rate": 0.00010947747726978086, + "loss": 2.6779, + "step": 22102 + }, + { + "epoch": 0.94, + "grad_norm": 0.380859375, + "learning_rate": 0.00010947079951107163, + "loss": 2.5588, + "step": 22103 + }, + { + "epoch": 0.94, + "grad_norm": 0.51171875, + "learning_rate": 0.00010946412170974724, + "loss": 2.6072, + "step": 22104 + }, + { + "epoch": 0.94, + "grad_norm": 0.423828125, + "learning_rate": 0.00010945744386583778, + "loss": 2.5781, + "step": 22105 + }, + { + "epoch": 0.94, + "grad_norm": 0.48046875, + "learning_rate": 0.00010945076597937328, + "loss": 2.56, + "step": 22106 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00010944408805038375, + "loss": 2.4254, + "step": 22107 + }, + { + "epoch": 0.94, + "grad_norm": 0.412109375, + "learning_rate": 0.00010943741007889933, + "loss": 2.5715, + "step": 22108 + }, + { + "epoch": 0.94, + "grad_norm": 0.4921875, + "learning_rate": 0.00010943073206494997, + "loss": 2.6287, + "step": 22109 + }, + { + "epoch": 0.94, + "grad_norm": 0.4296875, + "learning_rate": 0.00010942405400856581, + "loss": 2.4125, + "step": 22110 + }, + { + "epoch": 0.94, + "grad_norm": 0.458984375, + "learning_rate": 0.00010941737590977683, + "loss": 2.5309, + "step": 22111 + }, + { + "epoch": 0.94, + "grad_norm": 0.54296875, + "learning_rate": 0.00010941069776861308, + "loss": 2.5899, + "step": 22112 + }, + { + "epoch": 0.94, + "grad_norm": 0.4375, + "learning_rate": 0.00010940401958510467, + "loss": 2.587, + "step": 22113 + }, + { + "epoch": 0.94, + "grad_norm": 0.47265625, + "learning_rate": 0.00010939734135928156, + "loss": 2.6016, + "step": 22114 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.00010939066309117387, + "loss": 2.5654, + "step": 22115 + }, + { + "epoch": 0.94, + "grad_norm": 0.46484375, + "learning_rate": 0.00010938398478081165, + "loss": 2.5833, + "step": 22116 + }, + { + "epoch": 0.94, + "grad_norm": 0.447265625, + "learning_rate": 0.00010937730642822488, + "loss": 2.4877, + "step": 22117 + }, + { + "epoch": 0.94, + "grad_norm": 0.427734375, + "learning_rate": 0.0001093706280334437, + "loss": 2.5776, + "step": 22118 + }, + { + "epoch": 0.94, + "grad_norm": 0.4921875, + "learning_rate": 0.00010936394959649809, + "loss": 2.6151, + "step": 22119 + }, + { + "epoch": 0.94, + "grad_norm": 0.4375, + "learning_rate": 0.00010935727111741816, + "loss": 2.5252, + "step": 22120 + }, + { + "epoch": 0.94, + "grad_norm": 0.490234375, + "learning_rate": 0.0001093505925962339, + "loss": 2.5744, + "step": 22121 + }, + { + "epoch": 0.94, + "grad_norm": 0.421875, + "learning_rate": 0.00010934391403297542, + "loss": 2.4596, + "step": 22122 + }, + { + "epoch": 0.94, + "grad_norm": 0.48828125, + "learning_rate": 0.00010933723542767272, + "loss": 2.5351, + "step": 22123 + }, + { + "epoch": 0.94, + "grad_norm": 0.421875, + "learning_rate": 0.00010933055678035588, + "loss": 2.4704, + "step": 22124 + }, + { + "epoch": 0.94, + "grad_norm": 0.451171875, + "learning_rate": 0.00010932387809105493, + "loss": 2.5037, + "step": 22125 + }, + { + "epoch": 0.94, + "grad_norm": 0.4765625, + "learning_rate": 0.00010931719935979996, + "loss": 2.5155, + "step": 22126 + }, + { + "epoch": 0.94, + "grad_norm": 0.447265625, + "learning_rate": 0.00010931052058662097, + "loss": 2.5609, + "step": 22127 + }, + { + "epoch": 0.94, + "grad_norm": 0.4375, + "learning_rate": 0.00010930384177154806, + "loss": 2.5286, + "step": 22128 + }, + { + "epoch": 0.94, + "grad_norm": 0.423828125, + "learning_rate": 0.00010929716291461128, + "loss": 2.6671, + "step": 22129 + }, + { + "epoch": 0.94, + "grad_norm": 0.439453125, + "learning_rate": 0.00010929048401584064, + "loss": 2.5372, + "step": 22130 + }, + { + "epoch": 0.94, + "grad_norm": 0.4765625, + "learning_rate": 0.00010928380507526618, + "loss": 2.5859, + "step": 22131 + }, + { + "epoch": 0.94, + "grad_norm": 0.482421875, + "learning_rate": 0.00010927712609291807, + "loss": 2.6177, + "step": 22132 + }, + { + "epoch": 0.94, + "grad_norm": 0.455078125, + "learning_rate": 0.00010927044706882623, + "loss": 2.5841, + "step": 22133 + }, + { + "epoch": 0.94, + "grad_norm": 0.48046875, + "learning_rate": 0.00010926376800302079, + "loss": 2.552, + "step": 22134 + }, + { + "epoch": 0.94, + "grad_norm": 0.466796875, + "learning_rate": 0.00010925708889553177, + "loss": 2.5688, + "step": 22135 + }, + { + "epoch": 0.94, + "grad_norm": 0.466796875, + "learning_rate": 0.00010925040974638923, + "loss": 2.4382, + "step": 22136 + }, + { + "epoch": 0.94, + "grad_norm": 0.4140625, + "learning_rate": 0.00010924373055562322, + "loss": 2.5948, + "step": 22137 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.0001092370513232638, + "loss": 2.5622, + "step": 22138 + }, + { + "epoch": 0.94, + "grad_norm": 0.458984375, + "learning_rate": 0.00010923037204934104, + "loss": 2.4253, + "step": 22139 + }, + { + "epoch": 0.94, + "grad_norm": 0.439453125, + "learning_rate": 0.00010922369273388499, + "loss": 2.4483, + "step": 22140 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.00010921701337692565, + "loss": 2.5251, + "step": 22141 + }, + { + "epoch": 0.94, + "grad_norm": 0.41796875, + "learning_rate": 0.00010921033397849316, + "loss": 2.5654, + "step": 22142 + }, + { + "epoch": 0.94, + "grad_norm": 0.470703125, + "learning_rate": 0.00010920365453861751, + "loss": 2.65, + "step": 22143 + }, + { + "epoch": 0.94, + "grad_norm": 0.462890625, + "learning_rate": 0.0001091969750573288, + "loss": 2.5378, + "step": 22144 + }, + { + "epoch": 0.94, + "grad_norm": 0.427734375, + "learning_rate": 0.00010919029553465705, + "loss": 2.6264, + "step": 22145 + }, + { + "epoch": 0.94, + "grad_norm": 0.4453125, + "learning_rate": 0.00010918361597063232, + "loss": 2.5454, + "step": 22146 + }, + { + "epoch": 0.94, + "grad_norm": 0.431640625, + "learning_rate": 0.00010917693636528468, + "loss": 2.5727, + "step": 22147 + }, + { + "epoch": 0.94, + "grad_norm": 0.5234375, + "learning_rate": 0.0001091702567186442, + "loss": 2.5973, + "step": 22148 + }, + { + "epoch": 0.94, + "grad_norm": 0.470703125, + "learning_rate": 0.00010916357703074087, + "loss": 2.5749, + "step": 22149 + }, + { + "epoch": 0.94, + "grad_norm": 0.41796875, + "learning_rate": 0.00010915689730160483, + "loss": 2.4708, + "step": 22150 + }, + { + "epoch": 0.94, + "grad_norm": 0.4296875, + "learning_rate": 0.00010915021753126608, + "loss": 2.5202, + "step": 22151 + }, + { + "epoch": 0.94, + "grad_norm": 0.4453125, + "learning_rate": 0.00010914353771975472, + "loss": 2.4386, + "step": 22152 + }, + { + "epoch": 0.94, + "grad_norm": 0.478515625, + "learning_rate": 0.00010913685786710077, + "loss": 2.5115, + "step": 22153 + }, + { + "epoch": 0.94, + "grad_norm": 0.4296875, + "learning_rate": 0.00010913017797333429, + "loss": 2.4876, + "step": 22154 + }, + { + "epoch": 0.94, + "grad_norm": 0.41796875, + "learning_rate": 0.00010912349803848534, + "loss": 2.521, + "step": 22155 + }, + { + "epoch": 0.94, + "grad_norm": 0.419921875, + "learning_rate": 0.00010911681806258403, + "loss": 2.5736, + "step": 22156 + }, + { + "epoch": 0.94, + "grad_norm": 0.44140625, + "learning_rate": 0.0001091101380456603, + "loss": 2.5039, + "step": 22157 + }, + { + "epoch": 0.94, + "grad_norm": 0.423828125, + "learning_rate": 0.00010910345798774433, + "loss": 2.4708, + "step": 22158 + }, + { + "epoch": 0.94, + "grad_norm": 0.4609375, + "learning_rate": 0.0001090967778888661, + "loss": 2.6614, + "step": 22159 + }, + { + "epoch": 0.94, + "grad_norm": 0.48046875, + "learning_rate": 0.0001090900977490557, + "loss": 2.4322, + "step": 22160 + }, + { + "epoch": 0.94, + "grad_norm": 0.451171875, + "learning_rate": 0.0001090834175683432, + "loss": 2.6038, + "step": 22161 + }, + { + "epoch": 0.94, + "grad_norm": 0.53515625, + "learning_rate": 0.00010907673734675863, + "loss": 2.6069, + "step": 22162 + }, + { + "epoch": 0.94, + "grad_norm": 0.462890625, + "learning_rate": 0.00010907005708433202, + "loss": 2.5275, + "step": 22163 + }, + { + "epoch": 0.94, + "grad_norm": 0.447265625, + "learning_rate": 0.00010906337678109353, + "loss": 2.5269, + "step": 22164 + }, + { + "epoch": 0.94, + "grad_norm": 0.41015625, + "learning_rate": 0.00010905669643707313, + "loss": 2.5324, + "step": 22165 + }, + { + "epoch": 0.94, + "grad_norm": 0.498046875, + "learning_rate": 0.0001090500160523009, + "loss": 2.5423, + "step": 22166 + }, + { + "epoch": 0.94, + "grad_norm": 0.478515625, + "learning_rate": 0.00010904333562680691, + "loss": 2.5458, + "step": 22167 + }, + { + "epoch": 0.94, + "grad_norm": 0.404296875, + "learning_rate": 0.00010903665516062122, + "loss": 2.4563, + "step": 22168 + }, + { + "epoch": 0.94, + "grad_norm": 0.431640625, + "learning_rate": 0.00010902997465377387, + "loss": 2.724, + "step": 22169 + }, + { + "epoch": 0.94, + "grad_norm": 0.4609375, + "learning_rate": 0.00010902329410629494, + "loss": 2.4871, + "step": 22170 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.00010901661351821448, + "loss": 2.4705, + "step": 22171 + }, + { + "epoch": 0.94, + "grad_norm": 0.6015625, + "learning_rate": 0.00010900993288956258, + "loss": 2.7305, + "step": 22172 + }, + { + "epoch": 0.94, + "grad_norm": 0.423828125, + "learning_rate": 0.00010900325222036925, + "loss": 2.5675, + "step": 22173 + }, + { + "epoch": 0.94, + "grad_norm": 0.51171875, + "learning_rate": 0.00010899657151066458, + "loss": 2.6343, + "step": 22174 + }, + { + "epoch": 0.94, + "grad_norm": 0.453125, + "learning_rate": 0.00010898989076047863, + "loss": 2.5208, + "step": 22175 + }, + { + "epoch": 0.94, + "grad_norm": 0.447265625, + "learning_rate": 0.00010898320996984146, + "loss": 2.5789, + "step": 22176 + }, + { + "epoch": 0.95, + "grad_norm": 0.4921875, + "learning_rate": 0.00010897652913878313, + "loss": 2.5943, + "step": 22177 + }, + { + "epoch": 0.95, + "grad_norm": 0.416015625, + "learning_rate": 0.0001089698482673337, + "loss": 2.4151, + "step": 22178 + }, + { + "epoch": 0.95, + "grad_norm": 0.380859375, + "learning_rate": 0.00010896316735552319, + "loss": 2.5337, + "step": 22179 + }, + { + "epoch": 0.95, + "grad_norm": 0.392578125, + "learning_rate": 0.00010895648640338177, + "loss": 2.4045, + "step": 22180 + }, + { + "epoch": 0.95, + "grad_norm": 0.5234375, + "learning_rate": 0.00010894980541093937, + "loss": 2.6551, + "step": 22181 + }, + { + "epoch": 0.95, + "grad_norm": 0.54296875, + "learning_rate": 0.00010894312437822616, + "loss": 2.5713, + "step": 22182 + }, + { + "epoch": 0.95, + "grad_norm": 0.49609375, + "learning_rate": 0.00010893644330527217, + "loss": 2.6111, + "step": 22183 + }, + { + "epoch": 0.95, + "grad_norm": 0.474609375, + "learning_rate": 0.0001089297621921074, + "loss": 2.5277, + "step": 22184 + }, + { + "epoch": 0.95, + "grad_norm": 0.498046875, + "learning_rate": 0.000108923081038762, + "loss": 2.6578, + "step": 22185 + }, + { + "epoch": 0.95, + "grad_norm": 0.48046875, + "learning_rate": 0.00010891639984526599, + "loss": 2.5428, + "step": 22186 + }, + { + "epoch": 0.95, + "grad_norm": 0.470703125, + "learning_rate": 0.00010890971861164944, + "loss": 2.6633, + "step": 22187 + }, + { + "epoch": 0.95, + "grad_norm": 0.45703125, + "learning_rate": 0.0001089030373379424, + "loss": 2.4738, + "step": 22188 + }, + { + "epoch": 0.95, + "grad_norm": 0.42578125, + "learning_rate": 0.00010889635602417497, + "loss": 2.5135, + "step": 22189 + }, + { + "epoch": 0.95, + "grad_norm": 0.46875, + "learning_rate": 0.00010888967467037714, + "loss": 2.498, + "step": 22190 + }, + { + "epoch": 0.95, + "grad_norm": 0.4765625, + "learning_rate": 0.00010888299327657909, + "loss": 2.5585, + "step": 22191 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010887631184281075, + "loss": 2.6457, + "step": 22192 + }, + { + "epoch": 0.95, + "grad_norm": 0.494140625, + "learning_rate": 0.0001088696303691023, + "loss": 2.5751, + "step": 22193 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010886294885548374, + "loss": 2.6838, + "step": 22194 + }, + { + "epoch": 0.95, + "grad_norm": 0.42578125, + "learning_rate": 0.00010885626730198516, + "loss": 2.5767, + "step": 22195 + }, + { + "epoch": 0.95, + "grad_norm": 0.427734375, + "learning_rate": 0.0001088495857086366, + "loss": 2.4808, + "step": 22196 + }, + { + "epoch": 0.95, + "grad_norm": 0.484375, + "learning_rate": 0.00010884290407546813, + "loss": 2.6201, + "step": 22197 + }, + { + "epoch": 0.95, + "grad_norm": 0.50390625, + "learning_rate": 0.00010883622240250982, + "loss": 2.5711, + "step": 22198 + }, + { + "epoch": 0.95, + "grad_norm": 0.43359375, + "learning_rate": 0.00010882954068979177, + "loss": 2.5875, + "step": 22199 + }, + { + "epoch": 0.95, + "grad_norm": 0.439453125, + "learning_rate": 0.00010882285893734399, + "loss": 2.4671, + "step": 22200 + }, + { + "epoch": 0.95, + "grad_norm": 0.41796875, + "learning_rate": 0.00010881617714519657, + "loss": 2.5156, + "step": 22201 + }, + { + "epoch": 0.95, + "grad_norm": 0.4453125, + "learning_rate": 0.00010880949531337958, + "loss": 2.6322, + "step": 22202 + }, + { + "epoch": 0.95, + "grad_norm": 0.484375, + "learning_rate": 0.00010880281344192309, + "loss": 2.6421, + "step": 22203 + }, + { + "epoch": 0.95, + "grad_norm": 0.4453125, + "learning_rate": 0.00010879613153085715, + "loss": 2.6379, + "step": 22204 + }, + { + "epoch": 0.95, + "grad_norm": 0.48828125, + "learning_rate": 0.00010878944958021182, + "loss": 2.5506, + "step": 22205 + }, + { + "epoch": 0.95, + "grad_norm": 0.39453125, + "learning_rate": 0.00010878276759001717, + "loss": 2.4294, + "step": 22206 + }, + { + "epoch": 0.95, + "grad_norm": 0.451171875, + "learning_rate": 0.00010877608556030333, + "loss": 2.5025, + "step": 22207 + }, + { + "epoch": 0.95, + "grad_norm": 0.455078125, + "learning_rate": 0.00010876940349110025, + "loss": 2.5377, + "step": 22208 + }, + { + "epoch": 0.95, + "grad_norm": 0.470703125, + "learning_rate": 0.00010876272138243808, + "loss": 2.5601, + "step": 22209 + }, + { + "epoch": 0.95, + "grad_norm": 0.412109375, + "learning_rate": 0.00010875603923434688, + "loss": 2.4074, + "step": 22210 + }, + { + "epoch": 0.95, + "grad_norm": 0.490234375, + "learning_rate": 0.00010874935704685669, + "loss": 2.5635, + "step": 22211 + }, + { + "epoch": 0.95, + "grad_norm": 0.4375, + "learning_rate": 0.00010874267481999761, + "loss": 2.5739, + "step": 22212 + }, + { + "epoch": 0.95, + "grad_norm": 0.54296875, + "learning_rate": 0.00010873599255379967, + "loss": 2.6468, + "step": 22213 + }, + { + "epoch": 0.95, + "grad_norm": 0.4375, + "learning_rate": 0.00010872931024829294, + "loss": 2.4757, + "step": 22214 + }, + { + "epoch": 0.95, + "grad_norm": 0.40625, + "learning_rate": 0.00010872262790350756, + "loss": 2.4364, + "step": 22215 + }, + { + "epoch": 0.95, + "grad_norm": 0.48046875, + "learning_rate": 0.00010871594551947349, + "loss": 2.5442, + "step": 22216 + }, + { + "epoch": 0.95, + "grad_norm": 0.419921875, + "learning_rate": 0.00010870926309622085, + "loss": 2.6312, + "step": 22217 + }, + { + "epoch": 0.95, + "grad_norm": 0.43359375, + "learning_rate": 0.00010870258063377973, + "loss": 2.6135, + "step": 22218 + }, + { + "epoch": 0.95, + "grad_norm": 0.421875, + "learning_rate": 0.00010869589813218017, + "loss": 2.5231, + "step": 22219 + }, + { + "epoch": 0.95, + "grad_norm": 0.490234375, + "learning_rate": 0.00010868921559145226, + "loss": 2.7412, + "step": 22220 + }, + { + "epoch": 0.95, + "grad_norm": 0.46875, + "learning_rate": 0.00010868253301162605, + "loss": 2.6698, + "step": 22221 + }, + { + "epoch": 0.95, + "grad_norm": 0.431640625, + "learning_rate": 0.00010867585039273159, + "loss": 2.574, + "step": 22222 + }, + { + "epoch": 0.95, + "grad_norm": 0.5078125, + "learning_rate": 0.00010866916773479903, + "loss": 2.6366, + "step": 22223 + }, + { + "epoch": 0.95, + "grad_norm": 0.435546875, + "learning_rate": 0.00010866248503785832, + "loss": 2.5234, + "step": 22224 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010865580230193962, + "loss": 2.5718, + "step": 22225 + }, + { + "epoch": 0.95, + "grad_norm": 0.404296875, + "learning_rate": 0.00010864911952707297, + "loss": 2.579, + "step": 22226 + }, + { + "epoch": 0.95, + "grad_norm": 0.5546875, + "learning_rate": 0.00010864243671328844, + "loss": 2.7238, + "step": 22227 + }, + { + "epoch": 0.95, + "grad_norm": 0.443359375, + "learning_rate": 0.00010863575386061614, + "loss": 2.6359, + "step": 22228 + }, + { + "epoch": 0.95, + "grad_norm": 0.4296875, + "learning_rate": 0.00010862907096908606, + "loss": 2.5679, + "step": 22229 + }, + { + "epoch": 0.95, + "grad_norm": 0.40234375, + "learning_rate": 0.00010862238803872831, + "loss": 2.5163, + "step": 22230 + }, + { + "epoch": 0.95, + "grad_norm": 0.45703125, + "learning_rate": 0.00010861570506957299, + "loss": 2.46, + "step": 22231 + }, + { + "epoch": 0.95, + "grad_norm": 0.466796875, + "learning_rate": 0.00010860902206165013, + "loss": 2.5597, + "step": 22232 + }, + { + "epoch": 0.95, + "grad_norm": 0.396484375, + "learning_rate": 0.00010860233901498982, + "loss": 2.536, + "step": 22233 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010859565592962213, + "loss": 2.6066, + "step": 22234 + }, + { + "epoch": 0.95, + "grad_norm": 0.40625, + "learning_rate": 0.00010858897280557714, + "loss": 2.6152, + "step": 22235 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010858228964288491, + "loss": 2.5116, + "step": 22236 + }, + { + "epoch": 0.95, + "grad_norm": 0.498046875, + "learning_rate": 0.00010857560644157552, + "loss": 2.584, + "step": 22237 + }, + { + "epoch": 0.95, + "grad_norm": 0.4296875, + "learning_rate": 0.000108568923201679, + "loss": 2.4363, + "step": 22238 + }, + { + "epoch": 0.95, + "grad_norm": 0.40234375, + "learning_rate": 0.0001085622399232255, + "loss": 2.5219, + "step": 22239 + }, + { + "epoch": 0.95, + "grad_norm": 0.44921875, + "learning_rate": 0.00010855555660624502, + "loss": 2.5416, + "step": 22240 + }, + { + "epoch": 0.95, + "grad_norm": 0.474609375, + "learning_rate": 0.00010854887325076765, + "loss": 2.5985, + "step": 22241 + }, + { + "epoch": 0.95, + "grad_norm": 0.458984375, + "learning_rate": 0.00010854218985682352, + "loss": 2.5186, + "step": 22242 + }, + { + "epoch": 0.95, + "grad_norm": 0.392578125, + "learning_rate": 0.00010853550642444264, + "loss": 2.6065, + "step": 22243 + }, + { + "epoch": 0.95, + "grad_norm": 0.54296875, + "learning_rate": 0.00010852882295365508, + "loss": 2.55, + "step": 22244 + }, + { + "epoch": 0.95, + "grad_norm": 0.47265625, + "learning_rate": 0.00010852213944449095, + "loss": 2.5154, + "step": 22245 + }, + { + "epoch": 0.95, + "grad_norm": 0.43359375, + "learning_rate": 0.00010851545589698028, + "loss": 2.6243, + "step": 22246 + }, + { + "epoch": 0.95, + "grad_norm": 0.52734375, + "learning_rate": 0.00010850877231115322, + "loss": 2.6935, + "step": 22247 + }, + { + "epoch": 0.95, + "grad_norm": 0.474609375, + "learning_rate": 0.00010850208868703975, + "loss": 2.5531, + "step": 22248 + }, + { + "epoch": 0.95, + "grad_norm": 0.458984375, + "learning_rate": 0.00010849540502467002, + "loss": 2.5726, + "step": 22249 + }, + { + "epoch": 0.95, + "grad_norm": 0.486328125, + "learning_rate": 0.00010848872132407406, + "loss": 2.5256, + "step": 22250 + }, + { + "epoch": 0.95, + "grad_norm": 0.46875, + "learning_rate": 0.00010848203758528194, + "loss": 2.4823, + "step": 22251 + }, + { + "epoch": 0.95, + "grad_norm": 0.431640625, + "learning_rate": 0.00010847535380832379, + "loss": 2.5631, + "step": 22252 + }, + { + "epoch": 0.95, + "grad_norm": 0.404296875, + "learning_rate": 0.00010846866999322962, + "loss": 2.4288, + "step": 22253 + }, + { + "epoch": 0.95, + "grad_norm": 0.41015625, + "learning_rate": 0.00010846198614002952, + "loss": 2.6247, + "step": 22254 + }, + { + "epoch": 0.95, + "grad_norm": 0.408203125, + "learning_rate": 0.0001084553022487536, + "loss": 2.4572, + "step": 22255 + }, + { + "epoch": 0.95, + "grad_norm": 0.4921875, + "learning_rate": 0.00010844861831943187, + "loss": 2.6045, + "step": 22256 + }, + { + "epoch": 0.95, + "grad_norm": 0.462890625, + "learning_rate": 0.00010844193435209446, + "loss": 2.641, + "step": 22257 + }, + { + "epoch": 0.95, + "grad_norm": 0.46875, + "learning_rate": 0.00010843525034677147, + "loss": 2.6182, + "step": 22258 + }, + { + "epoch": 0.95, + "grad_norm": 0.427734375, + "learning_rate": 0.0001084285663034929, + "loss": 2.4816, + "step": 22259 + }, + { + "epoch": 0.95, + "grad_norm": 0.47265625, + "learning_rate": 0.00010842188222228889, + "loss": 2.5132, + "step": 22260 + }, + { + "epoch": 0.95, + "grad_norm": 0.4375, + "learning_rate": 0.00010841519810318945, + "loss": 2.4364, + "step": 22261 + }, + { + "epoch": 0.95, + "grad_norm": 0.435546875, + "learning_rate": 0.00010840851394622469, + "loss": 2.5061, + "step": 22262 + }, + { + "epoch": 0.95, + "grad_norm": 0.48046875, + "learning_rate": 0.00010840182975142475, + "loss": 2.5588, + "step": 22263 + }, + { + "epoch": 0.95, + "grad_norm": 0.482421875, + "learning_rate": 0.00010839514551881958, + "loss": 2.5304, + "step": 22264 + }, + { + "epoch": 0.95, + "grad_norm": 0.458984375, + "learning_rate": 0.00010838846124843935, + "loss": 2.5169, + "step": 22265 + }, + { + "epoch": 0.95, + "grad_norm": 0.466796875, + "learning_rate": 0.00010838177694031413, + "loss": 2.5375, + "step": 22266 + }, + { + "epoch": 0.95, + "grad_norm": 0.453125, + "learning_rate": 0.00010837509259447396, + "loss": 2.6815, + "step": 22267 + }, + { + "epoch": 0.95, + "grad_norm": 0.50390625, + "learning_rate": 0.00010836840821094893, + "loss": 2.5129, + "step": 22268 + }, + { + "epoch": 0.95, + "grad_norm": 0.451171875, + "learning_rate": 0.00010836172378976911, + "loss": 2.5428, + "step": 22269 + }, + { + "epoch": 0.95, + "grad_norm": 0.41015625, + "learning_rate": 0.00010835503933096461, + "loss": 2.4173, + "step": 22270 + }, + { + "epoch": 0.95, + "grad_norm": 0.431640625, + "learning_rate": 0.00010834835483456548, + "loss": 2.3526, + "step": 22271 + }, + { + "epoch": 0.95, + "grad_norm": 0.46875, + "learning_rate": 0.00010834167030060178, + "loss": 2.5432, + "step": 22272 + }, + { + "epoch": 0.95, + "grad_norm": 0.44140625, + "learning_rate": 0.00010833498572910366, + "loss": 2.4632, + "step": 22273 + }, + { + "epoch": 0.95, + "grad_norm": 0.44140625, + "learning_rate": 0.00010832830112010113, + "loss": 2.3096, + "step": 22274 + }, + { + "epoch": 0.95, + "grad_norm": 0.392578125, + "learning_rate": 0.00010832161647362426, + "loss": 2.374, + "step": 22275 + }, + { + "epoch": 0.95, + "grad_norm": 0.478515625, + "learning_rate": 0.0001083149317897032, + "loss": 2.6096, + "step": 22276 + }, + { + "epoch": 0.95, + "grad_norm": 0.423828125, + "learning_rate": 0.00010830824706836797, + "loss": 2.6138, + "step": 22277 + }, + { + "epoch": 0.95, + "grad_norm": 0.427734375, + "learning_rate": 0.00010830156230964865, + "loss": 2.4677, + "step": 22278 + }, + { + "epoch": 0.95, + "grad_norm": 0.4453125, + "learning_rate": 0.00010829487751357537, + "loss": 2.5326, + "step": 22279 + }, + { + "epoch": 0.95, + "grad_norm": 0.46484375, + "learning_rate": 0.00010828819268017812, + "loss": 2.6964, + "step": 22280 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010828150780948708, + "loss": 2.5073, + "step": 22281 + }, + { + "epoch": 0.95, + "grad_norm": 0.455078125, + "learning_rate": 0.00010827482290153226, + "loss": 2.4761, + "step": 22282 + }, + { + "epoch": 0.95, + "grad_norm": 0.421875, + "learning_rate": 0.00010826813795634376, + "loss": 2.5173, + "step": 22283 + }, + { + "epoch": 0.95, + "grad_norm": 0.494140625, + "learning_rate": 0.00010826145297395165, + "loss": 2.5406, + "step": 22284 + }, + { + "epoch": 0.95, + "grad_norm": 0.49609375, + "learning_rate": 0.00010825476795438605, + "loss": 2.5922, + "step": 22285 + }, + { + "epoch": 0.95, + "grad_norm": 0.5234375, + "learning_rate": 0.00010824808289767696, + "loss": 2.6295, + "step": 22286 + }, + { + "epoch": 0.95, + "grad_norm": 0.462890625, + "learning_rate": 0.00010824139780385458, + "loss": 2.5557, + "step": 22287 + }, + { + "epoch": 0.95, + "grad_norm": 0.462890625, + "learning_rate": 0.00010823471267294887, + "loss": 2.6176, + "step": 22288 + }, + { + "epoch": 0.95, + "grad_norm": 0.5078125, + "learning_rate": 0.00010822802750498999, + "loss": 2.5769, + "step": 22289 + }, + { + "epoch": 0.95, + "grad_norm": 0.408203125, + "learning_rate": 0.00010822134230000798, + "loss": 2.4254, + "step": 22290 + }, + { + "epoch": 0.95, + "grad_norm": 0.515625, + "learning_rate": 0.00010821465705803293, + "loss": 2.502, + "step": 22291 + }, + { + "epoch": 0.95, + "grad_norm": 0.453125, + "learning_rate": 0.00010820797177909494, + "loss": 2.4068, + "step": 22292 + }, + { + "epoch": 0.95, + "grad_norm": 0.4375, + "learning_rate": 0.00010820128646322408, + "loss": 2.5903, + "step": 22293 + }, + { + "epoch": 0.95, + "grad_norm": 0.46484375, + "learning_rate": 0.00010819460111045038, + "loss": 2.5899, + "step": 22294 + }, + { + "epoch": 0.95, + "grad_norm": 0.494140625, + "learning_rate": 0.00010818791572080405, + "loss": 2.4224, + "step": 22295 + }, + { + "epoch": 0.95, + "grad_norm": 0.41796875, + "learning_rate": 0.00010818123029431504, + "loss": 2.4744, + "step": 22296 + }, + { + "epoch": 0.95, + "grad_norm": 0.47265625, + "learning_rate": 0.00010817454483101347, + "loss": 2.5368, + "step": 22297 + }, + { + "epoch": 0.95, + "grad_norm": 0.462890625, + "learning_rate": 0.00010816785933092948, + "loss": 2.3936, + "step": 22298 + }, + { + "epoch": 0.95, + "grad_norm": 0.498046875, + "learning_rate": 0.00010816117379409307, + "loss": 2.3992, + "step": 22299 + }, + { + "epoch": 0.95, + "grad_norm": 0.44921875, + "learning_rate": 0.0001081544882205344, + "loss": 2.5654, + "step": 22300 + }, + { + "epoch": 0.95, + "grad_norm": 0.44140625, + "learning_rate": 0.00010814780261028347, + "loss": 2.5647, + "step": 22301 + }, + { + "epoch": 0.95, + "grad_norm": 0.41796875, + "learning_rate": 0.00010814111696337042, + "loss": 2.5324, + "step": 22302 + }, + { + "epoch": 0.95, + "grad_norm": 0.4375, + "learning_rate": 0.00010813443127982535, + "loss": 2.502, + "step": 22303 + }, + { + "epoch": 0.95, + "grad_norm": 0.515625, + "learning_rate": 0.00010812774555967829, + "loss": 2.5546, + "step": 22304 + }, + { + "epoch": 0.95, + "grad_norm": 0.4375, + "learning_rate": 0.00010812105980295933, + "loss": 2.6729, + "step": 22305 + }, + { + "epoch": 0.95, + "grad_norm": 0.55078125, + "learning_rate": 0.0001081143740096986, + "loss": 2.6134, + "step": 22306 + }, + { + "epoch": 0.95, + "grad_norm": 0.45703125, + "learning_rate": 0.00010810768817992612, + "loss": 2.599, + "step": 22307 + }, + { + "epoch": 0.95, + "grad_norm": 0.4921875, + "learning_rate": 0.00010810100231367201, + "loss": 2.6432, + "step": 22308 + }, + { + "epoch": 0.95, + "grad_norm": 0.43359375, + "learning_rate": 0.00010809431641096637, + "loss": 2.5259, + "step": 22309 + }, + { + "epoch": 0.95, + "grad_norm": 0.45703125, + "learning_rate": 0.00010808763047183926, + "loss": 2.4902, + "step": 22310 + }, + { + "epoch": 0.95, + "grad_norm": 0.4140625, + "learning_rate": 0.00010808094449632076, + "loss": 2.7234, + "step": 22311 + }, + { + "epoch": 0.95, + "grad_norm": 0.40625, + "learning_rate": 0.00010807425848444097, + "loss": 2.5942, + "step": 22312 + }, + { + "epoch": 0.95, + "grad_norm": 0.462890625, + "learning_rate": 0.00010806757243622994, + "loss": 2.4384, + "step": 22313 + }, + { + "epoch": 0.95, + "grad_norm": 0.4375, + "learning_rate": 0.00010806088635171784, + "loss": 2.4242, + "step": 22314 + }, + { + "epoch": 0.95, + "grad_norm": 0.435546875, + "learning_rate": 0.00010805420023093465, + "loss": 2.5498, + "step": 22315 + }, + { + "epoch": 0.95, + "grad_norm": 0.466796875, + "learning_rate": 0.00010804751407391052, + "loss": 2.4577, + "step": 22316 + }, + { + "epoch": 0.95, + "grad_norm": 0.515625, + "learning_rate": 0.00010804082788067551, + "loss": 2.6549, + "step": 22317 + }, + { + "epoch": 0.95, + "grad_norm": 0.439453125, + "learning_rate": 0.00010803414165125971, + "loss": 2.5206, + "step": 22318 + }, + { + "epoch": 0.95, + "grad_norm": 0.4453125, + "learning_rate": 0.00010802745538569323, + "loss": 2.3927, + "step": 22319 + }, + { + "epoch": 0.95, + "grad_norm": 0.51171875, + "learning_rate": 0.00010802076908400611, + "loss": 2.5169, + "step": 22320 + }, + { + "epoch": 0.95, + "grad_norm": 0.41796875, + "learning_rate": 0.00010801408274622846, + "loss": 2.5445, + "step": 22321 + }, + { + "epoch": 0.95, + "grad_norm": 0.48828125, + "learning_rate": 0.00010800739637239039, + "loss": 2.6547, + "step": 22322 + }, + { + "epoch": 0.95, + "grad_norm": 0.419921875, + "learning_rate": 0.00010800070996252194, + "loss": 2.472, + "step": 22323 + }, + { + "epoch": 0.95, + "grad_norm": 0.427734375, + "learning_rate": 0.00010799402351665322, + "loss": 2.5502, + "step": 22324 + }, + { + "epoch": 0.95, + "grad_norm": 0.47265625, + "learning_rate": 0.00010798733703481434, + "loss": 2.6777, + "step": 22325 + }, + { + "epoch": 0.95, + "grad_norm": 0.482421875, + "learning_rate": 0.00010798065051703535, + "loss": 2.6525, + "step": 22326 + }, + { + "epoch": 0.95, + "grad_norm": 0.419921875, + "learning_rate": 0.00010797396396334634, + "loss": 2.4971, + "step": 22327 + }, + { + "epoch": 0.95, + "grad_norm": 0.44921875, + "learning_rate": 0.00010796727737377741, + "loss": 2.6264, + "step": 22328 + }, + { + "epoch": 0.95, + "grad_norm": 0.4609375, + "learning_rate": 0.00010796059074835863, + "loss": 2.5547, + "step": 22329 + }, + { + "epoch": 0.95, + "grad_norm": 0.4921875, + "learning_rate": 0.00010795390408712014, + "loss": 2.483, + "step": 22330 + }, + { + "epoch": 0.95, + "grad_norm": 0.470703125, + "learning_rate": 0.00010794721739009195, + "loss": 2.5223, + "step": 22331 + }, + { + "epoch": 0.95, + "grad_norm": 0.453125, + "learning_rate": 0.0001079405306573042, + "loss": 2.4071, + "step": 22332 + }, + { + "epoch": 0.95, + "grad_norm": 0.48046875, + "learning_rate": 0.00010793384388878697, + "loss": 2.6606, + "step": 22333 + }, + { + "epoch": 0.95, + "grad_norm": 0.435546875, + "learning_rate": 0.00010792715708457033, + "loss": 2.6482, + "step": 22334 + }, + { + "epoch": 0.95, + "grad_norm": 0.4140625, + "learning_rate": 0.00010792047024468439, + "loss": 2.4512, + "step": 22335 + }, + { + "epoch": 0.95, + "grad_norm": 0.5078125, + "learning_rate": 0.00010791378336915925, + "loss": 2.5432, + "step": 22336 + }, + { + "epoch": 0.95, + "grad_norm": 0.5390625, + "learning_rate": 0.00010790709645802493, + "loss": 2.499, + "step": 22337 + }, + { + "epoch": 0.95, + "grad_norm": 0.458984375, + "learning_rate": 0.0001079004095113116, + "loss": 2.5117, + "step": 22338 + }, + { + "epoch": 0.95, + "grad_norm": 0.416015625, + "learning_rate": 0.00010789372252904928, + "loss": 2.6304, + "step": 22339 + }, + { + "epoch": 0.95, + "grad_norm": 0.439453125, + "learning_rate": 0.00010788703551126813, + "loss": 2.6472, + "step": 22340 + }, + { + "epoch": 0.95, + "grad_norm": 0.486328125, + "learning_rate": 0.00010788034845799819, + "loss": 2.5102, + "step": 22341 + }, + { + "epoch": 0.95, + "grad_norm": 0.4765625, + "learning_rate": 0.00010787366136926955, + "loss": 2.5616, + "step": 22342 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010786697424511235, + "loss": 2.5624, + "step": 22343 + }, + { + "epoch": 0.95, + "grad_norm": 0.41796875, + "learning_rate": 0.0001078602870855566, + "loss": 2.5747, + "step": 22344 + }, + { + "epoch": 0.95, + "grad_norm": 0.462890625, + "learning_rate": 0.00010785359989063242, + "loss": 2.7245, + "step": 22345 + }, + { + "epoch": 0.95, + "grad_norm": 0.40625, + "learning_rate": 0.00010784691266036996, + "loss": 2.5564, + "step": 22346 + }, + { + "epoch": 0.95, + "grad_norm": 0.46484375, + "learning_rate": 0.0001078402253947992, + "loss": 2.5706, + "step": 22347 + }, + { + "epoch": 0.95, + "grad_norm": 0.43359375, + "learning_rate": 0.00010783353809395033, + "loss": 2.445, + "step": 22348 + }, + { + "epoch": 0.95, + "grad_norm": 0.43359375, + "learning_rate": 0.00010782685075785341, + "loss": 2.5044, + "step": 22349 + }, + { + "epoch": 0.95, + "grad_norm": 0.47265625, + "learning_rate": 0.0001078201633865385, + "loss": 2.6239, + "step": 22350 + }, + { + "epoch": 0.95, + "grad_norm": 0.4140625, + "learning_rate": 0.00010781347598003575, + "loss": 2.545, + "step": 22351 + }, + { + "epoch": 0.95, + "grad_norm": 0.44140625, + "learning_rate": 0.00010780678853837518, + "loss": 2.5326, + "step": 22352 + }, + { + "epoch": 0.95, + "grad_norm": 0.478515625, + "learning_rate": 0.0001078001010615869, + "loss": 2.6394, + "step": 22353 + }, + { + "epoch": 0.95, + "grad_norm": 0.42578125, + "learning_rate": 0.00010779341354970108, + "loss": 2.4566, + "step": 22354 + }, + { + "epoch": 0.95, + "grad_norm": 0.451171875, + "learning_rate": 0.00010778672600274767, + "loss": 2.5564, + "step": 22355 + }, + { + "epoch": 0.95, + "grad_norm": 0.470703125, + "learning_rate": 0.00010778003842075688, + "loss": 2.5557, + "step": 22356 + }, + { + "epoch": 0.95, + "grad_norm": 0.470703125, + "learning_rate": 0.00010777335080375875, + "loss": 2.6684, + "step": 22357 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010776666315178339, + "loss": 2.5211, + "step": 22358 + }, + { + "epoch": 0.95, + "grad_norm": 0.51171875, + "learning_rate": 0.00010775997546486087, + "loss": 2.618, + "step": 22359 + }, + { + "epoch": 0.95, + "grad_norm": 0.44921875, + "learning_rate": 0.00010775328774302131, + "loss": 2.6429, + "step": 22360 + }, + { + "epoch": 0.95, + "grad_norm": 0.494140625, + "learning_rate": 0.00010774659998629477, + "loss": 2.7426, + "step": 22361 + }, + { + "epoch": 0.95, + "grad_norm": 0.51171875, + "learning_rate": 0.0001077399121947114, + "loss": 2.5609, + "step": 22362 + }, + { + "epoch": 0.95, + "grad_norm": 0.470703125, + "learning_rate": 0.0001077332243683012, + "loss": 2.6041, + "step": 22363 + }, + { + "epoch": 0.95, + "grad_norm": 0.439453125, + "learning_rate": 0.00010772653650709434, + "loss": 2.4441, + "step": 22364 + }, + { + "epoch": 0.95, + "grad_norm": 0.47265625, + "learning_rate": 0.00010771984861112093, + "loss": 2.6158, + "step": 22365 + }, + { + "epoch": 0.95, + "grad_norm": 0.490234375, + "learning_rate": 0.00010771316068041097, + "loss": 2.5325, + "step": 22366 + }, + { + "epoch": 0.95, + "grad_norm": 0.4765625, + "learning_rate": 0.00010770647271499463, + "loss": 2.4581, + "step": 22367 + }, + { + "epoch": 0.95, + "grad_norm": 0.4921875, + "learning_rate": 0.00010769978471490196, + "loss": 2.6376, + "step": 22368 + }, + { + "epoch": 0.95, + "grad_norm": 0.4375, + "learning_rate": 0.00010769309668016308, + "loss": 2.5792, + "step": 22369 + }, + { + "epoch": 0.95, + "grad_norm": 0.416015625, + "learning_rate": 0.0001076864086108081, + "loss": 2.4815, + "step": 22370 + }, + { + "epoch": 0.95, + "grad_norm": 0.470703125, + "learning_rate": 0.00010767972050686706, + "loss": 2.4953, + "step": 22371 + }, + { + "epoch": 0.95, + "grad_norm": 0.46484375, + "learning_rate": 0.0001076730323683701, + "loss": 2.4689, + "step": 22372 + }, + { + "epoch": 0.95, + "grad_norm": 0.421875, + "learning_rate": 0.0001076663441953473, + "loss": 2.5221, + "step": 22373 + }, + { + "epoch": 0.95, + "grad_norm": 0.458984375, + "learning_rate": 0.00010765965598782875, + "loss": 2.4305, + "step": 22374 + }, + { + "epoch": 0.95, + "grad_norm": 0.4921875, + "learning_rate": 0.00010765296774584455, + "loss": 2.5074, + "step": 22375 + }, + { + "epoch": 0.95, + "grad_norm": 0.439453125, + "learning_rate": 0.00010764627946942478, + "loss": 2.5384, + "step": 22376 + }, + { + "epoch": 0.95, + "grad_norm": 0.44921875, + "learning_rate": 0.00010763959115859953, + "loss": 2.607, + "step": 22377 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.00010763290281339897, + "loss": 2.448, + "step": 22378 + }, + { + "epoch": 0.95, + "grad_norm": 0.498046875, + "learning_rate": 0.00010762621443385308, + "loss": 2.5905, + "step": 22379 + }, + { + "epoch": 0.95, + "grad_norm": 0.4609375, + "learning_rate": 0.00010761952601999204, + "loss": 2.3987, + "step": 22380 + }, + { + "epoch": 0.95, + "grad_norm": 0.5234375, + "learning_rate": 0.0001076128375718459, + "loss": 2.7286, + "step": 22381 + }, + { + "epoch": 0.95, + "grad_norm": 0.45703125, + "learning_rate": 0.0001076061490894448, + "loss": 2.643, + "step": 22382 + }, + { + "epoch": 0.95, + "grad_norm": 0.462890625, + "learning_rate": 0.0001075994605728188, + "loss": 2.5895, + "step": 22383 + }, + { + "epoch": 0.95, + "grad_norm": 0.41015625, + "learning_rate": 0.000107592772021998, + "loss": 2.5024, + "step": 22384 + }, + { + "epoch": 0.95, + "grad_norm": 0.478515625, + "learning_rate": 0.00010758608343701246, + "loss": 2.6191, + "step": 22385 + }, + { + "epoch": 0.95, + "grad_norm": 0.431640625, + "learning_rate": 0.00010757939481789239, + "loss": 2.5746, + "step": 22386 + }, + { + "epoch": 0.95, + "grad_norm": 0.470703125, + "learning_rate": 0.00010757270616466776, + "loss": 2.528, + "step": 22387 + }, + { + "epoch": 0.95, + "grad_norm": 0.46484375, + "learning_rate": 0.00010756601747736875, + "loss": 2.5019, + "step": 22388 + }, + { + "epoch": 0.95, + "grad_norm": 0.49609375, + "learning_rate": 0.00010755932875602542, + "loss": 2.6886, + "step": 22389 + }, + { + "epoch": 0.95, + "grad_norm": 0.4453125, + "learning_rate": 0.00010755264000066788, + "loss": 2.5009, + "step": 22390 + }, + { + "epoch": 0.95, + "grad_norm": 0.46484375, + "learning_rate": 0.0001075459512113262, + "loss": 2.5216, + "step": 22391 + }, + { + "epoch": 0.95, + "grad_norm": 0.47265625, + "learning_rate": 0.0001075392623880305, + "loss": 2.5323, + "step": 22392 + }, + { + "epoch": 0.95, + "grad_norm": 0.49609375, + "learning_rate": 0.00010753257353081086, + "loss": 2.6707, + "step": 22393 + }, + { + "epoch": 0.95, + "grad_norm": 0.486328125, + "learning_rate": 0.00010752588463969745, + "loss": 2.5786, + "step": 22394 + }, + { + "epoch": 0.95, + "grad_norm": 0.453125, + "learning_rate": 0.00010751919571472024, + "loss": 2.5849, + "step": 22395 + }, + { + "epoch": 0.95, + "grad_norm": 0.46875, + "learning_rate": 0.00010751250675590942, + "loss": 2.597, + "step": 22396 + }, + { + "epoch": 0.95, + "grad_norm": 0.431640625, + "learning_rate": 0.00010750581776329509, + "loss": 2.5528, + "step": 22397 + }, + { + "epoch": 0.95, + "grad_norm": 0.447265625, + "learning_rate": 0.0001074991287369073, + "loss": 2.499, + "step": 22398 + }, + { + "epoch": 0.95, + "grad_norm": 0.44921875, + "learning_rate": 0.00010749243967677621, + "loss": 2.5191, + "step": 22399 + }, + { + "epoch": 0.95, + "grad_norm": 0.498046875, + "learning_rate": 0.00010748575058293185, + "loss": 2.6729, + "step": 22400 + }, + { + "epoch": 0.95, + "grad_norm": 0.474609375, + "learning_rate": 0.00010747906145540432, + "loss": 2.5743, + "step": 22401 + }, + { + "epoch": 0.95, + "grad_norm": 0.466796875, + "learning_rate": 0.0001074723722942238, + "loss": 2.4674, + "step": 22402 + }, + { + "epoch": 0.95, + "grad_norm": 0.427734375, + "learning_rate": 0.00010746568309942029, + "loss": 2.5275, + "step": 22403 + }, + { + "epoch": 0.95, + "grad_norm": 0.416015625, + "learning_rate": 0.00010745899387102393, + "loss": 2.4386, + "step": 22404 + }, + { + "epoch": 0.95, + "grad_norm": 0.47265625, + "learning_rate": 0.00010745230460906486, + "loss": 2.5753, + "step": 22405 + }, + { + "epoch": 0.95, + "grad_norm": 0.4296875, + "learning_rate": 0.00010744561531357314, + "loss": 2.5897, + "step": 22406 + }, + { + "epoch": 0.95, + "grad_norm": 0.4296875, + "learning_rate": 0.00010743892598457885, + "loss": 2.4743, + "step": 22407 + }, + { + "epoch": 0.95, + "grad_norm": 0.41796875, + "learning_rate": 0.00010743223662211215, + "loss": 2.5443, + "step": 22408 + }, + { + "epoch": 0.95, + "grad_norm": 0.40234375, + "learning_rate": 0.00010742554722620304, + "loss": 2.4831, + "step": 22409 + }, + { + "epoch": 0.95, + "grad_norm": 0.44921875, + "learning_rate": 0.00010741885779688175, + "loss": 2.5543, + "step": 22410 + }, + { + "epoch": 0.95, + "grad_norm": 0.443359375, + "learning_rate": 0.00010741216833417824, + "loss": 2.5796, + "step": 22411 + }, + { + "epoch": 0.96, + "grad_norm": 0.43359375, + "learning_rate": 0.00010740547883812274, + "loss": 2.4587, + "step": 22412 + }, + { + "epoch": 0.96, + "grad_norm": 0.482421875, + "learning_rate": 0.00010739878930874527, + "loss": 2.5657, + "step": 22413 + }, + { + "epoch": 0.96, + "grad_norm": 0.4453125, + "learning_rate": 0.00010739209974607596, + "loss": 2.6573, + "step": 22414 + }, + { + "epoch": 0.96, + "grad_norm": 0.486328125, + "learning_rate": 0.00010738541015014489, + "loss": 2.6203, + "step": 22415 + }, + { + "epoch": 0.96, + "grad_norm": 0.4453125, + "learning_rate": 0.0001073787205209822, + "loss": 2.5178, + "step": 22416 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 0.00010737203085861791, + "loss": 2.5722, + "step": 22417 + }, + { + "epoch": 0.96, + "grad_norm": 0.447265625, + "learning_rate": 0.00010736534116308223, + "loss": 2.6228, + "step": 22418 + }, + { + "epoch": 0.96, + "grad_norm": 0.474609375, + "learning_rate": 0.00010735865143440516, + "loss": 2.4816, + "step": 22419 + }, + { + "epoch": 0.96, + "grad_norm": 0.45703125, + "learning_rate": 0.00010735196167261687, + "loss": 2.5743, + "step": 22420 + }, + { + "epoch": 0.96, + "grad_norm": 0.55078125, + "learning_rate": 0.00010734527187774746, + "loss": 2.5937, + "step": 22421 + }, + { + "epoch": 0.96, + "grad_norm": 0.416015625, + "learning_rate": 0.00010733858204982699, + "loss": 2.6688, + "step": 22422 + }, + { + "epoch": 0.96, + "grad_norm": 0.4921875, + "learning_rate": 0.00010733189218888558, + "loss": 2.4993, + "step": 22423 + }, + { + "epoch": 0.96, + "grad_norm": 0.390625, + "learning_rate": 0.00010732520229495335, + "loss": 2.4311, + "step": 22424 + }, + { + "epoch": 0.96, + "grad_norm": 0.443359375, + "learning_rate": 0.00010731851236806034, + "loss": 2.6138, + "step": 22425 + }, + { + "epoch": 0.96, + "grad_norm": 0.4921875, + "learning_rate": 0.00010731182240823676, + "loss": 2.5607, + "step": 22426 + }, + { + "epoch": 0.96, + "grad_norm": 0.4375, + "learning_rate": 0.00010730513241551263, + "loss": 2.5067, + "step": 22427 + }, + { + "epoch": 0.96, + "grad_norm": 0.44921875, + "learning_rate": 0.00010729844238991806, + "loss": 2.7341, + "step": 22428 + }, + { + "epoch": 0.96, + "grad_norm": 0.470703125, + "learning_rate": 0.0001072917523314832, + "loss": 2.4845, + "step": 22429 + }, + { + "epoch": 0.96, + "grad_norm": 0.3828125, + "learning_rate": 0.00010728506224023808, + "loss": 2.3673, + "step": 22430 + }, + { + "epoch": 0.96, + "grad_norm": 0.453125, + "learning_rate": 0.00010727837211621286, + "loss": 2.5549, + "step": 22431 + }, + { + "epoch": 0.96, + "grad_norm": 0.48046875, + "learning_rate": 0.00010727168195943762, + "loss": 2.4287, + "step": 22432 + }, + { + "epoch": 0.96, + "grad_norm": 0.51171875, + "learning_rate": 0.00010726499176994249, + "loss": 2.4456, + "step": 22433 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010725830154775753, + "loss": 2.3865, + "step": 22434 + }, + { + "epoch": 0.96, + "grad_norm": 0.48046875, + "learning_rate": 0.00010725161129291288, + "loss": 2.7187, + "step": 22435 + }, + { + "epoch": 0.96, + "grad_norm": 0.412109375, + "learning_rate": 0.00010724492100543859, + "loss": 2.3381, + "step": 22436 + }, + { + "epoch": 0.96, + "grad_norm": 0.484375, + "learning_rate": 0.00010723823068536487, + "loss": 2.4866, + "step": 22437 + }, + { + "epoch": 0.96, + "grad_norm": 0.453125, + "learning_rate": 0.0001072315403327217, + "loss": 2.5994, + "step": 22438 + }, + { + "epoch": 0.96, + "grad_norm": 0.453125, + "learning_rate": 0.00010722484994753925, + "loss": 2.5421, + "step": 22439 + }, + { + "epoch": 0.96, + "grad_norm": 0.51171875, + "learning_rate": 0.00010721815952984764, + "loss": 2.5332, + "step": 22440 + }, + { + "epoch": 0.96, + "grad_norm": 0.42578125, + "learning_rate": 0.00010721146907967694, + "loss": 2.4317, + "step": 22441 + }, + { + "epoch": 0.96, + "grad_norm": 0.42578125, + "learning_rate": 0.00010720477859705727, + "loss": 2.5434, + "step": 22442 + }, + { + "epoch": 0.96, + "grad_norm": 0.412109375, + "learning_rate": 0.00010719808808201872, + "loss": 2.5133, + "step": 22443 + }, + { + "epoch": 0.96, + "grad_norm": 0.4453125, + "learning_rate": 0.00010719139753459137, + "loss": 2.6132, + "step": 22444 + }, + { + "epoch": 0.96, + "grad_norm": 0.47265625, + "learning_rate": 0.00010718470695480543, + "loss": 2.4459, + "step": 22445 + }, + { + "epoch": 0.96, + "grad_norm": 0.42578125, + "learning_rate": 0.0001071780163426909, + "loss": 2.5095, + "step": 22446 + }, + { + "epoch": 0.96, + "grad_norm": 0.49609375, + "learning_rate": 0.0001071713256982779, + "loss": 2.5927, + "step": 22447 + }, + { + "epoch": 0.96, + "grad_norm": 0.453125, + "learning_rate": 0.00010716463502159656, + "loss": 2.6374, + "step": 22448 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 0.00010715794431267702, + "loss": 2.5783, + "step": 22449 + }, + { + "epoch": 0.96, + "grad_norm": 0.4765625, + "learning_rate": 0.00010715125357154931, + "loss": 2.5434, + "step": 22450 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010714456279824361, + "loss": 2.5147, + "step": 22451 + }, + { + "epoch": 0.96, + "grad_norm": 0.416015625, + "learning_rate": 0.00010713787199278993, + "loss": 2.4824, + "step": 22452 + }, + { + "epoch": 0.96, + "grad_norm": 0.46484375, + "learning_rate": 0.0001071311811552185, + "loss": 2.6118, + "step": 22453 + }, + { + "epoch": 0.96, + "grad_norm": 0.4921875, + "learning_rate": 0.0001071244902855593, + "loss": 2.563, + "step": 22454 + }, + { + "epoch": 0.96, + "grad_norm": 0.498046875, + "learning_rate": 0.00010711779938384252, + "loss": 2.6254, + "step": 22455 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010711110845009827, + "loss": 2.5744, + "step": 22456 + }, + { + "epoch": 0.96, + "grad_norm": 0.455078125, + "learning_rate": 0.0001071044174843566, + "loss": 2.5718, + "step": 22457 + }, + { + "epoch": 0.96, + "grad_norm": 0.490234375, + "learning_rate": 0.00010709772648664766, + "loss": 2.5302, + "step": 22458 + }, + { + "epoch": 0.96, + "grad_norm": 0.447265625, + "learning_rate": 0.00010709103545700155, + "loss": 2.592, + "step": 22459 + }, + { + "epoch": 0.96, + "grad_norm": 0.474609375, + "learning_rate": 0.00010708434439544834, + "loss": 2.617, + "step": 22460 + }, + { + "epoch": 0.96, + "grad_norm": 0.474609375, + "learning_rate": 0.00010707765330201823, + "loss": 2.5523, + "step": 22461 + }, + { + "epoch": 0.96, + "grad_norm": 0.49609375, + "learning_rate": 0.0001070709621767412, + "loss": 2.6911, + "step": 22462 + }, + { + "epoch": 0.96, + "grad_norm": 0.5, + "learning_rate": 0.00010706427101964747, + "loss": 2.5462, + "step": 22463 + }, + { + "epoch": 0.96, + "grad_norm": 0.48828125, + "learning_rate": 0.00010705757983076707, + "loss": 2.4198, + "step": 22464 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010705088861013017, + "loss": 2.5465, + "step": 22465 + }, + { + "epoch": 0.96, + "grad_norm": 0.46484375, + "learning_rate": 0.00010704419735776683, + "loss": 2.6234, + "step": 22466 + }, + { + "epoch": 0.96, + "grad_norm": 0.4453125, + "learning_rate": 0.00010703750607370719, + "loss": 2.5443, + "step": 22467 + }, + { + "epoch": 0.96, + "grad_norm": 0.419921875, + "learning_rate": 0.00010703081475798131, + "loss": 2.6179, + "step": 22468 + }, + { + "epoch": 0.96, + "grad_norm": 0.4375, + "learning_rate": 0.00010702412341061938, + "loss": 2.5584, + "step": 22469 + }, + { + "epoch": 0.96, + "grad_norm": 0.45703125, + "learning_rate": 0.00010701743203165144, + "loss": 2.5715, + "step": 22470 + }, + { + "epoch": 0.96, + "grad_norm": 0.427734375, + "learning_rate": 0.00010701074062110761, + "loss": 2.4947, + "step": 22471 + }, + { + "epoch": 0.96, + "grad_norm": 0.408203125, + "learning_rate": 0.00010700404917901803, + "loss": 2.6672, + "step": 22472 + }, + { + "epoch": 0.96, + "grad_norm": 0.4609375, + "learning_rate": 0.00010699735770541277, + "loss": 2.5151, + "step": 22473 + }, + { + "epoch": 0.96, + "grad_norm": 0.46484375, + "learning_rate": 0.00010699066620032199, + "loss": 2.5171, + "step": 22474 + }, + { + "epoch": 0.96, + "grad_norm": 0.4375, + "learning_rate": 0.00010698397466377574, + "loss": 2.512, + "step": 22475 + }, + { + "epoch": 0.96, + "grad_norm": 0.4375, + "learning_rate": 0.00010697728309580413, + "loss": 2.5401, + "step": 22476 + }, + { + "epoch": 0.96, + "grad_norm": 0.498046875, + "learning_rate": 0.00010697059149643736, + "loss": 2.4972, + "step": 22477 + }, + { + "epoch": 0.96, + "grad_norm": 0.466796875, + "learning_rate": 0.00010696389986570541, + "loss": 2.4397, + "step": 22478 + }, + { + "epoch": 0.96, + "grad_norm": 0.43359375, + "learning_rate": 0.0001069572082036385, + "loss": 2.5811, + "step": 22479 + }, + { + "epoch": 0.96, + "grad_norm": 0.451171875, + "learning_rate": 0.00010695051651026668, + "loss": 2.569, + "step": 22480 + }, + { + "epoch": 0.96, + "grad_norm": 0.486328125, + "learning_rate": 0.00010694382478562009, + "loss": 2.6758, + "step": 22481 + }, + { + "epoch": 0.96, + "grad_norm": 0.427734375, + "learning_rate": 0.0001069371330297288, + "loss": 2.6701, + "step": 22482 + }, + { + "epoch": 0.96, + "grad_norm": 0.46484375, + "learning_rate": 0.00010693044124262297, + "loss": 2.6551, + "step": 22483 + }, + { + "epoch": 0.96, + "grad_norm": 0.48046875, + "learning_rate": 0.00010692374942433266, + "loss": 2.6315, + "step": 22484 + }, + { + "epoch": 0.96, + "grad_norm": 0.4140625, + "learning_rate": 0.00010691705757488806, + "loss": 2.5923, + "step": 22485 + }, + { + "epoch": 0.96, + "grad_norm": 0.478515625, + "learning_rate": 0.00010691036569431916, + "loss": 2.5718, + "step": 22486 + }, + { + "epoch": 0.96, + "grad_norm": 0.44921875, + "learning_rate": 0.00010690367378265618, + "loss": 2.6448, + "step": 22487 + }, + { + "epoch": 0.96, + "grad_norm": 0.412109375, + "learning_rate": 0.0001068969818399292, + "loss": 2.4912, + "step": 22488 + }, + { + "epoch": 0.96, + "grad_norm": 0.48046875, + "learning_rate": 0.00010689028986616831, + "loss": 2.5458, + "step": 22489 + }, + { + "epoch": 0.96, + "grad_norm": 0.51953125, + "learning_rate": 0.00010688359786140363, + "loss": 2.7396, + "step": 22490 + }, + { + "epoch": 0.96, + "grad_norm": 0.423828125, + "learning_rate": 0.00010687690582566527, + "loss": 2.649, + "step": 22491 + }, + { + "epoch": 0.96, + "grad_norm": 0.40625, + "learning_rate": 0.00010687021375898334, + "loss": 2.3523, + "step": 22492 + }, + { + "epoch": 0.96, + "grad_norm": 0.39453125, + "learning_rate": 0.000106863521661388, + "loss": 2.4906, + "step": 22493 + }, + { + "epoch": 0.96, + "grad_norm": 0.40234375, + "learning_rate": 0.00010685682953290928, + "loss": 2.5235, + "step": 22494 + }, + { + "epoch": 0.96, + "grad_norm": 0.44921875, + "learning_rate": 0.00010685013737357734, + "loss": 2.5528, + "step": 22495 + }, + { + "epoch": 0.96, + "grad_norm": 0.4140625, + "learning_rate": 0.0001068434451834223, + "loss": 2.6345, + "step": 22496 + }, + { + "epoch": 0.96, + "grad_norm": 0.447265625, + "learning_rate": 0.00010683675296247426, + "loss": 2.4992, + "step": 22497 + }, + { + "epoch": 0.96, + "grad_norm": 0.443359375, + "learning_rate": 0.00010683006071076331, + "loss": 2.5539, + "step": 22498 + }, + { + "epoch": 0.96, + "grad_norm": 0.47265625, + "learning_rate": 0.0001068233684283196, + "loss": 2.6353, + "step": 22499 + }, + { + "epoch": 0.96, + "grad_norm": 0.55859375, + "learning_rate": 0.00010681667611517321, + "loss": 2.5846, + "step": 22500 + }, + { + "epoch": 0.96, + "grad_norm": 0.400390625, + "learning_rate": 0.0001068099837713543, + "loss": 2.4938, + "step": 22501 + }, + { + "epoch": 0.96, + "grad_norm": 0.46484375, + "learning_rate": 0.00010680329139689289, + "loss": 2.4574, + "step": 22502 + }, + { + "epoch": 0.96, + "grad_norm": 0.478515625, + "learning_rate": 0.0001067965989918192, + "loss": 2.4753, + "step": 22503 + }, + { + "epoch": 0.96, + "grad_norm": 0.482421875, + "learning_rate": 0.0001067899065561633, + "loss": 2.5261, + "step": 22504 + }, + { + "epoch": 0.96, + "grad_norm": 0.41796875, + "learning_rate": 0.0001067832140899553, + "loss": 2.5873, + "step": 22505 + }, + { + "epoch": 0.96, + "grad_norm": 0.423828125, + "learning_rate": 0.00010677652159322531, + "loss": 2.6516, + "step": 22506 + }, + { + "epoch": 0.96, + "grad_norm": 0.48046875, + "learning_rate": 0.00010676982906600345, + "loss": 2.5233, + "step": 22507 + }, + { + "epoch": 0.96, + "grad_norm": 0.439453125, + "learning_rate": 0.00010676313650831982, + "loss": 2.4422, + "step": 22508 + }, + { + "epoch": 0.96, + "grad_norm": 0.427734375, + "learning_rate": 0.00010675644392020458, + "loss": 2.4713, + "step": 22509 + }, + { + "epoch": 0.96, + "grad_norm": 0.482421875, + "learning_rate": 0.00010674975130168778, + "loss": 2.5969, + "step": 22510 + }, + { + "epoch": 0.96, + "grad_norm": 0.37890625, + "learning_rate": 0.00010674305865279957, + "loss": 2.4906, + "step": 22511 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 0.00010673636597357007, + "loss": 2.522, + "step": 22512 + }, + { + "epoch": 0.96, + "grad_norm": 0.466796875, + "learning_rate": 0.00010672967326402938, + "loss": 2.6052, + "step": 22513 + }, + { + "epoch": 0.96, + "grad_norm": 0.46484375, + "learning_rate": 0.00010672298052420764, + "loss": 2.5965, + "step": 22514 + }, + { + "epoch": 0.96, + "grad_norm": 0.427734375, + "learning_rate": 0.00010671628775413492, + "loss": 2.4011, + "step": 22515 + }, + { + "epoch": 0.96, + "grad_norm": 0.451171875, + "learning_rate": 0.00010670959495384134, + "loss": 2.6273, + "step": 22516 + }, + { + "epoch": 0.96, + "grad_norm": 0.453125, + "learning_rate": 0.00010670290212335709, + "loss": 2.6304, + "step": 22517 + }, + { + "epoch": 0.96, + "grad_norm": 0.396484375, + "learning_rate": 0.0001066962092627122, + "loss": 2.5799, + "step": 22518 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 0.00010668951637193681, + "loss": 2.5667, + "step": 22519 + }, + { + "epoch": 0.96, + "grad_norm": 0.486328125, + "learning_rate": 0.00010668282345106105, + "loss": 2.3763, + "step": 22520 + }, + { + "epoch": 0.96, + "grad_norm": 0.462890625, + "learning_rate": 0.00010667613050011502, + "loss": 2.5276, + "step": 22521 + }, + { + "epoch": 0.96, + "grad_norm": 0.474609375, + "learning_rate": 0.00010666943751912885, + "loss": 2.4597, + "step": 22522 + }, + { + "epoch": 0.96, + "grad_norm": 0.498046875, + "learning_rate": 0.00010666274450813265, + "loss": 2.6359, + "step": 22523 + }, + { + "epoch": 0.96, + "grad_norm": 0.4765625, + "learning_rate": 0.00010665605146715651, + "loss": 2.5122, + "step": 22524 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 0.00010664935839623062, + "loss": 2.6678, + "step": 22525 + }, + { + "epoch": 0.96, + "grad_norm": 0.44921875, + "learning_rate": 0.000106642665295385, + "loss": 2.3053, + "step": 22526 + }, + { + "epoch": 0.96, + "grad_norm": 0.42578125, + "learning_rate": 0.00010663597216464985, + "loss": 2.558, + "step": 22527 + }, + { + "epoch": 0.96, + "grad_norm": 0.5078125, + "learning_rate": 0.00010662927900405521, + "loss": 2.5424, + "step": 22528 + }, + { + "epoch": 0.96, + "grad_norm": 0.45703125, + "learning_rate": 0.00010662258581363127, + "loss": 2.5528, + "step": 22529 + }, + { + "epoch": 0.96, + "grad_norm": 0.4375, + "learning_rate": 0.00010661589259340812, + "loss": 2.4816, + "step": 22530 + }, + { + "epoch": 0.96, + "grad_norm": 0.494140625, + "learning_rate": 0.00010660919934341585, + "loss": 2.594, + "step": 22531 + }, + { + "epoch": 0.96, + "grad_norm": 0.46875, + "learning_rate": 0.00010660250606368458, + "loss": 2.5543, + "step": 22532 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 0.00010659581275424451, + "loss": 2.5399, + "step": 22533 + }, + { + "epoch": 0.96, + "grad_norm": 0.4453125, + "learning_rate": 0.00010658911941512562, + "loss": 2.4453, + "step": 22534 + }, + { + "epoch": 0.96, + "grad_norm": 0.4921875, + "learning_rate": 0.00010658242604635815, + "loss": 2.5532, + "step": 22535 + }, + { + "epoch": 0.96, + "grad_norm": 0.47265625, + "learning_rate": 0.00010657573264797213, + "loss": 2.6084, + "step": 22536 + }, + { + "epoch": 0.96, + "grad_norm": 0.51171875, + "learning_rate": 0.00010656903921999775, + "loss": 2.445, + "step": 22537 + }, + { + "epoch": 0.96, + "grad_norm": 0.408203125, + "learning_rate": 0.00010656234576246506, + "loss": 2.4596, + "step": 22538 + }, + { + "epoch": 0.96, + "grad_norm": 0.484375, + "learning_rate": 0.00010655565227540427, + "loss": 2.7133, + "step": 22539 + }, + { + "epoch": 0.96, + "grad_norm": 0.494140625, + "learning_rate": 0.00010654895875884538, + "loss": 2.5673, + "step": 22540 + }, + { + "epoch": 0.96, + "grad_norm": 0.49609375, + "learning_rate": 0.00010654226521281862, + "loss": 2.71, + "step": 22541 + }, + { + "epoch": 0.96, + "grad_norm": 0.53125, + "learning_rate": 0.00010653557163735401, + "loss": 2.6979, + "step": 22542 + }, + { + "epoch": 0.96, + "grad_norm": 0.4921875, + "learning_rate": 0.00010652887803248174, + "loss": 2.6787, + "step": 22543 + }, + { + "epoch": 0.96, + "grad_norm": 0.51953125, + "learning_rate": 0.00010652218439823192, + "loss": 2.5252, + "step": 22544 + }, + { + "epoch": 0.96, + "grad_norm": 0.421875, + "learning_rate": 0.00010651549073463463, + "loss": 2.8227, + "step": 22545 + }, + { + "epoch": 0.96, + "grad_norm": 0.474609375, + "learning_rate": 0.00010650879704172001, + "loss": 2.6289, + "step": 22546 + }, + { + "epoch": 0.96, + "grad_norm": 0.470703125, + "learning_rate": 0.0001065021033195182, + "loss": 2.6171, + "step": 22547 + }, + { + "epoch": 0.96, + "grad_norm": 0.421875, + "learning_rate": 0.00010649540956805931, + "loss": 2.562, + "step": 22548 + }, + { + "epoch": 0.96, + "grad_norm": 0.455078125, + "learning_rate": 0.00010648871578737343, + "loss": 2.5724, + "step": 22549 + }, + { + "epoch": 0.96, + "grad_norm": 0.4765625, + "learning_rate": 0.00010648202197749071, + "loss": 2.3974, + "step": 22550 + }, + { + "epoch": 0.96, + "grad_norm": 0.423828125, + "learning_rate": 0.00010647532813844124, + "loss": 2.4895, + "step": 22551 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 0.00010646863427025517, + "loss": 2.5237, + "step": 22552 + }, + { + "epoch": 0.96, + "grad_norm": 0.45703125, + "learning_rate": 0.0001064619403729626, + "loss": 2.6728, + "step": 22553 + }, + { + "epoch": 0.96, + "grad_norm": 0.4609375, + "learning_rate": 0.00010645524644659368, + "loss": 2.5269, + "step": 22554 + }, + { + "epoch": 0.96, + "grad_norm": 0.4609375, + "learning_rate": 0.00010644855249117851, + "loss": 2.649, + "step": 22555 + }, + { + "epoch": 0.96, + "grad_norm": 0.43359375, + "learning_rate": 0.00010644185850674721, + "loss": 2.4017, + "step": 22556 + }, + { + "epoch": 0.96, + "grad_norm": 0.451171875, + "learning_rate": 0.00010643516449332988, + "loss": 2.5282, + "step": 22557 + }, + { + "epoch": 0.96, + "grad_norm": 0.47265625, + "learning_rate": 0.00010642847045095669, + "loss": 2.5897, + "step": 22558 + }, + { + "epoch": 0.96, + "grad_norm": 0.53125, + "learning_rate": 0.00010642177637965771, + "loss": 2.678, + "step": 22559 + }, + { + "epoch": 0.96, + "grad_norm": 0.369140625, + "learning_rate": 0.00010641508227946311, + "loss": 2.3798, + "step": 22560 + }, + { + "epoch": 0.96, + "grad_norm": 0.4296875, + "learning_rate": 0.00010640838815040295, + "loss": 2.6449, + "step": 22561 + }, + { + "epoch": 0.96, + "grad_norm": 0.4296875, + "learning_rate": 0.0001064016939925074, + "loss": 2.5739, + "step": 22562 + }, + { + "epoch": 0.96, + "grad_norm": 0.51171875, + "learning_rate": 0.00010639499980580656, + "loss": 2.6193, + "step": 22563 + }, + { + "epoch": 0.96, + "grad_norm": 0.462890625, + "learning_rate": 0.00010638830559033057, + "loss": 2.4888, + "step": 22564 + }, + { + "epoch": 0.96, + "grad_norm": 0.546875, + "learning_rate": 0.00010638161134610953, + "loss": 2.716, + "step": 22565 + }, + { + "epoch": 0.96, + "grad_norm": 0.44140625, + "learning_rate": 0.0001063749170731736, + "loss": 2.6751, + "step": 22566 + }, + { + "epoch": 0.96, + "grad_norm": 0.4375, + "learning_rate": 0.00010636822277155283, + "loss": 2.5816, + "step": 22567 + }, + { + "epoch": 0.96, + "grad_norm": 0.447265625, + "learning_rate": 0.00010636152844127742, + "loss": 2.5872, + "step": 22568 + }, + { + "epoch": 0.96, + "grad_norm": 0.5, + "learning_rate": 0.00010635483408237743, + "loss": 2.6782, + "step": 22569 + }, + { + "epoch": 0.96, + "grad_norm": 0.482421875, + "learning_rate": 0.00010634813969488301, + "loss": 2.5031, + "step": 22570 + }, + { + "epoch": 0.96, + "grad_norm": 0.42578125, + "learning_rate": 0.0001063414452788243, + "loss": 2.4395, + "step": 22571 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010633475083423139, + "loss": 2.476, + "step": 22572 + }, + { + "epoch": 0.96, + "grad_norm": 0.45703125, + "learning_rate": 0.0001063280563611344, + "loss": 2.5921, + "step": 22573 + }, + { + "epoch": 0.96, + "grad_norm": 0.484375, + "learning_rate": 0.00010632136185956351, + "loss": 2.5653, + "step": 22574 + }, + { + "epoch": 0.96, + "grad_norm": 0.515625, + "learning_rate": 0.00010631466732954878, + "loss": 2.6022, + "step": 22575 + }, + { + "epoch": 0.96, + "grad_norm": 0.5234375, + "learning_rate": 0.0001063079727711204, + "loss": 2.606, + "step": 22576 + }, + { + "epoch": 0.96, + "grad_norm": 0.44921875, + "learning_rate": 0.00010630127818430838, + "loss": 2.4691, + "step": 22577 + }, + { + "epoch": 0.96, + "grad_norm": 0.408203125, + "learning_rate": 0.00010629458356914293, + "loss": 2.4827, + "step": 22578 + }, + { + "epoch": 0.96, + "grad_norm": 0.47265625, + "learning_rate": 0.00010628788892565415, + "loss": 2.4547, + "step": 22579 + }, + { + "epoch": 0.96, + "grad_norm": 0.4609375, + "learning_rate": 0.00010628119425387219, + "loss": 2.3944, + "step": 22580 + }, + { + "epoch": 0.96, + "grad_norm": 0.4453125, + "learning_rate": 0.00010627449955382712, + "loss": 2.553, + "step": 22581 + }, + { + "epoch": 0.96, + "grad_norm": 0.47265625, + "learning_rate": 0.00010626780482554915, + "loss": 2.5777, + "step": 22582 + }, + { + "epoch": 0.96, + "grad_norm": 0.49609375, + "learning_rate": 0.0001062611100690683, + "loss": 2.527, + "step": 22583 + }, + { + "epoch": 0.96, + "grad_norm": 0.486328125, + "learning_rate": 0.00010625441528441476, + "loss": 2.5629, + "step": 22584 + }, + { + "epoch": 0.96, + "grad_norm": 0.41015625, + "learning_rate": 0.00010624772047161862, + "loss": 2.4105, + "step": 22585 + }, + { + "epoch": 0.96, + "grad_norm": 0.439453125, + "learning_rate": 0.00010624102563071004, + "loss": 2.5974, + "step": 22586 + }, + { + "epoch": 0.96, + "grad_norm": 0.404296875, + "learning_rate": 0.00010623433076171912, + "loss": 2.4905, + "step": 22587 + }, + { + "epoch": 0.96, + "grad_norm": 0.466796875, + "learning_rate": 0.00010622763586467599, + "loss": 2.5918, + "step": 22588 + }, + { + "epoch": 0.96, + "grad_norm": 0.4765625, + "learning_rate": 0.00010622094093961074, + "loss": 2.417, + "step": 22589 + }, + { + "epoch": 0.96, + "grad_norm": 0.4921875, + "learning_rate": 0.0001062142459865536, + "loss": 2.5073, + "step": 22590 + }, + { + "epoch": 0.96, + "grad_norm": 0.494140625, + "learning_rate": 0.00010620755100553457, + "loss": 2.5322, + "step": 22591 + }, + { + "epoch": 0.96, + "grad_norm": 0.48046875, + "learning_rate": 0.00010620085599658384, + "loss": 2.5753, + "step": 22592 + }, + { + "epoch": 0.96, + "grad_norm": 0.431640625, + "learning_rate": 0.00010619416095973154, + "loss": 2.64, + "step": 22593 + }, + { + "epoch": 0.96, + "grad_norm": 0.455078125, + "learning_rate": 0.00010618746589500779, + "loss": 2.5862, + "step": 22594 + }, + { + "epoch": 0.96, + "grad_norm": 0.53515625, + "learning_rate": 0.00010618077080244268, + "loss": 2.6033, + "step": 22595 + }, + { + "epoch": 0.96, + "grad_norm": 0.3984375, + "learning_rate": 0.00010617407568206637, + "loss": 2.5201, + "step": 22596 + }, + { + "epoch": 0.96, + "grad_norm": 0.490234375, + "learning_rate": 0.00010616738053390895, + "loss": 2.6491, + "step": 22597 + }, + { + "epoch": 0.96, + "grad_norm": 0.48828125, + "learning_rate": 0.00010616068535800062, + "loss": 2.6269, + "step": 22598 + }, + { + "epoch": 0.96, + "grad_norm": 0.40625, + "learning_rate": 0.0001061539901543714, + "loss": 2.512, + "step": 22599 + }, + { + "epoch": 0.96, + "grad_norm": 0.390625, + "learning_rate": 0.00010614729492305151, + "loss": 2.4677, + "step": 22600 + }, + { + "epoch": 0.96, + "grad_norm": 0.455078125, + "learning_rate": 0.00010614059966407104, + "loss": 2.5893, + "step": 22601 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010613390437746011, + "loss": 2.6233, + "step": 22602 + }, + { + "epoch": 0.96, + "grad_norm": 0.41796875, + "learning_rate": 0.00010612720906324885, + "loss": 2.4942, + "step": 22603 + }, + { + "epoch": 0.96, + "grad_norm": 0.455078125, + "learning_rate": 0.00010612051372146739, + "loss": 2.4802, + "step": 22604 + }, + { + "epoch": 0.96, + "grad_norm": 0.3984375, + "learning_rate": 0.00010611381835214585, + "loss": 2.4372, + "step": 22605 + }, + { + "epoch": 0.96, + "grad_norm": 0.466796875, + "learning_rate": 0.00010610712295531438, + "loss": 2.5263, + "step": 22606 + }, + { + "epoch": 0.96, + "grad_norm": 0.4921875, + "learning_rate": 0.00010610042753100305, + "loss": 2.6328, + "step": 22607 + }, + { + "epoch": 0.96, + "grad_norm": 0.4609375, + "learning_rate": 0.00010609373207924205, + "loss": 2.6299, + "step": 22608 + }, + { + "epoch": 0.96, + "grad_norm": 0.400390625, + "learning_rate": 0.0001060870366000615, + "loss": 2.3931, + "step": 22609 + }, + { + "epoch": 0.96, + "grad_norm": 0.453125, + "learning_rate": 0.00010608034109349149, + "loss": 2.5019, + "step": 22610 + }, + { + "epoch": 0.96, + "grad_norm": 0.431640625, + "learning_rate": 0.00010607364555956216, + "loss": 2.5092, + "step": 22611 + }, + { + "epoch": 0.96, + "grad_norm": 0.474609375, + "learning_rate": 0.00010606694999830367, + "loss": 2.6279, + "step": 22612 + }, + { + "epoch": 0.96, + "grad_norm": 0.46875, + "learning_rate": 0.00010606025440974608, + "loss": 2.6345, + "step": 22613 + }, + { + "epoch": 0.96, + "grad_norm": 0.453125, + "learning_rate": 0.0001060535587939196, + "loss": 2.5217, + "step": 22614 + }, + { + "epoch": 0.96, + "grad_norm": 0.447265625, + "learning_rate": 0.0001060468631508543, + "loss": 2.5782, + "step": 22615 + }, + { + "epoch": 0.96, + "grad_norm": 0.486328125, + "learning_rate": 0.00010604016748058032, + "loss": 2.5396, + "step": 22616 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010603347178312782, + "loss": 2.5994, + "step": 22617 + }, + { + "epoch": 0.96, + "grad_norm": 0.423828125, + "learning_rate": 0.0001060267760585269, + "loss": 2.4623, + "step": 22618 + }, + { + "epoch": 0.96, + "grad_norm": 0.5390625, + "learning_rate": 0.00010602008030680766, + "loss": 2.6198, + "step": 22619 + }, + { + "epoch": 0.96, + "grad_norm": 0.42578125, + "learning_rate": 0.00010601338452800029, + "loss": 2.5082, + "step": 22620 + }, + { + "epoch": 0.96, + "grad_norm": 0.4296875, + "learning_rate": 0.00010600668872213484, + "loss": 2.514, + "step": 22621 + }, + { + "epoch": 0.96, + "grad_norm": 0.486328125, + "learning_rate": 0.00010599999288924154, + "loss": 2.4262, + "step": 22622 + }, + { + "epoch": 0.96, + "grad_norm": 0.490234375, + "learning_rate": 0.00010599329702935043, + "loss": 2.6516, + "step": 22623 + }, + { + "epoch": 0.96, + "grad_norm": 0.45703125, + "learning_rate": 0.00010598660114249169, + "loss": 2.5271, + "step": 22624 + }, + { + "epoch": 0.96, + "grad_norm": 0.43359375, + "learning_rate": 0.00010597990522869542, + "loss": 2.5632, + "step": 22625 + }, + { + "epoch": 0.96, + "grad_norm": 0.40625, + "learning_rate": 0.00010597320928799179, + "loss": 2.5004, + "step": 22626 + }, + { + "epoch": 0.96, + "grad_norm": 0.435546875, + "learning_rate": 0.00010596651332041087, + "loss": 2.4701, + "step": 22627 + }, + { + "epoch": 0.96, + "grad_norm": 0.57421875, + "learning_rate": 0.00010595981732598283, + "loss": 2.6866, + "step": 22628 + }, + { + "epoch": 0.96, + "grad_norm": 0.48828125, + "learning_rate": 0.00010595312130473777, + "loss": 2.6175, + "step": 22629 + }, + { + "epoch": 0.96, + "grad_norm": 0.470703125, + "learning_rate": 0.00010594642525670591, + "loss": 2.4185, + "step": 22630 + }, + { + "epoch": 0.96, + "grad_norm": 0.4453125, + "learning_rate": 0.00010593972918191723, + "loss": 2.4438, + "step": 22631 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010593303308040199, + "loss": 2.4874, + "step": 22632 + }, + { + "epoch": 0.96, + "grad_norm": 0.494140625, + "learning_rate": 0.00010592633695219026, + "loss": 2.6694, + "step": 22633 + }, + { + "epoch": 0.96, + "grad_norm": 0.490234375, + "learning_rate": 0.00010591964079731216, + "loss": 2.6229, + "step": 22634 + }, + { + "epoch": 0.96, + "grad_norm": 0.455078125, + "learning_rate": 0.00010591294461579787, + "loss": 2.5936, + "step": 22635 + }, + { + "epoch": 0.96, + "grad_norm": 0.439453125, + "learning_rate": 0.00010590624840767744, + "loss": 2.5465, + "step": 22636 + }, + { + "epoch": 0.96, + "grad_norm": 0.50390625, + "learning_rate": 0.00010589955217298107, + "loss": 2.5909, + "step": 22637 + }, + { + "epoch": 0.96, + "grad_norm": 0.451171875, + "learning_rate": 0.00010589285591173891, + "loss": 2.6355, + "step": 22638 + }, + { + "epoch": 0.96, + "grad_norm": 0.458984375, + "learning_rate": 0.00010588615962398101, + "loss": 2.6082, + "step": 22639 + }, + { + "epoch": 0.96, + "grad_norm": 0.447265625, + "learning_rate": 0.00010587946330973755, + "loss": 2.5083, + "step": 22640 + }, + { + "epoch": 0.96, + "grad_norm": 0.435546875, + "learning_rate": 0.00010587276696903867, + "loss": 2.3762, + "step": 22641 + }, + { + "epoch": 0.96, + "grad_norm": 0.4453125, + "learning_rate": 0.00010586607060191447, + "loss": 2.4913, + "step": 22642 + }, + { + "epoch": 0.96, + "grad_norm": 0.46484375, + "learning_rate": 0.00010585937420839511, + "loss": 2.5294, + "step": 22643 + }, + { + "epoch": 0.96, + "grad_norm": 0.451171875, + "learning_rate": 0.00010585267778851072, + "loss": 2.4914, + "step": 22644 + }, + { + "epoch": 0.96, + "grad_norm": 0.42578125, + "learning_rate": 0.00010584598134229135, + "loss": 2.4798, + "step": 22645 + }, + { + "epoch": 0.97, + "grad_norm": 0.53125, + "learning_rate": 0.00010583928486976729, + "loss": 2.6783, + "step": 22646 + }, + { + "epoch": 0.97, + "grad_norm": 0.4375, + "learning_rate": 0.00010583258837096852, + "loss": 2.5511, + "step": 22647 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.00010582589184592526, + "loss": 2.4939, + "step": 22648 + }, + { + "epoch": 0.97, + "grad_norm": 0.4921875, + "learning_rate": 0.00010581919529466762, + "loss": 2.4067, + "step": 22649 + }, + { + "epoch": 0.97, + "grad_norm": 0.431640625, + "learning_rate": 0.00010581249871722572, + "loss": 2.5502, + "step": 22650 + }, + { + "epoch": 0.97, + "grad_norm": 0.44921875, + "learning_rate": 0.00010580580211362969, + "loss": 2.5255, + "step": 22651 + }, + { + "epoch": 0.97, + "grad_norm": 0.427734375, + "learning_rate": 0.00010579910548390967, + "loss": 2.4555, + "step": 22652 + }, + { + "epoch": 0.97, + "grad_norm": 0.58984375, + "learning_rate": 0.0001057924088280958, + "loss": 2.66, + "step": 22653 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.00010578571214621823, + "loss": 2.6069, + "step": 22654 + }, + { + "epoch": 0.97, + "grad_norm": 0.50390625, + "learning_rate": 0.00010577901543830704, + "loss": 2.4512, + "step": 22655 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.00010577231870439239, + "loss": 2.6053, + "step": 22656 + }, + { + "epoch": 0.97, + "grad_norm": 0.482421875, + "learning_rate": 0.00010576562194450446, + "loss": 2.7227, + "step": 22657 + }, + { + "epoch": 0.97, + "grad_norm": 0.443359375, + "learning_rate": 0.00010575892515867329, + "loss": 2.603, + "step": 22658 + }, + { + "epoch": 0.97, + "grad_norm": 0.439453125, + "learning_rate": 0.00010575222834692908, + "loss": 2.3895, + "step": 22659 + }, + { + "epoch": 0.97, + "grad_norm": 0.53125, + "learning_rate": 0.00010574553150930196, + "loss": 2.7854, + "step": 22660 + }, + { + "epoch": 0.97, + "grad_norm": 0.484375, + "learning_rate": 0.00010573883464582202, + "loss": 2.533, + "step": 22661 + }, + { + "epoch": 0.97, + "grad_norm": 0.412109375, + "learning_rate": 0.00010573213775651945, + "loss": 2.5717, + "step": 22662 + }, + { + "epoch": 0.97, + "grad_norm": 0.408203125, + "learning_rate": 0.0001057254408414243, + "loss": 2.5645, + "step": 22663 + }, + { + "epoch": 0.97, + "grad_norm": 0.408203125, + "learning_rate": 0.00010571874390056682, + "loss": 2.5337, + "step": 22664 + }, + { + "epoch": 0.97, + "grad_norm": 0.50390625, + "learning_rate": 0.00010571204693397704, + "loss": 2.6259, + "step": 22665 + }, + { + "epoch": 0.97, + "grad_norm": 0.482421875, + "learning_rate": 0.00010570534994168517, + "loss": 2.6147, + "step": 22666 + }, + { + "epoch": 0.97, + "grad_norm": 0.40234375, + "learning_rate": 0.00010569865292372129, + "loss": 2.4785, + "step": 22667 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.00010569195588011558, + "loss": 2.5224, + "step": 22668 + }, + { + "epoch": 0.97, + "grad_norm": 0.51171875, + "learning_rate": 0.00010568525881089813, + "loss": 2.4965, + "step": 22669 + }, + { + "epoch": 0.97, + "grad_norm": 0.48046875, + "learning_rate": 0.00010567856171609909, + "loss": 2.6549, + "step": 22670 + }, + { + "epoch": 0.97, + "grad_norm": 0.416015625, + "learning_rate": 0.0001056718645957486, + "loss": 2.3661, + "step": 22671 + }, + { + "epoch": 0.97, + "grad_norm": 0.48828125, + "learning_rate": 0.00010566516744987679, + "loss": 2.5338, + "step": 22672 + }, + { + "epoch": 0.97, + "grad_norm": 0.45703125, + "learning_rate": 0.00010565847027851381, + "loss": 2.4873, + "step": 22673 + }, + { + "epoch": 0.97, + "grad_norm": 0.458984375, + "learning_rate": 0.00010565177308168976, + "loss": 2.4633, + "step": 22674 + }, + { + "epoch": 0.97, + "grad_norm": 0.494140625, + "learning_rate": 0.00010564507585943482, + "loss": 2.473, + "step": 22675 + }, + { + "epoch": 0.97, + "grad_norm": 0.484375, + "learning_rate": 0.00010563837861177907, + "loss": 2.6241, + "step": 22676 + }, + { + "epoch": 0.97, + "grad_norm": 0.390625, + "learning_rate": 0.0001056316813387527, + "loss": 2.3258, + "step": 22677 + }, + { + "epoch": 0.97, + "grad_norm": 0.42578125, + "learning_rate": 0.00010562498404038583, + "loss": 2.5017, + "step": 22678 + }, + { + "epoch": 0.97, + "grad_norm": 0.396484375, + "learning_rate": 0.00010561828671670857, + "loss": 2.4796, + "step": 22679 + }, + { + "epoch": 0.97, + "grad_norm": 0.41796875, + "learning_rate": 0.00010561158936775106, + "loss": 2.4313, + "step": 22680 + }, + { + "epoch": 0.97, + "grad_norm": 0.48046875, + "learning_rate": 0.00010560489199354348, + "loss": 2.4723, + "step": 22681 + }, + { + "epoch": 0.97, + "grad_norm": 0.48046875, + "learning_rate": 0.00010559819459411593, + "loss": 2.5148, + "step": 22682 + }, + { + "epoch": 0.97, + "grad_norm": 0.4921875, + "learning_rate": 0.00010559149716949855, + "loss": 2.6358, + "step": 22683 + }, + { + "epoch": 0.97, + "grad_norm": 0.41796875, + "learning_rate": 0.00010558479971972146, + "loss": 2.6651, + "step": 22684 + }, + { + "epoch": 0.97, + "grad_norm": 0.46875, + "learning_rate": 0.00010557810224481481, + "loss": 2.4792, + "step": 22685 + }, + { + "epoch": 0.97, + "grad_norm": 0.478515625, + "learning_rate": 0.00010557140474480875, + "loss": 2.5617, + "step": 22686 + }, + { + "epoch": 0.97, + "grad_norm": 0.484375, + "learning_rate": 0.00010556470721973341, + "loss": 2.5411, + "step": 22687 + }, + { + "epoch": 0.97, + "grad_norm": 0.478515625, + "learning_rate": 0.0001055580096696189, + "loss": 2.7121, + "step": 22688 + }, + { + "epoch": 0.97, + "grad_norm": 0.439453125, + "learning_rate": 0.00010555131209449542, + "loss": 2.5758, + "step": 22689 + }, + { + "epoch": 0.97, + "grad_norm": 0.390625, + "learning_rate": 0.00010554461449439301, + "loss": 2.5216, + "step": 22690 + }, + { + "epoch": 0.97, + "grad_norm": 0.421875, + "learning_rate": 0.00010553791686934189, + "loss": 2.4584, + "step": 22691 + }, + { + "epoch": 0.97, + "grad_norm": 0.494140625, + "learning_rate": 0.00010553121921937216, + "loss": 2.5515, + "step": 22692 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.00010552452154451397, + "loss": 2.3411, + "step": 22693 + }, + { + "epoch": 0.97, + "grad_norm": 0.427734375, + "learning_rate": 0.00010551782384479748, + "loss": 2.7152, + "step": 22694 + }, + { + "epoch": 0.97, + "grad_norm": 0.46875, + "learning_rate": 0.00010551112612025275, + "loss": 2.6652, + "step": 22695 + }, + { + "epoch": 0.97, + "grad_norm": 0.48046875, + "learning_rate": 0.00010550442837090999, + "loss": 2.627, + "step": 22696 + }, + { + "epoch": 0.97, + "grad_norm": 0.39453125, + "learning_rate": 0.00010549773059679932, + "loss": 2.5378, + "step": 22697 + }, + { + "epoch": 0.97, + "grad_norm": 0.408203125, + "learning_rate": 0.00010549103279795085, + "loss": 2.4142, + "step": 22698 + }, + { + "epoch": 0.97, + "grad_norm": 0.484375, + "learning_rate": 0.00010548433497439476, + "loss": 2.5197, + "step": 22699 + }, + { + "epoch": 0.97, + "grad_norm": 0.451171875, + "learning_rate": 0.00010547763712616116, + "loss": 2.501, + "step": 22700 + }, + { + "epoch": 0.97, + "grad_norm": 0.47265625, + "learning_rate": 0.0001054709392532802, + "loss": 2.5063, + "step": 22701 + }, + { + "epoch": 0.97, + "grad_norm": 0.44921875, + "learning_rate": 0.00010546424135578202, + "loss": 2.4426, + "step": 22702 + }, + { + "epoch": 0.97, + "grad_norm": 0.447265625, + "learning_rate": 0.00010545754343369674, + "loss": 2.5251, + "step": 22703 + }, + { + "epoch": 0.97, + "grad_norm": 0.453125, + "learning_rate": 0.00010545084548705449, + "loss": 2.541, + "step": 22704 + }, + { + "epoch": 0.97, + "grad_norm": 0.451171875, + "learning_rate": 0.00010544414751588547, + "loss": 2.4783, + "step": 22705 + }, + { + "epoch": 0.97, + "grad_norm": 0.470703125, + "learning_rate": 0.00010543744952021973, + "loss": 2.4999, + "step": 22706 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.00010543075150008747, + "loss": 2.5678, + "step": 22707 + }, + { + "epoch": 0.97, + "grad_norm": 0.482421875, + "learning_rate": 0.00010542405345551882, + "loss": 2.645, + "step": 22708 + }, + { + "epoch": 0.97, + "grad_norm": 0.4609375, + "learning_rate": 0.00010541735538654392, + "loss": 2.5233, + "step": 22709 + }, + { + "epoch": 0.97, + "grad_norm": 0.4140625, + "learning_rate": 0.00010541065729319288, + "loss": 2.5677, + "step": 22710 + }, + { + "epoch": 0.97, + "grad_norm": 0.462890625, + "learning_rate": 0.00010540395917549589, + "loss": 2.4873, + "step": 22711 + }, + { + "epoch": 0.97, + "grad_norm": 0.443359375, + "learning_rate": 0.00010539726103348299, + "loss": 2.4172, + "step": 22712 + }, + { + "epoch": 0.97, + "grad_norm": 0.46484375, + "learning_rate": 0.00010539056286718447, + "loss": 2.5749, + "step": 22713 + }, + { + "epoch": 0.97, + "grad_norm": 0.435546875, + "learning_rate": 0.00010538386467663035, + "loss": 2.5956, + "step": 22714 + }, + { + "epoch": 0.97, + "grad_norm": 0.4765625, + "learning_rate": 0.0001053771664618508, + "loss": 2.5884, + "step": 22715 + }, + { + "epoch": 0.97, + "grad_norm": 0.4375, + "learning_rate": 0.00010537046822287597, + "loss": 2.4552, + "step": 22716 + }, + { + "epoch": 0.97, + "grad_norm": 0.435546875, + "learning_rate": 0.00010536376995973602, + "loss": 2.5521, + "step": 22717 + }, + { + "epoch": 0.97, + "grad_norm": 0.458984375, + "learning_rate": 0.00010535707167246104, + "loss": 2.6358, + "step": 22718 + }, + { + "epoch": 0.97, + "grad_norm": 0.384765625, + "learning_rate": 0.00010535037336108123, + "loss": 2.6202, + "step": 22719 + }, + { + "epoch": 0.97, + "grad_norm": 0.404296875, + "learning_rate": 0.00010534367502562663, + "loss": 2.5843, + "step": 22720 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.0001053369766661275, + "loss": 2.5315, + "step": 22721 + }, + { + "epoch": 0.97, + "grad_norm": 0.439453125, + "learning_rate": 0.00010533027828261391, + "loss": 2.3379, + "step": 22722 + }, + { + "epoch": 0.97, + "grad_norm": 0.490234375, + "learning_rate": 0.000105323579875116, + "loss": 2.6315, + "step": 22723 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.00010531688144366394, + "loss": 2.5548, + "step": 22724 + }, + { + "epoch": 0.97, + "grad_norm": 0.43359375, + "learning_rate": 0.00010531018298828787, + "loss": 2.5748, + "step": 22725 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.0001053034845090179, + "loss": 2.5943, + "step": 22726 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.0001052967860058842, + "loss": 2.3669, + "step": 22727 + }, + { + "epoch": 0.97, + "grad_norm": 0.39453125, + "learning_rate": 0.00010529008747891687, + "loss": 2.4975, + "step": 22728 + }, + { + "epoch": 0.97, + "grad_norm": 0.470703125, + "learning_rate": 0.00010528338892814613, + "loss": 2.4368, + "step": 22729 + }, + { + "epoch": 0.97, + "grad_norm": 0.494140625, + "learning_rate": 0.00010527669035360202, + "loss": 2.5106, + "step": 22730 + }, + { + "epoch": 0.97, + "grad_norm": 0.478515625, + "learning_rate": 0.00010526999175531475, + "loss": 2.5526, + "step": 22731 + }, + { + "epoch": 0.97, + "grad_norm": 0.4140625, + "learning_rate": 0.00010526329313331444, + "loss": 2.4107, + "step": 22732 + }, + { + "epoch": 0.97, + "grad_norm": 0.447265625, + "learning_rate": 0.00010525659448763125, + "loss": 2.6725, + "step": 22733 + }, + { + "epoch": 0.97, + "grad_norm": 0.478515625, + "learning_rate": 0.0001052498958182953, + "loss": 2.6463, + "step": 22734 + }, + { + "epoch": 0.97, + "grad_norm": 0.50390625, + "learning_rate": 0.00010524319712533675, + "loss": 2.6744, + "step": 22735 + }, + { + "epoch": 0.97, + "grad_norm": 0.453125, + "learning_rate": 0.00010523649840878567, + "loss": 2.4334, + "step": 22736 + }, + { + "epoch": 0.97, + "grad_norm": 0.46484375, + "learning_rate": 0.00010522979966867232, + "loss": 2.5204, + "step": 22737 + }, + { + "epoch": 0.97, + "grad_norm": 0.4921875, + "learning_rate": 0.00010522310090502676, + "loss": 2.6001, + "step": 22738 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.00010521640211787917, + "loss": 2.5431, + "step": 22739 + }, + { + "epoch": 0.97, + "grad_norm": 0.478515625, + "learning_rate": 0.00010520970330725965, + "loss": 2.4569, + "step": 22740 + }, + { + "epoch": 0.97, + "grad_norm": 0.486328125, + "learning_rate": 0.0001052030044731984, + "loss": 2.5539, + "step": 22741 + }, + { + "epoch": 0.97, + "grad_norm": 0.451171875, + "learning_rate": 0.00010519630561572553, + "loss": 2.5428, + "step": 22742 + }, + { + "epoch": 0.97, + "grad_norm": 0.4765625, + "learning_rate": 0.00010518960673487116, + "loss": 2.4768, + "step": 22743 + }, + { + "epoch": 0.97, + "grad_norm": 0.4296875, + "learning_rate": 0.00010518290783066547, + "loss": 2.6127, + "step": 22744 + }, + { + "epoch": 0.97, + "grad_norm": 0.5234375, + "learning_rate": 0.0001051762089031386, + "loss": 2.5914, + "step": 22745 + }, + { + "epoch": 0.97, + "grad_norm": 0.4609375, + "learning_rate": 0.00010516950995232064, + "loss": 2.512, + "step": 22746 + }, + { + "epoch": 0.97, + "grad_norm": 0.40625, + "learning_rate": 0.00010516281097824182, + "loss": 2.5129, + "step": 22747 + }, + { + "epoch": 0.97, + "grad_norm": 0.45703125, + "learning_rate": 0.00010515611198093223, + "loss": 2.572, + "step": 22748 + }, + { + "epoch": 0.97, + "grad_norm": 0.486328125, + "learning_rate": 0.000105149412960422, + "loss": 2.5819, + "step": 22749 + }, + { + "epoch": 0.97, + "grad_norm": 0.431640625, + "learning_rate": 0.00010514271391674131, + "loss": 2.4703, + "step": 22750 + }, + { + "epoch": 0.97, + "grad_norm": 0.39453125, + "learning_rate": 0.00010513601484992029, + "loss": 2.3936, + "step": 22751 + }, + { + "epoch": 0.97, + "grad_norm": 0.443359375, + "learning_rate": 0.00010512931575998905, + "loss": 2.6329, + "step": 22752 + }, + { + "epoch": 0.97, + "grad_norm": 0.421875, + "learning_rate": 0.0001051226166469778, + "loss": 2.5243, + "step": 22753 + }, + { + "epoch": 0.97, + "grad_norm": 0.384765625, + "learning_rate": 0.00010511591751091663, + "loss": 2.3707, + "step": 22754 + }, + { + "epoch": 0.97, + "grad_norm": 0.5703125, + "learning_rate": 0.00010510921835183569, + "loss": 2.6135, + "step": 22755 + }, + { + "epoch": 0.97, + "grad_norm": 0.421875, + "learning_rate": 0.00010510251916976515, + "loss": 2.5167, + "step": 22756 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.00010509581996473514, + "loss": 2.495, + "step": 22757 + }, + { + "epoch": 0.97, + "grad_norm": 0.48046875, + "learning_rate": 0.00010508912073677581, + "loss": 2.5201, + "step": 22758 + }, + { + "epoch": 0.97, + "grad_norm": 0.478515625, + "learning_rate": 0.00010508242148591729, + "loss": 2.5399, + "step": 22759 + }, + { + "epoch": 0.97, + "grad_norm": 0.443359375, + "learning_rate": 0.0001050757222121897, + "loss": 2.6132, + "step": 22760 + }, + { + "epoch": 0.97, + "grad_norm": 0.50390625, + "learning_rate": 0.00010506902291562326, + "loss": 2.4449, + "step": 22761 + }, + { + "epoch": 0.97, + "grad_norm": 0.427734375, + "learning_rate": 0.00010506232359624804, + "loss": 2.413, + "step": 22762 + }, + { + "epoch": 0.97, + "grad_norm": 0.435546875, + "learning_rate": 0.00010505562425409423, + "loss": 2.4671, + "step": 22763 + }, + { + "epoch": 0.97, + "grad_norm": 0.41015625, + "learning_rate": 0.00010504892488919195, + "loss": 2.4169, + "step": 22764 + }, + { + "epoch": 0.97, + "grad_norm": 0.44921875, + "learning_rate": 0.00010504222550157137, + "loss": 2.4642, + "step": 22765 + }, + { + "epoch": 0.97, + "grad_norm": 0.458984375, + "learning_rate": 0.0001050355260912626, + "loss": 2.699, + "step": 22766 + }, + { + "epoch": 0.97, + "grad_norm": 0.431640625, + "learning_rate": 0.00010502882665829581, + "loss": 2.4786, + "step": 22767 + }, + { + "epoch": 0.97, + "grad_norm": 0.47265625, + "learning_rate": 0.0001050221272027011, + "loss": 2.6316, + "step": 22768 + }, + { + "epoch": 0.97, + "grad_norm": 0.47265625, + "learning_rate": 0.00010501542772450872, + "loss": 2.5992, + "step": 22769 + }, + { + "epoch": 0.97, + "grad_norm": 0.439453125, + "learning_rate": 0.0001050087282237487, + "loss": 2.364, + "step": 22770 + }, + { + "epoch": 0.97, + "grad_norm": 0.4140625, + "learning_rate": 0.00010500202870045126, + "loss": 2.5509, + "step": 22771 + }, + { + "epoch": 0.97, + "grad_norm": 0.4765625, + "learning_rate": 0.0001049953291546465, + "loss": 2.5826, + "step": 22772 + }, + { + "epoch": 0.97, + "grad_norm": 0.4140625, + "learning_rate": 0.0001049886295863646, + "loss": 2.3904, + "step": 22773 + }, + { + "epoch": 0.97, + "grad_norm": 0.458984375, + "learning_rate": 0.00010498192999563568, + "loss": 2.5198, + "step": 22774 + }, + { + "epoch": 0.97, + "grad_norm": 0.4765625, + "learning_rate": 0.0001049752303824899, + "loss": 2.5746, + "step": 22775 + }, + { + "epoch": 0.97, + "grad_norm": 0.48046875, + "learning_rate": 0.00010496853074695739, + "loss": 2.6311, + "step": 22776 + }, + { + "epoch": 0.97, + "grad_norm": 0.4765625, + "learning_rate": 0.00010496183108906835, + "loss": 2.6368, + "step": 22777 + }, + { + "epoch": 0.97, + "grad_norm": 0.43359375, + "learning_rate": 0.00010495513140885282, + "loss": 2.4213, + "step": 22778 + }, + { + "epoch": 0.97, + "grad_norm": 0.453125, + "learning_rate": 0.00010494843170634103, + "loss": 2.5582, + "step": 22779 + }, + { + "epoch": 0.97, + "grad_norm": 0.44921875, + "learning_rate": 0.00010494173198156314, + "loss": 2.4813, + "step": 22780 + }, + { + "epoch": 0.97, + "grad_norm": 0.4453125, + "learning_rate": 0.00010493503223454924, + "loss": 2.5873, + "step": 22781 + }, + { + "epoch": 0.97, + "grad_norm": 0.482421875, + "learning_rate": 0.00010492833246532949, + "loss": 2.5609, + "step": 22782 + }, + { + "epoch": 0.97, + "grad_norm": 0.515625, + "learning_rate": 0.00010492163267393404, + "loss": 2.4909, + "step": 22783 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.00010491493286039305, + "loss": 2.5983, + "step": 22784 + }, + { + "epoch": 0.97, + "grad_norm": 0.462890625, + "learning_rate": 0.00010490823302473669, + "loss": 2.4292, + "step": 22785 + }, + { + "epoch": 0.97, + "grad_norm": 0.451171875, + "learning_rate": 0.00010490153316699503, + "loss": 2.7311, + "step": 22786 + }, + { + "epoch": 0.97, + "grad_norm": 0.443359375, + "learning_rate": 0.00010489483328719826, + "loss": 2.5722, + "step": 22787 + }, + { + "epoch": 0.97, + "grad_norm": 0.447265625, + "learning_rate": 0.00010488813338537656, + "loss": 2.6357, + "step": 22788 + }, + { + "epoch": 0.97, + "grad_norm": 0.4609375, + "learning_rate": 0.00010488143346156004, + "loss": 2.6802, + "step": 22789 + }, + { + "epoch": 0.97, + "grad_norm": 0.412109375, + "learning_rate": 0.00010487473351577885, + "loss": 2.6095, + "step": 22790 + }, + { + "epoch": 0.97, + "grad_norm": 0.45703125, + "learning_rate": 0.00010486803354806314, + "loss": 2.6012, + "step": 22791 + }, + { + "epoch": 0.97, + "grad_norm": 0.43359375, + "learning_rate": 0.00010486133355844306, + "loss": 2.5105, + "step": 22792 + }, + { + "epoch": 0.97, + "grad_norm": 0.47265625, + "learning_rate": 0.00010485463354694876, + "loss": 2.5812, + "step": 22793 + }, + { + "epoch": 0.97, + "grad_norm": 0.50390625, + "learning_rate": 0.00010484793351361037, + "loss": 2.6569, + "step": 22794 + }, + { + "epoch": 0.97, + "grad_norm": 0.42578125, + "learning_rate": 0.00010484123345845805, + "loss": 2.4476, + "step": 22795 + }, + { + "epoch": 0.97, + "grad_norm": 0.396484375, + "learning_rate": 0.00010483453338152195, + "loss": 2.4642, + "step": 22796 + }, + { + "epoch": 0.97, + "grad_norm": 0.447265625, + "learning_rate": 0.00010482783328283221, + "loss": 2.5444, + "step": 22797 + }, + { + "epoch": 0.97, + "grad_norm": 0.43359375, + "learning_rate": 0.00010482113316241902, + "loss": 2.4165, + "step": 22798 + }, + { + "epoch": 0.97, + "grad_norm": 0.54296875, + "learning_rate": 0.00010481443302031247, + "loss": 2.5839, + "step": 22799 + }, + { + "epoch": 0.97, + "grad_norm": 0.494140625, + "learning_rate": 0.00010480773285654273, + "loss": 2.5832, + "step": 22800 + }, + { + "epoch": 0.97, + "grad_norm": 0.462890625, + "learning_rate": 0.00010480103267113994, + "loss": 2.6406, + "step": 22801 + }, + { + "epoch": 0.97, + "grad_norm": 0.49609375, + "learning_rate": 0.00010479433246413428, + "loss": 2.551, + "step": 22802 + }, + { + "epoch": 0.97, + "grad_norm": 0.4375, + "learning_rate": 0.00010478763223555583, + "loss": 2.5679, + "step": 22803 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.00010478093198543484, + "loss": 2.5699, + "step": 22804 + }, + { + "epoch": 0.97, + "grad_norm": 0.4765625, + "learning_rate": 0.00010477423171380137, + "loss": 2.5639, + "step": 22805 + }, + { + "epoch": 0.97, + "grad_norm": 0.470703125, + "learning_rate": 0.00010476753142068561, + "loss": 2.5903, + "step": 22806 + }, + { + "epoch": 0.97, + "grad_norm": 0.482421875, + "learning_rate": 0.00010476083110611769, + "loss": 2.6155, + "step": 22807 + }, + { + "epoch": 0.97, + "grad_norm": 0.49609375, + "learning_rate": 0.0001047541307701278, + "loss": 2.6143, + "step": 22808 + }, + { + "epoch": 0.97, + "grad_norm": 0.478515625, + "learning_rate": 0.00010474743041274604, + "loss": 2.4874, + "step": 22809 + }, + { + "epoch": 0.97, + "grad_norm": 0.4765625, + "learning_rate": 0.0001047407300340026, + "loss": 2.5527, + "step": 22810 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.00010473402963392756, + "loss": 2.6088, + "step": 22811 + }, + { + "epoch": 0.97, + "grad_norm": 0.42578125, + "learning_rate": 0.00010472732921255117, + "loss": 2.4872, + "step": 22812 + }, + { + "epoch": 0.97, + "grad_norm": 0.458984375, + "learning_rate": 0.00010472062876990347, + "loss": 2.5573, + "step": 22813 + }, + { + "epoch": 0.97, + "grad_norm": 0.4296875, + "learning_rate": 0.0001047139283060147, + "loss": 2.6027, + "step": 22814 + }, + { + "epoch": 0.97, + "grad_norm": 0.484375, + "learning_rate": 0.00010470722782091498, + "loss": 2.6137, + "step": 22815 + }, + { + "epoch": 0.97, + "grad_norm": 0.486328125, + "learning_rate": 0.00010470052731463445, + "loss": 2.4246, + "step": 22816 + }, + { + "epoch": 0.97, + "grad_norm": 0.41796875, + "learning_rate": 0.00010469382678720326, + "loss": 2.5957, + "step": 22817 + }, + { + "epoch": 0.97, + "grad_norm": 0.388671875, + "learning_rate": 0.00010468712623865158, + "loss": 2.5434, + "step": 22818 + }, + { + "epoch": 0.97, + "grad_norm": 0.46875, + "learning_rate": 0.0001046804256690095, + "loss": 2.5772, + "step": 22819 + }, + { + "epoch": 0.97, + "grad_norm": 0.3984375, + "learning_rate": 0.00010467372507830728, + "loss": 2.5039, + "step": 22820 + }, + { + "epoch": 0.97, + "grad_norm": 0.43359375, + "learning_rate": 0.00010466702446657494, + "loss": 2.5648, + "step": 22821 + }, + { + "epoch": 0.97, + "grad_norm": 0.43359375, + "learning_rate": 0.00010466032383384275, + "loss": 2.5091, + "step": 22822 + }, + { + "epoch": 0.97, + "grad_norm": 0.48828125, + "learning_rate": 0.00010465362318014078, + "loss": 2.5214, + "step": 22823 + }, + { + "epoch": 0.97, + "grad_norm": 0.431640625, + "learning_rate": 0.00010464692250549921, + "loss": 2.5974, + "step": 22824 + }, + { + "epoch": 0.97, + "grad_norm": 0.474609375, + "learning_rate": 0.00010464022180994819, + "loss": 2.444, + "step": 22825 + }, + { + "epoch": 0.97, + "grad_norm": 0.46484375, + "learning_rate": 0.00010463352109351785, + "loss": 2.5324, + "step": 22826 + }, + { + "epoch": 0.97, + "grad_norm": 0.421875, + "learning_rate": 0.00010462682035623837, + "loss": 2.5042, + "step": 22827 + }, + { + "epoch": 0.97, + "grad_norm": 0.458984375, + "learning_rate": 0.00010462011959813992, + "loss": 2.5514, + "step": 22828 + }, + { + "epoch": 0.97, + "grad_norm": 0.4453125, + "learning_rate": 0.00010461341881925255, + "loss": 2.5951, + "step": 22829 + }, + { + "epoch": 0.97, + "grad_norm": 0.4609375, + "learning_rate": 0.00010460671801960653, + "loss": 2.6554, + "step": 22830 + }, + { + "epoch": 0.97, + "grad_norm": 0.458984375, + "learning_rate": 0.00010460001719923196, + "loss": 2.6167, + "step": 22831 + }, + { + "epoch": 0.97, + "grad_norm": 0.46875, + "learning_rate": 0.00010459331635815898, + "loss": 2.6218, + "step": 22832 + }, + { + "epoch": 0.97, + "grad_norm": 0.443359375, + "learning_rate": 0.00010458661549641777, + "loss": 2.5502, + "step": 22833 + }, + { + "epoch": 0.97, + "grad_norm": 0.447265625, + "learning_rate": 0.00010457991461403849, + "loss": 2.6263, + "step": 22834 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.0001045732137110512, + "loss": 2.56, + "step": 22835 + }, + { + "epoch": 0.97, + "grad_norm": 0.458984375, + "learning_rate": 0.0001045665127874862, + "loss": 2.6005, + "step": 22836 + }, + { + "epoch": 0.97, + "grad_norm": 0.51953125, + "learning_rate": 0.00010455981184337349, + "loss": 2.6816, + "step": 22837 + }, + { + "epoch": 0.97, + "grad_norm": 0.47265625, + "learning_rate": 0.00010455311087874333, + "loss": 2.5344, + "step": 22838 + }, + { + "epoch": 0.97, + "grad_norm": 0.474609375, + "learning_rate": 0.00010454640989362583, + "loss": 2.6062, + "step": 22839 + }, + { + "epoch": 0.97, + "grad_norm": 0.4921875, + "learning_rate": 0.00010453970888805113, + "loss": 2.5996, + "step": 22840 + }, + { + "epoch": 0.97, + "grad_norm": 0.443359375, + "learning_rate": 0.00010453300786204943, + "loss": 2.4655, + "step": 22841 + }, + { + "epoch": 0.97, + "grad_norm": 0.400390625, + "learning_rate": 0.00010452630681565083, + "loss": 2.5414, + "step": 22842 + }, + { + "epoch": 0.97, + "grad_norm": 0.5, + "learning_rate": 0.0001045196057488855, + "loss": 2.487, + "step": 22843 + }, + { + "epoch": 0.97, + "grad_norm": 0.494140625, + "learning_rate": 0.00010451290466178363, + "loss": 2.4863, + "step": 22844 + }, + { + "epoch": 0.97, + "grad_norm": 0.439453125, + "learning_rate": 0.00010450620355437528, + "loss": 2.4293, + "step": 22845 + }, + { + "epoch": 0.97, + "grad_norm": 0.53125, + "learning_rate": 0.00010449950242669071, + "loss": 2.61, + "step": 22846 + }, + { + "epoch": 0.97, + "grad_norm": 0.5078125, + "learning_rate": 0.00010449280127876, + "loss": 2.6333, + "step": 22847 + }, + { + "epoch": 0.97, + "grad_norm": 0.435546875, + "learning_rate": 0.00010448610011061335, + "loss": 2.4374, + "step": 22848 + }, + { + "epoch": 0.97, + "grad_norm": 0.404296875, + "learning_rate": 0.00010447939892228087, + "loss": 2.5752, + "step": 22849 + }, + { + "epoch": 0.97, + "grad_norm": 0.404296875, + "learning_rate": 0.00010447269771379273, + "loss": 2.4035, + "step": 22850 + }, + { + "epoch": 0.97, + "grad_norm": 0.390625, + "learning_rate": 0.00010446599648517908, + "loss": 2.4444, + "step": 22851 + }, + { + "epoch": 0.97, + "grad_norm": 0.404296875, + "learning_rate": 0.00010445929523647012, + "loss": 2.467, + "step": 22852 + }, + { + "epoch": 0.97, + "grad_norm": 0.45703125, + "learning_rate": 0.00010445259396769589, + "loss": 2.5957, + "step": 22853 + }, + { + "epoch": 0.97, + "grad_norm": 0.46875, + "learning_rate": 0.00010444589267888667, + "loss": 2.6697, + "step": 22854 + }, + { + "epoch": 0.97, + "grad_norm": 0.498046875, + "learning_rate": 0.00010443919137007255, + "loss": 2.541, + "step": 22855 + }, + { + "epoch": 0.97, + "grad_norm": 0.4375, + "learning_rate": 0.00010443249004128365, + "loss": 2.5146, + "step": 22856 + }, + { + "epoch": 0.97, + "grad_norm": 0.51171875, + "learning_rate": 0.00010442578869255022, + "loss": 2.4461, + "step": 22857 + }, + { + "epoch": 0.97, + "grad_norm": 0.42578125, + "learning_rate": 0.00010441908732390233, + "loss": 2.4446, + "step": 22858 + }, + { + "epoch": 0.97, + "grad_norm": 0.47265625, + "learning_rate": 0.00010441238593537014, + "loss": 2.6568, + "step": 22859 + }, + { + "epoch": 0.97, + "grad_norm": 0.482421875, + "learning_rate": 0.0001044056845269839, + "loss": 2.5018, + "step": 22860 + }, + { + "epoch": 0.97, + "grad_norm": 0.44140625, + "learning_rate": 0.00010439898309877361, + "loss": 2.6791, + "step": 22861 + }, + { + "epoch": 0.97, + "grad_norm": 0.47265625, + "learning_rate": 0.00010439228165076953, + "loss": 2.5545, + "step": 22862 + }, + { + "epoch": 0.97, + "grad_norm": 0.451171875, + "learning_rate": 0.00010438558018300179, + "loss": 2.4294, + "step": 22863 + }, + { + "epoch": 0.97, + "grad_norm": 0.388671875, + "learning_rate": 0.00010437887869550055, + "loss": 2.6801, + "step": 22864 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.00010437217718829597, + "loss": 2.5242, + "step": 22865 + }, + { + "epoch": 0.97, + "grad_norm": 0.4296875, + "learning_rate": 0.00010436547566141817, + "loss": 2.5245, + "step": 22866 + }, + { + "epoch": 0.97, + "grad_norm": 0.48046875, + "learning_rate": 0.00010435877411489731, + "loss": 2.4657, + "step": 22867 + }, + { + "epoch": 0.97, + "grad_norm": 0.466796875, + "learning_rate": 0.00010435207254876361, + "loss": 2.4323, + "step": 22868 + }, + { + "epoch": 0.97, + "grad_norm": 0.41796875, + "learning_rate": 0.00010434537096304711, + "loss": 2.3983, + "step": 22869 + }, + { + "epoch": 0.97, + "grad_norm": 0.439453125, + "learning_rate": 0.00010433866935777807, + "loss": 2.577, + "step": 22870 + }, + { + "epoch": 0.97, + "grad_norm": 0.443359375, + "learning_rate": 0.00010433196773298659, + "loss": 2.5764, + "step": 22871 + }, + { + "epoch": 0.97, + "grad_norm": 0.47265625, + "learning_rate": 0.00010432526608870286, + "loss": 2.6726, + "step": 22872 + }, + { + "epoch": 0.97, + "grad_norm": 0.54296875, + "learning_rate": 0.00010431856442495698, + "loss": 2.6195, + "step": 22873 + }, + { + "epoch": 0.97, + "grad_norm": 0.4375, + "learning_rate": 0.00010431186274177915, + "loss": 2.4818, + "step": 22874 + }, + { + "epoch": 0.97, + "grad_norm": 0.9296875, + "learning_rate": 0.0001043051610391995, + "loss": 2.6622, + "step": 22875 + }, + { + "epoch": 0.97, + "grad_norm": 0.435546875, + "learning_rate": 0.00010429845931724824, + "loss": 2.5571, + "step": 22876 + }, + { + "epoch": 0.97, + "grad_norm": 0.396484375, + "learning_rate": 0.00010429175757595544, + "loss": 2.4951, + "step": 22877 + }, + { + "epoch": 0.97, + "grad_norm": 0.43359375, + "learning_rate": 0.00010428505581535133, + "loss": 2.4431, + "step": 22878 + }, + { + "epoch": 0.97, + "grad_norm": 0.44921875, + "learning_rate": 0.00010427835403546602, + "loss": 2.5048, + "step": 22879 + }, + { + "epoch": 0.97, + "grad_norm": 0.46484375, + "learning_rate": 0.00010427165223632969, + "loss": 2.6523, + "step": 22880 + }, + { + "epoch": 0.98, + "grad_norm": 0.466796875, + "learning_rate": 0.00010426495041797249, + "loss": 2.5218, + "step": 22881 + }, + { + "epoch": 0.98, + "grad_norm": 0.462890625, + "learning_rate": 0.00010425824858042455, + "loss": 2.5183, + "step": 22882 + }, + { + "epoch": 0.98, + "grad_norm": 0.421875, + "learning_rate": 0.00010425154672371604, + "loss": 2.4917, + "step": 22883 + }, + { + "epoch": 0.98, + "grad_norm": 0.42578125, + "learning_rate": 0.00010424484484787718, + "loss": 2.5799, + "step": 22884 + }, + { + "epoch": 0.98, + "grad_norm": 0.421875, + "learning_rate": 0.00010423814295293801, + "loss": 2.4866, + "step": 22885 + }, + { + "epoch": 0.98, + "grad_norm": 0.470703125, + "learning_rate": 0.00010423144103892877, + "loss": 2.4365, + "step": 22886 + }, + { + "epoch": 0.98, + "grad_norm": 0.56640625, + "learning_rate": 0.00010422473910587958, + "loss": 2.5884, + "step": 22887 + }, + { + "epoch": 0.98, + "grad_norm": 0.470703125, + "learning_rate": 0.00010421803715382061, + "loss": 2.5518, + "step": 22888 + }, + { + "epoch": 0.98, + "grad_norm": 0.4453125, + "learning_rate": 0.00010421133518278204, + "loss": 2.5052, + "step": 22889 + }, + { + "epoch": 0.98, + "grad_norm": 0.44921875, + "learning_rate": 0.000104204633192794, + "loss": 2.4509, + "step": 22890 + }, + { + "epoch": 0.98, + "grad_norm": 0.451171875, + "learning_rate": 0.00010419793118388658, + "loss": 2.6492, + "step": 22891 + }, + { + "epoch": 0.98, + "grad_norm": 0.470703125, + "learning_rate": 0.00010419122915609009, + "loss": 2.5539, + "step": 22892 + }, + { + "epoch": 0.98, + "grad_norm": 0.484375, + "learning_rate": 0.00010418452710943454, + "loss": 2.547, + "step": 22893 + }, + { + "epoch": 0.98, + "grad_norm": 0.466796875, + "learning_rate": 0.00010417782504395017, + "loss": 2.6109, + "step": 22894 + }, + { + "epoch": 0.98, + "grad_norm": 0.482421875, + "learning_rate": 0.00010417112295966714, + "loss": 2.5127, + "step": 22895 + }, + { + "epoch": 0.98, + "grad_norm": 0.435546875, + "learning_rate": 0.00010416442085661555, + "loss": 2.5729, + "step": 22896 + }, + { + "epoch": 0.98, + "grad_norm": 0.453125, + "learning_rate": 0.0001041577187348256, + "loss": 2.5123, + "step": 22897 + }, + { + "epoch": 0.98, + "grad_norm": 0.466796875, + "learning_rate": 0.00010415101659432742, + "loss": 2.5879, + "step": 22898 + }, + { + "epoch": 0.98, + "grad_norm": 0.4140625, + "learning_rate": 0.00010414431443515119, + "loss": 2.5146, + "step": 22899 + }, + { + "epoch": 0.98, + "grad_norm": 0.46484375, + "learning_rate": 0.00010413761225732708, + "loss": 2.5159, + "step": 22900 + }, + { + "epoch": 0.98, + "grad_norm": 0.4609375, + "learning_rate": 0.00010413091006088518, + "loss": 2.6118, + "step": 22901 + }, + { + "epoch": 0.98, + "grad_norm": 0.482421875, + "learning_rate": 0.00010412420784585573, + "loss": 2.5835, + "step": 22902 + }, + { + "epoch": 0.98, + "grad_norm": 0.482421875, + "learning_rate": 0.00010411750561226886, + "loss": 2.6613, + "step": 22903 + }, + { + "epoch": 0.98, + "grad_norm": 0.44140625, + "learning_rate": 0.00010411080336015469, + "loss": 2.5611, + "step": 22904 + }, + { + "epoch": 0.98, + "grad_norm": 0.40625, + "learning_rate": 0.00010410410108954342, + "loss": 2.4862, + "step": 22905 + }, + { + "epoch": 0.98, + "grad_norm": 0.4453125, + "learning_rate": 0.0001040973988004652, + "loss": 2.4376, + "step": 22906 + }, + { + "epoch": 0.98, + "grad_norm": 0.4453125, + "learning_rate": 0.00010409069649295018, + "loss": 2.4885, + "step": 22907 + }, + { + "epoch": 0.98, + "grad_norm": 0.4375, + "learning_rate": 0.00010408399416702854, + "loss": 2.4572, + "step": 22908 + }, + { + "epoch": 0.98, + "grad_norm": 0.462890625, + "learning_rate": 0.00010407729182273038, + "loss": 2.6434, + "step": 22909 + }, + { + "epoch": 0.98, + "grad_norm": 0.455078125, + "learning_rate": 0.00010407058946008592, + "loss": 2.5268, + "step": 22910 + }, + { + "epoch": 0.98, + "grad_norm": 0.4609375, + "learning_rate": 0.00010406388707912531, + "loss": 2.4815, + "step": 22911 + }, + { + "epoch": 0.98, + "grad_norm": 0.486328125, + "learning_rate": 0.00010405718467987866, + "loss": 2.5016, + "step": 22912 + }, + { + "epoch": 0.98, + "grad_norm": 0.46484375, + "learning_rate": 0.00010405048226237617, + "loss": 2.5579, + "step": 22913 + }, + { + "epoch": 0.98, + "grad_norm": 0.416015625, + "learning_rate": 0.00010404377982664799, + "loss": 2.542, + "step": 22914 + }, + { + "epoch": 0.98, + "grad_norm": 0.46875, + "learning_rate": 0.0001040370773727243, + "loss": 2.6077, + "step": 22915 + }, + { + "epoch": 0.98, + "grad_norm": 0.47265625, + "learning_rate": 0.00010403037490063523, + "loss": 2.5236, + "step": 22916 + }, + { + "epoch": 0.98, + "grad_norm": 0.451171875, + "learning_rate": 0.00010402367241041094, + "loss": 2.5247, + "step": 22917 + }, + { + "epoch": 0.98, + "grad_norm": 0.421875, + "learning_rate": 0.00010401696990208156, + "loss": 2.5431, + "step": 22918 + }, + { + "epoch": 0.98, + "grad_norm": 0.423828125, + "learning_rate": 0.00010401026737567735, + "loss": 2.5945, + "step": 22919 + }, + { + "epoch": 0.98, + "grad_norm": 0.50390625, + "learning_rate": 0.00010400356483122835, + "loss": 2.5537, + "step": 22920 + }, + { + "epoch": 0.98, + "grad_norm": 0.43359375, + "learning_rate": 0.0001039968622687648, + "loss": 2.6179, + "step": 22921 + }, + { + "epoch": 0.98, + "grad_norm": 0.4296875, + "learning_rate": 0.0001039901596883168, + "loss": 2.5709, + "step": 22922 + }, + { + "epoch": 0.98, + "grad_norm": 0.48046875, + "learning_rate": 0.00010398345708991457, + "loss": 2.6773, + "step": 22923 + }, + { + "epoch": 0.98, + "grad_norm": 0.50390625, + "learning_rate": 0.00010397675447358823, + "loss": 2.4562, + "step": 22924 + }, + { + "epoch": 0.98, + "grad_norm": 0.46484375, + "learning_rate": 0.00010397005183936794, + "loss": 2.6128, + "step": 22925 + }, + { + "epoch": 0.98, + "grad_norm": 0.44140625, + "learning_rate": 0.00010396334918728387, + "loss": 2.4693, + "step": 22926 + }, + { + "epoch": 0.98, + "grad_norm": 0.435546875, + "learning_rate": 0.0001039566465173662, + "loss": 2.5474, + "step": 22927 + }, + { + "epoch": 0.98, + "grad_norm": 0.4375, + "learning_rate": 0.00010394994382964503, + "loss": 2.5714, + "step": 22928 + }, + { + "epoch": 0.98, + "grad_norm": 0.4609375, + "learning_rate": 0.00010394324112415057, + "loss": 2.4814, + "step": 22929 + }, + { + "epoch": 0.98, + "grad_norm": 0.486328125, + "learning_rate": 0.00010393653840091298, + "loss": 2.6411, + "step": 22930 + }, + { + "epoch": 0.98, + "grad_norm": 0.453125, + "learning_rate": 0.0001039298356599624, + "loss": 2.5708, + "step": 22931 + }, + { + "epoch": 0.98, + "grad_norm": 0.423828125, + "learning_rate": 0.000103923132901329, + "loss": 2.5781, + "step": 22932 + }, + { + "epoch": 0.98, + "grad_norm": 0.49609375, + "learning_rate": 0.00010391643012504294, + "loss": 2.6894, + "step": 22933 + }, + { + "epoch": 0.98, + "grad_norm": 0.46875, + "learning_rate": 0.00010390972733113434, + "loss": 2.5489, + "step": 22934 + }, + { + "epoch": 0.98, + "grad_norm": 0.423828125, + "learning_rate": 0.00010390302451963343, + "loss": 2.5165, + "step": 22935 + }, + { + "epoch": 0.98, + "grad_norm": 0.484375, + "learning_rate": 0.00010389632169057031, + "loss": 2.5262, + "step": 22936 + }, + { + "epoch": 0.98, + "grad_norm": 0.51953125, + "learning_rate": 0.00010388961884397519, + "loss": 2.5345, + "step": 22937 + }, + { + "epoch": 0.98, + "grad_norm": 0.4140625, + "learning_rate": 0.0001038829159798782, + "loss": 2.5221, + "step": 22938 + }, + { + "epoch": 0.98, + "grad_norm": 0.466796875, + "learning_rate": 0.0001038762130983095, + "loss": 2.5543, + "step": 22939 + }, + { + "epoch": 0.98, + "grad_norm": 0.515625, + "learning_rate": 0.00010386951019929926, + "loss": 2.6502, + "step": 22940 + }, + { + "epoch": 0.98, + "grad_norm": 0.466796875, + "learning_rate": 0.00010386280728287765, + "loss": 2.5653, + "step": 22941 + }, + { + "epoch": 0.98, + "grad_norm": 0.44921875, + "learning_rate": 0.0001038561043490748, + "loss": 2.6738, + "step": 22942 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.00010384940139792093, + "loss": 2.4544, + "step": 22943 + }, + { + "epoch": 0.98, + "grad_norm": 0.44921875, + "learning_rate": 0.0001038426984294461, + "loss": 2.347, + "step": 22944 + }, + { + "epoch": 0.98, + "grad_norm": 0.462890625, + "learning_rate": 0.00010383599544368057, + "loss": 2.407, + "step": 22945 + }, + { + "epoch": 0.98, + "grad_norm": 0.474609375, + "learning_rate": 0.00010382929244065444, + "loss": 2.7029, + "step": 22946 + }, + { + "epoch": 0.98, + "grad_norm": 0.47265625, + "learning_rate": 0.00010382258942039793, + "loss": 2.5162, + "step": 22947 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.00010381588638294113, + "loss": 2.4304, + "step": 22948 + }, + { + "epoch": 0.98, + "grad_norm": 0.466796875, + "learning_rate": 0.00010380918332831426, + "loss": 2.5013, + "step": 22949 + }, + { + "epoch": 0.98, + "grad_norm": 0.431640625, + "learning_rate": 0.00010380248025654741, + "loss": 2.4866, + "step": 22950 + }, + { + "epoch": 0.98, + "grad_norm": 0.486328125, + "learning_rate": 0.00010379577716767087, + "loss": 2.569, + "step": 22951 + }, + { + "epoch": 0.98, + "grad_norm": 0.47265625, + "learning_rate": 0.00010378907406171465, + "loss": 2.6847, + "step": 22952 + }, + { + "epoch": 0.98, + "grad_norm": 0.392578125, + "learning_rate": 0.00010378237093870902, + "loss": 2.478, + "step": 22953 + }, + { + "epoch": 0.98, + "grad_norm": 0.462890625, + "learning_rate": 0.00010377566779868408, + "loss": 2.5424, + "step": 22954 + }, + { + "epoch": 0.98, + "grad_norm": 0.44921875, + "learning_rate": 0.00010376896464167004, + "loss": 2.4236, + "step": 22955 + }, + { + "epoch": 0.98, + "grad_norm": 0.474609375, + "learning_rate": 0.00010376226146769701, + "loss": 2.5726, + "step": 22956 + }, + { + "epoch": 0.98, + "grad_norm": 0.455078125, + "learning_rate": 0.0001037555582767952, + "loss": 2.4694, + "step": 22957 + }, + { + "epoch": 0.98, + "grad_norm": 0.4375, + "learning_rate": 0.0001037488550689947, + "loss": 2.5947, + "step": 22958 + }, + { + "epoch": 0.98, + "grad_norm": 0.466796875, + "learning_rate": 0.00010374215184432578, + "loss": 2.717, + "step": 22959 + }, + { + "epoch": 0.98, + "grad_norm": 0.515625, + "learning_rate": 0.00010373544860281852, + "loss": 2.5425, + "step": 22960 + }, + { + "epoch": 0.98, + "grad_norm": 0.515625, + "learning_rate": 0.00010372874534450312, + "loss": 2.5472, + "step": 22961 + }, + { + "epoch": 0.98, + "grad_norm": 0.4609375, + "learning_rate": 0.00010372204206940972, + "loss": 2.5648, + "step": 22962 + }, + { + "epoch": 0.98, + "grad_norm": 0.466796875, + "learning_rate": 0.00010371533877756849, + "loss": 2.4215, + "step": 22963 + }, + { + "epoch": 0.98, + "grad_norm": 0.453125, + "learning_rate": 0.0001037086354690096, + "loss": 2.5621, + "step": 22964 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.00010370193214376321, + "loss": 2.6103, + "step": 22965 + }, + { + "epoch": 0.98, + "grad_norm": 0.4453125, + "learning_rate": 0.00010369522880185942, + "loss": 2.5602, + "step": 22966 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.00010368852544332854, + "loss": 2.4917, + "step": 22967 + }, + { + "epoch": 0.98, + "grad_norm": 0.470703125, + "learning_rate": 0.00010368182206820058, + "loss": 2.5686, + "step": 22968 + }, + { + "epoch": 0.98, + "grad_norm": 0.478515625, + "learning_rate": 0.00010367511867650578, + "loss": 2.6096, + "step": 22969 + }, + { + "epoch": 0.98, + "grad_norm": 0.498046875, + "learning_rate": 0.00010366841526827431, + "loss": 2.5546, + "step": 22970 + }, + { + "epoch": 0.98, + "grad_norm": 0.453125, + "learning_rate": 0.0001036617118435363, + "loss": 2.5603, + "step": 22971 + }, + { + "epoch": 0.98, + "grad_norm": 0.490234375, + "learning_rate": 0.00010365500840232194, + "loss": 2.5666, + "step": 22972 + }, + { + "epoch": 0.98, + "grad_norm": 0.447265625, + "learning_rate": 0.00010364830494466135, + "loss": 2.506, + "step": 22973 + }, + { + "epoch": 0.98, + "grad_norm": 0.49609375, + "learning_rate": 0.00010364160147058473, + "loss": 2.5679, + "step": 22974 + }, + { + "epoch": 0.98, + "grad_norm": 0.404296875, + "learning_rate": 0.00010363489798012224, + "loss": 2.4843, + "step": 22975 + }, + { + "epoch": 0.98, + "grad_norm": 0.4453125, + "learning_rate": 0.00010362819447330403, + "loss": 2.4705, + "step": 22976 + }, + { + "epoch": 0.98, + "grad_norm": 0.462890625, + "learning_rate": 0.00010362149095016028, + "loss": 2.5861, + "step": 22977 + }, + { + "epoch": 0.98, + "grad_norm": 0.423828125, + "learning_rate": 0.00010361478741072115, + "loss": 2.3727, + "step": 22978 + }, + { + "epoch": 0.98, + "grad_norm": 0.474609375, + "learning_rate": 0.00010360808385501679, + "loss": 2.593, + "step": 22979 + }, + { + "epoch": 0.98, + "grad_norm": 0.439453125, + "learning_rate": 0.00010360138028307735, + "loss": 2.6103, + "step": 22980 + }, + { + "epoch": 0.98, + "grad_norm": 0.44921875, + "learning_rate": 0.00010359467669493305, + "loss": 2.4963, + "step": 22981 + }, + { + "epoch": 0.98, + "grad_norm": 0.474609375, + "learning_rate": 0.00010358797309061398, + "loss": 2.7204, + "step": 22982 + }, + { + "epoch": 0.98, + "grad_norm": 0.435546875, + "learning_rate": 0.00010358126947015039, + "loss": 2.5632, + "step": 22983 + }, + { + "epoch": 0.98, + "grad_norm": 0.419921875, + "learning_rate": 0.00010357456583357235, + "loss": 2.6036, + "step": 22984 + }, + { + "epoch": 0.98, + "grad_norm": 0.439453125, + "learning_rate": 0.0001035678621809101, + "loss": 2.5676, + "step": 22985 + }, + { + "epoch": 0.98, + "grad_norm": 0.4375, + "learning_rate": 0.00010356115851219376, + "loss": 2.5783, + "step": 22986 + }, + { + "epoch": 0.98, + "grad_norm": 0.43359375, + "learning_rate": 0.00010355445482745352, + "loss": 2.597, + "step": 22987 + }, + { + "epoch": 0.98, + "grad_norm": 0.453125, + "learning_rate": 0.00010354775112671953, + "loss": 2.4649, + "step": 22988 + }, + { + "epoch": 0.98, + "grad_norm": 0.48046875, + "learning_rate": 0.00010354104741002196, + "loss": 2.4405, + "step": 22989 + }, + { + "epoch": 0.98, + "grad_norm": 0.494140625, + "learning_rate": 0.00010353434367739094, + "loss": 2.5791, + "step": 22990 + }, + { + "epoch": 0.98, + "grad_norm": 0.46875, + "learning_rate": 0.00010352763992885672, + "loss": 2.536, + "step": 22991 + }, + { + "epoch": 0.98, + "grad_norm": 0.4609375, + "learning_rate": 0.00010352093616444936, + "loss": 2.5763, + "step": 22992 + }, + { + "epoch": 0.98, + "grad_norm": 0.412109375, + "learning_rate": 0.00010351423238419908, + "loss": 2.5737, + "step": 22993 + }, + { + "epoch": 0.98, + "grad_norm": 0.44140625, + "learning_rate": 0.00010350752858813605, + "loss": 2.5183, + "step": 22994 + }, + { + "epoch": 0.98, + "grad_norm": 0.46875, + "learning_rate": 0.00010350082477629045, + "loss": 2.4738, + "step": 22995 + }, + { + "epoch": 0.98, + "grad_norm": 0.416015625, + "learning_rate": 0.00010349412094869237, + "loss": 2.5786, + "step": 22996 + }, + { + "epoch": 0.98, + "grad_norm": 0.46875, + "learning_rate": 0.00010348741710537206, + "loss": 2.4574, + "step": 22997 + }, + { + "epoch": 0.98, + "grad_norm": 0.451171875, + "learning_rate": 0.00010348071324635962, + "loss": 2.6276, + "step": 22998 + }, + { + "epoch": 0.98, + "grad_norm": 0.404296875, + "learning_rate": 0.00010347400937168528, + "loss": 2.572, + "step": 22999 + }, + { + "epoch": 0.98, + "grad_norm": 0.458984375, + "learning_rate": 0.0001034673054813791, + "loss": 2.623, + "step": 23000 + }, + { + "epoch": 0.98, + "grad_norm": 0.50390625, + "learning_rate": 0.00010346060157547138, + "loss": 2.5508, + "step": 23001 + }, + { + "epoch": 0.98, + "grad_norm": 0.4453125, + "learning_rate": 0.00010345389765399218, + "loss": 2.4378, + "step": 23002 + }, + { + "epoch": 0.98, + "grad_norm": 0.5078125, + "learning_rate": 0.00010344719371697172, + "loss": 2.6125, + "step": 23003 + }, + { + "epoch": 0.98, + "grad_norm": 0.416015625, + "learning_rate": 0.00010344048976444014, + "loss": 2.5611, + "step": 23004 + }, + { + "epoch": 0.98, + "grad_norm": 0.470703125, + "learning_rate": 0.00010343378579642761, + "loss": 2.5898, + "step": 23005 + }, + { + "epoch": 0.98, + "grad_norm": 0.435546875, + "learning_rate": 0.0001034270818129643, + "loss": 2.4886, + "step": 23006 + }, + { + "epoch": 0.98, + "grad_norm": 0.447265625, + "learning_rate": 0.0001034203778140804, + "loss": 2.5181, + "step": 23007 + }, + { + "epoch": 0.98, + "grad_norm": 0.474609375, + "learning_rate": 0.00010341367379980602, + "loss": 2.5668, + "step": 23008 + }, + { + "epoch": 0.98, + "grad_norm": 0.439453125, + "learning_rate": 0.00010340696977017137, + "loss": 2.6939, + "step": 23009 + }, + { + "epoch": 0.98, + "grad_norm": 0.421875, + "learning_rate": 0.00010340026572520657, + "loss": 2.4383, + "step": 23010 + }, + { + "epoch": 0.98, + "grad_norm": 0.53515625, + "learning_rate": 0.00010339356166494187, + "loss": 2.5013, + "step": 23011 + }, + { + "epoch": 0.98, + "grad_norm": 0.439453125, + "learning_rate": 0.00010338685758940734, + "loss": 2.5349, + "step": 23012 + }, + { + "epoch": 0.98, + "grad_norm": 0.490234375, + "learning_rate": 0.00010338015349863321, + "loss": 2.4957, + "step": 23013 + }, + { + "epoch": 0.98, + "grad_norm": 0.462890625, + "learning_rate": 0.00010337344939264961, + "loss": 2.4617, + "step": 23014 + }, + { + "epoch": 0.98, + "grad_norm": 0.482421875, + "learning_rate": 0.00010336674527148675, + "loss": 2.5803, + "step": 23015 + }, + { + "epoch": 0.98, + "grad_norm": 0.4921875, + "learning_rate": 0.00010336004113517472, + "loss": 2.5034, + "step": 23016 + }, + { + "epoch": 0.98, + "grad_norm": 0.4296875, + "learning_rate": 0.00010335333698374376, + "loss": 2.5229, + "step": 23017 + }, + { + "epoch": 0.98, + "grad_norm": 0.44140625, + "learning_rate": 0.000103346632817224, + "loss": 2.5536, + "step": 23018 + }, + { + "epoch": 0.98, + "grad_norm": 0.44921875, + "learning_rate": 0.00010333992863564563, + "loss": 2.6333, + "step": 23019 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.00010333322443903879, + "loss": 2.5387, + "step": 23020 + }, + { + "epoch": 0.98, + "grad_norm": 0.48046875, + "learning_rate": 0.00010332652022743367, + "loss": 2.3712, + "step": 23021 + }, + { + "epoch": 0.98, + "grad_norm": 0.451171875, + "learning_rate": 0.0001033198160008604, + "loss": 2.5614, + "step": 23022 + }, + { + "epoch": 0.98, + "grad_norm": 0.458984375, + "learning_rate": 0.00010331311175934923, + "loss": 2.5519, + "step": 23023 + }, + { + "epoch": 0.98, + "grad_norm": 0.4140625, + "learning_rate": 0.00010330640750293022, + "loss": 2.5983, + "step": 23024 + }, + { + "epoch": 0.98, + "grad_norm": 0.404296875, + "learning_rate": 0.00010329970323163358, + "loss": 2.5047, + "step": 23025 + }, + { + "epoch": 0.98, + "grad_norm": 0.4609375, + "learning_rate": 0.00010329299894548949, + "loss": 2.563, + "step": 23026 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.00010328629464452812, + "loss": 2.6502, + "step": 23027 + }, + { + "epoch": 0.98, + "grad_norm": 0.439453125, + "learning_rate": 0.00010327959032877959, + "loss": 2.5026, + "step": 23028 + }, + { + "epoch": 0.98, + "grad_norm": 0.470703125, + "learning_rate": 0.00010327288599827416, + "loss": 2.5701, + "step": 23029 + }, + { + "epoch": 0.98, + "grad_norm": 0.421875, + "learning_rate": 0.00010326618165304189, + "loss": 2.4828, + "step": 23030 + }, + { + "epoch": 0.98, + "grad_norm": 0.51171875, + "learning_rate": 0.00010325947729311303, + "loss": 2.6391, + "step": 23031 + }, + { + "epoch": 0.98, + "grad_norm": 0.43359375, + "learning_rate": 0.00010325277291851769, + "loss": 2.4769, + "step": 23032 + }, + { + "epoch": 0.98, + "grad_norm": 0.431640625, + "learning_rate": 0.00010324606852928607, + "loss": 2.5452, + "step": 23033 + }, + { + "epoch": 0.98, + "grad_norm": 0.447265625, + "learning_rate": 0.00010323936412544835, + "loss": 2.5513, + "step": 23034 + }, + { + "epoch": 0.98, + "grad_norm": 0.453125, + "learning_rate": 0.00010323265970703462, + "loss": 2.4364, + "step": 23035 + }, + { + "epoch": 0.98, + "grad_norm": 0.5234375, + "learning_rate": 0.00010322595527407512, + "loss": 2.5982, + "step": 23036 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.00010321925082660003, + "loss": 2.5645, + "step": 23037 + }, + { + "epoch": 0.98, + "grad_norm": 0.419921875, + "learning_rate": 0.00010321254636463949, + "loss": 2.5148, + "step": 23038 + }, + { + "epoch": 0.98, + "grad_norm": 0.40625, + "learning_rate": 0.00010320584188822362, + "loss": 2.426, + "step": 23039 + }, + { + "epoch": 0.98, + "grad_norm": 0.431640625, + "learning_rate": 0.00010319913739738267, + "loss": 2.4882, + "step": 23040 + }, + { + "epoch": 0.98, + "grad_norm": 0.484375, + "learning_rate": 0.00010319243289214675, + "loss": 2.6592, + "step": 23041 + }, + { + "epoch": 0.98, + "grad_norm": 0.46484375, + "learning_rate": 0.00010318572837254607, + "loss": 2.8009, + "step": 23042 + }, + { + "epoch": 0.98, + "grad_norm": 0.416015625, + "learning_rate": 0.00010317902383861076, + "loss": 2.5677, + "step": 23043 + }, + { + "epoch": 0.98, + "grad_norm": 0.41796875, + "learning_rate": 0.000103172319290371, + "loss": 2.4398, + "step": 23044 + }, + { + "epoch": 0.98, + "grad_norm": 0.498046875, + "learning_rate": 0.000103165614727857, + "loss": 2.5118, + "step": 23045 + }, + { + "epoch": 0.98, + "grad_norm": 0.46875, + "learning_rate": 0.00010315891015109886, + "loss": 2.6021, + "step": 23046 + }, + { + "epoch": 0.98, + "grad_norm": 0.443359375, + "learning_rate": 0.00010315220556012675, + "loss": 2.7199, + "step": 23047 + }, + { + "epoch": 0.98, + "grad_norm": 0.474609375, + "learning_rate": 0.00010314550095497094, + "loss": 2.5558, + "step": 23048 + }, + { + "epoch": 0.98, + "grad_norm": 0.40625, + "learning_rate": 0.00010313879633566146, + "loss": 2.4771, + "step": 23049 + }, + { + "epoch": 0.98, + "grad_norm": 0.4453125, + "learning_rate": 0.00010313209170222859, + "loss": 2.4226, + "step": 23050 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.0001031253870547024, + "loss": 2.6136, + "step": 23051 + }, + { + "epoch": 0.98, + "grad_norm": 0.458984375, + "learning_rate": 0.00010311868239311315, + "loss": 2.6155, + "step": 23052 + }, + { + "epoch": 0.98, + "grad_norm": 0.52734375, + "learning_rate": 0.00010311197771749098, + "loss": 2.5493, + "step": 23053 + }, + { + "epoch": 0.98, + "grad_norm": 0.458984375, + "learning_rate": 0.00010310527302786602, + "loss": 2.462, + "step": 23054 + }, + { + "epoch": 0.98, + "grad_norm": 0.421875, + "learning_rate": 0.00010309856832426845, + "loss": 2.4864, + "step": 23055 + }, + { + "epoch": 0.98, + "grad_norm": 0.482421875, + "learning_rate": 0.00010309186360672851, + "loss": 2.5372, + "step": 23056 + }, + { + "epoch": 0.98, + "grad_norm": 0.5, + "learning_rate": 0.00010308515887527625, + "loss": 2.5751, + "step": 23057 + }, + { + "epoch": 0.98, + "grad_norm": 0.4375, + "learning_rate": 0.00010307845412994198, + "loss": 2.5601, + "step": 23058 + }, + { + "epoch": 0.98, + "grad_norm": 0.458984375, + "learning_rate": 0.00010307174937075572, + "loss": 2.5248, + "step": 23059 + }, + { + "epoch": 0.98, + "grad_norm": 0.435546875, + "learning_rate": 0.00010306504459774773, + "loss": 2.5541, + "step": 23060 + }, + { + "epoch": 0.98, + "grad_norm": 0.447265625, + "learning_rate": 0.00010305833981094817, + "loss": 2.5449, + "step": 23061 + }, + { + "epoch": 0.98, + "grad_norm": 0.42578125, + "learning_rate": 0.0001030516350103872, + "loss": 2.5308, + "step": 23062 + }, + { + "epoch": 0.98, + "grad_norm": 0.484375, + "learning_rate": 0.00010304493019609495, + "loss": 2.6814, + "step": 23063 + }, + { + "epoch": 0.98, + "grad_norm": 0.400390625, + "learning_rate": 0.0001030382253681017, + "loss": 2.3575, + "step": 23064 + }, + { + "epoch": 0.98, + "grad_norm": 0.4921875, + "learning_rate": 0.00010303152052643747, + "loss": 2.6215, + "step": 23065 + }, + { + "epoch": 0.98, + "grad_norm": 0.4453125, + "learning_rate": 0.00010302481567113256, + "loss": 2.3974, + "step": 23066 + }, + { + "epoch": 0.98, + "grad_norm": 0.4296875, + "learning_rate": 0.00010301811080221706, + "loss": 2.4831, + "step": 23067 + }, + { + "epoch": 0.98, + "grad_norm": 0.421875, + "learning_rate": 0.00010301140591972118, + "loss": 2.6335, + "step": 23068 + }, + { + "epoch": 0.98, + "grad_norm": 0.4296875, + "learning_rate": 0.00010300470102367505, + "loss": 2.484, + "step": 23069 + }, + { + "epoch": 0.98, + "grad_norm": 0.439453125, + "learning_rate": 0.0001029979961141089, + "loss": 2.5307, + "step": 23070 + }, + { + "epoch": 0.98, + "grad_norm": 0.498046875, + "learning_rate": 0.00010299129119105281, + "loss": 2.5131, + "step": 23071 + }, + { + "epoch": 0.98, + "grad_norm": 0.419921875, + "learning_rate": 0.00010298458625453706, + "loss": 2.6099, + "step": 23072 + }, + { + "epoch": 0.98, + "grad_norm": 0.4296875, + "learning_rate": 0.00010297788130459172, + "loss": 2.4976, + "step": 23073 + }, + { + "epoch": 0.98, + "grad_norm": 0.482421875, + "learning_rate": 0.00010297117634124703, + "loss": 2.442, + "step": 23074 + }, + { + "epoch": 0.98, + "grad_norm": 0.4765625, + "learning_rate": 0.00010296447136453311, + "loss": 2.6231, + "step": 23075 + }, + { + "epoch": 0.98, + "grad_norm": 0.4140625, + "learning_rate": 0.00010295776637448016, + "loss": 2.4796, + "step": 23076 + }, + { + "epoch": 0.98, + "grad_norm": 0.482421875, + "learning_rate": 0.00010295106137111836, + "loss": 2.4718, + "step": 23077 + }, + { + "epoch": 0.98, + "grad_norm": 0.4609375, + "learning_rate": 0.00010294435635447785, + "loss": 2.6026, + "step": 23078 + }, + { + "epoch": 0.98, + "grad_norm": 0.5, + "learning_rate": 0.0001029376513245888, + "loss": 2.5756, + "step": 23079 + }, + { + "epoch": 0.98, + "grad_norm": 0.4765625, + "learning_rate": 0.00010293094628148141, + "loss": 2.5154, + "step": 23080 + }, + { + "epoch": 0.98, + "grad_norm": 0.408203125, + "learning_rate": 0.0001029242412251858, + "loss": 2.5142, + "step": 23081 + }, + { + "epoch": 0.98, + "grad_norm": 0.462890625, + "learning_rate": 0.0001029175361557322, + "loss": 2.6188, + "step": 23082 + }, + { + "epoch": 0.98, + "grad_norm": 0.419921875, + "learning_rate": 0.00010291083107315076, + "loss": 2.5334, + "step": 23083 + }, + { + "epoch": 0.98, + "grad_norm": 0.443359375, + "learning_rate": 0.00010290412597747165, + "loss": 2.5578, + "step": 23084 + }, + { + "epoch": 0.98, + "grad_norm": 0.4296875, + "learning_rate": 0.00010289742086872503, + "loss": 2.3965, + "step": 23085 + }, + { + "epoch": 0.98, + "grad_norm": 0.447265625, + "learning_rate": 0.00010289071574694104, + "loss": 2.502, + "step": 23086 + }, + { + "epoch": 0.98, + "grad_norm": 0.447265625, + "learning_rate": 0.00010288401061214991, + "loss": 2.3591, + "step": 23087 + }, + { + "epoch": 0.98, + "grad_norm": 0.49609375, + "learning_rate": 0.00010287730546438181, + "loss": 2.5144, + "step": 23088 + }, + { + "epoch": 0.98, + "grad_norm": 0.474609375, + "learning_rate": 0.00010287060030366684, + "loss": 2.5509, + "step": 23089 + }, + { + "epoch": 0.98, + "grad_norm": 0.455078125, + "learning_rate": 0.00010286389513003525, + "loss": 2.602, + "step": 23090 + }, + { + "epoch": 0.98, + "grad_norm": 0.58984375, + "learning_rate": 0.00010285718994351717, + "loss": 2.5409, + "step": 23091 + }, + { + "epoch": 0.98, + "grad_norm": 0.470703125, + "learning_rate": 0.00010285048474414278, + "loss": 2.4245, + "step": 23092 + }, + { + "epoch": 0.98, + "grad_norm": 0.484375, + "learning_rate": 0.00010284377953194227, + "loss": 2.5121, + "step": 23093 + }, + { + "epoch": 0.98, + "grad_norm": 0.416015625, + "learning_rate": 0.00010283707430694576, + "loss": 2.3472, + "step": 23094 + }, + { + "epoch": 0.98, + "grad_norm": 0.4296875, + "learning_rate": 0.00010283036906918343, + "loss": 2.5332, + "step": 23095 + }, + { + "epoch": 0.98, + "grad_norm": 0.46484375, + "learning_rate": 0.00010282366381868555, + "loss": 2.4977, + "step": 23096 + }, + { + "epoch": 0.98, + "grad_norm": 0.451171875, + "learning_rate": 0.00010281695855548214, + "loss": 2.6414, + "step": 23097 + }, + { + "epoch": 0.98, + "grad_norm": 0.50390625, + "learning_rate": 0.0001028102532796035, + "loss": 2.5953, + "step": 23098 + }, + { + "epoch": 0.98, + "grad_norm": 0.439453125, + "learning_rate": 0.00010280354799107972, + "loss": 2.5398, + "step": 23099 + }, + { + "epoch": 0.98, + "grad_norm": 0.46484375, + "learning_rate": 0.000102796842689941, + "loss": 2.4541, + "step": 23100 + }, + { + "epoch": 0.98, + "grad_norm": 0.46484375, + "learning_rate": 0.00010279013737621752, + "loss": 2.6552, + "step": 23101 + }, + { + "epoch": 0.98, + "grad_norm": 0.416015625, + "learning_rate": 0.00010278343204993945, + "loss": 2.5096, + "step": 23102 + }, + { + "epoch": 0.98, + "grad_norm": 0.4921875, + "learning_rate": 0.00010277672671113691, + "loss": 2.5592, + "step": 23103 + }, + { + "epoch": 0.98, + "grad_norm": 0.486328125, + "learning_rate": 0.00010277002135984017, + "loss": 2.4125, + "step": 23104 + }, + { + "epoch": 0.98, + "grad_norm": 0.46875, + "learning_rate": 0.00010276331599607931, + "loss": 2.5173, + "step": 23105 + }, + { + "epoch": 0.98, + "grad_norm": 0.447265625, + "learning_rate": 0.00010275661061988455, + "loss": 2.5907, + "step": 23106 + }, + { + "epoch": 0.98, + "grad_norm": 0.458984375, + "learning_rate": 0.00010274990523128605, + "loss": 2.54, + "step": 23107 + }, + { + "epoch": 0.98, + "grad_norm": 0.490234375, + "learning_rate": 0.00010274319983031397, + "loss": 2.7005, + "step": 23108 + }, + { + "epoch": 0.98, + "grad_norm": 0.44921875, + "learning_rate": 0.00010273649441699851, + "loss": 2.5554, + "step": 23109 + }, + { + "epoch": 0.98, + "grad_norm": 0.45703125, + "learning_rate": 0.00010272978899136982, + "loss": 2.5978, + "step": 23110 + }, + { + "epoch": 0.98, + "grad_norm": 0.462890625, + "learning_rate": 0.00010272308355345806, + "loss": 2.5641, + "step": 23111 + }, + { + "epoch": 0.98, + "grad_norm": 0.392578125, + "learning_rate": 0.00010271637810329346, + "loss": 2.4724, + "step": 23112 + }, + { + "epoch": 0.98, + "grad_norm": 0.423828125, + "learning_rate": 0.00010270967264090609, + "loss": 2.4897, + "step": 23113 + }, + { + "epoch": 0.98, + "grad_norm": 0.443359375, + "learning_rate": 0.00010270296716632624, + "loss": 2.6614, + "step": 23114 + }, + { + "epoch": 0.98, + "grad_norm": 0.484375, + "learning_rate": 0.00010269626167958399, + "loss": 2.5199, + "step": 23115 + }, + { + "epoch": 0.99, + "grad_norm": 0.42578125, + "learning_rate": 0.00010268955618070957, + "loss": 2.5414, + "step": 23116 + }, + { + "epoch": 0.99, + "grad_norm": 0.458984375, + "learning_rate": 0.00010268285066973313, + "loss": 2.5567, + "step": 23117 + }, + { + "epoch": 0.99, + "grad_norm": 0.4375, + "learning_rate": 0.00010267614514668485, + "loss": 2.658, + "step": 23118 + }, + { + "epoch": 0.99, + "grad_norm": 0.466796875, + "learning_rate": 0.00010266943961159485, + "loss": 2.6072, + "step": 23119 + }, + { + "epoch": 0.99, + "grad_norm": 0.396484375, + "learning_rate": 0.00010266273406449341, + "loss": 2.5812, + "step": 23120 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.00010265602850541056, + "loss": 2.5533, + "step": 23121 + }, + { + "epoch": 0.99, + "grad_norm": 0.458984375, + "learning_rate": 0.00010264932293437661, + "loss": 2.7516, + "step": 23122 + }, + { + "epoch": 0.99, + "grad_norm": 0.57421875, + "learning_rate": 0.00010264261735142166, + "loss": 2.795, + "step": 23123 + }, + { + "epoch": 0.99, + "grad_norm": 0.466796875, + "learning_rate": 0.00010263591175657591, + "loss": 2.5451, + "step": 23124 + }, + { + "epoch": 0.99, + "grad_norm": 0.48046875, + "learning_rate": 0.00010262920614986952, + "loss": 2.4887, + "step": 23125 + }, + { + "epoch": 0.99, + "grad_norm": 0.4765625, + "learning_rate": 0.00010262250053133267, + "loss": 2.5402, + "step": 23126 + }, + { + "epoch": 0.99, + "grad_norm": 0.455078125, + "learning_rate": 0.00010261579490099547, + "loss": 2.6354, + "step": 23127 + }, + { + "epoch": 0.99, + "grad_norm": 0.474609375, + "learning_rate": 0.00010260908925888821, + "loss": 2.5892, + "step": 23128 + }, + { + "epoch": 0.99, + "grad_norm": 0.431640625, + "learning_rate": 0.00010260238360504098, + "loss": 2.5805, + "step": 23129 + }, + { + "epoch": 0.99, + "grad_norm": 0.439453125, + "learning_rate": 0.00010259567793948397, + "loss": 2.5687, + "step": 23130 + }, + { + "epoch": 0.99, + "grad_norm": 0.470703125, + "learning_rate": 0.00010258897226224737, + "loss": 2.6592, + "step": 23131 + }, + { + "epoch": 0.99, + "grad_norm": 0.47265625, + "learning_rate": 0.00010258226657336134, + "loss": 2.5854, + "step": 23132 + }, + { + "epoch": 0.99, + "grad_norm": 0.470703125, + "learning_rate": 0.00010257556087285605, + "loss": 2.5455, + "step": 23133 + }, + { + "epoch": 0.99, + "grad_norm": 0.427734375, + "learning_rate": 0.00010256885516076169, + "loss": 2.5369, + "step": 23134 + }, + { + "epoch": 0.99, + "grad_norm": 0.427734375, + "learning_rate": 0.00010256214943710838, + "loss": 2.5728, + "step": 23135 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010255544370192638, + "loss": 2.6024, + "step": 23136 + }, + { + "epoch": 0.99, + "grad_norm": 0.51171875, + "learning_rate": 0.00010254873795524577, + "loss": 2.5832, + "step": 23137 + }, + { + "epoch": 0.99, + "grad_norm": 0.458984375, + "learning_rate": 0.00010254203219709682, + "loss": 2.4477, + "step": 23138 + }, + { + "epoch": 0.99, + "grad_norm": 0.4453125, + "learning_rate": 0.00010253532642750962, + "loss": 2.5935, + "step": 23139 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010252862064651441, + "loss": 2.6107, + "step": 23140 + }, + { + "epoch": 0.99, + "grad_norm": 0.431640625, + "learning_rate": 0.00010252191485414131, + "loss": 2.5147, + "step": 23141 + }, + { + "epoch": 0.99, + "grad_norm": 0.416015625, + "learning_rate": 0.00010251520905042052, + "loss": 2.5385, + "step": 23142 + }, + { + "epoch": 0.99, + "grad_norm": 0.494140625, + "learning_rate": 0.00010250850323538219, + "loss": 2.7078, + "step": 23143 + }, + { + "epoch": 0.99, + "grad_norm": 0.44140625, + "learning_rate": 0.00010250179740905654, + "loss": 2.5521, + "step": 23144 + }, + { + "epoch": 0.99, + "grad_norm": 0.44140625, + "learning_rate": 0.0001024950915714737, + "loss": 2.5278, + "step": 23145 + }, + { + "epoch": 0.99, + "grad_norm": 0.490234375, + "learning_rate": 0.00010248838572266386, + "loss": 2.5459, + "step": 23146 + }, + { + "epoch": 0.99, + "grad_norm": 0.390625, + "learning_rate": 0.0001024816798626572, + "loss": 2.5423, + "step": 23147 + }, + { + "epoch": 0.99, + "grad_norm": 0.44921875, + "learning_rate": 0.0001024749739914839, + "loss": 2.564, + "step": 23148 + }, + { + "epoch": 0.99, + "grad_norm": 0.455078125, + "learning_rate": 0.0001024682681091741, + "loss": 2.5333, + "step": 23149 + }, + { + "epoch": 0.99, + "grad_norm": 0.412109375, + "learning_rate": 0.000102461562215758, + "loss": 2.6211, + "step": 23150 + }, + { + "epoch": 0.99, + "grad_norm": 0.5078125, + "learning_rate": 0.00010245485631126574, + "loss": 2.7145, + "step": 23151 + }, + { + "epoch": 0.99, + "grad_norm": 0.48046875, + "learning_rate": 0.00010244815039572759, + "loss": 2.3874, + "step": 23152 + }, + { + "epoch": 0.99, + "grad_norm": 0.474609375, + "learning_rate": 0.00010244144446917361, + "loss": 2.5442, + "step": 23153 + }, + { + "epoch": 0.99, + "grad_norm": 0.447265625, + "learning_rate": 0.00010243473853163403, + "loss": 2.6813, + "step": 23154 + }, + { + "epoch": 0.99, + "grad_norm": 0.416015625, + "learning_rate": 0.00010242803258313904, + "loss": 2.4678, + "step": 23155 + }, + { + "epoch": 0.99, + "grad_norm": 0.4609375, + "learning_rate": 0.00010242132662371875, + "loss": 2.5338, + "step": 23156 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.00010241462065340341, + "loss": 2.5894, + "step": 23157 + }, + { + "epoch": 0.99, + "grad_norm": 0.4921875, + "learning_rate": 0.00010240791467222315, + "loss": 2.5109, + "step": 23158 + }, + { + "epoch": 0.99, + "grad_norm": 0.4375, + "learning_rate": 0.00010240120868020815, + "loss": 2.444, + "step": 23159 + }, + { + "epoch": 0.99, + "grad_norm": 0.4140625, + "learning_rate": 0.00010239450267738859, + "loss": 2.5943, + "step": 23160 + }, + { + "epoch": 0.99, + "grad_norm": 0.421875, + "learning_rate": 0.00010238779666379466, + "loss": 2.5227, + "step": 23161 + }, + { + "epoch": 0.99, + "grad_norm": 0.431640625, + "learning_rate": 0.00010238109063945647, + "loss": 2.6749, + "step": 23162 + }, + { + "epoch": 0.99, + "grad_norm": 0.44921875, + "learning_rate": 0.00010237438460440429, + "loss": 2.6528, + "step": 23163 + }, + { + "epoch": 0.99, + "grad_norm": 0.439453125, + "learning_rate": 0.00010236767855866823, + "loss": 2.491, + "step": 23164 + }, + { + "epoch": 0.99, + "grad_norm": 0.462890625, + "learning_rate": 0.00010236097250227847, + "loss": 2.4718, + "step": 23165 + }, + { + "epoch": 0.99, + "grad_norm": 0.51171875, + "learning_rate": 0.0001023542664352652, + "loss": 2.4917, + "step": 23166 + }, + { + "epoch": 0.99, + "grad_norm": 0.50390625, + "learning_rate": 0.00010234756035765862, + "loss": 2.6805, + "step": 23167 + }, + { + "epoch": 0.99, + "grad_norm": 0.4609375, + "learning_rate": 0.00010234085426948885, + "loss": 2.5324, + "step": 23168 + }, + { + "epoch": 0.99, + "grad_norm": 0.4609375, + "learning_rate": 0.0001023341481707861, + "loss": 2.4304, + "step": 23169 + }, + { + "epoch": 0.99, + "grad_norm": 0.48046875, + "learning_rate": 0.00010232744206158051, + "loss": 2.6828, + "step": 23170 + }, + { + "epoch": 0.99, + "grad_norm": 0.515625, + "learning_rate": 0.00010232073594190232, + "loss": 2.5545, + "step": 23171 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.00010231402981178165, + "loss": 2.55, + "step": 23172 + }, + { + "epoch": 0.99, + "grad_norm": 0.42578125, + "learning_rate": 0.00010230732367124869, + "loss": 2.4861, + "step": 23173 + }, + { + "epoch": 0.99, + "grad_norm": 0.447265625, + "learning_rate": 0.00010230061752033362, + "loss": 2.5395, + "step": 23174 + }, + { + "epoch": 0.99, + "grad_norm": 0.4921875, + "learning_rate": 0.00010229391135906661, + "loss": 2.6254, + "step": 23175 + }, + { + "epoch": 0.99, + "grad_norm": 0.4921875, + "learning_rate": 0.00010228720518747785, + "loss": 2.5887, + "step": 23176 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010228049900559748, + "loss": 2.5839, + "step": 23177 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.0001022737928134557, + "loss": 2.5208, + "step": 23178 + }, + { + "epoch": 0.99, + "grad_norm": 0.439453125, + "learning_rate": 0.0001022670866110827, + "loss": 2.5774, + "step": 23179 + }, + { + "epoch": 0.99, + "grad_norm": 0.439453125, + "learning_rate": 0.00010226038039850863, + "loss": 2.6854, + "step": 23180 + }, + { + "epoch": 0.99, + "grad_norm": 0.51171875, + "learning_rate": 0.00010225367417576368, + "loss": 2.5724, + "step": 23181 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.000102246967942878, + "loss": 2.528, + "step": 23182 + }, + { + "epoch": 0.99, + "grad_norm": 0.466796875, + "learning_rate": 0.00010224026169988181, + "loss": 2.6122, + "step": 23183 + }, + { + "epoch": 0.99, + "grad_norm": 0.435546875, + "learning_rate": 0.00010223355544680526, + "loss": 2.4633, + "step": 23184 + }, + { + "epoch": 0.99, + "grad_norm": 0.490234375, + "learning_rate": 0.00010222684918367851, + "loss": 2.5622, + "step": 23185 + }, + { + "epoch": 0.99, + "grad_norm": 0.51953125, + "learning_rate": 0.00010222014291053175, + "loss": 2.6094, + "step": 23186 + }, + { + "epoch": 0.99, + "grad_norm": 0.451171875, + "learning_rate": 0.0001022134366273952, + "loss": 2.4316, + "step": 23187 + }, + { + "epoch": 0.99, + "grad_norm": 0.421875, + "learning_rate": 0.00010220673033429893, + "loss": 2.5085, + "step": 23188 + }, + { + "epoch": 0.99, + "grad_norm": 0.478515625, + "learning_rate": 0.00010220002403127322, + "loss": 2.5525, + "step": 23189 + }, + { + "epoch": 0.99, + "grad_norm": 0.48046875, + "learning_rate": 0.00010219331771834819, + "loss": 2.4663, + "step": 23190 + }, + { + "epoch": 0.99, + "grad_norm": 0.423828125, + "learning_rate": 0.00010218661139555406, + "loss": 2.4522, + "step": 23191 + }, + { + "epoch": 0.99, + "grad_norm": 0.466796875, + "learning_rate": 0.00010217990506292097, + "loss": 2.5936, + "step": 23192 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.00010217319872047908, + "loss": 2.6063, + "step": 23193 + }, + { + "epoch": 0.99, + "grad_norm": 0.42578125, + "learning_rate": 0.00010216649236825859, + "loss": 2.5668, + "step": 23194 + }, + { + "epoch": 0.99, + "grad_norm": 0.5546875, + "learning_rate": 0.0001021597860062897, + "loss": 2.6027, + "step": 23195 + }, + { + "epoch": 0.99, + "grad_norm": 0.51171875, + "learning_rate": 0.00010215307963460254, + "loss": 2.6263, + "step": 23196 + }, + { + "epoch": 0.99, + "grad_norm": 0.4921875, + "learning_rate": 0.00010214637325322731, + "loss": 2.5931, + "step": 23197 + }, + { + "epoch": 0.99, + "grad_norm": 0.5, + "learning_rate": 0.0001021396668621942, + "loss": 2.6853, + "step": 23198 + }, + { + "epoch": 0.99, + "grad_norm": 0.435546875, + "learning_rate": 0.00010213296046153338, + "loss": 2.5483, + "step": 23199 + }, + { + "epoch": 0.99, + "grad_norm": 0.435546875, + "learning_rate": 0.000102126254051275, + "loss": 2.5043, + "step": 23200 + }, + { + "epoch": 0.99, + "grad_norm": 0.474609375, + "learning_rate": 0.00010211954763144926, + "loss": 2.7165, + "step": 23201 + }, + { + "epoch": 0.99, + "grad_norm": 0.482421875, + "learning_rate": 0.0001021128412020863, + "loss": 2.61, + "step": 23202 + }, + { + "epoch": 0.99, + "grad_norm": 0.41796875, + "learning_rate": 0.00010210613476321638, + "loss": 2.4124, + "step": 23203 + }, + { + "epoch": 0.99, + "grad_norm": 0.4375, + "learning_rate": 0.00010209942831486959, + "loss": 2.5361, + "step": 23204 + }, + { + "epoch": 0.99, + "grad_norm": 0.462890625, + "learning_rate": 0.00010209272185707615, + "loss": 2.5558, + "step": 23205 + }, + { + "epoch": 0.99, + "grad_norm": 0.4375, + "learning_rate": 0.00010208601538986622, + "loss": 2.495, + "step": 23206 + }, + { + "epoch": 0.99, + "grad_norm": 0.515625, + "learning_rate": 0.00010207930891326997, + "loss": 2.6256, + "step": 23207 + }, + { + "epoch": 0.99, + "grad_norm": 0.57421875, + "learning_rate": 0.00010207260242731762, + "loss": 2.4728, + "step": 23208 + }, + { + "epoch": 0.99, + "grad_norm": 0.470703125, + "learning_rate": 0.0001020658959320393, + "loss": 2.5489, + "step": 23209 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.0001020591894274652, + "loss": 2.5077, + "step": 23210 + }, + { + "epoch": 0.99, + "grad_norm": 0.431640625, + "learning_rate": 0.00010205248291362553, + "loss": 2.4829, + "step": 23211 + }, + { + "epoch": 0.99, + "grad_norm": 0.498046875, + "learning_rate": 0.00010204577639055038, + "loss": 2.5881, + "step": 23212 + }, + { + "epoch": 0.99, + "grad_norm": 0.46875, + "learning_rate": 0.00010203906985827004, + "loss": 2.5654, + "step": 23213 + }, + { + "epoch": 0.99, + "grad_norm": 0.48046875, + "learning_rate": 0.00010203236331681458, + "loss": 2.4546, + "step": 23214 + }, + { + "epoch": 0.99, + "grad_norm": 0.4921875, + "learning_rate": 0.00010202565676621428, + "loss": 2.5732, + "step": 23215 + }, + { + "epoch": 0.99, + "grad_norm": 0.482421875, + "learning_rate": 0.00010201895020649923, + "loss": 2.5238, + "step": 23216 + }, + { + "epoch": 0.99, + "grad_norm": 0.53125, + "learning_rate": 0.00010201224363769965, + "loss": 2.5872, + "step": 23217 + }, + { + "epoch": 0.99, + "grad_norm": 0.416015625, + "learning_rate": 0.0001020055370598457, + "loss": 2.4002, + "step": 23218 + }, + { + "epoch": 0.99, + "grad_norm": 0.45703125, + "learning_rate": 0.00010199883047296761, + "loss": 2.5386, + "step": 23219 + }, + { + "epoch": 0.99, + "grad_norm": 0.4453125, + "learning_rate": 0.00010199212387709545, + "loss": 2.4748, + "step": 23220 + }, + { + "epoch": 0.99, + "grad_norm": 0.455078125, + "learning_rate": 0.0001019854172722595, + "loss": 2.4303, + "step": 23221 + }, + { + "epoch": 0.99, + "grad_norm": 0.462890625, + "learning_rate": 0.00010197871065848989, + "loss": 2.4704, + "step": 23222 + }, + { + "epoch": 0.99, + "grad_norm": 0.482421875, + "learning_rate": 0.00010197200403581681, + "loss": 2.5193, + "step": 23223 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010196529740427043, + "loss": 2.4558, + "step": 23224 + }, + { + "epoch": 0.99, + "grad_norm": 0.45703125, + "learning_rate": 0.00010195859076388092, + "loss": 2.686, + "step": 23225 + }, + { + "epoch": 0.99, + "grad_norm": 0.474609375, + "learning_rate": 0.00010195188411467847, + "loss": 2.6268, + "step": 23226 + }, + { + "epoch": 0.99, + "grad_norm": 0.494140625, + "learning_rate": 0.00010194517745669327, + "loss": 2.4313, + "step": 23227 + }, + { + "epoch": 0.99, + "grad_norm": 0.412109375, + "learning_rate": 0.00010193847078995544, + "loss": 2.4634, + "step": 23228 + }, + { + "epoch": 0.99, + "grad_norm": 0.44921875, + "learning_rate": 0.00010193176411449525, + "loss": 2.6277, + "step": 23229 + }, + { + "epoch": 0.99, + "grad_norm": 0.447265625, + "learning_rate": 0.0001019250574303428, + "loss": 2.5694, + "step": 23230 + }, + { + "epoch": 0.99, + "grad_norm": 0.3984375, + "learning_rate": 0.00010191835073752832, + "loss": 2.4822, + "step": 23231 + }, + { + "epoch": 0.99, + "grad_norm": 0.419921875, + "learning_rate": 0.00010191164403608193, + "loss": 2.4345, + "step": 23232 + }, + { + "epoch": 0.99, + "grad_norm": 0.5625, + "learning_rate": 0.00010190493732603388, + "loss": 2.5633, + "step": 23233 + }, + { + "epoch": 0.99, + "grad_norm": 0.455078125, + "learning_rate": 0.00010189823060741426, + "loss": 2.4923, + "step": 23234 + }, + { + "epoch": 0.99, + "grad_norm": 0.5078125, + "learning_rate": 0.00010189152388025334, + "loss": 2.5705, + "step": 23235 + }, + { + "epoch": 0.99, + "grad_norm": 0.486328125, + "learning_rate": 0.00010188481714458122, + "loss": 2.7006, + "step": 23236 + }, + { + "epoch": 0.99, + "grad_norm": 0.48828125, + "learning_rate": 0.00010187811040042811, + "loss": 2.446, + "step": 23237 + }, + { + "epoch": 0.99, + "grad_norm": 0.416015625, + "learning_rate": 0.00010187140364782422, + "loss": 2.4529, + "step": 23238 + }, + { + "epoch": 0.99, + "grad_norm": 0.42578125, + "learning_rate": 0.00010186469688679969, + "loss": 2.4675, + "step": 23239 + }, + { + "epoch": 0.99, + "grad_norm": 0.45703125, + "learning_rate": 0.00010185799011738471, + "loss": 2.5592, + "step": 23240 + }, + { + "epoch": 0.99, + "grad_norm": 0.482421875, + "learning_rate": 0.00010185128333960942, + "loss": 2.6344, + "step": 23241 + }, + { + "epoch": 0.99, + "grad_norm": 0.44140625, + "learning_rate": 0.00010184457655350406, + "loss": 2.5666, + "step": 23242 + }, + { + "epoch": 0.99, + "grad_norm": 0.48046875, + "learning_rate": 0.00010183786975909879, + "loss": 2.6458, + "step": 23243 + }, + { + "epoch": 0.99, + "grad_norm": 0.42578125, + "learning_rate": 0.00010183116295642374, + "loss": 2.4477, + "step": 23244 + }, + { + "epoch": 0.99, + "grad_norm": 0.443359375, + "learning_rate": 0.00010182445614550915, + "loss": 2.6093, + "step": 23245 + }, + { + "epoch": 0.99, + "grad_norm": 0.47265625, + "learning_rate": 0.00010181774932638517, + "loss": 2.5122, + "step": 23246 + }, + { + "epoch": 0.99, + "grad_norm": 0.44140625, + "learning_rate": 0.000101811042499082, + "loss": 2.5749, + "step": 23247 + }, + { + "epoch": 0.99, + "grad_norm": 0.482421875, + "learning_rate": 0.00010180433566362977, + "loss": 2.5337, + "step": 23248 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010179762882005871, + "loss": 2.6067, + "step": 23249 + }, + { + "epoch": 0.99, + "grad_norm": 0.431640625, + "learning_rate": 0.00010179092196839895, + "loss": 2.546, + "step": 23250 + }, + { + "epoch": 0.99, + "grad_norm": 0.3984375, + "learning_rate": 0.00010178421510868075, + "loss": 2.5493, + "step": 23251 + }, + { + "epoch": 0.99, + "grad_norm": 0.423828125, + "learning_rate": 0.00010177750824093417, + "loss": 2.46, + "step": 23252 + }, + { + "epoch": 0.99, + "grad_norm": 0.431640625, + "learning_rate": 0.00010177080136518948, + "loss": 2.5566, + "step": 23253 + }, + { + "epoch": 0.99, + "grad_norm": 0.498046875, + "learning_rate": 0.00010176409448147685, + "loss": 2.6263, + "step": 23254 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.00010175738758982643, + "loss": 2.547, + "step": 23255 + }, + { + "epoch": 0.99, + "grad_norm": 0.4140625, + "learning_rate": 0.00010175068069026841, + "loss": 2.5745, + "step": 23256 + }, + { + "epoch": 0.99, + "grad_norm": 0.44140625, + "learning_rate": 0.00010174397378283295, + "loss": 2.4624, + "step": 23257 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010173726686755023, + "loss": 2.5393, + "step": 23258 + }, + { + "epoch": 0.99, + "grad_norm": 0.427734375, + "learning_rate": 0.00010173055994445048, + "loss": 2.4721, + "step": 23259 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.00010172385301356382, + "loss": 2.6912, + "step": 23260 + }, + { + "epoch": 0.99, + "grad_norm": 0.482421875, + "learning_rate": 0.00010171714607492047, + "loss": 2.7076, + "step": 23261 + }, + { + "epoch": 0.99, + "grad_norm": 0.404296875, + "learning_rate": 0.00010171043912855058, + "loss": 2.577, + "step": 23262 + }, + { + "epoch": 0.99, + "grad_norm": 0.4453125, + "learning_rate": 0.00010170373217448434, + "loss": 2.5854, + "step": 23263 + }, + { + "epoch": 0.99, + "grad_norm": 0.5390625, + "learning_rate": 0.00010169702521275193, + "loss": 2.5568, + "step": 23264 + }, + { + "epoch": 0.99, + "grad_norm": 0.44140625, + "learning_rate": 0.00010169031824338353, + "loss": 2.5776, + "step": 23265 + }, + { + "epoch": 0.99, + "grad_norm": 0.515625, + "learning_rate": 0.00010168361126640931, + "loss": 2.6354, + "step": 23266 + }, + { + "epoch": 0.99, + "grad_norm": 0.515625, + "learning_rate": 0.00010167690428185947, + "loss": 2.6633, + "step": 23267 + }, + { + "epoch": 0.99, + "grad_norm": 0.478515625, + "learning_rate": 0.00010167019728976415, + "loss": 2.5721, + "step": 23268 + }, + { + "epoch": 0.99, + "grad_norm": 0.498046875, + "learning_rate": 0.00010166349029015356, + "loss": 2.5833, + "step": 23269 + }, + { + "epoch": 0.99, + "grad_norm": 0.490234375, + "learning_rate": 0.00010165678328305787, + "loss": 2.5571, + "step": 23270 + }, + { + "epoch": 0.99, + "grad_norm": 0.4609375, + "learning_rate": 0.00010165007626850726, + "loss": 2.4694, + "step": 23271 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010164336924653193, + "loss": 2.5629, + "step": 23272 + }, + { + "epoch": 0.99, + "grad_norm": 0.45703125, + "learning_rate": 0.00010163666221716202, + "loss": 2.6092, + "step": 23273 + }, + { + "epoch": 0.99, + "grad_norm": 0.43359375, + "learning_rate": 0.0001016299551804277, + "loss": 2.2654, + "step": 23274 + }, + { + "epoch": 0.99, + "grad_norm": 0.4921875, + "learning_rate": 0.00010162324813635924, + "loss": 2.4685, + "step": 23275 + }, + { + "epoch": 0.99, + "grad_norm": 0.47265625, + "learning_rate": 0.00010161654108498672, + "loss": 2.5195, + "step": 23276 + }, + { + "epoch": 0.99, + "grad_norm": 0.4140625, + "learning_rate": 0.00010160983402634035, + "loss": 2.4735, + "step": 23277 + }, + { + "epoch": 0.99, + "grad_norm": 0.48046875, + "learning_rate": 0.00010160312696045035, + "loss": 2.6327, + "step": 23278 + }, + { + "epoch": 0.99, + "grad_norm": 0.419921875, + "learning_rate": 0.00010159641988734682, + "loss": 2.5223, + "step": 23279 + }, + { + "epoch": 0.99, + "grad_norm": 0.439453125, + "learning_rate": 0.00010158971280706, + "loss": 2.6143, + "step": 23280 + }, + { + "epoch": 0.99, + "grad_norm": 0.44140625, + "learning_rate": 0.00010158300571962007, + "loss": 2.5453, + "step": 23281 + }, + { + "epoch": 0.99, + "grad_norm": 0.45703125, + "learning_rate": 0.00010157629862505718, + "loss": 2.5408, + "step": 23282 + }, + { + "epoch": 0.99, + "grad_norm": 0.455078125, + "learning_rate": 0.00010156959152340153, + "loss": 2.5554, + "step": 23283 + }, + { + "epoch": 0.99, + "grad_norm": 0.474609375, + "learning_rate": 0.0001015628844146833, + "loss": 2.7113, + "step": 23284 + }, + { + "epoch": 0.99, + "grad_norm": 0.44921875, + "learning_rate": 0.0001015561772989326, + "loss": 2.4733, + "step": 23285 + }, + { + "epoch": 0.99, + "grad_norm": 0.435546875, + "learning_rate": 0.00010154947017617975, + "loss": 2.5142, + "step": 23286 + }, + { + "epoch": 0.99, + "grad_norm": 0.41015625, + "learning_rate": 0.00010154276304645481, + "loss": 2.4631, + "step": 23287 + }, + { + "epoch": 0.99, + "grad_norm": 0.4296875, + "learning_rate": 0.00010153605590978801, + "loss": 2.5445, + "step": 23288 + }, + { + "epoch": 0.99, + "grad_norm": 0.44140625, + "learning_rate": 0.00010152934876620953, + "loss": 2.5356, + "step": 23289 + }, + { + "epoch": 0.99, + "grad_norm": 0.439453125, + "learning_rate": 0.00010152264161574951, + "loss": 2.4344, + "step": 23290 + }, + { + "epoch": 0.99, + "grad_norm": 0.462890625, + "learning_rate": 0.00010151593445843819, + "loss": 2.5807, + "step": 23291 + }, + { + "epoch": 0.99, + "grad_norm": 0.48046875, + "learning_rate": 0.0001015092272943057, + "loss": 2.59, + "step": 23292 + }, + { + "epoch": 0.99, + "grad_norm": 0.462890625, + "learning_rate": 0.00010150252012338222, + "loss": 2.528, + "step": 23293 + }, + { + "epoch": 0.99, + "grad_norm": 0.5234375, + "learning_rate": 0.00010149581294569799, + "loss": 2.638, + "step": 23294 + }, + { + "epoch": 0.99, + "grad_norm": 0.427734375, + "learning_rate": 0.00010148910576128312, + "loss": 2.6008, + "step": 23295 + }, + { + "epoch": 0.99, + "grad_norm": 0.49609375, + "learning_rate": 0.00010148239857016784, + "loss": 2.5942, + "step": 23296 + }, + { + "epoch": 0.99, + "grad_norm": 0.466796875, + "learning_rate": 0.00010147569137238229, + "loss": 2.6054, + "step": 23297 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010146898416795667, + "loss": 2.5656, + "step": 23298 + }, + { + "epoch": 0.99, + "grad_norm": 0.447265625, + "learning_rate": 0.00010146227695692117, + "loss": 2.4847, + "step": 23299 + }, + { + "epoch": 0.99, + "grad_norm": 0.44921875, + "learning_rate": 0.00010145556973930597, + "loss": 2.7183, + "step": 23300 + }, + { + "epoch": 0.99, + "grad_norm": 0.451171875, + "learning_rate": 0.00010144886251514119, + "loss": 2.484, + "step": 23301 + }, + { + "epoch": 0.99, + "grad_norm": 0.416015625, + "learning_rate": 0.00010144215528445711, + "loss": 2.4646, + "step": 23302 + }, + { + "epoch": 0.99, + "grad_norm": 0.470703125, + "learning_rate": 0.00010143544804728382, + "loss": 2.517, + "step": 23303 + }, + { + "epoch": 0.99, + "grad_norm": 0.50390625, + "learning_rate": 0.00010142874080365154, + "loss": 2.5591, + "step": 23304 + }, + { + "epoch": 0.99, + "grad_norm": 0.404296875, + "learning_rate": 0.00010142203355359048, + "loss": 2.6089, + "step": 23305 + }, + { + "epoch": 0.99, + "grad_norm": 0.50390625, + "learning_rate": 0.00010141532629713077, + "loss": 2.6394, + "step": 23306 + }, + { + "epoch": 0.99, + "grad_norm": 0.490234375, + "learning_rate": 0.00010140861903430261, + "loss": 2.579, + "step": 23307 + }, + { + "epoch": 0.99, + "grad_norm": 0.455078125, + "learning_rate": 0.00010140191176513618, + "loss": 2.5066, + "step": 23308 + }, + { + "epoch": 0.99, + "grad_norm": 0.490234375, + "learning_rate": 0.00010139520448966164, + "loss": 2.4742, + "step": 23309 + }, + { + "epoch": 0.99, + "grad_norm": 0.5078125, + "learning_rate": 0.00010138849720790924, + "loss": 2.6805, + "step": 23310 + }, + { + "epoch": 0.99, + "grad_norm": 0.466796875, + "learning_rate": 0.00010138178991990905, + "loss": 2.6086, + "step": 23311 + }, + { + "epoch": 0.99, + "grad_norm": 0.5, + "learning_rate": 0.00010137508262569135, + "loss": 2.6454, + "step": 23312 + }, + { + "epoch": 0.99, + "grad_norm": 0.41796875, + "learning_rate": 0.00010136837532528627, + "loss": 2.5542, + "step": 23313 + }, + { + "epoch": 0.99, + "grad_norm": 0.48828125, + "learning_rate": 0.000101361668018724, + "loss": 2.6873, + "step": 23314 + }, + { + "epoch": 0.99, + "grad_norm": 0.3984375, + "learning_rate": 0.00010135496070603474, + "loss": 2.5281, + "step": 23315 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.00010134825338724862, + "loss": 2.5905, + "step": 23316 + }, + { + "epoch": 0.99, + "grad_norm": 0.408203125, + "learning_rate": 0.00010134154606239585, + "loss": 2.5633, + "step": 23317 + }, + { + "epoch": 0.99, + "grad_norm": 0.47265625, + "learning_rate": 0.00010133483873150667, + "loss": 2.523, + "step": 23318 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010132813139461112, + "loss": 2.4741, + "step": 23319 + }, + { + "epoch": 0.99, + "grad_norm": 0.443359375, + "learning_rate": 0.00010132142405173952, + "loss": 2.3964, + "step": 23320 + }, + { + "epoch": 0.99, + "grad_norm": 0.462890625, + "learning_rate": 0.00010131471670292199, + "loss": 2.5884, + "step": 23321 + }, + { + "epoch": 0.99, + "grad_norm": 0.4296875, + "learning_rate": 0.00010130800934818871, + "loss": 2.4562, + "step": 23322 + }, + { + "epoch": 0.99, + "grad_norm": 0.40625, + "learning_rate": 0.00010130130198756986, + "loss": 2.535, + "step": 23323 + }, + { + "epoch": 0.99, + "grad_norm": 0.4453125, + "learning_rate": 0.00010129459462109563, + "loss": 2.3727, + "step": 23324 + }, + { + "epoch": 0.99, + "grad_norm": 0.49609375, + "learning_rate": 0.00010128788724879618, + "loss": 2.6007, + "step": 23325 + }, + { + "epoch": 0.99, + "grad_norm": 0.42578125, + "learning_rate": 0.00010128117987070176, + "loss": 2.5119, + "step": 23326 + }, + { + "epoch": 0.99, + "grad_norm": 0.421875, + "learning_rate": 0.00010127447248684244, + "loss": 2.5169, + "step": 23327 + }, + { + "epoch": 0.99, + "grad_norm": 0.4609375, + "learning_rate": 0.00010126776509724849, + "loss": 2.4776, + "step": 23328 + }, + { + "epoch": 0.99, + "grad_norm": 0.453125, + "learning_rate": 0.00010126105770195005, + "loss": 2.6161, + "step": 23329 + }, + { + "epoch": 0.99, + "grad_norm": 0.443359375, + "learning_rate": 0.00010125435030097732, + "loss": 2.5028, + "step": 23330 + }, + { + "epoch": 0.99, + "grad_norm": 0.44921875, + "learning_rate": 0.00010124764289436048, + "loss": 2.4284, + "step": 23331 + }, + { + "epoch": 0.99, + "grad_norm": 0.404296875, + "learning_rate": 0.0001012409354821297, + "loss": 2.533, + "step": 23332 + }, + { + "epoch": 0.99, + "grad_norm": 0.482421875, + "learning_rate": 0.00010123422806431513, + "loss": 2.6035, + "step": 23333 + }, + { + "epoch": 0.99, + "grad_norm": 0.470703125, + "learning_rate": 0.00010122752064094703, + "loss": 2.574, + "step": 23334 + }, + { + "epoch": 0.99, + "grad_norm": 0.443359375, + "learning_rate": 0.0001012208132120555, + "loss": 2.5856, + "step": 23335 + }, + { + "epoch": 0.99, + "grad_norm": 0.427734375, + "learning_rate": 0.00010121410577767076, + "loss": 2.4748, + "step": 23336 + }, + { + "epoch": 0.99, + "grad_norm": 0.466796875, + "learning_rate": 0.000101207398337823, + "loss": 2.4799, + "step": 23337 + }, + { + "epoch": 0.99, + "grad_norm": 0.439453125, + "learning_rate": 0.0001012006908925424, + "loss": 2.4144, + "step": 23338 + }, + { + "epoch": 0.99, + "grad_norm": 0.470703125, + "learning_rate": 0.00010119398344185911, + "loss": 2.6762, + "step": 23339 + }, + { + "epoch": 0.99, + "grad_norm": 0.46484375, + "learning_rate": 0.00010118727598580335, + "loss": 2.6631, + "step": 23340 + }, + { + "epoch": 0.99, + "grad_norm": 0.42578125, + "learning_rate": 0.00010118056852440525, + "loss": 2.5334, + "step": 23341 + }, + { + "epoch": 0.99, + "grad_norm": 0.447265625, + "learning_rate": 0.00010117386105769506, + "loss": 2.6036, + "step": 23342 + }, + { + "epoch": 0.99, + "grad_norm": 0.482421875, + "learning_rate": 0.0001011671535857029, + "loss": 2.448, + "step": 23343 + }, + { + "epoch": 0.99, + "grad_norm": 0.46875, + "learning_rate": 0.00010116044610845898, + "loss": 2.5421, + "step": 23344 + }, + { + "epoch": 0.99, + "grad_norm": 0.45703125, + "learning_rate": 0.00010115373862599348, + "loss": 2.5619, + "step": 23345 + }, + { + "epoch": 0.99, + "grad_norm": 0.404296875, + "learning_rate": 0.00010114703113833659, + "loss": 2.5107, + "step": 23346 + }, + { + "epoch": 0.99, + "grad_norm": 0.5390625, + "learning_rate": 0.00010114032364551846, + "loss": 2.5679, + "step": 23347 + }, + { + "epoch": 0.99, + "grad_norm": 0.43359375, + "learning_rate": 0.00010113361614756931, + "loss": 2.4744, + "step": 23348 + }, + { + "epoch": 0.99, + "grad_norm": 0.45703125, + "learning_rate": 0.00010112690864451925, + "loss": 2.5527, + "step": 23349 + }, + { + "epoch": 1.0, + "grad_norm": 0.498046875, + "learning_rate": 0.0001011202011363986, + "loss": 2.6751, + "step": 23350 + }, + { + "epoch": 1.0, + "grad_norm": 0.44140625, + "learning_rate": 0.00010111349362323737, + "loss": 2.619, + "step": 23351 + }, + { + "epoch": 1.0, + "grad_norm": 0.443359375, + "learning_rate": 0.00010110678610506588, + "loss": 2.4602, + "step": 23352 + }, + { + "epoch": 1.0, + "grad_norm": 0.421875, + "learning_rate": 0.00010110007858191424, + "loss": 2.4987, + "step": 23353 + }, + { + "epoch": 1.0, + "grad_norm": 0.49609375, + "learning_rate": 0.00010109337105381265, + "loss": 2.4741, + "step": 23354 + }, + { + "epoch": 1.0, + "grad_norm": 0.466796875, + "learning_rate": 0.00010108666352079128, + "loss": 2.4169, + "step": 23355 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 0.00010107995598288034, + "loss": 2.4553, + "step": 23356 + }, + { + "epoch": 1.0, + "grad_norm": 0.466796875, + "learning_rate": 0.00010107324844010994, + "loss": 2.5045, + "step": 23357 + }, + { + "epoch": 1.0, + "grad_norm": 0.498046875, + "learning_rate": 0.0001010665408925104, + "loss": 2.5277, + "step": 23358 + }, + { + "epoch": 1.0, + "grad_norm": 0.5078125, + "learning_rate": 0.00010105983334011176, + "loss": 2.5922, + "step": 23359 + }, + { + "epoch": 1.0, + "grad_norm": 0.40625, + "learning_rate": 0.00010105312578294426, + "loss": 2.5415, + "step": 23360 + }, + { + "epoch": 1.0, + "grad_norm": 0.41796875, + "learning_rate": 0.00010104641822103809, + "loss": 2.4093, + "step": 23361 + }, + { + "epoch": 1.0, + "grad_norm": 0.48046875, + "learning_rate": 0.00010103971065442342, + "loss": 2.5391, + "step": 23362 + }, + { + "epoch": 1.0, + "grad_norm": 0.439453125, + "learning_rate": 0.00010103300308313042, + "loss": 2.4208, + "step": 23363 + }, + { + "epoch": 1.0, + "grad_norm": 0.462890625, + "learning_rate": 0.0001010262955071893, + "loss": 2.4566, + "step": 23364 + }, + { + "epoch": 1.0, + "grad_norm": 0.40625, + "learning_rate": 0.0001010195879266302, + "loss": 2.6322, + "step": 23365 + }, + { + "epoch": 1.0, + "grad_norm": 0.5390625, + "learning_rate": 0.00010101288034148337, + "loss": 2.615, + "step": 23366 + }, + { + "epoch": 1.0, + "grad_norm": 0.45703125, + "learning_rate": 0.00010100617275177891, + "loss": 2.564, + "step": 23367 + }, + { + "epoch": 1.0, + "grad_norm": 0.4296875, + "learning_rate": 0.00010099946515754706, + "loss": 2.5347, + "step": 23368 + }, + { + "epoch": 1.0, + "grad_norm": 0.47265625, + "learning_rate": 0.00010099275755881796, + "loss": 2.6223, + "step": 23369 + }, + { + "epoch": 1.0, + "grad_norm": 0.412109375, + "learning_rate": 0.00010098604995562185, + "loss": 2.5733, + "step": 23370 + }, + { + "epoch": 1.0, + "grad_norm": 0.48046875, + "learning_rate": 0.00010097934234798885, + "loss": 2.5219, + "step": 23371 + }, + { + "epoch": 1.0, + "grad_norm": 0.45703125, + "learning_rate": 0.00010097263473594918, + "loss": 2.5748, + "step": 23372 + }, + { + "epoch": 1.0, + "grad_norm": 0.49609375, + "learning_rate": 0.000100965927119533, + "loss": 2.5625, + "step": 23373 + }, + { + "epoch": 1.0, + "grad_norm": 0.421875, + "learning_rate": 0.00010095921949877051, + "loss": 2.5193, + "step": 23374 + }, + { + "epoch": 1.0, + "grad_norm": 0.4296875, + "learning_rate": 0.00010095251187369186, + "loss": 2.5713, + "step": 23375 + }, + { + "epoch": 1.0, + "grad_norm": 0.427734375, + "learning_rate": 0.00010094580424432727, + "loss": 2.4325, + "step": 23376 + }, + { + "epoch": 1.0, + "grad_norm": 0.49609375, + "learning_rate": 0.0001009390966107069, + "loss": 2.5332, + "step": 23377 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 0.00010093238897286095, + "loss": 2.5799, + "step": 23378 + }, + { + "epoch": 1.0, + "grad_norm": 0.39453125, + "learning_rate": 0.00010092568133081959, + "loss": 2.5346, + "step": 23379 + }, + { + "epoch": 1.0, + "grad_norm": 0.43359375, + "learning_rate": 0.00010091897368461301, + "loss": 2.4999, + "step": 23380 + }, + { + "epoch": 1.0, + "grad_norm": 0.4921875, + "learning_rate": 0.00010091226603427135, + "loss": 2.5379, + "step": 23381 + }, + { + "epoch": 1.0, + "grad_norm": 0.4609375, + "learning_rate": 0.00010090555837982488, + "loss": 2.4512, + "step": 23382 + }, + { + "epoch": 1.0, + "grad_norm": 0.439453125, + "learning_rate": 0.00010089885072130367, + "loss": 2.543, + "step": 23383 + }, + { + "epoch": 1.0, + "grad_norm": 0.48046875, + "learning_rate": 0.000100892143058738, + "loss": 2.6844, + "step": 23384 + }, + { + "epoch": 1.0, + "grad_norm": 0.4296875, + "learning_rate": 0.000100885435392158, + "loss": 2.5288, + "step": 23385 + }, + { + "epoch": 1.0, + "grad_norm": 0.4296875, + "learning_rate": 0.00010087872772159386, + "loss": 2.6243, + "step": 23386 + }, + { + "epoch": 1.0, + "grad_norm": 0.498046875, + "learning_rate": 0.00010087202004707577, + "loss": 2.6318, + "step": 23387 + }, + { + "epoch": 1.0, + "grad_norm": 0.45703125, + "learning_rate": 0.0001008653123686339, + "loss": 2.5683, + "step": 23388 + }, + { + "epoch": 1.0, + "grad_norm": 0.45703125, + "learning_rate": 0.00010085860468629844, + "loss": 2.547, + "step": 23389 + }, + { + "epoch": 1.0, + "grad_norm": 0.466796875, + "learning_rate": 0.0001008518970000996, + "loss": 2.6381, + "step": 23390 + }, + { + "epoch": 1.0, + "grad_norm": 0.458984375, + "learning_rate": 0.0001008451893100675, + "loss": 2.5295, + "step": 23391 + }, + { + "epoch": 1.0, + "grad_norm": 0.48828125, + "learning_rate": 0.00010083848161623237, + "loss": 2.6347, + "step": 23392 + }, + { + "epoch": 1.0, + "grad_norm": 0.44140625, + "learning_rate": 0.0001008317739186244, + "loss": 2.4228, + "step": 23393 + }, + { + "epoch": 1.0, + "grad_norm": 0.4296875, + "learning_rate": 0.00010082506621727374, + "loss": 2.5647, + "step": 23394 + }, + { + "epoch": 1.0, + "grad_norm": 0.43359375, + "learning_rate": 0.00010081835851221057, + "loss": 2.5073, + "step": 23395 + }, + { + "epoch": 1.0, + "grad_norm": 0.45703125, + "learning_rate": 0.00010081165080346511, + "loss": 2.5281, + "step": 23396 + }, + { + "epoch": 1.0, + "grad_norm": 0.388671875, + "learning_rate": 0.0001008049430910675, + "loss": 2.5004, + "step": 23397 + }, + { + "epoch": 1.0, + "grad_norm": 0.40234375, + "learning_rate": 0.00010079823537504794, + "loss": 2.6039, + "step": 23398 + }, + { + "epoch": 1.0, + "grad_norm": 0.45703125, + "learning_rate": 0.00010079152765543663, + "loss": 2.6074, + "step": 23399 + }, + { + "epoch": 1.0, + "grad_norm": 0.458984375, + "learning_rate": 0.0001007848199322637, + "loss": 2.6889, + "step": 23400 + }, + { + "epoch": 1.0, + "grad_norm": 0.4453125, + "learning_rate": 0.0001007781122055594, + "loss": 2.5351, + "step": 23401 + }, + { + "epoch": 1.0, + "grad_norm": 0.4296875, + "learning_rate": 0.00010077140447535385, + "loss": 2.5329, + "step": 23402 + }, + { + "epoch": 1.0, + "grad_norm": 0.4453125, + "learning_rate": 0.00010076469674167728, + "loss": 2.4466, + "step": 23403 + }, + { + "epoch": 1.0, + "grad_norm": 0.455078125, + "learning_rate": 0.00010075798900455986, + "loss": 2.6564, + "step": 23404 + }, + { + "epoch": 1.0, + "grad_norm": 0.423828125, + "learning_rate": 0.00010075128126403174, + "loss": 2.5243, + "step": 23405 + }, + { + "epoch": 1.0, + "grad_norm": 0.447265625, + "learning_rate": 0.00010074457352012317, + "loss": 2.607, + "step": 23406 + }, + { + "epoch": 1.0, + "grad_norm": 0.435546875, + "learning_rate": 0.00010073786577286426, + "loss": 2.6038, + "step": 23407 + }, + { + "epoch": 1.0, + "grad_norm": 0.4609375, + "learning_rate": 0.00010073115802228521, + "loss": 2.4522, + "step": 23408 + }, + { + "epoch": 1.0, + "grad_norm": 0.470703125, + "learning_rate": 0.00010072445026841625, + "loss": 2.5367, + "step": 23409 + }, + { + "epoch": 1.0, + "grad_norm": 0.4765625, + "learning_rate": 0.00010071774251128751, + "loss": 2.5261, + "step": 23410 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 0.0001007110347509292, + "loss": 2.6263, + "step": 23411 + }, + { + "epoch": 1.0, + "grad_norm": 0.443359375, + "learning_rate": 0.0001007043269873715, + "loss": 2.5707, + "step": 23412 + }, + { + "epoch": 1.0, + "grad_norm": 0.44921875, + "learning_rate": 0.00010069761922064458, + "loss": 2.6164, + "step": 23413 + }, + { + "epoch": 1.0, + "grad_norm": 0.451171875, + "learning_rate": 0.00010069091145077862, + "loss": 2.4238, + "step": 23414 + }, + { + "epoch": 1.0, + "grad_norm": 0.54296875, + "learning_rate": 0.00010068420367780382, + "loss": 2.6007, + "step": 23415 + }, + { + "epoch": 1.0, + "grad_norm": 0.408203125, + "learning_rate": 0.00010067749590175033, + "loss": 2.4781, + "step": 23416 + }, + { + "epoch": 1.0, + "grad_norm": 0.466796875, + "learning_rate": 0.00010067078812264841, + "loss": 2.5708, + "step": 23417 + }, + { + "epoch": 1.0, + "grad_norm": 0.48828125, + "learning_rate": 0.00010066408034052812, + "loss": 2.5794, + "step": 23418 + }, + { + "epoch": 1.0, + "grad_norm": 0.43359375, + "learning_rate": 0.00010065737255541976, + "loss": 2.5034, + "step": 23419 + }, + { + "epoch": 1.0, + "grad_norm": 0.421875, + "learning_rate": 0.00010065066476735345, + "loss": 2.5321, + "step": 23420 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 0.0001006439569763594, + "loss": 2.6419, + "step": 23421 + }, + { + "epoch": 1.0, + "grad_norm": 0.40234375, + "learning_rate": 0.00010063724918246776, + "loss": 2.4734, + "step": 23422 + }, + { + "epoch": 1.0, + "grad_norm": 0.458984375, + "learning_rate": 0.00010063054138570873, + "loss": 2.441, + "step": 23423 + }, + { + "epoch": 1.0, + "grad_norm": 0.447265625, + "learning_rate": 0.00010062383358611249, + "loss": 2.4888, + "step": 23424 + }, + { + "epoch": 1.0, + "grad_norm": 0.47265625, + "learning_rate": 0.00010061712578370926, + "loss": 2.6004, + "step": 23425 + }, + { + "epoch": 1.0, + "grad_norm": 0.412109375, + "learning_rate": 0.00010061041797852915, + "loss": 2.4759, + "step": 23426 + }, + { + "epoch": 1.0, + "grad_norm": 0.41015625, + "learning_rate": 0.00010060371017060239, + "loss": 2.554, + "step": 23427 + }, + { + "epoch": 1.0, + "grad_norm": 0.41015625, + "learning_rate": 0.00010059700235995916, + "loss": 2.4767, + "step": 23428 + }, + { + "epoch": 1.0, + "grad_norm": 0.455078125, + "learning_rate": 0.00010059029454662966, + "loss": 2.3969, + "step": 23429 + }, + { + "epoch": 1.0, + "grad_norm": 0.4921875, + "learning_rate": 0.00010058358673064403, + "loss": 2.469, + "step": 23430 + }, + { + "epoch": 1.0, + "grad_norm": 0.44140625, + "learning_rate": 0.00010057687891203249, + "loss": 2.63, + "step": 23431 + }, + { + "epoch": 1.0, + "grad_norm": 0.51171875, + "learning_rate": 0.00010057017109082516, + "loss": 2.54, + "step": 23432 + }, + { + "epoch": 1.0, + "grad_norm": 0.470703125, + "learning_rate": 0.00010056346326705233, + "loss": 2.402, + "step": 23433 + }, + { + "epoch": 1.0, + "grad_norm": 0.390625, + "learning_rate": 0.00010055675544074408, + "loss": 2.3763, + "step": 23434 + }, + { + "epoch": 1.0, + "grad_norm": 0.4609375, + "learning_rate": 0.00010055004761193062, + "loss": 2.4915, + "step": 23435 + }, + { + "epoch": 1.0, + "grad_norm": 0.4765625, + "learning_rate": 0.00010054333978064218, + "loss": 2.5707, + "step": 23436 + }, + { + "epoch": 1.0, + "grad_norm": 0.5078125, + "learning_rate": 0.00010053663194690892, + "loss": 2.4757, + "step": 23437 + }, + { + "epoch": 1.0, + "grad_norm": 0.498046875, + "learning_rate": 0.00010052992411076099, + "loss": 2.5192, + "step": 23438 + }, + { + "epoch": 1.0, + "grad_norm": 0.40625, + "learning_rate": 0.0001005232162722286, + "loss": 2.4536, + "step": 23439 + }, + { + "epoch": 1.0, + "grad_norm": 0.46875, + "learning_rate": 0.00010051650843134191, + "loss": 2.8198, + "step": 23440 + }, + { + "epoch": 1.0, + "grad_norm": 0.41015625, + "learning_rate": 0.00010050980058813116, + "loss": 2.5411, + "step": 23441 + }, + { + "epoch": 1.0, + "grad_norm": 0.4609375, + "learning_rate": 0.00010050309274262645, + "loss": 2.5957, + "step": 23442 + }, + { + "epoch": 1.0, + "grad_norm": 0.466796875, + "learning_rate": 0.00010049638489485802, + "loss": 2.5711, + "step": 23443 + }, + { + "epoch": 1.0, + "grad_norm": 0.453125, + "learning_rate": 0.00010048967704485606, + "loss": 2.5764, + "step": 23444 + }, + { + "epoch": 1.0, + "grad_norm": 0.431640625, + "learning_rate": 0.00010048296919265073, + "loss": 2.6594, + "step": 23445 + }, + { + "epoch": 1.0, + "grad_norm": 0.478515625, + "learning_rate": 0.0001004762613382722, + "loss": 2.6228, + "step": 23446 + }, + { + "epoch": 1.0, + "grad_norm": 0.455078125, + "learning_rate": 0.00010046955348175067, + "loss": 2.5357, + "step": 23447 + }, + { + "epoch": 1.0, + "grad_norm": 0.482421875, + "learning_rate": 0.00010046284562311629, + "loss": 2.5363, + "step": 23448 + }, + { + "epoch": 1.0, + "grad_norm": 0.435546875, + "learning_rate": 0.00010045613776239935, + "loss": 2.5589, + "step": 23449 + }, + { + "epoch": 1.0, + "grad_norm": 0.416015625, + "learning_rate": 0.00010044942989962989, + "loss": 2.5265, + "step": 23450 + }, + { + "epoch": 1.0, + "grad_norm": 0.421875, + "learning_rate": 0.00010044272203483818, + "loss": 2.5572, + "step": 23451 + }, + { + "epoch": 1.0, + "grad_norm": 0.451171875, + "learning_rate": 0.0001004360141680544, + "loss": 2.5455, + "step": 23452 + }, + { + "epoch": 1.0, + "grad_norm": 0.43359375, + "learning_rate": 0.0001004293062993087, + "loss": 2.2308, + "step": 23453 + }, + { + "epoch": 1.0, + "grad_norm": 0.4375, + "learning_rate": 0.00010042259842863128, + "loss": 2.6207, + "step": 23454 + }, + { + "epoch": 1.0, + "grad_norm": 0.466796875, + "learning_rate": 0.00010041589055605232, + "loss": 2.5421, + "step": 23455 + }, + { + "epoch": 1.0, + "grad_norm": 0.43359375, + "learning_rate": 0.000100409182681602, + "loss": 2.4753, + "step": 23456 + }, + { + "epoch": 1.0, + "grad_norm": 0.48828125, + "learning_rate": 0.00010040247480531053, + "loss": 2.4071, + "step": 23457 + }, + { + "epoch": 1.0, + "grad_norm": 0.515625, + "learning_rate": 0.00010039576692720805, + "loss": 2.6906, + "step": 23458 + }, + { + "epoch": 1.0, + "grad_norm": 0.42578125, + "learning_rate": 0.00010038905904732476, + "loss": 2.4664, + "step": 23459 + }, + { + "epoch": 1.0, + "grad_norm": 0.482421875, + "learning_rate": 0.00010038235116569086, + "loss": 2.5753, + "step": 23460 + }, + { + "epoch": 1.0, + "grad_norm": 0.412109375, + "learning_rate": 0.00010037564328233651, + "loss": 2.5479, + "step": 23461 + }, + { + "epoch": 1.0, + "grad_norm": 0.478515625, + "learning_rate": 0.0001003689353972919, + "loss": 2.5655, + "step": 23462 + }, + { + "epoch": 1.0, + "grad_norm": 0.4453125, + "learning_rate": 0.00010036222751058723, + "loss": 2.5011, + "step": 23463 + }, + { + "epoch": 1.0, + "grad_norm": 0.423828125, + "learning_rate": 0.00010035551962225265, + "loss": 2.5394, + "step": 23464 + }, + { + "epoch": 1.0, + "grad_norm": 0.42578125, + "learning_rate": 0.00010034881173231839, + "loss": 2.5387, + "step": 23465 + }, + { + "epoch": 1.0, + "grad_norm": 0.4453125, + "learning_rate": 0.00010034210384081458, + "loss": 2.4938, + "step": 23466 + }, + { + "epoch": 1.0, + "grad_norm": 0.51953125, + "learning_rate": 0.00010033539594777142, + "loss": 2.5845, + "step": 23467 + }, + { + "epoch": 1.0, + "grad_norm": 0.416015625, + "learning_rate": 0.00010032868805321911, + "loss": 2.532, + "step": 23468 + }, + { + "epoch": 1.0, + "eval_loss": 2.5235140323638916, + "eval_runtime": 82.4214, + "eval_samples_per_second": 18.248, + "eval_steps_per_second": 2.281, + "step": 23468 + }, + { + "epoch": 1.0, + "grad_norm": 0.478515625, + "learning_rate": 0.00010032198015718784, + "loss": 2.4416, + "step": 23469 + }, + { + "epoch": 1.0, + "grad_norm": 0.384765625, + "learning_rate": 0.00010031527225970776, + "loss": 2.4076, + "step": 23470 + }, + { + "epoch": 1.0, + "grad_norm": 0.453125, + "learning_rate": 0.00010030856436080909, + "loss": 2.4153, + "step": 23471 + }, + { + "epoch": 1.0, + "grad_norm": 0.421875, + "learning_rate": 0.00010030185646052194, + "loss": 2.3556, + "step": 23472 + }, + { + "epoch": 1.0, + "grad_norm": 0.416015625, + "learning_rate": 0.00010029514855887663, + "loss": 2.3711, + "step": 23473 + }, + { + "epoch": 1.0, + "grad_norm": 0.44140625, + "learning_rate": 0.00010028844065590319, + "loss": 2.4819, + "step": 23474 + }, + { + "epoch": 1.0, + "grad_norm": 0.4921875, + "learning_rate": 0.00010028173275163191, + "loss": 2.4433, + "step": 23475 + }, + { + "epoch": 1.0, + "grad_norm": 0.447265625, + "learning_rate": 0.00010027502484609293, + "loss": 2.5178, + "step": 23476 + }, + { + "epoch": 1.0, + "grad_norm": 0.42578125, + "learning_rate": 0.00010026831693931642, + "loss": 2.5236, + "step": 23477 + }, + { + "epoch": 1.0, + "grad_norm": 0.470703125, + "learning_rate": 0.00010026160903133259, + "loss": 2.3521, + "step": 23478 + }, + { + "epoch": 1.0, + "grad_norm": 0.4921875, + "learning_rate": 0.00010025490112217163, + "loss": 2.4952, + "step": 23479 + }, + { + "epoch": 1.0, + "grad_norm": 0.470703125, + "learning_rate": 0.00010024819321186367, + "loss": 2.4485, + "step": 23480 + }, + { + "epoch": 1.0, + "grad_norm": 0.453125, + "learning_rate": 0.00010024148530043897, + "loss": 2.338, + "step": 23481 + }, + { + "epoch": 1.0, + "grad_norm": 0.4375, + "learning_rate": 0.00010023477738792766, + "loss": 2.4657, + "step": 23482 + }, + { + "epoch": 1.0, + "grad_norm": 0.419921875, + "learning_rate": 0.00010022806947435994, + "loss": 2.385, + "step": 23483 + }, + { + "epoch": 1.0, + "grad_norm": 0.43359375, + "learning_rate": 0.00010022136155976597, + "loss": 2.3956, + "step": 23484 + }, + { + "epoch": 1.0, + "grad_norm": 0.447265625, + "learning_rate": 0.00010021465364417599, + "loss": 2.4977, + "step": 23485 + }, + { + "epoch": 1.0, + "grad_norm": 0.51171875, + "learning_rate": 0.00010020794572762013, + "loss": 2.2801, + "step": 23486 + }, + { + "epoch": 1.0, + "grad_norm": 0.462890625, + "learning_rate": 0.0001002012378101286, + "loss": 2.5413, + "step": 23487 + }, + { + "epoch": 1.0, + "grad_norm": 0.427734375, + "learning_rate": 0.00010019452989173155, + "loss": 2.4194, + "step": 23488 + }, + { + "epoch": 1.0, + "grad_norm": 0.5234375, + "learning_rate": 0.00010018782197245922, + "loss": 2.4982, + "step": 23489 + }, + { + "epoch": 1.0, + "grad_norm": 0.458984375, + "learning_rate": 0.00010018111405234172, + "loss": 2.3837, + "step": 23490 + }, + { + "epoch": 1.0, + "grad_norm": 0.515625, + "learning_rate": 0.0001001744061314093, + "loss": 2.5083, + "step": 23491 + }, + { + "epoch": 1.0, + "grad_norm": 0.423828125, + "learning_rate": 0.00010016769820969212, + "loss": 2.6237, + "step": 23492 + }, + { + "epoch": 1.0, + "grad_norm": 0.48828125, + "learning_rate": 0.00010016099028722035, + "loss": 2.5597, + "step": 23493 + }, + { + "epoch": 1.0, + "grad_norm": 0.51953125, + "learning_rate": 0.00010015428236402416, + "loss": 2.6014, + "step": 23494 + }, + { + "epoch": 1.0, + "grad_norm": 0.435546875, + "learning_rate": 0.00010014757444013381, + "loss": 2.275, + "step": 23495 + }, + { + "epoch": 1.0, + "grad_norm": 0.49609375, + "learning_rate": 0.00010014086651557937, + "loss": 2.6574, + "step": 23496 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 0.00010013415859039112, + "loss": 2.5118, + "step": 23497 + }, + { + "epoch": 1.0, + "grad_norm": 0.447265625, + "learning_rate": 0.00010012745066459918, + "loss": 2.6077, + "step": 23498 + }, + { + "epoch": 1.0, + "grad_norm": 0.5234375, + "learning_rate": 0.00010012074273823377, + "loss": 2.4713, + "step": 23499 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 0.00010011403481132508, + "loss": 2.3758, + "step": 23500 + }, + { + "epoch": 1.0, + "grad_norm": 0.435546875, + "learning_rate": 0.00010010732688390326, + "loss": 2.5504, + "step": 23501 + }, + { + "epoch": 1.0, + "grad_norm": 0.451171875, + "learning_rate": 0.0001001006189559985, + "loss": 2.4423, + "step": 23502 + }, + { + "epoch": 1.0, + "grad_norm": 0.462890625, + "learning_rate": 0.00010009391102764101, + "loss": 2.481, + "step": 23503 + }, + { + "epoch": 1.0, + "grad_norm": 0.51171875, + "learning_rate": 0.00010008720309886092, + "loss": 2.5331, + "step": 23504 + }, + { + "epoch": 1.0, + "grad_norm": 0.42578125, + "learning_rate": 0.00010008049516968851, + "loss": 2.4212, + "step": 23505 + }, + { + "epoch": 1.0, + "grad_norm": 0.451171875, + "learning_rate": 0.00010007378724015383, + "loss": 2.3929, + "step": 23506 + }, + { + "epoch": 1.0, + "grad_norm": 0.439453125, + "learning_rate": 0.00010006707931028718, + "loss": 2.4953, + "step": 23507 + }, + { + "epoch": 1.0, + "grad_norm": 0.49609375, + "learning_rate": 0.00010006037138011868, + "loss": 2.4164, + "step": 23508 + }, + { + "epoch": 1.0, + "grad_norm": 0.515625, + "learning_rate": 0.00010005366344967855, + "loss": 2.528, + "step": 23509 + }, + { + "epoch": 1.0, + "grad_norm": 0.431640625, + "learning_rate": 0.00010004695551899692, + "loss": 2.4981, + "step": 23510 + }, + { + "epoch": 1.0, + "grad_norm": 0.46484375, + "learning_rate": 0.00010004024758810404, + "loss": 2.5546, + "step": 23511 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 0.00010003353965703004, + "loss": 2.5003, + "step": 23512 + }, + { + "epoch": 1.0, + "grad_norm": 0.427734375, + "learning_rate": 0.00010002683172580517, + "loss": 2.4013, + "step": 23513 + }, + { + "epoch": 1.0, + "grad_norm": 0.462890625, + "learning_rate": 0.00010002012379445949, + "loss": 2.3601, + "step": 23514 + }, + { + "epoch": 1.0, + "grad_norm": 0.51171875, + "learning_rate": 0.0001000134158630233, + "loss": 2.5579, + "step": 23515 + }, + { + "epoch": 1.0, + "grad_norm": 0.48828125, + "learning_rate": 0.00010000670793152676, + "loss": 2.4221, + "step": 23516 + }, + { + "epoch": 1.0, + "grad_norm": 0.43359375, + "learning_rate": 0.0001, + "loss": 2.5197, + "step": 23517 + }, + { + "epoch": 1.0, + "grad_norm": 0.453125, + "learning_rate": 9.999329206847328e-05, + "loss": 2.5627, + "step": 23518 + }, + { + "epoch": 1.0, + "grad_norm": 0.44921875, + "learning_rate": 9.998658413697673e-05, + "loss": 2.5432, + "step": 23519 + }, + { + "epoch": 1.0, + "grad_norm": 0.455078125, + "learning_rate": 9.997987620554051e-05, + "loss": 2.5697, + "step": 23520 + }, + { + "epoch": 1.0, + "grad_norm": 0.5234375, + "learning_rate": 9.997316827419487e-05, + "loss": 2.4003, + "step": 23521 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 9.996646034296999e-05, + "loss": 2.5035, + "step": 23522 + }, + { + "epoch": 1.0, + "grad_norm": 0.49609375, + "learning_rate": 9.9959752411896e-05, + "loss": 2.4699, + "step": 23523 + }, + { + "epoch": 1.0, + "grad_norm": 0.451171875, + "learning_rate": 9.995304448100309e-05, + "loss": 2.3637, + "step": 23524 + }, + { + "epoch": 1.0, + "grad_norm": 0.5, + "learning_rate": 9.994633655032147e-05, + "loss": 2.5812, + "step": 23525 + }, + { + "epoch": 1.0, + "grad_norm": 0.46875, + "learning_rate": 9.993962861988133e-05, + "loss": 2.4855, + "step": 23526 + }, + { + "epoch": 1.0, + "grad_norm": 0.462890625, + "learning_rate": 9.993292068971286e-05, + "loss": 2.4411, + "step": 23527 + }, + { + "epoch": 1.0, + "grad_norm": 0.4609375, + "learning_rate": 9.992621275984616e-05, + "loss": 2.3466, + "step": 23528 + }, + { + "epoch": 1.0, + "grad_norm": 0.43359375, + "learning_rate": 9.991950483031152e-05, + "loss": 2.4127, + "step": 23529 + }, + { + "epoch": 1.0, + "grad_norm": 0.44921875, + "learning_rate": 9.99127969011391e-05, + "loss": 2.4617, + "step": 23530 + }, + { + "epoch": 1.0, + "grad_norm": 0.466796875, + "learning_rate": 9.990608897235902e-05, + "loss": 2.4809, + "step": 23531 + }, + { + "epoch": 1.0, + "grad_norm": 0.484375, + "learning_rate": 9.98993810440015e-05, + "loss": 2.462, + "step": 23532 + }, + { + "epoch": 1.0, + "grad_norm": 0.51171875, + "learning_rate": 9.989267311609675e-05, + "loss": 2.4136, + "step": 23533 + }, + { + "epoch": 1.0, + "grad_norm": 0.5234375, + "learning_rate": 9.988596518867495e-05, + "loss": 2.3646, + "step": 23534 + }, + { + "epoch": 1.0, + "grad_norm": 0.44140625, + "learning_rate": 9.987925726176626e-05, + "loss": 2.6204, + "step": 23535 + }, + { + "epoch": 1.0, + "grad_norm": 0.48046875, + "learning_rate": 9.987254933540081e-05, + "loss": 2.4627, + "step": 23536 + }, + { + "epoch": 1.0, + "grad_norm": 0.474609375, + "learning_rate": 9.986584140960889e-05, + "loss": 2.4607, + "step": 23537 + }, + { + "epoch": 1.0, + "grad_norm": 0.453125, + "learning_rate": 9.985913348442065e-05, + "loss": 2.4299, + "step": 23538 + }, + { + "epoch": 1.0, + "grad_norm": 0.431640625, + "learning_rate": 9.985242555986624e-05, + "loss": 2.3972, + "step": 23539 + }, + { + "epoch": 1.0, + "grad_norm": 0.44140625, + "learning_rate": 9.984571763597585e-05, + "loss": 2.4446, + "step": 23540 + }, + { + "epoch": 1.0, + "grad_norm": 0.451171875, + "learning_rate": 9.983900971277967e-05, + "loss": 2.5235, + "step": 23541 + }, + { + "epoch": 1.0, + "grad_norm": 0.478515625, + "learning_rate": 9.98323017903079e-05, + "loss": 2.505, + "step": 23542 + }, + { + "epoch": 1.0, + "grad_norm": 0.490234375, + "learning_rate": 9.982559386859074e-05, + "loss": 2.4366, + "step": 23543 + }, + { + "epoch": 1.0, + "grad_norm": 0.427734375, + "learning_rate": 9.981888594765827e-05, + "loss": 2.3633, + "step": 23544 + }, + { + "epoch": 1.0, + "grad_norm": 0.4609375, + "learning_rate": 9.981217802754079e-05, + "loss": 2.4206, + "step": 23545 + }, + { + "epoch": 1.0, + "grad_norm": 0.484375, + "learning_rate": 9.980547010826847e-05, + "loss": 2.5124, + "step": 23546 + }, + { + "epoch": 1.0, + "grad_norm": 0.46875, + "learning_rate": 9.979876218987145e-05, + "loss": 2.4939, + "step": 23547 + }, + { + "epoch": 1.0, + "grad_norm": 0.423828125, + "learning_rate": 9.979205427237988e-05, + "loss": 2.5872, + "step": 23548 + }, + { + "epoch": 1.0, + "grad_norm": 0.5078125, + "learning_rate": 9.978534635582403e-05, + "loss": 2.5229, + "step": 23549 + }, + { + "epoch": 1.0, + "grad_norm": 0.45703125, + "learning_rate": 9.977863844023404e-05, + "loss": 2.3267, + "step": 23550 + }, + { + "epoch": 1.0, + "grad_norm": 0.390625, + "learning_rate": 9.977193052564011e-05, + "loss": 2.4738, + "step": 23551 + }, + { + "epoch": 1.0, + "grad_norm": 0.4375, + "learning_rate": 9.976522261207234e-05, + "loss": 2.3292, + "step": 23552 + }, + { + "epoch": 1.0, + "grad_norm": 0.451171875, + "learning_rate": 9.975851469956104e-05, + "loss": 2.4766, + "step": 23553 + }, + { + "epoch": 1.0, + "grad_norm": 0.44140625, + "learning_rate": 9.975180678813635e-05, + "loss": 2.4037, + "step": 23554 + }, + { + "epoch": 1.0, + "grad_norm": 0.484375, + "learning_rate": 9.974509887782843e-05, + "loss": 2.5793, + "step": 23555 + }, + { + "epoch": 1.0, + "grad_norm": 0.458984375, + "learning_rate": 9.973839096866743e-05, + "loss": 2.392, + "step": 23556 + }, + { + "epoch": 1.0, + "grad_norm": 0.453125, + "learning_rate": 9.97316830606836e-05, + "loss": 2.4928, + "step": 23557 + }, + { + "epoch": 1.0, + "grad_norm": 0.447265625, + "learning_rate": 9.972497515390711e-05, + "loss": 2.3866, + "step": 23558 + }, + { + "epoch": 1.0, + "grad_norm": 0.5234375, + "learning_rate": 9.971826724836815e-05, + "loss": 2.4396, + "step": 23559 + }, + { + "epoch": 1.0, + "grad_norm": 0.439453125, + "learning_rate": 9.97115593440968e-05, + "loss": 2.3788, + "step": 23560 + }, + { + "epoch": 1.0, + "grad_norm": 0.498046875, + "learning_rate": 9.97048514411234e-05, + "loss": 2.4191, + "step": 23561 + }, + { + "epoch": 1.0, + "grad_norm": 0.45703125, + "learning_rate": 9.969814353947807e-05, + "loss": 2.3261, + "step": 23562 + }, + { + "epoch": 1.0, + "grad_norm": 0.4296875, + "learning_rate": 9.969143563919096e-05, + "loss": 2.4782, + "step": 23563 + }, + { + "epoch": 1.0, + "grad_norm": 0.46875, + "learning_rate": 9.968472774029225e-05, + "loss": 2.4714, + "step": 23564 + }, + { + "epoch": 1.0, + "grad_norm": 0.50390625, + "learning_rate": 9.967801984281218e-05, + "loss": 2.405, + "step": 23565 + }, + { + "epoch": 1.0, + "grad_norm": 0.453125, + "learning_rate": 9.96713119467809e-05, + "loss": 2.429, + "step": 23566 + }, + { + "epoch": 1.0, + "grad_norm": 0.5078125, + "learning_rate": 9.966460405222863e-05, + "loss": 2.4792, + "step": 23567 + }, + { + "epoch": 1.0, + "grad_norm": 0.5234375, + "learning_rate": 9.965789615918543e-05, + "loss": 2.5584, + "step": 23568 + }, + { + "epoch": 1.0, + "grad_norm": 0.515625, + "learning_rate": 9.965118826768164e-05, + "loss": 2.6328, + "step": 23569 + }, + { + "epoch": 1.0, + "grad_norm": 0.5078125, + "learning_rate": 9.964448037774738e-05, + "loss": 2.4534, + "step": 23570 + }, + { + "epoch": 1.0, + "grad_norm": 0.455078125, + "learning_rate": 9.963777248941282e-05, + "loss": 2.4305, + "step": 23571 + }, + { + "epoch": 1.0, + "grad_norm": 0.48046875, + "learning_rate": 9.96310646027081e-05, + "loss": 2.4388, + "step": 23572 + }, + { + "epoch": 1.0, + "grad_norm": 0.51171875, + "learning_rate": 9.962435671766351e-05, + "loss": 2.4572, + "step": 23573 + }, + { + "epoch": 1.0, + "grad_norm": 0.462890625, + "learning_rate": 9.961764883430918e-05, + "loss": 2.5413, + "step": 23574 + }, + { + "epoch": 1.0, + "grad_norm": 0.49609375, + "learning_rate": 9.961094095267526e-05, + "loss": 2.5742, + "step": 23575 + }, + { + "epoch": 1.0, + "grad_norm": 0.462890625, + "learning_rate": 9.960423307279196e-05, + "loss": 2.5105, + "step": 23576 + }, + { + "epoch": 1.0, + "grad_norm": 0.455078125, + "learning_rate": 9.959752519468948e-05, + "loss": 2.3689, + "step": 23577 + }, + { + "epoch": 1.0, + "grad_norm": 0.435546875, + "learning_rate": 9.959081731839803e-05, + "loss": 2.4714, + "step": 23578 + }, + { + "epoch": 1.0, + "grad_norm": 0.50390625, + "learning_rate": 9.958410944394772e-05, + "loss": 2.4715, + "step": 23579 + }, + { + "epoch": 1.0, + "grad_norm": 0.498046875, + "learning_rate": 9.957740157136874e-05, + "loss": 2.59, + "step": 23580 + }, + { + "epoch": 1.0, + "grad_norm": 0.49609375, + "learning_rate": 9.957069370069132e-05, + "loss": 2.6403, + "step": 23581 + }, + { + "epoch": 1.0, + "grad_norm": 0.5078125, + "learning_rate": 9.956398583194563e-05, + "loss": 2.4332, + "step": 23582 + }, + { + "epoch": 1.0, + "grad_norm": 0.515625, + "learning_rate": 9.955727796516183e-05, + "loss": 2.4524, + "step": 23583 + }, + { + "epoch": 1.0, + "grad_norm": 0.412109375, + "learning_rate": 9.95505701003701e-05, + "loss": 2.5382, + "step": 23584 + }, + { + "epoch": 1.01, + "grad_norm": 0.490234375, + "learning_rate": 9.954386223760067e-05, + "loss": 2.5603, + "step": 23585 + }, + { + "epoch": 1.01, + "grad_norm": 0.4453125, + "learning_rate": 9.953715437688372e-05, + "loss": 2.3969, + "step": 23586 + }, + { + "epoch": 1.01, + "grad_norm": 0.46875, + "learning_rate": 9.953044651824938e-05, + "loss": 2.5237, + "step": 23587 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.952373866172783e-05, + "loss": 2.4519, + "step": 23588 + }, + { + "epoch": 1.01, + "grad_norm": 0.494140625, + "learning_rate": 9.95170308073493e-05, + "loss": 2.4833, + "step": 23589 + }, + { + "epoch": 1.01, + "grad_norm": 0.45703125, + "learning_rate": 9.951032295514397e-05, + "loss": 2.5164, + "step": 23590 + }, + { + "epoch": 1.01, + "grad_norm": 0.5625, + "learning_rate": 9.950361510514199e-05, + "loss": 2.442, + "step": 23591 + }, + { + "epoch": 1.01, + "grad_norm": 0.412109375, + "learning_rate": 9.949690725737355e-05, + "loss": 2.4029, + "step": 23592 + }, + { + "epoch": 1.01, + "grad_norm": 0.49609375, + "learning_rate": 9.949019941186886e-05, + "loss": 2.4675, + "step": 23593 + }, + { + "epoch": 1.01, + "grad_norm": 0.431640625, + "learning_rate": 9.948349156865812e-05, + "loss": 2.4652, + "step": 23594 + }, + { + "epoch": 1.01, + "grad_norm": 0.478515625, + "learning_rate": 9.947678372777144e-05, + "loss": 2.5477, + "step": 23595 + }, + { + "epoch": 1.01, + "grad_norm": 0.5078125, + "learning_rate": 9.947007588923903e-05, + "loss": 2.6268, + "step": 23596 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.946336805309112e-05, + "loss": 2.3654, + "step": 23597 + }, + { + "epoch": 1.01, + "grad_norm": 0.46484375, + "learning_rate": 9.945666021935784e-05, + "loss": 2.4337, + "step": 23598 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.944995238806939e-05, + "loss": 2.3508, + "step": 23599 + }, + { + "epoch": 1.01, + "grad_norm": 0.4921875, + "learning_rate": 9.944324455925593e-05, + "loss": 2.4232, + "step": 23600 + }, + { + "epoch": 1.01, + "grad_norm": 0.490234375, + "learning_rate": 9.94365367329477e-05, + "loss": 2.5135, + "step": 23601 + }, + { + "epoch": 1.01, + "grad_norm": 0.408203125, + "learning_rate": 9.942982890917486e-05, + "loss": 2.435, + "step": 23602 + }, + { + "epoch": 1.01, + "grad_norm": 0.443359375, + "learning_rate": 9.942312108796756e-05, + "loss": 2.5208, + "step": 23603 + }, + { + "epoch": 1.01, + "grad_norm": 0.458984375, + "learning_rate": 9.941641326935598e-05, + "loss": 2.3868, + "step": 23604 + }, + { + "epoch": 1.01, + "grad_norm": 0.478515625, + "learning_rate": 9.940970545337036e-05, + "loss": 2.5483, + "step": 23605 + }, + { + "epoch": 1.01, + "grad_norm": 0.48828125, + "learning_rate": 9.940299764004084e-05, + "loss": 2.5059, + "step": 23606 + }, + { + "epoch": 1.01, + "grad_norm": 0.4921875, + "learning_rate": 9.939628982939762e-05, + "loss": 2.4706, + "step": 23607 + }, + { + "epoch": 1.01, + "grad_norm": 0.4296875, + "learning_rate": 9.938958202147085e-05, + "loss": 2.566, + "step": 23608 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.938287421629075e-05, + "loss": 2.4288, + "step": 23609 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.937616641388753e-05, + "loss": 2.6308, + "step": 23610 + }, + { + "epoch": 1.01, + "grad_norm": 0.49609375, + "learning_rate": 9.93694586142913e-05, + "loss": 2.4639, + "step": 23611 + }, + { + "epoch": 1.01, + "grad_norm": 0.453125, + "learning_rate": 9.936275081753225e-05, + "loss": 2.5534, + "step": 23612 + }, + { + "epoch": 1.01, + "grad_norm": 0.466796875, + "learning_rate": 9.935604302364063e-05, + "loss": 2.412, + "step": 23613 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.934933523264657e-05, + "loss": 2.4564, + "step": 23614 + }, + { + "epoch": 1.01, + "grad_norm": 0.498046875, + "learning_rate": 9.934262744458026e-05, + "loss": 2.4414, + "step": 23615 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.933591965947186e-05, + "loss": 2.4567, + "step": 23616 + }, + { + "epoch": 1.01, + "grad_norm": 0.462890625, + "learning_rate": 9.932921187735161e-05, + "loss": 2.4922, + "step": 23617 + }, + { + "epoch": 1.01, + "grad_norm": 0.423828125, + "learning_rate": 9.932250409824968e-05, + "loss": 2.4882, + "step": 23618 + }, + { + "epoch": 1.01, + "grad_norm": 0.5, + "learning_rate": 9.931579632219622e-05, + "loss": 2.5049, + "step": 23619 + }, + { + "epoch": 1.01, + "grad_norm": 0.46484375, + "learning_rate": 9.93090885492214e-05, + "loss": 2.4847, + "step": 23620 + }, + { + "epoch": 1.01, + "grad_norm": 0.4453125, + "learning_rate": 9.930238077935544e-05, + "loss": 2.4178, + "step": 23621 + }, + { + "epoch": 1.01, + "grad_norm": 0.421875, + "learning_rate": 9.929567301262851e-05, + "loss": 2.3847, + "step": 23622 + }, + { + "epoch": 1.01, + "grad_norm": 0.5, + "learning_rate": 9.928896524907082e-05, + "loss": 2.4502, + "step": 23623 + }, + { + "epoch": 1.01, + "grad_norm": 0.515625, + "learning_rate": 9.928225748871248e-05, + "loss": 2.4973, + "step": 23624 + }, + { + "epoch": 1.01, + "grad_norm": 0.439453125, + "learning_rate": 9.927554973158375e-05, + "loss": 2.4728, + "step": 23625 + }, + { + "epoch": 1.01, + "grad_norm": 0.484375, + "learning_rate": 9.926884197771482e-05, + "loss": 2.4744, + "step": 23626 + }, + { + "epoch": 1.01, + "grad_norm": 0.458984375, + "learning_rate": 9.926213422713578e-05, + "loss": 2.4054, + "step": 23627 + }, + { + "epoch": 1.01, + "grad_norm": 0.5, + "learning_rate": 9.925542647987686e-05, + "loss": 2.4809, + "step": 23628 + }, + { + "epoch": 1.01, + "grad_norm": 0.46875, + "learning_rate": 9.924871873596827e-05, + "loss": 2.4317, + "step": 23629 + }, + { + "epoch": 1.01, + "grad_norm": 0.52734375, + "learning_rate": 9.924201099544018e-05, + "loss": 2.4551, + "step": 23630 + }, + { + "epoch": 1.01, + "grad_norm": 0.46875, + "learning_rate": 9.923530325832274e-05, + "loss": 2.4859, + "step": 23631 + }, + { + "epoch": 1.01, + "grad_norm": 0.451171875, + "learning_rate": 9.922859552464615e-05, + "loss": 2.5979, + "step": 23632 + }, + { + "epoch": 1.01, + "grad_norm": 0.42578125, + "learning_rate": 9.922188779444063e-05, + "loss": 2.2683, + "step": 23633 + }, + { + "epoch": 1.01, + "grad_norm": 0.48046875, + "learning_rate": 9.921518006773633e-05, + "loss": 2.5199, + "step": 23634 + }, + { + "epoch": 1.01, + "grad_norm": 0.494140625, + "learning_rate": 9.920847234456344e-05, + "loss": 2.3624, + "step": 23635 + }, + { + "epoch": 1.01, + "grad_norm": 0.4453125, + "learning_rate": 9.920176462495207e-05, + "loss": 2.4508, + "step": 23636 + }, + { + "epoch": 1.01, + "grad_norm": 0.48828125, + "learning_rate": 9.919505690893252e-05, + "loss": 2.4998, + "step": 23637 + }, + { + "epoch": 1.01, + "grad_norm": 0.408203125, + "learning_rate": 9.918834919653493e-05, + "loss": 2.5327, + "step": 23638 + }, + { + "epoch": 1.01, + "grad_norm": 0.46484375, + "learning_rate": 9.918164148778944e-05, + "loss": 2.3981, + "step": 23639 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.917493378272631e-05, + "loss": 2.6015, + "step": 23640 + }, + { + "epoch": 1.01, + "grad_norm": 0.439453125, + "learning_rate": 9.916822608137561e-05, + "loss": 2.4075, + "step": 23641 + }, + { + "epoch": 1.01, + "grad_norm": 0.51171875, + "learning_rate": 9.916151838376764e-05, + "loss": 2.5332, + "step": 23642 + }, + { + "epoch": 1.01, + "grad_norm": 0.484375, + "learning_rate": 9.915481068993252e-05, + "loss": 2.5061, + "step": 23643 + }, + { + "epoch": 1.01, + "grad_norm": 0.453125, + "learning_rate": 9.914810299990044e-05, + "loss": 2.5038, + "step": 23644 + }, + { + "epoch": 1.01, + "grad_norm": 0.482421875, + "learning_rate": 9.914139531370155e-05, + "loss": 2.5299, + "step": 23645 + }, + { + "epoch": 1.01, + "grad_norm": 0.482421875, + "learning_rate": 9.91346876313661e-05, + "loss": 2.5053, + "step": 23646 + }, + { + "epoch": 1.01, + "grad_norm": 0.474609375, + "learning_rate": 9.912797995292426e-05, + "loss": 2.4008, + "step": 23647 + }, + { + "epoch": 1.01, + "grad_norm": 0.453125, + "learning_rate": 9.912127227840619e-05, + "loss": 2.4992, + "step": 23648 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.9114564607842e-05, + "loss": 2.4056, + "step": 23649 + }, + { + "epoch": 1.01, + "grad_norm": 0.46875, + "learning_rate": 9.910785694126204e-05, + "loss": 2.4455, + "step": 23650 + }, + { + "epoch": 1.01, + "grad_norm": 0.462890625, + "learning_rate": 9.910114927869635e-05, + "loss": 2.3724, + "step": 23651 + }, + { + "epoch": 1.01, + "grad_norm": 0.46484375, + "learning_rate": 9.909444162017517e-05, + "loss": 2.472, + "step": 23652 + }, + { + "epoch": 1.01, + "grad_norm": 0.427734375, + "learning_rate": 9.908773396572866e-05, + "loss": 2.4188, + "step": 23653 + }, + { + "epoch": 1.01, + "grad_norm": 0.474609375, + "learning_rate": 9.908102631538701e-05, + "loss": 2.385, + "step": 23654 + }, + { + "epoch": 1.01, + "grad_norm": 0.451171875, + "learning_rate": 9.907431866918043e-05, + "loss": 2.5408, + "step": 23655 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.906761102713909e-05, + "loss": 2.5637, + "step": 23656 + }, + { + "epoch": 1.01, + "grad_norm": 0.43359375, + "learning_rate": 9.906090338929309e-05, + "loss": 2.3581, + "step": 23657 + }, + { + "epoch": 1.01, + "grad_norm": 0.5078125, + "learning_rate": 9.905419575567276e-05, + "loss": 2.6331, + "step": 23658 + }, + { + "epoch": 1.01, + "grad_norm": 0.49609375, + "learning_rate": 9.904748812630816e-05, + "loss": 2.5442, + "step": 23659 + }, + { + "epoch": 1.01, + "grad_norm": 0.44140625, + "learning_rate": 9.904078050122954e-05, + "loss": 2.4828, + "step": 23660 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.903407288046702e-05, + "loss": 2.5029, + "step": 23661 + }, + { + "epoch": 1.01, + "grad_norm": 0.458984375, + "learning_rate": 9.902736526405084e-05, + "loss": 2.4474, + "step": 23662 + }, + { + "epoch": 1.01, + "grad_norm": 0.443359375, + "learning_rate": 9.902065765201117e-05, + "loss": 2.4899, + "step": 23663 + }, + { + "epoch": 1.01, + "grad_norm": 0.45703125, + "learning_rate": 9.90139500443782e-05, + "loss": 2.3648, + "step": 23664 + }, + { + "epoch": 1.01, + "grad_norm": 0.466796875, + "learning_rate": 9.900724244118202e-05, + "loss": 2.388, + "step": 23665 + }, + { + "epoch": 1.01, + "grad_norm": 0.44921875, + "learning_rate": 9.900053484245298e-05, + "loss": 2.3394, + "step": 23666 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.899382724822112e-05, + "loss": 2.4418, + "step": 23667 + }, + { + "epoch": 1.01, + "grad_norm": 0.4765625, + "learning_rate": 9.898711965851668e-05, + "loss": 2.4574, + "step": 23668 + }, + { + "epoch": 1.01, + "grad_norm": 0.478515625, + "learning_rate": 9.89804120733698e-05, + "loss": 2.6334, + "step": 23669 + }, + { + "epoch": 1.01, + "grad_norm": 0.474609375, + "learning_rate": 9.897370449281073e-05, + "loss": 2.5308, + "step": 23670 + }, + { + "epoch": 1.01, + "grad_norm": 0.421875, + "learning_rate": 9.896699691686959e-05, + "loss": 2.4659, + "step": 23671 + }, + { + "epoch": 1.01, + "grad_norm": 0.421875, + "learning_rate": 9.896028934557663e-05, + "loss": 2.4851, + "step": 23672 + }, + { + "epoch": 1.01, + "grad_norm": 0.494140625, + "learning_rate": 9.895358177896192e-05, + "loss": 2.3395, + "step": 23673 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.894687421705577e-05, + "loss": 2.4679, + "step": 23674 + }, + { + "epoch": 1.01, + "grad_norm": 0.4375, + "learning_rate": 9.894016665988828e-05, + "loss": 2.4428, + "step": 23675 + }, + { + "epoch": 1.01, + "grad_norm": 0.466796875, + "learning_rate": 9.893345910748965e-05, + "loss": 2.4024, + "step": 23676 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.892675155989005e-05, + "loss": 2.4189, + "step": 23677 + }, + { + "epoch": 1.01, + "grad_norm": 0.474609375, + "learning_rate": 9.89200440171197e-05, + "loss": 2.4337, + "step": 23678 + }, + { + "epoch": 1.01, + "grad_norm": 0.49609375, + "learning_rate": 9.891333647920875e-05, + "loss": 2.4197, + "step": 23679 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.890662894618741e-05, + "loss": 2.5106, + "step": 23680 + }, + { + "epoch": 1.01, + "grad_norm": 0.48828125, + "learning_rate": 9.889992141808577e-05, + "loss": 2.4802, + "step": 23681 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.889321389493416e-05, + "loss": 2.3977, + "step": 23682 + }, + { + "epoch": 1.01, + "grad_norm": 0.478515625, + "learning_rate": 9.888650637676264e-05, + "loss": 2.5503, + "step": 23683 + }, + { + "epoch": 1.01, + "grad_norm": 0.43359375, + "learning_rate": 9.887979886360146e-05, + "loss": 2.441, + "step": 23684 + }, + { + "epoch": 1.01, + "grad_norm": 0.48828125, + "learning_rate": 9.887309135548074e-05, + "loss": 2.5421, + "step": 23685 + }, + { + "epoch": 1.01, + "grad_norm": 0.451171875, + "learning_rate": 9.886638385243073e-05, + "loss": 2.5054, + "step": 23686 + }, + { + "epoch": 1.01, + "grad_norm": 0.55078125, + "learning_rate": 9.885967635448156e-05, + "loss": 2.4251, + "step": 23687 + }, + { + "epoch": 1.01, + "grad_norm": 0.4296875, + "learning_rate": 9.885296886166346e-05, + "loss": 2.3653, + "step": 23688 + }, + { + "epoch": 1.01, + "grad_norm": 0.49609375, + "learning_rate": 9.884626137400653e-05, + "loss": 2.5146, + "step": 23689 + }, + { + "epoch": 1.01, + "grad_norm": 0.484375, + "learning_rate": 9.883955389154104e-05, + "loss": 2.408, + "step": 23690 + }, + { + "epoch": 1.01, + "grad_norm": 0.458984375, + "learning_rate": 9.883284641429714e-05, + "loss": 2.3676, + "step": 23691 + }, + { + "epoch": 1.01, + "grad_norm": 0.515625, + "learning_rate": 9.882613894230497e-05, + "loss": 2.51, + "step": 23692 + }, + { + "epoch": 1.01, + "grad_norm": 0.46875, + "learning_rate": 9.881943147559475e-05, + "loss": 2.4555, + "step": 23693 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.881272401419667e-05, + "loss": 2.3598, + "step": 23694 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.88060165581409e-05, + "loss": 2.3975, + "step": 23695 + }, + { + "epoch": 1.01, + "grad_norm": 0.51953125, + "learning_rate": 9.879930910745765e-05, + "loss": 2.6095, + "step": 23696 + }, + { + "epoch": 1.01, + "grad_norm": 0.443359375, + "learning_rate": 9.8792601662177e-05, + "loss": 2.3985, + "step": 23697 + }, + { + "epoch": 1.01, + "grad_norm": 0.474609375, + "learning_rate": 9.878589422232925e-05, + "loss": 2.3722, + "step": 23698 + }, + { + "epoch": 1.01, + "grad_norm": 0.451171875, + "learning_rate": 9.877918678794454e-05, + "loss": 2.5841, + "step": 23699 + }, + { + "epoch": 1.01, + "grad_norm": 0.46484375, + "learning_rate": 9.877247935905302e-05, + "loss": 2.4964, + "step": 23700 + }, + { + "epoch": 1.01, + "grad_norm": 0.55078125, + "learning_rate": 9.876577193568488e-05, + "loss": 2.4914, + "step": 23701 + }, + { + "epoch": 1.01, + "grad_norm": 0.427734375, + "learning_rate": 9.875906451787034e-05, + "loss": 2.4812, + "step": 23702 + }, + { + "epoch": 1.01, + "grad_norm": 0.54296875, + "learning_rate": 9.875235710563955e-05, + "loss": 2.4472, + "step": 23703 + }, + { + "epoch": 1.01, + "grad_norm": 0.3984375, + "learning_rate": 9.874564969902272e-05, + "loss": 2.3788, + "step": 23704 + }, + { + "epoch": 1.01, + "grad_norm": 0.45703125, + "learning_rate": 9.873894229804995e-05, + "loss": 2.5624, + "step": 23705 + }, + { + "epoch": 1.01, + "grad_norm": 0.427734375, + "learning_rate": 9.873223490275154e-05, + "loss": 2.3561, + "step": 23706 + }, + { + "epoch": 1.01, + "grad_norm": 0.486328125, + "learning_rate": 9.872552751315757e-05, + "loss": 2.4545, + "step": 23707 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.871882012929828e-05, + "loss": 2.4252, + "step": 23708 + }, + { + "epoch": 1.01, + "grad_norm": 0.48046875, + "learning_rate": 9.871211275120381e-05, + "loss": 2.5217, + "step": 23709 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.870540537890438e-05, + "loss": 2.5202, + "step": 23710 + }, + { + "epoch": 1.01, + "grad_norm": 0.4140625, + "learning_rate": 9.869869801243016e-05, + "loss": 2.4656, + "step": 23711 + }, + { + "epoch": 1.01, + "grad_norm": 0.419921875, + "learning_rate": 9.869199065181134e-05, + "loss": 2.5014, + "step": 23712 + }, + { + "epoch": 1.01, + "grad_norm": 0.474609375, + "learning_rate": 9.868528329707801e-05, + "loss": 2.4877, + "step": 23713 + }, + { + "epoch": 1.01, + "grad_norm": 0.53515625, + "learning_rate": 9.86785759482605e-05, + "loss": 2.5631, + "step": 23714 + }, + { + "epoch": 1.01, + "grad_norm": 0.478515625, + "learning_rate": 9.867186860538889e-05, + "loss": 2.4852, + "step": 23715 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.866516126849338e-05, + "loss": 2.4831, + "step": 23716 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.865845393760414e-05, + "loss": 2.5443, + "step": 23717 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.865174661275139e-05, + "loss": 2.5753, + "step": 23718 + }, + { + "epoch": 1.01, + "grad_norm": 0.5, + "learning_rate": 9.864503929396529e-05, + "loss": 2.5563, + "step": 23719 + }, + { + "epoch": 1.01, + "grad_norm": 0.46484375, + "learning_rate": 9.863833198127603e-05, + "loss": 2.4975, + "step": 23720 + }, + { + "epoch": 1.01, + "grad_norm": 0.53125, + "learning_rate": 9.863162467471373e-05, + "loss": 2.5023, + "step": 23721 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.862491737430866e-05, + "loss": 2.4564, + "step": 23722 + }, + { + "epoch": 1.01, + "grad_norm": 0.439453125, + "learning_rate": 9.861821008009096e-05, + "loss": 2.4128, + "step": 23723 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.861150279209081e-05, + "loss": 2.5084, + "step": 23724 + }, + { + "epoch": 1.01, + "grad_norm": 0.5, + "learning_rate": 9.860479551033835e-05, + "loss": 2.4465, + "step": 23725 + }, + { + "epoch": 1.01, + "grad_norm": 0.435546875, + "learning_rate": 9.859808823486384e-05, + "loss": 2.4616, + "step": 23726 + }, + { + "epoch": 1.01, + "grad_norm": 0.5078125, + "learning_rate": 9.859138096569742e-05, + "loss": 2.5178, + "step": 23727 + }, + { + "epoch": 1.01, + "grad_norm": 0.5546875, + "learning_rate": 9.858467370286927e-05, + "loss": 2.5225, + "step": 23728 + }, + { + "epoch": 1.01, + "grad_norm": 0.4765625, + "learning_rate": 9.857796644640952e-05, + "loss": 2.4055, + "step": 23729 + }, + { + "epoch": 1.01, + "grad_norm": 0.443359375, + "learning_rate": 9.857125919634847e-05, + "loss": 2.3705, + "step": 23730 + }, + { + "epoch": 1.01, + "grad_norm": 0.474609375, + "learning_rate": 9.856455195271621e-05, + "loss": 2.5504, + "step": 23731 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.855784471554294e-05, + "loss": 2.5886, + "step": 23732 + }, + { + "epoch": 1.01, + "grad_norm": 0.453125, + "learning_rate": 9.855113748485881e-05, + "loss": 2.3768, + "step": 23733 + }, + { + "epoch": 1.01, + "grad_norm": 0.482421875, + "learning_rate": 9.854443026069407e-05, + "loss": 2.5784, + "step": 23734 + }, + { + "epoch": 1.01, + "grad_norm": 0.431640625, + "learning_rate": 9.853772304307884e-05, + "loss": 2.5624, + "step": 23735 + }, + { + "epoch": 1.01, + "grad_norm": 0.416015625, + "learning_rate": 9.853101583204335e-05, + "loss": 2.3753, + "step": 23736 + }, + { + "epoch": 1.01, + "grad_norm": 0.53125, + "learning_rate": 9.85243086276177e-05, + "loss": 2.485, + "step": 23737 + }, + { + "epoch": 1.01, + "grad_norm": 0.578125, + "learning_rate": 9.85176014298322e-05, + "loss": 2.4979, + "step": 23738 + }, + { + "epoch": 1.01, + "grad_norm": 0.5234375, + "learning_rate": 9.85108942387169e-05, + "loss": 2.483, + "step": 23739 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.850418705430204e-05, + "loss": 2.4539, + "step": 23740 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.849747987661779e-05, + "loss": 2.5195, + "step": 23741 + }, + { + "epoch": 1.01, + "grad_norm": 0.42578125, + "learning_rate": 9.849077270569433e-05, + "loss": 2.5121, + "step": 23742 + }, + { + "epoch": 1.01, + "grad_norm": 0.53125, + "learning_rate": 9.848406554156184e-05, + "loss": 2.3639, + "step": 23743 + }, + { + "epoch": 1.01, + "grad_norm": 0.515625, + "learning_rate": 9.847735838425053e-05, + "loss": 2.4569, + "step": 23744 + }, + { + "epoch": 1.01, + "grad_norm": 0.439453125, + "learning_rate": 9.847065123379048e-05, + "loss": 2.4298, + "step": 23745 + }, + { + "epoch": 1.01, + "grad_norm": 0.5234375, + "learning_rate": 9.846394409021202e-05, + "loss": 2.5621, + "step": 23746 + }, + { + "epoch": 1.01, + "grad_norm": 0.451171875, + "learning_rate": 9.845723695354521e-05, + "loss": 2.5567, + "step": 23747 + }, + { + "epoch": 1.01, + "grad_norm": 0.458984375, + "learning_rate": 9.84505298238203e-05, + "loss": 2.5529, + "step": 23748 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.844382270106739e-05, + "loss": 2.6571, + "step": 23749 + }, + { + "epoch": 1.01, + "grad_norm": 0.44140625, + "learning_rate": 9.843711558531674e-05, + "loss": 2.4098, + "step": 23750 + }, + { + "epoch": 1.01, + "grad_norm": 0.41015625, + "learning_rate": 9.843040847659848e-05, + "loss": 2.4037, + "step": 23751 + }, + { + "epoch": 1.01, + "grad_norm": 0.466796875, + "learning_rate": 9.842370137494285e-05, + "loss": 2.3227, + "step": 23752 + }, + { + "epoch": 1.01, + "grad_norm": 0.494140625, + "learning_rate": 9.841699428037993e-05, + "loss": 2.6301, + "step": 23753 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.841028719294e-05, + "loss": 2.6097, + "step": 23754 + }, + { + "epoch": 1.01, + "grad_norm": 0.41796875, + "learning_rate": 9.84035801126532e-05, + "loss": 2.3794, + "step": 23755 + }, + { + "epoch": 1.01, + "grad_norm": 0.462890625, + "learning_rate": 9.83968730395497e-05, + "loss": 2.4048, + "step": 23756 + }, + { + "epoch": 1.01, + "grad_norm": 0.482421875, + "learning_rate": 9.839016597365966e-05, + "loss": 2.5564, + "step": 23757 + }, + { + "epoch": 1.01, + "grad_norm": 0.462890625, + "learning_rate": 9.838345891501329e-05, + "loss": 2.2431, + "step": 23758 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.837675186364079e-05, + "loss": 2.5099, + "step": 23759 + }, + { + "epoch": 1.01, + "grad_norm": 0.48828125, + "learning_rate": 9.837004481957231e-05, + "loss": 2.4657, + "step": 23760 + }, + { + "epoch": 1.01, + "grad_norm": 0.486328125, + "learning_rate": 9.836333778283803e-05, + "loss": 2.3495, + "step": 23761 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.835663075346811e-05, + "loss": 2.3628, + "step": 23762 + }, + { + "epoch": 1.01, + "grad_norm": 0.53125, + "learning_rate": 9.834992373149275e-05, + "loss": 2.598, + "step": 23763 + }, + { + "epoch": 1.01, + "grad_norm": 0.45703125, + "learning_rate": 9.834321671694215e-05, + "loss": 2.3576, + "step": 23764 + }, + { + "epoch": 1.01, + "grad_norm": 0.453125, + "learning_rate": 9.833650970984648e-05, + "loss": 2.5238, + "step": 23765 + }, + { + "epoch": 1.01, + "grad_norm": 0.5390625, + "learning_rate": 9.832980271023586e-05, + "loss": 2.5644, + "step": 23766 + }, + { + "epoch": 1.01, + "grad_norm": 0.47265625, + "learning_rate": 9.832309571814055e-05, + "loss": 2.4776, + "step": 23767 + }, + { + "epoch": 1.01, + "grad_norm": 0.416015625, + "learning_rate": 9.831638873359073e-05, + "loss": 2.3369, + "step": 23768 + }, + { + "epoch": 1.01, + "grad_norm": 0.44921875, + "learning_rate": 9.830968175661652e-05, + "loss": 2.5384, + "step": 23769 + }, + { + "epoch": 1.01, + "grad_norm": 0.443359375, + "learning_rate": 9.830297478724808e-05, + "loss": 2.5033, + "step": 23770 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.829626782551568e-05, + "loss": 2.42, + "step": 23771 + }, + { + "epoch": 1.01, + "grad_norm": 0.46484375, + "learning_rate": 9.828956087144944e-05, + "loss": 2.3772, + "step": 23772 + }, + { + "epoch": 1.01, + "grad_norm": 0.455078125, + "learning_rate": 9.828285392507955e-05, + "loss": 2.5836, + "step": 23773 + }, + { + "epoch": 1.01, + "grad_norm": 0.494140625, + "learning_rate": 9.827614698643618e-05, + "loss": 2.5061, + "step": 23774 + }, + { + "epoch": 1.01, + "grad_norm": 0.498046875, + "learning_rate": 9.826944005554953e-05, + "loss": 2.4456, + "step": 23775 + }, + { + "epoch": 1.01, + "grad_norm": 0.447265625, + "learning_rate": 9.82627331324498e-05, + "loss": 2.4894, + "step": 23776 + }, + { + "epoch": 1.01, + "grad_norm": 0.5078125, + "learning_rate": 9.82560262171671e-05, + "loss": 2.5359, + "step": 23777 + }, + { + "epoch": 1.01, + "grad_norm": 0.53515625, + "learning_rate": 9.824931930973162e-05, + "loss": 2.4816, + "step": 23778 + }, + { + "epoch": 1.01, + "grad_norm": 0.5078125, + "learning_rate": 9.82426124101736e-05, + "loss": 2.6141, + "step": 23779 + }, + { + "epoch": 1.01, + "grad_norm": 0.53125, + "learning_rate": 9.823590551852317e-05, + "loss": 2.4604, + "step": 23780 + }, + { + "epoch": 1.01, + "grad_norm": 0.431640625, + "learning_rate": 9.822919863481053e-05, + "loss": 2.5195, + "step": 23781 + }, + { + "epoch": 1.01, + "grad_norm": 0.48046875, + "learning_rate": 9.822249175906582e-05, + "loss": 2.6356, + "step": 23782 + }, + { + "epoch": 1.01, + "grad_norm": 0.4453125, + "learning_rate": 9.821578489131928e-05, + "loss": 2.5135, + "step": 23783 + }, + { + "epoch": 1.01, + "grad_norm": 0.515625, + "learning_rate": 9.820907803160106e-05, + "loss": 2.5281, + "step": 23784 + }, + { + "epoch": 1.01, + "grad_norm": 0.40625, + "learning_rate": 9.820237117994134e-05, + "loss": 2.3309, + "step": 23785 + }, + { + "epoch": 1.01, + "grad_norm": 0.48046875, + "learning_rate": 9.819566433637023e-05, + "loss": 2.4312, + "step": 23786 + }, + { + "epoch": 1.01, + "grad_norm": 0.453125, + "learning_rate": 9.818895750091803e-05, + "loss": 2.4147, + "step": 23787 + }, + { + "epoch": 1.01, + "grad_norm": 0.482421875, + "learning_rate": 9.818225067361486e-05, + "loss": 2.4408, + "step": 23788 + }, + { + "epoch": 1.01, + "grad_norm": 0.455078125, + "learning_rate": 9.817554385449088e-05, + "loss": 2.5549, + "step": 23789 + }, + { + "epoch": 1.01, + "grad_norm": 0.462890625, + "learning_rate": 9.816883704357626e-05, + "loss": 2.3775, + "step": 23790 + }, + { + "epoch": 1.01, + "grad_norm": 0.431640625, + "learning_rate": 9.816213024090123e-05, + "loss": 2.4546, + "step": 23791 + }, + { + "epoch": 1.01, + "grad_norm": 0.462890625, + "learning_rate": 9.815542344649598e-05, + "loss": 2.4563, + "step": 23792 + }, + { + "epoch": 1.01, + "grad_norm": 0.51171875, + "learning_rate": 9.81487166603906e-05, + "loss": 2.5093, + "step": 23793 + }, + { + "epoch": 1.01, + "grad_norm": 0.44921875, + "learning_rate": 9.814200988261531e-05, + "loss": 2.395, + "step": 23794 + }, + { + "epoch": 1.01, + "grad_norm": 0.5234375, + "learning_rate": 9.813530311320033e-05, + "loss": 2.3234, + "step": 23795 + }, + { + "epoch": 1.01, + "grad_norm": 0.51953125, + "learning_rate": 9.812859635217579e-05, + "loss": 2.4872, + "step": 23796 + }, + { + "epoch": 1.01, + "grad_norm": 0.4921875, + "learning_rate": 9.81218895995719e-05, + "loss": 2.4421, + "step": 23797 + }, + { + "epoch": 1.01, + "grad_norm": 0.5234375, + "learning_rate": 9.811518285541879e-05, + "loss": 2.4885, + "step": 23798 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.810847611974669e-05, + "loss": 2.3791, + "step": 23799 + }, + { + "epoch": 1.01, + "grad_norm": 0.46875, + "learning_rate": 9.810176939258576e-05, + "loss": 2.4609, + "step": 23800 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.809506267396617e-05, + "loss": 2.4453, + "step": 23801 + }, + { + "epoch": 1.01, + "grad_norm": 0.515625, + "learning_rate": 9.808835596391808e-05, + "loss": 2.5052, + "step": 23802 + }, + { + "epoch": 1.01, + "grad_norm": 0.44921875, + "learning_rate": 9.80816492624717e-05, + "loss": 2.5005, + "step": 23803 + }, + { + "epoch": 1.01, + "grad_norm": 0.48046875, + "learning_rate": 9.807494256965721e-05, + "loss": 2.5374, + "step": 23804 + }, + { + "epoch": 1.01, + "grad_norm": 0.49609375, + "learning_rate": 9.806823588550478e-05, + "loss": 2.456, + "step": 23805 + }, + { + "epoch": 1.01, + "grad_norm": 0.48046875, + "learning_rate": 9.806152921004454e-05, + "loss": 2.4797, + "step": 23806 + }, + { + "epoch": 1.01, + "grad_norm": 0.43359375, + "learning_rate": 9.805482254330675e-05, + "loss": 2.475, + "step": 23807 + }, + { + "epoch": 1.01, + "grad_norm": 0.447265625, + "learning_rate": 9.804811588532157e-05, + "loss": 2.373, + "step": 23808 + }, + { + "epoch": 1.01, + "grad_norm": 0.455078125, + "learning_rate": 9.804140923611912e-05, + "loss": 2.5837, + "step": 23809 + }, + { + "epoch": 1.01, + "grad_norm": 0.427734375, + "learning_rate": 9.80347025957296e-05, + "loss": 2.3626, + "step": 23810 + }, + { + "epoch": 1.01, + "grad_norm": 0.52734375, + "learning_rate": 9.802799596418321e-05, + "loss": 2.416, + "step": 23811 + }, + { + "epoch": 1.01, + "grad_norm": 0.51171875, + "learning_rate": 9.802128934151013e-05, + "loss": 2.4161, + "step": 23812 + }, + { + "epoch": 1.01, + "grad_norm": 0.51171875, + "learning_rate": 9.801458272774052e-05, + "loss": 2.5719, + "step": 23813 + }, + { + "epoch": 1.01, + "grad_norm": 0.50390625, + "learning_rate": 9.800787612290456e-05, + "loss": 2.5377, + "step": 23814 + }, + { + "epoch": 1.01, + "grad_norm": 0.4609375, + "learning_rate": 9.800116952703243e-05, + "loss": 2.3377, + "step": 23815 + }, + { + "epoch": 1.01, + "grad_norm": 0.482421875, + "learning_rate": 9.799446294015432e-05, + "loss": 2.411, + "step": 23816 + }, + { + "epoch": 1.01, + "grad_norm": 0.55078125, + "learning_rate": 9.798775636230038e-05, + "loss": 2.4932, + "step": 23817 + }, + { + "epoch": 1.01, + "grad_norm": 0.466796875, + "learning_rate": 9.79810497935008e-05, + "loss": 2.7101, + "step": 23818 + }, + { + "epoch": 1.01, + "grad_norm": 0.470703125, + "learning_rate": 9.797434323378575e-05, + "loss": 2.4911, + "step": 23819 + }, + { + "epoch": 1.02, + "grad_norm": 0.46875, + "learning_rate": 9.796763668318543e-05, + "loss": 2.621, + "step": 23820 + }, + { + "epoch": 1.02, + "grad_norm": 0.466796875, + "learning_rate": 9.796093014173e-05, + "loss": 2.5436, + "step": 23821 + }, + { + "epoch": 1.02, + "grad_norm": 0.41015625, + "learning_rate": 9.795422360944962e-05, + "loss": 2.4513, + "step": 23822 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.79475170863745e-05, + "loss": 2.5286, + "step": 23823 + }, + { + "epoch": 1.02, + "grad_norm": 0.4375, + "learning_rate": 9.794081057253483e-05, + "loss": 2.5315, + "step": 23824 + }, + { + "epoch": 1.02, + "grad_norm": 0.4765625, + "learning_rate": 9.793410406796074e-05, + "loss": 2.4934, + "step": 23825 + }, + { + "epoch": 1.02, + "grad_norm": 0.474609375, + "learning_rate": 9.79273975726824e-05, + "loss": 2.461, + "step": 23826 + }, + { + "epoch": 1.02, + "grad_norm": 0.482421875, + "learning_rate": 9.792069108673005e-05, + "loss": 2.4791, + "step": 23827 + }, + { + "epoch": 1.02, + "grad_norm": 0.4609375, + "learning_rate": 9.791398461013382e-05, + "loss": 2.4984, + "step": 23828 + }, + { + "epoch": 1.02, + "grad_norm": 0.478515625, + "learning_rate": 9.790727814292389e-05, + "loss": 2.487, + "step": 23829 + }, + { + "epoch": 1.02, + "grad_norm": 0.482421875, + "learning_rate": 9.790057168513041e-05, + "loss": 2.4778, + "step": 23830 + }, + { + "epoch": 1.02, + "grad_norm": 0.482421875, + "learning_rate": 9.789386523678365e-05, + "loss": 2.5504, + "step": 23831 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.788715879791371e-05, + "loss": 2.5522, + "step": 23832 + }, + { + "epoch": 1.02, + "grad_norm": 0.443359375, + "learning_rate": 9.788045236855079e-05, + "loss": 2.5149, + "step": 23833 + }, + { + "epoch": 1.02, + "grad_norm": 0.447265625, + "learning_rate": 9.787374594872502e-05, + "loss": 2.4243, + "step": 23834 + }, + { + "epoch": 1.02, + "grad_norm": 0.53515625, + "learning_rate": 9.786703953846665e-05, + "loss": 2.6454, + "step": 23835 + }, + { + "epoch": 1.02, + "grad_norm": 0.51171875, + "learning_rate": 9.786033313780581e-05, + "loss": 2.5392, + "step": 23836 + }, + { + "epoch": 1.02, + "grad_norm": 0.4765625, + "learning_rate": 9.78536267467727e-05, + "loss": 2.4573, + "step": 23837 + }, + { + "epoch": 1.02, + "grad_norm": 0.48046875, + "learning_rate": 9.784692036539747e-05, + "loss": 2.4033, + "step": 23838 + }, + { + "epoch": 1.02, + "grad_norm": 0.4765625, + "learning_rate": 9.784021399371032e-05, + "loss": 2.4016, + "step": 23839 + }, + { + "epoch": 1.02, + "grad_norm": 0.466796875, + "learning_rate": 9.783350763174143e-05, + "loss": 2.384, + "step": 23840 + }, + { + "epoch": 1.02, + "grad_norm": 0.486328125, + "learning_rate": 9.782680127952096e-05, + "loss": 2.544, + "step": 23841 + }, + { + "epoch": 1.02, + "grad_norm": 0.443359375, + "learning_rate": 9.782009493707907e-05, + "loss": 2.3478, + "step": 23842 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.781338860444597e-05, + "loss": 2.6227, + "step": 23843 + }, + { + "epoch": 1.02, + "grad_norm": 0.494140625, + "learning_rate": 9.780668228165182e-05, + "loss": 2.3157, + "step": 23844 + }, + { + "epoch": 1.02, + "grad_norm": 0.4765625, + "learning_rate": 9.77999759687268e-05, + "loss": 2.5469, + "step": 23845 + }, + { + "epoch": 1.02, + "grad_norm": 0.412109375, + "learning_rate": 9.779326966570105e-05, + "loss": 2.4459, + "step": 23846 + }, + { + "epoch": 1.02, + "grad_norm": 0.52734375, + "learning_rate": 9.778656337260484e-05, + "loss": 2.526, + "step": 23847 + }, + { + "epoch": 1.02, + "grad_norm": 0.498046875, + "learning_rate": 9.777985708946827e-05, + "loss": 2.4827, + "step": 23848 + }, + { + "epoch": 1.02, + "grad_norm": 0.4765625, + "learning_rate": 9.777315081632152e-05, + "loss": 2.4899, + "step": 23849 + }, + { + "epoch": 1.02, + "grad_norm": 0.458984375, + "learning_rate": 9.776644455319476e-05, + "loss": 2.3979, + "step": 23850 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.775973830011821e-05, + "loss": 2.477, + "step": 23851 + }, + { + "epoch": 1.02, + "grad_norm": 0.419921875, + "learning_rate": 9.775303205712201e-05, + "loss": 2.3026, + "step": 23852 + }, + { + "epoch": 1.02, + "grad_norm": 0.5546875, + "learning_rate": 9.774632582423635e-05, + "loss": 2.612, + "step": 23853 + }, + { + "epoch": 1.02, + "grad_norm": 0.41796875, + "learning_rate": 9.773961960149138e-05, + "loss": 2.4444, + "step": 23854 + }, + { + "epoch": 1.02, + "grad_norm": 0.4296875, + "learning_rate": 9.773291338891731e-05, + "loss": 2.3987, + "step": 23855 + }, + { + "epoch": 1.02, + "grad_norm": 0.423828125, + "learning_rate": 9.772620718654432e-05, + "loss": 2.4252, + "step": 23856 + }, + { + "epoch": 1.02, + "grad_norm": 0.474609375, + "learning_rate": 9.771950099440256e-05, + "loss": 2.4843, + "step": 23857 + }, + { + "epoch": 1.02, + "grad_norm": 0.5390625, + "learning_rate": 9.771279481252218e-05, + "loss": 2.5176, + "step": 23858 + }, + { + "epoch": 1.02, + "grad_norm": 0.462890625, + "learning_rate": 9.770608864093341e-05, + "loss": 2.3908, + "step": 23859 + }, + { + "epoch": 1.02, + "grad_norm": 0.5546875, + "learning_rate": 9.76993824796664e-05, + "loss": 2.5674, + "step": 23860 + }, + { + "epoch": 1.02, + "grad_norm": 0.490234375, + "learning_rate": 9.769267632875133e-05, + "loss": 2.4014, + "step": 23861 + }, + { + "epoch": 1.02, + "grad_norm": 0.48828125, + "learning_rate": 9.768597018821836e-05, + "loss": 2.4833, + "step": 23862 + }, + { + "epoch": 1.02, + "grad_norm": 0.490234375, + "learning_rate": 9.767926405809769e-05, + "loss": 2.4504, + "step": 23863 + }, + { + "epoch": 1.02, + "grad_norm": 0.46875, + "learning_rate": 9.76725579384195e-05, + "loss": 2.5139, + "step": 23864 + }, + { + "epoch": 1.02, + "grad_norm": 0.44921875, + "learning_rate": 9.766585182921394e-05, + "loss": 2.4208, + "step": 23865 + }, + { + "epoch": 1.02, + "grad_norm": 0.4765625, + "learning_rate": 9.765914573051118e-05, + "loss": 2.4527, + "step": 23866 + }, + { + "epoch": 1.02, + "grad_norm": 0.5703125, + "learning_rate": 9.76524396423414e-05, + "loss": 2.5723, + "step": 23867 + }, + { + "epoch": 1.02, + "grad_norm": 0.423828125, + "learning_rate": 9.764573356473481e-05, + "loss": 2.4206, + "step": 23868 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.763902749772155e-05, + "loss": 2.496, + "step": 23869 + }, + { + "epoch": 1.02, + "grad_norm": 0.494140625, + "learning_rate": 9.763232144133178e-05, + "loss": 2.4637, + "step": 23870 + }, + { + "epoch": 1.02, + "grad_norm": 0.46875, + "learning_rate": 9.762561539559573e-05, + "loss": 2.3673, + "step": 23871 + }, + { + "epoch": 1.02, + "grad_norm": 0.482421875, + "learning_rate": 9.761890936054356e-05, + "loss": 2.506, + "step": 23872 + }, + { + "epoch": 1.02, + "grad_norm": 0.51171875, + "learning_rate": 9.76122033362054e-05, + "loss": 2.612, + "step": 23873 + }, + { + "epoch": 1.02, + "grad_norm": 0.421875, + "learning_rate": 9.760549732261142e-05, + "loss": 2.3138, + "step": 23874 + }, + { + "epoch": 1.02, + "grad_norm": 0.486328125, + "learning_rate": 9.759879131979188e-05, + "loss": 2.4821, + "step": 23875 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.759208532777688e-05, + "loss": 2.4459, + "step": 23876 + }, + { + "epoch": 1.02, + "grad_norm": 0.54296875, + "learning_rate": 9.758537934659661e-05, + "loss": 2.4014, + "step": 23877 + }, + { + "epoch": 1.02, + "grad_norm": 0.4296875, + "learning_rate": 9.757867337628124e-05, + "loss": 2.5628, + "step": 23878 + }, + { + "epoch": 1.02, + "grad_norm": 0.466796875, + "learning_rate": 9.757196741686098e-05, + "loss": 2.4324, + "step": 23879 + }, + { + "epoch": 1.02, + "grad_norm": 0.474609375, + "learning_rate": 9.756526146836599e-05, + "loss": 2.5828, + "step": 23880 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.755855553082642e-05, + "loss": 2.5207, + "step": 23881 + }, + { + "epoch": 1.02, + "grad_norm": 0.48828125, + "learning_rate": 9.755184960427245e-05, + "loss": 2.5154, + "step": 23882 + }, + { + "epoch": 1.02, + "grad_norm": 0.431640625, + "learning_rate": 9.754514368873425e-05, + "loss": 2.5591, + "step": 23883 + }, + { + "epoch": 1.02, + "grad_norm": 0.458984375, + "learning_rate": 9.753843778424202e-05, + "loss": 2.4496, + "step": 23884 + }, + { + "epoch": 1.02, + "grad_norm": 0.52734375, + "learning_rate": 9.753173189082592e-05, + "loss": 2.4824, + "step": 23885 + }, + { + "epoch": 1.02, + "grad_norm": 0.52734375, + "learning_rate": 9.752502600851617e-05, + "loss": 2.5769, + "step": 23886 + }, + { + "epoch": 1.02, + "grad_norm": 0.4609375, + "learning_rate": 9.751832013734281e-05, + "loss": 2.4479, + "step": 23887 + }, + { + "epoch": 1.02, + "grad_norm": 0.431640625, + "learning_rate": 9.751161427733616e-05, + "loss": 2.4152, + "step": 23888 + }, + { + "epoch": 1.02, + "grad_norm": 0.400390625, + "learning_rate": 9.750490842852633e-05, + "loss": 2.45, + "step": 23889 + }, + { + "epoch": 1.02, + "grad_norm": 0.431640625, + "learning_rate": 9.74982025909435e-05, + "loss": 2.4268, + "step": 23890 + }, + { + "epoch": 1.02, + "grad_norm": 0.431640625, + "learning_rate": 9.749149676461782e-05, + "loss": 2.486, + "step": 23891 + }, + { + "epoch": 1.02, + "grad_norm": 0.443359375, + "learning_rate": 9.74847909495795e-05, + "loss": 2.4195, + "step": 23892 + }, + { + "epoch": 1.02, + "grad_norm": 0.48828125, + "learning_rate": 9.747808514585871e-05, + "loss": 2.455, + "step": 23893 + }, + { + "epoch": 1.02, + "grad_norm": 0.484375, + "learning_rate": 9.747137935348564e-05, + "loss": 2.4256, + "step": 23894 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.746467357249037e-05, + "loss": 2.4884, + "step": 23895 + }, + { + "epoch": 1.02, + "grad_norm": 0.4765625, + "learning_rate": 9.74579678029032e-05, + "loss": 2.5137, + "step": 23896 + }, + { + "epoch": 1.02, + "grad_norm": 0.44921875, + "learning_rate": 9.745126204475425e-05, + "loss": 2.3579, + "step": 23897 + }, + { + "epoch": 1.02, + "grad_norm": 0.466796875, + "learning_rate": 9.744455629807366e-05, + "loss": 2.4539, + "step": 23898 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.743785056289163e-05, + "loss": 2.4809, + "step": 23899 + }, + { + "epoch": 1.02, + "grad_norm": 0.4140625, + "learning_rate": 9.743114483923834e-05, + "loss": 2.4533, + "step": 23900 + }, + { + "epoch": 1.02, + "grad_norm": 0.47265625, + "learning_rate": 9.742443912714397e-05, + "loss": 2.5518, + "step": 23901 + }, + { + "epoch": 1.02, + "grad_norm": 0.443359375, + "learning_rate": 9.74177334266387e-05, + "loss": 2.5172, + "step": 23902 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.741102773775263e-05, + "loss": 2.3525, + "step": 23903 + }, + { + "epoch": 1.02, + "grad_norm": 0.439453125, + "learning_rate": 9.740432206051605e-05, + "loss": 2.3732, + "step": 23904 + }, + { + "epoch": 1.02, + "grad_norm": 0.435546875, + "learning_rate": 9.739761639495904e-05, + "loss": 2.4019, + "step": 23905 + }, + { + "epoch": 1.02, + "grad_norm": 0.47265625, + "learning_rate": 9.739091074111181e-05, + "loss": 2.5224, + "step": 23906 + }, + { + "epoch": 1.02, + "grad_norm": 0.48828125, + "learning_rate": 9.738420509900453e-05, + "loss": 2.5363, + "step": 23907 + }, + { + "epoch": 1.02, + "grad_norm": 0.396484375, + "learning_rate": 9.737749946866737e-05, + "loss": 2.3994, + "step": 23908 + }, + { + "epoch": 1.02, + "grad_norm": 0.474609375, + "learning_rate": 9.73707938501305e-05, + "loss": 2.3527, + "step": 23909 + }, + { + "epoch": 1.02, + "grad_norm": 0.427734375, + "learning_rate": 9.736408824342414e-05, + "loss": 2.4542, + "step": 23910 + }, + { + "epoch": 1.02, + "grad_norm": 0.484375, + "learning_rate": 9.735738264857833e-05, + "loss": 2.5553, + "step": 23911 + }, + { + "epoch": 1.02, + "grad_norm": 0.462890625, + "learning_rate": 9.735067706562341e-05, + "loss": 2.4573, + "step": 23912 + }, + { + "epoch": 1.02, + "grad_norm": 0.486328125, + "learning_rate": 9.734397149458945e-05, + "loss": 2.427, + "step": 23913 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.733726593550665e-05, + "loss": 2.5667, + "step": 23914 + }, + { + "epoch": 1.02, + "grad_norm": 0.419921875, + "learning_rate": 9.733056038840516e-05, + "loss": 2.4631, + "step": 23915 + }, + { + "epoch": 1.02, + "grad_norm": 0.419921875, + "learning_rate": 9.732385485331519e-05, + "loss": 2.4786, + "step": 23916 + }, + { + "epoch": 1.02, + "grad_norm": 0.48828125, + "learning_rate": 9.73171493302669e-05, + "loss": 2.4847, + "step": 23917 + }, + { + "epoch": 1.02, + "grad_norm": 0.453125, + "learning_rate": 9.731044381929047e-05, + "loss": 2.5555, + "step": 23918 + }, + { + "epoch": 1.02, + "grad_norm": 0.41796875, + "learning_rate": 9.730373832041601e-05, + "loss": 2.4145, + "step": 23919 + }, + { + "epoch": 1.02, + "grad_norm": 0.458984375, + "learning_rate": 9.729703283367378e-05, + "loss": 2.4464, + "step": 23920 + }, + { + "epoch": 1.02, + "grad_norm": 0.4453125, + "learning_rate": 9.729032735909392e-05, + "loss": 2.5573, + "step": 23921 + }, + { + "epoch": 1.02, + "grad_norm": 0.48046875, + "learning_rate": 9.728362189670659e-05, + "loss": 2.4237, + "step": 23922 + }, + { + "epoch": 1.02, + "grad_norm": 0.482421875, + "learning_rate": 9.727691644654194e-05, + "loss": 2.4216, + "step": 23923 + }, + { + "epoch": 1.02, + "grad_norm": 0.4765625, + "learning_rate": 9.72702110086302e-05, + "loss": 2.3986, + "step": 23924 + }, + { + "epoch": 1.02, + "grad_norm": 0.474609375, + "learning_rate": 9.726350558300151e-05, + "loss": 2.3654, + "step": 23925 + }, + { + "epoch": 1.02, + "grad_norm": 0.48828125, + "learning_rate": 9.725680016968606e-05, + "loss": 2.4293, + "step": 23926 + }, + { + "epoch": 1.02, + "grad_norm": 0.44921875, + "learning_rate": 9.725009476871395e-05, + "loss": 2.4164, + "step": 23927 + }, + { + "epoch": 1.02, + "grad_norm": 0.51953125, + "learning_rate": 9.724338938011548e-05, + "loss": 2.4236, + "step": 23928 + }, + { + "epoch": 1.02, + "grad_norm": 0.498046875, + "learning_rate": 9.723668400392073e-05, + "loss": 2.3855, + "step": 23929 + }, + { + "epoch": 1.02, + "grad_norm": 0.4296875, + "learning_rate": 9.722997864015986e-05, + "loss": 2.4283, + "step": 23930 + }, + { + "epoch": 1.02, + "grad_norm": 0.451171875, + "learning_rate": 9.722327328886308e-05, + "loss": 2.5278, + "step": 23931 + }, + { + "epoch": 1.02, + "grad_norm": 0.48046875, + "learning_rate": 9.721656795006058e-05, + "loss": 2.3988, + "step": 23932 + }, + { + "epoch": 1.02, + "grad_norm": 0.4609375, + "learning_rate": 9.720986262378249e-05, + "loss": 2.2928, + "step": 23933 + }, + { + "epoch": 1.02, + "grad_norm": 0.421875, + "learning_rate": 9.720315731005903e-05, + "loss": 2.3763, + "step": 23934 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.719645200892028e-05, + "loss": 2.5704, + "step": 23935 + }, + { + "epoch": 1.02, + "grad_norm": 0.3984375, + "learning_rate": 9.718974672039653e-05, + "loss": 2.2853, + "step": 23936 + }, + { + "epoch": 1.02, + "grad_norm": 0.494140625, + "learning_rate": 9.718304144451787e-05, + "loss": 2.4478, + "step": 23937 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.71763361813145e-05, + "loss": 2.3973, + "step": 23938 + }, + { + "epoch": 1.02, + "grad_norm": 0.47265625, + "learning_rate": 9.716963093081655e-05, + "loss": 2.3806, + "step": 23939 + }, + { + "epoch": 1.02, + "grad_norm": 0.453125, + "learning_rate": 9.716292569305426e-05, + "loss": 2.3856, + "step": 23940 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.715622046805775e-05, + "loss": 2.4639, + "step": 23941 + }, + { + "epoch": 1.02, + "grad_norm": 0.453125, + "learning_rate": 9.714951525585726e-05, + "loss": 2.3828, + "step": 23942 + }, + { + "epoch": 1.02, + "grad_norm": 0.51171875, + "learning_rate": 9.714281005648283e-05, + "loss": 2.535, + "step": 23943 + }, + { + "epoch": 1.02, + "grad_norm": 0.4140625, + "learning_rate": 9.713610486996478e-05, + "loss": 2.4652, + "step": 23944 + }, + { + "epoch": 1.02, + "grad_norm": 0.42578125, + "learning_rate": 9.712939969633318e-05, + "loss": 2.4172, + "step": 23945 + }, + { + "epoch": 1.02, + "grad_norm": 0.458984375, + "learning_rate": 9.712269453561824e-05, + "loss": 2.4477, + "step": 23946 + }, + { + "epoch": 1.02, + "grad_norm": 0.44921875, + "learning_rate": 9.71159893878501e-05, + "loss": 2.4228, + "step": 23947 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.710928425305897e-05, + "loss": 2.5033, + "step": 23948 + }, + { + "epoch": 1.02, + "grad_norm": 0.419921875, + "learning_rate": 9.7102579131275e-05, + "loss": 2.56, + "step": 23949 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.70958740225284e-05, + "loss": 2.3836, + "step": 23950 + }, + { + "epoch": 1.02, + "grad_norm": 0.482421875, + "learning_rate": 9.708916892684924e-05, + "loss": 2.3113, + "step": 23951 + }, + { + "epoch": 1.02, + "grad_norm": 0.486328125, + "learning_rate": 9.708246384426781e-05, + "loss": 2.5883, + "step": 23952 + }, + { + "epoch": 1.02, + "grad_norm": 0.490234375, + "learning_rate": 9.707575877481421e-05, + "loss": 2.5425, + "step": 23953 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.706905371851864e-05, + "loss": 2.5044, + "step": 23954 + }, + { + "epoch": 1.02, + "grad_norm": 0.4609375, + "learning_rate": 9.706234867541122e-05, + "loss": 2.5725, + "step": 23955 + }, + { + "epoch": 1.02, + "grad_norm": 0.494140625, + "learning_rate": 9.705564364552217e-05, + "loss": 2.5244, + "step": 23956 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.704893862888167e-05, + "loss": 2.4684, + "step": 23957 + }, + { + "epoch": 1.02, + "grad_norm": 0.439453125, + "learning_rate": 9.704223362551987e-05, + "loss": 2.5075, + "step": 23958 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.703552863546688e-05, + "loss": 2.5314, + "step": 23959 + }, + { + "epoch": 1.02, + "grad_norm": 0.49609375, + "learning_rate": 9.7028823658753e-05, + "loss": 2.4778, + "step": 23960 + }, + { + "epoch": 1.02, + "grad_norm": 0.43359375, + "learning_rate": 9.70221186954083e-05, + "loss": 2.3921, + "step": 23961 + }, + { + "epoch": 1.02, + "grad_norm": 0.462890625, + "learning_rate": 9.701541374546299e-05, + "loss": 2.4557, + "step": 23962 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.70087088089472e-05, + "loss": 2.337, + "step": 23963 + }, + { + "epoch": 1.02, + "grad_norm": 0.47265625, + "learning_rate": 9.700200388589114e-05, + "loss": 2.5039, + "step": 23964 + }, + { + "epoch": 1.02, + "grad_norm": 0.447265625, + "learning_rate": 9.699529897632497e-05, + "loss": 2.4895, + "step": 23965 + }, + { + "epoch": 1.02, + "grad_norm": 0.478515625, + "learning_rate": 9.698859408027887e-05, + "loss": 2.4964, + "step": 23966 + }, + { + "epoch": 1.02, + "grad_norm": 0.462890625, + "learning_rate": 9.698188919778294e-05, + "loss": 2.4287, + "step": 23967 + }, + { + "epoch": 1.02, + "grad_norm": 0.462890625, + "learning_rate": 9.697518432886748e-05, + "loss": 2.4661, + "step": 23968 + }, + { + "epoch": 1.02, + "grad_norm": 0.423828125, + "learning_rate": 9.696847947356255e-05, + "loss": 2.3979, + "step": 23969 + }, + { + "epoch": 1.02, + "grad_norm": 0.45703125, + "learning_rate": 9.696177463189835e-05, + "loss": 2.3011, + "step": 23970 + }, + { + "epoch": 1.02, + "grad_norm": 0.486328125, + "learning_rate": 9.695506980390504e-05, + "loss": 2.5089, + "step": 23971 + }, + { + "epoch": 1.02, + "grad_norm": 0.498046875, + "learning_rate": 9.694836498961283e-05, + "loss": 2.5346, + "step": 23972 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.694166018905186e-05, + "loss": 2.5178, + "step": 23973 + }, + { + "epoch": 1.02, + "grad_norm": 0.458984375, + "learning_rate": 9.693495540225231e-05, + "loss": 2.4955, + "step": 23974 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.692825062924429e-05, + "loss": 2.4768, + "step": 23975 + }, + { + "epoch": 1.02, + "grad_norm": 0.4609375, + "learning_rate": 9.692154587005806e-05, + "loss": 2.3952, + "step": 23976 + }, + { + "epoch": 1.02, + "grad_norm": 0.462890625, + "learning_rate": 9.691484112472376e-05, + "loss": 2.5025, + "step": 23977 + }, + { + "epoch": 1.02, + "grad_norm": 0.494140625, + "learning_rate": 9.690813639327154e-05, + "loss": 2.4099, + "step": 23978 + }, + { + "epoch": 1.02, + "grad_norm": 0.42578125, + "learning_rate": 9.690143167573155e-05, + "loss": 2.4797, + "step": 23979 + }, + { + "epoch": 1.02, + "grad_norm": 0.515625, + "learning_rate": 9.6894726972134e-05, + "loss": 2.4851, + "step": 23980 + }, + { + "epoch": 1.02, + "grad_norm": 0.447265625, + "learning_rate": 9.688802228250905e-05, + "loss": 2.4358, + "step": 23981 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.688131760688689e-05, + "loss": 2.4556, + "step": 23982 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.68746129452976e-05, + "loss": 2.3827, + "step": 23983 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.686790829777143e-05, + "loss": 2.4933, + "step": 23984 + }, + { + "epoch": 1.02, + "grad_norm": 0.46875, + "learning_rate": 9.686120366433856e-05, + "loss": 2.678, + "step": 23985 + }, + { + "epoch": 1.02, + "grad_norm": 0.484375, + "learning_rate": 9.685449904502911e-05, + "loss": 2.4498, + "step": 23986 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.684779443987324e-05, + "loss": 2.4896, + "step": 23987 + }, + { + "epoch": 1.02, + "grad_norm": 0.447265625, + "learning_rate": 9.684108984890116e-05, + "loss": 2.7462, + "step": 23988 + }, + { + "epoch": 1.02, + "grad_norm": 0.49609375, + "learning_rate": 9.683438527214303e-05, + "loss": 2.4883, + "step": 23989 + }, + { + "epoch": 1.02, + "grad_norm": 0.5546875, + "learning_rate": 9.682768070962903e-05, + "loss": 2.6246, + "step": 23990 + }, + { + "epoch": 1.02, + "grad_norm": 0.44140625, + "learning_rate": 9.682097616138924e-05, + "loss": 2.4173, + "step": 23991 + }, + { + "epoch": 1.02, + "grad_norm": 0.55078125, + "learning_rate": 9.681427162745394e-05, + "loss": 2.4045, + "step": 23992 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.680756710785327e-05, + "loss": 2.5064, + "step": 23993 + }, + { + "epoch": 1.02, + "grad_norm": 0.42578125, + "learning_rate": 9.680086260261737e-05, + "loss": 2.4377, + "step": 23994 + }, + { + "epoch": 1.02, + "grad_norm": 0.486328125, + "learning_rate": 9.679415811177639e-05, + "loss": 2.383, + "step": 23995 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.678745363536055e-05, + "loss": 2.4598, + "step": 23996 + }, + { + "epoch": 1.02, + "grad_norm": 0.490234375, + "learning_rate": 9.678074917339998e-05, + "loss": 2.4478, + "step": 23997 + }, + { + "epoch": 1.02, + "grad_norm": 0.486328125, + "learning_rate": 9.67740447259249e-05, + "loss": 2.396, + "step": 23998 + }, + { + "epoch": 1.02, + "grad_norm": 0.44921875, + "learning_rate": 9.676734029296539e-05, + "loss": 2.4871, + "step": 23999 + }, + { + "epoch": 1.02, + "grad_norm": 0.453125, + "learning_rate": 9.676063587455169e-05, + "loss": 2.4735, + "step": 24000 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.675393147071397e-05, + "loss": 2.4938, + "step": 24001 + }, + { + "epoch": 1.02, + "grad_norm": 0.478515625, + "learning_rate": 9.674722708148233e-05, + "loss": 2.5554, + "step": 24002 + }, + { + "epoch": 1.02, + "grad_norm": 0.53515625, + "learning_rate": 9.674052270688701e-05, + "loss": 2.4961, + "step": 24003 + }, + { + "epoch": 1.02, + "grad_norm": 0.455078125, + "learning_rate": 9.673381834695812e-05, + "loss": 2.4829, + "step": 24004 + }, + { + "epoch": 1.02, + "grad_norm": 0.4140625, + "learning_rate": 9.672711400172586e-05, + "loss": 2.5959, + "step": 24005 + }, + { + "epoch": 1.02, + "grad_norm": 0.52734375, + "learning_rate": 9.672040967122042e-05, + "loss": 2.5222, + "step": 24006 + }, + { + "epoch": 1.02, + "grad_norm": 0.5234375, + "learning_rate": 9.671370535547193e-05, + "loss": 2.5502, + "step": 24007 + }, + { + "epoch": 1.02, + "grad_norm": 0.5, + "learning_rate": 9.67070010545105e-05, + "loss": 2.3549, + "step": 24008 + }, + { + "epoch": 1.02, + "grad_norm": 0.494140625, + "learning_rate": 9.670029676836644e-05, + "loss": 2.5005, + "step": 24009 + }, + { + "epoch": 1.02, + "grad_norm": 0.447265625, + "learning_rate": 9.669359249706982e-05, + "loss": 2.4226, + "step": 24010 + }, + { + "epoch": 1.02, + "grad_norm": 0.453125, + "learning_rate": 9.668688824065082e-05, + "loss": 2.4243, + "step": 24011 + }, + { + "epoch": 1.02, + "grad_norm": 0.439453125, + "learning_rate": 9.668018399913959e-05, + "loss": 2.4865, + "step": 24012 + }, + { + "epoch": 1.02, + "grad_norm": 0.427734375, + "learning_rate": 9.667347977256633e-05, + "loss": 2.5249, + "step": 24013 + }, + { + "epoch": 1.02, + "grad_norm": 0.453125, + "learning_rate": 9.666677556096124e-05, + "loss": 2.4129, + "step": 24014 + }, + { + "epoch": 1.02, + "grad_norm": 0.5234375, + "learning_rate": 9.66600713643544e-05, + "loss": 2.5694, + "step": 24015 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.665336718277598e-05, + "loss": 2.5104, + "step": 24016 + }, + { + "epoch": 1.02, + "grad_norm": 0.46484375, + "learning_rate": 9.664666301625626e-05, + "loss": 2.4499, + "step": 24017 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.66399588648253e-05, + "loss": 2.5134, + "step": 24018 + }, + { + "epoch": 1.02, + "grad_norm": 0.451171875, + "learning_rate": 9.663325472851329e-05, + "loss": 2.5889, + "step": 24019 + }, + { + "epoch": 1.02, + "grad_norm": 0.48046875, + "learning_rate": 9.66265506073504e-05, + "loss": 2.5997, + "step": 24020 + }, + { + "epoch": 1.02, + "grad_norm": 0.49609375, + "learning_rate": 9.661984650136681e-05, + "loss": 2.4496, + "step": 24021 + }, + { + "epoch": 1.02, + "grad_norm": 0.43359375, + "learning_rate": 9.66131424105927e-05, + "loss": 2.3798, + "step": 24022 + }, + { + "epoch": 1.02, + "grad_norm": 0.50390625, + "learning_rate": 9.660643833505818e-05, + "loss": 2.4118, + "step": 24023 + }, + { + "epoch": 1.02, + "grad_norm": 0.458984375, + "learning_rate": 9.659973427479341e-05, + "loss": 2.5327, + "step": 24024 + }, + { + "epoch": 1.02, + "grad_norm": 0.515625, + "learning_rate": 9.659303022982867e-05, + "loss": 2.4479, + "step": 24025 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.658632620019401e-05, + "loss": 2.5529, + "step": 24026 + }, + { + "epoch": 1.02, + "grad_norm": 0.48828125, + "learning_rate": 9.657962218591963e-05, + "loss": 2.3931, + "step": 24027 + }, + { + "epoch": 1.02, + "grad_norm": 0.5, + "learning_rate": 9.65729181870357e-05, + "loss": 2.4733, + "step": 24028 + }, + { + "epoch": 1.02, + "grad_norm": 0.466796875, + "learning_rate": 9.65662142035724e-05, + "loss": 2.4578, + "step": 24029 + }, + { + "epoch": 1.02, + "grad_norm": 0.455078125, + "learning_rate": 9.655951023555988e-05, + "loss": 2.4655, + "step": 24030 + }, + { + "epoch": 1.02, + "grad_norm": 0.453125, + "learning_rate": 9.655280628302832e-05, + "loss": 2.445, + "step": 24031 + }, + { + "epoch": 1.02, + "grad_norm": 0.470703125, + "learning_rate": 9.654610234600782e-05, + "loss": 2.4341, + "step": 24032 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.653939842452864e-05, + "loss": 2.4839, + "step": 24033 + }, + { + "epoch": 1.02, + "grad_norm": 0.49609375, + "learning_rate": 9.653269451862091e-05, + "loss": 2.5167, + "step": 24034 + }, + { + "epoch": 1.02, + "grad_norm": 0.484375, + "learning_rate": 9.652599062831477e-05, + "loss": 2.4342, + "step": 24035 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.651928675364039e-05, + "loss": 2.5059, + "step": 24036 + }, + { + "epoch": 1.02, + "grad_norm": 0.447265625, + "learning_rate": 9.651258289462796e-05, + "loss": 2.4401, + "step": 24037 + }, + { + "epoch": 1.02, + "grad_norm": 0.435546875, + "learning_rate": 9.650587905130764e-05, + "loss": 2.4933, + "step": 24038 + }, + { + "epoch": 1.02, + "grad_norm": 0.45703125, + "learning_rate": 9.64991752237096e-05, + "loss": 2.6237, + "step": 24039 + }, + { + "epoch": 1.02, + "grad_norm": 0.462890625, + "learning_rate": 9.649247141186394e-05, + "loss": 2.4543, + "step": 24040 + }, + { + "epoch": 1.02, + "grad_norm": 0.46484375, + "learning_rate": 9.648576761580094e-05, + "loss": 2.4183, + "step": 24041 + }, + { + "epoch": 1.02, + "grad_norm": 0.46875, + "learning_rate": 9.647906383555066e-05, + "loss": 2.4323, + "step": 24042 + }, + { + "epoch": 1.02, + "grad_norm": 0.4921875, + "learning_rate": 9.647236007114333e-05, + "loss": 2.5484, + "step": 24043 + }, + { + "epoch": 1.02, + "grad_norm": 0.431640625, + "learning_rate": 9.646565632260906e-05, + "loss": 2.3762, + "step": 24044 + }, + { + "epoch": 1.02, + "grad_norm": 0.46484375, + "learning_rate": 9.645895258997807e-05, + "loss": 2.3538, + "step": 24045 + }, + { + "epoch": 1.02, + "grad_norm": 0.490234375, + "learning_rate": 9.64522488732805e-05, + "loss": 2.5179, + "step": 24046 + }, + { + "epoch": 1.02, + "grad_norm": 0.5078125, + "learning_rate": 9.644554517254651e-05, + "loss": 2.4707, + "step": 24047 + }, + { + "epoch": 1.02, + "grad_norm": 0.416015625, + "learning_rate": 9.643884148780623e-05, + "loss": 2.4759, + "step": 24048 + }, + { + "epoch": 1.02, + "grad_norm": 0.4453125, + "learning_rate": 9.643213781908992e-05, + "loss": 2.4625, + "step": 24049 + }, + { + "epoch": 1.02, + "grad_norm": 0.5234375, + "learning_rate": 9.642543416642766e-05, + "loss": 2.2793, + "step": 24050 + }, + { + "epoch": 1.02, + "grad_norm": 0.51171875, + "learning_rate": 9.641873052984965e-05, + "loss": 2.5184, + "step": 24051 + }, + { + "epoch": 1.02, + "grad_norm": 0.453125, + "learning_rate": 9.641202690938603e-05, + "loss": 2.3978, + "step": 24052 + }, + { + "epoch": 1.02, + "grad_norm": 0.458984375, + "learning_rate": 9.640532330506697e-05, + "loss": 2.4412, + "step": 24053 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.639861971692266e-05, + "loss": 2.3359, + "step": 24054 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.639191614498326e-05, + "loss": 2.5134, + "step": 24055 + }, + { + "epoch": 1.03, + "grad_norm": 0.453125, + "learning_rate": 9.638521258927886e-05, + "loss": 2.5692, + "step": 24056 + }, + { + "epoch": 1.03, + "grad_norm": 0.4765625, + "learning_rate": 9.637850904983975e-05, + "loss": 2.5136, + "step": 24057 + }, + { + "epoch": 1.03, + "grad_norm": 0.447265625, + "learning_rate": 9.6371805526696e-05, + "loss": 2.4107, + "step": 24058 + }, + { + "epoch": 1.03, + "grad_norm": 0.52734375, + "learning_rate": 9.636510201987779e-05, + "loss": 2.4189, + "step": 24059 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.635839852941528e-05, + "loss": 2.501, + "step": 24060 + }, + { + "epoch": 1.03, + "grad_norm": 0.490234375, + "learning_rate": 9.635169505533867e-05, + "loss": 2.5412, + "step": 24061 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.63449915976781e-05, + "loss": 2.433, + "step": 24062 + }, + { + "epoch": 1.03, + "grad_norm": 0.53125, + "learning_rate": 9.633828815646374e-05, + "loss": 2.477, + "step": 24063 + }, + { + "epoch": 1.03, + "grad_norm": 0.466796875, + "learning_rate": 9.633158473172568e-05, + "loss": 2.4903, + "step": 24064 + }, + { + "epoch": 1.03, + "grad_norm": 0.453125, + "learning_rate": 9.632488132349423e-05, + "loss": 2.3884, + "step": 24065 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.631817793179944e-05, + "loss": 2.4713, + "step": 24066 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.631147455667151e-05, + "loss": 2.4944, + "step": 24067 + }, + { + "epoch": 1.03, + "grad_norm": 0.54296875, + "learning_rate": 9.630477119814056e-05, + "loss": 2.5676, + "step": 24068 + }, + { + "epoch": 1.03, + "grad_norm": 0.447265625, + "learning_rate": 9.629806785623683e-05, + "loss": 2.5604, + "step": 24069 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.629136453099044e-05, + "loss": 2.4972, + "step": 24070 + }, + { + "epoch": 1.03, + "grad_norm": 0.53515625, + "learning_rate": 9.628466122243156e-05, + "loss": 2.5298, + "step": 24071 + }, + { + "epoch": 1.03, + "grad_norm": 0.462890625, + "learning_rate": 9.627795793059029e-05, + "loss": 2.4126, + "step": 24072 + }, + { + "epoch": 1.03, + "grad_norm": 0.43359375, + "learning_rate": 9.627125465549691e-05, + "loss": 2.465, + "step": 24073 + }, + { + "epoch": 1.03, + "grad_norm": 0.45703125, + "learning_rate": 9.626455139718151e-05, + "loss": 2.5124, + "step": 24074 + }, + { + "epoch": 1.03, + "grad_norm": 0.44921875, + "learning_rate": 9.625784815567426e-05, + "loss": 2.4853, + "step": 24075 + }, + { + "epoch": 1.03, + "grad_norm": 0.53125, + "learning_rate": 9.62511449310053e-05, + "loss": 2.5544, + "step": 24076 + }, + { + "epoch": 1.03, + "grad_norm": 0.5390625, + "learning_rate": 9.624444172320484e-05, + "loss": 2.4427, + "step": 24077 + }, + { + "epoch": 1.03, + "grad_norm": 0.43359375, + "learning_rate": 9.623773853230301e-05, + "loss": 2.4592, + "step": 24078 + }, + { + "epoch": 1.03, + "grad_norm": 0.458984375, + "learning_rate": 9.623103535833003e-05, + "loss": 2.4944, + "step": 24079 + }, + { + "epoch": 1.03, + "grad_norm": 0.470703125, + "learning_rate": 9.622433220131593e-05, + "loss": 2.4827, + "step": 24080 + }, + { + "epoch": 1.03, + "grad_norm": 0.53125, + "learning_rate": 9.621762906129102e-05, + "loss": 2.4892, + "step": 24081 + }, + { + "epoch": 1.03, + "grad_norm": 0.43359375, + "learning_rate": 9.621092593828538e-05, + "loss": 2.4406, + "step": 24082 + }, + { + "epoch": 1.03, + "grad_norm": 0.4453125, + "learning_rate": 9.620422283232918e-05, + "loss": 2.5016, + "step": 24083 + }, + { + "epoch": 1.03, + "grad_norm": 0.51171875, + "learning_rate": 9.619751974345257e-05, + "loss": 2.472, + "step": 24084 + }, + { + "epoch": 1.03, + "grad_norm": 0.55078125, + "learning_rate": 9.619081667168576e-05, + "loss": 2.4771, + "step": 24085 + }, + { + "epoch": 1.03, + "grad_norm": 0.5, + "learning_rate": 9.618411361705888e-05, + "loss": 2.5546, + "step": 24086 + }, + { + "epoch": 1.03, + "grad_norm": 0.427734375, + "learning_rate": 9.617741057960212e-05, + "loss": 2.4496, + "step": 24087 + }, + { + "epoch": 1.03, + "grad_norm": 0.443359375, + "learning_rate": 9.617070755934554e-05, + "loss": 2.5377, + "step": 24088 + }, + { + "epoch": 1.03, + "grad_norm": 0.5234375, + "learning_rate": 9.616400455631946e-05, + "loss": 2.5366, + "step": 24089 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.615730157055391e-05, + "loss": 2.5626, + "step": 24090 + }, + { + "epoch": 1.03, + "grad_norm": 0.48046875, + "learning_rate": 9.615059860207913e-05, + "loss": 2.3762, + "step": 24091 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.614389565092522e-05, + "loss": 2.5929, + "step": 24092 + }, + { + "epoch": 1.03, + "grad_norm": 0.486328125, + "learning_rate": 9.613719271712238e-05, + "loss": 2.393, + "step": 24093 + }, + { + "epoch": 1.03, + "grad_norm": 0.52734375, + "learning_rate": 9.613048980070075e-05, + "loss": 2.4039, + "step": 24094 + }, + { + "epoch": 1.03, + "grad_norm": 0.423828125, + "learning_rate": 9.612378690169054e-05, + "loss": 2.3575, + "step": 24095 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.611708402012181e-05, + "loss": 2.4817, + "step": 24096 + }, + { + "epoch": 1.03, + "grad_norm": 0.412109375, + "learning_rate": 9.611038115602485e-05, + "loss": 2.4834, + "step": 24097 + }, + { + "epoch": 1.03, + "grad_norm": 0.447265625, + "learning_rate": 9.610367830942971e-05, + "loss": 2.5763, + "step": 24098 + }, + { + "epoch": 1.03, + "grad_norm": 0.43359375, + "learning_rate": 9.609697548036662e-05, + "loss": 2.3602, + "step": 24099 + }, + { + "epoch": 1.03, + "grad_norm": 0.443359375, + "learning_rate": 9.609027266886567e-05, + "loss": 2.4112, + "step": 24100 + }, + { + "epoch": 1.03, + "grad_norm": 0.51953125, + "learning_rate": 9.608356987495709e-05, + "loss": 2.3951, + "step": 24101 + }, + { + "epoch": 1.03, + "grad_norm": 0.46875, + "learning_rate": 9.607686709867103e-05, + "loss": 2.5257, + "step": 24102 + }, + { + "epoch": 1.03, + "grad_norm": 0.447265625, + "learning_rate": 9.607016434003764e-05, + "loss": 2.3591, + "step": 24103 + }, + { + "epoch": 1.03, + "grad_norm": 0.486328125, + "learning_rate": 9.606346159908702e-05, + "loss": 2.3628, + "step": 24104 + }, + { + "epoch": 1.03, + "grad_norm": 0.47265625, + "learning_rate": 9.605675887584944e-05, + "loss": 2.4598, + "step": 24105 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.605005617035498e-05, + "loss": 2.4512, + "step": 24106 + }, + { + "epoch": 1.03, + "grad_norm": 0.4921875, + "learning_rate": 9.604335348263383e-05, + "loss": 2.418, + "step": 24107 + }, + { + "epoch": 1.03, + "grad_norm": 0.474609375, + "learning_rate": 9.603665081271614e-05, + "loss": 2.5091, + "step": 24108 + }, + { + "epoch": 1.03, + "grad_norm": 0.4609375, + "learning_rate": 9.602994816063206e-05, + "loss": 2.4423, + "step": 24109 + }, + { + "epoch": 1.03, + "grad_norm": 0.44921875, + "learning_rate": 9.602324552641178e-05, + "loss": 2.5215, + "step": 24110 + }, + { + "epoch": 1.03, + "grad_norm": 0.458984375, + "learning_rate": 9.601654291008547e-05, + "loss": 2.449, + "step": 24111 + }, + { + "epoch": 1.03, + "grad_norm": 0.4375, + "learning_rate": 9.600984031168319e-05, + "loss": 2.5327, + "step": 24112 + }, + { + "epoch": 1.03, + "grad_norm": 0.474609375, + "learning_rate": 9.600313773123524e-05, + "loss": 2.5329, + "step": 24113 + }, + { + "epoch": 1.03, + "grad_norm": 0.53125, + "learning_rate": 9.599643516877168e-05, + "loss": 2.4607, + "step": 24114 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.59897326243227e-05, + "loss": 2.5581, + "step": 24115 + }, + { + "epoch": 1.03, + "grad_norm": 0.49609375, + "learning_rate": 9.598303009791844e-05, + "loss": 2.4075, + "step": 24116 + }, + { + "epoch": 1.03, + "grad_norm": 0.4453125, + "learning_rate": 9.597632758958909e-05, + "loss": 2.4606, + "step": 24117 + }, + { + "epoch": 1.03, + "grad_norm": 0.5, + "learning_rate": 9.59696250993648e-05, + "loss": 2.5494, + "step": 24118 + }, + { + "epoch": 1.03, + "grad_norm": 0.48046875, + "learning_rate": 9.596292262727574e-05, + "loss": 2.5978, + "step": 24119 + }, + { + "epoch": 1.03, + "grad_norm": 0.46875, + "learning_rate": 9.595622017335199e-05, + "loss": 2.4631, + "step": 24120 + }, + { + "epoch": 1.03, + "grad_norm": 0.4609375, + "learning_rate": 9.594951773762385e-05, + "loss": 2.4784, + "step": 24121 + }, + { + "epoch": 1.03, + "grad_norm": 0.4375, + "learning_rate": 9.594281532012138e-05, + "loss": 2.457, + "step": 24122 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.593611292087473e-05, + "loss": 2.3724, + "step": 24123 + }, + { + "epoch": 1.03, + "grad_norm": 0.427734375, + "learning_rate": 9.592941053991411e-05, + "loss": 2.4062, + "step": 24124 + }, + { + "epoch": 1.03, + "grad_norm": 0.4140625, + "learning_rate": 9.592270817726963e-05, + "loss": 2.5238, + "step": 24125 + }, + { + "epoch": 1.03, + "grad_norm": 0.4765625, + "learning_rate": 9.591600583297148e-05, + "loss": 2.4093, + "step": 24126 + }, + { + "epoch": 1.03, + "grad_norm": 0.490234375, + "learning_rate": 9.590930350704985e-05, + "loss": 2.4854, + "step": 24127 + }, + { + "epoch": 1.03, + "grad_norm": 0.53515625, + "learning_rate": 9.590260119953483e-05, + "loss": 2.5195, + "step": 24128 + }, + { + "epoch": 1.03, + "grad_norm": 0.51171875, + "learning_rate": 9.58958989104566e-05, + "loss": 2.5188, + "step": 24129 + }, + { + "epoch": 1.03, + "grad_norm": 0.5078125, + "learning_rate": 9.588919663984532e-05, + "loss": 2.539, + "step": 24130 + }, + { + "epoch": 1.03, + "grad_norm": 0.50390625, + "learning_rate": 9.588249438773117e-05, + "loss": 2.4708, + "step": 24131 + }, + { + "epoch": 1.03, + "grad_norm": 0.466796875, + "learning_rate": 9.58757921541443e-05, + "loss": 2.5064, + "step": 24132 + }, + { + "epoch": 1.03, + "grad_norm": 0.42578125, + "learning_rate": 9.58690899391148e-05, + "loss": 2.4337, + "step": 24133 + }, + { + "epoch": 1.03, + "grad_norm": 0.462890625, + "learning_rate": 9.586238774267294e-05, + "loss": 2.5119, + "step": 24134 + }, + { + "epoch": 1.03, + "grad_norm": 0.462890625, + "learning_rate": 9.585568556484884e-05, + "loss": 2.3968, + "step": 24135 + }, + { + "epoch": 1.03, + "grad_norm": 0.486328125, + "learning_rate": 9.584898340567262e-05, + "loss": 2.439, + "step": 24136 + }, + { + "epoch": 1.03, + "grad_norm": 0.474609375, + "learning_rate": 9.584228126517442e-05, + "loss": 2.5555, + "step": 24137 + }, + { + "epoch": 1.03, + "grad_norm": 0.45703125, + "learning_rate": 9.583557914338447e-05, + "loss": 2.418, + "step": 24138 + }, + { + "epoch": 1.03, + "grad_norm": 0.47265625, + "learning_rate": 9.58288770403329e-05, + "loss": 2.3556, + "step": 24139 + }, + { + "epoch": 1.03, + "grad_norm": 0.48828125, + "learning_rate": 9.582217495604984e-05, + "loss": 2.475, + "step": 24140 + }, + { + "epoch": 1.03, + "grad_norm": 0.4765625, + "learning_rate": 9.581547289056545e-05, + "loss": 2.5332, + "step": 24141 + }, + { + "epoch": 1.03, + "grad_norm": 0.462890625, + "learning_rate": 9.580877084390993e-05, + "loss": 2.3996, + "step": 24142 + }, + { + "epoch": 1.03, + "grad_norm": 0.447265625, + "learning_rate": 9.580206881611343e-05, + "loss": 2.4686, + "step": 24143 + }, + { + "epoch": 1.03, + "grad_norm": 0.3828125, + "learning_rate": 9.579536680720606e-05, + "loss": 2.47, + "step": 24144 + }, + { + "epoch": 1.03, + "grad_norm": 0.44140625, + "learning_rate": 9.578866481721799e-05, + "loss": 2.5586, + "step": 24145 + }, + { + "epoch": 1.03, + "grad_norm": 0.5390625, + "learning_rate": 9.57819628461794e-05, + "loss": 2.5349, + "step": 24146 + }, + { + "epoch": 1.03, + "grad_norm": 0.50390625, + "learning_rate": 9.577526089412043e-05, + "loss": 2.4113, + "step": 24147 + }, + { + "epoch": 1.03, + "grad_norm": 0.4375, + "learning_rate": 9.576855896107126e-05, + "loss": 2.341, + "step": 24148 + }, + { + "epoch": 1.03, + "grad_norm": 0.470703125, + "learning_rate": 9.5761857047062e-05, + "loss": 2.4121, + "step": 24149 + }, + { + "epoch": 1.03, + "grad_norm": 0.470703125, + "learning_rate": 9.575515515212286e-05, + "loss": 2.6149, + "step": 24150 + }, + { + "epoch": 1.03, + "grad_norm": 0.431640625, + "learning_rate": 9.574845327628397e-05, + "loss": 2.369, + "step": 24151 + }, + { + "epoch": 1.03, + "grad_norm": 0.443359375, + "learning_rate": 9.574175141957549e-05, + "loss": 2.3828, + "step": 24152 + }, + { + "epoch": 1.03, + "grad_norm": 0.515625, + "learning_rate": 9.573504958202755e-05, + "loss": 2.546, + "step": 24153 + }, + { + "epoch": 1.03, + "grad_norm": 0.48828125, + "learning_rate": 9.572834776367033e-05, + "loss": 2.5377, + "step": 24154 + }, + { + "epoch": 1.03, + "grad_norm": 0.515625, + "learning_rate": 9.5721645964534e-05, + "loss": 2.4749, + "step": 24155 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.57149441846487e-05, + "loss": 2.4981, + "step": 24156 + }, + { + "epoch": 1.03, + "grad_norm": 0.546875, + "learning_rate": 9.570824242404456e-05, + "loss": 2.5859, + "step": 24157 + }, + { + "epoch": 1.03, + "grad_norm": 0.515625, + "learning_rate": 9.570154068275178e-05, + "loss": 2.4384, + "step": 24158 + }, + { + "epoch": 1.03, + "grad_norm": 0.5234375, + "learning_rate": 9.569483896080051e-05, + "loss": 2.438, + "step": 24159 + }, + { + "epoch": 1.03, + "grad_norm": 0.44921875, + "learning_rate": 9.568813725822088e-05, + "loss": 2.5456, + "step": 24160 + }, + { + "epoch": 1.03, + "grad_norm": 0.458984375, + "learning_rate": 9.568143557504304e-05, + "loss": 2.4857, + "step": 24161 + }, + { + "epoch": 1.03, + "grad_norm": 0.53125, + "learning_rate": 9.567473391129718e-05, + "loss": 2.4402, + "step": 24162 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.566803226701344e-05, + "loss": 2.5769, + "step": 24163 + }, + { + "epoch": 1.03, + "grad_norm": 0.416015625, + "learning_rate": 9.566133064222195e-05, + "loss": 2.3903, + "step": 24164 + }, + { + "epoch": 1.03, + "grad_norm": 0.478515625, + "learning_rate": 9.565462903695289e-05, + "loss": 2.4619, + "step": 24165 + }, + { + "epoch": 1.03, + "grad_norm": 0.49609375, + "learning_rate": 9.564792745123642e-05, + "loss": 2.5758, + "step": 24166 + }, + { + "epoch": 1.03, + "grad_norm": 0.5390625, + "learning_rate": 9.564122588510271e-05, + "loss": 2.5337, + "step": 24167 + }, + { + "epoch": 1.03, + "grad_norm": 0.494140625, + "learning_rate": 9.563452433858187e-05, + "loss": 2.4491, + "step": 24168 + }, + { + "epoch": 1.03, + "grad_norm": 0.46875, + "learning_rate": 9.562782281170405e-05, + "loss": 2.4308, + "step": 24169 + }, + { + "epoch": 1.03, + "grad_norm": 0.4765625, + "learning_rate": 9.562112130449946e-05, + "loss": 2.3468, + "step": 24170 + }, + { + "epoch": 1.03, + "grad_norm": 0.486328125, + "learning_rate": 9.561441981699822e-05, + "loss": 2.4771, + "step": 24171 + }, + { + "epoch": 1.03, + "grad_norm": 0.48828125, + "learning_rate": 9.560771834923048e-05, + "loss": 2.4485, + "step": 24172 + }, + { + "epoch": 1.03, + "grad_norm": 0.44140625, + "learning_rate": 9.56010169012264e-05, + "loss": 2.4123, + "step": 24173 + }, + { + "epoch": 1.03, + "grad_norm": 0.453125, + "learning_rate": 9.559431547301614e-05, + "loss": 2.4609, + "step": 24174 + }, + { + "epoch": 1.03, + "grad_norm": 0.45703125, + "learning_rate": 9.558761406462987e-05, + "loss": 2.4584, + "step": 24175 + }, + { + "epoch": 1.03, + "grad_norm": 0.50390625, + "learning_rate": 9.55809126760977e-05, + "loss": 2.4607, + "step": 24176 + }, + { + "epoch": 1.03, + "grad_norm": 0.474609375, + "learning_rate": 9.55742113074498e-05, + "loss": 2.2964, + "step": 24177 + }, + { + "epoch": 1.03, + "grad_norm": 0.50390625, + "learning_rate": 9.556750995871636e-05, + "loss": 2.5746, + "step": 24178 + }, + { + "epoch": 1.03, + "grad_norm": 0.48046875, + "learning_rate": 9.556080862992749e-05, + "loss": 2.6107, + "step": 24179 + }, + { + "epoch": 1.03, + "grad_norm": 0.423828125, + "learning_rate": 9.555410732111337e-05, + "loss": 2.3332, + "step": 24180 + }, + { + "epoch": 1.03, + "grad_norm": 0.50390625, + "learning_rate": 9.55474060323041e-05, + "loss": 2.4441, + "step": 24181 + }, + { + "epoch": 1.03, + "grad_norm": 0.47265625, + "learning_rate": 9.554070476352992e-05, + "loss": 2.5103, + "step": 24182 + }, + { + "epoch": 1.03, + "grad_norm": 0.38671875, + "learning_rate": 9.553400351482095e-05, + "loss": 2.2677, + "step": 24183 + }, + { + "epoch": 1.03, + "grad_norm": 0.5390625, + "learning_rate": 9.55273022862073e-05, + "loss": 2.6224, + "step": 24184 + }, + { + "epoch": 1.03, + "grad_norm": 0.4453125, + "learning_rate": 9.552060107771915e-05, + "loss": 2.4967, + "step": 24185 + }, + { + "epoch": 1.03, + "grad_norm": 0.5078125, + "learning_rate": 9.551389988938669e-05, + "loss": 2.4778, + "step": 24186 + }, + { + "epoch": 1.03, + "grad_norm": 0.56640625, + "learning_rate": 9.550719872124e-05, + "loss": 2.6994, + "step": 24187 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.550049757330931e-05, + "loss": 2.3182, + "step": 24188 + }, + { + "epoch": 1.03, + "grad_norm": 0.578125, + "learning_rate": 9.54937964456247e-05, + "loss": 2.4329, + "step": 24189 + }, + { + "epoch": 1.03, + "grad_norm": 0.4609375, + "learning_rate": 9.54870953382164e-05, + "loss": 2.3666, + "step": 24190 + }, + { + "epoch": 1.03, + "grad_norm": 0.474609375, + "learning_rate": 9.548039425111453e-05, + "loss": 2.4002, + "step": 24191 + }, + { + "epoch": 1.03, + "grad_norm": 0.45703125, + "learning_rate": 9.54736931843492e-05, + "loss": 2.5278, + "step": 24192 + }, + { + "epoch": 1.03, + "grad_norm": 0.419921875, + "learning_rate": 9.54669921379506e-05, + "loss": 2.3252, + "step": 24193 + }, + { + "epoch": 1.03, + "grad_norm": 0.4453125, + "learning_rate": 9.546029111194888e-05, + "loss": 2.4839, + "step": 24194 + }, + { + "epoch": 1.03, + "grad_norm": 0.462890625, + "learning_rate": 9.545359010637419e-05, + "loss": 2.4541, + "step": 24195 + }, + { + "epoch": 1.03, + "grad_norm": 0.46875, + "learning_rate": 9.54468891212567e-05, + "loss": 2.4519, + "step": 24196 + }, + { + "epoch": 1.03, + "grad_norm": 0.470703125, + "learning_rate": 9.54401881566265e-05, + "loss": 2.4485, + "step": 24197 + }, + { + "epoch": 1.03, + "grad_norm": 0.484375, + "learning_rate": 9.543348721251383e-05, + "loss": 2.3641, + "step": 24198 + }, + { + "epoch": 1.03, + "grad_norm": 0.419921875, + "learning_rate": 9.54267862889488e-05, + "loss": 2.5239, + "step": 24199 + }, + { + "epoch": 1.03, + "grad_norm": 0.439453125, + "learning_rate": 9.542008538596157e-05, + "loss": 2.441, + "step": 24200 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.541338450358224e-05, + "loss": 2.4185, + "step": 24201 + }, + { + "epoch": 1.03, + "grad_norm": 0.455078125, + "learning_rate": 9.540668364184103e-05, + "loss": 2.4149, + "step": 24202 + }, + { + "epoch": 1.03, + "grad_norm": 0.44140625, + "learning_rate": 9.539998280076807e-05, + "loss": 2.4603, + "step": 24203 + }, + { + "epoch": 1.03, + "grad_norm": 0.419921875, + "learning_rate": 9.539328198039349e-05, + "loss": 2.4053, + "step": 24204 + }, + { + "epoch": 1.03, + "grad_norm": 0.5078125, + "learning_rate": 9.538658118074743e-05, + "loss": 2.5043, + "step": 24205 + }, + { + "epoch": 1.03, + "grad_norm": 0.494140625, + "learning_rate": 9.537988040186011e-05, + "loss": 2.5021, + "step": 24206 + }, + { + "epoch": 1.03, + "grad_norm": 0.48828125, + "learning_rate": 9.537317964376165e-05, + "loss": 2.4452, + "step": 24207 + }, + { + "epoch": 1.03, + "grad_norm": 0.439453125, + "learning_rate": 9.536647890648219e-05, + "loss": 2.5045, + "step": 24208 + }, + { + "epoch": 1.03, + "grad_norm": 0.494140625, + "learning_rate": 9.535977819005185e-05, + "loss": 2.5391, + "step": 24209 + }, + { + "epoch": 1.03, + "grad_norm": 0.474609375, + "learning_rate": 9.535307749450081e-05, + "loss": 2.5422, + "step": 24210 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.534637681985926e-05, + "loss": 2.4747, + "step": 24211 + }, + { + "epoch": 1.03, + "grad_norm": 0.455078125, + "learning_rate": 9.533967616615729e-05, + "loss": 2.5, + "step": 24212 + }, + { + "epoch": 1.03, + "grad_norm": 0.462890625, + "learning_rate": 9.533297553342505e-05, + "loss": 2.4681, + "step": 24213 + }, + { + "epoch": 1.03, + "grad_norm": 0.51953125, + "learning_rate": 9.532627492169275e-05, + "loss": 2.5587, + "step": 24214 + }, + { + "epoch": 1.03, + "grad_norm": 0.515625, + "learning_rate": 9.531957433099052e-05, + "loss": 2.4921, + "step": 24215 + }, + { + "epoch": 1.03, + "grad_norm": 0.43359375, + "learning_rate": 9.531287376134847e-05, + "loss": 2.4648, + "step": 24216 + }, + { + "epoch": 1.03, + "grad_norm": 0.486328125, + "learning_rate": 9.530617321279675e-05, + "loss": 2.6327, + "step": 24217 + }, + { + "epoch": 1.03, + "grad_norm": 0.515625, + "learning_rate": 9.529947268536558e-05, + "loss": 2.6069, + "step": 24218 + }, + { + "epoch": 1.03, + "grad_norm": 0.47265625, + "learning_rate": 9.529277217908505e-05, + "loss": 2.5435, + "step": 24219 + }, + { + "epoch": 1.03, + "grad_norm": 0.4609375, + "learning_rate": 9.528607169398531e-05, + "loss": 2.4338, + "step": 24220 + }, + { + "epoch": 1.03, + "grad_norm": 0.478515625, + "learning_rate": 9.527937123009653e-05, + "loss": 2.4401, + "step": 24221 + }, + { + "epoch": 1.03, + "grad_norm": 0.455078125, + "learning_rate": 9.527267078744885e-05, + "loss": 2.5521, + "step": 24222 + }, + { + "epoch": 1.03, + "grad_norm": 0.455078125, + "learning_rate": 9.526597036607246e-05, + "loss": 2.5321, + "step": 24223 + }, + { + "epoch": 1.03, + "grad_norm": 0.44921875, + "learning_rate": 9.525926996599745e-05, + "loss": 2.3774, + "step": 24224 + }, + { + "epoch": 1.03, + "grad_norm": 0.48828125, + "learning_rate": 9.525256958725398e-05, + "loss": 2.4606, + "step": 24225 + }, + { + "epoch": 1.03, + "grad_norm": 0.466796875, + "learning_rate": 9.524586922987222e-05, + "loss": 2.4995, + "step": 24226 + }, + { + "epoch": 1.03, + "grad_norm": 0.427734375, + "learning_rate": 9.523916889388232e-05, + "loss": 2.309, + "step": 24227 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.523246857931441e-05, + "loss": 2.6024, + "step": 24228 + }, + { + "epoch": 1.03, + "grad_norm": 0.52734375, + "learning_rate": 9.522576828619864e-05, + "loss": 2.4614, + "step": 24229 + }, + { + "epoch": 1.03, + "grad_norm": 0.46875, + "learning_rate": 9.521906801456518e-05, + "loss": 2.5823, + "step": 24230 + }, + { + "epoch": 1.03, + "grad_norm": 0.41796875, + "learning_rate": 9.521236776444419e-05, + "loss": 2.421, + "step": 24231 + }, + { + "epoch": 1.03, + "grad_norm": 0.486328125, + "learning_rate": 9.520566753586577e-05, + "loss": 2.5365, + "step": 24232 + }, + { + "epoch": 1.03, + "grad_norm": 0.466796875, + "learning_rate": 9.519896732886007e-05, + "loss": 2.3402, + "step": 24233 + }, + { + "epoch": 1.03, + "grad_norm": 0.458984375, + "learning_rate": 9.51922671434573e-05, + "loss": 2.4124, + "step": 24234 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.518556697968757e-05, + "loss": 2.3994, + "step": 24235 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.517886683758102e-05, + "loss": 2.4009, + "step": 24236 + }, + { + "epoch": 1.03, + "grad_norm": 0.4609375, + "learning_rate": 9.517216671716777e-05, + "loss": 2.4939, + "step": 24237 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.516546661847806e-05, + "loss": 2.6131, + "step": 24238 + }, + { + "epoch": 1.03, + "grad_norm": 0.4453125, + "learning_rate": 9.515876654154199e-05, + "loss": 2.4782, + "step": 24239 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.515206648638966e-05, + "loss": 2.5195, + "step": 24240 + }, + { + "epoch": 1.03, + "grad_norm": 0.458984375, + "learning_rate": 9.514536645305126e-05, + "loss": 2.5246, + "step": 24241 + }, + { + "epoch": 1.03, + "grad_norm": 0.439453125, + "learning_rate": 9.513866644155697e-05, + "loss": 2.4098, + "step": 24242 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.513196645193688e-05, + "loss": 2.4535, + "step": 24243 + }, + { + "epoch": 1.03, + "grad_norm": 0.51953125, + "learning_rate": 9.512526648422117e-05, + "loss": 2.586, + "step": 24244 + }, + { + "epoch": 1.03, + "grad_norm": 0.37890625, + "learning_rate": 9.511856653844e-05, + "loss": 2.2632, + "step": 24245 + }, + { + "epoch": 1.03, + "grad_norm": 0.4609375, + "learning_rate": 9.511186661462344e-05, + "loss": 2.4104, + "step": 24246 + }, + { + "epoch": 1.03, + "grad_norm": 0.416015625, + "learning_rate": 9.510516671280176e-05, + "loss": 2.5609, + "step": 24247 + }, + { + "epoch": 1.03, + "grad_norm": 0.400390625, + "learning_rate": 9.5098466833005e-05, + "loss": 2.5407, + "step": 24248 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.509176697526336e-05, + "loss": 2.4759, + "step": 24249 + }, + { + "epoch": 1.03, + "grad_norm": 0.458984375, + "learning_rate": 9.508506713960696e-05, + "loss": 2.5548, + "step": 24250 + }, + { + "epoch": 1.03, + "grad_norm": 0.455078125, + "learning_rate": 9.507836732606597e-05, + "loss": 2.4796, + "step": 24251 + }, + { + "epoch": 1.03, + "grad_norm": 0.38671875, + "learning_rate": 9.507166753467052e-05, + "loss": 2.431, + "step": 24252 + }, + { + "epoch": 1.03, + "grad_norm": 0.4140625, + "learning_rate": 9.506496776545082e-05, + "loss": 2.6722, + "step": 24253 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.505826801843687e-05, + "loss": 2.3591, + "step": 24254 + }, + { + "epoch": 1.03, + "grad_norm": 0.4609375, + "learning_rate": 9.505156829365898e-05, + "loss": 2.451, + "step": 24255 + }, + { + "epoch": 1.03, + "grad_norm": 0.453125, + "learning_rate": 9.50448685911472e-05, + "loss": 2.4902, + "step": 24256 + }, + { + "epoch": 1.03, + "grad_norm": 0.482421875, + "learning_rate": 9.50381689109317e-05, + "loss": 2.6036, + "step": 24257 + }, + { + "epoch": 1.03, + "grad_norm": 0.5390625, + "learning_rate": 9.503146925304261e-05, + "loss": 2.4107, + "step": 24258 + }, + { + "epoch": 1.03, + "grad_norm": 0.48828125, + "learning_rate": 9.502476961751012e-05, + "loss": 2.5235, + "step": 24259 + }, + { + "epoch": 1.03, + "grad_norm": 0.4296875, + "learning_rate": 9.501807000436433e-05, + "loss": 2.5061, + "step": 24260 + }, + { + "epoch": 1.03, + "grad_norm": 0.498046875, + "learning_rate": 9.501137041363545e-05, + "loss": 2.5206, + "step": 24261 + }, + { + "epoch": 1.03, + "grad_norm": 0.4765625, + "learning_rate": 9.50046708453535e-05, + "loss": 2.4068, + "step": 24262 + }, + { + "epoch": 1.03, + "grad_norm": 0.5078125, + "learning_rate": 9.499797129954876e-05, + "loss": 2.5667, + "step": 24263 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.499127177625132e-05, + "loss": 2.5215, + "step": 24264 + }, + { + "epoch": 1.03, + "grad_norm": 0.49609375, + "learning_rate": 9.498457227549133e-05, + "loss": 2.3574, + "step": 24265 + }, + { + "epoch": 1.03, + "grad_norm": 0.46875, + "learning_rate": 9.497787279729888e-05, + "loss": 2.515, + "step": 24266 + }, + { + "epoch": 1.03, + "grad_norm": 0.44921875, + "learning_rate": 9.497117334170421e-05, + "loss": 2.5556, + "step": 24267 + }, + { + "epoch": 1.03, + "grad_norm": 0.486328125, + "learning_rate": 9.496447390873743e-05, + "loss": 2.3583, + "step": 24268 + }, + { + "epoch": 1.03, + "grad_norm": 0.40625, + "learning_rate": 9.495777449842868e-05, + "loss": 2.5007, + "step": 24269 + }, + { + "epoch": 1.03, + "grad_norm": 0.51171875, + "learning_rate": 9.495107511080806e-05, + "loss": 2.5965, + "step": 24270 + }, + { + "epoch": 1.03, + "grad_norm": 0.515625, + "learning_rate": 9.49443757459058e-05, + "loss": 2.5971, + "step": 24271 + }, + { + "epoch": 1.03, + "grad_norm": 0.46484375, + "learning_rate": 9.493767640375198e-05, + "loss": 2.3327, + "step": 24272 + }, + { + "epoch": 1.03, + "grad_norm": 0.466796875, + "learning_rate": 9.493097708437677e-05, + "loss": 2.4842, + "step": 24273 + }, + { + "epoch": 1.03, + "grad_norm": 0.470703125, + "learning_rate": 9.49242777878103e-05, + "loss": 2.4369, + "step": 24274 + }, + { + "epoch": 1.03, + "grad_norm": 0.47265625, + "learning_rate": 9.491757851408273e-05, + "loss": 2.4104, + "step": 24275 + }, + { + "epoch": 1.03, + "grad_norm": 0.451171875, + "learning_rate": 9.491087926322421e-05, + "loss": 2.4091, + "step": 24276 + }, + { + "epoch": 1.03, + "grad_norm": 0.47265625, + "learning_rate": 9.49041800352649e-05, + "loss": 2.3849, + "step": 24277 + }, + { + "epoch": 1.03, + "grad_norm": 0.4375, + "learning_rate": 9.489748083023484e-05, + "loss": 2.4042, + "step": 24278 + }, + { + "epoch": 1.03, + "grad_norm": 0.486328125, + "learning_rate": 9.489078164816433e-05, + "loss": 2.4758, + "step": 24279 + }, + { + "epoch": 1.03, + "grad_norm": 0.4921875, + "learning_rate": 9.48840824890834e-05, + "loss": 2.5265, + "step": 24280 + }, + { + "epoch": 1.03, + "grad_norm": 0.4453125, + "learning_rate": 9.487738335302223e-05, + "loss": 2.352, + "step": 24281 + }, + { + "epoch": 1.03, + "grad_norm": 0.54296875, + "learning_rate": 9.487068424001096e-05, + "loss": 2.4832, + "step": 24282 + }, + { + "epoch": 1.03, + "grad_norm": 0.46875, + "learning_rate": 9.486398515007973e-05, + "loss": 2.5283, + "step": 24283 + }, + { + "epoch": 1.03, + "grad_norm": 0.40625, + "learning_rate": 9.485728608325872e-05, + "loss": 2.4345, + "step": 24284 + }, + { + "epoch": 1.03, + "grad_norm": 0.50390625, + "learning_rate": 9.485058703957804e-05, + "loss": 2.4858, + "step": 24285 + }, + { + "epoch": 1.03, + "grad_norm": 0.462890625, + "learning_rate": 9.484388801906778e-05, + "loss": 2.4999, + "step": 24286 + }, + { + "epoch": 1.03, + "grad_norm": 0.50390625, + "learning_rate": 9.48371890217582e-05, + "loss": 2.4557, + "step": 24287 + }, + { + "epoch": 1.03, + "grad_norm": 0.50390625, + "learning_rate": 9.483049004767937e-05, + "loss": 2.33, + "step": 24288 + }, + { + "epoch": 1.04, + "grad_norm": 0.455078125, + "learning_rate": 9.482379109686145e-05, + "loss": 2.5204, + "step": 24289 + }, + { + "epoch": 1.04, + "grad_norm": 0.490234375, + "learning_rate": 9.481709216933454e-05, + "loss": 2.5719, + "step": 24290 + }, + { + "epoch": 1.04, + "grad_norm": 0.46484375, + "learning_rate": 9.481039326512884e-05, + "loss": 2.4505, + "step": 24291 + }, + { + "epoch": 1.04, + "grad_norm": 0.490234375, + "learning_rate": 9.480369438427449e-05, + "loss": 2.5313, + "step": 24292 + }, + { + "epoch": 1.04, + "grad_norm": 0.4453125, + "learning_rate": 9.479699552680164e-05, + "loss": 2.4521, + "step": 24293 + }, + { + "epoch": 1.04, + "grad_norm": 0.4765625, + "learning_rate": 9.479029669274033e-05, + "loss": 2.4709, + "step": 24294 + }, + { + "epoch": 1.04, + "grad_norm": 0.45703125, + "learning_rate": 9.478359788212085e-05, + "loss": 2.3404, + "step": 24295 + }, + { + "epoch": 1.04, + "grad_norm": 0.51171875, + "learning_rate": 9.477689909497327e-05, + "loss": 2.4591, + "step": 24296 + }, + { + "epoch": 1.04, + "grad_norm": 0.486328125, + "learning_rate": 9.47702003313277e-05, + "loss": 2.6379, + "step": 24297 + }, + { + "epoch": 1.04, + "grad_norm": 0.45703125, + "learning_rate": 9.476350159121432e-05, + "loss": 2.2766, + "step": 24298 + }, + { + "epoch": 1.04, + "grad_norm": 0.51171875, + "learning_rate": 9.475680287466329e-05, + "loss": 2.529, + "step": 24299 + }, + { + "epoch": 1.04, + "grad_norm": 0.44921875, + "learning_rate": 9.475010418170473e-05, + "loss": 2.4589, + "step": 24300 + }, + { + "epoch": 1.04, + "grad_norm": 0.462890625, + "learning_rate": 9.47434055123688e-05, + "loss": 2.4702, + "step": 24301 + }, + { + "epoch": 1.04, + "grad_norm": 0.478515625, + "learning_rate": 9.473670686668556e-05, + "loss": 2.4729, + "step": 24302 + }, + { + "epoch": 1.04, + "grad_norm": 0.5390625, + "learning_rate": 9.473000824468527e-05, + "loss": 2.408, + "step": 24303 + }, + { + "epoch": 1.04, + "grad_norm": 0.48828125, + "learning_rate": 9.472330964639801e-05, + "loss": 2.3265, + "step": 24304 + }, + { + "epoch": 1.04, + "grad_norm": 0.462890625, + "learning_rate": 9.471661107185392e-05, + "loss": 2.4754, + "step": 24305 + }, + { + "epoch": 1.04, + "grad_norm": 0.421875, + "learning_rate": 9.470991252108313e-05, + "loss": 2.4846, + "step": 24306 + }, + { + "epoch": 1.04, + "grad_norm": 0.4765625, + "learning_rate": 9.470321399411582e-05, + "loss": 2.3919, + "step": 24307 + }, + { + "epoch": 1.04, + "grad_norm": 0.44140625, + "learning_rate": 9.469651549098212e-05, + "loss": 2.4662, + "step": 24308 + }, + { + "epoch": 1.04, + "grad_norm": 0.451171875, + "learning_rate": 9.468981701171218e-05, + "loss": 2.4877, + "step": 24309 + }, + { + "epoch": 1.04, + "grad_norm": 0.484375, + "learning_rate": 9.468311855633606e-05, + "loss": 2.4509, + "step": 24310 + }, + { + "epoch": 1.04, + "grad_norm": 0.44921875, + "learning_rate": 9.467642012488402e-05, + "loss": 2.4457, + "step": 24311 + }, + { + "epoch": 1.04, + "grad_norm": 0.49609375, + "learning_rate": 9.466972171738613e-05, + "loss": 2.5199, + "step": 24312 + }, + { + "epoch": 1.04, + "grad_norm": 0.484375, + "learning_rate": 9.466302333387254e-05, + "loss": 2.4459, + "step": 24313 + }, + { + "epoch": 1.04, + "grad_norm": 0.44921875, + "learning_rate": 9.465632497437336e-05, + "loss": 2.5372, + "step": 24314 + }, + { + "epoch": 1.04, + "grad_norm": 0.51953125, + "learning_rate": 9.464962663891881e-05, + "loss": 2.4554, + "step": 24315 + }, + { + "epoch": 1.04, + "grad_norm": 0.453125, + "learning_rate": 9.464292832753897e-05, + "loss": 2.6258, + "step": 24316 + }, + { + "epoch": 1.04, + "grad_norm": 0.44140625, + "learning_rate": 9.463623004026403e-05, + "loss": 2.5976, + "step": 24317 + }, + { + "epoch": 1.04, + "grad_norm": 0.4609375, + "learning_rate": 9.462953177712401e-05, + "loss": 2.5328, + "step": 24318 + }, + { + "epoch": 1.04, + "grad_norm": 0.4921875, + "learning_rate": 9.462283353814922e-05, + "loss": 2.4789, + "step": 24319 + }, + { + "epoch": 1.04, + "grad_norm": 0.498046875, + "learning_rate": 9.461613532336969e-05, + "loss": 2.4544, + "step": 24320 + }, + { + "epoch": 1.04, + "grad_norm": 0.498046875, + "learning_rate": 9.460943713281557e-05, + "loss": 2.4568, + "step": 24321 + }, + { + "epoch": 1.04, + "grad_norm": 0.447265625, + "learning_rate": 9.460273896651699e-05, + "loss": 2.48, + "step": 24322 + }, + { + "epoch": 1.04, + "grad_norm": 0.44140625, + "learning_rate": 9.459604082450414e-05, + "loss": 2.409, + "step": 24323 + }, + { + "epoch": 1.04, + "grad_norm": 0.447265625, + "learning_rate": 9.458934270680713e-05, + "loss": 2.3774, + "step": 24324 + }, + { + "epoch": 1.04, + "grad_norm": 0.458984375, + "learning_rate": 9.458264461345613e-05, + "loss": 2.4555, + "step": 24325 + }, + { + "epoch": 1.04, + "grad_norm": 0.46875, + "learning_rate": 9.457594654448118e-05, + "loss": 2.388, + "step": 24326 + }, + { + "epoch": 1.04, + "grad_norm": 0.52734375, + "learning_rate": 9.456924849991255e-05, + "loss": 2.4878, + "step": 24327 + }, + { + "epoch": 1.04, + "grad_norm": 0.439453125, + "learning_rate": 9.45625504797803e-05, + "loss": 2.2999, + "step": 24328 + }, + { + "epoch": 1.04, + "grad_norm": 0.482421875, + "learning_rate": 9.455585248411458e-05, + "loss": 2.4684, + "step": 24329 + }, + { + "epoch": 1.04, + "grad_norm": 0.4765625, + "learning_rate": 9.454915451294552e-05, + "loss": 2.4465, + "step": 24330 + }, + { + "epoch": 1.04, + "grad_norm": 0.515625, + "learning_rate": 9.454245656630328e-05, + "loss": 2.6349, + "step": 24331 + }, + { + "epoch": 1.04, + "grad_norm": 0.490234375, + "learning_rate": 9.4535758644218e-05, + "loss": 2.5034, + "step": 24332 + }, + { + "epoch": 1.04, + "grad_norm": 0.52734375, + "learning_rate": 9.452906074671984e-05, + "loss": 2.5192, + "step": 24333 + }, + { + "epoch": 