diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13373 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 475,
+  "global_step": 1900,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005263157894736842,
+      "grad_norm": 1.3774428367614746,
+      "learning_rate": 7.5e-07,
+      "loss": 0.7461,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005263157894736842,
+      "eval_loss": 1.3115713596343994,
+      "eval_runtime": 13.3879,
+      "eval_samples_per_second": 7.469,
+      "eval_steps_per_second": 7.469,
+      "step": 1
+    },
+    {
+      "epoch": 0.0010526315789473684,
+      "grad_norm": 0.40974903106689453,
+      "learning_rate": 1.5e-06,
+      "loss": 1.2756,
+      "step": 2
+    },
+    {
+      "epoch": 0.0015789473684210526,
+      "grad_norm": 0.5177225470542908,
+      "learning_rate": 2.25e-06,
+      "loss": 1.7994,
+      "step": 3
+    },
+    {
+      "epoch": 0.002105263157894737,
+      "grad_norm": 0.5397021174430847,
+      "learning_rate": 3e-06,
+      "loss": 1.3129,
+      "step": 4
+    },
+    {
+      "epoch": 0.002631578947368421,
+      "grad_norm": 0.37644293904304504,
+      "learning_rate": 3.75e-06,
+      "loss": 1.2963,
+      "step": 5
+    },
+    {
+      "epoch": 0.003157894736842105,
+      "grad_norm": 0.6736767292022705,
+      "learning_rate": 4.5e-06,
+      "loss": 1.0839,
+      "step": 6
+    },
+    {
+      "epoch": 0.0036842105263157894,
+      "grad_norm": 0.9168409109115601,
+      "learning_rate": 5.25e-06,
+      "loss": 1.2321,
+      "step": 7
+    },
+    {
+      "epoch": 0.004210526315789474,
+      "grad_norm": 0.5211862325668335,
+      "learning_rate": 6e-06,
+      "loss": 1.567,
+      "step": 8
+    },
+    {
+      "epoch": 0.004736842105263158,
+      "grad_norm": 0.6043412685394287,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 1.236,
+      "step": 9
+    },
+    {
+      "epoch": 0.005263157894736842,
+      "grad_norm": 0.9153957366943359,
+      "learning_rate": 7.5e-06,
+      "loss": 1.9274,
+      "step": 10
+    },
+    {
+      "epoch": 0.005789473684210527,
+      "grad_norm": 0.3442760407924652,
+      "learning_rate": 8.25e-06,
+      "loss": 1.7819,
+      "step": 11
+    },
+    {
+      "epoch": 0.00631578947368421,
+      "grad_norm": 3.2623486518859863,
+      "learning_rate": 9e-06,
+      "loss": 3.5503,
+      "step": 12
+    },
+    {
+      "epoch": 0.006842105263157895,
+      "grad_norm": 0.5294060111045837,
+      "learning_rate": 9.75e-06,
+      "loss": 1.0113,
+      "step": 13
+    },
+    {
+      "epoch": 0.007368421052631579,
+      "grad_norm": 0.7050935626029968,
+      "learning_rate": 1.05e-05,
+      "loss": 0.8552,
+      "step": 14
+    },
+    {
+      "epoch": 0.007894736842105263,
+      "grad_norm": 0.3988778591156006,
+      "learning_rate": 1.125e-05,
+      "loss": 1.2478,
+      "step": 15
+    },
+    {
+      "epoch": 0.008421052631578947,
+      "grad_norm": 1.1330819129943848,
+      "learning_rate": 1.2e-05,
+      "loss": 0.5481,
+      "step": 16
+    },
+    {
+      "epoch": 0.008947368421052631,
+      "grad_norm": 0.3928195834159851,
+      "learning_rate": 1.275e-05,
+      "loss": 1.2705,
+      "step": 17
+    },
+    {
+      "epoch": 0.009473684210526316,
+      "grad_norm": 0.5050056576728821,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 1.0001,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.40563464164733887,
+      "learning_rate": 1.4249999999999999e-05,
+      "loss": 1.3443,
+      "step": 19
+    },
+    {
+      "epoch": 0.010526315789473684,
+      "grad_norm": 0.802213728427887,
+      "learning_rate": 1.5e-05,
+      "loss": 1.6883,
+      "step": 20
+    },
+    {
+      "epoch": 0.011052631578947368,
+      "grad_norm": 0.6669230461120605,
+      "learning_rate": 1.575e-05,
+      "loss": 1.8126,
+      "step": 21
+    },
+    {
+      "epoch": 0.011578947368421053,
+      "grad_norm": 0.7659299373626709,
+      "learning_rate": 1.65e-05,
+      "loss": 0.892,
+      "step": 22
+    },
+    {
+      "epoch": 0.012105263157894737,
+      "grad_norm": 0.42965367436408997,
+      "learning_rate": 1.725e-05,
+      "loss": 1.5116,
+      "step": 23
+    },
+    {
+      "epoch": 0.01263157894736842,
+      "grad_norm": 0.7102161049842834,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6337,
+      "step": 24
+    },
+    {
+      "epoch": 0.013157894736842105,
+      "grad_norm": 1.3091274499893188,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 1.0829,
+      "step": 25
+    },
+    {
+      "epoch": 0.01368421052631579,
+      "grad_norm": 4.387722969055176,
+      "learning_rate": 1.95e-05,
+      "loss": 0.7469,
+      "step": 26
+    },
+    {
+      "epoch": 0.014210526315789474,
+      "grad_norm": 1.5182483196258545,
+      "learning_rate": 2.025e-05,
+      "loss": 0.7235,
+      "step": 27
+    },
+    {
+      "epoch": 0.014736842105263158,
+      "grad_norm": 4.085071563720703,
+      "learning_rate": 2.1e-05,
+      "loss": 2.2212,
+      "step": 28
+    },
+    {
+      "epoch": 0.015263157894736841,
+      "grad_norm": 0.7777957916259766,
+      "learning_rate": 2.175e-05,
+      "loss": 1.114,
+      "step": 29
+    },
+    {
+      "epoch": 0.015789473684210527,
+      "grad_norm": 2.115023136138916,
+      "learning_rate": 2.25e-05,
+      "loss": 0.3802,
+      "step": 30
+    },
+    {
+      "epoch": 0.01631578947368421,
+      "grad_norm": 6.118307113647461,
+      "learning_rate": 2.3250000000000003e-05,
+      "loss": 1.2828,
+      "step": 31
+    },
+    {
+      "epoch": 0.016842105263157894,
+      "grad_norm": 1.7932595014572144,
+      "learning_rate": 2.4e-05,
+      "loss": 1.1272,
+      "step": 32
+    },
+    {
+      "epoch": 0.017368421052631578,
+      "grad_norm": 1.0127062797546387,
+      "learning_rate": 2.475e-05,
+      "loss": 1.3309,
+      "step": 33
+    },
+    {
+      "epoch": 0.017894736842105262,
+      "grad_norm": 0.7217763066291809,
+      "learning_rate": 2.55e-05,
+      "loss": 1.2299,
+      "step": 34
+    },
+    {
+      "epoch": 0.018421052631578946,
+      "grad_norm": 4.896430492401123,
+      "learning_rate": 2.625e-05,
+      "loss": 0.8715,
+      "step": 35
+    },
+    {
+      "epoch": 0.018947368421052633,
+      "grad_norm": 1.978946328163147,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.078,
+      "step": 36
+    },
+    {
+      "epoch": 0.019473684210526317,
+      "grad_norm": 4.97585916519165,
+      "learning_rate": 2.7750000000000004e-05,
+      "loss": 1.4184,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3499672412872314,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 1.373,
+      "step": 38
+    },
+    {
+      "epoch": 0.020526315789473684,
+      "grad_norm": 3.067195415496826,
+      "learning_rate": 2.925e-05,
+      "loss": 0.4262,
+      "step": 39
+    },
+    {
+      "epoch": 0.021052631578947368,
+      "grad_norm": 11.364575386047363,
+      "learning_rate": 3e-05,
+      "loss": 1.8748,
+      "step": 40
+    },
+    {
+      "epoch": 0.02157894736842105,
+      "grad_norm": 1.4466383457183838,
+      "learning_rate": 2.999999979408673e-05,
+      "loss": 1.0143,
+      "step": 41
+    },
+    {
+      "epoch": 0.022105263157894735,
+      "grad_norm": 1.8457032442092896,
+      "learning_rate": 2.9999999176346915e-05,
+      "loss": 1.176,
+      "step": 42
+    },
+    {
+      "epoch": 0.022631578947368423,
+      "grad_norm": 1.4849117994308472,
+      "learning_rate": 2.9999998146780576e-05,
+      "loss": 1.1838,
+      "step": 43
+    },
+    {
+      "epoch": 0.023157894736842106,
+      "grad_norm": 1.6621780395507812,
+      "learning_rate": 2.9999996705387744e-05,
+      "loss": 1.0695,
+      "step": 44
+    },
+    {
+      "epoch": 0.02368421052631579,
+      "grad_norm": 2.543567180633545,
+      "learning_rate": 2.9999994852168458e-05,
+      "loss": 1.2559,
+      "step": 45
+    },
+    {
+      "epoch": 0.024210526315789474,
+      "grad_norm": 2.46481990814209,
+      "learning_rate": 2.999999258712277e-05,
+      "loss": 1.2556,
+      "step": 46
+    },
+    {
+      "epoch": 0.024736842105263158,
+      "grad_norm": 3.249171733856201,
+      "learning_rate": 2.999998991025073e-05,
+      "loss": 0.1304,
+      "step": 47
+    },
+    {
+      "epoch": 0.02526315789473684,
+      "grad_norm": 8.809581756591797,
+      "learning_rate": 2.9999986821552427e-05,
+      "loss": 3.5048,
+      "step": 48
+    },
+    {
+      "epoch": 0.025789473684210525,
+      "grad_norm": 2.1939406394958496,
+      "learning_rate": 2.999998332102794e-05,
+      "loss": 1.024,
+      "step": 49
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 5.13323450088501,
+      "learning_rate": 2.9999979408677368e-05,
+      "loss": 0.7764,
+      "step": 50
+    },
+    {
+      "epoch": 0.026842105263157896,
+      "grad_norm": 1.4928818941116333,
+      "learning_rate": 2.999997508450081e-05,
+      "loss": 1.3143,
+      "step": 51
+    },
+    {
+      "epoch": 0.02736842105263158,
+      "grad_norm": 5.20947265625,
+      "learning_rate": 2.999997034849839e-05,
+      "loss": 2.3539,
+      "step": 52
+    },
+    {
+      "epoch": 0.027894736842105264,
+      "grad_norm": 1.5518254041671753,
+      "learning_rate": 2.999996520067024e-05,
+      "loss": 1.0206,
+      "step": 53
+    },
+    {
+      "epoch": 0.028421052631578948,
+      "grad_norm": 1.5228031873703003,
+      "learning_rate": 2.9999959641016498e-05,
+      "loss": 1.1746,
+      "step": 54
+    },
+    {
+      "epoch": 0.02894736842105263,
+      "grad_norm": 4.204188823699951,
+      "learning_rate": 2.999995366953732e-05,
+      "loss": 1.6748,
+      "step": 55
+    },
+    {
+      "epoch": 0.029473684210526315,
+      "grad_norm": 12.567036628723145,
+      "learning_rate": 2.999994728623287e-05,
+      "loss": 1.6706,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.302267074584961,
+      "learning_rate": 2.9999940491103316e-05,
+      "loss": 1.2736,
+      "step": 57
+    },
+    {
+      "epoch": 0.030526315789473683,
+      "grad_norm": 2.4936375617980957,
+      "learning_rate": 2.999993328414885e-05,
+      "loss": 1.1823,
+      "step": 58
+    },
+    {
+      "epoch": 0.03105263157894737,
+      "grad_norm": 2.828848361968994,
+      "learning_rate": 2.9999925665369675e-05,
+      "loss": 2.1706,
+      "step": 59
+    },
+    {
+      "epoch": 0.031578947368421054,
+      "grad_norm": 3.864727258682251,
+      "learning_rate": 2.999991763476599e-05,
+      "loss": 1.4049,
+      "step": 60
+    },
+    {
+      "epoch": 0.032105263157894734,
+      "grad_norm": 2.246717691421509,
+      "learning_rate": 2.9999909192338023e-05,
+      "loss": 1.7979,
+      "step": 61
+    },
+    {
+      "epoch": 0.03263157894736842,
+      "grad_norm": 2.686671257019043,
+      "learning_rate": 2.9999900338086e-05,
+      "loss": 1.2399,
+      "step": 62
+    },
+    {
+      "epoch": 0.03315789473684211,
+      "grad_norm": 4.649468898773193,
+      "learning_rate": 2.9999891072010173e-05,