1.04, + "grad_norm": 0.404296875, + "learning_rate": 9.452236287383883e-05, + "loss": 2.4068, + "step": 24334 + }, + { + "epoch": 1.04, + "grad_norm": 0.46484375, + "learning_rate": 9.451566502560526e-05, + "loss": 2.458, + "step": 24335 + }, + { + "epoch": 1.04, + "grad_norm": 0.458984375, + "learning_rate": 9.450896720204917e-05, + "loss": 2.4519, + "step": 24336 + }, + { + "epoch": 1.04, + "grad_norm": 0.48046875, + "learning_rate": 9.450226940320071e-05, + "loss": 2.6465, + "step": 24337 + }, + { + "epoch": 1.04, + "grad_norm": 0.41015625, + "learning_rate": 9.449557162909002e-05, + "loss": 2.4141, + "step": 24338 + }, + { + "epoch": 1.04, + "grad_norm": 0.4609375, + "learning_rate": 9.448887387974726e-05, + "loss": 2.5886, + "step": 24339 + }, + { + "epoch": 1.04, + "grad_norm": 0.439453125, + "learning_rate": 9.448217615520255e-05, + "loss": 2.5381, + "step": 24340 + }, + { + "epoch": 1.04, + "grad_norm": 0.47265625, + "learning_rate": 9.447547845548605e-05, + "loss": 2.3912, + "step": 24341 + }, + { + "epoch": 1.04, + "grad_norm": 0.42578125, + "learning_rate": 9.446878078062783e-05, + "loss": 2.368, + "step": 24342 + }, + { + "epoch": 1.04, + "grad_norm": 0.46484375, + "learning_rate": 9.446208313065812e-05, + "loss": 2.4585, + "step": 24343 + }, + { + "epoch": 1.04, + "grad_norm": 0.455078125, + "learning_rate": 9.445538550560701e-05, + "loss": 2.3279, + "step": 24344 + }, + { + "epoch": 1.04, + "grad_norm": 0.470703125, + "learning_rate": 9.444868790550463e-05, + "loss": 2.4415, + "step": 24345 + }, + { + "epoch": 1.04, + "grad_norm": 0.51171875, + "learning_rate": 9.444199033038111e-05, + "loss": 2.5116, + "step": 24346 + }, + { + "epoch": 1.04, + "grad_norm": 0.47265625, + "learning_rate": 9.443529278026661e-05, + "loss": 2.4631, + "step": 24347 + }, + { + "epoch": 1.04, + "grad_norm": 0.52734375, + "learning_rate": 9.442859525519126e-05, + "loss": 2.6056, + "step": 24348 + }, + { + "epoch": 1.04, + "grad_norm": 0.478515625, + "learning_rate": 9.442189775518522e-05, + "loss": 2.6327, + "step": 24349 + }, + { + "epoch": 1.04, + "grad_norm": 0.462890625, + "learning_rate": 9.441520028027855e-05, + "loss": 2.4399, + "step": 24350 + }, + { + "epoch": 1.04, + "grad_norm": 0.447265625, + "learning_rate": 9.440850283050149e-05, + "loss": 2.3407, + "step": 24351 + }, + { + "epoch": 1.04, + "grad_norm": 0.455078125, + "learning_rate": 9.440180540588411e-05, + "loss": 2.489, + "step": 24352 + }, + { + "epoch": 1.04, + "grad_norm": 0.42578125, + "learning_rate": 9.439510800645654e-05, + "loss": 2.4908, + "step": 24353 + }, + { + "epoch": 1.04, + "grad_norm": 0.421875, + "learning_rate": 9.438841063224894e-05, + "loss": 2.3275, + "step": 24354 + }, + { + "epoch": 1.04, + "grad_norm": 0.439453125, + "learning_rate": 9.438171328329146e-05, + "loss": 2.4812, + "step": 24355 + }, + { + "epoch": 1.04, + "grad_norm": 0.48046875, + "learning_rate": 9.43750159596142e-05, + "loss": 2.3811, + "step": 24356 + }, + { + "epoch": 1.04, + "grad_norm": 0.56640625, + "learning_rate": 9.436831866124733e-05, + "loss": 2.4522, + "step": 24357 + }, + { + "epoch": 1.04, + "grad_norm": 0.50390625, + "learning_rate": 9.436162138822092e-05, + "loss": 2.5162, + "step": 24358 + }, + { + "epoch": 1.04, + "grad_norm": 0.53125, + "learning_rate": 9.435492414056522e-05, + "loss": 2.5605, + "step": 24359 + }, + { + "epoch": 1.04, + "grad_norm": 0.51171875, + "learning_rate": 9.434822691831027e-05, + "loss": 2.4046, + "step": 24360 + }, + { + "epoch": 1.04, + "grad_norm": 0.494140625, + "learning_rate": 9.434152972148624e-05, + "loss": 2.4526, + "step": 24361 + }, + { + "epoch": 1.04, + "grad_norm": 0.4609375, + "learning_rate": 9.433483255012322e-05, + "loss": 2.4639, + "step": 24362 + }, + { + "epoch": 1.04, + "grad_norm": 0.4921875, + "learning_rate": 9.432813540425142e-05, + "loss": 2.4503, + "step": 24363 + }, + { + "epoch": 1.04, + "grad_norm": 0.515625, + "learning_rate": 9.432143828390093e-05, + "loss": 2.5141, + "step": 24364 + }, + { + "epoch": 1.04, + "grad_norm": 0.44921875, + "learning_rate": 9.431474118910192e-05, + "loss": 2.3558, + "step": 24365 + }, + { + "epoch": 1.04, + "grad_norm": 0.4609375, + "learning_rate": 9.430804411988441e-05, + "loss": 2.5761, + "step": 24366 + }, + { + "epoch": 1.04, + "grad_norm": 0.451171875, + "learning_rate": 9.430134707627872e-05, + "loss": 2.3547, + "step": 24367 + }, + { + "epoch": 1.04, + "grad_norm": 0.478515625, + "learning_rate": 9.429465005831485e-05, + "loss": 2.3755, + "step": 24368 + }, + { + "epoch": 1.04, + "grad_norm": 0.470703125, + "learning_rate": 9.428795306602297e-05, + "loss": 2.3959, + "step": 24369 + }, + { + "epoch": 1.04, + "grad_norm": 0.45703125, + "learning_rate": 9.42812560994332e-05, + "loss": 2.385, + "step": 24370 + }, + { + "epoch": 1.04, + "grad_norm": 0.4140625, + "learning_rate": 9.427455915857568e-05, + "loss": 2.4766, + "step": 24371 + }, + { + "epoch": 1.04, + "grad_norm": 0.46875, + "learning_rate": 9.426786224348058e-05, + "loss": 2.4181, + "step": 24372 + }, + { + "epoch": 1.04, + "grad_norm": 0.48828125, + "learning_rate": 9.426116535417802e-05, + "loss": 2.5652, + "step": 24373 + }, + { + "epoch": 1.04, + "grad_norm": 0.435546875, + "learning_rate": 9.42544684906981e-05, + "loss": 2.471, + "step": 24374 + }, + { + "epoch": 1.04, + "grad_norm": 0.490234375, + "learning_rate": 9.424777165307094e-05, + "loss": 2.3466, + "step": 24375 + }, + { + "epoch": 1.04, + "grad_norm": 0.50390625, + "learning_rate": 9.424107484132672e-05, + "loss": 2.572, + "step": 24376 + }, + { + "epoch": 1.04, + "grad_norm": 0.443359375, + "learning_rate": 9.423437805549558e-05, + "loss": 2.4176, + "step": 24377 + }, + { + "epoch": 1.04, + "grad_norm": 0.4140625, + "learning_rate": 9.422768129560762e-05, + "loss": 2.422, + "step": 24378 + }, + { + "epoch": 1.04, + "grad_norm": 0.4609375, + "learning_rate": 9.422098456169296e-05, + "loss": 2.5081, + "step": 24379 + }, + { + "epoch": 1.04, + "grad_norm": 0.423828125, + "learning_rate": 9.421428785378179e-05, + "loss": 2.3747, + "step": 24380 + }, + { + "epoch": 1.04, + "grad_norm": 0.5234375, + "learning_rate": 9.420759117190424e-05, + "loss": 2.4306, + "step": 24381 + }, + { + "epoch": 1.04, + "grad_norm": 0.486328125, + "learning_rate": 9.420089451609036e-05, + "loss": 2.5133, + "step": 24382 + }, + { + "epoch": 1.04, + "grad_norm": 0.5, + "learning_rate": 9.419419788637034e-05, + "loss": 2.5656, + "step": 24383 + }, + { + "epoch": 1.04, + "grad_norm": 0.4375, + "learning_rate": 9.418750128277432e-05, + "loss": 2.4293, + "step": 24384 + }, + { + "epoch": 1.04, + "grad_norm": 0.44921875, + "learning_rate": 9.418080470533242e-05, + "loss": 2.4278, + "step": 24385 + }, + { + "epoch": 1.04, + "grad_norm": 0.51953125, + "learning_rate": 9.417410815407476e-05, + "loss": 2.4218, + "step": 24386 + }, + { + "epoch": 1.04, + "grad_norm": 0.5234375, + "learning_rate": 9.416741162903149e-05, + "loss": 2.5652, + "step": 24387 + }, + { + "epoch": 1.04, + "grad_norm": 0.486328125, + "learning_rate": 9.416071513023273e-05, + "loss": 2.4078, + "step": 24388 + }, + { + "epoch": 1.04, + "grad_norm": 0.416015625, + "learning_rate": 9.415401865770866e-05, + "loss": 2.3561, + "step": 24389 + }, + { + "epoch": 1.04, + "grad_norm": 0.498046875, + "learning_rate": 9.414732221148935e-05, + "loss": 2.5048, + "step": 24390 + }, + { + "epoch": 1.04, + "grad_norm": 0.498046875, + "learning_rate": 9.414062579160491e-05, + "loss": 2.4023, + "step": 24391 + }, + { + "epoch": 1.04, + "grad_norm": 0.44921875, + "learning_rate": 9.413392939808554e-05, + "loss": 2.3333, + "step": 24392 + }, + { + "epoch": 1.04, + "grad_norm": 0.4765625, + "learning_rate": 9.412723303096136e-05, + "loss": 2.5052, + "step": 24393 + }, + { + "epoch": 1.04, + "grad_norm": 0.48828125, + "learning_rate": 9.412053669026246e-05, + "loss": 2.4707, + "step": 24394 + }, + { + "epoch": 1.04, + "grad_norm": 0.470703125, + "learning_rate": 9.411384037601899e-05, + "loss": 2.5942, + "step": 24395 + }, + { + "epoch": 1.04, + "grad_norm": 0.5078125, + "learning_rate": 9.410714408826111e-05, + "loss": 2.3619, + "step": 24396 + }, + { + "epoch": 1.04, + "grad_norm": 0.423828125, + "learning_rate": 9.410044782701894e-05, + "loss": 2.4256, + "step": 24397 + }, + { + "epoch": 1.04, + "grad_norm": 0.45703125, + "learning_rate": 9.40937515923226e-05, + "loss": 2.3819, + "step": 24398 + }, + { + "epoch": 1.04, + "grad_norm": 0.48828125, + "learning_rate": 9.408705538420217e-05, + "loss": 2.3494, + "step": 24399 + }, + { + "epoch": 1.04, + "grad_norm": 0.46484375, + "learning_rate": 9.408035920268786e-05, + "loss": 2.4659, + "step": 24400 + }, + { + "epoch": 1.04, + "grad_norm": 0.453125, + "learning_rate": 9.407366304780978e-05, + "loss": 2.4367, + "step": 24401 + }, + { + "epoch": 1.04, + "grad_norm": 0.5078125, + "learning_rate": 9.406696691959805e-05, + "loss": 2.6227, + "step": 24402 + }, + { + "epoch": 1.04, + "grad_norm": 0.51953125, + "learning_rate": 9.406027081808277e-05, + "loss": 2.6146, + "step": 24403 + }, + { + "epoch": 1.04, + "grad_norm": 0.5, + "learning_rate": 9.405357474329411e-05, + "loss": 2.5352, + "step": 24404 + }, + { + "epoch": 1.04, + "grad_norm": 0.48828125, + "learning_rate": 9.404687869526224e-05, + "loss": 2.4667, + "step": 24405 + }, + { + "epoch": 1.04, + "grad_norm": 0.44140625, + "learning_rate": 9.404018267401719e-05, + "loss": 2.3628, + "step": 24406 + }, + { + "epoch": 1.04, + "grad_norm": 0.474609375, + "learning_rate": 9.403348667958914e-05, + "loss": 2.4269, + "step": 24407 + }, + { + "epoch": 1.04, + "grad_norm": 0.515625, + "learning_rate": 9.402679071200824e-05, + "loss": 2.4541, + "step": 24408 + }, + { + "epoch": 1.04, + "grad_norm": 0.49609375, + "learning_rate": 9.402009477130458e-05, + "loss": 2.3848, + "step": 24409 + }, + { + "epoch": 1.04, + "grad_norm": 0.5546875, + "learning_rate": 9.401339885750832e-05, + "loss": 2.5832, + "step": 24410 + }, + { + "epoch": 1.04, + "grad_norm": 0.482421875, + "learning_rate": 9.400670297064958e-05, + "loss": 2.6012, + "step": 24411 + }, + { + "epoch": 1.04, + "grad_norm": 0.453125, + "learning_rate": 9.400000711075847e-05, + "loss": 2.4629, + "step": 24412 + }, + { + "epoch": 1.04, + "grad_norm": 0.44921875, + "learning_rate": 9.399331127786518e-05, + "loss": 2.4424, + "step": 24413 + }, + { + "epoch": 1.04, + "grad_norm": 0.482421875, + "learning_rate": 9.398661547199976e-05, + "loss": 2.425, + "step": 24414 + }, + { + "epoch": 1.04, + "grad_norm": 0.5546875, + "learning_rate": 9.397991969319235e-05, + "loss": 2.42, + "step": 24415 + }, + { + "epoch": 1.04, + "grad_norm": 0.455078125, + "learning_rate": 9.397322394147313e-05, + "loss": 2.4649, + "step": 24416 + }, + { + "epoch": 1.04, + "grad_norm": 0.478515625, + "learning_rate": 9.396652821687221e-05, + "loss": 2.4231, + "step": 24417 + }, + { + "epoch": 1.04, + "grad_norm": 0.400390625, + "learning_rate": 9.39598325194197e-05, + "loss": 2.332, + "step": 24418 + }, + { + "epoch": 1.04, + "grad_norm": 0.48828125, + "learning_rate": 9.395313684914569e-05, + "loss": 2.5175, + "step": 24419 + }, + { + "epoch": 1.04, + "grad_norm": 0.51171875, + "learning_rate": 9.39464412060804e-05, + "loss": 2.5242, + "step": 24420 + }, + { + "epoch": 1.04, + "grad_norm": 0.443359375, + "learning_rate": 9.393974559025393e-05, + "loss": 2.3942, + "step": 24421 + }, + { + "epoch": 1.04, + "grad_norm": 0.47265625, + "learning_rate": 9.393305000169637e-05, + "loss": 2.445, + "step": 24422 + }, + { + "epoch": 1.04, + "grad_norm": 0.50390625, + "learning_rate": 9.392635444043785e-05, + "loss": 2.4081, + "step": 24423 + }, + { + "epoch": 1.04, + "grad_norm": 0.46875, + "learning_rate": 9.391965890650853e-05, + "loss": 2.3816, + "step": 24424 + }, + { + "epoch": 1.04, + "grad_norm": 0.439453125, + "learning_rate": 9.391296339993852e-05, + "loss": 2.5716, + "step": 24425 + }, + { + "epoch": 1.04, + "grad_norm": 0.54296875, + "learning_rate": 9.390626792075796e-05, + "loss": 2.4997, + "step": 24426 + }, + { + "epoch": 1.04, + "grad_norm": 0.4609375, + "learning_rate": 9.389957246899695e-05, + "loss": 2.6484, + "step": 24427 + }, + { + "epoch": 1.04, + "grad_norm": 0.5078125, + "learning_rate": 9.389287704468564e-05, + "loss": 2.6026, + "step": 24428 + }, + { + "epoch": 1.04, + "grad_norm": 0.427734375, + "learning_rate": 9.388618164785419e-05, + "loss": 2.271, + "step": 24429 + }, + { + "epoch": 1.04, + "grad_norm": 0.455078125, + "learning_rate": 9.387948627853265e-05, + "loss": 2.4584, + "step": 24430 + }, + { + "epoch": 1.04, + "grad_norm": 0.4921875, + "learning_rate": 9.387279093675114e-05, + "loss": 2.4548, + "step": 24431 + }, + { + "epoch": 1.04, + "grad_norm": 0.431640625, + "learning_rate": 9.386609562253991e-05, + "loss": 2.4918, + "step": 24432 + }, + { + "epoch": 1.04, + "grad_norm": 0.455078125, + "learning_rate": 9.385940033592898e-05, + "loss": 2.4063, + "step": 24433 + }, + { + "epoch": 1.04, + "grad_norm": 0.431640625, + "learning_rate": 9.385270507694851e-05, + "loss": 2.3059, + "step": 24434 + }, + { + "epoch": 1.04, + "grad_norm": 0.466796875, + "learning_rate": 9.38460098456286e-05, + "loss": 2.5251, + "step": 24435 + }, + { + "epoch": 1.04, + "grad_norm": 0.482421875, + "learning_rate": 9.38393146419994e-05, + "loss": 2.3839, + "step": 24436 + }, + { + "epoch": 1.04, + "grad_norm": 0.48828125, + "learning_rate": 9.383261946609107e-05, + "loss": 2.4708, + "step": 24437 + }, + { + "epoch": 1.04, + "grad_norm": 0.466796875, + "learning_rate": 9.382592431793368e-05, + "loss": 2.4922, + "step": 24438 + }, + { + "epoch": 1.04, + "grad_norm": 0.4296875, + "learning_rate": 9.381922919755733e-05, + "loss": 2.3951, + "step": 24439 + }, + { + "epoch": 1.04, + "grad_norm": 0.4375, + "learning_rate": 9.381253410499225e-05, + "loss": 2.4864, + "step": 24440 + }, + { + "epoch": 1.04, + "grad_norm": 0.478515625, + "learning_rate": 9.380583904026847e-05, + "loss": 2.3749, + "step": 24441 + }, + { + "epoch": 1.04, + "grad_norm": 0.54296875, + "learning_rate": 9.379914400341617e-05, + "loss": 2.6355, + "step": 24442 + }, + { + "epoch": 1.04, + "grad_norm": 0.5390625, + "learning_rate": 9.379244899446543e-05, + "loss": 2.459, + "step": 24443 + }, + { + "epoch": 1.04, + "grad_norm": 0.44140625, + "learning_rate": 9.378575401344642e-05, + "loss": 2.4434, + "step": 24444 + }, + { + "epoch": 1.04, + "grad_norm": 0.5078125, + "learning_rate": 9.377905906038927e-05, + "loss": 2.4508, + "step": 24445 + }, + { + "epoch": 1.04, + "grad_norm": 0.49609375, + "learning_rate": 9.377236413532405e-05, + "loss": 2.641, + "step": 24446 + }, + { + "epoch": 1.04, + "grad_norm": 0.470703125, + "learning_rate": 9.376566923828089e-05, + "loss": 2.5182, + "step": 24447 + }, + { + "epoch": 1.04, + "grad_norm": 0.515625, + "learning_rate": 9.375897436928998e-05, + "loss": 2.509, + "step": 24448 + }, + { + "epoch": 1.04, + "grad_norm": 0.40234375, + "learning_rate": 9.37522795283814e-05, + "loss": 2.5817, + "step": 24449 + }, + { + "epoch": 1.04, + "grad_norm": 0.46484375, + "learning_rate": 9.374558471558528e-05, + "loss": 2.3528, + "step": 24450 + }, + { + "epoch": 1.04, + "grad_norm": 0.48046875, + "learning_rate": 9.373888993093171e-05, + "loss": 2.6035, + "step": 24451 + }, + { + "epoch": 1.04, + "grad_norm": 0.490234375, + "learning_rate": 9.373219517445089e-05, + "loss": 2.575, + "step": 24452 + }, + { + "epoch": 1.04, + "grad_norm": 0.498046875, + "learning_rate": 9.372550044617289e-05, + "loss": 2.4283, + "step": 24453 + }, + { + "epoch": 1.04, + "grad_norm": 0.474609375, + "learning_rate": 9.371880574612786e-05, + "loss": 2.4107, + "step": 24454 + }, + { + "epoch": 1.04, + "grad_norm": 0.474609375, + "learning_rate": 9.371211107434585e-05, + "loss": 2.3005, + "step": 24455 + }, + { + "epoch": 1.04, + "grad_norm": 0.498046875, + "learning_rate": 9.37054164308571e-05, + "loss": 2.4798, + "step": 24456 + }, + { + "epoch": 1.04, + "grad_norm": 0.482421875, + "learning_rate": 9.369872181569166e-05, + "loss": 2.5021, + "step": 24457 + }, + { + "epoch": 1.04, + "grad_norm": 0.48046875, + "learning_rate": 9.369202722887967e-05, + "loss": 2.5855, + "step": 24458 + }, + { + "epoch": 1.04, + "grad_norm": 0.478515625, + "learning_rate": 9.368533267045123e-05, + "loss": 2.4986, + "step": 24459 + }, + { + "epoch": 1.04, + "grad_norm": 0.48046875, + "learning_rate": 9.36786381404365e-05, + "loss": 2.452, + "step": 24460 + }, + { + "epoch": 1.04, + "grad_norm": 0.48046875, + "learning_rate": 9.36719436388656e-05, + "loss": 2.6029, + "step": 24461 + }, + { + "epoch": 1.04, + "grad_norm": 0.451171875, + "learning_rate": 9.366524916576864e-05, + "loss": 2.5145, + "step": 24462 + }, + { + "epoch": 1.04, + "grad_norm": 0.474609375, + "learning_rate": 9.36585547211757e-05, + "loss": 2.5351, + "step": 24463 + }, + { + "epoch": 1.04, + "grad_norm": 0.4453125, + "learning_rate": 9.3651860305117e-05, + "loss": 2.4256, + "step": 24464 + }, + { + "epoch": 1.04, + "grad_norm": 0.50390625, + "learning_rate": 9.36451659176226e-05, + "loss": 2.5726, + "step": 24465 + }, + { + "epoch": 1.04, + "grad_norm": 0.451171875, + "learning_rate": 9.363847155872261e-05, + "loss": 2.5597, + "step": 24466 + }, + { + "epoch": 1.04, + "grad_norm": 0.46484375, + "learning_rate": 9.363177722844718e-05, + "loss": 2.5443, + "step": 24467 + }, + { + "epoch": 1.04, + "grad_norm": 0.486328125, + "learning_rate": 9.362508292682643e-05, + "loss": 2.5239, + "step": 24468 + }, + { + "epoch": 1.04, + "grad_norm": 0.5234375, + "learning_rate": 9.36183886538905e-05, + "loss": 2.4524, + "step": 24469 + }, + { + "epoch": 1.04, + "grad_norm": 0.423828125, + "learning_rate": 9.361169440966946e-05, + "loss": 2.3906, + "step": 24470 + }, + { + "epoch": 1.04, + "grad_norm": 0.55859375, + "learning_rate": 9.360500019419343e-05, + "loss": 2.5843, + "step": 24471 + }, + { + "epoch": 1.04, + "grad_norm": 0.458984375, + "learning_rate": 9.359830600749261e-05, + "loss": 2.5409, + "step": 24472 + }, + { + "epoch": 1.04, + "grad_norm": 0.46484375, + "learning_rate": 9.359161184959709e-05, + "loss": 2.326, + "step": 24473 + }, + { + "epoch": 1.04, + "grad_norm": 0.484375, + "learning_rate": 9.358491772053694e-05, + "loss": 2.6357, + "step": 24474 + }, + { + "epoch": 1.04, + "grad_norm": 0.453125, + "learning_rate": 9.35782236203423e-05, + "loss": 2.4882, + "step": 24475 + }, + { + "epoch": 1.04, + "grad_norm": 0.4296875, + "learning_rate": 9.357152954904333e-05, + "loss": 2.4453, + "step": 24476 + }, + { + "epoch": 1.04, + "grad_norm": 0.515625, + "learning_rate": 9.356483550667013e-05, + "loss": 2.3428, + "step": 24477 + }, + { + "epoch": 1.04, + "grad_norm": 0.50390625, + "learning_rate": 9.355814149325284e-05, + "loss": 2.4099, + "step": 24478 + }, + { + "epoch": 1.04, + "grad_norm": 0.43359375, + "learning_rate": 9.355144750882149e-05, + "loss": 2.4207, + "step": 24479 + }, + { + "epoch": 1.04, + "grad_norm": 0.455078125, + "learning_rate": 9.354475355340634e-05, + "loss": 2.4992, + "step": 24480 + }, + { + "epoch": 1.04, + "grad_norm": 0.4921875, + "learning_rate": 9.353805962703741e-05, + "loss": 2.5235, + "step": 24481 + }, + { + "epoch": 1.04, + "grad_norm": 0.431640625, + "learning_rate": 9.353136572974485e-05, + "loss": 2.5201, + "step": 24482 + }, + { + "epoch": 1.04, + "grad_norm": 0.453125, + "learning_rate": 9.352467186155877e-05, + "loss": 2.4739, + "step": 24483 + }, + { + "epoch": 1.04, + "grad_norm": 0.400390625, + "learning_rate": 9.351797802250933e-05, + "loss": 2.4159, + "step": 24484 + }, + { + "epoch": 1.04, + "grad_norm": 0.466796875, + "learning_rate": 9.35112842126266e-05, + "loss": 2.4587, + "step": 24485 + }, + { + "epoch": 1.04, + "grad_norm": 0.4921875, + "learning_rate": 9.350459043194074e-05, + "loss": 2.4071, + "step": 24486 + }, + { + "epoch": 1.04, + "grad_norm": 0.51953125, + "learning_rate": 9.34978966804818e-05, + "loss": 2.4985, + "step": 24487 + }, + { + "epoch": 1.04, + "grad_norm": 0.486328125, + "learning_rate": 9.349120295828e-05, + "loss": 2.4676, + "step": 24488 + }, + { + "epoch": 1.04, + "grad_norm": 0.455078125, + "learning_rate": 9.34845092653654e-05, + "loss": 2.4845, + "step": 24489 + }, + { + "epoch": 1.04, + "grad_norm": 0.5234375, + "learning_rate": 9.347781560176812e-05, + "loss": 2.5216, + "step": 24490 + }, + { + "epoch": 1.04, + "grad_norm": 0.47265625, + "learning_rate": 9.34711219675183e-05, + "loss": 2.61, + "step": 24491 + }, + { + "epoch": 1.04, + "grad_norm": 0.55859375, + "learning_rate": 9.346442836264598e-05, + "loss": 2.5934, + "step": 24492 + }, + { + "epoch": 1.04, + "grad_norm": 0.41015625, + "learning_rate": 9.34577347871814e-05, + "loss": 2.5102, + "step": 24493 + }, + { + "epoch": 1.04, + "grad_norm": 0.4765625, + "learning_rate": 9.345104124115464e-05, + "loss": 2.3939, + "step": 24494 + }, + { + "epoch": 1.04, + "grad_norm": 0.47265625, + "learning_rate": 9.344434772459578e-05, + "loss": 2.4577, + "step": 24495 + }, + { + "epoch": 1.04, + "grad_norm": 0.48828125, + "learning_rate": 9.343765423753494e-05, + "loss": 2.498, + "step": 24496 + }, + { + "epoch": 1.04, + "grad_norm": 0.478515625, + "learning_rate": 9.343096078000226e-05, + "loss": 2.4175, + "step": 24497 + }, + { + "epoch": 1.04, + "grad_norm": 0.400390625, + "learning_rate": 9.342426735202788e-05, + "loss": 2.4357, + "step": 24498 + }, + { + "epoch": 1.04, + "grad_norm": 0.416015625, + "learning_rate": 9.34175739536419e-05, + "loss": 2.455, + "step": 24499 + }, + { + "epoch": 1.04, + "grad_norm": 0.4375, + "learning_rate": 9.341088058487438e-05, + "loss": 2.6202, + "step": 24500 + }, + { + "epoch": 1.04, + "grad_norm": 0.4609375, + "learning_rate": 9.340418724575552e-05, + "loss": 2.5277, + "step": 24501 + }, + { + "epoch": 1.04, + "grad_norm": 0.51953125, + "learning_rate": 9.339749393631543e-05, + "loss": 2.473, + "step": 24502 + }, + { + "epoch": 1.04, + "grad_norm": 0.447265625, + "learning_rate": 9.339080065658419e-05, + "loss": 2.4311, + "step": 24503 + }, + { + "epoch": 1.04, + "grad_norm": 0.4453125, + "learning_rate": 9.33841074065919e-05, + "loss": 2.3636, + "step": 24504 + }, + { + "epoch": 1.04, + "grad_norm": 0.4375, + "learning_rate": 9.337741418636874e-05, + "loss": 2.5355, + "step": 24505 + }, + { + "epoch": 1.04, + "grad_norm": 0.494140625, + "learning_rate": 9.33707209959448e-05, + "loss": 2.4527, + "step": 24506 + }, + { + "epoch": 1.04, + "grad_norm": 0.49609375, + "learning_rate": 9.33640278353502e-05, + "loss": 2.5785, + "step": 24507 + }, + { + "epoch": 1.04, + "grad_norm": 0.447265625, + "learning_rate": 9.3357334704615e-05, + "loss": 2.4797, + "step": 24508 + }, + { + "epoch": 1.04, + "grad_norm": 0.4453125, + "learning_rate": 9.33506416037694e-05, + "loss": 2.4612, + "step": 24509 + }, + { + "epoch": 1.04, + "grad_norm": 0.421875, + "learning_rate": 9.33439485328435e-05, + "loss": 2.3998, + "step": 24510 + }, + { + "epoch": 1.04, + "grad_norm": 0.44921875, + "learning_rate": 9.333725549186739e-05, + "loss": 2.4966, + "step": 24511 + }, + { + "epoch": 1.04, + "grad_norm": 0.478515625, + "learning_rate": 9.333056248087117e-05, + "loss": 2.4419, + "step": 24512 + }, + { + "epoch": 1.04, + "grad_norm": 0.515625, + "learning_rate": 9.3323869499885e-05, + "loss": 2.5257, + "step": 24513 + }, + { + "epoch": 1.04, + "grad_norm": 0.45703125, + "learning_rate": 9.331717654893896e-05, + "loss": 2.4513, + "step": 24514 + }, + { + "epoch": 1.04, + "grad_norm": 0.470703125, + "learning_rate": 9.331048362806324e-05, + "loss": 2.4201, + "step": 24515 + }, + { + "epoch": 1.04, + "grad_norm": 0.43359375, + "learning_rate": 9.33037907372878e-05, + "loss": 2.471, + "step": 24516 + }, + { + "epoch": 1.04, + "grad_norm": 0.5, + "learning_rate": 9.329709787664292e-05, + "loss": 2.4984, + "step": 24517 + }, + { + "epoch": 1.04, + "grad_norm": 0.52734375, + "learning_rate": 9.329040504615866e-05, + "loss": 2.3901, + "step": 24518 + }, + { + "epoch": 1.04, + "grad_norm": 0.466796875, + "learning_rate": 9.328371224586512e-05, + "loss": 2.5777, + "step": 24519 + }, + { + "epoch": 1.04, + "grad_norm": 0.53515625, + "learning_rate": 9.327701947579239e-05, + "loss": 2.5453, + "step": 24520 + }, + { + "epoch": 1.04, + "grad_norm": 0.46875, + "learning_rate": 9.327032673597064e-05, + "loss": 2.49, + "step": 24521 + }, + { + "epoch": 1.04, + "grad_norm": 0.470703125, + "learning_rate": 9.326363402642994e-05, + "loss": 2.4815, + "step": 24522 + }, + { + "epoch": 1.04, + "grad_norm": 0.462890625, + "learning_rate": 9.325694134720047e-05, + "loss": 2.2457, + "step": 24523 + }, + { + "epoch": 1.05, + "grad_norm": 0.416015625, + "learning_rate": 9.325024869831223e-05, + "loss": 2.4878, + "step": 24524 + }, + { + "epoch": 1.05, + "grad_norm": 0.43359375, + "learning_rate": 9.324355607979545e-05, + "loss": 2.31, + "step": 24525 + }, + { + "epoch": 1.05, + "grad_norm": 0.46875, + "learning_rate": 9.32368634916802e-05, + "loss": 2.5349, + "step": 24526 + }, + { + "epoch": 1.05, + "grad_norm": 0.416015625, + "learning_rate": 9.32301709339966e-05, + "loss": 2.4933, + "step": 24527 + }, + { + "epoch": 1.05, + "grad_norm": 0.490234375, + "learning_rate": 9.322347840677471e-05, + "loss": 2.554, + "step": 24528 + }, + { + "epoch": 1.05, + "grad_norm": 0.396484375, + "learning_rate": 9.321678591004472e-05, + "loss": 2.3341, + "step": 24529 + }, + { + "epoch": 1.05, + "grad_norm": 0.43359375, + "learning_rate": 9.321009344383672e-05, + "loss": 2.456, + "step": 24530 + }, + { + "epoch": 1.05, + "grad_norm": 0.455078125, + "learning_rate": 9.320340100818081e-05, + "loss": 2.4142, + "step": 24531 + }, + { + "epoch": 1.05, + "grad_norm": 0.458984375, + "learning_rate": 9.319670860310709e-05, + "loss": 2.384, + "step": 24532 + }, + { + "epoch": 1.05, + "grad_norm": 0.458984375, + "learning_rate": 9.319001622864574e-05, + "loss": 2.4644, + "step": 24533 + }, + { + "epoch": 1.05, + "grad_norm": 0.439453125, + "learning_rate": 9.318332388482683e-05, + "loss": 2.467, + "step": 24534 + }, + { + "epoch": 1.05, + "grad_norm": 0.453125, + "learning_rate": 9.317663157168044e-05, + "loss": 2.4316, + "step": 24535 + }, + { + "epoch": 1.05, + "grad_norm": 0.47265625, + "learning_rate": 9.316993928923671e-05, + "loss": 2.4242, + "step": 24536 + }, + { + "epoch": 1.05, + "grad_norm": 0.53515625, + "learning_rate": 9.316324703752577e-05, + "loss": 2.4854, + "step": 24537 + }, + { + "epoch": 1.05, + "grad_norm": 0.52734375, + "learning_rate": 9.315655481657772e-05, + "loss": 2.4627, + "step": 24538 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.314986262642268e-05, + "loss": 2.5061, + "step": 24539 + }, + { + "epoch": 1.05, + "grad_norm": 0.48828125, + "learning_rate": 9.314317046709072e-05, + "loss": 2.5269, + "step": 24540 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.313647833861202e-05, + "loss": 2.452, + "step": 24541 + }, + { + "epoch": 1.05, + "grad_norm": 0.466796875, + "learning_rate": 9.312978624101668e-05, + "loss": 2.419, + "step": 24542 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.312309417433476e-05, + "loss": 2.4089, + "step": 24543 + }, + { + "epoch": 1.05, + "grad_norm": 0.486328125, + "learning_rate": 9.31164021385964e-05, + "loss": 2.5164, + "step": 24544 + }, + { + "epoch": 1.05, + "grad_norm": 0.400390625, + "learning_rate": 9.310971013383172e-05, + "loss": 2.5356, + "step": 24545 + }, + { + "epoch": 1.05, + "grad_norm": 0.46875, + "learning_rate": 9.310301816007084e-05, + "loss": 2.5481, + "step": 24546 + }, + { + "epoch": 1.05, + "grad_norm": 0.390625, + "learning_rate": 9.309632621734383e-05, + "loss": 2.4783, + "step": 24547 + }, + { + "epoch": 1.05, + "grad_norm": 0.53125, + "learning_rate": 9.308963430568083e-05, + "loss": 2.4943, + "step": 24548 + }, + { + "epoch": 1.05, + "grad_norm": 0.5, + "learning_rate": 9.308294242511198e-05, + "loss": 2.5036, + "step": 24549 + }, + { + "epoch": 1.05, + "grad_norm": 0.453125, + "learning_rate": 9.307625057566735e-05, + "loss": 2.3916, + "step": 24550 + }, + { + "epoch": 1.05, + "grad_norm": 0.5, + "learning_rate": 9.306955875737708e-05, + "loss": 2.5951, + "step": 24551 + }, + { + "epoch": 1.05, + "grad_norm": 0.453125, + "learning_rate": 9.30628669702712e-05, + "loss": 2.417, + "step": 24552 + }, + { + "epoch": 1.05, + "grad_norm": 0.412109375, + "learning_rate": 9.305617521437994e-05, + "loss": 2.4528, + "step": 24553 + }, + { + "epoch": 1.05, + "grad_norm": 0.51171875, + "learning_rate": 9.304948348973335e-05, + "loss": 2.578, + "step": 24554 + }, + { + "epoch": 1.05, + "grad_norm": 0.435546875, + "learning_rate": 9.304279179636153e-05, + "loss": 2.6201, + "step": 24555 + }, + { + "epoch": 1.05, + "grad_norm": 0.50390625, + "learning_rate": 9.303610013429458e-05, + "loss": 2.3845, + "step": 24556 + }, + { + "epoch": 1.05, + "grad_norm": 0.47265625, + "learning_rate": 9.302940850356266e-05, + "loss": 2.4023, + "step": 24557 + }, + { + "epoch": 1.05, + "grad_norm": 0.455078125, + "learning_rate": 9.302271690419588e-05, + "loss": 2.4058, + "step": 24558 + }, + { + "epoch": 1.05, + "grad_norm": 0.478515625, + "learning_rate": 9.301602533622431e-05, + "loss": 2.458, + "step": 24559 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.300933379967805e-05, + "loss": 2.5195, + "step": 24560 + }, + { + "epoch": 1.05, + "grad_norm": 0.474609375, + "learning_rate": 9.300264229458724e-05, + "loss": 2.5354, + "step": 24561 + }, + { + "epoch": 1.05, + "grad_norm": 0.42578125, + "learning_rate": 9.2995950820982e-05, + "loss": 2.5364, + "step": 24562 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.29892593788924e-05, + "loss": 2.459, + "step": 24563 + }, + { + "epoch": 1.05, + "grad_norm": 0.44921875, + "learning_rate": 9.298256796834857e-05, + "loss": 2.3797, + "step": 24564 + }, + { + "epoch": 1.05, + "grad_norm": 0.49609375, + "learning_rate": 9.297587658938063e-05, + "loss": 2.5252, + "step": 24565 + }, + { + "epoch": 1.05, + "grad_norm": 0.53125, + "learning_rate": 9.29691852420187e-05, + "loss": 2.4887, + "step": 24566 + }, + { + "epoch": 1.05, + "grad_norm": 0.494140625, + "learning_rate": 9.296249392629285e-05, + "loss": 2.4059, + "step": 24567 + }, + { + "epoch": 1.05, + "grad_norm": 0.419921875, + "learning_rate": 9.295580264223317e-05, + "loss": 2.4299, + "step": 24568 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.294911138986984e-05, + "loss": 2.3912, + "step": 24569 + }, + { + "epoch": 1.05, + "grad_norm": 0.48828125, + "learning_rate": 9.294242016923294e-05, + "loss": 2.5545, + "step": 24570 + }, + { + "epoch": 1.05, + "grad_norm": 0.515625, + "learning_rate": 9.293572898035257e-05, + "loss": 2.4271, + "step": 24571 + }, + { + "epoch": 1.05, + "grad_norm": 0.45703125, + "learning_rate": 9.292903782325878e-05, + "loss": 2.3746, + "step": 24572 + }, + { + "epoch": 1.05, + "grad_norm": 0.494140625, + "learning_rate": 9.29223466979818e-05, + "loss": 2.5172, + "step": 24573 + }, + { + "epoch": 1.05, + "grad_norm": 0.51953125, + "learning_rate": 9.291565560455167e-05, + "loss": 2.6314, + "step": 24574 + }, + { + "epoch": 1.05, + "grad_norm": 0.52734375, + "learning_rate": 9.290896454299849e-05, + "loss": 2.4694, + "step": 24575 + }, + { + "epoch": 1.05, + "grad_norm": 0.474609375, + "learning_rate": 9.290227351335236e-05, + "loss": 2.5075, + "step": 24576 + }, + { + "epoch": 1.05, + "grad_norm": 0.484375, + "learning_rate": 9.289558251564343e-05, + "loss": 2.2993, + "step": 24577 + }, + { + "epoch": 1.05, + "grad_norm": 0.4609375, + "learning_rate": 9.288889154990176e-05, + "loss": 2.5375, + "step": 24578 + }, + { + "epoch": 1.05, + "grad_norm": 0.44921875, + "learning_rate": 9.288220061615748e-05, + "loss": 2.3903, + "step": 24579 + }, + { + "epoch": 1.05, + "grad_norm": 0.51953125, + "learning_rate": 9.287550971444071e-05, + "loss": 2.4987, + "step": 24580 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.286881884478152e-05, + "loss": 2.4169, + "step": 24581 + }, + { + "epoch": 1.05, + "grad_norm": 0.58203125, + "learning_rate": 9.286212800721009e-05, + "loss": 2.5237, + "step": 24582 + }, + { + "epoch": 1.05, + "grad_norm": 0.462890625, + "learning_rate": 9.285543720175645e-05, + "loss": 2.4965, + "step": 24583 + }, + { + "epoch": 1.05, + "grad_norm": 0.53125, + "learning_rate": 9.28487464284507e-05, + "loss": 2.3796, + "step": 24584 + }, + { + "epoch": 1.05, + "grad_norm": 0.490234375, + "learning_rate": 9.2842055687323e-05, + "loss": 2.5501, + "step": 24585 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.283536497840344e-05, + "loss": 2.4805, + "step": 24586 + }, + { + "epoch": 1.05, + "grad_norm": 0.50390625, + "learning_rate": 9.282867430172212e-05, + "loss": 2.5485, + "step": 24587 + }, + { + "epoch": 1.05, + "grad_norm": 0.43359375, + "learning_rate": 9.282198365730911e-05, + "loss": 2.2897, + "step": 24588 + }, + { + "epoch": 1.05, + "grad_norm": 0.4609375, + "learning_rate": 9.281529304519458e-05, + "loss": 2.4429, + "step": 24589 + }, + { + "epoch": 1.05, + "grad_norm": 0.4453125, + "learning_rate": 9.280860246540864e-05, + "loss": 2.4099, + "step": 24590 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.280191191798133e-05, + "loss": 2.4265, + "step": 24591 + }, + { + "epoch": 1.05, + "grad_norm": 0.484375, + "learning_rate": 9.279522140294275e-05, + "loss": 2.4876, + "step": 24592 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.278853092032308e-05, + "loss": 2.4675, + "step": 24593 + }, + { + "epoch": 1.05, + "grad_norm": 0.494140625, + "learning_rate": 9.278184047015238e-05, + "loss": 2.4924, + "step": 24594 + }, + { + "epoch": 1.05, + "grad_norm": 0.53515625, + "learning_rate": 9.277515005246076e-05, + "loss": 2.5527, + "step": 24595 + }, + { + "epoch": 1.05, + "grad_norm": 0.458984375, + "learning_rate": 9.276845966727831e-05, + "loss": 2.4295, + "step": 24596 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.276176931463516e-05, + "loss": 2.5026, + "step": 24597 + }, + { + "epoch": 1.05, + "grad_norm": 0.50390625, + "learning_rate": 9.275507899456142e-05, + "loss": 2.5547, + "step": 24598 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.274838870708718e-05, + "loss": 2.5202, + "step": 24599 + }, + { + "epoch": 1.05, + "grad_norm": 0.455078125, + "learning_rate": 9.274169845224249e-05, + "loss": 2.5502, + "step": 24600 + }, + { + "epoch": 1.05, + "grad_norm": 0.453125, + "learning_rate": 9.273500823005754e-05, + "loss": 2.3634, + "step": 24601 + }, + { + "epoch": 1.05, + "grad_norm": 0.50390625, + "learning_rate": 9.272831804056239e-05, + "loss": 2.5151, + "step": 24602 + }, + { + "epoch": 1.05, + "grad_norm": 0.50390625, + "learning_rate": 9.272162788378715e-05, + "loss": 2.4548, + "step": 24603 + }, + { + "epoch": 1.05, + "grad_norm": 0.484375, + "learning_rate": 9.271493775976193e-05, + "loss": 2.4273, + "step": 24604 + }, + { + "epoch": 1.05, + "grad_norm": 0.423828125, + "learning_rate": 9.270824766851682e-05, + "loss": 2.4103, + "step": 24605 + }, + { + "epoch": 1.05, + "grad_norm": 0.462890625, + "learning_rate": 9.270155761008196e-05, + "loss": 2.4955, + "step": 24606 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.26948675844874e-05, + "loss": 2.6257, + "step": 24607 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.268817759176325e-05, + "loss": 2.4223, + "step": 24608 + }, + { + "epoch": 1.05, + "grad_norm": 0.494140625, + "learning_rate": 9.268148763193964e-05, + "loss": 2.6152, + "step": 24609 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.267479770504667e-05, + "loss": 2.3747, + "step": 24610 + }, + { + "epoch": 1.05, + "grad_norm": 0.466796875, + "learning_rate": 9.266810781111443e-05, + "loss": 2.4427, + "step": 24611 + }, + { + "epoch": 1.05, + "grad_norm": 0.4296875, + "learning_rate": 9.266141795017306e-05, + "loss": 2.3959, + "step": 24612 + }, + { + "epoch": 1.05, + "grad_norm": 0.462890625, + "learning_rate": 9.265472812225255e-05, + "loss": 2.3589, + "step": 24613 + }, + { + "epoch": 1.05, + "grad_norm": 0.47265625, + "learning_rate": 9.264803832738315e-05, + "loss": 2.4753, + "step": 24614 + }, + { + "epoch": 1.05, + "grad_norm": 0.4609375, + "learning_rate": 9.264134856559487e-05, + "loss": 2.4899, + "step": 24615 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.263465883691782e-05, + "loss": 2.4037, + "step": 24616 + }, + { + "epoch": 1.05, + "grad_norm": 0.404296875, + "learning_rate": 9.26279691413821e-05, + "loss": 2.4064, + "step": 24617 + }, + { + "epoch": 1.05, + "grad_norm": 0.48046875, + "learning_rate": 9.262127947901784e-05, + "loss": 2.4631, + "step": 24618 + }, + { + "epoch": 1.05, + "grad_norm": 0.4296875, + "learning_rate": 9.261458984985513e-05, + "loss": 2.3906, + "step": 24619 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.260790025392409e-05, + "loss": 2.4354, + "step": 24620 + }, + { + "epoch": 1.05, + "grad_norm": 0.490234375, + "learning_rate": 9.260121069125474e-05, + "loss": 2.4084, + "step": 24621 + }, + { + "epoch": 1.05, + "grad_norm": 0.5, + "learning_rate": 9.259452116187729e-05, + "loss": 2.6175, + "step": 24622 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.258783166582177e-05, + "loss": 2.4899, + "step": 24623 + }, + { + "epoch": 1.05, + "grad_norm": 0.482421875, + "learning_rate": 9.25811422031183e-05, + "loss": 2.4253, + "step": 24624 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.257445277379696e-05, + "loss": 2.5144, + "step": 24625 + }, + { + "epoch": 1.05, + "grad_norm": 0.419921875, + "learning_rate": 9.256776337788789e-05, + "loss": 2.4643, + "step": 24626 + }, + { + "epoch": 1.05, + "grad_norm": 0.4453125, + "learning_rate": 9.256107401542116e-05, + "loss": 2.3515, + "step": 24627 + }, + { + "epoch": 1.05, + "grad_norm": 0.5390625, + "learning_rate": 9.255438468642691e-05, + "loss": 2.4528, + "step": 24628 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.254769539093514e-05, + "loss": 2.5205, + "step": 24629 + }, + { + "epoch": 1.05, + "grad_norm": 0.54296875, + "learning_rate": 9.254100612897608e-05, + "loss": 2.4503, + "step": 24630 + }, + { + "epoch": 1.05, + "grad_norm": 0.482421875, + "learning_rate": 9.253431690057974e-05, + "loss": 2.5103, + "step": 24631 + }, + { + "epoch": 1.05, + "grad_norm": 0.42578125, + "learning_rate": 9.252762770577625e-05, + "loss": 2.4708, + "step": 24632 + }, + { + "epoch": 1.05, + "grad_norm": 0.50390625, + "learning_rate": 9.25209385445957e-05, + "loss": 2.4644, + "step": 24633 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.251424941706819e-05, + "loss": 2.5314, + "step": 24634 + }, + { + "epoch": 1.05, + "grad_norm": 0.408203125, + "learning_rate": 9.250756032322383e-05, + "loss": 2.3726, + "step": 24635 + }, + { + "epoch": 1.05, + "grad_norm": 0.458984375, + "learning_rate": 9.250087126309273e-05, + "loss": 2.4538, + "step": 24636 + }, + { + "epoch": 1.05, + "grad_norm": 0.455078125, + "learning_rate": 9.24941822367049e-05, + "loss": 2.6044, + "step": 24637 + }, + { + "epoch": 1.05, + "grad_norm": 0.47265625, + "learning_rate": 9.248749324409059e-05, + "loss": 2.5073, + "step": 24638 + }, + { + "epoch": 1.05, + "grad_norm": 0.43359375, + "learning_rate": 9.248080428527977e-05, + "loss": 2.3484, + "step": 24639 + }, + { + "epoch": 1.05, + "grad_norm": 0.53125, + "learning_rate": 9.24741153603026e-05, + "loss": 2.4591, + "step": 24640 + }, + { + "epoch": 1.05, + "grad_norm": 0.423828125, + "learning_rate": 9.246742646918914e-05, + "loss": 2.3645, + "step": 24641 + }, + { + "epoch": 1.05, + "grad_norm": 0.486328125, + "learning_rate": 9.246073761196952e-05, + "loss": 2.4999, + "step": 24642 + }, + { + "epoch": 1.05, + "grad_norm": 0.490234375, + "learning_rate": 9.245404878867383e-05, + "loss": 2.5555, + "step": 24643 + }, + { + "epoch": 1.05, + "grad_norm": 0.486328125, + "learning_rate": 9.244735999933219e-05, + "loss": 2.3024, + "step": 24644 + }, + { + "epoch": 1.05, + "grad_norm": 0.419921875, + "learning_rate": 9.244067124397459e-05, + "loss": 2.4853, + "step": 24645 + }, + { + "epoch": 1.05, + "grad_norm": 0.5390625, + "learning_rate": 9.243398252263128e-05, + "loss": 2.4726, + "step": 24646 + }, + { + "epoch": 1.05, + "grad_norm": 0.490234375, + "learning_rate": 9.242729383533226e-05, + "loss": 2.6396, + "step": 24647 + }, + { + "epoch": 1.05, + "grad_norm": 0.478515625, + "learning_rate": 9.242060518210766e-05, + "loss": 2.6076, + "step": 24648 + }, + { + "epoch": 1.05, + "grad_norm": 0.494140625, + "learning_rate": 9.241391656298752e-05, + "loss": 2.5702, + "step": 24649 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.240722797800203e-05, + "loss": 2.5969, + "step": 24650 + }, + { + "epoch": 1.05, + "grad_norm": 0.48828125, + "learning_rate": 9.240053942718124e-05, + "loss": 2.4746, + "step": 24651 + }, + { + "epoch": 1.05, + "grad_norm": 0.486328125, + "learning_rate": 9.239385091055524e-05, + "loss": 2.423, + "step": 24652 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.238716242815409e-05, + "loss": 2.4053, + "step": 24653 + }, + { + "epoch": 1.05, + "grad_norm": 0.48828125, + "learning_rate": 9.238047398000799e-05, + "loss": 2.5698, + "step": 24654 + }, + { + "epoch": 1.05, + "grad_norm": 0.51953125, + "learning_rate": 9.237378556614694e-05, + "loss": 2.5732, + "step": 24655 + }, + { + "epoch": 1.05, + "grad_norm": 0.47265625, + "learning_rate": 9.236709718660108e-05, + "loss": 2.5468, + "step": 24656 + }, + { + "epoch": 1.05, + "grad_norm": 0.439453125, + "learning_rate": 9.236040884140048e-05, + "loss": 2.4896, + "step": 24657 + }, + { + "epoch": 1.05, + "grad_norm": 0.462890625, + "learning_rate": 9.235372053057524e-05, + "loss": 2.4135, + "step": 24658 + }, + { + "epoch": 1.05, + "grad_norm": 0.478515625, + "learning_rate": 9.234703225415548e-05, + "loss": 2.4814, + "step": 24659 + }, + { + "epoch": 1.05, + "grad_norm": 0.462890625, + "learning_rate": 9.23403440121713e-05, + "loss": 2.4067, + "step": 24660 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.233365580465271e-05, + "loss": 2.3881, + "step": 24661 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.232696763162993e-05, + "loss": 2.4844, + "step": 24662 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.232027949313297e-05, + "loss": 2.4684, + "step": 24663 + }, + { + "epoch": 1.05, + "grad_norm": 0.486328125, + "learning_rate": 9.231359138919194e-05, + "loss": 2.3425, + "step": 24664 + }, + { + "epoch": 1.05, + "grad_norm": 0.455078125, + "learning_rate": 9.230690331983692e-05, + "loss": 2.3911, + "step": 24665 + }, + { + "epoch": 1.05, + "grad_norm": 0.451171875, + "learning_rate": 9.230021528509805e-05, + "loss": 2.4935, + "step": 24666 + }, + { + "epoch": 1.05, + "grad_norm": 0.5, + "learning_rate": 9.22935272850054e-05, + "loss": 2.5354, + "step": 24667 + }, + { + "epoch": 1.05, + "grad_norm": 0.435546875, + "learning_rate": 9.228683931958907e-05, + "loss": 2.3968, + "step": 24668 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.228015138887908e-05, + "loss": 2.2962, + "step": 24669 + }, + { + "epoch": 1.05, + "grad_norm": 0.48828125, + "learning_rate": 9.227346349290567e-05, + "loss": 2.5024, + "step": 24670 + }, + { + "epoch": 1.05, + "grad_norm": 0.45703125, + "learning_rate": 9.226677563169882e-05, + "loss": 2.3934, + "step": 24671 + }, + { + "epoch": 1.05, + "grad_norm": 0.51171875, + "learning_rate": 9.226008780528865e-05, + "loss": 2.5496, + "step": 24672 + }, + { + "epoch": 1.05, + "grad_norm": 0.458984375, + "learning_rate": 9.225340001370522e-05, + "loss": 2.4468, + "step": 24673 + }, + { + "epoch": 1.05, + "grad_norm": 0.46875, + "learning_rate": 9.224671225697871e-05, + "loss": 2.4416, + "step": 24674 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.224002453513915e-05, + "loss": 2.5744, + "step": 24675 + }, + { + "epoch": 1.05, + "grad_norm": 0.49609375, + "learning_rate": 9.223333684821666e-05, + "loss": 2.6184, + "step": 24676 + }, + { + "epoch": 1.05, + "grad_norm": 0.44140625, + "learning_rate": 9.222664919624126e-05, + "loss": 2.3776, + "step": 24677 + }, + { + "epoch": 1.05, + "grad_norm": 0.498046875, + "learning_rate": 9.221996157924316e-05, + "loss": 2.5239, + "step": 24678 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.221327399725236e-05, + "loss": 2.3839, + "step": 24679 + }, + { + "epoch": 1.05, + "grad_norm": 0.42578125, + "learning_rate": 9.220658645029899e-05, + "loss": 2.409, + "step": 24680 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.21998989384131e-05, + "loss": 2.434, + "step": 24681 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.219321146162484e-05, + "loss": 2.412, + "step": 24682 + }, + { + "epoch": 1.05, + "grad_norm": 0.466796875, + "learning_rate": 9.218652401996428e-05, + "loss": 2.3645, + "step": 24683 + }, + { + "epoch": 1.05, + "grad_norm": 0.5234375, + "learning_rate": 9.217983661346152e-05, + "loss": 2.4542, + "step": 24684 + }, + { + "epoch": 1.05, + "grad_norm": 0.427734375, + "learning_rate": 9.217314924214658e-05, + "loss": 2.4453, + "step": 24685 + }, + { + "epoch": 1.05, + "grad_norm": 0.484375, + "learning_rate": 9.216646190604968e-05, + "loss": 2.4794, + "step": 24686 + }, + { + "epoch": 1.05, + "grad_norm": 0.45703125, + "learning_rate": 9.21597746052008e-05, + "loss": 2.5654, + "step": 24687 + }, + { + "epoch": 1.05, + "grad_norm": 0.48828125, + "learning_rate": 9.215308733963009e-05, + "loss": 2.491, + "step": 24688 + }, + { + "epoch": 1.05, + "grad_norm": 0.46875, + "learning_rate": 9.214640010936758e-05, + "loss": 2.4289, + "step": 24689 + }, + { + "epoch": 1.05, + "grad_norm": 0.484375, + "learning_rate": 9.213971291444341e-05, + "loss": 2.6323, + "step": 24690 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.213302575488769e-05, + "loss": 2.5284, + "step": 24691 + }, + { + "epoch": 1.05, + "grad_norm": 0.474609375, + "learning_rate": 9.212633863073048e-05, + "loss": 2.5772, + "step": 24692 + }, + { + "epoch": 1.05, + "grad_norm": 0.447265625, + "learning_rate": 9.211965154200182e-05, + "loss": 2.4299, + "step": 24693 + }, + { + "epoch": 1.05, + "grad_norm": 0.53515625, + "learning_rate": 9.21129644887319e-05, + "loss": 2.5841, + "step": 24694 + }, + { + "epoch": 1.05, + "grad_norm": 0.427734375, + "learning_rate": 9.210627747095073e-05, + "loss": 2.355, + "step": 24695 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.209959048868844e-05, + "loss": 2.473, + "step": 24696 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.209290354197508e-05, + "loss": 2.6826, + "step": 24697 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.208621663084079e-05, + "loss": 2.5382, + "step": 24698 + }, + { + "epoch": 1.05, + "grad_norm": 0.53515625, + "learning_rate": 9.207952975531562e-05, + "loss": 2.5698, + "step": 24699 + }, + { + "epoch": 1.05, + "grad_norm": 0.51171875, + "learning_rate": 9.20728429154297e-05, + "loss": 2.5537, + "step": 24700 + }, + { + "epoch": 1.05, + "grad_norm": 0.458984375, + "learning_rate": 9.206615611121304e-05, + "loss": 2.5327, + "step": 24701 + }, + { + "epoch": 1.05, + "grad_norm": 0.44921875, + "learning_rate": 9.205946934269582e-05, + "loss": 2.4825, + "step": 24702 + }, + { + "epoch": 1.05, + "grad_norm": 0.515625, + "learning_rate": 9.205278260990808e-05, + "loss": 2.5093, + "step": 24703 + }, + { + "epoch": 1.05, + "grad_norm": 0.466796875, + "learning_rate": 9.20460959128799e-05, + "loss": 2.4591, + "step": 24704 + }, + { + "epoch": 1.05, + "grad_norm": 0.5234375, + "learning_rate": 9.203940925164136e-05, + "loss": 2.5577, + "step": 24705 + }, + { + "epoch": 1.05, + "grad_norm": 0.515625, + "learning_rate": 9.20327226262226e-05, + "loss": 2.5899, + "step": 24706 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.202603603665369e-05, + "loss": 2.467, + "step": 24707 + }, + { + "epoch": 1.05, + "grad_norm": 0.515625, + "learning_rate": 9.20193494829647e-05, + "loss": 2.2889, + "step": 24708 + }, + { + "epoch": 1.05, + "grad_norm": 0.44921875, + "learning_rate": 9.201266296518566e-05, + "loss": 2.6132, + "step": 24709 + }, + { + "epoch": 1.05, + "grad_norm": 0.47265625, + "learning_rate": 9.200597648334679e-05, + "loss": 2.4542, + "step": 24710 + }, + { + "epoch": 1.05, + "grad_norm": 0.498046875, + "learning_rate": 9.199929003747808e-05, + "loss": 2.3796, + "step": 24711 + }, + { + "epoch": 1.05, + "grad_norm": 0.46875, + "learning_rate": 9.199260362760964e-05, + "loss": 2.3847, + "step": 24712 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.198591725377155e-05, + "loss": 2.5286, + "step": 24713 + }, + { + "epoch": 1.05, + "grad_norm": 0.421875, + "learning_rate": 9.197923091599391e-05, + "loss": 2.4332, + "step": 24714 + }, + { + "epoch": 1.05, + "grad_norm": 0.47265625, + "learning_rate": 9.19725446143068e-05, + "loss": 2.5869, + "step": 24715 + }, + { + "epoch": 1.05, + "grad_norm": 0.486328125, + "learning_rate": 9.196585834874033e-05, + "loss": 2.553, + "step": 24716 + }, + { + "epoch": 1.05, + "grad_norm": 0.5, + "learning_rate": 9.19591721193245e-05, + "loss": 2.6889, + "step": 24717 + }, + { + "epoch": 1.05, + "grad_norm": 0.44140625, + "learning_rate": 9.195248592608952e-05, + "loss": 2.3896, + "step": 24718 + }, + { + "epoch": 1.05, + "grad_norm": 0.52734375, + "learning_rate": 9.194579976906537e-05, + "loss": 2.468, + "step": 24719 + }, + { + "epoch": 1.05, + "grad_norm": 0.5234375, + "learning_rate": 9.19391136482822e-05, + "loss": 2.296, + "step": 24720 + }, + { + "epoch": 1.05, + "grad_norm": 0.4453125, + "learning_rate": 9.193242756377005e-05, + "loss": 2.4845, + "step": 24721 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.192574151555905e-05, + "loss": 2.4229, + "step": 24722 + }, + { + "epoch": 1.05, + "grad_norm": 0.46484375, + "learning_rate": 9.191905550367925e-05, + "loss": 2.5445, + "step": 24723 + }, + { + "epoch": 1.05, + "grad_norm": 0.453125, + "learning_rate": 9.191236952816078e-05, + "loss": 2.5458, + "step": 24724 + }, + { + "epoch": 1.05, + "grad_norm": 0.41015625, + "learning_rate": 9.190568358903363e-05, + "loss": 2.4163, + "step": 24725 + }, + { + "epoch": 1.05, + "grad_norm": 0.498046875, + "learning_rate": 9.189899768632801e-05, + "loss": 2.4212, + "step": 24726 + }, + { + "epoch": 1.05, + "grad_norm": 0.4453125, + "learning_rate": 9.18923118200739e-05, + "loss": 2.3172, + "step": 24727 + }, + { + "epoch": 1.05, + "grad_norm": 0.5078125, + "learning_rate": 9.188562599030144e-05, + "loss": 2.5955, + "step": 24728 + }, + { + "epoch": 1.05, + "grad_norm": 0.455078125, + "learning_rate": 9.187894019704067e-05, + "loss": 2.5294, + "step": 24729 + }, + { + "epoch": 1.05, + "grad_norm": 0.44921875, + "learning_rate": 9.187225444032172e-05, + "loss": 2.3946, + "step": 24730 + }, + { + "epoch": 1.05, + "grad_norm": 0.427734375, + "learning_rate": 9.186556872017466e-05, + "loss": 2.5272, + "step": 24731 + }, + { + "epoch": 1.05, + "grad_norm": 0.47265625, + "learning_rate": 9.185888303662959e-05, + "loss": 2.4819, + "step": 24732 + }, + { + "epoch": 1.05, + "grad_norm": 0.498046875, + "learning_rate": 9.185219738971655e-05, + "loss": 2.5, + "step": 24733 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.184551177946563e-05, + "loss": 2.455, + "step": 24734 + }, + { + "epoch": 1.05, + "grad_norm": 0.458984375, + "learning_rate": 9.183882620590694e-05, + "loss": 2.4032, + "step": 24735 + }, + { + "epoch": 1.05, + "grad_norm": 0.515625, + "learning_rate": 9.183214066907055e-05, + "loss": 2.4883, + "step": 24736 + }, + { + "epoch": 1.05, + "grad_norm": 0.5234375, + "learning_rate": 9.182545516898654e-05, + "loss": 2.4264, + "step": 24737 + }, + { + "epoch": 1.05, + "grad_norm": 0.4609375, + "learning_rate": 9.181876970568497e-05, + "loss": 2.5143, + "step": 24738 + }, + { + "epoch": 1.05, + "grad_norm": 0.4921875, + "learning_rate": 9.181208427919598e-05, + "loss": 2.4584, + "step": 24739 + }, + { + "epoch": 1.05, + "grad_norm": 0.494140625, + "learning_rate": 9.180539888954963e-05, + "loss": 2.3799, + "step": 24740 + }, + { + "epoch": 1.05, + "grad_norm": 0.41796875, + "learning_rate": 9.179871353677597e-05, + "loss": 2.3696, + "step": 24741 + }, + { + "epoch": 1.05, + "grad_norm": 0.380859375, + "learning_rate": 9.179202822090508e-05, + "loss": 2.4578, + "step": 24742 + }, + { + "epoch": 1.05, + "grad_norm": 0.466796875, + "learning_rate": 9.178534294196708e-05, + "loss": 2.41, + "step": 24743 + }, + { + "epoch": 1.05, + "grad_norm": 0.466796875, + "learning_rate": 9.177865769999204e-05, + "loss": 2.3801, + "step": 24744 + }, + { + "epoch": 1.05, + "grad_norm": 0.490234375, + "learning_rate": 9.177197249501004e-05, + "loss": 2.6038, + "step": 24745 + }, + { + "epoch": 1.05, + "grad_norm": 0.458984375, + "learning_rate": 9.176528732705114e-05, + "loss": 2.5121, + "step": 24746 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.175860219614543e-05, + "loss": 2.4538, + "step": 24747 + }, + { + "epoch": 1.05, + "grad_norm": 0.466796875, + "learning_rate": 9.175191710232305e-05, + "loss": 2.2969, + "step": 24748 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.1745232045614e-05, + "loss": 2.5182, + "step": 24749 + }, + { + "epoch": 1.05, + "grad_norm": 0.44921875, + "learning_rate": 9.173854702604835e-05, + "loss": 2.4305, + "step": 24750 + }, + { + "epoch": 1.05, + "grad_norm": 0.470703125, + "learning_rate": 9.173186204365627e-05, + "loss": 2.5131, + "step": 24751 + }, + { + "epoch": 1.05, + "grad_norm": 0.54296875, + "learning_rate": 9.172517709846777e-05, + "loss": 2.5136, + "step": 24752 + }, + { + "epoch": 1.05, + "grad_norm": 0.462890625, + "learning_rate": 9.171849219051296e-05, + "loss": 2.5198, + "step": 24753 + }, + { + "epoch": 1.05, + "grad_norm": 0.4765625, + "learning_rate": 9.171180731982188e-05, + "loss": 2.6134, + "step": 24754 + }, + { + "epoch": 1.05, + "grad_norm": 0.490234375, + "learning_rate": 9.170512248642465e-05, + "loss": 2.4032, + "step": 24755 + }, + { + "epoch": 1.05, + "grad_norm": 0.48046875, + "learning_rate": 9.169843769035139e-05, + "loss": 2.5537, + "step": 24756 + }, + { + "epoch": 1.05, + "grad_norm": 0.462890625, + "learning_rate": 9.169175293163207e-05, + "loss": 2.3825, + "step": 24757 + }, + { + "epoch": 1.06, + "grad_norm": 0.52734375, + "learning_rate": 9.168506821029683e-05, + "loss": 2.4524, + "step": 24758 + }, + { + "epoch": 1.06, + "grad_norm": 0.43359375, + "learning_rate": 9.167838352637575e-05, + "loss": 2.3385, + "step": 24759 + }, + { + "epoch": 1.06, + "grad_norm": 0.466796875, + "learning_rate": 9.167169887989891e-05, + "loss": 2.3697, + "step": 24760 + }, + { + "epoch": 1.06, + "grad_norm": 0.431640625, + "learning_rate": 9.166501427089638e-05, + "loss": 2.5674, + "step": 24761 + }, + { + "epoch": 1.06, + "grad_norm": 0.41015625, + "learning_rate": 9.16583296993982e-05, + "loss": 2.5945, + "step": 24762 + }, + { + "epoch": 1.06, + "grad_norm": 0.4453125, + "learning_rate": 9.165164516543452e-05, + "loss": 2.3158, + "step": 24763 + }, + { + "epoch": 1.06, + "grad_norm": 0.48046875, + "learning_rate": 9.164496066903543e-05, + "loss": 2.4299, + "step": 24764 + }, + { + "epoch": 1.06, + "grad_norm": 0.451171875, + "learning_rate": 9.163827621023092e-05, + "loss": 2.3952, + "step": 24765 + }, + { + "epoch": 1.06, + "grad_norm": 0.4375, + "learning_rate": 9.16315917890511e-05, + "loss": 2.522, + "step": 24766 + }, + { + "epoch": 1.06, + "grad_norm": 0.474609375, + "learning_rate": 9.162490740552607e-05, + "loss": 2.4618, + "step": 24767 + }, + { + "epoch": 1.06, + "grad_norm": 0.48828125, + "learning_rate": 9.16182230596859e-05, + "loss": 2.4179, + "step": 24768 + }, + { + "epoch": 1.06, + "grad_norm": 0.494140625, + "learning_rate": 9.161153875156067e-05, + "loss": 2.4785, + "step": 24769 + }, + { + "epoch": 1.06, + "grad_norm": 0.48046875, + "learning_rate": 9.160485448118042e-05, + "loss": 2.4365, + "step": 24770 + }, + { + "epoch": 1.06, + "grad_norm": 0.5, + "learning_rate": 9.159817024857527e-05, + "loss": 2.5639, + "step": 24771 + }, + { + "epoch": 1.06, + "grad_norm": 0.40234375, + "learning_rate": 9.159148605377532e-05, + "loss": 2.4547, + "step": 24772 + }, + { + "epoch": 1.06, + "grad_norm": 0.44921875, + "learning_rate": 9.158480189681059e-05, + "loss": 2.4773, + "step": 24773 + }, + { + "epoch": 1.06, + "grad_norm": 0.4921875, + "learning_rate": 9.157811777771115e-05, + "loss": 2.5401, + "step": 24774 + }, + { + "epoch": 1.06, + "grad_norm": 0.53515625, + "learning_rate": 9.157143369650713e-05, + "loss": 2.4804, + "step": 24775 + }, + { + "epoch": 1.06, + "grad_norm": 0.48046875, + "learning_rate": 9.156474965322856e-05, + "loss": 2.4542, + "step": 24776 + }, + { + "epoch": 1.06, + "grad_norm": 0.486328125, + "learning_rate": 9.155806564790555e-05, + "loss": 2.5338, + "step": 24777 + }, + { + "epoch": 1.06, + "grad_norm": 0.4296875, + "learning_rate": 9.155138168056812e-05, + "loss": 2.4154, + "step": 24778 + }, + { + "epoch": 1.06, + "grad_norm": 0.44140625, + "learning_rate": 9.154469775124642e-05, + "loss": 2.4946, + "step": 24779 + }, + { + "epoch": 1.06, + "grad_norm": 0.4453125, + "learning_rate": 9.153801385997052e-05, + "loss": 2.4282, + "step": 24780 + }, + { + "epoch": 1.06, + "grad_norm": 0.4921875, + "learning_rate": 9.153133000677043e-05, + "loss": 2.4314, + "step": 24781 + }, + { + "epoch": 1.06, + "grad_norm": 0.51171875, + "learning_rate": 9.152464619167624e-05, + "loss": 2.422, + "step": 24782 + }, + { + "epoch": 1.06, + "grad_norm": 0.46484375, + "learning_rate": 9.151796241471807e-05, + "loss": 2.3844, + "step": 24783 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.151127867592598e-05, + "loss": 2.5741, + "step": 24784 + }, + { + "epoch": 1.06, + "grad_norm": 0.53125, + "learning_rate": 9.150459497533e-05, + "loss": 2.434, + "step": 24785 + }, + { + "epoch": 1.06, + "grad_norm": 0.435546875, + "learning_rate": 9.149791131296024e-05, + "loss": 2.4157, + "step": 24786 + }, + { + "epoch": 1.06, + "grad_norm": 0.4296875, + "learning_rate": 9.149122768884679e-05, + "loss": 2.4159, + "step": 24787 + }, + { + "epoch": 1.06, + "grad_norm": 0.4453125, + "learning_rate": 9.148454410301973e-05, + "loss": 2.5355, + "step": 24788 + }, + { + "epoch": 1.06, + "grad_norm": 0.470703125, + "learning_rate": 9.147786055550909e-05, + "loss": 2.4709, + "step": 24789 + }, + { + "epoch": 1.06, + "grad_norm": 0.53515625, + "learning_rate": 9.147117704634493e-05, + "loss": 2.6397, + "step": 24790 + }, + { + "epoch": 1.06, + "grad_norm": 0.494140625, + "learning_rate": 9.14644935755574e-05, + "loss": 2.532, + "step": 24791 + }, + { + "epoch": 1.06, + "grad_norm": 0.4765625, + "learning_rate": 9.145781014317652e-05, + "loss": 2.618, + "step": 24792 + }, + { + "epoch": 1.06, + "grad_norm": 0.4296875, + "learning_rate": 9.145112674923237e-05, + "loss": 2.5329, + "step": 24793 + }, + { + "epoch": 1.06, + "grad_norm": 0.51171875, + "learning_rate": 9.144444339375498e-05, + "loss": 2.4563, + "step": 24794 + }, + { + "epoch": 1.06, + "grad_norm": 0.4609375, + "learning_rate": 9.143776007677452e-05, + "loss": 2.526, + "step": 24795 + }, + { + "epoch": 1.06, + "grad_norm": 0.53125, + "learning_rate": 9.143107679832102e-05, + "loss": 2.655, + "step": 24796 + }, + { + "epoch": 1.06, + "grad_norm": 0.47265625, + "learning_rate": 9.142439355842452e-05, + "loss": 2.4943, + "step": 24797 + }, + { + "epoch": 1.06, + "grad_norm": 0.486328125, + "learning_rate": 9.14177103571151e-05, + "loss": 2.5078, + "step": 24798 + }, + { + "epoch": 1.06, + "grad_norm": 0.455078125, + "learning_rate": 9.141102719442287e-05, + "loss": 2.3897, + "step": 24799 + }, + { + "epoch": 1.06, + "grad_norm": 0.51171875, + "learning_rate": 9.140434407037788e-05, + "loss": 2.5121, + "step": 24800 + }, + { + "epoch": 1.06, + "grad_norm": 0.416015625, + "learning_rate": 9.13976609850102e-05, + "loss": 2.4633, + "step": 24801 + }, + { + "epoch": 1.06, + "grad_norm": 0.55859375, + "learning_rate": 9.139097793834988e-05, + "loss": 2.3626, + "step": 24802 + }, + { + "epoch": 1.06, + "grad_norm": 0.435546875, + "learning_rate": 9.138429493042702e-05, + "loss": 2.3599, + "step": 24803 + }, + { + "epoch": 1.06, + "grad_norm": 0.47265625, + "learning_rate": 9.137761196127171e-05, + "loss": 2.5193, + "step": 24804 + }, + { + "epoch": 1.06, + "grad_norm": 0.4453125, + "learning_rate": 9.137092903091398e-05, + "loss": 2.5457, + "step": 24805 + }, + { + "epoch": 1.06, + "grad_norm": 0.466796875, + "learning_rate": 9.13642461393839e-05, + "loss": 2.4735, + "step": 24806 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.135756328671157e-05, + "loss": 2.3638, + "step": 24807 + }, + { + "epoch": 1.06, + "grad_norm": 0.4609375, + "learning_rate": 9.135088047292704e-05, + "loss": 2.4975, + "step": 24808 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.134419769806041e-05, + "loss": 2.2871, + "step": 24809 + }, + { + "epoch": 1.06, + "grad_norm": 0.4453125, + "learning_rate": 9.133751496214169e-05, + "loss": 2.4921, + "step": 24810 + }, + { + "epoch": 1.06, + "grad_norm": 0.439453125, + "learning_rate": 9.1330832265201e-05, + "loss": 2.4355, + "step": 24811 + }, + { + "epoch": 1.06, + "grad_norm": 0.470703125, + "learning_rate": 9.132414960726844e-05, + "loss": 2.4542, + "step": 24812 + }, + { + "epoch": 1.06, + "grad_norm": 0.435546875, + "learning_rate": 9.1317466988374e-05, + "loss": 2.4967, + "step": 24813 + }, + { + "epoch": 1.06, + "grad_norm": 0.4296875, + "learning_rate": 9.131078440854776e-05, + "loss": 2.5658, + "step": 24814 + }, + { + "epoch": 1.06, + "grad_norm": 0.4609375, + "learning_rate": 9.130410186781985e-05, + "loss": 2.4617, + "step": 24815 + }, + { + "epoch": 1.06, + "grad_norm": 0.494140625, + "learning_rate": 9.129741936622029e-05, + "loss": 2.4588, + "step": 24816 + }, + { + "epoch": 1.06, + "grad_norm": 0.451171875, + "learning_rate": 9.129073690377917e-05, + "loss": 2.3398, + "step": 24817 + }, + { + "epoch": 1.06, + "grad_norm": 0.462890625, + "learning_rate": 9.128405448052652e-05, + "loss": 2.47, + "step": 24818 + }, + { + "epoch": 1.06, + "grad_norm": 0.466796875, + "learning_rate": 9.127737209649248e-05, + "loss": 2.5463, + "step": 24819 + }, + { + "epoch": 1.06, + "grad_norm": 0.490234375, + "learning_rate": 9.127068975170708e-05, + "loss": 2.5076, + "step": 24820 + }, + { + "epoch": 1.06, + "grad_norm": 0.439453125, + "learning_rate": 9.126400744620037e-05, + "loss": 2.5322, + "step": 24821 + }, + { + "epoch": 1.06, + "grad_norm": 0.48828125, + "learning_rate": 9.125732518000242e-05, + "loss": 2.4773, + "step": 24822 + }, + { + "epoch": 1.06, + "grad_norm": 0.421875, + "learning_rate": 9.125064295314333e-05, + "loss": 2.4672, + "step": 24823 + }, + { + "epoch": 1.06, + "grad_norm": 0.431640625, + "learning_rate": 9.124396076565315e-05, + "loss": 2.4115, + "step": 24824 + }, + { + "epoch": 1.06, + "grad_norm": 0.470703125, + "learning_rate": 9.123727861756194e-05, + "loss": 2.574, + "step": 24825 + }, + { + "epoch": 1.06, + "grad_norm": 0.423828125, + "learning_rate": 9.123059650889976e-05, + "loss": 2.417, + "step": 24826 + }, + { + "epoch": 1.06, + "grad_norm": 0.431640625, + "learning_rate": 9.12239144396967e-05, + "loss": 2.3722, + "step": 24827 + }, + { + "epoch": 1.06, + "grad_norm": 0.462890625, + "learning_rate": 9.121723240998285e-05, + "loss": 2.5697, + "step": 24828 + }, + { + "epoch": 1.06, + "grad_norm": 0.42578125, + "learning_rate": 9.121055041978821e-05, + "loss": 2.5043, + "step": 24829 + }, + { + "epoch": 1.06, + "grad_norm": 0.41796875, + "learning_rate": 9.120386846914289e-05, + "loss": 2.3225, + "step": 24830 + }, + { + "epoch": 1.06, + "grad_norm": 0.46875, + "learning_rate": 9.119718655807694e-05, + "loss": 2.5876, + "step": 24831 + }, + { + "epoch": 1.06, + "grad_norm": 0.515625, + "learning_rate": 9.119050468662044e-05, + "loss": 2.4262, + "step": 24832 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.118382285480346e-05, + "loss": 2.4537, + "step": 24833 + }, + { + "epoch": 1.06, + "grad_norm": 0.494140625, + "learning_rate": 9.117714106265602e-05, + "loss": 2.6119, + "step": 24834 + }, + { + "epoch": 1.06, + "grad_norm": 0.453125, + "learning_rate": 9.117045931020824e-05, + "loss": 2.5187, + "step": 24835 + }, + { + "epoch": 1.06, + "grad_norm": 0.4375, + "learning_rate": 9.116377759749019e-05, + "loss": 2.5111, + "step": 24836 + }, + { + "epoch": 1.06, + "grad_norm": 0.39453125, + "learning_rate": 9.115709592453192e-05, + "loss": 2.5747, + "step": 24837 + }, + { + "epoch": 1.06, + "grad_norm": 0.439453125, + "learning_rate": 9.115041429136343e-05, + "loss": 2.393, + "step": 24838 + }, + { + "epoch": 1.06, + "grad_norm": 0.478515625, + "learning_rate": 9.114373269801487e-05, + "loss": 2.5333, + "step": 24839 + }, + { + "epoch": 1.06, + "grad_norm": 0.400390625, + "learning_rate": 9.113705114451628e-05, + "loss": 2.4708, + "step": 24840 + }, + { + "epoch": 1.06, + "grad_norm": 0.455078125, + "learning_rate": 9.113036963089771e-05, + "loss": 2.4882, + "step": 24841 + }, + { + "epoch": 1.06, + "grad_norm": 0.423828125, + "learning_rate": 9.112368815718923e-05, + "loss": 2.5161, + "step": 24842 + }, + { + "epoch": 1.06, + "grad_norm": 0.447265625, + "learning_rate": 9.111700672342092e-05, + "loss": 2.423, + "step": 24843 + }, + { + "epoch": 1.06, + "grad_norm": 0.373046875, + "learning_rate": 9.111032532962287e-05, + "loss": 2.3036, + "step": 24844 + }, + { + "epoch": 1.06, + "grad_norm": 0.423828125, + "learning_rate": 9.110364397582508e-05, + "loss": 2.5253, + "step": 24845 + }, + { + "epoch": 1.06, + "grad_norm": 0.486328125, + "learning_rate": 9.109696266205762e-05, + "loss": 2.4841, + "step": 24846 + }, + { + "epoch": 1.06, + "grad_norm": 0.46875, + "learning_rate": 9.109028138835058e-05, + "loss": 2.4235, + "step": 24847 + }, + { + "epoch": 1.06, + "grad_norm": 0.5078125, + "learning_rate": 9.108360015473402e-05, + "loss": 2.4514, + "step": 24848 + }, + { + "epoch": 1.06, + "grad_norm": 0.421875, + "learning_rate": 9.107691896123802e-05, + "loss": 2.4367, + "step": 24849 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.10702378078926e-05, + "loss": 2.3873, + "step": 24850 + }, + { + "epoch": 1.06, + "grad_norm": 0.49609375, + "learning_rate": 9.106355669472786e-05, + "loss": 2.3293, + "step": 24851 + }, + { + "epoch": 1.06, + "grad_norm": 0.46484375, + "learning_rate": 9.105687562177386e-05, + "loss": 2.537, + "step": 24852 + }, + { + "epoch": 1.06, + "grad_norm": 0.51171875, + "learning_rate": 9.105019458906065e-05, + "loss": 2.4135, + "step": 24853 + }, + { + "epoch": 1.06, + "grad_norm": 0.546875, + "learning_rate": 9.104351359661828e-05, + "loss": 2.4434, + "step": 24854 + }, + { + "epoch": 1.06, + "grad_norm": 0.458984375, + "learning_rate": 9.10368326444768e-05, + "loss": 2.4179, + "step": 24855 + }, + { + "epoch": 1.06, + "grad_norm": 0.482421875, + "learning_rate": 9.103015173266634e-05, + "loss": 2.6148, + "step": 24856 + }, + { + "epoch": 1.06, + "grad_norm": 0.51953125, + "learning_rate": 9.102347086121689e-05, + "loss": 2.5625, + "step": 24857 + }, + { + "epoch": 1.06, + "grad_norm": 0.48046875, + "learning_rate": 9.101679003015859e-05, + "loss": 2.5098, + "step": 24858 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.101010923952137e-05, + "loss": 2.4089, + "step": 24859 + }, + { + "epoch": 1.06, + "grad_norm": 0.45703125, + "learning_rate": 9.100342848933544e-05, + "loss": 2.5011, + "step": 24860 + }, + { + "epoch": 1.06, + "grad_norm": 0.44140625, + "learning_rate": 9.099674777963077e-05, + "loss": 2.5567, + "step": 24861 + }, + { + "epoch": 1.06, + "grad_norm": 0.447265625, + "learning_rate": 9.099006711043745e-05, + "loss": 2.4055, + "step": 24862 + }, + { + "epoch": 1.06, + "grad_norm": 0.46875, + "learning_rate": 9.098338648178552e-05, + "loss": 2.4895, + "step": 24863 + }, + { + "epoch": 1.06, + "grad_norm": 0.546875, + "learning_rate": 9.097670589370507e-05, + "loss": 2.5635, + "step": 24864 + }, + { + "epoch": 1.06, + "grad_norm": 0.45703125, + "learning_rate": 9.097002534622614e-05, + "loss": 2.4307, + "step": 24865 + }, + { + "epoch": 1.06, + "grad_norm": 0.435546875, + "learning_rate": 9.096334483937883e-05, + "loss": 2.373, + "step": 24866 + }, + { + "epoch": 1.06, + "grad_norm": 0.466796875, + "learning_rate": 9.095666437319309e-05, + "loss": 2.6033, + "step": 24867 + }, + { + "epoch": 1.06, + "grad_norm": 0.494140625, + "learning_rate": 9.094998394769913e-05, + "loss": 2.3868, + "step": 24868 + }, + { + "epoch": 1.06, + "grad_norm": 0.462890625, + "learning_rate": 9.094330356292691e-05, + "loss": 2.309, + "step": 24869 + }, + { + "epoch": 1.06, + "grad_norm": 0.431640625, + "learning_rate": 9.09366232189065e-05, + "loss": 2.3823, + "step": 24870 + }, + { + "epoch": 1.06, + "grad_norm": 0.431640625, + "learning_rate": 9.092994291566797e-05, + "loss": 2.2711, + "step": 24871 + }, + { + "epoch": 1.06, + "grad_norm": 0.53515625, + "learning_rate": 9.09232626532414e-05, + "loss": 2.4878, + "step": 24872 + }, + { + "epoch": 1.06, + "grad_norm": 0.40625, + "learning_rate": 9.091658243165683e-05, + "loss": 2.4157, + "step": 24873 + }, + { + "epoch": 1.06, + "grad_norm": 0.46484375, + "learning_rate": 9.090990225094433e-05, + "loss": 2.5715, + "step": 24874 + }, + { + "epoch": 1.06, + "grad_norm": 0.49609375, + "learning_rate": 9.09032221111339e-05, + "loss": 2.4534, + "step": 24875 + }, + { + "epoch": 1.06, + "grad_norm": 0.54296875, + "learning_rate": 9.089654201225571e-05, + "loss": 2.6394, + "step": 24876 + }, + { + "epoch": 1.06, + "grad_norm": 0.482421875, + "learning_rate": 9.088986195433972e-05, + "loss": 2.696, + "step": 24877 + }, + { + "epoch": 1.06, + "grad_norm": 0.470703125, + "learning_rate": 9.088318193741602e-05, + "loss": 2.5609, + "step": 24878 + }, + { + "epoch": 1.06, + "grad_norm": 0.4765625, + "learning_rate": 9.087650196151466e-05, + "loss": 2.4487, + "step": 24879 + }, + { + "epoch": 1.06, + "grad_norm": 0.498046875, + "learning_rate": 9.086982202666573e-05, + "loss": 2.4068, + "step": 24880 + }, + { + "epoch": 1.06, + "grad_norm": 0.478515625, + "learning_rate": 9.086314213289926e-05, + "loss": 2.4754, + "step": 24881 + }, + { + "epoch": 1.06, + "grad_norm": 0.478515625, + "learning_rate": 9.085646228024532e-05, + "loss": 2.4143, + "step": 24882 + }, + { + "epoch": 1.06, + "grad_norm": 0.55859375, + "learning_rate": 9.084978246873391e-05, + "loss": 2.6071, + "step": 24883 + }, + { + "epoch": 1.06, + "grad_norm": 0.3984375, + "learning_rate": 9.08431026983952e-05, + "loss": 2.4707, + "step": 24884 + }, + { + "epoch": 1.06, + "grad_norm": 0.5234375, + "learning_rate": 9.083642296925915e-05, + "loss": 2.579, + "step": 24885 + }, + { + "epoch": 1.06, + "grad_norm": 0.49609375, + "learning_rate": 9.082974328135585e-05, + "loss": 2.4073, + "step": 24886 + }, + { + "epoch": 1.06, + "grad_norm": 0.5, + "learning_rate": 9.082306363471533e-05, + "loss": 2.4146, + "step": 24887 + }, + { + "epoch": 1.06, + "grad_norm": 0.462890625, + "learning_rate": 9.081638402936769e-05, + "loss": 2.5827, + "step": 24888 + }, + { + "epoch": 1.06, + "grad_norm": 0.51953125, + "learning_rate": 9.080970446534298e-05, + "loss": 2.5719, + "step": 24889 + }, + { + "epoch": 1.06, + "grad_norm": 0.515625, + "learning_rate": 9.080302494267125e-05, + "loss": 2.5236, + "step": 24890 + }, + { + "epoch": 1.06, + "grad_norm": 0.458984375, + "learning_rate": 9.07963454613825e-05, + "loss": 2.4806, + "step": 24891 + }, + { + "epoch": 1.06, + "grad_norm": 0.46484375, + "learning_rate": 9.078966602150687e-05, + "loss": 2.472, + "step": 24892 + }, + { + "epoch": 1.06, + "grad_norm": 0.5234375, + "learning_rate": 9.078298662307436e-05, + "loss": 2.5448, + "step": 24893 + }, + { + "epoch": 1.06, + "grad_norm": 0.431640625, + "learning_rate": 9.077630726611506e-05, + "loss": 2.3654, + "step": 24894 + }, + { + "epoch": 1.06, + "grad_norm": 0.453125, + "learning_rate": 9.076962795065897e-05, + "loss": 2.4758, + "step": 24895 + }, + { + "epoch": 1.06, + "grad_norm": 0.447265625, + "learning_rate": 9.07629486767362e-05, + "loss": 2.4683, + "step": 24896 + }, + { + "epoch": 1.06, + "grad_norm": 0.43359375, + "learning_rate": 9.07562694443768e-05, + "loss": 2.439, + "step": 24897 + }, + { + "epoch": 1.06, + "grad_norm": 0.5078125, + "learning_rate": 9.074959025361082e-05, + "loss": 2.4205, + "step": 24898 + }, + { + "epoch": 1.06, + "grad_norm": 0.466796875, + "learning_rate": 9.074291110446824e-05, + "loss": 2.5346, + "step": 24899 + }, + { + "epoch": 1.06, + "grad_norm": 0.45703125, + "learning_rate": 9.073623199697925e-05, + "loss": 2.3865, + "step": 24900 + }, + { + "epoch": 1.06, + "grad_norm": 0.4453125, + "learning_rate": 9.072955293117381e-05, + "loss": 2.4367, + "step": 24901 + }, + { + "epoch": 1.06, + "grad_norm": 0.5234375, + "learning_rate": 9.072287390708197e-05, + "loss": 2.3532, + "step": 24902 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.07161949247338e-05, + "loss": 2.5223, + "step": 24903 + }, + { + "epoch": 1.06, + "grad_norm": 0.466796875, + "learning_rate": 9.070951598415939e-05, + "loss": 2.4375, + "step": 24904 + }, + { + "epoch": 1.06, + "grad_norm": 0.53125, + "learning_rate": 9.070283708538876e-05, + "loss": 2.4923, + "step": 24905 + }, + { + "epoch": 1.06, + "grad_norm": 0.4375, + "learning_rate": 9.069615822845197e-05, + "loss": 2.4145, + "step": 24906 + }, + { + "epoch": 1.06, + "grad_norm": 0.48828125, + "learning_rate": 9.068947941337902e-05, + "loss": 2.4399, + "step": 24907 + }, + { + "epoch": 1.06, + "grad_norm": 0.5546875, + "learning_rate": 9.068280064020008e-05, + "loss": 2.6158, + "step": 24908 + }, + { + "epoch": 1.06, + "grad_norm": 0.47265625, + "learning_rate": 9.06761219089451e-05, + "loss": 2.647, + "step": 24909 + }, + { + "epoch": 1.06, + "grad_norm": 0.443359375, + "learning_rate": 9.066944321964415e-05, + "loss": 2.5159, + "step": 24910 + }, + { + "epoch": 1.06, + "grad_norm": 0.55078125, + "learning_rate": 9.066276457232729e-05, + "loss": 2.4827, + "step": 24911 + }, + { + "epoch": 1.06, + "grad_norm": 0.4375, + "learning_rate": 9.06560859670246e-05, + "loss": 2.4773, + "step": 24912 + }, + { + "epoch": 1.06, + "grad_norm": 0.423828125, + "learning_rate": 9.064940740376611e-05, + "loss": 2.4993, + "step": 24913 + }, + { + "epoch": 1.06, + "grad_norm": 0.53125, + "learning_rate": 9.064272888258189e-05, + "loss": 2.4163, + "step": 24914 + }, + { + "epoch": 1.06, + "grad_norm": 0.431640625, + "learning_rate": 9.063605040350189e-05, + "loss": 2.457, + "step": 24915 + }, + { + "epoch": 1.06, + "grad_norm": 0.400390625, + "learning_rate": 9.062937196655633e-05, + "loss": 2.4609, + "step": 24916 + }, + { + "epoch": 1.06, + "grad_norm": 0.486328125, + "learning_rate": 9.062269357177513e-05, + "loss": 2.4584, + "step": 24917 + }, + { + "epoch": 1.06, + "grad_norm": 0.498046875, + "learning_rate": 9.06160152191884e-05, + "loss": 2.3003, + "step": 24918 + }, + { + "epoch": 1.06, + "grad_norm": 0.41796875, + "learning_rate": 9.060933690882614e-05, + "loss": 2.4604, + "step": 24919 + }, + { + "epoch": 1.06, + "grad_norm": 0.4375, + "learning_rate": 9.060265864071845e-05, + "loss": 2.4601, + "step": 24920 + }, + { + "epoch": 1.06, + "grad_norm": 0.46484375, + "learning_rate": 9.059598041489537e-05, + "loss": 2.3808, + "step": 24921 + }, + { + "epoch": 1.06, + "grad_norm": 0.447265625, + "learning_rate": 9.058930223138695e-05, + "loss": 2.4006, + "step": 24922 + }, + { + "epoch": 1.06, + "grad_norm": 0.5078125, + "learning_rate": 9.058262409022318e-05, + "loss": 2.5694, + "step": 24923 + }, + { + "epoch": 1.06, + "grad_norm": 0.490234375, + "learning_rate": 9.057594599143421e-05, + "loss": 2.4521, + "step": 24924 + }, + { + "epoch": 1.06, + "grad_norm": 0.470703125, + "learning_rate": 9.056926793505003e-05, + "loss": 2.4734, + "step": 24925 + }, + { + "epoch": 1.06, + "grad_norm": 0.53515625, + "learning_rate": 9.05625899211007e-05, + "loss": 2.4036, + "step": 24926 + }, + { + "epoch": 1.06, + "grad_norm": 0.478515625, + "learning_rate": 9.055591194961623e-05, + "loss": 2.47, + "step": 24927 + }, + { + "epoch": 1.06, + "grad_norm": 0.435546875, + "learning_rate": 9.054923402062673e-05, + "loss": 2.4679, + "step": 24928 + }, + { + "epoch": 1.06, + "grad_norm": 0.462890625, + "learning_rate": 9.054255613416224e-05, + "loss": 2.426, + "step": 24929 + }, + { + "epoch": 1.06, + "grad_norm": 0.54296875, + "learning_rate": 9.053587829025281e-05, + "loss": 2.3881, + "step": 24930 + }, + { + "epoch": 1.06, + "grad_norm": 0.52734375, + "learning_rate": 9.052920048892838e-05, + "loss": 2.5193, + "step": 24931 + }, + { + "epoch": 1.06, + "grad_norm": 0.44140625, + "learning_rate": 9.052252273021915e-05, + "loss": 2.5056, + "step": 24932 + }, + { + "epoch": 1.06, + "grad_norm": 0.421875, + "learning_rate": 9.051584501415513e-05, + "loss": 2.4172, + "step": 24933 + }, + { + "epoch": 1.06, + "grad_norm": 0.431640625, + "learning_rate": 9.050916734076631e-05, + "loss": 2.4426, + "step": 24934 + }, + { + "epoch": 1.06, + "grad_norm": 0.46875, + "learning_rate": 9.050248971008276e-05, + "loss": 2.5347, + "step": 24935 + }, + { + "epoch": 1.06, + "grad_norm": 0.45703125, + "learning_rate": 9.049581212213452e-05, + "loss": 2.5158, + "step": 24936 + }, + { + "epoch": 1.06, + "grad_norm": 0.44140625, + "learning_rate": 9.048913457695168e-05, + "loss": 2.2984, + "step": 24937 + }, + { + "epoch": 1.06, + "grad_norm": 0.462890625, + "learning_rate": 9.048245707456429e-05, + "loss": 2.3804, + "step": 24938 + }, + { + "epoch": 1.06, + "grad_norm": 0.43359375, + "learning_rate": 9.047577961500229e-05, + "loss": 2.5245, + "step": 24939 + }, + { + "epoch": 1.06, + "grad_norm": 0.62890625, + "learning_rate": 9.046910219829582e-05, + "loss": 2.5024, + "step": 24940 + }, + { + "epoch": 1.06, + "grad_norm": 0.44921875, + "learning_rate": 9.046242482447497e-05, + "loss": 2.3679, + "step": 24941 + }, + { + "epoch": 1.06, + "grad_norm": 0.455078125, + "learning_rate": 9.045574749356967e-05, + "loss": 2.546, + "step": 24942 + }, + { + "epoch": 1.06, + "grad_norm": 0.5, + "learning_rate": 9.044907020561e-05, + "loss": 2.6696, + "step": 24943 + }, + { + "epoch": 1.06, + "grad_norm": 0.421875, + "learning_rate": 9.044239296062606e-05, + "loss": 2.4527, + "step": 24944 + }, + { + "epoch": 1.06, + "grad_norm": 0.478515625, + "learning_rate": 9.043571575864785e-05, + "loss": 2.4347, + "step": 24945 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.042903859970545e-05, + "loss": 2.4202, + "step": 24946 + }, + { + "epoch": 1.06, + "grad_norm": 0.427734375, + "learning_rate": 9.042236148382881e-05, + "loss": 2.4453, + "step": 24947 + }, + { + "epoch": 1.06, + "grad_norm": 0.447265625, + "learning_rate": 9.041568441104808e-05, + "loss": 2.3639, + "step": 24948 + }, + { + "epoch": 1.06, + "grad_norm": 0.44140625, + "learning_rate": 9.04090073813933e-05, + "loss": 2.5212, + "step": 24949 + }, + { + "epoch": 1.06, + "grad_norm": 0.4921875, + "learning_rate": 9.040233039489446e-05, + "loss": 2.4951, + "step": 24950 + }, + { + "epoch": 1.06, + "grad_norm": 0.48046875, + "learning_rate": 9.039565345158161e-05, + "loss": 2.487, + "step": 24951 + }, + { + "epoch": 1.06, + "grad_norm": 0.47265625, + "learning_rate": 9.038897655148482e-05, + "loss": 2.5319, + "step": 24952 + }, + { + "epoch": 1.06, + "grad_norm": 0.458984375, + "learning_rate": 9.038229969463412e-05, + "loss": 2.3892, + "step": 24953 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.037562288105958e-05, + "loss": 2.4353, + "step": 24954 + }, + { + "epoch": 1.06, + "grad_norm": 0.5234375, + "learning_rate": 9.036894611079116e-05, + "loss": 2.4597, + "step": 24955 + }, + { + "epoch": 1.06, + "grad_norm": 0.51953125, + "learning_rate": 9.0362269383859e-05, + "loss": 2.5701, + "step": 24956 + }, + { + "epoch": 1.06, + "grad_norm": 0.482421875, + "learning_rate": 9.035559270029314e-05, + "loss": 2.5039, + "step": 24957 + }, + { + "epoch": 1.06, + "grad_norm": 0.46875, + "learning_rate": 9.034891606012355e-05, + "loss": 2.5862, + "step": 24958 + }, + { + "epoch": 1.06, + "grad_norm": 0.5, + "learning_rate": 9.03422394633803e-05, + "loss": 2.5381, + "step": 24959 + }, + { + "epoch": 1.06, + "grad_norm": 0.470703125, + "learning_rate": 9.033556291009346e-05, + "loss": 2.417, + "step": 24960 + }, + { + "epoch": 1.06, + "grad_norm": 0.4296875, + "learning_rate": 9.032888640029306e-05, + "loss": 2.4607, + "step": 24961 + }, + { + "epoch": 1.06, + "grad_norm": 0.4765625, + "learning_rate": 9.032220993400915e-05, + "loss": 2.4859, + "step": 24962 + }, + { + "epoch": 1.06, + "grad_norm": 0.498046875, + "learning_rate": 9.03155335112717e-05, + "loss": 2.4657, + "step": 24963 + }, + { + "epoch": 1.06, + "grad_norm": 0.50390625, + "learning_rate": 9.030885713211084e-05, + "loss": 2.4602, + "step": 24964 + }, + { + "epoch": 1.06, + "grad_norm": 0.46484375, + "learning_rate": 9.030218079655662e-05, + "loss": 2.5031, + "step": 24965 + }, + { + "epoch": 1.06, + "grad_norm": 0.423828125, + "learning_rate": 9.029550450463902e-05, + "loss": 2.4359, + "step": 24966 + }, + { + "epoch": 1.06, + "grad_norm": 0.46875, + "learning_rate": 9.028882825638808e-05, + "loss": 2.3964, + "step": 24967 + }, + { + "epoch": 1.06, + "grad_norm": 0.439453125, + "learning_rate": 9.028215205183387e-05, + "loss": 2.5839, + "step": 24968 + }, + { + "epoch": 1.06, + "grad_norm": 0.462890625, + "learning_rate": 9.027547589100644e-05, + "loss": 2.5282, + "step": 24969 + }, + { + "epoch": 1.06, + "grad_norm": 0.43359375, + "learning_rate": 9.026879977393584e-05, + "loss": 2.3589, + "step": 24970 + }, + { + "epoch": 1.06, + "grad_norm": 0.48046875, + "learning_rate": 9.0262123700652e-05, + "loss": 2.3986, + "step": 24971 + }, + { + "epoch": 1.06, + "grad_norm": 0.466796875, + "learning_rate": 9.02554476711851e-05, + "loss": 2.444, + "step": 24972 + }, + { + "epoch": 1.06, + "grad_norm": 0.443359375, + "learning_rate": 9.024877168556515e-05, + "loss": 2.5216, + "step": 24973 + }, + { + "epoch": 1.06, + "grad_norm": 0.46484375, + "learning_rate": 9.024209574382213e-05, + "loss": 2.3947, + "step": 24974 + }, + { + "epoch": 1.06, + "grad_norm": 0.474609375, + "learning_rate": 9.023541984598612e-05, + "loss": 2.5534, + "step": 24975 + }, + { + "epoch": 1.06, + "grad_norm": 0.50390625, + "learning_rate": 9.022874399208714e-05, + "loss": 2.4425, + "step": 24976 + }, + { + "epoch": 1.06, + "grad_norm": 0.474609375, + "learning_rate": 9.022206818215525e-05, + "loss": 2.4518, + "step": 24977 + }, + { + "epoch": 1.06, + "grad_norm": 0.4921875, + "learning_rate": 9.02153924162205e-05, + "loss": 2.5349, + "step": 24978 + }, + { + "epoch": 1.06, + "grad_norm": 0.443359375, + "learning_rate": 9.02087166943129e-05, + "loss": 2.4058, + "step": 24979 + }, + { + "epoch": 1.06, + "grad_norm": 0.455078125, + "learning_rate": 9.020204101646243e-05, + "loss": 2.5296, + "step": 24980 + }, + { + "epoch": 1.06, + "grad_norm": 0.484375, + "learning_rate": 9.019536538269928e-05, + "loss": 2.4827, + "step": 24981 + }, + { + "epoch": 1.06, + "grad_norm": 0.478515625, + "learning_rate": 9.018868979305338e-05, + "loss": 2.44, + "step": 24982 + }, + { + "epoch": 1.06, + "grad_norm": 0.48046875, + "learning_rate": 9.01820142475548e-05, + "loss": 2.4217, + "step": 24983 + }, + { + "epoch": 1.06, + "grad_norm": 0.48046875, + "learning_rate": 9.017533874623352e-05, + "loss": 2.5032, + "step": 24984 + }, + { + "epoch": 1.06, + "grad_norm": 0.458984375, + "learning_rate": 9.016866328911965e-05, + "loss": 2.417, + "step": 24985 + }, + { + "epoch": 1.06, + "grad_norm": 0.458984375, + "learning_rate": 9.016198787624322e-05, + "loss": 2.4046, + "step": 24986 + }, + { + "epoch": 1.06, + "grad_norm": 0.515625, + "learning_rate": 9.015531250763427e-05, + "loss": 2.4122, + "step": 24987 + }, + { + "epoch": 1.06, + "grad_norm": 0.51171875, + "learning_rate": 9.014863718332274e-05, + "loss": 2.6192, + "step": 24988 + }, + { + "epoch": 1.06, + "grad_norm": 0.435546875, + "learning_rate": 9.01419619033388e-05, + "loss": 2.5037, + "step": 24989 + }, + { + "epoch": 1.06, + "grad_norm": 0.474609375, + "learning_rate": 9.013528666771243e-05, + "loss": 2.3383, + "step": 24990 + }, + { + "epoch": 1.06, + "grad_norm": 0.41796875, + "learning_rate": 9.012861147647365e-05, + "loss": 2.3847, + "step": 24991 + }, + { + "epoch": 1.06, + "grad_norm": 0.51953125, + "learning_rate": 9.012193632965248e-05, + "loss": 2.4821, + "step": 24992 + }, + { + "epoch": 1.07, + "grad_norm": 0.392578125, + "learning_rate": 9.011526122727903e-05, + "loss": 2.3405, + "step": 24993 + }, + { + "epoch": 1.07, + "grad_norm": 0.435546875, + "learning_rate": 9.010858616938328e-05, + "loss": 2.38, + "step": 24994 + }, + { + "epoch": 1.07, + "grad_norm": 0.48046875, + "learning_rate": 9.01019111559953e-05, + "loss": 2.5255, + "step": 24995 + }, + { + "epoch": 1.07, + "grad_norm": 0.46875, + "learning_rate": 9.009523618714503e-05, + "loss": 2.4665, + "step": 24996 + }, + { + "epoch": 1.07, + "grad_norm": 0.416015625, + "learning_rate": 9.008856126286267e-05, + "loss": 2.4703, + "step": 24997 + }, + { + "epoch": 1.07, + "grad_norm": 0.482421875, + "learning_rate": 9.00818863831781e-05, + "loss": 2.3662, + "step": 24998 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 9.007521154812144e-05, + "loss": 2.4393, + "step": 24999 + }, + { + "epoch": 1.07, + "grad_norm": 0.4765625, + "learning_rate": 9.006853675772268e-05, + "loss": 2.5011, + "step": 25000 + }, + { + "epoch": 1.07, + "grad_norm": 0.48828125, + "learning_rate": 9.00618620120119e-05, + "loss": 2.6739, + "step": 25001 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 9.005518731101909e-05, + "loss": 2.5617, + "step": 25002 + }, + { + "epoch": 1.07, + "grad_norm": 0.470703125, + "learning_rate": 9.004851265477435e-05, + "loss": 2.372, + "step": 25003 + }, + { + "epoch": 1.07, + "grad_norm": 0.5234375, + "learning_rate": 9.004183804330757e-05, + "loss": 2.5667, + "step": 25004 + }, + { + "epoch": 1.07, + "grad_norm": 0.47265625, + "learning_rate": 9.003516347664898e-05, + "loss": 2.583, + "step": 25005 + }, + { + "epoch": 1.07, + "grad_norm": 0.49609375, + "learning_rate": 9.002848895482848e-05, + "loss": 2.422, + "step": 25006 + }, + { + "epoch": 1.07, + "grad_norm": 0.458984375, + "learning_rate": 9.002181447787612e-05, + "loss": 2.5315, + "step": 25007 + }, + { + "epoch": 1.07, + "grad_norm": 0.45703125, + "learning_rate": 9.001514004582193e-05, + "loss": 2.5254, + "step": 25008 + }, + { + "epoch": 1.07, + "grad_norm": 0.482421875, + "learning_rate": 9.000846565869598e-05, + "loss": 2.4945, + "step": 25009 + }, + { + "epoch": 1.07, + "grad_norm": 0.50390625, + "learning_rate": 9.00017913165283e-05, + "loss": 2.4561, + "step": 25010 + }, + { + "epoch": 1.07, + "grad_norm": 0.453125, + "learning_rate": 8.999511701934892e-05, + "loss": 2.4539, + "step": 25011 + }, + { + "epoch": 1.07, + "grad_norm": 0.42578125, + "learning_rate": 8.998844276718779e-05, + "loss": 2.4336, + "step": 25012 + }, + { + "epoch": 1.07, + "grad_norm": 0.46875, + "learning_rate": 8.998176856007507e-05, + "loss": 2.5247, + "step": 25013 + }, + { + "epoch": 1.07, + "grad_norm": 0.482421875, + "learning_rate": 8.997509439804071e-05, + "loss": 2.5647, + "step": 25014 + }, + { + "epoch": 1.07, + "grad_norm": 0.453125, + "learning_rate": 8.996842028111476e-05, + "loss": 2.4403, + "step": 25015 + }, + { + "epoch": 1.07, + "grad_norm": 0.4609375, + "learning_rate": 8.996174620932724e-05, + "loss": 2.3718, + "step": 25016 + }, + { + "epoch": 1.07, + "grad_norm": 0.50390625, + "learning_rate": 8.995507218270822e-05, + "loss": 2.4409, + "step": 25017 + }, + { + "epoch": 1.07, + "grad_norm": 0.44921875, + "learning_rate": 8.994839820128769e-05, + "loss": 2.2916, + "step": 25018 + }, + { + "epoch": 1.07, + "grad_norm": 0.455078125, + "learning_rate": 8.994172426509572e-05, + "loss": 2.4481, + "step": 25019 + }, + { + "epoch": 1.07, + "grad_norm": 0.482421875, + "learning_rate": 8.993505037416224e-05, + "loss": 2.5026, + "step": 25020 + }, + { + "epoch": 1.07, + "grad_norm": 0.44921875, + "learning_rate": 8.992837652851743e-05, + "loss": 2.4973, + "step": 25021 + }, + { + "epoch": 1.07, + "grad_norm": 0.427734375, + "learning_rate": 8.992170272819123e-05, + "loss": 2.345, + "step": 25022 + }, + { + "epoch": 1.07, + "grad_norm": 0.494140625, + "learning_rate": 8.991502897321368e-05, + "loss": 2.5025, + "step": 25023 + }, + { + "epoch": 1.07, + "grad_norm": 0.515625, + "learning_rate": 8.990835526361478e-05, + "loss": 2.4362, + "step": 25024 + }, + { + "epoch": 1.07, + "grad_norm": 0.5234375, + "learning_rate": 8.990168159942464e-05, + "loss": 2.5954, + "step": 25025 + }, + { + "epoch": 1.07, + "grad_norm": 0.47265625, + "learning_rate": 8.989500798067322e-05, + "loss": 2.4836, + "step": 25026 + }, + { + "epoch": 1.07, + "grad_norm": 0.443359375, + "learning_rate": 8.988833440739059e-05, + "loss": 2.3774, + "step": 25027 + }, + { + "epoch": 1.07, + "grad_norm": 0.46484375, + "learning_rate": 8.988166087960671e-05, + "loss": 2.576, + "step": 25028 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 8.987498739735173e-05, + "loss": 2.4836, + "step": 25029 + }, + { + "epoch": 1.07, + "grad_norm": 0.470703125, + "learning_rate": 8.986831396065557e-05, + "loss": 2.4497, + "step": 25030 + }, + { + "epoch": 1.07, + "grad_norm": 0.484375, + "learning_rate": 8.986164056954829e-05, + "loss": 2.4411, + "step": 25031 + }, + { + "epoch": 1.07, + "grad_norm": 0.42578125, + "learning_rate": 8.985496722405992e-05, + "loss": 2.314, + "step": 25032 + }, + { + "epoch": 1.07, + "grad_norm": 0.447265625, + "learning_rate": 8.984829392422049e-05, + "loss": 2.4799, + "step": 25033 + }, + { + "epoch": 1.07, + "grad_norm": 0.490234375, + "learning_rate": 8.984162067006003e-05, + "loss": 2.3984, + "step": 25034 + }, + { + "epoch": 1.07, + "grad_norm": 0.390625, + "learning_rate": 8.98349474616086e-05, + "loss": 2.558, + "step": 25035 + }, + { + "epoch": 1.07, + "grad_norm": 0.439453125, + "learning_rate": 8.982827429889612e-05, + "loss": 2.356, + "step": 25036 + }, + { + "epoch": 1.07, + "grad_norm": 0.498046875, + "learning_rate": 8.982160118195275e-05, + "loss": 2.4344, + "step": 25037 + }, + { + "epoch": 1.07, + "grad_norm": 0.5, + "learning_rate": 8.981492811080844e-05, + "loss": 2.6532, + "step": 25038 + }, + { + "epoch": 1.07, + "grad_norm": 0.5078125, + "learning_rate": 8.980825508549323e-05, + "loss": 2.4776, + "step": 25039 + }, + { + "epoch": 1.07, + "grad_norm": 0.45703125, + "learning_rate": 8.980158210603712e-05, + "loss": 2.4497, + "step": 25040 + }, + { + "epoch": 1.07, + "grad_norm": 0.4609375, + "learning_rate": 8.97949091724702e-05, + "loss": 2.4017, + "step": 25041 + }, + { + "epoch": 1.07, + "grad_norm": 0.498046875, + "learning_rate": 8.978823628482244e-05, + "loss": 2.6091, + "step": 25042 + }, + { + "epoch": 1.07, + "grad_norm": 0.494140625, + "learning_rate": 8.978156344312392e-05, + "loss": 2.4456, + "step": 25043 + }, + { + "epoch": 1.07, + "grad_norm": 0.4375, + "learning_rate": 8.977489064740455e-05, + "loss": 2.4441, + "step": 25044 + }, + { + "epoch": 1.07, + "grad_norm": 0.5, + "learning_rate": 8.97682178976945e-05, + "loss": 2.5851, + "step": 25045 + }, + { + "epoch": 1.07, + "grad_norm": 0.46484375, + "learning_rate": 8.976154519402374e-05, + "loss": 2.3807, + "step": 25046 + }, + { + "epoch": 1.07, + "grad_norm": 0.388671875, + "learning_rate": 8.975487253642225e-05, + "loss": 2.407, + "step": 25047 + }, + { + "epoch": 1.07, + "grad_norm": 0.396484375, + "learning_rate": 8.974819992492008e-05, + "loss": 2.5092, + "step": 25048 + }, + { + "epoch": 1.07, + "grad_norm": 0.470703125, + "learning_rate": 8.974152735954729e-05, + "loss": 2.4876, + "step": 25049 + }, + { + "epoch": 1.07, + "grad_norm": 0.4375, + "learning_rate": 8.973485484033387e-05, + "loss": 2.4284, + "step": 25050 + }, + { + "epoch": 1.07, + "grad_norm": 0.482421875, + "learning_rate": 8.972818236730987e-05, + "loss": 2.2801, + "step": 25051 + }, + { + "epoch": 1.07, + "grad_norm": 0.515625, + "learning_rate": 8.972150994050524e-05, + "loss": 2.5463, + "step": 25052 + }, + { + "epoch": 1.07, + "grad_norm": 0.5546875, + "learning_rate": 8.971483755995012e-05, + "loss": 2.608, + "step": 25053 + }, + { + "epoch": 1.07, + "grad_norm": 0.47265625, + "learning_rate": 8.970816522567445e-05, + "loss": 2.469, + "step": 25054 + }, + { + "epoch": 1.07, + "grad_norm": 0.55859375, + "learning_rate": 8.970149293770828e-05, + "loss": 2.4868, + "step": 25055 + }, + { + "epoch": 1.07, + "grad_norm": 0.5078125, + "learning_rate": 8.969482069608162e-05, + "loss": 2.4715, + "step": 25056 + }, + { + "epoch": 1.07, + "grad_norm": 0.474609375, + "learning_rate": 8.96881485008245e-05, + "loss": 2.3471, + "step": 25057 + }, + { + "epoch": 1.07, + "grad_norm": 0.400390625, + "learning_rate": 8.968147635196694e-05, + "loss": 2.3259, + "step": 25058 + }, + { + "epoch": 1.07, + "grad_norm": 0.455078125, + "learning_rate": 8.9674804249539e-05, + "loss": 2.4179, + "step": 25059 + }, + { + "epoch": 1.07, + "grad_norm": 0.5234375, + "learning_rate": 8.96681321935706e-05, + "loss": 2.537, + "step": 25060 + }, + { + "epoch": 1.07, + "grad_norm": 0.53515625, + "learning_rate": 8.966146018409189e-05, + "loss": 2.4915, + "step": 25061 + }, + { + "epoch": 1.07, + "grad_norm": 0.44140625, + "learning_rate": 8.965478822113282e-05, + "loss": 2.5076, + "step": 25062 + }, + { + "epoch": 1.07, + "grad_norm": 0.474609375, + "learning_rate": 8.96481163047234e-05, + "loss": 2.3436, + "step": 25063 + }, + { + "epoch": 1.07, + "grad_norm": 0.478515625, + "learning_rate": 8.964144443489368e-05, + "loss": 2.5318, + "step": 25064 + }, + { + "epoch": 1.07, + "grad_norm": 0.46875, + "learning_rate": 8.963477261167368e-05, + "loss": 2.4478, + "step": 25065 + }, + { + "epoch": 1.07, + "grad_norm": 0.478515625, + "learning_rate": 8.962810083509342e-05, + "loss": 2.5328, + "step": 25066 + }, + { + "epoch": 1.07, + "grad_norm": 0.453125, + "learning_rate": 8.962142910518293e-05, + "loss": 2.4615, + "step": 25067 + }, + { + "epoch": 1.07, + "grad_norm": 0.44921875, + "learning_rate": 8.961475742197216e-05, + "loss": 2.4526, + "step": 25068 + }, + { + "epoch": 1.07, + "grad_norm": 0.41015625, + "learning_rate": 8.960808578549125e-05, + "loss": 2.4915, + "step": 25069 + }, + { + "epoch": 1.07, + "grad_norm": 0.41796875, + "learning_rate": 8.960141419577013e-05, + "loss": 2.4362, + "step": 25070 + }, + { + "epoch": 1.07, + "grad_norm": 0.47265625, + "learning_rate": 8.959474265283885e-05, + "loss": 2.4159, + "step": 25071 + }, + { + "epoch": 1.07, + "grad_norm": 0.455078125, + "learning_rate": 8.95880711567274e-05, + "loss": 2.5777, + "step": 25072 + }, + { + "epoch": 1.07, + "grad_norm": 0.482421875, + "learning_rate": 8.958139970746584e-05, + "loss": 2.5351, + "step": 25073 + }, + { + "epoch": 1.07, + "grad_norm": 0.4296875, + "learning_rate": 8.957472830508419e-05, + "loss": 2.4448, + "step": 25074 + }, + { + "epoch": 1.07, + "grad_norm": 0.515625, + "learning_rate": 8.956805694961247e-05, + "loss": 2.4141, + "step": 25075 + }, + { + "epoch": 1.07, + "grad_norm": 0.45703125, + "learning_rate": 8.956138564108061e-05, + "loss": 2.5704, + "step": 25076 + }, + { + "epoch": 1.07, + "grad_norm": 0.427734375, + "learning_rate": 8.955471437951877e-05, + "loss": 2.4467, + "step": 25077 + }, + { + "epoch": 1.07, + "grad_norm": 0.53125, + "learning_rate": 8.954804316495687e-05, + "loss": 2.4885, + "step": 25078 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 8.954137199742494e-05, + "loss": 2.4096, + "step": 25079 + }, + { + "epoch": 1.07, + "grad_norm": 0.447265625, + "learning_rate": 8.953470087695303e-05, + "loss": 2.3197, + "step": 25080 + }, + { + "epoch": 1.07, + "grad_norm": 0.53125, + "learning_rate": 8.952802980357114e-05, + "loss": 2.5338, + "step": 25081 + }, + { + "epoch": 1.07, + "grad_norm": 0.41796875, + "learning_rate": 8.952135877730929e-05, + "loss": 2.3435, + "step": 25082 + }, + { + "epoch": 1.07, + "grad_norm": 0.4921875, + "learning_rate": 8.951468779819753e-05, + "loss": 2.4338, + "step": 25083 + }, + { + "epoch": 1.07, + "grad_norm": 0.435546875, + "learning_rate": 8.950801686626576e-05, + "loss": 2.475, + "step": 25084 + }, + { + "epoch": 1.07, + "grad_norm": 0.451171875, + "learning_rate": 8.950134598154416e-05, + "loss": 2.4675, + "step": 25085 + }, + { + "epoch": 1.07, + "grad_norm": 0.478515625, + "learning_rate": 8.949467514406264e-05, + "loss": 2.4764, + "step": 25086 + }, + { + "epoch": 1.07, + "grad_norm": 0.515625, + "learning_rate": 8.948800435385123e-05, + "loss": 2.5687, + "step": 25087 + }, + { + "epoch": 1.07, + "grad_norm": 0.48046875, + "learning_rate": 8.948133361093996e-05, + "loss": 2.4183, + "step": 25088 + }, + { + "epoch": 1.07, + "grad_norm": 0.515625, + "learning_rate": 8.947466291535884e-05, + "loss": 2.5181, + "step": 25089 + }, + { + "epoch": 1.07, + "grad_norm": 0.49609375, + "learning_rate": 8.946799226713791e-05, + "loss": 2.5563, + "step": 25090 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 8.946132166630719e-05, + "loss": 2.4288, + "step": 25091 + }, + { + "epoch": 1.07, + "grad_norm": 0.3984375, + "learning_rate": 8.945465111289659e-05, + "loss": 2.3585, + "step": 25092 + }, + { + "epoch": 1.07, + "grad_norm": 0.48828125, + "learning_rate": 8.944798060693628e-05, + "loss": 2.4687, + "step": 25093 + }, + { + "epoch": 1.07, + "grad_norm": 0.44921875, + "learning_rate": 8.944131014845618e-05, + "loss": 2.4025, + "step": 25094 + }, + { + "epoch": 1.07, + "grad_norm": 0.443359375, + "learning_rate": 8.943463973748633e-05, + "loss": 2.5009, + "step": 25095 + }, + { + "epoch": 1.07, + "grad_norm": 0.51953125, + "learning_rate": 8.942796937405673e-05, + "loss": 2.5955, + "step": 25096 + }, + { + "epoch": 1.07, + "grad_norm": 0.48828125, + "learning_rate": 8.942129905819738e-05, + "loss": 2.5947, + "step": 25097 + }, + { + "epoch": 1.07, + "grad_norm": 0.46875, + "learning_rate": 8.941462878993835e-05, + "loss": 2.4984, + "step": 25098 + }, + { + "epoch": 1.07, + "grad_norm": 0.466796875, + "learning_rate": 8.940795856930964e-05, + "loss": 2.4503, + "step": 25099 + }, + { + "epoch": 1.07, + "grad_norm": 0.404296875, + "learning_rate": 8.940128839634123e-05, + "loss": 2.3898, + "step": 25100 + }, + { + "epoch": 1.07, + "grad_norm": 0.41796875, + "learning_rate": 8.939461827106311e-05, + "loss": 2.458, + "step": 25101 + }, + { + "epoch": 1.07, + "grad_norm": 0.46484375, + "learning_rate": 8.938794819350536e-05, + "loss": 2.4212, + "step": 25102 + }, + { + "epoch": 1.07, + "grad_norm": 0.53125, + "learning_rate": 8.938127816369798e-05, + "loss": 2.6149, + "step": 25103 + }, + { + "epoch": 1.07, + "grad_norm": 0.466796875, + "learning_rate": 8.937460818167095e-05, + "loss": 2.4724, + "step": 25104 + }, + { + "epoch": 1.07, + "grad_norm": 0.431640625, + "learning_rate": 8.936793824745428e-05, + "loss": 2.4515, + "step": 25105 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 8.936126836107803e-05, + "loss": 2.6295, + "step": 25106 + }, + { + "epoch": 1.07, + "grad_norm": 0.515625, + "learning_rate": 8.935459852257221e-05, + "loss": 2.4744, + "step": 25107 + }, + { + "epoch": 1.07, + "grad_norm": 0.40625, + "learning_rate": 8.934792873196678e-05, + "loss": 2.4556, + "step": 25108 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 8.934125898929175e-05, + "loss": 2.5634, + "step": 25109 + }, + { + "epoch": 1.07, + "grad_norm": 0.474609375, + "learning_rate": 8.93345892945772e-05, + "loss": 2.5267, + "step": 25110 + }, + { + "epoch": 1.07, + "grad_norm": 0.45703125, + "learning_rate": 8.932791964785309e-05, + "loss": 2.4311, + "step": 25111 + }, + { + "epoch": 1.07, + "grad_norm": 0.484375, + "learning_rate": 8.932125004914943e-05, + "loss": 2.5089, + "step": 25112 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 8.931458049849624e-05, + "loss": 2.3175, + "step": 25113 + }, + { + "epoch": 1.07, + "grad_norm": 0.478515625, + "learning_rate": 8.930791099592354e-05, + "loss": 2.4405, + "step": 25114 + }, + { + "epoch": 1.07, + "grad_norm": 0.4375, + "learning_rate": 8.930124154146136e-05, + "loss": 2.5453, + "step": 25115 + }, + { + "epoch": 1.07, + "grad_norm": 0.4453125, + "learning_rate": 8.929457213513966e-05, + "loss": 2.3326, + "step": 25116 + }, + { + "epoch": 1.07, + "grad_norm": 0.49609375, + "learning_rate": 8.928790277698845e-05, + "loss": 2.5426, + "step": 25117 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 8.928123346703778e-05, + "loss": 2.5199, + "step": 25118 + }, + { + "epoch": 1.07, + "grad_norm": 0.431640625, + "learning_rate": 8.927456420531765e-05, + "loss": 2.7116, + "step": 25119 + }, + { + "epoch": 1.07, + "grad_norm": 0.466796875, + "learning_rate": 8.926789499185806e-05, + "loss": 2.4007, + "step": 25120 + }, + { + "epoch": 1.07, + "grad_norm": 0.40625, + "learning_rate": 8.926122582668898e-05, + "loss": 2.3968, + "step": 25121 + }, + { + "epoch": 1.07, + "grad_norm": 0.44921875, + "learning_rate": 8.925455670984051e-05, + "loss": 2.4896, + "step": 25122 + }, + { + "epoch": 1.07, + "grad_norm": 0.5234375, + "learning_rate": 8.924788764134262e-05, + "loss": 2.5782, + "step": 25123 + }, + { + "epoch": 1.07, + "grad_norm": 0.443359375, + "learning_rate": 8.924121862122527e-05, + "loss": 2.5831, + "step": 25124 + }, + { + "epoch": 1.07, + "grad_norm": 0.478515625, + "learning_rate": 8.92345496495185e-05, + "loss": 2.5418, + "step": 25125 + }, + { + "epoch": 1.07, + "grad_norm": 0.435546875, + "learning_rate": 8.922788072625235e-05, + "loss": 2.4638, + "step": 25126 + }, + { + "epoch": 1.07, + "grad_norm": 0.4609375, + "learning_rate": 8.922121185145679e-05, + "loss": 2.5162, + "step": 25127 + }, + { + "epoch": 1.07, + "grad_norm": 0.5, + "learning_rate": 8.921454302516184e-05, + "loss": 2.4686, + "step": 25128 + }, + { + "epoch": 1.07, + "grad_norm": 0.447265625, + "learning_rate": 8.920787424739747e-05, + "loss": 2.3155, + "step": 25129 + }, + { + "epoch": 1.07, + "grad_norm": 0.474609375, + "learning_rate": 8.920120551819375e-05, + "loss": 2.4299, + "step": 25130 + }, + { + "epoch": 1.07, + "grad_norm": 0.51953125, + "learning_rate": 8.919453683758068e-05, + "loss": 2.4695, + "step": 25131 + }, + { + "epoch": 1.07, + "grad_norm": 0.53515625, + "learning_rate": 8.918786820558823e-05, + "loss": 2.4012, + "step": 25132 + }, + { + "epoch": 1.07, + "grad_norm": 0.453125, + "learning_rate": 8.91811996222464e-05, + "loss": 2.4307, + "step": 25133 + }, + { + "epoch": 1.07, + "grad_norm": 0.46875, + "learning_rate": 8.917453108758523e-05, + "loss": 2.4679, + "step": 25134 + }, + { + "epoch": 1.07, + "grad_norm": 0.4453125, + "learning_rate": 8.916786260163472e-05, + "loss": 2.4226, + "step": 25135 + }, + { + "epoch": 1.07, + "grad_norm": 0.53515625, + "learning_rate": 8.916119416442488e-05, + "loss": 2.3672, + "step": 25136 + }, + { + "epoch": 1.07, + "grad_norm": 0.40234375, + "learning_rate": 8.915452577598566e-05, + "loss": 2.4124, + "step": 25137 + }, + { + "epoch": 1.07, + "grad_norm": 0.46875, + "learning_rate": 8.914785743634715e-05, + "loss": 2.2891, + "step": 25138 + }, + { + "epoch": 1.07, + "grad_norm": 0.4765625, + "learning_rate": 8.914118914553934e-05, + "loss": 2.481, + "step": 25139 + }, + { + "epoch": 1.07, + "grad_norm": 0.46484375, + "learning_rate": 8.913452090359218e-05, + "loss": 2.4592, + "step": 25140 + }, + { + "epoch": 1.07, + "grad_norm": 0.431640625, + "learning_rate": 8.912785271053569e-05, + "loss": 2.3146, + "step": 25141 + }, + { + "epoch": 1.07, + "grad_norm": 0.56640625, + "learning_rate": 8.91211845663999e-05, + "loss": 2.5241, + "step": 25142 + }, + { + "epoch": 1.07, + "grad_norm": 0.4765625, + "learning_rate": 8.911451647121481e-05, + "loss": 2.5467, + "step": 25143 + }, + { + "epoch": 1.07, + "grad_norm": 0.4453125, + "learning_rate": 8.910784842501042e-05, + "loss": 2.4217, + "step": 25144 + }, + { + "epoch": 1.07, + "grad_norm": 0.4609375, + "learning_rate": 8.91011804278167e-05, + "loss": 2.4374, + "step": 25145 + }, + { + "epoch": 1.07, + "grad_norm": 0.5234375, + "learning_rate": 8.909451247966371e-05, + "loss": 2.5872, + "step": 25146 + }, + { + "epoch": 1.07, + "grad_norm": 0.466796875, + "learning_rate": 8.908784458058145e-05, + "loss": 2.455, + "step": 25147 + }, + { + "epoch": 1.07, + "grad_norm": 0.4140625, + "learning_rate": 8.908117673059989e-05, + "loss": 2.507, + "step": 25148 + }, + { + "epoch": 1.07, + "grad_norm": 0.453125, + "learning_rate": 8.907450892974902e-05, + "loss": 2.4934, + "step": 25149 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 8.906784117805888e-05, + "loss": 2.3571, + "step": 25150 + }, + { + "epoch": 1.07, + "grad_norm": 0.48046875, + "learning_rate": 8.906117347555947e-05, + "loss": 2.5261, + "step": 25151 + }, + { + "epoch": 1.07, + "grad_norm": 0.41796875, + "learning_rate": 8.905450582228078e-05, + "loss": 2.3045, + "step": 25152 + }, + { + "epoch": 1.07, + "grad_norm": 0.458984375, + "learning_rate": 8.904783821825279e-05, + "loss": 2.5111, + "step": 25153 + }, + { + "epoch": 1.07, + "grad_norm": 0.4609375, + "learning_rate": 8.904117066350553e-05, + "loss": 2.4447, + "step": 25154 + }, + { + "epoch": 1.07, + "grad_norm": 0.47265625, + "learning_rate": 8.903450315806904e-05, + "loss": 2.4559, + "step": 25155 + }, + { + "epoch": 1.07, + "grad_norm": 0.490234375, + "learning_rate": 8.902783570197325e-05, + "loss": 2.3998, + "step": 25156 + }, + { + "epoch": 1.07, + "grad_norm": 0.47265625, + "learning_rate": 8.902116829524817e-05, + "loss": 2.4648, + "step": 25157 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 8.901450093792384e-05, + "loss": 2.4654, + "step": 25158 + }, + { + "epoch": 1.07, + "grad_norm": 0.466796875, + "learning_rate": 8.900783363003025e-05, + "loss": 2.4869, + "step": 25159 + }, + { + "epoch": 1.07, + "grad_norm": 0.53125, + "learning_rate": 8.900116637159738e-05, + "loss": 2.4711, + "step": 25160 + }, + { + "epoch": 1.07, + "grad_norm": 0.46484375, + "learning_rate": 8.899449916265522e-05, + "loss": 2.5983, + "step": 25161 + }, + { + "epoch": 1.07, + "grad_norm": 0.48828125, + "learning_rate": 8.898783200323381e-05, + "loss": 2.3597, + "step": 25162 + }, + { + "epoch": 1.07, + "grad_norm": 0.453125, + "learning_rate": 8.898116489336317e-05, + "loss": 2.4742, + "step": 25163 + }, + { + "epoch": 1.07, + "grad_norm": 0.484375, + "learning_rate": 8.897449783307322e-05, + "loss": 2.4437, + "step": 25164 + }, + { + "epoch": 1.07, + "grad_norm": 0.47265625, + "learning_rate": 8.8967830822394e-05, + "loss": 2.5615, + "step": 25165 + }, + { + "epoch": 1.07, + "grad_norm": 0.4453125, + "learning_rate": 8.89611638613555e-05, + "loss": 2.4411, + "step": 25166 + }, + { + "epoch": 1.07, + "grad_norm": 0.52734375, + "learning_rate": 8.895449694998777e-05, + "loss": 2.4743, + "step": 25167 + }, + { + "epoch": 1.07, + "grad_norm": 0.48828125, + "learning_rate": 8.894783008832073e-05, + "loss": 2.4741, + "step": 25168 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 8.894116327638441e-05, + "loss": 2.3736, + "step": 25169 + }, + { + "epoch": 1.07, + "grad_norm": 0.494140625, + "learning_rate": 8.893449651420883e-05, + "loss": 2.4339, + "step": 25170 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 8.8927829801824e-05, + "loss": 2.4333, + "step": 25171 + }, + { + "epoch": 1.07, + "grad_norm": 0.52734375, + "learning_rate": 8.892116313925987e-05, + "loss": 2.4251, + "step": 25172 + }, + { + "epoch": 1.07, + "grad_norm": 0.45703125, + "learning_rate": 8.891449652654642e-05, + "loss": 2.5628, + "step": 25173 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 8.890782996371372e-05, + "loss": 2.4543, + "step": 25174 + }, + { + "epoch": 1.07, + "grad_norm": 0.466796875, + "learning_rate": 8.890116345079174e-05, + "loss": 2.4883, + "step": 25175 + }, + { + "epoch": 1.07, + "grad_norm": 0.546875, + "learning_rate": 8.889449698781045e-05, + "loss": 2.3962, + "step": 25176 + }, + { + "epoch": 1.07, + "grad_norm": 0.427734375, + "learning_rate": 8.888783057479985e-05, + "loss": 2.3721, + "step": 25177 + }, + { + "epoch": 1.07, + "grad_norm": 0.482421875, + "learning_rate": 8.888116421178996e-05, + "loss": 2.4518, + "step": 25178 + }, + { + "epoch": 1.07, + "grad_norm": 0.4765625, + "learning_rate": 8.887449789881081e-05, + "loss": 2.4963, + "step": 25179 + }, + { + "epoch": 1.07, + "grad_norm": 0.48828125, + "learning_rate": 8.886783163589231e-05, + "loss": 2.6225, + "step": 25180 + }, + { + "epoch": 1.07, + "grad_norm": 0.43359375, + "learning_rate": 8.88611654230645e-05, + "loss": 2.4902, + "step": 25181 + }, + { + "epoch": 1.07, + "grad_norm": 0.470703125, + "learning_rate": 8.88544992603574e-05, + "loss": 2.4375, + "step": 25182 + }, + { + "epoch": 1.07, + "grad_norm": 0.490234375, + "learning_rate": 8.884783314780096e-05, + "loss": 2.4219, + "step": 25183 + }, + { + "epoch": 1.07, + "grad_norm": 0.46484375, + "learning_rate": 8.88411670854252e-05, + "loss": 2.4765, + "step": 25184 + }, + { + "epoch": 1.07, + "grad_norm": 0.451171875, + "learning_rate": 8.883450107326008e-05, + "loss": 2.4926, + "step": 25185 + }, + { + "epoch": 1.07, + "grad_norm": 0.4375, + "learning_rate": 8.882783511133567e-05, + "loss": 2.4852, + "step": 25186 + }, + { + "epoch": 1.07, + "grad_norm": 0.4609375, + "learning_rate": 8.882116919968193e-05, + "loss": 2.5744, + "step": 25187 + }, + { + "epoch": 1.07, + "grad_norm": 0.484375, + "learning_rate": 8.881450333832881e-05, + "loss": 2.4224, + "step": 25188 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 8.880783752730633e-05, + "loss": 2.406, + "step": 25189 + }, + { + "epoch": 1.07, + "grad_norm": 0.439453125, + "learning_rate": 8.88011717666445e-05, + "loss": 2.4116, + "step": 25190 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 8.879450605637331e-05, + "loss": 2.5097, + "step": 25191 + }, + { + "epoch": 1.07, + "grad_norm": 0.453125, + "learning_rate": 8.878784039652276e-05, + "loss": 2.4801, + "step": 25192 + }, + { + "epoch": 1.07, + "grad_norm": 0.46484375, + "learning_rate": 8.878117478712281e-05, + "loss": 2.4848, + "step": 25193 + }, + { + "epoch": 1.07, + "grad_norm": 0.431640625, + "learning_rate": 8.877450922820347e-05, + "loss": 2.4273, + "step": 25194 + }, + { + "epoch": 1.07, + "grad_norm": 0.451171875, + "learning_rate": 8.876784371979478e-05, + "loss": 2.5137, + "step": 25195 + }, + { + "epoch": 1.07, + "grad_norm": 0.435546875, + "learning_rate": 8.876117826192665e-05, + "loss": 2.3164, + "step": 25196 + }, + { + "epoch": 1.07, + "grad_norm": 0.478515625, + "learning_rate": 8.87545128546291e-05, + "loss": 2.4674, + "step": 25197 + }, + { + "epoch": 1.07, + "grad_norm": 0.47265625, + "learning_rate": 8.874784749793215e-05, + "loss": 2.4275, + "step": 25198 + }, + { + "epoch": 1.07, + "grad_norm": 0.51171875, + "learning_rate": 8.874118219186578e-05, + "loss": 2.5257, + "step": 25199 + }, + { + "epoch": 1.07, + "grad_norm": 0.4921875, + "learning_rate": 8.873451693645997e-05, + "loss": 2.4331, + "step": 25200 + }, + { + "epoch": 1.07, + "grad_norm": 0.45703125, + "learning_rate": 8.87278517317447e-05, + "loss": 2.4488, + "step": 25201 + }, + { + "epoch": 1.07, + "grad_norm": 0.453125, + "learning_rate": 8.872118657774999e-05, + "loss": 2.3908, + "step": 25202 + }, + { + "epoch": 1.07, + "grad_norm": 0.51953125, + "learning_rate": 8.871452147450583e-05, + "loss": 2.4068, + "step": 25203 + }, + { + "epoch": 1.07, + "grad_norm": 0.447265625, + "learning_rate": 8.87078564220422e-05, + "loss": 2.4715, + "step": 25204 + }, + { + "epoch": 1.07, + "grad_norm": 0.462890625, + "learning_rate": 8.870119142038906e-05, + "loss": 2.4707, + "step": 25205 + }, + { + "epoch": 1.07, + "grad_norm": 0.474609375, + "learning_rate": 8.869452646957646e-05, + "loss": 2.569, + "step": 25206 + }, + { + "epoch": 1.07, + "grad_norm": 0.48046875, + "learning_rate": 8.868786156963435e-05, + "loss": 2.4897, + "step": 25207 + }, + { + "epoch": 1.07, + "grad_norm": 0.455078125, + "learning_rate": 8.868119672059272e-05, + "loss": 2.4986, + "step": 25208 + }, + { + "epoch": 1.07, + "grad_norm": 0.44921875, + "learning_rate": 8.867453192248156e-05, + "loss": 2.4847, + "step": 25209 + }, + { + "epoch": 1.07, + "grad_norm": 0.4453125, + "learning_rate": 8.866786717533089e-05, + "loss": 2.4932, + "step": 25210 + }, + { + "epoch": 1.07, + "grad_norm": 0.458984375, + "learning_rate": 8.866120247917069e-05, + "loss": 2.3331, + "step": 25211 + }, + { + "epoch": 1.07, + "grad_norm": 0.4921875, + "learning_rate": 8.865453783403091e-05, + "loss": 2.4484, + "step": 25212 + }, + { + "epoch": 1.07, + "grad_norm": 0.4296875, + "learning_rate": 8.864787323994155e-05, + "loss": 2.4505, + "step": 25213 + }, + { + "epoch": 1.07, + "grad_norm": 0.50390625, + "learning_rate": 8.864120869693263e-05, + "loss": 2.5921, + "step": 25214 + }, + { + "epoch": 1.07, + "grad_norm": 0.484375, + "learning_rate": 8.863454420503412e-05, + "loss": 2.5093, + "step": 25215 + }, + { + "epoch": 1.07, + "grad_norm": 0.44140625, + "learning_rate": 8.862787976427601e-05, + "loss": 2.5164, + "step": 25216 + }, + { + "epoch": 1.07, + "grad_norm": 0.5078125, + "learning_rate": 8.86212153746883e-05, + "loss": 2.4541, + "step": 25217 + }, + { + "epoch": 1.07, + "grad_norm": 0.474609375, + "learning_rate": 8.861455103630091e-05, + "loss": 2.5019, + "step": 25218 + }, + { + "epoch": 1.07, + "grad_norm": 0.451171875, + "learning_rate": 8.860788674914394e-05, + "loss": 2.398, + "step": 25219 + }, + { + "epoch": 1.07, + "grad_norm": 0.44140625, + "learning_rate": 8.860122251324729e-05, + "loss": 2.4242, + "step": 25220 + }, + { + "epoch": 1.07, + "grad_norm": 0.486328125, + "learning_rate": 8.859455832864097e-05, + "loss": 2.4316, + "step": 25221 + }, + { + "epoch": 1.07, + "grad_norm": 0.44140625, + "learning_rate": 8.858789419535495e-05, + "loss": 2.4585, + "step": 25222 + }, + { + "epoch": 1.07, + "grad_norm": 0.421875, + "learning_rate": 8.858123011341926e-05, + "loss": 2.6285, + "step": 25223 + }, + { + "epoch": 1.07, + "grad_norm": 0.46875, + "learning_rate": 8.857456608286386e-05, + "loss": 2.5613, + "step": 25224 + }, + { + "epoch": 1.07, + "grad_norm": 0.423828125, + "learning_rate": 8.856790210371876e-05, + "loss": 2.4223, + "step": 25225 + }, + { + "epoch": 1.07, + "grad_norm": 0.45703125, + "learning_rate": 8.856123817601386e-05, + "loss": 2.38, + "step": 25226 + }, + { + "epoch": 1.07, + "grad_norm": 0.466796875, + "learning_rate": 8.855457429977927e-05, + "loss": 2.4088, + "step": 25227 + }, + { + "epoch": 1.08, + "grad_norm": 0.482421875, + "learning_rate": 8.854791047504489e-05, + "loss": 2.5793, + "step": 25228 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.854124670184072e-05, + "loss": 2.3736, + "step": 25229 + }, + { + "epoch": 1.08, + "grad_norm": 0.4765625, + "learning_rate": 8.853458298019674e-05, + "loss": 2.4033, + "step": 25230 + }, + { + "epoch": 1.08, + "grad_norm": 0.453125, + "learning_rate": 8.852791931014297e-05, + "loss": 2.4184, + "step": 25231 + }, + { + "epoch": 1.08, + "grad_norm": 0.478515625, + "learning_rate": 8.852125569170937e-05, + "loss": 2.3992, + "step": 25232 + }, + { + "epoch": 1.08, + "grad_norm": 0.47265625, + "learning_rate": 8.851459212492594e-05, + "loss": 2.3136, + "step": 25233 + }, + { + "epoch": 1.08, + "grad_norm": 0.435546875, + "learning_rate": 8.85079286098226e-05, + "loss": 2.5234, + "step": 25234 + }, + { + "epoch": 1.08, + "grad_norm": 0.447265625, + "learning_rate": 8.850126514642943e-05, + "loss": 2.4678, + "step": 25235 + }, + { + "epoch": 1.08, + "grad_norm": 0.431640625, + "learning_rate": 8.849460173477634e-05, + "loss": 2.4475, + "step": 25236 + }, + { + "epoch": 1.08, + "grad_norm": 0.4453125, + "learning_rate": 8.848793837489334e-05, + "loss": 2.5321, + "step": 25237 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.848127506681039e-05, + "loss": 2.515, + "step": 25238 + }, + { + "epoch": 1.08, + "grad_norm": 0.48828125, + "learning_rate": 8.847461181055752e-05, + "loss": 2.5092, + "step": 25239 + }, + { + "epoch": 1.08, + "grad_norm": 0.453125, + "learning_rate": 8.846794860616468e-05, + "loss": 2.4948, + "step": 25240 + }, + { + "epoch": 1.08, + "grad_norm": 0.546875, + "learning_rate": 8.846128545366188e-05, + "loss": 2.4808, + "step": 25241 + }, + { + "epoch": 1.08, + "grad_norm": 0.49609375, + "learning_rate": 8.845462235307901e-05, + "loss": 2.5043, + "step": 25242 + }, + { + "epoch": 1.08, + "grad_norm": 0.486328125, + "learning_rate": 8.84479593044462e-05, + "loss": 2.4362, + "step": 25243 + }, + { + "epoch": 1.08, + "grad_norm": 0.49609375, + "learning_rate": 8.844129630779331e-05, + "loss": 2.562, + "step": 25244 + }, + { + "epoch": 1.08, + "grad_norm": 0.453125, + "learning_rate": 8.843463336315037e-05, + "loss": 2.5551, + "step": 25245 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.842797047054733e-05, + "loss": 2.5122, + "step": 25246 + }, + { + "epoch": 1.08, + "grad_norm": 0.478515625, + "learning_rate": 8.842130763001422e-05, + "loss": 2.4516, + "step": 25247 + }, + { + "epoch": 1.08, + "grad_norm": 0.455078125, + "learning_rate": 8.841464484158098e-05, + "loss": 2.421, + "step": 25248 + }, + { + "epoch": 1.08, + "grad_norm": 0.52734375, + "learning_rate": 8.840798210527765e-05, + "loss": 2.5889, + "step": 25249 + }, + { + "epoch": 1.08, + "grad_norm": 0.53125, + "learning_rate": 8.840131942113407e-05, + "loss": 2.6602, + "step": 25250 + }, + { + "epoch": 1.08, + "grad_norm": 0.515625, + "learning_rate": 8.839465678918041e-05, + "loss": 2.568, + "step": 25251 + }, + { + "epoch": 1.08, + "grad_norm": 0.5390625, + "learning_rate": 8.83879942094465e-05, + "loss": 2.4943, + "step": 25252 + }, + { + "epoch": 1.08, + "grad_norm": 0.361328125, + "learning_rate": 8.838133168196239e-05, + "loss": 2.3176, + "step": 25253 + }, + { + "epoch": 1.08, + "grad_norm": 0.46875, + "learning_rate": 8.837466920675803e-05, + "loss": 2.5111, + "step": 25254 + }, + { + "epoch": 1.08, + "grad_norm": 0.45703125, + "learning_rate": 8.836800678386342e-05, + "loss": 2.4673, + "step": 25255 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.836134441330852e-05, + "loss": 2.3924, + "step": 25256 + }, + { + "epoch": 1.08, + "grad_norm": 0.470703125, + "learning_rate": 8.835468209512336e-05, + "loss": 2.4755, + "step": 25257 + }, + { + "epoch": 1.08, + "grad_norm": 0.478515625, + "learning_rate": 8.834801982933779e-05, + "loss": 2.4511, + "step": 25258 + }, + { + "epoch": 1.08, + "grad_norm": 0.462890625, + "learning_rate": 8.834135761598194e-05, + "loss": 2.3557, + "step": 25259 + }, + { + "epoch": 1.08, + "grad_norm": 0.43359375, + "learning_rate": 8.83346954550857e-05, + "loss": 2.4163, + "step": 25260 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.832803334667907e-05, + "loss": 2.4456, + "step": 25261 + }, + { + "epoch": 1.08, + "grad_norm": 0.439453125, + "learning_rate": 8.832137129079201e-05, + "loss": 2.364, + "step": 25262 + }, + { + "epoch": 1.08, + "grad_norm": 0.435546875, + "learning_rate": 8.831470928745452e-05, + "loss": 2.4255, + "step": 25263 + }, + { + "epoch": 1.08, + "grad_norm": 0.478515625, + "learning_rate": 8.830804733669657e-05, + "loss": 2.4814, + "step": 25264 + }, + { + "epoch": 1.08, + "grad_norm": 0.44140625, + "learning_rate": 8.830138543854816e-05, + "loss": 2.5427, + "step": 25265 + }, + { + "epoch": 1.08, + "grad_norm": 0.4609375, + "learning_rate": 8.829472359303917e-05, + "loss": 2.5592, + "step": 25266 + }, + { + "epoch": 1.08, + "grad_norm": 0.4609375, + "learning_rate": 8.828806180019973e-05, + "loss": 2.5304, + "step": 25267 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.82814000600597e-05, + "loss": 2.425, + "step": 25268 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.827473837264906e-05, + "loss": 2.4211, + "step": 25269 + }, + { + "epoch": 1.08, + "grad_norm": 0.494140625, + "learning_rate": 8.826807673799782e-05, + "loss": 2.4687, + "step": 25270 + }, + { + "epoch": 1.08, + "grad_norm": 0.4609375, + "learning_rate": 8.826141515613596e-05, + "loss": 2.4485, + "step": 25271 + }, + { + "epoch": 1.08, + "grad_norm": 0.4765625, + "learning_rate": 8.825475362709344e-05, + "loss": 2.4224, + "step": 25272 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.824809215090026e-05, + "loss": 2.6394, + "step": 25273 + }, + { + "epoch": 1.08, + "grad_norm": 0.48046875, + "learning_rate": 8.824143072758631e-05, + "loss": 2.508, + "step": 25274 + }, + { + "epoch": 1.08, + "grad_norm": 0.5546875, + "learning_rate": 8.823476935718169e-05, + "loss": 2.4963, + "step": 25275 + }, + { + "epoch": 1.08, + "grad_norm": 0.51171875, + "learning_rate": 8.822810803971629e-05, + "loss": 2.4818, + "step": 25276 + }, + { + "epoch": 1.08, + "grad_norm": 0.421875, + "learning_rate": 8.822144677522009e-05, + "loss": 2.4513, + "step": 25277 + }, + { + "epoch": 1.08, + "grad_norm": 0.58984375, + "learning_rate": 8.821478556372306e-05, + "loss": 2.5135, + "step": 25278 + }, + { + "epoch": 1.08, + "grad_norm": 0.4921875, + "learning_rate": 8.82081244052552e-05, + "loss": 2.5866, + "step": 25279 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.820146329984648e-05, + "loss": 2.4563, + "step": 25280 + }, + { + "epoch": 1.08, + "grad_norm": 0.451171875, + "learning_rate": 8.819480224752689e-05, + "loss": 2.5524, + "step": 25281 + }, + { + "epoch": 1.08, + "grad_norm": 0.44140625, + "learning_rate": 8.81881412483263e-05, + "loss": 2.4988, + "step": 25282 + }, + { + "epoch": 1.08, + "grad_norm": 0.48046875, + "learning_rate": 8.818148030227481e-05, + "loss": 2.5914, + "step": 25283 + }, + { + "epoch": 1.08, + "grad_norm": 0.57421875, + "learning_rate": 8.817481940940234e-05, + "loss": 2.6225, + "step": 25284 + }, + { + "epoch": 1.08, + "grad_norm": 0.451171875, + "learning_rate": 8.816815856973886e-05, + "loss": 2.5531, + "step": 25285 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.816149778331432e-05, + "loss": 2.4819, + "step": 25286 + }, + { + "epoch": 1.08, + "grad_norm": 0.439453125, + "learning_rate": 8.815483705015873e-05, + "loss": 2.5083, + "step": 25287 + }, + { + "epoch": 1.08, + "grad_norm": 0.478515625, + "learning_rate": 8.814817637030204e-05, + "loss": 2.4878, + "step": 25288 + }, + { + "epoch": 1.08, + "grad_norm": 0.44140625, + "learning_rate": 8.814151574377425e-05, + "loss": 2.4577, + "step": 25289 + }, + { + "epoch": 1.08, + "grad_norm": 0.427734375, + "learning_rate": 8.813485517060525e-05, + "loss": 2.42, + "step": 25290 + }, + { + "epoch": 1.08, + "grad_norm": 0.515625, + "learning_rate": 8.812819465082512e-05, + "loss": 2.4791, + "step": 25291 + }, + { + "epoch": 1.08, + "grad_norm": 0.486328125, + "learning_rate": 8.812153418446374e-05, + "loss": 2.5367, + "step": 25292 + }, + { + "epoch": 1.08, + "grad_norm": 0.451171875, + "learning_rate": 8.811487377155114e-05, + "loss": 2.4772, + "step": 25293 + }, + { + "epoch": 1.08, + "grad_norm": 0.4765625, + "learning_rate": 8.810821341211723e-05, + "loss": 2.448, + "step": 25294 + }, + { + "epoch": 1.08, + "grad_norm": 0.5, + "learning_rate": 8.810155310619205e-05, + "loss": 2.4625, + "step": 25295 + }, + { + "epoch": 1.08, + "grad_norm": 0.44921875, + "learning_rate": 8.809489285380553e-05, + "loss": 2.5241, + "step": 25296 + }, + { + "epoch": 1.08, + "grad_norm": 0.484375, + "learning_rate": 8.808823265498764e-05, + "loss": 2.3664, + "step": 25297 + }, + { + "epoch": 1.08, + "grad_norm": 0.498046875, + "learning_rate": 8.808157250976831e-05, + "loss": 2.4642, + "step": 25298 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.80749124181776e-05, + "loss": 2.4217, + "step": 25299 + }, + { + "epoch": 1.08, + "grad_norm": 0.5078125, + "learning_rate": 8.80682523802454e-05, + "loss": 2.5162, + "step": 25300 + }, + { + "epoch": 1.08, + "grad_norm": 0.46875, + "learning_rate": 8.806159239600173e-05, + "loss": 2.5986, + "step": 25301 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.805493246547648e-05, + "loss": 2.4409, + "step": 25302 + }, + { + "epoch": 1.08, + "grad_norm": 0.51171875, + "learning_rate": 8.80482725886997e-05, + "loss": 2.4873, + "step": 25303 + }, + { + "epoch": 1.08, + "grad_norm": 0.54296875, + "learning_rate": 8.804161276570133e-05, + "loss": 2.4917, + "step": 25304 + }, + { + "epoch": 1.08, + "grad_norm": 0.4921875, + "learning_rate": 8.803495299651134e-05, + "loss": 2.5332, + "step": 25305 + }, + { + "epoch": 1.08, + "grad_norm": 0.427734375, + "learning_rate": 8.802829328115964e-05, + "loss": 2.4093, + "step": 25306 + }, + { + "epoch": 1.08, + "grad_norm": 0.49609375, + "learning_rate": 8.80216336196763e-05, + "loss": 2.4393, + "step": 25307 + }, + { + "epoch": 1.08, + "grad_norm": 0.484375, + "learning_rate": 8.801497401209122e-05, + "loss": 2.4112, + "step": 25308 + }, + { + "epoch": 1.08, + "grad_norm": 0.47265625, + "learning_rate": 8.800831445843438e-05, + "loss": 2.5129, + "step": 25309 + }, + { + "epoch": 1.08, + "grad_norm": 0.423828125, + "learning_rate": 8.80016549587357e-05, + "loss": 2.6126, + "step": 25310 + }, + { + "epoch": 1.08, + "grad_norm": 0.412109375, + "learning_rate": 8.799499551302523e-05, + "loss": 2.3805, + "step": 25311 + }, + { + "epoch": 1.08, + "grad_norm": 0.4921875, + "learning_rate": 8.798833612133289e-05, + "loss": 2.383, + "step": 25312 + }, + { + "epoch": 1.08, + "grad_norm": 0.48046875, + "learning_rate": 8.798167678368864e-05, + "loss": 2.2558, + "step": 25313 + }, + { + "epoch": 1.08, + "grad_norm": 0.5234375, + "learning_rate": 8.797501750012243e-05, + "loss": 2.5257, + "step": 25314 + }, + { + "epoch": 1.08, + "grad_norm": 0.5390625, + "learning_rate": 8.796835827066428e-05, + "loss": 2.4978, + "step": 25315 + }, + { + "epoch": 1.08, + "grad_norm": 0.462890625, + "learning_rate": 8.796169909534411e-05, + "loss": 2.4497, + "step": 25316 + }, + { + "epoch": 1.08, + "grad_norm": 0.4609375, + "learning_rate": 8.79550399741919e-05, + "loss": 2.4867, + "step": 25317 + }, + { + "epoch": 1.08, + "grad_norm": 0.47265625, + "learning_rate": 8.794838090723759e-05, + "loss": 2.4664, + "step": 25318 + }, + { + "epoch": 1.08, + "grad_norm": 0.515625, + "learning_rate": 8.794172189451116e-05, + "loss": 2.4639, + "step": 25319 + }, + { + "epoch": 1.08, + "grad_norm": 0.4609375, + "learning_rate": 8.793506293604259e-05, + "loss": 2.4099, + "step": 25320 + }, + { + "epoch": 1.08, + "grad_norm": 0.4921875, + "learning_rate": 8.792840403186186e-05, + "loss": 2.5664, + "step": 25321 + }, + { + "epoch": 1.08, + "grad_norm": 0.484375, + "learning_rate": 8.792174518199882e-05, + "loss": 2.3791, + "step": 25322 + }, + { + "epoch": 1.08, + "grad_norm": 0.46875, + "learning_rate": 8.791508638648358e-05, + "loss": 2.2982, + "step": 25323 + }, + { + "epoch": 1.08, + "grad_norm": 0.443359375, + "learning_rate": 8.790842764534601e-05, + "loss": 2.5354, + "step": 25324 + }, + { + "epoch": 1.08, + "grad_norm": 0.48046875, + "learning_rate": 8.79017689586161e-05, + "loss": 2.6403, + "step": 25325 + }, + { + "epoch": 1.08, + "grad_norm": 0.4375, + "learning_rate": 8.78951103263238e-05, + "loss": 2.4313, + "step": 25326 + }, + { + "epoch": 1.08, + "grad_norm": 0.54296875, + "learning_rate": 8.788845174849909e-05, + "loss": 2.5527, + "step": 25327 + }, + { + "epoch": 1.08, + "grad_norm": 0.43359375, + "learning_rate": 8.788179322517192e-05, + "loss": 2.3747, + "step": 25328 + }, + { + "epoch": 1.08, + "grad_norm": 0.419921875, + "learning_rate": 8.787513475637227e-05, + "loss": 2.5566, + "step": 25329 + }, + { + "epoch": 1.08, + "grad_norm": 0.45703125, + "learning_rate": 8.786847634213003e-05, + "loss": 2.4454, + "step": 25330 + }, + { + "epoch": 1.08, + "grad_norm": 0.51171875, + "learning_rate": 8.786181798247526e-05, + "loss": 2.4311, + "step": 25331 + }, + { + "epoch": 1.08, + "grad_norm": 0.451171875, + "learning_rate": 8.785515967743786e-05, + "loss": 2.3902, + "step": 25332 + }, + { + "epoch": 1.08, + "grad_norm": 0.443359375, + "learning_rate": 8.784850142704781e-05, + "loss": 2.5261, + "step": 25333 + }, + { + "epoch": 1.08, + "grad_norm": 0.4609375, + "learning_rate": 8.784184323133503e-05, + "loss": 2.6292, + "step": 25334 + }, + { + "epoch": 1.08, + "grad_norm": 0.482421875, + "learning_rate": 8.783518509032954e-05, + "loss": 2.4946, + "step": 25335 + }, + { + "epoch": 1.08, + "grad_norm": 0.494140625, + "learning_rate": 8.782852700406127e-05, + "loss": 2.4749, + "step": 25336 + }, + { + "epoch": 1.08, + "grad_norm": 0.478515625, + "learning_rate": 8.782186897256021e-05, + "loss": 2.5629, + "step": 25337 + }, + { + "epoch": 1.08, + "grad_norm": 0.4765625, + "learning_rate": 8.781521099585626e-05, + "loss": 2.433, + "step": 25338 + }, + { + "epoch": 1.08, + "grad_norm": 0.470703125, + "learning_rate": 8.780855307397937e-05, + "loss": 2.2845, + "step": 25339 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.780189520695958e-05, + "loss": 2.4731, + "step": 25340 + }, + { + "epoch": 1.08, + "grad_norm": 0.5078125, + "learning_rate": 8.77952373948268e-05, + "loss": 2.5064, + "step": 25341 + }, + { + "epoch": 1.08, + "grad_norm": 0.38671875, + "learning_rate": 8.778857963761097e-05, + "loss": 2.363, + "step": 25342 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.778192193534208e-05, + "loss": 2.4991, + "step": 25343 + }, + { + "epoch": 1.08, + "grad_norm": 0.4921875, + "learning_rate": 8.777526428805007e-05, + "loss": 2.3874, + "step": 25344 + }, + { + "epoch": 1.08, + "grad_norm": 0.478515625, + "learning_rate": 8.776860669576495e-05, + "loss": 2.4919, + "step": 25345 + }, + { + "epoch": 1.08, + "grad_norm": 0.5234375, + "learning_rate": 8.776194915851658e-05, + "loss": 2.584, + "step": 25346 + }, + { + "epoch": 1.08, + "grad_norm": 0.5703125, + "learning_rate": 8.775529167633497e-05, + "loss": 2.5522, + "step": 25347 + }, + { + "epoch": 1.08, + "grad_norm": 0.427734375, + "learning_rate": 8.774863424925009e-05, + "loss": 2.61, + "step": 25348 + }, + { + "epoch": 1.08, + "grad_norm": 0.482421875, + "learning_rate": 8.774197687729187e-05, + "loss": 2.3594, + "step": 25349 + }, + { + "epoch": 1.08, + "grad_norm": 0.4609375, + "learning_rate": 8.773531956049028e-05, + "loss": 2.5267, + "step": 25350 + }, + { + "epoch": 1.08, + "grad_norm": 0.486328125, + "learning_rate": 8.772866229887525e-05, + "loss": 2.5649, + "step": 25351 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.772200509247677e-05, + "loss": 2.4597, + "step": 25352 + }, + { + "epoch": 1.08, + "grad_norm": 0.494140625, + "learning_rate": 8.77153479413248e-05, + "loss": 2.4511, + "step": 25353 + }, + { + "epoch": 1.08, + "grad_norm": 0.412109375, + "learning_rate": 8.770869084544926e-05, + "loss": 2.4604, + "step": 25354 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.770203380488009e-05, + "loss": 2.5792, + "step": 25355 + }, + { + "epoch": 1.08, + "grad_norm": 0.5390625, + "learning_rate": 8.76953768196473e-05, + "loss": 2.6576, + "step": 25356 + }, + { + "epoch": 1.08, + "grad_norm": 0.5, + "learning_rate": 8.768871988978081e-05, + "loss": 2.5613, + "step": 25357 + }, + { + "epoch": 1.08, + "grad_norm": 0.515625, + "learning_rate": 8.768206301531059e-05, + "loss": 2.5564, + "step": 25358 + }, + { + "epoch": 1.08, + "grad_norm": 0.46484375, + "learning_rate": 8.767540619626657e-05, + "loss": 2.4524, + "step": 25359 + }, + { + "epoch": 1.08, + "grad_norm": 0.51953125, + "learning_rate": 8.766874943267872e-05, + "loss": 2.5447, + "step": 25360 + }, + { + "epoch": 1.08, + "grad_norm": 0.5390625, + "learning_rate": 8.766209272457703e-05, + "loss": 2.4543, + "step": 25361 + }, + { + "epoch": 1.08, + "grad_norm": 0.443359375, + "learning_rate": 8.765543607199138e-05, + "loss": 2.3896, + "step": 25362 + }, + { + "epoch": 1.08, + "grad_norm": 0.46484375, + "learning_rate": 8.764877947495175e-05, + "loss": 2.3333, + "step": 25363 + }, + { + "epoch": 1.08, + "grad_norm": 0.451171875, + "learning_rate": 8.764212293348811e-05, + "loss": 2.6784, + "step": 25364 + }, + { + "epoch": 1.08, + "grad_norm": 0.458984375, + "learning_rate": 8.76354664476304e-05, + "loss": 2.4614, + "step": 25365 + }, + { + "epoch": 1.08, + "grad_norm": 0.419921875, + "learning_rate": 8.762881001740857e-05, + "loss": 2.3647, + "step": 25366 + }, + { + "epoch": 1.08, + "grad_norm": 0.4765625, + "learning_rate": 8.762215364285256e-05, + "loss": 2.5091, + "step": 25367 + }, + { + "epoch": 1.08, + "grad_norm": 0.431640625, + "learning_rate": 8.761549732399235e-05, + "loss": 2.4329, + "step": 25368 + }, + { + "epoch": 1.08, + "grad_norm": 0.48046875, + "learning_rate": 8.76088410608579e-05, + "loss": 2.4513, + "step": 25369 + }, + { + "epoch": 1.08, + "grad_norm": 0.435546875, + "learning_rate": 8.760218485347911e-05, + "loss": 2.4566, + "step": 25370 + }, + { + "epoch": 1.08, + "grad_norm": 0.43359375, + "learning_rate": 8.759552870188595e-05, + "loss": 2.5645, + "step": 25371 + }, + { + "epoch": 1.08, + "grad_norm": 0.44921875, + "learning_rate": 8.758887260610837e-05, + "loss": 2.4698, + "step": 25372 + }, + { + "epoch": 1.08, + "grad_norm": 0.44140625, + "learning_rate": 8.758221656617636e-05, + "loss": 2.4217, + "step": 25373 + }, + { + "epoch": 1.08, + "grad_norm": 0.470703125, + "learning_rate": 8.757556058211981e-05, + "loss": 2.4855, + "step": 25374 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.756890465396869e-05, + "loss": 2.4956, + "step": 25375 + }, + { + "epoch": 1.08, + "grad_norm": 0.486328125, + "learning_rate": 8.756224878175298e-05, + "loss": 2.4171, + "step": 25376 + }, + { + "epoch": 1.08, + "grad_norm": 0.431640625, + "learning_rate": 8.755559296550262e-05, + "loss": 2.396, + "step": 25377 + }, + { + "epoch": 1.08, + "grad_norm": 0.54296875, + "learning_rate": 8.754893720524754e-05, + "loss": 2.4456, + "step": 25378 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.754228150101766e-05, + "loss": 2.4429, + "step": 25379 + }, + { + "epoch": 1.08, + "grad_norm": 0.47265625, + "learning_rate": 8.753562585284297e-05, + "loss": 2.4027, + "step": 25380 + }, + { + "epoch": 1.08, + "grad_norm": 0.48046875, + "learning_rate": 8.752897026075342e-05, + "loss": 2.4677, + "step": 25381 + }, + { + "epoch": 1.08, + "grad_norm": 0.48828125, + "learning_rate": 8.752231472477895e-05, + "loss": 2.3442, + "step": 25382 + }, + { + "epoch": 1.08, + "grad_norm": 0.451171875, + "learning_rate": 8.751565924494946e-05, + "loss": 2.5006, + "step": 25383 + }, + { + "epoch": 1.08, + "grad_norm": 0.5, + "learning_rate": 8.750900382129498e-05, + "loss": 2.425, + "step": 25384 + }, + { + "epoch": 1.08, + "grad_norm": 0.4375, + "learning_rate": 8.750234845384544e-05, + "loss": 2.601, + "step": 25385 + }, + { + "epoch": 1.08, + "grad_norm": 0.5390625, + "learning_rate": 8.749569314263074e-05, + "loss": 2.4972, + "step": 25386 + }, + { + "epoch": 1.08, + "grad_norm": 0.455078125, + "learning_rate": 8.748903788768081e-05, + "loss": 2.65, + "step": 25387 + }, + { + "epoch": 1.08, + "grad_norm": 0.45703125, + "learning_rate": 8.74823826890257e-05, + "loss": 2.4667, + "step": 25388 + }, + { + "epoch": 1.08, + "grad_norm": 0.46875, + "learning_rate": 8.747572754669526e-05, + "loss": 2.3943, + "step": 25389 + }, + { + "epoch": 1.08, + "grad_norm": 0.44140625, + "learning_rate": 8.74690724607195e-05, + "loss": 2.3627, + "step": 25390 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.746241743112829e-05, + "loss": 2.5009, + "step": 25391 + }, + { + "epoch": 1.08, + "grad_norm": 0.455078125, + "learning_rate": 8.745576245795164e-05, + "loss": 2.4299, + "step": 25392 + }, + { + "epoch": 1.08, + "grad_norm": 0.490234375, + "learning_rate": 8.74491075412195e-05, + "loss": 2.4638, + "step": 25393 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.744245268096176e-05, + "loss": 2.42, + "step": 25394 + }, + { + "epoch": 1.08, + "grad_norm": 0.48828125, + "learning_rate": 8.743579787720837e-05, + "loss": 2.4853, + "step": 25395 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.742914312998934e-05, + "loss": 2.5666, + "step": 25396 + }, + { + "epoch": 1.08, + "grad_norm": 0.421875, + "learning_rate": 8.742248843933457e-05, + "loss": 2.4853, + "step": 25397 + }, + { + "epoch": 1.08, + "grad_norm": 0.42578125, + "learning_rate": 8.7415833805274e-05, + "loss": 2.4199, + "step": 25398 + }, + { + "epoch": 1.08, + "grad_norm": 0.482421875, + "learning_rate": 8.740917922783756e-05, + "loss": 2.4828, + "step": 25399 + }, + { + "epoch": 1.08, + "grad_norm": 0.42578125, + "learning_rate": 8.740252470705522e-05, + "loss": 2.3922, + "step": 25400 + }, + { + "epoch": 1.08, + "grad_norm": 0.52734375, + "learning_rate": 8.739587024295695e-05, + "loss": 2.5493, + "step": 25401 + }, + { + "epoch": 1.08, + "grad_norm": 0.515625, + "learning_rate": 8.738921583557263e-05, + "loss": 2.3776, + "step": 25402 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.738256148493218e-05, + "loss": 2.4141, + "step": 25403 + }, + { + "epoch": 1.08, + "grad_norm": 0.462890625, + "learning_rate": 8.737590719106566e-05, + "loss": 2.4477, + "step": 25404 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.736925295400293e-05, + "loss": 2.3962, + "step": 25405 + }, + { + "epoch": 1.08, + "grad_norm": 0.43359375, + "learning_rate": 8.736259877377395e-05, + "loss": 2.4835, + "step": 25406 + }, + { + "epoch": 1.08, + "grad_norm": 0.451171875, + "learning_rate": 8.735594465040863e-05, + "loss": 2.3404, + "step": 25407 + }, + { + "epoch": 1.08, + "grad_norm": 0.4765625, + "learning_rate": 8.734929058393695e-05, + "loss": 2.4429, + "step": 25408 + }, + { + "epoch": 1.08, + "grad_norm": 0.498046875, + "learning_rate": 8.734263657438887e-05, + "loss": 2.4448, + "step": 25409 + }, + { + "epoch": 1.08, + "grad_norm": 0.4921875, + "learning_rate": 8.733598262179428e-05, + "loss": 2.4832, + "step": 25410 + }, + { + "epoch": 1.08, + "grad_norm": 0.400390625, + "learning_rate": 8.732932872618309e-05, + "loss": 2.49, + "step": 25411 + }, + { + "epoch": 1.08, + "grad_norm": 0.46484375, + "learning_rate": 8.732267488758537e-05, + "loss": 2.3871, + "step": 25412 + }, + { + "epoch": 1.08, + "grad_norm": 0.51953125, + "learning_rate": 8.731602110603093e-05, + "loss": 2.57, + "step": 25413 + }, + { + "epoch": 1.08, + "grad_norm": 0.431640625, + "learning_rate": 8.730936738154978e-05, + "loss": 2.5729, + "step": 25414 + }, + { + "epoch": 1.08, + "grad_norm": 0.4296875, + "learning_rate": 8.730271371417181e-05, + "loss": 2.4326, + "step": 25415 + }, + { + "epoch": 1.08, + "grad_norm": 0.48046875, + "learning_rate": 8.729606010392702e-05, + "loss": 2.437, + "step": 25416 + }, + { + "epoch": 1.08, + "grad_norm": 0.5234375, + "learning_rate": 8.728940655084533e-05, + "loss": 2.6963, + "step": 25417 + }, + { + "epoch": 1.08, + "grad_norm": 0.494140625, + "learning_rate": 8.728275305495664e-05, + "loss": 2.4034, + "step": 25418 + }, + { + "epoch": 1.08, + "grad_norm": 0.494140625, + "learning_rate": 8.727609961629088e-05, + "loss": 2.4842, + "step": 25419 + }, + { + "epoch": 1.08, + "grad_norm": 0.447265625, + "learning_rate": 8.726944623487808e-05, + "loss": 2.3942, + "step": 25420 + }, + { + "epoch": 1.08, + "grad_norm": 0.51171875, + "learning_rate": 8.72627929107481e-05, + "loss": 2.5113, + "step": 25421 + }, + { + "epoch": 1.08, + "grad_norm": 0.51171875, + "learning_rate": 8.72561396439309e-05, + "loss": 2.4922, + "step": 25422 + }, + { + "epoch": 1.08, + "grad_norm": 0.478515625, + "learning_rate": 8.724948643445639e-05, + "loss": 2.4648, + "step": 25423 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.724283328235455e-05, + "loss": 2.5355, + "step": 25424 + }, + { + "epoch": 1.08, + "grad_norm": 0.453125, + "learning_rate": 8.723618018765533e-05, + "loss": 2.4549, + "step": 25425 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.72295271503886e-05, + "loss": 2.4271, + "step": 25426 + }, + { + "epoch": 1.08, + "grad_norm": 0.42578125, + "learning_rate": 8.72228741705843e-05, + "loss": 2.3574, + "step": 25427 + }, + { + "epoch": 1.08, + "grad_norm": 0.51953125, + "learning_rate": 8.721622124827244e-05, + "loss": 2.5304, + "step": 25428 + }, + { + "epoch": 1.08, + "grad_norm": 0.453125, + "learning_rate": 8.720956838348291e-05, + "loss": 2.4008, + "step": 25429 + }, + { + "epoch": 1.08, + "grad_norm": 0.5078125, + "learning_rate": 8.720291557624565e-05, + "loss": 2.4105, + "step": 25430 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.719626282659055e-05, + "loss": 2.4722, + "step": 25431 + }, + { + "epoch": 1.08, + "grad_norm": 0.51171875, + "learning_rate": 8.718961013454763e-05, + "loss": 2.4279, + "step": 25432 + }, + { + "epoch": 1.08, + "grad_norm": 0.462890625, + "learning_rate": 8.71829575001468e-05, + "loss": 2.4665, + "step": 25433 + }, + { + "epoch": 1.08, + "grad_norm": 0.494140625, + "learning_rate": 8.717630492341796e-05, + "loss": 2.5047, + "step": 25434 + }, + { + "epoch": 1.08, + "grad_norm": 0.4140625, + "learning_rate": 8.716965240439101e-05, + "loss": 2.3714, + "step": 25435 + }, + { + "epoch": 1.08, + "grad_norm": 0.44921875, + "learning_rate": 8.716299994309601e-05, + "loss": 2.3522, + "step": 25436 + }, + { + "epoch": 1.08, + "grad_norm": 0.439453125, + "learning_rate": 8.715634753956278e-05, + "loss": 2.4084, + "step": 25437 + }, + { + "epoch": 1.08, + "grad_norm": 0.45703125, + "learning_rate": 8.714969519382131e-05, + "loss": 2.5163, + "step": 25438 + }, + { + "epoch": 1.08, + "grad_norm": 0.484375, + "learning_rate": 8.714304290590148e-05, + "loss": 2.523, + "step": 25439 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.713639067583328e-05, + "loss": 2.4455, + "step": 25440 + }, + { + "epoch": 1.08, + "grad_norm": 0.4296875, + "learning_rate": 8.712973850364664e-05, + "loss": 2.4618, + "step": 25441 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.712308638937147e-05, + "loss": 2.4833, + "step": 25442 + }, + { + "epoch": 1.08, + "grad_norm": 0.48046875, + "learning_rate": 8.711643433303765e-05, + "loss": 2.4397, + "step": 25443 + }, + { + "epoch": 1.08, + "grad_norm": 0.455078125, + "learning_rate": 8.710978233467524e-05, + "loss": 2.5035, + "step": 25444 + }, + { + "epoch": 1.08, + "grad_norm": 0.62109375, + "learning_rate": 8.710313039431406e-05, + "loss": 2.4443, + "step": 25445 + }, + { + "epoch": 1.08, + "grad_norm": 0.41796875, + "learning_rate": 8.709647851198409e-05, + "loss": 2.5811, + "step": 25446 + }, + { + "epoch": 1.08, + "grad_norm": 0.470703125, + "learning_rate": 8.708982668771524e-05, + "loss": 2.3697, + "step": 25447 + }, + { + "epoch": 1.08, + "grad_norm": 0.4921875, + "learning_rate": 8.708317492153745e-05, + "loss": 2.3693, + "step": 25448 + }, + { + "epoch": 1.08, + "grad_norm": 0.486328125, + "learning_rate": 8.707652321348067e-05, + "loss": 2.3529, + "step": 25449 + }, + { + "epoch": 1.08, + "grad_norm": 0.4140625, + "learning_rate": 8.706987156357484e-05, + "loss": 2.4378, + "step": 25450 + }, + { + "epoch": 1.08, + "grad_norm": 0.431640625, + "learning_rate": 8.706321997184978e-05, + "loss": 2.3889, + "step": 25451 + }, + { + "epoch": 1.08, + "grad_norm": 0.458984375, + "learning_rate": 8.705656843833557e-05, + "loss": 2.4412, + "step": 25452 + }, + { + "epoch": 1.08, + "grad_norm": 0.470703125, + "learning_rate": 8.704991696306207e-05, + "loss": 2.3543, + "step": 25453 + }, + { + "epoch": 1.08, + "grad_norm": 0.48828125, + "learning_rate": 8.704326554605918e-05, + "loss": 2.4975, + "step": 25454 + }, + { + "epoch": 1.08, + "grad_norm": 0.466796875, + "learning_rate": 8.703661418735686e-05, + "loss": 2.3858, + "step": 25455 + }, + { + "epoch": 1.08, + "grad_norm": 0.515625, + "learning_rate": 8.702996288698505e-05, + "loss": 2.544, + "step": 25456 + }, + { + "epoch": 1.08, + "grad_norm": 0.47265625, + "learning_rate": 8.702331164497367e-05, + "loss": 2.4106, + "step": 25457 + }, + { + "epoch": 1.08, + "grad_norm": 0.50390625, + "learning_rate": 8.701666046135268e-05, + "loss": 2.5981, + "step": 25458 + }, + { + "epoch": 1.08, + "grad_norm": 0.474609375, + "learning_rate": 8.701000933615194e-05, + "loss": 2.4916, + "step": 25459 + }, + { + "epoch": 1.08, + "grad_norm": 0.498046875, + "learning_rate": 8.700335826940138e-05, + "loss": 2.4355, + "step": 25460 + }, + { + "epoch": 1.08, + "grad_norm": 0.46484375, + "learning_rate": 8.699670726113099e-05, + "loss": 2.4508, + "step": 25461 + }, + { + "epoch": 1.09, + "grad_norm": 0.48046875, + "learning_rate": 8.699005631137067e-05, + "loss": 2.3883, + "step": 25462 + }, + { + "epoch": 1.09, + "grad_norm": 0.44140625, + "learning_rate": 8.698340542015035e-05, + "loss": 2.4078, + "step": 25463 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.697675458749988e-05, + "loss": 2.3613, + "step": 25464 + }, + { + "epoch": 1.09, + "grad_norm": 0.462890625, + "learning_rate": 8.697010381344931e-05, + "loss": 2.3742, + "step": 25465 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.696345309802854e-05, + "loss": 2.5144, + "step": 25466 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.695680244126743e-05, + "loss": 2.5348, + "step": 25467 + }, + { + "epoch": 1.09, + "grad_norm": 0.4765625, + "learning_rate": 8.695015184319592e-05, + "loss": 2.5387, + "step": 25468 + }, + { + "epoch": 1.09, + "grad_norm": 0.4296875, + "learning_rate": 8.694350130384399e-05, + "loss": 2.4822, + "step": 25469 + }, + { + "epoch": 1.09, + "grad_norm": 0.44921875, + "learning_rate": 8.693685082324153e-05, + "loss": 2.4637, + "step": 25470 + }, + { + "epoch": 1.09, + "grad_norm": 0.439453125, + "learning_rate": 8.693020040141849e-05, + "loss": 2.4041, + "step": 25471 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.692355003840471e-05, + "loss": 2.5637, + "step": 25472 + }, + { + "epoch": 1.09, + "grad_norm": 0.51953125, + "learning_rate": 8.69168997342302e-05, + "loss": 2.5552, + "step": 25473 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.69102494889249e-05, + "loss": 2.4275, + "step": 25474 + }, + { + "epoch": 1.09, + "grad_norm": 0.46875, + "learning_rate": 8.690359930251867e-05, + "loss": 2.3925, + "step": 25475 + }, + { + "epoch": 1.09, + "grad_norm": 0.478515625, + "learning_rate": 8.689694917504144e-05, + "loss": 2.4953, + "step": 25476 + }, + { + "epoch": 1.09, + "grad_norm": 0.51171875, + "learning_rate": 8.689029910652315e-05, + "loss": 2.5092, + "step": 25477 + }, + { + "epoch": 1.09, + "grad_norm": 0.53125, + "learning_rate": 8.688364909699375e-05, + "loss": 2.411, + "step": 25478 + }, + { + "epoch": 1.09, + "grad_norm": 0.51953125, + "learning_rate": 8.687699914648315e-05, + "loss": 2.458, + "step": 25479 + }, + { + "epoch": 1.09, + "grad_norm": 0.44921875, + "learning_rate": 8.687034925502117e-05, + "loss": 2.436, + "step": 25480 + }, + { + "epoch": 1.09, + "grad_norm": 0.453125, + "learning_rate": 8.686369942263789e-05, + "loss": 2.4544, + "step": 25481 + }, + { + "epoch": 1.09, + "grad_norm": 0.427734375, + "learning_rate": 8.685704964936316e-05, + "loss": 2.4822, + "step": 25482 + }, + { + "epoch": 1.09, + "grad_norm": 0.453125, + "learning_rate": 8.68503999352269e-05, + "loss": 2.451, + "step": 25483 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.684375028025902e-05, + "loss": 2.5186, + "step": 25484 + }, + { + "epoch": 1.09, + "grad_norm": 0.53515625, + "learning_rate": 8.683710068448945e-05, + "loss": 2.4897, + "step": 25485 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.683045114794815e-05, + "loss": 2.5991, + "step": 25486 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.682380167066501e-05, + "loss": 2.5316, + "step": 25487 + }, + { + "epoch": 1.09, + "grad_norm": 0.419921875, + "learning_rate": 8.681715225266988e-05, + "loss": 2.538, + "step": 25488 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.68105028939928e-05, + "loss": 2.574, + "step": 25489 + }, + { + "epoch": 1.09, + "grad_norm": 0.46875, + "learning_rate": 8.680385359466364e-05, + "loss": 2.4108, + "step": 25490 + }, + { + "epoch": 1.09, + "grad_norm": 0.486328125, + "learning_rate": 8.67972043547123e-05, + "loss": 2.5501, + "step": 25491 + }, + { + "epoch": 1.09, + "grad_norm": 0.478515625, + "learning_rate": 8.679055517416871e-05, + "loss": 2.4219, + "step": 25492 + }, + { + "epoch": 1.09, + "grad_norm": 0.453125, + "learning_rate": 8.67839060530628e-05, + "loss": 2.5789, + "step": 25493 + }, + { + "epoch": 1.09, + "grad_norm": 0.478515625, + "learning_rate": 8.677725699142449e-05, + "loss": 2.3556, + "step": 25494 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.677060798928369e-05, + "loss": 2.3906, + "step": 25495 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.67639590466703e-05, + "loss": 2.6315, + "step": 25496 + }, + { + "epoch": 1.09, + "grad_norm": 0.46875, + "learning_rate": 8.675731016361428e-05, + "loss": 2.3545, + "step": 25497 + }, + { + "epoch": 1.09, + "grad_norm": 0.4453125, + "learning_rate": 8.675066134014555e-05, + "loss": 2.4636, + "step": 25498 + }, + { + "epoch": 1.09, + "grad_norm": 0.53515625, + "learning_rate": 8.674401257629396e-05, + "loss": 2.6722, + "step": 25499 + }, + { + "epoch": 1.09, + "grad_norm": 0.494140625, + "learning_rate": 8.673736387208946e-05, + "loss": 2.5371, + "step": 25500 + }, + { + "epoch": 1.09, + "grad_norm": 0.51171875, + "learning_rate": 8.6730715227562e-05, + "loss": 2.514, + "step": 25501 + }, + { + "epoch": 1.09, + "grad_norm": 0.44140625, + "learning_rate": 8.672406664274147e-05, + "loss": 2.4986, + "step": 25502 + }, + { + "epoch": 1.09, + "grad_norm": 0.435546875, + "learning_rate": 8.67174181176578e-05, + "loss": 2.3724, + "step": 25503 + }, + { + "epoch": 1.09, + "grad_norm": 0.451171875, + "learning_rate": 8.671076965234086e-05, + "loss": 2.5835, + "step": 25504 + }, + { + "epoch": 1.09, + "grad_norm": 0.482421875, + "learning_rate": 8.670412124682061e-05, + "loss": 2.3554, + "step": 25505 + }, + { + "epoch": 1.09, + "grad_norm": 0.47265625, + "learning_rate": 8.6697472901127e-05, + "loss": 2.3596, + "step": 25506 + }, + { + "epoch": 1.09, + "grad_norm": 0.439453125, + "learning_rate": 8.669082461528985e-05, + "loss": 2.4998, + "step": 25507 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.668417638933913e-05, + "loss": 2.3953, + "step": 25508 + }, + { + "epoch": 1.09, + "grad_norm": 0.4921875, + "learning_rate": 8.667752822330476e-05, + "loss": 2.5725, + "step": 25509 + }, + { + "epoch": 1.09, + "grad_norm": 0.48828125, + "learning_rate": 8.667088011721665e-05, + "loss": 2.4542, + "step": 25510 + }, + { + "epoch": 1.09, + "grad_norm": 0.59375, + "learning_rate": 8.66642320711047e-05, + "loss": 2.5249, + "step": 25511 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.665758408499883e-05, + "loss": 2.5875, + "step": 25512 + }, + { + "epoch": 1.09, + "grad_norm": 0.484375, + "learning_rate": 8.665093615892895e-05, + "loss": 2.5552, + "step": 25513 + }, + { + "epoch": 1.09, + "grad_norm": 0.5703125, + "learning_rate": 8.6644288292925e-05, + "loss": 2.4277, + "step": 25514 + }, + { + "epoch": 1.09, + "grad_norm": 0.49609375, + "learning_rate": 8.663764048701687e-05, + "loss": 2.3962, + "step": 25515 + }, + { + "epoch": 1.09, + "grad_norm": 0.53515625, + "learning_rate": 8.663099274123445e-05, + "loss": 2.5539, + "step": 25516 + }, + { + "epoch": 1.09, + "grad_norm": 0.498046875, + "learning_rate": 8.66243450556077e-05, + "loss": 2.5079, + "step": 25517 + }, + { + "epoch": 1.09, + "grad_norm": 0.462890625, + "learning_rate": 8.66176974301665e-05, + "loss": 2.3969, + "step": 25518 + }, + { + "epoch": 1.09, + "grad_norm": 0.486328125, + "learning_rate": 8.661104986494078e-05, + "loss": 2.5373, + "step": 25519 + }, + { + "epoch": 1.09, + "grad_norm": 0.451171875, + "learning_rate": 8.660440235996041e-05, + "loss": 2.4546, + "step": 25520 + }, + { + "epoch": 1.09, + "grad_norm": 0.44140625, + "learning_rate": 8.659775491525537e-05, + "loss": 2.5051, + "step": 25521 + }, + { + "epoch": 1.09, + "grad_norm": 0.42578125, + "learning_rate": 8.659110753085554e-05, + "loss": 2.3898, + "step": 25522 + }, + { + "epoch": 1.09, + "grad_norm": 0.447265625, + "learning_rate": 8.658446020679082e-05, + "loss": 2.5603, + "step": 25523 + }, + { + "epoch": 1.09, + "grad_norm": 0.5234375, + "learning_rate": 8.657781294309108e-05, + "loss": 2.6587, + "step": 25524 + }, + { + "epoch": 1.09, + "grad_norm": 0.48046875, + "learning_rate": 8.657116573978632e-05, + "loss": 2.4495, + "step": 25525 + }, + { + "epoch": 1.09, + "grad_norm": 0.45703125, + "learning_rate": 8.65645185969064e-05, + "loss": 2.5139, + "step": 25526 + }, + { + "epoch": 1.09, + "grad_norm": 0.4609375, + "learning_rate": 8.655787151448125e-05, + "loss": 2.4217, + "step": 25527 + }, + { + "epoch": 1.09, + "grad_norm": 0.470703125, + "learning_rate": 8.655122449254073e-05, + "loss": 2.3423, + "step": 25528 + }, + { + "epoch": 1.09, + "grad_norm": 0.466796875, + "learning_rate": 8.65445775311148e-05, + "loss": 2.3938, + "step": 25529 + }, + { + "epoch": 1.09, + "grad_norm": 0.53515625, + "learning_rate": 8.653793063023339e-05, + "loss": 2.4906, + "step": 25530 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.653128378992635e-05, + "loss": 2.4576, + "step": 25531 + }, + { + "epoch": 1.09, + "grad_norm": 0.431640625, + "learning_rate": 8.652463701022357e-05, + "loss": 2.4891, + "step": 25532 + }, + { + "epoch": 1.09, + "grad_norm": 0.45703125, + "learning_rate": 8.651799029115504e-05, + "loss": 2.4448, + "step": 25533 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.651134363275062e-05, + "loss": 2.5174, + "step": 25534 + }, + { + "epoch": 1.09, + "grad_norm": 0.439453125, + "learning_rate": 8.650469703504021e-05, + "loss": 2.3632, + "step": 25535 + }, + { + "epoch": 1.09, + "grad_norm": 0.439453125, + "learning_rate": 8.649805049805373e-05, + "loss": 2.3829, + "step": 25536 + }, + { + "epoch": 1.09, + "grad_norm": 0.54296875, + "learning_rate": 8.64914040218211e-05, + "loss": 2.4178, + "step": 25537 + }, + { + "epoch": 1.09, + "grad_norm": 0.5, + "learning_rate": 8.648475760637224e-05, + "loss": 2.5448, + "step": 25538 + }, + { + "epoch": 1.09, + "grad_norm": 0.404296875, + "learning_rate": 8.647811125173701e-05, + "loss": 2.4582, + "step": 25539 + }, + { + "epoch": 1.09, + "grad_norm": 0.44140625, + "learning_rate": 8.647146495794532e-05, + "loss": 2.4417, + "step": 25540 + }, + { + "epoch": 1.09, + "grad_norm": 0.46484375, + "learning_rate": 8.646481872502711e-05, + "loss": 2.4403, + "step": 25541 + }, + { + "epoch": 1.09, + "grad_norm": 0.43359375, + "learning_rate": 8.645817255301228e-05, + "loss": 2.472, + "step": 25542 + }, + { + "epoch": 1.09, + "grad_norm": 0.4921875, + "learning_rate": 8.645152644193071e-05, + "loss": 2.5709, + "step": 25543 + }, + { + "epoch": 1.09, + "grad_norm": 0.494140625, + "learning_rate": 8.64448803918123e-05, + "loss": 2.5612, + "step": 25544 + }, + { + "epoch": 1.09, + "grad_norm": 0.451171875, + "learning_rate": 8.6438234402687e-05, + "loss": 2.3678, + "step": 25545 + }, + { + "epoch": 1.09, + "grad_norm": 0.44921875, + "learning_rate": 8.643158847458471e-05, + "loss": 2.391, + "step": 25546 + }, + { + "epoch": 1.09, + "grad_norm": 0.51171875, + "learning_rate": 8.642494260753529e-05, + "loss": 2.564, + "step": 25547 + }, + { + "epoch": 1.09, + "grad_norm": 0.447265625, + "learning_rate": 8.641829680156864e-05, + "loss": 2.4874, + "step": 25548 + }, + { + "epoch": 1.09, + "grad_norm": 0.54296875, + "learning_rate": 8.641165105671472e-05, + "loss": 2.5576, + "step": 25549 + }, + { + "epoch": 1.09, + "grad_norm": 0.455078125, + "learning_rate": 8.640500537300342e-05, + "loss": 2.4301, + "step": 25550 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.63983597504646e-05, + "loss": 2.4947, + "step": 25551 + }, + { + "epoch": 1.09, + "grad_norm": 0.462890625, + "learning_rate": 8.639171418912819e-05, + "loss": 2.4419, + "step": 25552 + }, + { + "epoch": 1.09, + "grad_norm": 0.427734375, + "learning_rate": 8.63850686890241e-05, + "loss": 2.45, + "step": 25553 + }, + { + "epoch": 1.09, + "grad_norm": 0.484375, + "learning_rate": 8.637842325018224e-05, + "loss": 2.4059, + "step": 25554 + }, + { + "epoch": 1.09, + "grad_norm": 0.416015625, + "learning_rate": 8.637177787263248e-05, + "loss": 2.5491, + "step": 25555 + }, + { + "epoch": 1.09, + "grad_norm": 0.494140625, + "learning_rate": 8.636513255640474e-05, + "loss": 2.526, + "step": 25556 + }, + { + "epoch": 1.09, + "grad_norm": 0.474609375, + "learning_rate": 8.635848730152892e-05, + "loss": 2.5318, + "step": 25557 + }, + { + "epoch": 1.09, + "grad_norm": 0.447265625, + "learning_rate": 8.635184210803493e-05, + "loss": 2.5112, + "step": 25558 + }, + { + "epoch": 1.09, + "grad_norm": 0.451171875, + "learning_rate": 8.634519697595265e-05, + "loss": 2.4081, + "step": 25559 + }, + { + "epoch": 1.09, + "grad_norm": 0.455078125, + "learning_rate": 8.633855190531199e-05, + "loss": 2.4903, + "step": 25560 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.633190689614287e-05, + "loss": 2.548, + "step": 25561 + }, + { + "epoch": 1.09, + "grad_norm": 0.439453125, + "learning_rate": 8.632526194847518e-05, + "loss": 2.4957, + "step": 25562 + }, + { + "epoch": 1.09, + "grad_norm": 0.54296875, + "learning_rate": 8.631861706233882e-05, + "loss": 2.5649, + "step": 25563 + }, + { + "epoch": 1.09, + "grad_norm": 0.46484375, + "learning_rate": 8.631197223776364e-05, + "loss": 2.4587, + "step": 25564 + }, + { + "epoch": 1.09, + "grad_norm": 0.443359375, + "learning_rate": 8.630532747477961e-05, + "loss": 2.4847, + "step": 25565 + }, + { + "epoch": 1.09, + "grad_norm": 0.443359375, + "learning_rate": 8.62986827734166e-05, + "loss": 2.4172, + "step": 25566 + }, + { + "epoch": 1.09, + "grad_norm": 0.455078125, + "learning_rate": 8.629203813370452e-05, + "loss": 2.4275, + "step": 25567 + }, + { + "epoch": 1.09, + "grad_norm": 0.484375, + "learning_rate": 8.628539355567322e-05, + "loss": 2.4825, + "step": 25568 + }, + { + "epoch": 1.09, + "grad_norm": 0.4921875, + "learning_rate": 8.627874903935268e-05, + "loss": 2.608, + "step": 25569 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.627210458477276e-05, + "loss": 2.4818, + "step": 25570 + }, + { + "epoch": 1.09, + "grad_norm": 0.447265625, + "learning_rate": 8.626546019196333e-05, + "loss": 2.5372, + "step": 25571 + }, + { + "epoch": 1.09, + "grad_norm": 0.5, + "learning_rate": 8.62588158609543e-05, + "loss": 2.4799, + "step": 25572 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.625217159177559e-05, + "loss": 2.5926, + "step": 25573 + }, + { + "epoch": 1.09, + "grad_norm": 0.435546875, + "learning_rate": 8.624552738445708e-05, + "loss": 2.4651, + "step": 25574 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.623888323902869e-05, + "loss": 2.4821, + "step": 25575 + }, + { + "epoch": 1.09, + "grad_norm": 0.44921875, + "learning_rate": 8.623223915552025e-05, + "loss": 2.4567, + "step": 25576 + }, + { + "epoch": 1.09, + "grad_norm": 0.451171875, + "learning_rate": 8.622559513396173e-05, + "loss": 2.4202, + "step": 25577 + }, + { + "epoch": 1.09, + "grad_norm": 0.486328125, + "learning_rate": 8.621895117438304e-05, + "loss": 2.379, + "step": 25578 + }, + { + "epoch": 1.09, + "grad_norm": 0.455078125, + "learning_rate": 8.621230727681399e-05, + "loss": 2.3861, + "step": 25579 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.620566344128453e-05, + "loss": 2.5544, + "step": 25580 + }, + { + "epoch": 1.09, + "grad_norm": 0.431640625, + "learning_rate": 8.61990196678245e-05, + "loss": 2.3503, + "step": 25581 + }, + { + "epoch": 1.09, + "grad_norm": 0.4765625, + "learning_rate": 8.619237595646388e-05, + "loss": 2.4742, + "step": 25582 + }, + { + "epoch": 1.09, + "grad_norm": 0.515625, + "learning_rate": 8.618573230723252e-05, + "loss": 2.5332, + "step": 25583 + }, + { + "epoch": 1.09, + "grad_norm": 0.482421875, + "learning_rate": 8.617908872016034e-05, + "loss": 2.4617, + "step": 25584 + }, + { + "epoch": 1.09, + "grad_norm": 0.52734375, + "learning_rate": 8.617244519527713e-05, + "loss": 2.4198, + "step": 25585 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.616580173261294e-05, + "loss": 2.5083, + "step": 25586 + }, + { + "epoch": 1.09, + "grad_norm": 0.421875, + "learning_rate": 8.615915833219756e-05, + "loss": 2.5874, + "step": 25587 + }, + { + "epoch": 1.09, + "grad_norm": 0.4921875, + "learning_rate": 8.615251499406092e-05, + "loss": 2.5249, + "step": 25588 + }, + { + "epoch": 1.09, + "grad_norm": 0.482421875, + "learning_rate": 8.614587171823286e-05, + "loss": 2.4435, + "step": 25589 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.613922850474335e-05, + "loss": 2.4524, + "step": 25590 + }, + { + "epoch": 1.09, + "grad_norm": 0.478515625, + "learning_rate": 8.613258535362226e-05, + "loss": 2.5081, + "step": 25591 + }, + { + "epoch": 1.09, + "grad_norm": 0.482421875, + "learning_rate": 8.612594226489947e-05, + "loss": 2.5059, + "step": 25592 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.61192992386048e-05, + "loss": 2.4443, + "step": 25593 + }, + { + "epoch": 1.09, + "grad_norm": 0.53125, + "learning_rate": 8.611265627476827e-05, + "loss": 2.5119, + "step": 25594 + }, + { + "epoch": 1.09, + "grad_norm": 0.46875, + "learning_rate": 8.610601337341971e-05, + "loss": 2.4725, + "step": 25595 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.6099370534589e-05, + "loss": 2.5052, + "step": 25596 + }, + { + "epoch": 1.09, + "grad_norm": 0.439453125, + "learning_rate": 8.609272775830604e-05, + "loss": 2.4003, + "step": 25597 + }, + { + "epoch": 1.09, + "grad_norm": 0.435546875, + "learning_rate": 8.608608504460073e-05, + "loss": 2.4152, + "step": 25598 + }, + { + "epoch": 1.09, + "grad_norm": 0.458984375, + "learning_rate": 8.607944239350295e-05, + "loss": 2.4841, + "step": 25599 + }, + { + "epoch": 1.09, + "grad_norm": 0.5, + "learning_rate": 8.607279980504261e-05, + "loss": 2.4407, + "step": 25600 + }, + { + "epoch": 1.09, + "grad_norm": 0.484375, + "learning_rate": 8.606615727924953e-05, + "loss": 2.4445, + "step": 25601 + }, + { + "epoch": 1.09, + "grad_norm": 0.45703125, + "learning_rate": 8.605951481615371e-05, + "loss": 2.4754, + "step": 25602 + }, + { + "epoch": 1.09, + "grad_norm": 0.52734375, + "learning_rate": 8.605287241578495e-05, + "loss": 2.5208, + "step": 25603 + }, + { + "epoch": 1.09, + "grad_norm": 0.53515625, + "learning_rate": 8.604623007817317e-05, + "loss": 2.5131, + "step": 25604 + }, + { + "epoch": 1.09, + "grad_norm": 0.54296875, + "learning_rate": 8.603958780334824e-05, + "loss": 2.4443, + "step": 25605 + }, + { + "epoch": 1.09, + "grad_norm": 0.466796875, + "learning_rate": 8.603294559134007e-05, + "loss": 2.5221, + "step": 25606 + }, + { + "epoch": 1.09, + "grad_norm": 0.443359375, + "learning_rate": 8.602630344217856e-05, + "loss": 2.4127, + "step": 25607 + }, + { + "epoch": 1.09, + "grad_norm": 0.4765625, + "learning_rate": 8.601966135589358e-05, + "loss": 2.4421, + "step": 25608 + }, + { + "epoch": 1.09, + "grad_norm": 0.478515625, + "learning_rate": 8.601301933251496e-05, + "loss": 2.4591, + "step": 25609 + }, + { + "epoch": 1.09, + "grad_norm": 0.490234375, + "learning_rate": 8.600637737207271e-05, + "loss": 2.4619, + "step": 25610 + }, + { + "epoch": 1.09, + "grad_norm": 0.51171875, + "learning_rate": 8.599973547459663e-05, + "loss": 2.4619, + "step": 25611 + }, + { + "epoch": 1.09, + "grad_norm": 0.45703125, + "learning_rate": 8.59930936401166e-05, + "loss": 2.5114, + "step": 25612 + }, + { + "epoch": 1.09, + "grad_norm": 0.47265625, + "learning_rate": 8.598645186866254e-05, + "loss": 2.4589, + "step": 25613 + }, + { + "epoch": 1.09, + "grad_norm": 0.4921875, + "learning_rate": 8.597981016026433e-05, + "loss": 2.3046, + "step": 25614 + }, + { + "epoch": 1.09, + "grad_norm": 0.421875, + "learning_rate": 8.597316851495186e-05, + "loss": 2.4894, + "step": 25615 + }, + { + "epoch": 1.09, + "grad_norm": 0.45703125, + "learning_rate": 8.5966526932755e-05, + "loss": 2.5618, + "step": 25616 + }, + { + "epoch": 1.09, + "grad_norm": 0.46875, + "learning_rate": 8.59598854137036e-05, + "loss": 2.384, + "step": 25617 + }, + { + "epoch": 1.09, + "grad_norm": 0.4609375, + "learning_rate": 8.595324395782765e-05, + "loss": 2.4663, + "step": 25618 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.594660256515693e-05, + "loss": 2.4653, + "step": 25619 + }, + { + "epoch": 1.09, + "grad_norm": 0.5, + "learning_rate": 8.593996123572137e-05, + "loss": 2.5114, + "step": 25620 + }, + { + "epoch": 1.09, + "grad_norm": 0.4921875, + "learning_rate": 8.593331996955083e-05, + "loss": 2.4065, + "step": 25621 + }, + { + "epoch": 1.09, + "grad_norm": 0.48046875, + "learning_rate": 8.592667876667523e-05, + "loss": 2.3237, + "step": 25622 + }, + { + "epoch": 1.09, + "grad_norm": 0.455078125, + "learning_rate": 8.592003762712443e-05, + "loss": 2.6032, + "step": 25623 + }, + { + "epoch": 1.09, + "grad_norm": 0.482421875, + "learning_rate": 8.591339655092834e-05, + "loss": 2.4353, + "step": 25624 + }, + { + "epoch": 1.09, + "grad_norm": 0.43359375, + "learning_rate": 8.590675553811674e-05, + "loss": 2.2792, + "step": 25625 + }, + { + "epoch": 1.09, + "grad_norm": 0.484375, + "learning_rate": 8.590011458871967e-05, + "loss": 2.4997, + "step": 25626 + }, + { + "epoch": 1.09, + "grad_norm": 0.50390625, + "learning_rate": 8.58934737027669e-05, + "loss": 2.4547, + "step": 25627 + }, + { + "epoch": 1.09, + "grad_norm": 0.474609375, + "learning_rate": 8.588683288028835e-05, + "loss": 2.4725, + "step": 25628 + }, + { + "epoch": 1.09, + "grad_norm": 0.51171875, + "learning_rate": 8.588019212131386e-05, + "loss": 2.5434, + "step": 25629 + }, + { + "epoch": 1.09, + "grad_norm": 0.466796875, + "learning_rate": 8.587355142587338e-05, + "loss": 2.4717, + "step": 25630 + }, + { + "epoch": 1.09, + "grad_norm": 0.49609375, + "learning_rate": 8.586691079399676e-05, + "loss": 2.3596, + "step": 25631 + }, + { + "epoch": 1.09, + "grad_norm": 0.49609375, + "learning_rate": 8.586027022571388e-05, + "loss": 2.4469, + "step": 25632 + }, + { + "epoch": 1.09, + "grad_norm": 0.46875, + "learning_rate": 8.585362972105456e-05, + "loss": 2.5325, + "step": 25633 + }, + { + "epoch": 1.09, + "grad_norm": 0.408203125, + "learning_rate": 8.584698928004881e-05, + "loss": 2.4203, + "step": 25634 + }, + { + "epoch": 1.09, + "grad_norm": 0.46484375, + "learning_rate": 8.58403489027264e-05, + "loss": 2.4408, + "step": 25635 + }, + { + "epoch": 1.09, + "grad_norm": 0.455078125, + "learning_rate": 8.583370858911726e-05, + "loss": 2.4133, + "step": 25636 + }, + { + "epoch": 1.09, + "grad_norm": 0.474609375, + "learning_rate": 8.582706833925123e-05, + "loss": 2.5312, + "step": 25637 + }, + { + "epoch": 1.09, + "grad_norm": 0.48046875, + "learning_rate": 8.582042815315823e-05, + "loss": 2.4454, + "step": 25638 + }, + { + "epoch": 1.09, + "grad_norm": 0.462890625, + "learning_rate": 8.581378803086811e-05, + "loss": 2.4171, + "step": 25639 + }, + { + "epoch": 1.09, + "grad_norm": 0.59375, + "learning_rate": 8.58071479724108e-05, + "loss": 2.5326, + "step": 25640 + }, + { + "epoch": 1.09, + "grad_norm": 0.470703125, + "learning_rate": 8.580050797781606e-05, + "loss": 2.5903, + "step": 25641 + }, + { + "epoch": 1.09, + "grad_norm": 0.474609375, + "learning_rate": 8.579386804711391e-05, + "loss": 2.5391, + "step": 25642 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.578722818033415e-05, + "loss": 2.5382, + "step": 25643 + }, + { + "epoch": 1.09, + "grad_norm": 0.443359375, + "learning_rate": 8.578058837750666e-05, + "loss": 2.4695, + "step": 25644 + }, + { + "epoch": 1.09, + "grad_norm": 0.46875, + "learning_rate": 8.577394863866131e-05, + "loss": 2.5454, + "step": 25645 + }, + { + "epoch": 1.09, + "grad_norm": 0.56640625, + "learning_rate": 8.576730896382801e-05, + "loss": 2.6554, + "step": 25646 + }, + { + "epoch": 1.09, + "grad_norm": 0.5, + "learning_rate": 8.576066935303661e-05, + "loss": 2.4686, + "step": 25647 + }, + { + "epoch": 1.09, + "grad_norm": 0.51953125, + "learning_rate": 8.575402980631703e-05, + "loss": 2.3934, + "step": 25648 + }, + { + "epoch": 1.09, + "grad_norm": 0.44921875, + "learning_rate": 8.574739032369904e-05, + "loss": 2.4524, + "step": 25649 + }, + { + "epoch": 1.09, + "grad_norm": 0.4609375, + "learning_rate": 8.574075090521265e-05, + "loss": 2.5234, + "step": 25650 + }, + { + "epoch": 1.09, + "grad_norm": 0.451171875, + "learning_rate": 8.573411155088764e-05, + "loss": 2.5018, + "step": 25651 + }, + { + "epoch": 1.09, + "grad_norm": 0.474609375, + "learning_rate": 8.572747226075393e-05, + "loss": 2.4943, + "step": 25652 + }, + { + "epoch": 1.09, + "grad_norm": 0.4375, + "learning_rate": 8.572083303484136e-05, + "loss": 2.3668, + "step": 25653 + }, + { + "epoch": 1.09, + "grad_norm": 0.52734375, + "learning_rate": 8.571419387317983e-05, + "loss": 2.5349, + "step": 25654 + }, + { + "epoch": 1.09, + "grad_norm": 0.48828125, + "learning_rate": 8.57075547757992e-05, + "loss": 2.5411, + "step": 25655 + }, + { + "epoch": 1.09, + "grad_norm": 0.53125, + "learning_rate": 8.570091574272938e-05, + "loss": 2.4934, + "step": 25656 + }, + { + "epoch": 1.09, + "grad_norm": 0.416015625, + "learning_rate": 8.569427677400015e-05, + "loss": 2.3956, + "step": 25657 + }, + { + "epoch": 1.09, + "grad_norm": 0.4765625, + "learning_rate": 8.568763786964151e-05, + "loss": 2.4061, + "step": 25658 + }, + { + "epoch": 1.09, + "grad_norm": 0.41796875, + "learning_rate": 8.568099902968325e-05, + "loss": 2.4062, + "step": 25659 + }, + { + "epoch": 1.09, + "grad_norm": 0.52734375, + "learning_rate": 8.567436025415526e-05, + "loss": 2.5051, + "step": 25660 + }, + { + "epoch": 1.09, + "grad_norm": 0.51953125, + "learning_rate": 8.56677215430874e-05, + "loss": 2.5283, + "step": 25661 + }, + { + "epoch": 1.09, + "grad_norm": 0.48828125, + "learning_rate": 8.566108289650955e-05, + "loss": 2.5653, + "step": 25662 + }, + { + "epoch": 1.09, + "grad_norm": 0.51171875, + "learning_rate": 8.565444431445162e-05, + "loss": 2.367, + "step": 25663 + }, + { + "epoch": 1.09, + "grad_norm": 0.4375, + "learning_rate": 8.564780579694345e-05, + "loss": 2.4617, + "step": 25664 + }, + { + "epoch": 1.09, + "grad_norm": 0.4453125, + "learning_rate": 8.564116734401485e-05, + "loss": 2.6158, + "step": 25665 + }, + { + "epoch": 1.09, + "grad_norm": 0.482421875, + "learning_rate": 8.563452895569581e-05, + "loss": 2.4654, + "step": 25666 + }, + { + "epoch": 1.09, + "grad_norm": 0.484375, + "learning_rate": 8.562789063201613e-05, + "loss": 2.5651, + "step": 25667 + }, + { + "epoch": 1.09, + "grad_norm": 0.43359375, + "learning_rate": 8.562125237300568e-05, + "loss": 2.421, + "step": 25668 + }, + { + "epoch": 1.09, + "grad_norm": 0.47265625, + "learning_rate": 8.561461417869432e-05, + "loss": 2.488, + "step": 25669 + }, + { + "epoch": 1.09, + "grad_norm": 0.462890625, + "learning_rate": 8.560797604911196e-05, + "loss": 2.5497, + "step": 25670 + }, + { + "epoch": 1.09, + "grad_norm": 0.498046875, + "learning_rate": 8.560133798428845e-05, + "loss": 2.5197, + "step": 25671 + }, + { + "epoch": 1.09, + "grad_norm": 0.44140625, + "learning_rate": 8.559469998425367e-05, + "loss": 2.385, + "step": 25672 + }, + { + "epoch": 1.09, + "grad_norm": 0.40234375, + "learning_rate": 8.558806204903742e-05, + "loss": 2.3462, + "step": 25673 + }, + { + "epoch": 1.09, + "grad_norm": 0.53515625, + "learning_rate": 8.55814241786697e-05, + "loss": 2.4739, + "step": 25674 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.557478637318025e-05, + "loss": 2.4522, + "step": 25675 + }, + { + "epoch": 1.09, + "grad_norm": 0.4765625, + "learning_rate": 8.5568148632599e-05, + "loss": 2.3609, + "step": 25676 + }, + { + "epoch": 1.09, + "grad_norm": 0.5078125, + "learning_rate": 8.55615109569558e-05, + "loss": 2.6114, + "step": 25677 + }, + { + "epoch": 1.09, + "grad_norm": 0.474609375, + "learning_rate": 8.555487334628054e-05, + "loss": 2.496, + "step": 25678 + }, + { + "epoch": 1.09, + "grad_norm": 0.57421875, + "learning_rate": 8.554823580060307e-05, + "loss": 2.5434, + "step": 25679 + }, + { + "epoch": 1.09, + "grad_norm": 0.443359375, + "learning_rate": 8.554159831995327e-05, + "loss": 2.4093, + "step": 25680 + }, + { + "epoch": 1.09, + "grad_norm": 0.478515625, + "learning_rate": 8.553496090436093e-05, + "loss": 2.3458, + "step": 25681 + }, + { + "epoch": 1.09, + "grad_norm": 0.49609375, + "learning_rate": 8.552832355385606e-05, + "loss": 2.4602, + "step": 25682 + }, + { + "epoch": 1.09, + "grad_norm": 0.396484375, + "learning_rate": 8.552168626846841e-05, + "loss": 2.3931, + "step": 25683 + }, + { + "epoch": 1.09, + "grad_norm": 0.494140625, + "learning_rate": 8.551504904822789e-05, + "loss": 2.3627, + "step": 25684 + }, + { + "epoch": 1.09, + "grad_norm": 0.48828125, + "learning_rate": 8.550841189316432e-05, + "loss": 2.5669, + "step": 25685 + }, + { + "epoch": 1.09, + "grad_norm": 0.482421875, + "learning_rate": 8.550177480330763e-05, + "loss": 2.528, + "step": 25686 + }, + { + "epoch": 1.09, + "grad_norm": 0.54296875, + "learning_rate": 8.549513777868765e-05, + "loss": 2.4662, + "step": 25687 + }, + { + "epoch": 1.09, + "grad_norm": 0.55078125, + "learning_rate": 8.548850081933428e-05, + "loss": 2.5645, + "step": 25688 + }, + { + "epoch": 1.09, + "grad_norm": 0.48046875, + "learning_rate": 8.548186392527728e-05, + "loss": 2.5635, + "step": 25689 + }, + { + "epoch": 1.09, + "grad_norm": 0.52734375, + "learning_rate": 8.547522709654666e-05, + "loss": 2.5732, + "step": 25690 + }, + { + "epoch": 1.09, + "grad_norm": 0.412109375, + "learning_rate": 8.546859033317219e-05, + "loss": 2.5413, + "step": 25691 + }, + { + "epoch": 1.09, + "grad_norm": 0.546875, + "learning_rate": 8.546195363518375e-05, + "loss": 2.527, + "step": 25692 + }, + { + "epoch": 1.09, + "grad_norm": 0.484375, + "learning_rate": 8.545531700261117e-05, + "loss": 2.3989, + "step": 25693 + }, + { + "epoch": 1.09, + "grad_norm": 0.46875, + "learning_rate": 8.544868043548438e-05, + "loss": 2.5748, + "step": 25694 + }, + { + "epoch": 1.09, + "grad_norm": 0.484375, + "learning_rate": 8.54420439338332e-05, + "loss": 2.5377, + "step": 25695 + }, + { + "epoch": 1.09, + "grad_norm": 0.486328125, + "learning_rate": 8.543540749768754e-05, + "loss": 2.457, + "step": 25696 + }, + { + "epoch": 1.1, + "grad_norm": 0.451171875, + "learning_rate": 8.542877112707716e-05, + "loss": 2.4627, + "step": 25697 + }, + { + "epoch": 1.1, + "grad_norm": 0.443359375, + "learning_rate": 8.542213482203203e-05, + "loss": 2.4172, + "step": 25698 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.541549858258197e-05, + "loss": 2.386, + "step": 25699 + }, + { + "epoch": 1.1, + "grad_norm": 0.44140625, + "learning_rate": 8.540886240875681e-05, + "loss": 2.3494, + "step": 25700 + }, + { + "epoch": 1.1, + "grad_norm": 0.46875, + "learning_rate": 8.540222630058646e-05, + "loss": 2.5574, + "step": 25701 + }, + { + "epoch": 1.1, + "grad_norm": 0.427734375, + "learning_rate": 8.539559025810072e-05, + "loss": 2.488, + "step": 25702 + }, + { + "epoch": 1.1, + "grad_norm": 0.54296875, + "learning_rate": 8.538895428132951e-05, + "loss": 2.5664, + "step": 25703 + }, + { + "epoch": 1.1, + "grad_norm": 0.451171875, + "learning_rate": 8.538231837030269e-05, + "loss": 2.4907, + "step": 25704 + }, + { + "epoch": 1.1, + "grad_norm": 0.490234375, + "learning_rate": 8.537568252505007e-05, + "loss": 2.5565, + "step": 25705 + }, + { + "epoch": 1.1, + "grad_norm": 0.48828125, + "learning_rate": 8.53690467456015e-05, + "loss": 2.4736, + "step": 25706 + }, + { + "epoch": 1.1, + "grad_norm": 0.6171875, + "learning_rate": 8.536241103198692e-05, + "loss": 2.5241, + "step": 25707 + }, + { + "epoch": 1.1, + "grad_norm": 0.46484375, + "learning_rate": 8.535577538423613e-05, + "loss": 2.565, + "step": 25708 + }, + { + "epoch": 1.1, + "grad_norm": 0.51171875, + "learning_rate": 8.534913980237901e-05, + "loss": 2.3959, + "step": 25709 + }, + { + "epoch": 1.1, + "grad_norm": 0.48046875, + "learning_rate": 8.534250428644536e-05, + "loss": 2.52, + "step": 25710 + }, + { + "epoch": 1.1, + "grad_norm": 0.41796875, + "learning_rate": 8.533586883646514e-05, + "loss": 2.4819, + "step": 25711 + }, + { + "epoch": 1.1, + "grad_norm": 0.435546875, + "learning_rate": 8.532923345246815e-05, + "loss": 2.3583, + "step": 25712 + }, + { + "epoch": 1.1, + "grad_norm": 0.498046875, + "learning_rate": 8.532259813448422e-05, + "loss": 2.6053, + "step": 25713 + }, + { + "epoch": 1.1, + "grad_norm": 0.40234375, + "learning_rate": 8.531596288254322e-05, + "loss": 2.2126, + "step": 25714 + }, + { + "epoch": 1.1, + "grad_norm": 0.486328125, + "learning_rate": 8.530932769667506e-05, + "loss": 2.53, + "step": 25715 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.530269257690952e-05, + "loss": 2.4895, + "step": 25716 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.529605752327652e-05, + "loss": 2.4772, + "step": 25717 + }, + { + "epoch": 1.1, + "grad_norm": 0.482421875, + "learning_rate": 8.528942253580587e-05, + "loss": 2.3285, + "step": 25718 + }, + { + "epoch": 1.1, + "grad_norm": 0.4140625, + "learning_rate": 8.528278761452745e-05, + "loss": 2.4693, + "step": 25719 + }, + { + "epoch": 1.1, + "grad_norm": 0.474609375, + "learning_rate": 8.527615275947113e-05, + "loss": 2.3721, + "step": 25720 + }, + { + "epoch": 1.1, + "grad_norm": 0.52734375, + "learning_rate": 8.526951797066672e-05, + "loss": 2.6091, + "step": 25721 + }, + { + "epoch": 1.1, + "grad_norm": 0.4765625, + "learning_rate": 8.526288324814408e-05, + "loss": 2.5481, + "step": 25722 + }, + { + "epoch": 1.1, + "grad_norm": 0.482421875, + "learning_rate": 8.52562485919331e-05, + "loss": 2.5893, + "step": 25723 + }, + { + "epoch": 1.1, + "grad_norm": 0.51171875, + "learning_rate": 8.52496140020636e-05, + "loss": 2.5421, + "step": 25724 + }, + { + "epoch": 1.1, + "grad_norm": 0.447265625, + "learning_rate": 8.524297947856547e-05, + "loss": 2.4297, + "step": 25725 + }, + { + "epoch": 1.1, + "grad_norm": 0.5, + "learning_rate": 8.52363450214685e-05, + "loss": 2.3687, + "step": 25726 + }, + { + "epoch": 1.1, + "grad_norm": 0.43359375, + "learning_rate": 8.522971063080261e-05, + "loss": 2.5059, + "step": 25727 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.522307630659764e-05, + "loss": 2.4266, + "step": 25728 + }, + { + "epoch": 1.1, + "grad_norm": 0.46484375, + "learning_rate": 8.521644204888341e-05, + "loss": 2.5096, + "step": 25729 + }, + { + "epoch": 1.1, + "grad_norm": 0.49609375, + "learning_rate": 8.520980785768976e-05, + "loss": 2.5158, + "step": 25730 + }, + { + "epoch": 1.1, + "grad_norm": 0.51953125, + "learning_rate": 8.52031737330466e-05, + "loss": 2.4927, + "step": 25731 + }, + { + "epoch": 1.1, + "grad_norm": 0.435546875, + "learning_rate": 8.519653967498374e-05, + "loss": 2.3984, + "step": 25732 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.518990568353105e-05, + "loss": 2.6136, + "step": 25733 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.518327175871835e-05, + "loss": 2.4004, + "step": 25734 + }, + { + "epoch": 1.1, + "grad_norm": 0.498046875, + "learning_rate": 8.517663790057553e-05, + "loss": 2.4157, + "step": 25735 + }, + { + "epoch": 1.1, + "grad_norm": 0.53515625, + "learning_rate": 8.517000410913245e-05, + "loss": 2.5816, + "step": 25736 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.51633703844189e-05, + "loss": 2.4486, + "step": 25737 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.515673672646474e-05, + "loss": 2.4204, + "step": 25738 + }, + { + "epoch": 1.1, + "grad_norm": 0.51171875, + "learning_rate": 8.515010313529987e-05, + "loss": 2.4964, + "step": 25739 + }, + { + "epoch": 1.1, + "grad_norm": 0.498046875, + "learning_rate": 8.514346961095412e-05, + "loss": 2.5592, + "step": 25740 + }, + { + "epoch": 1.1, + "grad_norm": 0.53515625, + "learning_rate": 8.513683615345731e-05, + "loss": 2.6, + "step": 25741 + }, + { + "epoch": 1.1, + "grad_norm": 0.486328125, + "learning_rate": 8.513020276283928e-05, + "loss": 2.5176, + "step": 25742 + }, + { + "epoch": 1.1, + "grad_norm": 0.5234375, + "learning_rate": 8.512356943912994e-05, + "loss": 2.7934, + "step": 25743 + }, + { + "epoch": 1.1, + "grad_norm": 0.494140625, + "learning_rate": 8.511693618235911e-05, + "loss": 2.5079, + "step": 25744 + }, + { + "epoch": 1.1, + "grad_norm": 0.490234375, + "learning_rate": 8.511030299255661e-05, + "loss": 2.477, + "step": 25745 + }, + { + "epoch": 1.1, + "grad_norm": 0.484375, + "learning_rate": 8.51036698697523e-05, + "loss": 2.5222, + "step": 25746 + }, + { + "epoch": 1.1, + "grad_norm": 0.53125, + "learning_rate": 8.509703681397602e-05, + "loss": 2.4469, + "step": 25747 + }, + { + "epoch": 1.1, + "grad_norm": 0.515625, + "learning_rate": 8.509040382525764e-05, + "loss": 2.5014, + "step": 25748 + }, + { + "epoch": 1.1, + "grad_norm": 0.439453125, + "learning_rate": 8.5083770903627e-05, + "loss": 2.382, + "step": 25749 + }, + { + "epoch": 1.1, + "grad_norm": 0.5390625, + "learning_rate": 8.507713804911393e-05, + "loss": 2.4844, + "step": 25750 + }, + { + "epoch": 1.1, + "grad_norm": 0.43359375, + "learning_rate": 8.507050526174829e-05, + "loss": 2.3779, + "step": 25751 + }, + { + "epoch": 1.1, + "grad_norm": 0.53125, + "learning_rate": 8.506387254155994e-05, + "loss": 2.6523, + "step": 25752 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.50572398885787e-05, + "loss": 2.4555, + "step": 25753 + }, + { + "epoch": 1.1, + "grad_norm": 0.427734375, + "learning_rate": 8.505060730283438e-05, + "loss": 2.4199, + "step": 25754 + }, + { + "epoch": 1.1, + "grad_norm": 0.44921875, + "learning_rate": 8.504397478435689e-05, + "loss": 2.3482, + "step": 25755 + }, + { + "epoch": 1.1, + "grad_norm": 0.4375, + "learning_rate": 8.503734233317607e-05, + "loss": 2.4802, + "step": 25756 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.50307099493217e-05, + "loss": 2.4158, + "step": 25757 + }, + { + "epoch": 1.1, + "grad_norm": 0.44140625, + "learning_rate": 8.502407763282368e-05, + "loss": 2.333, + "step": 25758 + }, + { + "epoch": 1.1, + "grad_norm": 0.46875, + "learning_rate": 8.501744538371183e-05, + "loss": 2.3238, + "step": 25759 + }, + { + "epoch": 1.1, + "grad_norm": 0.498046875, + "learning_rate": 8.501081320201605e-05, + "loss": 2.509, + "step": 25760 + }, + { + "epoch": 1.1, + "grad_norm": 0.4375, + "learning_rate": 8.50041810877661e-05, + "loss": 2.3879, + "step": 25761 + }, + { + "epoch": 1.1, + "grad_norm": 0.45703125, + "learning_rate": 8.499754904099182e-05, + "loss": 2.5246, + "step": 25762 + }, + { + "epoch": 1.1, + "grad_norm": 0.41796875, + "learning_rate": 8.499091706172312e-05, + "loss": 2.5135, + "step": 25763 + }, + { + "epoch": 1.1, + "grad_norm": 0.4921875, + "learning_rate": 8.49842851499898e-05, + "loss": 2.4372, + "step": 25764 + }, + { + "epoch": 1.1, + "grad_norm": 0.48828125, + "learning_rate": 8.497765330582172e-05, + "loss": 2.5467, + "step": 25765 + }, + { + "epoch": 1.1, + "grad_norm": 0.5078125, + "learning_rate": 8.497102152924869e-05, + "loss": 2.5279, + "step": 25766 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.496438982030057e-05, + "loss": 2.4877, + "step": 25767 + }, + { + "epoch": 1.1, + "grad_norm": 0.474609375, + "learning_rate": 8.495775817900724e-05, + "loss": 2.3804, + "step": 25768 + }, + { + "epoch": 1.1, + "grad_norm": 0.44921875, + "learning_rate": 8.495112660539848e-05, + "loss": 2.354, + "step": 25769 + }, + { + "epoch": 1.1, + "grad_norm": 0.48828125, + "learning_rate": 8.494449509950412e-05, + "loss": 2.3964, + "step": 25770 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.493786366135407e-05, + "loss": 2.5904, + "step": 25771 + }, + { + "epoch": 1.1, + "grad_norm": 0.458984375, + "learning_rate": 8.493123229097811e-05, + "loss": 2.443, + "step": 25772 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.492460098840611e-05, + "loss": 2.4774, + "step": 25773 + }, + { + "epoch": 1.1, + "grad_norm": 0.44921875, + "learning_rate": 8.491796975366787e-05, + "loss": 2.458, + "step": 25774 + }, + { + "epoch": 1.1, + "grad_norm": 0.515625, + "learning_rate": 8.491133858679327e-05, + "loss": 2.4137, + "step": 25775 + }, + { + "epoch": 1.1, + "grad_norm": 0.474609375, + "learning_rate": 8.490470748781216e-05, + "loss": 2.5245, + "step": 25776 + }, + { + "epoch": 1.1, + "grad_norm": 0.5, + "learning_rate": 8.489807645675434e-05, + "loss": 2.5418, + "step": 25777 + }, + { + "epoch": 1.1, + "grad_norm": 0.486328125, + "learning_rate": 8.489144549364962e-05, + "loss": 2.4642, + "step": 25778 + }, + { + "epoch": 1.1, + "grad_norm": 0.48046875, + "learning_rate": 8.48848145985279e-05, + "loss": 2.5767, + "step": 25779 + }, + { + "epoch": 1.1, + "grad_norm": 0.5078125, + "learning_rate": 8.487818377141899e-05, + "loss": 2.4604, + "step": 25780 + }, + { + "epoch": 1.1, + "grad_norm": 0.46484375, + "learning_rate": 8.487155301235272e-05, + "loss": 2.3404, + "step": 25781 + }, + { + "epoch": 1.1, + "grad_norm": 0.474609375, + "learning_rate": 8.486492232135891e-05, + "loss": 2.4727, + "step": 25782 + }, + { + "epoch": 1.1, + "grad_norm": 0.51171875, + "learning_rate": 8.485829169846744e-05, + "loss": 2.3877, + "step": 25783 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.485166114370814e-05, + "loss": 2.4696, + "step": 25784 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.484503065711083e-05, + "loss": 2.3394, + "step": 25785 + }, + { + "epoch": 1.1, + "grad_norm": 0.435546875, + "learning_rate": 8.483840023870531e-05, + "loss": 2.4895, + "step": 25786 + }, + { + "epoch": 1.1, + "grad_norm": 0.44921875, + "learning_rate": 8.483176988852147e-05, + "loss": 2.4419, + "step": 25787 + }, + { + "epoch": 1.1, + "grad_norm": 0.48046875, + "learning_rate": 8.482513960658914e-05, + "loss": 2.4944, + "step": 25788 + }, + { + "epoch": 1.1, + "grad_norm": 0.431640625, + "learning_rate": 8.481850939293811e-05, + "loss": 2.4146, + "step": 25789 + }, + { + "epoch": 1.1, + "grad_norm": 0.490234375, + "learning_rate": 8.481187924759822e-05, + "loss": 2.4369, + "step": 25790 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.480524917059935e-05, + "loss": 2.5807, + "step": 25791 + }, + { + "epoch": 1.1, + "grad_norm": 0.435546875, + "learning_rate": 8.479861916197132e-05, + "loss": 2.3664, + "step": 25792 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.479198922174394e-05, + "loss": 2.3933, + "step": 25793 + }, + { + "epoch": 1.1, + "grad_norm": 0.404296875, + "learning_rate": 8.478535934994702e-05, + "loss": 2.36, + "step": 25794 + }, + { + "epoch": 1.1, + "grad_norm": 0.53125, + "learning_rate": 8.477872954661044e-05, + "loss": 2.4723, + "step": 25795 + }, + { + "epoch": 1.1, + "grad_norm": 0.4296875, + "learning_rate": 8.477209981176403e-05, + "loss": 2.4309, + "step": 25796 + }, + { + "epoch": 1.1, + "grad_norm": 0.498046875, + "learning_rate": 8.47654701454376e-05, + "loss": 2.5808, + "step": 25797 + }, + { + "epoch": 1.1, + "grad_norm": 0.412109375, + "learning_rate": 8.475884054766096e-05, + "loss": 2.4313, + "step": 25798 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.475221101846399e-05, + "loss": 2.4552, + "step": 25799 + }, + { + "epoch": 1.1, + "grad_norm": 0.462890625, + "learning_rate": 8.474558155787652e-05, + "loss": 2.4536, + "step": 25800 + }, + { + "epoch": 1.1, + "grad_norm": 0.412109375, + "learning_rate": 8.473895216592834e-05, + "loss": 2.359, + "step": 25801 + }, + { + "epoch": 1.1, + "grad_norm": 0.458984375, + "learning_rate": 8.473232284264928e-05, + "loss": 2.4589, + "step": 25802 + }, + { + "epoch": 1.1, + "grad_norm": 0.57421875, + "learning_rate": 8.47256935880692e-05, + "loss": 2.5148, + "step": 25803 + }, + { + "epoch": 1.1, + "grad_norm": 0.44140625, + "learning_rate": 8.471906440221792e-05, + "loss": 2.3908, + "step": 25804 + }, + { + "epoch": 1.1, + "grad_norm": 0.490234375, + "learning_rate": 8.471243528512527e-05, + "loss": 2.5475, + "step": 25805 + }, + { + "epoch": 1.1, + "grad_norm": 0.482421875, + "learning_rate": 8.470580623682103e-05, + "loss": 2.3663, + "step": 25806 + }, + { + "epoch": 1.1, + "grad_norm": 0.443359375, + "learning_rate": 8.469917725733512e-05, + "loss": 2.5088, + "step": 25807 + }, + { + "epoch": 1.1, + "grad_norm": 0.462890625, + "learning_rate": 8.469254834669734e-05, + "loss": 2.3991, + "step": 25808 + }, + { + "epoch": 1.1, + "grad_norm": 0.474609375, + "learning_rate": 8.468591950493748e-05, + "loss": 2.5204, + "step": 25809 + }, + { + "epoch": 1.1, + "grad_norm": 0.5078125, + "learning_rate": 8.467929073208533e-05, + "loss": 2.4548, + "step": 25810 + }, + { + "epoch": 1.1, + "grad_norm": 0.482421875, + "learning_rate": 8.467266202817084e-05, + "loss": 2.508, + "step": 25811 + }, + { + "epoch": 1.1, + "grad_norm": 0.46875, + "learning_rate": 8.466603339322374e-05, + "loss": 2.5081, + "step": 25812 + }, + { + "epoch": 1.1, + "grad_norm": 0.4921875, + "learning_rate": 8.46594048272739e-05, + "loss": 2.3201, + "step": 25813 + }, + { + "epoch": 1.1, + "grad_norm": 0.5078125, + "learning_rate": 8.465277633035111e-05, + "loss": 2.5321, + "step": 25814 + }, + { + "epoch": 1.1, + "grad_norm": 0.423828125, + "learning_rate": 8.464614790248522e-05, + "loss": 2.5321, + "step": 25815 + }, + { + "epoch": 1.1, + "grad_norm": 0.515625, + "learning_rate": 8.463951954370609e-05, + "loss": 2.4045, + "step": 25816 + }, + { + "epoch": 1.1, + "grad_norm": 0.546875, + "learning_rate": 8.463289125404349e-05, + "loss": 2.5882, + "step": 25817 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.462626303352724e-05, + "loss": 2.5616, + "step": 25818 + }, + { + "epoch": 1.1, + "grad_norm": 0.51171875, + "learning_rate": 8.46196348821872e-05, + "loss": 2.5033, + "step": 25819 + }, + { + "epoch": 1.1, + "grad_norm": 0.443359375, + "learning_rate": 8.46130068000532e-05, + "loss": 2.5839, + "step": 25820 + }, + { + "epoch": 1.1, + "grad_norm": 0.43359375, + "learning_rate": 8.460637878715503e-05, + "loss": 2.4546, + "step": 25821 + }, + { + "epoch": 1.1, + "grad_norm": 0.53515625, + "learning_rate": 8.459975084352251e-05, + "loss": 2.5235, + "step": 25822 + }, + { + "epoch": 1.1, + "grad_norm": 0.41796875, + "learning_rate": 8.459312296918552e-05, + "loss": 2.3277, + "step": 25823 + }, + { + "epoch": 1.1, + "grad_norm": 0.40234375, + "learning_rate": 8.458649516417384e-05, + "loss": 2.4615, + "step": 25824 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.457986742851728e-05, + "loss": 2.4755, + "step": 25825 + }, + { + "epoch": 1.1, + "grad_norm": 0.4765625, + "learning_rate": 8.457323976224572e-05, + "loss": 2.5132, + "step": 25826 + }, + { + "epoch": 1.1, + "grad_norm": 0.419921875, + "learning_rate": 8.456661216538889e-05, + "loss": 2.3718, + "step": 25827 + }, + { + "epoch": 1.1, + "grad_norm": 0.4765625, + "learning_rate": 8.455998463797669e-05, + "loss": 2.4638, + "step": 25828 + }, + { + "epoch": 1.1, + "grad_norm": 0.439453125, + "learning_rate": 8.455335718003891e-05, + "loss": 2.384, + "step": 25829 + }, + { + "epoch": 1.1, + "grad_norm": 0.494140625, + "learning_rate": 8.454672979160542e-05, + "loss": 2.5488, + "step": 25830 + }, + { + "epoch": 1.1, + "grad_norm": 0.44921875, + "learning_rate": 8.454010247270591e-05, + "loss": 2.4885, + "step": 25831 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.453347522337038e-05, + "loss": 2.3788, + "step": 25832 + }, + { + "epoch": 1.1, + "grad_norm": 0.44140625, + "learning_rate": 8.452684804362851e-05, + "loss": 2.5207, + "step": 25833 + }, + { + "epoch": 1.1, + "grad_norm": 0.447265625, + "learning_rate": 8.452022093351019e-05, + "loss": 2.4738, + "step": 25834 + }, + { + "epoch": 1.1, + "grad_norm": 0.55859375, + "learning_rate": 8.451359389304518e-05, + "loss": 2.3636, + "step": 25835 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.450696692226337e-05, + "loss": 2.5119, + "step": 25836 + }, + { + "epoch": 1.1, + "grad_norm": 0.447265625, + "learning_rate": 8.450034002119454e-05, + "loss": 2.4209, + "step": 25837 + }, + { + "epoch": 1.1, + "grad_norm": 0.4296875, + "learning_rate": 8.449371318986855e-05, + "loss": 2.3686, + "step": 25838 + }, + { + "epoch": 1.1, + "grad_norm": 0.51171875, + "learning_rate": 8.448708642831511e-05, + "loss": 2.4565, + "step": 25839 + }, + { + "epoch": 1.1, + "grad_norm": 0.498046875, + "learning_rate": 8.448045973656417e-05, + "loss": 2.4719, + "step": 25840 + }, + { + "epoch": 1.1, + "grad_norm": 0.404296875, + "learning_rate": 8.447383311464548e-05, + "loss": 2.4778, + "step": 25841 + }, + { + "epoch": 1.1, + "grad_norm": 0.494140625, + "learning_rate": 8.446720656258887e-05, + "loss": 2.3871, + "step": 25842 + }, + { + "epoch": 1.1, + "grad_norm": 0.431640625, + "learning_rate": 8.446058008042413e-05, + "loss": 2.4506, + "step": 25843 + }, + { + "epoch": 1.1, + "grad_norm": 0.494140625, + "learning_rate": 8.445395366818111e-05, + "loss": 2.5492, + "step": 25844 + }, + { + "epoch": 1.1, + "grad_norm": 0.4765625, + "learning_rate": 8.444732732588962e-05, + "loss": 2.4598, + "step": 25845 + }, + { + "epoch": 1.1, + "grad_norm": 0.466796875, + "learning_rate": 8.44407010535795e-05, + "loss": 2.6328, + "step": 25846 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.443407485128047e-05, + "loss": 2.4351, + "step": 25847 + }, + { + "epoch": 1.1, + "grad_norm": 0.515625, + "learning_rate": 8.442744871902248e-05, + "loss": 2.4549, + "step": 25848 + }, + { + "epoch": 1.1, + "grad_norm": 0.515625, + "learning_rate": 8.442082265683527e-05, + "loss": 2.5397, + "step": 25849 + }, + { + "epoch": 1.1, + "grad_norm": 0.4453125, + "learning_rate": 8.441419666474865e-05, + "loss": 2.5034, + "step": 25850 + }, + { + "epoch": 1.1, + "grad_norm": 0.443359375, + "learning_rate": 8.440757074279243e-05, + "loss": 2.6434, + "step": 25851 + }, + { + "epoch": 1.1, + "grad_norm": 0.515625, + "learning_rate": 8.440094489099647e-05, + "loss": 2.4376, + "step": 25852 + }, + { + "epoch": 1.1, + "grad_norm": 0.443359375, + "learning_rate": 8.439431910939056e-05, + "loss": 2.3517, + "step": 25853 + }, + { + "epoch": 1.1, + "grad_norm": 0.45703125, + "learning_rate": 8.438769339800452e-05, + "loss": 2.5316, + "step": 25854 + }, + { + "epoch": 1.1, + "grad_norm": 0.4765625, + "learning_rate": 8.438106775686809e-05, + "loss": 2.594, + "step": 25855 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.437444218601121e-05, + "loss": 2.5971, + "step": 25856 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.436781668546363e-05, + "loss": 2.5756, + "step": 25857 + }, + { + "epoch": 1.1, + "grad_norm": 0.51171875, + "learning_rate": 8.436119125525515e-05, + "loss": 2.5438, + "step": 25858 + }, + { + "epoch": 1.1, + "grad_norm": 0.44921875, + "learning_rate": 8.435456589541556e-05, + "loss": 2.4671, + "step": 25859 + }, + { + "epoch": 1.1, + "grad_norm": 0.46484375, + "learning_rate": 8.434794060597474e-05, + "loss": 2.4822, + "step": 25860 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.434131538696246e-05, + "loss": 2.3823, + "step": 25861 + }, + { + "epoch": 1.1, + "grad_norm": 0.46875, + "learning_rate": 8.433469023840857e-05, + "loss": 2.4222, + "step": 25862 + }, + { + "epoch": 1.1, + "grad_norm": 0.486328125, + "learning_rate": 8.432806516034278e-05, + "loss": 2.5637, + "step": 25863 + }, + { + "epoch": 1.1, + "grad_norm": 0.41796875, + "learning_rate": 8.432144015279502e-05, + "loss": 2.5299, + "step": 25864 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.431481521579505e-05, + "loss": 2.4883, + "step": 25865 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.430819034937268e-05, + "loss": 2.4306, + "step": 25866 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.430156555355767e-05, + "loss": 2.5206, + "step": 25867 + }, + { + "epoch": 1.1, + "grad_norm": 0.486328125, + "learning_rate": 8.429494082837991e-05, + "loss": 2.4832, + "step": 25868 + }, + { + "epoch": 1.1, + "grad_norm": 0.43359375, + "learning_rate": 8.428831617386918e-05, + "loss": 2.4073, + "step": 25869 + }, + { + "epoch": 1.1, + "grad_norm": 0.5234375, + "learning_rate": 8.428169159005533e-05, + "loss": 2.4966, + "step": 25870 + }, + { + "epoch": 1.1, + "grad_norm": 0.482421875, + "learning_rate": 8.427506707696803e-05, + "loss": 2.4504, + "step": 25871 + }, + { + "epoch": 1.1, + "grad_norm": 0.458984375, + "learning_rate": 8.426844263463726e-05, + "loss": 2.4902, + "step": 25872 + }, + { + "epoch": 1.1, + "grad_norm": 0.3984375, + "learning_rate": 8.426181826309271e-05, + "loss": 2.3247, + "step": 25873 + }, + { + "epoch": 1.1, + "grad_norm": 0.48046875, + "learning_rate": 8.425519396236425e-05, + "loss": 2.4431, + "step": 25874 + }, + { + "epoch": 1.1, + "grad_norm": 0.486328125, + "learning_rate": 8.424856973248163e-05, + "loss": 2.488, + "step": 25875 + }, + { + "epoch": 1.1, + "grad_norm": 0.427734375, + "learning_rate": 8.42419455734747e-05, + "loss": 2.4417, + "step": 25876 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.423532148537327e-05, + "loss": 2.5513, + "step": 25877 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.422869746820715e-05, + "loss": 2.512, + "step": 25878 + }, + { + "epoch": 1.1, + "grad_norm": 0.43359375, + "learning_rate": 8.422207352200606e-05, + "loss": 2.4837, + "step": 25879 + }, + { + "epoch": 1.1, + "grad_norm": 0.400390625, + "learning_rate": 8.421544964679994e-05, + "loss": 2.5205, + "step": 25880 + }, + { + "epoch": 1.1, + "grad_norm": 0.451171875, + "learning_rate": 8.420882584261851e-05, + "loss": 2.4428, + "step": 25881 + }, + { + "epoch": 1.1, + "grad_norm": 0.458984375, + "learning_rate": 8.42022021094916e-05, + "loss": 2.4537, + "step": 25882 + }, + { + "epoch": 1.1, + "grad_norm": 0.46875, + "learning_rate": 8.419557844744896e-05, + "loss": 2.415, + "step": 25883 + }, + { + "epoch": 1.1, + "grad_norm": 0.458984375, + "learning_rate": 8.418895485652049e-05, + "loss": 2.5129, + "step": 25884 + }, + { + "epoch": 1.1, + "grad_norm": 0.482421875, + "learning_rate": 8.418233133673594e-05, + "loss": 2.5396, + "step": 25885 + }, + { + "epoch": 1.1, + "grad_norm": 0.5234375, + "learning_rate": 8.417570788812512e-05, + "loss": 2.6109, + "step": 25886 + }, + { + "epoch": 1.1, + "grad_norm": 0.43359375, + "learning_rate": 8.416908451071779e-05, + "loss": 2.4644, + "step": 25887 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.416246120454385e-05, + "loss": 2.5261, + "step": 25888 + }, + { + "epoch": 1.1, + "grad_norm": 0.49609375, + "learning_rate": 8.415583796963302e-05, + "loss": 2.5274, + "step": 25889 + }, + { + "epoch": 1.1, + "grad_norm": 0.490234375, + "learning_rate": 8.414921480601514e-05, + "loss": 2.3898, + "step": 25890 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.414259171371997e-05, + "loss": 2.425, + "step": 25891 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.413596869277735e-05, + "loss": 2.4268, + "step": 25892 + }, + { + "epoch": 1.1, + "grad_norm": 0.4375, + "learning_rate": 8.412934574321709e-05, + "loss": 2.4661, + "step": 25893 + }, + { + "epoch": 1.1, + "grad_norm": 0.5078125, + "learning_rate": 8.4122722865069e-05, + "loss": 2.5533, + "step": 25894 + }, + { + "epoch": 1.1, + "grad_norm": 0.490234375, + "learning_rate": 8.411610005836278e-05, + "loss": 2.5206, + "step": 25895 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.410947732312835e-05, + "loss": 2.4295, + "step": 25896 + }, + { + "epoch": 1.1, + "grad_norm": 0.5390625, + "learning_rate": 8.410285465939549e-05, + "loss": 2.489, + "step": 25897 + }, + { + "epoch": 1.1, + "grad_norm": 0.462890625, + "learning_rate": 8.409623206719394e-05, + "loss": 2.5203, + "step": 25898 + }, + { + "epoch": 1.1, + "grad_norm": 0.52734375, + "learning_rate": 8.408960954655353e-05, + "loss": 2.5381, + "step": 25899 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.408298709750409e-05, + "loss": 2.4993, + "step": 25900 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.407636472007536e-05, + "loss": 2.4357, + "step": 25901 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.406974241429722e-05, + "loss": 2.4775, + "step": 25902 + }, + { + "epoch": 1.1, + "grad_norm": 0.462890625, + "learning_rate": 8.406312018019934e-05, + "loss": 2.6134, + "step": 25903 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.405649801781164e-05, + "loss": 2.5431, + "step": 25904 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.404987592716388e-05, + "loss": 2.483, + "step": 25905 + }, + { + "epoch": 1.1, + "grad_norm": 0.451171875, + "learning_rate": 8.404325390828584e-05, + "loss": 2.5411, + "step": 25906 + }, + { + "epoch": 1.1, + "grad_norm": 0.453125, + "learning_rate": 8.403663196120733e-05, + "loss": 2.4422, + "step": 25907 + }, + { + "epoch": 1.1, + "grad_norm": 0.486328125, + "learning_rate": 8.403001008595812e-05, + "loss": 2.5697, + "step": 25908 + }, + { + "epoch": 1.1, + "grad_norm": 0.451171875, + "learning_rate": 8.402338828256805e-05, + "loss": 2.4962, + "step": 25909 + }, + { + "epoch": 1.1, + "grad_norm": 0.51171875, + "learning_rate": 8.401676655106692e-05, + "loss": 2.3616, + "step": 25910 + }, + { + "epoch": 1.1, + "grad_norm": 0.44140625, + "learning_rate": 8.401014489148442e-05, + "loss": 2.2538, + "step": 25911 + }, + { + "epoch": 1.1, + "grad_norm": 0.443359375, + "learning_rate": 8.400352330385049e-05, + "loss": 2.5779, + "step": 25912 + }, + { + "epoch": 1.1, + "grad_norm": 0.4609375, + "learning_rate": 8.399690178819486e-05, + "loss": 2.4982, + "step": 25913 + }, + { + "epoch": 1.1, + "grad_norm": 0.44921875, + "learning_rate": 8.39902803445473e-05, + "loss": 2.4497, + "step": 25914 + }, + { + "epoch": 1.1, + "grad_norm": 0.42578125, + "learning_rate": 8.398365897293762e-05, + "loss": 2.4516, + "step": 25915 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.397703767339563e-05, + "loss": 2.5074, + "step": 25916 + }, + { + "epoch": 1.1, + "grad_norm": 0.470703125, + "learning_rate": 8.397041644595111e-05, + "loss": 2.469, + "step": 25917 + }, + { + "epoch": 1.1, + "grad_norm": 0.458984375, + "learning_rate": 8.396379529063391e-05, + "loss": 2.3638, + "step": 25918 + }, + { + "epoch": 1.1, + "grad_norm": 0.49609375, + "learning_rate": 8.395717420747368e-05, + "loss": 2.5166, + "step": 25919 + }, + { + "epoch": 1.1, + "grad_norm": 0.4375, + "learning_rate": 8.395055319650036e-05, + "loss": 2.4222, + "step": 25920 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.394393225774369e-05, + "loss": 2.4886, + "step": 25921 + }, + { + "epoch": 1.1, + "grad_norm": 0.482421875, + "learning_rate": 8.393731139123344e-05, + "loss": 2.3968, + "step": 25922 + }, + { + "epoch": 1.1, + "grad_norm": 0.435546875, + "learning_rate": 8.39306905969994e-05, + "loss": 2.486, + "step": 25923 + }, + { + "epoch": 1.1, + "grad_norm": 0.490234375, + "learning_rate": 8.39240698750714e-05, + "loss": 2.5826, + "step": 25924 + }, + { + "epoch": 1.1, + "grad_norm": 0.484375, + "learning_rate": 8.391744922547923e-05, + "loss": 2.5845, + "step": 25925 + }, + { + "epoch": 1.1, + "grad_norm": 0.46875, + "learning_rate": 8.391082864825266e-05, + "loss": 2.5095, + "step": 25926 + }, + { + "epoch": 1.1, + "grad_norm": 0.515625, + "learning_rate": 8.390420814342143e-05, + "loss": 2.6148, + "step": 25927 + }, + { + "epoch": 1.1, + "grad_norm": 0.42578125, + "learning_rate": 8.38975877110154e-05, + "loss": 2.2875, + "step": 25928 + }, + { + "epoch": 1.1, + "grad_norm": 0.478515625, + "learning_rate": 8.389096735106439e-05, + "loss": 2.4708, + "step": 25929 + }, + { + "epoch": 1.1, + "grad_norm": 0.50390625, + "learning_rate": 8.388434706359811e-05, + "loss": 2.5253, + "step": 25930 + }, + { + "epoch": 1.1, + "grad_norm": 0.44921875, + "learning_rate": 8.387772684864636e-05, + "loss": 2.5215, + "step": 25931 + }, + { + "epoch": 1.11, + "grad_norm": 0.498046875, + "learning_rate": 8.387110670623896e-05, + "loss": 2.4509, + "step": 25932 + }, + { + "epoch": 1.11, + "grad_norm": 0.48046875, + "learning_rate": 8.386448663640569e-05, + "loss": 2.4068, + "step": 25933 + }, + { + "epoch": 1.11, + "grad_norm": 0.451171875, + "learning_rate": 8.385786663917634e-05, + "loss": 2.3062, + "step": 25934 + }, + { + "epoch": 1.11, + "grad_norm": 0.423828125, + "learning_rate": 8.385124671458064e-05, + "loss": 2.6196, + "step": 25935 + }, + { + "epoch": 1.11, + "grad_norm": 0.44921875, + "learning_rate": 8.384462686264847e-05, + "loss": 2.3626, + "step": 25936 + }, + { + "epoch": 1.11, + "grad_norm": 0.5, + "learning_rate": 8.38380070834096e-05, + "loss": 2.475, + "step": 25937 + }, + { + "epoch": 1.11, + "grad_norm": 0.46484375, + "learning_rate": 8.383138737689376e-05, + "loss": 2.5101, + "step": 25938 + }, + { + "epoch": 1.11, + "grad_norm": 0.455078125, + "learning_rate": 8.382476774313076e-05, + "loss": 2.4873, + "step": 25939 + }, + { + "epoch": 1.11, + "grad_norm": 0.455078125, + "learning_rate": 8.381814818215042e-05, + "loss": 2.5172, + "step": 25940 + }, + { + "epoch": 1.11, + "grad_norm": 0.44921875, + "learning_rate": 8.381152869398247e-05, + "loss": 2.4773, + "step": 25941 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.380490927865675e-05, + "loss": 2.5453, + "step": 25942 + }, + { + "epoch": 1.11, + "grad_norm": 0.478515625, + "learning_rate": 8.379828993620299e-05, + "loss": 2.419, + "step": 25943 + }, + { + "epoch": 1.11, + "grad_norm": 0.52734375, + "learning_rate": 8.379167066665102e-05, + "loss": 2.4518, + "step": 25944 + }, + { + "epoch": 1.11, + "grad_norm": 0.5, + "learning_rate": 8.378505147003064e-05, + "loss": 2.4284, + "step": 25945 + }, + { + "epoch": 1.11, + "grad_norm": 0.435546875, + "learning_rate": 8.377843234637158e-05, + "loss": 2.5329, + "step": 25946 + }, + { + "epoch": 1.11, + "grad_norm": 0.451171875, + "learning_rate": 8.377181329570365e-05, + "loss": 2.6548, + "step": 25947 + }, + { + "epoch": 1.11, + "grad_norm": 0.478515625, + "learning_rate": 8.37651943180566e-05, + "loss": 2.3773, + "step": 25948 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.375857541346025e-05, + "loss": 2.4316, + "step": 25949 + }, + { + "epoch": 1.11, + "grad_norm": 0.443359375, + "learning_rate": 8.37519565819444e-05, + "loss": 2.4862, + "step": 25950 + }, + { + "epoch": 1.11, + "grad_norm": 0.478515625, + "learning_rate": 8.374533782353882e-05, + "loss": 2.4867, + "step": 25951 + }, + { + "epoch": 1.11, + "grad_norm": 0.486328125, + "learning_rate": 8.373871913827321e-05, + "loss": 2.5878, + "step": 25952 + }, + { + "epoch": 1.11, + "grad_norm": 0.470703125, + "learning_rate": 8.37321005261775e-05, + "loss": 2.3681, + "step": 25953 + }, + { + "epoch": 1.11, + "grad_norm": 0.4375, + "learning_rate": 8.372548198728136e-05, + "loss": 2.4061, + "step": 25954 + }, + { + "epoch": 1.11, + "grad_norm": 0.451171875, + "learning_rate": 8.371886352161459e-05, + "loss": 2.5168, + "step": 25955 + }, + { + "epoch": 1.11, + "grad_norm": 0.44140625, + "learning_rate": 8.371224512920699e-05, + "loss": 2.4032, + "step": 25956 + }, + { + "epoch": 1.11, + "grad_norm": 0.462890625, + "learning_rate": 8.370562681008834e-05, + "loss": 2.4017, + "step": 25957 + }, + { + "epoch": 1.11, + "grad_norm": 0.435546875, + "learning_rate": 8.369900856428841e-05, + "loss": 2.5117, + "step": 25958 + }, + { + "epoch": 1.11, + "grad_norm": 0.431640625, + "learning_rate": 8.369239039183703e-05, + "loss": 2.3839, + "step": 25959 + }, + { + "epoch": 1.11, + "grad_norm": 0.5234375, + "learning_rate": 8.368577229276385e-05, + "loss": 2.3804, + "step": 25960 + }, + { + "epoch": 1.11, + "grad_norm": 0.453125, + "learning_rate": 8.367915426709879e-05, + "loss": 2.406, + "step": 25961 + }, + { + "epoch": 1.11, + "grad_norm": 0.54296875, + "learning_rate": 8.367253631487156e-05, + "loss": 2.5866, + "step": 25962 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.366591843611195e-05, + "loss": 2.4375, + "step": 25963 + }, + { + "epoch": 1.11, + "grad_norm": 0.515625, + "learning_rate": 8.365930063084972e-05, + "loss": 2.4328, + "step": 25964 + }, + { + "epoch": 1.11, + "grad_norm": 0.52734375, + "learning_rate": 8.365268289911468e-05, + "loss": 2.5704, + "step": 25965 + }, + { + "epoch": 1.11, + "grad_norm": 0.45703125, + "learning_rate": 8.364606524093659e-05, + "loss": 2.4136, + "step": 25966 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.363944765634526e-05, + "loss": 2.5466, + "step": 25967 + }, + { + "epoch": 1.11, + "grad_norm": 0.439453125, + "learning_rate": 8.363283014537036e-05, + "loss": 2.5012, + "step": 25968 + }, + { + "epoch": 1.11, + "grad_norm": 0.49609375, + "learning_rate": 8.362621270804182e-05, + "loss": 2.5217, + "step": 25969 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.36195953443893e-05, + "loss": 2.4582, + "step": 25970 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.361297805444264e-05, + "loss": 2.5458, + "step": 25971 + }, + { + "epoch": 1.11, + "grad_norm": 0.466796875, + "learning_rate": 8.360636083823154e-05, + "loss": 2.5114, + "step": 25972 + }, + { + "epoch": 1.11, + "grad_norm": 0.4453125, + "learning_rate": 8.359974369578587e-05, + "loss": 2.5948, + "step": 25973 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.359312662713534e-05, + "loss": 2.4831, + "step": 25974 + }, + { + "epoch": 1.11, + "grad_norm": 0.5, + "learning_rate": 8.358650963230978e-05, + "loss": 2.5061, + "step": 25975 + }, + { + "epoch": 1.11, + "grad_norm": 0.443359375, + "learning_rate": 8.357989271133887e-05, + "loss": 2.4973, + "step": 25976 + }, + { + "epoch": 1.11, + "grad_norm": 0.435546875, + "learning_rate": 8.35732758642525e-05, + "loss": 2.3024, + "step": 25977 + }, + { + "epoch": 1.11, + "grad_norm": 0.455078125, + "learning_rate": 8.356665909108036e-05, + "loss": 2.4572, + "step": 25978 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.356004239185224e-05, + "loss": 2.4955, + "step": 25979 + }, + { + "epoch": 1.11, + "grad_norm": 0.48828125, + "learning_rate": 8.355342576659791e-05, + "loss": 2.3744, + "step": 25980 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.354680921534717e-05, + "loss": 2.474, + "step": 25981 + }, + { + "epoch": 1.11, + "grad_norm": 0.4453125, + "learning_rate": 8.354019273812978e-05, + "loss": 2.3629, + "step": 25982 + }, + { + "epoch": 1.11, + "grad_norm": 0.474609375, + "learning_rate": 8.353357633497554e-05, + "loss": 2.2733, + "step": 25983 + }, + { + "epoch": 1.11, + "grad_norm": 0.421875, + "learning_rate": 8.35269600059141e-05, + "loss": 2.515, + "step": 25984 + }, + { + "epoch": 1.11, + "grad_norm": 0.474609375, + "learning_rate": 8.352034375097541e-05, + "loss": 2.5568, + "step": 25985 + }, + { + "epoch": 1.11, + "grad_norm": 0.453125, + "learning_rate": 8.35137275701891e-05, + "loss": 2.5592, + "step": 25986 + }, + { + "epoch": 1.11, + "grad_norm": 0.48828125, + "learning_rate": 8.350711146358501e-05, + "loss": 2.4277, + "step": 25987 + }, + { + "epoch": 1.11, + "grad_norm": 0.5546875, + "learning_rate": 8.350049543119287e-05, + "loss": 2.4351, + "step": 25988 + }, + { + "epoch": 1.11, + "grad_norm": 0.45703125, + "learning_rate": 8.34938794730425e-05, + "loss": 2.5346, + "step": 25989 + }, + { + "epoch": 1.11, + "grad_norm": 0.451171875, + "learning_rate": 8.348726358916361e-05, + "loss": 2.4804, + "step": 25990 + }, + { + "epoch": 1.11, + "grad_norm": 0.484375, + "learning_rate": 8.348064777958604e-05, + "loss": 2.4234, + "step": 25991 + }, + { + "epoch": 1.11, + "grad_norm": 0.41796875, + "learning_rate": 8.347403204433945e-05, + "loss": 2.3741, + "step": 25992 + }, + { + "epoch": 1.11, + "grad_norm": 0.57421875, + "learning_rate": 8.346741638345375e-05, + "loss": 2.616, + "step": 25993 + }, + { + "epoch": 1.11, + "grad_norm": 0.40234375, + "learning_rate": 8.34608007969586e-05, + "loss": 2.338, + "step": 25994 + }, + { + "epoch": 1.11, + "grad_norm": 0.46875, + "learning_rate": 8.345418528488381e-05, + "loss": 2.5017, + "step": 25995 + }, + { + "epoch": 1.11, + "grad_norm": 0.431640625, + "learning_rate": 8.344756984725912e-05, + "loss": 2.4516, + "step": 25996 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.344095448411433e-05, + "loss": 2.5606, + "step": 25997 + }, + { + "epoch": 1.11, + "grad_norm": 0.4453125, + "learning_rate": 8.34343391954792e-05, + "loss": 2.4965, + "step": 25998 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.342772398138351e-05, + "loss": 2.6331, + "step": 25999 + }, + { + "epoch": 1.11, + "grad_norm": 0.4921875, + "learning_rate": 8.342110884185693e-05, + "loss": 2.4216, + "step": 26000 + }, + { + "epoch": 1.11, + "grad_norm": 0.470703125, + "learning_rate": 8.341449377692938e-05, + "loss": 2.549, + "step": 26001 + }, + { + "epoch": 1.11, + "grad_norm": 0.498046875, + "learning_rate": 8.340787878663052e-05, + "loss": 2.5055, + "step": 26002 + }, + { + "epoch": 1.11, + "grad_norm": 0.53515625, + "learning_rate": 8.340126387099015e-05, + "loss": 2.5272, + "step": 26003 + }, + { + "epoch": 1.11, + "grad_norm": 0.515625, + "learning_rate": 8.339464903003799e-05, + "loss": 2.4818, + "step": 26004 + }, + { + "epoch": 1.11, + "grad_norm": 0.453125, + "learning_rate": 8.338803426380387e-05, + "loss": 2.4331, + "step": 26005 + }, + { + "epoch": 1.11, + "grad_norm": 0.443359375, + "learning_rate": 8.338141957231752e-05, + "loss": 2.5338, + "step": 26006 + }, + { + "epoch": 1.11, + "grad_norm": 0.41796875, + "learning_rate": 8.337480495560874e-05, + "loss": 2.5821, + "step": 26007 + }, + { + "epoch": 1.11, + "grad_norm": 0.474609375, + "learning_rate": 8.336819041370719e-05, + "loss": 2.5012, + "step": 26008 + }, + { + "epoch": 1.11, + "grad_norm": 0.447265625, + "learning_rate": 8.336157594664277e-05, + "loss": 2.412, + "step": 26009 + }, + { + "epoch": 1.11, + "grad_norm": 0.490234375, + "learning_rate": 8.335496155444516e-05, + "loss": 2.4922, + "step": 26010 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.334834723714414e-05, + "loss": 2.5672, + "step": 26011 + }, + { + "epoch": 1.11, + "grad_norm": 0.484375, + "learning_rate": 8.334173299476944e-05, + "loss": 2.3367, + "step": 26012 + }, + { + "epoch": 1.11, + "grad_norm": 0.478515625, + "learning_rate": 8.333511882735087e-05, + "loss": 2.4988, + "step": 26013 + }, + { + "epoch": 1.11, + "grad_norm": 0.423828125, + "learning_rate": 8.33285047349182e-05, + "loss": 2.3974, + "step": 26014 + }, + { + "epoch": 1.11, + "grad_norm": 0.462890625, + "learning_rate": 8.332189071750118e-05, + "loss": 2.6048, + "step": 26015 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.331527677512948e-05, + "loss": 2.5877, + "step": 26016 + }, + { + "epoch": 1.11, + "grad_norm": 0.478515625, + "learning_rate": 8.3308662907833e-05, + "loss": 2.4224, + "step": 26017 + }, + { + "epoch": 1.11, + "grad_norm": 0.42578125, + "learning_rate": 8.330204911564143e-05, + "loss": 2.3384, + "step": 26018 + }, + { + "epoch": 1.11, + "grad_norm": 0.4375, + "learning_rate": 8.329543539858453e-05, + "loss": 2.4626, + "step": 26019 + }, + { + "epoch": 1.11, + "grad_norm": 0.49609375, + "learning_rate": 8.328882175669205e-05, + "loss": 2.4321, + "step": 26020 + }, + { + "epoch": 1.11, + "grad_norm": 0.482421875, + "learning_rate": 8.328220818999377e-05, + "loss": 2.5202, + "step": 26021 + }, + { + "epoch": 1.11, + "grad_norm": 0.498046875, + "learning_rate": 8.327559469851947e-05, + "loss": 2.5559, + "step": 26022 + }, + { + "epoch": 1.11, + "grad_norm": 0.44140625, + "learning_rate": 8.326898128229887e-05, + "loss": 2.4713, + "step": 26023 + }, + { + "epoch": 1.11, + "grad_norm": 0.498046875, + "learning_rate": 8.32623679413617e-05, + "loss": 2.403, + "step": 26024 + }, + { + "epoch": 1.11, + "grad_norm": 0.427734375, + "learning_rate": 8.325575467573782e-05, + "loss": 2.4686, + "step": 26025 + }, + { + "epoch": 1.11, + "grad_norm": 0.498046875, + "learning_rate": 8.32491414854569e-05, + "loss": 2.4934, + "step": 26026 + }, + { + "epoch": 1.11, + "grad_norm": 0.515625, + "learning_rate": 8.324252837054872e-05, + "loss": 2.4372, + "step": 26027 + }, + { + "epoch": 1.11, + "grad_norm": 0.4609375, + "learning_rate": 8.323591533104302e-05, + "loss": 2.5015, + "step": 26028 + }, + { + "epoch": 1.11, + "grad_norm": 0.423828125, + "learning_rate": 8.322930236696957e-05, + "loss": 2.3433, + "step": 26029 + }, + { + "epoch": 1.11, + "grad_norm": 0.41796875, + "learning_rate": 8.322268947835817e-05, + "loss": 2.4993, + "step": 26030 + }, + { + "epoch": 1.11, + "grad_norm": 0.498046875, + "learning_rate": 8.321607666523854e-05, + "loss": 2.5968, + "step": 26031 + }, + { + "epoch": 1.11, + "grad_norm": 0.455078125, + "learning_rate": 8.320946392764036e-05, + "loss": 2.4481, + "step": 26032 + }, + { + "epoch": 1.11, + "grad_norm": 0.453125, + "learning_rate": 8.320285126559353e-05, + "loss": 2.45, + "step": 26033 + }, + { + "epoch": 1.11, + "grad_norm": 0.427734375, + "learning_rate": 8.31962386791277e-05, + "loss": 2.3845, + "step": 26034 + }, + { + "epoch": 1.11, + "grad_norm": 0.43359375, + "learning_rate": 8.318962616827266e-05, + "loss": 2.4962, + "step": 26035 + }, + { + "epoch": 1.11, + "grad_norm": 0.4453125, + "learning_rate": 8.318301373305814e-05, + "loss": 2.4225, + "step": 26036 + }, + { + "epoch": 1.11, + "grad_norm": 0.447265625, + "learning_rate": 8.317640137351394e-05, + "loss": 2.4337, + "step": 26037 + }, + { + "epoch": 1.11, + "grad_norm": 0.4296875, + "learning_rate": 8.316978908966978e-05, + "loss": 2.4547, + "step": 26038 + }, + { + "epoch": 1.11, + "grad_norm": 0.474609375, + "learning_rate": 8.316317688155544e-05, + "loss": 2.3464, + "step": 26039 + }, + { + "epoch": 1.11, + "grad_norm": 0.380859375, + "learning_rate": 8.315656474920057e-05, + "loss": 2.4434, + "step": 26040 + }, + { + "epoch": 1.11, + "grad_norm": 0.474609375, + "learning_rate": 8.314995269263508e-05, + "loss": 2.4905, + "step": 26041 + }, + { + "epoch": 1.11, + "grad_norm": 0.466796875, + "learning_rate": 8.314334071188863e-05, + "loss": 2.4615, + "step": 26042 + }, + { + "epoch": 1.11, + "grad_norm": 0.5, + "learning_rate": 8.313672880699098e-05, + "loss": 2.3709, + "step": 26043 + }, + { + "epoch": 1.11, + "grad_norm": 0.5390625, + "learning_rate": 8.313011697797188e-05, + "loss": 2.4883, + "step": 26044 + }, + { + "epoch": 1.11, + "grad_norm": 0.5234375, + "learning_rate": 8.312350522486109e-05, + "loss": 2.6625, + "step": 26045 + }, + { + "epoch": 1.11, + "grad_norm": 0.4921875, + "learning_rate": 8.311689354768836e-05, + "loss": 2.5514, + "step": 26046 + }, + { + "epoch": 1.11, + "grad_norm": 0.453125, + "learning_rate": 8.311028194648348e-05, + "loss": 2.5027, + "step": 26047 + }, + { + "epoch": 1.11, + "grad_norm": 0.53515625, + "learning_rate": 8.310367042127606e-05, + "loss": 2.6241, + "step": 26048 + }, + { + "epoch": 1.11, + "grad_norm": 0.486328125, + "learning_rate": 8.309705897209604e-05, + "loss": 2.453, + "step": 26049 + }, + { + "epoch": 1.11, + "grad_norm": 0.4921875, + "learning_rate": 8.309044759897306e-05, + "loss": 2.3853, + "step": 26050 + }, + { + "epoch": 1.11, + "grad_norm": 0.53125, + "learning_rate": 8.308383630193687e-05, + "loss": 2.5493, + "step": 26051 + }, + { + "epoch": 1.11, + "grad_norm": 0.490234375, + "learning_rate": 8.307722508101722e-05, + "loss": 2.4742, + "step": 26052 + }, + { + "epoch": 1.11, + "grad_norm": 0.44921875, + "learning_rate": 8.307061393624389e-05, + "loss": 2.2856, + "step": 26053 + }, + { + "epoch": 1.11, + "grad_norm": 0.4375, + "learning_rate": 8.306400286764662e-05, + "loss": 2.3534, + "step": 26054 + }, + { + "epoch": 1.11, + "grad_norm": 0.51171875, + "learning_rate": 8.305739187525515e-05, + "loss": 2.3844, + "step": 26055 + }, + { + "epoch": 1.11, + "grad_norm": 0.482421875, + "learning_rate": 8.305078095909917e-05, + "loss": 2.5383, + "step": 26056 + }, + { + "epoch": 1.11, + "grad_norm": 0.52734375, + "learning_rate": 8.304417011920856e-05, + "loss": 2.4358, + "step": 26057 + }, + { + "epoch": 1.11, + "grad_norm": 0.5234375, + "learning_rate": 8.303755935561295e-05, + "loss": 2.4493, + "step": 26058 + }, + { + "epoch": 1.11, + "grad_norm": 0.419921875, + "learning_rate": 8.303094866834213e-05, + "loss": 2.4592, + "step": 26059 + }, + { + "epoch": 1.11, + "grad_norm": 0.453125, + "learning_rate": 8.302433805742583e-05, + "loss": 2.4255, + "step": 26060 + }, + { + "epoch": 1.11, + "grad_norm": 0.451171875, + "learning_rate": 8.30177275228938e-05, + "loss": 2.3711, + "step": 26061 + }, + { + "epoch": 1.11, + "grad_norm": 0.48046875, + "learning_rate": 8.30111170647758e-05, + "loss": 2.4809, + "step": 26062 + }, + { + "epoch": 1.11, + "grad_norm": 0.546875, + "learning_rate": 8.30045066831016e-05, + "loss": 2.5975, + "step": 26063 + }, + { + "epoch": 1.11, + "grad_norm": 0.46875, + "learning_rate": 8.299789637790082e-05, + "loss": 2.4272, + "step": 26064 + }, + { + "epoch": 1.11, + "grad_norm": 0.466796875, + "learning_rate": 8.299128614920338e-05, + "loss": 2.4715, + "step": 26065 + }, + { + "epoch": 1.11, + "grad_norm": 0.4609375, + "learning_rate": 8.29846759970389e-05, + "loss": 2.3926, + "step": 26066 + }, + { + "epoch": 1.11, + "grad_norm": 0.40234375, + "learning_rate": 8.297806592143716e-05, + "loss": 2.408, + "step": 26067 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.29714559224279e-05, + "loss": 2.4562, + "step": 26068 + }, + { + "epoch": 1.11, + "grad_norm": 0.482421875, + "learning_rate": 8.296484600004087e-05, + "loss": 2.4107, + "step": 26069 + }, + { + "epoch": 1.11, + "grad_norm": 0.4453125, + "learning_rate": 8.29582361543058e-05, + "loss": 2.6028, + "step": 26070 + }, + { + "epoch": 1.11, + "grad_norm": 0.51953125, + "learning_rate": 8.295162638525246e-05, + "loss": 2.4973, + "step": 26071 + }, + { + "epoch": 1.11, + "grad_norm": 0.427734375, + "learning_rate": 8.294501669291058e-05, + "loss": 2.3066, + "step": 26072 + }, + { + "epoch": 1.11, + "grad_norm": 0.49609375, + "learning_rate": 8.293840707730985e-05, + "loss": 2.3789, + "step": 26073 + }, + { + "epoch": 1.11, + "grad_norm": 0.43359375, + "learning_rate": 8.293179753848006e-05, + "loss": 2.4739, + "step": 26074 + }, + { + "epoch": 1.11, + "grad_norm": 0.48046875, + "learning_rate": 8.292518807645095e-05, + "loss": 2.4861, + "step": 26075 + }, + { + "epoch": 1.11, + "grad_norm": 0.5078125, + "learning_rate": 8.291857869125227e-05, + "loss": 2.5676, + "step": 26076 + }, + { + "epoch": 1.11, + "grad_norm": 0.490234375, + "learning_rate": 8.291196938291371e-05, + "loss": 2.442, + "step": 26077 + }, + { + "epoch": 1.11, + "grad_norm": 0.482421875, + "learning_rate": 8.290536015146507e-05, + "loss": 2.587, + "step": 26078 + }, + { + "epoch": 1.11, + "grad_norm": 0.453125, + "learning_rate": 8.289875099693608e-05, + "loss": 2.47, + "step": 26079 + }, + { + "epoch": 1.11, + "grad_norm": 0.52734375, + "learning_rate": 8.289214191935645e-05, + "loss": 2.6764, + "step": 26080 + }, + { + "epoch": 1.11, + "grad_norm": 0.447265625, + "learning_rate": 8.28855329187559e-05, + "loss": 2.4264, + "step": 26081 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.287892399516421e-05, + "loss": 2.4057, + "step": 26082 + }, + { + "epoch": 1.11, + "grad_norm": 0.46484375, + "learning_rate": 8.287231514861111e-05, + "loss": 2.4724, + "step": 26083 + }, + { + "epoch": 1.11, + "grad_norm": 0.484375, + "learning_rate": 8.286570637912633e-05, + "loss": 2.4325, + "step": 26084 + }, + { + "epoch": 1.11, + "grad_norm": 0.44140625, + "learning_rate": 8.28590976867396e-05, + "loss": 2.3045, + "step": 26085 + }, + { + "epoch": 1.11, + "grad_norm": 0.494140625, + "learning_rate": 8.285248907148067e-05, + "loss": 2.3897, + "step": 26086 + }, + { + "epoch": 1.11, + "grad_norm": 0.408203125, + "learning_rate": 8.284588053337931e-05, + "loss": 2.5616, + "step": 26087 + }, + { + "epoch": 1.11, + "grad_norm": 0.46484375, + "learning_rate": 8.28392720724652e-05, + "loss": 2.4073, + "step": 26088 + }, + { + "epoch": 1.11, + "grad_norm": 0.46875, + "learning_rate": 8.283266368876806e-05, + "loss": 2.4624, + "step": 26089 + }, + { + "epoch": 1.11, + "grad_norm": 0.451171875, + "learning_rate": 8.282605538231767e-05, + "loss": 2.5225, + "step": 26090 + }, + { + "epoch": 1.11, + "grad_norm": 0.4140625, + "learning_rate": 8.281944715314377e-05, + "loss": 2.5022, + "step": 26091 + }, + { + "epoch": 1.11, + "grad_norm": 0.50390625, + "learning_rate": 8.281283900127608e-05, + "loss": 2.4776, + "step": 26092 + }, + { + "epoch": 1.11, + "grad_norm": 0.48828125, + "learning_rate": 8.280623092674431e-05, + "loss": 2.311, + "step": 26093 + }, + { + "epoch": 1.11, + "grad_norm": 0.42578125, + "learning_rate": 8.279962292957823e-05, + "loss": 2.4727, + "step": 26094 + }, + { + "epoch": 1.11, + "grad_norm": 0.47265625, + "learning_rate": 8.279301500980761e-05, + "loss": 2.4613, + "step": 26095 + }, + { + "epoch": 1.11, + "grad_norm": 0.515625, + "learning_rate": 8.278640716746209e-05, + "loss": 2.5803, + "step": 26096 + }, + { + "epoch": 1.11, + "grad_norm": 0.5, + "learning_rate": 8.277979940257143e-05, + "loss": 2.322, + "step": 26097 + }, + { + "epoch": 1.11, + "grad_norm": 0.5, + "learning_rate": 8.277319171516539e-05, + "loss": 2.4949, + "step": 26098 + }, + { + "epoch": 1.11, + "grad_norm": 0.478515625, + "learning_rate": 8.27665841052737e-05, + "loss": 2.4782, + "step": 26099 + }, + { + "epoch": 1.11, + "grad_norm": 0.53125, + "learning_rate": 8.27599765729261e-05, + "loss": 2.4893, + "step": 26100 + }, + { + "epoch": 1.11, + "grad_norm": 0.45703125, + "learning_rate": 8.275336911815226e-05, + "loss": 2.5105, + "step": 26101 + }, + { + "epoch": 1.11, + "grad_norm": 0.4375, + "learning_rate": 8.274676174098197e-05, + "loss": 2.4023, + "step": 26102 + }, + { + "epoch": 1.11, + "grad_norm": 0.455078125, + "learning_rate": 8.2740154441445e-05, + "loss": 2.4449, + "step": 26103 + }, + { + "epoch": 1.11, + "grad_norm": 0.5078125, + "learning_rate": 8.273354721957099e-05, + "loss": 2.5804, + "step": 26104 + }, + { + "epoch": 1.11, + "grad_norm": 0.46875, + "learning_rate": 8.272694007538968e-05, + "loss": 2.3738, + "step": 26105 + }, + { + "epoch": 1.11, + "grad_norm": 0.439453125, + "learning_rate": 8.272033300893085e-05, + "loss": 2.4039, + "step": 26106 + }, + { + "epoch": 1.11, + "grad_norm": 0.416015625, + "learning_rate": 8.27137260202242e-05, + "loss": 2.4304, + "step": 26107 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.27071191092995e-05, + "loss": 2.5264, + "step": 26108 + }, + { + "epoch": 1.11, + "grad_norm": 0.46484375, + "learning_rate": 8.270051227618639e-05, + "loss": 2.524, + "step": 26109 + }, + { + "epoch": 1.11, + "grad_norm": 0.400390625, + "learning_rate": 8.269390552091467e-05, + "loss": 2.4849, + "step": 26110 + }, + { + "epoch": 1.11, + "grad_norm": 0.41015625, + "learning_rate": 8.268729884351409e-05, + "loss": 2.4487, + "step": 26111 + }, + { + "epoch": 1.11, + "grad_norm": 0.470703125, + "learning_rate": 8.268069224401431e-05, + "loss": 2.4322, + "step": 26112 + }, + { + "epoch": 1.11, + "grad_norm": 0.484375, + "learning_rate": 8.267408572244505e-05, + "loss": 2.609, + "step": 26113 + }, + { + "epoch": 1.11, + "grad_norm": 0.4609375, + "learning_rate": 8.266747927883613e-05, + "loss": 2.5671, + "step": 26114 + }, + { + "epoch": 1.11, + "grad_norm": 0.43359375, + "learning_rate": 8.266087291321718e-05, + "loss": 2.4949, + "step": 26115 + }, + { + "epoch": 1.11, + "grad_norm": 0.412109375, + "learning_rate": 8.2654266625618e-05, + "loss": 2.3165, + "step": 26116 + }, + { + "epoch": 1.11, + "grad_norm": 0.5078125, + "learning_rate": 8.264766041606824e-05, + "loss": 2.5114, + "step": 26117 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.264105428459769e-05, + "loss": 2.5174, + "step": 26118 + }, + { + "epoch": 1.11, + "grad_norm": 0.4296875, + "learning_rate": 8.263444823123607e-05, + "loss": 2.5896, + "step": 26119 + }, + { + "epoch": 1.11, + "grad_norm": 0.482421875, + "learning_rate": 8.262784225601308e-05, + "loss": 2.4429, + "step": 26120 + }, + { + "epoch": 1.11, + "grad_norm": 0.451171875, + "learning_rate": 8.262123635895841e-05, + "loss": 2.4166, + "step": 26121 + }, + { + "epoch": 1.11, + "grad_norm": 0.466796875, + "learning_rate": 8.261463054010185e-05, + "loss": 2.4792, + "step": 26122 + }, + { + "epoch": 1.11, + "grad_norm": 0.408203125, + "learning_rate": 8.260802479947311e-05, + "loss": 2.4454, + "step": 26123 + }, + { + "epoch": 1.11, + "grad_norm": 0.490234375, + "learning_rate": 8.260141913710189e-05, + "loss": 2.417, + "step": 26124 + }, + { + "epoch": 1.11, + "grad_norm": 0.447265625, + "learning_rate": 8.259481355301791e-05, + "loss": 2.2576, + "step": 26125 + }, + { + "epoch": 1.11, + "grad_norm": 0.4609375, + "learning_rate": 8.258820804725094e-05, + "loss": 2.4844, + "step": 26126 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.258160261983068e-05, + "loss": 2.4723, + "step": 26127 + }, + { + "epoch": 1.11, + "grad_norm": 0.484375, + "learning_rate": 8.257499727078681e-05, + "loss": 2.3869, + "step": 26128 + }, + { + "epoch": 1.11, + "grad_norm": 0.515625, + "learning_rate": 8.256839200014908e-05, + "loss": 2.3238, + "step": 26129 + }, + { + "epoch": 1.11, + "grad_norm": 0.447265625, + "learning_rate": 8.256178680794724e-05, + "loss": 2.3914, + "step": 26130 + }, + { + "epoch": 1.11, + "grad_norm": 0.46484375, + "learning_rate": 8.255518169421097e-05, + "loss": 2.5113, + "step": 26131 + }, + { + "epoch": 1.11, + "grad_norm": 0.54296875, + "learning_rate": 8.254857665897003e-05, + "loss": 2.4383, + "step": 26132 + }, + { + "epoch": 1.11, + "grad_norm": 0.46875, + "learning_rate": 8.254197170225407e-05, + "loss": 2.4934, + "step": 26133 + }, + { + "epoch": 1.11, + "grad_norm": 0.462890625, + "learning_rate": 8.253536682409289e-05, + "loss": 2.3967, + "step": 26134 + }, + { + "epoch": 1.11, + "grad_norm": 0.51171875, + "learning_rate": 8.252876202451621e-05, + "loss": 2.7396, + "step": 26135 + }, + { + "epoch": 1.11, + "grad_norm": 0.46875, + "learning_rate": 8.252215730355368e-05, + "loss": 2.4685, + "step": 26136 + }, + { + "epoch": 1.11, + "grad_norm": 0.4453125, + "learning_rate": 8.251555266123503e-05, + "loss": 2.3985, + "step": 26137 + }, + { + "epoch": 1.11, + "grad_norm": 0.43359375, + "learning_rate": 8.250894809759004e-05, + "loss": 2.4317, + "step": 26138 + }, + { + "epoch": 1.11, + "grad_norm": 0.4453125, + "learning_rate": 8.250234361264838e-05, + "loss": 2.4231, + "step": 26139 + }, + { + "epoch": 1.11, + "grad_norm": 0.50390625, + "learning_rate": 8.249573920643978e-05, + "loss": 2.529, + "step": 26140 + }, + { + "epoch": 1.11, + "grad_norm": 0.453125, + "learning_rate": 8.248913487899394e-05, + "loss": 2.5105, + "step": 26141 + }, + { + "epoch": 1.11, + "grad_norm": 0.45703125, + "learning_rate": 8.24825306303406e-05, + "loss": 2.5176, + "step": 26142 + }, + { + "epoch": 1.11, + "grad_norm": 0.4296875, + "learning_rate": 8.247592646050952e-05, + "loss": 2.6028, + "step": 26143 + }, + { + "epoch": 1.11, + "grad_norm": 0.4765625, + "learning_rate": 8.246932236953031e-05, + "loss": 2.4257, + "step": 26144 + }, + { + "epoch": 1.11, + "grad_norm": 0.42578125, + "learning_rate": 8.246271835743274e-05, + "loss": 2.598, + "step": 26145 + }, + { + "epoch": 1.11, + "grad_norm": 0.46875, + "learning_rate": 8.245611442424655e-05, + "loss": 2.4526, + "step": 26146 + }, + { + "epoch": 1.11, + "grad_norm": 0.458984375, + "learning_rate": 8.244951057000143e-05, + "loss": 2.4995, + "step": 26147 + }, + { + "epoch": 1.11, + "grad_norm": 0.46875, + "learning_rate": 8.244290679472711e-05, + "loss": 2.4793, + "step": 26148 + }, + { + "epoch": 1.11, + "grad_norm": 0.50390625, + "learning_rate": 8.243630309845326e-05, + "loss": 2.4782, + "step": 26149 + }, + { + "epoch": 1.11, + "grad_norm": 0.51953125, + "learning_rate": 8.242969948120965e-05, + "loss": 2.4726, + "step": 26150 + }, + { + "epoch": 1.11, + "grad_norm": 0.478515625, + "learning_rate": 8.242309594302599e-05, + "loss": 2.4322, + "step": 26151 + }, + { + "epoch": 1.11, + "grad_norm": 0.4921875, + "learning_rate": 8.241649248393196e-05, + "loss": 2.5065, + "step": 26152 + }, + { + "epoch": 1.11, + "grad_norm": 0.5078125, + "learning_rate": 8.240988910395726e-05, + "loss": 2.3617, + "step": 26153 + }, + { + "epoch": 1.11, + "grad_norm": 0.451171875, + "learning_rate": 8.240328580313165e-05, + "loss": 2.5061, + "step": 26154 + }, + { + "epoch": 1.11, + "grad_norm": 0.494140625, + "learning_rate": 8.239668258148484e-05, + "loss": 2.6199, + "step": 26155 + }, + { + "epoch": 1.11, + "grad_norm": 0.51171875, + "learning_rate": 8.23900794390465e-05, + "loss": 2.584, + "step": 26156 + }, + { + "epoch": 1.11, + "grad_norm": 0.42578125, + "learning_rate": 8.238347637584636e-05, + "loss": 2.4488, + "step": 26157 + }, + { + "epoch": 1.11, + "grad_norm": 0.4609375, + "learning_rate": 8.237687339191418e-05, + "loss": 2.467, + "step": 26158 + }, + { + "epoch": 1.11, + "grad_norm": 0.40625, + "learning_rate": 8.237027048727963e-05, + "loss": 2.3219, + "step": 26159 + }, + { + "epoch": 1.11, + "grad_norm": 0.466796875, + "learning_rate": 8.236366766197239e-05, + "loss": 2.5259, + "step": 26160 + }, + { + "epoch": 1.11, + "grad_norm": 0.486328125, + "learning_rate": 8.235706491602218e-05, + "loss": 2.586, + "step": 26161 + }, + { + "epoch": 1.11, + "grad_norm": 0.47265625, + "learning_rate": 8.235046224945876e-05, + "loss": 2.4796, + "step": 26162 + }, + { + "epoch": 1.11, + "grad_norm": 0.4296875, + "learning_rate": 8.234385966231182e-05, + "loss": 2.4826, + "step": 26163 + }, + { + "epoch": 1.11, + "grad_norm": 0.455078125, + "learning_rate": 8.233725715461104e-05, + "loss": 2.4493, + "step": 26164 + }, + { + "epoch": 1.11, + "grad_norm": 0.47265625, + "learning_rate": 8.233065472638612e-05, + "loss": 2.4263, + "step": 26165 + }, + { + "epoch": 1.12, + "grad_norm": 0.455078125, + "learning_rate": 8.232405237766684e-05, + "loss": 2.4347, + "step": 26166 + }, + { + "epoch": 1.12, + "grad_norm": 0.50390625, + "learning_rate": 8.231745010848288e-05, + "loss": 2.489, + "step": 26167 + }, + { + "epoch": 1.12, + "grad_norm": 0.447265625, + "learning_rate": 8.231084791886391e-05, + "loss": 2.3969, + "step": 26168 + }, + { + "epoch": 1.12, + "grad_norm": 0.515625, + "learning_rate": 8.230424580883964e-05, + "loss": 2.6608, + "step": 26169 + }, + { + "epoch": 1.12, + "grad_norm": 0.5234375, + "learning_rate": 8.22976437784398e-05, + "loss": 2.604, + "step": 26170 + }, + { + "epoch": 1.12, + "grad_norm": 0.443359375, + "learning_rate": 8.229104182769412e-05, + "loss": 2.4179, + "step": 26171 + }, + { + "epoch": 1.12, + "grad_norm": 0.466796875, + "learning_rate": 8.228443995663227e-05, + "loss": 2.4706, + "step": 26172 + }, + { + "epoch": 1.12, + "grad_norm": 0.48046875, + "learning_rate": 8.227783816528394e-05, + "loss": 2.5741, + "step": 26173 + }, + { + "epoch": 1.12, + "grad_norm": 0.498046875, + "learning_rate": 8.227123645367887e-05, + "loss": 2.666, + "step": 26174 + }, + { + "epoch": 1.12, + "grad_norm": 0.4609375, + "learning_rate": 8.226463482184681e-05, + "loss": 2.4948, + "step": 26175 + }, + { + "epoch": 1.12, + "grad_norm": 0.486328125, + "learning_rate": 8.225803326981737e-05, + "loss": 2.5182, + "step": 26176 + }, + { + "epoch": 1.12, + "grad_norm": 0.47265625, + "learning_rate": 8.225143179762027e-05, + "loss": 2.5441, + "step": 26177 + }, + { + "epoch": 1.12, + "grad_norm": 0.427734375, + "learning_rate": 8.224483040528527e-05, + "loss": 2.3337, + "step": 26178 + }, + { + "epoch": 1.12, + "grad_norm": 0.490234375, + "learning_rate": 8.223822909284205e-05, + "loss": 2.4144, + "step": 26179 + }, + { + "epoch": 1.12, + "grad_norm": 0.474609375, + "learning_rate": 8.22316278603203e-05, + "loss": 2.3892, + "step": 26180 + }, + { + "epoch": 1.12, + "grad_norm": 0.48828125, + "learning_rate": 8.222502670774972e-05, + "loss": 2.313, + "step": 26181 + }, + { + "epoch": 1.12, + "grad_norm": 0.421875, + "learning_rate": 8.221842563516003e-05, + "loss": 2.5179, + "step": 26182 + }, + { + "epoch": 1.12, + "grad_norm": 0.451171875, + "learning_rate": 8.221182464258096e-05, + "loss": 2.4383, + "step": 26183 + }, + { + "epoch": 1.12, + "grad_norm": 0.478515625, + "learning_rate": 8.220522373004215e-05, + "loss": 2.4753, + "step": 26184 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.219862289757331e-05, + "loss": 2.6446, + "step": 26185 + }, + { + "epoch": 1.12, + "grad_norm": 0.44140625, + "learning_rate": 8.219202214520418e-05, + "loss": 2.4405, + "step": 26186 + }, + { + "epoch": 1.12, + "grad_norm": 0.58203125, + "learning_rate": 8.218542147296445e-05, + "loss": 2.5772, + "step": 26187 + }, + { + "epoch": 1.12, + "grad_norm": 0.4921875, + "learning_rate": 8.217882088088381e-05, + "loss": 2.5711, + "step": 26188 + }, + { + "epoch": 1.12, + "grad_norm": 0.451171875, + "learning_rate": 8.217222036899199e-05, + "loss": 2.4, + "step": 26189 + }, + { + "epoch": 1.12, + "grad_norm": 0.4453125, + "learning_rate": 8.216561993731859e-05, + "loss": 2.4967, + "step": 26190 + }, + { + "epoch": 1.12, + "grad_norm": 0.427734375, + "learning_rate": 8.215901958589345e-05, + "loss": 2.5532, + "step": 26191 + }, + { + "epoch": 1.12, + "grad_norm": 0.427734375, + "learning_rate": 8.215241931474618e-05, + "loss": 2.4657, + "step": 26192 + }, + { + "epoch": 1.12, + "grad_norm": 0.47265625, + "learning_rate": 8.214581912390649e-05, + "loss": 2.4158, + "step": 26193 + }, + { + "epoch": 1.12, + "grad_norm": 0.43359375, + "learning_rate": 8.213921901340407e-05, + "loss": 2.5879, + "step": 26194 + }, + { + "epoch": 1.12, + "grad_norm": 0.427734375, + "learning_rate": 8.213261898326867e-05, + "loss": 2.4472, + "step": 26195 + }, + { + "epoch": 1.12, + "grad_norm": 0.515625, + "learning_rate": 8.212601903352995e-05, + "loss": 2.4564, + "step": 26196 + }, + { + "epoch": 1.12, + "grad_norm": 0.47265625, + "learning_rate": 8.211941916421764e-05, + "loss": 2.4949, + "step": 26197 + }, + { + "epoch": 1.12, + "grad_norm": 0.408203125, + "learning_rate": 8.211281937536133e-05, + "loss": 2.4181, + "step": 26198 + }, + { + "epoch": 1.12, + "grad_norm": 0.51953125, + "learning_rate": 8.210621966699089e-05, + "loss": 2.5275, + "step": 26199 + }, + { + "epoch": 1.12, + "grad_norm": 0.5078125, + "learning_rate": 8.209962003913587e-05, + "loss": 2.5513, + "step": 26200 + }, + { + "epoch": 1.12, + "grad_norm": 0.57421875, + "learning_rate": 8.209302049182602e-05, + "loss": 2.5239, + "step": 26201 + }, + { + "epoch": 1.12, + "grad_norm": 0.48046875, + "learning_rate": 8.208642102509102e-05, + "loss": 2.5573, + "step": 26202 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.20798216389606e-05, + "loss": 2.4498, + "step": 26203 + }, + { + "epoch": 1.12, + "grad_norm": 0.4375, + "learning_rate": 8.207322233346444e-05, + "loss": 2.353, + "step": 26204 + }, + { + "epoch": 1.12, + "grad_norm": 0.37890625, + "learning_rate": 8.206662310863225e-05, + "loss": 2.5274, + "step": 26205 + }, + { + "epoch": 1.12, + "grad_norm": 0.453125, + "learning_rate": 8.206002396449362e-05, + "loss": 2.4551, + "step": 26206 + }, + { + "epoch": 1.12, + "grad_norm": 0.49609375, + "learning_rate": 8.20534249010784e-05, + "loss": 2.5586, + "step": 26207 + }, + { + "epoch": 1.12, + "grad_norm": 0.466796875, + "learning_rate": 8.20468259184162e-05, + "loss": 2.4214, + "step": 26208 + }, + { + "epoch": 1.12, + "grad_norm": 0.4921875, + "learning_rate": 8.20402270165367e-05, + "loss": 2.5777, + "step": 26209 + }, + { + "epoch": 1.12, + "grad_norm": 0.49609375, + "learning_rate": 8.203362819546961e-05, + "loss": 2.3877, + "step": 26210 + }, + { + "epoch": 1.12, + "grad_norm": 0.474609375, + "learning_rate": 8.202702945524463e-05, + "loss": 2.5167, + "step": 26211 + }, + { + "epoch": 1.12, + "grad_norm": 0.53125, + "learning_rate": 8.202043079589146e-05, + "loss": 2.6629, + "step": 26212 + }, + { + "epoch": 1.12, + "grad_norm": 0.44921875, + "learning_rate": 8.20138322174398e-05, + "loss": 2.3928, + "step": 26213 + }, + { + "epoch": 1.12, + "grad_norm": 0.4921875, + "learning_rate": 8.200723371991928e-05, + "loss": 2.5111, + "step": 26214 + }, + { + "epoch": 1.12, + "grad_norm": 0.421875, + "learning_rate": 8.200063530335967e-05, + "loss": 2.3754, + "step": 26215 + }, + { + "epoch": 1.12, + "grad_norm": 0.4921875, + "learning_rate": 8.19940369677906e-05, + "loss": 2.5767, + "step": 26216 + }, + { + "epoch": 1.12, + "grad_norm": 0.453125, + "learning_rate": 8.19874387132418e-05, + "loss": 2.4549, + "step": 26217 + }, + { + "epoch": 1.12, + "grad_norm": 0.41796875, + "learning_rate": 8.198084053974292e-05, + "loss": 2.3332, + "step": 26218 + }, + { + "epoch": 1.12, + "grad_norm": 0.478515625, + "learning_rate": 8.197424244732369e-05, + "loss": 2.5288, + "step": 26219 + }, + { + "epoch": 1.12, + "grad_norm": 0.4609375, + "learning_rate": 8.196764443601379e-05, + "loss": 2.4472, + "step": 26220 + }, + { + "epoch": 1.12, + "grad_norm": 0.455078125, + "learning_rate": 8.19610465058429e-05, + "loss": 2.4682, + "step": 26221 + }, + { + "epoch": 1.12, + "grad_norm": 0.50390625, + "learning_rate": 8.195444865684065e-05, + "loss": 2.4976, + "step": 26222 + }, + { + "epoch": 1.12, + "grad_norm": 0.48828125, + "learning_rate": 8.194785088903686e-05, + "loss": 2.5278, + "step": 26223 + }, + { + "epoch": 1.12, + "grad_norm": 0.486328125, + "learning_rate": 8.194125320246112e-05, + "loss": 2.5267, + "step": 26224 + }, + { + "epoch": 1.12, + "grad_norm": 0.51171875, + "learning_rate": 8.193465559714313e-05, + "loss": 2.4728, + "step": 26225 + }, + { + "epoch": 1.12, + "grad_norm": 0.5703125, + "learning_rate": 8.19280580731126e-05, + "loss": 2.4888, + "step": 26226 + }, + { + "epoch": 1.12, + "grad_norm": 0.52734375, + "learning_rate": 8.192146063039918e-05, + "loss": 2.509, + "step": 26227 + }, + { + "epoch": 1.12, + "grad_norm": 0.5, + "learning_rate": 8.191486326903261e-05, + "loss": 2.5494, + "step": 26228 + }, + { + "epoch": 1.12, + "grad_norm": 0.45703125, + "learning_rate": 8.190826598904255e-05, + "loss": 2.4835, + "step": 26229 + }, + { + "epoch": 1.12, + "grad_norm": 0.474609375, + "learning_rate": 8.190166879045863e-05, + "loss": 2.4637, + "step": 26230 + }, + { + "epoch": 1.12, + "grad_norm": 0.51953125, + "learning_rate": 8.189507167331065e-05, + "loss": 2.495, + "step": 26231 + }, + { + "epoch": 1.12, + "grad_norm": 0.486328125, + "learning_rate": 8.18884746376282e-05, + "loss": 2.435, + "step": 26232 + }, + { + "epoch": 1.12, + "grad_norm": 0.44140625, + "learning_rate": 8.1881877683441e-05, + "loss": 2.5272, + "step": 26233 + }, + { + "epoch": 1.12, + "grad_norm": 0.373046875, + "learning_rate": 8.187528081077871e-05, + "loss": 2.3822, + "step": 26234 + }, + { + "epoch": 1.12, + "grad_norm": 0.46484375, + "learning_rate": 8.186868401967105e-05, + "loss": 2.3523, + "step": 26235 + }, + { + "epoch": 1.12, + "grad_norm": 0.51171875, + "learning_rate": 8.186208731014769e-05, + "loss": 2.4392, + "step": 26236 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.185549068223833e-05, + "loss": 2.4421, + "step": 26237 + }, + { + "epoch": 1.12, + "grad_norm": 0.46875, + "learning_rate": 8.184889413597257e-05, + "loss": 2.3892, + "step": 26238 + }, + { + "epoch": 1.12, + "grad_norm": 0.45703125, + "learning_rate": 8.184229767138021e-05, + "loss": 2.4799, + "step": 26239 + }, + { + "epoch": 1.12, + "grad_norm": 0.439453125, + "learning_rate": 8.183570128849085e-05, + "loss": 2.4218, + "step": 26240 + }, + { + "epoch": 1.12, + "grad_norm": 0.470703125, + "learning_rate": 8.18291049873342e-05, + "loss": 2.4745, + "step": 26241 + }, + { + "epoch": 1.12, + "grad_norm": 0.44921875, + "learning_rate": 8.18225087679399e-05, + "loss": 2.4251, + "step": 26242 + }, + { + "epoch": 1.12, + "grad_norm": 0.44140625, + "learning_rate": 8.181591263033772e-05, + "loss": 2.3902, + "step": 26243 + }, + { + "epoch": 1.12, + "grad_norm": 0.53515625, + "learning_rate": 8.180931657455727e-05, + "loss": 2.5723, + "step": 26244 + }, + { + "epoch": 1.12, + "grad_norm": 0.443359375, + "learning_rate": 8.180272060062827e-05, + "loss": 2.4259, + "step": 26245 + }, + { + "epoch": 1.12, + "grad_norm": 0.458984375, + "learning_rate": 8.179612470858031e-05, + "loss": 2.5347, + "step": 26246 + }, + { + "epoch": 1.12, + "grad_norm": 0.427734375, + "learning_rate": 8.178952889844322e-05, + "loss": 2.4537, + "step": 26247 + }, + { + "epoch": 1.12, + "grad_norm": 0.5, + "learning_rate": 8.178293317024654e-05, + "loss": 2.3712, + "step": 26248 + }, + { + "epoch": 1.12, + "grad_norm": 0.4296875, + "learning_rate": 8.177633752402003e-05, + "loss": 2.6133, + "step": 26249 + }, + { + "epoch": 1.12, + "grad_norm": 0.45703125, + "learning_rate": 8.176974195979331e-05, + "loss": 2.4699, + "step": 26250 + }, + { + "epoch": 1.12, + "grad_norm": 0.546875, + "learning_rate": 8.176314647759611e-05, + "loss": 2.4499, + "step": 26251 + }, + { + "epoch": 1.12, + "grad_norm": 0.453125, + "learning_rate": 8.175655107745809e-05, + "loss": 2.5814, + "step": 26252 + }, + { + "epoch": 1.12, + "grad_norm": 0.478515625, + "learning_rate": 8.174995575940893e-05, + "loss": 2.4183, + "step": 26253 + }, + { + "epoch": 1.12, + "grad_norm": 0.5, + "learning_rate": 8.174336052347825e-05, + "loss": 2.5129, + "step": 26254 + }, + { + "epoch": 1.12, + "grad_norm": 0.4921875, + "learning_rate": 8.173676536969583e-05, + "loss": 2.5251, + "step": 26255 + }, + { + "epoch": 1.12, + "grad_norm": 0.54296875, + "learning_rate": 8.17301702980913e-05, + "loss": 2.5656, + "step": 26256 + }, + { + "epoch": 1.12, + "grad_norm": 0.412109375, + "learning_rate": 8.172357530869428e-05, + "loss": 2.5723, + "step": 26257 + }, + { + "epoch": 1.12, + "grad_norm": 0.53125, + "learning_rate": 8.171698040153451e-05, + "loss": 2.4967, + "step": 26258 + }, + { + "epoch": 1.12, + "grad_norm": 0.5078125, + "learning_rate": 8.171038557664165e-05, + "loss": 2.4495, + "step": 26259 + }, + { + "epoch": 1.12, + "grad_norm": 0.478515625, + "learning_rate": 8.170379083404536e-05, + "loss": 2.4383, + "step": 26260 + }, + { + "epoch": 1.12, + "grad_norm": 0.439453125, + "learning_rate": 8.169719617377536e-05, + "loss": 2.4508, + "step": 26261 + }, + { + "epoch": 1.12, + "grad_norm": 0.443359375, + "learning_rate": 8.169060159586123e-05, + "loss": 2.4055, + "step": 26262 + }, + { + "epoch": 1.12, + "grad_norm": 0.494140625, + "learning_rate": 8.168400710033275e-05, + "loss": 2.5367, + "step": 26263 + }, + { + "epoch": 1.12, + "grad_norm": 0.47265625, + "learning_rate": 8.167741268721953e-05, + "loss": 2.5212, + "step": 26264 + }, + { + "epoch": 1.12, + "grad_norm": 0.48828125, + "learning_rate": 8.167081835655126e-05, + "loss": 2.5665, + "step": 26265 + }, + { + "epoch": 1.12, + "grad_norm": 0.4375, + "learning_rate": 8.166422410835759e-05, + "loss": 2.3407, + "step": 26266 + }, + { + "epoch": 1.12, + "grad_norm": 0.423828125, + "learning_rate": 8.165762994266823e-05, + "loss": 2.4774, + "step": 26267 + }, + { + "epoch": 1.12, + "grad_norm": 0.5234375, + "learning_rate": 8.165103585951282e-05, + "loss": 2.623, + "step": 26268 + }, + { + "epoch": 1.12, + "grad_norm": 0.51953125, + "learning_rate": 8.164444185892107e-05, + "loss": 2.4484, + "step": 26269 + }, + { + "epoch": 1.12, + "grad_norm": 0.4375, + "learning_rate": 8.163784794092257e-05, + "loss": 2.382, + "step": 26270 + }, + { + "epoch": 1.12, + "grad_norm": 0.47265625, + "learning_rate": 8.16312541055471e-05, + "loss": 2.3064, + "step": 26271 + }, + { + "epoch": 1.12, + "grad_norm": 0.46875, + "learning_rate": 8.162466035282425e-05, + "loss": 2.5886, + "step": 26272 + }, + { + "epoch": 1.12, + "grad_norm": 0.458984375, + "learning_rate": 8.161806668278372e-05, + "loss": 2.349, + "step": 26273 + }, + { + "epoch": 1.12, + "grad_norm": 0.486328125, + "learning_rate": 8.161147309545513e-05, + "loss": 2.53, + "step": 26274 + }, + { + "epoch": 1.12, + "grad_norm": 0.462890625, + "learning_rate": 8.160487959086824e-05, + "loss": 2.3971, + "step": 26275 + }, + { + "epoch": 1.12, + "grad_norm": 0.3984375, + "learning_rate": 8.159828616905266e-05, + "loss": 2.3342, + "step": 26276 + }, + { + "epoch": 1.12, + "grad_norm": 0.4375, + "learning_rate": 8.159169283003809e-05, + "loss": 2.4541, + "step": 26277 + }, + { + "epoch": 1.12, + "grad_norm": 0.404296875, + "learning_rate": 8.15850995738541e-05, + "loss": 2.3547, + "step": 26278 + }, + { + "epoch": 1.12, + "grad_norm": 0.470703125, + "learning_rate": 8.15785064005305e-05, + "loss": 2.5148, + "step": 26279 + }, + { + "epoch": 1.12, + "grad_norm": 0.5, + "learning_rate": 8.157191331009688e-05, + "loss": 2.4896, + "step": 26280 + }, + { + "epoch": 1.12, + "grad_norm": 0.4609375, + "learning_rate": 8.15653203025829e-05, + "loss": 2.4383, + "step": 26281 + }, + { + "epoch": 1.12, + "grad_norm": 0.515625, + "learning_rate": 8.155872737801823e-05, + "loss": 2.477, + "step": 26282 + }, + { + "epoch": 1.12, + "grad_norm": 0.51953125, + "learning_rate": 8.155213453643255e-05, + "loss": 2.4701, + "step": 26283 + }, + { + "epoch": 1.12, + "grad_norm": 0.51171875, + "learning_rate": 8.154554177785555e-05, + "loss": 2.4729, + "step": 26284 + }, + { + "epoch": 1.12, + "grad_norm": 0.49609375, + "learning_rate": 8.153894910231687e-05, + "loss": 2.3986, + "step": 26285 + }, + { + "epoch": 1.12, + "grad_norm": 0.462890625, + "learning_rate": 8.153235650984612e-05, + "loss": 2.4319, + "step": 26286 + }, + { + "epoch": 1.12, + "grad_norm": 0.4765625, + "learning_rate": 8.152576400047306e-05, + "loss": 2.5674, + "step": 26287 + }, + { + "epoch": 1.12, + "grad_norm": 0.458984375, + "learning_rate": 8.151917157422732e-05, + "loss": 2.4301, + "step": 26288 + }, + { + "epoch": 1.12, + "grad_norm": 0.4609375, + "learning_rate": 8.151257923113855e-05, + "loss": 2.5225, + "step": 26289 + }, + { + "epoch": 1.12, + "grad_norm": 0.47265625, + "learning_rate": 8.150598697123637e-05, + "loss": 2.3985, + "step": 26290 + }, + { + "epoch": 1.12, + "grad_norm": 0.5078125, + "learning_rate": 8.149939479455053e-05, + "loss": 2.3428, + "step": 26291 + }, + { + "epoch": 1.12, + "grad_norm": 0.44921875, + "learning_rate": 8.149280270111064e-05, + "loss": 2.4327, + "step": 26292 + }, + { + "epoch": 1.12, + "grad_norm": 0.51953125, + "learning_rate": 8.14862106909464e-05, + "loss": 2.5606, + "step": 26293 + }, + { + "epoch": 1.12, + "grad_norm": 0.453125, + "learning_rate": 8.14796187640874e-05, + "loss": 2.3888, + "step": 26294 + }, + { + "epoch": 1.12, + "grad_norm": 0.51953125, + "learning_rate": 8.147302692056339e-05, + "loss": 2.4292, + "step": 26295 + }, + { + "epoch": 1.12, + "grad_norm": 0.482421875, + "learning_rate": 8.146643516040398e-05, + "loss": 2.5173, + "step": 26296 + }, + { + "epoch": 1.12, + "grad_norm": 0.474609375, + "learning_rate": 8.145984348363882e-05, + "loss": 2.3737, + "step": 26297 + }, + { + "epoch": 1.12, + "grad_norm": 0.474609375, + "learning_rate": 8.145325189029758e-05, + "loss": 2.3286, + "step": 26298 + }, + { + "epoch": 1.12, + "grad_norm": 0.4453125, + "learning_rate": 8.144666038040996e-05, + "loss": 2.3743, + "step": 26299 + }, + { + "epoch": 1.12, + "grad_norm": 0.490234375, + "learning_rate": 8.144006895400557e-05, + "loss": 2.4736, + "step": 26300 + }, + { + "epoch": 1.12, + "grad_norm": 0.458984375, + "learning_rate": 8.14334776111141e-05, + "loss": 2.5077, + "step": 26301 + }, + { + "epoch": 1.12, + "grad_norm": 0.478515625, + "learning_rate": 8.142688635176515e-05, + "loss": 2.397, + "step": 26302 + }, + { + "epoch": 1.12, + "grad_norm": 0.46484375, + "learning_rate": 8.142029517598849e-05, + "loss": 2.3952, + "step": 26303 + }, + { + "epoch": 1.12, + "grad_norm": 0.41796875, + "learning_rate": 8.141370408381367e-05, + "loss": 2.3557, + "step": 26304 + }, + { + "epoch": 1.12, + "grad_norm": 0.46484375, + "learning_rate": 8.140711307527041e-05, + "loss": 2.3177, + "step": 26305 + }, + { + "epoch": 1.12, + "grad_norm": 0.46484375, + "learning_rate": 8.140052215038833e-05, + "loss": 2.5509, + "step": 26306 + }, + { + "epoch": 1.12, + "grad_norm": 0.546875, + "learning_rate": 8.13939313091971e-05, + "loss": 2.4157, + "step": 26307 + }, + { + "epoch": 1.12, + "grad_norm": 0.51171875, + "learning_rate": 8.13873405517264e-05, + "loss": 2.533, + "step": 26308 + }, + { + "epoch": 1.12, + "grad_norm": 0.5, + "learning_rate": 8.138074987800588e-05, + "loss": 2.565, + "step": 26309 + }, + { + "epoch": 1.12, + "grad_norm": 0.53125, + "learning_rate": 8.137415928806516e-05, + "loss": 2.4768, + "step": 26310 + }, + { + "epoch": 1.12, + "grad_norm": 0.48046875, + "learning_rate": 8.136756878193389e-05, + "loss": 2.4219, + "step": 26311 + }, + { + "epoch": 1.12, + "grad_norm": 0.5078125, + "learning_rate": 8.136097835964178e-05, + "loss": 2.6351, + "step": 26312 + }, + { + "epoch": 1.12, + "grad_norm": 0.4453125, + "learning_rate": 8.135438802121845e-05, + "loss": 2.4539, + "step": 26313 + }, + { + "epoch": 1.12, + "grad_norm": 0.470703125, + "learning_rate": 8.134779776669356e-05, + "loss": 2.3783, + "step": 26314 + }, + { + "epoch": 1.12, + "grad_norm": 0.5078125, + "learning_rate": 8.134120759609673e-05, + "loss": 2.4592, + "step": 26315 + }, + { + "epoch": 1.12, + "grad_norm": 0.4296875, + "learning_rate": 8.133461750945769e-05, + "loss": 2.4906, + "step": 26316 + }, + { + "epoch": 1.12, + "grad_norm": 0.515625, + "learning_rate": 8.132802750680606e-05, + "loss": 2.5026, + "step": 26317 + }, + { + "epoch": 1.12, + "grad_norm": 0.435546875, + "learning_rate": 8.132143758817146e-05, + "loss": 2.4379, + "step": 26318 + }, + { + "epoch": 1.12, + "grad_norm": 0.46875, + "learning_rate": 8.131484775358355e-05, + "loss": 2.4036, + "step": 26319 + }, + { + "epoch": 1.12, + "grad_norm": 0.46875, + "learning_rate": 8.130825800307202e-05, + "loss": 2.4681, + "step": 26320 + }, + { + "epoch": 1.12, + "grad_norm": 0.470703125, + "learning_rate": 8.130166833666649e-05, + "loss": 2.469, + "step": 26321 + }, + { + "epoch": 1.12, + "grad_norm": 0.53125, + "learning_rate": 8.129507875439661e-05, + "loss": 2.624, + "step": 26322 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.128848925629203e-05, + "loss": 2.5388, + "step": 26323 + }, + { + "epoch": 1.12, + "grad_norm": 0.482421875, + "learning_rate": 8.128189984238243e-05, + "loss": 2.5234, + "step": 26324 + }, + { + "epoch": 1.12, + "grad_norm": 0.443359375, + "learning_rate": 8.127531051269746e-05, + "loss": 2.3936, + "step": 26325 + }, + { + "epoch": 1.12, + "grad_norm": 0.4609375, + "learning_rate": 8.126872126726675e-05, + "loss": 2.501, + "step": 26326 + }, + { + "epoch": 1.12, + "grad_norm": 0.48046875, + "learning_rate": 8.126213210611989e-05, + "loss": 2.3966, + "step": 26327 + }, + { + "epoch": 1.12, + "grad_norm": 0.455078125, + "learning_rate": 8.125554302928664e-05, + "loss": 2.376, + "step": 26328 + }, + { + "epoch": 1.12, + "grad_norm": 0.52734375, + "learning_rate": 8.124895403679658e-05, + "loss": 2.5237, + "step": 26329 + }, + { + "epoch": 1.12, + "grad_norm": 0.45703125, + "learning_rate": 8.124236512867937e-05, + "loss": 2.512, + "step": 26330 + }, + { + "epoch": 1.12, + "grad_norm": 0.443359375, + "learning_rate": 8.123577630496465e-05, + "loss": 2.3797, + "step": 26331 + }, + { + "epoch": 1.12, + "grad_norm": 0.423828125, + "learning_rate": 8.122918756568209e-05, + "loss": 2.4226, + "step": 26332 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.122259891086136e-05, + "loss": 2.4425, + "step": 26333 + }, + { + "epoch": 1.12, + "grad_norm": 0.494140625, + "learning_rate": 8.121601034053205e-05, + "loss": 2.5276, + "step": 26334 + }, + { + "epoch": 1.12, + "grad_norm": 0.482421875, + "learning_rate": 8.12094218547238e-05, + "loss": 2.4155, + "step": 26335 + }, + { + "epoch": 1.12, + "grad_norm": 0.427734375, + "learning_rate": 8.12028334534663e-05, + "loss": 2.5395, + "step": 26336 + }, + { + "epoch": 1.12, + "grad_norm": 0.42578125, + "learning_rate": 8.11962451367892e-05, + "loss": 2.4831, + "step": 26337 + }, + { + "epoch": 1.12, + "grad_norm": 0.46875, + "learning_rate": 8.118965690472212e-05, + "loss": 2.4937, + "step": 26338 + }, + { + "epoch": 1.12, + "grad_norm": 0.4609375, + "learning_rate": 8.118306875729468e-05, + "loss": 2.4231, + "step": 26339 + }, + { + "epoch": 1.12, + "grad_norm": 0.49609375, + "learning_rate": 8.117648069453658e-05, + "loss": 2.4812, + "step": 26340 + }, + { + "epoch": 1.12, + "grad_norm": 0.48828125, + "learning_rate": 8.116989271647746e-05, + "loss": 2.5007, + "step": 26341 + }, + { + "epoch": 1.12, + "grad_norm": 0.4375, + "learning_rate": 8.116330482314692e-05, + "loss": 2.3901, + "step": 26342 + }, + { + "epoch": 1.12, + "grad_norm": 0.515625, + "learning_rate": 8.115671701457457e-05, + "loss": 2.5726, + "step": 26343 + }, + { + "epoch": 1.12, + "grad_norm": 0.5, + "learning_rate": 8.115012929079017e-05, + "loss": 2.33, + "step": 26344 + }, + { + "epoch": 1.12, + "grad_norm": 0.51171875, + "learning_rate": 8.11435416518233e-05, + "loss": 2.5284, + "step": 26345 + }, + { + "epoch": 1.12, + "grad_norm": 0.43359375, + "learning_rate": 8.113695409770358e-05, + "loss": 2.4332, + "step": 26346 + }, + { + "epoch": 1.12, + "grad_norm": 0.44140625, + "learning_rate": 8.113036662846067e-05, + "loss": 2.4824, + "step": 26347 + }, + { + "epoch": 1.12, + "grad_norm": 0.44921875, + "learning_rate": 8.112377924412423e-05, + "loss": 2.3282, + "step": 26348 + }, + { + "epoch": 1.12, + "grad_norm": 0.478515625, + "learning_rate": 8.111719194472388e-05, + "loss": 2.4965, + "step": 26349 + }, + { + "epoch": 1.12, + "grad_norm": 0.455078125, + "learning_rate": 8.111060473028927e-05, + "loss": 2.3738, + "step": 26350 + }, + { + "epoch": 1.12, + "grad_norm": 0.40625, + "learning_rate": 8.110401760085e-05, + "loss": 2.525, + "step": 26351 + }, + { + "epoch": 1.12, + "grad_norm": 0.4375, + "learning_rate": 8.10974305564358e-05, + "loss": 2.5772, + "step": 26352 + }, + { + "epoch": 1.12, + "grad_norm": 0.4921875, + "learning_rate": 8.109084359707623e-05, + "loss": 2.5124, + "step": 26353 + }, + { + "epoch": 1.12, + "grad_norm": 0.474609375, + "learning_rate": 8.108425672280096e-05, + "loss": 2.3508, + "step": 26354 + }, + { + "epoch": 1.12, + "grad_norm": 0.48828125, + "learning_rate": 8.107766993363959e-05, + "loss": 2.455, + "step": 26355 + }, + { + "epoch": 1.12, + "grad_norm": 0.52734375, + "learning_rate": 8.107108322962181e-05, + "loss": 2.5302, + "step": 26356 + }, + { + "epoch": 1.12, + "grad_norm": 0.466796875, + "learning_rate": 8.106449661077726e-05, + "loss": 2.4702, + "step": 26357 + }, + { + "epoch": 1.12, + "grad_norm": 0.482421875, + "learning_rate": 8.105791007713555e-05, + "loss": 2.4831, + "step": 26358 + }, + { + "epoch": 1.12, + "grad_norm": 0.455078125, + "learning_rate": 8.105132362872626e-05, + "loss": 2.514, + "step": 26359 + }, + { + "epoch": 1.12, + "grad_norm": 0.5, + "learning_rate": 8.104473726557916e-05, + "loss": 2.4955, + "step": 26360 + }, + { + "epoch": 1.12, + "grad_norm": 0.50390625, + "learning_rate": 8.10381509877238e-05, + "loss": 2.556, + "step": 26361 + }, + { + "epoch": 1.12, + "grad_norm": 0.412109375, + "learning_rate": 8.103156479518983e-05, + "loss": 2.4278, + "step": 26362 + }, + { + "epoch": 1.12, + "grad_norm": 0.375, + "learning_rate": 8.102497868800685e-05, + "loss": 2.3637, + "step": 26363 + }, + { + "epoch": 1.12, + "grad_norm": 0.42578125, + "learning_rate": 8.101839266620457e-05, + "loss": 2.372, + "step": 26364 + }, + { + "epoch": 1.12, + "grad_norm": 0.447265625, + "learning_rate": 8.10118067298126e-05, + "loss": 2.3708, + "step": 26365 + }, + { + "epoch": 1.12, + "grad_norm": 0.470703125, + "learning_rate": 8.100522087886053e-05, + "loss": 2.3762, + "step": 26366 + }, + { + "epoch": 1.12, + "grad_norm": 0.50390625, + "learning_rate": 8.0998635113378e-05, + "loss": 2.3836, + "step": 26367 + }, + { + "epoch": 1.12, + "grad_norm": 0.43359375, + "learning_rate": 8.099204943339471e-05, + "loss": 2.4771, + "step": 26368 + }, + { + "epoch": 1.12, + "grad_norm": 0.50390625, + "learning_rate": 8.098546383894025e-05, + "loss": 2.4834, + "step": 26369 + }, + { + "epoch": 1.12, + "grad_norm": 0.5234375, + "learning_rate": 8.097887833004424e-05, + "loss": 2.47, + "step": 26370 + }, + { + "epoch": 1.12, + "grad_norm": 0.5078125, + "learning_rate": 8.097229290673631e-05, + "loss": 2.5405, + "step": 26371 + }, + { + "epoch": 1.12, + "grad_norm": 0.447265625, + "learning_rate": 8.096570756904613e-05, + "loss": 2.4095, + "step": 26372 + }, + { + "epoch": 1.12, + "grad_norm": 0.44140625, + "learning_rate": 8.095912231700334e-05, + "loss": 2.5502, + "step": 26373 + }, + { + "epoch": 1.12, + "grad_norm": 0.5, + "learning_rate": 8.095253715063751e-05, + "loss": 2.4205, + "step": 26374 + }, + { + "epoch": 1.12, + "grad_norm": 0.49609375, + "learning_rate": 8.094595206997825e-05, + "loss": 2.518, + "step": 26375 + }, + { + "epoch": 1.12, + "grad_norm": 0.53125, + "learning_rate": 8.093936707505532e-05, + "loss": 2.4229, + "step": 26376 + }, + { + "epoch": 1.12, + "grad_norm": 0.4296875, + "learning_rate": 8.093278216589826e-05, + "loss": 2.4938, + "step": 26377 + }, + { + "epoch": 1.12, + "grad_norm": 0.44140625, + "learning_rate": 8.09261973425367e-05, + "loss": 2.3024, + "step": 26378 + }, + { + "epoch": 1.12, + "grad_norm": 0.447265625, + "learning_rate": 8.091961260500024e-05, + "loss": 2.4803, + "step": 26379 + }, + { + "epoch": 1.12, + "grad_norm": 0.515625, + "learning_rate": 8.09130279533186e-05, + "loss": 2.4985, + "step": 26380 + }, + { + "epoch": 1.12, + "grad_norm": 0.41796875, + "learning_rate": 8.090644338752137e-05, + "loss": 2.4423, + "step": 26381 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.089985890763815e-05, + "loss": 2.4294, + "step": 26382 + }, + { + "epoch": 1.12, + "grad_norm": 0.44921875, + "learning_rate": 8.089327451369854e-05, + "loss": 2.4106, + "step": 26383 + }, + { + "epoch": 1.12, + "grad_norm": 0.46484375, + "learning_rate": 8.088669020573227e-05, + "loss": 2.5919, + "step": 26384 + }, + { + "epoch": 1.12, + "grad_norm": 0.44140625, + "learning_rate": 8.088010598376889e-05, + "loss": 2.4249, + "step": 26385 + }, + { + "epoch": 1.12, + "grad_norm": 0.431640625, + "learning_rate": 8.087352184783805e-05, + "loss": 2.5333, + "step": 26386 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.086693779796935e-05, + "loss": 2.4007, + "step": 26387 + }, + { + "epoch": 1.12, + "grad_norm": 0.462890625, + "learning_rate": 8.086035383419245e-05, + "loss": 2.5411, + "step": 26388 + }, + { + "epoch": 1.12, + "grad_norm": 0.44921875, + "learning_rate": 8.0853769956537e-05, + "loss": 2.4678, + "step": 26389 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.084718616503255e-05, + "loss": 2.523, + "step": 26390 + }, + { + "epoch": 1.12, + "grad_norm": 0.40234375, + "learning_rate": 8.084060245970873e-05, + "loss": 2.4186, + "step": 26391 + }, + { + "epoch": 1.12, + "grad_norm": 0.462890625, + "learning_rate": 8.083401884059525e-05, + "loss": 2.5224, + "step": 26392 + }, + { + "epoch": 1.12, + "grad_norm": 0.53515625, + "learning_rate": 8.082743530772168e-05, + "loss": 2.4244, + "step": 26393 + }, + { + "epoch": 1.12, + "grad_norm": 0.46484375, + "learning_rate": 8.082085186111763e-05, + "loss": 2.5716, + "step": 26394 + }, + { + "epoch": 1.12, + "grad_norm": 0.482421875, + "learning_rate": 8.081426850081272e-05, + "loss": 2.4857, + "step": 26395 + }, + { + "epoch": 1.12, + "grad_norm": 0.56640625, + "learning_rate": 8.08076852268366e-05, + "loss": 2.4347, + "step": 26396 + }, + { + "epoch": 1.12, + "grad_norm": 0.4609375, + "learning_rate": 8.08011020392189e-05, + "loss": 2.5134, + "step": 26397 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.079451893798923e-05, + "loss": 2.4055, + "step": 26398 + }, + { + "epoch": 1.12, + "grad_norm": 0.484375, + "learning_rate": 8.078793592317716e-05, + "loss": 2.4828, + "step": 26399 + }, + { + "epoch": 1.12, + "grad_norm": 0.44140625, + "learning_rate": 8.078135299481241e-05, + "loss": 2.5127, + "step": 26400 + }, + { + "epoch": 1.13, + "grad_norm": 0.439453125, + "learning_rate": 8.077477015292453e-05, + "loss": 2.6201, + "step": 26401 + }, + { + "epoch": 1.13, + "grad_norm": 0.451171875, + "learning_rate": 8.076818739754314e-05, + "loss": 2.5111, + "step": 26402 + }, + { + "epoch": 1.13, + "grad_norm": 0.46875, + "learning_rate": 8.076160472869788e-05, + "loss": 2.4744, + "step": 26403 + }, + { + "epoch": 1.13, + "grad_norm": 0.486328125, + "learning_rate": 8.075502214641839e-05, + "loss": 2.4916, + "step": 26404 + }, + { + "epoch": 1.13, + "grad_norm": 0.447265625, + "learning_rate": 8.074843965073427e-05, + "loss": 2.5182, + "step": 26405 + }, + { + "epoch": 1.13, + "grad_norm": 0.5234375, + "learning_rate": 8.074185724167514e-05, + "loss": 2.4391, + "step": 26406 + }, + { + "epoch": 1.13, + "grad_norm": 0.5546875, + "learning_rate": 8.073527491927055e-05, + "loss": 2.5591, + "step": 26407 + }, + { + "epoch": 1.13, + "grad_norm": 0.46875, + "learning_rate": 8.072869268355025e-05, + "loss": 2.3773, + "step": 26408 + }, + { + "epoch": 1.13, + "grad_norm": 0.470703125, + "learning_rate": 8.072211053454379e-05, + "loss": 2.3715, + "step": 26409 + }, + { + "epoch": 1.13, + "grad_norm": 0.494140625, + "learning_rate": 8.071552847228075e-05, + "loss": 2.6158, + "step": 26410 + }, + { + "epoch": 1.13, + "grad_norm": 0.50390625, + "learning_rate": 8.070894649679079e-05, + "loss": 2.387, + "step": 26411 + }, + { + "epoch": 1.13, + "grad_norm": 0.470703125, + "learning_rate": 8.070236460810353e-05, + "loss": 2.3581, + "step": 26412 + }, + { + "epoch": 1.13, + "grad_norm": 0.50390625, + "learning_rate": 8.069578280624859e-05, + "loss": 2.6333, + "step": 26413 + }, + { + "epoch": 1.13, + "grad_norm": 0.44140625, + "learning_rate": 8.068920109125559e-05, + "loss": 2.4441, + "step": 26414 + }, + { + "epoch": 1.13, + "grad_norm": 0.44921875, + "learning_rate": 8.068261946315406e-05, + "loss": 2.4803, + "step": 26415 + }, + { + "epoch": 1.13, + "grad_norm": 0.4609375, + "learning_rate": 8.067603792197373e-05, + "loss": 2.5008, + "step": 26416 + }, + { + "epoch": 1.13, + "grad_norm": 0.484375, + "learning_rate": 8.066945646774416e-05, + "loss": 2.4481, + "step": 26417 + }, + { + "epoch": 1.13, + "grad_norm": 0.443359375, + "learning_rate": 8.066287510049498e-05, + "loss": 2.4026, + "step": 26418 + }, + { + "epoch": 1.13, + "grad_norm": 0.55859375, + "learning_rate": 8.065629382025577e-05, + "loss": 2.5505, + "step": 26419 + }, + { + "epoch": 1.13, + "grad_norm": 0.48828125, + "learning_rate": 8.064971262705618e-05, + "loss": 2.5073, + "step": 26420 + }, + { + "epoch": 1.13, + "grad_norm": 0.54296875, + "learning_rate": 8.064313152092582e-05, + "loss": 2.4376, + "step": 26421 + }, + { + "epoch": 1.13, + "grad_norm": 0.466796875, + "learning_rate": 8.063655050189432e-05, + "loss": 2.5662, + "step": 26422 + }, + { + "epoch": 1.13, + "grad_norm": 0.484375, + "learning_rate": 8.062996956999119e-05, + "loss": 2.4514, + "step": 26423 + }, + { + "epoch": 1.13, + "grad_norm": 0.484375, + "learning_rate": 8.062338872524618e-05, + "loss": 2.463, + "step": 26424 + }, + { + "epoch": 1.13, + "grad_norm": 0.515625, + "learning_rate": 8.061680796768884e-05, + "loss": 2.4225, + "step": 26425 + }, + { + "epoch": 1.13, + "grad_norm": 0.4453125, + "learning_rate": 8.061022729734876e-05, + "loss": 2.4272, + "step": 26426 + }, + { + "epoch": 1.13, + "grad_norm": 0.59765625, + "learning_rate": 8.060364671425555e-05, + "loss": 2.5594, + "step": 26427 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 8.059706621843888e-05, + "loss": 2.4833, + "step": 26428 + }, + { + "epoch": 1.13, + "grad_norm": 0.431640625, + "learning_rate": 8.059048580992831e-05, + "loss": 2.4667, + "step": 26429 + }, + { + "epoch": 1.13, + "grad_norm": 0.453125, + "learning_rate": 8.058390548875347e-05, + "loss": 2.53, + "step": 26430 + }, + { + "epoch": 1.13, + "grad_norm": 0.4296875, + "learning_rate": 8.057732525494396e-05, + "loss": 2.3297, + "step": 26431 + }, + { + "epoch": 1.13, + "grad_norm": 0.5, + "learning_rate": 8.057074510852936e-05, + "loss": 2.4563, + "step": 26432 + }, + { + "epoch": 1.13, + "grad_norm": 0.515625, + "learning_rate": 8.056416504953933e-05, + "loss": 2.5382, + "step": 26433 + }, + { + "epoch": 1.13, + "grad_norm": 0.52734375, + "learning_rate": 8.055758507800344e-05, + "loss": 2.4893, + "step": 26434 + }, + { + "epoch": 1.13, + "grad_norm": 0.4375, + "learning_rate": 8.055100519395133e-05, + "loss": 2.4374, + "step": 26435 + }, + { + "epoch": 1.13, + "grad_norm": 0.470703125, + "learning_rate": 8.054442539741253e-05, + "loss": 2.5545, + "step": 26436 + }, + { + "epoch": 1.13, + "grad_norm": 0.51171875, + "learning_rate": 8.053784568841677e-05, + "loss": 2.5608, + "step": 26437 + }, + { + "epoch": 1.13, + "grad_norm": 0.42578125, + "learning_rate": 8.05312660669936e-05, + "loss": 2.3504, + "step": 26438 + }, + { + "epoch": 1.13, + "grad_norm": 0.51953125, + "learning_rate": 8.052468653317259e-05, + "loss": 2.6233, + "step": 26439 + }, + { + "epoch": 1.13, + "grad_norm": 0.443359375, + "learning_rate": 8.051810708698335e-05, + "loss": 2.4763, + "step": 26440 + }, + { + "epoch": 1.13, + "grad_norm": 0.435546875, + "learning_rate": 8.051152772845554e-05, + "loss": 2.5466, + "step": 26441 + }, + { + "epoch": 1.13, + "grad_norm": 0.4375, + "learning_rate": 8.050494845761873e-05, + "loss": 2.6096, + "step": 26442 + }, + { + "epoch": 1.13, + "grad_norm": 0.439453125, + "learning_rate": 8.049836927450254e-05, + "loss": 2.3066, + "step": 26443 + }, + { + "epoch": 1.13, + "grad_norm": 0.49609375, + "learning_rate": 8.04917901791365e-05, + "loss": 2.6468, + "step": 26444 + }, + { + "epoch": 1.13, + "grad_norm": 0.51953125, + "learning_rate": 8.048521117155032e-05, + "loss": 2.456, + "step": 26445 + }, + { + "epoch": 1.13, + "grad_norm": 0.44921875, + "learning_rate": 8.047863225177358e-05, + "loss": 2.3979, + "step": 26446 + }, + { + "epoch": 1.13, + "grad_norm": 0.427734375, + "learning_rate": 8.047205341983585e-05, + "loss": 2.4194, + "step": 26447 + }, + { + "epoch": 1.13, + "grad_norm": 0.47265625, + "learning_rate": 8.046547467576672e-05, + "loss": 2.4581, + "step": 26448 + }, + { + "epoch": 1.13, + "grad_norm": 0.478515625, + "learning_rate": 8.045889601959582e-05, + "loss": 2.4311, + "step": 26449 + }, + { + "epoch": 1.13, + "grad_norm": 0.462890625, + "learning_rate": 8.045231745135276e-05, + "loss": 2.4997, + "step": 26450 + }, + { + "epoch": 1.13, + "grad_norm": 0.45703125, + "learning_rate": 8.044573897106712e-05, + "loss": 2.4726, + "step": 26451 + }, + { + "epoch": 1.13, + "grad_norm": 0.400390625, + "learning_rate": 8.04391605787685e-05, + "loss": 2.3325, + "step": 26452 + }, + { + "epoch": 1.13, + "grad_norm": 0.443359375, + "learning_rate": 8.043258227448651e-05, + "loss": 2.6166, + "step": 26453 + }, + { + "epoch": 1.13, + "grad_norm": 0.458984375, + "learning_rate": 8.042600405825078e-05, + "loss": 2.4207, + "step": 26454 + }, + { + "epoch": 1.13, + "grad_norm": 0.4296875, + "learning_rate": 8.041942593009086e-05, + "loss": 2.3716, + "step": 26455 + }, + { + "epoch": 1.13, + "grad_norm": 0.447265625, + "learning_rate": 8.041284789003634e-05, + "loss": 2.4881, + "step": 26456 + }, + { + "epoch": 1.13, + "grad_norm": 0.4296875, + "learning_rate": 8.040626993811686e-05, + "loss": 2.5176, + "step": 26457 + }, + { + "epoch": 1.13, + "grad_norm": 0.447265625, + "learning_rate": 8.039969207436203e-05, + "loss": 2.3939, + "step": 26458 + }, + { + "epoch": 1.13, + "grad_norm": 0.52734375, + "learning_rate": 8.03931142988014e-05, + "loss": 2.5906, + "step": 26459 + }, + { + "epoch": 1.13, + "grad_norm": 0.490234375, + "learning_rate": 8.038653661146457e-05, + "loss": 2.4043, + "step": 26460 + }, + { + "epoch": 1.13, + "grad_norm": 0.458984375, + "learning_rate": 8.037995901238118e-05, + "loss": 2.4551, + "step": 26461 + }, + { + "epoch": 1.13, + "grad_norm": 0.4375, + "learning_rate": 8.037338150158084e-05, + "loss": 2.5177, + "step": 26462 + }, + { + "epoch": 1.13, + "grad_norm": 0.53125, + "learning_rate": 8.036680407909307e-05, + "loss": 2.396, + "step": 26463 + }, + { + "epoch": 1.13, + "grad_norm": 0.53125, + "learning_rate": 8.03602267449475e-05, + "loss": 2.5349, + "step": 26464 + }, + { + "epoch": 1.13, + "grad_norm": 0.53125, + "learning_rate": 8.035364949917374e-05, + "loss": 2.5388, + "step": 26465 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 8.03470723418014e-05, + "loss": 2.3721, + "step": 26466 + }, + { + "epoch": 1.13, + "grad_norm": 0.50390625, + "learning_rate": 8.034049527286003e-05, + "loss": 2.3966, + "step": 26467 + }, + { + "epoch": 1.13, + "grad_norm": 0.498046875, + "learning_rate": 8.033391829237924e-05, + "loss": 2.4843, + "step": 26468 + }, + { + "epoch": 1.13, + "grad_norm": 0.4375, + "learning_rate": 8.032734140038864e-05, + "loss": 2.4931, + "step": 26469 + }, + { + "epoch": 1.13, + "grad_norm": 0.5234375, + "learning_rate": 8.032076459691785e-05, + "loss": 2.5916, + "step": 26470 + }, + { + "epoch": 1.13, + "grad_norm": 0.5, + "learning_rate": 8.031418788199638e-05, + "loss": 2.3804, + "step": 26471 + }, + { + "epoch": 1.13, + "grad_norm": 0.42578125, + "learning_rate": 8.030761125565387e-05, + "loss": 2.4856, + "step": 26472 + }, + { + "epoch": 1.13, + "grad_norm": 0.466796875, + "learning_rate": 8.030103471791992e-05, + "loss": 2.3882, + "step": 26473 + }, + { + "epoch": 1.13, + "grad_norm": 0.498046875, + "learning_rate": 8.029445826882413e-05, + "loss": 2.4366, + "step": 26474 + }, + { + "epoch": 1.13, + "grad_norm": 0.46484375, + "learning_rate": 8.028788190839607e-05, + "loss": 2.4206, + "step": 26475 + }, + { + "epoch": 1.13, + "grad_norm": 0.48046875, + "learning_rate": 8.028130563666532e-05, + "loss": 2.3755, + "step": 26476 + }, + { + "epoch": 1.13, + "grad_norm": 0.49609375, + "learning_rate": 8.027472945366151e-05, + "loss": 2.3753, + "step": 26477 + }, + { + "epoch": 1.13, + "grad_norm": 0.45703125, + "learning_rate": 8.026815335941422e-05, + "loss": 2.4047, + "step": 26478 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 8.026157735395301e-05, + "loss": 2.6322, + "step": 26479 + }, + { + "epoch": 1.13, + "grad_norm": 0.5234375, + "learning_rate": 8.025500143730746e-05, + "loss": 2.5787, + "step": 26480 + }, + { + "epoch": 1.13, + "grad_norm": 0.44921875, + "learning_rate": 8.024842560950723e-05, + "loss": 2.3725, + "step": 26481 + }, + { + "epoch": 1.13, + "grad_norm": 0.478515625, + "learning_rate": 8.024184987058185e-05, + "loss": 2.6211, + "step": 26482 + }, + { + "epoch": 1.13, + "grad_norm": 0.45703125, + "learning_rate": 8.023527422056092e-05, + "loss": 2.4812, + "step": 26483 + }, + { + "epoch": 1.13, + "grad_norm": 0.4453125, + "learning_rate": 8.022869865947403e-05, + "loss": 2.4983, + "step": 26484 + }, + { + "epoch": 1.13, + "grad_norm": 0.455078125, + "learning_rate": 8.022212318735079e-05, + "loss": 2.5463, + "step": 26485 + }, + { + "epoch": 1.13, + "grad_norm": 0.4765625, + "learning_rate": 8.021554780422078e-05, + "loss": 2.5784, + "step": 26486 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 8.020897251011355e-05, + "loss": 2.3408, + "step": 26487 + }, + { + "epoch": 1.13, + "grad_norm": 0.46875, + "learning_rate": 8.02023973050587e-05, + "loss": 2.4358, + "step": 26488 + }, + { + "epoch": 1.13, + "grad_norm": 0.55859375, + "learning_rate": 8.019582218908584e-05, + "loss": 2.5226, + "step": 26489 + }, + { + "epoch": 1.13, + "grad_norm": 0.451171875, + "learning_rate": 8.018924716222454e-05, + "loss": 2.3268, + "step": 26490 + }, + { + "epoch": 1.13, + "grad_norm": 0.51171875, + "learning_rate": 8.01826722245044e-05, + "loss": 2.4505, + "step": 26491 + }, + { + "epoch": 1.13, + "grad_norm": 0.466796875, + "learning_rate": 8.017609737595496e-05, + "loss": 2.4341, + "step": 26492 + }, + { + "epoch": 1.13, + "grad_norm": 0.46484375, + "learning_rate": 8.016952261660586e-05, + "loss": 2.407, + "step": 26493 + }, + { + "epoch": 1.13, + "grad_norm": 0.50390625, + "learning_rate": 8.016294794648668e-05, + "loss": 2.528, + "step": 26494 + }, + { + "epoch": 1.13, + "grad_norm": 0.4609375, + "learning_rate": 8.015637336562699e-05, + "loss": 2.5527, + "step": 26495 + }, + { + "epoch": 1.13, + "grad_norm": 0.515625, + "learning_rate": 8.014979887405631e-05, + "loss": 2.4139, + "step": 26496 + }, + { + "epoch": 1.13, + "grad_norm": 0.46484375, + "learning_rate": 8.014322447180433e-05, + "loss": 2.5258, + "step": 26497 + }, + { + "epoch": 1.13, + "grad_norm": 0.490234375, + "learning_rate": 8.013665015890056e-05, + "loss": 2.4714, + "step": 26498 + }, + { + "epoch": 1.13, + "grad_norm": 0.48828125, + "learning_rate": 8.013007593537463e-05, + "loss": 2.4773, + "step": 26499 + }, + { + "epoch": 1.13, + "grad_norm": 0.458984375, + "learning_rate": 8.012350180125606e-05, + "loss": 2.6851, + "step": 26500 + }, + { + "epoch": 1.13, + "grad_norm": 0.48046875, + "learning_rate": 8.011692775657448e-05, + "loss": 2.3968, + "step": 26501 + }, + { + "epoch": 1.13, + "grad_norm": 0.453125, + "learning_rate": 8.01103538013595e-05, + "loss": 2.508, + "step": 26502 + }, + { + "epoch": 1.13, + "grad_norm": 0.453125, + "learning_rate": 8.010377993564062e-05, + "loss": 2.5402, + "step": 26503 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 8.009720615944745e-05, + "loss": 2.563, + "step": 26504 + }, + { + "epoch": 1.13, + "grad_norm": 0.48828125, + "learning_rate": 8.009063247280959e-05, + "loss": 2.4608, + "step": 26505 + }, + { + "epoch": 1.13, + "grad_norm": 0.482421875, + "learning_rate": 8.00840588757566e-05, + "loss": 2.5212, + "step": 26506 + }, + { + "epoch": 1.13, + "grad_norm": 0.55859375, + "learning_rate": 8.00774853683181e-05, + "loss": 2.5788, + "step": 26507 + }, + { + "epoch": 1.13, + "grad_norm": 0.4921875, + "learning_rate": 8.007091195052358e-05, + "loss": 2.3177, + "step": 26508 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 8.006433862240271e-05, + "loss": 2.3555, + "step": 26509 + }, + { + "epoch": 1.13, + "grad_norm": 0.466796875, + "learning_rate": 8.005776538398506e-05, + "loss": 2.4822, + "step": 26510 + }, + { + "epoch": 1.13, + "grad_norm": 0.5, + "learning_rate": 8.005119223530013e-05, + "loss": 2.4692, + "step": 26511 + }, + { + "epoch": 1.13, + "grad_norm": 0.423828125, + "learning_rate": 8.004461917637755e-05, + "loss": 2.4065, + "step": 26512 + }, + { + "epoch": 1.13, + "grad_norm": 0.466796875, + "learning_rate": 8.00380462072469e-05, + "loss": 2.4925, + "step": 26513 + }, + { + "epoch": 1.13, + "grad_norm": 0.45703125, + "learning_rate": 8.003147332793775e-05, + "loss": 2.4141, + "step": 26514 + }, + { + "epoch": 1.13, + "grad_norm": 0.47265625, + "learning_rate": 8.002490053847966e-05, + "loss": 2.4148, + "step": 26515 + }, + { + "epoch": 1.13, + "grad_norm": 0.484375, + "learning_rate": 8.001832783890222e-05, + "loss": 2.3522, + "step": 26516 + }, + { + "epoch": 1.13, + "grad_norm": 0.494140625, + "learning_rate": 8.0011755229235e-05, + "loss": 2.4524, + "step": 26517 + }, + { + "epoch": 1.13, + "grad_norm": 0.44921875, + "learning_rate": 8.00051827095076e-05, + "loss": 2.5169, + "step": 26518 + }, + { + "epoch": 1.13, + "grad_norm": 0.48828125, + "learning_rate": 7.999861027974958e-05, + "loss": 2.3014, + "step": 26519 + }, + { + "epoch": 1.13, + "grad_norm": 0.482421875, + "learning_rate": 7.999203793999045e-05, + "loss": 2.457, + "step": 26520 + }, + { + "epoch": 1.13, + "grad_norm": 0.4375, + "learning_rate": 7.998546569025986e-05, + "loss": 2.5637, + "step": 26521 + }, + { + "epoch": 1.13, + "grad_norm": 0.470703125, + "learning_rate": 7.997889353058737e-05, + "loss": 2.4893, + "step": 26522 + }, + { + "epoch": 1.13, + "grad_norm": 0.474609375, + "learning_rate": 7.997232146100255e-05, + "loss": 2.496, + "step": 26523 + }, + { + "epoch": 1.13, + "grad_norm": 0.484375, + "learning_rate": 7.996574948153492e-05, + "loss": 2.4259, + "step": 26524 + }, + { + "epoch": 1.13, + "grad_norm": 0.484375, + "learning_rate": 7.995917759221414e-05, + "loss": 2.5004, + "step": 26525 + }, + { + "epoch": 1.13, + "grad_norm": 0.51953125, + "learning_rate": 7.995260579306976e-05, + "loss": 2.5069, + "step": 26526 + }, + { + "epoch": 1.13, + "grad_norm": 0.4140625, + "learning_rate": 7.994603408413128e-05, + "loss": 2.3832, + "step": 26527 + }, + { + "epoch": 1.13, + "grad_norm": 0.482421875, + "learning_rate": 7.993946246542832e-05, + "loss": 2.5094, + "step": 26528 + }, + { + "epoch": 1.13, + "grad_norm": 0.5390625, + "learning_rate": 7.993289093699046e-05, + "loss": 2.518, + "step": 26529 + }, + { + "epoch": 1.13, + "grad_norm": 0.416015625, + "learning_rate": 7.992631949884725e-05, + "loss": 2.6036, + "step": 26530 + }, + { + "epoch": 1.13, + "grad_norm": 0.494140625, + "learning_rate": 7.991974815102828e-05, + "loss": 2.5657, + "step": 26531 + }, + { + "epoch": 1.13, + "grad_norm": 0.4609375, + "learning_rate": 7.991317689356308e-05, + "loss": 2.5233, + "step": 26532 + }, + { + "epoch": 1.13, + "grad_norm": 0.486328125, + "learning_rate": 7.990660572648125e-05, + "loss": 2.4676, + "step": 26533 + }, + { + "epoch": 1.13, + "grad_norm": 0.48046875, + "learning_rate": 7.990003464981239e-05, + "loss": 2.5789, + "step": 26534 + }, + { + "epoch": 1.13, + "grad_norm": 0.48828125, + "learning_rate": 7.9893463663586e-05, + "loss": 2.3864, + "step": 26535 + }, + { + "epoch": 1.13, + "grad_norm": 0.4921875, + "learning_rate": 7.988689276783165e-05, + "loss": 2.328, + "step": 26536 + }, + { + "epoch": 1.13, + "grad_norm": 0.458984375, + "learning_rate": 7.988032196257897e-05, + "loss": 2.5237, + "step": 26537 + }, + { + "epoch": 1.13, + "grad_norm": 0.482421875, + "learning_rate": 7.987375124785746e-05, + "loss": 2.4618, + "step": 26538 + }, + { + "epoch": 1.13, + "grad_norm": 0.48046875, + "learning_rate": 7.986718062369673e-05, + "loss": 2.4401, + "step": 26539 + }, + { + "epoch": 1.13, + "grad_norm": 0.443359375, + "learning_rate": 7.986061009012632e-05, + "loss": 2.4279, + "step": 26540 + }, + { + "epoch": 1.13, + "grad_norm": 0.494140625, + "learning_rate": 7.985403964717581e-05, + "loss": 2.5536, + "step": 26541 + }, + { + "epoch": 1.13, + "grad_norm": 0.478515625, + "learning_rate": 7.98474692948748e-05, + "loss": 2.5058, + "step": 26542 + }, + { + "epoch": 1.13, + "grad_norm": 0.44921875, + "learning_rate": 7.984089903325276e-05, + "loss": 2.3896, + "step": 26543 + }, + { + "epoch": 1.13, + "grad_norm": 0.48828125, + "learning_rate": 7.98343288623393e-05, + "loss": 2.5135, + "step": 26544 + }, + { + "epoch": 1.13, + "grad_norm": 0.5234375, + "learning_rate": 7.982775878216402e-05, + "loss": 2.6119, + "step": 26545 + }, + { + "epoch": 1.13, + "grad_norm": 0.421875, + "learning_rate": 7.982118879275645e-05, + "loss": 2.2954, + "step": 26546 + }, + { + "epoch": 1.13, + "grad_norm": 0.5390625, + "learning_rate": 7.981461889414616e-05, + "loss": 2.5267, + "step": 26547 + }, + { + "epoch": 1.13, + "grad_norm": 0.4375, + "learning_rate": 7.980804908636267e-05, + "loss": 2.3832, + "step": 26548 + }, + { + "epoch": 1.13, + "grad_norm": 0.46875, + "learning_rate": 7.98014793694356e-05, + "loss": 2.5346, + "step": 26549 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 7.979490974339454e-05, + "loss": 2.4761, + "step": 26550 + }, + { + "epoch": 1.13, + "grad_norm": 0.494140625, + "learning_rate": 7.978834020826896e-05, + "loss": 2.5645, + "step": 26551 + }, + { + "epoch": 1.13, + "grad_norm": 0.484375, + "learning_rate": 7.978177076408847e-05, + "loss": 2.4501, + "step": 26552 + }, + { + "epoch": 1.13, + "grad_norm": 0.490234375, + "learning_rate": 7.977520141088259e-05, + "loss": 2.3337, + "step": 26553 + }, + { + "epoch": 1.13, + "grad_norm": 0.5, + "learning_rate": 7.976863214868095e-05, + "loss": 2.5337, + "step": 26554 + }, + { + "epoch": 1.13, + "grad_norm": 0.423828125, + "learning_rate": 7.976206297751307e-05, + "loss": 2.405, + "step": 26555 + }, + { + "epoch": 1.13, + "grad_norm": 0.478515625, + "learning_rate": 7.975549389740852e-05, + "loss": 2.6097, + "step": 26556 + }, + { + "epoch": 1.13, + "grad_norm": 0.51171875, + "learning_rate": 7.974892490839679e-05, + "loss": 2.4315, + "step": 26557 + }, + { + "epoch": 1.13, + "grad_norm": 0.44140625, + "learning_rate": 7.974235601050756e-05, + "loss": 2.3965, + "step": 26558 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 7.97357872037703e-05, + "loss": 2.5786, + "step": 26559 + }, + { + "epoch": 1.13, + "grad_norm": 0.4375, + "learning_rate": 7.972921848821459e-05, + "loss": 2.5136, + "step": 26560 + }, + { + "epoch": 1.13, + "grad_norm": 0.47265625, + "learning_rate": 7.972264986386999e-05, + "loss": 2.46, + "step": 26561 + }, + { + "epoch": 1.13, + "grad_norm": 0.451171875, + "learning_rate": 7.971608133076606e-05, + "loss": 2.4034, + "step": 26562 + }, + { + "epoch": 1.13, + "grad_norm": 0.453125, + "learning_rate": 7.970951288893235e-05, + "loss": 2.4374, + "step": 26563 + }, + { + "epoch": 1.13, + "grad_norm": 0.5, + "learning_rate": 7.970294453839844e-05, + "loss": 2.5154, + "step": 26564 + }, + { + "epoch": 1.13, + "grad_norm": 0.443359375, + "learning_rate": 7.96963762791938e-05, + "loss": 2.4843, + "step": 26565 + }, + { + "epoch": 1.13, + "grad_norm": 0.5546875, + "learning_rate": 7.968980811134809e-05, + "loss": 2.6815, + "step": 26566 + }, + { + "epoch": 1.13, + "grad_norm": 0.50390625, + "learning_rate": 7.968324003489082e-05, + "loss": 2.4357, + "step": 26567 + }, + { + "epoch": 1.13, + "grad_norm": 0.439453125, + "learning_rate": 7.967667204985155e-05, + "loss": 2.4294, + "step": 26568 + }, + { + "epoch": 1.13, + "grad_norm": 0.50390625, + "learning_rate": 7.967010415625979e-05, + "loss": 2.4588, + "step": 26569 + }, + { + "epoch": 1.13, + "grad_norm": 0.4453125, + "learning_rate": 7.966353635414518e-05, + "loss": 2.5665, + "step": 26570 + }, + { + "epoch": 1.13, + "grad_norm": 0.48046875, + "learning_rate": 7.96569686435372e-05, + "loss": 2.436, + "step": 26571 + }, + { + "epoch": 1.13, + "grad_norm": 0.4609375, + "learning_rate": 7.965040102446546e-05, + "loss": 2.4693, + "step": 26572 + }, + { + "epoch": 1.13, + "grad_norm": 0.44921875, + "learning_rate": 7.964383349695941e-05, + "loss": 2.5397, + "step": 26573 + }, + { + "epoch": 1.13, + "grad_norm": 0.484375, + "learning_rate": 7.963726606104872e-05, + "loss": 2.4634, + "step": 26574 + }, + { + "epoch": 1.13, + "grad_norm": 0.49609375, + "learning_rate": 7.963069871676288e-05, + "loss": 2.3974, + "step": 26575 + }, + { + "epoch": 1.13, + "grad_norm": 0.44140625, + "learning_rate": 7.962413146413146e-05, + "loss": 2.4374, + "step": 26576 + }, + { + "epoch": 1.13, + "grad_norm": 0.455078125, + "learning_rate": 7.961756430318398e-05, + "loss": 2.4498, + "step": 26577 + }, + { + "epoch": 1.13, + "grad_norm": 0.4375, + "learning_rate": 7.961099723395002e-05, + "loss": 2.5122, + "step": 26578 + }, + { + "epoch": 1.13, + "grad_norm": 0.494140625, + "learning_rate": 7.960443025645914e-05, + "loss": 2.5279, + "step": 26579 + }, + { + "epoch": 1.13, + "grad_norm": 0.49609375, + "learning_rate": 7.959786337074087e-05, + "loss": 2.476, + "step": 26580 + }, + { + "epoch": 1.13, + "grad_norm": 0.515625, + "learning_rate": 7.959129657682471e-05, + "loss": 2.5767, + "step": 26581 + }, + { + "epoch": 1.13, + "grad_norm": 0.47265625, + "learning_rate": 7.958472987474032e-05, + "loss": 