+      "loss": 1.6921,
+      "step": 63
+    },
+    {
+      "epoch": 0.03368421052631579,
+      "grad_norm": 6.839999675750732,
+      "learning_rate": 2.9999881394110785e-05,
+      "loss": 1.2485,
+      "step": 64
+    },
+    {
+      "epoch": 0.034210526315789476,
+      "grad_norm": 2.245439052581787,
+      "learning_rate": 2.9999871304388115e-05,
+      "loss": 1.8246,
+      "step": 65
+    },
+    {
+      "epoch": 0.034736842105263156,
+      "grad_norm": 3.225515127182007,
+      "learning_rate": 2.9999860802842423e-05,
+      "loss": 1.4347,
+      "step": 66
+    },
+    {
+      "epoch": 0.035263157894736843,
+      "grad_norm": 5.565622806549072,
+      "learning_rate": 2.9999849889474012e-05,
+      "loss": 1.2189,
+      "step": 67
+    },
+    {
+      "epoch": 0.035789473684210524,
+      "grad_norm": 10.95042896270752,
+      "learning_rate": 2.9999838564283172e-05,
+      "loss": 0.4937,
+      "step": 68
+    },
+    {
+      "epoch": 0.03631578947368421,
+      "grad_norm": 2.187350034713745,
+      "learning_rate": 2.9999826827270223e-05,
+      "loss": 1.4298,
+      "step": 69
+    },
+    {
+      "epoch": 0.03684210526315789,
+      "grad_norm": 2.14497447013855,
+      "learning_rate": 2.999981467843548e-05,
+      "loss": 1.1944,
+      "step": 70
+    },
+    {
+      "epoch": 0.03736842105263158,
+      "grad_norm": 1.1492475271224976,
+      "learning_rate": 2.9999802117779277e-05,
+      "loss": 0.6928,
+      "step": 71
+    },
+    {
+      "epoch": 0.037894736842105266,
+      "grad_norm": 4.218234062194824,
+      "learning_rate": 2.9999789145301967e-05,
+      "loss": 0.0767,
+      "step": 72
+    },
+    {
+      "epoch": 0.038421052631578946,
+      "grad_norm": 7.308555603027344,
+      "learning_rate": 2.9999775761003895e-05,
+      "loss": 1.2755,
+      "step": 73
+    },
+    {
+      "epoch": 0.03894736842105263,
+      "grad_norm": 2.922872543334961,
+      "learning_rate": 2.9999761964885436e-05,
+      "loss": 1.1238,
+      "step": 74
+    },
+    {
+      "epoch": 0.039473684210526314,
+      "grad_norm": 2.961275339126587,
+      "learning_rate": 2.9999747756946967e-05,
+      "loss": 1.0357,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.174813270568848,
+      "learning_rate": 2.9999733137188872e-05,
+      "loss": 0.7883,
+      "step": 76
+    },
+    {
+      "epoch": 0.04052631578947368,
+      "grad_norm": 2.3447883129119873,
+      "learning_rate": 2.9999718105611564e-05,
+      "loss": 3.1787,
+      "step": 77
+    },
+    {
+      "epoch": 0.04105263157894737,
+      "grad_norm": 1.4516093730926514,
+      "learning_rate": 2.9999702662215446e-05,
+      "loss": 1.1768,
+      "step": 78
+    },
+    {
+      "epoch": 0.041578947368421056,
+      "grad_norm": 2.680931568145752,
+      "learning_rate": 2.9999686807000945e-05,
+      "loss": 1.6191,
+      "step": 79
+    },
+    {
+      "epoch": 0.042105263157894736,
+      "grad_norm": 2.4612650871276855,
+      "learning_rate": 2.99996705399685e-05,
+      "loss": 0.9811,
+      "step": 80
+    },
+    {
+      "epoch": 0.04263157894736842,
+      "grad_norm": 10.88182258605957,
+      "learning_rate": 2.999965386111855e-05,
+      "loss": 1.5144,
+      "step": 81
+    },
+    {
+      "epoch": 0.0431578947368421,
+      "grad_norm": 1.9869924783706665,
+      "learning_rate": 2.9999636770451562e-05,
+      "loss": 0.7623,
+      "step": 82
+    },
+    {
+      "epoch": 0.04368421052631579,
+      "grad_norm": 25.559877395629883,
+      "learning_rate": 2.9999619267968e-05,
+      "loss": 0.4775,
+      "step": 83
+    },
+    {
+      "epoch": 0.04421052631578947,
+      "grad_norm": 2.2357089519500732,
+      "learning_rate": 2.9999601353668344e-05,
+      "loss": 0.9626,
+      "step": 84
+    },
+    {
+      "epoch": 0.04473684210526316,
+      "grad_norm": 2.5922367572784424,
+      "learning_rate": 2.9999583027553084e-05,
+      "loss": 1.161,
+      "step": 85
+    },
+    {
+      "epoch": 0.045263157894736845,
+      "grad_norm": 3.0082902908325195,
+      "learning_rate": 2.999956428962273e-05,
+      "loss": 0.6429,
+      "step": 86
+    },
+    {
+      "epoch": 0.045789473684210526,
+      "grad_norm": 1.8411225080490112,
+      "learning_rate": 2.9999545139877787e-05,
+      "loss": 1.2799,
+      "step": 87
+    },
+    {
+      "epoch": 0.04631578947368421,
+      "grad_norm": 61.66618347167969,
+      "learning_rate": 2.999952557831879e-05,
+      "loss": 3.875,
+      "step": 88
+    },
+    {
+      "epoch": 0.04684210526315789,
+      "grad_norm": 1.83926522731781,
+      "learning_rate": 2.9999505604946272e-05,
+      "loss": 1.1743,
+      "step": 89
+    },
+    {
+      "epoch": 0.04736842105263158,
+      "grad_norm": 3.020488977432251,
+      "learning_rate": 2.9999485219760786e-05,
+      "loss": 2.1196,
+      "step": 90
+    },
+    {
+      "epoch": 0.04789473684210526,
+      "grad_norm": 3.429885149002075,
+      "learning_rate": 2.999946442276288e-05,
+      "loss": 2.4292,
+      "step": 91
+    },
+    {
+      "epoch": 0.04842105263157895,
+      "grad_norm": 2.8225109577178955,
+      "learning_rate": 2.9999443213953137e-05,
+      "loss": 1.2811,
+      "step": 92
+    },
+    {
+      "epoch": 0.04894736842105263,
+      "grad_norm": 3.0931406021118164,
+      "learning_rate": 2.9999421593332133e-05,
+      "loss": 0.8613,
+      "step": 93
+    },
+    {
+      "epoch": 0.049473684210526316,
+      "grad_norm": 1.8076380491256714,
+      "learning_rate": 2.999939956090046e-05,
+      "loss": 0.9606,
+      "step": 94
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.066877365112305,
+      "learning_rate": 2.999937711665873e-05,
+      "loss": 1.0512,
+      "step": 95
+    },
+    {
+      "epoch": 0.05052631578947368,
+      "grad_norm": 1.641595482826233,
+      "learning_rate": 2.9999354260607556e-05,
+      "loss": 0.7993,
+      "step": 96
+    },
+    {
+      "epoch": 0.05105263157894737,
+      "grad_norm": 9.759380340576172,
+      "learning_rate": 2.9999330992747566e-05,
+      "loss": 0.7986,
+      "step": 97
+    },
+    {
+      "epoch": 0.05157894736842105,
+      "grad_norm": 1.395878791809082,
+      "learning_rate": 2.999930731307939e-05,
+      "loss": 1.129,
+      "step": 98
+    },
+    {
+      "epoch": 0.05210526315789474,
+      "grad_norm": 5.108509540557861,
+      "learning_rate": 2.9999283221603697e-05,
+      "loss": 0.0843,
+      "step": 99
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.187808513641357,
+      "learning_rate": 2.999925871832113e-05,
+      "loss": 1.9898,
+      "step": 100
+    },
+    {
+      "epoch": 0.053157894736842105,
+      "grad_norm": 2.7410473823547363,
+      "learning_rate": 2.9999233803232368e-05,
+      "loss": 1.4947,
+      "step": 101
+    },
+    {
+      "epoch": 0.05368421052631579,
+      "grad_norm": 3.0927412509918213,
+      "learning_rate": 2.99992084763381e-05,
+      "loss": 0.5582,
+      "step": 102
+    },
+    {
+      "epoch": 0.05421052631578947,
+      "grad_norm": 28.332054138183594,
+      "learning_rate": 2.9999182737639015e-05,
+      "loss": 2.1906,
+      "step": 103
+    },
+    {
+      "epoch": 0.05473684210526316,
+      "grad_norm": 1.7883048057556152,
+      "learning_rate": 2.9999156587135824e-05,
+      "loss": 1.2287,
+      "step": 104
+    },
+    {
+      "epoch": 0.05526315789473684,
+      "grad_norm": 3.060166597366333,
+      "learning_rate": 2.999913002482924e-05,
+      "loss": 1.476,
+      "step": 105
+    },
+    {
+      "epoch": 0.05578947368421053,
+      "grad_norm": 1.8946037292480469,
+      "learning_rate": 2.9999103050719998e-05,
+      "loss": 0.8644,
+      "step": 106
+    },
+    {
+      "epoch": 0.05631578947368421,
+      "grad_norm": 2.7290570735931396,
+      "learning_rate": 2.9999075664808832e-05,
+      "loss": 0.1352,
+      "step": 107
+    },
+    {
+      "epoch": 0.056842105263157895,
+      "grad_norm": 6.147144794464111,
+      "learning_rate": 2.9999047867096502e-05,
+      "loss": 0.2661,
+      "step": 108
+    },
+    {
+      "epoch": 0.057368421052631575,
+      "grad_norm": 1.951554298400879,
+      "learning_rate": 2.999901965758377e-05,
+      "loss": 1.4944,
+      "step": 109
+    },
+    {
+      "epoch": 0.05789473684210526,
+      "grad_norm": 2.1169261932373047,
+      "learning_rate": 2.99989910362714e-05,
+      "loss": 1.0947,
+      "step": 110
+    },
+    {
+      "epoch": 0.05842105263157895,
+      "grad_norm": 1.7119637727737427,
+      "learning_rate": 2.9998962003160186e-05,
+      "loss": 1.2013,
+      "step": 111
+    },
+    {
+      "epoch": 0.05894736842105263,
+      "grad_norm": 1.4529640674591064,
+      "learning_rate": 2.9998932558250927e-05,
+      "loss": 1.1557,
+      "step": 112
+    },
+    {
+      "epoch": 0.05947368421052632,
+      "grad_norm": 2.9316415786743164,
+      "learning_rate": 2.9998902701544427e-05,
+      "loss": 1.3037,
+      "step": 113
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.239468574523926,
+      "learning_rate": 2.999887243304151e-05,
+      "loss": 1.135,
+      "step": 114
+    },
+    {
+      "epoch": 0.060526315789473685,
+      "grad_norm": 2.1971309185028076,
+      "learning_rate": 2.9998841752743002e-05,
+      "loss": 1.227,
+      "step": 115
+    },
+    {
+      "epoch": 0.061052631578947365,
+      "grad_norm": 4.265197277069092,
+      "learning_rate": 2.999881066064975e-05,
+      "loss": 0.4307,
+      "step": 116
+    },
+    {
+      "epoch": 0.06157894736842105,
+      "grad_norm": 9.106600761413574,
+      "learning_rate": 2.9998779156762604e-05,
+      "loss": 1.7782,
+      "step": 117
+    },
+    {
+      "epoch": 0.06210526315789474,
+      "grad_norm": 5.628964424133301,
+      "learning_rate": 2.9998747241082433e-05,
+      "loss": 1.1438,
+      "step": 118
+    },
+    {
+      "epoch": 0.06263157894736843,
+      "grad_norm": 33.68876647949219,
+      "learning_rate": 2.9998714913610106e-05,
+      "loss": 1.8069,
+      "step": 119
+    },
+    {
+      "epoch": 0.06315789473684211,
+      "grad_norm": 2.011239767074585,
+      "learning_rate": 2.9998682174346518e-05,
+      "loss": 1.168,
+      "step": 120
+    },
+    {
+      "epoch": 0.06368421052631579,
+      "grad_norm": 2.534121036529541,
+      "learning_rate": 2.9998649023292564e-05,
+      "loss": 1.0316,
+      "step": 121
+    },
+    {
+      "epoch": 0.06421052631578947,
+      "grad_norm": 7.767317295074463,
+      "learning_rate": 2.9998615460449155e-05,
+      "loss": 2.0673,
+      "step": 122
+    },
+    {
+      "epoch": 0.06473684210526316,
+      "grad_norm": 2.559743642807007,
+      "learning_rate": 2.9998581485817213e-05,
+      "loss": 1.1438,
+      "step": 123
+    },
+    {
+      "epoch": 0.06526315789473684,
+      "grad_norm": 5.9045186042785645,
+      "learning_rate": 2.9998547099397673e-05,
+      "loss": 0.527,
+      "step": 124
+    },
+    {
+      "epoch": 0.06578947368421052,
+      "grad_norm": 2.093289375305176,
+      "learning_rate": 2.9998512301191472e-05,
+      "loss": 1.4109,
+      "step": 125
+    },
+    {
+      "epoch": 0.06631578947368422,
+      "grad_norm": 3.366138458251953,
+      "learning_rate": 2.9998477091199575e-05,
+      "loss": 1.536,
+      "step": 126
+    },
+    {
+      "epoch": 0.0668421052631579,
+      "grad_norm": 13.098956108093262,
+      "learning_rate": 2.9998441469422938e-05,
+      "loss": 1.4469,
+      "step": 127
+    },
+    {
+      "epoch": 0.06736842105263158,
+      "grad_norm": 1.7754340171813965,
+      "learning_rate": 2.999840543586255e-05,
+      "loss": 1.2916,
+      "step": 128
+    },
+    {
+      "epoch": 0.06789473684210526,
+      "grad_norm": 2.7748563289642334,
+      "learning_rate": 2.9998368990519393e-05,
+      "loss": 1.0124,
+      "step": 129
+    },
+    {
+      "epoch": 0.06842105263157895,
+      "grad_norm": 16.237028121948242,
+      "learning_rate": 2.9998332133394467e-05,
+      "loss": 0.7667,
+      "step": 130
+    },
+    {
+      "epoch": 0.06894736842105263,
+      "grad_norm": 1.7698700428009033,
+      "learning_rate": 2.9998294864488786e-05,
+      "loss": 1.7484,
+      "step": 131
+    },
+    {
+      "epoch": 0.06947368421052631,
+      "grad_norm": 2.1112704277038574,
+      "learning_rate": 2.9998257183803378e-05,
+      "loss": 1.3866,
+      "step": 132
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5450336933135986,
+      "learning_rate": 2.999821909133927e-05,
+      "loss": 1.1476,
+      "step": 133
+    },
+    {
+      "epoch": 0.07052631578947369,
+      "grad_norm": 2.787191390991211,
+      "learning_rate": 2.9998180587097518e-05,
+      "loss": 0.7392,
+      "step": 134
+    },
+    {
+      "epoch": 0.07105263157894737,
+      "grad_norm": 7.418491840362549,
+      "learning_rate": 2.999814167107916e-05,
+      "loss": 1.0287,
+      "step": 135
+    },
+    {
+      "epoch": 0.07157894736842105,
+      "grad_norm": 2.0604071617126465,
+      "learning_rate": 2.9998102343285288e-05,
+      "loss": 0.9736,
+      "step": 136
+    },
+    {
+      "epoch": 0.07210526315789474,
+      "grad_norm": 3.1856720447540283,
+      "learning_rate": 2.9998062603716966e-05,
+      "loss": 1.1029,
+      "step": 137
+    },
+    {
+      "epoch": 0.07263157894736842,
+      "grad_norm": 44.8514289855957,
+      "learning_rate": 2.9998022452375286e-05,
+      "loss": 2.1055,
+      "step": 138
+    },
+    {
+      "epoch": 0.0731578947368421,
+      "grad_norm": 2.2738101482391357,
+      "learning_rate": 2.999798188926136e-05,
+      "loss": 1.135,
+      "step": 139
+    },
+    {
+      "epoch": 0.07368421052631578,
+      "grad_norm": 4.192337512969971,
+      "learning_rate": 2.9997940914376287e-05,
+      "loss": 0.9468,
+      "step": 140
+    },
+    {
+      "epoch": 0.07421052631578948,
+      "grad_norm": 1.970901370048523,
+      "learning_rate": 2.9997899527721208e-05,
+      "loss": 1.0999,
+      "step": 141
+    },
+    {
+      "epoch": 0.07473684210526316,
+      "grad_norm": 7.665932655334473,
+      "learning_rate": 2.999785772929725e-05,
+      "loss": 0.8222,
+      "step": 142
+    },
+    {
+      "epoch": 0.07526315789473684,
+      "grad_norm": 1.8983104228973389,
+      "learning_rate": 2.9997815519105562e-05,
+      "loss": 1.2967,
+      "step": 143
+    },
+    {
+      "epoch": 0.07578947368421053,
+      "grad_norm": 1.8003042936325073,
+      "learning_rate": 2.9997772897147302e-05,
+      "loss": 1.4046,
+      "step": 144
+    },
+    {
+      "epoch": 0.07631578947368421,
+      "grad_norm": 9.573161125183105,
+      "learning_rate": 2.999772986342364e-05,
+      "loss": 0.5591,
+      "step": 145
+    },
+    {
+      "epoch": 0.07684210526315789,
+      "grad_norm": 4.940485954284668,
+      "learning_rate": 2.9997686417935764e-05,
+      "loss": 1.9492,
+      "step": 146
+    },
+    {
+      "epoch": 0.07736842105263157,
+      "grad_norm": 3.1047379970550537,
+      "learning_rate": 2.9997642560684854e-05,
+      "loss": 1.631,
+      "step": 147
+    },
+    {
+      "epoch": 0.07789473684210527,
+      "grad_norm": 2.6074681282043457,
+      "learning_rate": 2.999759829167213e-05,
+      "loss": 1.4144,
+      "step": 148
+    },
+    {
+      "epoch": 0.07842105263157895,
+      "grad_norm": 6.4516448974609375,
+      "learning_rate": 2.9997553610898793e-05,
+      "loss": 0.7626,
+      "step": 149
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 5.060396671295166,
+      "learning_rate": 2.999750851836608e-05,
+      "loss": 1.457,
+      "step": 150
+    },
+    {
+      "epoch": 0.07947368421052632,
+      "grad_norm": 2.47971510887146,
+      "learning_rate": 2.9997463014075222e-05,
+      "loss": 1.3472,
+      "step": 151
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.023242473602295,
+      "learning_rate": 2.999741709802747e-05,
+      "loss": 1.045,
+      "step": 152
+    },
+    {
+      "epoch": 0.08052631578947368,
+      "grad_norm": 2.3610341548919678,
+      "learning_rate": 2.999737077022409e-05,
+      "loss": 1.0414,
+      "step": 153
+    },
+    {
+      "epoch": 0.08105263157894736,
+      "grad_norm": 6.400054931640625,
+      "learning_rate": 2.9997324030666347e-05,
+      "loss": 2.2875,
+      "step": 154
+    },
+    {
+      "epoch": 0.08157894736842106,
+      "grad_norm": 3.4902946949005127,
+      "learning_rate": 2.999727687935553e-05,
+      "loss": 0.851,
+      "step": 155
+    },
+    {
+      "epoch": 0.08210526315789474,
+      "grad_norm": 17.07858657836914,
+      "learning_rate": 2.9997229316292928e-05,
+      "loss": 1.6888,
+      "step": 156
+    },
+    {
+      "epoch": 0.08263157894736842,
+      "grad_norm": 0.5993421077728271,
+      "learning_rate": 2.999718134147985e-05,
+      "loss": 0.0141,
+      "step": 157
+    },
+    {
+      "epoch": 0.08315789473684211,
+      "grad_norm": 2.3327324390411377,
+      "learning_rate": 2.9997132954917615e-05,
+      "loss": 1.4322,
+      "step": 158
+    },
+    {
+      "epoch": 0.08368421052631579,
+      "grad_norm": 2.2541940212249756,
+      "learning_rate": 2.9997084156607543e-05,
+      "loss": 1.0867,
+      "step": 159
+    },
+    {
+      "epoch": 0.08421052631578947,
+      "grad_norm": 5.866251468658447,
+      "learning_rate": 2.9997034946550984e-05,
+      "loss": 1.4221,
+      "step": 160
+    },
+    {
+      "epoch": 0.08473684210526315,
+      "grad_norm": 4.001148223876953,
+      "learning_rate": 2.9996985324749288e-05,
+      "loss": 1.6462,
+      "step": 161
+    },
+    {
+      "epoch": 0.08526315789473685,
+      "grad_norm": 10.700677871704102,
+      "learning_rate": 2.9996935291203805e-05,
+      "loss": 1.456,
+      "step": 162
+    },
+    {
+      "epoch": 0.08578947368421053,
+      "grad_norm": 1.5370553731918335,
+      "learning_rate": 2.9996884845915925e-05,
+      "loss": 0.4811,
+      "step": 163
+    },
+    {
+      "epoch": 0.0863157894736842,
+      "grad_norm": 4.019623756408691,
+      "learning_rate": 2.9996833988887026e-05,
+      "loss": 0.0901,
+      "step": 164
+    },
+    {
+      "epoch": 0.0868421052631579,
+      "grad_norm": 1.621778130531311,
+      "learning_rate": 2.9996782720118502e-05,
+      "loss": 1.005,
+      "step": 165
+    },
+    {
+      "epoch": 0.08736842105263158,
+      "grad_norm": 2.1922006607055664,
+      "learning_rate": 2.999673103961176e-05,
+      "loss": 1.1868,
+      "step": 166
+    },
+    {
+      "epoch": 0.08789473684210526,
+      "grad_norm": 2.594514846801758,
+      "learning_rate": 2.999667894736823e-05,
+      "loss": 1.2468,
+      "step": 167
+    },
+    {
+      "epoch": 0.08842105263157894,
+      "grad_norm": 1.5204415321350098,
+      "learning_rate": 2.9996626443389325e-05,
+      "loss": 1.098,
+      "step": 168
+    },
+    {
+      "epoch": 0.08894736842105264,
+      "grad_norm": 14.031874656677246,
+      "learning_rate": 2.9996573527676498e-05,
+      "loss": 1.7613,
+      "step": 169
+    },
+    {
+      "epoch": 0.08947368421052632,
+      "grad_norm": 2.578813076019287,
+      "learning_rate": 2.99965202002312e-05,
+      "loss": 0.8607,
+      "step": 170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.223609447479248,
+      "learning_rate": 2.9996466461054897e-05,
+      "loss": 0.3025,
+      "step": 171
+    },
+    {
+      "epoch": 0.09052631578947369,
+      "grad_norm": 3.067014694213867,
+      "learning_rate": 2.9996412310149058e-05,
+      "loss": 1.1488,
+      "step": 172
+    },
+    {
+      "epoch": 0.09105263157894737,
+      "grad_norm": 1.7577857971191406,
+      "learning_rate": 2.9996357747515174e-05,
+      "loss": 1.3616,
+      "step": 173
+    },
+    {
+      "epoch": 0.09157894736842105,
+      "grad_norm": 4.0033278465271,
+      "learning_rate": 2.9996302773154742e-05,
+      "loss": 0.5106,
+      "step": 174
+    },
+    {
+      "epoch": 0.09210526315789473,
+      "grad_norm": 2.1187245845794678,
+      "learning_rate": 2.999624738706927e-05,
+      "loss": 1.1571,
+      "step": 175
+    },
+    {
+      "epoch": 0.09263157894736843,
+      "grad_norm": 10.947097778320312,
+      "learning_rate": 2.9996191589260284e-05,
+      "loss": 1.2541,
+      "step": 176
+    },
+    {
+      "epoch": 0.0931578947368421,
+      "grad_norm": 1.7811756134033203,
+      "learning_rate": 2.9996135379729314e-05,
+      "loss": 1.5256,
+      "step": 177
+    },
+    {
+      "epoch": 0.09368421052631579,
+      "grad_norm": 2.26837420463562,
+      "learning_rate": 2.9996078758477898e-05,
+      "loss": 0.9947,
+      "step": 178
+    },
+    {
+      "epoch": 0.09421052631578947,
+      "grad_norm": 4.844338417053223,
+      "learning_rate": 2.9996021725507595e-05,
+      "loss": 1.7111,
+      "step": 179
+    },
+    {
+      "epoch": 0.09473684210526316,
+      "grad_norm": 1.7822426557540894,
+      "learning_rate": 2.9995964280819967e-05,
+      "loss": 0.824,
+      "step": 180
+    },
+    {
+      "epoch": 0.09526315789473684,
+      "grad_norm": 2.7125401496887207,
+      "learning_rate": 2.99959064244166e-05,
+      "loss": 1.2433,
+      "step": 181
+    },
+    {
+      "epoch": 0.09578947368421052,
+      "grad_norm": 1.5130306482315063,
+      "learning_rate": 2.9995848156299076e-05,
+      "loss": 1.2715,
+      "step": 182
+    },
+    {
+      "epoch": 0.09631578947368422,
+      "grad_norm": 8.074607849121094,
+      "learning_rate": 2.999578947646899e-05,
+      "loss": 1.9851,
+      "step": 183
+    },
+    {
+      "epoch": 0.0968421052631579,
+      "grad_norm": 5.693285942077637,
+      "learning_rate": 2.999573038492796e-05,
+      "loss": 1.5789,
+      "step": 184
+    },
+    {
+      "epoch": 0.09736842105263158,
+      "grad_norm": 1.201459288597107,
+      "learning_rate": 2.9995670881677607e-05,
+      "loss": 1.1029,
+      "step": 185
+    },
+    {
+      "epoch": 0.09789473684210526,
+      "grad_norm": 2.4890646934509277,
+      "learning_rate": 2.9995610966719565e-05,
+      "loss": 1.1719,
+      "step": 186
+    },
+    {
+      "epoch": 0.09842105263157895,
+      "grad_norm": 6.686047077178955,
+      "learning_rate": 2.999555064005548e-05,