2.4935, + "step": 26582 + }, + { + "epoch": 1.13, + "grad_norm": 0.47265625, + "learning_rate": 7.957816326451716e-05, + "loss": 2.4451, + "step": 26583 + }, + { + "epoch": 1.13, + "grad_norm": 0.490234375, + "learning_rate": 7.957159674618479e-05, + "loss": 2.4479, + "step": 26584 + }, + { + "epoch": 1.13, + "grad_norm": 0.4765625, + "learning_rate": 7.956503031977276e-05, + "loss": 2.4145, + "step": 26585 + }, + { + "epoch": 1.13, + "grad_norm": 0.42578125, + "learning_rate": 7.955846398531065e-05, + "loss": 2.5005, + "step": 26586 + }, + { + "epoch": 1.13, + "grad_norm": 0.5234375, + "learning_rate": 7.955189774282798e-05, + "loss": 2.5881, + "step": 26587 + }, + { + "epoch": 1.13, + "grad_norm": 0.51171875, + "learning_rate": 7.954533159235431e-05, + "loss": 2.4987, + "step": 26588 + }, + { + "epoch": 1.13, + "grad_norm": 0.4921875, + "learning_rate": 7.953876553391913e-05, + "loss": 2.5511, + "step": 26589 + }, + { + "epoch": 1.13, + "grad_norm": 0.546875, + "learning_rate": 7.953219956755207e-05, + "loss": 2.469, + "step": 26590 + }, + { + "epoch": 1.13, + "grad_norm": 0.498046875, + "learning_rate": 7.95256336932826e-05, + "loss": 2.4048, + "step": 26591 + }, + { + "epoch": 1.13, + "grad_norm": 0.470703125, + "learning_rate": 7.951906791114032e-05, + "loss": 2.6521, + "step": 26592 + }, + { + "epoch": 1.13, + "grad_norm": 0.419921875, + "learning_rate": 7.951250222115471e-05, + "loss": 2.3203, + "step": 26593 + }, + { + "epoch": 1.13, + "grad_norm": 0.4609375, + "learning_rate": 7.950593662335539e-05, + "loss": 2.5746, + "step": 26594 + }, + { + "epoch": 1.13, + "grad_norm": 0.48828125, + "learning_rate": 7.949937111777185e-05, + "loss": 2.535, + "step": 26595 + }, + { + "epoch": 1.13, + "grad_norm": 0.486328125, + "learning_rate": 7.949280570443369e-05, + "loss": 2.6489, + "step": 26596 + }, + { + "epoch": 1.13, + "grad_norm": 0.53125, + "learning_rate": 7.948624038337033e-05, + "loss": 2.5593, + "step": 26597 + }, + { + "epoch": 1.13, + "grad_norm": 0.427734375, + "learning_rate": 7.947967515461146e-05, + "loss": 2.3989, + "step": 26598 + }, + { + "epoch": 1.13, + "grad_norm": 0.490234375, + "learning_rate": 7.947311001818654e-05, + "loss": 2.4665, + "step": 26599 + }, + { + "epoch": 1.13, + "grad_norm": 0.490234375, + "learning_rate": 7.946654497412511e-05, + "loss": 2.362, + "step": 26600 + }, + { + "epoch": 1.13, + "grad_norm": 0.421875, + "learning_rate": 7.945998002245671e-05, + "loss": 2.2624, + "step": 26601 + }, + { + "epoch": 1.13, + "grad_norm": 0.46484375, + "learning_rate": 7.945341516321091e-05, + "loss": 2.4733, + "step": 26602 + }, + { + "epoch": 1.13, + "grad_norm": 0.451171875, + "learning_rate": 7.944685039641724e-05, + "loss": 2.4135, + "step": 26603 + }, + { + "epoch": 1.13, + "grad_norm": 0.486328125, + "learning_rate": 7.944028572210525e-05, + "loss": 2.2952, + "step": 26604 + }, + { + "epoch": 1.13, + "grad_norm": 0.48046875, + "learning_rate": 7.943372114030438e-05, + "loss": 2.4718, + "step": 26605 + }, + { + "epoch": 1.13, + "grad_norm": 0.462890625, + "learning_rate": 7.942715665104433e-05, + "loss": 2.5426, + "step": 26606 + }, + { + "epoch": 1.13, + "grad_norm": 0.4921875, + "learning_rate": 7.942059225435454e-05, + "loss": 2.5157, + "step": 26607 + }, + { + "epoch": 1.13, + "grad_norm": 0.46484375, + "learning_rate": 7.941402795026456e-05, + "loss": 2.3797, + "step": 26608 + }, + { + "epoch": 1.13, + "grad_norm": 0.5078125, + "learning_rate": 7.94074637388039e-05, + "loss": 2.4898, + "step": 26609 + }, + { + "epoch": 1.13, + "grad_norm": 0.52734375, + "learning_rate": 7.940089962000214e-05, + "loss": 2.5558, + "step": 26610 + }, + { + "epoch": 1.13, + "grad_norm": 0.400390625, + "learning_rate": 7.939433559388883e-05, + "loss": 2.4045, + "step": 26611 + }, + { + "epoch": 1.13, + "grad_norm": 0.40625, + "learning_rate": 7.93877716604935e-05, + "loss": 2.3584, + "step": 26612 + }, + { + "epoch": 1.13, + "grad_norm": 0.4609375, + "learning_rate": 7.938120781984557e-05, + "loss": 2.6534, + "step": 26613 + }, + { + "epoch": 1.13, + "grad_norm": 0.54296875, + "learning_rate": 7.937464407197475e-05, + "loss": 2.5196, + "step": 26614 + }, + { + "epoch": 1.13, + "grad_norm": 0.51171875, + "learning_rate": 7.936808041691048e-05, + "loss": 2.5409, + "step": 26615 + }, + { + "epoch": 1.13, + "grad_norm": 0.48046875, + "learning_rate": 7.93615168546823e-05, + "loss": 2.5429, + "step": 26616 + }, + { + "epoch": 1.13, + "grad_norm": 0.4609375, + "learning_rate": 7.935495338531975e-05, + "loss": 2.4431, + "step": 26617 + }, + { + "epoch": 1.13, + "grad_norm": 0.4921875, + "learning_rate": 7.934839000885236e-05, + "loss": 2.4397, + "step": 26618 + }, + { + "epoch": 1.13, + "grad_norm": 0.44921875, + "learning_rate": 7.934182672530968e-05, + "loss": 2.4277, + "step": 26619 + }, + { + "epoch": 1.13, + "grad_norm": 0.55078125, + "learning_rate": 7.933526353472125e-05, + "loss": 2.419, + "step": 26620 + }, + { + "epoch": 1.13, + "grad_norm": 0.4921875, + "learning_rate": 7.932870043711651e-05, + "loss": 2.4248, + "step": 26621 + }, + { + "epoch": 1.13, + "grad_norm": 0.474609375, + "learning_rate": 7.932213743252513e-05, + "loss": 2.4666, + "step": 26622 + }, + { + "epoch": 1.13, + "grad_norm": 0.470703125, + "learning_rate": 7.931557452097656e-05, + "loss": 2.5151, + "step": 26623 + }, + { + "epoch": 1.13, + "grad_norm": 0.474609375, + "learning_rate": 7.930901170250035e-05, + "loss": 2.3774, + "step": 26624 + }, + { + "epoch": 1.13, + "grad_norm": 0.443359375, + "learning_rate": 7.930244897712599e-05, + "loss": 2.4801, + "step": 26625 + }, + { + "epoch": 1.13, + "grad_norm": 0.48046875, + "learning_rate": 7.929588634488307e-05, + "loss": 2.4769, + "step": 26626 + }, + { + "epoch": 1.13, + "grad_norm": 0.4609375, + "learning_rate": 7.928932380580111e-05, + "loss": 2.4627, + "step": 26627 + }, + { + "epoch": 1.13, + "grad_norm": 0.458984375, + "learning_rate": 7.928276135990962e-05, + "loss": 2.3574, + "step": 26628 + }, + { + "epoch": 1.13, + "grad_norm": 0.462890625, + "learning_rate": 7.927619900723809e-05, + "loss": 2.5618, + "step": 26629 + }, + { + "epoch": 1.13, + "grad_norm": 0.53125, + "learning_rate": 7.926963674781615e-05, + "loss": 2.5724, + "step": 26630 + }, + { + "epoch": 1.13, + "grad_norm": 0.48828125, + "learning_rate": 7.926307458167325e-05, + "loss": 2.3928, + "step": 26631 + }, + { + "epoch": 1.13, + "grad_norm": 0.482421875, + "learning_rate": 7.925651250883893e-05, + "loss": 2.3949, + "step": 26632 + }, + { + "epoch": 1.13, + "grad_norm": 0.47265625, + "learning_rate": 7.92499505293427e-05, + "loss": 2.5325, + "step": 26633 + }, + { + "epoch": 1.13, + "grad_norm": 0.46484375, + "learning_rate": 7.924338864321413e-05, + "loss": 2.5359, + "step": 26634 + }, + { + "epoch": 1.13, + "grad_norm": 0.466796875, + "learning_rate": 7.923682685048273e-05, + "loss": 2.4553, + "step": 26635 + }, + { + "epoch": 1.14, + "grad_norm": 0.421875, + "learning_rate": 7.923026515117804e-05, + "loss": 2.3757, + "step": 26636 + }, + { + "epoch": 1.14, + "grad_norm": 0.45703125, + "learning_rate": 7.922370354532951e-05, + "loss": 2.433, + "step": 26637 + }, + { + "epoch": 1.14, + "grad_norm": 0.5, + "learning_rate": 7.921714203296678e-05, + "loss": 2.5355, + "step": 26638 + }, + { + "epoch": 1.14, + "grad_norm": 0.44921875, + "learning_rate": 7.921058061411929e-05, + "loss": 2.4685, + "step": 26639 + }, + { + "epoch": 1.14, + "grad_norm": 0.447265625, + "learning_rate": 7.920401928881659e-05, + "loss": 2.4573, + "step": 26640 + }, + { + "epoch": 1.14, + "grad_norm": 0.44140625, + "learning_rate": 7.919745805708818e-05, + "loss": 2.6039, + "step": 26641 + }, + { + "epoch": 1.14, + "grad_norm": 0.3984375, + "learning_rate": 7.919089691896364e-05, + "loss": 2.4278, + "step": 26642 + }, + { + "epoch": 1.14, + "grad_norm": 0.44921875, + "learning_rate": 7.918433587447245e-05, + "loss": 2.4108, + "step": 26643 + }, + { + "epoch": 1.14, + "grad_norm": 0.5390625, + "learning_rate": 7.917777492364417e-05, + "loss": 2.4876, + "step": 26644 + }, + { + "epoch": 1.14, + "grad_norm": 0.455078125, + "learning_rate": 7.917121406650822e-05, + "loss": 2.3745, + "step": 26645 + }, + { + "epoch": 1.14, + "grad_norm": 0.455078125, + "learning_rate": 7.916465330309426e-05, + "loss": 2.4209, + "step": 26646 + }, + { + "epoch": 1.14, + "grad_norm": 0.49609375, + "learning_rate": 7.915809263343172e-05, + "loss": 2.4797, + "step": 26647 + }, + { + "epoch": 1.14, + "grad_norm": 0.439453125, + "learning_rate": 7.915153205755015e-05, + "loss": 2.2932, + "step": 26648 + }, + { + "epoch": 1.14, + "grad_norm": 0.451171875, + "learning_rate": 7.914497157547905e-05, + "loss": 2.4922, + "step": 26649 + }, + { + "epoch": 1.14, + "grad_norm": 0.447265625, + "learning_rate": 7.913841118724797e-05, + "loss": 2.5426, + "step": 26650 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.913185089288642e-05, + "loss": 2.3506, + "step": 26651 + }, + { + "epoch": 1.14, + "grad_norm": 0.474609375, + "learning_rate": 7.912529069242393e-05, + "loss": 2.5235, + "step": 26652 + }, + { + "epoch": 1.14, + "grad_norm": 0.4609375, + "learning_rate": 7.911873058588996e-05, + "loss": 2.4397, + "step": 26653 + }, + { + "epoch": 1.14, + "grad_norm": 0.44140625, + "learning_rate": 7.911217057331411e-05, + "loss": 2.4763, + "step": 26654 + }, + { + "epoch": 1.14, + "grad_norm": 0.498046875, + "learning_rate": 7.910561065472586e-05, + "loss": 2.3819, + "step": 26655 + }, + { + "epoch": 1.14, + "grad_norm": 0.486328125, + "learning_rate": 7.909905083015471e-05, + "loss": 2.45, + "step": 26656 + }, + { + "epoch": 1.14, + "grad_norm": 0.47265625, + "learning_rate": 7.909249109963017e-05, + "loss": 2.4484, + "step": 26657 + }, + { + "epoch": 1.14, + "grad_norm": 0.5078125, + "learning_rate": 7.90859314631818e-05, + "loss": 2.4003, + "step": 26658 + }, + { + "epoch": 1.14, + "grad_norm": 0.404296875, + "learning_rate": 7.90793719208391e-05, + "loss": 2.3901, + "step": 26659 + }, + { + "epoch": 1.14, + "grad_norm": 0.498046875, + "learning_rate": 7.90728124726316e-05, + "loss": 2.433, + "step": 26660 + }, + { + "epoch": 1.14, + "grad_norm": 0.423828125, + "learning_rate": 7.906625311858876e-05, + "loss": 2.417, + "step": 26661 + }, + { + "epoch": 1.14, + "grad_norm": 0.462890625, + "learning_rate": 7.905969385874016e-05, + "loss": 2.6761, + "step": 26662 + }, + { + "epoch": 1.14, + "grad_norm": 0.4921875, + "learning_rate": 7.905313469311527e-05, + "loss": 2.5132, + "step": 26663 + }, + { + "epoch": 1.14, + "grad_norm": 0.466796875, + "learning_rate": 7.904657562174362e-05, + "loss": 2.4069, + "step": 26664 + }, + { + "epoch": 1.14, + "grad_norm": 0.482421875, + "learning_rate": 7.90400166446547e-05, + "loss": 2.5216, + "step": 26665 + }, + { + "epoch": 1.14, + "grad_norm": 0.458984375, + "learning_rate": 7.903345776187807e-05, + "loss": 2.4397, + "step": 26666 + }, + { + "epoch": 1.14, + "grad_norm": 0.439453125, + "learning_rate": 7.902689897344322e-05, + "loss": 2.4563, + "step": 26667 + }, + { + "epoch": 1.14, + "grad_norm": 0.453125, + "learning_rate": 7.902034027937968e-05, + "loss": 2.4916, + "step": 26668 + }, + { + "epoch": 1.14, + "grad_norm": 0.458984375, + "learning_rate": 7.901378167971687e-05, + "loss": 2.4763, + "step": 26669 + }, + { + "epoch": 1.14, + "grad_norm": 0.51953125, + "learning_rate": 7.900722317448445e-05, + "loss": 2.463, + "step": 26670 + }, + { + "epoch": 1.14, + "grad_norm": 0.51953125, + "learning_rate": 7.900066476371182e-05, + "loss": 2.5139, + "step": 26671 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.899410644742853e-05, + "loss": 2.4307, + "step": 26672 + }, + { + "epoch": 1.14, + "grad_norm": 0.5, + "learning_rate": 7.898754822566408e-05, + "loss": 2.5202, + "step": 26673 + }, + { + "epoch": 1.14, + "grad_norm": 0.443359375, + "learning_rate": 7.898099009844798e-05, + "loss": 2.4234, + "step": 26674 + }, + { + "epoch": 1.14, + "grad_norm": 0.53125, + "learning_rate": 7.897443206580973e-05, + "loss": 2.6391, + "step": 26675 + }, + { + "epoch": 1.14, + "grad_norm": 0.40625, + "learning_rate": 7.89678741277789e-05, + "loss": 2.4159, + "step": 26676 + }, + { + "epoch": 1.14, + "grad_norm": 0.49609375, + "learning_rate": 7.896131628438492e-05, + "loss": 2.5197, + "step": 26677 + }, + { + "epoch": 1.14, + "grad_norm": 0.482421875, + "learning_rate": 7.89547585356573e-05, + "loss": 2.5667, + "step": 26678 + }, + { + "epoch": 1.14, + "grad_norm": 0.53125, + "learning_rate": 7.894820088162561e-05, + "loss": 2.483, + "step": 26679 + }, + { + "epoch": 1.14, + "grad_norm": 0.482421875, + "learning_rate": 7.894164332231933e-05, + "loss": 2.4493, + "step": 26680 + }, + { + "epoch": 1.14, + "grad_norm": 0.478515625, + "learning_rate": 7.893508585776795e-05, + "loss": 2.5243, + "step": 26681 + }, + { + "epoch": 1.14, + "grad_norm": 0.490234375, + "learning_rate": 7.892852848800095e-05, + "loss": 2.5412, + "step": 26682 + }, + { + "epoch": 1.14, + "grad_norm": 0.3984375, + "learning_rate": 7.892197121304789e-05, + "loss": 2.3838, + "step": 26683 + }, + { + "epoch": 1.14, + "grad_norm": 0.462890625, + "learning_rate": 7.89154140329383e-05, + "loss": 2.3992, + "step": 26684 + }, + { + "epoch": 1.14, + "grad_norm": 0.470703125, + "learning_rate": 7.890885694770161e-05, + "loss": 2.5464, + "step": 26685 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.890229995736733e-05, + "loss": 2.5206, + "step": 26686 + }, + { + "epoch": 1.14, + "grad_norm": 0.43359375, + "learning_rate": 7.889574306196502e-05, + "loss": 2.4884, + "step": 26687 + }, + { + "epoch": 1.14, + "grad_norm": 0.51171875, + "learning_rate": 7.888918626152415e-05, + "loss": 2.5279, + "step": 26688 + }, + { + "epoch": 1.14, + "grad_norm": 0.443359375, + "learning_rate": 7.888262955607423e-05, + "loss": 2.6085, + "step": 26689 + }, + { + "epoch": 1.14, + "grad_norm": 0.462890625, + "learning_rate": 7.887607294564475e-05, + "loss": 2.4729, + "step": 26690 + }, + { + "epoch": 1.14, + "grad_norm": 0.443359375, + "learning_rate": 7.886951643026522e-05, + "loss": 2.479, + "step": 26691 + }, + { + "epoch": 1.14, + "grad_norm": 0.48046875, + "learning_rate": 7.886296000996516e-05, + "loss": 2.3838, + "step": 26692 + }, + { + "epoch": 1.14, + "grad_norm": 0.48828125, + "learning_rate": 7.885640368477405e-05, + "loss": 2.4223, + "step": 26693 + }, + { + "epoch": 1.14, + "grad_norm": 0.44140625, + "learning_rate": 7.884984745472136e-05, + "loss": 2.3956, + "step": 26694 + }, + { + "epoch": 1.14, + "grad_norm": 0.494140625, + "learning_rate": 7.884329131983668e-05, + "loss": 2.4698, + "step": 26695 + }, + { + "epoch": 1.14, + "grad_norm": 0.45703125, + "learning_rate": 7.883673528014943e-05, + "loss": 2.4503, + "step": 26696 + }, + { + "epoch": 1.14, + "grad_norm": 0.458984375, + "learning_rate": 7.883017933568915e-05, + "loss": 2.4875, + "step": 26697 + }, + { + "epoch": 1.14, + "grad_norm": 0.490234375, + "learning_rate": 7.88236234864853e-05, + "loss": 2.4513, + "step": 26698 + }, + { + "epoch": 1.14, + "grad_norm": 0.451171875, + "learning_rate": 7.881706773256741e-05, + "loss": 2.4797, + "step": 26699 + }, + { + "epoch": 1.14, + "grad_norm": 0.4375, + "learning_rate": 7.881051207396503e-05, + "loss": 2.4934, + "step": 26700 + }, + { + "epoch": 1.14, + "grad_norm": 0.48046875, + "learning_rate": 7.880395651070755e-05, + "loss": 2.6274, + "step": 26701 + }, + { + "epoch": 1.14, + "grad_norm": 0.431640625, + "learning_rate": 7.879740104282452e-05, + "loss": 2.359, + "step": 26702 + }, + { + "epoch": 1.14, + "grad_norm": 0.52734375, + "learning_rate": 7.879084567034543e-05, + "loss": 2.4156, + "step": 26703 + }, + { + "epoch": 1.14, + "grad_norm": 0.4453125, + "learning_rate": 7.87842903932998e-05, + "loss": 2.4017, + "step": 26704 + }, + { + "epoch": 1.14, + "grad_norm": 0.423828125, + "learning_rate": 7.877773521171712e-05, + "loss": 2.4168, + "step": 26705 + }, + { + "epoch": 1.14, + "grad_norm": 0.404296875, + "learning_rate": 7.877118012562684e-05, + "loss": 2.5461, + "step": 26706 + }, + { + "epoch": 1.14, + "grad_norm": 0.470703125, + "learning_rate": 7.876462513505853e-05, + "loss": 2.538, + "step": 26707 + }, + { + "epoch": 1.14, + "grad_norm": 0.494140625, + "learning_rate": 7.875807024004165e-05, + "loss": 2.4847, + "step": 26708 + }, + { + "epoch": 1.14, + "grad_norm": 0.431640625, + "learning_rate": 7.875151544060568e-05, + "loss": 2.4121, + "step": 26709 + }, + { + "epoch": 1.14, + "grad_norm": 0.51953125, + "learning_rate": 7.874496073678009e-05, + "loss": 2.5638, + "step": 26710 + }, + { + "epoch": 1.14, + "grad_norm": 0.435546875, + "learning_rate": 7.873840612859445e-05, + "loss": 2.5386, + "step": 26711 + }, + { + "epoch": 1.14, + "grad_norm": 0.458984375, + "learning_rate": 7.873185161607818e-05, + "loss": 2.4431, + "step": 26712 + }, + { + "epoch": 1.14, + "grad_norm": 0.45703125, + "learning_rate": 7.872529719926084e-05, + "loss": 2.384, + "step": 26713 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.871874287817185e-05, + "loss": 2.5775, + "step": 26714 + }, + { + "epoch": 1.14, + "grad_norm": 0.5078125, + "learning_rate": 7.871218865284076e-05, + "loss": 2.4323, + "step": 26715 + }, + { + "epoch": 1.14, + "grad_norm": 0.4765625, + "learning_rate": 7.870563452329706e-05, + "loss": 2.5005, + "step": 26716 + }, + { + "epoch": 1.14, + "grad_norm": 0.484375, + "learning_rate": 7.869908048957022e-05, + "loss": 2.4752, + "step": 26717 + }, + { + "epoch": 1.14, + "grad_norm": 0.494140625, + "learning_rate": 7.86925265516897e-05, + "loss": 2.3655, + "step": 26718 + }, + { + "epoch": 1.14, + "grad_norm": 0.494140625, + "learning_rate": 7.868597270968503e-05, + "loss": 2.5156, + "step": 26719 + }, + { + "epoch": 1.14, + "grad_norm": 0.494140625, + "learning_rate": 7.867941896358572e-05, + "loss": 2.4095, + "step": 26720 + }, + { + "epoch": 1.14, + "grad_norm": 0.431640625, + "learning_rate": 7.867286531342121e-05, + "loss": 2.4728, + "step": 26721 + }, + { + "epoch": 1.14, + "grad_norm": 0.4921875, + "learning_rate": 7.8666311759221e-05, + "loss": 2.4524, + "step": 26722 + }, + { + "epoch": 1.14, + "grad_norm": 0.48828125, + "learning_rate": 7.865975830101463e-05, + "loss": 2.5131, + "step": 26723 + }, + { + "epoch": 1.14, + "grad_norm": 0.53125, + "learning_rate": 7.865320493883154e-05, + "loss": 2.3773, + "step": 26724 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.864665167270123e-05, + "loss": 2.4776, + "step": 26725 + }, + { + "epoch": 1.14, + "grad_norm": 0.515625, + "learning_rate": 7.864009850265314e-05, + "loss": 2.45, + "step": 26726 + }, + { + "epoch": 1.14, + "grad_norm": 0.49609375, + "learning_rate": 7.863354542871683e-05, + "loss": 2.409, + "step": 26727 + }, + { + "epoch": 1.14, + "grad_norm": 0.466796875, + "learning_rate": 7.862699245092175e-05, + "loss": 2.5848, + "step": 26728 + }, + { + "epoch": 1.14, + "grad_norm": 0.4921875, + "learning_rate": 7.862043956929741e-05, + "loss": 2.4698, + "step": 26729 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.861388678387325e-05, + "loss": 2.5243, + "step": 26730 + }, + { + "epoch": 1.14, + "grad_norm": 0.44921875, + "learning_rate": 7.86073340946788e-05, + "loss": 2.324, + "step": 26731 + }, + { + "epoch": 1.14, + "grad_norm": 0.443359375, + "learning_rate": 7.860078150174355e-05, + "loss": 2.4355, + "step": 26732 + }, + { + "epoch": 1.14, + "grad_norm": 0.470703125, + "learning_rate": 7.859422900509694e-05, + "loss": 2.4277, + "step": 26733 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.858767660476847e-05, + "loss": 2.5121, + "step": 26734 + }, + { + "epoch": 1.14, + "grad_norm": 0.474609375, + "learning_rate": 7.858112430078764e-05, + "loss": 2.482, + "step": 26735 + }, + { + "epoch": 1.14, + "grad_norm": 0.455078125, + "learning_rate": 7.857457209318393e-05, + "loss": 2.4068, + "step": 26736 + }, + { + "epoch": 1.14, + "grad_norm": 0.44140625, + "learning_rate": 7.85680199819868e-05, + "loss": 2.467, + "step": 26737 + }, + { + "epoch": 1.14, + "grad_norm": 0.4453125, + "learning_rate": 7.856146796722574e-05, + "loss": 2.4621, + "step": 26738 + }, + { + "epoch": 1.14, + "grad_norm": 0.515625, + "learning_rate": 7.855491604893027e-05, + "loss": 2.5006, + "step": 26739 + }, + { + "epoch": 1.14, + "grad_norm": 0.484375, + "learning_rate": 7.854836422712983e-05, + "loss": 2.5456, + "step": 26740 + }, + { + "epoch": 1.14, + "grad_norm": 0.486328125, + "learning_rate": 7.854181250185392e-05, + "loss": 2.5421, + "step": 26741 + }, + { + "epoch": 1.14, + "grad_norm": 0.455078125, + "learning_rate": 7.853526087313198e-05, + "loss": 2.487, + "step": 26742 + }, + { + "epoch": 1.14, + "grad_norm": 0.423828125, + "learning_rate": 7.852870934099354e-05, + "loss": 2.4553, + "step": 26743 + }, + { + "epoch": 1.14, + "grad_norm": 0.46875, + "learning_rate": 7.852215790546806e-05, + "loss": 2.4444, + "step": 26744 + }, + { + "epoch": 1.14, + "grad_norm": 0.53125, + "learning_rate": 7.851560656658502e-05, + "loss": 2.421, + "step": 26745 + }, + { + "epoch": 1.14, + "grad_norm": 0.5390625, + "learning_rate": 7.850905532437388e-05, + "loss": 2.4966, + "step": 26746 + }, + { + "epoch": 1.14, + "grad_norm": 0.54296875, + "learning_rate": 7.850250417886417e-05, + "loss": 2.4816, + "step": 26747 + }, + { + "epoch": 1.14, + "grad_norm": 0.490234375, + "learning_rate": 7.849595313008534e-05, + "loss": 2.4894, + "step": 26748 + }, + { + "epoch": 1.14, + "grad_norm": 0.46875, + "learning_rate": 7.848940217806684e-05, + "loss": 2.5916, + "step": 26749 + }, + { + "epoch": 1.14, + "grad_norm": 0.484375, + "learning_rate": 7.848285132283816e-05, + "loss": 2.4824, + "step": 26750 + }, + { + "epoch": 1.14, + "grad_norm": 0.45703125, + "learning_rate": 7.84763005644288e-05, + "loss": 2.4523, + "step": 26751 + }, + { + "epoch": 1.14, + "grad_norm": 0.46875, + "learning_rate": 7.846974990286822e-05, + "loss": 2.4477, + "step": 26752 + }, + { + "epoch": 1.14, + "grad_norm": 0.478515625, + "learning_rate": 7.84631993381859e-05, + "loss": 2.5229, + "step": 26753 + }, + { + "epoch": 1.14, + "grad_norm": 0.4296875, + "learning_rate": 7.845664887041128e-05, + "loss": 2.6317, + "step": 26754 + }, + { + "epoch": 1.14, + "grad_norm": 0.44140625, + "learning_rate": 7.84500984995739e-05, + "loss": 2.5917, + "step": 26755 + }, + { + "epoch": 1.14, + "grad_norm": 0.48046875, + "learning_rate": 7.844354822570323e-05, + "loss": 2.3651, + "step": 26756 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.843699804882867e-05, + "loss": 2.4949, + "step": 26757 + }, + { + "epoch": 1.14, + "grad_norm": 0.51171875, + "learning_rate": 7.843044796897973e-05, + "loss": 2.5166, + "step": 26758 + }, + { + "epoch": 1.14, + "grad_norm": 0.443359375, + "learning_rate": 7.842389798618592e-05, + "loss": 2.4677, + "step": 26759 + }, + { + "epoch": 1.14, + "grad_norm": 0.43359375, + "learning_rate": 7.841734810047668e-05, + "loss": 2.3069, + "step": 26760 + }, + { + "epoch": 1.14, + "grad_norm": 0.4375, + "learning_rate": 7.841079831188149e-05, + "loss": 2.42, + "step": 26761 + }, + { + "epoch": 1.14, + "grad_norm": 0.44140625, + "learning_rate": 7.840424862042979e-05, + "loss": 2.531, + "step": 26762 + }, + { + "epoch": 1.14, + "grad_norm": 0.494140625, + "learning_rate": 7.839769902615109e-05, + "loss": 2.4603, + "step": 26763 + }, + { + "epoch": 1.14, + "grad_norm": 0.466796875, + "learning_rate": 7.839114952907489e-05, + "loss": 2.5285, + "step": 26764 + }, + { + "epoch": 1.14, + "grad_norm": 0.431640625, + "learning_rate": 7.838460012923059e-05, + "loss": 2.3509, + "step": 26765 + }, + { + "epoch": 1.14, + "grad_norm": 0.443359375, + "learning_rate": 7.837805082664766e-05, + "loss": 2.3241, + "step": 26766 + }, + { + "epoch": 1.14, + "grad_norm": 0.4453125, + "learning_rate": 7.837150162135563e-05, + "loss": 2.3972, + "step": 26767 + }, + { + "epoch": 1.14, + "grad_norm": 0.443359375, + "learning_rate": 7.836495251338393e-05, + "loss": 2.3613, + "step": 26768 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.835840350276206e-05, + "loss": 2.4583, + "step": 26769 + }, + { + "epoch": 1.14, + "grad_norm": 0.470703125, + "learning_rate": 7.835185458951943e-05, + "loss": 2.4969, + "step": 26770 + }, + { + "epoch": 1.14, + "grad_norm": 0.4765625, + "learning_rate": 7.834530577368556e-05, + "loss": 2.5411, + "step": 26771 + }, + { + "epoch": 1.14, + "grad_norm": 0.439453125, + "learning_rate": 7.833875705528993e-05, + "loss": 2.2954, + "step": 26772 + }, + { + "epoch": 1.14, + "grad_norm": 0.47265625, + "learning_rate": 7.833220843436194e-05, + "loss": 2.4381, + "step": 26773 + }, + { + "epoch": 1.14, + "grad_norm": 0.498046875, + "learning_rate": 7.832565991093108e-05, + "loss": 2.3384, + "step": 26774 + }, + { + "epoch": 1.14, + "grad_norm": 0.44921875, + "learning_rate": 7.831911148502686e-05, + "loss": 2.4166, + "step": 26775 + }, + { + "epoch": 1.14, + "grad_norm": 0.55078125, + "learning_rate": 7.83125631566787e-05, + "loss": 2.4052, + "step": 26776 + }, + { + "epoch": 1.14, + "grad_norm": 0.466796875, + "learning_rate": 7.83060149259161e-05, + "loss": 2.4738, + "step": 26777 + }, + { + "epoch": 1.14, + "grad_norm": 0.51171875, + "learning_rate": 7.829946679276846e-05, + "loss": 2.5131, + "step": 26778 + }, + { + "epoch": 1.14, + "grad_norm": 0.47265625, + "learning_rate": 7.829291875726532e-05, + "loss": 2.399, + "step": 26779 + }, + { + "epoch": 1.14, + "grad_norm": 0.427734375, + "learning_rate": 7.828637081943614e-05, + "loss": 2.448, + "step": 26780 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.827982297931032e-05, + "loss": 2.499, + "step": 26781 + }, + { + "epoch": 1.14, + "grad_norm": 0.46484375, + "learning_rate": 7.827327523691735e-05, + "loss": 2.5087, + "step": 26782 + }, + { + "epoch": 1.14, + "grad_norm": 0.48828125, + "learning_rate": 7.826672759228673e-05, + "loss": 2.5286, + "step": 26783 + }, + { + "epoch": 1.14, + "grad_norm": 0.458984375, + "learning_rate": 7.826018004544787e-05, + "loss": 2.5822, + "step": 26784 + }, + { + "epoch": 1.14, + "grad_norm": 0.486328125, + "learning_rate": 7.825363259643027e-05, + "loss": 2.4915, + "step": 26785 + }, + { + "epoch": 1.14, + "grad_norm": 0.494140625, + "learning_rate": 7.824708524526335e-05, + "loss": 2.4307, + "step": 26786 + }, + { + "epoch": 1.14, + "grad_norm": 0.455078125, + "learning_rate": 7.824053799197662e-05, + "loss": 2.4774, + "step": 26787 + }, + { + "epoch": 1.14, + "grad_norm": 0.5703125, + "learning_rate": 7.823399083659953e-05, + "loss": 2.441, + "step": 26788 + }, + { + "epoch": 1.14, + "grad_norm": 0.455078125, + "learning_rate": 7.822744377916153e-05, + "loss": 2.4318, + "step": 26789 + }, + { + "epoch": 1.14, + "grad_norm": 0.52734375, + "learning_rate": 7.822089681969203e-05, + "loss": 2.4993, + "step": 26790 + }, + { + "epoch": 1.14, + "grad_norm": 0.482421875, + "learning_rate": 7.821434995822057e-05, + "loss": 2.3887, + "step": 26791 + }, + { + "epoch": 1.14, + "grad_norm": 0.451171875, + "learning_rate": 7.820780319477656e-05, + "loss": 2.4809, + "step": 26792 + }, + { + "epoch": 1.14, + "grad_norm": 0.4375, + "learning_rate": 7.820125652938948e-05, + "loss": 2.5035, + "step": 26793 + }, + { + "epoch": 1.14, + "grad_norm": 0.4609375, + "learning_rate": 7.81947099620888e-05, + "loss": 2.4116, + "step": 26794 + }, + { + "epoch": 1.14, + "grad_norm": 0.53515625, + "learning_rate": 7.818816349290388e-05, + "loss": 2.5023, + "step": 26795 + }, + { + "epoch": 1.14, + "grad_norm": 0.4765625, + "learning_rate": 7.818161712186434e-05, + "loss": 2.5081, + "step": 26796 + }, + { + "epoch": 1.14, + "grad_norm": 0.44921875, + "learning_rate": 7.817507084899952e-05, + "loss": 2.4148, + "step": 26797 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.816852467433889e-05, + "loss": 2.5267, + "step": 26798 + }, + { + "epoch": 1.14, + "grad_norm": 0.43359375, + "learning_rate": 7.816197859791192e-05, + "loss": 2.2689, + "step": 26799 + }, + { + "epoch": 1.14, + "grad_norm": 0.52734375, + "learning_rate": 7.815543261974808e-05, + "loss": 2.426, + "step": 26800 + }, + { + "epoch": 1.14, + "grad_norm": 0.451171875, + "learning_rate": 7.814888673987681e-05, + "loss": 2.5636, + "step": 26801 + }, + { + "epoch": 1.14, + "grad_norm": 0.43359375, + "learning_rate": 7.814234095832757e-05, + "loss": 2.3958, + "step": 26802 + }, + { + "epoch": 1.14, + "grad_norm": 0.53125, + "learning_rate": 7.813579527512976e-05, + "loss": 2.4925, + "step": 26803 + }, + { + "epoch": 1.14, + "grad_norm": 0.5, + "learning_rate": 7.812924969031295e-05, + "loss": 2.5747, + "step": 26804 + }, + { + "epoch": 1.14, + "grad_norm": 0.515625, + "learning_rate": 7.812270420390649e-05, + "loss": 2.3793, + "step": 26805 + }, + { + "epoch": 1.14, + "grad_norm": 0.51953125, + "learning_rate": 7.811615881593986e-05, + "loss": 2.4569, + "step": 26806 + }, + { + "epoch": 1.14, + "grad_norm": 0.470703125, + "learning_rate": 7.810961352644252e-05, + "loss": 2.4762, + "step": 26807 + }, + { + "epoch": 1.14, + "grad_norm": 0.435546875, + "learning_rate": 7.810306833544392e-05, + "loss": 2.4732, + "step": 26808 + }, + { + "epoch": 1.14, + "grad_norm": 0.490234375, + "learning_rate": 7.809652324297352e-05, + "loss": 2.4206, + "step": 26809 + }, + { + "epoch": 1.14, + "grad_norm": 0.427734375, + "learning_rate": 7.808997824906077e-05, + "loss": 2.4021, + "step": 26810 + }, + { + "epoch": 1.14, + "grad_norm": 0.478515625, + "learning_rate": 7.808343335373506e-05, + "loss": 2.4164, + "step": 26811 + }, + { + "epoch": 1.14, + "grad_norm": 0.486328125, + "learning_rate": 7.807688855702595e-05, + "loss": 2.5208, + "step": 26812 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.80703438589628e-05, + "loss": 2.4961, + "step": 26813 + }, + { + "epoch": 1.14, + "grad_norm": 0.51953125, + "learning_rate": 7.80637992595751e-05, + "loss": 2.4758, + "step": 26814 + }, + { + "epoch": 1.14, + "grad_norm": 0.455078125, + "learning_rate": 7.805725475889226e-05, + "loss": 2.4232, + "step": 26815 + }, + { + "epoch": 1.14, + "grad_norm": 0.4921875, + "learning_rate": 7.805071035694379e-05, + "loss": 2.4585, + "step": 26816 + }, + { + "epoch": 1.14, + "grad_norm": 0.5546875, + "learning_rate": 7.804416605375908e-05, + "loss": 2.5403, + "step": 26817 + }, + { + "epoch": 1.14, + "grad_norm": 0.474609375, + "learning_rate": 7.803762184936762e-05, + "loss": 2.5101, + "step": 26818 + }, + { + "epoch": 1.14, + "grad_norm": 0.484375, + "learning_rate": 7.803107774379878e-05, + "loss": 2.5215, + "step": 26819 + }, + { + "epoch": 1.14, + "grad_norm": 0.48046875, + "learning_rate": 7.802453373708213e-05, + "loss": 2.4982, + "step": 26820 + }, + { + "epoch": 1.14, + "grad_norm": 0.4609375, + "learning_rate": 7.801798982924702e-05, + "loss": 2.3824, + "step": 26821 + }, + { + "epoch": 1.14, + "grad_norm": 0.46875, + "learning_rate": 7.801144602032292e-05, + "loss": 2.4902, + "step": 26822 + }, + { + "epoch": 1.14, + "grad_norm": 0.546875, + "learning_rate": 7.800490231033926e-05, + "loss": 2.3737, + "step": 26823 + }, + { + "epoch": 1.14, + "grad_norm": 0.55859375, + "learning_rate": 7.799835869932553e-05, + "loss": 2.558, + "step": 26824 + }, + { + "epoch": 1.14, + "grad_norm": 0.5234375, + "learning_rate": 7.799181518731113e-05, + "loss": 2.4919, + "step": 26825 + }, + { + "epoch": 1.14, + "grad_norm": 0.435546875, + "learning_rate": 7.798527177432556e-05, + "loss": 2.4223, + "step": 26826 + }, + { + "epoch": 1.14, + "grad_norm": 0.42578125, + "learning_rate": 7.797872846039813e-05, + "loss": 2.4024, + "step": 26827 + }, + { + "epoch": 1.14, + "grad_norm": 0.462890625, + "learning_rate": 7.797218524555845e-05, + "loss": 2.5265, + "step": 26828 + }, + { + "epoch": 1.14, + "grad_norm": 0.4296875, + "learning_rate": 7.796564212983587e-05, + "loss": 2.3803, + "step": 26829 + }, + { + "epoch": 1.14, + "grad_norm": 0.470703125, + "learning_rate": 7.795909911325985e-05, + "loss": 2.4199, + "step": 26830 + }, + { + "epoch": 1.14, + "grad_norm": 0.5078125, + "learning_rate": 7.79525561958598e-05, + "loss": 2.4687, + "step": 26831 + }, + { + "epoch": 1.14, + "grad_norm": 0.447265625, + "learning_rate": 7.794601337766522e-05, + "loss": 2.4324, + "step": 26832 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.79394706587055e-05, + "loss": 2.4234, + "step": 26833 + }, + { + "epoch": 1.14, + "grad_norm": 0.45703125, + "learning_rate": 7.793292803901014e-05, + "loss": 2.6174, + "step": 26834 + }, + { + "epoch": 1.14, + "grad_norm": 0.453125, + "learning_rate": 7.792638551860846e-05, + "loss": 2.566, + "step": 26835 + }, + { + "epoch": 1.14, + "grad_norm": 0.42578125, + "learning_rate": 7.791984309753006e-05, + "loss": 2.5534, + "step": 26836 + }, + { + "epoch": 1.14, + "grad_norm": 0.474609375, + "learning_rate": 7.791330077580427e-05, + "loss": 2.4504, + "step": 26837 + }, + { + "epoch": 1.14, + "grad_norm": 0.5078125, + "learning_rate": 7.790675855346056e-05, + "loss": 2.6989, + "step": 26838 + }, + { + "epoch": 1.14, + "grad_norm": 0.498046875, + "learning_rate": 7.790021643052832e-05, + "loss": 2.4888, + "step": 26839 + }, + { + "epoch": 1.14, + "grad_norm": 0.458984375, + "learning_rate": 7.789367440703707e-05, + "loss": 2.4125, + "step": 26840 + }, + { + "epoch": 1.14, + "grad_norm": 0.4609375, + "learning_rate": 7.78871324830162e-05, + "loss": 2.4126, + "step": 26841 + }, + { + "epoch": 1.14, + "grad_norm": 0.51171875, + "learning_rate": 7.788059065849518e-05, + "loss": 2.5332, + "step": 26842 + }, + { + "epoch": 1.14, + "grad_norm": 0.47265625, + "learning_rate": 7.787404893350335e-05, + "loss": 2.459, + "step": 26843 + }, + { + "epoch": 1.14, + "grad_norm": 0.474609375, + "learning_rate": 7.786750730807026e-05, + "loss": 2.3641, + "step": 26844 + }, + { + "epoch": 1.14, + "grad_norm": 0.466796875, + "learning_rate": 7.78609657822253e-05, + "loss": 2.4585, + "step": 26845 + }, + { + "epoch": 1.14, + "grad_norm": 0.47265625, + "learning_rate": 7.78544243559979e-05, + "loss": 2.4075, + "step": 26846 + }, + { + "epoch": 1.14, + "grad_norm": 0.5234375, + "learning_rate": 7.784788302941748e-05, + "loss": 2.3671, + "step": 26847 + }, + { + "epoch": 1.14, + "grad_norm": 0.49609375, + "learning_rate": 7.78413418025135e-05, + "loss": 2.4307, + "step": 26848 + }, + { + "epoch": 1.14, + "grad_norm": 0.462890625, + "learning_rate": 7.783480067531538e-05, + "loss": 2.5405, + "step": 26849 + }, + { + "epoch": 1.14, + "grad_norm": 0.443359375, + "learning_rate": 7.782825964785259e-05, + "loss": 2.5356, + "step": 26850 + }, + { + "epoch": 1.14, + "grad_norm": 0.48046875, + "learning_rate": 7.782171872015446e-05, + "loss": 2.3732, + "step": 26851 + }, + { + "epoch": 1.14, + "grad_norm": 0.45703125, + "learning_rate": 7.781517789225053e-05, + "loss": 2.5052, + "step": 26852 + }, + { + "epoch": 1.14, + "grad_norm": 0.50390625, + "learning_rate": 7.780863716417022e-05, + "loss": 2.5659, + "step": 26853 + }, + { + "epoch": 1.14, + "grad_norm": 0.470703125, + "learning_rate": 7.780209653594292e-05, + "loss": 2.3764, + "step": 26854 + }, + { + "epoch": 1.14, + "grad_norm": 0.486328125, + "learning_rate": 7.779555600759803e-05, + "loss": 2.3143, + "step": 26855 + }, + { + "epoch": 1.14, + "grad_norm": 0.58203125, + "learning_rate": 7.778901557916506e-05, + "loss": 2.5651, + "step": 26856 + }, + { + "epoch": 1.14, + "grad_norm": 0.4453125, + "learning_rate": 7.778247525067339e-05, + "loss": 2.4517, + "step": 26857 + }, + { + "epoch": 1.14, + "grad_norm": 0.46484375, + "learning_rate": 7.77759350221525e-05, + "loss": 2.5457, + "step": 26858 + }, + { + "epoch": 1.14, + "grad_norm": 0.484375, + "learning_rate": 7.77693948936317e-05, + "loss": 2.5306, + "step": 26859 + }, + { + "epoch": 1.14, + "grad_norm": 0.490234375, + "learning_rate": 7.776285486514055e-05, + "loss": 2.416, + "step": 26860 + }, + { + "epoch": 1.14, + "grad_norm": 0.51953125, + "learning_rate": 7.775631493670846e-05, + "loss": 2.4346, + "step": 26861 + }, + { + "epoch": 1.14, + "grad_norm": 0.546875, + "learning_rate": 7.774977510836477e-05, + "loss": 2.4689, + "step": 26862 + }, + { + "epoch": 1.14, + "grad_norm": 0.455078125, + "learning_rate": 7.774323538013897e-05, + "loss": 2.4522, + "step": 26863 + }, + { + "epoch": 1.14, + "grad_norm": 0.498046875, + "learning_rate": 7.773669575206049e-05, + "loss": 2.4487, + "step": 26864 + }, + { + "epoch": 1.14, + "grad_norm": 0.5, + "learning_rate": 7.773015622415873e-05, + "loss": 2.3825, + "step": 26865 + }, + { + "epoch": 1.14, + "grad_norm": 0.4609375, + "learning_rate": 7.772361679646316e-05, + "loss": 2.5738, + "step": 26866 + }, + { + "epoch": 1.14, + "grad_norm": 0.421875, + "learning_rate": 7.771707746900311e-05, + "loss": 2.3955, + "step": 26867 + }, + { + "epoch": 1.14, + "grad_norm": 0.466796875, + "learning_rate": 7.771053824180811e-05, + "loss": 2.4731, + "step": 26868 + }, + { + "epoch": 1.14, + "grad_norm": 0.484375, + "learning_rate": 7.770399911490756e-05, + "loss": 2.5692, + "step": 26869 + }, + { + "epoch": 1.15, + "grad_norm": 0.451171875, + "learning_rate": 7.769746008833085e-05, + "loss": 2.3689, + "step": 26870 + }, + { + "epoch": 1.15, + "grad_norm": 0.56640625, + "learning_rate": 7.769092116210739e-05, + "loss": 2.4967, + "step": 26871 + }, + { + "epoch": 1.15, + "grad_norm": 0.482421875, + "learning_rate": 7.768438233626665e-05, + "loss": 2.4986, + "step": 26872 + }, + { + "epoch": 1.15, + "grad_norm": 0.4765625, + "learning_rate": 7.767784361083804e-05, + "loss": 2.4095, + "step": 26873 + }, + { + "epoch": 1.15, + "grad_norm": 0.474609375, + "learning_rate": 7.7671304985851e-05, + "loss": 2.4593, + "step": 26874 + }, + { + "epoch": 1.15, + "grad_norm": 0.4921875, + "learning_rate": 7.766476646133486e-05, + "loss": 2.5701, + "step": 26875 + }, + { + "epoch": 1.15, + "grad_norm": 0.47265625, + "learning_rate": 7.765822803731915e-05, + "loss": 2.4063, + "step": 26876 + }, + { + "epoch": 1.15, + "grad_norm": 0.494140625, + "learning_rate": 7.765168971383328e-05, + "loss": 2.4943, + "step": 26877 + }, + { + "epoch": 1.15, + "grad_norm": 0.46484375, + "learning_rate": 7.76451514909066e-05, + "loss": 2.4538, + "step": 26878 + }, + { + "epoch": 1.15, + "grad_norm": 0.53515625, + "learning_rate": 7.763861336856857e-05, + "loss": 2.357, + "step": 26879 + }, + { + "epoch": 1.15, + "grad_norm": 0.474609375, + "learning_rate": 7.763207534684861e-05, + "loss": 2.3421, + "step": 26880 + }, + { + "epoch": 1.15, + "grad_norm": 0.46875, + "learning_rate": 7.762553742577614e-05, + "loss": 2.4851, + "step": 26881 + }, + { + "epoch": 1.15, + "grad_norm": 0.48046875, + "learning_rate": 7.761899960538061e-05, + "loss": 2.4428, + "step": 26882 + }, + { + "epoch": 1.15, + "grad_norm": 0.52734375, + "learning_rate": 7.761246188569132e-05, + "loss": 2.5404, + "step": 26883 + }, + { + "epoch": 1.15, + "grad_norm": 0.484375, + "learning_rate": 7.760592426673781e-05, + "loss": 2.5225, + "step": 26884 + }, + { + "epoch": 1.15, + "grad_norm": 0.49609375, + "learning_rate": 7.75993867485495e-05, + "loss": 2.3805, + "step": 26885 + }, + { + "epoch": 1.15, + "grad_norm": 0.4609375, + "learning_rate": 7.759284933115572e-05, + "loss": 2.4377, + "step": 26886 + }, + { + "epoch": 1.15, + "grad_norm": 0.490234375, + "learning_rate": 7.758631201458593e-05, + "loss": 2.4695, + "step": 26887 + }, + { + "epoch": 1.15, + "grad_norm": 0.462890625, + "learning_rate": 7.757977479886955e-05, + "loss": 2.5192, + "step": 26888 + }, + { + "epoch": 1.15, + "grad_norm": 0.54296875, + "learning_rate": 7.7573237684036e-05, + "loss": 2.6913, + "step": 26889 + }, + { + "epoch": 1.15, + "grad_norm": 0.515625, + "learning_rate": 7.756670067011468e-05, + "loss": 2.3751, + "step": 26890 + }, + { + "epoch": 1.15, + "grad_norm": 0.466796875, + "learning_rate": 7.756016375713498e-05, + "loss": 2.5577, + "step": 26891 + }, + { + "epoch": 1.15, + "grad_norm": 0.431640625, + "learning_rate": 7.755362694512636e-05, + "loss": 2.4582, + "step": 26892 + }, + { + "epoch": 1.15, + "grad_norm": 0.46484375, + "learning_rate": 7.754709023411824e-05, + "loss": 2.5407, + "step": 26893 + }, + { + "epoch": 1.15, + "grad_norm": 0.462890625, + "learning_rate": 7.754055362414e-05, + "loss": 2.4377, + "step": 26894 + }, + { + "epoch": 1.15, + "grad_norm": 0.4765625, + "learning_rate": 7.753401711522103e-05, + "loss": 2.3855, + "step": 26895 + }, + { + "epoch": 1.15, + "grad_norm": 0.453125, + "learning_rate": 7.75274807073908e-05, + "loss": 2.4663, + "step": 26896 + }, + { + "epoch": 1.15, + "grad_norm": 0.498046875, + "learning_rate": 7.752094440067869e-05, + "loss": 2.46, + "step": 26897 + }, + { + "epoch": 1.15, + "grad_norm": 0.48828125, + "learning_rate": 7.751440819511413e-05, + "loss": 2.536, + "step": 26898 + }, + { + "epoch": 1.15, + "grad_norm": 0.478515625, + "learning_rate": 7.750787209072646e-05, + "loss": 2.463, + "step": 26899 + }, + { + "epoch": 1.15, + "grad_norm": 0.48046875, + "learning_rate": 7.750133608754518e-05, + "loss": 2.4441, + "step": 26900 + }, + { + "epoch": 1.15, + "grad_norm": 0.484375, + "learning_rate": 7.749480018559969e-05, + "loss": 2.5029, + "step": 26901 + }, + { + "epoch": 1.15, + "grad_norm": 0.435546875, + "learning_rate": 7.748826438491937e-05, + "loss": 2.5198, + "step": 26902 + }, + { + "epoch": 1.15, + "grad_norm": 0.40625, + "learning_rate": 7.748172868553358e-05, + "loss": 2.5065, + "step": 26903 + }, + { + "epoch": 1.15, + "grad_norm": 0.462890625, + "learning_rate": 7.747519308747183e-05, + "loss": 2.4818, + "step": 26904 + }, + { + "epoch": 1.15, + "grad_norm": 0.44140625, + "learning_rate": 7.746865759076347e-05, + "loss": 2.5195, + "step": 26905 + }, + { + "epoch": 1.15, + "grad_norm": 0.498046875, + "learning_rate": 7.746212219543792e-05, + "loss": 2.5591, + "step": 26906 + }, + { + "epoch": 1.15, + "grad_norm": 0.498046875, + "learning_rate": 7.745558690152454e-05, + "loss": 2.4779, + "step": 26907 + }, + { + "epoch": 1.15, + "grad_norm": 0.4453125, + "learning_rate": 7.744905170905283e-05, + "loss": 2.3248, + "step": 26908 + }, + { + "epoch": 1.15, + "grad_norm": 0.45703125, + "learning_rate": 7.744251661805215e-05, + "loss": 2.4425, + "step": 26909 + }, + { + "epoch": 1.15, + "grad_norm": 0.53515625, + "learning_rate": 7.74359816285519e-05, + "loss": 2.5318, + "step": 26910 + }, + { + "epoch": 1.15, + "grad_norm": 0.453125, + "learning_rate": 7.742944674058145e-05, + "loss": 2.5354, + "step": 26911 + }, + { + "epoch": 1.15, + "grad_norm": 0.53125, + "learning_rate": 7.742291195417027e-05, + "loss": 2.4351, + "step": 26912 + }, + { + "epoch": 1.15, + "grad_norm": 0.478515625, + "learning_rate": 7.741637726934771e-05, + "loss": 2.4224, + "step": 26913 + }, + { + "epoch": 1.15, + "grad_norm": 0.50390625, + "learning_rate": 7.740984268614323e-05, + "loss": 2.4728, + "step": 26914 + }, + { + "epoch": 1.15, + "grad_norm": 0.4296875, + "learning_rate": 7.74033082045862e-05, + "loss": 2.4311, + "step": 26915 + }, + { + "epoch": 1.15, + "grad_norm": 0.41796875, + "learning_rate": 7.739677382470598e-05, + "loss": 2.423, + "step": 26916 + }, + { + "epoch": 1.15, + "grad_norm": 0.484375, + "learning_rate": 7.739023954653207e-05, + "loss": 2.4344, + "step": 26917 + }, + { + "epoch": 1.15, + "grad_norm": 0.5, + "learning_rate": 7.738370537009381e-05, + "loss": 2.4904, + "step": 26918 + }, + { + "epoch": 1.15, + "grad_norm": 0.453125, + "learning_rate": 7.73771712954206e-05, + "loss": 2.2952, + "step": 26919 + }, + { + "epoch": 1.15, + "grad_norm": 0.466796875, + "learning_rate": 7.737063732254184e-05, + "loss": 2.336, + "step": 26920 + }, + { + "epoch": 1.15, + "grad_norm": 0.451171875, + "learning_rate": 7.736410345148695e-05, + "loss": 2.5157, + "step": 26921 + }, + { + "epoch": 1.15, + "grad_norm": 0.427734375, + "learning_rate": 7.735756968228532e-05, + "loss": 2.3998, + "step": 26922 + }, + { + "epoch": 1.15, + "grad_norm": 0.478515625, + "learning_rate": 7.735103601496638e-05, + "loss": 2.5036, + "step": 26923 + }, + { + "epoch": 1.15, + "grad_norm": 0.490234375, + "learning_rate": 7.734450244955944e-05, + "loss": 2.3167, + "step": 26924 + }, + { + "epoch": 1.15, + "grad_norm": 0.47265625, + "learning_rate": 7.733796898609402e-05, + "loss": 2.4089, + "step": 26925 + }, + { + "epoch": 1.15, + "grad_norm": 0.45703125, + "learning_rate": 7.733143562459943e-05, + "loss": 2.3487, + "step": 26926 + }, + { + "epoch": 1.15, + "grad_norm": 0.4609375, + "learning_rate": 7.732490236510512e-05, + "loss": 2.4772, + "step": 26927 + }, + { + "epoch": 1.15, + "grad_norm": 0.43359375, + "learning_rate": 7.731836920764042e-05, + "loss": 2.4903, + "step": 26928 + }, + { + "epoch": 1.15, + "grad_norm": 0.51171875, + "learning_rate": 7.731183615223478e-05, + "loss": 2.5048, + "step": 26929 + }, + { + "epoch": 1.15, + "grad_norm": 0.46875, + "learning_rate": 7.730530319891761e-05, + "loss": 2.3847, + "step": 26930 + }, + { + "epoch": 1.15, + "grad_norm": 0.5234375, + "learning_rate": 7.729877034771829e-05, + "loss": 2.4319, + "step": 26931 + }, + { + "epoch": 1.15, + "grad_norm": 0.48828125, + "learning_rate": 7.729223759866616e-05, + "loss": 2.6268, + "step": 26932 + }, + { + "epoch": 1.15, + "grad_norm": 0.462890625, + "learning_rate": 7.728570495179071e-05, + "loss": 2.5095, + "step": 26933 + }, + { + "epoch": 1.15, + "grad_norm": 0.56640625, + "learning_rate": 7.727917240712126e-05, + "loss": 2.5432, + "step": 26934 + }, + { + "epoch": 1.15, + "grad_norm": 0.4765625, + "learning_rate": 7.727263996468723e-05, + "loss": 2.5216, + "step": 26935 + }, + { + "epoch": 1.15, + "grad_norm": 0.51171875, + "learning_rate": 7.7266107624518e-05, + "loss": 2.5613, + "step": 26936 + }, + { + "epoch": 1.15, + "grad_norm": 0.447265625, + "learning_rate": 7.725957538664301e-05, + "loss": 2.3625, + "step": 26937 + }, + { + "epoch": 1.15, + "grad_norm": 0.484375, + "learning_rate": 7.72530432510916e-05, + "loss": 2.4913, + "step": 26938 + }, + { + "epoch": 1.15, + "grad_norm": 0.427734375, + "learning_rate": 7.724651121789321e-05, + "loss": 2.3303, + "step": 26939 + }, + { + "epoch": 1.15, + "grad_norm": 0.494140625, + "learning_rate": 7.723997928707715e-05, + "loss": 2.4404, + "step": 26940 + }, + { + "epoch": 1.15, + "grad_norm": 0.494140625, + "learning_rate": 7.723344745867292e-05, + "loss": 2.5041, + "step": 26941 + }, + { + "epoch": 1.15, + "grad_norm": 0.49609375, + "learning_rate": 7.722691573270983e-05, + "loss": 2.5286, + "step": 26942 + }, + { + "epoch": 1.15, + "grad_norm": 0.5234375, + "learning_rate": 7.722038410921732e-05, + "loss": 2.5041, + "step": 26943 + }, + { + "epoch": 1.15, + "grad_norm": 0.52734375, + "learning_rate": 7.72138525882247e-05, + "loss": 2.5209, + "step": 26944 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.720732116976148e-05, + "loss": 2.2866, + "step": 26945 + }, + { + "epoch": 1.15, + "grad_norm": 0.5625, + "learning_rate": 7.720078985385695e-05, + "loss": 2.4563, + "step": 26946 + }, + { + "epoch": 1.15, + "grad_norm": 0.421875, + "learning_rate": 7.719425864054058e-05, + "loss": 2.4727, + "step": 26947 + }, + { + "epoch": 1.15, + "grad_norm": 0.4609375, + "learning_rate": 7.718772752984162e-05, + "loss": 2.4795, + "step": 26948 + }, + { + "epoch": 1.15, + "grad_norm": 0.431640625, + "learning_rate": 7.718119652178964e-05, + "loss": 2.4803, + "step": 26949 + }, + { + "epoch": 1.15, + "grad_norm": 0.48046875, + "learning_rate": 7.717466561641389e-05, + "loss": 2.5239, + "step": 26950 + }, + { + "epoch": 1.15, + "grad_norm": 0.416015625, + "learning_rate": 7.716813481374382e-05, + "loss": 2.3618, + "step": 26951 + }, + { + "epoch": 1.15, + "grad_norm": 0.53515625, + "learning_rate": 7.716160411380877e-05, + "loss": 2.4557, + "step": 26952 + }, + { + "epoch": 1.15, + "grad_norm": 0.447265625, + "learning_rate": 7.715507351663817e-05, + "loss": 2.6263, + "step": 26953 + }, + { + "epoch": 1.15, + "grad_norm": 0.427734375, + "learning_rate": 7.714854302226141e-05, + "loss": 2.3569, + "step": 26954 + }, + { + "epoch": 1.15, + "grad_norm": 0.486328125, + "learning_rate": 7.714201263070787e-05, + "loss": 2.5267, + "step": 26955 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.713548234200684e-05, + "loss": 2.5112, + "step": 26956 + }, + { + "epoch": 1.15, + "grad_norm": 0.5, + "learning_rate": 7.712895215618788e-05, + "loss": 2.4125, + "step": 26957 + }, + { + "epoch": 1.15, + "grad_norm": 0.451171875, + "learning_rate": 7.712242207328021e-05, + "loss": 2.2926, + "step": 26958 + }, + { + "epoch": 1.15, + "grad_norm": 0.490234375, + "learning_rate": 7.711589209331332e-05, + "loss": 2.4684, + "step": 26959 + }, + { + "epoch": 1.15, + "grad_norm": 0.423828125, + "learning_rate": 7.710936221631652e-05, + "loss": 2.4754, + "step": 26960 + }, + { + "epoch": 1.15, + "grad_norm": 0.47265625, + "learning_rate": 7.710283244231924e-05, + "loss": 2.5492, + "step": 26961 + }, + { + "epoch": 1.15, + "grad_norm": 0.53515625, + "learning_rate": 7.709630277135085e-05, + "loss": 2.5731, + "step": 26962 + }, + { + "epoch": 1.15, + "grad_norm": 0.455078125, + "learning_rate": 7.708977320344075e-05, + "loss": 2.4256, + "step": 26963 + }, + { + "epoch": 1.15, + "grad_norm": 0.486328125, + "learning_rate": 7.708324373861822e-05, + "loss": 2.651, + "step": 26964 + }, + { + "epoch": 1.15, + "grad_norm": 0.427734375, + "learning_rate": 7.70767143769128e-05, + "loss": 2.3815, + "step": 26965 + }, + { + "epoch": 1.15, + "grad_norm": 0.458984375, + "learning_rate": 7.707018511835376e-05, + "loss": 2.3783, + "step": 26966 + }, + { + "epoch": 1.15, + "grad_norm": 0.447265625, + "learning_rate": 7.70636559629705e-05, + "loss": 2.6259, + "step": 26967 + }, + { + "epoch": 1.15, + "grad_norm": 0.4296875, + "learning_rate": 7.705712691079239e-05, + "loss": 2.5657, + "step": 26968 + }, + { + "epoch": 1.15, + "grad_norm": 0.52734375, + "learning_rate": 7.705059796184883e-05, + "loss": 2.4335, + "step": 26969 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.704406911616921e-05, + "loss": 2.5328, + "step": 26970 + }, + { + "epoch": 1.15, + "grad_norm": 0.423828125, + "learning_rate": 7.703754037378291e-05, + "loss": 2.3618, + "step": 26971 + }, + { + "epoch": 1.15, + "grad_norm": 0.46484375, + "learning_rate": 7.703101173471922e-05, + "loss": 2.487, + "step": 26972 + }, + { + "epoch": 1.15, + "grad_norm": 0.52734375, + "learning_rate": 7.702448319900766e-05, + "loss": 2.4723, + "step": 26973 + }, + { + "epoch": 1.15, + "grad_norm": 0.494140625, + "learning_rate": 7.701795476667749e-05, + "loss": 2.5376, + "step": 26974 + }, + { + "epoch": 1.15, + "grad_norm": 0.51953125, + "learning_rate": 7.701142643775814e-05, + "loss": 2.5839, + "step": 26975 + }, + { + "epoch": 1.15, + "grad_norm": 0.43359375, + "learning_rate": 7.700489821227892e-05, + "loss": 2.4195, + "step": 26976 + }, + { + "epoch": 1.15, + "grad_norm": 0.46484375, + "learning_rate": 7.69983700902693e-05, + "loss": 2.4376, + "step": 26977 + }, + { + "epoch": 1.15, + "grad_norm": 0.455078125, + "learning_rate": 7.69918420717586e-05, + "loss": 2.6482, + "step": 26978 + }, + { + "epoch": 1.15, + "grad_norm": 0.50390625, + "learning_rate": 7.698531415677623e-05, + "loss": 2.4688, + "step": 26979 + }, + { + "epoch": 1.15, + "grad_norm": 0.455078125, + "learning_rate": 7.697878634535147e-05, + "loss": 2.49, + "step": 26980 + }, + { + "epoch": 1.15, + "grad_norm": 0.486328125, + "learning_rate": 7.697225863751383e-05, + "loss": 2.5003, + "step": 26981 + }, + { + "epoch": 1.15, + "grad_norm": 0.455078125, + "learning_rate": 7.696573103329257e-05, + "loss": 2.3718, + "step": 26982 + }, + { + "epoch": 1.15, + "grad_norm": 0.455078125, + "learning_rate": 7.695920353271711e-05, + "loss": 2.4374, + "step": 26983 + }, + { + "epoch": 1.15, + "grad_norm": 0.490234375, + "learning_rate": 7.695267613581679e-05, + "loss": 2.4851, + "step": 26984 + }, + { + "epoch": 1.15, + "grad_norm": 0.498046875, + "learning_rate": 7.694614884262104e-05, + "loss": 2.5319, + "step": 26985 + }, + { + "epoch": 1.15, + "grad_norm": 0.515625, + "learning_rate": 7.693962165315918e-05, + "loss": 2.4642, + "step": 26986 + }, + { + "epoch": 1.15, + "grad_norm": 0.47265625, + "learning_rate": 7.693309456746062e-05, + "loss": 2.3605, + "step": 26987 + }, + { + "epoch": 1.15, + "grad_norm": 0.43359375, + "learning_rate": 7.692656758555465e-05, + "loss": 2.4954, + "step": 26988 + }, + { + "epoch": 1.15, + "grad_norm": 0.431640625, + "learning_rate": 7.692004070747074e-05, + "loss": 2.4892, + "step": 26989 + }, + { + "epoch": 1.15, + "grad_norm": 0.474609375, + "learning_rate": 7.69135139332382e-05, + "loss": 2.3716, + "step": 26990 + }, + { + "epoch": 1.15, + "grad_norm": 0.470703125, + "learning_rate": 7.69069872628864e-05, + "loss": 2.4522, + "step": 26991 + }, + { + "epoch": 1.15, + "grad_norm": 0.53125, + "learning_rate": 7.690046069644472e-05, + "loss": 2.5151, + "step": 26992 + }, + { + "epoch": 1.15, + "grad_norm": 0.455078125, + "learning_rate": 7.689393423394254e-05, + "loss": 2.5427, + "step": 26993 + }, + { + "epoch": 1.15, + "grad_norm": 0.400390625, + "learning_rate": 7.688740787540919e-05, + "loss": 2.5599, + "step": 26994 + }, + { + "epoch": 1.15, + "grad_norm": 0.421875, + "learning_rate": 7.688088162087409e-05, + "loss": 2.3837, + "step": 26995 + }, + { + "epoch": 1.15, + "grad_norm": 0.4921875, + "learning_rate": 7.687435547036652e-05, + "loss": 2.518, + "step": 26996 + }, + { + "epoch": 1.15, + "grad_norm": 0.494140625, + "learning_rate": 7.686782942391595e-05, + "loss": 2.4965, + "step": 26997 + }, + { + "epoch": 1.15, + "grad_norm": 0.466796875, + "learning_rate": 7.686130348155168e-05, + "loss": 2.4529, + "step": 26998 + }, + { + "epoch": 1.15, + "grad_norm": 0.490234375, + "learning_rate": 7.685477764330308e-05, + "loss": 2.4297, + "step": 26999 + }, + { + "epoch": 1.15, + "grad_norm": 0.478515625, + "learning_rate": 7.684825190919951e-05, + "loss": 2.615, + "step": 27000 + }, + { + "epoch": 1.15, + "grad_norm": 0.482421875, + "learning_rate": 7.684172627927036e-05, + "loss": 2.4528, + "step": 27001 + }, + { + "epoch": 1.15, + "grad_norm": 0.466796875, + "learning_rate": 7.683520075354499e-05, + "loss": 2.4715, + "step": 27002 + }, + { + "epoch": 1.15, + "grad_norm": 0.48828125, + "learning_rate": 7.682867533205276e-05, + "loss": 2.4609, + "step": 27003 + }, + { + "epoch": 1.15, + "grad_norm": 0.45703125, + "learning_rate": 7.682215001482297e-05, + "loss": 2.3784, + "step": 27004 + }, + { + "epoch": 1.15, + "grad_norm": 0.50390625, + "learning_rate": 7.681562480188509e-05, + "loss": 2.5851, + "step": 27005 + }, + { + "epoch": 1.15, + "grad_norm": 0.484375, + "learning_rate": 7.68090996932684e-05, + "loss": 2.3557, + "step": 27006 + }, + { + "epoch": 1.15, + "grad_norm": 0.50390625, + "learning_rate": 7.680257468900229e-05, + "loss": 2.4458, + "step": 27007 + }, + { + "epoch": 1.15, + "grad_norm": 0.447265625, + "learning_rate": 7.67960497891161e-05, + "loss": 2.4039, + "step": 27008 + }, + { + "epoch": 1.15, + "grad_norm": 0.42578125, + "learning_rate": 7.678952499363923e-05, + "loss": 2.4842, + "step": 27009 + }, + { + "epoch": 1.15, + "grad_norm": 0.462890625, + "learning_rate": 7.6783000302601e-05, + "loss": 2.4895, + "step": 27010 + }, + { + "epoch": 1.15, + "grad_norm": 0.48046875, + "learning_rate": 7.677647571603082e-05, + "loss": 2.3129, + "step": 27011 + }, + { + "epoch": 1.15, + "grad_norm": 0.47265625, + "learning_rate": 7.676995123395793e-05, + "loss": 2.3183, + "step": 27012 + }, + { + "epoch": 1.15, + "grad_norm": 0.48828125, + "learning_rate": 7.676342685641185e-05, + "loss": 2.4286, + "step": 27013 + }, + { + "epoch": 1.15, + "grad_norm": 0.51953125, + "learning_rate": 7.675690258342183e-05, + "loss": 2.4348, + "step": 27014 + }, + { + "epoch": 1.15, + "grad_norm": 0.43359375, + "learning_rate": 7.675037841501727e-05, + "loss": 2.3607, + "step": 27015 + }, + { + "epoch": 1.15, + "grad_norm": 0.4609375, + "learning_rate": 7.674385435122747e-05, + "loss": 2.3493, + "step": 27016 + }, + { + "epoch": 1.15, + "grad_norm": 0.5390625, + "learning_rate": 7.673733039208185e-05, + "loss": 2.5168, + "step": 27017 + }, + { + "epoch": 1.15, + "grad_norm": 0.47265625, + "learning_rate": 7.673080653760975e-05, + "loss": 2.3567, + "step": 27018 + }, + { + "epoch": 1.15, + "grad_norm": 0.41015625, + "learning_rate": 7.672428278784053e-05, + "loss": 2.3401, + "step": 27019 + }, + { + "epoch": 1.15, + "grad_norm": 0.486328125, + "learning_rate": 7.671775914280346e-05, + "loss": 2.5209, + "step": 27020 + }, + { + "epoch": 1.15, + "grad_norm": 0.48828125, + "learning_rate": 7.671123560252804e-05, + "loss": 2.3701, + "step": 27021 + }, + { + "epoch": 1.15, + "grad_norm": 0.458984375, + "learning_rate": 7.670471216704353e-05, + "loss": 2.4304, + "step": 27022 + }, + { + "epoch": 1.15, + "grad_norm": 0.53515625, + "learning_rate": 7.669818883637929e-05, + "loss": 2.4391, + "step": 27023 + }, + { + "epoch": 1.15, + "grad_norm": 0.5078125, + "learning_rate": 7.669166561056467e-05, + "loss": 2.5531, + "step": 27024 + }, + { + "epoch": 1.15, + "grad_norm": 0.443359375, + "learning_rate": 7.668514248962905e-05, + "loss": 2.4542, + "step": 27025 + }, + { + "epoch": 1.15, + "grad_norm": 0.404296875, + "learning_rate": 7.667861947360178e-05, + "loss": 2.3666, + "step": 27026 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.667209656251223e-05, + "loss": 2.4108, + "step": 27027 + }, + { + "epoch": 1.15, + "grad_norm": 0.451171875, + "learning_rate": 7.666557375638965e-05, + "loss": 2.2971, + "step": 27028 + }, + { + "epoch": 1.15, + "grad_norm": 0.45703125, + "learning_rate": 7.66590510552635e-05, + "loss": 2.4226, + "step": 27029 + }, + { + "epoch": 1.15, + "grad_norm": 0.4921875, + "learning_rate": 7.665252845916309e-05, + "loss": 2.4625, + "step": 27030 + }, + { + "epoch": 1.15, + "grad_norm": 0.435546875, + "learning_rate": 7.664600596811778e-05, + "loss": 2.406, + "step": 27031 + }, + { + "epoch": 1.15, + "grad_norm": 0.4921875, + "learning_rate": 7.663948358215687e-05, + "loss": 2.3891, + "step": 27032 + }, + { + "epoch": 1.15, + "grad_norm": 0.396484375, + "learning_rate": 7.663296130130977e-05, + "loss": 2.4514, + "step": 27033 + }, + { + "epoch": 1.15, + "grad_norm": 0.439453125, + "learning_rate": 7.66264391256058e-05, + "loss": 2.4424, + "step": 27034 + }, + { + "epoch": 1.15, + "grad_norm": 0.50390625, + "learning_rate": 7.661991705507434e-05, + "loss": 2.5001, + "step": 27035 + }, + { + "epoch": 1.15, + "grad_norm": 0.57421875, + "learning_rate": 7.661339508974469e-05, + "loss": 2.5471, + "step": 27036 + }, + { + "epoch": 1.15, + "grad_norm": 0.4609375, + "learning_rate": 7.660687322964618e-05, + "loss": 2.3974, + "step": 27037 + }, + { + "epoch": 1.15, + "grad_norm": 0.490234375, + "learning_rate": 7.660035147480823e-05, + "loss": 2.5164, + "step": 27038 + }, + { + "epoch": 1.15, + "grad_norm": 0.51953125, + "learning_rate": 7.659382982526013e-05, + "loss": 2.6509, + "step": 27039 + }, + { + "epoch": 1.15, + "grad_norm": 0.453125, + "learning_rate": 7.658730828103125e-05, + "loss": 2.5047, + "step": 27040 + }, + { + "epoch": 1.15, + "grad_norm": 0.435546875, + "learning_rate": 7.658078684215091e-05, + "loss": 2.4285, + "step": 27041 + }, + { + "epoch": 1.15, + "grad_norm": 0.4921875, + "learning_rate": 7.657426550864849e-05, + "loss": 2.3851, + "step": 27042 + }, + { + "epoch": 1.15, + "grad_norm": 0.46875, + "learning_rate": 7.656774428055333e-05, + "loss": 2.4697, + "step": 27043 + }, + { + "epoch": 1.15, + "grad_norm": 0.4296875, + "learning_rate": 7.656122315789475e-05, + "loss": 2.3611, + "step": 27044 + }, + { + "epoch": 1.15, + "grad_norm": 0.466796875, + "learning_rate": 7.655470214070207e-05, + "loss": 2.5148, + "step": 27045 + }, + { + "epoch": 1.15, + "grad_norm": 0.4453125, + "learning_rate": 7.654818122900466e-05, + "loss": 2.4293, + "step": 27046 + }, + { + "epoch": 1.15, + "grad_norm": 0.490234375, + "learning_rate": 7.65416604228319e-05, + "loss": 2.449, + "step": 27047 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.653513972221307e-05, + "loss": 2.5563, + "step": 27048 + }, + { + "epoch": 1.15, + "grad_norm": 0.46484375, + "learning_rate": 7.652861912717753e-05, + "loss": 2.5451, + "step": 27049 + }, + { + "epoch": 1.15, + "grad_norm": 0.412109375, + "learning_rate": 7.652209863775464e-05, + "loss": 2.4452, + "step": 27050 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.651557825397375e-05, + "loss": 2.4383, + "step": 27051 + }, + { + "epoch": 1.15, + "grad_norm": 0.51171875, + "learning_rate": 7.650905797586416e-05, + "loss": 2.3861, + "step": 27052 + }, + { + "epoch": 1.15, + "grad_norm": 0.515625, + "learning_rate": 7.65025378034552e-05, + "loss": 2.5794, + "step": 27053 + }, + { + "epoch": 1.15, + "grad_norm": 0.470703125, + "learning_rate": 7.649601773677623e-05, + "loss": 2.4994, + "step": 27054 + }, + { + "epoch": 1.15, + "grad_norm": 0.51953125, + "learning_rate": 7.648949777585662e-05, + "loss": 2.5479, + "step": 27055 + }, + { + "epoch": 1.15, + "grad_norm": 0.453125, + "learning_rate": 7.648297792072567e-05, + "loss": 2.5261, + "step": 27056 + }, + { + "epoch": 1.15, + "grad_norm": 0.46484375, + "learning_rate": 7.64764581714127e-05, + "loss": 2.394, + "step": 27057 + }, + { + "epoch": 1.15, + "grad_norm": 0.515625, + "learning_rate": 7.646993852794709e-05, + "loss": 2.4649, + "step": 27058 + }, + { + "epoch": 1.15, + "grad_norm": 0.466796875, + "learning_rate": 7.64634189903582e-05, + "loss": 2.4299, + "step": 27059 + }, + { + "epoch": 1.15, + "grad_norm": 0.45703125, + "learning_rate": 7.645689955867527e-05, + "loss": 2.4, + "step": 27060 + }, + { + "epoch": 1.15, + "grad_norm": 0.46484375, + "learning_rate": 7.645038023292769e-05, + "loss": 2.4825, + "step": 27061 + }, + { + "epoch": 1.15, + "grad_norm": 0.466796875, + "learning_rate": 7.64438610131448e-05, + "loss": 2.3652, + "step": 27062 + }, + { + "epoch": 1.15, + "grad_norm": 0.5078125, + "learning_rate": 7.643734189935593e-05, + "loss": 2.4411, + "step": 27063 + }, + { + "epoch": 1.15, + "grad_norm": 0.478515625, + "learning_rate": 7.643082289159041e-05, + "loss": 2.5199, + "step": 27064 + }, + { + "epoch": 1.15, + "grad_norm": 0.50390625, + "learning_rate": 7.642430398987757e-05, + "loss": 2.3919, + "step": 27065 + }, + { + "epoch": 1.15, + "grad_norm": 0.49609375, + "learning_rate": 7.641778519424674e-05, + "loss": 2.6163, + "step": 27066 + }, + { + "epoch": 1.15, + "grad_norm": 0.462890625, + "learning_rate": 7.64112665047273e-05, + "loss": 2.4602, + "step": 27067 + }, + { + "epoch": 1.15, + "grad_norm": 0.419921875, + "learning_rate": 7.640474792134851e-05, + "loss": 2.4697, + "step": 27068 + }, + { + "epoch": 1.15, + "grad_norm": 0.451171875, + "learning_rate": 7.63982294441397e-05, + "loss": 2.4749, + "step": 27069 + }, + { + "epoch": 1.15, + "grad_norm": 0.48828125, + "learning_rate": 7.639171107313026e-05, + "loss": 2.5474, + "step": 27070 + }, + { + "epoch": 1.15, + "grad_norm": 0.48046875, + "learning_rate": 7.638519280834951e-05, + "loss": 2.3662, + "step": 27071 + }, + { + "epoch": 1.15, + "grad_norm": 0.52734375, + "learning_rate": 7.637867464982675e-05, + "loss": 2.5679, + "step": 27072 + }, + { + "epoch": 1.15, + "grad_norm": 0.443359375, + "learning_rate": 7.637215659759129e-05, + "loss": 2.2952, + "step": 27073 + }, + { + "epoch": 1.15, + "grad_norm": 0.4453125, + "learning_rate": 7.636563865167253e-05, + "loss": 2.4867, + "step": 27074 + }, + { + "epoch": 1.15, + "grad_norm": 0.470703125, + "learning_rate": 7.635912081209976e-05, + "loss": 2.4948, + "step": 27075 + }, + { + "epoch": 1.15, + "grad_norm": 0.478515625, + "learning_rate": 7.635260307890229e-05, + "loss": 2.3732, + "step": 27076 + }, + { + "epoch": 1.15, + "grad_norm": 0.486328125, + "learning_rate": 7.634608545210945e-05, + "loss": 2.4272, + "step": 27077 + }, + { + "epoch": 1.15, + "grad_norm": 0.439453125, + "learning_rate": 7.63395679317506e-05, + "loss": 2.4908, + "step": 27078 + }, + { + "epoch": 1.15, + "grad_norm": 0.4609375, + "learning_rate": 7.633305051785504e-05, + "loss": 2.359, + "step": 27079 + }, + { + "epoch": 1.15, + "grad_norm": 0.494140625, + "learning_rate": 7.632653321045211e-05, + "loss": 2.4664, + "step": 27080 + }, + { + "epoch": 1.15, + "grad_norm": 0.46484375, + "learning_rate": 7.632001600957111e-05, + "loss": 2.5, + "step": 27081 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.631349891524141e-05, + "loss": 2.3869, + "step": 27082 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.630698192749233e-05, + "loss": 2.5484, + "step": 27083 + }, + { + "epoch": 1.15, + "grad_norm": 0.486328125, + "learning_rate": 7.630046504635314e-05, + "loss": 2.4612, + "step": 27084 + }, + { + "epoch": 1.15, + "grad_norm": 0.50390625, + "learning_rate": 7.629394827185318e-05, + "loss": 2.4965, + "step": 27085 + }, + { + "epoch": 1.15, + "grad_norm": 0.474609375, + "learning_rate": 7.62874316040218e-05, + "loss": 2.3842, + "step": 27086 + }, + { + "epoch": 1.15, + "grad_norm": 0.4453125, + "learning_rate": 7.628091504288833e-05, + "loss": 2.409, + "step": 27087 + }, + { + "epoch": 1.15, + "grad_norm": 0.466796875, + "learning_rate": 7.627439858848207e-05, + "loss": 2.5298, + "step": 27088 + }, + { + "epoch": 1.15, + "grad_norm": 0.5, + "learning_rate": 7.626788224083232e-05, + "loss": 2.5701, + "step": 27089 + }, + { + "epoch": 1.15, + "grad_norm": 0.453125, + "learning_rate": 7.626136599996844e-05, + "loss": 2.386, + "step": 27090 + }, + { + "epoch": 1.15, + "grad_norm": 0.474609375, + "learning_rate": 7.625484986591978e-05, + "loss": 2.4451, + "step": 27091 + }, + { + "epoch": 1.15, + "grad_norm": 0.5390625, + "learning_rate": 7.624833383871558e-05, + "loss": 2.4415, + "step": 27092 + }, + { + "epoch": 1.15, + "grad_norm": 0.4453125, + "learning_rate": 7.624181791838518e-05, + "loss": 2.4766, + "step": 27093 + }, + { + "epoch": 1.15, + "grad_norm": 0.4296875, + "learning_rate": 7.623530210495793e-05, + "loss": 2.4716, + "step": 27094 + }, + { + "epoch": 1.15, + "grad_norm": 0.49609375, + "learning_rate": 7.622878639846314e-05, + "loss": 2.4287, + "step": 27095 + }, + { + "epoch": 1.15, + "grad_norm": 0.42578125, + "learning_rate": 7.622227079893013e-05, + "loss": 2.6842, + "step": 27096 + }, + { + "epoch": 1.15, + "grad_norm": 0.486328125, + "learning_rate": 7.621575530638818e-05, + "loss": 2.5753, + "step": 27097 + }, + { + "epoch": 1.15, + "grad_norm": 0.40625, + "learning_rate": 7.620923992086668e-05, + "loss": 2.4873, + "step": 27098 + }, + { + "epoch": 1.15, + "grad_norm": 0.45703125, + "learning_rate": 7.62027246423949e-05, + "loss": 2.3815, + "step": 27099 + }, + { + "epoch": 1.15, + "grad_norm": 0.427734375, + "learning_rate": 7.619620947100216e-05, + "loss": 2.4296, + "step": 27100 + }, + { + "epoch": 1.15, + "grad_norm": 0.443359375, + "learning_rate": 7.618969440671775e-05, + "loss": 2.5037, + "step": 27101 + }, + { + "epoch": 1.15, + "grad_norm": 0.43359375, + "learning_rate": 7.618317944957105e-05, + "loss": 2.4551, + "step": 27102 + }, + { + "epoch": 1.15, + "grad_norm": 0.44921875, + "learning_rate": 7.617666459959131e-05, + "loss": 2.6343, + "step": 27103 + }, + { + "epoch": 1.15, + "grad_norm": 0.4765625, + "learning_rate": 7.61701498568079e-05, + "loss": 2.586, + "step": 27104 + }, + { + "epoch": 1.16, + "grad_norm": 0.4296875, + "learning_rate": 7.616363522125007e-05, + "loss": 2.4784, + "step": 27105 + }, + { + "epoch": 1.16, + "grad_norm": 0.546875, + "learning_rate": 7.61571206929472e-05, + "loss": 2.4178, + "step": 27106 + }, + { + "epoch": 1.16, + "grad_norm": 0.474609375, + "learning_rate": 7.615060627192858e-05, + "loss": 2.5459, + "step": 27107 + }, + { + "epoch": 1.16, + "grad_norm": 0.49609375, + "learning_rate": 7.614409195822351e-05, + "loss": 2.5787, + "step": 27108 + }, + { + "epoch": 1.16, + "grad_norm": 0.494140625, + "learning_rate": 7.613757775186128e-05, + "loss": 2.5179, + "step": 27109 + }, + { + "epoch": 1.16, + "grad_norm": 0.419921875, + "learning_rate": 7.613106365287126e-05, + "loss": 2.4381, + "step": 27110 + }, + { + "epoch": 1.16, + "grad_norm": 0.4375, + "learning_rate": 7.612454966128272e-05, + "loss": 2.5124, + "step": 27111 + }, + { + "epoch": 1.16, + "grad_norm": 0.55078125, + "learning_rate": 7.611803577712497e-05, + "loss": 2.603, + "step": 27112 + }, + { + "epoch": 1.16, + "grad_norm": 0.484375, + "learning_rate": 7.611152200042733e-05, + "loss": 2.363, + "step": 27113 + }, + { + "epoch": 1.16, + "grad_norm": 0.484375, + "learning_rate": 7.610500833121913e-05, + "loss": 2.4258, + "step": 27114 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.609849476952968e-05, + "loss": 2.439, + "step": 27115 + }, + { + "epoch": 1.16, + "grad_norm": 0.431640625, + "learning_rate": 7.609198131538824e-05, + "loss": 2.42, + "step": 27116 + }, + { + "epoch": 1.16, + "grad_norm": 0.4296875, + "learning_rate": 7.608546796882413e-05, + "loss": 2.422, + "step": 27117 + }, + { + "epoch": 1.16, + "grad_norm": 0.474609375, + "learning_rate": 7.60789547298667e-05, + "loss": 2.3482, + "step": 27118 + }, + { + "epoch": 1.16, + "grad_norm": 0.47265625, + "learning_rate": 7.607244159854522e-05, + "loss": 2.4666, + "step": 27119 + }, + { + "epoch": 1.16, + "grad_norm": 0.5, + "learning_rate": 7.606592857488902e-05, + "loss": 2.5328, + "step": 27120 + }, + { + "epoch": 1.16, + "grad_norm": 0.515625, + "learning_rate": 7.605941565892736e-05, + "loss": 2.4324, + "step": 27121 + }, + { + "epoch": 1.16, + "grad_norm": 0.53125, + "learning_rate": 7.60529028506896e-05, + "loss": 2.5024, + "step": 27122 + }, + { + "epoch": 1.16, + "grad_norm": 0.47265625, + "learning_rate": 7.604639015020507e-05, + "loss": 2.4774, + "step": 27123 + }, + { + "epoch": 1.16, + "grad_norm": 0.458984375, + "learning_rate": 7.603987755750299e-05, + "loss": 2.3099, + "step": 27124 + }, + { + "epoch": 1.16, + "grad_norm": 0.56640625, + "learning_rate": 7.603336507261269e-05, + "loss": 2.4679, + "step": 27125 + }, + { + "epoch": 1.16, + "grad_norm": 0.5, + "learning_rate": 7.602685269556351e-05, + "loss": 2.3896, + "step": 27126 + }, + { + "epoch": 1.16, + "grad_norm": 0.51171875, + "learning_rate": 7.602034042638473e-05, + "loss": 2.4542, + "step": 27127 + }, + { + "epoch": 1.16, + "grad_norm": 0.51953125, + "learning_rate": 7.601382826510566e-05, + "loss": 2.5024, + "step": 27128 + }, + { + "epoch": 1.16, + "grad_norm": 0.45703125, + "learning_rate": 7.600731621175556e-05, + "loss": 2.4315, + "step": 27129 + }, + { + "epoch": 1.16, + "grad_norm": 0.40625, + "learning_rate": 7.60008042663638e-05, + "loss": 2.4453, + "step": 27130 + }, + { + "epoch": 1.16, + "grad_norm": 0.51171875, + "learning_rate": 7.599429242895967e-05, + "loss": 2.4723, + "step": 27131 + }, + { + "epoch": 1.16, + "grad_norm": 0.4609375, + "learning_rate": 7.598778069957244e-05, + "loss": 2.4904, + "step": 27132 + }, + { + "epoch": 1.16, + "grad_norm": 0.53515625, + "learning_rate": 7.598126907823139e-05, + "loss": 2.555, + "step": 27133 + }, + { + "epoch": 1.16, + "grad_norm": 0.515625, + "learning_rate": 7.597475756496587e-05, + "loss": 2.4638, + "step": 27134 + }, + { + "epoch": 1.16, + "grad_norm": 0.46875, + "learning_rate": 7.596824615980517e-05, + "loss": 2.7252, + "step": 27135 + }, + { + "epoch": 1.16, + "grad_norm": 0.486328125, + "learning_rate": 7.596173486277857e-05, + "loss": 2.5167, + "step": 27136 + }, + { + "epoch": 1.16, + "grad_norm": 0.421875, + "learning_rate": 7.595522367391537e-05, + "loss": 2.2836, + "step": 27137 + }, + { + "epoch": 1.16, + "grad_norm": 0.515625, + "learning_rate": 7.594871259324488e-05, + "loss": 2.521, + "step": 27138 + }, + { + "epoch": 1.16, + "grad_norm": 0.46484375, + "learning_rate": 7.594220162079643e-05, + "loss": 2.521, + "step": 27139 + }, + { + "epoch": 1.16, + "grad_norm": 0.443359375, + "learning_rate": 7.593569075659924e-05, + "loss": 2.5051, + "step": 27140 + }, + { + "epoch": 1.16, + "grad_norm": 0.470703125, + "learning_rate": 7.592918000068264e-05, + "loss": 2.4447, + "step": 27141 + }, + { + "epoch": 1.16, + "grad_norm": 0.392578125, + "learning_rate": 7.592266935307596e-05, + "loss": 2.5058, + "step": 27142 + }, + { + "epoch": 1.16, + "grad_norm": 0.53125, + "learning_rate": 7.591615881380845e-05, + "loss": 2.645, + "step": 27143 + }, + { + "epoch": 1.16, + "grad_norm": 0.47265625, + "learning_rate": 7.590964838290944e-05, + "loss": 2.3851, + "step": 27144 + }, + { + "epoch": 1.16, + "grad_norm": 0.42578125, + "learning_rate": 7.590313806040818e-05, + "loss": 2.5218, + "step": 27145 + }, + { + "epoch": 1.16, + "grad_norm": 0.5078125, + "learning_rate": 7.589662784633402e-05, + "loss": 2.4884, + "step": 27146 + }, + { + "epoch": 1.16, + "grad_norm": 0.478515625, + "learning_rate": 7.589011774071623e-05, + "loss": 2.5349, + "step": 27147 + }, + { + "epoch": 1.16, + "grad_norm": 0.4140625, + "learning_rate": 7.58836077435841e-05, + "loss": 2.3843, + "step": 27148 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.587709785496689e-05, + "loss": 2.5109, + "step": 27149 + }, + { + "epoch": 1.16, + "grad_norm": 0.416015625, + "learning_rate": 7.587058807489396e-05, + "loss": 2.5281, + "step": 27150 + }, + { + "epoch": 1.16, + "grad_norm": 0.46875, + "learning_rate": 7.586407840339454e-05, + "loss": 2.5613, + "step": 27151 + }, + { + "epoch": 1.16, + "grad_norm": 0.404296875, + "learning_rate": 7.585756884049796e-05, + "loss": 2.3626, + "step": 27152 + }, + { + "epoch": 1.16, + "grad_norm": 0.48046875, + "learning_rate": 7.585105938623347e-05, + "loss": 2.4251, + "step": 27153 + }, + { + "epoch": 1.16, + "grad_norm": 0.474609375, + "learning_rate": 7.584455004063041e-05, + "loss": 2.485, + "step": 27154 + }, + { + "epoch": 1.16, + "grad_norm": 0.43359375, + "learning_rate": 7.583804080371809e-05, + "loss": 2.5059, + "step": 27155 + }, + { + "epoch": 1.16, + "grad_norm": 0.462890625, + "learning_rate": 7.583153167552571e-05, + "loss": 2.4263, + "step": 27156 + }, + { + "epoch": 1.16, + "grad_norm": 0.46875, + "learning_rate": 7.58250226560826e-05, + "loss": 2.4695, + "step": 27157 + }, + { + "epoch": 1.16, + "grad_norm": 0.478515625, + "learning_rate": 7.581851374541804e-05, + "loss": 2.4425, + "step": 27158 + }, + { + "epoch": 1.16, + "grad_norm": 0.4765625, + "learning_rate": 7.581200494356136e-05, + "loss": 2.4635, + "step": 27159 + }, + { + "epoch": 1.16, + "grad_norm": 0.453125, + "learning_rate": 7.580549625054181e-05, + "loss": 2.4174, + "step": 27160 + }, + { + "epoch": 1.16, + "grad_norm": 0.435546875, + "learning_rate": 7.57989876663887e-05, + "loss": 2.5603, + "step": 27161 + }, + { + "epoch": 1.16, + "grad_norm": 0.48828125, + "learning_rate": 7.579247919113126e-05, + "loss": 2.4067, + "step": 27162 + }, + { + "epoch": 1.16, + "grad_norm": 0.4765625, + "learning_rate": 7.578597082479887e-05, + "loss": 2.3838, + "step": 27163 + }, + { + "epoch": 1.16, + "grad_norm": 0.48046875, + "learning_rate": 7.577946256742074e-05, + "loss": 2.3891, + "step": 27164 + }, + { + "epoch": 1.16, + "grad_norm": 0.498046875, + "learning_rate": 7.577295441902617e-05, + "loss": 2.4155, + "step": 27165 + }, + { + "epoch": 1.16, + "grad_norm": 0.455078125, + "learning_rate": 7.576644637964444e-05, + "loss": 2.3567, + "step": 27166 + }, + { + "epoch": 1.16, + "grad_norm": 0.462890625, + "learning_rate": 7.575993844930484e-05, + "loss": 2.377, + "step": 27167 + }, + { + "epoch": 1.16, + "grad_norm": 0.455078125, + "learning_rate": 7.575343062803669e-05, + "loss": 2.5007, + "step": 27168 + }, + { + "epoch": 1.16, + "grad_norm": 0.515625, + "learning_rate": 7.574692291586925e-05, + "loss": 2.3242, + "step": 27169 + }, + { + "epoch": 1.16, + "grad_norm": 0.4609375, + "learning_rate": 7.574041531283174e-05, + "loss": 2.6064, + "step": 27170 + }, + { + "epoch": 1.16, + "grad_norm": 0.412109375, + "learning_rate": 7.573390781895354e-05, + "loss": 2.4372, + "step": 27171 + }, + { + "epoch": 1.16, + "grad_norm": 0.50390625, + "learning_rate": 7.572740043426387e-05, + "loss": 2.4497, + "step": 27172 + }, + { + "epoch": 1.16, + "grad_norm": 0.5234375, + "learning_rate": 7.572089315879204e-05, + "loss": 2.5788, + "step": 27173 + }, + { + "epoch": 1.16, + "grad_norm": 0.458984375, + "learning_rate": 7.571438599256728e-05, + "loss": 2.5546, + "step": 27174 + }, + { + "epoch": 1.16, + "grad_norm": 0.5078125, + "learning_rate": 7.570787893561895e-05, + "loss": 2.4872, + "step": 27175 + }, + { + "epoch": 1.16, + "grad_norm": 0.46484375, + "learning_rate": 7.570137198797627e-05, + "loss": 2.4512, + "step": 27176 + }, + { + "epoch": 1.16, + "grad_norm": 0.484375, + "learning_rate": 7.569486514966856e-05, + "loss": 2.4294, + "step": 27177 + }, + { + "epoch": 1.16, + "grad_norm": 0.50390625, + "learning_rate": 7.568835842072501e-05, + "loss": 2.4497, + "step": 27178 + }, + { + "epoch": 1.16, + "grad_norm": 0.486328125, + "learning_rate": 7.568185180117503e-05, + "loss": 2.6321, + "step": 27179 + }, + { + "epoch": 1.16, + "grad_norm": 0.427734375, + "learning_rate": 7.567534529104781e-05, + "loss": 2.393, + "step": 27180 + }, + { + "epoch": 1.16, + "grad_norm": 0.51171875, + "learning_rate": 7.566883889037264e-05, + "loss": 2.4071, + "step": 27181 + }, + { + "epoch": 1.16, + "grad_norm": 0.4609375, + "learning_rate": 7.566233259917878e-05, + "loss": 2.3609, + "step": 27182 + }, + { + "epoch": 1.16, + "grad_norm": 0.466796875, + "learning_rate": 7.565582641749556e-05, + "loss": 2.5823, + "step": 27183 + }, + { + "epoch": 1.16, + "grad_norm": 0.4765625, + "learning_rate": 7.564932034535222e-05, + "loss": 2.6499, + "step": 27184 + }, + { + "epoch": 1.16, + "grad_norm": 0.435546875, + "learning_rate": 7.564281438277806e-05, + "loss": 2.5215, + "step": 27185 + }, + { + "epoch": 1.16, + "grad_norm": 0.453125, + "learning_rate": 7.563630852980227e-05, + "loss": 2.4465, + "step": 27186 + }, + { + "epoch": 1.16, + "grad_norm": 0.46875, + "learning_rate": 7.562980278645425e-05, + "loss": 2.4017, + "step": 27187 + }, + { + "epoch": 1.16, + "grad_norm": 0.4765625, + "learning_rate": 7.56232971527632e-05, + "loss": 2.3956, + "step": 27188 + }, + { + "epoch": 1.16, + "grad_norm": 0.5625, + "learning_rate": 7.561679162875838e-05, + "loss": 2.584, + "step": 27189 + }, + { + "epoch": 1.16, + "grad_norm": 0.462890625, + "learning_rate": 7.561028621446908e-05, + "loss": 2.4283, + "step": 27190 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.56037809099246e-05, + "loss": 2.5158, + "step": 27191 + }, + { + "epoch": 1.16, + "grad_norm": 0.4375, + "learning_rate": 7.55972757151542e-05, + "loss": 2.4454, + "step": 27192 + }, + { + "epoch": 1.16, + "grad_norm": 0.46875, + "learning_rate": 7.559077063018716e-05, + "loss": 2.5187, + "step": 27193 + }, + { + "epoch": 1.16, + "grad_norm": 0.443359375, + "learning_rate": 7.558426565505266e-05, + "loss": 2.4235, + "step": 27194 + }, + { + "epoch": 1.16, + "grad_norm": 0.48046875, + "learning_rate": 7.557776078978009e-05, + "loss": 2.5958, + "step": 27195 + }, + { + "epoch": 1.16, + "grad_norm": 0.478515625, + "learning_rate": 7.557125603439868e-05, + "loss": 2.2808, + "step": 27196 + }, + { + "epoch": 1.16, + "grad_norm": 0.455078125, + "learning_rate": 7.556475138893767e-05, + "loss": 2.4586, + "step": 27197 + }, + { + "epoch": 1.16, + "grad_norm": 0.44921875, + "learning_rate": 7.555824685342634e-05, + "loss": 2.406, + "step": 27198 + }, + { + "epoch": 1.16, + "grad_norm": 0.427734375, + "learning_rate": 7.555174242789398e-05, + "loss": 2.5186, + "step": 27199 + }, + { + "epoch": 1.16, + "grad_norm": 0.48828125, + "learning_rate": 7.554523811236984e-05, + "loss": 2.4652, + "step": 27200 + }, + { + "epoch": 1.16, + "grad_norm": 0.5, + "learning_rate": 7.553873390688322e-05, + "loss": 2.5844, + "step": 27201 + }, + { + "epoch": 1.16, + "grad_norm": 0.43359375, + "learning_rate": 7.55322298114633e-05, + "loss": 2.5243, + "step": 27202 + }, + { + "epoch": 1.16, + "grad_norm": 0.43359375, + "learning_rate": 7.552572582613946e-05, + "loss": 2.4882, + "step": 27203 + }, + { + "epoch": 1.16, + "grad_norm": 0.5, + "learning_rate": 7.551922195094088e-05, + "loss": 2.3981, + "step": 27204 + }, + { + "epoch": 1.16, + "grad_norm": 0.51171875, + "learning_rate": 7.551271818589687e-05, + "loss": 2.4966, + "step": 27205 + }, + { + "epoch": 1.16, + "grad_norm": 0.44140625, + "learning_rate": 7.550621453103664e-05, + "loss": 2.4823, + "step": 27206 + }, + { + "epoch": 1.16, + "grad_norm": 0.4609375, + "learning_rate": 7.549971098638954e-05, + "loss": 2.5305, + "step": 27207 + }, + { + "epoch": 1.16, + "grad_norm": 0.498046875, + "learning_rate": 7.549320755198477e-05, + "loss": 2.676, + "step": 27208 + }, + { + "epoch": 1.16, + "grad_norm": 0.51953125, + "learning_rate": 7.548670422785163e-05, + "loss": 2.6168, + "step": 27209 + }, + { + "epoch": 1.16, + "grad_norm": 0.419921875, + "learning_rate": 7.54802010140193e-05, + "loss": 2.4952, + "step": 27210 + }, + { + "epoch": 1.16, + "grad_norm": 0.53125, + "learning_rate": 7.547369791051717e-05, + "loss": 2.5118, + "step": 27211 + }, + { + "epoch": 1.16, + "grad_norm": 0.44921875, + "learning_rate": 7.546719491737442e-05, + "loss": 2.3808, + "step": 27212 + }, + { + "epoch": 1.16, + "grad_norm": 0.494140625, + "learning_rate": 7.546069203462032e-05, + "loss": 2.5662, + "step": 27213 + }, + { + "epoch": 1.16, + "grad_norm": 0.486328125, + "learning_rate": 7.545418926228411e-05, + "loss": 2.5449, + "step": 27214 + }, + { + "epoch": 1.16, + "grad_norm": 0.56640625, + "learning_rate": 7.544768660039512e-05, + "loss": 2.4653, + "step": 27215 + }, + { + "epoch": 1.16, + "grad_norm": 0.51953125, + "learning_rate": 7.544118404898255e-05, + "loss": 2.3968, + "step": 27216 + }, + { + "epoch": 1.16, + "grad_norm": 0.5078125, + "learning_rate": 7.54346816080757e-05, + "loss": 2.414, + "step": 27217 + }, + { + "epoch": 1.16, + "grad_norm": 0.5, + "learning_rate": 7.542817927770375e-05, + "loss": 2.471, + "step": 27218 + }, + { + "epoch": 1.16, + "grad_norm": 0.4765625, + "learning_rate": 7.542167705789607e-05, + "loss": 2.5686, + "step": 27219 + }, + { + "epoch": 1.16, + "grad_norm": 0.490234375, + "learning_rate": 7.541517494868184e-05, + "loss": 2.487, + "step": 27220 + }, + { + "epoch": 1.16, + "grad_norm": 0.4375, + "learning_rate": 7.540867295009034e-05, + "loss": 2.5351, + "step": 27221 + }, + { + "epoch": 1.16, + "grad_norm": 0.447265625, + "learning_rate": 7.54021710621508e-05, + "loss": 2.4566, + "step": 27222 + }, + { + "epoch": 1.16, + "grad_norm": 0.515625, + "learning_rate": 7.539566928489253e-05, + "loss": 2.4817, + "step": 27223 + }, + { + "epoch": 1.16, + "grad_norm": 0.546875, + "learning_rate": 7.538916761834475e-05, + "loss": 2.4708, + "step": 27224 + }, + { + "epoch": 1.16, + "grad_norm": 0.423828125, + "learning_rate": 7.538266606253674e-05, + "loss": 2.3008, + "step": 27225 + }, + { + "epoch": 1.16, + "grad_norm": 0.443359375, + "learning_rate": 7.537616461749768e-05, + "loss": 2.3971, + "step": 27226 + }, + { + "epoch": 1.16, + "grad_norm": 0.48046875, + "learning_rate": 7.536966328325693e-05, + "loss": 2.4776, + "step": 27227 + }, + { + "epoch": 1.16, + "grad_norm": 0.48828125, + "learning_rate": 7.536316205984368e-05, + "loss": 2.4985, + "step": 27228 + }, + { + "epoch": 1.16, + "grad_norm": 0.46484375, + "learning_rate": 7.53566609472872e-05, + "loss": 2.4714, + "step": 27229 + }, + { + "epoch": 1.16, + "grad_norm": 0.453125, + "learning_rate": 7.535015994561673e-05, + "loss": 2.4286, + "step": 27230 + }, + { + "epoch": 1.16, + "grad_norm": 0.49609375, + "learning_rate": 7.534365905486153e-05, + "loss": 2.4614, + "step": 27231 + }, + { + "epoch": 1.16, + "grad_norm": 0.412109375, + "learning_rate": 7.533715827505086e-05, + "loss": 2.3039, + "step": 27232 + }, + { + "epoch": 1.16, + "grad_norm": 0.451171875, + "learning_rate": 7.5330657606214e-05, + "loss": 2.4668, + "step": 27233 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.532415704838009e-05, + "loss": 2.682, + "step": 27234 + }, + { + "epoch": 1.16, + "grad_norm": 0.5078125, + "learning_rate": 7.531765660157853e-05, + "loss": 2.528, + "step": 27235 + }, + { + "epoch": 1.16, + "grad_norm": 0.50390625, + "learning_rate": 7.531115626583846e-05, + "loss": 2.3516, + "step": 27236 + }, + { + "epoch": 1.16, + "grad_norm": 0.515625, + "learning_rate": 7.530465604118917e-05, + "loss": 2.5185, + "step": 27237 + }, + { + "epoch": 1.16, + "grad_norm": 0.4296875, + "learning_rate": 7.529815592765987e-05, + "loss": 2.5565, + "step": 27238 + }, + { + "epoch": 1.16, + "grad_norm": 0.4453125, + "learning_rate": 7.529165592527987e-05, + "loss": 2.4221, + "step": 27239 + }, + { + "epoch": 1.16, + "grad_norm": 0.443359375, + "learning_rate": 7.52851560340784e-05, + "loss": 2.5234, + "step": 27240 + }, + { + "epoch": 1.16, + "grad_norm": 0.494140625, + "learning_rate": 7.52786562540847e-05, + "loss": 2.4226, + "step": 27241 + }, + { + "epoch": 1.16, + "grad_norm": 0.4140625, + "learning_rate": 7.527215658532796e-05, + "loss": 2.5585, + "step": 27242 + }, + { + "epoch": 1.16, + "grad_norm": 0.478515625, + "learning_rate": 7.526565702783754e-05, + "loss": 2.4316, + "step": 27243 + }, + { + "epoch": 1.16, + "grad_norm": 0.48046875, + "learning_rate": 7.525915758164259e-05, + "loss": 2.4708, + "step": 27244 + }, + { + "epoch": 1.16, + "grad_norm": 0.46484375, + "learning_rate": 7.52526582467724e-05, + "loss": 2.3173, + "step": 27245 + }, + { + "epoch": 1.16, + "grad_norm": 0.453125, + "learning_rate": 7.524615902325618e-05, + "loss": 2.4465, + "step": 27246 + }, + { + "epoch": 1.16, + "grad_norm": 0.427734375, + "learning_rate": 7.523965991112323e-05, + "loss": 2.3121, + "step": 27247 + }, + { + "epoch": 1.16, + "grad_norm": 0.5078125, + "learning_rate": 7.523316091040276e-05, + "loss": 2.4104, + "step": 27248 + }, + { + "epoch": 1.16, + "grad_norm": 0.515625, + "learning_rate": 7.522666202112401e-05, + "loss": 2.3576, + "step": 27249 + }, + { + "epoch": 1.16, + "grad_norm": 0.5390625, + "learning_rate": 7.522016324331619e-05, + "loss": 2.5266, + "step": 27250 + }, + { + "epoch": 1.16, + "grad_norm": 0.44921875, + "learning_rate": 7.521366457700863e-05, + "loss": 2.4676, + "step": 27251 + }, + { + "epoch": 1.16, + "grad_norm": 0.466796875, + "learning_rate": 7.520716602223051e-05, + "loss": 2.4208, + "step": 27252 + }, + { + "epoch": 1.16, + "grad_norm": 0.490234375, + "learning_rate": 7.520066757901107e-05, + "loss": 2.5569, + "step": 27253 + }, + { + "epoch": 1.16, + "grad_norm": 0.42578125, + "learning_rate": 7.519416924737956e-05, + "loss": 2.5029, + "step": 27254 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.518767102736523e-05, + "loss": 2.5725, + "step": 27255 + }, + { + "epoch": 1.16, + "grad_norm": 0.43359375, + "learning_rate": 7.518117291899731e-05, + "loss": 2.3197, + "step": 27256 + }, + { + "epoch": 1.16, + "grad_norm": 0.44140625, + "learning_rate": 7.517467492230508e-05, + "loss": 2.4247, + "step": 27257 + }, + { + "epoch": 1.16, + "grad_norm": 0.4375, + "learning_rate": 7.516817703731765e-05, + "loss": 2.6014, + "step": 27258 + }, + { + "epoch": 1.16, + "grad_norm": 0.408203125, + "learning_rate": 7.516167926406444e-05, + "loss": 2.4314, + "step": 27259 + }, + { + "epoch": 1.16, + "grad_norm": 0.51171875, + "learning_rate": 7.515518160257455e-05, + "loss": 2.4454, + "step": 27260 + }, + { + "epoch": 1.16, + "grad_norm": 0.46484375, + "learning_rate": 7.514868405287728e-05, + "loss": 2.4215, + "step": 27261 + }, + { + "epoch": 1.16, + "grad_norm": 0.443359375, + "learning_rate": 7.514218661500183e-05, + "loss": 2.4484, + "step": 27262 + }, + { + "epoch": 1.16, + "grad_norm": 0.447265625, + "learning_rate": 7.513568928897746e-05, + "loss": 2.5353, + "step": 27263 + }, + { + "epoch": 1.16, + "grad_norm": 0.408203125, + "learning_rate": 7.512919207483342e-05, + "loss": 2.3636, + "step": 27264 + }, + { + "epoch": 1.16, + "grad_norm": 0.466796875, + "learning_rate": 7.512269497259894e-05, + "loss": 2.3717, + "step": 27265 + }, + { + "epoch": 1.16, + "grad_norm": 0.419921875, + "learning_rate": 7.511619798230318e-05, + "loss": 2.4456, + "step": 27266 + }, + { + "epoch": 1.16, + "grad_norm": 0.451171875, + "learning_rate": 7.510970110397548e-05, + "loss": 2.3684, + "step": 27267 + }, + { + "epoch": 1.16, + "grad_norm": 0.46875, + "learning_rate": 7.510320433764503e-05, + "loss": 2.5918, + "step": 27268 + }, + { + "epoch": 1.16, + "grad_norm": 0.40625, + "learning_rate": 7.509670768334105e-05, + "loss": 2.528, + "step": 27269 + }, + { + "epoch": 1.16, + "grad_norm": 0.439453125, + "learning_rate": 7.509021114109277e-05, + "loss": 2.474, + "step": 27270 + }, + { + "epoch": 1.16, + "grad_norm": 0.466796875, + "learning_rate": 7.508371471092945e-05, + "loss": 2.4931, + "step": 27271 + }, + { + "epoch": 1.16, + "grad_norm": 0.515625, + "learning_rate": 7.507721839288032e-05, + "loss": 2.5339, + "step": 27272 + }, + { + "epoch": 1.16, + "grad_norm": 0.44140625, + "learning_rate": 7.507072218697462e-05, + "loss": 2.451, + "step": 27273 + }, + { + "epoch": 1.16, + "grad_norm": 0.4765625, + "learning_rate": 7.506422609324148e-05, + "loss": 2.5655, + "step": 27274 + }, + { + "epoch": 1.16, + "grad_norm": 0.44140625, + "learning_rate": 7.505773011171029e-05, + "loss": 2.4303, + "step": 27275 + }, + { + "epoch": 1.16, + "grad_norm": 0.46875, + "learning_rate": 7.505123424241016e-05, + "loss": 2.3893, + "step": 27276 + }, + { + "epoch": 1.16, + "grad_norm": 0.478515625, + "learning_rate": 7.504473848537037e-05, + "loss": 2.1983, + "step": 27277 + }, + { + "epoch": 1.16, + "grad_norm": 0.47265625, + "learning_rate": 7.503824284062011e-05, + "loss": 2.5462, + "step": 27278 + }, + { + "epoch": 1.16, + "grad_norm": 0.455078125, + "learning_rate": 7.503174730818867e-05, + "loss": 2.412, + "step": 27279 + }, + { + "epoch": 1.16, + "grad_norm": 0.4765625, + "learning_rate": 7.502525188810523e-05, + "loss": 2.4389, + "step": 27280 + }, + { + "epoch": 1.16, + "grad_norm": 0.458984375, + "learning_rate": 7.501875658039906e-05, + "loss": 2.4559, + "step": 27281 + }, + { + "epoch": 1.16, + "grad_norm": 0.53125, + "learning_rate": 7.501226138509931e-05, + "loss": 2.3703, + "step": 27282 + }, + { + "epoch": 1.16, + "grad_norm": 0.51953125, + "learning_rate": 7.500576630223525e-05, + "loss": 2.6086, + "step": 27283 + }, + { + "epoch": 1.16, + "grad_norm": 0.47265625, + "learning_rate": 7.499927133183613e-05, + "loss": 2.4946, + "step": 27284 + }, + { + "epoch": 1.16, + "grad_norm": 0.466796875, + "learning_rate": 7.499277647393115e-05, + "loss": 2.4211, + "step": 27285 + }, + { + "epoch": 1.16, + "grad_norm": 0.443359375, + "learning_rate": 7.498628172854954e-05, + "loss": 2.4287, + "step": 27286 + }, + { + "epoch": 1.16, + "grad_norm": 0.4375, + "learning_rate": 7.497978709572048e-05, + "loss": 2.5651, + "step": 27287 + }, + { + "epoch": 1.16, + "grad_norm": 0.4609375, + "learning_rate": 7.497329257547328e-05, + "loss": 2.585, + "step": 27288 + }, + { + "epoch": 1.16, + "grad_norm": 0.412109375, + "learning_rate": 7.496679816783714e-05, + "loss": 2.3626, + "step": 27289 + }, + { + "epoch": 1.16, + "grad_norm": 0.431640625, + "learning_rate": 7.496030387284121e-05, + "loss": 2.4365, + "step": 27290 + }, + { + "epoch": 1.16, + "grad_norm": 0.498046875, + "learning_rate": 7.495380969051476e-05, + "loss": 2.4758, + "step": 27291 + }, + { + "epoch": 1.16, + "grad_norm": 0.443359375, + "learning_rate": 7.494731562088704e-05, + "loss": 2.4112, + "step": 27292 + }, + { + "epoch": 1.16, + "grad_norm": 0.4609375, + "learning_rate": 7.494082166398724e-05, + "loss": 2.4445, + "step": 27293 + }, + { + "epoch": 1.16, + "grad_norm": 0.458984375, + "learning_rate": 7.493432781984458e-05, + "loss": 2.5366, + "step": 27294 + }, + { + "epoch": 1.16, + "grad_norm": 0.44921875, + "learning_rate": 7.492783408848826e-05, + "loss": 2.4079, + "step": 27295 + }, + { + "epoch": 1.16, + "grad_norm": 0.451171875, + "learning_rate": 7.492134046994755e-05, + "loss": 2.5247, + "step": 27296 + }, + { + "epoch": 1.16, + "grad_norm": 0.486328125, + "learning_rate": 7.491484696425166e-05, + "loss": 2.4771, + "step": 27297 + }, + { + "epoch": 1.16, + "grad_norm": 0.435546875, + "learning_rate": 7.490835357142977e-05, + "loss": 2.48, + "step": 27298 + }, + { + "epoch": 1.16, + "grad_norm": 0.5078125, + "learning_rate": 7.490186029151109e-05, + "loss": 2.4645, + "step": 27299 + }, + { + "epoch": 1.16, + "grad_norm": 0.494140625, + "learning_rate": 7.48953671245249e-05, + "loss": 2.5952, + "step": 27300 + }, + { + "epoch": 1.16, + "grad_norm": 0.51953125, + "learning_rate": 7.488887407050038e-05, + "loss": 2.4133, + "step": 27301 + }, + { + "epoch": 1.16, + "grad_norm": 0.478515625, + "learning_rate": 7.488238112946675e-05, + "loss": 2.5034, + "step": 27302 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.487588830145322e-05, + "loss": 2.4054, + "step": 27303 + }, + { + "epoch": 1.16, + "grad_norm": 0.47265625, + "learning_rate": 7.4869395586489e-05, + "loss": 2.4426, + "step": 27304 + }, + { + "epoch": 1.16, + "grad_norm": 0.48046875, + "learning_rate": 7.486290298460337e-05, + "loss": 2.4691, + "step": 27305 + }, + { + "epoch": 1.16, + "grad_norm": 0.4140625, + "learning_rate": 7.485641049582544e-05, + "loss": 2.3284, + "step": 27306 + }, + { + "epoch": 1.16, + "grad_norm": 0.4765625, + "learning_rate": 7.484991812018445e-05, + "loss": 2.4216, + "step": 27307 + }, + { + "epoch": 1.16, + "grad_norm": 0.51171875, + "learning_rate": 7.48434258577097e-05, + "loss": 2.6189, + "step": 27308 + }, + { + "epoch": 1.16, + "grad_norm": 0.51171875, + "learning_rate": 7.48369337084303e-05, + "loss": 2.4603, + "step": 27309 + }, + { + "epoch": 1.16, + "grad_norm": 0.478515625, + "learning_rate": 7.483044167237551e-05, + "loss": 2.5004, + "step": 27310 + }, + { + "epoch": 1.16, + "grad_norm": 0.4609375, + "learning_rate": 7.482394974957453e-05, + "loss": 2.4999, + "step": 27311 + }, + { + "epoch": 1.16, + "grad_norm": 0.45703125, + "learning_rate": 7.481745794005659e-05, + "loss": 2.4718, + "step": 27312 + }, + { + "epoch": 1.16, + "grad_norm": 0.4375, + "learning_rate": 7.481096624385089e-05, + "loss": 2.374, + "step": 27313 + }, + { + "epoch": 1.16, + "grad_norm": 0.4453125, + "learning_rate": 7.480447466098663e-05, + "loss": 2.4741, + "step": 27314 + }, + { + "epoch": 1.16, + "grad_norm": 0.5, + "learning_rate": 7.479798319149298e-05, + "loss": 2.4839, + "step": 27315 + }, + { + "epoch": 1.16, + "grad_norm": 0.41796875, + "learning_rate": 7.479149183539927e-05, + "loss": 2.4042, + "step": 27316 + }, + { + "epoch": 1.16, + "grad_norm": 0.48046875, + "learning_rate": 7.47850005927346e-05, + "loss": 2.508, + "step": 27317 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.477850946352822e-05, + "loss": 2.6038, + "step": 27318 + }, + { + "epoch": 1.16, + "grad_norm": 0.52734375, + "learning_rate": 7.47720184478093e-05, + "loss": 2.415, + "step": 27319 + }, + { + "epoch": 1.16, + "grad_norm": 0.490234375, + "learning_rate": 7.47655275456071e-05, + "loss": 2.5427, + "step": 27320 + }, + { + "epoch": 1.16, + "grad_norm": 0.5234375, + "learning_rate": 7.475903675695083e-05, + "loss": 2.4341, + "step": 27321 + }, + { + "epoch": 1.16, + "grad_norm": 0.453125, + "learning_rate": 7.475254608186964e-05, + "loss": 2.5092, + "step": 27322 + }, + { + "epoch": 1.16, + "grad_norm": 0.451171875, + "learning_rate": 7.474605552039273e-05, + "loss": 2.4528, + "step": 27323 + }, + { + "epoch": 1.16, + "grad_norm": 0.421875, + "learning_rate": 7.47395650725494e-05, + "loss": 2.5014, + "step": 27324 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.473307473836879e-05, + "loss": 2.4862, + "step": 27325 + }, + { + "epoch": 1.16, + "grad_norm": 0.47265625, + "learning_rate": 7.472658451788009e-05, + "loss": 2.3974, + "step": 27326 + }, + { + "epoch": 1.16, + "grad_norm": 0.484375, + "learning_rate": 7.472009441111252e-05, + "loss": 2.5456, + "step": 27327 + }, + { + "epoch": 1.16, + "grad_norm": 0.447265625, + "learning_rate": 7.47136044180953e-05, + "loss": 2.4768, + "step": 27328 + }, + { + "epoch": 1.16, + "grad_norm": 0.458984375, + "learning_rate": 7.470711453885763e-05, + "loss": 2.3146, + "step": 27329 + }, + { + "epoch": 1.16, + "grad_norm": 0.466796875, + "learning_rate": 7.47006247734287e-05, + "loss": 2.5863, + "step": 27330 + }, + { + "epoch": 1.16, + "grad_norm": 0.466796875, + "learning_rate": 7.469413512183766e-05, + "loss": 2.4657, + "step": 27331 + }, + { + "epoch": 1.16, + "grad_norm": 0.466796875, + "learning_rate": 7.468764558411381e-05, + "loss": 2.5981, + "step": 27332 + }, + { + "epoch": 1.16, + "grad_norm": 0.47265625, + "learning_rate": 7.46811561602863e-05, + "loss": 2.3392, + "step": 27333 + }, + { + "epoch": 1.16, + "grad_norm": 0.44140625, + "learning_rate": 7.467466685038433e-05, + "loss": 2.4355, + "step": 27334 + }, + { + "epoch": 1.16, + "grad_norm": 0.486328125, + "learning_rate": 7.466817765443708e-05, + "loss": 2.3588, + "step": 27335 + }, + { + "epoch": 1.16, + "grad_norm": 0.4921875, + "learning_rate": 7.466168857247382e-05, + "loss": 2.5457, + "step": 27336 + }, + { + "epoch": 1.16, + "grad_norm": 0.40625, + "learning_rate": 7.46551996045237e-05, + "loss": 2.4714, + "step": 27337 + }, + { + "epoch": 1.16, + "grad_norm": 0.45703125, + "learning_rate": 7.46487107506159e-05, + "loss": 2.4107, + "step": 27338 + }, + { + "epoch": 1.16, + "grad_norm": 0.458984375, + "learning_rate": 7.46422220107796e-05, + "loss": 2.4826, + "step": 27339 + }, + { + "epoch": 1.17, + "grad_norm": 0.4765625, + "learning_rate": 7.463573338504408e-05, + "loss": 2.6587, + "step": 27340 + }, + { + "epoch": 1.17, + "grad_norm": 0.466796875, + "learning_rate": 7.462924487343848e-05, + "loss": 2.5196, + "step": 27341 + }, + { + "epoch": 1.17, + "grad_norm": 0.44921875, + "learning_rate": 7.4622756475992e-05, + "loss": 2.3959, + "step": 27342 + }, + { + "epoch": 1.17, + "grad_norm": 0.47265625, + "learning_rate": 7.461626819273384e-05, + "loss": 2.4871, + "step": 27343 + }, + { + "epoch": 1.17, + "grad_norm": 0.48828125, + "learning_rate": 7.46097800236932e-05, + "loss": 2.4239, + "step": 27344 + }, + { + "epoch": 1.17, + "grad_norm": 0.5078125, + "learning_rate": 7.46032919688993e-05, + "loss": 2.4893, + "step": 27345 + }, + { + "epoch": 1.17, + "grad_norm": 0.435546875, + "learning_rate": 7.459680402838127e-05, + "loss": 2.3644, + "step": 27346 + }, + { + "epoch": 1.17, + "grad_norm": 0.55859375, + "learning_rate": 7.459031620216832e-05, + "loss": 2.5684, + "step": 27347 + }, + { + "epoch": 1.17, + "grad_norm": 0.470703125, + "learning_rate": 7.45838284902897e-05, + "loss": 2.4659, + "step": 27348 + }, + { + "epoch": 1.17, + "grad_norm": 0.4609375, + "learning_rate": 7.457734089277457e-05, + "loss": 2.2987, + "step": 27349 + }, + { + "epoch": 1.17, + "grad_norm": 0.5078125, + "learning_rate": 7.457085340965208e-05, + "loss": 2.5268, + "step": 27350 + }, + { + "epoch": 1.17, + "grad_norm": 0.57421875, + "learning_rate": 7.456436604095146e-05, + "loss": 2.5351, + "step": 27351 + }, + { + "epoch": 1.17, + "grad_norm": 0.47265625, + "learning_rate": 7.455787878670191e-05, + "loss": 2.3234, + "step": 27352 + }, + { + "epoch": 1.17, + "grad_norm": 0.435546875, + "learning_rate": 7.455139164693263e-05, + "loss": 2.3295, + "step": 27353 + }, + { + "epoch": 1.17, + "grad_norm": 0.44921875, + "learning_rate": 7.454490462167276e-05, + "loss": 2.4579, + "step": 27354 + }, + { + "epoch": 1.17, + "grad_norm": 0.439453125, + "learning_rate": 7.453841771095149e-05, + "loss": 2.5478, + "step": 27355 + }, + { + "epoch": 1.17, + "grad_norm": 0.466796875, + "learning_rate": 7.453193091479808e-05, + "loss": 2.5112, + "step": 27356 + }, + { + "epoch": 1.17, + "grad_norm": 0.470703125, + "learning_rate": 7.452544423324166e-05, + "loss": 2.4843, + "step": 27357 + }, + { + "epoch": 1.17, + "grad_norm": 0.494140625, + "learning_rate": 7.451895766631144e-05, + "loss": 2.5595, + "step": 27358 + }, + { + "epoch": 1.17, + "grad_norm": 0.49609375, + "learning_rate": 7.451247121403658e-05, + "loss": 2.4254, + "step": 27359 + }, + { + "epoch": 1.17, + "grad_norm": 0.423828125, + "learning_rate": 7.450598487644629e-05, + "loss": 2.4363, + "step": 27360 + }, + { + "epoch": 1.17, + "grad_norm": 0.439453125, + "learning_rate": 7.449949865356975e-05, + "loss": 2.4783, + "step": 27361 + }, + { + "epoch": 1.17, + "grad_norm": 0.443359375, + "learning_rate": 7.449301254543617e-05, + "loss": 2.4245, + "step": 27362 + }, + { + "epoch": 1.17, + "grad_norm": 0.498046875, + "learning_rate": 7.448652655207467e-05, + "loss": 2.4854, + "step": 27363 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.44800406735145e-05, + "loss": 2.4814, + "step": 27364 + }, + { + "epoch": 1.17, + "grad_norm": 0.462890625, + "learning_rate": 7.447355490978482e-05, + "loss": 2.497, + "step": 27365 + }, + { + "epoch": 1.17, + "grad_norm": 0.484375, + "learning_rate": 7.446706926091482e-05, + "loss": 2.4272, + "step": 27366 + }, + { + "epoch": 1.17, + "grad_norm": 0.4765625, + "learning_rate": 7.446058372693365e-05, + "loss": 2.4635, + "step": 27367 + }, + { + "epoch": 1.17, + "grad_norm": 0.484375, + "learning_rate": 7.445409830787055e-05, + "loss": 2.5517, + "step": 27368 + }, + { + "epoch": 1.17, + "grad_norm": 0.5390625, + "learning_rate": 7.444761300375466e-05, + "loss": 2.4619, + "step": 27369 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.44411278146152e-05, + "loss": 2.3768, + "step": 27370 + }, + { + "epoch": 1.17, + "grad_norm": 0.482421875, + "learning_rate": 7.443464274048127e-05, + "loss": 2.4776, + "step": 27371 + }, + { + "epoch": 1.17, + "grad_norm": 0.482421875, + "learning_rate": 7.442815778138214e-05, + "loss": 2.5388, + "step": 27372 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.442167293734697e-05, + "loss": 2.4733, + "step": 27373 + }, + { + "epoch": 1.17, + "grad_norm": 0.458984375, + "learning_rate": 7.44151882084049e-05, + "loss": 2.4465, + "step": 27374 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.440870359458511e-05, + "loss": 2.559, + "step": 27375 + }, + { + "epoch": 1.17, + "grad_norm": 0.5, + "learning_rate": 7.440221909591682e-05, + "loss": 2.3922, + "step": 27376 + }, + { + "epoch": 1.17, + "grad_norm": 0.458984375, + "learning_rate": 7.439573471242921e-05, + "loss": 2.5205, + "step": 27377 + }, + { + "epoch": 1.17, + "grad_norm": 0.44921875, + "learning_rate": 7.438925044415144e-05, + "loss": 2.5126, + "step": 27378 + }, + { + "epoch": 1.17, + "grad_norm": 0.44921875, + "learning_rate": 7.438276629111263e-05, + "loss": 2.5292, + "step": 27379 + }, + { + "epoch": 1.17, + "grad_norm": 0.443359375, + "learning_rate": 7.437628225334207e-05, + "loss": 2.4957, + "step": 27380 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.436979833086885e-05, + "loss": 2.3433, + "step": 27381 + }, + { + "epoch": 1.17, + "grad_norm": 0.423828125, + "learning_rate": 7.436331452372218e-05, + "loss": 2.528, + "step": 27382 + }, + { + "epoch": 1.17, + "grad_norm": 0.41015625, + "learning_rate": 7.435683083193118e-05, + "loss": 2.5237, + "step": 27383 + }, + { + "epoch": 1.17, + "grad_norm": 0.45703125, + "learning_rate": 7.435034725552512e-05, + "loss": 2.3792, + "step": 27384 + }, + { + "epoch": 1.17, + "grad_norm": 0.46875, + "learning_rate": 7.434386379453311e-05, + "loss": 2.3407, + "step": 27385 + }, + { + "epoch": 1.17, + "grad_norm": 0.455078125, + "learning_rate": 7.433738044898436e-05, + "loss": 2.5167, + "step": 27386 + }, + { + "epoch": 1.17, + "grad_norm": 0.447265625, + "learning_rate": 7.433089721890796e-05, + "loss": 2.564, + "step": 27387 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.43244141043332e-05, + "loss": 2.4504, + "step": 27388 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.431793110528917e-05, + "loss": 2.3932, + "step": 27389 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.431144822180506e-05, + "loss": 2.4022, + "step": 27390 + }, + { + "epoch": 1.17, + "grad_norm": 0.482421875, + "learning_rate": 7.430496545391004e-05, + "loss": 2.4719, + "step": 27391 + }, + { + "epoch": 1.17, + "grad_norm": 0.451171875, + "learning_rate": 7.42984828016333e-05, + "loss": 2.4484, + "step": 27392 + }, + { + "epoch": 1.17, + "grad_norm": 0.455078125, + "learning_rate": 7.429200026500399e-05, + "loss": 2.3878, + "step": 27393 + }, + { + "epoch": 1.17, + "grad_norm": 0.5234375, + "learning_rate": 7.428551784405132e-05, + "loss": 2.4858, + "step": 27394 + }, + { + "epoch": 1.17, + "grad_norm": 0.515625, + "learning_rate": 7.427903553880433e-05, + "loss": 2.3787, + "step": 27395 + }, + { + "epoch": 1.17, + "grad_norm": 0.48046875, + "learning_rate": 7.427255334929237e-05, + "loss": 2.5911, + "step": 27396 + }, + { + "epoch": 1.17, + "grad_norm": 0.51171875, + "learning_rate": 7.426607127554448e-05, + "loss": 2.4383, + "step": 27397 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.425958931758987e-05, + "loss": 2.3544, + "step": 27398 + }, + { + "epoch": 1.17, + "grad_norm": 0.419921875, + "learning_rate": 7.425310747545769e-05, + "loss": 2.3633, + "step": 27399 + }, + { + "epoch": 1.17, + "grad_norm": 0.51171875, + "learning_rate": 7.424662574917712e-05, + "loss": 2.4993, + "step": 27400 + }, + { + "epoch": 1.17, + "grad_norm": 0.435546875, + "learning_rate": 7.424014413877732e-05, + "loss": 2.3221, + "step": 27401 + }, + { + "epoch": 1.17, + "grad_norm": 0.439453125, + "learning_rate": 7.42336626442875e-05, + "loss": 2.5499, + "step": 27402 + }, + { + "epoch": 1.17, + "grad_norm": 0.5234375, + "learning_rate": 7.422718126573675e-05, + "loss": 2.4721, + "step": 27403 + }, + { + "epoch": 1.17, + "grad_norm": 0.44140625, + "learning_rate": 7.422070000315424e-05, + "loss": 2.4339, + "step": 27404 + }, + { + "epoch": 1.17, + "grad_norm": 0.47265625, + "learning_rate": 7.421421885656918e-05, + "loss": 2.5017, + "step": 27405 + }, + { + "epoch": 1.17, + "grad_norm": 0.443359375, + "learning_rate": 7.420773782601071e-05, + "loss": 2.3912, + "step": 27406 + }, + { + "epoch": 1.17, + "grad_norm": 0.484375, + "learning_rate": 7.420125691150801e-05, + "loss": 2.4587, + "step": 27407 + }, + { + "epoch": 1.17, + "grad_norm": 0.494140625, + "learning_rate": 7.419477611309018e-05, + "loss": 2.5191, + "step": 27408 + }, + { + "epoch": 1.17, + "grad_norm": 0.400390625, + "learning_rate": 7.418829543078647e-05, + "loss": 2.3225, + "step": 27409 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.418181486462601e-05, + "loss": 2.5048, + "step": 27410 + }, + { + "epoch": 1.17, + "grad_norm": 0.4921875, + "learning_rate": 7.417533441463792e-05, + "loss": 2.5017, + "step": 27411 + }, + { + "epoch": 1.17, + "grad_norm": 0.4453125, + "learning_rate": 7.416885408085137e-05, + "loss": 2.4436, + "step": 27412 + }, + { + "epoch": 1.17, + "grad_norm": 0.46484375, + "learning_rate": 7.416237386329556e-05, + "loss": 2.515, + "step": 27413 + }, + { + "epoch": 1.17, + "grad_norm": 0.416015625, + "learning_rate": 7.415589376199963e-05, + "loss": 2.3822, + "step": 27414 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.414941377699273e-05, + "loss": 2.4752, + "step": 27415 + }, + { + "epoch": 1.17, + "grad_norm": 0.470703125, + "learning_rate": 7.4142933908304e-05, + "loss": 2.4251, + "step": 27416 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.413645415596263e-05, + "loss": 2.4035, + "step": 27417 + }, + { + "epoch": 1.17, + "grad_norm": 0.41796875, + "learning_rate": 7.412997451999779e-05, + "loss": 2.322, + "step": 27418 + }, + { + "epoch": 1.17, + "grad_norm": 0.474609375, + "learning_rate": 7.41234950004386e-05, + "loss": 2.4626, + "step": 27419 + }, + { + "epoch": 1.17, + "grad_norm": 0.458984375, + "learning_rate": 7.41170155973142e-05, + "loss": 2.476, + "step": 27420 + }, + { + "epoch": 1.17, + "grad_norm": 0.42578125, + "learning_rate": 7.411053631065377e-05, + "loss": 2.5101, + "step": 27421 + }, + { + "epoch": 1.17, + "grad_norm": 0.52734375, + "learning_rate": 7.41040571404865e-05, + "loss": 2.5521, + "step": 27422 + }, + { + "epoch": 1.17, + "grad_norm": 0.4921875, + "learning_rate": 7.409757808684148e-05, + "loss": 2.5225, + "step": 27423 + }, + { + "epoch": 1.17, + "grad_norm": 0.46484375, + "learning_rate": 7.409109914974788e-05, + "loss": 2.4062, + "step": 27424 + }, + { + "epoch": 1.17, + "grad_norm": 0.43359375, + "learning_rate": 7.408462032923488e-05, + "loss": 2.4758, + "step": 27425 + }, + { + "epoch": 1.17, + "grad_norm": 0.458984375, + "learning_rate": 7.407814162533167e-05, + "loss": 2.4476, + "step": 27426 + }, + { + "epoch": 1.17, + "grad_norm": 0.4453125, + "learning_rate": 7.40716630380673e-05, + "loss": 2.5165, + "step": 27427 + }, + { + "epoch": 1.17, + "grad_norm": 0.5078125, + "learning_rate": 7.406518456747094e-05, + "loss": 2.4938, + "step": 27428 + }, + { + "epoch": 1.17, + "grad_norm": 0.44921875, + "learning_rate": 7.405870621357182e-05, + "loss": 2.4246, + "step": 27429 + }, + { + "epoch": 1.17, + "grad_norm": 0.470703125, + "learning_rate": 7.405222797639902e-05, + "loss": 2.5055, + "step": 27430 + }, + { + "epoch": 1.17, + "grad_norm": 0.4453125, + "learning_rate": 7.404574985598172e-05, + "loss": 2.5244, + "step": 27431 + }, + { + "epoch": 1.17, + "grad_norm": 0.5703125, + "learning_rate": 7.403927185234905e-05, + "loss": 2.4825, + "step": 27432 + }, + { + "epoch": 1.17, + "grad_norm": 0.53125, + "learning_rate": 7.403279396553017e-05, + "loss": 2.5888, + "step": 27433 + }, + { + "epoch": 1.17, + "grad_norm": 0.498046875, + "learning_rate": 7.402631619555426e-05, + "loss": 2.4652, + "step": 27434 + }, + { + "epoch": 1.17, + "grad_norm": 0.49609375, + "learning_rate": 7.401983854245041e-05, + "loss": 2.4769, + "step": 27435 + }, + { + "epoch": 1.17, + "grad_norm": 0.4375, + "learning_rate": 7.401336100624779e-05, + "loss": 2.4364, + "step": 27436 + }, + { + "epoch": 1.17, + "grad_norm": 0.47265625, + "learning_rate": 7.400688358697554e-05, + "loss": 2.4813, + "step": 27437 + }, + { + "epoch": 1.17, + "grad_norm": 0.51953125, + "learning_rate": 7.400040628466282e-05, + "loss": 2.3569, + "step": 27438 + }, + { + "epoch": 1.17, + "grad_norm": 0.458984375, + "learning_rate": 7.399392909933879e-05, + "loss": 2.4319, + "step": 27439 + }, + { + "epoch": 1.17, + "grad_norm": 0.451171875, + "learning_rate": 7.398745203103253e-05, + "loss": 2.3734, + "step": 27440 + }, + { + "epoch": 1.17, + "grad_norm": 0.486328125, + "learning_rate": 7.398097507977327e-05, + "loss": 2.5122, + "step": 27441 + }, + { + "epoch": 1.17, + "grad_norm": 0.380859375, + "learning_rate": 7.397449824559012e-05, + "loss": 2.3373, + "step": 27442 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.396802152851222e-05, + "loss": 2.4619, + "step": 27443 + }, + { + "epoch": 1.17, + "grad_norm": 0.396484375, + "learning_rate": 7.396154492856866e-05, + "loss": 2.4541, + "step": 27444 + }, + { + "epoch": 1.17, + "grad_norm": 0.41796875, + "learning_rate": 7.395506844578866e-05, + "loss": 2.4758, + "step": 27445 + }, + { + "epoch": 1.17, + "grad_norm": 0.451171875, + "learning_rate": 7.394859208020133e-05, + "loss": 2.3643, + "step": 27446 + }, + { + "epoch": 1.17, + "grad_norm": 0.462890625, + "learning_rate": 7.394211583183583e-05, + "loss": 2.4899, + "step": 27447 + }, + { + "epoch": 1.17, + "grad_norm": 0.51953125, + "learning_rate": 7.393563970072124e-05, + "loss": 2.4684, + "step": 27448 + }, + { + "epoch": 1.17, + "grad_norm": 0.447265625, + "learning_rate": 7.392916368688679e-05, + "loss": 2.5434, + "step": 27449 + }, + { + "epoch": 1.17, + "grad_norm": 0.50390625, + "learning_rate": 7.392268779036159e-05, + "loss": 2.5145, + "step": 27450 + }, + { + "epoch": 1.17, + "grad_norm": 0.427734375, + "learning_rate": 7.391621201117474e-05, + "loss": 2.3969, + "step": 27451 + }, + { + "epoch": 1.17, + "grad_norm": 0.484375, + "learning_rate": 7.390973634935539e-05, + "loss": 2.4632, + "step": 27452 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.39032608049327e-05, + "loss": 2.5411, + "step": 27453 + }, + { + "epoch": 1.17, + "grad_norm": 0.53125, + "learning_rate": 7.389678537793581e-05, + "loss": 2.5332, + "step": 27454 + }, + { + "epoch": 1.17, + "grad_norm": 0.484375, + "learning_rate": 7.389031006839384e-05, + "loss": 2.4425, + "step": 27455 + }, + { + "epoch": 1.17, + "grad_norm": 0.447265625, + "learning_rate": 7.388383487633592e-05, + "loss": 2.4458, + "step": 27456 + }, + { + "epoch": 1.17, + "grad_norm": 0.498046875, + "learning_rate": 7.387735980179122e-05, + "loss": 2.4929, + "step": 27457 + }, + { + "epoch": 1.17, + "grad_norm": 0.50390625, + "learning_rate": 7.387088484478887e-05, + "loss": 2.4722, + "step": 27458 + }, + { + "epoch": 1.17, + "grad_norm": 0.396484375, + "learning_rate": 7.386441000535796e-05, + "loss": 2.2868, + "step": 27459 + }, + { + "epoch": 1.17, + "grad_norm": 0.47265625, + "learning_rate": 7.385793528352765e-05, + "loss": 2.4303, + "step": 27460 + }, + { + "epoch": 1.17, + "grad_norm": 0.466796875, + "learning_rate": 7.385146067932709e-05, + "loss": 2.4441, + "step": 27461 + }, + { + "epoch": 1.17, + "grad_norm": 0.431640625, + "learning_rate": 7.384498619278539e-05, + "loss": 2.4574, + "step": 27462 + }, + { + "epoch": 1.17, + "grad_norm": 0.4921875, + "learning_rate": 7.383851182393171e-05, + "loss": 2.4455, + "step": 27463 + }, + { + "epoch": 1.17, + "grad_norm": 0.546875, + "learning_rate": 7.383203757279513e-05, + "loss": 2.5156, + "step": 27464 + }, + { + "epoch": 1.17, + "grad_norm": 0.443359375, + "learning_rate": 7.382556343940485e-05, + "loss": 2.5303, + "step": 27465 + }, + { + "epoch": 1.17, + "grad_norm": 0.466796875, + "learning_rate": 7.381908942379e-05, + "loss": 2.3614, + "step": 27466 + }, + { + "epoch": 1.17, + "grad_norm": 0.4609375, + "learning_rate": 7.381261552597964e-05, + "loss": 2.4006, + "step": 27467 + }, + { + "epoch": 1.17, + "grad_norm": 0.494140625, + "learning_rate": 7.38061417460029e-05, + "loss": 2.5151, + "step": 27468 + }, + { + "epoch": 1.17, + "grad_norm": 0.466796875, + "learning_rate": 7.379966808388901e-05, + "loss": 2.4395, + "step": 27469 + }, + { + "epoch": 1.17, + "grad_norm": 0.498046875, + "learning_rate": 7.379319453966703e-05, + "loss": 2.5087, + "step": 27470 + }, + { + "epoch": 1.17, + "grad_norm": 0.5, + "learning_rate": 7.378672111336608e-05, + "loss": 2.6614, + "step": 27471 + }, + { + "epoch": 1.17, + "grad_norm": 0.51953125, + "learning_rate": 7.378024780501529e-05, + "loss": 2.5614, + "step": 27472 + }, + { + "epoch": 1.17, + "grad_norm": 0.53515625, + "learning_rate": 7.377377461464383e-05, + "loss": 2.5522, + "step": 27473 + }, + { + "epoch": 1.17, + "grad_norm": 0.5234375, + "learning_rate": 7.376730154228082e-05, + "loss": 2.5758, + "step": 27474 + }, + { + "epoch": 1.17, + "grad_norm": 0.50390625, + "learning_rate": 7.376082858795534e-05, + "loss": 2.6079, + "step": 27475 + }, + { + "epoch": 1.17, + "grad_norm": 0.55078125, + "learning_rate": 7.375435575169653e-05, + "loss": 2.5859, + "step": 27476 + }, + { + "epoch": 1.17, + "grad_norm": 0.455078125, + "learning_rate": 7.374788303353355e-05, + "loss": 2.3268, + "step": 27477 + }, + { + "epoch": 1.17, + "grad_norm": 0.48828125, + "learning_rate": 7.374141043349548e-05, + "loss": 2.5213, + "step": 27478 + }, + { + "epoch": 1.17, + "grad_norm": 0.52734375, + "learning_rate": 7.373493795161148e-05, + "loss": 2.4722, + "step": 27479 + }, + { + "epoch": 1.17, + "grad_norm": 0.44140625, + "learning_rate": 7.372846558791064e-05, + "loss": 2.4438, + "step": 27480 + }, + { + "epoch": 1.17, + "grad_norm": 0.470703125, + "learning_rate": 7.372199334242213e-05, + "loss": 2.5424, + "step": 27481 + }, + { + "epoch": 1.17, + "grad_norm": 0.53515625, + "learning_rate": 7.371552121517506e-05, + "loss": 2.5612, + "step": 27482 + }, + { + "epoch": 1.17, + "grad_norm": 0.494140625, + "learning_rate": 7.370904920619852e-05, + "loss": 2.3833, + "step": 27483 + }, + { + "epoch": 1.17, + "grad_norm": 0.451171875, + "learning_rate": 7.370257731552163e-05, + "loss": 2.5707, + "step": 27484 + }, + { + "epoch": 1.17, + "grad_norm": 0.47265625, + "learning_rate": 7.369610554317356e-05, + "loss": 2.4854, + "step": 27485 + }, + { + "epoch": 1.17, + "grad_norm": 0.421875, + "learning_rate": 7.368963388918339e-05, + "loss": 2.5133, + "step": 27486 + }, + { + "epoch": 1.17, + "grad_norm": 0.46875, + "learning_rate": 7.368316235358027e-05, + "loss": 2.4873, + "step": 27487 + }, + { + "epoch": 1.17, + "grad_norm": 0.52734375, + "learning_rate": 7.367669093639327e-05, + "loss": 2.5333, + "step": 27488 + }, + { + "epoch": 1.17, + "grad_norm": 0.48828125, + "learning_rate": 7.367021963765155e-05, + "loss": 2.6165, + "step": 27489 + }, + { + "epoch": 1.17, + "grad_norm": 0.455078125, + "learning_rate": 7.366374845738425e-05, + "loss": 2.4897, + "step": 27490 + }, + { + "epoch": 1.17, + "grad_norm": 0.46484375, + "learning_rate": 7.365727739562043e-05, + "loss": 2.3823, + "step": 27491 + }, + { + "epoch": 1.17, + "grad_norm": 0.48046875, + "learning_rate": 7.365080645238922e-05, + "loss": 2.5242, + "step": 27492 + }, + { + "epoch": 1.17, + "grad_norm": 0.51953125, + "learning_rate": 7.364433562771978e-05, + "loss": 2.4707, + "step": 27493 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.363786492164118e-05, + "loss": 2.4648, + "step": 27494 + }, + { + "epoch": 1.17, + "grad_norm": 0.423828125, + "learning_rate": 7.363139433418256e-05, + "loss": 2.4089, + "step": 27495 + }, + { + "epoch": 1.17, + "grad_norm": 0.462890625, + "learning_rate": 7.362492386537302e-05, + "loss": 2.437, + "step": 27496 + }, + { + "epoch": 1.17, + "grad_norm": 0.46875, + "learning_rate": 7.361845351524169e-05, + "loss": 2.3678, + "step": 27497 + }, + { + "epoch": 1.17, + "grad_norm": 0.45703125, + "learning_rate": 7.361198328381771e-05, + "loss": 2.4416, + "step": 27498 + }, + { + "epoch": 1.17, + "grad_norm": 0.546875, + "learning_rate": 7.360551317113014e-05, + "loss": 2.6194, + "step": 27499 + }, + { + "epoch": 1.17, + "grad_norm": 0.455078125, + "learning_rate": 7.359904317720808e-05, + "loss": 2.5129, + "step": 27500 + }, + { + "epoch": 1.17, + "grad_norm": 0.47265625, + "learning_rate": 7.359257330208072e-05, + "loss": 2.624, + "step": 27501 + }, + { + "epoch": 1.17, + "grad_norm": 0.494140625, + "learning_rate": 7.35861035457771e-05, + "loss": 2.5453, + "step": 27502 + }, + { + "epoch": 1.17, + "grad_norm": 0.439453125, + "learning_rate": 7.357963390832638e-05, + "loss": 2.4492, + "step": 27503 + }, + { + "epoch": 1.17, + "grad_norm": 0.490234375, + "learning_rate": 7.357316438975763e-05, + "loss": 2.4052, + "step": 27504 + }, + { + "epoch": 1.17, + "grad_norm": 0.490234375, + "learning_rate": 7.356669499010001e-05, + "loss": 2.4001, + "step": 27505 + }, + { + "epoch": 1.17, + "grad_norm": 0.490234375, + "learning_rate": 7.356022570938261e-05, + "loss": 2.6253, + "step": 27506 + }, + { + "epoch": 1.17, + "grad_norm": 0.48828125, + "learning_rate": 7.355375654763451e-05, + "loss": 2.4814, + "step": 27507 + }, + { + "epoch": 1.17, + "grad_norm": 0.44140625, + "learning_rate": 7.354728750488483e-05, + "loss": 2.2587, + "step": 27508 + }, + { + "epoch": 1.17, + "grad_norm": 0.4765625, + "learning_rate": 7.354081858116271e-05, + "loss": 2.5444, + "step": 27509 + }, + { + "epoch": 1.17, + "grad_norm": 0.53515625, + "learning_rate": 7.353434977649723e-05, + "loss": 2.3805, + "step": 27510 + }, + { + "epoch": 1.17, + "grad_norm": 0.46875, + "learning_rate": 7.35278810909175e-05, + "loss": 2.3843, + "step": 27511 + }, + { + "epoch": 1.17, + "grad_norm": 0.45703125, + "learning_rate": 7.352141252445262e-05, + "loss": 2.5125, + "step": 27512 + }, + { + "epoch": 1.17, + "grad_norm": 0.4375, + "learning_rate": 7.351494407713171e-05, + "loss": 2.4342, + "step": 27513 + }, + { + "epoch": 1.17, + "grad_norm": 0.53515625, + "learning_rate": 7.350847574898392e-05, + "loss": 2.5494, + "step": 27514 + }, + { + "epoch": 1.17, + "grad_norm": 0.51171875, + "learning_rate": 7.350200754003827e-05, + "loss": 2.5004, + "step": 27515 + }, + { + "epoch": 1.17, + "grad_norm": 0.4453125, + "learning_rate": 7.349553945032388e-05, + "loss": 2.5496, + "step": 27516 + }, + { + "epoch": 1.17, + "grad_norm": 0.470703125, + "learning_rate": 7.34890714798699e-05, + "loss": 2.4378, + "step": 27517 + }, + { + "epoch": 1.17, + "grad_norm": 0.44921875, + "learning_rate": 7.348260362870541e-05, + "loss": 2.4798, + "step": 27518 + }, + { + "epoch": 1.17, + "grad_norm": 0.46484375, + "learning_rate": 7.34761358968595e-05, + "loss": 2.5673, + "step": 27519 + }, + { + "epoch": 1.17, + "grad_norm": 0.470703125, + "learning_rate": 7.346966828436128e-05, + "loss": 2.5497, + "step": 27520 + }, + { + "epoch": 1.17, + "grad_norm": 0.42578125, + "learning_rate": 7.346320079123989e-05, + "loss": 2.4079, + "step": 27521 + }, + { + "epoch": 1.17, + "grad_norm": 0.50390625, + "learning_rate": 7.34567334175244e-05, + "loss": 2.4077, + "step": 27522 + }, + { + "epoch": 1.17, + "grad_norm": 0.439453125, + "learning_rate": 7.345026616324388e-05, + "loss": 2.4529, + "step": 27523 + }, + { + "epoch": 1.17, + "grad_norm": 0.462890625, + "learning_rate": 7.344379902842748e-05, + "loss": 2.537, + "step": 27524 + }, + { + "epoch": 1.17, + "grad_norm": 0.458984375, + "learning_rate": 7.343733201310426e-05, + "loss": 2.4541, + "step": 27525 + }, + { + "epoch": 1.17, + "grad_norm": 0.46484375, + "learning_rate": 7.343086511730335e-05, + "loss": 2.4807, + "step": 27526 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.342439834105383e-05, + "loss": 2.4525, + "step": 27527 + }, + { + "epoch": 1.17, + "grad_norm": 0.494140625, + "learning_rate": 7.341793168438484e-05, + "loss": 2.4535, + "step": 27528 + }, + { + "epoch": 1.17, + "grad_norm": 0.578125, + "learning_rate": 7.341146514732538e-05, + "loss": 2.4094, + "step": 27529 + }, + { + "epoch": 1.17, + "grad_norm": 0.51171875, + "learning_rate": 7.340499872990465e-05, + "loss": 2.5011, + "step": 27530 + }, + { + "epoch": 1.17, + "grad_norm": 0.515625, + "learning_rate": 7.339853243215172e-05, + "loss": 2.3438, + "step": 27531 + }, + { + "epoch": 1.17, + "grad_norm": 0.408203125, + "learning_rate": 7.339206625409564e-05, + "loss": 2.3238, + "step": 27532 + }, + { + "epoch": 1.17, + "grad_norm": 0.5390625, + "learning_rate": 7.338560019576553e-05, + "loss": 2.5007, + "step": 27533 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.337913425719052e-05, + "loss": 2.5204, + "step": 27534 + }, + { + "epoch": 1.17, + "grad_norm": 0.4765625, + "learning_rate": 7.337266843839967e-05, + "loss": 2.5499, + "step": 27535 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.33662027394221e-05, + "loss": 2.411, + "step": 27536 + }, + { + "epoch": 1.17, + "grad_norm": 0.46484375, + "learning_rate": 7.335973716028682e-05, + "loss": 2.5025, + "step": 27537 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.335327170102306e-05, + "loss": 2.3543, + "step": 27538 + }, + { + "epoch": 1.17, + "grad_norm": 0.490234375, + "learning_rate": 7.33468063616598e-05, + "loss": 2.4791, + "step": 27539 + }, + { + "epoch": 1.17, + "grad_norm": 0.55859375, + "learning_rate": 7.334034114222619e-05, + "loss": 2.5202, + "step": 27540 + }, + { + "epoch": 1.17, + "grad_norm": 0.482421875, + "learning_rate": 7.333387604275128e-05, + "loss": 2.605, + "step": 27541 + }, + { + "epoch": 1.17, + "grad_norm": 0.453125, + "learning_rate": 7.33274110632642e-05, + "loss": 2.3789, + "step": 27542 + }, + { + "epoch": 1.17, + "grad_norm": 0.47265625, + "learning_rate": 7.332094620379401e-05, + "loss": 2.6274, + "step": 27543 + }, + { + "epoch": 1.17, + "grad_norm": 0.498046875, + "learning_rate": 7.331448146436985e-05, + "loss": 2.5319, + "step": 27544 + }, + { + "epoch": 1.17, + "grad_norm": 0.50390625, + "learning_rate": 7.330801684502068e-05, + "loss": 2.4037, + "step": 27545 + }, + { + "epoch": 1.17, + "grad_norm": 0.4921875, + "learning_rate": 7.330155234577579e-05, + "loss": 2.6162, + "step": 27546 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.329508796666409e-05, + "loss": 2.4441, + "step": 27547 + }, + { + "epoch": 1.17, + "grad_norm": 0.45703125, + "learning_rate": 7.328862370771477e-05, + "loss": 2.3978, + "step": 27548 + }, + { + "epoch": 1.17, + "grad_norm": 0.482421875, + "learning_rate": 7.328215956895684e-05, + "loss": 2.4358, + "step": 27549 + }, + { + "epoch": 1.17, + "grad_norm": 0.4375, + "learning_rate": 7.327569555041944e-05, + "loss": 2.4714, + "step": 27550 + }, + { + "epoch": 1.17, + "grad_norm": 0.49609375, + "learning_rate": 7.326923165213166e-05, + "loss": 2.5374, + "step": 27551 + }, + { + "epoch": 1.17, + "grad_norm": 0.52734375, + "learning_rate": 7.326276787412258e-05, + "loss": 2.5117, + "step": 27552 + }, + { + "epoch": 1.17, + "grad_norm": 0.427734375, + "learning_rate": 7.325630421642122e-05, + "loss": 2.5235, + "step": 27553 + }, + { + "epoch": 1.17, + "grad_norm": 0.435546875, + "learning_rate": 7.324984067905678e-05, + "loss": 2.4891, + "step": 27554 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.324337726205825e-05, + "loss": 2.461, + "step": 27555 + }, + { + "epoch": 1.17, + "grad_norm": 0.451171875, + "learning_rate": 7.323691396545475e-05, + "loss": 2.3869, + "step": 27556 + }, + { + "epoch": 1.17, + "grad_norm": 0.52734375, + "learning_rate": 7.323045078927535e-05, + "loss": 2.4587, + "step": 27557 + }, + { + "epoch": 1.17, + "grad_norm": 0.4453125, + "learning_rate": 7.322398773354914e-05, + "loss": 2.5682, + "step": 27558 + }, + { + "epoch": 1.17, + "grad_norm": 0.51171875, + "learning_rate": 7.321752479830521e-05, + "loss": 2.4228, + "step": 27559 + }, + { + "epoch": 1.17, + "grad_norm": 0.4921875, + "learning_rate": 7.321106198357264e-05, + "loss": 2.4818, + "step": 27560 + }, + { + "epoch": 1.17, + "grad_norm": 0.515625, + "learning_rate": 7.320459928938044e-05, + "loss": 2.528, + "step": 27561 + }, + { + "epoch": 1.17, + "grad_norm": 0.447265625, + "learning_rate": 7.319813671575781e-05, + "loss": 2.4018, + "step": 27562 + }, + { + "epoch": 1.17, + "grad_norm": 0.478515625, + "learning_rate": 7.319167426273377e-05, + "loss": 2.5016, + "step": 27563 + }, + { + "epoch": 1.17, + "grad_norm": 0.48828125, + "learning_rate": 7.318521193033739e-05, + "loss": 2.5743, + "step": 27564 + }, + { + "epoch": 1.17, + "grad_norm": 0.46875, + "learning_rate": 7.317874971859773e-05, + "loss": 2.5059, + "step": 27565 + }, + { + "epoch": 1.17, + "grad_norm": 0.494140625, + "learning_rate": 7.317228762754392e-05, + "loss": 2.3663, + "step": 27566 + }, + { + "epoch": 1.17, + "grad_norm": 0.53125, + "learning_rate": 7.3165825657205e-05, + "loss": 2.422, + "step": 27567 + }, + { + "epoch": 1.17, + "grad_norm": 0.458984375, + "learning_rate": 7.315936380761008e-05, + "loss": 2.4993, + "step": 27568 + }, + { + "epoch": 1.17, + "grad_norm": 0.486328125, + "learning_rate": 7.315290207878815e-05, + "loss": 2.5744, + "step": 27569 + }, + { + "epoch": 1.17, + "grad_norm": 0.462890625, + "learning_rate": 7.314644047076843e-05, + "loss": 2.3413, + "step": 27570 + }, + { + "epoch": 1.17, + "grad_norm": 0.462890625, + "learning_rate": 7.313997898357988e-05, + "loss": 2.5264, + "step": 27571 + }, + { + "epoch": 1.17, + "grad_norm": 0.404296875, + "learning_rate": 7.31335176172516e-05, + "loss": 2.4851, + "step": 27572 + }, + { + "epoch": 1.17, + "grad_norm": 0.451171875, + "learning_rate": 7.312705637181267e-05, + "loss": 2.4434, + "step": 27573 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.312059524729218e-05, + "loss": 2.4507, + "step": 27574 + }, + { + "epoch": 1.18, + "grad_norm": 0.462890625, + "learning_rate": 7.311413424371917e-05, + "loss": 2.4789, + "step": 27575 + }, + { + "epoch": 1.18, + "grad_norm": 0.4921875, + "learning_rate": 7.310767336112279e-05, + "loss": 2.332, + "step": 27576 + }, + { + "epoch": 1.18, + "grad_norm": 0.453125, + "learning_rate": 7.310121259953195e-05, + "loss": 2.4439, + "step": 27577 + }, + { + "epoch": 1.18, + "grad_norm": 0.53125, + "learning_rate": 7.309475195897591e-05, + "loss": 2.526, + "step": 27578 + }, + { + "epoch": 1.18, + "grad_norm": 0.546875, + "learning_rate": 7.308829143948362e-05, + "loss": 2.4921, + "step": 27579 + }, + { + "epoch": 1.18, + "grad_norm": 0.453125, + "learning_rate": 7.308183104108419e-05, + "loss": 2.606, + "step": 27580 + }, + { + "epoch": 1.18, + "grad_norm": 0.484375, + "learning_rate": 7.307537076380666e-05, + "loss": 2.5915, + "step": 27581 + }, + { + "epoch": 1.18, + "grad_norm": 0.48828125, + "learning_rate": 7.306891060768015e-05, + "loss": 2.4916, + "step": 27582 + }, + { + "epoch": 1.18, + "grad_norm": 0.51171875, + "learning_rate": 7.306245057273369e-05, + "loss": 2.5066, + "step": 27583 + }, + { + "epoch": 1.18, + "grad_norm": 0.5078125, + "learning_rate": 7.305599065899638e-05, + "loss": 2.4764, + "step": 27584 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.304953086649719e-05, + "loss": 2.4588, + "step": 27585 + }, + { + "epoch": 1.18, + "grad_norm": 0.4375, + "learning_rate": 7.304307119526534e-05, + "loss": 2.5829, + "step": 27586 + }, + { + "epoch": 1.18, + "grad_norm": 0.455078125, + "learning_rate": 7.30366116453298e-05, + "loss": 2.4276, + "step": 27587 + }, + { + "epoch": 1.18, + "grad_norm": 0.546875, + "learning_rate": 7.303015221671965e-05, + "loss": 2.5694, + "step": 27588 + }, + { + "epoch": 1.18, + "grad_norm": 0.46484375, + "learning_rate": 7.302369290946393e-05, + "loss": 2.4487, + "step": 27589 + }, + { + "epoch": 1.18, + "grad_norm": 0.4609375, + "learning_rate": 7.301723372359177e-05, + "loss": 2.425, + "step": 27590 + }, + { + "epoch": 1.18, + "grad_norm": 0.412109375, + "learning_rate": 7.301077465913218e-05, + "loss": 2.3804, + "step": 27591 + }, + { + "epoch": 1.18, + "grad_norm": 0.45703125, + "learning_rate": 7.300431571611428e-05, + "loss": 2.4304, + "step": 27592 + }, + { + "epoch": 1.18, + "grad_norm": 0.4921875, + "learning_rate": 7.299785689456701e-05, + "loss": 2.5144, + "step": 27593 + }, + { + "epoch": 1.18, + "grad_norm": 0.451171875, + "learning_rate": 7.299139819451959e-05, + "loss": 2.5637, + "step": 27594 + }, + { + "epoch": 1.18, + "grad_norm": 0.470703125, + "learning_rate": 7.298493961600099e-05, + "loss": 2.4685, + "step": 27595 + }, + { + "epoch": 1.18, + "grad_norm": 0.392578125, + "learning_rate": 7.297848115904028e-05, + "loss": 2.4637, + "step": 27596 + }, + { + "epoch": 1.18, + "grad_norm": 0.54296875, + "learning_rate": 7.29720228236665e-05, + "loss": 2.5255, + "step": 27597 + }, + { + "epoch": 1.18, + "grad_norm": 0.484375, + "learning_rate": 7.296556460990878e-05, + "loss": 2.4573, + "step": 27598 + }, + { + "epoch": 1.18, + "grad_norm": 0.4609375, + "learning_rate": 7.295910651779613e-05, + "loss": 2.547, + "step": 27599 + }, + { + "epoch": 1.18, + "grad_norm": 0.462890625, + "learning_rate": 7.295264854735765e-05, + "loss": 2.5526, + "step": 27600 + }, + { + "epoch": 1.18, + "grad_norm": 0.4921875, + "learning_rate": 7.29461906986223e-05, + "loss": 2.5944, + "step": 27601 + }, + { + "epoch": 1.18, + "grad_norm": 0.498046875, + "learning_rate": 7.293973297161927e-05, + "loss": 2.4506, + "step": 27602 + }, + { + "epoch": 1.18, + "grad_norm": 0.53125, + "learning_rate": 7.29332753663775e-05, + "loss": 2.5261, + "step": 27603 + }, + { + "epoch": 1.18, + "grad_norm": 0.48828125, + "learning_rate": 7.292681788292615e-05, + "loss": 2.353, + "step": 27604 + }, + { + "epoch": 1.18, + "grad_norm": 0.423828125, + "learning_rate": 7.292036052129416e-05, + "loss": 2.443, + "step": 27605 + }, + { + "epoch": 1.18, + "grad_norm": 0.470703125, + "learning_rate": 7.29139032815107e-05, + "loss": 2.5033, + "step": 27606 + }, + { + "epoch": 1.18, + "grad_norm": 0.46875, + "learning_rate": 7.290744616360476e-05, + "loss": 2.4382, + "step": 27607 + }, + { + "epoch": 1.18, + "grad_norm": 0.408203125, + "learning_rate": 7.290098916760543e-05, + "loss": 2.4717, + "step": 27608 + }, + { + "epoch": 1.18, + "grad_norm": 0.490234375, + "learning_rate": 7.28945322935417e-05, + "loss": 2.4664, + "step": 27609 + }, + { + "epoch": 1.18, + "grad_norm": 0.498046875, + "learning_rate": 7.288807554144273e-05, + "loss": 2.4502, + "step": 27610 + }, + { + "epoch": 1.18, + "grad_norm": 0.47265625, + "learning_rate": 7.288161891133749e-05, + "loss": 2.3521, + "step": 27611 + }, + { + "epoch": 1.18, + "grad_norm": 0.5078125, + "learning_rate": 7.287516240325504e-05, + "loss": 2.577, + "step": 27612 + }, + { + "epoch": 1.18, + "grad_norm": 0.494140625, + "learning_rate": 7.286870601722443e-05, + "loss": 2.514, + "step": 27613 + }, + { + "epoch": 1.18, + "grad_norm": 0.4765625, + "learning_rate": 7.286224975327476e-05, + "loss": 2.485, + "step": 27614 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.285579361143504e-05, + "loss": 2.5213, + "step": 27615 + }, + { + "epoch": 1.18, + "grad_norm": 0.435546875, + "learning_rate": 7.284933759173435e-05, + "loss": 2.3674, + "step": 27616 + }, + { + "epoch": 1.18, + "grad_norm": 0.4921875, + "learning_rate": 7.284288169420165e-05, + "loss": 2.5842, + "step": 27617 + }, + { + "epoch": 1.18, + "grad_norm": 0.4609375, + "learning_rate": 7.283642591886613e-05, + "loss": 2.6411, + "step": 27618 + }, + { + "epoch": 1.18, + "grad_norm": 0.490234375, + "learning_rate": 7.282997026575673e-05, + "loss": 2.4332, + "step": 27619 + }, + { + "epoch": 1.18, + "grad_norm": 0.515625, + "learning_rate": 7.282351473490256e-05, + "loss": 2.3419, + "step": 27620 + }, + { + "epoch": 1.18, + "grad_norm": 0.416015625, + "learning_rate": 7.28170593263326e-05, + "loss": 2.4087, + "step": 27621 + }, + { + "epoch": 1.18, + "grad_norm": 0.45703125, + "learning_rate": 7.281060404007596e-05, + "loss": 2.4797, + "step": 27622 + }, + { + "epoch": 1.18, + "grad_norm": 0.451171875, + "learning_rate": 7.280414887616167e-05, + "loss": 2.4429, + "step": 27623 + }, + { + "epoch": 1.18, + "grad_norm": 0.47265625, + "learning_rate": 7.27976938346188e-05, + "loss": 2.5149, + "step": 27624 + }, + { + "epoch": 1.18, + "grad_norm": 0.423828125, + "learning_rate": 7.27912389154763e-05, + "loss": 2.4717, + "step": 27625 + }, + { + "epoch": 1.18, + "grad_norm": 0.3984375, + "learning_rate": 7.278478411876334e-05, + "loss": 2.5178, + "step": 27626 + }, + { + "epoch": 1.18, + "grad_norm": 0.47265625, + "learning_rate": 7.277832944450889e-05, + "loss": 2.4884, + "step": 27627 + }, + { + "epoch": 1.18, + "grad_norm": 0.4453125, + "learning_rate": 7.277187489274199e-05, + "loss": 2.4574, + "step": 27628 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.276542046349171e-05, + "loss": 2.5725, + "step": 27629 + }, + { + "epoch": 1.18, + "grad_norm": 0.3984375, + "learning_rate": 7.275896615678707e-05, + "loss": 2.5003, + "step": 27630 + }, + { + "epoch": 1.18, + "grad_norm": 0.419921875, + "learning_rate": 7.275251197265715e-05, + "loss": 2.4296, + "step": 27631 + }, + { + "epoch": 1.18, + "grad_norm": 0.427734375, + "learning_rate": 7.274605791113101e-05, + "loss": 2.405, + "step": 27632 + }, + { + "epoch": 1.18, + "grad_norm": 0.470703125, + "learning_rate": 7.273960397223756e-05, + "loss": 2.5714, + "step": 27633 + }, + { + "epoch": 1.18, + "grad_norm": 0.515625, + "learning_rate": 7.273315015600599e-05, + "loss": 2.4121, + "step": 27634 + }, + { + "epoch": 1.18, + "grad_norm": 0.47265625, + "learning_rate": 7.272669646246526e-05, + "loss": 2.5175, + "step": 27635 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.272024289164444e-05, + "loss": 2.5233, + "step": 27636 + }, + { + "epoch": 1.18, + "grad_norm": 0.48046875, + "learning_rate": 7.271378944357252e-05, + "loss": 2.4701, + "step": 27637 + }, + { + "epoch": 1.18, + "grad_norm": 0.482421875, + "learning_rate": 7.270733611827862e-05, + "loss": 2.3267, + "step": 27638 + }, + { + "epoch": 1.18, + "grad_norm": 0.51171875, + "learning_rate": 7.270088291579171e-05, + "loss": 2.4589, + "step": 27639 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.269442983614089e-05, + "loss": 2.4835, + "step": 27640 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.268797687935508e-05, + "loss": 2.4192, + "step": 27641 + }, + { + "epoch": 1.18, + "grad_norm": 0.451171875, + "learning_rate": 7.268152404546347e-05, + "loss": 2.4547, + "step": 27642 + }, + { + "epoch": 1.18, + "grad_norm": 0.53125, + "learning_rate": 7.267507133449498e-05, + "loss": 2.4914, + "step": 27643 + }, + { + "epoch": 1.18, + "grad_norm": 0.5078125, + "learning_rate": 7.266861874647868e-05, + "loss": 2.4775, + "step": 27644 + }, + { + "epoch": 1.18, + "grad_norm": 0.48046875, + "learning_rate": 7.266216628144363e-05, + "loss": 2.3876, + "step": 27645 + }, + { + "epoch": 1.18, + "grad_norm": 0.498046875, + "learning_rate": 7.26557139394188e-05, + "loss": 2.3923, + "step": 27646 + }, + { + "epoch": 1.18, + "grad_norm": 0.466796875, + "learning_rate": 7.264926172043329e-05, + "loss": 2.4566, + "step": 27647 + }, + { + "epoch": 1.18, + "grad_norm": 0.47265625, + "learning_rate": 7.264280962451612e-05, + "loss": 2.4526, + "step": 27648 + }, + { + "epoch": 1.18, + "grad_norm": 0.478515625, + "learning_rate": 7.263635765169632e-05, + "loss": 2.5128, + "step": 27649 + }, + { + "epoch": 1.18, + "grad_norm": 0.40234375, + "learning_rate": 7.262990580200286e-05, + "loss": 2.3703, + "step": 27650 + }, + { + "epoch": 1.18, + "grad_norm": 0.4609375, + "learning_rate": 7.262345407546485e-05, + "loss": 2.4195, + "step": 27651 + }, + { + "epoch": 1.18, + "grad_norm": 0.44921875, + "learning_rate": 7.26170024721113e-05, + "loss": 2.5044, + "step": 27652 + }, + { + "epoch": 1.18, + "grad_norm": 0.52734375, + "learning_rate": 7.26105509919712e-05, + "loss": 2.5683, + "step": 27653 + }, + { + "epoch": 1.18, + "grad_norm": 0.494140625, + "learning_rate": 7.260409963507361e-05, + "loss": 2.5513, + "step": 27654 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.25976484014476e-05, + "loss": 2.4337, + "step": 27655 + }, + { + "epoch": 1.18, + "grad_norm": 0.431640625, + "learning_rate": 7.259119729112216e-05, + "loss": 2.3924, + "step": 27656 + }, + { + "epoch": 1.18, + "grad_norm": 0.44921875, + "learning_rate": 7.25847463041263e-05, + "loss": 2.3966, + "step": 27657 + }, + { + "epoch": 1.18, + "grad_norm": 0.53515625, + "learning_rate": 7.257829544048903e-05, + "loss": 2.4538, + "step": 27658 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.257184470023944e-05, + "loss": 2.3867, + "step": 27659 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.256539408340653e-05, + "loss": 2.5062, + "step": 27660 + }, + { + "epoch": 1.18, + "grad_norm": 0.5625, + "learning_rate": 7.255894359001932e-05, + "loss": 2.4735, + "step": 27661 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.25524932201068e-05, + "loss": 2.5355, + "step": 27662 + }, + { + "epoch": 1.18, + "grad_norm": 0.41796875, + "learning_rate": 7.254604297369805e-05, + "loss": 2.4667, + "step": 27663 + }, + { + "epoch": 1.18, + "grad_norm": 0.48828125, + "learning_rate": 7.25395928508221e-05, + "loss": 2.5392, + "step": 27664 + }, + { + "epoch": 1.18, + "grad_norm": 0.42578125, + "learning_rate": 7.253314285150793e-05, + "loss": 2.3941, + "step": 27665 + }, + { + "epoch": 1.18, + "grad_norm": 0.44140625, + "learning_rate": 7.252669297578456e-05, + "loss": 2.4008, + "step": 27666 + }, + { + "epoch": 1.18, + "grad_norm": 0.4453125, + "learning_rate": 7.252024322368105e-05, + "loss": 2.4123, + "step": 27667 + }, + { + "epoch": 1.18, + "grad_norm": 0.48828125, + "learning_rate": 7.25137935952264e-05, + "loss": 2.4614, + "step": 27668 + }, + { + "epoch": 1.18, + "grad_norm": 0.46484375, + "learning_rate": 7.250734409044963e-05, + "loss": 2.3702, + "step": 27669 + }, + { + "epoch": 1.18, + "grad_norm": 0.462890625, + "learning_rate": 7.250089470937975e-05, + "loss": 2.403, + "step": 27670 + }, + { + "epoch": 1.18, + "grad_norm": 0.4765625, + "learning_rate": 7.249444545204582e-05, + "loss": 2.5114, + "step": 27671 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.248799631847684e-05, + "loss": 2.5206, + "step": 27672 + }, + { + "epoch": 1.18, + "grad_norm": 0.466796875, + "learning_rate": 7.24815473087018e-05, + "loss": 2.5099, + "step": 27673 + }, + { + "epoch": 1.18, + "grad_norm": 0.5078125, + "learning_rate": 7.247509842274973e-05, + "loss": 2.5893, + "step": 27674 + }, + { + "epoch": 1.18, + "grad_norm": 0.4765625, + "learning_rate": 7.246864966064968e-05, + "loss": 2.4057, + "step": 27675 + }, + { + "epoch": 1.18, + "grad_norm": 0.451171875, + "learning_rate": 7.246220102243065e-05, + "loss": 2.3905, + "step": 27676 + }, + { + "epoch": 1.18, + "grad_norm": 0.439453125, + "learning_rate": 7.245575250812165e-05, + "loss": 2.4949, + "step": 27677 + }, + { + "epoch": 1.18, + "grad_norm": 0.4375, + "learning_rate": 7.244930411775166e-05, + "loss": 2.4754, + "step": 27678 + }, + { + "epoch": 1.18, + "grad_norm": 0.462890625, + "learning_rate": 7.244285585134976e-05, + "loss": 2.3901, + "step": 27679 + }, + { + "epoch": 1.18, + "grad_norm": 0.447265625, + "learning_rate": 7.243640770894497e-05, + "loss": 2.4338, + "step": 27680 + }, + { + "epoch": 1.18, + "grad_norm": 0.4375, + "learning_rate": 7.242995969056623e-05, + "loss": 2.4716, + "step": 27681 + }, + { + "epoch": 1.18, + "grad_norm": 0.4453125, + "learning_rate": 7.242351179624259e-05, + "loss": 2.527, + "step": 27682 + }, + { + "epoch": 1.18, + "grad_norm": 0.52734375, + "learning_rate": 7.24170640260031e-05, + "loss": 2.5413, + "step": 27683 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.241061637987671e-05, + "loss": 2.5361, + "step": 27684 + }, + { + "epoch": 1.18, + "grad_norm": 0.443359375, + "learning_rate": 7.240416885789249e-05, + "loss": 2.3825, + "step": 27685 + }, + { + "epoch": 1.18, + "grad_norm": 0.48046875, + "learning_rate": 7.239772146007939e-05, + "loss": 2.567, + "step": 27686 + }, + { + "epoch": 1.18, + "grad_norm": 0.392578125, + "learning_rate": 7.239127418646649e-05, + "loss": 2.3255, + "step": 27687 + }, + { + "epoch": 1.18, + "grad_norm": 0.4296875, + "learning_rate": 7.238482703708278e-05, + "loss": 2.3975, + "step": 27688 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.237838001195725e-05, + "loss": 2.5344, + "step": 27689 + }, + { + "epoch": 1.18, + "grad_norm": 0.494140625, + "learning_rate": 7.237193311111888e-05, + "loss": 2.5349, + "step": 27690 + }, + { + "epoch": 1.18, + "grad_norm": 0.478515625, + "learning_rate": 7.236548633459672e-05, + "loss": 2.4802, + "step": 27691 + }, + { + "epoch": 1.18, + "grad_norm": 0.498046875, + "learning_rate": 7.235903968241979e-05, + "loss": 2.4813, + "step": 27692 + }, + { + "epoch": 1.18, + "grad_norm": 0.515625, + "learning_rate": 7.235259315461709e-05, + "loss": 2.5512, + "step": 27693 + }, + { + "epoch": 1.18, + "grad_norm": 0.5078125, + "learning_rate": 7.234614675121758e-05, + "loss": 2.377, + "step": 27694 + }, + { + "epoch": 1.18, + "grad_norm": 0.4609375, + "learning_rate": 7.233970047225031e-05, + "loss": 2.4599, + "step": 27695 + }, + { + "epoch": 1.18, + "grad_norm": 0.47265625, + "learning_rate": 7.233325431774434e-05, + "loss": 2.3657, + "step": 27696 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.232680828772855e-05, + "loss": 2.482, + "step": 27697 + }, + { + "epoch": 1.18, + "grad_norm": 0.48046875, + "learning_rate": 7.232036238223202e-05, + "loss": 2.5353, + "step": 27698 + }, + { + "epoch": 1.18, + "grad_norm": 0.4453125, + "learning_rate": 7.231391660128375e-05, + "loss": 2.5372, + "step": 27699 + }, + { + "epoch": 1.18, + "grad_norm": 0.4765625, + "learning_rate": 7.230747094491274e-05, + "loss": 2.4564, + "step": 27700 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.230102541314801e-05, + "loss": 2.2725, + "step": 27701 + }, + { + "epoch": 1.18, + "grad_norm": 0.51171875, + "learning_rate": 7.22945800060185e-05, + "loss": 2.4969, + "step": 27702 + }, + { + "epoch": 1.18, + "grad_norm": 0.427734375, + "learning_rate": 7.228813472355327e-05, + "loss": 2.3698, + "step": 27703 + }, + { + "epoch": 1.18, + "grad_norm": 0.546875, + "learning_rate": 7.228168956578134e-05, + "loss": 2.5004, + "step": 27704 + }, + { + "epoch": 1.18, + "grad_norm": 0.427734375, + "learning_rate": 7.227524453273165e-05, + "loss": 2.5191, + "step": 27705 + }, + { + "epoch": 1.18, + "grad_norm": 0.4375, + "learning_rate": 7.22687996244332e-05, + "loss": 2.3949, + "step": 27706 + }, + { + "epoch": 1.18, + "grad_norm": 0.5234375, + "learning_rate": 7.226235484091505e-05, + "loss": 2.3915, + "step": 27707 + }, + { + "epoch": 1.18, + "grad_norm": 0.470703125, + "learning_rate": 7.225591018220617e-05, + "loss": 2.5283, + "step": 27708 + }, + { + "epoch": 1.18, + "grad_norm": 0.466796875, + "learning_rate": 7.224946564833554e-05, + "loss": 2.4607, + "step": 27709 + }, + { + "epoch": 1.18, + "grad_norm": 0.5078125, + "learning_rate": 7.224302123933217e-05, + "loss": 2.4206, + "step": 27710 + }, + { + "epoch": 1.18, + "grad_norm": 0.474609375, + "learning_rate": 7.223657695522508e-05, + "loss": 2.4448, + "step": 27711 + }, + { + "epoch": 1.18, + "grad_norm": 0.4921875, + "learning_rate": 7.223013279604325e-05, + "loss": 2.5441, + "step": 27712 + }, + { + "epoch": 1.18, + "grad_norm": 0.47265625, + "learning_rate": 7.222368876181566e-05, + "loss": 2.4594, + "step": 27713 + }, + { + "epoch": 1.18, + "grad_norm": 0.439453125, + "learning_rate": 7.221724485257131e-05, + "loss": 2.5469, + "step": 27714 + }, + { + "epoch": 1.18, + "grad_norm": 0.455078125, + "learning_rate": 7.221080106833921e-05, + "loss": 2.4776, + "step": 27715 + }, + { + "epoch": 1.18, + "grad_norm": 0.470703125, + "learning_rate": 7.220435740914835e-05, + "loss": 2.5769, + "step": 27716 + }, + { + "epoch": 1.18, + "grad_norm": 0.431640625, + "learning_rate": 7.219791387502774e-05, + "loss": 2.4203, + "step": 27717 + }, + { + "epoch": 1.18, + "grad_norm": 0.46875, + "learning_rate": 7.219147046600632e-05, + "loss": 2.4215, + "step": 27718 + }, + { + "epoch": 1.18, + "grad_norm": 0.404296875, + "learning_rate": 7.218502718211316e-05, + "loss": 2.4313, + "step": 27719 + }, + { + "epoch": 1.18, + "grad_norm": 0.54296875, + "learning_rate": 7.217858402337721e-05, + "loss": 2.5179, + "step": 27720 + }, + { + "epoch": 1.18, + "grad_norm": 0.4453125, + "learning_rate": 7.217214098982745e-05, + "loss": 2.4117, + "step": 27721 + }, + { + "epoch": 1.18, + "grad_norm": 0.451171875, + "learning_rate": 7.216569808149288e-05, + "loss": 2.417, + "step": 27722 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.215925529840249e-05, + "loss": 2.4128, + "step": 27723 + }, + { + "epoch": 1.18, + "grad_norm": 0.462890625, + "learning_rate": 7.21528126405853e-05, + "loss": 2.4327, + "step": 27724 + }, + { + "epoch": 1.18, + "grad_norm": 0.43359375, + "learning_rate": 7.214637010807026e-05, + "loss": 2.476, + "step": 27725 + }, + { + "epoch": 1.18, + "grad_norm": 0.474609375, + "learning_rate": 7.213992770088636e-05, + "loss": 2.4737, + "step": 27726 + }, + { + "epoch": 1.18, + "grad_norm": 0.46875, + "learning_rate": 7.213348541906263e-05, + "loss": 2.4822, + "step": 27727 + }, + { + "epoch": 1.18, + "grad_norm": 0.435546875, + "learning_rate": 7.212704326262805e-05, + "loss": 2.3499, + "step": 27728 + }, + { + "epoch": 1.18, + "grad_norm": 0.49609375, + "learning_rate": 7.212060123161157e-05, + "loss": 2.5161, + "step": 27729 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.211415932604217e-05, + "loss": 2.5348, + "step": 27730 + }, + { + "epoch": 1.18, + "grad_norm": 0.416015625, + "learning_rate": 7.21077175459489e-05, + "loss": 2.3667, + "step": 27731 + }, + { + "epoch": 1.18, + "grad_norm": 0.51953125, + "learning_rate": 7.210127589136068e-05, + "loss": 2.3765, + "step": 27732 + }, + { + "epoch": 1.18, + "grad_norm": 0.443359375, + "learning_rate": 7.209483436230655e-05, + "loss": 2.4511, + "step": 27733 + }, + { + "epoch": 1.18, + "grad_norm": 0.453125, + "learning_rate": 7.208839295881542e-05, + "loss": 2.5961, + "step": 27734 + }, + { + "epoch": 1.18, + "grad_norm": 0.439453125, + "learning_rate": 7.208195168091635e-05, + "loss": 2.4713, + "step": 27735 + }, + { + "epoch": 1.18, + "grad_norm": 0.4375, + "learning_rate": 7.207551052863833e-05, + "loss": 2.4307, + "step": 27736 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.206906950201027e-05, + "loss": 2.5495, + "step": 27737 + }, + { + "epoch": 1.18, + "grad_norm": 0.451171875, + "learning_rate": 7.206262860106118e-05, + "loss": 2.4364, + "step": 27738 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.205618782582007e-05, + "loss": 2.6067, + "step": 27739 + }, + { + "epoch": 1.18, + "grad_norm": 0.498046875, + "learning_rate": 7.20497471763159e-05, + "loss": 2.5203, + "step": 27740 + }, + { + "epoch": 1.18, + "grad_norm": 0.45703125, + "learning_rate": 7.204330665257764e-05, + "loss": 2.3719, + "step": 27741 + }, + { + "epoch": 1.18, + "grad_norm": 0.4140625, + "learning_rate": 7.203686625463428e-05, + "loss": 2.4726, + "step": 27742 + }, + { + "epoch": 1.18, + "grad_norm": 0.47265625, + "learning_rate": 7.203042598251482e-05, + "loss": 2.5524, + "step": 27743 + }, + { + "epoch": 1.18, + "grad_norm": 0.4296875, + "learning_rate": 7.202398583624823e-05, + "loss": 2.4734, + "step": 27744 + }, + { + "epoch": 1.18, + "grad_norm": 0.423828125, + "learning_rate": 7.201754581586348e-05, + "loss": 2.4036, + "step": 27745 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.201110592138951e-05, + "loss": 2.5202, + "step": 27746 + }, + { + "epoch": 1.18, + "grad_norm": 0.45703125, + "learning_rate": 7.200466615285535e-05, + "loss": 2.3528, + "step": 27747 + }, + { + "epoch": 1.18, + "grad_norm": 0.435546875, + "learning_rate": 7.199822651028997e-05, + "loss": 2.3634, + "step": 27748 + }, + { + "epoch": 1.18, + "grad_norm": 0.447265625, + "learning_rate": 7.199178699372236e-05, + "loss": 2.2953, + "step": 27749 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.198534760318142e-05, + "loss": 2.4668, + "step": 27750 + }, + { + "epoch": 1.18, + "grad_norm": 0.498046875, + "learning_rate": 7.197890833869622e-05, + "loss": 2.4444, + "step": 27751 + }, + { + "epoch": 1.18, + "grad_norm": 0.392578125, + "learning_rate": 7.19724692002957e-05, + "loss": 2.4695, + "step": 27752 + }, + { + "epoch": 1.18, + "grad_norm": 0.46484375, + "learning_rate": 7.19660301880088e-05, + "loss": 2.4284, + "step": 27753 + }, + { + "epoch": 1.18, + "grad_norm": 0.474609375, + "learning_rate": 7.19595913018645e-05, + "loss": 2.4245, + "step": 27754 + }, + { + "epoch": 1.18, + "grad_norm": 0.419921875, + "learning_rate": 7.195315254189183e-05, + "loss": 2.4666, + "step": 27755 + }, + { + "epoch": 1.18, + "grad_norm": 0.478515625, + "learning_rate": 7.19467139081197e-05, + "loss": 2.4604, + "step": 27756 + }, + { + "epoch": 1.18, + "grad_norm": 0.44921875, + "learning_rate": 7.194027540057713e-05, + "loss": 2.4347, + "step": 27757 + }, + { + "epoch": 1.18, + "grad_norm": 0.48046875, + "learning_rate": 7.193383701929303e-05, + "loss": 2.507, + "step": 27758 + }, + { + "epoch": 1.18, + "grad_norm": 0.453125, + "learning_rate": 7.192739876429642e-05, + "loss": 2.3952, + "step": 27759 + }, + { + "epoch": 1.18, + "grad_norm": 0.45703125, + "learning_rate": 7.192096063561629e-05, + "loss": 2.5004, + "step": 27760 + }, + { + "epoch": 1.18, + "grad_norm": 0.494140625, + "learning_rate": 7.191452263328156e-05, + "loss": 2.4546, + "step": 27761 + }, + { + "epoch": 1.18, + "grad_norm": 0.4765625, + "learning_rate": 7.190808475732116e-05, + "loss": 2.4906, + "step": 27762 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.190164700776417e-05, + "loss": 2.4491, + "step": 27763 + }, + { + "epoch": 1.18, + "grad_norm": 0.48828125, + "learning_rate": 7.18952093846395e-05, + "loss": 2.4963, + "step": 27764 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.188877188797609e-05, + "loss": 2.5269, + "step": 27765 + }, + { + "epoch": 1.18, + "grad_norm": 0.482421875, + "learning_rate": 7.188233451780297e-05, + "loss": 2.3808, + "step": 27766 + }, + { + "epoch": 1.18, + "grad_norm": 0.482421875, + "learning_rate": 7.187589727414901e-05, + "loss": 2.4544, + "step": 27767 + }, + { + "epoch": 1.18, + "grad_norm": 0.470703125, + "learning_rate": 7.186946015704327e-05, + "loss": 2.4926, + "step": 27768 + }, + { + "epoch": 1.18, + "grad_norm": 0.46484375, + "learning_rate": 7.186302316651469e-05, + "loss": 2.3108, + "step": 27769 + }, + { + "epoch": 1.18, + "grad_norm": 0.404296875, + "learning_rate": 7.185658630259221e-05, + "loss": 2.3749, + "step": 27770 + }, + { + "epoch": 1.18, + "grad_norm": 0.470703125, + "learning_rate": 7.185014956530479e-05, + "loss": 2.442, + "step": 27771 + }, + { + "epoch": 1.18, + "grad_norm": 0.44921875, + "learning_rate": 7.184371295468141e-05, + "loss": 2.3239, + "step": 27772 + }, + { + "epoch": 1.18, + "grad_norm": 0.490234375, + "learning_rate": 7.183727647075106e-05, + "loss": 2.5754, + "step": 27773 + }, + { + "epoch": 1.18, + "grad_norm": 0.5546875, + "learning_rate": 7.183084011354268e-05, + "loss": 2.6143, + "step": 27774 + }, + { + "epoch": 1.18, + "grad_norm": 0.435546875, + "learning_rate": 7.182440388308516e-05, + "loss": 2.3643, + "step": 27775 + }, + { + "epoch": 1.18, + "grad_norm": 0.4921875, + "learning_rate": 7.18179677794076e-05, + "loss": 2.3744, + "step": 27776 + }, + { + "epoch": 1.18, + "grad_norm": 0.5625, + "learning_rate": 7.181153180253886e-05, + "loss": 2.4458, + "step": 27777 + }, + { + "epoch": 1.18, + "grad_norm": 0.46875, + "learning_rate": 7.180509595250792e-05, + "loss": 2.3752, + "step": 27778 + }, + { + "epoch": 1.18, + "grad_norm": 0.5, + "learning_rate": 7.179866022934372e-05, + "loss": 2.4753, + "step": 27779 + }, + { + "epoch": 1.18, + "grad_norm": 0.4375, + "learning_rate": 7.179222463307528e-05, + "loss": 2.5225, + "step": 27780 + }, + { + "epoch": 1.18, + "grad_norm": 0.447265625, + "learning_rate": 7.17857891637315e-05, + "loss": 2.595, + "step": 27781 + }, + { + "epoch": 1.18, + "grad_norm": 0.4140625, + "learning_rate": 7.17793538213414e-05, + "loss": 2.3422, + "step": 27782 + }, + { + "epoch": 1.18, + "grad_norm": 0.41796875, + "learning_rate": 7.177291860593381e-05, + "loss": 2.3593, + "step": 27783 + }, + { + "epoch": 1.18, + "grad_norm": 0.38671875, + "learning_rate": 7.176648351753785e-05, + "loss": 2.4336, + "step": 27784 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.176004855618237e-05, + "loss": 2.5748, + "step": 27785 + }, + { + "epoch": 1.18, + "grad_norm": 0.4140625, + "learning_rate": 7.175361372189636e-05, + "loss": 2.4596, + "step": 27786 + }, + { + "epoch": 1.18, + "grad_norm": 0.4765625, + "learning_rate": 7.174717901470873e-05, + "loss": 2.4436, + "step": 27787 + }, + { + "epoch": 1.18, + "grad_norm": 0.41796875, + "learning_rate": 7.174074443464849e-05, + "loss": 2.3535, + "step": 27788 + }, + { + "epoch": 1.18, + "grad_norm": 0.439453125, + "learning_rate": 7.173430998174458e-05, + "loss": 2.4198, + "step": 27789 + }, + { + "epoch": 1.18, + "grad_norm": 0.498046875, + "learning_rate": 7.172787565602596e-05, + "loss": 2.5222, + "step": 27790 + }, + { + "epoch": 1.18, + "grad_norm": 0.44921875, + "learning_rate": 7.172144145752151e-05, + "loss": 2.53, + "step": 27791 + }, + { + "epoch": 1.18, + "grad_norm": 0.498046875, + "learning_rate": 7.171500738626029e-05, + "loss": 2.583, + "step": 27792 + }, + { + "epoch": 1.18, + "grad_norm": 0.50390625, + "learning_rate": 7.170857344227119e-05, + "loss": 2.4842, + "step": 27793 + }, + { + "epoch": 1.18, + "grad_norm": 0.48046875, + "learning_rate": 7.170213962558316e-05, + "loss": 2.6774, + "step": 27794 + }, + { + "epoch": 1.18, + "grad_norm": 0.49609375, + "learning_rate": 7.169570593622515e-05, + "loss": 2.4381, + "step": 27795 + }, + { + "epoch": 1.18, + "grad_norm": 0.423828125, + "learning_rate": 7.168927237422613e-05, + "loss": 2.4837, + "step": 27796 + }, + { + "epoch": 1.18, + "grad_norm": 0.486328125, + "learning_rate": 7.168283893961505e-05, + "loss": 2.5712, + "step": 27797 + }, + { + "epoch": 1.18, + "grad_norm": 0.490234375, + "learning_rate": 7.167640563242086e-05, + "loss": 2.5689, + "step": 27798 + }, + { + "epoch": 1.18, + "grad_norm": 0.5234375, + "learning_rate": 7.166997245267245e-05, + "loss": 2.4977, + "step": 27799 + }, + { + "epoch": 1.18, + "grad_norm": 0.478515625, + "learning_rate": 7.166353940039885e-05, + "loss": 2.3678, + "step": 27800 + }, + { + "epoch": 1.18, + "grad_norm": 0.44921875, + "learning_rate": 7.165710647562895e-05, + "loss": 2.4929, + "step": 27801 + }, + { + "epoch": 1.18, + "grad_norm": 0.466796875, + "learning_rate": 7.165067367839171e-05, + "loss": 2.4776, + "step": 27802 + }, + { + "epoch": 1.18, + "grad_norm": 0.470703125, + "learning_rate": 7.164424100871606e-05, + "loss": 2.4576, + "step": 27803 + }, + { + "epoch": 1.18, + "grad_norm": 0.435546875, + "learning_rate": 7.1637808466631e-05, + "loss": 2.4673, + "step": 27804 + }, + { + "epoch": 1.18, + "grad_norm": 0.4375, + "learning_rate": 7.16313760521654e-05, + "loss": 2.4811, + "step": 27805 + }, + { + "epoch": 1.18, + "grad_norm": 0.51953125, + "learning_rate": 7.162494376534829e-05, + "loss": 2.4318, + "step": 27806 + }, + { + "epoch": 1.18, + "grad_norm": 0.462890625, + "learning_rate": 7.161851160620849e-05, + "loss": 2.4712, + "step": 27807 + }, + { + "epoch": 1.18, + "grad_norm": 0.474609375, + "learning_rate": 7.161207957477506e-05, + "loss": 2.3945, + "step": 27808 + }, + { + "epoch": 1.19, + "grad_norm": 0.470703125, + "learning_rate": 7.160564767107693e-05, + "loss": 2.3948, + "step": 27809 + }, + { + "epoch": 1.19, + "grad_norm": 0.375, + "learning_rate": 7.159921589514298e-05, + "loss": 2.3663, + "step": 27810 + }, + { + "epoch": 1.19, + "grad_norm": 0.5625, + "learning_rate": 7.159278424700215e-05, + "loss": 2.5405, + "step": 27811 + }, + { + "epoch": 1.19, + "grad_norm": 0.45703125, + "learning_rate": 7.158635272668343e-05, + "loss": 2.4673, + "step": 27812 + }, + { + "epoch": 1.19, + "grad_norm": 0.439453125, + "learning_rate": 7.157992133421575e-05, + "loss": 2.39, + "step": 27813 + }, + { + "epoch": 1.19, + "grad_norm": 0.462890625, + "learning_rate": 7.157349006962805e-05, + "loss": 2.4477, + "step": 27814 + }, + { + "epoch": 1.19, + "grad_norm": 0.53515625, + "learning_rate": 7.156705893294919e-05, + "loss": 2.3862, + "step": 27815 + }, + { + "epoch": 1.19, + "grad_norm": 0.443359375, + "learning_rate": 7.156062792420823e-05, + "loss": 2.3888, + "step": 27816 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.155419704343407e-05, + "loss": 2.4437, + "step": 27817 + }, + { + "epoch": 1.19, + "grad_norm": 0.5546875, + "learning_rate": 7.154776629065559e-05, + "loss": 2.5987, + "step": 27818 + }, + { + "epoch": 1.19, + "grad_norm": 0.435546875, + "learning_rate": 7.154133566590176e-05, + "loss": 2.3898, + "step": 27819 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.153490516920152e-05, + "loss": 2.4215, + "step": 27820 + }, + { + "epoch": 1.19, + "grad_norm": 0.48828125, + "learning_rate": 7.152847480058383e-05, + "loss": 2.4785, + "step": 27821 + }, + { + "epoch": 1.19, + "grad_norm": 0.4921875, + "learning_rate": 7.152204456007759e-05, + "loss": 2.4096, + "step": 27822 + }, + { + "epoch": 1.19, + "grad_norm": 0.40234375, + "learning_rate": 7.15156144477117e-05, + "loss": 2.4673, + "step": 27823 + }, + { + "epoch": 1.19, + "grad_norm": 0.43359375, + "learning_rate": 7.150918446351518e-05, + "loss": 2.5271, + "step": 27824 + }, + { + "epoch": 1.19, + "grad_norm": 0.486328125, + "learning_rate": 7.150275460751692e-05, + "loss": 2.5269, + "step": 27825 + }, + { + "epoch": 1.19, + "grad_norm": 0.48046875, + "learning_rate": 7.149632487974585e-05, + "loss": 2.48, + "step": 27826 + }, + { + "epoch": 1.19, + "grad_norm": 0.4453125, + "learning_rate": 7.148989528023087e-05, + "loss": 2.6022, + "step": 27827 + }, + { + "epoch": 1.19, + "grad_norm": 0.5, + "learning_rate": 7.148346580900096e-05, + "loss": 2.5734, + "step": 27828 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.147703646608503e-05, + "loss": 2.3744, + "step": 27829 + }, + { + "epoch": 1.19, + "grad_norm": 0.396484375, + "learning_rate": 7.147060725151203e-05, + "loss": 2.3831, + "step": 27830 + }, + { + "epoch": 1.19, + "grad_norm": 0.458984375, + "learning_rate": 7.146417816531081e-05, + "loss": 2.457, + "step": 27831 + }, + { + "epoch": 1.19, + "grad_norm": 0.48046875, + "learning_rate": 7.145774920751042e-05, + "loss": 2.5139, + "step": 27832 + }, + { + "epoch": 1.19, + "grad_norm": 0.470703125, + "learning_rate": 7.145132037813973e-05, + "loss": 2.4673, + "step": 27833 + }, + { + "epoch": 1.19, + "grad_norm": 0.474609375, + "learning_rate": 7.144489167722765e-05, + "loss": 2.5813, + "step": 27834 + }, + { + "epoch": 1.19, + "grad_norm": 0.486328125, + "learning_rate": 7.143846310480311e-05, + "loss": 2.4221, + "step": 27835 + }, + { + "epoch": 1.19, + "grad_norm": 0.41796875, + "learning_rate": 7.143203466089505e-05, + "loss": 2.3494, + "step": 27836 + }, + { + "epoch": 1.19, + "grad_norm": 0.45703125, + "learning_rate": 7.14256063455324e-05, + "loss": 2.4219, + "step": 27837 + }, + { + "epoch": 1.19, + "grad_norm": 0.498046875, + "learning_rate": 7.141917815874411e-05, + "loss": 2.4617, + "step": 27838 + }, + { + "epoch": 1.19, + "grad_norm": 0.48828125, + "learning_rate": 7.1412750100559e-05, + "loss": 2.381, + "step": 27839 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.140632217100611e-05, + "loss": 2.451, + "step": 27840 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.139989437011434e-05, + "loss": 2.4412, + "step": 27841 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.139346669791259e-05, + "loss": 2.328, + "step": 27842 + }, + { + "epoch": 1.19, + "grad_norm": 0.52734375, + "learning_rate": 7.138703915442975e-05, + "loss": 2.5081, + "step": 27843 + }, + { + "epoch": 1.19, + "grad_norm": 0.49609375, + "learning_rate": 7.138061173969479e-05, + "loss": 2.4902, + "step": 27844 + }, + { + "epoch": 1.19, + "grad_norm": 0.470703125, + "learning_rate": 7.137418445373663e-05, + "loss": 2.5251, + "step": 27845 + }, + { + "epoch": 1.19, + "grad_norm": 0.478515625, + "learning_rate": 7.136775729658419e-05, + "loss": 2.468, + "step": 27846 + }, + { + "epoch": 1.19, + "grad_norm": 0.435546875, + "learning_rate": 7.136133026826632e-05, + "loss": 2.4948, + "step": 27847 + }, + { + "epoch": 1.19, + "grad_norm": 0.482421875, + "learning_rate": 7.135490336881204e-05, + "loss": 2.4143, + "step": 27848 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.134847659825024e-05, + "loss": 2.4288, + "step": 27849 + }, + { + "epoch": 1.19, + "grad_norm": 0.498046875, + "learning_rate": 7.134204995660982e-05, + "loss": 2.4319, + "step": 27850 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.133562344391967e-05, + "loss": 2.4801, + "step": 27851 + }, + { + "epoch": 1.19, + "grad_norm": 0.423828125, + "learning_rate": 7.132919706020876e-05, + "loss": 2.3429, + "step": 27852 + }, + { + "epoch": 1.19, + "grad_norm": 0.49609375, + "learning_rate": 7.132277080550599e-05, + "loss": 2.399, + "step": 27853 + }, + { + "epoch": 1.19, + "grad_norm": 0.474609375, + "learning_rate": 7.13163446798403e-05, + "loss": 2.4058, + "step": 27854 + }, + { + "epoch": 1.19, + "grad_norm": 0.447265625, + "learning_rate": 7.130991868324049e-05, + "loss": 2.406, + "step": 27855 + }, + { + "epoch": 1.19, + "grad_norm": 0.4921875, + "learning_rate": 7.130349281573561e-05, + "loss": 2.3804, + "step": 27856 + }, + { + "epoch": 1.19, + "grad_norm": 0.474609375, + "learning_rate": 7.129706707735455e-05, + "loss": 2.4136, + "step": 27857 + }, + { + "epoch": 1.19, + "grad_norm": 0.404296875, + "learning_rate": 7.129064146812618e-05, + "loss": 2.4575, + "step": 27858 + }, + { + "epoch": 1.19, + "grad_norm": 0.494140625, + "learning_rate": 7.128421598807942e-05, + "loss": 2.3657, + "step": 27859 + }, + { + "epoch": 1.19, + "grad_norm": 0.470703125, + "learning_rate": 7.12777906372432e-05, + "loss": 2.4178, + "step": 27860 + }, + { + "epoch": 1.19, + "grad_norm": 0.447265625, + "learning_rate": 7.127136541564643e-05, + "loss": 2.4696, + "step": 27861 + }, + { + "epoch": 1.19, + "grad_norm": 0.451171875, + "learning_rate": 7.1264940323318e-05, + "loss": 2.3275, + "step": 27862 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.125851536028685e-05, + "loss": 2.5263, + "step": 27863 + }, + { + "epoch": 1.19, + "grad_norm": 0.41015625, + "learning_rate": 7.125209052658188e-05, + "loss": 2.3577, + "step": 27864 + }, + { + "epoch": 1.19, + "grad_norm": 0.515625, + "learning_rate": 7.124566582223201e-05, + "loss": 2.4871, + "step": 27865 + }, + { + "epoch": 1.19, + "grad_norm": 0.4609375, + "learning_rate": 7.123924124726611e-05, + "loss": 2.6439, + "step": 27866 + }, + { + "epoch": 1.19, + "grad_norm": 0.484375, + "learning_rate": 7.123281680171312e-05, + "loss": 2.5824, + "step": 27867 + }, + { + "epoch": 1.19, + "grad_norm": 0.53125, + "learning_rate": 7.122639248560195e-05, + "loss": 2.4548, + "step": 27868 + }, + { + "epoch": 1.19, + "grad_norm": 0.58203125, + "learning_rate": 7.121996829896148e-05, + "loss": 2.5229, + "step": 27869 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.121354424182065e-05, + "loss": 2.3126, + "step": 27870 + }, + { + "epoch": 1.19, + "grad_norm": 0.412109375, + "learning_rate": 7.120712031420833e-05, + "loss": 2.4116, + "step": 27871 + }, + { + "epoch": 1.19, + "grad_norm": 0.451171875, + "learning_rate": 7.120069651615347e-05, + "loss": 2.2909, + "step": 27872 + }, + { + "epoch": 1.19, + "grad_norm": 0.5390625, + "learning_rate": 7.119427284768497e-05, + "loss": 2.4641, + "step": 27873 + }, + { + "epoch": 1.19, + "grad_norm": 0.44140625, + "learning_rate": 7.118784930883169e-05, + "loss": 2.3455, + "step": 27874 + }, + { + "epoch": 1.19, + "grad_norm": 0.498046875, + "learning_rate": 7.118142589962253e-05, + "loss": 2.4176, + "step": 27875 + }, + { + "epoch": 1.19, + "grad_norm": 0.494140625, + "learning_rate": 7.117500262008645e-05, + "loss": 2.4051, + "step": 27876 + }, + { + "epoch": 1.19, + "grad_norm": 0.439453125, + "learning_rate": 7.116857947025233e-05, + "loss": 2.441, + "step": 27877 + }, + { + "epoch": 1.19, + "grad_norm": 0.474609375, + "learning_rate": 7.116215645014905e-05, + "loss": 2.6348, + "step": 27878 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.115573355980553e-05, + "loss": 2.5627, + "step": 27879 + }, + { + "epoch": 1.19, + "grad_norm": 0.546875, + "learning_rate": 7.114931079925067e-05, + "loss": 2.5561, + "step": 27880 + }, + { + "epoch": 1.19, + "grad_norm": 0.447265625, + "learning_rate": 7.11428881685134e-05, + "loss": 2.523, + "step": 27881 + }, + { + "epoch": 1.19, + "grad_norm": 0.494140625, + "learning_rate": 7.113646566762255e-05, + "loss": 2.447, + "step": 27882 + }, + { + "epoch": 1.19, + "grad_norm": 0.47265625, + "learning_rate": 7.113004329660705e-05, + "loss": 2.3587, + "step": 27883 + }, + { + "epoch": 1.19, + "grad_norm": 0.546875, + "learning_rate": 7.112362105549581e-05, + "loss": 2.481, + "step": 27884 + }, + { + "epoch": 1.19, + "grad_norm": 0.466796875, + "learning_rate": 7.111719894431774e-05, + "loss": 2.4524, + "step": 27885 + }, + { + "epoch": 1.19, + "grad_norm": 0.4453125, + "learning_rate": 7.111077696310171e-05, + "loss": 2.4986, + "step": 27886 + }, + { + "epoch": 1.19, + "grad_norm": 0.47265625, + "learning_rate": 7.110435511187665e-05, + "loss": 2.4523, + "step": 27887 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.109793339067137e-05, + "loss": 2.4652, + "step": 27888 + }, + { + "epoch": 1.19, + "grad_norm": 0.4609375, + "learning_rate": 7.109151179951489e-05, + "loss": 2.5119, + "step": 27889 + }, + { + "epoch": 1.19, + "grad_norm": 0.423828125, + "learning_rate": 7.108509033843602e-05, + "loss": 2.5361, + "step": 27890 + }, + { + "epoch": 1.19, + "grad_norm": 0.45703125, + "learning_rate": 7.107866900746367e-05, + "loss": 2.4129, + "step": 27891 + }, + { + "epoch": 1.19, + "grad_norm": 0.453125, + "learning_rate": 7.107224780662674e-05, + "loss": 2.4708, + "step": 27892 + }, + { + "epoch": 1.19, + "grad_norm": 0.51171875, + "learning_rate": 7.106582673595412e-05, + "loss": 2.3814, + "step": 27893 + }, + { + "epoch": 1.19, + "grad_norm": 0.4140625, + "learning_rate": 7.105940579547471e-05, + "loss": 2.4808, + "step": 27894 + }, + { + "epoch": 1.19, + "grad_norm": 0.5625, + "learning_rate": 7.105298498521742e-05, + "loss": 2.6381, + "step": 27895 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.104656430521106e-05, + "loss": 2.5123, + "step": 27896 + }, + { + "epoch": 1.19, + "grad_norm": 0.4921875, + "learning_rate": 7.104014375548465e-05, + "loss": 2.496, + "step": 27897 + }, + { + "epoch": 1.19, + "grad_norm": 0.486328125, + "learning_rate": 7.103372333606697e-05, + "loss": 2.4102, + "step": 27898 + }, + { + "epoch": 1.19, + "grad_norm": 0.453125, + "learning_rate": 7.102730304698696e-05, + "loss": 2.4384, + "step": 27899 + }, + { + "epoch": 1.19, + "grad_norm": 0.431640625, + "learning_rate": 7.102088288827347e-05, + "loss": 2.4815, + "step": 27900 + }, + { + "epoch": 1.19, + "grad_norm": 0.5078125, + "learning_rate": 7.101446285995545e-05, + "loss": 2.4989, + "step": 27901 + }, + { + "epoch": 1.19, + "grad_norm": 0.42578125, + "learning_rate": 7.100804296206173e-05, + "loss": 2.4574, + "step": 27902 + }, + { + "epoch": 1.19, + "grad_norm": 0.435546875, + "learning_rate": 7.100162319462127e-05, + "loss": 2.4353, + "step": 27903 + }, + { + "epoch": 1.19, + "grad_norm": 0.47265625, + "learning_rate": 7.099520355766283e-05, + "loss": 2.4486, + "step": 27904 + }, + { + "epoch": 1.19, + "grad_norm": 0.419921875, + "learning_rate": 7.098878405121544e-05, + "loss": 2.4174, + "step": 27905 + }, + { + "epoch": 1.19, + "grad_norm": 0.431640625, + "learning_rate": 7.09823646753079e-05, + "loss": 2.5312, + "step": 27906 + }, + { + "epoch": 1.19, + "grad_norm": 0.48828125, + "learning_rate": 7.097594542996911e-05, + "loss": 2.5986, + "step": 27907 + }, + { + "epoch": 1.19, + "grad_norm": 0.51171875, + "learning_rate": 7.096952631522795e-05, + "loss": 2.502, + "step": 27908 + }, + { + "epoch": 1.19, + "grad_norm": 0.484375, + "learning_rate": 7.096310733111331e-05, + "loss": 2.5045, + "step": 27909 + }, + { + "epoch": 1.19, + "grad_norm": 0.55078125, + "learning_rate": 7.09566884776541e-05, + "loss": 2.474, + "step": 27910 + }, + { + "epoch": 1.19, + "grad_norm": 0.400390625, + "learning_rate": 7.095026975487919e-05, + "loss": 2.4717, + "step": 27911 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.094385116281738e-05, + "loss": 2.3979, + "step": 27912 + }, + { + "epoch": 1.19, + "grad_norm": 0.435546875, + "learning_rate": 7.093743270149768e-05, + "loss": 2.3553, + "step": 27913 + }, + { + "epoch": 1.19, + "grad_norm": 0.47265625, + "learning_rate": 7.093101437094892e-05, + "loss": 2.3851, + "step": 27914 + }, + { + "epoch": 1.19, + "grad_norm": 0.56640625, + "learning_rate": 7.092459617119993e-05, + "loss": 2.6607, + "step": 27915 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.091817810227963e-05, + "loss": 2.4482, + "step": 27916 + }, + { + "epoch": 1.19, + "grad_norm": 0.48828125, + "learning_rate": 7.091176016421692e-05, + "loss": 2.5287, + "step": 27917 + }, + { + "epoch": 1.19, + "grad_norm": 0.486328125, + "learning_rate": 7.090534235704067e-05, + "loss": 2.6261, + "step": 27918 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.089892468077977e-05, + "loss": 2.3486, + "step": 27919 + }, + { + "epoch": 1.19, + "grad_norm": 0.42578125, + "learning_rate": 7.089250713546298e-05, + "loss": 2.4153, + "step": 27920 + }, + { + "epoch": 1.19, + "grad_norm": 0.478515625, + "learning_rate": 7.088608972111937e-05, + "loss": 2.5135, + "step": 27921 + }, + { + "epoch": 1.19, + "grad_norm": 0.45703125, + "learning_rate": 7.087967243777767e-05, + "loss": 2.5265, + "step": 27922 + }, + { + "epoch": 1.19, + "grad_norm": 0.478515625, + "learning_rate": 7.08732552854668e-05, + "loss": 2.6204, + "step": 27923 + }, + { + "epoch": 1.19, + "grad_norm": 0.47265625, + "learning_rate": 7.086683826421563e-05, + "loss": 2.5098, + "step": 27924 + }, + { + "epoch": 1.19, + "grad_norm": 0.578125, + "learning_rate": 7.086042137405306e-05, + "loss": 2.6103, + "step": 27925 + }, + { + "epoch": 1.19, + "grad_norm": 0.45703125, + "learning_rate": 7.085400461500794e-05, + "loss": 2.4766, + "step": 27926 + }, + { + "epoch": 1.19, + "grad_norm": 0.515625, + "learning_rate": 7.084758798710918e-05, + "loss": 2.4274, + "step": 27927 + }, + { + "epoch": 1.19, + "grad_norm": 0.421875, + "learning_rate": 7.084117149038554e-05, + "loss": 2.4689, + "step": 27928 + }, + { + "epoch": 1.19, + "grad_norm": 0.486328125, + "learning_rate": 7.083475512486603e-05, + "loss": 2.5292, + "step": 27929 + }, + { + "epoch": 1.19, + "grad_norm": 0.4296875, + "learning_rate": 7.082833889057946e-05, + "loss": 2.3421, + "step": 27930 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.082192278755471e-05, + "loss": 2.4106, + "step": 27931 + }, + { + "epoch": 1.19, + "grad_norm": 0.455078125, + "learning_rate": 7.081550681582061e-05, + "loss": 2.4547, + "step": 27932 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.080909097540609e-05, + "loss": 2.456, + "step": 27933 + }, + { + "epoch": 1.19, + "grad_norm": 0.4453125, + "learning_rate": 7.080267526633997e-05, + "loss": 2.3933, + "step": 27934 + }, + { + "epoch": 1.19, + "grad_norm": 0.453125, + "learning_rate": 7.079625968865119e-05, + "loss": 2.4445, + "step": 27935 + }, + { + "epoch": 1.19, + "grad_norm": 0.43359375, + "learning_rate": 7.07898442423685e-05, + "loss": 2.5911, + "step": 27936 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.07834289275209e-05, + "loss": 2.5537, + "step": 27937 + }, + { + "epoch": 1.19, + "grad_norm": 0.43359375, + "learning_rate": 7.077701374413717e-05, + "loss": 2.4212, + "step": 27938 + }, + { + "epoch": 1.19, + "grad_norm": 0.482421875, + "learning_rate": 7.077059869224618e-05, + "loss": 2.3483, + "step": 27939 + }, + { + "epoch": 1.19, + "grad_norm": 0.5078125, + "learning_rate": 7.076418377187681e-05, + "loss": 2.4217, + "step": 27940 + }, + { + "epoch": 1.19, + "grad_norm": 0.443359375, + "learning_rate": 7.075776898305795e-05, + "loss": 2.42, + "step": 27941 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.075135432581843e-05, + "loss": 2.3794, + "step": 27942 + }, + { + "epoch": 1.19, + "grad_norm": 0.494140625, + "learning_rate": 7.074493980018717e-05, + "loss": 2.5319, + "step": 27943 + }, + { + "epoch": 1.19, + "grad_norm": 0.51953125, + "learning_rate": 7.073852540619292e-05, + "loss": 2.5183, + "step": 27944 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.073211114386466e-05, + "loss": 2.4666, + "step": 27945 + }, + { + "epoch": 1.19, + "grad_norm": 0.5078125, + "learning_rate": 7.07256970132312e-05, + "loss": 2.5387, + "step": 27946 + }, + { + "epoch": 1.19, + "grad_norm": 0.5, + "learning_rate": 7.071928301432138e-05, + "loss": 2.5565, + "step": 27947 + }, + { + "epoch": 1.19, + "grad_norm": 0.490234375, + "learning_rate": 7.071286914716408e-05, + "loss": 2.5235, + "step": 27948 + }, + { + "epoch": 1.19, + "grad_norm": 0.52734375, + "learning_rate": 7.070645541178819e-05, + "loss": 2.5028, + "step": 27949 + }, + { + "epoch": 1.19, + "grad_norm": 0.515625, + "learning_rate": 7.070004180822254e-05, + "loss": 2.5656, + "step": 27950 + }, + { + "epoch": 1.19, + "grad_norm": 0.51953125, + "learning_rate": 7.069362833649602e-05, + "loss": 2.452, + "step": 27951 + }, + { + "epoch": 1.19, + "grad_norm": 0.478515625, + "learning_rate": 7.068721499663739e-05, + "loss": 2.4532, + "step": 27952 + }, + { + "epoch": 1.19, + "grad_norm": 0.51953125, + "learning_rate": 7.068080178867566e-05, + "loss": 2.5791, + "step": 27953 + }, + { + "epoch": 1.19, + "grad_norm": 0.490234375, + "learning_rate": 7.067438871263959e-05, + "loss": 2.5021, + "step": 27954 + }, + { + "epoch": 1.19, + "grad_norm": 0.490234375, + "learning_rate": 7.066797576855803e-05, + "loss": 2.4428, + "step": 27955 + }, + { + "epoch": 1.19, + "grad_norm": 0.470703125, + "learning_rate": 7.066156295645984e-05, + "loss": 2.4255, + "step": 27956 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.065515027637393e-05, + "loss": 2.5546, + "step": 27957 + }, + { + "epoch": 1.19, + "grad_norm": 0.5234375, + "learning_rate": 7.06487377283291e-05, + "loss": 2.5504, + "step": 27958 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.064232531235426e-05, + "loss": 2.4015, + "step": 27959 + }, + { + "epoch": 1.19, + "grad_norm": 0.474609375, + "learning_rate": 7.063591302847816e-05, + "loss": 2.5475, + "step": 27960 + }, + { + "epoch": 1.19, + "grad_norm": 0.51953125, + "learning_rate": 7.06295008767298e-05, + "loss": 2.3968, + "step": 27961 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.06230888571379e-05, + "loss": 2.5588, + "step": 27962 + }, + { + "epoch": 1.19, + "grad_norm": 0.51953125, + "learning_rate": 7.061667696973139e-05, + "loss": 2.5667, + "step": 27963 + }, + { + "epoch": 1.19, + "grad_norm": 0.453125, + "learning_rate": 7.061026521453907e-05, + "loss": 2.3901, + "step": 27964 + }, + { + "epoch": 1.19, + "grad_norm": 0.51953125, + "learning_rate": 7.060385359158981e-05, + "loss": 2.5894, + "step": 27965 + }, + { + "epoch": 1.19, + "grad_norm": 0.4375, + "learning_rate": 7.059744210091251e-05, + "loss": 2.3979, + "step": 27966 + }, + { + "epoch": 1.19, + "grad_norm": 0.466796875, + "learning_rate": 7.059103074253597e-05, + "loss": 2.4833, + "step": 27967 + }, + { + "epoch": 1.19, + "grad_norm": 0.49609375, + "learning_rate": 7.058461951648899e-05, + "loss": 2.4897, + "step": 27968 + }, + { + "epoch": 1.19, + "grad_norm": 0.4765625, + "learning_rate": 7.057820842280052e-05, + "loss": 2.5694, + "step": 27969 + }, + { + "epoch": 1.19, + "grad_norm": 0.52734375, + "learning_rate": 7.057179746149933e-05, + "loss": 2.4351, + "step": 27970 + }, + { + "epoch": 1.19, + "grad_norm": 0.46484375, + "learning_rate": 7.056538663261431e-05, + "loss": 2.4423, + "step": 27971 + }, + { + "epoch": 1.19, + "grad_norm": 0.455078125, + "learning_rate": 7.055897593617429e-05, + "loss": 2.4787, + "step": 27972 + }, + { + "epoch": 1.19, + "grad_norm": 0.53125, + "learning_rate": 7.055256537220813e-05, + "loss": 2.6767, + "step": 27973 + }, + { + "epoch": 1.19, + "grad_norm": 0.451171875, + "learning_rate": 7.054615494074465e-05, + "loss": 2.5141, + "step": 27974 + }, + { + "epoch": 1.19, + "grad_norm": 0.49609375, + "learning_rate": 7.053974464181274e-05, + "loss": 2.4879, + "step": 27975 + }, + { + "epoch": 1.19, + "grad_norm": 0.478515625, + "learning_rate": 7.053333447544114e-05, + "loss": 2.4598, + "step": 27976 + }, + { + "epoch": 1.19, + "grad_norm": 0.53125, + "learning_rate": 7.052692444165885e-05, + "loss": 2.4577, + "step": 27977 + }, + { + "epoch": 1.19, + "grad_norm": 0.49609375, + "learning_rate": 7.05205145404946e-05, + "loss": 2.4323, + "step": 27978 + }, + { + "epoch": 1.19, + "grad_norm": 0.416015625, + "learning_rate": 7.051410477197726e-05, + "loss": 2.5138, + "step": 27979 + }, + { + "epoch": 1.19, + "grad_norm": 0.515625, + "learning_rate": 7.050769513613565e-05, + "loss": 2.5353, + "step": 27980 + }, + { + "epoch": 1.19, + "grad_norm": 0.46484375, + "learning_rate": 7.050128563299865e-05, + "loss": 2.2508, + "step": 27981 + }, + { + "epoch": 1.19, + "grad_norm": 0.4921875, + "learning_rate": 7.049487626259508e-05, + "loss": 2.3979, + "step": 27982 + }, + { + "epoch": 1.19, + "grad_norm": 0.458984375, + "learning_rate": 7.048846702495382e-05, + "loss": 2.3805, + "step": 27983 + }, + { + "epoch": 1.19, + "grad_norm": 0.427734375, + "learning_rate": 7.04820579201036e-05, + "loss": 2.252, + "step": 27984 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.04756489480734e-05, + "loss": 2.4509, + "step": 27985 + }, + { + "epoch": 1.19, + "grad_norm": 0.423828125, + "learning_rate": 7.046924010889196e-05, + "loss": 2.4462, + "step": 27986 + }, + { + "epoch": 1.19, + "grad_norm": 0.4296875, + "learning_rate": 7.046283140258816e-05, + "loss": 2.4414, + "step": 27987 + }, + { + "epoch": 1.19, + "grad_norm": 0.494140625, + "learning_rate": 7.04564228291908e-05, + "loss": 2.4506, + "step": 27988 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.045001438872875e-05, + "loss": 2.3972, + "step": 27989 + }, + { + "epoch": 1.19, + "grad_norm": 0.458984375, + "learning_rate": 7.044360608123083e-05, + "loss": 2.4897, + "step": 27990 + }, + { + "epoch": 1.19, + "grad_norm": 0.48046875, + "learning_rate": 7.043719790672594e-05, + "loss": 2.5292, + "step": 27991 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.043078986524277e-05, + "loss": 2.3951, + "step": 27992 + }, + { + "epoch": 1.19, + "grad_norm": 0.54296875, + "learning_rate": 7.04243819568103e-05, + "loss": 2.3967, + "step": 27993 + }, + { + "epoch": 1.19, + "grad_norm": 0.4921875, + "learning_rate": 7.041797418145728e-05, + "loss": 2.4221, + "step": 27994 + }, + { + "epoch": 1.19, + "grad_norm": 0.48828125, + "learning_rate": 7.041156653921256e-05, + "loss": 2.3918, + "step": 27995 + }, + { + "epoch": 1.19, + "grad_norm": 0.49609375, + "learning_rate": 7.040515903010496e-05, + "loss": 2.489, + "step": 27996 + }, + { + "epoch": 1.19, + "grad_norm": 0.453125, + "learning_rate": 7.039875165416336e-05, + "loss": 2.3818, + "step": 27997 + }, + { + "epoch": 1.19, + "grad_norm": 0.53515625, + "learning_rate": 7.039234441141655e-05, + "loss": 2.4472, + "step": 27998 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.03859373018934e-05, + "loss": 2.4627, + "step": 27999 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.037953032562263e-05, + "loss": 2.5272, + "step": 28000 + }, + { + "epoch": 1.19, + "grad_norm": 0.486328125, + "learning_rate": 7.037312348263323e-05, + "loss": 2.5199, + "step": 28001 + }, + { + "epoch": 1.19, + "grad_norm": 0.55859375, + "learning_rate": 7.036671677295391e-05, + "loss": 2.4511, + "step": 28002 + }, + { + "epoch": 1.19, + "grad_norm": 0.392578125, + "learning_rate": 7.036031019661354e-05, + "loss": 2.4687, + "step": 28003 + }, + { + "epoch": 1.19, + "grad_norm": 0.48828125, + "learning_rate": 7.035390375364093e-05, + "loss": 2.3041, + "step": 28004 + }, + { + "epoch": 1.19, + "grad_norm": 0.439453125, + "learning_rate": 7.034749744406493e-05, + "loss": 2.4872, + "step": 28005 + }, + { + "epoch": 1.19, + "grad_norm": 0.5078125, + "learning_rate": 7.034109126791436e-05, + "loss": 2.4975, + "step": 28006 + }, + { + "epoch": 1.19, + "grad_norm": 0.474609375, + "learning_rate": 7.033468522521806e-05, + "loss": 2.3984, + "step": 28007 + }, + { + "epoch": 1.19, + "grad_norm": 0.46484375, + "learning_rate": 7.032827931600481e-05, + "loss": 2.4593, + "step": 28008 + }, + { + "epoch": 1.19, + "grad_norm": 0.490234375, + "learning_rate": 7.032187354030344e-05, + "loss": 2.5228, + "step": 28009 + }, + { + "epoch": 1.19, + "grad_norm": 0.578125, + "learning_rate": 7.031546789814282e-05, + "loss": 2.5294, + "step": 28010 + }, + { + "epoch": 1.19, + "grad_norm": 0.4765625, + "learning_rate": 7.030906238955174e-05, + "loss": 2.4029, + "step": 28011 + }, + { + "epoch": 1.19, + "grad_norm": 0.5546875, + "learning_rate": 7.030265701455903e-05, + "loss": 2.5312, + "step": 28012 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.029625177319349e-05, + "loss": 2.5763, + "step": 28013 + }, + { + "epoch": 1.19, + "grad_norm": 0.443359375, + "learning_rate": 7.028984666548397e-05, + "loss": 2.547, + "step": 28014 + }, + { + "epoch": 1.19, + "grad_norm": 0.51953125, + "learning_rate": 7.028344169145933e-05, + "loss": 2.5093, + "step": 28015 + }, + { + "epoch": 1.19, + "grad_norm": 0.498046875, + "learning_rate": 7.02770368511483e-05, + "loss": 2.4776, + "step": 28016 + }, + { + "epoch": 1.19, + "grad_norm": 0.498046875, + "learning_rate": 7.027063214457973e-05, + "loss": 2.4752, + "step": 28017 + }, + { + "epoch": 1.19, + "grad_norm": 0.404296875, + "learning_rate": 7.026422757178245e-05, + "loss": 2.441, + "step": 28018 + }, + { + "epoch": 1.19, + "grad_norm": 0.421875, + "learning_rate": 7.025782313278531e-05, + "loss": 2.4472, + "step": 28019 + }, + { + "epoch": 1.19, + "grad_norm": 0.482421875, + "learning_rate": 7.025141882761708e-05, + "loss": 2.4507, + "step": 28020 + }, + { + "epoch": 1.19, + "grad_norm": 0.447265625, + "learning_rate": 7.024501465630658e-05, + "loss": 2.3904, + "step": 28021 + }, + { + "epoch": 1.19, + "grad_norm": 0.466796875, + "learning_rate": 7.023861061888266e-05, + "loss": 2.5577, + "step": 28022 + }, + { + "epoch": 1.19, + "grad_norm": 0.53515625, + "learning_rate": 7.023220671537414e-05, + "loss": 2.6054, + "step": 28023 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.022580294580978e-05, + "loss": 2.4215, + "step": 28024 + }, + { + "epoch": 1.19, + "grad_norm": 0.484375, + "learning_rate": 7.021939931021841e-05, + "loss": 2.4024, + "step": 28025 + }, + { + "epoch": 1.19, + "grad_norm": 0.46875, + "learning_rate": 7.021299580862887e-05, + "loss": 2.4993, + "step": 28026 + }, + { + "epoch": 1.19, + "grad_norm": 0.453125, + "learning_rate": 7.020659244106997e-05, + "loss": 2.4145, + "step": 28027 + }, + { + "epoch": 1.19, + "grad_norm": 0.443359375, + "learning_rate": 7.020018920757052e-05, + "loss": 2.5074, + "step": 28028 + }, + { + "epoch": 1.19, + "grad_norm": 0.46484375, + "learning_rate": 7.01937861081593e-05, + "loss": 2.4146, + "step": 28029 + }, + { + "epoch": 1.19, + "grad_norm": 0.47265625, + "learning_rate": 7.018738314286517e-05, + "loss": 2.5201, + "step": 28030 + }, + { + "epoch": 1.19, + "grad_norm": 0.41796875, + "learning_rate": 7.018098031171696e-05, + "loss": 2.3335, + "step": 28031 + }, + { + "epoch": 1.19, + "grad_norm": 0.4296875, + "learning_rate": 7.01745776147434e-05, + "loss": 2.4157, + "step": 28032 + }, + { + "epoch": 1.19, + "grad_norm": 0.44921875, + "learning_rate": 7.016817505197333e-05, + "loss": 2.4204, + "step": 28033 + }, + { + "epoch": 1.19, + "grad_norm": 0.46484375, + "learning_rate": 7.016177262343557e-05, + "loss": 2.478, + "step": 28034 + }, + { + "epoch": 1.19, + "grad_norm": 0.498046875, + "learning_rate": 7.015537032915895e-05, + "loss": 2.5658, + "step": 28035 + }, + { + "epoch": 1.19, + "grad_norm": 0.37109375, + "learning_rate": 7.014896816917225e-05, + "loss": 2.425, + "step": 28036 + }, + { + "epoch": 1.19, + "grad_norm": 0.52734375, + "learning_rate": 7.014256614350428e-05, + "loss": 2.4542, + "step": 28037 + }, + { + "epoch": 1.19, + "grad_norm": 0.498046875, + "learning_rate": 7.013616425218386e-05, + "loss": 2.3166, + "step": 28038 + }, + { + "epoch": 1.19, + "grad_norm": 0.4140625, + "learning_rate": 7.012976249523979e-05, + "loss": 2.3132, + "step": 28039 + }, + { + "epoch": 1.19, + "grad_norm": 0.50390625, + "learning_rate": 7.012336087270087e-05, + "loss": 2.5206, + "step": 28040 + }, + { + "epoch": 1.19, + "grad_norm": 0.46484375, + "learning_rate": 7.011695938459589e-05, + "loss": 2.4786, + "step": 28041 + }, + { + "epoch": 1.19, + "grad_norm": 0.5234375, + "learning_rate": 7.011055803095367e-05, + "loss": 2.428, + "step": 28042 + }, + { + "epoch": 1.19, + "grad_norm": 0.51171875, + "learning_rate": 7.010415681180305e-05, + "loss": 2.4413, + "step": 28043 + }, + { + "epoch": 1.2, + "grad_norm": 0.4921875, + "learning_rate": 7.009775572717277e-05, + "loss": 2.457, + "step": 28044 + }, + { + "epoch": 1.2, + "grad_norm": 0.447265625, + "learning_rate": 7.009135477709165e-05, + "loss": 2.4169, + "step": 28045 + }, + { + "epoch": 1.2, + "grad_norm": 0.455078125, + "learning_rate": 7.008495396158852e-05, + "loss": 2.3131, + "step": 28046 + }, + { + "epoch": 1.2, + "grad_norm": 0.46484375, + "learning_rate": 7.007855328069219e-05, + "loss": 2.3935, + "step": 28047 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 7.007215273443141e-05, + "loss": 2.5132, + "step": 28048 + }, + { + "epoch": 1.2, + "grad_norm": 0.5, + "learning_rate": 7.006575232283498e-05, + "loss": 2.4132, + "step": 28049 + }, + { + "epoch": 1.2, + "grad_norm": 0.490234375, + "learning_rate": 7.005935204593177e-05, + "loss": 2.4807, + "step": 28050 + }, + { + "epoch": 1.2, + "grad_norm": 0.5078125, + "learning_rate": 7.005295190375053e-05, + "loss": 2.4714, + "step": 28051 + }, + { + "epoch": 1.2, + "grad_norm": 0.462890625, + "learning_rate": 7.004655189632004e-05, + "loss": 2.4568, + "step": 28052 + }, + { + "epoch": 1.2, + "grad_norm": 0.482421875, + "learning_rate": 7.00401520236691e-05, + "loss": 2.4802, + "step": 28053 + }, + { + "epoch": 1.2, + "grad_norm": 0.4296875, + "learning_rate": 7.003375228582655e-05, + "loss": 2.4268, + "step": 28054 + }, + { + "epoch": 1.2, + "grad_norm": 0.478515625, + "learning_rate": 7.002735268282119e-05, + "loss": 2.5733, + "step": 28055 + }, + { + "epoch": 1.2, + "grad_norm": 0.427734375, + "learning_rate": 7.002095321468177e-05, + "loss": 2.401, + "step": 28056 + }, + { + "epoch": 1.2, + "grad_norm": 0.478515625, + "learning_rate": 7.001455388143708e-05, + "loss": 2.39, + "step": 28057 + }, + { + "epoch": 1.2, + "grad_norm": 0.478515625, + "learning_rate": 7.000815468311596e-05, + "loss": 2.5019, + "step": 28058 + }, + { + "epoch": 1.2, + "grad_norm": 0.419921875, + "learning_rate": 7.000175561974719e-05, + "loss": 2.3341, + "step": 28059 + }, + { + "epoch": 1.2, + "grad_norm": 0.4453125, + "learning_rate": 6.999535669135955e-05, + "loss": 2.4976, + "step": 28060 + }, + { + "epoch": 1.2, + "grad_norm": 0.490234375, + "learning_rate": 6.998895789798182e-05, + "loss": 2.3989, + "step": 28061 + }, + { + "epoch": 1.2, + "grad_norm": 0.5234375, + "learning_rate": 6.998255923964283e-05, + "loss": 2.4851, + "step": 28062 + }, + { + "epoch": 1.2, + "grad_norm": 0.54296875, + "learning_rate": 6.997616071637139e-05, + "loss": 2.3894, + "step": 28063 + }, + { + "epoch": 1.2, + "grad_norm": 0.5078125, + "learning_rate": 6.996976232819621e-05, + "loss": 2.4193, + "step": 28064 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.996336407514612e-05, + "loss": 2.4711, + "step": 28065 + }, + { + "epoch": 1.2, + "grad_norm": 0.494140625, + "learning_rate": 6.995696595724992e-05, + "loss": 2.5818, + "step": 28066 + }, + { + "epoch": 1.2, + "grad_norm": 0.5078125, + "learning_rate": 6.995056797453641e-05, + "loss": 2.4444, + "step": 28067 + }, + { + "epoch": 1.2, + "grad_norm": 0.474609375, + "learning_rate": 6.994417012703436e-05, + "loss": 2.5971, + "step": 28068 + }, + { + "epoch": 1.2, + "grad_norm": 0.42578125, + "learning_rate": 6.993777241477253e-05, + "loss": 2.4192, + "step": 28069 + }, + { + "epoch": 1.2, + "grad_norm": 0.5078125, + "learning_rate": 6.993137483777976e-05, + "loss": 2.4856, + "step": 28070 + }, + { + "epoch": 1.2, + "grad_norm": 0.458984375, + "learning_rate": 6.992497739608483e-05, + "loss": 2.5243, + "step": 28071 + }, + { + "epoch": 1.2, + "grad_norm": 0.41796875, + "learning_rate": 6.991858008971651e-05, + "loss": 2.4244, + "step": 28072 + }, + { + "epoch": 1.2, + "grad_norm": 0.51171875, + "learning_rate": 6.991218291870356e-05, + "loss": 2.5546, + "step": 28073 + }, + { + "epoch": 1.2, + "grad_norm": 0.498046875, + "learning_rate": 6.990578588307479e-05, + "loss": 2.4679, + "step": 28074 + }, + { + "epoch": 1.2, + "grad_norm": 0.419921875, + "learning_rate": 6.989938898285901e-05, + "loss": 2.467, + "step": 28075 + }, + { + "epoch": 1.2, + "grad_norm": 0.451171875, + "learning_rate": 6.989299221808496e-05, + "loss": 2.4115, + "step": 28076 + }, + { + "epoch": 1.2, + "grad_norm": 0.490234375, + "learning_rate": 6.988659558878143e-05, + "loss": 2.5486, + "step": 28077 + }, + { + "epoch": 1.2, + "grad_norm": 0.466796875, + "learning_rate": 6.988019909497725e-05, + "loss": 2.4279, + "step": 28078 + }, + { + "epoch": 1.2, + "grad_norm": 0.458984375, + "learning_rate": 6.987380273670117e-05, + "loss": 2.428, + "step": 28079 + }, + { + "epoch": 1.2, + "grad_norm": 0.478515625, + "learning_rate": 6.986740651398195e-05, + "loss": 2.4678, + "step": 28080 + }, + { + "epoch": 1.2, + "grad_norm": 0.4765625, + "learning_rate": 6.986101042684838e-05, + "loss": 2.4732, + "step": 28081 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.985461447532925e-05, + "loss": 2.4393, + "step": 28082 + }, + { + "epoch": 1.2, + "grad_norm": 0.4609375, + "learning_rate": 6.984821865945335e-05, + "loss": 2.5652, + "step": 28083 + }, + { + "epoch": 1.2, + "grad_norm": 0.51953125, + "learning_rate": 6.984182297924944e-05, + "loss": 2.4627, + "step": 28084 + }, + { + "epoch": 1.2, + "grad_norm": 0.42578125, + "learning_rate": 6.983542743474629e-05, + "loss": 2.3653, + "step": 28085 + }, + { + "epoch": 1.2, + "grad_norm": 0.443359375, + "learning_rate": 6.98290320259727e-05, + "loss": 2.5742, + "step": 28086 + }, + { + "epoch": 1.2, + "grad_norm": 0.50390625, + "learning_rate": 6.982263675295749e-05, + "loss": 2.5274, + "step": 28087 + }, + { + "epoch": 1.2, + "grad_norm": 0.5390625, + "learning_rate": 6.981624161572933e-05, + "loss": 2.5332, + "step": 28088 + }, + { + "epoch": 1.2, + "grad_norm": 0.427734375, + "learning_rate": 6.980984661431706e-05, + "loss": 2.4663, + "step": 28089 + }, + { + "epoch": 1.2, + "grad_norm": 0.48046875, + "learning_rate": 6.980345174874945e-05, + "loss": 2.3874, + "step": 28090 + }, + { + "epoch": 1.2, + "grad_norm": 0.52734375, + "learning_rate": 6.979705701905527e-05, + "loss": 2.5656, + "step": 28091 + }, + { + "epoch": 1.2, + "grad_norm": 0.447265625, + "learning_rate": 6.97906624252633e-05, + "loss": 2.5282, + "step": 28092 + }, + { + "epoch": 1.2, + "grad_norm": 0.474609375, + "learning_rate": 6.978426796740229e-05, + "loss": 2.3471, + "step": 28093 + }, + { + "epoch": 1.2, + "grad_norm": 0.53515625, + "learning_rate": 6.977787364550106e-05, + "loss": 2.5478, + "step": 28094 + }, + { + "epoch": 1.2, + "grad_norm": 0.4765625, + "learning_rate": 6.977147945958837e-05, + "loss": 2.4589, + "step": 28095 + }, + { + "epoch": 1.2, + "grad_norm": 0.51171875, + "learning_rate": 6.976508540969296e-05, + "loss": 2.3489, + "step": 28096 + }, + { + "epoch": 1.2, + "grad_norm": 0.470703125, + "learning_rate": 6.975869149584357e-05, + "loss": 2.3967, + "step": 28097 + }, + { + "epoch": 1.2, + "grad_norm": 0.484375, + "learning_rate": 6.975229771806906e-05, + "loss": 2.4275, + "step": 28098 + }, + { + "epoch": 1.2, + "grad_norm": 0.49609375, + "learning_rate": 6.974590407639815e-05, + "loss": 2.4989, + "step": 28099 + }, + { + "epoch": 1.2, + "grad_norm": 0.462890625, + "learning_rate": 6.973951057085962e-05, + "loss": 2.5158, + "step": 28100 + }, + { + "epoch": 1.2, + "grad_norm": 0.486328125, + "learning_rate": 6.973311720148221e-05, + "loss": 2.4826, + "step": 28101 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.972672396829471e-05, + "loss": 2.394, + "step": 28102 + }, + { + "epoch": 1.2, + "grad_norm": 0.4765625, + "learning_rate": 6.972033087132595e-05, + "loss": 2.3772, + "step": 28103 + }, + { + "epoch": 1.2, + "grad_norm": 0.482421875, + "learning_rate": 6.971393791060458e-05, + "loss": 2.5019, + "step": 28104 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.970754508615943e-05, + "loss": 2.4963, + "step": 28105 + }, + { + "epoch": 1.2, + "grad_norm": 0.478515625, + "learning_rate": 6.970115239801924e-05, + "loss": 2.4772, + "step": 28106 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.969475984621282e-05, + "loss": 2.4144, + "step": 28107 + }, + { + "epoch": 1.2, + "grad_norm": 0.462890625, + "learning_rate": 6.96883674307689e-05, + "loss": 2.415, + "step": 28108 + }, + { + "epoch": 1.2, + "grad_norm": 0.48828125, + "learning_rate": 6.968197515171622e-05, + "loss": 2.5193, + "step": 28109 + }, + { + "epoch": 1.2, + "grad_norm": 0.53125, + "learning_rate": 6.967558300908361e-05, + "loss": 2.612, + "step": 28110 + }, + { + "epoch": 1.2, + "grad_norm": 0.515625, + "learning_rate": 6.966919100289979e-05, + "loss": 2.5018, + "step": 28111 + }, + { + "epoch": 1.2, + "grad_norm": 0.4453125, + "learning_rate": 6.966279913319352e-05, + "loss": 2.4799, + "step": 28112 + }, + { + "epoch": 1.2, + "grad_norm": 0.423828125, + "learning_rate": 6.965640739999355e-05, + "loss": 2.5114, + "step": 28113 + }, + { + "epoch": 1.2, + "grad_norm": 0.42578125, + "learning_rate": 6.965001580332867e-05, + "loss": 2.4714, + "step": 28114 + }, + { + "epoch": 1.2, + "grad_norm": 0.51953125, + "learning_rate": 6.964362434322763e-05, + "loss": 2.4105, + "step": 28115 + }, + { + "epoch": 1.2, + "grad_norm": 0.462890625, + "learning_rate": 6.963723301971919e-05, + "loss": 2.4844, + "step": 28116 + }, + { + "epoch": 1.2, + "grad_norm": 0.515625, + "learning_rate": 6.963084183283208e-05, + "loss": 2.429, + "step": 28117 + }, + { + "epoch": 1.2, + "grad_norm": 0.447265625, + "learning_rate": 6.96244507825951e-05, + "loss": 2.6166, + "step": 28118 + }, + { + "epoch": 1.2, + "grad_norm": 0.455078125, + "learning_rate": 6.961805986903702e-05, + "loss": 2.4252, + "step": 28119 + }, + { + "epoch": 1.2, + "grad_norm": 0.4375, + "learning_rate": 6.961166909218654e-05, + "loss": 2.4348, + "step": 28120 + }, + { + "epoch": 1.2, + "grad_norm": 0.4609375, + "learning_rate": 6.960527845207243e-05, + "loss": 2.4384, + "step": 28121 + }, + { + "epoch": 1.2, + "grad_norm": 0.486328125, + "learning_rate": 6.959888794872347e-05, + "loss": 2.4338, + "step": 28122 + }, + { + "epoch": 1.2, + "grad_norm": 0.431640625, + "learning_rate": 6.959249758216841e-05, + "loss": 2.4961, + "step": 28123 + }, + { + "epoch": 1.2, + "grad_norm": 0.42578125, + "learning_rate": 6.9586107352436e-05, + "loss": 2.4221, + "step": 28124 + }, + { + "epoch": 1.2, + "grad_norm": 0.45703125, + "learning_rate": 6.957971725955496e-05, + "loss": 2.3593, + "step": 28125 + }, + { + "epoch": 1.2, + "grad_norm": 0.490234375, + "learning_rate": 6.95733273035541e-05, + "loss": 2.3669, + "step": 28126 + }, + { + "epoch": 1.2, + "grad_norm": 0.46875, + "learning_rate": 6.956693748446218e-05, + "loss": 2.4399, + "step": 28127 + }, + { + "epoch": 1.2, + "grad_norm": 0.486328125, + "learning_rate": 6.95605478023079e-05, + "loss": 2.4162, + "step": 28128 + }, + { + "epoch": 1.2, + "grad_norm": 0.45703125, + "learning_rate": 6.955415825712002e-05, + "loss": 2.5425, + "step": 28129 + }, + { + "epoch": 1.2, + "grad_norm": 0.50390625, + "learning_rate": 6.954776884892728e-05, + "loss": 2.5329, + "step": 28130 + }, + { + "epoch": 1.2, + "grad_norm": 0.412109375, + "learning_rate": 6.954137957775847e-05, + "loss": 2.3635, + "step": 28131 + }, + { + "epoch": 1.2, + "grad_norm": 0.458984375, + "learning_rate": 6.953499044364232e-05, + "loss": 2.5219, + "step": 28132 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.95286014466076e-05, + "loss": 2.5461, + "step": 28133 + }, + { + "epoch": 1.2, + "grad_norm": 0.45703125, + "learning_rate": 6.952221258668299e-05, + "loss": 2.3364, + "step": 28134 + }, + { + "epoch": 1.2, + "grad_norm": 0.486328125, + "learning_rate": 6.951582386389734e-05, + "loss": 2.4832, + "step": 28135 + }, + { + "epoch": 1.2, + "grad_norm": 0.55078125, + "learning_rate": 6.950943527827931e-05, + "loss": 2.5198, + "step": 28136 + }, + { + "epoch": 1.2, + "grad_norm": 0.49609375, + "learning_rate": 6.950304682985769e-05, + "loss": 2.5524, + "step": 28137 + }, + { + "epoch": 1.2, + "grad_norm": 0.48046875, + "learning_rate": 6.949665851866118e-05, + "loss": 2.4084, + "step": 28138 + }, + { + "epoch": 1.2, + "grad_norm": 0.484375, + "learning_rate": 6.94902703447186e-05, + "loss": 2.6777, + "step": 28139 + }, + { + "epoch": 1.2, + "grad_norm": 0.4453125, + "learning_rate": 6.948388230805863e-05, + "loss": 2.472, + "step": 28140 + }, + { + "epoch": 1.2, + "grad_norm": 0.49609375, + "learning_rate": 6.947749440871007e-05, + "loss": 2.5305, + "step": 28141 + }, + { + "epoch": 1.2, + "grad_norm": 0.4375, + "learning_rate": 6.947110664670157e-05, + "loss": 2.4076, + "step": 28142 + }, + { + "epoch": 1.2, + "grad_norm": 0.46875, + "learning_rate": 6.9464719022062e-05, + "loss": 2.4405, + "step": 28143 + }, + { + "epoch": 1.2, + "grad_norm": 0.51953125, + "learning_rate": 6.945833153482e-05, + "loss": 2.4503, + "step": 28144 + }, + { + "epoch": 1.2, + "grad_norm": 0.498046875, + "learning_rate": 6.945194418500436e-05, + "loss": 2.4257, + "step": 28145 + }, + { + "epoch": 1.2, + "grad_norm": 0.5078125, + "learning_rate": 6.944555697264378e-05, + "loss": 2.359, + "step": 28146 + }, + { + "epoch": 1.2, + "grad_norm": 0.455078125, + "learning_rate": 6.943916989776705e-05, + "loss": 2.5449, + "step": 28147 + }, + { + "epoch": 1.2, + "grad_norm": 0.41796875, + "learning_rate": 6.943278296040289e-05, + "loss": 2.4579, + "step": 28148 + }, + { + "epoch": 1.2, + "grad_norm": 0.48046875, + "learning_rate": 6.942639616058004e-05, + "loss": 2.483, + "step": 28149 + }, + { + "epoch": 1.2, + "grad_norm": 0.44140625, + "learning_rate": 6.942000949832718e-05, + "loss": 2.4145, + "step": 28150 + }, + { + "epoch": 1.2, + "grad_norm": 0.466796875, + "learning_rate": 6.941362297367317e-05, + "loss": 2.4461, + "step": 28151 + }, + { + "epoch": 1.2, + "grad_norm": 0.478515625, + "learning_rate": 6.940723658664666e-05, + "loss": 2.4006, + "step": 28152 + }, + { + "epoch": 1.2, + "grad_norm": 0.443359375, + "learning_rate": 6.94008503372764e-05, + "loss": 2.4687, + "step": 28153 + }, + { + "epoch": 1.2, + "grad_norm": 0.46875, + "learning_rate": 6.939446422559111e-05, + "loss": 2.5884, + "step": 28154 + }, + { + "epoch": 1.2, + "grad_norm": 0.48828125, + "learning_rate": 6.938807825161957e-05, + "loss": 2.6113, + "step": 28155 + }, + { + "epoch": 1.2, + "grad_norm": 0.4296875, + "learning_rate": 6.938169241539049e-05, + "loss": 2.4544, + "step": 28156 + }, + { + "epoch": 1.2, + "grad_norm": 0.498046875, + "learning_rate": 6.937530671693262e-05, + "loss": 2.4722, + "step": 28157 + }, + { + "epoch": 1.2, + "grad_norm": 0.4609375, + "learning_rate": 6.936892115627462e-05, + "loss": 2.3653, + "step": 28158 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.936253573344534e-05, + "loss": 2.4331, + "step": 28159 + }, + { + "epoch": 1.2, + "grad_norm": 0.5, + "learning_rate": 6.935615044847343e-05, + "loss": 2.5423, + "step": 28160 + }, + { + "epoch": 1.2, + "grad_norm": 0.451171875, + "learning_rate": 6.934976530138765e-05, + "loss": 2.5279, + "step": 28161 + }, + { + "epoch": 1.2, + "grad_norm": 0.453125, + "learning_rate": 6.934338029221669e-05, + "loss": 2.4385, + "step": 28162 + }, + { + "epoch": 1.2, + "grad_norm": 0.52734375, + "learning_rate": 6.933699542098934e-05, + "loss": 2.4943, + "step": 28163 + }, + { + "epoch": 1.2, + "grad_norm": 0.46484375, + "learning_rate": 6.933061068773432e-05, + "loss": 2.3748, + "step": 28164 + }, + { + "epoch": 1.2, + "grad_norm": 0.470703125, + "learning_rate": 6.932422609248037e-05, + "loss": 2.3005, + "step": 28165 + }, + { + "epoch": 1.2, + "grad_norm": 0.46875, + "learning_rate": 6.93178416352561e-05, + "loss": 2.5042, + "step": 28166 + }, + { + "epoch": 1.2, + "grad_norm": 0.49609375, + "learning_rate": 6.93114573160904e-05, + "loss": 2.5637, + "step": 28167 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.930507313501192e-05, + "loss": 2.5954, + "step": 28168 + }, + { + "epoch": 1.2, + "grad_norm": 0.482421875, + "learning_rate": 6.92986890920494e-05, + "loss": 2.4888, + "step": 28169 + }, + { + "epoch": 1.2, + "grad_norm": 0.384765625, + "learning_rate": 6.929230518723151e-05, + "loss": 2.3999, + "step": 28170 + }, + { + "epoch": 1.2, + "grad_norm": 0.4453125, + "learning_rate": 6.928592142058706e-05, + "loss": 2.4108, + "step": 28171 + }, + { + "epoch": 1.2, + "grad_norm": 0.419921875, + "learning_rate": 6.927953779214473e-05, + "loss": 2.2835, + "step": 28172 + }, + { + "epoch": 1.2, + "grad_norm": 0.484375, + "learning_rate": 6.927315430193329e-05, + "loss": 2.3312, + "step": 28173 + }, + { + "epoch": 1.2, + "grad_norm": 0.54296875, + "learning_rate": 6.926677094998136e-05, + "loss": 2.5861, + "step": 28174 + }, + { + "epoch": 1.2, + "grad_norm": 0.470703125, + "learning_rate": 6.926038773631778e-05, + "loss": 2.4497, + "step": 28175 + }, + { + "epoch": 1.2, + "grad_norm": 0.48828125, + "learning_rate": 6.92540046609712e-05, + "loss": 2.5078, + "step": 28176 + }, + { + "epoch": 1.2, + "grad_norm": 0.435546875, + "learning_rate": 6.924762172397038e-05, + "loss": 2.4039, + "step": 28177 + }, + { + "epoch": 1.2, + "grad_norm": 0.4921875, + "learning_rate": 6.924123892534398e-05, + "loss": 2.4584, + "step": 28178 + }, + { + "epoch": 1.2, + "grad_norm": 0.4453125, + "learning_rate": 6.92348562651208e-05, + "loss": 2.4169, + "step": 28179 + }, + { + "epoch": 1.2, + "grad_norm": 0.462890625, + "learning_rate": 6.922847374332953e-05, + "loss": 2.4456, + "step": 28180 + }, + { + "epoch": 1.2, + "grad_norm": 0.46875, + "learning_rate": 6.922209135999888e-05, + "loss": 2.4412, + "step": 28181 + }, + { + "epoch": 1.2, + "grad_norm": 0.484375, + "learning_rate": 6.921570911515754e-05, + "loss": 2.5697, + "step": 28182 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.920932700883428e-05, + "loss": 2.5242, + "step": 28183 + }, + { + "epoch": 1.2, + "grad_norm": 0.41796875, + "learning_rate": 6.920294504105783e-05, + "loss": 2.3059, + "step": 28184 + }, + { + "epoch": 1.2, + "grad_norm": 0.49609375, + "learning_rate": 6.919656321185682e-05, + "loss": 2.5019, + "step": 28185 + }, + { + "epoch": 1.2, + "grad_norm": 0.515625, + "learning_rate": 6.919018152126003e-05, + "loss": 2.4485, + "step": 28186 + }, + { + "epoch": 1.2, + "grad_norm": 0.4921875, + "learning_rate": 6.918379996929616e-05, + "loss": 2.5921, + "step": 28187 + }, + { + "epoch": 1.2, + "grad_norm": 0.4765625, + "learning_rate": 6.917741855599395e-05, + "loss": 2.4606, + "step": 28188 + }, + { + "epoch": 1.2, + "grad_norm": 0.51953125, + "learning_rate": 6.91710372813821e-05, + "loss": 2.4281, + "step": 28189 + }, + { + "epoch": 1.2, + "grad_norm": 0.439453125, + "learning_rate": 6.916465614548926e-05, + "loss": 2.3127, + "step": 28190 + }, + { + "epoch": 1.2, + "grad_norm": 0.51953125, + "learning_rate": 6.915827514834426e-05, + "loss": 2.4094, + "step": 28191 + }, + { + "epoch": 1.2, + "grad_norm": 0.4765625, + "learning_rate": 6.915189428997572e-05, + "loss": 2.4296, + "step": 28192 + }, + { + "epoch": 1.2, + "grad_norm": 0.51171875, + "learning_rate": 6.91455135704124e-05, + "loss": 2.4715, + "step": 28193 + }, + { + "epoch": 1.2, + "grad_norm": 0.41796875, + "learning_rate": 6.913913298968295e-05, + "loss": 2.4608, + "step": 28194 + }, + { + "epoch": 1.2, + "grad_norm": 0.46875, + "learning_rate": 6.913275254781615e-05, + "loss": 2.5001, + "step": 28195 + }, + { + "epoch": 1.2, + "grad_norm": 0.51171875, + "learning_rate": 6.91263722448407e-05, + "loss": 2.4392, + "step": 28196 + }, + { + "epoch": 1.2, + "grad_norm": 0.494140625, + "learning_rate": 6.91199920807853e-05, + "loss": 2.6003, + "step": 28197 + }, + { + "epoch": 1.2, + "grad_norm": 0.5546875, + "learning_rate": 6.911361205567857e-05, + "loss": 2.598, + "step": 28198 + }, + { + "epoch": 1.2, + "grad_norm": 0.451171875, + "learning_rate": 6.910723216954939e-05, + "loss": 2.4372, + "step": 28199 + }, + { + "epoch": 1.2, + "grad_norm": 0.4609375, + "learning_rate": 6.910085242242635e-05, + "loss": 2.5358, + "step": 28200 + }, + { + "epoch": 1.2, + "grad_norm": 0.498046875, + "learning_rate": 6.909447281433816e-05, + "loss": 2.3926, + "step": 28201 + }, + { + "epoch": 1.2, + "grad_norm": 0.458984375, + "learning_rate": 6.908809334531355e-05, + "loss": 2.4232, + "step": 28202 + }, + { + "epoch": 1.2, + "grad_norm": 0.40625, + "learning_rate": 6.908171401538124e-05, + "loss": 2.5284, + "step": 28203 + }, + { + "epoch": 1.2, + "grad_norm": 0.494140625, + "learning_rate": 6.907533482456991e-05, + "loss": 2.4705, + "step": 28204 + }, + { + "epoch": 1.2, + "grad_norm": 0.478515625, + "learning_rate": 6.906895577290828e-05, + "loss": 2.5253, + "step": 28205 + }, + { + "epoch": 1.2, + "grad_norm": 0.50390625, + "learning_rate": 6.906257686042501e-05, + "loss": 2.4954, + "step": 28206 + }, + { + "epoch": 1.2, + "grad_norm": 0.4296875, + "learning_rate": 6.905619808714888e-05, + "loss": 2.4689, + "step": 28207 + }, + { + "epoch": 1.2, + "grad_norm": 0.494140625, + "learning_rate": 6.904981945310854e-05, + "loss": 2.5435, + "step": 28208 + }, + { + "epoch": 1.2, + "grad_norm": 0.45703125, + "learning_rate": 6.904344095833271e-05, + "loss": 2.5151, + "step": 28209 + }, + { + "epoch": 1.2, + "grad_norm": 0.416015625, + "learning_rate": 6.903706260285004e-05, + "loss": 2.4392, + "step": 28210 + }, + { + "epoch": 1.2, + "grad_norm": 0.453125, + "learning_rate": 6.903068438668931e-05, + "loss": 2.3594, + "step": 28211 + }, + { + "epoch": 1.2, + "grad_norm": 0.43359375, + "learning_rate": 6.902430630987918e-05, + "loss": 2.3638, + "step": 28212 + }, + { + "epoch": 1.2, + "grad_norm": 0.470703125, + "learning_rate": 6.901792837244835e-05, + "loss": 2.5252, + "step": 28213 + }, + { + "epoch": 1.2, + "grad_norm": 0.484375, + "learning_rate": 6.901155057442548e-05, + "loss": 2.4702, + "step": 28214 + }, + { + "epoch": 1.2, + "grad_norm": 0.50390625, + "learning_rate": 6.900517291583937e-05, + "loss": 2.3833, + "step": 28215 + }, + { + "epoch": 1.2, + "grad_norm": 0.4609375, + "learning_rate": 6.899879539671863e-05, + "loss": 2.4644, + "step": 28216 + }, + { + "epoch": 1.2, + "grad_norm": 0.470703125, + "learning_rate": 6.899241801709198e-05, + "loss": 2.4092, + "step": 28217 + }, + { + "epoch": 1.2, + "grad_norm": 0.453125, + "learning_rate": 6.898604077698809e-05, + "loss": 2.4386, + "step": 28218 + }, + { + "epoch": 1.2, + "grad_norm": 0.458984375, + "learning_rate": 6.897966367643571e-05, + "loss": 2.6271, + "step": 28219 + }, + { + "epoch": 1.2, + "grad_norm": 0.50390625, + "learning_rate": 6.89732867154635e-05, + "loss": 2.5351, + "step": 28220 + }, + { + "epoch": 1.2, + "grad_norm": 0.44140625, + "learning_rate": 6.89669098941002e-05, + "loss": 2.3749, + "step": 28221 + }, + { + "epoch": 1.2, + "grad_norm": 0.48828125, + "learning_rate": 6.896053321237439e-05, + "loss": 2.4619, + "step": 28222 + }, + { + "epoch": 1.2, + "grad_norm": 0.46875, + "learning_rate": 6.89541566703149e-05, + "loss": 2.4431, + "step": 28223 + }, + { + "epoch": 1.2, + "grad_norm": 0.45703125, + "learning_rate": 6.894778026795032e-05, + "loss": 2.5104, + "step": 28224 + }, + { + "epoch": 1.2, + "grad_norm": 0.451171875, + "learning_rate": 6.894140400530941e-05, + "loss": 2.4277, + "step": 28225 + }, + { + "epoch": 1.2, + "grad_norm": 0.42578125, + "learning_rate": 6.89350278824208e-05, + "loss": 2.4338, + "step": 28226 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.892865189931323e-05, + "loss": 2.3829, + "step": 28227 + }, + { + "epoch": 1.2, + "grad_norm": 0.51171875, + "learning_rate": 6.892227605601537e-05, + "loss": 2.3962, + "step": 28228 + }, + { + "epoch": 1.2, + "grad_norm": 0.478515625, + "learning_rate": 6.891590035255592e-05, + "loss": 2.5208, + "step": 28229 + }, + { + "epoch": 1.2, + "grad_norm": 0.435546875, + "learning_rate": 6.890952478896352e-05, + "loss": 2.4558, + "step": 28230 + }, + { + "epoch": 1.2, + "grad_norm": 0.45703125, + "learning_rate": 6.890314936526694e-05, + "loss": 2.4045, + "step": 28231 + }, + { + "epoch": 1.2, + "grad_norm": 0.474609375, + "learning_rate": 6.889677408149483e-05, + "loss": 2.5342, + "step": 28232 + }, + { + "epoch": 1.2, + "grad_norm": 0.486328125, + "learning_rate": 6.889039893767584e-05, + "loss": 2.4256, + "step": 28233 + }, + { + "epoch": 1.2, + "grad_norm": 0.498046875, + "learning_rate": 6.888402393383868e-05, + "loss": 2.3695, + "step": 28234 + }, + { + "epoch": 1.2, + "grad_norm": 0.48046875, + "learning_rate": 6.887764907001204e-05, + "loss": 2.4247, + "step": 28235 + }, + { + "epoch": 1.2, + "grad_norm": 0.44921875, + "learning_rate": 6.887127434622462e-05, + "loss": 2.4558, + "step": 28236 + }, + { + "epoch": 1.2, + "grad_norm": 0.50390625, + "learning_rate": 6.886489976250511e-05, + "loss": 2.4727, + "step": 28237 + }, + { + "epoch": 1.2, + "grad_norm": 0.453125, + "learning_rate": 6.885852531888211e-05, + "loss": 2.4164, + "step": 28238 + }, + { + "epoch": 1.2, + "grad_norm": 0.515625, + "learning_rate": 6.885215101538444e-05, + "loss": 2.4449, + "step": 28239 + }, + { + "epoch": 1.2, + "grad_norm": 0.4375, + "learning_rate": 6.884577685204067e-05, + "loss": 2.4447, + "step": 28240 + }, + { + "epoch": 1.2, + "grad_norm": 0.486328125, + "learning_rate": 6.883940282887953e-05, + "loss": 2.4427, + "step": 28241 + }, + { + "epoch": 1.2, + "grad_norm": 0.49609375, + "learning_rate": 6.883302894592966e-05, + "loss": 2.4708, + "step": 28242 + }, + { + "epoch": 1.2, + "grad_norm": 0.486328125, + "learning_rate": 6.882665520321979e-05, + "loss": 2.4458, + "step": 28243 + }, + { + "epoch": 1.2, + "grad_norm": 0.4375, + "learning_rate": 6.882028160077858e-05, + "loss": 2.5678, + "step": 28244 + }, + { + "epoch": 1.2, + "grad_norm": 0.48046875, + "learning_rate": 6.881390813863473e-05, + "loss": 2.4261, + "step": 28245 + }, + { + "epoch": 1.2, + "grad_norm": 0.458984375, + "learning_rate": 6.880753481681683e-05, + "loss": 2.4692, + "step": 28246 + }, + { + "epoch": 1.2, + "grad_norm": 0.5078125, + "learning_rate": 6.88011616353537e-05, + "loss": 2.3651, + "step": 28247 + }, + { + "epoch": 1.2, + "grad_norm": 0.4765625, + "learning_rate": 6.879478859427391e-05, + "loss": 2.5455, + "step": 28248 + }, + { + "epoch": 1.2, + "grad_norm": 0.4765625, + "learning_rate": 6.878841569360617e-05, + "loss": 2.5495, + "step": 28249 + }, + { + "epoch": 1.2, + "grad_norm": 0.44921875, + "learning_rate": 6.878204293337915e-05, + "loss": 2.5378, + "step": 28250 + }, + { + "epoch": 1.2, + "grad_norm": 0.392578125, + "learning_rate": 6.87756703136215e-05, + "loss": 2.3874, + "step": 28251 + }, + { + "epoch": 1.2, + "grad_norm": 0.431640625, + "learning_rate": 6.876929783436194e-05, + "loss": 2.5218, + "step": 28252 + }, + { + "epoch": 1.2, + "grad_norm": 0.431640625, + "learning_rate": 6.876292549562917e-05, + "loss": 2.4714, + "step": 28253 + }, + { + "epoch": 1.2, + "grad_norm": 0.443359375, + "learning_rate": 6.875655329745179e-05, + "loss": 2.5049, + "step": 28254 + }, + { + "epoch": 1.2, + "grad_norm": 0.4921875, + "learning_rate": 6.875018123985847e-05, + "loss": 2.5177, + "step": 28255 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.874380932287794e-05, + "loss": 2.532, + "step": 28256 + }, + { + "epoch": 1.2, + "grad_norm": 0.46484375, + "learning_rate": 6.873743754653885e-05, + "loss": 2.5769, + "step": 28257 + }, + { + "epoch": 1.2, + "grad_norm": 0.53125, + "learning_rate": 6.873106591086986e-05, + "loss": 2.4787, + "step": 28258 + }, + { + "epoch": 1.2, + "grad_norm": 0.50390625, + "learning_rate": 6.872469441589962e-05, + "loss": 2.5592, + "step": 28259 + }, + { + "epoch": 1.2, + "grad_norm": 0.453125, + "learning_rate": 6.871832306165687e-05, + "loss": 2.4125, + "step": 28260 + }, + { + "epoch": 1.2, + "grad_norm": 0.427734375, + "learning_rate": 6.871195184817023e-05, + "loss": 2.259, + "step": 28261 + }, + { + "epoch": 1.2, + "grad_norm": 0.47265625, + "learning_rate": 6.870558077546835e-05, + "loss": 2.5075, + "step": 28262 + }, + { + "epoch": 1.2, + "grad_norm": 0.494140625, + "learning_rate": 6.869920984357989e-05, + "loss": 2.514, + "step": 28263 + }, + { + "epoch": 1.2, + "grad_norm": 0.466796875, + "learning_rate": 6.869283905253361e-05, + "loss": 2.513, + "step": 28264 + }, + { + "epoch": 1.2, + "grad_norm": 0.48828125, + "learning_rate": 6.86864684023581e-05, + "loss": 2.4522, + "step": 28265 + }, + { + "epoch": 1.2, + "grad_norm": 0.435546875, + "learning_rate": 6.868009789308203e-05, + "loss": 2.4345, + "step": 28266 + }, + { + "epoch": 1.2, + "grad_norm": 0.494140625, + "learning_rate": 6.867372752473404e-05, + "loss": 2.5081, + "step": 28267 + }, + { + "epoch": 1.2, + "grad_norm": 0.53125, + "learning_rate": 6.866735729734286e-05, + "loss": 2.4651, + "step": 28268 + }, + { + "epoch": 1.2, + "grad_norm": 0.55078125, + "learning_rate": 6.866098721093714e-05, + "loss": 2.4483, + "step": 28269 + }, + { + "epoch": 1.2, + "grad_norm": 0.46875, + "learning_rate": 6.865461726554552e-05, + "loss": 2.4866, + "step": 28270 + }, + { + "epoch": 1.2, + "grad_norm": 0.44140625, + "learning_rate": 6.86482474611966e-05, + "loss": 2.4752, + "step": 28271 + }, + { + "epoch": 1.2, + "grad_norm": 0.49609375, + "learning_rate": 6.864187779791919e-05, + "loss": 2.53, + "step": 28272 + }, + { + "epoch": 1.2, + "grad_norm": 0.498046875, + "learning_rate": 6.863550827574183e-05, + "loss": 2.4636, + "step": 28273 + }, + { + "epoch": 1.2, + "grad_norm": 0.451171875, + "learning_rate": 6.862913889469323e-05, + "loss": 2.4561, + "step": 28274 + }, + { + "epoch": 1.2, + "grad_norm": 0.46484375, + "learning_rate": 6.862276965480202e-05, + "loss": 2.5456, + "step": 28275 + }, + { + "epoch": 1.2, + "grad_norm": 0.5078125, + "learning_rate": 6.86164005560969e-05, + "loss": 2.3933, + "step": 28276 + }, + { + "epoch": 1.2, + "grad_norm": 0.455078125, + "learning_rate": 6.861003159860653e-05, + "loss": 2.3468, + "step": 28277 + }, + { + "epoch": 1.21, + "grad_norm": 0.44921875, + "learning_rate": 6.860366278235952e-05, + "loss": 2.3871, + "step": 28278 + }, + { + "epoch": 1.21, + "grad_norm": 0.451171875, + "learning_rate": 6.859729410738452e-05, + "loss": 2.3334, + "step": 28279 + }, + { + "epoch": 1.21, + "grad_norm": 0.431640625, + "learning_rate": 6.859092557371029e-05, + "loss": 2.4262, + "step": 28280 + }, + { + "epoch": 1.21, + "grad_norm": 0.5, + "learning_rate": 6.858455718136538e-05, + "loss": 2.4864, + "step": 28281 + }, + { + "epoch": 1.21, + "grad_norm": 0.41796875, + "learning_rate": 6.857818893037848e-05, + "loss": 2.4558, + "step": 28282 + }, + { + "epoch": 1.21, + "grad_norm": 0.421875, + "learning_rate": 6.857182082077823e-05, + "loss": 2.3425, + "step": 28283 + }, + { + "epoch": 1.21, + "grad_norm": 0.490234375, + "learning_rate": 6.856545285259332e-05, + "loss": 2.4521, + "step": 28284 + }, + { + "epoch": 1.21, + "grad_norm": 0.44140625, + "learning_rate": 6.855908502585243e-05, + "loss": 2.4958, + "step": 28285 + }, + { + "epoch": 1.21, + "grad_norm": 0.46875, + "learning_rate": 6.85527173405841e-05, + "loss": 2.3066, + "step": 28286 + }, + { + "epoch": 1.21, + "grad_norm": 0.427734375, + "learning_rate": 6.854634979681704e-05, + "loss": 2.4904, + "step": 28287 + }, + { + "epoch": 1.21, + "grad_norm": 0.494140625, + "learning_rate": 6.853998239457997e-05, + "loss": 2.5509, + "step": 28288 + }, + { + "epoch": 1.21, + "grad_norm": 0.51953125, + "learning_rate": 6.853361513390145e-05, + "loss": 2.4464, + "step": 28289 + }, + { + "epoch": 1.21, + "grad_norm": 0.5, + "learning_rate": 6.852724801481018e-05, + "loss": 2.5395, + "step": 28290 + }, + { + "epoch": 1.21, + "grad_norm": 0.447265625, + "learning_rate": 6.852088103733475e-05, + "loss": 2.431, + "step": 28291 + }, + { + "epoch": 1.21, + "grad_norm": 0.486328125, + "learning_rate": 6.851451420150387e-05, + "loss": 2.4748, + "step": 28292 + }, + { + "epoch": 1.21, + "grad_norm": 0.44140625, + "learning_rate": 6.850814750734622e-05, + "loss": 2.4593, + "step": 28293 + }, + { + "epoch": 1.21, + "grad_norm": 0.470703125, + "learning_rate": 6.850178095489034e-05, + "loss": 2.4678, + "step": 28294 + }, + { + "epoch": 1.21, + "grad_norm": 0.466796875, + "learning_rate": 6.849541454416493e-05, + "loss": 2.4942, + "step": 28295 + }, + { + "epoch": 1.21, + "grad_norm": 0.4453125, + "learning_rate": 6.848904827519869e-05, + "loss": 2.5397, + "step": 28296 + }, + { + "epoch": 1.21, + "grad_norm": 0.5078125, + "learning_rate": 6.848268214802018e-05, + "loss": 2.4895, + "step": 28297 + }, + { + "epoch": 1.21, + "grad_norm": 0.4609375, + "learning_rate": 6.847631616265809e-05, + "loss": 2.5032, + "step": 28298 + }, + { + "epoch": 1.21, + "grad_norm": 0.474609375, + "learning_rate": 6.846995031914105e-05, + "loss": 2.4721, + "step": 28299 + }, + { + "epoch": 1.21, + "grad_norm": 0.427734375, + "learning_rate": 6.846358461749771e-05, + "loss": 2.5289, + "step": 28300 + }, + { + "epoch": 1.21, + "grad_norm": 0.478515625, + "learning_rate": 6.845721905775676e-05, + "loss": 2.4366, + "step": 28301 + }, + { + "epoch": 1.21, + "grad_norm": 0.498046875, + "learning_rate": 6.845085363994674e-05, + "loss": 2.3262, + "step": 28302 + }, + { + "epoch": 1.21, + "grad_norm": 0.515625, + "learning_rate": 6.844448836409634e-05, + "loss": 2.5604, + "step": 28303 + }, + { + "epoch": 1.21, + "grad_norm": 0.47265625, + "learning_rate": 6.843812323023426e-05, + "loss": 2.4112, + "step": 28304 + }, + { + "epoch": 1.21, + "grad_norm": 0.55078125, + "learning_rate": 6.843175823838906e-05, + "loss": 2.4597, + "step": 28305 + }, + { + "epoch": 1.21, + "grad_norm": 0.51171875, + "learning_rate": 6.842539338858943e-05, + "loss": 2.4855, + "step": 28306 + }, + { + "epoch": 1.21, + "grad_norm": 0.43359375, + "learning_rate": 6.841902868086395e-05, + "loss": 2.3314, + "step": 28307 + }, + { + "epoch": 1.21, + "grad_norm": 0.494140625, + "learning_rate": 6.841266411524134e-05, + "loss": 2.4845, + "step": 28308 + }, + { + "epoch": 1.21, + "grad_norm": 0.443359375, + "learning_rate": 6.840629969175021e-05, + "loss": 2.5862, + "step": 28309 + }, + { + "epoch": 1.21, + "grad_norm": 0.484375, + "learning_rate": 6.839993541041916e-05, + "loss": 2.4436, + "step": 28310 + }, + { + "epoch": 1.21, + "grad_norm": 0.4609375, + "learning_rate": 6.839357127127682e-05, + "loss": 2.3428, + "step": 28311 + }, + { + "epoch": 1.21, + "grad_norm": 0.470703125, + "learning_rate": 6.83872072743519e-05, + "loss": 2.4662, + "step": 28312 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.8380843419673e-05, + "loss": 2.4364, + "step": 28313 + }, + { + "epoch": 1.21, + "grad_norm": 0.43359375, + "learning_rate": 6.837447970726873e-05, + "loss": 2.4603, + "step": 28314 + }, + { + "epoch": 1.21, + "grad_norm": 0.470703125, + "learning_rate": 6.836811613716773e-05, + "loss": 2.4512, + "step": 28315 + }, + { + "epoch": 1.21, + "grad_norm": 0.458984375, + "learning_rate": 6.836175270939868e-05, + "loss": 2.5466, + "step": 28316 + }, + { + "epoch": 1.21, + "grad_norm": 0.484375, + "learning_rate": 6.835538942399017e-05, + "loss": 2.4865, + "step": 28317 + }, + { + "epoch": 1.21, + "grad_norm": 0.45703125, + "learning_rate": 6.834902628097087e-05, + "loss": 2.621, + "step": 28318 + }, + { + "epoch": 1.21, + "grad_norm": 0.50390625, + "learning_rate": 6.834266328036931e-05, + "loss": 2.4423, + "step": 28319 + }, + { + "epoch": 1.21, + "grad_norm": 0.44921875, + "learning_rate": 6.833630042221426e-05, + "loss": 2.4417, + "step": 28320 + }, + { + "epoch": 1.21, + "grad_norm": 0.5078125, + "learning_rate": 6.832993770653426e-05, + "loss": 2.3831, + "step": 28321 + }, + { + "epoch": 1.21, + "grad_norm": 0.427734375, + "learning_rate": 6.832357513335799e-05, + "loss": 2.4166, + "step": 28322 + }, + { + "epoch": 1.21, + "grad_norm": 0.451171875, + "learning_rate": 6.8317212702714e-05, + "loss": 2.5318, + "step": 28323 + }, + { + "epoch": 1.21, + "grad_norm": 0.48828125, + "learning_rate": 6.831085041463104e-05, + "loss": 2.3814, + "step": 28324 + }, + { + "epoch": 1.21, + "grad_norm": 0.5, + "learning_rate": 6.830448826913763e-05, + "loss": 2.372, + "step": 28325 + }, + { + "epoch": 1.21, + "grad_norm": 0.58203125, + "learning_rate": 6.829812626626249e-05, + "loss": 2.5332, + "step": 28326 + }, + { + "epoch": 1.21, + "grad_norm": 0.5, + "learning_rate": 6.82917644060341e-05, + "loss": 2.4353, + "step": 28327 + }, + { + "epoch": 1.21, + "grad_norm": 0.443359375, + "learning_rate": 6.828540268848128e-05, + "loss": 2.5297, + "step": 28328 + }, + { + "epoch": 1.21, + "grad_norm": 0.482421875, + "learning_rate": 6.827904111363252e-05, + "loss": 2.4461, + "step": 28329 + }, + { + "epoch": 1.21, + "grad_norm": 0.5, + "learning_rate": 6.827267968151649e-05, + "loss": 2.5079, + "step": 28330 + }, + { + "epoch": 1.21, + "grad_norm": 0.40625, + "learning_rate": 6.826631839216177e-05, + "loss": 2.4782, + "step": 28331 + }, + { + "epoch": 1.21, + "grad_norm": 0.443359375, + "learning_rate": 6.825995724559705e-05, + "loss": 2.3526, + "step": 28332 + }, + { + "epoch": 1.21, + "grad_norm": 0.44921875, + "learning_rate": 6.825359624185092e-05, + "loss": 2.554, + "step": 28333 + }, + { + "epoch": 1.21, + "grad_norm": 0.474609375, + "learning_rate": 6.824723538095203e-05, + "loss": 2.3955, + "step": 28334 + }, + { + "epoch": 1.21, + "grad_norm": 0.443359375, + "learning_rate": 6.824087466292893e-05, + "loss": 2.39, + "step": 28335 + }, + { + "epoch": 1.21, + "grad_norm": 0.41796875, + "learning_rate": 6.823451408781032e-05, + "loss": 2.3074, + "step": 28336 + }, + { + "epoch": 1.21, + "grad_norm": 0.484375, + "learning_rate": 6.822815365562479e-05, + "loss": 2.5279, + "step": 28337 + }, + { + "epoch": 1.21, + "grad_norm": 0.474609375, + "learning_rate": 6.822179336640093e-05, + "loss": 2.2845, + "step": 28338 + }, + { + "epoch": 1.21, + "grad_norm": 0.5078125, + "learning_rate": 6.821543322016739e-05, + "loss": 2.5826, + "step": 28339 + }, + { + "epoch": 1.21, + "grad_norm": 0.458984375, + "learning_rate": 6.82090732169528e-05, + "loss": 2.4906, + "step": 28340 + }, + { + "epoch": 1.21, + "grad_norm": 0.4453125, + "learning_rate": 6.820271335678575e-05, + "loss": 2.493, + "step": 28341 + }, + { + "epoch": 1.21, + "grad_norm": 0.50390625, + "learning_rate": 6.81963536396949e-05, + "loss": 2.6135, + "step": 28342 + }, + { + "epoch": 1.21, + "grad_norm": 0.51171875, + "learning_rate": 6.818999406570878e-05, + "loss": 2.4536, + "step": 28343 + }, + { + "epoch": 1.21, + "grad_norm": 0.46484375, + "learning_rate": 6.818363463485611e-05, + "loss": 2.4731, + "step": 28344 + }, + { + "epoch": 1.21, + "grad_norm": 0.5390625, + "learning_rate": 6.817727534716545e-05, + "loss": 2.6523, + "step": 28345 + }, + { + "epoch": 1.21, + "grad_norm": 0.4140625, + "learning_rate": 6.817091620266542e-05, + "loss": 2.4608, + "step": 28346 + }, + { + "epoch": 1.21, + "grad_norm": 0.54296875, + "learning_rate": 6.81645572013846e-05, + "loss": 2.5701, + "step": 28347 + }, + { + "epoch": 1.21, + "grad_norm": 0.478515625, + "learning_rate": 6.815819834335166e-05, + "loss": 2.4865, + "step": 28348 + }, + { + "epoch": 1.21, + "grad_norm": 0.5234375, + "learning_rate": 6.81518396285952e-05, + "loss": 2.496, + "step": 28349 + }, + { + "epoch": 1.21, + "grad_norm": 0.55859375, + "learning_rate": 6.814548105714385e-05, + "loss": 2.5992, + "step": 28350 + }, + { + "epoch": 1.21, + "grad_norm": 0.447265625, + "learning_rate": 6.813912262902612e-05, + "loss": 2.3959, + "step": 28351 + }, + { + "epoch": 1.21, + "grad_norm": 0.51171875, + "learning_rate": 6.813276434427076e-05, + "loss": 2.5149, + "step": 28352 + }, + { + "epoch": 1.21, + "grad_norm": 0.490234375, + "learning_rate": 6.812640620290629e-05, + "loss": 2.4291, + "step": 28353 + }, + { + "epoch": 1.21, + "grad_norm": 0.466796875, + "learning_rate": 6.812004820496134e-05, + "loss": 2.3695, + "step": 28354 + }, + { + "epoch": 1.21, + "grad_norm": 0.451171875, + "learning_rate": 6.811369035046451e-05, + "loss": 2.4019, + "step": 28355 + }, + { + "epoch": 1.21, + "grad_norm": 0.421875, + "learning_rate": 6.810733263944445e-05, + "loss": 2.4417, + "step": 28356 + }, + { + "epoch": 1.21, + "grad_norm": 0.5, + "learning_rate": 6.810097507192972e-05, + "loss": 2.5426, + "step": 28357 + }, + { + "epoch": 1.21, + "grad_norm": 0.435546875, + "learning_rate": 6.809461764794898e-05, + "loss": 2.5114, + "step": 28358 + }, + { + "epoch": 1.21, + "grad_norm": 0.4765625, + "learning_rate": 6.808826036753072e-05, + "loss": 2.4852, + "step": 28359 + }, + { + "epoch": 1.21, + "grad_norm": 0.482421875, + "learning_rate": 6.808190323070369e-05, + "loss": 2.509, + "step": 28360 + }, + { + "epoch": 1.21, + "grad_norm": 0.490234375, + "learning_rate": 6.807554623749644e-05, + "loss": 2.4482, + "step": 28361 + }, + { + "epoch": 1.21, + "grad_norm": 0.470703125, + "learning_rate": 6.806918938793753e-05, + "loss": 2.6678, + "step": 28362 + }, + { + "epoch": 1.21, + "grad_norm": 0.40625, + "learning_rate": 6.80628326820556e-05, + "loss": 2.4214, + "step": 28363 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.805647611987925e-05, + "loss": 2.4597, + "step": 28364 + }, + { + "epoch": 1.21, + "grad_norm": 0.48828125, + "learning_rate": 6.80501197014371e-05, + "loss": 2.41, + "step": 28365 + }, + { + "epoch": 1.21, + "grad_norm": 0.5546875, + "learning_rate": 6.804376342675775e-05, + "loss": 2.6343, + "step": 28366 + }, + { + "epoch": 1.21, + "grad_norm": 0.462890625, + "learning_rate": 6.803740729586973e-05, + "loss": 2.4821, + "step": 28367 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.803105130880176e-05, + "loss": 2.4939, + "step": 28368 + }, + { + "epoch": 1.21, + "grad_norm": 0.48828125, + "learning_rate": 6.802469546558234e-05, + "loss": 2.4992, + "step": 28369 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.801833976624011e-05, + "loss": 2.308, + "step": 28370 + }, + { + "epoch": 1.21, + "grad_norm": 0.515625, + "learning_rate": 6.801198421080367e-05, + "loss": 2.5424, + "step": 28371 + }, + { + "epoch": 1.21, + "grad_norm": 0.498046875, + "learning_rate": 6.800562879930159e-05, + "loss": 2.4626, + "step": 28372 + }, + { + "epoch": 1.21, + "grad_norm": 0.52734375, + "learning_rate": 6.799927353176249e-05, + "loss": 2.6427, + "step": 28373 + }, + { + "epoch": 1.21, + "grad_norm": 0.423828125, + "learning_rate": 6.7992918408215e-05, + "loss": 2.3736, + "step": 28374 + }, + { + "epoch": 1.21, + "grad_norm": 0.455078125, + "learning_rate": 6.798656342868768e-05, + "loss": 2.421, + "step": 28375 + }, + { + "epoch": 1.21, + "grad_norm": 0.435546875, + "learning_rate": 6.798020859320908e-05, + "loss": 2.4234, + "step": 28376 + }, + { + "epoch": 1.21, + "grad_norm": 0.462890625, + "learning_rate": 6.797385390180786e-05, + "loss": 2.4515, + "step": 28377 + }, + { + "epoch": 1.21, + "grad_norm": 0.435546875, + "learning_rate": 6.79674993545126e-05, + "loss": 2.3951, + "step": 28378 + }, + { + "epoch": 1.21, + "grad_norm": 0.427734375, + "learning_rate": 6.796114495135189e-05, + "loss": 2.3828, + "step": 28379 + }, + { + "epoch": 1.21, + "grad_norm": 0.462890625, + "learning_rate": 6.795479069235429e-05, + "loss": 2.4706, + "step": 28380 + }, + { + "epoch": 1.21, + "grad_norm": 0.48828125, + "learning_rate": 6.794843657754844e-05, + "loss": 2.4114, + "step": 28381 + }, + { + "epoch": 1.21, + "grad_norm": 0.515625, + "learning_rate": 6.794208260696293e-05, + "loss": 2.5733, + "step": 28382 + }, + { + "epoch": 1.21, + "grad_norm": 0.45703125, + "learning_rate": 6.793572878062632e-05, + "loss": 2.3373, + "step": 28383 + }, + { + "epoch": 1.21, + "grad_norm": 0.50390625, + "learning_rate": 6.792937509856718e-05, + "loss": 2.4131, + "step": 28384 + }, + { + "epoch": 1.21, + "grad_norm": 0.408203125, + "learning_rate": 6.792302156081418e-05, + "loss": 2.364, + "step": 28385 + }, + { + "epoch": 1.21, + "grad_norm": 0.4609375, + "learning_rate": 6.791666816739583e-05, + "loss": 2.4581, + "step": 28386 + }, + { + "epoch": 1.21, + "grad_norm": 0.498046875, + "learning_rate": 6.791031491834077e-05, + "loss": 2.5071, + "step": 28387 + }, + { + "epoch": 1.21, + "grad_norm": 0.56640625, + "learning_rate": 6.790396181367753e-05, + "loss": 2.4949, + "step": 28388 + }, + { + "epoch": 1.21, + "grad_norm": 0.470703125, + "learning_rate": 6.789760885343474e-05, + "loss": 2.4865, + "step": 28389 + }, + { + "epoch": 1.21, + "grad_norm": 0.4296875, + "learning_rate": 6.789125603764103e-05, + "loss": 2.4388, + "step": 28390 + }, + { + "epoch": 1.21, + "grad_norm": 0.4375, + "learning_rate": 6.788490336632489e-05, + "loss": 2.49, + "step": 28391 + }, + { + "epoch": 1.21, + "grad_norm": 0.498046875, + "learning_rate": 6.787855083951494e-05, + "loss": 2.5196, + "step": 28392 + }, + { + "epoch": 1.21, + "grad_norm": 0.455078125, + "learning_rate": 6.78721984572398e-05, + "loss": 2.4345, + "step": 28393 + }, + { + "epoch": 1.21, + "grad_norm": 0.439453125, + "learning_rate": 6.7865846219528e-05, + "loss": 2.3819, + "step": 28394 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.785949412640817e-05, + "loss": 2.6158, + "step": 28395 + }, + { + "epoch": 1.21, + "grad_norm": 0.4296875, + "learning_rate": 6.785314217790884e-05, + "loss": 2.4955, + "step": 28396 + }, + { + "epoch": 1.21, + "grad_norm": 0.5078125, + "learning_rate": 6.784679037405864e-05, + "loss": 2.4678, + "step": 28397 + }, + { + "epoch": 1.21, + "grad_norm": 0.49609375, + "learning_rate": 6.784043871488617e-05, + "loss": 2.4979, + "step": 28398 + }, + { + "epoch": 1.21, + "grad_norm": 0.48046875, + "learning_rate": 6.783408720041993e-05, + "loss": 2.5485, + "step": 28399 + }, + { + "epoch": 1.21, + "grad_norm": 0.423828125, + "learning_rate": 6.782773583068854e-05, + "loss": 2.361, + "step": 28400 + }, + { + "epoch": 1.21, + "grad_norm": 0.482421875, + "learning_rate": 6.782138460572059e-05, + "loss": 2.497, + "step": 28401 + }, + { + "epoch": 1.21, + "grad_norm": 0.5078125, + "learning_rate": 6.781503352554466e-05, + "loss": 2.5934, + "step": 28402 + }, + { + "epoch": 1.21, + "grad_norm": 0.478515625, + "learning_rate": 6.780868259018931e-05, + "loss": 2.5081, + "step": 28403 + }, + { + "epoch": 1.21, + "grad_norm": 0.443359375, + "learning_rate": 6.780233179968309e-05, + "loss": 2.4436, + "step": 28404 + }, + { + "epoch": 1.21, + "grad_norm": 0.4453125, + "learning_rate": 6.779598115405466e-05, + "loss": 2.4079, + "step": 28405 + }, + { + "epoch": 1.21, + "grad_norm": 0.5078125, + "learning_rate": 6.778963065333255e-05, + "loss": 2.3758, + "step": 28406 + }, + { + "epoch": 1.21, + "grad_norm": 0.455078125, + "learning_rate": 6.77832802975453e-05, + "loss": 2.3705, + "step": 28407 + }, + { + "epoch": 1.21, + "grad_norm": 0.5078125, + "learning_rate": 6.77769300867215e-05, + "loss": 2.5949, + "step": 28408 + }, + { + "epoch": 1.21, + "grad_norm": 0.5, + "learning_rate": 6.777058002088977e-05, + "loss": 2.493, + "step": 28409 + }, + { + "epoch": 1.21, + "grad_norm": 0.494140625, + "learning_rate": 6.776423010007864e-05, + "loss": 2.3312, + "step": 28410 + }, + { + "epoch": 1.21, + "grad_norm": 0.49609375, + "learning_rate": 6.775788032431671e-05, + "loss": 2.5899, + "step": 28411 + }, + { + "epoch": 1.21, + "grad_norm": 0.384765625, + "learning_rate": 6.77515306936325e-05, + "loss": 2.309, + "step": 28412 + }, + { + "epoch": 1.21, + "grad_norm": 0.484375, + "learning_rate": 6.774518120805465e-05, + "loss": 2.443, + "step": 28413 + }, + { + "epoch": 1.21, + "grad_norm": 0.4765625, + "learning_rate": 6.77388318676117e-05, + "loss": 2.5313, + "step": 28414 + }, + { + "epoch": 1.21, + "grad_norm": 0.462890625, + "learning_rate": 6.773248267233219e-05, + "loss": 2.4732, + "step": 28415 + }, + { + "epoch": 1.21, + "grad_norm": 0.470703125, + "learning_rate": 6.772613362224472e-05, + "loss": 2.3945, + "step": 28416 + }, + { + "epoch": 1.21, + "grad_norm": 0.478515625, + "learning_rate": 6.771978471737785e-05, + "loss": 2.4769, + "step": 28417 + }, + { + "epoch": 1.21, + "grad_norm": 0.419921875, + "learning_rate": 6.771343595776017e-05, + "loss": 2.5245, + "step": 28418 + }, + { + "epoch": 1.21, + "grad_norm": 0.482421875, + "learning_rate": 6.770708734342021e-05, + "loss": 2.4446, + "step": 28419 + }, + { + "epoch": 1.21, + "grad_norm": 0.412109375, + "learning_rate": 6.770073887438657e-05, + "loss": 2.4844, + "step": 28420 + }, + { + "epoch": 1.21, + "grad_norm": 0.44921875, + "learning_rate": 6.76943905506878e-05, + "loss": 2.6271, + "step": 28421 + }, + { + "epoch": 1.21, + "grad_norm": 0.515625, + "learning_rate": 6.768804237235247e-05, + "loss": 2.6573, + "step": 28422 + }, + { + "epoch": 1.21, + "grad_norm": 0.43359375, + "learning_rate": 6.768169433940913e-05, + "loss": 2.4269, + "step": 28423 + }, + { + "epoch": 1.21, + "grad_norm": 0.44140625, + "learning_rate": 6.767534645188636e-05, + "loss": 2.5287, + "step": 28424 + }, + { + "epoch": 1.21, + "grad_norm": 0.435546875, + "learning_rate": 6.766899870981272e-05, + "loss": 2.4154, + "step": 28425 + }, + { + "epoch": 1.21, + "grad_norm": 0.44140625, + "learning_rate": 6.766265111321677e-05, + "loss": 2.517, + "step": 28426 + }, + { + "epoch": 1.21, + "grad_norm": 0.447265625, + "learning_rate": 6.765630366212706e-05, + "loss": 2.4864, + "step": 28427 + }, + { + "epoch": 1.21, + "grad_norm": 0.455078125, + "learning_rate": 6.764995635657217e-05, + "loss": 2.4116, + "step": 28428 + }, + { + "epoch": 1.21, + "grad_norm": 0.56640625, + "learning_rate": 6.764360919658065e-05, + "loss": 2.4966, + "step": 28429 + }, + { + "epoch": 1.21, + "grad_norm": 0.470703125, + "learning_rate": 6.763726218218109e-05, + "loss": 2.4892, + "step": 28430 + }, + { + "epoch": 1.21, + "grad_norm": 0.5546875, + "learning_rate": 6.7630915313402e-05, + "loss": 2.5802, + "step": 28431 + }, + { + "epoch": 1.21, + "grad_norm": 0.40234375, + "learning_rate": 6.762456859027195e-05, + "loss": 2.2903, + "step": 28432 + }, + { + "epoch": 1.21, + "grad_norm": 0.46875, + "learning_rate": 6.761822201281953e-05, + "loss": 2.4864, + "step": 28433 + }, + { + "epoch": 1.21, + "grad_norm": 0.4765625, + "learning_rate": 6.761187558107328e-05, + "loss": 2.4454, + "step": 28434 + }, + { + "epoch": 1.21, + "grad_norm": 0.423828125, + "learning_rate": 6.760552929506173e-05, + "loss": 2.4278, + "step": 28435 + }, + { + "epoch": 1.21, + "grad_norm": 0.466796875, + "learning_rate": 6.759918315481346e-05, + "loss": 2.4047, + "step": 28436 + }, + { + "epoch": 1.21, + "grad_norm": 0.482421875, + "learning_rate": 6.759283716035704e-05, + "loss": 2.3516, + "step": 28437 + }, + { + "epoch": 1.21, + "grad_norm": 0.451171875, + "learning_rate": 6.758649131172103e-05, + "loss": 2.3158, + "step": 28438 + }, + { + "epoch": 1.21, + "grad_norm": 0.53125, + "learning_rate": 6.758014560893393e-05, + "loss": 2.5413, + "step": 28439 + }, + { + "epoch": 1.21, + "grad_norm": 0.423828125, + "learning_rate": 6.757380005202433e-05, + "loss": 2.4125, + "step": 28440 + }, + { + "epoch": 1.21, + "grad_norm": 0.435546875, + "learning_rate": 6.756745464102079e-05, + "loss": 2.3109, + "step": 28441 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.756110937595184e-05, + "loss": 2.464, + "step": 28442 + }, + { + "epoch": 1.21, + "grad_norm": 0.47265625, + "learning_rate": 6.755476425684607e-05, + "loss": 2.3756, + "step": 28443 + }, + { + "epoch": 1.21, + "grad_norm": 0.482421875, + "learning_rate": 6.754841928373194e-05, + "loss": 2.6066, + "step": 28444 + }, + { + "epoch": 1.21, + "grad_norm": 0.447265625, + "learning_rate": 6.754207445663812e-05, + "loss": 2.3376, + "step": 28445 + }, + { + "epoch": 1.21, + "grad_norm": 0.494140625, + "learning_rate": 6.753572977559311e-05, + "loss": 2.3525, + "step": 28446 + }, + { + "epoch": 1.21, + "grad_norm": 0.478515625, + "learning_rate": 6.752938524062545e-05, + "loss": 2.508, + "step": 28447 + }, + { + "epoch": 1.21, + "grad_norm": 0.53125, + "learning_rate": 6.752304085176365e-05, + "loss": 2.4697, + "step": 28448 + }, + { + "epoch": 1.21, + "grad_norm": 0.40234375, + "learning_rate": 6.751669660903632e-05, + "loss": 2.5073, + "step": 28449 + }, + { + "epoch": 1.21, + "grad_norm": 0.427734375, + "learning_rate": 6.751035251247198e-05, + "loss": 2.4614, + "step": 28450 + }, + { + "epoch": 1.21, + "grad_norm": 0.412109375, + "learning_rate": 6.750400856209919e-05, + "loss": 2.493, + "step": 28451 + }, + { + "epoch": 1.21, + "grad_norm": 0.46875, + "learning_rate": 6.749766475794645e-05, + "loss": 2.4691, + "step": 28452 + }, + { + "epoch": 1.21, + "grad_norm": 0.421875, + "learning_rate": 6.749132110004237e-05, + "loss": 2.5542, + "step": 28453 + }, + { + "epoch": 1.21, + "grad_norm": 0.41796875, + "learning_rate": 6.74849775884155e-05, + "loss": 2.511, + "step": 28454 + }, + { + "epoch": 1.21, + "grad_norm": 0.41796875, + "learning_rate": 6.747863422309431e-05, + "loss": 2.44, + "step": 28455 + }, + { + "epoch": 1.21, + "grad_norm": 0.43359375, + "learning_rate": 6.747229100410736e-05, + "loss": 2.3691, + "step": 28456 + }, + { + "epoch": 1.21, + "grad_norm": 0.484375, + "learning_rate": 6.746594793148325e-05, + "loss": 2.4886, + "step": 28457 + }, + { + "epoch": 1.21, + "grad_norm": 0.4375, + "learning_rate": 6.745960500525045e-05, + "loss": 2.4291, + "step": 28458 + }, + { + "epoch": 1.21, + "grad_norm": 0.50390625, + "learning_rate": 6.745326222543756e-05, + "loss": 2.5575, + "step": 28459 + }, + { + "epoch": 1.21, + "grad_norm": 0.474609375, + "learning_rate": 6.744691959207308e-05, + "loss": 2.4663, + "step": 28460 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.744057710518557e-05, + "loss": 2.523, + "step": 28461 + }, + { + "epoch": 1.21, + "grad_norm": 0.4453125, + "learning_rate": 6.743423476480359e-05, + "loss": 2.4731, + "step": 28462 + }, + { + "epoch": 1.21, + "grad_norm": 0.48828125, + "learning_rate": 6.742789257095564e-05, + "loss": 2.4252, + "step": 28463 + }, + { + "epoch": 1.21, + "grad_norm": 0.466796875, + "learning_rate": 6.742155052367025e-05, + "loss": 2.5482, + "step": 28464 + }, + { + "epoch": 1.21, + "grad_norm": 0.47265625, + "learning_rate": 6.7415208622976e-05, + "loss": 2.4653, + "step": 28465 + }, + { + "epoch": 1.21, + "grad_norm": 0.443359375, + "learning_rate": 6.740886686890137e-05, + "loss": 2.5461, + "step": 28466 + }, + { + "epoch": 1.21, + "grad_norm": 0.4765625, + "learning_rate": 6.740252526147496e-05, + "loss": 2.4868, + "step": 28467 + }, + { + "epoch": 1.21, + "grad_norm": 0.455078125, + "learning_rate": 6.739618380072524e-05, + "loss": 2.4701, + "step": 28468 + }, + { + "epoch": 1.21, + "grad_norm": 0.490234375, + "learning_rate": 6.738984248668082e-05, + "loss": 2.3708, + "step": 28469 + }, + { + "epoch": 1.21, + "grad_norm": 0.498046875, + "learning_rate": 6.738350131937017e-05, + "loss": 2.5162, + "step": 28470 + }, + { + "epoch": 1.21, + "grad_norm": 0.51171875, + "learning_rate": 6.737716029882187e-05, + "loss": 2.5059, + "step": 28471 + }, + { + "epoch": 1.21, + "grad_norm": 0.4765625, + "learning_rate": 6.737081942506437e-05, + "loss": 2.4318, + "step": 28472 + }, + { + "epoch": 1.21, + "grad_norm": 0.50390625, + "learning_rate": 6.736447869812629e-05, + "loss": 2.5076, + "step": 28473 + }, + { + "epoch": 1.21, + "grad_norm": 0.443359375, + "learning_rate": 6.735813811803613e-05, + "loss": 2.3503, + "step": 28474 + }, + { + "epoch": 1.21, + "grad_norm": 0.48046875, + "learning_rate": 6.735179768482242e-05, + "loss": 2.3499, + "step": 28475 + }, + { + "epoch": 1.21, + "grad_norm": 0.451171875, + "learning_rate": 6.734545739851366e-05, + "loss": 2.5337, + "step": 28476 + }, + { + "epoch": 1.21, + "grad_norm": 0.48828125, + "learning_rate": 6.733911725913843e-05, + "loss": 2.6033, + "step": 28477 + }, + { + "epoch": 1.21, + "grad_norm": 0.4375, + "learning_rate": 6.733277726672528e-05, + "loss": 2.4099, + "step": 28478 + }, + { + "epoch": 1.21, + "grad_norm": 0.423828125, + "learning_rate": 6.732643742130264e-05, + "loss": 2.4366, + "step": 28479 + }, + { + "epoch": 1.21, + "grad_norm": 0.42578125, + "learning_rate": 6.73200977228991e-05, + "loss": 2.347, + "step": 28480 + }, + { + "epoch": 1.21, + "grad_norm": 0.482421875, + "learning_rate": 6.731375817154319e-05, + "loss": 2.5234, + "step": 28481 + }, + { + "epoch": 1.21, + "grad_norm": 0.458984375, + "learning_rate": 6.73074187672634e-05, + "loss": 2.3764, + "step": 28482 + }, + { + "epoch": 1.21, + "grad_norm": 0.4921875, + "learning_rate": 6.73010795100883e-05, + "loss": 2.4665, + "step": 28483 + }, + { + "epoch": 1.21, + "grad_norm": 0.515625, + "learning_rate": 6.729474040004638e-05, + "loss": 2.3933, + "step": 28484 + }, + { + "epoch": 1.21, + "grad_norm": 0.46484375, + "learning_rate": 6.728840143716617e-05, + "loss": 2.6028, + "step": 28485 + }, + { + "epoch": 1.21, + "grad_norm": 0.51953125, + "learning_rate": 6.728206262147624e-05, + "loss": 2.4485, + "step": 28486 + }, + { + "epoch": 1.21, + "grad_norm": 0.478515625, + "learning_rate": 6.727572395300504e-05, + "loss": 2.5275, + "step": 28487 + }, + { + "epoch": 1.21, + "grad_norm": 0.427734375, + "learning_rate": 6.72693854317811e-05, + "loss": 2.3973, + "step": 28488 + }, + { + "epoch": 1.21, + "grad_norm": 0.5234375, + "learning_rate": 6.7263047057833e-05, + "loss": 2.4098, + "step": 28489 + }, + { + "epoch": 1.21, + "grad_norm": 0.462890625, + "learning_rate": 6.725670883118921e-05, + "loss": 2.6963, + "step": 28490 + }, + { + "epoch": 1.21, + "grad_norm": 0.4453125, + "learning_rate": 6.725037075187826e-05, + "loss": 2.4369, + "step": 28491 + }, + { + "epoch": 1.21, + "grad_norm": 0.48046875, + "learning_rate": 6.724403281992871e-05, + "loss": 2.4068, + "step": 28492 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.723769503536896e-05, + "loss": 2.3771, + "step": 28493 + }, + { + "epoch": 1.21, + "grad_norm": 0.4921875, + "learning_rate": 6.723135739822769e-05, + "loss": 2.4806, + "step": 28494 + }, + { + "epoch": 1.21, + "grad_norm": 0.427734375, + "learning_rate": 6.722501990853331e-05, + "loss": 2.4076, + "step": 28495 + }, + { + "epoch": 1.21, + "grad_norm": 0.490234375, + "learning_rate": 6.721868256631436e-05, + "loss": 2.5147, + "step": 28496 + }, + { + "epoch": 1.21, + "grad_norm": 0.453125, + "learning_rate": 6.721234537159934e-05, + "loss": 2.3416, + "step": 28497 + }, + { + "epoch": 1.21, + "grad_norm": 0.458984375, + "learning_rate": 6.720600832441679e-05, + "loss": 2.3984, + "step": 28498 + }, + { + "epoch": 1.21, + "grad_norm": 0.4453125, + "learning_rate": 6.719967142479524e-05, + "loss": 2.3939, + "step": 28499 + }, + { + "epoch": 1.21, + "grad_norm": 0.515625, + "learning_rate": 6.719333467276318e-05, + "loss": 2.4759, + "step": 28500 + }, + { + "epoch": 1.21, + "grad_norm": 0.423828125, + "learning_rate": 6.718699806834908e-05, + "loss": 2.3694, + "step": 28501 + }, + { + "epoch": 1.21, + "grad_norm": 0.51171875, + "learning_rate": 6.718066161158155e-05, + "loss": 2.5559, + "step": 28502 + }, + { + "epoch": 1.21, + "grad_norm": 0.4921875, + "learning_rate": 6.717432530248904e-05, + "loss": 2.4849, + "step": 28503 + }, + { + "epoch": 1.21, + "grad_norm": 0.44921875, + "learning_rate": 6.716798914110006e-05, + "loss": 2.3937, + "step": 28504 + }, + { + "epoch": 1.21, + "grad_norm": 0.4921875, + "learning_rate": 6.716165312744311e-05, + "loss": 2.5758, + "step": 28505 + }, + { + "epoch": 1.21, + "grad_norm": 0.462890625, + "learning_rate": 6.715531726154674e-05, + "loss": 2.4767, + "step": 28506 + }, + { + "epoch": 1.21, + "grad_norm": 0.46484375, + "learning_rate": 6.714898154343944e-05, + "loss": 2.336, + "step": 28507 + }, + { + "epoch": 1.21, + "grad_norm": 0.48828125, + "learning_rate": 6.714264597314973e-05, + "loss": 2.4903, + "step": 28508 + }, + { + "epoch": 1.21, + "grad_norm": 0.4921875, + "learning_rate": 6.713631055070606e-05, + "loss": 2.5203, + "step": 28509 + }, + { + "epoch": 1.21, + "grad_norm": 0.50390625, + "learning_rate": 6.712997527613703e-05, + "loss": 2.5104, + "step": 28510 + }, + { + "epoch": 1.21, + "grad_norm": 0.50390625, + "learning_rate": 6.712364014947108e-05, + "loss": 2.4543, + "step": 28511 + }, + { + "epoch": 1.21, + "grad_norm": 0.55078125, + "learning_rate": 6.711730517073673e-05, + "loss": 2.5389, + "step": 28512 + }, + { + "epoch": 1.22, + "grad_norm": 0.466796875, + "learning_rate": 6.711097033996247e-05, + "loss": 2.6176, + "step": 28513 + }, + { + "epoch": 1.22, + "grad_norm": 0.453125, + "learning_rate": 6.710463565717687e-05, + "loss": 2.4028, + "step": 28514 + }, + { + "epoch": 1.22, + "grad_norm": 0.5234375, + "learning_rate": 6.709830112240835e-05, + "loss": 2.5497, + "step": 28515 + }, + { + "epoch": 1.22, + "grad_norm": 0.470703125, + "learning_rate": 6.70919667356855e-05, + "loss": 2.5137, + "step": 28516 + }, + { + "epoch": 1.22, + "grad_norm": 0.458984375, + "learning_rate": 6.70856324970367e-05, + "loss": 2.4886, + "step": 28517 + }, + { + "epoch": 1.22, + "grad_norm": 0.482421875, + "learning_rate": 6.707929840649058e-05, + "loss": 2.4012, + "step": 28518 + }, + { + "epoch": 1.22, + "grad_norm": 0.46484375, + "learning_rate": 6.707296446407558e-05, + "loss": 2.5158, + "step": 28519 + }, + { + "epoch": 1.22, + "grad_norm": 0.54296875, + "learning_rate": 6.70666306698202e-05, + "loss": 2.3464, + "step": 28520 + }, + { + "epoch": 1.22, + "grad_norm": 0.421875, + "learning_rate": 6.706029702375292e-05, + "loss": 2.453, + "step": 28521 + }, + { + "epoch": 1.22, + "grad_norm": 0.470703125, + "learning_rate": 6.70539635259023e-05, + "loss": 2.5138, + "step": 28522 + }, + { + "epoch": 1.22, + "grad_norm": 0.462890625, + "learning_rate": 6.704763017629679e-05, + "loss": 2.5179, + "step": 28523 + }, + { + "epoch": 1.22, + "grad_norm": 0.447265625, + "learning_rate": 6.704129697496493e-05, + "loss": 2.3539, + "step": 28524 + }, + { + "epoch": 1.22, + "grad_norm": 0.447265625, + "learning_rate": 6.703496392193512e-05, + "loss": 2.5016, + "step": 28525 + }, + { + "epoch": 1.22, + "grad_norm": 0.478515625, + "learning_rate": 6.7028631017236e-05, + "loss": 2.3681, + "step": 28526 + }, + { + "epoch": 1.22, + "grad_norm": 0.48046875, + "learning_rate": 6.702229826089596e-05, + "loss": 2.4899, + "step": 28527 + }, + { + "epoch": 1.22, + "grad_norm": 0.435546875, + "learning_rate": 6.701596565294352e-05, + "loss": 2.4882, + "step": 28528 + }, + { + "epoch": 1.22, + "grad_norm": 0.482421875, + "learning_rate": 6.700963319340717e-05, + "loss": 2.4571, + "step": 28529 + }, + { + "epoch": 1.22, + "grad_norm": 0.4921875, + "learning_rate": 6.700330088231544e-05, + "loss": 2.4793, + "step": 28530 + }, + { + "epoch": 1.22, + "grad_norm": 0.51953125, + "learning_rate": 6.699696871969678e-05, + "loss": 2.4721, + "step": 28531 + }, + { + "epoch": 1.22, + "grad_norm": 0.52734375, + "learning_rate": 6.699063670557972e-05, + "loss": 2.4907, + "step": 28532 + }, + { + "epoch": 1.22, + "grad_norm": 0.482421875, + "learning_rate": 6.698430483999268e-05, + "loss": 2.4521, + "step": 28533 + }, + { + "epoch": 1.22, + "grad_norm": 0.470703125, + "learning_rate": 6.697797312296427e-05, + "loss": 2.5108, + "step": 28534 + }, + { + "epoch": 1.22, + "grad_norm": 0.4375, + "learning_rate": 6.697164155452288e-05, + "loss": 2.4369, + "step": 28535 + }, + { + "epoch": 1.22, + "grad_norm": 0.43359375, + "learning_rate": 6.696531013469704e-05, + "loss": 2.5729, + "step": 28536 + }, + { + "epoch": 1.22, + "grad_norm": 0.46484375, + "learning_rate": 6.69589788635152e-05, + "loss": 2.4818, + "step": 28537 + }, + { + "epoch": 1.22, + "grad_norm": 0.47265625, + "learning_rate": 6.695264774100591e-05, + "loss": 2.4495, + "step": 28538 + }, + { + "epoch": 1.22, + "grad_norm": 0.51171875, + "learning_rate": 6.694631676719763e-05, + "loss": 2.4329, + "step": 28539 + }, + { + "epoch": 1.22, + "grad_norm": 0.51953125, + "learning_rate": 6.693998594211886e-05, + "loss": 2.4132, + "step": 28540 + }, + { + "epoch": 1.22, + "grad_norm": 0.46484375, + "learning_rate": 6.6933655265798e-05, + "loss": 2.4491, + "step": 28541 + }, + { + "epoch": 1.22, + "grad_norm": 0.490234375, + "learning_rate": 6.692732473826368e-05, + "loss": 2.3977, + "step": 28542 + }, + { + "epoch": 1.22, + "grad_norm": 0.427734375, + "learning_rate": 6.692099435954429e-05, + "loss": 2.3705, + "step": 28543 + }, + { + "epoch": 1.22, + "grad_norm": 0.5546875, + "learning_rate": 6.691466412966834e-05, + "loss": 2.4098, + "step": 28544 + }, + { + "epoch": 1.22, + "grad_norm": 0.43359375, + "learning_rate": 6.690833404866429e-05, + "loss": 2.4539, + "step": 28545 + }, + { + "epoch": 1.22, + "grad_norm": 0.44140625, + "learning_rate": 6.690200411656067e-05, + "loss": 2.4744, + "step": 28546 + }, + { + "epoch": 1.22, + "grad_norm": 0.515625, + "learning_rate": 6.689567433338591e-05, + "loss": 2.436, + "step": 28547 + }, + { + "epoch": 1.22, + "grad_norm": 0.5078125, + "learning_rate": 6.688934469916857e-05, + "loss": 2.6349, + "step": 28548 + }, + { + "epoch": 1.22, + "grad_norm": 0.53125, + "learning_rate": 6.688301521393698e-05, + "loss": 2.4848, + "step": 28549 + }, + { + "epoch": 1.22, + "grad_norm": 0.46875, + "learning_rate": 6.687668587771981e-05, + "loss": 2.4846, + "step": 28550 + }, + { + "epoch": 1.22, + "grad_norm": 0.51171875, + "learning_rate": 6.687035669054542e-05, + "loss": 2.4849, + "step": 28551 + }, + { + "epoch": 1.22, + "grad_norm": 0.466796875, + "learning_rate": 6.686402765244232e-05, + "loss": 2.43, + "step": 28552 + }, + { + "epoch": 1.22, + "grad_norm": 0.53125, + "learning_rate": 6.685769876343896e-05, + "loss": 2.474, + "step": 28553 + }, + { + "epoch": 1.22, + "grad_norm": 0.47265625, + "learning_rate": 6.685137002356386e-05, + "loss": 2.5243, + "step": 28554 + }, + { + "epoch": 1.22, + "grad_norm": 0.48828125, + "learning_rate": 6.684504143284549e-05, + "loss": 2.5432, + "step": 28555 + }, + { + "epoch": 1.22, + "grad_norm": 0.421875, + "learning_rate": 6.683871299131233e-05, + "loss": 2.2734, + "step": 28556 + }, + { + "epoch": 1.22, + "grad_norm": 0.482421875, + "learning_rate": 6.683238469899278e-05, + "loss": 2.5049, + "step": 28557 + }, + { + "epoch": 1.22, + "grad_norm": 0.46875, + "learning_rate": 6.682605655591546e-05, + "loss": 2.5086, + "step": 28558 + }, + { + "epoch": 1.22, + "grad_norm": 0.486328125, + "learning_rate": 6.681972856210871e-05, + "loss": 2.4988, + "step": 28559 + }, + { + "epoch": 1.22, + "grad_norm": 0.435546875, + "learning_rate": 6.681340071760107e-05, + "loss": 2.3879, + "step": 28560 + }, + { + "epoch": 1.22, + "grad_norm": 0.49609375, + "learning_rate": 6.680707302242097e-05, + "loss": 2.3836, + "step": 28561 + }, + { + "epoch": 1.22, + "grad_norm": 0.4140625, + "learning_rate": 6.680074547659695e-05, + "loss": 2.5763, + "step": 28562 + }, + { + "epoch": 1.22, + "grad_norm": 0.4296875, + "learning_rate": 6.679441808015743e-05, + "loss": 2.4867, + "step": 28563 + }, + { + "epoch": 1.22, + "grad_norm": 0.466796875, + "learning_rate": 6.678809083313092e-05, + "loss": 2.4336, + "step": 28564 + }, + { + "epoch": 1.22, + "grad_norm": 0.423828125, + "learning_rate": 6.678176373554579e-05, + "loss": 2.342, + "step": 28565 + }, + { + "epoch": 1.22, + "grad_norm": 0.46875, + "learning_rate": 6.677543678743066e-05, + "loss": 2.4582, + "step": 28566 + }, + { + "epoch": 1.22, + "grad_norm": 0.466796875, + "learning_rate": 6.67691099888139e-05, + "loss": 2.5187, + "step": 28567 + }, + { + "epoch": 1.22, + "grad_norm": 0.5078125, + "learning_rate": 6.676278333972397e-05, + "loss": 2.4517, + "step": 28568 + }, + { + "epoch": 1.22, + "grad_norm": 0.43359375, + "learning_rate": 6.675645684018939e-05, + "loss": 2.4787, + "step": 28569 + }, + { + "epoch": 1.22, + "grad_norm": 0.4375, + "learning_rate": 6.675013049023859e-05, + "loss": 2.2654, + "step": 28570 + }, + { + "epoch": 1.22, + "grad_norm": 0.44921875, + "learning_rate": 6.674380428990008e-05, + "loss": 2.3895, + "step": 28571 + }, + { + "epoch": 1.22, + "grad_norm": 0.44140625, + "learning_rate": 6.673747823920232e-05, + "loss": 2.5235, + "step": 28572 + }, + { + "epoch": 1.22, + "grad_norm": 0.423828125, + "learning_rate": 6.673115233817366e-05, + "loss": 2.5526, + "step": 28573 + }, + { + "epoch": 1.22, + "grad_norm": 0.451171875, + "learning_rate": 6.672482658684273e-05, + "loss": 2.4183, + "step": 28574 + }, + { + "epoch": 1.22, + "grad_norm": 0.4921875, + "learning_rate": 6.671850098523791e-05, + "loss": 2.3536, + "step": 28575 + }, + { + "epoch": 1.22, + "grad_norm": 0.515625, + "learning_rate": 6.671217553338766e-05, + "loss": 2.5239, + "step": 28576 + }, + { + "epoch": 1.22, + "grad_norm": 0.482421875, + "learning_rate": 6.670585023132044e-05, + "loss": 2.4238, + "step": 28577 + }, + { + "epoch": 1.22, + "grad_norm": 0.451171875, + "learning_rate": 6.669952507906476e-05, + "loss": 2.3257, + "step": 28578 + }, + { + "epoch": 1.22, + "grad_norm": 0.546875, + "learning_rate": 6.669320007664904e-05, + "loss": 2.5094, + "step": 28579 + }, + { + "epoch": 1.22, + "grad_norm": 0.45703125, + "learning_rate": 6.668687522410175e-05, + "loss": 2.4823, + "step": 28580 + }, + { + "epoch": 1.22, + "grad_norm": 0.423828125, + "learning_rate": 6.668055052145131e-05, + "loss": 2.397, + "step": 28581 + }, + { + "epoch": 1.22, + "grad_norm": 0.478515625, + "learning_rate": 6.667422596872625e-05, + "loss": 2.4386, + "step": 28582 + }, + { + "epoch": 1.22, + "grad_norm": 0.458984375, + "learning_rate": 6.666790156595499e-05, + "loss": 2.4282, + "step": 28583 + }, + { + "epoch": 1.22, + "grad_norm": 0.45703125, + "learning_rate": 6.666157731316599e-05, + "loss": 2.5296, + "step": 28584 + }, + { + "epoch": 1.22, + "grad_norm": 0.423828125, + "learning_rate": 6.66552532103877e-05, + "loss": 2.4283, + "step": 28585 + }, + { + "epoch": 1.22, + "grad_norm": 0.4296875, + "learning_rate": 6.664892925764859e-05, + "loss": 2.3578, + "step": 28586 + }, + { + "epoch": 1.22, + "grad_norm": 0.51953125, + "learning_rate": 6.66426054549771e-05, + "loss": 2.5126, + "step": 28587 + }, + { + "epoch": 1.22, + "grad_norm": 0.4921875, + "learning_rate": 6.663628180240173e-05, + "loss": 2.5413, + "step": 28588 + }, + { + "epoch": 1.22, + "grad_norm": 0.498046875, + "learning_rate": 6.662995829995084e-05, + "loss": 2.5007, + "step": 28589 + }, + { + "epoch": 1.22, + "grad_norm": 0.462890625, + "learning_rate": 6.6623634947653e-05, + "loss": 2.3197, + "step": 28590 + }, + { + "epoch": 1.22, + "grad_norm": 0.435546875, + "learning_rate": 6.661731174553659e-05, + "loss": 2.4182, + "step": 28591 + }, + { + "epoch": 1.22, + "grad_norm": 0.46484375, + "learning_rate": 6.661098869363006e-05, + "loss": 2.5482, + "step": 28592 + }, + { + "epoch": 1.22, + "grad_norm": 0.50390625, + "learning_rate": 6.660466579196187e-05, + "loss": 2.5174, + "step": 28593 + }, + { + "epoch": 1.22, + "grad_norm": 0.443359375, + "learning_rate": 6.659834304056049e-05, + "loss": 2.4514, + "step": 28594 + }, + { + "epoch": 1.22, + "grad_norm": 0.5703125, + "learning_rate": 6.659202043945438e-05, + "loss": 2.547, + "step": 28595 + }, + { + "epoch": 1.22, + "grad_norm": 0.46484375, + "learning_rate": 6.658569798867197e-05, + "loss": 2.5135, + "step": 28596 + }, + { + "epoch": 1.22, + "grad_norm": 0.50390625, + "learning_rate": 6.657937568824167e-05, + "loss": 2.4422, + "step": 28597 + }, + { + "epoch": 1.22, + "grad_norm": 0.4375, + "learning_rate": 6.6573053538192e-05, + "loss": 2.5, + "step": 28598 + }, + { + "epoch": 1.22, + "grad_norm": 0.482421875, + "learning_rate": 6.656673153855136e-05, + "loss": 2.5839, + "step": 28599 + }, + { + "epoch": 1.22, + "grad_norm": 0.45703125, + "learning_rate": 6.656040968934821e-05, + "loss": 2.3961, + "step": 28600 + }, + { + "epoch": 1.22, + "grad_norm": 0.490234375, + "learning_rate": 6.655408799061099e-05, + "loss": 2.4309, + "step": 28601 + }, + { + "epoch": 1.22, + "grad_norm": 0.474609375, + "learning_rate": 6.654776644236815e-05, + "loss": 2.5143, + "step": 28602 + }, + { + "epoch": 1.22, + "grad_norm": 0.546875, + "learning_rate": 6.654144504464815e-05, + "loss": 2.4812, + "step": 28603 + }, + { + "epoch": 1.22, + "grad_norm": 0.46875, + "learning_rate": 6.653512379747944e-05, + "loss": 2.4608, + "step": 28604 + }, + { + "epoch": 1.22, + "grad_norm": 0.4921875, + "learning_rate": 6.652880270089037e-05, + "loss": 2.4776, + "step": 28605 + }, + { + "epoch": 1.22, + "grad_norm": 0.427734375, + "learning_rate": 6.652248175490952e-05, + "loss": 2.4346, + "step": 28606 + }, + { + "epoch": 1.22, + "grad_norm": 0.47265625, + "learning_rate": 6.651616095956525e-05, + "loss": 2.4156, + "step": 28607 + }, + { + "epoch": 1.22, + "grad_norm": 0.4453125, + "learning_rate": 6.650984031488602e-05, + "loss": 2.4411, + "step": 28608 + }, + { + "epoch": 1.22, + "grad_norm": 0.427734375, + "learning_rate": 6.650351982090024e-05, + "loss": 2.4721, + "step": 28609 + }, + { + "epoch": 1.22, + "grad_norm": 0.466796875, + "learning_rate": 6.649719947763642e-05, + "loss": 2.5355, + "step": 28610 + }, + { + "epoch": 1.22, + "grad_norm": 0.455078125, + "learning_rate": 6.649087928512294e-05, + "loss": 2.516, + "step": 28611 + }, + { + "epoch": 1.22, + "grad_norm": 0.4296875, + "learning_rate": 6.648455924338827e-05, + "loss": 2.3547, + "step": 28612 + }, + { + "epoch": 1.22, + "grad_norm": 0.427734375, + "learning_rate": 6.647823935246085e-05, + "loss": 2.3704, + "step": 28613 + }, + { + "epoch": 1.22, + "grad_norm": 0.431640625, + "learning_rate": 6.647191961236905e-05, + "loss": 2.4149, + "step": 28614 + }, + { + "epoch": 1.22, + "grad_norm": 0.4765625, + "learning_rate": 6.64656000231414e-05, + "loss": 2.413, + "step": 28615 + }, + { + "epoch": 1.22, + "grad_norm": 0.447265625, + "learning_rate": 6.645928058480627e-05, + "loss": 2.546, + "step": 28616 + }, + { + "epoch": 1.22, + "grad_norm": 0.4375, + "learning_rate": 6.645296129739213e-05, + "loss": 2.6165, + "step": 28617 + }, + { + "epoch": 1.22, + "grad_norm": 0.44140625, + "learning_rate": 6.64466421609274e-05, + "loss": 2.4152, + "step": 28618 + }, + { + "epoch": 1.22, + "grad_norm": 0.50390625, + "learning_rate": 6.64403231754405e-05, + "loss": 2.4739, + "step": 28619 + }, + { + "epoch": 1.22, + "grad_norm": 0.45703125, + "learning_rate": 6.643400434095995e-05, + "loss": 2.534, + "step": 28620 + }, + { + "epoch": 1.22, + "grad_norm": 0.435546875, + "learning_rate": 6.642768565751408e-05, + "loss": 2.2967, + "step": 28621 + }, + { + "epoch": 1.22, + "grad_norm": 0.53125, + "learning_rate": 6.64213671251313e-05, + "loss": 2.4442, + "step": 28622 + }, + { + "epoch": 1.22, + "grad_norm": 0.44140625, + "learning_rate": 6.641504874384015e-05, + "loss": 2.4805, + "step": 28623 + }, + { + "epoch": 1.22, + "grad_norm": 0.439453125, + "learning_rate": 6.6408730513669e-05, + "loss": 2.427, + "step": 28624 + }, + { + "epoch": 1.22, + "grad_norm": 0.4453125, + "learning_rate": 6.640241243464628e-05, + "loss": 2.4302, + "step": 28625 + }, + { + "epoch": 1.22, + "grad_norm": 0.484375, + "learning_rate": 6.639609450680041e-05, + "loss": 2.4245, + "step": 28626 + }, + { + "epoch": 1.22, + "grad_norm": 0.455078125, + "learning_rate": 6.638977673015985e-05, + "loss": 2.4221, + "step": 28627 + }, + { + "epoch": 1.22, + "grad_norm": 0.458984375, + "learning_rate": 6.638345910475304e-05, + "loss": 2.3604, + "step": 28628 + }, + { + "epoch": 1.22, + "grad_norm": 0.470703125, + "learning_rate": 6.637714163060836e-05, + "loss": 2.4256, + "step": 28629 + }, + { + "epoch": 1.22, + "grad_norm": 0.53125, + "learning_rate": 6.637082430775422e-05, + "loss": 2.5273, + "step": 28630 + }, + { + "epoch": 1.22, + "grad_norm": 0.4375, + "learning_rate": 6.636450713621912e-05, + "loss": 2.4435, + "step": 28631 + }, + { + "epoch": 1.22, + "grad_norm": 0.51171875, + "learning_rate": 6.635819011603143e-05, + "loss": 2.5791, + "step": 28632 + }, + { + "epoch": 1.22, + "grad_norm": 0.4609375, + "learning_rate": 6.635187324721959e-05, + "loss": 2.5269, + "step": 28633 + }, + { + "epoch": 1.22, + "grad_norm": 0.48046875, + "learning_rate": 6.6345556529812e-05, + "loss": 2.5782, + "step": 28634 + }, + { + "epoch": 1.22, + "grad_norm": 0.4453125, + "learning_rate": 6.633923996383715e-05, + "loss": 2.4415, + "step": 28635 + }, + { + "epoch": 1.22, + "grad_norm": 0.43359375, + "learning_rate": 6.633292354932342e-05, + "loss": 2.5157, + "step": 28636 + }, + { + "epoch": 1.22, + "grad_norm": 0.490234375, + "learning_rate": 6.63266072862992e-05, + "loss": 2.4319, + "step": 28637 + }, + { + "epoch": 1.22, + "grad_norm": 0.470703125, + "learning_rate": 6.632029117479295e-05, + "loss": 2.5849, + "step": 28638 + }, + { + "epoch": 1.22, + "grad_norm": 0.46484375, + "learning_rate": 6.631397521483308e-05, + "loss": 2.4272, + "step": 28639 + }, + { + "epoch": 1.22, + "grad_norm": 0.4921875, + "learning_rate": 6.630765940644801e-05, + "loss": 2.5417, + "step": 28640 + }, + { + "epoch": 1.22, + "grad_norm": 0.435546875, + "learning_rate": 6.630134374966617e-05, + "loss": 2.3946, + "step": 28641 + }, + { + "epoch": 1.22, + "grad_norm": 0.498046875, + "learning_rate": 6.629502824451594e-05, + "loss": 2.4308, + "step": 28642 + }, + { + "epoch": 1.22, + "grad_norm": 0.53125, + "learning_rate": 6.628871289102578e-05, + "loss": 2.3486, + "step": 28643 + }, + { + "epoch": 1.22, + "grad_norm": 0.5078125, + "learning_rate": 6.628239768922414e-05, + "loss": 2.4485, + "step": 28644 + }, + { + "epoch": 1.22, + "grad_norm": 0.447265625, + "learning_rate": 6.627608263913933e-05, + "loss": 2.5478, + "step": 28645 + }, + { + "epoch": 1.22, + "grad_norm": 0.4765625, + "learning_rate": 6.626976774079982e-05, + "loss": 2.4088, + "step": 28646 + }, + { + "epoch": 1.22, + "grad_norm": 0.51171875, + "learning_rate": 6.626345299423407e-05, + "loss": 2.5344, + "step": 28647 + }, + { + "epoch": 1.22, + "grad_norm": 0.419921875, + "learning_rate": 6.625713839947043e-05, + "loss": 2.3577, + "step": 28648 + }, + { + "epoch": 1.22, + "grad_norm": 0.50390625, + "learning_rate": 6.625082395653733e-05, + "loss": 2.4285, + "step": 28649 + }, + { + "epoch": 1.22, + "grad_norm": 0.65625, + "learning_rate": 6.624450966546318e-05, + "loss": 2.5164, + "step": 28650 + }, + { + "epoch": 1.22, + "grad_norm": 0.49609375, + "learning_rate": 6.623819552627642e-05, + "loss": 2.4671, + "step": 28651 + }, + { + "epoch": 1.22, + "grad_norm": 0.416015625, + "learning_rate": 6.623188153900546e-05, + "loss": 2.2527, + "step": 28652 + }, + { + "epoch": 1.22, + "grad_norm": 0.4140625, + "learning_rate": 6.622556770367866e-05, + "loss": 2.6891, + "step": 28653 + }, + { + "epoch": 1.22, + "grad_norm": 0.4453125, + "learning_rate": 6.621925402032445e-05, + "loss": 2.5062, + "step": 28654 + }, + { + "epoch": 1.22, + "grad_norm": 0.419921875, + "learning_rate": 6.621294048897127e-05, + "loss": 2.3184, + "step": 28655 + }, + { + "epoch": 1.22, + "grad_norm": 0.515625, + "learning_rate": 6.620662710964752e-05, + "loss": 2.5618, + "step": 28656 + }, + { + "epoch": 1.22, + "grad_norm": 0.45703125, + "learning_rate": 6.620031388238159e-05, + "loss": 2.4467, + "step": 28657 + }, + { + "epoch": 1.22, + "grad_norm": 0.443359375, + "learning_rate": 6.619400080720188e-05, + "loss": 2.4275, + "step": 28658 + }, + { + "epoch": 1.22, + "grad_norm": 0.4765625, + "learning_rate": 6.618768788413683e-05, + "loss": 2.5018, + "step": 28659 + }, + { + "epoch": 1.22, + "grad_norm": 0.48828125, + "learning_rate": 6.618137511321485e-05, + "loss": 2.5314, + "step": 28660 + }, + { + "epoch": 1.22, + "grad_norm": 0.462890625, + "learning_rate": 6.617506249446428e-05, + "loss": 2.4355, + "step": 28661 + }, + { + "epoch": 1.22, + "grad_norm": 0.447265625, + "learning_rate": 6.616875002791359e-05, + "loss": 2.4022, + "step": 28662 + }, + { + "epoch": 1.22, + "grad_norm": 0.4453125, + "learning_rate": 6.616243771359114e-05, + "loss": 2.3436, + "step": 28663 + }, + { + "epoch": 1.22, + "grad_norm": 0.47265625, + "learning_rate": 6.615612555152537e-05, + "loss": 2.4302, + "step": 28664 + }, + { + "epoch": 1.22, + "grad_norm": 0.47265625, + "learning_rate": 6.614981354174468e-05, + "loss": 2.4175, + "step": 28665 + }, + { + "epoch": 1.22, + "grad_norm": 0.431640625, + "learning_rate": 6.614350168427742e-05, + "loss": 2.2756, + "step": 28666 + }, + { + "epoch": 1.22, + "grad_norm": 0.455078125, + "learning_rate": 6.613718997915206e-05, + "loss": 2.4444, + "step": 28667 + }, + { + "epoch": 1.22, + "grad_norm": 0.51953125, + "learning_rate": 6.613087842639698e-05, + "loss": 2.4721, + "step": 28668 + }, + { + "epoch": 1.22, + "grad_norm": 0.53125, + "learning_rate": 6.612456702604055e-05, + "loss": 2.5874, + "step": 28669 + }, + { + "epoch": 1.22, + "grad_norm": 0.470703125, + "learning_rate": 6.611825577811117e-05, + "loss": 2.446, + "step": 28670 + }, + { + "epoch": 1.22, + "grad_norm": 0.46875, + "learning_rate": 6.611194468263728e-05, + "loss": 2.5325, + "step": 28671 + }, + { + "epoch": 1.22, + "grad_norm": 0.546875, + "learning_rate": 6.610563373964726e-05, + "loss": 2.4808, + "step": 28672 + }, + { + "epoch": 1.22, + "grad_norm": 0.447265625, + "learning_rate": 6.60993229491695e-05, + "loss": 2.3974, + "step": 28673 + }, + { + "epoch": 1.22, + "grad_norm": 0.58203125, + "learning_rate": 6.609301231123236e-05, + "loss": 2.5747, + "step": 28674 + }, + { + "epoch": 1.22, + "grad_norm": 0.4921875, + "learning_rate": 6.60867018258643e-05, + "loss": 2.4927, + "step": 28675 + }, + { + "epoch": 1.22, + "grad_norm": 0.470703125, + "learning_rate": 6.608039149309373e-05, + "loss": 2.5535, + "step": 28676 + }, + { + "epoch": 1.22, + "grad_norm": 0.453125, + "learning_rate": 6.607408131294896e-05, + "loss": 2.5288, + "step": 28677 + }, + { + "epoch": 1.22, + "grad_norm": 0.50390625, + "learning_rate": 6.60677712854584e-05, + "loss": 2.4518, + "step": 28678 + }, + { + "epoch": 1.22, + "grad_norm": 0.400390625, + "learning_rate": 6.606146141065051e-05, + "loss": 2.351, + "step": 28679 + }, + { + "epoch": 1.22, + "grad_norm": 0.455078125, + "learning_rate": 6.605515168855364e-05, + "loss": 2.3937, + "step": 28680 + }, + { + "epoch": 1.22, + "grad_norm": 0.546875, + "learning_rate": 6.604884211919616e-05, + "loss": 2.5316, + "step": 28681 + }, + { + "epoch": 1.22, + "grad_norm": 0.482421875, + "learning_rate": 6.604253270260647e-05, + "loss": 2.42, + "step": 28682 + }, + { + "epoch": 1.22, + "grad_norm": 0.458984375, + "learning_rate": 6.6036223438813e-05, + "loss": 2.457, + "step": 28683 + }, + { + "epoch": 1.22, + "grad_norm": 0.55078125, + "learning_rate": 6.602991432784414e-05, + "loss": 2.5088, + "step": 28684 + }, + { + "epoch": 1.22, + "grad_norm": 0.5, + "learning_rate": 6.60236053697282e-05, + "loss": 2.5358, + "step": 28685 + }, + { + "epoch": 1.22, + "grad_norm": 0.44140625, + "learning_rate": 6.601729656449363e-05, + "loss": 2.3838, + "step": 28686 + }, + { + "epoch": 1.22, + "grad_norm": 0.4765625, + "learning_rate": 6.60109879121688e-05, + "loss": 2.5221, + "step": 28687 + }, + { + "epoch": 1.22, + "grad_norm": 0.470703125, + "learning_rate": 6.600467941278212e-05, + "loss": 2.4577, + "step": 28688 + }, + { + "epoch": 1.22, + "grad_norm": 0.5234375, + "learning_rate": 6.599837106636196e-05, + "loss": 2.4659, + "step": 28689 + }, + { + "epoch": 1.22, + "grad_norm": 0.50390625, + "learning_rate": 6.599206287293669e-05, + "loss": 2.5439, + "step": 28690 + }, + { + "epoch": 1.22, + "grad_norm": 0.5234375, + "learning_rate": 6.59857548325347e-05, + "loss": 2.4289, + "step": 28691 + }, + { + "epoch": 1.22, + "grad_norm": 0.5, + "learning_rate": 6.597944694518443e-05, + "loss": 2.426, + "step": 28692 + }, + { + "epoch": 1.22, + "grad_norm": 0.494140625, + "learning_rate": 6.597313921091419e-05, + "loss": 2.513, + "step": 28693 + }, + { + "epoch": 1.22, + "grad_norm": 0.474609375, + "learning_rate": 6.596683162975237e-05, + "loss": 2.3568, + "step": 28694 + }, + { + "epoch": 1.22, + "grad_norm": 0.51171875, + "learning_rate": 6.596052420172738e-05, + "loss": 2.4894, + "step": 28695 + }, + { + "epoch": 1.22, + "grad_norm": 0.4453125, + "learning_rate": 6.595421692686761e-05, + "loss": 2.441, + "step": 28696 + }, + { + "epoch": 1.22, + "grad_norm": 0.48828125, + "learning_rate": 6.59479098052014e-05, + "loss": 2.4454, + "step": 28697 + }, + { + "epoch": 1.22, + "grad_norm": 0.45703125, + "learning_rate": 6.594160283675714e-05, + "loss": 2.3846, + "step": 28698 + }, + { + "epoch": 1.22, + "grad_norm": 0.46875, + "learning_rate": 6.593529602156323e-05, + "loss": 2.5446, + "step": 28699 + }, + { + "epoch": 1.22, + "grad_norm": 0.486328125, + "learning_rate": 6.592898935964808e-05, + "loss": 2.5926, + "step": 28700 + }, + { + "epoch": 1.22, + "grad_norm": 0.40625, + "learning_rate": 6.592268285103999e-05, + "loss": 2.5535, + "step": 28701 + }, + { + "epoch": 1.22, + "grad_norm": 0.4296875, + "learning_rate": 6.591637649576737e-05, + "loss": 2.4967, + "step": 28702 + }, + { + "epoch": 1.22, + "grad_norm": 0.486328125, + "learning_rate": 6.59100702938586e-05, + "loss": 2.4101, + "step": 28703 + }, + { + "epoch": 1.22, + "grad_norm": 0.435546875, + "learning_rate": 6.590376424534206e-05, + "loss": 2.4073, + "step": 28704 + }, + { + "epoch": 1.22, + "grad_norm": 0.43359375, + "learning_rate": 6.589745835024613e-05, + "loss": 2.3548, + "step": 28705 + }, + { + "epoch": 1.22, + "grad_norm": 0.443359375, + "learning_rate": 6.589115260859913e-05, + "loss": 2.4531, + "step": 28706 + }, + { + "epoch": 1.22, + "grad_norm": 0.515625, + "learning_rate": 6.588484702042953e-05, + "loss": 2.5211, + "step": 28707 + }, + { + "epoch": 1.22, + "grad_norm": 0.484375, + "learning_rate": 6.587854158576565e-05, + "loss": 2.3476, + "step": 28708 + }, + { + "epoch": 1.22, + "grad_norm": 0.431640625, + "learning_rate": 6.587223630463586e-05, + "loss": 2.1849, + "step": 28709 + }, + { + "epoch": 1.22, + "grad_norm": 0.447265625, + "learning_rate": 6.58659311770685e-05, + "loss": 2.2557, + "step": 28710 + }, + { + "epoch": 1.22, + "grad_norm": 0.53125, + "learning_rate": 6.585962620309198e-05, + "loss": 2.5391, + "step": 28711 + }, + { + "epoch": 1.22, + "grad_norm": 0.53515625, + "learning_rate": 6.58533213827347e-05, + "loss": 2.5115, + "step": 28712 + }, + { + "epoch": 1.22, + "grad_norm": 0.431640625, + "learning_rate": 6.584701671602498e-05, + "loss": 2.4796, + "step": 28713 + }, + { + "epoch": 1.22, + "grad_norm": 0.4921875, + "learning_rate": 6.584071220299117e-05, + "loss": 2.5551, + "step": 28714 + }, + { + "epoch": 1.22, + "grad_norm": 0.49609375, + "learning_rate": 6.583440784366171e-05, + "loss": 2.4134, + "step": 28715 + }, + { + "epoch": 1.22, + "grad_norm": 0.48046875, + "learning_rate": 6.582810363806495e-05, + "loss": 2.4744, + "step": 28716 + }, + { + "epoch": 1.22, + "grad_norm": 0.486328125, + "learning_rate": 6.582179958622921e-05, + "loss": 2.5906, + "step": 28717 + }, + { + "epoch": 1.22, + "grad_norm": 0.51171875, + "learning_rate": 6.581549568818282e-05, + "loss": 2.5523, + "step": 28718 + }, + { + "epoch": 1.22, + "grad_norm": 0.466796875, + "learning_rate": 6.58091919439543e-05, + "loss": 2.5116, + "step": 28719 + }, + { + "epoch": 1.22, + "grad_norm": 0.515625, + "learning_rate": 6.580288835357186e-05, + "loss": 2.4606, + "step": 28720 + }, + { + "epoch": 1.22, + "grad_norm": 0.447265625, + "learning_rate": 6.579658491706397e-05, + "loss": 2.3672, + "step": 28721 + }, + { + "epoch": 1.22, + "grad_norm": 0.5546875, + "learning_rate": 6.579028163445891e-05, + "loss": 2.3833, + "step": 28722 + }, + { + "epoch": 1.22, + "grad_norm": 0.416015625, + "learning_rate": 6.578397850578508e-05, + "loss": 2.3705, + "step": 28723 + }, + { + "epoch": 1.22, + "grad_norm": 0.55859375, + "learning_rate": 6.577767553107089e-05, + "loss": 2.5953, + "step": 28724 + }, + { + "epoch": 1.22, + "grad_norm": 0.3984375, + "learning_rate": 6.577137271034462e-05, + "loss": 2.3857, + "step": 28725 + }, + { + "epoch": 1.22, + "grad_norm": 0.453125, + "learning_rate": 6.576507004363463e-05, + "loss": 2.3939, + "step": 28726 + }, + { + "epoch": 1.22, + "grad_norm": 0.439453125, + "learning_rate": 6.575876753096937e-05, + "loss": 2.42, + "step": 28727 + }, + { + "epoch": 1.22, + "grad_norm": 0.486328125, + "learning_rate": 6.575246517237712e-05, + "loss": 2.4616, + "step": 28728 + }, + { + "epoch": 1.22, + "grad_norm": 0.43359375, + "learning_rate": 6.574616296788627e-05, + "loss": 2.37, + "step": 28729 + }, + { + "epoch": 1.22, + "grad_norm": 0.47265625, + "learning_rate": 6.573986091752514e-05, + "loss": 2.4638, + "step": 28730 + }, + { + "epoch": 1.22, + "grad_norm": 0.412109375, + "learning_rate": 6.573355902132214e-05, + "loss": 2.4007, + "step": 28731 + }, + { + "epoch": 1.22, + "grad_norm": 0.51171875, + "learning_rate": 6.572725727930562e-05, + "loss": 2.4314, + "step": 28732 + }, + { + "epoch": 1.22, + "grad_norm": 0.474609375, + "learning_rate": 6.57209556915039e-05, + "loss": 2.3889, + "step": 28733 + }, + { + "epoch": 1.22, + "grad_norm": 0.4296875, + "learning_rate": 6.57146542579453e-05, + "loss": 2.4692, + "step": 28734 + }, + { + "epoch": 1.22, + "grad_norm": 0.482421875, + "learning_rate": 6.57083529786583e-05, + "loss": 2.3562, + "step": 28735 + }, + { + "epoch": 1.22, + "grad_norm": 0.451171875, + "learning_rate": 6.570205185367115e-05, + "loss": 2.3923, + "step": 28736 + }, + { + "epoch": 1.22, + "grad_norm": 0.474609375, + "learning_rate": 6.569575088301224e-05, + "loss": 2.4887, + "step": 28737 + }, + { + "epoch": 1.22, + "grad_norm": 0.462890625, + "learning_rate": 6.568945006670992e-05, + "loss": 2.3693, + "step": 28738 + }, + { + "epoch": 1.22, + "grad_norm": 0.451171875, + "learning_rate": 6.56831494047925e-05, + "loss": 2.3925, + "step": 28739 + }, + { + "epoch": 1.22, + "grad_norm": 0.50390625, + "learning_rate": 6.56768488972884e-05, + "loss": 2.4729, + "step": 28740 + }, + { + "epoch": 1.22, + "grad_norm": 0.48046875, + "learning_rate": 6.567054854422592e-05, + "loss": 2.4249, + "step": 28741 + }, + { + "epoch": 1.22, + "grad_norm": 0.466796875, + "learning_rate": 6.566424834563345e-05, + "loss": 2.3801, + "step": 28742 + }, + { + "epoch": 1.22, + "grad_norm": 0.5, + "learning_rate": 6.565794830153926e-05, + "loss": 2.4842, + "step": 28743 + }, + { + "epoch": 1.22, + "grad_norm": 0.5, + "learning_rate": 6.56516484119718e-05, + "loss": 2.5311, + "step": 28744 + }, + { + "epoch": 1.22, + "grad_norm": 0.431640625, + "learning_rate": 6.564534867695935e-05, + "loss": 2.3906, + "step": 28745 + }, + { + "epoch": 1.22, + "grad_norm": 0.439453125, + "learning_rate": 6.563904909653031e-05, + "loss": 2.4259, + "step": 28746 + }, + { + "epoch": 1.22, + "grad_norm": 0.47265625, + "learning_rate": 6.563274967071293e-05, + "loss": 2.4212, + "step": 28747 + }, + { + "epoch": 1.23, + "grad_norm": 0.443359375, + "learning_rate": 6.562645039953567e-05, + "loss": 2.4373, + "step": 28748 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.562015128302681e-05, + "loss": 2.3554, + "step": 28749 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.56138523212147e-05, + "loss": 2.516, + "step": 28750 + }, + { + "epoch": 1.23, + "grad_norm": 0.53125, + "learning_rate": 6.560755351412766e-05, + "loss": 2.5467, + "step": 28751 + }, + { + "epoch": 1.23, + "grad_norm": 0.46484375, + "learning_rate": 6.560125486179409e-05, + "loss": 2.4325, + "step": 28752 + }, + { + "epoch": 1.23, + "grad_norm": 0.478515625, + "learning_rate": 6.55949563642423e-05, + "loss": 2.5342, + "step": 28753 + }, + { + "epoch": 1.23, + "grad_norm": 0.474609375, + "learning_rate": 6.558865802150065e-05, + "loss": 2.4827, + "step": 28754 + }, + { + "epoch": 1.23, + "grad_norm": 0.490234375, + "learning_rate": 6.558235983359739e-05, + "loss": 2.3843, + "step": 28755 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.557606180056102e-05, + "loss": 2.4983, + "step": 28756 + }, + { + "epoch": 1.23, + "grad_norm": 0.482421875, + "learning_rate": 6.556976392241976e-05, + "loss": 2.4697, + "step": 28757 + }, + { + "epoch": 1.23, + "grad_norm": 0.47265625, + "learning_rate": 6.556346619920197e-05, + "loss": 2.5546, + "step": 28758 + }, + { + "epoch": 1.23, + "grad_norm": 0.455078125, + "learning_rate": 6.555716863093599e-05, + "loss": 2.3493, + "step": 28759 + }, + { + "epoch": 1.23, + "grad_norm": 0.455078125, + "learning_rate": 6.555087121765018e-05, + "loss": 2.4552, + "step": 28760 + }, + { + "epoch": 1.23, + "grad_norm": 0.384765625, + "learning_rate": 6.554457395937286e-05, + "loss": 2.4969, + "step": 28761 + }, + { + "epoch": 1.23, + "grad_norm": 0.458984375, + "learning_rate": 6.553827685613239e-05, + "loss": 2.4837, + "step": 28762 + }, + { + "epoch": 1.23, + "grad_norm": 0.431640625, + "learning_rate": 6.5531979907957e-05, + "loss": 2.4461, + "step": 28763 + }, + { + "epoch": 1.23, + "grad_norm": 0.453125, + "learning_rate": 6.552568311487518e-05, + "loss": 2.5793, + "step": 28764 + }, + { + "epoch": 1.23, + "grad_norm": 0.482421875, + "learning_rate": 6.551938647691516e-05, + "loss": 2.5349, + "step": 28765 + }, + { + "epoch": 1.23, + "grad_norm": 0.51171875, + "learning_rate": 6.55130899941053e-05, + "loss": 2.4556, + "step": 28766 + }, + { + "epoch": 1.23, + "grad_norm": 0.484375, + "learning_rate": 6.550679366647392e-05, + "loss": 2.6257, + "step": 28767 + }, + { + "epoch": 1.23, + "grad_norm": 0.45703125, + "learning_rate": 6.550049749404938e-05, + "loss": 2.3835, + "step": 28768 + }, + { + "epoch": 1.23, + "grad_norm": 0.5078125, + "learning_rate": 6.549420147686e-05, + "loss": 2.4227, + "step": 28769 + }, + { + "epoch": 1.23, + "grad_norm": 0.5078125, + "learning_rate": 6.548790561493411e-05, + "loss": 2.4349, + "step": 28770 + }, + { + "epoch": 1.23, + "grad_norm": 0.51171875, + "learning_rate": 6.548160990829996e-05, + "loss": 2.5189, + "step": 28771 + }, + { + "epoch": 1.23, + "grad_norm": 0.44921875, + "learning_rate": 6.547531435698602e-05, + "loss": 2.4261, + "step": 28772 + }, + { + "epoch": 1.23, + "grad_norm": 0.5546875, + "learning_rate": 6.546901896102056e-05, + "loss": 2.4269, + "step": 28773 + }, + { + "epoch": 1.23, + "grad_norm": 0.498046875, + "learning_rate": 6.546272372043187e-05, + "loss": 2.4587, + "step": 28774 + }, + { + "epoch": 1.23, + "grad_norm": 0.451171875, + "learning_rate": 6.545642863524829e-05, + "loss": 2.4866, + "step": 28775 + }, + { + "epoch": 1.23, + "grad_norm": 0.515625, + "learning_rate": 6.545013370549815e-05, + "loss": 2.4309, + "step": 28776 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.54438389312098e-05, + "loss": 2.5491, + "step": 28777 + }, + { + "epoch": 1.23, + "grad_norm": 0.5390625, + "learning_rate": 6.543754431241157e-05, + "loss": 2.6022, + "step": 28778 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.543124984913169e-05, + "loss": 2.5041, + "step": 28779 + }, + { + "epoch": 1.23, + "grad_norm": 0.54296875, + "learning_rate": 6.54249555413986e-05, + "loss": 2.5742, + "step": 28780 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.54186613892406e-05, + "loss": 2.4169, + "step": 28781 + }, + { + "epoch": 1.23, + "grad_norm": 0.4296875, + "learning_rate": 6.541236739268595e-05, + "loss": 2.5002, + "step": 28782 + }, + { + "epoch": 1.23, + "grad_norm": 0.48046875, + "learning_rate": 6.5406073551763e-05, + "loss": 2.6227, + "step": 28783 + }, + { + "epoch": 1.23, + "grad_norm": 0.392578125, + "learning_rate": 6.539977986650009e-05, + "loss": 2.4259, + "step": 28784 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.539348633692553e-05, + "loss": 2.4869, + "step": 28785 + }, + { + "epoch": 1.23, + "grad_norm": 0.40625, + "learning_rate": 6.538719296306765e-05, + "loss": 2.4765, + "step": 28786 + }, + { + "epoch": 1.23, + "grad_norm": 0.466796875, + "learning_rate": 6.538089974495468e-05, + "loss": 2.4259, + "step": 28787 + }, + { + "epoch": 1.23, + "grad_norm": 0.49609375, + "learning_rate": 6.537460668261506e-05, + "loss": 2.5047, + "step": 28788 + }, + { + "epoch": 1.23, + "grad_norm": 0.421875, + "learning_rate": 6.536831377607709e-05, + "loss": 2.5055, + "step": 28789 + }, + { + "epoch": 1.23, + "grad_norm": 0.49609375, + "learning_rate": 6.536202102536902e-05, + "loss": 2.5445, + "step": 28790 + }, + { + "epoch": 1.23, + "grad_norm": 0.4765625, + "learning_rate": 6.535572843051916e-05, + "loss": 2.4468, + "step": 28791 + }, + { + "epoch": 1.23, + "grad_norm": 0.478515625, + "learning_rate": 6.53494359915559e-05, + "loss": 2.519, + "step": 28792 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.534314370850751e-05, + "loss": 2.4399, + "step": 28793 + }, + { + "epoch": 1.23, + "grad_norm": 0.462890625, + "learning_rate": 6.533685158140233e-05, + "loss": 2.4529, + "step": 28794 + }, + { + "epoch": 1.23, + "grad_norm": 0.4453125, + "learning_rate": 6.533055961026859e-05, + "loss": 2.4367, + "step": 28795 + }, + { + "epoch": 1.23, + "grad_norm": 0.458984375, + "learning_rate": 6.53242677951347e-05, + "loss": 2.5245, + "step": 28796 + }, + { + "epoch": 1.23, + "grad_norm": 0.45703125, + "learning_rate": 6.531797613602895e-05, + "loss": 2.4911, + "step": 28797 + }, + { + "epoch": 1.23, + "grad_norm": 0.490234375, + "learning_rate": 6.531168463297963e-05, + "loss": 2.5479, + "step": 28798 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.530539328601501e-05, + "loss": 2.393, + "step": 28799 + }, + { + "epoch": 1.23, + "grad_norm": 0.396484375, + "learning_rate": 6.529910209516347e-05, + "loss": 2.4913, + "step": 28800 + }, + { + "epoch": 1.23, + "grad_norm": 0.52734375, + "learning_rate": 6.529281106045328e-05, + "loss": 2.5092, + "step": 28801 + }, + { + "epoch": 1.23, + "grad_norm": 0.482421875, + "learning_rate": 6.528652018191279e-05, + "loss": 2.2885, + "step": 28802 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.52802294595702e-05, + "loss": 2.2867, + "step": 28803 + }, + { + "epoch": 1.23, + "grad_norm": 0.4609375, + "learning_rate": 6.527393889345395e-05, + "loss": 2.5428, + "step": 28804 + }, + { + "epoch": 1.23, + "grad_norm": 0.486328125, + "learning_rate": 6.526764848359227e-05, + "loss": 2.3774, + "step": 28805 + }, + { + "epoch": 1.23, + "grad_norm": 0.453125, + "learning_rate": 6.52613582300135e-05, + "loss": 2.4865, + "step": 28806 + }, + { + "epoch": 1.23, + "grad_norm": 0.5390625, + "learning_rate": 6.525506813274586e-05, + "loss": 2.5135, + "step": 28807 + }, + { + "epoch": 1.23, + "grad_norm": 0.439453125, + "learning_rate": 6.524877819181776e-05, + "loss": 2.4129, + "step": 28808 + }, + { + "epoch": 1.23, + "grad_norm": 0.51953125, + "learning_rate": 6.524248840725746e-05, + "loss": 2.3745, + "step": 28809 + }, + { + "epoch": 1.23, + "grad_norm": 0.400390625, + "learning_rate": 6.523619877909327e-05, + "loss": 2.3591, + "step": 28810 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.522990930735342e-05, + "loss": 2.4617, + "step": 28811 + }, + { + "epoch": 1.23, + "grad_norm": 0.4375, + "learning_rate": 6.522361999206632e-05, + "loss": 2.3284, + "step": 28812 + }, + { + "epoch": 1.23, + "grad_norm": 0.4296875, + "learning_rate": 6.521733083326022e-05, + "loss": 2.4586, + "step": 28813 + }, + { + "epoch": 1.23, + "grad_norm": 0.40625, + "learning_rate": 6.521104183096343e-05, + "loss": 2.4303, + "step": 28814 + }, + { + "epoch": 1.23, + "grad_norm": 0.5, + "learning_rate": 6.520475298520421e-05, + "loss": 2.4547, + "step": 28815 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.519846429601088e-05, + "loss": 2.5035, + "step": 28816 + }, + { + "epoch": 1.23, + "grad_norm": 0.466796875, + "learning_rate": 6.519217576341177e-05, + "loss": 2.4012, + "step": 28817 + }, + { + "epoch": 1.23, + "grad_norm": 0.419921875, + "learning_rate": 6.518588738743516e-05, + "loss": 2.4671, + "step": 28818 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.517959916810927e-05, + "loss": 2.618, + "step": 28819 + }, + { + "epoch": 1.23, + "grad_norm": 0.451171875, + "learning_rate": 6.51733111054625e-05, + "loss": 2.481, + "step": 28820 + }, + { + "epoch": 1.23, + "grad_norm": 0.435546875, + "learning_rate": 6.516702319952312e-05, + "loss": 2.4472, + "step": 28821 + }, + { + "epoch": 1.23, + "grad_norm": 0.44921875, + "learning_rate": 6.51607354503194e-05, + "loss": 2.5321, + "step": 28822 + }, + { + "epoch": 1.23, + "grad_norm": 0.45703125, + "learning_rate": 6.515444785787961e-05, + "loss": 2.3873, + "step": 28823 + }, + { + "epoch": 1.23, + "grad_norm": 0.453125, + "learning_rate": 6.514816042223209e-05, + "loss": 2.4012, + "step": 28824 + }, + { + "epoch": 1.23, + "grad_norm": 0.48046875, + "learning_rate": 6.514187314340509e-05, + "loss": 2.5397, + "step": 28825 + }, + { + "epoch": 1.23, + "grad_norm": 0.4453125, + "learning_rate": 6.513558602142695e-05, + "loss": 2.5137, + "step": 28826 + }, + { + "epoch": 1.23, + "grad_norm": 0.54296875, + "learning_rate": 6.512929905632589e-05, + "loss": 2.4477, + "step": 28827 + }, + { + "epoch": 1.23, + "grad_norm": 0.404296875, + "learning_rate": 6.512301224813029e-05, + "loss": 2.4446, + "step": 28828 + }, + { + "epoch": 1.23, + "grad_norm": 0.4375, + "learning_rate": 6.511672559686839e-05, + "loss": 2.4393, + "step": 28829 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.511043910256845e-05, + "loss": 2.4525, + "step": 28830 + }, + { + "epoch": 1.23, + "grad_norm": 0.53125, + "learning_rate": 6.510415276525877e-05, + "loss": 2.5113, + "step": 28831 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.509786658496769e-05, + "loss": 2.4566, + "step": 28832 + }, + { + "epoch": 1.23, + "grad_norm": 0.44140625, + "learning_rate": 6.509158056172341e-05, + "loss": 2.4276, + "step": 28833 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.50852946955543e-05, + "loss": 2.4314, + "step": 28834 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.507900898648856e-05, + "loss": 2.5277, + "step": 28835 + }, + { + "epoch": 1.23, + "grad_norm": 0.408203125, + "learning_rate": 6.507272343455455e-05, + "loss": 2.4302, + "step": 28836 + }, + { + "epoch": 1.23, + "grad_norm": 0.478515625, + "learning_rate": 6.506643803978054e-05, + "loss": 2.5168, + "step": 28837 + }, + { + "epoch": 1.23, + "grad_norm": 0.423828125, + "learning_rate": 6.506015280219477e-05, + "loss": 2.6272, + "step": 28838 + }, + { + "epoch": 1.23, + "grad_norm": 0.6484375, + "learning_rate": 6.505386772182552e-05, + "loss": 2.5212, + "step": 28839 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.50475827987011e-05, + "loss": 2.5327, + "step": 28840 + }, + { + "epoch": 1.23, + "grad_norm": 0.4765625, + "learning_rate": 6.504129803284981e-05, + "loss": 2.5205, + "step": 28841 + }, + { + "epoch": 1.23, + "grad_norm": 0.458984375, + "learning_rate": 6.50350134242999e-05, + "loss": 2.5553, + "step": 28842 + }, + { + "epoch": 1.23, + "grad_norm": 0.5234375, + "learning_rate": 6.502872897307963e-05, + "loss": 2.4843, + "step": 28843 + }, + { + "epoch": 1.23, + "grad_norm": 0.490234375, + "learning_rate": 6.50224446792173e-05, + "loss": 2.3916, + "step": 28844 + }, + { + "epoch": 1.23, + "grad_norm": 0.4609375, + "learning_rate": 6.501616054274123e-05, + "loss": 2.483, + "step": 28845 + }, + { + "epoch": 1.23, + "grad_norm": 0.484375, + "learning_rate": 6.500987656367962e-05, + "loss": 2.5802, + "step": 28846 + }, + { + "epoch": 1.23, + "grad_norm": 0.458984375, + "learning_rate": 6.500359274206076e-05, + "loss": 2.4082, + "step": 28847 + }, + { + "epoch": 1.23, + "grad_norm": 0.484375, + "learning_rate": 6.499730907791296e-05, + "loss": 2.5249, + "step": 28848 + }, + { + "epoch": 1.23, + "grad_norm": 0.443359375, + "learning_rate": 6.499102557126448e-05, + "loss": 2.3396, + "step": 28849 + }, + { + "epoch": 1.23, + "grad_norm": 0.474609375, + "learning_rate": 6.49847422221436e-05, + "loss": 2.2831, + "step": 28850 + }, + { + "epoch": 1.23, + "grad_norm": 0.466796875, + "learning_rate": 6.497845903057855e-05, + "loss": 2.3826, + "step": 28851 + }, + { + "epoch": 1.23, + "grad_norm": 0.515625, + "learning_rate": 6.497217599659767e-05, + "loss": 2.4813, + "step": 28852 + }, + { + "epoch": 1.23, + "grad_norm": 0.482421875, + "learning_rate": 6.496589312022919e-05, + "loss": 2.3785, + "step": 28853 + }, + { + "epoch": 1.23, + "grad_norm": 0.44140625, + "learning_rate": 6.495961040150138e-05, + "loss": 2.4353, + "step": 28854 + }, + { + "epoch": 1.23, + "grad_norm": 0.4453125, + "learning_rate": 6.49533278404425e-05, + "loss": 2.6045, + "step": 28855 + }, + { + "epoch": 1.23, + "grad_norm": 0.5078125, + "learning_rate": 6.494704543708086e-05, + "loss": 2.5698, + "step": 28856 + }, + { + "epoch": 1.23, + "grad_norm": 0.51953125, + "learning_rate": 6.49407631914447e-05, + "loss": 2.4876, + "step": 28857 + }, + { + "epoch": 1.23, + "grad_norm": 0.5546875, + "learning_rate": 6.49344811035623e-05, + "loss": 2.5723, + "step": 28858 + }, + { + "epoch": 1.23, + "grad_norm": 0.45703125, + "learning_rate": 6.492819917346193e-05, + "loss": 2.4834, + "step": 28859 + }, + { + "epoch": 1.23, + "grad_norm": 0.4921875, + "learning_rate": 6.492191740117179e-05, + "loss": 2.4818, + "step": 28860 + }, + { + "epoch": 1.23, + "grad_norm": 0.4921875, + "learning_rate": 6.491563578672028e-05, + "loss": 2.4238, + "step": 28861 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.490935433013554e-05, + "loss": 2.501, + "step": 28862 + }, + { + "epoch": 1.23, + "grad_norm": 0.431640625, + "learning_rate": 6.490307303144589e-05, + "loss": 2.5003, + "step": 28863 + }, + { + "epoch": 1.23, + "grad_norm": 0.43359375, + "learning_rate": 6.489679189067957e-05, + "loss": 2.4801, + "step": 28864 + }, + { + "epoch": 1.23, + "grad_norm": 0.46484375, + "learning_rate": 6.48905109078649e-05, + "loss": 2.4057, + "step": 28865 + }, + { + "epoch": 1.23, + "grad_norm": 0.478515625, + "learning_rate": 6.488423008303006e-05, + "loss": 2.5055, + "step": 28866 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.487794941620339e-05, + "loss": 2.3898, + "step": 28867 + }, + { + "epoch": 1.23, + "grad_norm": 0.53125, + "learning_rate": 6.487166890741304e-05, + "loss": 2.4472, + "step": 28868 + }, + { + "epoch": 1.23, + "grad_norm": 0.453125, + "learning_rate": 6.486538855668742e-05, + "loss": 2.4597, + "step": 28869 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.485910836405468e-05, + "loss": 2.3768, + "step": 28870 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.485282832954311e-05, + "loss": 2.5387, + "step": 28871 + }, + { + "epoch": 1.23, + "grad_norm": 0.44921875, + "learning_rate": 6.484654845318094e-05, + "loss": 2.5044, + "step": 28872 + }, + { + "epoch": 1.23, + "grad_norm": 0.46484375, + "learning_rate": 6.484026873499649e-05, + "loss": 2.438, + "step": 28873 + }, + { + "epoch": 1.23, + "grad_norm": 0.54296875, + "learning_rate": 6.483398917501797e-05, + "loss": 2.5322, + "step": 28874 + }, + { + "epoch": 1.23, + "grad_norm": 0.447265625, + "learning_rate": 6.482770977327368e-05, + "loss": 2.4559, + "step": 28875 + }, + { + "epoch": 1.23, + "grad_norm": 0.4609375, + "learning_rate": 6.482143052979177e-05, + "loss": 2.4911, + "step": 28876 + }, + { + "epoch": 1.23, + "grad_norm": 0.439453125, + "learning_rate": 6.481515144460064e-05, + "loss": 2.437, + "step": 28877 + }, + { + "epoch": 1.23, + "grad_norm": 0.515625, + "learning_rate": 6.480887251772843e-05, + "loss": 2.5564, + "step": 28878 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.480259374920346e-05, + "loss": 2.4845, + "step": 28879 + }, + { + "epoch": 1.23, + "grad_norm": 0.51171875, + "learning_rate": 6.479631513905391e-05, + "loss": 2.3884, + "step": 28880 + }, + { + "epoch": 1.23, + "grad_norm": 0.490234375, + "learning_rate": 6.47900366873081e-05, + "loss": 2.4639, + "step": 28881 + }, + { + "epoch": 1.23, + "grad_norm": 0.4140625, + "learning_rate": 6.478375839399427e-05, + "loss": 2.4745, + "step": 28882 + }, + { + "epoch": 1.23, + "grad_norm": 0.447265625, + "learning_rate": 6.477748025914065e-05, + "loss": 2.5763, + "step": 28883 + }, + { + "epoch": 1.23, + "grad_norm": 0.4375, + "learning_rate": 6.477120228277547e-05, + "loss": 2.4841, + "step": 28884 + }, + { + "epoch": 1.23, + "grad_norm": 0.474609375, + "learning_rate": 6.476492446492705e-05, + "loss": 2.4374, + "step": 28885 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.475864680562358e-05, + "loss": 2.464, + "step": 28886 + }, + { + "epoch": 1.23, + "grad_norm": 0.447265625, + "learning_rate": 6.47523693048933e-05, + "loss": 2.4739, + "step": 28887 + }, + { + "epoch": 1.23, + "grad_norm": 0.4453125, + "learning_rate": 6.474609196276448e-05, + "loss": 2.524, + "step": 28888 + }, + { + "epoch": 1.23, + "grad_norm": 0.4296875, + "learning_rate": 6.473981477926536e-05, + "loss": 2.3343, + "step": 28889 + }, + { + "epoch": 1.23, + "grad_norm": 0.53515625, + "learning_rate": 6.47335377544242e-05, + "loss": 2.5241, + "step": 28890 + }, + { + "epoch": 1.23, + "grad_norm": 0.5, + "learning_rate": 6.472726088826926e-05, + "loss": 2.5036, + "step": 28891 + }, + { + "epoch": 1.23, + "grad_norm": 0.5, + "learning_rate": 6.472098418082868e-05, + "loss": 2.6445, + "step": 28892 + }, + { + "epoch": 1.23, + "grad_norm": 0.484375, + "learning_rate": 6.471470763213086e-05, + "loss": 2.4381, + "step": 28893 + }, + { + "epoch": 1.23, + "grad_norm": 0.51171875, + "learning_rate": 6.470843124220391e-05, + "loss": 2.5945, + "step": 28894 + }, + { + "epoch": 1.23, + "grad_norm": 0.455078125, + "learning_rate": 6.470215501107615e-05, + "loss": 2.2859, + "step": 28895 + }, + { + "epoch": 1.23, + "grad_norm": 0.41796875, + "learning_rate": 6.469587893877575e-05, + "loss": 2.4662, + "step": 28896 + }, + { + "epoch": 1.23, + "grad_norm": 0.515625, + "learning_rate": 6.468960302533102e-05, + "loss": 2.5752, + "step": 28897 + }, + { + "epoch": 1.23, + "grad_norm": 0.419921875, + "learning_rate": 6.468332727077018e-05, + "loss": 2.3594, + "step": 28898 + }, + { + "epoch": 1.23, + "grad_norm": 0.443359375, + "learning_rate": 6.467705167512148e-05, + "loss": 2.4362, + "step": 28899 + }, + { + "epoch": 1.23, + "grad_norm": 0.484375, + "learning_rate": 6.467077623841307e-05, + "loss": 2.4157, + "step": 28900 + }, + { + "epoch": 1.23, + "grad_norm": 0.515625, + "learning_rate": 6.466450096067331e-05, + "loss": 2.4016, + "step": 28901 + }, + { + "epoch": 1.23, + "grad_norm": 0.443359375, + "learning_rate": 6.465822584193038e-05, + "loss": 2.3804, + "step": 28902 + }, + { + "epoch": 1.23, + "grad_norm": 0.4765625, + "learning_rate": 6.46519508822125e-05, + "loss": 2.414, + "step": 28903 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.464567608154793e-05, + "loss": 2.5067, + "step": 28904 + }, + { + "epoch": 1.23, + "grad_norm": 0.490234375, + "learning_rate": 6.46394014399649e-05, + "loss": 2.4426, + "step": 28905 + }, + { + "epoch": 1.23, + "grad_norm": 0.462890625, + "learning_rate": 6.463312695749163e-05, + "loss": 2.4798, + "step": 28906 + }, + { + "epoch": 1.23, + "grad_norm": 0.486328125, + "learning_rate": 6.46268526341564e-05, + "loss": 2.3304, + "step": 28907 + }, + { + "epoch": 1.23, + "grad_norm": 0.474609375, + "learning_rate": 6.462057846998733e-05, + "loss": 2.395, + "step": 28908 + }, + { + "epoch": 1.23, + "grad_norm": 0.455078125, + "learning_rate": 6.46143044650128e-05, + "loss": 2.51, + "step": 28909 + }, + { + "epoch": 1.23, + "grad_norm": 0.474609375, + "learning_rate": 6.460803061926094e-05, + "loss": 2.5116, + "step": 28910 + }, + { + "epoch": 1.23, + "grad_norm": 0.423828125, + "learning_rate": 6.460175693276002e-05, + "loss": 2.4945, + "step": 28911 + }, + { + "epoch": 1.23, + "grad_norm": 0.486328125, + "learning_rate": 6.459548340553823e-05, + "loss": 2.4887, + "step": 28912 + }, + { + "epoch": 1.23, + "grad_norm": 0.4375, + "learning_rate": 6.458921003762384e-05, + "loss": 2.4491, + "step": 28913 + }, + { + "epoch": 1.23, + "grad_norm": 0.455078125, + "learning_rate": 6.458293682904506e-05, + "loss": 2.4657, + "step": 28914 + }, + { + "epoch": 1.23, + "grad_norm": 0.451171875, + "learning_rate": 6.457666377983016e-05, + "loss": 2.421, + "step": 28915 + }, + { + "epoch": 1.23, + "grad_norm": 0.43359375, + "learning_rate": 6.457039089000725e-05, + "loss": 2.4643, + "step": 28916 + }, + { + "epoch": 1.23, + "grad_norm": 0.443359375, + "learning_rate": 6.45641181596047e-05, + "loss": 2.6451, + "step": 28917 + }, + { + "epoch": 1.23, + "grad_norm": 0.48046875, + "learning_rate": 6.455784558865064e-05, + "loss": 2.4198, + "step": 28918 + }, + { + "epoch": 1.23, + "grad_norm": 0.447265625, + "learning_rate": 6.455157317717333e-05, + "loss": 2.5621, + "step": 28919 + }, + { + "epoch": 1.23, + "grad_norm": 0.546875, + "learning_rate": 6.454530092520096e-05, + "loss": 2.4164, + "step": 28920 + }, + { + "epoch": 1.23, + "grad_norm": 0.408203125, + "learning_rate": 6.453902883276182e-05, + "loss": 2.4922, + "step": 28921 + }, + { + "epoch": 1.23, + "grad_norm": 0.486328125, + "learning_rate": 6.453275689988408e-05, + "loss": 2.448, + "step": 28922 + }, + { + "epoch": 1.23, + "grad_norm": 0.45703125, + "learning_rate": 6.452648512659598e-05, + "loss": 2.429, + "step": 28923 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.452021351292568e-05, + "loss": 2.4693, + "step": 28924 + }, + { + "epoch": 1.23, + "grad_norm": 0.498046875, + "learning_rate": 6.45139420589015e-05, + "loss": 2.4218, + "step": 28925 + }, + { + "epoch": 1.23, + "grad_norm": 0.5078125, + "learning_rate": 6.450767076455162e-05, + "loss": 2.3952, + "step": 28926 + }, + { + "epoch": 1.23, + "grad_norm": 0.42578125, + "learning_rate": 6.450139962990423e-05, + "loss": 2.3858, + "step": 28927 + }, + { + "epoch": 1.23, + "grad_norm": 0.435546875, + "learning_rate": 6.449512865498754e-05, + "loss": 2.5536, + "step": 28928 + }, + { + "epoch": 1.23, + "grad_norm": 0.4921875, + "learning_rate": 6.448885783982985e-05, + "loss": 2.5837, + "step": 28929 + }, + { + "epoch": 1.23, + "grad_norm": 0.46484375, + "learning_rate": 6.44825871844593e-05, + "loss": 2.4418, + "step": 28930 + }, + { + "epoch": 1.23, + "grad_norm": 0.474609375, + "learning_rate": 6.447631668890414e-05, + "loss": 2.3726, + "step": 28931 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.447004635319252e-05, + "loss": 2.4384, + "step": 28932 + }, + { + "epoch": 1.23, + "grad_norm": 0.48046875, + "learning_rate": 6.446377617735277e-05, + "loss": 2.4054, + "step": 28933 + }, + { + "epoch": 1.23, + "grad_norm": 0.451171875, + "learning_rate": 6.445750616141302e-05, + "loss": 2.5556, + "step": 28934 + }, + { + "epoch": 1.23, + "grad_norm": 0.51953125, + "learning_rate": 6.445123630540151e-05, + "loss": 2.4953, + "step": 28935 + }, + { + "epoch": 1.23, + "grad_norm": 0.42578125, + "learning_rate": 6.44449666093464e-05, + "loss": 2.3942, + "step": 28936 + }, + { + "epoch": 1.23, + "grad_norm": 0.486328125, + "learning_rate": 6.443869707327598e-05, + "loss": 2.4284, + "step": 28937 + }, + { + "epoch": 1.23, + "grad_norm": 0.474609375, + "learning_rate": 6.443242769721844e-05, + "loss": 2.4811, + "step": 28938 + }, + { + "epoch": 1.23, + "grad_norm": 0.4765625, + "learning_rate": 6.4426158481202e-05, + "loss": 2.4668, + "step": 28939 + }, + { + "epoch": 1.23, + "grad_norm": 0.453125, + "learning_rate": 6.441988942525475e-05, + "loss": 2.5017, + "step": 28940 + }, + { + "epoch": 1.23, + "grad_norm": 0.4609375, + "learning_rate": 6.441362052940509e-05, + "loss": 2.4141, + "step": 28941 + }, + { + "epoch": 1.23, + "grad_norm": 0.462890625, + "learning_rate": 6.44073517936811e-05, + "loss": 2.4908, + "step": 28942 + }, + { + "epoch": 1.23, + "grad_norm": 0.478515625, + "learning_rate": 6.440108321811101e-05, + "loss": 2.484, + "step": 28943 + }, + { + "epoch": 1.23, + "grad_norm": 0.404296875, + "learning_rate": 6.439481480272302e-05, + "loss": 2.3682, + "step": 28944 + }, + { + "epoch": 1.23, + "grad_norm": 0.458984375, + "learning_rate": 6.438854654754537e-05, + "loss": 2.5067, + "step": 28945 + }, + { + "epoch": 1.23, + "grad_norm": 0.44140625, + "learning_rate": 6.438227845260626e-05, + "loss": 2.3582, + "step": 28946 + }, + { + "epoch": 1.23, + "grad_norm": 0.46484375, + "learning_rate": 6.437601051793389e-05, + "loss": 2.3767, + "step": 28947 + }, + { + "epoch": 1.23, + "grad_norm": 0.54296875, + "learning_rate": 6.43697427435564e-05, + "loss": 2.3515, + "step": 28948 + }, + { + "epoch": 1.23, + "grad_norm": 0.439453125, + "learning_rate": 6.436347512950209e-05, + "loss": 2.4105, + "step": 28949 + }, + { + "epoch": 1.23, + "grad_norm": 0.46484375, + "learning_rate": 6.435720767579912e-05, + "loss": 2.4713, + "step": 28950 + }, + { + "epoch": 1.23, + "grad_norm": 0.3984375, + "learning_rate": 6.435094038247567e-05, + "loss": 2.487, + "step": 28951 + }, + { + "epoch": 1.23, + "grad_norm": 0.48828125, + "learning_rate": 6.434467324955995e-05, + "loss": 2.4794, + "step": 28952 + }, + { + "epoch": 1.23, + "grad_norm": 0.4921875, + "learning_rate": 6.43384062770802e-05, + "loss": 2.644, + "step": 28953 + }, + { + "epoch": 1.23, + "grad_norm": 0.474609375, + "learning_rate": 6.433213946506457e-05, + "loss": 2.5098, + "step": 28954 + }, + { + "epoch": 1.23, + "grad_norm": 0.458984375, + "learning_rate": 6.432587281354131e-05, + "loss": 2.5426, + "step": 28955 + }, + { + "epoch": 1.23, + "grad_norm": 0.41796875, + "learning_rate": 6.431960632253851e-05, + "loss": 2.3876, + "step": 28956 + }, + { + "epoch": 1.23, + "grad_norm": 0.484375, + "learning_rate": 6.431333999208453e-05, + "loss": 2.4354, + "step": 28957 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.430707382220743e-05, + "loss": 2.4125, + "step": 28958 + }, + { + "epoch": 1.23, + "grad_norm": 0.486328125, + "learning_rate": 6.430080781293546e-05, + "loss": 2.319, + "step": 28959 + }, + { + "epoch": 1.23, + "grad_norm": 0.494140625, + "learning_rate": 6.42945419642968e-05, + "loss": 2.4601, + "step": 28960 + }, + { + "epoch": 1.23, + "grad_norm": 0.416015625, + "learning_rate": 6.428827627631966e-05, + "loss": 2.3402, + "step": 28961 + }, + { + "epoch": 1.23, + "grad_norm": 0.447265625, + "learning_rate": 6.428201074903223e-05, + "loss": 2.4756, + "step": 28962 + }, + { + "epoch": 1.23, + "grad_norm": 0.5, + "learning_rate": 6.427574538246273e-05, + "loss": 2.3907, + "step": 28963 + }, + { + "epoch": 1.23, + "grad_norm": 0.50390625, + "learning_rate": 6.426948017663923e-05, + "loss": 2.5703, + "step": 28964 + }, + { + "epoch": 1.23, + "grad_norm": 0.46875, + "learning_rate": 6.42632151315901e-05, + "loss": 2.4462, + "step": 28965 + }, + { + "epoch": 1.23, + "grad_norm": 0.431640625, + "learning_rate": 6.425695024734341e-05, + "loss": 2.5765, + "step": 28966 + }, + { + "epoch": 1.23, + "grad_norm": 0.458984375, + "learning_rate": 6.425068552392738e-05, + "loss": 2.5711, + "step": 28967 + }, + { + "epoch": 1.23, + "grad_norm": 0.455078125, + "learning_rate": 6.424442096137018e-05, + "loss": 2.4969, + "step": 28968 + }, + { + "epoch": 1.23, + "grad_norm": 0.404296875, + "learning_rate": 6.423815655970005e-05, + "loss": 2.4223, + "step": 28969 + }, + { + "epoch": 1.23, + "grad_norm": 0.5546875, + "learning_rate": 6.423189231894514e-05, + "loss": 2.5025, + "step": 28970 + }, + { + "epoch": 1.23, + "grad_norm": 0.431640625, + "learning_rate": 6.422562823913365e-05, + "loss": 2.5103, + "step": 28971 + }, + { + "epoch": 1.23, + "grad_norm": 0.421875, + "learning_rate": 6.42193643202937e-05, + "loss": 2.313, + "step": 28972 + }, + { + "epoch": 1.23, + "grad_norm": 0.515625, + "learning_rate": 6.42131005624536e-05, + "loss": 2.394, + "step": 28973 + }, + { + "epoch": 1.23, + "grad_norm": 0.470703125, + "learning_rate": 6.420683696564148e-05, + "loss": 2.3357, + "step": 28974 + }, + { + "epoch": 1.23, + "grad_norm": 0.4921875, + "learning_rate": 6.420057352988546e-05, + "loss": 2.4456, + "step": 28975 + }, + { + "epoch": 1.23, + "grad_norm": 0.462890625, + "learning_rate": 6.419431025521379e-05, + "loss": 2.4708, + "step": 28976 + }, + { + "epoch": 1.23, + "grad_norm": 0.4765625, + "learning_rate": 6.418804714165465e-05, + "loss": 2.5831, + "step": 28977 + }, + { + "epoch": 1.23, + "grad_norm": 0.4453125, + "learning_rate": 6.41817841892362e-05, + "loss": 2.3579, + "step": 28978 + }, + { + "epoch": 1.23, + "grad_norm": 0.439453125, + "learning_rate": 6.417552139798666e-05, + "loss": 2.4557, + "step": 28979 + }, + { + "epoch": 1.23, + "grad_norm": 0.46484375, + "learning_rate": 6.416925876793417e-05, + "loss": 2.3957, + "step": 28980 + }, + { + "epoch": 1.23, + "grad_norm": 0.48046875, + "learning_rate": 6.41629962991069e-05, + "loss": 2.4602, + "step": 28981 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.415673399153305e-05, + "loss": 2.4389, + "step": 28982 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.415047184524082e-05, + "loss": 2.5321, + "step": 28983 + }, + { + "epoch": 1.24, + "grad_norm": 0.439453125, + "learning_rate": 6.414420986025835e-05, + "loss": 2.4476, + "step": 28984 + }, + { + "epoch": 1.24, + "grad_norm": 0.48046875, + "learning_rate": 6.413794803661381e-05, + "loss": 2.4526, + "step": 28985 + }, + { + "epoch": 1.24, + "grad_norm": 0.478515625, + "learning_rate": 6.413168637433542e-05, + "loss": 2.3886, + "step": 28986 + }, + { + "epoch": 1.24, + "grad_norm": 0.474609375, + "learning_rate": 6.412542487345136e-05, + "loss": 2.4287, + "step": 28987 + }, + { + "epoch": 1.24, + "grad_norm": 0.4375, + "learning_rate": 6.411916353398976e-05, + "loss": 2.4697, + "step": 28988 + }, + { + "epoch": 1.24, + "grad_norm": 0.462890625, + "learning_rate": 6.411290235597878e-05, + "loss": 2.4971, + "step": 28989 + }, + { + "epoch": 1.24, + "grad_norm": 0.455078125, + "learning_rate": 6.410664133944666e-05, + "loss": 2.3821, + "step": 28990 + }, + { + "epoch": 1.24, + "grad_norm": 0.43359375, + "learning_rate": 6.410038048442153e-05, + "loss": 2.4534, + "step": 28991 + }, + { + "epoch": 1.24, + "grad_norm": 0.458984375, + "learning_rate": 6.409411979093156e-05, + "loss": 2.6368, + "step": 28992 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.408785925900492e-05, + "loss": 2.4422, + "step": 28993 + }, + { + "epoch": 1.24, + "grad_norm": 0.48828125, + "learning_rate": 6.408159888866979e-05, + "loss": 2.4717, + "step": 28994 + }, + { + "epoch": 1.24, + "grad_norm": 0.462890625, + "learning_rate": 6.407533867995436e-05, + "loss": 2.4484, + "step": 28995 + }, + { + "epoch": 1.24, + "grad_norm": 0.474609375, + "learning_rate": 6.406907863288678e-05, + "loss": 2.4248, + "step": 28996 + }, + { + "epoch": 1.24, + "grad_norm": 0.58203125, + "learning_rate": 6.406281874749517e-05, + "loss": 2.5229, + "step": 28997 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.405655902380778e-05, + "loss": 2.4177, + "step": 28998 + }, + { + "epoch": 1.24, + "grad_norm": 0.482421875, + "learning_rate": 6.405029946185275e-05, + "loss": 2.4899, + "step": 28999 + }, + { + "epoch": 1.24, + "grad_norm": 0.404296875, + "learning_rate": 6.40440400616582e-05, + "loss": 2.4668, + "step": 29000 + }, + { + "epoch": 1.24, + "grad_norm": 0.51171875, + "learning_rate": 6.403778082325234e-05, + "loss": 2.4374, + "step": 29001 + }, + { + "epoch": 1.24, + "grad_norm": 0.498046875, + "learning_rate": 6.403152174666333e-05, + "loss": 2.5348, + "step": 29002 + }, + { + "epoch": 1.24, + "grad_norm": 0.494140625, + "learning_rate": 6.402526283191934e-05, + "loss": 2.589, + "step": 29003 + }, + { + "epoch": 1.24, + "grad_norm": 0.416015625, + "learning_rate": 6.401900407904851e-05, + "loss": 2.5125, + "step": 29004 + }, + { + "epoch": 1.24, + "grad_norm": 0.484375, + "learning_rate": 6.401274548807898e-05, + "loss": 2.4784, + "step": 29005 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.400648705903898e-05, + "loss": 2.4563, + "step": 29006 + }, + { + "epoch": 1.24, + "grad_norm": 0.447265625, + "learning_rate": 6.400022879195664e-05, + "loss": 2.4674, + "step": 29007 + }, + { + "epoch": 1.24, + "grad_norm": 0.494140625, + "learning_rate": 6.39939706868601e-05, + "loss": 2.4561, + "step": 29008 + }, + { + "epoch": 1.24, + "grad_norm": 0.5078125, + "learning_rate": 6.39877127437775e-05, + "loss": 2.6112, + "step": 29009 + }, + { + "epoch": 1.24, + "grad_norm": 0.5703125, + "learning_rate": 6.398145496273708e-05, + "loss": 2.5022, + "step": 29010 + }, + { + "epoch": 1.24, + "grad_norm": 0.50390625, + "learning_rate": 6.397519734376696e-05, + "loss": 2.6253, + "step": 29011 + }, + { + "epoch": 1.24, + "grad_norm": 0.4609375, + "learning_rate": 6.396893988689525e-05, + "loss": 2.4712, + "step": 29012 + }, + { + "epoch": 1.24, + "grad_norm": 0.484375, + "learning_rate": 6.396268259215013e-05, + "loss": 2.4807, + "step": 29013 + }, + { + "epoch": 1.24, + "grad_norm": 0.5234375, + "learning_rate": 6.39564254595598e-05, + "loss": 2.5016, + "step": 29014 + }, + { + "epoch": 1.24, + "grad_norm": 0.462890625, + "learning_rate": 6.395016848915238e-05, + "loss": 2.5524, + "step": 29015 + }, + { + "epoch": 1.24, + "grad_norm": 0.447265625, + "learning_rate": 6.394391168095601e-05, + "loss": 2.4259, + "step": 29016 + }, + { + "epoch": 1.24, + "grad_norm": 0.4921875, + "learning_rate": 6.393765503499887e-05, + "loss": 2.5069, + "step": 29017 + }, + { + "epoch": 1.24, + "grad_norm": 0.484375, + "learning_rate": 6.39313985513091e-05, + "loss": 2.4618, + "step": 29018 + }, + { + "epoch": 1.24, + "grad_norm": 0.462890625, + "learning_rate": 6.392514222991488e-05, + "loss": 2.4359, + "step": 29019 + }, + { + "epoch": 1.24, + "grad_norm": 0.4609375, + "learning_rate": 6.391888607084432e-05, + "loss": 2.6065, + "step": 29020 + }, + { + "epoch": 1.24, + "grad_norm": 0.515625, + "learning_rate": 6.391263007412556e-05, + "loss": 2.3686, + "step": 29021 + }, + { + "epoch": 1.24, + "grad_norm": 0.478515625, + "learning_rate": 6.39063742397868e-05, + "loss": 2.5052, + "step": 29022 + }, + { + "epoch": 1.24, + "grad_norm": 0.416015625, + "learning_rate": 6.390011856785617e-05, + "loss": 2.3422, + "step": 29023 + }, + { + "epoch": 1.24, + "grad_norm": 0.52734375, + "learning_rate": 6.38938630583618e-05, + "loss": 2.4619, + "step": 29024 + }, + { + "epoch": 1.24, + "grad_norm": 0.5, + "learning_rate": 6.388760771133184e-05, + "loss": 2.5016, + "step": 29025 + }, + { + "epoch": 1.24, + "grad_norm": 0.46484375, + "learning_rate": 6.388135252679448e-05, + "loss": 2.4938, + "step": 29026 + }, + { + "epoch": 1.24, + "grad_norm": 0.5390625, + "learning_rate": 6.387509750477783e-05, + "loss": 2.4631, + "step": 29027 + }, + { + "epoch": 1.24, + "grad_norm": 0.482421875, + "learning_rate": 6.386884264531003e-05, + "loss": 2.4702, + "step": 29028 + }, + { + "epoch": 1.24, + "grad_norm": 0.462890625, + "learning_rate": 6.386258794841922e-05, + "loss": 2.5261, + "step": 29029 + }, + { + "epoch": 1.24, + "grad_norm": 0.455078125, + "learning_rate": 6.385633341413358e-05, + "loss": 2.432, + "step": 29030 + }, + { + "epoch": 1.24, + "grad_norm": 0.494140625, + "learning_rate": 6.385007904248121e-05, + "loss": 2.4717, + "step": 29031 + }, + { + "epoch": 1.24, + "grad_norm": 0.4375, + "learning_rate": 6.384382483349029e-05, + "loss": 2.5238, + "step": 29032 + }, + { + "epoch": 1.24, + "grad_norm": 0.5234375, + "learning_rate": 6.383757078718892e-05, + "loss": 2.4461, + "step": 29033 + }, + { + "epoch": 1.24, + "grad_norm": 0.4296875, + "learning_rate": 6.383131690360529e-05, + "loss": 2.5404, + "step": 29034 + }, + { + "epoch": 1.24, + "grad_norm": 0.427734375, + "learning_rate": 6.382506318276754e-05, + "loss": 2.4181, + "step": 29035 + }, + { + "epoch": 1.24, + "grad_norm": 0.462890625, + "learning_rate": 6.381880962470375e-05, + "loss": 2.6147, + "step": 29036 + }, + { + "epoch": 1.24, + "grad_norm": 0.4921875, + "learning_rate": 6.381255622944208e-05, + "loss": 2.4687, + "step": 29037 + }, + { + "epoch": 1.24, + "grad_norm": 0.470703125, + "learning_rate": 6.38063029970107e-05, + "loss": 2.4546, + "step": 29038 + }, + { + "epoch": 1.24, + "grad_norm": 0.5234375, + "learning_rate": 6.380004992743773e-05, + "loss": 2.5145, + "step": 29039 + }, + { + "epoch": 1.24, + "grad_norm": 0.50390625, + "learning_rate": 6.379379702075131e-05, + "loss": 2.5571, + "step": 29040 + }, + { + "epoch": 1.24, + "grad_norm": 0.43359375, + "learning_rate": 6.378754427697956e-05, + "loss": 2.585, + "step": 29041 + }, + { + "epoch": 1.24, + "grad_norm": 0.44140625, + "learning_rate": 6.378129169615062e-05, + "loss": 2.554, + "step": 29042 + }, + { + "epoch": 1.24, + "grad_norm": 0.51953125, + "learning_rate": 6.377503927829267e-05, + "loss": 2.3702, + "step": 29043 + }, + { + "epoch": 1.24, + "grad_norm": 0.470703125, + "learning_rate": 6.376878702343379e-05, + "loss": 2.6109, + "step": 29044 + }, + { + "epoch": 1.24, + "grad_norm": 0.5390625, + "learning_rate": 6.37625349316021e-05, + "loss": 2.482, + "step": 29045 + }, + { + "epoch": 1.24, + "grad_norm": 0.470703125, + "learning_rate": 6.375628300282577e-05, + "loss": 2.6068, + "step": 29046 + }, + { + "epoch": 1.24, + "grad_norm": 0.447265625, + "learning_rate": 6.375003123713292e-05, + "loss": 2.3876, + "step": 29047 + }, + { + "epoch": 1.24, + "grad_norm": 0.474609375, + "learning_rate": 6.37437796345517e-05, + "loss": 2.4206, + "step": 29048 + }, + { + "epoch": 1.24, + "grad_norm": 0.4453125, + "learning_rate": 6.373752819511018e-05, + "loss": 2.3795, + "step": 29049 + }, + { + "epoch": 1.24, + "grad_norm": 0.427734375, + "learning_rate": 6.373127691883658e-05, + "loss": 2.5277, + "step": 29050 + }, + { + "epoch": 1.24, + "grad_norm": 0.44140625, + "learning_rate": 6.372502580575898e-05, + "loss": 2.4526, + "step": 29051 + }, + { + "epoch": 1.24, + "grad_norm": 0.50390625, + "learning_rate": 6.371877485590549e-05, + "loss": 2.4663, + "step": 29052 + }, + { + "epoch": 1.24, + "grad_norm": 0.447265625, + "learning_rate": 6.371252406930423e-05, + "loss": 2.4355, + "step": 29053 + }, + { + "epoch": 1.24, + "grad_norm": 0.458984375, + "learning_rate": 6.370627344598336e-05, + "loss": 2.3819, + "step": 29054 + }, + { + "epoch": 1.24, + "grad_norm": 0.466796875, + "learning_rate": 6.370002298597101e-05, + "loss": 2.4971, + "step": 29055 + }, + { + "epoch": 1.24, + "grad_norm": 0.515625, + "learning_rate": 6.369377268929531e-05, + "loss": 2.514, + "step": 29056 + }, + { + "epoch": 1.24, + "grad_norm": 0.4765625, + "learning_rate": 6.368752255598431e-05, + "loss": 2.5292, + "step": 29057 + }, + { + "epoch": 1.24, + "grad_norm": 0.478515625, + "learning_rate": 6.368127258606622e-05, + "loss": 2.4124, + "step": 29058 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.367502277956915e-05, + "loss": 2.4439, + "step": 29059 + }, + { + "epoch": 1.24, + "grad_norm": 0.51171875, + "learning_rate": 6.36687731365212e-05, + "loss": 2.5466, + "step": 29060 + }, + { + "epoch": 1.24, + "grad_norm": 0.48828125, + "learning_rate": 6.366252365695044e-05, + "loss": 2.4288, + "step": 29061 + }, + { + "epoch": 1.24, + "grad_norm": 0.45703125, + "learning_rate": 6.36562743408851e-05, + "loss": 2.3793, + "step": 29062 + }, + { + "epoch": 1.24, + "grad_norm": 0.482421875, + "learning_rate": 6.365002518835321e-05, + "loss": 2.4731, + "step": 29063 + }, + { + "epoch": 1.24, + "grad_norm": 0.486328125, + "learning_rate": 6.364377619938294e-05, + "loss": 2.4133, + "step": 29064 + }, + { + "epoch": 1.24, + "grad_norm": 0.48046875, + "learning_rate": 6.363752737400237e-05, + "loss": 2.449, + "step": 29065 + }, + { + "epoch": 1.24, + "grad_norm": 0.41796875, + "learning_rate": 6.363127871223966e-05, + "loss": 2.5141, + "step": 29066 + }, + { + "epoch": 1.24, + "grad_norm": 0.51953125, + "learning_rate": 6.362503021412291e-05, + "loss": 2.5866, + "step": 29067 + }, + { + "epoch": 1.24, + "grad_norm": 0.51953125, + "learning_rate": 6.361878187968022e-05, + "loss": 2.4195, + "step": 29068 + }, + { + "epoch": 1.24, + "grad_norm": 0.51171875, + "learning_rate": 6.36125337089397e-05, + "loss": 2.4352, + "step": 29069 + }, + { + "epoch": 1.24, + "grad_norm": 0.427734375, + "learning_rate": 6.360628570192951e-05, + "loss": 2.4321, + "step": 29070 + }, + { + "epoch": 1.24, + "grad_norm": 0.423828125, + "learning_rate": 6.360003785867771e-05, + "loss": 2.2652, + "step": 29071 + }, + { + "epoch": 1.24, + "grad_norm": 0.458984375, + "learning_rate": 6.359379017921244e-05, + "loss": 2.3785, + "step": 29072 + }, + { + "epoch": 1.24, + "grad_norm": 0.5234375, + "learning_rate": 6.358754266356179e-05, + "loss": 2.5303, + "step": 29073 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.358129531175392e-05, + "loss": 2.4588, + "step": 29074 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.357504812381694e-05, + "loss": 2.493, + "step": 29075 + }, + { + "epoch": 1.24, + "grad_norm": 0.404296875, + "learning_rate": 6.356880109977889e-05, + "loss": 2.3983, + "step": 29076 + }, + { + "epoch": 1.24, + "grad_norm": 0.451171875, + "learning_rate": 6.356255423966791e-05, + "loss": 2.5128, + "step": 29077 + }, + { + "epoch": 1.24, + "grad_norm": 0.498046875, + "learning_rate": 6.355630754351215e-05, + "loss": 2.4741, + "step": 29078 + }, + { + "epoch": 1.24, + "grad_norm": 0.5, + "learning_rate": 6.355006101133968e-05, + "loss": 2.5496, + "step": 29079 + }, + { + "epoch": 1.24, + "grad_norm": 0.5, + "learning_rate": 6.35438146431786e-05, + "loss": 2.4838, + "step": 29080 + }, + { + "epoch": 1.24, + "grad_norm": 0.41796875, + "learning_rate": 6.353756843905703e-05, + "loss": 2.4414, + "step": 29081 + }, + { + "epoch": 1.24, + "grad_norm": 0.48046875, + "learning_rate": 6.35313223990031e-05, + "loss": 2.3602, + "step": 29082 + }, + { + "epoch": 1.24, + "grad_norm": 0.4921875, + "learning_rate": 6.352507652304488e-05, + "loss": 2.5428, + "step": 29083 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.35188308112105e-05, + "loss": 2.5957, + "step": 29084 + }, + { + "epoch": 1.24, + "grad_norm": 0.478515625, + "learning_rate": 6.351258526352801e-05, + "loss": 2.4524, + "step": 29085 + }, + { + "epoch": 1.24, + "grad_norm": 0.4609375, + "learning_rate": 6.350633988002558e-05, + "loss": 2.3579, + "step": 29086 + }, + { + "epoch": 1.24, + "grad_norm": 0.4296875, + "learning_rate": 6.350009466073128e-05, + "loss": 2.4803, + "step": 29087 + }, + { + "epoch": 1.24, + "grad_norm": 0.41796875, + "learning_rate": 6.349384960567322e-05, + "loss": 2.4552, + "step": 29088 + }, + { + "epoch": 1.24, + "grad_norm": 0.5078125, + "learning_rate": 6.348760471487947e-05, + "loss": 2.3589, + "step": 29089 + }, + { + "epoch": 1.24, + "grad_norm": 0.4296875, + "learning_rate": 6.348135998837819e-05, + "loss": 2.5744, + "step": 29090 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.347511542619743e-05, + "loss": 2.3656, + "step": 29091 + }, + { + "epoch": 1.24, + "grad_norm": 0.58984375, + "learning_rate": 6.346887102836532e-05, + "loss": 2.5928, + "step": 29092 + }, + { + "epoch": 1.24, + "grad_norm": 0.443359375, + "learning_rate": 6.346262679490989e-05, + "loss": 2.5541, + "step": 29093 + }, + { + "epoch": 1.24, + "grad_norm": 0.45703125, + "learning_rate": 6.345638272585932e-05, + "loss": 2.4326, + "step": 29094 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.345013882124168e-05, + "loss": 2.5129, + "step": 29095 + }, + { + "epoch": 1.24, + "grad_norm": 0.5, + "learning_rate": 6.344389508108505e-05, + "loss": 2.5625, + "step": 29096 + }, + { + "epoch": 1.24, + "grad_norm": 0.486328125, + "learning_rate": 6.343765150541752e-05, + "loss": 2.5217, + "step": 29097 + }, + { + "epoch": 1.24, + "grad_norm": 0.470703125, + "learning_rate": 6.34314080942672e-05, + "loss": 2.4016, + "step": 29098 + }, + { + "epoch": 1.24, + "grad_norm": 0.47265625, + "learning_rate": 6.342516484766221e-05, + "loss": 2.5465, + "step": 29099 + }, + { + "epoch": 1.24, + "grad_norm": 0.44921875, + "learning_rate": 6.341892176563057e-05, + "loss": 2.4295, + "step": 29100 + }, + { + "epoch": 1.24, + "grad_norm": 0.490234375, + "learning_rate": 6.341267884820044e-05, + "loss": 2.3889, + "step": 29101 + }, + { + "epoch": 1.24, + "grad_norm": 0.4765625, + "learning_rate": 6.340643609539987e-05, + "loss": 2.4809, + "step": 29102 + }, + { + "epoch": 1.24, + "grad_norm": 0.4765625, + "learning_rate": 6.340019350725697e-05, + "loss": 2.4232, + "step": 29103 + }, + { + "epoch": 1.24, + "grad_norm": 0.44921875, + "learning_rate": 6.339395108379982e-05, + "loss": 2.4979, + "step": 29104 + }, + { + "epoch": 1.24, + "grad_norm": 0.451171875, + "learning_rate": 6.338770882505653e-05, + "loss": 2.4754, + "step": 29105 + }, + { + "epoch": 1.24, + "grad_norm": 0.474609375, + "learning_rate": 6.338146673105511e-05, + "loss": 2.5253, + "step": 29106 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.337522480182379e-05, + "loss": 2.3609, + "step": 29107 + }, + { + "epoch": 1.24, + "grad_norm": 0.42578125, + "learning_rate": 6.336898303739054e-05, + "loss": 2.4167, + "step": 29108 + }, + { + "epoch": 1.24, + "grad_norm": 0.50390625, + "learning_rate": 6.336274143778346e-05, + "loss": 2.446, + "step": 29109 + }, + { + "epoch": 1.24, + "grad_norm": 0.50390625, + "learning_rate": 6.335650000303066e-05, + "loss": 2.4652, + "step": 29110 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.335025873316023e-05, + "loss": 2.4407, + "step": 29111 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.334401762820023e-05, + "loss": 2.541, + "step": 29112 + }, + { + "epoch": 1.24, + "grad_norm": 0.435546875, + "learning_rate": 6.33377766881788e-05, + "loss": 2.3006, + "step": 29113 + }, + { + "epoch": 1.24, + "grad_norm": 0.42578125, + "learning_rate": 6.333153591312389e-05, + "loss": 2.4015, + "step": 29114 + }, + { + "epoch": 1.24, + "grad_norm": 0.50390625, + "learning_rate": 6.332529530306375e-05, + "loss": 2.5208, + "step": 29115 + }, + { + "epoch": 1.24, + "grad_norm": 0.4921875, + "learning_rate": 6.331905485802633e-05, + "loss": 2.5055, + "step": 29116 + }, + { + "epoch": 1.24, + "grad_norm": 0.49609375, + "learning_rate": 6.331281457803979e-05, + "loss": 2.3117, + "step": 29117 + }, + { + "epoch": 1.24, + "grad_norm": 0.40234375, + "learning_rate": 6.330657446313213e-05, + "loss": 2.4115, + "step": 29118 + }, + { + "epoch": 1.24, + "grad_norm": 0.43359375, + "learning_rate": 6.330033451333151e-05, + "loss": 2.4959, + "step": 29119 + }, + { + "epoch": 1.24, + "grad_norm": 0.466796875, + "learning_rate": 6.329409472866596e-05, + "loss": 2.3102, + "step": 29120 + }, + { + "epoch": 1.24, + "grad_norm": 0.5234375, + "learning_rate": 6.32878551091636e-05, + "loss": 2.4288, + "step": 29121 + }, + { + "epoch": 1.24, + "grad_norm": 0.482421875, + "learning_rate": 6.328161565485242e-05, + "loss": 2.5181, + "step": 29122 + }, + { + "epoch": 1.24, + "grad_norm": 0.56640625, + "learning_rate": 6.327537636576061e-05, + "loss": 2.5047, + "step": 29123 + }, + { + "epoch": 1.24, + "grad_norm": 0.4140625, + "learning_rate": 6.326913724191615e-05, + "loss": 2.3732, + "step": 29124 + }, + { + "epoch": 1.24, + "grad_norm": 0.494140625, + "learning_rate": 6.326289828334717e-05, + "loss": 2.3823, + "step": 29125 + }, + { + "epoch": 1.24, + "grad_norm": 0.43359375, + "learning_rate": 6.325665949008168e-05, + "loss": 2.3808, + "step": 29126 + }, + { + "epoch": 1.24, + "grad_norm": 0.4921875, + "learning_rate": 6.325042086214783e-05, + "loss": 2.5299, + "step": 29127 + }, + { + "epoch": 1.24, + "grad_norm": 0.435546875, + "learning_rate": 6.324418239957364e-05, + "loss": 2.4407, + "step": 29128 + }, + { + "epoch": 1.24, + "grad_norm": 0.431640625, + "learning_rate": 6.323794410238724e-05, + "loss": 2.4937, + "step": 29129 + }, + { + "epoch": 1.24, + "grad_norm": 0.48828125, + "learning_rate": 6.323170597061658e-05, + "loss": 2.5774, + "step": 29130 + }, + { + "epoch": 1.24, + "grad_norm": 0.48828125, + "learning_rate": 6.322546800428985e-05, + "loss": 2.3964, + "step": 29131 + }, + { + "epoch": 1.24, + "grad_norm": 0.40625, + "learning_rate": 6.321923020343508e-05, + "loss": 2.2079, + "step": 29132 + }, + { + "epoch": 1.24, + "grad_norm": 0.451171875, + "learning_rate": 6.321299256808033e-05, + "loss": 2.3961, + "step": 29133 + }, + { + "epoch": 1.24, + "grad_norm": 0.40234375, + "learning_rate": 6.320675509825365e-05, + "loss": 2.3765, + "step": 29134 + }, + { + "epoch": 1.24, + "grad_norm": 0.482421875, + "learning_rate": 6.320051779398314e-05, + "loss": 2.5061, + "step": 29135 + }, + { + "epoch": 1.24, + "grad_norm": 0.48828125, + "learning_rate": 6.319428065529685e-05, + "loss": 2.3461, + "step": 29136 + }, + { + "epoch": 1.24, + "grad_norm": 0.474609375, + "learning_rate": 6.318804368222287e-05, + "loss": 2.3719, + "step": 29137 + }, + { + "epoch": 1.24, + "grad_norm": 0.482421875, + "learning_rate": 6.318180687478917e-05, + "loss": 2.5193, + "step": 29138 + }, + { + "epoch": 1.24, + "grad_norm": 0.431640625, + "learning_rate": 6.317557023302396e-05, + "loss": 2.4468, + "step": 29139 + }, + { + "epoch": 1.24, + "grad_norm": 0.423828125, + "learning_rate": 6.31693337569552e-05, + "loss": 2.4824, + "step": 29140 + }, + { + "epoch": 1.24, + "grad_norm": 0.4296875, + "learning_rate": 6.316309744661098e-05, + "loss": 2.3168, + "step": 29141 + }, + { + "epoch": 1.24, + "grad_norm": 0.44921875, + "learning_rate": 6.315686130201932e-05, + "loss": 2.4906, + "step": 29142 + }, + { + "epoch": 1.24, + "grad_norm": 0.388671875, + "learning_rate": 6.315062532320835e-05, + "loss": 2.3513, + "step": 29143 + }, + { + "epoch": 1.24, + "grad_norm": 0.455078125, + "learning_rate": 6.314438951020612e-05, + "loss": 2.4636, + "step": 29144 + }, + { + "epoch": 1.24, + "grad_norm": 0.435546875, + "learning_rate": 6.313815386304067e-05, + "loss": 2.4818, + "step": 29145 + }, + { + "epoch": 1.24, + "grad_norm": 0.55078125, + "learning_rate": 6.313191838173999e-05, + "loss": 2.6147, + "step": 29146 + }, + { + "epoch": 1.24, + "grad_norm": 0.435546875, + "learning_rate": 6.312568306633226e-05, + "loss": 2.4314, + "step": 29147 + }, + { + "epoch": 1.24, + "grad_norm": 0.431640625, + "learning_rate": 6.311944791684548e-05, + "loss": 2.3625, + "step": 29148 + }, + { + "epoch": 1.24, + "grad_norm": 0.419921875, + "learning_rate": 6.31132129333077e-05, + "loss": 2.5134, + "step": 29149 + }, + { + "epoch": 1.24, + "grad_norm": 0.490234375, + "learning_rate": 6.310697811574695e-05, + "loss": 2.4636, + "step": 29150 + }, + { + "epoch": 1.24, + "grad_norm": 0.41796875, + "learning_rate": 6.310074346419135e-05, + "loss": 2.4334, + "step": 29151 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.30945089786689e-05, + "loss": 2.3889, + "step": 29152 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.308827465920772e-05, + "loss": 2.601, + "step": 29153 + }, + { + "epoch": 1.24, + "grad_norm": 0.404296875, + "learning_rate": 6.308204050583574e-05, + "loss": 2.3046, + "step": 29154 + }, + { + "epoch": 1.24, + "grad_norm": 0.466796875, + "learning_rate": 6.307580651858114e-05, + "loss": 2.4769, + "step": 29155 + }, + { + "epoch": 1.24, + "grad_norm": 0.462890625, + "learning_rate": 6.306957269747191e-05, + "loss": 2.47, + "step": 29156 + }, + { + "epoch": 1.24, + "grad_norm": 0.439453125, + "learning_rate": 6.30633390425361e-05, + "loss": 2.5653, + "step": 29157 + }, + { + "epoch": 1.24, + "grad_norm": 0.46484375, + "learning_rate": 6.305710555380173e-05, + "loss": 2.3819, + "step": 29158 + }, + { + "epoch": 1.24, + "grad_norm": 0.49609375, + "learning_rate": 6.305087223129693e-05, + "loss": 2.4903, + "step": 29159 + }, + { + "epoch": 1.24, + "grad_norm": 0.45703125, + "learning_rate": 6.304463907504968e-05, + "loss": 2.4566, + "step": 29160 + }, + { + "epoch": 1.24, + "grad_norm": 0.4609375, + "learning_rate": 6.303840608508808e-05, + "loss": 2.4342, + "step": 29161 + }, + { + "epoch": 1.24, + "grad_norm": 0.466796875, + "learning_rate": 6.303217326144008e-05, + "loss": 2.4139, + "step": 29162 + }, + { + "epoch": 1.24, + "grad_norm": 0.4140625, + "learning_rate": 6.302594060413385e-05, + "loss": 2.4124, + "step": 29163 + }, + { + "epoch": 1.24, + "grad_norm": 0.478515625, + "learning_rate": 6.301970811319737e-05, + "loss": 2.4753, + "step": 29164 + }, + { + "epoch": 1.24, + "grad_norm": 0.3828125, + "learning_rate": 6.301347578865868e-05, + "loss": 2.4144, + "step": 29165 + }, + { + "epoch": 1.24, + "grad_norm": 0.494140625, + "learning_rate": 6.300724363054579e-05, + "loss": 2.4629, + "step": 29166 + }, + { + "epoch": 1.24, + "grad_norm": 0.47265625, + "learning_rate": 6.300101163888684e-05, + "loss": 2.5495, + "step": 29167 + }, + { + "epoch": 1.24, + "grad_norm": 0.447265625, + "learning_rate": 6.299477981370978e-05, + "loss": 2.3881, + "step": 29168 + }, + { + "epoch": 1.24, + "grad_norm": 0.46484375, + "learning_rate": 6.298854815504273e-05, + "loss": 2.5724, + "step": 29169 + }, + { + "epoch": 1.24, + "grad_norm": 0.484375, + "learning_rate": 6.298231666291364e-05, + "loss": 2.4376, + "step": 29170 + }, + { + "epoch": 1.24, + "grad_norm": 0.515625, + "learning_rate": 6.297608533735062e-05, + "loss": 2.4053, + "step": 29171 + }, + { + "epoch": 1.24, + "grad_norm": 0.4609375, + "learning_rate": 6.29698541783817e-05, + "loss": 2.4138, + "step": 29172 + }, + { + "epoch": 1.24, + "grad_norm": 0.4453125, + "learning_rate": 6.296362318603488e-05, + "loss": 2.5022, + "step": 29173 + }, + { + "epoch": 1.24, + "grad_norm": 0.5, + "learning_rate": 6.29573923603382e-05, + "loss": 2.5177, + "step": 29174 + }, + { + "epoch": 1.24, + "grad_norm": 0.453125, + "learning_rate": 6.295116170131975e-05, + "loss": 2.3797, + "step": 29175 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.294493120900753e-05, + "loss": 2.4491, + "step": 29176 + }, + { + "epoch": 1.24, + "grad_norm": 0.45703125, + "learning_rate": 6.29387008834296e-05, + "loss": 2.5221, + "step": 29177 + }, + { + "epoch": 1.24, + "grad_norm": 0.5078125, + "learning_rate": 6.29324707246139e-05, + "loss": 2.4534, + "step": 29178 + }, + { + "epoch": 1.24, + "grad_norm": 0.486328125, + "learning_rate": 6.292624073258861e-05, + "loss": 2.3479, + "step": 29179 + }, + { + "epoch": 1.24, + "grad_norm": 0.5, + "learning_rate": 6.292001090738166e-05, + "loss": 2.4611, + "step": 29180 + }, + { + "epoch": 1.24, + "grad_norm": 0.40234375, + "learning_rate": 6.29137812490211e-05, + "loss": 2.5555, + "step": 29181 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.290755175753495e-05, + "loss": 2.5254, + "step": 29182 + }, + { + "epoch": 1.24, + "grad_norm": 0.50390625, + "learning_rate": 6.290132243295131e-05, + "loss": 2.4496, + "step": 29183 + }, + { + "epoch": 1.24, + "grad_norm": 0.4296875, + "learning_rate": 6.289509327529814e-05, + "loss": 2.5699, + "step": 29184 + }, + { + "epoch": 1.24, + "grad_norm": 0.490234375, + "learning_rate": 6.28888642846035e-05, + "loss": 2.4609, + "step": 29185 + }, + { + "epoch": 1.24, + "grad_norm": 0.484375, + "learning_rate": 6.288263546089537e-05, + "loss": 2.4401, + "step": 29186 + }, + { + "epoch": 1.24, + "grad_norm": 0.486328125, + "learning_rate": 6.287640680420186e-05, + "loss": 2.5164, + "step": 29187 + }, + { + "epoch": 1.24, + "grad_norm": 0.4140625, + "learning_rate": 6.287017831455095e-05, + "loss": 2.4883, + "step": 29188 + }, + { + "epoch": 1.24, + "grad_norm": 0.44921875, + "learning_rate": 6.286394999197065e-05, + "loss": 2.4785, + "step": 29189 + }, + { + "epoch": 1.24, + "grad_norm": 0.5234375, + "learning_rate": 6.285772183648899e-05, + "loss": 2.5775, + "step": 29190 + }, + { + "epoch": 1.24, + "grad_norm": 0.5078125, + "learning_rate": 6.285149384813403e-05, + "loss": 2.3787, + "step": 29191 + }, + { + "epoch": 1.24, + "grad_norm": 0.44921875, + "learning_rate": 6.284526602693377e-05, + "loss": 2.4039, + "step": 29192 + }, + { + "epoch": 1.24, + "grad_norm": 0.404296875, + "learning_rate": 6.283903837291625e-05, + "loss": 2.3864, + "step": 29193 + }, + { + "epoch": 1.24, + "grad_norm": 0.4921875, + "learning_rate": 6.283281088610942e-05, + "loss": 2.614, + "step": 29194 + }, + { + "epoch": 1.24, + "grad_norm": 0.49609375, + "learning_rate": 6.282658356654143e-05, + "loss": 2.4533, + "step": 29195 + }, + { + "epoch": 1.24, + "grad_norm": 0.46875, + "learning_rate": 6.282035641424019e-05, + "loss": 2.4067, + "step": 29196 + }, + { + "epoch": 1.24, + "grad_norm": 0.416015625, + "learning_rate": 6.281412942923377e-05, + "loss": 2.5782, + "step": 29197 + }, + { + "epoch": 1.24, + "grad_norm": 0.5078125, + "learning_rate": 6.280790261155016e-05, + "loss": 2.452, + "step": 29198 + }, + { + "epoch": 1.24, + "grad_norm": 0.462890625, + "learning_rate": 6.28016759612174e-05, + "loss": 2.4175, + "step": 29199 + }, + { + "epoch": 1.24, + "grad_norm": 0.48046875, + "learning_rate": 6.279544947826351e-05, + "loss": 2.3474, + "step": 29200 + }, + { + "epoch": 1.24, + "grad_norm": 0.43359375, + "learning_rate": 6.278922316271652e-05, + "loss": 2.3555, + "step": 29201 + }, + { + "epoch": 1.24, + "grad_norm": 0.49609375, + "learning_rate": 6.278299701460438e-05, + "loss": 2.491, + "step": 29202 + }, + { + "epoch": 1.24, + "grad_norm": 0.44921875, + "learning_rate": 6.27767710339552e-05, + "loss": 2.5162, + "step": 29203 + }, + { + "epoch": 1.24, + "grad_norm": 0.490234375, + "learning_rate": 6.277054522079693e-05, + "loss": 2.4442, + "step": 29204 + }, + { + "epoch": 1.24, + "grad_norm": 0.478515625, + "learning_rate": 6.276431957515761e-05, + "loss": 2.4193, + "step": 29205 + }, + { + "epoch": 1.24, + "grad_norm": 0.423828125, + "learning_rate": 6.275809409706521e-05, + "loss": 2.5675, + "step": 29206 + }, + { + "epoch": 1.24, + "grad_norm": 0.44921875, + "learning_rate": 6.27518687865478e-05, + "loss": 2.4485, + "step": 29207 + }, + { + "epoch": 1.24, + "grad_norm": 0.421875, + "learning_rate": 6.274564364363336e-05, + "loss": 2.401, + "step": 29208 + }, + { + "epoch": 1.24, + "grad_norm": 0.49609375, + "learning_rate": 6.273941866834993e-05, + "loss": 2.5643, + "step": 29209 + }, + { + "epoch": 1.24, + "grad_norm": 0.458984375, + "learning_rate": 6.273319386072544e-05, + "loss": 2.5496, + "step": 29210 + }, + { + "epoch": 1.24, + "grad_norm": 0.408203125, + "learning_rate": 6.272696922078803e-05, + "loss": 2.3704, + "step": 29211 + }, + { + "epoch": 1.24, + "grad_norm": 0.43359375, + "learning_rate": 6.272074474856562e-05, + "loss": 2.5039, + "step": 29212 + }, + { + "epoch": 1.24, + "grad_norm": 0.435546875, + "learning_rate": 6.271452044408622e-05, + "loss": 2.5979, + "step": 29213 + }, + { + "epoch": 1.24, + "grad_norm": 0.5703125, + "learning_rate": 6.270829630737784e-05, + "loss": 2.3746, + "step": 29214 + }, + { + "epoch": 1.24, + "grad_norm": 0.4609375, + "learning_rate": 6.27020723384685e-05, + "loss": 2.4453, + "step": 29215 + }, + { + "epoch": 1.24, + "grad_norm": 0.494140625, + "learning_rate": 6.269584853738624e-05, + "loss": 2.7313, + "step": 29216 + }, + { + "epoch": 1.25, + "grad_norm": 0.53125, + "learning_rate": 6.268962490415901e-05, + "loss": 2.484, + "step": 29217 + }, + { + "epoch": 1.25, + "grad_norm": 0.42578125, + "learning_rate": 6.268340143881479e-05, + "loss": 2.4133, + "step": 29218 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.267717814138167e-05, + "loss": 2.5171, + "step": 29219 + }, + { + "epoch": 1.25, + "grad_norm": 0.4765625, + "learning_rate": 6.267095501188763e-05, + "loss": 2.4422, + "step": 29220 + }, + { + "epoch": 1.25, + "grad_norm": 0.56640625, + "learning_rate": 6.266473205036062e-05, + "loss": 2.3136, + "step": 29221 + }, + { + "epoch": 1.25, + "grad_norm": 0.447265625, + "learning_rate": 6.265850925682868e-05, + "loss": 2.5512, + "step": 29222 + }, + { + "epoch": 1.25, + "grad_norm": 0.466796875, + "learning_rate": 6.265228663131976e-05, + "loss": 2.5332, + "step": 29223 + }, + { + "epoch": 1.25, + "grad_norm": 0.447265625, + "learning_rate": 6.264606417386195e-05, + "loss": 2.3136, + "step": 29224 + }, + { + "epoch": 1.25, + "grad_norm": 0.515625, + "learning_rate": 6.263984188448321e-05, + "loss": 2.5075, + "step": 29225 + }, + { + "epoch": 1.25, + "grad_norm": 0.50390625, + "learning_rate": 6.26336197632115e-05, + "loss": 2.4944, + "step": 29226 + }, + { + "epoch": 1.25, + "grad_norm": 0.484375, + "learning_rate": 6.262739781007482e-05, + "loss": 2.4685, + "step": 29227 + }, + { + "epoch": 1.25, + "grad_norm": 0.44140625, + "learning_rate": 6.262117602510124e-05, + "loss": 2.4939, + "step": 29228 + }, + { + "epoch": 1.25, + "grad_norm": 0.466796875, + "learning_rate": 6.261495440831869e-05, + "loss": 2.4323, + "step": 29229 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.260873295975518e-05, + "loss": 2.4089, + "step": 29230 + }, + { + "epoch": 1.25, + "grad_norm": 0.458984375, + "learning_rate": 6.26025116794387e-05, + "loss": 2.4565, + "step": 29231 + }, + { + "epoch": 1.25, + "grad_norm": 0.44140625, + "learning_rate": 6.259629056739726e-05, + "loss": 2.5551, + "step": 29232 + }, + { + "epoch": 1.25, + "grad_norm": 0.447265625, + "learning_rate": 6.259006962365886e-05, + "loss": 2.5208, + "step": 29233 + }, + { + "epoch": 1.25, + "grad_norm": 0.53125, + "learning_rate": 6.258384884825146e-05, + "loss": 2.4764, + "step": 29234 + }, + { + "epoch": 1.25, + "grad_norm": 0.4765625, + "learning_rate": 6.257762824120302e-05, + "loss": 2.4225, + "step": 29235 + }, + { + "epoch": 1.25, + "grad_norm": 0.43359375, + "learning_rate": 6.257140780254163e-05, + "loss": 2.3963, + "step": 29236 + }, + { + "epoch": 1.25, + "grad_norm": 0.50390625, + "learning_rate": 6.256518753229522e-05, + "loss": 2.4672, + "step": 29237 + }, + { + "epoch": 1.25, + "grad_norm": 0.44140625, + "learning_rate": 6.255896743049179e-05, + "loss": 2.3955, + "step": 29238 + }, + { + "epoch": 1.25, + "grad_norm": 0.419921875, + "learning_rate": 6.25527474971593e-05, + "loss": 2.2867, + "step": 29239 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.254652773232577e-05, + "loss": 2.5888, + "step": 29240 + }, + { + "epoch": 1.25, + "grad_norm": 0.423828125, + "learning_rate": 6.25403081360192e-05, + "loss": 2.4911, + "step": 29241 + }, + { + "epoch": 1.25, + "grad_norm": 0.474609375, + "learning_rate": 6.253408870826753e-05, + "loss": 2.4245, + "step": 29242 + }, + { + "epoch": 1.25, + "grad_norm": 0.490234375, + "learning_rate": 6.252786944909874e-05, + "loss": 2.5223, + "step": 29243 + }, + { + "epoch": 1.25, + "grad_norm": 0.5078125, + "learning_rate": 6.25216503585409e-05, + "loss": 2.5276, + "step": 29244 + }, + { + "epoch": 1.25, + "grad_norm": 0.44140625, + "learning_rate": 6.251543143662192e-05, + "loss": 2.5039, + "step": 29245 + }, + { + "epoch": 1.25, + "grad_norm": 0.494140625, + "learning_rate": 6.25092126833698e-05, + "loss": 2.4959, + "step": 29246 + }, + { + "epoch": 1.25, + "grad_norm": 0.48828125, + "learning_rate": 6.250299409881248e-05, + "loss": 2.4767, + "step": 29247 + }, + { + "epoch": 1.25, + "grad_norm": 0.435546875, + "learning_rate": 6.249677568297802e-05, + "loss": 2.3814, + "step": 29248 + }, + { + "epoch": 1.25, + "grad_norm": 0.447265625, + "learning_rate": 6.249055743589439e-05, + "loss": 2.4152, + "step": 29249 + }, + { + "epoch": 1.25, + "grad_norm": 0.478515625, + "learning_rate": 6.248433935758952e-05, + "loss": 2.469, + "step": 29250 + }, + { + "epoch": 1.25, + "grad_norm": 0.43359375, + "learning_rate": 6.247812144809136e-05, + "loss": 2.5305, + "step": 29251 + }, + { + "epoch": 1.25, + "grad_norm": 0.4765625, + "learning_rate": 6.247190370742801e-05, + "loss": 2.3739, + "step": 29252 + }, + { + "epoch": 1.25, + "grad_norm": 0.404296875, + "learning_rate": 6.246568613562737e-05, + "loss": 2.4701, + "step": 29253 + }, + { + "epoch": 1.25, + "grad_norm": 0.4140625, + "learning_rate": 6.245946873271741e-05, + "loss": 2.3435, + "step": 29254 + }, + { + "epoch": 1.25, + "grad_norm": 0.46875, + "learning_rate": 6.24532514987261e-05, + "loss": 2.526, + "step": 29255 + }, + { + "epoch": 1.25, + "grad_norm": 0.515625, + "learning_rate": 6.244703443368146e-05, + "loss": 2.3703, + "step": 29256 + }, + { + "epoch": 1.25, + "grad_norm": 0.439453125, + "learning_rate": 6.244081753761147e-05, + "loss": 2.3609, + "step": 29257 + }, + { + "epoch": 1.25, + "grad_norm": 0.52734375, + "learning_rate": 6.243460081054404e-05, + "loss": 2.5606, + "step": 29258 + }, + { + "epoch": 1.25, + "grad_norm": 0.451171875, + "learning_rate": 6.242838425250714e-05, + "loss": 2.4279, + "step": 29259 + }, + { + "epoch": 1.25, + "grad_norm": 0.45703125, + "learning_rate": 6.242216786352886e-05, + "loss": 2.5288, + "step": 29260 + }, + { + "epoch": 1.25, + "grad_norm": 0.498046875, + "learning_rate": 6.241595164363705e-05, + "loss": 2.5123, + "step": 29261 + }, + { + "epoch": 1.25, + "grad_norm": 0.451171875, + "learning_rate": 6.240973559285971e-05, + "loss": 2.4302, + "step": 29262 + }, + { + "epoch": 1.25, + "grad_norm": 0.498046875, + "learning_rate": 6.240351971122484e-05, + "loss": 2.5653, + "step": 29263 + }, + { + "epoch": 1.25, + "grad_norm": 0.52734375, + "learning_rate": 6.239730399876037e-05, + "loss": 2.4636, + "step": 29264 + }, + { + "epoch": 1.25, + "grad_norm": 0.48046875, + "learning_rate": 6.239108845549435e-05, + "loss": 2.447, + "step": 29265 + }, + { + "epoch": 1.25, + "grad_norm": 0.435546875, + "learning_rate": 6.238487308145464e-05, + "loss": 2.5724, + "step": 29266 + }, + { + "epoch": 1.25, + "grad_norm": 0.4296875, + "learning_rate": 6.237865787666923e-05, + "loss": 2.4286, + "step": 29267 + }, + { + "epoch": 1.25, + "grad_norm": 0.453125, + "learning_rate": 6.237244284116617e-05, + "loss": 2.3913, + "step": 29268 + }, + { + "epoch": 1.25, + "grad_norm": 0.478515625, + "learning_rate": 6.236622797497334e-05, + "loss": 2.545, + "step": 29269 + }, + { + "epoch": 1.25, + "grad_norm": 0.419921875, + "learning_rate": 6.236001327811872e-05, + "loss": 2.3713, + "step": 29270 + }, + { + "epoch": 1.25, + "grad_norm": 0.48828125, + "learning_rate": 6.235379875063027e-05, + "loss": 2.3797, + "step": 29271 + }, + { + "epoch": 1.25, + "grad_norm": 0.451171875, + "learning_rate": 6.234758439253601e-05, + "loss": 2.3885, + "step": 29272 + }, + { + "epoch": 1.25, + "grad_norm": 0.5, + "learning_rate": 6.234137020386384e-05, + "loss": 2.6529, + "step": 29273 + }, + { + "epoch": 1.25, + "grad_norm": 0.52734375, + "learning_rate": 6.233515618464176e-05, + "loss": 2.5359, + "step": 29274 + }, + { + "epoch": 1.25, + "grad_norm": 0.443359375, + "learning_rate": 6.232894233489766e-05, + "loss": 2.557, + "step": 29275 + }, + { + "epoch": 1.25, + "grad_norm": 0.486328125, + "learning_rate": 6.23227286546596e-05, + "loss": 2.5416, + "step": 29276 + }, + { + "epoch": 1.25, + "grad_norm": 0.4453125, + "learning_rate": 6.231651514395548e-05, + "loss": 2.4245, + "step": 29277 + }, + { + "epoch": 1.25, + "grad_norm": 0.50390625, + "learning_rate": 6.231030180281328e-05, + "loss": 2.4404, + "step": 29278 + }, + { + "epoch": 1.25, + "grad_norm": 0.44921875, + "learning_rate": 6.230408863126092e-05, + "loss": 2.5294, + "step": 29279 + }, + { + "epoch": 1.25, + "grad_norm": 0.4296875, + "learning_rate": 6.229787562932639e-05, + "loss": 2.4827, + "step": 29280 + }, + { + "epoch": 1.25, + "grad_norm": 0.451171875, + "learning_rate": 6.229166279703765e-05, + "loss": 2.2881, + "step": 29281 + }, + { + "epoch": 1.25, + "grad_norm": 0.478515625, + "learning_rate": 6.228545013442267e-05, + "loss": 2.4027, + "step": 29282 + }, + { + "epoch": 1.25, + "grad_norm": 0.55078125, + "learning_rate": 6.227923764150933e-05, + "loss": 2.5144, + "step": 29283 + }, + { + "epoch": 1.25, + "grad_norm": 0.48046875, + "learning_rate": 6.227302531832569e-05, + "loss": 2.5235, + "step": 29284 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.226681316489962e-05, + "loss": 2.4541, + "step": 29285 + }, + { + "epoch": 1.25, + "grad_norm": 0.48046875, + "learning_rate": 6.22606011812591e-05, + "loss": 2.3861, + "step": 29286 + }, + { + "epoch": 1.25, + "grad_norm": 0.515625, + "learning_rate": 6.225438936743206e-05, + "loss": 2.5802, + "step": 29287 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.22481777234465e-05, + "loss": 2.3499, + "step": 29288 + }, + { + "epoch": 1.25, + "grad_norm": 0.42578125, + "learning_rate": 6.224196624933036e-05, + "loss": 2.3569, + "step": 29289 + }, + { + "epoch": 1.25, + "grad_norm": 0.486328125, + "learning_rate": 6.223575494511158e-05, + "loss": 2.4973, + "step": 29290 + }, + { + "epoch": 1.25, + "grad_norm": 0.490234375, + "learning_rate": 6.222954381081804e-05, + "loss": 2.4087, + "step": 29291 + }, + { + "epoch": 1.25, + "grad_norm": 0.46875, + "learning_rate": 6.222333284647781e-05, + "loss": 2.3071, + "step": 29292 + }, + { + "epoch": 1.25, + "grad_norm": 0.5546875, + "learning_rate": 6.221712205211875e-05, + "loss": 2.3907, + "step": 29293 + }, + { + "epoch": 1.25, + "grad_norm": 0.4453125, + "learning_rate": 6.221091142776884e-05, + "loss": 2.32, + "step": 29294 + }, + { + "epoch": 1.25, + "grad_norm": 0.5234375, + "learning_rate": 6.2204700973456e-05, + "loss": 2.5145, + "step": 29295 + }, + { + "epoch": 1.25, + "grad_norm": 0.44921875, + "learning_rate": 6.21984906892082e-05, + "loss": 2.5447, + "step": 29296 + }, + { + "epoch": 1.25, + "grad_norm": 0.482421875, + "learning_rate": 6.21922805750534e-05, + "loss": 2.489, + "step": 29297 + }, + { + "epoch": 1.25, + "grad_norm": 0.50390625, + "learning_rate": 6.218607063101952e-05, + "loss": 2.549, + "step": 29298 + }, + { + "epoch": 1.25, + "grad_norm": 0.5703125, + "learning_rate": 6.217986085713445e-05, + "loss": 2.4889, + "step": 29299 + }, + { + "epoch": 1.25, + "grad_norm": 0.5078125, + "learning_rate": 6.217365125342624e-05, + "loss": 2.5114, + "step": 29300 + }, + { + "epoch": 1.25, + "grad_norm": 0.41015625, + "learning_rate": 6.216744181992277e-05, + "loss": 2.5361, + "step": 29301 + }, + { + "epoch": 1.25, + "grad_norm": 0.48828125, + "learning_rate": 6.216123255665197e-05, + "loss": 2.4759, + "step": 29302 + }, + { + "epoch": 1.25, + "grad_norm": 0.447265625, + "learning_rate": 6.215502346364179e-05, + "loss": 2.4656, + "step": 29303 + }, + { + "epoch": 1.25, + "grad_norm": 0.53515625, + "learning_rate": 6.214881454092018e-05, + "loss": 2.5412, + "step": 29304 + }, + { + "epoch": 1.25, + "grad_norm": 0.4765625, + "learning_rate": 6.214260578851507e-05, + "loss": 2.4974, + "step": 29305 + }, + { + "epoch": 1.25, + "grad_norm": 0.482421875, + "learning_rate": 6.213639720645442e-05, + "loss": 2.3378, + "step": 29306 + }, + { + "epoch": 1.25, + "grad_norm": 0.431640625, + "learning_rate": 6.213018879476609e-05, + "loss": 2.3466, + "step": 29307 + }, + { + "epoch": 1.25, + "grad_norm": 0.478515625, + "learning_rate": 6.212398055347814e-05, + "loss": 2.4917, + "step": 29308 + }, + { + "epoch": 1.25, + "grad_norm": 0.5234375, + "learning_rate": 6.21177724826184e-05, + "loss": 2.5298, + "step": 29309 + }, + { + "epoch": 1.25, + "grad_norm": 0.453125, + "learning_rate": 6.211156458221484e-05, + "loss": 2.4284, + "step": 29310 + }, + { + "epoch": 1.25, + "grad_norm": 0.51953125, + "learning_rate": 6.210535685229537e-05, + "loss": 2.4496, + "step": 29311 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.209914929288798e-05, + "loss": 2.4317, + "step": 29312 + }, + { + "epoch": 1.25, + "grad_norm": 0.48046875, + "learning_rate": 6.209294190402056e-05, + "loss": 2.608, + "step": 29313 + }, + { + "epoch": 1.25, + "grad_norm": 0.515625, + "learning_rate": 6.208673468572107e-05, + "loss": 2.3949, + "step": 29314 + }, + { + "epoch": 1.25, + "grad_norm": 0.57421875, + "learning_rate": 6.208052763801735e-05, + "loss": 2.4006, + "step": 29315 + }, + { + "epoch": 1.25, + "grad_norm": 0.50390625, + "learning_rate": 6.207432076093747e-05, + "loss": 2.5381, + "step": 29316 + }, + { + "epoch": 1.25, + "grad_norm": 0.51953125, + "learning_rate": 6.206811405450925e-05, + "loss": 2.4537, + "step": 29317 + }, + { + "epoch": 1.25, + "grad_norm": 0.4609375, + "learning_rate": 6.206190751876066e-05, + "loss": 2.4217, + "step": 29318 + }, + { + "epoch": 1.25, + "grad_norm": 0.421875, + "learning_rate": 6.205570115371961e-05, + "loss": 2.5223, + "step": 29319 + }, + { + "epoch": 1.25, + "grad_norm": 0.48046875, + "learning_rate": 6.204949495941406e-05, + "loss": 2.4548, + "step": 29320 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.204328893587191e-05, + "loss": 2.413, + "step": 29321 + }, + { + "epoch": 1.25, + "grad_norm": 0.5234375, + "learning_rate": 6.20370830831211e-05, + "loss": 2.4571, + "step": 29322 + }, + { + "epoch": 1.25, + "grad_norm": 0.482421875, + "learning_rate": 6.203087740118949e-05, + "loss": 2.5208, + "step": 29323 + }, + { + "epoch": 1.25, + "grad_norm": 0.4609375, + "learning_rate": 6.20246718901051e-05, + "loss": 2.5387, + "step": 29324 + }, + { + "epoch": 1.25, + "grad_norm": 0.404296875, + "learning_rate": 6.20184665498958e-05, + "loss": 2.4126, + "step": 29325 + }, + { + "epoch": 1.25, + "grad_norm": 0.458984375, + "learning_rate": 6.201226138058954e-05, + "loss": 2.4495, + "step": 29326 + }, + { + "epoch": 1.25, + "grad_norm": 0.37890625, + "learning_rate": 6.200605638221417e-05, + "loss": 2.4113, + "step": 29327 + }, + { + "epoch": 1.25, + "grad_norm": 0.47265625, + "learning_rate": 6.199985155479769e-05, + "loss": 2.5273, + "step": 29328 + }, + { + "epoch": 1.25, + "grad_norm": 0.46484375, + "learning_rate": 6.199364689836799e-05, + "loss": 2.4017, + "step": 29329 + }, + { + "epoch": 1.25, + "grad_norm": 0.408203125, + "learning_rate": 6.198744241295302e-05, + "loss": 2.4201, + "step": 29330 + }, + { + "epoch": 1.25, + "grad_norm": 0.474609375, + "learning_rate": 6.19812380985806e-05, + "loss": 2.402, + "step": 29331 + }, + { + "epoch": 1.25, + "grad_norm": 0.41796875, + "learning_rate": 6.197503395527878e-05, + "loss": 2.5769, + "step": 29332 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.196882998307538e-05, + "loss": 2.5078, + "step": 29333 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.196262618199836e-05, + "loss": 2.4973, + "step": 29334 + }, + { + "epoch": 1.25, + "grad_norm": 0.478515625, + "learning_rate": 6.195642255207558e-05, + "loss": 2.5319, + "step": 29335 + }, + { + "epoch": 1.25, + "eval_loss": 2.5066845417022705, + "eval_runtime": 82.8093, + "eval_samples_per_second": 18.162, + "eval_steps_per_second": 2.27, + "step": 29335 + }, + { + "epoch": 1.25, + "grad_norm": 0.458984375, + "learning_rate": 6.195021909333503e-05, + "loss": 2.4241, + "step": 29336 + }, + { + "epoch": 1.25, + "grad_norm": 0.50390625, + "learning_rate": 6.194401580580458e-05, + "loss": 2.5697, + "step": 29337 + }, + { + "epoch": 1.25, + "grad_norm": 0.408203125, + "learning_rate": 6.193781268951218e-05, + "loss": 2.3692, + "step": 29338 + }, + { + "epoch": 1.25, + "grad_norm": 0.4453125, + "learning_rate": 6.193160974448564e-05, + "loss": 2.5536, + "step": 29339 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.192540697075302e-05, + "loss": 2.4939, + "step": 29340 + }, + { + "epoch": 1.25, + "grad_norm": 0.47265625, + "learning_rate": 6.191920436834212e-05, + "loss": 2.5672, + "step": 29341 + }, + { + "epoch": 1.25, + "grad_norm": 0.46875, + "learning_rate": 6.191300193728089e-05, + "loss": 2.4736, + "step": 29342 + }, + { + "epoch": 1.25, + "grad_norm": 0.486328125, + "learning_rate": 6.190679967759724e-05, + "loss": 2.4397, + "step": 29343 + }, + { + "epoch": 1.25, + "grad_norm": 0.4609375, + "learning_rate": 6.190059758931903e-05, + "loss": 2.4805, + "step": 29344 + }, + { + "epoch": 1.25, + "grad_norm": 0.490234375, + "learning_rate": 6.189439567247425e-05, + "loss": 2.4033, + "step": 29345 + }, + { + "epoch": 1.25, + "grad_norm": 0.49609375, + "learning_rate": 6.188819392709078e-05, + "loss": 2.5134, + "step": 29346 + }, + { + "epoch": 1.25, + "grad_norm": 0.443359375, + "learning_rate": 6.188199235319647e-05, + "loss": 2.3838, + "step": 29347 + }, + { + "epoch": 1.25, + "grad_norm": 0.458984375, + "learning_rate": 6.187579095081926e-05, + "loss": 2.3614, + "step": 29348 + }, + { + "epoch": 1.25, + "grad_norm": 0.5, + "learning_rate": 6.186958971998706e-05, + "loss": 2.379, + "step": 29349 + }, + { + "epoch": 1.25, + "grad_norm": 0.458984375, + "learning_rate": 6.186338866072779e-05, + "loss": 2.3593, + "step": 29350 + }, + { + "epoch": 1.25, + "grad_norm": 0.5390625, + "learning_rate": 6.185718777306934e-05, + "loss": 2.5622, + "step": 29351 + }, + { + "epoch": 1.25, + "grad_norm": 0.408203125, + "learning_rate": 6.185098705703956e-05, + "loss": 2.5113, + "step": 29352 + }, + { + "epoch": 1.25, + "grad_norm": 0.50390625, + "learning_rate": 6.184478651266644e-05, + "loss": 2.4015, + "step": 29353 + }, + { + "epoch": 1.25, + "grad_norm": 0.52734375, + "learning_rate": 6.183858613997784e-05, + "loss": 2.4791, + "step": 29354 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.183238593900165e-05, + "loss": 2.4411, + "step": 29355 + }, + { + "epoch": 1.25, + "grad_norm": 0.546875, + "learning_rate": 6.182618590976575e-05, + "loss": 2.5872, + "step": 29356 + }, + { + "epoch": 1.25, + "grad_norm": 0.44921875, + "learning_rate": 6.181998605229807e-05, + "loss": 2.3848, + "step": 29357 + }, + { + "epoch": 1.25, + "grad_norm": 0.4453125, + "learning_rate": 6.181378636662652e-05, + "loss": 2.3188, + "step": 29358 + }, + { + "epoch": 1.25, + "grad_norm": 0.48046875, + "learning_rate": 6.180758685277896e-05, + "loss": 2.4638, + "step": 29359 + }, + { + "epoch": 1.25, + "grad_norm": 0.453125, + "learning_rate": 6.180138751078328e-05, + "loss": 2.5186, + "step": 29360 + }, + { + "epoch": 1.25, + "grad_norm": 0.498046875, + "learning_rate": 6.179518834066742e-05, + "loss": 2.5713, + "step": 29361 + }, + { + "epoch": 1.25, + "grad_norm": 0.47265625, + "learning_rate": 6.178898934245927e-05, + "loss": 2.616, + "step": 29362 + }, + { + "epoch": 1.25, + "grad_norm": 0.48828125, + "learning_rate": 6.178279051618669e-05, + "loss": 2.5646, + "step": 29363 + }, + { + "epoch": 1.25, + "grad_norm": 0.474609375, + "learning_rate": 6.177659186187757e-05, + "loss": 2.4177, + "step": 29364 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.177039337955983e-05, + "loss": 2.4763, + "step": 29365 + }, + { + "epoch": 1.25, + "grad_norm": 0.466796875, + "learning_rate": 6.176419506926135e-05, + "loss": 2.5034, + "step": 29366 + }, + { + "epoch": 1.25, + "grad_norm": 0.44140625, + "learning_rate": 6.175799693101e-05, + "loss": 2.3663, + "step": 29367 + }, + { + "epoch": 1.25, + "grad_norm": 0.5703125, + "learning_rate": 6.175179896483369e-05, + "loss": 2.6317, + "step": 29368 + }, + { + "epoch": 1.25, + "grad_norm": 0.4296875, + "learning_rate": 6.174560117076033e-05, + "loss": 2.3803, + "step": 29369 + }, + { + "epoch": 1.25, + "grad_norm": 0.455078125, + "learning_rate": 6.17394035488178e-05, + "loss": 2.4265, + "step": 29370 + }, + { + "epoch": 1.25, + "grad_norm": 0.5390625, + "learning_rate": 6.173320609903394e-05, + "loss": 2.4473, + "step": 29371 + }, + { + "epoch": 1.25, + "grad_norm": 0.53515625, + "learning_rate": 6.172700882143666e-05, + "loss": 2.4156, + "step": 29372 + }, + { + "epoch": 1.25, + "grad_norm": 0.484375, + "learning_rate": 6.172081171605387e-05, + "loss": 2.5021, + "step": 29373 + }, + { + "epoch": 1.25, + "grad_norm": 0.40234375, + "learning_rate": 6.171461478291343e-05, + "loss": 2.3465, + "step": 29374 + }, + { + "epoch": 1.25, + "grad_norm": 0.51953125, + "learning_rate": 6.170841802204324e-05, + "loss": 2.5086, + "step": 29375 + }, + { + "epoch": 1.25, + "grad_norm": 0.515625, + "learning_rate": 6.170222143347115e-05, + "loss": 2.5471, + "step": 29376 + }, + { + "epoch": 1.25, + "grad_norm": 0.53125, + "learning_rate": 6.16960250172251e-05, + "loss": 2.5317, + "step": 29377 + }, + { + "epoch": 1.25, + "grad_norm": 0.44921875, + "learning_rate": 6.168982877333293e-05, + "loss": 2.5252, + "step": 29378 + }, + { + "epoch": 1.25, + "grad_norm": 0.4765625, + "learning_rate": 6.168363270182253e-05, + "loss": 2.5272, + "step": 29379 + }, + { + "epoch": 1.25, + "grad_norm": 0.48046875, + "learning_rate": 6.167743680272177e-05, + "loss": 2.4653, + "step": 29380 + }, + { + "epoch": 1.25, + "grad_norm": 0.46484375, + "learning_rate": 6.167124107605856e-05, + "loss": 2.521, + "step": 29381 + }, + { + "epoch": 1.25, + "grad_norm": 0.4609375, + "learning_rate": 6.166504552186074e-05, + "loss": 2.4969, + "step": 29382 + }, + { + "epoch": 1.25, + "grad_norm": 0.478515625, + "learning_rate": 6.16588501401562e-05, + "loss": 2.5005, + "step": 29383 + }, + { + "epoch": 1.25, + "grad_norm": 0.45703125, + "learning_rate": 6.16526549309728e-05, + "loss": 2.4718, + "step": 29384 + }, + { + "epoch": 1.25, + "grad_norm": 0.47265625, + "learning_rate": 6.164645989433849e-05, + "loss": 2.348, + "step": 29385 + }, + { + "epoch": 1.25, + "grad_norm": 0.40625, + "learning_rate": 6.164026503028109e-05, + "loss": 2.3483, + "step": 29386 + }, + { + "epoch": 1.25, + "grad_norm": 0.486328125, + "learning_rate": 6.163407033882847e-05, + "loss": 2.5574, + "step": 29387 + }, + { + "epoch": 1.25, + "grad_norm": 0.478515625, + "learning_rate": 6.162787582000849e-05, + "loss": 2.5526, + "step": 29388 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.162168147384906e-05, + "loss": 2.458, + "step": 29389 + }, + { + "epoch": 1.25, + "grad_norm": 0.466796875, + "learning_rate": 6.161548730037803e-05, + "loss": 2.3492, + "step": 29390 + }, + { + "epoch": 1.25, + "grad_norm": 0.419921875, + "learning_rate": 6.160929329962329e-05, + "loss": 2.4289, + "step": 29391 + }, + { + "epoch": 1.25, + "grad_norm": 0.46484375, + "learning_rate": 6.160309947161268e-05, + "loss": 2.4493, + "step": 29392 + }, + { + "epoch": 1.25, + "grad_norm": 0.51953125, + "learning_rate": 6.159690581637411e-05, + "loss": 2.5976, + "step": 29393 + }, + { + "epoch": 1.25, + "grad_norm": 0.439453125, + "learning_rate": 6.159071233393545e-05, + "loss": 2.2813, + "step": 29394 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.158451902432452e-05, + "loss": 2.4306, + "step": 29395 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.15783258875692e-05, + "loss": 2.501, + "step": 29396 + }, + { + "epoch": 1.25, + "grad_norm": 0.45703125, + "learning_rate": 6.157213292369738e-05, + "loss": 2.5266, + "step": 29397 + }, + { + "epoch": 1.25, + "grad_norm": 0.48046875, + "learning_rate": 6.156594013273693e-05, + "loss": 2.5517, + "step": 29398 + }, + { + "epoch": 1.25, + "grad_norm": 0.52734375, + "learning_rate": 6.155974751471572e-05, + "loss": 2.4195, + "step": 29399 + }, + { + "epoch": 1.25, + "grad_norm": 0.54296875, + "learning_rate": 6.155355506966155e-05, + "loss": 2.3853, + "step": 29400 + }, + { + "epoch": 1.25, + "grad_norm": 0.4921875, + "learning_rate": 6.154736279760236e-05, + "loss": 2.4713, + "step": 29401 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.154117069856603e-05, + "loss": 2.6129, + "step": 29402 + }, + { + "epoch": 1.25, + "grad_norm": 0.455078125, + "learning_rate": 6.153497877258032e-05, + "loss": 2.4504, + "step": 29403 + }, + { + "epoch": 1.25, + "grad_norm": 0.439453125, + "learning_rate": 6.152878701967314e-05, + "loss": 2.4899, + "step": 29404 + }, + { + "epoch": 1.25, + "grad_norm": 0.45703125, + "learning_rate": 6.152259543987237e-05, + "loss": 2.5217, + "step": 29405 + }, + { + "epoch": 1.25, + "grad_norm": 0.4921875, + "learning_rate": 6.151640403320588e-05, + "loss": 2.5028, + "step": 29406 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.151021279970151e-05, + "loss": 2.4921, + "step": 29407 + }, + { + "epoch": 1.25, + "grad_norm": 0.419921875, + "learning_rate": 6.150402173938708e-05, + "loss": 2.4693, + "step": 29408 + }, + { + "epoch": 1.25, + "grad_norm": 0.4296875, + "learning_rate": 6.149783085229049e-05, + "loss": 2.2846, + "step": 29409 + }, + { + "epoch": 1.25, + "grad_norm": 0.45703125, + "learning_rate": 6.149164013843964e-05, + "loss": 2.4156, + "step": 29410 + }, + { + "epoch": 1.25, + "grad_norm": 0.486328125, + "learning_rate": 6.14854495978623e-05, + "loss": 2.5085, + "step": 29411 + }, + { + "epoch": 1.25, + "grad_norm": 0.50390625, + "learning_rate": 6.147925923058635e-05, + "loss": 2.4764, + "step": 29412 + }, + { + "epoch": 1.25, + "grad_norm": 0.462890625, + "learning_rate": 6.147306903663966e-05, + "loss": 2.462, + "step": 29413 + }, + { + "epoch": 1.25, + "grad_norm": 0.390625, + "learning_rate": 6.146687901605009e-05, + "loss": 2.2052, + "step": 29414 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.146068916884548e-05, + "loss": 2.4212, + "step": 29415 + }, + { + "epoch": 1.25, + "grad_norm": 0.494140625, + "learning_rate": 6.145449949505366e-05, + "loss": 2.4894, + "step": 29416 + }, + { + "epoch": 1.25, + "grad_norm": 0.45703125, + "learning_rate": 6.144830999470254e-05, + "loss": 2.4423, + "step": 29417 + }, + { + "epoch": 1.25, + "grad_norm": 0.48828125, + "learning_rate": 6.144212066781995e-05, + "loss": 2.4224, + "step": 29418 + }, + { + "epoch": 1.25, + "grad_norm": 0.453125, + "learning_rate": 6.14359315144337e-05, + "loss": 2.352, + "step": 29419 + }, + { + "epoch": 1.25, + "grad_norm": 0.4296875, + "learning_rate": 6.142974253457166e-05, + "loss": 2.4644, + "step": 29420 + }, + { + "epoch": 1.25, + "grad_norm": 0.515625, + "learning_rate": 6.142355372826167e-05, + "loss": 2.5259, + "step": 29421 + }, + { + "epoch": 1.25, + "grad_norm": 0.498046875, + "learning_rate": 6.141736509553162e-05, + "loss": 2.4414, + "step": 29422 + }, + { + "epoch": 1.25, + "grad_norm": 0.48828125, + "learning_rate": 6.141117663640931e-05, + "loss": 2.4264, + "step": 29423 + }, + { + "epoch": 1.25, + "grad_norm": 0.470703125, + "learning_rate": 6.14049883509226e-05, + "loss": 2.5244, + "step": 29424 + }, + { + "epoch": 1.25, + "grad_norm": 0.458984375, + "learning_rate": 6.139880023909933e-05, + "loss": 2.3095, + "step": 29425 + }, + { + "epoch": 1.25, + "grad_norm": 0.4296875, + "learning_rate": 6.13926123009674e-05, + "loss": 2.3994, + "step": 29426 + }, + { + "epoch": 1.25, + "grad_norm": 0.431640625, + "learning_rate": 6.138642453655459e-05, + "loss": 2.4095, + "step": 29427 + }, + { + "epoch": 1.25, + "grad_norm": 0.44921875, + "learning_rate": 6.138023694588872e-05, + "loss": 2.4557, + "step": 29428 + }, + { + "epoch": 1.25, + "grad_norm": 0.421875, + "learning_rate": 6.137404952899768e-05, + "loss": 2.4526, + "step": 29429 + }, + { + "epoch": 1.25, + "grad_norm": 0.4453125, + "learning_rate": 6.136786228590931e-05, + "loss": 2.4113, + "step": 29430 + }, + { + "epoch": 1.25, + "grad_norm": 0.44921875, + "learning_rate": 6.136167521665146e-05, + "loss": 2.4262, + "step": 29431 + }, + { + "epoch": 1.25, + "grad_norm": 0.5078125, + "learning_rate": 6.135548832125191e-05, + "loss": 2.5667, + "step": 29432 + }, + { + "epoch": 1.25, + "grad_norm": 0.482421875, + "learning_rate": 6.134930159973856e-05, + "loss": 2.4379, + "step": 29433 + }, + { + "epoch": 1.25, + "grad_norm": 0.45703125, + "learning_rate": 6.134311505213926e-05, + "loss": 2.4875, + "step": 29434 + }, + { + "epoch": 1.25, + "grad_norm": 0.4609375, + "learning_rate": 6.133692867848178e-05, + "loss": 2.4718, + "step": 29435 + }, + { + "epoch": 1.25, + "grad_norm": 0.486328125, + "learning_rate": 6.133074247879398e-05, + "loss": 2.5773, + "step": 29436 + }, + { + "epoch": 1.25, + "grad_norm": 0.453125, + "learning_rate": 6.132455645310373e-05, + "loss": 2.3831, + "step": 29437 + }, + { + "epoch": 1.25, + "grad_norm": 0.396484375, + "learning_rate": 6.131837060143883e-05, + "loss": 2.4721, + "step": 29438 + }, + { + "epoch": 1.25, + "grad_norm": 0.51171875, + "learning_rate": 6.131218492382714e-05, + "loss": 2.433, + "step": 29439 + }, + { + "epoch": 1.25, + "grad_norm": 0.447265625, + "learning_rate": 6.130599942029645e-05, + "loss": 2.387, + "step": 29440 + }, + { + "epoch": 1.25, + "grad_norm": 0.435546875, + "learning_rate": 6.129981409087464e-05, + "loss": 2.4233, + "step": 29441 + }, + { + "epoch": 1.25, + "grad_norm": 0.474609375, + "learning_rate": 6.129362893558956e-05, + "loss": 2.4893, + "step": 29442 + }, + { + "epoch": 1.25, + "grad_norm": 0.51953125, + "learning_rate": 6.128744395446897e-05, + "loss": 2.4152, + "step": 29443 + }, + { + "epoch": 1.25, + "grad_norm": 0.46875, + "learning_rate": 6.12812591475407e-05, + "loss": 2.4997, + "step": 29444 + }, + { + "epoch": 1.25, + "grad_norm": 0.482421875, + "learning_rate": 6.127507451483266e-05, + "loss": 2.3256, + "step": 29445 + }, + { + "epoch": 1.25, + "grad_norm": 0.53125, + "learning_rate": 6.126889005637262e-05, + "loss": 2.4516, + "step": 29446 + }, + { + "epoch": 1.25, + "grad_norm": 0.447265625, + "learning_rate": 6.126270577218842e-05, + "loss": 2.4819, + "step": 29447 + }, + { + "epoch": 1.25, + "grad_norm": 0.5390625, + "learning_rate": 6.125652166230786e-05, + "loss": 2.5452, + "step": 29448 + }, + { + "epoch": 1.25, + "grad_norm": 0.458984375, + "learning_rate": 6.125033772675883e-05, + "loss": 2.4677, + "step": 29449 + }, + { + "epoch": 1.25, + "grad_norm": 0.546875, + "learning_rate": 6.124415396556913e-05, + "loss": 2.5094, + "step": 29450 + }, + { + "epoch": 1.25, + "grad_norm": 0.466796875, + "learning_rate": 6.123797037876657e-05, + "loss": 2.4808, + "step": 29451 + }, + { + "epoch": 1.26, + "grad_norm": 0.4765625, + "learning_rate": 6.123178696637893e-05, + "loss": 2.44, + "step": 29452 + }, + { + "epoch": 1.26, + "grad_norm": 0.5078125, + "learning_rate": 6.12256037284341e-05, + "loss": 2.4919, + "step": 29453 + }, + { + "epoch": 1.26, + "grad_norm": 0.423828125, + "learning_rate": 6.121942066495991e-05, + "loss": 2.2865, + "step": 29454 + }, + { + "epoch": 1.26, + "grad_norm": 0.4453125, + "learning_rate": 6.121323777598414e-05, + "loss": 2.3812, + "step": 29455 + }, + { + "epoch": 1.26, + "grad_norm": 0.458984375, + "learning_rate": 6.120705506153461e-05, + "loss": 2.4291, + "step": 29456 + }, + { + "epoch": 1.26, + "grad_norm": 0.54296875, + "learning_rate": 6.120087252163916e-05, + "loss": 2.5256, + "step": 29457 + }, + { + "epoch": 1.26, + "grad_norm": 0.439453125, + "learning_rate": 6.119469015632564e-05, + "loss": 2.2681, + "step": 29458 + }, + { + "epoch": 1.26, + "grad_norm": 0.47265625, + "learning_rate": 6.11885079656218e-05, + "loss": 2.4347, + "step": 29459 + }, + { + "epoch": 1.26, + "grad_norm": 0.46484375, + "learning_rate": 6.118232594955548e-05, + "loss": 2.4413, + "step": 29460 + }, + { + "epoch": 1.26, + "grad_norm": 0.43359375, + "learning_rate": 6.117614410815453e-05, + "loss": 2.4098, + "step": 29461 + }, + { + "epoch": 1.26, + "grad_norm": 0.5234375, + "learning_rate": 6.116996244144673e-05, + "loss": 2.5865, + "step": 29462 + }, + { + "epoch": 1.26, + "grad_norm": 0.48828125, + "learning_rate": 6.116378094945991e-05, + "loss": 2.4524, + "step": 29463 + }, + { + "epoch": 1.26, + "grad_norm": 0.443359375, + "learning_rate": 6.115759963222191e-05, + "loss": 2.4641, + "step": 29464 + }, + { + "epoch": 1.26, + "grad_norm": 0.4765625, + "learning_rate": 6.115141848976046e-05, + "loss": 2.3897, + "step": 29465 + }, + { + "epoch": 1.26, + "grad_norm": 0.4921875, + "learning_rate": 6.114523752210348e-05, + "loss": 2.389, + "step": 29466 + }, + { + "epoch": 1.26, + "grad_norm": 0.49609375, + "learning_rate": 6.113905672927872e-05, + "loss": 2.4577, + "step": 29467 + }, + { + "epoch": 1.26, + "grad_norm": 0.443359375, + "learning_rate": 6.1132876111314e-05, + "loss": 2.4233, + "step": 29468 + }, + { + "epoch": 1.26, + "grad_norm": 0.52734375, + "learning_rate": 6.11266956682371e-05, + "loss": 2.6262, + "step": 29469 + }, + { + "epoch": 1.26, + "grad_norm": 0.51171875, + "learning_rate": 6.11205154000759e-05, + "loss": 2.4676, + "step": 29470 + }, + { + "epoch": 1.26, + "grad_norm": 0.45703125, + "learning_rate": 6.111433530685818e-05, + "loss": 2.5595, + "step": 29471 + }, + { + "epoch": 1.26, + "grad_norm": 0.44921875, + "learning_rate": 6.110815538861173e-05, + "loss": 2.4237, + "step": 29472 + }, + { + "epoch": 1.26, + "grad_norm": 0.474609375, + "learning_rate": 6.110197564536433e-05, + "loss": 2.3727, + "step": 29473 + }, + { + "epoch": 1.26, + "grad_norm": 0.43359375, + "learning_rate": 6.109579607714386e-05, + "loss": 2.485, + "step": 29474 + }, + { + "epoch": 1.26, + "grad_norm": 0.42578125, + "learning_rate": 6.10896166839781e-05, + "loss": 2.481, + "step": 29475 + }, + { + "epoch": 1.26, + "grad_norm": 0.3984375, + "learning_rate": 6.108343746589482e-05, + "loss": 2.4224, + "step": 29476 + }, + { + "epoch": 1.26, + "grad_norm": 0.4453125, + "learning_rate": 6.107725842292184e-05, + "loss": 2.3114, + "step": 29477 + }, + { + "epoch": 1.26, + "grad_norm": 0.5078125, + "learning_rate": 6.107107955508699e-05, + "loss": 2.3275, + "step": 29478 + }, + { + "epoch": 1.26, + "grad_norm": 0.466796875, + "learning_rate": 6.106490086241807e-05, + "loss": 2.4038, + "step": 29479 + }, + { + "epoch": 1.26, + "grad_norm": 0.5234375, + "learning_rate": 6.105872234494288e-05, + "loss": 2.6028, + "step": 29480 + }, + { + "epoch": 1.26, + "grad_norm": 0.48828125, + "learning_rate": 6.105254400268914e-05, + "loss": 2.4287, + "step": 29481 + }, + { + "epoch": 1.26, + "grad_norm": 0.486328125, + "learning_rate": 6.104636583568478e-05, + "loss": 2.3523, + "step": 29482 + }, + { + "epoch": 1.26, + "grad_norm": 0.400390625, + "learning_rate": 6.104018784395752e-05, + "loss": 2.2896, + "step": 29483 + }, + { + "epoch": 1.26, + "grad_norm": 0.470703125, + "learning_rate": 6.103401002753517e-05, + "loss": 2.4477, + "step": 29484 + }, + { + "epoch": 1.26, + "grad_norm": 0.453125, + "learning_rate": 6.102783238644553e-05, + "loss": 2.4111, + "step": 29485 + }, + { + "epoch": 1.26, + "grad_norm": 0.45703125, + "learning_rate": 6.102165492071641e-05, + "loss": 2.4607, + "step": 29486 + }, + { + "epoch": 1.26, + "grad_norm": 0.45703125, + "learning_rate": 6.1015477630375596e-05, + "loss": 2.4774, + "step": 29487 + }, + { + "epoch": 1.26, + "grad_norm": 0.466796875, + "learning_rate": 6.100930051545091e-05, + "loss": 2.5631, + "step": 29488 + }, + { + "epoch": 1.26, + "grad_norm": 0.498046875, + "learning_rate": 6.1003123575970066e-05, + "loss": 2.3995, + "step": 29489 + }, + { + "epoch": 1.26, + "grad_norm": 0.443359375, + "learning_rate": 6.099694681196096e-05, + "loss": 2.4455, + "step": 29490 + }, + { + "epoch": 1.26, + "grad_norm": 0.443359375, + "learning_rate": 6.0990770223451334e-05, + "loss": 2.3797, + "step": 29491 + }, + { + "epoch": 1.26, + "grad_norm": 0.4296875, + "learning_rate": 6.098459381046898e-05, + "loss": 2.4021, + "step": 29492 + }, + { + "epoch": 1.26, + "grad_norm": 0.51171875, + "learning_rate": 6.097841757304168e-05, + "loss": 2.2975, + "step": 29493 + }, + { + "epoch": 1.26, + "grad_norm": 0.482421875, + "learning_rate": 6.097224151119725e-05, + "loss": 2.5198, + "step": 29494 + }, + { + "epoch": 1.26, + "grad_norm": 0.470703125, + "learning_rate": 6.096606562496348e-05, + "loss": 2.4078, + "step": 29495 + }, + { + "epoch": 1.26, + "grad_norm": 0.4921875, + "learning_rate": 6.0959889914368166e-05, + "loss": 2.6011, + "step": 29496 + }, + { + "epoch": 1.26, + "grad_norm": 0.435546875, + "learning_rate": 6.095371437943902e-05, + "loss": 2.4347, + "step": 29497 + }, + { + "epoch": 1.26, + "grad_norm": 0.45703125, + "learning_rate": 6.094753902020395e-05, + "loss": 2.3197, + "step": 29498 + }, + { + "epoch": 1.26, + "grad_norm": 0.435546875, + "learning_rate": 6.094136383669066e-05, + "loss": 2.4824, + "step": 29499 + }, + { + "epoch": 1.26, + "grad_norm": 0.4296875, + "learning_rate": 6.093518882892696e-05, + "loss": 2.4744, + "step": 29500 + }, + { + "epoch": 1.26, + "grad_norm": 0.47265625, + "learning_rate": 6.092901399694061e-05, + "loss": 2.3955, + "step": 29501 + }, + { + "epoch": 1.26, + "grad_norm": 0.4921875, + "learning_rate": 6.092283934075944e-05, + "loss": 2.5542, + "step": 29502 + }, + { + "epoch": 1.26, + "grad_norm": 0.46484375, + "learning_rate": 6.091666486041121e-05, + "loss": 2.5062, + "step": 29503 + }, + { + "epoch": 1.26, + "grad_norm": 0.439453125, + "learning_rate": 6.091049055592374e-05, + "loss": 2.467, + "step": 29504 + }, + { + "epoch": 1.26, + "grad_norm": 0.48828125, + "learning_rate": 6.09043164273247e-05, + "loss": 2.4346, + "step": 29505 + }, + { + "epoch": 1.26, + "grad_norm": 0.51171875, + "learning_rate": 6.0898142474642014e-05, + "loss": 2.5868, + "step": 29506 + }, + { + "epoch": 1.26, + "grad_norm": 0.5078125, + "learning_rate": 6.0891968697903376e-05, + "loss": 2.4559, + "step": 29507 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.088579509713659e-05, + "loss": 2.5007, + "step": 29508 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.0879621672369403e-05, + "loss": 2.4387, + "step": 29509 + }, + { + "epoch": 1.26, + "grad_norm": 0.5, + "learning_rate": 6.0873448423629654e-05, + "loss": 2.4232, + "step": 29510 + }, + { + "epoch": 1.26, + "grad_norm": 0.462890625, + "learning_rate": 6.086727535094507e-05, + "loss": 2.5175, + "step": 29511 + }, + { + "epoch": 1.26, + "grad_norm": 0.46875, + "learning_rate": 6.086110245434349e-05, + "loss": 2.3608, + "step": 29512 + }, + { + "epoch": 1.26, + "grad_norm": 0.47265625, + "learning_rate": 6.085492973385256e-05, + "loss": 2.3386, + "step": 29513 + }, + { + "epoch": 1.26, + "grad_norm": 0.4453125, + "learning_rate": 6.084875718950022e-05, + "loss": 2.3539, + "step": 29514 + }, + { + "epoch": 1.26, + "grad_norm": 0.458984375, + "learning_rate": 6.084258482131413e-05, + "loss": 2.4994, + "step": 29515 + }, + { + "epoch": 1.26, + "grad_norm": 0.390625, + "learning_rate": 6.0836412629322114e-05, + "loss": 2.347, + "step": 29516 + }, + { + "epoch": 1.26, + "grad_norm": 0.4296875, + "learning_rate": 6.0830240613551895e-05, + "loss": 2.4679, + "step": 29517 + }, + { + "epoch": 1.26, + "grad_norm": 0.4921875, + "learning_rate": 6.082406877403131e-05, + "loss": 2.4068, + "step": 29518 + }, + { + "epoch": 1.26, + "grad_norm": 0.46875, + "learning_rate": 6.081789711078808e-05, + "loss": 2.453, + "step": 29519 + }, + { + "epoch": 1.26, + "grad_norm": 0.52734375, + "learning_rate": 6.0811725623850026e-05, + "loss": 2.4089, + "step": 29520 + }, + { + "epoch": 1.26, + "grad_norm": 0.46875, + "learning_rate": 6.080555431324482e-05, + "loss": 2.5097, + "step": 29521 + }, + { + "epoch": 1.26, + "grad_norm": 0.478515625, + "learning_rate": 6.0799383179000355e-05, + "loss": 2.3409, + "step": 29522 + }, + { + "epoch": 1.26, + "grad_norm": 0.431640625, + "learning_rate": 6.0793212221144336e-05, + "loss": 2.4477, + "step": 29523 + }, + { + "epoch": 1.26, + "grad_norm": 0.48828125, + "learning_rate": 6.078704143970452e-05, + "loss": 2.3824, + "step": 29524 + }, + { + "epoch": 1.26, + "grad_norm": 0.435546875, + "learning_rate": 6.078087083470867e-05, + "loss": 2.5673, + "step": 29525 + }, + { + "epoch": 1.26, + "grad_norm": 0.51171875, + "learning_rate": 6.077470040618458e-05, + "loss": 2.3841, + "step": 29526 + }, + { + "epoch": 1.26, + "grad_norm": 0.53125, + "learning_rate": 6.076853015416001e-05, + "loss": 2.5099, + "step": 29527 + }, + { + "epoch": 1.26, + "grad_norm": 0.451171875, + "learning_rate": 6.076236007866274e-05, + "loss": 2.3657, + "step": 29528 + }, + { + "epoch": 1.26, + "grad_norm": 0.46484375, + "learning_rate": 6.075619017972044e-05, + "loss": 2.3979, + "step": 29529 + }, + { + "epoch": 1.26, + "grad_norm": 0.455078125, + "learning_rate": 6.075002045736101e-05, + "loss": 2.5331, + "step": 29530 + }, + { + "epoch": 1.26, + "grad_norm": 0.498046875, + "learning_rate": 6.0743850911612124e-05, + "loss": 2.5145, + "step": 29531 + }, + { + "epoch": 1.26, + "grad_norm": 0.494140625, + "learning_rate": 6.073768154250156e-05, + "loss": 2.6146, + "step": 29532 + }, + { + "epoch": 1.26, + "grad_norm": 0.478515625, + "learning_rate": 6.073151235005705e-05, + "loss": 2.4803, + "step": 29533 + }, + { + "epoch": 1.26, + "grad_norm": 0.49609375, + "learning_rate": 6.072534333430642e-05, + "loss": 2.5121, + "step": 29534 + }, + { + "epoch": 1.26, + "grad_norm": 0.46484375, + "learning_rate": 6.0719174495277385e-05, + "loss": 2.3982, + "step": 29535 + }, + { + "epoch": 1.26, + "grad_norm": 0.46484375, + "learning_rate": 6.0713005832997726e-05, + "loss": 2.314, + "step": 29536 + }, + { + "epoch": 1.26, + "grad_norm": 0.447265625, + "learning_rate": 6.070683734749514e-05, + "loss": 2.4553, + "step": 29537 + }, + { + "epoch": 1.26, + "grad_norm": 0.51171875, + "learning_rate": 6.070066903879748e-05, + "loss": 2.529, + "step": 29538 + }, + { + "epoch": 1.26, + "grad_norm": 0.44140625, + "learning_rate": 6.0694500906932406e-05, + "loss": 2.3957, + "step": 29539 + }, + { + "epoch": 1.26, + "grad_norm": 0.447265625, + "learning_rate": 6.0688332951927726e-05, + "loss": 2.5182, + "step": 29540 + }, + { + "epoch": 1.26, + "grad_norm": 0.4765625, + "learning_rate": 6.0682165173811165e-05, + "loss": 2.4761, + "step": 29541 + }, + { + "epoch": 1.26, + "grad_norm": 0.40625, + "learning_rate": 6.067599757261051e-05, + "loss": 2.3493, + "step": 29542 + }, + { + "epoch": 1.26, + "grad_norm": 0.45703125, + "learning_rate": 6.06698301483535e-05, + "loss": 2.3528, + "step": 29543 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.0663662901067885e-05, + "loss": 2.4703, + "step": 29544 + }, + { + "epoch": 1.26, + "grad_norm": 0.453125, + "learning_rate": 6.0657495830781366e-05, + "loss": 2.4556, + "step": 29545 + }, + { + "epoch": 1.26, + "grad_norm": 0.51171875, + "learning_rate": 6.065132893752179e-05, + "loss": 2.258, + "step": 29546 + }, + { + "epoch": 1.26, + "grad_norm": 0.484375, + "learning_rate": 6.0645162221316844e-05, + "loss": 2.5943, + "step": 29547 + }, + { + "epoch": 1.26, + "grad_norm": 0.4921875, + "learning_rate": 6.0638995682194264e-05, + "loss": 2.4891, + "step": 29548 + }, + { + "epoch": 1.26, + "grad_norm": 0.490234375, + "learning_rate": 6.0632829320181814e-05, + "loss": 2.4596, + "step": 29549 + }, + { + "epoch": 1.26, + "grad_norm": 0.466796875, + "learning_rate": 6.062666313530727e-05, + "loss": 2.4523, + "step": 29550 + }, + { + "epoch": 1.26, + "grad_norm": 0.5, + "learning_rate": 6.062049712759834e-05, + "loss": 2.4847, + "step": 29551 + }, + { + "epoch": 1.26, + "grad_norm": 0.380859375, + "learning_rate": 6.0614331297082805e-05, + "loss": 2.5162, + "step": 29552 + }, + { + "epoch": 1.26, + "grad_norm": 0.462890625, + "learning_rate": 6.060816564378833e-05, + "loss": 2.4754, + "step": 29553 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.060200016774278e-05, + "loss": 2.4767, + "step": 29554 + }, + { + "epoch": 1.26, + "grad_norm": 0.498046875, + "learning_rate": 6.05958348689738e-05, + "loss": 2.4843, + "step": 29555 + }, + { + "epoch": 1.26, + "grad_norm": 0.515625, + "learning_rate": 6.058966974750916e-05, + "loss": 2.5186, + "step": 29556 + }, + { + "epoch": 1.26, + "grad_norm": 0.439453125, + "learning_rate": 6.0583504803376576e-05, + "loss": 2.5272, + "step": 29557 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.057734003660386e-05, + "loss": 2.3831, + "step": 29558 + }, + { + "epoch": 1.26, + "grad_norm": 0.470703125, + "learning_rate": 6.0571175447218684e-05, + "loss": 2.5004, + "step": 29559 + }, + { + "epoch": 1.26, + "grad_norm": 0.400390625, + "learning_rate": 6.056501103524884e-05, + "loss": 2.3027, + "step": 29560 + }, + { + "epoch": 1.26, + "grad_norm": 0.50390625, + "learning_rate": 6.055884680072197e-05, + "loss": 2.6123, + "step": 29561 + }, + { + "epoch": 1.26, + "grad_norm": 0.4375, + "learning_rate": 6.0552682743665943e-05, + "loss": 2.3858, + "step": 29562 + }, + { + "epoch": 1.26, + "grad_norm": 0.416015625, + "learning_rate": 6.05465188641084e-05, + "loss": 2.332, + "step": 29563 + }, + { + "epoch": 1.26, + "grad_norm": 0.435546875, + "learning_rate": 6.054035516207711e-05, + "loss": 2.3682, + "step": 29564 + }, + { + "epoch": 1.26, + "grad_norm": 0.47265625, + "learning_rate": 6.053419163759978e-05, + "loss": 2.445, + "step": 29565 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.0528028290704185e-05, + "loss": 2.4449, + "step": 29566 + }, + { + "epoch": 1.26, + "grad_norm": 0.52734375, + "learning_rate": 6.052186512141802e-05, + "loss": 2.5419, + "step": 29567 + }, + { + "epoch": 1.26, + "grad_norm": 0.4375, + "learning_rate": 6.051570212976907e-05, + "loss": 2.4021, + "step": 29568 + }, + { + "epoch": 1.26, + "grad_norm": 0.451171875, + "learning_rate": 6.050953931578498e-05, + "loss": 2.4923, + "step": 29569 + }, + { + "epoch": 1.26, + "grad_norm": 0.474609375, + "learning_rate": 6.050337667949357e-05, + "loss": 2.4391, + "step": 29570 + }, + { + "epoch": 1.26, + "grad_norm": 0.451171875, + "learning_rate": 6.0497214220922516e-05, + "loss": 2.4192, + "step": 29571 + }, + { + "epoch": 1.26, + "grad_norm": 0.494140625, + "learning_rate": 6.0491051940099574e-05, + "loss": 2.3988, + "step": 29572 + }, + { + "epoch": 1.26, + "grad_norm": 0.46875, + "learning_rate": 6.048488983705243e-05, + "loss": 2.514, + "step": 29573 + }, + { + "epoch": 1.26, + "grad_norm": 0.486328125, + "learning_rate": 6.0478727911808865e-05, + "loss": 2.6072, + "step": 29574 + }, + { + "epoch": 1.26, + "grad_norm": 0.486328125, + "learning_rate": 6.0472566164396574e-05, + "loss": 2.4322, + "step": 29575 + }, + { + "epoch": 1.26, + "grad_norm": 0.50390625, + "learning_rate": 6.046640459484332e-05, + "loss": 2.5604, + "step": 29576 + }, + { + "epoch": 1.26, + "grad_norm": 0.482421875, + "learning_rate": 6.046024320317673e-05, + "loss": 2.598, + "step": 29577 + }, + { + "epoch": 1.26, + "grad_norm": 0.50390625, + "learning_rate": 6.045408198942465e-05, + "loss": 2.373, + "step": 29578 + }, + { + "epoch": 1.26, + "grad_norm": 0.42578125, + "learning_rate": 6.044792095361473e-05, + "loss": 2.4361, + "step": 29579 + }, + { + "epoch": 1.26, + "grad_norm": 0.455078125, + "learning_rate": 6.044176009577471e-05, + "loss": 2.4136, + "step": 29580 + }, + { + "epoch": 1.26, + "grad_norm": 0.435546875, + "learning_rate": 6.043559941593231e-05, + "loss": 2.3696, + "step": 29581 + }, + { + "epoch": 1.26, + "grad_norm": 0.48828125, + "learning_rate": 6.042943891411526e-05, + "loss": 2.4606, + "step": 29582 + }, + { + "epoch": 1.26, + "grad_norm": 0.4765625, + "learning_rate": 6.042327859035128e-05, + "loss": 2.4486, + "step": 29583 + }, + { + "epoch": 1.26, + "grad_norm": 0.51171875, + "learning_rate": 6.0417118444668096e-05, + "loss": 2.4791, + "step": 29584 + }, + { + "epoch": 1.26, + "grad_norm": 0.51953125, + "learning_rate": 6.0410958477093395e-05, + "loss": 2.4458, + "step": 29585 + }, + { + "epoch": 1.26, + "grad_norm": 0.4453125, + "learning_rate": 6.04047986876549e-05, + "loss": 2.5039, + "step": 29586 + }, + { + "epoch": 1.26, + "grad_norm": 0.423828125, + "learning_rate": 6.039863907638035e-05, + "loss": 2.4319, + "step": 29587 + }, + { + "epoch": 1.26, + "grad_norm": 0.56640625, + "learning_rate": 6.039247964329745e-05, + "loss": 2.5187, + "step": 29588 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.0386320388433924e-05, + "loss": 2.5083, + "step": 29589 + }, + { + "epoch": 1.26, + "grad_norm": 0.45703125, + "learning_rate": 6.0380161311817455e-05, + "loss": 2.3478, + "step": 29590 + }, + { + "epoch": 1.26, + "grad_norm": 0.5390625, + "learning_rate": 6.0374002413475807e-05, + "loss": 2.5238, + "step": 29591 + }, + { + "epoch": 1.26, + "grad_norm": 0.447265625, + "learning_rate": 6.0367843693436674e-05, + "loss": 2.5776, + "step": 29592 + }, + { + "epoch": 1.26, + "grad_norm": 0.47265625, + "learning_rate": 6.036168515172775e-05, + "loss": 2.4647, + "step": 29593 + }, + { + "epoch": 1.26, + "grad_norm": 0.578125, + "learning_rate": 6.0355526788376735e-05, + "loss": 2.5694, + "step": 29594 + }, + { + "epoch": 1.26, + "grad_norm": 0.5, + "learning_rate": 6.0349368603411385e-05, + "loss": 2.3794, + "step": 29595 + }, + { + "epoch": 1.26, + "grad_norm": 0.486328125, + "learning_rate": 6.034321059685939e-05, + "loss": 2.4368, + "step": 29596 + }, + { + "epoch": 1.26, + "grad_norm": 0.5390625, + "learning_rate": 6.033705276874844e-05, + "loss": 2.4371, + "step": 29597 + }, + { + "epoch": 1.26, + "grad_norm": 0.478515625, + "learning_rate": 6.033089511910624e-05, + "loss": 2.4268, + "step": 29598 + }, + { + "epoch": 1.26, + "grad_norm": 0.3984375, + "learning_rate": 6.0324737647960536e-05, + "loss": 2.528, + "step": 29599 + }, + { + "epoch": 1.26, + "grad_norm": 0.4453125, + "learning_rate": 6.0318580355339036e-05, + "loss": 2.3433, + "step": 29600 + }, + { + "epoch": 1.26, + "grad_norm": 0.494140625, + "learning_rate": 6.03124232412694e-05, + "loss": 2.5393, + "step": 29601 + }, + { + "epoch": 1.26, + "grad_norm": 0.49609375, + "learning_rate": 6.030626630577934e-05, + "loss": 2.4511, + "step": 29602 + }, + { + "epoch": 1.26, + "grad_norm": 0.484375, + "learning_rate": 6.03001095488966e-05, + "loss": 2.556, + "step": 29603 + }, + { + "epoch": 1.26, + "grad_norm": 0.376953125, + "learning_rate": 6.029395297064884e-05, + "loss": 2.3949, + "step": 29604 + }, + { + "epoch": 1.26, + "grad_norm": 0.490234375, + "learning_rate": 6.0287796571063804e-05, + "loss": 2.386, + "step": 29605 + }, + { + "epoch": 1.26, + "grad_norm": 0.41015625, + "learning_rate": 6.0281640350169124e-05, + "loss": 2.4155, + "step": 29606 + }, + { + "epoch": 1.26, + "grad_norm": 0.52734375, + "learning_rate": 6.027548430799259e-05, + "loss": 2.475, + "step": 29607 + }, + { + "epoch": 1.26, + "grad_norm": 0.451171875, + "learning_rate": 6.026932844456187e-05, + "loss": 2.5211, + "step": 29608 + }, + { + "epoch": 1.26, + "grad_norm": 0.498046875, + "learning_rate": 6.026317275990463e-05, + "loss": 2.4792, + "step": 29609 + }, + { + "epoch": 1.26, + "grad_norm": 0.515625, + "learning_rate": 6.0257017254048575e-05, + "loss": 2.5357, + "step": 29610 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.025086192702143e-05, + "loss": 2.4529, + "step": 29611 + }, + { + "epoch": 1.26, + "grad_norm": 0.5625, + "learning_rate": 6.024470677885088e-05, + "loss": 2.468, + "step": 29612 + }, + { + "epoch": 1.26, + "grad_norm": 0.474609375, + "learning_rate": 6.023855180956464e-05, + "loss": 2.6001, + "step": 29613 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 6.0232397019190346e-05, + "loss": 2.5451, + "step": 29614 + }, + { + "epoch": 1.26, + "grad_norm": 0.482421875, + "learning_rate": 6.022624240775575e-05, + "loss": 2.4155, + "step": 29615 + }, + { + "epoch": 1.26, + "grad_norm": 0.490234375, + "learning_rate": 6.0220087975288556e-05, + "loss": 2.5529, + "step": 29616 + }, + { + "epoch": 1.26, + "grad_norm": 0.451171875, + "learning_rate": 6.0213933721816405e-05, + "loss": 2.4032, + "step": 29617 + }, + { + "epoch": 1.26, + "grad_norm": 0.447265625, + "learning_rate": 6.020777964736699e-05, + "loss": 2.5965, + "step": 29618 + }, + { + "epoch": 1.26, + "grad_norm": 0.404296875, + "learning_rate": 6.020162575196806e-05, + "loss": 2.4978, + "step": 29619 + }, + { + "epoch": 1.26, + "grad_norm": 0.451171875, + "learning_rate": 6.019547203564726e-05, + "loss": 2.4389, + "step": 29620 + }, + { + "epoch": 1.26, + "grad_norm": 0.451171875, + "learning_rate": 6.018931849843228e-05, + "loss": 2.3755, + "step": 29621 + }, + { + "epoch": 1.26, + "grad_norm": 0.490234375, + "learning_rate": 6.0183165140350806e-05, + "loss": 2.4738, + "step": 29622 + }, + { + "epoch": 1.26, + "grad_norm": 0.41015625, + "learning_rate": 6.0177011961430554e-05, + "loss": 2.4434, + "step": 29623 + }, + { + "epoch": 1.26, + "grad_norm": 0.4765625, + "learning_rate": 6.0170858961699215e-05, + "loss": 2.5028, + "step": 29624 + }, + { + "epoch": 1.26, + "grad_norm": 0.44140625, + "learning_rate": 6.0164706141184434e-05, + "loss": 2.4069, + "step": 29625 + }, + { + "epoch": 1.26, + "grad_norm": 0.41796875, + "learning_rate": 6.015855349991391e-05, + "loss": 2.4078, + "step": 29626 + }, + { + "epoch": 1.26, + "grad_norm": 0.44921875, + "learning_rate": 6.015240103791533e-05, + "loss": 2.3012, + "step": 29627 + }, + { + "epoch": 1.26, + "grad_norm": 0.44140625, + "learning_rate": 6.014624875521641e-05, + "loss": 2.4305, + "step": 29628 + }, + { + "epoch": 1.26, + "grad_norm": 0.45703125, + "learning_rate": 6.014009665184479e-05, + "loss": 2.5195, + "step": 29629 + }, + { + "epoch": 1.26, + "grad_norm": 0.51953125, + "learning_rate": 6.013394472782814e-05, + "loss": 2.558, + "step": 29630 + }, + { + "epoch": 1.26, + "grad_norm": 0.50390625, + "learning_rate": 6.01277929831942e-05, + "loss": 2.4946, + "step": 29631 + }, + { + "epoch": 1.26, + "grad_norm": 0.515625, + "learning_rate": 6.0121641417970653e-05, + "loss": 2.4615, + "step": 29632 + }, + { + "epoch": 1.26, + "grad_norm": 0.5, + "learning_rate": 6.011549003218511e-05, + "loss": 2.5889, + "step": 29633 + }, + { + "epoch": 1.26, + "grad_norm": 0.462890625, + "learning_rate": 6.010933882586527e-05, + "loss": 2.5122, + "step": 29634 + }, + { + "epoch": 1.26, + "grad_norm": 0.5234375, + "learning_rate": 6.010318779903884e-05, + "loss": 2.5821, + "step": 29635 + }, + { + "epoch": 1.26, + "grad_norm": 0.50390625, + "learning_rate": 6.0097036951733486e-05, + "loss": 2.4842, + "step": 29636 + }, + { + "epoch": 1.26, + "grad_norm": 0.455078125, + "learning_rate": 6.009088628397689e-05, + "loss": 2.4313, + "step": 29637 + }, + { + "epoch": 1.26, + "grad_norm": 0.4453125, + "learning_rate": 6.00847357957967e-05, + "loss": 2.4882, + "step": 29638 + }, + { + "epoch": 1.26, + "grad_norm": 0.494140625, + "learning_rate": 6.0078585487220626e-05, + "loss": 2.5255, + "step": 29639 + }, + { + "epoch": 1.26, + "grad_norm": 0.47265625, + "learning_rate": 6.007243535827635e-05, + "loss": 2.4727, + "step": 29640 + }, + { + "epoch": 1.26, + "grad_norm": 0.486328125, + "learning_rate": 6.006628540899151e-05, + "loss": 2.4056, + "step": 29641 + }, + { + "epoch": 1.26, + "grad_norm": 0.470703125, + "learning_rate": 6.006013563939376e-05, + "loss": 2.4507, + "step": 29642 + }, + { + "epoch": 1.26, + "grad_norm": 0.43359375, + "learning_rate": 6.0053986049510824e-05, + "loss": 2.3615, + "step": 29643 + }, + { + "epoch": 1.26, + "grad_norm": 0.490234375, + "learning_rate": 6.0047836639370355e-05, + "loss": 2.4216, + "step": 29644 + }, + { + "epoch": 1.26, + "grad_norm": 0.4765625, + "learning_rate": 6.0041687409000026e-05, + "loss": 2.4756, + "step": 29645 + }, + { + "epoch": 1.26, + "grad_norm": 0.490234375, + "learning_rate": 6.0035538358427476e-05, + "loss": 2.3697, + "step": 29646 + }, + { + "epoch": 1.26, + "grad_norm": 0.494140625, + "learning_rate": 6.002938948768042e-05, + "loss": 2.5494, + "step": 29647 + }, + { + "epoch": 1.26, + "grad_norm": 0.4921875, + "learning_rate": 6.002324079678652e-05, + "loss": 2.4104, + "step": 29648 + }, + { + "epoch": 1.26, + "grad_norm": 0.419921875, + "learning_rate": 6.0017092285773416e-05, + "loss": 2.4666, + "step": 29649 + }, + { + "epoch": 1.26, + "grad_norm": 0.462890625, + "learning_rate": 6.0010943954668754e-05, + "loss": 2.4894, + "step": 29650 + }, + { + "epoch": 1.26, + "grad_norm": 0.5546875, + "learning_rate": 6.000479580350025e-05, + "loss": 2.5662, + "step": 29651 + }, + { + "epoch": 1.26, + "grad_norm": 0.4453125, + "learning_rate": 5.999864783229554e-05, + "loss": 2.3207, + "step": 29652 + }, + { + "epoch": 1.26, + "grad_norm": 0.47265625, + "learning_rate": 5.999250004108231e-05, + "loss": 2.5134, + "step": 29653 + }, + { + "epoch": 1.26, + "grad_norm": 0.439453125, + "learning_rate": 5.9986352429888195e-05, + "loss": 2.4359, + "step": 29654 + }, + { + "epoch": 1.26, + "grad_norm": 0.53515625, + "learning_rate": 5.9980204998740874e-05, + "loss": 2.5849, + "step": 29655 + }, + { + "epoch": 1.26, + "grad_norm": 0.46875, + "learning_rate": 5.9974057747668035e-05, + "loss": 2.4332, + "step": 29656 + }, + { + "epoch": 1.26, + "grad_norm": 0.53515625, + "learning_rate": 5.996791067669729e-05, + "loss": 2.4569, + "step": 29657 + }, + { + "epoch": 1.26, + "grad_norm": 0.4765625, + "learning_rate": 5.9961763785856295e-05, + "loss": 2.4874, + "step": 29658 + }, + { + "epoch": 1.26, + "grad_norm": 0.447265625, + "learning_rate": 5.995561707517274e-05, + "loss": 2.4895, + "step": 29659 + }, + { + "epoch": 1.26, + "grad_norm": 0.48828125, + "learning_rate": 5.994947054467428e-05, + "loss": 2.4505, + "step": 29660 + }, + { + "epoch": 1.26, + "grad_norm": 0.53125, + "learning_rate": 5.994332419438857e-05, + "loss": 2.4527, + "step": 29661 + }, + { + "epoch": 1.26, + "grad_norm": 0.52734375, + "learning_rate": 5.9937178024343245e-05, + "loss": 2.5565, + "step": 29662 + }, + { + "epoch": 1.26, + "grad_norm": 0.5, + "learning_rate": 5.993103203456598e-05, + "loss": 2.5593, + "step": 29663 + }, + { + "epoch": 1.26, + "grad_norm": 0.443359375, + "learning_rate": 5.992488622508448e-05, + "loss": 2.4136, + "step": 29664 + }, + { + "epoch": 1.26, + "grad_norm": 0.42578125, + "learning_rate": 5.991874059592629e-05, + "loss": 2.6418, + "step": 29665 + }, + { + "epoch": 1.26, + "grad_norm": 0.421875, + "learning_rate": 5.991259514711912e-05, + "loss": 2.43, + "step": 29666 + }, + { + "epoch": 1.26, + "grad_norm": 0.412109375, + "learning_rate": 5.990644987869063e-05, + "loss": 2.5156, + "step": 29667 + }, + { + "epoch": 1.26, + "grad_norm": 0.419921875, + "learning_rate": 5.9900304790668486e-05, + "loss": 2.4233, + "step": 29668 + }, + { + "epoch": 1.26, + "grad_norm": 0.4921875, + "learning_rate": 5.989415988308029e-05, + "loss": 2.442, + "step": 29669 + }, + { + "epoch": 1.26, + "grad_norm": 0.478515625, + "learning_rate": 5.98880151559537e-05, + "loss": 2.4567, + "step": 29670 + }, + { + "epoch": 1.26, + "grad_norm": 0.41015625, + "learning_rate": 5.9881870609316415e-05, + "loss": 2.532, + "step": 29671 + }, + { + "epoch": 1.26, + "grad_norm": 0.5, + "learning_rate": 5.987572624319606e-05, + "loss": 2.44, + "step": 29672 + }, + { + "epoch": 1.26, + "grad_norm": 0.466796875, + "learning_rate": 5.986958205762027e-05, + "loss": 2.44, + "step": 29673 + }, + { + "epoch": 1.26, + "grad_norm": 0.474609375, + "learning_rate": 5.986343805261664e-05, + "loss": 2.5747, + "step": 29674 + }, + { + "epoch": 1.26, + "grad_norm": 0.4609375, + "learning_rate": 5.985729422821293e-05, + "loss": 2.4174, + "step": 29675 + }, + { + "epoch": 1.26, + "grad_norm": 0.435546875, + "learning_rate": 5.985115058443671e-05, + "loss": 2.4634, + "step": 29676 + }, + { + "epoch": 1.26, + "grad_norm": 0.451171875, + "learning_rate": 5.984500712131564e-05, + "loss": 2.4189, + "step": 29677 + }, + { + "epoch": 1.26, + "grad_norm": 0.494140625, + "learning_rate": 5.983886383887735e-05, + "loss": 2.5632, + "step": 29678 + }, + { + "epoch": 1.26, + "grad_norm": 0.478515625, + "learning_rate": 5.983272073714949e-05, + "loss": 2.5067, + "step": 29679 + }, + { + "epoch": 1.26, + "grad_norm": 0.447265625, + "learning_rate": 5.982657781615976e-05, + "loss": 2.4409, + "step": 29680 + }, + { + "epoch": 1.26, + "grad_norm": 0.486328125, + "learning_rate": 5.982043507593571e-05, + "loss": 2.5574, + "step": 29681 + }, + { + "epoch": 1.26, + "grad_norm": 0.48828125, + "learning_rate": 5.9814292516504976e-05, + "loss": 2.3942, + "step": 29682 + }, + { + "epoch": 1.26, + "grad_norm": 0.44140625, + "learning_rate": 5.9808150137895315e-05, + "loss": 2.4676, + "step": 29683 + }, + { + "epoch": 1.26, + "grad_norm": 0.443359375, + "learning_rate": 5.9802007940134255e-05, + "loss": 2.4175, + "step": 29684 + }, + { + "epoch": 1.26, + "grad_norm": 0.46875, + "learning_rate": 5.979586592324947e-05, + "loss": 2.4534, + "step": 29685 + }, + { + "epoch": 1.27, + "grad_norm": 0.546875, + "learning_rate": 5.978972408726857e-05, + "loss": 2.5135, + "step": 29686 + }, + { + "epoch": 1.27, + "grad_norm": 0.421875, + "learning_rate": 5.9783582432219255e-05, + "loss": 2.4902, + "step": 29687 + }, + { + "epoch": 1.27, + "grad_norm": 0.458984375, + "learning_rate": 5.9777440958129135e-05, + "loss": 2.5258, + "step": 29688 + }, + { + "epoch": 1.27, + "grad_norm": 0.46875, + "learning_rate": 5.9771299665025806e-05, + "loss": 2.4154, + "step": 29689 + }, + { + "epoch": 1.27, + "grad_norm": 0.494140625, + "learning_rate": 5.976515855293688e-05, + "loss": 2.6203, + "step": 29690 + }, + { + "epoch": 1.27, + "grad_norm": 0.4765625, + "learning_rate": 5.9759017621890114e-05, + "loss": 2.547, + "step": 29691 + }, + { + "epoch": 1.27, + "grad_norm": 0.458984375, + "learning_rate": 5.975287687191302e-05, + "loss": 2.3336, + "step": 29692 + }, + { + "epoch": 1.27, + "grad_norm": 0.5, + "learning_rate": 5.974673630303329e-05, + "loss": 2.3899, + "step": 29693 + }, + { + "epoch": 1.27, + "grad_norm": 0.48828125, + "learning_rate": 5.974059591527852e-05, + "loss": 2.5328, + "step": 29694 + }, + { + "epoch": 1.27, + "grad_norm": 0.53125, + "learning_rate": 5.973445570867635e-05, + "loss": 2.6259, + "step": 29695 + }, + { + "epoch": 1.27, + "grad_norm": 0.52734375, + "learning_rate": 5.9728315683254454e-05, + "loss": 2.6591, + "step": 29696 + }, + { + "epoch": 1.27, + "grad_norm": 0.427734375, + "learning_rate": 5.9722175839040406e-05, + "loss": 2.4002, + "step": 29697 + }, + { + "epoch": 1.27, + "grad_norm": 0.50390625, + "learning_rate": 5.971603617606178e-05, + "loss": 2.4605, + "step": 29698 + }, + { + "epoch": 1.27, + "grad_norm": 0.431640625, + "learning_rate": 5.9709896694346346e-05, + "loss": 2.4499, + "step": 29699 + }, + { + "epoch": 1.27, + "grad_norm": 0.44140625, + "learning_rate": 5.970375739392164e-05, + "loss": 2.4797, + "step": 29700 + }, + { + "epoch": 1.27, + "grad_norm": 0.46875, + "learning_rate": 5.969761827481529e-05, + "loss": 2.3902, + "step": 29701 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.96914793370549e-05, + "loss": 2.3798, + "step": 29702 + }, + { + "epoch": 1.27, + "grad_norm": 0.45703125, + "learning_rate": 5.968534058066816e-05, + "loss": 2.449, + "step": 29703 + }, + { + "epoch": 1.27, + "grad_norm": 0.4296875, + "learning_rate": 5.9679202005682664e-05, + "loss": 2.4945, + "step": 29704 + }, + { + "epoch": 1.27, + "grad_norm": 0.49609375, + "learning_rate": 5.967306361212601e-05, + "loss": 2.453, + "step": 29705 + }, + { + "epoch": 1.27, + "grad_norm": 0.455078125, + "learning_rate": 5.966692540002583e-05, + "loss": 2.573, + "step": 29706 + }, + { + "epoch": 1.27, + "grad_norm": 0.462890625, + "learning_rate": 5.966078736940972e-05, + "loss": 2.4925, + "step": 29707 + }, + { + "epoch": 1.27, + "grad_norm": 0.4375, + "learning_rate": 5.9654649520305353e-05, + "loss": 2.4575, + "step": 29708 + }, + { + "epoch": 1.27, + "grad_norm": 0.4296875, + "learning_rate": 5.9648511852740296e-05, + "loss": 2.4672, + "step": 29709 + }, + { + "epoch": 1.27, + "grad_norm": 0.494140625, + "learning_rate": 5.9642374366742226e-05, + "loss": 2.4159, + "step": 29710 + }, + { + "epoch": 1.27, + "grad_norm": 0.44921875, + "learning_rate": 5.9636237062338675e-05, + "loss": 2.4851, + "step": 29711 + }, + { + "epoch": 1.27, + "grad_norm": 0.474609375, + "learning_rate": 5.963009993955735e-05, + "loss": 2.4978, + "step": 29712 + }, + { + "epoch": 1.27, + "grad_norm": 0.498046875, + "learning_rate": 5.962396299842581e-05, + "loss": 2.3752, + "step": 29713 + }, + { + "epoch": 1.27, + "grad_norm": 0.50390625, + "learning_rate": 5.961782623897167e-05, + "loss": 2.3799, + "step": 29714 + }, + { + "epoch": 1.27, + "grad_norm": 0.4921875, + "learning_rate": 5.9611689661222535e-05, + "loss": 2.48, + "step": 29715 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.960555326520606e-05, + "loss": 2.4757, + "step": 29716 + }, + { + "epoch": 1.27, + "grad_norm": 0.484375, + "learning_rate": 5.9599417050949835e-05, + "loss": 2.5097, + "step": 29717 + }, + { + "epoch": 1.27, + "grad_norm": 0.45703125, + "learning_rate": 5.959328101848149e-05, + "loss": 2.532, + "step": 29718 + }, + { + "epoch": 1.27, + "grad_norm": 0.455078125, + "learning_rate": 5.9587145167828554e-05, + "loss": 2.5743, + "step": 29719 + }, + { + "epoch": 1.27, + "grad_norm": 0.50390625, + "learning_rate": 5.958100949901875e-05, + "loss": 2.4596, + "step": 29720 + }, + { + "epoch": 1.27, + "grad_norm": 0.458984375, + "learning_rate": 5.9574874012079605e-05, + "loss": 2.5572, + "step": 29721 + }, + { + "epoch": 1.27, + "grad_norm": 0.484375, + "learning_rate": 5.9568738707038765e-05, + "loss": 2.5741, + "step": 29722 + }, + { + "epoch": 1.27, + "grad_norm": 0.478515625, + "learning_rate": 5.9562603583923805e-05, + "loss": 2.4552, + "step": 29723 + }, + { + "epoch": 1.27, + "grad_norm": 0.46875, + "learning_rate": 5.9556468642762364e-05, + "loss": 2.3944, + "step": 29724 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.955033388358204e-05, + "loss": 2.4522, + "step": 29725 + }, + { + "epoch": 1.27, + "grad_norm": 0.484375, + "learning_rate": 5.9544199306410444e-05, + "loss": 2.5503, + "step": 29726 + }, + { + "epoch": 1.27, + "grad_norm": 0.4609375, + "learning_rate": 5.953806491127512e-05, + "loss": 2.41, + "step": 29727 + }, + { + "epoch": 1.27, + "grad_norm": 0.490234375, + "learning_rate": 5.953193069820374e-05, + "loss": 2.4693, + "step": 29728 + }, + { + "epoch": 1.27, + "grad_norm": 0.494140625, + "learning_rate": 5.9525796667223906e-05, + "loss": 2.4748, + "step": 29729 + }, + { + "epoch": 1.27, + "grad_norm": 0.392578125, + "learning_rate": 5.951966281836319e-05, + "loss": 2.4365, + "step": 29730 + }, + { + "epoch": 1.27, + "grad_norm": 0.46484375, + "learning_rate": 5.951352915164916e-05, + "loss": 2.4218, + "step": 29731 + }, + { + "epoch": 1.27, + "grad_norm": 0.494140625, + "learning_rate": 5.950739566710948e-05, + "loss": 2.368, + "step": 29732 + }, + { + "epoch": 1.27, + "grad_norm": 0.53125, + "learning_rate": 5.950126236477172e-05, + "loss": 2.4854, + "step": 29733 + }, + { + "epoch": 1.27, + "grad_norm": 0.4609375, + "learning_rate": 5.949512924466349e-05, + "loss": 2.4274, + "step": 29734 + }, + { + "epoch": 1.27, + "grad_norm": 0.46875, + "learning_rate": 5.948899630681233e-05, + "loss": 2.4442, + "step": 29735 + }, + { + "epoch": 1.27, + "grad_norm": 0.427734375, + "learning_rate": 5.9482863551245916e-05, + "loss": 2.4649, + "step": 29736 + }, + { + "epoch": 1.27, + "grad_norm": 0.482421875, + "learning_rate": 5.9476730977991825e-05, + "loss": 2.5449, + "step": 29737 + }, + { + "epoch": 1.27, + "grad_norm": 0.490234375, + "learning_rate": 5.947059858707762e-05, + "loss": 2.4816, + "step": 29738 + }, + { + "epoch": 1.27, + "grad_norm": 0.51953125, + "learning_rate": 5.946446637853089e-05, + "loss": 2.5311, + "step": 29739 + }, + { + "epoch": 1.27, + "grad_norm": 0.44921875, + "learning_rate": 5.9458334352379265e-05, + "loss": 2.4338, + "step": 29740 + }, + { + "epoch": 1.27, + "grad_norm": 0.498046875, + "learning_rate": 5.9452202508650324e-05, + "loss": 2.4394, + "step": 29741 + }, + { + "epoch": 1.27, + "grad_norm": 0.474609375, + "learning_rate": 5.944607084737167e-05, + "loss": 2.4945, + "step": 29742 + }, + { + "epoch": 1.27, + "grad_norm": 0.466796875, + "learning_rate": 5.9439939368570815e-05, + "loss": 2.4809, + "step": 29743 + }, + { + "epoch": 1.27, + "grad_norm": 0.427734375, + "learning_rate": 5.943380807227545e-05, + "loss": 2.5101, + "step": 29744 + }, + { + "epoch": 1.27, + "grad_norm": 0.451171875, + "learning_rate": 5.9427676958513143e-05, + "loss": 2.4529, + "step": 29745 + }, + { + "epoch": 1.27, + "grad_norm": 0.466796875, + "learning_rate": 5.9421546027311445e-05, + "loss": 2.452, + "step": 29746 + }, + { + "epoch": 1.27, + "grad_norm": 0.451171875, + "learning_rate": 5.941541527869794e-05, + "loss": 2.4753, + "step": 29747 + }, + { + "epoch": 1.27, + "grad_norm": 0.4765625, + "learning_rate": 5.940928471270024e-05, + "loss": 2.4274, + "step": 29748 + }, + { + "epoch": 1.27, + "grad_norm": 0.44921875, + "learning_rate": 5.940315432934595e-05, + "loss": 2.5262, + "step": 29749 + }, + { + "epoch": 1.27, + "grad_norm": 0.462890625, + "learning_rate": 5.939702412866264e-05, + "loss": 2.4364, + "step": 29750 + }, + { + "epoch": 1.27, + "grad_norm": 0.41015625, + "learning_rate": 5.939089411067781e-05, + "loss": 2.5531, + "step": 29751 + }, + { + "epoch": 1.27, + "grad_norm": 0.474609375, + "learning_rate": 5.938476427541917e-05, + "loss": 2.5098, + "step": 29752 + }, + { + "epoch": 1.27, + "grad_norm": 0.458984375, + "learning_rate": 5.937863462291425e-05, + "loss": 2.5134, + "step": 29753 + }, + { + "epoch": 1.27, + "grad_norm": 0.4453125, + "learning_rate": 5.9372505153190626e-05, + "loss": 2.3561, + "step": 29754 + }, + { + "epoch": 1.27, + "grad_norm": 0.5234375, + "learning_rate": 5.936637586627585e-05, + "loss": 2.4159, + "step": 29755 + }, + { + "epoch": 1.27, + "grad_norm": 0.4765625, + "learning_rate": 5.9360246762197557e-05, + "loss": 2.5168, + "step": 29756 + }, + { + "epoch": 1.27, + "grad_norm": 0.453125, + "learning_rate": 5.9354117840983304e-05, + "loss": 2.4714, + "step": 29757 + }, + { + "epoch": 1.27, + "grad_norm": 0.431640625, + "learning_rate": 5.934798910266068e-05, + "loss": 2.3844, + "step": 29758 + }, + { + "epoch": 1.27, + "grad_norm": 0.494140625, + "learning_rate": 5.934186054725721e-05, + "loss": 2.4218, + "step": 29759 + }, + { + "epoch": 1.27, + "grad_norm": 0.474609375, + "learning_rate": 5.933573217480053e-05, + "loss": 2.4525, + "step": 29760 + }, + { + "epoch": 1.27, + "grad_norm": 0.435546875, + "learning_rate": 5.9329603985318215e-05, + "loss": 2.5283, + "step": 29761 + }, + { + "epoch": 1.27, + "grad_norm": 0.50390625, + "learning_rate": 5.9323475978837807e-05, + "loss": 2.3941, + "step": 29762 + }, + { + "epoch": 1.27, + "grad_norm": 0.5078125, + "learning_rate": 5.931734815538685e-05, + "loss": 2.5397, + "step": 29763 + }, + { + "epoch": 1.27, + "grad_norm": 0.474609375, + "learning_rate": 5.931122051499301e-05, + "loss": 2.5175, + "step": 29764 + }, + { + "epoch": 1.27, + "grad_norm": 0.455078125, + "learning_rate": 5.930509305768378e-05, + "loss": 2.4763, + "step": 29765 + }, + { + "epoch": 1.27, + "grad_norm": 0.5, + "learning_rate": 5.9298965783486794e-05, + "loss": 2.4624, + "step": 29766 + }, + { + "epoch": 1.27, + "grad_norm": 0.48046875, + "learning_rate": 5.929283869242952e-05, + "loss": 2.3983, + "step": 29767 + }, + { + "epoch": 1.27, + "grad_norm": 0.458984375, + "learning_rate": 5.928671178453964e-05, + "loss": 2.3615, + "step": 29768 + }, + { + "epoch": 1.27, + "grad_norm": 0.451171875, + "learning_rate": 5.9280585059844696e-05, + "loss": 2.3815, + "step": 29769 + }, + { + "epoch": 1.27, + "grad_norm": 0.478515625, + "learning_rate": 5.927445851837221e-05, + "loss": 2.4423, + "step": 29770 + }, + { + "epoch": 1.27, + "grad_norm": 0.462890625, + "learning_rate": 5.926833216014978e-05, + "loss": 2.3277, + "step": 29771 + }, + { + "epoch": 1.27, + "grad_norm": 0.421875, + "learning_rate": 5.926220598520496e-05, + "loss": 2.3974, + "step": 29772 + }, + { + "epoch": 1.27, + "grad_norm": 0.53515625, + "learning_rate": 5.925607999356535e-05, + "loss": 2.5576, + "step": 29773 + }, + { + "epoch": 1.27, + "grad_norm": 0.38671875, + "learning_rate": 5.92499541852585e-05, + "loss": 2.3792, + "step": 29774 + }, + { + "epoch": 1.27, + "grad_norm": 0.5703125, + "learning_rate": 5.9243828560311895e-05, + "loss": 2.5596, + "step": 29775 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.923770311875323e-05, + "loss": 2.4055, + "step": 29776 + }, + { + "epoch": 1.27, + "grad_norm": 0.451171875, + "learning_rate": 5.923157786061e-05, + "loss": 2.3806, + "step": 29777 + }, + { + "epoch": 1.27, + "grad_norm": 0.49609375, + "learning_rate": 5.922545278590976e-05, + "loss": 2.4155, + "step": 29778 + }, + { + "epoch": 1.27, + "grad_norm": 0.427734375, + "learning_rate": 5.9219327894680066e-05, + "loss": 2.4478, + "step": 29779 + }, + { + "epoch": 1.27, + "grad_norm": 0.46484375, + "learning_rate": 5.9213203186948505e-05, + "loss": 2.4504, + "step": 29780 + }, + { + "epoch": 1.27, + "grad_norm": 0.470703125, + "learning_rate": 5.9207078662742624e-05, + "loss": 2.4442, + "step": 29781 + }, + { + "epoch": 1.27, + "grad_norm": 0.49609375, + "learning_rate": 5.920095432208998e-05, + "loss": 2.4303, + "step": 29782 + }, + { + "epoch": 1.27, + "grad_norm": 0.4140625, + "learning_rate": 5.9194830165018125e-05, + "loss": 2.4743, + "step": 29783 + }, + { + "epoch": 1.27, + "grad_norm": 0.47265625, + "learning_rate": 5.918870619155463e-05, + "loss": 2.4961, + "step": 29784 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.918258240172706e-05, + "loss": 2.4429, + "step": 29785 + }, + { + "epoch": 1.27, + "grad_norm": 0.490234375, + "learning_rate": 5.917645879556294e-05, + "loss": 2.4797, + "step": 29786 + }, + { + "epoch": 1.27, + "grad_norm": 0.4765625, + "learning_rate": 5.9170335373089804e-05, + "loss": 2.4826, + "step": 29787 + }, + { + "epoch": 1.27, + "grad_norm": 0.462890625, + "learning_rate": 5.916421213433527e-05, + "loss": 2.49, + "step": 29788 + }, + { + "epoch": 1.27, + "grad_norm": 0.431640625, + "learning_rate": 5.915808907932685e-05, + "loss": 2.5527, + "step": 29789 + }, + { + "epoch": 1.27, + "grad_norm": 0.453125, + "learning_rate": 5.9151966208092115e-05, + "loss": 2.371, + "step": 29790 + }, + { + "epoch": 1.27, + "grad_norm": 0.45703125, + "learning_rate": 5.914584352065857e-05, + "loss": 2.4144, + "step": 29791 + }, + { + "epoch": 1.27, + "grad_norm": 0.49609375, + "learning_rate": 5.9139721017053826e-05, + "loss": 2.2667, + "step": 29792 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.9133598697305414e-05, + "loss": 2.5381, + "step": 29793 + }, + { + "epoch": 1.27, + "grad_norm": 0.53515625, + "learning_rate": 5.912747656144086e-05, + "loss": 2.5306, + "step": 29794 + }, + { + "epoch": 1.27, + "grad_norm": 0.474609375, + "learning_rate": 5.9121354609487714e-05, + "loss": 2.5115, + "step": 29795 + }, + { + "epoch": 1.27, + "grad_norm": 0.494140625, + "learning_rate": 5.9115232841473545e-05, + "loss": 2.4714, + "step": 29796 + }, + { + "epoch": 1.27, + "grad_norm": 0.54296875, + "learning_rate": 5.9109111257425886e-05, + "loss": 2.2604, + "step": 29797 + }, + { + "epoch": 1.27, + "grad_norm": 0.44140625, + "learning_rate": 5.910298985737228e-05, + "loss": 2.6539, + "step": 29798 + }, + { + "epoch": 1.27, + "grad_norm": 0.431640625, + "learning_rate": 5.909686864134025e-05, + "loss": 2.3445, + "step": 29799 + }, + { + "epoch": 1.27, + "grad_norm": 0.439453125, + "learning_rate": 5.90907476093574e-05, + "loss": 2.4325, + "step": 29800 + }, + { + "epoch": 1.27, + "grad_norm": 0.5, + "learning_rate": 5.9084626761451237e-05, + "loss": 2.5354, + "step": 29801 + }, + { + "epoch": 1.27, + "grad_norm": 0.51171875, + "learning_rate": 5.907850609764929e-05, + "loss": 2.5361, + "step": 29802 + }, + { + "epoch": 1.27, + "grad_norm": 0.447265625, + "learning_rate": 5.907238561797909e-05, + "loss": 2.3911, + "step": 29803 + }, + { + "epoch": 1.27, + "grad_norm": 0.470703125, + "learning_rate": 5.906626532246822e-05, + "loss": 2.2682, + "step": 29804 + }, + { + "epoch": 1.27, + "grad_norm": 0.490234375, + "learning_rate": 5.906014521114418e-05, + "loss": 2.4913, + "step": 29805 + }, + { + "epoch": 1.27, + "grad_norm": 0.470703125, + "learning_rate": 5.9054025284034544e-05, + "loss": 2.3453, + "step": 29806 + }, + { + "epoch": 1.27, + "grad_norm": 0.46875, + "learning_rate": 5.904790554116679e-05, + "loss": 2.5172, + "step": 29807 + }, + { + "epoch": 1.27, + "grad_norm": 0.5, + "learning_rate": 5.904178598256853e-05, + "loss": 2.609, + "step": 29808 + }, + { + "epoch": 1.27, + "grad_norm": 0.439453125, + "learning_rate": 5.903566660826729e-05, + "loss": 2.4434, + "step": 29809 + }, + { + "epoch": 1.27, + "grad_norm": 0.546875, + "learning_rate": 5.902954741829054e-05, + "loss": 2.4137, + "step": 29810 + }, + { + "epoch": 1.27, + "grad_norm": 0.43359375, + "learning_rate": 5.902342841266586e-05, + "loss": 2.5051, + "step": 29811 + }, + { + "epoch": 1.27, + "grad_norm": 0.46875, + "learning_rate": 5.901730959142078e-05, + "loss": 2.5884, + "step": 29812 + }, + { + "epoch": 1.27, + "grad_norm": 0.482421875, + "learning_rate": 5.901119095458283e-05, + "loss": 2.5751, + "step": 29813 + }, + { + "epoch": 1.27, + "grad_norm": 0.484375, + "learning_rate": 5.900507250217955e-05, + "loss": 2.575, + "step": 29814 + }, + { + "epoch": 1.27, + "grad_norm": 0.486328125, + "learning_rate": 5.8998954234238426e-05, + "loss": 2.4661, + "step": 29815 + }, + { + "epoch": 1.27, + "grad_norm": 0.482421875, + "learning_rate": 5.899283615078706e-05, + "loss": 2.4556, + "step": 29816 + }, + { + "epoch": 1.27, + "grad_norm": 0.5078125, + "learning_rate": 5.898671825185296e-05, + "loss": 2.4814, + "step": 29817 + }, + { + "epoch": 1.27, + "grad_norm": 0.48828125, + "learning_rate": 5.8980600537463604e-05, + "loss": 2.4342, + "step": 29818 + }, + { + "epoch": 1.27, + "grad_norm": 0.50390625, + "learning_rate": 5.897448300764654e-05, + "loss": 2.6452, + "step": 29819 + }, + { + "epoch": 1.27, + "grad_norm": 0.498046875, + "learning_rate": 5.8968365662429333e-05, + "loss": 2.5688, + "step": 29820 + }, + { + "epoch": 1.27, + "grad_norm": 0.50390625, + "learning_rate": 5.8962248501839504e-05, + "loss": 2.4562, + "step": 29821 + }, + { + "epoch": 1.27, + "grad_norm": 0.47265625, + "learning_rate": 5.8956131525904537e-05, + "loss": 2.3717, + "step": 29822 + }, + { + "epoch": 1.27, + "grad_norm": 0.470703125, + "learning_rate": 5.895001473465196e-05, + "loss": 2.4792, + "step": 29823 + }, + { + "epoch": 1.27, + "grad_norm": 0.53125, + "learning_rate": 5.894389812810934e-05, + "loss": 2.5452, + "step": 29824 + }, + { + "epoch": 1.27, + "grad_norm": 0.4296875, + "learning_rate": 5.8937781706304195e-05, + "loss": 2.5788, + "step": 29825 + }, + { + "epoch": 1.27, + "grad_norm": 0.5234375, + "learning_rate": 5.8931665469264006e-05, + "loss": 2.376, + "step": 29826 + }, + { + "epoch": 1.27, + "grad_norm": 0.41015625, + "learning_rate": 5.892554941701631e-05, + "loss": 2.5266, + "step": 29827 + }, + { + "epoch": 1.27, + "grad_norm": 0.474609375, + "learning_rate": 5.891943354958862e-05, + "loss": 2.4858, + "step": 29828 + }, + { + "epoch": 1.27, + "grad_norm": 0.56640625, + "learning_rate": 5.8913317867008475e-05, + "loss": 2.5662, + "step": 29829 + }, + { + "epoch": 1.27, + "grad_norm": 0.490234375, + "learning_rate": 5.890720236930338e-05, + "loss": 2.4441, + "step": 29830 + }, + { + "epoch": 1.27, + "grad_norm": 0.4453125, + "learning_rate": 5.8901087056500894e-05, + "loss": 2.508, + "step": 29831 + }, + { + "epoch": 1.27, + "grad_norm": 0.47265625, + "learning_rate": 5.889497192862843e-05, + "loss": 2.4751, + "step": 29832 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.8888856985713624e-05, + "loss": 2.4494, + "step": 29833 + }, + { + "epoch": 1.27, + "grad_norm": 0.3984375, + "learning_rate": 5.888274222778392e-05, + "loss": 2.4097, + "step": 29834 + }, + { + "epoch": 1.27, + "grad_norm": 0.4375, + "learning_rate": 5.887662765486686e-05, + "loss": 2.3529, + "step": 29835 + }, + { + "epoch": 1.27, + "grad_norm": 0.474609375, + "learning_rate": 5.8870513266989915e-05, + "loss": 2.3959, + "step": 29836 + }, + { + "epoch": 1.27, + "grad_norm": 0.427734375, + "learning_rate": 5.8864399064180644e-05, + "loss": 2.5677, + "step": 29837 + }, + { + "epoch": 1.27, + "grad_norm": 0.47265625, + "learning_rate": 5.8858285046466544e-05, + "loss": 2.5059, + "step": 29838 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.8852171213875155e-05, + "loss": 2.409, + "step": 29839 + }, + { + "epoch": 1.27, + "grad_norm": 0.431640625, + "learning_rate": 5.8846057566433884e-05, + "loss": 2.427, + "step": 29840 + }, + { + "epoch": 1.27, + "grad_norm": 0.453125, + "learning_rate": 5.8839944104170395e-05, + "loss": 2.4141, + "step": 29841 + }, + { + "epoch": 1.27, + "grad_norm": 0.421875, + "learning_rate": 5.883383082711209e-05, + "loss": 2.2797, + "step": 29842 + }, + { + "epoch": 1.27, + "grad_norm": 0.47265625, + "learning_rate": 5.882771773528649e-05, + "loss": 2.4209, + "step": 29843 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.88216048287211e-05, + "loss": 2.3905, + "step": 29844 + }, + { + "epoch": 1.27, + "grad_norm": 0.5, + "learning_rate": 5.881549210744346e-05, + "loss": 2.4801, + "step": 29845 + }, + { + "epoch": 1.27, + "grad_norm": 0.51171875, + "learning_rate": 5.880937957148105e-05, + "loss": 2.4861, + "step": 29846 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.8803267220861405e-05, + "loss": 2.4873, + "step": 29847 + }, + { + "epoch": 1.27, + "grad_norm": 0.462890625, + "learning_rate": 5.879715505561194e-05, + "loss": 2.374, + "step": 29848 + }, + { + "epoch": 1.27, + "grad_norm": 0.4609375, + "learning_rate": 5.879104307576028e-05, + "loss": 2.4341, + "step": 29849 + }, + { + "epoch": 1.27, + "grad_norm": 0.470703125, + "learning_rate": 5.878493128133385e-05, + "loss": 2.4254, + "step": 29850 + }, + { + "epoch": 1.27, + "grad_norm": 0.4921875, + "learning_rate": 5.877881967236015e-05, + "loss": 2.4284, + "step": 29851 + }, + { + "epoch": 1.27, + "grad_norm": 0.486328125, + "learning_rate": 5.877270824886669e-05, + "loss": 2.3618, + "step": 29852 + }, + { + "epoch": 1.27, + "grad_norm": 0.486328125, + "learning_rate": 5.8766597010880984e-05, + "loss": 2.4571, + "step": 29853 + }, + { + "epoch": 1.27, + "grad_norm": 0.40625, + "learning_rate": 5.876048595843053e-05, + "loss": 2.5176, + "step": 29854 + }, + { + "epoch": 1.27, + "grad_norm": 0.4609375, + "learning_rate": 5.875437509154285e-05, + "loss": 2.5797, + "step": 29855 + }, + { + "epoch": 1.27, + "grad_norm": 0.44921875, + "learning_rate": 5.874826441024535e-05, + "loss": 2.5064, + "step": 29856 + }, + { + "epoch": 1.27, + "grad_norm": 0.48828125, + "learning_rate": 5.874215391456562e-05, + "loss": 2.5393, + "step": 29857 + }, + { + "epoch": 1.27, + "grad_norm": 0.412109375, + "learning_rate": 5.873604360453111e-05, + "loss": 2.4987, + "step": 29858 + }, + { + "epoch": 1.27, + "grad_norm": 0.4609375, + "learning_rate": 5.872993348016933e-05, + "loss": 2.5441, + "step": 29859 + }, + { + "epoch": 1.27, + "grad_norm": 0.51953125, + "learning_rate": 5.872382354150774e-05, + "loss": 2.2444, + "step": 29860 + }, + { + "epoch": 1.27, + "grad_norm": 0.490234375, + "learning_rate": 5.871771378857389e-05, + "loss": 2.541, + "step": 29861 + }, + { + "epoch": 1.27, + "grad_norm": 0.51953125, + "learning_rate": 5.8711604221395235e-05, + "loss": 2.4698, + "step": 29862 + }, + { + "epoch": 1.27, + "grad_norm": 0.455078125, + "learning_rate": 5.870549483999927e-05, + "loss": 2.4588, + "step": 29863 + }, + { + "epoch": 1.27, + "grad_norm": 0.462890625, + "learning_rate": 5.8699385644413466e-05, + "loss": 2.5678, + "step": 29864 + }, + { + "epoch": 1.27, + "grad_norm": 0.484375, + "learning_rate": 5.869327663466537e-05, + "loss": 2.4819, + "step": 29865 + }, + { + "epoch": 1.27, + "grad_norm": 0.5390625, + "learning_rate": 5.868716781078241e-05, + "loss": 2.5151, + "step": 29866 + }, + { + "epoch": 1.27, + "grad_norm": 0.37890625, + "learning_rate": 5.868105917279211e-05, + "loss": 2.4838, + "step": 29867 + }, + { + "epoch": 1.27, + "grad_norm": 0.482421875, + "learning_rate": 5.867495072072191e-05, + "loss": 2.6365, + "step": 29868 + }, + { + "epoch": 1.27, + "grad_norm": 0.49609375, + "learning_rate": 5.866884245459936e-05, + "loss": 2.4663, + "step": 29869 + }, + { + "epoch": 1.27, + "grad_norm": 0.51953125, + "learning_rate": 5.866273437445191e-05, + "loss": 2.4646, + "step": 29870 + }, + { + "epoch": 1.27, + "grad_norm": 0.447265625, + "learning_rate": 5.865662648030707e-05, + "loss": 2.5102, + "step": 29871 + }, + { + "epoch": 1.27, + "grad_norm": 0.44140625, + "learning_rate": 5.865051877219223e-05, + "loss": 2.3267, + "step": 29872 + }, + { + "epoch": 1.27, + "grad_norm": 0.53515625, + "learning_rate": 5.8644411250135023e-05, + "loss": 2.4393, + "step": 29873 + }, + { + "epoch": 1.27, + "grad_norm": 0.49609375, + "learning_rate": 5.863830391416283e-05, + "loss": 2.5348, + "step": 29874 + }, + { + "epoch": 1.27, + "grad_norm": 0.45703125, + "learning_rate": 5.863219676430314e-05, + "loss": 2.3986, + "step": 29875 + }, + { + "epoch": 1.27, + "grad_norm": 0.416015625, + "learning_rate": 5.862608980058344e-05, + "loss": 2.565, + "step": 29876 + }, + { + "epoch": 1.27, + "grad_norm": 0.470703125, + "learning_rate": 5.8619983023031224e-05, + "loss": 2.4099, + "step": 29877 + }, + { + "epoch": 1.27, + "grad_norm": 0.482421875, + "learning_rate": 5.8613876431673956e-05, + "loss": 2.4866, + "step": 29878 + }, + { + "epoch": 1.27, + "grad_norm": 0.515625, + "learning_rate": 5.860777002653916e-05, + "loss": 2.4961, + "step": 29879 + }, + { + "epoch": 1.27, + "grad_norm": 0.45703125, + "learning_rate": 5.860166380765419e-05, + "loss": 2.5007, + "step": 29880 + }, + { + "epoch": 1.27, + "grad_norm": 0.435546875, + "learning_rate": 5.859555777504668e-05, + "loss": 2.5278, + "step": 29881 + }, + { + "epoch": 1.27, + "grad_norm": 0.5234375, + "learning_rate": 5.8589451928744e-05, + "loss": 2.5777, + "step": 29882 + }, + { + "epoch": 1.27, + "grad_norm": 0.458984375, + "learning_rate": 5.858334626877366e-05, + "loss": 2.3328, + "step": 29883 + }, + { + "epoch": 1.27, + "grad_norm": 0.466796875, + "learning_rate": 5.857724079516309e-05, + "loss": 2.4232, + "step": 29884 + }, + { + "epoch": 1.27, + "grad_norm": 0.42578125, + "learning_rate": 5.8571135507939836e-05, + "loss": 2.488, + "step": 29885 + }, + { + "epoch": 1.27, + "grad_norm": 0.4765625, + "learning_rate": 5.856503040713133e-05, + "loss": 2.478, + "step": 29886 + }, + { + "epoch": 1.27, + "grad_norm": 0.51171875, + "learning_rate": 5.855892549276505e-05, + "loss": 2.548, + "step": 29887 + }, + { + "epoch": 1.27, + "grad_norm": 0.443359375, + "learning_rate": 5.855282076486841e-05, + "loss": 2.3775, + "step": 29888 + }, + { + "epoch": 1.27, + "grad_norm": 0.4921875, + "learning_rate": 5.8546716223468986e-05, + "loss": 2.4513, + "step": 29889 + }, + { + "epoch": 1.27, + "grad_norm": 0.470703125, + "learning_rate": 5.854061186859419e-05, + "loss": 2.2886, + "step": 29890 + }, + { + "epoch": 1.27, + "grad_norm": 0.46484375, + "learning_rate": 5.853450770027146e-05, + "loss": 2.4587, + "step": 29891 + }, + { + "epoch": 1.27, + "grad_norm": 0.462890625, + "learning_rate": 5.852840371852828e-05, + "loss": 2.3731, + "step": 29892 + }, + { + "epoch": 1.27, + "grad_norm": 0.466796875, + "learning_rate": 5.852229992339215e-05, + "loss": 2.4789, + "step": 29893 + }, + { + "epoch": 1.27, + "grad_norm": 0.5078125, + "learning_rate": 5.851619631489051e-05, + "loss": 2.3657, + "step": 29894 + }, + { + "epoch": 1.27, + "grad_norm": 0.482421875, + "learning_rate": 5.851009289305085e-05, + "loss": 2.3785, + "step": 29895 + }, + { + "epoch": 1.27, + "grad_norm": 0.4296875, + "learning_rate": 5.850398965790055e-05, + "loss": 2.4541, + "step": 29896 + }, + { + "epoch": 1.27, + "grad_norm": 0.4921875, + "learning_rate": 5.849788660946718e-05, + "loss": 2.3848, + "step": 29897 + }, + { + "epoch": 1.27, + "grad_norm": 0.462890625, + "learning_rate": 5.849178374777815e-05, + "loss": 2.5635, + "step": 29898 + }, + { + "epoch": 1.27, + "grad_norm": 0.431640625, + "learning_rate": 5.848568107286092e-05, + "loss": 2.4094, + "step": 29899 + }, + { + "epoch": 1.27, + "grad_norm": 0.51171875, + "learning_rate": 5.847957858474292e-05, + "loss": 2.5132, + "step": 29900 + }, + { + "epoch": 1.27, + "grad_norm": 0.44140625, + "learning_rate": 5.847347628345168e-05, + "loss": 2.4124, + "step": 29901 + }, + { + "epoch": 1.27, + "grad_norm": 0.46484375, + "learning_rate": 5.84673741690146e-05, + "loss": 2.3396, + "step": 29902 + }, + { + "epoch": 1.27, + "grad_norm": 0.5078125, + "learning_rate": 5.846127224145919e-05, + "loss": 2.4773, + "step": 29903 + }, + { + "epoch": 1.27, + "grad_norm": 0.5625, + "learning_rate": 5.845517050081282e-05, + "loss": 2.5698, + "step": 29904 + }, + { + "epoch": 1.27, + "grad_norm": 0.50390625, + "learning_rate": 5.844906894710305e-05, + "loss": 2.4594, + "step": 29905 + }, + { + "epoch": 1.27, + "grad_norm": 0.494140625, + "learning_rate": 5.844296758035727e-05, + "loss": 2.3786, + "step": 29906 + }, + { + "epoch": 1.27, + "grad_norm": 0.466796875, + "learning_rate": 5.843686640060294e-05, + "loss": 2.4459, + "step": 29907 + }, + { + "epoch": 1.27, + "grad_norm": 0.50390625, + "learning_rate": 5.8430765407867513e-05, + "loss": 2.3466, + "step": 29908 + }, + { + "epoch": 1.27, + "grad_norm": 0.48046875, + "learning_rate": 5.842466460217846e-05, + "loss": 2.4355, + "step": 29909 + }, + { + "epoch": 1.27, + "grad_norm": 0.4453125, + "learning_rate": 5.841856398356322e-05, + "loss": 2.5771, + "step": 29910 + }, + { + "epoch": 1.27, + "grad_norm": 0.4296875, + "learning_rate": 5.841246355204928e-05, + "loss": 2.4563, + "step": 29911 + }, + { + "epoch": 1.27, + "grad_norm": 0.4921875, + "learning_rate": 5.840636330766399e-05, + "loss": 2.382, + "step": 29912 + }, + { + "epoch": 1.27, + "grad_norm": 0.54296875, + "learning_rate": 5.840026325043493e-05, + "loss": 2.6464, + "step": 29913 + }, + { + "epoch": 1.27, + "grad_norm": 0.451171875, + "learning_rate": 5.8394163380389455e-05, + "loss": 2.4481, + "step": 29914 + }, + { + "epoch": 1.27, + "grad_norm": 0.546875, + "learning_rate": 5.838806369755503e-05, + "loss": 2.5697, + "step": 29915 + }, + { + "epoch": 1.27, + "grad_norm": 0.5234375, + "learning_rate": 5.838196420195908e-05, + "loss": 2.5179, + "step": 29916 + }, + { + "epoch": 1.27, + "grad_norm": 0.44140625, + "learning_rate": 5.837586489362914e-05, + "loss": 2.4561, + "step": 29917 + }, + { + "epoch": 1.27, + "grad_norm": 0.4453125, + "learning_rate": 5.836976577259261e-05, + "loss": 2.4582, + "step": 29918 + }, + { + "epoch": 1.27, + "grad_norm": 0.404296875, + "learning_rate": 5.8363666838876866e-05, + "loss": 2.4835, + "step": 29919 + }, + { + "epoch": 1.27, + "grad_norm": 0.443359375, + "learning_rate": 5.8357568092509406e-05, + "loss": 2.6021, + "step": 29920 + }, + { + "epoch": 1.28, + "grad_norm": 0.4765625, + "learning_rate": 5.835146953351771e-05, + "loss": 2.3609, + "step": 29921 + }, + { + "epoch": 1.28, + "grad_norm": 0.48828125, + "learning_rate": 5.834537116192914e-05, + "loss": 2.497, + "step": 29922 + }, + { + "epoch": 1.28, + "grad_norm": 0.48828125, + "learning_rate": 5.833927297777122e-05, + "loss": 2.5379, + "step": 29923 + }, + { + "epoch": 1.28, + "grad_norm": 0.50390625, + "learning_rate": 5.8333174981071305e-05, + "loss": 2.4396, + "step": 29924 + }, + { + "epoch": 1.28, + "grad_norm": 0.515625, + "learning_rate": 5.8327077171856917e-05, + "loss": 2.4781, + "step": 29925 + }, + { + "epoch": 1.28, + "grad_norm": 0.474609375, + "learning_rate": 5.832097955015541e-05, + "loss": 2.4187, + "step": 29926 + }, + { + "epoch": 1.28, + "grad_norm": 0.43359375, + "learning_rate": 5.8314882115994295e-05, + "loss": 2.412, + "step": 29927 + }, + { + "epoch": 1.28, + "grad_norm": 0.4296875, + "learning_rate": 5.830878486940094e-05, + "loss": 2.3646, + "step": 29928 + }, + { + "epoch": 1.28, + "grad_norm": 0.373046875, + "learning_rate": 5.8302687810402836e-05, + "loss": 2.3233, + "step": 29929 + }, + { + "epoch": 1.28, + "grad_norm": 0.423828125, + "learning_rate": 5.829659093902742e-05, + "loss": 2.3722, + "step": 29930 + }, + { + "epoch": 1.28, + "grad_norm": 0.474609375, + "learning_rate": 5.829049425530211e-05, + "loss": 2.4684, + "step": 29931 + }, + { + "epoch": 1.28, + "grad_norm": 0.55078125, + "learning_rate": 5.828439775925425e-05, + "loss": 2.6073, + "step": 29932 + }, + { + "epoch": 1.28, + "grad_norm": 0.421875, + "learning_rate": 5.827830145091144e-05, + "loss": 2.3966, + "step": 29933 + }, + { + "epoch": 1.28, + "grad_norm": 0.458984375, + "learning_rate": 5.827220533030102e-05, + "loss": 2.5083, + "step": 29934 + }, + { + "epoch": 1.28, + "grad_norm": 0.48046875, + "learning_rate": 5.826610939745039e-05, + "loss": 2.4804, + "step": 29935 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.8260013652387e-05, + "loss": 2.423, + "step": 29936 + }, + { + "epoch": 1.28, + "grad_norm": 0.51171875, + "learning_rate": 5.825391809513836e-05, + "loss": 2.5723, + "step": 29937 + }, + { + "epoch": 1.28, + "grad_norm": 0.3984375, + "learning_rate": 5.824782272573177e-05, + "loss": 2.3794, + "step": 29938 + }, + { + "epoch": 1.28, + "grad_norm": 0.5078125, + "learning_rate": 5.824172754419477e-05, + "loss": 2.4023, + "step": 29939 + }, + { + "epoch": 1.28, + "grad_norm": 0.421875, + "learning_rate": 5.8235632550554695e-05, + "loss": 2.385, + "step": 29940 + }, + { + "epoch": 1.28, + "grad_norm": 0.435546875, + "learning_rate": 5.8229537744839056e-05, + "loss": 2.4567, + "step": 29941 + }, + { + "epoch": 1.28, + "grad_norm": 0.392578125, + "learning_rate": 5.822344312707518e-05, + "loss": 2.2479, + "step": 29942 + }, + { + "epoch": 1.28, + "grad_norm": 0.498046875, + "learning_rate": 5.821734869729059e-05, + "loss": 2.3118, + "step": 29943 + }, + { + "epoch": 1.28, + "grad_norm": 0.486328125, + "learning_rate": 5.821125445551261e-05, + "loss": 2.5104, + "step": 29944 + }, + { + "epoch": 1.28, + "grad_norm": 0.51953125, + "learning_rate": 5.8205160401768734e-05, + "loss": 2.4636, + "step": 29945 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.8199066536086377e-05, + "loss": 2.4849, + "step": 29946 + }, + { + "epoch": 1.28, + "grad_norm": 0.5, + "learning_rate": 5.8192972858492964e-05, + "loss": 2.3868, + "step": 29947 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.8186879369015846e-05, + "loss": 2.4481, + "step": 29948 + }, + { + "epoch": 1.28, + "grad_norm": 0.47265625, + "learning_rate": 5.818078606768249e-05, + "loss": 2.5421, + "step": 29949 + }, + { + "epoch": 1.28, + "grad_norm": 0.4453125, + "learning_rate": 5.817469295452036e-05, + "loss": 2.2809, + "step": 29950 + }, + { + "epoch": 1.28, + "grad_norm": 0.47265625, + "learning_rate": 5.816860002955677e-05, + "loss": 2.4706, + "step": 29951 + }, + { + "epoch": 1.28, + "grad_norm": 0.439453125, + "learning_rate": 5.8162507292819246e-05, + "loss": 2.4452, + "step": 29952 + }, + { + "epoch": 1.28, + "grad_norm": 0.53515625, + "learning_rate": 5.815641474433511e-05, + "loss": 2.4768, + "step": 29953 + }, + { + "epoch": 1.28, + "grad_norm": 0.443359375, + "learning_rate": 5.8150322384131805e-05, + "loss": 2.4716, + "step": 29954 + }, + { + "epoch": 1.28, + "grad_norm": 0.4140625, + "learning_rate": 5.8144230212236804e-05, + "loss": 2.4559, + "step": 29955 + }, + { + "epoch": 1.28, + "grad_norm": 0.5, + "learning_rate": 5.8138138228677465e-05, + "loss": 2.6083, + "step": 29956 + }, + { + "epoch": 1.28, + "grad_norm": 0.41796875, + "learning_rate": 5.813204643348118e-05, + "loss": 2.3094, + "step": 29957 + }, + { + "epoch": 1.28, + "grad_norm": 0.46484375, + "learning_rate": 5.8125954826675364e-05, + "loss": 2.4283, + "step": 29958 + }, + { + "epoch": 1.28, + "grad_norm": 0.470703125, + "learning_rate": 5.81198634082875e-05, + "loss": 2.4975, + "step": 29959 + }, + { + "epoch": 1.28, + "grad_norm": 0.4765625, + "learning_rate": 5.8113772178344904e-05, + "loss": 2.4068, + "step": 29960 + }, + { + "epoch": 1.28, + "grad_norm": 0.486328125, + "learning_rate": 5.810768113687502e-05, + "loss": 2.4922, + "step": 29961 + }, + { + "epoch": 1.28, + "grad_norm": 0.46484375, + "learning_rate": 5.810159028390532e-05, + "loss": 2.3176, + "step": 29962 + }, + { + "epoch": 1.28, + "grad_norm": 0.478515625, + "learning_rate": 5.809549961946312e-05, + "loss": 2.476, + "step": 29963 + }, + { + "epoch": 1.28, + "grad_norm": 0.44921875, + "learning_rate": 5.8089409143575836e-05, + "loss": 2.4129, + "step": 29964 + }, + { + "epoch": 1.28, + "grad_norm": 0.470703125, + "learning_rate": 5.8083318856270895e-05, + "loss": 2.3624, + "step": 29965 + }, + { + "epoch": 1.28, + "grad_norm": 0.462890625, + "learning_rate": 5.807722875757574e-05, + "loss": 2.3954, + "step": 29966 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.807113884751769e-05, + "loss": 2.504, + "step": 29967 + }, + { + "epoch": 1.28, + "grad_norm": 0.46484375, + "learning_rate": 5.806504912612424e-05, + "loss": 2.4635, + "step": 29968 + }, + { + "epoch": 1.28, + "grad_norm": 0.5078125, + "learning_rate": 5.805895959342269e-05, + "loss": 2.4161, + "step": 29969 + }, + { + "epoch": 1.28, + "grad_norm": 0.4765625, + "learning_rate": 5.80528702494405e-05, + "loss": 2.4349, + "step": 29970 + }, + { + "epoch": 1.28, + "grad_norm": 0.474609375, + "learning_rate": 5.8046781094205117e-05, + "loss": 2.3911, + "step": 29971 + }, + { + "epoch": 1.28, + "grad_norm": 0.435546875, + "learning_rate": 5.804069212774387e-05, + "loss": 2.423, + "step": 29972 + }, + { + "epoch": 1.28, + "grad_norm": 0.50390625, + "learning_rate": 5.803460335008414e-05, + "loss": 2.3886, + "step": 29973 + }, + { + "epoch": 1.28, + "grad_norm": 0.427734375, + "learning_rate": 5.802851476125335e-05, + "loss": 2.4495, + "step": 29974 + }, + { + "epoch": 1.28, + "grad_norm": 0.484375, + "learning_rate": 5.802242636127896e-05, + "loss": 2.3982, + "step": 29975 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.8016338150188265e-05, + "loss": 2.4828, + "step": 29976 + }, + { + "epoch": 1.28, + "grad_norm": 0.53515625, + "learning_rate": 5.80102501280087e-05, + "loss": 2.4694, + "step": 29977 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.800416229476771e-05, + "loss": 2.5356, + "step": 29978 + }, + { + "epoch": 1.28, + "grad_norm": 0.458984375, + "learning_rate": 5.799807465049264e-05, + "loss": 2.5089, + "step": 29979 + }, + { + "epoch": 1.28, + "grad_norm": 0.515625, + "learning_rate": 5.799198719521084e-05, + "loss": 2.5065, + "step": 29980 + }, + { + "epoch": 1.28, + "grad_norm": 0.462890625, + "learning_rate": 5.798589992894975e-05, + "loss": 2.3844, + "step": 29981 + }, + { + "epoch": 1.28, + "grad_norm": 0.4609375, + "learning_rate": 5.7979812851736795e-05, + "loss": 2.4193, + "step": 29982 + }, + { + "epoch": 1.28, + "grad_norm": 0.51171875, + "learning_rate": 5.797372596359929e-05, + "loss": 2.6347, + "step": 29983 + }, + { + "epoch": 1.28, + "grad_norm": 0.431640625, + "learning_rate": 5.796763926456469e-05, + "loss": 2.3553, + "step": 29984 + }, + { + "epoch": 1.28, + "grad_norm": 0.4609375, + "learning_rate": 5.796155275466032e-05, + "loss": 2.4765, + "step": 29985 + }, + { + "epoch": 1.28, + "grad_norm": 0.453125, + "learning_rate": 5.7955466433913584e-05, + "loss": 2.3387, + "step": 29986 + }, + { + "epoch": 1.28, + "grad_norm": 0.53125, + "learning_rate": 5.794938030235195e-05, + "loss": 2.4468, + "step": 29987 + }, + { + "epoch": 1.28, + "grad_norm": 0.486328125, + "learning_rate": 5.794329436000272e-05, + "loss": 2.4631, + "step": 29988 + }, + { + "epoch": 1.28, + "grad_norm": 0.44140625, + "learning_rate": 5.793720860689326e-05, + "loss": 2.3378, + "step": 29989 + }, + { + "epoch": 1.28, + "grad_norm": 0.5234375, + "learning_rate": 5.7931123043050986e-05, + "loss": 2.344, + "step": 29990 + }, + { + "epoch": 1.28, + "grad_norm": 0.45703125, + "learning_rate": 5.7925037668503325e-05, + "loss": 2.3567, + "step": 29991 + }, + { + "epoch": 1.28, + "grad_norm": 0.49609375, + "learning_rate": 5.791895248327757e-05, + "loss": 2.4532, + "step": 29992 + }, + { + "epoch": 1.28, + "grad_norm": 0.423828125, + "learning_rate": 5.791286748740117e-05, + "loss": 2.4576, + "step": 29993 + }, + { + "epoch": 1.28, + "grad_norm": 0.4453125, + "learning_rate": 5.7906782680901506e-05, + "loss": 2.3573, + "step": 29994 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.7900698063805935e-05, + "loss": 2.5052, + "step": 29995 + }, + { + "epoch": 1.28, + "grad_norm": 0.486328125, + "learning_rate": 5.7894613636141817e-05, + "loss": 2.4813, + "step": 29996 + }, + { + "epoch": 1.28, + "grad_norm": 0.5234375, + "learning_rate": 5.788852939793654e-05, + "loss": 2.5251, + "step": 29997 + }, + { + "epoch": 1.28, + "grad_norm": 0.49609375, + "learning_rate": 5.788244534921752e-05, + "loss": 2.449, + "step": 29998 + }, + { + "epoch": 1.28, + "grad_norm": 0.421875, + "learning_rate": 5.787636149001208e-05, + "loss": 2.3108, + "step": 29999 + }, + { + "epoch": 1.28, + "grad_norm": 0.54296875, + "learning_rate": 5.787027782034765e-05, + "loss": 2.5337, + "step": 30000 + }, + { + "epoch": 1.28, + "grad_norm": 0.46875, + "learning_rate": 5.786419434025153e-05, + "loss": 2.3826, + "step": 30001 + }, + { + "epoch": 1.28, + "grad_norm": 0.48828125, + "learning_rate": 5.785811104975114e-05, + "loss": 2.5203, + "step": 30002 + }, + { + "epoch": 1.28, + "grad_norm": 0.486328125, + "learning_rate": 5.78520279488739e-05, + "loss": 2.4676, + "step": 30003 + }, + { + "epoch": 1.28, + "grad_norm": 0.396484375, + "learning_rate": 5.784594503764712e-05, + "loss": 2.4586, + "step": 30004 + }, + { + "epoch": 1.28, + "grad_norm": 0.51171875, + "learning_rate": 5.7839862316098145e-05, + "loss": 2.4634, + "step": 30005 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.783377978425437e-05, + "loss": 2.3203, + "step": 30006 + }, + { + "epoch": 1.28, + "grad_norm": 0.427734375, + "learning_rate": 5.782769744214324e-05, + "loss": 2.3849, + "step": 30007 + }, + { + "epoch": 1.28, + "grad_norm": 0.43359375, + "learning_rate": 5.7821615289792e-05, + "loss": 2.4178, + "step": 30008 + }, + { + "epoch": 1.28, + "grad_norm": 0.435546875, + "learning_rate": 5.781553332722809e-05, + "loss": 2.3224, + "step": 30009 + }, + { + "epoch": 1.28, + "grad_norm": 0.447265625, + "learning_rate": 5.780945155447889e-05, + "loss": 2.5175, + "step": 30010 + }, + { + "epoch": 1.28, + "grad_norm": 0.48046875, + "learning_rate": 5.780336997157174e-05, + "loss": 2.4433, + "step": 30011 + }, + { + "epoch": 1.28, + "grad_norm": 0.49609375, + "learning_rate": 5.779728857853397e-05, + "loss": 2.7198, + "step": 30012 + }, + { + "epoch": 1.28, + "grad_norm": 0.48046875, + "learning_rate": 5.7791207375392984e-05, + "loss": 2.6114, + "step": 30013 + }, + { + "epoch": 1.28, + "grad_norm": 0.46875, + "learning_rate": 5.7785126362176165e-05, + "loss": 2.4458, + "step": 30014 + }, + { + "epoch": 1.28, + "grad_norm": 0.42578125, + "learning_rate": 5.777904553891083e-05, + "loss": 2.4957, + "step": 30015 + }, + { + "epoch": 1.28, + "grad_norm": 0.478515625, + "learning_rate": 5.7772964905624383e-05, + "loss": 2.4068, + "step": 30016 + }, + { + "epoch": 1.28, + "grad_norm": 0.4609375, + "learning_rate": 5.776688446234412e-05, + "loss": 2.479, + "step": 30017 + }, + { + "epoch": 1.28, + "grad_norm": 0.4921875, + "learning_rate": 5.7760804209097444e-05, + "loss": 2.4121, + "step": 30018 + }, + { + "epoch": 1.28, + "grad_norm": 0.494140625, + "learning_rate": 5.775472414591176e-05, + "loss": 2.3968, + "step": 30019 + }, + { + "epoch": 1.28, + "grad_norm": 0.44921875, + "learning_rate": 5.774864427281437e-05, + "loss": 2.4181, + "step": 30020 + }, + { + "epoch": 1.28, + "grad_norm": 0.421875, + "learning_rate": 5.7742564589832606e-05, + "loss": 2.4514, + "step": 30021 + }, + { + "epoch": 1.28, + "grad_norm": 0.48828125, + "learning_rate": 5.7736485096993855e-05, + "loss": 2.4618, + "step": 30022 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.7730405794325515e-05, + "loss": 2.4649, + "step": 30023 + }, + { + "epoch": 1.28, + "grad_norm": 0.5, + "learning_rate": 5.7724326681854865e-05, + "loss": 2.3065, + "step": 30024 + }, + { + "epoch": 1.28, + "grad_norm": 0.486328125, + "learning_rate": 5.77182477596093e-05, + "loss": 2.3414, + "step": 30025 + }, + { + "epoch": 1.28, + "grad_norm": 0.470703125, + "learning_rate": 5.77121690276162e-05, + "loss": 2.4262, + "step": 30026 + }, + { + "epoch": 1.28, + "grad_norm": 0.443359375, + "learning_rate": 5.770609048590289e-05, + "loss": 2.2935, + "step": 30027 + }, + { + "epoch": 1.28, + "grad_norm": 0.53125, + "learning_rate": 5.7700012134496675e-05, + "loss": 2.4342, + "step": 30028 + }, + { + "epoch": 1.28, + "grad_norm": 0.43359375, + "learning_rate": 5.769393397342493e-05, + "loss": 2.3501, + "step": 30029 + }, + { + "epoch": 1.28, + "grad_norm": 0.55859375, + "learning_rate": 5.768785600271508e-05, + "loss": 2.4906, + "step": 30030 + }, + { + "epoch": 1.28, + "grad_norm": 0.466796875, + "learning_rate": 5.7681778222394376e-05, + "loss": 2.453, + "step": 30031 + }, + { + "epoch": 1.28, + "grad_norm": 0.404296875, + "learning_rate": 5.7675700632490236e-05, + "loss": 2.5434, + "step": 30032 + }, + { + "epoch": 1.28, + "grad_norm": 0.4921875, + "learning_rate": 5.7669623233029935e-05, + "loss": 2.5039, + "step": 30033 + }, + { + "epoch": 1.28, + "grad_norm": 0.392578125, + "learning_rate": 5.766354602404086e-05, + "loss": 2.4463, + "step": 30034 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.76574690055504e-05, + "loss": 2.4479, + "step": 30035 + }, + { + "epoch": 1.28, + "grad_norm": 0.484375, + "learning_rate": 5.765139217758585e-05, + "loss": 2.4961, + "step": 30036 + }, + { + "epoch": 1.28, + "grad_norm": 0.50390625, + "learning_rate": 5.7645315540174516e-05, + "loss": 2.5793, + "step": 30037 + }, + { + "epoch": 1.28, + "grad_norm": 0.46875, + "learning_rate": 5.763923909334379e-05, + "loss": 2.4954, + "step": 30038 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.763316283712105e-05, + "loss": 2.4137, + "step": 30039 + }, + { + "epoch": 1.28, + "grad_norm": 0.47265625, + "learning_rate": 5.7627086771533544e-05, + "loss": 2.3298, + "step": 30040 + }, + { + "epoch": 1.28, + "grad_norm": 0.400390625, + "learning_rate": 5.762101089660867e-05, + "loss": 2.4117, + "step": 30041 + }, + { + "epoch": 1.28, + "grad_norm": 0.494140625, + "learning_rate": 5.76149352123738e-05, + "loss": 2.401, + "step": 30042 + }, + { + "epoch": 1.28, + "grad_norm": 0.40234375, + "learning_rate": 5.760885971885622e-05, + "loss": 2.3863, + "step": 30043 + }, + { + "epoch": 1.28, + "grad_norm": 0.482421875, + "learning_rate": 5.7602784416083244e-05, + "loss": 2.5048, + "step": 30044 + }, + { + "epoch": 1.28, + "grad_norm": 0.49609375, + "learning_rate": 5.7596709304082255e-05, + "loss": 2.5802, + "step": 30045 + }, + { + "epoch": 1.28, + "grad_norm": 0.484375, + "learning_rate": 5.7590634382880614e-05, + "loss": 2.3928, + "step": 30046 + }, + { + "epoch": 1.28, + "grad_norm": 0.5078125, + "learning_rate": 5.7584559652505575e-05, + "loss": 2.5359, + "step": 30047 + }, + { + "epoch": 1.28, + "grad_norm": 0.4375, + "learning_rate": 5.7578485112984556e-05, + "loss": 2.4436, + "step": 30048 + }, + { + "epoch": 1.28, + "grad_norm": 0.46484375, + "learning_rate": 5.757241076434481e-05, + "loss": 2.4961, + "step": 30049 + }, + { + "epoch": 1.28, + "grad_norm": 0.458984375, + "learning_rate": 5.75663366066137e-05, + "loss": 2.4065, + "step": 30050 + }, + { + "epoch": 1.28, + "grad_norm": 0.4453125, + "learning_rate": 5.7560262639818616e-05, + "loss": 2.4411, + "step": 30051 + }, + { + "epoch": 1.28, + "grad_norm": 0.47265625, + "learning_rate": 5.7554188863986846e-05, + "loss": 2.6542, + "step": 30052 + }, + { + "epoch": 1.28, + "grad_norm": 0.494140625, + "learning_rate": 5.754811527914568e-05, + "loss": 2.4629, + "step": 30053 + }, + { + "epoch": 1.28, + "grad_norm": 0.47265625, + "learning_rate": 5.754204188532246e-05, + "loss": 2.479, + "step": 30054 + }, + { + "epoch": 1.28, + "grad_norm": 0.4765625, + "learning_rate": 5.7535968682544595e-05, + "loss": 2.5538, + "step": 30055 + }, + { + "epoch": 1.28, + "grad_norm": 0.458984375, + "learning_rate": 5.7529895670839306e-05, + "loss": 2.4219, + "step": 30056 + }, + { + "epoch": 1.28, + "grad_norm": 0.474609375, + "learning_rate": 5.7523822850233966e-05, + "loss": 2.487, + "step": 30057 + }, + { + "epoch": 1.28, + "grad_norm": 0.4140625, + "learning_rate": 5.7517750220755926e-05, + "loss": 2.4261, + "step": 30058 + }, + { + "epoch": 1.28, + "grad_norm": 0.4453125, + "learning_rate": 5.751167778243249e-05, + "loss": 2.4875, + "step": 30059 + }, + { + "epoch": 1.28, + "grad_norm": 0.458984375, + "learning_rate": 5.750560553529095e-05, + "loss": 2.4731, + "step": 30060 + }, + { + "epoch": 1.28, + "grad_norm": 0.451171875, + "learning_rate": 5.749953347935865e-05, + "loss": 2.3863, + "step": 30061 + }, + { + "epoch": 1.28, + "grad_norm": 0.490234375, + "learning_rate": 5.749346161466296e-05, + "loss": 2.5004, + "step": 30062 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.7487389941231106e-05, + "loss": 2.4197, + "step": 30063 + }, + { + "epoch": 1.28, + "grad_norm": 0.404296875, + "learning_rate": 5.7481318459090506e-05, + "loss": 2.4792, + "step": 30064 + }, + { + "epoch": 1.28, + "grad_norm": 0.443359375, + "learning_rate": 5.74752471682684e-05, + "loss": 2.459, + "step": 30065 + }, + { + "epoch": 1.28, + "grad_norm": 0.4453125, + "learning_rate": 5.746917606879213e-05, + "loss": 2.3287, + "step": 30066 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.746310516068907e-05, + "loss": 2.4572, + "step": 30067 + }, + { + "epoch": 1.28, + "grad_norm": 0.45703125, + "learning_rate": 5.745703444398649e-05, + "loss": 2.4771, + "step": 30068 + }, + { + "epoch": 1.28, + "grad_norm": 0.482421875, + "learning_rate": 5.7450963918711675e-05, + "loss": 2.4944, + "step": 30069 + }, + { + "epoch": 1.28, + "grad_norm": 0.466796875, + "learning_rate": 5.7444893584891954e-05, + "loss": 2.5194, + "step": 30070 + }, + { + "epoch": 1.28, + "grad_norm": 0.498046875, + "learning_rate": 5.7438823442554724e-05, + "loss": 2.4348, + "step": 30071 + }, + { + "epoch": 1.28, + "grad_norm": 0.494140625, + "learning_rate": 5.7432753491727185e-05, + "loss": 2.3871, + "step": 30072 + }, + { + "epoch": 1.28, + "grad_norm": 0.47265625, + "learning_rate": 5.742668373243674e-05, + "loss": 2.4053, + "step": 30073 + }, + { + "epoch": 1.28, + "grad_norm": 0.439453125, + "learning_rate": 5.742061416471061e-05, + "loss": 2.4049, + "step": 30074 + }, + { + "epoch": 1.28, + "grad_norm": 0.498046875, + "learning_rate": 5.74145447885762e-05, + "loss": 2.5495, + "step": 30075 + }, + { + "epoch": 1.28, + "grad_norm": 0.5234375, + "learning_rate": 5.740847560406073e-05, + "loss": 2.4522, + "step": 30076 + }, + { + "epoch": 1.28, + "grad_norm": 0.484375, + "learning_rate": 5.7402406611191606e-05, + "loss": 2.5683, + "step": 30077 + }, + { + "epoch": 1.28, + "grad_norm": 0.466796875, + "learning_rate": 5.739633780999604e-05, + "loss": 2.4631, + "step": 30078 + }, + { + "epoch": 1.28, + "grad_norm": 0.5078125, + "learning_rate": 5.739026920050138e-05, + "loss": 2.5338, + "step": 30079 + }, + { + "epoch": 1.28, + "grad_norm": 0.466796875, + "learning_rate": 5.7384200782734986e-05, + "loss": 2.4482, + "step": 30080 + }, + { + "epoch": 1.28, + "grad_norm": 0.40625, + "learning_rate": 5.737813255672411e-05, + "loss": 2.5069, + "step": 30081 + }, + { + "epoch": 1.28, + "grad_norm": 0.53515625, + "learning_rate": 5.7372064522495974e-05, + "loss": 2.7402, + "step": 30082 + }, + { + "epoch": 1.28, + "grad_norm": 0.443359375, + "learning_rate": 5.736599668007805e-05, + "loss": 2.5042, + "step": 30083 + }, + { + "epoch": 1.28, + "grad_norm": 0.40625, + "learning_rate": 5.735992902949756e-05, + "loss": 2.4929, + "step": 30084 + }, + { + "epoch": 1.28, + "grad_norm": 0.51171875, + "learning_rate": 5.735386157078178e-05, + "loss": 2.6335, + "step": 30085 + }, + { + "epoch": 1.28, + "grad_norm": 0.435546875, + "learning_rate": 5.734779430395803e-05, + "loss": 2.3886, + "step": 30086 + }, + { + "epoch": 1.28, + "grad_norm": 0.443359375, + "learning_rate": 5.734172722905364e-05, + "loss": 2.5816, + "step": 30087 + }, + { + "epoch": 1.28, + "grad_norm": 0.482421875, + "learning_rate": 5.733566034609588e-05, + "loss": 2.4763, + "step": 30088 + }, + { + "epoch": 1.28, + "grad_norm": 0.490234375, + "learning_rate": 5.732959365511207e-05, + "loss": 2.4993, + "step": 30089 + }, + { + "epoch": 1.28, + "grad_norm": 0.421875, + "learning_rate": 5.7323527156129465e-05, + "loss": 2.6208, + "step": 30090 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.731746084917542e-05, + "loss": 2.5435, + "step": 30091 + }, + { + "epoch": 1.28, + "grad_norm": 0.435546875, + "learning_rate": 5.731139473427718e-05, + "loss": 2.5527, + "step": 30092 + }, + { + "epoch": 1.28, + "grad_norm": 0.46484375, + "learning_rate": 5.730532881146209e-05, + "loss": 2.5921, + "step": 30093 + }, + { + "epoch": 1.28, + "grad_norm": 0.494140625, + "learning_rate": 5.7299263080757396e-05, + "loss": 2.494, + "step": 30094 + }, + { + "epoch": 1.28, + "grad_norm": 0.546875, + "learning_rate": 5.729319754219039e-05, + "loss": 2.5739, + "step": 30095 + }, + { + "epoch": 1.28, + "grad_norm": 0.453125, + "learning_rate": 5.728713219578844e-05, + "loss": 2.461, + "step": 30096 + }, + { + "epoch": 1.28, + "grad_norm": 0.49609375, + "learning_rate": 5.7281067041578787e-05, + "loss": 2.4481, + "step": 30097 + }, + { + "epoch": 1.28, + "grad_norm": 0.45703125, + "learning_rate": 5.727500207958863e-05, + "loss": 2.4447, + "step": 30098 + }, + { + "epoch": 1.28, + "grad_norm": 0.4296875, + "learning_rate": 5.7268937309845436e-05, + "loss": 2.4289, + "step": 30099 + }, + { + "epoch": 1.28, + "grad_norm": 0.458984375, + "learning_rate": 5.7262872732376406e-05, + "loss": 2.3901, + "step": 30100 + }, + { + "epoch": 1.28, + "grad_norm": 0.4765625, + "learning_rate": 5.725680834720879e-05, + "loss": 2.4217, + "step": 30101 + }, + { + "epoch": 1.28, + "grad_norm": 0.46875, + "learning_rate": 5.7250744154369916e-05, + "loss": 2.5733, + "step": 30102 + }, + { + "epoch": 1.28, + "grad_norm": 0.392578125, + "learning_rate": 5.724468015388711e-05, + "loss": 2.4527, + "step": 30103 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.7238616345787575e-05, + "loss": 2.4307, + "step": 30104 + }, + { + "epoch": 1.28, + "grad_norm": 0.470703125, + "learning_rate": 5.723255273009869e-05, + "loss": 2.4618, + "step": 30105 + }, + { + "epoch": 1.28, + "grad_norm": 0.51953125, + "learning_rate": 5.722648930684764e-05, + "loss": 2.5394, + "step": 30106 + }, + { + "epoch": 1.28, + "grad_norm": 0.53125, + "learning_rate": 5.722042607606179e-05, + "loss": 2.5111, + "step": 30107 + }, + { + "epoch": 1.28, + "grad_norm": 0.466796875, + "learning_rate": 5.7214363037768357e-05, + "loss": 2.451, + "step": 30108 + }, + { + "epoch": 1.28, + "grad_norm": 0.4453125, + "learning_rate": 5.72083001919947e-05, + "loss": 2.5611, + "step": 30109 + }, + { + "epoch": 1.28, + "grad_norm": 0.484375, + "learning_rate": 5.7202237538768e-05, + "loss": 2.4126, + "step": 30110 + }, + { + "epoch": 1.28, + "grad_norm": 0.4609375, + "learning_rate": 5.71961750781156e-05, + "loss": 2.4698, + "step": 30111 + }, + { + "epoch": 1.28, + "grad_norm": 0.49609375, + "learning_rate": 5.719011281006481e-05, + "loss": 2.5846, + "step": 30112 + }, + { + "epoch": 1.28, + "grad_norm": 0.490234375, + "learning_rate": 5.718405073464286e-05, + "loss": 2.4671, + "step": 30113 + }, + { + "epoch": 1.28, + "grad_norm": 0.5, + "learning_rate": 5.7177988851876953e-05, + "loss": 2.5205, + "step": 30114 + }, + { + "epoch": 1.28, + "grad_norm": 0.41015625, + "learning_rate": 5.7171927161794536e-05, + "loss": 2.4233, + "step": 30115 + }, + { + "epoch": 1.28, + "grad_norm": 0.45703125, + "learning_rate": 5.716586566442278e-05, + "loss": 2.5005, + "step": 30116 + }, + { + "epoch": 1.28, + "grad_norm": 0.498046875, + "learning_rate": 5.715980435978894e-05, + "loss": 2.4012, + "step": 30117 + }, + { + "epoch": 1.28, + "grad_norm": 0.466796875, + "learning_rate": 5.7153743247920336e-05, + "loss": 2.4337, + "step": 30118 + }, + { + "epoch": 1.28, + "grad_norm": 0.41015625, + "learning_rate": 5.7147682328844264e-05, + "loss": 2.4333, + "step": 30119 + }, + { + "epoch": 1.28, + "grad_norm": 0.4296875, + "learning_rate": 5.714162160258791e-05, + "loss": 2.4405, + "step": 30120 + }, + { + "epoch": 1.28, + "grad_norm": 0.482421875, + "learning_rate": 5.713556106917865e-05, + "loss": 2.5249, + "step": 30121 + }, + { + "epoch": 1.28, + "grad_norm": 0.451171875, + "learning_rate": 5.712950072864365e-05, + "loss": 2.3984, + "step": 30122 + }, + { + "epoch": 1.28, + "grad_norm": 0.451171875, + "learning_rate": 5.712344058101027e-05, + "loss": 2.3043, + "step": 30123 + }, + { + "epoch": 1.28, + "grad_norm": 0.462890625, + "learning_rate": 5.71173806263057e-05, + "loss": 2.4841, + "step": 30124 + }, + { + "epoch": 1.28, + "grad_norm": 0.404296875, + "learning_rate": 5.7111320864557284e-05, + "loss": 2.4531, + "step": 30125 + }, + { + "epoch": 1.28, + "grad_norm": 0.578125, + "learning_rate": 5.710526129579221e-05, + "loss": 2.3151, + "step": 30126 + }, + { + "epoch": 1.28, + "grad_norm": 0.51171875, + "learning_rate": 5.7099201920037795e-05, + "loss": 2.4807, + "step": 30127 + }, + { + "epoch": 1.28, + "grad_norm": 0.39453125, + "learning_rate": 5.7093142737321315e-05, + "loss": 2.4808, + "step": 30128 + }, + { + "epoch": 1.28, + "grad_norm": 0.439453125, + "learning_rate": 5.7087083747670024e-05, + "loss": 2.4823, + "step": 30129 + }, + { + "epoch": 1.28, + "grad_norm": 0.419921875, + "learning_rate": 5.7081024951111094e-05, + "loss": 2.484, + "step": 30130 + }, + { + "epoch": 1.28, + "grad_norm": 0.44140625, + "learning_rate": 5.707496634767193e-05, + "loss": 2.4921, + "step": 30131 + }, + { + "epoch": 1.28, + "grad_norm": 0.373046875, + "learning_rate": 5.7068907937379756e-05, + "loss": 2.3602, + "step": 30132 + }, + { + "epoch": 1.28, + "grad_norm": 0.5234375, + "learning_rate": 5.706284972026175e-05, + "loss": 2.5898, + "step": 30133 + }, + { + "epoch": 1.28, + "grad_norm": 0.482421875, + "learning_rate": 5.705679169634524e-05, + "loss": 2.5287, + "step": 30134 + }, + { + "epoch": 1.28, + "grad_norm": 0.451171875, + "learning_rate": 5.705073386565749e-05, + "loss": 2.6155, + "step": 30135 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.704467622822572e-05, + "loss": 2.4088, + "step": 30136 + }, + { + "epoch": 1.28, + "grad_norm": 0.44921875, + "learning_rate": 5.7038618784077246e-05, + "loss": 2.5858, + "step": 30137 + }, + { + "epoch": 1.28, + "grad_norm": 0.50390625, + "learning_rate": 5.703256153323925e-05, + "loss": 2.3536, + "step": 30138 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.702650447573905e-05, + "loss": 2.4916, + "step": 30139 + }, + { + "epoch": 1.28, + "grad_norm": 0.462890625, + "learning_rate": 5.7020447611603855e-05, + "loss": 2.4747, + "step": 30140 + }, + { + "epoch": 1.28, + "grad_norm": 0.40625, + "learning_rate": 5.701439094086096e-05, + "loss": 2.3296, + "step": 30141 + }, + { + "epoch": 1.28, + "grad_norm": 0.478515625, + "learning_rate": 5.7008334463537575e-05, + "loss": 2.5428, + "step": 30142 + }, + { + "epoch": 1.28, + "grad_norm": 0.4375, + "learning_rate": 5.700227817966096e-05, + "loss": 2.3962, + "step": 30143 + }, + { + "epoch": 1.28, + "grad_norm": 0.427734375, + "learning_rate": 5.699622208925842e-05, + "loss": 2.5379, + "step": 30144 + }, + { + "epoch": 1.28, + "grad_norm": 0.498046875, + "learning_rate": 5.699016619235717e-05, + "loss": 2.4738, + "step": 30145 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.698411048898438e-05, + "loss": 2.4872, + "step": 30146 + }, + { + "epoch": 1.28, + "grad_norm": 0.52734375, + "learning_rate": 5.697805497916744e-05, + "loss": 2.4902, + "step": 30147 + }, + { + "epoch": 1.28, + "grad_norm": 0.455078125, + "learning_rate": 5.697199966293353e-05, + "loss": 2.5229, + "step": 30148 + }, + { + "epoch": 1.28, + "grad_norm": 0.470703125, + "learning_rate": 5.696594454030986e-05, + "loss": 2.3573, + "step": 30149 + }, + { + "epoch": 1.28, + "grad_norm": 0.58203125, + "learning_rate": 5.6959889611323715e-05, + "loss": 2.499, + "step": 30150 + }, + { + "epoch": 1.28, + "grad_norm": 0.466796875, + "learning_rate": 5.6953834876002365e-05, + "loss": 2.3596, + "step": 30151 + }, + { + "epoch": 1.28, + "grad_norm": 0.53515625, + "learning_rate": 5.6947780334373e-05, + "loss": 2.5236, + "step": 30152 + }, + { + "epoch": 1.28, + "grad_norm": 0.4921875, + "learning_rate": 5.6941725986462925e-05, + "loss": 2.3671, + "step": 30153 + }, + { + "epoch": 1.28, + "grad_norm": 0.421875, + "learning_rate": 5.693567183229931e-05, + "loss": 2.4944, + "step": 30154 + }, + { + "epoch": 1.28, + "grad_norm": 0.494140625, + "learning_rate": 5.692961787190947e-05, + "loss": 2.4971, + "step": 30155 + }, + { + "epoch": 1.29, + "grad_norm": 0.443359375, + "learning_rate": 5.692356410532057e-05, + "loss": 2.4601, + "step": 30156 + }, + { + "epoch": 1.29, + "grad_norm": 0.478515625, + "learning_rate": 5.691751053255994e-05, + "loss": 2.4497, + "step": 30157 + }, + { + "epoch": 1.29, + "grad_norm": 0.490234375, + "learning_rate": 5.691145715365471e-05, + "loss": 2.3547, + "step": 30158 + }, + { + "epoch": 1.29, + "grad_norm": 0.447265625, + "learning_rate": 5.690540396863218e-05, + "loss": 2.4737, + "step": 30159 + }, + { + "epoch": 1.29, + "grad_norm": 0.515625, + "learning_rate": 5.689935097751963e-05, + "loss": 2.4774, + "step": 30160 + }, + { + "epoch": 1.29, + "grad_norm": 0.470703125, + "learning_rate": 5.6893298180344245e-05, + "loss": 2.4497, + "step": 30161 + }, + { + "epoch": 1.29, + "grad_norm": 0.50390625, + "learning_rate": 5.6887245577133186e-05, + "loss": 2.4956, + "step": 30162 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.688119316791385e-05, + "loss": 2.5687, + "step": 30163 + }, + { + "epoch": 1.29, + "grad_norm": 0.46875, + "learning_rate": 5.687514095271338e-05, + "loss": 2.4578, + "step": 30164 + }, + { + "epoch": 1.29, + "grad_norm": 0.52734375, + "learning_rate": 5.686908893155897e-05, + "loss": 2.6463, + "step": 30165 + }, + { + "epoch": 1.29, + "grad_norm": 0.46875, + "learning_rate": 5.68630371044779e-05, + "loss": 2.4146, + "step": 30166 + }, + { + "epoch": 1.29, + "grad_norm": 0.47265625, + "learning_rate": 5.685698547149745e-05, + "loss": 2.5312, + "step": 30167 + }, + { + "epoch": 1.29, + "grad_norm": 0.44921875, + "learning_rate": 5.6850934032644746e-05, + "loss": 2.583, + "step": 30168 + }, + { + "epoch": 1.29, + "grad_norm": 0.43359375, + "learning_rate": 5.6844882787947106e-05, + "loss": 2.4301, + "step": 30169 + }, + { + "epoch": 1.29, + "grad_norm": 0.490234375, + "learning_rate": 5.6838831737431695e-05, + "loss": 2.4607, + "step": 30170 + }, + { + "epoch": 1.29, + "grad_norm": 0.4296875, + "learning_rate": 5.68327808811258e-05, + "loss": 2.3831, + "step": 30171 + }, + { + "epoch": 1.29, + "grad_norm": 0.458984375, + "learning_rate": 5.6826730219056576e-05, + "loss": 2.5467, + "step": 30172 + }, + { + "epoch": 1.29, + "grad_norm": 0.45703125, + "learning_rate": 5.6820679751251335e-05, + "loss": 2.5927, + "step": 30173 + }, + { + "epoch": 1.29, + "grad_norm": 0.458984375, + "learning_rate": 5.681462947773722e-05, + "loss": 2.5127, + "step": 30174 + }, + { + "epoch": 1.29, + "grad_norm": 0.484375, + "learning_rate": 5.680857939854147e-05, + "loss": 2.4688, + "step": 30175 + }, + { + "epoch": 1.29, + "grad_norm": 0.4453125, + "learning_rate": 5.680252951369138e-05, + "loss": 2.4471, + "step": 30176 + }, + { + "epoch": 1.29, + "grad_norm": 0.474609375, + "learning_rate": 5.679647982321412e-05, + "loss": 2.5957, + "step": 30177 + }, + { + "epoch": 1.29, + "grad_norm": 0.54296875, + "learning_rate": 5.679043032713682e-05, + "loss": 2.5459, + "step": 30178 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.678438102548689e-05, + "loss": 2.3628, + "step": 30179 + }, + { + "epoch": 1.29, + "grad_norm": 0.42578125, + "learning_rate": 5.677833191829145e-05, + "loss": 2.5352, + "step": 30180 + }, + { + "epoch": 1.29, + "grad_norm": 0.47265625, + "learning_rate": 5.677228300557766e-05, + "loss": 2.4764, + "step": 30181 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.676623428737282e-05, + "loss": 2.3385, + "step": 30182 + }, + { + "epoch": 1.29, + "grad_norm": 0.44140625, + "learning_rate": 5.676018576370417e-05, + "loss": 2.4644, + "step": 30183 + }, + { + "epoch": 1.29, + "grad_norm": 0.470703125, + "learning_rate": 5.675413743459882e-05, + "loss": 2.473, + "step": 30184 + }, + { + "epoch": 1.29, + "grad_norm": 0.443359375, + "learning_rate": 5.6748089300084105e-05, + "loss": 2.4769, + "step": 30185 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.6742041360187126e-05, + "loss": 2.4171, + "step": 30186 + }, + { + "epoch": 1.29, + "grad_norm": 0.4375, + "learning_rate": 5.673599361493516e-05, + "loss": 2.4934, + "step": 30187 + }, + { + "epoch": 1.29, + "grad_norm": 0.412109375, + "learning_rate": 5.6729946064355455e-05, + "loss": 2.4712, + "step": 30188 + }, + { + "epoch": 1.29, + "grad_norm": 0.470703125, + "learning_rate": 5.672389870847518e-05, + "loss": 2.4658, + "step": 30189 + }, + { + "epoch": 1.29, + "grad_norm": 0.4921875, + "learning_rate": 5.67178515473215e-05, + "loss": 2.5134, + "step": 30190 + }, + { + "epoch": 1.29, + "grad_norm": 0.486328125, + "learning_rate": 5.671180458092167e-05, + "loss": 2.4217, + "step": 30191 + }, + { + "epoch": 1.29, + "grad_norm": 0.51171875, + "learning_rate": 5.670575780930294e-05, + "loss": 2.5857, + "step": 30192 + }, + { + "epoch": 1.29, + "grad_norm": 0.431640625, + "learning_rate": 5.669971123249247e-05, + "loss": 2.5946, + "step": 30193 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.669366485051745e-05, + "loss": 2.5335, + "step": 30194 + }, + { + "epoch": 1.29, + "grad_norm": 0.52734375, + "learning_rate": 5.668761866340511e-05, + "loss": 2.4377, + "step": 30195 + }, + { + "epoch": 1.29, + "grad_norm": 0.50390625, + "learning_rate": 5.668157267118268e-05, + "loss": 2.4825, + "step": 30196 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.667552687387731e-05, + "loss": 2.5046, + "step": 30197 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.666948127151629e-05, + "loss": 2.4917, + "step": 30198 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.6663435864126715e-05, + "loss": 2.4084, + "step": 30199 + }, + { + "epoch": 1.29, + "grad_norm": 0.48828125, + "learning_rate": 5.6657390651735844e-05, + "loss": 2.5732, + "step": 30200 + }, + { + "epoch": 1.29, + "grad_norm": 0.46875, + "learning_rate": 5.665134563437091e-05, + "loss": 2.456, + "step": 30201 + }, + { + "epoch": 1.29, + "grad_norm": 0.53125, + "learning_rate": 5.664530081205909e-05, + "loss": 2.4282, + "step": 30202 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.663925618482748e-05, + "loss": 2.3473, + "step": 30203 + }, + { + "epoch": 1.29, + "grad_norm": 0.482421875, + "learning_rate": 5.6633211752703464e-05, + "loss": 2.3862, + "step": 30204 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.662716751571413e-05, + "loss": 2.6367, + "step": 30205 + }, + { + "epoch": 1.29, + "grad_norm": 0.486328125, + "learning_rate": 5.662112347388666e-05, + "loss": 2.2926, + "step": 30206 + }, + { + "epoch": 1.29, + "grad_norm": 0.43359375, + "learning_rate": 5.661507962724828e-05, + "loss": 2.5679, + "step": 30207 + }, + { + "epoch": 1.29, + "grad_norm": 0.51171875, + "learning_rate": 5.6609035975826233e-05, + "loss": 2.4276, + "step": 30208 + }, + { + "epoch": 1.29, + "grad_norm": 0.53515625, + "learning_rate": 5.660299251964767e-05, + "loss": 2.4718, + "step": 30209 + }, + { + "epoch": 1.29, + "grad_norm": 0.46484375, + "learning_rate": 5.6596949258739726e-05, + "loss": 2.3863, + "step": 30210 + }, + { + "epoch": 1.29, + "grad_norm": 0.43359375, + "learning_rate": 5.6590906193129655e-05, + "loss": 2.5725, + "step": 30211 + }, + { + "epoch": 1.29, + "grad_norm": 0.5, + "learning_rate": 5.658486332284467e-05, + "loss": 2.5066, + "step": 30212 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.657882064791191e-05, + "loss": 2.4709, + "step": 30213 + }, + { + "epoch": 1.29, + "grad_norm": 0.484375, + "learning_rate": 5.657277816835862e-05, + "loss": 2.4343, + "step": 30214 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.656673588421193e-05, + "loss": 2.4629, + "step": 30215 + }, + { + "epoch": 1.29, + "grad_norm": 0.44140625, + "learning_rate": 5.6560693795499046e-05, + "loss": 2.4421, + "step": 30216 + }, + { + "epoch": 1.29, + "grad_norm": 0.50390625, + "learning_rate": 5.6554651902247205e-05, + "loss": 2.4865, + "step": 30217 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.654861020448354e-05, + "loss": 2.3754, + "step": 30218 + }, + { + "epoch": 1.29, + "grad_norm": 0.412109375, + "learning_rate": 5.65425687022352e-05, + "loss": 2.4222, + "step": 30219 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.653652739552948e-05, + "loss": 2.3681, + "step": 30220 + }, + { + "epoch": 1.29, + "grad_norm": 0.455078125, + "learning_rate": 5.65304862843935e-05, + "loss": 2.4628, + "step": 30221 + }, + { + "epoch": 1.29, + "grad_norm": 0.421875, + "learning_rate": 5.652444536885441e-05, + "loss": 2.4543, + "step": 30222 + }, + { + "epoch": 1.29, + "grad_norm": 0.453125, + "learning_rate": 5.651840464893943e-05, + "loss": 2.4986, + "step": 30223 + }, + { + "epoch": 1.29, + "grad_norm": 0.482421875, + "learning_rate": 5.6512364124675775e-05, + "loss": 2.3927, + "step": 30224 + }, + { + "epoch": 1.29, + "grad_norm": 0.5234375, + "learning_rate": 5.650632379609058e-05, + "loss": 2.4534, + "step": 30225 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.650028366321101e-05, + "loss": 2.4292, + "step": 30226 + }, + { + "epoch": 1.29, + "grad_norm": 0.498046875, + "learning_rate": 5.6494243726064245e-05, + "loss": 2.3702, + "step": 30227 + }, + { + "epoch": 1.29, + "grad_norm": 0.486328125, + "learning_rate": 5.6488203984677536e-05, + "loss": 2.492, + "step": 30228 + }, + { + "epoch": 1.29, + "grad_norm": 0.453125, + "learning_rate": 5.648216443907797e-05, + "loss": 2.566, + "step": 30229 + }, + { + "epoch": 1.29, + "grad_norm": 0.427734375, + "learning_rate": 5.6476125089292805e-05, + "loss": 2.3926, + "step": 30230 + }, + { + "epoch": 1.29, + "grad_norm": 0.462890625, + "learning_rate": 5.647008593534913e-05, + "loss": 2.338, + "step": 30231 + }, + { + "epoch": 1.29, + "grad_norm": 0.46875, + "learning_rate": 5.646404697727414e-05, + "loss": 2.553, + "step": 30232 + }, + { + "epoch": 1.29, + "grad_norm": 0.462890625, + "learning_rate": 5.645800821509508e-05, + "loss": 2.5392, + "step": 30233 + }, + { + "epoch": 1.29, + "grad_norm": 0.40625, + "learning_rate": 5.645196964883908e-05, + "loss": 2.3967, + "step": 30234 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.644593127853322e-05, + "loss": 2.3883, + "step": 30235 + }, + { + "epoch": 1.29, + "grad_norm": 0.4140625, + "learning_rate": 5.643989310420481e-05, + "loss": 2.3364, + "step": 30236 + }, + { + "epoch": 1.29, + "grad_norm": 0.482421875, + "learning_rate": 5.6433855125880974e-05, + "loss": 2.3646, + "step": 30237 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.642781734358884e-05, + "loss": 2.4542, + "step": 30238 + }, + { + "epoch": 1.29, + "grad_norm": 0.470703125, + "learning_rate": 5.642177975735559e-05, + "loss": 2.5579, + "step": 30239 + }, + { + "epoch": 1.29, + "grad_norm": 0.451171875, + "learning_rate": 5.641574236720846e-05, + "loss": 2.4921, + "step": 30240 + }, + { + "epoch": 1.29, + "grad_norm": 0.46484375, + "learning_rate": 5.64097051731745e-05, + "loss": 2.4171, + "step": 30241 + }, + { + "epoch": 1.29, + "grad_norm": 0.443359375, + "learning_rate": 5.6403668175280996e-05, + "loss": 2.4409, + "step": 30242 + }, + { + "epoch": 1.29, + "grad_norm": 0.4140625, + "learning_rate": 5.6397631373555e-05, + "loss": 2.5523, + "step": 30243 + }, + { + "epoch": 1.29, + "grad_norm": 0.53515625, + "learning_rate": 5.639159476802378e-05, + "loss": 2.4587, + "step": 30244 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.63855583587144e-05, + "loss": 2.3104, + "step": 30245 + }, + { + "epoch": 1.29, + "grad_norm": 0.49609375, + "learning_rate": 5.6379522145654096e-05, + "loss": 2.4565, + "step": 30246 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.637348612886998e-05, + "loss": 2.5355, + "step": 30247 + }, + { + "epoch": 1.29, + "grad_norm": 0.482421875, + "learning_rate": 5.636745030838923e-05, + "loss": 2.4928, + "step": 30248 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.636141468423903e-05, + "loss": 2.3967, + "step": 30249 + }, + { + "epoch": 1.29, + "grad_norm": 0.470703125, + "learning_rate": 5.635537925644654e-05, + "loss": 2.4358, + "step": 30250 + }, + { + "epoch": 1.29, + "grad_norm": 0.50390625, + "learning_rate": 5.6349344025038795e-05, + "loss": 2.4514, + "step": 30251 + }, + { + "epoch": 1.29, + "grad_norm": 0.515625, + "learning_rate": 5.634330899004314e-05, + "loss": 2.6431, + "step": 30252 + }, + { + "epoch": 1.29, + "grad_norm": 0.5078125, + "learning_rate": 5.633727415148663e-05, + "loss": 2.4219, + "step": 30253 + }, + { + "epoch": 1.29, + "grad_norm": 0.474609375, + "learning_rate": 5.633123950939639e-05, + "loss": 2.4804, + "step": 30254 + }, + { + "epoch": 1.29, + "grad_norm": 0.4453125, + "learning_rate": 5.632520506379963e-05, + "loss": 2.2833, + "step": 30255 + }, + { + "epoch": 1.29, + "grad_norm": 0.427734375, + "learning_rate": 5.631917081472351e-05, + "loss": 2.4109, + "step": 30256 + }, + { + "epoch": 1.29, + "grad_norm": 0.470703125, + "learning_rate": 5.6313136762195116e-05, + "loss": 2.4381, + "step": 30257 + }, + { + "epoch": 1.29, + "grad_norm": 0.55078125, + "learning_rate": 5.6307102906241696e-05, + "loss": 2.5126, + "step": 30258 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.6301069246890294e-05, + "loss": 2.489, + "step": 30259 + }, + { + "epoch": 1.29, + "grad_norm": 0.416015625, + "learning_rate": 5.6295035784168145e-05, + "loss": 2.3653, + "step": 30260 + }, + { + "epoch": 1.29, + "grad_norm": 0.447265625, + "learning_rate": 5.628900251810234e-05, + "loss": 2.447, + "step": 30261 + }, + { + "epoch": 1.29, + "grad_norm": 0.5, + "learning_rate": 5.6282969448720066e-05, + "loss": 2.4537, + "step": 30262 + }, + { + "epoch": 1.29, + "grad_norm": 0.439453125, + "learning_rate": 5.6276936576048434e-05, + "loss": 2.3691, + "step": 30263 + }, + { + "epoch": 1.29, + "grad_norm": 0.412109375, + "learning_rate": 5.62709039001146e-05, + "loss": 2.3616, + "step": 30264 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.6264871420945754e-05, + "loss": 2.521, + "step": 30265 + }, + { + "epoch": 1.29, + "grad_norm": 0.474609375, + "learning_rate": 5.625883913856901e-05, + "loss": 2.3895, + "step": 30266 + }, + { + "epoch": 1.29, + "grad_norm": 0.5078125, + "learning_rate": 5.625280705301141e-05, + "loss": 2.5024, + "step": 30267 + }, + { + "epoch": 1.29, + "grad_norm": 0.51953125, + "learning_rate": 5.624677516430029e-05, + "loss": 2.5874, + "step": 30268 + }, + { + "epoch": 1.29, + "grad_norm": 0.48046875, + "learning_rate": 5.624074347246269e-05, + "loss": 2.4693, + "step": 30269 + }, + { + "epoch": 1.29, + "grad_norm": 0.453125, + "learning_rate": 5.62347119775257e-05, + "loss": 2.256, + "step": 30270 + }, + { + "epoch": 1.29, + "grad_norm": 0.412109375, + "learning_rate": 5.622868067951652e-05, + "loss": 2.3323, + "step": 30271 + }, + { + "epoch": 1.29, + "grad_norm": 0.421875, + "learning_rate": 5.622264957846231e-05, + "loss": 2.3966, + "step": 30272 + }, + { + "epoch": 1.29, + "grad_norm": 0.50390625, + "learning_rate": 5.6216618674390145e-05, + "loss": 2.4567, + "step": 30273 + }, + { + "epoch": 1.29, + "grad_norm": 0.462890625, + "learning_rate": 5.6210587967327236e-05, + "loss": 2.422, + "step": 30274 + }, + { + "epoch": 1.29, + "grad_norm": 0.443359375, + "learning_rate": 5.620455745730064e-05, + "loss": 2.4004, + "step": 30275 + }, + { + "epoch": 1.29, + "grad_norm": 0.439453125, + "learning_rate": 5.6198527144337566e-05, + "loss": 2.5299, + "step": 30276 + }, + { + "epoch": 1.29, + "grad_norm": 0.40625, + "learning_rate": 5.619249702846508e-05, + "loss": 2.2843, + "step": 30277 + }, + { + "epoch": 1.29, + "grad_norm": 0.498046875, + "learning_rate": 5.618646710971038e-05, + "loss": 2.429, + "step": 30278 + }, + { + "epoch": 1.29, + "grad_norm": 0.486328125, + "learning_rate": 5.6180437388100525e-05, + "loss": 2.5523, + "step": 30279 + }, + { + "epoch": 1.29, + "grad_norm": 0.5546875, + "learning_rate": 5.617440786366268e-05, + "loss": 2.5537, + "step": 30280 + }, + { + "epoch": 1.29, + "grad_norm": 0.443359375, + "learning_rate": 5.616837853642405e-05, + "loss": 2.477, + "step": 30281 + }, + { + "epoch": 1.29, + "grad_norm": 0.451171875, + "learning_rate": 5.616234940641168e-05, + "loss": 2.4327, + "step": 30282 + }, + { + "epoch": 1.29, + "grad_norm": 0.51953125, + "learning_rate": 5.615632047365262e-05, + "loss": 2.4657, + "step": 30283 + }, + { + "epoch": 1.29, + "grad_norm": 0.470703125, + "learning_rate": 5.615029173817421e-05, + "loss": 2.4272, + "step": 30284 + }, + { + "epoch": 1.29, + "grad_norm": 0.494140625, + "learning_rate": 5.6144263200003435e-05, + "loss": 2.419, + "step": 30285 + }, + { + "epoch": 1.29, + "grad_norm": 0.46484375, + "learning_rate": 5.613823485916743e-05, + "loss": 2.3717, + "step": 30286 + }, + { + "epoch": 1.29, + "grad_norm": 0.42578125, + "learning_rate": 5.6132206715693314e-05, + "loss": 2.4667, + "step": 30287 + }, + { + "epoch": 1.29, + "grad_norm": 0.46875, + "learning_rate": 5.6126178769608286e-05, + "loss": 2.4634, + "step": 30288 + }, + { + "epoch": 1.29, + "grad_norm": 0.453125, + "learning_rate": 5.612015102093939e-05, + "loss": 2.428, + "step": 30289 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.611412346971381e-05, + "loss": 2.5552, + "step": 30290 + }, + { + "epoch": 1.29, + "grad_norm": 0.486328125, + "learning_rate": 5.6108096115958595e-05, + "loss": 2.5342, + "step": 30291 + }, + { + "epoch": 1.29, + "grad_norm": 0.52734375, + "learning_rate": 5.610206895970094e-05, + "loss": 2.3605, + "step": 30292 + }, + { + "epoch": 1.29, + "grad_norm": 0.4921875, + "learning_rate": 5.6096042000967896e-05, + "loss": 2.3836, + "step": 30293 + }, + { + "epoch": 1.29, + "grad_norm": 0.47265625, + "learning_rate": 5.609001523978667e-05, + "loss": 2.2457, + "step": 30294 + }, + { + "epoch": 1.29, + "grad_norm": 0.416015625, + "learning_rate": 5.6083988676184275e-05, + "loss": 2.3072, + "step": 30295 + }, + { + "epoch": 1.29, + "grad_norm": 0.51953125, + "learning_rate": 5.6077962310187894e-05, + "loss": 2.5133, + "step": 30296 + }, + { + "epoch": 1.29, + "grad_norm": 0.4453125, + "learning_rate": 5.6071936141824655e-05, + "loss": 2.4808, + "step": 30297 + }, + { + "epoch": 1.29, + "grad_norm": 0.42578125, + "learning_rate": 5.6065910171121664e-05, + "loss": 2.3401, + "step": 30298 + }, + { + "epoch": 1.29, + "grad_norm": 0.48046875, + "learning_rate": 5.6059884398105924e-05, + "loss": 2.4732, + "step": 30299 + }, + { + "epoch": 1.29, + "grad_norm": 0.474609375, + "learning_rate": 5.6053858822804736e-05, + "loss": 2.426, + "step": 30300 + }, + { + "epoch": 1.29, + "grad_norm": 0.5078125, + "learning_rate": 5.604783344524514e-05, + "loss": 2.3821, + "step": 30301 + }, + { + "epoch": 1.29, + "grad_norm": 0.44140625, + "learning_rate": 5.604180826545417e-05, + "loss": 2.5742, + "step": 30302 + }, + { + "epoch": 1.29, + "grad_norm": 0.498046875, + "learning_rate": 5.6035783283458996e-05, + "loss": 2.553, + "step": 30303 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.602975849928678e-05, + "loss": 2.4389, + "step": 30304 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.6023733912964536e-05, + "loss": 2.373, + "step": 30305 + }, + { + "epoch": 1.29, + "grad_norm": 0.412109375, + "learning_rate": 5.6017709524519456e-05, + "loss": 2.5618, + "step": 30306 + }, + { + "epoch": 1.29, + "grad_norm": 0.50390625, + "learning_rate": 5.601168533397858e-05, + "loss": 2.6117, + "step": 30307 + }, + { + "epoch": 1.29, + "grad_norm": 0.46875, + "learning_rate": 5.600566134136906e-05, + "loss": 2.428, + "step": 30308 + }, + { + "epoch": 1.29, + "grad_norm": 0.53515625, + "learning_rate": 5.599963754671798e-05, + "loss": 2.4586, + "step": 30309 + }, + { + "epoch": 1.29, + "grad_norm": 0.419921875, + "learning_rate": 5.599361395005247e-05, + "loss": 2.4135, + "step": 30310 + }, + { + "epoch": 1.29, + "grad_norm": 0.478515625, + "learning_rate": 5.598759055139957e-05, + "loss": 2.4173, + "step": 30311 + }, + { + "epoch": 1.29, + "grad_norm": 0.43359375, + "learning_rate": 5.598156735078643e-05, + "loss": 2.4426, + "step": 30312 + }, + { + "epoch": 1.29, + "grad_norm": 0.49609375, + "learning_rate": 5.597554434824021e-05, + "loss": 2.485, + "step": 30313 + }, + { + "epoch": 1.29, + "grad_norm": 0.515625, + "learning_rate": 5.596952154378794e-05, + "loss": 2.4876, + "step": 30314 + }, + { + "epoch": 1.29, + "grad_norm": 0.51171875, + "learning_rate": 5.5963498937456684e-05, + "loss": 2.4163, + "step": 30315 + }, + { + "epoch": 1.29, + "grad_norm": 0.41796875, + "learning_rate": 5.595747652927359e-05, + "loss": 2.3003, + "step": 30316 + }, + { + "epoch": 1.29, + "grad_norm": 0.482421875, + "learning_rate": 5.595145431926581e-05, + "loss": 2.4669, + "step": 30317 + }, + { + "epoch": 1.29, + "grad_norm": 0.51171875, + "learning_rate": 5.594543230746033e-05, + "loss": 2.6349, + "step": 30318 + }, + { + "epoch": 1.29, + "grad_norm": 0.458984375, + "learning_rate": 5.593941049388436e-05, + "loss": 2.4247, + "step": 30319 + }, + { + "epoch": 1.29, + "grad_norm": 0.455078125, + "learning_rate": 5.59333888785649e-05, + "loss": 2.4542, + "step": 30320 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.592736746152907e-05, + "loss": 2.4597, + "step": 30321 + }, + { + "epoch": 1.29, + "grad_norm": 0.53125, + "learning_rate": 5.592134624280403e-05, + "loss": 2.4895, + "step": 30322 + }, + { + "epoch": 1.29, + "grad_norm": 0.546875, + "learning_rate": 5.591532522241682e-05, + "loss": 2.4056, + "step": 30323 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.59093044003945e-05, + "loss": 2.5164, + "step": 30324 + }, + { + "epoch": 1.29, + "grad_norm": 0.5234375, + "learning_rate": 5.5903283776764196e-05, + "loss": 2.5513, + "step": 30325 + }, + { + "epoch": 1.29, + "grad_norm": 0.51953125, + "learning_rate": 5.589726335155303e-05, + "loss": 2.4141, + "step": 30326 + }, + { + "epoch": 1.29, + "grad_norm": 0.48828125, + "learning_rate": 5.5891243124788027e-05, + "loss": 2.5963, + "step": 30327 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.588522309649631e-05, + "loss": 2.4102, + "step": 30328 + }, + { + "epoch": 1.29, + "grad_norm": 0.47265625, + "learning_rate": 5.5879203266705014e-05, + "loss": 2.5597, + "step": 30329 + }, + { + "epoch": 1.29, + "grad_norm": 0.52734375, + "learning_rate": 5.5873183635441175e-05, + "loss": 2.4855, + "step": 30330 + }, + { + "epoch": 1.29, + "grad_norm": 0.470703125, + "learning_rate": 5.5867164202731834e-05, + "loss": 2.4871, + "step": 30331 + }, + { + "epoch": 1.29, + "grad_norm": 0.408203125, + "learning_rate": 5.586114496860413e-05, + "loss": 2.4158, + "step": 30332 + }, + { + "epoch": 1.29, + "grad_norm": 0.46875, + "learning_rate": 5.5855125933085194e-05, + "loss": 2.4358, + "step": 30333 + }, + { + "epoch": 1.29, + "grad_norm": 0.5390625, + "learning_rate": 5.584910709620201e-05, + "loss": 2.4939, + "step": 30334 + }, + { + "epoch": 1.29, + "grad_norm": 0.490234375, + "learning_rate": 5.584308845798174e-05, + "loss": 2.4536, + "step": 30335 + }, + { + "epoch": 1.29, + "grad_norm": 0.431640625, + "learning_rate": 5.583707001845141e-05, + "loss": 2.4443, + "step": 30336 + }, + { + "epoch": 1.29, + "grad_norm": 0.427734375, + "learning_rate": 5.583105177763812e-05, + "loss": 2.373, + "step": 30337 + }, + { + "epoch": 1.29, + "grad_norm": 0.443359375, + "learning_rate": 5.5825033735569e-05, + "loss": 2.3699, + "step": 30338 + }, + { + "epoch": 1.29, + "grad_norm": 0.435546875, + "learning_rate": 5.5819015892271076e-05, + "loss": 2.4276, + "step": 30339 + }, + { + "epoch": 1.29, + "grad_norm": 0.419921875, + "learning_rate": 5.5812998247771384e-05, + "loss": 2.3694, + "step": 30340 + }, + { + "epoch": 1.29, + "grad_norm": 0.51953125, + "learning_rate": 5.580698080209708e-05, + "loss": 2.5587, + "step": 30341 + }, + { + "epoch": 1.29, + "grad_norm": 0.45703125, + "learning_rate": 5.580096355527522e-05, + "loss": 2.4706, + "step": 30342 + }, + { + "epoch": 1.29, + "grad_norm": 0.53515625, + "learning_rate": 5.5794946507332846e-05, + "loss": 2.4383, + "step": 30343 + }, + { + "epoch": 1.29, + "grad_norm": 0.48828125, + "learning_rate": 5.578892965829706e-05, + "loss": 2.5097, + "step": 30344 + }, + { + "epoch": 1.29, + "grad_norm": 0.44140625, + "learning_rate": 5.578291300819497e-05, + "loss": 2.384, + "step": 30345 + }, + { + "epoch": 1.29, + "grad_norm": 0.494140625, + "learning_rate": 5.577689655705362e-05, + "loss": 2.3701, + "step": 30346 + }, + { + "epoch": 1.29, + "grad_norm": 0.40625, + "learning_rate": 5.5770880304900016e-05, + "loss": 2.3093, + "step": 30347 + }, + { + "epoch": 1.29, + "grad_norm": 0.47265625, + "learning_rate": 5.5764864251761306e-05, + "loss": 2.4697, + "step": 30348 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.5758848397664564e-05, + "loss": 2.469, + "step": 30349 + }, + { + "epoch": 1.29, + "grad_norm": 0.498046875, + "learning_rate": 5.575283274263681e-05, + "loss": 2.522, + "step": 30350 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.574681728670517e-05, + "loss": 2.4101, + "step": 30351 + }, + { + "epoch": 1.29, + "grad_norm": 0.44140625, + "learning_rate": 5.574080202989664e-05, + "loss": 2.3795, + "step": 30352 + }, + { + "epoch": 1.29, + "grad_norm": 0.49609375, + "learning_rate": 5.573478697223833e-05, + "loss": 2.4396, + "step": 30353 + }, + { + "epoch": 1.29, + "grad_norm": 0.447265625, + "learning_rate": 5.5728772113757335e-05, + "loss": 2.487, + "step": 30354 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.572275745448069e-05, + "loss": 2.6075, + "step": 30355 + }, + { + "epoch": 1.29, + "grad_norm": 0.474609375, + "learning_rate": 5.571674299443541e-05, + "loss": 2.4998, + "step": 30356 + }, + { + "epoch": 1.29, + "grad_norm": 0.484375, + "learning_rate": 5.571072873364862e-05, + "loss": 2.586, + "step": 30357 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.57047146721474e-05, + "loss": 2.3445, + "step": 30358 + }, + { + "epoch": 1.29, + "grad_norm": 0.51171875, + "learning_rate": 5.569870080995872e-05, + "loss": 2.4575, + "step": 30359 + }, + { + "epoch": 1.29, + "grad_norm": 0.443359375, + "learning_rate": 5.56926871471097e-05, + "loss": 2.5225, + "step": 30360 + }, + { + "epoch": 1.29, + "grad_norm": 0.478515625, + "learning_rate": 5.568667368362745e-05, + "loss": 2.4981, + "step": 30361 + }, + { + "epoch": 1.29, + "grad_norm": 0.42578125, + "learning_rate": 5.568066041953896e-05, + "loss": 2.5056, + "step": 30362 + }, + { + "epoch": 1.29, + "grad_norm": 0.447265625, + "learning_rate": 5.567464735487127e-05, + "loss": 2.4797, + "step": 30363 + }, + { + "epoch": 1.29, + "grad_norm": 0.400390625, + "learning_rate": 5.566863448965145e-05, + "loss": 2.4301, + "step": 30364 + }, + { + "epoch": 1.29, + "grad_norm": 0.50390625, + "learning_rate": 5.5662621823906626e-05, + "loss": 2.5654, + "step": 30365 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.565660935766377e-05, + "loss": 2.469, + "step": 30366 + }, + { + "epoch": 1.29, + "grad_norm": 0.435546875, + "learning_rate": 5.565059709095e-05, + "loss": 2.3019, + "step": 30367 + }, + { + "epoch": 1.29, + "grad_norm": 0.482421875, + "learning_rate": 5.5644585023792286e-05, + "loss": 2.6009, + "step": 30368 + }, + { + "epoch": 1.29, + "grad_norm": 0.4921875, + "learning_rate": 5.563857315621774e-05, + "loss": 2.4838, + "step": 30369 + }, + { + "epoch": 1.29, + "grad_norm": 0.494140625, + "learning_rate": 5.563256148825343e-05, + "loss": 2.5011, + "step": 30370 + }, + { + "epoch": 1.29, + "grad_norm": 0.474609375, + "learning_rate": 5.562655001992639e-05, + "loss": 2.4316, + "step": 30371 + }, + { + "epoch": 1.29, + "grad_norm": 0.43359375, + "learning_rate": 5.562053875126362e-05, + "loss": 2.4814, + "step": 30372 + }, + { + "epoch": 1.29, + "grad_norm": 0.4765625, + "learning_rate": 5.56145276822922e-05, + "loss": 2.4363, + "step": 30373 + }, + { + "epoch": 1.29, + "grad_norm": 0.421875, + "learning_rate": 5.560851681303922e-05, + "loss": 2.3584, + "step": 30374 + }, + { + "epoch": 1.29, + "grad_norm": 0.4609375, + "learning_rate": 5.560250614353166e-05, + "loss": 2.4634, + "step": 30375 + }, + { + "epoch": 1.29, + "grad_norm": 0.51171875, + "learning_rate": 5.55964956737966e-05, + "loss": 2.5985, + "step": 30376 + }, + { + "epoch": 1.29, + "grad_norm": 0.458984375, + "learning_rate": 5.5590485403861116e-05, + "loss": 2.3468, + "step": 30377 + }, + { + "epoch": 1.29, + "grad_norm": 0.4453125, + "learning_rate": 5.558447533375222e-05, + "loss": 2.5268, + "step": 30378 + }, + { + "epoch": 1.29, + "grad_norm": 0.5234375, + "learning_rate": 5.557846546349691e-05, + "loss": 2.4943, + "step": 30379 + }, + { + "epoch": 1.29, + "grad_norm": 0.49609375, + "learning_rate": 5.557245579312227e-05, + "loss": 2.4738, + "step": 30380 + }, + { + "epoch": 1.29, + "grad_norm": 0.478515625, + "learning_rate": 5.5566446322655385e-05, + "loss": 2.5305, + "step": 30381 + }, + { + "epoch": 1.29, + "grad_norm": 0.427734375, + "learning_rate": 5.5560437052123215e-05, + "loss": 2.4278, + "step": 30382 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.5554427981552884e-05, + "loss": 2.3383, + "step": 30383 + }, + { + "epoch": 1.29, + "grad_norm": 0.453125, + "learning_rate": 5.554841911097133e-05, + "loss": 2.5211, + "step": 30384 + }, + { + "epoch": 1.29, + "grad_norm": 0.466796875, + "learning_rate": 5.5542410440405646e-05, + "loss": 2.3584, + "step": 30385 + }, + { + "epoch": 1.29, + "grad_norm": 0.53515625, + "learning_rate": 5.5536401969882915e-05, + "loss": 2.5486, + "step": 30386 + }, + { + "epoch": 1.29, + "grad_norm": 0.52734375, + "learning_rate": 5.553039369943012e-05, + "loss": 2.5585, + "step": 30387 + }, + { + "epoch": 1.29, + "grad_norm": 0.494140625, + "learning_rate": 5.552438562907427e-05, + "loss": 2.4963, + "step": 30388 + }, + { + "epoch": 1.29, + "grad_norm": 0.474609375, + "learning_rate": 5.551837775884242e-05, + "loss": 2.534, + "step": 30389 + }, + { + "epoch": 1.3, + "grad_norm": 0.421875, + "learning_rate": 5.551237008876166e-05, + "loss": 2.5172, + "step": 30390 + }, + { + "epoch": 1.3, + "grad_norm": 0.490234375, + "learning_rate": 5.550636261885893e-05, + "loss": 2.4566, + "step": 30391 + }, + { + "epoch": 1.3, + "grad_norm": 0.4453125, + "learning_rate": 5.550035534916132e-05, + "loss": 2.4462, + "step": 30392 + }, + { + "epoch": 1.3, + "grad_norm": 0.482421875, + "learning_rate": 5.549434827969588e-05, + "loss": 2.5104, + "step": 30393 + }, + { + "epoch": 1.3, + "grad_norm": 0.4609375, + "learning_rate": 5.548834141048961e-05, + "loss": 2.4967, + "step": 30394 + }, + { + "epoch": 1.3, + "grad_norm": 0.48046875, + "learning_rate": 5.54823347415695e-05, + "loss": 2.4613, + "step": 30395 + }, + { + "epoch": 1.3, + "grad_norm": 0.41796875, + "learning_rate": 5.547632827296261e-05, + "loss": 2.3546, + "step": 30396 + }, + { + "epoch": 1.3, + "grad_norm": 0.51953125, + "learning_rate": 5.5470322004695995e-05, + "loss": 2.4517, + "step": 30397 + }, + { + "epoch": 1.3, + "grad_norm": 0.51953125, + "learning_rate": 5.546431593679664e-05, + "loss": 2.4513, + "step": 30398 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.5458310069291616e-05, + "loss": 2.5266, + "step": 30399 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.545230440220788e-05, + "loss": 2.4644, + "step": 30400 + }, + { + "epoch": 1.3, + "grad_norm": 0.46484375, + "learning_rate": 5.544629893557251e-05, + "loss": 2.4085, + "step": 30401 + }, + { + "epoch": 1.3, + "grad_norm": 0.52734375, + "learning_rate": 5.544029366941252e-05, + "loss": 2.5979, + "step": 30402 + }, + { + "epoch": 1.3, + "grad_norm": 0.44921875, + "learning_rate": 5.543428860375495e-05, + "loss": 2.4461, + "step": 30403 + }, + { + "epoch": 1.3, + "grad_norm": 0.482421875, + "learning_rate": 5.542828373862675e-05, + "loss": 2.475, + "step": 30404 + }, + { + "epoch": 1.3, + "grad_norm": 0.46875, + "learning_rate": 5.542227907405498e-05, + "loss": 2.5543, + "step": 30405 + }, + { + "epoch": 1.3, + "grad_norm": 0.51953125, + "learning_rate": 5.541627461006671e-05, + "loss": 2.4758, + "step": 30406 + }, + { + "epoch": 1.3, + "grad_norm": 0.447265625, + "learning_rate": 5.541027034668886e-05, + "loss": 2.2005, + "step": 30407 + }, + { + "epoch": 1.3, + "grad_norm": 0.5, + "learning_rate": 5.5404266283948505e-05, + "loss": 2.5519, + "step": 30408 + }, + { + "epoch": 1.3, + "grad_norm": 0.5078125, + "learning_rate": 5.539826242187269e-05, + "loss": 2.5262, + "step": 30409 + }, + { + "epoch": 1.3, + "grad_norm": 0.443359375, + "learning_rate": 5.539225876048839e-05, + "loss": 2.4909, + "step": 30410 + }, + { + "epoch": 1.3, + "grad_norm": 0.44921875, + "learning_rate": 5.53862552998226e-05, + "loss": 2.4676, + "step": 30411 + }, + { + "epoch": 1.3, + "grad_norm": 0.466796875, + "learning_rate": 5.538025203990234e-05, + "loss": 2.3837, + "step": 30412 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.537424898075468e-05, + "loss": 2.4518, + "step": 30413 + }, + { + "epoch": 1.3, + "grad_norm": 0.48046875, + "learning_rate": 5.536824612240656e-05, + "loss": 2.3682, + "step": 30414 + }, + { + "epoch": 1.3, + "grad_norm": 0.4140625, + "learning_rate": 5.5362243464885055e-05, + "loss": 2.5852, + "step": 30415 + }, + { + "epoch": 1.3, + "grad_norm": 0.498046875, + "learning_rate": 5.53562410082171e-05, + "loss": 2.3844, + "step": 30416 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.535023875242975e-05, + "loss": 2.5165, + "step": 30417 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.5344236697550046e-05, + "loss": 2.4945, + "step": 30418 + }, + { + "epoch": 1.3, + "grad_norm": 0.5546875, + "learning_rate": 5.533823484360496e-05, + "loss": 2.4761, + "step": 30419 + }, + { + "epoch": 1.3, + "grad_norm": 0.4765625, + "learning_rate": 5.5332233190621453e-05, + "loss": 2.497, + "step": 30420 + }, + { + "epoch": 1.3, + "grad_norm": 0.515625, + "learning_rate": 5.5326231738626564e-05, + "loss": 2.5192, + "step": 30421 + }, + { + "epoch": 1.3, + "grad_norm": 0.4375, + "learning_rate": 5.532023048764736e-05, + "loss": 2.4646, + "step": 30422 + }, + { + "epoch": 1.3, + "grad_norm": 0.388671875, + "learning_rate": 5.531422943771074e-05, + "loss": 2.4802, + "step": 30423 + }, + { + "epoch": 1.3, + "grad_norm": 0.4296875, + "learning_rate": 5.530822858884377e-05, + "loss": 2.4429, + "step": 30424 + }, + { + "epoch": 1.3, + "grad_norm": 0.5, + "learning_rate": 5.530222794107347e-05, + "loss": 2.4581, + "step": 30425 + }, + { + "epoch": 1.3, + "grad_norm": 0.49609375, + "learning_rate": 5.52962274944268e-05, + "loss": 2.4399, + "step": 30426 + }, + { + "epoch": 1.3, + "grad_norm": 0.4765625, + "learning_rate": 5.529022724893074e-05, + "loss": 2.5557, + "step": 30427 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.52842272046123e-05, + "loss": 2.5055, + "step": 30428 + }, + { + "epoch": 1.3, + "grad_norm": 0.46484375, + "learning_rate": 5.527822736149857e-05, + "loss": 2.5384, + "step": 30429 + }, + { + "epoch": 1.3, + "grad_norm": 0.4453125, + "learning_rate": 5.52722277196164e-05, + "loss": 2.4182, + "step": 30430 + }, + { + "epoch": 1.3, + "grad_norm": 0.458984375, + "learning_rate": 5.5266228278992927e-05, + "loss": 2.4525, + "step": 30431 + }, + { + "epoch": 1.3, + "grad_norm": 0.423828125, + "learning_rate": 5.526022903965502e-05, + "loss": 2.6446, + "step": 30432 + }, + { + "epoch": 1.3, + "grad_norm": 0.470703125, + "learning_rate": 5.5254230001629724e-05, + "loss": 2.5271, + "step": 30433 + }, + { + "epoch": 1.3, + "grad_norm": 0.498046875, + "learning_rate": 5.52482311649441e-05, + "loss": 2.4622, + "step": 30434 + }, + { + "epoch": 1.3, + "grad_norm": 0.482421875, + "learning_rate": 5.524223252962507e-05, + "loss": 2.4906, + "step": 30435 + }, + { + "epoch": 1.3, + "grad_norm": 0.46875, + "learning_rate": 5.52362340956996e-05, + "loss": 2.4552, + "step": 30436 + }, + { + "epoch": 1.3, + "grad_norm": 0.43359375, + "learning_rate": 5.523023586319471e-05, + "loss": 2.5099, + "step": 30437 + }, + { + "epoch": 1.3, + "grad_norm": 0.380859375, + "learning_rate": 5.522423783213746e-05, + "loss": 2.4216, + "step": 30438 + }, + { + "epoch": 1.3, + "grad_norm": 0.4609375, + "learning_rate": 5.521824000255471e-05, + "loss": 2.243, + "step": 30439 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.5212242374473565e-05, + "loss": 2.3518, + "step": 30440 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.520624494792093e-05, + "loss": 2.5528, + "step": 30441 + }, + { + "epoch": 1.3, + "grad_norm": 0.443359375, + "learning_rate": 5.5200247722923846e-05, + "loss": 2.3445, + "step": 30442 + }, + { + "epoch": 1.3, + "grad_norm": 0.423828125, + "learning_rate": 5.519425069950924e-05, + "loss": 2.4478, + "step": 30443 + }, + { + "epoch": 1.3, + "grad_norm": 0.435546875, + "learning_rate": 5.5188253877704166e-05, + "loss": 2.5277, + "step": 30444 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.518225725753554e-05, + "loss": 2.4951, + "step": 30445 + }, + { + "epoch": 1.3, + "grad_norm": 0.49609375, + "learning_rate": 5.517626083903038e-05, + "loss": 2.5396, + "step": 30446 + }, + { + "epoch": 1.3, + "grad_norm": 0.48828125, + "learning_rate": 5.51702646222157e-05, + "loss": 2.4398, + "step": 30447 + }, + { + "epoch": 1.3, + "grad_norm": 0.47265625, + "learning_rate": 5.516426860711845e-05, + "loss": 2.4468, + "step": 30448 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.515827279376553e-05, + "loss": 2.4505, + "step": 30449 + }, + { + "epoch": 1.3, + "grad_norm": 0.44140625, + "learning_rate": 5.515227718218406e-05, + "loss": 2.4418, + "step": 30450 + }, + { + "epoch": 1.3, + "grad_norm": 0.53125, + "learning_rate": 5.514628177240097e-05, + "loss": 2.5401, + "step": 30451 + }, + { + "epoch": 1.3, + "grad_norm": 0.5078125, + "learning_rate": 5.5140286564443165e-05, + "loss": 2.4966, + "step": 30452 + }, + { + "epoch": 1.3, + "grad_norm": 0.455078125, + "learning_rate": 5.513429155833768e-05, + "loss": 2.3536, + "step": 30453 + }, + { + "epoch": 1.3, + "grad_norm": 0.453125, + "learning_rate": 5.512829675411153e-05, + "loss": 2.559, + "step": 30454 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.51223021517916e-05, + "loss": 2.4848, + "step": 30455 + }, + { + "epoch": 1.3, + "grad_norm": 0.439453125, + "learning_rate": 5.511630775140495e-05, + "loss": 2.4956, + "step": 30456 + }, + { + "epoch": 1.3, + "grad_norm": 0.466796875, + "learning_rate": 5.511031355297849e-05, + "loss": 2.588, + "step": 30457 + }, + { + "epoch": 1.3, + "grad_norm": 0.5234375, + "learning_rate": 5.510431955653923e-05, + "loss": 2.5276, + "step": 30458 + }, + { + "epoch": 1.3, + "grad_norm": 0.447265625, + "learning_rate": 5.509832576211409e-05, + "loss": 2.48, + "step": 30459 + }, + { + "epoch": 1.3, + "grad_norm": 0.46875, + "learning_rate": 5.509233216973013e-05, + "loss": 2.4513, + "step": 30460 + }, + { + "epoch": 1.3, + "grad_norm": 0.4453125, + "learning_rate": 5.5086338779414204e-05, + "loss": 2.5954, + "step": 30461 + }, + { + "epoch": 1.3, + "grad_norm": 0.412109375, + "learning_rate": 5.508034559119335e-05, + "loss": 2.4464, + "step": 30462 + }, + { + "epoch": 1.3, + "grad_norm": 0.46875, + "learning_rate": 5.507435260509456e-05, + "loss": 2.6312, + "step": 30463 + }, + { + "epoch": 1.3, + "grad_norm": 0.439453125, + "learning_rate": 5.5068359821144765e-05, + "loss": 2.3808, + "step": 30464 + }, + { + "epoch": 1.3, + "grad_norm": 0.482421875, + "learning_rate": 5.506236723937086e-05, + "loss": 2.4896, + "step": 30465 + }, + { + "epoch": 1.3, + "grad_norm": 0.494140625, + "learning_rate": 5.505637485979994e-05, + "loss": 2.5089, + "step": 30466 + }, + { + "epoch": 1.3, + "grad_norm": 0.53125, + "learning_rate": 5.5050382682458926e-05, + "loss": 2.3993, + "step": 30467 + }, + { + "epoch": 1.3, + "grad_norm": 0.494140625, + "learning_rate": 5.5044390707374714e-05, + "loss": 2.4213, + "step": 30468 + }, + { + "epoch": 1.3, + "grad_norm": 0.412109375, + "learning_rate": 5.5038398934574323e-05, + "loss": 2.3887, + "step": 30469 + }, + { + "epoch": 1.3, + "grad_norm": 0.474609375, + "learning_rate": 5.503240736408472e-05, + "loss": 2.5274, + "step": 30470 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.5026415995932835e-05, + "loss": 2.5222, + "step": 30471 + }, + { + "epoch": 1.3, + "grad_norm": 0.421875, + "learning_rate": 5.502042483014568e-05, + "loss": 2.1944, + "step": 30472 + }, + { + "epoch": 1.3, + "grad_norm": 0.53125, + "learning_rate": 5.5014433866750136e-05, + "loss": 2.4514, + "step": 30473 + }, + { + "epoch": 1.3, + "grad_norm": 0.427734375, + "learning_rate": 5.500844310577322e-05, + "loss": 2.4454, + "step": 30474 + }, + { + "epoch": 1.3, + "grad_norm": 0.478515625, + "learning_rate": 5.5002452547241854e-05, + "loss": 2.4597, + "step": 30475 + }, + { + "epoch": 1.3, + "grad_norm": 0.443359375, + "learning_rate": 5.499646219118303e-05, + "loss": 2.3591, + "step": 30476 + }, + { + "epoch": 1.3, + "grad_norm": 0.466796875, + "learning_rate": 5.499047203762364e-05, + "loss": 2.4436, + "step": 30477 + }, + { + "epoch": 1.3, + "grad_norm": 0.447265625, + "learning_rate": 5.498448208659068e-05, + "loss": 2.4603, + "step": 30478 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.4978492338111145e-05, + "loss": 2.4342, + "step": 30479 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.497250279221193e-05, + "loss": 2.3637, + "step": 30480 + }, + { + "epoch": 1.3, + "grad_norm": 0.5078125, + "learning_rate": 5.496651344891992e-05, + "loss": 2.5176, + "step": 30481 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.496052430826223e-05, + "loss": 2.5467, + "step": 30482 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.4954535370265714e-05, + "loss": 2.3615, + "step": 30483 + }, + { + "epoch": 1.3, + "grad_norm": 0.45703125, + "learning_rate": 5.4948546634957287e-05, + "loss": 2.5463, + "step": 30484 + }, + { + "epoch": 1.3, + "grad_norm": 0.4375, + "learning_rate": 5.494255810236394e-05, + "loss": 2.5452, + "step": 30485 + }, + { + "epoch": 1.3, + "grad_norm": 0.490234375, + "learning_rate": 5.493656977251265e-05, + "loss": 2.4565, + "step": 30486 + }, + { + "epoch": 1.3, + "grad_norm": 0.458984375, + "learning_rate": 5.4930581645430277e-05, + "loss": 2.3968, + "step": 30487 + }, + { + "epoch": 1.3, + "grad_norm": 0.431640625, + "learning_rate": 5.492459372114387e-05, + "loss": 2.4523, + "step": 30488 + }, + { + "epoch": 1.3, + "grad_norm": 0.39453125, + "learning_rate": 5.4918605999680285e-05, + "loss": 2.424, + "step": 30489 + }, + { + "epoch": 1.3, + "grad_norm": 0.458984375, + "learning_rate": 5.4912618481066524e-05, + "loss": 2.4752, + "step": 30490 + }, + { + "epoch": 1.3, + "grad_norm": 0.423828125, + "learning_rate": 5.490663116532946e-05, + "loss": 2.419, + "step": 30491 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.490064405249613e-05, + "loss": 2.4261, + "step": 30492 + }, + { + "epoch": 1.3, + "grad_norm": 0.435546875, + "learning_rate": 5.4894657142593366e-05, + "loss": 2.5111, + "step": 30493 + }, + { + "epoch": 1.3, + "grad_norm": 0.4609375, + "learning_rate": 5.4888670435648174e-05, + "loss": 2.395, + "step": 30494 + }, + { + "epoch": 1.3, + "grad_norm": 0.455078125, + "learning_rate": 5.488268393168751e-05, + "loss": 2.4627, + "step": 30495 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.487669763073828e-05, + "loss": 2.4994, + "step": 30496 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.4870711532827336e-05, + "loss": 2.4302, + "step": 30497 + }, + { + "epoch": 1.3, + "grad_norm": 0.5234375, + "learning_rate": 5.4864725637981785e-05, + "loss": 2.4277, + "step": 30498 + }, + { + "epoch": 1.3, + "grad_norm": 0.443359375, + "learning_rate": 5.4858739946228465e-05, + "loss": 2.2889, + "step": 30499 + }, + { + "epoch": 1.3, + "grad_norm": 0.490234375, + "learning_rate": 5.485275445759428e-05, + "loss": 2.4281, + "step": 30500 + }, + { + "epoch": 1.3, + "grad_norm": 0.458984375, + "learning_rate": 5.4846769172106205e-05, + "loss": 2.5232, + "step": 30501 + }, + { + "epoch": 1.3, + "grad_norm": 0.470703125, + "learning_rate": 5.4840784089791206e-05, + "loss": 2.3786, + "step": 30502 + }, + { + "epoch": 1.3, + "grad_norm": 0.439453125, + "learning_rate": 5.483479921067614e-05, + "loss": 2.4295, + "step": 30503 + }, + { + "epoch": 1.3, + "grad_norm": 0.447265625, + "learning_rate": 5.482881453478801e-05, + "loss": 2.4783, + "step": 30504 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.482283006215366e-05, + "loss": 2.531, + "step": 30505 + }, + { + "epoch": 1.3, + "grad_norm": 0.458984375, + "learning_rate": 5.481684579280011e-05, + "loss": 2.4611, + "step": 30506 + }, + { + "epoch": 1.3, + "grad_norm": 0.49609375, + "learning_rate": 5.48108617267542e-05, + "loss": 2.5572, + "step": 30507 + }, + { + "epoch": 1.3, + "grad_norm": 0.5234375, + "learning_rate": 5.480487786404294e-05, + "loss": 2.6363, + "step": 30508 + }, + { + "epoch": 1.3, + "grad_norm": 0.400390625, + "learning_rate": 5.479889420469316e-05, + "loss": 2.2976, + "step": 30509 + }, + { + "epoch": 1.3, + "grad_norm": 0.515625, + "learning_rate": 5.479291074873185e-05, + "loss": 2.7256, + "step": 30510 + }, + { + "epoch": 1.3, + "grad_norm": 0.44140625, + "learning_rate": 5.478692749618595e-05, + "loss": 2.367, + "step": 30511 + }, + { + "epoch": 1.3, + "grad_norm": 0.455078125, + "learning_rate": 5.478094444708236e-05, + "loss": 2.3594, + "step": 30512 + }, + { + "epoch": 1.3, + "grad_norm": 0.435546875, + "learning_rate": 5.477496160144792e-05, + "loss": 2.597, + "step": 30513 + }, + { + "epoch": 1.3, + "grad_norm": 0.451171875, + "learning_rate": 5.476897895930969e-05, + "loss": 2.5369, + "step": 30514 + }, + { + "epoch": 1.3, + "grad_norm": 0.4453125, + "learning_rate": 5.4762996520694544e-05, + "loss": 2.318, + "step": 30515 + }, + { + "epoch": 1.3, + "grad_norm": 0.51171875, + "learning_rate": 5.475701428562932e-05, + "loss": 2.5909, + "step": 30516 + }, + { + "epoch": 1.3, + "grad_norm": 0.435546875, + "learning_rate": 5.475103225414101e-05, + "loss": 2.4171, + "step": 30517 + }, + { + "epoch": 1.3, + "grad_norm": 0.453125, + "learning_rate": 5.474505042625655e-05, + "loss": 2.5163, + "step": 30518 + }, + { + "epoch": 1.3, + "grad_norm": 0.51171875, + "learning_rate": 5.473906880200279e-05, + "loss": 2.4381, + "step": 30519 + }, + { + "epoch": 1.3, + "grad_norm": 0.46484375, + "learning_rate": 5.473308738140672e-05, + "loss": 2.5103, + "step": 30520 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.472710616449518e-05, + "loss": 2.4409, + "step": 30521 + }, + { + "epoch": 1.3, + "grad_norm": 0.4921875, + "learning_rate": 5.472112515129515e-05, + "loss": 2.4591, + "step": 30522 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.471514434183347e-05, + "loss": 2.5458, + "step": 30523 + }, + { + "epoch": 1.3, + "grad_norm": 0.482421875, + "learning_rate": 5.4709163736137124e-05, + "loss": 2.5699, + "step": 30524 + }, + { + "epoch": 1.3, + "grad_norm": 0.51953125, + "learning_rate": 5.470318333423297e-05, + "loss": 2.3579, + "step": 30525 + }, + { + "epoch": 1.3, + "grad_norm": 0.52734375, + "learning_rate": 5.469720313614791e-05, + "loss": 2.3999, + "step": 30526 + }, + { + "epoch": 1.3, + "grad_norm": 0.466796875, + "learning_rate": 5.469122314190893e-05, + "loss": 2.5701, + "step": 30527 + }, + { + "epoch": 1.3, + "grad_norm": 0.455078125, + "learning_rate": 5.468524335154289e-05, + "loss": 2.3315, + "step": 30528 + }, + { + "epoch": 1.3, + "grad_norm": 0.443359375, + "learning_rate": 5.4679263765076617e-05, + "loss": 2.3873, + "step": 30529 + }, + { + "epoch": 1.3, + "grad_norm": 0.458984375, + "learning_rate": 5.467328438253717e-05, + "loss": 2.3402, + "step": 30530 + }, + { + "epoch": 1.3, + "grad_norm": 0.43359375, + "learning_rate": 5.466730520395138e-05, + "loss": 2.5038, + "step": 30531 + }, + { + "epoch": 1.3, + "grad_norm": 0.44140625, + "learning_rate": 5.46613262293461e-05, + "loss": 2.4688, + "step": 30532 + }, + { + "epoch": 1.3, + "grad_norm": 0.52734375, + "learning_rate": 5.4655347458748294e-05, + "loss": 2.367, + "step": 30533 + }, + { + "epoch": 1.3, + "grad_norm": 0.41796875, + "learning_rate": 5.464936889218489e-05, + "loss": 2.3646, + "step": 30534 + }, + { + "epoch": 1.3, + "grad_norm": 0.46875, + "learning_rate": 5.4643390529682714e-05, + "loss": 2.5344, + "step": 30535 + }, + { + "epoch": 1.3, + "grad_norm": 0.458984375, + "learning_rate": 5.4637412371268736e-05, + "loss": 2.4385, + "step": 30536 + }, + { + "epoch": 1.3, + "grad_norm": 0.447265625, + "learning_rate": 5.463143441696981e-05, + "loss": 2.4537, + "step": 30537 + }, + { + "epoch": 1.3, + "grad_norm": 0.4609375, + "learning_rate": 5.462545666681286e-05, + "loss": 2.4224, + "step": 30538 + }, + { + "epoch": 1.3, + "grad_norm": 0.44140625, + "learning_rate": 5.461947912082475e-05, + "loss": 2.5568, + "step": 30539 + }, + { + "epoch": 1.3, + "grad_norm": 0.4453125, + "learning_rate": 5.4613501779032415e-05, + "loss": 2.4937, + "step": 30540 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.460752464146273e-05, + "loss": 2.4652, + "step": 30541 + }, + { + "epoch": 1.3, + "grad_norm": 0.46484375, + "learning_rate": 5.460154770814257e-05, + "loss": 2.4169, + "step": 30542 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.4595570979098896e-05, + "loss": 2.3721, + "step": 30543 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.458959445435856e-05, + "loss": 2.4439, + "step": 30544 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.458361813394839e-05, + "loss": 2.5021, + "step": 30545 + }, + { + "epoch": 1.3, + "grad_norm": 0.494140625, + "learning_rate": 5.457764201789539e-05, + "loss": 2.4783, + "step": 30546 + }, + { + "epoch": 1.3, + "grad_norm": 0.45703125, + "learning_rate": 5.457166610622643e-05, + "loss": 2.5107, + "step": 30547 + }, + { + "epoch": 1.3, + "grad_norm": 0.4453125, + "learning_rate": 5.456569039896832e-05, + "loss": 2.3331, + "step": 30548 + }, + { + "epoch": 1.3, + "grad_norm": 0.431640625, + "learning_rate": 5.4559714896147994e-05, + "loss": 2.4372, + "step": 30549 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.45537395977924e-05, + "loss": 2.3786, + "step": 30550 + }, + { + "epoch": 1.3, + "grad_norm": 0.48828125, + "learning_rate": 5.4547764503928314e-05, + "loss": 2.4003, + "step": 30551 + }, + { + "epoch": 1.3, + "grad_norm": 0.53515625, + "learning_rate": 5.4541789614582715e-05, + "loss": 2.5249, + "step": 30552 + }, + { + "epoch": 1.3, + "grad_norm": 0.46484375, + "learning_rate": 5.4535814929782414e-05, + "loss": 2.5296, + "step": 30553 + }, + { + "epoch": 1.3, + "grad_norm": 0.51171875, + "learning_rate": 5.452984044955437e-05, + "loss": 2.5738, + "step": 30554 + }, + { + "epoch": 1.3, + "grad_norm": 0.4609375, + "learning_rate": 5.45238661739254e-05, + "loss": 2.4842, + "step": 30555 + }, + { + "epoch": 1.3, + "grad_norm": 0.48046875, + "learning_rate": 5.451789210292243e-05, + "loss": 2.4493, + "step": 30556 + }, + { + "epoch": 1.3, + "grad_norm": 0.482421875, + "learning_rate": 5.451191823657229e-05, + "loss": 2.3941, + "step": 30557 + }, + { + "epoch": 1.3, + "grad_norm": 0.482421875, + "learning_rate": 5.4505944574901905e-05, + "loss": 2.4893, + "step": 30558 + }, + { + "epoch": 1.3, + "grad_norm": 0.494140625, + "learning_rate": 5.4499971117938175e-05, + "loss": 2.4971, + "step": 30559 + }, + { + "epoch": 1.3, + "grad_norm": 0.470703125, + "learning_rate": 5.4493997865707946e-05, + "loss": 2.4743, + "step": 30560 + }, + { + "epoch": 1.3, + "grad_norm": 0.48046875, + "learning_rate": 5.448802481823806e-05, + "loss": 2.5407, + "step": 30561 + }, + { + "epoch": 1.3, + "grad_norm": 0.49609375, + "learning_rate": 5.448205197555543e-05, + "loss": 2.5208, + "step": 30562 + }, + { + "epoch": 1.3, + "grad_norm": 0.5234375, + "learning_rate": 5.447607933768697e-05, + "loss": 2.4343, + "step": 30563 + }, + { + "epoch": 1.3, + "grad_norm": 0.447265625, + "learning_rate": 5.447010690465948e-05, + "loss": 2.4149, + "step": 30564 + }, + { + "epoch": 1.3, + "grad_norm": 0.5078125, + "learning_rate": 5.446413467649989e-05, + "loss": 2.5831, + "step": 30565 + }, + { + "epoch": 1.3, + "grad_norm": 0.494140625, + "learning_rate": 5.445816265323504e-05, + "loss": 2.457, + "step": 30566 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.445219083489179e-05, + "loss": 2.3895, + "step": 30567 + }, + { + "epoch": 1.3, + "grad_norm": 0.482421875, + "learning_rate": 5.444621922149709e-05, + "loss": 2.4255, + "step": 30568 + }, + { + "epoch": 1.3, + "grad_norm": 0.494140625, + "learning_rate": 5.444024781307775e-05, + "loss": 2.583, + "step": 30569 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.4434276609660606e-05, + "loss": 2.3513, + "step": 30570 + }, + { + "epoch": 1.3, + "grad_norm": 0.47265625, + "learning_rate": 5.442830561127257e-05, + "loss": 2.5197, + "step": 30571 + }, + { + "epoch": 1.3, + "grad_norm": 0.466796875, + "learning_rate": 5.4422334817940546e-05, + "loss": 2.3866, + "step": 30572 + }, + { + "epoch": 1.3, + "grad_norm": 0.451171875, + "learning_rate": 5.441636422969131e-05, + "loss": 2.4712, + "step": 30573 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.4410393846551776e-05, + "loss": 2.3366, + "step": 30574 + }, + { + "epoch": 1.3, + "grad_norm": 0.453125, + "learning_rate": 5.4404423668548855e-05, + "loss": 2.2774, + "step": 30575 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.439845369570936e-05, + "loss": 2.4881, + "step": 30576 + }, + { + "epoch": 1.3, + "grad_norm": 0.451171875, + "learning_rate": 5.439248392806011e-05, + "loss": 2.3296, + "step": 30577 + }, + { + "epoch": 1.3, + "grad_norm": 0.4375, + "learning_rate": 5.438651436562803e-05, + "loss": 2.3284, + "step": 30578 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.438054500844e-05, + "loss": 2.4141, + "step": 30579 + }, + { + "epoch": 1.3, + "grad_norm": 0.47265625, + "learning_rate": 5.4374575856522804e-05, + "loss": 2.512, + "step": 30580 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.436860690990339e-05, + "loss": 2.5402, + "step": 30581 + }, + { + "epoch": 1.3, + "grad_norm": 0.51171875, + "learning_rate": 5.436263816860851e-05, + "loss": 2.4451, + "step": 30582 + }, + { + "epoch": 1.3, + "grad_norm": 0.453125, + "learning_rate": 5.43566696326651e-05, + "loss": 2.3425, + "step": 30583 + }, + { + "epoch": 1.3, + "grad_norm": 0.49609375, + "learning_rate": 5.4350701302100024e-05, + "loss": 2.5202, + "step": 30584 + }, + { + "epoch": 1.3, + "grad_norm": 0.490234375, + "learning_rate": 5.434473317694011e-05, + "loss": 2.4944, + "step": 30585 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.433876525721218e-05, + "loss": 2.404, + "step": 30586 + }, + { + "epoch": 1.3, + "grad_norm": 0.48046875, + "learning_rate": 5.433279754294311e-05, + "loss": 2.4549, + "step": 30587 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.43268300341598e-05, + "loss": 2.4062, + "step": 30588 + }, + { + "epoch": 1.3, + "grad_norm": 0.51953125, + "learning_rate": 5.4320862730889036e-05, + "loss": 2.5219, + "step": 30589 + }, + { + "epoch": 1.3, + "grad_norm": 0.451171875, + "learning_rate": 5.431489563315768e-05, + "loss": 2.6143, + "step": 30590 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.430892874099265e-05, + "loss": 2.4882, + "step": 30591 + }, + { + "epoch": 1.3, + "grad_norm": 0.474609375, + "learning_rate": 5.4302962054420736e-05, + "loss": 2.3868, + "step": 30592 + }, + { + "epoch": 1.3, + "grad_norm": 0.48828125, + "learning_rate": 5.429699557346875e-05, + "loss": 2.4762, + "step": 30593 + }, + { + "epoch": 1.3, + "grad_norm": 0.48046875, + "learning_rate": 5.4291029298163585e-05, + "loss": 2.4759, + "step": 30594 + }, + { + "epoch": 1.3, + "grad_norm": 0.470703125, + "learning_rate": 5.428506322853213e-05, + "loss": 2.497, + "step": 30595 + }, + { + "epoch": 1.3, + "grad_norm": 0.49609375, + "learning_rate": 5.427909736460113e-05, + "loss": 2.5556, + "step": 30596 + }, + { + "epoch": 1.3, + "grad_norm": 0.494140625, + "learning_rate": 5.427313170639754e-05, + "loss": 2.3724, + "step": 30597 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.4267166253948096e-05, + "loss": 2.7324, + "step": 30598 + }, + { + "epoch": 1.3, + "grad_norm": 0.4375, + "learning_rate": 5.42612010072797e-05, + "loss": 2.3102, + "step": 30599 + }, + { + "epoch": 1.3, + "grad_norm": 0.474609375, + "learning_rate": 5.4255235966419196e-05, + "loss": 2.4266, + "step": 30600 + }, + { + "epoch": 1.3, + "grad_norm": 0.400390625, + "learning_rate": 5.424927113139344e-05, + "loss": 2.4329, + "step": 30601 + }, + { + "epoch": 1.3, + "grad_norm": 0.4765625, + "learning_rate": 5.42433065022292e-05, + "loss": 2.3986, + "step": 30602 + }, + { + "epoch": 1.3, + "grad_norm": 0.4921875, + "learning_rate": 5.423734207895335e-05, + "loss": 2.4044, + "step": 30603 + }, + { + "epoch": 1.3, + "grad_norm": 0.470703125, + "learning_rate": 5.423137786159278e-05, + "loss": 2.64, + "step": 30604 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.422541385017423e-05, + "loss": 2.4335, + "step": 30605 + }, + { + "epoch": 1.3, + "grad_norm": 0.458984375, + "learning_rate": 5.421945004472461e-05, + "loss": 2.4412, + "step": 30606 + }, + { + "epoch": 1.3, + "grad_norm": 0.423828125, + "learning_rate": 5.421348644527075e-05, + "loss": 2.4233, + "step": 30607 + }, + { + "epoch": 1.3, + "grad_norm": 0.47265625, + "learning_rate": 5.4207523051839496e-05, + "loss": 2.3845, + "step": 30608 + }, + { + "epoch": 1.3, + "grad_norm": 0.515625, + "learning_rate": 5.420155986445761e-05, + "loss": 2.3641, + "step": 30609 + }, + { + "epoch": 1.3, + "grad_norm": 0.435546875, + "learning_rate": 5.4195596883151946e-05, + "loss": 2.4252, + "step": 30610 + }, + { + "epoch": 1.3, + "grad_norm": 0.453125, + "learning_rate": 5.4189634107949416e-05, + "loss": 2.4519, + "step": 30611 + }, + { + "epoch": 1.3, + "grad_norm": 0.474609375, + "learning_rate": 5.418367153887673e-05, + "loss": 2.5649, + "step": 30612 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.4177709175960835e-05, + "loss": 2.5093, + "step": 30613 + }, + { + "epoch": 1.3, + "grad_norm": 0.462890625, + "learning_rate": 5.417174701922847e-05, + "loss": 2.4309, + "step": 30614 + }, + { + "epoch": 1.3, + "grad_norm": 0.3984375, + "learning_rate": 5.416578506870649e-05, + "loss": 2.4799, + "step": 30615 + }, + { + "epoch": 1.3, + "grad_norm": 0.50390625, + "learning_rate": 5.4159823324421756e-05, + "loss": 2.304, + "step": 30616 + }, + { + "epoch": 1.3, + "grad_norm": 0.484375, + "learning_rate": 5.415386178640106e-05, + "loss": 2.4813, + "step": 30617 + }, + { + "epoch": 1.3, + "grad_norm": 0.486328125, + "learning_rate": 5.4147900454671196e-05, + "loss": 2.3667, + "step": 30618 + }, + { + "epoch": 1.3, + "grad_norm": 0.478515625, + "learning_rate": 5.414193932925904e-05, + "loss": 2.5373, + "step": 30619 + }, + { + "epoch": 1.3, + "grad_norm": 0.43359375, + "learning_rate": 5.4135978410191423e-05, + "loss": 2.4129, + "step": 30620 + }, + { + "epoch": 1.3, + "grad_norm": 0.423828125, + "learning_rate": 5.4130017697495106e-05, + "loss": 2.2701, + "step": 30621 + }, + { + "epoch": 1.3, + "grad_norm": 0.46875, + "learning_rate": 5.4124057191196934e-05, + "loss": 2.4858, + "step": 30622 + }, + { + "epoch": 1.3, + "grad_norm": 0.447265625, + "learning_rate": 5.411809689132379e-05, + "loss": 2.4763, + "step": 30623 + }, + { + "epoch": 1.3, + "grad_norm": 0.48828125, + "learning_rate": 5.411213679790243e-05, + "loss": 2.459, + "step": 30624 + }, + { + "epoch": 1.31, + "grad_norm": 0.408203125, + "learning_rate": 5.410617691095967e-05, + "loss": 2.4788, + "step": 30625 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.410021723052232e-05, + "loss": 2.5493, + "step": 30626 + }, + { + "epoch": 1.31, + "grad_norm": 0.48828125, + "learning_rate": 5.409425775661727e-05, + "loss": 2.4682, + "step": 30627 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.408829848927124e-05, + "loss": 2.4123, + "step": 30628 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.408233942851112e-05, + "loss": 2.4629, + "step": 30629 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.407638057436366e-05, + "loss": 2.4926, + "step": 30630 + }, + { + "epoch": 1.31, + "grad_norm": 0.453125, + "learning_rate": 5.40704219268557e-05, + "loss": 2.4423, + "step": 30631 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.406446348601409e-05, + "loss": 2.4174, + "step": 30632 + }, + { + "epoch": 1.31, + "grad_norm": 0.47265625, + "learning_rate": 5.4058505251865624e-05, + "loss": 2.7152, + "step": 30633 + }, + { + "epoch": 1.31, + "grad_norm": 0.51953125, + "learning_rate": 5.4052547224437e-05, + "loss": 2.457, + "step": 30634 + }, + { + "epoch": 1.31, + "grad_norm": 0.484375, + "learning_rate": 5.404658940375521e-05, + "loss": 2.3744, + "step": 30635 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.404063178984697e-05, + "loss": 2.4629, + "step": 30636 + }, + { + "epoch": 1.31, + "grad_norm": 0.439453125, + "learning_rate": 5.403467438273905e-05, + "loss": 2.4205, + "step": 30637 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.4028717182458296e-05, + "loss": 2.4568, + "step": 30638 + }, + { + "epoch": 1.31, + "grad_norm": 0.55859375, + "learning_rate": 5.402276018903156e-05, + "loss": 2.6755, + "step": 30639 + }, + { + "epoch": 1.31, + "grad_norm": 0.435546875, + "learning_rate": 5.40168034024856e-05, + "loss": 2.418, + "step": 30640 + }, + { + "epoch": 1.31, + "grad_norm": 0.4375, + "learning_rate": 5.401084682284717e-05, + "loss": 2.3661, + "step": 30641 + }, + { + "epoch": 1.31, + "grad_norm": 0.4921875, + "learning_rate": 5.400489045014313e-05, + "loss": 2.5015, + "step": 30642 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.3998934284400306e-05, + "loss": 2.4587, + "step": 30643 + }, + { + "epoch": 1.31, + "grad_norm": 0.4453125, + "learning_rate": 5.399297832564544e-05, + "loss": 2.4907, + "step": 30644 + }, + { + "epoch": 1.31, + "grad_norm": 0.439453125, + "learning_rate": 5.398702257390539e-05, + "loss": 2.3953, + "step": 30645 + }, + { + "epoch": 1.31, + "grad_norm": 0.443359375, + "learning_rate": 5.3981067029206875e-05, + "loss": 2.5456, + "step": 30646 + }, + { + "epoch": 1.31, + "grad_norm": 0.4765625, + "learning_rate": 5.3975111691576754e-05, + "loss": 2.4703, + "step": 30647 + }, + { + "epoch": 1.31, + "grad_norm": 0.462890625, + "learning_rate": 5.3969156561041845e-05, + "loss": 2.5341, + "step": 30648 + }, + { + "epoch": 1.31, + "grad_norm": 0.498046875, + "learning_rate": 5.3963201637628915e-05, + "loss": 2.5828, + "step": 30649 + }, + { + "epoch": 1.31, + "grad_norm": 0.416015625, + "learning_rate": 5.3957246921364654e-05, + "loss": 2.4557, + "step": 30650 + }, + { + "epoch": 1.31, + "grad_norm": 0.546875, + "learning_rate": 5.395129241227605e-05, + "loss": 2.5923, + "step": 30651 + }, + { + "epoch": 1.31, + "grad_norm": 0.48828125, + "learning_rate": 5.3945338110389796e-05, + "loss": 2.5027, + "step": 30652 + }, + { + "epoch": 1.31, + "grad_norm": 0.474609375, + "learning_rate": 5.3939384015732665e-05, + "loss": 2.5525, + "step": 30653 + }, + { + "epoch": 1.31, + "grad_norm": 0.4765625, + "learning_rate": 5.393343012833145e-05, + "loss": 2.5508, + "step": 30654 + }, + { + "epoch": 1.31, + "grad_norm": 0.4453125, + "learning_rate": 5.392747644821301e-05, + "loss": 2.422, + "step": 30655 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.392152297540409e-05, + "loss": 2.4577, + "step": 30656 + }, + { + "epoch": 1.31, + "grad_norm": 0.484375, + "learning_rate": 5.391556970993146e-05, + "loss": 2.5135, + "step": 30657 + }, + { + "epoch": 1.31, + "grad_norm": 0.486328125, + "learning_rate": 5.390961665182189e-05, + "loss": 2.4687, + "step": 30658 + }, + { + "epoch": 1.31, + "grad_norm": 0.51171875, + "learning_rate": 5.3903663801102256e-05, + "loss": 2.5312, + "step": 30659 + }, + { + "epoch": 1.31, + "grad_norm": 0.375, + "learning_rate": 5.3897711157799244e-05, + "loss": 2.2646, + "step": 30660 + }, + { + "epoch": 1.31, + "grad_norm": 0.52734375, + "learning_rate": 5.3891758721939737e-05, + "loss": 2.452, + "step": 30661 + }, + { + "epoch": 1.31, + "grad_norm": 0.455078125, + "learning_rate": 5.388580649355042e-05, + "loss": 2.5265, + "step": 30662 + }, + { + "epoch": 1.31, + "grad_norm": 0.474609375, + "learning_rate": 5.3879854472658106e-05, + "loss": 2.4507, + "step": 30663 + }, + { + "epoch": 1.31, + "grad_norm": 0.443359375, + "learning_rate": 5.3873902659289644e-05, + "loss": 2.509, + "step": 30664 + }, + { + "epoch": 1.31, + "grad_norm": 0.41796875, + "learning_rate": 5.386795105347175e-05, + "loss": 2.3567, + "step": 30665 + }, + { + "epoch": 1.31, + "grad_norm": 0.4140625, + "learning_rate": 5.3861999655231157e-05, + "loss": 2.4922, + "step": 30666 + }, + { + "epoch": 1.31, + "grad_norm": 0.46484375, + "learning_rate": 5.385604846459478e-05, + "loss": 2.4445, + "step": 30667 + }, + { + "epoch": 1.31, + "grad_norm": 0.47265625, + "learning_rate": 5.385009748158931e-05, + "loss": 2.4052, + "step": 30668 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.3844146706241497e-05, + "loss": 2.4277, + "step": 30669 + }, + { + "epoch": 1.31, + "grad_norm": 0.451171875, + "learning_rate": 5.3838196138578166e-05, + "loss": 2.5634, + "step": 30670 + }, + { + "epoch": 1.31, + "grad_norm": 0.43359375, + "learning_rate": 5.383224577862611e-05, + "loss": 2.4442, + "step": 30671 + }, + { + "epoch": 1.31, + "grad_norm": 0.419921875, + "learning_rate": 5.382629562641208e-05, + "loss": 2.3248, + "step": 30672 + }, + { + "epoch": 1.31, + "grad_norm": 0.49609375, + "learning_rate": 5.3820345681962814e-05, + "loss": 2.4132, + "step": 30673 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.3814395945305104e-05, + "loss": 2.4568, + "step": 30674 + }, + { + "epoch": 1.31, + "grad_norm": 0.455078125, + "learning_rate": 5.380844641646578e-05, + "loss": 2.4435, + "step": 30675 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.380249709547153e-05, + "loss": 2.4766, + "step": 30676 + }, + { + "epoch": 1.31, + "grad_norm": 0.44921875, + "learning_rate": 5.3796547982349186e-05, + "loss": 2.4094, + "step": 30677 + }, + { + "epoch": 1.31, + "grad_norm": 0.4609375, + "learning_rate": 5.379059907712549e-05, + "loss": 2.4936, + "step": 30678 + }, + { + "epoch": 1.31, + "grad_norm": 0.39453125, + "learning_rate": 5.378465037982715e-05, + "loss": 2.4812, + "step": 30679 + }, + { + "epoch": 1.31, + "grad_norm": 0.453125, + "learning_rate": 5.3778701890481065e-05, + "loss": 2.5463, + "step": 30680 + }, + { + "epoch": 1.31, + "grad_norm": 0.439453125, + "learning_rate": 5.377275360911392e-05, + "loss": 2.5069, + "step": 30681 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.376680553575246e-05, + "loss": 2.4513, + "step": 30682 + }, + { + "epoch": 1.31, + "grad_norm": 0.47265625, + "learning_rate": 5.376085767042347e-05, + "loss": 2.5325, + "step": 30683 + }, + { + "epoch": 1.31, + "grad_norm": 0.462890625, + "learning_rate": 5.375491001315378e-05, + "loss": 2.4324, + "step": 30684 + }, + { + "epoch": 1.31, + "grad_norm": 0.51171875, + "learning_rate": 5.374896256397004e-05, + "loss": 2.4714, + "step": 30685 + }, + { + "epoch": 1.31, + "grad_norm": 0.455078125, + "learning_rate": 5.37430153228991e-05, + "loss": 2.4495, + "step": 30686 + }, + { + "epoch": 1.31, + "grad_norm": 0.4609375, + "learning_rate": 5.373706828996765e-05, + "loss": 2.3769, + "step": 30687 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.37311214652025e-05, + "loss": 2.5475, + "step": 30688 + }, + { + "epoch": 1.31, + "grad_norm": 0.52734375, + "learning_rate": 5.3725174848630425e-05, + "loss": 2.4307, + "step": 30689 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.3719228440278147e-05, + "loss": 2.5784, + "step": 30690 + }, + { + "epoch": 1.31, + "grad_norm": 0.486328125, + "learning_rate": 5.371328224017238e-05, + "loss": 2.5216, + "step": 30691 + }, + { + "epoch": 1.31, + "grad_norm": 0.51171875, + "learning_rate": 5.370733624833995e-05, + "loss": 2.5424, + "step": 30692 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.370139046480761e-05, + "loss": 2.4608, + "step": 30693 + }, + { + "epoch": 1.31, + "grad_norm": 0.431640625, + "learning_rate": 5.36954448896021e-05, + "loss": 2.2453, + "step": 30694 + }, + { + "epoch": 1.31, + "grad_norm": 0.515625, + "learning_rate": 5.368949952275009e-05, + "loss": 2.4371, + "step": 30695 + }, + { + "epoch": 1.31, + "grad_norm": 0.451171875, + "learning_rate": 5.3683554364278475e-05, + "loss": 2.4686, + "step": 30696 + }, + { + "epoch": 1.31, + "grad_norm": 0.484375, + "learning_rate": 5.3677609414213945e-05, + "loss": 2.4562, + "step": 30697 + }, + { + "epoch": 1.31, + "grad_norm": 0.466796875, + "learning_rate": 5.36716646725832e-05, + "loss": 2.4031, + "step": 30698 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.366572013941304e-05, + "loss": 2.6001, + "step": 30699 + }, + { + "epoch": 1.31, + "grad_norm": 0.431640625, + "learning_rate": 5.365977581473025e-05, + "loss": 2.3618, + "step": 30700 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.3653831698561485e-05, + "loss": 2.4009, + "step": 30701 + }, + { + "epoch": 1.31, + "grad_norm": 0.484375, + "learning_rate": 5.364788779093357e-05, + "loss": 2.5561, + "step": 30702 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.36419440918732e-05, + "loss": 2.5617, + "step": 30703 + }, + { + "epoch": 1.31, + "grad_norm": 0.45703125, + "learning_rate": 5.3636000601407134e-05, + "loss": 2.4887, + "step": 30704 + }, + { + "epoch": 1.31, + "grad_norm": 0.5078125, + "learning_rate": 5.3630057319562155e-05, + "loss": 2.4452, + "step": 30705 + }, + { + "epoch": 1.31, + "grad_norm": 0.48046875, + "learning_rate": 5.362411424636497e-05, + "loss": 2.576, + "step": 30706 + }, + { + "epoch": 1.31, + "grad_norm": 0.41796875, + "learning_rate": 5.3618171381842285e-05, + "loss": 2.3358, + "step": 30707 + }, + { + "epoch": 1.31, + "grad_norm": 0.421875, + "learning_rate": 5.36122287260209e-05, + "loss": 2.486, + "step": 30708 + }, + { + "epoch": 1.31, + "grad_norm": 0.45703125, + "learning_rate": 5.360628627892754e-05, + "loss": 2.5066, + "step": 30709 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.3600344040588954e-05, + "loss": 2.4372, + "step": 30710 + }, + { + "epoch": 1.31, + "grad_norm": 0.462890625, + "learning_rate": 5.3594402011031786e-05, + "loss": 2.3911, + "step": 30711 + }, + { + "epoch": 1.31, + "grad_norm": 0.4609375, + "learning_rate": 5.358846019028292e-05, + "loss": 2.4959, + "step": 30712 + }, + { + "epoch": 1.31, + "grad_norm": 0.54296875, + "learning_rate": 5.3582518578369024e-05, + "loss": 2.5988, + "step": 30713 + }, + { + "epoch": 1.31, + "grad_norm": 0.490234375, + "learning_rate": 5.357657717531679e-05, + "loss": 2.3893, + "step": 30714 + }, + { + "epoch": 1.31, + "grad_norm": 0.474609375, + "learning_rate": 5.357063598115299e-05, + "loss": 2.3726, + "step": 30715 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.3564694995904405e-05, + "loss": 2.5133, + "step": 30716 + }, + { + "epoch": 1.31, + "grad_norm": 0.51953125, + "learning_rate": 5.355875421959768e-05, + "loss": 2.5031, + "step": 30717 + }, + { + "epoch": 1.31, + "grad_norm": 0.515625, + "learning_rate": 5.3552813652259634e-05, + "loss": 2.5712, + "step": 30718 + }, + { + "epoch": 1.31, + "grad_norm": 0.46484375, + "learning_rate": 5.354687329391691e-05, + "loss": 2.5822, + "step": 30719 + }, + { + "epoch": 1.31, + "grad_norm": 0.52734375, + "learning_rate": 5.3540933144596274e-05, + "loss": 2.5802, + "step": 30720 + }, + { + "epoch": 1.31, + "grad_norm": 0.5078125, + "learning_rate": 5.3534993204324515e-05, + "loss": 2.4076, + "step": 30721 + }, + { + "epoch": 1.31, + "grad_norm": 0.51171875, + "learning_rate": 5.3529053473128285e-05, + "loss": 2.4715, + "step": 30722 + }, + { + "epoch": 1.31, + "grad_norm": 0.44140625, + "learning_rate": 5.352311395103431e-05, + "loss": 2.3905, + "step": 30723 + }, + { + "epoch": 1.31, + "grad_norm": 0.4609375, + "learning_rate": 5.3517174638069315e-05, + "loss": 2.491, + "step": 30724 + }, + { + "epoch": 1.31, + "grad_norm": 0.50390625, + "learning_rate": 5.3511235534260095e-05, + "loss": 2.5021, + "step": 30725 + }, + { + "epoch": 1.31, + "grad_norm": 0.490234375, + "learning_rate": 5.350529663963334e-05, + "loss": 2.4315, + "step": 30726 + }, + { + "epoch": 1.31, + "grad_norm": 0.498046875, + "learning_rate": 5.349935795421567e-05, + "loss": 2.5283, + "step": 30727 + }, + { + "epoch": 1.31, + "grad_norm": 0.4609375, + "learning_rate": 5.349341947803397e-05, + "loss": 2.4087, + "step": 30728 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.3487481211114885e-05, + "loss": 2.5418, + "step": 30729 + }, + { + "epoch": 1.31, + "grad_norm": 0.5078125, + "learning_rate": 5.3481543153485105e-05, + "loss": 2.5106, + "step": 30730 + }, + { + "epoch": 1.31, + "grad_norm": 0.419921875, + "learning_rate": 5.3475605305171375e-05, + "loss": 2.4942, + "step": 30731 + }, + { + "epoch": 1.31, + "grad_norm": 0.4921875, + "learning_rate": 5.346966766620045e-05, + "loss": 2.4115, + "step": 30732 + }, + { + "epoch": 1.31, + "grad_norm": 0.4375, + "learning_rate": 5.3463730236598986e-05, + "loss": 2.3781, + "step": 30733 + }, + { + "epoch": 1.31, + "grad_norm": 0.482421875, + "learning_rate": 5.3457793016393776e-05, + "loss": 2.5457, + "step": 30734 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.3451856005611446e-05, + "loss": 2.4676, + "step": 30735 + }, + { + "epoch": 1.31, + "grad_norm": 0.48828125, + "learning_rate": 5.344591920427874e-05, + "loss": 2.5052, + "step": 30736 + }, + { + "epoch": 1.31, + "grad_norm": 0.423828125, + "learning_rate": 5.3439982612422426e-05, + "loss": 2.4434, + "step": 30737 + }, + { + "epoch": 1.31, + "grad_norm": 0.458984375, + "learning_rate": 5.343404623006918e-05, + "loss": 2.443, + "step": 30738 + }, + { + "epoch": 1.31, + "grad_norm": 0.458984375, + "learning_rate": 5.342811005724566e-05, + "loss": 2.4602, + "step": 30739 + }, + { + "epoch": 1.31, + "grad_norm": 0.51953125, + "learning_rate": 5.342217409397864e-05, + "loss": 2.4865, + "step": 30740 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.3416238340294845e-05, + "loss": 2.3212, + "step": 30741 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.341030279622091e-05, + "loss": 2.4493, + "step": 30742 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.340436746178359e-05, + "loss": 2.4971, + "step": 30743 + }, + { + "epoch": 1.31, + "grad_norm": 0.48046875, + "learning_rate": 5.339843233700963e-05, + "loss": 2.5306, + "step": 30744 + }, + { + "epoch": 1.31, + "grad_norm": 0.4296875, + "learning_rate": 5.339249742192569e-05, + "loss": 2.4954, + "step": 30745 + }, + { + "epoch": 1.31, + "grad_norm": 0.578125, + "learning_rate": 5.338656271655844e-05, + "loss": 2.3849, + "step": 30746 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.3380628220934624e-05, + "loss": 2.4351, + "step": 30747 + }, + { + "epoch": 1.31, + "grad_norm": 0.48828125, + "learning_rate": 5.3374693935080975e-05, + "loss": 2.4039, + "step": 30748 + }, + { + "epoch": 1.31, + "grad_norm": 0.4921875, + "learning_rate": 5.336875985902414e-05, + "loss": 2.5543, + "step": 30749 + }, + { + "epoch": 1.31, + "grad_norm": 0.455078125, + "learning_rate": 5.336282599279088e-05, + "loss": 2.4559, + "step": 30750 + }, + { + "epoch": 1.31, + "grad_norm": 0.51953125, + "learning_rate": 5.3356892336407804e-05, + "loss": 2.5408, + "step": 30751 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.335095888990167e-05, + "loss": 2.3984, + "step": 30752 + }, + { + "epoch": 1.31, + "grad_norm": 0.482421875, + "learning_rate": 5.334502565329922e-05, + "loss": 2.4816, + "step": 30753 + }, + { + "epoch": 1.31, + "grad_norm": 0.41796875, + "learning_rate": 5.3339092626627104e-05, + "loss": 2.2745, + "step": 30754 + }, + { + "epoch": 1.31, + "grad_norm": 0.42578125, + "learning_rate": 5.333315980991198e-05, + "loss": 2.3652, + "step": 30755 + }, + { + "epoch": 1.31, + "grad_norm": 0.474609375, + "learning_rate": 5.332722720318057e-05, + "loss": 2.4746, + "step": 30756 + }, + { + "epoch": 1.31, + "grad_norm": 0.44921875, + "learning_rate": 5.332129480645964e-05, + "loss": 2.4061, + "step": 30757 + }, + { + "epoch": 1.31, + "grad_norm": 0.515625, + "learning_rate": 5.3315362619775767e-05, + "loss": 2.4701, + "step": 30758 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.330943064315571e-05, + "loss": 2.4783, + "step": 30759 + }, + { + "epoch": 1.31, + "grad_norm": 0.47265625, + "learning_rate": 5.330349887662619e-05, + "loss": 2.5541, + "step": 30760 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.329756732021386e-05, + "loss": 2.5184, + "step": 30761 + }, + { + "epoch": 1.31, + "grad_norm": 0.484375, + "learning_rate": 5.329163597394536e-05, + "loss": 2.4305, + "step": 30762 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.328570483784745e-05, + "loss": 2.4412, + "step": 30763 + }, + { + "epoch": 1.31, + "grad_norm": 0.47265625, + "learning_rate": 5.327977391194682e-05, + "loss": 2.4324, + "step": 30764 + }, + { + "epoch": 1.31, + "grad_norm": 0.54296875, + "learning_rate": 5.327384319627009e-05, + "loss": 2.5055, + "step": 30765 + }, + { + "epoch": 1.31, + "grad_norm": 0.490234375, + "learning_rate": 5.3267912690844034e-05, + "loss": 2.6052, + "step": 30766 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.326198239569525e-05, + "loss": 2.4395, + "step": 30767 + }, + { + "epoch": 1.31, + "grad_norm": 0.462890625, + "learning_rate": 5.325605231085047e-05, + "loss": 2.6121, + "step": 30768 + }, + { + "epoch": 1.31, + "grad_norm": 0.482421875, + "learning_rate": 5.325012243633641e-05, + "loss": 2.4794, + "step": 30769 + }, + { + "epoch": 1.31, + "grad_norm": 0.443359375, + "learning_rate": 5.324419277217971e-05, + "loss": 2.5739, + "step": 30770 + }, + { + "epoch": 1.31, + "grad_norm": 0.4296875, + "learning_rate": 5.3238263318407024e-05, + "loss": 2.4435, + "step": 30771 + }, + { + "epoch": 1.31, + "grad_norm": 0.51953125, + "learning_rate": 5.3232334075045054e-05, + "loss": 2.4251, + "step": 30772 + }, + { + "epoch": 1.31, + "grad_norm": 0.54296875, + "learning_rate": 5.322640504212053e-05, + "loss": 2.595, + "step": 30773 + }, + { + "epoch": 1.31, + "grad_norm": 0.53125, + "learning_rate": 5.322047621966006e-05, + "loss": 2.4529, + "step": 30774 + }, + { + "epoch": 1.31, + "grad_norm": 0.478515625, + "learning_rate": 5.3214547607690336e-05, + "loss": 2.4207, + "step": 30775 + }, + { + "epoch": 1.31, + "grad_norm": 0.494140625, + "learning_rate": 5.32086192062381e-05, + "loss": 2.4281, + "step": 30776 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.3202691015329975e-05, + "loss": 2.3491, + "step": 30777 + }, + { + "epoch": 1.31, + "grad_norm": 0.4140625, + "learning_rate": 5.319676303499259e-05, + "loss": 2.3999, + "step": 30778 + }, + { + "epoch": 1.31, + "grad_norm": 0.44140625, + "learning_rate": 5.319083526525268e-05, + "loss": 2.4598, + "step": 30779 + }, + { + "epoch": 1.31, + "grad_norm": 0.51171875, + "learning_rate": 5.318490770613693e-05, + "loss": 2.438, + "step": 30780 + }, + { + "epoch": 1.31, + "grad_norm": 0.435546875, + "learning_rate": 5.317898035767197e-05, + "loss": 2.313, + "step": 30781 + }, + { + "epoch": 1.31, + "grad_norm": 0.44140625, + "learning_rate": 5.317305321988451e-05, + "loss": 2.4837, + "step": 30782 + }, + { + "epoch": 1.31, + "grad_norm": 0.498046875, + "learning_rate": 5.316712629280116e-05, + "loss": 2.3063, + "step": 30783 + }, + { + "epoch": 1.31, + "grad_norm": 0.51171875, + "learning_rate": 5.316119957644863e-05, + "loss": 2.3902, + "step": 30784 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.315527307085362e-05, + "loss": 2.4801, + "step": 30785 + }, + { + "epoch": 1.31, + "grad_norm": 0.482421875, + "learning_rate": 5.314934677604275e-05, + "loss": 2.4978, + "step": 30786 + }, + { + "epoch": 1.31, + "grad_norm": 0.515625, + "learning_rate": 5.314342069204268e-05, + "loss": 2.6321, + "step": 30787 + }, + { + "epoch": 1.31, + "grad_norm": 0.5078125, + "learning_rate": 5.3137494818880074e-05, + "loss": 2.5504, + "step": 30788 + }, + { + "epoch": 1.31, + "grad_norm": 0.3984375, + "learning_rate": 5.3131569156581663e-05, + "loss": 2.4314, + "step": 30789 + }, + { + "epoch": 1.31, + "grad_norm": 0.396484375, + "learning_rate": 5.312564370517402e-05, + "loss": 2.409, + "step": 30790 + }, + { + "epoch": 1.31, + "grad_norm": 0.5234375, + "learning_rate": 5.3119718464683855e-05, + "loss": 2.4764, + "step": 30791 + }, + { + "epoch": 1.31, + "grad_norm": 0.451171875, + "learning_rate": 5.311379343513786e-05, + "loss": 2.5075, + "step": 30792 + }, + { + "epoch": 1.31, + "grad_norm": 0.486328125, + "learning_rate": 5.310786861656266e-05, + "loss": 2.4688, + "step": 30793 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.310194400898487e-05, + "loss": 2.4743, + "step": 30794 + }, + { + "epoch": 1.31, + "grad_norm": 0.45703125, + "learning_rate": 5.30960196124312e-05, + "loss": 2.3737, + "step": 30795 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.309009542692833e-05, + "loss": 2.4595, + "step": 30796 + }, + { + "epoch": 1.31, + "grad_norm": 0.44140625, + "learning_rate": 5.308417145250286e-05, + "loss": 2.5316, + "step": 30797 + }, + { + "epoch": 1.31, + "grad_norm": 0.5078125, + "learning_rate": 5.3078247689181505e-05, + "loss": 2.4567, + "step": 30798 + }, + { + "epoch": 1.31, + "grad_norm": 0.453125, + "learning_rate": 5.307232413699089e-05, + "loss": 2.4405, + "step": 30799 + }, + { + "epoch": 1.31, + "grad_norm": 0.490234375, + "learning_rate": 5.3066400795957596e-05, + "loss": 2.477, + "step": 30800 + }, + { + "epoch": 1.31, + "grad_norm": 0.46484375, + "learning_rate": 5.306047766610841e-05, + "loss": 2.5269, + "step": 30801 + }, + { + "epoch": 1.31, + "grad_norm": 0.490234375, + "learning_rate": 5.305455474746992e-05, + "loss": 2.48, + "step": 30802 + }, + { + "epoch": 1.31, + "grad_norm": 0.453125, + "learning_rate": 5.304863204006876e-05, + "loss": 2.4627, + "step": 30803 + }, + { + "epoch": 1.31, + "grad_norm": 0.49609375, + "learning_rate": 5.3042709543931576e-05, + "loss": 2.5147, + "step": 30804 + }, + { + "epoch": 1.31, + "grad_norm": 0.51171875, + "learning_rate": 5.3036787259085076e-05, + "loss": 2.5291, + "step": 30805 + }, + { + "epoch": 1.31, + "grad_norm": 0.51953125, + "learning_rate": 5.303086518555585e-05, + "loss": 2.4409, + "step": 30806 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.302494332337058e-05, + "loss": 2.4283, + "step": 30807 + }, + { + "epoch": 1.31, + "grad_norm": 0.4765625, + "learning_rate": 5.301902167255587e-05, + "loss": 2.3959, + "step": 30808 + }, + { + "epoch": 1.31, + "grad_norm": 0.4453125, + "learning_rate": 5.3013100233138435e-05, + "loss": 2.4807, + "step": 30809 + }, + { + "epoch": 1.31, + "grad_norm": 0.490234375, + "learning_rate": 5.300717900514482e-05, + "loss": 2.4999, + "step": 30810 + }, + { + "epoch": 1.31, + "grad_norm": 0.494140625, + "learning_rate": 5.300125798860179e-05, + "loss": 2.3513, + "step": 30811 + }, + { + "epoch": 1.31, + "grad_norm": 0.5078125, + "learning_rate": 5.299533718353588e-05, + "loss": 2.4027, + "step": 30812 + }, + { + "epoch": 1.31, + "grad_norm": 0.474609375, + "learning_rate": 5.2989416589973764e-05, + "loss": 2.4026, + "step": 30813 + }, + { + "epoch": 1.31, + "grad_norm": 0.44140625, + "learning_rate": 5.298349620794214e-05, + "loss": 2.5064, + "step": 30814 + }, + { + "epoch": 1.31, + "grad_norm": 0.47265625, + "learning_rate": 5.297757603746759e-05, + "loss": 2.408, + "step": 30815 + }, + { + "epoch": 1.31, + "grad_norm": 0.44921875, + "learning_rate": 5.297165607857669e-05, + "loss": 2.4443, + "step": 30816 + }, + { + "epoch": 1.31, + "grad_norm": 0.515625, + "learning_rate": 5.2965736331296234e-05, + "loss": 2.5166, + "step": 30817 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.2959816795652773e-05, + "loss": 2.3613, + "step": 30818 + }, + { + "epoch": 1.31, + "grad_norm": 0.484375, + "learning_rate": 5.29538974716729e-05, + "loss": 2.474, + "step": 30819 + }, + { + "epoch": 1.31, + "grad_norm": 0.486328125, + "learning_rate": 5.29479783593833e-05, + "loss": 2.4956, + "step": 30820 + }, + { + "epoch": 1.31, + "grad_norm": 0.484375, + "learning_rate": 5.2942059458810646e-05, + "loss": 2.3442, + "step": 30821 + }, + { + "epoch": 1.31, + "grad_norm": 0.49609375, + "learning_rate": 5.2936140769981476e-05, + "loss": 2.5543, + "step": 30822 + }, + { + "epoch": 1.31, + "grad_norm": 0.4375, + "learning_rate": 5.293022229292253e-05, + "loss": 2.5576, + "step": 30823 + }, + { + "epoch": 1.31, + "grad_norm": 0.45703125, + "learning_rate": 5.292430402766033e-05, + "loss": 2.4375, + "step": 30824 + }, + { + "epoch": 1.31, + "grad_norm": 0.4921875, + "learning_rate": 5.291838597422158e-05, + "loss": 2.3824, + "step": 30825 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.2912468132632875e-05, + "loss": 2.459, + "step": 30826 + }, + { + "epoch": 1.31, + "grad_norm": 0.40625, + "learning_rate": 5.290655050292088e-05, + "loss": 2.5016, + "step": 30827 + }, + { + "epoch": 1.31, + "grad_norm": 0.427734375, + "learning_rate": 5.290063308511217e-05, + "loss": 2.428, + "step": 30828 + }, + { + "epoch": 1.31, + "grad_norm": 0.4609375, + "learning_rate": 5.289471587923338e-05, + "loss": 2.4239, + "step": 30829 + }, + { + "epoch": 1.31, + "grad_norm": 0.435546875, + "learning_rate": 5.288879888531122e-05, + "loss": 2.3622, + "step": 30830 + }, + { + "epoch": 1.31, + "grad_norm": 0.466796875, + "learning_rate": 5.2882882103372225e-05, + "loss": 2.6012, + "step": 30831 + }, + { + "epoch": 1.31, + "grad_norm": 0.462890625, + "learning_rate": 5.2876965533442966e-05, + "loss": 2.4966, + "step": 30832 + }, + { + "epoch": 1.31, + "grad_norm": 0.4375, + "learning_rate": 5.2871049175550224e-05, + "loss": 2.492, + "step": 30833 + }, + { + "epoch": 1.31, + "grad_norm": 0.400390625, + "learning_rate": 5.2865133029720536e-05, + "loss": 2.3002, + "step": 30834 + }, + { + "epoch": 1.31, + "grad_norm": 0.46484375, + "learning_rate": 5.285921709598047e-05, + "loss": 2.4965, + "step": 30835 + }, + { + "epoch": 1.31, + "grad_norm": 0.466796875, + "learning_rate": 5.285330137435672e-05, + "loss": 2.3823, + "step": 30836 + }, + { + "epoch": 1.31, + "grad_norm": 0.455078125, + "learning_rate": 5.284738586487592e-05, + "loss": 2.5067, + "step": 30837 + }, + { + "epoch": 1.31, + "grad_norm": 0.48046875, + "learning_rate": 5.284147056756461e-05, + "loss": 2.4114, + "step": 30838 + }, + { + "epoch": 1.31, + "grad_norm": 0.482421875, + "learning_rate": 5.283555548244947e-05, + "loss": 2.4931, + "step": 30839 + }, + { + "epoch": 1.31, + "grad_norm": 0.4765625, + "learning_rate": 5.2829640609557085e-05, + "loss": 2.419, + "step": 30840 + }, + { + "epoch": 1.31, + "grad_norm": 0.546875, + "learning_rate": 5.282372594891409e-05, + "loss": 2.4872, + "step": 30841 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.2817811500547064e-05, + "loss": 2.438, + "step": 30842 + }, + { + "epoch": 1.31, + "grad_norm": 0.46875, + "learning_rate": 5.281189726448268e-05, + "loss": 2.3716, + "step": 30843 + }, + { + "epoch": 1.31, + "grad_norm": 0.40625, + "learning_rate": 5.280598324074747e-05, + "loss": 2.4559, + "step": 30844 + }, + { + "epoch": 1.31, + "grad_norm": 0.447265625, + "learning_rate": 5.2800069429368094e-05, + "loss": 2.4606, + "step": 30845 + }, + { + "epoch": 1.31, + "grad_norm": 0.4140625, + "learning_rate": 5.279415583037119e-05, + "loss": 2.3997, + "step": 30846 + }, + { + "epoch": 1.31, + "grad_norm": 0.4375, + "learning_rate": 5.278824244378333e-05, + "loss": 2.461, + "step": 30847 + }, + { + "epoch": 1.31, + "grad_norm": 0.470703125, + "learning_rate": 5.278232926963105e-05, + "loss": 2.6026, + "step": 30848 + }, + { + "epoch": 1.31, + "grad_norm": 0.40234375, + "learning_rate": 5.2776416307941114e-05, + "loss": 2.4583, + "step": 30849 + }, + { + "epoch": 1.31, + "grad_norm": 0.462890625, + "learning_rate": 5.2770503558740046e-05, + "loss": 2.503, + "step": 30850 + }, + { + "epoch": 1.31, + "grad_norm": 0.5078125, + "learning_rate": 5.27645910220544e-05, + "loss": 2.3849, + "step": 30851 + }, + { + "epoch": 1.31, + "grad_norm": 0.47265625, + "learning_rate": 5.2758678697910845e-05, + "loss": 2.5129, + "step": 30852 + }, + { + "epoch": 1.31, + "grad_norm": 0.5546875, + "learning_rate": 5.2752766586336014e-05, + "loss": 2.5253, + "step": 30853 + }, + { + "epoch": 1.31, + "grad_norm": 0.52734375, + "learning_rate": 5.274685468735642e-05, + "loss": 2.4801, + "step": 30854 + }, + { + "epoch": 1.31, + "grad_norm": 0.53125, + "learning_rate": 5.2740943000998746e-05, + "loss": 2.4206, + "step": 30855 + }, + { + "epoch": 1.31, + "grad_norm": 0.48046875, + "learning_rate": 5.273503152728952e-05, + "loss": 2.5852, + "step": 30856 + }, + { + "epoch": 1.31, + "grad_norm": 0.4375, + "learning_rate": 5.2729120266255426e-05, + "loss": 2.4984, + "step": 30857 + }, + { + "epoch": 1.31, + "grad_norm": 0.443359375, + "learning_rate": 5.272320921792298e-05, + "loss": 2.5345, + "step": 30858 + }, + { + "epoch": 1.31, + "grad_norm": 0.5, + "learning_rate": 5.2717298382318836e-05, + "loss": 2.3955, + "step": 30859 + }, + { + "epoch": 1.32, + "grad_norm": 0.50390625, + "learning_rate": 5.271138775946953e-05, + "loss": 2.5019, + "step": 30860 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.270547734940171e-05, + "loss": 2.5663, + "step": 30861 + }, + { + "epoch": 1.32, + "grad_norm": 0.416015625, + "learning_rate": 5.2699567152141994e-05, + "loss": 2.4319, + "step": 30862 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.269365716771695e-05, + "loss": 2.5215, + "step": 30863 + }, + { + "epoch": 1.32, + "grad_norm": 0.45703125, + "learning_rate": 5.2687747396153056e-05, + "loss": 2.4385, + "step": 30864 + }, + { + "epoch": 1.32, + "grad_norm": 0.4921875, + "learning_rate": 5.268183783747711e-05, + "loss": 2.506, + "step": 30865 + }, + { + "epoch": 1.32, + "grad_norm": 0.46484375, + "learning_rate": 5.267592849171559e-05, + "loss": 2.4436, + "step": 30866 + }, + { + "epoch": 1.32, + "grad_norm": 0.490234375, + "learning_rate": 5.2670019358895064e-05, + "loss": 2.4123, + "step": 30867 + }, + { + "epoch": 1.32, + "grad_norm": 0.498046875, + "learning_rate": 5.266411043904214e-05, + "loss": 2.5768, + "step": 30868 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.265820173218348e-05, + "loss": 2.3887, + "step": 30869 + }, + { + "epoch": 1.32, + "grad_norm": 0.431640625, + "learning_rate": 5.265229323834555e-05, + "loss": 2.4727, + "step": 30870 + }, + { + "epoch": 1.32, + "grad_norm": 0.466796875, + "learning_rate": 5.264638495755505e-05, + "loss": 2.3507, + "step": 30871 + }, + { + "epoch": 1.32, + "grad_norm": 0.46875, + "learning_rate": 5.264047688983846e-05, + "loss": 2.4567, + "step": 30872 + }, + { + "epoch": 1.32, + "grad_norm": 0.443359375, + "learning_rate": 5.263456903522246e-05, + "loss": 2.4742, + "step": 30873 + }, + { + "epoch": 1.32, + "grad_norm": 0.474609375, + "learning_rate": 5.2628661393733555e-05, + "loss": 2.4347, + "step": 30874 + }, + { + "epoch": 1.32, + "grad_norm": 0.4453125, + "learning_rate": 5.2622753965398394e-05, + "loss": 2.3922, + "step": 30875 + }, + { + "epoch": 1.32, + "grad_norm": 0.458984375, + "learning_rate": 5.2616846750243495e-05, + "loss": 2.3535, + "step": 30876 + }, + { + "epoch": 1.32, + "grad_norm": 0.53125, + "learning_rate": 5.261093974829548e-05, + "loss": 2.4718, + "step": 30877 + }, + { + "epoch": 1.32, + "grad_norm": 0.51953125, + "learning_rate": 5.2605032959580935e-05, + "loss": 2.3718, + "step": 30878 + }, + { + "epoch": 1.32, + "grad_norm": 0.451171875, + "learning_rate": 5.2599126384126427e-05, + "loss": 2.5854, + "step": 30879 + }, + { + "epoch": 1.32, + "grad_norm": 0.435546875, + "learning_rate": 5.259322002195845e-05, + "loss": 2.4929, + "step": 30880 + }, + { + "epoch": 1.32, + "grad_norm": 0.447265625, + "learning_rate": 5.258731387310374e-05, + "loss": 2.3945, + "step": 30881 + }, + { + "epoch": 1.32, + "grad_norm": 0.46484375, + "learning_rate": 5.258140793758879e-05, + "loss": 2.4361, + "step": 30882 + }, + { + "epoch": 1.32, + "grad_norm": 0.515625, + "learning_rate": 5.257550221544012e-05, + "loss": 2.3717, + "step": 30883 + }, + { + "epoch": 1.32, + "grad_norm": 0.443359375, + "learning_rate": 5.256959670668437e-05, + "loss": 2.4235, + "step": 30884 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.256369141134815e-05, + "loss": 2.3834, + "step": 30885 + }, + { + "epoch": 1.32, + "grad_norm": 0.52734375, + "learning_rate": 5.255778632945794e-05, + "loss": 2.4392, + "step": 30886 + }, + { + "epoch": 1.32, + "grad_norm": 0.5078125, + "learning_rate": 5.2551881461040374e-05, + "loss": 2.6616, + "step": 30887 + }, + { + "epoch": 1.32, + "grad_norm": 0.443359375, + "learning_rate": 5.254597680612198e-05, + "loss": 2.302, + "step": 30888 + }, + { + "epoch": 1.32, + "grad_norm": 0.435546875, + "learning_rate": 5.254007236472938e-05, + "loss": 2.3794, + "step": 30889 + }, + { + "epoch": 1.32, + "grad_norm": 0.4140625, + "learning_rate": 5.253416813688907e-05, + "loss": 2.4097, + "step": 30890 + }, + { + "epoch": 1.32, + "grad_norm": 0.482421875, + "learning_rate": 5.25282641226277e-05, + "loss": 2.3967, + "step": 30891 + }, + { + "epoch": 1.32, + "grad_norm": 0.47265625, + "learning_rate": 5.252236032197175e-05, + "loss": 2.5517, + "step": 30892 + }, + { + "epoch": 1.32, + "grad_norm": 0.4921875, + "learning_rate": 5.2516456734947826e-05, + "loss": 2.4004, + "step": 30893 + }, + { + "epoch": 1.32, + "grad_norm": 0.41015625, + "learning_rate": 5.251055336158254e-05, + "loss": 2.2712, + "step": 30894 + }, + { + "epoch": 1.32, + "grad_norm": 0.46484375, + "learning_rate": 5.250465020190241e-05, + "loss": 2.457, + "step": 30895 + }, + { + "epoch": 1.32, + "grad_norm": 0.484375, + "learning_rate": 5.249874725593391e-05, + "loss": 2.5493, + "step": 30896 + }, + { + "epoch": 1.32, + "grad_norm": 0.5390625, + "learning_rate": 5.2492844523703755e-05, + "loss": 2.52, + "step": 30897 + }, + { + "epoch": 1.32, + "grad_norm": 0.453125, + "learning_rate": 5.248694200523846e-05, + "loss": 2.4939, + "step": 30898 + }, + { + "epoch": 1.32, + "grad_norm": 0.47265625, + "learning_rate": 5.2481039700564506e-05, + "loss": 2.4717, + "step": 30899 + }, + { + "epoch": 1.32, + "grad_norm": 0.4609375, + "learning_rate": 5.247513760970852e-05, + "loss": 2.4108, + "step": 30900 + }, + { + "epoch": 1.32, + "grad_norm": 0.46484375, + "learning_rate": 5.2469235732697084e-05, + "loss": 2.55, + "step": 30901 + }, + { + "epoch": 1.32, + "grad_norm": 0.380859375, + "learning_rate": 5.246333406955667e-05, + "loss": 2.5135, + "step": 30902 + }, + { + "epoch": 1.32, + "grad_norm": 0.484375, + "learning_rate": 5.245743262031393e-05, + "loss": 2.5308, + "step": 30903 + }, + { + "epoch": 1.32, + "grad_norm": 0.453125, + "learning_rate": 5.245153138499532e-05, + "loss": 2.4934, + "step": 30904 + }, + { + "epoch": 1.32, + "grad_norm": 0.466796875, + "learning_rate": 5.2445630363627484e-05, + "loss": 2.2878, + "step": 30905 + }, + { + "epoch": 1.32, + "grad_norm": 0.4609375, + "learning_rate": 5.2439729556236885e-05, + "loss": 2.4975, + "step": 30906 + }, + { + "epoch": 1.32, + "grad_norm": 0.4765625, + "learning_rate": 5.2433828962850164e-05, + "loss": 2.411, + "step": 30907 + }, + { + "epoch": 1.32, + "grad_norm": 0.486328125, + "learning_rate": 5.24279285834938e-05, + "loss": 2.4208, + "step": 30908 + }, + { + "epoch": 1.32, + "grad_norm": 0.439453125, + "learning_rate": 5.2422028418194366e-05, + "loss": 2.3912, + "step": 30909 + }, + { + "epoch": 1.32, + "grad_norm": 0.412109375, + "learning_rate": 5.241612846697846e-05, + "loss": 2.5921, + "step": 30910 + }, + { + "epoch": 1.32, + "grad_norm": 0.451171875, + "learning_rate": 5.241022872987258e-05, + "loss": 2.5224, + "step": 30911 + }, + { + "epoch": 1.32, + "grad_norm": 0.455078125, + "learning_rate": 5.240432920690319e-05, + "loss": 2.4581, + "step": 30912 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.239842989809702e-05, + "loss": 2.4303, + "step": 30913 + }, + { + "epoch": 1.32, + "grad_norm": 0.462890625, + "learning_rate": 5.2392530803480525e-05, + "loss": 2.4605, + "step": 30914 + }, + { + "epoch": 1.32, + "grad_norm": 0.46875, + "learning_rate": 5.238663192308018e-05, + "loss": 2.4907, + "step": 30915 + }, + { + "epoch": 1.32, + "grad_norm": 0.5234375, + "learning_rate": 5.238073325692261e-05, + "loss": 2.6414, + "step": 30916 + }, + { + "epoch": 1.32, + "grad_norm": 0.462890625, + "learning_rate": 5.237483480503437e-05, + "loss": 2.391, + "step": 30917 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.236893656744193e-05, + "loss": 2.3359, + "step": 30918 + }, + { + "epoch": 1.32, + "grad_norm": 0.431640625, + "learning_rate": 5.23630385441719e-05, + "loss": 2.4687, + "step": 30919 + }, + { + "epoch": 1.32, + "grad_norm": 0.44140625, + "learning_rate": 5.23571407352508e-05, + "loss": 2.5086, + "step": 30920 + }, + { + "epoch": 1.32, + "grad_norm": 0.466796875, + "learning_rate": 5.23512431407051e-05, + "loss": 2.4378, + "step": 30921 + }, + { + "epoch": 1.32, + "grad_norm": 0.48828125, + "learning_rate": 5.2345345760561405e-05, + "loss": 2.4779, + "step": 30922 + }, + { + "epoch": 1.32, + "grad_norm": 0.498046875, + "learning_rate": 5.2339448594846275e-05, + "loss": 2.367, + "step": 30923 + }, + { + "epoch": 1.32, + "grad_norm": 0.46875, + "learning_rate": 5.233355164358617e-05, + "loss": 2.3241, + "step": 30924 + }, + { + "epoch": 1.32, + "grad_norm": 0.408203125, + "learning_rate": 5.232765490680767e-05, + "loss": 2.4797, + "step": 30925 + }, + { + "epoch": 1.32, + "grad_norm": 0.453125, + "learning_rate": 5.232175838453732e-05, + "loss": 2.4194, + "step": 30926 + }, + { + "epoch": 1.32, + "grad_norm": 0.466796875, + "learning_rate": 5.231586207680164e-05, + "loss": 2.4016, + "step": 30927 + }, + { + "epoch": 1.32, + "grad_norm": 0.455078125, + "learning_rate": 5.230996598362712e-05, + "loss": 2.4063, + "step": 30928 + }, + { + "epoch": 1.32, + "grad_norm": 0.484375, + "learning_rate": 5.230407010504032e-05, + "loss": 2.4598, + "step": 30929 + }, + { + "epoch": 1.32, + "grad_norm": 0.447265625, + "learning_rate": 5.229817444106782e-05, + "loss": 2.4847, + "step": 30930 + }, + { + "epoch": 1.32, + "grad_norm": 0.43359375, + "learning_rate": 5.229227899173605e-05, + "loss": 2.3232, + "step": 30931 + }, + { + "epoch": 1.32, + "grad_norm": 0.5234375, + "learning_rate": 5.2286383757071646e-05, + "loss": 2.4283, + "step": 30932 + }, + { + "epoch": 1.32, + "grad_norm": 0.474609375, + "learning_rate": 5.228048873710103e-05, + "loss": 2.4585, + "step": 30933 + }, + { + "epoch": 1.32, + "grad_norm": 0.51171875, + "learning_rate": 5.227459393185077e-05, + "loss": 2.5228, + "step": 30934 + }, + { + "epoch": 1.32, + "grad_norm": 0.498046875, + "learning_rate": 5.226869934134744e-05, + "loss": 2.5524, + "step": 30935 + }, + { + "epoch": 1.32, + "grad_norm": 0.494140625, + "learning_rate": 5.226280496561753e-05, + "loss": 2.5199, + "step": 30936 + }, + { + "epoch": 1.32, + "grad_norm": 0.51171875, + "learning_rate": 5.2256910804687485e-05, + "loss": 2.4107, + "step": 30937 + }, + { + "epoch": 1.32, + "grad_norm": 0.474609375, + "learning_rate": 5.225101685858391e-05, + "loss": 2.4711, + "step": 30938 + }, + { + "epoch": 1.32, + "grad_norm": 0.5234375, + "learning_rate": 5.224512312733336e-05, + "loss": 2.4682, + "step": 30939 + }, + { + "epoch": 1.32, + "grad_norm": 0.484375, + "learning_rate": 5.223922961096225e-05, + "loss": 2.4112, + "step": 30940 + }, + { + "epoch": 1.32, + "grad_norm": 0.515625, + "learning_rate": 5.223333630949715e-05, + "loss": 2.5754, + "step": 30941 + }, + { + "epoch": 1.32, + "grad_norm": 0.515625, + "learning_rate": 5.222744322296462e-05, + "loss": 2.5109, + "step": 30942 + }, + { + "epoch": 1.32, + "grad_norm": 0.416015625, + "learning_rate": 5.222155035139115e-05, + "loss": 2.5644, + "step": 30943 + }, + { + "epoch": 1.32, + "grad_norm": 0.546875, + "learning_rate": 5.221565769480319e-05, + "loss": 2.5328, + "step": 30944 + }, + { + "epoch": 1.32, + "grad_norm": 0.419921875, + "learning_rate": 5.220976525322732e-05, + "loss": 2.4692, + "step": 30945 + }, + { + "epoch": 1.32, + "grad_norm": 0.5078125, + "learning_rate": 5.220387302669005e-05, + "loss": 2.4285, + "step": 30946 + }, + { + "epoch": 1.32, + "grad_norm": 0.47265625, + "learning_rate": 5.219798101521787e-05, + "loss": 2.5271, + "step": 30947 + }, + { + "epoch": 1.32, + "grad_norm": 0.47265625, + "learning_rate": 5.219208921883734e-05, + "loss": 2.5598, + "step": 30948 + }, + { + "epoch": 1.32, + "grad_norm": 0.408203125, + "learning_rate": 5.218619763757491e-05, + "loss": 2.3641, + "step": 30949 + }, + { + "epoch": 1.32, + "grad_norm": 0.4765625, + "learning_rate": 5.2180306271457094e-05, + "loss": 2.4127, + "step": 30950 + }, + { + "epoch": 1.32, + "grad_norm": 0.462890625, + "learning_rate": 5.217441512051047e-05, + "loss": 2.4522, + "step": 30951 + }, + { + "epoch": 1.32, + "grad_norm": 0.51171875, + "learning_rate": 5.21685241847615e-05, + "loss": 2.4309, + "step": 30952 + }, + { + "epoch": 1.32, + "grad_norm": 0.416015625, + "learning_rate": 5.2162633464236666e-05, + "loss": 2.517, + "step": 30953 + }, + { + "epoch": 1.32, + "grad_norm": 0.5, + "learning_rate": 5.215674295896249e-05, + "loss": 2.3946, + "step": 30954 + }, + { + "epoch": 1.32, + "grad_norm": 0.484375, + "learning_rate": 5.215085266896551e-05, + "loss": 2.3098, + "step": 30955 + }, + { + "epoch": 1.32, + "grad_norm": 0.494140625, + "learning_rate": 5.214496259427218e-05, + "loss": 2.4801, + "step": 30956 + }, + { + "epoch": 1.32, + "grad_norm": 0.466796875, + "learning_rate": 5.2139072734909035e-05, + "loss": 2.4938, + "step": 30957 + }, + { + "epoch": 1.32, + "grad_norm": 0.44921875, + "learning_rate": 5.2133183090902605e-05, + "loss": 2.5926, + "step": 30958 + }, + { + "epoch": 1.32, + "grad_norm": 0.474609375, + "learning_rate": 5.212729366227935e-05, + "loss": 2.5141, + "step": 30959 + }, + { + "epoch": 1.32, + "grad_norm": 0.5546875, + "learning_rate": 5.212140444906575e-05, + "loss": 2.4241, + "step": 30960 + }, + { + "epoch": 1.32, + "grad_norm": 0.44140625, + "learning_rate": 5.211551545128833e-05, + "loss": 2.2511, + "step": 30961 + }, + { + "epoch": 1.32, + "grad_norm": 0.53125, + "learning_rate": 5.210962666897363e-05, + "loss": 2.3545, + "step": 30962 + }, + { + "epoch": 1.32, + "grad_norm": 0.5078125, + "learning_rate": 5.210373810214807e-05, + "loss": 2.5585, + "step": 30963 + }, + { + "epoch": 1.32, + "grad_norm": 0.443359375, + "learning_rate": 5.209784975083822e-05, + "loss": 2.491, + "step": 30964 + }, + { + "epoch": 1.32, + "grad_norm": 0.4296875, + "learning_rate": 5.20919616150705e-05, + "loss": 2.3622, + "step": 30965 + }, + { + "epoch": 1.32, + "grad_norm": 0.421875, + "learning_rate": 5.208607369487144e-05, + "loss": 2.4082, + "step": 30966 + }, + { + "epoch": 1.32, + "grad_norm": 0.453125, + "learning_rate": 5.208018599026758e-05, + "loss": 2.5248, + "step": 30967 + }, + { + "epoch": 1.32, + "grad_norm": 0.51171875, + "learning_rate": 5.207429850128538e-05, + "loss": 2.2605, + "step": 30968 + }, + { + "epoch": 1.32, + "grad_norm": 0.494140625, + "learning_rate": 5.206841122795127e-05, + "loss": 2.3692, + "step": 30969 + }, + { + "epoch": 1.32, + "grad_norm": 0.45703125, + "learning_rate": 5.2062524170291804e-05, + "loss": 2.4168, + "step": 30970 + }, + { + "epoch": 1.32, + "grad_norm": 0.48828125, + "learning_rate": 5.2056637328333504e-05, + "loss": 2.4118, + "step": 30971 + }, + { + "epoch": 1.32, + "grad_norm": 0.4765625, + "learning_rate": 5.205075070210276e-05, + "loss": 2.4005, + "step": 30972 + }, + { + "epoch": 1.32, + "grad_norm": 0.46875, + "learning_rate": 5.2044864291626114e-05, + "loss": 2.4912, + "step": 30973 + }, + { + "epoch": 1.32, + "grad_norm": 0.40234375, + "learning_rate": 5.20389780969301e-05, + "loss": 2.4674, + "step": 30974 + }, + { + "epoch": 1.32, + "grad_norm": 0.50390625, + "learning_rate": 5.203309211804115e-05, + "loss": 2.5048, + "step": 30975 + }, + { + "epoch": 1.32, + "grad_norm": 0.54296875, + "learning_rate": 5.202720635498573e-05, + "loss": 2.4681, + "step": 30976 + }, + { + "epoch": 1.32, + "grad_norm": 0.5390625, + "learning_rate": 5.202132080779034e-05, + "loss": 2.518, + "step": 30977 + }, + { + "epoch": 1.32, + "grad_norm": 0.4609375, + "learning_rate": 5.201543547648151e-05, + "loss": 2.4088, + "step": 30978 + }, + { + "epoch": 1.32, + "grad_norm": 0.455078125, + "learning_rate": 5.200955036108565e-05, + "loss": 2.4137, + "step": 30979 + }, + { + "epoch": 1.32, + "grad_norm": 0.462890625, + "learning_rate": 5.200366546162931e-05, + "loss": 2.5789, + "step": 30980 + }, + { + "epoch": 1.32, + "grad_norm": 0.4609375, + "learning_rate": 5.1997780778138896e-05, + "loss": 2.3155, + "step": 30981 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.199189631064092e-05, + "loss": 2.4901, + "step": 30982 + }, + { + "epoch": 1.32, + "grad_norm": 0.447265625, + "learning_rate": 5.198601205916192e-05, + "loss": 2.4146, + "step": 30983 + }, + { + "epoch": 1.32, + "grad_norm": 0.44921875, + "learning_rate": 5.1980128023728304e-05, + "loss": 2.3444, + "step": 30984 + }, + { + "epoch": 1.32, + "grad_norm": 0.4453125, + "learning_rate": 5.197424420436653e-05, + "loss": 2.4368, + "step": 30985 + }, + { + "epoch": 1.32, + "grad_norm": 0.5078125, + "learning_rate": 5.1968360601103103e-05, + "loss": 2.5148, + "step": 30986 + }, + { + "epoch": 1.32, + "grad_norm": 0.498046875, + "learning_rate": 5.196247721396455e-05, + "loss": 2.507, + "step": 30987 + }, + { + "epoch": 1.32, + "grad_norm": 0.4453125, + "learning_rate": 5.195659404297724e-05, + "loss": 2.4793, + "step": 30988 + }, + { + "epoch": 1.32, + "grad_norm": 0.4375, + "learning_rate": 5.19507110881677e-05, + "loss": 2.4064, + "step": 30989 + }, + { + "epoch": 1.32, + "grad_norm": 0.48828125, + "learning_rate": 5.194482834956245e-05, + "loss": 2.3375, + "step": 30990 + }, + { + "epoch": 1.32, + "grad_norm": 0.462890625, + "learning_rate": 5.19389458271879e-05, + "loss": 2.4094, + "step": 30991 + }, + { + "epoch": 1.32, + "grad_norm": 0.4609375, + "learning_rate": 5.193306352107049e-05, + "loss": 2.4989, + "step": 30992 + }, + { + "epoch": 1.32, + "grad_norm": 0.51953125, + "learning_rate": 5.192718143123674e-05, + "loss": 2.4953, + "step": 30993 + }, + { + "epoch": 1.32, + "grad_norm": 0.4765625, + "learning_rate": 5.192129955771312e-05, + "loss": 2.5777, + "step": 30994 + }, + { + "epoch": 1.32, + "grad_norm": 0.451171875, + "learning_rate": 5.1915417900526074e-05, + "loss": 2.4655, + "step": 30995 + }, + { + "epoch": 1.32, + "grad_norm": 0.5, + "learning_rate": 5.1909536459702094e-05, + "loss": 2.394, + "step": 30996 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.190365523526759e-05, + "loss": 2.4753, + "step": 30997 + }, + { + "epoch": 1.32, + "grad_norm": 0.49609375, + "learning_rate": 5.1897774227249065e-05, + "loss": 2.352, + "step": 30998 + }, + { + "epoch": 1.32, + "grad_norm": 0.46484375, + "learning_rate": 5.1891893435673e-05, + "loss": 2.4566, + "step": 30999 + }, + { + "epoch": 1.32, + "grad_norm": 0.462890625, + "learning_rate": 5.188601286056585e-05, + "loss": 2.3749, + "step": 31000 + }, + { + "epoch": 1.32, + "grad_norm": 0.482421875, + "learning_rate": 5.188013250195401e-05, + "loss": 2.4446, + "step": 31001 + }, + { + "epoch": 1.32, + "grad_norm": 0.50390625, + "learning_rate": 5.187425235986398e-05, + "loss": 2.5853, + "step": 31002 + }, + { + "epoch": 1.32, + "grad_norm": 0.482421875, + "learning_rate": 5.186837243432228e-05, + "loss": 2.4622, + "step": 31003 + }, + { + "epoch": 1.32, + "grad_norm": 0.51953125, + "learning_rate": 5.186249272535528e-05, + "loss": 2.5847, + "step": 31004 + }, + { + "epoch": 1.32, + "grad_norm": 0.462890625, + "learning_rate": 5.185661323298946e-05, + "loss": 2.5229, + "step": 31005 + }, + { + "epoch": 1.32, + "grad_norm": 0.435546875, + "learning_rate": 5.185073395725133e-05, + "loss": 2.421, + "step": 31006 + }, + { + "epoch": 1.32, + "grad_norm": 0.4453125, + "learning_rate": 5.1844854898167286e-05, + "loss": 2.4015, + "step": 31007 + }, + { + "epoch": 1.32, + "grad_norm": 0.5, + "learning_rate": 5.1838976055763774e-05, + "loss": 2.4575, + "step": 31008 + }, + { + "epoch": 1.32, + "grad_norm": 0.48046875, + "learning_rate": 5.183309743006726e-05, + "loss": 2.4374, + "step": 31009 + }, + { + "epoch": 1.32, + "grad_norm": 0.482421875, + "learning_rate": 5.1827219021104256e-05, + "loss": 2.5257, + "step": 31010 + }, + { + "epoch": 1.32, + "grad_norm": 0.44921875, + "learning_rate": 5.182134082890111e-05, + "loss": 2.3658, + "step": 31011 + }, + { + "epoch": 1.32, + "grad_norm": 0.466796875, + "learning_rate": 5.1815462853484365e-05, + "loss": 2.4398, + "step": 31012 + }, + { + "epoch": 1.32, + "grad_norm": 0.48046875, + "learning_rate": 5.180958509488039e-05, + "loss": 2.2905, + "step": 31013 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.1803707553115665e-05, + "loss": 2.3602, + "step": 31014 + }, + { + "epoch": 1.32, + "grad_norm": 0.4765625, + "learning_rate": 5.179783022821669e-05, + "loss": 2.3396, + "step": 31015 + }, + { + "epoch": 1.32, + "grad_norm": 0.486328125, + "learning_rate": 5.179195312020986e-05, + "loss": 2.4881, + "step": 31016 + }, + { + "epoch": 1.32, + "grad_norm": 0.494140625, + "learning_rate": 5.1786076229121575e-05, + "loss": 2.4255, + "step": 31017 + }, + { + "epoch": 1.32, + "grad_norm": 0.392578125, + "learning_rate": 5.178019955497834e-05, + "loss": 2.3763, + "step": 31018 + }, + { + "epoch": 1.32, + "grad_norm": 0.453125, + "learning_rate": 5.177432309780661e-05, + "loss": 2.494, + "step": 31019 + }, + { + "epoch": 1.32, + "grad_norm": 0.494140625, + "learning_rate": 5.176844685763278e-05, + "loss": 2.5604, + "step": 31020 + }, + { + "epoch": 1.32, + "grad_norm": 0.52734375, + "learning_rate": 5.17625708344833e-05, + "loss": 2.4552, + "step": 31021 + }, + { + "epoch": 1.32, + "grad_norm": 0.482421875, + "learning_rate": 5.175669502838467e-05, + "loss": 2.5729, + "step": 31022 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.175081943936327e-05, + "loss": 2.4894, + "step": 31023 + }, + { + "epoch": 1.32, + "grad_norm": 0.453125, + "learning_rate": 5.174494406744552e-05, + "loss": 2.4446, + "step": 31024 + }, + { + "epoch": 1.32, + "grad_norm": 0.451171875, + "learning_rate": 5.173906891265788e-05, + "loss": 2.4652, + "step": 31025 + }, + { + "epoch": 1.32, + "grad_norm": 0.423828125, + "learning_rate": 5.173319397502684e-05, + "loss": 2.3626, + "step": 31026 + }, + { + "epoch": 1.32, + "grad_norm": 0.46875, + "learning_rate": 5.172731925457874e-05, + "loss": 2.4494, + "step": 31027 + }, + { + "epoch": 1.32, + "grad_norm": 0.42578125, + "learning_rate": 5.172144475134011e-05, + "loss": 2.4382, + "step": 31028 + }, + { + "epoch": 1.32, + "grad_norm": 0.439453125, + "learning_rate": 5.171557046533728e-05, + "loss": 2.4118, + "step": 31029 + }, + { + "epoch": 1.32, + "grad_norm": 0.4765625, + "learning_rate": 5.170969639659675e-05, + "loss": 2.4546, + "step": 31030 + }, + { + "epoch": 1.32, + "grad_norm": 0.48828125, + "learning_rate": 5.1703822545144975e-05, + "loss": 2.3976, + "step": 31031 + }, + { + "epoch": 1.32, + "grad_norm": 0.458984375, + "learning_rate": 5.169794891100834e-05, + "loss": 2.5821, + "step": 31032 + }, + { + "epoch": 1.32, + "grad_norm": 0.48046875, + "learning_rate": 5.169207549421326e-05, + "loss": 2.5544, + "step": 31033 + }, + { + "epoch": 1.32, + "grad_norm": 0.515625, + "learning_rate": 5.168620229478618e-05, + "loss": 2.4803, + "step": 31034 + }, + { + "epoch": 1.32, + "grad_norm": 0.494140625, + "learning_rate": 5.168032931275356e-05, + "loss": 2.4955, + "step": 31035 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.167445654814178e-05, + "loss": 2.3811, + "step": 31036 + }, + { + "epoch": 1.32, + "grad_norm": 0.50390625, + "learning_rate": 5.166858400097727e-05, + "loss": 2.3844, + "step": 31037 + }, + { + "epoch": 1.32, + "grad_norm": 0.515625, + "learning_rate": 5.166271167128651e-05, + "loss": 2.4867, + "step": 31038 + }, + { + "epoch": 1.32, + "grad_norm": 0.546875, + "learning_rate": 5.1656839559095884e-05, + "loss": 2.5949, + "step": 31039 + }, + { + "epoch": 1.32, + "grad_norm": 0.4609375, + "learning_rate": 5.1650967664431774e-05, + "loss": 2.5569, + "step": 31040 + }, + { + "epoch": 1.32, + "grad_norm": 0.50390625, + "learning_rate": 5.164509598732068e-05, + "loss": 2.4056, + "step": 31041 + }, + { + "epoch": 1.32, + "grad_norm": 0.48046875, + "learning_rate": 5.163922452778893e-05, + "loss": 2.5141, + "step": 31042 + }, + { + "epoch": 1.32, + "grad_norm": 0.435546875, + "learning_rate": 5.1633353285863005e-05, + "loss": 2.41, + "step": 31043 + }, + { + "epoch": 1.32, + "grad_norm": 0.50390625, + "learning_rate": 5.162748226156937e-05, + "loss": 2.6025, + "step": 31044 + }, + { + "epoch": 1.32, + "grad_norm": 0.462890625, + "learning_rate": 5.162161145493436e-05, + "loss": 2.4578, + "step": 31045 + }, + { + "epoch": 1.32, + "grad_norm": 0.408203125, + "learning_rate": 5.161574086598435e-05, + "loss": 2.3773, + "step": 31046 + }, + { + "epoch": 1.32, + "grad_norm": 0.4375, + "learning_rate": 5.1609870494745914e-05, + "loss": 2.4225, + "step": 31047 + }, + { + "epoch": 1.32, + "grad_norm": 0.439453125, + "learning_rate": 5.160400034124537e-05, + "loss": 2.5744, + "step": 31048 + }, + { + "epoch": 1.32, + "grad_norm": 0.42578125, + "learning_rate": 5.1598130405509094e-05, + "loss": 2.3834, + "step": 31049 + }, + { + "epoch": 1.32, + "grad_norm": 0.515625, + "learning_rate": 5.159226068756353e-05, + "loss": 2.5019, + "step": 31050 + }, + { + "epoch": 1.32, + "grad_norm": 0.4921875, + "learning_rate": 5.158639118743516e-05, + "loss": 2.5529, + "step": 31051 + }, + { + "epoch": 1.32, + "grad_norm": 0.51171875, + "learning_rate": 5.1580521905150306e-05, + "loss": 2.4744, + "step": 31052 + }, + { + "epoch": 1.32, + "grad_norm": 0.494140625, + "learning_rate": 5.1574652840735415e-05, + "loss": 2.5584, + "step": 31053 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.1568783994216875e-05, + "loss": 2.5277, + "step": 31054 + }, + { + "epoch": 1.32, + "grad_norm": 0.453125, + "learning_rate": 5.156291536562114e-05, + "loss": 2.488, + "step": 31055 + }, + { + "epoch": 1.32, + "grad_norm": 0.515625, + "learning_rate": 5.1557046954974544e-05, + "loss": 2.6354, + "step": 31056 + }, + { + "epoch": 1.32, + "grad_norm": 0.46484375, + "learning_rate": 5.155117876230357e-05, + "loss": 2.3593, + "step": 31057 + }, + { + "epoch": 1.32, + "grad_norm": 0.474609375, + "learning_rate": 5.154531078763455e-05, + "loss": 2.5223, + "step": 31058 + }, + { + "epoch": 1.32, + "grad_norm": 0.41015625, + "learning_rate": 5.153944303099392e-05, + "loss": 2.6161, + "step": 31059 + }, + { + "epoch": 1.32, + "grad_norm": 0.44921875, + "learning_rate": 5.153357549240814e-05, + "loss": 2.4047, + "step": 31060 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.1527708171903534e-05, + "loss": 2.4595, + "step": 31061 + }, + { + "epoch": 1.32, + "grad_norm": 0.498046875, + "learning_rate": 5.152184106950646e-05, + "loss": 2.5396, + "step": 31062 + }, + { + "epoch": 1.32, + "grad_norm": 0.4453125, + "learning_rate": 5.151597418524345e-05, + "loss": 2.4651, + "step": 31063 + }, + { + "epoch": 1.32, + "grad_norm": 0.431640625, + "learning_rate": 5.151010751914086e-05, + "loss": 2.4144, + "step": 31064 + }, + { + "epoch": 1.32, + "grad_norm": 0.484375, + "learning_rate": 5.150424107122502e-05, + "loss": 2.4546, + "step": 31065 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.149837484152236e-05, + "loss": 2.5493, + "step": 31066 + }, + { + "epoch": 1.32, + "grad_norm": 0.40234375, + "learning_rate": 5.149250883005934e-05, + "loss": 2.3552, + "step": 31067 + }, + { + "epoch": 1.32, + "grad_norm": 0.49609375, + "learning_rate": 5.148664303686226e-05, + "loss": 2.5669, + "step": 31068 + }, + { + "epoch": 1.32, + "grad_norm": 0.49609375, + "learning_rate": 5.148077746195759e-05, + "loss": 2.6527, + "step": 31069 + }, + { + "epoch": 1.32, + "grad_norm": 0.470703125, + "learning_rate": 5.1474912105371654e-05, + "loss": 2.375, + "step": 31070 + }, + { + "epoch": 1.32, + "grad_norm": 0.4140625, + "learning_rate": 5.146904696713092e-05, + "loss": 2.5098, + "step": 31071 + }, + { + "epoch": 1.32, + "grad_norm": 0.50390625, + "learning_rate": 5.1463182047261704e-05, + "loss": 2.5462, + "step": 31072 + }, + { + "epoch": 1.32, + "grad_norm": 0.447265625, + "learning_rate": 5.145731734579046e-05, + "loss": 2.4197, + "step": 31073 + }, + { + "epoch": 1.32, + "grad_norm": 0.4375, + "learning_rate": 5.145145286274351e-05, + "loss": 2.3599, + "step": 31074 + }, + { + "epoch": 1.32, + "grad_norm": 0.4609375, + "learning_rate": 5.144558859814728e-05, + "loss": 2.5207, + "step": 31075 + }, + { + "epoch": 1.32, + "grad_norm": 0.419921875, + "learning_rate": 5.1439724552028204e-05, + "loss": 2.4542, + "step": 31076 + }, + { + "epoch": 1.32, + "grad_norm": 0.4140625, + "learning_rate": 5.143386072441261e-05, + "loss": 2.3974, + "step": 31077 + }, + { + "epoch": 1.32, + "grad_norm": 0.453125, + "learning_rate": 5.1427997115326824e-05, + "loss": 2.4588, + "step": 31078 + }, + { + "epoch": 1.32, + "grad_norm": 0.42578125, + "learning_rate": 5.142213372479737e-05, + "loss": 2.445, + "step": 31079 + }, + { + "epoch": 1.32, + "grad_norm": 0.49609375, + "learning_rate": 5.141627055285057e-05, + "loss": 2.4718, + "step": 31080 + }, + { + "epoch": 1.32, + "grad_norm": 0.455078125, + "learning_rate": 5.141040759951275e-05, + "loss": 2.4203, + "step": 31081 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.1404544864810345e-05, + "loss": 2.6198, + "step": 31082 + }, + { + "epoch": 1.32, + "grad_norm": 0.48828125, + "learning_rate": 5.1398682348769764e-05, + "loss": 2.4376, + "step": 31083 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.1392820051417314e-05, + "loss": 2.2618, + "step": 31084 + }, + { + "epoch": 1.32, + "grad_norm": 0.46875, + "learning_rate": 5.138695797277945e-05, + "loss": 2.4294, + "step": 31085 + }, + { + "epoch": 1.32, + "grad_norm": 0.5625, + "learning_rate": 5.1381096112882466e-05, + "loss": 2.3625, + "step": 31086 + }, + { + "epoch": 1.32, + "grad_norm": 0.498046875, + "learning_rate": 5.137523447175282e-05, + "loss": 2.4371, + "step": 31087 + }, + { + "epoch": 1.32, + "grad_norm": 0.478515625, + "learning_rate": 5.136937304941681e-05, + "loss": 2.4525, + "step": 31088 + }, + { + "epoch": 1.32, + "grad_norm": 0.46484375, + "learning_rate": 5.136351184590089e-05, + "loss": 2.4857, + "step": 31089 + }, + { + "epoch": 1.32, + "grad_norm": 0.53515625, + "learning_rate": 5.1357650861231355e-05, + "loss": 2.4451, + "step": 31090 + }, + { + "epoch": 1.32, + "grad_norm": 0.5546875, + "learning_rate": 5.1351790095434604e-05, + "loss": 2.4568, + "step": 31091 + }, + { + "epoch": 1.32, + "grad_norm": 0.47265625, + "learning_rate": 5.134592954853708e-05, + "loss": 2.375, + "step": 31092 + }, + { + "epoch": 1.32, + "grad_norm": 0.4453125, + "learning_rate": 5.134006922056507e-05, + "loss": 2.3901, + "step": 31093 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.133420911154489e-05, + "loss": 2.5573, + "step": 31094 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.132834922150307e-05, + "loss": 2.5549, + "step": 31095 + }, + { + "epoch": 1.33, + "grad_norm": 0.46484375, + "learning_rate": 5.132248955046588e-05, + "loss": 2.43, + "step": 31096 + }, + { + "epoch": 1.33, + "grad_norm": 0.462890625, + "learning_rate": 5.131663009845967e-05, + "loss": 2.4526, + "step": 31097 + }, + { + "epoch": 1.33, + "grad_norm": 0.515625, + "learning_rate": 5.131077086551084e-05, + "loss": 2.5679, + "step": 31098 + }, + { + "epoch": 1.33, + "grad_norm": 0.55078125, + "learning_rate": 5.130491185164577e-05, + "loss": 2.5388, + "step": 31099 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.1299053056890754e-05, + "loss": 2.5803, + "step": 31100 + }, + { + "epoch": 1.33, + "grad_norm": 0.5, + "learning_rate": 5.129319448127226e-05, + "loss": 2.5036, + "step": 31101 + }, + { + "epoch": 1.33, + "grad_norm": 0.462890625, + "learning_rate": 5.128733612481656e-05, + "loss": 2.5207, + "step": 31102 + }, + { + "epoch": 1.33, + "grad_norm": 0.4609375, + "learning_rate": 5.128147798755005e-05, + "loss": 2.4052, + "step": 31103 + }, + { + "epoch": 1.33, + "grad_norm": 0.46875, + "learning_rate": 5.127562006949906e-05, + "loss": 2.4638, + "step": 31104 + }, + { + "epoch": 1.33, + "grad_norm": 0.482421875, + "learning_rate": 5.126976237069002e-05, + "loss": 2.5244, + "step": 31105 + }, + { + "epoch": 1.33, + "grad_norm": 0.53515625, + "learning_rate": 5.126390489114921e-05, + "loss": 2.3601, + "step": 31106 + }, + { + "epoch": 1.33, + "grad_norm": 0.466796875, + "learning_rate": 5.1258047630902995e-05, + "loss": 2.4499, + "step": 31107 + }, + { + "epoch": 1.33, + "grad_norm": 0.5078125, + "learning_rate": 5.12521905899778e-05, + "loss": 2.4768, + "step": 31108 + }, + { + "epoch": 1.33, + "grad_norm": 0.498046875, + "learning_rate": 5.124633376839993e-05, + "loss": 2.6302, + "step": 31109 + }, + { + "epoch": 1.33, + "grad_norm": 0.419921875, + "learning_rate": 5.1240477166195666e-05, + "loss": 2.4636, + "step": 31110 + }, + { + "epoch": 1.33, + "grad_norm": 0.458984375, + "learning_rate": 5.123462078339152e-05, + "loss": 2.4929, + "step": 31111 + }, + { + "epoch": 1.33, + "grad_norm": 0.515625, + "learning_rate": 5.1228764620013746e-05, + "loss": 2.4243, + "step": 31112 + }, + { + "epoch": 1.33, + "grad_norm": 0.50390625, + "learning_rate": 5.122290867608868e-05, + "loss": 2.3065, + "step": 31113 + }, + { + "epoch": 1.33, + "grad_norm": 0.49609375, + "learning_rate": 5.1217052951642685e-05, + "loss": 2.4559, + "step": 31114 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.121119744670218e-05, + "loss": 2.4003, + "step": 31115 + }, + { + "epoch": 1.33, + "grad_norm": 0.443359375, + "learning_rate": 5.12053421612934e-05, + "loss": 2.2506, + "step": 31116 + }, + { + "epoch": 1.33, + "grad_norm": 0.486328125, + "learning_rate": 5.11994870954428e-05, + "loss": 2.5039, + "step": 31117 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.119363224917661e-05, + "loss": 2.4083, + "step": 31118 + }, + { + "epoch": 1.33, + "grad_norm": 0.486328125, + "learning_rate": 5.11877776225213e-05, + "loss": 2.4747, + "step": 31119 + }, + { + "epoch": 1.33, + "grad_norm": 0.54296875, + "learning_rate": 5.1181923215503113e-05, + "loss": 2.4301, + "step": 31120 + }, + { + "epoch": 1.33, + "grad_norm": 0.431640625, + "learning_rate": 5.1176069028148465e-05, + "loss": 2.3981, + "step": 31121 + }, + { + "epoch": 1.33, + "grad_norm": 0.439453125, + "learning_rate": 5.117021506048363e-05, + "loss": 2.4314, + "step": 31122 + }, + { + "epoch": 1.33, + "grad_norm": 0.4609375, + "learning_rate": 5.116436131253497e-05, + "loss": 2.4183, + "step": 31123 + }, + { + "epoch": 1.33, + "grad_norm": 0.50390625, + "learning_rate": 5.115850778432889e-05, + "loss": 2.5072, + "step": 31124 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.115265447589166e-05, + "loss": 2.5566, + "step": 31125 + }, + { + "epoch": 1.33, + "grad_norm": 0.451171875, + "learning_rate": 5.1146801387249564e-05, + "loss": 2.4664, + "step": 31126 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.1140948518429085e-05, + "loss": 2.4625, + "step": 31127 + }, + { + "epoch": 1.33, + "grad_norm": 0.453125, + "learning_rate": 5.113509586945648e-05, + "loss": 2.4766, + "step": 31128 + }, + { + "epoch": 1.33, + "grad_norm": 0.439453125, + "learning_rate": 5.112924344035805e-05, + "loss": 2.4431, + "step": 31129 + }, + { + "epoch": 1.33, + "grad_norm": 0.462890625, + "learning_rate": 5.1123391231160165e-05, + "loss": 2.4392, + "step": 31130 + }, + { + "epoch": 1.33, + "grad_norm": 0.470703125, + "learning_rate": 5.11175392418892e-05, + "loss": 2.4707, + "step": 31131 + }, + { + "epoch": 1.33, + "grad_norm": 0.453125, + "learning_rate": 5.11116874725714e-05, + "loss": 2.327, + "step": 31132 + }, + { + "epoch": 1.33, + "grad_norm": 0.470703125, + "learning_rate": 5.1105835923233166e-05, + "loss": 2.3539, + "step": 31133 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.1099984593900774e-05, + "loss": 2.3768, + "step": 31134 + }, + { + "epoch": 1.33, + "grad_norm": 0.48828125, + "learning_rate": 5.109413348460061e-05, + "loss": 2.4051, + "step": 31135 + }, + { + "epoch": 1.33, + "grad_norm": 0.48046875, + "learning_rate": 5.108828259535896e-05, + "loss": 2.4682, + "step": 31136 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.108243192620218e-05, + "loss": 2.59, + "step": 31137 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.107658147715655e-05, + "loss": 2.6255, + "step": 31138 + }, + { + "epoch": 1.33, + "grad_norm": 0.4921875, + "learning_rate": 5.107073124824841e-05, + "loss": 2.4671, + "step": 31139 + }, + { + "epoch": 1.33, + "grad_norm": 0.45703125, + "learning_rate": 5.106488123950415e-05, + "loss": 2.5224, + "step": 31140 + }, + { + "epoch": 1.33, + "grad_norm": 0.43359375, + "learning_rate": 5.105903145095003e-05, + "loss": 2.4706, + "step": 31141 + }, + { + "epoch": 1.33, + "grad_norm": 0.48046875, + "learning_rate": 5.1053181882612325e-05, + "loss": 2.3916, + "step": 31142 + }, + { + "epoch": 1.33, + "grad_norm": 0.44140625, + "learning_rate": 5.104733253451747e-05, + "loss": 2.5236, + "step": 31143 + }, + { + "epoch": 1.33, + "grad_norm": 0.48046875, + "learning_rate": 5.1041483406691725e-05, + "loss": 2.3347, + "step": 31144 + }, + { + "epoch": 1.33, + "grad_norm": 0.423828125, + "learning_rate": 5.103563449916139e-05, + "loss": 2.4705, + "step": 31145 + }, + { + "epoch": 1.33, + "grad_norm": 0.50390625, + "learning_rate": 5.10297858119528e-05, + "loss": 2.5469, + "step": 31146 + }, + { + "epoch": 1.33, + "grad_norm": 0.451171875, + "learning_rate": 5.102393734509232e-05, + "loss": 2.4148, + "step": 31147 + }, + { + "epoch": 1.33, + "grad_norm": 0.53125, + "learning_rate": 5.101808909860617e-05, + "loss": 2.4977, + "step": 31148 + }, + { + "epoch": 1.33, + "grad_norm": 0.5, + "learning_rate": 5.101224107252077e-05, + "loss": 2.4393, + "step": 31149 + }, + { + "epoch": 1.33, + "grad_norm": 0.42578125, + "learning_rate": 5.100639326686234e-05, + "loss": 2.2712, + "step": 31150 + }, + { + "epoch": 1.33, + "grad_norm": 0.48046875, + "learning_rate": 5.100054568165724e-05, + "loss": 2.5215, + "step": 31151 + }, + { + "epoch": 1.33, + "grad_norm": 0.443359375, + "learning_rate": 5.0994698316931813e-05, + "loss": 2.5045, + "step": 31152 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.0988851172712324e-05, + "loss": 2.4849, + "step": 31153 + }, + { + "epoch": 1.33, + "grad_norm": 0.46875, + "learning_rate": 5.098300424902507e-05, + "loss": 2.5298, + "step": 31154 + }, + { + "epoch": 1.33, + "grad_norm": 0.5234375, + "learning_rate": 5.097715754589638e-05, + "loss": 2.3877, + "step": 31155 + }, + { + "epoch": 1.33, + "grad_norm": 0.45703125, + "learning_rate": 5.09713110633526e-05, + "loss": 2.4829, + "step": 31156 + }, + { + "epoch": 1.33, + "grad_norm": 0.4765625, + "learning_rate": 5.096546480141999e-05, + "loss": 2.2984, + "step": 31157 + }, + { + "epoch": 1.33, + "grad_norm": 0.443359375, + "learning_rate": 5.095961876012481e-05, + "loss": 2.5037, + "step": 31158 + }, + { + "epoch": 1.33, + "grad_norm": 0.40234375, + "learning_rate": 5.0953772939493495e-05, + "loss": 2.5365, + "step": 31159 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.0947927339552274e-05, + "loss": 2.4997, + "step": 31160 + }, + { + "epoch": 1.33, + "grad_norm": 0.423828125, + "learning_rate": 5.0942081960327415e-05, + "loss": 2.4195, + "step": 31161 + }, + { + "epoch": 1.33, + "grad_norm": 0.48046875, + "learning_rate": 5.0936236801845305e-05, + "loss": 2.4358, + "step": 31162 + }, + { + "epoch": 1.33, + "grad_norm": 0.42578125, + "learning_rate": 5.0930391864132155e-05, + "loss": 2.4347, + "step": 31163 + }, + { + "epoch": 1.33, + "grad_norm": 0.498046875, + "learning_rate": 5.0924547147214305e-05, + "loss": 2.4756, + "step": 31164 + }, + { + "epoch": 1.33, + "grad_norm": 0.48828125, + "learning_rate": 5.09187026511181e-05, + "loss": 2.4847, + "step": 31165 + }, + { + "epoch": 1.33, + "grad_norm": 0.443359375, + "learning_rate": 5.0912858375869796e-05, + "loss": 2.4068, + "step": 31166 + }, + { + "epoch": 1.33, + "grad_norm": 0.45703125, + "learning_rate": 5.0907014321495615e-05, + "loss": 2.5534, + "step": 31167 + }, + { + "epoch": 1.33, + "grad_norm": 0.4921875, + "learning_rate": 5.090117048802201e-05, + "loss": 2.6298, + "step": 31168 + }, + { + "epoch": 1.33, + "grad_norm": 0.49609375, + "learning_rate": 5.089532687547518e-05, + "loss": 2.5854, + "step": 31169 + }, + { + "epoch": 1.33, + "grad_norm": 0.474609375, + "learning_rate": 5.08894834838814e-05, + "loss": 2.4569, + "step": 31170 + }, + { + "epoch": 1.33, + "grad_norm": 0.453125, + "learning_rate": 5.0883640313267e-05, + "loss": 2.5783, + "step": 31171 + }, + { + "epoch": 1.33, + "grad_norm": 0.478515625, + "learning_rate": 5.0877797363658296e-05, + "loss": 2.4597, + "step": 31172 + }, + { + "epoch": 1.33, + "grad_norm": 0.44140625, + "learning_rate": 5.0871954635081564e-05, + "loss": 2.3463, + "step": 31173 + }, + { + "epoch": 1.33, + "grad_norm": 0.4453125, + "learning_rate": 5.086611212756305e-05, + "loss": 2.4787, + "step": 31174 + }, + { + "epoch": 1.33, + "grad_norm": 0.50390625, + "learning_rate": 5.086026984112906e-05, + "loss": 2.4932, + "step": 31175 + }, + { + "epoch": 1.33, + "grad_norm": 0.482421875, + "learning_rate": 5.085442777580593e-05, + "loss": 2.3961, + "step": 31176 + }, + { + "epoch": 1.33, + "grad_norm": 0.48046875, + "learning_rate": 5.08485859316199e-05, + "loss": 2.4935, + "step": 31177 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.08427443085973e-05, + "loss": 2.4247, + "step": 31178 + }, + { + "epoch": 1.33, + "grad_norm": 0.453125, + "learning_rate": 5.083690290676434e-05, + "loss": 2.4976, + "step": 31179 + }, + { + "epoch": 1.33, + "grad_norm": 0.470703125, + "learning_rate": 5.0831061726147355e-05, + "loss": 2.3223, + "step": 31180 + }, + { + "epoch": 1.33, + "grad_norm": 0.48828125, + "learning_rate": 5.082522076677265e-05, + "loss": 2.4894, + "step": 31181 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.08193800286665e-05, + "loss": 2.3655, + "step": 31182 + }, + { + "epoch": 1.33, + "grad_norm": 0.4296875, + "learning_rate": 5.081353951185508e-05, + "loss": 2.3651, + "step": 31183 + }, + { + "epoch": 1.33, + "grad_norm": 0.427734375, + "learning_rate": 5.080769921636484e-05, + "loss": 2.3811, + "step": 31184 + }, + { + "epoch": 1.33, + "grad_norm": 0.451171875, + "learning_rate": 5.0801859142221955e-05, + "loss": 2.2253, + "step": 31185 + }, + { + "epoch": 1.33, + "grad_norm": 0.421875, + "learning_rate": 5.07960192894527e-05, + "loss": 2.3647, + "step": 31186 + }, + { + "epoch": 1.33, + "grad_norm": 0.5546875, + "learning_rate": 5.079017965808338e-05, + "loss": 2.4181, + "step": 31187 + }, + { + "epoch": 1.33, + "grad_norm": 0.404296875, + "learning_rate": 5.078434024814029e-05, + "loss": 2.4558, + "step": 31188 + }, + { + "epoch": 1.33, + "grad_norm": 0.5078125, + "learning_rate": 5.07785010596497e-05, + "loss": 2.5713, + "step": 31189 + }, + { + "epoch": 1.33, + "grad_norm": 0.474609375, + "learning_rate": 5.0772662092637824e-05, + "loss": 2.4197, + "step": 31190 + }, + { + "epoch": 1.33, + "grad_norm": 0.46875, + "learning_rate": 5.0766823347130974e-05, + "loss": 2.4933, + "step": 31191 + }, + { + "epoch": 1.33, + "grad_norm": 0.5234375, + "learning_rate": 5.0760984823155475e-05, + "loss": 2.5111, + "step": 31192 + }, + { + "epoch": 1.33, + "grad_norm": 0.5234375, + "learning_rate": 5.0755146520737496e-05, + "loss": 2.4737, + "step": 31193 + }, + { + "epoch": 1.33, + "grad_norm": 0.44921875, + "learning_rate": 5.074930843990341e-05, + "loss": 2.5352, + "step": 31194 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.0743470580679396e-05, + "loss": 2.4983, + "step": 31195 + }, + { + "epoch": 1.33, + "grad_norm": 0.5234375, + "learning_rate": 5.073763294309176e-05, + "loss": 2.4425, + "step": 31196 + }, + { + "epoch": 1.33, + "grad_norm": 0.498046875, + "learning_rate": 5.073179552716681e-05, + "loss": 2.382, + "step": 31197 + }, + { + "epoch": 1.33, + "grad_norm": 0.46484375, + "learning_rate": 5.072595833293079e-05, + "loss": 2.3677, + "step": 31198 + }, + { + "epoch": 1.33, + "grad_norm": 0.5078125, + "learning_rate": 5.072012136040984e-05, + "loss": 2.4312, + "step": 31199 + }, + { + "epoch": 1.33, + "grad_norm": 0.490234375, + "learning_rate": 5.071428460963045e-05, + "loss": 2.5163, + "step": 31200 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.070844808061874e-05, + "loss": 2.4236, + "step": 31201 + }, + { + "epoch": 1.33, + "grad_norm": 0.4921875, + "learning_rate": 5.0702611773400964e-05, + "loss": 2.4737, + "step": 31202 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.069677568800344e-05, + "loss": 2.3731, + "step": 31203 + }, + { + "epoch": 1.33, + "grad_norm": 0.4921875, + "learning_rate": 5.069093982445243e-05, + "loss": 2.3972, + "step": 31204 + }, + { + "epoch": 1.33, + "grad_norm": 0.43359375, + "learning_rate": 5.068510418277412e-05, + "loss": 2.3084, + "step": 31205 + }, + { + "epoch": 1.33, + "grad_norm": 0.453125, + "learning_rate": 5.067926876299487e-05, + "loss": 2.5604, + "step": 31206 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.067343356514085e-05, + "loss": 2.3624, + "step": 31207 + }, + { + "epoch": 1.33, + "grad_norm": 0.50390625, + "learning_rate": 5.06675985892384e-05, + "loss": 2.5536, + "step": 31208 + }, + { + "epoch": 1.33, + "grad_norm": 0.47265625, + "learning_rate": 5.0661763835313666e-05, + "loss": 2.4662, + "step": 31209 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.065592930339302e-05, + "loss": 2.4265, + "step": 31210 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 5.065009499350262e-05, + "loss": 2.4188, + "step": 31211 + }, + { + "epoch": 1.33, + "grad_norm": 0.4375, + "learning_rate": 5.064426090566874e-05, + "loss": 2.3833, + "step": 31212 + }, + { + "epoch": 1.33, + "grad_norm": 0.46484375, + "learning_rate": 5.06384270399177e-05, + "loss": 2.5028, + "step": 31213 + }, + { + "epoch": 1.33, + "grad_norm": 0.451171875, + "learning_rate": 5.06325933962757e-05, + "loss": 2.3749, + "step": 31214 + }, + { + "epoch": 1.33, + "grad_norm": 0.451171875, + "learning_rate": 5.06267599747689e-05, + "loss": 2.6472, + "step": 31215 + }, + { + "epoch": 1.33, + "grad_norm": 0.419921875, + "learning_rate": 5.062092677542372e-05, + "loss": 2.5093, + "step": 31216 + }, + { + "epoch": 1.33, + "grad_norm": 0.49609375, + "learning_rate": 5.061509379826632e-05, + "loss": 2.3456, + "step": 31217 + }, + { + "epoch": 1.33, + "grad_norm": 0.478515625, + "learning_rate": 5.060926104332292e-05, + "loss": 2.4495, + "step": 31218 + }, + { + "epoch": 1.33, + "grad_norm": 0.48046875, + "learning_rate": 5.0603428510619786e-05, + "loss": 2.4334, + "step": 31219 + }, + { + "epoch": 1.33, + "grad_norm": 0.462890625, + "learning_rate": 5.059759620018322e-05, + "loss": 2.382, + "step": 31220 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.0591764112039366e-05, + "loss": 2.5069, + "step": 31221 + }, + { + "epoch": 1.33, + "grad_norm": 0.453125, + "learning_rate": 5.0585932246214575e-05, + "loss": 2.4305, + "step": 31222 + }, + { + "epoch": 1.33, + "grad_norm": 0.5625, + "learning_rate": 5.058010060273497e-05, + "loss": 2.6071, + "step": 31223 + }, + { + "epoch": 1.33, + "grad_norm": 0.46484375, + "learning_rate": 5.05742691816269e-05, + "loss": 2.5167, + "step": 31224 + }, + { + "epoch": 1.33, + "grad_norm": 0.4765625, + "learning_rate": 5.05684379829165e-05, + "loss": 2.456, + "step": 31225 + }, + { + "epoch": 1.33, + "grad_norm": 0.5, + "learning_rate": 5.0562607006630135e-05, + "loss": 2.3894, + "step": 31226 + }, + { + "epoch": 1.33, + "grad_norm": 0.46484375, + "learning_rate": 5.055677625279391e-05, + "loss": 2.5606, + "step": 31227 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.055094572143413e-05, + "loss": 2.4014, + "step": 31228 + }, + { + "epoch": 1.33, + "grad_norm": 0.478515625, + "learning_rate": 5.054511541257706e-05, + "loss": 2.4728, + "step": 31229 + }, + { + "epoch": 1.33, + "grad_norm": 0.44140625, + "learning_rate": 5.05392853262489e-05, + "loss": 2.3578, + "step": 31230 + }, + { + "epoch": 1.33, + "grad_norm": 0.470703125, + "learning_rate": 5.0533455462475785e-05, + "loss": 2.4816, + "step": 31231 + }, + { + "epoch": 1.33, + "grad_norm": 0.412109375, + "learning_rate": 5.0527625821284144e-05, + "loss": 2.3564, + "step": 31232 + }, + { + "epoch": 1.33, + "grad_norm": 0.43359375, + "learning_rate": 5.0521796402700094e-05, + "loss": 2.5661, + "step": 31233 + }, + { + "epoch": 1.33, + "grad_norm": 0.53515625, + "learning_rate": 5.0515967206749845e-05, + "loss": 2.6724, + "step": 31234 + }, + { + "epoch": 1.33, + "grad_norm": 0.42578125, + "learning_rate": 5.051013823345965e-05, + "loss": 2.4765, + "step": 31235 + }, + { + "epoch": 1.33, + "grad_norm": 0.458984375, + "learning_rate": 5.0504309482855795e-05, + "loss": 2.5074, + "step": 31236 + }, + { + "epoch": 1.33, + "grad_norm": 0.50390625, + "learning_rate": 5.049848095496441e-05, + "loss": 2.414, + "step": 31237 + }, + { + "epoch": 1.33, + "grad_norm": 0.5078125, + "learning_rate": 5.049265264981182e-05, + "loss": 2.4199, + "step": 31238 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.048682456742415e-05, + "loss": 2.2571, + "step": 31239 + }, + { + "epoch": 1.33, + "grad_norm": 0.49609375, + "learning_rate": 5.048099670782773e-05, + "loss": 2.4347, + "step": 31240 + }, + { + "epoch": 1.33, + "grad_norm": 0.515625, + "learning_rate": 5.047516907104868e-05, + "loss": 2.5633, + "step": 31241 + }, + { + "epoch": 1.33, + "grad_norm": 0.46484375, + "learning_rate": 5.04693416571133e-05, + "loss": 2.3363, + "step": 31242 + }, + { + "epoch": 1.33, + "grad_norm": 0.45703125, + "learning_rate": 5.0463514466047754e-05, + "loss": 2.5496, + "step": 31243 + }, + { + "epoch": 1.33, + "grad_norm": 0.474609375, + "learning_rate": 5.045768749787828e-05, + "loss": 2.5457, + "step": 31244 + }, + { + "epoch": 1.33, + "grad_norm": 0.46484375, + "learning_rate": 5.0451860752631154e-05, + "loss": 2.4607, + "step": 31245 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.044603423033254e-05, + "loss": 2.4173, + "step": 31246 + }, + { + "epoch": 1.33, + "grad_norm": 0.48828125, + "learning_rate": 5.0440207931008584e-05, + "loss": 2.5435, + "step": 31247 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.043438185468567e-05, + "loss": 2.397, + "step": 31248 + }, + { + "epoch": 1.33, + "grad_norm": 0.490234375, + "learning_rate": 5.042855600138991e-05, + "loss": 2.5368, + "step": 31249 + }, + { + "epoch": 1.33, + "grad_norm": 0.5078125, + "learning_rate": 5.04227303711475e-05, + "loss": 2.5624, + "step": 31250 + }, + { + "epoch": 1.33, + "grad_norm": 0.49609375, + "learning_rate": 5.0416904963984676e-05, + "loss": 2.5713, + "step": 31251 + }, + { + "epoch": 1.33, + "grad_norm": 0.482421875, + "learning_rate": 5.0411079779927716e-05, + "loss": 2.5627, + "step": 31252 + }, + { + "epoch": 1.33, + "grad_norm": 0.5078125, + "learning_rate": 5.0405254819002714e-05, + "loss": 2.4052, + "step": 31253 + }, + { + "epoch": 1.33, + "grad_norm": 0.47265625, + "learning_rate": 5.0399430081236e-05, + "loss": 2.494, + "step": 31254 + }, + { + "epoch": 1.33, + "grad_norm": 0.470703125, + "learning_rate": 5.039360556665367e-05, + "loss": 2.5325, + "step": 31255 + }, + { + "epoch": 1.33, + "grad_norm": 0.515625, + "learning_rate": 5.038778127528203e-05, + "loss": 2.5868, + "step": 31256 + }, + { + "epoch": 1.33, + "grad_norm": 0.45703125, + "learning_rate": 5.03819572071472e-05, + "loss": 2.4583, + "step": 31257 + }, + { + "epoch": 1.33, + "grad_norm": 0.4765625, + "learning_rate": 5.037613336227547e-05, + "loss": 2.3946, + "step": 31258 + }, + { + "epoch": 1.33, + "grad_norm": 0.404296875, + "learning_rate": 5.037030974069297e-05, + "loss": 2.453, + "step": 31259 + }, + { + "epoch": 1.33, + "grad_norm": 0.451171875, + "learning_rate": 5.0364486342425934e-05, + "loss": 2.4612, + "step": 31260 + }, + { + "epoch": 1.33, + "grad_norm": 0.48828125, + "learning_rate": 5.035866316750061e-05, + "loss": 2.4729, + "step": 31261 + }, + { + "epoch": 1.33, + "grad_norm": 0.474609375, + "learning_rate": 5.0352840215943155e-05, + "loss": 2.3272, + "step": 31262 + }, + { + "epoch": 1.33, + "grad_norm": 0.435546875, + "learning_rate": 5.03470174877797e-05, + "loss": 2.3669, + "step": 31263 + }, + { + "epoch": 1.33, + "grad_norm": 0.458984375, + "learning_rate": 5.034119498303661e-05, + "loss": 2.4835, + "step": 31264 + }, + { + "epoch": 1.33, + "grad_norm": 0.498046875, + "learning_rate": 5.033537270173998e-05, + "loss": 2.5887, + "step": 31265 + }, + { + "epoch": 1.33, + "grad_norm": 0.4921875, + "learning_rate": 5.032955064391599e-05, + "loss": 2.5407, + "step": 31266 + }, + { + "epoch": 1.33, + "grad_norm": 0.48828125, + "learning_rate": 5.032372880959086e-05, + "loss": 2.4502, + "step": 31267 + }, + { + "epoch": 1.33, + "grad_norm": 0.4140625, + "learning_rate": 5.031790719879086e-05, + "loss": 2.3004, + "step": 31268 + }, + { + "epoch": 1.33, + "grad_norm": 0.470703125, + "learning_rate": 5.031208581154205e-05, + "loss": 2.4932, + "step": 31269 + }, + { + "epoch": 1.33, + "grad_norm": 0.455078125, + "learning_rate": 5.030626464787076e-05, + "loss": 2.3282, + "step": 31270 + }, + { + "epoch": 1.33, + "grad_norm": 0.470703125, + "learning_rate": 5.030044370780307e-05, + "loss": 2.4415, + "step": 31271 + }, + { + "epoch": 1.33, + "grad_norm": 0.43359375, + "learning_rate": 5.0294622991365245e-05, + "loss": 2.5409, + "step": 31272 + }, + { + "epoch": 1.33, + "grad_norm": 0.486328125, + "learning_rate": 5.028880249858343e-05, + "loss": 2.4574, + "step": 31273 + }, + { + "epoch": 1.33, + "grad_norm": 0.48046875, + "learning_rate": 5.028298222948387e-05, + "loss": 2.5676, + "step": 31274 + }, + { + "epoch": 1.33, + "grad_norm": 0.44140625, + "learning_rate": 5.027716218409269e-05, + "loss": 2.5762, + "step": 31275 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.027134236243609e-05, + "loss": 2.4673, + "step": 31276 + }, + { + "epoch": 1.33, + "grad_norm": 0.5234375, + "learning_rate": 5.026552276454033e-05, + "loss": 2.5137, + "step": 31277 + }, + { + "epoch": 1.33, + "grad_norm": 0.494140625, + "learning_rate": 5.0259703390431534e-05, + "loss": 2.5207, + "step": 31278 + }, + { + "epoch": 1.33, + "grad_norm": 0.380859375, + "learning_rate": 5.025388424013582e-05, + "loss": 2.4361, + "step": 31279 + }, + { + "epoch": 1.33, + "grad_norm": 0.42578125, + "learning_rate": 5.024806531367952e-05, + "loss": 2.5337, + "step": 31280 + }, + { + "epoch": 1.33, + "grad_norm": 0.490234375, + "learning_rate": 5.024224661108874e-05, + "loss": 2.4468, + "step": 31281 + }, + { + "epoch": 1.33, + "grad_norm": 0.490234375, + "learning_rate": 5.0236428132389646e-05, + "loss": 2.4711, + "step": 31282 + }, + { + "epoch": 1.33, + "grad_norm": 0.48828125, + "learning_rate": 5.0230609877608456e-05, + "loss": 2.569, + "step": 31283 + }, + { + "epoch": 1.33, + "grad_norm": 0.482421875, + "learning_rate": 5.022479184677131e-05, + "loss": 2.5992, + "step": 31284 + }, + { + "epoch": 1.33, + "grad_norm": 0.427734375, + "learning_rate": 5.02189740399044e-05, + "loss": 2.5329, + "step": 31285 + }, + { + "epoch": 1.33, + "grad_norm": 0.421875, + "learning_rate": 5.021315645703395e-05, + "loss": 2.4637, + "step": 31286 + }, + { + "epoch": 1.33, + "grad_norm": 0.5078125, + "learning_rate": 5.020733909818611e-05, + "loss": 2.438, + "step": 31287 + }, + { + "epoch": 1.33, + "grad_norm": 0.427734375, + "learning_rate": 5.020152196338701e-05, + "loss": 2.3681, + "step": 31288 + }, + { + "epoch": 1.33, + "grad_norm": 0.4296875, + "learning_rate": 5.0195705052662855e-05, + "loss": 2.4124, + "step": 31289 + }, + { + "epoch": 1.33, + "grad_norm": 0.5234375, + "learning_rate": 5.018988836603987e-05, + "loss": 2.5731, + "step": 31290 + }, + { + "epoch": 1.33, + "grad_norm": 0.490234375, + "learning_rate": 5.0184071903544146e-05, + "loss": 2.4583, + "step": 31291 + }, + { + "epoch": 1.33, + "grad_norm": 0.47265625, + "learning_rate": 5.017825566520189e-05, + "loss": 2.5149, + "step": 31292 + }, + { + "epoch": 1.33, + "grad_norm": 0.515625, + "learning_rate": 5.01724396510393e-05, + "loss": 2.4339, + "step": 31293 + }, + { + "epoch": 1.33, + "grad_norm": 0.51953125, + "learning_rate": 5.016662386108255e-05, + "loss": 2.5031, + "step": 31294 + }, + { + "epoch": 1.33, + "grad_norm": 0.458984375, + "learning_rate": 5.016080829535772e-05, + "loss": 2.4729, + "step": 31295 + }, + { + "epoch": 1.33, + "grad_norm": 0.47265625, + "learning_rate": 5.0154992953891054e-05, + "loss": 2.4458, + "step": 31296 + }, + { + "epoch": 1.33, + "grad_norm": 0.484375, + "learning_rate": 5.014917783670873e-05, + "loss": 2.3773, + "step": 31297 + }, + { + "epoch": 1.33, + "grad_norm": 0.427734375, + "learning_rate": 5.014336294383686e-05, + "loss": 2.4788, + "step": 31298 + }, + { + "epoch": 1.33, + "grad_norm": 0.47265625, + "learning_rate": 5.013754827530166e-05, + "loss": 2.4336, + "step": 31299 + }, + { + "epoch": 1.33, + "grad_norm": 0.46484375, + "learning_rate": 5.0131733831129236e-05, + "loss": 2.5743, + "step": 31300 + }, + { + "epoch": 1.33, + "grad_norm": 0.451171875, + "learning_rate": 5.0125919611345785e-05, + "loss": 2.47, + "step": 31301 + }, + { + "epoch": 1.33, + "grad_norm": 0.478515625, + "learning_rate": 5.0120105615977506e-05, + "loss": 2.4397, + "step": 31302 + }, + { + "epoch": 1.33, + "grad_norm": 0.4375, + "learning_rate": 5.011429184505054e-05, + "loss": 2.4735, + "step": 31303 + }, + { + "epoch": 1.33, + "grad_norm": 0.46875, + "learning_rate": 5.0108478298590975e-05, + "loss": 2.4425, + "step": 31304 + }, + { + "epoch": 1.33, + "grad_norm": 0.5390625, + "learning_rate": 5.010266497662502e-05, + "loss": 2.5061, + "step": 31305 + }, + { + "epoch": 1.33, + "grad_norm": 0.484375, + "learning_rate": 5.0096851879178895e-05, + "loss": 2.4505, + "step": 31306 + }, + { + "epoch": 1.33, + "grad_norm": 0.4453125, + "learning_rate": 5.009103900627865e-05, + "loss": 2.4004, + "step": 31307 + }, + { + "epoch": 1.33, + "grad_norm": 0.52734375, + "learning_rate": 5.0085226357950476e-05, + "loss": 2.4783, + "step": 31308 + }, + { + "epoch": 1.33, + "grad_norm": 0.4765625, + "learning_rate": 5.00794139342206e-05, + "loss": 2.5236, + "step": 31309 + }, + { + "epoch": 1.33, + "grad_norm": 0.478515625, + "learning_rate": 5.0073601735115105e-05, + "loss": 2.4704, + "step": 31310 + }, + { + "epoch": 1.33, + "grad_norm": 0.4765625, + "learning_rate": 5.0067789760660136e-05, + "loss": 2.4856, + "step": 31311 + }, + { + "epoch": 1.33, + "grad_norm": 0.478515625, + "learning_rate": 5.0061978010881835e-05, + "loss": 2.3586, + "step": 31312 + }, + { + "epoch": 1.33, + "grad_norm": 0.451171875, + "learning_rate": 5.005616648580645e-05, + "loss": 2.5079, + "step": 31313 + }, + { + "epoch": 1.33, + "grad_norm": 0.453125, + "learning_rate": 5.0050355185460006e-05, + "loss": 2.4927, + "step": 31314 + }, + { + "epoch": 1.33, + "grad_norm": 0.439453125, + "learning_rate": 5.0044544109868764e-05, + "loss": 2.2927, + "step": 31315 + }, + { + "epoch": 1.33, + "grad_norm": 0.4453125, + "learning_rate": 5.003873325905877e-05, + "loss": 2.4555, + "step": 31316 + }, + { + "epoch": 1.33, + "grad_norm": 0.44140625, + "learning_rate": 5.003292263305621e-05, + "loss": 2.4822, + "step": 31317 + }, + { + "epoch": 1.33, + "grad_norm": 0.41015625, + "learning_rate": 5.002711223188728e-05, + "loss": 2.5166, + "step": 31318 + }, + { + "epoch": 1.33, + "grad_norm": 0.458984375, + "learning_rate": 5.002130205557809e-05, + "loss": 2.3897, + "step": 31319 + }, + { + "epoch": 1.33, + "grad_norm": 0.470703125, + "learning_rate": 5.001549210415473e-05, + "loss": 2.4076, + "step": 31320 + }, + { + "epoch": 1.33, + "grad_norm": 0.43359375, + "learning_rate": 5.0009682377643386e-05, + "loss": 2.413, + "step": 31321 + }, + { + "epoch": 1.33, + "grad_norm": 0.5, + "learning_rate": 5.000387287607025e-05, + "loss": 2.4276, + "step": 31322 + }, + { + "epoch": 1.33, + "grad_norm": 0.4609375, + "learning_rate": 4.999806359946136e-05, + "loss": 2.4149, + "step": 31323 + }, + { + "epoch": 1.33, + "grad_norm": 0.5078125, + "learning_rate": 4.999225454784292e-05, + "loss": 2.3874, + "step": 31324 + }, + { + "epoch": 1.33, + "grad_norm": 0.52734375, + "learning_rate": 4.9986445721241104e-05, + "loss": 2.5537, + "step": 31325 + }, + { + "epoch": 1.33, + "grad_norm": 0.462890625, + "learning_rate": 4.9980637119681996e-05, + "loss": 2.4349, + "step": 31326 + }, + { + "epoch": 1.33, + "grad_norm": 0.51171875, + "learning_rate": 4.997482874319171e-05, + "loss": 2.4615, + "step": 31327 + }, + { + "epoch": 1.33, + "grad_norm": 0.427734375, + "learning_rate": 4.99690205917964e-05, + "loss": 2.5583, + "step": 31328 + }, + { + "epoch": 1.34, + "grad_norm": 0.41015625, + "learning_rate": 4.9963212665522266e-05, + "loss": 2.4748, + "step": 31329 + }, + { + "epoch": 1.34, + "grad_norm": 0.53515625, + "learning_rate": 4.995740496439534e-05, + "loss": 2.4928, + "step": 31330 + }, + { + "epoch": 1.34, + "grad_norm": 0.443359375, + "learning_rate": 4.995159748844184e-05, + "loss": 2.4281, + "step": 31331 + }, + { + "epoch": 1.34, + "grad_norm": 0.431640625, + "learning_rate": 4.994579023768783e-05, + "loss": 2.4678, + "step": 31332 + }, + { + "epoch": 1.34, + "grad_norm": 0.451171875, + "learning_rate": 4.9939983212159466e-05, + "loss": 2.3603, + "step": 31333 + }, + { + "epoch": 1.34, + "grad_norm": 0.439453125, + "learning_rate": 4.9934176411882925e-05, + "loss": 2.4844, + "step": 31334 + }, + { + "epoch": 1.34, + "grad_norm": 0.4296875, + "learning_rate": 4.992836983688429e-05, + "loss": 2.4351, + "step": 31335 + }, + { + "epoch": 1.34, + "grad_norm": 0.458984375, + "learning_rate": 4.9922563487189656e-05, + "loss": 2.4705, + "step": 31336 + }, + { + "epoch": 1.34, + "grad_norm": 0.4609375, + "learning_rate": 4.9916757362825186e-05, + "loss": 2.5061, + "step": 31337 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.991095146381704e-05, + "loss": 2.5385, + "step": 31338 + }, + { + "epoch": 1.34, + "grad_norm": 0.53125, + "learning_rate": 4.9905145790191264e-05, + "loss": 2.4557, + "step": 31339 + }, + { + "epoch": 1.34, + "grad_norm": 0.494140625, + "learning_rate": 4.989934034197403e-05, + "loss": 2.4122, + "step": 31340 + }, + { + "epoch": 1.34, + "grad_norm": 0.462890625, + "learning_rate": 4.989353511919149e-05, + "loss": 2.3764, + "step": 31341 + }, + { + "epoch": 1.34, + "grad_norm": 0.515625, + "learning_rate": 4.9887730121869734e-05, + "loss": 2.4276, + "step": 31342 + }, + { + "epoch": 1.34, + "grad_norm": 0.455078125, + "learning_rate": 4.9881925350034843e-05, + "loss": 2.4441, + "step": 31343 + }, + { + "epoch": 1.34, + "grad_norm": 0.51953125, + "learning_rate": 4.987612080371297e-05, + "loss": 2.4689, + "step": 31344 + }, + { + "epoch": 1.34, + "grad_norm": 0.4453125, + "learning_rate": 4.987031648293027e-05, + "loss": 2.5275, + "step": 31345 + }, + { + "epoch": 1.34, + "grad_norm": 0.5, + "learning_rate": 4.986451238771278e-05, + "loss": 2.4892, + "step": 31346 + }, + { + "epoch": 1.34, + "grad_norm": 0.515625, + "learning_rate": 4.985870851808672e-05, + "loss": 2.5199, + "step": 31347 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.985290487407809e-05, + "loss": 2.6186, + "step": 31348 + }, + { + "epoch": 1.34, + "grad_norm": 0.416015625, + "learning_rate": 4.9847101455713065e-05, + "loss": 2.3788, + "step": 31349 + }, + { + "epoch": 1.34, + "grad_norm": 0.4375, + "learning_rate": 4.98412982630178e-05, + "loss": 2.4722, + "step": 31350 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.9835495296018364e-05, + "loss": 2.6339, + "step": 31351 + }, + { + "epoch": 1.34, + "grad_norm": 0.486328125, + "learning_rate": 4.9829692554740817e-05, + "loss": 2.344, + "step": 31352 + }, + { + "epoch": 1.34, + "grad_norm": 0.478515625, + "learning_rate": 4.982389003921133e-05, + "loss": 2.506, + "step": 31353 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.981808774945602e-05, + "loss": 2.4831, + "step": 31354 + }, + { + "epoch": 1.34, + "grad_norm": 0.4765625, + "learning_rate": 4.981228568550096e-05, + "loss": 2.4431, + "step": 31355 + }, + { + "epoch": 1.34, + "grad_norm": 0.447265625, + "learning_rate": 4.980648384737227e-05, + "loss": 2.2903, + "step": 31356 + }, + { + "epoch": 1.34, + "grad_norm": 0.474609375, + "learning_rate": 4.9800682235096085e-05, + "loss": 2.4651, + "step": 31357 + }, + { + "epoch": 1.34, + "grad_norm": 0.46875, + "learning_rate": 4.979488084869849e-05, + "loss": 2.508, + "step": 31358 + }, + { + "epoch": 1.34, + "grad_norm": 0.498046875, + "learning_rate": 4.9789079688205545e-05, + "loss": 2.5114, + "step": 31359 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.978327875364338e-05, + "loss": 2.3442, + "step": 31360 + }, + { + "epoch": 1.34, + "grad_norm": 0.4765625, + "learning_rate": 4.977747804503816e-05, + "loss": 2.3931, + "step": 31361 + }, + { + "epoch": 1.34, + "grad_norm": 0.447265625, + "learning_rate": 4.97716775624159e-05, + "loss": 2.3434, + "step": 31362 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.976587730580277e-05, + "loss": 2.3855, + "step": 31363 + }, + { + "epoch": 1.34, + "grad_norm": 0.478515625, + "learning_rate": 4.976007727522479e-05, + "loss": 2.5891, + "step": 31364 + }, + { + "epoch": 1.34, + "grad_norm": 0.51171875, + "learning_rate": 4.97542774707081e-05, + "loss": 2.4075, + "step": 31365 + }, + { + "epoch": 1.34, + "grad_norm": 0.455078125, + "learning_rate": 4.974847789227884e-05, + "loss": 2.3542, + "step": 31366 + }, + { + "epoch": 1.34, + "grad_norm": 0.490234375, + "learning_rate": 4.9742678539963074e-05, + "loss": 2.4163, + "step": 31367 + }, + { + "epoch": 1.34, + "grad_norm": 0.47265625, + "learning_rate": 4.973687941378683e-05, + "loss": 2.537, + "step": 31368 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.973108051377627e-05, + "loss": 2.5447, + "step": 31369 + }, + { + "epoch": 1.34, + "grad_norm": 0.435546875, + "learning_rate": 4.9725281839957516e-05, + "loss": 2.3479, + "step": 31370 + }, + { + "epoch": 1.34, + "grad_norm": 0.466796875, + "learning_rate": 4.971948339235659e-05, + "loss": 2.4625, + "step": 31371 + }, + { + "epoch": 1.34, + "grad_norm": 0.44140625, + "learning_rate": 4.9713685170999605e-05, + "loss": 2.3639, + "step": 31372 + }, + { + "epoch": 1.34, + "grad_norm": 0.478515625, + "learning_rate": 4.9707887175912694e-05, + "loss": 2.6438, + "step": 31373 + }, + { + "epoch": 1.34, + "grad_norm": 0.50390625, + "learning_rate": 4.970208940712191e-05, + "loss": 2.4706, + "step": 31374 + }, + { + "epoch": 1.34, + "grad_norm": 0.453125, + "learning_rate": 4.969629186465331e-05, + "loss": 2.4845, + "step": 31375 + }, + { + "epoch": 1.34, + "grad_norm": 0.47265625, + "learning_rate": 4.969049454853302e-05, + "loss": 2.5552, + "step": 31376 + }, + { + "epoch": 1.34, + "grad_norm": 0.484375, + "learning_rate": 4.968469745878714e-05, + "loss": 2.4217, + "step": 31377 + }, + { + "epoch": 1.34, + "grad_norm": 0.4765625, + "learning_rate": 4.96789005954417e-05, + "loss": 2.5104, + "step": 31378 + }, + { + "epoch": 1.34, + "grad_norm": 0.404296875, + "learning_rate": 4.967310395852286e-05, + "loss": 2.3482, + "step": 31379 + }, + { + "epoch": 1.34, + "grad_norm": 0.5, + "learning_rate": 4.966730754805662e-05, + "loss": 2.5733, + "step": 31380 + }, + { + "epoch": 1.34, + "grad_norm": 0.380859375, + "learning_rate": 4.96615113640691e-05, + "loss": 2.2748, + "step": 31381 + }, + { + "epoch": 1.34, + "grad_norm": 0.470703125, + "learning_rate": 4.965571540658642e-05, + "loss": 2.4864, + "step": 31382 + }, + { + "epoch": 1.34, + "grad_norm": 0.46875, + "learning_rate": 4.964991967563462e-05, + "loss": 2.5072, + "step": 31383 + }, + { + "epoch": 1.34, + "grad_norm": 0.50390625, + "learning_rate": 4.9644124171239745e-05, + "loss": 2.5132, + "step": 31384 + }, + { + "epoch": 1.34, + "grad_norm": 0.5390625, + "learning_rate": 4.963832889342791e-05, + "loss": 2.4373, + "step": 31385 + }, + { + "epoch": 1.34, + "grad_norm": 0.447265625, + "learning_rate": 4.963253384222523e-05, + "loss": 2.2585, + "step": 31386 + }, + { + "epoch": 1.34, + "grad_norm": 0.45703125, + "learning_rate": 4.96267390176577e-05, + "loss": 2.4577, + "step": 31387 + }, + { + "epoch": 1.34, + "grad_norm": 0.484375, + "learning_rate": 4.962094441975144e-05, + "loss": 2.4497, + "step": 31388 + }, + { + "epoch": 1.34, + "grad_norm": 0.5078125, + "learning_rate": 4.9615150048532544e-05, + "loss": 2.4627, + "step": 31389 + }, + { + "epoch": 1.34, + "grad_norm": 0.52734375, + "learning_rate": 4.960935590402707e-05, + "loss": 2.4238, + "step": 31390 + }, + { + "epoch": 1.34, + "grad_norm": 0.48828125, + "learning_rate": 4.960356198626104e-05, + "loss": 2.4308, + "step": 31391 + }, + { + "epoch": 1.34, + "grad_norm": 0.419921875, + "learning_rate": 4.959776829526056e-05, + "loss": 2.5129, + "step": 31392 + }, + { + "epoch": 1.34, + "grad_norm": 0.443359375, + "learning_rate": 4.959197483105174e-05, + "loss": 2.4767, + "step": 31393 + }, + { + "epoch": 1.34, + "grad_norm": 0.53125, + "learning_rate": 4.9586181593660576e-05, + "loss": 2.5599, + "step": 31394 + }, + { + "epoch": 1.34, + "grad_norm": 0.494140625, + "learning_rate": 4.95803885831132e-05, + "loss": 2.495, + "step": 31395 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.957459579943562e-05, + "loss": 2.3289, + "step": 31396 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.9568803242653927e-05, + "loss": 2.4276, + "step": 31397 + }, + { + "epoch": 1.34, + "grad_norm": 0.482421875, + "learning_rate": 4.956301091279423e-05, + "loss": 2.4689, + "step": 31398 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.955721880988255e-05, + "loss": 2.2713, + "step": 31399 + }, + { + "epoch": 1.34, + "grad_norm": 0.43359375, + "learning_rate": 4.955142693394491e-05, + "loss": 2.3958, + "step": 31400 + }, + { + "epoch": 1.34, + "grad_norm": 0.478515625, + "learning_rate": 4.954563528500742e-05, + "loss": 2.5124, + "step": 31401 + }, + { + "epoch": 1.34, + "grad_norm": 0.51171875, + "learning_rate": 4.9539843863096155e-05, + "loss": 2.373, + "step": 31402 + }, + { + "epoch": 1.34, + "grad_norm": 0.42578125, + "learning_rate": 4.953405266823713e-05, + "loss": 2.3935, + "step": 31403 + }, + { + "epoch": 1.34, + "grad_norm": 0.47265625, + "learning_rate": 4.9528261700456455e-05, + "loss": 2.5528, + "step": 31404 + }, + { + "epoch": 1.34, + "grad_norm": 0.416015625, + "learning_rate": 4.952247095978012e-05, + "loss": 2.4418, + "step": 31405 + }, + { + "epoch": 1.34, + "grad_norm": 0.478515625, + "learning_rate": 4.951668044623426e-05, + "loss": 2.4771, + "step": 31406 + }, + { + "epoch": 1.34, + "grad_norm": 0.455078125, + "learning_rate": 4.951089015984485e-05, + "loss": 2.5009, + "step": 31407 + }, + { + "epoch": 1.34, + "grad_norm": 0.4609375, + "learning_rate": 4.950510010063801e-05, + "loss": 2.5009, + "step": 31408 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.949931026863974e-05, + "loss": 2.4652, + "step": 31409 + }, + { + "epoch": 1.34, + "grad_norm": 0.5078125, + "learning_rate": 4.949352066387612e-05, + "loss": 2.54, + "step": 31410 + }, + { + "epoch": 1.34, + "grad_norm": 0.466796875, + "learning_rate": 4.9487731286373215e-05, + "loss": 2.3832, + "step": 31411 + }, + { + "epoch": 1.34, + "grad_norm": 0.51953125, + "learning_rate": 4.9481942136157076e-05, + "loss": 2.5636, + "step": 31412 + }, + { + "epoch": 1.34, + "grad_norm": 0.4609375, + "learning_rate": 4.947615321325366e-05, + "loss": 2.4936, + "step": 31413 + }, + { + "epoch": 1.34, + "grad_norm": 0.439453125, + "learning_rate": 4.947036451768917e-05, + "loss": 2.3986, + "step": 31414 + }, + { + "epoch": 1.34, + "grad_norm": 0.5, + "learning_rate": 4.946457604948956e-05, + "loss": 2.4322, + "step": 31415 + }, + { + "epoch": 1.34, + "grad_norm": 0.474609375, + "learning_rate": 4.945878780868086e-05, + "loss": 2.5175, + "step": 31416 + }, + { + "epoch": 1.34, + "grad_norm": 0.5546875, + "learning_rate": 4.945299979528915e-05, + "loss": 2.4078, + "step": 31417 + }, + { + "epoch": 1.34, + "grad_norm": 0.498046875, + "learning_rate": 4.944721200934049e-05, + "loss": 2.5322, + "step": 31418 + }, + { + "epoch": 1.34, + "grad_norm": 0.5078125, + "learning_rate": 4.9441424450860875e-05, + "loss": 2.4972, + "step": 31419 + }, + { + "epoch": 1.34, + "grad_norm": 0.431640625, + "learning_rate": 4.943563711987641e-05, + "loss": 2.3306, + "step": 31420 + }, + { + "epoch": 1.34, + "grad_norm": 0.447265625, + "learning_rate": 4.942985001641305e-05, + "loss": 2.4457, + "step": 31421 + }, + { + "epoch": 1.34, + "grad_norm": 0.431640625, + "learning_rate": 4.9424063140496934e-05, + "loss": 2.3994, + "step": 31422 + }, + { + "epoch": 1.34, + "grad_norm": 0.439453125, + "learning_rate": 4.9418276492154e-05, + "loss": 2.4773, + "step": 31423 + }, + { + "epoch": 1.34, + "grad_norm": 0.46875, + "learning_rate": 4.941249007141039e-05, + "loss": 2.3535, + "step": 31424 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.940670387829204e-05, + "loss": 2.4706, + "step": 31425 + }, + { + "epoch": 1.34, + "grad_norm": 0.4765625, + "learning_rate": 4.940091791282504e-05, + "loss": 2.4607, + "step": 31426 + }, + { + "epoch": 1.34, + "grad_norm": 0.462890625, + "learning_rate": 4.939513217503544e-05, + "loss": 2.4179, + "step": 31427 + }, + { + "epoch": 1.34, + "grad_norm": 0.41015625, + "learning_rate": 4.938934666494927e-05, + "loss": 2.4908, + "step": 31428 + }, + { + "epoch": 1.34, + "grad_norm": 0.486328125, + "learning_rate": 4.9383561382592456e-05, + "loss": 2.7134, + "step": 31429 + }, + { + "epoch": 1.34, + "grad_norm": 0.453125, + "learning_rate": 4.9377776327991196e-05, + "loss": 2.349, + "step": 31430 + }, + { + "epoch": 1.34, + "grad_norm": 0.43359375, + "learning_rate": 4.937199150117144e-05, + "loss": 2.3844, + "step": 31431 + }, + { + "epoch": 1.34, + "grad_norm": 0.44140625, + "learning_rate": 4.9366206902159185e-05, + "loss": 2.435, + "step": 31432 + }, + { + "epoch": 1.34, + "grad_norm": 0.50390625, + "learning_rate": 4.93604225309805e-05, + "loss": 2.5872, + "step": 31433 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.9354638387661434e-05, + "loss": 2.5207, + "step": 31434 + }, + { + "epoch": 1.34, + "grad_norm": 0.451171875, + "learning_rate": 4.9348854472227955e-05, + "loss": 2.5046, + "step": 31435 + }, + { + "epoch": 1.34, + "grad_norm": 0.41796875, + "learning_rate": 4.934307078470616e-05, + "loss": 2.4648, + "step": 31436 + }, + { + "epoch": 1.34, + "grad_norm": 0.455078125, + "learning_rate": 4.9337287325121986e-05, + "loss": 2.3708, + "step": 31437 + }, + { + "epoch": 1.34, + "grad_norm": 0.4921875, + "learning_rate": 4.9331504093501546e-05, + "loss": 2.5062, + "step": 31438 + }, + { + "epoch": 1.34, + "grad_norm": 0.490234375, + "learning_rate": 4.932572108987079e-05, + "loss": 2.4633, + "step": 31439 + }, + { + "epoch": 1.34, + "grad_norm": 0.49609375, + "learning_rate": 4.9319938314255796e-05, + "loss": 2.4715, + "step": 31440 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.9314155766682534e-05, + "loss": 2.327, + "step": 31441 + }, + { + "epoch": 1.34, + "grad_norm": 0.52734375, + "learning_rate": 4.9308373447177034e-05, + "loss": 2.486, + "step": 31442 + }, + { + "epoch": 1.34, + "grad_norm": 0.474609375, + "learning_rate": 4.930259135576537e-05, + "loss": 2.4221, + "step": 31443 + }, + { + "epoch": 1.34, + "grad_norm": 0.439453125, + "learning_rate": 4.929680949247352e-05, + "loss": 2.412, + "step": 31444 + }, + { + "epoch": 1.34, + "grad_norm": 0.51171875, + "learning_rate": 4.929102785732742e-05, + "loss": 2.5555, + "step": 31445 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.928524645035324e-05, + "loss": 2.6108, + "step": 31446 + }, + { + "epoch": 1.34, + "grad_norm": 0.5078125, + "learning_rate": 4.927946527157692e-05, + "loss": 2.4241, + "step": 31447 + }, + { + "epoch": 1.34, + "grad_norm": 0.458984375, + "learning_rate": 4.927368432102443e-05, + "loss": 2.4641, + "step": 31448 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.926790359872181e-05, + "loss": 2.4089, + "step": 31449 + }, + { + "epoch": 1.34, + "grad_norm": 0.453125, + "learning_rate": 4.9262123104695136e-05, + "loss": 2.5786, + "step": 31450 + }, + { + "epoch": 1.34, + "grad_norm": 0.49609375, + "learning_rate": 4.925634283897032e-05, + "loss": 2.4583, + "step": 31451 + }, + { + "epoch": 1.34, + "grad_norm": 0.423828125, + "learning_rate": 4.925056280157346e-05, + "loss": 2.4297, + "step": 31452 + }, + { + "epoch": 1.34, + "grad_norm": 0.435546875, + "learning_rate": 4.924478299253048e-05, + "loss": 2.4398, + "step": 31453 + }, + { + "epoch": 1.34, + "grad_norm": 0.474609375, + "learning_rate": 4.923900341186747e-05, + "loss": 2.5024, + "step": 31454 + }, + { + "epoch": 1.34, + "grad_norm": 0.466796875, + "learning_rate": 4.923322405961034e-05, + "loss": 2.3848, + "step": 31455 + }, + { + "epoch": 1.34, + "grad_norm": 0.4609375, + "learning_rate": 4.9227444935785205e-05, + "loss": 2.4206, + "step": 31456 + }, + { + "epoch": 1.34, + "grad_norm": 0.451171875, + "learning_rate": 4.922166604041797e-05, + "loss": 2.438, + "step": 31457 + }, + { + "epoch": 1.34, + "grad_norm": 0.462890625, + "learning_rate": 4.9215887373534674e-05, + "loss": 2.4955, + "step": 31458 + }, + { + "epoch": 1.34, + "grad_norm": 0.462890625, + "learning_rate": 4.921010893516136e-05, + "loss": 2.4281, + "step": 31459 + }, + { + "epoch": 1.34, + "grad_norm": 0.4609375, + "learning_rate": 4.9204330725324e-05, + "loss": 2.4664, + "step": 31460 + }, + { + "epoch": 1.34, + "grad_norm": 0.486328125, + "learning_rate": 4.919855274404852e-05, + "loss": 2.4708, + "step": 31461 + }, + { + "epoch": 1.34, + "grad_norm": 0.431640625, + "learning_rate": 4.9192774991361056e-05, + "loss": 2.6343, + "step": 31462 + }, + { + "epoch": 1.34, + "grad_norm": 0.47265625, + "learning_rate": 4.918699746728753e-05, + "loss": 2.4716, + "step": 31463 + }, + { + "epoch": 1.34, + "grad_norm": 0.494140625, + "learning_rate": 4.9181220171853916e-05, + "loss": 2.4309, + "step": 31464 + }, + { + "epoch": 1.34, + "grad_norm": 0.453125, + "learning_rate": 4.9175443105086236e-05, + "loss": 2.4965, + "step": 31465 + }, + { + "epoch": 1.34, + "grad_norm": 0.4375, + "learning_rate": 4.9169666267010526e-05, + "loss": 2.4425, + "step": 31466 + }, + { + "epoch": 1.34, + "grad_norm": 0.47265625, + "learning_rate": 4.9163889657652706e-05, + "loss": 2.4861, + "step": 31467 + }, + { + "epoch": 1.34, + "grad_norm": 0.490234375, + "learning_rate": 4.9158113277038844e-05, + "loss": 2.5208, + "step": 31468 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.9152337125194846e-05, + "loss": 2.5467, + "step": 31469 + }, + { + "epoch": 1.34, + "grad_norm": 0.443359375, + "learning_rate": 4.914656120214679e-05, + "loss": 2.4548, + "step": 31470 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.9140785507920584e-05, + "loss": 2.3926, + "step": 31471 + }, + { + "epoch": 1.34, + "grad_norm": 0.462890625, + "learning_rate": 4.913501004254231e-05, + "loss": 2.482, + "step": 31472 + }, + { + "epoch": 1.34, + "grad_norm": 0.482421875, + "learning_rate": 4.912923480603784e-05, + "loss": 2.4601, + "step": 31473 + }, + { + "epoch": 1.34, + "grad_norm": 0.451171875, + "learning_rate": 4.912345979843324e-05, + "loss": 2.4599, + "step": 31474 + }, + { + "epoch": 1.34, + "grad_norm": 0.5234375, + "learning_rate": 4.91176850197545e-05, + "loss": 2.7004, + "step": 31475 + }, + { + "epoch": 1.34, + "grad_norm": 0.50390625, + "learning_rate": 4.911191047002759e-05, + "loss": 2.4657, + "step": 31476 + }, + { + "epoch": 1.34, + "grad_norm": 0.484375, + "learning_rate": 4.9106136149278416e-05, + "loss": 2.5993, + "step": 31477 + }, + { + "epoch": 1.34, + "grad_norm": 0.4609375, + "learning_rate": 4.910036205753309e-05, + "loss": 2.3878, + "step": 31478 + }, + { + "epoch": 1.34, + "grad_norm": 0.5078125, + "learning_rate": 4.909458819481755e-05, + "loss": 2.445, + "step": 31479 + }, + { + "epoch": 1.34, + "grad_norm": 0.41796875, + "learning_rate": 4.9088814561157716e-05, + "loss": 2.4133, + "step": 31480 + }, + { + "epoch": 1.34, + "grad_norm": 0.45703125, + "learning_rate": 4.90830411565796e-05, + "loss": 2.4888, + "step": 31481 + }, + { + "epoch": 1.34, + "grad_norm": 0.458984375, + "learning_rate": 4.9077267981109245e-05, + "loss": 2.4621, + "step": 31482 + }, + { + "epoch": 1.34, + "grad_norm": 0.546875, + "learning_rate": 4.907149503477252e-05, + "loss": 2.5537, + "step": 31483 + }, + { + "epoch": 1.34, + "grad_norm": 0.4375, + "learning_rate": 4.9065722317595505e-05, + "loss": 2.4208, + "step": 31484 + }, + { + "epoch": 1.34, + "grad_norm": 0.47265625, + "learning_rate": 4.905994982960408e-05, + "loss": 2.4606, + "step": 31485 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.905417757082431e-05, + "loss": 2.4215, + "step": 31486 + }, + { + "epoch": 1.34, + "grad_norm": 0.5, + "learning_rate": 4.904840554128207e-05, + "loss": 2.4493, + "step": 31487 + }, + { + "epoch": 1.34, + "grad_norm": 0.474609375, + "learning_rate": 4.9042633741003444e-05, + "loss": 2.4995, + "step": 31488 + }, + { + "epoch": 1.34, + "grad_norm": 0.427734375, + "learning_rate": 4.9036862170014286e-05, + "loss": 2.4605, + "step": 31489 + }, + { + "epoch": 1.34, + "grad_norm": 0.466796875, + "learning_rate": 4.903109082834063e-05, + "loss": 2.4943, + "step": 31490 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.9025319716008475e-05, + "loss": 2.5306, + "step": 31491 + }, + { + "epoch": 1.34, + "grad_norm": 0.486328125, + "learning_rate": 4.901954883304376e-05, + "loss": 2.5546, + "step": 31492 + }, + { + "epoch": 1.34, + "grad_norm": 0.5078125, + "learning_rate": 4.901377817947236e-05, + "loss": 2.6198, + "step": 31493 + }, + { + "epoch": 1.34, + "grad_norm": 0.474609375, + "learning_rate": 4.900800775532041e-05, + "loss": 2.3571, + "step": 31494 + }, + { + "epoch": 1.34, + "grad_norm": 0.47265625, + "learning_rate": 4.9002237560613775e-05, + "loss": 2.4714, + "step": 31495 + }, + { + "epoch": 1.34, + "grad_norm": 0.443359375, + "learning_rate": 4.89964675953784e-05, + "loss": 2.5003, + "step": 31496 + }, + { + "epoch": 1.34, + "grad_norm": 0.53125, + "learning_rate": 4.899069785964029e-05, + "loss": 2.4643, + "step": 31497 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.898492835342543e-05, + "loss": 2.5061, + "step": 31498 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.897915907675972e-05, + "loss": 2.42, + "step": 31499 + }, + { + "epoch": 1.34, + "grad_norm": 0.478515625, + "learning_rate": 4.897339002966917e-05, + "loss": 2.4932, + "step": 31500 + }, + { + "epoch": 1.34, + "grad_norm": 0.4296875, + "learning_rate": 4.896762121217968e-05, + "loss": 2.4428, + "step": 31501 + }, + { + "epoch": 1.34, + "grad_norm": 0.51171875, + "learning_rate": 4.8961852624317294e-05, + "loss": 2.3326, + "step": 31502 + }, + { + "epoch": 1.34, + "grad_norm": 0.41015625, + "learning_rate": 4.895608426610787e-05, + "loss": 2.5247, + "step": 31503 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.895031613757746e-05, + "loss": 2.4405, + "step": 31504 + }, + { + "epoch": 1.34, + "grad_norm": 0.43359375, + "learning_rate": 4.8944548238751933e-05, + "loss": 2.3235, + "step": 31505 + }, + { + "epoch": 1.34, + "grad_norm": 0.4296875, + "learning_rate": 4.8938780569657275e-05, + "loss": 2.5112, + "step": 31506 + }, + { + "epoch": 1.34, + "grad_norm": 0.50390625, + "learning_rate": 4.8933013130319484e-05, + "loss": 2.4527, + "step": 31507 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.8927245920764466e-05, + "loss": 2.4737, + "step": 31508 + }, + { + "epoch": 1.34, + "grad_norm": 0.453125, + "learning_rate": 4.8921478941018106e-05, + "loss": 2.335, + "step": 31509 + }, + { + "epoch": 1.34, + "grad_norm": 0.46875, + "learning_rate": 4.8915712191106485e-05, + "loss": 2.4095, + "step": 31510 + }, + { + "epoch": 1.34, + "grad_norm": 0.46875, + "learning_rate": 4.890994567105551e-05, + "loss": 2.5165, + "step": 31511 + }, + { + "epoch": 1.34, + "grad_norm": 0.4453125, + "learning_rate": 4.8904179380891066e-05, + "loss": 2.4438, + "step": 31512 + }, + { + "epoch": 1.34, + "grad_norm": 0.52734375, + "learning_rate": 4.889841332063913e-05, + "loss": 2.5054, + "step": 31513 + }, + { + "epoch": 1.34, + "grad_norm": 0.546875, + "learning_rate": 4.8892647490325705e-05, + "loss": 2.5595, + "step": 31514 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.888688188997665e-05, + "loss": 2.5039, + "step": 31515 + }, + { + "epoch": 1.34, + "grad_norm": 0.453125, + "learning_rate": 4.8881116519617985e-05, + "loss": 2.4744, + "step": 31516 + }, + { + "epoch": 1.34, + "grad_norm": 0.478515625, + "learning_rate": 4.887535137927558e-05, + "loss": 2.6258, + "step": 31517 + }, + { + "epoch": 1.34, + "grad_norm": 0.482421875, + "learning_rate": 4.886958646897545e-05, + "loss": 2.5056, + "step": 31518 + }, + { + "epoch": 1.34, + "grad_norm": 0.412109375, + "learning_rate": 4.886382178874345e-05, + "loss": 2.4219, + "step": 31519 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.88580573386056e-05, + "loss": 2.3386, + "step": 31520 + }, + { + "epoch": 1.34, + "grad_norm": 0.50390625, + "learning_rate": 4.8852293118587764e-05, + "loss": 2.4848, + "step": 31521 + }, + { + "epoch": 1.34, + "grad_norm": 0.404296875, + "learning_rate": 4.8846529128715926e-05, + "loss": 2.4192, + "step": 31522 + }, + { + "epoch": 1.34, + "grad_norm": 0.4296875, + "learning_rate": 4.884076536901604e-05, + "loss": 2.4698, + "step": 31523 + }, + { + "epoch": 1.34, + "grad_norm": 0.42578125, + "learning_rate": 4.883500183951402e-05, + "loss": 2.5196, + "step": 31524 + }, + { + "epoch": 1.34, + "grad_norm": 0.5078125, + "learning_rate": 4.8829238540235744e-05, + "loss": 2.5007, + "step": 31525 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.88234754712072e-05, + "loss": 2.4319, + "step": 31526 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.881771263245436e-05, + "loss": 2.4623, + "step": 31527 + }, + { + "epoch": 1.34, + "grad_norm": 0.494140625, + "learning_rate": 4.881195002400306e-05, + "loss": 2.6007, + "step": 31528 + }, + { + "epoch": 1.34, + "grad_norm": 0.51171875, + "learning_rate": 4.880618764587932e-05, + "loss": 2.3649, + "step": 31529 + }, + { + "epoch": 1.34, + "grad_norm": 0.48046875, + "learning_rate": 4.880042549810898e-05, + "loss": 2.5419, + "step": 31530 + }, + { + "epoch": 1.34, + "grad_norm": 0.494140625, + "learning_rate": 4.8794663580718024e-05, + "loss": 2.5048, + "step": 31531 + }, + { + "epoch": 1.34, + "grad_norm": 0.5078125, + "learning_rate": 4.878890189373241e-05, + "loss": 2.4374, + "step": 31532 + }, + { + "epoch": 1.34, + "grad_norm": 0.490234375, + "learning_rate": 4.878314043717801e-05, + "loss": 2.3609, + "step": 31533 + }, + { + "epoch": 1.34, + "grad_norm": 0.515625, + "learning_rate": 4.877737921108073e-05, + "loss": 2.5199, + "step": 31534 + }, + { + "epoch": 1.34, + "grad_norm": 0.5, + "learning_rate": 4.877161821546652e-05, + "loss": 2.3739, + "step": 31535 + }, + { + "epoch": 1.34, + "grad_norm": 0.466796875, + "learning_rate": 4.876585745036135e-05, + "loss": 2.5424, + "step": 31536 + }, + { + "epoch": 1.34, + "grad_norm": 0.47265625, + "learning_rate": 4.8760096915791053e-05, + "loss": 2.572, + "step": 31537 + }, + { + "epoch": 1.34, + "grad_norm": 0.451171875, + "learning_rate": 4.8754336611781595e-05, + "loss": 2.5493, + "step": 31538 + }, + { + "epoch": 1.34, + "grad_norm": 0.515625, + "learning_rate": 4.874857653835894e-05, + "loss": 2.4732, + "step": 31539 + }, + { + "epoch": 1.34, + "grad_norm": 0.55078125, + "learning_rate": 4.874281669554895e-05, + "loss": 2.4991, + "step": 31540 + }, + { + "epoch": 1.34, + "grad_norm": 0.431640625, + "learning_rate": 4.873705708337751e-05, + "loss": 2.4894, + "step": 31541 + }, + { + "epoch": 1.34, + "grad_norm": 0.478515625, + "learning_rate": 4.873129770187059e-05, + "loss": 2.5345, + "step": 31542 + }, + { + "epoch": 1.34, + "grad_norm": 0.498046875, + "learning_rate": 4.8725538551054126e-05, + "loss": 2.3827, + "step": 31543 + }, + { + "epoch": 1.34, + "grad_norm": 0.41796875, + "learning_rate": 4.871977963095398e-05, + "loss": 2.4825, + "step": 31544 + }, + { + "epoch": 1.34, + "grad_norm": 0.421875, + "learning_rate": 4.871402094159611e-05, + "loss": 2.4165, + "step": 31545 + }, + { + "epoch": 1.34, + "grad_norm": 0.470703125, + "learning_rate": 4.870826248300635e-05, + "loss": 2.4497, + "step": 31546 + }, + { + "epoch": 1.34, + "grad_norm": 0.44140625, + "learning_rate": 4.870250425521068e-05, + "loss": 2.5315, + "step": 31547 + }, + { + "epoch": 1.34, + "grad_norm": 0.42578125, + "learning_rate": 4.8696746258235026e-05, + "loss": 2.4151, + "step": 31548 + }, + { + "epoch": 1.34, + "grad_norm": 0.4609375, + "learning_rate": 4.869098849210527e-05, + "loss": 2.4548, + "step": 31549 + }, + { + "epoch": 1.34, + "grad_norm": 0.49609375, + "learning_rate": 4.8685230956847274e-05, + "loss": 2.4897, + "step": 31550 + }, + { + "epoch": 1.34, + "grad_norm": 0.52734375, + "learning_rate": 4.867947365248698e-05, + "loss": 2.516, + "step": 31551 + }, + { + "epoch": 1.34, + "grad_norm": 0.431640625, + "learning_rate": 4.867371657905034e-05, + "loss": 2.444, + "step": 31552 + }, + { + "epoch": 1.34, + "grad_norm": 0.46875, + "learning_rate": 4.8667959736563185e-05, + "loss": 2.5034, + "step": 31553 + }, + { + "epoch": 1.34, + "grad_norm": 0.482421875, + "learning_rate": 4.866220312505144e-05, + "loss": 2.5446, + "step": 31554 + }, + { + "epoch": 1.34, + "grad_norm": 0.53515625, + "learning_rate": 4.8656446744541076e-05, + "loss": 2.6346, + "step": 31555 + }, + { + "epoch": 1.34, + "grad_norm": 0.466796875, + "learning_rate": 4.865069059505792e-05, + "loss": 2.3677, + "step": 31556 + }, + { + "epoch": 1.34, + "grad_norm": 0.490234375, + "learning_rate": 4.864493467662786e-05, + "loss": 2.3824, + "step": 31557 + }, + { + "epoch": 1.34, + "grad_norm": 0.439453125, + "learning_rate": 4.863917898927682e-05, + "loss": 2.3161, + "step": 31558 + }, + { + "epoch": 1.34, + "grad_norm": 0.44140625, + "learning_rate": 4.8633423533030756e-05, + "loss": 2.5354, + "step": 31559 + }, + { + "epoch": 1.34, + "grad_norm": 0.5, + "learning_rate": 4.8627668307915465e-05, + "loss": 2.47, + "step": 31560 + }, + { + "epoch": 1.34, + "grad_norm": 0.482421875, + "learning_rate": 4.8621913313956934e-05, + "loss": 2.4098, + "step": 31561 + }, + { + "epoch": 1.34, + "grad_norm": 0.44921875, + "learning_rate": 4.861615855118098e-05, + "loss": 2.4718, + "step": 31562 + }, + { + "epoch": 1.34, + "grad_norm": 0.46484375, + "learning_rate": 4.861040401961352e-05, + "loss": 2.3346, + "step": 31563 + }, + { + "epoch": 1.35, + "grad_norm": 0.5, + "learning_rate": 4.860464971928052e-05, + "loss": 2.4777, + "step": 31564 + }, + { + "epoch": 1.35, + "grad_norm": 0.478515625, + "learning_rate": 4.85988956502078e-05, + "loss": 2.6198, + "step": 31565 + }, + { + "epoch": 1.35, + "grad_norm": 0.474609375, + "learning_rate": 4.859314181242123e-05, + "loss": 2.4379, + "step": 31566 + }, + { + "epoch": 1.35, + "grad_norm": 0.486328125, + "learning_rate": 4.8587388205946726e-05, + "loss": 2.5277, + "step": 31567 + }, + { + "epoch": 1.35, + "grad_norm": 0.43359375, + "learning_rate": 4.858163483081024e-05, + "loss": 2.362, + "step": 31568 + }, + { + "epoch": 1.35, + "grad_norm": 0.380859375, + "learning_rate": 4.857588168703756e-05, + "loss": 2.4235, + "step": 31569 + }, + { + "epoch": 1.35, + "grad_norm": 0.46484375, + "learning_rate": 4.857012877465462e-05, + "loss": 2.5848, + "step": 31570 + }, + { + "epoch": 1.35, + "grad_norm": 0.462890625, + "learning_rate": 4.856437609368734e-05, + "loss": 2.4839, + "step": 31571 + }, + { + "epoch": 1.35, + "grad_norm": 0.484375, + "learning_rate": 4.855862364416157e-05, + "loss": 2.5623, + "step": 31572 + }, + { + "epoch": 1.35, + "grad_norm": 0.46875, + "learning_rate": 4.8552871426103164e-05, + "loss": 2.424, + "step": 31573 + }, + { + "epoch": 1.35, + "grad_norm": 0.4921875, + "learning_rate": 4.854711943953803e-05, + "loss": 2.4009, + "step": 31574 + }, + { + "epoch": 1.35, + "grad_norm": 0.4609375, + "learning_rate": 4.854136768449209e-05, + "loss": 2.4736, + "step": 31575 + }, + { + "epoch": 1.35, + "grad_norm": 0.4375, + "learning_rate": 4.853561616099115e-05, + "loss": 2.4675, + "step": 31576 + }, + { + "epoch": 1.35, + "grad_norm": 0.447265625, + "learning_rate": 4.852986486906117e-05, + "loss": 2.3969, + "step": 31577 + }, + { + "epoch": 1.35, + "grad_norm": 0.5078125, + "learning_rate": 4.8524113808727956e-05, + "loss": 2.4311, + "step": 31578 + }, + { + "epoch": 1.35, + "grad_norm": 0.5234375, + "learning_rate": 4.851836298001741e-05, + "loss": 2.3879, + "step": 31579 + }, + { + "epoch": 1.35, + "grad_norm": 0.515625, + "learning_rate": 4.851261238295546e-05, + "loss": 2.4459, + "step": 31580 + }, + { + "epoch": 1.35, + "grad_norm": 0.50390625, + "learning_rate": 4.8506862017567935e-05, + "loss": 2.4741, + "step": 31581 + }, + { + "epoch": 1.35, + "grad_norm": 0.484375, + "learning_rate": 4.850111188388067e-05, + "loss": 2.4247, + "step": 31582 + }, + { + "epoch": 1.35, + "grad_norm": 0.4921875, + "learning_rate": 4.8495361981919586e-05, + "loss": 2.4425, + "step": 31583 + }, + { + "epoch": 1.35, + "grad_norm": 0.458984375, + "learning_rate": 4.848961231171059e-05, + "loss": 2.432, + "step": 31584 + }, + { + "epoch": 1.35, + "grad_norm": 0.43359375, + "learning_rate": 4.8483862873279464e-05, + "loss": 2.4777, + "step": 31585 + }, + { + "epoch": 1.35, + "grad_norm": 0.44140625, + "learning_rate": 4.847811366665214e-05, + "loss": 2.2378, + "step": 31586 + }, + { + "epoch": 1.35, + "grad_norm": 0.466796875, + "learning_rate": 4.847236469185451e-05, + "loss": 2.4106, + "step": 31587 + }, + { + "epoch": 1.35, + "grad_norm": 0.46484375, + "learning_rate": 4.84666159489124e-05, + "loss": 2.4005, + "step": 31588 + }, + { + "epoch": 1.35, + "grad_norm": 0.44921875, + "learning_rate": 4.846086743785165e-05, + "loss": 2.4776, + "step": 31589 + }, + { + "epoch": 1.35, + "grad_norm": 0.54296875, + "learning_rate": 4.845511915869817e-05, + "loss": 2.5637, + "step": 31590 + }, + { + "epoch": 1.35, + "grad_norm": 0.45703125, + "learning_rate": 4.844937111147785e-05, + "loss": 2.3572, + "step": 31591 + }, + { + "epoch": 1.35, + "grad_norm": 0.443359375, + "learning_rate": 4.8443623296216476e-05, + "loss": 2.3513, + "step": 31592 + }, + { + "epoch": 1.35, + "grad_norm": 0.51953125, + "learning_rate": 4.8437875712940006e-05, + "loss": 2.4752, + "step": 31593 + }, + { + "epoch": 1.35, + "grad_norm": 0.5234375, + "learning_rate": 4.843212836167421e-05, + "loss": 2.4063, + "step": 31594 + }, + { + "epoch": 1.35, + "grad_norm": 0.41015625, + "learning_rate": 4.842638124244498e-05, + "loss": 2.3544, + "step": 31595 + }, + { + "epoch": 1.35, + "grad_norm": 0.451171875, + "learning_rate": 4.842063435527824e-05, + "loss": 2.486, + "step": 31596 + }, + { + "epoch": 1.35, + "grad_norm": 0.5, + "learning_rate": 4.8414887700199787e-05, + "loss": 2.4533, + "step": 31597 + }, + { + "epoch": 1.35, + "grad_norm": 0.455078125, + "learning_rate": 4.8409141277235404e-05, + "loss": 2.4217, + "step": 31598 + }, + { + "epoch": 1.35, + "grad_norm": 0.451171875, + "learning_rate": 4.840339508641112e-05, + "loss": 2.5964, + "step": 31599 + }, + { + "epoch": 1.35, + "grad_norm": 0.5, + "learning_rate": 4.83976491277527e-05, + "loss": 2.5115, + "step": 31600 + }, + { + "epoch": 1.35, + "grad_norm": 0.486328125, + "learning_rate": 4.839190340128595e-05, + "loss": 2.4764, + "step": 31601 + }, + { + "epoch": 1.35, + "grad_norm": 0.5390625, + "learning_rate": 4.838615790703679e-05, + "loss": 2.5357, + "step": 31602 + }, + { + "epoch": 1.35, + "grad_norm": 0.46484375, + "learning_rate": 4.838041264503108e-05, + "loss": 2.4834, + "step": 31603 + }, + { + "epoch": 1.35, + "grad_norm": 0.451171875, + "learning_rate": 4.837466761529465e-05, + "loss": 2.4971, + "step": 31604 + }, + { + "epoch": 1.35, + "grad_norm": 0.400390625, + "learning_rate": 4.8368922817853323e-05, + "loss": 2.4296, + "step": 31605 + }, + { + "epoch": 1.35, + "grad_norm": 0.486328125, + "learning_rate": 4.8363178252732966e-05, + "loss": 2.4482, + "step": 31606 + }, + { + "epoch": 1.35, + "grad_norm": 0.447265625, + "learning_rate": 4.8357433919959485e-05, + "loss": 2.3642, + "step": 31607 + }, + { + "epoch": 1.35, + "grad_norm": 0.455078125, + "learning_rate": 4.8351689819558624e-05, + "loss": 2.5794, + "step": 31608 + }, + { + "epoch": 1.35, + "grad_norm": 0.451171875, + "learning_rate": 4.834594595155633e-05, + "loss": 2.3807, + "step": 31609 + }, + { + "epoch": 1.35, + "grad_norm": 0.462890625, + "learning_rate": 4.834020231597836e-05, + "loss": 2.558, + "step": 31610 + }, + { + "epoch": 1.35, + "grad_norm": 0.4921875, + "learning_rate": 4.83344589128506e-05, + "loss": 2.4725, + "step": 31611 + }, + { + "epoch": 1.35, + "grad_norm": 0.4609375, + "learning_rate": 4.832871574219894e-05, + "loss": 2.4948, + "step": 31612 + }, + { + "epoch": 1.35, + "grad_norm": 0.43359375, + "learning_rate": 4.832297280404916e-05, + "loss": 2.4484, + "step": 31613 + }, + { + "epoch": 1.35, + "grad_norm": 0.50390625, + "learning_rate": 4.831723009842706e-05, + "loss": 2.4828, + "step": 31614 + }, + { + "epoch": 1.35, + "grad_norm": 0.453125, + "learning_rate": 4.8311487625358595e-05, + "loss": 2.3941, + "step": 31615 + }, + { + "epoch": 1.35, + "grad_norm": 0.4140625, + "learning_rate": 4.830574538486955e-05, + "loss": 2.4748, + "step": 31616 + }, + { + "epoch": 1.35, + "grad_norm": 0.494140625, + "learning_rate": 4.8300003376985725e-05, + "loss": 2.3234, + "step": 31617 + }, + { + "epoch": 1.35, + "grad_norm": 0.443359375, + "learning_rate": 4.8294261601732995e-05, + "loss": 2.4069, + "step": 31618 + }, + { + "epoch": 1.35, + "grad_norm": 0.48828125, + "learning_rate": 4.828852005913722e-05, + "loss": 2.6088, + "step": 31619 + }, + { + "epoch": 1.35, + "grad_norm": 0.51171875, + "learning_rate": 4.8282778749224214e-05, + "loss": 2.5895, + "step": 31620 + }, + { + "epoch": 1.35, + "grad_norm": 0.392578125, + "learning_rate": 4.827703767201975e-05, + "loss": 2.2754, + "step": 31621 + }, + { + "epoch": 1.35, + "grad_norm": 0.486328125, + "learning_rate": 4.827129682754973e-05, + "loss": 2.4987, + "step": 31622 + }, + { + "epoch": 1.35, + "grad_norm": 0.453125, + "learning_rate": 4.826555621584e-05, + "loss": 2.4636, + "step": 31623 + }, + { + "epoch": 1.35, + "grad_norm": 0.40625, + "learning_rate": 4.8259815836916334e-05, + "loss": 2.4701, + "step": 31624 + }, + { + "epoch": 1.35, + "grad_norm": 0.474609375, + "learning_rate": 4.825407569080461e-05, + "loss": 2.5889, + "step": 31625 + }, + { + "epoch": 1.35, + "grad_norm": 0.498046875, + "learning_rate": 4.824833577753061e-05, + "loss": 2.481, + "step": 31626 + }, + { + "epoch": 1.35, + "grad_norm": 0.44921875, + "learning_rate": 4.8242596097120175e-05, + "loss": 2.4947, + "step": 31627 + }, + { + "epoch": 1.35, + "grad_norm": 0.47265625, + "learning_rate": 4.823685664959918e-05, + "loss": 2.4301, + "step": 31628 + }, + { + "epoch": 1.35, + "grad_norm": 0.515625, + "learning_rate": 4.8231117434993424e-05, + "loss": 2.3475, + "step": 31629 + }, + { + "epoch": 1.35, + "grad_norm": 0.443359375, + "learning_rate": 4.822537845332863e-05, + "loss": 2.4192, + "step": 31630 + }, + { + "epoch": 1.35, + "grad_norm": 0.396484375, + "learning_rate": 4.821963970463079e-05, + "loss": 2.3391, + "step": 31631 + }, + { + "epoch": 1.35, + "grad_norm": 0.447265625, + "learning_rate": 4.8213901188925656e-05, + "loss": 2.64, + "step": 31632 + }, + { + "epoch": 1.35, + "grad_norm": 0.474609375, + "learning_rate": 4.8208162906239005e-05, + "loss": 2.4714, + "step": 31633 + }, + { + "epoch": 1.35, + "grad_norm": 0.443359375, + "learning_rate": 4.8202424856596685e-05, + "loss": 2.382, + "step": 31634 + }, + { + "epoch": 1.35, + "grad_norm": 0.478515625, + "learning_rate": 4.8196687040024567e-05, + "loss": 2.5162, + "step": 31635 + }, + { + "epoch": 1.35, + "grad_norm": 0.48046875, + "learning_rate": 4.8190949456548414e-05, + "loss": 2.6349, + "step": 31636 + }, + { + "epoch": 1.35, + "grad_norm": 0.48828125, + "learning_rate": 4.818521210619403e-05, + "loss": 2.5604, + "step": 31637 + }, + { + "epoch": 1.35, + "grad_norm": 0.46875, + "learning_rate": 4.817947498898725e-05, + "loss": 2.4321, + "step": 31638 + }, + { + "epoch": 1.35, + "grad_norm": 0.474609375, + "learning_rate": 4.8173738104953935e-05, + "loss": 2.3883, + "step": 31639 + }, + { + "epoch": 1.35, + "grad_norm": 0.455078125, + "learning_rate": 4.8168001454119814e-05, + "loss": 2.4342, + "step": 31640 + }, + { + "epoch": 1.35, + "grad_norm": 0.4375, + "learning_rate": 4.816226503651079e-05, + "loss": 2.4179, + "step": 31641 + }, + { + "epoch": 1.35, + "grad_norm": 0.4453125, + "learning_rate": 4.8156528852152586e-05, + "loss": 2.4714, + "step": 31642 + }, + { + "epoch": 1.35, + "grad_norm": 0.427734375, + "learning_rate": 4.815079290107105e-05, + "loss": 2.4832, + "step": 31643 + }, + { + "epoch": 1.35, + "grad_norm": 0.5078125, + "learning_rate": 4.814505718329205e-05, + "loss": 2.4383, + "step": 31644 + }, + { + "epoch": 1.35, + "grad_norm": 0.453125, + "learning_rate": 4.8139321698841334e-05, + "loss": 2.4985, + "step": 31645 + }, + { + "epoch": 1.35, + "grad_norm": 0.4375, + "learning_rate": 4.8133586447744636e-05, + "loss": 2.4784, + "step": 31646 + }, + { + "epoch": 1.35, + "grad_norm": 0.427734375, + "learning_rate": 4.8127851430027924e-05, + "loss": 2.63, + "step": 31647 + }, + { + "epoch": 1.35, + "grad_norm": 0.439453125, + "learning_rate": 4.812211664571693e-05, + "loss": 2.4417, + "step": 31648 + }, + { + "epoch": 1.35, + "grad_norm": 0.48046875, + "learning_rate": 4.811638209483741e-05, + "loss": 2.4748, + "step": 31649 + }, + { + "epoch": 1.35, + "grad_norm": 0.44921875, + "learning_rate": 4.811064777741524e-05, + "loss": 2.382, + "step": 31650 + }, + { + "epoch": 1.35, + "grad_norm": 0.45703125, + "learning_rate": 4.8104913693476163e-05, + "loss": 2.406, + "step": 31651 + }, + { + "epoch": 1.35, + "grad_norm": 0.5, + "learning_rate": 4.809917984304601e-05, + "loss": 2.596, + "step": 31652 + }, + { + "epoch": 1.35, + "grad_norm": 0.44140625, + "learning_rate": 4.809344622615061e-05, + "loss": 2.514, + "step": 31653 + }, + { + "epoch": 1.35, + "grad_norm": 0.51953125, + "learning_rate": 4.808771284281572e-05, + "loss": 2.5632, + "step": 31654 + }, + { + "epoch": 1.35, + "grad_norm": 0.4765625, + "learning_rate": 4.808197969306714e-05, + "loss": 2.5073, + "step": 31655 + }, + { + "epoch": 1.35, + "grad_norm": 0.498046875, + "learning_rate": 4.807624677693067e-05, + "loss": 2.3689, + "step": 31656 + }, + { + "epoch": 1.35, + "grad_norm": 0.439453125, + "learning_rate": 4.807051409443214e-05, + "loss": 2.3346, + "step": 31657 + }, + { + "epoch": 1.35, + "grad_norm": 0.546875, + "learning_rate": 4.806478164559732e-05, + "loss": 2.4395, + "step": 31658 + }, + { + "epoch": 1.35, + "grad_norm": 0.451171875, + "learning_rate": 4.805904943045193e-05, + "loss": 2.2785, + "step": 31659 + }, + { + "epoch": 1.35, + "grad_norm": 0.5234375, + "learning_rate": 4.805331744902192e-05, + "loss": 2.5308, + "step": 31660 + }, + { + "epoch": 1.35, + "grad_norm": 0.51953125, + "learning_rate": 4.804758570133298e-05, + "loss": 2.4636, + "step": 31661 + }, + { + "epoch": 1.35, + "grad_norm": 0.4453125, + "learning_rate": 4.80418541874109e-05, + "loss": 2.4577, + "step": 31662 + }, + { + "epoch": 1.35, + "grad_norm": 0.466796875, + "learning_rate": 4.803612290728148e-05, + "loss": 2.4913, + "step": 31663 + }, + { + "epoch": 1.35, + "grad_norm": 0.44140625, + "learning_rate": 4.8030391860970545e-05, + "loss": 2.3931, + "step": 31664 + }, + { + "epoch": 1.35, + "grad_norm": 0.482421875, + "learning_rate": 4.8024661048503825e-05, + "loss": 2.36, + "step": 31665 + }, + { + "epoch": 1.35, + "grad_norm": 0.45703125, + "learning_rate": 4.801893046990718e-05, + "loss": 2.4269, + "step": 31666 + }, + { + "epoch": 1.35, + "grad_norm": 0.470703125, + "learning_rate": 4.801320012520629e-05, + "loss": 2.5772, + "step": 31667 + }, + { + "epoch": 1.35, + "grad_norm": 0.47265625, + "learning_rate": 4.8007470014427026e-05, + "loss": 2.3293, + "step": 31668 + }, + { + "epoch": 1.35, + "grad_norm": 0.43359375, + "learning_rate": 4.8001740137595164e-05, + "loss": 2.4027, + "step": 31669 + }, + { + "epoch": 1.35, + "grad_norm": 0.458984375, + "learning_rate": 4.7996010494736474e-05, + "loss": 2.5147, + "step": 31670 + }, + { + "epoch": 1.35, + "grad_norm": 0.419921875, + "learning_rate": 4.79902810858767e-05, + "loss": 2.4732, + "step": 31671 + }, + { + "epoch": 1.35, + "grad_norm": 0.482421875, + "learning_rate": 4.7984551911041656e-05, + "loss": 2.4408, + "step": 31672 + }, + { + "epoch": 1.35, + "grad_norm": 0.4921875, + "learning_rate": 4.797882297025714e-05, + "loss": 2.5318, + "step": 31673 + }, + { + "epoch": 1.35, + "grad_norm": 0.51171875, + "learning_rate": 4.797309426354892e-05, + "loss": 2.4938, + "step": 31674 + }, + { + "epoch": 1.35, + "grad_norm": 0.4296875, + "learning_rate": 4.79673657909427e-05, + "loss": 2.4313, + "step": 31675 + }, + { + "epoch": 1.35, + "grad_norm": 0.474609375, + "learning_rate": 4.796163755246438e-05, + "loss": 2.4965, + "step": 31676 + }, + { + "epoch": 1.35, + "grad_norm": 0.5, + "learning_rate": 4.795590954813966e-05, + "loss": 2.3952, + "step": 31677 + }, + { + "epoch": 1.35, + "grad_norm": 0.45703125, + "learning_rate": 4.795018177799431e-05, + "loss": 2.5364, + "step": 31678 + }, + { + "epoch": 1.35, + "grad_norm": 0.47265625, + "learning_rate": 4.7944454242054114e-05, + "loss": 2.4238, + "step": 31679 + }, + { + "epoch": 1.35, + "grad_norm": 0.48046875, + "learning_rate": 4.793872694034488e-05, + "loss": 2.5069, + "step": 31680 + }, + { + "epoch": 1.35, + "grad_norm": 0.49609375, + "learning_rate": 4.793299987289233e-05, + "loss": 2.6015, + "step": 31681 + }, + { + "epoch": 1.35, + "grad_norm": 0.486328125, + "learning_rate": 4.7927273039722264e-05, + "loss": 2.4775, + "step": 31682 + }, + { + "epoch": 1.35, + "grad_norm": 0.515625, + "learning_rate": 4.792154644086042e-05, + "loss": 2.5279, + "step": 31683 + }, + { + "epoch": 1.35, + "grad_norm": 0.55859375, + "learning_rate": 4.791582007633257e-05, + "loss": 2.566, + "step": 31684 + }, + { + "epoch": 1.35, + "grad_norm": 0.5390625, + "learning_rate": 4.791009394616455e-05, + "loss": 2.5494, + "step": 31685 + }, + { + "epoch": 1.35, + "grad_norm": 0.41796875, + "learning_rate": 4.790436805038206e-05, + "loss": 2.332, + "step": 31686 + }, + { + "epoch": 1.35, + "grad_norm": 0.3984375, + "learning_rate": 4.789864238901084e-05, + "loss": 2.4007, + "step": 31687 + }, + { + "epoch": 1.35, + "grad_norm": 0.3984375, + "learning_rate": 4.7892916962076685e-05, + "loss": 2.3621, + "step": 31688 + }, + { + "epoch": 1.35, + "grad_norm": 0.423828125, + "learning_rate": 4.788719176960541e-05, + "loss": 2.3294, + "step": 31689 + }, + { + "epoch": 1.35, + "grad_norm": 0.546875, + "learning_rate": 4.7881466811622714e-05, + "loss": 2.5252, + "step": 31690 + }, + { + "epoch": 1.35, + "grad_norm": 0.44140625, + "learning_rate": 4.78757420881543e-05, + "loss": 2.3682, + "step": 31691 + }, + { + "epoch": 1.35, + "grad_norm": 0.455078125, + "learning_rate": 4.7870017599226066e-05, + "loss": 2.461, + "step": 31692 + }, + { + "epoch": 1.35, + "grad_norm": 0.52734375, + "learning_rate": 4.786429334486371e-05, + "loss": 2.4284, + "step": 31693 + }, + { + "epoch": 1.35, + "grad_norm": 0.482421875, + "learning_rate": 4.785856932509294e-05, + "loss": 2.4331, + "step": 31694 + }, + { + "epoch": 1.35, + "grad_norm": 0.427734375, + "learning_rate": 4.785284553993955e-05, + "loss": 2.3556, + "step": 31695 + }, + { + "epoch": 1.35, + "grad_norm": 0.4765625, + "learning_rate": 4.7847121989429345e-05, + "loss": 2.4944, + "step": 31696 + }, + { + "epoch": 1.35, + "grad_norm": 0.54296875, + "learning_rate": 4.7841398673587976e-05, + "loss": 2.5408, + "step": 31697 + }, + { + "epoch": 1.35, + "grad_norm": 0.453125, + "learning_rate": 4.783567559244131e-05, + "loss": 2.598, + "step": 31698 + }, + { + "epoch": 1.35, + "grad_norm": 0.42578125, + "learning_rate": 4.782995274601498e-05, + "loss": 2.3702, + "step": 31699 + }, + { + "epoch": 1.35, + "grad_norm": 0.423828125, + "learning_rate": 4.7824230134334815e-05, + "loss": 2.4282, + "step": 31700 + }, + { + "epoch": 1.35, + "grad_norm": 0.490234375, + "learning_rate": 4.781850775742658e-05, + "loss": 2.521, + "step": 31701 + }, + { + "epoch": 1.35, + "grad_norm": 0.50390625, + "learning_rate": 4.781278561531598e-05, + "loss": 2.4029, + "step": 31702 + }, + { + "epoch": 1.35, + "grad_norm": 0.51953125, + "learning_rate": 4.7807063708028746e-05, + "loss": 2.4702, + "step": 31703 + }, + { + "epoch": 1.35, + "grad_norm": 0.51171875, + "learning_rate": 4.7801342035590645e-05, + "loss": 2.5355, + "step": 31704 + }, + { + "epoch": 1.35, + "grad_norm": 0.5234375, + "learning_rate": 4.779562059802747e-05, + "loss": 2.4371, + "step": 31705 + }, + { + "epoch": 1.35, + "grad_norm": 0.470703125, + "learning_rate": 4.778989939536489e-05, + "loss": 2.5157, + "step": 31706 + }, + { + "epoch": 1.35, + "grad_norm": 0.49609375, + "learning_rate": 4.7784178427628665e-05, + "loss": 2.5281, + "step": 31707 + }, + { + "epoch": 1.35, + "grad_norm": 0.51953125, + "learning_rate": 4.777845769484461e-05, + "loss": 2.4966, + "step": 31708 + }, + { + "epoch": 1.35, + "grad_norm": 0.380859375, + "learning_rate": 4.777273719703841e-05, + "loss": 2.4162, + "step": 31709 + }, + { + "epoch": 1.35, + "grad_norm": 0.51953125, + "learning_rate": 4.776701693423575e-05, + "loss": 2.3125, + "step": 31710 + }, + { + "epoch": 1.35, + "grad_norm": 0.439453125, + "learning_rate": 4.776129690646244e-05, + "loss": 2.3593, + "step": 31711 + }, + { + "epoch": 1.35, + "grad_norm": 0.46484375, + "learning_rate": 4.7755577113744244e-05, + "loss": 2.4564, + "step": 31712 + }, + { + "epoch": 1.35, + "grad_norm": 0.484375, + "learning_rate": 4.77498575561068e-05, + "loss": 2.5211, + "step": 31713 + }, + { + "epoch": 1.35, + "grad_norm": 0.50390625, + "learning_rate": 4.774413823357596e-05, + "loss": 2.386, + "step": 31714 + }, + { + "epoch": 1.35, + "grad_norm": 0.462890625, + "learning_rate": 4.773841914617736e-05, + "loss": 2.4382, + "step": 31715 + }, + { + "epoch": 1.35, + "grad_norm": 0.447265625, + "learning_rate": 4.7732700293936775e-05, + "loss": 2.3726, + "step": 31716 + }, + { + "epoch": 1.35, + "grad_norm": 0.486328125, + "learning_rate": 4.7726981676879966e-05, + "loss": 2.5059, + "step": 31717 + }, + { + "epoch": 1.35, + "grad_norm": 0.482421875, + "learning_rate": 4.772126329503264e-05, + "loss": 2.492, + "step": 31718 + }, + { + "epoch": 1.35, + "grad_norm": 0.470703125, + "learning_rate": 4.7715545148420485e-05, + "loss": 2.5622, + "step": 31719 + }, + { + "epoch": 1.35, + "grad_norm": 0.46484375, + "learning_rate": 4.770982723706927e-05, + "loss": 2.2992, + "step": 31720 + }, + { + "epoch": 1.35, + "grad_norm": 0.3984375, + "learning_rate": 4.770410956100476e-05, + "loss": 2.3981, + "step": 31721 + }, + { + "epoch": 1.35, + "grad_norm": 0.52734375, + "learning_rate": 4.76983921202526e-05, + "loss": 2.38, + "step": 31722 + }, + { + "epoch": 1.35, + "grad_norm": 0.4140625, + "learning_rate": 4.769267491483857e-05, + "loss": 2.2721, + "step": 31723 + }, + { + "epoch": 1.35, + "grad_norm": 0.498046875, + "learning_rate": 4.7686957944788425e-05, + "loss": 2.4867, + "step": 31724 + }, + { + "epoch": 1.35, + "grad_norm": 0.416015625, + "learning_rate": 4.7681241210127846e-05, + "loss": 2.3185, + "step": 31725 + }, + { + "epoch": 1.35, + "grad_norm": 0.466796875, + "learning_rate": 4.767552471088254e-05, + "loss": 2.417, + "step": 31726 + }, + { + "epoch": 1.35, + "grad_norm": 0.5, + "learning_rate": 4.7669808447078224e-05, + "loss": 2.5394, + "step": 31727 + }, + { + "epoch": 1.35, + "grad_norm": 0.4921875, + "learning_rate": 4.76640924187407e-05, + "loss": 2.4576, + "step": 31728 + }, + { + "epoch": 1.35, + "grad_norm": 0.447265625, + "learning_rate": 4.7658376625895595e-05, + "loss": 2.3662, + "step": 31729 + }, + { + "epoch": 1.35, + "grad_norm": 0.453125, + "learning_rate": 4.765266106856869e-05, + "loss": 2.4496, + "step": 31730 + }, + { + "epoch": 1.35, + "grad_norm": 0.484375, + "learning_rate": 4.7646945746785665e-05, + "loss": 2.5496, + "step": 31731 + }, + { + "epoch": 1.35, + "grad_norm": 0.51953125, + "learning_rate": 4.764123066057223e-05, + "loss": 2.5295, + "step": 31732 + }, + { + "epoch": 1.35, + "grad_norm": 0.462890625, + "learning_rate": 4.763551580995418e-05, + "loss": 2.4235, + "step": 31733 + }, + { + "epoch": 1.35, + "grad_norm": 0.4765625, + "learning_rate": 4.762980119495716e-05, + "loss": 2.3675, + "step": 31734 + }, + { + "epoch": 1.35, + "grad_norm": 0.474609375, + "learning_rate": 4.762408681560685e-05, + "loss": 2.4309, + "step": 31735 + }, + { + "epoch": 1.35, + "grad_norm": 0.4765625, + "learning_rate": 4.7618372671929026e-05, + "loss": 2.4101, + "step": 31736 + }, + { + "epoch": 1.35, + "grad_norm": 0.455078125, + "learning_rate": 4.76126587639494e-05, + "loss": 2.3407, + "step": 31737 + }, + { + "epoch": 1.35, + "grad_norm": 0.41796875, + "learning_rate": 4.7606945091693644e-05, + "loss": 2.3418, + "step": 31738 + }, + { + "epoch": 1.35, + "grad_norm": 0.462890625, + "learning_rate": 4.7601231655187485e-05, + "loss": 2.3386, + "step": 31739 + }, + { + "epoch": 1.35, + "grad_norm": 0.494140625, + "learning_rate": 4.759551845445667e-05, + "loss": 2.5129, + "step": 31740 + }, + { + "epoch": 1.35, + "grad_norm": 0.38671875, + "learning_rate": 4.758980548952686e-05, + "loss": 2.4587, + "step": 31741 + }, + { + "epoch": 1.35, + "grad_norm": 0.458984375, + "learning_rate": 4.7584092760423736e-05, + "loss": 2.4908, + "step": 31742 + }, + { + "epoch": 1.35, + "grad_norm": 0.44921875, + "learning_rate": 4.7578380267173054e-05, + "loss": 2.5113, + "step": 31743 + }, + { + "epoch": 1.35, + "grad_norm": 0.4765625, + "learning_rate": 4.757266800980052e-05, + "loss": 2.5189, + "step": 31744 + }, + { + "epoch": 1.35, + "grad_norm": 0.490234375, + "learning_rate": 4.7566955988331796e-05, + "loss": 2.5735, + "step": 31745 + }, + { + "epoch": 1.35, + "grad_norm": 0.51171875, + "learning_rate": 4.7561244202792645e-05, + "loss": 2.4753, + "step": 31746 + }, + { + "epoch": 1.35, + "grad_norm": 0.458984375, + "learning_rate": 4.755553265320869e-05, + "loss": 2.4202, + "step": 31747 + }, + { + "epoch": 1.35, + "grad_norm": 0.49609375, + "learning_rate": 4.754982133960567e-05, + "loss": 2.3572, + "step": 31748 + }, + { + "epoch": 1.35, + "grad_norm": 0.53515625, + "learning_rate": 4.754411026200933e-05, + "loss": 2.4151, + "step": 31749 + }, + { + "epoch": 1.35, + "grad_norm": 0.53125, + "learning_rate": 4.753839942044531e-05, + "loss": 2.4739, + "step": 31750 + }, + { + "epoch": 1.35, + "grad_norm": 0.439453125, + "learning_rate": 4.7532688814939296e-05, + "loss": 2.3258, + "step": 31751 + }, + { + "epoch": 1.35, + "grad_norm": 0.427734375, + "learning_rate": 4.752697844551701e-05, + "loss": 2.3667, + "step": 31752 + }, + { + "epoch": 1.35, + "grad_norm": 0.48046875, + "learning_rate": 4.752126831220417e-05, + "loss": 2.4487, + "step": 31753 + }, + { + "epoch": 1.35, + "grad_norm": 0.45703125, + "learning_rate": 4.751555841502642e-05, + "loss": 2.4239, + "step": 31754 + }, + { + "epoch": 1.35, + "grad_norm": 0.48046875, + "learning_rate": 4.750984875400945e-05, + "loss": 2.4769, + "step": 31755 + }, + { + "epoch": 1.35, + "grad_norm": 0.43359375, + "learning_rate": 4.7504139329179045e-05, + "loss": 2.4309, + "step": 31756 + }, + { + "epoch": 1.35, + "grad_norm": 0.4921875, + "learning_rate": 4.749843014056081e-05, + "loss": 2.4784, + "step": 31757 + }, + { + "epoch": 1.35, + "grad_norm": 0.447265625, + "learning_rate": 4.749272118818042e-05, + "loss": 2.5128, + "step": 31758 + }, + { + "epoch": 1.35, + "grad_norm": 0.49609375, + "learning_rate": 4.74870124720636e-05, + "loss": 2.4799, + "step": 31759 + }, + { + "epoch": 1.35, + "grad_norm": 0.490234375, + "learning_rate": 4.748130399223607e-05, + "loss": 2.4001, + "step": 31760 + }, + { + "epoch": 1.35, + "grad_norm": 0.5, + "learning_rate": 4.747559574872343e-05, + "loss": 2.5508, + "step": 31761 + }, + { + "epoch": 1.35, + "grad_norm": 0.384765625, + "learning_rate": 4.746988774155147e-05, + "loss": 2.3719, + "step": 31762 + }, + { + "epoch": 1.35, + "grad_norm": 0.4765625, + "learning_rate": 4.7464179970745756e-05, + "loss": 2.3439, + "step": 31763 + }, + { + "epoch": 1.35, + "grad_norm": 0.421875, + "learning_rate": 4.7458472436332045e-05, + "loss": 2.337, + "step": 31764 + }, + { + "epoch": 1.35, + "grad_norm": 0.47265625, + "learning_rate": 4.745276513833605e-05, + "loss": 2.5376, + "step": 31765 + }, + { + "epoch": 1.35, + "grad_norm": 0.51953125, + "learning_rate": 4.74470580767834e-05, + "loss": 2.4479, + "step": 31766 + }, + { + "epoch": 1.35, + "grad_norm": 0.46484375, + "learning_rate": 4.744135125169975e-05, + "loss": 2.4443, + "step": 31767 + }, + { + "epoch": 1.35, + "grad_norm": 0.51171875, + "learning_rate": 4.743564466311081e-05, + "loss": 2.5459, + "step": 31768 + }, + { + "epoch": 1.35, + "grad_norm": 0.484375, + "learning_rate": 4.742993831104229e-05, + "loss": 2.5839, + "step": 31769 + }, + { + "epoch": 1.35, + "grad_norm": 0.5078125, + "learning_rate": 4.74242321955198e-05, + "loss": 2.3434, + "step": 31770 + }, + { + "epoch": 1.35, + "grad_norm": 0.4453125, + "learning_rate": 4.74185263165691e-05, + "loss": 2.2834, + "step": 31771 + }, + { + "epoch": 1.35, + "grad_norm": 0.427734375, + "learning_rate": 4.7412820674215765e-05, + "loss": 2.4295, + "step": 31772 + }, + { + "epoch": 1.35, + "grad_norm": 0.431640625, + "learning_rate": 4.740711526848556e-05, + "loss": 2.3961, + "step": 31773 + }, + { + "epoch": 1.35, + "grad_norm": 0.46484375, + "learning_rate": 4.740141009940408e-05, + "loss": 2.3954, + "step": 31774 + }, + { + "epoch": 1.35, + "grad_norm": 0.458984375, + "learning_rate": 4.739570516699707e-05, + "loss": 2.4703, + "step": 31775 + }, + { + "epoch": 1.35, + "grad_norm": 0.4765625, + "learning_rate": 4.739000047129012e-05, + "loss": 2.4038, + "step": 31776 + }, + { + "epoch": 1.35, + "grad_norm": 0.47265625, + "learning_rate": 4.738429601230895e-05, + "loss": 2.4736, + "step": 31777 + }, + { + "epoch": 1.35, + "grad_norm": 0.447265625, + "learning_rate": 4.737859179007924e-05, + "loss": 2.5352, + "step": 31778 + }, + { + "epoch": 1.35, + "grad_norm": 0.478515625, + "learning_rate": 4.7372887804626656e-05, + "loss": 2.4381, + "step": 31779 + }, + { + "epoch": 1.35, + "grad_norm": 0.482421875, + "learning_rate": 4.7367184055976754e-05, + "loss": 2.5007, + "step": 31780 + }, + { + "epoch": 1.35, + "grad_norm": 0.44140625, + "learning_rate": 4.736148054415537e-05, + "loss": 2.5893, + "step": 31781 + }, + { + "epoch": 1.35, + "grad_norm": 0.48046875, + "learning_rate": 4.735577726918808e-05, + "loss": 2.5129, + "step": 31782 + }, + { + "epoch": 1.35, + "grad_norm": 0.4296875, + "learning_rate": 4.735007423110053e-05, + "loss": 2.4619, + "step": 31783 + }, + { + "epoch": 1.35, + "grad_norm": 0.4609375, + "learning_rate": 4.734437142991839e-05, + "loss": 2.3848, + "step": 31784 + }, + { + "epoch": 1.35, + "grad_norm": 0.443359375, + "learning_rate": 4.733866886566738e-05, + "loss": 2.4319, + "step": 31785 + }, + { + "epoch": 1.35, + "grad_norm": 0.4453125, + "learning_rate": 4.7332966538373067e-05, + "loss": 2.4838, + "step": 31786 + }, + { + "epoch": 1.35, + "grad_norm": 0.4453125, + "learning_rate": 4.73272644480612e-05, + "loss": 2.3843, + "step": 31787 + }, + { + "epoch": 1.35, + "grad_norm": 0.47265625, + "learning_rate": 4.7321562594757364e-05, + "loss": 2.3348, + "step": 31788 + }, + { + "epoch": 1.35, + "grad_norm": 0.51171875, + "learning_rate": 4.731586097848728e-05, + "loss": 2.3954, + "step": 31789 + }, + { + "epoch": 1.35, + "grad_norm": 0.48828125, + "learning_rate": 4.731015959927653e-05, + "loss": 2.4119, + "step": 31790 + }, + { + "epoch": 1.35, + "grad_norm": 0.4453125, + "learning_rate": 4.730445845715084e-05, + "loss": 2.4778, + "step": 31791 + }, + { + "epoch": 1.35, + "grad_norm": 0.48828125, + "learning_rate": 4.729875755213579e-05, + "loss": 2.4024, + "step": 31792 + }, + { + "epoch": 1.35, + "grad_norm": 0.45703125, + "learning_rate": 4.729305688425707e-05, + "loss": 2.4802, + "step": 31793 + }, + { + "epoch": 1.35, + "grad_norm": 0.396484375, + "learning_rate": 4.7287356453540366e-05, + "loss": 2.4651, + "step": 31794 + }, + { + "epoch": 1.35, + "grad_norm": 0.462890625, + "learning_rate": 4.7281656260011296e-05, + "loss": 2.4402, + "step": 31795 + }, + { + "epoch": 1.35, + "grad_norm": 0.474609375, + "learning_rate": 4.727595630369544e-05, + "loss": 2.4821, + "step": 31796 + }, + { + "epoch": 1.35, + "grad_norm": 0.466796875, + "learning_rate": 4.727025658461859e-05, + "loss": 2.4183, + "step": 31797 + }, + { + "epoch": 1.36, + "grad_norm": 0.4921875, + "learning_rate": 4.7264557102806304e-05, + "loss": 2.4234, + "step": 31798 + }, + { + "epoch": 1.36, + "grad_norm": 0.50390625, + "learning_rate": 4.72588578582842e-05, + "loss": 2.3643, + "step": 31799 + }, + { + "epoch": 1.36, + "grad_norm": 0.466796875, + "learning_rate": 4.725315885107796e-05, + "loss": 2.4535, + "step": 31800 + }, + { + "epoch": 1.36, + "grad_norm": 0.50390625, + "learning_rate": 4.724746008121328e-05, + "loss": 2.5399, + "step": 31801 + }, + { + "epoch": 1.36, + "grad_norm": 0.3828125, + "learning_rate": 4.72417615487157e-05, + "loss": 2.4341, + "step": 31802 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.723606325361095e-05, + "loss": 2.375, + "step": 31803 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.7230365195924594e-05, + "loss": 2.5257, + "step": 31804 + }, + { + "epoch": 1.36, + "grad_norm": 0.515625, + "learning_rate": 4.722466737568236e-05, + "loss": 2.4427, + "step": 31805 + }, + { + "epoch": 1.36, + "grad_norm": 0.484375, + "learning_rate": 4.721896979290978e-05, + "loss": 2.459, + "step": 31806 + }, + { + "epoch": 1.36, + "grad_norm": 0.46484375, + "learning_rate": 4.721327244763261e-05, + "loss": 2.4069, + "step": 31807 + }, + { + "epoch": 1.36, + "grad_norm": 0.515625, + "learning_rate": 4.7207575339876365e-05, + "loss": 2.5313, + "step": 31808 + }, + { + "epoch": 1.36, + "grad_norm": 0.46484375, + "learning_rate": 4.720187846966674e-05, + "loss": 2.4749, + "step": 31809 + }, + { + "epoch": 1.36, + "grad_norm": 0.443359375, + "learning_rate": 4.719618183702941e-05, + "loss": 2.4038, + "step": 31810 + }, + { + "epoch": 1.36, + "grad_norm": 0.404296875, + "learning_rate": 4.719048544198995e-05, + "loss": 2.4389, + "step": 31811 + }, + { + "epoch": 1.36, + "grad_norm": 0.486328125, + "learning_rate": 4.718478928457394e-05, + "loss": 2.5095, + "step": 31812 + }, + { + "epoch": 1.36, + "grad_norm": 0.46484375, + "learning_rate": 4.717909336480716e-05, + "loss": 2.4707, + "step": 31813 + }, + { + "epoch": 1.36, + "grad_norm": 0.46875, + "learning_rate": 4.717339768271514e-05, + "loss": 2.4893, + "step": 31814 + }, + { + "epoch": 1.36, + "grad_norm": 0.486328125, + "learning_rate": 4.71677022383235e-05, + "loss": 2.4343, + "step": 31815 + }, + { + "epoch": 1.36, + "grad_norm": 0.4765625, + "learning_rate": 4.716200703165788e-05, + "loss": 2.3927, + "step": 31816 + }, + { + "epoch": 1.36, + "grad_norm": 0.478515625, + "learning_rate": 4.7156312062743966e-05, + "loss": 2.4418, + "step": 31817 + }, + { + "epoch": 1.36, + "grad_norm": 0.453125, + "learning_rate": 4.7150617331607295e-05, + "loss": 2.3694, + "step": 31818 + }, + { + "epoch": 1.36, + "grad_norm": 0.453125, + "learning_rate": 4.7144922838273566e-05, + "loss": 2.4623, + "step": 31819 + }, + { + "epoch": 1.36, + "grad_norm": 0.4375, + "learning_rate": 4.713922858276834e-05, + "loss": 2.3311, + "step": 31820 + }, + { + "epoch": 1.36, + "grad_norm": 0.466796875, + "learning_rate": 4.713353456511729e-05, + "loss": 2.4282, + "step": 31821 + }, + { + "epoch": 1.36, + "grad_norm": 0.4140625, + "learning_rate": 4.7127840785345986e-05, + "loss": 2.381, + "step": 31822 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.712214724348011e-05, + "loss": 2.5371, + "step": 31823 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.7116453939545215e-05, + "loss": 2.5038, + "step": 31824 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.711076087356694e-05, + "loss": 2.4649, + "step": 31825 + }, + { + "epoch": 1.36, + "grad_norm": 0.431640625, + "learning_rate": 4.7105068045570965e-05, + "loss": 2.4342, + "step": 31826 + }, + { + "epoch": 1.36, + "grad_norm": 0.4921875, + "learning_rate": 4.709937545558284e-05, + "loss": 2.4319, + "step": 31827 + }, + { + "epoch": 1.36, + "grad_norm": 0.49609375, + "learning_rate": 4.709368310362812e-05, + "loss": 2.3696, + "step": 31828 + }, + { + "epoch": 1.36, + "grad_norm": 0.46875, + "learning_rate": 4.708799098973257e-05, + "loss": 2.4659, + "step": 31829 + }, + { + "epoch": 1.36, + "grad_norm": 0.466796875, + "learning_rate": 4.708229911392172e-05, + "loss": 2.4677, + "step": 31830 + }, + { + "epoch": 1.36, + "grad_norm": 0.48046875, + "learning_rate": 4.707660747622115e-05, + "loss": 2.4979, + "step": 31831 + }, + { + "epoch": 1.36, + "grad_norm": 0.427734375, + "learning_rate": 4.70709160766565e-05, + "loss": 2.463, + "step": 31832 + }, + { + "epoch": 1.36, + "grad_norm": 0.498046875, + "learning_rate": 4.7065224915253425e-05, + "loss": 2.5883, + "step": 31833 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.7059533992037455e-05, + "loss": 2.3648, + "step": 31834 + }, + { + "epoch": 1.36, + "grad_norm": 0.5078125, + "learning_rate": 4.7053843307034276e-05, + "loss": 2.454, + "step": 31835 + }, + { + "epoch": 1.36, + "grad_norm": 0.48828125, + "learning_rate": 4.704815286026941e-05, + "loss": 2.487, + "step": 31836 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.7042462651768534e-05, + "loss": 2.4209, + "step": 31837 + }, + { + "epoch": 1.36, + "grad_norm": 0.46875, + "learning_rate": 4.7036772681557194e-05, + "loss": 2.4678, + "step": 31838 + }, + { + "epoch": 1.36, + "grad_norm": 0.490234375, + "learning_rate": 4.703108294966107e-05, + "loss": 2.5217, + "step": 31839 + }, + { + "epoch": 1.36, + "grad_norm": 0.5078125, + "learning_rate": 4.702539345610566e-05, + "loss": 2.5301, + "step": 31840 + }, + { + "epoch": 1.36, + "grad_norm": 0.427734375, + "learning_rate": 4.7019704200916626e-05, + "loss": 2.4009, + "step": 31841 + }, + { + "epoch": 1.36, + "grad_norm": 0.458984375, + "learning_rate": 4.7014015184119594e-05, + "loss": 2.4639, + "step": 31842 + }, + { + "epoch": 1.36, + "grad_norm": 0.5, + "learning_rate": 4.700832640574013e-05, + "loss": 2.56, + "step": 31843 + }, + { + "epoch": 1.36, + "grad_norm": 0.44140625, + "learning_rate": 4.7002637865803764e-05, + "loss": 2.6324, + "step": 31844 + }, + { + "epoch": 1.36, + "grad_norm": 0.474609375, + "learning_rate": 4.6996949564336234e-05, + "loss": 2.5657, + "step": 31845 + }, + { + "epoch": 1.36, + "grad_norm": 0.482421875, + "learning_rate": 4.699126150136305e-05, + "loss": 2.4377, + "step": 31846 + }, + { + "epoch": 1.36, + "grad_norm": 0.4765625, + "learning_rate": 4.698557367690979e-05, + "loss": 2.4051, + "step": 31847 + }, + { + "epoch": 1.36, + "grad_norm": 0.458984375, + "learning_rate": 4.6979886091002064e-05, + "loss": 2.4727, + "step": 31848 + }, + { + "epoch": 1.36, + "grad_norm": 0.431640625, + "learning_rate": 4.697419874366552e-05, + "loss": 2.4965, + "step": 31849 + }, + { + "epoch": 1.36, + "grad_norm": 0.412109375, + "learning_rate": 4.696851163492566e-05, + "loss": 2.4024, + "step": 31850 + }, + { + "epoch": 1.36, + "grad_norm": 0.458984375, + "learning_rate": 4.6962824764808155e-05, + "loss": 2.4368, + "step": 31851 + }, + { + "epoch": 1.36, + "grad_norm": 0.4375, + "learning_rate": 4.695713813333852e-05, + "loss": 2.4572, + "step": 31852 + }, + { + "epoch": 1.36, + "grad_norm": 0.458984375, + "learning_rate": 4.695145174054242e-05, + "loss": 2.5381, + "step": 31853 + }, + { + "epoch": 1.36, + "grad_norm": 0.435546875, + "learning_rate": 4.694576558644535e-05, + "loss": 2.5769, + "step": 31854 + }, + { + "epoch": 1.36, + "grad_norm": 0.515625, + "learning_rate": 4.6940079671073003e-05, + "loss": 2.4982, + "step": 31855 + }, + { + "epoch": 1.36, + "grad_norm": 0.486328125, + "learning_rate": 4.6934393994450856e-05, + "loss": 2.5835, + "step": 31856 + }, + { + "epoch": 1.36, + "grad_norm": 0.51171875, + "learning_rate": 4.692870855660454e-05, + "loss": 2.515, + "step": 31857 + }, + { + "epoch": 1.36, + "grad_norm": 0.4921875, + "learning_rate": 4.692302335755968e-05, + "loss": 2.4336, + "step": 31858 + }, + { + "epoch": 1.36, + "grad_norm": 0.4296875, + "learning_rate": 4.691733839734182e-05, + "loss": 2.5565, + "step": 31859 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.6911653675976465e-05, + "loss": 2.6634, + "step": 31860 + }, + { + "epoch": 1.36, + "grad_norm": 0.478515625, + "learning_rate": 4.690596919348933e-05, + "loss": 2.4183, + "step": 31861 + }, + { + "epoch": 1.36, + "grad_norm": 0.455078125, + "learning_rate": 4.690028494990595e-05, + "loss": 2.5096, + "step": 31862 + }, + { + "epoch": 1.36, + "grad_norm": 0.4921875, + "learning_rate": 4.689460094525182e-05, + "loss": 2.4346, + "step": 31863 + }, + { + "epoch": 1.36, + "grad_norm": 0.431640625, + "learning_rate": 4.688891717955259e-05, + "loss": 2.3819, + "step": 31864 + }, + { + "epoch": 1.36, + "grad_norm": 0.41015625, + "learning_rate": 4.688323365283387e-05, + "loss": 2.4512, + "step": 31865 + }, + { + "epoch": 1.36, + "grad_norm": 0.494140625, + "learning_rate": 4.687755036512114e-05, + "loss": 2.4651, + "step": 31866 + }, + { + "epoch": 1.36, + "grad_norm": 0.455078125, + "learning_rate": 4.6871867316440054e-05, + "loss": 2.3986, + "step": 31867 + }, + { + "epoch": 1.36, + "grad_norm": 0.439453125, + "learning_rate": 4.686618450681611e-05, + "loss": 2.3873, + "step": 31868 + }, + { + "epoch": 1.36, + "grad_norm": 0.466796875, + "learning_rate": 4.686050193627497e-05, + "loss": 2.6621, + "step": 31869 + }, + { + "epoch": 1.36, + "grad_norm": 0.5, + "learning_rate": 4.685481960484211e-05, + "loss": 2.4689, + "step": 31870 + }, + { + "epoch": 1.36, + "grad_norm": 0.478515625, + "learning_rate": 4.684913751254316e-05, + "loss": 2.389, + "step": 31871 + }, + { + "epoch": 1.36, + "grad_norm": 0.484375, + "learning_rate": 4.684345565940363e-05, + "loss": 2.5377, + "step": 31872 + }, + { + "epoch": 1.36, + "grad_norm": 0.458984375, + "learning_rate": 4.683777404544915e-05, + "loss": 2.5763, + "step": 31873 + }, + { + "epoch": 1.36, + "grad_norm": 0.412109375, + "learning_rate": 4.683209267070527e-05, + "loss": 2.2863, + "step": 31874 + }, + { + "epoch": 1.36, + "grad_norm": 0.5390625, + "learning_rate": 4.682641153519756e-05, + "loss": 2.5818, + "step": 31875 + }, + { + "epoch": 1.36, + "grad_norm": 0.46484375, + "learning_rate": 4.682073063895147e-05, + "loss": 2.3979, + "step": 31876 + }, + { + "epoch": 1.36, + "grad_norm": 0.46484375, + "learning_rate": 4.6815049981992745e-05, + "loss": 2.4162, + "step": 31877 + }, + { + "epoch": 1.36, + "grad_norm": 0.47265625, + "learning_rate": 4.680936956434686e-05, + "loss": 2.4088, + "step": 31878 + }, + { + "epoch": 1.36, + "grad_norm": 0.439453125, + "learning_rate": 4.680368938603933e-05, + "loss": 2.3084, + "step": 31879 + }, + { + "epoch": 1.36, + "grad_norm": 0.44921875, + "learning_rate": 4.6798009447095756e-05, + "loss": 2.2852, + "step": 31880 + }, + { + "epoch": 1.36, + "grad_norm": 0.51171875, + "learning_rate": 4.679232974754174e-05, + "loss": 2.4878, + "step": 31881 + }, + { + "epoch": 1.36, + "grad_norm": 0.53125, + "learning_rate": 4.678665028740276e-05, + "loss": 2.4997, + "step": 31882 + }, + { + "epoch": 1.36, + "grad_norm": 0.50390625, + "learning_rate": 4.6780971066704426e-05, + "loss": 2.4099, + "step": 31883 + }, + { + "epoch": 1.36, + "grad_norm": 0.51953125, + "learning_rate": 4.6775292085472256e-05, + "loss": 2.5178, + "step": 31884 + }, + { + "epoch": 1.36, + "grad_norm": 0.48046875, + "learning_rate": 4.676961334373184e-05, + "loss": 2.4915, + "step": 31885 + }, + { + "epoch": 1.36, + "grad_norm": 0.42578125, + "learning_rate": 4.676393484150868e-05, + "loss": 2.5111, + "step": 31886 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.675825657882841e-05, + "loss": 2.6657, + "step": 31887 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.6752578555716473e-05, + "loss": 2.4876, + "step": 31888 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.674690077219848e-05, + "loss": 2.3308, + "step": 31889 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.67412232283e-05, + "loss": 2.3527, + "step": 31890 + }, + { + "epoch": 1.36, + "grad_norm": 0.482421875, + "learning_rate": 4.673554592404656e-05, + "loss": 2.4054, + "step": 31891 + }, + { + "epoch": 1.36, + "grad_norm": 0.453125, + "learning_rate": 4.672986885946366e-05, + "loss": 2.4204, + "step": 31892 + }, + { + "epoch": 1.36, + "grad_norm": 0.53125, + "learning_rate": 4.6724192034576875e-05, + "loss": 2.3592, + "step": 31893 + }, + { + "epoch": 1.36, + "grad_norm": 0.486328125, + "learning_rate": 4.671851544941181e-05, + "loss": 2.4696, + "step": 31894 + }, + { + "epoch": 1.36, + "grad_norm": 0.49609375, + "learning_rate": 4.67128391039939e-05, + "loss": 2.3399, + "step": 31895 + }, + { + "epoch": 1.36, + "grad_norm": 0.455078125, + "learning_rate": 4.670716299834877e-05, + "loss": 2.4593, + "step": 31896 + }, + { + "epoch": 1.36, + "grad_norm": 0.490234375, + "learning_rate": 4.670148713250191e-05, + "loss": 2.5167, + "step": 31897 + }, + { + "epoch": 1.36, + "grad_norm": 0.4921875, + "learning_rate": 4.669581150647888e-05, + "loss": 2.5339, + "step": 31898 + }, + { + "epoch": 1.36, + "grad_norm": 0.50390625, + "learning_rate": 4.669013612030525e-05, + "loss": 2.4707, + "step": 31899 + }, + { + "epoch": 1.36, + "grad_norm": 0.51171875, + "learning_rate": 4.6684460974006514e-05, + "loss": 2.5639, + "step": 31900 + }, + { + "epoch": 1.36, + "grad_norm": 0.423828125, + "learning_rate": 4.667878606760819e-05, + "loss": 2.4013, + "step": 31901 + }, + { + "epoch": 1.36, + "grad_norm": 0.435546875, + "learning_rate": 4.667311140113585e-05, + "loss": 2.3716, + "step": 31902 + }, + { + "epoch": 1.36, + "grad_norm": 0.466796875, + "learning_rate": 4.6667436974615054e-05, + "loss": 2.5142, + "step": 31903 + }, + { + "epoch": 1.36, + "grad_norm": 0.486328125, + "learning_rate": 4.666176278807126e-05, + "loss": 2.5068, + "step": 31904 + }, + { + "epoch": 1.36, + "grad_norm": 0.427734375, + "learning_rate": 4.665608884153004e-05, + "loss": 2.3653, + "step": 31905 + }, + { + "epoch": 1.36, + "grad_norm": 0.466796875, + "learning_rate": 4.665041513501697e-05, + "loss": 2.4981, + "step": 31906 + }, + { + "epoch": 1.36, + "grad_norm": 0.44921875, + "learning_rate": 4.6644741668557525e-05, + "loss": 2.4659, + "step": 31907 + }, + { + "epoch": 1.36, + "grad_norm": 0.462890625, + "learning_rate": 4.6639068442177204e-05, + "loss": 2.582, + "step": 31908 + }, + { + "epoch": 1.36, + "grad_norm": 0.447265625, + "learning_rate": 4.6633395455901585e-05, + "loss": 2.5444, + "step": 31909 + }, + { + "epoch": 1.36, + "grad_norm": 0.5546875, + "learning_rate": 4.66277227097562e-05, + "loss": 2.453, + "step": 31910 + }, + { + "epoch": 1.36, + "grad_norm": 0.466796875, + "learning_rate": 4.662205020376654e-05, + "loss": 2.4882, + "step": 31911 + }, + { + "epoch": 1.36, + "grad_norm": 0.46875, + "learning_rate": 4.661637793795819e-05, + "loss": 2.3509, + "step": 31912 + }, + { + "epoch": 1.36, + "grad_norm": 0.50390625, + "learning_rate": 4.661070591235658e-05, + "loss": 2.3866, + "step": 31913 + }, + { + "epoch": 1.36, + "grad_norm": 0.416015625, + "learning_rate": 4.660503412698728e-05, + "loss": 2.3612, + "step": 31914 + }, + { + "epoch": 1.36, + "grad_norm": 0.486328125, + "learning_rate": 4.659936258187585e-05, + "loss": 2.5411, + "step": 31915 + }, + { + "epoch": 1.36, + "grad_norm": 0.408203125, + "learning_rate": 4.6593691277047776e-05, + "loss": 2.4169, + "step": 31916 + }, + { + "epoch": 1.36, + "grad_norm": 0.4140625, + "learning_rate": 4.658802021252854e-05, + "loss": 2.5019, + "step": 31917 + }, + { + "epoch": 1.36, + "grad_norm": 0.48046875, + "learning_rate": 4.6582349388343685e-05, + "loss": 2.4715, + "step": 31918 + }, + { + "epoch": 1.36, + "grad_norm": 0.462890625, + "learning_rate": 4.657667880451878e-05, + "loss": 2.5291, + "step": 31919 + }, + { + "epoch": 1.36, + "grad_norm": 0.51171875, + "learning_rate": 4.6571008461079255e-05, + "loss": 2.6003, + "step": 31920 + }, + { + "epoch": 1.36, + "grad_norm": 0.423828125, + "learning_rate": 4.656533835805067e-05, + "loss": 2.4044, + "step": 31921 + }, + { + "epoch": 1.36, + "grad_norm": 0.451171875, + "learning_rate": 4.6559668495458566e-05, + "loss": 2.3481, + "step": 31922 + }, + { + "epoch": 1.36, + "grad_norm": 0.431640625, + "learning_rate": 4.655399887332843e-05, + "loss": 2.5264, + "step": 31923 + }, + { + "epoch": 1.36, + "grad_norm": 0.482421875, + "learning_rate": 4.654832949168573e-05, + "loss": 2.5314, + "step": 31924 + }, + { + "epoch": 1.36, + "grad_norm": 0.44921875, + "learning_rate": 4.6542660350556e-05, + "loss": 2.5198, + "step": 31925 + }, + { + "epoch": 1.36, + "grad_norm": 0.49609375, + "learning_rate": 4.653699144996482e-05, + "loss": 2.4191, + "step": 31926 + }, + { + "epoch": 1.36, + "grad_norm": 0.458984375, + "learning_rate": 4.653132278993758e-05, + "loss": 2.461, + "step": 31927 + }, + { + "epoch": 1.36, + "grad_norm": 0.4375, + "learning_rate": 4.65256543704999e-05, + "loss": 2.4052, + "step": 31928 + }, + { + "epoch": 1.36, + "grad_norm": 0.4453125, + "learning_rate": 4.651998619167719e-05, + "loss": 2.446, + "step": 31929 + }, + { + "epoch": 1.36, + "grad_norm": 0.423828125, + "learning_rate": 4.651431825349499e-05, + "loss": 2.435, + "step": 31930 + }, + { + "epoch": 1.36, + "grad_norm": 0.498046875, + "learning_rate": 4.650865055597886e-05, + "loss": 2.5742, + "step": 31931 + }, + { + "epoch": 1.36, + "grad_norm": 0.4609375, + "learning_rate": 4.650298309915425e-05, + "loss": 2.4055, + "step": 31932 + }, + { + "epoch": 1.36, + "grad_norm": 0.43359375, + "learning_rate": 4.6497315883046634e-05, + "loss": 2.4275, + "step": 31933 + }, + { + "epoch": 1.36, + "grad_norm": 0.412109375, + "learning_rate": 4.649164890768153e-05, + "loss": 2.4486, + "step": 31934 + }, + { + "epoch": 1.36, + "grad_norm": 0.421875, + "learning_rate": 4.6485982173084494e-05, + "loss": 2.3874, + "step": 31935 + }, + { + "epoch": 1.36, + "grad_norm": 0.43359375, + "learning_rate": 4.648031567928095e-05, + "loss": 2.4409, + "step": 31936 + }, + { + "epoch": 1.36, + "grad_norm": 0.455078125, + "learning_rate": 4.647464942629643e-05, + "loss": 2.4879, + "step": 31937 + }, + { + "epoch": 1.36, + "grad_norm": 0.51953125, + "learning_rate": 4.646898341415645e-05, + "loss": 2.5718, + "step": 31938 + }, + { + "epoch": 1.36, + "grad_norm": 0.498046875, + "learning_rate": 4.646331764288649e-05, + "loss": 2.4352, + "step": 31939 + }, + { + "epoch": 1.36, + "grad_norm": 0.44921875, + "learning_rate": 4.6457652112511986e-05, + "loss": 2.3493, + "step": 31940 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.645198682305849e-05, + "loss": 2.3631, + "step": 31941 + }, + { + "epoch": 1.36, + "grad_norm": 0.431640625, + "learning_rate": 4.644632177455152e-05, + "loss": 2.5164, + "step": 31942 + }, + { + "epoch": 1.36, + "grad_norm": 0.421875, + "learning_rate": 4.6440656967016496e-05, + "loss": 2.3351, + "step": 31943 + }, + { + "epoch": 1.36, + "grad_norm": 0.53125, + "learning_rate": 4.643499240047897e-05, + "loss": 2.4741, + "step": 31944 + }, + { + "epoch": 1.36, + "grad_norm": 0.41015625, + "learning_rate": 4.642932807496437e-05, + "loss": 2.5145, + "step": 31945 + }, + { + "epoch": 1.36, + "grad_norm": 0.400390625, + "learning_rate": 4.6423663990498224e-05, + "loss": 2.3892, + "step": 31946 + }, + { + "epoch": 1.36, + "grad_norm": 0.427734375, + "learning_rate": 4.6418000147106037e-05, + "loss": 2.3353, + "step": 31947 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.6412336544813273e-05, + "loss": 2.4253, + "step": 31948 + }, + { + "epoch": 1.36, + "grad_norm": 0.474609375, + "learning_rate": 4.6406673183645366e-05, + "loss": 2.3036, + "step": 31949 + }, + { + "epoch": 1.36, + "grad_norm": 0.4765625, + "learning_rate": 4.640101006362785e-05, + "loss": 2.3884, + "step": 31950 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.639534718478623e-05, + "loss": 2.4605, + "step": 31951 + }, + { + "epoch": 1.36, + "grad_norm": 0.48828125, + "learning_rate": 4.638968454714593e-05, + "loss": 2.4822, + "step": 31952 + }, + { + "epoch": 1.36, + "grad_norm": 0.458984375, + "learning_rate": 4.6384022150732456e-05, + "loss": 2.3953, + "step": 31953 + }, + { + "epoch": 1.36, + "grad_norm": 0.51171875, + "learning_rate": 4.637835999557132e-05, + "loss": 2.404, + "step": 31954 + }, + { + "epoch": 1.36, + "grad_norm": 0.51953125, + "learning_rate": 4.637269808168797e-05, + "loss": 2.5433, + "step": 31955 + }, + { + "epoch": 1.36, + "grad_norm": 0.484375, + "learning_rate": 4.6367036409107844e-05, + "loss": 2.4389, + "step": 31956 + }, + { + "epoch": 1.36, + "grad_norm": 0.494140625, + "learning_rate": 4.6361374977856456e-05, + "loss": 2.4672, + "step": 31957 + }, + { + "epoch": 1.36, + "grad_norm": 0.46484375, + "learning_rate": 4.635571378795931e-05, + "loss": 2.3162, + "step": 31958 + }, + { + "epoch": 1.36, + "grad_norm": 0.50390625, + "learning_rate": 4.63500528394418e-05, + "loss": 2.4404, + "step": 31959 + }, + { + "epoch": 1.36, + "grad_norm": 0.47265625, + "learning_rate": 4.6344392132329505e-05, + "loss": 2.6045, + "step": 31960 + }, + { + "epoch": 1.36, + "grad_norm": 0.44921875, + "learning_rate": 4.633873166664778e-05, + "loss": 2.4634, + "step": 31961 + }, + { + "epoch": 1.36, + "grad_norm": 0.453125, + "learning_rate": 4.6333071442422174e-05, + "loss": 2.4642, + "step": 31962 + }, + { + "epoch": 1.36, + "grad_norm": 0.451171875, + "learning_rate": 4.632741145967815e-05, + "loss": 2.3469, + "step": 31963 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.632175171844116e-05, + "loss": 2.3815, + "step": 31964 + }, + { + "epoch": 1.36, + "grad_norm": 0.462890625, + "learning_rate": 4.6316092218736636e-05, + "loss": 2.3112, + "step": 31965 + }, + { + "epoch": 1.36, + "grad_norm": 0.408203125, + "learning_rate": 4.631043296059009e-05, + "loss": 2.5233, + "step": 31966 + }, + { + "epoch": 1.36, + "grad_norm": 0.4453125, + "learning_rate": 4.6304773944027006e-05, + "loss": 2.3603, + "step": 31967 + }, + { + "epoch": 1.36, + "grad_norm": 0.482421875, + "learning_rate": 4.629911516907277e-05, + "loss": 2.3844, + "step": 31968 + }, + { + "epoch": 1.36, + "grad_norm": 0.4453125, + "learning_rate": 4.62934566357529e-05, + "loss": 2.4331, + "step": 31969 + }, + { + "epoch": 1.36, + "grad_norm": 0.49609375, + "learning_rate": 4.628779834409288e-05, + "loss": 2.4934, + "step": 31970 + }, + { + "epoch": 1.36, + "grad_norm": 0.4921875, + "learning_rate": 4.6282140294118135e-05, + "loss": 2.4127, + "step": 31971 + }, + { + "epoch": 1.36, + "grad_norm": 0.416015625, + "learning_rate": 4.627648248585409e-05, + "loss": 2.463, + "step": 31972 + }, + { + "epoch": 1.36, + "grad_norm": 0.51171875, + "learning_rate": 4.627082491932625e-05, + "loss": 2.4939, + "step": 31973 + }, + { + "epoch": 1.36, + "grad_norm": 0.359375, + "learning_rate": 4.626516759456009e-05, + "loss": 2.4004, + "step": 31974 + }, + { + "epoch": 1.36, + "grad_norm": 0.392578125, + "learning_rate": 4.6259510511581006e-05, + "loss": 2.5566, + "step": 31975 + }, + { + "epoch": 1.36, + "grad_norm": 0.453125, + "learning_rate": 4.6253853670414515e-05, + "loss": 2.571, + "step": 31976 + }, + { + "epoch": 1.36, + "grad_norm": 0.43359375, + "learning_rate": 4.624819707108601e-05, + "loss": 2.5322, + "step": 31977 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.6242540713620975e-05, + "loss": 2.4372, + "step": 31978 + }, + { + "epoch": 1.36, + "grad_norm": 0.498046875, + "learning_rate": 4.6236884598044896e-05, + "loss": 2.582, + "step": 31979 + }, + { + "epoch": 1.36, + "grad_norm": 0.50390625, + "learning_rate": 4.623122872438318e-05, + "loss": 2.4955, + "step": 31980 + }, + { + "epoch": 1.36, + "grad_norm": 0.494140625, + "learning_rate": 4.622557309266126e-05, + "loss": 2.5455, + "step": 31981 + }, + { + "epoch": 1.36, + "grad_norm": 0.5, + "learning_rate": 4.621991770290459e-05, + "loss": 2.4999, + "step": 31982 + }, + { + "epoch": 1.36, + "grad_norm": 0.484375, + "learning_rate": 4.621426255513869e-05, + "loss": 2.6225, + "step": 31983 + }, + { + "epoch": 1.36, + "grad_norm": 0.44140625, + "learning_rate": 4.620860764938891e-05, + "loss": 2.4927, + "step": 31984 + }, + { + "epoch": 1.36, + "grad_norm": 0.490234375, + "learning_rate": 4.6202952985680735e-05, + "loss": 2.4795, + "step": 31985 + }, + { + "epoch": 1.36, + "grad_norm": 0.44921875, + "learning_rate": 4.619729856403966e-05, + "loss": 2.6183, + "step": 31986 + }, + { + "epoch": 1.36, + "grad_norm": 0.5, + "learning_rate": 4.6191644384491064e-05, + "loss": 2.4984, + "step": 31987 + }, + { + "epoch": 1.36, + "grad_norm": 0.45703125, + "learning_rate": 4.618599044706038e-05, + "loss": 2.4691, + "step": 31988 + }, + { + "epoch": 1.36, + "grad_norm": 0.52734375, + "learning_rate": 4.6180336751773055e-05, + "loss": 2.5207, + "step": 31989 + }, + { + "epoch": 1.36, + "grad_norm": 0.453125, + "learning_rate": 4.61746832986546e-05, + "loss": 2.4594, + "step": 31990 + }, + { + "epoch": 1.36, + "grad_norm": 0.44140625, + "learning_rate": 4.6169030087730347e-05, + "loss": 2.4377, + "step": 31991 + }, + { + "epoch": 1.36, + "grad_norm": 0.3984375, + "learning_rate": 4.616337711902583e-05, + "loss": 2.5279, + "step": 31992 + }, + { + "epoch": 1.36, + "grad_norm": 0.421875, + "learning_rate": 4.6157724392566404e-05, + "loss": 2.5027, + "step": 31993 + }, + { + "epoch": 1.36, + "grad_norm": 0.498046875, + "learning_rate": 4.615207190837755e-05, + "loss": 2.3493, + "step": 31994 + }, + { + "epoch": 1.36, + "grad_norm": 0.49609375, + "learning_rate": 4.614641966648472e-05, + "loss": 2.5781, + "step": 31995 + }, + { + "epoch": 1.36, + "grad_norm": 0.408203125, + "learning_rate": 4.614076766691332e-05, + "loss": 2.5195, + "step": 31996 + }, + { + "epoch": 1.36, + "grad_norm": 0.5078125, + "learning_rate": 4.613511590968874e-05, + "loss": 2.3945, + "step": 31997 + }, + { + "epoch": 1.36, + "grad_norm": 0.3984375, + "learning_rate": 4.612946439483646e-05, + "loss": 2.4908, + "step": 31998 + }, + { + "epoch": 1.36, + "grad_norm": 0.498046875, + "learning_rate": 4.6123813122381945e-05, + "loss": 2.5393, + "step": 31999 + }, + { + "epoch": 1.36, + "grad_norm": 0.53125, + "learning_rate": 4.6118162092350536e-05, + "loss": 2.5862, + "step": 32000 + }, + { + "epoch": 1.36, + "grad_norm": 0.52734375, + "learning_rate": 4.611251130476771e-05, + "loss": 2.5352, + "step": 32001 + }, + { + "epoch": 1.36, + "grad_norm": 0.494140625, + "learning_rate": 4.610686075965893e-05, + "loss": 2.5041, + "step": 32002 + }, + { + "epoch": 1.36, + "grad_norm": 0.46484375, + "learning_rate": 4.610121045704958e-05, + "loss": 2.3924, + "step": 32003 + }, + { + "epoch": 1.36, + "grad_norm": 0.46875, + "learning_rate": 4.6095560396965044e-05, + "loss": 2.4606, + "step": 32004 + }, + { + "epoch": 1.36, + "grad_norm": 0.4453125, + "learning_rate": 4.6089910579430785e-05, + "loss": 2.3743, + "step": 32005 + }, + { + "epoch": 1.36, + "grad_norm": 0.462890625, + "learning_rate": 4.608426100447226e-05, + "loss": 2.4372, + "step": 32006 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.6078611672114836e-05, + "loss": 2.4611, + "step": 32007 + }, + { + "epoch": 1.36, + "grad_norm": 0.51171875, + "learning_rate": 4.6072962582383974e-05, + "loss": 2.5516, + "step": 32008 + }, + { + "epoch": 1.36, + "grad_norm": 0.447265625, + "learning_rate": 4.606731373530504e-05, + "loss": 2.4881, + "step": 32009 + }, + { + "epoch": 1.36, + "grad_norm": 0.435546875, + "learning_rate": 4.6061665130903484e-05, + "loss": 2.3646, + "step": 32010 + }, + { + "epoch": 1.36, + "grad_norm": 0.443359375, + "learning_rate": 4.6056016769204766e-05, + "loss": 2.504, + "step": 32011 + }, + { + "epoch": 1.36, + "grad_norm": 0.4609375, + "learning_rate": 4.6050368650234255e-05, + "loss": 2.4004, + "step": 32012 + }, + { + "epoch": 1.36, + "grad_norm": 0.47265625, + "learning_rate": 4.604472077401732e-05, + "loss": 2.4965, + "step": 32013 + }, + { + "epoch": 1.36, + "grad_norm": 0.52734375, + "learning_rate": 4.603907314057943e-05, + "loss": 2.464, + "step": 32014 + }, + { + "epoch": 1.36, + "grad_norm": 0.498046875, + "learning_rate": 4.603342574994604e-05, + "loss": 2.4421, + "step": 32015 + }, + { + "epoch": 1.36, + "grad_norm": 0.490234375, + "learning_rate": 4.602777860214247e-05, + "loss": 2.4234, + "step": 32016 + }, + { + "epoch": 1.36, + "grad_norm": 0.498046875, + "learning_rate": 4.6022131697194194e-05, + "loss": 2.7088, + "step": 32017 + }, + { + "epoch": 1.36, + "grad_norm": 0.48828125, + "learning_rate": 4.601648503512658e-05, + "loss": 2.4749, + "step": 32018 + }, + { + "epoch": 1.36, + "grad_norm": 0.453125, + "learning_rate": 4.601083861596508e-05, + "loss": 2.3285, + "step": 32019 + }, + { + "epoch": 1.36, + "grad_norm": 0.474609375, + "learning_rate": 4.6005192439735036e-05, + "loss": 2.4434, + "step": 32020 + }, + { + "epoch": 1.36, + "grad_norm": 0.455078125, + "learning_rate": 4.599954650646194e-05, + "loss": 2.3005, + "step": 32021 + }, + { + "epoch": 1.36, + "grad_norm": 0.482421875, + "learning_rate": 4.599390081617111e-05, + "loss": 2.3194, + "step": 32022 + }, + { + "epoch": 1.36, + "grad_norm": 0.490234375, + "learning_rate": 4.598825536888799e-05, + "loss": 2.3692, + "step": 32023 + }, + { + "epoch": 1.36, + "grad_norm": 0.482421875, + "learning_rate": 4.598261016463801e-05, + "loss": 2.3296, + "step": 32024 + }, + { + "epoch": 1.36, + "grad_norm": 0.470703125, + "learning_rate": 4.5976965203446555e-05, + "loss": 2.5872, + "step": 32025 + }, + { + "epoch": 1.36, + "grad_norm": 0.5, + "learning_rate": 4.597132048533893e-05, + "loss": 2.5212, + "step": 32026 + }, + { + "epoch": 1.36, + "grad_norm": 0.462890625, + "learning_rate": 4.5965676010340694e-05, + "loss": 2.4497, + "step": 32027 + }, + { + "epoch": 1.36, + "grad_norm": 0.447265625, + "learning_rate": 4.5960031778477175e-05, + "loss": 2.4648, + "step": 32028 + }, + { + "epoch": 1.36, + "grad_norm": 0.482421875, + "learning_rate": 4.595438778977372e-05, + "loss": 2.3506, + "step": 32029 + }, + { + "epoch": 1.36, + "grad_norm": 0.5, + "learning_rate": 4.594874404425577e-05, + "loss": 2.5886, + "step": 32030 + }, + { + "epoch": 1.36, + "grad_norm": 0.46875, + "learning_rate": 4.594310054194876e-05, + "loss": 2.4488, + "step": 32031 + }, + { + "epoch": 1.36, + "grad_norm": 0.443359375, + "learning_rate": 4.593745728287799e-05, + "loss": 2.4702, + "step": 32032 + }, + { + "epoch": 1.37, + "grad_norm": 0.431640625, + "learning_rate": 4.5931814267068964e-05, + "loss": 2.5062, + "step": 32033 + }, + { + "epoch": 1.37, + "grad_norm": 0.4453125, + "learning_rate": 4.5926171494546964e-05, + "loss": 2.4131, + "step": 32034 + }, + { + "epoch": 1.37, + "grad_norm": 0.498046875, + "learning_rate": 4.5920528965337475e-05, + "loss": 2.4442, + "step": 32035 + }, + { + "epoch": 1.37, + "grad_norm": 0.4921875, + "learning_rate": 4.591488667946579e-05, + "loss": 2.5161, + "step": 32036 + }, + { + "epoch": 1.37, + "grad_norm": 0.53515625, + "learning_rate": 4.59092446369574e-05, + "loss": 2.548, + "step": 32037 + }, + { + "epoch": 1.37, + "grad_norm": 0.48828125, + "learning_rate": 4.59036028378376e-05, + "loss": 2.332, + "step": 32038 + }, + { + "epoch": 1.37, + "grad_norm": 0.451171875, + "learning_rate": 4.589796128213181e-05, + "loss": 2.5357, + "step": 32039 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.589231996986547e-05, + "loss": 2.261, + "step": 32040 + }, + { + "epoch": 1.37, + "grad_norm": 0.46875, + "learning_rate": 4.588667890106391e-05, + "loss": 2.5043, + "step": 32041 + }, + { + "epoch": 1.37, + "grad_norm": 0.51953125, + "learning_rate": 4.588103807575245e-05, + "loss": 2.4632, + "step": 32042 + }, + { + "epoch": 1.37, + "grad_norm": 0.48046875, + "learning_rate": 4.5875397493956604e-05, + "loss": 2.4939, + "step": 32043 + }, + { + "epoch": 1.37, + "grad_norm": 0.49609375, + "learning_rate": 4.58697571557017e-05, + "loss": 2.4778, + "step": 32044 + }, + { + "epoch": 1.37, + "grad_norm": 0.455078125, + "learning_rate": 4.586411706101307e-05, + "loss": 2.4267, + "step": 32045 + }, + { + "epoch": 1.37, + "grad_norm": 0.4609375, + "learning_rate": 4.5858477209916115e-05, + "loss": 2.4293, + "step": 32046 + }, + { + "epoch": 1.37, + "grad_norm": 0.48828125, + "learning_rate": 4.5852837602436284e-05, + "loss": 2.5295, + "step": 32047 + }, + { + "epoch": 1.37, + "grad_norm": 0.466796875, + "learning_rate": 4.5847198238598846e-05, + "loss": 2.4744, + "step": 32048 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.584155911842925e-05, + "loss": 2.4828, + "step": 32049 + }, + { + "epoch": 1.37, + "grad_norm": 0.4609375, + "learning_rate": 4.5835920241952834e-05, + "loss": 2.4343, + "step": 32050 + }, + { + "epoch": 1.37, + "grad_norm": 0.50390625, + "learning_rate": 4.5830281609195e-05, + "loss": 2.487, + "step": 32051 + }, + { + "epoch": 1.37, + "grad_norm": 0.45703125, + "learning_rate": 4.582464322018107e-05, + "loss": 2.3704, + "step": 32052 + }, + { + "epoch": 1.37, + "grad_norm": 0.5234375, + "learning_rate": 4.581900507493649e-05, + "loss": 2.4911, + "step": 32053 + }, + { + "epoch": 1.37, + "grad_norm": 0.44921875, + "learning_rate": 4.5813367173486544e-05, + "loss": 2.4746, + "step": 32054 + }, + { + "epoch": 1.37, + "grad_norm": 0.466796875, + "learning_rate": 4.580772951585664e-05, + "loss": 2.4727, + "step": 32055 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.580209210207219e-05, + "loss": 2.427, + "step": 32056 + }, + { + "epoch": 1.37, + "grad_norm": 0.46875, + "learning_rate": 4.5796454932158514e-05, + "loss": 2.4851, + "step": 32057 + }, + { + "epoch": 1.37, + "grad_norm": 0.5234375, + "learning_rate": 4.579081800614091e-05, + "loss": 2.6526, + "step": 32058 + }, + { + "epoch": 1.37, + "grad_norm": 0.4765625, + "learning_rate": 4.578518132404489e-05, + "loss": 2.399, + "step": 32059 + }, + { + "epoch": 1.37, + "grad_norm": 0.5, + "learning_rate": 4.5779544885895745e-05, + "loss": 2.5873, + "step": 32060 + }, + { + "epoch": 1.37, + "grad_norm": 0.4296875, + "learning_rate": 4.577390869171879e-05, + "loss": 2.4612, + "step": 32061 + }, + { + "epoch": 1.37, + "grad_norm": 0.4765625, + "learning_rate": 4.576827274153943e-05, + "loss": 2.3533, + "step": 32062 + }, + { + "epoch": 1.37, + "grad_norm": 0.484375, + "learning_rate": 4.576263703538306e-05, + "loss": 2.3672, + "step": 32063 + }, + { + "epoch": 1.37, + "grad_norm": 0.4609375, + "learning_rate": 4.575700157327496e-05, + "loss": 2.2872, + "step": 32064 + }, + { + "epoch": 1.37, + "grad_norm": 0.44921875, + "learning_rate": 4.575136635524057e-05, + "loss": 2.3793, + "step": 32065 + }, + { + "epoch": 1.37, + "grad_norm": 0.5234375, + "learning_rate": 4.5745731381305166e-05, + "loss": 2.484, + "step": 32066 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.5740096651494194e-05, + "loss": 2.388, + "step": 32067 + }, + { + "epoch": 1.37, + "grad_norm": 0.4375, + "learning_rate": 4.57344621658329e-05, + "loss": 2.4799, + "step": 32068 + }, + { + "epoch": 1.37, + "grad_norm": 0.490234375, + "learning_rate": 4.572882792434674e-05, + "loss": 2.4487, + "step": 32069 + }, + { + "epoch": 1.37, + "grad_norm": 0.462890625, + "learning_rate": 4.572319392706098e-05, + "loss": 2.3989, + "step": 32070 + }, + { + "epoch": 1.37, + "grad_norm": 0.44921875, + "learning_rate": 4.571756017400102e-05, + "loss": 2.4481, + "step": 32071 + }, + { + "epoch": 1.37, + "grad_norm": 0.451171875, + "learning_rate": 4.571192666519224e-05, + "loss": 2.4763, + "step": 32072 + }, + { + "epoch": 1.37, + "grad_norm": 0.498046875, + "learning_rate": 4.570629340065994e-05, + "loss": 2.4009, + "step": 32073 + }, + { + "epoch": 1.37, + "grad_norm": 0.5234375, + "learning_rate": 4.5700660380429403e-05, + "loss": 2.3781, + "step": 32074 + }, + { + "epoch": 1.37, + "grad_norm": 0.53515625, + "learning_rate": 4.569502760452613e-05, + "loss": 2.4947, + "step": 32075 + }, + { + "epoch": 1.37, + "grad_norm": 0.470703125, + "learning_rate": 4.568939507297538e-05, + "loss": 2.3937, + "step": 32076 + }, + { + "epoch": 1.37, + "grad_norm": 0.45703125, + "learning_rate": 4.5683762785802474e-05, + "loss": 2.4895, + "step": 32077 + }, + { + "epoch": 1.37, + "grad_norm": 0.4375, + "learning_rate": 4.5678130743032786e-05, + "loss": 2.4868, + "step": 32078 + }, + { + "epoch": 1.37, + "grad_norm": 0.484375, + "learning_rate": 4.56724989446917e-05, + "loss": 2.4349, + "step": 32079 + }, + { + "epoch": 1.37, + "grad_norm": 0.404296875, + "learning_rate": 4.5666867390804466e-05, + "loss": 2.4117, + "step": 32080 + }, + { + "epoch": 1.37, + "grad_norm": 0.443359375, + "learning_rate": 4.5661236081396506e-05, + "loss": 2.409, + "step": 32081 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.56556050164931e-05, + "loss": 2.406, + "step": 32082 + }, + { + "epoch": 1.37, + "grad_norm": 0.49609375, + "learning_rate": 4.564997419611964e-05, + "loss": 2.3851, + "step": 32083 + }, + { + "epoch": 1.37, + "grad_norm": 0.41796875, + "learning_rate": 4.56443436203014e-05, + "loss": 2.3514, + "step": 32084 + }, + { + "epoch": 1.37, + "grad_norm": 0.515625, + "learning_rate": 4.5638713289063786e-05, + "loss": 2.4234, + "step": 32085 + }, + { + "epoch": 1.37, + "grad_norm": 0.43359375, + "learning_rate": 4.563308320243206e-05, + "loss": 2.251, + "step": 32086 + }, + { + "epoch": 1.37, + "grad_norm": 0.5625, + "learning_rate": 4.562745336043158e-05, + "loss": 2.6256, + "step": 32087 + }, + { + "epoch": 1.37, + "grad_norm": 0.396484375, + "learning_rate": 4.5621823763087734e-05, + "loss": 2.3484, + "step": 32088 + }, + { + "epoch": 1.37, + "grad_norm": 0.455078125, + "learning_rate": 4.5616194410425805e-05, + "loss": 2.4761, + "step": 32089 + }, + { + "epoch": 1.37, + "grad_norm": 0.451171875, + "learning_rate": 4.561056530247105e-05, + "loss": 2.4023, + "step": 32090 + }, + { + "epoch": 1.37, + "grad_norm": 0.44140625, + "learning_rate": 4.560493643924896e-05, + "loss": 2.4077, + "step": 32091 + }, + { + "epoch": 1.37, + "grad_norm": 0.4296875, + "learning_rate": 4.559930782078477e-05, + "loss": 2.3813, + "step": 32092 + }, + { + "epoch": 1.37, + "grad_norm": 0.400390625, + "learning_rate": 4.559367944710378e-05, + "loss": 2.2893, + "step": 32093 + }, + { + "epoch": 1.37, + "grad_norm": 0.458984375, + "learning_rate": 4.558805131823134e-05, + "loss": 2.5049, + "step": 32094 + }, + { + "epoch": 1.37, + "grad_norm": 0.404296875, + "learning_rate": 4.558242343419283e-05, + "loss": 2.2806, + "step": 32095 + }, + { + "epoch": 1.37, + "grad_norm": 0.53515625, + "learning_rate": 4.557679579501348e-05, + "loss": 2.4371, + "step": 32096 + }, + { + "epoch": 1.37, + "grad_norm": 0.466796875, + "learning_rate": 4.557116840071871e-05, + "loss": 2.4727, + "step": 32097 + }, + { + "epoch": 1.37, + "grad_norm": 0.466796875, + "learning_rate": 4.5565541251333745e-05, + "loss": 2.403, + "step": 32098 + }, + { + "epoch": 1.37, + "grad_norm": 0.421875, + "learning_rate": 4.5559914346883955e-05, + "loss": 2.5879, + "step": 32099 + }, + { + "epoch": 1.37, + "grad_norm": 0.52734375, + "learning_rate": 4.555428768739469e-05, + "loss": 2.5246, + "step": 32100 + }, + { + "epoch": 1.37, + "grad_norm": 0.5078125, + "learning_rate": 4.5548661272891234e-05, + "loss": 2.5363, + "step": 32101 + }, + { + "epoch": 1.37, + "grad_norm": 0.478515625, + "learning_rate": 4.554303510339887e-05, + "loss": 2.5161, + "step": 32102 + }, + { + "epoch": 1.37, + "grad_norm": 0.412109375, + "learning_rate": 4.5537409178942934e-05, + "loss": 2.3111, + "step": 32103 + }, + { + "epoch": 1.37, + "grad_norm": 0.4140625, + "learning_rate": 4.55317834995488e-05, + "loss": 2.489, + "step": 32104 + }, + { + "epoch": 1.37, + "grad_norm": 0.423828125, + "learning_rate": 4.552615806524173e-05, + "loss": 2.4493, + "step": 32105 + }, + { + "epoch": 1.37, + "grad_norm": 0.51953125, + "learning_rate": 4.552053287604696e-05, + "loss": 2.6118, + "step": 32106 + }, + { + "epoch": 1.37, + "grad_norm": 0.458984375, + "learning_rate": 4.551490793198997e-05, + "loss": 2.335, + "step": 32107 + }, + { + "epoch": 1.37, + "grad_norm": 0.41796875, + "learning_rate": 4.550928323309598e-05, + "loss": 2.475, + "step": 32108 + }, + { + "epoch": 1.37, + "grad_norm": 0.466796875, + "learning_rate": 4.550365877939026e-05, + "loss": 2.3882, + "step": 32109 + }, + { + "epoch": 1.37, + "grad_norm": 0.48828125, + "learning_rate": 4.549803457089815e-05, + "loss": 2.5064, + "step": 32110 + }, + { + "epoch": 1.37, + "grad_norm": 0.53125, + "learning_rate": 4.5492410607645016e-05, + "loss": 2.4927, + "step": 32111 + }, + { + "epoch": 1.37, + "grad_norm": 0.50390625, + "learning_rate": 4.5486786889656073e-05, + "loss": 2.4921, + "step": 32112 + }, + { + "epoch": 1.37, + "grad_norm": 0.546875, + "learning_rate": 4.5481163416956705e-05, + "loss": 2.5354, + "step": 32113 + }, + { + "epoch": 1.37, + "grad_norm": 0.41796875, + "learning_rate": 4.547554018957214e-05, + "loss": 2.4228, + "step": 32114 + }, + { + "epoch": 1.37, + "grad_norm": 0.478515625, + "learning_rate": 4.546991720752771e-05, + "loss": 2.4432, + "step": 32115 + }, + { + "epoch": 1.37, + "grad_norm": 0.41015625, + "learning_rate": 4.546429447084878e-05, + "loss": 2.4307, + "step": 32116 + }, + { + "epoch": 1.37, + "grad_norm": 0.447265625, + "learning_rate": 4.5458671979560576e-05, + "loss": 2.4342, + "step": 32117 + }, + { + "epoch": 1.37, + "grad_norm": 0.45703125, + "learning_rate": 4.545304973368839e-05, + "loss": 2.5421, + "step": 32118 + }, + { + "epoch": 1.37, + "grad_norm": 0.45703125, + "learning_rate": 4.544742773325754e-05, + "loss": 2.5256, + "step": 32119 + }, + { + "epoch": 1.37, + "grad_norm": 0.435546875, + "learning_rate": 4.5441805978293363e-05, + "loss": 2.5696, + "step": 32120 + }, + { + "epoch": 1.37, + "grad_norm": 0.470703125, + "learning_rate": 4.543618446882113e-05, + "loss": 2.4851, + "step": 32121 + }, + { + "epoch": 1.37, + "grad_norm": 0.494140625, + "learning_rate": 4.543056320486604e-05, + "loss": 2.4498, + "step": 32122 + }, + { + "epoch": 1.37, + "grad_norm": 0.466796875, + "learning_rate": 4.5424942186453555e-05, + "loss": 2.4485, + "step": 32123 + }, + { + "epoch": 1.37, + "grad_norm": 0.4296875, + "learning_rate": 4.541932141360888e-05, + "loss": 2.3258, + "step": 32124 + }, + { + "epoch": 1.37, + "grad_norm": 0.55078125, + "learning_rate": 4.541370088635728e-05, + "loss": 2.498, + "step": 32125 + }, + { + "epoch": 1.37, + "grad_norm": 0.423828125, + "learning_rate": 4.540808060472408e-05, + "loss": 2.3601, + "step": 32126 + }, + { + "epoch": 1.37, + "grad_norm": 0.4296875, + "learning_rate": 4.5402460568734596e-05, + "loss": 2.554, + "step": 32127 + }, + { + "epoch": 1.37, + "grad_norm": 0.5, + "learning_rate": 4.539684077841405e-05, + "loss": 2.5486, + "step": 32128 + }, + { + "epoch": 1.37, + "grad_norm": 0.5390625, + "learning_rate": 4.5391221233787806e-05, + "loss": 2.6757, + "step": 32129 + }, + { + "epoch": 1.37, + "grad_norm": 0.4765625, + "learning_rate": 4.538560193488106e-05, + "loss": 2.5234, + "step": 32130 + }, + { + "epoch": 1.37, + "grad_norm": 0.4453125, + "learning_rate": 4.537998288171916e-05, + "loss": 2.4171, + "step": 32131 + }, + { + "epoch": 1.37, + "grad_norm": 0.46484375, + "learning_rate": 4.537436407432741e-05, + "loss": 2.56, + "step": 32132 + }, + { + "epoch": 1.37, + "grad_norm": 0.447265625, + "learning_rate": 4.536874551273106e-05, + "loss": 2.4387, + "step": 32133 + }, + { + "epoch": 1.37, + "grad_norm": 0.470703125, + "learning_rate": 4.536312719695535e-05, + "loss": 2.3193, + "step": 32134 + }, + { + "epoch": 1.37, + "grad_norm": 0.4921875, + "learning_rate": 4.53575091270256e-05, + "loss": 2.403, + "step": 32135 + }, + { + "epoch": 1.37, + "grad_norm": 0.46875, + "learning_rate": 4.535189130296712e-05, + "loss": 2.5557, + "step": 32136 + }, + { + "epoch": 1.37, + "grad_norm": 0.46875, + "learning_rate": 4.534627372480517e-05, + "loss": 2.3804, + "step": 32137 + }, + { + "epoch": 1.37, + "grad_norm": 0.486328125, + "learning_rate": 4.534065639256497e-05, + "loss": 2.4698, + "step": 32138 + }, + { + "epoch": 1.37, + "grad_norm": 0.4296875, + "learning_rate": 4.5335039306271846e-05, + "loss": 2.3918, + "step": 32139 + }, + { + "epoch": 1.37, + "grad_norm": 0.462890625, + "learning_rate": 4.532942246595109e-05, + "loss": 2.484, + "step": 32140 + }, + { + "epoch": 1.37, + "grad_norm": 0.427734375, + "learning_rate": 4.532380587162793e-05, + "loss": 2.5327, + "step": 32141 + }, + { + "epoch": 1.37, + "grad_norm": 0.44921875, + "learning_rate": 4.531818952332769e-05, + "loss": 2.5181, + "step": 32142 + }, + { + "epoch": 1.37, + "grad_norm": 0.5234375, + "learning_rate": 4.5312573421075586e-05, + "loss": 2.5282, + "step": 32143 + }, + { + "epoch": 1.37, + "grad_norm": 0.53515625, + "learning_rate": 4.530695756489691e-05, + "loss": 2.4877, + "step": 32144 + }, + { + "epoch": 1.37, + "grad_norm": 0.46875, + "learning_rate": 4.530134195481698e-05, + "loss": 2.5934, + "step": 32145 + }, + { + "epoch": 1.37, + "grad_norm": 0.419921875, + "learning_rate": 4.529572659086101e-05, + "loss": 2.3748, + "step": 32146 + }, + { + "epoch": 1.37, + "grad_norm": 0.44921875, + "learning_rate": 4.529011147305421e-05, + "loss": 2.5672, + "step": 32147 + }, + { + "epoch": 1.37, + "grad_norm": 0.515625, + "learning_rate": 4.528449660142199e-05, + "loss": 2.4426, + "step": 32148 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.527888197598954e-05, + "loss": 2.5573, + "step": 32149 + }, + { + "epoch": 1.37, + "grad_norm": 0.498046875, + "learning_rate": 4.527326759678208e-05, + "loss": 2.4668, + "step": 32150 + }, + { + "epoch": 1.37, + "grad_norm": 0.447265625, + "learning_rate": 4.526765346382492e-05, + "loss": 2.5946, + "step": 32151 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.5262039577143356e-05, + "loss": 2.4152, + "step": 32152 + }, + { + "epoch": 1.37, + "grad_norm": 0.478515625, + "learning_rate": 4.525642593676256e-05, + "loss": 2.4654, + "step": 32153 + }, + { + "epoch": 1.37, + "grad_norm": 0.48828125, + "learning_rate": 4.525081254270789e-05, + "loss": 2.4876, + "step": 32154 + }, + { + "epoch": 1.37, + "grad_norm": 0.4609375, + "learning_rate": 4.5245199395004514e-05, + "loss": 2.3753, + "step": 32155 + }, + { + "epoch": 1.37, + "grad_norm": 0.408203125, + "learning_rate": 4.5239586493677766e-05, + "loss": 2.524, + "step": 32156 + }, + { + "epoch": 1.37, + "grad_norm": 0.50390625, + "learning_rate": 4.523397383875283e-05, + "loss": 2.439, + "step": 32157 + }, + { + "epoch": 1.37, + "grad_norm": 0.46484375, + "learning_rate": 4.5228361430255036e-05, + "loss": 2.2999, + "step": 32158 + }, + { + "epoch": 1.37, + "grad_norm": 0.462890625, + "learning_rate": 4.5222749268209565e-05, + "loss": 2.4039, + "step": 32159 + }, + { + "epoch": 1.37, + "grad_norm": 0.47265625, + "learning_rate": 4.52171373526417e-05, + "loss": 2.4458, + "step": 32160 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.5211525683576724e-05, + "loss": 2.4777, + "step": 32161 + }, + { + "epoch": 1.37, + "grad_norm": 0.4609375, + "learning_rate": 4.520591426103987e-05, + "loss": 2.2991, + "step": 32162 + }, + { + "epoch": 1.37, + "grad_norm": 0.486328125, + "learning_rate": 4.520030308505631e-05, + "loss": 2.3195, + "step": 32163 + }, + { + "epoch": 1.37, + "grad_norm": 0.5, + "learning_rate": 4.519469215565142e-05, + "loss": 2.489, + "step": 32164 + }, + { + "epoch": 1.37, + "grad_norm": 0.470703125, + "learning_rate": 4.5189081472850395e-05, + "loss": 2.4452, + "step": 32165 + }, + { + "epoch": 1.37, + "grad_norm": 0.52734375, + "learning_rate": 4.518347103667842e-05, + "loss": 2.5388, + "step": 32166 + }, + { + "epoch": 1.37, + "grad_norm": 0.4921875, + "learning_rate": 4.517786084716081e-05, + "loss": 2.5511, + "step": 32167 + }, + { + "epoch": 1.37, + "grad_norm": 0.482421875, + "learning_rate": 4.5172250904322825e-05, + "loss": 2.4806, + "step": 32168 + }, + { + "epoch": 1.37, + "grad_norm": 0.47265625, + "learning_rate": 4.5166641208189634e-05, + "loss": 2.5627, + "step": 32169 + }, + { + "epoch": 1.37, + "grad_norm": 0.46484375, + "learning_rate": 4.516103175878654e-05, + "loss": 2.5232, + "step": 32170 + }, + { + "epoch": 1.37, + "grad_norm": 0.54296875, + "learning_rate": 4.515542255613875e-05, + "loss": 2.5189, + "step": 32171 + }, + { + "epoch": 1.37, + "grad_norm": 0.51953125, + "learning_rate": 4.514981360027154e-05, + "loss": 2.5272, + "step": 32172 + }, + { + "epoch": 1.37, + "grad_norm": 0.4609375, + "learning_rate": 4.5144204891210084e-05, + "loss": 2.3965, + "step": 32173 + }, + { + "epoch": 1.37, + "grad_norm": 0.478515625, + "learning_rate": 4.51385964289797e-05, + "loss": 2.4797, + "step": 32174 + }, + { + "epoch": 1.37, + "grad_norm": 0.482421875, + "learning_rate": 4.5132988213605545e-05, + "loss": 2.4356, + "step": 32175 + }, + { + "epoch": 1.37, + "grad_norm": 0.515625, + "learning_rate": 4.512738024511288e-05, + "loss": 2.4764, + "step": 32176 + }, + { + "epoch": 1.37, + "grad_norm": 0.5, + "learning_rate": 4.5121772523527005e-05, + "loss": 2.5988, + "step": 32177 + }, + { + "epoch": 1.37, + "grad_norm": 0.482421875, + "learning_rate": 4.5116165048873094e-05, + "loss": 2.5708, + "step": 32178 + }, + { + "epoch": 1.37, + "grad_norm": 0.48046875, + "learning_rate": 4.51105578211763e-05, + "loss": 2.413, + "step": 32179 + }, + { + "epoch": 1.37, + "grad_norm": 0.52734375, + "learning_rate": 4.510495084046201e-05, + "loss": 2.2942, + "step": 32180 + }, + { + "epoch": 1.37, + "grad_norm": 0.47265625, + "learning_rate": 4.5099344106755395e-05, + "loss": 2.5285, + "step": 32181 + }, + { + "epoch": 1.37, + "grad_norm": 0.443359375, + "learning_rate": 4.509373762008162e-05, + "loss": 2.4109, + "step": 32182 + }, + { + "epoch": 1.37, + "grad_norm": 0.48828125, + "learning_rate": 4.508813138046596e-05, + "loss": 2.611, + "step": 32183 + }, + { + "epoch": 1.37, + "grad_norm": 0.4609375, + "learning_rate": 4.508252538793367e-05, + "loss": 2.4905, + "step": 32184 + }, + { + "epoch": 1.37, + "grad_norm": 0.50390625, + "learning_rate": 4.507691964250992e-05, + "loss": 2.4815, + "step": 32185 + }, + { + "epoch": 1.37, + "grad_norm": 0.44140625, + "learning_rate": 4.507131414421999e-05, + "loss": 2.2595, + "step": 32186 + }, + { + "epoch": 1.37, + "grad_norm": 0.59765625, + "learning_rate": 4.506570889308904e-05, + "loss": 2.511, + "step": 32187 + }, + { + "epoch": 1.37, + "grad_norm": 0.45703125, + "learning_rate": 4.506010388914236e-05, + "loss": 2.4331, + "step": 32188 + }, + { + "epoch": 1.37, + "grad_norm": 0.427734375, + "learning_rate": 4.505449913240509e-05, + "loss": 2.4187, + "step": 32189 + }, + { + "epoch": 1.37, + "grad_norm": 0.45703125, + "learning_rate": 4.5048894622902546e-05, + "loss": 2.6022, + "step": 32190 + }, + { + "epoch": 1.37, + "grad_norm": 0.45703125, + "learning_rate": 4.504329036065984e-05, + "loss": 2.4065, + "step": 32191 + }, + { + "epoch": 1.37, + "grad_norm": 0.49609375, + "learning_rate": 4.503768634570226e-05, + "loss": 2.418, + "step": 32192 + }, + { + "epoch": 1.37, + "grad_norm": 0.498046875, + "learning_rate": 4.503208257805502e-05, + "loss": 2.5765, + "step": 32193 + }, + { + "epoch": 1.37, + "grad_norm": 0.498046875, + "learning_rate": 4.502647905774334e-05, + "loss": 2.5448, + "step": 32194 + }, + { + "epoch": 1.37, + "grad_norm": 0.4765625, + "learning_rate": 4.5020875784792336e-05, + "loss": 2.4154, + "step": 32195 + }, + { + "epoch": 1.37, + "grad_norm": 0.4375, + "learning_rate": 4.501527275922737e-05, + "loss": 2.5175, + "step": 32196 + }, + { + "epoch": 1.37, + "grad_norm": 0.5078125, + "learning_rate": 4.500966998107358e-05, + "loss": 2.514, + "step": 32197 + }, + { + "epoch": 1.37, + "grad_norm": 0.5078125, + "learning_rate": 4.500406745035614e-05, + "loss": 2.4338, + "step": 32198 + }, + { + "epoch": 1.37, + "grad_norm": 0.4296875, + "learning_rate": 4.49984651671003e-05, + "loss": 2.416, + "step": 32199 + }, + { + "epoch": 1.37, + "grad_norm": 0.4765625, + "learning_rate": 4.49928631313313e-05, + "loss": 2.5034, + "step": 32200 + }, + { + "epoch": 1.37, + "grad_norm": 0.4921875, + "learning_rate": 4.4987261343074285e-05, + "loss": 2.2966, + "step": 32201 + }, + { + "epoch": 1.37, + "grad_norm": 0.47265625, + "learning_rate": 4.498165980235451e-05, + "loss": 2.3441, + "step": 32202 + }, + { + "epoch": 1.37, + "grad_norm": 0.4921875, + "learning_rate": 4.497605850919714e-05, + "loss": 2.4687, + "step": 32203 + }, + { + "epoch": 1.37, + "grad_norm": 0.451171875, + "learning_rate": 4.497045746362742e-05, + "loss": 2.4476, + "step": 32204 + }, + { + "epoch": 1.37, + "grad_norm": 0.53125, + "learning_rate": 4.496485666567051e-05, + "loss": 2.6081, + "step": 32205 + }, + { + "epoch": 1.37, + "grad_norm": 0.482421875, + "learning_rate": 4.495925611535166e-05, + "loss": 2.4708, + "step": 32206 + }, + { + "epoch": 1.37, + "grad_norm": 0.50390625, + "learning_rate": 4.495365581269599e-05, + "loss": 2.5853, + "step": 32207 + }, + { + "epoch": 1.37, + "grad_norm": 0.5, + "learning_rate": 4.494805575772878e-05, + "loss": 2.3948, + "step": 32208 + }, + { + "epoch": 1.37, + "grad_norm": 0.455078125, + "learning_rate": 4.4942455950475215e-05, + "loss": 2.515, + "step": 32209 + }, + { + "epoch": 1.37, + "grad_norm": 0.455078125, + "learning_rate": 4.493685639096049e-05, + "loss": 2.5504, + "step": 32210 + }, + { + "epoch": 1.37, + "grad_norm": 0.46875, + "learning_rate": 4.493125707920971e-05, + "loss": 2.4743, + "step": 32211 + }, + { + "epoch": 1.37, + "grad_norm": 0.51171875, + "learning_rate": 4.492565801524822e-05, + "loss": 2.4893, + "step": 32212 + }, + { + "epoch": 1.37, + "grad_norm": 0.4765625, + "learning_rate": 4.492005919910114e-05, + "loss": 2.3974, + "step": 32213 + }, + { + "epoch": 1.37, + "grad_norm": 0.443359375, + "learning_rate": 4.491446063079363e-05, + "loss": 2.4348, + "step": 32214 + }, + { + "epoch": 1.37, + "grad_norm": 0.50390625, + "learning_rate": 4.4908862310350916e-05, + "loss": 2.4891, + "step": 32215 + }, + { + "epoch": 1.37, + "grad_norm": 0.462890625, + "learning_rate": 4.490326423779823e-05, + "loss": 2.4115, + "step": 32216 + }, + { + "epoch": 1.37, + "grad_norm": 0.486328125, + "learning_rate": 4.4897666413160665e-05, + "loss": 2.4424, + "step": 32217 + }, + { + "epoch": 1.37, + "grad_norm": 0.4609375, + "learning_rate": 4.489206883646352e-05, + "loss": 2.381, + "step": 32218 + }, + { + "epoch": 1.37, + "grad_norm": 0.458984375, + "learning_rate": 4.4886471507731876e-05, + "loss": 2.4185, + "step": 32219 + }, + { + "epoch": 1.37, + "grad_norm": 0.47265625, + "learning_rate": 4.488087442699102e-05, + "loss": 2.478, + "step": 32220 + }, + { + "epoch": 1.37, + "grad_norm": 0.482421875, + "learning_rate": 4.487527759426603e-05, + "loss": 2.4297, + "step": 32221 + }, + { + "epoch": 1.37, + "grad_norm": 0.47265625, + "learning_rate": 4.4869681009582195e-05, + "loss": 2.5308, + "step": 32222 + }, + { + "epoch": 1.37, + "grad_norm": 0.494140625, + "learning_rate": 4.48640846729646e-05, + "loss": 2.4073, + "step": 32223 + }, + { + "epoch": 1.37, + "grad_norm": 0.46484375, + "learning_rate": 4.485848858443849e-05, + "loss": 2.4671, + "step": 32224 + }, + { + "epoch": 1.37, + "grad_norm": 0.50390625, + "learning_rate": 4.485289274402905e-05, + "loss": 2.4202, + "step": 32225 + }, + { + "epoch": 1.37, + "grad_norm": 0.4140625, + "learning_rate": 4.484729715176145e-05, + "loss": 2.4659, + "step": 32226 + }, + { + "epoch": 1.37, + "grad_norm": 0.482421875, + "learning_rate": 4.484170180766077e-05, + "loss": 2.4771, + "step": 32227 + }, + { + "epoch": 1.37, + "grad_norm": 0.447265625, + "learning_rate": 4.4836106711752366e-05, + "loss": 2.4997, + "step": 32228 + }, + { + "epoch": 1.37, + "grad_norm": 0.52734375, + "learning_rate": 4.483051186406131e-05, + "loss": 2.5455, + "step": 32229 + }, + { + "epoch": 1.37, + "grad_norm": 0.474609375, + "learning_rate": 4.482491726461275e-05, + "loss": 2.5114, + "step": 32230 + }, + { + "epoch": 1.37, + "grad_norm": 0.44140625, + "learning_rate": 4.481932291343191e-05, + "loss": 2.3699, + "step": 32231 + }, + { + "epoch": 1.37, + "grad_norm": 0.494140625, + "learning_rate": 4.481372881054398e-05, + "loss": 2.4399, + "step": 32232 + }, + { + "epoch": 1.37, + "grad_norm": 0.4375, + "learning_rate": 4.480813495597406e-05, + "loss": 2.4344, + "step": 32233 + }, + { + "epoch": 1.37, + "grad_norm": 0.48828125, + "learning_rate": 4.480254134974741e-05, + "loss": 2.4762, + "step": 32234 + }, + { + "epoch": 1.37, + "grad_norm": 0.443359375, + "learning_rate": 4.4796947991889107e-05, + "loss": 2.3328, + "step": 32235 + }, + { + "epoch": 1.37, + "grad_norm": 0.48046875, + "learning_rate": 4.4791354882424396e-05, + "loss": 2.6452, + "step": 32236 + }, + { + "epoch": 1.37, + "grad_norm": 0.447265625, + "learning_rate": 4.4785762021378376e-05, + "loss": 2.2964, + "step": 32237 + }, + { + "epoch": 1.37, + "grad_norm": 0.44921875, + "learning_rate": 4.4780169408776295e-05, + "loss": 2.4709, + "step": 32238 + }, + { + "epoch": 1.37, + "grad_norm": 0.51171875, + "learning_rate": 4.477457704464323e-05, + "loss": 2.4791, + "step": 32239 + }, + { + "epoch": 1.37, + "grad_norm": 0.55078125, + "learning_rate": 4.4768984929004385e-05, + "loss": 2.4855, + "step": 32240 + }, + { + "epoch": 1.37, + "grad_norm": 0.49609375, + "learning_rate": 4.476339306188496e-05, + "loss": 2.4529, + "step": 32241 + }, + { + "epoch": 1.37, + "grad_norm": 0.5078125, + "learning_rate": 4.475780144331008e-05, + "loss": 2.5254, + "step": 32242 + }, + { + "epoch": 1.37, + "grad_norm": 0.46484375, + "learning_rate": 4.475221007330482e-05, + "loss": 2.4165, + "step": 32243 + }, + { + "epoch": 1.37, + "grad_norm": 0.498046875, + "learning_rate": 4.4746618951894514e-05, + "loss": 2.4312, + "step": 32244 + }, + { + "epoch": 1.37, + "grad_norm": 0.466796875, + "learning_rate": 4.474102807910423e-05, + "loss": 2.4304, + "step": 32245 + }, + { + "epoch": 1.37, + "grad_norm": 0.4296875, + "learning_rate": 4.4735437454959075e-05, + "loss": 2.3949, + "step": 32246 + }, + { + "epoch": 1.37, + "grad_norm": 0.44140625, + "learning_rate": 4.4729847079484266e-05, + "loss": 2.4499, + "step": 32247 + }, + { + "epoch": 1.37, + "grad_norm": 0.43359375, + "learning_rate": 4.472425695270498e-05, + "loss": 2.3311, + "step": 32248 + }, + { + "epoch": 1.37, + "grad_norm": 0.45703125, + "learning_rate": 4.4718667074646295e-05, + "loss": 2.6225, + "step": 32249 + }, + { + "epoch": 1.37, + "grad_norm": 0.46875, + "learning_rate": 4.471307744533345e-05, + "loss": 2.2884, + "step": 32250 + }, + { + "epoch": 1.37, + "grad_norm": 0.478515625, + "learning_rate": 4.47074880647915e-05, + "loss": 2.5012, + "step": 32251 + }, + { + "epoch": 1.37, + "grad_norm": 0.421875, + "learning_rate": 4.47018989330457e-05, + "loss": 2.5314, + "step": 32252 + }, + { + "epoch": 1.37, + "grad_norm": 0.443359375, + "learning_rate": 4.46963100501211e-05, + "loss": 2.2917, + "step": 32253 + }, + { + "epoch": 1.37, + "grad_norm": 0.486328125, + "learning_rate": 4.469072141604293e-05, + "loss": 2.4625, + "step": 32254 + }, + { + "epoch": 1.37, + "grad_norm": 0.4765625, + "learning_rate": 4.468513303083627e-05, + "loss": 2.5513, + "step": 32255 + }, + { + "epoch": 1.37, + "grad_norm": 0.48828125, + "learning_rate": 4.467954489452629e-05, + "loss": 2.3936, + "step": 32256 + }, + { + "epoch": 1.37, + "grad_norm": 0.482421875, + "learning_rate": 4.467395700713819e-05, + "loss": 2.4818, + "step": 32257 + }, + { + "epoch": 1.37, + "grad_norm": 0.4375, + "learning_rate": 4.466836936869705e-05, + "loss": 2.3398, + "step": 32258 + }, + { + "epoch": 1.37, + "grad_norm": 0.4453125, + "learning_rate": 4.4662781979227994e-05, + "loss": 2.4463, + "step": 32259 + }, + { + "epoch": 1.37, + "grad_norm": 0.51171875, + "learning_rate": 4.465719483875619e-05, + "loss": 2.5852, + "step": 32260 + }, + { + "epoch": 1.37, + "grad_norm": 0.486328125, + "learning_rate": 4.4651607947306825e-05, + "loss": 2.4471, + "step": 32261 + }, + { + "epoch": 1.37, + "grad_norm": 0.4453125, + "learning_rate": 4.464602130490496e-05, + "loss": 2.583, + "step": 32262 + }, + { + "epoch": 1.37, + "grad_norm": 0.462890625, + "learning_rate": 4.4640434911575814e-05, + "loss": 2.3817, + "step": 32263 + }, + { + "epoch": 1.37, + "grad_norm": 0.494140625, + "learning_rate": 4.463484876734444e-05, + "loss": 2.3877, + "step": 32264 + }, + { + "epoch": 1.37, + "grad_norm": 0.49609375, + "learning_rate": 4.4629262872236e-05, + "loss": 2.5613, + "step": 32265 + }, + { + "epoch": 1.37, + "grad_norm": 0.546875, + "learning_rate": 4.46236772262757e-05, + "loss": 2.5003, + "step": 32266 + }, + { + "epoch": 1.37, + "grad_norm": 0.546875, + "learning_rate": 4.461809182948859e-05, + "loss": 2.3293, + "step": 32267 + }, + { + "epoch": 1.38, + "grad_norm": 0.435546875, + "learning_rate": 4.4612506681899805e-05, + "loss": 2.3942, + "step": 32268 + }, + { + "epoch": 1.38, + "grad_norm": 0.451171875, + "learning_rate": 4.460692178353449e-05, + "loss": 2.5457, + "step": 32269 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.460133713441783e-05, + "loss": 2.485, + "step": 32270 + }, + { + "epoch": 1.38, + "grad_norm": 0.408203125, + "learning_rate": 4.4595752734574856e-05, + "loss": 2.4885, + "step": 32271 + }, + { + "epoch": 1.38, + "grad_norm": 0.51171875, + "learning_rate": 4.459016858403077e-05, + "loss": 2.4867, + "step": 32272 + }, + { + "epoch": 1.38, + "grad_norm": 0.43359375, + "learning_rate": 4.458458468281069e-05, + "loss": 2.4552, + "step": 32273 + }, + { + "epoch": 1.38, + "grad_norm": 0.46484375, + "learning_rate": 4.457900103093974e-05, + "loss": 2.4296, + "step": 32274 + }, + { + "epoch": 1.38, + "grad_norm": 0.46875, + "learning_rate": 4.457341762844298e-05, + "loss": 2.4093, + "step": 32275 + }, + { + "epoch": 1.38, + "grad_norm": 0.486328125, + "learning_rate": 4.4567834475345596e-05, + "loss": 2.4926, + "step": 32276 + }, + { + "epoch": 1.38, + "grad_norm": 0.46875, + "learning_rate": 4.456225157167273e-05, + "loss": 2.6244, + "step": 32277 + }, + { + "epoch": 1.38, + "grad_norm": 0.443359375, + "learning_rate": 4.455666891744944e-05, + "loss": 2.4632, + "step": 32278 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.4551086512700914e-05, + "loss": 2.5048, + "step": 32279 + }, + { + "epoch": 1.38, + "grad_norm": 0.4765625, + "learning_rate": 4.454550435745219e-05, + "loss": 2.3764, + "step": 32280 + }, + { + "epoch": 1.38, + "grad_norm": 0.458984375, + "learning_rate": 4.453992245172844e-05, + "loss": 2.3572, + "step": 32281 + }, + { + "epoch": 1.38, + "grad_norm": 0.486328125, + "learning_rate": 4.4534340795554805e-05, + "loss": 2.4884, + "step": 32282 + }, + { + "epoch": 1.38, + "grad_norm": 0.4375, + "learning_rate": 4.4528759388956365e-05, + "loss": 2.358, + "step": 32283 + }, + { + "epoch": 1.38, + "grad_norm": 0.443359375, + "learning_rate": 4.4523178231958205e-05, + "loss": 2.4706, + "step": 32284 + }, + { + "epoch": 1.38, + "grad_norm": 0.4765625, + "learning_rate": 4.451759732458546e-05, + "loss": 2.6335, + "step": 32285 + }, + { + "epoch": 1.38, + "grad_norm": 0.39453125, + "learning_rate": 4.451201666686329e-05, + "loss": 2.2793, + "step": 32286 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.450643625881673e-05, + "loss": 2.437, + "step": 32287 + }, + { + "epoch": 1.38, + "grad_norm": 0.546875, + "learning_rate": 4.450085610047092e-05, + "loss": 2.5422, + "step": 32288 + }, + { + "epoch": 1.38, + "grad_norm": 0.439453125, + "learning_rate": 4.449527619185102e-05, + "loss": 2.3354, + "step": 32289 + }, + { + "epoch": 1.38, + "grad_norm": 0.490234375, + "learning_rate": 4.4489696532982097e-05, + "loss": 2.5372, + "step": 32290 + }, + { + "epoch": 1.38, + "grad_norm": 0.455078125, + "learning_rate": 4.4484117123889204e-05, + "loss": 2.5096, + "step": 32291 + }, + { + "epoch": 1.38, + "grad_norm": 0.431640625, + "learning_rate": 4.44785379645975e-05, + "loss": 2.4664, + "step": 32292 + }, + { + "epoch": 1.38, + "grad_norm": 0.431640625, + "learning_rate": 4.447295905513212e-05, + "loss": 2.4437, + "step": 32293 + }, + { + "epoch": 1.38, + "grad_norm": 0.400390625, + "learning_rate": 4.44673803955181e-05, + "loss": 2.3641, + "step": 32294 + }, + { + "epoch": 1.38, + "grad_norm": 0.50390625, + "learning_rate": 4.446180198578061e-05, + "loss": 2.5035, + "step": 32295 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.4456223825944665e-05, + "loss": 2.4912, + "step": 32296 + }, + { + "epoch": 1.38, + "grad_norm": 0.458984375, + "learning_rate": 4.4450645916035416e-05, + "loss": 2.4118, + "step": 32297 + }, + { + "epoch": 1.38, + "grad_norm": 0.4765625, + "learning_rate": 4.444506825607801e-05, + "loss": 2.4881, + "step": 32298 + }, + { + "epoch": 1.38, + "grad_norm": 0.455078125, + "learning_rate": 4.4439490846097484e-05, + "loss": 2.3653, + "step": 32299 + }, + { + "epoch": 1.38, + "grad_norm": 0.47265625, + "learning_rate": 4.443391368611891e-05, + "loss": 2.477, + "step": 32300 + }, + { + "epoch": 1.38, + "grad_norm": 0.486328125, + "learning_rate": 4.4428336776167414e-05, + "loss": 2.4931, + "step": 32301 + }, + { + "epoch": 1.38, + "grad_norm": 0.400390625, + "learning_rate": 4.4422760116268135e-05, + "loss": 2.33, + "step": 32302 + }, + { + "epoch": 1.38, + "grad_norm": 0.51171875, + "learning_rate": 4.4417183706446084e-05, + "loss": 2.4911, + "step": 32303 + }, + { + "epoch": 1.38, + "grad_norm": 0.451171875, + "learning_rate": 4.4411607546726394e-05, + "loss": 2.4371, + "step": 32304 + }, + { + "epoch": 1.38, + "grad_norm": 0.5, + "learning_rate": 4.440603163713419e-05, + "loss": 2.5252, + "step": 32305 + }, + { + "epoch": 1.38, + "grad_norm": 0.40625, + "learning_rate": 4.4400455977694535e-05, + "loss": 2.6101, + "step": 32306 + }, + { + "epoch": 1.38, + "grad_norm": 0.439453125, + "learning_rate": 4.4394880568432463e-05, + "loss": 2.4112, + "step": 32307 + }, + { + "epoch": 1.38, + "grad_norm": 0.41015625, + "learning_rate": 4.43893054093731e-05, + "loss": 2.465, + "step": 32308 + }, + { + "epoch": 1.38, + "grad_norm": 0.44921875, + "learning_rate": 4.43837305005416e-05, + "loss": 2.3887, + "step": 32309 + }, + { + "epoch": 1.38, + "grad_norm": 0.439453125, + "learning_rate": 4.437815584196293e-05, + "loss": 2.4109, + "step": 32310 + }, + { + "epoch": 1.38, + "grad_norm": 0.484375, + "learning_rate": 4.437258143366227e-05, + "loss": 2.4873, + "step": 32311 + }, + { + "epoch": 1.38, + "grad_norm": 0.5546875, + "learning_rate": 4.436700727566464e-05, + "loss": 2.6495, + "step": 32312 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.436143336799513e-05, + "loss": 2.392, + "step": 32313 + }, + { + "epoch": 1.38, + "grad_norm": 0.462890625, + "learning_rate": 4.435585971067888e-05, + "loss": 2.5216, + "step": 32314 + }, + { + "epoch": 1.38, + "grad_norm": 0.4921875, + "learning_rate": 4.435028630374093e-05, + "loss": 2.5173, + "step": 32315 + }, + { + "epoch": 1.38, + "grad_norm": 0.3984375, + "learning_rate": 4.434471314720632e-05, + "loss": 2.3555, + "step": 32316 + }, + { + "epoch": 1.38, + "grad_norm": 0.50390625, + "learning_rate": 4.433914024110015e-05, + "loss": 2.343, + "step": 32317 + }, + { + "epoch": 1.38, + "grad_norm": 0.4921875, + "learning_rate": 4.4333567585447545e-05, + "loss": 2.4156, + "step": 32318 + }, + { + "epoch": 1.38, + "grad_norm": 0.41796875, + "learning_rate": 4.4327995180273507e-05, + "loss": 2.4487, + "step": 32319 + }, + { + "epoch": 1.38, + "grad_norm": 0.4609375, + "learning_rate": 4.4322423025603154e-05, + "loss": 2.4708, + "step": 32320 + }, + { + "epoch": 1.38, + "grad_norm": 0.53125, + "learning_rate": 4.431685112146158e-05, + "loss": 2.4348, + "step": 32321 + }, + { + "epoch": 1.38, + "grad_norm": 0.49609375, + "learning_rate": 4.431127946787382e-05, + "loss": 2.531, + "step": 32322 + }, + { + "epoch": 1.38, + "grad_norm": 0.52734375, + "learning_rate": 4.4305708064864916e-05, + "loss": 2.5667, + "step": 32323 + }, + { + "epoch": 1.38, + "grad_norm": 0.421875, + "learning_rate": 4.430013691245998e-05, + "loss": 2.465, + "step": 32324 + }, + { + "epoch": 1.38, + "grad_norm": 0.546875, + "learning_rate": 4.42945660106841e-05, + "loss": 2.5857, + "step": 32325 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.428899535956228e-05, + "loss": 2.5068, + "step": 32326 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.4283424959119654e-05, + "loss": 2.3029, + "step": 32327 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.427785480938123e-05, + "loss": 2.3862, + "step": 32328 + }, + { + "epoch": 1.38, + "grad_norm": 0.482421875, + "learning_rate": 4.427228491037209e-05, + "loss": 2.374, + "step": 32329 + }, + { + "epoch": 1.38, + "grad_norm": 0.443359375, + "learning_rate": 4.426671526211733e-05, + "loss": 2.5913, + "step": 32330 + }, + { + "epoch": 1.38, + "grad_norm": 0.4609375, + "learning_rate": 4.4261145864641995e-05, + "loss": 2.4708, + "step": 32331 + }, + { + "epoch": 1.38, + "grad_norm": 0.462890625, + "learning_rate": 4.4255576717971084e-05, + "loss": 2.5311, + "step": 32332 + }, + { + "epoch": 1.38, + "grad_norm": 0.451171875, + "learning_rate": 4.4250007822129715e-05, + "loss": 2.4982, + "step": 32333 + }, + { + "epoch": 1.38, + "grad_norm": 0.5, + "learning_rate": 4.4244439177142974e-05, + "loss": 2.4104, + "step": 32334 + }, + { + "epoch": 1.38, + "grad_norm": 0.498046875, + "learning_rate": 4.423887078303583e-05, + "loss": 2.5853, + "step": 32335 + }, + { + "epoch": 1.38, + "grad_norm": 0.498046875, + "learning_rate": 4.423330263983341e-05, + "loss": 2.4866, + "step": 32336 + }, + { + "epoch": 1.38, + "grad_norm": 0.4453125, + "learning_rate": 4.422773474756077e-05, + "loss": 2.3904, + "step": 32337 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.422216710624294e-05, + "loss": 2.479, + "step": 32338 + }, + { + "epoch": 1.38, + "grad_norm": 0.498046875, + "learning_rate": 4.421659971590494e-05, + "loss": 2.4539, + "step": 32339 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.4211032576571866e-05, + "loss": 2.5834, + "step": 32340 + }, + { + "epoch": 1.38, + "grad_norm": 0.46875, + "learning_rate": 4.420546568826878e-05, + "loss": 2.4964, + "step": 32341 + }, + { + "epoch": 1.38, + "grad_norm": 0.54296875, + "learning_rate": 4.419989905102068e-05, + "loss": 2.5019, + "step": 32342 + }, + { + "epoch": 1.38, + "grad_norm": 0.435546875, + "learning_rate": 4.4194332664852666e-05, + "loss": 2.4131, + "step": 32343 + }, + { + "epoch": 1.38, + "grad_norm": 0.44921875, + "learning_rate": 4.4188766529789726e-05, + "loss": 2.4783, + "step": 32344 + }, + { + "epoch": 1.38, + "grad_norm": 0.4765625, + "learning_rate": 4.418320064585695e-05, + "loss": 2.3524, + "step": 32345 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.41776350130794e-05, + "loss": 2.5687, + "step": 32346 + }, + { + "epoch": 1.38, + "grad_norm": 0.439453125, + "learning_rate": 4.41720696314821e-05, + "loss": 2.5004, + "step": 32347 + }, + { + "epoch": 1.38, + "grad_norm": 0.498046875, + "learning_rate": 4.4166504501090044e-05, + "loss": 2.6584, + "step": 32348 + }, + { + "epoch": 1.38, + "grad_norm": 0.466796875, + "learning_rate": 4.416093962192831e-05, + "loss": 2.4527, + "step": 32349 + }, + { + "epoch": 1.38, + "grad_norm": 0.41796875, + "learning_rate": 4.415537499402198e-05, + "loss": 2.4179, + "step": 32350 + }, + { + "epoch": 1.38, + "grad_norm": 0.423828125, + "learning_rate": 4.4149810617396016e-05, + "loss": 2.4382, + "step": 32351 + }, + { + "epoch": 1.38, + "grad_norm": 0.435546875, + "learning_rate": 4.414424649207549e-05, + "loss": 2.4021, + "step": 32352 + }, + { + "epoch": 1.38, + "grad_norm": 0.498046875, + "learning_rate": 4.413868261808548e-05, + "loss": 2.5433, + "step": 32353 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.4133118995451e-05, + "loss": 2.4074, + "step": 32354 + }, + { + "epoch": 1.38, + "grad_norm": 0.50390625, + "learning_rate": 4.4127555624197e-05, + "loss": 2.3096, + "step": 32355 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.412199250434861e-05, + "loss": 2.4212, + "step": 32356 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.4116429635930845e-05, + "loss": 2.4323, + "step": 32357 + }, + { + "epoch": 1.38, + "grad_norm": 0.498046875, + "learning_rate": 4.411086701896871e-05, + "loss": 2.5112, + "step": 32358 + }, + { + "epoch": 1.38, + "grad_norm": 0.484375, + "learning_rate": 4.4105304653487264e-05, + "loss": 2.4711, + "step": 32359 + }, + { + "epoch": 1.38, + "grad_norm": 0.5, + "learning_rate": 4.409974253951149e-05, + "loss": 2.5568, + "step": 32360 + }, + { + "epoch": 1.38, + "grad_norm": 0.486328125, + "learning_rate": 4.4094180677066456e-05, + "loss": 2.5335, + "step": 32361 + }, + { + "epoch": 1.38, + "grad_norm": 0.4609375, + "learning_rate": 4.408861906617721e-05, + "loss": 2.6286, + "step": 32362 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.4083057706868734e-05, + "loss": 2.4356, + "step": 32363 + }, + { + "epoch": 1.38, + "grad_norm": 0.470703125, + "learning_rate": 4.4077496599166037e-05, + "loss": 2.4285, + "step": 32364 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.407193574309416e-05, + "loss": 2.3744, + "step": 32365 + }, + { + "epoch": 1.38, + "grad_norm": 0.5234375, + "learning_rate": 4.406637513867817e-05, + "loss": 2.545, + "step": 32366 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.4060814785943015e-05, + "loss": 2.4987, + "step": 32367 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.405525468491376e-05, + "loss": 2.3392, + "step": 32368 + }, + { + "epoch": 1.38, + "grad_norm": 0.474609375, + "learning_rate": 4.404969483561544e-05, + "loss": 2.5294, + "step": 32369 + }, + { + "epoch": 1.38, + "grad_norm": 0.51171875, + "learning_rate": 4.4044135238073046e-05, + "loss": 2.5533, + "step": 32370 + }, + { + "epoch": 1.38, + "grad_norm": 0.458984375, + "learning_rate": 4.403857589231156e-05, + "loss": 2.3744, + "step": 32371 + }, + { + "epoch": 1.38, + "grad_norm": 0.484375, + "learning_rate": 4.403301679835603e-05, + "loss": 2.326, + "step": 32372 + }, + { + "epoch": 1.38, + "grad_norm": 0.56640625, + "learning_rate": 4.402745795623151e-05, + "loss": 2.4525, + "step": 32373 + }, + { + "epoch": 1.38, + "grad_norm": 0.490234375, + "learning_rate": 4.402189936596294e-05, + "loss": 2.4708, + "step": 32374 + }, + { + "epoch": 1.38, + "grad_norm": 0.427734375, + "learning_rate": 4.4016341027575394e-05, + "loss": 2.3542, + "step": 32375 + }, + { + "epoch": 1.38, + "grad_norm": 0.412109375, + "learning_rate": 4.4010782941093855e-05, + "loss": 2.3967, + "step": 32376 + }, + { + "epoch": 1.38, + "grad_norm": 0.431640625, + "learning_rate": 4.400522510654326e-05, + "loss": 2.4559, + "step": 32377 + }, + { + "epoch": 1.38, + "grad_norm": 0.4453125, + "learning_rate": 4.3999667523948765e-05, + "loss": 2.4718, + "step": 32378 + }, + { + "epoch": 1.38, + "grad_norm": 0.43359375, + "learning_rate": 4.399411019333529e-05, + "loss": 2.4924, + "step": 32379 + }, + { + "epoch": 1.38, + "grad_norm": 0.455078125, + "learning_rate": 4.398855311472783e-05, + "loss": 2.3563, + "step": 32380 + }, + { + "epoch": 1.38, + "grad_norm": 0.474609375, + "learning_rate": 4.398299628815139e-05, + "loss": 2.4652, + "step": 32381 + }, + { + "epoch": 1.38, + "grad_norm": 0.453125, + "learning_rate": 4.3977439713631055e-05, + "loss": 2.366, + "step": 32382 + }, + { + "epoch": 1.38, + "grad_norm": 0.439453125, + "learning_rate": 4.3971883391191716e-05, + "loss": 2.5446, + "step": 32383 + }, + { + "epoch": 1.38, + "grad_norm": 0.455078125, + "learning_rate": 4.396632732085846e-05, + "loss": 2.5113, + "step": 32384 + }, + { + "epoch": 1.38, + "grad_norm": 0.5625, + "learning_rate": 4.396077150265622e-05, + "loss": 2.4992, + "step": 32385 + }, + { + "epoch": 1.38, + "grad_norm": 0.41015625, + "learning_rate": 4.395521593661007e-05, + "loss": 2.4001, + "step": 32386 + }, + { + "epoch": 1.38, + "grad_norm": 0.5078125, + "learning_rate": 4.3949660622744926e-05, + "loss": 2.5047, + "step": 32387 + }, + { + "epoch": 1.38, + "grad_norm": 0.44921875, + "learning_rate": 4.394410556108586e-05, + "loss": 2.5887, + "step": 32388 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.3938550751657806e-05, + "loss": 2.4159, + "step": 32389 + }, + { + "epoch": 1.38, + "grad_norm": 0.44140625, + "learning_rate": 4.3932996194485775e-05, + "loss": 2.3657, + "step": 32390 + }, + { + "epoch": 1.38, + "grad_norm": 0.439453125, + "learning_rate": 4.392744188959481e-05, + "loss": 2.4889, + "step": 32391 + }, + { + "epoch": 1.38, + "grad_norm": 0.5625, + "learning_rate": 4.3921887837009866e-05, + "loss": 2.5148, + "step": 32392 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.3916334036755854e-05, + "loss": 2.5802, + "step": 32393 + }, + { + "epoch": 1.38, + "grad_norm": 0.5078125, + "learning_rate": 4.3910780488857915e-05, + "loss": 2.5411, + "step": 32394 + }, + { + "epoch": 1.38, + "grad_norm": 0.470703125, + "learning_rate": 4.3905227193340984e-05, + "loss": 2.4236, + "step": 32395 + }, + { + "epoch": 1.38, + "grad_norm": 0.404296875, + "learning_rate": 4.389967415022997e-05, + "loss": 2.4545, + "step": 32396 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.3894121359549936e-05, + "loss": 2.4294, + "step": 32397 + }, + { + "epoch": 1.38, + "grad_norm": 0.421875, + "learning_rate": 4.388856882132588e-05, + "loss": 2.3932, + "step": 32398 + }, + { + "epoch": 1.38, + "grad_norm": 0.486328125, + "learning_rate": 4.3883016535582724e-05, + "loss": 2.5552, + "step": 32399 + }, + { + "epoch": 1.38, + "grad_norm": 0.4140625, + "learning_rate": 4.387746450234552e-05, + "loss": 2.5272, + "step": 32400 + }, + { + "epoch": 1.38, + "grad_norm": 0.427734375, + "learning_rate": 4.387191272163919e-05, + "loss": 2.3527, + "step": 32401 + }, + { + "epoch": 1.38, + "grad_norm": 0.48828125, + "learning_rate": 4.386636119348876e-05, + "loss": 2.3911, + "step": 32402 + }, + { + "epoch": 1.38, + "grad_norm": 0.44921875, + "learning_rate": 4.3860809917919166e-05, + "loss": 2.4568, + "step": 32403 + }, + { + "epoch": 1.38, + "grad_norm": 0.515625, + "learning_rate": 4.385525889495544e-05, + "loss": 2.5432, + "step": 32404 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.384970812462249e-05, + "loss": 2.4388, + "step": 32405 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.3844157606945346e-05, + "loss": 2.5083, + "step": 32406 + }, + { + "epoch": 1.38, + "grad_norm": 0.427734375, + "learning_rate": 4.3838607341949e-05, + "loss": 2.3634, + "step": 32407 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.383305732965839e-05, + "loss": 2.6223, + "step": 32408 + }, + { + "epoch": 1.38, + "grad_norm": 0.455078125, + "learning_rate": 4.3827507570098425e-05, + "loss": 2.6086, + "step": 32409 + }, + { + "epoch": 1.38, + "grad_norm": 0.498046875, + "learning_rate": 4.382195806329423e-05, + "loss": 2.5076, + "step": 32410 + }, + { + "epoch": 1.38, + "grad_norm": 0.55078125, + "learning_rate": 4.381640880927068e-05, + "loss": 2.5343, + "step": 32411 + }, + { + "epoch": 1.38, + "grad_norm": 0.41796875, + "learning_rate": 4.381085980805273e-05, + "loss": 2.3828, + "step": 32412 + }, + { + "epoch": 1.38, + "grad_norm": 0.458984375, + "learning_rate": 4.3805311059665376e-05, + "loss": 2.418, + "step": 32413 + }, + { + "epoch": 1.38, + "grad_norm": 0.46875, + "learning_rate": 4.3799762564133616e-05, + "loss": 2.5216, + "step": 32414 + }, + { + "epoch": 1.38, + "grad_norm": 0.408203125, + "learning_rate": 4.379421432148235e-05, + "loss": 2.3621, + "step": 32415 + }, + { + "epoch": 1.38, + "grad_norm": 0.4296875, + "learning_rate": 4.378866633173663e-05, + "loss": 2.5535, + "step": 32416 + }, + { + "epoch": 1.38, + "grad_norm": 0.47265625, + "learning_rate": 4.3783118594921316e-05, + "loss": 2.4993, + "step": 32417 + }, + { + "epoch": 1.38, + "grad_norm": 0.482421875, + "learning_rate": 4.377757111106147e-05, + "loss": 2.4161, + "step": 32418 + }, + { + "epoch": 1.38, + "grad_norm": 0.5390625, + "learning_rate": 4.377202388018196e-05, + "loss": 2.439, + "step": 32419 + }, + { + "epoch": 1.38, + "grad_norm": 0.484375, + "learning_rate": 4.376647690230783e-05, + "loss": 2.3556, + "step": 32420 + }, + { + "epoch": 1.38, + "grad_norm": 0.439453125, + "learning_rate": 4.376093017746398e-05, + "loss": 2.4466, + "step": 32421 + }, + { + "epoch": 1.38, + "grad_norm": 0.431640625, + "learning_rate": 4.375538370567537e-05, + "loss": 2.4315, + "step": 32422 + }, + { + "epoch": 1.38, + "grad_norm": 0.42578125, + "learning_rate": 4.374983748696703e-05, + "loss": 2.4519, + "step": 32423 + }, + { + "epoch": 1.38, + "grad_norm": 0.47265625, + "learning_rate": 4.3744291521363866e-05, + "loss": 2.4807, + "step": 32424 + }, + { + "epoch": 1.38, + "grad_norm": 0.439453125, + "learning_rate": 4.373874580889075e-05, + "loss": 2.4412, + "step": 32425 + }, + { + "epoch": 1.38, + "grad_norm": 0.392578125, + "learning_rate": 4.373320034957279e-05, + "loss": 2.3882, + "step": 32426 + }, + { + "epoch": 1.38, + "grad_norm": 0.44921875, + "learning_rate": 4.372765514343487e-05, + "loss": 2.5267, + "step": 32427 + }, + { + "epoch": 1.38, + "grad_norm": 0.44140625, + "learning_rate": 4.37221101905019e-05, + "loss": 2.4606, + "step": 32428 + }, + { + "epoch": 1.38, + "grad_norm": 0.4921875, + "learning_rate": 4.371656549079887e-05, + "loss": 2.3753, + "step": 32429 + }, + { + "epoch": 1.38, + "grad_norm": 0.46484375, + "learning_rate": 4.371102104435075e-05, + "loss": 2.4805, + "step": 32430 + }, + { + "epoch": 1.38, + "grad_norm": 0.4375, + "learning_rate": 4.370547685118244e-05, + "loss": 2.4594, + "step": 32431 + }, + { + "epoch": 1.38, + "grad_norm": 0.3828125, + "learning_rate": 4.369993291131894e-05, + "loss": 2.571, + "step": 32432 + }, + { + "epoch": 1.38, + "grad_norm": 0.4765625, + "learning_rate": 4.3694389224785125e-05, + "loss": 2.4124, + "step": 32433 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.3688845791606036e-05, + "loss": 2.3632, + "step": 32434 + }, + { + "epoch": 1.38, + "grad_norm": 0.470703125, + "learning_rate": 4.368330261180651e-05, + "loss": 2.5265, + "step": 32435 + }, + { + "epoch": 1.38, + "grad_norm": 0.5546875, + "learning_rate": 4.367775968541158e-05, + "loss": 2.5675, + "step": 32436 + }, + { + "epoch": 1.38, + "grad_norm": 0.46484375, + "learning_rate": 4.367221701244612e-05, + "loss": 2.4105, + "step": 32437 + }, + { + "epoch": 1.38, + "grad_norm": 0.494140625, + "learning_rate": 4.3666674592935086e-05, + "loss": 2.4882, + "step": 32438 + }, + { + "epoch": 1.38, + "grad_norm": 0.44921875, + "learning_rate": 4.3661132426903476e-05, + "loss": 2.5783, + "step": 32439 + }, + { + "epoch": 1.38, + "grad_norm": 0.4140625, + "learning_rate": 4.365559051437617e-05, + "loss": 2.3312, + "step": 32440 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.3650048855378046e-05, + "loss": 2.5082, + "step": 32441 + }, + { + "epoch": 1.38, + "grad_norm": 0.50390625, + "learning_rate": 4.364450744993418e-05, + "loss": 2.5491, + "step": 32442 + }, + { + "epoch": 1.38, + "grad_norm": 0.455078125, + "learning_rate": 4.363896629806944e-05, + "loss": 2.3572, + "step": 32443 + }, + { + "epoch": 1.38, + "grad_norm": 0.48828125, + "learning_rate": 4.3633425399808716e-05, + "loss": 2.4149, + "step": 32444 + }, + { + "epoch": 1.38, + "grad_norm": 0.43359375, + "learning_rate": 4.362788475517699e-05, + "loss": 2.3712, + "step": 32445 + }, + { + "epoch": 1.38, + "grad_norm": 0.474609375, + "learning_rate": 4.3622344364199195e-05, + "loss": 2.3428, + "step": 32446 + }, + { + "epoch": 1.38, + "grad_norm": 0.53125, + "learning_rate": 4.361680422690022e-05, + "loss": 2.5038, + "step": 32447 + }, + { + "epoch": 1.38, + "grad_norm": 0.458984375, + "learning_rate": 4.361126434330506e-05, + "loss": 2.4416, + "step": 32448 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.360572471343856e-05, + "loss": 2.5026, + "step": 32449 + }, + { + "epoch": 1.38, + "grad_norm": 0.5078125, + "learning_rate": 4.360018533732573e-05, + "loss": 2.4008, + "step": 32450 + }, + { + "epoch": 1.38, + "grad_norm": 0.4453125, + "learning_rate": 4.359464621499142e-05, + "loss": 2.4056, + "step": 32451 + }, + { + "epoch": 1.38, + "grad_norm": 0.455078125, + "learning_rate": 4.358910734646061e-05, + "loss": 2.4233, + "step": 32452 + }, + { + "epoch": 1.38, + "grad_norm": 0.458984375, + "learning_rate": 4.358356873175819e-05, + "loss": 2.4522, + "step": 32453 + }, + { + "epoch": 1.38, + "grad_norm": 0.46875, + "learning_rate": 4.357803037090907e-05, + "loss": 2.4476, + "step": 32454 + }, + { + "epoch": 1.38, + "grad_norm": 0.4296875, + "learning_rate": 4.357249226393824e-05, + "loss": 2.3977, + "step": 32455 + }, + { + "epoch": 1.38, + "grad_norm": 0.419921875, + "learning_rate": 4.3566954410870576e-05, + "loss": 2.4569, + "step": 32456 + }, + { + "epoch": 1.38, + "grad_norm": 0.46484375, + "learning_rate": 4.3561416811730914e-05, + "loss": 2.525, + "step": 32457 + }, + { + "epoch": 1.38, + "grad_norm": 0.451171875, + "learning_rate": 4.355587946654433e-05, + "loss": 2.4145, + "step": 32458 + }, + { + "epoch": 1.38, + "grad_norm": 0.470703125, + "learning_rate": 4.355034237533565e-05, + "loss": 2.4627, + "step": 32459 + }, + { + "epoch": 1.38, + "grad_norm": 0.47265625, + "learning_rate": 4.354480553812976e-05, + "loss": 2.4582, + "step": 32460 + }, + { + "epoch": 1.38, + "grad_norm": 0.443359375, + "learning_rate": 4.353926895495163e-05, + "loss": 2.4699, + "step": 32461 + }, + { + "epoch": 1.38, + "grad_norm": 0.482421875, + "learning_rate": 4.353373262582619e-05, + "loss": 2.4775, + "step": 32462 + }, + { + "epoch": 1.38, + "grad_norm": 0.431640625, + "learning_rate": 4.3528196550778265e-05, + "loss": 2.3846, + "step": 32463 + }, + { + "epoch": 1.38, + "grad_norm": 0.4453125, + "learning_rate": 4.352266072983287e-05, + "loss": 2.4728, + "step": 32464 + }, + { + "epoch": 1.38, + "grad_norm": 0.453125, + "learning_rate": 4.351712516301482e-05, + "loss": 2.4793, + "step": 32465 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.351158985034909e-05, + "loss": 2.4103, + "step": 32466 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.350605479186054e-05, + "loss": 2.4662, + "step": 32467 + }, + { + "epoch": 1.38, + "grad_norm": 0.54296875, + "learning_rate": 4.350051998757413e-05, + "loss": 2.3198, + "step": 32468 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.349498543751469e-05, + "loss": 2.5202, + "step": 32469 + }, + { + "epoch": 1.38, + "grad_norm": 0.423828125, + "learning_rate": 4.348945114170717e-05, + "loss": 2.4788, + "step": 32470 + }, + { + "epoch": 1.38, + "grad_norm": 0.44921875, + "learning_rate": 4.34839171001765e-05, + "loss": 2.4764, + "step": 32471 + }, + { + "epoch": 1.38, + "grad_norm": 0.455078125, + "learning_rate": 4.3478383312947544e-05, + "loss": 2.3381, + "step": 32472 + }, + { + "epoch": 1.38, + "grad_norm": 0.52734375, + "learning_rate": 4.3472849780045146e-05, + "loss": 2.4841, + "step": 32473 + }, + { + "epoch": 1.38, + "grad_norm": 0.482421875, + "learning_rate": 4.346731650149434e-05, + "loss": 2.5149, + "step": 32474 + }, + { + "epoch": 1.38, + "grad_norm": 0.419921875, + "learning_rate": 4.346178347731994e-05, + "loss": 2.3213, + "step": 32475 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.345625070754682e-05, + "loss": 2.4803, + "step": 32476 + }, + { + "epoch": 1.38, + "grad_norm": 0.4921875, + "learning_rate": 4.3450718192199904e-05, + "loss": 2.4409, + "step": 32477 + }, + { + "epoch": 1.38, + "grad_norm": 0.46484375, + "learning_rate": 4.3445185931304134e-05, + "loss": 2.3776, + "step": 32478 + }, + { + "epoch": 1.38, + "grad_norm": 0.408203125, + "learning_rate": 4.3439653924884316e-05, + "loss": 2.4395, + "step": 32479 + }, + { + "epoch": 1.38, + "grad_norm": 0.4609375, + "learning_rate": 4.343412217296542e-05, + "loss": 2.4196, + "step": 32480 + }, + { + "epoch": 1.38, + "grad_norm": 0.462890625, + "learning_rate": 4.342859067557227e-05, + "loss": 2.498, + "step": 32481 + }, + { + "epoch": 1.38, + "grad_norm": 0.4375, + "learning_rate": 4.3423059432729816e-05, + "loss": 2.4131, + "step": 32482 + }, + { + "epoch": 1.38, + "grad_norm": 0.478515625, + "learning_rate": 4.34175284444629e-05, + "loss": 2.5199, + "step": 32483 + }, + { + "epoch": 1.38, + "grad_norm": 0.453125, + "learning_rate": 4.341199771079645e-05, + "loss": 2.4556, + "step": 32484 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.340646723175529e-05, + "loss": 2.3517, + "step": 32485 + }, + { + "epoch": 1.38, + "grad_norm": 0.427734375, + "learning_rate": 4.340093700736435e-05, + "loss": 2.5256, + "step": 32486 + }, + { + "epoch": 1.38, + "grad_norm": 0.431640625, + "learning_rate": 4.339540703764854e-05, + "loss": 2.5097, + "step": 32487 + }, + { + "epoch": 1.38, + "grad_norm": 0.408203125, + "learning_rate": 4.338987732263272e-05, + "loss": 2.4364, + "step": 32488 + }, + { + "epoch": 1.38, + "grad_norm": 0.4765625, + "learning_rate": 4.338434786234169e-05, + "loss": 2.5245, + "step": 32489 + }, + { + "epoch": 1.38, + "grad_norm": 0.51953125, + "learning_rate": 4.337881865680047e-05, + "loss": 2.377, + "step": 32490 + }, + { + "epoch": 1.38, + "grad_norm": 0.443359375, + "learning_rate": 4.337328970603388e-05, + "loss": 2.2817, + "step": 32491 + }, + { + "epoch": 1.38, + "grad_norm": 0.447265625, + "learning_rate": 4.336776101006675e-05, + "loss": 2.4438, + "step": 32492 + }, + { + "epoch": 1.38, + "grad_norm": 0.51953125, + "learning_rate": 4.3362232568924e-05, + "loss": 2.4761, + "step": 32493 + }, + { + "epoch": 1.38, + "grad_norm": 0.466796875, + "learning_rate": 4.335670438263054e-05, + "loss": 2.4148, + "step": 32494 + }, + { + "epoch": 1.38, + "grad_norm": 0.45703125, + "learning_rate": 4.3351176451211185e-05, + "loss": 2.4669, + "step": 32495 + }, + { + "epoch": 1.38, + "grad_norm": 0.474609375, + "learning_rate": 4.334564877469085e-05, + "loss": 2.5302, + "step": 32496 + }, + { + "epoch": 1.38, + "grad_norm": 0.48046875, + "learning_rate": 4.334012135309441e-05, + "loss": 2.5698, + "step": 32497 + }, + { + "epoch": 1.38, + "grad_norm": 0.474609375, + "learning_rate": 4.333459418644666e-05, + "loss": 2.4315, + "step": 32498 + }, + { + "epoch": 1.38, + "grad_norm": 0.55078125, + "learning_rate": 4.332906727477255e-05, + "loss": 2.4456, + "step": 32499 + }, + { + "epoch": 1.38, + "grad_norm": 0.466796875, + "learning_rate": 4.3323540618096946e-05, + "loss": 2.435, + "step": 32500 + }, + { + "epoch": 1.38, + "grad_norm": 0.48828125, + "learning_rate": 4.3318014216444666e-05, + "loss": 2.5072, + "step": 32501 + }, + { + "epoch": 1.39, + "grad_norm": 0.44140625, + "learning_rate": 4.331248806984061e-05, + "loss": 2.4645, + "step": 32502 + }, + { + "epoch": 1.39, + "grad_norm": 0.447265625, + "learning_rate": 4.330696217830967e-05, + "loss": 2.4857, + "step": 32503 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.330143654187668e-05, + "loss": 2.3695, + "step": 32504 + }, + { + "epoch": 1.39, + "grad_norm": 0.435546875, + "learning_rate": 4.329591116056648e-05, + "loss": 2.5053, + "step": 32505 + }, + { + "epoch": 1.39, + "grad_norm": 0.453125, + "learning_rate": 4.3290386034403954e-05, + "loss": 2.5286, + "step": 32506 + }, + { + "epoch": 1.39, + "grad_norm": 0.515625, + "learning_rate": 4.328486116341399e-05, + "loss": 2.3297, + "step": 32507 + }, + { + "epoch": 1.39, + "grad_norm": 0.53125, + "learning_rate": 4.32793365476214e-05, + "loss": 2.6168, + "step": 32508 + }, + { + "epoch": 1.39, + "grad_norm": 0.5, + "learning_rate": 4.32738121870511e-05, + "loss": 2.4693, + "step": 32509 + }, + { + "epoch": 1.39, + "grad_norm": 0.462890625, + "learning_rate": 4.326828808172788e-05, + "loss": 2.4074, + "step": 32510 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.326276423167662e-05, + "loss": 2.479, + "step": 32511 + }, + { + "epoch": 1.39, + "grad_norm": 0.484375, + "learning_rate": 4.3257240636922236e-05, + "loss": 2.5275, + "step": 32512 + }, + { + "epoch": 1.39, + "grad_norm": 0.458984375, + "learning_rate": 4.325171729748952e-05, + "loss": 2.471, + "step": 32513 + }, + { + "epoch": 1.39, + "grad_norm": 0.44921875, + "learning_rate": 4.32461942134033e-05, + "loss": 2.5721, + "step": 32514 + }, + { + "epoch": 1.39, + "grad_norm": 0.5, + "learning_rate": 4.324067138468848e-05, + "loss": 2.4337, + "step": 32515 + }, + { + "epoch": 1.39, + "grad_norm": 0.51171875, + "learning_rate": 4.323514881136992e-05, + "loss": 2.5147, + "step": 32516 + }, + { + "epoch": 1.39, + "grad_norm": 0.5390625, + "learning_rate": 4.322962649347241e-05, + "loss": 2.4979, + "step": 32517 + }, + { + "epoch": 1.39, + "grad_norm": 0.5078125, + "learning_rate": 4.322410443102084e-05, + "loss": 2.2297, + "step": 32518 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.321858262404008e-05, + "loss": 2.5743, + "step": 32519 + }, + { + "epoch": 1.39, + "grad_norm": 0.482421875, + "learning_rate": 4.321306107255495e-05, + "loss": 2.4517, + "step": 32520 + }, + { + "epoch": 1.39, + "grad_norm": 0.4921875, + "learning_rate": 4.3207539776590254e-05, + "loss": 2.5151, + "step": 32521 + }, + { + "epoch": 1.39, + "grad_norm": 0.51953125, + "learning_rate": 4.3202018736170876e-05, + "loss": 2.5908, + "step": 32522 + }, + { + "epoch": 1.39, + "grad_norm": 0.45703125, + "learning_rate": 4.319649795132169e-05, + "loss": 2.4725, + "step": 32523 + }, + { + "epoch": 1.39, + "grad_norm": 0.4453125, + "learning_rate": 4.319097742206747e-05, + "loss": 2.409, + "step": 32524 + }, + { + "epoch": 1.39, + "grad_norm": 0.458984375, + "learning_rate": 4.318545714843313e-05, + "loss": 2.4029, + "step": 32525 + }, + { + "epoch": 1.39, + "grad_norm": 0.4375, + "learning_rate": 4.3179937130443424e-05, + "loss": 2.4345, + "step": 32526 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.317441736812324e-05, + "loss": 2.5788, + "step": 32527 + }, + { + "epoch": 1.39, + "grad_norm": 0.51953125, + "learning_rate": 4.316889786149745e-05, + "loss": 2.5234, + "step": 32528 + }, + { + "epoch": 1.39, + "grad_norm": 0.51171875, + "learning_rate": 4.316337861059084e-05, + "loss": 2.5488, + "step": 32529 + }, + { + "epoch": 1.39, + "grad_norm": 0.453125, + "learning_rate": 4.315785961542823e-05, + "loss": 2.456, + "step": 32530 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.315234087603447e-05, + "loss": 2.4125, + "step": 32531 + }, + { + "epoch": 1.39, + "grad_norm": 0.404296875, + "learning_rate": 4.314682239243445e-05, + "loss": 2.2955, + "step": 32532 + }, + { + "epoch": 1.39, + "grad_norm": 0.4609375, + "learning_rate": 4.3141304164652896e-05, + "loss": 2.3076, + "step": 32533 + }, + { + "epoch": 1.39, + "grad_norm": 0.41796875, + "learning_rate": 4.313578619271471e-05, + "loss": 2.4017, + "step": 32534 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.313026847664473e-05, + "loss": 2.564, + "step": 32535 + }, + { + "epoch": 1.39, + "grad_norm": 0.423828125, + "learning_rate": 4.312475101646777e-05, + "loss": 2.506, + "step": 32536 + }, + { + "epoch": 1.39, + "grad_norm": 0.470703125, + "learning_rate": 4.3119233812208606e-05, + "loss": 2.4666, + "step": 32537 + }, + { + "epoch": 1.39, + "grad_norm": 0.470703125, + "learning_rate": 4.311371686389209e-05, + "loss": 2.2924, + "step": 32538 + }, + { + "epoch": 1.39, + "grad_norm": 0.466796875, + "learning_rate": 4.310820017154311e-05, + "loss": 2.5365, + "step": 32539 + }, + { + "epoch": 1.39, + "grad_norm": 0.49609375, + "learning_rate": 4.310268373518639e-05, + "loss": 2.5345, + "step": 32540 + }, + { + "epoch": 1.39, + "grad_norm": 0.435546875, + "learning_rate": 4.309716755484684e-05, + "loss": 2.4266, + "step": 32541 + }, + { + "epoch": 1.39, + "grad_norm": 0.51171875, + "learning_rate": 4.30916516305492e-05, + "loss": 2.3703, + "step": 32542 + }, + { + "epoch": 1.39, + "grad_norm": 0.43359375, + "learning_rate": 4.308613596231834e-05, + "loss": 2.4945, + "step": 32543 + }, + { + "epoch": 1.39, + "grad_norm": 0.4609375, + "learning_rate": 4.3080620550179094e-05, + "loss": 2.5471, + "step": 32544 + }, + { + "epoch": 1.39, + "grad_norm": 0.484375, + "learning_rate": 4.307510539415627e-05, + "loss": 2.491, + "step": 32545 + }, + { + "epoch": 1.39, + "grad_norm": 0.4296875, + "learning_rate": 4.3069590494274615e-05, + "loss": 2.4363, + "step": 32546 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.306407585055901e-05, + "loss": 2.5057, + "step": 32547 + }, + { + "epoch": 1.39, + "grad_norm": 0.455078125, + "learning_rate": 4.305856146303429e-05, + "loss": 2.4838, + "step": 32548 + }, + { + "epoch": 1.39, + "grad_norm": 0.439453125, + "learning_rate": 4.3053047331725185e-05, + "loss": 2.4017, + "step": 32549 + }, + { + "epoch": 1.39, + "grad_norm": 0.490234375, + "learning_rate": 4.3047533456656564e-05, + "loss": 2.5546, + "step": 32550 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.304201983785326e-05, + "loss": 2.5503, + "step": 32551 + }, + { + "epoch": 1.39, + "grad_norm": 0.453125, + "learning_rate": 4.3036506475340053e-05, + "loss": 2.6042, + "step": 32552 + }, + { + "epoch": 1.39, + "grad_norm": 0.482421875, + "learning_rate": 4.303099336914171e-05, + "loss": 2.3544, + "step": 32553 + }, + { + "epoch": 1.39, + "grad_norm": 0.416015625, + "learning_rate": 4.302548051928309e-05, + "loss": 2.3966, + "step": 32554 + }, + { + "epoch": 1.39, + "grad_norm": 0.47265625, + "learning_rate": 4.3019967925789015e-05, + "loss": 2.4854, + "step": 32555 + }, + { + "epoch": 1.39, + "grad_norm": 0.453125, + "learning_rate": 4.301445558868422e-05, + "loss": 2.4363, + "step": 32556 + }, + { + "epoch": 1.39, + "grad_norm": 0.455078125, + "learning_rate": 4.3008943507993595e-05, + "loss": 2.5398, + "step": 32557 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.300343168374185e-05, + "loss": 2.4617, + "step": 32558 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.2997920115953835e-05, + "loss": 2.3562, + "step": 32559 + }, + { + "epoch": 1.39, + "grad_norm": 0.498046875, + "learning_rate": 4.2992408804654385e-05, + "loss": 2.456, + "step": 32560 + }, + { + "epoch": 1.39, + "grad_norm": 0.470703125, + "learning_rate": 4.298689774986826e-05, + "loss": 2.3018, + "step": 32561 + }, + { + "epoch": 1.39, + "grad_norm": 0.439453125, + "learning_rate": 4.2981386951620195e-05, + "loss": 2.5928, + "step": 32562 + }, + { + "epoch": 1.39, + "grad_norm": 0.5625, + "learning_rate": 4.297587640993511e-05, + "loss": 2.4182, + "step": 32563 + }, + { + "epoch": 1.39, + "grad_norm": 0.439453125, + "learning_rate": 4.297036612483776e-05, + "loss": 2.443, + "step": 32564 + }, + { + "epoch": 1.39, + "grad_norm": 0.4765625, + "learning_rate": 4.296485609635287e-05, + "loss": 2.3681, + "step": 32565 + }, + { + "epoch": 1.39, + "grad_norm": 0.53125, + "learning_rate": 4.295934632450528e-05, + "loss": 2.4951, + "step": 32566 + }, + { + "epoch": 1.39, + "grad_norm": 0.478515625, + "learning_rate": 4.295383680931982e-05, + "loss": 2.4984, + "step": 32567 + }, + { + "epoch": 1.39, + "grad_norm": 0.474609375, + "learning_rate": 4.294832755082127e-05, + "loss": 2.5218, + "step": 32568 + }, + { + "epoch": 1.39, + "grad_norm": 0.443359375, + "learning_rate": 4.2942818549034336e-05, + "loss": 2.4998, + "step": 32569 + }, + { + "epoch": 1.39, + "grad_norm": 0.455078125, + "learning_rate": 4.293730980398387e-05, + "loss": 2.4709, + "step": 32570 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.29318013156947e-05, + "loss": 2.4866, + "step": 32571 + }, + { + "epoch": 1.39, + "grad_norm": 0.453125, + "learning_rate": 4.292629308419153e-05, + "loss": 2.3279, + "step": 32572 + }, + { + "epoch": 1.39, + "grad_norm": 0.46875, + "learning_rate": 4.292078510949922e-05, + "loss": 2.5001, + "step": 32573 + }, + { + "epoch": 1.39, + "grad_norm": 0.4765625, + "learning_rate": 4.291527739164247e-05, + "loss": 2.5423, + "step": 32574 + }, + { + "epoch": 1.39, + "grad_norm": 0.435546875, + "learning_rate": 4.2909769930646126e-05, + "loss": 2.6591, + "step": 32575 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.290426272653498e-05, + "loss": 2.6013, + "step": 32576 + }, + { + "epoch": 1.39, + "grad_norm": 0.431640625, + "learning_rate": 4.2898755779333785e-05, + "loss": 2.3746, + "step": 32577 + }, + { + "epoch": 1.39, + "grad_norm": 0.484375, + "learning_rate": 4.289324908906725e-05, + "loss": 2.5128, + "step": 32578 + }, + { + "epoch": 1.39, + "grad_norm": 0.515625, + "learning_rate": 4.28877426557603e-05, + "loss": 2.4696, + "step": 32579 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.288223647943763e-05, + "loss": 2.3218, + "step": 32580 + }, + { + "epoch": 1.39, + "grad_norm": 0.50390625, + "learning_rate": 4.287673056012399e-05, + "loss": 2.4547, + "step": 32581 + }, + { + "epoch": 1.39, + "grad_norm": 0.484375, + "learning_rate": 4.2871224897844187e-05, + "loss": 2.4269, + "step": 32582 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.286571949262302e-05, + "loss": 2.346, + "step": 32583 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.286021434448525e-05, + "loss": 2.4553, + "step": 32584 + }, + { + "epoch": 1.39, + "grad_norm": 0.52734375, + "learning_rate": 4.285470945345559e-05, + "loss": 2.6077, + "step": 32585 + }, + { + "epoch": 1.39, + "grad_norm": 0.427734375, + "learning_rate": 4.284920481955885e-05, + "loss": 2.3237, + "step": 32586 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.284370044281984e-05, + "loss": 2.4797, + "step": 32587 + }, + { + "epoch": 1.39, + "grad_norm": 0.41015625, + "learning_rate": 4.2838196323263255e-05, + "loss": 2.5153, + "step": 32588 + }, + { + "epoch": 1.39, + "grad_norm": 0.455078125, + "learning_rate": 4.2832692460913926e-05, + "loss": 2.4127, + "step": 32589 + }, + { + "epoch": 1.39, + "grad_norm": 0.408203125, + "learning_rate": 4.282718885579656e-05, + "loss": 2.4122, + "step": 32590 + }, + { + "epoch": 1.39, + "grad_norm": 0.466796875, + "learning_rate": 4.282168550793595e-05, + "loss": 2.397, + "step": 32591 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.28161824173569e-05, + "loss": 2.4476, + "step": 32592 + }, + { + "epoch": 1.39, + "grad_norm": 0.5078125, + "learning_rate": 4.281067958408412e-05, + "loss": 2.4212, + "step": 32593 + }, + { + "epoch": 1.39, + "grad_norm": 0.5390625, + "learning_rate": 4.2805177008142316e-05, + "loss": 2.5004, + "step": 32594 + }, + { + "epoch": 1.39, + "grad_norm": 0.51171875, + "learning_rate": 4.279967468955638e-05, + "loss": 2.4615, + "step": 32595 + }, + { + "epoch": 1.39, + "grad_norm": 0.5, + "learning_rate": 4.2794172628351e-05, + "loss": 2.5135, + "step": 32596 + }, + { + "epoch": 1.39, + "grad_norm": 0.4453125, + "learning_rate": 4.2788670824550914e-05, + "loss": 2.5985, + "step": 32597 + }, + { + "epoch": 1.39, + "grad_norm": 0.4765625, + "learning_rate": 4.278316927818089e-05, + "loss": 2.6936, + "step": 32598 + }, + { + "epoch": 1.39, + "grad_norm": 0.515625, + "learning_rate": 4.2777667989265736e-05, + "loss": 2.4909, + "step": 32599 + }, + { + "epoch": 1.39, + "grad_norm": 0.462890625, + "learning_rate": 4.277216695783016e-05, + "loss": 2.3834, + "step": 32600 + }, + { + "epoch": 1.39, + "grad_norm": 0.4609375, + "learning_rate": 4.276666618389888e-05, + "loss": 2.5293, + "step": 32601 + }, + { + "epoch": 1.39, + "grad_norm": 0.47265625, + "learning_rate": 4.2761165667496695e-05, + "loss": 2.3702, + "step": 32602 + }, + { + "epoch": 1.39, + "grad_norm": 0.54296875, + "learning_rate": 4.275566540864838e-05, + "loss": 2.5075, + "step": 32603 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.275016540737862e-05, + "loss": 2.4359, + "step": 32604 + }, + { + "epoch": 1.39, + "grad_norm": 0.462890625, + "learning_rate": 4.274466566371221e-05, + "loss": 2.5388, + "step": 32605 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.273916617767386e-05, + "loss": 2.4858, + "step": 32606 + }, + { + "epoch": 1.39, + "grad_norm": 0.50390625, + "learning_rate": 4.273366694928833e-05, + "loss": 2.5355, + "step": 32607 + }, + { + "epoch": 1.39, + "grad_norm": 0.484375, + "learning_rate": 4.27281679785804e-05, + "loss": 2.4961, + "step": 32608 + }, + { + "epoch": 1.39, + "grad_norm": 0.443359375, + "learning_rate": 4.27226692655748e-05, + "loss": 2.4752, + "step": 32609 + }, + { + "epoch": 1.39, + "grad_norm": 0.47265625, + "learning_rate": 4.271717081029617e-05, + "loss": 2.4396, + "step": 32610 + }, + { + "epoch": 1.39, + "grad_norm": 0.498046875, + "learning_rate": 4.2711672612769414e-05, + "loss": 2.4114, + "step": 32611 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.270617467301919e-05, + "loss": 2.4527, + "step": 32612 + }, + { + "epoch": 1.39, + "grad_norm": 0.42578125, + "learning_rate": 4.270067699107021e-05, + "loss": 2.5093, + "step": 32613 + }, + { + "epoch": 1.39, + "grad_norm": 0.490234375, + "learning_rate": 4.269517956694724e-05, + "loss": 2.521, + "step": 32614 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.2689682400675054e-05, + "loss": 2.5289, + "step": 32615 + }, + { + "epoch": 1.39, + "grad_norm": 0.453125, + "learning_rate": 4.268418549227833e-05, + "loss": 2.4944, + "step": 32616 + }, + { + "epoch": 1.39, + "grad_norm": 0.494140625, + "learning_rate": 4.2678688841781854e-05, + "loss": 2.5273, + "step": 32617 + }, + { + "epoch": 1.39, + "grad_norm": 0.482421875, + "learning_rate": 4.267319244921032e-05, + "loss": 2.3933, + "step": 32618 + }, + { + "epoch": 1.39, + "grad_norm": 0.4609375, + "learning_rate": 4.2667696314588444e-05, + "loss": 2.4725, + "step": 32619 + }, + { + "epoch": 1.39, + "grad_norm": 0.439453125, + "learning_rate": 4.266220043794098e-05, + "loss": 2.4927, + "step": 32620 + }, + { + "epoch": 1.39, + "grad_norm": 0.466796875, + "learning_rate": 4.265670481929269e-05, + "loss": 2.1945, + "step": 32621 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.2651209458668274e-05, + "loss": 2.3849, + "step": 32622 + }, + { + "epoch": 1.39, + "grad_norm": 0.48828125, + "learning_rate": 4.264571435609239e-05, + "loss": 2.4238, + "step": 32623 + }, + { + "epoch": 1.39, + "grad_norm": 0.51953125, + "learning_rate": 4.264021951158989e-05, + "loss": 2.5371, + "step": 32624 + }, + { + "epoch": 1.39, + "grad_norm": 0.47265625, + "learning_rate": 4.263472492518545e-05, + "loss": 2.533, + "step": 32625 + }, + { + "epoch": 1.39, + "grad_norm": 0.443359375, + "learning_rate": 4.2629230596903756e-05, + "loss": 2.3818, + "step": 32626 + }, + { + "epoch": 1.39, + "grad_norm": 0.486328125, + "learning_rate": 4.262373652676954e-05, + "loss": 2.4141, + "step": 32627 + }, + { + "epoch": 1.39, + "grad_norm": 0.5546875, + "learning_rate": 4.261824271480759e-05, + "loss": 2.5594, + "step": 32628 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.261274916104253e-05, + "loss": 2.437, + "step": 32629 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.260725586549917e-05, + "loss": 2.4233, + "step": 32630 + }, + { + "epoch": 1.39, + "grad_norm": 0.447265625, + "learning_rate": 4.260176282820215e-05, + "loss": 2.5528, + "step": 32631 + }, + { + "epoch": 1.39, + "grad_norm": 0.4609375, + "learning_rate": 4.2596270049176214e-05, + "loss": 2.4755, + "step": 32632 + }, + { + "epoch": 1.39, + "grad_norm": 0.423828125, + "learning_rate": 4.2590777528446125e-05, + "loss": 2.4148, + "step": 32633 + }, + { + "epoch": 1.39, + "grad_norm": 0.412109375, + "learning_rate": 4.2585285266036555e-05, + "loss": 2.4349, + "step": 32634 + }, + { + "epoch": 1.39, + "grad_norm": 0.453125, + "learning_rate": 4.2579793261972175e-05, + "loss": 2.4531, + "step": 32635 + }, + { + "epoch": 1.39, + "grad_norm": 0.37890625, + "learning_rate": 4.257430151627775e-05, + "loss": 2.4333, + "step": 32636 + }, + { + "epoch": 1.39, + "grad_norm": 0.42578125, + "learning_rate": 4.2568810028978026e-05, + "loss": 2.3408, + "step": 32637 + }, + { + "epoch": 1.39, + "grad_norm": 0.4921875, + "learning_rate": 4.2563318800097654e-05, + "loss": 2.4282, + "step": 32638 + }, + { + "epoch": 1.39, + "grad_norm": 0.470703125, + "learning_rate": 4.255782782966128e-05, + "loss": 2.3665, + "step": 32639 + }, + { + "epoch": 1.39, + "grad_norm": 0.5078125, + "learning_rate": 4.2552337117693775e-05, + "loss": 2.4366, + "step": 32640 + }, + { + "epoch": 1.39, + "grad_norm": 0.421875, + "learning_rate": 4.2546846664219754e-05, + "loss": 2.4004, + "step": 32641 + }, + { + "epoch": 1.39, + "grad_norm": 0.4609375, + "learning_rate": 4.254135646926388e-05, + "loss": 2.5379, + "step": 32642 + }, + { + "epoch": 1.39, + "grad_norm": 0.419921875, + "learning_rate": 4.253586653285091e-05, + "loss": 2.569, + "step": 32643 + }, + { + "epoch": 1.39, + "grad_norm": 0.427734375, + "learning_rate": 4.2530376855005583e-05, + "loss": 2.3404, + "step": 32644 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.2524887435752516e-05, + "loss": 2.3961, + "step": 32645 + }, + { + "epoch": 1.39, + "grad_norm": 0.486328125, + "learning_rate": 4.251939827511647e-05, + "loss": 2.3899, + "step": 32646 + }, + { + "epoch": 1.39, + "grad_norm": 0.484375, + "learning_rate": 4.251390937312212e-05, + "loss": 2.4777, + "step": 32647 + }, + { + "epoch": 1.39, + "grad_norm": 0.478515625, + "learning_rate": 4.250842072979415e-05, + "loss": 2.4035, + "step": 32648 + }, + { + "epoch": 1.39, + "grad_norm": 0.60546875, + "learning_rate": 4.250293234515731e-05, + "loss": 2.5809, + "step": 32649 + }, + { + "epoch": 1.39, + "grad_norm": 0.51171875, + "learning_rate": 4.249744421923627e-05, + "loss": 2.5071, + "step": 32650 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.249195635205567e-05, + "loss": 2.4029, + "step": 32651 + }, + { + "epoch": 1.39, + "grad_norm": 0.412109375, + "learning_rate": 4.248646874364026e-05, + "loss": 2.2885, + "step": 32652 + }, + { + "epoch": 1.39, + "grad_norm": 0.443359375, + "learning_rate": 4.248098139401476e-05, + "loss": 2.5448, + "step": 32653 + }, + { + "epoch": 1.39, + "grad_norm": 0.474609375, + "learning_rate": 4.247549430320382e-05, + "loss": 2.3922, + "step": 32654 + }, + { + "epoch": 1.39, + "grad_norm": 0.4921875, + "learning_rate": 4.247000747123205e-05, + "loss": 2.5513, + "step": 32655 + }, + { + "epoch": 1.39, + "grad_norm": 0.42578125, + "learning_rate": 4.24645208981243e-05, + "loss": 2.351, + "step": 32656 + }, + { + "epoch": 1.39, + "grad_norm": 0.45703125, + "learning_rate": 4.2459034583905185e-05, + "loss": 2.4984, + "step": 32657 + }, + { + "epoch": 1.39, + "grad_norm": 0.458984375, + "learning_rate": 4.2453548528599355e-05, + "loss": 2.5114, + "step": 32658 + }, + { + "epoch": 1.39, + "grad_norm": 0.427734375, + "learning_rate": 4.244806273223151e-05, + "loss": 2.3816, + "step": 32659 + }, + { + "epoch": 1.39, + "grad_norm": 0.4765625, + "learning_rate": 4.24425771948264e-05, + "loss": 2.4747, + "step": 32660 + }, + { + "epoch": 1.39, + "grad_norm": 0.51171875, + "learning_rate": 4.2437091916408614e-05, + "loss": 2.3753, + "step": 32661 + }, + { + "epoch": 1.39, + "grad_norm": 0.53125, + "learning_rate": 4.243160689700291e-05, + "loss": 2.5797, + "step": 32662 + }, + { + "epoch": 1.39, + "grad_norm": 0.45703125, + "learning_rate": 4.2426122136633906e-05, + "loss": 2.3333, + "step": 32663 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.242063763532631e-05, + "loss": 2.4928, + "step": 32664 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.241515339310483e-05, + "loss": 2.5504, + "step": 32665 + }, + { + "epoch": 1.39, + "grad_norm": 0.466796875, + "learning_rate": 4.2409669409994125e-05, + "loss": 2.4427, + "step": 32666 + }, + { + "epoch": 1.39, + "grad_norm": 0.412109375, + "learning_rate": 4.240418568601883e-05, + "loss": 2.4528, + "step": 32667 + }, + { + "epoch": 1.39, + "grad_norm": 0.482421875, + "learning_rate": 4.239870222120363e-05, + "loss": 2.4685, + "step": 32668 + }, + { + "epoch": 1.39, + "grad_norm": 0.427734375, + "learning_rate": 4.239321901557327e-05, + "loss": 2.4229, + "step": 32669 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.238773606915234e-05, + "loss": 2.5325, + "step": 32670 + }, + { + "epoch": 1.39, + "grad_norm": 0.4375, + "learning_rate": 4.238225338196553e-05, + "loss": 2.297, + "step": 32671 + }, + { + "epoch": 1.39, + "grad_norm": 0.5078125, + "learning_rate": 4.237677095403757e-05, + "loss": 2.3883, + "step": 32672 + }, + { + "epoch": 1.39, + "grad_norm": 0.4609375, + "learning_rate": 4.2371288785393085e-05, + "loss": 2.4314, + "step": 32673 + }, + { + "epoch": 1.39, + "grad_norm": 0.458984375, + "learning_rate": 4.236580687605669e-05, + "loss": 2.6617, + "step": 32674 + }, + { + "epoch": 1.39, + "grad_norm": 0.515625, + "learning_rate": 4.23603252260531e-05, + "loss": 2.4961, + "step": 32675 + }, + { + "epoch": 1.39, + "grad_norm": 0.490234375, + "learning_rate": 4.2354843835407035e-05, + "loss": 2.4817, + "step": 32676 + }, + { + "epoch": 1.39, + "grad_norm": 0.53515625, + "learning_rate": 4.234936270414305e-05, + "loss": 2.58, + "step": 32677 + }, + { + "epoch": 1.39, + "grad_norm": 0.412109375, + "learning_rate": 4.2343881832285915e-05, + "loss": 2.394, + "step": 32678 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.233840121986019e-05, + "loss": 2.4033, + "step": 32679 + }, + { + "epoch": 1.39, + "grad_norm": 0.498046875, + "learning_rate": 4.23329208668906e-05, + "loss": 2.4414, + "step": 32680 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.232744077340183e-05, + "loss": 2.3295, + "step": 32681 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.232196093941848e-05, + "loss": 2.3966, + "step": 32682 + }, + { + "epoch": 1.39, + "grad_norm": 0.5, + "learning_rate": 4.231648136496521e-05, + "loss": 2.4129, + "step": 32683 + }, + { + "epoch": 1.39, + "grad_norm": 0.47265625, + "learning_rate": 4.231100205006668e-05, + "loss": 2.3796, + "step": 32684 + }, + { + "epoch": 1.39, + "grad_norm": 0.458984375, + "learning_rate": 4.23055229947476e-05, + "loss": 2.4588, + "step": 32685 + }, + { + "epoch": 1.39, + "grad_norm": 0.45703125, + "learning_rate": 4.2300044199032555e-05, + "loss": 2.5084, + "step": 32686 + }, + { + "epoch": 1.39, + "grad_norm": 0.5078125, + "learning_rate": 4.2294565662946204e-05, + "loss": 2.4634, + "step": 32687 + }, + { + "epoch": 1.39, + "grad_norm": 0.43359375, + "learning_rate": 4.228908738651327e-05, + "loss": 2.3823, + "step": 32688 + }, + { + "epoch": 1.39, + "grad_norm": 0.51953125, + "learning_rate": 4.228360936975835e-05, + "loss": 2.601, + "step": 32689 + }, + { + "epoch": 1.39, + "grad_norm": 0.431640625, + "learning_rate": 4.227813161270605e-05, + "loss": 2.4249, + "step": 32690 + }, + { + "epoch": 1.39, + "grad_norm": 0.45703125, + "learning_rate": 4.227265411538107e-05, + "loss": 2.4942, + "step": 32691 + }, + { + "epoch": 1.39, + "grad_norm": 0.455078125, + "learning_rate": 4.226717687780809e-05, + "loss": 2.338, + "step": 32692 + }, + { + "epoch": 1.39, + "grad_norm": 0.51953125, + "learning_rate": 4.226169990001167e-05, + "loss": 2.4597, + "step": 32693 + }, + { + "epoch": 1.39, + "grad_norm": 0.48828125, + "learning_rate": 4.225622318201654e-05, + "loss": 2.3617, + "step": 32694 + }, + { + "epoch": 1.39, + "grad_norm": 0.51171875, + "learning_rate": 4.225074672384727e-05, + "loss": 2.4811, + "step": 32695 + }, + { + "epoch": 1.39, + "grad_norm": 0.59375, + "learning_rate": 4.2245270525528526e-05, + "loss": 2.4098, + "step": 32696 + }, + { + "epoch": 1.39, + "grad_norm": 0.435546875, + "learning_rate": 4.2239794587085e-05, + "loss": 2.3466, + "step": 32697 + }, + { + "epoch": 1.39, + "grad_norm": 0.51953125, + "learning_rate": 4.2234318908541294e-05, + "loss": 2.4415, + "step": 32698 + }, + { + "epoch": 1.39, + "grad_norm": 0.44140625, + "learning_rate": 4.2228843489921996e-05, + "loss": 2.4268, + "step": 32699 + }, + { + "epoch": 1.39, + "grad_norm": 0.50390625, + "learning_rate": 4.222336833125179e-05, + "loss": 2.4738, + "step": 32700 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.2217893432555354e-05, + "loss": 2.3415, + "step": 32701 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.2212418793857234e-05, + "loss": 2.3885, + "step": 32702 + }, + { + "epoch": 1.39, + "grad_norm": 0.482421875, + "learning_rate": 4.220694441518213e-05, + "loss": 2.5333, + "step": 32703 + }, + { + "epoch": 1.39, + "grad_norm": 0.439453125, + "learning_rate": 4.220147029655467e-05, + "loss": 2.4735, + "step": 32704 + }, + { + "epoch": 1.39, + "grad_norm": 0.4765625, + "learning_rate": 4.2195996437999486e-05, + "loss": 2.521, + "step": 32705 + }, + { + "epoch": 1.39, + "grad_norm": 0.498046875, + "learning_rate": 4.2190522839541146e-05, + "loss": 2.4574, + "step": 32706 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.2185049501204334e-05, + "loss": 2.4455, + "step": 32707 + }, + { + "epoch": 1.39, + "grad_norm": 0.466796875, + "learning_rate": 4.21795764230137e-05, + "loss": 2.427, + "step": 32708 + }, + { + "epoch": 1.39, + "grad_norm": 0.48046875, + "learning_rate": 4.2174103604993806e-05, + "loss": 2.4591, + "step": 32709 + }, + { + "epoch": 1.39, + "grad_norm": 0.47265625, + "learning_rate": 4.216863104716935e-05, + "loss": 2.4879, + "step": 32710 + }, + { + "epoch": 1.39, + "grad_norm": 0.46484375, + "learning_rate": 4.216315874956489e-05, + "loss": 2.4806, + "step": 32711 + }, + { + "epoch": 1.39, + "grad_norm": 0.470703125, + "learning_rate": 4.215768671220507e-05, + "loss": 2.4573, + "step": 32712 + }, + { + "epoch": 1.39, + "grad_norm": 0.5234375, + "learning_rate": 4.2152214935114566e-05, + "loss": 2.5227, + "step": 32713 + }, + { + "epoch": 1.39, + "grad_norm": 0.458984375, + "learning_rate": 4.2146743418317946e-05, + "loss": 2.5133, + "step": 32714 + }, + { + "epoch": 1.39, + "grad_norm": 0.470703125, + "learning_rate": 4.2141272161839806e-05, + "loss": 2.4725, + "step": 32715 + }, + { + "epoch": 1.39, + "grad_norm": 0.482421875, + "learning_rate": 4.213580116570479e-05, + "loss": 2.4515, + "step": 32716 + }, + { + "epoch": 1.39, + "grad_norm": 0.494140625, + "learning_rate": 4.213033042993756e-05, + "loss": 2.4176, + "step": 32717 + }, + { + "epoch": 1.39, + "grad_norm": 0.451171875, + "learning_rate": 4.212485995456266e-05, + "loss": 2.4597, + "step": 32718 + }, + { + "epoch": 1.39, + "grad_norm": 0.431640625, + "learning_rate": 4.211938973960474e-05, + "loss": 2.4533, + "step": 32719 + }, + { + "epoch": 1.39, + "grad_norm": 0.46875, + "learning_rate": 4.2113919785088454e-05, + "loss": 2.549, + "step": 32720 + }, + { + "epoch": 1.39, + "grad_norm": 0.39453125, + "learning_rate": 4.2108450091038364e-05, + "loss": 2.334, + "step": 32721 + }, + { + "epoch": 1.39, + "grad_norm": 0.4765625, + "learning_rate": 4.210298065747906e-05, + "loss": 2.4064, + "step": 32722 + }, + { + "epoch": 1.39, + "grad_norm": 0.51171875, + "learning_rate": 4.209751148443518e-05, + "loss": 2.5851, + "step": 32723 + }, + { + "epoch": 1.39, + "grad_norm": 0.447265625, + "learning_rate": 4.2092042571931375e-05, + "loss": 2.5193, + "step": 32724 + }, + { + "epoch": 1.39, + "grad_norm": 0.42578125, + "learning_rate": 4.208657391999219e-05, + "loss": 2.4487, + "step": 32725 + }, + { + "epoch": 1.39, + "grad_norm": 0.47265625, + "learning_rate": 4.208110552864228e-05, + "loss": 2.3658, + "step": 32726 + }, + { + "epoch": 1.39, + "grad_norm": 0.46875, + "learning_rate": 4.20756373979062e-05, + "loss": 2.4941, + "step": 32727 + }, + { + "epoch": 1.39, + "grad_norm": 0.482421875, + "learning_rate": 4.2070169527808575e-05, + "loss": 2.4777, + "step": 32728 + }, + { + "epoch": 1.39, + "grad_norm": 0.52734375, + "learning_rate": 4.206470191837406e-05, + "loss": 2.4411, + "step": 32729 + }, + { + "epoch": 1.39, + "grad_norm": 0.474609375, + "learning_rate": 4.2059234569627216e-05, + "loss": 2.3567, + "step": 32730 + }, + { + "epoch": 1.39, + "grad_norm": 0.4296875, + "learning_rate": 4.20537674815926e-05, + "loss": 2.4565, + "step": 32731 + }, + { + "epoch": 1.39, + "grad_norm": 0.466796875, + "learning_rate": 4.204830065429486e-05, + "loss": 2.5002, + "step": 32732 + }, + { + "epoch": 1.39, + "grad_norm": 0.4375, + "learning_rate": 4.204283408775861e-05, + "loss": 2.3839, + "step": 32733 + }, + { + "epoch": 1.39, + "grad_norm": 0.4921875, + "learning_rate": 4.20373677820084e-05, + "loss": 2.4863, + "step": 32734 + }, + { + "epoch": 1.39, + "grad_norm": 0.49609375, + "learning_rate": 4.203190173706885e-05, + "loss": 2.4169, + "step": 32735 + }, + { + "epoch": 1.39, + "grad_norm": 0.435546875, + "learning_rate": 4.202643595296459e-05, + "loss": 2.3932, + "step": 32736 + }, + { + "epoch": 1.4, + "grad_norm": 0.45703125, + "learning_rate": 4.2020970429720185e-05, + "loss": 2.3582, + "step": 32737 + }, + { + "epoch": 1.4, + "grad_norm": 0.46875, + "learning_rate": 4.201550516736019e-05, + "loss": 2.4579, + "step": 32738 + }, + { + "epoch": 1.4, + "grad_norm": 0.474609375, + "learning_rate": 4.201004016590926e-05, + "loss": 2.4957, + "step": 32739 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.200457542539193e-05, + "loss": 2.3156, + "step": 32740 + }, + { + "epoch": 1.4, + "grad_norm": 0.458984375, + "learning_rate": 4.199911094583281e-05, + "loss": 2.5506, + "step": 32741 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.199364672725653e-05, + "loss": 2.3513, + "step": 32742 + }, + { + "epoch": 1.4, + "grad_norm": 0.46875, + "learning_rate": 4.198818276968765e-05, + "loss": 2.452, + "step": 32743 + }, + { + "epoch": 1.4, + "grad_norm": 0.50390625, + "learning_rate": 4.198271907315066e-05, + "loss": 2.2742, + "step": 32744 + }, + { + "epoch": 1.4, + "grad_norm": 0.41796875, + "learning_rate": 4.197725563767031e-05, + "loss": 2.5304, + "step": 32745 + }, + { + "epoch": 1.4, + "grad_norm": 0.4375, + "learning_rate": 4.19717924632711e-05, + "loss": 2.491, + "step": 32746 + }, + { + "epoch": 1.4, + "grad_norm": 0.49609375, + "learning_rate": 4.1966329549977575e-05, + "loss": 2.4983, + "step": 32747 + }, + { + "epoch": 1.4, + "grad_norm": 0.451171875, + "learning_rate": 4.1960866897814375e-05, + "loss": 2.4601, + "step": 32748 + }, + { + "epoch": 1.4, + "grad_norm": 0.51953125, + "learning_rate": 4.1955404506806097e-05, + "loss": 2.574, + "step": 32749 + }, + { + "epoch": 1.4, + "grad_norm": 0.48046875, + "learning_rate": 4.194994237697725e-05, + "loss": 2.4587, + "step": 32750 + }, + { + "epoch": 1.4, + "grad_norm": 0.431640625, + "learning_rate": 4.194448050835247e-05, + "loss": 2.4784, + "step": 32751 + }, + { + "epoch": 1.4, + "grad_norm": 0.41015625, + "learning_rate": 4.1939018900956286e-05, + "loss": 2.4579, + "step": 32752 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.1933557554813326e-05, + "loss": 2.5047, + "step": 32753 + }, + { + "epoch": 1.4, + "grad_norm": 0.5, + "learning_rate": 4.192809646994811e-05, + "loss": 2.2616, + "step": 32754 + }, + { + "epoch": 1.4, + "grad_norm": 0.427734375, + "learning_rate": 4.192263564638527e-05, + "loss": 2.4192, + "step": 32755 + }, + { + "epoch": 1.4, + "grad_norm": 0.47265625, + "learning_rate": 4.1917175084149305e-05, + "loss": 2.4457, + "step": 32756 + }, + { + "epoch": 1.4, + "grad_norm": 0.443359375, + "learning_rate": 4.191171478326482e-05, + "loss": 2.5813, + "step": 32757 + }, + { + "epoch": 1.4, + "grad_norm": 0.47265625, + "learning_rate": 4.190625474375643e-05, + "loss": 2.475, + "step": 32758 + }, + { + "epoch": 1.4, + "grad_norm": 0.4375, + "learning_rate": 4.190079496564867e-05, + "loss": 2.3465, + "step": 32759 + }, + { + "epoch": 1.4, + "grad_norm": 0.455078125, + "learning_rate": 4.1895335448966033e-05, + "loss": 2.4998, + "step": 32760 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.1889876193733214e-05, + "loss": 2.4049, + "step": 32761 + }, + { + "epoch": 1.4, + "grad_norm": 0.50390625, + "learning_rate": 4.1884417199974716e-05, + "loss": 2.4801, + "step": 32762 + }, + { + "epoch": 1.4, + "grad_norm": 0.43359375, + "learning_rate": 4.1878958467715054e-05, + "loss": 2.503, + "step": 32763 + }, + { + "epoch": 1.4, + "grad_norm": 0.439453125, + "learning_rate": 4.187349999697886e-05, + "loss": 2.4592, + "step": 32764 + }, + { + "epoch": 1.4, + "grad_norm": 0.4375, + "learning_rate": 4.18680417877907e-05, + "loss": 2.5374, + "step": 32765 + }, + { + "epoch": 1.4, + "grad_norm": 0.482421875, + "learning_rate": 4.1862583840175065e-05, + "loss": 2.3567, + "step": 32766 + }, + { + "epoch": 1.4, + "grad_norm": 0.462890625, + "learning_rate": 4.18571261541566e-05, + "loss": 2.4544, + "step": 32767 + }, + { + "epoch": 1.4, + "grad_norm": 0.48828125, + "learning_rate": 4.1851668729759776e-05, + "loss": 2.5043, + "step": 32768 + }, + { + "epoch": 1.4, + "grad_norm": 0.455078125, + "learning_rate": 4.184621156700923e-05, + "loss": 2.3128, + "step": 32769 + }, + { + "epoch": 1.4, + "grad_norm": 0.49609375, + "learning_rate": 4.184075466592945e-05, + "loss": 2.5762, + "step": 32770 + }, + { + "epoch": 1.4, + "grad_norm": 0.484375, + "learning_rate": 4.183529802654505e-05, + "loss": 2.5921, + "step": 32771 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.182984164888051e-05, + "loss": 2.522, + "step": 32772 + }, + { + "epoch": 1.4, + "grad_norm": 0.46875, + "learning_rate": 4.182438553296043e-05, + "loss": 2.4097, + "step": 32773 + }, + { + "epoch": 1.4, + "grad_norm": 0.5390625, + "learning_rate": 4.181892967880939e-05, + "loss": 2.5617, + "step": 32774 + }, + { + "epoch": 1.4, + "grad_norm": 0.474609375, + "learning_rate": 4.181347408645191e-05, + "loss": 2.4702, + "step": 32775 + }, + { + "epoch": 1.4, + "grad_norm": 0.4375, + "learning_rate": 4.180801875591245e-05, + "loss": 2.3876, + "step": 32776 + }, + { + "epoch": 1.4, + "grad_norm": 0.44921875, + "learning_rate": 4.180256368721571e-05, + "loss": 2.4432, + "step": 32777 + }, + { + "epoch": 1.4, + "grad_norm": 0.4609375, + "learning_rate": 4.179710888038617e-05, + "loss": 2.5239, + "step": 32778 + }, + { + "epoch": 1.4, + "grad_norm": 0.45703125, + "learning_rate": 4.179165433544834e-05, + "loss": 2.4408, + "step": 32779 + }, + { + "epoch": 1.4, + "grad_norm": 0.5078125, + "learning_rate": 4.178620005242678e-05, + "loss": 2.4466, + "step": 32780 + }, + { + "epoch": 1.4, + "grad_norm": 0.388671875, + "learning_rate": 4.1780746031346084e-05, + "loss": 2.5027, + "step": 32781 + }, + { + "epoch": 1.4, + "grad_norm": 0.50390625, + "learning_rate": 4.177529227223073e-05, + "loss": 2.4179, + "step": 32782 + }, + { + "epoch": 1.4, + "grad_norm": 0.48828125, + "learning_rate": 4.17698387751053e-05, + "loss": 2.5231, + "step": 32783 + }, + { + "epoch": 1.4, + "grad_norm": 0.48046875, + "learning_rate": 4.176438553999429e-05, + "loss": 2.4629, + "step": 32784 + }, + { + "epoch": 1.4, + "grad_norm": 0.46875, + "learning_rate": 4.1758932566922294e-05, + "loss": 2.4298, + "step": 32785 + }, + { + "epoch": 1.4, + "grad_norm": 0.408203125, + "learning_rate": 4.175347985591378e-05, + "loss": 2.4725, + "step": 32786 + }, + { + "epoch": 1.4, + "grad_norm": 0.46875, + "learning_rate": 4.174802740699336e-05, + "loss": 2.4448, + "step": 32787 + }, + { + "epoch": 1.4, + "grad_norm": 0.466796875, + "learning_rate": 4.174257522018549e-05, + "loss": 2.3672, + "step": 32788 + }, + { + "epoch": 1.4, + "grad_norm": 0.43359375, + "learning_rate": 4.1737123295514745e-05, + "loss": 2.5742, + "step": 32789 + }, + { + "epoch": 1.4, + "grad_norm": 0.484375, + "learning_rate": 4.1731671633005685e-05, + "loss": 2.3709, + "step": 32790 + }, + { + "epoch": 1.4, + "grad_norm": 0.447265625, + "learning_rate": 4.172622023268281e-05, + "loss": 2.4594, + "step": 32791 + }, + { + "epoch": 1.4, + "grad_norm": 0.91796875, + "learning_rate": 4.172076909457057e-05, + "loss": 2.5336, + "step": 32792 + }, + { + "epoch": 1.4, + "grad_norm": 0.462890625, + "learning_rate": 4.171531821869365e-05, + "loss": 2.4616, + "step": 32793 + }, + { + "epoch": 1.4, + "grad_norm": 0.5, + "learning_rate": 4.170986760507649e-05, + "loss": 2.5776, + "step": 32794 + }, + { + "epoch": 1.4, + "grad_norm": 0.421875, + "learning_rate": 4.1704417253743586e-05, + "loss": 2.4414, + "step": 32795 + }, + { + "epoch": 1.4, + "grad_norm": 0.4921875, + "learning_rate": 4.169896716471951e-05, + "loss": 2.4537, + "step": 32796 + }, + { + "epoch": 1.4, + "grad_norm": 0.40625, + "learning_rate": 4.1693517338028796e-05, + "loss": 2.2963, + "step": 32797 + }, + { + "epoch": 1.4, + "grad_norm": 0.43359375, + "learning_rate": 4.168806777369593e-05, + "loss": 2.4806, + "step": 32798 + }, + { + "epoch": 1.4, + "grad_norm": 0.453125, + "learning_rate": 4.1682618471745475e-05, + "loss": 2.4325, + "step": 32799 + }, + { + "epoch": 1.4, + "grad_norm": 0.458984375, + "learning_rate": 4.167716943220188e-05, + "loss": 2.5193, + "step": 32800 + }, + { + "epoch": 1.4, + "grad_norm": 0.51171875, + "learning_rate": 4.167172065508975e-05, + "loss": 2.5103, + "step": 32801 + }, + { + "epoch": 1.4, + "grad_norm": 0.52734375, + "learning_rate": 4.1666272140433524e-05, + "loss": 2.6338, + "step": 32802 + }, + { + "epoch": 1.4, + "grad_norm": 0.462890625, + "learning_rate": 4.166082388825779e-05, + "loss": 2.5425, + "step": 32803 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.165537589858699e-05, + "loss": 2.4456, + "step": 32804 + }, + { + "epoch": 1.4, + "grad_norm": 0.435546875, + "learning_rate": 4.164992817144568e-05, + "loss": 2.3097, + "step": 32805 + }, + { + "epoch": 1.4, + "grad_norm": 0.5078125, + "learning_rate": 4.16444807068584e-05, + "loss": 2.5352, + "step": 32806 + }, + { + "epoch": 1.4, + "grad_norm": 0.490234375, + "learning_rate": 4.1639033504849644e-05, + "loss": 2.3893, + "step": 32807 + }, + { + "epoch": 1.4, + "grad_norm": 0.51953125, + "learning_rate": 4.1633586565443826e-05, + "loss": 2.5295, + "step": 32808 + }, + { + "epoch": 1.4, + "grad_norm": 0.431640625, + "learning_rate": 4.162813988866561e-05, + "loss": 2.3148, + "step": 32809 + }, + { + "epoch": 1.4, + "grad_norm": 0.439453125, + "learning_rate": 4.162269347453943e-05, + "loss": 2.3612, + "step": 32810 + }, + { + "epoch": 1.4, + "grad_norm": 0.44140625, + "learning_rate": 4.161724732308977e-05, + "loss": 2.5012, + "step": 32811 + }, + { + "epoch": 1.4, + "grad_norm": 0.443359375, + "learning_rate": 4.1611801434341144e-05, + "loss": 2.4264, + "step": 32812 + }, + { + "epoch": 1.4, + "grad_norm": 0.498046875, + "learning_rate": 4.160635580831812e-05, + "loss": 2.3128, + "step": 32813 + }, + { + "epoch": 1.4, + "grad_norm": 0.455078125, + "learning_rate": 4.1600910445045114e-05, + "loss": 2.4932, + "step": 32814 + }, + { + "epoch": 1.4, + "grad_norm": 0.490234375, + "learning_rate": 4.1595465344546714e-05, + "loss": 2.3897, + "step": 32815 + }, + { + "epoch": 1.4, + "grad_norm": 0.458984375, + "learning_rate": 4.1590020506847326e-05, + "loss": 2.4934, + "step": 32816 + }, + { + "epoch": 1.4, + "grad_norm": 0.439453125, + "learning_rate": 4.158457593197155e-05, + "loss": 2.4777, + "step": 32817 + }, + { + "epoch": 1.4, + "grad_norm": 0.40625, + "learning_rate": 4.157913161994379e-05, + "loss": 2.4328, + "step": 32818 + }, + { + "epoch": 1.4, + "grad_norm": 0.45703125, + "learning_rate": 4.1573687570788624e-05, + "loss": 2.5538, + "step": 32819 + }, + { + "epoch": 1.4, + "grad_norm": 0.455078125, + "learning_rate": 4.156824378453047e-05, + "loss": 2.4151, + "step": 32820 + }, + { + "epoch": 1.4, + "grad_norm": 0.52734375, + "learning_rate": 4.1562800261193866e-05, + "loss": 2.3056, + "step": 32821 + }, + { + "epoch": 1.4, + "grad_norm": 0.46875, + "learning_rate": 4.1557357000803345e-05, + "loss": 2.4862, + "step": 32822 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.1551914003383364e-05, + "loss": 2.5031, + "step": 32823 + }, + { + "epoch": 1.4, + "grad_norm": 0.490234375, + "learning_rate": 4.154647126895832e-05, + "loss": 2.5887, + "step": 32824 + }, + { + "epoch": 1.4, + "grad_norm": 0.451171875, + "learning_rate": 4.1541028797552885e-05, + "loss": 2.4157, + "step": 32825 + }, + { + "epoch": 1.4, + "grad_norm": 0.48046875, + "learning_rate": 4.153558658919145e-05, + "loss": 2.5274, + "step": 32826 + }, + { + "epoch": 1.4, + "grad_norm": 0.427734375, + "learning_rate": 4.153014464389846e-05, + "loss": 2.2837, + "step": 32827 + }, + { + "epoch": 1.4, + "grad_norm": 0.41015625, + "learning_rate": 4.152470296169847e-05, + "loss": 2.3603, + "step": 32828 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.1519261542615975e-05, + "loss": 2.4358, + "step": 32829 + }, + { + "epoch": 1.4, + "grad_norm": 0.53125, + "learning_rate": 4.151382038667538e-05, + "loss": 2.5074, + "step": 32830 + }, + { + "epoch": 1.4, + "grad_norm": 0.498046875, + "learning_rate": 4.150837949390127e-05, + "loss": 2.4396, + "step": 32831 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.150293886431804e-05, + "loss": 2.4452, + "step": 32832 + }, + { + "epoch": 1.4, + "grad_norm": 0.478515625, + "learning_rate": 4.149749849795025e-05, + "loss": 2.4488, + "step": 32833 + }, + { + "epoch": 1.4, + "grad_norm": 0.490234375, + "learning_rate": 4.1492058394822284e-05, + "loss": 2.4756, + "step": 32834 + }, + { + "epoch": 1.4, + "grad_norm": 0.50390625, + "learning_rate": 4.148661855495872e-05, + "loss": 2.5838, + "step": 32835 + }, + { + "epoch": 1.4, + "grad_norm": 0.466796875, + "learning_rate": 4.148117897838395e-05, + "loss": 2.5337, + "step": 32836 + }, + { + "epoch": 1.4, + "grad_norm": 0.5, + "learning_rate": 4.147573966512249e-05, + "loss": 2.5608, + "step": 32837 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.1470300615198845e-05, + "loss": 2.3475, + "step": 32838 + }, + { + "epoch": 1.4, + "grad_norm": 0.404296875, + "learning_rate": 4.1464861828637455e-05, + "loss": 2.3479, + "step": 32839 + }, + { + "epoch": 1.4, + "grad_norm": 0.5, + "learning_rate": 4.145942330546273e-05, + "loss": 2.2101, + "step": 32840 + }, + { + "epoch": 1.4, + "grad_norm": 0.4921875, + "learning_rate": 4.1453985045699276e-05, + "loss": 2.3912, + "step": 32841 + }, + { + "epoch": 1.4, + "grad_norm": 0.44921875, + "learning_rate": 4.144854704937149e-05, + "loss": 2.4379, + "step": 32842 + }, + { + "epoch": 1.4, + "grad_norm": 0.482421875, + "learning_rate": 4.1443109316503816e-05, + "loss": 2.4349, + "step": 32843 + }, + { + "epoch": 1.4, + "grad_norm": 0.515625, + "learning_rate": 4.143767184712074e-05, + "loss": 2.494, + "step": 32844 + }, + { + "epoch": 1.4, + "grad_norm": 0.4609375, + "learning_rate": 4.143223464124678e-05, + "loss": 2.2495, + "step": 32845 + }, + { + "epoch": 1.4, + "grad_norm": 0.46875, + "learning_rate": 4.142679769890632e-05, + "loss": 2.459, + "step": 32846 + }, + { + "epoch": 1.4, + "grad_norm": 0.44921875, + "learning_rate": 4.1421361020123907e-05, + "loss": 2.4179, + "step": 32847 + }, + { + "epoch": 1.4, + "grad_norm": 0.4609375, + "learning_rate": 4.141592460492392e-05, + "loss": 2.4832, + "step": 32848 + }, + { + "epoch": 1.4, + "grad_norm": 0.48046875, + "learning_rate": 4.1410488453330895e-05, + "loss": 2.4978, + "step": 32849 + }, + { + "epoch": 1.4, + "grad_norm": 0.40234375, + "learning_rate": 4.140505256536923e-05, + "loss": 2.3598, + "step": 32850 + }, + { + "epoch": 1.4, + "grad_norm": 0.498046875, + "learning_rate": 4.139961694106344e-05, + "loss": 2.5105, + "step": 32851 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.139418158043794e-05, + "loss": 2.3547, + "step": 32852 + }, + { + "epoch": 1.4, + "grad_norm": 0.484375, + "learning_rate": 4.1388746483517196e-05, + "loss": 2.4018, + "step": 32853 + }, + { + "epoch": 1.4, + "grad_norm": 0.453125, + "learning_rate": 4.1383311650325704e-05, + "loss": 2.5041, + "step": 32854 + }, + { + "epoch": 1.4, + "grad_norm": 0.46484375, + "learning_rate": 4.137787708088789e-05, + "loss": 2.2965, + "step": 32855 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.137244277522815e-05, + "loss": 2.5217, + "step": 32856 + }, + { + "epoch": 1.4, + "grad_norm": 0.54296875, + "learning_rate": 4.136700873337104e-05, + "loss": 2.3949, + "step": 32857 + }, + { + "epoch": 1.4, + "grad_norm": 0.44140625, + "learning_rate": 4.136157495534097e-05, + "loss": 2.5191, + "step": 32858 + }, + { + "epoch": 1.4, + "grad_norm": 0.482421875, + "learning_rate": 4.135614144116235e-05, + "loss": 2.4151, + "step": 32859 + }, + { + "epoch": 1.4, + "grad_norm": 0.47265625, + "learning_rate": 4.13507081908597e-05, + "loss": 2.3425, + "step": 32860 + }, + { + "epoch": 1.4, + "grad_norm": 0.546875, + "learning_rate": 4.134527520445739e-05, + "loss": 2.4339, + "step": 32861 + }, + { + "epoch": 1.4, + "grad_norm": 0.515625, + "learning_rate": 4.133984248197991e-05, + "loss": 2.5108, + "step": 32862 + }, + { + "epoch": 1.4, + "grad_norm": 0.455078125, + "learning_rate": 4.133441002345173e-05, + "loss": 2.4233, + "step": 32863 + }, + { + "epoch": 1.4, + "grad_norm": 0.44140625, + "learning_rate": 4.132897782889726e-05, + "loss": 2.3825, + "step": 32864 + }, + { + "epoch": 1.4, + "grad_norm": 0.48046875, + "learning_rate": 4.132354589834092e-05, + "loss": 2.5617, + "step": 32865 + }, + { + "epoch": 1.4, + "grad_norm": 0.423828125, + "learning_rate": 4.131811423180718e-05, + "loss": 2.4359, + "step": 32866 + }, + { + "epoch": 1.4, + "grad_norm": 0.47265625, + "learning_rate": 4.131268282932051e-05, + "loss": 2.4373, + "step": 32867 + }, + { + "epoch": 1.4, + "grad_norm": 0.453125, + "learning_rate": 4.130725169090529e-05, + "loss": 2.5431, + "step": 32868 + }, + { + "epoch": 1.4, + "grad_norm": 0.44921875, + "learning_rate": 4.130182081658599e-05, + "loss": 2.4421, + "step": 32869 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.129639020638708e-05, + "loss": 2.5022, + "step": 32870 + }, + { + "epoch": 1.4, + "grad_norm": 0.51953125, + "learning_rate": 4.129095986033295e-05, + "loss": 2.4529, + "step": 32871 + }, + { + "epoch": 1.4, + "grad_norm": 0.40625, + "learning_rate": 4.128552977844802e-05, + "loss": 2.3188, + "step": 32872 + }, + { + "epoch": 1.4, + "grad_norm": 0.48046875, + "learning_rate": 4.1280099960756745e-05, + "loss": 2.4618, + "step": 32873 + }, + { + "epoch": 1.4, + "grad_norm": 0.431640625, + "learning_rate": 4.12746704072836e-05, + "loss": 2.3423, + "step": 32874 + }, + { + "epoch": 1.4, + "grad_norm": 0.50390625, + "learning_rate": 4.126924111805294e-05, + "loss": 2.4049, + "step": 32875 + }, + { + "epoch": 1.4, + "grad_norm": 0.4375, + "learning_rate": 4.126381209308926e-05, + "loss": 2.2537, + "step": 32876 + }, + { + "epoch": 1.4, + "grad_norm": 0.4296875, + "learning_rate": 4.1258383332416926e-05, + "loss": 2.5103, + "step": 32877 + }, + { + "epoch": 1.4, + "grad_norm": 0.482421875, + "learning_rate": 4.1252954836060395e-05, + "loss": 2.3683, + "step": 32878 + }, + { + "epoch": 1.4, + "grad_norm": 0.4765625, + "learning_rate": 4.124752660404414e-05, + "loss": 2.453, + "step": 32879 + }, + { + "epoch": 1.4, + "grad_norm": 0.498046875, + "learning_rate": 4.124209863639255e-05, + "loss": 2.4418, + "step": 32880 + }, + { + "epoch": 1.4, + "grad_norm": 0.478515625, + "learning_rate": 4.123667093312999e-05, + "loss": 2.4281, + "step": 32881 + }, + { + "epoch": 1.4, + "grad_norm": 0.5234375, + "learning_rate": 4.123124349428096e-05, + "loss": 2.3776, + "step": 32882 + }, + { + "epoch": 1.4, + "grad_norm": 0.4921875, + "learning_rate": 4.122581631986987e-05, + "loss": 2.5062, + "step": 32883 + }, + { + "epoch": 1.4, + "grad_norm": 0.388671875, + "learning_rate": 4.12203894099211e-05, + "loss": 2.2567, + "step": 32884 + }, + { + "epoch": 1.4, + "grad_norm": 0.51171875, + "learning_rate": 4.12149627644591e-05, + "loss": 2.4323, + "step": 32885 + }, + { + "epoch": 1.4, + "grad_norm": 0.52734375, + "learning_rate": 4.1209536383508306e-05, + "loss": 2.4993, + "step": 32886 + }, + { + "epoch": 1.4, + "grad_norm": 0.4921875, + "learning_rate": 4.120411026709313e-05, + "loss": 2.5841, + "step": 32887 + }, + { + "epoch": 1.4, + "grad_norm": 0.486328125, + "learning_rate": 4.119868441523792e-05, + "loss": 2.4866, + "step": 32888 + }, + { + "epoch": 1.4, + "grad_norm": 0.52734375, + "learning_rate": 4.119325882796715e-05, + "loss": 2.543, + "step": 32889 + }, + { + "epoch": 1.4, + "grad_norm": 0.5078125, + "learning_rate": 4.118783350530526e-05, + "loss": 2.4908, + "step": 32890 + }, + { + "epoch": 1.4, + "grad_norm": 0.52734375, + "learning_rate": 4.118240844727659e-05, + "loss": 2.5001, + "step": 32891 + }, + { + "epoch": 1.4, + "grad_norm": 0.482421875, + "learning_rate": 4.117698365390562e-05, + "loss": 2.4928, + "step": 32892 + }, + { + "epoch": 1.4, + "grad_norm": 0.40234375, + "learning_rate": 4.117155912521668e-05, + "loss": 2.4837, + "step": 32893 + }, + { + "epoch": 1.4, + "grad_norm": 0.478515625, + "learning_rate": 4.116613486123424e-05, + "loss": 2.3917, + "step": 32894 + }, + { + "epoch": 1.4, + "grad_norm": 0.435546875, + "learning_rate": 4.116071086198272e-05, + "loss": 2.4142, + "step": 32895 + }, + { + "epoch": 1.4, + "grad_norm": 0.5234375, + "learning_rate": 4.115528712748651e-05, + "loss": 2.4218, + "step": 32896 + }, + { + "epoch": 1.4, + "grad_norm": 0.5, + "learning_rate": 4.114986365776995e-05, + "loss": 2.4856, + "step": 32897 + }, + { + "epoch": 1.4, + "grad_norm": 0.453125, + "learning_rate": 4.11444404528575e-05, + "loss": 2.4623, + "step": 32898 + }, + { + "epoch": 1.4, + "grad_norm": 0.484375, + "learning_rate": 4.11390175127736e-05, + "loss": 2.5137, + "step": 32899 + }, + { + "epoch": 1.4, + "grad_norm": 0.49609375, + "learning_rate": 4.113359483754258e-05, + "loss": 2.3842, + "step": 32900 + }, + { + "epoch": 1.4, + "grad_norm": 0.466796875, + "learning_rate": 4.112817242718885e-05, + "loss": 2.3471, + "step": 32901 + }, + { + "epoch": 1.4, + "grad_norm": 0.44921875, + "learning_rate": 4.1122750281736875e-05, + "loss": 2.4933, + "step": 32902 + }, + { + "epoch": 1.4, + "grad_norm": 0.50390625, + "learning_rate": 4.111732840121101e-05, + "loss": 2.6151, + "step": 32903 + }, + { + "epoch": 1.4, + "grad_norm": 0.412109375, + "learning_rate": 4.111190678563561e-05, + "loss": 2.3099, + "step": 32904 + }, + { + "epoch": 1.4, + "grad_norm": 0.47265625, + "learning_rate": 4.1106485435035115e-05, + "loss": 2.3856, + "step": 32905 + }, + { + "epoch": 1.4, + "grad_norm": 0.48046875, + "learning_rate": 4.1101064349433935e-05, + "loss": 2.5255, + "step": 32906 + }, + { + "epoch": 1.4, + "grad_norm": 0.43359375, + "learning_rate": 4.109564352885642e-05, + "loss": 2.4099, + "step": 32907 + }, + { + "epoch": 1.4, + "grad_norm": 0.466796875, + "learning_rate": 4.1090222973327005e-05, + "loss": 2.5709, + "step": 32908 + }, + { + "epoch": 1.4, + "grad_norm": 0.54296875, + "learning_rate": 4.108480268287003e-05, + "loss": 2.3031, + "step": 32909 + }, + { + "epoch": 1.4, + "grad_norm": 0.453125, + "learning_rate": 4.1079382657509904e-05, + "loss": 2.4237, + "step": 32910 + }, + { + "epoch": 1.4, + "grad_norm": 0.431640625, + "learning_rate": 4.107396289727106e-05, + "loss": 2.4864, + "step": 32911 + }, + { + "epoch": 1.4, + "grad_norm": 0.46484375, + "learning_rate": 4.106854340217785e-05, + "loss": 2.4522, + "step": 32912 + }, + { + "epoch": 1.4, + "grad_norm": 0.4609375, + "learning_rate": 4.106312417225463e-05, + "loss": 2.3341, + "step": 32913 + }, + { + "epoch": 1.4, + "grad_norm": 0.494140625, + "learning_rate": 4.10577052075258e-05, + "loss": 2.4586, + "step": 32914 + }, + { + "epoch": 1.4, + "grad_norm": 0.5390625, + "learning_rate": 4.105228650801579e-05, + "loss": 2.6462, + "step": 32915 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.104686807374892e-05, + "loss": 2.4682, + "step": 32916 + }, + { + "epoch": 1.4, + "grad_norm": 0.466796875, + "learning_rate": 4.104144990474959e-05, + "loss": 2.5292, + "step": 32917 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.103603200104225e-05, + "loss": 2.4219, + "step": 32918 + }, + { + "epoch": 1.4, + "grad_norm": 0.462890625, + "learning_rate": 4.1030614362651185e-05, + "loss": 2.4278, + "step": 32919 + }, + { + "epoch": 1.4, + "grad_norm": 0.486328125, + "learning_rate": 4.102519698960079e-05, + "loss": 2.4812, + "step": 32920 + }, + { + "epoch": 1.4, + "grad_norm": 0.51953125, + "learning_rate": 4.1019779881915444e-05, + "loss": 2.5301, + "step": 32921 + }, + { + "epoch": 1.4, + "grad_norm": 0.43359375, + "learning_rate": 4.101436303961958e-05, + "loss": 2.4481, + "step": 32922 + }, + { + "epoch": 1.4, + "grad_norm": 0.458984375, + "learning_rate": 4.100894646273747e-05, + "loss": 2.4635, + "step": 32923 + }, + { + "epoch": 1.4, + "grad_norm": 0.435546875, + "learning_rate": 4.10035301512936e-05, + "loss": 2.5676, + "step": 32924 + }, + { + "epoch": 1.4, + "grad_norm": 0.423828125, + "learning_rate": 4.099811410531224e-05, + "loss": 2.3721, + "step": 32925 + }, + { + "epoch": 1.4, + "grad_norm": 0.5078125, + "learning_rate": 4.09926983248178e-05, + "loss": 2.4671, + "step": 32926 + }, + { + "epoch": 1.4, + "grad_norm": 0.453125, + "learning_rate": 4.098728280983469e-05, + "loss": 2.5117, + "step": 32927 + }, + { + "epoch": 1.4, + "grad_norm": 0.4921875, + "learning_rate": 4.0981867560387245e-05, + "loss": 2.3478, + "step": 32928 + }, + { + "epoch": 1.4, + "grad_norm": 0.53515625, + "learning_rate": 4.097645257649979e-05, + "loss": 2.5628, + "step": 32929 + }, + { + "epoch": 1.4, + "grad_norm": 0.46875, + "learning_rate": 4.0971037858196725e-05, + "loss": 2.3039, + "step": 32930 + }, + { + "epoch": 1.4, + "grad_norm": 0.412109375, + "learning_rate": 4.096562340550245e-05, + "loss": 2.4219, + "step": 32931 + }, + { + "epoch": 1.4, + "grad_norm": 0.5, + "learning_rate": 4.096020921844127e-05, + "loss": 2.3999, + "step": 32932 + }, + { + "epoch": 1.4, + "grad_norm": 0.47265625, + "learning_rate": 4.0954795297037564e-05, + "loss": 2.4613, + "step": 32933 + }, + { + "epoch": 1.4, + "grad_norm": 0.43359375, + "learning_rate": 4.094938164131573e-05, + "loss": 2.5333, + "step": 32934 + }, + { + "epoch": 1.4, + "grad_norm": 0.439453125, + "learning_rate": 4.094396825130011e-05, + "loss": 2.5188, + "step": 32935 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.093855512701501e-05, + "loss": 2.4132, + "step": 32936 + }, + { + "epoch": 1.4, + "grad_norm": 0.4453125, + "learning_rate": 4.0933142268484824e-05, + "loss": 2.4886, + "step": 32937 + }, + { + "epoch": 1.4, + "grad_norm": 0.4765625, + "learning_rate": 4.092772967573395e-05, + "loss": 2.4846, + "step": 32938 + }, + { + "epoch": 1.4, + "grad_norm": 0.4140625, + "learning_rate": 4.092231734878668e-05, + "loss": 2.3448, + "step": 32939 + }, + { + "epoch": 1.4, + "grad_norm": 0.431640625, + "learning_rate": 4.091690528766741e-05, + "loss": 2.4285, + "step": 32940 + }, + { + "epoch": 1.4, + "grad_norm": 0.4375, + "learning_rate": 4.091149349240044e-05, + "loss": 2.4033, + "step": 32941 + }, + { + "epoch": 1.4, + "grad_norm": 0.478515625, + "learning_rate": 4.090608196301017e-05, + "loss": 2.3416, + "step": 32942 + }, + { + "epoch": 1.4, + "grad_norm": 0.482421875, + "learning_rate": 4.090067069952096e-05, + "loss": 2.3971, + "step": 32943 + }, + { + "epoch": 1.4, + "grad_norm": 0.5078125, + "learning_rate": 4.089525970195713e-05, + "loss": 2.5206, + "step": 32944 + }, + { + "epoch": 1.4, + "grad_norm": 0.412109375, + "learning_rate": 4.0889848970342995e-05, + "loss": 2.4399, + "step": 32945 + }, + { + "epoch": 1.4, + "grad_norm": 0.482421875, + "learning_rate": 4.088443850470294e-05, + "loss": 2.3612, + "step": 32946 + }, + { + "epoch": 1.4, + "grad_norm": 0.4296875, + "learning_rate": 4.087902830506135e-05, + "loss": 2.3959, + "step": 32947 + }, + { + "epoch": 1.4, + "grad_norm": 0.51953125, + "learning_rate": 4.087361837144248e-05, + "loss": 2.4425, + "step": 32948 + }, + { + "epoch": 1.4, + "grad_norm": 0.51953125, + "learning_rate": 4.086820870387072e-05, + "loss": 2.4831, + "step": 32949 + }, + { + "epoch": 1.4, + "grad_norm": 0.498046875, + "learning_rate": 4.0862799302370455e-05, + "loss": 2.4887, + "step": 32950 + }, + { + "epoch": 1.4, + "grad_norm": 0.486328125, + "learning_rate": 4.085739016696598e-05, + "loss": 2.478, + "step": 32951 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.0851981297681594e-05, + "loss": 2.5163, + "step": 32952 + }, + { + "epoch": 1.4, + "grad_norm": 0.4375, + "learning_rate": 4.084657269454167e-05, + "loss": 2.5555, + "step": 32953 + }, + { + "epoch": 1.4, + "grad_norm": 0.5, + "learning_rate": 4.084116435757061e-05, + "loss": 2.5408, + "step": 32954 + }, + { + "epoch": 1.4, + "grad_norm": 0.466796875, + "learning_rate": 4.083575628679262e-05, + "loss": 2.3305, + "step": 32955 + }, + { + "epoch": 1.4, + "grad_norm": 0.443359375, + "learning_rate": 4.083034848223217e-05, + "loss": 2.3594, + "step": 32956 + }, + { + "epoch": 1.4, + "grad_norm": 0.47265625, + "learning_rate": 4.082494094391348e-05, + "loss": 2.4626, + "step": 32957 + }, + { + "epoch": 1.4, + "grad_norm": 0.40234375, + "learning_rate": 4.081953367186093e-05, + "loss": 2.3632, + "step": 32958 + }, + { + "epoch": 1.4, + "grad_norm": 0.48046875, + "learning_rate": 4.081412666609889e-05, + "loss": 2.4911, + "step": 32959 + }, + { + "epoch": 1.4, + "grad_norm": 0.44921875, + "learning_rate": 4.080871992665165e-05, + "loss": 2.4299, + "step": 32960 + }, + { + "epoch": 1.4, + "grad_norm": 0.451171875, + "learning_rate": 4.08033134535435e-05, + "loss": 2.4929, + "step": 32961 + }, + { + "epoch": 1.4, + "grad_norm": 0.47265625, + "learning_rate": 4.079790724679881e-05, + "loss": 2.2841, + "step": 32962 + }, + { + "epoch": 1.4, + "grad_norm": 0.48828125, + "learning_rate": 4.079250130644192e-05, + "loss": 2.4791, + "step": 32963 + }, + { + "epoch": 1.4, + "grad_norm": 0.5390625, + "learning_rate": 4.078709563249712e-05, + "loss": 2.5144, + "step": 32964 + }, + { + "epoch": 1.4, + "grad_norm": 0.41796875, + "learning_rate": 4.0781690224988745e-05, + "loss": 2.5349, + "step": 32965 + }, + { + "epoch": 1.4, + "grad_norm": 0.470703125, + "learning_rate": 4.077628508394116e-05, + "loss": 2.4071, + "step": 32966 + }, + { + "epoch": 1.4, + "grad_norm": 0.453125, + "learning_rate": 4.0770880209378635e-05, + "loss": 2.5653, + "step": 32967 + }, + { + "epoch": 1.4, + "grad_norm": 0.451171875, + "learning_rate": 4.0765475601325474e-05, + "loss": 2.4301, + "step": 32968 + }, + { + "epoch": 1.4, + "grad_norm": 0.53125, + "learning_rate": 4.076007125980603e-05, + "loss": 2.5598, + "step": 32969 + }, + { + "epoch": 1.4, + "grad_norm": 0.48828125, + "learning_rate": 4.0754667184844654e-05, + "loss": 2.4897, + "step": 32970 + }, + { + "epoch": 1.4, + "grad_norm": 0.412109375, + "learning_rate": 4.074926337646559e-05, + "loss": 2.402, + "step": 32971 + }, + { + "epoch": 1.41, + "grad_norm": 0.515625, + "learning_rate": 4.074385983469321e-05, + "loss": 2.5155, + "step": 32972 + }, + { + "epoch": 1.41, + "grad_norm": 0.439453125, + "learning_rate": 4.073845655955178e-05, + "loss": 2.5309, + "step": 32973 + }, + { + "epoch": 1.41, + "grad_norm": 0.486328125, + "learning_rate": 4.0733053551065625e-05, + "loss": 2.4257, + "step": 32974 + }, + { + "epoch": 1.41, + "grad_norm": 0.40625, + "learning_rate": 4.072765080925911e-05, + "loss": 2.4339, + "step": 32975 + }, + { + "epoch": 1.41, + "grad_norm": 0.431640625, + "learning_rate": 4.072224833415651e-05, + "loss": 2.3932, + "step": 32976 + }, + { + "epoch": 1.41, + "grad_norm": 0.46875, + "learning_rate": 4.071684612578209e-05, + "loss": 2.4978, + "step": 32977 + }, + { + "epoch": 1.41, + "grad_norm": 0.490234375, + "learning_rate": 4.071144418416021e-05, + "loss": 2.5465, + "step": 32978 + }, + { + "epoch": 1.41, + "grad_norm": 0.4765625, + "learning_rate": 4.070604250931518e-05, + "loss": 2.6022, + "step": 32979 + }, + { + "epoch": 1.41, + "grad_norm": 0.4453125, + "learning_rate": 4.0700641101271266e-05, + "loss": 2.551, + "step": 32980 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 4.069523996005282e-05, + "loss": 2.3754, + "step": 32981 + }, + { + "epoch": 1.41, + "grad_norm": 0.453125, + "learning_rate": 4.06898390856841e-05, + "loss": 2.3372, + "step": 32982 + }, + { + "epoch": 1.41, + "grad_norm": 0.5, + "learning_rate": 4.068443847818946e-05, + "loss": 2.5146, + "step": 32983 + }, + { + "epoch": 1.41, + "grad_norm": 0.40625, + "learning_rate": 4.067903813759313e-05, + "loss": 2.4614, + "step": 32984 + }, + { + "epoch": 1.41, + "grad_norm": 0.462890625, + "learning_rate": 4.0673638063919485e-05, + "loss": 2.3592, + "step": 32985 + }, + { + "epoch": 1.41, + "grad_norm": 0.4453125, + "learning_rate": 4.066823825719275e-05, + "loss": 2.4003, + "step": 32986 + }, + { + "epoch": 1.41, + "grad_norm": 0.50390625, + "learning_rate": 4.0662838717437256e-05, + "loss": 2.3919, + "step": 32987 + }, + { + "epoch": 1.41, + "grad_norm": 0.423828125, + "learning_rate": 4.065743944467735e-05, + "loss": 2.4304, + "step": 32988 + }, + { + "epoch": 1.41, + "grad_norm": 0.52734375, + "learning_rate": 4.065204043893727e-05, + "loss": 2.492, + "step": 32989 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 4.0646641700241253e-05, + "loss": 2.3867, + "step": 32990 + }, + { + "epoch": 1.41, + "grad_norm": 0.5234375, + "learning_rate": 4.064124322861372e-05, + "loss": 2.5488, + "step": 32991 + }, + { + "epoch": 1.41, + "grad_norm": 0.55859375, + "learning_rate": 4.06358450240789e-05, + "loss": 2.5555, + "step": 32992 + }, + { + "epoch": 1.41, + "grad_norm": 0.423828125, + "learning_rate": 4.0630447086661064e-05, + "loss": 2.3414, + "step": 32993 + }, + { + "epoch": 1.41, + "grad_norm": 0.44140625, + "learning_rate": 4.0625049416384496e-05, + "loss": 2.4646, + "step": 32994 + }, + { + "epoch": 1.41, + "grad_norm": 0.50390625, + "learning_rate": 4.061965201327356e-05, + "loss": 2.4862, + "step": 32995 + }, + { + "epoch": 1.41, + "grad_norm": 0.50390625, + "learning_rate": 4.061425487735244e-05, + "loss": 2.4273, + "step": 32996 + }, + { + "epoch": 1.41, + "grad_norm": 0.45703125, + "learning_rate": 4.060885800864551e-05, + "loss": 2.4278, + "step": 32997 + }, + { + "epoch": 1.41, + "grad_norm": 0.515625, + "learning_rate": 4.0603461407176976e-05, + "loss": 2.4731, + "step": 32998 + }, + { + "epoch": 1.41, + "grad_norm": 0.47265625, + "learning_rate": 4.05980650729712e-05, + "loss": 2.3508, + "step": 32999 + }, + { + "epoch": 1.41, + "grad_norm": 0.455078125, + "learning_rate": 4.0592669006052386e-05, + "loss": 2.4161, + "step": 33000 + }, + { + "epoch": 1.41, + "grad_norm": 0.45703125, + "learning_rate": 4.0587273206444883e-05, + "loss": 2.4247, + "step": 33001 + }, + { + "epoch": 1.41, + "grad_norm": 0.439453125, + "learning_rate": 4.0581877674172905e-05, + "loss": 2.4753, + "step": 33002 + }, + { + "epoch": 1.41, + "grad_norm": 0.455078125, + "learning_rate": 4.057648240926077e-05, + "loss": 2.4826, + "step": 33003 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 4.0571087411732775e-05, + "loss": 2.5054, + "step": 33004 + }, + { + "epoch": 1.41, + "grad_norm": 0.453125, + "learning_rate": 4.056569268161317e-05, + "loss": 2.4101, + "step": 33005 + }, + { + "epoch": 1.41, + "grad_norm": 0.490234375, + "learning_rate": 4.0560298218926155e-05, + "loss": 2.4849, + "step": 33006 + }, + { + "epoch": 1.41, + "grad_norm": 0.447265625, + "learning_rate": 4.055490402369614e-05, + "loss": 2.4341, + "step": 33007 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 4.054951009594734e-05, + "loss": 2.511, + "step": 33008 + }, + { + "epoch": 1.41, + "grad_norm": 0.515625, + "learning_rate": 4.0544116435703984e-05, + "loss": 2.3548, + "step": 33009 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 4.0538723042990366e-05, + "loss": 2.5557, + "step": 33010 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 4.053332991783081e-05, + "loss": 2.4752, + "step": 33011 + }, + { + "epoch": 1.41, + "grad_norm": 0.466796875, + "learning_rate": 4.052793706024949e-05, + "loss": 2.5339, + "step": 33012 + }, + { + "epoch": 1.41, + "grad_norm": 0.482421875, + "learning_rate": 4.0522544470270765e-05, + "loss": 2.5314, + "step": 33013 + }, + { + "epoch": 1.41, + "grad_norm": 0.4453125, + "learning_rate": 4.0517152147918815e-05, + "loss": 2.4483, + "step": 33014 + }, + { + "epoch": 1.41, + "grad_norm": 0.451171875, + "learning_rate": 4.0511760093217966e-05, + "loss": 2.4624, + "step": 33015 + }, + { + "epoch": 1.41, + "grad_norm": 0.486328125, + "learning_rate": 4.050636830619243e-05, + "loss": 2.425, + "step": 33016 + }, + { + "epoch": 1.41, + "grad_norm": 0.55859375, + "learning_rate": 4.050097678686653e-05, + "loss": 2.4944, + "step": 33017 + }, + { + "epoch": 1.41, + "grad_norm": 0.478515625, + "learning_rate": 4.049558553526446e-05, + "loss": 2.5007, + "step": 33018 + }, + { + "epoch": 1.41, + "grad_norm": 0.451171875, + "learning_rate": 4.04901945514105e-05, + "loss": 2.2588, + "step": 33019 + }, + { + "epoch": 1.41, + "grad_norm": 0.498046875, + "learning_rate": 4.0484803835328955e-05, + "loss": 2.447, + "step": 33020 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 4.047941338704404e-05, + "loss": 2.4286, + "step": 33021 + }, + { + "epoch": 1.41, + "grad_norm": 0.453125, + "learning_rate": 4.047402320657995e-05, + "loss": 2.4147, + "step": 33022 + }, + { + "epoch": 1.41, + "grad_norm": 0.47265625, + "learning_rate": 4.0468633293961054e-05, + "loss": 2.4266, + "step": 33023 + }, + { + "epoch": 1.41, + "grad_norm": 0.447265625, + "learning_rate": 4.046324364921157e-05, + "loss": 2.4586, + "step": 33024 + }, + { + "epoch": 1.41, + "grad_norm": 0.51953125, + "learning_rate": 4.045785427235569e-05, + "loss": 2.4855, + "step": 33025 + }, + { + "epoch": 1.41, + "grad_norm": 0.4296875, + "learning_rate": 4.0452465163417704e-05, + "loss": 2.4435, + "step": 33026 + }, + { + "epoch": 1.41, + "grad_norm": 0.484375, + "learning_rate": 4.0447076322421894e-05, + "loss": 2.454, + "step": 33027 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 4.044168774939245e-05, + "loss": 2.4244, + "step": 33028 + }, + { + "epoch": 1.41, + "grad_norm": 0.46875, + "learning_rate": 4.0436299444353676e-05, + "loss": 2.5343, + "step": 33029 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 4.043091140732975e-05, + "loss": 2.6135, + "step": 33030 + }, + { + "epoch": 1.41, + "grad_norm": 0.4765625, + "learning_rate": 4.0425523638344985e-05, + "loss": 2.4831, + "step": 33031 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 4.042013613742357e-05, + "loss": 2.3681, + "step": 33032 + }, + { + "epoch": 1.41, + "grad_norm": 0.47265625, + "learning_rate": 4.0414748904589794e-05, + "loss": 2.4364, + "step": 33033 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 4.040936193986784e-05, + "loss": 2.5702, + "step": 33034 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 4.040397524328198e-05, + "loss": 2.38, + "step": 33035 + }, + { + "epoch": 1.41, + "grad_norm": 0.5390625, + "learning_rate": 4.039858881485647e-05, + "loss": 2.5751, + "step": 33036 + }, + { + "epoch": 1.41, + "grad_norm": 0.427734375, + "learning_rate": 4.0393202654615546e-05, + "loss": 2.4777, + "step": 33037 + }, + { + "epoch": 1.41, + "grad_norm": 0.42578125, + "learning_rate": 4.038781676258335e-05, + "loss": 2.3921, + "step": 33038 + }, + { + "epoch": 1.41, + "grad_norm": 0.37109375, + "learning_rate": 4.0382431138784274e-05, + "loss": 2.3323, + "step": 33039 + }, + { + "epoch": 1.41, + "grad_norm": 0.447265625, + "learning_rate": 4.037704578324246e-05, + "loss": 2.4492, + "step": 33040 + }, + { + "epoch": 1.41, + "grad_norm": 0.48046875, + "learning_rate": 4.037166069598212e-05, + "loss": 2.5222, + "step": 33041 + }, + { + "epoch": 1.41, + "grad_norm": 0.5546875, + "learning_rate": 4.036627587702751e-05, + "loss": 2.6273, + "step": 33042 + }, + { + "epoch": 1.41, + "grad_norm": 0.435546875, + "learning_rate": 4.0360891326402896e-05, + "loss": 2.4144, + "step": 33043 + }, + { + "epoch": 1.41, + "grad_norm": 0.52734375, + "learning_rate": 4.0355507044132445e-05, + "loss": 2.4322, + "step": 33044 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 4.0350123030240446e-05, + "loss": 2.5003, + "step": 33045 + }, + { + "epoch": 1.41, + "grad_norm": 0.453125, + "learning_rate": 4.0344739284751064e-05, + "loss": 2.4155, + "step": 33046 + }, + { + "epoch": 1.41, + "grad_norm": 0.5, + "learning_rate": 4.03393558076886e-05, + "loss": 2.4831, + "step": 33047 + }, + { + "epoch": 1.41, + "grad_norm": 0.46875, + "learning_rate": 4.033397259907717e-05, + "loss": 2.3964, + "step": 33048 + }, + { + "epoch": 1.41, + "grad_norm": 0.451171875, + "learning_rate": 4.032858965894112e-05, + "loss": 2.3901, + "step": 33049 + }, + { + "epoch": 1.41, + "grad_norm": 0.443359375, + "learning_rate": 4.032320698730456e-05, + "loss": 2.5285, + "step": 33050 + }, + { + "epoch": 1.41, + "grad_norm": 0.466796875, + "learning_rate": 4.0317824584191754e-05, + "loss": 2.3887, + "step": 33051 + }, + { + "epoch": 1.41, + "grad_norm": 0.48828125, + "learning_rate": 4.0312442449626974e-05, + "loss": 2.5403, + "step": 33052 + }, + { + "epoch": 1.41, + "grad_norm": 0.41796875, + "learning_rate": 4.030706058363438e-05, + "loss": 2.3781, + "step": 33053 + }, + { + "epoch": 1.41, + "grad_norm": 0.482421875, + "learning_rate": 4.030167898623814e-05, + "loss": 2.401, + "step": 33054 + }, + { + "epoch": 1.41, + "grad_norm": 0.412109375, + "learning_rate": 4.029629765746259e-05, + "loss": 2.3482, + "step": 33055 + }, + { + "epoch": 1.41, + "grad_norm": 0.55859375, + "learning_rate": 4.029091659733187e-05, + "loss": 2.5558, + "step": 33056 + }, + { + "epoch": 1.41, + "grad_norm": 0.462890625, + "learning_rate": 4.028553580587019e-05, + "loss": 2.482, + "step": 33057 + }, + { + "epoch": 1.41, + "grad_norm": 0.494140625, + "learning_rate": 4.028015528310176e-05, + "loss": 2.5987, + "step": 33058 + }, + { + "epoch": 1.41, + "grad_norm": 0.498046875, + "learning_rate": 4.0274775029050846e-05, + "loss": 2.4723, + "step": 33059 + }, + { + "epoch": 1.41, + "grad_norm": 0.421875, + "learning_rate": 4.0269395043741585e-05, + "loss": 2.3961, + "step": 33060 + }, + { + "epoch": 1.41, + "grad_norm": 0.50390625, + "learning_rate": 4.0264015327198243e-05, + "loss": 2.4484, + "step": 33061 + }, + { + "epoch": 1.41, + "grad_norm": 0.455078125, + "learning_rate": 4.025863587944497e-05, + "loss": 2.4629, + "step": 33062 + }, + { + "epoch": 1.41, + "grad_norm": 0.52734375, + "learning_rate": 4.0253256700506015e-05, + "loss": 2.4438, + "step": 33063 + }, + { + "epoch": 1.41, + "grad_norm": 0.42578125, + "learning_rate": 4.02478777904056e-05, + "loss": 2.3945, + "step": 33064 + }, + { + "epoch": 1.41, + "grad_norm": 0.453125, + "learning_rate": 4.0242499149167886e-05, + "loss": 2.4116, + "step": 33065 + }, + { + "epoch": 1.41, + "grad_norm": 0.423828125, + "learning_rate": 4.023712077681706e-05, + "loss": 2.335, + "step": 33066 + }, + { + "epoch": 1.41, + "grad_norm": 0.4296875, + "learning_rate": 4.023174267337735e-05, + "loss": 2.3647, + "step": 33067 + }, + { + "epoch": 1.41, + "grad_norm": 0.48046875, + "learning_rate": 4.022636483887299e-05, + "loss": 2.3102, + "step": 33068 + }, + { + "epoch": 1.41, + "grad_norm": 0.435546875, + "learning_rate": 4.022098727332814e-05, + "loss": 2.4494, + "step": 33069 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 4.021560997676693e-05, + "loss": 2.3964, + "step": 33070 + }, + { + "epoch": 1.41, + "grad_norm": 0.482421875, + "learning_rate": 4.02102329492137e-05, + "loss": 2.4822, + "step": 33071 + }, + { + "epoch": 1.41, + "grad_norm": 0.52734375, + "learning_rate": 4.020485619069257e-05, + "loss": 2.6887, + "step": 33072 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 4.019947970122771e-05, + "loss": 2.3739, + "step": 33073 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 4.019410348084333e-05, + "loss": 2.4416, + "step": 33074 + }, + { + "epoch": 1.41, + "grad_norm": 0.423828125, + "learning_rate": 4.0188727529563654e-05, + "loss": 2.4901, + "step": 33075 + }, + { + "epoch": 1.41, + "grad_norm": 0.51171875, + "learning_rate": 4.018335184741282e-05, + "loss": 2.6037, + "step": 33076 + }, + { + "epoch": 1.41, + "grad_norm": 0.451171875, + "learning_rate": 4.0177976434415065e-05, + "loss": 2.5617, + "step": 33077 + }, + { + "epoch": 1.41, + "grad_norm": 0.470703125, + "learning_rate": 4.017260129059454e-05, + "loss": 2.345, + "step": 33078 + }, + { + "epoch": 1.41, + "grad_norm": 0.45703125, + "learning_rate": 4.016722641597542e-05, + "loss": 2.4649, + "step": 33079 + }, + { + "epoch": 1.41, + "grad_norm": 0.4609375, + "learning_rate": 4.016185181058196e-05, + "loss": 2.4537, + "step": 33080 + }, + { + "epoch": 1.41, + "grad_norm": 0.4765625, + "learning_rate": 4.01564774744383e-05, + "loss": 2.3794, + "step": 33081 + }, + { + "epoch": 1.41, + "grad_norm": 0.470703125, + "learning_rate": 4.015110340756859e-05, + "loss": 2.4565, + "step": 33082 + }, + { + "epoch": 1.41, + "grad_norm": 0.439453125, + "learning_rate": 4.014572960999704e-05, + "loss": 2.3764, + "step": 33083 + }, + { + "epoch": 1.41, + "grad_norm": 0.482421875, + "learning_rate": 4.014035608174787e-05, + "loss": 2.4711, + "step": 33084 + }, + { + "epoch": 1.41, + "grad_norm": 0.484375, + "learning_rate": 4.0134982822845215e-05, + "loss": 2.5061, + "step": 33085 + }, + { + "epoch": 1.41, + "grad_norm": 0.4765625, + "learning_rate": 4.012960983331319e-05, + "loss": 2.3661, + "step": 33086 + }, + { + "epoch": 1.41, + "grad_norm": 0.427734375, + "learning_rate": 4.012423711317611e-05, + "loss": 2.4837, + "step": 33087 + }, + { + "epoch": 1.41, + "grad_norm": 0.466796875, + "learning_rate": 4.0118864662458086e-05, + "loss": 2.6261, + "step": 33088 + }, + { + "epoch": 1.41, + "grad_norm": 0.451171875, + "learning_rate": 4.011349248118325e-05, + "loss": 2.4338, + "step": 33089 + }, + { + "epoch": 1.41, + "grad_norm": 0.482421875, + "learning_rate": 4.010812056937581e-05, + "loss": 2.3495, + "step": 33090 + }, + { + "epoch": 1.41, + "grad_norm": 0.4296875, + "learning_rate": 4.010274892705997e-05, + "loss": 2.4837, + "step": 33091 + }, + { + "epoch": 1.41, + "grad_norm": 0.4765625, + "learning_rate": 4.0097377554259855e-05, + "loss": 2.4458, + "step": 33092 + }, + { + "epoch": 1.41, + "grad_norm": 0.58984375, + "learning_rate": 4.009200645099966e-05, + "loss": 2.4332, + "step": 33093 + }, + { + "epoch": 1.41, + "grad_norm": 0.423828125, + "learning_rate": 4.0086635617303516e-05, + "loss": 2.4279, + "step": 33094 + }, + { + "epoch": 1.41, + "grad_norm": 0.455078125, + "learning_rate": 4.008126505319562e-05, + "loss": 2.516, + "step": 33095 + }, + { + "epoch": 1.41, + "grad_norm": 0.45703125, + "learning_rate": 4.007589475870016e-05, + "loss": 2.3962, + "step": 33096 + }, + { + "epoch": 1.41, + "grad_norm": 0.45703125, + "learning_rate": 4.0070524733841286e-05, + "loss": 2.466, + "step": 33097 + }, + { + "epoch": 1.41, + "grad_norm": 0.44921875, + "learning_rate": 4.00651549786431e-05, + "loss": 2.4517, + "step": 33098 + }, + { + "epoch": 1.41, + "grad_norm": 0.451171875, + "learning_rate": 4.005978549312982e-05, + "loss": 2.4445, + "step": 33099 + }, + { + "epoch": 1.41, + "grad_norm": 0.50390625, + "learning_rate": 4.005441627732563e-05, + "loss": 2.4619, + "step": 33100 + }, + { + "epoch": 1.41, + "grad_norm": 0.46484375, + "learning_rate": 4.0049047331254674e-05, + "loss": 2.4589, + "step": 33101 + }, + { + "epoch": 1.41, + "grad_norm": 0.470703125, + "learning_rate": 4.004367865494101e-05, + "loss": 2.273, + "step": 33102 + }, + { + "epoch": 1.41, + "grad_norm": 0.4375, + "learning_rate": 4.003831024840896e-05, + "loss": 2.5302, + "step": 33103 + }, + { + "epoch": 1.41, + "grad_norm": 0.408203125, + "learning_rate": 4.003294211168259e-05, + "loss": 2.398, + "step": 33104 + }, + { + "epoch": 1.41, + "grad_norm": 0.515625, + "learning_rate": 4.0027574244786035e-05, + "loss": 2.4959, + "step": 33105 + }, + { + "epoch": 1.41, + "grad_norm": 0.427734375, + "learning_rate": 4.0022206647743524e-05, + "loss": 2.4455, + "step": 33106 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 4.0016839320579114e-05, + "loss": 2.338, + "step": 33107 + }, + { + "epoch": 1.41, + "grad_norm": 0.4765625, + "learning_rate": 4.001147226331701e-05, + "loss": 2.5286, + "step": 33108 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 4.000610547598139e-05, + "loss": 2.4346, + "step": 33109 + }, + { + "epoch": 1.41, + "grad_norm": 0.453125, + "learning_rate": 4.000073895859637e-05, + "loss": 2.4121, + "step": 33110 + }, + { + "epoch": 1.41, + "grad_norm": 0.42578125, + "learning_rate": 3.9995372711186044e-05, + "loss": 2.3378, + "step": 33111 + }, + { + "epoch": 1.41, + "grad_norm": 0.451171875, + "learning_rate": 3.9990006733774666e-05, + "loss": 2.4094, + "step": 33112 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 3.9984641026386325e-05, + "loss": 2.3135, + "step": 33113 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 3.997927558904514e-05, + "loss": 2.4322, + "step": 33114 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 3.997391042177527e-05, + "loss": 2.3622, + "step": 33115 + }, + { + "epoch": 1.41, + "grad_norm": 0.439453125, + "learning_rate": 3.996854552460091e-05, + "loss": 2.4131, + "step": 33116 + }, + { + "epoch": 1.41, + "grad_norm": 0.490234375, + "learning_rate": 3.996318089754611e-05, + "loss": 2.4494, + "step": 33117 + }, + { + "epoch": 1.41, + "grad_norm": 0.431640625, + "learning_rate": 3.9957816540635105e-05, + "loss": 2.3697, + "step": 33118 + }, + { + "epoch": 1.41, + "grad_norm": 0.462890625, + "learning_rate": 3.9952452453891945e-05, + "loss": 2.4643, + "step": 33119 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 3.994708863734082e-05, + "loss": 2.4265, + "step": 33120 + }, + { + "epoch": 1.41, + "grad_norm": 0.39453125, + "learning_rate": 3.9941725091005836e-05, + "loss": 2.3563, + "step": 33121 + }, + { + "epoch": 1.41, + "grad_norm": 0.51171875, + "learning_rate": 3.9936361814911174e-05, + "loss": 2.464, + "step": 33122 + }, + { + "epoch": 1.41, + "grad_norm": 0.515625, + "learning_rate": 3.9930998809080886e-05, + "loss": 2.4255, + "step": 33123 + }, + { + "epoch": 1.41, + "grad_norm": 0.56640625, + "learning_rate": 3.992563607353915e-05, + "loss": 2.4349, + "step": 33124 + }, + { + "epoch": 1.41, + "grad_norm": 0.49609375, + "learning_rate": 3.9920273608310146e-05, + "loss": 2.5133, + "step": 33125 + }, + { + "epoch": 1.41, + "grad_norm": 0.484375, + "learning_rate": 3.991491141341795e-05, + "loss": 2.4775, + "step": 33126 + }, + { + "epoch": 1.41, + "grad_norm": 0.421875, + "learning_rate": 3.990954948888661e-05, + "loss": 2.6176, + "step": 33127 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 3.990418783474043e-05, + "loss": 2.4133, + "step": 33128 + }, + { + "epoch": 1.41, + "grad_norm": 0.466796875, + "learning_rate": 3.989882645100342e-05, + "loss": 2.3879, + "step": 33129 + }, + { + "epoch": 1.41, + "grad_norm": 0.45703125, + "learning_rate": 3.989346533769971e-05, + "loss": 2.5122, + "step": 33130 + }, + { + "epoch": 1.41, + "grad_norm": 0.50390625, + "learning_rate": 3.988810449485343e-05, + "loss": 2.4107, + "step": 33131 + }, + { + "epoch": 1.41, + "grad_norm": 0.46484375, + "learning_rate": 3.9882743922488744e-05, + "loss": 2.4003, + "step": 33132 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 3.9877383620629706e-05, + "loss": 2.3661, + "step": 33133 + }, + { + "epoch": 1.41, + "grad_norm": 0.484375, + "learning_rate": 3.987202358930051e-05, + "loss": 2.3852, + "step": 33134 + }, + { + "epoch": 1.41, + "grad_norm": 0.427734375, + "learning_rate": 3.98666638285252e-05, + "loss": 2.3917, + "step": 33135 + }, + { + "epoch": 1.41, + "grad_norm": 0.41796875, + "learning_rate": 3.986130433832795e-05, + "loss": 2.47, + "step": 33136 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 3.985594511873282e-05, + "loss": 2.4998, + "step": 33137 + }, + { + "epoch": 1.41, + "grad_norm": 0.443359375, + "learning_rate": 3.9850586169764006e-05, + "loss": 2.3919, + "step": 33138 + }, + { + "epoch": 1.41, + "grad_norm": 0.439453125, + "learning_rate": 3.9845227491445524e-05, + "loss": 2.479, + "step": 33139 + }, + { + "epoch": 1.41, + "grad_norm": 0.44921875, + "learning_rate": 3.983986908380153e-05, + "loss": 2.4822, + "step": 33140 + }, + { + "epoch": 1.41, + "grad_norm": 0.474609375, + "learning_rate": 3.983451094685619e-05, + "loss": 2.4144, + "step": 33141 + }, + { + "epoch": 1.41, + "grad_norm": 0.515625, + "learning_rate": 3.982915308063355e-05, + "loss": 2.5689, + "step": 33142 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 3.9823795485157666e-05, + "loss": 2.4395, + "step": 33143 + }, + { + "epoch": 1.41, + "grad_norm": 0.462890625, + "learning_rate": 3.9818438160452776e-05, + "loss": 2.2538, + "step": 33144 + }, + { + "epoch": 1.41, + "grad_norm": 0.44921875, + "learning_rate": 3.981308110654293e-05, + "loss": 2.5323, + "step": 33145 + }, + { + "epoch": 1.41, + "grad_norm": 0.44921875, + "learning_rate": 3.9807724323452176e-05, + "loss": 2.4491, + "step": 33146 + }, + { + "epoch": 1.41, + "grad_norm": 0.48046875, + "learning_rate": 3.980236781120468e-05, + "loss": 2.3611, + "step": 33147 + }, + { + "epoch": 1.41, + "grad_norm": 0.486328125, + "learning_rate": 3.979701156982455e-05, + "loss": 2.4724, + "step": 33148 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 3.979165559933583e-05, + "loss": 2.3057, + "step": 33149 + }, + { + "epoch": 1.41, + "grad_norm": 0.49609375, + "learning_rate": 3.97862998997627e-05, + "loss": 2.4743, + "step": 33150 + }, + { + "epoch": 1.41, + "grad_norm": 0.53515625, + "learning_rate": 3.9780944471129176e-05, + "loss": 2.614, + "step": 33151 + }, + { + "epoch": 1.41, + "grad_norm": 0.490234375, + "learning_rate": 3.977558931345944e-05, + "loss": 2.4711, + "step": 33152 + }, + { + "epoch": 1.41, + "grad_norm": 0.453125, + "learning_rate": 3.97702344267775e-05, + "loss": 2.4483, + "step": 33153 + }, + { + "epoch": 1.41, + "grad_norm": 0.53125, + "learning_rate": 3.976487981110753e-05, + "loss": 2.565, + "step": 33154 + }, + { + "epoch": 1.41, + "grad_norm": 0.4140625, + "learning_rate": 3.975952546647353e-05, + "loss": 2.4163, + "step": 33155 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 3.9754171392899675e-05, + "loss": 2.3809, + "step": 33156 + }, + { + "epoch": 1.41, + "grad_norm": 0.462890625, + "learning_rate": 3.974881759041007e-05, + "loss": 2.5394, + "step": 33157 + }, + { + "epoch": 1.41, + "grad_norm": 0.4765625, + "learning_rate": 3.9743464059028755e-05, + "loss": 2.4962, + "step": 33158 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 3.973811079877976e-05, + "loss": 2.6482, + "step": 33159 + }, + { + "epoch": 1.41, + "grad_norm": 0.416015625, + "learning_rate": 3.9732757809687315e-05, + "loss": 2.2926, + "step": 33160 + }, + { + "epoch": 1.41, + "grad_norm": 0.48828125, + "learning_rate": 3.972740509177544e-05, + "loss": 2.4438, + "step": 33161 + }, + { + "epoch": 1.41, + "grad_norm": 0.447265625, + "learning_rate": 3.972205264506817e-05, + "loss": 2.3994, + "step": 33162 + }, + { + "epoch": 1.41, + "grad_norm": 0.447265625, + "learning_rate": 3.9716700469589644e-05, + "loss": 2.3275, + "step": 33163 + }, + { + "epoch": 1.41, + "grad_norm": 0.443359375, + "learning_rate": 3.971134856536396e-05, + "loss": 2.3898, + "step": 33164 + }, + { + "epoch": 1.41, + "grad_norm": 0.443359375, + "learning_rate": 3.970599693241515e-05, + "loss": 2.3355, + "step": 33165 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 3.9700645570767345e-05, + "loss": 2.4276, + "step": 33166 + }, + { + "epoch": 1.41, + "grad_norm": 0.453125, + "learning_rate": 3.9695294480444565e-05, + "loss": 2.4843, + "step": 33167 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 3.9689943661470965e-05, + "loss": 2.3983, + "step": 33168 + }, + { + "epoch": 1.41, + "grad_norm": 0.5234375, + "learning_rate": 3.968459311387054e-05, + "loss": 2.5381, + "step": 33169 + }, + { + "epoch": 1.41, + "grad_norm": 0.4453125, + "learning_rate": 3.967924283766743e-05, + "loss": 2.2842, + "step": 33170 + }, + { + "epoch": 1.41, + "grad_norm": 0.486328125, + "learning_rate": 3.967389283288565e-05, + "loss": 2.5846, + "step": 33171 + }, + { + "epoch": 1.41, + "grad_norm": 0.53125, + "learning_rate": 3.96685430995493e-05, + "loss": 2.4667, + "step": 33172 + }, + { + "epoch": 1.41, + "grad_norm": 0.439453125, + "learning_rate": 3.9663193637682504e-05, + "loss": 2.4249, + "step": 33173 + }, + { + "epoch": 1.41, + "grad_norm": 0.48046875, + "learning_rate": 3.965784444730929e-05, + "loss": 2.3479, + "step": 33174 + }, + { + "epoch": 1.41, + "grad_norm": 0.458984375, + "learning_rate": 3.965249552845364e-05, + "loss": 2.5114, + "step": 33175 + }, + { + "epoch": 1.41, + "grad_norm": 0.46875, + "learning_rate": 3.964714688113979e-05, + "loss": 2.5128, + "step": 33176 + }, + { + "epoch": 1.41, + "grad_norm": 0.455078125, + "learning_rate": 3.96417985053917e-05, + "loss": 2.4425, + "step": 33177 + }, + { + "epoch": 1.41, + "grad_norm": 0.46875, + "learning_rate": 3.963645040123344e-05, + "loss": 2.5736, + "step": 33178 + }, + { + "epoch": 1.41, + "grad_norm": 0.4453125, + "learning_rate": 3.963110256868909e-05, + "loss": 2.3775, + "step": 33179 + }, + { + "epoch": 1.41, + "grad_norm": 0.50390625, + "learning_rate": 3.962575500778275e-05, + "loss": 2.3019, + "step": 33180 + }, + { + "epoch": 1.41, + "grad_norm": 0.51953125, + "learning_rate": 3.962040771853841e-05, + "loss": 2.4463, + "step": 33181 + }, + { + "epoch": 1.41, + "grad_norm": 0.56640625, + "learning_rate": 3.961506070098021e-05, + "loss": 2.5878, + "step": 33182 + }, + { + "epoch": 1.41, + "grad_norm": 0.4375, + "learning_rate": 3.9609713955132136e-05, + "loss": 2.3907, + "step": 33183 + }, + { + "epoch": 1.41, + "grad_norm": 0.5078125, + "learning_rate": 3.9604367481018314e-05, + "loss": 2.4361, + "step": 33184 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 3.9599021278662715e-05, + "loss": 2.4377, + "step": 33185 + }, + { + "epoch": 1.41, + "grad_norm": 0.54296875, + "learning_rate": 3.9593675348089496e-05, + "loss": 2.6203, + "step": 33186 + }, + { + "epoch": 1.41, + "grad_norm": 0.439453125, + "learning_rate": 3.9588329689322624e-05, + "loss": 2.3531, + "step": 33187 + }, + { + "epoch": 1.41, + "grad_norm": 0.466796875, + "learning_rate": 3.9582984302386195e-05, + "loss": 2.4812, + "step": 33188 + }, + { + "epoch": 1.41, + "grad_norm": 0.49609375, + "learning_rate": 3.9577639187304286e-05, + "loss": 2.5038, + "step": 33189 + }, + { + "epoch": 1.41, + "grad_norm": 0.49609375, + "learning_rate": 3.9572294344100925e-05, + "loss": 2.4298, + "step": 33190 + }, + { + "epoch": 1.41, + "grad_norm": 0.470703125, + "learning_rate": 3.9566949772800074e-05, + "loss": 2.4632, + "step": 33191 + }, + { + "epoch": 1.41, + "grad_norm": 0.478515625, + "learning_rate": 3.956160547342594e-05, + "loss": 2.3506, + "step": 33192 + }, + { + "epoch": 1.41, + "grad_norm": 0.4375, + "learning_rate": 3.9556261446002494e-05, + "loss": 2.4709, + "step": 33193 + }, + { + "epoch": 1.41, + "grad_norm": 0.43359375, + "learning_rate": 3.955091769055375e-05, + "loss": 2.5069, + "step": 33194 + }, + { + "epoch": 1.41, + "grad_norm": 0.44921875, + "learning_rate": 3.9545574207103776e-05, + "loss": 2.3457, + "step": 33195 + }, + { + "epoch": 1.41, + "grad_norm": 0.46484375, + "learning_rate": 3.9540230995676665e-05, + "loss": 2.4291, + "step": 33196 + }, + { + "epoch": 1.41, + "grad_norm": 0.484375, + "learning_rate": 3.9534888056296385e-05, + "loss": 2.4268, + "step": 33197 + }, + { + "epoch": 1.41, + "grad_norm": 0.45703125, + "learning_rate": 3.9529545388987045e-05, + "loss": 2.3302, + "step": 33198 + }, + { + "epoch": 1.41, + "grad_norm": 0.4921875, + "learning_rate": 3.952420299377262e-05, + "loss": 2.4975, + "step": 33199 + }, + { + "epoch": 1.41, + "grad_norm": 0.494140625, + "learning_rate": 3.951886087067721e-05, + "loss": 2.4276, + "step": 33200 + }, + { + "epoch": 1.41, + "grad_norm": 0.482421875, + "learning_rate": 3.951351901972479e-05, + "loss": 2.3678, + "step": 33201 + }, + { + "epoch": 1.41, + "grad_norm": 0.4921875, + "learning_rate": 3.9508177440939466e-05, + "loss": 2.6077, + "step": 33202 + }, + { + "epoch": 1.41, + "grad_norm": 0.478515625, + "learning_rate": 3.950283613434519e-05, + "loss": 2.4805, + "step": 33203 + }, + { + "epoch": 1.41, + "grad_norm": 0.435546875, + "learning_rate": 3.9497495099966055e-05, + "loss": 2.4256, + "step": 33204 + }, + { + "epoch": 1.41, + "grad_norm": 0.419921875, + "learning_rate": 3.9492154337826116e-05, + "loss": 2.5097, + "step": 33205 + }, + { + "epoch": 1.42, + "grad_norm": 0.462890625, + "learning_rate": 3.948681384794937e-05, + "loss": 2.4398, + "step": 33206 + }, + { + "epoch": 1.42, + "grad_norm": 0.396484375, + "learning_rate": 3.948147363035977e-05, + "loss": 2.4655, + "step": 33207 + }, + { + "epoch": 1.42, + "grad_norm": 0.4921875, + "learning_rate": 3.9476133685081484e-05, + "loss": 2.5223, + "step": 33208 + }, + { + "epoch": 1.42, + "grad_norm": 0.515625, + "learning_rate": 3.947079401213849e-05, + "loss": 2.5712, + "step": 33209 + }, + { + "epoch": 1.42, + "grad_norm": 0.470703125, + "learning_rate": 3.946545461155476e-05, + "loss": 2.5507, + "step": 33210 + }, + { + "epoch": 1.42, + "grad_norm": 0.45703125, + "learning_rate": 3.9460115483354364e-05, + "loss": 2.5417, + "step": 33211 + }, + { + "epoch": 1.42, + "grad_norm": 0.4765625, + "learning_rate": 3.9454776627561366e-05, + "loss": 2.318, + "step": 33212 + }, + { + "epoch": 1.42, + "grad_norm": 0.4765625, + "learning_rate": 3.94494380441997e-05, + "loss": 2.5757, + "step": 33213 + }, + { + "epoch": 1.42, + "grad_norm": 0.48828125, + "learning_rate": 3.944409973329348e-05, + "loss": 2.4967, + "step": 33214 + }, + { + "epoch": 1.42, + "grad_norm": 0.4609375, + "learning_rate": 3.943876169486664e-05, + "loss": 2.3789, + "step": 33215 + }, + { + "epoch": 1.42, + "grad_norm": 0.443359375, + "learning_rate": 3.9433423928943267e-05, + "loss": 2.5363, + "step": 33216 + }, + { + "epoch": 1.42, + "grad_norm": 0.458984375, + "learning_rate": 3.942808643554733e-05, + "loss": 2.4477, + "step": 33217 + }, + { + "epoch": 1.42, + "grad_norm": 0.45703125, + "learning_rate": 3.94227492147029e-05, + "loss": 2.3815, + "step": 33218 + }, + { + "epoch": 1.42, + "grad_norm": 0.4609375, + "learning_rate": 3.9417412266433915e-05, + "loss": 2.5098, + "step": 33219 + }, + { + "epoch": 1.42, + "grad_norm": 0.46484375, + "learning_rate": 3.941207559076445e-05, + "loss": 2.5517, + "step": 33220 + }, + { + "epoch": 1.42, + "grad_norm": 0.52734375, + "learning_rate": 3.9406739187718524e-05, + "loss": 2.4943, + "step": 33221 + }, + { + "epoch": 1.42, + "grad_norm": 0.53515625, + "learning_rate": 3.940140305732013e-05, + "loss": 2.407, + "step": 33222 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.9396067199593214e-05, + "loss": 2.5599, + "step": 33223 + }, + { + "epoch": 1.42, + "grad_norm": 0.41796875, + "learning_rate": 3.9390731614561916e-05, + "loss": 2.4846, + "step": 33224 + }, + { + "epoch": 1.42, + "grad_norm": 0.484375, + "learning_rate": 3.938539630225017e-05, + "loss": 2.3768, + "step": 33225 + }, + { + "epoch": 1.42, + "grad_norm": 0.3984375, + "learning_rate": 3.938006126268196e-05, + "loss": 2.4132, + "step": 33226 + }, + { + "epoch": 1.42, + "grad_norm": 0.48828125, + "learning_rate": 3.937472649588134e-05, + "loss": 2.4976, + "step": 33227 + }, + { + "epoch": 1.42, + "grad_norm": 0.419921875, + "learning_rate": 3.9369392001872276e-05, + "loss": 2.3498, + "step": 33228 + }, + { + "epoch": 1.42, + "grad_norm": 0.46875, + "learning_rate": 3.936405778067878e-05, + "loss": 2.469, + "step": 33229 + }, + { + "epoch": 1.42, + "grad_norm": 0.453125, + "learning_rate": 3.935872383232491e-05, + "loss": 2.6008, + "step": 33230 + }, + { + "epoch": 1.42, + "grad_norm": 0.55078125, + "learning_rate": 3.93533901568346e-05, + "loss": 2.5812, + "step": 33231 + }, + { + "epoch": 1.42, + "grad_norm": 0.49609375, + "learning_rate": 3.934805675423185e-05, + "loss": 2.4527, + "step": 33232 + }, + { + "epoch": 1.42, + "grad_norm": 0.421875, + "learning_rate": 3.934272362454069e-05, + "loss": 2.4654, + "step": 33233 + }, + { + "epoch": 1.42, + "grad_norm": 0.494140625, + "learning_rate": 3.933739076778512e-05, + "loss": 2.4454, + "step": 33234 + }, + { + "epoch": 1.42, + "grad_norm": 0.458984375, + "learning_rate": 3.9332058183989096e-05, + "loss": 2.3899, + "step": 33235 + }, + { + "epoch": 1.42, + "grad_norm": 0.484375, + "learning_rate": 3.932672587317664e-05, + "loss": 2.3552, + "step": 33236 + }, + { + "epoch": 1.42, + "grad_norm": 0.484375, + "learning_rate": 3.932139383537178e-05, + "loss": 2.3878, + "step": 33237 + }, + { + "epoch": 1.42, + "grad_norm": 0.435546875, + "learning_rate": 3.931606207059847e-05, + "loss": 2.57, + "step": 33238 + }, + { + "epoch": 1.42, + "grad_norm": 0.51171875, + "learning_rate": 3.931073057888067e-05, + "loss": 2.5294, + "step": 33239 + }, + { + "epoch": 1.42, + "grad_norm": 0.5, + "learning_rate": 3.93053993602424e-05, + "loss": 2.4864, + "step": 33240 + }, + { + "epoch": 1.42, + "grad_norm": 0.42578125, + "learning_rate": 3.930006841470769e-05, + "loss": 2.4636, + "step": 33241 + }, + { + "epoch": 1.42, + "grad_norm": 0.4296875, + "learning_rate": 3.929473774230045e-05, + "loss": 2.3786, + "step": 33242 + }, + { + "epoch": 1.42, + "grad_norm": 0.439453125, + "learning_rate": 3.9289407343044735e-05, + "loss": 2.43, + "step": 33243 + }, + { + "epoch": 1.42, + "grad_norm": 0.43359375, + "learning_rate": 3.928407721696448e-05, + "loss": 2.4463, + "step": 33244 + }, + { + "epoch": 1.42, + "grad_norm": 0.515625, + "learning_rate": 3.927874736408368e-05, + "loss": 2.4077, + "step": 33245 + }, + { + "epoch": 1.42, + "grad_norm": 0.478515625, + "learning_rate": 3.927341778442636e-05, + "loss": 2.5295, + "step": 33246 + }, + { + "epoch": 1.42, + "grad_norm": 0.48828125, + "learning_rate": 3.926808847801646e-05, + "loss": 2.4293, + "step": 33247 + }, + { + "epoch": 1.42, + "grad_norm": 0.41796875, + "learning_rate": 3.926275944487795e-05, + "loss": 2.4466, + "step": 33248 + }, + { + "epoch": 1.42, + "grad_norm": 0.41796875, + "learning_rate": 3.925743068503481e-05, + "loss": 2.3518, + "step": 33249 + }, + { + "epoch": 1.42, + "grad_norm": 0.51171875, + "learning_rate": 3.925210219851108e-05, + "loss": 2.4752, + "step": 33250 + }, + { + "epoch": 1.42, + "grad_norm": 0.4296875, + "learning_rate": 3.9246773985330644e-05, + "loss": 2.3829, + "step": 33251 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.9241446045517524e-05, + "loss": 2.4155, + "step": 33252 + }, + { + "epoch": 1.42, + "grad_norm": 0.4609375, + "learning_rate": 3.923611837909573e-05, + "loss": 2.327, + "step": 33253 + }, + { + "epoch": 1.42, + "grad_norm": 0.4453125, + "learning_rate": 3.92307909860892e-05, + "loss": 2.484, + "step": 33254 + }, + { + "epoch": 1.42, + "grad_norm": 0.5234375, + "learning_rate": 3.922546386652185e-05, + "loss": 2.3899, + "step": 33255 + }, + { + "epoch": 1.42, + "grad_norm": 0.51171875, + "learning_rate": 3.9220137020417724e-05, + "loss": 2.4595, + "step": 33256 + }, + { + "epoch": 1.42, + "grad_norm": 0.515625, + "learning_rate": 3.921481044780079e-05, + "loss": 2.4215, + "step": 33257 + }, + { + "epoch": 1.42, + "grad_norm": 0.396484375, + "learning_rate": 3.920948414869496e-05, + "loss": 2.4511, + "step": 33258 + }, + { + "epoch": 1.42, + "grad_norm": 0.4453125, + "learning_rate": 3.920415812312428e-05, + "loss": 2.4484, + "step": 33259 + }, + { + "epoch": 1.42, + "grad_norm": 0.4375, + "learning_rate": 3.919883237111263e-05, + "loss": 2.4549, + "step": 33260 + }, + { + "epoch": 1.42, + "grad_norm": 0.474609375, + "learning_rate": 3.9193506892684005e-05, + "loss": 2.4576, + "step": 33261 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.9188181687862426e-05, + "loss": 2.3938, + "step": 33262 + }, + { + "epoch": 1.42, + "grad_norm": 0.46875, + "learning_rate": 3.918285675667179e-05, + "loss": 2.4421, + "step": 33263 + }, + { + "epoch": 1.42, + "grad_norm": 0.455078125, + "learning_rate": 3.917753209913605e-05, + "loss": 2.5068, + "step": 33264 + }, + { + "epoch": 1.42, + "grad_norm": 0.484375, + "learning_rate": 3.9172207715279185e-05, + "loss": 2.3231, + "step": 33265 + }, + { + "epoch": 1.42, + "grad_norm": 0.4609375, + "learning_rate": 3.9166883605125196e-05, + "loss": 2.4874, + "step": 33266 + }, + { + "epoch": 1.42, + "grad_norm": 0.46484375, + "learning_rate": 3.9161559768697954e-05, + "loss": 2.4891, + "step": 33267 + }, + { + "epoch": 1.42, + "grad_norm": 0.53125, + "learning_rate": 3.915623620602146e-05, + "loss": 2.6637, + "step": 33268 + }, + { + "epoch": 1.42, + "grad_norm": 0.416015625, + "learning_rate": 3.915091291711971e-05, + "loss": 2.3081, + "step": 33269 + }, + { + "epoch": 1.42, + "grad_norm": 0.5390625, + "learning_rate": 3.91455899020166e-05, + "loss": 2.4701, + "step": 33270 + }, + { + "epoch": 1.42, + "grad_norm": 0.4453125, + "learning_rate": 3.914026716073607e-05, + "loss": 2.4088, + "step": 33271 + }, + { + "epoch": 1.42, + "grad_norm": 0.4921875, + "learning_rate": 3.91349446933021e-05, + "loss": 2.4169, + "step": 33272 + }, + { + "epoch": 1.42, + "grad_norm": 0.40234375, + "learning_rate": 3.9129622499738664e-05, + "loss": 2.5008, + "step": 33273 + }, + { + "epoch": 1.42, + "grad_norm": 0.52734375, + "learning_rate": 3.912430058006965e-05, + "loss": 2.4598, + "step": 33274 + }, + { + "epoch": 1.42, + "grad_norm": 0.451171875, + "learning_rate": 3.9118978934319065e-05, + "loss": 2.4595, + "step": 33275 + }, + { + "epoch": 1.42, + "grad_norm": 0.447265625, + "learning_rate": 3.91136575625108e-05, + "loss": 2.4435, + "step": 33276 + }, + { + "epoch": 1.42, + "grad_norm": 0.498046875, + "learning_rate": 3.9108336464668815e-05, + "loss": 2.4566, + "step": 33277 + }, + { + "epoch": 1.42, + "grad_norm": 0.44140625, + "learning_rate": 3.91030156408171e-05, + "loss": 2.3099, + "step": 33278 + }, + { + "epoch": 1.42, + "grad_norm": 0.451171875, + "learning_rate": 3.909769509097956e-05, + "loss": 2.4719, + "step": 33279 + }, + { + "epoch": 1.42, + "grad_norm": 0.44140625, + "learning_rate": 3.90923748151801e-05, + "loss": 2.4409, + "step": 33280 + }, + { + "epoch": 1.42, + "grad_norm": 0.53515625, + "learning_rate": 3.908705481344269e-05, + "loss": 2.3724, + "step": 33281 + }, + { + "epoch": 1.42, + "grad_norm": 0.478515625, + "learning_rate": 3.908173508579132e-05, + "loss": 2.3959, + "step": 33282 + }, + { + "epoch": 1.42, + "grad_norm": 0.4921875, + "learning_rate": 3.907641563224983e-05, + "loss": 2.5352, + "step": 33283 + }, + { + "epoch": 1.42, + "grad_norm": 0.443359375, + "learning_rate": 3.9071096452842206e-05, + "loss": 2.553, + "step": 33284 + }, + { + "epoch": 1.42, + "grad_norm": 0.435546875, + "learning_rate": 3.906577754759242e-05, + "loss": 2.342, + "step": 33285 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.9060458916524365e-05, + "loss": 2.2983, + "step": 33286 + }, + { + "epoch": 1.42, + "grad_norm": 0.43359375, + "learning_rate": 3.905514055966194e-05, + "loss": 2.3691, + "step": 33287 + }, + { + "epoch": 1.42, + "grad_norm": 0.5078125, + "learning_rate": 3.904982247702911e-05, + "loss": 2.4163, + "step": 33288 + }, + { + "epoch": 1.42, + "grad_norm": 0.431640625, + "learning_rate": 3.904450466864984e-05, + "loss": 2.4709, + "step": 33289 + }, + { + "epoch": 1.42, + "grad_norm": 0.474609375, + "learning_rate": 3.903918713454798e-05, + "loss": 2.4322, + "step": 33290 + }, + { + "epoch": 1.42, + "grad_norm": 0.4921875, + "learning_rate": 3.9033869874747555e-05, + "loss": 2.4925, + "step": 33291 + }, + { + "epoch": 1.42, + "grad_norm": 0.42578125, + "learning_rate": 3.9028552889272376e-05, + "loss": 2.4422, + "step": 33292 + }, + { + "epoch": 1.42, + "grad_norm": 0.482421875, + "learning_rate": 3.902323617814644e-05, + "loss": 2.5177, + "step": 33293 + }, + { + "epoch": 1.42, + "grad_norm": 0.51953125, + "learning_rate": 3.9017919741393685e-05, + "loss": 2.4697, + "step": 33294 + }, + { + "epoch": 1.42, + "grad_norm": 0.4765625, + "learning_rate": 3.901260357903801e-05, + "loss": 2.4634, + "step": 33295 + }, + { + "epoch": 1.42, + "grad_norm": 0.4296875, + "learning_rate": 3.900728769110329e-05, + "loss": 2.3489, + "step": 33296 + }, + { + "epoch": 1.42, + "grad_norm": 0.470703125, + "learning_rate": 3.9001972077613494e-05, + "loss": 2.4697, + "step": 33297 + }, + { + "epoch": 1.42, + "grad_norm": 0.498046875, + "learning_rate": 3.899665673859255e-05, + "loss": 2.3868, + "step": 33298 + }, + { + "epoch": 1.42, + "grad_norm": 0.515625, + "learning_rate": 3.899134167406434e-05, + "loss": 2.5684, + "step": 33299 + }, + { + "epoch": 1.42, + "grad_norm": 0.494140625, + "learning_rate": 3.898602688405279e-05, + "loss": 2.5164, + "step": 33300 + }, + { + "epoch": 1.42, + "grad_norm": 0.478515625, + "learning_rate": 3.898071236858185e-05, + "loss": 2.4688, + "step": 33301 + }, + { + "epoch": 1.42, + "grad_norm": 0.484375, + "learning_rate": 3.897539812767541e-05, + "loss": 2.4954, + "step": 33302 + }, + { + "epoch": 1.42, + "grad_norm": 0.5390625, + "learning_rate": 3.897008416135735e-05, + "loss": 2.3972, + "step": 33303 + }, + { + "epoch": 1.42, + "grad_norm": 0.44921875, + "learning_rate": 3.896477046965159e-05, + "loss": 2.5015, + "step": 33304 + }, + { + "epoch": 1.42, + "grad_norm": 0.48828125, + "learning_rate": 3.895945705258209e-05, + "loss": 2.4453, + "step": 33305 + }, + { + "epoch": 1.42, + "grad_norm": 0.44921875, + "learning_rate": 3.895414391017271e-05, + "loss": 2.395, + "step": 33306 + }, + { + "epoch": 1.42, + "grad_norm": 0.4140625, + "learning_rate": 3.8948831042447396e-05, + "loss": 2.303, + "step": 33307 + }, + { + "epoch": 1.42, + "grad_norm": 0.494140625, + "learning_rate": 3.8943518449429997e-05, + "loss": 2.3725, + "step": 33308 + }, + { + "epoch": 1.42, + "grad_norm": 0.470703125, + "learning_rate": 3.8938206131144453e-05, + "loss": 2.4936, + "step": 33309 + }, + { + "epoch": 1.42, + "grad_norm": 0.47265625, + "learning_rate": 3.89328940876147e-05, + "loss": 2.5147, + "step": 33310 + }, + { + "epoch": 1.42, + "grad_norm": 0.44140625, + "learning_rate": 3.892758231886461e-05, + "loss": 2.5523, + "step": 33311 + }, + { + "epoch": 1.42, + "grad_norm": 0.474609375, + "learning_rate": 3.892227082491804e-05, + "loss": 2.5648, + "step": 33312 + }, + { + "epoch": 1.42, + "grad_norm": 0.515625, + "learning_rate": 3.891695960579893e-05, + "loss": 2.4964, + "step": 33313 + }, + { + "epoch": 1.42, + "grad_norm": 0.515625, + "learning_rate": 3.891164866153121e-05, + "loss": 2.6375, + "step": 33314 + }, + { + "epoch": 1.42, + "grad_norm": 0.5234375, + "learning_rate": 3.890633799213872e-05, + "loss": 2.4539, + "step": 33315 + }, + { + "epoch": 1.42, + "grad_norm": 0.482421875, + "learning_rate": 3.890102759764538e-05, + "loss": 2.4619, + "step": 33316 + }, + { + "epoch": 1.42, + "grad_norm": 0.484375, + "learning_rate": 3.8895717478075113e-05, + "loss": 2.486, + "step": 33317 + }, + { + "epoch": 1.42, + "grad_norm": 0.48046875, + "learning_rate": 3.8890407633451796e-05, + "loss": 2.5504, + "step": 33318 + }, + { + "epoch": 1.42, + "grad_norm": 0.4453125, + "learning_rate": 3.888509806379927e-05, + "loss": 2.4273, + "step": 33319 + }, + { + "epoch": 1.42, + "grad_norm": 0.400390625, + "learning_rate": 3.8879788769141465e-05, + "loss": 2.3311, + "step": 33320 + }, + { + "epoch": 1.42, + "grad_norm": 0.478515625, + "learning_rate": 3.8874479749502326e-05, + "loss": 2.3761, + "step": 33321 + }, + { + "epoch": 1.42, + "grad_norm": 0.435546875, + "learning_rate": 3.886917100490563e-05, + "loss": 2.4256, + "step": 33322 + }, + { + "epoch": 1.42, + "grad_norm": 0.453125, + "learning_rate": 3.8863862535375375e-05, + "loss": 2.5431, + "step": 33323 + }, + { + "epoch": 1.42, + "grad_norm": 0.408203125, + "learning_rate": 3.885855434093536e-05, + "loss": 2.4087, + "step": 33324 + }, + { + "epoch": 1.42, + "grad_norm": 0.49609375, + "learning_rate": 3.88532464216095e-05, + "loss": 2.4281, + "step": 33325 + }, + { + "epoch": 1.42, + "grad_norm": 0.48828125, + "learning_rate": 3.884793877742172e-05, + "loss": 2.5111, + "step": 33326 + }, + { + "epoch": 1.42, + "grad_norm": 0.455078125, + "learning_rate": 3.884263140839587e-05, + "loss": 2.4194, + "step": 33327 + }, + { + "epoch": 1.42, + "grad_norm": 0.451171875, + "learning_rate": 3.88373243145558e-05, + "loss": 2.3662, + "step": 33328 + }, + { + "epoch": 1.42, + "grad_norm": 0.490234375, + "learning_rate": 3.883201749592541e-05, + "loss": 2.5208, + "step": 33329 + }, + { + "epoch": 1.42, + "grad_norm": 0.49609375, + "learning_rate": 3.8826710952528623e-05, + "loss": 2.6328, + "step": 33330 + }, + { + "epoch": 1.42, + "grad_norm": 0.408203125, + "learning_rate": 3.8821404684389237e-05, + "loss": 2.3914, + "step": 33331 + }, + { + "epoch": 1.42, + "grad_norm": 0.478515625, + "learning_rate": 3.881609869153119e-05, + "loss": 2.666, + "step": 33332 + }, + { + "epoch": 1.42, + "grad_norm": 0.55859375, + "learning_rate": 3.881079297397836e-05, + "loss": 2.5975, + "step": 33333 + }, + { + "epoch": 1.42, + "grad_norm": 0.5078125, + "learning_rate": 3.88054875317546e-05, + "loss": 2.2987, + "step": 33334 + }, + { + "epoch": 1.42, + "grad_norm": 0.50390625, + "learning_rate": 3.8800182364883754e-05, + "loss": 2.3718, + "step": 33335 + }, + { + "epoch": 1.42, + "grad_norm": 0.365234375, + "learning_rate": 3.879487747338971e-05, + "loss": 2.3135, + "step": 33336 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.8789572857296393e-05, + "loss": 2.4629, + "step": 33337 + }, + { + "epoch": 1.42, + "grad_norm": 0.52734375, + "learning_rate": 3.87842685166276e-05, + "loss": 2.4561, + "step": 33338 + }, + { + "epoch": 1.42, + "grad_norm": 0.49609375, + "learning_rate": 3.8778964451407253e-05, + "loss": 2.5201, + "step": 33339 + }, + { + "epoch": 1.42, + "grad_norm": 0.49609375, + "learning_rate": 3.877366066165916e-05, + "loss": 2.4921, + "step": 33340 + }, + { + "epoch": 1.42, + "grad_norm": 0.4765625, + "learning_rate": 3.876835714740721e-05, + "loss": 2.4468, + "step": 33341 + }, + { + "epoch": 1.42, + "grad_norm": 0.466796875, + "learning_rate": 3.8763053908675316e-05, + "loss": 2.4169, + "step": 33342 + }, + { + "epoch": 1.42, + "grad_norm": 0.4609375, + "learning_rate": 3.87577509454873e-05, + "loss": 2.3777, + "step": 33343 + }, + { + "epoch": 1.42, + "grad_norm": 0.408203125, + "learning_rate": 3.8752448257866993e-05, + "loss": 2.3588, + "step": 33344 + }, + { + "epoch": 1.42, + "grad_norm": 0.423828125, + "learning_rate": 3.874714584583828e-05, + "loss": 2.445, + "step": 33345 + }, + { + "epoch": 1.42, + "grad_norm": 0.431640625, + "learning_rate": 3.8741843709425066e-05, + "loss": 2.4082, + "step": 33346 + }, + { + "epoch": 1.42, + "grad_norm": 0.4453125, + "learning_rate": 3.873654184865113e-05, + "loss": 2.4836, + "step": 33347 + }, + { + "epoch": 1.42, + "grad_norm": 0.423828125, + "learning_rate": 3.8731240263540406e-05, + "loss": 2.4754, + "step": 33348 + }, + { + "epoch": 1.42, + "grad_norm": 0.458984375, + "learning_rate": 3.872593895411666e-05, + "loss": 2.4846, + "step": 33349 + }, + { + "epoch": 1.42, + "grad_norm": 0.40625, + "learning_rate": 3.872063792040383e-05, + "loss": 2.4165, + "step": 33350 + }, + { + "epoch": 1.42, + "grad_norm": 0.423828125, + "learning_rate": 3.871533716242572e-05, + "loss": 2.517, + "step": 33351 + }, + { + "epoch": 1.42, + "grad_norm": 0.43359375, + "learning_rate": 3.8710036680206216e-05, + "loss": 2.5413, + "step": 33352 + }, + { + "epoch": 1.42, + "grad_norm": 0.4609375, + "learning_rate": 3.870473647376912e-05, + "loss": 2.533, + "step": 33353 + }, + { + "epoch": 1.42, + "grad_norm": 0.4140625, + "learning_rate": 3.86994365431383e-05, + "loss": 2.4357, + "step": 33354 + }, + { + "epoch": 1.42, + "grad_norm": 0.46875, + "learning_rate": 3.869413688833766e-05, + "loss": 2.5051, + "step": 33355 + }, + { + "epoch": 1.42, + "grad_norm": 0.46484375, + "learning_rate": 3.8688837509390994e-05, + "loss": 2.4592, + "step": 33356 + }, + { + "epoch": 1.42, + "grad_norm": 0.44921875, + "learning_rate": 3.8683538406322085e-05, + "loss": 2.4277, + "step": 33357 + }, + { + "epoch": 1.42, + "grad_norm": 0.4765625, + "learning_rate": 3.867823957915491e-05, + "loss": 2.4462, + "step": 33358 + }, + { + "epoch": 1.42, + "grad_norm": 0.451171875, + "learning_rate": 3.8672941027913245e-05, + "loss": 2.467, + "step": 33359 + }, + { + "epoch": 1.42, + "grad_norm": 0.421875, + "learning_rate": 3.8667642752620904e-05, + "loss": 2.3666, + "step": 33360 + }, + { + "epoch": 1.42, + "grad_norm": 0.4296875, + "learning_rate": 3.866234475330175e-05, + "loss": 2.407, + "step": 33361 + }, + { + "epoch": 1.42, + "grad_norm": 0.52734375, + "learning_rate": 3.865704702997966e-05, + "loss": 2.5624, + "step": 33362 + }, + { + "epoch": 1.42, + "grad_norm": 0.46484375, + "learning_rate": 3.865174958267841e-05, + "loss": 2.4459, + "step": 33363 + }, + { + "epoch": 1.42, + "grad_norm": 0.47265625, + "learning_rate": 3.8646452411421906e-05, + "loss": 2.3975, + "step": 33364 + }, + { + "epoch": 1.42, + "grad_norm": 0.4140625, + "learning_rate": 3.864115551623391e-05, + "loss": 2.4816, + "step": 33365 + }, + { + "epoch": 1.42, + "grad_norm": 0.455078125, + "learning_rate": 3.863585889713831e-05, + "loss": 2.4115, + "step": 33366 + }, + { + "epoch": 1.42, + "grad_norm": 0.5703125, + "learning_rate": 3.86305625541589e-05, + "loss": 2.4955, + "step": 33367 + }, + { + "epoch": 1.42, + "grad_norm": 0.484375, + "learning_rate": 3.8625266487319554e-05, + "loss": 2.346, + "step": 33368 + }, + { + "epoch": 1.42, + "grad_norm": 0.431640625, + "learning_rate": 3.861997069664406e-05, + "loss": 2.3351, + "step": 33369 + }, + { + "epoch": 1.42, + "grad_norm": 0.435546875, + "learning_rate": 3.8614675182156244e-05, + "loss": 2.3614, + "step": 33370 + }, + { + "epoch": 1.42, + "grad_norm": 0.4453125, + "learning_rate": 3.860937994388001e-05, + "loss": 2.4099, + "step": 33371 + }, + { + "epoch": 1.42, + "grad_norm": 0.46484375, + "learning_rate": 3.8604084981839116e-05, + "loss": 2.4675, + "step": 33372 + }, + { + "epoch": 1.42, + "grad_norm": 0.482421875, + "learning_rate": 3.859879029605734e-05, + "loss": 2.5203, + "step": 33373 + }, + { + "epoch": 1.42, + "grad_norm": 0.494140625, + "learning_rate": 3.859349588655864e-05, + "loss": 2.5562, + "step": 33374 + }, + { + "epoch": 1.42, + "grad_norm": 0.427734375, + "learning_rate": 3.858820175336677e-05, + "loss": 2.4029, + "step": 33375 + }, + { + "epoch": 1.42, + "grad_norm": 0.462890625, + "learning_rate": 3.8582907896505504e-05, + "loss": 2.3813, + "step": 33376 + }, + { + "epoch": 1.42, + "grad_norm": 0.4609375, + "learning_rate": 3.857761431599871e-05, + "loss": 2.4047, + "step": 33377 + }, + { + "epoch": 1.42, + "grad_norm": 0.470703125, + "learning_rate": 3.857232101187024e-05, + "loss": 2.5833, + "step": 33378 + }, + { + "epoch": 1.42, + "grad_norm": 0.52734375, + "learning_rate": 3.856702798414384e-05, + "loss": 2.452, + "step": 33379 + }, + { + "epoch": 1.42, + "grad_norm": 0.482421875, + "learning_rate": 3.85617352328434e-05, + "loss": 2.4562, + "step": 33380 + }, + { + "epoch": 1.42, + "grad_norm": 0.404296875, + "learning_rate": 3.855644275799266e-05, + "loss": 2.4117, + "step": 33381 + }, + { + "epoch": 1.42, + "grad_norm": 0.52734375, + "learning_rate": 3.855115055961551e-05, + "loss": 2.4652, + "step": 33382 + }, + { + "epoch": 1.42, + "grad_norm": 0.416015625, + "learning_rate": 3.85458586377357e-05, + "loss": 2.6067, + "step": 33383 + }, + { + "epoch": 1.42, + "grad_norm": 0.490234375, + "learning_rate": 3.8540566992377094e-05, + "loss": 2.5481, + "step": 33384 + }, + { + "epoch": 1.42, + "grad_norm": 0.470703125, + "learning_rate": 3.853527562356344e-05, + "loss": 2.4259, + "step": 33385 + }, + { + "epoch": 1.42, + "grad_norm": 0.48828125, + "learning_rate": 3.8529984531318575e-05, + "loss": 2.4066, + "step": 33386 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.852469371566636e-05, + "loss": 2.3739, + "step": 33387 + }, + { + "epoch": 1.42, + "grad_norm": 0.5, + "learning_rate": 3.851940317663055e-05, + "loss": 2.4571, + "step": 33388 + }, + { + "epoch": 1.42, + "grad_norm": 0.515625, + "learning_rate": 3.85141129142349e-05, + "loss": 2.5213, + "step": 33389 + }, + { + "epoch": 1.42, + "grad_norm": 0.4765625, + "learning_rate": 3.850882292850333e-05, + "loss": 2.4932, + "step": 33390 + }, + { + "epoch": 1.42, + "grad_norm": 0.50390625, + "learning_rate": 3.850353321945959e-05, + "loss": 2.444, + "step": 33391 + }, + { + "epoch": 1.42, + "grad_norm": 0.443359375, + "learning_rate": 3.849824378712743e-05, + "loss": 2.3828, + "step": 33392 + }, + { + "epoch": 1.42, + "grad_norm": 0.484375, + "learning_rate": 3.849295463153071e-05, + "loss": 2.4843, + "step": 33393 + }, + { + "epoch": 1.42, + "grad_norm": 0.439453125, + "learning_rate": 3.848766575269325e-05, + "loss": 2.4264, + "step": 33394 + }, + { + "epoch": 1.42, + "grad_norm": 0.458984375, + "learning_rate": 3.848237715063877e-05, + "loss": 2.3991, + "step": 33395 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.847708882539115e-05, + "loss": 2.4695, + "step": 33396 + }, + { + "epoch": 1.42, + "grad_norm": 0.51171875, + "learning_rate": 3.847180077697411e-05, + "loss": 2.4586, + "step": 33397 + }, + { + "epoch": 1.42, + "grad_norm": 0.466796875, + "learning_rate": 3.846651300541152e-05, + "loss": 2.4917, + "step": 33398 + }, + { + "epoch": 1.42, + "grad_norm": 0.435546875, + "learning_rate": 3.846122551072709e-05, + "loss": 2.4321, + "step": 33399 + }, + { + "epoch": 1.42, + "grad_norm": 0.50390625, + "learning_rate": 3.84559382929447e-05, + "loss": 2.3828, + "step": 33400 + }, + { + "epoch": 1.42, + "grad_norm": 0.46484375, + "learning_rate": 3.845065135208805e-05, + "loss": 2.6457, + "step": 33401 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.844536468818098e-05, + "loss": 2.4965, + "step": 33402 + }, + { + "epoch": 1.42, + "grad_norm": 0.474609375, + "learning_rate": 3.844007830124732e-05, + "loss": 2.4511, + "step": 33403 + }, + { + "epoch": 1.42, + "grad_norm": 0.515625, + "learning_rate": 3.84347921913108e-05, + "loss": 2.5138, + "step": 33404 + }, + { + "epoch": 1.42, + "grad_norm": 0.439453125, + "learning_rate": 3.842950635839514e-05, + "loss": 2.4687, + "step": 33405 + }, + { + "epoch": 1.42, + "grad_norm": 0.5, + "learning_rate": 3.842422080252428e-05, + "loss": 2.4127, + "step": 33406 + }, + { + "epoch": 1.42, + "grad_norm": 0.53515625, + "learning_rate": 3.841893552372192e-05, + "loss": 2.5989, + "step": 33407 + }, + { + "epoch": 1.42, + "grad_norm": 0.46484375, + "learning_rate": 3.8413650522011814e-05, + "loss": 2.4903, + "step": 33408 + }, + { + "epoch": 1.42, + "grad_norm": 0.5078125, + "learning_rate": 3.840836579741777e-05, + "loss": 2.4178, + "step": 33409 + }, + { + "epoch": 1.42, + "grad_norm": 0.447265625, + "learning_rate": 3.84030813499636e-05, + "loss": 2.5134, + "step": 33410 + }, + { + "epoch": 1.42, + "grad_norm": 0.38671875, + "learning_rate": 3.839779717967302e-05, + "loss": 2.4043, + "step": 33411 + }, + { + "epoch": 1.42, + "grad_norm": 0.439453125, + "learning_rate": 3.839251328656988e-05, + "loss": 2.6138, + "step": 33412 + }, + { + "epoch": 1.42, + "grad_norm": 0.4765625, + "learning_rate": 3.838722967067787e-05, + "loss": 2.5506, + "step": 33413 + }, + { + "epoch": 1.42, + "grad_norm": 0.439453125, + "learning_rate": 3.8381946332020854e-05, + "loss": 2.5131, + "step": 33414 + }, + { + "epoch": 1.42, + "grad_norm": 0.47265625, + "learning_rate": 3.837666327062253e-05, + "loss": 2.554, + "step": 33415 + }, + { + "epoch": 1.42, + "grad_norm": 0.51171875, + "learning_rate": 3.8371380486506724e-05, + "loss": 2.4124, + "step": 33416 + }, + { + "epoch": 1.42, + "grad_norm": 0.455078125, + "learning_rate": 3.836609797969715e-05, + "loss": 2.336, + "step": 33417 + }, + { + "epoch": 1.42, + "grad_norm": 0.421875, + "learning_rate": 3.836081575021762e-05, + "loss": 2.5276, + "step": 33418 + }, + { + "epoch": 1.42, + "grad_norm": 0.447265625, + "learning_rate": 3.835553379809191e-05, + "loss": 2.5123, + "step": 33419 + }, + { + "epoch": 1.42, + "grad_norm": 0.482421875, + "learning_rate": 3.835025212334378e-05, + "loss": 2.4974, + "step": 33420 + }, + { + "epoch": 1.42, + "grad_norm": 0.490234375, + "learning_rate": 3.834497072599692e-05, + "loss": 2.4476, + "step": 33421 + }, + { + "epoch": 1.42, + "grad_norm": 0.5234375, + "learning_rate": 3.8339689606075226e-05, + "loss": 2.3876, + "step": 33422 + }, + { + "epoch": 1.42, + "grad_norm": 0.46875, + "learning_rate": 3.833440876360239e-05, + "loss": 2.5051, + "step": 33423 + }, + { + "epoch": 1.42, + "grad_norm": 0.423828125, + "learning_rate": 3.832912819860215e-05, + "loss": 2.4011, + "step": 33424 + }, + { + "epoch": 1.42, + "grad_norm": 0.4453125, + "learning_rate": 3.832384791109829e-05, + "loss": 2.4887, + "step": 33425 + }, + { + "epoch": 1.42, + "grad_norm": 0.486328125, + "learning_rate": 3.831856790111461e-05, + "loss": 2.4812, + "step": 33426 + }, + { + "epoch": 1.42, + "grad_norm": 0.5234375, + "learning_rate": 3.83132881686748e-05, + "loss": 2.2681, + "step": 33427 + }, + { + "epoch": 1.42, + "grad_norm": 0.494140625, + "learning_rate": 3.8308008713802676e-05, + "loss": 2.4348, + "step": 33428 + }, + { + "epoch": 1.42, + "grad_norm": 0.4765625, + "learning_rate": 3.830272953652194e-05, + "loss": 2.5228, + "step": 33429 + }, + { + "epoch": 1.42, + "grad_norm": 0.466796875, + "learning_rate": 3.82974506368564e-05, + "loss": 2.4781, + "step": 33430 + }, + { + "epoch": 1.42, + "grad_norm": 0.453125, + "learning_rate": 3.829217201482976e-05, + "loss": 2.43, + "step": 33431 + }, + { + "epoch": 1.42, + "grad_norm": 0.50390625, + "learning_rate": 3.828689367046583e-05, + "loss": 2.4467, + "step": 33432 + }, + { + "epoch": 1.42, + "grad_norm": 0.5078125, + "learning_rate": 3.828161560378828e-05, + "loss": 2.4651, + "step": 33433 + }, + { + "epoch": 1.42, + "grad_norm": 0.44140625, + "learning_rate": 3.827633781482091e-05, + "loss": 2.3882, + "step": 33434 + }, + { + "epoch": 1.42, + "grad_norm": 0.51171875, + "learning_rate": 3.827106030358749e-05, + "loss": 2.4399, + "step": 33435 + }, + { + "epoch": 1.42, + "grad_norm": 0.427734375, + "learning_rate": 3.826578307011175e-05, + "loss": 2.4504, + "step": 33436 + }, + { + "epoch": 1.42, + "grad_norm": 0.51171875, + "learning_rate": 3.8260506114417346e-05, + "loss": 2.478, + "step": 33437 + }, + { + "epoch": 1.42, + "grad_norm": 0.423828125, + "learning_rate": 3.825522943652817e-05, + "loss": 2.4081, + "step": 33438 + }, + { + "epoch": 1.42, + "grad_norm": 0.46875, + "learning_rate": 3.82499530364679e-05, + "loss": 2.4601, + "step": 33439 + }, + { + "epoch": 1.42, + "grad_norm": 0.45703125, + "learning_rate": 3.824467691426025e-05, + "loss": 2.486, + "step": 33440 + }, + { + "epoch": 1.43, + "grad_norm": 0.5, + "learning_rate": 3.823940106992897e-05, + "loss": 2.4532, + "step": 33441 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.823412550349785e-05, + "loss": 2.6235, + "step": 33442 + }, + { + "epoch": 1.43, + "grad_norm": 0.46484375, + "learning_rate": 3.822885021499056e-05, + "loss": 2.3938, + "step": 33443 + }, + { + "epoch": 1.43, + "grad_norm": 0.453125, + "learning_rate": 3.822357520443092e-05, + "loss": 2.4, + "step": 33444 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.821830047184256e-05, + "loss": 2.5203, + "step": 33445 + }, + { + "epoch": 1.43, + "grad_norm": 0.53515625, + "learning_rate": 3.821302601724931e-05, + "loss": 2.5459, + "step": 33446 + }, + { + "epoch": 1.43, + "grad_norm": 0.546875, + "learning_rate": 3.820775184067482e-05, + "loss": 2.5179, + "step": 33447 + }, + { + "epoch": 1.43, + "grad_norm": 0.478515625, + "learning_rate": 3.820247794214291e-05, + "loss": 2.4603, + "step": 33448 + }, + { + "epoch": 1.43, + "grad_norm": 0.51953125, + "learning_rate": 3.8197204321677226e-05, + "loss": 2.4675, + "step": 33449 + }, + { + "epoch": 1.43, + "grad_norm": 0.466796875, + "learning_rate": 3.819193097930154e-05, + "loss": 2.3975, + "step": 33450 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.8186657915039604e-05, + "loss": 2.4844, + "step": 33451 + }, + { + "epoch": 1.43, + "grad_norm": 0.44921875, + "learning_rate": 3.818138512891513e-05, + "loss": 2.2694, + "step": 33452 + }, + { + "epoch": 1.43, + "grad_norm": 0.474609375, + "learning_rate": 3.8176112620951744e-05, + "loss": 2.4196, + "step": 33453 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.8170840391173334e-05, + "loss": 2.4441, + "step": 33454 + }, + { + "epoch": 1.43, + "grad_norm": 0.53125, + "learning_rate": 3.816556843960355e-05, + "loss": 2.4733, + "step": 33455 + }, + { + "epoch": 1.43, + "grad_norm": 0.455078125, + "learning_rate": 3.816029676626608e-05, + "loss": 2.3807, + "step": 33456 + }, + { + "epoch": 1.43, + "grad_norm": 0.4921875, + "learning_rate": 3.815502537118466e-05, + "loss": 2.5341, + "step": 33457 + }, + { + "epoch": 1.43, + "grad_norm": 0.458984375, + "learning_rate": 3.8149754254383075e-05, + "loss": 2.5391, + "step": 33458 + }, + { + "epoch": 1.43, + "grad_norm": 0.435546875, + "learning_rate": 3.8144483415884945e-05, + "loss": 2.2885, + "step": 33459 + }, + { + "epoch": 1.43, + "grad_norm": 0.416015625, + "learning_rate": 3.813921285571407e-05, + "loss": 2.4532, + "step": 33460 + }, + { + "epoch": 1.43, + "grad_norm": 0.47265625, + "learning_rate": 3.81339425738941e-05, + "loss": 2.4971, + "step": 33461 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.812867257044881e-05, + "loss": 2.4282, + "step": 33462 + }, + { + "epoch": 1.43, + "grad_norm": 0.484375, + "learning_rate": 3.812340284540185e-05, + "loss": 2.3988, + "step": 33463 + }, + { + "epoch": 1.43, + "grad_norm": 0.4609375, + "learning_rate": 3.8118133398777e-05, + "loss": 2.3832, + "step": 33464 + }, + { + "epoch": 1.43, + "grad_norm": 0.515625, + "learning_rate": 3.811286423059789e-05, + "loss": 2.4551, + "step": 33465 + }, + { + "epoch": 1.43, + "grad_norm": 0.478515625, + "learning_rate": 3.810759534088828e-05, + "loss": 2.5567, + "step": 33466 + }, + { + "epoch": 1.43, + "grad_norm": 0.515625, + "learning_rate": 3.810232672967191e-05, + "loss": 2.5765, + "step": 33467 + }, + { + "epoch": 1.43, + "grad_norm": 0.44140625, + "learning_rate": 3.809705839697245e-05, + "loss": 2.4981, + "step": 33468 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.809179034281356e-05, + "loss": 2.468, + "step": 33469 + }, + { + "epoch": 1.43, + "grad_norm": 0.484375, + "learning_rate": 3.8086522567219e-05, + "loss": 2.3772, + "step": 33470 + }, + { + "epoch": 1.43, + "grad_norm": 0.474609375, + "learning_rate": 3.8081255070212486e-05, + "loss": 2.4908, + "step": 33471 + }, + { + "epoch": 1.43, + "grad_norm": 0.46875, + "learning_rate": 3.8075987851817664e-05, + "loss": 2.5434, + "step": 33472 + }, + { + "epoch": 1.43, + "grad_norm": 0.42578125, + "learning_rate": 3.80707209120583e-05, + "loss": 2.5026, + "step": 33473 + }, + { + "epoch": 1.43, + "grad_norm": 0.478515625, + "learning_rate": 3.806545425095803e-05, + "loss": 2.4177, + "step": 33474 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.8060187868540574e-05, + "loss": 2.3835, + "step": 33475 + }, + { + "epoch": 1.43, + "grad_norm": 0.48828125, + "learning_rate": 3.805492176482968e-05, + "loss": 2.4666, + "step": 33476 + }, + { + "epoch": 1.43, + "grad_norm": 0.46484375, + "learning_rate": 3.8049655939848994e-05, + "loss": 2.259, + "step": 33477 + }, + { + "epoch": 1.43, + "grad_norm": 0.443359375, + "learning_rate": 3.8044390393622184e-05, + "loss": 2.6021, + "step": 33478 + }, + { + "epoch": 1.43, + "grad_norm": 0.462890625, + "learning_rate": 3.803912512617297e-05, + "loss": 2.3513, + "step": 33479 + }, + { + "epoch": 1.43, + "grad_norm": 0.41015625, + "learning_rate": 3.803386013752509e-05, + "loss": 2.3611, + "step": 33480 + }, + { + "epoch": 1.43, + "grad_norm": 0.39453125, + "learning_rate": 3.8028595427702155e-05, + "loss": 2.3735, + "step": 33481 + }, + { + "epoch": 1.43, + "grad_norm": 0.46484375, + "learning_rate": 3.802333099672789e-05, + "loss": 2.4083, + "step": 33482 + }, + { + "epoch": 1.43, + "grad_norm": 0.47265625, + "learning_rate": 3.801806684462602e-05, + "loss": 2.4577, + "step": 33483 + }, + { + "epoch": 1.43, + "grad_norm": 0.482421875, + "learning_rate": 3.8012802971420205e-05, + "loss": 2.4335, + "step": 33484 + }, + { + "epoch": 1.43, + "grad_norm": 0.427734375, + "learning_rate": 3.800753937713408e-05, + "loss": 2.379, + "step": 33485 + }, + { + "epoch": 1.43, + "grad_norm": 0.43359375, + "learning_rate": 3.800227606179137e-05, + "loss": 2.3989, + "step": 33486 + }, + { + "epoch": 1.43, + "grad_norm": 0.478515625, + "learning_rate": 3.7997013025415804e-05, + "loss": 2.3969, + "step": 33487 + }, + { + "epoch": 1.43, + "grad_norm": 0.427734375, + "learning_rate": 3.799175026803099e-05, + "loss": 2.3552, + "step": 33488 + }, + { + "epoch": 1.43, + "grad_norm": 0.46484375, + "learning_rate": 3.7986487789660665e-05, + "loss": 2.3902, + "step": 33489 + }, + { + "epoch": 1.43, + "grad_norm": 0.48046875, + "learning_rate": 3.798122559032844e-05, + "loss": 2.475, + "step": 33490 + }, + { + "epoch": 1.43, + "grad_norm": 0.55859375, + "learning_rate": 3.7975963670058046e-05, + "loss": 2.6599, + "step": 33491 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.7970702028873176e-05, + "loss": 2.3832, + "step": 33492 + }, + { + "epoch": 1.43, + "grad_norm": 0.498046875, + "learning_rate": 3.796544066679748e-05, + "loss": 2.5026, + "step": 33493 + }, + { + "epoch": 1.43, + "grad_norm": 0.41015625, + "learning_rate": 3.796017958385459e-05, + "loss": 2.4274, + "step": 33494 + }, + { + "epoch": 1.43, + "grad_norm": 0.474609375, + "learning_rate": 3.7954918780068224e-05, + "loss": 2.438, + "step": 33495 + }, + { + "epoch": 1.43, + "grad_norm": 0.49609375, + "learning_rate": 3.794965825546209e-05, + "loss": 2.4416, + "step": 33496 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.794439801005978e-05, + "loss": 2.3228, + "step": 33497 + }, + { + "epoch": 1.43, + "grad_norm": 0.5703125, + "learning_rate": 3.7939138043884995e-05, + "loss": 2.607, + "step": 33498 + }, + { + "epoch": 1.43, + "grad_norm": 0.431640625, + "learning_rate": 3.793387835696144e-05, + "loss": 2.3559, + "step": 33499 + }, + { + "epoch": 1.43, + "grad_norm": 0.48046875, + "learning_rate": 3.7928618949312756e-05, + "loss": 2.4684, + "step": 33500 + }, + { + "epoch": 1.43, + "grad_norm": 0.5234375, + "learning_rate": 3.7923359820962565e-05, + "loss": 2.512, + "step": 33501 + }, + { + "epoch": 1.43, + "grad_norm": 0.48046875, + "learning_rate": 3.791810097193457e-05, + "loss": 2.3796, + "step": 33502 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.7912842402252466e-05, + "loss": 2.4575, + "step": 33503 + }, + { + "epoch": 1.43, + "grad_norm": 0.4765625, + "learning_rate": 3.790758411193984e-05, + "loss": 2.3631, + "step": 33504 + }, + { + "epoch": 1.43, + "grad_norm": 0.4609375, + "learning_rate": 3.790232610102044e-05, + "loss": 2.3695, + "step": 33505 + }, + { + "epoch": 1.43, + "grad_norm": 0.451171875, + "learning_rate": 3.789706836951784e-05, + "loss": 2.471, + "step": 33506 + }, + { + "epoch": 1.43, + "grad_norm": 0.470703125, + "learning_rate": 3.789181091745573e-05, + "loss": 2.5347, + "step": 33507 + }, + { + "epoch": 1.43, + "grad_norm": 0.396484375, + "learning_rate": 3.788655374485781e-05, + "loss": 2.2375, + "step": 33508 + }, + { + "epoch": 1.43, + "grad_norm": 0.458984375, + "learning_rate": 3.7881296851747704e-05, + "loss": 2.3869, + "step": 33509 + }, + { + "epoch": 1.43, + "grad_norm": 0.453125, + "learning_rate": 3.787604023814899e-05, + "loss": 2.3912, + "step": 33510 + }, + { + "epoch": 1.43, + "grad_norm": 0.484375, + "learning_rate": 3.787078390408545e-05, + "loss": 2.4423, + "step": 33511 + }, + { + "epoch": 1.43, + "grad_norm": 0.44921875, + "learning_rate": 3.7865527849580686e-05, + "loss": 2.3601, + "step": 33512 + }, + { + "epoch": 1.43, + "grad_norm": 0.466796875, + "learning_rate": 3.786027207465831e-05, + "loss": 2.4065, + "step": 33513 + }, + { + "epoch": 1.43, + "grad_norm": 0.4609375, + "learning_rate": 3.785501657934199e-05, + "loss": 2.4251, + "step": 33514 + }, + { + "epoch": 1.43, + "grad_norm": 0.453125, + "learning_rate": 3.784976136365542e-05, + "loss": 2.3148, + "step": 33515 + }, + { + "epoch": 1.43, + "grad_norm": 0.412109375, + "learning_rate": 3.784450642762221e-05, + "loss": 2.4369, + "step": 33516 + }, + { + "epoch": 1.43, + "grad_norm": 0.408203125, + "learning_rate": 3.7839251771265973e-05, + "loss": 2.5325, + "step": 33517 + }, + { + "epoch": 1.43, + "grad_norm": 0.47265625, + "learning_rate": 3.7833997394610386e-05, + "loss": 2.4776, + "step": 33518 + }, + { + "epoch": 1.43, + "grad_norm": 0.47265625, + "learning_rate": 3.782874329767913e-05, + "loss": 2.4084, + "step": 33519 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.782348948049577e-05, + "loss": 2.3805, + "step": 33520 + }, + { + "epoch": 1.43, + "grad_norm": 0.484375, + "learning_rate": 3.7818235943084016e-05, + "loss": 2.3952, + "step": 33521 + }, + { + "epoch": 1.43, + "grad_norm": 0.4140625, + "learning_rate": 3.7812982685467446e-05, + "loss": 2.5499, + "step": 33522 + }, + { + "epoch": 1.43, + "grad_norm": 0.466796875, + "learning_rate": 3.780772970766973e-05, + "loss": 2.3528, + "step": 33523 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.780247700971453e-05, + "loss": 2.3964, + "step": 33524 + }, + { + "epoch": 1.43, + "grad_norm": 0.447265625, + "learning_rate": 3.7797224591625446e-05, + "loss": 2.4114, + "step": 33525 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.7791972453426064e-05, + "loss": 2.4456, + "step": 33526 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.778672059514014e-05, + "loss": 2.4679, + "step": 33527 + }, + { + "epoch": 1.43, + "grad_norm": 0.458984375, + "learning_rate": 3.778146901679123e-05, + "loss": 2.5267, + "step": 33528 + }, + { + "epoch": 1.43, + "grad_norm": 0.466796875, + "learning_rate": 3.777621771840293e-05, + "loss": 2.4808, + "step": 33529 + }, + { + "epoch": 1.43, + "grad_norm": 0.4375, + "learning_rate": 3.777096669999892e-05, + "loss": 2.3836, + "step": 33530 + }, + { + "epoch": 1.43, + "grad_norm": 0.443359375, + "learning_rate": 3.776571596160285e-05, + "loss": 2.4444, + "step": 33531 + }, + { + "epoch": 1.43, + "grad_norm": 0.46875, + "learning_rate": 3.7760465503238316e-05, + "loss": 2.5041, + "step": 33532 + }, + { + "epoch": 1.43, + "grad_norm": 0.484375, + "learning_rate": 3.775521532492891e-05, + "loss": 2.4813, + "step": 33533 + }, + { + "epoch": 1.43, + "grad_norm": 0.482421875, + "learning_rate": 3.7749965426698275e-05, + "loss": 2.5314, + "step": 33534 + }, + { + "epoch": 1.43, + "grad_norm": 0.46875, + "learning_rate": 3.774471580857009e-05, + "loss": 2.4433, + "step": 33535 + }, + { + "epoch": 1.43, + "grad_norm": 0.5234375, + "learning_rate": 3.7739466470567894e-05, + "loss": 2.5063, + "step": 33536 + }, + { + "epoch": 1.43, + "grad_norm": 0.44921875, + "learning_rate": 3.773421741271539e-05, + "loss": 2.5064, + "step": 33537 + }, + { + "epoch": 1.43, + "grad_norm": 0.56640625, + "learning_rate": 3.77289686350361e-05, + "loss": 2.6728, + "step": 33538 + }, + { + "epoch": 1.43, + "grad_norm": 0.498046875, + "learning_rate": 3.7723720137553706e-05, + "loss": 2.4104, + "step": 33539 + }, + { + "epoch": 1.43, + "grad_norm": 0.447265625, + "learning_rate": 3.771847192029184e-05, + "loss": 2.3174, + "step": 33540 + }, + { + "epoch": 1.43, + "grad_norm": 0.455078125, + "learning_rate": 3.771322398327409e-05, + "loss": 2.381, + "step": 33541 + }, + { + "epoch": 1.43, + "grad_norm": 0.5078125, + "learning_rate": 3.770797632652401e-05, + "loss": 2.4033, + "step": 33542 + }, + { + "epoch": 1.43, + "grad_norm": 0.49609375, + "learning_rate": 3.770272895006533e-05, + "loss": 2.5546, + "step": 33543 + }, + { + "epoch": 1.43, + "grad_norm": 0.4609375, + "learning_rate": 3.76974818539216e-05, + "loss": 2.3768, + "step": 33544 + }, + { + "epoch": 1.43, + "grad_norm": 0.5078125, + "learning_rate": 3.769223503811641e-05, + "loss": 2.5383, + "step": 33545 + }, + { + "epoch": 1.43, + "grad_norm": 0.5234375, + "learning_rate": 3.7686988502673384e-05, + "loss": 2.5588, + "step": 33546 + }, + { + "epoch": 1.43, + "grad_norm": 0.392578125, + "learning_rate": 3.768174224761617e-05, + "loss": 2.3838, + "step": 33547 + }, + { + "epoch": 1.43, + "grad_norm": 0.46484375, + "learning_rate": 3.767649627296834e-05, + "loss": 2.4173, + "step": 33548 + }, + { + "epoch": 1.43, + "grad_norm": 0.515625, + "learning_rate": 3.767125057875347e-05, + "loss": 2.5942, + "step": 33549 + }, + { + "epoch": 1.43, + "grad_norm": 0.474609375, + "learning_rate": 3.7666005164995196e-05, + "loss": 2.4583, + "step": 33550 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.766076003171714e-05, + "loss": 2.6103, + "step": 33551 + }, + { + "epoch": 1.43, + "grad_norm": 0.451171875, + "learning_rate": 3.7655515178942866e-05, + "loss": 2.3648, + "step": 33552 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.7650270606696014e-05, + "loss": 2.516, + "step": 33553 + }, + { + "epoch": 1.43, + "grad_norm": 0.4765625, + "learning_rate": 3.764502631500012e-05, + "loss": 2.4234, + "step": 33554 + }, + { + "epoch": 1.43, + "grad_norm": 0.482421875, + "learning_rate": 3.7639782303878825e-05, + "loss": 2.3215, + "step": 33555 + }, + { + "epoch": 1.43, + "grad_norm": 0.5859375, + "learning_rate": 3.763453857335576e-05, + "loss": 2.4206, + "step": 33556 + }, + { + "epoch": 1.43, + "grad_norm": 0.462890625, + "learning_rate": 3.7629295123454465e-05, + "loss": 2.5068, + "step": 33557 + }, + { + "epoch": 1.43, + "grad_norm": 0.43359375, + "learning_rate": 3.762405195419849e-05, + "loss": 2.4313, + "step": 33558 + }, + { + "epoch": 1.43, + "grad_norm": 0.48046875, + "learning_rate": 3.761880906561155e-05, + "loss": 2.3839, + "step": 33559 + }, + { + "epoch": 1.43, + "grad_norm": 0.5625, + "learning_rate": 3.761356645771717e-05, + "loss": 2.5444, + "step": 33560 + }, + { + "epoch": 1.43, + "grad_norm": 0.51953125, + "learning_rate": 3.7608324130538907e-05, + "loss": 2.4798, + "step": 33561 + }, + { + "epoch": 1.43, + "grad_norm": 0.46484375, + "learning_rate": 3.760308208410038e-05, + "loss": 2.6113, + "step": 33562 + }, + { + "epoch": 1.43, + "grad_norm": 0.451171875, + "learning_rate": 3.759784031842523e-05, + "loss": 2.4877, + "step": 33563 + }, + { + "epoch": 1.43, + "grad_norm": 0.423828125, + "learning_rate": 3.759259883353695e-05, + "loss": 2.5035, + "step": 33564 + }, + { + "epoch": 1.43, + "grad_norm": 0.443359375, + "learning_rate": 3.758735762945919e-05, + "loss": 2.4811, + "step": 33565 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.75821167062155e-05, + "loss": 2.4357, + "step": 33566 + }, + { + "epoch": 1.43, + "grad_norm": 0.53515625, + "learning_rate": 3.757687606382949e-05, + "loss": 2.4043, + "step": 33567 + }, + { + "epoch": 1.43, + "grad_norm": 0.49609375, + "learning_rate": 3.75716357023247e-05, + "loss": 2.5287, + "step": 33568 + }, + { + "epoch": 1.43, + "grad_norm": 0.44140625, + "learning_rate": 3.756639562172477e-05, + "loss": 2.4925, + "step": 33569 + }, + { + "epoch": 1.43, + "grad_norm": 0.40625, + "learning_rate": 3.756115582205319e-05, + "loss": 2.3968, + "step": 33570 + }, + { + "epoch": 1.43, + "grad_norm": 0.482421875, + "learning_rate": 3.755591630333361e-05, + "loss": 2.4647, + "step": 33571 + }, + { + "epoch": 1.43, + "grad_norm": 0.55078125, + "learning_rate": 3.755067706558961e-05, + "loss": 2.5238, + "step": 33572 + }, + { + "epoch": 1.43, + "grad_norm": 0.47265625, + "learning_rate": 3.7545438108844756e-05, + "loss": 2.3675, + "step": 33573 + }, + { + "epoch": 1.43, + "grad_norm": 0.5234375, + "learning_rate": 3.7540199433122526e-05, + "loss": 2.6223, + "step": 33574 + }, + { + "epoch": 1.43, + "grad_norm": 0.546875, + "learning_rate": 3.7534961038446635e-05, + "loss": 2.4689, + "step": 33575 + }, + { + "epoch": 1.43, + "grad_norm": 0.451171875, + "learning_rate": 3.75297229248406e-05, + "loss": 2.491, + "step": 33576 + }, + { + "epoch": 1.43, + "grad_norm": 0.478515625, + "learning_rate": 3.752448509232794e-05, + "loss": 2.428, + "step": 33577 + }, + { + "epoch": 1.43, + "grad_norm": 0.5078125, + "learning_rate": 3.751924754093227e-05, + "loss": 2.3923, + "step": 33578 + }, + { + "epoch": 1.43, + "grad_norm": 0.462890625, + "learning_rate": 3.751401027067719e-05, + "loss": 2.5854, + "step": 33579 + }, + { + "epoch": 1.43, + "grad_norm": 0.48046875, + "learning_rate": 3.7508773281586185e-05, + "loss": 2.3182, + "step": 33580 + }, + { + "epoch": 1.43, + "grad_norm": 0.4921875, + "learning_rate": 3.750353657368291e-05, + "loss": 2.491, + "step": 33581 + }, + { + "epoch": 1.43, + "grad_norm": 0.4765625, + "learning_rate": 3.749830014699083e-05, + "loss": 2.6706, + "step": 33582 + }, + { + "epoch": 1.43, + "grad_norm": 0.515625, + "learning_rate": 3.7493064001533594e-05, + "loss": 2.4993, + "step": 33583 + }, + { + "epoch": 1.43, + "grad_norm": 0.451171875, + "learning_rate": 3.7487828137334704e-05, + "loss": 2.4158, + "step": 33584 + }, + { + "epoch": 1.43, + "grad_norm": 0.447265625, + "learning_rate": 3.748259255441776e-05, + "loss": 2.583, + "step": 33585 + }, + { + "epoch": 1.43, + "grad_norm": 0.40625, + "learning_rate": 3.747735725280628e-05, + "loss": 2.4431, + "step": 33586 + }, + { + "epoch": 1.43, + "grad_norm": 0.470703125, + "learning_rate": 3.747212223252384e-05, + "loss": 2.3746, + "step": 33587 + }, + { + "epoch": 1.43, + "grad_norm": 0.48046875, + "learning_rate": 3.7466887493594027e-05, + "loss": 2.3575, + "step": 33588 + }, + { + "epoch": 1.43, + "grad_norm": 0.58203125, + "learning_rate": 3.746165303604038e-05, + "loss": 2.4636, + "step": 33589 + }, + { + "epoch": 1.43, + "grad_norm": 0.458984375, + "learning_rate": 3.74564188598864e-05, + "loss": 2.4932, + "step": 33590 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.745118496515567e-05, + "loss": 2.5363, + "step": 33591 + }, + { + "epoch": 1.43, + "grad_norm": 0.4375, + "learning_rate": 3.744595135187179e-05, + "loss": 2.3511, + "step": 33592 + }, + { + "epoch": 1.43, + "grad_norm": 0.439453125, + "learning_rate": 3.744071802005823e-05, + "loss": 2.2733, + "step": 33593 + }, + { + "epoch": 1.43, + "grad_norm": 0.490234375, + "learning_rate": 3.743548496973861e-05, + "loss": 2.5396, + "step": 33594 + }, + { + "epoch": 1.43, + "grad_norm": 0.443359375, + "learning_rate": 3.743025220093641e-05, + "loss": 2.4379, + "step": 33595 + }, + { + "epoch": 1.43, + "grad_norm": 0.451171875, + "learning_rate": 3.742501971367521e-05, + "loss": 2.3853, + "step": 33596 + }, + { + "epoch": 1.43, + "grad_norm": 0.498046875, + "learning_rate": 3.741978750797859e-05, + "loss": 2.3778, + "step": 33597 + }, + { + "epoch": 1.43, + "grad_norm": 0.482421875, + "learning_rate": 3.7414555583870046e-05, + "loss": 2.471, + "step": 33598 + }, + { + "epoch": 1.43, + "grad_norm": 0.474609375, + "learning_rate": 3.740932394137311e-05, + "loss": 2.5097, + "step": 33599 + }, + { + "epoch": 1.43, + "grad_norm": 0.408203125, + "learning_rate": 3.740409258051133e-05, + "loss": 2.4909, + "step": 33600 + }, + { + "epoch": 1.43, + "grad_norm": 0.470703125, + "learning_rate": 3.739886150130829e-05, + "loss": 2.499, + "step": 33601 + }, + { + "epoch": 1.43, + "grad_norm": 0.54296875, + "learning_rate": 3.739363070378751e-05, + "loss": 2.5199, + "step": 33602 + }, + { + "epoch": 1.43, + "grad_norm": 0.515625, + "learning_rate": 3.738840018797243e-05, + "loss": 2.5289, + "step": 33603 + }, + { + "epoch": 1.43, + "grad_norm": 0.462890625, + "learning_rate": 3.738316995388674e-05, + "loss": 2.5401, + "step": 33604 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.73779400015539e-05, + "loss": 2.5543, + "step": 33605 + }, + { + "epoch": 1.43, + "grad_norm": 0.470703125, + "learning_rate": 3.7372710330997405e-05, + "loss": 2.4764, + "step": 33606 + }, + { + "epoch": 1.43, + "grad_norm": 0.4921875, + "learning_rate": 3.736748094224084e-05, + "loss": 2.4417, + "step": 33607 + }, + { + "epoch": 1.43, + "grad_norm": 0.478515625, + "learning_rate": 3.736225183530774e-05, + "loss": 2.4727, + "step": 33608 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.735702301022159e-05, + "loss": 2.5626, + "step": 33609 + }, + { + "epoch": 1.43, + "grad_norm": 0.474609375, + "learning_rate": 3.735179446700597e-05, + "loss": 2.4718, + "step": 33610 + }, + { + "epoch": 1.43, + "grad_norm": 0.490234375, + "learning_rate": 3.734656620568435e-05, + "loss": 2.5779, + "step": 33611 + }, + { + "epoch": 1.43, + "grad_norm": 0.458984375, + "learning_rate": 3.734133822628029e-05, + "loss": 2.4816, + "step": 33612 + }, + { + "epoch": 1.43, + "grad_norm": 0.4609375, + "learning_rate": 3.733611052881734e-05, + "loss": 2.4681, + "step": 33613 + }, + { + "epoch": 1.43, + "grad_norm": 0.4140625, + "learning_rate": 3.7330883113318994e-05, + "loss": 2.4913, + "step": 33614 + }, + { + "epoch": 1.43, + "grad_norm": 0.48828125, + "learning_rate": 3.732565597980873e-05, + "loss": 2.4605, + "step": 33615 + }, + { + "epoch": 1.43, + "grad_norm": 0.439453125, + "learning_rate": 3.732042912831011e-05, + "loss": 2.3612, + "step": 33616 + }, + { + "epoch": 1.43, + "grad_norm": 0.4375, + "learning_rate": 3.731520255884669e-05, + "loss": 2.4888, + "step": 33617 + }, + { + "epoch": 1.43, + "grad_norm": 0.4921875, + "learning_rate": 3.7309976271441914e-05, + "loss": 2.5061, + "step": 33618 + }, + { + "epoch": 1.43, + "grad_norm": 0.546875, + "learning_rate": 3.730475026611934e-05, + "loss": 2.4708, + "step": 33619 + }, + { + "epoch": 1.43, + "grad_norm": 0.482421875, + "learning_rate": 3.72995245429025e-05, + "loss": 2.3993, + "step": 33620 + }, + { + "epoch": 1.43, + "grad_norm": 0.48828125, + "learning_rate": 3.729429910181489e-05, + "loss": 2.4996, + "step": 33621 + }, + { + "epoch": 1.43, + "grad_norm": 0.435546875, + "learning_rate": 3.7289073942879985e-05, + "loss": 2.3516, + "step": 33622 + }, + { + "epoch": 1.43, + "grad_norm": 0.490234375, + "learning_rate": 3.728384906612132e-05, + "loss": 2.4104, + "step": 33623 + }, + { + "epoch": 1.43, + "grad_norm": 0.421875, + "learning_rate": 3.7278624471562464e-05, + "loss": 2.3619, + "step": 33624 + }, + { + "epoch": 1.43, + "grad_norm": 0.53515625, + "learning_rate": 3.7273400159226834e-05, + "loss": 2.4948, + "step": 33625 + }, + { + "epoch": 1.43, + "grad_norm": 0.494140625, + "learning_rate": 3.726817612913801e-05, + "loss": 2.4072, + "step": 33626 + }, + { + "epoch": 1.43, + "grad_norm": 0.44921875, + "learning_rate": 3.726295238131944e-05, + "loss": 2.5269, + "step": 33627 + }, + { + "epoch": 1.43, + "grad_norm": 0.4375, + "learning_rate": 3.725772891579465e-05, + "loss": 2.3561, + "step": 33628 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.72525057325872e-05, + "loss": 2.5401, + "step": 33629 + }, + { + "epoch": 1.43, + "grad_norm": 0.5078125, + "learning_rate": 3.724728283172052e-05, + "loss": 2.3919, + "step": 33630 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.724206021321811e-05, + "loss": 2.4957, + "step": 33631 + }, + { + "epoch": 1.43, + "grad_norm": 0.494140625, + "learning_rate": 3.72368378771035e-05, + "loss": 2.4839, + "step": 33632 + }, + { + "epoch": 1.43, + "grad_norm": 0.478515625, + "learning_rate": 3.723161582340021e-05, + "loss": 2.5851, + "step": 33633 + }, + { + "epoch": 1.43, + "grad_norm": 0.484375, + "learning_rate": 3.7226394052131676e-05, + "loss": 2.4911, + "step": 33634 + }, + { + "epoch": 1.43, + "grad_norm": 0.498046875, + "learning_rate": 3.722117256332143e-05, + "loss": 2.6067, + "step": 33635 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.7215951356992996e-05, + "loss": 2.484, + "step": 33636 + }, + { + "epoch": 1.43, + "grad_norm": 0.51953125, + "learning_rate": 3.721073043316984e-05, + "loss": 2.5711, + "step": 33637 + }, + { + "epoch": 1.43, + "grad_norm": 0.4609375, + "learning_rate": 3.72055097918754e-05, + "loss": 2.3994, + "step": 33638 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.720028943313324e-05, + "loss": 2.3751, + "step": 33639 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.719506935696685e-05, + "loss": 2.4752, + "step": 33640 + }, + { + "epoch": 1.43, + "grad_norm": 0.494140625, + "learning_rate": 3.718984956339966e-05, + "loss": 2.5073, + "step": 33641 + }, + { + "epoch": 1.43, + "grad_norm": 0.466796875, + "learning_rate": 3.7184630052455224e-05, + "loss": 2.5508, + "step": 33642 + }, + { + "epoch": 1.43, + "grad_norm": 0.451171875, + "learning_rate": 3.7179410824156965e-05, + "loss": 2.3942, + "step": 33643 + }, + { + "epoch": 1.43, + "grad_norm": 0.484375, + "learning_rate": 3.71741918785284e-05, + "loss": 2.5627, + "step": 33644 + }, + { + "epoch": 1.43, + "grad_norm": 0.47265625, + "learning_rate": 3.716897321559305e-05, + "loss": 2.3912, + "step": 33645 + }, + { + "epoch": 1.43, + "grad_norm": 0.4921875, + "learning_rate": 3.716375483537435e-05, + "loss": 2.5252, + "step": 33646 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.7158536737895766e-05, + "loss": 2.5613, + "step": 33647 + }, + { + "epoch": 1.43, + "grad_norm": 0.46875, + "learning_rate": 3.7153318923180794e-05, + "loss": 2.4953, + "step": 33648 + }, + { + "epoch": 1.43, + "grad_norm": 0.431640625, + "learning_rate": 3.714810139125294e-05, + "loss": 2.5151, + "step": 33649 + }, + { + "epoch": 1.43, + "grad_norm": 0.47265625, + "learning_rate": 3.714288414213565e-05, + "loss": 2.4893, + "step": 33650 + }, + { + "epoch": 1.43, + "grad_norm": 0.4765625, + "learning_rate": 3.71376671758524e-05, + "loss": 2.3658, + "step": 33651 + }, + { + "epoch": 1.43, + "grad_norm": 0.490234375, + "learning_rate": 3.7132450492426704e-05, + "loss": 2.4749, + "step": 33652 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.712723409188201e-05, + "loss": 2.5064, + "step": 33653 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.712201797424175e-05, + "loss": 2.5963, + "step": 33654 + }, + { + "epoch": 1.43, + "grad_norm": 0.486328125, + "learning_rate": 3.711680213952943e-05, + "loss": 2.5331, + "step": 33655 + }, + { + "epoch": 1.43, + "grad_norm": 0.546875, + "learning_rate": 3.7111586587768556e-05, + "loss": 2.463, + "step": 33656 + }, + { + "epoch": 1.43, + "grad_norm": 0.42578125, + "learning_rate": 3.710637131898251e-05, + "loss": 2.3261, + "step": 33657 + }, + { + "epoch": 1.43, + "grad_norm": 0.45703125, + "learning_rate": 3.710115633319485e-05, + "loss": 2.4593, + "step": 33658 + }, + { + "epoch": 1.43, + "grad_norm": 0.486328125, + "learning_rate": 3.7095941630428974e-05, + "loss": 2.4617, + "step": 33659 + }, + { + "epoch": 1.43, + "grad_norm": 0.5546875, + "learning_rate": 3.709072721070838e-05, + "loss": 2.5214, + "step": 33660 + }, + { + "epoch": 1.43, + "grad_norm": 0.439453125, + "learning_rate": 3.7085513074056536e-05, + "loss": 2.4152, + "step": 33661 + }, + { + "epoch": 1.43, + "grad_norm": 0.4375, + "learning_rate": 3.7080299220496895e-05, + "loss": 2.4786, + "step": 33662 + }, + { + "epoch": 1.43, + "grad_norm": 0.50390625, + "learning_rate": 3.7075085650052887e-05, + "loss": 2.5631, + "step": 33663 + }, + { + "epoch": 1.43, + "grad_norm": 0.48828125, + "learning_rate": 3.706987236274799e-05, + "loss": 2.5609, + "step": 33664 + }, + { + "epoch": 1.43, + "grad_norm": 0.474609375, + "learning_rate": 3.70646593586057e-05, + "loss": 2.4212, + "step": 33665 + }, + { + "epoch": 1.43, + "grad_norm": 0.462890625, + "learning_rate": 3.705944663764942e-05, + "loss": 2.4392, + "step": 33666 + }, + { + "epoch": 1.43, + "grad_norm": 0.5, + "learning_rate": 3.705423419990263e-05, + "loss": 2.4689, + "step": 33667 + }, + { + "epoch": 1.43, + "grad_norm": 0.44140625, + "learning_rate": 3.704902204538881e-05, + "loss": 2.4987, + "step": 33668 + }, + { + "epoch": 1.43, + "grad_norm": 0.51171875, + "learning_rate": 3.704381017413138e-05, + "loss": 2.4542, + "step": 33669 + }, + { + "epoch": 1.43, + "grad_norm": 0.451171875, + "learning_rate": 3.703859858615376e-05, + "loss": 2.4167, + "step": 33670 + }, + { + "epoch": 1.43, + "grad_norm": 0.48828125, + "learning_rate": 3.703338728147945e-05, + "loss": 2.342, + "step": 33671 + }, + { + "epoch": 1.43, + "grad_norm": 0.51953125, + "learning_rate": 3.702817626013191e-05, + "loss": 2.5055, + "step": 33672 + }, + { + "epoch": 1.43, + "grad_norm": 0.427734375, + "learning_rate": 3.7022965522134526e-05, + "loss": 2.3549, + "step": 33673 + }, + { + "epoch": 1.43, + "grad_norm": 0.5234375, + "learning_rate": 3.701775506751083e-05, + "loss": 2.3144, + "step": 33674 + }, + { + "epoch": 1.43, + "grad_norm": 0.515625, + "learning_rate": 3.701254489628417e-05, + "loss": 2.4855, + "step": 33675 + }, + { + "epoch": 1.44, + "grad_norm": 0.427734375, + "learning_rate": 3.700733500847804e-05, + "loss": 2.432, + "step": 33676 + }, + { + "epoch": 1.44, + "grad_norm": 0.46875, + "learning_rate": 3.700212540411593e-05, + "loss": 2.4741, + "step": 33677 + }, + { + "epoch": 1.44, + "grad_norm": 0.51171875, + "learning_rate": 3.699691608322121e-05, + "loss": 2.3973, + "step": 33678 + }, + { + "epoch": 1.44, + "grad_norm": 0.455078125, + "learning_rate": 3.6991707045817326e-05, + "loss": 2.3711, + "step": 33679 + }, + { + "epoch": 1.44, + "grad_norm": 0.49609375, + "learning_rate": 3.698649829192773e-05, + "loss": 2.5098, + "step": 33680 + }, + { + "epoch": 1.44, + "grad_norm": 0.49609375, + "learning_rate": 3.6981289821575895e-05, + "loss": 2.4606, + "step": 33681 + }, + { + "epoch": 1.44, + "grad_norm": 0.404296875, + "learning_rate": 3.697608163478519e-05, + "loss": 2.3945, + "step": 33682 + }, + { + "epoch": 1.44, + "grad_norm": 0.41796875, + "learning_rate": 3.697087373157908e-05, + "loss": 2.5352, + "step": 33683 + }, + { + "epoch": 1.44, + "grad_norm": 0.490234375, + "learning_rate": 3.696566611198105e-05, + "loss": 2.5306, + "step": 33684 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.6960458776014484e-05, + "loss": 2.5306, + "step": 33685 + }, + { + "epoch": 1.44, + "grad_norm": 0.5, + "learning_rate": 3.6955251723702776e-05, + "loss": 2.5182, + "step": 33686 + }, + { + "epoch": 1.44, + "grad_norm": 0.4921875, + "learning_rate": 3.6950044955069395e-05, + "loss": 2.4699, + "step": 33687 + }, + { + "epoch": 1.44, + "grad_norm": 0.47265625, + "learning_rate": 3.69448384701378e-05, + "loss": 2.4679, + "step": 33688 + }, + { + "epoch": 1.44, + "grad_norm": 0.45703125, + "learning_rate": 3.693963226893137e-05, + "loss": 2.5901, + "step": 33689 + }, + { + "epoch": 1.44, + "grad_norm": 0.421875, + "learning_rate": 3.693442635147356e-05, + "loss": 2.387, + "step": 33690 + }, + { + "epoch": 1.44, + "grad_norm": 0.451171875, + "learning_rate": 3.692922071778776e-05, + "loss": 2.4457, + "step": 33691 + }, + { + "epoch": 1.44, + "grad_norm": 0.4375, + "learning_rate": 3.6924015367897426e-05, + "loss": 2.4792, + "step": 33692 + }, + { + "epoch": 1.44, + "grad_norm": 0.474609375, + "learning_rate": 3.691881030182599e-05, + "loss": 2.5005, + "step": 33693 + }, + { + "epoch": 1.44, + "grad_norm": 0.421875, + "learning_rate": 3.691360551959687e-05, + "loss": 2.3069, + "step": 33694 + }, + { + "epoch": 1.44, + "grad_norm": 0.5390625, + "learning_rate": 3.690840102123343e-05, + "loss": 2.4912, + "step": 33695 + }, + { + "epoch": 1.44, + "grad_norm": 0.4375, + "learning_rate": 3.6903196806759124e-05, + "loss": 2.4623, + "step": 33696 + }, + { + "epoch": 1.44, + "grad_norm": 0.45703125, + "learning_rate": 3.6897992876197406e-05, + "loss": 2.3808, + "step": 33697 + }, + { + "epoch": 1.44, + "grad_norm": 0.484375, + "learning_rate": 3.689278922957162e-05, + "loss": 2.4782, + "step": 33698 + }, + { + "epoch": 1.44, + "grad_norm": 0.43359375, + "learning_rate": 3.688758586690523e-05, + "loss": 2.4432, + "step": 33699 + }, + { + "epoch": 1.44, + "grad_norm": 0.4375, + "learning_rate": 3.688238278822167e-05, + "loss": 2.405, + "step": 33700 + }, + { + "epoch": 1.44, + "grad_norm": 0.43359375, + "learning_rate": 3.687717999354432e-05, + "loss": 2.4303, + "step": 33701 + }, + { + "epoch": 1.44, + "grad_norm": 0.486328125, + "learning_rate": 3.687197748289656e-05, + "loss": 2.577, + "step": 33702 + }, + { + "epoch": 1.44, + "grad_norm": 0.51953125, + "learning_rate": 3.6866775256301825e-05, + "loss": 2.4878, + "step": 33703 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.686157331378357e-05, + "loss": 2.5709, + "step": 33704 + }, + { + "epoch": 1.44, + "grad_norm": 0.4609375, + "learning_rate": 3.685637165536512e-05, + "loss": 2.3999, + "step": 33705 + }, + { + "epoch": 1.44, + "grad_norm": 0.44921875, + "learning_rate": 3.6851170281069955e-05, + "loss": 2.3935, + "step": 33706 + }, + { + "epoch": 1.44, + "grad_norm": 0.474609375, + "learning_rate": 3.684596919092142e-05, + "loss": 2.5223, + "step": 33707 + }, + { + "epoch": 1.44, + "grad_norm": 0.46875, + "learning_rate": 3.684076838494294e-05, + "loss": 2.5487, + "step": 33708 + }, + { + "epoch": 1.44, + "grad_norm": 0.515625, + "learning_rate": 3.683556786315795e-05, + "loss": 2.5132, + "step": 33709 + }, + { + "epoch": 1.44, + "grad_norm": 0.40234375, + "learning_rate": 3.6830367625589826e-05, + "loss": 2.4154, + "step": 33710 + }, + { + "epoch": 1.44, + "grad_norm": 0.4609375, + "learning_rate": 3.682516767226192e-05, + "loss": 2.4817, + "step": 33711 + }, + { + "epoch": 1.44, + "grad_norm": 0.4921875, + "learning_rate": 3.681996800319767e-05, + "loss": 2.4531, + "step": 33712 + }, + { + "epoch": 1.44, + "grad_norm": 0.51953125, + "learning_rate": 3.6814768618420514e-05, + "loss": 2.4438, + "step": 33713 + }, + { + "epoch": 1.44, + "grad_norm": 0.48046875, + "learning_rate": 3.680956951795378e-05, + "loss": 2.6769, + "step": 33714 + }, + { + "epoch": 1.44, + "grad_norm": 0.490234375, + "learning_rate": 3.680437070182091e-05, + "loss": 2.567, + "step": 33715 + }, + { + "epoch": 1.44, + "grad_norm": 0.498046875, + "learning_rate": 3.679917217004526e-05, + "loss": 2.4921, + "step": 33716 + }, + { + "epoch": 1.44, + "grad_norm": 0.470703125, + "learning_rate": 3.679397392265026e-05, + "loss": 2.427, + "step": 33717 + }, + { + "epoch": 1.44, + "grad_norm": 0.4375, + "learning_rate": 3.678877595965924e-05, + "loss": 2.4407, + "step": 33718 + }, + { + "epoch": 1.44, + "grad_norm": 0.51171875, + "learning_rate": 3.6783578281095674e-05, + "loss": 2.3476, + "step": 33719 + }, + { + "epoch": 1.44, + "grad_norm": 0.5546875, + "learning_rate": 3.6778380886982855e-05, + "loss": 2.5198, + "step": 33720 + }, + { + "epoch": 1.44, + "grad_norm": 0.478515625, + "learning_rate": 3.6773183777344236e-05, + "loss": 2.513, + "step": 33721 + }, + { + "epoch": 1.44, + "grad_norm": 0.48046875, + "learning_rate": 3.676798695220321e-05, + "loss": 2.5366, + "step": 33722 + }, + { + "epoch": 1.44, + "grad_norm": 0.52734375, + "learning_rate": 3.6762790411583113e-05, + "loss": 2.4952, + "step": 33723 + }, + { + "epoch": 1.44, + "grad_norm": 0.474609375, + "learning_rate": 3.67575941555073e-05, + "loss": 2.5499, + "step": 33724 + }, + { + "epoch": 1.44, + "grad_norm": 0.47265625, + "learning_rate": 3.675239818399927e-05, + "loss": 2.4977, + "step": 33725 + }, + { + "epoch": 1.44, + "grad_norm": 0.49609375, + "learning_rate": 3.6747202497082334e-05, + "loss": 2.5522, + "step": 33726 + }, + { + "epoch": 1.44, + "grad_norm": 0.5078125, + "learning_rate": 3.674200709477983e-05, + "loss": 2.4289, + "step": 33727 + }, + { + "epoch": 1.44, + "grad_norm": 0.498046875, + "learning_rate": 3.673681197711517e-05, + "loss": 2.5521, + "step": 33728 + }, + { + "epoch": 1.44, + "grad_norm": 0.447265625, + "learning_rate": 3.6731617144111786e-05, + "loss": 2.4582, + "step": 33729 + }, + { + "epoch": 1.44, + "grad_norm": 0.42578125, + "learning_rate": 3.672642259579296e-05, + "loss": 2.3821, + "step": 33730 + }, + { + "epoch": 1.44, + "grad_norm": 0.515625, + "learning_rate": 3.6721228332182144e-05, + "loss": 2.4101, + "step": 33731 + }, + { + "epoch": 1.44, + "grad_norm": 0.5078125, + "learning_rate": 3.6716034353302645e-05, + "loss": 2.3952, + "step": 33732 + }, + { + "epoch": 1.44, + "grad_norm": 0.427734375, + "learning_rate": 3.671084065917789e-05, + "loss": 2.4704, + "step": 33733 + }, + { + "epoch": 1.44, + "grad_norm": 0.421875, + "learning_rate": 3.670564724983119e-05, + "loss": 2.4587, + "step": 33734 + }, + { + "epoch": 1.44, + "grad_norm": 0.44140625, + "learning_rate": 3.670045412528598e-05, + "loss": 2.4573, + "step": 33735 + }, + { + "epoch": 1.44, + "grad_norm": 0.462890625, + "learning_rate": 3.6695261285565565e-05, + "loss": 2.4623, + "step": 33736 + }, + { + "epoch": 1.44, + "grad_norm": 0.48046875, + "learning_rate": 3.6690068730693326e-05, + "loss": 2.4409, + "step": 33737 + }, + { + "epoch": 1.44, + "grad_norm": 0.4296875, + "learning_rate": 3.668487646069269e-05, + "loss": 2.437, + "step": 33738 + }, + { + "epoch": 1.44, + "grad_norm": 0.4765625, + "learning_rate": 3.667968447558696e-05, + "loss": 2.4545, + "step": 33739 + }, + { + "epoch": 1.44, + "grad_norm": 0.423828125, + "learning_rate": 3.6674492775399446e-05, + "loss": 2.422, + "step": 33740 + }, + { + "epoch": 1.44, + "grad_norm": 0.46484375, + "learning_rate": 3.666930136015364e-05, + "loss": 2.467, + "step": 33741 + }, + { + "epoch": 1.44, + "grad_norm": 0.443359375, + "learning_rate": 3.666411022987284e-05, + "loss": 2.5266, + "step": 33742 + }, + { + "epoch": 1.44, + "grad_norm": 0.4765625, + "learning_rate": 3.665891938458035e-05, + "loss": 2.4716, + "step": 33743 + }, + { + "epoch": 1.44, + "grad_norm": 0.44921875, + "learning_rate": 3.665372882429958e-05, + "loss": 2.4569, + "step": 33744 + }, + { + "epoch": 1.44, + "grad_norm": 0.46875, + "learning_rate": 3.664853854905391e-05, + "loss": 2.4726, + "step": 33745 + }, + { + "epoch": 1.44, + "grad_norm": 0.5234375, + "learning_rate": 3.664334855886664e-05, + "loss": 2.5003, + "step": 33746 + }, + { + "epoch": 1.44, + "grad_norm": 0.462890625, + "learning_rate": 3.663815885376117e-05, + "loss": 2.3207, + "step": 33747 + }, + { + "epoch": 1.44, + "grad_norm": 0.53125, + "learning_rate": 3.66329694337608e-05, + "loss": 2.4919, + "step": 33748 + }, + { + "epoch": 1.44, + "grad_norm": 0.4453125, + "learning_rate": 3.6627780298888946e-05, + "loss": 2.5042, + "step": 33749 + }, + { + "epoch": 1.44, + "grad_norm": 0.55078125, + "learning_rate": 3.6622591449168886e-05, + "loss": 2.5901, + "step": 33750 + }, + { + "epoch": 1.44, + "grad_norm": 0.48828125, + "learning_rate": 3.661740288462403e-05, + "loss": 2.3646, + "step": 33751 + }, + { + "epoch": 1.44, + "grad_norm": 0.439453125, + "learning_rate": 3.6612214605277675e-05, + "loss": 2.3921, + "step": 33752 + }, + { + "epoch": 1.44, + "grad_norm": 0.4375, + "learning_rate": 3.6607026611153185e-05, + "loss": 2.4233, + "step": 33753 + }, + { + "epoch": 1.44, + "grad_norm": 0.5078125, + "learning_rate": 3.660183890227393e-05, + "loss": 2.3713, + "step": 33754 + }, + { + "epoch": 1.44, + "grad_norm": 0.470703125, + "learning_rate": 3.6596651478663246e-05, + "loss": 2.3974, + "step": 33755 + }, + { + "epoch": 1.44, + "grad_norm": 0.494140625, + "learning_rate": 3.659146434034438e-05, + "loss": 2.5544, + "step": 33756 + }, + { + "epoch": 1.44, + "grad_norm": 0.4609375, + "learning_rate": 3.658627748734082e-05, + "loss": 2.5084, + "step": 33757 + }, + { + "epoch": 1.44, + "grad_norm": 0.388671875, + "learning_rate": 3.658109091967584e-05, + "loss": 2.4675, + "step": 33758 + }, + { + "epoch": 1.44, + "grad_norm": 0.5, + "learning_rate": 3.657590463737275e-05, + "loss": 2.3964, + "step": 33759 + }, + { + "epoch": 1.44, + "grad_norm": 0.44140625, + "learning_rate": 3.6570718640454894e-05, + "loss": 2.5141, + "step": 33760 + }, + { + "epoch": 1.44, + "grad_norm": 0.47265625, + "learning_rate": 3.656553292894567e-05, + "loss": 2.4999, + "step": 33761 + }, + { + "epoch": 1.44, + "grad_norm": 0.52734375, + "learning_rate": 3.656034750286832e-05, + "loss": 2.3869, + "step": 33762 + }, + { + "epoch": 1.44, + "grad_norm": 0.482421875, + "learning_rate": 3.6555162362246265e-05, + "loss": 2.6163, + "step": 33763 + }, + { + "epoch": 1.44, + "grad_norm": 0.51171875, + "learning_rate": 3.654997750710275e-05, + "loss": 2.614, + "step": 33764 + }, + { + "epoch": 1.44, + "grad_norm": 0.466796875, + "learning_rate": 3.654479293746118e-05, + "loss": 2.5322, + "step": 33765 + }, + { + "epoch": 1.44, + "grad_norm": 0.478515625, + "learning_rate": 3.6539608653344824e-05, + "loss": 2.4312, + "step": 33766 + }, + { + "epoch": 1.44, + "grad_norm": 0.51953125, + "learning_rate": 3.6534424654777074e-05, + "loss": 2.4365, + "step": 33767 + }, + { + "epoch": 1.44, + "grad_norm": 0.48046875, + "learning_rate": 3.6529240941781184e-05, + "loss": 2.5672, + "step": 33768 + }, + { + "epoch": 1.44, + "grad_norm": 0.466796875, + "learning_rate": 3.652405751438052e-05, + "loss": 2.4848, + "step": 33769 + }, + { + "epoch": 1.44, + "grad_norm": 0.478515625, + "learning_rate": 3.651887437259842e-05, + "loss": 2.539, + "step": 33770 + }, + { + "epoch": 1.44, + "grad_norm": 0.4453125, + "learning_rate": 3.6513691516458185e-05, + "loss": 2.3875, + "step": 33771 + }, + { + "epoch": 1.44, + "grad_norm": 0.546875, + "learning_rate": 3.650850894598308e-05, + "loss": 2.5638, + "step": 33772 + }, + { + "epoch": 1.44, + "grad_norm": 0.4765625, + "learning_rate": 3.650332666119653e-05, + "loss": 2.4241, + "step": 33773 + }, + { + "epoch": 1.44, + "grad_norm": 0.5390625, + "learning_rate": 3.649814466212182e-05, + "loss": 2.5185, + "step": 33774 + }, + { + "epoch": 1.44, + "grad_norm": 0.5, + "learning_rate": 3.6492962948782225e-05, + "loss": 2.3991, + "step": 33775 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.648778152120108e-05, + "loss": 2.5152, + "step": 33776 + }, + { + "epoch": 1.44, + "grad_norm": 0.46484375, + "learning_rate": 3.648260037940173e-05, + "loss": 2.4163, + "step": 33777 + }, + { + "epoch": 1.44, + "grad_norm": 0.431640625, + "learning_rate": 3.647741952340745e-05, + "loss": 2.4629, + "step": 33778 + }, + { + "epoch": 1.44, + "grad_norm": 0.458984375, + "learning_rate": 3.6472238953241585e-05, + "loss": 2.4347, + "step": 33779 + }, + { + "epoch": 1.44, + "grad_norm": 0.453125, + "learning_rate": 3.64670586689274e-05, + "loss": 2.4183, + "step": 33780 + }, + { + "epoch": 1.44, + "grad_norm": 0.392578125, + "learning_rate": 3.646187867048827e-05, + "loss": 2.3473, + "step": 33781 + }, + { + "epoch": 1.44, + "grad_norm": 0.4453125, + "learning_rate": 3.645669895794743e-05, + "loss": 2.3894, + "step": 33782 + }, + { + "epoch": 1.44, + "grad_norm": 0.490234375, + "learning_rate": 3.645151953132826e-05, + "loss": 2.3571, + "step": 33783 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.644634039065399e-05, + "loss": 2.3295, + "step": 33784 + }, + { + "epoch": 1.44, + "grad_norm": 0.478515625, + "learning_rate": 3.6441161535947954e-05, + "loss": 2.4938, + "step": 33785 + }, + { + "epoch": 1.44, + "grad_norm": 0.443359375, + "learning_rate": 3.643598296723351e-05, + "loss": 2.4384, + "step": 33786 + }, + { + "epoch": 1.44, + "grad_norm": 0.44921875, + "learning_rate": 3.6430804684533906e-05, + "loss": 2.4235, + "step": 33787 + }, + { + "epoch": 1.44, + "grad_norm": 0.4765625, + "learning_rate": 3.6425626687872384e-05, + "loss": 2.5187, + "step": 33788 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.642044897727237e-05, + "loss": 2.5162, + "step": 33789 + }, + { + "epoch": 1.44, + "grad_norm": 0.47265625, + "learning_rate": 3.6415271552757104e-05, + "loss": 2.3869, + "step": 33790 + }, + { + "epoch": 1.44, + "grad_norm": 0.4453125, + "learning_rate": 3.641009441434985e-05, + "loss": 2.4075, + "step": 33791 + }, + { + "epoch": 1.44, + "grad_norm": 0.419921875, + "learning_rate": 3.640491756207393e-05, + "loss": 2.4477, + "step": 33792 + }, + { + "epoch": 1.44, + "grad_norm": 0.484375, + "learning_rate": 3.6399740995952655e-05, + "loss": 2.4647, + "step": 33793 + }, + { + "epoch": 1.44, + "grad_norm": 0.51953125, + "learning_rate": 3.6394564716009283e-05, + "loss": 2.4903, + "step": 33794 + }, + { + "epoch": 1.44, + "grad_norm": 0.43359375, + "learning_rate": 3.638938872226716e-05, + "loss": 2.4287, + "step": 33795 + }, + { + "epoch": 1.44, + "grad_norm": 0.44140625, + "learning_rate": 3.638421301474951e-05, + "loss": 2.3317, + "step": 33796 + }, + { + "epoch": 1.44, + "grad_norm": 0.474609375, + "learning_rate": 3.637903759347968e-05, + "loss": 2.5142, + "step": 33797 + }, + { + "epoch": 1.44, + "grad_norm": 0.39453125, + "learning_rate": 3.63738624584809e-05, + "loss": 2.4233, + "step": 33798 + }, + { + "epoch": 1.44, + "grad_norm": 0.451171875, + "learning_rate": 3.636868760977652e-05, + "loss": 2.3977, + "step": 33799 + }, + { + "epoch": 1.44, + "grad_norm": 0.470703125, + "learning_rate": 3.636351304738975e-05, + "loss": 2.6476, + "step": 33800 + }, + { + "epoch": 1.44, + "grad_norm": 0.455078125, + "learning_rate": 3.6358338771343924e-05, + "loss": 2.3755, + "step": 33801 + }, + { + "epoch": 1.44, + "grad_norm": 0.53125, + "learning_rate": 3.6353164781662344e-05, + "loss": 2.5347, + "step": 33802 + }, + { + "epoch": 1.44, + "grad_norm": 0.490234375, + "learning_rate": 3.634799107836826e-05, + "loss": 2.4339, + "step": 33803 + }, + { + "epoch": 1.44, + "grad_norm": 0.482421875, + "learning_rate": 3.63428176614849e-05, + "loss": 2.391, + "step": 33804 + }, + { + "epoch": 1.44, + "grad_norm": 0.4765625, + "learning_rate": 3.633764453103564e-05, + "loss": 2.4001, + "step": 33805 + }, + { + "epoch": 1.44, + "grad_norm": 0.4921875, + "learning_rate": 3.633247168704374e-05, + "loss": 2.4058, + "step": 33806 + }, + { + "epoch": 1.44, + "grad_norm": 0.484375, + "learning_rate": 3.63272991295324e-05, + "loss": 2.5293, + "step": 33807 + }, + { + "epoch": 1.44, + "grad_norm": 0.55078125, + "learning_rate": 3.632212685852494e-05, + "loss": 2.4788, + "step": 33808 + }, + { + "epoch": 1.44, + "grad_norm": 0.494140625, + "learning_rate": 3.6316954874044685e-05, + "loss": 2.5152, + "step": 33809 + }, + { + "epoch": 1.44, + "grad_norm": 0.455078125, + "learning_rate": 3.631178317611482e-05, + "loss": 2.4661, + "step": 33810 + }, + { + "epoch": 1.44, + "grad_norm": 0.44921875, + "learning_rate": 3.630661176475869e-05, + "loss": 2.427, + "step": 33811 + }, + { + "epoch": 1.44, + "grad_norm": 0.474609375, + "learning_rate": 3.6301440639999495e-05, + "loss": 2.5523, + "step": 33812 + }, + { + "epoch": 1.44, + "grad_norm": 0.484375, + "learning_rate": 3.6296269801860574e-05, + "loss": 2.5207, + "step": 33813 + }, + { + "epoch": 1.44, + "grad_norm": 0.443359375, + "learning_rate": 3.629109925036511e-05, + "loss": 2.3902, + "step": 33814 + }, + { + "epoch": 1.44, + "grad_norm": 0.462890625, + "learning_rate": 3.628592898553647e-05, + "loss": 2.4132, + "step": 33815 + }, + { + "epoch": 1.44, + "grad_norm": 0.44140625, + "learning_rate": 3.6280759007397824e-05, + "loss": 2.5051, + "step": 33816 + }, + { + "epoch": 1.44, + "grad_norm": 0.4296875, + "learning_rate": 3.6275589315972467e-05, + "loss": 2.3475, + "step": 33817 + }, + { + "epoch": 1.44, + "grad_norm": 0.447265625, + "learning_rate": 3.627041991128371e-05, + "loss": 2.5674, + "step": 33818 + }, + { + "epoch": 1.44, + "grad_norm": 0.427734375, + "learning_rate": 3.626525079335478e-05, + "loss": 2.445, + "step": 33819 + }, + { + "epoch": 1.44, + "grad_norm": 0.462890625, + "learning_rate": 3.6260081962208845e-05, + "loss": 2.3507, + "step": 33820 + }, + { + "epoch": 1.44, + "grad_norm": 0.4453125, + "learning_rate": 3.625491341786933e-05, + "loss": 2.4274, + "step": 33821 + }, + { + "epoch": 1.44, + "grad_norm": 0.458984375, + "learning_rate": 3.6249745160359406e-05, + "loss": 2.3359, + "step": 33822 + }, + { + "epoch": 1.44, + "grad_norm": 0.4375, + "learning_rate": 3.624457718970229e-05, + "loss": 2.4253, + "step": 33823 + }, + { + "epoch": 1.44, + "grad_norm": 0.40234375, + "learning_rate": 3.623940950592128e-05, + "loss": 2.4213, + "step": 33824 + }, + { + "epoch": 1.44, + "grad_norm": 0.458984375, + "learning_rate": 3.6234242109039664e-05, + "loss": 2.5533, + "step": 33825 + }, + { + "epoch": 1.44, + "grad_norm": 0.484375, + "learning_rate": 3.622907499908062e-05, + "loss": 2.4334, + "step": 33826 + }, + { + "epoch": 1.44, + "grad_norm": 0.4921875, + "learning_rate": 3.622390817606746e-05, + "loss": 2.3813, + "step": 33827 + }, + { + "epoch": 1.44, + "grad_norm": 0.474609375, + "learning_rate": 3.621874164002338e-05, + "loss": 2.4775, + "step": 33828 + }, + { + "epoch": 1.44, + "grad_norm": 0.48046875, + "learning_rate": 3.621357539097168e-05, + "loss": 2.4567, + "step": 33829 + }, + { + "epoch": 1.44, + "grad_norm": 0.462890625, + "learning_rate": 3.620840942893554e-05, + "loss": 2.3237, + "step": 33830 + }, + { + "epoch": 1.44, + "grad_norm": 0.49609375, + "learning_rate": 3.620324375393829e-05, + "loss": 2.4069, + "step": 33831 + }, + { + "epoch": 1.44, + "grad_norm": 0.42578125, + "learning_rate": 3.6198078366003094e-05, + "loss": 2.4701, + "step": 33832 + }, + { + "epoch": 1.44, + "grad_norm": 0.46875, + "learning_rate": 3.619291326515323e-05, + "loss": 2.4019, + "step": 33833 + }, + { + "epoch": 1.44, + "grad_norm": 0.486328125, + "learning_rate": 3.618774845141196e-05, + "loss": 2.5376, + "step": 33834 + }, + { + "epoch": 1.44, + "grad_norm": 0.5078125, + "learning_rate": 3.6182583924802506e-05, + "loss": 2.4105, + "step": 33835 + }, + { + "epoch": 1.44, + "grad_norm": 0.5, + "learning_rate": 3.617741968534807e-05, + "loss": 2.4334, + "step": 33836 + }, + { + "epoch": 1.44, + "grad_norm": 0.51171875, + "learning_rate": 3.6172255733071924e-05, + "loss": 2.5008, + "step": 33837 + }, + { + "epoch": 1.44, + "grad_norm": 0.498046875, + "learning_rate": 3.616709206799733e-05, + "loss": 2.4681, + "step": 33838 + }, + { + "epoch": 1.44, + "grad_norm": 0.482421875, + "learning_rate": 3.616192869014746e-05, + "loss": 2.4515, + "step": 33839 + }, + { + "epoch": 1.44, + "grad_norm": 0.435546875, + "learning_rate": 3.615676559954562e-05, + "loss": 2.3812, + "step": 33840 + }, + { + "epoch": 1.44, + "grad_norm": 0.482421875, + "learning_rate": 3.6151602796214955e-05, + "loss": 2.5266, + "step": 33841 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.614644028017876e-05, + "loss": 2.5149, + "step": 33842 + }, + { + "epoch": 1.44, + "grad_norm": 0.4921875, + "learning_rate": 3.614127805146027e-05, + "loss": 2.5153, + "step": 33843 + }, + { + "epoch": 1.44, + "grad_norm": 0.453125, + "learning_rate": 3.61361161100827e-05, + "loss": 2.4989, + "step": 33844 + }, + { + "epoch": 1.44, + "grad_norm": 0.458984375, + "learning_rate": 3.613095445606922e-05, + "loss": 2.3294, + "step": 33845 + }, + { + "epoch": 1.44, + "grad_norm": 0.453125, + "learning_rate": 3.6125793089443115e-05, + "loss": 2.4667, + "step": 33846 + }, + { + "epoch": 1.44, + "grad_norm": 0.466796875, + "learning_rate": 3.612063201022763e-05, + "loss": 2.5368, + "step": 33847 + }, + { + "epoch": 1.44, + "grad_norm": 0.453125, + "learning_rate": 3.611547121844593e-05, + "loss": 2.4541, + "step": 33848 + }, + { + "epoch": 1.44, + "grad_norm": 0.484375, + "learning_rate": 3.611031071412125e-05, + "loss": 2.5301, + "step": 33849 + }, + { + "epoch": 1.44, + "grad_norm": 0.4765625, + "learning_rate": 3.6105150497276866e-05, + "loss": 2.3603, + "step": 33850 + }, + { + "epoch": 1.44, + "grad_norm": 0.416015625, + "learning_rate": 3.609999056793594e-05, + "loss": 2.455, + "step": 33851 + }, + { + "epoch": 1.44, + "grad_norm": 0.458984375, + "learning_rate": 3.609483092612168e-05, + "loss": 2.3521, + "step": 33852 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.608967157185732e-05, + "loss": 2.4242, + "step": 33853 + }, + { + "epoch": 1.44, + "grad_norm": 0.453125, + "learning_rate": 3.608451250516612e-05, + "loss": 2.4197, + "step": 33854 + }, + { + "epoch": 1.44, + "grad_norm": 0.48828125, + "learning_rate": 3.607935372607123e-05, + "loss": 2.4728, + "step": 33855 + }, + { + "epoch": 1.44, + "grad_norm": 0.458984375, + "learning_rate": 3.6074195234595906e-05, + "loss": 2.3323, + "step": 33856 + }, + { + "epoch": 1.44, + "grad_norm": 0.51171875, + "learning_rate": 3.606903703076332e-05, + "loss": 2.429, + "step": 33857 + }, + { + "epoch": 1.44, + "grad_norm": 0.447265625, + "learning_rate": 3.60638791145967e-05, + "loss": 2.4238, + "step": 33858 + }, + { + "epoch": 1.44, + "grad_norm": 0.4765625, + "learning_rate": 3.605872148611929e-05, + "loss": 2.475, + "step": 33859 + }, + { + "epoch": 1.44, + "grad_norm": 0.451171875, + "learning_rate": 3.605356414535426e-05, + "loss": 2.3982, + "step": 33860 + }, + { + "epoch": 1.44, + "grad_norm": 0.51171875, + "learning_rate": 3.6048407092324787e-05, + "loss": 2.4913, + "step": 33861 + }, + { + "epoch": 1.44, + "grad_norm": 0.408203125, + "learning_rate": 3.604325032705412e-05, + "loss": 2.279, + "step": 33862 + }, + { + "epoch": 1.44, + "grad_norm": 0.45703125, + "learning_rate": 3.603809384956548e-05, + "loss": 2.4374, + "step": 33863 + }, + { + "epoch": 1.44, + "grad_norm": 0.458984375, + "learning_rate": 3.6032937659882013e-05, + "loss": 2.3634, + "step": 33864 + }, + { + "epoch": 1.44, + "grad_norm": 0.4921875, + "learning_rate": 3.6027781758026944e-05, + "loss": 2.2803, + "step": 33865 + }, + { + "epoch": 1.44, + "grad_norm": 0.4296875, + "learning_rate": 3.602262614402352e-05, + "loss": 2.5363, + "step": 33866 + }, + { + "epoch": 1.44, + "grad_norm": 0.42578125, + "learning_rate": 3.6017470817894895e-05, + "loss": 2.3665, + "step": 33867 + }, + { + "epoch": 1.44, + "grad_norm": 0.42578125, + "learning_rate": 3.601231577966423e-05, + "loss": 2.5303, + "step": 33868 + }, + { + "epoch": 1.44, + "grad_norm": 0.5, + "learning_rate": 3.6007161029354763e-05, + "loss": 2.491, + "step": 33869 + }, + { + "epoch": 1.44, + "grad_norm": 0.431640625, + "learning_rate": 3.600200656698971e-05, + "loss": 2.4425, + "step": 33870 + }, + { + "epoch": 1.44, + "grad_norm": 0.478515625, + "learning_rate": 3.599685239259222e-05, + "loss": 2.4714, + "step": 33871 + }, + { + "epoch": 1.44, + "grad_norm": 0.48828125, + "learning_rate": 3.5991698506185525e-05, + "loss": 2.4465, + "step": 33872 + }, + { + "epoch": 1.44, + "grad_norm": 0.43359375, + "learning_rate": 3.598654490779276e-05, + "loss": 2.4653, + "step": 33873 + }, + { + "epoch": 1.44, + "grad_norm": 0.5078125, + "learning_rate": 3.598139159743716e-05, + "loss": 2.5298, + "step": 33874 + }, + { + "epoch": 1.44, + "grad_norm": 0.455078125, + "learning_rate": 3.597623857514193e-05, + "loss": 2.3741, + "step": 33875 + }, + { + "epoch": 1.44, + "grad_norm": 0.5390625, + "learning_rate": 3.597108584093023e-05, + "loss": 2.4203, + "step": 33876 + }, + { + "epoch": 1.44, + "grad_norm": 0.44140625, + "learning_rate": 3.59659333948252e-05, + "loss": 2.5028, + "step": 33877 + }, + { + "epoch": 1.44, + "grad_norm": 0.55859375, + "learning_rate": 3.596078123685008e-05, + "loss": 2.418, + "step": 33878 + }, + { + "epoch": 1.44, + "grad_norm": 0.482421875, + "learning_rate": 3.595562936702807e-05, + "loss": 2.4338, + "step": 33879 + }, + { + "epoch": 1.44, + "grad_norm": 0.49609375, + "learning_rate": 3.595047778538229e-05, + "loss": 2.3937, + "step": 33880 + }, + { + "epoch": 1.44, + "grad_norm": 0.44140625, + "learning_rate": 3.594532649193595e-05, + "loss": 2.2107, + "step": 33881 + }, + { + "epoch": 1.44, + "grad_norm": 0.484375, + "learning_rate": 3.594017548671227e-05, + "loss": 2.371, + "step": 33882 + }, + { + "epoch": 1.44, + "grad_norm": 0.453125, + "learning_rate": 3.593502476973438e-05, + "loss": 2.3746, + "step": 33883 + }, + { + "epoch": 1.44, + "grad_norm": 0.474609375, + "learning_rate": 3.592987434102544e-05, + "loss": 2.4882, + "step": 33884 + }, + { + "epoch": 1.44, + "grad_norm": 0.51171875, + "learning_rate": 3.592472420060865e-05, + "loss": 2.5965, + "step": 33885 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.591957434850721e-05, + "loss": 2.403, + "step": 33886 + }, + { + "epoch": 1.44, + "grad_norm": 0.57421875, + "learning_rate": 3.591442478474424e-05, + "loss": 2.5489, + "step": 33887 + }, + { + "epoch": 1.44, + "grad_norm": 0.5, + "learning_rate": 3.5909275509342965e-05, + "loss": 2.4539, + "step": 33888 + }, + { + "epoch": 1.44, + "grad_norm": 0.470703125, + "learning_rate": 3.590412652232651e-05, + "loss": 2.4127, + "step": 33889 + }, + { + "epoch": 1.44, + "grad_norm": 0.43359375, + "learning_rate": 3.589897782371805e-05, + "loss": 2.3355, + "step": 33890 + }, + { + "epoch": 1.44, + "grad_norm": 0.47265625, + "learning_rate": 3.589382941354079e-05, + "loss": 2.509, + "step": 33891 + }, + { + "epoch": 1.44, + "grad_norm": 0.431640625, + "learning_rate": 3.588868129181788e-05, + "loss": 2.3859, + "step": 33892 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.5883533458572436e-05, + "loss": 2.615, + "step": 33893 + }, + { + "epoch": 1.44, + "grad_norm": 0.435546875, + "learning_rate": 3.587838591382766e-05, + "loss": 2.5601, + "step": 33894 + }, + { + "epoch": 1.44, + "grad_norm": 0.4453125, + "learning_rate": 3.587323865760676e-05, + "loss": 2.5737, + "step": 33895 + }, + { + "epoch": 1.44, + "grad_norm": 0.458984375, + "learning_rate": 3.586809168993279e-05, + "loss": 2.5023, + "step": 33896 + }, + { + "epoch": 1.44, + "grad_norm": 0.51171875, + "learning_rate": 3.5862945010829e-05, + "loss": 2.4338, + "step": 33897 + }, + { + "epoch": 1.44, + "grad_norm": 0.451171875, + "learning_rate": 3.585779862031853e-05, + "loss": 2.4352, + "step": 33898 + }, + { + "epoch": 1.44, + "grad_norm": 0.435546875, + "learning_rate": 3.5852652518424535e-05, + "loss": 2.3666, + "step": 33899 + }, + { + "epoch": 1.44, + "grad_norm": 0.48828125, + "learning_rate": 3.584750670517012e-05, + "loss": 2.4105, + "step": 33900 + }, + { + "epoch": 1.44, + "grad_norm": 0.486328125, + "learning_rate": 3.584236118057849e-05, + "loss": 2.482, + "step": 33901 + }, + { + "epoch": 1.44, + "grad_norm": 0.46484375, + "learning_rate": 3.583721594467281e-05, + "loss": 2.4255, + "step": 33902 + }, + { + "epoch": 1.44, + "grad_norm": 0.439453125, + "learning_rate": 3.583207099747617e-05, + "loss": 2.4062, + "step": 33903 + }, + { + "epoch": 1.44, + "grad_norm": 0.455078125, + "learning_rate": 3.582692633901181e-05, + "loss": 2.4494, + "step": 33904 + }, + { + "epoch": 1.44, + "grad_norm": 0.494140625, + "learning_rate": 3.5821781969302795e-05, + "loss": 2.4654, + "step": 33905 + }, + { + "epoch": 1.44, + "grad_norm": 0.48046875, + "learning_rate": 3.5816637888372295e-05, + "loss": 2.4187, + "step": 33906 + }, + { + "epoch": 1.44, + "grad_norm": 0.50390625, + "learning_rate": 3.581149409624351e-05, + "loss": 2.463, + "step": 33907 + }, + { + "epoch": 1.44, + "grad_norm": 0.396484375, + "learning_rate": 3.580635059293954e-05, + "loss": 2.5208, + "step": 33908 + }, + { + "epoch": 1.44, + "grad_norm": 0.484375, + "learning_rate": 3.5801207378483504e-05, + "loss": 2.3277, + "step": 33909 + }, + { + "epoch": 1.45, + "grad_norm": 0.515625, + "learning_rate": 3.579606445289857e-05, + "loss": 2.5274, + "step": 33910 + }, + { + "epoch": 1.45, + "grad_norm": 0.431640625, + "learning_rate": 3.5790921816207915e-05, + "loss": 2.4418, + "step": 33911 + }, + { + "epoch": 1.45, + "grad_norm": 0.50390625, + "learning_rate": 3.578577946843461e-05, + "loss": 2.5593, + "step": 33912 + }, + { + "epoch": 1.45, + "grad_norm": 0.484375, + "learning_rate": 3.578063740960184e-05, + "loss": 2.4654, + "step": 33913 + }, + { + "epoch": 1.45, + "grad_norm": 0.4375, + "learning_rate": 3.577549563973275e-05, + "loss": 2.3704, + "step": 33914 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.577035415885046e-05, + "loss": 2.4351, + "step": 33915 + }, + { + "epoch": 1.45, + "grad_norm": 0.458984375, + "learning_rate": 3.576521296697808e-05, + "loss": 2.418, + "step": 33916 + }, + { + "epoch": 1.45, + "grad_norm": 0.455078125, + "learning_rate": 3.576007206413876e-05, + "loss": 2.4889, + "step": 33917 + }, + { + "epoch": 1.45, + "grad_norm": 0.484375, + "learning_rate": 3.575493145035568e-05, + "loss": 2.4396, + "step": 33918 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.5749791125651886e-05, + "loss": 2.5656, + "step": 33919 + }, + { + "epoch": 1.45, + "grad_norm": 0.4609375, + "learning_rate": 3.574465109005058e-05, + "loss": 2.4935, + "step": 33920 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.573951134357484e-05, + "loss": 2.4367, + "step": 33921 + }, + { + "epoch": 1.45, + "grad_norm": 0.48828125, + "learning_rate": 3.573437188624782e-05, + "loss": 2.5073, + "step": 33922 + }, + { + "epoch": 1.45, + "grad_norm": 0.466796875, + "learning_rate": 3.572923271809266e-05, + "loss": 2.4405, + "step": 33923 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.572409383913247e-05, + "loss": 2.3915, + "step": 33924 + }, + { + "epoch": 1.45, + "grad_norm": 0.4140625, + "learning_rate": 3.571895524939033e-05, + "loss": 2.2298, + "step": 33925 + }, + { + "epoch": 1.45, + "grad_norm": 0.455078125, + "learning_rate": 3.57138169488894e-05, + "loss": 2.2995, + "step": 33926 + }, + { + "epoch": 1.45, + "grad_norm": 0.51171875, + "learning_rate": 3.570867893765284e-05, + "loss": 2.674, + "step": 33927 + }, + { + "epoch": 1.45, + "grad_norm": 0.453125, + "learning_rate": 3.5703541215703696e-05, + "loss": 2.4278, + "step": 33928 + }, + { + "epoch": 1.45, + "grad_norm": 0.51953125, + "learning_rate": 3.5698403783065106e-05, + "loss": 2.4694, + "step": 33929 + }, + { + "epoch": 1.45, + "grad_norm": 0.453125, + "learning_rate": 3.569326663976025e-05, + "loss": 2.3772, + "step": 33930 + }, + { + "epoch": 1.45, + "grad_norm": 0.4765625, + "learning_rate": 3.568812978581217e-05, + "loss": 2.5031, + "step": 33931 + }, + { + "epoch": 1.45, + "grad_norm": 0.4609375, + "learning_rate": 3.568299322124399e-05, + "loss": 2.3123, + "step": 33932 + }, + { + "epoch": 1.45, + "grad_norm": 0.458984375, + "learning_rate": 3.5677856946078826e-05, + "loss": 2.3355, + "step": 33933 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.5672720960339836e-05, + "loss": 2.463, + "step": 33934 + }, + { + "epoch": 1.45, + "grad_norm": 0.494140625, + "learning_rate": 3.566758526405006e-05, + "loss": 2.5116, + "step": 33935 + }, + { + "epoch": 1.45, + "grad_norm": 0.5234375, + "learning_rate": 3.5662449857232674e-05, + "loss": 2.4287, + "step": 33936 + }, + { + "epoch": 1.45, + "grad_norm": 0.48828125, + "learning_rate": 3.565731473991071e-05, + "loss": 2.3991, + "step": 33937 + }, + { + "epoch": 1.45, + "grad_norm": 0.42578125, + "learning_rate": 3.565217991210732e-05, + "loss": 2.4641, + "step": 33938 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.564704537384563e-05, + "loss": 2.3836, + "step": 33939 + }, + { + "epoch": 1.45, + "grad_norm": 0.455078125, + "learning_rate": 3.564191112514873e-05, + "loss": 2.4725, + "step": 33940 + }, + { + "epoch": 1.45, + "grad_norm": 0.431640625, + "learning_rate": 3.563677716603967e-05, + "loss": 2.4957, + "step": 33941 + }, + { + "epoch": 1.45, + "grad_norm": 0.5, + "learning_rate": 3.563164349654159e-05, + "loss": 2.5624, + "step": 33942 + }, + { + "epoch": 1.45, + "grad_norm": 0.4375, + "learning_rate": 3.5626510116677626e-05, + "loss": 2.3921, + "step": 33943 + }, + { + "epoch": 1.45, + "grad_norm": 0.51171875, + "learning_rate": 3.5621377026470804e-05, + "loss": 2.6212, + "step": 33944 + }, + { + "epoch": 1.45, + "grad_norm": 0.462890625, + "learning_rate": 3.561624422594426e-05, + "loss": 2.4327, + "step": 33945 + }, + { + "epoch": 1.45, + "grad_norm": 0.462890625, + "learning_rate": 3.5611111715121124e-05, + "loss": 2.449, + "step": 33946 + }, + { + "epoch": 1.45, + "grad_norm": 0.4453125, + "learning_rate": 3.560597949402445e-05, + "loss": 2.4774, + "step": 33947 + }, + { + "epoch": 1.45, + "grad_norm": 0.53125, + "learning_rate": 3.560084756267731e-05, + "loss": 2.501, + "step": 33948 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.559571592110281e-05, + "loss": 2.4553, + "step": 33949 + }, + { + "epoch": 1.45, + "grad_norm": 0.416015625, + "learning_rate": 3.559058456932408e-05, + "loss": 2.4536, + "step": 33950 + }, + { + "epoch": 1.45, + "grad_norm": 0.427734375, + "learning_rate": 3.558545350736416e-05, + "loss": 2.4111, + "step": 33951 + }, + { + "epoch": 1.45, + "grad_norm": 0.59375, + "learning_rate": 3.558032273524618e-05, + "loss": 2.4563, + "step": 33952 + }, + { + "epoch": 1.45, + "grad_norm": 0.458984375, + "learning_rate": 3.557519225299322e-05, + "loss": 2.2621, + "step": 33953 + }, + { + "epoch": 1.45, + "grad_norm": 0.404296875, + "learning_rate": 3.557006206062827e-05, + "loss": 2.404, + "step": 33954 + }, + { + "epoch": 1.45, + "grad_norm": 0.431640625, + "learning_rate": 3.556493215817457e-05, + "loss": 2.5993, + "step": 33955 + }, + { + "epoch": 1.45, + "grad_norm": 0.46484375, + "learning_rate": 3.555980254565512e-05, + "loss": 2.2214, + "step": 33956 + }, + { + "epoch": 1.45, + "grad_norm": 0.48046875, + "learning_rate": 3.555467322309298e-05, + "loss": 2.4062, + "step": 33957 + }, + { + "epoch": 1.45, + "grad_norm": 0.48046875, + "learning_rate": 3.554954419051125e-05, + "loss": 2.3767, + "step": 33958 + }, + { + "epoch": 1.45, + "grad_norm": 0.46484375, + "learning_rate": 3.5544415447933065e-05, + "loss": 2.4409, + "step": 33959 + }, + { + "epoch": 1.45, + "grad_norm": 0.478515625, + "learning_rate": 3.553928699538142e-05, + "loss": 2.4949, + "step": 33960 + }, + { + "epoch": 1.45, + "grad_norm": 0.5, + "learning_rate": 3.553415883287944e-05, + "loss": 2.415, + "step": 33961 + }, + { + "epoch": 1.45, + "grad_norm": 0.5078125, + "learning_rate": 3.552903096045017e-05, + "loss": 2.3559, + "step": 33962 + }, + { + "epoch": 1.45, + "grad_norm": 0.490234375, + "learning_rate": 3.552390337811673e-05, + "loss": 2.6109, + "step": 33963 + }, + { + "epoch": 1.45, + "grad_norm": 0.52734375, + "learning_rate": 3.5518776085902115e-05, + "loss": 2.5124, + "step": 33964 + }, + { + "epoch": 1.45, + "grad_norm": 0.435546875, + "learning_rate": 3.5513649083829495e-05, + "loss": 2.433, + "step": 33965 + }, + { + "epoch": 1.45, + "grad_norm": 0.46484375, + "learning_rate": 3.5508522371921834e-05, + "loss": 2.52, + "step": 33966 + }, + { + "epoch": 1.45, + "grad_norm": 0.43359375, + "learning_rate": 3.550339595020227e-05, + "loss": 2.4311, + "step": 33967 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.549826981869389e-05, + "loss": 2.4278, + "step": 33968 + }, + { + "epoch": 1.45, + "grad_norm": 0.421875, + "learning_rate": 3.549314397741971e-05, + "loss": 2.3757, + "step": 33969 + }, + { + "epoch": 1.45, + "grad_norm": 0.46484375, + "learning_rate": 3.548801842640275e-05, + "loss": 2.3224, + "step": 33970 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.5482893165666186e-05, + "loss": 2.4393, + "step": 33971 + }, + { + "epoch": 1.45, + "grad_norm": 0.47265625, + "learning_rate": 3.547776819523303e-05, + "loss": 2.4607, + "step": 33972 + }, + { + "epoch": 1.45, + "grad_norm": 0.451171875, + "learning_rate": 3.54726435151263e-05, + "loss": 2.3124, + "step": 33973 + }, + { + "epoch": 1.45, + "grad_norm": 0.484375, + "learning_rate": 3.54675191253691e-05, + "loss": 2.622, + "step": 33974 + }, + { + "epoch": 1.45, + "grad_norm": 0.41796875, + "learning_rate": 3.5462395025984505e-05, + "loss": 2.251, + "step": 33975 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.545727121699553e-05, + "loss": 2.4194, + "step": 33976 + }, + { + "epoch": 1.45, + "grad_norm": 0.453125, + "learning_rate": 3.5452147698425266e-05, + "loss": 2.5076, + "step": 33977 + }, + { + "epoch": 1.45, + "grad_norm": 0.48828125, + "learning_rate": 3.544702447029671e-05, + "loss": 2.4663, + "step": 33978 + }, + { + "epoch": 1.45, + "grad_norm": 0.51171875, + "learning_rate": 3.544190153263299e-05, + "loss": 2.4048, + "step": 33979 + }, + { + "epoch": 1.45, + "grad_norm": 0.494140625, + "learning_rate": 3.54367788854571e-05, + "loss": 2.572, + "step": 33980 + }, + { + "epoch": 1.45, + "grad_norm": 0.453125, + "learning_rate": 3.5431656528792126e-05, + "loss": 2.4037, + "step": 33981 + }, + { + "epoch": 1.45, + "grad_norm": 0.388671875, + "learning_rate": 3.5426534462661075e-05, + "loss": 2.3804, + "step": 33982 + }, + { + "epoch": 1.45, + "grad_norm": 0.388671875, + "learning_rate": 3.542141268708702e-05, + "loss": 2.5181, + "step": 33983 + }, + { + "epoch": 1.45, + "grad_norm": 0.3671875, + "learning_rate": 3.541629120209303e-05, + "loss": 2.3673, + "step": 33984 + }, + { + "epoch": 1.45, + "grad_norm": 0.46484375, + "learning_rate": 3.541117000770214e-05, + "loss": 2.5939, + "step": 33985 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.540604910393731e-05, + "loss": 2.5016, + "step": 33986 + }, + { + "epoch": 1.45, + "grad_norm": 0.435546875, + "learning_rate": 3.540092849082173e-05, + "loss": 2.6012, + "step": 33987 + }, + { + "epoch": 1.45, + "grad_norm": 0.52734375, + "learning_rate": 3.539580816837834e-05, + "loss": 2.4393, + "step": 33988 + }, + { + "epoch": 1.45, + "grad_norm": 0.41796875, + "learning_rate": 3.539068813663019e-05, + "loss": 2.4959, + "step": 33989 + }, + { + "epoch": 1.45, + "grad_norm": 0.443359375, + "learning_rate": 3.538556839560032e-05, + "loss": 2.3152, + "step": 33990 + }, + { + "epoch": 1.45, + "grad_norm": 0.546875, + "learning_rate": 3.538044894531182e-05, + "loss": 2.3229, + "step": 33991 + }, + { + "epoch": 1.45, + "grad_norm": 0.546875, + "learning_rate": 3.5375329785787646e-05, + "loss": 2.442, + "step": 33992 + }, + { + "epoch": 1.45, + "grad_norm": 0.435546875, + "learning_rate": 3.53702109170509e-05, + "loss": 2.4839, + "step": 33993 + }, + { + "epoch": 1.45, + "grad_norm": 0.484375, + "learning_rate": 3.5365092339124565e-05, + "loss": 2.5539, + "step": 33994 + }, + { + "epoch": 1.45, + "grad_norm": 0.498046875, + "learning_rate": 3.5359974052031705e-05, + "loss": 2.4057, + "step": 33995 + }, + { + "epoch": 1.45, + "grad_norm": 0.4921875, + "learning_rate": 3.535485605579533e-05, + "loss": 2.4108, + "step": 33996 + }, + { + "epoch": 1.45, + "grad_norm": 0.439453125, + "learning_rate": 3.534973835043849e-05, + "loss": 2.4333, + "step": 33997 + }, + { + "epoch": 1.45, + "grad_norm": 0.4609375, + "learning_rate": 3.5344620935984175e-05, + "loss": 2.4218, + "step": 33998 + }, + { + "epoch": 1.45, + "grad_norm": 0.453125, + "learning_rate": 3.533950381245544e-05, + "loss": 2.4605, + "step": 33999 + }, + { + "epoch": 1.45, + "grad_norm": 0.46484375, + "learning_rate": 3.5334386979875334e-05, + "loss": 2.4278, + "step": 34000 + }, + { + "epoch": 1.45, + "grad_norm": 0.466796875, + "learning_rate": 3.532927043826686e-05, + "loss": 2.3963, + "step": 34001 + }, + { + "epoch": 1.45, + "grad_norm": 0.494140625, + "learning_rate": 3.532415418765297e-05, + "loss": 2.5067, + "step": 34002 + }, + { + "epoch": 1.45, + "grad_norm": 0.40234375, + "learning_rate": 3.531903822805681e-05, + "loss": 2.5428, + "step": 34003 + }, + { + "epoch": 1.45, + "grad_norm": 0.46484375, + "learning_rate": 3.5313922559501346e-05, + "loss": 2.4785, + "step": 34004 + }, + { + "epoch": 1.45, + "grad_norm": 0.53515625, + "learning_rate": 3.5308807182009554e-05, + "loss": 2.4357, + "step": 34005 + }, + { + "epoch": 1.45, + "grad_norm": 0.44921875, + "learning_rate": 3.530369209560448e-05, + "loss": 2.4426, + "step": 34006 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.529857730030919e-05, + "loss": 2.3433, + "step": 34007 + }, + { + "epoch": 1.45, + "grad_norm": 0.431640625, + "learning_rate": 3.529346279614662e-05, + "loss": 2.402, + "step": 34008 + }, + { + "epoch": 1.45, + "grad_norm": 0.48828125, + "learning_rate": 3.528834858313985e-05, + "loss": 2.4186, + "step": 34009 + }, + { + "epoch": 1.45, + "grad_norm": 0.498046875, + "learning_rate": 3.528323466131184e-05, + "loss": 2.5008, + "step": 34010 + }, + { + "epoch": 1.45, + "grad_norm": 0.4609375, + "learning_rate": 3.527812103068565e-05, + "loss": 2.4984, + "step": 34011 + }, + { + "epoch": 1.45, + "grad_norm": 0.44921875, + "learning_rate": 3.5273007691284234e-05, + "loss": 2.539, + "step": 34012 + }, + { + "epoch": 1.45, + "grad_norm": 0.4609375, + "learning_rate": 3.526789464313065e-05, + "loss": 2.5102, + "step": 34013 + }, + { + "epoch": 1.45, + "grad_norm": 0.4765625, + "learning_rate": 3.5262781886247854e-05, + "loss": 2.3465, + "step": 34014 + }, + { + "epoch": 1.45, + "grad_norm": 0.490234375, + "learning_rate": 3.525766942065888e-05, + "loss": 2.5187, + "step": 34015 + }, + { + "epoch": 1.45, + "grad_norm": 0.47265625, + "learning_rate": 3.525255724638678e-05, + "loss": 2.5575, + "step": 34016 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.524744536345449e-05, + "loss": 2.451, + "step": 34017 + }, + { + "epoch": 1.45, + "grad_norm": 0.455078125, + "learning_rate": 3.5242333771884974e-05, + "loss": 2.3448, + "step": 34018 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.523722247170136e-05, + "loss": 2.3813, + "step": 34019 + }, + { + "epoch": 1.45, + "grad_norm": 0.5234375, + "learning_rate": 3.5232111462926566e-05, + "loss": 2.5622, + "step": 34020 + }, + { + "epoch": 1.45, + "grad_norm": 0.5234375, + "learning_rate": 3.522700074558358e-05, + "loss": 2.3866, + "step": 34021 + }, + { + "epoch": 1.45, + "grad_norm": 0.4921875, + "learning_rate": 3.52218903196954e-05, + "loss": 2.4404, + "step": 34022 + }, + { + "epoch": 1.45, + "grad_norm": 0.5, + "learning_rate": 3.521678018528509e-05, + "loss": 2.656, + "step": 34023 + }, + { + "epoch": 1.45, + "grad_norm": 0.4765625, + "learning_rate": 3.521167034237556e-05, + "loss": 2.4325, + "step": 34024 + }, + { + "epoch": 1.45, + "grad_norm": 0.5703125, + "learning_rate": 3.520656079098987e-05, + "loss": 2.5846, + "step": 34025 + }, + { + "epoch": 1.45, + "grad_norm": 0.44921875, + "learning_rate": 3.520145153115094e-05, + "loss": 2.5504, + "step": 34026 + }, + { + "epoch": 1.45, + "grad_norm": 0.40234375, + "learning_rate": 3.519634256288179e-05, + "loss": 2.422, + "step": 34027 + }, + { + "epoch": 1.45, + "grad_norm": 0.4453125, + "learning_rate": 3.519123388620546e-05, + "loss": 2.3759, + "step": 34028 + }, + { + "epoch": 1.45, + "grad_norm": 0.482421875, + "learning_rate": 3.518612550114488e-05, + "loss": 2.5258, + "step": 34029 + }, + { + "epoch": 1.45, + "grad_norm": 0.408203125, + "learning_rate": 3.518101740772303e-05, + "loss": 2.5226, + "step": 34030 + }, + { + "epoch": 1.45, + "grad_norm": 0.453125, + "learning_rate": 3.5175909605962906e-05, + "loss": 2.4639, + "step": 34031 + }, + { + "epoch": 1.45, + "grad_norm": 0.494140625, + "learning_rate": 3.5170802095887536e-05, + "loss": 2.3553, + "step": 34032 + }, + { + "epoch": 1.45, + "grad_norm": 0.4765625, + "learning_rate": 3.516569487751986e-05, + "loss": 2.5194, + "step": 34033 + }, + { + "epoch": 1.45, + "grad_norm": 0.466796875, + "learning_rate": 3.516058795088279e-05, + "loss": 2.3731, + "step": 34034 + }, + { + "epoch": 1.45, + "grad_norm": 0.50390625, + "learning_rate": 3.515548131599945e-05, + "loss": 2.4844, + "step": 34035 + }, + { + "epoch": 1.45, + "grad_norm": 0.48046875, + "learning_rate": 3.515037497289275e-05, + "loss": 2.3477, + "step": 34036 + }, + { + "epoch": 1.45, + "grad_norm": 0.494140625, + "learning_rate": 3.5145268921585626e-05, + "loss": 2.5552, + "step": 34037 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.514016316210109e-05, + "loss": 2.4649, + "step": 34038 + }, + { + "epoch": 1.45, + "grad_norm": 0.453125, + "learning_rate": 3.5135057694462145e-05, + "loss": 2.4444, + "step": 34039 + }, + { + "epoch": 1.45, + "grad_norm": 0.52734375, + "learning_rate": 3.512995251869171e-05, + "loss": 2.4827, + "step": 34040 + }, + { + "epoch": 1.45, + "grad_norm": 0.50390625, + "learning_rate": 3.512484763481282e-05, + "loss": 2.3676, + "step": 34041 + }, + { + "epoch": 1.45, + "grad_norm": 0.42578125, + "learning_rate": 3.511974304284836e-05, + "loss": 2.4526, + "step": 34042 + }, + { + "epoch": 1.45, + "grad_norm": 0.4921875, + "learning_rate": 3.511463874282135e-05, + "loss": 2.4931, + "step": 34043 + }, + { + "epoch": 1.45, + "grad_norm": 0.4765625, + "learning_rate": 3.510953473475479e-05, + "loss": 2.4631, + "step": 34044 + }, + { + "epoch": 1.45, + "grad_norm": 0.4765625, + "learning_rate": 3.51044310186716e-05, + "loss": 2.3312, + "step": 34045 + }, + { + "epoch": 1.45, + "grad_norm": 0.431640625, + "learning_rate": 3.5099327594594725e-05, + "loss": 2.3109, + "step": 34046 + }, + { + "epoch": 1.45, + "grad_norm": 0.51953125, + "learning_rate": 3.5094224462547155e-05, + "loss": 2.4131, + "step": 34047 + }, + { + "epoch": 1.45, + "grad_norm": 0.447265625, + "learning_rate": 3.508912162255189e-05, + "loss": 2.5338, + "step": 34048 + }, + { + "epoch": 1.45, + "grad_norm": 0.462890625, + "learning_rate": 3.508401907463186e-05, + "loss": 2.368, + "step": 34049 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.507891681880995e-05, + "loss": 2.5105, + "step": 34050 + }, + { + "epoch": 1.45, + "grad_norm": 0.451171875, + "learning_rate": 3.507381485510925e-05, + "loss": 2.3981, + "step": 34051 + }, + { + "epoch": 1.45, + "grad_norm": 0.451171875, + "learning_rate": 3.506871318355266e-05, + "loss": 2.4365, + "step": 34052 + }, + { + "epoch": 1.45, + "grad_norm": 0.453125, + "learning_rate": 3.5063611804163096e-05, + "loss": 2.3758, + "step": 34053 + }, + { + "epoch": 1.45, + "grad_norm": 0.451171875, + "learning_rate": 3.5058510716963554e-05, + "loss": 2.4429, + "step": 34054 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.505340992197701e-05, + "loss": 2.4253, + "step": 34055 + }, + { + "epoch": 1.45, + "grad_norm": 0.50390625, + "learning_rate": 3.504830941922635e-05, + "loss": 2.5243, + "step": 34056 + }, + { + "epoch": 1.45, + "grad_norm": 0.50390625, + "learning_rate": 3.5043209208734595e-05, + "loss": 2.4065, + "step": 34057 + }, + { + "epoch": 1.45, + "grad_norm": 0.4609375, + "learning_rate": 3.503810929052462e-05, + "loss": 2.3952, + "step": 34058 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.5033009664619435e-05, + "loss": 2.322, + "step": 34059 + }, + { + "epoch": 1.45, + "grad_norm": 0.484375, + "learning_rate": 3.502791033104197e-05, + "loss": 2.5826, + "step": 34060 + }, + { + "epoch": 1.45, + "grad_norm": 0.44140625, + "learning_rate": 3.5022811289815185e-05, + "loss": 2.5506, + "step": 34061 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.501771254096196e-05, + "loss": 2.4078, + "step": 34062 + }, + { + "epoch": 1.45, + "grad_norm": 0.4453125, + "learning_rate": 3.501261408450527e-05, + "loss": 2.4198, + "step": 34063 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.5007515920468116e-05, + "loss": 2.4636, + "step": 34064 + }, + { + "epoch": 1.45, + "grad_norm": 0.47265625, + "learning_rate": 3.500241804887337e-05, + "loss": 2.4857, + "step": 34065 + }, + { + "epoch": 1.45, + "grad_norm": 0.50390625, + "learning_rate": 3.4997320469743946e-05, + "loss": 2.3483, + "step": 34066 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.4992223183102865e-05, + "loss": 2.4157, + "step": 34067 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.498712618897304e-05, + "loss": 2.5184, + "step": 34068 + }, + { + "epoch": 1.45, + "grad_norm": 0.443359375, + "learning_rate": 3.4982029487377355e-05, + "loss": 2.3364, + "step": 34069 + }, + { + "epoch": 1.45, + "grad_norm": 0.5234375, + "learning_rate": 3.497693307833877e-05, + "loss": 2.4097, + "step": 34070 + }, + { + "epoch": 1.45, + "grad_norm": 0.48046875, + "learning_rate": 3.4971836961880255e-05, + "loss": 2.457, + "step": 34071 + }, + { + "epoch": 1.45, + "grad_norm": 0.4765625, + "learning_rate": 3.496674113802468e-05, + "loss": 2.5192, + "step": 34072 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.4961645606795034e-05, + "loss": 2.4552, + "step": 34073 + }, + { + "epoch": 1.45, + "grad_norm": 0.42578125, + "learning_rate": 3.4956550368214214e-05, + "loss": 2.4344, + "step": 34074 + }, + { + "epoch": 1.45, + "grad_norm": 0.5390625, + "learning_rate": 3.495145542230508e-05, + "loss": 2.5137, + "step": 34075 + }, + { + "epoch": 1.45, + "grad_norm": 0.462890625, + "learning_rate": 3.494636076909069e-05, + "loss": 2.4103, + "step": 34076 + }, + { + "epoch": 1.45, + "grad_norm": 0.41796875, + "learning_rate": 3.49412664085939e-05, + "loss": 2.3996, + "step": 34077 + }, + { + "epoch": 1.45, + "grad_norm": 0.482421875, + "learning_rate": 3.4936172340837604e-05, + "loss": 2.4826, + "step": 34078 + }, + { + "epoch": 1.45, + "grad_norm": 0.404296875, + "learning_rate": 3.4931078565844756e-05, + "loss": 2.4054, + "step": 34079 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.49259850836383e-05, + "loss": 2.3844, + "step": 34080 + }, + { + "epoch": 1.45, + "grad_norm": 0.439453125, + "learning_rate": 3.492089189424112e-05, + "loss": 2.5002, + "step": 34081 + }, + { + "epoch": 1.45, + "grad_norm": 0.53125, + "learning_rate": 3.4915798997676154e-05, + "loss": 2.4575, + "step": 34082 + }, + { + "epoch": 1.45, + "grad_norm": 0.5, + "learning_rate": 3.491070639396627e-05, + "loss": 2.4576, + "step": 34083 + }, + { + "epoch": 1.45, + "grad_norm": 0.421875, + "learning_rate": 3.4905614083134466e-05, + "loss": 2.3382, + "step": 34084 + }, + { + "epoch": 1.45, + "grad_norm": 0.451171875, + "learning_rate": 3.490052206520357e-05, + "loss": 2.3848, + "step": 34085 + }, + { + "epoch": 1.45, + "grad_norm": 0.51953125, + "learning_rate": 3.489543034019657e-05, + "loss": 2.418, + "step": 34086 + }, + { + "epoch": 1.45, + "grad_norm": 0.43359375, + "learning_rate": 3.48903389081363e-05, + "loss": 2.415, + "step": 34087 + }, + { + "epoch": 1.45, + "grad_norm": 0.431640625, + "learning_rate": 3.4885247769045714e-05, + "loss": 2.4161, + "step": 34088 + }, + { + "epoch": 1.45, + "grad_norm": 0.515625, + "learning_rate": 3.488015692294776e-05, + "loss": 2.5561, + "step": 34089 + }, + { + "epoch": 1.45, + "grad_norm": 0.46875, + "learning_rate": 3.4875066369865286e-05, + "loss": 2.5674, + "step": 34090 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.486997610982115e-05, + "loss": 2.3647, + "step": 34091 + }, + { + "epoch": 1.45, + "grad_norm": 0.44140625, + "learning_rate": 3.486488614283839e-05, + "loss": 2.4755, + "step": 34092 + }, + { + "epoch": 1.45, + "grad_norm": 0.42578125, + "learning_rate": 3.4859796468939846e-05, + "loss": 2.5363, + "step": 34093 + }, + { + "epoch": 1.45, + "grad_norm": 0.50390625, + "learning_rate": 3.4854707088148363e-05, + "loss": 2.5288, + "step": 34094 + }, + { + "epoch": 1.45, + "grad_norm": 0.455078125, + "learning_rate": 3.48496180004869e-05, + "loss": 2.5059, + "step": 34095 + }, + { + "epoch": 1.45, + "grad_norm": 0.4921875, + "learning_rate": 3.484452920597837e-05, + "loss": 2.4918, + "step": 34096 + }, + { + "epoch": 1.45, + "grad_norm": 0.5, + "learning_rate": 3.483944070464562e-05, + "loss": 2.4819, + "step": 34097 + }, + { + "epoch": 1.45, + "grad_norm": 0.56640625, + "learning_rate": 3.48343524965116e-05, + "loss": 2.5949, + "step": 34098 + }, + { + "epoch": 1.45, + "grad_norm": 0.5234375, + "learning_rate": 3.482926458159915e-05, + "loss": 2.3908, + "step": 34099 + }, + { + "epoch": 1.45, + "grad_norm": 0.49609375, + "learning_rate": 3.482417695993123e-05, + "loss": 2.4231, + "step": 34100 + }, + { + "epoch": 1.45, + "grad_norm": 0.3984375, + "learning_rate": 3.4819089631530666e-05, + "loss": 2.4928, + "step": 34101 + }, + { + "epoch": 1.45, + "grad_norm": 0.44140625, + "learning_rate": 3.48140025964204e-05, + "loss": 2.4245, + "step": 34102 + }, + { + "epoch": 1.45, + "grad_norm": 0.498046875, + "learning_rate": 3.480891585462327e-05, + "loss": 2.3553, + "step": 34103 + }, + { + "epoch": 1.45, + "grad_norm": 0.466796875, + "learning_rate": 3.480382940616219e-05, + "loss": 2.4117, + "step": 34104 + }, + { + "epoch": 1.45, + "grad_norm": 0.419921875, + "learning_rate": 3.47987432510601e-05, + "loss": 2.4811, + "step": 34105 + }, + { + "epoch": 1.45, + "grad_norm": 0.53515625, + "learning_rate": 3.479365738933982e-05, + "loss": 2.4993, + "step": 34106 + }, + { + "epoch": 1.45, + "grad_norm": 0.4921875, + "learning_rate": 3.4788571821024186e-05, + "loss": 2.4429, + "step": 34107 + }, + { + "epoch": 1.45, + "grad_norm": 0.466796875, + "learning_rate": 3.4783486546136214e-05, + "loss": 2.459, + "step": 34108 + }, + { + "epoch": 1.45, + "grad_norm": 0.49609375, + "learning_rate": 3.477840156469873e-05, + "loss": 2.4933, + "step": 34109 + }, + { + "epoch": 1.45, + "grad_norm": 0.490234375, + "learning_rate": 3.477331687673456e-05, + "loss": 2.6094, + "step": 34110 + }, + { + "epoch": 1.45, + "grad_norm": 0.54296875, + "learning_rate": 3.476823248226663e-05, + "loss": 2.4959, + "step": 34111 + }, + { + "epoch": 1.45, + "grad_norm": 0.419921875, + "learning_rate": 3.4763148381317845e-05, + "loss": 2.482, + "step": 34112 + }, + { + "epoch": 1.45, + "grad_norm": 0.5390625, + "learning_rate": 3.475806457391102e-05, + "loss": 2.4941, + "step": 34113 + }, + { + "epoch": 1.45, + "grad_norm": 0.482421875, + "learning_rate": 3.475298106006909e-05, + "loss": 2.4742, + "step": 34114 + }, + { + "epoch": 1.45, + "grad_norm": 0.490234375, + "learning_rate": 3.474789783981487e-05, + "loss": 2.3291, + "step": 34115 + }, + { + "epoch": 1.45, + "grad_norm": 0.48828125, + "learning_rate": 3.474281491317129e-05, + "loss": 2.3377, + "step": 34116 + }, + { + "epoch": 1.45, + "grad_norm": 0.435546875, + "learning_rate": 3.4737732280161174e-05, + "loss": 2.5014, + "step": 34117 + }, + { + "epoch": 1.45, + "grad_norm": 0.451171875, + "learning_rate": 3.473264994080744e-05, + "loss": 2.3514, + "step": 34118 + }, + { + "epoch": 1.45, + "grad_norm": 0.484375, + "learning_rate": 3.47275678951329e-05, + "loss": 2.4528, + "step": 34119 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.472248614316044e-05, + "loss": 2.5109, + "step": 34120 + }, + { + "epoch": 1.45, + "grad_norm": 0.4296875, + "learning_rate": 3.471740468491297e-05, + "loss": 2.324, + "step": 34121 + }, + { + "epoch": 1.45, + "grad_norm": 0.431640625, + "learning_rate": 3.471232352041333e-05, + "loss": 2.5026, + "step": 34122 + }, + { + "epoch": 1.45, + "grad_norm": 0.416015625, + "learning_rate": 3.470724264968429e-05, + "loss": 2.3962, + "step": 34123 + }, + { + "epoch": 1.45, + "grad_norm": 0.474609375, + "learning_rate": 3.470216207274888e-05, + "loss": 2.5368, + "step": 34124 + }, + { + "epoch": 1.45, + "grad_norm": 0.42578125, + "learning_rate": 3.469708178962987e-05, + "loss": 2.4625, + "step": 34125 + }, + { + "epoch": 1.45, + "grad_norm": 0.51171875, + "learning_rate": 3.469200180035009e-05, + "loss": 2.454, + "step": 34126 + }, + { + "epoch": 1.45, + "grad_norm": 0.5390625, + "learning_rate": 3.468692210493245e-05, + "loss": 2.5797, + "step": 34127 + }, + { + "epoch": 1.45, + "grad_norm": 0.482421875, + "learning_rate": 3.4681842703399814e-05, + "loss": 2.3736, + "step": 34128 + }, + { + "epoch": 1.45, + "grad_norm": 0.4375, + "learning_rate": 3.4676763595774974e-05, + "loss": 2.5497, + "step": 34129 + }, + { + "epoch": 1.45, + "grad_norm": 0.455078125, + "learning_rate": 3.4671684782080874e-05, + "loss": 2.405, + "step": 34130 + }, + { + "epoch": 1.45, + "grad_norm": 0.466796875, + "learning_rate": 3.4666606262340274e-05, + "loss": 2.4027, + "step": 34131 + }, + { + "epoch": 1.45, + "grad_norm": 0.498046875, + "learning_rate": 3.4661528036576106e-05, + "loss": 2.4253, + "step": 34132 + }, + { + "epoch": 1.45, + "grad_norm": 0.45703125, + "learning_rate": 3.4656450104811145e-05, + "loss": 2.4418, + "step": 34133 + }, + { + "epoch": 1.45, + "grad_norm": 0.490234375, + "learning_rate": 3.465137246706832e-05, + "loss": 2.4104, + "step": 34134 + }, + { + "epoch": 1.45, + "grad_norm": 0.48828125, + "learning_rate": 3.4646295123370396e-05, + "loss": 2.4248, + "step": 34135 + }, + { + "epoch": 1.45, + "grad_norm": 0.51171875, + "learning_rate": 3.464121807374027e-05, + "loss": 2.4587, + "step": 34136 + }, + { + "epoch": 1.45, + "grad_norm": 0.427734375, + "learning_rate": 3.46361413182008e-05, + "loss": 2.3926, + "step": 34137 + }, + { + "epoch": 1.45, + "grad_norm": 0.4375, + "learning_rate": 3.463106485677481e-05, + "loss": 2.5189, + "step": 34138 + }, + { + "epoch": 1.45, + "grad_norm": 0.458984375, + "learning_rate": 3.462598868948507e-05, + "loss": 2.3761, + "step": 34139 + }, + { + "epoch": 1.45, + "grad_norm": 0.4609375, + "learning_rate": 3.462091281635456e-05, + "loss": 2.4089, + "step": 34140 + }, + { + "epoch": 1.45, + "grad_norm": 0.435546875, + "learning_rate": 3.461583723740603e-05, + "loss": 2.4889, + "step": 34141 + }, + { + "epoch": 1.45, + "grad_norm": 0.5078125, + "learning_rate": 3.461076195266233e-05, + "loss": 2.4627, + "step": 34142 + }, + { + "epoch": 1.45, + "grad_norm": 0.416015625, + "learning_rate": 3.460568696214629e-05, + "loss": 2.4605, + "step": 34143 + }, + { + "epoch": 1.45, + "grad_norm": 0.4453125, + "learning_rate": 3.4600612265880804e-05, + "loss": 2.551, + "step": 34144 + }, + { + "epoch": 1.46, + "grad_norm": 0.5234375, + "learning_rate": 3.459553786388862e-05, + "loss": 2.4953, + "step": 34145 + }, + { + "epoch": 1.46, + "grad_norm": 0.431640625, + "learning_rate": 3.459046375619265e-05, + "loss": 2.5068, + "step": 34146 + }, + { + "epoch": 1.46, + "grad_norm": 0.466796875, + "learning_rate": 3.458538994281565e-05, + "loss": 2.4277, + "step": 34147 + }, + { + "epoch": 1.46, + "grad_norm": 0.431640625, + "learning_rate": 3.4580316423780526e-05, + "loss": 2.4527, + "step": 34148 + }, + { + "epoch": 1.46, + "grad_norm": 0.55078125, + "learning_rate": 3.457524319911004e-05, + "loss": 2.4033, + "step": 34149 + }, + { + "epoch": 1.46, + "grad_norm": 0.490234375, + "learning_rate": 3.4570170268827076e-05, + "loss": 2.4978, + "step": 34150 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.45650976329544e-05, + "loss": 2.4419, + "step": 34151 + }, + { + "epoch": 1.46, + "grad_norm": 0.4609375, + "learning_rate": 3.4560025291514876e-05, + "loss": 2.5027, + "step": 34152 + }, + { + "epoch": 1.46, + "grad_norm": 0.46484375, + "learning_rate": 3.455495324453136e-05, + "loss": 2.4157, + "step": 34153 + }, + { + "epoch": 1.46, + "grad_norm": 0.48828125, + "learning_rate": 3.454988149202663e-05, + "loss": 2.5245, + "step": 34154 + }, + { + "epoch": 1.46, + "grad_norm": 0.412109375, + "learning_rate": 3.4544810034023446e-05, + "loss": 2.4008, + "step": 34155 + }, + { + "epoch": 1.46, + "grad_norm": 0.462890625, + "learning_rate": 3.453973887054477e-05, + "loss": 2.4935, + "step": 34156 + }, + { + "epoch": 1.46, + "grad_norm": 0.44921875, + "learning_rate": 3.453466800161333e-05, + "loss": 2.4075, + "step": 34157 + }, + { + "epoch": 1.46, + "grad_norm": 0.470703125, + "learning_rate": 3.452959742725195e-05, + "loss": 2.402, + "step": 34158 + }, + { + "epoch": 1.46, + "grad_norm": 0.48046875, + "learning_rate": 3.4524527147483444e-05, + "loss": 2.37, + "step": 34159 + }, + { + "epoch": 1.46, + "grad_norm": 0.51171875, + "learning_rate": 3.451945716233067e-05, + "loss": 2.5913, + "step": 34160 + }, + { + "epoch": 1.46, + "grad_norm": 0.435546875, + "learning_rate": 3.4514387471816366e-05, + "loss": 2.4412, + "step": 34161 + }, + { + "epoch": 1.46, + "grad_norm": 0.58984375, + "learning_rate": 3.450931807596343e-05, + "loss": 2.5629, + "step": 34162 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.45042489747946e-05, + "loss": 2.5386, + "step": 34163 + }, + { + "epoch": 1.46, + "grad_norm": 0.443359375, + "learning_rate": 3.4499180168332737e-05, + "loss": 2.4528, + "step": 34164 + }, + { + "epoch": 1.46, + "grad_norm": 0.474609375, + "learning_rate": 3.449411165660059e-05, + "loss": 2.4829, + "step": 34165 + }, + { + "epoch": 1.46, + "grad_norm": 0.4609375, + "learning_rate": 3.4489043439621035e-05, + "loss": 2.4154, + "step": 34166 + }, + { + "epoch": 1.46, + "grad_norm": 0.4375, + "learning_rate": 3.448397551741682e-05, + "loss": 2.4323, + "step": 34167 + }, + { + "epoch": 1.46, + "grad_norm": 0.44921875, + "learning_rate": 3.447890789001077e-05, + "loss": 2.4083, + "step": 34168 + }, + { + "epoch": 1.46, + "grad_norm": 0.45703125, + "learning_rate": 3.4473840557425706e-05, + "loss": 2.5218, + "step": 34169 + }, + { + "epoch": 1.46, + "grad_norm": 0.455078125, + "learning_rate": 3.4468773519684424e-05, + "loss": 2.3488, + "step": 34170 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.446370677680966e-05, + "loss": 2.4209, + "step": 34171 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.445864032882431e-05, + "loss": 2.4765, + "step": 34172 + }, + { + "epoch": 1.46, + "grad_norm": 0.4296875, + "learning_rate": 3.445357417575114e-05, + "loss": 2.4685, + "step": 34173 + }, + { + "epoch": 1.46, + "grad_norm": 0.435546875, + "learning_rate": 3.444850831761288e-05, + "loss": 2.3443, + "step": 34174 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.4443442754432396e-05, + "loss": 2.4607, + "step": 34175 + }, + { + "epoch": 1.46, + "grad_norm": 0.546875, + "learning_rate": 3.443837748623249e-05, + "loss": 2.5666, + "step": 34176 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.443331251303589e-05, + "loss": 2.4735, + "step": 34177 + }, + { + "epoch": 1.46, + "grad_norm": 0.46875, + "learning_rate": 3.4428247834865446e-05, + "loss": 2.5187, + "step": 34178 + }, + { + "epoch": 1.46, + "grad_norm": 0.4375, + "learning_rate": 3.44231834517439e-05, + "loss": 2.4324, + "step": 34179 + }, + { + "epoch": 1.46, + "grad_norm": 0.482421875, + "learning_rate": 3.441811936369409e-05, + "loss": 2.547, + "step": 34180 + }, + { + "epoch": 1.46, + "grad_norm": 0.396484375, + "learning_rate": 3.441305557073875e-05, + "loss": 2.3793, + "step": 34181 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.440799207290072e-05, + "loss": 2.4362, + "step": 34182 + }, + { + "epoch": 1.46, + "grad_norm": 0.443359375, + "learning_rate": 3.440292887020272e-05, + "loss": 2.3787, + "step": 34183 + }, + { + "epoch": 1.46, + "grad_norm": 0.474609375, + "learning_rate": 3.439786596266756e-05, + "loss": 2.5749, + "step": 34184 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.4392803350318084e-05, + "loss": 2.4512, + "step": 34185 + }, + { + "epoch": 1.46, + "grad_norm": 0.4609375, + "learning_rate": 3.4387741033177e-05, + "loss": 2.5366, + "step": 34186 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.438267901126704e-05, + "loss": 2.3813, + "step": 34187 + }, + { + "epoch": 1.46, + "grad_norm": 0.484375, + "learning_rate": 3.437761728461111e-05, + "loss": 2.5312, + "step": 34188 + }, + { + "epoch": 1.46, + "grad_norm": 0.447265625, + "learning_rate": 3.4372555853231914e-05, + "loss": 2.4803, + "step": 34189 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.4367494717152204e-05, + "loss": 2.4596, + "step": 34190 + }, + { + "epoch": 1.46, + "grad_norm": 0.4765625, + "learning_rate": 3.436243387639479e-05, + "loss": 2.4211, + "step": 34191 + }, + { + "epoch": 1.46, + "grad_norm": 0.470703125, + "learning_rate": 3.435737333098246e-05, + "loss": 2.5198, + "step": 34192 + }, + { + "epoch": 1.46, + "grad_norm": 0.47265625, + "learning_rate": 3.435231308093792e-05, + "loss": 2.4911, + "step": 34193 + }, + { + "epoch": 1.46, + "grad_norm": 0.46875, + "learning_rate": 3.434725312628402e-05, + "loss": 2.474, + "step": 34194 + }, + { + "epoch": 1.46, + "grad_norm": 0.419921875, + "learning_rate": 3.434219346704349e-05, + "loss": 2.4891, + "step": 34195 + }, + { + "epoch": 1.46, + "grad_norm": 0.47265625, + "learning_rate": 3.433713410323905e-05, + "loss": 2.4456, + "step": 34196 + }, + { + "epoch": 1.46, + "grad_norm": 0.494140625, + "learning_rate": 3.433207503489353e-05, + "loss": 2.5489, + "step": 34197 + }, + { + "epoch": 1.46, + "grad_norm": 0.474609375, + "learning_rate": 3.43270162620297e-05, + "loss": 2.387, + "step": 34198 + }, + { + "epoch": 1.46, + "grad_norm": 0.51171875, + "learning_rate": 3.432195778467026e-05, + "loss": 2.4681, + "step": 34199 + }, + { + "epoch": 1.46, + "grad_norm": 0.3828125, + "learning_rate": 3.4316899602838004e-05, + "loss": 2.2882, + "step": 34200 + }, + { + "epoch": 1.46, + "grad_norm": 0.498046875, + "learning_rate": 3.4311841716555726e-05, + "loss": 2.5301, + "step": 34201 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.4306784125846156e-05, + "loss": 2.4647, + "step": 34202 + }, + { + "epoch": 1.46, + "grad_norm": 0.462890625, + "learning_rate": 3.430172683073202e-05, + "loss": 2.4281, + "step": 34203 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.42966698312361e-05, + "loss": 2.3592, + "step": 34204 + }, + { + "epoch": 1.46, + "grad_norm": 0.4921875, + "learning_rate": 3.4291613127381196e-05, + "loss": 2.3998, + "step": 34205 + }, + { + "epoch": 1.46, + "grad_norm": 0.5078125, + "learning_rate": 3.428655671918998e-05, + "loss": 2.4451, + "step": 34206 + }, + { + "epoch": 1.46, + "grad_norm": 0.435546875, + "learning_rate": 3.428150060668527e-05, + "loss": 2.3739, + "step": 34207 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.427644478988976e-05, + "loss": 2.3969, + "step": 34208 + }, + { + "epoch": 1.46, + "grad_norm": 0.376953125, + "learning_rate": 3.4271389268826236e-05, + "loss": 2.4669, + "step": 34209 + }, + { + "epoch": 1.46, + "grad_norm": 0.443359375, + "learning_rate": 3.426633404351747e-05, + "loss": 2.4778, + "step": 34210 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.426127911398618e-05, + "loss": 2.5085, + "step": 34211 + }, + { + "epoch": 1.46, + "grad_norm": 0.431640625, + "learning_rate": 3.425622448025506e-05, + "loss": 2.4253, + "step": 34212 + }, + { + "epoch": 1.46, + "grad_norm": 0.484375, + "learning_rate": 3.425117014234692e-05, + "loss": 2.452, + "step": 34213 + }, + { + "epoch": 1.46, + "grad_norm": 0.484375, + "learning_rate": 3.424611610028452e-05, + "loss": 2.5488, + "step": 34214 + }, + { + "epoch": 1.46, + "grad_norm": 0.423828125, + "learning_rate": 3.4241062354090525e-05, + "loss": 2.4745, + "step": 34215 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.4236008903787716e-05, + "loss": 2.3781, + "step": 34216 + }, + { + "epoch": 1.46, + "grad_norm": 0.47265625, + "learning_rate": 3.4230955749398876e-05, + "loss": 2.5234, + "step": 34217 + }, + { + "epoch": 1.46, + "grad_norm": 0.48828125, + "learning_rate": 3.422590289094669e-05, + "loss": 2.5032, + "step": 34218 + }, + { + "epoch": 1.46, + "grad_norm": 0.427734375, + "learning_rate": 3.422085032845388e-05, + "loss": 2.4234, + "step": 34219 + }, + { + "epoch": 1.46, + "grad_norm": 0.46875, + "learning_rate": 3.4215798061943194e-05, + "loss": 2.4319, + "step": 34220 + }, + { + "epoch": 1.46, + "grad_norm": 0.431640625, + "learning_rate": 3.421074609143742e-05, + "loss": 2.2912, + "step": 34221 + }, + { + "epoch": 1.46, + "grad_norm": 0.447265625, + "learning_rate": 3.42056944169592e-05, + "loss": 2.6133, + "step": 34222 + }, + { + "epoch": 1.46, + "grad_norm": 0.416015625, + "learning_rate": 3.420064303853134e-05, + "loss": 2.4719, + "step": 34223 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.419559195617652e-05, + "loss": 2.4037, + "step": 34224 + }, + { + "epoch": 1.46, + "grad_norm": 0.4609375, + "learning_rate": 3.4190541169917475e-05, + "loss": 2.5513, + "step": 34225 + }, + { + "epoch": 1.46, + "grad_norm": 0.41015625, + "learning_rate": 3.418549067977699e-05, + "loss": 2.3439, + "step": 34226 + }, + { + "epoch": 1.46, + "grad_norm": 0.474609375, + "learning_rate": 3.4180440485777744e-05, + "loss": 2.4652, + "step": 34227 + }, + { + "epoch": 1.46, + "grad_norm": 0.46875, + "learning_rate": 3.4175390587942425e-05, + "loss": 2.5852, + "step": 34228 + }, + { + "epoch": 1.46, + "grad_norm": 0.474609375, + "learning_rate": 3.417034098629378e-05, + "loss": 2.5027, + "step": 34229 + }, + { + "epoch": 1.46, + "grad_norm": 0.47265625, + "learning_rate": 3.4165291680854595e-05, + "loss": 2.2446, + "step": 34230 + }, + { + "epoch": 1.46, + "grad_norm": 0.455078125, + "learning_rate": 3.416024267164749e-05, + "loss": 2.41, + "step": 34231 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.415519395869524e-05, + "loss": 2.6075, + "step": 34232 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.4150145542020584e-05, + "loss": 2.4703, + "step": 34233 + }, + { + "epoch": 1.46, + "grad_norm": 0.474609375, + "learning_rate": 3.41450974216462e-05, + "loss": 2.5619, + "step": 34234 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.414004959759479e-05, + "loss": 2.548, + "step": 34235 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.413500206988908e-05, + "loss": 2.5426, + "step": 34236 + }, + { + "epoch": 1.46, + "grad_norm": 0.373046875, + "learning_rate": 3.4129954838551824e-05, + "loss": 2.4196, + "step": 34237 + }, + { + "epoch": 1.46, + "grad_norm": 0.5078125, + "learning_rate": 3.412490790360566e-05, + "loss": 2.4437, + "step": 34238 + }, + { + "epoch": 1.46, + "grad_norm": 0.396484375, + "learning_rate": 3.411986126507338e-05, + "loss": 2.4535, + "step": 34239 + }, + { + "epoch": 1.46, + "grad_norm": 0.453125, + "learning_rate": 3.411481492297761e-05, + "loss": 2.5043, + "step": 34240 + }, + { + "epoch": 1.46, + "grad_norm": 0.451171875, + "learning_rate": 3.4109768877341096e-05, + "loss": 2.6018, + "step": 34241 + }, + { + "epoch": 1.46, + "grad_norm": 0.466796875, + "learning_rate": 3.410472312818658e-05, + "loss": 2.3178, + "step": 34242 + }, + { + "epoch": 1.46, + "grad_norm": 0.462890625, + "learning_rate": 3.4099677675536726e-05, + "loss": 2.3998, + "step": 34243 + }, + { + "epoch": 1.46, + "grad_norm": 0.482421875, + "learning_rate": 3.4094632519414194e-05, + "loss": 2.4911, + "step": 34244 + }, + { + "epoch": 1.46, + "grad_norm": 0.45703125, + "learning_rate": 3.408958765984174e-05, + "loss": 2.524, + "step": 34245 + }, + { + "epoch": 1.46, + "grad_norm": 0.43359375, + "learning_rate": 3.408454309684208e-05, + "loss": 2.5041, + "step": 34246 + }, + { + "epoch": 1.46, + "grad_norm": 0.46484375, + "learning_rate": 3.407949883043786e-05, + "loss": 2.4784, + "step": 34247 + }, + { + "epoch": 1.46, + "grad_norm": 0.494140625, + "learning_rate": 3.407445486065181e-05, + "loss": 2.4547, + "step": 34248 + }, + { + "epoch": 1.46, + "grad_norm": 0.4296875, + "learning_rate": 3.406941118750664e-05, + "loss": 2.4997, + "step": 34249 + }, + { + "epoch": 1.46, + "grad_norm": 0.4296875, + "learning_rate": 3.406436781102502e-05, + "loss": 2.5648, + "step": 34250 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.405932473122964e-05, + "loss": 2.466, + "step": 34251 + }, + { + "epoch": 1.46, + "grad_norm": 0.48828125, + "learning_rate": 3.4054281948143174e-05, + "loss": 2.3613, + "step": 34252 + }, + { + "epoch": 1.46, + "grad_norm": 0.482421875, + "learning_rate": 3.404923946178838e-05, + "loss": 2.4431, + "step": 34253 + }, + { + "epoch": 1.46, + "grad_norm": 0.482421875, + "learning_rate": 3.4044197272187885e-05, + "loss": 2.4219, + "step": 34254 + }, + { + "epoch": 1.46, + "grad_norm": 0.4140625, + "learning_rate": 3.4039155379364415e-05, + "loss": 2.49, + "step": 34255 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.403411378334062e-05, + "loss": 2.4907, + "step": 34256 + }, + { + "epoch": 1.46, + "grad_norm": 0.490234375, + "learning_rate": 3.4029072484139204e-05, + "loss": 2.3556, + "step": 34257 + }, + { + "epoch": 1.46, + "grad_norm": 0.4921875, + "learning_rate": 3.4024031481782884e-05, + "loss": 2.4018, + "step": 34258 + }, + { + "epoch": 1.46, + "grad_norm": 0.52734375, + "learning_rate": 3.40189907762943e-05, + "loss": 2.5123, + "step": 34259 + }, + { + "epoch": 1.46, + "grad_norm": 0.484375, + "learning_rate": 3.401395036769612e-05, + "loss": 2.5153, + "step": 34260 + }, + { + "epoch": 1.46, + "grad_norm": 0.4609375, + "learning_rate": 3.400891025601105e-05, + "loss": 2.3901, + "step": 34261 + }, + { + "epoch": 1.46, + "grad_norm": 0.50390625, + "learning_rate": 3.4003870441261786e-05, + "loss": 2.3898, + "step": 34262 + }, + { + "epoch": 1.46, + "grad_norm": 0.427734375, + "learning_rate": 3.399883092347097e-05, + "loss": 2.5207, + "step": 34263 + }, + { + "epoch": 1.46, + "grad_norm": 0.38671875, + "learning_rate": 3.399379170266129e-05, + "loss": 2.4472, + "step": 34264 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.398875277885546e-05, + "loss": 2.4678, + "step": 34265 + }, + { + "epoch": 1.46, + "grad_norm": 0.4765625, + "learning_rate": 3.398371415207611e-05, + "loss": 2.508, + "step": 34266 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.397867582234589e-05, + "loss": 2.3843, + "step": 34267 + }, + { + "epoch": 1.46, + "grad_norm": 0.49609375, + "learning_rate": 3.3973637789687506e-05, + "loss": 2.327, + "step": 34268 + }, + { + "epoch": 1.46, + "grad_norm": 0.419921875, + "learning_rate": 3.396860005412364e-05, + "loss": 2.5545, + "step": 34269 + }, + { + "epoch": 1.46, + "grad_norm": 0.515625, + "learning_rate": 3.396356261567693e-05, + "loss": 2.4722, + "step": 34270 + }, + { + "epoch": 1.46, + "grad_norm": 0.47265625, + "learning_rate": 3.3958525474370074e-05, + "loss": 2.3842, + "step": 34271 + }, + { + "epoch": 1.46, + "grad_norm": 0.5078125, + "learning_rate": 3.3953488630225696e-05, + "loss": 2.3548, + "step": 34272 + }, + { + "epoch": 1.46, + "grad_norm": 0.494140625, + "learning_rate": 3.394845208326647e-05, + "loss": 2.4816, + "step": 34273 + }, + { + "epoch": 1.46, + "grad_norm": 0.46875, + "learning_rate": 3.394341583351511e-05, + "loss": 2.462, + "step": 34274 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.393837988099423e-05, + "loss": 2.5197, + "step": 34275 + }, + { + "epoch": 1.46, + "grad_norm": 0.44921875, + "learning_rate": 3.393334422572647e-05, + "loss": 2.5122, + "step": 34276 + }, + { + "epoch": 1.46, + "grad_norm": 0.4609375, + "learning_rate": 3.392830886773453e-05, + "loss": 2.4399, + "step": 34277 + }, + { + "epoch": 1.46, + "grad_norm": 0.40625, + "learning_rate": 3.3923273807041066e-05, + "loss": 2.4238, + "step": 34278 + }, + { + "epoch": 1.46, + "grad_norm": 0.41796875, + "learning_rate": 3.39182390436687e-05, + "loss": 2.3353, + "step": 34279 + }, + { + "epoch": 1.46, + "grad_norm": 0.435546875, + "learning_rate": 3.39132045776401e-05, + "loss": 2.4058, + "step": 34280 + }, + { + "epoch": 1.46, + "grad_norm": 0.474609375, + "learning_rate": 3.3908170408977966e-05, + "loss": 2.4681, + "step": 34281 + }, + { + "epoch": 1.46, + "grad_norm": 0.44140625, + "learning_rate": 3.390313653770491e-05, + "loss": 2.4245, + "step": 34282 + }, + { + "epoch": 1.46, + "grad_norm": 0.42578125, + "learning_rate": 3.3898102963843546e-05, + "loss": 2.396, + "step": 34283 + }, + { + "epoch": 1.46, + "grad_norm": 0.5625, + "learning_rate": 3.389306968741658e-05, + "loss": 2.6288, + "step": 34284 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.388803670844666e-05, + "loss": 2.5695, + "step": 34285 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.3883004026956375e-05, + "loss": 2.3567, + "step": 34286 + }, + { + "epoch": 1.46, + "grad_norm": 0.466796875, + "learning_rate": 3.387797164296845e-05, + "loss": 2.5014, + "step": 34287 + }, + { + "epoch": 1.46, + "grad_norm": 0.48828125, + "learning_rate": 3.3872939556505456e-05, + "loss": 2.3258, + "step": 34288 + }, + { + "epoch": 1.46, + "grad_norm": 0.515625, + "learning_rate": 3.386790776759006e-05, + "loss": 2.5576, + "step": 34289 + }, + { + "epoch": 1.46, + "grad_norm": 0.462890625, + "learning_rate": 3.386287627624495e-05, + "loss": 2.3758, + "step": 34290 + }, + { + "epoch": 1.46, + "grad_norm": 0.486328125, + "learning_rate": 3.3857845082492724e-05, + "loss": 2.5247, + "step": 34291 + }, + { + "epoch": 1.46, + "grad_norm": 0.466796875, + "learning_rate": 3.3852814186355983e-05, + "loss": 2.4139, + "step": 34292 + }, + { + "epoch": 1.46, + "grad_norm": 0.451171875, + "learning_rate": 3.384778358785742e-05, + "loss": 2.6236, + "step": 34293 + }, + { + "epoch": 1.46, + "grad_norm": 0.421875, + "learning_rate": 3.384275328701968e-05, + "loss": 2.4871, + "step": 34294 + }, + { + "epoch": 1.46, + "grad_norm": 0.451171875, + "learning_rate": 3.3837723283865344e-05, + "loss": 2.4417, + "step": 34295 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.383269357841706e-05, + "loss": 2.5204, + "step": 34296 + }, + { + "epoch": 1.46, + "grad_norm": 0.5, + "learning_rate": 3.382766417069752e-05, + "loss": 2.462, + "step": 34297 + }, + { + "epoch": 1.46, + "grad_norm": 0.453125, + "learning_rate": 3.38226350607293e-05, + "loss": 2.4674, + "step": 34298 + }, + { + "epoch": 1.46, + "grad_norm": 0.408203125, + "learning_rate": 3.381760624853501e-05, + "loss": 2.2862, + "step": 34299 + }, + { + "epoch": 1.46, + "grad_norm": 0.4140625, + "learning_rate": 3.38125777341373e-05, + "loss": 2.2699, + "step": 34300 + }, + { + "epoch": 1.46, + "grad_norm": 0.470703125, + "learning_rate": 3.3807549517558834e-05, + "loss": 2.4455, + "step": 34301 + }, + { + "epoch": 1.46, + "grad_norm": 0.474609375, + "learning_rate": 3.380252159882218e-05, + "loss": 2.4494, + "step": 34302 + }, + { + "epoch": 1.46, + "grad_norm": 0.447265625, + "learning_rate": 3.379749397795001e-05, + "loss": 2.3727, + "step": 34303 + }, + { + "epoch": 1.46, + "grad_norm": 0.455078125, + "learning_rate": 3.379246665496489e-05, + "loss": 2.4273, + "step": 34304 + }, + { + "epoch": 1.46, + "grad_norm": 0.53515625, + "learning_rate": 3.3787439629889486e-05, + "loss": 2.5122, + "step": 34305 + }, + { + "epoch": 1.46, + "grad_norm": 0.484375, + "learning_rate": 3.378241290274642e-05, + "loss": 2.5376, + "step": 34306 + }, + { + "epoch": 1.46, + "grad_norm": 0.43359375, + "learning_rate": 3.3777386473558314e-05, + "loss": 2.4162, + "step": 34307 + }, + { + "epoch": 1.46, + "grad_norm": 0.52734375, + "learning_rate": 3.377236034234772e-05, + "loss": 2.5904, + "step": 34308 + }, + { + "epoch": 1.46, + "grad_norm": 0.515625, + "learning_rate": 3.376733450913732e-05, + "loss": 2.5386, + "step": 34309 + }, + { + "epoch": 1.46, + "grad_norm": 0.443359375, + "learning_rate": 3.376230897394972e-05, + "loss": 2.4533, + "step": 34310 + }, + { + "epoch": 1.46, + "grad_norm": 0.455078125, + "learning_rate": 3.3757283736807496e-05, + "loss": 2.4353, + "step": 34311 + }, + { + "epoch": 1.46, + "grad_norm": 0.50390625, + "learning_rate": 3.3752258797733296e-05, + "loss": 2.5637, + "step": 34312 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.374723415674974e-05, + "loss": 2.4673, + "step": 34313 + }, + { + "epoch": 1.46, + "grad_norm": 0.486328125, + "learning_rate": 3.3742209813879436e-05, + "loss": 2.5475, + "step": 34314 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.373718576914493e-05, + "loss": 2.5069, + "step": 34315 + }, + { + "epoch": 1.46, + "grad_norm": 0.490234375, + "learning_rate": 3.37321620225689e-05, + "loss": 2.4365, + "step": 34316 + }, + { + "epoch": 1.46, + "grad_norm": 0.453125, + "learning_rate": 3.372713857417389e-05, + "loss": 2.4173, + "step": 34317 + }, + { + "epoch": 1.46, + "grad_norm": 0.45703125, + "learning_rate": 3.3722115423982546e-05, + "loss": 2.4409, + "step": 34318 + }, + { + "epoch": 1.46, + "grad_norm": 0.419921875, + "learning_rate": 3.3717092572017484e-05, + "loss": 2.4098, + "step": 34319 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.371207001830129e-05, + "loss": 2.4033, + "step": 34320 + }, + { + "epoch": 1.46, + "grad_norm": 0.451171875, + "learning_rate": 3.370704776285648e-05, + "loss": 2.4688, + "step": 34321 + }, + { + "epoch": 1.46, + "grad_norm": 0.4609375, + "learning_rate": 3.370202580570581e-05, + "loss": 2.3772, + "step": 34322 + }, + { + "epoch": 1.46, + "grad_norm": 0.447265625, + "learning_rate": 3.369700414687178e-05, + "loss": 2.3767, + "step": 34323 + }, + { + "epoch": 1.46, + "grad_norm": 0.453125, + "learning_rate": 3.369198278637697e-05, + "loss": 2.4504, + "step": 34324 + }, + { + "epoch": 1.46, + "grad_norm": 0.47265625, + "learning_rate": 3.3686961724244006e-05, + "loss": 2.4481, + "step": 34325 + }, + { + "epoch": 1.46, + "grad_norm": 0.546875, + "learning_rate": 3.3681940960495514e-05, + "loss": 2.4537, + "step": 34326 + }, + { + "epoch": 1.46, + "grad_norm": 0.490234375, + "learning_rate": 3.367692049515402e-05, + "loss": 2.4583, + "step": 34327 + }, + { + "epoch": 1.46, + "grad_norm": 0.494140625, + "learning_rate": 3.3671900328242166e-05, + "loss": 2.3543, + "step": 34328 + }, + { + "epoch": 1.46, + "grad_norm": 0.49609375, + "learning_rate": 3.36668804597825e-05, + "loss": 2.5844, + "step": 34329 + }, + { + "epoch": 1.46, + "grad_norm": 0.423828125, + "learning_rate": 3.3661860889797656e-05, + "loss": 2.4226, + "step": 34330 + }, + { + "epoch": 1.46, + "grad_norm": 0.51953125, + "learning_rate": 3.365684161831018e-05, + "loss": 2.5916, + "step": 34331 + }, + { + "epoch": 1.46, + "grad_norm": 0.49609375, + "learning_rate": 3.365182264534268e-05, + "loss": 2.4453, + "step": 34332 + }, + { + "epoch": 1.46, + "grad_norm": 0.51953125, + "learning_rate": 3.364680397091771e-05, + "loss": 2.5029, + "step": 34333 + }, + { + "epoch": 1.46, + "grad_norm": 0.50390625, + "learning_rate": 3.364178559505786e-05, + "loss": 2.4888, + "step": 34334 + }, + { + "epoch": 1.46, + "grad_norm": 0.447265625, + "learning_rate": 3.363676751778577e-05, + "loss": 2.4418, + "step": 34335 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.363174973912397e-05, + "loss": 2.4406, + "step": 34336 + }, + { + "epoch": 1.46, + "grad_norm": 0.431640625, + "learning_rate": 3.362673225909497e-05, + "loss": 2.4227, + "step": 34337 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.362171507772149e-05, + "loss": 2.4962, + "step": 34338 + }, + { + "epoch": 1.46, + "grad_norm": 0.498046875, + "learning_rate": 3.3616698195026044e-05, + "loss": 2.564, + "step": 34339 + }, + { + "epoch": 1.46, + "grad_norm": 0.53125, + "learning_rate": 3.361168161103114e-05, + "loss": 2.5963, + "step": 34340 + }, + { + "epoch": 1.46, + "grad_norm": 0.53515625, + "learning_rate": 3.360666532575942e-05, + "loss": 2.5176, + "step": 34341 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.360164933923348e-05, + "loss": 2.4159, + "step": 34342 + }, + { + "epoch": 1.46, + "grad_norm": 0.431640625, + "learning_rate": 3.359663365147581e-05, + "loss": 2.5178, + "step": 34343 + }, + { + "epoch": 1.46, + "grad_norm": 0.466796875, + "learning_rate": 3.359161826250905e-05, + "loss": 2.4565, + "step": 34344 + }, + { + "epoch": 1.46, + "grad_norm": 0.46484375, + "learning_rate": 3.358660317235571e-05, + "loss": 2.5399, + "step": 34345 + }, + { + "epoch": 1.46, + "grad_norm": 0.46875, + "learning_rate": 3.358158838103842e-05, + "loss": 2.4889, + "step": 34346 + }, + { + "epoch": 1.46, + "grad_norm": 0.4296875, + "learning_rate": 3.3576573888579675e-05, + "loss": 2.3605, + "step": 34347 + }, + { + "epoch": 1.46, + "grad_norm": 0.49609375, + "learning_rate": 3.357155969500211e-05, + "loss": 2.3269, + "step": 34348 + }, + { + "epoch": 1.46, + "grad_norm": 0.416015625, + "learning_rate": 3.356654580032821e-05, + "loss": 2.3226, + "step": 34349 + }, + { + "epoch": 1.46, + "grad_norm": 0.4453125, + "learning_rate": 3.356153220458058e-05, + "loss": 2.4224, + "step": 34350 + }, + { + "epoch": 1.46, + "grad_norm": 0.515625, + "learning_rate": 3.35565189077818e-05, + "loss": 2.4135, + "step": 34351 + }, + { + "epoch": 1.46, + "grad_norm": 0.466796875, + "learning_rate": 3.35515059099544e-05, + "loss": 2.3649, + "step": 34352 + }, + { + "epoch": 1.46, + "grad_norm": 0.439453125, + "learning_rate": 3.354649321112089e-05, + "loss": 2.4394, + "step": 34353 + }, + { + "epoch": 1.46, + "grad_norm": 0.546875, + "learning_rate": 3.354148081130393e-05, + "loss": 2.5905, + "step": 34354 + }, + { + "epoch": 1.46, + "grad_norm": 0.486328125, + "learning_rate": 3.3536468710526024e-05, + "loss": 2.3544, + "step": 34355 + }, + { + "epoch": 1.46, + "grad_norm": 0.4765625, + "learning_rate": 3.353145690880968e-05, + "loss": 2.5579, + "step": 34356 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.352644540617749e-05, + "loss": 2.4655, + "step": 34357 + }, + { + "epoch": 1.46, + "grad_norm": 0.443359375, + "learning_rate": 3.3521434202652037e-05, + "loss": 2.5316, + "step": 34358 + }, + { + "epoch": 1.46, + "grad_norm": 0.51953125, + "learning_rate": 3.35164232982558e-05, + "loss": 2.5444, + "step": 34359 + }, + { + "epoch": 1.46, + "grad_norm": 0.546875, + "learning_rate": 3.351141269301139e-05, + "loss": 2.4345, + "step": 34360 + }, + { + "epoch": 1.46, + "grad_norm": 0.44921875, + "learning_rate": 3.350640238694129e-05, + "loss": 2.4929, + "step": 34361 + }, + { + "epoch": 1.46, + "grad_norm": 0.486328125, + "learning_rate": 3.350139238006813e-05, + "loss": 2.5154, + "step": 34362 + }, + { + "epoch": 1.46, + "grad_norm": 0.42578125, + "learning_rate": 3.349638267241435e-05, + "loss": 2.3919, + "step": 34363 + }, + { + "epoch": 1.46, + "grad_norm": 0.423828125, + "learning_rate": 3.349137326400259e-05, + "loss": 2.3852, + "step": 34364 + }, + { + "epoch": 1.46, + "grad_norm": 0.455078125, + "learning_rate": 3.34863641548553e-05, + "loss": 2.5337, + "step": 34365 + }, + { + "epoch": 1.46, + "grad_norm": 0.490234375, + "learning_rate": 3.348135534499507e-05, + "loss": 2.4395, + "step": 34366 + }, + { + "epoch": 1.46, + "grad_norm": 0.458984375, + "learning_rate": 3.347634683444446e-05, + "loss": 2.4655, + "step": 34367 + }, + { + "epoch": 1.46, + "grad_norm": 0.55078125, + "learning_rate": 3.347133862322598e-05, + "loss": 2.5038, + "step": 34368 + }, + { + "epoch": 1.46, + "grad_norm": 0.451171875, + "learning_rate": 3.346633071136209e-05, + "loss": 2.281, + "step": 34369 + }, + { + "epoch": 1.46, + "grad_norm": 0.470703125, + "learning_rate": 3.346132309887547e-05, + "loss": 2.3118, + "step": 34370 + }, + { + "epoch": 1.46, + "grad_norm": 0.453125, + "learning_rate": 3.3456315785788584e-05, + "loss": 2.4138, + "step": 34371 + }, + { + "epoch": 1.46, + "grad_norm": 0.49609375, + "learning_rate": 3.345130877212391e-05, + "loss": 2.4119, + "step": 34372 + }, + { + "epoch": 1.46, + "grad_norm": 0.46484375, + "learning_rate": 3.3446302057904044e-05, + "loss": 2.4162, + "step": 34373 + }, + { + "epoch": 1.46, + "grad_norm": 0.51171875, + "learning_rate": 3.3441295643151515e-05, + "loss": 2.379, + "step": 34374 + }, + { + "epoch": 1.46, + "grad_norm": 0.4609375, + "learning_rate": 3.3436289527888795e-05, + "loss": 2.4632, + "step": 34375 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.34312837121385e-05, + "loss": 2.5086, + "step": 34376 + }, + { + "epoch": 1.46, + "grad_norm": 0.486328125, + "learning_rate": 3.3426278195923046e-05, + "loss": 2.4526, + "step": 34377 + }, + { + "epoch": 1.46, + "grad_norm": 0.478515625, + "learning_rate": 3.342127297926505e-05, + "loss": 2.5345, + "step": 34378 + }, + { + "epoch": 1.46, + "grad_norm": 0.43359375, + "learning_rate": 3.341626806218696e-05, + "loss": 2.3415, + "step": 34379 + }, + { + "epoch": 1.47, + "grad_norm": 0.4609375, + "learning_rate": 3.3411263444711374e-05, + "loss": 2.4073, + "step": 34380 + }, + { + "epoch": 1.47, + "grad_norm": 0.44140625, + "learning_rate": 3.3406259126860714e-05, + "loss": 2.5082, + "step": 34381 + }, + { + "epoch": 1.47, + "grad_norm": 0.47265625, + "learning_rate": 3.340125510865757e-05, + "loss": 2.3904, + "step": 34382 + }, + { + "epoch": 1.47, + "grad_norm": 0.484375, + "learning_rate": 3.3396251390124466e-05, + "loss": 2.5069, + "step": 34383 + }, + { + "epoch": 1.47, + "grad_norm": 0.48828125, + "learning_rate": 3.339124797128388e-05, + "loss": 2.4453, + "step": 34384 + }, + { + "epoch": 1.47, + "grad_norm": 0.490234375, + "learning_rate": 3.338624485215828e-05, + "loss": 2.4206, + "step": 34385 + }, + { + "epoch": 1.47, + "grad_norm": 0.46875, + "learning_rate": 3.33812420327703e-05, + "loss": 2.3736, + "step": 34386 + }, + { + "epoch": 1.47, + "grad_norm": 0.453125, + "learning_rate": 3.337623951314238e-05, + "loss": 2.4869, + "step": 34387 + }, + { + "epoch": 1.47, + "grad_norm": 0.46484375, + "learning_rate": 3.3371237293297e-05, + "loss": 2.4742, + "step": 34388 + }, + { + "epoch": 1.47, + "grad_norm": 0.47265625, + "learning_rate": 3.3366235373256705e-05, + "loss": 2.4263, + "step": 34389 + }, + { + "epoch": 1.47, + "grad_norm": 0.4609375, + "learning_rate": 3.336123375304403e-05, + "loss": 2.5094, + "step": 34390 + }, + { + "epoch": 1.47, + "grad_norm": 0.5234375, + "learning_rate": 3.335623243268142e-05, + "loss": 2.3817, + "step": 34391 + }, + { + "epoch": 1.47, + "grad_norm": 0.4765625, + "learning_rate": 3.3351231412191445e-05, + "loss": 2.3752, + "step": 34392 + }, + { + "epoch": 1.47, + "grad_norm": 0.498046875, + "learning_rate": 3.3346230691596524e-05, + "loss": 2.4575, + "step": 34393 + }, + { + "epoch": 1.47, + "grad_norm": 0.455078125, + "learning_rate": 3.3341230270919254e-05, + "loss": 2.4187, + "step": 34394 + }, + { + "epoch": 1.47, + "grad_norm": 0.54296875, + "learning_rate": 3.333623015018205e-05, + "loss": 2.4374, + "step": 34395 + }, + { + "epoch": 1.47, + "grad_norm": 0.482421875, + "learning_rate": 3.333123032940747e-05, + "loss": 2.5002, + "step": 34396 + }, + { + "epoch": 1.47, + "grad_norm": 0.47265625, + "learning_rate": 3.3326230808617965e-05, + "loss": 2.5125, + "step": 34397 + }, + { + "epoch": 1.47, + "grad_norm": 0.443359375, + "learning_rate": 3.3321231587836054e-05, + "loss": 2.5196, + "step": 34398 + }, + { + "epoch": 1.47, + "grad_norm": 0.462890625, + "learning_rate": 3.331623266708426e-05, + "loss": 2.5297, + "step": 34399 + }, + { + "epoch": 1.47, + "grad_norm": 0.48828125, + "learning_rate": 3.3311234046385056e-05, + "loss": 2.4187, + "step": 34400 + }, + { + "epoch": 1.47, + "grad_norm": 0.451171875, + "learning_rate": 3.330623572576086e-05, + "loss": 2.3584, + "step": 34401 + }, + { + "epoch": 1.47, + "grad_norm": 0.423828125, + "learning_rate": 3.330123770523429e-05, + "loss": 2.4104, + "step": 34402 + }, + { + "epoch": 1.47, + "grad_norm": 0.4453125, + "learning_rate": 3.329623998482778e-05, + "loss": 2.5045, + "step": 34403 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.329124256456378e-05, + "loss": 2.4079, + "step": 34404 + }, + { + "epoch": 1.47, + "grad_norm": 0.5078125, + "learning_rate": 3.32862454444648e-05, + "loss": 2.5182, + "step": 34405 + }, + { + "epoch": 1.47, + "grad_norm": 0.466796875, + "learning_rate": 3.328124862455339e-05, + "loss": 2.4784, + "step": 34406 + }, + { + "epoch": 1.47, + "grad_norm": 0.46484375, + "learning_rate": 3.3276252104851926e-05, + "loss": 2.4289, + "step": 34407 + }, + { + "epoch": 1.47, + "grad_norm": 0.4375, + "learning_rate": 3.327125588538298e-05, + "loss": 2.3472, + "step": 34408 + }, + { + "epoch": 1.47, + "grad_norm": 0.5, + "learning_rate": 3.3266259966168956e-05, + "loss": 2.521, + "step": 34409 + }, + { + "epoch": 1.47, + "grad_norm": 0.427734375, + "learning_rate": 3.326126434723241e-05, + "loss": 2.5238, + "step": 34410 + }, + { + "epoch": 1.47, + "grad_norm": 0.51171875, + "learning_rate": 3.325626902859577e-05, + "loss": 2.3445, + "step": 34411 + }, + { + "epoch": 1.47, + "grad_norm": 0.5, + "learning_rate": 3.3251274010281544e-05, + "loss": 2.4994, + "step": 34412 + }, + { + "epoch": 1.47, + "grad_norm": 0.5, + "learning_rate": 3.324627929231216e-05, + "loss": 2.3578, + "step": 34413 + }, + { + "epoch": 1.47, + "grad_norm": 0.48828125, + "learning_rate": 3.324128487471011e-05, + "loss": 2.3858, + "step": 34414 + }, + { + "epoch": 1.47, + "grad_norm": 0.408203125, + "learning_rate": 3.323629075749794e-05, + "loss": 2.4594, + "step": 34415 + }, + { + "epoch": 1.47, + "grad_norm": 0.427734375, + "learning_rate": 3.3231296940698044e-05, + "loss": 2.5417, + "step": 34416 + }, + { + "epoch": 1.47, + "grad_norm": 0.453125, + "learning_rate": 3.322630342433285e-05, + "loss": 2.3896, + "step": 34417 + }, + { + "epoch": 1.47, + "grad_norm": 0.43359375, + "learning_rate": 3.322131020842495e-05, + "loss": 2.4522, + "step": 34418 + }, + { + "epoch": 1.47, + "grad_norm": 0.48046875, + "learning_rate": 3.321631729299675e-05, + "loss": 2.421, + "step": 34419 + }, + { + "epoch": 1.47, + "grad_norm": 0.494140625, + "learning_rate": 3.321132467807069e-05, + "loss": 2.5591, + "step": 34420 + }, + { + "epoch": 1.47, + "grad_norm": 0.50390625, + "learning_rate": 3.320633236366925e-05, + "loss": 2.4037, + "step": 34421 + }, + { + "epoch": 1.47, + "grad_norm": 0.5, + "learning_rate": 3.3201340349814944e-05, + "loss": 2.5516, + "step": 34422 + }, + { + "epoch": 1.47, + "grad_norm": 0.5078125, + "learning_rate": 3.319634863653016e-05, + "loss": 2.49, + "step": 34423 + }, + { + "epoch": 1.47, + "grad_norm": 0.439453125, + "learning_rate": 3.3191357223837426e-05, + "loss": 2.555, + "step": 34424 + }, + { + "epoch": 1.47, + "grad_norm": 0.4296875, + "learning_rate": 3.318636611175914e-05, + "loss": 2.4795, + "step": 34425 + }, + { + "epoch": 1.47, + "grad_norm": 0.4140625, + "learning_rate": 3.318137530031781e-05, + "loss": 2.495, + "step": 34426 + }, + { + "epoch": 1.47, + "grad_norm": 0.451171875, + "learning_rate": 3.317638478953584e-05, + "loss": 2.3501, + "step": 34427 + }, + { + "epoch": 1.47, + "grad_norm": 0.427734375, + "learning_rate": 3.317139457943577e-05, + "loss": 2.5157, + "step": 34428 + }, + { + "epoch": 1.47, + "grad_norm": 0.466796875, + "learning_rate": 3.316640467003995e-05, + "loss": 2.2743, + "step": 34429 + }, + { + "epoch": 1.47, + "grad_norm": 0.421875, + "learning_rate": 3.316141506137088e-05, + "loss": 2.522, + "step": 34430 + }, + { + "epoch": 1.47, + "grad_norm": 0.4921875, + "learning_rate": 3.315642575345106e-05, + "loss": 2.4861, + "step": 34431 + }, + { + "epoch": 1.47, + "grad_norm": 0.4453125, + "learning_rate": 3.31514367463029e-05, + "loss": 2.4342, + "step": 34432 + }, + { + "epoch": 1.47, + "grad_norm": 0.478515625, + "learning_rate": 3.314644803994876e-05, + "loss": 2.4448, + "step": 34433 + }, + { + "epoch": 1.47, + "grad_norm": 0.427734375, + "learning_rate": 3.3141459634411254e-05, + "loss": 2.5612, + "step": 34434 + }, + { + "epoch": 1.47, + "grad_norm": 0.48046875, + "learning_rate": 3.3136471529712746e-05, + "loss": 2.415, + "step": 34435 + }, + { + "epoch": 1.47, + "grad_norm": 0.48046875, + "learning_rate": 3.313148372587563e-05, + "loss": 2.4619, + "step": 34436 + }, + { + "epoch": 1.47, + "grad_norm": 0.474609375, + "learning_rate": 3.3126496222922454e-05, + "loss": 2.4791, + "step": 34437 + }, + { + "epoch": 1.47, + "grad_norm": 0.48046875, + "learning_rate": 3.312150902087556e-05, + "loss": 2.4963, + "step": 34438 + }, + { + "epoch": 1.47, + "grad_norm": 0.5390625, + "learning_rate": 3.311652211975744e-05, + "loss": 2.5475, + "step": 34439 + }, + { + "epoch": 1.47, + "grad_norm": 0.43359375, + "learning_rate": 3.3111535519590554e-05, + "loss": 2.4052, + "step": 34440 + }, + { + "epoch": 1.47, + "grad_norm": 0.5390625, + "learning_rate": 3.310654922039732e-05, + "loss": 2.4916, + "step": 34441 + }, + { + "epoch": 1.47, + "grad_norm": 0.48046875, + "learning_rate": 3.310156322220013e-05, + "loss": 2.4913, + "step": 34442 + }, + { + "epoch": 1.47, + "grad_norm": 0.419921875, + "learning_rate": 3.3096577525021454e-05, + "loss": 2.2968, + "step": 34443 + }, + { + "epoch": 1.47, + "grad_norm": 0.451171875, + "learning_rate": 3.3091592128883766e-05, + "loss": 2.4257, + "step": 34444 + }, + { + "epoch": 1.47, + "grad_norm": 0.48828125, + "learning_rate": 3.308660703380944e-05, + "loss": 2.6463, + "step": 34445 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.308162223982091e-05, + "loss": 2.5029, + "step": 34446 + }, + { + "epoch": 1.47, + "grad_norm": 0.466796875, + "learning_rate": 3.3076637746940665e-05, + "loss": 2.352, + "step": 34447 + }, + { + "epoch": 1.47, + "grad_norm": 0.474609375, + "learning_rate": 3.307165355519108e-05, + "loss": 2.4461, + "step": 34448 + }, + { + "epoch": 1.47, + "grad_norm": 0.44921875, + "learning_rate": 3.3066669664594575e-05, + "loss": 2.4106, + "step": 34449 + }, + { + "epoch": 1.47, + "grad_norm": 0.4296875, + "learning_rate": 3.3061686075173595e-05, + "loss": 2.3604, + "step": 34450 + }, + { + "epoch": 1.47, + "grad_norm": 0.482421875, + "learning_rate": 3.30567027869506e-05, + "loss": 2.4203, + "step": 34451 + }, + { + "epoch": 1.47, + "grad_norm": 0.439453125, + "learning_rate": 3.305171979994793e-05, + "loss": 2.4081, + "step": 34452 + }, + { + "epoch": 1.47, + "grad_norm": 0.400390625, + "learning_rate": 3.3046737114188096e-05, + "loss": 2.369, + "step": 34453 + }, + { + "epoch": 1.47, + "grad_norm": 0.4296875, + "learning_rate": 3.304175472969345e-05, + "loss": 2.4623, + "step": 34454 + }, + { + "epoch": 1.47, + "grad_norm": 0.45703125, + "learning_rate": 3.303677264648644e-05, + "loss": 2.5271, + "step": 34455 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.30317908645895e-05, + "loss": 2.3652, + "step": 34456 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.3026809384025024e-05, + "loss": 2.4376, + "step": 34457 + }, + { + "epoch": 1.47, + "grad_norm": 0.462890625, + "learning_rate": 3.3021828204815405e-05, + "loss": 2.489, + "step": 34458 + }, + { + "epoch": 1.47, + "grad_norm": 0.451171875, + "learning_rate": 3.301684732698308e-05, + "loss": 2.4876, + "step": 34459 + }, + { + "epoch": 1.47, + "grad_norm": 0.466796875, + "learning_rate": 3.30118667505505e-05, + "loss": 2.4813, + "step": 34460 + }, + { + "epoch": 1.47, + "grad_norm": 0.52734375, + "learning_rate": 3.300688647554e-05, + "loss": 2.6016, + "step": 34461 + }, + { + "epoch": 1.47, + "grad_norm": 0.462890625, + "learning_rate": 3.3001906501974035e-05, + "loss": 2.3967, + "step": 34462 + }, + { + "epoch": 1.47, + "grad_norm": 0.416015625, + "learning_rate": 3.299692682987503e-05, + "loss": 2.4422, + "step": 34463 + }, + { + "epoch": 1.47, + "grad_norm": 0.5078125, + "learning_rate": 3.299194745926537e-05, + "loss": 2.4574, + "step": 34464 + }, + { + "epoch": 1.47, + "grad_norm": 0.470703125, + "learning_rate": 3.2986968390167425e-05, + "loss": 2.4894, + "step": 34465 + }, + { + "epoch": 1.47, + "grad_norm": 0.43359375, + "learning_rate": 3.298198962260364e-05, + "loss": 2.5368, + "step": 34466 + }, + { + "epoch": 1.47, + "grad_norm": 0.470703125, + "learning_rate": 3.2977011156596425e-05, + "loss": 2.3377, + "step": 34467 + }, + { + "epoch": 1.47, + "grad_norm": 0.5703125, + "learning_rate": 3.297203299216814e-05, + "loss": 2.571, + "step": 34468 + }, + { + "epoch": 1.47, + "grad_norm": 0.44921875, + "learning_rate": 3.296705512934125e-05, + "loss": 2.4212, + "step": 34469 + }, + { + "epoch": 1.47, + "grad_norm": 0.50390625, + "learning_rate": 3.2962077568138084e-05, + "loss": 2.3588, + "step": 34470 + }, + { + "epoch": 1.47, + "grad_norm": 0.41796875, + "learning_rate": 3.295710030858106e-05, + "loss": 2.4561, + "step": 34471 + }, + { + "epoch": 1.47, + "grad_norm": 0.4765625, + "learning_rate": 3.295212335069262e-05, + "loss": 2.4873, + "step": 34472 + }, + { + "epoch": 1.47, + "grad_norm": 0.404296875, + "learning_rate": 3.294714669449512e-05, + "loss": 2.4518, + "step": 34473 + }, + { + "epoch": 1.47, + "grad_norm": 0.462890625, + "learning_rate": 3.294217034001089e-05, + "loss": 2.5964, + "step": 34474 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.293719428726245e-05, + "loss": 2.4898, + "step": 34475 + }, + { + "epoch": 1.47, + "grad_norm": 0.53125, + "learning_rate": 3.293221853627213e-05, + "loss": 2.6193, + "step": 34476 + }, + { + "epoch": 1.47, + "grad_norm": 0.515625, + "learning_rate": 3.2927243087062285e-05, + "loss": 2.3857, + "step": 34477 + }, + { + "epoch": 1.47, + "grad_norm": 0.451171875, + "learning_rate": 3.292226793965533e-05, + "loss": 2.374, + "step": 34478 + }, + { + "epoch": 1.47, + "grad_norm": 0.435546875, + "learning_rate": 3.2917293094073686e-05, + "loss": 2.3034, + "step": 34479 + }, + { + "epoch": 1.47, + "grad_norm": 0.49609375, + "learning_rate": 3.291231855033972e-05, + "loss": 2.4439, + "step": 34480 + }, + { + "epoch": 1.47, + "grad_norm": 0.451171875, + "learning_rate": 3.2907344308475755e-05, + "loss": 2.4264, + "step": 34481 + }, + { + "epoch": 1.47, + "grad_norm": 0.482421875, + "learning_rate": 3.2902370368504235e-05, + "loss": 2.554, + "step": 34482 + }, + { + "epoch": 1.47, + "grad_norm": 0.45703125, + "learning_rate": 3.289739673044754e-05, + "loss": 2.3835, + "step": 34483 + }, + { + "epoch": 1.47, + "grad_norm": 0.48828125, + "learning_rate": 3.289242339432802e-05, + "loss": 2.5072, + "step": 34484 + }, + { + "epoch": 1.47, + "grad_norm": 0.490234375, + "learning_rate": 3.2887450360168096e-05, + "loss": 2.4296, + "step": 34485 + }, + { + "epoch": 1.47, + "grad_norm": 0.423828125, + "learning_rate": 3.2882477627990084e-05, + "loss": 2.3516, + "step": 34486 + }, + { + "epoch": 1.47, + "grad_norm": 0.494140625, + "learning_rate": 3.2877505197816405e-05, + "loss": 2.5044, + "step": 34487 + }, + { + "epoch": 1.47, + "grad_norm": 0.484375, + "learning_rate": 3.287253306966944e-05, + "loss": 2.3793, + "step": 34488 + }, + { + "epoch": 1.47, + "grad_norm": 0.474609375, + "learning_rate": 3.286756124357155e-05, + "loss": 2.5065, + "step": 34489 + }, + { + "epoch": 1.47, + "grad_norm": 0.478515625, + "learning_rate": 3.286258971954503e-05, + "loss": 2.4109, + "step": 34490 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.28576184976124e-05, + "loss": 2.5903, + "step": 34491 + }, + { + "epoch": 1.47, + "grad_norm": 0.45703125, + "learning_rate": 3.2852647577795935e-05, + "loss": 2.3564, + "step": 34492 + }, + { + "epoch": 1.47, + "grad_norm": 0.390625, + "learning_rate": 3.2847676960118e-05, + "loss": 2.2754, + "step": 34493 + }, + { + "epoch": 1.47, + "grad_norm": 0.4453125, + "learning_rate": 3.284270664460096e-05, + "loss": 2.4542, + "step": 34494 + }, + { + "epoch": 1.47, + "grad_norm": 0.48828125, + "learning_rate": 3.283773663126723e-05, + "loss": 2.4948, + "step": 34495 + }, + { + "epoch": 1.47, + "grad_norm": 0.447265625, + "learning_rate": 3.283276692013916e-05, + "loss": 2.4449, + "step": 34496 + }, + { + "epoch": 1.47, + "grad_norm": 0.486328125, + "learning_rate": 3.2827797511239044e-05, + "loss": 2.5744, + "step": 34497 + }, + { + "epoch": 1.47, + "grad_norm": 0.419921875, + "learning_rate": 3.282282840458929e-05, + "loss": 2.4604, + "step": 34498 + }, + { + "epoch": 1.47, + "grad_norm": 0.4375, + "learning_rate": 3.2817859600212294e-05, + "loss": 2.4078, + "step": 34499 + }, + { + "epoch": 1.47, + "grad_norm": 0.4375, + "learning_rate": 3.281289109813034e-05, + "loss": 2.4684, + "step": 34500 + }, + { + "epoch": 1.47, + "grad_norm": 0.43359375, + "learning_rate": 3.280792289836586e-05, + "loss": 2.4323, + "step": 34501 + }, + { + "epoch": 1.47, + "grad_norm": 0.4609375, + "learning_rate": 3.2802955000941136e-05, + "loss": 2.4057, + "step": 34502 + }, + { + "epoch": 1.47, + "grad_norm": 0.478515625, + "learning_rate": 3.279798740587856e-05, + "loss": 2.4703, + "step": 34503 + }, + { + "epoch": 1.47, + "grad_norm": 0.4609375, + "learning_rate": 3.279302011320051e-05, + "loss": 2.3975, + "step": 34504 + }, + { + "epoch": 1.47, + "grad_norm": 0.45703125, + "learning_rate": 3.2788053122929306e-05, + "loss": 2.5038, + "step": 34505 + }, + { + "epoch": 1.47, + "grad_norm": 0.45703125, + "learning_rate": 3.278308643508724e-05, + "loss": 2.4778, + "step": 34506 + }, + { + "epoch": 1.47, + "grad_norm": 0.4375, + "learning_rate": 3.2778120049696784e-05, + "loss": 2.5472, + "step": 34507 + }, + { + "epoch": 1.47, + "grad_norm": 0.494140625, + "learning_rate": 3.277315396678022e-05, + "loss": 2.5047, + "step": 34508 + }, + { + "epoch": 1.47, + "grad_norm": 0.47265625, + "learning_rate": 3.276818818635986e-05, + "loss": 2.5268, + "step": 34509 + }, + { + "epoch": 1.47, + "grad_norm": 0.443359375, + "learning_rate": 3.2763222708458085e-05, + "loss": 2.398, + "step": 34510 + }, + { + "epoch": 1.47, + "grad_norm": 0.478515625, + "learning_rate": 3.275825753309727e-05, + "loss": 2.4925, + "step": 34511 + }, + { + "epoch": 1.47, + "grad_norm": 0.474609375, + "learning_rate": 3.275329266029972e-05, + "loss": 2.3458, + "step": 34512 + }, + { + "epoch": 1.47, + "grad_norm": 0.48046875, + "learning_rate": 3.2748328090087744e-05, + "loss": 2.4532, + "step": 34513 + }, + { + "epoch": 1.47, + "grad_norm": 0.482421875, + "learning_rate": 3.274336382248372e-05, + "loss": 2.4681, + "step": 34514 + }, + { + "epoch": 1.47, + "grad_norm": 0.515625, + "learning_rate": 3.273839985751e-05, + "loss": 2.4845, + "step": 34515 + }, + { + "epoch": 1.47, + "grad_norm": 0.546875, + "learning_rate": 3.273343619518887e-05, + "loss": 2.5086, + "step": 34516 + }, + { + "epoch": 1.47, + "grad_norm": 0.455078125, + "learning_rate": 3.2728472835542735e-05, + "loss": 2.4132, + "step": 34517 + }, + { + "epoch": 1.47, + "grad_norm": 0.5234375, + "learning_rate": 3.2723509778593855e-05, + "loss": 2.3656, + "step": 34518 + }, + { + "epoch": 1.47, + "grad_norm": 0.412109375, + "learning_rate": 3.271854702436459e-05, + "loss": 2.3159, + "step": 34519 + }, + { + "epoch": 1.47, + "grad_norm": 0.39453125, + "learning_rate": 3.27135845728773e-05, + "loss": 2.3051, + "step": 34520 + }, + { + "epoch": 1.47, + "grad_norm": 0.482421875, + "learning_rate": 3.2708622424154304e-05, + "loss": 2.4581, + "step": 34521 + }, + { + "epoch": 1.47, + "grad_norm": 0.484375, + "learning_rate": 3.270366057821783e-05, + "loss": 2.5302, + "step": 34522 + }, + { + "epoch": 1.47, + "grad_norm": 0.435546875, + "learning_rate": 3.269869903509036e-05, + "loss": 2.4834, + "step": 34523 + }, + { + "epoch": 1.47, + "grad_norm": 0.451171875, + "learning_rate": 3.2693737794794165e-05, + "loss": 2.57, + "step": 34524 + }, + { + "epoch": 1.47, + "grad_norm": 0.515625, + "learning_rate": 3.268877685735151e-05, + "loss": 2.5365, + "step": 34525 + }, + { + "epoch": 1.47, + "grad_norm": 0.5234375, + "learning_rate": 3.268381622278476e-05, + "loss": 2.4411, + "step": 34526 + }, + { + "epoch": 1.47, + "grad_norm": 0.52734375, + "learning_rate": 3.267885589111627e-05, + "loss": 2.5782, + "step": 34527 + }, + { + "epoch": 1.47, + "grad_norm": 0.423828125, + "learning_rate": 3.267389586236829e-05, + "loss": 2.4935, + "step": 34528 + }, + { + "epoch": 1.47, + "grad_norm": 0.46484375, + "learning_rate": 3.26689361365632e-05, + "loss": 2.3802, + "step": 34529 + }, + { + "epoch": 1.47, + "grad_norm": 0.4921875, + "learning_rate": 3.266397671372327e-05, + "loss": 2.3634, + "step": 34530 + }, + { + "epoch": 1.47, + "grad_norm": 0.4765625, + "learning_rate": 3.265901759387085e-05, + "loss": 2.4738, + "step": 34531 + }, + { + "epoch": 1.47, + "grad_norm": 0.515625, + "learning_rate": 3.265405877702823e-05, + "loss": 2.49, + "step": 34532 + }, + { + "epoch": 1.47, + "grad_norm": 0.5, + "learning_rate": 3.264910026321776e-05, + "loss": 2.4798, + "step": 34533 + }, + { + "epoch": 1.47, + "grad_norm": 0.443359375, + "learning_rate": 3.264414205246168e-05, + "loss": 2.4188, + "step": 34534 + }, + { + "epoch": 1.47, + "grad_norm": 0.51953125, + "learning_rate": 3.263918414478235e-05, + "loss": 2.5611, + "step": 34535 + }, + { + "epoch": 1.47, + "grad_norm": 0.515625, + "learning_rate": 3.263422654020212e-05, + "loss": 2.6663, + "step": 34536 + }, + { + "epoch": 1.47, + "grad_norm": 0.4453125, + "learning_rate": 3.262926923874324e-05, + "loss": 2.4064, + "step": 34537 + }, + { + "epoch": 1.47, + "grad_norm": 0.423828125, + "learning_rate": 3.262431224042797e-05, + "loss": 2.5082, + "step": 34538 + }, + { + "epoch": 1.47, + "grad_norm": 0.484375, + "learning_rate": 3.261935554527873e-05, + "loss": 2.5363, + "step": 34539 + }, + { + "epoch": 1.47, + "grad_norm": 0.4296875, + "learning_rate": 3.261439915331776e-05, + "loss": 2.3765, + "step": 34540 + }, + { + "epoch": 1.47, + "grad_norm": 0.421875, + "learning_rate": 3.260944306456735e-05, + "loss": 2.4303, + "step": 34541 + }, + { + "epoch": 1.47, + "grad_norm": 0.490234375, + "learning_rate": 3.26044872790498e-05, + "loss": 2.4586, + "step": 34542 + }, + { + "epoch": 1.47, + "grad_norm": 0.447265625, + "learning_rate": 3.2599531796787464e-05, + "loss": 2.4886, + "step": 34543 + }, + { + "epoch": 1.47, + "grad_norm": 0.4453125, + "learning_rate": 3.259457661780258e-05, + "loss": 2.6146, + "step": 34544 + }, + { + "epoch": 1.47, + "grad_norm": 0.439453125, + "learning_rate": 3.2589621742117496e-05, + "loss": 2.4559, + "step": 34545 + }, + { + "epoch": 1.47, + "grad_norm": 0.498046875, + "learning_rate": 3.2584667169754445e-05, + "loss": 2.4287, + "step": 34546 + }, + { + "epoch": 1.47, + "grad_norm": 0.4296875, + "learning_rate": 3.257971290073578e-05, + "loss": 2.3252, + "step": 34547 + }, + { + "epoch": 1.47, + "grad_norm": 0.453125, + "learning_rate": 3.257475893508375e-05, + "loss": 2.5131, + "step": 34548 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.256980527282069e-05, + "loss": 2.4885, + "step": 34549 + }, + { + "epoch": 1.47, + "grad_norm": 0.45703125, + "learning_rate": 3.256485191396882e-05, + "loss": 2.3372, + "step": 34550 + }, + { + "epoch": 1.47, + "grad_norm": 0.43359375, + "learning_rate": 3.2559898858550475e-05, + "loss": 2.5317, + "step": 34551 + }, + { + "epoch": 1.47, + "grad_norm": 0.45703125, + "learning_rate": 3.255494610658798e-05, + "loss": 2.5137, + "step": 34552 + }, + { + "epoch": 1.47, + "grad_norm": 0.4765625, + "learning_rate": 3.2549993658103576e-05, + "loss": 2.3887, + "step": 34553 + }, + { + "epoch": 1.47, + "grad_norm": 0.447265625, + "learning_rate": 3.254504151311949e-05, + "loss": 2.3962, + "step": 34554 + }, + { + "epoch": 1.47, + "grad_norm": 0.48828125, + "learning_rate": 3.254008967165812e-05, + "loss": 2.595, + "step": 34555 + }, + { + "epoch": 1.47, + "grad_norm": 0.443359375, + "learning_rate": 3.2535138133741695e-05, + "loss": 2.4213, + "step": 34556 + }, + { + "epoch": 1.47, + "grad_norm": 0.4609375, + "learning_rate": 3.253018689939247e-05, + "loss": 2.3895, + "step": 34557 + }, + { + "epoch": 1.47, + "grad_norm": 0.435546875, + "learning_rate": 3.252523596863274e-05, + "loss": 2.3987, + "step": 34558 + }, + { + "epoch": 1.47, + "grad_norm": 0.453125, + "learning_rate": 3.2520285341484816e-05, + "loss": 2.3297, + "step": 34559 + }, + { + "epoch": 1.47, + "grad_norm": 0.478515625, + "learning_rate": 3.2515335017970915e-05, + "loss": 2.4403, + "step": 34560 + }, + { + "epoch": 1.47, + "grad_norm": 0.466796875, + "learning_rate": 3.2510384998113385e-05, + "loss": 2.3725, + "step": 34561 + }, + { + "epoch": 1.47, + "grad_norm": 0.515625, + "learning_rate": 3.250543528193445e-05, + "loss": 2.3992, + "step": 34562 + }, + { + "epoch": 1.47, + "grad_norm": 0.435546875, + "learning_rate": 3.2500485869456366e-05, + "loss": 2.4736, + "step": 34563 + }, + { + "epoch": 1.47, + "grad_norm": 0.4765625, + "learning_rate": 3.249553676070142e-05, + "loss": 2.4917, + "step": 34564 + }, + { + "epoch": 1.47, + "grad_norm": 0.447265625, + "learning_rate": 3.249058795569192e-05, + "loss": 2.362, + "step": 34565 + }, + { + "epoch": 1.47, + "grad_norm": 0.41796875, + "learning_rate": 3.248563945445011e-05, + "loss": 2.3222, + "step": 34566 + }, + { + "epoch": 1.47, + "grad_norm": 0.46875, + "learning_rate": 3.2480691256998176e-05, + "loss": 2.4421, + "step": 34567 + }, + { + "epoch": 1.47, + "grad_norm": 0.498046875, + "learning_rate": 3.247574336335852e-05, + "loss": 2.5317, + "step": 34568 + }, + { + "epoch": 1.47, + "grad_norm": 0.427734375, + "learning_rate": 3.2470795773553355e-05, + "loss": 2.4832, + "step": 34569 + }, + { + "epoch": 1.47, + "grad_norm": 0.4375, + "learning_rate": 3.246584848760488e-05, + "loss": 2.4933, + "step": 34570 + }, + { + "epoch": 1.47, + "grad_norm": 0.453125, + "learning_rate": 3.2460901505535415e-05, + "loss": 2.4167, + "step": 34571 + }, + { + "epoch": 1.47, + "grad_norm": 0.453125, + "learning_rate": 3.2455954827367244e-05, + "loss": 2.5072, + "step": 34572 + }, + { + "epoch": 1.47, + "grad_norm": 0.466796875, + "learning_rate": 3.245100845312256e-05, + "loss": 2.5108, + "step": 34573 + }, + { + "epoch": 1.47, + "grad_norm": 0.46484375, + "learning_rate": 3.244606238282367e-05, + "loss": 2.4758, + "step": 34574 + }, + { + "epoch": 1.47, + "grad_norm": 0.43359375, + "learning_rate": 3.244111661649279e-05, + "loss": 2.4652, + "step": 34575 + }, + { + "epoch": 1.47, + "grad_norm": 0.48828125, + "learning_rate": 3.2436171154152204e-05, + "loss": 2.4343, + "step": 34576 + }, + { + "epoch": 1.47, + "grad_norm": 0.400390625, + "learning_rate": 3.2431225995824175e-05, + "loss": 2.3772, + "step": 34577 + }, + { + "epoch": 1.47, + "grad_norm": 0.43359375, + "learning_rate": 3.2426281141530945e-05, + "loss": 2.3427, + "step": 34578 + }, + { + "epoch": 1.47, + "grad_norm": 0.453125, + "learning_rate": 3.242133659129471e-05, + "loss": 2.4398, + "step": 34579 + }, + { + "epoch": 1.47, + "grad_norm": 0.42578125, + "learning_rate": 3.241639234513777e-05, + "loss": 2.4126, + "step": 34580 + }, + { + "epoch": 1.47, + "grad_norm": 0.5078125, + "learning_rate": 3.24114484030824e-05, + "loss": 2.5198, + "step": 34581 + }, + { + "epoch": 1.47, + "grad_norm": 0.462890625, + "learning_rate": 3.240650476515078e-05, + "loss": 2.5161, + "step": 34582 + }, + { + "epoch": 1.47, + "grad_norm": 0.5234375, + "learning_rate": 3.240156143136518e-05, + "loss": 2.5019, + "step": 34583 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.239661840174789e-05, + "loss": 2.46, + "step": 34584 + }, + { + "epoch": 1.47, + "grad_norm": 0.412109375, + "learning_rate": 3.2391675676321095e-05, + "loss": 2.3504, + "step": 34585 + }, + { + "epoch": 1.47, + "grad_norm": 0.486328125, + "learning_rate": 3.2386733255107035e-05, + "loss": 2.438, + "step": 34586 + }, + { + "epoch": 1.47, + "grad_norm": 0.490234375, + "learning_rate": 3.238179113812796e-05, + "loss": 2.5646, + "step": 34587 + }, + { + "epoch": 1.47, + "grad_norm": 0.4921875, + "learning_rate": 3.2376849325406144e-05, + "loss": 2.4749, + "step": 34588 + }, + { + "epoch": 1.47, + "grad_norm": 0.408203125, + "learning_rate": 3.2371907816963766e-05, + "loss": 2.4104, + "step": 34589 + }, + { + "epoch": 1.47, + "grad_norm": 0.435546875, + "learning_rate": 3.236696661282311e-05, + "loss": 2.4978, + "step": 34590 + }, + { + "epoch": 1.47, + "grad_norm": 0.490234375, + "learning_rate": 3.236202571300637e-05, + "loss": 2.5498, + "step": 34591 + }, + { + "epoch": 1.47, + "grad_norm": 0.43359375, + "learning_rate": 3.235708511753578e-05, + "loss": 2.3583, + "step": 34592 + }, + { + "epoch": 1.47, + "grad_norm": 0.5078125, + "learning_rate": 3.2352144826433626e-05, + "loss": 2.5021, + "step": 34593 + }, + { + "epoch": 1.47, + "grad_norm": 0.490234375, + "learning_rate": 3.23472048397221e-05, + "loss": 2.4585, + "step": 34594 + }, + { + "epoch": 1.47, + "grad_norm": 0.482421875, + "learning_rate": 3.234226515742339e-05, + "loss": 2.5029, + "step": 34595 + }, + { + "epoch": 1.47, + "grad_norm": 0.423828125, + "learning_rate": 3.233732577955977e-05, + "loss": 2.4835, + "step": 34596 + }, + { + "epoch": 1.47, + "grad_norm": 0.427734375, + "learning_rate": 3.233238670615347e-05, + "loss": 2.3547, + "step": 34597 + }, + { + "epoch": 1.47, + "grad_norm": 0.44140625, + "learning_rate": 3.232744793722668e-05, + "loss": 2.4069, + "step": 34598 + }, + { + "epoch": 1.47, + "grad_norm": 0.421875, + "learning_rate": 3.232250947280164e-05, + "loss": 2.3769, + "step": 34599 + }, + { + "epoch": 1.47, + "grad_norm": 0.474609375, + "learning_rate": 3.2317571312900616e-05, + "loss": 2.4843, + "step": 34600 + }, + { + "epoch": 1.47, + "grad_norm": 0.453125, + "learning_rate": 3.231263345754578e-05, + "loss": 2.5911, + "step": 34601 + }, + { + "epoch": 1.47, + "grad_norm": 0.44921875, + "learning_rate": 3.230769590675933e-05, + "loss": 2.3302, + "step": 34602 + }, + { + "epoch": 1.47, + "grad_norm": 0.46875, + "learning_rate": 3.230275866056349e-05, + "loss": 2.534, + "step": 34603 + }, + { + "epoch": 1.47, + "grad_norm": 0.458984375, + "learning_rate": 3.2297821718980544e-05, + "loss": 2.4348, + "step": 34604 + }, + { + "epoch": 1.47, + "grad_norm": 0.447265625, + "learning_rate": 3.229288508203262e-05, + "loss": 2.4959, + "step": 34605 + }, + { + "epoch": 1.47, + "grad_norm": 0.47265625, + "learning_rate": 3.2287948749742e-05, + "loss": 2.4604, + "step": 34606 + }, + { + "epoch": 1.47, + "grad_norm": 0.474609375, + "learning_rate": 3.2283012722130846e-05, + "loss": 2.332, + "step": 34607 + }, + { + "epoch": 1.47, + "grad_norm": 0.474609375, + "learning_rate": 3.2278076999221376e-05, + "loss": 2.4435, + "step": 34608 + }, + { + "epoch": 1.47, + "grad_norm": 0.455078125, + "learning_rate": 3.2273141581035836e-05, + "loss": 2.3035, + "step": 34609 + }, + { + "epoch": 1.47, + "grad_norm": 0.4375, + "learning_rate": 3.226820646759642e-05, + "loss": 2.3494, + "step": 34610 + }, + { + "epoch": 1.47, + "grad_norm": 0.427734375, + "learning_rate": 3.226327165892529e-05, + "loss": 2.464, + "step": 34611 + }, + { + "epoch": 1.47, + "grad_norm": 0.4296875, + "learning_rate": 3.225833715504467e-05, + "loss": 2.3961, + "step": 34612 + }, + { + "epoch": 1.47, + "grad_norm": 0.44140625, + "learning_rate": 3.225340295597682e-05, + "loss": 2.4235, + "step": 34613 + }, + { + "epoch": 1.47, + "grad_norm": 0.44140625, + "learning_rate": 3.224846906174386e-05, + "loss": 2.3084, + "step": 34614 + }, + { + "epoch": 1.48, + "grad_norm": 0.421875, + "learning_rate": 3.224353547236803e-05, + "loss": 2.4044, + "step": 34615 + }, + { + "epoch": 1.48, + "grad_norm": 0.52734375, + "learning_rate": 3.223860218787156e-05, + "loss": 2.6004, + "step": 34616 + }, + { + "epoch": 1.48, + "grad_norm": 0.435546875, + "learning_rate": 3.2233669208276604e-05, + "loss": 2.5094, + "step": 34617 + }, + { + "epoch": 1.48, + "grad_norm": 0.431640625, + "learning_rate": 3.222873653360535e-05, + "loss": 2.5009, + "step": 34618 + }, + { + "epoch": 1.48, + "grad_norm": 0.43359375, + "learning_rate": 3.2223804163880024e-05, + "loss": 2.4825, + "step": 34619 + }, + { + "epoch": 1.48, + "grad_norm": 0.494140625, + "learning_rate": 3.221887209912283e-05, + "loss": 2.5437, + "step": 34620 + }, + { + "epoch": 1.48, + "grad_norm": 0.4453125, + "learning_rate": 3.2213940339355906e-05, + "loss": 2.3737, + "step": 34621 + }, + { + "epoch": 1.48, + "grad_norm": 0.455078125, + "learning_rate": 3.220900888460152e-05, + "loss": 2.4207, + "step": 34622 + }, + { + "epoch": 1.48, + "grad_norm": 0.421875, + "learning_rate": 3.220407773488179e-05, + "loss": 2.4244, + "step": 34623 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.219914689021892e-05, + "loss": 2.3474, + "step": 34624 + }, + { + "epoch": 1.48, + "grad_norm": 0.43359375, + "learning_rate": 3.2194216350635145e-05, + "loss": 2.5027, + "step": 34625 + }, + { + "epoch": 1.48, + "grad_norm": 0.48828125, + "learning_rate": 3.218928611615262e-05, + "loss": 2.537, + "step": 34626 + }, + { + "epoch": 1.48, + "grad_norm": 0.408203125, + "learning_rate": 3.2184356186793495e-05, + "loss": 2.4434, + "step": 34627 + }, + { + "epoch": 1.48, + "grad_norm": 0.419921875, + "learning_rate": 3.217942656257998e-05, + "loss": 2.42, + "step": 34628 + }, + { + "epoch": 1.48, + "grad_norm": 0.546875, + "learning_rate": 3.21744972435343e-05, + "loss": 2.5317, + "step": 34629 + }, + { + "epoch": 1.48, + "grad_norm": 0.478515625, + "learning_rate": 3.216956822967855e-05, + "loss": 2.4597, + "step": 34630 + }, + { + "epoch": 1.48, + "grad_norm": 0.41015625, + "learning_rate": 3.216463952103496e-05, + "loss": 2.4676, + "step": 34631 + }, + { + "epoch": 1.48, + "grad_norm": 0.4765625, + "learning_rate": 3.215971111762574e-05, + "loss": 2.4112, + "step": 34632 + }, + { + "epoch": 1.48, + "grad_norm": 0.470703125, + "learning_rate": 3.215478301947301e-05, + "loss": 2.327, + "step": 34633 + }, + { + "epoch": 1.48, + "grad_norm": 0.470703125, + "learning_rate": 3.2149855226598944e-05, + "loss": 2.3035, + "step": 34634 + }, + { + "epoch": 1.48, + "grad_norm": 0.435546875, + "learning_rate": 3.2144927739025735e-05, + "loss": 2.4244, + "step": 34635 + }, + { + "epoch": 1.48, + "grad_norm": 0.46484375, + "learning_rate": 3.2140000556775574e-05, + "loss": 2.4532, + "step": 34636 + }, + { + "epoch": 1.48, + "grad_norm": 0.466796875, + "learning_rate": 3.2135073679870585e-05, + "loss": 2.4376, + "step": 34637 + }, + { + "epoch": 1.48, + "grad_norm": 0.466796875, + "learning_rate": 3.2130147108332994e-05, + "loss": 2.3888, + "step": 34638 + }, + { + "epoch": 1.48, + "grad_norm": 0.4609375, + "learning_rate": 3.21252208421849e-05, + "loss": 2.4487, + "step": 34639 + }, + { + "epoch": 1.48, + "grad_norm": 0.54296875, + "learning_rate": 3.212029488144852e-05, + "loss": 2.4732, + "step": 34640 + }, + { + "epoch": 1.48, + "grad_norm": 0.50390625, + "learning_rate": 3.211536922614603e-05, + "loss": 2.4406, + "step": 34641 + }, + { + "epoch": 1.48, + "grad_norm": 0.44140625, + "learning_rate": 3.211044387629957e-05, + "loss": 2.4355, + "step": 34642 + }, + { + "epoch": 1.48, + "grad_norm": 0.51171875, + "learning_rate": 3.210551883193126e-05, + "loss": 2.5229, + "step": 34643 + }, + { + "epoch": 1.48, + "grad_norm": 0.53125, + "learning_rate": 3.210059409306332e-05, + "loss": 2.5779, + "step": 34644 + }, + { + "epoch": 1.48, + "grad_norm": 0.46875, + "learning_rate": 3.2095669659717906e-05, + "loss": 2.4354, + "step": 34645 + }, + { + "epoch": 1.48, + "grad_norm": 0.42578125, + "learning_rate": 3.2090745531917145e-05, + "loss": 2.4884, + "step": 34646 + }, + { + "epoch": 1.48, + "grad_norm": 0.46875, + "learning_rate": 3.208582170968321e-05, + "loss": 2.5187, + "step": 34647 + }, + { + "epoch": 1.48, + "grad_norm": 0.498046875, + "learning_rate": 3.208089819303828e-05, + "loss": 2.5416, + "step": 34648 + }, + { + "epoch": 1.48, + "grad_norm": 0.482421875, + "learning_rate": 3.2075974982004494e-05, + "loss": 2.5862, + "step": 34649 + }, + { + "epoch": 1.48, + "grad_norm": 0.466796875, + "learning_rate": 3.207105207660396e-05, + "loss": 2.4298, + "step": 34650 + }, + { + "epoch": 1.48, + "grad_norm": 0.462890625, + "learning_rate": 3.2066129476858875e-05, + "loss": 2.4167, + "step": 34651 + }, + { + "epoch": 1.48, + "grad_norm": 0.50390625, + "learning_rate": 3.206120718279141e-05, + "loss": 2.4955, + "step": 34652 + }, + { + "epoch": 1.48, + "grad_norm": 0.44921875, + "learning_rate": 3.205628519442364e-05, + "loss": 2.3962, + "step": 34653 + }, + { + "epoch": 1.48, + "grad_norm": 0.4296875, + "learning_rate": 3.2051363511777796e-05, + "loss": 2.5544, + "step": 34654 + }, + { + "epoch": 1.48, + "grad_norm": 0.435546875, + "learning_rate": 3.204644213487596e-05, + "loss": 2.3642, + "step": 34655 + }, + { + "epoch": 1.48, + "grad_norm": 0.45703125, + "learning_rate": 3.2041521063740285e-05, + "loss": 2.3744, + "step": 34656 + }, + { + "epoch": 1.48, + "grad_norm": 0.490234375, + "learning_rate": 3.2036600298392974e-05, + "loss": 2.3865, + "step": 34657 + }, + { + "epoch": 1.48, + "grad_norm": 0.494140625, + "learning_rate": 3.203167983885612e-05, + "loss": 2.4825, + "step": 34658 + }, + { + "epoch": 1.48, + "grad_norm": 0.416015625, + "learning_rate": 3.202675968515183e-05, + "loss": 2.4831, + "step": 34659 + }, + { + "epoch": 1.48, + "grad_norm": 0.466796875, + "learning_rate": 3.202183983730229e-05, + "loss": 2.4803, + "step": 34660 + }, + { + "epoch": 1.48, + "grad_norm": 0.46875, + "learning_rate": 3.2016920295329655e-05, + "loss": 2.3907, + "step": 34661 + }, + { + "epoch": 1.48, + "grad_norm": 0.47265625, + "learning_rate": 3.2012001059256e-05, + "loss": 2.4193, + "step": 34662 + }, + { + "epoch": 1.48, + "grad_norm": 0.51953125, + "learning_rate": 3.200708212910349e-05, + "loss": 2.5038, + "step": 34663 + }, + { + "epoch": 1.48, + "grad_norm": 0.4921875, + "learning_rate": 3.20021635048943e-05, + "loss": 2.5376, + "step": 34664 + }, + { + "epoch": 1.48, + "grad_norm": 0.486328125, + "learning_rate": 3.199724518665053e-05, + "loss": 2.4549, + "step": 34665 + }, + { + "epoch": 1.48, + "grad_norm": 0.45703125, + "learning_rate": 3.1992327174394254e-05, + "loss": 2.464, + "step": 34666 + }, + { + "epoch": 1.48, + "grad_norm": 0.5234375, + "learning_rate": 3.1987409468147666e-05, + "loss": 2.477, + "step": 34667 + }, + { + "epoch": 1.48, + "grad_norm": 0.5, + "learning_rate": 3.1982492067932915e-05, + "loss": 2.4251, + "step": 34668 + }, + { + "epoch": 1.48, + "grad_norm": 0.447265625, + "learning_rate": 3.197757497377204e-05, + "loss": 2.4217, + "step": 34669 + }, + { + "epoch": 1.48, + "grad_norm": 0.45703125, + "learning_rate": 3.1972658185687266e-05, + "loss": 2.3503, + "step": 34670 + }, + { + "epoch": 1.48, + "grad_norm": 0.41796875, + "learning_rate": 3.1967741703700635e-05, + "loss": 2.423, + "step": 34671 + }, + { + "epoch": 1.48, + "grad_norm": 0.5, + "learning_rate": 3.196282552783431e-05, + "loss": 2.5529, + "step": 34672 + }, + { + "epoch": 1.48, + "grad_norm": 0.4609375, + "learning_rate": 3.195790965811042e-05, + "loss": 2.5044, + "step": 34673 + }, + { + "epoch": 1.48, + "grad_norm": 0.376953125, + "learning_rate": 3.1952994094551084e-05, + "loss": 2.3781, + "step": 34674 + }, + { + "epoch": 1.48, + "grad_norm": 0.51953125, + "learning_rate": 3.1948078837178376e-05, + "loss": 2.4551, + "step": 34675 + }, + { + "epoch": 1.48, + "grad_norm": 0.478515625, + "learning_rate": 3.194316388601443e-05, + "loss": 2.319, + "step": 34676 + }, + { + "epoch": 1.48, + "grad_norm": 0.482421875, + "learning_rate": 3.193824924108142e-05, + "loss": 2.5022, + "step": 34677 + }, + { + "epoch": 1.48, + "grad_norm": 0.5703125, + "learning_rate": 3.193333490240137e-05, + "loss": 2.4752, + "step": 34678 + }, + { + "epoch": 1.48, + "grad_norm": 0.453125, + "learning_rate": 3.192842086999644e-05, + "loss": 2.4267, + "step": 34679 + }, + { + "epoch": 1.48, + "grad_norm": 0.515625, + "learning_rate": 3.192350714388877e-05, + "loss": 2.5504, + "step": 34680 + }, + { + "epoch": 1.48, + "grad_norm": 0.4921875, + "learning_rate": 3.191859372410045e-05, + "loss": 2.6283, + "step": 34681 + }, + { + "epoch": 1.48, + "grad_norm": 0.421875, + "learning_rate": 3.191368061065354e-05, + "loss": 2.4357, + "step": 34682 + }, + { + "epoch": 1.48, + "grad_norm": 0.5390625, + "learning_rate": 3.1908767803570214e-05, + "loss": 2.5522, + "step": 34683 + }, + { + "epoch": 1.48, + "grad_norm": 0.53515625, + "learning_rate": 3.190385530287252e-05, + "loss": 2.5289, + "step": 34684 + }, + { + "epoch": 1.48, + "grad_norm": 0.51171875, + "learning_rate": 3.189894310858259e-05, + "loss": 2.5035, + "step": 34685 + }, + { + "epoch": 1.48, + "grad_norm": 0.48046875, + "learning_rate": 3.189403122072255e-05, + "loss": 2.4472, + "step": 34686 + }, + { + "epoch": 1.48, + "grad_norm": 0.40625, + "learning_rate": 3.18891196393145e-05, + "loss": 2.3749, + "step": 34687 + }, + { + "epoch": 1.48, + "grad_norm": 0.490234375, + "learning_rate": 3.188420836438044e-05, + "loss": 2.3822, + "step": 34688 + }, + { + "epoch": 1.48, + "grad_norm": 0.484375, + "learning_rate": 3.187929739594262e-05, + "loss": 2.4979, + "step": 34689 + }, + { + "epoch": 1.48, + "grad_norm": 0.49609375, + "learning_rate": 3.187438673402307e-05, + "loss": 2.5002, + "step": 34690 + }, + { + "epoch": 1.48, + "grad_norm": 0.451171875, + "learning_rate": 3.186947637864384e-05, + "loss": 2.4879, + "step": 34691 + }, + { + "epoch": 1.48, + "grad_norm": 0.453125, + "learning_rate": 3.186456632982706e-05, + "loss": 2.4942, + "step": 34692 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.1859656587594876e-05, + "loss": 2.4798, + "step": 34693 + }, + { + "epoch": 1.48, + "grad_norm": 0.490234375, + "learning_rate": 3.185474715196931e-05, + "loss": 2.4024, + "step": 34694 + }, + { + "epoch": 1.48, + "grad_norm": 0.4765625, + "learning_rate": 3.18498380229725e-05, + "loss": 2.3765, + "step": 34695 + }, + { + "epoch": 1.48, + "grad_norm": 0.458984375, + "learning_rate": 3.1844929200626486e-05, + "loss": 2.3556, + "step": 34696 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.18400206849534e-05, + "loss": 2.533, + "step": 34697 + }, + { + "epoch": 1.48, + "grad_norm": 0.44921875, + "learning_rate": 3.183511247597529e-05, + "loss": 2.3933, + "step": 34698 + }, + { + "epoch": 1.48, + "grad_norm": 0.443359375, + "learning_rate": 3.18302045737143e-05, + "loss": 2.4804, + "step": 34699 + }, + { + "epoch": 1.48, + "grad_norm": 0.4453125, + "learning_rate": 3.182529697819243e-05, + "loss": 2.4807, + "step": 34700 + }, + { + "epoch": 1.48, + "grad_norm": 0.515625, + "learning_rate": 3.1820389689431816e-05, + "loss": 2.3424, + "step": 34701 + }, + { + "epoch": 1.48, + "grad_norm": 0.49609375, + "learning_rate": 3.181548270745457e-05, + "loss": 2.5003, + "step": 34702 + }, + { + "epoch": 1.48, + "grad_norm": 0.427734375, + "learning_rate": 3.181057603228273e-05, + "loss": 2.4181, + "step": 34703 + }, + { + "epoch": 1.48, + "grad_norm": 0.4609375, + "learning_rate": 3.180566966393831e-05, + "loss": 2.4401, + "step": 34704 + }, + { + "epoch": 1.48, + "grad_norm": 0.40625, + "learning_rate": 3.180076360244352e-05, + "loss": 2.3311, + "step": 34705 + }, + { + "epoch": 1.48, + "grad_norm": 0.486328125, + "learning_rate": 3.179585784782036e-05, + "loss": 2.5364, + "step": 34706 + }, + { + "epoch": 1.48, + "grad_norm": 0.515625, + "learning_rate": 3.17909524000909e-05, + "loss": 2.3574, + "step": 34707 + }, + { + "epoch": 1.48, + "grad_norm": 0.46484375, + "learning_rate": 3.178604725927722e-05, + "loss": 2.3311, + "step": 34708 + }, + { + "epoch": 1.48, + "grad_norm": 0.4296875, + "learning_rate": 3.178114242540142e-05, + "loss": 2.4336, + "step": 34709 + }, + { + "epoch": 1.48, + "grad_norm": 0.4375, + "learning_rate": 3.177623789848553e-05, + "loss": 2.4716, + "step": 34710 + }, + { + "epoch": 1.48, + "grad_norm": 0.47265625, + "learning_rate": 3.1771333678551666e-05, + "loss": 2.3963, + "step": 34711 + }, + { + "epoch": 1.48, + "grad_norm": 0.48828125, + "learning_rate": 3.176642976562183e-05, + "loss": 2.5211, + "step": 34712 + }, + { + "epoch": 1.48, + "grad_norm": 0.4921875, + "learning_rate": 3.176152615971815e-05, + "loss": 2.4288, + "step": 34713 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.175662286086264e-05, + "loss": 2.514, + "step": 34714 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.1751719869077415e-05, + "loss": 2.4859, + "step": 34715 + }, + { + "epoch": 1.48, + "grad_norm": 0.453125, + "learning_rate": 3.174681718438448e-05, + "loss": 2.4689, + "step": 34716 + }, + { + "epoch": 1.48, + "grad_norm": 0.466796875, + "learning_rate": 3.1741914806805916e-05, + "loss": 2.3645, + "step": 34717 + }, + { + "epoch": 1.48, + "grad_norm": 0.455078125, + "learning_rate": 3.1737012736363835e-05, + "loss": 2.485, + "step": 34718 + }, + { + "epoch": 1.48, + "grad_norm": 0.5, + "learning_rate": 3.173211097308024e-05, + "loss": 2.3644, + "step": 34719 + }, + { + "epoch": 1.48, + "grad_norm": 0.45703125, + "learning_rate": 3.172720951697714e-05, + "loss": 2.5238, + "step": 34720 + }, + { + "epoch": 1.48, + "grad_norm": 0.4140625, + "learning_rate": 3.1722308368076705e-05, + "loss": 2.4689, + "step": 34721 + }, + { + "epoch": 1.48, + "grad_norm": 0.5078125, + "learning_rate": 3.171740752640093e-05, + "loss": 2.3496, + "step": 34722 + }, + { + "epoch": 1.48, + "grad_norm": 0.4765625, + "learning_rate": 3.171250699197183e-05, + "loss": 2.4156, + "step": 34723 + }, + { + "epoch": 1.48, + "grad_norm": 0.4140625, + "learning_rate": 3.1707606764811504e-05, + "loss": 2.4795, + "step": 34724 + }, + { + "epoch": 1.48, + "grad_norm": 0.443359375, + "learning_rate": 3.170270684494201e-05, + "loss": 2.4373, + "step": 34725 + }, + { + "epoch": 1.48, + "grad_norm": 0.47265625, + "learning_rate": 3.1697807232385345e-05, + "loss": 2.6103, + "step": 34726 + }, + { + "epoch": 1.48, + "grad_norm": 0.447265625, + "learning_rate": 3.169290792716363e-05, + "loss": 2.437, + "step": 34727 + }, + { + "epoch": 1.48, + "grad_norm": 0.458984375, + "learning_rate": 3.168800892929882e-05, + "loss": 2.3473, + "step": 34728 + }, + { + "epoch": 1.48, + "grad_norm": 0.4921875, + "learning_rate": 3.168311023881304e-05, + "loss": 2.3653, + "step": 34729 + }, + { + "epoch": 1.48, + "grad_norm": 0.453125, + "learning_rate": 3.167821185572827e-05, + "loss": 2.3938, + "step": 34730 + }, + { + "epoch": 1.48, + "grad_norm": 0.49609375, + "learning_rate": 3.1673313780066607e-05, + "loss": 2.3751, + "step": 34731 + }, + { + "epoch": 1.48, + "grad_norm": 0.458984375, + "learning_rate": 3.166841601185003e-05, + "loss": 2.4563, + "step": 34732 + }, + { + "epoch": 1.48, + "grad_norm": 0.5078125, + "learning_rate": 3.16635185511006e-05, + "loss": 2.4031, + "step": 34733 + }, + { + "epoch": 1.48, + "grad_norm": 0.4921875, + "learning_rate": 3.1658621397840404e-05, + "loss": 2.4112, + "step": 34734 + }, + { + "epoch": 1.48, + "grad_norm": 0.494140625, + "learning_rate": 3.165372455209144e-05, + "loss": 2.5233, + "step": 34735 + }, + { + "epoch": 1.48, + "grad_norm": 0.443359375, + "learning_rate": 3.1648828013875664e-05, + "loss": 2.4529, + "step": 34736 + }, + { + "epoch": 1.48, + "grad_norm": 0.46875, + "learning_rate": 3.164393178321525e-05, + "loss": 2.3836, + "step": 34737 + }, + { + "epoch": 1.48, + "grad_norm": 0.498046875, + "learning_rate": 3.163903586013216e-05, + "loss": 2.5135, + "step": 34738 + }, + { + "epoch": 1.48, + "grad_norm": 0.478515625, + "learning_rate": 3.16341402446484e-05, + "loss": 2.4018, + "step": 34739 + }, + { + "epoch": 1.48, + "grad_norm": 0.55078125, + "learning_rate": 3.162924493678602e-05, + "loss": 2.4985, + "step": 34740 + }, + { + "epoch": 1.48, + "grad_norm": 0.478515625, + "learning_rate": 3.1624349936567076e-05, + "loss": 2.5377, + "step": 34741 + }, + { + "epoch": 1.48, + "grad_norm": 0.5390625, + "learning_rate": 3.161945524401354e-05, + "loss": 2.4962, + "step": 34742 + }, + { + "epoch": 1.48, + "grad_norm": 0.53125, + "learning_rate": 3.16145608591475e-05, + "loss": 2.5143, + "step": 34743 + }, + { + "epoch": 1.48, + "grad_norm": 0.4765625, + "learning_rate": 3.1609666781990924e-05, + "loss": 2.4198, + "step": 34744 + }, + { + "epoch": 1.48, + "grad_norm": 0.462890625, + "learning_rate": 3.160477301256587e-05, + "loss": 2.3602, + "step": 34745 + }, + { + "epoch": 1.48, + "grad_norm": 0.45703125, + "learning_rate": 3.1599879550894316e-05, + "loss": 2.5091, + "step": 34746 + }, + { + "epoch": 1.48, + "grad_norm": 0.484375, + "learning_rate": 3.159498639699834e-05, + "loss": 2.4526, + "step": 34747 + }, + { + "epoch": 1.48, + "grad_norm": 0.443359375, + "learning_rate": 3.159009355089988e-05, + "loss": 2.3935, + "step": 34748 + }, + { + "epoch": 1.48, + "grad_norm": 0.482421875, + "learning_rate": 3.1585201012621014e-05, + "loss": 2.6458, + "step": 34749 + }, + { + "epoch": 1.48, + "grad_norm": 0.458984375, + "learning_rate": 3.158030878218377e-05, + "loss": 2.4644, + "step": 34750 + }, + { + "epoch": 1.48, + "grad_norm": 0.48828125, + "learning_rate": 3.157541685961013e-05, + "loss": 2.4667, + "step": 34751 + }, + { + "epoch": 1.48, + "grad_norm": 0.435546875, + "learning_rate": 3.1570525244922044e-05, + "loss": 2.5696, + "step": 34752 + }, + { + "epoch": 1.48, + "grad_norm": 0.5546875, + "learning_rate": 3.1565633938141646e-05, + "loss": 2.3926, + "step": 34753 + }, + { + "epoch": 1.48, + "grad_norm": 0.4609375, + "learning_rate": 3.156074293929089e-05, + "loss": 2.4825, + "step": 34754 + }, + { + "epoch": 1.48, + "grad_norm": 0.43359375, + "learning_rate": 3.155585224839175e-05, + "loss": 2.4388, + "step": 34755 + }, + { + "epoch": 1.48, + "grad_norm": 0.484375, + "learning_rate": 3.155096186546626e-05, + "loss": 2.2456, + "step": 34756 + }, + { + "epoch": 1.48, + "grad_norm": 0.431640625, + "learning_rate": 3.1546071790536446e-05, + "loss": 2.3754, + "step": 34757 + }, + { + "epoch": 1.48, + "grad_norm": 0.41796875, + "learning_rate": 3.154118202362427e-05, + "loss": 2.3622, + "step": 34758 + }, + { + "epoch": 1.48, + "grad_norm": 0.48828125, + "learning_rate": 3.153629256475179e-05, + "loss": 2.5518, + "step": 34759 + }, + { + "epoch": 1.48, + "grad_norm": 0.50390625, + "learning_rate": 3.153140341394094e-05, + "loss": 2.4728, + "step": 34760 + }, + { + "epoch": 1.48, + "grad_norm": 0.37890625, + "learning_rate": 3.152651457121378e-05, + "loss": 2.4802, + "step": 34761 + }, + { + "epoch": 1.48, + "grad_norm": 0.50390625, + "learning_rate": 3.1521626036592255e-05, + "loss": 2.4459, + "step": 34762 + }, + { + "epoch": 1.48, + "grad_norm": 0.482421875, + "learning_rate": 3.1516737810098405e-05, + "loss": 2.4875, + "step": 34763 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.1511849891754186e-05, + "loss": 2.4891, + "step": 34764 + }, + { + "epoch": 1.48, + "grad_norm": 0.462890625, + "learning_rate": 3.150696228158162e-05, + "loss": 2.5498, + "step": 34765 + }, + { + "epoch": 1.48, + "grad_norm": 0.455078125, + "learning_rate": 3.150207497960271e-05, + "loss": 2.4638, + "step": 34766 + }, + { + "epoch": 1.48, + "grad_norm": 0.46875, + "learning_rate": 3.149718798583944e-05, + "loss": 2.4714, + "step": 34767 + }, + { + "epoch": 1.48, + "grad_norm": 0.5078125, + "learning_rate": 3.149230130031372e-05, + "loss": 2.5014, + "step": 34768 + }, + { + "epoch": 1.48, + "grad_norm": 0.46875, + "learning_rate": 3.1487414923047674e-05, + "loss": 2.5619, + "step": 34769 + }, + { + "epoch": 1.48, + "grad_norm": 0.52734375, + "learning_rate": 3.148252885406322e-05, + "loss": 2.5113, + "step": 34770 + }, + { + "epoch": 1.48, + "grad_norm": 0.48828125, + "learning_rate": 3.1477643093382315e-05, + "loss": 2.517, + "step": 34771 + }, + { + "epoch": 1.48, + "grad_norm": 0.4296875, + "learning_rate": 3.1472757641026976e-05, + "loss": 2.3429, + "step": 34772 + }, + { + "epoch": 1.48, + "grad_norm": 0.4765625, + "learning_rate": 3.1467872497019215e-05, + "loss": 2.3744, + "step": 34773 + }, + { + "epoch": 1.48, + "grad_norm": 0.5234375, + "learning_rate": 3.146298766138094e-05, + "loss": 2.4182, + "step": 34774 + }, + { + "epoch": 1.48, + "grad_norm": 0.486328125, + "learning_rate": 3.145810313413422e-05, + "loss": 2.4834, + "step": 34775 + }, + { + "epoch": 1.48, + "grad_norm": 0.48828125, + "learning_rate": 3.1453218915300945e-05, + "loss": 2.546, + "step": 34776 + }, + { + "epoch": 1.48, + "grad_norm": 0.5234375, + "learning_rate": 3.144833500490317e-05, + "loss": 2.6084, + "step": 34777 + }, + { + "epoch": 1.48, + "grad_norm": 0.515625, + "learning_rate": 3.1443451402962796e-05, + "loss": 2.4801, + "step": 34778 + }, + { + "epoch": 1.48, + "grad_norm": 0.490234375, + "learning_rate": 3.143856810950188e-05, + "loss": 2.6207, + "step": 34779 + }, + { + "epoch": 1.48, + "grad_norm": 0.462890625, + "learning_rate": 3.143368512454231e-05, + "loss": 2.427, + "step": 34780 + }, + { + "epoch": 1.48, + "grad_norm": 0.44140625, + "learning_rate": 3.1428802448106096e-05, + "loss": 2.4307, + "step": 34781 + }, + { + "epoch": 1.48, + "grad_norm": 0.51171875, + "learning_rate": 3.142392008021525e-05, + "loss": 2.4593, + "step": 34782 + }, + { + "epoch": 1.48, + "grad_norm": 0.41796875, + "learning_rate": 3.141903802089171e-05, + "loss": 2.4013, + "step": 34783 + }, + { + "epoch": 1.48, + "grad_norm": 0.443359375, + "learning_rate": 3.141415627015736e-05, + "loss": 2.3541, + "step": 34784 + }, + { + "epoch": 1.48, + "grad_norm": 0.51953125, + "learning_rate": 3.140927482803431e-05, + "loss": 2.4138, + "step": 34785 + }, + { + "epoch": 1.48, + "grad_norm": 0.5078125, + "learning_rate": 3.140439369454445e-05, + "loss": 2.4507, + "step": 34786 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.139951286970972e-05, + "loss": 2.4606, + "step": 34787 + }, + { + "epoch": 1.48, + "grad_norm": 0.4921875, + "learning_rate": 3.139463235355212e-05, + "loss": 2.4912, + "step": 34788 + }, + { + "epoch": 1.48, + "grad_norm": 0.45703125, + "learning_rate": 3.138975214609362e-05, + "loss": 2.4507, + "step": 34789 + }, + { + "epoch": 1.48, + "grad_norm": 0.45703125, + "learning_rate": 3.138487224735615e-05, + "loss": 2.4603, + "step": 34790 + }, + { + "epoch": 1.48, + "grad_norm": 0.44140625, + "learning_rate": 3.137999265736169e-05, + "loss": 2.3238, + "step": 34791 + }, + { + "epoch": 1.48, + "grad_norm": 0.48046875, + "learning_rate": 3.137511337613216e-05, + "loss": 2.3532, + "step": 34792 + }, + { + "epoch": 1.48, + "grad_norm": 0.447265625, + "learning_rate": 3.1370234403689566e-05, + "loss": 2.5071, + "step": 34793 + }, + { + "epoch": 1.48, + "grad_norm": 0.490234375, + "learning_rate": 3.136535574005581e-05, + "loss": 2.4636, + "step": 34794 + }, + { + "epoch": 1.48, + "grad_norm": 0.5546875, + "learning_rate": 3.136047738525291e-05, + "loss": 2.5406, + "step": 34795 + }, + { + "epoch": 1.48, + "grad_norm": 0.443359375, + "learning_rate": 3.135559933930274e-05, + "loss": 2.3577, + "step": 34796 + }, + { + "epoch": 1.48, + "grad_norm": 0.49609375, + "learning_rate": 3.135072160222728e-05, + "loss": 2.5149, + "step": 34797 + }, + { + "epoch": 1.48, + "grad_norm": 0.4453125, + "learning_rate": 3.134584417404852e-05, + "loss": 2.4055, + "step": 34798 + }, + { + "epoch": 1.48, + "grad_norm": 0.50390625, + "learning_rate": 3.1340967054788375e-05, + "loss": 2.264, + "step": 34799 + }, + { + "epoch": 1.48, + "grad_norm": 0.427734375, + "learning_rate": 3.133609024446872e-05, + "loss": 2.2139, + "step": 34800 + }, + { + "epoch": 1.48, + "grad_norm": 0.490234375, + "learning_rate": 3.133121374311162e-05, + "loss": 2.4342, + "step": 34801 + }, + { + "epoch": 1.48, + "grad_norm": 0.4375, + "learning_rate": 3.132633755073897e-05, + "loss": 2.2526, + "step": 34802 + }, + { + "epoch": 1.48, + "grad_norm": 0.46484375, + "learning_rate": 3.132146166737267e-05, + "loss": 2.4807, + "step": 34803 + }, + { + "epoch": 1.48, + "grad_norm": 0.4765625, + "learning_rate": 3.131658609303474e-05, + "loss": 2.482, + "step": 34804 + }, + { + "epoch": 1.48, + "grad_norm": 0.384765625, + "learning_rate": 3.131171082774702e-05, + "loss": 2.3235, + "step": 34805 + }, + { + "epoch": 1.48, + "grad_norm": 0.451171875, + "learning_rate": 3.1306835871531504e-05, + "loss": 2.5515, + "step": 34806 + }, + { + "epoch": 1.48, + "grad_norm": 0.5078125, + "learning_rate": 3.130196122441016e-05, + "loss": 2.6089, + "step": 34807 + }, + { + "epoch": 1.48, + "grad_norm": 0.48046875, + "learning_rate": 3.129708688640488e-05, + "loss": 2.4814, + "step": 34808 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.129221285753757e-05, + "loss": 2.508, + "step": 34809 + }, + { + "epoch": 1.48, + "grad_norm": 0.453125, + "learning_rate": 3.12873391378302e-05, + "loss": 2.3923, + "step": 34810 + }, + { + "epoch": 1.48, + "grad_norm": 0.55078125, + "learning_rate": 3.128246572730471e-05, + "loss": 2.5556, + "step": 34811 + }, + { + "epoch": 1.48, + "grad_norm": 0.498046875, + "learning_rate": 3.1277592625982974e-05, + "loss": 2.4215, + "step": 34812 + }, + { + "epoch": 1.48, + "grad_norm": 0.44921875, + "learning_rate": 3.127271983388698e-05, + "loss": 2.4464, + "step": 34813 + }, + { + "epoch": 1.48, + "grad_norm": 0.5, + "learning_rate": 3.126784735103864e-05, + "loss": 2.3824, + "step": 34814 + }, + { + "epoch": 1.48, + "grad_norm": 0.431640625, + "learning_rate": 3.1262975177459874e-05, + "loss": 2.4035, + "step": 34815 + }, + { + "epoch": 1.48, + "grad_norm": 0.453125, + "learning_rate": 3.125810331317257e-05, + "loss": 2.3351, + "step": 34816 + }, + { + "epoch": 1.48, + "grad_norm": 0.451171875, + "learning_rate": 3.1253231758198685e-05, + "loss": 2.4641, + "step": 34817 + }, + { + "epoch": 1.48, + "grad_norm": 0.412109375, + "learning_rate": 3.124836051256016e-05, + "loss": 2.383, + "step": 34818 + }, + { + "epoch": 1.48, + "grad_norm": 0.48828125, + "learning_rate": 3.124348957627885e-05, + "loss": 2.5252, + "step": 34819 + }, + { + "epoch": 1.48, + "grad_norm": 0.4296875, + "learning_rate": 3.123861894937674e-05, + "loss": 2.4197, + "step": 34820 + }, + { + "epoch": 1.48, + "grad_norm": 0.47265625, + "learning_rate": 3.123374863187569e-05, + "loss": 2.4261, + "step": 34821 + }, + { + "epoch": 1.48, + "grad_norm": 0.490234375, + "learning_rate": 3.122887862379764e-05, + "loss": 2.5135, + "step": 34822 + }, + { + "epoch": 1.48, + "grad_norm": 0.4453125, + "learning_rate": 3.1224008925164526e-05, + "loss": 2.3305, + "step": 34823 + }, + { + "epoch": 1.48, + "grad_norm": 0.474609375, + "learning_rate": 3.1219139535998255e-05, + "loss": 2.4906, + "step": 34824 + }, + { + "epoch": 1.48, + "grad_norm": 0.51171875, + "learning_rate": 3.121427045632067e-05, + "loss": 2.4296, + "step": 34825 + }, + { + "epoch": 1.48, + "grad_norm": 0.478515625, + "learning_rate": 3.120940168615374e-05, + "loss": 2.4245, + "step": 34826 + }, + { + "epoch": 1.48, + "grad_norm": 0.45703125, + "learning_rate": 3.1204533225519384e-05, + "loss": 2.5074, + "step": 34827 + }, + { + "epoch": 1.48, + "grad_norm": 0.40234375, + "learning_rate": 3.119966507443947e-05, + "loss": 2.3377, + "step": 34828 + }, + { + "epoch": 1.48, + "grad_norm": 0.498046875, + "learning_rate": 3.1194797232935904e-05, + "loss": 2.3629, + "step": 34829 + }, + { + "epoch": 1.48, + "grad_norm": 0.52734375, + "learning_rate": 3.1189929701030654e-05, + "loss": 2.5677, + "step": 34830 + }, + { + "epoch": 1.48, + "grad_norm": 0.46875, + "learning_rate": 3.118506247874555e-05, + "loss": 2.471, + "step": 34831 + }, + { + "epoch": 1.48, + "grad_norm": 0.46875, + "learning_rate": 3.11801955661025e-05, + "loss": 2.5402, + "step": 34832 + }, + { + "epoch": 1.48, + "grad_norm": 0.4375, + "learning_rate": 3.1175328963123417e-05, + "loss": 2.3746, + "step": 34833 + }, + { + "epoch": 1.48, + "grad_norm": 0.4453125, + "learning_rate": 3.117046266983024e-05, + "loss": 2.348, + "step": 34834 + }, + { + "epoch": 1.48, + "grad_norm": 0.37890625, + "learning_rate": 3.1165596686244783e-05, + "loss": 2.4197, + "step": 34835 + }, + { + "epoch": 1.48, + "grad_norm": 0.419921875, + "learning_rate": 3.116073101238902e-05, + "loss": 2.3474, + "step": 34836 + }, + { + "epoch": 1.48, + "grad_norm": 0.427734375, + "learning_rate": 3.115586564828478e-05, + "loss": 2.394, + "step": 34837 + }, + { + "epoch": 1.48, + "grad_norm": 0.46484375, + "learning_rate": 3.1151000593953984e-05, + "loss": 2.4518, + "step": 34838 + }, + { + "epoch": 1.48, + "grad_norm": 0.44140625, + "learning_rate": 3.114613584941854e-05, + "loss": 2.4775, + "step": 34839 + }, + { + "epoch": 1.48, + "grad_norm": 0.453125, + "learning_rate": 3.1141271414700345e-05, + "loss": 2.5041, + "step": 34840 + }, + { + "epoch": 1.48, + "grad_norm": 0.4453125, + "learning_rate": 3.113640728982122e-05, + "loss": 2.5873, + "step": 34841 + }, + { + "epoch": 1.48, + "grad_norm": 0.470703125, + "learning_rate": 3.11315434748031e-05, + "loss": 2.4177, + "step": 34842 + }, + { + "epoch": 1.48, + "grad_norm": 0.4453125, + "learning_rate": 3.1126679969667906e-05, + "loss": 2.3691, + "step": 34843 + }, + { + "epoch": 1.48, + "grad_norm": 0.53125, + "learning_rate": 3.112181677443743e-05, + "loss": 2.4964, + "step": 34844 + }, + { + "epoch": 1.48, + "grad_norm": 0.443359375, + "learning_rate": 3.1116953889133614e-05, + "loss": 2.5703, + "step": 34845 + }, + { + "epoch": 1.48, + "grad_norm": 0.419921875, + "learning_rate": 3.1112091313778366e-05, + "loss": 2.3429, + "step": 34846 + }, + { + "epoch": 1.48, + "grad_norm": 0.384765625, + "learning_rate": 3.110722904839354e-05, + "loss": 2.3666, + "step": 34847 + }, + { + "epoch": 1.48, + "grad_norm": 0.466796875, + "learning_rate": 3.1102367093000964e-05, + "loss": 2.4118, + "step": 34848 + }, + { + "epoch": 1.49, + "grad_norm": 0.431640625, + "learning_rate": 3.1097505447622556e-05, + "loss": 2.2467, + "step": 34849 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 3.1092644112280236e-05, + "loss": 2.5781, + "step": 34850 + }, + { + "epoch": 1.49, + "grad_norm": 0.48046875, + "learning_rate": 3.108778308699579e-05, + "loss": 2.4646, + "step": 34851 + }, + { + "epoch": 1.49, + "grad_norm": 0.46875, + "learning_rate": 3.108292237179117e-05, + "loss": 2.4828, + "step": 34852 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.107806196668818e-05, + "loss": 2.4146, + "step": 34853 + }, + { + "epoch": 1.49, + "grad_norm": 0.48828125, + "learning_rate": 3.107320187170873e-05, + "loss": 2.4735, + "step": 34854 + }, + { + "epoch": 1.49, + "grad_norm": 0.478515625, + "learning_rate": 3.1068342086874705e-05, + "loss": 2.449, + "step": 34855 + }, + { + "epoch": 1.49, + "grad_norm": 0.515625, + "learning_rate": 3.106348261220795e-05, + "loss": 2.4938, + "step": 34856 + }, + { + "epoch": 1.49, + "grad_norm": 0.423828125, + "learning_rate": 3.1058623447730304e-05, + "loss": 2.3866, + "step": 34857 + }, + { + "epoch": 1.49, + "grad_norm": 0.44140625, + "learning_rate": 3.105376459346365e-05, + "loss": 2.4102, + "step": 34858 + }, + { + "epoch": 1.49, + "grad_norm": 0.5, + "learning_rate": 3.104890604942989e-05, + "loss": 2.4983, + "step": 34859 + }, + { + "epoch": 1.49, + "grad_norm": 0.57421875, + "learning_rate": 3.1044047815650835e-05, + "loss": 2.5047, + "step": 34860 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.103918989214836e-05, + "loss": 2.4647, + "step": 34861 + }, + { + "epoch": 1.49, + "grad_norm": 0.435546875, + "learning_rate": 3.103433227894436e-05, + "loss": 2.3914, + "step": 34862 + }, + { + "epoch": 1.49, + "grad_norm": 0.421875, + "learning_rate": 3.102947497606064e-05, + "loss": 2.425, + "step": 34863 + }, + { + "epoch": 1.49, + "grad_norm": 0.44140625, + "learning_rate": 3.102461798351907e-05, + "loss": 2.3237, + "step": 34864 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.1019761301341496e-05, + "loss": 2.3433, + "step": 34865 + }, + { + "epoch": 1.49, + "grad_norm": 0.484375, + "learning_rate": 3.101490492954984e-05, + "loss": 2.5203, + "step": 34866 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.1010048868165844e-05, + "loss": 2.4478, + "step": 34867 + }, + { + "epoch": 1.49, + "grad_norm": 0.42578125, + "learning_rate": 3.1005193117211464e-05, + "loss": 2.3566, + "step": 34868 + }, + { + "epoch": 1.49, + "grad_norm": 0.443359375, + "learning_rate": 3.1000337676708455e-05, + "loss": 2.4116, + "step": 34869 + }, + { + "epoch": 1.49, + "grad_norm": 0.484375, + "learning_rate": 3.0995482546678726e-05, + "loss": 2.3919, + "step": 34870 + }, + { + "epoch": 1.49, + "grad_norm": 0.486328125, + "learning_rate": 3.099062772714413e-05, + "loss": 2.5391, + "step": 34871 + }, + { + "epoch": 1.49, + "grad_norm": 0.431640625, + "learning_rate": 3.098577321812648e-05, + "loss": 2.2573, + "step": 34872 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 3.098091901964761e-05, + "loss": 2.4674, + "step": 34873 + }, + { + "epoch": 1.49, + "grad_norm": 0.51171875, + "learning_rate": 3.097606513172937e-05, + "loss": 2.2974, + "step": 34874 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 3.097121155439365e-05, + "loss": 2.5562, + "step": 34875 + }, + { + "epoch": 1.49, + "grad_norm": 0.4453125, + "learning_rate": 3.096635828766222e-05, + "loss": 2.5089, + "step": 34876 + }, + { + "epoch": 1.49, + "grad_norm": 0.5078125, + "learning_rate": 3.096150533155695e-05, + "loss": 2.5134, + "step": 34877 + }, + { + "epoch": 1.49, + "grad_norm": 0.453125, + "learning_rate": 3.0956652686099706e-05, + "loss": 2.3819, + "step": 34878 + }, + { + "epoch": 1.49, + "grad_norm": 0.412109375, + "learning_rate": 3.09518003513123e-05, + "loss": 2.408, + "step": 34879 + }, + { + "epoch": 1.49, + "grad_norm": 0.51953125, + "learning_rate": 3.094694832721651e-05, + "loss": 2.6014, + "step": 34880 + }, + { + "epoch": 1.49, + "grad_norm": 0.466796875, + "learning_rate": 3.0942096613834225e-05, + "loss": 2.5379, + "step": 34881 + }, + { + "epoch": 1.49, + "grad_norm": 0.427734375, + "learning_rate": 3.0937245211187305e-05, + "loss": 2.3453, + "step": 34882 + }, + { + "epoch": 1.49, + "grad_norm": 0.4765625, + "learning_rate": 3.09323941192975e-05, + "loss": 2.3394, + "step": 34883 + }, + { + "epoch": 1.49, + "grad_norm": 0.44140625, + "learning_rate": 3.0927543338186716e-05, + "loss": 2.3095, + "step": 34884 + }, + { + "epoch": 1.49, + "grad_norm": 0.443359375, + "learning_rate": 3.092269286787671e-05, + "loss": 2.4398, + "step": 34885 + }, + { + "epoch": 1.49, + "grad_norm": 0.451171875, + "learning_rate": 3.0917842708389346e-05, + "loss": 2.5242, + "step": 34886 + }, + { + "epoch": 1.49, + "grad_norm": 0.494140625, + "learning_rate": 3.091299285974647e-05, + "loss": 2.5829, + "step": 34887 + }, + { + "epoch": 1.49, + "grad_norm": 0.4765625, + "learning_rate": 3.090814332196988e-05, + "loss": 2.4668, + "step": 34888 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.0903294095081355e-05, + "loss": 2.4431, + "step": 34889 + }, + { + "epoch": 1.49, + "grad_norm": 0.5, + "learning_rate": 3.089844517910278e-05, + "loss": 2.3984, + "step": 34890 + }, + { + "epoch": 1.49, + "grad_norm": 0.453125, + "learning_rate": 3.089359657405595e-05, + "loss": 2.4445, + "step": 34891 + }, + { + "epoch": 1.49, + "grad_norm": 0.490234375, + "learning_rate": 3.0888748279962664e-05, + "loss": 2.4906, + "step": 34892 + }, + { + "epoch": 1.49, + "grad_norm": 0.4375, + "learning_rate": 3.088390029684475e-05, + "loss": 2.4209, + "step": 34893 + }, + { + "epoch": 1.49, + "grad_norm": 0.50390625, + "learning_rate": 3.087905262472406e-05, + "loss": 2.403, + "step": 34894 + }, + { + "epoch": 1.49, + "grad_norm": 0.5078125, + "learning_rate": 3.0874205263622366e-05, + "loss": 2.5115, + "step": 34895 + }, + { + "epoch": 1.49, + "grad_norm": 0.48828125, + "learning_rate": 3.086935821356146e-05, + "loss": 2.3972, + "step": 34896 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 3.086451147456317e-05, + "loss": 2.5239, + "step": 34897 + }, + { + "epoch": 1.49, + "grad_norm": 0.478515625, + "learning_rate": 3.085966504664935e-05, + "loss": 2.5055, + "step": 34898 + }, + { + "epoch": 1.49, + "grad_norm": 0.515625, + "learning_rate": 3.0854818929841745e-05, + "loss": 2.5306, + "step": 34899 + }, + { + "epoch": 1.49, + "grad_norm": 0.48828125, + "learning_rate": 3.08499731241622e-05, + "loss": 2.4605, + "step": 34900 + }, + { + "epoch": 1.49, + "grad_norm": 0.486328125, + "learning_rate": 3.084512762963249e-05, + "loss": 2.5234, + "step": 34901 + }, + { + "epoch": 1.49, + "grad_norm": 0.408203125, + "learning_rate": 3.0840282446274413e-05, + "loss": 2.4838, + "step": 34902 + }, + { + "epoch": 1.49, + "grad_norm": 0.41796875, + "learning_rate": 3.083543757410984e-05, + "loss": 2.4051, + "step": 34903 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.083059301316052e-05, + "loss": 2.4223, + "step": 34904 + }, + { + "epoch": 1.49, + "grad_norm": 0.4375, + "learning_rate": 3.08257487634482e-05, + "loss": 2.3913, + "step": 34905 + }, + { + "epoch": 1.49, + "grad_norm": 0.400390625, + "learning_rate": 3.0820904824994755e-05, + "loss": 2.3536, + "step": 34906 + }, + { + "epoch": 1.49, + "grad_norm": 0.4453125, + "learning_rate": 3.0816061197821975e-05, + "loss": 2.5194, + "step": 34907 + }, + { + "epoch": 1.49, + "grad_norm": 0.486328125, + "learning_rate": 3.08112178819516e-05, + "loss": 2.4951, + "step": 34908 + }, + { + "epoch": 1.49, + "grad_norm": 0.45703125, + "learning_rate": 3.0806374877405474e-05, + "loss": 2.4717, + "step": 34909 + }, + { + "epoch": 1.49, + "grad_norm": 0.50390625, + "learning_rate": 3.0801532184205394e-05, + "loss": 2.4608, + "step": 34910 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.0796689802373134e-05, + "loss": 2.38, + "step": 34911 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.0791847731930454e-05, + "loss": 2.3349, + "step": 34912 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.078700597289915e-05, + "loss": 2.5007, + "step": 34913 + }, + { + "epoch": 1.49, + "grad_norm": 0.439453125, + "learning_rate": 3.078216452530107e-05, + "loss": 2.4842, + "step": 34914 + }, + { + "epoch": 1.49, + "grad_norm": 0.47265625, + "learning_rate": 3.077732338915792e-05, + "loss": 2.2797, + "step": 34915 + }, + { + "epoch": 1.49, + "grad_norm": 0.45703125, + "learning_rate": 3.077248256449154e-05, + "loss": 2.4447, + "step": 34916 + }, + { + "epoch": 1.49, + "grad_norm": 0.490234375, + "learning_rate": 3.0767642051323666e-05, + "loss": 2.4492, + "step": 34917 + }, + { + "epoch": 1.49, + "grad_norm": 0.482421875, + "learning_rate": 3.07628018496761e-05, + "loss": 2.4111, + "step": 34918 + }, + { + "epoch": 1.49, + "grad_norm": 0.482421875, + "learning_rate": 3.075796195957066e-05, + "loss": 2.4554, + "step": 34919 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.075312238102907e-05, + "loss": 2.3934, + "step": 34920 + }, + { + "epoch": 1.49, + "grad_norm": 0.408203125, + "learning_rate": 3.0748283114073076e-05, + "loss": 2.5794, + "step": 34921 + }, + { + "epoch": 1.49, + "grad_norm": 0.498046875, + "learning_rate": 3.074344415872454e-05, + "loss": 2.5554, + "step": 34922 + }, + { + "epoch": 1.49, + "grad_norm": 0.46875, + "learning_rate": 3.073860551500521e-05, + "loss": 2.4426, + "step": 34923 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.073376718293681e-05, + "loss": 2.3313, + "step": 34924 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.072892916254119e-05, + "loss": 2.4846, + "step": 34925 + }, + { + "epoch": 1.49, + "grad_norm": 0.490234375, + "learning_rate": 3.072409145384002e-05, + "loss": 2.4989, + "step": 34926 + }, + { + "epoch": 1.49, + "grad_norm": 0.5390625, + "learning_rate": 3.0719254056855165e-05, + "loss": 2.3133, + "step": 34927 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.071441697160831e-05, + "loss": 2.4414, + "step": 34928 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.070958019812129e-05, + "loss": 2.3858, + "step": 34929 + }, + { + "epoch": 1.49, + "grad_norm": 0.50390625, + "learning_rate": 3.07047437364158e-05, + "loss": 2.3878, + "step": 34930 + }, + { + "epoch": 1.49, + "grad_norm": 0.466796875, + "learning_rate": 3.0699907586513646e-05, + "loss": 2.4946, + "step": 34931 + }, + { + "epoch": 1.49, + "grad_norm": 0.451171875, + "learning_rate": 3.069507174843661e-05, + "loss": 2.5581, + "step": 34932 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.069023622220643e-05, + "loss": 2.4425, + "step": 34933 + }, + { + "epoch": 1.49, + "grad_norm": 0.484375, + "learning_rate": 3.0685401007844784e-05, + "loss": 2.3962, + "step": 34934 + }, + { + "epoch": 1.49, + "grad_norm": 0.443359375, + "learning_rate": 3.068056610537357e-05, + "loss": 2.5062, + "step": 34935 + }, + { + "epoch": 1.49, + "grad_norm": 0.5, + "learning_rate": 3.0675731514814475e-05, + "loss": 2.5077, + "step": 34936 + }, + { + "epoch": 1.49, + "grad_norm": 0.466796875, + "learning_rate": 3.0670897236189225e-05, + "loss": 2.4915, + "step": 34937 + }, + { + "epoch": 1.49, + "grad_norm": 0.38671875, + "learning_rate": 3.066606326951958e-05, + "loss": 2.397, + "step": 34938 + }, + { + "epoch": 1.49, + "grad_norm": 0.4609375, + "learning_rate": 3.066122961482737e-05, + "loss": 2.3001, + "step": 34939 + }, + { + "epoch": 1.49, + "grad_norm": 0.5, + "learning_rate": 3.0656396272134246e-05, + "loss": 2.3716, + "step": 34940 + }, + { + "epoch": 1.49, + "grad_norm": 0.390625, + "learning_rate": 3.0651563241462024e-05, + "loss": 2.421, + "step": 34941 + }, + { + "epoch": 1.49, + "grad_norm": 0.451171875, + "learning_rate": 3.064673052283239e-05, + "loss": 2.4882, + "step": 34942 + }, + { + "epoch": 1.49, + "grad_norm": 0.494140625, + "learning_rate": 3.064189811626717e-05, + "loss": 2.5864, + "step": 34943 + }, + { + "epoch": 1.49, + "grad_norm": 0.46875, + "learning_rate": 3.063706602178801e-05, + "loss": 2.502, + "step": 34944 + }, + { + "epoch": 1.49, + "grad_norm": 0.486328125, + "learning_rate": 3.063223423941674e-05, + "loss": 2.5817, + "step": 34945 + }, + { + "epoch": 1.49, + "grad_norm": 0.458984375, + "learning_rate": 3.062740276917504e-05, + "loss": 2.5748, + "step": 34946 + }, + { + "epoch": 1.49, + "grad_norm": 0.490234375, + "learning_rate": 3.062257161108467e-05, + "loss": 2.5715, + "step": 34947 + }, + { + "epoch": 1.49, + "grad_norm": 0.466796875, + "learning_rate": 3.06177407651674e-05, + "loss": 2.3565, + "step": 34948 + }, + { + "epoch": 1.49, + "grad_norm": 0.453125, + "learning_rate": 3.0612910231444957e-05, + "loss": 2.5174, + "step": 34949 + }, + { + "epoch": 1.49, + "grad_norm": 0.53515625, + "learning_rate": 3.060808000993898e-05, + "loss": 2.5032, + "step": 34950 + }, + { + "epoch": 1.49, + "grad_norm": 0.458984375, + "learning_rate": 3.060325010067135e-05, + "loss": 2.479, + "step": 34951 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 3.059842050366374e-05, + "loss": 2.4184, + "step": 34952 + }, + { + "epoch": 1.49, + "grad_norm": 0.41015625, + "learning_rate": 3.059359121893783e-05, + "loss": 2.2802, + "step": 34953 + }, + { + "epoch": 1.49, + "grad_norm": 0.52734375, + "learning_rate": 3.058876224651541e-05, + "loss": 2.5093, + "step": 34954 + }, + { + "epoch": 1.49, + "grad_norm": 0.52734375, + "learning_rate": 3.058393358641821e-05, + "loss": 2.5908, + "step": 34955 + }, + { + "epoch": 1.49, + "grad_norm": 0.5234375, + "learning_rate": 3.0579105238667925e-05, + "loss": 2.4483, + "step": 34956 + }, + { + "epoch": 1.49, + "grad_norm": 0.435546875, + "learning_rate": 3.0574277203286315e-05, + "loss": 2.4583, + "step": 34957 + }, + { + "epoch": 1.49, + "grad_norm": 0.47265625, + "learning_rate": 3.056944948029506e-05, + "loss": 2.5711, + "step": 34958 + }, + { + "epoch": 1.49, + "grad_norm": 0.466796875, + "learning_rate": 3.056462206971593e-05, + "loss": 2.4689, + "step": 34959 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 3.055979497157061e-05, + "loss": 2.3862, + "step": 34960 + }, + { + "epoch": 1.49, + "grad_norm": 0.53515625, + "learning_rate": 3.055496818588085e-05, + "loss": 2.396, + "step": 34961 + }, + { + "epoch": 1.49, + "grad_norm": 0.48046875, + "learning_rate": 3.055014171266834e-05, + "loss": 2.454, + "step": 34962 + }, + { + "epoch": 1.49, + "grad_norm": 0.482421875, + "learning_rate": 3.054531555195479e-05, + "loss": 2.435, + "step": 34963 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.054048970376198e-05, + "loss": 2.409, + "step": 34964 + }, + { + "epoch": 1.49, + "grad_norm": 0.4453125, + "learning_rate": 3.0535664168111586e-05, + "loss": 2.443, + "step": 34965 + }, + { + "epoch": 1.49, + "grad_norm": 0.423828125, + "learning_rate": 3.053083894502524e-05, + "loss": 2.4526, + "step": 34966 + }, + { + "epoch": 1.49, + "grad_norm": 0.5078125, + "learning_rate": 3.052601403452481e-05, + "loss": 2.4444, + "step": 34967 + }, + { + "epoch": 1.49, + "grad_norm": 0.4609375, + "learning_rate": 3.0521189436631917e-05, + "loss": 2.4316, + "step": 34968 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.0516365151368253e-05, + "loss": 2.5584, + "step": 34969 + }, + { + "epoch": 1.49, + "grad_norm": 0.400390625, + "learning_rate": 3.0511541178755553e-05, + "loss": 2.4683, + "step": 34970 + }, + { + "epoch": 1.49, + "grad_norm": 0.52734375, + "learning_rate": 3.0506717518815555e-05, + "loss": 2.5871, + "step": 34971 + }, + { + "epoch": 1.49, + "grad_norm": 0.494140625, + "learning_rate": 3.0501894171569888e-05, + "loss": 2.3941, + "step": 34972 + }, + { + "epoch": 1.49, + "grad_norm": 0.412109375, + "learning_rate": 3.0497071137040346e-05, + "loss": 2.4517, + "step": 34973 + }, + { + "epoch": 1.49, + "grad_norm": 0.408203125, + "learning_rate": 3.0492248415248538e-05, + "loss": 2.5153, + "step": 34974 + }, + { + "epoch": 1.49, + "grad_norm": 0.435546875, + "learning_rate": 3.0487426006216212e-05, + "loss": 2.4512, + "step": 34975 + }, + { + "epoch": 1.49, + "grad_norm": 0.439453125, + "learning_rate": 3.0482603909965113e-05, + "loss": 2.2947, + "step": 34976 + }, + { + "epoch": 1.49, + "grad_norm": 0.48046875, + "learning_rate": 3.0477782126516875e-05, + "loss": 2.6468, + "step": 34977 + }, + { + "epoch": 1.49, + "grad_norm": 0.490234375, + "learning_rate": 3.0472960655893178e-05, + "loss": 2.4285, + "step": 34978 + }, + { + "epoch": 1.49, + "grad_norm": 0.4375, + "learning_rate": 3.0468139498115757e-05, + "loss": 2.5108, + "step": 34979 + }, + { + "epoch": 1.49, + "grad_norm": 0.515625, + "learning_rate": 3.046331865320632e-05, + "loss": 2.5153, + "step": 34980 + }, + { + "epoch": 1.49, + "grad_norm": 0.396484375, + "learning_rate": 3.045849812118654e-05, + "loss": 2.471, + "step": 34981 + }, + { + "epoch": 1.49, + "grad_norm": 0.421875, + "learning_rate": 3.0453677902078047e-05, + "loss": 2.2684, + "step": 34982 + }, + { + "epoch": 1.49, + "grad_norm": 0.443359375, + "learning_rate": 3.0448857995902645e-05, + "loss": 2.346, + "step": 34983 + }, + { + "epoch": 1.49, + "grad_norm": 0.482421875, + "learning_rate": 3.0444038402681963e-05, + "loss": 2.4559, + "step": 34984 + }, + { + "epoch": 1.49, + "grad_norm": 0.46875, + "learning_rate": 3.0439219122437657e-05, + "loss": 2.3999, + "step": 34985 + }, + { + "epoch": 1.49, + "grad_norm": 0.427734375, + "learning_rate": 3.0434400155191443e-05, + "loss": 2.4428, + "step": 34986 + }, + { + "epoch": 1.49, + "grad_norm": 0.458984375, + "learning_rate": 3.0429581500965042e-05, + "loss": 2.5001, + "step": 34987 + }, + { + "epoch": 1.49, + "grad_norm": 0.48828125, + "learning_rate": 3.0424763159780066e-05, + "loss": 2.4207, + "step": 34988 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 3.0419945131658258e-05, + "loss": 2.4644, + "step": 34989 + }, + { + "epoch": 1.49, + "grad_norm": 0.48046875, + "learning_rate": 3.0415127416621236e-05, + "loss": 2.6135, + "step": 34990 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.0410310014690712e-05, + "loss": 2.4122, + "step": 34991 + }, + { + "epoch": 1.49, + "grad_norm": 0.41796875, + "learning_rate": 3.040549292588839e-05, + "loss": 2.3179, + "step": 34992 + }, + { + "epoch": 1.49, + "grad_norm": 0.4453125, + "learning_rate": 3.040067615023592e-05, + "loss": 2.4876, + "step": 34993 + }, + { + "epoch": 1.49, + "grad_norm": 0.51171875, + "learning_rate": 3.0395859687754937e-05, + "loss": 2.4848, + "step": 34994 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 3.0391043538467144e-05, + "loss": 2.526, + "step": 34995 + }, + { + "epoch": 1.49, + "grad_norm": 0.41796875, + "learning_rate": 3.0386227702394253e-05, + "loss": 2.528, + "step": 34996 + }, + { + "epoch": 1.49, + "grad_norm": 0.51171875, + "learning_rate": 3.03814121795579e-05, + "loss": 2.4949, + "step": 34997 + }, + { + "epoch": 1.49, + "grad_norm": 0.4609375, + "learning_rate": 3.0376596969979677e-05, + "loss": 2.4587, + "step": 34998 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.0371782073681387e-05, + "loss": 2.4226, + "step": 34999 + }, + { + "epoch": 1.49, + "grad_norm": 0.46875, + "learning_rate": 3.036696749068464e-05, + "loss": 2.3952, + "step": 35000 + }, + { + "epoch": 1.49, + "grad_norm": 0.4453125, + "learning_rate": 3.036215322101106e-05, + "loss": 2.4872, + "step": 35001 + }, + { + "epoch": 1.49, + "grad_norm": 0.4765625, + "learning_rate": 3.0357339264682327e-05, + "loss": 2.5414, + "step": 35002 + }, + { + "epoch": 1.49, + "grad_norm": 0.50390625, + "learning_rate": 3.0352525621720164e-05, + "loss": 2.4455, + "step": 35003 + }, + { + "epoch": 1.49, + "grad_norm": 0.453125, + "learning_rate": 3.0347712292146146e-05, + "loss": 2.4956, + "step": 35004 + }, + { + "epoch": 1.49, + "grad_norm": 0.51953125, + "learning_rate": 3.0342899275981994e-05, + "loss": 2.4723, + "step": 35005 + }, + { + "epoch": 1.49, + "grad_norm": 0.392578125, + "learning_rate": 3.0338086573249315e-05, + "loss": 2.413, + "step": 35006 + }, + { + "epoch": 1.49, + "grad_norm": 0.451171875, + "learning_rate": 3.0333274183969775e-05, + "loss": 2.385, + "step": 35007 + }, + { + "epoch": 1.49, + "grad_norm": 0.48828125, + "learning_rate": 3.032846210816508e-05, + "loss": 2.4613, + "step": 35008 + }, + { + "epoch": 1.49, + "grad_norm": 0.439453125, + "learning_rate": 3.0323650345856846e-05, + "loss": 2.3396, + "step": 35009 + }, + { + "epoch": 1.49, + "grad_norm": 0.4921875, + "learning_rate": 3.031883889706668e-05, + "loss": 2.5558, + "step": 35010 + }, + { + "epoch": 1.49, + "grad_norm": 0.490234375, + "learning_rate": 3.0314027761816288e-05, + "loss": 2.5222, + "step": 35011 + }, + { + "epoch": 1.49, + "grad_norm": 0.478515625, + "learning_rate": 3.0309216940127328e-05, + "loss": 2.5253, + "step": 35012 + }, + { + "epoch": 1.49, + "grad_norm": 0.455078125, + "learning_rate": 3.0304406432021427e-05, + "loss": 2.4201, + "step": 35013 + }, + { + "epoch": 1.49, + "grad_norm": 0.53125, + "learning_rate": 3.0299596237520157e-05, + "loss": 2.5789, + "step": 35014 + }, + { + "epoch": 1.49, + "grad_norm": 0.43359375, + "learning_rate": 3.0294786356645298e-05, + "loss": 2.4401, + "step": 35015 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.0289976789418417e-05, + "loss": 2.4771, + "step": 35016 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.0285167535861146e-05, + "loss": 2.3692, + "step": 35017 + }, + { + "epoch": 1.49, + "grad_norm": 0.408203125, + "learning_rate": 3.028035859599514e-05, + "loss": 2.5326, + "step": 35018 + }, + { + "epoch": 1.49, + "grad_norm": 0.48828125, + "learning_rate": 3.0275549969842065e-05, + "loss": 2.4272, + "step": 35019 + }, + { + "epoch": 1.49, + "grad_norm": 0.4921875, + "learning_rate": 3.0270741657423508e-05, + "loss": 2.5132, + "step": 35020 + }, + { + "epoch": 1.49, + "grad_norm": 0.451171875, + "learning_rate": 3.026593365876116e-05, + "loss": 2.4194, + "step": 35021 + }, + { + "epoch": 1.49, + "grad_norm": 0.48828125, + "learning_rate": 3.026112597387659e-05, + "loss": 2.4989, + "step": 35022 + }, + { + "epoch": 1.49, + "grad_norm": 0.41796875, + "learning_rate": 3.0256318602791478e-05, + "loss": 2.4725, + "step": 35023 + }, + { + "epoch": 1.49, + "grad_norm": 0.455078125, + "learning_rate": 3.0251511545527467e-05, + "loss": 2.4898, + "step": 35024 + }, + { + "epoch": 1.49, + "grad_norm": 0.486328125, + "learning_rate": 3.0246704802106175e-05, + "loss": 2.5768, + "step": 35025 + }, + { + "epoch": 1.49, + "grad_norm": 0.474609375, + "learning_rate": 3.024189837254917e-05, + "loss": 2.4325, + "step": 35026 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 3.023709225687814e-05, + "loss": 2.3539, + "step": 35027 + }, + { + "epoch": 1.49, + "grad_norm": 0.41796875, + "learning_rate": 3.0232286455114733e-05, + "loss": 2.3298, + "step": 35028 + }, + { + "epoch": 1.49, + "grad_norm": 0.53125, + "learning_rate": 3.0227480967280496e-05, + "loss": 2.5142, + "step": 35029 + }, + { + "epoch": 1.49, + "grad_norm": 0.474609375, + "learning_rate": 3.0222675793397103e-05, + "loss": 2.4095, + "step": 35030 + }, + { + "epoch": 1.49, + "grad_norm": 0.38671875, + "learning_rate": 3.0217870933486204e-05, + "loss": 2.3206, + "step": 35031 + }, + { + "epoch": 1.49, + "grad_norm": 0.49609375, + "learning_rate": 3.021306638756938e-05, + "loss": 2.434, + "step": 35032 + }, + { + "epoch": 1.49, + "grad_norm": 0.48046875, + "learning_rate": 3.020826215566822e-05, + "loss": 2.5387, + "step": 35033 + }, + { + "epoch": 1.49, + "grad_norm": 0.4609375, + "learning_rate": 3.0203458237804373e-05, + "loss": 2.419, + "step": 35034 + }, + { + "epoch": 1.49, + "grad_norm": 0.453125, + "learning_rate": 3.019865463399949e-05, + "loss": 2.482, + "step": 35035 + }, + { + "epoch": 1.49, + "grad_norm": 0.48828125, + "learning_rate": 3.0193851344275127e-05, + "loss": 2.5228, + "step": 35036 + }, + { + "epoch": 1.49, + "grad_norm": 0.5078125, + "learning_rate": 3.018904836865294e-05, + "loss": 2.4597, + "step": 35037 + }, + { + "epoch": 1.49, + "grad_norm": 0.4765625, + "learning_rate": 3.01842457071545e-05, + "loss": 2.4267, + "step": 35038 + }, + { + "epoch": 1.49, + "grad_norm": 0.451171875, + "learning_rate": 3.017944335980144e-05, + "loss": 2.5813, + "step": 35039 + }, + { + "epoch": 1.49, + "grad_norm": 0.419921875, + "learning_rate": 3.0174641326615392e-05, + "loss": 2.4312, + "step": 35040 + }, + { + "epoch": 1.49, + "grad_norm": 0.416015625, + "learning_rate": 3.0169839607617943e-05, + "loss": 2.3043, + "step": 35041 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.0165038202830676e-05, + "loss": 2.3519, + "step": 35042 + }, + { + "epoch": 1.49, + "grad_norm": 0.5234375, + "learning_rate": 3.0160237112275204e-05, + "loss": 2.3248, + "step": 35043 + }, + { + "epoch": 1.49, + "grad_norm": 0.439453125, + "learning_rate": 3.0155436335973173e-05, + "loss": 2.2997, + "step": 35044 + }, + { + "epoch": 1.49, + "grad_norm": 0.466796875, + "learning_rate": 3.015063587394612e-05, + "loss": 2.4245, + "step": 35045 + }, + { + "epoch": 1.49, + "grad_norm": 0.431640625, + "learning_rate": 3.014583572621571e-05, + "loss": 2.4427, + "step": 35046 + }, + { + "epoch": 1.49, + "grad_norm": 0.455078125, + "learning_rate": 3.014103589280348e-05, + "loss": 2.5164, + "step": 35047 + }, + { + "epoch": 1.49, + "grad_norm": 0.47265625, + "learning_rate": 3.0136236373731095e-05, + "loss": 2.5364, + "step": 35048 + }, + { + "epoch": 1.49, + "grad_norm": 0.42578125, + "learning_rate": 3.0131437169020084e-05, + "loss": 2.3376, + "step": 35049 + }, + { + "epoch": 1.49, + "grad_norm": 0.50390625, + "learning_rate": 3.01266382786921e-05, + "loss": 2.5049, + "step": 35050 + }, + { + "epoch": 1.49, + "grad_norm": 0.466796875, + "learning_rate": 3.0121839702768674e-05, + "loss": 2.4266, + "step": 35051 + }, + { + "epoch": 1.49, + "grad_norm": 0.5078125, + "learning_rate": 3.011704144127143e-05, + "loss": 2.4361, + "step": 35052 + }, + { + "epoch": 1.49, + "grad_norm": 0.48046875, + "learning_rate": 3.0112243494221994e-05, + "loss": 2.3649, + "step": 35053 + }, + { + "epoch": 1.49, + "grad_norm": 0.49609375, + "learning_rate": 3.0107445861641927e-05, + "loss": 2.4644, + "step": 35054 + }, + { + "epoch": 1.49, + "grad_norm": 0.443359375, + "learning_rate": 3.0102648543552735e-05, + "loss": 2.5571, + "step": 35055 + }, + { + "epoch": 1.49, + "grad_norm": 0.53125, + "learning_rate": 3.0097851539976153e-05, + "loss": 2.4228, + "step": 35056 + }, + { + "epoch": 1.49, + "grad_norm": 0.47265625, + "learning_rate": 3.009305485093369e-05, + "loss": 2.4451, + "step": 35057 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.0088258476446895e-05, + "loss": 2.4343, + "step": 35058 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.00834624165374e-05, + "loss": 2.4223, + "step": 35059 + }, + { + "epoch": 1.49, + "grad_norm": 0.455078125, + "learning_rate": 3.0078666671226798e-05, + "loss": 2.414, + "step": 35060 + }, + { + "epoch": 1.49, + "grad_norm": 0.4453125, + "learning_rate": 3.00738712405366e-05, + "loss": 2.4529, + "step": 35061 + }, + { + "epoch": 1.49, + "grad_norm": 0.439453125, + "learning_rate": 3.006907612448847e-05, + "loss": 2.461, + "step": 35062 + }, + { + "epoch": 1.49, + "grad_norm": 0.474609375, + "learning_rate": 3.0064281323103903e-05, + "loss": 2.371, + "step": 35063 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.0059486836404548e-05, + "loss": 2.4809, + "step": 35064 + }, + { + "epoch": 1.49, + "grad_norm": 0.44921875, + "learning_rate": 3.005469266441191e-05, + "loss": 2.362, + "step": 35065 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.0049898807147626e-05, + "loss": 2.4471, + "step": 35066 + }, + { + "epoch": 1.49, + "grad_norm": 0.50390625, + "learning_rate": 3.00451052646332e-05, + "loss": 2.4668, + "step": 35067 + }, + { + "epoch": 1.49, + "grad_norm": 0.439453125, + "learning_rate": 3.004031203689024e-05, + "loss": 2.3828, + "step": 35068 + }, + { + "epoch": 1.49, + "grad_norm": 0.453125, + "learning_rate": 3.0035519123940327e-05, + "loss": 2.5147, + "step": 35069 + }, + { + "epoch": 1.49, + "grad_norm": 0.52734375, + "learning_rate": 3.003072652580502e-05, + "loss": 2.579, + "step": 35070 + }, + { + "epoch": 1.49, + "grad_norm": 0.4296875, + "learning_rate": 3.0025934242505803e-05, + "loss": 2.4853, + "step": 35071 + }, + { + "epoch": 1.49, + "grad_norm": 0.46484375, + "learning_rate": 3.0021142274064372e-05, + "loss": 2.3514, + "step": 35072 + }, + { + "epoch": 1.49, + "grad_norm": 0.451171875, + "learning_rate": 3.001635062050223e-05, + "loss": 2.576, + "step": 35073 + }, + { + "epoch": 1.49, + "grad_norm": 0.416015625, + "learning_rate": 3.0011559281840906e-05, + "loss": 2.3861, + "step": 35074 + }, + { + "epoch": 1.49, + "grad_norm": 0.4609375, + "learning_rate": 3.000676825810198e-05, + "loss": 2.3382, + "step": 35075 + }, + { + "epoch": 1.49, + "grad_norm": 0.486328125, + "learning_rate": 3.000197754930706e-05, + "loss": 2.4502, + "step": 35076 + }, + { + "epoch": 1.49, + "grad_norm": 0.478515625, + "learning_rate": 2.9997187155477613e-05, + "loss": 2.3704, + "step": 35077 + }, + { + "epoch": 1.49, + "grad_norm": 0.453125, + "learning_rate": 2.999239707663528e-05, + "loss": 2.5045, + "step": 35078 + }, + { + "epoch": 1.49, + "grad_norm": 0.4609375, + "learning_rate": 2.9987607312801546e-05, + "loss": 2.4045, + "step": 35079 + }, + { + "epoch": 1.49, + "grad_norm": 0.47265625, + "learning_rate": 2.9982817863998025e-05, + "loss": 2.4745, + "step": 35080 + }, + { + "epoch": 1.49, + "grad_norm": 0.470703125, + "learning_rate": 2.99780287302462e-05, + "loss": 2.4024, + "step": 35081 + }, + { + "epoch": 1.49, + "grad_norm": 0.462890625, + "learning_rate": 2.9973239911567686e-05, + "loss": 2.4623, + "step": 35082 + }, + { + "epoch": 1.49, + "grad_norm": 0.47265625, + "learning_rate": 2.9968451407983965e-05, + "loss": 2.5784, + "step": 35083 + }, + { + "epoch": 1.5, + "grad_norm": 0.4765625, + "learning_rate": 2.9963663219516623e-05, + "loss": 2.5493, + "step": 35084 + }, + { + "epoch": 1.5, + "grad_norm": 0.5, + "learning_rate": 2.995887534618724e-05, + "loss": 2.3737, + "step": 35085 + }, + { + "epoch": 1.5, + "grad_norm": 0.462890625, + "learning_rate": 2.9954087788017315e-05, + "loss": 2.4914, + "step": 35086 + }, + { + "epoch": 1.5, + "grad_norm": 0.451171875, + "learning_rate": 2.9949300545028335e-05, + "loss": 2.4865, + "step": 35087 + }, + { + "epoch": 1.5, + "grad_norm": 0.53125, + "learning_rate": 2.9944513617241965e-05, + "loss": 2.4918, + "step": 35088 + }, + { + "epoch": 1.5, + "grad_norm": 0.435546875, + "learning_rate": 2.9939727004679675e-05, + "loss": 2.5693, + "step": 35089 + }, + { + "epoch": 1.5, + "grad_norm": 0.45703125, + "learning_rate": 2.993494070736298e-05, + "loss": 2.3673, + "step": 35090 + }, + { + "epoch": 1.5, + "grad_norm": 0.474609375, + "learning_rate": 2.9930154725313453e-05, + "loss": 2.3012, + "step": 35091 + }, + { + "epoch": 1.5, + "grad_norm": 0.408203125, + "learning_rate": 2.9925369058552645e-05, + "loss": 2.5439, + "step": 35092 + }, + { + "epoch": 1.5, + "grad_norm": 0.498046875, + "learning_rate": 2.9920583707102047e-05, + "loss": 2.3821, + "step": 35093 + }, + { + "epoch": 1.5, + "grad_norm": 0.474609375, + "learning_rate": 2.991579867098323e-05, + "loss": 2.4601, + "step": 35094 + }, + { + "epoch": 1.5, + "grad_norm": 0.447265625, + "learning_rate": 2.9911013950217682e-05, + "loss": 2.5143, + "step": 35095 + }, + { + "epoch": 1.5, + "grad_norm": 0.453125, + "learning_rate": 2.990622954482698e-05, + "loss": 2.5084, + "step": 35096 + }, + { + "epoch": 1.5, + "grad_norm": 0.5234375, + "learning_rate": 2.990144545483261e-05, + "loss": 2.3812, + "step": 35097 + }, + { + "epoch": 1.5, + "grad_norm": 0.490234375, + "learning_rate": 2.9896661680256142e-05, + "loss": 2.4981, + "step": 35098 + }, + { + "epoch": 1.5, + "grad_norm": 0.51953125, + "learning_rate": 2.989187822111904e-05, + "loss": 2.5721, + "step": 35099 + }, + { + "epoch": 1.5, + "grad_norm": 0.451171875, + "learning_rate": 2.9887095077442874e-05, + "loss": 2.4579, + "step": 35100 + }, + { + "epoch": 1.5, + "grad_norm": 0.48828125, + "learning_rate": 2.988231224924918e-05, + "loss": 2.4414, + "step": 35101 + }, + { + "epoch": 1.5, + "grad_norm": 0.55078125, + "learning_rate": 2.987752973655946e-05, + "loss": 2.4667, + "step": 35102 + }, + { + "epoch": 1.5, + "grad_norm": 0.5234375, + "learning_rate": 2.9872747539395163e-05, + "loss": 2.4401, + "step": 35103 + }, + { + "epoch": 1.5, + "grad_norm": 0.40234375, + "learning_rate": 2.986796565777794e-05, + "loss": 2.3064, + "step": 35104 + }, + { + "epoch": 1.5, + "grad_norm": 0.48046875, + "learning_rate": 2.9863184091729245e-05, + "loss": 2.4115, + "step": 35105 + }, + { + "epoch": 1.5, + "grad_norm": 0.4921875, + "learning_rate": 2.985840284127055e-05, + "loss": 2.4459, + "step": 35106 + }, + { + "epoch": 1.5, + "grad_norm": 0.44921875, + "learning_rate": 2.9853621906423414e-05, + "loss": 2.5148, + "step": 35107 + }, + { + "epoch": 1.5, + "grad_norm": 0.453125, + "learning_rate": 2.9848841287209374e-05, + "loss": 2.4485, + "step": 35108 + }, + { + "epoch": 1.5, + "grad_norm": 0.4453125, + "learning_rate": 2.984406098364988e-05, + "loss": 2.4751, + "step": 35109 + }, + { + "epoch": 1.5, + "grad_norm": 0.46875, + "learning_rate": 2.9839280995766504e-05, + "loss": 2.4408, + "step": 35110 + }, + { + "epoch": 1.5, + "grad_norm": 0.53515625, + "learning_rate": 2.9834501323580687e-05, + "loss": 2.5945, + "step": 35111 + }, + { + "epoch": 1.5, + "grad_norm": 0.5234375, + "learning_rate": 2.982972196711401e-05, + "loss": 2.4431, + "step": 35112 + }, + { + "epoch": 1.5, + "grad_norm": 0.515625, + "learning_rate": 2.98249429263879e-05, + "loss": 2.4939, + "step": 35113 + }, + { + "epoch": 1.5, + "grad_norm": 0.451171875, + "learning_rate": 2.9820164201423938e-05, + "loss": 2.5548, + "step": 35114 + }, + { + "epoch": 1.5, + "grad_norm": 0.49609375, + "learning_rate": 2.9815385792243554e-05, + "loss": 2.3851, + "step": 35115 + }, + { + "epoch": 1.5, + "grad_norm": 0.439453125, + "learning_rate": 2.981060769886829e-05, + "loss": 2.3072, + "step": 35116 + }, + { + "epoch": 1.5, + "grad_norm": 0.50390625, + "learning_rate": 2.9805829921319662e-05, + "loss": 2.51, + "step": 35117 + }, + { + "epoch": 1.5, + "grad_norm": 0.443359375, + "learning_rate": 2.980105245961916e-05, + "loss": 2.4367, + "step": 35118 + }, + { + "epoch": 1.5, + "grad_norm": 0.482421875, + "learning_rate": 2.979627531378819e-05, + "loss": 2.469, + "step": 35119 + }, + { + "epoch": 1.5, + "grad_norm": 0.47265625, + "learning_rate": 2.97914984838484e-05, + "loss": 2.4905, + "step": 35120 + }, + { + "epoch": 1.5, + "grad_norm": 0.408203125, + "learning_rate": 2.9786721969821196e-05, + "loss": 2.4206, + "step": 35121 + }, + { + "epoch": 1.5, + "grad_norm": 0.50390625, + "learning_rate": 2.978194577172806e-05, + "loss": 2.4763, + "step": 35122 + }, + { + "epoch": 1.5, + "grad_norm": 0.51953125, + "learning_rate": 2.9777169889590507e-05, + "loss": 2.4018, + "step": 35123 + }, + { + "epoch": 1.5, + "grad_norm": 0.4375, + "learning_rate": 2.9772394323430054e-05, + "loss": 2.5269, + "step": 35124 + }, + { + "epoch": 1.5, + "grad_norm": 0.4765625, + "learning_rate": 2.9767619073268138e-05, + "loss": 2.5161, + "step": 35125 + }, + { + "epoch": 1.5, + "grad_norm": 0.470703125, + "learning_rate": 2.976284413912629e-05, + "loss": 2.4886, + "step": 35126 + }, + { + "epoch": 1.5, + "grad_norm": 0.4609375, + "learning_rate": 2.9758069521025945e-05, + "loss": 2.3171, + "step": 35127 + }, + { + "epoch": 1.5, + "grad_norm": 0.466796875, + "learning_rate": 2.9753295218988654e-05, + "loss": 2.3597, + "step": 35128 + }, + { + "epoch": 1.5, + "grad_norm": 0.53125, + "learning_rate": 2.974852123303582e-05, + "loss": 2.3788, + "step": 35129 + }, + { + "epoch": 1.5, + "grad_norm": 0.58984375, + "learning_rate": 2.9743747563189005e-05, + "loss": 2.6632, + "step": 35130 + }, + { + "epoch": 1.5, + "grad_norm": 0.5390625, + "learning_rate": 2.9738974209469618e-05, + "loss": 2.4775, + "step": 35131 + }, + { + "epoch": 1.5, + "grad_norm": 0.412109375, + "learning_rate": 2.9734201171899177e-05, + "loss": 2.4785, + "step": 35132 + }, + { + "epoch": 1.5, + "grad_norm": 0.515625, + "learning_rate": 2.9729428450499176e-05, + "loss": 2.5524, + "step": 35133 + }, + { + "epoch": 1.5, + "grad_norm": 0.431640625, + "learning_rate": 2.9724656045291065e-05, + "loss": 2.4368, + "step": 35134 + }, + { + "epoch": 1.5, + "grad_norm": 0.447265625, + "learning_rate": 2.9719883956296257e-05, + "loss": 2.4037, + "step": 35135 + }, + { + "epoch": 1.5, + "grad_norm": 0.486328125, + "learning_rate": 2.9715112183536355e-05, + "loss": 2.4603, + "step": 35136 + }, + { + "epoch": 1.5, + "grad_norm": 0.53515625, + "learning_rate": 2.9710340727032748e-05, + "loss": 2.4545, + "step": 35137 + }, + { + "epoch": 1.5, + "grad_norm": 0.482421875, + "learning_rate": 2.9705569586806903e-05, + "loss": 2.4422, + "step": 35138 + }, + { + "epoch": 1.5, + "grad_norm": 0.5078125, + "learning_rate": 2.97007987628803e-05, + "loss": 2.4099, + "step": 35139 + }, + { + "epoch": 1.5, + "grad_norm": 0.494140625, + "learning_rate": 2.9696028255274434e-05, + "loss": 2.4333, + "step": 35140 + }, + { + "epoch": 1.5, + "grad_norm": 0.443359375, + "learning_rate": 2.969125806401073e-05, + "loss": 2.3092, + "step": 35141 + }, + { + "epoch": 1.5, + "grad_norm": 0.53515625, + "learning_rate": 2.9686488189110695e-05, + "loss": 2.4232, + "step": 35142 + }, + { + "epoch": 1.5, + "grad_norm": 0.46875, + "learning_rate": 2.9681718630595734e-05, + "loss": 2.413, + "step": 35143 + }, + { + "epoch": 1.5, + "grad_norm": 0.455078125, + "learning_rate": 2.9676949388487373e-05, + "loss": 2.4682, + "step": 35144 + }, + { + "epoch": 1.5, + "grad_norm": 0.46875, + "learning_rate": 2.9672180462807007e-05, + "loss": 2.2939, + "step": 35145 + }, + { + "epoch": 1.5, + "grad_norm": 0.498046875, + "learning_rate": 2.9667411853576165e-05, + "loss": 2.4624, + "step": 35146 + }, + { + "epoch": 1.5, + "grad_norm": 0.4921875, + "learning_rate": 2.966264356081623e-05, + "loss": 2.4049, + "step": 35147 + }, + { + "epoch": 1.5, + "grad_norm": 0.443359375, + "learning_rate": 2.965787558454869e-05, + "loss": 2.5679, + "step": 35148 + }, + { + "epoch": 1.5, + "grad_norm": 0.431640625, + "learning_rate": 2.9653107924795042e-05, + "loss": 2.2579, + "step": 35149 + }, + { + "epoch": 1.5, + "grad_norm": 0.423828125, + "learning_rate": 2.9648340581576696e-05, + "loss": 2.3254, + "step": 35150 + }, + { + "epoch": 1.5, + "grad_norm": 0.400390625, + "learning_rate": 2.964357355491504e-05, + "loss": 2.5121, + "step": 35151 + }, + { + "epoch": 1.5, + "grad_norm": 0.53515625, + "learning_rate": 2.9638806844831656e-05, + "loss": 2.5291, + "step": 35152 + }, + { + "epoch": 1.5, + "grad_norm": 0.48046875, + "learning_rate": 2.9634040451347934e-05, + "loss": 2.5153, + "step": 35153 + }, + { + "epoch": 1.5, + "grad_norm": 0.48046875, + "learning_rate": 2.9629274374485273e-05, + "loss": 2.5237, + "step": 35154 + }, + { + "epoch": 1.5, + "grad_norm": 0.51171875, + "learning_rate": 2.962450861426517e-05, + "loss": 2.4671, + "step": 35155 + }, + { + "epoch": 1.5, + "grad_norm": 0.5078125, + "learning_rate": 2.961974317070908e-05, + "loss": 2.4632, + "step": 35156 + }, + { + "epoch": 1.5, + "grad_norm": 0.431640625, + "learning_rate": 2.9614978043838394e-05, + "loss": 2.4103, + "step": 35157 + }, + { + "epoch": 1.5, + "grad_norm": 0.443359375, + "learning_rate": 2.9610213233674624e-05, + "loss": 2.4695, + "step": 35158 + }, + { + "epoch": 1.5, + "grad_norm": 0.439453125, + "learning_rate": 2.960544874023913e-05, + "loss": 2.4024, + "step": 35159 + }, + { + "epoch": 1.5, + "grad_norm": 0.44140625, + "learning_rate": 2.9600684563553428e-05, + "loss": 2.4014, + "step": 35160 + }, + { + "epoch": 1.5, + "grad_norm": 0.416015625, + "learning_rate": 2.959592070363888e-05, + "loss": 2.3668, + "step": 35161 + }, + { + "epoch": 1.5, + "grad_norm": 0.48046875, + "learning_rate": 2.959115716051699e-05, + "loss": 2.3771, + "step": 35162 + }, + { + "epoch": 1.5, + "grad_norm": 0.474609375, + "learning_rate": 2.9586393934209134e-05, + "loss": 2.3425, + "step": 35163 + }, + { + "epoch": 1.5, + "grad_norm": 0.484375, + "learning_rate": 2.9581631024736767e-05, + "loss": 2.3631, + "step": 35164 + }, + { + "epoch": 1.5, + "grad_norm": 0.419921875, + "learning_rate": 2.9576868432121363e-05, + "loss": 2.387, + "step": 35165 + }, + { + "epoch": 1.5, + "grad_norm": 0.5234375, + "learning_rate": 2.9572106156384316e-05, + "loss": 2.4597, + "step": 35166 + }, + { + "epoch": 1.5, + "grad_norm": 0.478515625, + "learning_rate": 2.956734419754702e-05, + "loss": 2.3678, + "step": 35167 + }, + { + "epoch": 1.5, + "grad_norm": 0.421875, + "learning_rate": 2.9562582555630924e-05, + "loss": 2.3228, + "step": 35168 + }, + { + "epoch": 1.5, + "grad_norm": 0.490234375, + "learning_rate": 2.955782123065751e-05, + "loss": 2.5717, + "step": 35169 + }, + { + "epoch": 1.5, + "grad_norm": 0.50390625, + "learning_rate": 2.955306022264812e-05, + "loss": 2.4131, + "step": 35170 + }, + { + "epoch": 1.5, + "grad_norm": 0.494140625, + "learning_rate": 2.9548299531624248e-05, + "loss": 2.4557, + "step": 35171 + }, + { + "epoch": 1.5, + "grad_norm": 0.53125, + "learning_rate": 2.9543539157607247e-05, + "loss": 2.5393, + "step": 35172 + }, + { + "epoch": 1.5, + "grad_norm": 0.5078125, + "learning_rate": 2.9538779100618564e-05, + "loss": 2.6355, + "step": 35173 + }, + { + "epoch": 1.5, + "grad_norm": 0.439453125, + "learning_rate": 2.9534019360679667e-05, + "loss": 2.433, + "step": 35174 + }, + { + "epoch": 1.5, + "grad_norm": 0.50390625, + "learning_rate": 2.9529259937811924e-05, + "loss": 2.3026, + "step": 35175 + }, + { + "epoch": 1.5, + "grad_norm": 0.5, + "learning_rate": 2.952450083203673e-05, + "loss": 2.3756, + "step": 35176 + }, + { + "epoch": 1.5, + "grad_norm": 0.5, + "learning_rate": 2.951974204337552e-05, + "loss": 2.4189, + "step": 35177 + }, + { + "epoch": 1.5, + "grad_norm": 0.40234375, + "learning_rate": 2.9514983571849752e-05, + "loss": 2.3, + "step": 35178 + }, + { + "epoch": 1.5, + "grad_norm": 0.5, + "learning_rate": 2.9510225417480763e-05, + "loss": 2.4284, + "step": 35179 + }, + { + "epoch": 1.5, + "grad_norm": 0.408203125, + "learning_rate": 2.9505467580290004e-05, + "loss": 2.4231, + "step": 35180 + }, + { + "epoch": 1.5, + "grad_norm": 0.466796875, + "learning_rate": 2.9500710060298898e-05, + "loss": 2.432, + "step": 35181 + }, + { + "epoch": 1.5, + "grad_norm": 0.416015625, + "learning_rate": 2.949595285752884e-05, + "loss": 2.545, + "step": 35182 + }, + { + "epoch": 1.5, + "grad_norm": 0.49609375, + "learning_rate": 2.9491195972001196e-05, + "loss": 2.4567, + "step": 35183 + }, + { + "epoch": 1.5, + "grad_norm": 0.5234375, + "learning_rate": 2.948643940373741e-05, + "loss": 2.4697, + "step": 35184 + }, + { + "epoch": 1.5, + "grad_norm": 0.44921875, + "learning_rate": 2.94816831527589e-05, + "loss": 2.3693, + "step": 35185 + }, + { + "epoch": 1.5, + "grad_norm": 0.451171875, + "learning_rate": 2.9476927219087003e-05, + "loss": 2.3791, + "step": 35186 + }, + { + "epoch": 1.5, + "grad_norm": 0.578125, + "learning_rate": 2.9472171602743203e-05, + "loss": 2.5231, + "step": 35187 + }, + { + "epoch": 1.5, + "grad_norm": 0.4921875, + "learning_rate": 2.946741630374883e-05, + "loss": 2.4705, + "step": 35188 + }, + { + "epoch": 1.5, + "grad_norm": 0.48828125, + "learning_rate": 2.9462661322125297e-05, + "loss": 2.4338, + "step": 35189 + }, + { + "epoch": 1.5, + "grad_norm": 0.51171875, + "learning_rate": 2.9457906657894042e-05, + "loss": 2.3907, + "step": 35190 + }, + { + "epoch": 1.5, + "grad_norm": 0.45703125, + "learning_rate": 2.9453152311076428e-05, + "loss": 2.4479, + "step": 35191 + }, + { + "epoch": 1.5, + "grad_norm": 0.458984375, + "learning_rate": 2.9448398281693822e-05, + "loss": 2.5182, + "step": 35192 + }, + { + "epoch": 1.5, + "grad_norm": 0.515625, + "learning_rate": 2.9443644569767626e-05, + "loss": 2.4858, + "step": 35193 + }, + { + "epoch": 1.5, + "grad_norm": 0.470703125, + "learning_rate": 2.943889117531928e-05, + "loss": 2.5615, + "step": 35194 + }, + { + "epoch": 1.5, + "grad_norm": 0.4140625, + "learning_rate": 2.943413809837011e-05, + "loss": 2.3986, + "step": 35195 + }, + { + "epoch": 1.5, + "grad_norm": 0.474609375, + "learning_rate": 2.942938533894154e-05, + "loss": 2.3617, + "step": 35196 + }, + { + "epoch": 1.5, + "grad_norm": 0.46875, + "learning_rate": 2.942463289705496e-05, + "loss": 2.3831, + "step": 35197 + }, + { + "epoch": 1.5, + "grad_norm": 0.53125, + "learning_rate": 2.9419880772731744e-05, + "loss": 2.5547, + "step": 35198 + }, + { + "epoch": 1.5, + "grad_norm": 0.49609375, + "learning_rate": 2.9415128965993245e-05, + "loss": 2.5955, + "step": 35199 + }, + { + "epoch": 1.5, + "grad_norm": 0.43359375, + "learning_rate": 2.9410377476860863e-05, + "loss": 2.5955, + "step": 35200 + }, + { + "epoch": 1.5, + "grad_norm": 0.419921875, + "learning_rate": 2.9405626305356014e-05, + "loss": 2.4599, + "step": 35201 + }, + { + "epoch": 1.5, + "grad_norm": 0.451171875, + "learning_rate": 2.940087545150002e-05, + "loss": 2.3336, + "step": 35202 + }, + { + "epoch": 1.5, + "eval_loss": 2.496812343597412, + "eval_runtime": 82.432, + "eval_samples_per_second": 18.245, + "eval_steps_per_second": 2.281, + "step": 35202 + }, + { + "epoch": 1.5, + "grad_norm": 0.466796875, + "learning_rate": 2.9396124915314316e-05, + "loss": 2.4868, + "step": 35203 + }, + { + "epoch": 1.5, + "grad_norm": 0.484375, + "learning_rate": 2.9391374696820216e-05, + "loss": 2.4453, + "step": 35204 + }, + { + "epoch": 1.5, + "grad_norm": 0.498046875, + "learning_rate": 2.938662479603913e-05, + "loss": 2.4851, + "step": 35205 + }, + { + "epoch": 1.5, + "grad_norm": 0.478515625, + "learning_rate": 2.9381875212992447e-05, + "loss": 2.4465, + "step": 35206 + }, + { + "epoch": 1.5, + "grad_norm": 0.423828125, + "learning_rate": 2.9377125947701534e-05, + "loss": 2.3288, + "step": 35207 + }, + { + "epoch": 1.5, + "grad_norm": 0.44140625, + "learning_rate": 2.93723770001877e-05, + "loss": 2.4313, + "step": 35208 + }, + { + "epoch": 1.5, + "grad_norm": 0.462890625, + "learning_rate": 2.9367628370472356e-05, + "loss": 2.4022, + "step": 35209 + }, + { + "epoch": 1.5, + "grad_norm": 0.390625, + "learning_rate": 2.9362880058576914e-05, + "loss": 2.3678, + "step": 35210 + }, + { + "epoch": 1.5, + "grad_norm": 0.44921875, + "learning_rate": 2.935813206452266e-05, + "loss": 2.6255, + "step": 35211 + }, + { + "epoch": 1.5, + "grad_norm": 0.478515625, + "learning_rate": 2.9353384388330995e-05, + "loss": 2.5624, + "step": 35212 + }, + { + "epoch": 1.5, + "grad_norm": 0.45703125, + "learning_rate": 2.9348637030023306e-05, + "loss": 2.509, + "step": 35213 + }, + { + "epoch": 1.5, + "grad_norm": 0.439453125, + "learning_rate": 2.9343889989620944e-05, + "loss": 2.4688, + "step": 35214 + }, + { + "epoch": 1.5, + "grad_norm": 0.50390625, + "learning_rate": 2.933914326714521e-05, + "loss": 2.4841, + "step": 35215 + }, + { + "epoch": 1.5, + "grad_norm": 0.5078125, + "learning_rate": 2.933439686261752e-05, + "loss": 2.5126, + "step": 35216 + }, + { + "epoch": 1.5, + "grad_norm": 0.416015625, + "learning_rate": 2.9329650776059238e-05, + "loss": 2.5057, + "step": 35217 + }, + { + "epoch": 1.5, + "grad_norm": 0.44921875, + "learning_rate": 2.9324905007491676e-05, + "loss": 2.5151, + "step": 35218 + }, + { + "epoch": 1.5, + "grad_norm": 0.404296875, + "learning_rate": 2.9320159556936244e-05, + "loss": 2.548, + "step": 35219 + }, + { + "epoch": 1.5, + "grad_norm": 0.423828125, + "learning_rate": 2.9315414424414233e-05, + "loss": 2.5364, + "step": 35220 + }, + { + "epoch": 1.5, + "grad_norm": 0.470703125, + "learning_rate": 2.931066960994703e-05, + "loss": 2.3856, + "step": 35221 + }, + { + "epoch": 1.5, + "grad_norm": 0.396484375, + "learning_rate": 2.9305925113556e-05, + "loss": 2.3895, + "step": 35222 + }, + { + "epoch": 1.5, + "grad_norm": 0.470703125, + "learning_rate": 2.9301180935262473e-05, + "loss": 2.4302, + "step": 35223 + }, + { + "epoch": 1.5, + "grad_norm": 0.5234375, + "learning_rate": 2.9296437075087767e-05, + "loss": 2.4933, + "step": 35224 + }, + { + "epoch": 1.5, + "grad_norm": 0.4765625, + "learning_rate": 2.9291693533053253e-05, + "loss": 2.4355, + "step": 35225 + }, + { + "epoch": 1.5, + "grad_norm": 0.515625, + "learning_rate": 2.9286950309180306e-05, + "loss": 2.505, + "step": 35226 + }, + { + "epoch": 1.5, + "grad_norm": 0.51953125, + "learning_rate": 2.9282207403490214e-05, + "loss": 2.4165, + "step": 35227 + }, + { + "epoch": 1.5, + "grad_norm": 0.498046875, + "learning_rate": 2.9277464816004342e-05, + "loss": 2.7651, + "step": 35228 + }, + { + "epoch": 1.5, + "grad_norm": 0.5, + "learning_rate": 2.927272254674407e-05, + "loss": 2.4451, + "step": 35229 + }, + { + "epoch": 1.5, + "grad_norm": 0.46484375, + "learning_rate": 2.9267980595730693e-05, + "loss": 2.5112, + "step": 35230 + }, + { + "epoch": 1.5, + "grad_norm": 0.4609375, + "learning_rate": 2.926323896298552e-05, + "loss": 2.4978, + "step": 35231 + }, + { + "epoch": 1.5, + "grad_norm": 0.40625, + "learning_rate": 2.9258497648529915e-05, + "loss": 2.4444, + "step": 35232 + }, + { + "epoch": 1.5, + "grad_norm": 0.5078125, + "learning_rate": 2.9253756652385254e-05, + "loss": 2.5189, + "step": 35233 + }, + { + "epoch": 1.5, + "grad_norm": 0.546875, + "learning_rate": 2.9249015974572814e-05, + "loss": 2.5821, + "step": 35234 + }, + { + "epoch": 1.5, + "grad_norm": 0.546875, + "learning_rate": 2.924427561511396e-05, + "loss": 2.4828, + "step": 35235 + }, + { + "epoch": 1.5, + "grad_norm": 0.482421875, + "learning_rate": 2.9239535574029986e-05, + "loss": 2.4337, + "step": 35236 + }, + { + "epoch": 1.5, + "grad_norm": 0.474609375, + "learning_rate": 2.9234795851342234e-05, + "loss": 2.4544, + "step": 35237 + }, + { + "epoch": 1.5, + "grad_norm": 0.54296875, + "learning_rate": 2.9230056447072083e-05, + "loss": 2.4823, + "step": 35238 + }, + { + "epoch": 1.5, + "grad_norm": 0.4921875, + "learning_rate": 2.92253173612408e-05, + "loss": 2.5599, + "step": 35239 + }, + { + "epoch": 1.5, + "grad_norm": 0.4453125, + "learning_rate": 2.9220578593869706e-05, + "loss": 2.41, + "step": 35240 + }, + { + "epoch": 1.5, + "grad_norm": 0.478515625, + "learning_rate": 2.9215840144980133e-05, + "loss": 2.3951, + "step": 35241 + }, + { + "epoch": 1.5, + "grad_norm": 0.515625, + "learning_rate": 2.921110201459344e-05, + "loss": 2.4529, + "step": 35242 + }, + { + "epoch": 1.5, + "grad_norm": 0.4765625, + "learning_rate": 2.9206364202730895e-05, + "loss": 2.4456, + "step": 35243 + }, + { + "epoch": 1.5, + "grad_norm": 0.48046875, + "learning_rate": 2.9201626709413833e-05, + "loss": 2.4541, + "step": 35244 + }, + { + "epoch": 1.5, + "grad_norm": 0.4609375, + "learning_rate": 2.9196889534663607e-05, + "loss": 2.4562, + "step": 35245 + }, + { + "epoch": 1.5, + "grad_norm": 0.4765625, + "learning_rate": 2.9192152678501506e-05, + "loss": 2.4075, + "step": 35246 + }, + { + "epoch": 1.5, + "grad_norm": 0.50390625, + "learning_rate": 2.9187416140948808e-05, + "loss": 2.5038, + "step": 35247 + }, + { + "epoch": 1.5, + "grad_norm": 0.490234375, + "learning_rate": 2.9182679922026856e-05, + "loss": 2.4071, + "step": 35248 + }, + { + "epoch": 1.5, + "grad_norm": 0.5546875, + "learning_rate": 2.9177944021756998e-05, + "loss": 2.439, + "step": 35249 + }, + { + "epoch": 1.5, + "grad_norm": 0.53125, + "learning_rate": 2.917320844016047e-05, + "loss": 2.4865, + "step": 35250 + }, + { + "epoch": 1.5, + "grad_norm": 0.5, + "learning_rate": 2.916847317725866e-05, + "loss": 2.4198, + "step": 35251 + }, + { + "epoch": 1.5, + "grad_norm": 0.4296875, + "learning_rate": 2.9163738233072802e-05, + "loss": 2.5229, + "step": 35252 + }, + { + "epoch": 1.5, + "grad_norm": 0.416015625, + "learning_rate": 2.9159003607624236e-05, + "loss": 2.4131, + "step": 35253 + }, + { + "epoch": 1.5, + "grad_norm": 0.486328125, + "learning_rate": 2.915426930093429e-05, + "loss": 2.4189, + "step": 35254 + }, + { + "epoch": 1.5, + "grad_norm": 0.5, + "learning_rate": 2.9149535313024246e-05, + "loss": 2.5281, + "step": 35255 + }, + { + "epoch": 1.5, + "grad_norm": 0.46875, + "learning_rate": 2.9144801643915377e-05, + "loss": 2.5193, + "step": 35256 + }, + { + "epoch": 1.5, + "grad_norm": 0.5, + "learning_rate": 2.9140068293628987e-05, + "loss": 2.4832, + "step": 35257 + }, + { + "epoch": 1.5, + "grad_norm": 0.4453125, + "learning_rate": 2.9135335262186436e-05, + "loss": 2.3418, + "step": 35258 + }, + { + "epoch": 1.5, + "grad_norm": 0.44921875, + "learning_rate": 2.9130602549608943e-05, + "loss": 2.4323, + "step": 35259 + }, + { + "epoch": 1.5, + "grad_norm": 0.41796875, + "learning_rate": 2.9125870155917844e-05, + "loss": 2.4457, + "step": 35260 + }, + { + "epoch": 1.5, + "grad_norm": 0.416015625, + "learning_rate": 2.912113808113446e-05, + "loss": 2.5628, + "step": 35261 + }, + { + "epoch": 1.5, + "grad_norm": 0.498046875, + "learning_rate": 2.9116406325280043e-05, + "loss": 2.5217, + "step": 35262 + }, + { + "epoch": 1.5, + "grad_norm": 0.451171875, + "learning_rate": 2.9111674888375863e-05, + "loss": 2.3938, + "step": 35263 + }, + { + "epoch": 1.5, + "grad_norm": 0.478515625, + "learning_rate": 2.910694377044324e-05, + "loss": 2.5921, + "step": 35264 + }, + { + "epoch": 1.5, + "grad_norm": 0.4453125, + "learning_rate": 2.9102212971503494e-05, + "loss": 2.4611, + "step": 35265 + }, + { + "epoch": 1.5, + "grad_norm": 0.5078125, + "learning_rate": 2.9097482491577856e-05, + "loss": 2.4943, + "step": 35266 + }, + { + "epoch": 1.5, + "grad_norm": 0.49609375, + "learning_rate": 2.9092752330687655e-05, + "loss": 2.3677, + "step": 35267 + }, + { + "epoch": 1.5, + "grad_norm": 0.5703125, + "learning_rate": 2.908802248885413e-05, + "loss": 2.5067, + "step": 35268 + }, + { + "epoch": 1.5, + "grad_norm": 0.44921875, + "learning_rate": 2.9083292966098584e-05, + "loss": 2.4793, + "step": 35269 + }, + { + "epoch": 1.5, + "grad_norm": 0.439453125, + "learning_rate": 2.9078563762442334e-05, + "loss": 2.5831, + "step": 35270 + }, + { + "epoch": 1.5, + "grad_norm": 0.482421875, + "learning_rate": 2.9073834877906637e-05, + "loss": 2.6094, + "step": 35271 + }, + { + "epoch": 1.5, + "grad_norm": 0.4765625, + "learning_rate": 2.9069106312512717e-05, + "loss": 2.4086, + "step": 35272 + }, + { + "epoch": 1.5, + "grad_norm": 0.44921875, + "learning_rate": 2.9064378066281906e-05, + "loss": 2.4083, + "step": 35273 + }, + { + "epoch": 1.5, + "grad_norm": 0.462890625, + "learning_rate": 2.9059650139235495e-05, + "loss": 2.4538, + "step": 35274 + }, + { + "epoch": 1.5, + "grad_norm": 0.462890625, + "learning_rate": 2.9054922531394714e-05, + "loss": 2.3689, + "step": 35275 + }, + { + "epoch": 1.5, + "grad_norm": 0.416015625, + "learning_rate": 2.9050195242780844e-05, + "loss": 2.3854, + "step": 35276 + }, + { + "epoch": 1.5, + "grad_norm": 0.5234375, + "learning_rate": 2.90454682734152e-05, + "loss": 2.4772, + "step": 35277 + }, + { + "epoch": 1.5, + "grad_norm": 0.43359375, + "learning_rate": 2.9040741623319013e-05, + "loss": 2.4513, + "step": 35278 + }, + { + "epoch": 1.5, + "grad_norm": 0.498046875, + "learning_rate": 2.903601529251353e-05, + "loss": 2.4485, + "step": 35279 + }, + { + "epoch": 1.5, + "grad_norm": 0.423828125, + "learning_rate": 2.9031289281020034e-05, + "loss": 2.3906, + "step": 35280 + }, + { + "epoch": 1.5, + "grad_norm": 0.482421875, + "learning_rate": 2.9026563588859846e-05, + "loss": 2.5137, + "step": 35281 + }, + { + "epoch": 1.5, + "grad_norm": 0.51953125, + "learning_rate": 2.9021838216054142e-05, + "loss": 2.3394, + "step": 35282 + }, + { + "epoch": 1.5, + "grad_norm": 0.47265625, + "learning_rate": 2.9017113162624264e-05, + "loss": 2.3779, + "step": 35283 + }, + { + "epoch": 1.5, + "grad_norm": 0.435546875, + "learning_rate": 2.9012388428591398e-05, + "loss": 2.5881, + "step": 35284 + }, + { + "epoch": 1.5, + "grad_norm": 0.462890625, + "learning_rate": 2.9007664013976855e-05, + "loss": 2.4681, + "step": 35285 + }, + { + "epoch": 1.5, + "grad_norm": 0.4609375, + "learning_rate": 2.9002939918801896e-05, + "loss": 2.5001, + "step": 35286 + }, + { + "epoch": 1.5, + "grad_norm": 0.46875, + "learning_rate": 2.8998216143087765e-05, + "loss": 2.4401, + "step": 35287 + }, + { + "epoch": 1.5, + "grad_norm": 0.427734375, + "learning_rate": 2.8993492686855684e-05, + "loss": 2.3291, + "step": 35288 + }, + { + "epoch": 1.5, + "grad_norm": 0.42578125, + "learning_rate": 2.8988769550126938e-05, + "loss": 2.2859, + "step": 35289 + }, + { + "epoch": 1.5, + "grad_norm": 0.53125, + "learning_rate": 2.8984046732922798e-05, + "loss": 2.6346, + "step": 35290 + }, + { + "epoch": 1.5, + "grad_norm": 0.48828125, + "learning_rate": 2.897932423526447e-05, + "loss": 2.4663, + "step": 35291 + }, + { + "epoch": 1.5, + "grad_norm": 0.466796875, + "learning_rate": 2.897460205717326e-05, + "loss": 2.4514, + "step": 35292 + }, + { + "epoch": 1.5, + "grad_norm": 0.439453125, + "learning_rate": 2.8969880198670352e-05, + "loss": 2.4366, + "step": 35293 + }, + { + "epoch": 1.5, + "grad_norm": 0.474609375, + "learning_rate": 2.896515865977706e-05, + "loss": 2.4471, + "step": 35294 + }, + { + "epoch": 1.5, + "grad_norm": 0.4765625, + "learning_rate": 2.8960437440514554e-05, + "loss": 2.4078, + "step": 35295 + }, + { + "epoch": 1.5, + "grad_norm": 0.435546875, + "learning_rate": 2.8955716540904156e-05, + "loss": 2.3254, + "step": 35296 + }, + { + "epoch": 1.5, + "grad_norm": 0.486328125, + "learning_rate": 2.895099596096703e-05, + "loss": 2.5765, + "step": 35297 + }, + { + "epoch": 1.5, + "grad_norm": 0.494140625, + "learning_rate": 2.8946275700724468e-05, + "loss": 2.3882, + "step": 35298 + }, + { + "epoch": 1.5, + "grad_norm": 0.45703125, + "learning_rate": 2.8941555760197726e-05, + "loss": 2.4517, + "step": 35299 + }, + { + "epoch": 1.5, + "grad_norm": 0.490234375, + "learning_rate": 2.893683613940801e-05, + "loss": 2.4901, + "step": 35300 + }, + { + "epoch": 1.5, + "grad_norm": 0.435546875, + "learning_rate": 2.89321168383765e-05, + "loss": 2.4284, + "step": 35301 + }, + { + "epoch": 1.5, + "grad_norm": 0.51171875, + "learning_rate": 2.892739785712455e-05, + "loss": 2.4353, + "step": 35302 + }, + { + "epoch": 1.5, + "grad_norm": 0.4375, + "learning_rate": 2.892267919567334e-05, + "loss": 2.4426, + "step": 35303 + }, + { + "epoch": 1.5, + "grad_norm": 0.51171875, + "learning_rate": 2.891796085404407e-05, + "loss": 2.4498, + "step": 35304 + }, + { + "epoch": 1.5, + "grad_norm": 0.39453125, + "learning_rate": 2.8913242832257993e-05, + "loss": 2.3097, + "step": 35305 + }, + { + "epoch": 1.5, + "grad_norm": 0.431640625, + "learning_rate": 2.8908525130336383e-05, + "loss": 2.4244, + "step": 35306 + }, + { + "epoch": 1.5, + "grad_norm": 0.400390625, + "learning_rate": 2.8903807748300393e-05, + "loss": 2.4894, + "step": 35307 + }, + { + "epoch": 1.5, + "grad_norm": 0.47265625, + "learning_rate": 2.8899090686171316e-05, + "loss": 2.4651, + "step": 35308 + }, + { + "epoch": 1.5, + "grad_norm": 0.4296875, + "learning_rate": 2.889437394397031e-05, + "loss": 2.5228, + "step": 35309 + }, + { + "epoch": 1.5, + "grad_norm": 0.404296875, + "learning_rate": 2.8889657521718673e-05, + "loss": 2.4411, + "step": 35310 + }, + { + "epoch": 1.5, + "grad_norm": 0.47265625, + "learning_rate": 2.888494141943756e-05, + "loss": 2.3992, + "step": 35311 + }, + { + "epoch": 1.5, + "grad_norm": 0.474609375, + "learning_rate": 2.888022563714825e-05, + "loss": 2.4782, + "step": 35312 + }, + { + "epoch": 1.5, + "grad_norm": 0.490234375, + "learning_rate": 2.88755101748719e-05, + "loss": 2.4242, + "step": 35313 + }, + { + "epoch": 1.5, + "grad_norm": 0.431640625, + "learning_rate": 2.887079503262976e-05, + "loss": 2.4123, + "step": 35314 + }, + { + "epoch": 1.5, + "grad_norm": 0.46484375, + "learning_rate": 2.886608021044308e-05, + "loss": 2.5439, + "step": 35315 + }, + { + "epoch": 1.5, + "grad_norm": 0.400390625, + "learning_rate": 2.8861365708333042e-05, + "loss": 2.4054, + "step": 35316 + }, + { + "epoch": 1.5, + "grad_norm": 0.4609375, + "learning_rate": 2.8856651526320812e-05, + "loss": 2.5184, + "step": 35317 + }, + { + "epoch": 1.5, + "grad_norm": 0.515625, + "learning_rate": 2.8851937664427696e-05, + "loss": 2.6001, + "step": 35318 + }, + { + "epoch": 1.51, + "grad_norm": 0.48828125, + "learning_rate": 2.8847224122674866e-05, + "loss": 2.4785, + "step": 35319 + }, + { + "epoch": 1.51, + "grad_norm": 0.439453125, + "learning_rate": 2.8842510901083486e-05, + "loss": 2.4188, + "step": 35320 + }, + { + "epoch": 1.51, + "grad_norm": 0.5234375, + "learning_rate": 2.883779799967481e-05, + "loss": 2.5317, + "step": 35321 + }, + { + "epoch": 1.51, + "grad_norm": 0.44140625, + "learning_rate": 2.8833085418470064e-05, + "loss": 2.4008, + "step": 35322 + }, + { + "epoch": 1.51, + "grad_norm": 0.51171875, + "learning_rate": 2.8828373157490384e-05, + "loss": 2.4855, + "step": 35323 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.882366121675706e-05, + "loss": 2.4747, + "step": 35324 + }, + { + "epoch": 1.51, + "grad_norm": 0.53125, + "learning_rate": 2.8818949596291212e-05, + "loss": 2.4354, + "step": 35325 + }, + { + "epoch": 1.51, + "grad_norm": 0.484375, + "learning_rate": 2.8814238296114106e-05, + "loss": 2.4062, + "step": 35326 + }, + { + "epoch": 1.51, + "grad_norm": 0.49609375, + "learning_rate": 2.8809527316246886e-05, + "loss": 2.4546, + "step": 35327 + }, + { + "epoch": 1.51, + "grad_norm": 0.462890625, + "learning_rate": 2.8804816656710796e-05, + "loss": 2.407, + "step": 35328 + }, + { + "epoch": 1.51, + "grad_norm": 0.46484375, + "learning_rate": 2.8800106317527e-05, + "loss": 2.5019, + "step": 35329 + }, + { + "epoch": 1.51, + "grad_norm": 0.52734375, + "learning_rate": 2.8795396298716694e-05, + "loss": 2.5525, + "step": 35330 + }, + { + "epoch": 1.51, + "grad_norm": 0.43359375, + "learning_rate": 2.879068660030111e-05, + "loss": 2.3827, + "step": 35331 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.8785977222301417e-05, + "loss": 2.3955, + "step": 35332 + }, + { + "epoch": 1.51, + "grad_norm": 0.51953125, + "learning_rate": 2.878126816473874e-05, + "loss": 2.4182, + "step": 35333 + }, + { + "epoch": 1.51, + "grad_norm": 0.435546875, + "learning_rate": 2.8776559427634388e-05, + "loss": 2.4665, + "step": 35334 + }, + { + "epoch": 1.51, + "grad_norm": 0.55078125, + "learning_rate": 2.8771851011009488e-05, + "loss": 2.4214, + "step": 35335 + }, + { + "epoch": 1.51, + "grad_norm": 0.466796875, + "learning_rate": 2.8767142914885202e-05, + "loss": 2.4737, + "step": 35336 + }, + { + "epoch": 1.51, + "grad_norm": 0.482421875, + "learning_rate": 2.876243513928274e-05, + "loss": 2.5313, + "step": 35337 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.8757727684223313e-05, + "loss": 2.4217, + "step": 35338 + }, + { + "epoch": 1.51, + "grad_norm": 0.453125, + "learning_rate": 2.8753020549728037e-05, + "loss": 2.377, + "step": 35339 + }, + { + "epoch": 1.51, + "grad_norm": 0.466796875, + "learning_rate": 2.8748313735818178e-05, + "loss": 2.5063, + "step": 35340 + }, + { + "epoch": 1.51, + "grad_norm": 0.435546875, + "learning_rate": 2.8743607242514826e-05, + "loss": 2.5214, + "step": 35341 + }, + { + "epoch": 1.51, + "grad_norm": 0.494140625, + "learning_rate": 2.873890106983923e-05, + "loss": 2.5617, + "step": 35342 + }, + { + "epoch": 1.51, + "grad_norm": 0.482421875, + "learning_rate": 2.873419521781251e-05, + "loss": 2.4653, + "step": 35343 + }, + { + "epoch": 1.51, + "grad_norm": 0.43359375, + "learning_rate": 2.872948968645589e-05, + "loss": 2.4461, + "step": 35344 + }, + { + "epoch": 1.51, + "grad_norm": 0.453125, + "learning_rate": 2.87247844757905e-05, + "loss": 2.365, + "step": 35345 + }, + { + "epoch": 1.51, + "grad_norm": 0.478515625, + "learning_rate": 2.8720079585837533e-05, + "loss": 2.463, + "step": 35346 + }, + { + "epoch": 1.51, + "grad_norm": 0.4921875, + "learning_rate": 2.8715375016618174e-05, + "loss": 2.4208, + "step": 35347 + }, + { + "epoch": 1.51, + "grad_norm": 0.4453125, + "learning_rate": 2.871067076815359e-05, + "loss": 2.2606, + "step": 35348 + }, + { + "epoch": 1.51, + "grad_norm": 0.447265625, + "learning_rate": 2.870596684046487e-05, + "loss": 2.647, + "step": 35349 + }, + { + "epoch": 1.51, + "grad_norm": 0.4296875, + "learning_rate": 2.8701263233573304e-05, + "loss": 2.3633, + "step": 35350 + }, + { + "epoch": 1.51, + "grad_norm": 0.48046875, + "learning_rate": 2.869655994749999e-05, + "loss": 2.4206, + "step": 35351 + }, + { + "epoch": 1.51, + "grad_norm": 0.443359375, + "learning_rate": 2.8691856982266076e-05, + "loss": 2.3541, + "step": 35352 + }, + { + "epoch": 1.51, + "grad_norm": 0.453125, + "learning_rate": 2.868715433789274e-05, + "loss": 2.4928, + "step": 35353 + }, + { + "epoch": 1.51, + "grad_norm": 0.478515625, + "learning_rate": 2.8682452014401184e-05, + "loss": 2.5557, + "step": 35354 + }, + { + "epoch": 1.51, + "grad_norm": 0.52734375, + "learning_rate": 2.86777500118125e-05, + "loss": 2.4948, + "step": 35355 + }, + { + "epoch": 1.51, + "grad_norm": 0.392578125, + "learning_rate": 2.86730483301479e-05, + "loss": 2.4117, + "step": 35356 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.8668346969428484e-05, + "loss": 2.404, + "step": 35357 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.8663645929675476e-05, + "loss": 2.5039, + "step": 35358 + }, + { + "epoch": 1.51, + "grad_norm": 0.427734375, + "learning_rate": 2.865894521090996e-05, + "loss": 2.4032, + "step": 35359 + }, + { + "epoch": 1.51, + "grad_norm": 0.47265625, + "learning_rate": 2.8654244813153142e-05, + "loss": 2.4571, + "step": 35360 + }, + { + "epoch": 1.51, + "grad_norm": 0.494140625, + "learning_rate": 2.8649544736426126e-05, + "loss": 2.4201, + "step": 35361 + }, + { + "epoch": 1.51, + "grad_norm": 0.4375, + "learning_rate": 2.8644844980750075e-05, + "loss": 2.5016, + "step": 35362 + }, + { + "epoch": 1.51, + "grad_norm": 0.43359375, + "learning_rate": 2.864014554614618e-05, + "loss": 2.3107, + "step": 35363 + }, + { + "epoch": 1.51, + "grad_norm": 0.46484375, + "learning_rate": 2.863544643263555e-05, + "loss": 2.3605, + "step": 35364 + }, + { + "epoch": 1.51, + "grad_norm": 0.44140625, + "learning_rate": 2.8630747640239264e-05, + "loss": 2.5188, + "step": 35365 + }, + { + "epoch": 1.51, + "grad_norm": 0.5703125, + "learning_rate": 2.8626049168978598e-05, + "loss": 2.4757, + "step": 35366 + }, + { + "epoch": 1.51, + "grad_norm": 0.451171875, + "learning_rate": 2.862135101887463e-05, + "loss": 2.5463, + "step": 35367 + }, + { + "epoch": 1.51, + "grad_norm": 0.490234375, + "learning_rate": 2.861665318994846e-05, + "loss": 2.4505, + "step": 35368 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.8611955682221257e-05, + "loss": 2.4488, + "step": 35369 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.8607258495714206e-05, + "loss": 2.5269, + "step": 35370 + }, + { + "epoch": 1.51, + "grad_norm": 0.46484375, + "learning_rate": 2.8602561630448367e-05, + "loss": 2.53, + "step": 35371 + }, + { + "epoch": 1.51, + "grad_norm": 0.42578125, + "learning_rate": 2.859786508644494e-05, + "loss": 2.4686, + "step": 35372 + }, + { + "epoch": 1.51, + "grad_norm": 0.51171875, + "learning_rate": 2.8593168863724994e-05, + "loss": 2.362, + "step": 35373 + }, + { + "epoch": 1.51, + "grad_norm": 0.478515625, + "learning_rate": 2.858847296230972e-05, + "loss": 2.4355, + "step": 35374 + }, + { + "epoch": 1.51, + "grad_norm": 0.490234375, + "learning_rate": 2.8583777382220188e-05, + "loss": 2.4829, + "step": 35375 + }, + { + "epoch": 1.51, + "grad_norm": 0.453125, + "learning_rate": 2.8579082123477596e-05, + "loss": 2.5982, + "step": 35376 + }, + { + "epoch": 1.51, + "grad_norm": 0.45703125, + "learning_rate": 2.857438718610299e-05, + "loss": 2.4432, + "step": 35377 + }, + { + "epoch": 1.51, + "grad_norm": 0.46484375, + "learning_rate": 2.8569692570117558e-05, + "loss": 2.5368, + "step": 35378 + }, + { + "epoch": 1.51, + "grad_norm": 0.48046875, + "learning_rate": 2.856499827554242e-05, + "loss": 2.5632, + "step": 35379 + }, + { + "epoch": 1.51, + "grad_norm": 0.47265625, + "learning_rate": 2.856030430239869e-05, + "loss": 2.4048, + "step": 35380 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.855561065070741e-05, + "loss": 2.3862, + "step": 35381 + }, + { + "epoch": 1.51, + "grad_norm": 0.486328125, + "learning_rate": 2.8550917320489845e-05, + "loss": 2.3548, + "step": 35382 + }, + { + "epoch": 1.51, + "grad_norm": 0.4921875, + "learning_rate": 2.8546224311767034e-05, + "loss": 2.5162, + "step": 35383 + }, + { + "epoch": 1.51, + "grad_norm": 0.5546875, + "learning_rate": 2.8541531624560068e-05, + "loss": 2.528, + "step": 35384 + }, + { + "epoch": 1.51, + "grad_norm": 0.46875, + "learning_rate": 2.8536839258890102e-05, + "loss": 2.4312, + "step": 35385 + }, + { + "epoch": 1.51, + "grad_norm": 0.474609375, + "learning_rate": 2.853214721477827e-05, + "loss": 2.3752, + "step": 35386 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.8527455492245626e-05, + "loss": 2.4632, + "step": 35387 + }, + { + "epoch": 1.51, + "grad_norm": 0.490234375, + "learning_rate": 2.852276409131335e-05, + "loss": 2.4669, + "step": 35388 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.8518073012002467e-05, + "loss": 2.4177, + "step": 35389 + }, + { + "epoch": 1.51, + "grad_norm": 0.451171875, + "learning_rate": 2.8513382254334176e-05, + "loss": 2.4743, + "step": 35390 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.8508691818329504e-05, + "loss": 2.4917, + "step": 35391 + }, + { + "epoch": 1.51, + "grad_norm": 0.439453125, + "learning_rate": 2.850400170400962e-05, + "loss": 2.462, + "step": 35392 + }, + { + "epoch": 1.51, + "grad_norm": 0.4609375, + "learning_rate": 2.8499311911395576e-05, + "loss": 2.3819, + "step": 35393 + }, + { + "epoch": 1.51, + "grad_norm": 0.45703125, + "learning_rate": 2.849462244050849e-05, + "loss": 2.463, + "step": 35394 + }, + { + "epoch": 1.51, + "grad_norm": 0.52734375, + "learning_rate": 2.8489933291369508e-05, + "loss": 2.4127, + "step": 35395 + }, + { + "epoch": 1.51, + "grad_norm": 0.455078125, + "learning_rate": 2.848524446399968e-05, + "loss": 2.5569, + "step": 35396 + }, + { + "epoch": 1.51, + "grad_norm": 0.484375, + "learning_rate": 2.8480555958420074e-05, + "loss": 2.6031, + "step": 35397 + }, + { + "epoch": 1.51, + "grad_norm": 0.439453125, + "learning_rate": 2.8475867774651877e-05, + "loss": 2.3444, + "step": 35398 + }, + { + "epoch": 1.51, + "grad_norm": 0.4765625, + "learning_rate": 2.847117991271614e-05, + "loss": 2.5178, + "step": 35399 + }, + { + "epoch": 1.51, + "grad_norm": 0.4765625, + "learning_rate": 2.8466492372633925e-05, + "loss": 2.3769, + "step": 35400 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.8461805154426347e-05, + "loss": 2.5254, + "step": 35401 + }, + { + "epoch": 1.51, + "grad_norm": 0.484375, + "learning_rate": 2.8457118258114536e-05, + "loss": 2.4752, + "step": 35402 + }, + { + "epoch": 1.51, + "grad_norm": 0.5, + "learning_rate": 2.8452431683719503e-05, + "loss": 2.556, + "step": 35403 + }, + { + "epoch": 1.51, + "grad_norm": 0.478515625, + "learning_rate": 2.844774543126243e-05, + "loss": 2.4677, + "step": 35404 + }, + { + "epoch": 1.51, + "grad_norm": 0.46484375, + "learning_rate": 2.8443059500764313e-05, + "loss": 2.5572, + "step": 35405 + }, + { + "epoch": 1.51, + "grad_norm": 0.5546875, + "learning_rate": 2.8438373892246307e-05, + "loss": 2.5267, + "step": 35406 + }, + { + "epoch": 1.51, + "grad_norm": 0.462890625, + "learning_rate": 2.8433688605729435e-05, + "loss": 2.489, + "step": 35407 + }, + { + "epoch": 1.51, + "grad_norm": 0.53515625, + "learning_rate": 2.842900364123484e-05, + "loss": 2.583, + "step": 35408 + }, + { + "epoch": 1.51, + "grad_norm": 0.55078125, + "learning_rate": 2.842431899878355e-05, + "loss": 2.4173, + "step": 35409 + }, + { + "epoch": 1.51, + "grad_norm": 0.4921875, + "learning_rate": 2.8419634678396657e-05, + "loss": 2.397, + "step": 35410 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.8414950680095277e-05, + "loss": 2.4093, + "step": 35411 + }, + { + "epoch": 1.51, + "grad_norm": 0.4453125, + "learning_rate": 2.841026700390047e-05, + "loss": 2.4475, + "step": 35412 + }, + { + "epoch": 1.51, + "grad_norm": 0.53125, + "learning_rate": 2.8405583649833257e-05, + "loss": 2.5615, + "step": 35413 + }, + { + "epoch": 1.51, + "grad_norm": 0.4296875, + "learning_rate": 2.840090061791476e-05, + "loss": 2.466, + "step": 35414 + }, + { + "epoch": 1.51, + "grad_norm": 0.443359375, + "learning_rate": 2.8396217908166067e-05, + "loss": 2.5312, + "step": 35415 + }, + { + "epoch": 1.51, + "grad_norm": 0.462890625, + "learning_rate": 2.839153552060819e-05, + "loss": 2.505, + "step": 35416 + }, + { + "epoch": 1.51, + "grad_norm": 0.515625, + "learning_rate": 2.8386853455262275e-05, + "loss": 2.4641, + "step": 35417 + }, + { + "epoch": 1.51, + "grad_norm": 0.5546875, + "learning_rate": 2.8382171712149318e-05, + "loss": 2.5569, + "step": 35418 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.837749029129041e-05, + "loss": 2.4301, + "step": 35419 + }, + { + "epoch": 1.51, + "grad_norm": 0.48046875, + "learning_rate": 2.837280919270665e-05, + "loss": 2.5031, + "step": 35420 + }, + { + "epoch": 1.51, + "grad_norm": 0.4375, + "learning_rate": 2.8368128416419083e-05, + "loss": 2.4247, + "step": 35421 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.836344796244872e-05, + "loss": 2.5688, + "step": 35422 + }, + { + "epoch": 1.51, + "grad_norm": 0.4609375, + "learning_rate": 2.8358767830816657e-05, + "loss": 2.5837, + "step": 35423 + }, + { + "epoch": 1.51, + "grad_norm": 0.427734375, + "learning_rate": 2.835408802154399e-05, + "loss": 2.4042, + "step": 35424 + }, + { + "epoch": 1.51, + "grad_norm": 0.5546875, + "learning_rate": 2.834940853465172e-05, + "loss": 2.4874, + "step": 35425 + }, + { + "epoch": 1.51, + "grad_norm": 0.43359375, + "learning_rate": 2.8344729370160915e-05, + "loss": 2.3609, + "step": 35426 + }, + { + "epoch": 1.51, + "grad_norm": 0.421875, + "learning_rate": 2.8340050528092676e-05, + "loss": 2.3728, + "step": 35427 + }, + { + "epoch": 1.51, + "grad_norm": 0.484375, + "learning_rate": 2.8335372008468032e-05, + "loss": 2.2375, + "step": 35428 + }, + { + "epoch": 1.51, + "grad_norm": 0.4609375, + "learning_rate": 2.8330693811307985e-05, + "loss": 2.4931, + "step": 35429 + }, + { + "epoch": 1.51, + "grad_norm": 0.5234375, + "learning_rate": 2.832601593663362e-05, + "loss": 2.5159, + "step": 35430 + }, + { + "epoch": 1.51, + "grad_norm": 0.478515625, + "learning_rate": 2.8321338384466022e-05, + "loss": 2.3696, + "step": 35431 + }, + { + "epoch": 1.51, + "grad_norm": 0.46875, + "learning_rate": 2.8316661154826174e-05, + "loss": 2.441, + "step": 35432 + }, + { + "epoch": 1.51, + "grad_norm": 0.451171875, + "learning_rate": 2.831198424773518e-05, + "loss": 2.4355, + "step": 35433 + }, + { + "epoch": 1.51, + "grad_norm": 0.435546875, + "learning_rate": 2.8307307663214033e-05, + "loss": 2.5027, + "step": 35434 + }, + { + "epoch": 1.51, + "grad_norm": 0.462890625, + "learning_rate": 2.8302631401283807e-05, + "loss": 2.4171, + "step": 35435 + }, + { + "epoch": 1.51, + "grad_norm": 0.4453125, + "learning_rate": 2.829795546196555e-05, + "loss": 2.4495, + "step": 35436 + }, + { + "epoch": 1.51, + "grad_norm": 0.4765625, + "learning_rate": 2.8293279845280306e-05, + "loss": 2.5799, + "step": 35437 + }, + { + "epoch": 1.51, + "grad_norm": 0.52734375, + "learning_rate": 2.828860455124903e-05, + "loss": 2.4442, + "step": 35438 + }, + { + "epoch": 1.51, + "grad_norm": 0.419921875, + "learning_rate": 2.8283929579892886e-05, + "loss": 2.4471, + "step": 35439 + }, + { + "epoch": 1.51, + "grad_norm": 0.44140625, + "learning_rate": 2.8279254931232857e-05, + "loss": 2.3088, + "step": 35440 + }, + { + "epoch": 1.51, + "grad_norm": 0.451171875, + "learning_rate": 2.8274580605289934e-05, + "loss": 2.3838, + "step": 35441 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.8269906602085183e-05, + "loss": 2.5728, + "step": 35442 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.8265232921639673e-05, + "loss": 2.6483, + "step": 35443 + }, + { + "epoch": 1.51, + "grad_norm": 0.484375, + "learning_rate": 2.8260559563974397e-05, + "loss": 2.5129, + "step": 35444 + }, + { + "epoch": 1.51, + "grad_norm": 0.4765625, + "learning_rate": 2.8255886529110352e-05, + "loss": 2.3442, + "step": 35445 + }, + { + "epoch": 1.51, + "grad_norm": 0.48046875, + "learning_rate": 2.82512138170686e-05, + "loss": 2.4696, + "step": 35446 + }, + { + "epoch": 1.51, + "grad_norm": 0.4609375, + "learning_rate": 2.82465414278702e-05, + "loss": 2.4749, + "step": 35447 + }, + { + "epoch": 1.51, + "grad_norm": 0.419921875, + "learning_rate": 2.8241869361536098e-05, + "loss": 2.3686, + "step": 35448 + }, + { + "epoch": 1.51, + "grad_norm": 0.412109375, + "learning_rate": 2.8237197618087395e-05, + "loss": 2.2757, + "step": 35449 + }, + { + "epoch": 1.51, + "grad_norm": 0.5234375, + "learning_rate": 2.823252619754505e-05, + "loss": 2.5555, + "step": 35450 + }, + { + "epoch": 1.51, + "grad_norm": 0.498046875, + "learning_rate": 2.8227855099930113e-05, + "loss": 2.4711, + "step": 35451 + }, + { + "epoch": 1.51, + "grad_norm": 0.42578125, + "learning_rate": 2.8223184325263617e-05, + "loss": 2.3573, + "step": 35452 + }, + { + "epoch": 1.51, + "grad_norm": 0.51171875, + "learning_rate": 2.8218513873566566e-05, + "loss": 2.5643, + "step": 35453 + }, + { + "epoch": 1.51, + "grad_norm": 0.466796875, + "learning_rate": 2.8213843744859915e-05, + "loss": 2.3763, + "step": 35454 + }, + { + "epoch": 1.51, + "grad_norm": 0.419921875, + "learning_rate": 2.820917393916479e-05, + "loss": 2.5116, + "step": 35455 + }, + { + "epoch": 1.51, + "grad_norm": 0.404296875, + "learning_rate": 2.8204504456502134e-05, + "loss": 2.3554, + "step": 35456 + }, + { + "epoch": 1.51, + "grad_norm": 0.51953125, + "learning_rate": 2.8199835296892952e-05, + "loss": 2.4336, + "step": 35457 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.8195166460358268e-05, + "loss": 2.4618, + "step": 35458 + }, + { + "epoch": 1.51, + "grad_norm": 0.44921875, + "learning_rate": 2.8190497946919127e-05, + "loss": 2.4642, + "step": 35459 + }, + { + "epoch": 1.51, + "grad_norm": 0.4296875, + "learning_rate": 2.8185829756596493e-05, + "loss": 2.384, + "step": 35460 + }, + { + "epoch": 1.51, + "grad_norm": 0.55078125, + "learning_rate": 2.8181161889411355e-05, + "loss": 2.5037, + "step": 35461 + }, + { + "epoch": 1.51, + "grad_norm": 0.4609375, + "learning_rate": 2.8176494345384742e-05, + "loss": 2.4681, + "step": 35462 + }, + { + "epoch": 1.51, + "grad_norm": 0.4296875, + "learning_rate": 2.8171827124537686e-05, + "loss": 2.4352, + "step": 35463 + }, + { + "epoch": 1.51, + "grad_norm": 0.4453125, + "learning_rate": 2.816716022689112e-05, + "loss": 2.5132, + "step": 35464 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.8162493652466126e-05, + "loss": 2.3372, + "step": 35465 + }, + { + "epoch": 1.51, + "grad_norm": 0.5234375, + "learning_rate": 2.815782740128362e-05, + "loss": 2.4861, + "step": 35466 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.815316147336463e-05, + "loss": 2.3621, + "step": 35467 + }, + { + "epoch": 1.51, + "grad_norm": 0.4921875, + "learning_rate": 2.8148495868730197e-05, + "loss": 2.4983, + "step": 35468 + }, + { + "epoch": 1.51, + "grad_norm": 0.4375, + "learning_rate": 2.8143830587401264e-05, + "loss": 2.5887, + "step": 35469 + }, + { + "epoch": 1.51, + "grad_norm": 0.4140625, + "learning_rate": 2.813916562939878e-05, + "loss": 2.4492, + "step": 35470 + }, + { + "epoch": 1.51, + "grad_norm": 0.482421875, + "learning_rate": 2.813450099474384e-05, + "loss": 2.3625, + "step": 35471 + }, + { + "epoch": 1.51, + "grad_norm": 0.462890625, + "learning_rate": 2.8129836683457388e-05, + "loss": 2.5613, + "step": 35472 + }, + { + "epoch": 1.51, + "grad_norm": 0.453125, + "learning_rate": 2.8125172695560376e-05, + "loss": 2.4534, + "step": 35473 + }, + { + "epoch": 1.51, + "grad_norm": 0.48046875, + "learning_rate": 2.8120509031073818e-05, + "loss": 2.4399, + "step": 35474 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.8115845690018727e-05, + "loss": 2.4451, + "step": 35475 + }, + { + "epoch": 1.51, + "grad_norm": 0.5234375, + "learning_rate": 2.811118267241606e-05, + "loss": 2.537, + "step": 35476 + }, + { + "epoch": 1.51, + "grad_norm": 0.443359375, + "learning_rate": 2.8106519978286772e-05, + "loss": 2.49, + "step": 35477 + }, + { + "epoch": 1.51, + "grad_norm": 0.486328125, + "learning_rate": 2.810185760765187e-05, + "loss": 2.4089, + "step": 35478 + }, + { + "epoch": 1.51, + "grad_norm": 0.49609375, + "learning_rate": 2.8097195560532363e-05, + "loss": 2.5447, + "step": 35479 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.8092533836949175e-05, + "loss": 2.501, + "step": 35480 + }, + { + "epoch": 1.51, + "grad_norm": 0.490234375, + "learning_rate": 2.8087872436923324e-05, + "loss": 2.4769, + "step": 35481 + }, + { + "epoch": 1.51, + "grad_norm": 0.486328125, + "learning_rate": 2.8083211360475737e-05, + "loss": 2.3845, + "step": 35482 + }, + { + "epoch": 1.51, + "grad_norm": 0.498046875, + "learning_rate": 2.807855060762742e-05, + "loss": 2.4911, + "step": 35483 + }, + { + "epoch": 1.51, + "grad_norm": 0.41796875, + "learning_rate": 2.8073890178399377e-05, + "loss": 2.4978, + "step": 35484 + }, + { + "epoch": 1.51, + "grad_norm": 0.47265625, + "learning_rate": 2.8069230072812535e-05, + "loss": 2.4548, + "step": 35485 + }, + { + "epoch": 1.51, + "grad_norm": 0.466796875, + "learning_rate": 2.8064570290887814e-05, + "loss": 2.5124, + "step": 35486 + }, + { + "epoch": 1.51, + "grad_norm": 0.419921875, + "learning_rate": 2.8059910832646285e-05, + "loss": 2.499, + "step": 35487 + }, + { + "epoch": 1.51, + "grad_norm": 0.48828125, + "learning_rate": 2.8055251698108887e-05, + "loss": 2.4694, + "step": 35488 + }, + { + "epoch": 1.51, + "grad_norm": 0.462890625, + "learning_rate": 2.8050592887296513e-05, + "loss": 2.3803, + "step": 35489 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.8045934400230188e-05, + "loss": 2.379, + "step": 35490 + }, + { + "epoch": 1.51, + "grad_norm": 0.462890625, + "learning_rate": 2.804127623693089e-05, + "loss": 2.4562, + "step": 35491 + }, + { + "epoch": 1.51, + "grad_norm": 0.48828125, + "learning_rate": 2.8036618397419522e-05, + "loss": 2.6179, + "step": 35492 + }, + { + "epoch": 1.51, + "grad_norm": 0.5234375, + "learning_rate": 2.8031960881717102e-05, + "loss": 2.539, + "step": 35493 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.802730368984452e-05, + "loss": 2.4174, + "step": 35494 + }, + { + "epoch": 1.51, + "grad_norm": 0.390625, + "learning_rate": 2.8022646821822808e-05, + "loss": 2.2928, + "step": 35495 + }, + { + "epoch": 1.51, + "grad_norm": 0.494140625, + "learning_rate": 2.8017990277672844e-05, + "loss": 2.5124, + "step": 35496 + }, + { + "epoch": 1.51, + "grad_norm": 0.45703125, + "learning_rate": 2.8013334057415652e-05, + "loss": 2.3342, + "step": 35497 + }, + { + "epoch": 1.51, + "grad_norm": 0.46875, + "learning_rate": 2.8008678161072112e-05, + "loss": 2.5203, + "step": 35498 + }, + { + "epoch": 1.51, + "grad_norm": 0.484375, + "learning_rate": 2.8004022588663214e-05, + "loss": 2.4278, + "step": 35499 + }, + { + "epoch": 1.51, + "grad_norm": 0.45703125, + "learning_rate": 2.7999367340209938e-05, + "loss": 2.3927, + "step": 35500 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.7994712415733194e-05, + "loss": 2.4368, + "step": 35501 + }, + { + "epoch": 1.51, + "grad_norm": 0.4921875, + "learning_rate": 2.7990057815253867e-05, + "loss": 2.4098, + "step": 35502 + }, + { + "epoch": 1.51, + "grad_norm": 0.3984375, + "learning_rate": 2.798540353879302e-05, + "loss": 2.4066, + "step": 35503 + }, + { + "epoch": 1.51, + "grad_norm": 0.455078125, + "learning_rate": 2.7980749586371546e-05, + "loss": 2.4282, + "step": 35504 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.797609595801034e-05, + "loss": 2.4011, + "step": 35505 + }, + { + "epoch": 1.51, + "grad_norm": 0.490234375, + "learning_rate": 2.7971442653730386e-05, + "loss": 2.5949, + "step": 35506 + }, + { + "epoch": 1.51, + "grad_norm": 0.5390625, + "learning_rate": 2.796678967355264e-05, + "loss": 2.5682, + "step": 35507 + }, + { + "epoch": 1.51, + "grad_norm": 0.4296875, + "learning_rate": 2.7962137017497992e-05, + "loss": 2.5172, + "step": 35508 + }, + { + "epoch": 1.51, + "grad_norm": 0.482421875, + "learning_rate": 2.795748468558743e-05, + "loss": 2.3569, + "step": 35509 + }, + { + "epoch": 1.51, + "grad_norm": 0.46875, + "learning_rate": 2.795283267784182e-05, + "loss": 2.4453, + "step": 35510 + }, + { + "epoch": 1.51, + "grad_norm": 0.447265625, + "learning_rate": 2.794818099428218e-05, + "loss": 2.4481, + "step": 35511 + }, + { + "epoch": 1.51, + "grad_norm": 0.4765625, + "learning_rate": 2.7943529634929355e-05, + "loss": 2.4463, + "step": 35512 + }, + { + "epoch": 1.51, + "grad_norm": 0.498046875, + "learning_rate": 2.7938878599804343e-05, + "loss": 2.5091, + "step": 35513 + }, + { + "epoch": 1.51, + "grad_norm": 0.427734375, + "learning_rate": 2.793422788892801e-05, + "loss": 2.474, + "step": 35514 + }, + { + "epoch": 1.51, + "grad_norm": 0.435546875, + "learning_rate": 2.7929577502321314e-05, + "loss": 2.4215, + "step": 35515 + }, + { + "epoch": 1.51, + "grad_norm": 0.4609375, + "learning_rate": 2.792492744000521e-05, + "loss": 2.5089, + "step": 35516 + }, + { + "epoch": 1.51, + "grad_norm": 0.419921875, + "learning_rate": 2.7920277702000596e-05, + "loss": 2.4302, + "step": 35517 + }, + { + "epoch": 1.51, + "grad_norm": 0.462890625, + "learning_rate": 2.7915628288328322e-05, + "loss": 2.4367, + "step": 35518 + }, + { + "epoch": 1.51, + "grad_norm": 0.451171875, + "learning_rate": 2.7910979199009446e-05, + "loss": 2.3413, + "step": 35519 + }, + { + "epoch": 1.51, + "grad_norm": 0.43359375, + "learning_rate": 2.7906330434064822e-05, + "loss": 2.2931, + "step": 35520 + }, + { + "epoch": 1.51, + "grad_norm": 0.416015625, + "learning_rate": 2.7901681993515317e-05, + "loss": 2.338, + "step": 35521 + }, + { + "epoch": 1.51, + "grad_norm": 0.4921875, + "learning_rate": 2.7897033877381905e-05, + "loss": 2.4934, + "step": 35522 + }, + { + "epoch": 1.51, + "grad_norm": 0.466796875, + "learning_rate": 2.7892386085685517e-05, + "loss": 2.5097, + "step": 35523 + }, + { + "epoch": 1.51, + "grad_norm": 0.5078125, + "learning_rate": 2.7887738618447013e-05, + "loss": 2.5042, + "step": 35524 + }, + { + "epoch": 1.51, + "grad_norm": 0.44140625, + "learning_rate": 2.7883091475687363e-05, + "loss": 2.4807, + "step": 35525 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.7878444657427415e-05, + "loss": 2.5366, + "step": 35526 + }, + { + "epoch": 1.51, + "grad_norm": 0.431640625, + "learning_rate": 2.7873798163688136e-05, + "loss": 2.3042, + "step": 35527 + }, + { + "epoch": 1.51, + "grad_norm": 0.44140625, + "learning_rate": 2.7869151994490372e-05, + "loss": 2.4966, + "step": 35528 + }, + { + "epoch": 1.51, + "grad_norm": 0.470703125, + "learning_rate": 2.7864506149855098e-05, + "loss": 2.4114, + "step": 35529 + }, + { + "epoch": 1.51, + "grad_norm": 0.50390625, + "learning_rate": 2.785986062980318e-05, + "loss": 2.4821, + "step": 35530 + }, + { + "epoch": 1.51, + "grad_norm": 0.478515625, + "learning_rate": 2.7855215434355475e-05, + "loss": 2.4098, + "step": 35531 + }, + { + "epoch": 1.51, + "grad_norm": 0.470703125, + "learning_rate": 2.7850570563532983e-05, + "loss": 2.4393, + "step": 35532 + }, + { + "epoch": 1.51, + "grad_norm": 0.439453125, + "learning_rate": 2.784592601735656e-05, + "loss": 2.4391, + "step": 35533 + }, + { + "epoch": 1.51, + "grad_norm": 0.52734375, + "learning_rate": 2.7841281795847073e-05, + "loss": 2.3815, + "step": 35534 + }, + { + "epoch": 1.51, + "grad_norm": 0.443359375, + "learning_rate": 2.783663789902544e-05, + "loss": 2.465, + "step": 35535 + }, + { + "epoch": 1.51, + "grad_norm": 0.51171875, + "learning_rate": 2.7831994326912604e-05, + "loss": 2.44, + "step": 35536 + }, + { + "epoch": 1.51, + "grad_norm": 0.490234375, + "learning_rate": 2.7827351079529384e-05, + "loss": 2.4234, + "step": 35537 + }, + { + "epoch": 1.51, + "grad_norm": 0.44921875, + "learning_rate": 2.782270815689674e-05, + "loss": 2.391, + "step": 35538 + }, + { + "epoch": 1.51, + "grad_norm": 0.484375, + "learning_rate": 2.78180655590355e-05, + "loss": 2.3063, + "step": 35539 + }, + { + "epoch": 1.51, + "grad_norm": 0.451171875, + "learning_rate": 2.7813423285966576e-05, + "loss": 2.3981, + "step": 35540 + }, + { + "epoch": 1.51, + "grad_norm": 0.51953125, + "learning_rate": 2.780878133771091e-05, + "loss": 2.5094, + "step": 35541 + }, + { + "epoch": 