+      "loss": 1.8928,
+      "step": 187
+    },
+    {
+      "epoch": 0.09894736842105263,
+      "grad_norm": 1.3713222742080688,
+      "learning_rate": 2.9995489901687004e-05,
+      "loss": 0.7986,
+      "step": 188
+    },
+    {
+      "epoch": 0.09947368421052631,
+      "grad_norm": 1.3539314270019531,
+      "learning_rate": 2.999542875161581e-05,
+      "loss": 1.0582,
+      "step": 189
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.031521797180176,
+      "learning_rate": 2.999536718984357e-05,
+      "loss": 1.2458,
+      "step": 190
+    },
+    {
+      "epoch": 0.10052631578947369,
+      "grad_norm": 2.317741870880127,
+      "learning_rate": 2.9995305216371986e-05,
+      "loss": 1.7422,
+      "step": 191
+    },
+    {
+      "epoch": 0.10105263157894737,
+      "grad_norm": 5.038298606872559,
+      "learning_rate": 2.9995242831202744e-05,
+      "loss": 1.2954,
+      "step": 192
+    },
+    {
+      "epoch": 0.10157894736842105,
+      "grad_norm": 16.174734115600586,
+      "learning_rate": 2.9995180034337566e-05,
+      "loss": 1.0241,
+      "step": 193
+    },
+    {
+      "epoch": 0.10210526315789474,
+      "grad_norm": 9.38874626159668,
+      "learning_rate": 2.9995116825778177e-05,
+      "loss": 1.0844,
+      "step": 194
+    },
+    {
+      "epoch": 0.10263157894736842,
+      "grad_norm": 2.1543664932250977,
+      "learning_rate": 2.9995053205526305e-05,
+      "loss": 1.1873,
+      "step": 195
+    },
+    {
+      "epoch": 0.1031578947368421,
+      "grad_norm": 3.171243906021118,
+      "learning_rate": 2.9994989173583712e-05,
+      "loss": 0.5879,
+      "step": 196
+    },
+    {
+      "epoch": 0.1036842105263158,
+      "grad_norm": 5.367387294769287,
+      "learning_rate": 2.9994924729952135e-05,
+      "loss": 1.1287,
+      "step": 197
+    },
+    {
+      "epoch": 0.10421052631578948,
+      "grad_norm": 6.392814636230469,
+      "learning_rate": 2.9994859874633358e-05,
+      "loss": 0.2617,
+      "step": 198
+    },
+    {
+      "epoch": 0.10473684210526316,
+      "grad_norm": 8.996861457824707,
+      "learning_rate": 2.999479460762916e-05,
+      "loss": 0.5161,
+      "step": 199
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 2.239112138748169,
+      "learning_rate": 2.9994728928941327e-05,
+      "loss": 1.4558,
+      "step": 200
+    },
+    {
+      "epoch": 0.10578947368421053,
+      "grad_norm": 1.4390355348587036,
+      "learning_rate": 2.9994662838571673e-05,
+      "loss": 1.2274,
+      "step": 201
+    },
+    {
+      "epoch": 0.10631578947368421,
+      "grad_norm": 20.712223052978516,
+      "learning_rate": 2.9994596336521997e-05,
+      "loss": 1.6812,
+      "step": 202
+    },
+    {
+      "epoch": 0.10684210526315789,
+      "grad_norm": 3.6506576538085938,
+      "learning_rate": 2.9994529422794136e-05,
+      "loss": 1.7615,
+      "step": 203
+    },
+    {
+      "epoch": 0.10736842105263159,
+      "grad_norm": 4.244356155395508,
+      "learning_rate": 2.9994462097389922e-05,
+      "loss": 2.1721,
+      "step": 204
+    },
+    {
+      "epoch": 0.10789473684210527,
+      "grad_norm": 5.399370193481445,
+      "learning_rate": 2.999439436031121e-05,
+      "loss": 1.2347,
+      "step": 205
+    },
+    {
+      "epoch": 0.10842105263157895,
+      "grad_norm": 13.014787673950195,
+      "learning_rate": 2.999432621155985e-05,
+      "loss": 1.0546,
+      "step": 206
+    },
+    {
+      "epoch": 0.10894736842105263,
+      "grad_norm": 6.201412200927734,
+      "learning_rate": 2.9994257651137727e-05,
+      "loss": 1.6362,
+      "step": 207
+    },
+    {
+      "epoch": 0.10947368421052632,
+      "grad_norm": 8.548736572265625,
+      "learning_rate": 2.999418867904671e-05,
+      "loss": 0.4087,
+      "step": 208
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0881563425064087,
+      "learning_rate": 2.9994119295288696e-05,
+      "loss": 0.0171,
+      "step": 209
+    },
+    {
+      "epoch": 0.11052631578947368,
+      "grad_norm": 5.160693168640137,
+      "learning_rate": 2.999404949986559e-05,
+      "loss": 0.7526,
+      "step": 210
+    },
+    {
+      "epoch": 0.11105263157894738,
+      "grad_norm": 3.1135430335998535,
+      "learning_rate": 2.999397929277931e-05,
+      "loss": 1.3041,
+      "step": 211
+    },
+    {
+      "epoch": 0.11157894736842106,
+      "grad_norm": 3.0570948123931885,
+      "learning_rate": 2.9993908674031787e-05,
+      "loss": 1.0342,
+      "step": 212
+    },
+    {
+      "epoch": 0.11210526315789474,
+      "grad_norm": 2.296029806137085,
+      "learning_rate": 2.9993837643624953e-05,
+      "loss": 1.2628,
+      "step": 213
+    },
+    {
+      "epoch": 0.11263157894736842,
+      "grad_norm": 34.04128646850586,
+      "learning_rate": 2.9993766201560764e-05,
+      "loss": 2.0594,
+      "step": 214
+    },
+    {
+      "epoch": 0.11315789473684211,
+      "grad_norm": 4.997979164123535,
+      "learning_rate": 2.999369434784118e-05,
+      "loss": 0.1955,
+      "step": 215
+    },
+    {
+      "epoch": 0.11368421052631579,
+      "grad_norm": 1.8177436590194702,
+      "learning_rate": 2.9993622082468165e-05,
+      "loss": 1.3383,
+      "step": 216
+    },
+    {
+      "epoch": 0.11421052631578947,
+      "grad_norm": 4.622316360473633,
+      "learning_rate": 2.9993549405443715e-05,
+      "loss": 0.4904,
+      "step": 217
+    },
+    {
+      "epoch": 0.11473684210526315,
+      "grad_norm": 1.736038327217102,
+      "learning_rate": 2.9993476316769822e-05,
+      "loss": 1.1311,
+      "step": 218
+    },
+    {
+      "epoch": 0.11526315789473685,
+      "grad_norm": 1.4792280197143555,
+      "learning_rate": 2.9993402816448487e-05,
+      "loss": 1.6061,
+      "step": 219
+    },
+    {
+      "epoch": 0.11578947368421053,
+      "grad_norm": 20.820289611816406,
+      "learning_rate": 2.9993328904481737e-05,
+      "loss": 3.1151,
+      "step": 220
+    },
+    {
+      "epoch": 0.1163157894736842,
+      "grad_norm": 3.1863715648651123,
+      "learning_rate": 2.999325458087159e-05,
+      "loss": 0.2512,
+      "step": 221
+    },
+    {
+      "epoch": 0.1168421052631579,
+      "grad_norm": 7.053295135498047,
+      "learning_rate": 2.99931798456201e-05,
+      "loss": 2.0884,
+      "step": 222
+    },
+    {
+      "epoch": 0.11736842105263158,
+      "grad_norm": 2.6936466693878174,
+      "learning_rate": 2.999310469872931e-05,
+      "loss": 1.0639,
+      "step": 223
+    },
+    {
+      "epoch": 0.11789473684210526,
+      "grad_norm": 6.108619213104248,
+      "learning_rate": 2.9993029140201288e-05,
+      "loss": 1.5187,
+      "step": 224
+    },
+    {
+      "epoch": 0.11842105263157894,
+      "grad_norm": 1.4771653413772583,
+      "learning_rate": 2.99929531700381e-05,
+      "loss": 0.8693,
+      "step": 225
+    },
+    {
+      "epoch": 0.11894736842105263,
+      "grad_norm": 5.0803632736206055,
+      "learning_rate": 2.999287678824184e-05,
+      "loss": 1.6351,
+      "step": 226
+    },
+    {
+      "epoch": 0.11947368421052632,
+      "grad_norm": 2.738546848297119,
+      "learning_rate": 2.9992799994814602e-05,
+      "loss": 0.076,
+      "step": 227
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8112925291061401,
+      "learning_rate": 2.9992722789758496e-05,
+      "loss": 1.2248,
+      "step": 228
+    },
+    {
+      "epoch": 0.12052631578947369,
+      "grad_norm": 2.918973684310913,
+      "learning_rate": 2.999264517307564e-05,
+      "loss": 1.0747,
+      "step": 229
+    },
+    {
+      "epoch": 0.12105263157894737,
+      "grad_norm": 26.450199127197266,
+      "learning_rate": 2.9992567144768167e-05,
+      "loss": 1.0269,
+      "step": 230
+    },
+    {
+      "epoch": 0.12157894736842105,
+      "grad_norm": 6.718221187591553,
+      "learning_rate": 2.9992488704838215e-05,
+      "loss": 1.3762,
+      "step": 231
+    },
+    {
+      "epoch": 0.12210526315789473,
+      "grad_norm": 12.591816902160645,
+      "learning_rate": 2.9992409853287942e-05,
+      "loss": 2.1439,
+      "step": 232
+    },
+    {
+      "epoch": 0.12263157894736842,
+      "grad_norm": 3.2106196880340576,
+      "learning_rate": 2.9992330590119516e-05,
+      "loss": 1.549,
+      "step": 233
+    },
+    {
+      "epoch": 0.1231578947368421,
+      "grad_norm": 3.1923153400421143,
+      "learning_rate": 2.9992250915335096e-05,
+      "loss": 0.052,
+      "step": 234
+    },
+    {
+      "epoch": 0.12368421052631579,
+      "grad_norm": 2.389961004257202,
+      "learning_rate": 2.999217082893689e-05,
+      "loss": 1.6529,
+      "step": 235
+    },
+    {
+      "epoch": 0.12421052631578948,
+      "grad_norm": 7.870512008666992,
+      "learning_rate": 2.9992090330927092e-05,
+      "loss": 0.9921,
+      "step": 236
+    },
+    {
+      "epoch": 0.12473684210526316,
+      "grad_norm": 7.302090167999268,
+      "learning_rate": 2.9992009421307904e-05,
+      "loss": 2.3439,
+      "step": 237
+    },
+    {
+      "epoch": 0.12526315789473685,
+      "grad_norm": 4.087291240692139,
+      "learning_rate": 2.9991928100081557e-05,
+      "loss": 0.301,
+      "step": 238
+    },
+    {
+      "epoch": 0.12578947368421053,
+      "grad_norm": 1.479772925376892,
+      "learning_rate": 2.9991846367250273e-05,
+      "loss": 1.1892,
+      "step": 239
+    },
+    {
+      "epoch": 0.12631578947368421,
+      "grad_norm": 2.7664432525634766,
+      "learning_rate": 2.9991764222816304e-05,
+      "loss": 0.0283,
+      "step": 240
+    },
+    {
+      "epoch": 0.1268421052631579,
+      "grad_norm": 8.476157188415527,
+      "learning_rate": 2.9991681666781904e-05,
+      "loss": 1.4744,
+      "step": 241
+    },
+    {
+      "epoch": 0.12736842105263158,
+      "grad_norm": 6.929798126220703,
+      "learning_rate": 2.9991598699149337e-05,
+      "loss": 0.122,
+      "step": 242
+    },
+    {
+      "epoch": 0.12789473684210526,
+      "grad_norm": 11.882857322692871,
+      "learning_rate": 2.9991515319920885e-05,
+      "loss": 2.2842,
+      "step": 243
+    },
+    {
+      "epoch": 0.12842105263157894,
+      "grad_norm": 3.38356351852417,
+      "learning_rate": 2.999143152909883e-05,
+      "loss": 0.5131,
+      "step": 244
+    },
+    {
+      "epoch": 0.12894736842105264,
+      "grad_norm": 2.857849359512329,
+      "learning_rate": 2.999134732668548e-05,
+      "loss": 1.4122,
+      "step": 245
+    },
+    {
+      "epoch": 0.12947368421052632,
+      "grad_norm": 5.270646572113037,
+      "learning_rate": 2.9991262712683142e-05,
+      "loss": 0.0852,
+      "step": 246
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.410548210144043,
+      "learning_rate": 2.9991177687094145e-05,
+      "loss": 0.3834,
+      "step": 247
+    },
+    {
+      "epoch": 0.13052631578947368,
+      "grad_norm": 9.727782249450684,
+      "learning_rate": 2.9991092249920818e-05,
+      "loss": 1.3174,
+      "step": 248
+    },
+    {
+      "epoch": 0.13105263157894737,
+      "grad_norm": 2.9647574424743652,
+      "learning_rate": 2.9991006401165505e-05,