1.51, + "grad_norm": 0.419921875, + "learning_rate": 2.7804139714289346e-05, + "loss": 2.4679, + "step": 35542 + }, + { + "epoch": 1.51, + "grad_norm": 0.458984375, + "learning_rate": 2.7799498415722725e-05, + "loss": 2.3968, + "step": 35543 + }, + { + "epoch": 1.51, + "grad_norm": 0.421875, + "learning_rate": 2.7794857442031987e-05, + "loss": 2.4054, + "step": 35544 + }, + { + "epoch": 1.51, + "grad_norm": 0.4453125, + "learning_rate": 2.779021679323802e-05, + "loss": 2.4327, + "step": 35545 + }, + { + "epoch": 1.51, + "grad_norm": 0.52734375, + "learning_rate": 2.7785576469361673e-05, + "loss": 2.526, + "step": 35546 + }, + { + "epoch": 1.51, + "grad_norm": 0.52734375, + "learning_rate": 2.7780936470423836e-05, + "loss": 2.4209, + "step": 35547 + }, + { + "epoch": 1.51, + "grad_norm": 0.4921875, + "learning_rate": 2.777629679644541e-05, + "loss": 2.4789, + "step": 35548 + }, + { + "epoch": 1.51, + "grad_norm": 0.474609375, + "learning_rate": 2.7771657447447275e-05, + "loss": 2.4509, + "step": 35549 + }, + { + "epoch": 1.51, + "grad_norm": 0.4609375, + "learning_rate": 2.7767018423450242e-05, + "loss": 2.391, + "step": 35550 + }, + { + "epoch": 1.51, + "grad_norm": 0.482421875, + "learning_rate": 2.7762379724475228e-05, + "loss": 2.5216, + "step": 35551 + }, + { + "epoch": 1.51, + "grad_norm": 0.46484375, + "learning_rate": 2.775774135054313e-05, + "loss": 2.3216, + "step": 35552 + }, + { + "epoch": 1.52, + "grad_norm": 0.421875, + "learning_rate": 2.7753103301674777e-05, + "loss": 2.4625, + "step": 35553 + }, + { + "epoch": 1.52, + "grad_norm": 0.51953125, + "learning_rate": 2.7748465577891082e-05, + "loss": 2.4366, + "step": 35554 + }, + { + "epoch": 1.52, + "grad_norm": 0.56640625, + "learning_rate": 2.774382817921285e-05, + "loss": 2.3426, + "step": 35555 + }, + { + "epoch": 1.52, + "grad_norm": 0.494140625, + "learning_rate": 2.773919110566099e-05, + "loss": 2.3261, + "step": 35556 + }, + { + "epoch": 1.52, + "grad_norm": 0.484375, + "learning_rate": 2.7734554357256394e-05, + "loss": 2.5868, + "step": 35557 + }, + { + "epoch": 1.52, + "grad_norm": 0.482421875, + "learning_rate": 2.772991793401989e-05, + "loss": 2.4531, + "step": 35558 + }, + { + "epoch": 1.52, + "grad_norm": 0.451171875, + "learning_rate": 2.7725281835972305e-05, + "loss": 2.4631, + "step": 35559 + }, + { + "epoch": 1.52, + "grad_norm": 0.41015625, + "learning_rate": 2.772064606313455e-05, + "loss": 2.5017, + "step": 35560 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.77160106155275e-05, + "loss": 2.5449, + "step": 35561 + }, + { + "epoch": 1.52, + "grad_norm": 0.458984375, + "learning_rate": 2.7711375493171952e-05, + "loss": 2.4412, + "step": 35562 + }, + { + "epoch": 1.52, + "grad_norm": 0.447265625, + "learning_rate": 2.7706740696088806e-05, + "loss": 2.4385, + "step": 35563 + }, + { + "epoch": 1.52, + "grad_norm": 0.4609375, + "learning_rate": 2.770210622429893e-05, + "loss": 2.3128, + "step": 35564 + }, + { + "epoch": 1.52, + "grad_norm": 0.498046875, + "learning_rate": 2.769747207782315e-05, + "loss": 2.4756, + "step": 35565 + }, + { + "epoch": 1.52, + "grad_norm": 0.5, + "learning_rate": 2.7692838256682297e-05, + "loss": 2.6207, + "step": 35566 + }, + { + "epoch": 1.52, + "grad_norm": 0.3984375, + "learning_rate": 2.7688204760897253e-05, + "loss": 2.3204, + "step": 35567 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.7683571590488878e-05, + "loss": 2.5214, + "step": 35568 + }, + { + "epoch": 1.52, + "grad_norm": 0.466796875, + "learning_rate": 2.7678938745477977e-05, + "loss": 2.5088, + "step": 35569 + }, + { + "epoch": 1.52, + "grad_norm": 0.443359375, + "learning_rate": 2.767430622588546e-05, + "loss": 2.4827, + "step": 35570 + }, + { + "epoch": 1.52, + "grad_norm": 0.494140625, + "learning_rate": 2.7669674031732086e-05, + "loss": 2.4142, + "step": 35571 + }, + { + "epoch": 1.52, + "grad_norm": 0.48828125, + "learning_rate": 2.7665042163038757e-05, + "loss": 2.5005, + "step": 35572 + }, + { + "epoch": 1.52, + "grad_norm": 0.46484375, + "learning_rate": 2.7660410619826328e-05, + "loss": 2.4359, + "step": 35573 + }, + { + "epoch": 1.52, + "grad_norm": 0.482421875, + "learning_rate": 2.7655779402115623e-05, + "loss": 2.5017, + "step": 35574 + }, + { + "epoch": 1.52, + "grad_norm": 0.4296875, + "learning_rate": 2.7651148509927426e-05, + "loss": 2.4754, + "step": 35575 + }, + { + "epoch": 1.52, + "grad_norm": 0.462890625, + "learning_rate": 2.7646517943282636e-05, + "loss": 2.4467, + "step": 35576 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.7641887702202095e-05, + "loss": 2.4897, + "step": 35577 + }, + { + "epoch": 1.52, + "grad_norm": 0.46484375, + "learning_rate": 2.7637257786706583e-05, + "loss": 2.3935, + "step": 35578 + }, + { + "epoch": 1.52, + "grad_norm": 0.5234375, + "learning_rate": 2.763262819681697e-05, + "loss": 2.5033, + "step": 35579 + }, + { + "epoch": 1.52, + "grad_norm": 0.494140625, + "learning_rate": 2.762799893255411e-05, + "loss": 2.4903, + "step": 35580 + }, + { + "epoch": 1.52, + "grad_norm": 0.494140625, + "learning_rate": 2.7623369993938807e-05, + "loss": 2.4239, + "step": 35581 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.7618741380991865e-05, + "loss": 2.5406, + "step": 35582 + }, + { + "epoch": 1.52, + "grad_norm": 0.46484375, + "learning_rate": 2.761411309373413e-05, + "loss": 2.3499, + "step": 35583 + }, + { + "epoch": 1.52, + "grad_norm": 0.49609375, + "learning_rate": 2.7609485132186464e-05, + "loss": 2.4983, + "step": 35584 + }, + { + "epoch": 1.52, + "grad_norm": 0.46484375, + "learning_rate": 2.7604857496369618e-05, + "loss": 2.3958, + "step": 35585 + }, + { + "epoch": 1.52, + "grad_norm": 0.44140625, + "learning_rate": 2.76002301863045e-05, + "loss": 2.3653, + "step": 35586 + }, + { + "epoch": 1.52, + "grad_norm": 0.46875, + "learning_rate": 2.7595603202011844e-05, + "loss": 2.4783, + "step": 35587 + }, + { + "epoch": 1.52, + "grad_norm": 0.5078125, + "learning_rate": 2.7590976543512526e-05, + "loss": 2.383, + "step": 35588 + }, + { + "epoch": 1.52, + "grad_norm": 0.44140625, + "learning_rate": 2.7586350210827372e-05, + "loss": 2.4635, + "step": 35589 + }, + { + "epoch": 1.52, + "grad_norm": 0.4375, + "learning_rate": 2.7581724203977178e-05, + "loss": 2.3181, + "step": 35590 + }, + { + "epoch": 1.52, + "grad_norm": 0.408203125, + "learning_rate": 2.7577098522982735e-05, + "loss": 2.3912, + "step": 35591 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.7572473167864878e-05, + "loss": 2.454, + "step": 35592 + }, + { + "epoch": 1.52, + "grad_norm": 0.490234375, + "learning_rate": 2.756784813864446e-05, + "loss": 2.5012, + "step": 35593 + }, + { + "epoch": 1.52, + "grad_norm": 0.427734375, + "learning_rate": 2.7563223435342213e-05, + "loss": 2.5146, + "step": 35594 + }, + { + "epoch": 1.52, + "grad_norm": 0.5390625, + "learning_rate": 2.7558599057978983e-05, + "loss": 2.4088, + "step": 35595 + }, + { + "epoch": 1.52, + "grad_norm": 0.5078125, + "learning_rate": 2.7553975006575628e-05, + "loss": 2.5919, + "step": 35596 + }, + { + "epoch": 1.52, + "grad_norm": 0.45703125, + "learning_rate": 2.7549351281152902e-05, + "loss": 2.4327, + "step": 35597 + }, + { + "epoch": 1.52, + "grad_norm": 0.453125, + "learning_rate": 2.7544727881731592e-05, + "loss": 2.4685, + "step": 35598 + }, + { + "epoch": 1.52, + "grad_norm": 0.447265625, + "learning_rate": 2.7540104808332522e-05, + "loss": 2.3672, + "step": 35599 + }, + { + "epoch": 1.52, + "grad_norm": 0.453125, + "learning_rate": 2.7535482060976535e-05, + "loss": 2.3573, + "step": 35600 + }, + { + "epoch": 1.52, + "grad_norm": 0.44921875, + "learning_rate": 2.753085963968436e-05, + "loss": 2.4239, + "step": 35601 + }, + { + "epoch": 1.52, + "grad_norm": 0.48046875, + "learning_rate": 2.7526237544476863e-05, + "loss": 2.5142, + "step": 35602 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.752161577537479e-05, + "loss": 2.464, + "step": 35603 + }, + { + "epoch": 1.52, + "grad_norm": 0.453125, + "learning_rate": 2.751699433239895e-05, + "loss": 2.3858, + "step": 35604 + }, + { + "epoch": 1.52, + "grad_norm": 0.447265625, + "learning_rate": 2.7512373215570174e-05, + "loss": 2.4329, + "step": 35605 + }, + { + "epoch": 1.52, + "grad_norm": 0.439453125, + "learning_rate": 2.750775242490923e-05, + "loss": 2.4234, + "step": 35606 + }, + { + "epoch": 1.52, + "grad_norm": 0.47265625, + "learning_rate": 2.750313196043688e-05, + "loss": 2.6022, + "step": 35607 + }, + { + "epoch": 1.52, + "grad_norm": 0.484375, + "learning_rate": 2.7498511822173934e-05, + "loss": 2.5623, + "step": 35608 + }, + { + "epoch": 1.52, + "grad_norm": 0.451171875, + "learning_rate": 2.749389201014122e-05, + "loss": 2.363, + "step": 35609 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.7489272524359467e-05, + "loss": 2.447, + "step": 35610 + }, + { + "epoch": 1.52, + "grad_norm": 0.474609375, + "learning_rate": 2.7484653364849487e-05, + "loss": 2.4594, + "step": 35611 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.7480034531632105e-05, + "loss": 2.2565, + "step": 35612 + }, + { + "epoch": 1.52, + "grad_norm": 0.458984375, + "learning_rate": 2.747541602472805e-05, + "loss": 2.4503, + "step": 35613 + }, + { + "epoch": 1.52, + "grad_norm": 0.52734375, + "learning_rate": 2.7470797844158104e-05, + "loss": 2.4679, + "step": 35614 + }, + { + "epoch": 1.52, + "grad_norm": 0.55078125, + "learning_rate": 2.7466179989943042e-05, + "loss": 2.523, + "step": 35615 + }, + { + "epoch": 1.52, + "grad_norm": 0.3828125, + "learning_rate": 2.7461562462103708e-05, + "loss": 2.2982, + "step": 35616 + }, + { + "epoch": 1.52, + "grad_norm": 0.49609375, + "learning_rate": 2.7456945260660794e-05, + "loss": 2.3519, + "step": 35617 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.7452328385635152e-05, + "loss": 2.6215, + "step": 35618 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.7447711837047484e-05, + "loss": 2.543, + "step": 35619 + }, + { + "epoch": 1.52, + "grad_norm": 0.3984375, + "learning_rate": 2.7443095614918602e-05, + "loss": 2.3924, + "step": 35620 + }, + { + "epoch": 1.52, + "grad_norm": 0.54296875, + "learning_rate": 2.7438479719269294e-05, + "loss": 2.4947, + "step": 35621 + }, + { + "epoch": 1.52, + "grad_norm": 0.58203125, + "learning_rate": 2.7433864150120313e-05, + "loss": 2.5048, + "step": 35622 + }, + { + "epoch": 1.52, + "grad_norm": 0.46484375, + "learning_rate": 2.74292489074924e-05, + "loss": 2.4863, + "step": 35623 + }, + { + "epoch": 1.52, + "grad_norm": 0.482421875, + "learning_rate": 2.7424633991406344e-05, + "loss": 2.5523, + "step": 35624 + }, + { + "epoch": 1.52, + "grad_norm": 0.4609375, + "learning_rate": 2.7420019401882934e-05, + "loss": 2.4636, + "step": 35625 + }, + { + "epoch": 1.52, + "grad_norm": 0.486328125, + "learning_rate": 2.7415405138942886e-05, + "loss": 2.4329, + "step": 35626 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.741079120260699e-05, + "loss": 2.471, + "step": 35627 + }, + { + "epoch": 1.52, + "grad_norm": 0.451171875, + "learning_rate": 2.740617759289603e-05, + "loss": 2.5445, + "step": 35628 + }, + { + "epoch": 1.52, + "grad_norm": 0.388671875, + "learning_rate": 2.740156430983075e-05, + "loss": 2.3585, + "step": 35629 + }, + { + "epoch": 1.52, + "grad_norm": 0.51171875, + "learning_rate": 2.7396951353431864e-05, + "loss": 2.4421, + "step": 35630 + }, + { + "epoch": 1.52, + "grad_norm": 0.47265625, + "learning_rate": 2.739233872372017e-05, + "loss": 2.4262, + "step": 35631 + }, + { + "epoch": 1.52, + "grad_norm": 0.470703125, + "learning_rate": 2.738772642071643e-05, + "loss": 2.4932, + "step": 35632 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.7383114444441373e-05, + "loss": 2.4313, + "step": 35633 + }, + { + "epoch": 1.52, + "grad_norm": 0.486328125, + "learning_rate": 2.7378502794915784e-05, + "loss": 2.6773, + "step": 35634 + }, + { + "epoch": 1.52, + "grad_norm": 0.447265625, + "learning_rate": 2.737389147216035e-05, + "loss": 2.3474, + "step": 35635 + }, + { + "epoch": 1.52, + "grad_norm": 0.447265625, + "learning_rate": 2.736928047619588e-05, + "loss": 2.3531, + "step": 35636 + }, + { + "epoch": 1.52, + "grad_norm": 0.478515625, + "learning_rate": 2.736466980704313e-05, + "loss": 2.4478, + "step": 35637 + }, + { + "epoch": 1.52, + "grad_norm": 0.453125, + "learning_rate": 2.736005946472282e-05, + "loss": 2.4312, + "step": 35638 + }, + { + "epoch": 1.52, + "grad_norm": 0.48046875, + "learning_rate": 2.7355449449255667e-05, + "loss": 2.5175, + "step": 35639 + }, + { + "epoch": 1.52, + "grad_norm": 0.484375, + "learning_rate": 2.735083976066244e-05, + "loss": 2.5547, + "step": 35640 + }, + { + "epoch": 1.52, + "grad_norm": 0.43359375, + "learning_rate": 2.7346230398963914e-05, + "loss": 2.4832, + "step": 35641 + }, + { + "epoch": 1.52, + "grad_norm": 0.435546875, + "learning_rate": 2.734162136418077e-05, + "loss": 2.3915, + "step": 35642 + }, + { + "epoch": 1.52, + "grad_norm": 0.435546875, + "learning_rate": 2.733701265633378e-05, + "loss": 2.4212, + "step": 35643 + }, + { + "epoch": 1.52, + "grad_norm": 0.494140625, + "learning_rate": 2.73324042754437e-05, + "loss": 2.5023, + "step": 35644 + }, + { + "epoch": 1.52, + "grad_norm": 0.4140625, + "learning_rate": 2.7327796221531254e-05, + "loss": 2.3814, + "step": 35645 + }, + { + "epoch": 1.52, + "grad_norm": 0.44921875, + "learning_rate": 2.7323188494617124e-05, + "loss": 2.3994, + "step": 35646 + }, + { + "epoch": 1.52, + "grad_norm": 0.453125, + "learning_rate": 2.7318581094722096e-05, + "loss": 2.6052, + "step": 35647 + }, + { + "epoch": 1.52, + "grad_norm": 0.44140625, + "learning_rate": 2.7313974021866916e-05, + "loss": 2.4912, + "step": 35648 + }, + { + "epoch": 1.52, + "grad_norm": 0.416015625, + "learning_rate": 2.7309367276072262e-05, + "loss": 2.4489, + "step": 35649 + }, + { + "epoch": 1.52, + "grad_norm": 0.443359375, + "learning_rate": 2.7304760857358913e-05, + "loss": 2.3249, + "step": 35650 + }, + { + "epoch": 1.52, + "grad_norm": 0.5234375, + "learning_rate": 2.7300154765747587e-05, + "loss": 2.5167, + "step": 35651 + }, + { + "epoch": 1.52, + "grad_norm": 0.46484375, + "learning_rate": 2.7295549001258925e-05, + "loss": 2.4737, + "step": 35652 + }, + { + "epoch": 1.52, + "grad_norm": 0.5625, + "learning_rate": 2.7290943563913785e-05, + "loss": 2.4544, + "step": 35653 + }, + { + "epoch": 1.52, + "grad_norm": 0.48046875, + "learning_rate": 2.728633845373282e-05, + "loss": 2.4205, + "step": 35654 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.728173367073674e-05, + "loss": 2.3525, + "step": 35655 + }, + { + "epoch": 1.52, + "grad_norm": 0.453125, + "learning_rate": 2.7277129214946272e-05, + "loss": 2.3669, + "step": 35656 + }, + { + "epoch": 1.52, + "grad_norm": 0.462890625, + "learning_rate": 2.7272525086382184e-05, + "loss": 2.5728, + "step": 35657 + }, + { + "epoch": 1.52, + "grad_norm": 0.47265625, + "learning_rate": 2.7267921285065124e-05, + "loss": 2.4092, + "step": 35658 + }, + { + "epoch": 1.52, + "grad_norm": 0.451171875, + "learning_rate": 2.7263317811015866e-05, + "loss": 2.3143, + "step": 35659 + }, + { + "epoch": 1.52, + "grad_norm": 0.447265625, + "learning_rate": 2.7258714664255057e-05, + "loss": 2.3436, + "step": 35660 + }, + { + "epoch": 1.52, + "grad_norm": 0.4375, + "learning_rate": 2.7254111844803486e-05, + "loss": 2.4475, + "step": 35661 + }, + { + "epoch": 1.52, + "grad_norm": 0.490234375, + "learning_rate": 2.72495093526818e-05, + "loss": 2.3252, + "step": 35662 + }, + { + "epoch": 1.52, + "grad_norm": 0.46875, + "learning_rate": 2.7244907187910752e-05, + "loss": 2.6222, + "step": 35663 + }, + { + "epoch": 1.52, + "grad_norm": 0.412109375, + "learning_rate": 2.724030535051102e-05, + "loss": 2.4443, + "step": 35664 + }, + { + "epoch": 1.52, + "grad_norm": 0.462890625, + "learning_rate": 2.723570384050331e-05, + "loss": 2.5383, + "step": 35665 + }, + { + "epoch": 1.52, + "grad_norm": 0.546875, + "learning_rate": 2.723110265790837e-05, + "loss": 2.3994, + "step": 35666 + }, + { + "epoch": 1.52, + "grad_norm": 0.51171875, + "learning_rate": 2.7226501802746884e-05, + "loss": 2.4951, + "step": 35667 + }, + { + "epoch": 1.52, + "grad_norm": 0.4609375, + "learning_rate": 2.7221901275039476e-05, + "loss": 2.499, + "step": 35668 + }, + { + "epoch": 1.52, + "grad_norm": 0.486328125, + "learning_rate": 2.721730107480698e-05, + "loss": 2.5123, + "step": 35669 + }, + { + "epoch": 1.52, + "grad_norm": 0.49609375, + "learning_rate": 2.7212701202070024e-05, + "loss": 2.3602, + "step": 35670 + }, + { + "epoch": 1.52, + "grad_norm": 0.443359375, + "learning_rate": 2.7208101656849284e-05, + "loss": 2.4698, + "step": 35671 + }, + { + "epoch": 1.52, + "grad_norm": 0.462890625, + "learning_rate": 2.7203502439165485e-05, + "loss": 2.4826, + "step": 35672 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.7198903549039356e-05, + "loss": 2.562, + "step": 35673 + }, + { + "epoch": 1.52, + "grad_norm": 0.435546875, + "learning_rate": 2.7194304986491514e-05, + "loss": 2.4686, + "step": 35674 + }, + { + "epoch": 1.52, + "grad_norm": 0.443359375, + "learning_rate": 2.7189706751542733e-05, + "loss": 2.3863, + "step": 35675 + }, + { + "epoch": 1.52, + "grad_norm": 0.482421875, + "learning_rate": 2.7185108844213636e-05, + "loss": 2.4575, + "step": 35676 + }, + { + "epoch": 1.52, + "grad_norm": 0.44921875, + "learning_rate": 2.7180511264524956e-05, + "loss": 2.3349, + "step": 35677 + }, + { + "epoch": 1.52, + "grad_norm": 0.458984375, + "learning_rate": 2.7175914012497338e-05, + "loss": 2.2645, + "step": 35678 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.7171317088151527e-05, + "loss": 2.3998, + "step": 35679 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.7166720491508134e-05, + "loss": 2.3878, + "step": 35680 + }, + { + "epoch": 1.52, + "grad_norm": 0.4609375, + "learning_rate": 2.7162124222587883e-05, + "loss": 2.377, + "step": 35681 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.7157528281411493e-05, + "loss": 2.3878, + "step": 35682 + }, + { + "epoch": 1.52, + "grad_norm": 0.494140625, + "learning_rate": 2.7152932667999598e-05, + "loss": 2.4009, + "step": 35683 + }, + { + "epoch": 1.52, + "grad_norm": 0.462890625, + "learning_rate": 2.7148337382372824e-05, + "loss": 2.4192, + "step": 35684 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.7143742424551965e-05, + "loss": 2.45, + "step": 35685 + }, + { + "epoch": 1.52, + "grad_norm": 0.5234375, + "learning_rate": 2.713914779455764e-05, + "loss": 2.5223, + "step": 35686 + }, + { + "epoch": 1.52, + "grad_norm": 0.48046875, + "learning_rate": 2.7134553492410496e-05, + "loss": 2.5913, + "step": 35687 + }, + { + "epoch": 1.52, + "grad_norm": 0.462890625, + "learning_rate": 2.712995951813123e-05, + "loss": 2.3175, + "step": 35688 + }, + { + "epoch": 1.52, + "grad_norm": 0.51171875, + "learning_rate": 2.712536587174055e-05, + "loss": 2.5522, + "step": 35689 + }, + { + "epoch": 1.52, + "grad_norm": 0.45703125, + "learning_rate": 2.7120772553259054e-05, + "loss": 2.4349, + "step": 35690 + }, + { + "epoch": 1.52, + "grad_norm": 0.46875, + "learning_rate": 2.711617956270748e-05, + "loss": 2.4238, + "step": 35691 + }, + { + "epoch": 1.52, + "grad_norm": 0.4609375, + "learning_rate": 2.7111586900106434e-05, + "loss": 2.4463, + "step": 35692 + }, + { + "epoch": 1.52, + "grad_norm": 0.482421875, + "learning_rate": 2.7106994565476627e-05, + "loss": 2.4631, + "step": 35693 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.710240255883869e-05, + "loss": 2.5446, + "step": 35694 + }, + { + "epoch": 1.52, + "grad_norm": 0.48828125, + "learning_rate": 2.7097810880213326e-05, + "loss": 2.5578, + "step": 35695 + }, + { + "epoch": 1.52, + "grad_norm": 0.494140625, + "learning_rate": 2.709321952962114e-05, + "loss": 2.4641, + "step": 35696 + }, + { + "epoch": 1.52, + "grad_norm": 0.458984375, + "learning_rate": 2.7088628507082827e-05, + "loss": 2.5489, + "step": 35697 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.7084037812619057e-05, + "loss": 2.3141, + "step": 35698 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.7079447446250484e-05, + "loss": 2.4642, + "step": 35699 + }, + { + "epoch": 1.52, + "grad_norm": 0.453125, + "learning_rate": 2.7074857407997678e-05, + "loss": 2.4036, + "step": 35700 + }, + { + "epoch": 1.52, + "grad_norm": 0.482421875, + "learning_rate": 2.7070267697881423e-05, + "loss": 2.4982, + "step": 35701 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.7065678315922317e-05, + "loss": 2.4554, + "step": 35702 + }, + { + "epoch": 1.52, + "grad_norm": 0.458984375, + "learning_rate": 2.706108926214096e-05, + "loss": 2.3826, + "step": 35703 + }, + { + "epoch": 1.52, + "grad_norm": 0.51171875, + "learning_rate": 2.705650053655806e-05, + "loss": 2.3648, + "step": 35704 + }, + { + "epoch": 1.52, + "grad_norm": 0.48046875, + "learning_rate": 2.705191213919427e-05, + "loss": 2.3793, + "step": 35705 + }, + { + "epoch": 1.52, + "grad_norm": 0.4453125, + "learning_rate": 2.704732407007019e-05, + "loss": 2.5205, + "step": 35706 + }, + { + "epoch": 1.52, + "grad_norm": 0.51953125, + "learning_rate": 2.704273632920652e-05, + "loss": 2.487, + "step": 35707 + }, + { + "epoch": 1.52, + "grad_norm": 0.5625, + "learning_rate": 2.7038148916623852e-05, + "loss": 2.4916, + "step": 35708 + }, + { + "epoch": 1.52, + "grad_norm": 0.416015625, + "learning_rate": 2.703356183234287e-05, + "loss": 2.4842, + "step": 35709 + }, + { + "epoch": 1.52, + "grad_norm": 0.5078125, + "learning_rate": 2.7028975076384167e-05, + "loss": 2.5226, + "step": 35710 + }, + { + "epoch": 1.52, + "grad_norm": 0.474609375, + "learning_rate": 2.702438864876844e-05, + "loss": 2.5791, + "step": 35711 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.701980254951626e-05, + "loss": 2.5638, + "step": 35712 + }, + { + "epoch": 1.52, + "grad_norm": 0.5, + "learning_rate": 2.70152167786483e-05, + "loss": 2.3623, + "step": 35713 + }, + { + "epoch": 1.52, + "grad_norm": 0.478515625, + "learning_rate": 2.7010631336185222e-05, + "loss": 2.4424, + "step": 35714 + }, + { + "epoch": 1.52, + "grad_norm": 0.423828125, + "learning_rate": 2.700604622214762e-05, + "loss": 2.3556, + "step": 35715 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.7001461436556087e-05, + "loss": 2.5409, + "step": 35716 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.6996876979431353e-05, + "loss": 2.5804, + "step": 35717 + }, + { + "epoch": 1.52, + "grad_norm": 0.546875, + "learning_rate": 2.6992292850793988e-05, + "loss": 2.5624, + "step": 35718 + }, + { + "epoch": 1.52, + "grad_norm": 0.46875, + "learning_rate": 2.69877090506646e-05, + "loss": 2.4966, + "step": 35719 + }, + { + "epoch": 1.52, + "grad_norm": 0.400390625, + "learning_rate": 2.698312557906384e-05, + "loss": 2.4462, + "step": 35720 + }, + { + "epoch": 1.52, + "grad_norm": 0.48046875, + "learning_rate": 2.6978542436012355e-05, + "loss": 2.4627, + "step": 35721 + }, + { + "epoch": 1.52, + "grad_norm": 0.40234375, + "learning_rate": 2.697395962153072e-05, + "loss": 2.4313, + "step": 35722 + }, + { + "epoch": 1.52, + "grad_norm": 0.486328125, + "learning_rate": 2.6969377135639594e-05, + "loss": 2.4169, + "step": 35723 + }, + { + "epoch": 1.52, + "grad_norm": 0.427734375, + "learning_rate": 2.6964794978359565e-05, + "loss": 2.491, + "step": 35724 + }, + { + "epoch": 1.52, + "grad_norm": 0.435546875, + "learning_rate": 2.69602131497113e-05, + "loss": 2.4579, + "step": 35725 + }, + { + "epoch": 1.52, + "grad_norm": 0.54296875, + "learning_rate": 2.695563164971534e-05, + "loss": 2.4133, + "step": 35726 + }, + { + "epoch": 1.52, + "grad_norm": 0.546875, + "learning_rate": 2.695105047839238e-05, + "loss": 2.4418, + "step": 35727 + }, + { + "epoch": 1.52, + "grad_norm": 0.4921875, + "learning_rate": 2.6946469635762973e-05, + "loss": 2.4669, + "step": 35728 + }, + { + "epoch": 1.52, + "grad_norm": 0.431640625, + "learning_rate": 2.694188912184774e-05, + "loss": 2.4459, + "step": 35729 + }, + { + "epoch": 1.52, + "grad_norm": 0.423828125, + "learning_rate": 2.6937308936667348e-05, + "loss": 2.4701, + "step": 35730 + }, + { + "epoch": 1.52, + "grad_norm": 0.419921875, + "learning_rate": 2.6932729080242358e-05, + "loss": 2.3126, + "step": 35731 + }, + { + "epoch": 1.52, + "grad_norm": 0.478515625, + "learning_rate": 2.692814955259333e-05, + "loss": 2.3797, + "step": 35732 + }, + { + "epoch": 1.52, + "grad_norm": 0.42578125, + "learning_rate": 2.6923570353740978e-05, + "loss": 2.4529, + "step": 35733 + }, + { + "epoch": 1.52, + "grad_norm": 0.5078125, + "learning_rate": 2.691899148370586e-05, + "loss": 2.498, + "step": 35734 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.6914412942508536e-05, + "loss": 2.4504, + "step": 35735 + }, + { + "epoch": 1.52, + "grad_norm": 0.48046875, + "learning_rate": 2.690983473016965e-05, + "loss": 2.4816, + "step": 35736 + }, + { + "epoch": 1.52, + "grad_norm": 0.5078125, + "learning_rate": 2.6905256846709815e-05, + "loss": 2.4402, + "step": 35737 + }, + { + "epoch": 1.52, + "grad_norm": 0.5078125, + "learning_rate": 2.690067929214959e-05, + "loss": 2.4579, + "step": 35738 + }, + { + "epoch": 1.52, + "grad_norm": 0.453125, + "learning_rate": 2.689610206650962e-05, + "loss": 2.453, + "step": 35739 + }, + { + "epoch": 1.52, + "grad_norm": 0.451171875, + "learning_rate": 2.689152516981045e-05, + "loss": 2.5014, + "step": 35740 + }, + { + "epoch": 1.52, + "grad_norm": 0.53125, + "learning_rate": 2.6886948602072726e-05, + "loss": 2.5373, + "step": 35741 + }, + { + "epoch": 1.52, + "grad_norm": 0.4609375, + "learning_rate": 2.6882372363316977e-05, + "loss": 2.3719, + "step": 35742 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.6877796453563874e-05, + "loss": 2.3298, + "step": 35743 + }, + { + "epoch": 1.52, + "grad_norm": 0.45703125, + "learning_rate": 2.6873220872833927e-05, + "loss": 2.5702, + "step": 35744 + }, + { + "epoch": 1.52, + "grad_norm": 0.52734375, + "learning_rate": 2.6868645621147758e-05, + "loss": 2.3331, + "step": 35745 + }, + { + "epoch": 1.52, + "grad_norm": 0.498046875, + "learning_rate": 2.6864070698525988e-05, + "loss": 2.5214, + "step": 35746 + }, + { + "epoch": 1.52, + "grad_norm": 0.5078125, + "learning_rate": 2.685949610498917e-05, + "loss": 2.4455, + "step": 35747 + }, + { + "epoch": 1.52, + "grad_norm": 0.423828125, + "learning_rate": 2.6854921840557834e-05, + "loss": 2.427, + "step": 35748 + }, + { + "epoch": 1.52, + "grad_norm": 0.44921875, + "learning_rate": 2.685034790525268e-05, + "loss": 2.2781, + "step": 35749 + }, + { + "epoch": 1.52, + "grad_norm": 0.451171875, + "learning_rate": 2.6845774299094227e-05, + "loss": 2.3679, + "step": 35750 + }, + { + "epoch": 1.52, + "grad_norm": 0.47265625, + "learning_rate": 2.6841201022103026e-05, + "loss": 2.573, + "step": 35751 + }, + { + "epoch": 1.52, + "grad_norm": 0.498046875, + "learning_rate": 2.6836628074299673e-05, + "loss": 2.4684, + "step": 35752 + }, + { + "epoch": 1.52, + "grad_norm": 0.462890625, + "learning_rate": 2.6832055455704797e-05, + "loss": 2.4439, + "step": 35753 + }, + { + "epoch": 1.52, + "grad_norm": 0.494140625, + "learning_rate": 2.6827483166338897e-05, + "loss": 2.6196, + "step": 35754 + }, + { + "epoch": 1.52, + "grad_norm": 0.451171875, + "learning_rate": 2.6822911206222613e-05, + "loss": 2.5062, + "step": 35755 + }, + { + "epoch": 1.52, + "grad_norm": 0.474609375, + "learning_rate": 2.681833957537645e-05, + "loss": 2.3148, + "step": 35756 + }, + { + "epoch": 1.52, + "grad_norm": 0.478515625, + "learning_rate": 2.681376827382105e-05, + "loss": 2.4925, + "step": 35757 + }, + { + "epoch": 1.52, + "grad_norm": 0.490234375, + "learning_rate": 2.6809197301576915e-05, + "loss": 2.3601, + "step": 35758 + }, + { + "epoch": 1.52, + "grad_norm": 0.466796875, + "learning_rate": 2.6804626658664677e-05, + "loss": 2.5367, + "step": 35759 + }, + { + "epoch": 1.52, + "grad_norm": 0.48828125, + "learning_rate": 2.680005634510483e-05, + "loss": 2.2949, + "step": 35760 + }, + { + "epoch": 1.52, + "grad_norm": 0.431640625, + "learning_rate": 2.6795486360917987e-05, + "loss": 2.4864, + "step": 35761 + }, + { + "epoch": 1.52, + "grad_norm": 0.419921875, + "learning_rate": 2.6790916706124725e-05, + "loss": 2.3761, + "step": 35762 + }, + { + "epoch": 1.52, + "grad_norm": 0.39453125, + "learning_rate": 2.6786347380745592e-05, + "loss": 2.4063, + "step": 35763 + }, + { + "epoch": 1.52, + "grad_norm": 0.4609375, + "learning_rate": 2.678177838480107e-05, + "loss": 2.4482, + "step": 35764 + }, + { + "epoch": 1.52, + "grad_norm": 0.51171875, + "learning_rate": 2.6777209718311857e-05, + "loss": 2.5011, + "step": 35765 + }, + { + "epoch": 1.52, + "grad_norm": 0.486328125, + "learning_rate": 2.6772641381298434e-05, + "loss": 2.3822, + "step": 35766 + }, + { + "epoch": 1.52, + "grad_norm": 0.423828125, + "learning_rate": 2.6768073373781333e-05, + "loss": 2.4471, + "step": 35767 + }, + { + "epoch": 1.52, + "grad_norm": 0.44140625, + "learning_rate": 2.6763505695781133e-05, + "loss": 2.375, + "step": 35768 + }, + { + "epoch": 1.52, + "grad_norm": 0.412109375, + "learning_rate": 2.675893834731843e-05, + "loss": 2.3935, + "step": 35769 + }, + { + "epoch": 1.52, + "grad_norm": 0.46484375, + "learning_rate": 2.6754371328413708e-05, + "loss": 2.4326, + "step": 35770 + }, + { + "epoch": 1.52, + "grad_norm": 0.443359375, + "learning_rate": 2.6749804639087572e-05, + "loss": 2.4038, + "step": 35771 + }, + { + "epoch": 1.52, + "grad_norm": 0.47265625, + "learning_rate": 2.6745238279360553e-05, + "loss": 2.541, + "step": 35772 + }, + { + "epoch": 1.52, + "grad_norm": 0.50390625, + "learning_rate": 2.6740672249253163e-05, + "loss": 2.4327, + "step": 35773 + }, + { + "epoch": 1.52, + "grad_norm": 0.48046875, + "learning_rate": 2.6736106548785968e-05, + "loss": 2.449, + "step": 35774 + }, + { + "epoch": 1.52, + "grad_norm": 0.5, + "learning_rate": 2.6731541177979547e-05, + "loss": 2.4568, + "step": 35775 + }, + { + "epoch": 1.52, + "grad_norm": 0.470703125, + "learning_rate": 2.6726976136854376e-05, + "loss": 2.4846, + "step": 35776 + }, + { + "epoch": 1.52, + "grad_norm": 0.447265625, + "learning_rate": 2.672241142543105e-05, + "loss": 2.4906, + "step": 35777 + }, + { + "epoch": 1.52, + "grad_norm": 0.478515625, + "learning_rate": 2.6717847043730114e-05, + "loss": 2.3193, + "step": 35778 + }, + { + "epoch": 1.52, + "grad_norm": 0.427734375, + "learning_rate": 2.6713282991772092e-05, + "loss": 2.4433, + "step": 35779 + }, + { + "epoch": 1.52, + "grad_norm": 0.5078125, + "learning_rate": 2.6708719269577474e-05, + "loss": 2.6067, + "step": 35780 + }, + { + "epoch": 1.52, + "grad_norm": 0.3828125, + "learning_rate": 2.6704155877166837e-05, + "loss": 2.5549, + "step": 35781 + }, + { + "epoch": 1.52, + "grad_norm": 0.427734375, + "learning_rate": 2.669959281456075e-05, + "loss": 2.4145, + "step": 35782 + }, + { + "epoch": 1.52, + "grad_norm": 0.404296875, + "learning_rate": 2.6695030081779672e-05, + "loss": 2.3977, + "step": 35783 + }, + { + "epoch": 1.52, + "grad_norm": 0.462890625, + "learning_rate": 2.669046767884419e-05, + "loss": 2.5005, + "step": 35784 + }, + { + "epoch": 1.52, + "grad_norm": 0.44140625, + "learning_rate": 2.6685905605774798e-05, + "loss": 2.4597, + "step": 35785 + }, + { + "epoch": 1.52, + "grad_norm": 0.455078125, + "learning_rate": 2.6681343862592023e-05, + "loss": 2.5507, + "step": 35786 + }, + { + "epoch": 1.52, + "grad_norm": 0.44921875, + "learning_rate": 2.667678244931644e-05, + "loss": 2.61, + "step": 35787 + }, + { + "epoch": 1.53, + "grad_norm": 0.412109375, + "learning_rate": 2.6672221365968543e-05, + "loss": 2.4117, + "step": 35788 + }, + { + "epoch": 1.53, + "grad_norm": 0.466796875, + "learning_rate": 2.6667660612568823e-05, + "loss": 2.306, + "step": 35789 + }, + { + "epoch": 1.53, + "grad_norm": 0.451171875, + "learning_rate": 2.6663100189137824e-05, + "loss": 2.5431, + "step": 35790 + }, + { + "epoch": 1.53, + "grad_norm": 0.494140625, + "learning_rate": 2.6658540095696104e-05, + "loss": 2.4867, + "step": 35791 + }, + { + "epoch": 1.53, + "grad_norm": 0.490234375, + "learning_rate": 2.6653980332264128e-05, + "loss": 2.4225, + "step": 35792 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.664942089886242e-05, + "loss": 2.3294, + "step": 35793 + }, + { + "epoch": 1.53, + "grad_norm": 0.470703125, + "learning_rate": 2.6644861795511555e-05, + "loss": 2.4803, + "step": 35794 + }, + { + "epoch": 1.53, + "grad_norm": 0.5234375, + "learning_rate": 2.6640303022232004e-05, + "loss": 2.5422, + "step": 35795 + }, + { + "epoch": 1.53, + "grad_norm": 0.384765625, + "learning_rate": 2.6635744579044242e-05, + "loss": 2.576, + "step": 35796 + }, + { + "epoch": 1.53, + "grad_norm": 0.4921875, + "learning_rate": 2.6631186465968828e-05, + "loss": 2.4837, + "step": 35797 + }, + { + "epoch": 1.53, + "grad_norm": 0.515625, + "learning_rate": 2.6626628683026278e-05, + "loss": 2.4404, + "step": 35798 + }, + { + "epoch": 1.53, + "grad_norm": 0.490234375, + "learning_rate": 2.6622071230237068e-05, + "loss": 2.4531, + "step": 35799 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.661751410762173e-05, + "loss": 2.3046, + "step": 35800 + }, + { + "epoch": 1.53, + "grad_norm": 0.53515625, + "learning_rate": 2.661295731520075e-05, + "loss": 2.38, + "step": 35801 + }, + { + "epoch": 1.53, + "grad_norm": 0.482421875, + "learning_rate": 2.660840085299464e-05, + "loss": 2.4908, + "step": 35802 + }, + { + "epoch": 1.53, + "grad_norm": 0.40234375, + "learning_rate": 2.660384472102393e-05, + "loss": 2.3713, + "step": 35803 + }, + { + "epoch": 1.53, + "grad_norm": 0.4140625, + "learning_rate": 2.6599288919309095e-05, + "loss": 2.2506, + "step": 35804 + }, + { + "epoch": 1.53, + "grad_norm": 0.482421875, + "learning_rate": 2.65947334478706e-05, + "loss": 2.5142, + "step": 35805 + }, + { + "epoch": 1.53, + "grad_norm": 0.4609375, + "learning_rate": 2.6590178306728985e-05, + "loss": 2.6064, + "step": 35806 + }, + { + "epoch": 1.53, + "grad_norm": 0.44140625, + "learning_rate": 2.6585623495904778e-05, + "loss": 2.509, + "step": 35807 + }, + { + "epoch": 1.53, + "grad_norm": 0.474609375, + "learning_rate": 2.6581069015418392e-05, + "loss": 2.5195, + "step": 35808 + }, + { + "epoch": 1.53, + "grad_norm": 0.458984375, + "learning_rate": 2.657651486529037e-05, + "loss": 2.4614, + "step": 35809 + }, + { + "epoch": 1.53, + "grad_norm": 0.51953125, + "learning_rate": 2.6571961045541227e-05, + "loss": 2.4848, + "step": 35810 + }, + { + "epoch": 1.53, + "grad_norm": 0.384765625, + "learning_rate": 2.6567407556191413e-05, + "loss": 2.5376, + "step": 35811 + }, + { + "epoch": 1.53, + "grad_norm": 0.4375, + "learning_rate": 2.6562854397261417e-05, + "loss": 2.3603, + "step": 35812 + }, + { + "epoch": 1.53, + "grad_norm": 0.474609375, + "learning_rate": 2.6558301568771727e-05, + "loss": 2.4401, + "step": 35813 + }, + { + "epoch": 1.53, + "grad_norm": 0.3828125, + "learning_rate": 2.6553749070742874e-05, + "loss": 2.3967, + "step": 35814 + }, + { + "epoch": 1.53, + "grad_norm": 0.447265625, + "learning_rate": 2.6549196903195273e-05, + "loss": 2.4301, + "step": 35815 + }, + { + "epoch": 1.53, + "grad_norm": 0.5, + "learning_rate": 2.6544645066149475e-05, + "loss": 2.4915, + "step": 35816 + }, + { + "epoch": 1.53, + "grad_norm": 0.52734375, + "learning_rate": 2.65400935596259e-05, + "loss": 2.4401, + "step": 35817 + }, + { + "epoch": 1.53, + "grad_norm": 0.498046875, + "learning_rate": 2.6535542383645063e-05, + "loss": 2.3249, + "step": 35818 + }, + { + "epoch": 1.53, + "grad_norm": 0.419921875, + "learning_rate": 2.653099153822747e-05, + "loss": 2.2414, + "step": 35819 + }, + { + "epoch": 1.53, + "grad_norm": 0.431640625, + "learning_rate": 2.6526441023393555e-05, + "loss": 2.2829, + "step": 35820 + }, + { + "epoch": 1.53, + "grad_norm": 0.451171875, + "learning_rate": 2.6521890839163778e-05, + "loss": 2.4916, + "step": 35821 + }, + { + "epoch": 1.53, + "grad_norm": 0.482421875, + "learning_rate": 2.6517340985558637e-05, + "loss": 2.3891, + "step": 35822 + }, + { + "epoch": 1.53, + "grad_norm": 0.435546875, + "learning_rate": 2.651279146259864e-05, + "loss": 2.4306, + "step": 35823 + }, + { + "epoch": 1.53, + "grad_norm": 0.4609375, + "learning_rate": 2.6508242270304185e-05, + "loss": 2.5076, + "step": 35824 + }, + { + "epoch": 1.53, + "grad_norm": 0.51953125, + "learning_rate": 2.6503693408695795e-05, + "loss": 2.3985, + "step": 35825 + }, + { + "epoch": 1.53, + "grad_norm": 0.4609375, + "learning_rate": 2.6499144877793947e-05, + "loss": 2.5937, + "step": 35826 + }, + { + "epoch": 1.53, + "grad_norm": 0.451171875, + "learning_rate": 2.6494596677619088e-05, + "loss": 2.3224, + "step": 35827 + }, + { + "epoch": 1.53, + "grad_norm": 0.4765625, + "learning_rate": 2.649004880819165e-05, + "loss": 2.4739, + "step": 35828 + }, + { + "epoch": 1.53, + "grad_norm": 0.47265625, + "learning_rate": 2.648550126953213e-05, + "loss": 2.461, + "step": 35829 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.648095406166101e-05, + "loss": 2.3473, + "step": 35830 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.6476407184598707e-05, + "loss": 2.6517, + "step": 35831 + }, + { + "epoch": 1.53, + "grad_norm": 0.5234375, + "learning_rate": 2.647186063836573e-05, + "loss": 2.4473, + "step": 35832 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.6467314422982482e-05, + "loss": 2.513, + "step": 35833 + }, + { + "epoch": 1.53, + "grad_norm": 0.4609375, + "learning_rate": 2.646276853846944e-05, + "loss": 2.4297, + "step": 35834 + }, + { + "epoch": 1.53, + "grad_norm": 0.5078125, + "learning_rate": 2.6458222984847092e-05, + "loss": 2.4126, + "step": 35835 + }, + { + "epoch": 1.53, + "grad_norm": 0.486328125, + "learning_rate": 2.6453677762135876e-05, + "loss": 2.4749, + "step": 35836 + }, + { + "epoch": 1.53, + "grad_norm": 0.44921875, + "learning_rate": 2.6449132870356187e-05, + "loss": 2.3721, + "step": 35837 + }, + { + "epoch": 1.53, + "grad_norm": 0.462890625, + "learning_rate": 2.6444588309528527e-05, + "loss": 2.3986, + "step": 35838 + }, + { + "epoch": 1.53, + "grad_norm": 0.451171875, + "learning_rate": 2.6440044079673377e-05, + "loss": 2.4485, + "step": 35839 + }, + { + "epoch": 1.53, + "grad_norm": 0.498046875, + "learning_rate": 2.64355001808111e-05, + "loss": 2.5687, + "step": 35840 + }, + { + "epoch": 1.53, + "grad_norm": 0.458984375, + "learning_rate": 2.64309566129622e-05, + "loss": 2.5827, + "step": 35841 + }, + { + "epoch": 1.53, + "grad_norm": 0.53515625, + "learning_rate": 2.642641337614713e-05, + "loss": 2.449, + "step": 35842 + }, + { + "epoch": 1.53, + "grad_norm": 0.4453125, + "learning_rate": 2.6421870470386323e-05, + "loss": 2.3639, + "step": 35843 + }, + { + "epoch": 1.53, + "grad_norm": 0.47265625, + "learning_rate": 2.6417327895700173e-05, + "loss": 2.5552, + "step": 35844 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.641278565210915e-05, + "loss": 2.4977, + "step": 35845 + }, + { + "epoch": 1.53, + "grad_norm": 0.431640625, + "learning_rate": 2.640824373963373e-05, + "loss": 2.3399, + "step": 35846 + }, + { + "epoch": 1.53, + "grad_norm": 0.453125, + "learning_rate": 2.6403702158294298e-05, + "loss": 2.3304, + "step": 35847 + }, + { + "epoch": 1.53, + "grad_norm": 0.484375, + "learning_rate": 2.639916090811133e-05, + "loss": 2.5701, + "step": 35848 + }, + { + "epoch": 1.53, + "grad_norm": 0.46875, + "learning_rate": 2.639461998910522e-05, + "loss": 2.4665, + "step": 35849 + }, + { + "epoch": 1.53, + "grad_norm": 0.421875, + "learning_rate": 2.6390079401296407e-05, + "loss": 2.3105, + "step": 35850 + }, + { + "epoch": 1.53, + "grad_norm": 0.486328125, + "learning_rate": 2.6385539144705374e-05, + "loss": 2.4924, + "step": 35851 + }, + { + "epoch": 1.53, + "grad_norm": 0.44921875, + "learning_rate": 2.6380999219352508e-05, + "loss": 2.4023, + "step": 35852 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.6376459625258222e-05, + "loss": 2.4129, + "step": 35853 + }, + { + "epoch": 1.53, + "grad_norm": 0.49609375, + "learning_rate": 2.6371920362442947e-05, + "loss": 2.4671, + "step": 35854 + }, + { + "epoch": 1.53, + "grad_norm": 0.546875, + "learning_rate": 2.6367381430927164e-05, + "loss": 2.3447, + "step": 35855 + }, + { + "epoch": 1.53, + "grad_norm": 0.4140625, + "learning_rate": 2.6362842830731215e-05, + "loss": 2.495, + "step": 35856 + }, + { + "epoch": 1.53, + "grad_norm": 0.486328125, + "learning_rate": 2.6358304561875567e-05, + "loss": 2.4271, + "step": 35857 + }, + { + "epoch": 1.53, + "grad_norm": 0.4609375, + "learning_rate": 2.635376662438066e-05, + "loss": 2.3509, + "step": 35858 + }, + { + "epoch": 1.53, + "grad_norm": 0.4296875, + "learning_rate": 2.6349229018266897e-05, + "loss": 2.3352, + "step": 35859 + }, + { + "epoch": 1.53, + "grad_norm": 0.49609375, + "learning_rate": 2.634469174355465e-05, + "loss": 2.5088, + "step": 35860 + }, + { + "epoch": 1.53, + "grad_norm": 0.494140625, + "learning_rate": 2.6340154800264382e-05, + "loss": 2.4631, + "step": 35861 + }, + { + "epoch": 1.53, + "grad_norm": 0.4375, + "learning_rate": 2.633561818841652e-05, + "loss": 2.519, + "step": 35862 + }, + { + "epoch": 1.53, + "grad_norm": 0.44140625, + "learning_rate": 2.633108190803142e-05, + "loss": 2.3484, + "step": 35863 + }, + { + "epoch": 1.53, + "grad_norm": 0.490234375, + "learning_rate": 2.6326545959129568e-05, + "loss": 2.5041, + "step": 35864 + }, + { + "epoch": 1.53, + "grad_norm": 0.43359375, + "learning_rate": 2.6322010341731306e-05, + "loss": 2.5631, + "step": 35865 + }, + { + "epoch": 1.53, + "grad_norm": 0.478515625, + "learning_rate": 2.6317475055857068e-05, + "loss": 2.4669, + "step": 35866 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.6312940101527283e-05, + "loss": 2.483, + "step": 35867 + }, + { + "epoch": 1.53, + "grad_norm": 0.4375, + "learning_rate": 2.6308405478762355e-05, + "loss": 2.4004, + "step": 35868 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.630387118758263e-05, + "loss": 2.4746, + "step": 35869 + }, + { + "epoch": 1.53, + "grad_norm": 0.498046875, + "learning_rate": 2.629933722800856e-05, + "loss": 2.3699, + "step": 35870 + }, + { + "epoch": 1.53, + "grad_norm": 0.484375, + "learning_rate": 2.629480360006056e-05, + "loss": 2.4512, + "step": 35871 + }, + { + "epoch": 1.53, + "grad_norm": 0.48046875, + "learning_rate": 2.629027030375898e-05, + "loss": 2.4595, + "step": 35872 + }, + { + "epoch": 1.53, + "grad_norm": 0.416015625, + "learning_rate": 2.628573733912425e-05, + "loss": 2.4621, + "step": 35873 + }, + { + "epoch": 1.53, + "grad_norm": 0.40625, + "learning_rate": 2.628120470617679e-05, + "loss": 2.3326, + "step": 35874 + }, + { + "epoch": 1.53, + "grad_norm": 0.455078125, + "learning_rate": 2.6276672404936964e-05, + "loss": 2.3522, + "step": 35875 + }, + { + "epoch": 1.53, + "grad_norm": 0.451171875, + "learning_rate": 2.6272140435425142e-05, + "loss": 2.4502, + "step": 35876 + }, + { + "epoch": 1.53, + "grad_norm": 0.421875, + "learning_rate": 2.6267608797661746e-05, + "loss": 2.3239, + "step": 35877 + }, + { + "epoch": 1.53, + "grad_norm": 0.498046875, + "learning_rate": 2.62630774916672e-05, + "loss": 2.4227, + "step": 35878 + }, + { + "epoch": 1.53, + "grad_norm": 0.41015625, + "learning_rate": 2.625854651746181e-05, + "loss": 2.5242, + "step": 35879 + }, + { + "epoch": 1.53, + "grad_norm": 0.443359375, + "learning_rate": 2.6254015875066052e-05, + "loss": 2.513, + "step": 35880 + }, + { + "epoch": 1.53, + "grad_norm": 0.43359375, + "learning_rate": 2.624948556450024e-05, + "loss": 2.3727, + "step": 35881 + }, + { + "epoch": 1.53, + "grad_norm": 0.515625, + "learning_rate": 2.6244955585784792e-05, + "loss": 2.5192, + "step": 35882 + }, + { + "epoch": 1.53, + "grad_norm": 0.470703125, + "learning_rate": 2.6240425938940117e-05, + "loss": 2.2992, + "step": 35883 + }, + { + "epoch": 1.53, + "grad_norm": 0.4296875, + "learning_rate": 2.6235896623986576e-05, + "loss": 2.4703, + "step": 35884 + }, + { + "epoch": 1.53, + "grad_norm": 0.462890625, + "learning_rate": 2.6231367640944472e-05, + "loss": 2.3826, + "step": 35885 + }, + { + "epoch": 1.53, + "grad_norm": 0.43359375, + "learning_rate": 2.6226838989834325e-05, + "loss": 2.4428, + "step": 35886 + }, + { + "epoch": 1.53, + "grad_norm": 0.455078125, + "learning_rate": 2.622231067067643e-05, + "loss": 2.5244, + "step": 35887 + }, + { + "epoch": 1.53, + "grad_norm": 0.51171875, + "learning_rate": 2.6217782683491155e-05, + "loss": 2.4894, + "step": 35888 + }, + { + "epoch": 1.53, + "grad_norm": 0.482421875, + "learning_rate": 2.621325502829889e-05, + "loss": 2.4771, + "step": 35889 + }, + { + "epoch": 1.53, + "grad_norm": 0.47265625, + "learning_rate": 2.620872770512004e-05, + "loss": 2.4798, + "step": 35890 + }, + { + "epoch": 1.53, + "grad_norm": 0.46875, + "learning_rate": 2.6204200713974946e-05, + "loss": 2.5311, + "step": 35891 + }, + { + "epoch": 1.53, + "grad_norm": 0.46875, + "learning_rate": 2.619967405488395e-05, + "loss": 2.5534, + "step": 35892 + }, + { + "epoch": 1.53, + "grad_norm": 0.4296875, + "learning_rate": 2.6195147727867485e-05, + "loss": 2.503, + "step": 35893 + }, + { + "epoch": 1.53, + "grad_norm": 0.4921875, + "learning_rate": 2.6190621732945853e-05, + "loss": 2.3158, + "step": 35894 + }, + { + "epoch": 1.53, + "grad_norm": 0.470703125, + "learning_rate": 2.6186096070139443e-05, + "loss": 2.3798, + "step": 35895 + }, + { + "epoch": 1.53, + "grad_norm": 0.474609375, + "learning_rate": 2.618157073946865e-05, + "loss": 2.5206, + "step": 35896 + }, + { + "epoch": 1.53, + "grad_norm": 0.466796875, + "learning_rate": 2.617704574095382e-05, + "loss": 2.4755, + "step": 35897 + }, + { + "epoch": 1.53, + "grad_norm": 0.453125, + "learning_rate": 2.6172521074615242e-05, + "loss": 2.4111, + "step": 35898 + }, + { + "epoch": 1.53, + "grad_norm": 0.486328125, + "learning_rate": 2.61679967404734e-05, + "loss": 2.4125, + "step": 35899 + }, + { + "epoch": 1.53, + "grad_norm": 0.45703125, + "learning_rate": 2.616347273854859e-05, + "loss": 2.4973, + "step": 35900 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.6158949068861138e-05, + "loss": 2.3549, + "step": 35901 + }, + { + "epoch": 1.53, + "grad_norm": 0.404296875, + "learning_rate": 2.6154425731431432e-05, + "loss": 2.38, + "step": 35902 + }, + { + "epoch": 1.53, + "grad_norm": 0.53125, + "learning_rate": 2.614990272627985e-05, + "loss": 2.4454, + "step": 35903 + }, + { + "epoch": 1.53, + "grad_norm": 0.48046875, + "learning_rate": 2.6145380053426682e-05, + "loss": 2.4979, + "step": 35904 + }, + { + "epoch": 1.53, + "grad_norm": 0.447265625, + "learning_rate": 2.6140857712892352e-05, + "loss": 2.4734, + "step": 35905 + }, + { + "epoch": 1.53, + "grad_norm": 0.453125, + "learning_rate": 2.6136335704697136e-05, + "loss": 2.489, + "step": 35906 + }, + { + "epoch": 1.53, + "grad_norm": 0.4140625, + "learning_rate": 2.6131814028861446e-05, + "loss": 2.4358, + "step": 35907 + }, + { + "epoch": 1.53, + "grad_norm": 0.498046875, + "learning_rate": 2.6127292685405567e-05, + "loss": 2.3596, + "step": 35908 + }, + { + "epoch": 1.53, + "grad_norm": 0.5, + "learning_rate": 2.61227716743499e-05, + "loss": 2.4332, + "step": 35909 + }, + { + "epoch": 1.53, + "grad_norm": 0.486328125, + "learning_rate": 2.611825099571473e-05, + "loss": 2.6247, + "step": 35910 + }, + { + "epoch": 1.53, + "grad_norm": 0.51171875, + "learning_rate": 2.611373064952044e-05, + "loss": 2.4064, + "step": 35911 + }, + { + "epoch": 1.53, + "grad_norm": 0.451171875, + "learning_rate": 2.6109210635787384e-05, + "loss": 2.3942, + "step": 35912 + }, + { + "epoch": 1.53, + "grad_norm": 0.40625, + "learning_rate": 2.6104690954535872e-05, + "loss": 2.5463, + "step": 35913 + }, + { + "epoch": 1.53, + "grad_norm": 0.453125, + "learning_rate": 2.6100171605786182e-05, + "loss": 2.4194, + "step": 35914 + }, + { + "epoch": 1.53, + "grad_norm": 0.453125, + "learning_rate": 2.6095652589558784e-05, + "loss": 2.5704, + "step": 35915 + }, + { + "epoch": 1.53, + "grad_norm": 0.435546875, + "learning_rate": 2.6091133905873933e-05, + "loss": 2.3721, + "step": 35916 + }, + { + "epoch": 1.53, + "grad_norm": 0.47265625, + "learning_rate": 2.6086615554751946e-05, + "loss": 2.4698, + "step": 35917 + }, + { + "epoch": 1.53, + "grad_norm": 0.51171875, + "learning_rate": 2.6082097536213168e-05, + "loss": 2.5172, + "step": 35918 + }, + { + "epoch": 1.53, + "grad_norm": 0.49609375, + "learning_rate": 2.6077579850277976e-05, + "loss": 2.4242, + "step": 35919 + }, + { + "epoch": 1.53, + "grad_norm": 0.44140625, + "learning_rate": 2.6073062496966626e-05, + "loss": 2.4329, + "step": 35920 + }, + { + "epoch": 1.53, + "grad_norm": 0.50390625, + "learning_rate": 2.60685454762995e-05, + "loss": 2.4013, + "step": 35921 + }, + { + "epoch": 1.53, + "grad_norm": 0.466796875, + "learning_rate": 2.606402878829688e-05, + "loss": 2.4914, + "step": 35922 + }, + { + "epoch": 1.53, + "grad_norm": 0.43359375, + "learning_rate": 2.605951243297914e-05, + "loss": 2.5278, + "step": 35923 + }, + { + "epoch": 1.53, + "grad_norm": 0.458984375, + "learning_rate": 2.6054996410366537e-05, + "loss": 2.4713, + "step": 35924 + }, + { + "epoch": 1.53, + "grad_norm": 0.470703125, + "learning_rate": 2.6050480720479455e-05, + "loss": 2.4027, + "step": 35925 + }, + { + "epoch": 1.53, + "grad_norm": 0.51171875, + "learning_rate": 2.6045965363338152e-05, + "loss": 2.5446, + "step": 35926 + }, + { + "epoch": 1.53, + "grad_norm": 0.44140625, + "learning_rate": 2.6041450338962993e-05, + "loss": 2.4286, + "step": 35927 + }, + { + "epoch": 1.53, + "grad_norm": 0.3984375, + "learning_rate": 2.6036935647374295e-05, + "loss": 2.2988, + "step": 35928 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.6032421288592357e-05, + "loss": 2.4006, + "step": 35929 + }, + { + "epoch": 1.53, + "grad_norm": 0.462890625, + "learning_rate": 2.602790726263743e-05, + "loss": 2.4581, + "step": 35930 + }, + { + "epoch": 1.53, + "grad_norm": 0.4140625, + "learning_rate": 2.602339356952995e-05, + "loss": 2.3779, + "step": 35931 + }, + { + "epoch": 1.53, + "grad_norm": 0.4609375, + "learning_rate": 2.6018880209290164e-05, + "loss": 2.4564, + "step": 35932 + }, + { + "epoch": 1.53, + "grad_norm": 0.4296875, + "learning_rate": 2.601436718193835e-05, + "loss": 2.3115, + "step": 35933 + }, + { + "epoch": 1.53, + "grad_norm": 0.42578125, + "learning_rate": 2.6009854487494845e-05, + "loss": 2.3651, + "step": 35934 + }, + { + "epoch": 1.53, + "grad_norm": 0.58984375, + "learning_rate": 2.600534212597999e-05, + "loss": 2.6292, + "step": 35935 + }, + { + "epoch": 1.53, + "grad_norm": 0.458984375, + "learning_rate": 2.600083009741402e-05, + "loss": 2.3257, + "step": 35936 + }, + { + "epoch": 1.53, + "grad_norm": 0.49609375, + "learning_rate": 2.59963184018173e-05, + "loss": 2.3535, + "step": 35937 + }, + { + "epoch": 1.53, + "grad_norm": 0.4921875, + "learning_rate": 2.599180703921007e-05, + "loss": 2.5015, + "step": 35938 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.5987296009612662e-05, + "loss": 2.4372, + "step": 35939 + }, + { + "epoch": 1.53, + "grad_norm": 0.40234375, + "learning_rate": 2.59827853130454e-05, + "loss": 2.3558, + "step": 35940 + }, + { + "epoch": 1.53, + "grad_norm": 0.416015625, + "learning_rate": 2.597827494952856e-05, + "loss": 2.4414, + "step": 35941 + }, + { + "epoch": 1.53, + "grad_norm": 0.46875, + "learning_rate": 2.59737649190824e-05, + "loss": 2.4359, + "step": 35942 + }, + { + "epoch": 1.53, + "grad_norm": 0.455078125, + "learning_rate": 2.596925522172724e-05, + "loss": 2.4269, + "step": 35943 + }, + { + "epoch": 1.53, + "grad_norm": 0.50390625, + "learning_rate": 2.5964745857483408e-05, + "loss": 2.3776, + "step": 35944 + }, + { + "epoch": 1.53, + "grad_norm": 0.451171875, + "learning_rate": 2.5960236826371165e-05, + "loss": 2.4395, + "step": 35945 + }, + { + "epoch": 1.53, + "grad_norm": 0.443359375, + "learning_rate": 2.595572812841074e-05, + "loss": 2.42, + "step": 35946 + }, + { + "epoch": 1.53, + "grad_norm": 0.490234375, + "learning_rate": 2.5951219763622525e-05, + "loss": 2.3452, + "step": 35947 + }, + { + "epoch": 1.53, + "grad_norm": 0.466796875, + "learning_rate": 2.5946711732026775e-05, + "loss": 2.4089, + "step": 35948 + }, + { + "epoch": 1.53, + "grad_norm": 0.515625, + "learning_rate": 2.5942204033643723e-05, + "loss": 2.5561, + "step": 35949 + }, + { + "epoch": 1.53, + "grad_norm": 0.5, + "learning_rate": 2.5937696668493692e-05, + "loss": 2.4725, + "step": 35950 + }, + { + "epoch": 1.53, + "grad_norm": 0.50390625, + "learning_rate": 2.5933189636596977e-05, + "loss": 2.4201, + "step": 35951 + }, + { + "epoch": 1.53, + "grad_norm": 0.494140625, + "learning_rate": 2.5928682937973825e-05, + "loss": 2.4167, + "step": 35952 + }, + { + "epoch": 1.53, + "grad_norm": 0.427734375, + "learning_rate": 2.5924176572644555e-05, + "loss": 2.4205, + "step": 35953 + }, + { + "epoch": 1.53, + "grad_norm": 0.435546875, + "learning_rate": 2.5919670540629383e-05, + "loss": 2.2375, + "step": 35954 + }, + { + "epoch": 1.53, + "grad_norm": 0.4921875, + "learning_rate": 2.5915164841948626e-05, + "loss": 2.3353, + "step": 35955 + }, + { + "epoch": 1.53, + "grad_norm": 0.416015625, + "learning_rate": 2.591065947662258e-05, + "loss": 2.5361, + "step": 35956 + }, + { + "epoch": 1.53, + "grad_norm": 0.47265625, + "learning_rate": 2.5906154444671494e-05, + "loss": 2.3414, + "step": 35957 + }, + { + "epoch": 1.53, + "grad_norm": 0.466796875, + "learning_rate": 2.5901649746115596e-05, + "loss": 2.4666, + "step": 35958 + }, + { + "epoch": 1.53, + "grad_norm": 0.5546875, + "learning_rate": 2.589714538097521e-05, + "loss": 2.6066, + "step": 35959 + }, + { + "epoch": 1.53, + "grad_norm": 0.50390625, + "learning_rate": 2.5892641349270596e-05, + "loss": 2.463, + "step": 35960 + }, + { + "epoch": 1.53, + "grad_norm": 0.40234375, + "learning_rate": 2.5888137651022027e-05, + "loss": 2.3424, + "step": 35961 + }, + { + "epoch": 1.53, + "grad_norm": 0.458984375, + "learning_rate": 2.5883634286249692e-05, + "loss": 2.518, + "step": 35962 + }, + { + "epoch": 1.53, + "grad_norm": 0.57421875, + "learning_rate": 2.5879131254973975e-05, + "loss": 2.482, + "step": 35963 + }, + { + "epoch": 1.53, + "grad_norm": 0.384765625, + "learning_rate": 2.5874628557215086e-05, + "loss": 2.3293, + "step": 35964 + }, + { + "epoch": 1.53, + "grad_norm": 0.447265625, + "learning_rate": 2.5870126192993238e-05, + "loss": 2.5141, + "step": 35965 + }, + { + "epoch": 1.53, + "grad_norm": 0.3984375, + "learning_rate": 2.586562416232874e-05, + "loss": 2.3674, + "step": 35966 + }, + { + "epoch": 1.53, + "grad_norm": 0.482421875, + "learning_rate": 2.586112246524187e-05, + "loss": 2.3519, + "step": 35967 + }, + { + "epoch": 1.53, + "grad_norm": 0.4609375, + "learning_rate": 2.585662110175282e-05, + "loss": 2.4883, + "step": 35968 + }, + { + "epoch": 1.53, + "grad_norm": 0.47265625, + "learning_rate": 2.5852120071881903e-05, + "loss": 2.3145, + "step": 35969 + }, + { + "epoch": 1.53, + "grad_norm": 0.490234375, + "learning_rate": 2.5847619375649322e-05, + "loss": 2.5563, + "step": 35970 + }, + { + "epoch": 1.53, + "grad_norm": 0.43359375, + "learning_rate": 2.5843119013075355e-05, + "loss": 2.4529, + "step": 35971 + }, + { + "epoch": 1.53, + "grad_norm": 0.404296875, + "learning_rate": 2.5838618984180286e-05, + "loss": 2.4595, + "step": 35972 + }, + { + "epoch": 1.53, + "grad_norm": 0.484375, + "learning_rate": 2.5834119288984327e-05, + "loss": 2.5225, + "step": 35973 + }, + { + "epoch": 1.53, + "grad_norm": 0.44921875, + "learning_rate": 2.5829619927507697e-05, + "loss": 2.3277, + "step": 35974 + }, + { + "epoch": 1.53, + "grad_norm": 0.44921875, + "learning_rate": 2.582512089977066e-05, + "loss": 2.4779, + "step": 35975 + }, + { + "epoch": 1.53, + "grad_norm": 0.53125, + "learning_rate": 2.582062220579352e-05, + "loss": 2.4209, + "step": 35976 + }, + { + "epoch": 1.53, + "grad_norm": 0.427734375, + "learning_rate": 2.5816123845596452e-05, + "loss": 2.4223, + "step": 35977 + }, + { + "epoch": 1.53, + "grad_norm": 0.498046875, + "learning_rate": 2.581162581919967e-05, + "loss": 2.4604, + "step": 35978 + }, + { + "epoch": 1.53, + "grad_norm": 0.42578125, + "learning_rate": 2.5807128126623513e-05, + "loss": 2.3289, + "step": 35979 + }, + { + "epoch": 1.53, + "grad_norm": 0.470703125, + "learning_rate": 2.580263076788816e-05, + "loss": 2.515, + "step": 35980 + }, + { + "epoch": 1.53, + "grad_norm": 0.51171875, + "learning_rate": 2.5798133743013832e-05, + "loss": 2.4516, + "step": 35981 + }, + { + "epoch": 1.53, + "grad_norm": 0.4453125, + "learning_rate": 2.579363705202078e-05, + "loss": 2.4622, + "step": 35982 + }, + { + "epoch": 1.53, + "grad_norm": 0.546875, + "learning_rate": 2.578914069492927e-05, + "loss": 2.5516, + "step": 35983 + }, + { + "epoch": 1.53, + "grad_norm": 0.4296875, + "learning_rate": 2.5784644671759483e-05, + "loss": 2.4646, + "step": 35984 + }, + { + "epoch": 1.53, + "grad_norm": 0.4296875, + "learning_rate": 2.5780148982531692e-05, + "loss": 2.4973, + "step": 35985 + }, + { + "epoch": 1.53, + "grad_norm": 0.447265625, + "learning_rate": 2.5775653627266084e-05, + "loss": 2.3682, + "step": 35986 + }, + { + "epoch": 1.53, + "grad_norm": 0.462890625, + "learning_rate": 2.5771158605982905e-05, + "loss": 2.3788, + "step": 35987 + }, + { + "epoch": 1.53, + "grad_norm": 0.4765625, + "learning_rate": 2.5766663918702415e-05, + "loss": 2.4861, + "step": 35988 + }, + { + "epoch": 1.53, + "grad_norm": 0.453125, + "learning_rate": 2.5762169565444817e-05, + "loss": 2.5136, + "step": 35989 + }, + { + "epoch": 1.53, + "grad_norm": 0.4453125, + "learning_rate": 2.5757675546230288e-05, + "loss": 2.3877, + "step": 35990 + }, + { + "epoch": 1.53, + "grad_norm": 0.45703125, + "learning_rate": 2.575318186107908e-05, + "loss": 2.5772, + "step": 35991 + }, + { + "epoch": 1.53, + "grad_norm": 0.427734375, + "learning_rate": 2.5748688510011453e-05, + "loss": 2.4892, + "step": 35992 + }, + { + "epoch": 1.53, + "grad_norm": 0.42578125, + "learning_rate": 2.574419549304756e-05, + "loss": 2.3084, + "step": 35993 + }, + { + "epoch": 1.53, + "grad_norm": 0.5078125, + "learning_rate": 2.5739702810207667e-05, + "loss": 2.4891, + "step": 35994 + }, + { + "epoch": 1.53, + "grad_norm": 0.42578125, + "learning_rate": 2.5735210461511984e-05, + "loss": 2.4557, + "step": 35995 + }, + { + "epoch": 1.53, + "grad_norm": 0.462890625, + "learning_rate": 2.5730718446980716e-05, + "loss": 2.5569, + "step": 35996 + }, + { + "epoch": 1.53, + "grad_norm": 0.51171875, + "learning_rate": 2.5726226766634053e-05, + "loss": 2.5447, + "step": 35997 + }, + { + "epoch": 1.53, + "grad_norm": 0.47265625, + "learning_rate": 2.572173542049221e-05, + "loss": 2.4869, + "step": 35998 + }, + { + "epoch": 1.53, + "grad_norm": 0.494140625, + "learning_rate": 2.5717244408575458e-05, + "loss": 2.5321, + "step": 35999 + }, + { + "epoch": 1.53, + "grad_norm": 0.46875, + "learning_rate": 2.5712753730903917e-05, + "loss": 2.4085, + "step": 36000 + }, + { + "epoch": 1.53, + "grad_norm": 0.53515625, + "learning_rate": 2.5708263387497866e-05, + "loss": 2.5641, + "step": 36001 + }, + { + "epoch": 1.53, + "grad_norm": 0.4765625, + "learning_rate": 2.570377337837746e-05, + "loss": 2.5509, + "step": 36002 + }, + { + "epoch": 1.53, + "grad_norm": 0.53125, + "learning_rate": 2.5699283703562917e-05, + "loss": 2.5148, + "step": 36003 + }, + { + "epoch": 1.53, + "grad_norm": 0.47265625, + "learning_rate": 2.569479436307447e-05, + "loss": 2.5347, + "step": 36004 + }, + { + "epoch": 1.53, + "grad_norm": 0.4609375, + "learning_rate": 2.569030535693229e-05, + "loss": 2.4433, + "step": 36005 + }, + { + "epoch": 1.53, + "grad_norm": 0.458984375, + "learning_rate": 2.568581668515655e-05, + "loss": 2.5384, + "step": 36006 + }, + { + "epoch": 1.53, + "grad_norm": 0.50390625, + "learning_rate": 2.5681328347767487e-05, + "loss": 2.4778, + "step": 36007 + }, + { + "epoch": 1.53, + "grad_norm": 0.45703125, + "learning_rate": 2.5676840344785304e-05, + "loss": 2.4097, + "step": 36008 + }, + { + "epoch": 1.53, + "grad_norm": 0.48828125, + "learning_rate": 2.567235267623015e-05, + "loss": 2.5239, + "step": 36009 + }, + { + "epoch": 1.53, + "grad_norm": 0.5, + "learning_rate": 2.5667865342122245e-05, + "loss": 2.3922, + "step": 36010 + }, + { + "epoch": 1.53, + "grad_norm": 0.478515625, + "learning_rate": 2.5663378342481813e-05, + "loss": 2.515, + "step": 36011 + }, + { + "epoch": 1.53, + "grad_norm": 0.408203125, + "learning_rate": 2.565889167732901e-05, + "loss": 2.3054, + "step": 36012 + }, + { + "epoch": 1.53, + "grad_norm": 0.4765625, + "learning_rate": 2.5654405346683995e-05, + "loss": 2.3609, + "step": 36013 + }, + { + "epoch": 1.53, + "grad_norm": 0.458984375, + "learning_rate": 2.5649919350566986e-05, + "loss": 2.3584, + "step": 36014 + }, + { + "epoch": 1.53, + "grad_norm": 0.439453125, + "learning_rate": 2.56454336889982e-05, + "loss": 2.48, + "step": 36015 + }, + { + "epoch": 1.53, + "grad_norm": 0.515625, + "learning_rate": 2.5640948361997763e-05, + "loss": 2.5758, + "step": 36016 + }, + { + "epoch": 1.53, + "grad_norm": 0.4765625, + "learning_rate": 2.5636463369585906e-05, + "loss": 2.4123, + "step": 36017 + }, + { + "epoch": 1.53, + "grad_norm": 0.4765625, + "learning_rate": 2.563197871178279e-05, + "loss": 2.5116, + "step": 36018 + }, + { + "epoch": 1.53, + "grad_norm": 0.458984375, + "learning_rate": 2.562749438860853e-05, + "loss": 2.3772, + "step": 36019 + }, + { + "epoch": 1.53, + "grad_norm": 0.462890625, + "learning_rate": 2.5623010400083435e-05, + "loss": 2.3975, + "step": 36020 + }, + { + "epoch": 1.53, + "grad_norm": 0.478515625, + "learning_rate": 2.56185267462276e-05, + "loss": 2.5394, + "step": 36021 + }, + { + "epoch": 1.53, + "grad_norm": 0.484375, + "learning_rate": 2.5614043427061185e-05, + "loss": 2.4375, + "step": 36022 + }, + { + "epoch": 1.54, + "grad_norm": 0.498046875, + "learning_rate": 2.5609560442604385e-05, + "loss": 2.4816, + "step": 36023 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.5605077792877408e-05, + "loss": 2.4395, + "step": 36024 + }, + { + "epoch": 1.54, + "grad_norm": 0.421875, + "learning_rate": 2.560059547790037e-05, + "loss": 2.4041, + "step": 36025 + }, + { + "epoch": 1.54, + "grad_norm": 0.421875, + "learning_rate": 2.5596113497693486e-05, + "loss": 2.4532, + "step": 36026 + }, + { + "epoch": 1.54, + "grad_norm": 0.4296875, + "learning_rate": 2.5591631852276876e-05, + "loss": 2.4937, + "step": 36027 + }, + { + "epoch": 1.54, + "grad_norm": 0.4375, + "learning_rate": 2.5587150541670747e-05, + "loss": 2.4998, + "step": 36028 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.5582669565895224e-05, + "loss": 2.5329, + "step": 36029 + }, + { + "epoch": 1.54, + "grad_norm": 0.3984375, + "learning_rate": 2.5578188924970525e-05, + "loss": 2.2949, + "step": 36030 + }, + { + "epoch": 1.54, + "grad_norm": 0.388671875, + "learning_rate": 2.5573708618916746e-05, + "loss": 2.5798, + "step": 36031 + }, + { + "epoch": 1.54, + "grad_norm": 0.46484375, + "learning_rate": 2.5569228647754073e-05, + "loss": 2.5283, + "step": 36032 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.5564749011502697e-05, + "loss": 2.3712, + "step": 36033 + }, + { + "epoch": 1.54, + "grad_norm": 0.470703125, + "learning_rate": 2.5560269710182762e-05, + "loss": 2.4761, + "step": 36034 + }, + { + "epoch": 1.54, + "grad_norm": 0.49609375, + "learning_rate": 2.5555790743814335e-05, + "loss": 2.4088, + "step": 36035 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.5551312112417715e-05, + "loss": 2.5654, + "step": 36036 + }, + { + "epoch": 1.54, + "grad_norm": 0.486328125, + "learning_rate": 2.5546833816012972e-05, + "loss": 2.7039, + "step": 36037 + }, + { + "epoch": 1.54, + "grad_norm": 0.421875, + "learning_rate": 2.5542355854620246e-05, + "loss": 2.4167, + "step": 36038 + }, + { + "epoch": 1.54, + "grad_norm": 0.474609375, + "learning_rate": 2.5537878228259704e-05, + "loss": 2.4824, + "step": 36039 + }, + { + "epoch": 1.54, + "grad_norm": 0.5234375, + "learning_rate": 2.5533400936951524e-05, + "loss": 2.457, + "step": 36040 + }, + { + "epoch": 1.54, + "grad_norm": 0.48828125, + "learning_rate": 2.5528923980715792e-05, + "loss": 2.496, + "step": 36041 + }, + { + "epoch": 1.54, + "grad_norm": 0.447265625, + "learning_rate": 2.5524447359572724e-05, + "loss": 2.5512, + "step": 36042 + }, + { + "epoch": 1.54, + "grad_norm": 0.486328125, + "learning_rate": 2.5519971073542392e-05, + "loss": 2.4488, + "step": 36043 + }, + { + "epoch": 1.54, + "grad_norm": 0.48828125, + "learning_rate": 2.5515495122645007e-05, + "loss": 2.5965, + "step": 36044 + }, + { + "epoch": 1.54, + "grad_norm": 0.47265625, + "learning_rate": 2.5511019506900635e-05, + "loss": 2.4159, + "step": 36045 + }, + { + "epoch": 1.54, + "grad_norm": 0.412109375, + "learning_rate": 2.5506544226329488e-05, + "loss": 2.2644, + "step": 36046 + }, + { + "epoch": 1.54, + "grad_norm": 0.4609375, + "learning_rate": 2.5502069280951636e-05, + "loss": 2.4517, + "step": 36047 + }, + { + "epoch": 1.54, + "grad_norm": 0.462890625, + "learning_rate": 2.549759467078724e-05, + "loss": 2.455, + "step": 36048 + }, + { + "epoch": 1.54, + "grad_norm": 0.439453125, + "learning_rate": 2.5493120395856473e-05, + "loss": 2.206, + "step": 36049 + }, + { + "epoch": 1.54, + "grad_norm": 0.5234375, + "learning_rate": 2.5488646456179444e-05, + "loss": 2.4121, + "step": 36050 + }, + { + "epoch": 1.54, + "grad_norm": 0.439453125, + "learning_rate": 2.5484172851776212e-05, + "loss": 2.4746, + "step": 36051 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.5479699582667015e-05, + "loss": 2.4318, + "step": 36052 + }, + { + "epoch": 1.54, + "grad_norm": 0.478515625, + "learning_rate": 2.547522664887195e-05, + "loss": 2.4709, + "step": 36053 + }, + { + "epoch": 1.54, + "grad_norm": 0.453125, + "learning_rate": 2.5470754050411106e-05, + "loss": 2.4584, + "step": 36054 + }, + { + "epoch": 1.54, + "grad_norm": 0.4921875, + "learning_rate": 2.5466281787304614e-05, + "loss": 2.5255, + "step": 36055 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.5461809859572662e-05, + "loss": 2.4286, + "step": 36056 + }, + { + "epoch": 1.54, + "grad_norm": 0.486328125, + "learning_rate": 2.5457338267235288e-05, + "loss": 2.5026, + "step": 36057 + }, + { + "epoch": 1.54, + "grad_norm": 0.419921875, + "learning_rate": 2.5452867010312675e-05, + "loss": 2.5768, + "step": 36058 + }, + { + "epoch": 1.54, + "grad_norm": 0.5390625, + "learning_rate": 2.5448396088824887e-05, + "loss": 2.3805, + "step": 36059 + }, + { + "epoch": 1.54, + "grad_norm": 0.50390625, + "learning_rate": 2.54439255027921e-05, + "loss": 2.5893, + "step": 36060 + }, + { + "epoch": 1.54, + "grad_norm": 0.421875, + "learning_rate": 2.543945525223438e-05, + "loss": 2.4076, + "step": 36061 + }, + { + "epoch": 1.54, + "grad_norm": 0.478515625, + "learning_rate": 2.5434985337171878e-05, + "loss": 2.4057, + "step": 36062 + }, + { + "epoch": 1.54, + "grad_norm": 0.4609375, + "learning_rate": 2.5430515757624673e-05, + "loss": 2.5941, + "step": 36063 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.54260465136129e-05, + "loss": 2.3928, + "step": 36064 + }, + { + "epoch": 1.54, + "grad_norm": 0.453125, + "learning_rate": 2.5421577605156687e-05, + "loss": 2.4832, + "step": 36065 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.541710903227612e-05, + "loss": 2.5867, + "step": 36066 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.5412640794991237e-05, + "loss": 2.5594, + "step": 36067 + }, + { + "epoch": 1.54, + "grad_norm": 0.439453125, + "learning_rate": 2.540817289332229e-05, + "loss": 2.3953, + "step": 36068 + }, + { + "epoch": 1.54, + "grad_norm": 0.474609375, + "learning_rate": 2.5403705327289295e-05, + "loss": 2.4361, + "step": 36069 + }, + { + "epoch": 1.54, + "grad_norm": 0.515625, + "learning_rate": 2.5399238096912326e-05, + "loss": 2.4726, + "step": 36070 + }, + { + "epoch": 1.54, + "grad_norm": 0.447265625, + "learning_rate": 2.5394771202211532e-05, + "loss": 2.3621, + "step": 36071 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.539030464320703e-05, + "loss": 2.3729, + "step": 36072 + }, + { + "epoch": 1.54, + "grad_norm": 0.42578125, + "learning_rate": 2.538583841991886e-05, + "loss": 2.3806, + "step": 36073 + }, + { + "epoch": 1.54, + "grad_norm": 0.5390625, + "learning_rate": 2.5381372532367188e-05, + "loss": 2.5053, + "step": 36074 + }, + { + "epoch": 1.54, + "grad_norm": 0.5234375, + "learning_rate": 2.5376906980572025e-05, + "loss": 2.5919, + "step": 36075 + }, + { + "epoch": 1.54, + "grad_norm": 0.435546875, + "learning_rate": 2.5372441764553544e-05, + "loss": 2.4778, + "step": 36076 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.5367976884331778e-05, + "loss": 2.3321, + "step": 36077 + }, + { + "epoch": 1.54, + "grad_norm": 0.486328125, + "learning_rate": 2.5363512339926866e-05, + "loss": 2.5321, + "step": 36078 + }, + { + "epoch": 1.54, + "grad_norm": 0.5, + "learning_rate": 2.5359048131358842e-05, + "loss": 2.6764, + "step": 36079 + }, + { + "epoch": 1.54, + "grad_norm": 0.435546875, + "learning_rate": 2.5354584258647818e-05, + "loss": 2.4919, + "step": 36080 + }, + { + "epoch": 1.54, + "grad_norm": 0.494140625, + "learning_rate": 2.535012072181392e-05, + "loss": 2.5529, + "step": 36081 + }, + { + "epoch": 1.54, + "grad_norm": 0.462890625, + "learning_rate": 2.5345657520877198e-05, + "loss": 2.3837, + "step": 36082 + }, + { + "epoch": 1.54, + "grad_norm": 0.478515625, + "learning_rate": 2.534119465585768e-05, + "loss": 2.4887, + "step": 36083 + }, + { + "epoch": 1.54, + "grad_norm": 0.416015625, + "learning_rate": 2.5336732126775552e-05, + "loss": 2.497, + "step": 36084 + }, + { + "epoch": 1.54, + "grad_norm": 0.388671875, + "learning_rate": 2.533226993365084e-05, + "loss": 2.3759, + "step": 36085 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.5327808076503613e-05, + "loss": 2.5004, + "step": 36086 + }, + { + "epoch": 1.54, + "grad_norm": 0.47265625, + "learning_rate": 2.5323346555353944e-05, + "loss": 2.5094, + "step": 36087 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.5318885370221955e-05, + "loss": 2.5997, + "step": 36088 + }, + { + "epoch": 1.54, + "grad_norm": 0.51171875, + "learning_rate": 2.5314424521127668e-05, + "loss": 2.5025, + "step": 36089 + }, + { + "epoch": 1.54, + "grad_norm": 0.4921875, + "learning_rate": 2.5309964008091202e-05, + "loss": 2.456, + "step": 36090 + }, + { + "epoch": 1.54, + "grad_norm": 0.4609375, + "learning_rate": 2.5305503831132572e-05, + "loss": 2.5131, + "step": 36091 + }, + { + "epoch": 1.54, + "grad_norm": 0.515625, + "learning_rate": 2.53010439902719e-05, + "loss": 2.54, + "step": 36092 + }, + { + "epoch": 1.54, + "grad_norm": 0.515625, + "learning_rate": 2.5296584485529206e-05, + "loss": 2.5091, + "step": 36093 + }, + { + "epoch": 1.54, + "grad_norm": 0.48046875, + "learning_rate": 2.529212531692462e-05, + "loss": 2.3908, + "step": 36094 + }, + { + "epoch": 1.54, + "grad_norm": 0.43359375, + "learning_rate": 2.528766648447812e-05, + "loss": 2.3447, + "step": 36095 + }, + { + "epoch": 1.54, + "grad_norm": 0.515625, + "learning_rate": 2.5283207988209834e-05, + "loss": 2.5054, + "step": 36096 + }, + { + "epoch": 1.54, + "grad_norm": 0.484375, + "learning_rate": 2.5278749828139826e-05, + "loss": 2.4012, + "step": 36097 + }, + { + "epoch": 1.54, + "grad_norm": 0.47265625, + "learning_rate": 2.5274292004288136e-05, + "loss": 2.5901, + "step": 36098 + }, + { + "epoch": 1.54, + "grad_norm": 0.478515625, + "learning_rate": 2.5269834516674773e-05, + "loss": 2.3741, + "step": 36099 + }, + { + "epoch": 1.54, + "grad_norm": 0.51953125, + "learning_rate": 2.5265377365319887e-05, + "loss": 2.5384, + "step": 36100 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.5260920550243504e-05, + "loss": 2.5182, + "step": 36101 + }, + { + "epoch": 1.54, + "grad_norm": 0.578125, + "learning_rate": 2.5256464071465623e-05, + "loss": 2.5541, + "step": 36102 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.525200792900635e-05, + "loss": 2.4251, + "step": 36103 + }, + { + "epoch": 1.54, + "grad_norm": 0.4609375, + "learning_rate": 2.5247552122885755e-05, + "loss": 2.4696, + "step": 36104 + }, + { + "epoch": 1.54, + "grad_norm": 0.466796875, + "learning_rate": 2.524309665312382e-05, + "loss": 2.3727, + "step": 36105 + }, + { + "epoch": 1.54, + "grad_norm": 0.50390625, + "learning_rate": 2.5238641519740658e-05, + "loss": 2.4521, + "step": 36106 + }, + { + "epoch": 1.54, + "grad_norm": 0.435546875, + "learning_rate": 2.5234186722756258e-05, + "loss": 2.4238, + "step": 36107 + }, + { + "epoch": 1.54, + "grad_norm": 0.451171875, + "learning_rate": 2.5229732262190732e-05, + "loss": 2.5827, + "step": 36108 + }, + { + "epoch": 1.54, + "grad_norm": 0.470703125, + "learning_rate": 2.5225278138064047e-05, + "loss": 2.392, + "step": 36109 + }, + { + "epoch": 1.54, + "grad_norm": 0.443359375, + "learning_rate": 2.522082435039631e-05, + "loss": 2.4435, + "step": 36110 + }, + { + "epoch": 1.54, + "grad_norm": 0.5, + "learning_rate": 2.521637089920751e-05, + "loss": 2.4983, + "step": 36111 + }, + { + "epoch": 1.54, + "grad_norm": 0.5234375, + "learning_rate": 2.5211917784517713e-05, + "loss": 2.5167, + "step": 36112 + }, + { + "epoch": 1.54, + "grad_norm": 0.427734375, + "learning_rate": 2.5207465006346977e-05, + "loss": 2.5412, + "step": 36113 + }, + { + "epoch": 1.54, + "grad_norm": 0.455078125, + "learning_rate": 2.5203012564715323e-05, + "loss": 2.4555, + "step": 36114 + }, + { + "epoch": 1.54, + "grad_norm": 0.494140625, + "learning_rate": 2.5198560459642717e-05, + "loss": 2.3814, + "step": 36115 + }, + { + "epoch": 1.54, + "grad_norm": 0.53125, + "learning_rate": 2.519410869114931e-05, + "loss": 2.574, + "step": 36116 + }, + { + "epoch": 1.54, + "grad_norm": 0.53125, + "learning_rate": 2.5189657259255073e-05, + "loss": 2.5023, + "step": 36117 + }, + { + "epoch": 1.54, + "grad_norm": 0.466796875, + "learning_rate": 2.5185206163980002e-05, + "loss": 2.4185, + "step": 36118 + }, + { + "epoch": 1.54, + "grad_norm": 0.48046875, + "learning_rate": 2.518075540534417e-05, + "loss": 2.4597, + "step": 36119 + }, + { + "epoch": 1.54, + "grad_norm": 0.431640625, + "learning_rate": 2.5176304983367626e-05, + "loss": 2.4361, + "step": 36120 + }, + { + "epoch": 1.54, + "grad_norm": 0.48828125, + "learning_rate": 2.5171854898070334e-05, + "loss": 2.4111, + "step": 36121 + }, + { + "epoch": 1.54, + "grad_norm": 0.435546875, + "learning_rate": 2.516740514947238e-05, + "loss": 2.5272, + "step": 36122 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.516295573759372e-05, + "loss": 2.4058, + "step": 36123 + }, + { + "epoch": 1.54, + "grad_norm": 0.4921875, + "learning_rate": 2.515850666245444e-05, + "loss": 2.4937, + "step": 36124 + }, + { + "epoch": 1.54, + "grad_norm": 0.486328125, + "learning_rate": 2.5154057924074503e-05, + "loss": 2.5312, + "step": 36125 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.5149609522473983e-05, + "loss": 2.4212, + "step": 36126 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.5145161457672828e-05, + "loss": 2.4777, + "step": 36127 + }, + { + "epoch": 1.54, + "grad_norm": 0.51171875, + "learning_rate": 2.514071372969109e-05, + "loss": 2.4922, + "step": 36128 + }, + { + "epoch": 1.54, + "grad_norm": 0.4140625, + "learning_rate": 2.513626633854882e-05, + "loss": 2.5607, + "step": 36129 + }, + { + "epoch": 1.54, + "grad_norm": 0.5, + "learning_rate": 2.513181928426599e-05, + "loss": 2.3586, + "step": 36130 + }, + { + "epoch": 1.54, + "grad_norm": 0.41796875, + "learning_rate": 2.512737256686255e-05, + "loss": 2.5084, + "step": 36131 + }, + { + "epoch": 1.54, + "grad_norm": 0.431640625, + "learning_rate": 2.512292618635863e-05, + "loss": 2.3747, + "step": 36132 + }, + { + "epoch": 1.54, + "grad_norm": 0.486328125, + "learning_rate": 2.511848014277418e-05, + "loss": 2.5197, + "step": 36133 + }, + { + "epoch": 1.54, + "grad_norm": 0.44921875, + "learning_rate": 2.5114034436129174e-05, + "loss": 2.5356, + "step": 36134 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.510958906644364e-05, + "loss": 2.411, + "step": 36135 + }, + { + "epoch": 1.54, + "grad_norm": 0.490234375, + "learning_rate": 2.5105144033737615e-05, + "loss": 2.5635, + "step": 36136 + }, + { + "epoch": 1.54, + "grad_norm": 0.41796875, + "learning_rate": 2.510069933803105e-05, + "loss": 2.4835, + "step": 36137 + }, + { + "epoch": 1.54, + "grad_norm": 0.49609375, + "learning_rate": 2.5096254979343993e-05, + "loss": 2.4032, + "step": 36138 + }, + { + "epoch": 1.54, + "grad_norm": 0.3828125, + "learning_rate": 2.5091810957696406e-05, + "loss": 2.4825, + "step": 36139 + }, + { + "epoch": 1.54, + "grad_norm": 0.431640625, + "learning_rate": 2.5087367273108277e-05, + "loss": 2.5077, + "step": 36140 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.50829239255996e-05, + "loss": 2.5097, + "step": 36141 + }, + { + "epoch": 1.54, + "grad_norm": 0.490234375, + "learning_rate": 2.5078480915190428e-05, + "loss": 2.4803, + "step": 36142 + }, + { + "epoch": 1.54, + "grad_norm": 0.4453125, + "learning_rate": 2.507403824190069e-05, + "loss": 2.4849, + "step": 36143 + }, + { + "epoch": 1.54, + "grad_norm": 0.48828125, + "learning_rate": 2.5069595905750377e-05, + "loss": 2.451, + "step": 36144 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.506515390675954e-05, + "loss": 2.481, + "step": 36145 + }, + { + "epoch": 1.54, + "grad_norm": 0.50390625, + "learning_rate": 2.5060712244948127e-05, + "loss": 2.3478, + "step": 36146 + }, + { + "epoch": 1.54, + "grad_norm": 0.4765625, + "learning_rate": 2.505627092033609e-05, + "loss": 2.4248, + "step": 36147 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.505182993294344e-05, + "loss": 2.5875, + "step": 36148 + }, + { + "epoch": 1.54, + "grad_norm": 0.474609375, + "learning_rate": 2.5047389282790203e-05, + "loss": 2.5065, + "step": 36149 + }, + { + "epoch": 1.54, + "grad_norm": 0.490234375, + "learning_rate": 2.5042948969896284e-05, + "loss": 2.3862, + "step": 36150 + }, + { + "epoch": 1.54, + "grad_norm": 0.48828125, + "learning_rate": 2.503850899428174e-05, + "loss": 2.6421, + "step": 36151 + }, + { + "epoch": 1.54, + "grad_norm": 0.54296875, + "learning_rate": 2.503406935596647e-05, + "loss": 2.4619, + "step": 36152 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.5029630054970497e-05, + "loss": 2.4461, + "step": 36153 + }, + { + "epoch": 1.54, + "grad_norm": 0.54296875, + "learning_rate": 2.5025191091313815e-05, + "loss": 2.5728, + "step": 36154 + }, + { + "epoch": 1.54, + "grad_norm": 0.45703125, + "learning_rate": 2.5020752465016383e-05, + "loss": 2.3576, + "step": 36155 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.501631417609813e-05, + "loss": 2.3518, + "step": 36156 + }, + { + "epoch": 1.54, + "grad_norm": 0.48046875, + "learning_rate": 2.5011876224579057e-05, + "loss": 2.5283, + "step": 36157 + }, + { + "epoch": 1.54, + "grad_norm": 0.48046875, + "learning_rate": 2.500743861047916e-05, + "loss": 2.4969, + "step": 36158 + }, + { + "epoch": 1.54, + "grad_norm": 0.439453125, + "learning_rate": 2.5003001333818365e-05, + "loss": 2.4285, + "step": 36159 + }, + { + "epoch": 1.54, + "grad_norm": 0.466796875, + "learning_rate": 2.499856439461665e-05, + "loss": 2.3314, + "step": 36160 + }, + { + "epoch": 1.54, + "grad_norm": 0.474609375, + "learning_rate": 2.4994127792894005e-05, + "loss": 2.5829, + "step": 36161 + }, + { + "epoch": 1.54, + "grad_norm": 0.470703125, + "learning_rate": 2.498969152867039e-05, + "loss": 2.4316, + "step": 36162 + }, + { + "epoch": 1.54, + "grad_norm": 0.470703125, + "learning_rate": 2.498525560196572e-05, + "loss": 2.4323, + "step": 36163 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.4980820012799977e-05, + "loss": 2.5343, + "step": 36164 + }, + { + "epoch": 1.54, + "grad_norm": 0.421875, + "learning_rate": 2.497638476119315e-05, + "loss": 2.3486, + "step": 36165 + }, + { + "epoch": 1.54, + "grad_norm": 0.447265625, + "learning_rate": 2.4971949847165153e-05, + "loss": 2.4519, + "step": 36166 + }, + { + "epoch": 1.54, + "grad_norm": 0.404296875, + "learning_rate": 2.4967515270735998e-05, + "loss": 2.4028, + "step": 36167 + }, + { + "epoch": 1.54, + "grad_norm": 0.47265625, + "learning_rate": 2.4963081031925562e-05, + "loss": 2.3931, + "step": 36168 + }, + { + "epoch": 1.54, + "grad_norm": 0.427734375, + "learning_rate": 2.4958647130753844e-05, + "loss": 2.4732, + "step": 36169 + }, + { + "epoch": 1.54, + "grad_norm": 0.48046875, + "learning_rate": 2.4954213567240815e-05, + "loss": 2.4464, + "step": 36170 + }, + { + "epoch": 1.54, + "grad_norm": 0.439453125, + "learning_rate": 2.4949780341406402e-05, + "loss": 2.5408, + "step": 36171 + }, + { + "epoch": 1.54, + "grad_norm": 0.478515625, + "learning_rate": 2.4945347453270517e-05, + "loss": 2.47, + "step": 36172 + }, + { + "epoch": 1.54, + "grad_norm": 0.41796875, + "learning_rate": 2.4940914902853152e-05, + "loss": 2.3565, + "step": 36173 + }, + { + "epoch": 1.54, + "grad_norm": 0.466796875, + "learning_rate": 2.4936482690174258e-05, + "loss": 2.4336, + "step": 36174 + }, + { + "epoch": 1.54, + "grad_norm": 0.4765625, + "learning_rate": 2.4932050815253726e-05, + "loss": 2.4411, + "step": 36175 + }, + { + "epoch": 1.54, + "grad_norm": 0.4375, + "learning_rate": 2.4927619278111535e-05, + "loss": 2.1906, + "step": 36176 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.4923188078767656e-05, + "loss": 2.3645, + "step": 36177 + }, + { + "epoch": 1.54, + "grad_norm": 0.451171875, + "learning_rate": 2.4918757217241984e-05, + "loss": 2.5031, + "step": 36178 + }, + { + "epoch": 1.54, + "grad_norm": 0.4609375, + "learning_rate": 2.4914326693554447e-05, + "loss": 2.4267, + "step": 36179 + }, + { + "epoch": 1.54, + "grad_norm": 0.4609375, + "learning_rate": 2.4909896507724995e-05, + "loss": 2.4579, + "step": 36180 + }, + { + "epoch": 1.54, + "grad_norm": 0.462890625, + "learning_rate": 2.490546665977359e-05, + "loss": 2.3999, + "step": 36181 + }, + { + "epoch": 1.54, + "grad_norm": 0.435546875, + "learning_rate": 2.490103714972012e-05, + "loss": 2.3365, + "step": 36182 + }, + { + "epoch": 1.54, + "grad_norm": 0.4375, + "learning_rate": 2.4896607977584564e-05, + "loss": 2.4366, + "step": 36183 + }, + { + "epoch": 1.54, + "grad_norm": 0.412109375, + "learning_rate": 2.4892179143386785e-05, + "loss": 2.5, + "step": 36184 + }, + { + "epoch": 1.54, + "grad_norm": 0.435546875, + "learning_rate": 2.4887750647146767e-05, + "loss": 2.3462, + "step": 36185 + }, + { + "epoch": 1.54, + "grad_norm": 0.447265625, + "learning_rate": 2.4883322488884442e-05, + "loss": 2.3946, + "step": 36186 + }, + { + "epoch": 1.54, + "grad_norm": 0.431640625, + "learning_rate": 2.487889466861971e-05, + "loss": 2.4419, + "step": 36187 + }, + { + "epoch": 1.54, + "grad_norm": 0.412109375, + "learning_rate": 2.4874467186372473e-05, + "loss": 2.42, + "step": 36188 + }, + { + "epoch": 1.54, + "grad_norm": 0.4921875, + "learning_rate": 2.4870040042162678e-05, + "loss": 2.5356, + "step": 36189 + }, + { + "epoch": 1.54, + "grad_norm": 0.4140625, + "learning_rate": 2.4865613236010278e-05, + "loss": 2.3709, + "step": 36190 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.486118676793513e-05, + "loss": 2.4699, + "step": 36191 + }, + { + "epoch": 1.54, + "grad_norm": 0.466796875, + "learning_rate": 2.485676063795718e-05, + "loss": 2.4037, + "step": 36192 + }, + { + "epoch": 1.54, + "grad_norm": 0.51953125, + "learning_rate": 2.4852334846096382e-05, + "loss": 2.4929, + "step": 36193 + }, + { + "epoch": 1.54, + "grad_norm": 0.51953125, + "learning_rate": 2.4847909392372605e-05, + "loss": 2.4641, + "step": 36194 + }, + { + "epoch": 1.54, + "grad_norm": 0.498046875, + "learning_rate": 2.484348427680574e-05, + "loss": 2.4123, + "step": 36195 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.483905949941574e-05, + "loss": 2.4986, + "step": 36196 + }, + { + "epoch": 1.54, + "grad_norm": 0.4765625, + "learning_rate": 2.483463506022252e-05, + "loss": 2.4988, + "step": 36197 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.4830210959245947e-05, + "loss": 2.4076, + "step": 36198 + }, + { + "epoch": 1.54, + "grad_norm": 0.42578125, + "learning_rate": 2.482578719650599e-05, + "loss": 2.4251, + "step": 36199 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.4821363772022476e-05, + "loss": 2.3712, + "step": 36200 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.481694068581536e-05, + "loss": 2.6007, + "step": 36201 + }, + { + "epoch": 1.54, + "grad_norm": 0.474609375, + "learning_rate": 2.4812517937904566e-05, + "loss": 2.4161, + "step": 36202 + }, + { + "epoch": 1.54, + "grad_norm": 0.396484375, + "learning_rate": 2.480809552830996e-05, + "loss": 2.4262, + "step": 36203 + }, + { + "epoch": 1.54, + "grad_norm": 0.443359375, + "learning_rate": 2.4803673457051413e-05, + "loss": 2.4555, + "step": 36204 + }, + { + "epoch": 1.54, + "grad_norm": 0.439453125, + "learning_rate": 2.4799251724148854e-05, + "loss": 2.6259, + "step": 36205 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.4794830329622208e-05, + "loss": 2.52, + "step": 36206 + }, + { + "epoch": 1.54, + "grad_norm": 0.474609375, + "learning_rate": 2.4790409273491323e-05, + "loss": 2.4491, + "step": 36207 + }, + { + "epoch": 1.54, + "grad_norm": 0.4609375, + "learning_rate": 2.4785988555776098e-05, + "loss": 2.521, + "step": 36208 + }, + { + "epoch": 1.54, + "grad_norm": 0.4453125, + "learning_rate": 2.478156817649647e-05, + "loss": 2.3328, + "step": 36209 + }, + { + "epoch": 1.54, + "grad_norm": 0.50390625, + "learning_rate": 2.477714813567229e-05, + "loss": 2.4166, + "step": 36210 + }, + { + "epoch": 1.54, + "grad_norm": 0.546875, + "learning_rate": 2.4772728433323432e-05, + "loss": 2.602, + "step": 36211 + }, + { + "epoch": 1.54, + "grad_norm": 0.421875, + "learning_rate": 2.476830906946981e-05, + "loss": 2.3691, + "step": 36212 + }, + { + "epoch": 1.54, + "grad_norm": 0.4453125, + "learning_rate": 2.4763890044131322e-05, + "loss": 2.4391, + "step": 36213 + }, + { + "epoch": 1.54, + "grad_norm": 0.51953125, + "learning_rate": 2.47594713573278e-05, + "loss": 2.3893, + "step": 36214 + }, + { + "epoch": 1.54, + "grad_norm": 0.431640625, + "learning_rate": 2.4755053009079198e-05, + "loss": 2.389, + "step": 36215 + }, + { + "epoch": 1.54, + "grad_norm": 0.51171875, + "learning_rate": 2.475063499940532e-05, + "loss": 2.45, + "step": 36216 + }, + { + "epoch": 1.54, + "grad_norm": 0.478515625, + "learning_rate": 2.4746217328326095e-05, + "loss": 2.3468, + "step": 36217 + }, + { + "epoch": 1.54, + "grad_norm": 0.423828125, + "learning_rate": 2.4741799995861415e-05, + "loss": 2.358, + "step": 36218 + }, + { + "epoch": 1.54, + "grad_norm": 0.51171875, + "learning_rate": 2.4737383002031122e-05, + "loss": 2.5857, + "step": 36219 + }, + { + "epoch": 1.54, + "grad_norm": 0.47265625, + "learning_rate": 2.473296634685508e-05, + "loss": 2.5065, + "step": 36220 + }, + { + "epoch": 1.54, + "grad_norm": 0.44140625, + "learning_rate": 2.472855003035318e-05, + "loss": 2.4742, + "step": 36221 + }, + { + "epoch": 1.54, + "grad_norm": 0.455078125, + "learning_rate": 2.4724134052545332e-05, + "loss": 2.5327, + "step": 36222 + }, + { + "epoch": 1.54, + "grad_norm": 0.4375, + "learning_rate": 2.4719718413451333e-05, + "loss": 2.3835, + "step": 36223 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.4715303113091094e-05, + "loss": 2.5715, + "step": 36224 + }, + { + "epoch": 1.54, + "grad_norm": 0.447265625, + "learning_rate": 2.4710888151484502e-05, + "loss": 2.48, + "step": 36225 + }, + { + "epoch": 1.54, + "grad_norm": 0.466796875, + "learning_rate": 2.4706473528651407e-05, + "loss": 2.3643, + "step": 36226 + }, + { + "epoch": 1.54, + "grad_norm": 0.4375, + "learning_rate": 2.470205924461163e-05, + "loss": 2.4065, + "step": 36227 + }, + { + "epoch": 1.54, + "grad_norm": 0.451171875, + "learning_rate": 2.4697645299385064e-05, + "loss": 2.3477, + "step": 36228 + }, + { + "epoch": 1.54, + "grad_norm": 0.49609375, + "learning_rate": 2.4693231692991604e-05, + "loss": 2.5028, + "step": 36229 + }, + { + "epoch": 1.54, + "grad_norm": 0.47265625, + "learning_rate": 2.4688818425451055e-05, + "loss": 2.3702, + "step": 36230 + }, + { + "epoch": 1.54, + "grad_norm": 0.474609375, + "learning_rate": 2.4684405496783323e-05, + "loss": 2.4408, + "step": 36231 + }, + { + "epoch": 1.54, + "grad_norm": 0.48046875, + "learning_rate": 2.4679992907008222e-05, + "loss": 2.3815, + "step": 36232 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.4675580656145626e-05, + "loss": 2.5147, + "step": 36233 + }, + { + "epoch": 1.54, + "grad_norm": 0.5, + "learning_rate": 2.467116874421541e-05, + "loss": 2.4636, + "step": 36234 + }, + { + "epoch": 1.54, + "grad_norm": 0.458984375, + "learning_rate": 2.466675717123741e-05, + "loss": 2.4004, + "step": 36235 + }, + { + "epoch": 1.54, + "grad_norm": 0.498046875, + "learning_rate": 2.4662345937231445e-05, + "loss": 2.3875, + "step": 36236 + }, + { + "epoch": 1.54, + "grad_norm": 0.5078125, + "learning_rate": 2.4657935042217385e-05, + "loss": 2.566, + "step": 36237 + }, + { + "epoch": 1.54, + "grad_norm": 0.455078125, + "learning_rate": 2.465352448621512e-05, + "loss": 2.4989, + "step": 36238 + }, + { + "epoch": 1.54, + "grad_norm": 0.47265625, + "learning_rate": 2.464911426924442e-05, + "loss": 2.5044, + "step": 36239 + }, + { + "epoch": 1.54, + "grad_norm": 0.4609375, + "learning_rate": 2.464470439132518e-05, + "loss": 2.4456, + "step": 36240 + }, + { + "epoch": 1.54, + "grad_norm": 0.447265625, + "learning_rate": 2.4640294852477254e-05, + "loss": 2.3596, + "step": 36241 + }, + { + "epoch": 1.54, + "grad_norm": 0.470703125, + "learning_rate": 2.463588565272046e-05, + "loss": 2.3711, + "step": 36242 + }, + { + "epoch": 1.54, + "grad_norm": 0.3984375, + "learning_rate": 2.4631476792074604e-05, + "loss": 2.4344, + "step": 36243 + }, + { + "epoch": 1.54, + "grad_norm": 0.466796875, + "learning_rate": 2.462706827055956e-05, + "loss": 2.3717, + "step": 36244 + }, + { + "epoch": 1.54, + "grad_norm": 0.4453125, + "learning_rate": 2.4622660088195193e-05, + "loss": 2.3246, + "step": 36245 + }, + { + "epoch": 1.54, + "grad_norm": 0.462890625, + "learning_rate": 2.4618252245001285e-05, + "loss": 2.3539, + "step": 36246 + }, + { + "epoch": 1.54, + "grad_norm": 0.443359375, + "learning_rate": 2.4613844740997716e-05, + "loss": 2.4391, + "step": 36247 + }, + { + "epoch": 1.54, + "grad_norm": 0.4375, + "learning_rate": 2.4609437576204266e-05, + "loss": 2.4611, + "step": 36248 + }, + { + "epoch": 1.54, + "grad_norm": 0.498046875, + "learning_rate": 2.4605030750640802e-05, + "loss": 2.4632, + "step": 36249 + }, + { + "epoch": 1.54, + "grad_norm": 0.482421875, + "learning_rate": 2.4600624264327167e-05, + "loss": 2.4591, + "step": 36250 + }, + { + "epoch": 1.54, + "grad_norm": 0.5, + "learning_rate": 2.4596218117283176e-05, + "loss": 2.5559, + "step": 36251 + }, + { + "epoch": 1.54, + "grad_norm": 0.447265625, + "learning_rate": 2.4591812309528607e-05, + "loss": 2.4819, + "step": 36252 + }, + { + "epoch": 1.54, + "grad_norm": 0.5, + "learning_rate": 2.4587406841083328e-05, + "loss": 2.3832, + "step": 36253 + }, + { + "epoch": 1.54, + "grad_norm": 0.5, + "learning_rate": 2.4583001711967192e-05, + "loss": 2.4438, + "step": 36254 + }, + { + "epoch": 1.54, + "grad_norm": 0.494140625, + "learning_rate": 2.4578596922199946e-05, + "loss": 2.4734, + "step": 36255 + }, + { + "epoch": 1.54, + "grad_norm": 0.53125, + "learning_rate": 2.457419247180146e-05, + "loss": 2.4132, + "step": 36256 + }, + { + "epoch": 1.55, + "grad_norm": 0.447265625, + "learning_rate": 2.4569788360791567e-05, + "loss": 2.5567, + "step": 36257 + }, + { + "epoch": 1.55, + "grad_norm": 0.482421875, + "learning_rate": 2.4565384589190067e-05, + "loss": 2.355, + "step": 36258 + }, + { + "epoch": 1.55, + "grad_norm": 0.470703125, + "learning_rate": 2.456098115701674e-05, + "loss": 2.483, + "step": 36259 + }, + { + "epoch": 1.55, + "grad_norm": 0.48046875, + "learning_rate": 2.4556578064291445e-05, + "loss": 2.4872, + "step": 36260 + }, + { + "epoch": 1.55, + "grad_norm": 0.494140625, + "learning_rate": 2.4552175311033954e-05, + "loss": 2.5142, + "step": 36261 + }, + { + "epoch": 1.55, + "grad_norm": 0.55078125, + "learning_rate": 2.454777289726411e-05, + "loss": 2.5707, + "step": 36262 + }, + { + "epoch": 1.55, + "grad_norm": 0.44921875, + "learning_rate": 2.4543370823001732e-05, + "loss": 2.4603, + "step": 36263 + }, + { + "epoch": 1.55, + "grad_norm": 0.44921875, + "learning_rate": 2.4538969088266615e-05, + "loss": 2.418, + "step": 36264 + }, + { + "epoch": 1.55, + "grad_norm": 0.404296875, + "learning_rate": 2.4534567693078492e-05, + "loss": 2.4549, + "step": 36265 + }, + { + "epoch": 1.55, + "grad_norm": 0.482421875, + "learning_rate": 2.4530166637457296e-05, + "loss": 2.5444, + "step": 36266 + }, + { + "epoch": 1.55, + "grad_norm": 0.451171875, + "learning_rate": 2.452576592142277e-05, + "loss": 2.4873, + "step": 36267 + }, + { + "epoch": 1.55, + "grad_norm": 0.4921875, + "learning_rate": 2.452136554499469e-05, + "loss": 2.516, + "step": 36268 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.451696550819288e-05, + "loss": 2.3729, + "step": 36269 + }, + { + "epoch": 1.55, + "grad_norm": 0.53125, + "learning_rate": 2.4512565811037168e-05, + "loss": 2.4352, + "step": 36270 + }, + { + "epoch": 1.55, + "grad_norm": 0.42578125, + "learning_rate": 2.45081664535473e-05, + "loss": 2.4639, + "step": 36271 + }, + { + "epoch": 1.55, + "grad_norm": 0.462890625, + "learning_rate": 2.4503767435743107e-05, + "loss": 2.447, + "step": 36272 + }, + { + "epoch": 1.55, + "grad_norm": 0.4140625, + "learning_rate": 2.4499368757644348e-05, + "loss": 2.4518, + "step": 36273 + }, + { + "epoch": 1.55, + "grad_norm": 0.53125, + "learning_rate": 2.449497041927087e-05, + "loss": 2.5313, + "step": 36274 + }, + { + "epoch": 1.55, + "grad_norm": 0.44921875, + "learning_rate": 2.44905724206424e-05, + "loss": 2.4661, + "step": 36275 + }, + { + "epoch": 1.55, + "grad_norm": 0.51171875, + "learning_rate": 2.448617476177879e-05, + "loss": 2.4481, + "step": 36276 + }, + { + "epoch": 1.55, + "grad_norm": 0.45703125, + "learning_rate": 2.448177744269976e-05, + "loss": 2.5008, + "step": 36277 + }, + { + "epoch": 1.55, + "grad_norm": 0.51171875, + "learning_rate": 2.4477380463425138e-05, + "loss": 2.4888, + "step": 36278 + }, + { + "epoch": 1.55, + "grad_norm": 0.451171875, + "learning_rate": 2.4472983823974726e-05, + "loss": 2.403, + "step": 36279 + }, + { + "epoch": 1.55, + "grad_norm": 0.435546875, + "learning_rate": 2.446858752436829e-05, + "loss": 2.46, + "step": 36280 + }, + { + "epoch": 1.55, + "grad_norm": 0.51171875, + "learning_rate": 2.446419156462555e-05, + "loss": 2.3994, + "step": 36281 + }, + { + "epoch": 1.55, + "grad_norm": 0.392578125, + "learning_rate": 2.4459795944766406e-05, + "loss": 2.3393, + "step": 36282 + }, + { + "epoch": 1.55, + "grad_norm": 0.455078125, + "learning_rate": 2.445540066481057e-05, + "loss": 2.4556, + "step": 36283 + }, + { + "epoch": 1.55, + "grad_norm": 0.466796875, + "learning_rate": 2.4451005724777786e-05, + "loss": 2.439, + "step": 36284 + }, + { + "epoch": 1.55, + "grad_norm": 0.4765625, + "learning_rate": 2.444661112468787e-05, + "loss": 2.4451, + "step": 36285 + }, + { + "epoch": 1.55, + "grad_norm": 0.478515625, + "learning_rate": 2.444221686456062e-05, + "loss": 2.4557, + "step": 36286 + }, + { + "epoch": 1.55, + "grad_norm": 0.4140625, + "learning_rate": 2.4437822944415755e-05, + "loss": 2.513, + "step": 36287 + }, + { + "epoch": 1.55, + "grad_norm": 0.474609375, + "learning_rate": 2.4433429364273097e-05, + "loss": 2.483, + "step": 36288 + }, + { + "epoch": 1.55, + "grad_norm": 0.443359375, + "learning_rate": 2.442903612415237e-05, + "loss": 2.4473, + "step": 36289 + }, + { + "epoch": 1.55, + "grad_norm": 0.4140625, + "learning_rate": 2.4424643224073385e-05, + "loss": 2.4413, + "step": 36290 + }, + { + "epoch": 1.55, + "grad_norm": 0.5, + "learning_rate": 2.4420250664055854e-05, + "loss": 2.5014, + "step": 36291 + }, + { + "epoch": 1.55, + "grad_norm": 0.416015625, + "learning_rate": 2.441585844411961e-05, + "loss": 2.5147, + "step": 36292 + }, + { + "epoch": 1.55, + "grad_norm": 0.447265625, + "learning_rate": 2.4411466564284346e-05, + "loss": 2.4565, + "step": 36293 + }, + { + "epoch": 1.55, + "grad_norm": 0.478515625, + "learning_rate": 2.4407075024569858e-05, + "loss": 2.4781, + "step": 36294 + }, + { + "epoch": 1.55, + "grad_norm": 0.435546875, + "learning_rate": 2.440268382499594e-05, + "loss": 2.4943, + "step": 36295 + }, + { + "epoch": 1.55, + "grad_norm": 0.53125, + "learning_rate": 2.4398292965582314e-05, + "loss": 2.5459, + "step": 36296 + }, + { + "epoch": 1.55, + "grad_norm": 0.48828125, + "learning_rate": 2.4393902446348683e-05, + "loss": 2.525, + "step": 36297 + }, + { + "epoch": 1.55, + "grad_norm": 0.4609375, + "learning_rate": 2.4389512267314918e-05, + "loss": 2.5448, + "step": 36298 + }, + { + "epoch": 1.55, + "grad_norm": 0.5078125, + "learning_rate": 2.438512242850072e-05, + "loss": 2.4076, + "step": 36299 + }, + { + "epoch": 1.55, + "grad_norm": 0.484375, + "learning_rate": 2.4380732929925797e-05, + "loss": 2.4856, + "step": 36300 + }, + { + "epoch": 1.55, + "grad_norm": 0.443359375, + "learning_rate": 2.4376343771609945e-05, + "loss": 2.3936, + "step": 36301 + }, + { + "epoch": 1.55, + "grad_norm": 0.455078125, + "learning_rate": 2.437195495357294e-05, + "loss": 2.5556, + "step": 36302 + }, + { + "epoch": 1.55, + "grad_norm": 0.484375, + "learning_rate": 2.436756647583447e-05, + "loss": 2.5103, + "step": 36303 + }, + { + "epoch": 1.55, + "grad_norm": 0.51953125, + "learning_rate": 2.436317833841434e-05, + "loss": 2.3432, + "step": 36304 + }, + { + "epoch": 1.55, + "grad_norm": 0.5078125, + "learning_rate": 2.435879054133223e-05, + "loss": 2.4898, + "step": 36305 + }, + { + "epoch": 1.55, + "grad_norm": 0.470703125, + "learning_rate": 2.4354403084607956e-05, + "loss": 2.4752, + "step": 36306 + }, + { + "epoch": 1.55, + "grad_norm": 0.4296875, + "learning_rate": 2.4350015968261187e-05, + "loss": 2.3522, + "step": 36307 + }, + { + "epoch": 1.55, + "grad_norm": 0.46875, + "learning_rate": 2.4345629192311735e-05, + "loss": 2.4642, + "step": 36308 + }, + { + "epoch": 1.55, + "grad_norm": 0.498046875, + "learning_rate": 2.434124275677927e-05, + "loss": 2.5029, + "step": 36309 + }, + { + "epoch": 1.55, + "grad_norm": 0.44140625, + "learning_rate": 2.433685666168356e-05, + "loss": 2.472, + "step": 36310 + }, + { + "epoch": 1.55, + "grad_norm": 0.431640625, + "learning_rate": 2.433247090704438e-05, + "loss": 2.4003, + "step": 36311 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.4328085492881425e-05, + "loss": 2.5429, + "step": 36312 + }, + { + "epoch": 1.55, + "grad_norm": 0.5078125, + "learning_rate": 2.432370041921437e-05, + "loss": 2.6193, + "step": 36313 + }, + { + "epoch": 1.55, + "grad_norm": 0.453125, + "learning_rate": 2.431931568606306e-05, + "loss": 2.4029, + "step": 36314 + }, + { + "epoch": 1.55, + "grad_norm": 0.4375, + "learning_rate": 2.4314931293447173e-05, + "loss": 2.5042, + "step": 36315 + }, + { + "epoch": 1.55, + "grad_norm": 0.48828125, + "learning_rate": 2.4310547241386415e-05, + "loss": 2.433, + "step": 36316 + }, + { + "epoch": 1.55, + "grad_norm": 0.43359375, + "learning_rate": 2.430616352990053e-05, + "loss": 2.3678, + "step": 36317 + }, + { + "epoch": 1.55, + "grad_norm": 0.431640625, + "learning_rate": 2.4301780159009267e-05, + "loss": 2.4055, + "step": 36318 + }, + { + "epoch": 1.55, + "grad_norm": 0.53125, + "learning_rate": 2.4297397128732312e-05, + "loss": 2.4312, + "step": 36319 + }, + { + "epoch": 1.55, + "grad_norm": 0.4375, + "learning_rate": 2.4293014439089423e-05, + "loss": 2.4488, + "step": 36320 + }, + { + "epoch": 1.55, + "grad_norm": 0.462890625, + "learning_rate": 2.428863209010027e-05, + "loss": 2.5026, + "step": 36321 + }, + { + "epoch": 1.55, + "grad_norm": 0.443359375, + "learning_rate": 2.4284250081784643e-05, + "loss": 2.3882, + "step": 36322 + }, + { + "epoch": 1.55, + "grad_norm": 0.4140625, + "learning_rate": 2.4279868414162188e-05, + "loss": 2.4485, + "step": 36323 + }, + { + "epoch": 1.55, + "grad_norm": 0.53125, + "learning_rate": 2.427548708725268e-05, + "loss": 2.407, + "step": 36324 + }, + { + "epoch": 1.55, + "grad_norm": 0.451171875, + "learning_rate": 2.427110610107578e-05, + "loss": 2.2786, + "step": 36325 + }, + { + "epoch": 1.55, + "grad_norm": 0.474609375, + "learning_rate": 2.4266725455651217e-05, + "loss": 2.4618, + "step": 36326 + }, + { + "epoch": 1.55, + "grad_norm": 0.48828125, + "learning_rate": 2.4262345150998744e-05, + "loss": 2.3898, + "step": 36327 + }, + { + "epoch": 1.55, + "grad_norm": 0.486328125, + "learning_rate": 2.4257965187138043e-05, + "loss": 2.5042, + "step": 36328 + }, + { + "epoch": 1.55, + "grad_norm": 0.51171875, + "learning_rate": 2.4253585564088764e-05, + "loss": 2.3799, + "step": 36329 + }, + { + "epoch": 1.55, + "grad_norm": 0.44921875, + "learning_rate": 2.4249206281870716e-05, + "loss": 2.4322, + "step": 36330 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.424482734050356e-05, + "loss": 2.453, + "step": 36331 + }, + { + "epoch": 1.55, + "grad_norm": 0.466796875, + "learning_rate": 2.424044874000697e-05, + "loss": 2.5495, + "step": 36332 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.4236070480400673e-05, + "loss": 2.2951, + "step": 36333 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.42316925617044e-05, + "loss": 2.3886, + "step": 36334 + }, + { + "epoch": 1.55, + "grad_norm": 0.462890625, + "learning_rate": 2.4227314983937787e-05, + "loss": 2.4656, + "step": 36335 + }, + { + "epoch": 1.55, + "grad_norm": 0.47265625, + "learning_rate": 2.42229377471206e-05, + "loss": 2.5714, + "step": 36336 + }, + { + "epoch": 1.55, + "grad_norm": 0.490234375, + "learning_rate": 2.4218560851272465e-05, + "loss": 2.4276, + "step": 36337 + }, + { + "epoch": 1.55, + "grad_norm": 0.5078125, + "learning_rate": 2.4214184296413134e-05, + "loss": 2.5842, + "step": 36338 + }, + { + "epoch": 1.55, + "grad_norm": 0.490234375, + "learning_rate": 2.4209808082562258e-05, + "loss": 2.4646, + "step": 36339 + }, + { + "epoch": 1.55, + "grad_norm": 0.50390625, + "learning_rate": 2.4205432209739576e-05, + "loss": 2.467, + "step": 36340 + }, + { + "epoch": 1.55, + "grad_norm": 0.5078125, + "learning_rate": 2.4201056677964718e-05, + "loss": 2.5231, + "step": 36341 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.41966814872574e-05, + "loss": 2.4064, + "step": 36342 + }, + { + "epoch": 1.55, + "grad_norm": 0.44921875, + "learning_rate": 2.419230663763735e-05, + "loss": 2.5606, + "step": 36343 + }, + { + "epoch": 1.55, + "grad_norm": 0.486328125, + "learning_rate": 2.4187932129124213e-05, + "loss": 2.4536, + "step": 36344 + }, + { + "epoch": 1.55, + "grad_norm": 0.423828125, + "learning_rate": 2.4183557961737613e-05, + "loss": 2.4435, + "step": 36345 + }, + { + "epoch": 1.55, + "grad_norm": 0.490234375, + "learning_rate": 2.4179184135497357e-05, + "loss": 2.5305, + "step": 36346 + }, + { + "epoch": 1.55, + "grad_norm": 0.50390625, + "learning_rate": 2.4174810650423063e-05, + "loss": 2.4873, + "step": 36347 + }, + { + "epoch": 1.55, + "grad_norm": 0.4765625, + "learning_rate": 2.4170437506534372e-05, + "loss": 2.2273, + "step": 36348 + }, + { + "epoch": 1.55, + "grad_norm": 0.5234375, + "learning_rate": 2.416606470385101e-05, + "loss": 2.4152, + "step": 36349 + }, + { + "epoch": 1.55, + "grad_norm": 0.439453125, + "learning_rate": 2.4161692242392665e-05, + "loss": 2.5426, + "step": 36350 + }, + { + "epoch": 1.55, + "grad_norm": 0.52734375, + "learning_rate": 2.415732012217897e-05, + "loss": 2.5918, + "step": 36351 + }, + { + "epoch": 1.55, + "grad_norm": 0.5, + "learning_rate": 2.415294834322964e-05, + "loss": 2.5865, + "step": 36352 + }, + { + "epoch": 1.55, + "grad_norm": 0.416015625, + "learning_rate": 2.414857690556429e-05, + "loss": 2.4628, + "step": 36353 + }, + { + "epoch": 1.55, + "grad_norm": 0.486328125, + "learning_rate": 2.414420580920266e-05, + "loss": 2.2789, + "step": 36354 + }, + { + "epoch": 1.55, + "grad_norm": 0.47265625, + "learning_rate": 2.413983505416435e-05, + "loss": 2.4587, + "step": 36355 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.4135464640469087e-05, + "loss": 2.3882, + "step": 36356 + }, + { + "epoch": 1.55, + "grad_norm": 0.51953125, + "learning_rate": 2.413109456813648e-05, + "loss": 2.4312, + "step": 36357 + }, + { + "epoch": 1.55, + "grad_norm": 0.439453125, + "learning_rate": 2.412672483718622e-05, + "loss": 2.5416, + "step": 36358 + }, + { + "epoch": 1.55, + "grad_norm": 0.5, + "learning_rate": 2.412235544763799e-05, + "loss": 2.3901, + "step": 36359 + }, + { + "epoch": 1.55, + "grad_norm": 0.453125, + "learning_rate": 2.4117986399511438e-05, + "loss": 2.29, + "step": 36360 + }, + { + "epoch": 1.55, + "grad_norm": 0.5234375, + "learning_rate": 2.4113617692826163e-05, + "loss": 2.5248, + "step": 36361 + }, + { + "epoch": 1.55, + "grad_norm": 0.482421875, + "learning_rate": 2.4109249327601922e-05, + "loss": 2.4624, + "step": 36362 + }, + { + "epoch": 1.55, + "grad_norm": 0.490234375, + "learning_rate": 2.410488130385834e-05, + "loss": 2.5596, + "step": 36363 + }, + { + "epoch": 1.55, + "grad_norm": 0.53515625, + "learning_rate": 2.4100513621615016e-05, + "loss": 2.4511, + "step": 36364 + }, + { + "epoch": 1.55, + "grad_norm": 0.44921875, + "learning_rate": 2.409614628089164e-05, + "loss": 2.4316, + "step": 36365 + }, + { + "epoch": 1.55, + "grad_norm": 0.5078125, + "learning_rate": 2.4091779281707893e-05, + "loss": 2.457, + "step": 36366 + }, + { + "epoch": 1.55, + "grad_norm": 0.54296875, + "learning_rate": 2.4087412624083384e-05, + "loss": 2.4876, + "step": 36367 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.4083046308037794e-05, + "loss": 2.4083, + "step": 36368 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.407868033359072e-05, + "loss": 2.4763, + "step": 36369 + }, + { + "epoch": 1.55, + "grad_norm": 0.49609375, + "learning_rate": 2.407431470076187e-05, + "loss": 2.4259, + "step": 36370 + }, + { + "epoch": 1.55, + "grad_norm": 0.46875, + "learning_rate": 2.4069949409570824e-05, + "loss": 2.4039, + "step": 36371 + }, + { + "epoch": 1.55, + "grad_norm": 0.53515625, + "learning_rate": 2.4065584460037294e-05, + "loss": 2.4889, + "step": 36372 + }, + { + "epoch": 1.55, + "grad_norm": 0.443359375, + "learning_rate": 2.406121985218085e-05, + "loss": 2.6169, + "step": 36373 + }, + { + "epoch": 1.55, + "grad_norm": 0.4375, + "learning_rate": 2.4056855586021154e-05, + "loss": 2.632, + "step": 36374 + }, + { + "epoch": 1.55, + "grad_norm": 0.52734375, + "learning_rate": 2.40524916615779e-05, + "loss": 2.5054, + "step": 36375 + }, + { + "epoch": 1.55, + "grad_norm": 0.482421875, + "learning_rate": 2.4048128078870667e-05, + "loss": 2.4414, + "step": 36376 + }, + { + "epoch": 1.55, + "grad_norm": 0.447265625, + "learning_rate": 2.404376483791905e-05, + "loss": 2.4682, + "step": 36377 + }, + { + "epoch": 1.55, + "grad_norm": 0.51953125, + "learning_rate": 2.4039401938742778e-05, + "loss": 2.412, + "step": 36378 + }, + { + "epoch": 1.55, + "grad_norm": 0.498046875, + "learning_rate": 2.4035039381361446e-05, + "loss": 2.375, + "step": 36379 + }, + { + "epoch": 1.55, + "grad_norm": 0.447265625, + "learning_rate": 2.4030677165794645e-05, + "loss": 2.5142, + "step": 36380 + }, + { + "epoch": 1.55, + "grad_norm": 0.50390625, + "learning_rate": 2.402631529206206e-05, + "loss": 2.505, + "step": 36381 + }, + { + "epoch": 1.55, + "grad_norm": 0.455078125, + "learning_rate": 2.402195376018327e-05, + "loss": 2.4133, + "step": 36382 + }, + { + "epoch": 1.55, + "grad_norm": 0.50390625, + "learning_rate": 2.4017592570177904e-05, + "loss": 2.3823, + "step": 36383 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.4013231722065654e-05, + "loss": 2.321, + "step": 36384 + }, + { + "epoch": 1.55, + "grad_norm": 0.400390625, + "learning_rate": 2.400887121586608e-05, + "loss": 2.2965, + "step": 36385 + }, + { + "epoch": 1.55, + "grad_norm": 0.490234375, + "learning_rate": 2.4004511051598754e-05, + "loss": 2.4713, + "step": 36386 + }, + { + "epoch": 1.55, + "grad_norm": 0.498046875, + "learning_rate": 2.4000151229283406e-05, + "loss": 2.5718, + "step": 36387 + }, + { + "epoch": 1.55, + "grad_norm": 0.49609375, + "learning_rate": 2.3995791748939624e-05, + "loss": 2.3899, + "step": 36388 + }, + { + "epoch": 1.55, + "grad_norm": 0.53125, + "learning_rate": 2.399143261058695e-05, + "loss": 2.5622, + "step": 36389 + }, + { + "epoch": 1.55, + "grad_norm": 0.4609375, + "learning_rate": 2.3987073814245063e-05, + "loss": 2.4023, + "step": 36390 + }, + { + "epoch": 1.55, + "grad_norm": 0.48046875, + "learning_rate": 2.3982715359933594e-05, + "loss": 2.5238, + "step": 36391 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.3978357247672122e-05, + "loss": 2.483, + "step": 36392 + }, + { + "epoch": 1.55, + "grad_norm": 0.44140625, + "learning_rate": 2.3973999477480236e-05, + "loss": 2.4805, + "step": 36393 + }, + { + "epoch": 1.55, + "grad_norm": 0.486328125, + "learning_rate": 2.3969642049377572e-05, + "loss": 2.3424, + "step": 36394 + }, + { + "epoch": 1.55, + "grad_norm": 0.484375, + "learning_rate": 2.396528496338375e-05, + "loss": 2.3572, + "step": 36395 + }, + { + "epoch": 1.55, + "grad_norm": 0.4765625, + "learning_rate": 2.3960928219518332e-05, + "loss": 2.476, + "step": 36396 + }, + { + "epoch": 1.55, + "grad_norm": 0.470703125, + "learning_rate": 2.3956571817800977e-05, + "loss": 2.4213, + "step": 36397 + }, + { + "epoch": 1.55, + "grad_norm": 0.4765625, + "learning_rate": 2.3952215758251228e-05, + "loss": 2.5061, + "step": 36398 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.3947860040888725e-05, + "loss": 2.3613, + "step": 36399 + }, + { + "epoch": 1.55, + "grad_norm": 0.45703125, + "learning_rate": 2.3943504665733073e-05, + "loss": 2.4994, + "step": 36400 + }, + { + "epoch": 1.55, + "grad_norm": 0.45703125, + "learning_rate": 2.393914963280387e-05, + "loss": 2.5829, + "step": 36401 + }, + { + "epoch": 1.55, + "grad_norm": 0.482421875, + "learning_rate": 2.3934794942120632e-05, + "loss": 2.4792, + "step": 36402 + }, + { + "epoch": 1.55, + "grad_norm": 0.48046875, + "learning_rate": 2.3930440593703075e-05, + "loss": 2.4532, + "step": 36403 + }, + { + "epoch": 1.55, + "grad_norm": 0.486328125, + "learning_rate": 2.392608658757073e-05, + "loss": 2.3339, + "step": 36404 + }, + { + "epoch": 1.55, + "grad_norm": 0.53515625, + "learning_rate": 2.3921732923743167e-05, + "loss": 2.2937, + "step": 36405 + }, + { + "epoch": 1.55, + "grad_norm": 0.484375, + "learning_rate": 2.391737960224001e-05, + "loss": 2.5671, + "step": 36406 + }, + { + "epoch": 1.55, + "grad_norm": 0.421875, + "learning_rate": 2.391302662308086e-05, + "loss": 2.2571, + "step": 36407 + }, + { + "epoch": 1.55, + "grad_norm": 0.5546875, + "learning_rate": 2.390867398628529e-05, + "loss": 2.5422, + "step": 36408 + }, + { + "epoch": 1.55, + "grad_norm": 0.482421875, + "learning_rate": 2.3904321691872844e-05, + "loss": 2.389, + "step": 36409 + }, + { + "epoch": 1.55, + "grad_norm": 0.443359375, + "learning_rate": 2.3899969739863137e-05, + "loss": 2.4723, + "step": 36410 + }, + { + "epoch": 1.55, + "grad_norm": 0.48046875, + "learning_rate": 2.38956181302758e-05, + "loss": 2.5192, + "step": 36411 + }, + { + "epoch": 1.55, + "grad_norm": 0.5078125, + "learning_rate": 2.3891266863130324e-05, + "loss": 2.3901, + "step": 36412 + }, + { + "epoch": 1.55, + "grad_norm": 0.51171875, + "learning_rate": 2.3886915938446363e-05, + "loss": 2.5021, + "step": 36413 + }, + { + "epoch": 1.55, + "grad_norm": 0.451171875, + "learning_rate": 2.3882565356243448e-05, + "loss": 2.5308, + "step": 36414 + }, + { + "epoch": 1.55, + "grad_norm": 0.43359375, + "learning_rate": 2.3878215116541158e-05, + "loss": 2.3936, + "step": 36415 + }, + { + "epoch": 1.55, + "grad_norm": 0.453125, + "learning_rate": 2.3873865219359114e-05, + "loss": 2.5115, + "step": 36416 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.3869515664716858e-05, + "loss": 2.4495, + "step": 36417 + }, + { + "epoch": 1.55, + "grad_norm": 0.447265625, + "learning_rate": 2.3865166452633902e-05, + "loss": 2.4424, + "step": 36418 + }, + { + "epoch": 1.55, + "grad_norm": 0.43359375, + "learning_rate": 2.3860817583129926e-05, + "loss": 2.3922, + "step": 36419 + }, + { + "epoch": 1.55, + "grad_norm": 0.474609375, + "learning_rate": 2.385646905622444e-05, + "loss": 2.3884, + "step": 36420 + }, + { + "epoch": 1.55, + "grad_norm": 0.455078125, + "learning_rate": 2.3852120871937e-05, + "loss": 2.4447, + "step": 36421 + }, + { + "epoch": 1.55, + "grad_norm": 0.4375, + "learning_rate": 2.384777303028718e-05, + "loss": 2.3542, + "step": 36422 + }, + { + "epoch": 1.55, + "grad_norm": 0.51171875, + "learning_rate": 2.3843425531294583e-05, + "loss": 2.393, + "step": 36423 + }, + { + "epoch": 1.55, + "grad_norm": 0.4375, + "learning_rate": 2.3839078374978728e-05, + "loss": 2.4406, + "step": 36424 + }, + { + "epoch": 1.55, + "grad_norm": 0.46875, + "learning_rate": 2.383473156135917e-05, + "loss": 2.4335, + "step": 36425 + }, + { + "epoch": 1.55, + "grad_norm": 0.443359375, + "learning_rate": 2.3830385090455486e-05, + "loss": 2.5781, + "step": 36426 + }, + { + "epoch": 1.55, + "grad_norm": 0.435546875, + "learning_rate": 2.3826038962287256e-05, + "loss": 2.4853, + "step": 36427 + }, + { + "epoch": 1.55, + "grad_norm": 0.41796875, + "learning_rate": 2.382169317687398e-05, + "loss": 2.3705, + "step": 36428 + }, + { + "epoch": 1.55, + "grad_norm": 0.48046875, + "learning_rate": 2.3817347734235272e-05, + "loss": 2.5334, + "step": 36429 + }, + { + "epoch": 1.55, + "grad_norm": 0.4140625, + "learning_rate": 2.3813002634390634e-05, + "loss": 2.3208, + "step": 36430 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.380865787735965e-05, + "loss": 2.4519, + "step": 36431 + }, + { + "epoch": 1.55, + "grad_norm": 0.5234375, + "learning_rate": 2.3804313463161875e-05, + "loss": 2.544, + "step": 36432 + }, + { + "epoch": 1.55, + "grad_norm": 0.482421875, + "learning_rate": 2.3799969391816856e-05, + "loss": 2.2988, + "step": 36433 + }, + { + "epoch": 1.55, + "grad_norm": 0.443359375, + "learning_rate": 2.3795625663344056e-05, + "loss": 2.4812, + "step": 36434 + }, + { + "epoch": 1.55, + "grad_norm": 0.45703125, + "learning_rate": 2.3791282277763162e-05, + "loss": 2.4068, + "step": 36435 + }, + { + "epoch": 1.55, + "grad_norm": 0.494140625, + "learning_rate": 2.378693923509363e-05, + "loss": 2.4402, + "step": 36436 + }, + { + "epoch": 1.55, + "grad_norm": 0.49609375, + "learning_rate": 2.3782596535355005e-05, + "loss": 2.5994, + "step": 36437 + }, + { + "epoch": 1.55, + "grad_norm": 0.435546875, + "learning_rate": 2.3778254178566837e-05, + "loss": 2.4339, + "step": 36438 + }, + { + "epoch": 1.55, + "grad_norm": 0.5078125, + "learning_rate": 2.377391216474869e-05, + "loss": 2.5185, + "step": 36439 + }, + { + "epoch": 1.55, + "grad_norm": 0.455078125, + "learning_rate": 2.3769570493920068e-05, + "loss": 2.4067, + "step": 36440 + }, + { + "epoch": 1.55, + "grad_norm": 0.455078125, + "learning_rate": 2.3765229166100532e-05, + "loss": 2.4806, + "step": 36441 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.3760888181309583e-05, + "loss": 2.5002, + "step": 36442 + }, + { + "epoch": 1.55, + "grad_norm": 0.4609375, + "learning_rate": 2.3756547539566808e-05, + "loss": 2.4699, + "step": 36443 + }, + { + "epoch": 1.55, + "grad_norm": 0.4296875, + "learning_rate": 2.3752207240891666e-05, + "loss": 2.4448, + "step": 36444 + }, + { + "epoch": 1.55, + "grad_norm": 0.470703125, + "learning_rate": 2.374786728530376e-05, + "loss": 2.5196, + "step": 36445 + }, + { + "epoch": 1.55, + "grad_norm": 0.515625, + "learning_rate": 2.3743527672822564e-05, + "loss": 2.4535, + "step": 36446 + }, + { + "epoch": 1.55, + "grad_norm": 0.44140625, + "learning_rate": 2.3739188403467616e-05, + "loss": 2.5807, + "step": 36447 + }, + { + "epoch": 1.55, + "grad_norm": 0.4765625, + "learning_rate": 2.373484947725848e-05, + "loss": 2.478, + "step": 36448 + }, + { + "epoch": 1.55, + "grad_norm": 0.455078125, + "learning_rate": 2.3730510894214654e-05, + "loss": 2.5234, + "step": 36449 + }, + { + "epoch": 1.55, + "grad_norm": 0.451171875, + "learning_rate": 2.3726172654355584e-05, + "loss": 2.3864, + "step": 36450 + }, + { + "epoch": 1.55, + "grad_norm": 0.447265625, + "learning_rate": 2.3721834757700933e-05, + "loss": 2.371, + "step": 36451 + }, + { + "epoch": 1.55, + "grad_norm": 0.48828125, + "learning_rate": 2.3717497204270143e-05, + "loss": 2.4663, + "step": 36452 + }, + { + "epoch": 1.55, + "grad_norm": 0.4140625, + "learning_rate": 2.371315999408271e-05, + "loss": 2.4233, + "step": 36453 + }, + { + "epoch": 1.55, + "grad_norm": 0.423828125, + "learning_rate": 2.3708823127158187e-05, + "loss": 2.4036, + "step": 36454 + }, + { + "epoch": 1.55, + "grad_norm": 0.46875, + "learning_rate": 2.3704486603516096e-05, + "loss": 2.5921, + "step": 36455 + }, + { + "epoch": 1.55, + "grad_norm": 0.4765625, + "learning_rate": 2.3700150423175906e-05, + "loss": 2.5078, + "step": 36456 + }, + { + "epoch": 1.55, + "grad_norm": 0.470703125, + "learning_rate": 2.3695814586157183e-05, + "loss": 2.4192, + "step": 36457 + }, + { + "epoch": 1.55, + "grad_norm": 0.494140625, + "learning_rate": 2.3691479092479383e-05, + "loss": 2.3829, + "step": 36458 + }, + { + "epoch": 1.55, + "grad_norm": 0.515625, + "learning_rate": 2.368714394216206e-05, + "loss": 2.425, + "step": 36459 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.368280913522468e-05, + "loss": 2.373, + "step": 36460 + }, + { + "epoch": 1.55, + "grad_norm": 0.470703125, + "learning_rate": 2.3678474671686803e-05, + "loss": 2.5451, + "step": 36461 + }, + { + "epoch": 1.55, + "grad_norm": 0.45703125, + "learning_rate": 2.367414055156786e-05, + "loss": 2.5095, + "step": 36462 + }, + { + "epoch": 1.55, + "grad_norm": 0.416015625, + "learning_rate": 2.366980677488738e-05, + "loss": 2.508, + "step": 36463 + }, + { + "epoch": 1.55, + "grad_norm": 0.474609375, + "learning_rate": 2.3665473341664924e-05, + "loss": 2.4294, + "step": 36464 + }, + { + "epoch": 1.55, + "grad_norm": 0.451171875, + "learning_rate": 2.3661140251919923e-05, + "loss": 2.3785, + "step": 36465 + }, + { + "epoch": 1.55, + "grad_norm": 0.494140625, + "learning_rate": 2.3656807505671842e-05, + "loss": 2.4816, + "step": 36466 + }, + { + "epoch": 1.55, + "grad_norm": 0.47265625, + "learning_rate": 2.3652475102940286e-05, + "loss": 2.4669, + "step": 36467 + }, + { + "epoch": 1.55, + "grad_norm": 0.470703125, + "learning_rate": 2.3648143043744687e-05, + "loss": 2.609, + "step": 36468 + }, + { + "epoch": 1.55, + "grad_norm": 0.484375, + "learning_rate": 2.3643811328104516e-05, + "loss": 2.3187, + "step": 36469 + }, + { + "epoch": 1.55, + "grad_norm": 0.5234375, + "learning_rate": 2.363947995603929e-05, + "loss": 2.5192, + "step": 36470 + }, + { + "epoch": 1.55, + "grad_norm": 0.474609375, + "learning_rate": 2.363514892756852e-05, + "loss": 2.5182, + "step": 36471 + }, + { + "epoch": 1.55, + "grad_norm": 0.400390625, + "learning_rate": 2.363081824271165e-05, + "loss": 2.3555, + "step": 36472 + }, + { + "epoch": 1.55, + "grad_norm": 0.515625, + "learning_rate": 2.3626487901488214e-05, + "loss": 2.6499, + "step": 36473 + }, + { + "epoch": 1.55, + "grad_norm": 0.486328125, + "learning_rate": 2.362215790391764e-05, + "loss": 2.4461, + "step": 36474 + }, + { + "epoch": 1.55, + "grad_norm": 0.48828125, + "learning_rate": 2.361782825001947e-05, + "loss": 2.4516, + "step": 36475 + }, + { + "epoch": 1.55, + "grad_norm": 0.474609375, + "learning_rate": 2.361349893981314e-05, + "loss": 2.4641, + "step": 36476 + }, + { + "epoch": 1.55, + "grad_norm": 0.466796875, + "learning_rate": 2.360916997331817e-05, + "loss": 2.3226, + "step": 36477 + }, + { + "epoch": 1.55, + "grad_norm": 0.46875, + "learning_rate": 2.3604841350553984e-05, + "loss": 2.6317, + "step": 36478 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.36005130715401e-05, + "loss": 2.4023, + "step": 36479 + }, + { + "epoch": 1.55, + "grad_norm": 0.458984375, + "learning_rate": 2.3596185136296012e-05, + "loss": 2.33, + "step": 36480 + }, + { + "epoch": 1.55, + "grad_norm": 0.48828125, + "learning_rate": 2.3591857544841166e-05, + "loss": 2.4414, + "step": 36481 + }, + { + "epoch": 1.55, + "grad_norm": 0.466796875, + "learning_rate": 2.358753029719498e-05, + "loss": 2.4382, + "step": 36482 + }, + { + "epoch": 1.55, + "grad_norm": 0.453125, + "learning_rate": 2.3583203393377052e-05, + "loss": 2.3972, + "step": 36483 + }, + { + "epoch": 1.55, + "grad_norm": 0.474609375, + "learning_rate": 2.3578876833406772e-05, + "loss": 2.4907, + "step": 36484 + }, + { + "epoch": 1.55, + "grad_norm": 0.455078125, + "learning_rate": 2.357455061730358e-05, + "loss": 2.4632, + "step": 36485 + }, + { + "epoch": 1.55, + "grad_norm": 0.5234375, + "learning_rate": 2.3570224745086988e-05, + "loss": 2.4561, + "step": 36486 + }, + { + "epoch": 1.55, + "grad_norm": 0.4765625, + "learning_rate": 2.356589921677649e-05, + "loss": 2.5305, + "step": 36487 + }, + { + "epoch": 1.55, + "grad_norm": 0.4453125, + "learning_rate": 2.3561574032391465e-05, + "loss": 2.4486, + "step": 36488 + }, + { + "epoch": 1.55, + "grad_norm": 0.47265625, + "learning_rate": 2.355724919195147e-05, + "loss": 2.3632, + "step": 36489 + }, + { + "epoch": 1.55, + "grad_norm": 0.451171875, + "learning_rate": 2.3552924695475875e-05, + "loss": 2.4485, + "step": 36490 + }, + { + "epoch": 1.55, + "grad_norm": 0.48828125, + "learning_rate": 2.3548600542984213e-05, + "loss": 2.5147, + "step": 36491 + }, + { + "epoch": 1.56, + "grad_norm": 0.40234375, + "learning_rate": 2.3544276734495884e-05, + "loss": 2.4113, + "step": 36492 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.353995327003039e-05, + "loss": 2.4524, + "step": 36493 + }, + { + "epoch": 1.56, + "grad_norm": 0.423828125, + "learning_rate": 2.3535630149607134e-05, + "loss": 2.3793, + "step": 36494 + }, + { + "epoch": 1.56, + "grad_norm": 0.412109375, + "learning_rate": 2.3531307373245602e-05, + "loss": 2.4313, + "step": 36495 + }, + { + "epoch": 1.56, + "grad_norm": 0.4921875, + "learning_rate": 2.352698494096528e-05, + "loss": 2.6129, + "step": 36496 + }, + { + "epoch": 1.56, + "grad_norm": 0.447265625, + "learning_rate": 2.3522662852785558e-05, + "loss": 2.4325, + "step": 36497 + }, + { + "epoch": 1.56, + "grad_norm": 0.423828125, + "learning_rate": 2.351834110872586e-05, + "loss": 2.5508, + "step": 36498 + }, + { + "epoch": 1.56, + "grad_norm": 0.4453125, + "learning_rate": 2.3514019708805733e-05, + "loss": 2.5291, + "step": 36499 + }, + { + "epoch": 1.56, + "grad_norm": 0.494140625, + "learning_rate": 2.3509698653044554e-05, + "loss": 2.4127, + "step": 36500 + }, + { + "epoch": 1.56, + "grad_norm": 0.5, + "learning_rate": 2.350537794146176e-05, + "loss": 2.3946, + "step": 36501 + }, + { + "epoch": 1.56, + "grad_norm": 0.443359375, + "learning_rate": 2.3501057574076834e-05, + "loss": 2.4277, + "step": 36502 + }, + { + "epoch": 1.56, + "grad_norm": 0.5546875, + "learning_rate": 2.3496737550909164e-05, + "loss": 2.3695, + "step": 36503 + }, + { + "epoch": 1.56, + "grad_norm": 0.4453125, + "learning_rate": 2.3492417871978212e-05, + "loss": 2.4912, + "step": 36504 + }, + { + "epoch": 1.56, + "grad_norm": 0.455078125, + "learning_rate": 2.348809853730345e-05, + "loss": 2.4231, + "step": 36505 + }, + { + "epoch": 1.56, + "grad_norm": 0.451171875, + "learning_rate": 2.3483779546904284e-05, + "loss": 2.4669, + "step": 36506 + }, + { + "epoch": 1.56, + "grad_norm": 0.423828125, + "learning_rate": 2.347946090080012e-05, + "loss": 2.304, + "step": 36507 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.3475142599010414e-05, + "loss": 2.3686, + "step": 36508 + }, + { + "epoch": 1.56, + "grad_norm": 0.52734375, + "learning_rate": 2.3470824641554634e-05, + "loss": 2.5153, + "step": 36509 + }, + { + "epoch": 1.56, + "grad_norm": 0.427734375, + "learning_rate": 2.3466507028452144e-05, + "loss": 2.4773, + "step": 36510 + }, + { + "epoch": 1.56, + "grad_norm": 0.47265625, + "learning_rate": 2.3462189759722397e-05, + "loss": 2.5279, + "step": 36511 + }, + { + "epoch": 1.56, + "grad_norm": 0.412109375, + "learning_rate": 2.3457872835384853e-05, + "loss": 2.4358, + "step": 36512 + }, + { + "epoch": 1.56, + "grad_norm": 0.4765625, + "learning_rate": 2.3453556255458918e-05, + "loss": 2.3998, + "step": 36513 + }, + { + "epoch": 1.56, + "grad_norm": 0.51171875, + "learning_rate": 2.344924001996397e-05, + "loss": 2.3507, + "step": 36514 + }, + { + "epoch": 1.56, + "grad_norm": 0.50390625, + "learning_rate": 2.3444924128919467e-05, + "loss": 2.4937, + "step": 36515 + }, + { + "epoch": 1.56, + "grad_norm": 0.484375, + "learning_rate": 2.3440608582344847e-05, + "loss": 2.4372, + "step": 36516 + }, + { + "epoch": 1.56, + "grad_norm": 0.40234375, + "learning_rate": 2.343629338025949e-05, + "loss": 2.3632, + "step": 36517 + }, + { + "epoch": 1.56, + "grad_norm": 0.478515625, + "learning_rate": 2.343197852268286e-05, + "loss": 2.3601, + "step": 36518 + }, + { + "epoch": 1.56, + "grad_norm": 0.4375, + "learning_rate": 2.342766400963432e-05, + "loss": 2.4298, + "step": 36519 + }, + { + "epoch": 1.56, + "grad_norm": 0.490234375, + "learning_rate": 2.3423349841133302e-05, + "loss": 2.434, + "step": 36520 + }, + { + "epoch": 1.56, + "grad_norm": 0.478515625, + "learning_rate": 2.3419036017199257e-05, + "loss": 2.486, + "step": 36521 + }, + { + "epoch": 1.56, + "grad_norm": 0.46875, + "learning_rate": 2.341472253785155e-05, + "loss": 2.4899, + "step": 36522 + }, + { + "epoch": 1.56, + "grad_norm": 0.46484375, + "learning_rate": 2.341040940310959e-05, + "loss": 2.557, + "step": 36523 + }, + { + "epoch": 1.56, + "grad_norm": 0.4765625, + "learning_rate": 2.340609661299279e-05, + "loss": 2.4477, + "step": 36524 + }, + { + "epoch": 1.56, + "grad_norm": 0.431640625, + "learning_rate": 2.340178416752059e-05, + "loss": 2.4058, + "step": 36525 + }, + { + "epoch": 1.56, + "grad_norm": 0.4765625, + "learning_rate": 2.3397472066712346e-05, + "loss": 2.5791, + "step": 36526 + }, + { + "epoch": 1.56, + "grad_norm": 0.5, + "learning_rate": 2.3393160310587482e-05, + "loss": 2.5317, + "step": 36527 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.3388848899165426e-05, + "loss": 2.5134, + "step": 36528 + }, + { + "epoch": 1.56, + "grad_norm": 0.498046875, + "learning_rate": 2.3384537832465547e-05, + "loss": 2.3633, + "step": 36529 + }, + { + "epoch": 1.56, + "grad_norm": 0.46875, + "learning_rate": 2.3380227110507226e-05, + "loss": 2.4893, + "step": 36530 + }, + { + "epoch": 1.56, + "grad_norm": 0.49609375, + "learning_rate": 2.3375916733309878e-05, + "loss": 2.6347, + "step": 36531 + }, + { + "epoch": 1.56, + "grad_norm": 0.5234375, + "learning_rate": 2.3371606700892935e-05, + "loss": 2.2397, + "step": 36532 + }, + { + "epoch": 1.56, + "grad_norm": 0.423828125, + "learning_rate": 2.3367297013275724e-05, + "loss": 2.4991, + "step": 36533 + }, + { + "epoch": 1.56, + "grad_norm": 0.5078125, + "learning_rate": 2.33629876704777e-05, + "loss": 2.3413, + "step": 36534 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.3358678672518197e-05, + "loss": 2.4869, + "step": 36535 + }, + { + "epoch": 1.56, + "grad_norm": 0.431640625, + "learning_rate": 2.3354370019416627e-05, + "loss": 2.3946, + "step": 36536 + }, + { + "epoch": 1.56, + "grad_norm": 0.439453125, + "learning_rate": 2.3350061711192415e-05, + "loss": 2.6291, + "step": 36537 + }, + { + "epoch": 1.56, + "grad_norm": 0.5625, + "learning_rate": 2.334575374786492e-05, + "loss": 2.5099, + "step": 36538 + }, + { + "epoch": 1.56, + "grad_norm": 0.439453125, + "learning_rate": 2.334144612945347e-05, + "loss": 2.4266, + "step": 36539 + }, + { + "epoch": 1.56, + "grad_norm": 0.462890625, + "learning_rate": 2.3337138855977515e-05, + "loss": 2.5531, + "step": 36540 + }, + { + "epoch": 1.56, + "grad_norm": 0.5234375, + "learning_rate": 2.3332831927456435e-05, + "loss": 2.4509, + "step": 36541 + }, + { + "epoch": 1.56, + "grad_norm": 0.470703125, + "learning_rate": 2.332852534390957e-05, + "loss": 2.577, + "step": 36542 + }, + { + "epoch": 1.56, + "grad_norm": 0.42578125, + "learning_rate": 2.3324219105356328e-05, + "loss": 2.4847, + "step": 36543 + }, + { + "epoch": 1.56, + "grad_norm": 0.46875, + "learning_rate": 2.331991321181609e-05, + "loss": 2.3518, + "step": 36544 + }, + { + "epoch": 1.56, + "grad_norm": 0.51171875, + "learning_rate": 2.3315607663308225e-05, + "loss": 2.3714, + "step": 36545 + }, + { + "epoch": 1.56, + "grad_norm": 0.51953125, + "learning_rate": 2.331130245985208e-05, + "loss": 2.5021, + "step": 36546 + }, + { + "epoch": 1.56, + "grad_norm": 0.50390625, + "learning_rate": 2.330699760146704e-05, + "loss": 2.428, + "step": 36547 + }, + { + "epoch": 1.56, + "grad_norm": 0.44140625, + "learning_rate": 2.3302693088172512e-05, + "loss": 2.3726, + "step": 36548 + }, + { + "epoch": 1.56, + "grad_norm": 0.470703125, + "learning_rate": 2.329838891998781e-05, + "loss": 2.4285, + "step": 36549 + }, + { + "epoch": 1.56, + "grad_norm": 0.484375, + "learning_rate": 2.329408509693235e-05, + "loss": 2.448, + "step": 36550 + }, + { + "epoch": 1.56, + "grad_norm": 0.494140625, + "learning_rate": 2.3289781619025454e-05, + "loss": 2.4825, + "step": 36551 + }, + { + "epoch": 1.56, + "grad_norm": 0.5078125, + "learning_rate": 2.3285478486286507e-05, + "loss": 2.5895, + "step": 36552 + }, + { + "epoch": 1.56, + "grad_norm": 0.43359375, + "learning_rate": 2.328117569873489e-05, + "loss": 2.4485, + "step": 36553 + }, + { + "epoch": 1.56, + "grad_norm": 0.4921875, + "learning_rate": 2.327687325638994e-05, + "loss": 2.4817, + "step": 36554 + }, + { + "epoch": 1.56, + "grad_norm": 0.48828125, + "learning_rate": 2.3272571159271005e-05, + "loss": 2.5586, + "step": 36555 + }, + { + "epoch": 1.56, + "grad_norm": 0.4453125, + "learning_rate": 2.3268269407397458e-05, + "loss": 2.3111, + "step": 36556 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.3263968000788685e-05, + "loss": 2.4624, + "step": 36557 + }, + { + "epoch": 1.56, + "grad_norm": 0.462890625, + "learning_rate": 2.3259666939463975e-05, + "loss": 2.4572, + "step": 36558 + }, + { + "epoch": 1.56, + "grad_norm": 0.41015625, + "learning_rate": 2.3255366223442732e-05, + "loss": 2.5251, + "step": 36559 + }, + { + "epoch": 1.56, + "grad_norm": 0.404296875, + "learning_rate": 2.3251065852744314e-05, + "loss": 2.3685, + "step": 36560 + }, + { + "epoch": 1.56, + "grad_norm": 0.470703125, + "learning_rate": 2.324676582738805e-05, + "loss": 2.3661, + "step": 36561 + }, + { + "epoch": 1.56, + "grad_norm": 0.447265625, + "learning_rate": 2.324246614739326e-05, + "loss": 2.3966, + "step": 36562 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.323816681277933e-05, + "loss": 2.488, + "step": 36563 + }, + { + "epoch": 1.56, + "grad_norm": 0.50390625, + "learning_rate": 2.3233867823565612e-05, + "loss": 2.3888, + "step": 36564 + }, + { + "epoch": 1.56, + "grad_norm": 0.458984375, + "learning_rate": 2.3229569179771417e-05, + "loss": 2.3476, + "step": 36565 + }, + { + "epoch": 1.56, + "grad_norm": 0.50390625, + "learning_rate": 2.322527088141613e-05, + "loss": 2.4975, + "step": 36566 + }, + { + "epoch": 1.56, + "grad_norm": 0.484375, + "learning_rate": 2.3220972928519035e-05, + "loss": 2.4412, + "step": 36567 + }, + { + "epoch": 1.56, + "grad_norm": 0.447265625, + "learning_rate": 2.32166753210995e-05, + "loss": 2.4808, + "step": 36568 + }, + { + "epoch": 1.56, + "grad_norm": 0.458984375, + "learning_rate": 2.3212378059176898e-05, + "loss": 2.4571, + "step": 36569 + }, + { + "epoch": 1.56, + "grad_norm": 0.4921875, + "learning_rate": 2.320808114277053e-05, + "loss": 2.2664, + "step": 36570 + }, + { + "epoch": 1.56, + "grad_norm": 0.4765625, + "learning_rate": 2.32037845718997e-05, + "loss": 2.4462, + "step": 36571 + }, + { + "epoch": 1.56, + "grad_norm": 0.470703125, + "learning_rate": 2.3199488346583786e-05, + "loss": 2.4353, + "step": 36572 + }, + { + "epoch": 1.56, + "grad_norm": 0.50390625, + "learning_rate": 2.319519246684212e-05, + "loss": 2.3656, + "step": 36573 + }, + { + "epoch": 1.56, + "grad_norm": 0.51171875, + "learning_rate": 2.3190896932694007e-05, + "loss": 2.4948, + "step": 36574 + }, + { + "epoch": 1.56, + "grad_norm": 0.48828125, + "learning_rate": 2.3186601744158775e-05, + "loss": 2.5103, + "step": 36575 + }, + { + "epoch": 1.56, + "grad_norm": 0.455078125, + "learning_rate": 2.318230690125579e-05, + "loss": 2.4042, + "step": 36576 + }, + { + "epoch": 1.56, + "grad_norm": 0.439453125, + "learning_rate": 2.317801240400436e-05, + "loss": 2.3316, + "step": 36577 + }, + { + "epoch": 1.56, + "grad_norm": 0.443359375, + "learning_rate": 2.317371825242376e-05, + "loss": 2.3998, + "step": 36578 + }, + { + "epoch": 1.56, + "grad_norm": 0.412109375, + "learning_rate": 2.3169424446533362e-05, + "loss": 2.314, + "step": 36579 + }, + { + "epoch": 1.56, + "grad_norm": 0.474609375, + "learning_rate": 2.31651309863525e-05, + "loss": 2.5303, + "step": 36580 + }, + { + "epoch": 1.56, + "grad_norm": 0.447265625, + "learning_rate": 2.316083787190043e-05, + "loss": 2.4664, + "step": 36581 + }, + { + "epoch": 1.56, + "grad_norm": 0.451171875, + "learning_rate": 2.315654510319655e-05, + "loss": 2.4494, + "step": 36582 + }, + { + "epoch": 1.56, + "grad_norm": 0.51953125, + "learning_rate": 2.31522526802601e-05, + "loss": 2.533, + "step": 36583 + }, + { + "epoch": 1.56, + "grad_norm": 0.48046875, + "learning_rate": 2.314796060311043e-05, + "loss": 2.4534, + "step": 36584 + }, + { + "epoch": 1.56, + "grad_norm": 0.498046875, + "learning_rate": 2.3143668871766878e-05, + "loss": 2.4696, + "step": 36585 + }, + { + "epoch": 1.56, + "grad_norm": 0.44921875, + "learning_rate": 2.3139377486248738e-05, + "loss": 2.3554, + "step": 36586 + }, + { + "epoch": 1.56, + "grad_norm": 0.515625, + "learning_rate": 2.3135086446575272e-05, + "loss": 2.4671, + "step": 36587 + }, + { + "epoch": 1.56, + "grad_norm": 0.42578125, + "learning_rate": 2.3130795752765822e-05, + "loss": 2.4897, + "step": 36588 + }, + { + "epoch": 1.56, + "grad_norm": 0.435546875, + "learning_rate": 2.312650540483974e-05, + "loss": 2.4695, + "step": 36589 + }, + { + "epoch": 1.56, + "grad_norm": 0.546875, + "learning_rate": 2.312221540281625e-05, + "loss": 2.5884, + "step": 36590 + }, + { + "epoch": 1.56, + "grad_norm": 0.47265625, + "learning_rate": 2.3117925746714698e-05, + "loss": 2.392, + "step": 36591 + }, + { + "epoch": 1.56, + "grad_norm": 0.431640625, + "learning_rate": 2.3113636436554408e-05, + "loss": 2.3542, + "step": 36592 + }, + { + "epoch": 1.56, + "grad_norm": 0.4375, + "learning_rate": 2.310934747235466e-05, + "loss": 2.4667, + "step": 36593 + }, + { + "epoch": 1.56, + "grad_norm": 0.5078125, + "learning_rate": 2.310505885413472e-05, + "loss": 2.4704, + "step": 36594 + }, + { + "epoch": 1.56, + "grad_norm": 0.55859375, + "learning_rate": 2.310077058191391e-05, + "loss": 2.4203, + "step": 36595 + }, + { + "epoch": 1.56, + "grad_norm": 0.46875, + "learning_rate": 2.309648265571156e-05, + "loss": 2.5128, + "step": 36596 + }, + { + "epoch": 1.56, + "grad_norm": 0.4609375, + "learning_rate": 2.3092195075546898e-05, + "loss": 2.4273, + "step": 36597 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.3087907841439293e-05, + "loss": 2.564, + "step": 36598 + }, + { + "epoch": 1.56, + "grad_norm": 0.515625, + "learning_rate": 2.3083620953407958e-05, + "loss": 2.4732, + "step": 36599 + }, + { + "epoch": 1.56, + "grad_norm": 0.50390625, + "learning_rate": 2.307933441147222e-05, + "loss": 2.4541, + "step": 36600 + }, + { + "epoch": 1.56, + "grad_norm": 0.484375, + "learning_rate": 2.3075048215651397e-05, + "loss": 2.4507, + "step": 36601 + }, + { + "epoch": 1.56, + "grad_norm": 0.44921875, + "learning_rate": 2.3070762365964737e-05, + "loss": 2.3987, + "step": 36602 + }, + { + "epoch": 1.56, + "grad_norm": 0.515625, + "learning_rate": 2.3066476862431506e-05, + "loss": 2.4929, + "step": 36603 + }, + { + "epoch": 1.56, + "grad_norm": 0.498046875, + "learning_rate": 2.306219170507101e-05, + "loss": 2.4765, + "step": 36604 + }, + { + "epoch": 1.56, + "grad_norm": 0.447265625, + "learning_rate": 2.305790689390257e-05, + "loss": 2.4338, + "step": 36605 + }, + { + "epoch": 1.56, + "grad_norm": 0.47265625, + "learning_rate": 2.30536224289454e-05, + "loss": 2.5441, + "step": 36606 + }, + { + "epoch": 1.56, + "grad_norm": 0.462890625, + "learning_rate": 2.3049338310218804e-05, + "loss": 2.5217, + "step": 36607 + }, + { + "epoch": 1.56, + "grad_norm": 0.42578125, + "learning_rate": 2.3045054537742094e-05, + "loss": 2.3785, + "step": 36608 + }, + { + "epoch": 1.56, + "grad_norm": 0.51171875, + "learning_rate": 2.304077111153452e-05, + "loss": 2.4364, + "step": 36609 + }, + { + "epoch": 1.56, + "grad_norm": 0.515625, + "learning_rate": 2.303648803161532e-05, + "loss": 2.5236, + "step": 36610 + }, + { + "epoch": 1.56, + "grad_norm": 0.52734375, + "learning_rate": 2.3032205298003795e-05, + "loss": 2.4697, + "step": 36611 + }, + { + "epoch": 1.56, + "grad_norm": 0.51953125, + "learning_rate": 2.3027922910719248e-05, + "loss": 2.4572, + "step": 36612 + }, + { + "epoch": 1.56, + "grad_norm": 0.451171875, + "learning_rate": 2.302364086978088e-05, + "loss": 2.4422, + "step": 36613 + }, + { + "epoch": 1.56, + "grad_norm": 0.48046875, + "learning_rate": 2.3019359175208033e-05, + "loss": 2.4708, + "step": 36614 + }, + { + "epoch": 1.56, + "grad_norm": 0.447265625, + "learning_rate": 2.3015077827019903e-05, + "loss": 2.3856, + "step": 36615 + }, + { + "epoch": 1.56, + "grad_norm": 0.47265625, + "learning_rate": 2.3010796825235793e-05, + "loss": 2.376, + "step": 36616 + }, + { + "epoch": 1.56, + "grad_norm": 0.53125, + "learning_rate": 2.3006516169874993e-05, + "loss": 2.5489, + "step": 36617 + }, + { + "epoch": 1.56, + "grad_norm": 0.53125, + "learning_rate": 2.3002235860956723e-05, + "loss": 2.6221, + "step": 36618 + }, + { + "epoch": 1.56, + "grad_norm": 0.53515625, + "learning_rate": 2.2997955898500234e-05, + "loss": 2.5821, + "step": 36619 + }, + { + "epoch": 1.56, + "grad_norm": 0.439453125, + "learning_rate": 2.299367628252479e-05, + "loss": 2.4191, + "step": 36620 + }, + { + "epoch": 1.56, + "grad_norm": 0.423828125, + "learning_rate": 2.2989397013049697e-05, + "loss": 2.379, + "step": 36621 + }, + { + "epoch": 1.56, + "grad_norm": 0.515625, + "learning_rate": 2.2985118090094148e-05, + "loss": 2.4742, + "step": 36622 + }, + { + "epoch": 1.56, + "grad_norm": 0.392578125, + "learning_rate": 2.2980839513677444e-05, + "loss": 2.3794, + "step": 36623 + }, + { + "epoch": 1.56, + "grad_norm": 0.48046875, + "learning_rate": 2.297656128381879e-05, + "loss": 2.5342, + "step": 36624 + }, + { + "epoch": 1.56, + "grad_norm": 0.4453125, + "learning_rate": 2.2972283400537487e-05, + "loss": 2.422, + "step": 36625 + }, + { + "epoch": 1.56, + "grad_norm": 0.435546875, + "learning_rate": 2.2968005863852726e-05, + "loss": 2.3768, + "step": 36626 + }, + { + "epoch": 1.56, + "grad_norm": 0.396484375, + "learning_rate": 2.2963728673783825e-05, + "loss": 2.3452, + "step": 36627 + }, + { + "epoch": 1.56, + "grad_norm": 0.484375, + "learning_rate": 2.2959451830349954e-05, + "loss": 2.4366, + "step": 36628 + }, + { + "epoch": 1.56, + "grad_norm": 0.5390625, + "learning_rate": 2.29551753335704e-05, + "loss": 2.5938, + "step": 36629 + }, + { + "epoch": 1.56, + "grad_norm": 0.515625, + "learning_rate": 2.2950899183464424e-05, + "loss": 2.6233, + "step": 36630 + }, + { + "epoch": 1.56, + "grad_norm": 0.44921875, + "learning_rate": 2.2946623380051247e-05, + "loss": 2.4027, + "step": 36631 + }, + { + "epoch": 1.56, + "grad_norm": 0.45703125, + "learning_rate": 2.294234792335005e-05, + "loss": 2.4848, + "step": 36632 + }, + { + "epoch": 1.56, + "grad_norm": 0.4453125, + "learning_rate": 2.2938072813380186e-05, + "loss": 2.4384, + "step": 36633 + }, + { + "epoch": 1.56, + "grad_norm": 0.48046875, + "learning_rate": 2.2933798050160825e-05, + "loss": 2.4355, + "step": 36634 + }, + { + "epoch": 1.56, + "grad_norm": 0.486328125, + "learning_rate": 2.292952363371118e-05, + "loss": 2.6243, + "step": 36635 + }, + { + "epoch": 1.56, + "grad_norm": 0.462890625, + "learning_rate": 2.292524956405051e-05, + "loss": 2.4132, + "step": 36636 + }, + { + "epoch": 1.56, + "grad_norm": 0.4453125, + "learning_rate": 2.2920975841198087e-05, + "loss": 2.4883, + "step": 36637 + }, + { + "epoch": 1.56, + "grad_norm": 0.435546875, + "learning_rate": 2.291670246517307e-05, + "loss": 2.5219, + "step": 36638 + }, + { + "epoch": 1.56, + "grad_norm": 0.41796875, + "learning_rate": 2.2912429435994743e-05, + "loss": 2.3589, + "step": 36639 + }, + { + "epoch": 1.56, + "grad_norm": 0.453125, + "learning_rate": 2.2908156753682298e-05, + "loss": 2.4383, + "step": 36640 + }, + { + "epoch": 1.56, + "grad_norm": 0.462890625, + "learning_rate": 2.2903884418254985e-05, + "loss": 2.425, + "step": 36641 + }, + { + "epoch": 1.56, + "grad_norm": 0.48828125, + "learning_rate": 2.2899612429731998e-05, + "loss": 2.4816, + "step": 36642 + }, + { + "epoch": 1.56, + "grad_norm": 0.404296875, + "learning_rate": 2.28953407881326e-05, + "loss": 2.3342, + "step": 36643 + }, + { + "epoch": 1.56, + "grad_norm": 0.419921875, + "learning_rate": 2.289106949347597e-05, + "loss": 2.3982, + "step": 36644 + }, + { + "epoch": 1.56, + "grad_norm": 0.48046875, + "learning_rate": 2.288679854578134e-05, + "loss": 2.5695, + "step": 36645 + }, + { + "epoch": 1.56, + "grad_norm": 0.46875, + "learning_rate": 2.288252794506798e-05, + "loss": 2.4176, + "step": 36646 + }, + { + "epoch": 1.56, + "grad_norm": 0.474609375, + "learning_rate": 2.2878257691355044e-05, + "loss": 2.4043, + "step": 36647 + }, + { + "epoch": 1.56, + "grad_norm": 0.482421875, + "learning_rate": 2.287398778466171e-05, + "loss": 2.3157, + "step": 36648 + }, + { + "epoch": 1.56, + "grad_norm": 0.5078125, + "learning_rate": 2.2869718225007297e-05, + "loss": 2.4315, + "step": 36649 + }, + { + "epoch": 1.56, + "grad_norm": 0.486328125, + "learning_rate": 2.2865449012410968e-05, + "loss": 2.5624, + "step": 36650 + }, + { + "epoch": 1.56, + "grad_norm": 0.453125, + "learning_rate": 2.2861180146891904e-05, + "loss": 2.4626, + "step": 36651 + }, + { + "epoch": 1.56, + "grad_norm": 0.443359375, + "learning_rate": 2.2856911628469334e-05, + "loss": 2.6491, + "step": 36652 + }, + { + "epoch": 1.56, + "grad_norm": 0.4609375, + "learning_rate": 2.2852643457162502e-05, + "loss": 2.3629, + "step": 36653 + }, + { + "epoch": 1.56, + "grad_norm": 0.515625, + "learning_rate": 2.284837563299055e-05, + "loss": 2.4026, + "step": 36654 + }, + { + "epoch": 1.56, + "grad_norm": 0.435546875, + "learning_rate": 2.2844108155972744e-05, + "loss": 2.3188, + "step": 36655 + }, + { + "epoch": 1.56, + "grad_norm": 0.474609375, + "learning_rate": 2.283984102612822e-05, + "loss": 2.4342, + "step": 36656 + }, + { + "epoch": 1.56, + "grad_norm": 0.484375, + "learning_rate": 2.2835574243476244e-05, + "loss": 2.4268, + "step": 36657 + }, + { + "epoch": 1.56, + "grad_norm": 0.455078125, + "learning_rate": 2.2831307808035952e-05, + "loss": 2.3843, + "step": 36658 + }, + { + "epoch": 1.56, + "grad_norm": 0.486328125, + "learning_rate": 2.2827041719826602e-05, + "loss": 2.6376, + "step": 36659 + }, + { + "epoch": 1.56, + "grad_norm": 0.443359375, + "learning_rate": 2.282277597886734e-05, + "loss": 2.4949, + "step": 36660 + }, + { + "epoch": 1.56, + "grad_norm": 0.400390625, + "learning_rate": 2.2818510585177366e-05, + "loss": 2.3662, + "step": 36661 + }, + { + "epoch": 1.56, + "grad_norm": 0.546875, + "learning_rate": 2.2814245538775925e-05, + "loss": 2.4745, + "step": 36662 + }, + { + "epoch": 1.56, + "grad_norm": 0.45703125, + "learning_rate": 2.2809980839682177e-05, + "loss": 2.3297, + "step": 36663 + }, + { + "epoch": 1.56, + "grad_norm": 0.482421875, + "learning_rate": 2.280571648791524e-05, + "loss": 2.3957, + "step": 36664 + }, + { + "epoch": 1.56, + "grad_norm": 0.4765625, + "learning_rate": 2.2801452483494423e-05, + "loss": 2.5055, + "step": 36665 + }, + { + "epoch": 1.56, + "grad_norm": 0.5078125, + "learning_rate": 2.2797188826438865e-05, + "loss": 2.4596, + "step": 36666 + }, + { + "epoch": 1.56, + "grad_norm": 0.443359375, + "learning_rate": 2.2792925516767705e-05, + "loss": 2.3802, + "step": 36667 + }, + { + "epoch": 1.56, + "grad_norm": 0.46484375, + "learning_rate": 2.278866255450017e-05, + "loss": 2.4431, + "step": 36668 + }, + { + "epoch": 1.56, + "grad_norm": 0.4375, + "learning_rate": 2.2784399939655467e-05, + "loss": 2.3774, + "step": 36669 + }, + { + "epoch": 1.56, + "grad_norm": 0.431640625, + "learning_rate": 2.278013767225271e-05, + "loss": 2.3789, + "step": 36670 + }, + { + "epoch": 1.56, + "grad_norm": 0.49609375, + "learning_rate": 2.2775875752311148e-05, + "loss": 2.378, + "step": 36671 + }, + { + "epoch": 1.56, + "grad_norm": 0.40625, + "learning_rate": 2.2771614179849888e-05, + "loss": 2.376, + "step": 36672 + }, + { + "epoch": 1.56, + "grad_norm": 0.431640625, + "learning_rate": 2.2767352954888177e-05, + "loss": 2.4467, + "step": 36673 + }, + { + "epoch": 1.56, + "grad_norm": 0.484375, + "learning_rate": 2.2763092077445114e-05, + "loss": 2.3942, + "step": 36674 + }, + { + "epoch": 1.56, + "grad_norm": 0.5234375, + "learning_rate": 2.275883154753995e-05, + "loss": 2.5805, + "step": 36675 + }, + { + "epoch": 1.56, + "grad_norm": 0.458984375, + "learning_rate": 2.275457136519178e-05, + "loss": 2.4512, + "step": 36676 + }, + { + "epoch": 1.56, + "grad_norm": 0.4453125, + "learning_rate": 2.275031153041981e-05, + "loss": 2.2539, + "step": 36677 + }, + { + "epoch": 1.56, + "grad_norm": 0.498046875, + "learning_rate": 2.2746052043243225e-05, + "loss": 2.4457, + "step": 36678 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.274179290368119e-05, + "loss": 2.5027, + "step": 36679 + }, + { + "epoch": 1.56, + "grad_norm": 0.423828125, + "learning_rate": 2.273753411175279e-05, + "loss": 2.4962, + "step": 36680 + }, + { + "epoch": 1.56, + "grad_norm": 0.4453125, + "learning_rate": 2.2733275667477293e-05, + "loss": 2.4371, + "step": 36681 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.2729017570873835e-05, + "loss": 2.4535, + "step": 36682 + }, + { + "epoch": 1.56, + "grad_norm": 0.54296875, + "learning_rate": 2.272475982196153e-05, + "loss": 2.3987, + "step": 36683 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.272050242075956e-05, + "loss": 2.4287, + "step": 36684 + }, + { + "epoch": 1.56, + "grad_norm": 0.470703125, + "learning_rate": 2.2716245367287104e-05, + "loss": 2.3464, + "step": 36685 + }, + { + "epoch": 1.56, + "grad_norm": 0.4375, + "learning_rate": 2.271198866156329e-05, + "loss": 2.3174, + "step": 36686 + }, + { + "epoch": 1.56, + "grad_norm": 0.404296875, + "learning_rate": 2.27077323036073e-05, + "loss": 2.3688, + "step": 36687 + }, + { + "epoch": 1.56, + "grad_norm": 0.443359375, + "learning_rate": 2.2703476293438242e-05, + "loss": 2.3445, + "step": 36688 + }, + { + "epoch": 1.56, + "grad_norm": 0.4296875, + "learning_rate": 2.2699220631075313e-05, + "loss": 2.3714, + "step": 36689 + }, + { + "epoch": 1.56, + "grad_norm": 0.52734375, + "learning_rate": 2.2694965316537618e-05, + "loss": 2.4652, + "step": 36690 + }, + { + "epoch": 1.56, + "grad_norm": 0.416015625, + "learning_rate": 2.2690710349844357e-05, + "loss": 2.4463, + "step": 36691 + }, + { + "epoch": 1.56, + "grad_norm": 0.51171875, + "learning_rate": 2.268645573101462e-05, + "loss": 2.4471, + "step": 36692 + }, + { + "epoch": 1.56, + "grad_norm": 0.51953125, + "learning_rate": 2.2682201460067576e-05, + "loss": 2.5311, + "step": 36693 + }, + { + "epoch": 1.56, + "grad_norm": 0.462890625, + "learning_rate": 2.2677947537022402e-05, + "loss": 2.3861, + "step": 36694 + }, + { + "epoch": 1.56, + "grad_norm": 0.447265625, + "learning_rate": 2.26736939618982e-05, + "loss": 2.4561, + "step": 36695 + }, + { + "epoch": 1.56, + "grad_norm": 0.400390625, + "learning_rate": 2.2669440734714054e-05, + "loss": 2.3576, + "step": 36696 + }, + { + "epoch": 1.56, + "grad_norm": 0.4296875, + "learning_rate": 2.266518785548922e-05, + "loss": 2.4989, + "step": 36697 + }, + { + "epoch": 1.56, + "grad_norm": 0.490234375, + "learning_rate": 2.2660935324242794e-05, + "loss": 2.4173, + "step": 36698 + }, + { + "epoch": 1.56, + "grad_norm": 0.43359375, + "learning_rate": 2.265668314099385e-05, + "loss": 2.4621, + "step": 36699 + }, + { + "epoch": 1.56, + "grad_norm": 0.474609375, + "learning_rate": 2.265243130576158e-05, + "loss": 2.6463, + "step": 36700 + }, + { + "epoch": 1.56, + "grad_norm": 0.46484375, + "learning_rate": 2.264817981856512e-05, + "loss": 2.5484, + "step": 36701 + }, + { + "epoch": 1.56, + "grad_norm": 0.462890625, + "learning_rate": 2.2643928679423553e-05, + "loss": 2.4706, + "step": 36702 + }, + { + "epoch": 1.56, + "grad_norm": 0.494140625, + "learning_rate": 2.263967788835607e-05, + "loss": 2.4255, + "step": 36703 + }, + { + "epoch": 1.56, + "grad_norm": 0.51953125, + "learning_rate": 2.263542744538173e-05, + "loss": 2.4218, + "step": 36704 + }, + { + "epoch": 1.56, + "grad_norm": 0.5234375, + "learning_rate": 2.2631177350519728e-05, + "loss": 2.5018, + "step": 36705 + }, + { + "epoch": 1.56, + "grad_norm": 0.466796875, + "learning_rate": 2.2626927603789126e-05, + "loss": 2.5177, + "step": 36706 + }, + { + "epoch": 1.56, + "grad_norm": 0.5078125, + "learning_rate": 2.2622678205209092e-05, + "loss": 2.4994, + "step": 36707 + }, + { + "epoch": 1.56, + "grad_norm": 0.47265625, + "learning_rate": 2.2618429154798703e-05, + "loss": 2.5711, + "step": 36708 + }, + { + "epoch": 1.56, + "grad_norm": 0.5546875, + "learning_rate": 2.2614180452577106e-05, + "loss": 2.5726, + "step": 36709 + }, + { + "epoch": 1.56, + "grad_norm": 0.48046875, + "learning_rate": 2.2609932098563447e-05, + "loss": 2.5411, + "step": 36710 + }, + { + "epoch": 1.56, + "grad_norm": 0.47265625, + "learning_rate": 2.2605684092776802e-05, + "loss": 2.4091, + "step": 36711 + }, + { + "epoch": 1.56, + "grad_norm": 0.48828125, + "learning_rate": 2.2601436435236244e-05, + "loss": 2.4753, + "step": 36712 + }, + { + "epoch": 1.56, + "grad_norm": 0.421875, + "learning_rate": 2.2597189125960993e-05, + "loss": 2.402, + "step": 36713 + }, + { + "epoch": 1.56, + "grad_norm": 0.490234375, + "learning_rate": 2.25929421649701e-05, + "loss": 2.5161, + "step": 36714 + }, + { + "epoch": 1.56, + "grad_norm": 0.412109375, + "learning_rate": 2.258869555228266e-05, + "loss": 2.4578, + "step": 36715 + }, + { + "epoch": 1.56, + "grad_norm": 0.46875, + "learning_rate": 2.2584449287917785e-05, + "loss": 2.4667, + "step": 36716 + }, + { + "epoch": 1.56, + "grad_norm": 0.453125, + "learning_rate": 2.2580203371894626e-05, + "loss": 2.2856, + "step": 36717 + }, + { + "epoch": 1.56, + "grad_norm": 0.4296875, + "learning_rate": 2.2575957804232238e-05, + "loss": 2.4711, + "step": 36718 + }, + { + "epoch": 1.56, + "grad_norm": 0.451171875, + "learning_rate": 2.2571712584949768e-05, + "loss": 2.5798, + "step": 36719 + }, + { + "epoch": 1.56, + "grad_norm": 0.435546875, + "learning_rate": 2.256746771406627e-05, + "loss": 2.325, + "step": 36720 + }, + { + "epoch": 1.56, + "grad_norm": 0.455078125, + "learning_rate": 2.256322319160089e-05, + "loss": 2.3025, + "step": 36721 + }, + { + "epoch": 1.56, + "grad_norm": 0.435546875, + "learning_rate": 2.255897901757268e-05, + "loss": 2.4989, + "step": 36722 + }, + { + "epoch": 1.56, + "grad_norm": 0.47265625, + "learning_rate": 2.25547351920008e-05, + "loss": 2.3872, + "step": 36723 + }, + { + "epoch": 1.56, + "grad_norm": 0.490234375, + "learning_rate": 2.2550491714904263e-05, + "loss": 2.4091, + "step": 36724 + }, + { + "epoch": 1.56, + "grad_norm": 0.427734375, + "learning_rate": 2.2546248586302222e-05, + "loss": 2.6341, + "step": 36725 + }, + { + "epoch": 1.56, + "grad_norm": 0.515625, + "learning_rate": 2.2542005806213785e-05, + "loss": 2.4251, + "step": 36726 + }, + { + "epoch": 1.57, + "grad_norm": 0.47265625, + "learning_rate": 2.2537763374658016e-05, + "loss": 2.6241, + "step": 36727 + }, + { + "epoch": 1.57, + "grad_norm": 0.4296875, + "learning_rate": 2.2533521291653936e-05, + "loss": 2.3892, + "step": 36728 + }, + { + "epoch": 1.57, + "grad_norm": 0.40234375, + "learning_rate": 2.252927955722076e-05, + "loss": 2.3696, + "step": 36729 + }, + { + "epoch": 1.57, + "grad_norm": 0.490234375, + "learning_rate": 2.2525038171377523e-05, + "loss": 2.5248, + "step": 36730 + }, + { + "epoch": 1.57, + "grad_norm": 0.484375, + "learning_rate": 2.2520797134143257e-05, + "loss": 2.4249, + "step": 36731 + }, + { + "epoch": 1.57, + "grad_norm": 0.53515625, + "learning_rate": 2.2516556445537094e-05, + "loss": 2.4937, + "step": 36732 + }, + { + "epoch": 1.57, + "grad_norm": 0.455078125, + "learning_rate": 2.2512316105578146e-05, + "loss": 2.4747, + "step": 36733 + }, + { + "epoch": 1.57, + "grad_norm": 0.41015625, + "learning_rate": 2.2508076114285415e-05, + "loss": 2.3822, + "step": 36734 + }, + { + "epoch": 1.57, + "grad_norm": 0.46875, + "learning_rate": 2.2503836471678064e-05, + "loss": 2.4739, + "step": 36735 + }, + { + "epoch": 1.57, + "grad_norm": 0.49609375, + "learning_rate": 2.2499597177775088e-05, + "loss": 2.3852, + "step": 36736 + }, + { + "epoch": 1.57, + "grad_norm": 0.4296875, + "learning_rate": 2.249535823259563e-05, + "loss": 2.4287, + "step": 36737 + }, + { + "epoch": 1.57, + "grad_norm": 0.42578125, + "learning_rate": 2.249111963615872e-05, + "loss": 2.4579, + "step": 36738 + }, + { + "epoch": 1.57, + "grad_norm": 0.5390625, + "learning_rate": 2.248688138848346e-05, + "loss": 2.4173, + "step": 36739 + }, + { + "epoch": 1.57, + "grad_norm": 0.44921875, + "learning_rate": 2.2482643489588884e-05, + "loss": 2.4341, + "step": 36740 + }, + { + "epoch": 1.57, + "grad_norm": 0.53125, + "learning_rate": 2.2478405939494073e-05, + "loss": 2.4965, + "step": 36741 + }, + { + "epoch": 1.57, + "grad_norm": 0.466796875, + "learning_rate": 2.247416873821815e-05, + "loss": 2.4327, + "step": 36742 + }, + { + "epoch": 1.57, + "grad_norm": 0.443359375, + "learning_rate": 2.2469931885780125e-05, + "loss": 2.3811, + "step": 36743 + }, + { + "epoch": 1.57, + "grad_norm": 0.439453125, + "learning_rate": 2.2465695382199037e-05, + "loss": 2.3982, + "step": 36744 + }, + { + "epoch": 1.57, + "grad_norm": 0.44140625, + "learning_rate": 2.2461459227493998e-05, + "loss": 2.2893, + "step": 36745 + }, + { + "epoch": 1.57, + "grad_norm": 0.435546875, + "learning_rate": 2.2457223421684082e-05, + "loss": 2.3766, + "step": 36746 + }, + { + "epoch": 1.57, + "grad_norm": 0.44921875, + "learning_rate": 2.2452987964788285e-05, + "loss": 2.3497, + "step": 36747 + }, + { + "epoch": 1.57, + "grad_norm": 0.44921875, + "learning_rate": 2.244875285682574e-05, + "loss": 2.3881, + "step": 36748 + }, + { + "epoch": 1.57, + "grad_norm": 0.419921875, + "learning_rate": 2.2444518097815427e-05, + "loss": 2.3583, + "step": 36749 + }, + { + "epoch": 1.57, + "grad_norm": 0.41015625, + "learning_rate": 2.244028368777644e-05, + "loss": 2.2709, + "step": 36750 + }, + { + "epoch": 1.57, + "grad_norm": 0.443359375, + "learning_rate": 2.2436049626727874e-05, + "loss": 2.383, + "step": 36751 + }, + { + "epoch": 1.57, + "grad_norm": 0.49609375, + "learning_rate": 2.2431815914688727e-05, + "loss": 2.3787, + "step": 36752 + }, + { + "epoch": 1.57, + "grad_norm": 0.50390625, + "learning_rate": 2.242758255167803e-05, + "loss": 2.5184, + "step": 36753 + }, + { + "epoch": 1.57, + "grad_norm": 0.46875, + "learning_rate": 2.2423349537714865e-05, + "loss": 2.3967, + "step": 36754 + }, + { + "epoch": 1.57, + "grad_norm": 0.453125, + "learning_rate": 2.24191168728183e-05, + "loss": 2.4367, + "step": 36755 + }, + { + "epoch": 1.57, + "grad_norm": 0.5234375, + "learning_rate": 2.2414884557007342e-05, + "loss": 2.4114, + "step": 36756 + }, + { + "epoch": 1.57, + "grad_norm": 0.447265625, + "learning_rate": 2.241065259030104e-05, + "loss": 2.6174, + "step": 36757 + }, + { + "epoch": 1.57, + "grad_norm": 0.515625, + "learning_rate": 2.240642097271848e-05, + "loss": 2.4848, + "step": 36758 + }, + { + "epoch": 1.57, + "grad_norm": 0.498046875, + "learning_rate": 2.2402189704278653e-05, + "loss": 2.5644, + "step": 36759 + }, + { + "epoch": 1.57, + "grad_norm": 0.5078125, + "learning_rate": 2.2397958785000594e-05, + "loss": 2.2811, + "step": 36760 + }, + { + "epoch": 1.57, + "grad_norm": 0.494140625, + "learning_rate": 2.2393728214903364e-05, + "loss": 2.3979, + "step": 36761 + }, + { + "epoch": 1.57, + "grad_norm": 0.435546875, + "learning_rate": 2.2389497994006026e-05, + "loss": 2.4567, + "step": 36762 + }, + { + "epoch": 1.57, + "grad_norm": 0.4609375, + "learning_rate": 2.2385268122327553e-05, + "loss": 2.3813, + "step": 36763 + }, + { + "epoch": 1.57, + "grad_norm": 0.490234375, + "learning_rate": 2.2381038599887028e-05, + "loss": 2.4711, + "step": 36764 + }, + { + "epoch": 1.57, + "grad_norm": 0.478515625, + "learning_rate": 2.237680942670344e-05, + "loss": 2.4729, + "step": 36765 + }, + { + "epoch": 1.57, + "grad_norm": 0.51953125, + "learning_rate": 2.2372580602795844e-05, + "loss": 2.4501, + "step": 36766 + }, + { + "epoch": 1.57, + "grad_norm": 0.470703125, + "learning_rate": 2.236835212818329e-05, + "loss": 2.4525, + "step": 36767 + }, + { + "epoch": 1.57, + "grad_norm": 0.462890625, + "learning_rate": 2.2364124002884778e-05, + "loss": 2.4027, + "step": 36768 + }, + { + "epoch": 1.57, + "grad_norm": 0.46875, + "learning_rate": 2.235989622691932e-05, + "loss": 2.4439, + "step": 36769 + }, + { + "epoch": 1.57, + "grad_norm": 0.462890625, + "learning_rate": 2.235566880030594e-05, + "loss": 2.2862, + "step": 36770 + }, + { + "epoch": 1.57, + "grad_norm": 0.5, + "learning_rate": 2.2351441723063715e-05, + "loss": 2.292, + "step": 36771 + }, + { + "epoch": 1.57, + "grad_norm": 0.40234375, + "learning_rate": 2.2347214995211587e-05, + "loss": 2.3093, + "step": 36772 + }, + { + "epoch": 1.57, + "grad_norm": 0.478515625, + "learning_rate": 2.2342988616768623e-05, + "loss": 2.4344, + "step": 36773 + }, + { + "epoch": 1.57, + "grad_norm": 0.408203125, + "learning_rate": 2.2338762587753848e-05, + "loss": 2.3509, + "step": 36774 + }, + { + "epoch": 1.57, + "grad_norm": 0.4453125, + "learning_rate": 2.2334536908186266e-05, + "loss": 2.2969, + "step": 36775 + }, + { + "epoch": 1.57, + "grad_norm": 0.4921875, + "learning_rate": 2.233031157808485e-05, + "loss": 2.4665, + "step": 36776 + }, + { + "epoch": 1.57, + "grad_norm": 0.53515625, + "learning_rate": 2.2326086597468653e-05, + "loss": 2.4585, + "step": 36777 + }, + { + "epoch": 1.57, + "grad_norm": 0.466796875, + "learning_rate": 2.2321861966356705e-05, + "loss": 2.4242, + "step": 36778 + }, + { + "epoch": 1.57, + "grad_norm": 0.412109375, + "learning_rate": 2.2317637684767956e-05, + "loss": 2.4086, + "step": 36779 + }, + { + "epoch": 1.57, + "grad_norm": 0.490234375, + "learning_rate": 2.231341375272149e-05, + "loss": 2.3593, + "step": 36780 + }, + { + "epoch": 1.57, + "grad_norm": 0.404296875, + "learning_rate": 2.2309190170236227e-05, + "loss": 2.4262, + "step": 36781 + }, + { + "epoch": 1.57, + "grad_norm": 0.443359375, + "learning_rate": 2.2304966937331228e-05, + "loss": 2.2914, + "step": 36782 + }, + { + "epoch": 1.57, + "grad_norm": 0.49609375, + "learning_rate": 2.23007440540255e-05, + "loss": 2.4631, + "step": 36783 + }, + { + "epoch": 1.57, + "grad_norm": 0.490234375, + "learning_rate": 2.229652152033804e-05, + "loss": 2.4303, + "step": 36784 + }, + { + "epoch": 1.57, + "grad_norm": 0.5078125, + "learning_rate": 2.22922993362878e-05, + "loss": 2.42, + "step": 36785 + }, + { + "epoch": 1.57, + "grad_norm": 0.458984375, + "learning_rate": 2.2288077501893823e-05, + "loss": 2.4814, + "step": 36786 + }, + { + "epoch": 1.57, + "grad_norm": 0.416015625, + "learning_rate": 2.228385601717511e-05, + "loss": 2.3501, + "step": 36787 + }, + { + "epoch": 1.57, + "grad_norm": 0.50390625, + "learning_rate": 2.2279634882150624e-05, + "loss": 2.5549, + "step": 36788 + }, + { + "epoch": 1.57, + "grad_norm": 0.4921875, + "learning_rate": 2.2275414096839375e-05, + "loss": 2.4317, + "step": 36789 + }, + { + "epoch": 1.57, + "grad_norm": 0.515625, + "learning_rate": 2.2271193661260392e-05, + "loss": 2.5871, + "step": 36790 + }, + { + "epoch": 1.57, + "grad_norm": 0.466796875, + "learning_rate": 2.2266973575432627e-05, + "loss": 2.4722, + "step": 36791 + }, + { + "epoch": 1.57, + "grad_norm": 0.51171875, + "learning_rate": 2.226275383937504e-05, + "loss": 2.4728, + "step": 36792 + }, + { + "epoch": 1.57, + "grad_norm": 0.43359375, + "learning_rate": 2.2258534453106662e-05, + "loss": 2.466, + "step": 36793 + }, + { + "epoch": 1.57, + "grad_norm": 0.474609375, + "learning_rate": 2.2254315416646487e-05, + "loss": 2.58, + "step": 36794 + }, + { + "epoch": 1.57, + "grad_norm": 0.490234375, + "learning_rate": 2.2250096730013446e-05, + "loss": 2.4577, + "step": 36795 + }, + { + "epoch": 1.57, + "grad_norm": 0.466796875, + "learning_rate": 2.2245878393226592e-05, + "loss": 2.4439, + "step": 36796 + }, + { + "epoch": 1.57, + "grad_norm": 0.482421875, + "learning_rate": 2.2241660406304844e-05, + "loss": 2.3954, + "step": 36797 + }, + { + "epoch": 1.57, + "grad_norm": 0.466796875, + "learning_rate": 2.22374427692672e-05, + "loss": 2.3732, + "step": 36798 + }, + { + "epoch": 1.57, + "grad_norm": 0.4921875, + "learning_rate": 2.223322548213268e-05, + "loss": 2.5273, + "step": 36799 + }, + { + "epoch": 1.57, + "grad_norm": 0.45703125, + "learning_rate": 2.2229008544920217e-05, + "loss": 2.4212, + "step": 36800 + }, + { + "epoch": 1.57, + "grad_norm": 0.5234375, + "learning_rate": 2.2224791957648772e-05, + "loss": 2.6474, + "step": 36801 + }, + { + "epoch": 1.57, + "grad_norm": 0.466796875, + "learning_rate": 2.2220575720337334e-05, + "loss": 2.3886, + "step": 36802 + }, + { + "epoch": 1.57, + "grad_norm": 0.4765625, + "learning_rate": 2.2216359833004907e-05, + "loss": 2.4598, + "step": 36803 + }, + { + "epoch": 1.57, + "grad_norm": 0.451171875, + "learning_rate": 2.221214429567041e-05, + "loss": 2.5021, + "step": 36804 + }, + { + "epoch": 1.57, + "grad_norm": 0.447265625, + "learning_rate": 2.220792910835284e-05, + "loss": 2.3721, + "step": 36805 + }, + { + "epoch": 1.57, + "grad_norm": 0.494140625, + "learning_rate": 2.2203714271071184e-05, + "loss": 2.4108, + "step": 36806 + }, + { + "epoch": 1.57, + "grad_norm": 0.494140625, + "learning_rate": 2.219949978384438e-05, + "loss": 2.4657, + "step": 36807 + }, + { + "epoch": 1.57, + "grad_norm": 0.4921875, + "learning_rate": 2.2195285646691366e-05, + "loss": 2.4994, + "step": 36808 + }, + { + "epoch": 1.57, + "grad_norm": 0.470703125, + "learning_rate": 2.2191071859631128e-05, + "loss": 2.3717, + "step": 36809 + }, + { + "epoch": 1.57, + "grad_norm": 0.447265625, + "learning_rate": 2.2186858422682666e-05, + "loss": 2.4044, + "step": 36810 + }, + { + "epoch": 1.57, + "grad_norm": 0.474609375, + "learning_rate": 2.218264533586487e-05, + "loss": 2.4459, + "step": 36811 + }, + { + "epoch": 1.57, + "grad_norm": 0.50390625, + "learning_rate": 2.2178432599196764e-05, + "loss": 2.4702, + "step": 36812 + }, + { + "epoch": 1.57, + "grad_norm": 0.4921875, + "learning_rate": 2.2174220212697237e-05, + "loss": 2.4915, + "step": 36813 + }, + { + "epoch": 1.57, + "grad_norm": 0.45703125, + "learning_rate": 2.2170008176385292e-05, + "loss": 2.4953, + "step": 36814 + }, + { + "epoch": 1.57, + "grad_norm": 0.4765625, + "learning_rate": 2.2165796490279878e-05, + "loss": 2.3165, + "step": 36815 + }, + { + "epoch": 1.57, + "grad_norm": 0.52734375, + "learning_rate": 2.2161585154399944e-05, + "loss": 2.4385, + "step": 36816 + }, + { + "epoch": 1.57, + "grad_norm": 0.443359375, + "learning_rate": 2.21573741687644e-05, + "loss": 2.5474, + "step": 36817 + }, + { + "epoch": 1.57, + "grad_norm": 0.484375, + "learning_rate": 2.2153163533392217e-05, + "loss": 2.4418, + "step": 36818 + }, + { + "epoch": 1.57, + "grad_norm": 0.443359375, + "learning_rate": 2.2148953248302385e-05, + "loss": 2.4043, + "step": 36819 + }, + { + "epoch": 1.57, + "grad_norm": 0.47265625, + "learning_rate": 2.214474331351378e-05, + "loss": 2.433, + "step": 36820 + }, + { + "epoch": 1.57, + "grad_norm": 0.41796875, + "learning_rate": 2.2140533729045377e-05, + "loss": 2.4679, + "step": 36821 + }, + { + "epoch": 1.57, + "grad_norm": 0.48828125, + "learning_rate": 2.213632449491615e-05, + "loss": 2.404, + "step": 36822 + }, + { + "epoch": 1.57, + "grad_norm": 0.37890625, + "learning_rate": 2.2132115611145e-05, + "loss": 2.3546, + "step": 36823 + }, + { + "epoch": 1.57, + "grad_norm": 0.46484375, + "learning_rate": 2.2127907077750853e-05, + "loss": 2.4405, + "step": 36824 + }, + { + "epoch": 1.57, + "grad_norm": 0.50390625, + "learning_rate": 2.2123698894752653e-05, + "loss": 2.4599, + "step": 36825 + }, + { + "epoch": 1.57, + "grad_norm": 0.5078125, + "learning_rate": 2.2119491062169385e-05, + "loss": 2.3932, + "step": 36826 + }, + { + "epoch": 1.57, + "grad_norm": 0.4609375, + "learning_rate": 2.2115283580019917e-05, + "loss": 2.3996, + "step": 36827 + }, + { + "epoch": 1.57, + "grad_norm": 0.50390625, + "learning_rate": 2.2111076448323243e-05, + "loss": 2.4537, + "step": 36828 + }, + { + "epoch": 1.57, + "grad_norm": 0.4921875, + "learning_rate": 2.2106869667098218e-05, + "loss": 2.5977, + "step": 36829 + }, + { + "epoch": 1.57, + "grad_norm": 0.48828125, + "learning_rate": 2.210266323636382e-05, + "loss": 2.6064, + "step": 36830 + }, + { + "epoch": 1.57, + "grad_norm": 0.478515625, + "learning_rate": 2.2098457156138995e-05, + "loss": 2.4054, + "step": 36831 + }, + { + "epoch": 1.57, + "grad_norm": 0.4453125, + "learning_rate": 2.2094251426442646e-05, + "loss": 2.3781, + "step": 36832 + }, + { + "epoch": 1.57, + "grad_norm": 0.4453125, + "learning_rate": 2.2090046047293667e-05, + "loss": 2.3887, + "step": 36833 + }, + { + "epoch": 1.57, + "grad_norm": 0.41796875, + "learning_rate": 2.2085841018711005e-05, + "loss": 2.4201, + "step": 36834 + }, + { + "epoch": 1.57, + "grad_norm": 0.439453125, + "learning_rate": 2.2081636340713606e-05, + "loss": 2.5718, + "step": 36835 + }, + { + "epoch": 1.57, + "grad_norm": 0.439453125, + "learning_rate": 2.2077432013320353e-05, + "loss": 2.4377, + "step": 36836 + }, + { + "epoch": 1.57, + "grad_norm": 0.42578125, + "learning_rate": 2.2073228036550163e-05, + "loss": 2.3548, + "step": 36837 + }, + { + "epoch": 1.57, + "grad_norm": 0.48046875, + "learning_rate": 2.2069024410422012e-05, + "loss": 2.4854, + "step": 36838 + }, + { + "epoch": 1.57, + "grad_norm": 0.462890625, + "learning_rate": 2.206482113495476e-05, + "loss": 2.4703, + "step": 36839 + }, + { + "epoch": 1.57, + "grad_norm": 0.5390625, + "learning_rate": 2.2060618210167305e-05, + "loss": 2.4778, + "step": 36840 + }, + { + "epoch": 1.57, + "grad_norm": 0.486328125, + "learning_rate": 2.2056415636078588e-05, + "loss": 2.4084, + "step": 36841 + }, + { + "epoch": 1.57, + "grad_norm": 0.44921875, + "learning_rate": 2.2052213412707535e-05, + "loss": 2.4609, + "step": 36842 + }, + { + "epoch": 1.57, + "grad_norm": 0.55859375, + "learning_rate": 2.204801154007302e-05, + "loss": 2.4106, + "step": 36843 + }, + { + "epoch": 1.57, + "grad_norm": 0.435546875, + "learning_rate": 2.2043810018193978e-05, + "loss": 2.4355, + "step": 36844 + }, + { + "epoch": 1.57, + "grad_norm": 0.462890625, + "learning_rate": 2.203960884708929e-05, + "loss": 2.3482, + "step": 36845 + }, + { + "epoch": 1.57, + "grad_norm": 0.462890625, + "learning_rate": 2.203540802677786e-05, + "loss": 2.4478, + "step": 36846 + }, + { + "epoch": 1.57, + "grad_norm": 0.45703125, + "learning_rate": 2.2031207557278632e-05, + "loss": 2.3579, + "step": 36847 + }, + { + "epoch": 1.57, + "grad_norm": 0.48046875, + "learning_rate": 2.202700743861047e-05, + "loss": 2.4484, + "step": 36848 + }, + { + "epoch": 1.57, + "grad_norm": 0.515625, + "learning_rate": 2.2022807670792233e-05, + "loss": 2.6009, + "step": 36849 + }, + { + "epoch": 1.57, + "grad_norm": 0.46875, + "learning_rate": 2.201860825384291e-05, + "loss": 2.4136, + "step": 36850 + }, + { + "epoch": 1.57, + "grad_norm": 0.59765625, + "learning_rate": 2.201440918778136e-05, + "loss": 2.5722, + "step": 36851 + }, + { + "epoch": 1.57, + "grad_norm": 0.48828125, + "learning_rate": 2.2010210472626436e-05, + "loss": 2.6259, + "step": 36852 + }, + { + "epoch": 1.57, + "grad_norm": 0.47265625, + "learning_rate": 2.2006012108397054e-05, + "loss": 2.4929, + "step": 36853 + }, + { + "epoch": 1.57, + "grad_norm": 0.392578125, + "learning_rate": 2.200181409511215e-05, + "loss": 2.443, + "step": 36854 + }, + { + "epoch": 1.57, + "grad_norm": 0.50390625, + "learning_rate": 2.199761643279058e-05, + "loss": 2.4444, + "step": 36855 + }, + { + "epoch": 1.57, + "grad_norm": 0.46484375, + "learning_rate": 2.1993419121451196e-05, + "loss": 2.43, + "step": 36856 + }, + { + "epoch": 1.57, + "grad_norm": 0.47265625, + "learning_rate": 2.198922216111292e-05, + "loss": 2.5438, + "step": 36857 + }, + { + "epoch": 1.57, + "grad_norm": 0.53125, + "learning_rate": 2.198502555179467e-05, + "loss": 2.6859, + "step": 36858 + }, + { + "epoch": 1.57, + "grad_norm": 0.462890625, + "learning_rate": 2.198082929351526e-05, + "loss": 2.5031, + "step": 36859 + }, + { + "epoch": 1.57, + "grad_norm": 0.50390625, + "learning_rate": 2.1976633386293634e-05, + "loss": 2.4789, + "step": 36860 + }, + { + "epoch": 1.57, + "grad_norm": 0.421875, + "learning_rate": 2.1972437830148607e-05, + "loss": 2.3038, + "step": 36861 + }, + { + "epoch": 1.57, + "grad_norm": 0.482421875, + "learning_rate": 2.19682426250991e-05, + "loss": 2.3628, + "step": 36862 + }, + { + "epoch": 1.57, + "grad_norm": 0.47265625, + "learning_rate": 2.196404777116402e-05, + "loss": 2.4352, + "step": 36863 + }, + { + "epoch": 1.57, + "grad_norm": 0.466796875, + "learning_rate": 2.1959853268362197e-05, + "loss": 2.4664, + "step": 36864 + }, + { + "epoch": 1.57, + "grad_norm": 0.46484375, + "learning_rate": 2.1955659116712492e-05, + "loss": 2.3875, + "step": 36865 + }, + { + "epoch": 1.57, + "grad_norm": 0.443359375, + "learning_rate": 2.1951465316233787e-05, + "loss": 2.4576, + "step": 36866 + }, + { + "epoch": 1.57, + "grad_norm": 0.5390625, + "learning_rate": 2.1947271866945008e-05, + "loss": 2.5348, + "step": 36867 + }, + { + "epoch": 1.57, + "grad_norm": 0.4609375, + "learning_rate": 2.1943078768864945e-05, + "loss": 2.3485, + "step": 36868 + }, + { + "epoch": 1.57, + "grad_norm": 0.48828125, + "learning_rate": 2.193888602201253e-05, + "loss": 2.3591, + "step": 36869 + }, + { + "epoch": 1.57, + "grad_norm": 0.55078125, + "learning_rate": 2.1934693626406566e-05, + "loss": 2.583, + "step": 36870 + }, + { + "epoch": 1.57, + "grad_norm": 0.51171875, + "learning_rate": 2.1930501582065987e-05, + "loss": 2.5032, + "step": 36871 + }, + { + "epoch": 1.57, + "grad_norm": 0.45703125, + "learning_rate": 2.1926309889009588e-05, + "loss": 2.3724, + "step": 36872 + }, + { + "epoch": 1.57, + "grad_norm": 0.44921875, + "learning_rate": 2.192211854725629e-05, + "loss": 2.4218, + "step": 36873 + }, + { + "epoch": 1.57, + "grad_norm": 0.484375, + "learning_rate": 2.1917927556824904e-05, + "loss": 2.4813, + "step": 36874 + }, + { + "epoch": 1.57, + "grad_norm": 0.474609375, + "learning_rate": 2.1913736917734294e-05, + "loss": 2.4027, + "step": 36875 + }, + { + "epoch": 1.57, + "grad_norm": 0.48828125, + "learning_rate": 2.1909546630003364e-05, + "loss": 2.4435, + "step": 36876 + }, + { + "epoch": 1.57, + "grad_norm": 0.609375, + "learning_rate": 2.1905356693650946e-05, + "loss": 2.5092, + "step": 36877 + }, + { + "epoch": 1.57, + "grad_norm": 0.48828125, + "learning_rate": 2.190116710869582e-05, + "loss": 2.4895, + "step": 36878 + }, + { + "epoch": 1.57, + "grad_norm": 0.45703125, + "learning_rate": 2.1896977875156943e-05, + "loss": 2.4237, + "step": 36879 + }, + { + "epoch": 1.57, + "grad_norm": 0.47265625, + "learning_rate": 2.189278899305314e-05, + "loss": 2.4364, + "step": 36880 + }, + { + "epoch": 1.57, + "grad_norm": 0.431640625, + "learning_rate": 2.188860046240321e-05, + "loss": 2.4875, + "step": 36881 + }, + { + "epoch": 1.57, + "grad_norm": 0.486328125, + "learning_rate": 2.1884412283226018e-05, + "loss": 2.4787, + "step": 36882 + }, + { + "epoch": 1.57, + "grad_norm": 0.451171875, + "learning_rate": 2.1880224455540465e-05, + "loss": 2.4407, + "step": 36883 + }, + { + "epoch": 1.57, + "grad_norm": 0.439453125, + "learning_rate": 2.1876036979365322e-05, + "loss": 2.425, + "step": 36884 + }, + { + "epoch": 1.57, + "grad_norm": 0.46875, + "learning_rate": 2.1871849854719484e-05, + "loss": 2.589, + "step": 36885 + }, + { + "epoch": 1.57, + "grad_norm": 0.4375, + "learning_rate": 2.1867663081621748e-05, + "loss": 2.5349, + "step": 36886 + }, + { + "epoch": 1.57, + "grad_norm": 0.474609375, + "learning_rate": 2.1863476660090996e-05, + "loss": 2.3989, + "step": 36887 + }, + { + "epoch": 1.57, + "grad_norm": 0.51953125, + "learning_rate": 2.185929059014602e-05, + "loss": 2.4746, + "step": 36888 + }, + { + "epoch": 1.57, + "grad_norm": 0.48828125, + "learning_rate": 2.185510487180571e-05, + "loss": 2.3198, + "step": 36889 + }, + { + "epoch": 1.57, + "grad_norm": 0.47265625, + "learning_rate": 2.185091950508884e-05, + "loss": 2.4684, + "step": 36890 + }, + { + "epoch": 1.57, + "grad_norm": 0.4921875, + "learning_rate": 2.1846734490014264e-05, + "loss": 2.5252, + "step": 36891 + }, + { + "epoch": 1.57, + "grad_norm": 0.515625, + "learning_rate": 2.1842549826600856e-05, + "loss": 2.4331, + "step": 36892 + }, + { + "epoch": 1.57, + "grad_norm": 0.478515625, + "learning_rate": 2.183836551486741e-05, + "loss": 2.4663, + "step": 36893 + }, + { + "epoch": 1.57, + "grad_norm": 0.427734375, + "learning_rate": 2.18341815548327e-05, + "loss": 2.5321, + "step": 36894 + }, + { + "epoch": 1.57, + "grad_norm": 0.46484375, + "learning_rate": 2.1829997946515657e-05, + "loss": 2.3811, + "step": 36895 + }, + { + "epoch": 1.57, + "grad_norm": 0.455078125, + "learning_rate": 2.1825814689935053e-05, + "loss": 2.45, + "step": 36896 + }, + { + "epoch": 1.57, + "grad_norm": 0.494140625, + "learning_rate": 2.18216317851097e-05, + "loss": 2.5191, + "step": 36897 + }, + { + "epoch": 1.57, + "grad_norm": 0.46875, + "learning_rate": 2.1817449232058418e-05, + "loss": 2.2844, + "step": 36898 + }, + { + "epoch": 1.57, + "grad_norm": 0.486328125, + "learning_rate": 2.1813267030800076e-05, + "loss": 2.4081, + "step": 36899 + }, + { + "epoch": 1.57, + "grad_norm": 0.51171875, + "learning_rate": 2.1809085181353427e-05, + "loss": 2.4759, + "step": 36900 + }, + { + "epoch": 1.57, + "grad_norm": 0.46484375, + "learning_rate": 2.1804903683737355e-05, + "loss": 2.3269, + "step": 36901 + }, + { + "epoch": 1.57, + "grad_norm": 0.484375, + "learning_rate": 2.1800722537970608e-05, + "loss": 2.5171, + "step": 36902 + }, + { + "epoch": 1.57, + "grad_norm": 0.482421875, + "learning_rate": 2.179654174407203e-05, + "loss": 2.5316, + "step": 36903 + }, + { + "epoch": 1.57, + "grad_norm": 0.443359375, + "learning_rate": 2.179236130206046e-05, + "loss": 2.395, + "step": 36904 + }, + { + "epoch": 1.57, + "grad_norm": 0.474609375, + "learning_rate": 2.1788181211954685e-05, + "loss": 2.5672, + "step": 36905 + }, + { + "epoch": 1.57, + "grad_norm": 0.51953125, + "learning_rate": 2.1784001473773486e-05, + "loss": 2.5245, + "step": 36906 + }, + { + "epoch": 1.57, + "grad_norm": 0.470703125, + "learning_rate": 2.1779822087535683e-05, + "loss": 2.4519, + "step": 36907 + }, + { + "epoch": 1.57, + "grad_norm": 0.43359375, + "learning_rate": 2.177564305326013e-05, + "loss": 2.4101, + "step": 36908 + }, + { + "epoch": 1.57, + "grad_norm": 0.439453125, + "learning_rate": 2.17714643709656e-05, + "loss": 2.4336, + "step": 36909 + }, + { + "epoch": 1.57, + "grad_norm": 0.494140625, + "learning_rate": 2.176728604067083e-05, + "loss": 2.5877, + "step": 36910 + }, + { + "epoch": 1.57, + "grad_norm": 0.4609375, + "learning_rate": 2.1763108062394744e-05, + "loss": 2.4465, + "step": 36911 + }, + { + "epoch": 1.57, + "grad_norm": 0.45703125, + "learning_rate": 2.1758930436156077e-05, + "loss": 2.3505, + "step": 36912 + }, + { + "epoch": 1.57, + "grad_norm": 0.4609375, + "learning_rate": 2.1754753161973596e-05, + "loss": 2.4635, + "step": 36913 + }, + { + "epoch": 1.57, + "grad_norm": 0.4296875, + "learning_rate": 2.175057623986614e-05, + "loss": 2.5398, + "step": 36914 + }, + { + "epoch": 1.57, + "grad_norm": 0.43359375, + "learning_rate": 2.1746399669852513e-05, + "loss": 2.4692, + "step": 36915 + }, + { + "epoch": 1.57, + "grad_norm": 0.458984375, + "learning_rate": 2.174222345195147e-05, + "loss": 2.4472, + "step": 36916 + }, + { + "epoch": 1.57, + "grad_norm": 0.43359375, + "learning_rate": 2.1738047586181842e-05, + "loss": 2.5351, + "step": 36917 + }, + { + "epoch": 1.57, + "grad_norm": 0.44140625, + "learning_rate": 2.1733872072562367e-05, + "loss": 2.3581, + "step": 36918 + }, + { + "epoch": 1.57, + "grad_norm": 0.5546875, + "learning_rate": 2.1729696911111873e-05, + "loss": 2.5148, + "step": 36919 + }, + { + "epoch": 1.57, + "grad_norm": 0.486328125, + "learning_rate": 2.172552210184916e-05, + "loss": 2.3934, + "step": 36920 + }, + { + "epoch": 1.57, + "grad_norm": 0.4453125, + "learning_rate": 2.1721347644793e-05, + "loss": 2.418, + "step": 36921 + }, + { + "epoch": 1.57, + "grad_norm": 0.421875, + "learning_rate": 2.1717173539962134e-05, + "loss": 2.5115, + "step": 36922 + }, + { + "epoch": 1.57, + "grad_norm": 0.42578125, + "learning_rate": 2.1712999787375378e-05, + "loss": 2.4318, + "step": 36923 + }, + { + "epoch": 1.57, + "grad_norm": 0.474609375, + "learning_rate": 2.1708826387051538e-05, + "loss": 2.3596, + "step": 36924 + }, + { + "epoch": 1.57, + "grad_norm": 0.423828125, + "learning_rate": 2.1704653339009373e-05, + "loss": 2.4449, + "step": 36925 + }, + { + "epoch": 1.57, + "grad_norm": 0.451171875, + "learning_rate": 2.1700480643267597e-05, + "loss": 2.353, + "step": 36926 + }, + { + "epoch": 1.57, + "grad_norm": 0.48046875, + "learning_rate": 2.1696308299845102e-05, + "loss": 2.4854, + "step": 36927 + }, + { + "epoch": 1.57, + "grad_norm": 0.515625, + "learning_rate": 2.1692136308760593e-05, + "loss": 2.3594, + "step": 36928 + }, + { + "epoch": 1.57, + "grad_norm": 0.447265625, + "learning_rate": 2.1687964670032834e-05, + "loss": 2.4738, + "step": 36929 + }, + { + "epoch": 1.57, + "grad_norm": 0.48046875, + "learning_rate": 2.1683793383680607e-05, + "loss": 2.384, + "step": 36930 + }, + { + "epoch": 1.57, + "grad_norm": 0.47265625, + "learning_rate": 2.1679622449722713e-05, + "loss": 2.5379, + "step": 36931 + }, + { + "epoch": 1.57, + "grad_norm": 0.4453125, + "learning_rate": 2.1675451868177877e-05, + "loss": 2.5854, + "step": 36932 + }, + { + "epoch": 1.57, + "grad_norm": 0.490234375, + "learning_rate": 2.167128163906491e-05, + "loss": 2.5328, + "step": 36933 + }, + { + "epoch": 1.57, + "grad_norm": 0.4765625, + "learning_rate": 2.1667111762402526e-05, + "loss": 2.5079, + "step": 36934 + }, + { + "epoch": 1.57, + "grad_norm": 0.55859375, + "learning_rate": 2.1662942238209504e-05, + "loss": 2.4268, + "step": 36935 + }, + { + "epoch": 1.57, + "grad_norm": 0.427734375, + "learning_rate": 2.1658773066504636e-05, + "loss": 2.4263, + "step": 36936 + }, + { + "epoch": 1.57, + "grad_norm": 0.5078125, + "learning_rate": 2.1654604247306663e-05, + "loss": 2.3679, + "step": 36937 + }, + { + "epoch": 1.57, + "grad_norm": 0.53125, + "learning_rate": 2.1650435780634313e-05, + "loss": 2.4739, + "step": 36938 + }, + { + "epoch": 1.57, + "grad_norm": 0.53515625, + "learning_rate": 2.164626766650637e-05, + "loss": 2.5423, + "step": 36939 + }, + { + "epoch": 1.57, + "grad_norm": 0.466796875, + "learning_rate": 2.1642099904941616e-05, + "loss": 2.5239, + "step": 36940 + }, + { + "epoch": 1.57, + "grad_norm": 0.478515625, + "learning_rate": 2.163793249595878e-05, + "loss": 2.3616, + "step": 36941 + }, + { + "epoch": 1.57, + "grad_norm": 0.44140625, + "learning_rate": 2.1633765439576547e-05, + "loss": 2.3769, + "step": 36942 + }, + { + "epoch": 1.57, + "grad_norm": 0.46875, + "learning_rate": 2.1629598735813782e-05, + "loss": 2.4625, + "step": 36943 + }, + { + "epoch": 1.57, + "grad_norm": 0.4609375, + "learning_rate": 2.162543238468918e-05, + "loss": 2.3242, + "step": 36944 + }, + { + "epoch": 1.57, + "grad_norm": 0.453125, + "learning_rate": 2.1621266386221474e-05, + "loss": 2.4416, + "step": 36945 + }, + { + "epoch": 1.57, + "grad_norm": 0.455078125, + "learning_rate": 2.161710074042942e-05, + "loss": 2.4219, + "step": 36946 + }, + { + "epoch": 1.57, + "grad_norm": 0.482421875, + "learning_rate": 2.1612935447331794e-05, + "loss": 2.5653, + "step": 36947 + }, + { + "epoch": 1.57, + "grad_norm": 0.451171875, + "learning_rate": 2.1608770506947284e-05, + "loss": 2.4655, + "step": 36948 + }, + { + "epoch": 1.57, + "grad_norm": 0.484375, + "learning_rate": 2.1604605919294683e-05, + "loss": 2.4886, + "step": 36949 + }, + { + "epoch": 1.57, + "grad_norm": 0.46484375, + "learning_rate": 2.160044168439268e-05, + "loss": 2.5079, + "step": 36950 + }, + { + "epoch": 1.57, + "grad_norm": 0.478515625, + "learning_rate": 2.1596277802260035e-05, + "loss": 2.5266, + "step": 36951 + }, + { + "epoch": 1.57, + "grad_norm": 0.458984375, + "learning_rate": 2.159211427291552e-05, + "loss": 2.4346, + "step": 36952 + }, + { + "epoch": 1.57, + "grad_norm": 0.53125, + "learning_rate": 2.1587951096377835e-05, + "loss": 2.512, + "step": 36953 + }, + { + "epoch": 1.57, + "grad_norm": 0.4453125, + "learning_rate": 2.158378827266567e-05, + "loss": 2.363, + "step": 36954 + }, + { + "epoch": 1.57, + "grad_norm": 0.43359375, + "learning_rate": 2.157962580179782e-05, + "loss": 2.4454, + "step": 36955 + }, + { + "epoch": 1.57, + "grad_norm": 0.44140625, + "learning_rate": 2.1575463683793018e-05, + "loss": 2.4501, + "step": 36956 + }, + { + "epoch": 1.57, + "grad_norm": 0.5, + "learning_rate": 2.157130191866994e-05, + "loss": 2.5666, + "step": 36957 + }, + { + "epoch": 1.57, + "grad_norm": 0.44140625, + "learning_rate": 2.1567140506447337e-05, + "loss": 2.414, + "step": 36958 + }, + { + "epoch": 1.57, + "grad_norm": 0.53125, + "learning_rate": 2.1562979447143972e-05, + "loss": 2.6276, + "step": 36959 + }, + { + "epoch": 1.57, + "grad_norm": 0.462890625, + "learning_rate": 2.155881874077853e-05, + "loss": 2.5639, + "step": 36960 + }, + { + "epoch": 1.58, + "grad_norm": 0.455078125, + "learning_rate": 2.155465838736971e-05, + "loss": 2.4328, + "step": 36961 + }, + { + "epoch": 1.58, + "grad_norm": 0.453125, + "learning_rate": 2.1550498386936245e-05, + "loss": 2.3934, + "step": 36962 + }, + { + "epoch": 1.58, + "grad_norm": 0.4453125, + "learning_rate": 2.154633873949692e-05, + "loss": 2.459, + "step": 36963 + }, + { + "epoch": 1.58, + "grad_norm": 0.48828125, + "learning_rate": 2.1542179445070353e-05, + "loss": 2.416, + "step": 36964 + }, + { + "epoch": 1.58, + "grad_norm": 0.478515625, + "learning_rate": 2.153802050367534e-05, + "loss": 2.4185, + "step": 36965 + }, + { + "epoch": 1.58, + "grad_norm": 0.4296875, + "learning_rate": 2.1533861915330532e-05, + "loss": 2.5328, + "step": 36966 + }, + { + "epoch": 1.58, + "grad_norm": 0.4921875, + "learning_rate": 2.152970368005467e-05, + "loss": 2.6142, + "step": 36967 + }, + { + "epoch": 1.58, + "grad_norm": 0.48046875, + "learning_rate": 2.1525545797866485e-05, + "loss": 2.4451, + "step": 36968 + }, + { + "epoch": 1.58, + "grad_norm": 0.5234375, + "learning_rate": 2.152138826878467e-05, + "loss": 2.3898, + "step": 36969 + }, + { + "epoch": 1.58, + "grad_norm": 0.52734375, + "learning_rate": 2.1517231092827905e-05, + "loss": 2.5266, + "step": 36970 + }, + { + "epoch": 1.58, + "grad_norm": 0.49609375, + "learning_rate": 2.151307427001491e-05, + "loss": 2.4141, + "step": 36971 + }, + { + "epoch": 1.58, + "grad_norm": 0.427734375, + "learning_rate": 2.1508917800364427e-05, + "loss": 2.4527, + "step": 36972 + }, + { + "epoch": 1.58, + "grad_norm": 0.431640625, + "learning_rate": 2.1504761683895104e-05, + "loss": 2.3943, + "step": 36973 + }, + { + "epoch": 1.58, + "grad_norm": 0.45703125, + "learning_rate": 2.150060592062566e-05, + "loss": 2.5396, + "step": 36974 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.1496450510574838e-05, + "loss": 2.3426, + "step": 36975 + }, + { + "epoch": 1.58, + "grad_norm": 0.4453125, + "learning_rate": 2.149229545376129e-05, + "loss": 2.4952, + "step": 36976 + }, + { + "epoch": 1.58, + "grad_norm": 0.46484375, + "learning_rate": 2.14881407502037e-05, + "loss": 2.5336, + "step": 36977 + }, + { + "epoch": 1.58, + "grad_norm": 0.46875, + "learning_rate": 2.1483986399920773e-05, + "loss": 2.5111, + "step": 36978 + }, + { + "epoch": 1.58, + "grad_norm": 0.52734375, + "learning_rate": 2.1479832402931254e-05, + "loss": 2.4819, + "step": 36979 + }, + { + "epoch": 1.58, + "grad_norm": 0.388671875, + "learning_rate": 2.147567875925376e-05, + "loss": 2.3458, + "step": 36980 + }, + { + "epoch": 1.58, + "grad_norm": 0.4453125, + "learning_rate": 2.147152546890704e-05, + "loss": 2.356, + "step": 36981 + }, + { + "epoch": 1.58, + "grad_norm": 0.486328125, + "learning_rate": 2.1467372531909735e-05, + "loss": 2.3999, + "step": 36982 + }, + { + "epoch": 1.58, + "grad_norm": 0.396484375, + "learning_rate": 2.1463219948280544e-05, + "loss": 2.4419, + "step": 36983 + }, + { + "epoch": 1.58, + "grad_norm": 0.40234375, + "learning_rate": 2.1459067718038194e-05, + "loss": 2.34, + "step": 36984 + }, + { + "epoch": 1.58, + "grad_norm": 0.41796875, + "learning_rate": 2.145491584120134e-05, + "loss": 2.3586, + "step": 36985 + }, + { + "epoch": 1.58, + "grad_norm": 0.43359375, + "learning_rate": 2.1450764317788628e-05, + "loss": 2.3752, + "step": 36986 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.1446613147818774e-05, + "loss": 2.4268, + "step": 36987 + }, + { + "epoch": 1.58, + "grad_norm": 0.5, + "learning_rate": 2.1442462331310476e-05, + "loss": 2.389, + "step": 36988 + }, + { + "epoch": 1.58, + "grad_norm": 0.51171875, + "learning_rate": 2.1438311868282356e-05, + "loss": 2.5785, + "step": 36989 + }, + { + "epoch": 1.58, + "grad_norm": 0.41796875, + "learning_rate": 2.1434161758753156e-05, + "loss": 2.5241, + "step": 36990 + }, + { + "epoch": 1.58, + "grad_norm": 0.4609375, + "learning_rate": 2.1430012002741486e-05, + "loss": 2.5111, + "step": 36991 + }, + { + "epoch": 1.58, + "grad_norm": 0.498046875, + "learning_rate": 2.1425862600266077e-05, + "loss": 2.5743, + "step": 36992 + }, + { + "epoch": 1.58, + "grad_norm": 0.486328125, + "learning_rate": 2.1421713551345547e-05, + "loss": 2.3795, + "step": 36993 + }, + { + "epoch": 1.58, + "grad_norm": 0.4765625, + "learning_rate": 2.1417564855998606e-05, + "loss": 2.469, + "step": 36994 + }, + { + "epoch": 1.58, + "grad_norm": 0.458984375, + "learning_rate": 2.1413416514243888e-05, + "loss": 2.4268, + "step": 36995 + }, + { + "epoch": 1.58, + "grad_norm": 0.482421875, + "learning_rate": 2.140926852610007e-05, + "loss": 2.4376, + "step": 36996 + }, + { + "epoch": 1.58, + "grad_norm": 0.49609375, + "learning_rate": 2.140512089158585e-05, + "loss": 2.4731, + "step": 36997 + }, + { + "epoch": 1.58, + "grad_norm": 0.494140625, + "learning_rate": 2.1400973610719856e-05, + "loss": 2.4704, + "step": 36998 + }, + { + "epoch": 1.58, + "grad_norm": 0.490234375, + "learning_rate": 2.139682668352071e-05, + "loss": 2.5209, + "step": 36999 + }, + { + "epoch": 1.58, + "grad_norm": 0.48828125, + "learning_rate": 2.139268011000717e-05, + "loss": 2.4485, + "step": 37000 + }, + { + "epoch": 1.58, + "grad_norm": 0.46875, + "learning_rate": 2.1388533890197838e-05, + "loss": 2.4338, + "step": 37001 + }, + { + "epoch": 1.58, + "grad_norm": 0.48046875, + "learning_rate": 2.1384388024111357e-05, + "loss": 2.4908, + "step": 37002 + }, + { + "epoch": 1.58, + "grad_norm": 0.46484375, + "learning_rate": 2.138024251176639e-05, + "loss": 2.4584, + "step": 37003 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.1376097353181635e-05, + "loss": 2.4722, + "step": 37004 + }, + { + "epoch": 1.58, + "grad_norm": 0.5703125, + "learning_rate": 2.1371952548375673e-05, + "loss": 2.4863, + "step": 37005 + }, + { + "epoch": 1.58, + "grad_norm": 0.52734375, + "learning_rate": 2.1367808097367214e-05, + "loss": 2.5764, + "step": 37006 + }, + { + "epoch": 1.58, + "grad_norm": 0.443359375, + "learning_rate": 2.1363664000174867e-05, + "loss": 2.463, + "step": 37007 + }, + { + "epoch": 1.58, + "grad_norm": 0.4921875, + "learning_rate": 2.1359520256817312e-05, + "loss": 2.3773, + "step": 37008 + }, + { + "epoch": 1.58, + "grad_norm": 0.48828125, + "learning_rate": 2.135537686731315e-05, + "loss": 2.36, + "step": 37009 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.1351233831681085e-05, + "loss": 2.4316, + "step": 37010 + }, + { + "epoch": 1.58, + "grad_norm": 0.484375, + "learning_rate": 2.1347091149939678e-05, + "loss": 2.4122, + "step": 37011 + }, + { + "epoch": 1.58, + "grad_norm": 0.462890625, + "learning_rate": 2.1342948822107634e-05, + "loss": 2.4939, + "step": 37012 + }, + { + "epoch": 1.58, + "grad_norm": 0.5, + "learning_rate": 2.1338806848203597e-05, + "loss": 2.4094, + "step": 37013 + }, + { + "epoch": 1.58, + "grad_norm": 0.5234375, + "learning_rate": 2.1334665228246186e-05, + "loss": 2.5006, + "step": 37014 + }, + { + "epoch": 1.58, + "grad_norm": 0.39453125, + "learning_rate": 2.133052396225398e-05, + "loss": 2.4402, + "step": 37015 + }, + { + "epoch": 1.58, + "grad_norm": 0.46875, + "learning_rate": 2.1326383050245713e-05, + "loss": 2.5125, + "step": 37016 + }, + { + "epoch": 1.58, + "grad_norm": 0.423828125, + "learning_rate": 2.1322242492239962e-05, + "loss": 2.3425, + "step": 37017 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.131810228825536e-05, + "loss": 2.4753, + "step": 37018 + }, + { + "epoch": 1.58, + "grad_norm": 0.484375, + "learning_rate": 2.131396243831053e-05, + "loss": 2.4308, + "step": 37019 + }, + { + "epoch": 1.58, + "grad_norm": 0.4296875, + "learning_rate": 2.130982294242414e-05, + "loss": 2.4281, + "step": 37020 + }, + { + "epoch": 1.58, + "grad_norm": 0.46484375, + "learning_rate": 2.130568380061476e-05, + "loss": 2.3502, + "step": 37021 + }, + { + "epoch": 1.58, + "grad_norm": 0.45703125, + "learning_rate": 2.130154501290108e-05, + "loss": 2.4541, + "step": 37022 + }, + { + "epoch": 1.58, + "grad_norm": 0.453125, + "learning_rate": 2.129740657930165e-05, + "loss": 2.3182, + "step": 37023 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.1293268499835163e-05, + "loss": 2.4171, + "step": 37024 + }, + { + "epoch": 1.58, + "grad_norm": 0.443359375, + "learning_rate": 2.1289130774520162e-05, + "loss": 2.294, + "step": 37025 + }, + { + "epoch": 1.58, + "grad_norm": 0.51953125, + "learning_rate": 2.1284993403375343e-05, + "loss": 2.4187, + "step": 37026 + }, + { + "epoch": 1.58, + "grad_norm": 0.5078125, + "learning_rate": 2.1280856386419268e-05, + "loss": 2.6477, + "step": 37027 + }, + { + "epoch": 1.58, + "grad_norm": 0.4375, + "learning_rate": 2.1276719723670558e-05, + "loss": 2.4723, + "step": 37028 + }, + { + "epoch": 1.58, + "grad_norm": 0.486328125, + "learning_rate": 2.1272583415147873e-05, + "loss": 2.5091, + "step": 37029 + }, + { + "epoch": 1.58, + "grad_norm": 0.5, + "learning_rate": 2.126844746086979e-05, + "loss": 2.5284, + "step": 37030 + }, + { + "epoch": 1.58, + "grad_norm": 0.5078125, + "learning_rate": 2.1264311860854867e-05, + "loss": 2.5415, + "step": 37031 + }, + { + "epoch": 1.58, + "grad_norm": 0.37890625, + "learning_rate": 2.126017661512182e-05, + "loss": 2.505, + "step": 37032 + }, + { + "epoch": 1.58, + "grad_norm": 0.416015625, + "learning_rate": 2.1256041723689202e-05, + "loss": 2.5717, + "step": 37033 + }, + { + "epoch": 1.58, + "grad_norm": 0.43359375, + "learning_rate": 2.1251907186575583e-05, + "loss": 2.3624, + "step": 37034 + }, + { + "epoch": 1.58, + "grad_norm": 0.5078125, + "learning_rate": 2.124777300379961e-05, + "loss": 2.567, + "step": 37035 + }, + { + "epoch": 1.58, + "grad_norm": 0.431640625, + "learning_rate": 2.1243639175379904e-05, + "loss": 2.4148, + "step": 37036 + }, + { + "epoch": 1.58, + "grad_norm": 0.45703125, + "learning_rate": 2.1239505701335016e-05, + "loss": 2.5045, + "step": 37037 + }, + { + "epoch": 1.58, + "grad_norm": 0.453125, + "learning_rate": 2.12353725816836e-05, + "loss": 2.5232, + "step": 37038 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.1231239816444183e-05, + "loss": 2.3913, + "step": 37039 + }, + { + "epoch": 1.58, + "grad_norm": 0.484375, + "learning_rate": 2.1227107405635426e-05, + "loss": 2.4196, + "step": 37040 + }, + { + "epoch": 1.58, + "grad_norm": 0.439453125, + "learning_rate": 2.1222975349275887e-05, + "loss": 2.392, + "step": 37041 + }, + { + "epoch": 1.58, + "grad_norm": 0.453125, + "learning_rate": 2.121884364738418e-05, + "loss": 2.4724, + "step": 37042 + }, + { + "epoch": 1.58, + "grad_norm": 0.4375, + "learning_rate": 2.1214712299978868e-05, + "loss": 2.4028, + "step": 37043 + }, + { + "epoch": 1.58, + "grad_norm": 0.48046875, + "learning_rate": 2.121058130707855e-05, + "loss": 2.486, + "step": 37044 + }, + { + "epoch": 1.58, + "grad_norm": 0.40234375, + "learning_rate": 2.120645066870186e-05, + "loss": 2.5818, + "step": 37045 + }, + { + "epoch": 1.58, + "grad_norm": 0.49609375, + "learning_rate": 2.1202320384867336e-05, + "loss": 2.5399, + "step": 37046 + }, + { + "epoch": 1.58, + "grad_norm": 0.435546875, + "learning_rate": 2.1198190455593526e-05, + "loss": 2.4256, + "step": 37047 + }, + { + "epoch": 1.58, + "grad_norm": 0.46484375, + "learning_rate": 2.11940608808991e-05, + "loss": 2.4248, + "step": 37048 + }, + { + "epoch": 1.58, + "grad_norm": 0.431640625, + "learning_rate": 2.118993166080262e-05, + "loss": 2.5105, + "step": 37049 + }, + { + "epoch": 1.58, + "grad_norm": 0.51171875, + "learning_rate": 2.1185802795322607e-05, + "loss": 2.4224, + "step": 37050 + }, + { + "epoch": 1.58, + "grad_norm": 0.439453125, + "learning_rate": 2.1181674284477682e-05, + "loss": 2.3392, + "step": 37051 + }, + { + "epoch": 1.58, + "grad_norm": 0.462890625, + "learning_rate": 2.117754612828644e-05, + "loss": 2.5294, + "step": 37052 + }, + { + "epoch": 1.58, + "grad_norm": 0.4609375, + "learning_rate": 2.1173418326767414e-05, + "loss": 2.532, + "step": 37053 + }, + { + "epoch": 1.58, + "grad_norm": 0.4296875, + "learning_rate": 2.116929087993922e-05, + "loss": 2.4638, + "step": 37054 + }, + { + "epoch": 1.58, + "grad_norm": 0.515625, + "learning_rate": 2.1165163787820375e-05, + "loss": 2.4535, + "step": 37055 + }, + { + "epoch": 1.58, + "grad_norm": 0.4453125, + "learning_rate": 2.1161037050429523e-05, + "loss": 2.4996, + "step": 37056 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.1156910667785158e-05, + "loss": 2.4975, + "step": 37057 + }, + { + "epoch": 1.58, + "grad_norm": 0.44921875, + "learning_rate": 2.115278463990591e-05, + "loss": 2.4817, + "step": 37058 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.1148658966810285e-05, + "loss": 2.4648, + "step": 37059 + }, + { + "epoch": 1.58, + "grad_norm": 0.390625, + "learning_rate": 2.1144533648516872e-05, + "loss": 2.4429, + "step": 37060 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.1140408685044276e-05, + "loss": 2.4982, + "step": 37061 + }, + { + "epoch": 1.58, + "grad_norm": 0.4921875, + "learning_rate": 2.1136284076411017e-05, + "loss": 2.3895, + "step": 37062 + }, + { + "epoch": 1.58, + "grad_norm": 0.5078125, + "learning_rate": 2.1132159822635612e-05, + "loss": 2.4676, + "step": 37063 + }, + { + "epoch": 1.58, + "grad_norm": 0.52734375, + "learning_rate": 2.112803592373671e-05, + "loss": 2.6482, + "step": 37064 + }, + { + "epoch": 1.58, + "grad_norm": 0.46484375, + "learning_rate": 2.1123912379732825e-05, + "loss": 2.4733, + "step": 37065 + }, + { + "epoch": 1.58, + "grad_norm": 0.51953125, + "learning_rate": 2.1119789190642482e-05, + "loss": 2.3933, + "step": 37066 + }, + { + "epoch": 1.58, + "grad_norm": 0.427734375, + "learning_rate": 2.1115666356484253e-05, + "loss": 2.5078, + "step": 37067 + }, + { + "epoch": 1.58, + "grad_norm": 0.4765625, + "learning_rate": 2.111154387727674e-05, + "loss": 2.5215, + "step": 37068 + }, + { + "epoch": 1.58, + "grad_norm": 0.447265625, + "learning_rate": 2.1107421753038404e-05, + "loss": 2.3742, + "step": 37069 + }, + { + "epoch": 1.58, + "grad_norm": 0.486328125, + "learning_rate": 2.1103299983787872e-05, + "loss": 2.3936, + "step": 37070 + }, + { + "epoch": 1.58, + "grad_norm": 0.484375, + "learning_rate": 2.109917856954363e-05, + "loss": 2.4455, + "step": 37071 + }, + { + "epoch": 1.58, + "grad_norm": 0.416015625, + "learning_rate": 2.109505751032428e-05, + "loss": 2.5579, + "step": 37072 + }, + { + "epoch": 1.58, + "grad_norm": 0.392578125, + "learning_rate": 2.109093680614831e-05, + "loss": 2.2871, + "step": 37073 + }, + { + "epoch": 1.58, + "grad_norm": 0.515625, + "learning_rate": 2.108681645703431e-05, + "loss": 2.4198, + "step": 37074 + }, + { + "epoch": 1.58, + "grad_norm": 0.44921875, + "learning_rate": 2.108269646300076e-05, + "loss": 2.5067, + "step": 37075 + }, + { + "epoch": 1.58, + "grad_norm": 0.427734375, + "learning_rate": 2.107857682406624e-05, + "loss": 2.4066, + "step": 37076 + }, + { + "epoch": 1.58, + "grad_norm": 0.43359375, + "learning_rate": 2.1074457540249315e-05, + "loss": 2.4785, + "step": 37077 + }, + { + "epoch": 1.58, + "grad_norm": 0.498046875, + "learning_rate": 2.1070338611568485e-05, + "loss": 2.5409, + "step": 37078 + }, + { + "epoch": 1.58, + "grad_norm": 0.44140625, + "learning_rate": 2.1066220038042238e-05, + "loss": 2.4267, + "step": 37079 + }, + { + "epoch": 1.58, + "grad_norm": 0.48828125, + "learning_rate": 2.1062101819689195e-05, + "loss": 2.516, + "step": 37080 + }, + { + "epoch": 1.58, + "grad_norm": 0.45703125, + "learning_rate": 2.105798395652785e-05, + "loss": 2.3247, + "step": 37081 + }, + { + "epoch": 1.58, + "grad_norm": 0.375, + "learning_rate": 2.1053866448576698e-05, + "loss": 2.3871, + "step": 37082 + }, + { + "epoch": 1.58, + "grad_norm": 0.44140625, + "learning_rate": 2.1049749295854304e-05, + "loss": 2.6151, + "step": 37083 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.104563249837921e-05, + "loss": 2.4499, + "step": 37084 + }, + { + "epoch": 1.58, + "grad_norm": 0.48828125, + "learning_rate": 2.1041516056169884e-05, + "loss": 2.5044, + "step": 37085 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.1037399969244907e-05, + "loss": 2.5001, + "step": 37086 + }, + { + "epoch": 1.58, + "grad_norm": 0.49609375, + "learning_rate": 2.103328423762274e-05, + "loss": 2.4724, + "step": 37087 + }, + { + "epoch": 1.58, + "grad_norm": 0.431640625, + "learning_rate": 2.1029168861321967e-05, + "loss": 2.5517, + "step": 37088 + }, + { + "epoch": 1.58, + "grad_norm": 0.451171875, + "learning_rate": 2.1025053840361053e-05, + "loss": 2.4673, + "step": 37089 + }, + { + "epoch": 1.58, + "grad_norm": 0.4453125, + "learning_rate": 2.1020939174758558e-05, + "loss": 2.4833, + "step": 37090 + }, + { + "epoch": 1.58, + "grad_norm": 0.455078125, + "learning_rate": 2.1016824864532937e-05, + "loss": 2.4499, + "step": 37091 + }, + { + "epoch": 1.58, + "grad_norm": 0.4765625, + "learning_rate": 2.1012710909702747e-05, + "loss": 2.3813, + "step": 37092 + }, + { + "epoch": 1.58, + "grad_norm": 0.5, + "learning_rate": 2.1008597310286525e-05, + "loss": 2.5559, + "step": 37093 + }, + { + "epoch": 1.58, + "grad_norm": 0.435546875, + "learning_rate": 2.1004484066302732e-05, + "loss": 2.243, + "step": 37094 + }, + { + "epoch": 1.58, + "grad_norm": 0.408203125, + "learning_rate": 2.100037117776985e-05, + "loss": 2.4213, + "step": 37095 + }, + { + "epoch": 1.58, + "grad_norm": 0.451171875, + "learning_rate": 2.099625864470648e-05, + "loss": 2.3483, + "step": 37096 + }, + { + "epoch": 1.58, + "grad_norm": 0.43359375, + "learning_rate": 2.099214646713107e-05, + "loss": 2.4482, + "step": 37097 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.0988034645062094e-05, + "loss": 2.3632, + "step": 37098 + }, + { + "epoch": 1.58, + "grad_norm": 0.416015625, + "learning_rate": 2.0983923178518096e-05, + "loss": 2.3649, + "step": 37099 + }, + { + "epoch": 1.58, + "grad_norm": 0.39453125, + "learning_rate": 2.097981206751758e-05, + "loss": 2.401, + "step": 37100 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.0975701312079023e-05, + "loss": 2.5593, + "step": 37101 + }, + { + "epoch": 1.58, + "grad_norm": 0.46875, + "learning_rate": 2.097159091222094e-05, + "loss": 2.4279, + "step": 37102 + }, + { + "epoch": 1.58, + "grad_norm": 0.498046875, + "learning_rate": 2.0967480867961797e-05, + "loss": 2.602, + "step": 37103 + }, + { + "epoch": 1.58, + "grad_norm": 0.5625, + "learning_rate": 2.0963371179320135e-05, + "loss": 2.4215, + "step": 37104 + }, + { + "epoch": 1.58, + "grad_norm": 0.48828125, + "learning_rate": 2.0959261846314393e-05, + "loss": 2.4788, + "step": 37105 + }, + { + "epoch": 1.58, + "grad_norm": 0.447265625, + "learning_rate": 2.0955152868963102e-05, + "loss": 2.2012, + "step": 37106 + }, + { + "epoch": 1.58, + "grad_norm": 0.49609375, + "learning_rate": 2.0951044247284723e-05, + "loss": 2.4946, + "step": 37107 + }, + { + "epoch": 1.58, + "grad_norm": 0.484375, + "learning_rate": 2.0946935981297742e-05, + "loss": 2.5883, + "step": 37108 + }, + { + "epoch": 1.58, + "grad_norm": 0.458984375, + "learning_rate": 2.0942828071020703e-05, + "loss": 2.4255, + "step": 37109 + }, + { + "epoch": 1.58, + "grad_norm": 0.478515625, + "learning_rate": 2.0938720516472032e-05, + "loss": 2.3705, + "step": 37110 + }, + { + "epoch": 1.58, + "grad_norm": 0.4296875, + "learning_rate": 2.0934613317670203e-05, + "loss": 2.4529, + "step": 37111 + }, + { + "epoch": 1.58, + "grad_norm": 0.455078125, + "learning_rate": 2.0930506474633715e-05, + "loss": 2.2775, + "step": 37112 + }, + { + "epoch": 1.58, + "grad_norm": 0.44921875, + "learning_rate": 2.0926399987381086e-05, + "loss": 2.4719, + "step": 37113 + }, + { + "epoch": 1.58, + "grad_norm": 0.44140625, + "learning_rate": 2.092229385593072e-05, + "loss": 2.3964, + "step": 37114 + }, + { + "epoch": 1.58, + "grad_norm": 0.46875, + "learning_rate": 2.0918188080301173e-05, + "loss": 2.3849, + "step": 37115 + }, + { + "epoch": 1.58, + "grad_norm": 0.435546875, + "learning_rate": 2.0914082660510835e-05, + "loss": 2.3407, + "step": 37116 + }, + { + "epoch": 1.58, + "grad_norm": 0.51171875, + "learning_rate": 2.0909977596578236e-05, + "loss": 2.4879, + "step": 37117 + }, + { + "epoch": 1.58, + "grad_norm": 0.4609375, + "learning_rate": 2.090587288852186e-05, + "loss": 2.3706, + "step": 37118 + }, + { + "epoch": 1.58, + "grad_norm": 0.443359375, + "learning_rate": 2.0901768536360144e-05, + "loss": 2.4821, + "step": 37119 + }, + { + "epoch": 1.58, + "grad_norm": 0.43359375, + "learning_rate": 2.0897664540111538e-05, + "loss": 2.3893, + "step": 37120 + }, + { + "epoch": 1.58, + "grad_norm": 0.431640625, + "learning_rate": 2.089356089979454e-05, + "loss": 2.4353, + "step": 37121 + }, + { + "epoch": 1.58, + "grad_norm": 0.42578125, + "learning_rate": 2.088945761542762e-05, + "loss": 2.408, + "step": 37122 + }, + { + "epoch": 1.58, + "grad_norm": 0.51171875, + "learning_rate": 2.0885354687029212e-05, + "loss": 2.4571, + "step": 37123 + }, + { + "epoch": 1.58, + "grad_norm": 0.458984375, + "learning_rate": 2.08812521146178e-05, + "loss": 2.3819, + "step": 37124 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.0877149898211855e-05, + "loss": 2.4816, + "step": 37125 + }, + { + "epoch": 1.58, + "grad_norm": 0.41796875, + "learning_rate": 2.0873048037829822e-05, + "loss": 2.3158, + "step": 37126 + }, + { + "epoch": 1.58, + "grad_norm": 0.478515625, + "learning_rate": 2.086894653349013e-05, + "loss": 2.4621, + "step": 37127 + }, + { + "epoch": 1.58, + "grad_norm": 0.451171875, + "learning_rate": 2.0864845385211262e-05, + "loss": 2.5996, + "step": 37128 + }, + { + "epoch": 1.58, + "grad_norm": 0.490234375, + "learning_rate": 2.0860744593011695e-05, + "loss": 2.5372, + "step": 37129 + }, + { + "epoch": 1.58, + "grad_norm": 0.44140625, + "learning_rate": 2.085664415690982e-05, + "loss": 2.4718, + "step": 37130 + }, + { + "epoch": 1.58, + "grad_norm": 0.4765625, + "learning_rate": 2.0852544076924162e-05, + "loss": 2.4131, + "step": 37131 + }, + { + "epoch": 1.58, + "grad_norm": 0.4296875, + "learning_rate": 2.0848444353073093e-05, + "loss": 2.5156, + "step": 37132 + }, + { + "epoch": 1.58, + "grad_norm": 0.47265625, + "learning_rate": 2.0844344985375108e-05, + "loss": 2.3736, + "step": 37133 + }, + { + "epoch": 1.58, + "grad_norm": 0.404296875, + "learning_rate": 2.084024597384866e-05, + "loss": 2.3151, + "step": 37134 + }, + { + "epoch": 1.58, + "grad_norm": 0.482421875, + "learning_rate": 2.083614731851219e-05, + "loss": 2.3474, + "step": 37135 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.0832049019384082e-05, + "loss": 2.5508, + "step": 37136 + }, + { + "epoch": 1.58, + "grad_norm": 0.427734375, + "learning_rate": 2.0827951076482834e-05, + "loss": 2.4503, + "step": 37137 + }, + { + "epoch": 1.58, + "grad_norm": 0.458984375, + "learning_rate": 2.0823853489826882e-05, + "loss": 2.4084, + "step": 37138 + }, + { + "epoch": 1.58, + "grad_norm": 0.45703125, + "learning_rate": 2.081975625943464e-05, + "loss": 2.3889, + "step": 37139 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.081565938532456e-05, + "loss": 2.5566, + "step": 37140 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.08115628675151e-05, + "loss": 2.5294, + "step": 37141 + }, + { + "epoch": 1.58, + "grad_norm": 0.58203125, + "learning_rate": 2.080746670602466e-05, + "loss": 2.4738, + "step": 37142 + }, + { + "epoch": 1.58, + "grad_norm": 0.484375, + "learning_rate": 2.0803370900871655e-05, + "loss": 2.3678, + "step": 37143 + }, + { + "epoch": 1.58, + "grad_norm": 0.453125, + "learning_rate": 2.079927545207454e-05, + "loss": 2.4362, + "step": 37144 + }, + { + "epoch": 1.58, + "grad_norm": 0.490234375, + "learning_rate": 2.0795180359651767e-05, + "loss": 2.6417, + "step": 37145 + }, + { + "epoch": 1.58, + "grad_norm": 0.447265625, + "learning_rate": 2.07910856236217e-05, + "loss": 2.4048, + "step": 37146 + }, + { + "epoch": 1.58, + "grad_norm": 0.470703125, + "learning_rate": 2.078699124400284e-05, + "loss": 2.2843, + "step": 37147 + }, + { + "epoch": 1.58, + "grad_norm": 0.435546875, + "learning_rate": 2.0782897220813536e-05, + "loss": 2.61, + "step": 37148 + }, + { + "epoch": 1.58, + "grad_norm": 0.41015625, + "learning_rate": 2.0778803554072247e-05, + "loss": 2.3524, + "step": 37149 + }, + { + "epoch": 1.58, + "grad_norm": 0.431640625, + "learning_rate": 2.0774710243797414e-05, + "loss": 2.4408, + "step": 37150 + }, + { + "epoch": 1.58, + "grad_norm": 0.4375, + "learning_rate": 2.0770617290007432e-05, + "loss": 2.4339, + "step": 37151 + }, + { + "epoch": 1.58, + "grad_norm": 0.462890625, + "learning_rate": 2.0766524692720703e-05, + "loss": 2.5453, + "step": 37152 + }, + { + "epoch": 1.58, + "grad_norm": 0.45703125, + "learning_rate": 2.076243245195564e-05, + "loss": 2.5361, + "step": 37153 + }, + { + "epoch": 1.58, + "grad_norm": 0.42578125, + "learning_rate": 2.0758340567730704e-05, + "loss": 2.5337, + "step": 37154 + }, + { + "epoch": 1.58, + "grad_norm": 0.48828125, + "learning_rate": 2.0754249040064244e-05, + "loss": 2.4997, + "step": 37155 + }, + { + "epoch": 1.58, + "grad_norm": 0.44140625, + "learning_rate": 2.0750157868974718e-05, + "loss": 2.4089, + "step": 37156 + }, + { + "epoch": 1.58, + "grad_norm": 0.5078125, + "learning_rate": 2.0746067054480533e-05, + "loss": 2.4717, + "step": 37157 + }, + { + "epoch": 1.58, + "grad_norm": 0.44921875, + "learning_rate": 2.0741976596600087e-05, + "loss": 2.4131, + "step": 37158 + }, + { + "epoch": 1.58, + "grad_norm": 0.470703125, + "learning_rate": 2.0737886495351754e-05, + "loss": 2.5496, + "step": 37159 + }, + { + "epoch": 1.58, + "grad_norm": 0.451171875, + "learning_rate": 2.0733796750753965e-05, + "loss": 2.4354, + "step": 37160 + }, + { + "epoch": 1.58, + "grad_norm": 0.43359375, + "learning_rate": 2.0729707362825145e-05, + "loss": 2.4854, + "step": 37161 + }, + { + "epoch": 1.58, + "grad_norm": 0.47265625, + "learning_rate": 2.072561833158364e-05, + "loss": 2.4852, + "step": 37162 + }, + { + "epoch": 1.58, + "grad_norm": 0.494140625, + "learning_rate": 2.0721529657047923e-05, + "loss": 2.4381, + "step": 37163 + }, + { + "epoch": 1.58, + "grad_norm": 0.46875, + "learning_rate": 2.071744133923631e-05, + "loss": 2.41, + "step": 37164 + }, + { + "epoch": 1.58, + "grad_norm": 0.4921875, + "learning_rate": 2.0713353378167243e-05, + "loss": 2.4609, + "step": 37165 + }, + { + "epoch": 1.58, + "grad_norm": 0.474609375, + "learning_rate": 2.0709265773859132e-05, + "loss": 2.436, + "step": 37166 + }, + { + "epoch": 1.58, + "grad_norm": 0.443359375, + "learning_rate": 2.070517852633036e-05, + "loss": 2.418, + "step": 37167 + }, + { + "epoch": 1.58, + "grad_norm": 0.494140625, + "learning_rate": 2.070109163559927e-05, + "loss": 2.4825, + "step": 37168 + }, + { + "epoch": 1.58, + "grad_norm": 0.4609375, + "learning_rate": 2.0697005101684285e-05, + "loss": 2.3334, + "step": 37169 + }, + { + "epoch": 1.58, + "grad_norm": 0.515625, + "learning_rate": 2.069291892460382e-05, + "loss": 2.457, + "step": 37170 + }, + { + "epoch": 1.58, + "grad_norm": 0.49609375, + "learning_rate": 2.0688833104376215e-05, + "loss": 2.5384, + "step": 37171 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.0684747641019875e-05, + "loss": 2.4839, + "step": 37172 + }, + { + "epoch": 1.58, + "grad_norm": 0.470703125, + "learning_rate": 2.068066253455321e-05, + "loss": 2.5252, + "step": 37173 + }, + { + "epoch": 1.58, + "grad_norm": 0.458984375, + "learning_rate": 2.067657778499458e-05, + "loss": 2.4326, + "step": 37174 + }, + { + "epoch": 1.58, + "grad_norm": 0.54296875, + "learning_rate": 2.0672493392362324e-05, + "loss": 2.49, + "step": 37175 + }, + { + "epoch": 1.58, + "grad_norm": 0.462890625, + "learning_rate": 2.0668409356674866e-05, + "loss": 2.5729, + "step": 37176 + }, + { + "epoch": 1.58, + "grad_norm": 0.4375, + "learning_rate": 2.0664325677950602e-05, + "loss": 2.511, + "step": 37177 + }, + { + "epoch": 1.58, + "grad_norm": 0.51953125, + "learning_rate": 2.0660242356207838e-05, + "loss": 2.5208, + "step": 37178 + }, + { + "epoch": 1.58, + "grad_norm": 0.41796875, + "learning_rate": 2.0656159391465023e-05, + "loss": 2.3336, + "step": 37179 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.065207678374047e-05, + "loss": 2.4588, + "step": 37180 + }, + { + "epoch": 1.58, + "grad_norm": 0.50390625, + "learning_rate": 2.0647994533052572e-05, + "loss": 2.3759, + "step": 37181 + }, + { + "epoch": 1.58, + "grad_norm": 0.515625, + "learning_rate": 2.0643912639419727e-05, + "loss": 2.4336, + "step": 37182 + }, + { + "epoch": 1.58, + "grad_norm": 0.5078125, + "learning_rate": 2.0639831102860273e-05, + "loss": 2.4994, + "step": 37183 + }, + { + "epoch": 1.58, + "grad_norm": 0.41015625, + "learning_rate": 2.0635749923392544e-05, + "loss": 2.4299, + "step": 37184 + }, + { + "epoch": 1.58, + "grad_norm": 0.466796875, + "learning_rate": 2.0631669101034947e-05, + "loss": 2.3387, + "step": 37185 + }, + { + "epoch": 1.58, + "grad_norm": 0.5546875, + "learning_rate": 2.062758863580585e-05, + "loss": 2.4864, + "step": 37186 + }, + { + "epoch": 1.58, + "grad_norm": 0.49609375, + "learning_rate": 2.0623508527723568e-05, + "loss": 2.4297, + "step": 37187 + }, + { + "epoch": 1.58, + "grad_norm": 0.4609375, + "learning_rate": 2.0619428776806494e-05, + "loss": 2.561, + "step": 37188 + }, + { + "epoch": 1.58, + "grad_norm": 0.494140625, + "learning_rate": 2.061534938307301e-05, + "loss": 2.5887, + "step": 37189 + }, + { + "epoch": 1.58, + "grad_norm": 0.5625, + "learning_rate": 2.0611270346541435e-05, + "loss": 2.5431, + "step": 37190 + }, + { + "epoch": 1.58, + "grad_norm": 0.451171875, + "learning_rate": 2.060719166723011e-05, + "loss": 2.4877, + "step": 37191 + }, + { + "epoch": 1.58, + "grad_norm": 0.48828125, + "learning_rate": 2.060311334515741e-05, + "loss": 2.49, + "step": 37192 + }, + { + "epoch": 1.58, + "grad_norm": 0.486328125, + "learning_rate": 2.0599035380341703e-05, + "loss": 2.5158, + "step": 37193 + }, + { + "epoch": 1.58, + "grad_norm": 0.515625, + "learning_rate": 2.0594957772801295e-05, + "loss": 2.4309, + "step": 37194 + }, + { + "epoch": 1.58, + "grad_norm": 0.439453125, + "learning_rate": 2.059088052255458e-05, + "loss": 2.4195, + "step": 37195 + }, + { + "epoch": 1.59, + "grad_norm": 0.4375, + "learning_rate": 2.0586803629619867e-05, + "loss": 2.4237, + "step": 37196 + }, + { + "epoch": 1.59, + "grad_norm": 0.447265625, + "learning_rate": 2.0582727094015496e-05, + "loss": 2.422, + "step": 37197 + }, + { + "epoch": 1.59, + "grad_norm": 0.5234375, + "learning_rate": 2.057865091575988e-05, + "loss": 2.6119, + "step": 37198 + }, + { + "epoch": 1.59, + "grad_norm": 0.447265625, + "learning_rate": 2.0574575094871294e-05, + "loss": 2.4326, + "step": 37199 + }, + { + "epoch": 1.59, + "grad_norm": 0.48046875, + "learning_rate": 2.057049963136807e-05, + "loss": 2.4985, + "step": 37200 + }, + { + "epoch": 1.59, + "grad_norm": 0.50390625, + "learning_rate": 2.056642452526857e-05, + "loss": 2.5062, + "step": 37201 + }, + { + "epoch": 1.59, + "grad_norm": 0.400390625, + "learning_rate": 2.0562349776591148e-05, + "loss": 2.4246, + "step": 37202 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 2.055827538535411e-05, + "loss": 2.3063, + "step": 37203 + }, + { + "epoch": 1.59, + "grad_norm": 0.47265625, + "learning_rate": 2.0554201351575797e-05, + "loss": 2.4275, + "step": 37204 + }, + { + "epoch": 1.59, + "grad_norm": 0.458984375, + "learning_rate": 2.0550127675274566e-05, + "loss": 2.4001, + "step": 37205 + }, + { + "epoch": 1.59, + "grad_norm": 0.48828125, + "learning_rate": 2.0546054356468724e-05, + "loss": 2.4966, + "step": 37206 + }, + { + "epoch": 1.59, + "grad_norm": 0.484375, + "learning_rate": 2.054198139517657e-05, + "loss": 2.5556, + "step": 37207 + }, + { + "epoch": 1.59, + "grad_norm": 0.48046875, + "learning_rate": 2.053790879141647e-05, + "loss": 2.4804, + "step": 37208 + }, + { + "epoch": 1.59, + "grad_norm": 0.47265625, + "learning_rate": 2.0533836545206765e-05, + "loss": 2.5531, + "step": 37209 + }, + { + "epoch": 1.59, + "grad_norm": 0.458984375, + "learning_rate": 2.052976465656572e-05, + "loss": 2.4656, + "step": 37210 + }, + { + "epoch": 1.59, + "grad_norm": 0.416015625, + "learning_rate": 2.0525693125511726e-05, + "loss": 2.5189, + "step": 37211 + }, + { + "epoch": 1.59, + "grad_norm": 0.51171875, + "learning_rate": 2.052162195206304e-05, + "loss": 2.4176, + "step": 37212 + }, + { + "epoch": 1.59, + "grad_norm": 0.42578125, + "learning_rate": 2.0517551136238e-05, + "loss": 2.4874, + "step": 37213 + }, + { + "epoch": 1.59, + "grad_norm": 0.478515625, + "learning_rate": 2.0513480678054975e-05, + "loss": 2.376, + "step": 37214 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 2.050941057753223e-05, + "loss": 2.5588, + "step": 37215 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 2.050534083468806e-05, + "loss": 2.494, + "step": 37216 + }, + { + "epoch": 1.59, + "grad_norm": 0.4609375, + "learning_rate": 2.0501271449540806e-05, + "loss": 2.5147, + "step": 37217 + }, + { + "epoch": 1.59, + "grad_norm": 0.45703125, + "learning_rate": 2.0497202422108796e-05, + "loss": 2.3103, + "step": 37218 + }, + { + "epoch": 1.59, + "grad_norm": 0.51171875, + "learning_rate": 2.04931337524103e-05, + "loss": 2.5783, + "step": 37219 + }, + { + "epoch": 1.59, + "grad_norm": 0.4296875, + "learning_rate": 2.048906544046365e-05, + "loss": 2.4514, + "step": 37220 + }, + { + "epoch": 1.59, + "grad_norm": 0.4453125, + "learning_rate": 2.048499748628717e-05, + "loss": 2.4298, + "step": 37221 + }, + { + "epoch": 1.59, + "grad_norm": 0.4296875, + "learning_rate": 2.0480929889899135e-05, + "loss": 2.362, + "step": 37222 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 2.047686265131784e-05, + "loss": 2.4313, + "step": 37223 + }, + { + "epoch": 1.59, + "grad_norm": 0.498046875, + "learning_rate": 2.047279577056159e-05, + "loss": 2.3994, + "step": 37224 + }, + { + "epoch": 1.59, + "grad_norm": 0.439453125, + "learning_rate": 2.0468729247648732e-05, + "loss": 2.4271, + "step": 37225 + }, + { + "epoch": 1.59, + "grad_norm": 0.392578125, + "learning_rate": 2.046466308259749e-05, + "loss": 2.3979, + "step": 37226 + }, + { + "epoch": 1.59, + "grad_norm": 0.44140625, + "learning_rate": 2.046059727542623e-05, + "loss": 2.5319, + "step": 37227 + }, + { + "epoch": 1.59, + "grad_norm": 0.423828125, + "learning_rate": 2.0456531826153214e-05, + "loss": 2.3587, + "step": 37228 + }, + { + "epoch": 1.59, + "grad_norm": 0.42578125, + "learning_rate": 2.0452466734796682e-05, + "loss": 2.4143, + "step": 37229 + }, + { + "epoch": 1.59, + "grad_norm": 0.515625, + "learning_rate": 2.0448402001375034e-05, + "loss": 2.3833, + "step": 37230 + }, + { + "epoch": 1.59, + "grad_norm": 0.5, + "learning_rate": 2.0444337625906497e-05, + "loss": 2.3213, + "step": 37231 + }, + { + "epoch": 1.59, + "grad_norm": 0.5, + "learning_rate": 2.0440273608409344e-05, + "loss": 2.5635, + "step": 37232 + }, + { + "epoch": 1.59, + "grad_norm": 0.470703125, + "learning_rate": 2.0436209948901887e-05, + "loss": 2.3635, + "step": 37233 + }, + { + "epoch": 1.59, + "grad_norm": 0.47265625, + "learning_rate": 2.043214664740243e-05, + "loss": 2.3331, + "step": 37234 + }, + { + "epoch": 1.59, + "grad_norm": 0.484375, + "learning_rate": 2.0428083703929203e-05, + "loss": 2.5664, + "step": 37235 + }, + { + "epoch": 1.59, + "grad_norm": 0.474609375, + "learning_rate": 2.0424021118500557e-05, + "loss": 2.4571, + "step": 37236 + }, + { + "epoch": 1.59, + "grad_norm": 0.4609375, + "learning_rate": 2.0419958891134706e-05, + "loss": 2.4841, + "step": 37237 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 2.041589702184997e-05, + "loss": 2.4292, + "step": 37238 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 2.04118355106646e-05, + "loss": 2.3445, + "step": 37239 + }, + { + "epoch": 1.59, + "grad_norm": 0.48046875, + "learning_rate": 2.040777435759691e-05, + "loss": 2.4387, + "step": 37240 + }, + { + "epoch": 1.59, + "grad_norm": 0.515625, + "learning_rate": 2.0403713562665116e-05, + "loss": 2.3924, + "step": 37241 + }, + { + "epoch": 1.59, + "grad_norm": 0.4609375, + "learning_rate": 2.039965312588752e-05, + "loss": 2.478, + "step": 37242 + }, + { + "epoch": 1.59, + "grad_norm": 0.478515625, + "learning_rate": 2.0395593047282425e-05, + "loss": 2.5347, + "step": 37243 + }, + { + "epoch": 1.59, + "grad_norm": 0.4921875, + "learning_rate": 2.0391533326868073e-05, + "loss": 2.3822, + "step": 37244 + }, + { + "epoch": 1.59, + "grad_norm": 0.427734375, + "learning_rate": 2.0387473964662663e-05, + "loss": 2.4489, + "step": 37245 + }, + { + "epoch": 1.59, + "grad_norm": 0.458984375, + "learning_rate": 2.038341496068459e-05, + "loss": 2.4099, + "step": 37246 + }, + { + "epoch": 1.59, + "grad_norm": 0.53515625, + "learning_rate": 2.0379356314952047e-05, + "loss": 2.336, + "step": 37247 + }, + { + "epoch": 1.59, + "grad_norm": 0.4296875, + "learning_rate": 2.037529802748328e-05, + "loss": 2.4173, + "step": 37248 + }, + { + "epoch": 1.59, + "grad_norm": 0.54296875, + "learning_rate": 2.0371240098296563e-05, + "loss": 2.388, + "step": 37249 + }, + { + "epoch": 1.59, + "grad_norm": 0.44140625, + "learning_rate": 2.03671825274102e-05, + "loss": 2.4022, + "step": 37250 + }, + { + "epoch": 1.59, + "grad_norm": 0.4375, + "learning_rate": 2.036312531484239e-05, + "loss": 2.5001, + "step": 37251 + }, + { + "epoch": 1.59, + "grad_norm": 0.50390625, + "learning_rate": 2.035906846061143e-05, + "loss": 2.5106, + "step": 37252 + }, + { + "epoch": 1.59, + "grad_norm": 0.447265625, + "learning_rate": 2.035501196473554e-05, + "loss": 2.4359, + "step": 37253 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 2.0350955827233e-05, + "loss": 2.3876, + "step": 37254 + }, + { + "epoch": 1.59, + "grad_norm": 0.4375, + "learning_rate": 2.0346900048122032e-05, + "loss": 2.4799, + "step": 37255 + }, + { + "epoch": 1.59, + "grad_norm": 0.40625, + "learning_rate": 2.034284462742093e-05, + "loss": 2.4466, + "step": 37256 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 2.0338789565147896e-05, + "loss": 2.391, + "step": 37257 + }, + { + "epoch": 1.59, + "grad_norm": 0.546875, + "learning_rate": 2.033473486132119e-05, + "loss": 2.4312, + "step": 37258 + }, + { + "epoch": 1.59, + "grad_norm": 0.498046875, + "learning_rate": 2.033068051595909e-05, + "loss": 2.4002, + "step": 37259 + }, + { + "epoch": 1.59, + "grad_norm": 0.447265625, + "learning_rate": 2.0326626529079805e-05, + "loss": 2.433, + "step": 37260 + }, + { + "epoch": 1.59, + "grad_norm": 0.490234375, + "learning_rate": 2.0322572900701542e-05, + "loss": 2.5643, + "step": 37261 + }, + { + "epoch": 1.59, + "grad_norm": 0.482421875, + "learning_rate": 2.0318519630842635e-05, + "loss": 2.5627, + "step": 37262 + }, + { + "epoch": 1.59, + "grad_norm": 0.474609375, + "learning_rate": 2.0314466719521265e-05, + "loss": 2.4094, + "step": 37263 + }, + { + "epoch": 1.59, + "grad_norm": 0.447265625, + "learning_rate": 2.0310414166755643e-05, + "loss": 2.4317, + "step": 37264 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 2.0306361972564036e-05, + "loss": 2.4183, + "step": 37265 + }, + { + "epoch": 1.59, + "grad_norm": 0.51953125, + "learning_rate": 2.0302310136964707e-05, + "loss": 2.502, + "step": 37266 + }, + { + "epoch": 1.59, + "grad_norm": 0.53515625, + "learning_rate": 2.029825865997582e-05, + "loss": 2.4118, + "step": 37267 + }, + { + "epoch": 1.59, + "grad_norm": 0.455078125, + "learning_rate": 2.0294207541615674e-05, + "loss": 2.4542, + "step": 37268 + }, + { + "epoch": 1.59, + "grad_norm": 0.455078125, + "learning_rate": 2.0290156781902436e-05, + "loss": 2.3331, + "step": 37269 + }, + { + "epoch": 1.59, + "grad_norm": 0.4609375, + "learning_rate": 2.0286106380854397e-05, + "loss": 2.3879, + "step": 37270 + }, + { + "epoch": 1.59, + "grad_norm": 0.453125, + "learning_rate": 2.0282056338489707e-05, + "loss": 2.2702, + "step": 37271 + }, + { + "epoch": 1.59, + "grad_norm": 0.4609375, + "learning_rate": 2.0278006654826665e-05, + "loss": 2.434, + "step": 37272 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 2.027395732988343e-05, + "loss": 2.4504, + "step": 37273 + }, + { + "epoch": 1.59, + "grad_norm": 0.412109375, + "learning_rate": 2.0269908363678257e-05, + "loss": 2.3222, + "step": 37274 + }, + { + "epoch": 1.59, + "grad_norm": 0.482421875, + "learning_rate": 2.026585975622938e-05, + "loss": 2.6914, + "step": 37275 + }, + { + "epoch": 1.59, + "grad_norm": 0.484375, + "learning_rate": 2.0261811507554983e-05, + "loss": 2.4103, + "step": 37276 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 2.0257763617673254e-05, + "loss": 2.4412, + "step": 37277 + }, + { + "epoch": 1.59, + "grad_norm": 0.435546875, + "learning_rate": 2.0253716086602505e-05, + "loss": 2.3179, + "step": 37278 + }, + { + "epoch": 1.59, + "grad_norm": 0.484375, + "learning_rate": 2.0249668914360875e-05, + "loss": 2.4581, + "step": 37279 + }, + { + "epoch": 1.59, + "grad_norm": 0.46484375, + "learning_rate": 2.0245622100966566e-05, + "loss": 2.4841, + "step": 37280 + }, + { + "epoch": 1.59, + "grad_norm": 0.5390625, + "learning_rate": 2.0241575646437816e-05, + "loss": 2.5538, + "step": 37281 + }, + { + "epoch": 1.59, + "grad_norm": 0.4765625, + "learning_rate": 2.0237529550792844e-05, + "loss": 2.5132, + "step": 37282 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 2.0233483814049825e-05, + "loss": 2.5543, + "step": 37283 + }, + { + "epoch": 1.59, + "grad_norm": 0.423828125, + "learning_rate": 2.0229438436227e-05, + "loss": 2.3684, + "step": 37284 + }, + { + "epoch": 1.59, + "grad_norm": 0.5234375, + "learning_rate": 2.022539341734252e-05, + "loss": 2.4982, + "step": 37285 + }, + { + "epoch": 1.59, + "grad_norm": 0.46484375, + "learning_rate": 2.022134875741464e-05, + "loss": 2.5365, + "step": 37286 + }, + { + "epoch": 1.59, + "grad_norm": 0.4921875, + "learning_rate": 2.0217304456461517e-05, + "loss": 2.355, + "step": 37287 + }, + { + "epoch": 1.59, + "grad_norm": 0.431640625, + "learning_rate": 2.0213260514501387e-05, + "loss": 2.4553, + "step": 37288 + }, + { + "epoch": 1.59, + "grad_norm": 0.451171875, + "learning_rate": 2.02092169315524e-05, + "loss": 2.2606, + "step": 37289 + }, + { + "epoch": 1.59, + "grad_norm": 0.4609375, + "learning_rate": 2.0205173707632775e-05, + "loss": 2.5219, + "step": 37290 + }, + { + "epoch": 1.59, + "grad_norm": 0.47265625, + "learning_rate": 2.020113084276073e-05, + "loss": 2.5611, + "step": 37291 + }, + { + "epoch": 1.59, + "grad_norm": 0.4453125, + "learning_rate": 2.019708833695444e-05, + "loss": 2.4651, + "step": 37292 + }, + { + "epoch": 1.59, + "grad_norm": 0.4765625, + "learning_rate": 2.0193046190232033e-05, + "loss": 2.4344, + "step": 37293 + }, + { + "epoch": 1.59, + "grad_norm": 0.435546875, + "learning_rate": 2.01890044026118e-05, + "loss": 2.492, + "step": 37294 + }, + { + "epoch": 1.59, + "grad_norm": 0.52734375, + "learning_rate": 2.0184962974111888e-05, + "loss": 2.4339, + "step": 37295 + }, + { + "epoch": 1.59, + "grad_norm": 0.408203125, + "learning_rate": 2.018092190475044e-05, + "loss": 2.4655, + "step": 37296 + }, + { + "epoch": 1.59, + "grad_norm": 0.51953125, + "learning_rate": 2.017688119454568e-05, + "loss": 2.5042, + "step": 37297 + }, + { + "epoch": 1.59, + "grad_norm": 0.421875, + "learning_rate": 2.0172840843515806e-05, + "loss": 2.3731, + "step": 37298 + }, + { + "epoch": 1.59, + "grad_norm": 0.48828125, + "learning_rate": 2.016880085167895e-05, + "loss": 2.4433, + "step": 37299 + }, + { + "epoch": 1.59, + "grad_norm": 0.44921875, + "learning_rate": 2.016476121905333e-05, + "loss": 2.4107, + "step": 37300 + }, + { + "epoch": 1.59, + "grad_norm": 0.419921875, + "learning_rate": 2.0160721945657092e-05, + "loss": 2.3214, + "step": 37301 + }, + { + "epoch": 1.59, + "grad_norm": 0.5, + "learning_rate": 2.015668303150846e-05, + "loss": 2.5579, + "step": 37302 + }, + { + "epoch": 1.59, + "grad_norm": 0.38671875, + "learning_rate": 2.015264447662554e-05, + "loss": 2.377, + "step": 37303 + }, + { + "epoch": 1.59, + "grad_norm": 0.490234375, + "learning_rate": 2.014860628102656e-05, + "loss": 2.4017, + "step": 37304 + }, + { + "epoch": 1.59, + "grad_norm": 0.498046875, + "learning_rate": 2.014456844472965e-05, + "loss": 2.4564, + "step": 37305 + }, + { + "epoch": 1.59, + "grad_norm": 0.546875, + "learning_rate": 2.0140530967753e-05, + "loss": 2.4866, + "step": 37306 + }, + { + "epoch": 1.59, + "grad_norm": 0.458984375, + "learning_rate": 2.0136493850114802e-05, + "loss": 2.3783, + "step": 37307 + }, + { + "epoch": 1.59, + "grad_norm": 0.5234375, + "learning_rate": 2.013245709183319e-05, + "loss": 2.4038, + "step": 37308 + }, + { + "epoch": 1.59, + "grad_norm": 0.435546875, + "learning_rate": 2.0128420692926287e-05, + "loss": 2.42, + "step": 37309 + }, + { + "epoch": 1.59, + "grad_norm": 0.4375, + "learning_rate": 2.012438465341234e-05, + "loss": 2.5047, + "step": 37310 + }, + { + "epoch": 1.59, + "grad_norm": 0.482421875, + "learning_rate": 2.012034897330949e-05, + "loss": 2.4877, + "step": 37311 + }, + { + "epoch": 1.59, + "grad_norm": 0.482421875, + "learning_rate": 2.011631365263583e-05, + "loss": 2.4768, + "step": 37312 + }, + { + "epoch": 1.59, + "grad_norm": 0.48046875, + "learning_rate": 2.011227869140958e-05, + "loss": 2.5131, + "step": 37313 + }, + { + "epoch": 1.59, + "grad_norm": 0.44921875, + "learning_rate": 2.01082440896489e-05, + "loss": 2.5116, + "step": 37314 + }, + { + "epoch": 1.59, + "grad_norm": 0.482421875, + "learning_rate": 2.01042098473719e-05, + "loss": 2.6581, + "step": 37315 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 2.010017596459679e-05, + "loss": 2.3645, + "step": 37316 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 2.009614244134165e-05, + "loss": 2.5511, + "step": 37317 + }, + { + "epoch": 1.59, + "grad_norm": 0.439453125, + "learning_rate": 2.0092109277624703e-05, + "loss": 2.485, + "step": 37318 + }, + { + "epoch": 1.59, + "grad_norm": 0.51953125, + "learning_rate": 2.0088076473464024e-05, + "loss": 2.3903, + "step": 37319 + }, + { + "epoch": 1.59, + "grad_norm": 0.48828125, + "learning_rate": 2.0084044028877837e-05, + "loss": 2.4619, + "step": 37320 + }, + { + "epoch": 1.59, + "grad_norm": 0.50390625, + "learning_rate": 2.0080011943884203e-05, + "loss": 2.448, + "step": 37321 + }, + { + "epoch": 1.59, + "grad_norm": 0.453125, + "learning_rate": 2.007598021850132e-05, + "loss": 2.5958, + "step": 37322 + }, + { + "epoch": 1.59, + "grad_norm": 0.455078125, + "learning_rate": 2.0071948852747348e-05, + "loss": 2.4224, + "step": 37323 + }, + { + "epoch": 1.59, + "grad_norm": 0.51171875, + "learning_rate": 2.0067917846640383e-05, + "loss": 2.4791, + "step": 37324 + }, + { + "epoch": 1.59, + "grad_norm": 0.46484375, + "learning_rate": 2.0063887200198528e-05, + "loss": 2.4808, + "step": 37325 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 2.0059856913440022e-05, + "loss": 2.4523, + "step": 37326 + }, + { + "epoch": 1.59, + "grad_norm": 0.50390625, + "learning_rate": 2.005582698638294e-05, + "loss": 2.3409, + "step": 37327 + }, + { + "epoch": 1.59, + "grad_norm": 0.4375, + "learning_rate": 2.0051797419045405e-05, + "loss": 2.3911, + "step": 37328 + }, + { + "epoch": 1.59, + "grad_norm": 0.423828125, + "learning_rate": 2.004776821144555e-05, + "loss": 2.4633, + "step": 37329 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 2.0043739363601553e-05, + "loss": 2.586, + "step": 37330 + }, + { + "epoch": 1.59, + "grad_norm": 0.5625, + "learning_rate": 2.003971087553148e-05, + "loss": 2.6017, + "step": 37331 + }, + { + "epoch": 1.59, + "grad_norm": 0.453125, + "learning_rate": 2.0035682747253516e-05, + "loss": 2.4689, + "step": 37332 + }, + { + "epoch": 1.59, + "grad_norm": 0.4921875, + "learning_rate": 2.0031654978785718e-05, + "loss": 2.4419, + "step": 37333 + }, + { + "epoch": 1.59, + "grad_norm": 0.4296875, + "learning_rate": 2.002762757014629e-05, + "loss": 2.2989, + "step": 37334 + }, + { + "epoch": 1.59, + "grad_norm": 0.51953125, + "learning_rate": 2.002360052135327e-05, + "loss": 2.477, + "step": 37335 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 2.0019573832424854e-05, + "loss": 2.4503, + "step": 37336 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 2.0015547503379107e-05, + "loss": 2.3964, + "step": 37337 + }, + { + "epoch": 1.59, + "grad_norm": 0.44140625, + "learning_rate": 2.0011521534234156e-05, + "loss": 2.4897, + "step": 37338 + }, + { + "epoch": 1.59, + "grad_norm": 0.4296875, + "learning_rate": 2.0007495925008145e-05, + "loss": 2.3535, + "step": 37339 + }, + { + "epoch": 1.59, + "grad_norm": 0.51171875, + "learning_rate": 2.0003470675719182e-05, + "loss": 2.5923, + "step": 37340 + }, + { + "epoch": 1.59, + "grad_norm": 0.4765625, + "learning_rate": 1.9999445786385307e-05, + "loss": 2.6468, + "step": 37341 + }, + { + "epoch": 1.59, + "grad_norm": 0.5390625, + "learning_rate": 1.999542125702475e-05, + "loss": 2.5546, + "step": 37342 + }, + { + "epoch": 1.59, + "grad_norm": 0.455078125, + "learning_rate": 1.999139708765555e-05, + "loss": 2.3704, + "step": 37343 + }, + { + "epoch": 1.59, + "grad_norm": 0.39453125, + "learning_rate": 1.9987373278295796e-05, + "loss": 2.4289, + "step": 37344 + }, + { + "epoch": 1.59, + "grad_norm": 0.51953125, + "learning_rate": 1.9983349828963617e-05, + "loss": 2.501, + "step": 37345 + }, + { + "epoch": 1.59, + "grad_norm": 0.458984375, + "learning_rate": 1.9979326739677164e-05, + "loss": 2.2172, + "step": 37346 + }, + { + "epoch": 1.59, + "grad_norm": 0.52734375, + "learning_rate": 1.9975304010454453e-05, + "loss": 2.5502, + "step": 37347 + }, + { + "epoch": 1.59, + "grad_norm": 0.435546875, + "learning_rate": 1.9971281641313667e-05, + "loss": 2.4185, + "step": 37348 + }, + { + "epoch": 1.59, + "grad_norm": 0.443359375, + "learning_rate": 1.9967259632272862e-05, + "loss": 2.4655, + "step": 37349 + }, + { + "epoch": 1.59, + "grad_norm": 0.44140625, + "learning_rate": 1.996323798335008e-05, + "loss": 2.2222, + "step": 37350 + }, + { + "epoch": 1.59, + "grad_norm": 0.451171875, + "learning_rate": 1.995921669456353e-05, + "loss": 2.6109, + "step": 37351 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 1.995519576593127e-05, + "loss": 2.5059, + "step": 37352 + }, + { + "epoch": 1.59, + "grad_norm": 0.49609375, + "learning_rate": 1.995117519747134e-05, + "loss": 2.4676, + "step": 37353 + }, + { + "epoch": 1.59, + "grad_norm": 0.435546875, + "learning_rate": 1.9947154989201866e-05, + "loss": 2.4104, + "step": 37354 + }, + { + "epoch": 1.59, + "grad_norm": 0.478515625, + "learning_rate": 1.9943135141140956e-05, + "loss": 2.3738, + "step": 37355 + }, + { + "epoch": 1.59, + "grad_norm": 0.392578125, + "learning_rate": 1.9939115653306695e-05, + "loss": 2.505, + "step": 37356 + }, + { + "epoch": 1.59, + "grad_norm": 0.4921875, + "learning_rate": 1.9935096525717122e-05, + "loss": 2.4553, + "step": 37357 + }, + { + "epoch": 1.59, + "grad_norm": 0.484375, + "learning_rate": 1.993107775839036e-05, + "loss": 2.3535, + "step": 37358 + }, + { + "epoch": 1.59, + "grad_norm": 0.49609375, + "learning_rate": 1.99270593513445e-05, + "loss": 2.4528, + "step": 37359 + }, + { + "epoch": 1.59, + "grad_norm": 0.46484375, + "learning_rate": 1.992304130459759e-05, + "loss": 2.3421, + "step": 37360 + }, + { + "epoch": 1.59, + "grad_norm": 0.52734375, + "learning_rate": 1.9919023618167765e-05, + "loss": 2.5071, + "step": 37361 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 1.9915006292073025e-05, + "loss": 2.4919, + "step": 37362 + }, + { + "epoch": 1.59, + "grad_norm": 0.390625, + "learning_rate": 1.9910989326331497e-05, + "loss": 2.4182, + "step": 37363 + }, + { + "epoch": 1.59, + "grad_norm": 0.478515625, + "learning_rate": 1.9906972720961283e-05, + "loss": 2.5067, + "step": 37364 + }, + { + "epoch": 1.59, + "grad_norm": 0.435546875, + "learning_rate": 1.9902956475980406e-05, + "loss": 2.4844, + "step": 37365 + }, + { + "epoch": 1.59, + "grad_norm": 0.47265625, + "learning_rate": 1.98989405914069e-05, + "loss": 2.5573, + "step": 37366 + }, + { + "epoch": 1.59, + "grad_norm": 0.44140625, + "learning_rate": 1.9894925067258953e-05, + "loss": 2.3776, + "step": 37367 + }, + { + "epoch": 1.59, + "grad_norm": 0.46484375, + "learning_rate": 1.9890909903554555e-05, + "loss": 2.4408, + "step": 37368 + }, + { + "epoch": 1.59, + "grad_norm": 0.5078125, + "learning_rate": 1.9886895100311775e-05, + "loss": 2.401, + "step": 37369 + }, + { + "epoch": 1.59, + "grad_norm": 0.44921875, + "learning_rate": 1.9882880657548676e-05, + "loss": 2.4193, + "step": 37370 + }, + { + "epoch": 1.59, + "grad_norm": 0.48828125, + "learning_rate": 1.9878866575283362e-05, + "loss": 2.4615, + "step": 37371 + }, + { + "epoch": 1.59, + "grad_norm": 0.427734375, + "learning_rate": 1.9874852853533875e-05, + "loss": 2.4552, + "step": 37372 + }, + { + "epoch": 1.59, + "grad_norm": 0.439453125, + "learning_rate": 1.987083949231824e-05, + "loss": 2.4515, + "step": 37373 + }, + { + "epoch": 1.59, + "grad_norm": 0.5078125, + "learning_rate": 1.9866826491654533e-05, + "loss": 2.5673, + "step": 37374 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 1.9862813851560858e-05, + "loss": 2.4921, + "step": 37375 + }, + { + "epoch": 1.59, + "grad_norm": 0.4375, + "learning_rate": 1.9858801572055207e-05, + "loss": 2.3921, + "step": 37376 + }, + { + "epoch": 1.59, + "grad_norm": 0.470703125, + "learning_rate": 1.9854789653155683e-05, + "loss": 2.338, + "step": 37377 + }, + { + "epoch": 1.59, + "grad_norm": 0.474609375, + "learning_rate": 1.9850778094880297e-05, + "loss": 2.4885, + "step": 37378 + }, + { + "epoch": 1.59, + "grad_norm": 0.443359375, + "learning_rate": 1.9846766897247105e-05, + "loss": 2.4352, + "step": 37379 + }, + { + "epoch": 1.59, + "grad_norm": 0.40625, + "learning_rate": 1.9842756060274203e-05, + "loss": 2.3496, + "step": 37380 + }, + { + "epoch": 1.59, + "grad_norm": 0.45703125, + "learning_rate": 1.983874558397961e-05, + "loss": 2.418, + "step": 37381 + }, + { + "epoch": 1.59, + "grad_norm": 0.49609375, + "learning_rate": 1.9834735468381306e-05, + "loss": 2.5798, + "step": 37382 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 1.9830725713497456e-05, + "loss": 2.4136, + "step": 37383 + }, + { + "epoch": 1.59, + "grad_norm": 0.482421875, + "learning_rate": 1.982671631934603e-05, + "loss": 2.427, + "step": 37384 + }, + { + "epoch": 1.59, + "grad_norm": 0.4296875, + "learning_rate": 1.9822707285945063e-05, + "loss": 2.4542, + "step": 37385 + }, + { + "epoch": 1.59, + "grad_norm": 0.4453125, + "learning_rate": 1.9818698613312615e-05, + "loss": 2.5194, + "step": 37386 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 1.9814690301466732e-05, + "loss": 2.4427, + "step": 37387 + }, + { + "epoch": 1.59, + "grad_norm": 0.5, + "learning_rate": 1.9810682350425446e-05, + "loss": 2.3484, + "step": 37388 + }, + { + "epoch": 1.59, + "grad_norm": 0.48046875, + "learning_rate": 1.9806674760206768e-05, + "loss": 2.5609, + "step": 37389 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 1.9802667530828734e-05, + "loss": 2.4551, + "step": 37390 + }, + { + "epoch": 1.59, + "grad_norm": 0.43359375, + "learning_rate": 1.979866066230942e-05, + "loss": 2.4423, + "step": 37391 + }, + { + "epoch": 1.59, + "grad_norm": 0.451171875, + "learning_rate": 1.9794654154666793e-05, + "loss": 2.3407, + "step": 37392 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 1.9790648007918933e-05, + "loss": 2.4783, + "step": 37393 + }, + { + "epoch": 1.59, + "grad_norm": 0.486328125, + "learning_rate": 1.978664222208382e-05, + "loss": 2.3251, + "step": 37394 + }, + { + "epoch": 1.59, + "grad_norm": 0.42578125, + "learning_rate": 1.97826367971795e-05, + "loss": 2.5205, + "step": 37395 + }, + { + "epoch": 1.59, + "grad_norm": 0.51953125, + "learning_rate": 1.977863173322403e-05, + "loss": 2.3858, + "step": 37396 + }, + { + "epoch": 1.59, + "grad_norm": 0.462890625, + "learning_rate": 1.9774627030235403e-05, + "loss": 2.3524, + "step": 37397 + }, + { + "epoch": 1.59, + "grad_norm": 0.494140625, + "learning_rate": 1.9770622688231576e-05, + "loss": 2.3683, + "step": 37398 + }, + { + "epoch": 1.59, + "grad_norm": 0.455078125, + "learning_rate": 1.976661870723068e-05, + "loss": 2.4768, + "step": 37399 + }, + { + "epoch": 1.59, + "grad_norm": 0.490234375, + "learning_rate": 1.976261508725068e-05, + "loss": 2.4284, + "step": 37400 + }, + { + "epoch": 1.59, + "grad_norm": 0.48046875, + "learning_rate": 1.975861182830956e-05, + "loss": 2.4336, + "step": 37401 + }, + { + "epoch": 1.59, + "grad_norm": 0.4609375, + "learning_rate": 1.9754608930425368e-05, + "loss": 2.3341, + "step": 37402 + }, + { + "epoch": 1.59, + "grad_norm": 0.52734375, + "learning_rate": 1.9750606393616123e-05, + "loss": 2.5828, + "step": 37403 + }, + { + "epoch": 1.59, + "grad_norm": 0.462890625, + "learning_rate": 1.97466042178998e-05, + "loss": 2.467, + "step": 37404 + }, + { + "epoch": 1.59, + "grad_norm": 0.4921875, + "learning_rate": 1.9742602403294465e-05, + "loss": 2.4922, + "step": 37405 + }, + { + "epoch": 1.59, + "grad_norm": 0.498046875, + "learning_rate": 1.9738600949818052e-05, + "loss": 2.4537, + "step": 37406 + }, + { + "epoch": 1.59, + "grad_norm": 0.4375, + "learning_rate": 1.9734599857488634e-05, + "loss": 2.4944, + "step": 37407 + }, + { + "epoch": 1.59, + "grad_norm": 0.41796875, + "learning_rate": 1.9730599126324157e-05, + "loss": 2.4807, + "step": 37408 + }, + { + "epoch": 1.59, + "grad_norm": 0.453125, + "learning_rate": 1.9726598756342674e-05, + "loss": 2.479, + "step": 37409 + }, + { + "epoch": 1.59, + "grad_norm": 0.51953125, + "learning_rate": 1.9722598747562125e-05, + "loss": 2.3603, + "step": 37410 + }, + { + "epoch": 1.59, + "grad_norm": 0.484375, + "learning_rate": 1.9718599100000556e-05, + "loss": 2.4541, + "step": 37411 + }, + { + "epoch": 1.59, + "grad_norm": 0.5234375, + "learning_rate": 1.971459981367597e-05, + "loss": 2.5507, + "step": 37412 + }, + { + "epoch": 1.59, + "grad_norm": 0.41796875, + "learning_rate": 1.9710600888606345e-05, + "loss": 2.4359, + "step": 37413 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 1.970660232480963e-05, + "loss": 2.5183, + "step": 37414 + }, + { + "epoch": 1.59, + "grad_norm": 0.455078125, + "learning_rate": 1.97026041223039e-05, + "loss": 2.4245, + "step": 37415 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 1.9698606281107112e-05, + "loss": 2.3606, + "step": 37416 + }, + { + "epoch": 1.59, + "grad_norm": 0.4609375, + "learning_rate": 1.9694608801237225e-05, + "loss": 2.448, + "step": 37417 + }, + { + "epoch": 1.59, + "grad_norm": 0.447265625, + "learning_rate": 1.9690611682712246e-05, + "loss": 2.3813, + "step": 37418 + }, + { + "epoch": 1.59, + "grad_norm": 0.416015625, + "learning_rate": 1.9686614925550195e-05, + "loss": 2.4386, + "step": 37419 + }, + { + "epoch": 1.59, + "grad_norm": 0.44921875, + "learning_rate": 1.9682618529768994e-05, + "loss": 2.4506, + "step": 37420 + }, + { + "epoch": 1.59, + "grad_norm": 0.44921875, + "learning_rate": 1.967862249538669e-05, + "loss": 2.4249, + "step": 37421 + }, + { + "epoch": 1.59, + "grad_norm": 0.42578125, + "learning_rate": 1.9674626822421206e-05, + "loss": 2.4654, + "step": 37422 + }, + { + "epoch": 1.59, + "grad_norm": 0.5078125, + "learning_rate": 1.9670631510890557e-05, + "loss": 2.4587, + "step": 37423 + }, + { + "epoch": 1.59, + "grad_norm": 0.45703125, + "learning_rate": 1.9666636560812702e-05, + "loss": 2.3564, + "step": 37424 + }, + { + "epoch": 1.59, + "grad_norm": 0.453125, + "learning_rate": 1.9662641972205632e-05, + "loss": 2.41, + "step": 37425 + }, + { + "epoch": 1.59, + "grad_norm": 0.45703125, + "learning_rate": 1.96586477450873e-05, + "loss": 2.3454, + "step": 37426 + }, + { + "epoch": 1.59, + "grad_norm": 0.466796875, + "learning_rate": 1.9654653879475693e-05, + "loss": 2.5054, + "step": 37427 + }, + { + "epoch": 1.59, + "grad_norm": 0.46875, + "learning_rate": 1.9650660375388795e-05, + "loss": 2.4503, + "step": 37428 + }, + { + "epoch": 1.59, + "grad_norm": 0.462890625, + "learning_rate": 1.9646667232844572e-05, + "loss": 2.4666, + "step": 37429 + }, + { + "epoch": 1.59, + "grad_norm": 0.48046875, + "learning_rate": 1.964267445186092e-05, + "loss": 2.4535, + "step": 37430 + }, + { + "epoch": 1.6, + "grad_norm": 0.4296875, + "learning_rate": 1.9638682032455924e-05, + "loss": 2.5274, + "step": 37431 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.963468997464749e-05, + "loss": 2.4133, + "step": 37432 + }, + { + "epoch": 1.6, + "grad_norm": 0.3984375, + "learning_rate": 1.9630698278453553e-05, + "loss": 2.3702, + "step": 37433 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.9626706943892103e-05, + "loss": 2.3645, + "step": 37434 + }, + { + "epoch": 1.6, + "grad_norm": 0.43359375, + "learning_rate": 1.962271597098112e-05, + "loss": 2.4421, + "step": 37435 + }, + { + "epoch": 1.6, + "grad_norm": 0.388671875, + "learning_rate": 1.9618725359738522e-05, + "loss": 2.3632, + "step": 37436 + }, + { + "epoch": 1.6, + "grad_norm": 0.49609375, + "learning_rate": 1.9614735110182303e-05, + "loss": 2.3943, + "step": 37437 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.9610745222330372e-05, + "loss": 2.4078, + "step": 37438 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.960675569620074e-05, + "loss": 2.4833, + "step": 37439 + }, + { + "epoch": 1.6, + "grad_norm": 0.44921875, + "learning_rate": 1.96027665318113e-05, + "loss": 2.3869, + "step": 37440 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.9598777729180052e-05, + "loss": 2.5977, + "step": 37441 + }, + { + "epoch": 1.6, + "grad_norm": 0.451171875, + "learning_rate": 1.95947892883249e-05, + "loss": 2.3989, + "step": 37442 + }, + { + "epoch": 1.6, + "grad_norm": 0.416015625, + "learning_rate": 1.9590801209263808e-05, + "loss": 2.3915, + "step": 37443 + }, + { + "epoch": 1.6, + "grad_norm": 0.431640625, + "learning_rate": 1.9586813492014757e-05, + "loss": 2.4185, + "step": 37444 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.9582826136595646e-05, + "loss": 2.3313, + "step": 37445 + }, + { + "epoch": 1.6, + "grad_norm": 0.486328125, + "learning_rate": 1.9578839143024386e-05, + "loss": 2.5285, + "step": 37446 + }, + { + "epoch": 1.6, + "grad_norm": 0.5, + "learning_rate": 1.9574852511319008e-05, + "loss": 2.5134, + "step": 37447 + }, + { + "epoch": 1.6, + "grad_norm": 0.455078125, + "learning_rate": 1.9570866241497397e-05, + "loss": 2.6313, + "step": 37448 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.9566880333577474e-05, + "loss": 2.4366, + "step": 37449 + }, + { + "epoch": 1.6, + "grad_norm": 0.482421875, + "learning_rate": 1.956289478757719e-05, + "loss": 2.5568, + "step": 37450 + }, + { + "epoch": 1.6, + "grad_norm": 0.4765625, + "learning_rate": 1.9558909603514507e-05, + "loss": 2.4573, + "step": 37451 + }, + { + "epoch": 1.6, + "grad_norm": 0.515625, + "learning_rate": 1.955492478140731e-05, + "loss": 2.4805, + "step": 37452 + }, + { + "epoch": 1.6, + "grad_norm": 0.46875, + "learning_rate": 1.9550940321273592e-05, + "loss": 2.4564, + "step": 37453 + }, + { + "epoch": 1.6, + "grad_norm": 0.474609375, + "learning_rate": 1.9546956223131196e-05, + "loss": 2.3658, + "step": 37454 + }, + { + "epoch": 1.6, + "grad_norm": 0.5078125, + "learning_rate": 1.9542972486998133e-05, + "loss": 2.4725, + "step": 37455 + }, + { + "epoch": 1.6, + "grad_norm": 0.546875, + "learning_rate": 1.953898911289226e-05, + "loss": 2.2469, + "step": 37456 + }, + { + "epoch": 1.6, + "grad_norm": 0.48046875, + "learning_rate": 1.9535006100831556e-05, + "loss": 2.4735, + "step": 37457 + }, + { + "epoch": 1.6, + "grad_norm": 0.56640625, + "learning_rate": 1.9531023450833895e-05, + "loss": 2.4125, + "step": 37458 + }, + { + "epoch": 1.6, + "grad_norm": 0.51953125, + "learning_rate": 1.9527041162917216e-05, + "loss": 2.4697, + "step": 37459 + }, + { + "epoch": 1.6, + "grad_norm": 0.4609375, + "learning_rate": 1.952305923709947e-05, + "loss": 2.4288, + "step": 37460 + }, + { + "epoch": 1.6, + "grad_norm": 0.41796875, + "learning_rate": 1.9519077673398543e-05, + "loss": 2.3417, + "step": 37461 + }, + { + "epoch": 1.6, + "grad_norm": 0.44140625, + "learning_rate": 1.9515096471832305e-05, + "loss": 2.3725, + "step": 37462 + }, + { + "epoch": 1.6, + "grad_norm": 0.408203125, + "learning_rate": 1.9511115632418774e-05, + "loss": 2.5294, + "step": 37463 + }, + { + "epoch": 1.6, + "grad_norm": 0.447265625, + "learning_rate": 1.9507135155175803e-05, + "loss": 2.471, + "step": 37464 + }, + { + "epoch": 1.6, + "grad_norm": 0.46875, + "learning_rate": 1.950315504012129e-05, + "loss": 2.4455, + "step": 37465 + }, + { + "epoch": 1.6, + "grad_norm": 0.42578125, + "learning_rate": 1.949917528727315e-05, + "loss": 2.3343, + "step": 37466 + }, + { + "epoch": 1.6, + "grad_norm": 0.427734375, + "learning_rate": 1.949519589664933e-05, + "loss": 2.4215, + "step": 37467 + }, + { + "epoch": 1.6, + "grad_norm": 0.470703125, + "learning_rate": 1.949121686826768e-05, + "loss": 2.2923, + "step": 37468 + }, + { + "epoch": 1.6, + "grad_norm": 0.45703125, + "learning_rate": 1.9487238202146164e-05, + "loss": 2.4171, + "step": 37469 + }, + { + "epoch": 1.6, + "grad_norm": 0.419921875, + "learning_rate": 1.948325989830262e-05, + "loss": 2.2206, + "step": 37470 + }, + { + "epoch": 1.6, + "grad_norm": 0.498046875, + "learning_rate": 1.9479281956755013e-05, + "loss": 2.4208, + "step": 37471 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.9475304377521176e-05, + "loss": 2.37, + "step": 37472 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.9471327160619078e-05, + "loss": 2.5115, + "step": 37473 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.9467350306066546e-05, + "loss": 2.4904, + "step": 37474 + }, + { + "epoch": 1.6, + "grad_norm": 0.427734375, + "learning_rate": 1.94633738138815e-05, + "loss": 2.4672, + "step": 37475 + }, + { + "epoch": 1.6, + "grad_norm": 0.5, + "learning_rate": 1.945939768408187e-05, + "loss": 2.4841, + "step": 37476 + }, + { + "epoch": 1.6, + "grad_norm": 0.455078125, + "learning_rate": 1.9455421916685525e-05, + "loss": 2.4908, + "step": 37477 + }, + { + "epoch": 1.6, + "grad_norm": 0.494140625, + "learning_rate": 1.9451446511710304e-05, + "loss": 2.4604, + "step": 37478 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.944747146917415e-05, + "loss": 2.4738, + "step": 37479 + }, + { + "epoch": 1.6, + "grad_norm": 0.474609375, + "learning_rate": 1.9443496789094952e-05, + "loss": 2.452, + "step": 37480 + }, + { + "epoch": 1.6, + "grad_norm": 0.498046875, + "learning_rate": 1.9439522471490556e-05, + "loss": 2.4707, + "step": 37481 + }, + { + "epoch": 1.6, + "grad_norm": 0.470703125, + "learning_rate": 1.94355485163789e-05, + "loss": 2.4637, + "step": 37482 + }, + { + "epoch": 1.6, + "grad_norm": 0.484375, + "learning_rate": 1.9431574923777796e-05, + "loss": 2.5339, + "step": 37483 + }, + { + "epoch": 1.6, + "grad_norm": 0.5, + "learning_rate": 1.9427601693705167e-05, + "loss": 2.5814, + "step": 37484 + }, + { + "epoch": 1.6, + "grad_norm": 0.478515625, + "learning_rate": 1.942362882617891e-05, + "loss": 2.5473, + "step": 37485 + }, + { + "epoch": 1.6, + "grad_norm": 0.45703125, + "learning_rate": 1.9419656321216872e-05, + "loss": 2.3099, + "step": 37486 + }, + { + "epoch": 1.6, + "grad_norm": 0.482421875, + "learning_rate": 1.9415684178836913e-05, + "loss": 2.5023, + "step": 37487 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.9411712399056915e-05, + "loss": 2.4332, + "step": 37488 + }, + { + "epoch": 1.6, + "grad_norm": 0.470703125, + "learning_rate": 1.9407740981894795e-05, + "loss": 2.5281, + "step": 37489 + }, + { + "epoch": 1.6, + "grad_norm": 0.47265625, + "learning_rate": 1.9403769927368355e-05, + "loss": 2.4047, + "step": 37490 + }, + { + "epoch": 1.6, + "grad_norm": 0.470703125, + "learning_rate": 1.9399799235495496e-05, + "loss": 2.4946, + "step": 37491 + }, + { + "epoch": 1.6, + "grad_norm": 0.5546875, + "learning_rate": 1.939582890629411e-05, + "loss": 2.4199, + "step": 37492 + }, + { + "epoch": 1.6, + "grad_norm": 0.4453125, + "learning_rate": 1.9391858939782047e-05, + "loss": 2.4252, + "step": 37493 + }, + { + "epoch": 1.6, + "grad_norm": 0.478515625, + "learning_rate": 1.9387889335977127e-05, + "loss": 2.5301, + "step": 37494 + }, + { + "epoch": 1.6, + "grad_norm": 0.4140625, + "learning_rate": 1.9383920094897245e-05, + "loss": 2.4452, + "step": 37495 + }, + { + "epoch": 1.6, + "grad_norm": 0.455078125, + "learning_rate": 1.9379951216560277e-05, + "loss": 2.3842, + "step": 37496 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.9375982700984053e-05, + "loss": 2.2876, + "step": 37497 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.9372014548186457e-05, + "loss": 2.5496, + "step": 37498 + }, + { + "epoch": 1.6, + "grad_norm": 0.46484375, + "learning_rate": 1.936804675818531e-05, + "loss": 2.4721, + "step": 37499 + }, + { + "epoch": 1.6, + "grad_norm": 0.486328125, + "learning_rate": 1.9364079330998486e-05, + "loss": 2.4422, + "step": 37500 + }, + { + "epoch": 1.6, + "grad_norm": 0.42578125, + "learning_rate": 1.9360112266643858e-05, + "loss": 2.3612, + "step": 37501 + }, + { + "epoch": 1.6, + "grad_norm": 0.515625, + "learning_rate": 1.935614556513926e-05, + "loss": 2.4635, + "step": 37502 + }, + { + "epoch": 1.6, + "grad_norm": 0.427734375, + "learning_rate": 1.935217922650251e-05, + "loss": 2.5611, + "step": 37503 + }, + { + "epoch": 1.6, + "grad_norm": 0.458984375, + "learning_rate": 1.9348213250751467e-05, + "loss": 2.389, + "step": 37504 + }, + { + "epoch": 1.6, + "grad_norm": 0.447265625, + "learning_rate": 1.9344247637904035e-05, + "loss": 2.4744, + "step": 37505 + }, + { + "epoch": 1.6, + "grad_norm": 0.42578125, + "learning_rate": 1.934028238797797e-05, + "loss": 2.4411, + "step": 37506 + }, + { + "epoch": 1.6, + "grad_norm": 0.486328125, + "learning_rate": 1.9336317500991165e-05, + "loss": 2.3805, + "step": 37507 + }, + { + "epoch": 1.6, + "grad_norm": 0.51171875, + "learning_rate": 1.9332352976961476e-05, + "loss": 2.5293, + "step": 37508 + }, + { + "epoch": 1.6, + "grad_norm": 0.458984375, + "learning_rate": 1.9328388815906716e-05, + "loss": 2.4292, + "step": 37509 + }, + { + "epoch": 1.6, + "grad_norm": 0.4375, + "learning_rate": 1.9324425017844693e-05, + "loss": 2.534, + "step": 37510 + }, + { + "epoch": 1.6, + "grad_norm": 0.4453125, + "learning_rate": 1.932046158279328e-05, + "loss": 2.4543, + "step": 37511 + }, + { + "epoch": 1.6, + "grad_norm": 0.44921875, + "learning_rate": 1.9316498510770332e-05, + "loss": 2.3412, + "step": 37512 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.931253580179363e-05, + "loss": 2.5496, + "step": 37513 + }, + { + "epoch": 1.6, + "grad_norm": 0.4765625, + "learning_rate": 1.9308573455881053e-05, + "loss": 2.4429, + "step": 37514 + }, + { + "epoch": 1.6, + "grad_norm": 0.51171875, + "learning_rate": 1.9304611473050372e-05, + "loss": 2.5208, + "step": 37515 + }, + { + "epoch": 1.6, + "grad_norm": 0.46484375, + "learning_rate": 1.9300649853319463e-05, + "loss": 2.5211, + "step": 37516 + }, + { + "epoch": 1.6, + "grad_norm": 0.46875, + "learning_rate": 1.9296688596706158e-05, + "loss": 2.5249, + "step": 37517 + }, + { + "epoch": 1.6, + "grad_norm": 0.43359375, + "learning_rate": 1.9292727703228254e-05, + "loss": 2.4743, + "step": 37518 + }, + { + "epoch": 1.6, + "grad_norm": 0.4375, + "learning_rate": 1.9288767172903555e-05, + "loss": 2.5354, + "step": 37519 + }, + { + "epoch": 1.6, + "grad_norm": 0.52734375, + "learning_rate": 1.9284807005749915e-05, + "loss": 2.4432, + "step": 37520 + }, + { + "epoch": 1.6, + "grad_norm": 0.51953125, + "learning_rate": 1.9280847201785168e-05, + "loss": 2.5267, + "step": 37521 + }, + { + "epoch": 1.6, + "grad_norm": 0.53515625, + "learning_rate": 1.9276887761027085e-05, + "loss": 2.3809, + "step": 37522 + }, + { + "epoch": 1.6, + "grad_norm": 0.57421875, + "learning_rate": 1.9272928683493497e-05, + "loss": 2.4416, + "step": 37523 + }, + { + "epoch": 1.6, + "grad_norm": 0.49609375, + "learning_rate": 1.9268969969202268e-05, + "loss": 2.5156, + "step": 37524 + }, + { + "epoch": 1.6, + "grad_norm": 0.482421875, + "learning_rate": 1.926501161817116e-05, + "loss": 2.5917, + "step": 37525 + }, + { + "epoch": 1.6, + "grad_norm": 0.427734375, + "learning_rate": 1.9261053630417968e-05, + "loss": 2.3337, + "step": 37526 + }, + { + "epoch": 1.6, + "grad_norm": 0.443359375, + "learning_rate": 1.9257096005960528e-05, + "loss": 2.4575, + "step": 37527 + }, + { + "epoch": 1.6, + "grad_norm": 0.4375, + "learning_rate": 1.9253138744816678e-05, + "loss": 2.4496, + "step": 37528 + }, + { + "epoch": 1.6, + "grad_norm": 0.49609375, + "learning_rate": 1.9249181847004172e-05, + "loss": 2.4561, + "step": 37529 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.924522531254086e-05, + "loss": 2.4632, + "step": 37530 + }, + { + "epoch": 1.6, + "grad_norm": 0.48828125, + "learning_rate": 1.9241269141444497e-05, + "loss": 2.4788, + "step": 37531 + }, + { + "epoch": 1.6, + "grad_norm": 0.55859375, + "learning_rate": 1.9237313333732897e-05, + "loss": 2.4791, + "step": 37532 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.9233357889423907e-05, + "loss": 2.3738, + "step": 37533 + }, + { + "epoch": 1.6, + "grad_norm": 0.486328125, + "learning_rate": 1.922940280853529e-05, + "loss": 2.3756, + "step": 37534 + }, + { + "epoch": 1.6, + "grad_norm": 0.51953125, + "learning_rate": 1.922544809108482e-05, + "loss": 2.4134, + "step": 37535 + }, + { + "epoch": 1.6, + "grad_norm": 0.4921875, + "learning_rate": 1.92214937370903e-05, + "loss": 2.4357, + "step": 37536 + }, + { + "epoch": 1.6, + "grad_norm": 0.443359375, + "learning_rate": 1.9217539746569568e-05, + "loss": 2.4746, + "step": 37537 + }, + { + "epoch": 1.6, + "grad_norm": 0.474609375, + "learning_rate": 1.921358611954037e-05, + "loss": 2.391, + "step": 37538 + }, + { + "epoch": 1.6, + "grad_norm": 0.51953125, + "learning_rate": 1.92096328560205e-05, + "loss": 2.5757, + "step": 37539 + }, + { + "epoch": 1.6, + "grad_norm": 0.4609375, + "learning_rate": 1.920567995602779e-05, + "loss": 2.5422, + "step": 37540 + }, + { + "epoch": 1.6, + "grad_norm": 0.46875, + "learning_rate": 1.9201727419579983e-05, + "loss": 2.4543, + "step": 37541 + }, + { + "epoch": 1.6, + "grad_norm": 0.5078125, + "learning_rate": 1.919777524669485e-05, + "loss": 2.3664, + "step": 37542 + }, + { + "epoch": 1.6, + "grad_norm": 0.4296875, + "learning_rate": 1.91938234373902e-05, + "loss": 2.3569, + "step": 37543 + }, + { + "epoch": 1.6, + "grad_norm": 0.478515625, + "learning_rate": 1.9189871991683848e-05, + "loss": 2.3799, + "step": 37544 + }, + { + "epoch": 1.6, + "grad_norm": 0.47265625, + "learning_rate": 1.9185920909593503e-05, + "loss": 2.4205, + "step": 37545 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.918197019113701e-05, + "loss": 2.4386, + "step": 37546 + }, + { + "epoch": 1.6, + "grad_norm": 0.494140625, + "learning_rate": 1.9178019836332096e-05, + "loss": 2.3163, + "step": 37547 + }, + { + "epoch": 1.6, + "grad_norm": 0.53125, + "learning_rate": 1.917406984519654e-05, + "loss": 2.5365, + "step": 37548 + }, + { + "epoch": 1.6, + "grad_norm": 0.451171875, + "learning_rate": 1.9170120217748165e-05, + "loss": 2.4636, + "step": 37549 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.9166170954004715e-05, + "loss": 2.4269, + "step": 37550 + }, + { + "epoch": 1.6, + "grad_norm": 0.484375, + "learning_rate": 1.9162222053983925e-05, + "loss": 2.5857, + "step": 37551 + }, + { + "epoch": 1.6, + "grad_norm": 0.478515625, + "learning_rate": 1.9158273517703585e-05, + "loss": 2.5216, + "step": 37552 + }, + { + "epoch": 1.6, + "grad_norm": 0.458984375, + "learning_rate": 1.91543253451815e-05, + "loss": 2.4546, + "step": 37553 + }, + { + "epoch": 1.6, + "grad_norm": 0.4609375, + "learning_rate": 1.9150377536435384e-05, + "loss": 2.3139, + "step": 37554 + }, + { + "epoch": 1.6, + "grad_norm": 0.5078125, + "learning_rate": 1.9146430091483027e-05, + "loss": 2.5617, + "step": 37555 + }, + { + "epoch": 1.6, + "grad_norm": 0.51171875, + "learning_rate": 1.9142483010342193e-05, + "loss": 2.4321, + "step": 37556 + }, + { + "epoch": 1.6, + "grad_norm": 0.47265625, + "learning_rate": 1.913853629303065e-05, + "loss": 2.4002, + "step": 37557 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.9134589939566118e-05, + "loss": 2.3295, + "step": 37558 + }, + { + "epoch": 1.6, + "grad_norm": 0.515625, + "learning_rate": 1.9130643949966377e-05, + "loss": 2.5349, + "step": 37559 + }, + { + "epoch": 1.6, + "grad_norm": 0.482421875, + "learning_rate": 1.9126698324249205e-05, + "loss": 2.4863, + "step": 37560 + }, + { + "epoch": 1.6, + "grad_norm": 0.546875, + "learning_rate": 1.912275306243231e-05, + "loss": 2.4493, + "step": 37561 + }, + { + "epoch": 1.6, + "grad_norm": 0.421875, + "learning_rate": 1.9118808164533486e-05, + "loss": 2.3789, + "step": 37562 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.9114863630570457e-05, + "loss": 2.3954, + "step": 37563 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.911091946056097e-05, + "loss": 2.4436, + "step": 37564 + }, + { + "epoch": 1.6, + "grad_norm": 0.4375, + "learning_rate": 1.9106975654522806e-05, + "loss": 2.3288, + "step": 37565 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.9103032212473703e-05, + "loss": 2.3754, + "step": 37566 + }, + { + "epoch": 1.6, + "grad_norm": 0.4609375, + "learning_rate": 1.9099089134431357e-05, + "loss": 2.4044, + "step": 37567 + }, + { + "epoch": 1.6, + "grad_norm": 0.46484375, + "learning_rate": 1.9095146420413547e-05, + "loss": 2.3453, + "step": 37568 + }, + { + "epoch": 1.6, + "grad_norm": 0.47265625, + "learning_rate": 1.909120407043803e-05, + "loss": 2.4189, + "step": 37569 + }, + { + "epoch": 1.6, + "grad_norm": 0.443359375, + "learning_rate": 1.9087262084522506e-05, + "loss": 2.4265, + "step": 37570 + }, + { + "epoch": 1.6, + "grad_norm": 0.44921875, + "learning_rate": 1.908332046268473e-05, + "loss": 2.4713, + "step": 37571 + }, + { + "epoch": 1.6, + "grad_norm": 0.44140625, + "learning_rate": 1.9079379204942473e-05, + "loss": 2.434, + "step": 37572 + }, + { + "epoch": 1.6, + "grad_norm": 0.5234375, + "learning_rate": 1.9075438311313422e-05, + "loss": 2.4068, + "step": 37573 + }, + { + "epoch": 1.6, + "grad_norm": 0.46484375, + "learning_rate": 1.9071497781815307e-05, + "loss": 2.5406, + "step": 37574 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.9067557616465882e-05, + "loss": 2.614, + "step": 37575 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.9063617815282884e-05, + "loss": 2.568, + "step": 37576 + }, + { + "epoch": 1.6, + "grad_norm": 0.42578125, + "learning_rate": 1.9059678378284006e-05, + "loss": 2.4223, + "step": 37577 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.905573930548703e-05, + "loss": 2.4156, + "step": 37578 + }, + { + "epoch": 1.6, + "grad_norm": 0.419921875, + "learning_rate": 1.905180059690961e-05, + "loss": 2.3971, + "step": 37579 + }, + { + "epoch": 1.6, + "grad_norm": 0.52734375, + "learning_rate": 1.9047862252569504e-05, + "loss": 2.5861, + "step": 37580 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.904392427248446e-05, + "loss": 2.3955, + "step": 37581 + }, + { + "epoch": 1.6, + "grad_norm": 0.423828125, + "learning_rate": 1.903998665667217e-05, + "loss": 2.4187, + "step": 37582 + }, + { + "epoch": 1.6, + "grad_norm": 0.48046875, + "learning_rate": 1.9036049405150325e-05, + "loss": 2.5149, + "step": 37583 + }, + { + "epoch": 1.6, + "grad_norm": 0.53125, + "learning_rate": 1.9032112517936673e-05, + "loss": 2.3733, + "step": 37584 + }, + { + "epoch": 1.6, + "grad_norm": 0.455078125, + "learning_rate": 1.9028175995048946e-05, + "loss": 2.4091, + "step": 37585 + }, + { + "epoch": 1.6, + "grad_norm": 0.515625, + "learning_rate": 1.902423983650482e-05, + "loss": 2.383, + "step": 37586 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.9020304042322014e-05, + "loss": 2.4741, + "step": 37587 + }, + { + "epoch": 1.6, + "grad_norm": 0.494140625, + "learning_rate": 1.9016368612518265e-05, + "loss": 2.5194, + "step": 37588 + }, + { + "epoch": 1.6, + "grad_norm": 0.498046875, + "learning_rate": 1.9012433547111264e-05, + "loss": 2.3711, + "step": 37589 + }, + { + "epoch": 1.6, + "grad_norm": 0.46875, + "learning_rate": 1.9008498846118694e-05, + "loss": 2.5062, + "step": 37590 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.9004564509558276e-05, + "loss": 2.4827, + "step": 37591 + }, + { + "epoch": 1.6, + "grad_norm": 0.392578125, + "learning_rate": 1.900063053744775e-05, + "loss": 2.4354, + "step": 37592 + }, + { + "epoch": 1.6, + "grad_norm": 0.53125, + "learning_rate": 1.8996696929804746e-05, + "loss": 2.6319, + "step": 37593 + }, + { + "epoch": 1.6, + "grad_norm": 0.46875, + "learning_rate": 1.8992763686647043e-05, + "loss": 2.4823, + "step": 37594 + }, + { + "epoch": 1.6, + "grad_norm": 0.498046875, + "learning_rate": 1.8988830807992286e-05, + "loss": 2.3981, + "step": 37595 + }, + { + "epoch": 1.6, + "grad_norm": 0.443359375, + "learning_rate": 1.8984898293858145e-05, + "loss": 2.523, + "step": 37596 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.898096614426239e-05, + "loss": 2.4448, + "step": 37597 + }, + { + "epoch": 1.6, + "grad_norm": 0.498046875, + "learning_rate": 1.897703435922269e-05, + "loss": 2.5793, + "step": 37598 + }, + { + "epoch": 1.6, + "grad_norm": 0.458984375, + "learning_rate": 1.8973102938756693e-05, + "loss": 2.3785, + "step": 37599 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.896917188288213e-05, + "loss": 2.5164, + "step": 37600 + }, + { + "epoch": 1.6, + "grad_norm": 0.439453125, + "learning_rate": 1.89652411916167e-05, + "loss": 2.6124, + "step": 37601 + }, + { + "epoch": 1.6, + "grad_norm": 0.56640625, + "learning_rate": 1.8961310864978043e-05, + "loss": 2.4856, + "step": 37602 + }, + { + "epoch": 1.6, + "grad_norm": 0.455078125, + "learning_rate": 1.8957380902983903e-05, + "loss": 2.5087, + "step": 37603 + }, + { + "epoch": 1.6, + "grad_norm": 0.451171875, + "learning_rate": 1.8953451305651893e-05, + "loss": 2.577, + "step": 37604 + }, + { + "epoch": 1.6, + "grad_norm": 0.435546875, + "learning_rate": 1.894952207299977e-05, + "loss": 2.3611, + "step": 37605 + }, + { + "epoch": 1.6, + "grad_norm": 0.47265625, + "learning_rate": 1.8945593205045143e-05, + "loss": 2.4012, + "step": 37606 + }, + { + "epoch": 1.6, + "grad_norm": 0.55859375, + "learning_rate": 1.894166470180575e-05, + "loss": 2.6265, + "step": 37607 + }, + { + "epoch": 1.6, + "grad_norm": 0.474609375, + "learning_rate": 1.8937736563299226e-05, + "loss": 2.5818, + "step": 37608 + }, + { + "epoch": 1.6, + "grad_norm": 0.458984375, + "learning_rate": 1.8933808789543263e-05, + "loss": 2.2264, + "step": 37609 + }, + { + "epoch": 1.6, + "grad_norm": 0.484375, + "learning_rate": 1.8929881380555546e-05, + "loss": 2.6057, + "step": 37610 + }, + { + "epoch": 1.6, + "grad_norm": 0.4609375, + "learning_rate": 1.8925954336353746e-05, + "loss": 2.4024, + "step": 37611 + }, + { + "epoch": 1.6, + "grad_norm": 0.46484375, + "learning_rate": 1.8922027656955464e-05, + "loss": 2.4383, + "step": 37612 + }, + { + "epoch": 1.6, + "grad_norm": 0.498046875, + "learning_rate": 1.891810134237849e-05, + "loss": 2.6538, + "step": 37613 + }, + { + "epoch": 1.6, + "grad_norm": 0.40625, + "learning_rate": 1.891417539264042e-05, + "loss": 2.431, + "step": 37614 + }, + { + "epoch": 1.6, + "grad_norm": 0.474609375, + "learning_rate": 1.8910249807758894e-05, + "loss": 2.5362, + "step": 37615 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.8906324587751623e-05, + "loss": 2.5001, + "step": 37616 + }, + { + "epoch": 1.6, + "grad_norm": 0.4453125, + "learning_rate": 1.8902399732636277e-05, + "loss": 2.5587, + "step": 37617 + }, + { + "epoch": 1.6, + "grad_norm": 0.51171875, + "learning_rate": 1.889847524243047e-05, + "loss": 2.4306, + "step": 37618 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.8894551117151903e-05, + "loss": 2.4436, + "step": 37619 + }, + { + "epoch": 1.6, + "grad_norm": 0.5234375, + "learning_rate": 1.889062735681819e-05, + "loss": 2.4506, + "step": 37620 + }, + { + "epoch": 1.6, + "grad_norm": 0.46875, + "learning_rate": 1.888670396144705e-05, + "loss": 2.3487, + "step": 37621 + }, + { + "epoch": 1.6, + "grad_norm": 0.451171875, + "learning_rate": 1.8882780931056054e-05, + "loss": 2.4364, + "step": 37622 + }, + { + "epoch": 1.6, + "grad_norm": 0.484375, + "learning_rate": 1.8878858265662948e-05, + "loss": 2.3729, + "step": 37623 + }, + { + "epoch": 1.6, + "grad_norm": 0.4765625, + "learning_rate": 1.887493596528529e-05, + "loss": 2.4201, + "step": 37624 + }, + { + "epoch": 1.6, + "grad_norm": 0.455078125, + "learning_rate": 1.887101402994077e-05, + "loss": 2.4533, + "step": 37625 + }, + { + "epoch": 1.6, + "grad_norm": 0.494140625, + "learning_rate": 1.8867092459647073e-05, + "loss": 2.3824, + "step": 37626 + }, + { + "epoch": 1.6, + "grad_norm": 0.474609375, + "learning_rate": 1.886317125442181e-05, + "loss": 2.4521, + "step": 37627 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.885925041428257e-05, + "loss": 2.5689, + "step": 37628 + }, + { + "epoch": 1.6, + "grad_norm": 0.439453125, + "learning_rate": 1.88553299392471e-05, + "loss": 2.4214, + "step": 37629 + }, + { + "epoch": 1.6, + "grad_norm": 0.5546875, + "learning_rate": 1.885140982933299e-05, + "loss": 2.4505, + "step": 37630 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.884749008455785e-05, + "loss": 2.5324, + "step": 37631 + }, + { + "epoch": 1.6, + "grad_norm": 0.53125, + "learning_rate": 1.884357070493935e-05, + "loss": 2.5219, + "step": 37632 + }, + { + "epoch": 1.6, + "grad_norm": 0.400390625, + "learning_rate": 1.883965169049515e-05, + "loss": 2.339, + "step": 37633 + }, + { + "epoch": 1.6, + "grad_norm": 0.44140625, + "learning_rate": 1.883573304124283e-05, + "loss": 2.3264, + "step": 37634 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.8831814757200085e-05, + "loss": 2.495, + "step": 37635 + }, + { + "epoch": 1.6, + "grad_norm": 0.5, + "learning_rate": 1.8827896838384472e-05, + "loss": 2.3443, + "step": 37636 + }, + { + "epoch": 1.6, + "grad_norm": 0.455078125, + "learning_rate": 1.8823979284813687e-05, + "loss": 2.4989, + "step": 37637 + }, + { + "epoch": 1.6, + "grad_norm": 0.453125, + "learning_rate": 1.88200620965053e-05, + "loss": 2.4178, + "step": 37638 + }, + { + "epoch": 1.6, + "grad_norm": 0.5078125, + "learning_rate": 1.8816145273476992e-05, + "loss": 2.4721, + "step": 37639 + }, + { + "epoch": 1.6, + "grad_norm": 0.46484375, + "learning_rate": 1.8812228815746334e-05, + "loss": 2.374, + "step": 37640 + }, + { + "epoch": 1.6, + "grad_norm": 0.484375, + "learning_rate": 1.880831272333099e-05, + "loss": 2.5449, + "step": 37641 + }, + { + "epoch": 1.6, + "grad_norm": 0.44921875, + "learning_rate": 1.8804396996248573e-05, + "loss": 2.4217, + "step": 37642 + }, + { + "epoch": 1.6, + "grad_norm": 0.482421875, + "learning_rate": 1.880048163451671e-05, + "loss": 2.4304, + "step": 37643 + }, + { + "epoch": 1.6, + "grad_norm": 0.4296875, + "learning_rate": 1.8796566638152945e-05, + "loss": 2.3113, + "step": 37644 + }, + { + "epoch": 1.6, + "grad_norm": 0.447265625, + "learning_rate": 1.8792652007175005e-05, + "loss": 2.4678, + "step": 37645 + }, + { + "epoch": 1.6, + "grad_norm": 0.5234375, + "learning_rate": 1.878873774160046e-05, + "loss": 2.5288, + "step": 37646 + }, + { + "epoch": 1.6, + "grad_norm": 0.48828125, + "learning_rate": 1.8784823841446887e-05, + "loss": 2.4962, + "step": 37647 + }, + { + "epoch": 1.6, + "grad_norm": 0.466796875, + "learning_rate": 1.8780910306731924e-05, + "loss": 2.4833, + "step": 37648 + }, + { + "epoch": 1.6, + "grad_norm": 0.419921875, + "learning_rate": 1.8776997137473205e-05, + "loss": 2.533, + "step": 37649 + }, + { + "epoch": 1.6, + "grad_norm": 0.44140625, + "learning_rate": 1.8773084333688295e-05, + "loss": 2.3485, + "step": 37650 + }, + { + "epoch": 1.6, + "grad_norm": 0.4375, + "learning_rate": 1.8769171895394843e-05, + "loss": 2.3743, + "step": 37651 + }, + { + "epoch": 1.6, + "grad_norm": 0.443359375, + "learning_rate": 1.8765259822610405e-05, + "loss": 2.4345, + "step": 37652 + }, + { + "epoch": 1.6, + "grad_norm": 0.43359375, + "learning_rate": 1.876134811535264e-05, + "loss": 2.4214, + "step": 37653 + }, + { + "epoch": 1.6, + "grad_norm": 0.47265625, + "learning_rate": 1.8757436773639094e-05, + "loss": 2.4199, + "step": 37654 + }, + { + "epoch": 1.6, + "grad_norm": 0.416015625, + "learning_rate": 1.8753525797487416e-05, + "loss": 2.5727, + "step": 37655 + }, + { + "epoch": 1.6, + "grad_norm": 0.478515625, + "learning_rate": 1.874961518691515e-05, + "loss": 2.4691, + "step": 37656 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.8745704941939923e-05, + "loss": 2.4869, + "step": 37657 + }, + { + "epoch": 1.6, + "grad_norm": 0.50390625, + "learning_rate": 1.874179506257936e-05, + "loss": 2.3979, + "step": 37658 + }, + { + "epoch": 1.6, + "grad_norm": 0.48046875, + "learning_rate": 1.873788554885101e-05, + "loss": 2.5068, + "step": 37659 + }, + { + "epoch": 1.6, + "grad_norm": 0.474609375, + "learning_rate": 1.8733976400772434e-05, + "loss": 2.4179, + "step": 37660 + }, + { + "epoch": 1.6, + "grad_norm": 0.49609375, + "learning_rate": 1.8730067618361324e-05, + "loss": 2.4975, + "step": 37661 + }, + { + "epoch": 1.6, + "grad_norm": 0.462890625, + "learning_rate": 1.8726159201635195e-05, + "loss": 2.5121, + "step": 37662 + }, + { + "epoch": 1.6, + "grad_norm": 0.5, + "learning_rate": 1.8722251150611634e-05, + "loss": 2.5955, + "step": 37663 + }, + { + "epoch": 1.6, + "grad_norm": 0.5078125, + "learning_rate": 1.871834346530823e-05, + "loss": 2.4975, + "step": 37664 + }, + { + "epoch": 1.61, + "grad_norm": 0.55078125, + "learning_rate": 1.871443614574261e-05, + "loss": 2.4535, + "step": 37665 + }, + { + "epoch": 1.61, + "grad_norm": 0.455078125, + "learning_rate": 1.871052919193228e-05, + "loss": 2.5964, + "step": 37666 + }, + { + "epoch": 1.61, + "grad_norm": 0.474609375, + "learning_rate": 1.8706622603894896e-05, + "loss": 2.6736, + "step": 37667 + }, + { + "epoch": 1.61, + "grad_norm": 0.427734375, + "learning_rate": 1.8702716381647977e-05, + "loss": 2.4114, + "step": 37668 + }, + { + "epoch": 1.61, + "grad_norm": 0.435546875, + "learning_rate": 1.8698810525209154e-05, + "loss": 2.417, + "step": 37669 + }, + { + "epoch": 1.61, + "grad_norm": 0.466796875, + "learning_rate": 1.8694905034595932e-05, + "loss": 2.3947, + "step": 37670 + }, + { + "epoch": 1.61, + "grad_norm": 0.470703125, + "learning_rate": 1.8690999909825958e-05, + "loss": 2.3724, + "step": 37671 + }, + { + "epoch": 1.61, + "grad_norm": 0.50390625, + "learning_rate": 1.8687095150916744e-05, + "loss": 2.4058, + "step": 37672 + }, + { + "epoch": 1.61, + "grad_norm": 0.453125, + "learning_rate": 1.868319075788588e-05, + "loss": 2.3713, + "step": 37673 + }, + { + "epoch": 1.61, + "grad_norm": 0.51171875, + "learning_rate": 1.8679286730750978e-05, + "loss": 2.3944, + "step": 37674 + }, + { + "epoch": 1.61, + "grad_norm": 0.40234375, + "learning_rate": 1.867538306952955e-05, + "loss": 2.3858, + "step": 37675 + }, + { + "epoch": 1.61, + "grad_norm": 0.474609375, + "learning_rate": 1.867147977423913e-05, + "loss": 2.5411, + "step": 37676 + }, + { + "epoch": 1.61, + "grad_norm": 0.50390625, + "learning_rate": 1.8667576844897396e-05, + "loss": 2.4622, + "step": 37677 + }, + { + "epoch": 1.61, + "grad_norm": 0.5, + "learning_rate": 1.8663674281521826e-05, + "loss": 2.5563, + "step": 37678 + }, + { + "epoch": 1.61, + "grad_norm": 0.51171875, + "learning_rate": 1.865977208412998e-05, + "loss": 2.4027, + "step": 37679 + }, + { + "epoch": 1.61, + "grad_norm": 0.48046875, + "learning_rate": 1.8655870252739426e-05, + "loss": 2.3548, + "step": 37680 + }, + { + "epoch": 1.61, + "grad_norm": 0.5078125, + "learning_rate": 1.8651968787367746e-05, + "loss": 2.3823, + "step": 37681 + }, + { + "epoch": 1.61, + "grad_norm": 0.490234375, + "learning_rate": 1.8648067688032468e-05, + "loss": 2.3589, + "step": 37682 + }, + { + "epoch": 1.61, + "grad_norm": 0.447265625, + "learning_rate": 1.8644166954751175e-05, + "loss": 2.2767, + "step": 37683 + }, + { + "epoch": 1.61, + "grad_norm": 0.439453125, + "learning_rate": 1.8640266587541367e-05, + "loss": 2.3762, + "step": 37684 + }, + { + "epoch": 1.61, + "grad_norm": 0.57421875, + "learning_rate": 1.8636366586420663e-05, + "loss": 2.497, + "step": 37685 + }, + { + "epoch": 1.61, + "grad_norm": 0.498046875, + "learning_rate": 1.8632466951406534e-05, + "loss": 2.4969, + "step": 37686 + }, + { + "epoch": 1.61, + "grad_norm": 0.55859375, + "learning_rate": 1.8628567682516607e-05, + "loss": 2.5009, + "step": 37687 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.862466877976835e-05, + "loss": 2.4243, + "step": 37688 + }, + { + "epoch": 1.61, + "grad_norm": 0.52734375, + "learning_rate": 1.8620770243179352e-05, + "loss": 2.3319, + "step": 37689 + }, + { + "epoch": 1.61, + "grad_norm": 0.416015625, + "learning_rate": 1.861687207276718e-05, + "loss": 2.452, + "step": 37690 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.861297426854932e-05, + "loss": 2.4508, + "step": 37691 + }, + { + "epoch": 1.61, + "grad_norm": 0.451171875, + "learning_rate": 1.8609076830543305e-05, + "loss": 2.4564, + "step": 37692 + }, + { + "epoch": 1.61, + "grad_norm": 0.458984375, + "learning_rate": 1.8605179758766732e-05, + "loss": 2.5278, + "step": 37693 + }, + { + "epoch": 1.61, + "grad_norm": 0.458984375, + "learning_rate": 1.8601283053237116e-05, + "loss": 2.3604, + "step": 37694 + }, + { + "epoch": 1.61, + "grad_norm": 0.482421875, + "learning_rate": 1.8597386713971943e-05, + "loss": 2.453, + "step": 37695 + }, + { + "epoch": 1.61, + "grad_norm": 0.50390625, + "learning_rate": 1.8593490740988783e-05, + "loss": 2.4919, + "step": 37696 + }, + { + "epoch": 1.61, + "grad_norm": 0.52734375, + "learning_rate": 1.8589595134305194e-05, + "loss": 2.605, + "step": 37697 + }, + { + "epoch": 1.61, + "grad_norm": 0.498046875, + "learning_rate": 1.8585699893938645e-05, + "loss": 2.3701, + "step": 37698 + }, + { + "epoch": 1.61, + "grad_norm": 0.52734375, + "learning_rate": 1.8581805019906717e-05, + "loss": 2.4907, + "step": 37699 + }, + { + "epoch": 1.61, + "grad_norm": 0.478515625, + "learning_rate": 1.8577910512226893e-05, + "loss": 2.4672, + "step": 37700 + }, + { + "epoch": 1.61, + "grad_norm": 0.451171875, + "learning_rate": 1.8574016370916736e-05, + "loss": 2.3686, + "step": 37701 + }, + { + "epoch": 1.61, + "grad_norm": 0.486328125, + "learning_rate": 1.857012259599372e-05, + "loss": 2.444, + "step": 37702 + }, + { + "epoch": 1.61, + "grad_norm": 0.5390625, + "learning_rate": 1.8566229187475426e-05, + "loss": 2.5796, + "step": 37703 + }, + { + "epoch": 1.61, + "grad_norm": 0.51953125, + "learning_rate": 1.8562336145379313e-05, + "loss": 2.7714, + "step": 37704 + }, + { + "epoch": 1.61, + "grad_norm": 0.421875, + "learning_rate": 1.855844346972292e-05, + "loss": 2.4149, + "step": 37705 + }, + { + "epoch": 1.61, + "grad_norm": 0.431640625, + "learning_rate": 1.855455116052379e-05, + "loss": 2.4774, + "step": 37706 + }, + { + "epoch": 1.61, + "grad_norm": 0.51953125, + "learning_rate": 1.8550659217799428e-05, + "loss": 2.681, + "step": 37707 + }, + { + "epoch": 1.61, + "grad_norm": 0.498046875, + "learning_rate": 1.8546767641567274e-05, + "loss": 2.4615, + "step": 37708 + }, + { + "epoch": 1.61, + "grad_norm": 0.51171875, + "learning_rate": 1.8542876431844948e-05, + "loss": 2.4661, + "step": 37709 + }, + { + "epoch": 1.61, + "grad_norm": 0.4140625, + "learning_rate": 1.8538985588649916e-05, + "loss": 2.422, + "step": 37710 + }, + { + "epoch": 1.61, + "grad_norm": 0.40234375, + "learning_rate": 1.8535095111999644e-05, + "loss": 2.2664, + "step": 37711 + }, + { + "epoch": 1.61, + "grad_norm": 0.5078125, + "learning_rate": 1.853120500191168e-05, + "loss": 2.4503, + "step": 37712 + }, + { + "epoch": 1.61, + "grad_norm": 0.4609375, + "learning_rate": 1.8527315258403534e-05, + "loss": 2.5325, + "step": 37713 + }, + { + "epoch": 1.61, + "grad_norm": 0.46484375, + "learning_rate": 1.8523425881492685e-05, + "loss": 2.4961, + "step": 37714 + }, + { + "epoch": 1.61, + "grad_norm": 0.484375, + "learning_rate": 1.8519536871196652e-05, + "loss": 2.3826, + "step": 37715 + }, + { + "epoch": 1.61, + "grad_norm": 0.453125, + "learning_rate": 1.8515648227532932e-05, + "loss": 2.3713, + "step": 37716 + }, + { + "epoch": 1.61, + "grad_norm": 0.5390625, + "learning_rate": 1.8511759950518993e-05, + "loss": 2.5419, + "step": 37717 + }, + { + "epoch": 1.61, + "grad_norm": 0.419921875, + "learning_rate": 1.850787204017236e-05, + "loss": 2.312, + "step": 37718 + }, + { + "epoch": 1.61, + "grad_norm": 0.474609375, + "learning_rate": 1.8503984496510528e-05, + "loss": 2.3959, + "step": 37719 + }, + { + "epoch": 1.61, + "grad_norm": 0.447265625, + "learning_rate": 1.850009731955097e-05, + "loss": 2.5404, + "step": 37720 + }, + { + "epoch": 1.61, + "grad_norm": 0.50390625, + "learning_rate": 1.8496210509311185e-05, + "loss": 2.5193, + "step": 37721 + }, + { + "epoch": 1.61, + "grad_norm": 0.486328125, + "learning_rate": 1.8492324065808698e-05, + "loss": 2.4826, + "step": 37722 + }, + { + "epoch": 1.61, + "grad_norm": 0.490234375, + "learning_rate": 1.8488437989060957e-05, + "loss": 2.3966, + "step": 37723 + }, + { + "epoch": 1.61, + "grad_norm": 0.435546875, + "learning_rate": 1.848455227908543e-05, + "loss": 2.5434, + "step": 37724 + }, + { + "epoch": 1.61, + "grad_norm": 0.49609375, + "learning_rate": 1.8480666935899637e-05, + "loss": 2.5259, + "step": 37725 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.847678195952107e-05, + "loss": 2.4138, + "step": 37726 + }, + { + "epoch": 1.61, + "grad_norm": 0.478515625, + "learning_rate": 1.8472897349967165e-05, + "loss": 2.5378, + "step": 37727 + }, + { + "epoch": 1.61, + "grad_norm": 0.490234375, + "learning_rate": 1.8469013107255463e-05, + "loss": 2.434, + "step": 37728 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.8465129231403367e-05, + "loss": 2.4342, + "step": 37729 + }, + { + "epoch": 1.61, + "grad_norm": 0.478515625, + "learning_rate": 1.84612457224284e-05, + "loss": 2.4314, + "step": 37730 + }, + { + "epoch": 1.61, + "grad_norm": 0.48046875, + "learning_rate": 1.845736258034805e-05, + "loss": 2.5014, + "step": 37731 + }, + { + "epoch": 1.61, + "grad_norm": 0.46484375, + "learning_rate": 1.8453479805179763e-05, + "loss": 2.3279, + "step": 37732 + }, + { + "epoch": 1.61, + "grad_norm": 0.447265625, + "learning_rate": 1.8449597396941e-05, + "loss": 2.4474, + "step": 37733 + }, + { + "epoch": 1.61, + "grad_norm": 0.41796875, + "learning_rate": 1.844571535564923e-05, + "loss": 2.4626, + "step": 37734 + }, + { + "epoch": 1.61, + "grad_norm": 0.51953125, + "learning_rate": 1.8441833681321973e-05, + "loss": 2.5139, + "step": 37735 + }, + { + "epoch": 1.61, + "grad_norm": 0.42578125, + "learning_rate": 1.843795237397663e-05, + "loss": 2.3294, + "step": 37736 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.8434071433630695e-05, + "loss": 2.4851, + "step": 37737 + }, + { + "epoch": 1.61, + "grad_norm": 0.419921875, + "learning_rate": 1.8430190860301655e-05, + "loss": 2.2988, + "step": 37738 + }, + { + "epoch": 1.61, + "grad_norm": 0.4921875, + "learning_rate": 1.842631065400695e-05, + "loss": 2.5827, + "step": 37739 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.8422430814764e-05, + "loss": 2.5049, + "step": 37740 + }, + { + "epoch": 1.61, + "grad_norm": 0.474609375, + "learning_rate": 1.8418551342590308e-05, + "loss": 2.4661, + "step": 37741 + }, + { + "epoch": 1.61, + "grad_norm": 0.4921875, + "learning_rate": 1.8414672237503338e-05, + "loss": 2.4418, + "step": 37742 + }, + { + "epoch": 1.61, + "grad_norm": 0.4765625, + "learning_rate": 1.8410793499520507e-05, + "loss": 2.4367, + "step": 37743 + }, + { + "epoch": 1.61, + "grad_norm": 0.515625, + "learning_rate": 1.840691512865932e-05, + "loss": 2.4354, + "step": 37744 + }, + { + "epoch": 1.61, + "grad_norm": 0.443359375, + "learning_rate": 1.840303712493716e-05, + "loss": 2.3838, + "step": 37745 + }, + { + "epoch": 1.61, + "grad_norm": 0.453125, + "learning_rate": 1.839915948837151e-05, + "loss": 2.3951, + "step": 37746 + }, + { + "epoch": 1.61, + "grad_norm": 0.462890625, + "learning_rate": 1.8395282218979847e-05, + "loss": 2.4348, + "step": 37747 + }, + { + "epoch": 1.61, + "grad_norm": 0.48046875, + "learning_rate": 1.8391405316779597e-05, + "loss": 2.3975, + "step": 37748 + }, + { + "epoch": 1.61, + "grad_norm": 0.39453125, + "learning_rate": 1.8387528781788178e-05, + "loss": 2.5597, + "step": 37749 + }, + { + "epoch": 1.61, + "grad_norm": 0.546875, + "learning_rate": 1.838365261402304e-05, + "loss": 2.449, + "step": 37750 + }, + { + "epoch": 1.61, + "grad_norm": 0.42578125, + "learning_rate": 1.837977681350167e-05, + "loss": 2.309, + "step": 37751 + }, + { + "epoch": 1.61, + "grad_norm": 0.4765625, + "learning_rate": 1.8375901380241456e-05, + "loss": 2.3493, + "step": 37752 + }, + { + "epoch": 1.61, + "grad_norm": 0.421875, + "learning_rate": 1.837202631425985e-05, + "loss": 2.453, + "step": 37753 + }, + { + "epoch": 1.61, + "grad_norm": 0.482421875, + "learning_rate": 1.8368151615574314e-05, + "loss": 2.5603, + "step": 37754 + }, + { + "epoch": 1.61, + "grad_norm": 0.48828125, + "learning_rate": 1.8364277284202268e-05, + "loss": 2.4606, + "step": 37755 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.8360403320161113e-05, + "loss": 2.4353, + "step": 37756 + }, + { + "epoch": 1.61, + "grad_norm": 0.45703125, + "learning_rate": 1.8356529723468308e-05, + "loss": 2.4376, + "step": 37757 + }, + { + "epoch": 1.61, + "grad_norm": 0.5078125, + "learning_rate": 1.8352656494141306e-05, + "loss": 2.394, + "step": 37758 + }, + { + "epoch": 1.61, + "grad_norm": 0.412109375, + "learning_rate": 1.8348783632197487e-05, + "loss": 2.2898, + "step": 37759 + }, + { + "epoch": 1.61, + "grad_norm": 0.451171875, + "learning_rate": 1.834491113765433e-05, + "loss": 2.3906, + "step": 37760 + }, + { + "epoch": 1.61, + "grad_norm": 0.51171875, + "learning_rate": 1.8341039010529205e-05, + "loss": 2.427, + "step": 37761 + }, + { + "epoch": 1.61, + "grad_norm": 0.447265625, + "learning_rate": 1.833716725083956e-05, + "loss": 2.3673, + "step": 37762 + }, + { + "epoch": 1.61, + "grad_norm": 0.51953125, + "learning_rate": 1.833329585860284e-05, + "loss": 2.4224, + "step": 37763 + }, + { + "epoch": 1.61, + "grad_norm": 0.57421875, + "learning_rate": 1.8329424833836438e-05, + "loss": 2.3977, + "step": 37764 + }, + { + "epoch": 1.61, + "grad_norm": 0.5078125, + "learning_rate": 1.8325554176557767e-05, + "loss": 2.3931, + "step": 37765 + }, + { + "epoch": 1.61, + "grad_norm": 0.466796875, + "learning_rate": 1.8321683886784245e-05, + "loss": 2.4427, + "step": 37766 + }, + { + "epoch": 1.61, + "grad_norm": 0.42578125, + "learning_rate": 1.8317813964533325e-05, + "loss": 2.4726, + "step": 37767 + }, + { + "epoch": 1.61, + "grad_norm": 0.455078125, + "learning_rate": 1.8313944409822358e-05, + "loss": 2.5435, + "step": 37768 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.8310075222668788e-05, + "loss": 2.3482, + "step": 37769 + }, + { + "epoch": 1.61, + "grad_norm": 0.46484375, + "learning_rate": 1.8306206403090065e-05, + "loss": 2.5128, + "step": 37770 + }, + { + "epoch": 1.61, + "grad_norm": 0.4296875, + "learning_rate": 1.8302337951103542e-05, + "loss": 2.3965, + "step": 37771 + }, + { + "epoch": 1.61, + "grad_norm": 0.447265625, + "learning_rate": 1.829846986672662e-05, + "loss": 2.4874, + "step": 37772 + }, + { + "epoch": 1.61, + "grad_norm": 0.46484375, + "learning_rate": 1.8294602149976726e-05, + "loss": 2.4678, + "step": 37773 + }, + { + "epoch": 1.61, + "grad_norm": 0.5234375, + "learning_rate": 1.829073480087129e-05, + "loss": 2.4872, + "step": 37774 + }, + { + "epoch": 1.61, + "grad_norm": 0.451171875, + "learning_rate": 