+      "loss": 1.0558,
+      "step": 249
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 6.312364101409912,
+      "learning_rate": 2.999092014083057e-05,
+      "loss": 1.5202,
+      "step": 250
+    },
+    {
+      "epoch": 0.13210526315789473,
+      "grad_norm": 6.21821928024292,
+      "learning_rate": 2.9990833468918376e-05,
+      "loss": 1.3892,
+      "step": 251
+    },
+    {
+      "epoch": 0.13263157894736843,
+      "grad_norm": 1.7169901132583618,
+      "learning_rate": 2.999074638543131e-05,
+      "loss": 1.3552,
+      "step": 252
+    },
+    {
+      "epoch": 0.13315789473684211,
+      "grad_norm": 1.8053324222564697,
+      "learning_rate": 2.9990658890371753e-05,
+      "loss": 1.3176,
+      "step": 253
+    },
+    {
+      "epoch": 0.1336842105263158,
+      "grad_norm": 2.0673868656158447,
+      "learning_rate": 2.9990570983742105e-05,
+      "loss": 1.1511,
+      "step": 254
+    },
+    {
+      "epoch": 0.13421052631578947,
+      "grad_norm": 13.948058128356934,
+      "learning_rate": 2.9990482665544792e-05,
+      "loss": 1.8145,
+      "step": 255
+    },
+    {
+      "epoch": 0.13473684210526315,
+      "grad_norm": 14.18269157409668,
+      "learning_rate": 2.999039393578223e-05,
+      "loss": 2.3278,
+      "step": 256
+    },
+    {
+      "epoch": 0.13526315789473684,
+      "grad_norm": 6.532087802886963,
+      "learning_rate": 2.9990304794456857e-05,
+      "loss": 1.6241,
+      "step": 257
+    },
+    {
+      "epoch": 0.13578947368421052,
+      "grad_norm": 5.4354987144470215,
+      "learning_rate": 2.9990215241571124e-05,
+      "loss": 1.094,
+      "step": 258
+    },
+    {
+      "epoch": 0.13631578947368422,
+      "grad_norm": 4.302351474761963,
+      "learning_rate": 2.9990125277127487e-05,
+      "loss": 0.8513,
+      "step": 259
+    },
+    {
+      "epoch": 0.1368421052631579,
+      "grad_norm": 5.297351837158203,
+      "learning_rate": 2.9990034901128414e-05,
+      "loss": 2.1124,
+      "step": 260
+    },
+    {
+      "epoch": 0.13736842105263158,
+      "grad_norm": 1.7843605279922485,
+      "learning_rate": 2.9989944113576387e-05,
+      "loss": 0.9198,
+      "step": 261
+    },
+    {
+      "epoch": 0.13789473684210526,
+      "grad_norm": 3.7895467281341553,
+      "learning_rate": 2.9989852914473898e-05,
+      "loss": 1.4874,
+      "step": 262
+    },
+    {
+      "epoch": 0.13842105263157894,
+      "grad_norm": 12.728391647338867,
+      "learning_rate": 2.9989761303823453e-05,
+      "loss": 1.3333,
+      "step": 263
+    },
+    {
+      "epoch": 0.13894736842105262,
+      "grad_norm": 4.476593017578125,
+      "learning_rate": 2.9989669281627566e-05,
+      "loss": 0.95,
+      "step": 264
+    },
+    {
+      "epoch": 0.1394736842105263,
+      "grad_norm": 2.2026288509368896,
+      "learning_rate": 2.998957684788877e-05,
+      "loss": 0.0549,
+      "step": 265
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.320668935775757,
+      "learning_rate": 2.998948400260959e-05,
+      "loss": 1.6905,
+      "step": 266
+    },
+    {
+      "epoch": 0.1405263157894737,
+      "grad_norm": 7.45945930480957,
+      "learning_rate": 2.9989390745792585e-05,
+      "loss": 0.9003,
+      "step": 267
+    },
+    {
+      "epoch": 0.14105263157894737,
+      "grad_norm": 1.9993150234222412,
+      "learning_rate": 2.998929707744031e-05,
+      "loss": 1.774,
+      "step": 268
+    },
+    {
+      "epoch": 0.14157894736842105,
+      "grad_norm": 2.8484346866607666,
+      "learning_rate": 2.998920299755534e-05,
+      "loss": 0.6502,
+      "step": 269
+    },
+    {
+      "epoch": 0.14210526315789473,
+      "grad_norm": 1.472853183746338,
+      "learning_rate": 2.9989108506140254e-05,
+      "loss": 1.0543,
+      "step": 270
+    },
+    {
+      "epoch": 0.14263157894736841,
+      "grad_norm": 2.5155389308929443,
+      "learning_rate": 2.9989013603197653e-05,
+      "loss": 1.6436,
+      "step": 271
+    },
+    {
+      "epoch": 0.1431578947368421,
+      "grad_norm": 3.893866777420044,
+      "learning_rate": 2.998891828873014e-05,
+      "loss": 0.9275,
+      "step": 272
+    },
+    {
+      "epoch": 0.1436842105263158,
+      "grad_norm": 5.068791389465332,
+      "learning_rate": 2.9988822562740325e-05,
+      "loss": 1.3801,
+      "step": 273
+    },
+    {
+      "epoch": 0.14421052631578948,
+      "grad_norm": 3.420543909072876,
+      "learning_rate": 2.9988726425230842e-05,
+      "loss": 1.4276,
+      "step": 274
+    },
+    {
+      "epoch": 0.14473684210526316,
+      "grad_norm": 3.8543524742126465,
+      "learning_rate": 2.9988629876204337e-05,
+      "loss": 0.1829,
+      "step": 275
+    },
+    {
+      "epoch": 0.14526315789473684,
+      "grad_norm": 2.018007755279541,
+      "learning_rate": 2.9988532915663446e-05,
+      "loss": 0.9781,
+      "step": 276
+    },
+    {
+      "epoch": 0.14578947368421052,
+      "grad_norm": 7.19284200668335,
+      "learning_rate": 2.9988435543610846e-05,
+      "loss": 1.5077,
+      "step": 277
+    },
+    {
+      "epoch": 0.1463157894736842,
+      "grad_norm": 4.468955039978027,
+      "learning_rate": 2.99883377600492e-05,
+      "loss": 1.7061,
+      "step": 278
+    },
+    {
+      "epoch": 0.14684210526315788,
+      "grad_norm": 2.5749592781066895,
+      "learning_rate": 2.9988239564981193e-05,
+      "loss": 1.1039,
+      "step": 279
+    },
+    {
+      "epoch": 0.14736842105263157,
+      "grad_norm": 1.5888959169387817,
+      "learning_rate": 2.9988140958409528e-05,
+      "loss": 1.132,
+      "step": 280
+    },
+    {
+      "epoch": 0.14789473684210527,
+      "grad_norm": 1.6728452444076538,
+      "learning_rate": 2.9988041940336906e-05,
+      "loss": 1.239,
+      "step": 281
+    },
+    {
+      "epoch": 0.14842105263157895,
+      "grad_norm": 1.9604003429412842,
+      "learning_rate": 2.9987942510766047e-05,
+      "loss": 1.2393,
+      "step": 282
+    },
+    {
+      "epoch": 0.14894736842105263,
+      "grad_norm": 12.022218704223633,
+      "learning_rate": 2.9987842669699687e-05,
+      "loss": 0.808,
+      "step": 283
+    },
+    {
+      "epoch": 0.14947368421052631,
+      "grad_norm": 1.27422034740448,
+      "learning_rate": 2.998774241714056e-05,
+      "loss": 1.1194,
+      "step": 284
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4916560649871826,
+      "learning_rate": 2.9987641753091416e-05,
+      "loss": 0.0263,
+      "step": 285
+    },
+    {
+      "epoch": 0.15052631578947367,
+      "grad_norm": 6.565960884094238,
+      "learning_rate": 2.9987540677555027e-05,
+      "loss": 1.2911,
+      "step": 286
+    },
+    {
+      "epoch": 0.15105263157894736,
+      "grad_norm": 1.564158320426941,
+      "learning_rate": 2.9987439190534163e-05,
+      "loss": 0.6841,
+      "step": 287
+    },
+    {
+      "epoch": 0.15157894736842106,
+      "grad_norm": 2.0669963359832764,
+      "learning_rate": 2.998733729203161e-05,
+      "loss": 0.9868,
+      "step": 288
+    },
+    {
+      "epoch": 0.15210526315789474,
+      "grad_norm": 7.10345983505249,
+      "learning_rate": 2.9987234982050168e-05,
+      "loss": 0.6493,
+      "step": 289
+    },
+    {
+      "epoch": 0.15263157894736842,
+      "grad_norm": 2.2110228538513184,
+      "learning_rate": 2.9987132260592645e-05,
+      "loss": 1.225,
+      "step": 290
+    },
+    {
+      "epoch": 0.1531578947368421,
+      "grad_norm": 2.3035542964935303,
+      "learning_rate": 2.9987029127661857e-05,
+      "loss": 1.3038,
+      "step": 291
+    },
+    {
+      "epoch": 0.15368421052631578,
+      "grad_norm": 2.7143070697784424,
+      "learning_rate": 2.9986925583260644e-05,
+      "loss": 1.0185,
+      "step": 292
+    },
+    {
+      "epoch": 0.15421052631578946,
+      "grad_norm": 1.633699655532837,
+      "learning_rate": 2.9986821627391845e-05,
+      "loss": 1.1361,
+      "step": 293
+    },
+    {
+      "epoch": 0.15473684210526314,
+      "grad_norm": 2.975994825363159,
+      "learning_rate": 2.9986717260058314e-05,
+      "loss": 1.8912,
+      "step": 294
+    },
+    {
+      "epoch": 0.15526315789473685,
+      "grad_norm": 2.5356616973876953,
+      "learning_rate": 2.9986612481262916e-05,
+      "loss": 1.2134,
+      "step": 295
+    },
+    {
+      "epoch": 0.15578947368421053,
+      "grad_norm": 1.880893588066101,
+      "learning_rate": 2.9986507291008524e-05,
+      "loss": 0.0698,
+      "step": 296
+    },
+    {
+      "epoch": 0.1563157894736842,
+      "grad_norm": 8.554359436035156,
+      "learning_rate": 2.9986401689298032e-05,
+      "loss": 1.3118,
+      "step": 297
+    },
+    {
+      "epoch": 0.1568421052631579,
+      "grad_norm": 1.8531193733215332,
+      "learning_rate": 2.9986295676134337e-05,
+      "loss": 1.1979,
+      "step": 298
+    },
+    {
+      "epoch": 0.15736842105263157,
+      "grad_norm": 44.23398971557617,
+      "learning_rate": 2.998618925152035e-05,
+      "loss": 2.0334,
+      "step": 299
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 4.069817066192627,
+      "learning_rate": 2.9986082415458993e-05,
+      "loss": 1.2089,
+      "step": 300
+    },
+    {
+      "epoch": 0.15842105263157893,
+      "grad_norm": 1.7818050384521484,
+      "learning_rate": 2.9985975167953198e-05,
+      "loss": 1.0545,
+      "step": 301
+    },
+    {
+      "epoch": 0.15894736842105264,
+      "grad_norm": 2.3045952320098877,
+      "learning_rate": 2.9985867509005906e-05,
+      "loss": 1.1889,
+      "step": 302
+    },
+    {
+      "epoch": 0.15947368421052632,
+      "grad_norm": 4.53651237487793,
+      "learning_rate": 2.9985759438620082e-05,
+      "loss": 0.78,
+      "step": 303
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.4460010528564453,
+      "learning_rate": 2.9985650956798686e-05,
+      "loss": 1.3093,
+      "step": 304
+    },
+    {
+      "epoch": 0.16052631578947368,
+      "grad_norm": 4.067452907562256,
+      "learning_rate": 2.99855420635447e-05,
+      "loss": 0.7694,
+      "step": 305
+    },
+    {
+      "epoch": 0.16105263157894736,
+      "grad_norm": 1.7187111377716064,
+      "learning_rate": 2.9985432758861114e-05,
+      "loss": 1.1114,
+      "step": 306
+    },
+    {
+      "epoch": 0.16157894736842104,
+      "grad_norm": 5.563487529754639,
+      "learning_rate": 2.9985323042750924e-05,
+      "loss": 0.2845,
+      "step": 307
+    },
+    {
+      "epoch": 0.16210526315789472,
+      "grad_norm": 13.657363891601562,
+      "learning_rate": 2.9985212915217146e-05,
+      "loss": 0.7273,
+      "step": 308
+    },
+    {
+      "epoch": 0.16263157894736843,
+      "grad_norm": 44.487571716308594,
+      "learning_rate": 2.9985102376262803e-05,
+      "loss": 1.0723,
+      "step": 309
+    },
+    {
+      "epoch": 0.1631578947368421,
+      "grad_norm": 3.729584217071533,
+      "learning_rate": 2.998499142589093e-05,
+      "loss": 1.0203,
+      "step": 310
+    },
+    {
+      "epoch": 0.1636842105263158,
+      "grad_norm": 3.7574782371520996,
+      "learning_rate": 2.9984880064104575e-05,
+      "loss": 1.1488,
+      "step": 311
+    },
+    {
+      "epoch": 0.16421052631578947,
+      "grad_norm": 10.508902549743652,
+      "learning_rate": 2.998476829090679e-05,
+      "loss": 0.3696,
+      "step": 312
+    },
+    {
+      "epoch": 0.16473684210526315,
+      "grad_norm": 7.688244342803955,
+      "learning_rate": 2.9984656106300647e-05,
+      "loss": 1.8374,
+      "step": 313
+    },
+    {
+      "epoch": 0.16526315789473683,
+      "grad_norm": 2.0016376972198486,
+      "learning_rate": 2.9984543510289227e-05,
+      "loss": 0.8432,
+      "step": 314
+    },
+    {
+      "epoch": 0.16578947368421051,
+      "grad_norm": 5.203074932098389,
+      "learning_rate": 2.998443050287562e-05,
+      "loss": 1.2669,
+      "step": 315
+    },
+    {
+      "epoch": 0.16631578947368422,
+      "grad_norm": 1.451202630996704,
+      "learning_rate": 2.9984317084062928e-05,
+      "loss": 1.0893,
+      "step": 316
+    },
+    {
+      "epoch": 0.1668421052631579,
+      "grad_norm": 1.7263648509979248,
+      "learning_rate": 2.998420325385427e-05,
+      "loss": 1.219,
+      "step": 317
+    },
+    {
+      "epoch": 0.16736842105263158,
+      "grad_norm": 8.742837905883789,
+      "learning_rate": 2.9984089012252765e-05,
+      "loss": 0.3662,
+      "step": 318
+    },
+    {
+      "epoch": 0.16789473684210526,
+      "grad_norm": 4.271378040313721,
+      "learning_rate": 2.9983974359261556e-05,
+      "loss": 0.9902,
+      "step": 319
+    },
+    {
+      "epoch": 0.16842105263157894,
+      "grad_norm": 1.9590401649475098,
+      "learning_rate": 2.998385929488378e-05,
+      "loss": 1.0124,
+      "step": 320
+    },
+    {
+      "epoch": 0.16894736842105262,
+      "grad_norm": 1.6438995599746704,
+      "learning_rate": 2.998374381912261e-05,
+      "loss": 1.0227,
+      "step": 321
+    },
+    {
+      "epoch": 0.1694736842105263,
+      "grad_norm": 8.173205375671387,
+      "learning_rate": 2.99836279319812e-05,
+      "loss": 0.3485,
+      "step": 322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.0618059635162354,
+      "learning_rate": 2.9983511633462746e-05,
+      "loss": 1.213,
+      "step": 323
+    },
+    {
+      "epoch": 0.1705263157894737,
+      "grad_norm": 2.2398312091827393,
+      "learning_rate": 2.9983394923570436e-05,
+      "loss": 0.7417,
+      "step": 324
+    },
+    {
+      "epoch": 0.17105263157894737,
+      "grad_norm": 2.8603358268737793,
+      "learning_rate": 2.9983277802307475e-05,
+      "loss": 0.7479,
+      "step": 325
+    },
+    {
+      "epoch": 0.17157894736842105,
+      "grad_norm": 6.809089660644531,
+      "learning_rate": 2.9983160269677074e-05,
+      "loss": 0.1203,
+      "step": 326
+    },
+    {
+      "epoch": 0.17210526315789473,
+      "grad_norm": 1.7018848657608032,
+      "learning_rate": 2.998304232568247e-05,
+      "loss": 0.7828,
+      "step": 327
+    },
+    {
+      "epoch": 0.1726315789473684,
+      "grad_norm": 1.4800243377685547,
+      "learning_rate": 2.9982923970326887e-05,
+      "loss": 0.0606,
+      "step": 328
+    },
+    {
+      "epoch": 0.1731578947368421,
+      "grad_norm": 3.1752805709838867,
+      "learning_rate": 2.998280520361359e-05,
+      "loss": 0.5135,
+      "step": 329
+    },
+    {
+      "epoch": 0.1736842105263158,
+      "grad_norm": 1.6199371814727783,
+      "learning_rate": 2.9982686025545824e-05,
+      "loss": 1.2659,
+      "step": 330
+    },
+    {
+      "epoch": 0.17421052631578948,
+      "grad_norm": 8.967512130737305,
+      "learning_rate": 2.9982566436126873e-05,
+      "loss": 1.8805,
+      "step": 331
+    },
+    {
+      "epoch": 0.17473684210526316,
+      "grad_norm": 20.32599449157715,
+      "learning_rate": 2.9982446435360016e-05,
+      "loss": 0.6175,
+      "step": 332
+    },
+    {
+      "epoch": 0.17526315789473684,
+      "grad_norm": 12.868733406066895,
+      "learning_rate": 2.9982326023248548e-05,
+      "loss": 1.1088,
+      "step": 333
+    },
+    {
+      "epoch": 0.17578947368421052,
+      "grad_norm": 9.508431434631348,
+      "learning_rate": 2.9982205199795773e-05,
+      "loss": 1.2409,
+      "step": 334
+    },
+    {
+      "epoch": 0.1763157894736842,
+      "grad_norm": 1.3848768472671509,
+      "learning_rate": 2.998208396500501e-05,
+      "loss": 1.1361,
+      "step": 335
+    },
+    {
+      "epoch": 0.17684210526315788,
+      "grad_norm": 1.4120675325393677,
+      "learning_rate": 2.998196231887959e-05,
+      "loss": 0.8192,
+      "step": 336
+    },
+    {
+      "epoch": 0.1773684210526316,
+      "grad_norm": 1.9072043895721436,
+      "learning_rate": 2.998184026142285e-05,
+      "loss": 1.1288,
+      "step": 337
+    },
+    {
+      "epoch": 0.17789473684210527,
+      "grad_norm": 17.120038986206055,
+      "learning_rate": 2.9981717792638143e-05,
+      "loss": 0.6453,
+      "step": 338
+    },
+    {
+      "epoch": 0.17842105263157895,
+      "grad_norm": 1.8352843523025513,
+      "learning_rate": 2.998159491252883e-05,
+      "loss": 1.7777,
+      "step": 339
+    },
+    {
+      "epoch": 0.17894736842105263,
+      "grad_norm": 1.8334228992462158,
+      "learning_rate": 2.9981471621098277e-05,
+      "loss": 1.1382,
+      "step": 340
+    },
+    {
+      "epoch": 0.1794736842105263,
+      "grad_norm": 2.4187421798706055,
+      "learning_rate": 2.9981347918349885e-05,
+      "loss": 1.1575,
+      "step": 341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.076684951782227,
+      "learning_rate": 2.998122380428704e-05,
+      "loss": 0.7906,
+      "step": 342
+    },
+    {
+      "epoch": 0.18052631578947367,
+      "grad_norm": 3.0878186225891113,
+      "learning_rate": 2.9981099278913147e-05,
+      "loss": 1.2896,
+      "step": 343
+    },
+    {
+      "epoch": 0.18105263157894738,
+      "grad_norm": 3.1805942058563232,
+      "learning_rate": 2.9980974342231633e-05,
+      "loss": 0.4216,
+      "step": 344
+    },
+    {
+      "epoch": 0.18157894736842106,
+      "grad_norm": 1.9451438188552856,
+      "learning_rate": 2.9980848994245926e-05,
+      "loss": 1.4878,
+      "step": 345
+    },
+    {
+      "epoch": 0.18210526315789474,
+      "grad_norm": 1.4234017133712769,
+      "learning_rate": 2.9980723234959464e-05,
+      "loss": 0.8428,
+      "step": 346
+    },
+    {
+      "epoch": 0.18263157894736842,
+      "grad_norm": 6.759546756744385,
+      "learning_rate": 2.99805970643757e-05,
+      "loss": 1.7714,
+      "step": 347
+    },
+    {
+      "epoch": 0.1831578947368421,
+      "grad_norm": 7.0381550788879395,
+      "learning_rate": 2.9980470482498105e-05,
+      "loss": 0.4636,
+      "step": 348
+    },
+    {
+      "epoch": 0.18368421052631578,
+      "grad_norm": 6.027357578277588,
+      "learning_rate": 2.9980343489330143e-05,
+      "loss": 1.5307,
+      "step": 349
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 4.1133012771606445,
+      "learning_rate": 2.998021608487531e-05,
+      "loss": 1.3673,
+      "step": 350
+    },
+    {
+      "epoch": 0.18473684210526317,
+      "grad_norm": 4.925607204437256,
+      "learning_rate": 2.9980088269137097e-05,
+      "loss": 2.2288,
+      "step": 351
+    },
+    {
+      "epoch": 0.18526315789473685,
+      "grad_norm": 2.568140983581543,
+      "learning_rate": 2.997996004211902e-05,
+      "loss": 1.3601,
+      "step": 352
+    },
+    {
+      "epoch": 0.18578947368421053,
+      "grad_norm": 2.9798226356506348,
+      "learning_rate": 2.9979831403824593e-05,
+      "loss": 1.1923,
+      "step": 353
+    },
+    {
+      "epoch": 0.1863157894736842,
+      "grad_norm": 1.192610263824463,
+      "learning_rate": 2.997970235425735e-05,
+      "loss": 0.0341,
+      "step": 354
+    },
+    {
+      "epoch": 0.1868421052631579,
+      "grad_norm": 1.2151992321014404,
+      "learning_rate": 2.997957289342084e-05,
+      "loss": 0.9861,
+      "step": 355
+    },
+    {
+      "epoch": 0.18736842105263157,
+      "grad_norm": 1.9701948165893555,
+      "learning_rate": 2.9979443021318607e-05,
+      "loss": 0.0926,
+      "step": 356
+    },
+    {
+      "epoch": 0.18789473684210525,
+      "grad_norm": 1.1996488571166992,
+      "learning_rate": 2.9979312737954225e-05,
+      "loss": 0.8892,
+      "step": 357
+    },
+    {
+      "epoch": 0.18842105263157893,
+      "grad_norm": 3.7606606483459473,
+      "learning_rate": 2.9979182043331268e-05,
+      "loss": 1.3173,
+      "step": 358
+    },
+    {
+      "epoch": 0.18894736842105264,
+      "grad_norm": 4.111024856567383,
+      "learning_rate": 2.9979050937453324e-05,
+      "loss": 1.2348,
+      "step": 359
+    },
+    {
+      "epoch": 0.18947368421052632,
+      "grad_norm": 16.119956970214844,
+      "learning_rate": 2.997891942032399e-05,
+      "loss": 1.7936,
+      "step": 360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.241058349609375,
+      "learning_rate": 2.9978787491946886e-05,
+      "loss": 0.2741,
+      "step": 361
+    },
+    {
+      "epoch": 0.19052631578947368,
+      "grad_norm": 1.398040533065796,
+      "learning_rate": 2.9978655152325623e-05,
+      "loss": 1.1073,
+      "step": 362
+    },
+    {
+      "epoch": 0.19105263157894736,
+      "grad_norm": 6.744707107543945,
+      "learning_rate": 2.997852240146384e-05,
+      "loss": 0.9413,
+      "step": 363
+    },
+    {
+      "epoch": 0.19157894736842104,
+      "grad_norm": 1.4302072525024414,
+      "learning_rate": 2.997838923936518e-05,
+      "loss": 1.1093,
+      "step": 364
+    },
+    {
+      "epoch": 0.19210526315789472,
+      "grad_norm": 1.3738486766815186,
+      "learning_rate": 2.99782556660333e-05,
+      "loss": 1.3803,
+      "step": 365
+    },
+    {
+      "epoch": 0.19263157894736843,
+      "grad_norm": 1.2413164377212524,
+      "learning_rate": 2.9978121681471868e-05,
+      "loss": 0.9807,
+      "step": 366
+    },
+    {
+      "epoch": 0.1931578947368421,
+      "grad_norm": 1.4329386949539185,
+      "learning_rate": 2.997798728568456e-05,
+      "loss": 1.3757,
+      "step": 367
+    },
+    {
+      "epoch": 0.1936842105263158,
+      "grad_norm": 1.2162076234817505,
+      "learning_rate": 2.9977852478675068e-05,
+      "loss": 0.7532,
+      "step": 368
+    },
+    {
+      "epoch": 0.19421052631578947,
+      "grad_norm": 2.649707078933716,
+      "learning_rate": 2.9977717260447092e-05,
+      "loss": 1.0267,
+      "step": 369
+    },
+    {
+      "epoch": 0.19473684210526315,
+      "grad_norm": 6.033254146575928,
+      "learning_rate": 2.9977581631004343e-05,
+      "loss": 2.927,
+      "step": 370
+    },
+    {
+      "epoch": 0.19526315789473683,
+      "grad_norm": 1.8854405879974365,
+      "learning_rate": 2.9977445590350548e-05,
+      "loss": 1.3284,
+      "step": 371
+    },
+    {
+      "epoch": 0.1957894736842105,
+      "grad_norm": 1.3157703876495361,
+      "learning_rate": 2.9977309138489443e-05,
+      "loss": 1.6263,
+      "step": 372
+    },
+    {
+      "epoch": 0.19631578947368422,
+      "grad_norm": 6.529455184936523,