1.8286867819427655e-05, + "loss": 2.5348, + "step": 37775 + }, + { + "epoch": 1.61, + "grad_norm": 0.453125, + "learning_rate": 1.8283001205663286e-05, + "loss": 2.5128, + "step": 37776 + }, + { + "epoch": 1.61, + "grad_norm": 0.470703125, + "learning_rate": 1.8279134959595524e-05, + "loss": 2.4977, + "step": 37777 + }, + { + "epoch": 1.61, + "grad_norm": 0.41796875, + "learning_rate": 1.8275269081241776e-05, + "loss": 2.3971, + "step": 37778 + }, + { + "epoch": 1.61, + "grad_norm": 0.443359375, + "learning_rate": 1.8271403570619484e-05, + "loss": 2.493, + "step": 37779 + }, + { + "epoch": 1.61, + "grad_norm": 0.388671875, + "learning_rate": 1.8267538427746012e-05, + "loss": 2.2909, + "step": 37780 + }, + { + "epoch": 1.61, + "grad_norm": 0.482421875, + "learning_rate": 1.8263673652638712e-05, + "loss": 2.5416, + "step": 37781 + }, + { + "epoch": 1.61, + "grad_norm": 0.46875, + "learning_rate": 1.8259809245315008e-05, + "loss": 2.464, + "step": 37782 + }, + { + "epoch": 1.61, + "grad_norm": 0.486328125, + "learning_rate": 1.8255945205792325e-05, + "loss": 2.4246, + "step": 37783 + }, + { + "epoch": 1.61, + "grad_norm": 0.4765625, + "learning_rate": 1.8252081534087984e-05, + "loss": 2.4573, + "step": 37784 + }, + { + "epoch": 1.61, + "grad_norm": 0.4765625, + "learning_rate": 1.8248218230219406e-05, + "loss": 2.4349, + "step": 37785 + }, + { + "epoch": 1.61, + "grad_norm": 0.42578125, + "learning_rate": 1.8244355294203987e-05, + "loss": 2.4666, + "step": 37786 + }, + { + "epoch": 1.61, + "grad_norm": 0.466796875, + "learning_rate": 1.824049272605909e-05, + "loss": 2.4381, + "step": 37787 + }, + { + "epoch": 1.61, + "grad_norm": 0.4921875, + "learning_rate": 1.8236630525802078e-05, + "loss": 2.388, + "step": 37788 + }, + { + "epoch": 1.61, + "grad_norm": 0.4609375, + "learning_rate": 1.823276869345034e-05, + "loss": 2.4691, + "step": 37789 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.8228907229021287e-05, + "loss": 2.509, + "step": 37790 + }, + { + "epoch": 1.61, + "grad_norm": 0.5078125, + "learning_rate": 1.822504613253224e-05, + "loss": 2.5145, + "step": 37791 + }, + { + "epoch": 1.61, + "grad_norm": 0.46875, + "learning_rate": 1.822118540400064e-05, + "loss": 2.5457, + "step": 37792 + }, + { + "epoch": 1.61, + "grad_norm": 0.443359375, + "learning_rate": 1.821732504344378e-05, + "loss": 2.3889, + "step": 37793 + }, + { + "epoch": 1.61, + "grad_norm": 0.5, + "learning_rate": 1.8213465050879076e-05, + "loss": 2.4183, + "step": 37794 + }, + { + "epoch": 1.61, + "grad_norm": 0.5078125, + "learning_rate": 1.8209605426323906e-05, + "loss": 2.4403, + "step": 37795 + }, + { + "epoch": 1.61, + "grad_norm": 0.439453125, + "learning_rate": 1.8205746169795635e-05, + "loss": 2.3272, + "step": 37796 + }, + { + "epoch": 1.61, + "grad_norm": 0.439453125, + "learning_rate": 1.8201887281311548e-05, + "loss": 2.3063, + "step": 37797 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.8198028760889128e-05, + "loss": 2.6197, + "step": 37798 + }, + { + "epoch": 1.61, + "grad_norm": 0.52734375, + "learning_rate": 1.8194170608545692e-05, + "loss": 2.4469, + "step": 37799 + }, + { + "epoch": 1.61, + "grad_norm": 0.42578125, + "learning_rate": 1.819031282429856e-05, + "loss": 2.4618, + "step": 37800 + }, + { + "epoch": 1.61, + "grad_norm": 0.478515625, + "learning_rate": 1.8186455408165126e-05, + "loss": 2.3693, + "step": 37801 + }, + { + "epoch": 1.61, + "grad_norm": 0.48046875, + "learning_rate": 1.8182598360162762e-05, + "loss": 2.3872, + "step": 37802 + }, + { + "epoch": 1.61, + "grad_norm": 0.46875, + "learning_rate": 1.8178741680308808e-05, + "loss": 2.5015, + "step": 37803 + }, + { + "epoch": 1.61, + "grad_norm": 0.478515625, + "learning_rate": 1.8174885368620597e-05, + "loss": 2.3272, + "step": 37804 + }, + { + "epoch": 1.61, + "grad_norm": 0.458984375, + "learning_rate": 1.817102942511548e-05, + "loss": 2.4718, + "step": 37805 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.816717384981087e-05, + "loss": 2.4479, + "step": 37806 + }, + { + "epoch": 1.61, + "grad_norm": 0.46875, + "learning_rate": 1.8163318642724035e-05, + "loss": 2.3778, + "step": 37807 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.815946380387239e-05, + "loss": 2.4484, + "step": 37808 + }, + { + "epoch": 1.61, + "grad_norm": 0.48828125, + "learning_rate": 1.8155609333273215e-05, + "loss": 2.4117, + "step": 37809 + }, + { + "epoch": 1.61, + "grad_norm": 0.474609375, + "learning_rate": 1.815175523094389e-05, + "loss": 2.5199, + "step": 37810 + }, + { + "epoch": 1.61, + "grad_norm": 0.42578125, + "learning_rate": 1.8147901496901788e-05, + "loss": 2.3842, + "step": 37811 + }, + { + "epoch": 1.61, + "grad_norm": 0.498046875, + "learning_rate": 1.8144048131164203e-05, + "loss": 2.482, + "step": 37812 + }, + { + "epoch": 1.61, + "grad_norm": 0.419921875, + "learning_rate": 1.814019513374845e-05, + "loss": 2.3819, + "step": 37813 + }, + { + "epoch": 1.61, + "grad_norm": 0.435546875, + "learning_rate": 1.8136342504671943e-05, + "loss": 2.3091, + "step": 37814 + }, + { + "epoch": 1.61, + "grad_norm": 0.466796875, + "learning_rate": 1.8132490243951973e-05, + "loss": 2.3052, + "step": 37815 + }, + { + "epoch": 1.61, + "grad_norm": 0.486328125, + "learning_rate": 1.8128638351605865e-05, + "loss": 2.4982, + "step": 37816 + }, + { + "epoch": 1.61, + "grad_norm": 0.4453125, + "learning_rate": 1.812478682765095e-05, + "loss": 2.4274, + "step": 37817 + }, + { + "epoch": 1.61, + "grad_norm": 0.515625, + "learning_rate": 1.812093567210461e-05, + "loss": 2.4083, + "step": 37818 + }, + { + "epoch": 1.61, + "grad_norm": 0.48046875, + "learning_rate": 1.8117084884984126e-05, + "loss": 2.4436, + "step": 37819 + }, + { + "epoch": 1.61, + "grad_norm": 0.52734375, + "learning_rate": 1.8113234466306817e-05, + "loss": 2.4315, + "step": 37820 + }, + { + "epoch": 1.61, + "grad_norm": 0.4375, + "learning_rate": 1.8109384416090024e-05, + "loss": 2.4591, + "step": 37821 + }, + { + "epoch": 1.61, + "grad_norm": 0.466796875, + "learning_rate": 1.8105534734351105e-05, + "loss": 2.4897, + "step": 37822 + }, + { + "epoch": 1.61, + "grad_norm": 0.439453125, + "learning_rate": 1.8101685421107318e-05, + "loss": 2.473, + "step": 37823 + }, + { + "epoch": 1.61, + "grad_norm": 0.466796875, + "learning_rate": 1.809783647637604e-05, + "loss": 2.379, + "step": 37824 + }, + { + "epoch": 1.61, + "grad_norm": 0.484375, + "learning_rate": 1.809398790017455e-05, + "loss": 2.3634, + "step": 37825 + }, + { + "epoch": 1.61, + "grad_norm": 0.4375, + "learning_rate": 1.809013969252017e-05, + "loss": 2.4935, + "step": 37826 + }, + { + "epoch": 1.61, + "grad_norm": 0.484375, + "learning_rate": 1.808629185343026e-05, + "loss": 2.5034, + "step": 37827 + }, + { + "epoch": 1.61, + "grad_norm": 0.51953125, + "learning_rate": 1.8082444382922103e-05, + "loss": 2.3576, + "step": 37828 + }, + { + "epoch": 1.61, + "grad_norm": 0.53515625, + "learning_rate": 1.8078597281012943e-05, + "loss": 2.5344, + "step": 37829 + }, + { + "epoch": 1.61, + "grad_norm": 0.453125, + "learning_rate": 1.807475054772022e-05, + "loss": 2.4249, + "step": 37830 + }, + { + "epoch": 1.61, + "grad_norm": 0.4609375, + "learning_rate": 1.807090418306118e-05, + "loss": 2.4804, + "step": 37831 + }, + { + "epoch": 1.61, + "grad_norm": 0.470703125, + "learning_rate": 1.8067058187053098e-05, + "loss": 2.5416, + "step": 37832 + }, + { + "epoch": 1.61, + "grad_norm": 0.5, + "learning_rate": 1.8063212559713317e-05, + "loss": 2.4389, + "step": 37833 + }, + { + "epoch": 1.61, + "grad_norm": 0.48046875, + "learning_rate": 1.805936730105916e-05, + "loss": 2.4901, + "step": 37834 + }, + { + "epoch": 1.61, + "grad_norm": 0.451171875, + "learning_rate": 1.8055522411107895e-05, + "loss": 2.4316, + "step": 37835 + }, + { + "epoch": 1.61, + "grad_norm": 0.439453125, + "learning_rate": 1.8051677889876818e-05, + "loss": 2.4669, + "step": 37836 + }, + { + "epoch": 1.61, + "grad_norm": 0.3984375, + "learning_rate": 1.804783373738326e-05, + "loss": 2.3378, + "step": 37837 + }, + { + "epoch": 1.61, + "grad_norm": 0.458984375, + "learning_rate": 1.804398995364448e-05, + "loss": 2.5347, + "step": 37838 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.8040146538677792e-05, + "loss": 2.4619, + "step": 37839 + }, + { + "epoch": 1.61, + "grad_norm": 0.41796875, + "learning_rate": 1.8036303492500518e-05, + "loss": 2.4093, + "step": 37840 + }, + { + "epoch": 1.61, + "grad_norm": 0.474609375, + "learning_rate": 1.8032460815129913e-05, + "loss": 2.44, + "step": 37841 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.8028618506583238e-05, + "loss": 2.5447, + "step": 37842 + }, + { + "epoch": 1.61, + "grad_norm": 0.51171875, + "learning_rate": 1.8024776566877876e-05, + "loss": 2.4716, + "step": 37843 + }, + { + "epoch": 1.61, + "grad_norm": 0.439453125, + "learning_rate": 1.802093499603106e-05, + "loss": 2.395, + "step": 37844 + }, + { + "epoch": 1.61, + "grad_norm": 0.462890625, + "learning_rate": 1.8017093794060047e-05, + "loss": 2.5393, + "step": 37845 + }, + { + "epoch": 1.61, + "grad_norm": 0.48828125, + "learning_rate": 1.8013252960982153e-05, + "loss": 2.4402, + "step": 37846 + }, + { + "epoch": 1.61, + "grad_norm": 0.515625, + "learning_rate": 1.800941249681467e-05, + "loss": 2.4702, + "step": 37847 + }, + { + "epoch": 1.61, + "grad_norm": 0.4296875, + "learning_rate": 1.8005572401574853e-05, + "loss": 2.3956, + "step": 37848 + }, + { + "epoch": 1.61, + "grad_norm": 0.443359375, + "learning_rate": 1.8001732675280024e-05, + "loss": 2.4538, + "step": 37849 + }, + { + "epoch": 1.61, + "grad_norm": 0.55078125, + "learning_rate": 1.79978933179474e-05, + "loss": 2.5777, + "step": 37850 + }, + { + "epoch": 1.61, + "grad_norm": 0.494140625, + "learning_rate": 1.799405432959429e-05, + "loss": 2.5493, + "step": 37851 + }, + { + "epoch": 1.61, + "grad_norm": 0.43359375, + "learning_rate": 1.799021571023799e-05, + "loss": 2.4222, + "step": 37852 + }, + { + "epoch": 1.61, + "grad_norm": 0.51171875, + "learning_rate": 1.7986377459895754e-05, + "loss": 2.4495, + "step": 37853 + }, + { + "epoch": 1.61, + "grad_norm": 0.462890625, + "learning_rate": 1.798253957858481e-05, + "loss": 2.4461, + "step": 37854 + }, + { + "epoch": 1.61, + "grad_norm": 0.4375, + "learning_rate": 1.7978702066322473e-05, + "loss": 2.4645, + "step": 37855 + }, + { + "epoch": 1.61, + "grad_norm": 0.53125, + "learning_rate": 1.797486492312602e-05, + "loss": 2.5154, + "step": 37856 + }, + { + "epoch": 1.61, + "grad_norm": 0.55078125, + "learning_rate": 1.7971028149012714e-05, + "loss": 2.5074, + "step": 37857 + }, + { + "epoch": 1.61, + "grad_norm": 0.451171875, + "learning_rate": 1.7967191743999733e-05, + "loss": 2.508, + "step": 37858 + }, + { + "epoch": 1.61, + "grad_norm": 0.455078125, + "learning_rate": 1.796335570810447e-05, + "loss": 2.5256, + "step": 37859 + }, + { + "epoch": 1.61, + "grad_norm": 0.4375, + "learning_rate": 1.7959520041344123e-05, + "loss": 2.5316, + "step": 37860 + }, + { + "epoch": 1.61, + "grad_norm": 0.46484375, + "learning_rate": 1.7955684743735935e-05, + "loss": 2.5873, + "step": 37861 + }, + { + "epoch": 1.61, + "grad_norm": 0.447265625, + "learning_rate": 1.795184981529717e-05, + "loss": 2.4782, + "step": 37862 + }, + { + "epoch": 1.61, + "grad_norm": 0.46875, + "learning_rate": 1.7948015256045127e-05, + "loss": 2.4357, + "step": 37863 + }, + { + "epoch": 1.61, + "grad_norm": 0.427734375, + "learning_rate": 1.7944181065997e-05, + "loss": 2.476, + "step": 37864 + }, + { + "epoch": 1.61, + "grad_norm": 0.431640625, + "learning_rate": 1.79403472451701e-05, + "loss": 2.5081, + "step": 37865 + }, + { + "epoch": 1.61, + "grad_norm": 0.466796875, + "learning_rate": 1.7936513793581623e-05, + "loss": 2.3613, + "step": 37866 + }, + { + "epoch": 1.61, + "grad_norm": 0.54296875, + "learning_rate": 1.793268071124884e-05, + "loss": 2.4471, + "step": 37867 + }, + { + "epoch": 1.61, + "grad_norm": 0.546875, + "learning_rate": 1.7928847998189025e-05, + "loss": 2.5197, + "step": 37868 + }, + { + "epoch": 1.61, + "grad_norm": 0.494140625, + "learning_rate": 1.792501565441941e-05, + "loss": 2.3676, + "step": 37869 + }, + { + "epoch": 1.61, + "grad_norm": 0.478515625, + "learning_rate": 1.79211836799572e-05, + "loss": 2.3518, + "step": 37870 + }, + { + "epoch": 1.61, + "grad_norm": 0.4921875, + "learning_rate": 1.791735207481966e-05, + "loss": 2.5169, + "step": 37871 + }, + { + "epoch": 1.61, + "grad_norm": 0.439453125, + "learning_rate": 1.7913520839024066e-05, + "loss": 2.3908, + "step": 37872 + }, + { + "epoch": 1.61, + "grad_norm": 0.39453125, + "learning_rate": 1.790968997258763e-05, + "loss": 2.3942, + "step": 37873 + }, + { + "epoch": 1.61, + "grad_norm": 0.41015625, + "learning_rate": 1.790585947552754e-05, + "loss": 2.4068, + "step": 37874 + }, + { + "epoch": 1.61, + "grad_norm": 0.498046875, + "learning_rate": 1.7902029347861126e-05, + "loss": 2.5059, + "step": 37875 + }, + { + "epoch": 1.61, + "grad_norm": 0.43359375, + "learning_rate": 1.7898199589605592e-05, + "loss": 2.5078, + "step": 37876 + }, + { + "epoch": 1.61, + "grad_norm": 0.37109375, + "learning_rate": 1.789437020077811e-05, + "loss": 2.3675, + "step": 37877 + }, + { + "epoch": 1.61, + "grad_norm": 0.4765625, + "learning_rate": 1.789054118139597e-05, + "loss": 2.4936, + "step": 37878 + }, + { + "epoch": 1.61, + "grad_norm": 0.396484375, + "learning_rate": 1.7886712531476402e-05, + "loss": 2.4198, + "step": 37879 + }, + { + "epoch": 1.61, + "grad_norm": 0.45703125, + "learning_rate": 1.7882884251036602e-05, + "loss": 2.4546, + "step": 37880 + }, + { + "epoch": 1.61, + "grad_norm": 0.5546875, + "learning_rate": 1.7879056340093834e-05, + "loss": 2.3194, + "step": 37881 + }, + { + "epoch": 1.61, + "grad_norm": 0.515625, + "learning_rate": 1.7875228798665277e-05, + "loss": 2.416, + "step": 37882 + }, + { + "epoch": 1.61, + "grad_norm": 0.482421875, + "learning_rate": 1.787140162676817e-05, + "loss": 2.5147, + "step": 37883 + }, + { + "epoch": 1.61, + "grad_norm": 0.51953125, + "learning_rate": 1.7867574824419773e-05, + "loss": 2.4869, + "step": 37884 + }, + { + "epoch": 1.61, + "grad_norm": 0.4609375, + "learning_rate": 1.7863748391637268e-05, + "loss": 2.4243, + "step": 37885 + }, + { + "epoch": 1.61, + "grad_norm": 0.486328125, + "learning_rate": 1.785992232843786e-05, + "loss": 2.4773, + "step": 37886 + }, + { + "epoch": 1.61, + "grad_norm": 0.53515625, + "learning_rate": 1.7856096634838782e-05, + "loss": 2.4702, + "step": 37887 + }, + { + "epoch": 1.61, + "grad_norm": 0.47265625, + "learning_rate": 1.7852271310857272e-05, + "loss": 2.405, + "step": 37888 + }, + { + "epoch": 1.61, + "grad_norm": 0.4453125, + "learning_rate": 1.784844635651052e-05, + "loss": 2.4112, + "step": 37889 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.7844621771815682e-05, + "loss": 2.3527, + "step": 37890 + }, + { + "epoch": 1.61, + "grad_norm": 0.4921875, + "learning_rate": 1.784079755679008e-05, + "loss": 2.3647, + "step": 37891 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.7836973711450855e-05, + "loss": 2.4861, + "step": 37892 + }, + { + "epoch": 1.61, + "grad_norm": 0.515625, + "learning_rate": 1.78331502358152e-05, + "loss": 2.4283, + "step": 37893 + }, + { + "epoch": 1.61, + "grad_norm": 0.44921875, + "learning_rate": 1.7829327129900342e-05, + "loss": 2.4392, + "step": 37894 + }, + { + "epoch": 1.61, + "grad_norm": 0.484375, + "learning_rate": 1.78255043937235e-05, + "loss": 2.4441, + "step": 37895 + }, + { + "epoch": 1.61, + "grad_norm": 0.474609375, + "learning_rate": 1.7821682027301844e-05, + "loss": 2.4475, + "step": 37896 + }, + { + "epoch": 1.61, + "grad_norm": 0.43359375, + "learning_rate": 1.78178600306526e-05, + "loss": 2.414, + "step": 37897 + }, + { + "epoch": 1.61, + "grad_norm": 0.45703125, + "learning_rate": 1.7814038403792932e-05, + "loss": 2.4577, + "step": 37898 + }, + { + "epoch": 1.61, + "grad_norm": 0.46875, + "learning_rate": 1.781021714674005e-05, + "loss": 2.391, + "step": 37899 + }, + { + "epoch": 1.62, + "grad_norm": 0.5, + "learning_rate": 1.780639625951118e-05, + "loss": 2.4127, + "step": 37900 + }, + { + "epoch": 1.62, + "grad_norm": 0.498046875, + "learning_rate": 1.7802575742123495e-05, + "loss": 2.5783, + "step": 37901 + }, + { + "epoch": 1.62, + "grad_norm": 0.4765625, + "learning_rate": 1.779875559459415e-05, + "loss": 2.5285, + "step": 37902 + }, + { + "epoch": 1.62, + "grad_norm": 0.486328125, + "learning_rate": 1.779493581694037e-05, + "loss": 2.4451, + "step": 37903 + }, + { + "epoch": 1.62, + "grad_norm": 0.48828125, + "learning_rate": 1.7791116409179354e-05, + "loss": 2.5708, + "step": 37904 + }, + { + "epoch": 1.62, + "grad_norm": 0.4609375, + "learning_rate": 1.778729737132825e-05, + "loss": 2.3306, + "step": 37905 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.7783478703404254e-05, + "loss": 2.5562, + "step": 37906 + }, + { + "epoch": 1.62, + "grad_norm": 0.494140625, + "learning_rate": 1.777966040542459e-05, + "loss": 2.6043, + "step": 37907 + }, + { + "epoch": 1.62, + "grad_norm": 0.53125, + "learning_rate": 1.7775842477406412e-05, + "loss": 2.6017, + "step": 37908 + }, + { + "epoch": 1.62, + "grad_norm": 0.47265625, + "learning_rate": 1.7772024919366858e-05, + "loss": 2.5039, + "step": 37909 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.776820773132314e-05, + "loss": 2.5665, + "step": 37910 + }, + { + "epoch": 1.62, + "grad_norm": 0.486328125, + "learning_rate": 1.776439091329246e-05, + "loss": 2.4586, + "step": 37911 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.776057446529196e-05, + "loss": 2.3137, + "step": 37912 + }, + { + "epoch": 1.62, + "grad_norm": 0.5234375, + "learning_rate": 1.7756758387338823e-05, + "loss": 2.4719, + "step": 37913 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.77529426794502e-05, + "loss": 2.5242, + "step": 37914 + }, + { + "epoch": 1.62, + "grad_norm": 0.486328125, + "learning_rate": 1.774912734164328e-05, + "loss": 2.5257, + "step": 37915 + }, + { + "epoch": 1.62, + "grad_norm": 0.453125, + "learning_rate": 1.774531237393524e-05, + "loss": 2.534, + "step": 37916 + }, + { + "epoch": 1.62, + "grad_norm": 0.490234375, + "learning_rate": 1.774149777634325e-05, + "loss": 2.4592, + "step": 37917 + }, + { + "epoch": 1.62, + "grad_norm": 0.451171875, + "learning_rate": 1.7737683548884422e-05, + "loss": 2.5303, + "step": 37918 + }, + { + "epoch": 1.62, + "grad_norm": 0.470703125, + "learning_rate": 1.7733869691575955e-05, + "loss": 2.4571, + "step": 37919 + }, + { + "epoch": 1.62, + "grad_norm": 0.49609375, + "learning_rate": 1.7730056204435042e-05, + "loss": 2.5394, + "step": 37920 + }, + { + "epoch": 1.62, + "grad_norm": 0.4609375, + "learning_rate": 1.772624308747878e-05, + "loss": 2.4584, + "step": 37921 + }, + { + "epoch": 1.62, + "grad_norm": 0.46484375, + "learning_rate": 1.772243034072436e-05, + "loss": 2.5365, + "step": 37922 + }, + { + "epoch": 1.62, + "grad_norm": 0.43359375, + "learning_rate": 1.7718617964188954e-05, + "loss": 2.4419, + "step": 37923 + }, + { + "epoch": 1.62, + "grad_norm": 0.51171875, + "learning_rate": 1.77148059578897e-05, + "loss": 2.435, + "step": 37924 + }, + { + "epoch": 1.62, + "grad_norm": 0.46484375, + "learning_rate": 1.771099432184372e-05, + "loss": 2.4559, + "step": 37925 + }, + { + "epoch": 1.62, + "grad_norm": 0.462890625, + "learning_rate": 1.77071830560682e-05, + "loss": 2.4343, + "step": 37926 + }, + { + "epoch": 1.62, + "grad_norm": 0.5, + "learning_rate": 1.77033721605803e-05, + "loss": 2.5276, + "step": 37927 + }, + { + "epoch": 1.62, + "grad_norm": 0.421875, + "learning_rate": 1.7699561635397123e-05, + "loss": 2.2921, + "step": 37928 + }, + { + "epoch": 1.62, + "grad_norm": 0.5078125, + "learning_rate": 1.7695751480535872e-05, + "loss": 2.4746, + "step": 37929 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.7691941696013625e-05, + "loss": 2.4656, + "step": 37930 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.768813228184757e-05, + "loss": 2.4212, + "step": 37931 + }, + { + "epoch": 1.62, + "grad_norm": 0.50390625, + "learning_rate": 1.7684323238054868e-05, + "loss": 2.4538, + "step": 37932 + }, + { + "epoch": 1.62, + "grad_norm": 0.46484375, + "learning_rate": 1.7680514564652616e-05, + "loss": 2.5686, + "step": 37933 + }, + { + "epoch": 1.62, + "grad_norm": 0.451171875, + "learning_rate": 1.767670626165794e-05, + "loss": 2.4467, + "step": 37934 + }, + { + "epoch": 1.62, + "grad_norm": 0.419921875, + "learning_rate": 1.7672898329088007e-05, + "loss": 2.4095, + "step": 37935 + }, + { + "epoch": 1.62, + "grad_norm": 0.5, + "learning_rate": 1.766909076695997e-05, + "loss": 2.5692, + "step": 37936 + }, + { + "epoch": 1.62, + "grad_norm": 0.50390625, + "learning_rate": 1.7665283575290902e-05, + "loss": 2.3944, + "step": 37937 + }, + { + "epoch": 1.62, + "grad_norm": 0.498046875, + "learning_rate": 1.766147675409797e-05, + "loss": 2.4717, + "step": 37938 + }, + { + "epoch": 1.62, + "grad_norm": 0.51171875, + "learning_rate": 1.765767030339833e-05, + "loss": 2.4198, + "step": 37939 + }, + { + "epoch": 1.62, + "grad_norm": 0.4296875, + "learning_rate": 1.765386422320908e-05, + "loss": 2.4665, + "step": 37940 + }, + { + "epoch": 1.62, + "grad_norm": 0.48046875, + "learning_rate": 1.765005851354733e-05, + "loss": 2.4558, + "step": 37941 + }, + { + "epoch": 1.62, + "grad_norm": 0.447265625, + "learning_rate": 1.7646253174430215e-05, + "loss": 2.3309, + "step": 37942 + }, + { + "epoch": 1.62, + "grad_norm": 0.5078125, + "learning_rate": 1.7642448205874885e-05, + "loss": 2.347, + "step": 37943 + }, + { + "epoch": 1.62, + "grad_norm": 0.466796875, + "learning_rate": 1.763864360789842e-05, + "loss": 2.4739, + "step": 37944 + }, + { + "epoch": 1.62, + "grad_norm": 0.4609375, + "learning_rate": 1.7634839380517987e-05, + "loss": 2.4976, + "step": 37945 + }, + { + "epoch": 1.62, + "grad_norm": 0.4296875, + "learning_rate": 1.763103552375064e-05, + "loss": 2.383, + "step": 37946 + }, + { + "epoch": 1.62, + "grad_norm": 0.4609375, + "learning_rate": 1.762723203761354e-05, + "loss": 2.3723, + "step": 37947 + }, + { + "epoch": 1.62, + "grad_norm": 0.482421875, + "learning_rate": 1.762342892212381e-05, + "loss": 2.4017, + "step": 37948 + }, + { + "epoch": 1.62, + "grad_norm": 0.5078125, + "learning_rate": 1.7619626177298552e-05, + "loss": 2.5081, + "step": 37949 + }, + { + "epoch": 1.62, + "grad_norm": 0.474609375, + "learning_rate": 1.761582380315484e-05, + "loss": 2.3902, + "step": 37950 + }, + { + "epoch": 1.62, + "grad_norm": 0.4296875, + "learning_rate": 1.761202179970981e-05, + "loss": 2.4401, + "step": 37951 + }, + { + "epoch": 1.62, + "grad_norm": 0.44140625, + "learning_rate": 1.7608220166980606e-05, + "loss": 2.4331, + "step": 37952 + }, + { + "epoch": 1.62, + "grad_norm": 0.484375, + "learning_rate": 1.7604418904984264e-05, + "loss": 2.2816, + "step": 37953 + }, + { + "epoch": 1.62, + "grad_norm": 0.43359375, + "learning_rate": 1.7600618013737934e-05, + "loss": 2.3097, + "step": 37954 + }, + { + "epoch": 1.62, + "grad_norm": 0.515625, + "learning_rate": 1.7596817493258734e-05, + "loss": 2.5276, + "step": 37955 + }, + { + "epoch": 1.62, + "grad_norm": 0.4375, + "learning_rate": 1.7593017343563735e-05, + "loss": 2.3731, + "step": 37956 + }, + { + "epoch": 1.62, + "grad_norm": 0.453125, + "learning_rate": 1.7589217564670026e-05, + "loss": 2.4577, + "step": 37957 + }, + { + "epoch": 1.62, + "grad_norm": 0.41015625, + "learning_rate": 1.7585418156594734e-05, + "loss": 2.3217, + "step": 37958 + }, + { + "epoch": 1.62, + "grad_norm": 0.474609375, + "learning_rate": 1.758161911935492e-05, + "loss": 2.436, + "step": 37959 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.757782045296771e-05, + "loss": 2.3695, + "step": 37960 + }, + { + "epoch": 1.62, + "grad_norm": 0.451171875, + "learning_rate": 1.75740221574502e-05, + "loss": 2.4871, + "step": 37961 + }, + { + "epoch": 1.62, + "grad_norm": 0.51953125, + "learning_rate": 1.7570224232819466e-05, + "loss": 2.4996, + "step": 37962 + }, + { + "epoch": 1.62, + "grad_norm": 0.46484375, + "learning_rate": 1.756642667909255e-05, + "loss": 2.548, + "step": 37963 + }, + { + "epoch": 1.62, + "grad_norm": 0.388671875, + "learning_rate": 1.7562629496286642e-05, + "loss": 2.3511, + "step": 37964 + }, + { + "epoch": 1.62, + "grad_norm": 0.466796875, + "learning_rate": 1.755883268441877e-05, + "loss": 2.419, + "step": 37965 + }, + { + "epoch": 1.62, + "grad_norm": 0.453125, + "learning_rate": 1.7555036243505996e-05, + "loss": 2.5752, + "step": 37966 + }, + { + "epoch": 1.62, + "grad_norm": 0.45703125, + "learning_rate": 1.755124017356543e-05, + "loss": 2.4274, + "step": 37967 + }, + { + "epoch": 1.62, + "grad_norm": 0.4765625, + "learning_rate": 1.7547444474614185e-05, + "loss": 2.4587, + "step": 37968 + }, + { + "epoch": 1.62, + "grad_norm": 0.466796875, + "learning_rate": 1.754364914666927e-05, + "loss": 2.2555, + "step": 37969 + }, + { + "epoch": 1.62, + "grad_norm": 0.443359375, + "learning_rate": 1.7539854189747827e-05, + "loss": 2.4694, + "step": 37970 + }, + { + "epoch": 1.62, + "grad_norm": 0.412109375, + "learning_rate": 1.7536059603866882e-05, + "loss": 2.4947, + "step": 37971 + }, + { + "epoch": 1.62, + "grad_norm": 0.462890625, + "learning_rate": 1.7532265389043557e-05, + "loss": 2.472, + "step": 37972 + }, + { + "epoch": 1.62, + "grad_norm": 0.48828125, + "learning_rate": 1.7528471545294865e-05, + "loss": 2.5791, + "step": 37973 + }, + { + "epoch": 1.62, + "grad_norm": 0.447265625, + "learning_rate": 1.7524678072637946e-05, + "loss": 2.4858, + "step": 37974 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.7520884971089813e-05, + "loss": 2.5953, + "step": 37975 + }, + { + "epoch": 1.62, + "grad_norm": 0.44140625, + "learning_rate": 1.7517092240667553e-05, + "loss": 2.3876, + "step": 37976 + }, + { + "epoch": 1.62, + "grad_norm": 0.455078125, + "learning_rate": 1.7513299881388256e-05, + "loss": 2.3746, + "step": 37977 + }, + { + "epoch": 1.62, + "grad_norm": 0.43359375, + "learning_rate": 1.750950789326896e-05, + "loss": 2.2858, + "step": 37978 + }, + { + "epoch": 1.62, + "grad_norm": 0.41015625, + "learning_rate": 1.750571627632669e-05, + "loss": 2.25, + "step": 37979 + }, + { + "epoch": 1.62, + "grad_norm": 0.51171875, + "learning_rate": 1.7501925030578592e-05, + "loss": 2.4934, + "step": 37980 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.7498134156041677e-05, + "loss": 2.4712, + "step": 37981 + }, + { + "epoch": 1.62, + "grad_norm": 0.478515625, + "learning_rate": 1.7494343652732992e-05, + "loss": 2.4604, + "step": 37982 + }, + { + "epoch": 1.62, + "grad_norm": 0.39453125, + "learning_rate": 1.7490553520669607e-05, + "loss": 2.4876, + "step": 37983 + }, + { + "epoch": 1.62, + "grad_norm": 0.416015625, + "learning_rate": 1.7486763759868597e-05, + "loss": 2.3155, + "step": 37984 + }, + { + "epoch": 1.62, + "grad_norm": 0.494140625, + "learning_rate": 1.7482974370346973e-05, + "loss": 2.4991, + "step": 37985 + }, + { + "epoch": 1.62, + "grad_norm": 0.451171875, + "learning_rate": 1.747918535212184e-05, + "loss": 2.4743, + "step": 37986 + }, + { + "epoch": 1.62, + "grad_norm": 0.3828125, + "learning_rate": 1.747539670521018e-05, + "loss": 2.3683, + "step": 37987 + }, + { + "epoch": 1.62, + "grad_norm": 0.53125, + "learning_rate": 1.7471608429629106e-05, + "loss": 2.4608, + "step": 37988 + }, + { + "epoch": 1.62, + "grad_norm": 0.49609375, + "learning_rate": 1.7467820525395605e-05, + "loss": 2.4489, + "step": 37989 + }, + { + "epoch": 1.62, + "grad_norm": 0.5078125, + "learning_rate": 1.746403299252678e-05, + "loss": 2.542, + "step": 37990 + }, + { + "epoch": 1.62, + "grad_norm": 0.51953125, + "learning_rate": 1.7460245831039612e-05, + "loss": 2.4626, + "step": 37991 + }, + { + "epoch": 1.62, + "grad_norm": 0.486328125, + "learning_rate": 1.745645904095117e-05, + "loss": 2.4458, + "step": 37992 + }, + { + "epoch": 1.62, + "grad_norm": 0.466796875, + "learning_rate": 1.745267262227853e-05, + "loss": 2.3919, + "step": 37993 + }, + { + "epoch": 1.62, + "grad_norm": 0.482421875, + "learning_rate": 1.744888657503869e-05, + "loss": 2.5389, + "step": 37994 + }, + { + "epoch": 1.62, + "grad_norm": 0.45703125, + "learning_rate": 1.7445100899248646e-05, + "loss": 2.5372, + "step": 37995 + }, + { + "epoch": 1.62, + "grad_norm": 0.546875, + "learning_rate": 1.7441315594925523e-05, + "loss": 2.5166, + "step": 37996 + }, + { + "epoch": 1.62, + "grad_norm": 0.478515625, + "learning_rate": 1.743753066208631e-05, + "loss": 2.4375, + "step": 37997 + }, + { + "epoch": 1.62, + "grad_norm": 0.47265625, + "learning_rate": 1.7433746100748004e-05, + "loss": 2.547, + "step": 37998 + }, + { + "epoch": 1.62, + "grad_norm": 0.447265625, + "learning_rate": 1.7429961910927672e-05, + "loss": 2.3591, + "step": 37999 + }, + { + "epoch": 1.62, + "grad_norm": 0.408203125, + "learning_rate": 1.7426178092642364e-05, + "loss": 2.3135, + "step": 38000 + }, + { + "epoch": 1.62, + "grad_norm": 0.498046875, + "learning_rate": 1.7422394645909044e-05, + "loss": 2.3729, + "step": 38001 + }, + { + "epoch": 1.62, + "grad_norm": 0.431640625, + "learning_rate": 1.7418611570744803e-05, + "loss": 2.587, + "step": 38002 + }, + { + "epoch": 1.62, + "grad_norm": 0.392578125, + "learning_rate": 1.741482886716661e-05, + "loss": 2.4042, + "step": 38003 + }, + { + "epoch": 1.62, + "grad_norm": 0.52734375, + "learning_rate": 1.7411046535191522e-05, + "loss": 2.4144, + "step": 38004 + }, + { + "epoch": 1.62, + "grad_norm": 0.50390625, + "learning_rate": 1.7407264574836514e-05, + "loss": 2.4484, + "step": 38005 + }, + { + "epoch": 1.62, + "grad_norm": 0.443359375, + "learning_rate": 1.740348298611866e-05, + "loss": 2.4179, + "step": 38006 + }, + { + "epoch": 1.62, + "grad_norm": 0.4765625, + "learning_rate": 1.7399701769054923e-05, + "loss": 2.4886, + "step": 38007 + }, + { + "epoch": 1.62, + "grad_norm": 0.427734375, + "learning_rate": 1.7395920923662333e-05, + "loss": 2.4337, + "step": 38008 + }, + { + "epoch": 1.62, + "grad_norm": 0.447265625, + "learning_rate": 1.739214044995795e-05, + "loss": 2.3781, + "step": 38009 + }, + { + "epoch": 1.62, + "grad_norm": 0.44140625, + "learning_rate": 1.738836034795873e-05, + "loss": 2.337, + "step": 38010 + }, + { + "epoch": 1.62, + "grad_norm": 0.48046875, + "learning_rate": 1.7384580617681655e-05, + "loss": 2.5722, + "step": 38011 + }, + { + "epoch": 1.62, + "grad_norm": 0.431640625, + "learning_rate": 1.7380801259143832e-05, + "loss": 2.4238, + "step": 38012 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.7377022272362197e-05, + "loss": 2.4529, + "step": 38013 + }, + { + "epoch": 1.62, + "grad_norm": 0.55078125, + "learning_rate": 1.7373243657353745e-05, + "loss": 2.5449, + "step": 38014 + }, + { + "epoch": 1.62, + "grad_norm": 0.50390625, + "learning_rate": 1.73694654141355e-05, + "loss": 2.5174, + "step": 38015 + }, + { + "epoch": 1.62, + "grad_norm": 0.466796875, + "learning_rate": 1.7365687542724485e-05, + "loss": 2.3943, + "step": 38016 + }, + { + "epoch": 1.62, + "grad_norm": 0.466796875, + "learning_rate": 1.7361910043137665e-05, + "loss": 2.2932, + "step": 38017 + }, + { + "epoch": 1.62, + "grad_norm": 0.45703125, + "learning_rate": 1.7358132915392057e-05, + "loss": 2.2415, + "step": 38018 + }, + { + "epoch": 1.62, + "grad_norm": 0.51171875, + "learning_rate": 1.7354356159504636e-05, + "loss": 2.5127, + "step": 38019 + }, + { + "epoch": 1.62, + "grad_norm": 0.458984375, + "learning_rate": 1.7350579775492425e-05, + "loss": 2.3827, + "step": 38020 + }, + { + "epoch": 1.62, + "grad_norm": 0.49609375, + "learning_rate": 1.7346803763372377e-05, + "loss": 2.4861, + "step": 38021 + }, + { + "epoch": 1.62, + "grad_norm": 0.451171875, + "learning_rate": 1.734302812316154e-05, + "loss": 2.4838, + "step": 38022 + }, + { + "epoch": 1.62, + "grad_norm": 0.51171875, + "learning_rate": 1.7339252854876832e-05, + "loss": 2.452, + "step": 38023 + }, + { + "epoch": 1.62, + "grad_norm": 0.51171875, + "learning_rate": 1.7335477958535284e-05, + "loss": 2.5593, + "step": 38024 + }, + { + "epoch": 1.62, + "grad_norm": 0.48046875, + "learning_rate": 1.73317034341539e-05, + "loss": 2.4943, + "step": 38025 + }, + { + "epoch": 1.62, + "grad_norm": 0.4375, + "learning_rate": 1.7327929281749633e-05, + "loss": 2.485, + "step": 38026 + }, + { + "epoch": 1.62, + "grad_norm": 0.546875, + "learning_rate": 1.7324155501339424e-05, + "loss": 2.476, + "step": 38027 + }, + { + "epoch": 1.62, + "grad_norm": 0.47265625, + "learning_rate": 1.7320382092940357e-05, + "loss": 2.4726, + "step": 38028 + }, + { + "epoch": 1.62, + "grad_norm": 0.470703125, + "learning_rate": 1.7316609056569344e-05, + "loss": 2.4048, + "step": 38029 + }, + { + "epoch": 1.62, + "grad_norm": 0.42578125, + "learning_rate": 1.731283639224335e-05, + "loss": 2.2834, + "step": 38030 + }, + { + "epoch": 1.62, + "grad_norm": 0.400390625, + "learning_rate": 1.7309064099979376e-05, + "loss": 2.367, + "step": 38031 + }, + { + "epoch": 1.62, + "grad_norm": 0.4609375, + "learning_rate": 1.730529217979442e-05, + "loss": 2.4316, + "step": 38032 + }, + { + "epoch": 1.62, + "grad_norm": 0.4921875, + "learning_rate": 1.7301520631705403e-05, + "loss": 2.4261, + "step": 38033 + }, + { + "epoch": 1.62, + "grad_norm": 0.4375, + "learning_rate": 1.729774945572934e-05, + "loss": 2.411, + "step": 38034 + }, + { + "epoch": 1.62, + "grad_norm": 0.515625, + "learning_rate": 1.7293978651883157e-05, + "loss": 2.4189, + "step": 38035 + }, + { + "epoch": 1.62, + "grad_norm": 0.54296875, + "learning_rate": 1.7290208220183878e-05, + "loss": 2.5453, + "step": 38036 + }, + { + "epoch": 1.62, + "grad_norm": 0.443359375, + "learning_rate": 1.72864381606484e-05, + "loss": 2.4378, + "step": 38037 + }, + { + "epoch": 1.62, + "grad_norm": 0.451171875, + "learning_rate": 1.728266847329375e-05, + "loss": 2.4755, + "step": 38038 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.7278899158136842e-05, + "loss": 2.5243, + "step": 38039 + }, + { + "epoch": 1.62, + "grad_norm": 0.474609375, + "learning_rate": 1.7275130215194645e-05, + "loss": 2.4791, + "step": 38040 + }, + { + "epoch": 1.62, + "grad_norm": 0.51171875, + "learning_rate": 1.7271361644484163e-05, + "loss": 2.6158, + "step": 38041 + }, + { + "epoch": 1.62, + "grad_norm": 0.462890625, + "learning_rate": 1.7267593446022323e-05, + "loss": 2.4692, + "step": 38042 + }, + { + "epoch": 1.62, + "grad_norm": 0.48046875, + "learning_rate": 1.7263825619826023e-05, + "loss": 2.6238, + "step": 38043 + }, + { + "epoch": 1.62, + "grad_norm": 0.43359375, + "learning_rate": 1.7260058165912317e-05, + "loss": 2.3746, + "step": 38044 + }, + { + "epoch": 1.62, + "grad_norm": 0.45703125, + "learning_rate": 1.7256291084298116e-05, + "loss": 2.4846, + "step": 38045 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.7252524375000344e-05, + "loss": 2.402, + "step": 38046 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.7248758038035962e-05, + "loss": 2.4274, + "step": 38047 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.724499207342196e-05, + "loss": 2.5439, + "step": 38048 + }, + { + "epoch": 1.62, + "grad_norm": 0.47265625, + "learning_rate": 1.724122648117522e-05, + "loss": 2.3876, + "step": 38049 + }, + { + "epoch": 1.62, + "grad_norm": 0.431640625, + "learning_rate": 1.723746126131275e-05, + "loss": 2.3816, + "step": 38050 + }, + { + "epoch": 1.62, + "grad_norm": 0.4453125, + "learning_rate": 1.7233696413851417e-05, + "loss": 2.4907, + "step": 38051 + }, + { + "epoch": 1.62, + "grad_norm": 0.455078125, + "learning_rate": 1.7229931938808243e-05, + "loss": 2.4795, + "step": 38052 + }, + { + "epoch": 1.62, + "grad_norm": 0.458984375, + "learning_rate": 1.722616783620009e-05, + "loss": 2.6374, + "step": 38053 + }, + { + "epoch": 1.62, + "grad_norm": 0.41796875, + "learning_rate": 1.7222404106043965e-05, + "loss": 2.528, + "step": 38054 + }, + { + "epoch": 1.62, + "grad_norm": 0.42578125, + "learning_rate": 1.721864074835674e-05, + "loss": 2.3501, + "step": 38055 + }, + { + "epoch": 1.62, + "grad_norm": 0.439453125, + "learning_rate": 1.7214877763155378e-05, + "loss": 2.3188, + "step": 38056 + }, + { + "epoch": 1.62, + "grad_norm": 0.51171875, + "learning_rate": 1.7211115150456836e-05, + "loss": 2.4762, + "step": 38057 + }, + { + "epoch": 1.62, + "grad_norm": 0.486328125, + "learning_rate": 1.720735291027803e-05, + "loss": 2.5261, + "step": 38058 + }, + { + "epoch": 1.62, + "grad_norm": 0.4765625, + "learning_rate": 1.7203591042635815e-05, + "loss": 2.4416, + "step": 38059 + }, + { + "epoch": 1.62, + "grad_norm": 0.46484375, + "learning_rate": 1.719982954754724e-05, + "loss": 2.4817, + "step": 38060 + }, + { + "epoch": 1.62, + "grad_norm": 0.435546875, + "learning_rate": 1.7196068425029165e-05, + "loss": 2.4124, + "step": 38061 + }, + { + "epoch": 1.62, + "grad_norm": 0.4765625, + "learning_rate": 1.7192307675098507e-05, + "loss": 2.5248, + "step": 38062 + }, + { + "epoch": 1.62, + "grad_norm": 0.5078125, + "learning_rate": 1.7188547297772195e-05, + "loss": 2.5442, + "step": 38063 + }, + { + "epoch": 1.62, + "grad_norm": 0.44140625, + "learning_rate": 1.7184787293067182e-05, + "loss": 2.4583, + "step": 38064 + }, + { + "epoch": 1.62, + "grad_norm": 0.5234375, + "learning_rate": 1.718102766100034e-05, + "loss": 2.4681, + "step": 38065 + }, + { + "epoch": 1.62, + "grad_norm": 0.45703125, + "learning_rate": 1.717726840158862e-05, + "loss": 2.5884, + "step": 38066 + }, + { + "epoch": 1.62, + "grad_norm": 0.435546875, + "learning_rate": 1.7173509514848908e-05, + "loss": 2.4202, + "step": 38067 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.716975100079815e-05, + "loss": 2.428, + "step": 38068 + }, + { + "epoch": 1.62, + "grad_norm": 0.515625, + "learning_rate": 1.7165992859453228e-05, + "loss": 2.5371, + "step": 38069 + }, + { + "epoch": 1.62, + "grad_norm": 0.44921875, + "learning_rate": 1.716223509083108e-05, + "loss": 2.5572, + "step": 38070 + }, + { + "epoch": 1.62, + "grad_norm": 0.44140625, + "learning_rate": 1.7158477694948584e-05, + "loss": 2.3549, + "step": 38071 + }, + { + "epoch": 1.62, + "grad_norm": 0.404296875, + "learning_rate": 1.7154720671822656e-05, + "loss": 2.356, + "step": 38072 + }, + { + "epoch": 1.62, + "grad_norm": 0.51953125, + "learning_rate": 1.715096402147024e-05, + "loss": 2.3997, + "step": 38073 + }, + { + "epoch": 1.62, + "grad_norm": 0.498046875, + "learning_rate": 1.71472077439082e-05, + "loss": 2.4939, + "step": 38074 + }, + { + "epoch": 1.62, + "grad_norm": 0.44140625, + "learning_rate": 1.71434518391534e-05, + "loss": 2.5041, + "step": 38075 + }, + { + "epoch": 1.62, + "grad_norm": 0.462890625, + "learning_rate": 1.713969630722284e-05, + "loss": 2.3316, + "step": 38076 + }, + { + "epoch": 1.62, + "grad_norm": 0.48828125, + "learning_rate": 1.7135941148133362e-05, + "loss": 2.4417, + "step": 38077 + }, + { + "epoch": 1.62, + "grad_norm": 0.43359375, + "learning_rate": 1.7132186361901847e-05, + "loss": 2.4313, + "step": 38078 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.712843194854522e-05, + "loss": 2.5592, + "step": 38079 + }, + { + "epoch": 1.62, + "grad_norm": 0.44140625, + "learning_rate": 1.7124677908080346e-05, + "loss": 2.5405, + "step": 38080 + }, + { + "epoch": 1.62, + "grad_norm": 0.453125, + "learning_rate": 1.712092424052414e-05, + "loss": 2.4801, + "step": 38081 + }, + { + "epoch": 1.62, + "grad_norm": 0.419921875, + "learning_rate": 1.71171709458935e-05, + "loss": 2.3811, + "step": 38082 + }, + { + "epoch": 1.62, + "grad_norm": 0.484375, + "learning_rate": 1.711341802420531e-05, + "loss": 2.4376, + "step": 38083 + }, + { + "epoch": 1.62, + "grad_norm": 0.515625, + "learning_rate": 1.710966547547641e-05, + "loss": 2.4044, + "step": 38084 + }, + { + "epoch": 1.62, + "grad_norm": 0.490234375, + "learning_rate": 1.710591329972373e-05, + "loss": 2.505, + "step": 38085 + }, + { + "epoch": 1.62, + "grad_norm": 0.5390625, + "learning_rate": 1.7102161496964175e-05, + "loss": 2.4925, + "step": 38086 + }, + { + "epoch": 1.62, + "grad_norm": 0.5078125, + "learning_rate": 1.7098410067214564e-05, + "loss": 2.4686, + "step": 38087 + }, + { + "epoch": 1.62, + "grad_norm": 0.4296875, + "learning_rate": 1.709465901049182e-05, + "loss": 2.4431, + "step": 38088 + }, + { + "epoch": 1.62, + "grad_norm": 0.400390625, + "learning_rate": 1.7090908326812827e-05, + "loss": 2.3361, + "step": 38089 + }, + { + "epoch": 1.62, + "grad_norm": 0.455078125, + "learning_rate": 1.708715801619445e-05, + "loss": 2.4115, + "step": 38090 + }, + { + "epoch": 1.62, + "grad_norm": 0.451171875, + "learning_rate": 1.708340807865354e-05, + "loss": 2.5067, + "step": 38091 + }, + { + "epoch": 1.62, + "grad_norm": 0.4921875, + "learning_rate": 1.7079658514206987e-05, + "loss": 2.6226, + "step": 38092 + }, + { + "epoch": 1.62, + "grad_norm": 0.458984375, + "learning_rate": 1.70759093228717e-05, + "loss": 2.5246, + "step": 38093 + }, + { + "epoch": 1.62, + "grad_norm": 0.5546875, + "learning_rate": 1.707216050466448e-05, + "loss": 2.5199, + "step": 38094 + }, + { + "epoch": 1.62, + "grad_norm": 0.42578125, + "learning_rate": 1.7068412059602256e-05, + "loss": 2.3324, + "step": 38095 + }, + { + "epoch": 1.62, + "grad_norm": 0.51953125, + "learning_rate": 1.7064663987701845e-05, + "loss": 2.5248, + "step": 38096 + }, + { + "epoch": 1.62, + "grad_norm": 0.47265625, + "learning_rate": 1.7060916288980145e-05, + "loss": 2.5401, + "step": 38097 + }, + { + "epoch": 1.62, + "grad_norm": 0.49609375, + "learning_rate": 1.705716896345403e-05, + "loss": 2.4907, + "step": 38098 + }, + { + "epoch": 1.62, + "grad_norm": 0.462890625, + "learning_rate": 1.7053422011140342e-05, + "loss": 2.3778, + "step": 38099 + }, + { + "epoch": 1.62, + "grad_norm": 0.4453125, + "learning_rate": 1.7049675432055923e-05, + "loss": 2.3611, + "step": 38100 + }, + { + "epoch": 1.62, + "grad_norm": 0.5390625, + "learning_rate": 1.704592922621764e-05, + "loss": 2.4794, + "step": 38101 + }, + { + "epoch": 1.62, + "grad_norm": 0.4375, + "learning_rate": 1.70421833936424e-05, + "loss": 2.4662, + "step": 38102 + }, + { + "epoch": 1.62, + "grad_norm": 0.455078125, + "learning_rate": 1.7038437934346974e-05, + "loss": 2.5088, + "step": 38103 + }, + { + "epoch": 1.62, + "grad_norm": 0.458984375, + "learning_rate": 1.7034692848348267e-05, + "loss": 2.5231, + "step": 38104 + }, + { + "epoch": 1.62, + "grad_norm": 0.455078125, + "learning_rate": 1.703094813566314e-05, + "loss": 2.5184, + "step": 38105 + }, + { + "epoch": 1.62, + "grad_norm": 0.455078125, + "learning_rate": 1.7027203796308434e-05, + "loss": 2.4682, + "step": 38106 + }, + { + "epoch": 1.62, + "grad_norm": 0.50390625, + "learning_rate": 1.702345983030096e-05, + "loss": 2.5545, + "step": 38107 + }, + { + "epoch": 1.62, + "grad_norm": 0.4140625, + "learning_rate": 1.701971623765759e-05, + "loss": 2.4046, + "step": 38108 + }, + { + "epoch": 1.62, + "grad_norm": 0.4296875, + "learning_rate": 1.701597301839519e-05, + "loss": 2.3606, + "step": 38109 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.7012230172530574e-05, + "loss": 2.2928, + "step": 38110 + }, + { + "epoch": 1.62, + "grad_norm": 0.4296875, + "learning_rate": 1.700848770008061e-05, + "loss": 2.4154, + "step": 38111 + }, + { + "epoch": 1.62, + "grad_norm": 0.404296875, + "learning_rate": 1.70047456010621e-05, + "loss": 2.4336, + "step": 38112 + }, + { + "epoch": 1.62, + "grad_norm": 0.427734375, + "learning_rate": 1.7001003875491896e-05, + "loss": 2.3852, + "step": 38113 + }, + { + "epoch": 1.62, + "grad_norm": 0.470703125, + "learning_rate": 1.699726252338687e-05, + "loss": 2.4644, + "step": 38114 + }, + { + "epoch": 1.62, + "grad_norm": 0.3828125, + "learning_rate": 1.6993521544763836e-05, + "loss": 2.5237, + "step": 38115 + }, + { + "epoch": 1.62, + "grad_norm": 0.458984375, + "learning_rate": 1.6989780939639588e-05, + "loss": 2.4918, + "step": 38116 + }, + { + "epoch": 1.62, + "grad_norm": 0.4453125, + "learning_rate": 1.698604070803099e-05, + "loss": 2.5469, + "step": 38117 + }, + { + "epoch": 1.62, + "grad_norm": 0.4765625, + "learning_rate": 1.698230084995489e-05, + "loss": 2.2927, + "step": 38118 + }, + { + "epoch": 1.62, + "grad_norm": 0.4453125, + "learning_rate": 1.697856136542808e-05, + "loss": 2.524, + "step": 38119 + }, + { + "epoch": 1.62, + "grad_norm": 0.48828125, + "learning_rate": 1.6974822254467406e-05, + "loss": 2.4346, + "step": 38120 + }, + { + "epoch": 1.62, + "grad_norm": 0.5234375, + "learning_rate": 1.6971083517089702e-05, + "loss": 2.4472, + "step": 38121 + }, + { + "epoch": 1.62, + "grad_norm": 0.439453125, + "learning_rate": 1.6967345153311787e-05, + "loss": 2.4856, + "step": 38122 + }, + { + "epoch": 1.62, + "grad_norm": 0.494140625, + "learning_rate": 1.6963607163150452e-05, + "loss": 2.3772, + "step": 38123 + }, + { + "epoch": 1.62, + "grad_norm": 0.51953125, + "learning_rate": 1.695986954662253e-05, + "loss": 2.579, + "step": 38124 + }, + { + "epoch": 1.62, + "grad_norm": 0.50390625, + "learning_rate": 1.6956132303744876e-05, + "loss": 2.5046, + "step": 38125 + }, + { + "epoch": 1.62, + "grad_norm": 0.431640625, + "learning_rate": 1.695239543453425e-05, + "loss": 2.5459, + "step": 38126 + }, + { + "epoch": 1.62, + "grad_norm": 0.498046875, + "learning_rate": 1.6948658939007523e-05, + "loss": 2.3651, + "step": 38127 + }, + { + "epoch": 1.62, + "grad_norm": 0.48046875, + "learning_rate": 1.694492281718145e-05, + "loss": 2.4871, + "step": 38128 + }, + { + "epoch": 1.62, + "grad_norm": 0.49609375, + "learning_rate": 1.6941187069072885e-05, + "loss": 2.5794, + "step": 38129 + }, + { + "epoch": 1.62, + "grad_norm": 0.46484375, + "learning_rate": 1.6937451694698637e-05, + "loss": 2.4296, + "step": 38130 + }, + { + "epoch": 1.62, + "grad_norm": 0.458984375, + "learning_rate": 1.6933716694075498e-05, + "loss": 2.3834, + "step": 38131 + }, + { + "epoch": 1.62, + "grad_norm": 0.4609375, + "learning_rate": 1.692998206722025e-05, + "loss": 2.5378, + "step": 38132 + }, + { + "epoch": 1.62, + "grad_norm": 0.46875, + "learning_rate": 1.692624781414973e-05, + "loss": 2.4818, + "step": 38133 + }, + { + "epoch": 1.62, + "grad_norm": 0.5078125, + "learning_rate": 1.692251393488076e-05, + "loss": 2.4461, + "step": 38134 + }, + { + "epoch": 1.63, + "grad_norm": 0.46484375, + "learning_rate": 1.6918780429430092e-05, + "loss": 2.4164, + "step": 38135 + }, + { + "epoch": 1.63, + "grad_norm": 0.4375, + "learning_rate": 1.6915047297814546e-05, + "loss": 2.4928, + "step": 38136 + }, + { + "epoch": 1.63, + "grad_norm": 0.498046875, + "learning_rate": 1.6911314540050948e-05, + "loss": 2.4478, + "step": 38137 + }, + { + "epoch": 1.63, + "grad_norm": 0.51953125, + "learning_rate": 1.6907582156156075e-05, + "loss": 2.5152, + "step": 38138 + }, + { + "epoch": 1.63, + "grad_norm": 0.419921875, + "learning_rate": 1.690385014614668e-05, + "loss": 2.4325, + "step": 38139 + }, + { + "epoch": 1.63, + "grad_norm": 0.478515625, + "learning_rate": 1.69001185100396e-05, + "loss": 2.4791, + "step": 38140 + }, + { + "epoch": 1.63, + "grad_norm": 0.515625, + "learning_rate": 1.6896387247851643e-05, + "loss": 2.5226, + "step": 38141 + }, + { + "epoch": 1.63, + "grad_norm": 0.51953125, + "learning_rate": 1.6892656359599545e-05, + "loss": 2.3644, + "step": 38142 + }, + { + "epoch": 1.63, + "grad_norm": 0.490234375, + "learning_rate": 1.688892584530014e-05, + "loss": 2.4709, + "step": 38143 + }, + { + "epoch": 1.63, + "grad_norm": 0.49609375, + "learning_rate": 1.688519570497018e-05, + "loss": 2.5337, + "step": 38144 + }, + { + "epoch": 1.63, + "grad_norm": 0.498046875, + "learning_rate": 1.688146593862645e-05, + "loss": 2.5022, + "step": 38145 + }, + { + "epoch": 1.63, + "grad_norm": 0.48828125, + "learning_rate": 1.6877736546285783e-05, + "loss": 2.5058, + "step": 38146 + }, + { + "epoch": 1.63, + "grad_norm": 0.4453125, + "learning_rate": 1.687400752796492e-05, + "loss": 2.4645, + "step": 38147 + }, + { + "epoch": 1.63, + "grad_norm": 0.486328125, + "learning_rate": 1.6870278883680624e-05, + "loss": 2.4787, + "step": 38148 + }, + { + "epoch": 1.63, + "grad_norm": 0.43359375, + "learning_rate": 1.6866550613449696e-05, + "loss": 2.4755, + "step": 38149 + }, + { + "epoch": 1.63, + "grad_norm": 0.427734375, + "learning_rate": 1.6862822717288917e-05, + "loss": 2.3931, + "step": 38150 + }, + { + "epoch": 1.63, + "grad_norm": 0.4453125, + "learning_rate": 1.6859095195215045e-05, + "loss": 2.3618, + "step": 38151 + }, + { + "epoch": 1.63, + "grad_norm": 0.490234375, + "learning_rate": 1.6855368047244845e-05, + "loss": 2.4547, + "step": 38152 + }, + { + "epoch": 1.63, + "grad_norm": 0.462890625, + "learning_rate": 1.6851641273395136e-05, + "loss": 2.4137, + "step": 38153 + }, + { + "epoch": 1.63, + "grad_norm": 0.4609375, + "learning_rate": 1.6847914873682657e-05, + "loss": 2.3876, + "step": 38154 + }, + { + "epoch": 1.63, + "grad_norm": 0.443359375, + "learning_rate": 1.6844188848124143e-05, + "loss": 2.3437, + "step": 38155 + }, + { + "epoch": 1.63, + "grad_norm": 0.5234375, + "learning_rate": 1.684046319673639e-05, + "loss": 2.3874, + "step": 38156 + }, + { + "epoch": 1.63, + "grad_norm": 0.46875, + "learning_rate": 1.6836737919536195e-05, + "loss": 2.3912, + "step": 38157 + }, + { + "epoch": 1.63, + "grad_norm": 0.42578125, + "learning_rate": 1.6833013016540245e-05, + "loss": 2.507, + "step": 38158 + }, + { + "epoch": 1.63, + "grad_norm": 0.443359375, + "learning_rate": 1.6829288487765382e-05, + "loss": 2.3875, + "step": 38159 + }, + { + "epoch": 1.63, + "grad_norm": 0.50390625, + "learning_rate": 1.6825564333228293e-05, + "loss": 2.5219, + "step": 38160 + }, + { + "epoch": 1.63, + "grad_norm": 0.5234375, + "learning_rate": 1.6821840552945766e-05, + "loss": 2.4825, + "step": 38161 + }, + { + "epoch": 1.63, + "grad_norm": 0.43359375, + "learning_rate": 1.6818117146934586e-05, + "loss": 2.3908, + "step": 38162 + }, + { + "epoch": 1.63, + "grad_norm": 0.46484375, + "learning_rate": 1.6814394115211483e-05, + "loss": 2.3109, + "step": 38163 + }, + { + "epoch": 1.63, + "grad_norm": 0.486328125, + "learning_rate": 1.6810671457793172e-05, + "loss": 2.443, + "step": 38164 + }, + { + "epoch": 1.63, + "grad_norm": 0.47265625, + "learning_rate": 1.680694917469644e-05, + "loss": 2.3935, + "step": 38165 + }, + { + "epoch": 1.63, + "grad_norm": 0.46484375, + "learning_rate": 1.6803227265938058e-05, + "loss": 2.4674, + "step": 38166 + }, + { + "epoch": 1.63, + "grad_norm": 0.427734375, + "learning_rate": 1.6799505731534737e-05, + "loss": 2.5554, + "step": 38167 + }, + { + "epoch": 1.63, + "grad_norm": 0.486328125, + "learning_rate": 1.6795784571503215e-05, + "loss": 2.3947, + "step": 38168 + }, + { + "epoch": 1.63, + "grad_norm": 0.49609375, + "learning_rate": 1.679206378586029e-05, + "loss": 2.5396, + "step": 38169 + }, + { + "epoch": 1.63, + "grad_norm": 0.4765625, + "learning_rate": 1.6788343374622663e-05, + "loss": 2.4569, + "step": 38170 + }, + { + "epoch": 1.63, + "grad_norm": 0.474609375, + "learning_rate": 1.678462333780706e-05, + "loss": 2.4359, + "step": 38171 + }, + { + "epoch": 1.63, + "grad_norm": 0.44921875, + "learning_rate": 1.6780903675430237e-05, + "loss": 2.3622, + "step": 38172 + }, + { + "epoch": 1.63, + "grad_norm": 0.53515625, + "learning_rate": 1.677718438750897e-05, + "loss": 2.5207, + "step": 38173 + }, + { + "epoch": 1.63, + "grad_norm": 0.48828125, + "learning_rate": 1.6773465474059925e-05, + "loss": 2.4434, + "step": 38174 + }, + { + "epoch": 1.63, + "grad_norm": 0.44921875, + "learning_rate": 1.6769746935099893e-05, + "loss": 2.4834, + "step": 38175 + }, + { + "epoch": 1.63, + "grad_norm": 0.49609375, + "learning_rate": 1.676602877064556e-05, + "loss": 2.432, + "step": 38176 + }, + { + "epoch": 1.63, + "grad_norm": 0.5234375, + "learning_rate": 1.676231098071368e-05, + "loss": 2.411, + "step": 38177 + }, + { + "epoch": 1.63, + "grad_norm": 0.48046875, + "learning_rate": 1.6758593565321e-05, + "loss": 2.3816, + "step": 38178 + }, + { + "epoch": 1.63, + "grad_norm": 0.435546875, + "learning_rate": 1.675487652448423e-05, + "loss": 2.2316, + "step": 38179 + }, + { + "epoch": 1.63, + "grad_norm": 0.478515625, + "learning_rate": 1.6751159858220068e-05, + "loss": 2.4785, + "step": 38180 + }, + { + "epoch": 1.63, + "grad_norm": 0.40234375, + "learning_rate": 1.6747443566545274e-05, + "loss": 2.5, + "step": 38181 + }, + { + "epoch": 1.63, + "grad_norm": 0.49609375, + "learning_rate": 1.674372764947657e-05, + "loss": 2.7598, + "step": 38182 + }, + { + "epoch": 1.63, + "grad_norm": 0.416015625, + "learning_rate": 1.6740012107030645e-05, + "loss": 2.4219, + "step": 38183 + }, + { + "epoch": 1.63, + "grad_norm": 0.42578125, + "learning_rate": 1.6736296939224237e-05, + "loss": 2.2318, + "step": 38184 + }, + { + "epoch": 1.63, + "grad_norm": 0.5078125, + "learning_rate": 1.673258214607408e-05, + "loss": 2.4523, + "step": 38185 + }, + { + "epoch": 1.63, + "grad_norm": 0.48828125, + "learning_rate": 1.6728867727596885e-05, + "loss": 2.4807, + "step": 38186 + }, + { + "epoch": 1.63, + "grad_norm": 0.484375, + "learning_rate": 1.6725153683809324e-05, + "loss": 2.4166, + "step": 38187 + }, + { + "epoch": 1.63, + "grad_norm": 0.439453125, + "learning_rate": 1.672144001472814e-05, + "loss": 2.514, + "step": 38188 + }, + { + "epoch": 1.63, + "grad_norm": 0.4765625, + "learning_rate": 1.6717726720370065e-05, + "loss": 2.4582, + "step": 38189 + }, + { + "epoch": 1.63, + "grad_norm": 0.400390625, + "learning_rate": 1.671401380075175e-05, + "loss": 2.4251, + "step": 38190 + }, + { + "epoch": 1.63, + "grad_norm": 0.435546875, + "learning_rate": 1.6710301255889972e-05, + "loss": 2.3487, + "step": 38191 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.670658908580137e-05, + "loss": 2.4527, + "step": 38192 + }, + { + "epoch": 1.63, + "grad_norm": 0.43359375, + "learning_rate": 1.670287729050267e-05, + "loss": 2.4624, + "step": 38193 + }, + { + "epoch": 1.63, + "grad_norm": 0.498046875, + "learning_rate": 1.6699165870010612e-05, + "loss": 2.5824, + "step": 38194 + }, + { + "epoch": 1.63, + "grad_norm": 0.47265625, + "learning_rate": 1.669545482434187e-05, + "loss": 2.3947, + "step": 38195 + }, + { + "epoch": 1.63, + "grad_norm": 0.4921875, + "learning_rate": 1.6691744153513112e-05, + "loss": 2.3362, + "step": 38196 + }, + { + "epoch": 1.63, + "grad_norm": 0.4609375, + "learning_rate": 1.6688033857541052e-05, + "loss": 2.4522, + "step": 38197 + }, + { + "epoch": 1.63, + "grad_norm": 0.47265625, + "learning_rate": 1.6684323936442413e-05, + "loss": 2.2816, + "step": 38198 + }, + { + "epoch": 1.63, + "grad_norm": 0.455078125, + "learning_rate": 1.668061439023385e-05, + "loss": 2.4326, + "step": 38199 + }, + { + "epoch": 1.63, + "grad_norm": 0.453125, + "learning_rate": 1.6676905218932104e-05, + "loss": 2.4798, + "step": 38200 + }, + { + "epoch": 1.63, + "grad_norm": 0.5234375, + "learning_rate": 1.6673196422553804e-05, + "loss": 2.4461, + "step": 38201 + }, + { + "epoch": 1.63, + "grad_norm": 0.451171875, + "learning_rate": 1.6669488001115685e-05, + "loss": 2.4383, + "step": 38202 + }, + { + "epoch": 1.63, + "grad_norm": 0.388671875, + "learning_rate": 1.6665779954634397e-05, + "loss": 2.5267, + "step": 38203 + }, + { + "epoch": 1.63, + "grad_norm": 0.451171875, + "learning_rate": 1.6662072283126673e-05, + "loss": 2.4671, + "step": 38204 + }, + { + "epoch": 1.63, + "grad_norm": 0.404296875, + "learning_rate": 1.665836498660913e-05, + "loss": 2.5138, + "step": 38205 + }, + { + "epoch": 1.63, + "grad_norm": 0.43359375, + "learning_rate": 1.6654658065098504e-05, + "loss": 2.4311, + "step": 38206 + }, + { + "epoch": 1.63, + "grad_norm": 0.4296875, + "learning_rate": 1.665095151861147e-05, + "loss": 2.4249, + "step": 38207 + }, + { + "epoch": 1.63, + "grad_norm": 0.443359375, + "learning_rate": 1.66472453471647e-05, + "loss": 2.3468, + "step": 38208 + }, + { + "epoch": 1.63, + "grad_norm": 0.46484375, + "learning_rate": 1.6643539550774813e-05, + "loss": 2.5375, + "step": 38209 + }, + { + "epoch": 1.63, + "grad_norm": 0.421875, + "learning_rate": 1.6639834129458577e-05, + "loss": 2.5631, + "step": 38210 + }, + { + "epoch": 1.63, + "grad_norm": 0.4921875, + "learning_rate": 1.6636129083232633e-05, + "loss": 2.4362, + "step": 38211 + }, + { + "epoch": 1.63, + "grad_norm": 0.462890625, + "learning_rate": 1.663242441211361e-05, + "loss": 2.5105, + "step": 38212 + }, + { + "epoch": 1.63, + "grad_norm": 0.40625, + "learning_rate": 1.6628720116118212e-05, + "loss": 2.5918, + "step": 38213 + }, + { + "epoch": 1.63, + "grad_norm": 0.482421875, + "learning_rate": 1.6625016195263132e-05, + "loss": 2.3465, + "step": 38214 + }, + { + "epoch": 1.63, + "grad_norm": 0.45703125, + "learning_rate": 1.662131264956498e-05, + "loss": 2.3472, + "step": 38215 + }, + { + "epoch": 1.63, + "grad_norm": 0.412109375, + "learning_rate": 1.6617609479040476e-05, + "loss": 2.3446, + "step": 38216 + }, + { + "epoch": 1.63, + "grad_norm": 0.447265625, + "learning_rate": 1.6613906683706225e-05, + "loss": 2.4666, + "step": 38217 + }, + { + "epoch": 1.63, + "grad_norm": 0.423828125, + "learning_rate": 1.6610204263578953e-05, + "loss": 2.5195, + "step": 38218 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.6606502218675267e-05, + "loss": 2.455, + "step": 38219 + }, + { + "epoch": 1.63, + "grad_norm": 0.50390625, + "learning_rate": 1.6602800549011855e-05, + "loss": 2.4989, + "step": 38220 + }, + { + "epoch": 1.63, + "grad_norm": 0.49609375, + "learning_rate": 1.6599099254605343e-05, + "loss": 2.47, + "step": 38221 + }, + { + "epoch": 1.63, + "grad_norm": 0.435546875, + "learning_rate": 1.659539833547241e-05, + "loss": 2.4391, + "step": 38222 + }, + { + "epoch": 1.63, + "grad_norm": 0.4765625, + "learning_rate": 1.6591697791629722e-05, + "loss": 2.5003, + "step": 38223 + }, + { + "epoch": 1.63, + "grad_norm": 0.482421875, + "learning_rate": 1.658799762309392e-05, + "loss": 2.5402, + "step": 38224 + }, + { + "epoch": 1.63, + "grad_norm": 0.43359375, + "learning_rate": 1.658429782988159e-05, + "loss": 2.3469, + "step": 38225 + }, + { + "epoch": 1.63, + "grad_norm": 0.423828125, + "learning_rate": 1.6580598412009483e-05, + "loss": 2.3913, + "step": 38226 + }, + { + "epoch": 1.63, + "grad_norm": 0.46484375, + "learning_rate": 1.6576899369494204e-05, + "loss": 2.4556, + "step": 38227 + }, + { + "epoch": 1.63, + "grad_norm": 0.4609375, + "learning_rate": 1.6573200702352353e-05, + "loss": 2.3434, + "step": 38228 + }, + { + "epoch": 1.63, + "grad_norm": 0.53125, + "learning_rate": 1.6569502410600624e-05, + "loss": 2.5507, + "step": 38229 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.6565804494255655e-05, + "loss": 2.485, + "step": 38230 + }, + { + "epoch": 1.63, + "grad_norm": 0.435546875, + "learning_rate": 1.6562106953334056e-05, + "loss": 2.4319, + "step": 38231 + }, + { + "epoch": 1.63, + "grad_norm": 0.455078125, + "learning_rate": 1.6558409787852512e-05, + "loss": 2.4452, + "step": 38232 + }, + { + "epoch": 1.63, + "grad_norm": 0.51953125, + "learning_rate": 1.6554712997827592e-05, + "loss": 2.4351, + "step": 38233 + }, + { + "epoch": 1.63, + "grad_norm": 0.48046875, + "learning_rate": 1.6551016583276003e-05, + "loss": 2.5181, + "step": 38234 + }, + { + "epoch": 1.63, + "grad_norm": 0.439453125, + "learning_rate": 1.6547320544214316e-05, + "loss": 2.4189, + "step": 38235 + }, + { + "epoch": 1.63, + "grad_norm": 0.44921875, + "learning_rate": 1.6543624880659214e-05, + "loss": 2.4015, + "step": 38236 + }, + { + "epoch": 1.63, + "grad_norm": 0.4453125, + "learning_rate": 1.653992959262727e-05, + "loss": 2.4507, + "step": 38237 + }, + { + "epoch": 1.63, + "grad_norm": 0.46875, + "learning_rate": 1.6536234680135143e-05, + "loss": 2.5817, + "step": 38238 + }, + { + "epoch": 1.63, + "grad_norm": 0.52734375, + "learning_rate": 1.6532540143199482e-05, + "loss": 2.4201, + "step": 38239 + }, + { + "epoch": 1.63, + "grad_norm": 0.478515625, + "learning_rate": 1.6528845981836892e-05, + "loss": 2.4074, + "step": 38240 + }, + { + "epoch": 1.63, + "grad_norm": 0.412109375, + "learning_rate": 1.6525152196063944e-05, + "loss": 2.3654, + "step": 38241 + }, + { + "epoch": 1.63, + "grad_norm": 0.484375, + "learning_rate": 1.6521458785897337e-05, + "loss": 2.4537, + "step": 38242 + }, + { + "epoch": 1.63, + "grad_norm": 0.443359375, + "learning_rate": 1.6517765751353675e-05, + "loss": 2.5636, + "step": 38243 + }, + { + "epoch": 1.63, + "grad_norm": 0.4921875, + "learning_rate": 1.6514073092449524e-05, + "loss": 2.517, + "step": 38244 + }, + { + "epoch": 1.63, + "grad_norm": 0.51953125, + "learning_rate": 1.651038080920153e-05, + "loss": 2.5614, + "step": 38245 + }, + { + "epoch": 1.63, + "grad_norm": 0.443359375, + "learning_rate": 1.6506688901626345e-05, + "loss": 2.363, + "step": 38246 + }, + { + "epoch": 1.63, + "grad_norm": 0.484375, + "learning_rate": 1.6502997369740513e-05, + "loss": 2.5435, + "step": 38247 + }, + { + "epoch": 1.63, + "grad_norm": 0.50390625, + "learning_rate": 1.6499306213560705e-05, + "loss": 2.7011, + "step": 38248 + }, + { + "epoch": 1.63, + "grad_norm": 0.474609375, + "learning_rate": 1.6495615433103483e-05, + "loss": 2.5841, + "step": 38249 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.6491925028385503e-05, + "loss": 2.6114, + "step": 38250 + }, + { + "epoch": 1.63, + "grad_norm": 0.443359375, + "learning_rate": 1.64882349994233e-05, + "loss": 2.5333, + "step": 38251 + }, + { + "epoch": 1.63, + "grad_norm": 0.427734375, + "learning_rate": 1.648454534623356e-05, + "loss": 2.4255, + "step": 38252 + }, + { + "epoch": 1.63, + "grad_norm": 0.439453125, + "learning_rate": 1.6480856068832818e-05, + "loss": 2.368, + "step": 38253 + }, + { + "epoch": 1.63, + "grad_norm": 0.458984375, + "learning_rate": 1.6477167167237696e-05, + "loss": 2.4142, + "step": 38254 + }, + { + "epoch": 1.63, + "grad_norm": 0.427734375, + "learning_rate": 1.6473478641464824e-05, + "loss": 2.4307, + "step": 38255 + }, + { + "epoch": 1.63, + "grad_norm": 0.45703125, + "learning_rate": 1.646979049153077e-05, + "loss": 2.4553, + "step": 38256 + }, + { + "epoch": 1.63, + "grad_norm": 0.427734375, + "learning_rate": 1.6466102717452082e-05, + "loss": 2.3377, + "step": 38257 + }, + { + "epoch": 1.63, + "grad_norm": 0.482421875, + "learning_rate": 1.646241531924545e-05, + "loss": 2.3378, + "step": 38258 + }, + { + "epoch": 1.63, + "grad_norm": 0.4375, + "learning_rate": 1.645872829692743e-05, + "loss": 2.3786, + "step": 38259 + }, + { + "epoch": 1.63, + "grad_norm": 0.431640625, + "learning_rate": 1.6455041650514568e-05, + "loss": 2.313, + "step": 38260 + }, + { + "epoch": 1.63, + "grad_norm": 0.4921875, + "learning_rate": 1.6451355380023493e-05, + "loss": 2.4928, + "step": 38261 + }, + { + "epoch": 1.63, + "grad_norm": 0.486328125, + "learning_rate": 1.6447669485470808e-05, + "loss": 2.5225, + "step": 38262 + }, + { + "epoch": 1.63, + "grad_norm": 0.431640625, + "learning_rate": 1.6443983966873056e-05, + "loss": 2.4378, + "step": 38263 + }, + { + "epoch": 1.63, + "grad_norm": 0.486328125, + "learning_rate": 1.6440298824246857e-05, + "loss": 2.4991, + "step": 38264 + }, + { + "epoch": 1.63, + "grad_norm": 0.42578125, + "learning_rate": 1.6436614057608746e-05, + "loss": 2.4103, + "step": 38265 + }, + { + "epoch": 1.63, + "grad_norm": 0.470703125, + "learning_rate": 1.643292966697537e-05, + "loss": 2.5012, + "step": 38266 + }, + { + "epoch": 1.63, + "grad_norm": 0.486328125, + "learning_rate": 1.6429245652363236e-05, + "loss": 2.4755, + "step": 38267 + }, + { + "epoch": 1.63, + "grad_norm": 0.47265625, + "learning_rate": 1.6425562013788985e-05, + "loss": 2.6218, + "step": 38268 + }, + { + "epoch": 1.63, + "grad_norm": 0.44140625, + "learning_rate": 1.6421878751269136e-05, + "loss": 2.4298, + "step": 38269 + }, + { + "epoch": 1.63, + "grad_norm": 0.427734375, + "learning_rate": 1.6418195864820295e-05, + "loss": 2.5009, + "step": 38270 + }, + { + "epoch": 1.63, + "grad_norm": 0.453125, + "learning_rate": 1.641451335445904e-05, + "loss": 2.2789, + "step": 38271 + }, + { + "epoch": 1.63, + "grad_norm": 0.48828125, + "learning_rate": 1.641083122020193e-05, + "loss": 2.318, + "step": 38272 + }, + { + "epoch": 1.63, + "grad_norm": 0.4453125, + "learning_rate": 1.6407149462065487e-05, + "loss": 2.3748, + "step": 38273 + }, + { + "epoch": 1.63, + "grad_norm": 0.435546875, + "learning_rate": 1.6403468080066364e-05, + "loss": 2.4341, + "step": 38274 + }, + { + "epoch": 1.63, + "grad_norm": 0.56640625, + "learning_rate": 1.6399787074221096e-05, + "loss": 2.4729, + "step": 38275 + }, + { + "epoch": 1.63, + "grad_norm": 0.400390625, + "learning_rate": 1.63961064445462e-05, + "loss": 2.4513, + "step": 38276 + }, + { + "epoch": 1.63, + "grad_norm": 0.48046875, + "learning_rate": 1.6392426191058273e-05, + "loss": 2.4068, + "step": 38277 + }, + { + "epoch": 1.63, + "grad_norm": 0.49609375, + "learning_rate": 1.6388746313773896e-05, + "loss": 2.4763, + "step": 38278 + }, + { + "epoch": 1.63, + "grad_norm": 0.4609375, + "learning_rate": 1.6385066812709584e-05, + "loss": 2.3796, + "step": 38279 + }, + { + "epoch": 1.63, + "grad_norm": 0.490234375, + "learning_rate": 1.6381387687881934e-05, + "loss": 2.383, + "step": 38280 + }, + { + "epoch": 1.63, + "grad_norm": 0.53125, + "learning_rate": 1.6377708939307455e-05, + "loss": 2.5589, + "step": 38281 + }, + { + "epoch": 1.63, + "grad_norm": 0.53125, + "learning_rate": 1.6374030567002752e-05, + "loss": 2.4787, + "step": 38282 + }, + { + "epoch": 1.63, + "grad_norm": 0.50390625, + "learning_rate": 1.6370352570984327e-05, + "loss": 2.5522, + "step": 38283 + }, + { + "epoch": 1.63, + "grad_norm": 0.46875, + "learning_rate": 1.6366674951268767e-05, + "loss": 2.3076, + "step": 38284 + }, + { + "epoch": 1.63, + "grad_norm": 0.412109375, + "learning_rate": 1.6362997707872597e-05, + "loss": 2.473, + "step": 38285 + }, + { + "epoch": 1.63, + "grad_norm": 0.431640625, + "learning_rate": 1.6359320840812363e-05, + "loss": 2.4441, + "step": 38286 + }, + { + "epoch": 1.63, + "grad_norm": 0.51953125, + "learning_rate": 1.6355644350104636e-05, + "loss": 2.4768, + "step": 38287 + }, + { + "epoch": 1.63, + "grad_norm": 0.431640625, + "learning_rate": 1.6351968235765947e-05, + "loss": 2.3104, + "step": 38288 + }, + { + "epoch": 1.63, + "grad_norm": 0.45703125, + "learning_rate": 1.6348292497812788e-05, + "loss": 2.332, + "step": 38289 + }, + { + "epoch": 1.63, + "grad_norm": 0.451171875, + "learning_rate": 1.6344617136261786e-05, + "loss": 2.5374, + "step": 38290 + }, + { + "epoch": 1.63, + "grad_norm": 0.47265625, + "learning_rate": 1.6340942151129434e-05, + "loss": 2.5761, + "step": 38291 + }, + { + "epoch": 1.63, + "grad_norm": 0.474609375, + "learning_rate": 1.6337267542432246e-05, + "loss": 2.4458, + "step": 38292 + }, + { + "epoch": 1.63, + "grad_norm": 0.439453125, + "learning_rate": 1.633359331018677e-05, + "loss": 2.5037, + "step": 38293 + }, + { + "epoch": 1.63, + "grad_norm": 0.4765625, + "learning_rate": 1.632991945440958e-05, + "loss": 2.4482, + "step": 38294 + }, + { + "epoch": 1.63, + "grad_norm": 0.46875, + "learning_rate": 1.632624597511715e-05, + "loss": 2.4731, + "step": 38295 + }, + { + "epoch": 1.63, + "grad_norm": 0.466796875, + "learning_rate": 1.6322572872326048e-05, + "loss": 2.4455, + "step": 38296 + }, + { + "epoch": 1.63, + "grad_norm": 0.435546875, + "learning_rate": 1.6318900146052774e-05, + "loss": 2.4722, + "step": 38297 + }, + { + "epoch": 1.63, + "grad_norm": 0.5390625, + "learning_rate": 1.631522779631389e-05, + "loss": 2.5507, + "step": 38298 + }, + { + "epoch": 1.63, + "grad_norm": 0.47265625, + "learning_rate": 1.6311555823125868e-05, + "loss": 2.5437, + "step": 38299 + }, + { + "epoch": 1.63, + "grad_norm": 0.453125, + "learning_rate": 1.63078842265053e-05, + "loss": 2.379, + "step": 38300 + }, + { + "epoch": 1.63, + "grad_norm": 0.48046875, + "learning_rate": 1.6304213006468626e-05, + "loss": 2.5527, + "step": 38301 + }, + { + "epoch": 1.63, + "grad_norm": 0.49609375, + "learning_rate": 1.6300542163032405e-05, + "loss": 2.5215, + "step": 38302 + }, + { + "epoch": 1.63, + "grad_norm": 0.4375, + "learning_rate": 1.629687169621319e-05, + "loss": 2.4312, + "step": 38303 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.6293201606027464e-05, + "loss": 2.4229, + "step": 38304 + }, + { + "epoch": 1.63, + "grad_norm": 0.59375, + "learning_rate": 1.6289531892491683e-05, + "loss": 2.4615, + "step": 38305 + }, + { + "epoch": 1.63, + "grad_norm": 0.515625, + "learning_rate": 1.6285862555622465e-05, + "loss": 2.4397, + "step": 38306 + }, + { + "epoch": 1.63, + "grad_norm": 0.48828125, + "learning_rate": 1.628219359543628e-05, + "loss": 2.4749, + "step": 38307 + }, + { + "epoch": 1.63, + "grad_norm": 0.462890625, + "learning_rate": 1.6278525011949597e-05, + "loss": 2.4261, + "step": 38308 + }, + { + "epoch": 1.63, + "grad_norm": 0.494140625, + "learning_rate": 1.627485680517895e-05, + "loss": 2.565, + "step": 38309 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.627118897514088e-05, + "loss": 2.4883, + "step": 38310 + }, + { + "epoch": 1.63, + "grad_norm": 0.4765625, + "learning_rate": 1.6267521521851836e-05, + "loss": 2.4486, + "step": 38311 + }, + { + "epoch": 1.63, + "grad_norm": 0.412109375, + "learning_rate": 1.6263854445328374e-05, + "loss": 2.5609, + "step": 38312 + }, + { + "epoch": 1.63, + "grad_norm": 0.408203125, + "learning_rate": 1.626018774558693e-05, + "loss": 2.3457, + "step": 38313 + }, + { + "epoch": 1.63, + "grad_norm": 0.494140625, + "learning_rate": 1.6256521422644043e-05, + "loss": 2.43, + "step": 38314 + }, + { + "epoch": 1.63, + "grad_norm": 0.494140625, + "learning_rate": 1.6252855476516228e-05, + "loss": 2.4828, + "step": 38315 + }, + { + "epoch": 1.63, + "grad_norm": 0.478515625, + "learning_rate": 1.6249189907219964e-05, + "loss": 2.5073, + "step": 38316 + }, + { + "epoch": 1.63, + "grad_norm": 0.4375, + "learning_rate": 1.6245524714771698e-05, + "loss": 2.4099, + "step": 38317 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.6241859899187973e-05, + "loss": 2.387, + "step": 38318 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.6238195460485294e-05, + "loss": 2.3946, + "step": 38319 + }, + { + "epoch": 1.63, + "grad_norm": 0.466796875, + "learning_rate": 1.623453139868012e-05, + "loss": 2.4775, + "step": 38320 + }, + { + "epoch": 1.63, + "grad_norm": 0.490234375, + "learning_rate": 1.6230867713788922e-05, + "loss": 2.572, + "step": 38321 + }, + { + "epoch": 1.63, + "grad_norm": 0.431640625, + "learning_rate": 1.62272044058282e-05, + "loss": 2.4066, + "step": 38322 + }, + { + "epoch": 1.63, + "grad_norm": 0.482421875, + "learning_rate": 1.6223541474814473e-05, + "loss": 2.3679, + "step": 38323 + }, + { + "epoch": 1.63, + "grad_norm": 0.4609375, + "learning_rate": 1.6219878920764164e-05, + "loss": 2.5814, + "step": 38324 + }, + { + "epoch": 1.63, + "grad_norm": 0.49609375, + "learning_rate": 1.6216216743693802e-05, + "loss": 2.4236, + "step": 38325 + }, + { + "epoch": 1.63, + "grad_norm": 0.48046875, + "learning_rate": 1.621255494361983e-05, + "loss": 2.4999, + "step": 38326 + }, + { + "epoch": 1.63, + "grad_norm": 0.435546875, + "learning_rate": 1.6208893520558732e-05, + "loss": 2.3421, + "step": 38327 + }, + { + "epoch": 1.63, + "grad_norm": 0.39453125, + "learning_rate": 1.620523247452702e-05, + "loss": 2.3275, + "step": 38328 + }, + { + "epoch": 1.63, + "grad_norm": 0.51171875, + "learning_rate": 1.6201571805541138e-05, + "loss": 2.5491, + "step": 38329 + }, + { + "epoch": 1.63, + "grad_norm": 0.453125, + "learning_rate": 1.6197911513617502e-05, + "loss": 2.3389, + "step": 38330 + }, + { + "epoch": 1.63, + "grad_norm": 0.470703125, + "learning_rate": 1.6194251598772704e-05, + "loss": 2.4463, + "step": 38331 + }, + { + "epoch": 1.63, + "grad_norm": 0.51171875, + "learning_rate": 1.6190592061023125e-05, + "loss": 2.4635, + "step": 38332 + }, + { + "epoch": 1.63, + "grad_norm": 0.4609375, + "learning_rate": 1.6186932900385242e-05, + "loss": 2.5518, + "step": 38333 + }, + { + "epoch": 1.63, + "grad_norm": 0.515625, + "learning_rate": 1.6183274116875524e-05, + "loss": 2.493, + "step": 38334 + }, + { + "epoch": 1.63, + "grad_norm": 0.486328125, + "learning_rate": 1.6179615710510475e-05, + "loss": 2.4473, + "step": 38335 + }, + { + "epoch": 1.63, + "grad_norm": 0.47265625, + "learning_rate": 1.617595768130652e-05, + "loss": 2.4219, + "step": 38336 + }, + { + "epoch": 1.63, + "grad_norm": 0.435546875, + "learning_rate": 1.6172300029280087e-05, + "loss": 2.4139, + "step": 38337 + }, + { + "epoch": 1.63, + "grad_norm": 0.50390625, + "learning_rate": 1.616864275444767e-05, + "loss": 2.3627, + "step": 38338 + }, + { + "epoch": 1.63, + "grad_norm": 0.482421875, + "learning_rate": 1.6164985856825754e-05, + "loss": 2.4236, + "step": 38339 + }, + { + "epoch": 1.63, + "grad_norm": 0.44140625, + "learning_rate": 1.6161329336430743e-05, + "loss": 2.4074, + "step": 38340 + }, + { + "epoch": 1.63, + "grad_norm": 0.451171875, + "learning_rate": 1.615767319327913e-05, + "loss": 2.3752, + "step": 38341 + }, + { + "epoch": 1.63, + "grad_norm": 0.431640625, + "learning_rate": 1.6154017427387324e-05, + "loss": 2.4382, + "step": 38342 + }, + { + "epoch": 1.63, + "grad_norm": 0.439453125, + "learning_rate": 1.6150362038771795e-05, + "loss": 2.3608, + "step": 38343 + }, + { + "epoch": 1.63, + "grad_norm": 0.400390625, + "learning_rate": 1.6146707027449016e-05, + "loss": 2.4441, + "step": 38344 + }, + { + "epoch": 1.63, + "grad_norm": 0.5234375, + "learning_rate": 1.614305239343541e-05, + "loss": 2.6253, + "step": 38345 + }, + { + "epoch": 1.63, + "grad_norm": 0.490234375, + "learning_rate": 1.6139398136747375e-05, + "loss": 2.3718, + "step": 38346 + }, + { + "epoch": 1.63, + "grad_norm": 0.484375, + "learning_rate": 1.6135744257401453e-05, + "loss": 2.5188, + "step": 38347 + }, + { + "epoch": 1.63, + "grad_norm": 0.484375, + "learning_rate": 1.6132090755414032e-05, + "loss": 2.4288, + "step": 38348 + }, + { + "epoch": 1.63, + "grad_norm": 0.5234375, + "learning_rate": 1.612843763080153e-05, + "loss": 2.4091, + "step": 38349 + }, + { + "epoch": 1.63, + "grad_norm": 0.451171875, + "learning_rate": 1.61247848835804e-05, + "loss": 2.3951, + "step": 38350 + }, + { + "epoch": 1.63, + "grad_norm": 0.453125, + "learning_rate": 1.612113251376711e-05, + "loss": 2.4059, + "step": 38351 + }, + { + "epoch": 1.63, + "grad_norm": 0.474609375, + "learning_rate": 1.611748052137808e-05, + "loss": 2.5322, + "step": 38352 + }, + { + "epoch": 1.63, + "grad_norm": 0.439453125, + "learning_rate": 1.6113828906429694e-05, + "loss": 2.354, + "step": 38353 + }, + { + "epoch": 1.63, + "grad_norm": 0.494140625, + "learning_rate": 1.611017766893843e-05, + "loss": 2.5088, + "step": 38354 + }, + { + "epoch": 1.63, + "grad_norm": 0.490234375, + "learning_rate": 1.610652680892073e-05, + "loss": 2.5361, + "step": 38355 + }, + { + "epoch": 1.63, + "grad_norm": 0.47265625, + "learning_rate": 1.6102876326392978e-05, + "loss": 2.3465, + "step": 38356 + }, + { + "epoch": 1.63, + "grad_norm": 0.4453125, + "learning_rate": 1.6099226221371645e-05, + "loss": 2.5281, + "step": 38357 + }, + { + "epoch": 1.63, + "grad_norm": 0.404296875, + "learning_rate": 1.60955764938731e-05, + "loss": 2.3965, + "step": 38358 + }, + { + "epoch": 1.63, + "grad_norm": 0.455078125, + "learning_rate": 1.60919271439138e-05, + "loss": 2.3189, + "step": 38359 + }, + { + "epoch": 1.63, + "grad_norm": 0.48046875, + "learning_rate": 1.6088278171510195e-05, + "loss": 2.4464, + "step": 38360 + }, + { + "epoch": 1.63, + "grad_norm": 0.5, + "learning_rate": 1.6084629576678666e-05, + "loss": 2.3873, + "step": 38361 + }, + { + "epoch": 1.63, + "grad_norm": 0.42578125, + "learning_rate": 1.6080981359435587e-05, + "loss": 2.4098, + "step": 38362 + }, + { + "epoch": 1.63, + "grad_norm": 0.41015625, + "learning_rate": 1.607733351979748e-05, + "loss": 2.3135, + "step": 38363 + }, + { + "epoch": 1.63, + "grad_norm": 0.4921875, + "learning_rate": 1.607368605778069e-05, + "loss": 2.3654, + "step": 38364 + }, + { + "epoch": 1.63, + "grad_norm": 0.453125, + "learning_rate": 1.6070038973401624e-05, + "loss": 2.4665, + "step": 38365 + }, + { + "epoch": 1.63, + "grad_norm": 0.470703125, + "learning_rate": 1.606639226667671e-05, + "loss": 2.3051, + "step": 38366 + }, + { + "epoch": 1.63, + "grad_norm": 0.53515625, + "learning_rate": 1.606274593762238e-05, + "loss": 2.5969, + "step": 38367 + }, + { + "epoch": 1.63, + "grad_norm": 0.5078125, + "learning_rate": 1.605909998625499e-05, + "loss": 2.4068, + "step": 38368 + }, + { + "epoch": 1.64, + "grad_norm": 0.462890625, + "learning_rate": 1.6055454412591e-05, + "loss": 2.3772, + "step": 38369 + }, + { + "epoch": 1.64, + "grad_norm": 0.4921875, + "learning_rate": 1.605180921664676e-05, + "loss": 2.5896, + "step": 38370 + }, + { + "epoch": 1.64, + "grad_norm": 0.45703125, + "learning_rate": 1.6048164398438726e-05, + "loss": 2.3905, + "step": 38371 + }, + { + "epoch": 1.64, + "grad_norm": 0.3984375, + "learning_rate": 1.604451995798325e-05, + "loss": 2.3631, + "step": 38372 + }, + { + "epoch": 1.64, + "grad_norm": 0.44921875, + "learning_rate": 1.604087589529677e-05, + "loss": 2.5131, + "step": 38373 + }, + { + "epoch": 1.64, + "grad_norm": 0.51953125, + "learning_rate": 1.6037232210395646e-05, + "loss": 2.4662, + "step": 38374 + }, + { + "epoch": 1.64, + "grad_norm": 0.46484375, + "learning_rate": 1.6033588903296293e-05, + "loss": 2.3704, + "step": 38375 + }, + { + "epoch": 1.64, + "grad_norm": 0.51171875, + "learning_rate": 1.6029945974015126e-05, + "loss": 2.4046, + "step": 38376 + }, + { + "epoch": 1.64, + "grad_norm": 0.46484375, + "learning_rate": 1.6026303422568513e-05, + "loss": 2.41, + "step": 38377 + }, + { + "epoch": 1.64, + "grad_norm": 0.44140625, + "learning_rate": 1.6022661248972794e-05, + "loss": 2.4867, + "step": 38378 + }, + { + "epoch": 1.64, + "grad_norm": 0.4453125, + "learning_rate": 1.6019019453244466e-05, + "loss": 2.3043, + "step": 38379 + }, + { + "epoch": 1.64, + "grad_norm": 0.48046875, + "learning_rate": 1.6015378035399854e-05, + "loss": 2.4264, + "step": 38380 + }, + { + "epoch": 1.64, + "grad_norm": 0.49609375, + "learning_rate": 1.601173699545532e-05, + "loss": 2.4142, + "step": 38381 + }, + { + "epoch": 1.64, + "grad_norm": 0.43359375, + "learning_rate": 1.6008096333427282e-05, + "loss": 2.4848, + "step": 38382 + }, + { + "epoch": 1.64, + "grad_norm": 0.466796875, + "learning_rate": 1.600445604933214e-05, + "loss": 2.4617, + "step": 38383 + }, + { + "epoch": 1.64, + "grad_norm": 0.486328125, + "learning_rate": 1.6000816143186215e-05, + "loss": 2.5085, + "step": 38384 + }, + { + "epoch": 1.64, + "grad_norm": 0.51953125, + "learning_rate": 1.5997176615005948e-05, + "loss": 2.4643, + "step": 38385 + }, + { + "epoch": 1.64, + "grad_norm": 0.52734375, + "learning_rate": 1.5993537464807663e-05, + "loss": 2.5035, + "step": 38386 + }, + { + "epoch": 1.64, + "grad_norm": 0.453125, + "learning_rate": 1.5989898692607784e-05, + "loss": 2.5365, + "step": 38387 + }, + { + "epoch": 1.64, + "grad_norm": 0.431640625, + "learning_rate": 1.598626029842264e-05, + "loss": 2.3814, + "step": 38388 + }, + { + "epoch": 1.64, + "grad_norm": 0.5078125, + "learning_rate": 1.5982622282268632e-05, + "loss": 2.4915, + "step": 38389 + }, + { + "epoch": 1.64, + "grad_norm": 0.484375, + "learning_rate": 1.597898464416211e-05, + "loss": 2.42, + "step": 38390 + }, + { + "epoch": 1.64, + "grad_norm": 0.5546875, + "learning_rate": 1.5975347384119444e-05, + "loss": 2.7492, + "step": 38391 + }, + { + "epoch": 1.64, + "grad_norm": 0.53125, + "learning_rate": 1.5971710502157034e-05, + "loss": 2.3901, + "step": 38392 + }, + { + "epoch": 1.64, + "grad_norm": 0.47265625, + "learning_rate": 1.5968073998291212e-05, + "loss": 2.4214, + "step": 38393 + }, + { + "epoch": 1.64, + "grad_norm": 0.482421875, + "learning_rate": 1.5964437872538296e-05, + "loss": 2.5687, + "step": 38394 + }, + { + "epoch": 1.64, + "grad_norm": 0.45703125, + "learning_rate": 1.5960802124914754e-05, + "loss": 2.3641, + "step": 38395 + }, + { + "epoch": 1.64, + "grad_norm": 0.451171875, + "learning_rate": 1.5957166755436892e-05, + "loss": 2.4705, + "step": 38396 + }, + { + "epoch": 1.64, + "grad_norm": 0.474609375, + "learning_rate": 1.5953531764121033e-05, + "loss": 2.3876, + "step": 38397 + }, + { + "epoch": 1.64, + "grad_norm": 0.55078125, + "learning_rate": 1.594989715098357e-05, + "loss": 2.369, + "step": 38398 + }, + { + "epoch": 1.64, + "grad_norm": 0.494140625, + "learning_rate": 1.5946262916040877e-05, + "loss": 2.4537, + "step": 38399 + }, + { + "epoch": 1.64, + "grad_norm": 0.408203125, + "learning_rate": 1.5942629059309255e-05, + "loss": 2.3793, + "step": 38400 + }, + { + "epoch": 1.64, + "grad_norm": 0.466796875, + "learning_rate": 1.5938995580805105e-05, + "loss": 2.3021, + "step": 38401 + }, + { + "epoch": 1.64, + "grad_norm": 0.46484375, + "learning_rate": 1.5935362480544735e-05, + "loss": 2.4329, + "step": 38402 + }, + { + "epoch": 1.64, + "grad_norm": 0.484375, + "learning_rate": 1.5931729758544545e-05, + "loss": 2.6261, + "step": 38403 + }, + { + "epoch": 1.64, + "grad_norm": 0.4921875, + "learning_rate": 1.592809741482081e-05, + "loss": 2.3748, + "step": 38404 + }, + { + "epoch": 1.64, + "grad_norm": 0.51171875, + "learning_rate": 1.5924465449389936e-05, + "loss": 2.4221, + "step": 38405 + }, + { + "epoch": 1.64, + "grad_norm": 0.451171875, + "learning_rate": 1.5920833862268226e-05, + "loss": 2.3243, + "step": 38406 + }, + { + "epoch": 1.64, + "grad_norm": 0.4296875, + "learning_rate": 1.591720265347203e-05, + "loss": 2.4228, + "step": 38407 + }, + { + "epoch": 1.64, + "grad_norm": 0.439453125, + "learning_rate": 1.5913571823017727e-05, + "loss": 2.3595, + "step": 38408 + }, + { + "epoch": 1.64, + "grad_norm": 0.44140625, + "learning_rate": 1.590994137092161e-05, + "loss": 2.4308, + "step": 38409 + }, + { + "epoch": 1.64, + "grad_norm": 0.470703125, + "learning_rate": 1.5906311297199972e-05, + "loss": 2.4316, + "step": 38410 + }, + { + "epoch": 1.64, + "grad_norm": 0.5078125, + "learning_rate": 1.5902681601869263e-05, + "loss": 2.4598, + "step": 38411 + }, + { + "epoch": 1.64, + "grad_norm": 0.53515625, + "learning_rate": 1.589905228494575e-05, + "loss": 2.5246, + "step": 38412 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.589542334644574e-05, + "loss": 2.3743, + "step": 38413 + }, + { + "epoch": 1.64, + "grad_norm": 0.48828125, + "learning_rate": 1.5891794786385584e-05, + "loss": 2.4935, + "step": 38414 + }, + { + "epoch": 1.64, + "grad_norm": 0.486328125, + "learning_rate": 1.588816660478164e-05, + "loss": 2.3138, + "step": 38415 + }, + { + "epoch": 1.64, + "grad_norm": 0.486328125, + "learning_rate": 1.5884538801650183e-05, + "loss": 2.5128, + "step": 38416 + }, + { + "epoch": 1.64, + "grad_norm": 0.50390625, + "learning_rate": 1.5880911377007577e-05, + "loss": 2.6832, + "step": 38417 + }, + { + "epoch": 1.64, + "grad_norm": 0.486328125, + "learning_rate": 1.587728433087011e-05, + "loss": 2.4352, + "step": 38418 + }, + { + "epoch": 1.64, + "grad_norm": 0.5, + "learning_rate": 1.587365766325415e-05, + "loss": 2.4698, + "step": 38419 + }, + { + "epoch": 1.64, + "grad_norm": 0.455078125, + "learning_rate": 1.5870031374175965e-05, + "loss": 2.5361, + "step": 38420 + }, + { + "epoch": 1.64, + "grad_norm": 0.4765625, + "learning_rate": 1.58664054636519e-05, + "loss": 2.409, + "step": 38421 + }, + { + "epoch": 1.64, + "grad_norm": 0.4375, + "learning_rate": 1.5862779931698257e-05, + "loss": 2.3234, + "step": 38422 + }, + { + "epoch": 1.64, + "grad_norm": 0.451171875, + "learning_rate": 1.585915477833134e-05, + "loss": 2.4267, + "step": 38423 + }, + { + "epoch": 1.64, + "grad_norm": 0.47265625, + "learning_rate": 1.585553000356751e-05, + "loss": 2.5077, + "step": 38424 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.5851905607423044e-05, + "loss": 2.4131, + "step": 38425 + }, + { + "epoch": 1.64, + "grad_norm": 0.48046875, + "learning_rate": 1.5848281589914206e-05, + "loss": 2.3675, + "step": 38426 + }, + { + "epoch": 1.64, + "grad_norm": 0.404296875, + "learning_rate": 1.5844657951057385e-05, + "loss": 2.6437, + "step": 38427 + }, + { + "epoch": 1.64, + "grad_norm": 0.455078125, + "learning_rate": 1.584103469086885e-05, + "loss": 2.4576, + "step": 38428 + }, + { + "epoch": 1.64, + "grad_norm": 0.515625, + "learning_rate": 1.5837411809364876e-05, + "loss": 2.459, + "step": 38429 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.5833789306561797e-05, + "loss": 2.4672, + "step": 38430 + }, + { + "epoch": 1.64, + "grad_norm": 0.46484375, + "learning_rate": 1.583016718247593e-05, + "loss": 2.4969, + "step": 38431 + }, + { + "epoch": 1.64, + "grad_norm": 0.41015625, + "learning_rate": 1.582654543712353e-05, + "loss": 2.3997, + "step": 38432 + }, + { + "epoch": 1.64, + "grad_norm": 0.384765625, + "learning_rate": 1.5822924070520927e-05, + "loss": 2.3067, + "step": 38433 + }, + { + "epoch": 1.64, + "grad_norm": 0.46484375, + "learning_rate": 1.581930308268439e-05, + "loss": 2.5322, + "step": 38434 + }, + { + "epoch": 1.64, + "grad_norm": 0.412109375, + "learning_rate": 1.5815682473630232e-05, + "loss": 2.443, + "step": 38435 + }, + { + "epoch": 1.64, + "grad_norm": 0.44140625, + "learning_rate": 1.5812062243374724e-05, + "loss": 2.427, + "step": 38436 + }, + { + "epoch": 1.64, + "grad_norm": 0.453125, + "learning_rate": 1.5808442391934197e-05, + "loss": 2.4132, + "step": 38437 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.580482291932488e-05, + "loss": 2.4086, + "step": 38438 + }, + { + "epoch": 1.64, + "grad_norm": 0.5, + "learning_rate": 1.5801203825563093e-05, + "loss": 2.4366, + "step": 38439 + }, + { + "epoch": 1.64, + "grad_norm": 0.52734375, + "learning_rate": 1.5797585110665148e-05, + "loss": 2.478, + "step": 38440 + }, + { + "epoch": 1.64, + "grad_norm": 0.41015625, + "learning_rate": 1.5793966774647285e-05, + "loss": 2.3475, + "step": 38441 + }, + { + "epoch": 1.64, + "grad_norm": 0.5390625, + "learning_rate": 1.579034881752578e-05, + "loss": 2.645, + "step": 38442 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.5786731239316932e-05, + "loss": 2.5013, + "step": 38443 + }, + { + "epoch": 1.64, + "grad_norm": 0.451171875, + "learning_rate": 1.5783114040037038e-05, + "loss": 2.2763, + "step": 38444 + }, + { + "epoch": 1.64, + "grad_norm": 0.49609375, + "learning_rate": 1.5779497219702334e-05, + "loss": 2.4854, + "step": 38445 + }, + { + "epoch": 1.64, + "grad_norm": 0.486328125, + "learning_rate": 1.5775880778329124e-05, + "loss": 2.4119, + "step": 38446 + }, + { + "epoch": 1.64, + "grad_norm": 0.412109375, + "learning_rate": 1.577226471593366e-05, + "loss": 2.3307, + "step": 38447 + }, + { + "epoch": 1.64, + "grad_norm": 0.44140625, + "learning_rate": 1.5768649032532212e-05, + "loss": 2.4887, + "step": 38448 + }, + { + "epoch": 1.64, + "grad_norm": 0.41015625, + "learning_rate": 1.5765033728141098e-05, + "loss": 2.5623, + "step": 38449 + }, + { + "epoch": 1.64, + "grad_norm": 0.439453125, + "learning_rate": 1.576141880277654e-05, + "loss": 2.4497, + "step": 38450 + }, + { + "epoch": 1.64, + "grad_norm": 0.46875, + "learning_rate": 1.5757804256454788e-05, + "loss": 2.3995, + "step": 38451 + }, + { + "epoch": 1.64, + "grad_norm": 0.48828125, + "learning_rate": 1.5754190089192124e-05, + "loss": 2.4823, + "step": 38452 + }, + { + "epoch": 1.64, + "grad_norm": 0.486328125, + "learning_rate": 1.575057630100485e-05, + "loss": 2.5249, + "step": 38453 + }, + { + "epoch": 1.64, + "grad_norm": 0.498046875, + "learning_rate": 1.5746962891909166e-05, + "loss": 2.363, + "step": 38454 + }, + { + "epoch": 1.64, + "grad_norm": 0.44921875, + "learning_rate": 1.5743349861921353e-05, + "loss": 2.4359, + "step": 38455 + }, + { + "epoch": 1.64, + "grad_norm": 0.44921875, + "learning_rate": 1.5739737211057702e-05, + "loss": 2.349, + "step": 38456 + }, + { + "epoch": 1.64, + "grad_norm": 0.482421875, + "learning_rate": 1.573612493933444e-05, + "loss": 2.5529, + "step": 38457 + }, + { + "epoch": 1.64, + "grad_norm": 0.447265625, + "learning_rate": 1.5732513046767794e-05, + "loss": 2.381, + "step": 38458 + }, + { + "epoch": 1.64, + "grad_norm": 0.53125, + "learning_rate": 1.5728901533374052e-05, + "loss": 2.5165, + "step": 38459 + }, + { + "epoch": 1.64, + "grad_norm": 0.4140625, + "learning_rate": 1.5725290399169466e-05, + "loss": 2.3917, + "step": 38460 + }, + { + "epoch": 1.64, + "grad_norm": 0.490234375, + "learning_rate": 1.5721679644170262e-05, + "loss": 2.4646, + "step": 38461 + }, + { + "epoch": 1.64, + "grad_norm": 0.4765625, + "learning_rate": 1.571806926839272e-05, + "loss": 2.3861, + "step": 38462 + }, + { + "epoch": 1.64, + "grad_norm": 0.47265625, + "learning_rate": 1.5714459271853034e-05, + "loss": 2.4816, + "step": 38463 + }, + { + "epoch": 1.64, + "grad_norm": 0.439453125, + "learning_rate": 1.5710849654567484e-05, + "loss": 2.4033, + "step": 38464 + }, + { + "epoch": 1.64, + "grad_norm": 0.44921875, + "learning_rate": 1.570724041655233e-05, + "loss": 2.4002, + "step": 38465 + }, + { + "epoch": 1.64, + "grad_norm": 0.439453125, + "learning_rate": 1.5703631557823794e-05, + "loss": 2.4087, + "step": 38466 + }, + { + "epoch": 1.64, + "grad_norm": 0.46484375, + "learning_rate": 1.5700023078398075e-05, + "loss": 2.4013, + "step": 38467 + }, + { + "epoch": 1.64, + "grad_norm": 0.44921875, + "learning_rate": 1.5696414978291452e-05, + "loss": 2.3707, + "step": 38468 + }, + { + "epoch": 1.64, + "grad_norm": 0.4453125, + "learning_rate": 1.569280725752017e-05, + "loss": 2.4383, + "step": 38469 + }, + { + "epoch": 1.64, + "grad_norm": 0.470703125, + "learning_rate": 1.5689199916100427e-05, + "loss": 2.4564, + "step": 38470 + }, + { + "epoch": 1.64, + "grad_norm": 0.482421875, + "learning_rate": 1.5685592954048468e-05, + "loss": 2.4342, + "step": 38471 + }, + { + "epoch": 1.64, + "grad_norm": 0.53515625, + "learning_rate": 1.5681986371380554e-05, + "loss": 2.4742, + "step": 38472 + }, + { + "epoch": 1.64, + "grad_norm": 0.474609375, + "learning_rate": 1.5678380168112883e-05, + "loss": 2.4519, + "step": 38473 + }, + { + "epoch": 1.64, + "grad_norm": 0.482421875, + "learning_rate": 1.5674774344261667e-05, + "loss": 2.4574, + "step": 38474 + }, + { + "epoch": 1.64, + "grad_norm": 0.490234375, + "learning_rate": 1.5671168899843137e-05, + "loss": 2.4042, + "step": 38475 + }, + { + "epoch": 1.64, + "grad_norm": 0.4296875, + "learning_rate": 1.5667563834873567e-05, + "loss": 2.3933, + "step": 38476 + }, + { + "epoch": 1.64, + "grad_norm": 0.4921875, + "learning_rate": 1.5663959149369113e-05, + "loss": 2.5655, + "step": 38477 + }, + { + "epoch": 1.64, + "grad_norm": 0.43359375, + "learning_rate": 1.5660354843346037e-05, + "loss": 2.4424, + "step": 38478 + }, + { + "epoch": 1.64, + "grad_norm": 0.546875, + "learning_rate": 1.5656750916820516e-05, + "loss": 2.565, + "step": 38479 + }, + { + "epoch": 1.64, + "grad_norm": 0.5, + "learning_rate": 1.5653147369808808e-05, + "loss": 2.4632, + "step": 38480 + }, + { + "epoch": 1.64, + "grad_norm": 0.447265625, + "learning_rate": 1.5649544202327128e-05, + "loss": 2.4877, + "step": 38481 + }, + { + "epoch": 1.64, + "grad_norm": 0.44921875, + "learning_rate": 1.5645941414391663e-05, + "loss": 2.4683, + "step": 38482 + }, + { + "epoch": 1.64, + "grad_norm": 0.43359375, + "learning_rate": 1.5642339006018614e-05, + "loss": 2.4273, + "step": 38483 + }, + { + "epoch": 1.64, + "grad_norm": 0.50390625, + "learning_rate": 1.5638736977224224e-05, + "loss": 2.5976, + "step": 38484 + }, + { + "epoch": 1.64, + "grad_norm": 0.443359375, + "learning_rate": 1.5635135328024698e-05, + "loss": 2.3476, + "step": 38485 + }, + { + "epoch": 1.64, + "grad_norm": 0.421875, + "learning_rate": 1.5631534058436204e-05, + "loss": 2.3876, + "step": 38486 + }, + { + "epoch": 1.64, + "grad_norm": 0.453125, + "learning_rate": 1.5627933168474986e-05, + "loss": 2.5197, + "step": 38487 + }, + { + "epoch": 1.64, + "grad_norm": 0.46484375, + "learning_rate": 1.5624332658157248e-05, + "loss": 2.3385, + "step": 38488 + }, + { + "epoch": 1.64, + "grad_norm": 0.458984375, + "learning_rate": 1.5620732527499182e-05, + "loss": 2.4701, + "step": 38489 + }, + { + "epoch": 1.64, + "grad_norm": 0.421875, + "learning_rate": 1.561713277651695e-05, + "loss": 2.4166, + "step": 38490 + }, + { + "epoch": 1.64, + "grad_norm": 0.4765625, + "learning_rate": 1.56135334052268e-05, + "loss": 2.4513, + "step": 38491 + }, + { + "epoch": 1.64, + "grad_norm": 0.5, + "learning_rate": 1.5609934413644922e-05, + "loss": 2.4791, + "step": 38492 + }, + { + "epoch": 1.64, + "grad_norm": 0.470703125, + "learning_rate": 1.5606335801787475e-05, + "loss": 2.2996, + "step": 38493 + }, + { + "epoch": 1.64, + "grad_norm": 0.50390625, + "learning_rate": 1.56027375696707e-05, + "loss": 2.4786, + "step": 38494 + }, + { + "epoch": 1.64, + "grad_norm": 0.455078125, + "learning_rate": 1.559913971731074e-05, + "loss": 2.3789, + "step": 38495 + }, + { + "epoch": 1.64, + "grad_norm": 0.53515625, + "learning_rate": 1.5595542244723792e-05, + "loss": 2.4703, + "step": 38496 + }, + { + "epoch": 1.64, + "grad_norm": 0.4375, + "learning_rate": 1.55919451519261e-05, + "loss": 2.4312, + "step": 38497 + }, + { + "epoch": 1.64, + "grad_norm": 0.52734375, + "learning_rate": 1.5588348438933797e-05, + "loss": 2.3776, + "step": 38498 + }, + { + "epoch": 1.64, + "grad_norm": 0.447265625, + "learning_rate": 1.5584752105763057e-05, + "loss": 2.4074, + "step": 38499 + }, + { + "epoch": 1.64, + "grad_norm": 0.41015625, + "learning_rate": 1.5581156152430086e-05, + "loss": 2.5301, + "step": 38500 + }, + { + "epoch": 1.64, + "grad_norm": 0.453125, + "learning_rate": 1.5577560578951077e-05, + "loss": 2.5273, + "step": 38501 + }, + { + "epoch": 1.64, + "grad_norm": 0.482421875, + "learning_rate": 1.557396538534217e-05, + "loss": 2.4156, + "step": 38502 + }, + { + "epoch": 1.64, + "grad_norm": 0.5390625, + "learning_rate": 1.5570370571619575e-05, + "loss": 2.427, + "step": 38503 + }, + { + "epoch": 1.64, + "grad_norm": 0.458984375, + "learning_rate": 1.5566776137799466e-05, + "loss": 2.5285, + "step": 38504 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.556318208389802e-05, + "loss": 2.3954, + "step": 38505 + }, + { + "epoch": 1.64, + "grad_norm": 0.447265625, + "learning_rate": 1.555958840993137e-05, + "loss": 2.4849, + "step": 38506 + }, + { + "epoch": 1.64, + "grad_norm": 0.455078125, + "learning_rate": 1.55559951159157e-05, + "loss": 2.3843, + "step": 38507 + }, + { + "epoch": 1.64, + "grad_norm": 0.455078125, + "learning_rate": 1.5552402201867232e-05, + "loss": 2.403, + "step": 38508 + }, + { + "epoch": 1.64, + "grad_norm": 0.54296875, + "learning_rate": 1.5548809667802057e-05, + "loss": 2.4722, + "step": 38509 + }, + { + "epoch": 1.64, + "grad_norm": 0.416015625, + "learning_rate": 1.5545217513736397e-05, + "loss": 2.4271, + "step": 38510 + }, + { + "epoch": 1.64, + "grad_norm": 0.43359375, + "learning_rate": 1.5541625739686384e-05, + "loss": 2.3616, + "step": 38511 + }, + { + "epoch": 1.64, + "grad_norm": 0.453125, + "learning_rate": 1.5538034345668175e-05, + "loss": 2.4602, + "step": 38512 + }, + { + "epoch": 1.64, + "grad_norm": 0.50390625, + "learning_rate": 1.553444333169798e-05, + "loss": 2.4051, + "step": 38513 + }, + { + "epoch": 1.64, + "grad_norm": 0.48828125, + "learning_rate": 1.5530852697791908e-05, + "loss": 2.4282, + "step": 38514 + }, + { + "epoch": 1.64, + "grad_norm": 0.46875, + "learning_rate": 1.5527262443966118e-05, + "loss": 2.4775, + "step": 38515 + }, + { + "epoch": 1.64, + "grad_norm": 0.48828125, + "learning_rate": 1.5523672570236768e-05, + "loss": 2.4293, + "step": 38516 + }, + { + "epoch": 1.64, + "grad_norm": 0.48828125, + "learning_rate": 1.552008307662004e-05, + "loss": 2.4569, + "step": 38517 + }, + { + "epoch": 1.64, + "grad_norm": 0.400390625, + "learning_rate": 1.551649396313205e-05, + "loss": 2.4049, + "step": 38518 + }, + { + "epoch": 1.64, + "grad_norm": 0.462890625, + "learning_rate": 1.5512905229788955e-05, + "loss": 2.4887, + "step": 38519 + }, + { + "epoch": 1.64, + "grad_norm": 0.443359375, + "learning_rate": 1.5509316876606938e-05, + "loss": 2.5639, + "step": 38520 + }, + { + "epoch": 1.64, + "grad_norm": 0.51953125, + "learning_rate": 1.5505728903602113e-05, + "loss": 2.5069, + "step": 38521 + }, + { + "epoch": 1.64, + "grad_norm": 0.455078125, + "learning_rate": 1.550214131079061e-05, + "loss": 2.3934, + "step": 38522 + }, + { + "epoch": 1.64, + "grad_norm": 0.412109375, + "learning_rate": 1.549855409818859e-05, + "loss": 2.3277, + "step": 38523 + }, + { + "epoch": 1.64, + "grad_norm": 0.5078125, + "learning_rate": 1.549496726581221e-05, + "loss": 2.4793, + "step": 38524 + }, + { + "epoch": 1.64, + "grad_norm": 0.490234375, + "learning_rate": 1.549138081367758e-05, + "loss": 2.4057, + "step": 38525 + }, + { + "epoch": 1.64, + "grad_norm": 0.4453125, + "learning_rate": 1.5487794741800865e-05, + "loss": 2.2935, + "step": 38526 + }, + { + "epoch": 1.64, + "grad_norm": 0.421875, + "learning_rate": 1.5484209050198172e-05, + "loss": 2.4871, + "step": 38527 + }, + { + "epoch": 1.64, + "grad_norm": 0.490234375, + "learning_rate": 1.548062373888566e-05, + "loss": 2.3121, + "step": 38528 + }, + { + "epoch": 1.64, + "grad_norm": 0.4375, + "learning_rate": 1.5477038807879464e-05, + "loss": 2.4447, + "step": 38529 + }, + { + "epoch": 1.64, + "grad_norm": 0.50390625, + "learning_rate": 1.5473454257195708e-05, + "loss": 2.3564, + "step": 38530 + }, + { + "epoch": 1.64, + "grad_norm": 0.478515625, + "learning_rate": 1.546987008685049e-05, + "loss": 2.5422, + "step": 38531 + }, + { + "epoch": 1.64, + "grad_norm": 0.435546875, + "learning_rate": 1.5466286296859965e-05, + "loss": 2.3821, + "step": 38532 + }, + { + "epoch": 1.64, + "grad_norm": 0.53515625, + "learning_rate": 1.546270288724029e-05, + "loss": 2.6379, + "step": 38533 + }, + { + "epoch": 1.64, + "grad_norm": 0.4375, + "learning_rate": 1.545911985800752e-05, + "loss": 2.5866, + "step": 38534 + }, + { + "epoch": 1.64, + "grad_norm": 0.47265625, + "learning_rate": 1.5455537209177828e-05, + "loss": 2.4641, + "step": 38535 + }, + { + "epoch": 1.64, + "grad_norm": 0.478515625, + "learning_rate": 1.5451954940767332e-05, + "loss": 2.4446, + "step": 38536 + }, + { + "epoch": 1.64, + "grad_norm": 0.50390625, + "learning_rate": 1.5448373052792154e-05, + "loss": 2.4831, + "step": 38537 + }, + { + "epoch": 1.64, + "grad_norm": 0.421875, + "learning_rate": 1.5444791545268356e-05, + "loss": 2.5032, + "step": 38538 + }, + { + "epoch": 1.64, + "grad_norm": 0.47265625, + "learning_rate": 1.5441210418212104e-05, + "loss": 2.5338, + "step": 38539 + }, + { + "epoch": 1.64, + "grad_norm": 0.4453125, + "learning_rate": 1.543762967163953e-05, + "loss": 2.394, + "step": 38540 + }, + { + "epoch": 1.64, + "grad_norm": 0.498046875, + "learning_rate": 1.5434049305566687e-05, + "loss": 2.5044, + "step": 38541 + }, + { + "epoch": 1.64, + "grad_norm": 0.5078125, + "learning_rate": 1.5430469320009755e-05, + "loss": 2.4812, + "step": 38542 + }, + { + "epoch": 1.64, + "grad_norm": 0.443359375, + "learning_rate": 1.5426889714984773e-05, + "loss": 2.45, + "step": 38543 + }, + { + "epoch": 1.64, + "grad_norm": 0.478515625, + "learning_rate": 1.5423310490507868e-05, + "loss": 2.4094, + "step": 38544 + }, + { + "epoch": 1.64, + "grad_norm": 0.400390625, + "learning_rate": 1.5419731646595193e-05, + "loss": 2.4736, + "step": 38545 + }, + { + "epoch": 1.64, + "grad_norm": 0.482421875, + "learning_rate": 1.5416153183262817e-05, + "loss": 2.396, + "step": 38546 + }, + { + "epoch": 1.64, + "grad_norm": 0.498046875, + "learning_rate": 1.5412575100526806e-05, + "loss": 2.3035, + "step": 38547 + }, + { + "epoch": 1.64, + "grad_norm": 0.494140625, + "learning_rate": 1.540899739840329e-05, + "loss": 2.4939, + "step": 38548 + }, + { + "epoch": 1.64, + "grad_norm": 0.412109375, + "learning_rate": 1.5405420076908404e-05, + "loss": 2.4067, + "step": 38549 + }, + { + "epoch": 1.64, + "grad_norm": 0.48828125, + "learning_rate": 1.540184313605819e-05, + "loss": 2.3837, + "step": 38550 + }, + { + "epoch": 1.64, + "grad_norm": 0.5703125, + "learning_rate": 1.5398266575868768e-05, + "loss": 2.4315, + "step": 38551 + }, + { + "epoch": 1.64, + "grad_norm": 0.44921875, + "learning_rate": 1.539469039635625e-05, + "loss": 2.4686, + "step": 38552 + }, + { + "epoch": 1.64, + "grad_norm": 0.48046875, + "learning_rate": 1.5391114597536693e-05, + "loss": 2.3673, + "step": 38553 + }, + { + "epoch": 1.64, + "grad_norm": 0.458984375, + "learning_rate": 1.538753917942618e-05, + "loss": 2.4988, + "step": 38554 + }, + { + "epoch": 1.64, + "grad_norm": 0.49609375, + "learning_rate": 1.538396414204083e-05, + "loss": 2.5184, + "step": 38555 + }, + { + "epoch": 1.64, + "grad_norm": 0.470703125, + "learning_rate": 1.538038948539673e-05, + "loss": 2.5449, + "step": 38556 + }, + { + "epoch": 1.64, + "grad_norm": 0.490234375, + "learning_rate": 1.5376815209509933e-05, + "loss": 2.5638, + "step": 38557 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.537324131439657e-05, + "loss": 2.4747, + "step": 38558 + }, + { + "epoch": 1.64, + "grad_norm": 0.462890625, + "learning_rate": 1.5369667800072664e-05, + "loss": 2.6417, + "step": 38559 + }, + { + "epoch": 1.64, + "grad_norm": 0.4453125, + "learning_rate": 1.536609466655432e-05, + "loss": 2.4063, + "step": 38560 + }, + { + "epoch": 1.64, + "grad_norm": 0.451171875, + "learning_rate": 1.536252191385765e-05, + "loss": 2.4933, + "step": 38561 + }, + { + "epoch": 1.64, + "grad_norm": 0.4140625, + "learning_rate": 1.53589495419987e-05, + "loss": 2.5198, + "step": 38562 + }, + { + "epoch": 1.64, + "grad_norm": 0.484375, + "learning_rate": 1.535537755099351e-05, + "loss": 2.5978, + "step": 38563 + }, + { + "epoch": 1.64, + "grad_norm": 0.466796875, + "learning_rate": 1.53518059408582e-05, + "loss": 2.5123, + "step": 38564 + }, + { + "epoch": 1.64, + "grad_norm": 0.453125, + "learning_rate": 1.5348234711608845e-05, + "loss": 2.4283, + "step": 38565 + }, + { + "epoch": 1.64, + "grad_norm": 0.48828125, + "learning_rate": 1.5344663863261476e-05, + "loss": 2.4025, + "step": 38566 + }, + { + "epoch": 1.64, + "grad_norm": 0.427734375, + "learning_rate": 1.5341093395832207e-05, + "loss": 2.4476, + "step": 38567 + }, + { + "epoch": 1.64, + "grad_norm": 0.453125, + "learning_rate": 1.5337523309337055e-05, + "loss": 2.3453, + "step": 38568 + }, + { + "epoch": 1.64, + "grad_norm": 0.421875, + "learning_rate": 1.533395360379213e-05, + "loss": 2.2406, + "step": 38569 + }, + { + "epoch": 1.64, + "grad_norm": 0.515625, + "learning_rate": 1.5330384279213438e-05, + "loss": 2.5071, + "step": 38570 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.5326815335617105e-05, + "loss": 2.5238, + "step": 38571 + }, + { + "epoch": 1.64, + "grad_norm": 0.455078125, + "learning_rate": 1.532324677301914e-05, + "loss": 2.4504, + "step": 38572 + }, + { + "epoch": 1.64, + "grad_norm": 0.447265625, + "learning_rate": 1.5319678591435617e-05, + "loss": 2.4651, + "step": 38573 + }, + { + "epoch": 1.64, + "grad_norm": 0.484375, + "learning_rate": 1.531611079088261e-05, + "loss": 2.3235, + "step": 38574 + }, + { + "epoch": 1.64, + "grad_norm": 0.423828125, + "learning_rate": 1.5312543371376165e-05, + "loss": 2.4535, + "step": 38575 + }, + { + "epoch": 1.64, + "grad_norm": 0.52734375, + "learning_rate": 1.5308976332932266e-05, + "loss": 2.3969, + "step": 38576 + }, + { + "epoch": 1.64, + "grad_norm": 0.498046875, + "learning_rate": 1.5305409675567084e-05, + "loss": 2.323, + "step": 38577 + }, + { + "epoch": 1.64, + "grad_norm": 0.54296875, + "learning_rate": 1.53018433992966e-05, + "loss": 2.4467, + "step": 38578 + }, + { + "epoch": 1.64, + "grad_norm": 0.4375, + "learning_rate": 1.529827750413684e-05, + "loss": 2.4874, + "step": 38579 + }, + { + "epoch": 1.64, + "grad_norm": 0.478515625, + "learning_rate": 1.5294711990103884e-05, + "loss": 2.495, + "step": 38580 + }, + { + "epoch": 1.64, + "grad_norm": 0.5234375, + "learning_rate": 1.5291146857213778e-05, + "loss": 2.4444, + "step": 38581 + }, + { + "epoch": 1.64, + "grad_norm": 0.53515625, + "learning_rate": 1.5287582105482544e-05, + "loss": 2.5113, + "step": 38582 + }, + { + "epoch": 1.64, + "grad_norm": 0.484375, + "learning_rate": 1.5284017734926247e-05, + "loss": 2.3997, + "step": 38583 + }, + { + "epoch": 1.64, + "grad_norm": 0.515625, + "learning_rate": 1.5280453745560896e-05, + "loss": 2.5212, + "step": 38584 + }, + { + "epoch": 1.64, + "grad_norm": 0.435546875, + "learning_rate": 1.5276890137402556e-05, + "loss": 2.4769, + "step": 38585 + }, + { + "epoch": 1.64, + "grad_norm": 0.431640625, + "learning_rate": 1.5273326910467224e-05, + "loss": 2.3515, + "step": 38586 + }, + { + "epoch": 1.64, + "grad_norm": 0.5, + "learning_rate": 1.5269764064770985e-05, + "loss": 2.377, + "step": 38587 + }, + { + "epoch": 1.64, + "grad_norm": 0.439453125, + "learning_rate": 1.526620160032981e-05, + "loss": 2.3658, + "step": 38588 + }, + { + "epoch": 1.64, + "grad_norm": 0.5078125, + "learning_rate": 1.5262639517159773e-05, + "loss": 2.3434, + "step": 38589 + }, + { + "epoch": 1.64, + "grad_norm": 0.4375, + "learning_rate": 1.525907781527691e-05, + "loss": 2.3659, + "step": 38590 + }, + { + "epoch": 1.64, + "grad_norm": 0.4609375, + "learning_rate": 1.525551649469722e-05, + "loss": 2.4714, + "step": 38591 + }, + { + "epoch": 1.64, + "grad_norm": 0.4921875, + "learning_rate": 1.5251955555436703e-05, + "loss": 2.4958, + "step": 38592 + }, + { + "epoch": 1.64, + "grad_norm": 0.466796875, + "learning_rate": 1.524839499751145e-05, + "loss": 2.5, + "step": 38593 + }, + { + "epoch": 1.64, + "grad_norm": 0.40625, + "learning_rate": 1.5244834820937448e-05, + "loss": 2.4784, + "step": 38594 + }, + { + "epoch": 1.64, + "grad_norm": 0.5, + "learning_rate": 1.5241275025730695e-05, + "loss": 2.4357, + "step": 38595 + }, + { + "epoch": 1.64, + "grad_norm": 0.4921875, + "learning_rate": 1.5237715611907222e-05, + "loss": 2.3743, + "step": 38596 + }, + { + "epoch": 1.64, + "grad_norm": 0.3984375, + "learning_rate": 1.5234156579483083e-05, + "loss": 2.488, + "step": 38597 + }, + { + "epoch": 1.64, + "grad_norm": 0.41796875, + "learning_rate": 1.5230597928474233e-05, + "loss": 2.4249, + "step": 38598 + }, + { + "epoch": 1.64, + "grad_norm": 0.392578125, + "learning_rate": 1.5227039658896735e-05, + "loss": 2.4013, + "step": 38599 + }, + { + "epoch": 1.64, + "grad_norm": 0.46875, + "learning_rate": 1.5223481770766557e-05, + "loss": 2.4141, + "step": 38600 + }, + { + "epoch": 1.64, + "grad_norm": 0.4765625, + "learning_rate": 1.521992426409975e-05, + "loss": 2.4375, + "step": 38601 + }, + { + "epoch": 1.64, + "grad_norm": 0.4765625, + "learning_rate": 1.5216367138912269e-05, + "loss": 2.3416, + "step": 38602 + }, + { + "epoch": 1.64, + "grad_norm": 0.46875, + "learning_rate": 1.521281039522019e-05, + "loss": 2.5081, + "step": 38603 + }, + { + "epoch": 1.65, + "grad_norm": 0.46484375, + "learning_rate": 1.5209254033039444e-05, + "loss": 2.3345, + "step": 38604 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.5205698052386074e-05, + "loss": 2.424, + "step": 38605 + }, + { + "epoch": 1.65, + "grad_norm": 0.45703125, + "learning_rate": 1.5202142453276091e-05, + "loss": 2.4562, + "step": 38606 + }, + { + "epoch": 1.65, + "grad_norm": 0.4609375, + "learning_rate": 1.5198587235725481e-05, + "loss": 2.5597, + "step": 38607 + }, + { + "epoch": 1.65, + "grad_norm": 0.490234375, + "learning_rate": 1.5195032399750186e-05, + "loss": 2.4839, + "step": 38608 + }, + { + "epoch": 1.65, + "grad_norm": 0.447265625, + "learning_rate": 1.5191477945366305e-05, + "loss": 2.32, + "step": 38609 + }, + { + "epoch": 1.65, + "grad_norm": 0.4609375, + "learning_rate": 1.5187923872589782e-05, + "loss": 2.4382, + "step": 38610 + }, + { + "epoch": 1.65, + "grad_norm": 0.4609375, + "learning_rate": 1.518437018143658e-05, + "loss": 2.484, + "step": 38611 + }, + { + "epoch": 1.65, + "grad_norm": 0.47265625, + "learning_rate": 1.518081687192271e-05, + "loss": 2.397, + "step": 38612 + }, + { + "epoch": 1.65, + "grad_norm": 0.416015625, + "learning_rate": 1.517726394406419e-05, + "loss": 2.4027, + "step": 38613 + }, + { + "epoch": 1.65, + "grad_norm": 0.494140625, + "learning_rate": 1.5173711397876966e-05, + "loss": 2.323, + "step": 38614 + }, + { + "epoch": 1.65, + "grad_norm": 0.42578125, + "learning_rate": 1.5170159233377057e-05, + "loss": 2.2892, + "step": 38615 + }, + { + "epoch": 1.65, + "grad_norm": 0.478515625, + "learning_rate": 1.5166607450580417e-05, + "loss": 2.579, + "step": 38616 + }, + { + "epoch": 1.65, + "grad_norm": 0.451171875, + "learning_rate": 1.5163056049503055e-05, + "loss": 2.4725, + "step": 38617 + }, + { + "epoch": 1.65, + "grad_norm": 0.53515625, + "learning_rate": 1.5159505030160914e-05, + "loss": 2.5505, + "step": 38618 + }, + { + "epoch": 1.65, + "grad_norm": 0.49609375, + "learning_rate": 1.5155954392570016e-05, + "loss": 2.5042, + "step": 38619 + }, + { + "epoch": 1.65, + "grad_norm": 0.45703125, + "learning_rate": 1.5152404136746302e-05, + "loss": 2.3457, + "step": 38620 + }, + { + "epoch": 1.65, + "grad_norm": 0.447265625, + "learning_rate": 1.5148854262705748e-05, + "loss": 2.5006, + "step": 38621 + }, + { + "epoch": 1.65, + "grad_norm": 0.51953125, + "learning_rate": 1.5145304770464375e-05, + "loss": 2.5139, + "step": 38622 + }, + { + "epoch": 1.65, + "grad_norm": 0.453125, + "learning_rate": 1.5141755660038115e-05, + "loss": 2.4657, + "step": 38623 + }, + { + "epoch": 1.65, + "grad_norm": 0.50390625, + "learning_rate": 1.5138206931442899e-05, + "loss": 2.514, + "step": 38624 + }, + { + "epoch": 1.65, + "grad_norm": 0.48046875, + "learning_rate": 1.5134658584694783e-05, + "loss": 2.5318, + "step": 38625 + }, + { + "epoch": 1.65, + "grad_norm": 0.486328125, + "learning_rate": 1.5131110619809685e-05, + "loss": 2.4565, + "step": 38626 + }, + { + "epoch": 1.65, + "grad_norm": 0.48046875, + "learning_rate": 1.5127563036803549e-05, + "loss": 2.5076, + "step": 38627 + }, + { + "epoch": 1.65, + "grad_norm": 0.53515625, + "learning_rate": 1.5124015835692363e-05, + "loss": 2.4849, + "step": 38628 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.5120469016492123e-05, + "loss": 2.547, + "step": 38629 + }, + { + "epoch": 1.65, + "grad_norm": 0.416015625, + "learning_rate": 1.5116922579218718e-05, + "loss": 2.4093, + "step": 38630 + }, + { + "epoch": 1.65, + "grad_norm": 0.5234375, + "learning_rate": 1.5113376523888156e-05, + "loss": 2.5075, + "step": 38631 + }, + { + "epoch": 1.65, + "grad_norm": 0.5546875, + "learning_rate": 1.5109830850516371e-05, + "loss": 2.526, + "step": 38632 + }, + { + "epoch": 1.65, + "grad_norm": 0.55078125, + "learning_rate": 1.5106285559119337e-05, + "loss": 2.5508, + "step": 38633 + }, + { + "epoch": 1.65, + "grad_norm": 0.5078125, + "learning_rate": 1.5102740649712966e-05, + "loss": 2.4411, + "step": 38634 + }, + { + "epoch": 1.65, + "grad_norm": 0.53515625, + "learning_rate": 1.5099196122313264e-05, + "loss": 2.423, + "step": 38635 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.5095651976936132e-05, + "loss": 2.5566, + "step": 38636 + }, + { + "epoch": 1.65, + "grad_norm": 0.3984375, + "learning_rate": 1.5092108213597533e-05, + "loss": 2.4172, + "step": 38637 + }, + { + "epoch": 1.65, + "grad_norm": 0.470703125, + "learning_rate": 1.5088564832313434e-05, + "loss": 2.4432, + "step": 38638 + }, + { + "epoch": 1.65, + "grad_norm": 0.3984375, + "learning_rate": 1.5085021833099766e-05, + "loss": 2.3744, + "step": 38639 + }, + { + "epoch": 1.65, + "grad_norm": 0.447265625, + "learning_rate": 1.5081479215972428e-05, + "loss": 2.5245, + "step": 38640 + }, + { + "epoch": 1.65, + "grad_norm": 0.4375, + "learning_rate": 1.5077936980947439e-05, + "loss": 2.3501, + "step": 38641 + }, + { + "epoch": 1.65, + "grad_norm": 0.423828125, + "learning_rate": 1.5074395128040698e-05, + "loss": 2.462, + "step": 38642 + }, + { + "epoch": 1.65, + "grad_norm": 0.48828125, + "learning_rate": 1.5070853657268124e-05, + "loss": 2.4694, + "step": 38643 + }, + { + "epoch": 1.65, + "grad_norm": 0.51953125, + "learning_rate": 1.5067312568645664e-05, + "loss": 2.4306, + "step": 38644 + }, + { + "epoch": 1.65, + "grad_norm": 0.431640625, + "learning_rate": 1.5063771862189279e-05, + "loss": 2.452, + "step": 38645 + }, + { + "epoch": 1.65, + "grad_norm": 0.46484375, + "learning_rate": 1.506023153791487e-05, + "loss": 2.5199, + "step": 38646 + }, + { + "epoch": 1.65, + "grad_norm": 0.45703125, + "learning_rate": 1.5056691595838391e-05, + "loss": 2.3781, + "step": 38647 + }, + { + "epoch": 1.65, + "grad_norm": 0.4765625, + "learning_rate": 1.5053152035975738e-05, + "loss": 2.6862, + "step": 38648 + }, + { + "epoch": 1.65, + "grad_norm": 0.48828125, + "learning_rate": 1.5049612858342877e-05, + "loss": 2.4707, + "step": 38649 + }, + { + "epoch": 1.65, + "grad_norm": 0.515625, + "learning_rate": 1.5046074062955684e-05, + "loss": 2.5136, + "step": 38650 + }, + { + "epoch": 1.65, + "grad_norm": 0.45703125, + "learning_rate": 1.5042535649830136e-05, + "loss": 2.4968, + "step": 38651 + }, + { + "epoch": 1.65, + "grad_norm": 0.427734375, + "learning_rate": 1.503899761898211e-05, + "loss": 2.4741, + "step": 38652 + }, + { + "epoch": 1.65, + "grad_norm": 0.482421875, + "learning_rate": 1.5035459970427546e-05, + "loss": 2.5366, + "step": 38653 + }, + { + "epoch": 1.65, + "grad_norm": 0.44140625, + "learning_rate": 1.5031922704182377e-05, + "loss": 2.4478, + "step": 38654 + }, + { + "epoch": 1.65, + "grad_norm": 0.453125, + "learning_rate": 1.5028385820262514e-05, + "loss": 2.4743, + "step": 38655 + }, + { + "epoch": 1.65, + "grad_norm": 0.498046875, + "learning_rate": 1.5024849318683798e-05, + "loss": 2.4167, + "step": 38656 + }, + { + "epoch": 1.65, + "grad_norm": 0.48828125, + "learning_rate": 1.5021313199462273e-05, + "loss": 2.488, + "step": 38657 + }, + { + "epoch": 1.65, + "grad_norm": 0.458984375, + "learning_rate": 1.5017777462613769e-05, + "loss": 2.553, + "step": 38658 + }, + { + "epoch": 1.65, + "grad_norm": 0.52734375, + "learning_rate": 1.5014242108154185e-05, + "loss": 2.5801, + "step": 38659 + }, + { + "epoch": 1.65, + "grad_norm": 0.4453125, + "learning_rate": 1.5010707136099456e-05, + "loss": 2.4669, + "step": 38660 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.500717254646551e-05, + "loss": 2.3886, + "step": 38661 + }, + { + "epoch": 1.65, + "grad_norm": 0.54296875, + "learning_rate": 1.50036383392682e-05, + "loss": 2.5992, + "step": 38662 + }, + { + "epoch": 1.65, + "grad_norm": 0.404296875, + "learning_rate": 1.5000104514523482e-05, + "loss": 2.5235, + "step": 38663 + }, + { + "epoch": 1.65, + "grad_norm": 0.44921875, + "learning_rate": 1.4996571072247201e-05, + "loss": 2.3473, + "step": 38664 + }, + { + "epoch": 1.65, + "grad_norm": 0.4609375, + "learning_rate": 1.4993038012455307e-05, + "loss": 2.2431, + "step": 38665 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.4989505335163668e-05, + "loss": 2.5508, + "step": 38666 + }, + { + "epoch": 1.65, + "grad_norm": 0.416015625, + "learning_rate": 1.498597304038819e-05, + "loss": 2.4075, + "step": 38667 + }, + { + "epoch": 1.65, + "grad_norm": 0.5234375, + "learning_rate": 1.4982441128144753e-05, + "loss": 2.5264, + "step": 38668 + }, + { + "epoch": 1.65, + "grad_norm": 0.5390625, + "learning_rate": 1.4978909598449253e-05, + "loss": 2.517, + "step": 38669 + }, + { + "epoch": 1.65, + "grad_norm": 0.515625, + "learning_rate": 1.4975378451317623e-05, + "loss": 2.5606, + "step": 38670 + }, + { + "epoch": 1.65, + "grad_norm": 0.4921875, + "learning_rate": 1.4971847686765706e-05, + "loss": 2.5716, + "step": 38671 + }, + { + "epoch": 1.65, + "grad_norm": 0.44921875, + "learning_rate": 1.4968317304809353e-05, + "loss": 2.4493, + "step": 38672 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.4964787305464556e-05, + "loss": 2.4212, + "step": 38673 + }, + { + "epoch": 1.65, + "grad_norm": 0.435546875, + "learning_rate": 1.4961257688747132e-05, + "loss": 2.4865, + "step": 38674 + }, + { + "epoch": 1.65, + "grad_norm": 0.458984375, + "learning_rate": 1.4957728454672936e-05, + "loss": 2.4577, + "step": 38675 + }, + { + "epoch": 1.65, + "grad_norm": 0.515625, + "learning_rate": 1.49541996032579e-05, + "loss": 2.5594, + "step": 38676 + }, + { + "epoch": 1.65, + "grad_norm": 0.474609375, + "learning_rate": 1.4950671134517902e-05, + "loss": 2.5648, + "step": 38677 + }, + { + "epoch": 1.65, + "grad_norm": 0.4765625, + "learning_rate": 1.494714304846877e-05, + "loss": 2.5167, + "step": 38678 + }, + { + "epoch": 1.65, + "grad_norm": 0.484375, + "learning_rate": 1.4943615345126449e-05, + "loss": 2.2659, + "step": 38679 + }, + { + "epoch": 1.65, + "grad_norm": 0.431640625, + "learning_rate": 1.4940088024506738e-05, + "loss": 2.2132, + "step": 38680 + }, + { + "epoch": 1.65, + "grad_norm": 0.47265625, + "learning_rate": 1.4936561086625579e-05, + "loss": 2.4342, + "step": 38681 + }, + { + "epoch": 1.65, + "grad_norm": 0.44921875, + "learning_rate": 1.493303453149878e-05, + "loss": 2.4622, + "step": 38682 + }, + { + "epoch": 1.65, + "grad_norm": 0.435546875, + "learning_rate": 1.4929508359142263e-05, + "loss": 2.4493, + "step": 38683 + }, + { + "epoch": 1.65, + "grad_norm": 0.5234375, + "learning_rate": 1.4925982569571838e-05, + "loss": 2.375, + "step": 38684 + }, + { + "epoch": 1.65, + "grad_norm": 0.47265625, + "learning_rate": 1.4922457162803404e-05, + "loss": 2.3628, + "step": 38685 + }, + { + "epoch": 1.65, + "grad_norm": 0.51171875, + "learning_rate": 1.4918932138852849e-05, + "loss": 2.4438, + "step": 38686 + }, + { + "epoch": 1.65, + "grad_norm": 0.4609375, + "learning_rate": 1.4915407497736e-05, + "loss": 2.4446, + "step": 38687 + }, + { + "epoch": 1.65, + "grad_norm": 0.5078125, + "learning_rate": 1.4911883239468705e-05, + "loss": 2.5615, + "step": 38688 + }, + { + "epoch": 1.65, + "grad_norm": 0.55078125, + "learning_rate": 1.4908359364066837e-05, + "loss": 2.5968, + "step": 38689 + }, + { + "epoch": 1.65, + "grad_norm": 0.443359375, + "learning_rate": 1.4904835871546275e-05, + "loss": 2.4281, + "step": 38690 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.4901312761922826e-05, + "loss": 2.3451, + "step": 38691 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.48977900352124e-05, + "loss": 2.5294, + "step": 38692 + }, + { + "epoch": 1.65, + "grad_norm": 0.44140625, + "learning_rate": 1.4894267691430786e-05, + "loss": 2.3535, + "step": 38693 + }, + { + "epoch": 1.65, + "grad_norm": 0.466796875, + "learning_rate": 1.4890745730593868e-05, + "loss": 2.5161, + "step": 38694 + }, + { + "epoch": 1.65, + "grad_norm": 0.44921875, + "learning_rate": 1.4887224152717527e-05, + "loss": 2.497, + "step": 38695 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.4883702957817559e-05, + "loss": 2.4626, + "step": 38696 + }, + { + "epoch": 1.65, + "grad_norm": 0.47265625, + "learning_rate": 1.4880182145909793e-05, + "loss": 2.5075, + "step": 38697 + }, + { + "epoch": 1.65, + "grad_norm": 0.515625, + "learning_rate": 1.487666171701011e-05, + "loss": 2.4236, + "step": 38698 + }, + { + "epoch": 1.65, + "grad_norm": 0.51953125, + "learning_rate": 1.487314167113436e-05, + "loss": 2.5088, + "step": 38699 + }, + { + "epoch": 1.65, + "grad_norm": 0.45703125, + "learning_rate": 1.4869622008298345e-05, + "loss": 2.5135, + "step": 38700 + }, + { + "epoch": 1.65, + "grad_norm": 0.47265625, + "learning_rate": 1.4866102728517905e-05, + "loss": 2.4679, + "step": 38701 + }, + { + "epoch": 1.65, + "grad_norm": 0.48046875, + "learning_rate": 1.486258383180893e-05, + "loss": 2.3626, + "step": 38702 + }, + { + "epoch": 1.65, + "grad_norm": 0.46875, + "learning_rate": 1.4859065318187204e-05, + "loss": 2.4199, + "step": 38703 + }, + { + "epoch": 1.65, + "grad_norm": 0.4921875, + "learning_rate": 1.4855547187668539e-05, + "loss": 2.4975, + "step": 38704 + }, + { + "epoch": 1.65, + "grad_norm": 0.470703125, + "learning_rate": 1.485202944026881e-05, + "loss": 2.4282, + "step": 38705 + }, + { + "epoch": 1.65, + "grad_norm": 0.5, + "learning_rate": 1.484851207600384e-05, + "loss": 2.504, + "step": 38706 + }, + { + "epoch": 1.65, + "grad_norm": 0.494140625, + "learning_rate": 1.4844995094889436e-05, + "loss": 2.5632, + "step": 38707 + }, + { + "epoch": 1.65, + "grad_norm": 0.515625, + "learning_rate": 1.4841478496941452e-05, + "loss": 2.5367, + "step": 38708 + }, + { + "epoch": 1.65, + "grad_norm": 0.439453125, + "learning_rate": 1.4837962282175666e-05, + "loss": 2.4388, + "step": 38709 + }, + { + "epoch": 1.65, + "grad_norm": 0.53125, + "learning_rate": 1.483444645060793e-05, + "loss": 2.5022, + "step": 38710 + }, + { + "epoch": 1.65, + "grad_norm": 0.4921875, + "learning_rate": 1.4830931002254079e-05, + "loss": 2.564, + "step": 38711 + }, + { + "epoch": 1.65, + "grad_norm": 0.498046875, + "learning_rate": 1.4827415937129918e-05, + "loss": 2.4152, + "step": 38712 + }, + { + "epoch": 1.65, + "grad_norm": 0.431640625, + "learning_rate": 1.4823901255251227e-05, + "loss": 2.4286, + "step": 38713 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.4820386956633859e-05, + "loss": 2.4912, + "step": 38714 + }, + { + "epoch": 1.65, + "grad_norm": 0.484375, + "learning_rate": 1.4816873041293644e-05, + "loss": 2.3582, + "step": 38715 + }, + { + "epoch": 1.65, + "grad_norm": 0.41796875, + "learning_rate": 1.4813359509246339e-05, + "loss": 2.4223, + "step": 38716 + }, + { + "epoch": 1.65, + "grad_norm": 0.46484375, + "learning_rate": 1.4809846360507796e-05, + "loss": 2.5182, + "step": 38717 + }, + { + "epoch": 1.65, + "grad_norm": 0.46484375, + "learning_rate": 1.4806333595093825e-05, + "loss": 2.3024, + "step": 38718 + }, + { + "epoch": 1.65, + "grad_norm": 0.478515625, + "learning_rate": 1.4802821213020223e-05, + "loss": 2.5731, + "step": 38719 + }, + { + "epoch": 1.65, + "grad_norm": 0.486328125, + "learning_rate": 1.479930921430276e-05, + "loss": 2.4419, + "step": 38720 + }, + { + "epoch": 1.65, + "grad_norm": 0.443359375, + "learning_rate": 1.4795797598957273e-05, + "loss": 2.3498, + "step": 38721 + }, + { + "epoch": 1.65, + "grad_norm": 0.490234375, + "learning_rate": 1.4792286366999587e-05, + "loss": 2.4654, + "step": 38722 + }, + { + "epoch": 1.65, + "grad_norm": 0.486328125, + "learning_rate": 1.4788775518445441e-05, + "loss": 2.5077, + "step": 38723 + }, + { + "epoch": 1.65, + "grad_norm": 0.4765625, + "learning_rate": 1.4785265053310681e-05, + "loss": 2.4078, + "step": 38724 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.4781754971611072e-05, + "loss": 2.3358, + "step": 38725 + }, + { + "epoch": 1.65, + "grad_norm": 0.470703125, + "learning_rate": 1.4778245273362423e-05, + "loss": 2.4384, + "step": 38726 + }, + { + "epoch": 1.65, + "grad_norm": 0.4765625, + "learning_rate": 1.4774735958580554e-05, + "loss": 2.3393, + "step": 38727 + }, + { + "epoch": 1.65, + "grad_norm": 0.4609375, + "learning_rate": 1.4771227027281209e-05, + "loss": 2.4218, + "step": 38728 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.4767718479480186e-05, + "loss": 2.4762, + "step": 38729 + }, + { + "epoch": 1.65, + "grad_norm": 0.5703125, + "learning_rate": 1.4764210315193283e-05, + "loss": 2.4324, + "step": 38730 + }, + { + "epoch": 1.65, + "grad_norm": 0.451171875, + "learning_rate": 1.47607025344363e-05, + "loss": 2.5226, + "step": 38731 + }, + { + "epoch": 1.65, + "grad_norm": 0.490234375, + "learning_rate": 1.4757195137224977e-05, + "loss": 2.3962, + "step": 38732 + }, + { + "epoch": 1.65, + "grad_norm": 0.45703125, + "learning_rate": 1.4753688123575127e-05, + "loss": 2.4587, + "step": 38733 + }, + { + "epoch": 1.65, + "grad_norm": 0.45703125, + "learning_rate": 1.4750181493502557e-05, + "loss": 2.4526, + "step": 38734 + }, + { + "epoch": 1.65, + "grad_norm": 0.45703125, + "learning_rate": 1.4746675247023e-05, + "loss": 2.4463, + "step": 38735 + }, + { + "epoch": 1.65, + "grad_norm": 0.44921875, + "learning_rate": 1.4743169384152233e-05, + "loss": 2.5193, + "step": 38736 + }, + { + "epoch": 1.65, + "grad_norm": 0.43359375, + "learning_rate": 1.4739663904906054e-05, + "loss": 2.4846, + "step": 38737 + }, + { + "epoch": 1.65, + "grad_norm": 0.4375, + "learning_rate": 1.4736158809300238e-05, + "loss": 2.5684, + "step": 38738 + }, + { + "epoch": 1.65, + "grad_norm": 0.5, + "learning_rate": 1.473265409735053e-05, + "loss": 2.497, + "step": 38739 + }, + { + "epoch": 1.65, + "grad_norm": 0.49609375, + "learning_rate": 1.4729149769072736e-05, + "loss": 2.3996, + "step": 38740 + }, + { + "epoch": 1.65, + "grad_norm": 0.4921875, + "learning_rate": 1.4725645824482592e-05, + "loss": 2.5359, + "step": 38741 + }, + { + "epoch": 1.65, + "grad_norm": 0.408203125, + "learning_rate": 1.4722142263595873e-05, + "loss": 2.4151, + "step": 38742 + }, + { + "epoch": 1.65, + "grad_norm": 0.466796875, + "learning_rate": 1.4718639086428365e-05, + "loss": 2.484, + "step": 38743 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.4715136292995824e-05, + "loss": 2.3322, + "step": 38744 + }, + { + "epoch": 1.65, + "grad_norm": 0.48828125, + "learning_rate": 1.4711633883313969e-05, + "loss": 2.2821, + "step": 38745 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.47081318573986e-05, + "loss": 2.3715, + "step": 38746 + }, + { + "epoch": 1.65, + "grad_norm": 0.5, + "learning_rate": 1.470463021526548e-05, + "loss": 2.5571, + "step": 38747 + }, + { + "epoch": 1.65, + "grad_norm": 0.474609375, + "learning_rate": 1.4701128956930343e-05, + "loss": 2.3802, + "step": 38748 + }, + { + "epoch": 1.65, + "grad_norm": 0.51171875, + "learning_rate": 1.4697628082408943e-05, + "loss": 2.4337, + "step": 38749 + }, + { + "epoch": 1.65, + "grad_norm": 0.54296875, + "learning_rate": 1.4694127591717067e-05, + "loss": 2.5707, + "step": 38750 + }, + { + "epoch": 1.65, + "grad_norm": 0.50390625, + "learning_rate": 1.4690627484870445e-05, + "loss": 2.4411, + "step": 38751 + }, + { + "epoch": 1.65, + "grad_norm": 0.48828125, + "learning_rate": 1.4687127761884788e-05, + "loss": 2.4268, + "step": 38752 + }, + { + "epoch": 1.65, + "grad_norm": 0.48046875, + "learning_rate": 1.4683628422775885e-05, + "loss": 2.3834, + "step": 38753 + }, + { + "epoch": 1.65, + "grad_norm": 0.42578125, + "learning_rate": 1.46801294675595e-05, + "loss": 2.4353, + "step": 38754 + }, + { + "epoch": 1.65, + "grad_norm": 0.44140625, + "learning_rate": 1.4676630896251331e-05, + "loss": 2.3251, + "step": 38755 + }, + { + "epoch": 1.65, + "grad_norm": 0.431640625, + "learning_rate": 1.4673132708867155e-05, + "loss": 2.4356, + "step": 38756 + }, + { + "epoch": 1.65, + "grad_norm": 0.4921875, + "learning_rate": 1.466963490542268e-05, + "loss": 2.5161, + "step": 38757 + }, + { + "epoch": 1.65, + "grad_norm": 0.41796875, + "learning_rate": 1.4666137485933662e-05, + "loss": 2.5437, + "step": 38758 + }, + { + "epoch": 1.65, + "grad_norm": 0.451171875, + "learning_rate": 1.4662640450415865e-05, + "loss": 2.4549, + "step": 38759 + }, + { + "epoch": 1.65, + "grad_norm": 0.4453125, + "learning_rate": 1.4659143798885e-05, + "loss": 2.4681, + "step": 38760 + }, + { + "epoch": 1.65, + "grad_norm": 0.46875, + "learning_rate": 1.4655647531356754e-05, + "loss": 2.5795, + "step": 38761 + }, + { + "epoch": 1.65, + "grad_norm": 0.421875, + "learning_rate": 1.4652151647846945e-05, + "loss": 2.3977, + "step": 38762 + }, + { + "epoch": 1.65, + "grad_norm": 0.423828125, + "learning_rate": 1.4648656148371265e-05, + "loss": 2.3923, + "step": 38763 + }, + { + "epoch": 1.65, + "grad_norm": 0.470703125, + "learning_rate": 1.4645161032945409e-05, + "loss": 2.4449, + "step": 38764 + }, + { + "epoch": 1.65, + "grad_norm": 0.5390625, + "learning_rate": 1.4641666301585143e-05, + "loss": 2.5288, + "step": 38765 + }, + { + "epoch": 1.65, + "grad_norm": 0.478515625, + "learning_rate": 1.4638171954306213e-05, + "loss": 2.5228, + "step": 38766 + }, + { + "epoch": 1.65, + "grad_norm": 0.46484375, + "learning_rate": 1.4634677991124301e-05, + "loss": 2.4295, + "step": 38767 + }, + { + "epoch": 1.65, + "grad_norm": 0.431640625, + "learning_rate": 1.4631184412055121e-05, + "loss": 2.4725, + "step": 38768 + }, + { + "epoch": 1.65, + "grad_norm": 0.44921875, + "learning_rate": 1.4627691217114415e-05, + "loss": 2.2516, + "step": 38769 + }, + { + "epoch": 1.65, + "grad_norm": 0.4453125, + "learning_rate": 1.4624198406317925e-05, + "loss": 2.4074, + "step": 38770 + }, + { + "epoch": 1.65, + "grad_norm": 0.486328125, + "learning_rate": 1.4620705979681315e-05, + "loss": 2.2585, + "step": 38771 + }, + { + "epoch": 1.65, + "grad_norm": 0.443359375, + "learning_rate": 1.4617213937220342e-05, + "loss": 2.5429, + "step": 38772 + }, + { + "epoch": 1.65, + "grad_norm": 0.5, + "learning_rate": 1.4613722278950681e-05, + "loss": 2.379, + "step": 38773 + }, + { + "epoch": 1.65, + "grad_norm": 0.48046875, + "learning_rate": 1.4610231004888076e-05, + "loss": 2.4783, + "step": 38774 + }, + { + "epoch": 1.65, + "grad_norm": 0.48828125, + "learning_rate": 1.4606740115048234e-05, + "loss": 2.4209, + "step": 38775 + }, + { + "epoch": 1.65, + "grad_norm": 0.470703125, + "learning_rate": 1.4603249609446867e-05, + "loss": 2.5244, + "step": 38776 + }, + { + "epoch": 1.65, + "grad_norm": 0.447265625, + "learning_rate": 1.4599759488099607e-05, + "loss": 2.5431, + "step": 38777 + }, + { + "epoch": 1.65, + "grad_norm": 0.4375, + "learning_rate": 1.4596269751022274e-05, + "loss": 2.3732, + "step": 38778 + }, + { + "epoch": 1.65, + "grad_norm": 0.486328125, + "learning_rate": 1.4592780398230521e-05, + "loss": 2.4384, + "step": 38779 + }, + { + "epoch": 1.65, + "grad_norm": 0.419921875, + "learning_rate": 1.4589291429740016e-05, + "loss": 2.4993, + "step": 38780 + }, + { + "epoch": 1.65, + "grad_norm": 0.48046875, + "learning_rate": 1.4585802845566476e-05, + "loss": 2.3071, + "step": 38781 + }, + { + "epoch": 1.65, + "grad_norm": 0.478515625, + "learning_rate": 1.4582314645725636e-05, + "loss": 2.3098, + "step": 38782 + }, + { + "epoch": 1.65, + "grad_norm": 0.451171875, + "learning_rate": 1.4578826830233171e-05, + "loss": 2.4939, + "step": 38783 + }, + { + "epoch": 1.65, + "grad_norm": 0.498046875, + "learning_rate": 1.4575339399104738e-05, + "loss": 2.5443, + "step": 38784 + }, + { + "epoch": 1.65, + "grad_norm": 0.451171875, + "learning_rate": 1.4571852352356064e-05, + "loss": 2.5231, + "step": 38785 + }, + { + "epoch": 1.65, + "grad_norm": 0.486328125, + "learning_rate": 1.456836569000285e-05, + "loss": 2.4767, + "step": 38786 + }, + { + "epoch": 1.65, + "grad_norm": 0.470703125, + "learning_rate": 1.456487941206075e-05, + "loss": 2.3073, + "step": 38787 + }, + { + "epoch": 1.65, + "grad_norm": 0.43359375, + "learning_rate": 1.4561393518545485e-05, + "loss": 2.4867, + "step": 38788 + }, + { + "epoch": 1.65, + "grad_norm": 0.46875, + "learning_rate": 1.455790800947271e-05, + "loss": 2.3941, + "step": 38789 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.4554422884858121e-05, + "loss": 2.4616, + "step": 38790 + }, + { + "epoch": 1.65, + "grad_norm": 0.54296875, + "learning_rate": 1.455093814471743e-05, + "loss": 2.5135, + "step": 38791 + }, + { + "epoch": 1.65, + "grad_norm": 0.49609375, + "learning_rate": 1.4547453789066279e-05, + "loss": 2.5166, + "step": 38792 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.4543969817920322e-05, + "loss": 2.4553, + "step": 38793 + }, + { + "epoch": 1.65, + "grad_norm": 0.482421875, + "learning_rate": 1.4540486231295303e-05, + "loss": 2.5403, + "step": 38794 + }, + { + "epoch": 1.65, + "grad_norm": 0.470703125, + "learning_rate": 1.4537003029206875e-05, + "loss": 2.4087, + "step": 38795 + }, + { + "epoch": 1.65, + "grad_norm": 0.52734375, + "learning_rate": 1.4533520211670681e-05, + "loss": 2.5309, + "step": 38796 + }, + { + "epoch": 1.65, + "grad_norm": 0.51171875, + "learning_rate": 1.4530037778702399e-05, + "loss": 2.526, + "step": 38797 + }, + { + "epoch": 1.65, + "grad_norm": 0.484375, + "learning_rate": 1.4526555730317737e-05, + "loss": 2.5602, + "step": 38798 + }, + { + "epoch": 1.65, + "grad_norm": 0.515625, + "learning_rate": 1.452307406653235e-05, + "loss": 2.4186, + "step": 38799 + }, + { + "epoch": 1.65, + "grad_norm": 0.52734375, + "learning_rate": 1.4519592787361857e-05, + "loss": 2.5158, + "step": 38800 + }, + { + "epoch": 1.65, + "grad_norm": 0.427734375, + "learning_rate": 1.451611189282196e-05, + "loss": 2.359, + "step": 38801 + }, + { + "epoch": 1.65, + "grad_norm": 0.46875, + "learning_rate": 1.4512631382928332e-05, + "loss": 2.5288, + "step": 38802 + }, + { + "epoch": 1.65, + "grad_norm": 0.47265625, + "learning_rate": 1.4509151257696607e-05, + "loss": 2.3702, + "step": 38803 + }, + { + "epoch": 1.65, + "grad_norm": 0.458984375, + "learning_rate": 1.4505671517142483e-05, + "loss": 2.4653, + "step": 38804 + }, + { + "epoch": 1.65, + "grad_norm": 0.515625, + "learning_rate": 1.4502192161281592e-05, + "loss": 2.4639, + "step": 38805 + }, + { + "epoch": 1.65, + "grad_norm": 0.431640625, + "learning_rate": 1.4498713190129543e-05, + "loss": 2.4761, + "step": 38806 + }, + { + "epoch": 1.65, + "grad_norm": 0.4765625, + "learning_rate": 1.4495234603702078e-05, + "loss": 2.4086, + "step": 38807 + }, + { + "epoch": 1.65, + "grad_norm": 0.427734375, + "learning_rate": 1.449175640201481e-05, + "loss": 2.4855, + "step": 38808 + }, + { + "epoch": 1.65, + "grad_norm": 0.40234375, + "learning_rate": 1.448827858508337e-05, + "loss": 2.5597, + "step": 38809 + }, + { + "epoch": 1.65, + "grad_norm": 0.458984375, + "learning_rate": 1.4484801152923433e-05, + "loss": 2.5268, + "step": 38810 + }, + { + "epoch": 1.65, + "grad_norm": 0.51171875, + "learning_rate": 1.4481324105550653e-05, + "loss": 2.4613, + "step": 38811 + }, + { + "epoch": 1.65, + "grad_norm": 0.55078125, + "learning_rate": 1.4477847442980629e-05, + "loss": 2.4991, + "step": 38812 + }, + { + "epoch": 1.65, + "grad_norm": 0.46875, + "learning_rate": 1.4474371165229073e-05, + "loss": 2.4012, + "step": 38813 + }, + { + "epoch": 1.65, + "grad_norm": 0.439453125, + "learning_rate": 1.4470895272311557e-05, + "loss": 2.4373, + "step": 38814 + }, + { + "epoch": 1.65, + "grad_norm": 0.478515625, + "learning_rate": 1.4467419764243773e-05, + "loss": 2.3835, + "step": 38815 + }, + { + "epoch": 1.65, + "grad_norm": 0.46875, + "learning_rate": 1.4463944641041349e-05, + "loss": 2.4561, + "step": 38816 + }, + { + "epoch": 1.65, + "grad_norm": 0.42578125, + "learning_rate": 1.446046990271992e-05, + "loss": 2.5638, + "step": 38817 + }, + { + "epoch": 1.65, + "grad_norm": 0.439453125, + "learning_rate": 1.4456995549295083e-05, + "loss": 2.5166, + "step": 38818 + }, + { + "epoch": 1.65, + "grad_norm": 0.421875, + "learning_rate": 1.4453521580782515e-05, + "loss": 2.4385, + "step": 38819 + }, + { + "epoch": 1.65, + "grad_norm": 0.4375, + "learning_rate": 1.4450047997197858e-05, + "loss": 2.5587, + "step": 38820 + }, + { + "epoch": 1.65, + "grad_norm": 0.48046875, + "learning_rate": 1.4446574798556712e-05, + "loss": 2.5361, + "step": 38821 + }, + { + "epoch": 1.65, + "grad_norm": 0.455078125, + "learning_rate": 1.4443101984874674e-05, + "loss": 2.507, + "step": 38822 + }, + { + "epoch": 1.65, + "grad_norm": 0.5234375, + "learning_rate": 1.4439629556167455e-05, + "loss": 2.4952, + "step": 38823 + }, + { + "epoch": 1.65, + "grad_norm": 0.431640625, + "learning_rate": 1.4436157512450632e-05, + "loss": 2.4792, + "step": 38824 + }, + { + "epoch": 1.65, + "grad_norm": 0.484375, + "learning_rate": 1.4432685853739814e-05, + "loss": 2.5586, + "step": 38825 + }, + { + "epoch": 1.65, + "grad_norm": 0.390625, + "learning_rate": 1.4429214580050631e-05, + "loss": 2.4114, + "step": 38826 + }, + { + "epoch": 1.65, + "grad_norm": 0.44921875, + "learning_rate": 1.442574369139873e-05, + "loss": 2.4702, + "step": 38827 + }, + { + "epoch": 1.65, + "grad_norm": 0.498046875, + "learning_rate": 1.4422273187799695e-05, + "loss": 2.4168, + "step": 38828 + }, + { + "epoch": 1.65, + "grad_norm": 0.47265625, + "learning_rate": 1.4418803069269172e-05, + "loss": 2.5266, + "step": 38829 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.4415333335822733e-05, + "loss": 2.4687, + "step": 38830 + }, + { + "epoch": 1.65, + "grad_norm": 0.41015625, + "learning_rate": 1.4411863987476027e-05, + "loss": 2.3369, + "step": 38831 + }, + { + "epoch": 1.65, + "grad_norm": 0.443359375, + "learning_rate": 1.440839502424467e-05, + "loss": 2.3563, + "step": 38832 + }, + { + "epoch": 1.65, + "grad_norm": 0.5234375, + "learning_rate": 1.4404926446144262e-05, + "loss": 2.3807, + "step": 38833 + }, + { + "epoch": 1.65, + "grad_norm": 0.43359375, + "learning_rate": 1.440145825319037e-05, + "loss": 2.3234, + "step": 38834 + }, + { + "epoch": 1.65, + "grad_norm": 0.462890625, + "learning_rate": 1.4397990445398634e-05, + "loss": 2.4105, + "step": 38835 + }, + { + "epoch": 1.65, + "grad_norm": 0.53125, + "learning_rate": 1.4394523022784689e-05, + "loss": 2.3435, + "step": 38836 + }, + { + "epoch": 1.65, + "grad_norm": 0.474609375, + "learning_rate": 1.4391055985364099e-05, + "loss": 2.5436, + "step": 38837 + }, + { + "epoch": 1.65, + "grad_norm": 0.44921875, + "learning_rate": 1.438758933315243e-05, + "loss": 2.612, + "step": 38838 + }, + { + "epoch": 1.66, + "grad_norm": 0.404296875, + "learning_rate": 1.4384123066165367e-05, + "loss": 2.2901, + "step": 38839 + }, + { + "epoch": 1.66, + "grad_norm": 0.51171875, + "learning_rate": 1.4380657184418456e-05, + "loss": 2.5434, + "step": 38840 + }, + { + "epoch": 1.66, + "grad_norm": 0.466796875, + "learning_rate": 1.4377191687927272e-05, + "loss": 2.4975, + "step": 38841 + }, + { + "epoch": 1.66, + "grad_norm": 0.498046875, + "learning_rate": 1.4373726576707436e-05, + "loss": 2.4245, + "step": 38842 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.4370261850774568e-05, + "loss": 2.5235, + "step": 38843 + }, + { + "epoch": 1.66, + "grad_norm": 0.474609375, + "learning_rate": 1.4366797510144193e-05, + "loss": 2.459, + "step": 38844 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.436333355483197e-05, + "loss": 2.4316, + "step": 38845 + }, + { + "epoch": 1.66, + "grad_norm": 0.51953125, + "learning_rate": 1.4359869984853414e-05, + "loss": 2.4109, + "step": 38846 + }, + { + "epoch": 1.66, + "grad_norm": 0.453125, + "learning_rate": 1.4356406800224153e-05, + "loss": 2.2997, + "step": 38847 + }, + { + "epoch": 1.66, + "grad_norm": 0.48046875, + "learning_rate": 1.4352944000959778e-05, + "loss": 2.5631, + "step": 38848 + }, + { + "epoch": 1.66, + "grad_norm": 0.48828125, + "learning_rate": 1.4349481587075864e-05, + "loss": 2.381, + "step": 38849 + }, + { + "epoch": 1.66, + "grad_norm": 0.458984375, + "learning_rate": 1.4346019558587954e-05, + "loss": 2.3759, + "step": 38850 + }, + { + "epoch": 1.66, + "grad_norm": 0.478515625, + "learning_rate": 1.4342557915511646e-05, + "loss": 2.3307, + "step": 38851 + }, + { + "epoch": 1.66, + "grad_norm": 0.443359375, + "learning_rate": 1.433909665786256e-05, + "loss": 2.52, + "step": 38852 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.4335635785656221e-05, + "loss": 2.37, + "step": 38853 + }, + { + "epoch": 1.66, + "grad_norm": 0.451171875, + "learning_rate": 1.4332175298908179e-05, + "loss": 2.4696, + "step": 38854 + }, + { + "epoch": 1.66, + "grad_norm": 0.484375, + "learning_rate": 1.432871519763408e-05, + "loss": 2.4554, + "step": 38855 + }, + { + "epoch": 1.66, + "grad_norm": 0.47265625, + "learning_rate": 1.4325255481849453e-05, + "loss": 2.4616, + "step": 38856 + }, + { + "epoch": 1.66, + "grad_norm": 0.44921875, + "learning_rate": 1.4321796151569844e-05, + "loss": 2.4891, + "step": 38857 + }, + { + "epoch": 1.66, + "grad_norm": 0.458984375, + "learning_rate": 1.4318337206810839e-05, + "loss": 2.4617, + "step": 38858 + }, + { + "epoch": 1.66, + "grad_norm": 0.4765625, + "learning_rate": 1.4314878647588026e-05, + "loss": 2.3819, + "step": 38859 + }, + { + "epoch": 1.66, + "grad_norm": 0.466796875, + "learning_rate": 1.4311420473916926e-05, + "loss": 2.425, + "step": 38860 + }, + { + "epoch": 1.66, + "grad_norm": 0.51171875, + "learning_rate": 1.4307962685813137e-05, + "loss": 2.5481, + "step": 38861 + }, + { + "epoch": 1.66, + "grad_norm": 0.421875, + "learning_rate": 1.430450528329218e-05, + "loss": 2.3705, + "step": 38862 + }, + { + "epoch": 1.66, + "grad_norm": 0.45703125, + "learning_rate": 1.4301048266369621e-05, + "loss": 2.4166, + "step": 38863 + }, + { + "epoch": 1.66, + "grad_norm": 0.486328125, + "learning_rate": 1.4297591635061059e-05, + "loss": 2.5085, + "step": 38864 + }, + { + "epoch": 1.66, + "grad_norm": 0.474609375, + "learning_rate": 1.4294135389382013e-05, + "loss": 2.3559, + "step": 38865 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.4290679529348017e-05, + "loss": 2.6265, + "step": 38866 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.4287224054974624e-05, + "loss": 2.5643, + "step": 38867 + }, + { + "epoch": 1.66, + "grad_norm": 0.4453125, + "learning_rate": 1.4283768966277434e-05, + "loss": 2.6044, + "step": 38868 + }, + { + "epoch": 1.66, + "grad_norm": 0.48828125, + "learning_rate": 1.4280314263271932e-05, + "loss": 2.4069, + "step": 38869 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.4276859945973697e-05, + "loss": 2.5038, + "step": 38870 + }, + { + "epoch": 1.66, + "grad_norm": 0.50390625, + "learning_rate": 1.4273406014398272e-05, + "loss": 2.4105, + "step": 38871 + }, + { + "epoch": 1.66, + "grad_norm": 0.50390625, + "learning_rate": 1.4269952468561199e-05, + "loss": 2.4071, + "step": 38872 + }, + { + "epoch": 1.66, + "grad_norm": 0.515625, + "learning_rate": 1.4266499308477987e-05, + "loss": 2.5707, + "step": 38873 + }, + { + "epoch": 1.66, + "grad_norm": 0.46875, + "learning_rate": 1.4263046534164192e-05, + "loss": 2.4515, + "step": 38874 + }, + { + "epoch": 1.66, + "grad_norm": 0.51171875, + "learning_rate": 1.425959414563538e-05, + "loss": 2.509, + "step": 38875 + }, + { + "epoch": 1.66, + "grad_norm": 0.470703125, + "learning_rate": 1.4256142142907037e-05, + "loss": 2.456, + "step": 38876 + }, + { + "epoch": 1.66, + "grad_norm": 0.4296875, + "learning_rate": 1.4252690525994739e-05, + "loss": 2.3007, + "step": 38877 + }, + { + "epoch": 1.66, + "grad_norm": 0.42578125, + "learning_rate": 1.4249239294913974e-05, + "loss": 2.514, + "step": 38878 + }, + { + "epoch": 1.66, + "grad_norm": 0.4296875, + "learning_rate": 1.4245788449680297e-05, + "loss": 2.4816, + "step": 38879 + }, + { + "epoch": 1.66, + "grad_norm": 0.42578125, + "learning_rate": 1.4242337990309252e-05, + "loss": 2.402, + "step": 38880 + }, + { + "epoch": 1.66, + "grad_norm": 0.46875, + "learning_rate": 1.4238887916816346e-05, + "loss": 2.4901, + "step": 38881 + }, + { + "epoch": 1.66, + "grad_norm": 0.474609375, + "learning_rate": 1.423543822921708e-05, + "loss": 2.462, + "step": 38882 + }, + { + "epoch": 1.66, + "grad_norm": 0.482421875, + "learning_rate": 1.4231988927526996e-05, + "loss": 2.3426, + "step": 38883 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.422854001176165e-05, + "loss": 2.5174, + "step": 38884 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.4225091481936493e-05, + "loss": 2.3849, + "step": 38885 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.422164333806708e-05, + "loss": 2.337, + "step": 38886 + }, + { + "epoch": 1.66, + "grad_norm": 0.50390625, + "learning_rate": 1.4218195580168948e-05, + "loss": 2.4443, + "step": 38887 + }, + { + "epoch": 1.66, + "grad_norm": 0.443359375, + "learning_rate": 1.421474820825759e-05, + "loss": 2.5245, + "step": 38888 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.4211301222348483e-05, + "loss": 2.2619, + "step": 38889 + }, + { + "epoch": 1.66, + "grad_norm": 0.4765625, + "learning_rate": 1.4207854622457183e-05, + "loss": 2.3335, + "step": 38890 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.4204408408599191e-05, + "loss": 2.4353, + "step": 38891 + }, + { + "epoch": 1.66, + "grad_norm": 0.4375, + "learning_rate": 1.420096258078999e-05, + "loss": 2.3586, + "step": 38892 + }, + { + "epoch": 1.66, + "grad_norm": 0.49609375, + "learning_rate": 1.4197517139045124e-05, + "loss": 2.6519, + "step": 38893 + }, + { + "epoch": 1.66, + "grad_norm": 0.4140625, + "learning_rate": 1.419407208338005e-05, + "loss": 2.3896, + "step": 38894 + }, + { + "epoch": 1.66, + "grad_norm": 0.48046875, + "learning_rate": 1.4190627413810297e-05, + "loss": 2.4451, + "step": 38895 + }, + { + "epoch": 1.66, + "grad_norm": 0.474609375, + "learning_rate": 1.4187183130351378e-05, + "loss": 2.4138, + "step": 38896 + }, + { + "epoch": 1.66, + "grad_norm": 0.416015625, + "learning_rate": 1.4183739233018778e-05, + "loss": 2.418, + "step": 38897 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.4180295721827962e-05, + "loss": 2.529, + "step": 38898 + }, + { + "epoch": 1.66, + "grad_norm": 0.48828125, + "learning_rate": 1.4176852596794454e-05, + "loss": 2.4088, + "step": 38899 + }, + { + "epoch": 1.66, + "grad_norm": 0.5234375, + "learning_rate": 1.4173409857933772e-05, + "loss": 2.3514, + "step": 38900 + }, + { + "epoch": 1.66, + "grad_norm": 0.3984375, + "learning_rate": 1.416996750526135e-05, + "loss": 2.3414, + "step": 38901 + }, + { + "epoch": 1.66, + "grad_norm": 0.47265625, + "learning_rate": 1.4166525538792708e-05, + "loss": 2.3427, + "step": 38902 + }, + { + "epoch": 1.66, + "grad_norm": 0.466796875, + "learning_rate": 1.4163083958543354e-05, + "loss": 2.3571, + "step": 38903 + }, + { + "epoch": 1.66, + "grad_norm": 0.41015625, + "learning_rate": 1.4159642764528757e-05, + "loss": 2.3695, + "step": 38904 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.4156201956764381e-05, + "loss": 2.4836, + "step": 38905 + }, + { + "epoch": 1.66, + "grad_norm": 0.494140625, + "learning_rate": 1.4152761535265713e-05, + "loss": 2.424, + "step": 38906 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.4149321500048263e-05, + "loss": 2.5467, + "step": 38907 + }, + { + "epoch": 1.66, + "grad_norm": 0.4921875, + "learning_rate": 1.4145881851127473e-05, + "loss": 2.4285, + "step": 38908 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.4142442588518866e-05, + "loss": 2.5695, + "step": 38909 + }, + { + "epoch": 1.66, + "grad_norm": 0.484375, + "learning_rate": 1.4139003712237863e-05, + "loss": 2.4786, + "step": 38910 + }, + { + "epoch": 1.66, + "grad_norm": 0.443359375, + "learning_rate": 1.4135565222299973e-05, + "loss": 2.3372, + "step": 38911 + }, + { + "epoch": 1.66, + "grad_norm": 0.46875, + "learning_rate": 1.4132127118720674e-05, + "loss": 2.4233, + "step": 38912 + }, + { + "epoch": 1.66, + "grad_norm": 0.421875, + "learning_rate": 1.4128689401515415e-05, + "loss": 2.5302, + "step": 38913 + }, + { + "epoch": 1.66, + "grad_norm": 0.546875, + "learning_rate": 1.4125252070699669e-05, + "loss": 2.4442, + "step": 38914 + }, + { + "epoch": 1.66, + "grad_norm": 0.453125, + "learning_rate": 1.4121815126288884e-05, + "loss": 2.4645, + "step": 38915 + }, + { + "epoch": 1.66, + "grad_norm": 0.49609375, + "learning_rate": 1.4118378568298585e-05, + "loss": 2.4532, + "step": 38916 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.4114942396744157e-05, + "loss": 2.394, + "step": 38917 + }, + { + "epoch": 1.66, + "grad_norm": 0.453125, + "learning_rate": 1.4111506611641112e-05, + "loss": 2.4555, + "step": 38918 + }, + { + "epoch": 1.66, + "grad_norm": 0.46484375, + "learning_rate": 1.4108071213004926e-05, + "loss": 2.3825, + "step": 38919 + }, + { + "epoch": 1.66, + "grad_norm": 0.46484375, + "learning_rate": 1.4104636200851018e-05, + "loss": 2.5002, + "step": 38920 + }, + { + "epoch": 1.66, + "grad_norm": 0.451171875, + "learning_rate": 1.4101201575194844e-05, + "loss": 2.323, + "step": 38921 + }, + { + "epoch": 1.66, + "grad_norm": 0.451171875, + "learning_rate": 1.4097767336051859e-05, + "loss": 2.427, + "step": 38922 + }, + { + "epoch": 1.66, + "grad_norm": 0.4921875, + "learning_rate": 1.4094333483437561e-05, + "loss": 2.3963, + "step": 38923 + }, + { + "epoch": 1.66, + "grad_norm": 0.45703125, + "learning_rate": 1.4090900017367337e-05, + "loss": 2.4761, + "step": 38924 + }, + { + "epoch": 1.66, + "grad_norm": 0.458984375, + "learning_rate": 1.4087466937856686e-05, + "loss": 2.4155, + "step": 38925 + }, + { + "epoch": 1.66, + "grad_norm": 0.49609375, + "learning_rate": 1.408403424492103e-05, + "loss": 2.4524, + "step": 38926 + }, + { + "epoch": 1.66, + "grad_norm": 0.443359375, + "learning_rate": 1.408060193857581e-05, + "loss": 2.4624, + "step": 38927 + }, + { + "epoch": 1.66, + "grad_norm": 0.51171875, + "learning_rate": 1.4077170018836506e-05, + "loss": 2.5271, + "step": 38928 + }, + { + "epoch": 1.66, + "grad_norm": 0.458984375, + "learning_rate": 1.4073738485718524e-05, + "loss": 2.3387, + "step": 38929 + }, + { + "epoch": 1.66, + "grad_norm": 0.470703125, + "learning_rate": 1.407030733923731e-05, + "loss": 2.495, + "step": 38930 + }, + { + "epoch": 1.66, + "grad_norm": 0.478515625, + "learning_rate": 1.4066876579408305e-05, + "loss": 2.3902, + "step": 38931 + }, + { + "epoch": 1.66, + "grad_norm": 0.470703125, + "learning_rate": 1.4063446206246967e-05, + "loss": 2.4176, + "step": 38932 + }, + { + "epoch": 1.66, + "grad_norm": 0.466796875, + "learning_rate": 1.4060016219768702e-05, + "loss": 2.3784, + "step": 38933 + }, + { + "epoch": 1.66, + "grad_norm": 0.5390625, + "learning_rate": 1.4056586619988964e-05, + "loss": 2.5143, + "step": 38934 + }, + { + "epoch": 1.66, + "grad_norm": 0.48046875, + "learning_rate": 1.4053157406923168e-05, + "loss": 2.5267, + "step": 38935 + }, + { + "epoch": 1.66, + "grad_norm": 0.49609375, + "learning_rate": 1.4049728580586774e-05, + "loss": 2.5266, + "step": 38936 + }, + { + "epoch": 1.66, + "grad_norm": 0.482421875, + "learning_rate": 1.4046300140995173e-05, + "loss": 2.3582, + "step": 38937 + }, + { + "epoch": 1.66, + "grad_norm": 0.486328125, + "learning_rate": 1.4042872088163816e-05, + "loss": 2.4883, + "step": 38938 + }, + { + "epoch": 1.66, + "grad_norm": 0.5, + "learning_rate": 1.4039444422108117e-05, + "loss": 2.4584, + "step": 38939 + }, + { + "epoch": 1.66, + "grad_norm": 0.42578125, + "learning_rate": 1.4036017142843495e-05, + "loss": 2.4528, + "step": 38940 + }, + { + "epoch": 1.66, + "grad_norm": 0.53125, + "learning_rate": 1.4032590250385392e-05, + "loss": 2.5188, + "step": 38941 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.402916374474923e-05, + "loss": 2.3809, + "step": 38942 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.4025737625950353e-05, + "loss": 2.4754, + "step": 38943 + }, + { + "epoch": 1.66, + "grad_norm": 0.484375, + "learning_rate": 1.4022311894004292e-05, + "loss": 2.5593, + "step": 38944 + }, + { + "epoch": 1.66, + "grad_norm": 0.4765625, + "learning_rate": 1.4018886548926414e-05, + "loss": 2.4446, + "step": 38945 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.4015461590732082e-05, + "loss": 2.3378, + "step": 38946 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.4012037019436764e-05, + "loss": 2.4088, + "step": 38947 + }, + { + "epoch": 1.66, + "grad_norm": 0.41796875, + "learning_rate": 1.4008612835055867e-05, + "loss": 2.3878, + "step": 38948 + }, + { + "epoch": 1.66, + "grad_norm": 0.46875, + "learning_rate": 1.400518903760477e-05, + "loss": 2.379, + "step": 38949 + }, + { + "epoch": 1.66, + "grad_norm": 0.453125, + "learning_rate": 1.4001765627098918e-05, + "loss": 2.3227, + "step": 38950 + }, + { + "epoch": 1.66, + "grad_norm": 0.462890625, + "learning_rate": 1.3998342603553671e-05, + "loss": 2.5189, + "step": 38951 + }, + { + "epoch": 1.66, + "grad_norm": 0.4375, + "learning_rate": 1.3994919966984465e-05, + "loss": 2.4057, + "step": 38952 + }, + { + "epoch": 1.66, + "grad_norm": 0.48046875, + "learning_rate": 1.3991497717406676e-05, + "loss": 2.4401, + "step": 38953 + }, + { + "epoch": 1.66, + "grad_norm": 0.43359375, + "learning_rate": 1.3988075854835726e-05, + "loss": 2.4501, + "step": 38954 + }, + { + "epoch": 1.66, + "grad_norm": 0.42578125, + "learning_rate": 1.3984654379286987e-05, + "loss": 2.3721, + "step": 38955 + }, + { + "epoch": 1.66, + "grad_norm": 0.484375, + "learning_rate": 1.3981233290775863e-05, + "loss": 2.4541, + "step": 38956 + }, + { + "epoch": 1.66, + "grad_norm": 0.474609375, + "learning_rate": 1.3977812589317784e-05, + "loss": 2.4288, + "step": 38957 + }, + { + "epoch": 1.66, + "grad_norm": 0.466796875, + "learning_rate": 1.3974392274928105e-05, + "loss": 2.4438, + "step": 38958 + }, + { + "epoch": 1.66, + "grad_norm": 0.546875, + "learning_rate": 1.397097234762218e-05, + "loss": 2.496, + "step": 38959 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.3967552807415473e-05, + "loss": 2.3859, + "step": 38960 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.3964133654323352e-05, + "loss": 2.5855, + "step": 38961 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.3960714888361159e-05, + "loss": 2.4081, + "step": 38962 + }, + { + "epoch": 1.66, + "grad_norm": 0.478515625, + "learning_rate": 1.3957296509544304e-05, + "loss": 2.5245, + "step": 38963 + }, + { + "epoch": 1.66, + "grad_norm": 0.5, + "learning_rate": 1.3953878517888197e-05, + "loss": 2.536, + "step": 38964 + }, + { + "epoch": 1.66, + "grad_norm": 0.458984375, + "learning_rate": 1.3950460913408159e-05, + "loss": 2.3776, + "step": 38965 + }, + { + "epoch": 1.66, + "grad_norm": 0.41796875, + "learning_rate": 1.3947043696119633e-05, + "loss": 2.4875, + "step": 38966 + }, + { + "epoch": 1.66, + "grad_norm": 0.427734375, + "learning_rate": 1.3943626866037928e-05, + "loss": 2.363, + "step": 38967 + }, + { + "epoch": 1.66, + "grad_norm": 0.4453125, + "learning_rate": 1.3940210423178478e-05, + "loss": 2.5026, + "step": 38968 + }, + { + "epoch": 1.66, + "grad_norm": 0.42578125, + "learning_rate": 1.3936794367556605e-05, + "loss": 2.3943, + "step": 38969 + }, + { + "epoch": 1.66, + "grad_norm": 0.458984375, + "learning_rate": 1.3933378699187738e-05, + "loss": 2.4928, + "step": 38970 + }, + { + "epoch": 1.66, + "grad_norm": 0.431640625, + "learning_rate": 1.3929963418087178e-05, + "loss": 2.4285, + "step": 38971 + }, + { + "epoch": 1.66, + "grad_norm": 0.451171875, + "learning_rate": 1.3926548524270334e-05, + "loss": 2.5106, + "step": 38972 + }, + { + "epoch": 1.66, + "grad_norm": 0.41796875, + "learning_rate": 1.3923134017752593e-05, + "loss": 2.2543, + "step": 38973 + }, + { + "epoch": 1.66, + "grad_norm": 0.458984375, + "learning_rate": 1.3919719898549288e-05, + "loss": 2.485, + "step": 38974 + }, + { + "epoch": 1.66, + "grad_norm": 0.453125, + "learning_rate": 1.391630616667573e-05, + "loss": 2.3494, + "step": 38975 + }, + { + "epoch": 1.66, + "grad_norm": 0.490234375, + "learning_rate": 1.3912892822147394e-05, + "loss": 2.3534, + "step": 38976 + }, + { + "epoch": 1.66, + "grad_norm": 0.419921875, + "learning_rate": 1.3909479864979568e-05, + "loss": 2.4194, + "step": 38977 + }, + { + "epoch": 1.66, + "grad_norm": 0.490234375, + "learning_rate": 1.3906067295187598e-05, + "loss": 2.555, + "step": 38978 + }, + { + "epoch": 1.66, + "grad_norm": 0.486328125, + "learning_rate": 1.3902655112786867e-05, + "loss": 2.4366, + "step": 38979 + }, + { + "epoch": 1.66, + "grad_norm": 0.419921875, + "learning_rate": 1.3899243317792743e-05, + "loss": 2.4164, + "step": 38980 + }, + { + "epoch": 1.66, + "grad_norm": 0.49609375, + "learning_rate": 1.3895831910220536e-05, + "loss": 2.4342, + "step": 38981 + }, + { + "epoch": 1.66, + "grad_norm": 0.51953125, + "learning_rate": 1.3892420890085644e-05, + "loss": 2.472, + "step": 38982 + }, + { + "epoch": 1.66, + "grad_norm": 0.546875, + "learning_rate": 1.3889010257403356e-05, + "loss": 2.4975, + "step": 38983 + }, + { + "epoch": 1.66, + "grad_norm": 0.41796875, + "learning_rate": 1.3885600012189082e-05, + "loss": 2.4147, + "step": 38984 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.388219015445812e-05, + "loss": 2.6211, + "step": 38985 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.3878780684225844e-05, + "loss": 2.5518, + "step": 38986 + }, + { + "epoch": 1.66, + "grad_norm": 0.40625, + "learning_rate": 1.3875371601507559e-05, + "loss": 2.4306, + "step": 38987 + }, + { + "epoch": 1.66, + "grad_norm": 0.431640625, + "learning_rate": 1.3871962906318624e-05, + "loss": 2.5227, + "step": 38988 + }, + { + "epoch": 1.66, + "grad_norm": 0.3828125, + "learning_rate": 1.386855459867441e-05, + "loss": 2.4338, + "step": 38989 + }, + { + "epoch": 1.66, + "grad_norm": 0.400390625, + "learning_rate": 1.3865146678590223e-05, + "loss": 2.3404, + "step": 38990 + }, + { + "epoch": 1.66, + "grad_norm": 0.453125, + "learning_rate": 1.3861739146081343e-05, + "loss": 2.3259, + "step": 38991 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.385833200116321e-05, + "loss": 2.4131, + "step": 38992 + }, + { + "epoch": 1.66, + "grad_norm": 0.44140625, + "learning_rate": 1.3854925243851102e-05, + "loss": 2.4588, + "step": 38993 + }, + { + "epoch": 1.66, + "grad_norm": 0.5234375, + "learning_rate": 1.3851518874160318e-05, + "loss": 2.4659, + "step": 38994 + }, + { + "epoch": 1.66, + "grad_norm": 0.451171875, + "learning_rate": 1.3848112892106225e-05, + "loss": 2.3267, + "step": 38995 + }, + { + "epoch": 1.66, + "grad_norm": 0.50390625, + "learning_rate": 1.3844707297704151e-05, + "loss": 2.4841, + "step": 38996 + }, + { + "epoch": 1.66, + "grad_norm": 0.4921875, + "learning_rate": 1.3841302090969398e-05, + "loss": 2.4563, + "step": 38997 + }, + { + "epoch": 1.66, + "grad_norm": 0.412109375, + "learning_rate": 1.383789727191731e-05, + "loss": 2.5075, + "step": 38998 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.3834492840563173e-05, + "loss": 2.4672, + "step": 38999 + }, + { + "epoch": 1.66, + "grad_norm": 0.48046875, + "learning_rate": 1.3831088796922354e-05, + "loss": 2.5192, + "step": 39000 + }, + { + "epoch": 1.66, + "grad_norm": 0.490234375, + "learning_rate": 1.3827685141010127e-05, + "loss": 2.5074, + "step": 39001 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.3824281872841848e-05, + "loss": 2.4016, + "step": 39002 + }, + { + "epoch": 1.66, + "grad_norm": 0.53125, + "learning_rate": 1.3820878992432773e-05, + "loss": 2.5335, + "step": 39003 + }, + { + "epoch": 1.66, + "grad_norm": 0.455078125, + "learning_rate": 1.3817476499798254e-05, + "loss": 2.4637, + "step": 39004 + }, + { + "epoch": 1.66, + "grad_norm": 0.45703125, + "learning_rate": 1.3814074394953614e-05, + "loss": 2.4256, + "step": 39005 + }, + { + "epoch": 1.66, + "grad_norm": 0.44140625, + "learning_rate": 1.3810672677914149e-05, + "loss": 2.5201, + "step": 39006 + }, + { + "epoch": 1.66, + "grad_norm": 0.47265625, + "learning_rate": 1.3807271348695105e-05, + "loss": 2.3736, + "step": 39007 + }, + { + "epoch": 1.66, + "grad_norm": 0.462890625, + "learning_rate": 1.3803870407311891e-05, + "loss": 2.4253, + "step": 39008 + }, + { + "epoch": 1.66, + "grad_norm": 0.45703125, + "learning_rate": 1.3800469853779763e-05, + "loss": 2.5453, + "step": 39009 + }, + { + "epoch": 1.66, + "grad_norm": 0.48046875, + "learning_rate": 1.3797069688113983e-05, + "loss": 2.5914, + "step": 39010 + }, + { + "epoch": 1.66, + "grad_norm": 0.455078125, + "learning_rate": 1.3793669910329898e-05, + "loss": 2.4819, + "step": 39011 + }, + { + "epoch": 1.66, + "grad_norm": 0.49609375, + "learning_rate": 1.3790270520442817e-05, + "loss": 2.657, + "step": 39012 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.378687151846798e-05, + "loss": 2.4974, + "step": 39013 + }, + { + "epoch": 1.66, + "grad_norm": 0.4453125, + "learning_rate": 1.3783472904420735e-05, + "loss": 2.3432, + "step": 39014 + }, + { + "epoch": 1.66, + "grad_norm": 0.43359375, + "learning_rate": 1.3780074678316347e-05, + "loss": 2.3802, + "step": 39015 + }, + { + "epoch": 1.66, + "grad_norm": 0.46484375, + "learning_rate": 1.3776676840170121e-05, + "loss": 2.3547, + "step": 39016 + }, + { + "epoch": 1.66, + "grad_norm": 0.5390625, + "learning_rate": 1.3773279389997318e-05, + "loss": 2.5808, + "step": 39017 + }, + { + "epoch": 1.66, + "grad_norm": 0.421875, + "learning_rate": 1.3769882327813266e-05, + "loss": 2.3749, + "step": 39018 + }, + { + "epoch": 1.66, + "grad_norm": 0.5, + "learning_rate": 1.376648565363321e-05, + "loss": 2.3466, + "step": 39019 + }, + { + "epoch": 1.66, + "grad_norm": 0.470703125, + "learning_rate": 1.376308936747246e-05, + "loss": 2.441, + "step": 39020 + }, + { + "epoch": 1.66, + "grad_norm": 0.52734375, + "learning_rate": 1.3759693469346313e-05, + "loss": 2.4923, + "step": 39021 + }, + { + "epoch": 1.66, + "grad_norm": 0.439453125, + "learning_rate": 1.3756297959270026e-05, + "loss": 2.3912, + "step": 39022 + }, + { + "epoch": 1.66, + "grad_norm": 0.51953125, + "learning_rate": 1.375290283725883e-05, + "loss": 2.4393, + "step": 39023 + }, + { + "epoch": 1.66, + "grad_norm": 0.423828125, + "learning_rate": 1.3749508103328102e-05, + "loss": 2.405, + "step": 39024 + }, + { + "epoch": 1.66, + "grad_norm": 0.53125, + "learning_rate": 1.3746113757493061e-05, + "loss": 2.4969, + "step": 39025 + }, + { + "epoch": 1.66, + "grad_norm": 0.494140625, + "learning_rate": 1.3742719799768966e-05, + "loss": 2.3992, + "step": 39026 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.3739326230171112e-05, + "loss": 2.6187, + "step": 39027 + }, + { + "epoch": 1.66, + "grad_norm": 0.40625, + "learning_rate": 1.3735933048714788e-05, + "loss": 2.4264, + "step": 39028 + }, + { + "epoch": 1.66, + "grad_norm": 0.46875, + "learning_rate": 1.3732540255415217e-05, + "loss": 2.3737, + "step": 39029 + }, + { + "epoch": 1.66, + "grad_norm": 0.43359375, + "learning_rate": 1.3729147850287704e-05, + "loss": 2.4189, + "step": 39030 + }, + { + "epoch": 1.66, + "grad_norm": 0.453125, + "learning_rate": 1.3725755833347475e-05, + "loss": 2.3696, + "step": 39031 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.3722364204609839e-05, + "loss": 2.3581, + "step": 39032 + }, + { + "epoch": 1.66, + "grad_norm": 0.392578125, + "learning_rate": 1.3718972964090016e-05, + "loss": 2.4497, + "step": 39033 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.3715582111803304e-05, + "loss": 2.4909, + "step": 39034 + }, + { + "epoch": 1.66, + "grad_norm": 0.4453125, + "learning_rate": 1.3712191647764905e-05, + "loss": 2.5129, + "step": 39035 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.3708801571990127e-05, + "loss": 2.479, + "step": 39036 + }, + { + "epoch": 1.66, + "grad_norm": 0.44140625, + "learning_rate": 1.3705411884494212e-05, + "loss": 2.3531, + "step": 39037 + }, + { + "epoch": 1.66, + "grad_norm": 0.50390625, + "learning_rate": 1.3702022585292418e-05, + "loss": 2.4558, + "step": 39038 + }, + { + "epoch": 1.66, + "grad_norm": 0.5078125, + "learning_rate": 1.3698633674399941e-05, + "loss": 2.5498, + "step": 39039 + }, + { + "epoch": 1.66, + "grad_norm": 0.498046875, + "learning_rate": 1.3695245151832103e-05, + "loss": 2.4426, + "step": 39040 + }, + { + "epoch": 1.66, + "grad_norm": 0.4765625, + "learning_rate": 1.3691857017604137e-05, + "loss": 2.4606, + "step": 39041 + }, + { + "epoch": 1.66, + "grad_norm": 0.5, + "learning_rate": 1.368846927173124e-05, + "loss": 2.5862, + "step": 39042 + }, + { + "epoch": 1.66, + "grad_norm": 0.431640625, + "learning_rate": 1.3685081914228693e-05, + "loss": 2.2954, + "step": 39043 + }, + { + "epoch": 1.66, + "grad_norm": 0.4921875, + "learning_rate": 1.3681694945111756e-05, + "loss": 2.5925, + "step": 39044 + }, + { + "epoch": 1.66, + "grad_norm": 0.421875, + "learning_rate": 1.367830836439562e-05, + "loss": 2.4308, + "step": 39045 + }, + { + "epoch": 1.66, + "grad_norm": 0.486328125, + "learning_rate": 1.3674922172095572e-05, + "loss": 2.4078, + "step": 39046 + }, + { + "epoch": 1.66, + "grad_norm": 0.546875, + "learning_rate": 1.36715363682268e-05, + "loss": 2.4582, + "step": 39047 + }, + { + "epoch": 1.66, + "grad_norm": 0.458984375, + "learning_rate": 1.3668150952804593e-05, + "loss": 2.3911, + "step": 39048 + }, + { + "epoch": 1.66, + "grad_norm": 0.4609375, + "learning_rate": 1.3664765925844136e-05, + "loss": 2.3798, + "step": 39049 + }, + { + "epoch": 1.66, + "grad_norm": 0.484375, + "learning_rate": 1.3661381287360697e-05, + "loss": 2.38, + "step": 39050 + }, + { + "epoch": 1.66, + "grad_norm": 0.474609375, + "learning_rate": 1.3657997037369475e-05, + "loss": 2.5291, + "step": 39051 + }, + { + "epoch": 1.66, + "grad_norm": 0.51171875, + "learning_rate": 1.36546131758857e-05, + "loss": 2.5509, + "step": 39052 + }, + { + "epoch": 1.66, + "grad_norm": 0.447265625, + "learning_rate": 1.365122970292464e-05, + "loss": 2.5201, + "step": 39053 + }, + { + "epoch": 1.66, + "grad_norm": 0.50390625, + "learning_rate": 1.364784661850148e-05, + "loss": 2.3824, + "step": 39054 + }, + { + "epoch": 1.66, + "grad_norm": 0.490234375, + "learning_rate": 1.3644463922631446e-05, + "loss": 2.4996, + "step": 39055 + }, + { + "epoch": 1.66, + "grad_norm": 0.478515625, + "learning_rate": 1.3641081615329755e-05, + "loss": 2.5318, + "step": 39056 + }, + { + "epoch": 1.66, + "grad_norm": 0.494140625, + "learning_rate": 1.3637699696611662e-05, + "loss": 2.4726, + "step": 39057 + }, + { + "epoch": 1.66, + "grad_norm": 0.48828125, + "learning_rate": 1.3634318166492332e-05, + "loss": 2.432, + "step": 39058 + }, + { + "epoch": 1.66, + "grad_norm": 0.455078125, + "learning_rate": 1.3630937024987035e-05, + "loss": 2.3089, + "step": 39059 + }, + { + "epoch": 1.66, + "grad_norm": 0.45703125, + "learning_rate": 1.3627556272110931e-05, + "loss": 2.4522, + "step": 39060 + }, + { + "epoch": 1.66, + "grad_norm": 0.4375, + "learning_rate": 1.3624175907879255e-05, + "loss": 2.4816, + "step": 39061 + }, + { + "epoch": 1.66, + "grad_norm": 0.4375, + "learning_rate": 1.362079593230725e-05, + "loss": 2.4306, + "step": 39062 + }, + { + "epoch": 1.66, + "grad_norm": 0.48046875, + "learning_rate": 1.3617416345410095e-05, + "loss": 2.4271, + "step": 39063 + }, + { + "epoch": 1.66, + "grad_norm": 0.4765625, + "learning_rate": 1.3614037147202962e-05, + "loss": 2.5166, + "step": 39064 + }, + { + "epoch": 1.66, + "grad_norm": 0.494140625, + "learning_rate": 1.361065833770111e-05, + "loss": 2.5086, + "step": 39065 + }, + { + "epoch": 1.66, + "grad_norm": 0.44921875, + "learning_rate": 1.3607279916919735e-05, + "loss": 2.4308, + "step": 39066 + }, + { + "epoch": 1.66, + "grad_norm": 0.478515625, + "learning_rate": 1.3603901884874003e-05, + "loss": 2.4971, + "step": 39067 + }, + { + "epoch": 1.66, + "grad_norm": 0.47265625, + "learning_rate": 1.3600524241579148e-05, + "loss": 2.4372, + "step": 39068 + }, + { + "epoch": 1.66, + "grad_norm": 0.482421875, + "learning_rate": 1.3597146987050368e-05, + "loss": 2.3846, + "step": 39069 + }, + { + "epoch": 1.66, + "grad_norm": 0.41015625, + "learning_rate": 1.3593770121302863e-05, + "loss": 2.494, + "step": 39070 + }, + { + "epoch": 1.66, + "grad_norm": 0.470703125, + "learning_rate": 1.3590393644351784e-05, + "loss": 2.4055, + "step": 39071 + }, + { + "epoch": 1.66, + "grad_norm": 0.46875, + "learning_rate": 1.3587017556212368e-05, + "loss": 2.4969, + "step": 39072 + }, + { + "epoch": 1.67, + "grad_norm": 0.455078125, + "learning_rate": 1.3583641856899797e-05, + "loss": 2.3596, + "step": 39073 + }, + { + "epoch": 1.67, + "grad_norm": 0.4453125, + "learning_rate": 1.3580266546429243e-05, + "loss": 2.4172, + "step": 39074 + }, + { + "epoch": 1.67, + "grad_norm": 0.474609375, + "learning_rate": 1.3576891624815924e-05, + "loss": 2.4515, + "step": 39075 + }, + { + "epoch": 1.67, + "grad_norm": 0.455078125, + "learning_rate": 1.3573517092074995e-05, + "loss": 2.3791, + "step": 39076 + }, + { + "epoch": 1.67, + "grad_norm": 0.451171875, + "learning_rate": 1.3570142948221642e-05, + "loss": 2.5677, + "step": 39077 + }, + { + "epoch": 1.67, + "grad_norm": 0.46484375, + "learning_rate": 1.356676919327109e-05, + "loss": 2.4406, + "step": 39078 + }, + { + "epoch": 1.67, + "grad_norm": 0.45703125, + "learning_rate": 1.356339582723849e-05, + "loss": 2.5953, + "step": 39079 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.3560022850138998e-05, + "loss": 2.5135, + "step": 39080 + }, + { + "epoch": 1.67, + "grad_norm": 0.5078125, + "learning_rate": 1.3556650261987802e-05, + "loss": 2.4798, + "step": 39081 + }, + { + "epoch": 1.67, + "grad_norm": 0.498046875, + "learning_rate": 1.3553278062800123e-05, + "loss": 2.5391, + "step": 39082 + }, + { + "epoch": 1.67, + "grad_norm": 0.47265625, + "learning_rate": 1.3549906252591071e-05, + "loss": 2.4158, + "step": 39083 + }, + { + "epoch": 1.67, + "grad_norm": 0.408203125, + "learning_rate": 1.3546534831375846e-05, + "loss": 2.3248, + "step": 39084 + }, + { + "epoch": 1.67, + "grad_norm": 0.470703125, + "learning_rate": 1.3543163799169644e-05, + "loss": 2.4073, + "step": 39085 + }, + { + "epoch": 1.67, + "grad_norm": 0.43359375, + "learning_rate": 1.3539793155987612e-05, + "loss": 2.4711, + "step": 39086 + }, + { + "epoch": 1.67, + "grad_norm": 0.494140625, + "learning_rate": 1.3536422901844891e-05, + "loss": 2.4594, + "step": 39087 + }, + { + "epoch": 1.67, + "grad_norm": 0.46484375, + "learning_rate": 1.353305303675666e-05, + "loss": 2.3353, + "step": 39088 + }, + { + "epoch": 1.67, + "grad_norm": 0.419921875, + "learning_rate": 1.3529683560738115e-05, + "loss": 2.3515, + "step": 39089 + }, + { + "epoch": 1.67, + "grad_norm": 0.423828125, + "learning_rate": 1.352631447380438e-05, + "loss": 2.3765, + "step": 39090 + }, + { + "epoch": 1.67, + "grad_norm": 0.5, + "learning_rate": 1.352294577597064e-05, + "loss": 2.5427, + "step": 39091 + }, + { + "epoch": 1.67, + "grad_norm": 0.421875, + "learning_rate": 1.3519577467252021e-05, + "loss": 2.3797, + "step": 39092 + }, + { + "epoch": 1.67, + "grad_norm": 0.44921875, + "learning_rate": 1.3516209547663695e-05, + "loss": 2.4114, + "step": 39093 + }, + { + "epoch": 1.67, + "grad_norm": 0.39453125, + "learning_rate": 1.3512842017220851e-05, + "loss": 2.3738, + "step": 39094 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.35094748759386e-05, + "loss": 2.5542, + "step": 39095 + }, + { + "epoch": 1.67, + "grad_norm": 0.4375, + "learning_rate": 1.3506108123832085e-05, + "loss": 2.5107, + "step": 39096 + }, + { + "epoch": 1.67, + "grad_norm": 0.447265625, + "learning_rate": 1.350274176091647e-05, + "loss": 2.3899, + "step": 39097 + }, + { + "epoch": 1.67, + "grad_norm": 0.482421875, + "learning_rate": 1.3499375787206935e-05, + "loss": 2.4169, + "step": 39098 + }, + { + "epoch": 1.67, + "grad_norm": 0.423828125, + "learning_rate": 1.3496010202718579e-05, + "loss": 2.4886, + "step": 39099 + }, + { + "epoch": 1.67, + "grad_norm": 0.4609375, + "learning_rate": 1.3492645007466553e-05, + "loss": 2.4153, + "step": 39100 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.3489280201466036e-05, + "loss": 2.4569, + "step": 39101 + }, + { + "epoch": 1.67, + "grad_norm": 0.5, + "learning_rate": 1.348591578473214e-05, + "loss": 2.4117, + "step": 39102 + }, + { + "epoch": 1.67, + "grad_norm": 0.47265625, + "learning_rate": 1.3482551757279982e-05, + "loss": 2.4338, + "step": 39103 + }, + { + "epoch": 1.67, + "grad_norm": 0.447265625, + "learning_rate": 1.347918811912473e-05, + "loss": 2.4598, + "step": 39104 + }, + { + "epoch": 1.67, + "grad_norm": 0.408203125, + "learning_rate": 1.3475824870281528e-05, + "loss": 2.4429, + "step": 39105 + }, + { + "epoch": 1.67, + "grad_norm": 0.515625, + "learning_rate": 1.3472462010765463e-05, + "loss": 2.4839, + "step": 39106 + }, + { + "epoch": 1.67, + "grad_norm": 0.447265625, + "learning_rate": 1.3469099540591735e-05, + "loss": 2.3263, + "step": 39107 + }, + { + "epoch": 1.67, + "grad_norm": 0.5078125, + "learning_rate": 1.3465737459775406e-05, + "loss": 2.4003, + "step": 39108 + }, + { + "epoch": 1.67, + "grad_norm": 0.3984375, + "learning_rate": 1.3462375768331637e-05, + "loss": 2.3695, + "step": 39109 + }, + { + "epoch": 1.67, + "grad_norm": 0.451171875, + "learning_rate": 1.3459014466275565e-05, + "loss": 2.4839, + "step": 39110 + }, + { + "epoch": 1.67, + "grad_norm": 0.478515625, + "learning_rate": 1.3455653553622305e-05, + "loss": 2.5201, + "step": 39111 + }, + { + "epoch": 1.67, + "grad_norm": 0.4296875, + "learning_rate": 1.3452293030386954e-05, + "loss": 2.4791, + "step": 39112 + }, + { + "epoch": 1.67, + "grad_norm": 0.48828125, + "learning_rate": 1.3448932896584654e-05, + "loss": 2.3798, + "step": 39113 + }, + { + "epoch": 1.67, + "grad_norm": 0.447265625, + "learning_rate": 1.344557315223055e-05, + "loss": 2.4472, + "step": 39114 + }, + { + "epoch": 1.67, + "grad_norm": 0.408203125, + "learning_rate": 1.344221379733972e-05, + "loss": 2.4401, + "step": 39115 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.3438854831927284e-05, + "loss": 2.3472, + "step": 39116 + }, + { + "epoch": 1.67, + "grad_norm": 0.419921875, + "learning_rate": 1.3435496256008384e-05, + "loss": 2.538, + "step": 39117 + }, + { + "epoch": 1.67, + "grad_norm": 0.484375, + "learning_rate": 1.3432138069598133e-05, + "loss": 2.4882, + "step": 39118 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.3428780272711583e-05, + "loss": 2.4695, + "step": 39119 + }, + { + "epoch": 1.67, + "grad_norm": 0.466796875, + "learning_rate": 1.3425422865363891e-05, + "loss": 2.4725, + "step": 39120 + }, + { + "epoch": 1.67, + "grad_norm": 0.54296875, + "learning_rate": 1.3422065847570187e-05, + "loss": 2.4119, + "step": 39121 + }, + { + "epoch": 1.67, + "grad_norm": 0.404296875, + "learning_rate": 1.3418709219345516e-05, + "loss": 2.4107, + "step": 39122 + }, + { + "epoch": 1.67, + "grad_norm": 0.4453125, + "learning_rate": 1.3415352980705032e-05, + "loss": 2.4111, + "step": 39123 + }, + { + "epoch": 1.67, + "grad_norm": 0.486328125, + "learning_rate": 1.3411997131663789e-05, + "loss": 2.4412, + "step": 39124 + }, + { + "epoch": 1.67, + "grad_norm": 0.453125, + "learning_rate": 1.340864167223692e-05, + "loss": 2.4805, + "step": 39125 + }, + { + "epoch": 1.67, + "grad_norm": 0.43359375, + "learning_rate": 1.3405286602439548e-05, + "loss": 2.4784, + "step": 39126 + }, + { + "epoch": 1.67, + "grad_norm": 0.455078125, + "learning_rate": 1.3401931922286737e-05, + "loss": 2.4881, + "step": 39127 + }, + { + "epoch": 1.67, + "grad_norm": 0.4296875, + "learning_rate": 1.3398577631793552e-05, + "loss": 2.5995, + "step": 39128 + }, + { + "epoch": 1.67, + "grad_norm": 0.451171875, + "learning_rate": 1.3395223730975127e-05, + "loss": 2.583, + "step": 39129 + }, + { + "epoch": 1.67, + "grad_norm": 0.5, + "learning_rate": 1.3391870219846569e-05, + "loss": 2.3231, + "step": 39130 + }, + { + "epoch": 1.67, + "grad_norm": 0.466796875, + "learning_rate": 1.3388517098422915e-05, + "loss": 2.6152, + "step": 39131 + }, + { + "epoch": 1.67, + "grad_norm": 0.4296875, + "learning_rate": 1.3385164366719282e-05, + "loss": 2.3778, + "step": 39132 + }, + { + "epoch": 1.67, + "grad_norm": 0.53515625, + "learning_rate": 1.3381812024750785e-05, + "loss": 2.4126, + "step": 39133 + }, + { + "epoch": 1.67, + "grad_norm": 0.4296875, + "learning_rate": 1.3378460072532473e-05, + "loss": 2.4061, + "step": 39134 + }, + { + "epoch": 1.67, + "grad_norm": 0.5234375, + "learning_rate": 1.3375108510079404e-05, + "loss": 2.6251, + "step": 39135 + }, + { + "epoch": 1.67, + "grad_norm": 0.494140625, + "learning_rate": 1.3371757337406698e-05, + "loss": 2.5142, + "step": 39136 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.3368406554529445e-05, + "loss": 2.3883, + "step": 39137 + }, + { + "epoch": 1.67, + "grad_norm": 0.56640625, + "learning_rate": 1.3365056161462686e-05, + "loss": 2.4098, + "step": 39138 + }, + { + "epoch": 1.67, + "grad_norm": 0.5, + "learning_rate": 1.3361706158221532e-05, + "loss": 2.3724, + "step": 39139 + }, + { + "epoch": 1.67, + "grad_norm": 0.427734375, + "learning_rate": 1.3358356544821004e-05, + "loss": 2.3103, + "step": 39140 + }, + { + "epoch": 1.67, + "grad_norm": 0.484375, + "learning_rate": 1.3355007321276225e-05, + "loss": 2.5628, + "step": 39141 + }, + { + "epoch": 1.67, + "grad_norm": 0.44921875, + "learning_rate": 1.335165848760226e-05, + "loss": 2.4288, + "step": 39142 + }, + { + "epoch": 1.67, + "grad_norm": 0.482421875, + "learning_rate": 1.3348310043814171e-05, + "loss": 2.4675, + "step": 39143 + }, + { + "epoch": 1.67, + "grad_norm": 0.51171875, + "learning_rate": 1.3344961989926985e-05, + "loss": 2.5203, + "step": 39144 + }, + { + "epoch": 1.67, + "grad_norm": 0.388671875, + "learning_rate": 1.334161432595582e-05, + "loss": 2.3562, + "step": 39145 + }, + { + "epoch": 1.67, + "grad_norm": 0.4921875, + "learning_rate": 1.3338267051915732e-05, + "loss": 2.5109, + "step": 39146 + }, + { + "epoch": 1.67, + "grad_norm": 0.498046875, + "learning_rate": 1.3334920167821741e-05, + "loss": 2.503, + "step": 39147 + }, + { + "epoch": 1.67, + "grad_norm": 0.52734375, + "learning_rate": 1.3331573673688946e-05, + "loss": 2.4511, + "step": 39148 + }, + { + "epoch": 1.67, + "grad_norm": 0.447265625, + "learning_rate": 1.3328227569532414e-05, + "loss": 2.4886, + "step": 39149 + }, + { + "epoch": 1.67, + "grad_norm": 0.4921875, + "learning_rate": 1.3324881855367189e-05, + "loss": 2.5481, + "step": 39150 + }, + { + "epoch": 1.67, + "grad_norm": 0.474609375, + "learning_rate": 1.3321536531208301e-05, + "loss": 2.4814, + "step": 39151 + }, + { + "epoch": 1.67, + "grad_norm": 0.453125, + "learning_rate": 1.3318191597070807e-05, + "loss": 2.5056, + "step": 39152 + }, + { + "epoch": 1.67, + "grad_norm": 0.45703125, + "learning_rate": 1.3314847052969804e-05, + "loss": 2.3761, + "step": 39153 + }, + { + "epoch": 1.67, + "grad_norm": 0.42578125, + "learning_rate": 1.3311502898920292e-05, + "loss": 2.4648, + "step": 39154 + }, + { + "epoch": 1.67, + "grad_norm": 0.439453125, + "learning_rate": 1.3308159134937348e-05, + "loss": 2.3618, + "step": 39155 + }, + { + "epoch": 1.67, + "grad_norm": 0.451171875, + "learning_rate": 1.3304815761035982e-05, + "loss": 2.4391, + "step": 39156 + }, + { + "epoch": 1.67, + "grad_norm": 0.455078125, + "learning_rate": 1.3301472777231272e-05, + "loss": 2.3443, + "step": 39157 + }, + { + "epoch": 1.67, + "grad_norm": 0.42578125, + "learning_rate": 1.329813018353826e-05, + "loss": 2.4071, + "step": 39158 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.329478797997199e-05, + "loss": 2.3477, + "step": 39159 + }, + { + "epoch": 1.67, + "grad_norm": 0.421875, + "learning_rate": 1.329144616654746e-05, + "loss": 2.3611, + "step": 39160 + }, + { + "epoch": 1.67, + "grad_norm": 0.453125, + "learning_rate": 1.3288104743279727e-05, + "loss": 2.4096, + "step": 39161 + }, + { + "epoch": 1.67, + "grad_norm": 0.498046875, + "learning_rate": 1.3284763710183855e-05, + "loss": 2.5263, + "step": 39162 + }, + { + "epoch": 1.67, + "grad_norm": 0.380859375, + "learning_rate": 1.3281423067274846e-05, + "loss": 2.4248, + "step": 39163 + }, + { + "epoch": 1.67, + "grad_norm": 0.421875, + "learning_rate": 1.3278082814567728e-05, + "loss": 2.5536, + "step": 39164 + }, + { + "epoch": 1.67, + "grad_norm": 0.46875, + "learning_rate": 1.3274742952077579e-05, + "loss": 2.3743, + "step": 39165 + }, + { + "epoch": 1.67, + "grad_norm": 0.435546875, + "learning_rate": 1.3271403479819378e-05, + "loss": 2.4449, + "step": 39166 + }, + { + "epoch": 1.67, + "grad_norm": 0.4453125, + "learning_rate": 1.3268064397808155e-05, + "loss": 2.5181, + "step": 39167 + }, + { + "epoch": 1.67, + "grad_norm": 0.419921875, + "learning_rate": 1.3264725706058933e-05, + "loss": 2.3211, + "step": 39168 + }, + { + "epoch": 1.67, + "grad_norm": 0.435546875, + "learning_rate": 1.3261387404586777e-05, + "loss": 2.5328, + "step": 39169 + }, + { + "epoch": 1.67, + "grad_norm": 0.43359375, + "learning_rate": 1.3258049493406666e-05, + "loss": 2.3801, + "step": 39170 + }, + { + "epoch": 1.67, + "grad_norm": 0.439453125, + "learning_rate": 1.325471197253364e-05, + "loss": 2.5069, + "step": 39171 + }, + { + "epoch": 1.67, + "grad_norm": 0.455078125, + "learning_rate": 1.325137484198271e-05, + "loss": 2.3827, + "step": 39172 + }, + { + "epoch": 1.67, + "grad_norm": 0.51171875, + "learning_rate": 1.3248038101768845e-05, + "loss": 2.4872, + "step": 39173 + }, + { + "epoch": 1.67, + "grad_norm": 0.458984375, + "learning_rate": 1.324470175190715e-05, + "loss": 2.5072, + "step": 39174 + }, + { + "epoch": 1.67, + "grad_norm": 0.44921875, + "learning_rate": 1.3241365792412596e-05, + "loss": 2.5063, + "step": 39175 + }, + { + "epoch": 1.67, + "grad_norm": 0.4140625, + "learning_rate": 1.3238030223300168e-05, + "loss": 2.3298, + "step": 39176 + }, + { + "epoch": 1.67, + "grad_norm": 0.515625, + "learning_rate": 1.3234695044584899e-05, + "loss": 2.4802, + "step": 39177 + }, + { + "epoch": 1.67, + "grad_norm": 0.453125, + "learning_rate": 1.32313602562818e-05, + "loss": 2.3515, + "step": 39178 + }, + { + "epoch": 1.67, + "grad_norm": 0.474609375, + "learning_rate": 1.3228025858405856e-05, + "loss": 2.4817, + "step": 39179 + }, + { + "epoch": 1.67, + "grad_norm": 0.466796875, + "learning_rate": 1.3224691850972116e-05, + "loss": 2.4624, + "step": 39180 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.3221358233995518e-05, + "loss": 2.4328, + "step": 39181 + }, + { + "epoch": 1.67, + "grad_norm": 0.515625, + "learning_rate": 1.3218025007491108e-05, + "loss": 2.5043, + "step": 39182 + }, + { + "epoch": 1.67, + "grad_norm": 0.46875, + "learning_rate": 1.3214692171473864e-05, + "loss": 2.6194, + "step": 39183 + }, + { + "epoch": 1.67, + "grad_norm": 0.49609375, + "learning_rate": 1.3211359725958805e-05, + "loss": 2.3776, + "step": 39184 + }, + { + "epoch": 1.67, + "grad_norm": 0.478515625, + "learning_rate": 1.3208027670960888e-05, + "loss": 2.4881, + "step": 39185 + }, + { + "epoch": 1.67, + "grad_norm": 0.498046875, + "learning_rate": 1.3204696006495121e-05, + "loss": 2.5314, + "step": 39186 + }, + { + "epoch": 1.67, + "grad_norm": 0.470703125, + "learning_rate": 1.3201364732576526e-05, + "loss": 2.4707, + "step": 39187 + }, + { + "epoch": 1.67, + "grad_norm": 0.470703125, + "learning_rate": 1.3198033849220071e-05, + "loss": 2.4828, + "step": 39188 + }, + { + "epoch": 1.67, + "grad_norm": 0.455078125, + "learning_rate": 1.3194703356440696e-05, + "loss": 2.4458, + "step": 39189 + }, + { + "epoch": 1.67, + "grad_norm": 0.451171875, + "learning_rate": 1.319137325425348e-05, + "loss": 2.3384, + "step": 39190 + }, + { + "epoch": 1.67, + "grad_norm": 0.494140625, + "learning_rate": 1.3188043542673356e-05, + "loss": 2.5152, + "step": 39191 + }, + { + "epoch": 1.67, + "grad_norm": 0.490234375, + "learning_rate": 1.3184714221715288e-05, + "loss": 2.4132, + "step": 39192 + }, + { + "epoch": 1.67, + "grad_norm": 0.4765625, + "learning_rate": 1.3181385291394289e-05, + "loss": 2.5799, + "step": 39193 + }, + { + "epoch": 1.67, + "grad_norm": 0.4609375, + "learning_rate": 1.3178056751725332e-05, + "loss": 2.4075, + "step": 39194 + }, + { + "epoch": 1.67, + "grad_norm": 0.45703125, + "learning_rate": 1.3174728602723384e-05, + "loss": 2.3459, + "step": 39195 + }, + { + "epoch": 1.67, + "grad_norm": 0.5078125, + "learning_rate": 1.3171400844403436e-05, + "loss": 2.653, + "step": 39196 + }, + { + "epoch": 1.67, + "grad_norm": 0.5, + "learning_rate": 1.3168073476780441e-05, + "loss": 2.4987, + "step": 39197 + }, + { + "epoch": 1.67, + "grad_norm": 0.4609375, + "learning_rate": 1.3164746499869396e-05, + "loss": 2.5046, + "step": 39198 + }, + { + "epoch": 1.67, + "grad_norm": 0.42578125, + "learning_rate": 1.3161419913685236e-05, + "loss": 2.3315, + "step": 39199 + }, + { + "epoch": 1.67, + "grad_norm": 0.48828125, + "learning_rate": 1.315809371824297e-05, + "loss": 2.3664, + "step": 39200 + }, + { + "epoch": 1.67, + "grad_norm": 0.431640625, + "learning_rate": 1.315476791355753e-05, + "loss": 2.5281, + "step": 39201 + }, + { + "epoch": 1.67, + "grad_norm": 0.4765625, + "learning_rate": 1.3151442499643884e-05, + "loss": 2.5083, + "step": 39202 + }, + { + "epoch": 1.67, + "grad_norm": 0.482421875, + "learning_rate": 1.3148117476517041e-05, + "loss": 2.5415, + "step": 39203 + }, + { + "epoch": 1.67, + "grad_norm": 0.486328125, + "learning_rate": 1.3144792844191922e-05, + "loss": 2.5282, + "step": 39204 + }, + { + "epoch": 1.67, + "grad_norm": 0.41796875, + "learning_rate": 1.314146860268346e-05, + "loss": 2.4189, + "step": 39205 + }, + { + "epoch": 1.67, + "grad_norm": 0.478515625, + "learning_rate": 1.3138144752006676e-05, + "loss": 2.5564, + "step": 39206 + }, + { + "epoch": 1.67, + "grad_norm": 0.48828125, + "learning_rate": 1.3134821292176503e-05, + "loss": 2.4384, + "step": 39207 + }, + { + "epoch": 1.67, + "grad_norm": 0.490234375, + "learning_rate": 1.3131498223207872e-05, + "loss": 2.3603, + "step": 39208 + }, + { + "epoch": 1.67, + "grad_norm": 0.421875, + "learning_rate": 1.312817554511574e-05, + "loss": 2.4732, + "step": 39209 + }, + { + "epoch": 1.67, + "grad_norm": 0.43359375, + "learning_rate": 1.3124853257915104e-05, + "loss": 2.4372, + "step": 39210 + }, + { + "epoch": 1.67, + "grad_norm": 0.423828125, + "learning_rate": 1.3121531361620853e-05, + "loss": 2.4908, + "step": 39211 + }, + { + "epoch": 1.67, + "grad_norm": 0.447265625, + "learning_rate": 1.3118209856247977e-05, + "loss": 2.39, + "step": 39212 + }, + { + "epoch": 1.67, + "grad_norm": 0.5234375, + "learning_rate": 1.3114888741811393e-05, + "loss": 2.4504, + "step": 39213 + }, + { + "epoch": 1.67, + "grad_norm": 0.4296875, + "learning_rate": 1.311156801832607e-05, + "loss": 2.5293, + "step": 39214 + }, + { + "epoch": 1.67, + "grad_norm": 0.494140625, + "learning_rate": 1.3108247685806919e-05, + "loss": 2.5686, + "step": 39215 + }, + { + "epoch": 1.67, + "grad_norm": 0.44921875, + "learning_rate": 1.3104927744268914e-05, + "loss": 2.5114, + "step": 39216 + }, + { + "epoch": 1.67, + "grad_norm": 0.41796875, + "learning_rate": 1.3101608193726966e-05, + "loss": 2.2935, + "step": 39217 + }, + { + "epoch": 1.67, + "grad_norm": 0.447265625, + "learning_rate": 1.3098289034196009e-05, + "loss": 2.5168, + "step": 39218 + }, + { + "epoch": 1.67, + "grad_norm": 0.439453125, + "learning_rate": 1.309497026569103e-05, + "loss": 2.5457, + "step": 39219 + }, + { + "epoch": 1.67, + "grad_norm": 0.435546875, + "learning_rate": 1.3091651888226919e-05, + "loss": 2.465, + "step": 39220 + }, + { + "epoch": 1.67, + "grad_norm": 0.4453125, + "learning_rate": 1.308833390181856e-05, + "loss": 2.5806, + "step": 39221 + }, + { + "epoch": 1.67, + "grad_norm": 0.46875, + "learning_rate": 1.308501630648099e-05, + "loss": 2.5531, + "step": 39222 + }, + { + "epoch": 1.67, + "grad_norm": 0.42578125, + "learning_rate": 1.3081699102229083e-05, + "loss": 2.3249, + "step": 39223 + }, + { + "epoch": 1.67, + "grad_norm": 0.41796875, + "learning_rate": 1.307838228907774e-05, + "loss": 2.514, + "step": 39224 + }, + { + "epoch": 1.67, + "grad_norm": 0.49609375, + "learning_rate": 1.3075065867041914e-05, + "loss": 2.4721, + "step": 39225 + }, + { + "epoch": 1.67, + "grad_norm": 0.498046875, + "learning_rate": 1.3071749836136537e-05, + "loss": 2.3793, + "step": 39226 + }, + { + "epoch": 1.67, + "grad_norm": 0.52734375, + "learning_rate": 1.3068434196376511e-05, + "loss": 2.5345, + "step": 39227 + }, + { + "epoch": 1.67, + "grad_norm": 0.455078125, + "learning_rate": 1.3065118947776766e-05, + "loss": 2.4343, + "step": 39228 + }, + { + "epoch": 1.67, + "grad_norm": 0.451171875, + "learning_rate": 1.3061804090352203e-05, + "loss": 2.4055, + "step": 39229 + }, + { + "epoch": 1.67, + "grad_norm": 0.46484375, + "learning_rate": 1.3058489624117764e-05, + "loss": 2.4641, + "step": 39230 + }, + { + "epoch": 1.67, + "grad_norm": 0.470703125, + "learning_rate": 1.3055175549088339e-05, + "loss": 2.3141, + "step": 39231 + }, + { + "epoch": 1.67, + "grad_norm": 0.44140625, + "learning_rate": 1.3051861865278858e-05, + "loss": 2.4479, + "step": 39232 + }, + { + "epoch": 1.67, + "grad_norm": 0.4921875, + "learning_rate": 1.304854857270421e-05, + "loss": 2.6325, + "step": 39233 + }, + { + "epoch": 1.67, + "grad_norm": 0.408203125, + "learning_rate": 1.3045235671379308e-05, + "loss": 2.38, + "step": 39234 + }, + { + "epoch": 1.67, + "grad_norm": 0.5234375, + "learning_rate": 1.3041923161319091e-05, + "loss": 2.4531, + "step": 39235 + }, + { + "epoch": 1.67, + "grad_norm": 0.4296875, + "learning_rate": 1.3038611042538451e-05, + "loss": 2.4693, + "step": 39236 + }, + { + "epoch": 1.67, + "grad_norm": 0.462890625, + "learning_rate": 1.303529931505223e-05, + "loss": 2.3871, + "step": 39237 + }, + { + "epoch": 1.67, + "grad_norm": 0.482421875, + "learning_rate": 1.3031987978875415e-05, + "loss": 2.5385, + "step": 39238 + }, + { + "epoch": 1.67, + "grad_norm": 0.447265625, + "learning_rate": 1.3028677034022884e-05, + "loss": 2.5232, + "step": 39239 + }, + { + "epoch": 1.67, + "grad_norm": 0.453125, + "learning_rate": 1.3025366480509493e-05, + "loss": 2.3796, + "step": 39240 + }, + { + "epoch": 1.67, + "grad_norm": 0.4921875, + "learning_rate": 1.3022056318350162e-05, + "loss": 2.5763, + "step": 39241 + }, + { + "epoch": 1.67, + "grad_norm": 0.47265625, + "learning_rate": 1.3018746547559824e-05, + "loss": 2.4222, + "step": 39242 + }, + { + "epoch": 1.67, + "grad_norm": 0.4375, + "learning_rate": 1.301543716815331e-05, + "loss": 2.4217, + "step": 39243 + }, + { + "epoch": 1.67, + "grad_norm": 0.419921875, + "learning_rate": 1.3012128180145566e-05, + "loss": 2.3088, + "step": 39244 + }, + { + "epoch": 1.67, + "grad_norm": 0.4921875, + "learning_rate": 1.3008819583551435e-05, + "loss": 2.3271, + "step": 39245 + }, + { + "epoch": 1.67, + "grad_norm": 0.44921875, + "learning_rate": 1.3005511378385838e-05, + "loss": 2.5395, + "step": 39246 + }, + { + "epoch": 1.67, + "grad_norm": 0.462890625, + "learning_rate": 1.3002203564663629e-05, + "loss": 2.3825, + "step": 39247 + }, + { + "epoch": 1.67, + "grad_norm": 0.5078125, + "learning_rate": 1.2998896142399731e-05, + "loss": 2.4051, + "step": 39248 + }, + { + "epoch": 1.67, + "grad_norm": 0.5546875, + "learning_rate": 1.2995589111608986e-05, + "loss": 2.449, + "step": 39249 + }, + { + "epoch": 1.67, + "grad_norm": 0.51953125, + "learning_rate": 1.2992282472306283e-05, + "loss": 2.4823, + "step": 39250 + }, + { + "epoch": 1.67, + "grad_norm": 0.462890625, + "learning_rate": 1.2988976224506543e-05, + "loss": 2.5259, + "step": 39251 + }, + { + "epoch": 1.67, + "grad_norm": 0.51171875, + "learning_rate": 1.2985670368224612e-05, + "loss": 2.6198, + "step": 39252 + }, + { + "epoch": 1.67, + "grad_norm": 0.48046875, + "learning_rate": 1.298236490347532e-05, + "loss": 2.4571, + "step": 39253 + }, + { + "epoch": 1.67, + "grad_norm": 0.453125, + "learning_rate": 1.2979059830273621e-05, + "loss": 2.5049, + "step": 39254 + }, + { + "epoch": 1.67, + "grad_norm": 0.40234375, + "learning_rate": 1.297575514863435e-05, + "loss": 2.5022, + "step": 39255 + }, + { + "epoch": 1.67, + "grad_norm": 0.490234375, + "learning_rate": 1.297245085857236e-05, + "loss": 2.4816, + "step": 39256 + }, + { + "epoch": 1.67, + "grad_norm": 0.50390625, + "learning_rate": 1.296914696010254e-05, + "loss": 2.4847, + "step": 39257 + }, + { + "epoch": 1.67, + "grad_norm": 0.4609375, + "learning_rate": 1.2965843453239767e-05, + "loss": 2.42, + "step": 39258 + }, + { + "epoch": 1.67, + "grad_norm": 0.42578125, + "learning_rate": 1.2962540337998875e-05, + "loss": 2.4738, + "step": 39259 + }, + { + "epoch": 1.67, + "grad_norm": 0.5234375, + "learning_rate": 1.295923761439477e-05, + "loss": 2.5098, + "step": 39260 + }, + { + "epoch": 1.67, + "grad_norm": 0.5, + "learning_rate": 1.2955935282442255e-05, + "loss": 2.4124, + "step": 39261 + }, + { + "epoch": 1.67, + "grad_norm": 0.470703125, + "learning_rate": 1.2952633342156217e-05, + "loss": 2.511, + "step": 39262 + }, + { + "epoch": 1.67, + "grad_norm": 0.421875, + "learning_rate": 1.2949331793551544e-05, + "loss": 2.3364, + "step": 39263 + }, + { + "epoch": 1.67, + "grad_norm": 0.46875, + "learning_rate": 1.294603063664307e-05, + "loss": 2.3681, + "step": 39264 + }, + { + "epoch": 1.67, + "grad_norm": 0.4609375, + "learning_rate": 1.2942729871445614e-05, + "loss": 2.3743, + "step": 39265 + }, + { + "epoch": 1.67, + "grad_norm": 0.412109375, + "learning_rate": 1.2939429497974064e-05, + "loss": 2.3842, + "step": 39266 + }, + { + "epoch": 1.67, + "grad_norm": 0.5078125, + "learning_rate": 1.2936129516243278e-05, + "loss": 2.5696, + "step": 39267 + }, + { + "epoch": 1.67, + "grad_norm": 0.466796875, + "learning_rate": 1.2932829926268097e-05, + "loss": 2.3791, + "step": 39268 + }, + { + "epoch": 1.67, + "grad_norm": 0.515625, + "learning_rate": 1.292953072806331e-05, + "loss": 2.425, + "step": 39269 + }, + { + "epoch": 1.67, + "grad_norm": 0.482421875, + "learning_rate": 1.2926231921643871e-05, + "loss": 2.5519, + "step": 39270 + }, + { + "epoch": 1.67, + "grad_norm": 0.5078125, + "learning_rate": 1.2922933507024559e-05, + "loss": 2.5669, + "step": 39271 + }, + { + "epoch": 1.67, + "grad_norm": 0.484375, + "learning_rate": 1.2919635484220204e-05, + "loss": 2.6038, + "step": 39272 + }, + { + "epoch": 1.67, + "grad_norm": 0.451171875, + "learning_rate": 1.2916337853245664e-05, + "loss": 2.4242, + "step": 39273 + }, + { + "epoch": 1.67, + "grad_norm": 0.4765625, + "learning_rate": 1.2913040614115801e-05, + "loss": 2.4437, + "step": 39274 + }, + { + "epoch": 1.67, + "grad_norm": 0.48828125, + "learning_rate": 1.2909743766845394e-05, + "loss": 2.4574, + "step": 39275 + }, + { + "epoch": 1.67, + "grad_norm": 0.498046875, + "learning_rate": 1.2906447311449355e-05, + "loss": 2.5299, + "step": 39276 + }, + { + "epoch": 1.67, + "grad_norm": 0.462890625, + "learning_rate": 1.2903151247942435e-05, + "loss": 2.3841, + "step": 39277 + }, + { + "epoch": 1.67, + "grad_norm": 0.443359375, + "learning_rate": 1.2899855576339515e-05, + "loss": 2.3748, + "step": 39278 + }, + { + "epoch": 1.67, + "grad_norm": 0.5, + "learning_rate": 1.2896560296655436e-05, + "loss": 2.5078, + "step": 39279 + }, + { + "epoch": 1.67, + "grad_norm": 0.45703125, + "learning_rate": 1.2893265408904998e-05, + "loss": 2.5211, + "step": 39280 + }, + { + "epoch": 1.67, + "grad_norm": 0.421875, + "learning_rate": 1.2889970913103011e-05, + "loss": 2.4944, + "step": 39281 + }, + { + "epoch": 1.67, + "grad_norm": 0.490234375, + "learning_rate": 1.288667680926432e-05, + "loss": 2.4489, + "step": 39282 + }, + { + "epoch": 1.67, + "grad_norm": 0.470703125, + "learning_rate": 1.2883383097403767e-05, + "loss": 2.3935, + "step": 39283 + }, + { + "epoch": 1.67, + "grad_norm": 0.4609375, + "learning_rate": 1.2880089777536153e-05, + "loss": 2.4137, + "step": 39284 + }, + { + "epoch": 1.67, + "grad_norm": 0.478515625, + "learning_rate": 1.2876796849676253e-05, + "loss": 2.3689, + "step": 39285 + }, + { + "epoch": 1.67, + "grad_norm": 0.44921875, + "learning_rate": 1.2873504313838969e-05, + "loss": 2.321, + "step": 39286 + }, + { + "epoch": 1.67, + "grad_norm": 0.44921875, + "learning_rate": 1.2870212170039076e-05, + "loss": 2.4179, + "step": 39287 + }, + { + "epoch": 1.67, + "grad_norm": 0.47265625, + "learning_rate": 1.2866920418291361e-05, + "loss": 2.5603, + "step": 39288 + }, + { + "epoch": 1.67, + "grad_norm": 0.45703125, + "learning_rate": 1.2863629058610671e-05, + "loss": 2.3344, + "step": 39289 + }, + { + "epoch": 1.67, + "grad_norm": 0.4453125, + "learning_rate": 1.2860338091011815e-05, + "loss": 2.2978, + "step": 39290 + }, + { + "epoch": 1.67, + "grad_norm": 0.50390625, + "learning_rate": 1.2857047515509579e-05, + "loss": 2.511, + "step": 39291 + }, + { + "epoch": 1.67, + "grad_norm": 0.40234375, + "learning_rate": 1.28537573321188e-05, + "loss": 2.3772, + "step": 39292 + }, + { + "epoch": 1.67, + "grad_norm": 0.46484375, + "learning_rate": 1.2850467540854272e-05, + "loss": 2.4038, + "step": 39293 + }, + { + "epoch": 1.67, + "grad_norm": 0.408203125, + "learning_rate": 1.2847178141730731e-05, + "loss": 2.4886, + "step": 39294 + }, + { + "epoch": 1.67, + "grad_norm": 0.396484375, + "learning_rate": 1.2843889134763098e-05, + "loss": 2.5394, + "step": 39295 + }, + { + "epoch": 1.67, + "grad_norm": 0.53515625, + "learning_rate": 1.2840600519966107e-05, + "loss": 2.3901, + "step": 39296 + }, + { + "epoch": 1.67, + "grad_norm": 0.474609375, + "learning_rate": 1.283731229735453e-05, + "loss": 2.4935, + "step": 39297 + }, + { + "epoch": 1.67, + "grad_norm": 0.443359375, + "learning_rate": 1.2834024466943206e-05, + "loss": 2.3871, + "step": 39298 + }, + { + "epoch": 1.67, + "grad_norm": 0.4375, + "learning_rate": 1.2830737028746931e-05, + "loss": 2.5112, + "step": 39299 + }, + { + "epoch": 1.67, + "grad_norm": 0.421875, + "learning_rate": 1.2827449982780481e-05, + "loss": 2.4468, + "step": 39300 + }, + { + "epoch": 1.67, + "grad_norm": 0.435546875, + "learning_rate": 1.2824163329058624e-05, + "loss": 2.3731, + "step": 39301 + }, + { + "epoch": 1.67, + "grad_norm": 0.50390625, + "learning_rate": 1.2820877067596182e-05, + "loss": 2.3547, + "step": 39302 + }, + { + "epoch": 1.67, + "grad_norm": 0.431640625, + "learning_rate": 1.2817591198407952e-05, + "loss": 2.399, + "step": 39303 + }, + { + "epoch": 1.67, + "grad_norm": 0.439453125, + "learning_rate": 1.2814305721508679e-05, + "loss": 2.315, + "step": 39304 + }, + { + "epoch": 1.67, + "grad_norm": 0.47265625, + "learning_rate": 1.2811020636913185e-05, + "loss": 2.3911, + "step": 39305 + }, + { + "epoch": 1.67, + "grad_norm": 0.4375, + "learning_rate": 1.2807735944636212e-05, + "loss": 2.6099, + "step": 39306 + }, + { + "epoch": 1.67, + "grad_norm": 0.462890625, + "learning_rate": 1.2804451644692562e-05, + "loss": 2.4075, + "step": 39307 + }, + { + "epoch": 1.68, + "grad_norm": 0.44140625, + "learning_rate": 1.2801167737097043e-05, + "loss": 2.46, + "step": 39308 + }, + { + "epoch": 1.68, + "grad_norm": 0.455078125, + "learning_rate": 1.2797884221864398e-05, + "loss": 2.48, + "step": 39309 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.2794601099009351e-05, + "loss": 2.4569, + "step": 39310 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.279131836854679e-05, + "loss": 2.5, + "step": 39311 + }, + { + "epoch": 1.68, + "grad_norm": 0.46484375, + "learning_rate": 1.2788036030491413e-05, + "loss": 2.4223, + "step": 39312 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2784754084857986e-05, + "loss": 2.39, + "step": 39313 + }, + { + "epoch": 1.68, + "grad_norm": 0.431640625, + "learning_rate": 1.2781472531661288e-05, + "loss": 2.4038, + "step": 39314 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2778191370916115e-05, + "loss": 2.3521, + "step": 39315 + }, + { + "epoch": 1.68, + "grad_norm": 0.49609375, + "learning_rate": 1.277491060263718e-05, + "loss": 2.4274, + "step": 39316 + }, + { + "epoch": 1.68, + "grad_norm": 0.486328125, + "learning_rate": 1.2771630226839293e-05, + "loss": 2.4455, + "step": 39317 + }, + { + "epoch": 1.68, + "grad_norm": 0.48828125, + "learning_rate": 1.2768350243537164e-05, + "loss": 2.3883, + "step": 39318 + }, + { + "epoch": 1.68, + "grad_norm": 0.435546875, + "learning_rate": 1.2765070652745614e-05, + "loss": 2.4367, + "step": 39319 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.2761791454479333e-05, + "loss": 2.6168, + "step": 39320 + }, + { + "epoch": 1.68, + "grad_norm": 0.427734375, + "learning_rate": 1.275851264875313e-05, + "loss": 2.3837, + "step": 39321 + }, + { + "epoch": 1.68, + "grad_norm": 0.50390625, + "learning_rate": 1.2755234235581725e-05, + "loss": 2.4709, + "step": 39322 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.2751956214979876e-05, + "loss": 2.556, + "step": 39323 + }, + { + "epoch": 1.68, + "grad_norm": 0.42578125, + "learning_rate": 1.2748678586962359e-05, + "loss": 2.5401, + "step": 39324 + }, + { + "epoch": 1.68, + "grad_norm": 0.486328125, + "learning_rate": 1.2745401351543895e-05, + "loss": 2.4845, + "step": 39325 + }, + { + "epoch": 1.68, + "grad_norm": 0.462890625, + "learning_rate": 1.2742124508739206e-05, + "loss": 2.4353, + "step": 39326 + }, + { + "epoch": 1.68, + "grad_norm": 0.43359375, + "learning_rate": 1.2738848058563103e-05, + "loss": 2.4225, + "step": 39327 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2735572001030294e-05, + "loss": 2.3281, + "step": 39328 + }, + { + "epoch": 1.68, + "grad_norm": 0.51171875, + "learning_rate": 1.2732296336155502e-05, + "loss": 2.5331, + "step": 39329 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2729021063953483e-05, + "loss": 2.4749, + "step": 39330 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2725746184438992e-05, + "loss": 2.4986, + "step": 39331 + }, + { + "epoch": 1.68, + "grad_norm": 0.498046875, + "learning_rate": 1.2722471697626736e-05, + "loss": 2.361, + "step": 39332 + }, + { + "epoch": 1.68, + "grad_norm": 0.51953125, + "learning_rate": 1.2719197603531473e-05, + "loss": 2.3773, + "step": 39333 + }, + { + "epoch": 1.68, + "grad_norm": 0.455078125, + "learning_rate": 1.2715923902167903e-05, + "loss": 2.3513, + "step": 39334 + }, + { + "epoch": 1.68, + "grad_norm": 0.44140625, + "learning_rate": 1.2712650593550802e-05, + "loss": 2.4594, + "step": 39335 + }, + { + "epoch": 1.68, + "grad_norm": 0.466796875, + "learning_rate": 1.2709377677694855e-05, + "loss": 2.4964, + "step": 39336 + }, + { + "epoch": 1.68, + "grad_norm": 0.484375, + "learning_rate": 1.2706105154614823e-05, + "loss": 2.4645, + "step": 39337 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.27028330243254e-05, + "loss": 2.5391, + "step": 39338 + }, + { + "epoch": 1.68, + "grad_norm": 0.453125, + "learning_rate": 1.2699561286841322e-05, + "loss": 2.4399, + "step": 39339 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2696289942177331e-05, + "loss": 2.4196, + "step": 39340 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2693018990348126e-05, + "loss": 2.445, + "step": 39341 + }, + { + "epoch": 1.68, + "grad_norm": 0.48046875, + "learning_rate": 1.2689748431368397e-05, + "loss": 2.473, + "step": 39342 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.2686478265252932e-05, + "loss": 2.4658, + "step": 39343 + }, + { + "epoch": 1.68, + "grad_norm": 0.52734375, + "learning_rate": 1.2683208492016397e-05, + "loss": 2.5105, + "step": 39344 + }, + { + "epoch": 1.68, + "grad_norm": 0.47265625, + "learning_rate": 1.26799391116735e-05, + "loss": 2.3262, + "step": 39345 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2676670124238966e-05, + "loss": 2.3897, + "step": 39346 + }, + { + "epoch": 1.68, + "grad_norm": 0.458984375, + "learning_rate": 1.2673401529727513e-05, + "loss": 2.3978, + "step": 39347 + }, + { + "epoch": 1.68, + "grad_norm": 0.4453125, + "learning_rate": 1.2670133328153833e-05, + "loss": 2.303, + "step": 39348 + }, + { + "epoch": 1.68, + "grad_norm": 0.427734375, + "learning_rate": 1.2666865519532656e-05, + "loss": 2.4832, + "step": 39349 + }, + { + "epoch": 1.68, + "grad_norm": 0.53125, + "learning_rate": 1.2663598103878638e-05, + "loss": 2.4834, + "step": 39350 + }, + { + "epoch": 1.68, + "grad_norm": 0.515625, + "learning_rate": 1.2660331081206534e-05, + "loss": 2.5066, + "step": 39351 + }, + { + "epoch": 1.68, + "grad_norm": 0.44921875, + "learning_rate": 1.2657064451530998e-05, + "loss": 2.5114, + "step": 39352 + }, + { + "epoch": 1.68, + "grad_norm": 0.455078125, + "learning_rate": 1.2653798214866785e-05, + "loss": 2.4311, + "step": 39353 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.265053237122853e-05, + "loss": 2.4334, + "step": 39354 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.2647266920630951e-05, + "loss": 2.5412, + "step": 39355 + }, + { + "epoch": 1.68, + "grad_norm": 0.44140625, + "learning_rate": 1.2644001863088773e-05, + "loss": 2.4457, + "step": 39356 + }, + { + "epoch": 1.68, + "grad_norm": 0.451171875, + "learning_rate": 1.264073719861667e-05, + "loss": 2.4013, + "step": 39357 + }, + { + "epoch": 1.68, + "grad_norm": 0.515625, + "learning_rate": 1.2637472927229265e-05, + "loss": 2.5012, + "step": 39358 + }, + { + "epoch": 1.68, + "grad_norm": 0.443359375, + "learning_rate": 1.2634209048941358e-05, + "loss": 2.4856, + "step": 39359 + }, + { + "epoch": 1.68, + "grad_norm": 0.46875, + "learning_rate": 1.2630945563767582e-05, + "loss": 2.4476, + "step": 39360 + }, + { + "epoch": 1.68, + "grad_norm": 0.51953125, + "learning_rate": 1.2627682471722591e-05, + "loss": 2.4387, + "step": 39361 + }, + { + "epoch": 1.68, + "grad_norm": 0.41796875, + "learning_rate": 1.2624419772821105e-05, + "loss": 2.3326, + "step": 39362 + }, + { + "epoch": 1.68, + "grad_norm": 0.44921875, + "learning_rate": 1.2621157467077816e-05, + "loss": 2.4219, + "step": 39363 + }, + { + "epoch": 1.68, + "grad_norm": 0.515625, + "learning_rate": 1.2617895554507376e-05, + "loss": 2.6184, + "step": 39364 + }, + { + "epoch": 1.68, + "grad_norm": 0.470703125, + "learning_rate": 1.2614634035124473e-05, + "loss": 2.4025, + "step": 39365 + }, + { + "epoch": 1.68, + "grad_norm": 0.4453125, + "learning_rate": 1.2611372908943776e-05, + "loss": 2.4349, + "step": 39366 + }, + { + "epoch": 1.68, + "grad_norm": 0.462890625, + "learning_rate": 1.260811217597998e-05, + "loss": 2.476, + "step": 39367 + }, + { + "epoch": 1.68, + "grad_norm": 0.482421875, + "learning_rate": 1.260485183624771e-05, + "loss": 2.404, + "step": 39368 + }, + { + "epoch": 1.68, + "grad_norm": 0.5, + "learning_rate": 1.2601591889761699e-05, + "loss": 2.4339, + "step": 39369 + }, + { + "epoch": 1.68, + "grad_norm": 0.486328125, + "learning_rate": 1.2598332336536566e-05, + "loss": 2.4183, + "step": 39370 + }, + { + "epoch": 1.68, + "grad_norm": 0.50390625, + "learning_rate": 1.2595073176586991e-05, + "loss": 2.3879, + "step": 39371 + }, + { + "epoch": 1.68, + "grad_norm": 0.49609375, + "learning_rate": 1.259181440992766e-05, + "loss": 2.3998, + "step": 39372 + }, + { + "epoch": 1.68, + "grad_norm": 0.484375, + "learning_rate": 1.2588556036573229e-05, + "loss": 2.4987, + "step": 39373 + }, + { + "epoch": 1.68, + "grad_norm": 0.5390625, + "learning_rate": 1.2585298056538308e-05, + "loss": 2.4114, + "step": 39374 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2582040469837631e-05, + "loss": 2.4706, + "step": 39375 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.2578783276485839e-05, + "loss": 2.3954, + "step": 39376 + }, + { + "epoch": 1.68, + "grad_norm": 0.46484375, + "learning_rate": 1.2575526476497534e-05, + "loss": 2.2786, + "step": 39377 + }, + { + "epoch": 1.68, + "grad_norm": 0.498046875, + "learning_rate": 1.2572270069887426e-05, + "loss": 2.3986, + "step": 39378 + }, + { + "epoch": 1.68, + "grad_norm": 0.5234375, + "learning_rate": 1.256901405667017e-05, + "loss": 2.5653, + "step": 39379 + }, + { + "epoch": 1.68, + "grad_norm": 0.4453125, + "learning_rate": 1.2565758436860376e-05, + "loss": 2.4129, + "step": 39380 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.256250321047273e-05, + "loss": 2.4557, + "step": 39381 + }, + { + "epoch": 1.68, + "grad_norm": 0.4453125, + "learning_rate": 1.2559248377521849e-05, + "loss": 2.3755, + "step": 39382 + }, + { + "epoch": 1.68, + "grad_norm": 0.443359375, + "learning_rate": 1.2555993938022415e-05, + "loss": 2.4346, + "step": 39383 + }, + { + "epoch": 1.68, + "grad_norm": 0.453125, + "learning_rate": 1.2552739891989029e-05, + "loss": 2.4126, + "step": 39384 + }, + { + "epoch": 1.68, + "grad_norm": 0.5, + "learning_rate": 1.254948623943638e-05, + "loss": 2.5654, + "step": 39385 + }, + { + "epoch": 1.68, + "grad_norm": 0.42578125, + "learning_rate": 1.2546232980379058e-05, + "loss": 2.4462, + "step": 39386 + }, + { + "epoch": 1.68, + "grad_norm": 0.443359375, + "learning_rate": 1.2542980114831727e-05, + "loss": 2.4063, + "step": 39387 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2539727642809052e-05, + "loss": 2.5185, + "step": 39388 + }, + { + "epoch": 1.68, + "grad_norm": 0.482421875, + "learning_rate": 1.2536475564325634e-05, + "loss": 2.4942, + "step": 39389 + }, + { + "epoch": 1.68, + "grad_norm": 0.455078125, + "learning_rate": 1.253322387939606e-05, + "loss": 2.3443, + "step": 39390 + }, + { + "epoch": 1.68, + "grad_norm": 0.44921875, + "learning_rate": 1.2529972588035067e-05, + "loss": 2.3979, + "step": 39391 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.2526721690257238e-05, + "loss": 2.5189, + "step": 39392 + }, + { + "epoch": 1.68, + "grad_norm": 0.44921875, + "learning_rate": 1.2523471186077152e-05, + "loss": 2.4682, + "step": 39393 + }, + { + "epoch": 1.68, + "grad_norm": 0.4453125, + "learning_rate": 1.2520221075509497e-05, + "loss": 2.4581, + "step": 39394 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2516971358568885e-05, + "loss": 2.5035, + "step": 39395 + }, + { + "epoch": 1.68, + "grad_norm": 0.482421875, + "learning_rate": 1.2513722035269915e-05, + "loss": 2.4901, + "step": 39396 + }, + { + "epoch": 1.68, + "grad_norm": 0.4375, + "learning_rate": 1.251047310562724e-05, + "loss": 2.4313, + "step": 39397 + }, + { + "epoch": 1.68, + "grad_norm": 0.44921875, + "learning_rate": 1.250722456965544e-05, + "loss": 2.4303, + "step": 39398 + }, + { + "epoch": 1.68, + "grad_norm": 0.490234375, + "learning_rate": 1.250397642736918e-05, + "loss": 2.334, + "step": 39399 + }, + { + "epoch": 1.68, + "grad_norm": 0.470703125, + "learning_rate": 1.2500728678783024e-05, + "loss": 2.4988, + "step": 39400 + }, + { + "epoch": 1.68, + "grad_norm": 0.47265625, + "learning_rate": 1.2497481323911642e-05, + "loss": 2.3665, + "step": 39401 + }, + { + "epoch": 1.68, + "grad_norm": 0.470703125, + "learning_rate": 1.2494234362769586e-05, + "loss": 2.3309, + "step": 39402 + }, + { + "epoch": 1.68, + "grad_norm": 0.431640625, + "learning_rate": 1.24909877953715e-05, + "loss": 2.3814, + "step": 39403 + }, + { + "epoch": 1.68, + "grad_norm": 0.439453125, + "learning_rate": 1.2487741621732008e-05, + "loss": 2.562, + "step": 39404 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.2484495841865696e-05, + "loss": 2.3156, + "step": 39405 + }, + { + "epoch": 1.68, + "grad_norm": 0.55078125, + "learning_rate": 1.248125045578713e-05, + "loss": 2.4514, + "step": 39406 + }, + { + "epoch": 1.68, + "grad_norm": 0.462890625, + "learning_rate": 1.247800546351099e-05, + "loss": 2.5363, + "step": 39407 + }, + { + "epoch": 1.68, + "grad_norm": 0.482421875, + "learning_rate": 1.2474760865051849e-05, + "loss": 2.377, + "step": 39408 + }, + { + "epoch": 1.68, + "grad_norm": 0.46875, + "learning_rate": 1.2471516660424265e-05, + "loss": 2.2634, + "step": 39409 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.246827284964287e-05, + "loss": 2.5493, + "step": 39410 + }, + { + "epoch": 1.68, + "grad_norm": 0.421875, + "learning_rate": 1.2465029432722286e-05, + "loss": 2.5462, + "step": 39411 + }, + { + "epoch": 1.68, + "grad_norm": 0.52734375, + "learning_rate": 1.2461786409677056e-05, + "loss": 2.393, + "step": 39412 + }, + { + "epoch": 1.68, + "grad_norm": 0.416015625, + "learning_rate": 1.2458543780521815e-05, + "loss": 2.4222, + "step": 39413 + }, + { + "epoch": 1.68, + "grad_norm": 0.5390625, + "learning_rate": 1.2455301545271126e-05, + "loss": 2.5045, + "step": 39414 + }, + { + "epoch": 1.68, + "grad_norm": 0.384765625, + "learning_rate": 1.2452059703939579e-05, + "loss": 2.4661, + "step": 39415 + }, + { + "epoch": 1.68, + "grad_norm": 0.54296875, + "learning_rate": 1.2448818256541761e-05, + "loss": 2.4512, + "step": 39416 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2445577203092285e-05, + "loss": 2.5034, + "step": 39417 + }, + { + "epoch": 1.68, + "grad_norm": 0.5078125, + "learning_rate": 1.2442336543605692e-05, + "loss": 2.4545, + "step": 39418 + }, + { + "epoch": 1.68, + "grad_norm": 0.4296875, + "learning_rate": 1.2439096278096585e-05, + "loss": 2.3651, + "step": 39419 + }, + { + "epoch": 1.68, + "grad_norm": 0.443359375, + "learning_rate": 1.243585640657957e-05, + "loss": 2.397, + "step": 39420 + }, + { + "epoch": 1.68, + "grad_norm": 0.396484375, + "learning_rate": 1.2432616929069207e-05, + "loss": 2.2643, + "step": 39421 + }, + { + "epoch": 1.68, + "grad_norm": 0.48828125, + "learning_rate": 1.2429377845580037e-05, + "loss": 2.3808, + "step": 39422 + }, + { + "epoch": 1.68, + "grad_norm": 0.4453125, + "learning_rate": 1.2426139156126659e-05, + "loss": 2.3137, + "step": 39423 + }, + { + "epoch": 1.68, + "grad_norm": 0.453125, + "learning_rate": 1.2422900860723674e-05, + "loss": 2.4974, + "step": 39424 + }, + { + "epoch": 1.68, + "grad_norm": 0.458984375, + "learning_rate": 1.2419662959385613e-05, + "loss": 2.4186, + "step": 39425 + }, + { + "epoch": 1.68, + "grad_norm": 0.494140625, + "learning_rate": 1.2416425452127078e-05, + "loss": 2.4519, + "step": 39426 + }, + { + "epoch": 1.68, + "grad_norm": 0.453125, + "learning_rate": 1.24131883389626e-05, + "loss": 2.4597, + "step": 39427 + }, + { + "epoch": 1.68, + "grad_norm": 0.5625, + "learning_rate": 1.2409951619906767e-05, + "loss": 2.3924, + "step": 39428 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2406715294974159e-05, + "loss": 2.4455, + "step": 39429 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.2403479364179327e-05, + "loss": 2.4443, + "step": 39430 + }, + { + "epoch": 1.68, + "grad_norm": 0.482421875, + "learning_rate": 1.2400243827536795e-05, + "loss": 2.3851, + "step": 39431 + }, + { + "epoch": 1.68, + "grad_norm": 0.46875, + "learning_rate": 1.2397008685061163e-05, + "loss": 2.5072, + "step": 39432 + }, + { + "epoch": 1.68, + "grad_norm": 0.453125, + "learning_rate": 1.2393773936766995e-05, + "loss": 2.3214, + "step": 39433 + }, + { + "epoch": 1.68, + "grad_norm": 0.408203125, + "learning_rate": 1.2390539582668792e-05, + "loss": 2.4414, + "step": 39434 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.2387305622781165e-05, + "loss": 2.4677, + "step": 39435 + }, + { + "epoch": 1.68, + "grad_norm": 0.41796875, + "learning_rate": 1.2384072057118646e-05, + "loss": 2.3127, + "step": 39436 + }, + { + "epoch": 1.68, + "grad_norm": 0.494140625, + "learning_rate": 1.2380838885695789e-05, + "loss": 2.4983, + "step": 39437 + }, + { + "epoch": 1.68, + "grad_norm": 0.4296875, + "learning_rate": 1.2377606108527118e-05, + "loss": 2.445, + "step": 39438 + }, + { + "epoch": 1.68, + "grad_norm": 0.46875, + "learning_rate": 1.2374373725627197e-05, + "loss": 2.3215, + "step": 39439 + }, + { + "epoch": 1.68, + "grad_norm": 0.458984375, + "learning_rate": 1.2371141737010594e-05, + "loss": 2.4435, + "step": 39440 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2367910142691807e-05, + "loss": 2.5026, + "step": 39441 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.2364678942685415e-05, + "loss": 2.3186, + "step": 39442 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.2361448137005927e-05, + "loss": 2.4772, + "step": 39443 + }, + { + "epoch": 1.68, + "grad_norm": 0.431640625, + "learning_rate": 1.2358217725667886e-05, + "loss": 2.4633, + "step": 39444 + }, + { + "epoch": 1.68, + "grad_norm": 0.5078125, + "learning_rate": 1.235498770868586e-05, + "loss": 2.4946, + "step": 39445 + }, + { + "epoch": 1.68, + "grad_norm": 0.466796875, + "learning_rate": 1.2351758086074349e-05, + "loss": 2.3528, + "step": 39446 + }, + { + "epoch": 1.68, + "grad_norm": 0.44921875, + "learning_rate": 1.2348528857847885e-05, + "loss": 2.5326, + "step": 39447 + }, + { + "epoch": 1.68, + "grad_norm": 0.462890625, + "learning_rate": 1.2345300024021011e-05, + "loss": 2.3943, + "step": 39448 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.2342071584608273e-05, + "loss": 2.3905, + "step": 39449 + }, + { + "epoch": 1.68, + "grad_norm": 0.470703125, + "learning_rate": 1.2338843539624146e-05, + "loss": 2.5283, + "step": 39450 + }, + { + "epoch": 1.68, + "grad_norm": 0.4453125, + "learning_rate": 1.2335615889083197e-05, + "loss": 2.4424, + "step": 39451 + }, + { + "epoch": 1.68, + "grad_norm": 0.42578125, + "learning_rate": 1.2332388632999959e-05, + "loss": 2.3926, + "step": 39452 + }, + { + "epoch": 1.68, + "grad_norm": 0.455078125, + "learning_rate": 1.2329161771388931e-05, + "loss": 2.568, + "step": 39453 + }, + { + "epoch": 1.68, + "grad_norm": 0.44921875, + "learning_rate": 1.2325935304264613e-05, + "loss": 2.4247, + "step": 39454 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2322709231641538e-05, + "loss": 2.4399, + "step": 39455 + }, + { + "epoch": 1.68, + "grad_norm": 0.4296875, + "learning_rate": 1.231948355353426e-05, + "loss": 2.4959, + "step": 39456 + }, + { + "epoch": 1.68, + "grad_norm": 0.48828125, + "learning_rate": 1.2316258269957236e-05, + "loss": 2.4681, + "step": 39457 + }, + { + "epoch": 1.68, + "grad_norm": 0.396484375, + "learning_rate": 1.2313033380925032e-05, + "loss": 2.2957, + "step": 39458 + }, + { + "epoch": 1.68, + "grad_norm": 0.48828125, + "learning_rate": 1.2309808886452113e-05, + "loss": 2.4248, + "step": 39459 + }, + { + "epoch": 1.68, + "grad_norm": 0.5078125, + "learning_rate": 1.2306584786552999e-05, + "loss": 2.4741, + "step": 39460 + }, + { + "epoch": 1.68, + "grad_norm": 0.48046875, + "learning_rate": 1.2303361081242226e-05, + "loss": 2.4211, + "step": 39461 + }, + { + "epoch": 1.68, + "grad_norm": 0.46484375, + "learning_rate": 1.2300137770534282e-05, + "loss": 2.4007, + "step": 39462 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2296914854443643e-05, + "loss": 2.4559, + "step": 39463 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.229369233298484e-05, + "loss": 2.4002, + "step": 39464 + }, + { + "epoch": 1.68, + "grad_norm": 0.4296875, + "learning_rate": 1.2290470206172388e-05, + "loss": 2.3648, + "step": 39465 + }, + { + "epoch": 1.68, + "grad_norm": 0.46875, + "learning_rate": 1.228724847402074e-05, + "loss": 2.3711, + "step": 39466 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.2284027136544429e-05, + "loss": 2.5709, + "step": 39467 + }, + { + "epoch": 1.68, + "grad_norm": 0.44921875, + "learning_rate": 1.2280806193757954e-05, + "loss": 2.4938, + "step": 39468 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.2277585645675794e-05, + "loss": 2.4906, + "step": 39469 + }, + { + "epoch": 1.68, + "grad_norm": 0.427734375, + "learning_rate": 1.2274365492312423e-05, + "loss": 2.4035, + "step": 39470 + }, + { + "epoch": 1.68, + "grad_norm": 0.4375, + "learning_rate": 1.2271145733682343e-05, + "loss": 2.3569, + "step": 39471 + }, + { + "epoch": 1.68, + "grad_norm": 0.515625, + "learning_rate": 1.2267926369800064e-05, + "loss": 2.5, + "step": 39472 + }, + { + "epoch": 1.68, + "grad_norm": 0.486328125, + "learning_rate": 1.226470740068003e-05, + "loss": 2.5692, + "step": 39473 + }, + { + "epoch": 1.68, + "grad_norm": 0.48828125, + "learning_rate": 1.2261488826336776e-05, + "loss": 2.3374, + "step": 39474 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.2258270646784742e-05, + "loss": 2.4281, + "step": 39475 + }, + { + "epoch": 1.68, + "grad_norm": 0.5234375, + "learning_rate": 1.2255052862038418e-05, + "loss": 2.4178, + "step": 39476 + }, + { + "epoch": 1.68, + "grad_norm": 0.41796875, + "learning_rate": 1.2251835472112305e-05, + "loss": 2.456, + "step": 39477 + }, + { + "epoch": 1.68, + "grad_norm": 0.484375, + "learning_rate": 1.224861847702088e-05, + "loss": 2.4525, + "step": 39478 + }, + { + "epoch": 1.68, + "grad_norm": 0.47265625, + "learning_rate": 1.2245401876778561e-05, + "loss": 2.5192, + "step": 39479 + }, + { + "epoch": 1.68, + "grad_norm": 0.451171875, + "learning_rate": 1.2242185671399875e-05, + "loss": 2.4013, + "step": 39480 + }, + { + "epoch": 1.68, + "grad_norm": 0.486328125, + "learning_rate": 1.2238969860899308e-05, + "loss": 2.4581, + "step": 39481 + }, + { + "epoch": 1.68, + "grad_norm": 0.443359375, + "learning_rate": 1.223575444529127e-05, + "loss": 2.5472, + "step": 39482 + }, + { + "epoch": 1.68, + "grad_norm": 0.466796875, + "learning_rate": 1.2232539424590273e-05, + "loss": 2.5206, + "step": 39483 + }, + { + "epoch": 1.68, + "grad_norm": 0.44140625, + "learning_rate": 1.2229324798810781e-05, + "loss": 2.5461, + "step": 39484 + }, + { + "epoch": 1.68, + "grad_norm": 0.482421875, + "learning_rate": 1.2226110567967264e-05, + "loss": 2.5241, + "step": 39485 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.222289673207414e-05, + "loss": 2.5099, + "step": 39486 + }, + { + "epoch": 1.68, + "grad_norm": 0.5, + "learning_rate": 1.2219683291145911e-05, + "loss": 2.3741, + "step": 39487 + }, + { + "epoch": 1.68, + "grad_norm": 0.44140625, + "learning_rate": 1.2216470245197043e-05, + "loss": 2.5881, + "step": 39488 + }, + { + "epoch": 1.68, + "grad_norm": 0.50390625, + "learning_rate": 1.2213257594241967e-05, + "loss": 2.5354, + "step": 39489 + }, + { + "epoch": 1.68, + "grad_norm": 0.470703125, + "learning_rate": 1.2210045338295162e-05, + "loss": 2.4248, + "step": 39490 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2206833477371049e-05, + "loss": 2.3314, + "step": 39491 + }, + { + "epoch": 1.68, + "grad_norm": 0.4609375, + "learning_rate": 1.2203622011484094e-05, + "loss": 2.5152, + "step": 39492 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.2200410940648787e-05, + "loss": 2.5045, + "step": 39493 + }, + { + "epoch": 1.68, + "grad_norm": 0.47265625, + "learning_rate": 1.2197200264879538e-05, + "loss": 2.4174, + "step": 39494 + }, + { + "epoch": 1.68, + "grad_norm": 0.44140625, + "learning_rate": 1.2193989984190779e-05, + "loss": 2.4549, + "step": 39495 + }, + { + "epoch": 1.68, + "grad_norm": 0.416015625, + "learning_rate": 1.2190780098596977e-05, + "loss": 2.3623, + "step": 39496 + }, + { + "epoch": 1.68, + "grad_norm": 0.439453125, + "learning_rate": 1.2187570608112586e-05, + "loss": 2.3392, + "step": 39497 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.2184361512752018e-05, + "loss": 2.5257, + "step": 39498 + }, + { + "epoch": 1.68, + "grad_norm": 0.515625, + "learning_rate": 1.218115281252974e-05, + "loss": 2.4886, + "step": 39499 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2177944507460193e-05, + "loss": 2.5372, + "step": 39500 + }, + { + "epoch": 1.68, + "grad_norm": 0.54296875, + "learning_rate": 1.2174736597557801e-05, + "loss": 2.4297, + "step": 39501 + }, + { + "epoch": 1.68, + "grad_norm": 0.46875, + "learning_rate": 1.2171529082836986e-05, + "loss": 2.5152, + "step": 39502 + }, + { + "epoch": 1.68, + "grad_norm": 0.486328125, + "learning_rate": 1.216832196331218e-05, + "loss": 2.4231, + "step": 39503 + }, + { + "epoch": 1.68, + "grad_norm": 0.427734375, + "learning_rate": 1.2165115238997848e-05, + "loss": 2.4542, + "step": 39504 + }, + { + "epoch": 1.68, + "grad_norm": 0.443359375, + "learning_rate": 1.2161908909908382e-05, + "loss": 2.4764, + "step": 39505 + }, + { + "epoch": 1.68, + "grad_norm": 0.453125, + "learning_rate": 1.2158702976058245e-05, + "loss": 2.543, + "step": 39506 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.2155497437461827e-05, + "loss": 2.4118, + "step": 39507 + }, + { + "epoch": 1.68, + "grad_norm": 0.50390625, + "learning_rate": 1.215229229413356e-05, + "loss": 2.5079, + "step": 39508 + }, + { + "epoch": 1.68, + "grad_norm": 0.515625, + "learning_rate": 1.2149087546087889e-05, + "loss": 2.4152, + "step": 39509 + }, + { + "epoch": 1.68, + "grad_norm": 0.462890625, + "learning_rate": 1.2145883193339224e-05, + "loss": 2.4047, + "step": 39510 + }, + { + "epoch": 1.68, + "grad_norm": 0.4140625, + "learning_rate": 1.2142679235901955e-05, + "loss": 2.4068, + "step": 39511 + }, + { + "epoch": 1.68, + "grad_norm": 0.486328125, + "learning_rate": 1.2139475673790523e-05, + "loss": 2.3524, + "step": 39512 + }, + { + "epoch": 1.68, + "grad_norm": 0.478515625, + "learning_rate": 1.2136272507019365e-05, + "loss": 2.4262, + "step": 39513 + }, + { + "epoch": 1.68, + "grad_norm": 0.42578125, + "learning_rate": 1.2133069735602842e-05, + "loss": 2.4593, + "step": 39514 + }, + { + "epoch": 1.68, + "grad_norm": 0.5078125, + "learning_rate": 1.2129867359555402e-05, + "loss": 2.5064, + "step": 39515 + }, + { + "epoch": 1.68, + "grad_norm": 0.4921875, + "learning_rate": 1.2126665378891466e-05, + "loss": 2.4545, + "step": 39516 + }, + { + "epoch": 1.68, + "grad_norm": 0.404296875, + "learning_rate": 1.2123463793625411e-05, + "loss": 2.2907, + "step": 39517 + }, + { + "epoch": 1.68, + "grad_norm": 0.435546875, + "learning_rate": 1.2120262603771648e-05, + "loss": 2.5224, + "step": 39518 + }, + { + "epoch": 1.68, + "grad_norm": 0.458984375, + "learning_rate": 1.2117061809344577e-05, + "loss": 2.4831, + "step": 39519 + }, + { + "epoch": 1.68, + "grad_norm": 0.53125, + "learning_rate": 1.211386141035863e-05, + "loss": 2.5608, + "step": 39520 + }, + { + "epoch": 1.68, + "grad_norm": 0.447265625, + "learning_rate": 1.2110661406828172e-05, + "loss": 2.4431, + "step": 39521 + }, + { + "epoch": 1.68, + "grad_norm": 0.51171875, + "learning_rate": 1.2107461798767638e-05, + "loss": 2.5369, + "step": 39522 + }, + { + "epoch": 1.68, + "grad_norm": 0.451171875, + "learning_rate": 1.2104262586191384e-05, + "loss": 2.4497, + "step": 39523 + }, + { + "epoch": 1.68, + "grad_norm": 0.50390625, + "learning_rate": 1.210106376911383e-05, + "loss": 2.3683, + "step": 39524 + }, + { + "epoch": 1.68, + "grad_norm": 0.412109375, + "learning_rate": 1.2097865347549376e-05, + "loss": 2.5472, + "step": 39525 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.209466732151241e-05, + "loss": 2.5231, + "step": 39526 + }, + { + "epoch": 1.68, + "grad_norm": 0.4375, + "learning_rate": 1.20914696910173e-05, + "loss": 2.4597, + "step": 39527 + }, + { + "epoch": 1.68, + "grad_norm": 0.462890625, + "learning_rate": 1.2088272456078432e-05, + "loss": 2.5759, + "step": 39528 + }, + { + "epoch": 1.68, + "grad_norm": 0.443359375, + "learning_rate": 1.2085075616710251e-05, + "loss": 2.3708, + "step": 39529 + }, + { + "epoch": 1.68, + "grad_norm": 0.484375, + "learning_rate": 1.2081879172927068e-05, + "loss": 2.4868, + "step": 39530 + }, + { + "epoch": 1.68, + "grad_norm": 0.44140625, + "learning_rate": 1.2078683124743295e-05, + "loss": 2.2948, + "step": 39531 + }, + { + "epoch": 1.68, + "grad_norm": 0.4765625, + "learning_rate": 1.2075487472173342e-05, + "loss": 2.4569, + "step": 39532 + }, + { + "epoch": 1.68, + "grad_norm": 0.51171875, + "learning_rate": 1.2072292215231573e-05, + "loss": 2.4918, + "step": 39533 + }, + { + "epoch": 1.68, + "grad_norm": 0.45703125, + "learning_rate": 1.2069097353932323e-05, + "loss": 2.4731, + "step": 39534 + }, + { + "epoch": 1.68, + "grad_norm": 0.494140625, + "learning_rate": 1.2065902888290014e-05, + "loss": 2.5182, + "step": 39535 + }, + { + "epoch": 1.68, + "grad_norm": 0.484375, + "learning_rate": 1.2062708818319002e-05, + "loss": 2.3878, + "step": 39536 + }, + { + "epoch": 1.68, + "grad_norm": 0.498046875, + "learning_rate": 1.2059515144033651e-05, + "loss": 2.3935, + "step": 39537 + }, + { + "epoch": 1.68, + "grad_norm": 0.466796875, + "learning_rate": 1.2056321865448373e-05, + "loss": 2.4304, + "step": 39538 + }, + { + "epoch": 1.68, + "grad_norm": 0.50390625, + "learning_rate": 1.2053128982577499e-05, + "loss": 2.5075, + "step": 39539 + }, + { + "epoch": 1.68, + "grad_norm": 0.474609375, + "learning_rate": 1.2049936495435366e-05, + "loss": 2.4171, + "step": 39540 + }, + { + "epoch": 1.68, + "grad_norm": 0.466796875, + "learning_rate": 1.2046744404036415e-05, + "loss": 2.435, + "step": 39541 + }, + { + "epoch": 1.68, + "grad_norm": 0.453125, + "learning_rate": 1.2043552708394967e-05, + "loss": 2.5873, + "step": 39542 + }, + { + "epoch": 1.69, + "grad_norm": 0.5, + "learning_rate": 1.204036140852538e-05, + "loss": 2.4789, + "step": 39543 + }, + { + "epoch": 1.69, + "grad_norm": 0.44921875, + "learning_rate": 1.2037170504442008e-05, + "loss": 2.5267, + "step": 39544 + }, + { + "epoch": 1.69, + "grad_norm": 0.46875, + "learning_rate": 1.203397999615925e-05, + "loss": 2.376, + "step": 39545 + }, + { + "epoch": 1.69, + "grad_norm": 0.412109375, + "learning_rate": 1.2030789883691407e-05, + "loss": 2.4198, + "step": 39546 + }, + { + "epoch": 1.69, + "grad_norm": 0.4375, + "learning_rate": 1.2027600167052888e-05, + "loss": 2.469, + "step": 39547 + }, + { + "epoch": 1.69, + "grad_norm": 0.50390625, + "learning_rate": 1.2024410846257994e-05, + "loss": 2.5274, + "step": 39548 + }, + { + "epoch": 1.69, + "grad_norm": 0.4375, + "learning_rate": 1.2021221921321123e-05, + "loss": 2.3937, + "step": 39549 + }, + { + "epoch": 1.69, + "grad_norm": 0.4921875, + "learning_rate": 1.2018033392256578e-05, + "loss": 2.4223, + "step": 39550 + }, + { + "epoch": 1.69, + "grad_norm": 0.423828125, + "learning_rate": 1.2014845259078756e-05, + "loss": 2.3125, + "step": 39551 + }, + { + "epoch": 1.69, + "grad_norm": 0.427734375, + "learning_rate": 1.2011657521801945e-05, + "loss": 2.4409, + "step": 39552 + }, + { + "epoch": 1.69, + "grad_norm": 0.43359375, + "learning_rate": 1.2008470180440523e-05, + "loss": 2.3955, + "step": 39553 + }, + { + "epoch": 1.69, + "grad_norm": 0.4375, + "learning_rate": 1.2005283235008846e-05, + "loss": 2.4209, + "step": 39554 + }, + { + "epoch": 1.69, + "grad_norm": 0.427734375, + "learning_rate": 1.2002096685521248e-05, + "loss": 2.3991, + "step": 39555 + }, + { + "epoch": 1.69, + "grad_norm": 0.4921875, + "learning_rate": 1.1998910531992002e-05, + "loss": 2.4927, + "step": 39556 + }, + { + "epoch": 1.69, + "grad_norm": 0.490234375, + "learning_rate": 1.1995724774435536e-05, + "loss": 2.6624, + "step": 39557 + }, + { + "epoch": 1.69, + "grad_norm": 0.478515625, + "learning_rate": 1.1992539412866155e-05, + "loss": 2.4165, + "step": 39558 + }, + { + "epoch": 1.69, + "grad_norm": 0.48828125, + "learning_rate": 1.1989354447298162e-05, + "loss": 2.31, + "step": 39559 + }, + { + "epoch": 1.69, + "grad_norm": 0.43359375, + "learning_rate": 1.1986169877745912e-05, + "loss": 2.4901, + "step": 39560 + }, + { + "epoch": 1.69, + "grad_norm": 0.455078125, + "learning_rate": 1.1982985704223749e-05, + "loss": 2.362, + "step": 39561 + }, + { + "epoch": 1.69, + "grad_norm": 0.5390625, + "learning_rate": 1.1979801926745959e-05, + "loss": 2.5434, + "step": 39562 + }, + { + "epoch": 1.69, + "grad_norm": 0.49609375, + "learning_rate": 1.1976618545326924e-05, + "loss": 2.4804, + "step": 39563 + }, + { + "epoch": 1.69, + "grad_norm": 0.447265625, + "learning_rate": 1.1973435559980905e-05, + "loss": 2.4358, + "step": 39564 + }, + { + "epoch": 1.69, + "grad_norm": 0.482421875, + "learning_rate": 1.1970252970722285e-05, + "loss": 2.378, + "step": 39565 + }, + { + "epoch": 1.69, + "grad_norm": 0.470703125, + "learning_rate": 1.1967070777565326e-05, + "loss": 2.4838, + "step": 39566 + }, + { + "epoch": 1.69, + "grad_norm": 0.470703125, + "learning_rate": 1.1963888980524408e-05, + "loss": 2.3554, + "step": 39567 + }, + { + "epoch": 1.69, + "grad_norm": 0.431640625, + "learning_rate": 1.1960707579613783e-05, + "loss": 2.4438, + "step": 39568 + }, + { + "epoch": 1.69, + "grad_norm": 0.443359375, + "learning_rate": 1.195752657484781e-05, + "loss": 2.2713, + "step": 39569 + }, + { + "epoch": 1.69, + "grad_norm": 0.51171875, + "learning_rate": 1.195434596624081e-05, + "loss": 2.5308, + "step": 39570 + }, + { + "epoch": 1.69, + "grad_norm": 0.4765625, + "learning_rate": 1.195116575380707e-05, + "loss": 2.4415, + "step": 39571 + }, + { + "epoch": 1.69, + "grad_norm": 0.48828125, + "learning_rate": 1.1947985937560868e-05, + "loss": 2.5654, + "step": 39572 + }, + { + "epoch": 1.69, + "grad_norm": 0.423828125, + "learning_rate": 1.1944806517516593e-05, + "loss": 2.3455, + "step": 39573 + }, + { + "epoch": 1.69, + "grad_norm": 0.482421875, + "learning_rate": 1.1941627493688501e-05, + "loss": 2.3327, + "step": 39574 + }, + { + "epoch": 1.69, + "grad_norm": 0.5546875, + "learning_rate": 1.193844886609089e-05, + "loss": 2.4019, + "step": 39575 + }, + { + "epoch": 1.69, + "grad_norm": 0.466796875, + "learning_rate": 1.1935270634738083e-05, + "loss": 2.5699, + "step": 39576 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.1932092799644378e-05, + "loss": 2.5894, + "step": 39577 + }, + { + "epoch": 1.69, + "grad_norm": 0.45703125, + "learning_rate": 1.1928915360824067e-05, + "loss": 2.4799, + "step": 39578 + }, + { + "epoch": 1.69, + "grad_norm": 0.51171875, + "learning_rate": 1.1925738318291457e-05, + "loss": 2.5779, + "step": 39579 + }, + { + "epoch": 1.69, + "grad_norm": 0.5234375, + "learning_rate": 1.1922561672060828e-05, + "loss": 2.4983, + "step": 39580 + }, + { + "epoch": 1.69, + "grad_norm": 0.4765625, + "learning_rate": 1.191938542214649e-05, + "loss": 2.3483, + "step": 39581 + }, + { + "epoch": 1.69, + "grad_norm": 0.494140625, + "learning_rate": 1.191620956856272e-05, + "loss": 2.496, + "step": 39582 + }, + { + "epoch": 1.69, + "grad_norm": 0.474609375, + "learning_rate": 1.1913034111323828e-05, + "loss": 2.4232, + "step": 39583 + }, + { + "epoch": 1.69, + "grad_norm": 0.46875, + "learning_rate": 1.1909859050444072e-05, + "loss": 2.4075, + "step": 39584 + }, + { + "epoch": 1.69, + "grad_norm": 0.49609375, + "learning_rate": 1.190668438593776e-05, + "loss": 2.4729, + "step": 39585 + }, + { + "epoch": 1.69, + "grad_norm": 0.46484375, + "learning_rate": 1.1903510117819194e-05, + "loss": 2.5076, + "step": 39586 + }, + { + "epoch": 1.69, + "grad_norm": 0.455078125, + "learning_rate": 1.190033624610264e-05, + "loss": 2.4906, + "step": 39587 + }, + { + "epoch": 1.69, + "grad_norm": 0.46875, + "learning_rate": 1.1897162770802328e-05, + "loss": 2.4174, + "step": 39588 + }, + { + "epoch": 1.69, + "grad_norm": 0.458984375, + "learning_rate": 1.1893989691932627e-05, + "loss": 2.4085, + "step": 39589 + }, + { + "epoch": 1.69, + "grad_norm": 0.482421875, + "learning_rate": 1.1890817009507782e-05, + "loss": 2.4602, + "step": 39590 + }, + { + "epoch": 1.69, + "grad_norm": 0.404296875, + "learning_rate": 1.1887644723542035e-05, + "loss": 2.4073, + "step": 39591 + }, + { + "epoch": 1.69, + "grad_norm": 0.46484375, + "learning_rate": 1.1884472834049677e-05, + "loss": 2.3808, + "step": 39592 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.1881301341045014e-05, + "loss": 2.3405, + "step": 39593 + }, + { + "epoch": 1.69, + "grad_norm": 0.486328125, + "learning_rate": 1.1878130244542274e-05, + "loss": 2.4544, + "step": 39594 + }, + { + "epoch": 1.69, + "grad_norm": 0.490234375, + "learning_rate": 1.1874959544555764e-05, + "loss": 2.5645, + "step": 39595 + }, + { + "epoch": 1.69, + "grad_norm": 0.5078125, + "learning_rate": 1.1871789241099707e-05, + "loss": 2.5545, + "step": 39596 + }, + { + "epoch": 1.69, + "grad_norm": 0.498046875, + "learning_rate": 1.1868619334188413e-05, + "loss": 2.5702, + "step": 39597 + }, + { + "epoch": 1.69, + "grad_norm": 0.625, + "learning_rate": 1.1865449823836105e-05, + "loss": 2.4889, + "step": 39598 + }, + { + "epoch": 1.69, + "grad_norm": 0.55859375, + "learning_rate": 1.1862280710057083e-05, + "loss": 2.3126, + "step": 39599 + }, + { + "epoch": 1.69, + "grad_norm": 0.419921875, + "learning_rate": 1.1859111992865568e-05, + "loss": 2.4415, + "step": 39600 + }, + { + "epoch": 1.69, + "grad_norm": 0.4609375, + "learning_rate": 1.1855943672275827e-05, + "loss": 2.3641, + "step": 39601 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.1852775748302158e-05, + "loss": 2.3654, + "step": 39602 + }, + { + "epoch": 1.69, + "grad_norm": 0.48046875, + "learning_rate": 1.1849608220958786e-05, + "loss": 2.4927, + "step": 39603 + }, + { + "epoch": 1.69, + "grad_norm": 0.546875, + "learning_rate": 1.1846441090259908e-05, + "loss": 2.5138, + "step": 39604 + }, + { + "epoch": 1.69, + "grad_norm": 0.470703125, + "learning_rate": 1.184327435621988e-05, + "loss": 2.6343, + "step": 39605 + }, + { + "epoch": 1.69, + "grad_norm": 0.51953125, + "learning_rate": 1.184010801885289e-05, + "loss": 2.5301, + "step": 39606 + }, + { + "epoch": 1.69, + "grad_norm": 0.498046875, + "learning_rate": 1.1836942078173186e-05, + "loss": 2.5237, + "step": 39607 + }, + { + "epoch": 1.69, + "grad_norm": 0.46484375, + "learning_rate": 1.1833776534195017e-05, + "loss": 2.4707, + "step": 39608 + }, + { + "epoch": 1.69, + "grad_norm": 0.44921875, + "learning_rate": 1.1830611386932655e-05, + "loss": 2.4603, + "step": 39609 + }, + { + "epoch": 1.69, + "grad_norm": 0.40625, + "learning_rate": 1.1827446636400297e-05, + "loss": 2.3015, + "step": 39610 + }, + { + "epoch": 1.69, + "grad_norm": 0.5234375, + "learning_rate": 1.1824282282612232e-05, + "loss": 2.5229, + "step": 39611 + }, + { + "epoch": 1.69, + "grad_norm": 0.474609375, + "learning_rate": 1.1821118325582648e-05, + "loss": 2.4667, + "step": 39612 + }, + { + "epoch": 1.69, + "grad_norm": 0.3828125, + "learning_rate": 1.1817954765325823e-05, + "loss": 2.279, + "step": 39613 + }, + { + "epoch": 1.69, + "grad_norm": 0.447265625, + "learning_rate": 1.1814791601855957e-05, + "loss": 2.4327, + "step": 39614 + }, + { + "epoch": 1.69, + "grad_norm": 0.41015625, + "learning_rate": 1.1811628835187316e-05, + "loss": 2.3916, + "step": 39615 + }, + { + "epoch": 1.69, + "grad_norm": 0.435546875, + "learning_rate": 1.18084664653341e-05, + "loss": 2.2476, + "step": 39616 + }, + { + "epoch": 1.69, + "grad_norm": 0.478515625, + "learning_rate": 1.1805304492310565e-05, + "loss": 2.4048, + "step": 39617 + }, + { + "epoch": 1.69, + "grad_norm": 0.48828125, + "learning_rate": 1.1802142916130932e-05, + "loss": 2.4318, + "step": 39618 + }, + { + "epoch": 1.69, + "grad_norm": 0.490234375, + "learning_rate": 1.1798981736809434e-05, + "loss": 2.4748, + "step": 39619 + }, + { + "epoch": 1.69, + "grad_norm": 0.478515625, + "learning_rate": 1.1795820954360237e-05, + "loss": 2.4748, + "step": 39620 + }, + { + "epoch": 1.69, + "grad_norm": 0.451171875, + "learning_rate": 1.1792660568797652e-05, + "loss": 2.4783, + "step": 39621 + }, + { + "epoch": 1.69, + "grad_norm": 0.46875, + "learning_rate": 1.1789500580135859e-05, + "loss": 2.2913, + "step": 39622 + }, + { + "epoch": 1.69, + "grad_norm": 0.443359375, + "learning_rate": 1.1786340988389067e-05, + "loss": 2.3902, + "step": 39623 + }, + { + "epoch": 1.69, + "grad_norm": 0.412109375, + "learning_rate": 1.1783181793571484e-05, + "loss": 2.481, + "step": 39624 + }, + { + "epoch": 1.69, + "grad_norm": 0.45703125, + "learning_rate": 1.1780022995697371e-05, + "loss": 2.3501, + "step": 39625 + }, + { + "epoch": 1.69, + "grad_norm": 0.443359375, + "learning_rate": 1.1776864594780902e-05, + "loss": 2.4471, + "step": 39626 + }, + { + "epoch": 1.69, + "grad_norm": 0.46875, + "learning_rate": 1.1773706590836308e-05, + "loss": 2.5059, + "step": 39627 + }, + { + "epoch": 1.69, + "grad_norm": 0.4921875, + "learning_rate": 1.1770548983877782e-05, + "loss": 2.4725, + "step": 39628 + }, + { + "epoch": 1.69, + "grad_norm": 0.458984375, + "learning_rate": 1.1767391773919556e-05, + "loss": 2.4947, + "step": 39629 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.1764234960975795e-05, + "loss": 2.3772, + "step": 39630 + }, + { + "epoch": 1.69, + "grad_norm": 0.419921875, + "learning_rate": 1.1761078545060766e-05, + "loss": 2.3538, + "step": 39631 + }, + { + "epoch": 1.69, + "grad_norm": 0.458984375, + "learning_rate": 1.1757922526188603e-05, + "loss": 2.2571, + "step": 39632 + }, + { + "epoch": 1.69, + "grad_norm": 0.4453125, + "learning_rate": 1.1754766904373537e-05, + "loss": 2.322, + "step": 39633 + }, + { + "epoch": 1.69, + "grad_norm": 0.4609375, + "learning_rate": 1.1751611679629804e-05, + "loss": 2.4692, + "step": 39634 + }, + { + "epoch": 1.69, + "grad_norm": 0.54296875, + "learning_rate": 1.1748456851971557e-05, + "loss": 2.5098, + "step": 39635 + }, + { + "epoch": 1.69, + "grad_norm": 0.53515625, + "learning_rate": 1.1745302421412962e-05, + "loss": 2.4781, + "step": 39636 + }, + { + "epoch": 1.69, + "grad_norm": 0.4609375, + "learning_rate": 1.1742148387968298e-05, + "loss": 2.4914, + "step": 39637 + }, + { + "epoch": 1.69, + "grad_norm": 0.5, + "learning_rate": 1.173899475165171e-05, + "loss": 2.4241, + "step": 39638 + }, + { + "epoch": 1.69, + "grad_norm": 0.48828125, + "learning_rate": 1.1735841512477375e-05, + "loss": 2.5265, + "step": 39639 + }, + { + "epoch": 1.69, + "grad_norm": 0.486328125, + "learning_rate": 1.1732688670459491e-05, + "loss": 2.5037, + "step": 39640 + }, + { + "epoch": 1.69, + "grad_norm": 0.44921875, + "learning_rate": 1.172953622561227e-05, + "loss": 2.5253, + "step": 39641 + }, + { + "epoch": 1.69, + "grad_norm": 0.470703125, + "learning_rate": 1.1726384177949857e-05, + "loss": 2.4818, + "step": 39642 + }, + { + "epoch": 1.69, + "grad_norm": 0.4453125, + "learning_rate": 1.1723232527486472e-05, + "loss": 2.3601, + "step": 39643 + }, + { + "epoch": 1.69, + "grad_norm": 0.5, + "learning_rate": 1.172008127423626e-05, + "loss": 2.4805, + "step": 39644 + }, + { + "epoch": 1.69, + "grad_norm": 0.49609375, + "learning_rate": 1.1716930418213434e-05, + "loss": 2.5402, + "step": 39645 + }, + { + "epoch": 1.69, + "grad_norm": 0.546875, + "learning_rate": 1.1713779959432148e-05, + "loss": 2.4522, + "step": 39646 + }, + { + "epoch": 1.69, + "grad_norm": 0.46484375, + "learning_rate": 1.171062989790661e-05, + "loss": 2.5362, + "step": 39647 + }, + { + "epoch": 1.69, + "grad_norm": 0.419921875, + "learning_rate": 1.1707480233650936e-05, + "loss": 2.3898, + "step": 39648 + }, + { + "epoch": 1.69, + "grad_norm": 0.48828125, + "learning_rate": 1.1704330966679344e-05, + "loss": 2.548, + "step": 39649 + }, + { + "epoch": 1.69, + "grad_norm": 0.4609375, + "learning_rate": 1.1701182097006014e-05, + "loss": 2.431, + "step": 39650 + }, + { + "epoch": 1.69, + "grad_norm": 0.466796875, + "learning_rate": 1.16980336246451e-05, + "loss": 2.4182, + "step": 39651 + }, + { + "epoch": 1.69, + "grad_norm": 0.498046875, + "learning_rate": 1.1694885549610712e-05, + "loss": 2.418, + "step": 39652 + }, + { + "epoch": 1.69, + "grad_norm": 0.5, + "learning_rate": 1.1691737871917118e-05, + "loss": 2.5087, + "step": 39653 + }, + { + "epoch": 1.69, + "grad_norm": 0.50390625, + "learning_rate": 1.1688590591578418e-05, + "loss": 2.5051, + "step": 39654 + }, + { + "epoch": 1.69, + "grad_norm": 0.50390625, + "learning_rate": 1.168544370860878e-05, + "loss": 2.3238, + "step": 39655 + }, + { + "epoch": 1.69, + "grad_norm": 0.478515625, + "learning_rate": 1.1682297223022388e-05, + "loss": 2.5445, + "step": 39656 + }, + { + "epoch": 1.69, + "grad_norm": 0.515625, + "learning_rate": 1.1679151134833355e-05, + "loss": 2.4886, + "step": 39657 + }, + { + "epoch": 1.69, + "grad_norm": 0.4921875, + "learning_rate": 1.167600544405587e-05, + "loss": 2.4732, + "step": 39658 + }, + { + "epoch": 1.69, + "grad_norm": 0.4453125, + "learning_rate": 1.16728601507041e-05, + "loss": 2.5562, + "step": 39659 + }, + { + "epoch": 1.69, + "grad_norm": 0.43359375, + "learning_rate": 1.166971525479218e-05, + "loss": 2.4852, + "step": 39660 + }, + { + "epoch": 1.69, + "grad_norm": 0.458984375, + "learning_rate": 1.1666570756334239e-05, + "loss": 2.363, + "step": 39661 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.1663426655344445e-05, + "loss": 2.4569, + "step": 39662 + }, + { + "epoch": 1.69, + "grad_norm": 0.5, + "learning_rate": 1.1660282951836966e-05, + "loss": 2.4623, + "step": 39663 + }, + { + "epoch": 1.69, + "grad_norm": 0.4296875, + "learning_rate": 1.1657139645825909e-05, + "loss": 2.5096, + "step": 39664 + }, + { + "epoch": 1.69, + "grad_norm": 0.435546875, + "learning_rate": 1.1653996737325446e-05, + "loss": 2.4967, + "step": 39665 + }, + { + "epoch": 1.69, + "grad_norm": 0.482421875, + "learning_rate": 1.1650854226349728e-05, + "loss": 2.3678, + "step": 39666 + }, + { + "epoch": 1.69, + "grad_norm": 0.376953125, + "learning_rate": 1.1647712112912878e-05, + "loss": 2.4342, + "step": 39667 + }, + { + "epoch": 1.69, + "grad_norm": 0.4453125, + "learning_rate": 1.164457039702902e-05, + "loss": 2.3609, + "step": 39668 + }, + { + "epoch": 1.69, + "grad_norm": 0.4296875, + "learning_rate": 1.1641429078712296e-05, + "loss": 2.4733, + "step": 39669 + }, + { + "epoch": 1.69, + "grad_norm": 0.462890625, + "learning_rate": 1.1638288157976873e-05, + "loss": 2.3836, + "step": 39670 + }, + { + "epoch": 1.69, + "grad_norm": 0.474609375, + "learning_rate": 1.163514763483684e-05, + "loss": 2.4052, + "step": 39671 + }, + { + "epoch": 1.69, + "grad_norm": 0.5390625, + "learning_rate": 1.1632007509306365e-05, + "loss": 2.4059, + "step": 39672 + }, + { + "epoch": 1.69, + "grad_norm": 0.451171875, + "learning_rate": 1.1628867781399543e-05, + "loss": 2.4496, + "step": 39673 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.1625728451130525e-05, + "loss": 2.3926, + "step": 39674 + }, + { + "epoch": 1.69, + "grad_norm": 0.4375, + "learning_rate": 1.1622589518513449e-05, + "loss": 2.465, + "step": 39675 + }, + { + "epoch": 1.69, + "grad_norm": 0.470703125, + "learning_rate": 1.161945098356243e-05, + "loss": 2.5549, + "step": 39676 + }, + { + "epoch": 1.69, + "grad_norm": 0.421875, + "learning_rate": 1.1616312846291555e-05, + "loss": 2.5019, + "step": 39677 + }, + { + "epoch": 1.69, + "grad_norm": 0.482421875, + "learning_rate": 1.1613175106714968e-05, + "loss": 2.5448, + "step": 39678 + }, + { + "epoch": 1.69, + "grad_norm": 0.51953125, + "learning_rate": 1.1610037764846814e-05, + "loss": 2.4673, + "step": 39679 + }, + { + "epoch": 1.69, + "grad_norm": 0.48046875, + "learning_rate": 1.1606900820701172e-05, + "loss": 2.5112, + "step": 39680 + }, + { + "epoch": 1.69, + "grad_norm": 0.431640625, + "learning_rate": 1.160376427429216e-05, + "loss": 2.3837, + "step": 39681 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.160062812563394e-05, + "loss": 2.4216, + "step": 39682 + }, + { + "epoch": 1.69, + "grad_norm": 0.42578125, + "learning_rate": 1.1597492374740571e-05, + "loss": 2.3347, + "step": 39683 + }, + { + "epoch": 1.69, + "grad_norm": 0.482421875, + "learning_rate": 1.1594357021626168e-05, + "loss": 2.4143, + "step": 39684 + }, + { + "epoch": 1.69, + "grad_norm": 0.484375, + "learning_rate": 1.1591222066304852e-05, + "loss": 2.4612, + "step": 39685 + }, + { + "epoch": 1.69, + "grad_norm": 0.45703125, + "learning_rate": 1.1588087508790746e-05, + "loss": 2.5174, + "step": 39686 + }, + { + "epoch": 1.69, + "grad_norm": 0.4375, + "learning_rate": 1.1584953349097915e-05, + "loss": 2.3022, + "step": 39687 + }, + { + "epoch": 1.69, + "grad_norm": 0.42578125, + "learning_rate": 1.1581819587240494e-05, + "loss": 2.4289, + "step": 39688 + }, + { + "epoch": 1.69, + "grad_norm": 0.50390625, + "learning_rate": 1.157868622323256e-05, + "loss": 2.4423, + "step": 39689 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.1575553257088223e-05, + "loss": 2.5302, + "step": 39690 + }, + { + "epoch": 1.69, + "grad_norm": 0.482421875, + "learning_rate": 1.1572420688821595e-05, + "loss": 2.3843, + "step": 39691 + }, + { + "epoch": 1.69, + "grad_norm": 0.458984375, + "learning_rate": 1.1569288518446763e-05, + "loss": 2.2811, + "step": 39692 + }, + { + "epoch": 1.69, + "grad_norm": 0.390625, + "learning_rate": 1.1566156745977796e-05, + "loss": 2.5197, + "step": 39693 + }, + { + "epoch": 1.69, + "grad_norm": 0.455078125, + "learning_rate": 1.1563025371428804e-05, + "loss": 2.3358, + "step": 39694 + }, + { + "epoch": 1.69, + "grad_norm": 0.439453125, + "learning_rate": 1.1559894394813898e-05, + "loss": 2.4916, + "step": 39695 + }, + { + "epoch": 1.69, + "grad_norm": 0.474609375, + "learning_rate": 1.155676381614712e-05, + "loss": 2.3745, + "step": 39696 + }, + { + "epoch": 1.69, + "grad_norm": 0.48828125, + "learning_rate": 1.1553633635442585e-05, + "loss": 2.5285, + "step": 39697 + }, + { + "epoch": 1.69, + "grad_norm": 0.431640625, + "learning_rate": 1.1550503852714401e-05, + "loss": 2.4524, + "step": 39698 + }, + { + "epoch": 1.69, + "grad_norm": 0.427734375, + "learning_rate": 1.1547374467976624e-05, + "loss": 2.3553, + "step": 39699 + }, + { + "epoch": 1.69, + "grad_norm": 0.4765625, + "learning_rate": 1.1544245481243309e-05, + "loss": 2.4534, + "step": 39700 + }, + { + "epoch": 1.69, + "grad_norm": 0.49609375, + "learning_rate": 1.154111689252857e-05, + "loss": 2.4749, + "step": 39701 + }, + { + "epoch": 1.69, + "grad_norm": 0.494140625, + "learning_rate": 1.1537988701846492e-05, + "loss": 2.3086, + "step": 39702 + }, + { + "epoch": 1.69, + "grad_norm": 0.44921875, + "learning_rate": 1.153486090921111e-05, + "loss": 2.4304, + "step": 39703 + }, + { + "epoch": 1.69, + "grad_norm": 0.447265625, + "learning_rate": 1.1531733514636555e-05, + "loss": 2.4586, + "step": 39704 + }, + { + "epoch": 1.69, + "grad_norm": 0.4375, + "learning_rate": 1.1528606518136842e-05, + "loss": 2.4229, + "step": 39705 + }, + { + "epoch": 1.69, + "grad_norm": 0.4375, + "learning_rate": 1.1525479919726068e-05, + "loss": 2.31, + "step": 39706 + }, + { + "epoch": 1.69, + "grad_norm": 0.390625, + "learning_rate": 1.1522353719418321e-05, + "loss": 2.4111, + "step": 39707 + }, + { + "epoch": 1.69, + "grad_norm": 0.4921875, + "learning_rate": 1.1519227917227649e-05, + "loss": 2.6156, + "step": 39708 + }, + { + "epoch": 1.69, + "grad_norm": 0.48046875, + "learning_rate": 1.1516102513168093e-05, + "loss": 2.3977, + "step": 39709 + }, + { + "epoch": 1.69, + "grad_norm": 0.44921875, + "learning_rate": 1.1512977507253741e-05, + "loss": 2.2551, + "step": 39710 + }, + { + "epoch": 1.69, + "grad_norm": 0.5078125, + "learning_rate": 1.1509852899498675e-05, + "loss": 2.4247, + "step": 39711 + }, + { + "epoch": 1.69, + "grad_norm": 0.49609375, + "learning_rate": 1.1506728689916902e-05, + "loss": 2.545, + "step": 39712 + }, + { + "epoch": 1.69, + "grad_norm": 0.51953125, + "learning_rate": 1.1503604878522512e-05, + "loss": 2.4007, + "step": 39713 + }, + { + "epoch": 1.69, + "grad_norm": 0.484375, + "learning_rate": 1.1500481465329582e-05, + "loss": 2.4219, + "step": 39714 + }, + { + "epoch": 1.69, + "grad_norm": 0.46484375, + "learning_rate": 1.1497358450352147e-05, + "loss": 2.2945, + "step": 39715 + }, + { + "epoch": 1.69, + "grad_norm": 0.427734375, + "learning_rate": 1.1494235833604227e-05, + "loss": 2.4417, + "step": 39716 + }, + { + "epoch": 1.69, + "grad_norm": 0.44921875, + "learning_rate": 1.1491113615099902e-05, + "loss": 2.5207, + "step": 39717 + }, + { + "epoch": 1.69, + "grad_norm": 0.5, + "learning_rate": 1.1487991794853237e-05, + "loss": 2.4588, + "step": 39718 + }, + { + "epoch": 1.69, + "grad_norm": 0.46875, + "learning_rate": 1.1484870372878242e-05, + "loss": 2.4437, + "step": 39719 + }, + { + "epoch": 1.69, + "grad_norm": 0.478515625, + "learning_rate": 1.1481749349189009e-05, + "loss": 2.4782, + "step": 39720 + }, + { + "epoch": 1.69, + "grad_norm": 0.45703125, + "learning_rate": 1.1478628723799534e-05, + "loss": 2.3588, + "step": 39721 + }, + { + "epoch": 1.69, + "grad_norm": 0.47265625, + "learning_rate": 1.1475508496723875e-05, + "loss": 2.4794, + "step": 39722 + }, + { + "epoch": 1.69, + "grad_norm": 0.443359375, + "learning_rate": 1.1472388667976097e-05, + "loss": 2.4833, + "step": 39723 + }, + { + "epoch": 1.69, + "grad_norm": 0.51171875, + "learning_rate": 1.1469269237570213e-05, + "loss": 2.5661, + "step": 39724 + }, + { + "epoch": 1.69, + "grad_norm": 0.46875, + "learning_rate": 1.146615020552022e-05, + "loss": 2.3428, + "step": 39725 + }, + { + "epoch": 1.69, + "grad_norm": 0.50390625, + "learning_rate": 1.1463031571840244e-05, + "loss": 2.3039, + "step": 39726 + }, + { + "epoch": 1.69, + "grad_norm": 0.48046875, + "learning_rate": 1.145991333654427e-05, + "loss": 2.3723, + "step": 39727 + }, + { + "epoch": 1.69, + "grad_norm": 0.478515625, + "learning_rate": 1.14567954996463e-05, + "loss": 2.4042, + "step": 39728 + }, + { + "epoch": 1.69, + "grad_norm": 0.423828125, + "learning_rate": 1.14536780611604e-05, + "loss": 2.434, + "step": 39729 + }, + { + "epoch": 1.69, + "grad_norm": 0.5078125, + "learning_rate": 1.1450561021100604e-05, + "loss": 2.2831, + "step": 39730 + }, + { + "epoch": 1.69, + "grad_norm": 0.484375, + "learning_rate": 1.144744437948092e-05, + "loss": 2.4585, + "step": 39731 + }, + { + "epoch": 1.69, + "grad_norm": 0.52734375, + "learning_rate": 1.1444328136315351e-05, + "loss": 2.4485, + "step": 39732 + }, + { + "epoch": 1.69, + "grad_norm": 0.494140625, + "learning_rate": 1.1441212291617953e-05, + "loss": 2.5214, + "step": 39733 + }, + { + "epoch": 1.69, + "grad_norm": 0.4296875, + "learning_rate": 1.1438096845402746e-05, + "loss": 2.4411, + "step": 39734 + }, + { + "epoch": 1.69, + "grad_norm": 0.5, + "learning_rate": 1.143498179768373e-05, + "loss": 2.5207, + "step": 39735 + }, + { + "epoch": 1.69, + "grad_norm": 0.46484375, + "learning_rate": 1.143186714847494e-05, + "loss": 2.4125, + "step": 39736 + }, + { + "epoch": 1.69, + "grad_norm": 0.5078125, + "learning_rate": 1.1428752897790362e-05, + "loss": 2.5258, + "step": 39737 + }, + { + "epoch": 1.69, + "grad_norm": 0.4765625, + "learning_rate": 1.1425639045644022e-05, + "loss": 2.3414, + "step": 39738 + }, + { + "epoch": 1.69, + "grad_norm": 0.44140625, + "learning_rate": 1.142252559204996e-05, + "loss": 2.525, + "step": 39739 + }, + { + "epoch": 1.69, + "grad_norm": 0.46875, + "learning_rate": 1.1419412537022155e-05, + "loss": 2.5528, + "step": 39740 + }, + { + "epoch": 1.69, + "grad_norm": 0.5, + "learning_rate": 1.1416299880574588e-05, + "loss": 2.5288, + "step": 39741 + }, + { + "epoch": 1.69, + "grad_norm": 0.439453125, + "learning_rate": 1.1413187622721344e-05, + "loss": 2.4358, + "step": 39742 + }, + { + "epoch": 1.69, + "grad_norm": 0.55859375, + "learning_rate": 1.1410075763476368e-05, + "loss": 2.5408, + "step": 39743 + }, + { + "epoch": 1.69, + "grad_norm": 0.416015625, + "learning_rate": 1.1406964302853662e-05, + "loss": 2.435, + "step": 39744 + }, + { + "epoch": 1.69, + "grad_norm": 0.484375, + "learning_rate": 1.1403853240867247e-05, + "loss": 2.4901, + "step": 39745 + }, + { + "epoch": 1.69, + "grad_norm": 0.51171875, + "learning_rate": 1.140074257753112e-05, + "loss": 2.4783, + "step": 39746 + }, + { + "epoch": 1.69, + "grad_norm": 0.5390625, + "learning_rate": 1.1397632312859275e-05, + "loss": 2.4413, + "step": 39747 + }, + { + "epoch": 1.69, + "grad_norm": 0.435546875, + "learning_rate": 1.1394522446865685e-05, + "loss": 2.4089, + "step": 39748 + }, + { + "epoch": 1.69, + "grad_norm": 0.486328125, + "learning_rate": 1.1391412979564364e-05, + "loss": 2.5039, + "step": 39749 + }, + { + "epoch": 1.69, + "grad_norm": 0.4921875, + "learning_rate": 1.1388303910969334e-05, + "loss": 2.5628, + "step": 39750 + }, + { + "epoch": 1.69, + "grad_norm": 0.447265625, + "learning_rate": 1.1385195241094515e-05, + "loss": 2.5852, + "step": 39751 + }, + { + "epoch": 1.69, + "grad_norm": 0.478515625, + "learning_rate": 1.1382086969953954e-05, + "loss": 2.342, + "step": 39752 + }, + { + "epoch": 1.69, + "grad_norm": 0.4765625, + "learning_rate": 1.1378979097561604e-05, + "loss": 2.4886, + "step": 39753 + }, + { + "epoch": 1.69, + "grad_norm": 0.458984375, + "learning_rate": 1.1375871623931455e-05, + "loss": 2.4667, + "step": 39754 + }, + { + "epoch": 1.69, + "grad_norm": 0.466796875, + "learning_rate": 1.137276454907752e-05, + "loss": 2.4174, + "step": 39755 + }, + { + "epoch": 1.69, + "grad_norm": 0.455078125, + "learning_rate": 1.136965787301375e-05, + "loss": 2.4429, + "step": 39756 + }, + { + "epoch": 1.69, + "grad_norm": 0.515625, + "learning_rate": 1.1366551595754094e-05, + "loss": 2.5109, + "step": 39757 + }, + { + "epoch": 1.69, + "grad_norm": 0.435546875, + "learning_rate": 1.1363445717312592e-05, + "loss": 2.4372, + "step": 39758 + }, + { + "epoch": 1.69, + "grad_norm": 0.45703125, + "learning_rate": 1.1360340237703205e-05, + "loss": 2.4712, + "step": 39759 + }, + { + "epoch": 1.69, + "grad_norm": 0.453125, + "learning_rate": 1.1357235156939871e-05, + "loss": 2.4423, + "step": 39760 + }, + { + "epoch": 1.69, + "grad_norm": 0.470703125, + "learning_rate": 1.1354130475036572e-05, + "loss": 2.4973, + "step": 39761 + }, + { + "epoch": 1.69, + "grad_norm": 0.46484375, + "learning_rate": 1.135102619200733e-05, + "loss": 2.549, + "step": 39762 + }, + { + "epoch": 1.69, + "grad_norm": 0.4453125, + "learning_rate": 1.1347922307866055e-05, + "loss": 2.3635, + "step": 39763 + }, + { + "epoch": 1.69, + "grad_norm": 0.427734375, + "learning_rate": 1.1344818822626712e-05, + "loss": 2.4992, + "step": 39764 + }, + { + "epoch": 1.69, + "grad_norm": 0.462890625, + "learning_rate": 1.1341715736303293e-05, + "loss": 2.4881, + "step": 39765 + }, + { + "epoch": 1.69, + "grad_norm": 0.48828125, + "learning_rate": 1.1338613048909774e-05, + "loss": 2.4276, + "step": 39766 + }, + { + "epoch": 1.69, + "grad_norm": 0.52734375, + "learning_rate": 1.1335510760460067e-05, + "loss": 2.4751, + "step": 39767 + }, + { + "epoch": 1.69, + "grad_norm": 0.39453125, + "learning_rate": 1.1332408870968181e-05, + "loss": 2.5002, + "step": 39768 + }, + { + "epoch": 1.69, + "grad_norm": 0.48828125, + "learning_rate": 1.1329307380448029e-05, + "loss": 2.416, + "step": 39769 + }, + { + "epoch": 1.69, + "grad_norm": 0.474609375, + "learning_rate": 1.1326206288913589e-05, + "loss": 2.5049, + "step": 39770 + }, + { + "epoch": 1.69, + "grad_norm": 0.490234375, + "learning_rate": 1.1323105596378836e-05, + "loss": 2.528, + "step": 39771 + }, + { + "epoch": 1.69, + "grad_norm": 0.439453125, + "learning_rate": 1.1320005302857695e-05, + "loss": 2.4932, + "step": 39772 + }, + { + "epoch": 1.69, + "grad_norm": 0.41796875, + "learning_rate": 1.1316905408364086e-05, + "loss": 2.3163, + "step": 39773 + }, + { + "epoch": 1.69, + "grad_norm": 0.4765625, + "learning_rate": 1.1313805912912035e-05, + "loss": 2.5666, + "step": 39774 + }, + { + "epoch": 1.69, + "grad_norm": 0.466796875, + "learning_rate": 1.1310706816515437e-05, + "loss": 2.496, + "step": 39775 + }, + { + "epoch": 1.69, + "grad_norm": 0.51171875, + "learning_rate": 1.1307608119188229e-05, + "loss": 2.4641, + "step": 39776 + }, + { + "epoch": 1.7, + "grad_norm": 0.443359375, + "learning_rate": 1.1304509820944398e-05, + "loss": 2.4264, + "step": 39777 + }, + { + "epoch": 1.7, + "grad_norm": 0.453125, + "learning_rate": 1.1301411921797833e-05, + "loss": 2.2923, + "step": 39778 + }, + { + "epoch": 1.7, + "grad_norm": 0.447265625, + "learning_rate": 1.1298314421762502e-05, + "loss": 2.4333, + "step": 39779 + }, + { + "epoch": 1.7, + "grad_norm": 0.515625, + "learning_rate": 1.1295217320852358e-05, + "loss": 2.5743, + "step": 39780 + }, + { + "epoch": 1.7, + "grad_norm": 0.46875, + "learning_rate": 1.1292120619081314e-05, + "loss": 2.5403, + "step": 39781 + }, + { + "epoch": 1.7, + "grad_norm": 0.515625, + "learning_rate": 1.1289024316463292e-05, + "loss": 2.568, + "step": 39782 + }, + { + "epoch": 1.7, + "grad_norm": 0.490234375, + "learning_rate": 1.1285928413012237e-05, + "loss": 2.3733, + "step": 39783 + }, + { + "epoch": 1.7, + "grad_norm": 0.50390625, + "learning_rate": 1.1282832908742114e-05, + "loss": 2.4844, + "step": 39784 + }, + { + "epoch": 1.7, + "grad_norm": 0.474609375, + "learning_rate": 1.1279737803666813e-05, + "loss": 2.4294, + "step": 39785 + }, + { + "epoch": 1.7, + "grad_norm": 0.4609375, + "learning_rate": 1.127664309780022e-05, + "loss": 2.4832, + "step": 39786 + }, + { + "epoch": 1.7, + "grad_norm": 0.486328125, + "learning_rate": 1.1273548791156363e-05, + "loss": 2.452, + "step": 39787 + }, + { + "epoch": 1.7, + "grad_norm": 0.5234375, + "learning_rate": 1.1270454883749104e-05, + "loss": 2.4312, + "step": 39788 + }, + { + "epoch": 1.7, + "grad_norm": 0.45703125, + "learning_rate": 1.1267361375592344e-05, + "loss": 2.2374, + "step": 39789 + }, + { + "epoch": 1.7, + "grad_norm": 0.46484375, + "learning_rate": 1.126426826670004e-05, + "loss": 2.3997, + "step": 39790 + }, + { + "epoch": 1.7, + "grad_norm": 0.42578125, + "learning_rate": 1.1261175557086112e-05, + "loss": 2.4406, + "step": 39791 + }, + { + "epoch": 1.7, + "grad_norm": 0.455078125, + "learning_rate": 1.125808324676445e-05, + "loss": 2.4127, + "step": 39792 + }, + { + "epoch": 1.7, + "grad_norm": 0.462890625, + "learning_rate": 1.1254991335748998e-05, + "loss": 2.3762, + "step": 39793 + }, + { + "epoch": 1.7, + "grad_norm": 0.48046875, + "learning_rate": 1.1251899824053635e-05, + "loss": 2.5128, + "step": 39794 + }, + { + "epoch": 1.7, + "grad_norm": 0.466796875, + "learning_rate": 1.1248808711692293e-05, + "loss": 2.4557, + "step": 39795 + }, + { + "epoch": 1.7, + "grad_norm": 0.4609375, + "learning_rate": 1.1245717998678895e-05, + "loss": 2.5721, + "step": 39796 + }, + { + "epoch": 1.7, + "grad_norm": 0.408203125, + "learning_rate": 1.1242627685027329e-05, + "loss": 2.3281, + "step": 39797 + }, + { + "epoch": 1.7, + "grad_norm": 0.515625, + "learning_rate": 1.1239537770751473e-05, + "loss": 2.3826, + "step": 39798 + }, + { + "epoch": 1.7, + "grad_norm": 0.462890625, + "learning_rate": 1.1236448255865273e-05, + "loss": 2.5381, + "step": 39799 + }, + { + "epoch": 1.7, + "grad_norm": 0.546875, + "learning_rate": 1.1233359140382627e-05, + "loss": 2.4445, + "step": 39800 + }, + { + "epoch": 1.7, + "grad_norm": 0.455078125, + "learning_rate": 1.1230270424317424e-05, + "loss": 2.405, + "step": 39801 + }, + { + "epoch": 1.7, + "grad_norm": 0.470703125, + "learning_rate": 1.1227182107683521e-05, + "loss": 2.3366, + "step": 39802 + }, + { + "epoch": 1.7, + "grad_norm": 0.484375, + "learning_rate": 1.1224094190494905e-05, + "loss": 2.4583, + "step": 39803 + }, + { + "epoch": 1.7, + "grad_norm": 0.4453125, + "learning_rate": 1.1221006672765411e-05, + "loss": 2.4215, + "step": 39804 + }, + { + "epoch": 1.7, + "grad_norm": 0.51953125, + "learning_rate": 1.1217919554508926e-05, + "loss": 2.4512, + "step": 39805 + }, + { + "epoch": 1.7, + "grad_norm": 0.431640625, + "learning_rate": 1.1214832835739364e-05, + "loss": 2.4466, + "step": 39806 + }, + { + "epoch": 1.7, + "grad_norm": 0.451171875, + "learning_rate": 1.1211746516470612e-05, + "loss": 2.5156, + "step": 39807 + }, + { + "epoch": 1.7, + "grad_norm": 0.515625, + "learning_rate": 1.1208660596716546e-05, + "loss": 2.6012, + "step": 39808 + }, + { + "epoch": 1.7, + "grad_norm": 0.50390625, + "learning_rate": 1.1205575076491071e-05, + "loss": 2.4099, + "step": 39809 + }, + { + "epoch": 1.7, + "grad_norm": 0.4296875, + "learning_rate": 1.1202489955808037e-05, + "loss": 2.3057, + "step": 39810 + }, + { + "epoch": 1.7, + "grad_norm": 0.49609375, + "learning_rate": 1.1199405234681349e-05, + "loss": 2.5799, + "step": 39811 + }, + { + "epoch": 1.7, + "grad_norm": 0.443359375, + "learning_rate": 1.1196320913124903e-05, + "loss": 2.6009, + "step": 39812 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.1193236991152567e-05, + "loss": 2.5438, + "step": 39813 + }, + { + "epoch": 1.7, + "grad_norm": 0.453125, + "learning_rate": 1.1190153468778175e-05, + "loss": 2.5349, + "step": 39814 + }, + { + "epoch": 1.7, + "grad_norm": 0.396484375, + "learning_rate": 1.1187070346015649e-05, + "loss": 2.2308, + "step": 39815 + }, + { + "epoch": 1.7, + "grad_norm": 0.5625, + "learning_rate": 1.1183987622878867e-05, + "loss": 2.5286, + "step": 39816 + }, + { + "epoch": 1.7, + "grad_norm": 0.486328125, + "learning_rate": 1.1180905299381673e-05, + "loss": 2.4587, + "step": 39817 + }, + { + "epoch": 1.7, + "grad_norm": 0.494140625, + "learning_rate": 1.117782337553791e-05, + "loss": 2.2975, + "step": 39818 + }, + { + "epoch": 1.7, + "grad_norm": 0.482421875, + "learning_rate": 1.1174741851361525e-05, + "loss": 2.453, + "step": 39819 + }, + { + "epoch": 1.7, + "grad_norm": 0.474609375, + "learning_rate": 1.117166072686634e-05, + "loss": 2.474, + "step": 39820 + }, + { + "epoch": 1.7, + "grad_norm": 0.431640625, + "learning_rate": 1.1168580002066186e-05, + "loss": 2.401, + "step": 39821 + }, + { + "epoch": 1.7, + "grad_norm": 0.51953125, + "learning_rate": 1.1165499676974978e-05, + "loss": 2.641, + "step": 39822 + }, + { + "epoch": 1.7, + "grad_norm": 0.44140625, + "learning_rate": 1.1162419751606556e-05, + "loss": 2.5014, + "step": 39823 + }, + { + "epoch": 1.7, + "grad_norm": 0.462890625, + "learning_rate": 1.1159340225974768e-05, + "loss": 2.588, + "step": 39824 + }, + { + "epoch": 1.7, + "grad_norm": 0.39453125, + "learning_rate": 1.11562611000935e-05, + "loss": 2.4757, + "step": 39825 + }, + { + "epoch": 1.7, + "grad_norm": 0.50390625, + "learning_rate": 1.1153182373976579e-05, + "loss": 2.5206, + "step": 39826 + }, + { + "epoch": 1.7, + "grad_norm": 0.453125, + "learning_rate": 1.1150104047637854e-05, + "loss": 2.3446, + "step": 39827 + }, + { + "epoch": 1.7, + "grad_norm": 0.482421875, + "learning_rate": 1.114702612109122e-05, + "loss": 2.4329, + "step": 39828 + }, + { + "epoch": 1.7, + "grad_norm": 0.474609375, + "learning_rate": 1.1143948594350484e-05, + "loss": 2.5079, + "step": 39829 + }, + { + "epoch": 1.7, + "grad_norm": 0.48828125, + "learning_rate": 1.1140871467429492e-05, + "loss": 2.4583, + "step": 39830 + }, + { + "epoch": 1.7, + "grad_norm": 0.50390625, + "learning_rate": 1.1137794740342111e-05, + "loss": 2.4477, + "step": 39831 + }, + { + "epoch": 1.7, + "grad_norm": 0.48046875, + "learning_rate": 1.1134718413102196e-05, + "loss": 2.4581, + "step": 39832 + }, + { + "epoch": 1.7, + "grad_norm": 0.4453125, + "learning_rate": 1.1131642485723548e-05, + "loss": 2.3913, + "step": 39833 + }, + { + "epoch": 1.7, + "grad_norm": 0.46875, + "learning_rate": 1.1128566958220022e-05, + "loss": 2.5657, + "step": 39834 + }, + { + "epoch": 1.7, + "grad_norm": 0.4296875, + "learning_rate": 1.1125491830605495e-05, + "loss": 2.3688, + "step": 39835 + }, + { + "epoch": 1.7, + "grad_norm": 0.46875, + "learning_rate": 1.1122417102893778e-05, + "loss": 2.4156, + "step": 39836 + }, + { + "epoch": 1.7, + "grad_norm": 0.474609375, + "learning_rate": 1.1119342775098685e-05, + "loss": 2.4931, + "step": 39837 + }, + { + "epoch": 1.7, + "grad_norm": 0.41796875, + "learning_rate": 1.1116268847234056e-05, + "loss": 2.6541, + "step": 39838 + }, + { + "epoch": 1.7, + "grad_norm": 0.423828125, + "learning_rate": 1.1113195319313762e-05, + "loss": 2.4513, + "step": 39839 + }, + { + "epoch": 1.7, + "grad_norm": 0.4609375, + "learning_rate": 1.1110122191351579e-05, + "loss": 2.5178, + "step": 39840 + }, + { + "epoch": 1.7, + "grad_norm": 0.4296875, + "learning_rate": 1.1107049463361386e-05, + "loss": 2.3334, + "step": 39841 + }, + { + "epoch": 1.7, + "grad_norm": 0.384765625, + "learning_rate": 1.1103977135356969e-05, + "loss": 2.3588, + "step": 39842 + }, + { + "epoch": 1.7, + "grad_norm": 0.486328125, + "learning_rate": 1.1100905207352152e-05, + "loss": 2.4215, + "step": 39843 + }, + { + "epoch": 1.7, + "grad_norm": 0.443359375, + "learning_rate": 1.1097833679360803e-05, + "loss": 2.3146, + "step": 39844 + }, + { + "epoch": 1.7, + "grad_norm": 0.57421875, + "learning_rate": 1.1094762551396709e-05, + "loss": 2.6099, + "step": 39845 + }, + { + "epoch": 1.7, + "grad_norm": 0.4765625, + "learning_rate": 1.109169182347367e-05, + "loss": 2.4985, + "step": 39846 + }, + { + "epoch": 1.7, + "grad_norm": 0.419921875, + "learning_rate": 1.1088621495605533e-05, + "loss": 2.3296, + "step": 39847 + }, + { + "epoch": 1.7, + "grad_norm": 0.44140625, + "learning_rate": 1.108555156780613e-05, + "loss": 2.4023, + "step": 39848 + }, + { + "epoch": 1.7, + "grad_norm": 0.451171875, + "learning_rate": 1.1082482040089216e-05, + "loss": 2.6048, + "step": 39849 + }, + { + "epoch": 1.7, + "grad_norm": 0.396484375, + "learning_rate": 1.1079412912468645e-05, + "loss": 2.3574, + "step": 39850 + }, + { + "epoch": 1.7, + "grad_norm": 0.466796875, + "learning_rate": 1.1076344184958243e-05, + "loss": 2.4546, + "step": 39851 + }, + { + "epoch": 1.7, + "grad_norm": 0.46484375, + "learning_rate": 1.1073275857571785e-05, + "loss": 2.4575, + "step": 39852 + }, + { + "epoch": 1.7, + "grad_norm": 0.54296875, + "learning_rate": 1.1070207930323073e-05, + "loss": 2.648, + "step": 39853 + }, + { + "epoch": 1.7, + "grad_norm": 0.52734375, + "learning_rate": 1.1067140403225917e-05, + "loss": 2.5593, + "step": 39854 + }, + { + "epoch": 1.7, + "grad_norm": 0.4765625, + "learning_rate": 1.106407327629415e-05, + "loss": 2.4405, + "step": 39855 + }, + { + "epoch": 1.7, + "grad_norm": 0.46875, + "learning_rate": 1.1061006549541542e-05, + "loss": 2.5937, + "step": 39856 + }, + { + "epoch": 1.7, + "grad_norm": 0.474609375, + "learning_rate": 1.105794022298191e-05, + "loss": 2.3836, + "step": 39857 + }, + { + "epoch": 1.7, + "grad_norm": 0.474609375, + "learning_rate": 1.1054874296629026e-05, + "loss": 2.5164, + "step": 39858 + }, + { + "epoch": 1.7, + "grad_norm": 0.466796875, + "learning_rate": 1.1051808770496707e-05, + "loss": 2.3744, + "step": 39859 + }, + { + "epoch": 1.7, + "grad_norm": 0.46484375, + "learning_rate": 1.1048743644598758e-05, + "loss": 2.4143, + "step": 39860 + }, + { + "epoch": 1.7, + "grad_norm": 0.48828125, + "learning_rate": 1.1045678918948955e-05, + "loss": 2.3451, + "step": 39861 + }, + { + "epoch": 1.7, + "grad_norm": 0.447265625, + "learning_rate": 1.1042614593561062e-05, + "loss": 2.4007, + "step": 39862 + }, + { + "epoch": 1.7, + "grad_norm": 0.451171875, + "learning_rate": 1.1039550668448895e-05, + "loss": 2.4971, + "step": 39863 + }, + { + "epoch": 1.7, + "grad_norm": 0.4609375, + "learning_rate": 1.1036487143626262e-05, + "loss": 2.4813, + "step": 39864 + }, + { + "epoch": 1.7, + "grad_norm": 0.484375, + "learning_rate": 1.1033424019106908e-05, + "loss": 2.4894, + "step": 39865 + }, + { + "epoch": 1.7, + "grad_norm": 0.51953125, + "learning_rate": 1.1030361294904623e-05, + "loss": 2.4714, + "step": 39866 + }, + { + "epoch": 1.7, + "grad_norm": 0.4609375, + "learning_rate": 1.1027298971033229e-05, + "loss": 2.523, + "step": 39867 + }, + { + "epoch": 1.7, + "grad_norm": 0.490234375, + "learning_rate": 1.1024237047506458e-05, + "loss": 2.5919, + "step": 39868 + }, + { + "epoch": 1.7, + "grad_norm": 0.42578125, + "learning_rate": 1.1021175524338101e-05, + "loss": 2.4321, + "step": 39869 + }, + { + "epoch": 1.7, + "grad_norm": 0.486328125, + "learning_rate": 1.1018114401541924e-05, + "loss": 2.393, + "step": 39870 + }, + { + "epoch": 1.7, + "grad_norm": 0.4453125, + "learning_rate": 1.1015053679131737e-05, + "loss": 2.2829, + "step": 39871 + }, + { + "epoch": 1.7, + "grad_norm": 0.515625, + "learning_rate": 1.1011993357121264e-05, + "loss": 2.6032, + "step": 39872 + }, + { + "epoch": 1.7, + "grad_norm": 0.455078125, + "learning_rate": 1.1008933435524327e-05, + "loss": 2.345, + "step": 39873 + }, + { + "epoch": 1.7, + "grad_norm": 0.46484375, + "learning_rate": 1.1005873914354647e-05, + "loss": 2.4719, + "step": 39874 + }, + { + "epoch": 1.7, + "grad_norm": 0.4765625, + "learning_rate": 1.1002814793626004e-05, + "loss": 2.5537, + "step": 39875 + }, + { + "epoch": 1.7, + "grad_norm": 0.484375, + "learning_rate": 1.0999756073352185e-05, + "loss": 2.5305, + "step": 39876 + }, + { + "epoch": 1.7, + "grad_norm": 0.546875, + "learning_rate": 1.0996697753546947e-05, + "loss": 2.4961, + "step": 39877 + }, + { + "epoch": 1.7, + "grad_norm": 0.55859375, + "learning_rate": 1.0993639834224023e-05, + "loss": 2.4709, + "step": 39878 + }, + { + "epoch": 1.7, + "grad_norm": 0.435546875, + "learning_rate": 1.099058231539719e-05, + "loss": 2.3127, + "step": 39879 + }, + { + "epoch": 1.7, + "grad_norm": 0.486328125, + "learning_rate": 1.0987525197080228e-05, + "loss": 2.4127, + "step": 39880 + }, + { + "epoch": 1.7, + "grad_norm": 0.462890625, + "learning_rate": 1.0984468479286846e-05, + "loss": 2.3724, + "step": 39881 + }, + { + "epoch": 1.7, + "grad_norm": 0.42578125, + "learning_rate": 1.0981412162030836e-05, + "loss": 2.3719, + "step": 39882 + }, + { + "epoch": 1.7, + "grad_norm": 0.453125, + "learning_rate": 1.0978356245325949e-05, + "loss": 2.405, + "step": 39883 + }, + { + "epoch": 1.7, + "grad_norm": 0.54296875, + "learning_rate": 1.0975300729185934e-05, + "loss": 2.5313, + "step": 39884 + }, + { + "epoch": 1.7, + "grad_norm": 0.5234375, + "learning_rate": 1.0972245613624499e-05, + "loss": 2.4748, + "step": 39885 + }, + { + "epoch": 1.7, + "grad_norm": 0.5546875, + "learning_rate": 1.0969190898655436e-05, + "loss": 2.6018, + "step": 39886 + }, + { + "epoch": 1.7, + "grad_norm": 0.51171875, + "learning_rate": 1.0966136584292485e-05, + "loss": 2.564, + "step": 39887 + }, + { + "epoch": 1.7, + "grad_norm": 0.546875, + "learning_rate": 1.0963082670549363e-05, + "loss": 2.5767, + "step": 39888 + }, + { + "epoch": 1.7, + "grad_norm": 0.443359375, + "learning_rate": 1.0960029157439855e-05, + "loss": 2.2549, + "step": 39889 + }, + { + "epoch": 1.7, + "grad_norm": 0.39453125, + "learning_rate": 1.095697604497765e-05, + "loss": 2.4097, + "step": 39890 + }, + { + "epoch": 1.7, + "grad_norm": 0.447265625, + "learning_rate": 1.0953923333176518e-05, + "loss": 2.3546, + "step": 39891 + }, + { + "epoch": 1.7, + "grad_norm": 0.5234375, + "learning_rate": 1.0950871022050202e-05, + "loss": 2.457, + "step": 39892 + }, + { + "epoch": 1.7, + "grad_norm": 0.408203125, + "learning_rate": 1.0947819111612434e-05, + "loss": 2.4493, + "step": 39893 + }, + { + "epoch": 1.7, + "grad_norm": 0.416015625, + "learning_rate": 1.0944767601876905e-05, + "loss": 2.3506, + "step": 39894 + }, + { + "epoch": 1.7, + "grad_norm": 0.486328125, + "learning_rate": 1.0941716492857378e-05, + "loss": 2.4499, + "step": 39895 + }, + { + "epoch": 1.7, + "grad_norm": 0.474609375, + "learning_rate": 1.0938665784567593e-05, + "loss": 2.3509, + "step": 39896 + }, + { + "epoch": 1.7, + "grad_norm": 0.46484375, + "learning_rate": 1.0935615477021255e-05, + "loss": 2.4895, + "step": 39897 + }, + { + "epoch": 1.7, + "grad_norm": 0.484375, + "learning_rate": 1.0932565570232111e-05, + "loss": 2.492, + "step": 39898 + }, + { + "epoch": 1.7, + "grad_norm": 0.52734375, + "learning_rate": 1.0929516064213852e-05, + "loss": 2.4556, + "step": 39899 + }, + { + "epoch": 1.7, + "grad_norm": 0.439453125, + "learning_rate": 1.0926466958980242e-05, + "loss": 2.4553, + "step": 39900 + }, + { + "epoch": 1.7, + "grad_norm": 0.40625, + "learning_rate": 1.092341825454496e-05, + "loss": 2.4311, + "step": 39901 + }, + { + "epoch": 1.7, + "grad_norm": 0.451171875, + "learning_rate": 1.092036995092176e-05, + "loss": 2.5214, + "step": 39902 + }, + { + "epoch": 1.7, + "grad_norm": 0.51953125, + "learning_rate": 1.0917322048124312e-05, + "loss": 2.5736, + "step": 39903 + }, + { + "epoch": 1.7, + "grad_norm": 0.45703125, + "learning_rate": 1.091427454616637e-05, + "loss": 2.3938, + "step": 39904 + }, + { + "epoch": 1.7, + "grad_norm": 0.478515625, + "learning_rate": 1.0911227445061645e-05, + "loss": 2.4065, + "step": 39905 + }, + { + "epoch": 1.7, + "grad_norm": 0.47265625, + "learning_rate": 1.0908180744823848e-05, + "loss": 2.4498, + "step": 39906 + }, + { + "epoch": 1.7, + "grad_norm": 0.431640625, + "learning_rate": 1.0905134445466636e-05, + "loss": 2.7831, + "step": 39907 + }, + { + "epoch": 1.7, + "grad_norm": 0.53515625, + "learning_rate": 1.0902088547003797e-05, + "loss": 2.4736, + "step": 39908 + }, + { + "epoch": 1.7, + "grad_norm": 0.427734375, + "learning_rate": 1.0899043049448999e-05, + "loss": 2.5161, + "step": 39909 + }, + { + "epoch": 1.7, + "grad_norm": 0.494140625, + "learning_rate": 1.0895997952815928e-05, + "loss": 2.4547, + "step": 39910 + }, + { + "epoch": 1.7, + "grad_norm": 0.416015625, + "learning_rate": 1.08929532571183e-05, + "loss": 2.5032, + "step": 39911 + }, + { + "epoch": 1.7, + "grad_norm": 0.408203125, + "learning_rate": 1.0889908962369833e-05, + "loss": 2.3846, + "step": 39912 + }, + { + "epoch": 1.7, + "grad_norm": 0.42578125, + "learning_rate": 1.0886865068584195e-05, + "loss": 2.1791, + "step": 39913 + }, + { + "epoch": 1.7, + "grad_norm": 0.439453125, + "learning_rate": 1.0883821575775111e-05, + "loss": 2.5302, + "step": 39914 + }, + { + "epoch": 1.7, + "grad_norm": 0.455078125, + "learning_rate": 1.0880778483956244e-05, + "loss": 2.5582, + "step": 39915 + }, + { + "epoch": 1.7, + "grad_norm": 0.47265625, + "learning_rate": 1.087773579314133e-05, + "loss": 2.5031, + "step": 39916 + }, + { + "epoch": 1.7, + "grad_norm": 0.478515625, + "learning_rate": 1.0874693503344003e-05, + "loss": 2.5195, + "step": 39917 + }, + { + "epoch": 1.7, + "grad_norm": 0.4609375, + "learning_rate": 1.0871651614578015e-05, + "loss": 2.4196, + "step": 39918 + }, + { + "epoch": 1.7, + "grad_norm": 0.494140625, + "learning_rate": 1.0868610126856992e-05, + "loss": 2.6312, + "step": 39919 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.0865569040194656e-05, + "loss": 2.3825, + "step": 39920 + }, + { + "epoch": 1.7, + "grad_norm": 0.48828125, + "learning_rate": 1.0862528354604707e-05, + "loss": 2.4374, + "step": 39921 + }, + { + "epoch": 1.7, + "grad_norm": 0.4921875, + "learning_rate": 1.085948807010081e-05, + "loss": 2.4534, + "step": 39922 + }, + { + "epoch": 1.7, + "grad_norm": 0.439453125, + "learning_rate": 1.0856448186696588e-05, + "loss": 2.4329, + "step": 39923 + }, + { + "epoch": 1.7, + "grad_norm": 0.48046875, + "learning_rate": 1.085340870440581e-05, + "loss": 2.3338, + "step": 39924 + }, + { + "epoch": 1.7, + "grad_norm": 0.44140625, + "learning_rate": 1.0850369623242118e-05, + "loss": 2.598, + "step": 39925 + }, + { + "epoch": 1.7, + "grad_norm": 0.42578125, + "learning_rate": 1.0847330943219158e-05, + "loss": 2.602, + "step": 39926 + }, + { + "epoch": 1.7, + "grad_norm": 0.490234375, + "learning_rate": 1.0844292664350631e-05, + "loss": 2.4005, + "step": 39927 + }, + { + "epoch": 1.7, + "grad_norm": 0.4765625, + "learning_rate": 1.0841254786650223e-05, + "loss": 2.4088, + "step": 39928 + }, + { + "epoch": 1.7, + "grad_norm": 0.443359375, + "learning_rate": 1.0838217310131571e-05, + "loss": 2.4695, + "step": 39929 + }, + { + "epoch": 1.7, + "grad_norm": 0.46875, + "learning_rate": 1.0835180234808373e-05, + "loss": 2.4116, + "step": 39930 + }, + { + "epoch": 1.7, + "grad_norm": 0.470703125, + "learning_rate": 1.0832143560694252e-05, + "loss": 2.3918, + "step": 39931 + }, + { + "epoch": 1.7, + "grad_norm": 0.478515625, + "learning_rate": 1.082910728780292e-05, + "loss": 2.5171, + "step": 39932 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.0826071416147999e-05, + "loss": 2.5191, + "step": 39933 + }, + { + "epoch": 1.7, + "grad_norm": 0.474609375, + "learning_rate": 1.0823035945743187e-05, + "loss": 2.4566, + "step": 39934 + }, + { + "epoch": 1.7, + "grad_norm": 0.3984375, + "learning_rate": 1.082000087660211e-05, + "loss": 2.4895, + "step": 39935 + }, + { + "epoch": 1.7, + "grad_norm": 0.470703125, + "learning_rate": 1.0816966208738432e-05, + "loss": 2.4472, + "step": 39936 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.0813931942165823e-05, + "loss": 2.4866, + "step": 39937 + }, + { + "epoch": 1.7, + "grad_norm": 0.458984375, + "learning_rate": 1.0810898076897935e-05, + "loss": 2.429, + "step": 39938 + }, + { + "epoch": 1.7, + "grad_norm": 0.490234375, + "learning_rate": 1.080786461294837e-05, + "loss": 2.484, + "step": 39939 + }, + { + "epoch": 1.7, + "grad_norm": 0.416015625, + "learning_rate": 1.0804831550330864e-05, + "loss": 2.4902, + "step": 39940 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.0801798889059012e-05, + "loss": 2.4766, + "step": 39941 + }, + { + "epoch": 1.7, + "grad_norm": 0.439453125, + "learning_rate": 1.079876662914645e-05, + "loss": 2.3845, + "step": 39942 + }, + { + "epoch": 1.7, + "grad_norm": 0.53515625, + "learning_rate": 1.0795734770606835e-05, + "loss": 2.528, + "step": 39943 + }, + { + "epoch": 1.7, + "grad_norm": 0.43359375, + "learning_rate": 1.0792703313453845e-05, + "loss": 2.6091, + "step": 39944 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.0789672257701056e-05, + "loss": 2.4498, + "step": 39945 + }, + { + "epoch": 1.7, + "grad_norm": 0.451171875, + "learning_rate": 1.0786641603362158e-05, + "loss": 2.4502, + "step": 39946 + }, + { + "epoch": 1.7, + "grad_norm": 0.423828125, + "learning_rate": 1.0783611350450762e-05, + "loss": 2.4475, + "step": 39947 + }, + { + "epoch": 1.7, + "grad_norm": 0.515625, + "learning_rate": 1.0780581498980514e-05, + "loss": 2.5447, + "step": 39948 + }, + { + "epoch": 1.7, + "grad_norm": 0.490234375, + "learning_rate": 1.0777552048965034e-05, + "loss": 2.5629, + "step": 39949 + }, + { + "epoch": 1.7, + "grad_norm": 0.484375, + "learning_rate": 1.077452300041798e-05, + "loss": 2.3813, + "step": 39950 + }, + { + "epoch": 1.7, + "grad_norm": 0.49609375, + "learning_rate": 1.077149435335294e-05, + "loss": 2.4882, + "step": 39951 + }, + { + "epoch": 1.7, + "grad_norm": 0.435546875, + "learning_rate": 1.0768466107783571e-05, + "loss": 2.4573, + "step": 39952 + }, + { + "epoch": 1.7, + "grad_norm": 0.380859375, + "learning_rate": 1.0765438263723516e-05, + "loss": 2.4585, + "step": 39953 + }, + { + "epoch": 1.7, + "grad_norm": 0.458984375, + "learning_rate": 1.0762410821186375e-05, + "loss": 2.3504, + "step": 39954 + }, + { + "epoch": 1.7, + "grad_norm": 0.43359375, + "learning_rate": 1.0759383780185728e-05, + "loss": 2.4611, + "step": 39955 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.0756357140735285e-05, + "loss": 2.4064, + "step": 39956 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.0753330902848623e-05, + "loss": 2.5097, + "step": 39957 + }, + { + "epoch": 1.7, + "grad_norm": 0.4140625, + "learning_rate": 1.0750305066539335e-05, + "loss": 2.4928, + "step": 39958 + }, + { + "epoch": 1.7, + "grad_norm": 0.5, + "learning_rate": 1.074727963182105e-05, + "loss": 2.3854, + "step": 39959 + }, + { + "epoch": 1.7, + "grad_norm": 0.53125, + "learning_rate": 1.0744254598707414e-05, + "loss": 2.3998, + "step": 39960 + }, + { + "epoch": 1.7, + "grad_norm": 0.439453125, + "learning_rate": 1.0741229967211997e-05, + "loss": 2.4096, + "step": 39961 + }, + { + "epoch": 1.7, + "grad_norm": 0.46875, + "learning_rate": 1.0738205737348439e-05, + "loss": 2.5214, + "step": 39962 + }, + { + "epoch": 1.7, + "grad_norm": 0.478515625, + "learning_rate": 1.073518190913031e-05, + "loss": 2.395, + "step": 39963 + }, + { + "epoch": 1.7, + "grad_norm": 0.486328125, + "learning_rate": 1.0732158482571264e-05, + "loss": 2.4288, + "step": 39964 + }, + { + "epoch": 1.7, + "grad_norm": 0.45703125, + "learning_rate": 1.0729135457684869e-05, + "loss": 2.5542, + "step": 39965 + }, + { + "epoch": 1.7, + "grad_norm": 0.55859375, + "learning_rate": 1.0726112834484747e-05, + "loss": 2.4448, + "step": 39966 + }, + { + "epoch": 1.7, + "grad_norm": 0.494140625, + "learning_rate": 1.0723090612984476e-05, + "loss": 2.5696, + "step": 39967 + }, + { + "epoch": 1.7, + "grad_norm": 0.455078125, + "learning_rate": 1.072006879319768e-05, + "loss": 2.2825, + "step": 39968 + }, + { + "epoch": 1.7, + "grad_norm": 0.45703125, + "learning_rate": 1.0717047375137956e-05, + "loss": 2.3294, + "step": 39969 + }, + { + "epoch": 1.7, + "grad_norm": 0.53515625, + "learning_rate": 1.0714026358818885e-05, + "loss": 2.5505, + "step": 39970 + }, + { + "epoch": 1.7, + "grad_norm": 0.4140625, + "learning_rate": 1.0711005744254032e-05, + "loss": 2.5426, + "step": 39971 + }, + { + "epoch": 1.7, + "grad_norm": 0.486328125, + "learning_rate": 1.0707985531457066e-05, + "loss": 2.3242, + "step": 39972 + }, + { + "epoch": 1.7, + "grad_norm": 0.515625, + "learning_rate": 1.0704965720441528e-05, + "loss": 2.4116, + "step": 39973 + }, + { + "epoch": 1.7, + "grad_norm": 0.443359375, + "learning_rate": 1.070194631122099e-05, + "loss": 2.3841, + "step": 39974 + }, + { + "epoch": 1.7, + "grad_norm": 0.4296875, + "learning_rate": 1.0698927303809059e-05, + "loss": 2.4572, + "step": 39975 + }, + { + "epoch": 1.7, + "grad_norm": 0.48828125, + "learning_rate": 1.0695908698219337e-05, + "loss": 2.3507, + "step": 39976 + }, + { + "epoch": 1.7, + "grad_norm": 0.416015625, + "learning_rate": 1.0692890494465368e-05, + "loss": 2.5231, + "step": 39977 + }, + { + "epoch": 1.7, + "grad_norm": 0.453125, + "learning_rate": 1.0689872692560776e-05, + "loss": 2.5576, + "step": 39978 + }, + { + "epoch": 1.7, + "grad_norm": 0.4609375, + "learning_rate": 1.0686855292519104e-05, + "loss": 2.6151, + "step": 39979 + }, + { + "epoch": 1.7, + "grad_norm": 0.458984375, + "learning_rate": 1.0683838294353954e-05, + "loss": 2.4695, + "step": 39980 + }, + { + "epoch": 1.7, + "grad_norm": 0.53125, + "learning_rate": 1.068082169807888e-05, + "loss": 2.4428, + "step": 39981 + }, + { + "epoch": 1.7, + "grad_norm": 0.5234375, + "learning_rate": 1.0677805503707483e-05, + "loss": 2.4032, + "step": 39982 + }, + { + "epoch": 1.7, + "grad_norm": 0.427734375, + "learning_rate": 1.0674789711253286e-05, + "loss": 2.2922, + "step": 39983 + }, + { + "epoch": 1.7, + "grad_norm": 0.50390625, + "learning_rate": 1.0671774320729899e-05, + "loss": 2.398, + "step": 39984 + }, + { + "epoch": 1.7, + "grad_norm": 0.515625, + "learning_rate": 1.0668759332150902e-05, + "loss": 2.4687, + "step": 39985 + }, + { + "epoch": 1.7, + "grad_norm": 0.412109375, + "learning_rate": 1.0665744745529849e-05, + "loss": 2.3395, + "step": 39986 + }, + { + "epoch": 1.7, + "grad_norm": 0.455078125, + "learning_rate": 1.0662730560880251e-05, + "loss": 2.4645, + "step": 39987 + }, + { + "epoch": 1.7, + "grad_norm": 0.56640625, + "learning_rate": 1.0659716778215757e-05, + "loss": 2.3681, + "step": 39988 + }, + { + "epoch": 1.7, + "grad_norm": 0.42578125, + "learning_rate": 1.0656703397549883e-05, + "loss": 2.3632, + "step": 39989 + }, + { + "epoch": 1.7, + "grad_norm": 0.41015625, + "learning_rate": 1.0653690418896167e-05, + "loss": 2.3116, + "step": 39990 + }, + { + "epoch": 1.7, + "grad_norm": 0.47265625, + "learning_rate": 1.0650677842268197e-05, + "loss": 2.5199, + "step": 39991 + }, + { + "epoch": 1.7, + "grad_norm": 0.396484375, + "learning_rate": 1.0647665667679552e-05, + "loss": 2.4357, + "step": 39992 + }, + { + "epoch": 1.7, + "grad_norm": 0.51953125, + "learning_rate": 1.064465389514373e-05, + "loss": 2.4011, + "step": 39993 + }, + { + "epoch": 1.7, + "grad_norm": 0.462890625, + "learning_rate": 1.0641642524674334e-05, + "loss": 2.452, + "step": 39994 + }, + { + "epoch": 1.7, + "grad_norm": 0.4609375, + "learning_rate": 1.0638631556284861e-05, + "loss": 2.4739, + "step": 39995 + }, + { + "epoch": 1.7, + "grad_norm": 0.42578125, + "learning_rate": 1.0635620989988915e-05, + "loss": 2.3715, + "step": 39996 + }, + { + "epoch": 1.7, + "grad_norm": 0.43359375, + "learning_rate": 1.0632610825799994e-05, + "loss": 2.4827, + "step": 39997 + }, + { + "epoch": 1.7, + "grad_norm": 0.478515625, + "learning_rate": 1.0629601063731687e-05, + "loss": 2.4487, + "step": 39998 + }, + { + "epoch": 1.7, + "grad_norm": 0.46875, + "learning_rate": 1.0626591703797483e-05, + "loss": 2.587, + "step": 39999 + }, + { + "epoch": 1.7, + "grad_norm": 0.478515625, + "learning_rate": 1.0623582746010973e-05, + "loss": 2.536, + "step": 40000 + } + ], + "logging_steps": 1, + "max_steps": 46934, + "num_input_tokens_seen": 0, + "num_train_epochs": 2, + "save_steps": 4000, + "total_flos": 9.948308023552246e+19, + "train_batch_size": 2, + "trial_name": null, + "trial_params": null +}