+      "learning_rate": 2.997717227542477e-05,
+      "loss": 0.5972,
+      "step": 373
+    },
+    {
+      "epoch": 0.1968421052631579,
+      "grad_norm": 5.371290683746338,
+      "learning_rate": 2.997703500116029e-05,
+      "loss": 0.2108,
+      "step": 374
+    },
+    {
+      "epoch": 0.19736842105263158,
+      "grad_norm": 11.562923431396484,
+      "learning_rate": 2.9976897315699767e-05,
+      "loss": 0.5319,
+      "step": 375
+    },
+    {
+      "epoch": 0.19789473684210526,
+      "grad_norm": 1.7732006311416626,
+      "learning_rate": 2.9976759219046988e-05,
+      "loss": 1.3096,
+      "step": 376
+    },
+    {
+      "epoch": 0.19842105263157894,
+      "grad_norm": 2.1219842433929443,
+      "learning_rate": 2.997662071120574e-05,
+      "loss": 1.5544,
+      "step": 377
+    },
+    {
+      "epoch": 0.19894736842105262,
+      "grad_norm": 2.662497043609619,
+      "learning_rate": 2.9976481792179827e-05,
+      "loss": 1.3716,
+      "step": 378
+    },
+    {
+      "epoch": 0.1994736842105263,
+      "grad_norm": 1.3942060470581055,
+      "learning_rate": 2.997634246197306e-05,
+      "loss": 1.8652,
+      "step": 379
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5858086347579956,
+      "learning_rate": 2.9976202720589273e-05,
+      "loss": 0.8704,
+      "step": 380
+    },
+    {
+      "epoch": 0.2005263157894737,
+      "grad_norm": 4.291787624359131,
+      "learning_rate": 2.997606256803229e-05,
+      "loss": 0.5378,
+      "step": 381
+    },
+    {
+      "epoch": 0.20105263157894737,
+      "grad_norm": 1.4939614534378052,
+      "learning_rate": 2.9975922004305972e-05,
+      "loss": 1.5748,
+      "step": 382
+    },
+    {
+      "epoch": 0.20157894736842105,
+      "grad_norm": 3.255887269973755,
+      "learning_rate": 2.9975781029414168e-05,
+      "loss": 0.5039,
+      "step": 383
+    },
+    {
+      "epoch": 0.20210526315789473,
+      "grad_norm": 1.5569735765457153,
+      "learning_rate": 2.9975639643360756e-05,
+      "loss": 1.2235,
+      "step": 384
+    },
+    {
+      "epoch": 0.2026315789473684,
+      "grad_norm": 10.412734985351562,
+      "learning_rate": 2.997549784614961e-05,
+      "loss": 1.7202,
+      "step": 385
+    },
+    {
+      "epoch": 0.2031578947368421,
+      "grad_norm": 1.6121785640716553,
+      "learning_rate": 2.997535563778463e-05,
+      "loss": 1.4258,
+      "step": 386
+    },
+    {
+      "epoch": 0.2036842105263158,
+      "grad_norm": 7.74535608291626,
+      "learning_rate": 2.9975213018269714e-05,
+      "loss": 0.3529,
+      "step": 387
+    },
+    {
+      "epoch": 0.20421052631578948,
+      "grad_norm": 1.4278053045272827,
+      "learning_rate": 2.997506998760878e-05,
+      "loss": 1.1017,
+      "step": 388
+    },
+    {
+      "epoch": 0.20473684210526316,
+      "grad_norm": 4.179811477661133,
+      "learning_rate": 2.9974926545805762e-05,
+      "loss": 0.3317,
+      "step": 389
+    },
+    {
+      "epoch": 0.20526315789473684,
+      "grad_norm": 5.233457088470459,
+      "learning_rate": 2.997478269286459e-05,
+      "loss": 2.0176,
+      "step": 390
+    },
+    {
+      "epoch": 0.20578947368421052,
+      "grad_norm": 10.88487720489502,
+      "learning_rate": 2.9974638428789216e-05,
+      "loss": 1.885,
+      "step": 391
+    },
+    {
+      "epoch": 0.2063157894736842,
+      "grad_norm": 1.2455674409866333,
+      "learning_rate": 2.9974493753583597e-05,
+      "loss": 1.3968,
+      "step": 392
+    },
+    {
+      "epoch": 0.20684210526315788,
+      "grad_norm": 1.7053565979003906,
+      "learning_rate": 2.9974348667251715e-05,
+      "loss": 1.0096,
+      "step": 393
+    },
+    {
+      "epoch": 0.2073684210526316,
+      "grad_norm": 4.857005596160889,
+      "learning_rate": 2.9974203169797543e-05,
+      "loss": 0.5129,
+      "step": 394
+    },
+    {
+      "epoch": 0.20789473684210527,
+      "grad_norm": 1.7618751525878906,
+      "learning_rate": 2.997405726122508e-05,
+      "loss": 0.8299,
+      "step": 395
+    },
+    {
+      "epoch": 0.20842105263157895,
+      "grad_norm": 6.38074254989624,
+      "learning_rate": 2.9973910941538333e-05,
+      "loss": 1.711,
+      "step": 396
+    },
+    {
+      "epoch": 0.20894736842105263,
+      "grad_norm": 4.114274024963379,
+      "learning_rate": 2.9973764210741312e-05,
+      "loss": 0.2308,
+      "step": 397
+    },
+    {
+      "epoch": 0.2094736842105263,
+      "grad_norm": 1.7079466581344604,
+      "learning_rate": 2.9973617068838056e-05,
+      "loss": 1.3165,
+      "step": 398
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.909496307373047,
+      "learning_rate": 2.9973469515832604e-05,
+      "loss": 0.4798,
+      "step": 399
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 1.0828051567077637,
+      "learning_rate": 2.9973321551728995e-05,
+      "loss": 0.751,
+      "step": 400
+    },
+    {
+      "epoch": 0.21105263157894738,
+      "grad_norm": 2.6114165782928467,
+      "learning_rate": 2.9973173176531305e-05,
+      "loss": 1.226,
+      "step": 401
+    },
+    {
+      "epoch": 0.21157894736842106,
+      "grad_norm": 2.137087106704712,
+      "learning_rate": 2.99730243902436e-05,
+      "loss": 1.0014,
+      "step": 402
+    },
+    {
+      "epoch": 0.21210526315789474,
+      "grad_norm": 34.939544677734375,
+      "learning_rate": 2.9972875192869965e-05,
+      "loss": 1.3549,
+      "step": 403
+    },
+    {
+      "epoch": 0.21263157894736842,
+      "grad_norm": 9.31963062286377,
+      "learning_rate": 2.9972725584414506e-05,
+      "loss": 1.6069,
+      "step": 404
+    },
+    {
+      "epoch": 0.2131578947368421,
+      "grad_norm": 1.6685644388198853,
+      "learning_rate": 2.9972575564881316e-05,
+      "loss": 0.6431,
+      "step": 405
+    },
+    {
+      "epoch": 0.21368421052631578,
+      "grad_norm": 18.59819984436035,
+      "learning_rate": 2.9972425134274522e-05,
+      "loss": 3.0248,
+      "step": 406
+    },
+    {
+      "epoch": 0.21421052631578946,
+      "grad_norm": 5.35085391998291,
+      "learning_rate": 2.9972274292598255e-05,
+      "loss": 1.6838,
+      "step": 407
+    },
+    {
+      "epoch": 0.21473684210526317,
+      "grad_norm": 2.953678846359253,
+      "learning_rate": 2.9972123039856654e-05,
+      "loss": 0.9957,
+      "step": 408
+    },
+    {
+      "epoch": 0.21526315789473685,
+      "grad_norm": 1.5719188451766968,
+      "learning_rate": 2.997197137605387e-05,
+      "loss": 1.0414,
+      "step": 409
+    },
+    {
+      "epoch": 0.21578947368421053,
+      "grad_norm": 5.360487937927246,
+      "learning_rate": 2.997181930119407e-05,
+      "loss": 1.5796,
+      "step": 410
+    },
+    {
+      "epoch": 0.2163157894736842,
+      "grad_norm": 1.611450433731079,
+      "learning_rate": 2.997166681528143e-05,
+      "loss": 1.025,
+      "step": 411
+    },
+    {
+      "epoch": 0.2168421052631579,
+      "grad_norm": 2.5470244884490967,
+      "learning_rate": 2.997151391832013e-05,
+      "loss": 1.0886,
+      "step": 412
+    },
+    {
+      "epoch": 0.21736842105263157,
+      "grad_norm": 1.2462735176086426,
+      "learning_rate": 2.9971360610314374e-05,
+      "loss": 0.5111,
+      "step": 413
+    },
+    {
+      "epoch": 0.21789473684210525,
+      "grad_norm": 1.5405879020690918,
+      "learning_rate": 2.9971206891268367e-05,
+      "loss": 0.9845,
+      "step": 414
+    },
+    {
+      "epoch": 0.21842105263157896,
+      "grad_norm": 1.282374620437622,
+      "learning_rate": 2.9971052761186338e-05,
+      "loss": 0.4869,
+      "step": 415
+    },
+    {
+      "epoch": 0.21894736842105264,
+      "grad_norm": 1.9443198442459106,
+      "learning_rate": 2.9970898220072505e-05,
+      "loss": 0.0399,
+      "step": 416
+    },
+    {
+      "epoch": 0.21947368421052632,
+      "grad_norm": 1.685855746269226,
+      "learning_rate": 2.9970743267931123e-05,
+      "loss": 1.3679,
+      "step": 417
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6044466495513916,
+      "learning_rate": 2.997058790476644e-05,
+      "loss": 1.4172,
+      "step": 418
+    },
+    {
+      "epoch": 0.22052631578947368,
+      "grad_norm": 3.274233818054199,
+      "learning_rate": 2.9970432130582727e-05,
+      "loss": 1.5899,
+      "step": 419
+    },
+    {
+      "epoch": 0.22105263157894736,
+      "grad_norm": 7.699522972106934,
+      "learning_rate": 2.9970275945384252e-05,
+      "loss": 1.5596,
+      "step": 420
+    },
+    {
+      "epoch": 0.22157894736842104,
+      "grad_norm": 1.1986091136932373,
+      "learning_rate": 2.997011934917531e-05,
+      "loss": 1.1682,
+      "step": 421
+    },
+    {
+      "epoch": 0.22210526315789475,
+      "grad_norm": 56.02631378173828,
+      "learning_rate": 2.9969962341960196e-05,
+      "loss": 1.7715,
+      "step": 422
+    },
+    {
+      "epoch": 0.22263157894736843,
+      "grad_norm": 2.560245990753174,
+      "learning_rate": 2.996980492374323e-05,
+      "loss": 0.9785,
+      "step": 423
+    },
+    {
+      "epoch": 0.2231578947368421,
+      "grad_norm": 11.875818252563477,
+      "learning_rate": 2.9969647094528718e-05,
+      "loss": 1.4328,
+      "step": 424
+    },
+    {
+      "epoch": 0.2236842105263158,
+      "grad_norm": 1.9637598991394043,
+      "learning_rate": 2.9969488854321007e-05,
+      "loss": 2.0752,
+      "step": 425
+    },
+    {
+      "epoch": 0.22421052631578947,
+      "grad_norm": 1.1943355798721313,
+      "learning_rate": 2.9969330203124433e-05,
+      "loss": 0.9533,
+      "step": 426
+    },
+    {
+      "epoch": 0.22473684210526315,
+      "grad_norm": 3.17193341255188,
+      "learning_rate": 2.996917114094336e-05,
+      "loss": 0.3756,
+      "step": 427
+    },
+    {
+      "epoch": 0.22526315789473683,
+      "grad_norm": 6.2271928787231445,
+      "learning_rate": 2.9969011667782152e-05,
+      "loss": 1.5742,
+      "step": 428
+    },
+    {
+      "epoch": 0.22578947368421054,
+      "grad_norm": 3.3042328357696533,
+      "learning_rate": 2.9968851783645182e-05,
+      "loss": 1.2712,
+      "step": 429
+    },
+    {
+      "epoch": 0.22631578947368422,
+      "grad_norm": 1.6236546039581299,
+      "learning_rate": 2.9968691488536842e-05,
+      "loss": 1.3309,
+      "step": 430
+    },
+    {
+      "epoch": 0.2268421052631579,
+      "grad_norm": 3.908189535140991,
+      "learning_rate": 2.9968530782461537e-05,
+      "loss": 2.0394,
+      "step": 431
+    },
+    {
+      "epoch": 0.22736842105263158,
+      "grad_norm": 4.952622413635254,
+      "learning_rate": 2.9968369665423677e-05,
+      "loss": 1.2584,
+      "step": 432
+    },
+    {
+      "epoch": 0.22789473684210526,
+      "grad_norm": 3.6018247604370117,
+      "learning_rate": 2.9968208137427685e-05,
+      "loss": 0.6666,
+      "step": 433
+    },
+    {
+      "epoch": 0.22842105263157894,
+      "grad_norm": 1.319495677947998,
+      "learning_rate": 2.9968046198477994e-05,
+      "loss": 1.0497,
+      "step": 434
+    },
+    {
+      "epoch": 0.22894736842105262,
+      "grad_norm": 9.26498031616211,
+      "learning_rate": 2.9967883848579052e-05,
+      "loss": 1.8237,
+      "step": 435
+    },
+    {
+      "epoch": 0.2294736842105263,
+      "grad_norm": 2.161882162094116,
+      "learning_rate": 2.996772108773532e-05,
+      "loss": 1.5562,
+      "step": 436
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2435736656188965,
+      "learning_rate": 2.9967557915951258e-05,
+      "loss": 1.292,
+      "step": 437
+    },
+    {
+      "epoch": 0.2305263157894737,
+      "grad_norm": 1.310845971107483,
+      "learning_rate": 2.996739433323135e-05,
+      "loss": 0.8349,
+      "step": 438
+    },
+    {
+      "epoch": 0.23105263157894737,
+      "grad_norm": 10.944952011108398,
+      "learning_rate": 2.9967230339580095e-05,
+      "loss": 0.5981,
+      "step": 439
+    },
+    {
+      "epoch": 0.23157894736842105,
+      "grad_norm": 3.8131868839263916,
+      "learning_rate": 2.996706593500198e-05,
+      "loss": 1.3155,
+      "step": 440
+    },
+    {
+      "epoch": 0.23210526315789473,
+      "grad_norm": 3.2014780044555664,
+      "learning_rate": 2.9966901119501535e-05,
+      "loss": 1.2769,
+      "step": 441
+    },
+    {
+      "epoch": 0.2326315789473684,
+      "grad_norm": 0.4154053330421448,
+      "learning_rate": 2.9966735893083274e-05,
+      "loss": 0.0089,
+      "step": 442
+    },
+    {
+      "epoch": 0.2331578947368421,
+      "grad_norm": 1.2587388753890991,
+      "learning_rate": 2.996657025575174e-05,
+      "loss": 1.34,
+      "step": 443
+    },
+    {
+      "epoch": 0.2336842105263158,
+      "grad_norm": 1.6259454488754272,
+      "learning_rate": 2.996640420751147e-05,
+      "loss": 1.2116,
+      "step": 444
+    },
+    {
+      "epoch": 0.23421052631578948,
+      "grad_norm": 24.173168182373047,
+      "learning_rate": 2.9966237748367032e-05,
+      "loss": 1.2547,
+      "step": 445
+    },
+    {
+      "epoch": 0.23473684210526316,
+      "grad_norm": 1.4230668544769287,
+      "learning_rate": 2.9966070878322994e-05,
+      "loss": 0.9427,
+      "step": 446
+    },
+    {
+      "epoch": 0.23526315789473684,
+      "grad_norm": 1.4833992719650269,
+      "learning_rate": 2.996590359738394e-05,
+      "loss": 0.9436,
+      "step": 447
+    },
+    {
+      "epoch": 0.23578947368421052,
+      "grad_norm": 2.0306854248046875,
+      "learning_rate": 2.996573590555446e-05,
+      "loss": 1.333,
+      "step": 448
+    },
+    {
+      "epoch": 0.2363157894736842,
+      "grad_norm": 3.2706172466278076,
+      "learning_rate": 2.996556780283916e-05,
+      "loss": 0.3585,
+      "step": 449
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 2.0781776905059814,
+      "learning_rate": 2.996539928924265e-05,
+      "loss": 1.1149,
+      "step": 450
+    },
+    {
+      "epoch": 0.2373684210526316,
+      "grad_norm": 3.0296876430511475,
+      "learning_rate": 2.996523036476956e-05,
+      "loss": 0.4218,
+      "step": 451
+    },
+    {
+      "epoch": 0.23789473684210527,
+      "grad_norm": 4.4293293952941895,
+      "learning_rate": 2.9965061029424524e-05,
+      "loss": 1.6268,
+      "step": 452
+    },
+    {
+      "epoch": 0.23842105263157895,
+      "grad_norm": 1.4754269123077393,
+      "learning_rate": 2.9964891283212202e-05,
+      "loss": 1.3374,
+      "step": 453
+    },
+    {
+      "epoch": 0.23894736842105263,
+      "grad_norm": 8.703702926635742,
+      "learning_rate": 2.9964721126137245e-05,
+      "loss": 1.5505,
+      "step": 454
+    },
+    {
+      "epoch": 0.2394736842105263,
+      "grad_norm": 1.779236078262329,
+      "learning_rate": 2.9964550558204332e-05,
+      "loss": 1.0386,
+      "step": 455
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.5311484336853027,
+      "learning_rate": 2.9964379579418136e-05,
+      "loss": 1.4634,
+      "step": 456
+    },
+    {
+      "epoch": 0.24052631578947367,
+      "grad_norm": 8.054793357849121,
+      "learning_rate": 2.9964208189783355e-05,
+      "loss": 0.4973,
+      "step": 457
+    },
+    {
+      "epoch": 0.24105263157894738,
+      "grad_norm": 4.909681797027588,
+      "learning_rate": 2.9964036389304696e-05,
+      "loss": 1.1552,
+      "step": 458
+    },
+    {
+      "epoch": 0.24157894736842106,
+      "grad_norm": 6.568388938903809,
+      "learning_rate": 2.996386417798688e-05,
+      "loss": 1.339,
+      "step": 459
+    },
+    {
+      "epoch": 0.24210526315789474,
+      "grad_norm": 1.7436282634735107,
+      "learning_rate": 2.9963691555834626e-05,
+      "loss": 1.6004,
+      "step": 460
+    },
+    {
+      "epoch": 0.24263157894736842,
+      "grad_norm": 2.2429983615875244,
+      "learning_rate": 2.996351852285268e-05,
+      "loss": 1.1068,
+      "step": 461
+    },
+    {
+      "epoch": 0.2431578947368421,
+      "grad_norm": 2.5454885959625244,
+      "learning_rate": 2.996334507904579e-05,
+      "loss": 0.2007,
+      "step": 462
+    },
+    {
+      "epoch": 0.24368421052631578,
+      "grad_norm": 1.5255674123764038,
+      "learning_rate": 2.9963171224418722e-05,
+      "loss": 1.0913,
+      "step": 463
+    },
+    {
+      "epoch": 0.24421052631578946,
+      "grad_norm": 1.427351713180542,
+      "learning_rate": 2.9962996958976242e-05,
+      "loss": 0.7462,
+      "step": 464
+    },
+    {
+      "epoch": 0.24473684210526317,
+      "grad_norm": 1.3530781269073486,
+      "learning_rate": 2.9962822282723136e-05,
+      "loss": 1.3773,
+      "step": 465
+    },
+    {
+      "epoch": 0.24526315789473685,
+      "grad_norm": 1.161293625831604,
+      "learning_rate": 2.996264719566421e-05,
+      "loss": 0.7285,
+      "step": 466
+    },
+    {
+      "epoch": 0.24578947368421053,
+      "grad_norm": 3.624866008758545,
+      "learning_rate": 2.996247169780426e-05,
+      "loss": 0.9183,
+      "step": 467
+    },
+    {
+      "epoch": 0.2463157894736842,
+      "grad_norm": 1.182325005531311,
+      "learning_rate": 2.9962295789148104e-05,
+      "loss": 1.1596,
+      "step": 468
+    },
+    {
+      "epoch": 0.2468421052631579,
+      "grad_norm": 2.112578868865967,
+      "learning_rate": 2.9962119469700577e-05,
+      "loss": 1.302,
+      "step": 469
+    },
+    {
+      "epoch": 0.24736842105263157,
+      "grad_norm": 2.4179584980010986,
+      "learning_rate": 2.9961942739466516e-05,
+      "loss": 1.1984,
+      "step": 470
+    },
+    {
+      "epoch": 0.24789473684210525,
+      "grad_norm": 5.681208610534668,
+      "learning_rate": 2.9961765598450782e-05,
+      "loss": 1.3961,
+      "step": 471
+    },
+    {
+      "epoch": 0.24842105263157896,
+      "grad_norm": 1.9745169878005981,
+      "learning_rate": 2.9961588046658222e-05,
+      "loss": 0.8747,
+      "step": 472
+    },
+    {
+      "epoch": 0.24894736842105264,
+      "grad_norm": 1.0899487733840942,
+      "learning_rate": 2.9961410084093727e-05,
+      "loss": 0.8849,
+      "step": 473
+    },
+    {
+      "epoch": 0.24947368421052632,
+      "grad_norm": 27.703815460205078,
+      "learning_rate": 2.9961231710762173e-05,
+      "loss": 1.362,
+      "step": 474
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 77.52987670898438,
+      "learning_rate": 2.996105292666846e-05,
+      "loss": 1.9388,
+      "step": 475
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.0506709814071655,
+      "eval_runtime": 12.9408,
+      "eval_samples_per_second": 7.727,
+      "eval_steps_per_second": 7.727,
+      "step": 475
+    },
+    {
+      "epoch": 0.2505263157894737,
+      "grad_norm": 1.5593270063400269,
+      "learning_rate": 2.99608737318175e-05,
+      "loss": 1.1288,
+      "step": 476
+    },
+    {
+      "epoch": 0.25105263157894736,
+      "grad_norm": 3.246509552001953,
+      "learning_rate": 2.9960694126214204e-05,
+      "loss": 1.1222,
+      "step": 477
+    },
+    {
+      "epoch": 0.25157894736842107,
+      "grad_norm": 1.3423675298690796,
+      "learning_rate": 2.996051410986351e-05,
+      "loss": 0.8958,
+      "step": 478
+    },
+    {
+      "epoch": 0.2521052631578947,
+      "grad_norm": 2.2557055950164795,
+      "learning_rate": 2.9960333682770367e-05,
+      "loss": 1.0483,
+      "step": 479
+    },
+    {
+      "epoch": 0.25263157894736843,
+      "grad_norm": 1.5628443956375122,
+      "learning_rate": 2.9960152844939716e-05,
+      "loss": 1.1525,
+      "step": 480
+    },
+    {
+      "epoch": 0.2531578947368421,
+      "grad_norm": 4.363856315612793,
+      "learning_rate": 2.9959971596376527e-05,
+      "loss": 2.8042,
+      "step": 481
+    },
+    {
+      "epoch": 0.2536842105263158,
+      "grad_norm": 5.757142543792725,
+      "learning_rate": 2.9959789937085774e-05,
+      "loss": 0.9964,
+      "step": 482
+    },
+    {
+      "epoch": 0.2542105263157895,
+      "grad_norm": 2.2105233669281006,
+      "learning_rate": 2.995960786707245e-05,
+      "loss": 0.844,
+      "step": 483
+    },
+    {
+      "epoch": 0.25473684210526315,
+      "grad_norm": 1.6293426752090454,
+      "learning_rate": 2.9959425386341544e-05,
+      "loss": 1.1284,
+      "step": 484
+    },
+    {
+      "epoch": 0.25526315789473686,
+      "grad_norm": 5.280759334564209,
+      "learning_rate": 2.9959242494898076e-05,
+      "loss": 1.1822,
+      "step": 485
+    },
+    {
+      "epoch": 0.2557894736842105,
+      "grad_norm": 3.082880735397339,
+      "learning_rate": 2.9959059192747064e-05,
+      "loss": 0.4506,
+      "step": 486
+    },
+    {
+      "epoch": 0.2563157894736842,
+      "grad_norm": 1.4250710010528564,
+      "learning_rate": 2.995887547989354e-05,
+      "loss": 0.8048,
+      "step": 487
+    },
+    {
+      "epoch": 0.25684210526315787,
+      "grad_norm": 2.168121814727783,
+      "learning_rate": 2.9958691356342546e-05,
+      "loss": 1.1605,
+      "step": 488
+    },
+    {
+      "epoch": 0.2573684210526316,
+      "grad_norm": 1.3707858324050903,
+      "learning_rate": 2.995850682209914e-05,
+      "loss": 0.9299,
+      "step": 489
+    },
+    {
+      "epoch": 0.2578947368421053,
+      "grad_norm": 4.050701141357422,
+      "learning_rate": 2.9958321877168392e-05,
+      "loss": 0.5374,
+      "step": 490
+    },
+    {
+      "epoch": 0.25842105263157894,
+      "grad_norm": 7.512131214141846,
+      "learning_rate": 2.9958136521555373e-05,
+      "loss": 1.5434,
+      "step": 491
+    },
+    {
+      "epoch": 0.25894736842105265,
+      "grad_norm": 2.668602466583252,
+      "learning_rate": 2.9957950755265167e-05,
+      "loss": 1.0796,
+      "step": 492
+    },
+    {
+      "epoch": 0.2594736842105263,
+      "grad_norm": 1.4869720935821533,
+      "learning_rate": 2.995776457830289e-05,
+      "loss": 1.2192,
+      "step": 493
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3927292823791504,
+      "learning_rate": 2.995757799067364e-05,
+      "loss": 1.1358,
+      "step": 494
+    },
+    {
+      "epoch": 0.26052631578947366,
+      "grad_norm": 1.5988667011260986,
+      "learning_rate": 2.995739099238255e-05,
+      "loss": 0.9472,
+      "step": 495
+    },
+    {
+      "epoch": 0.26105263157894737,
+      "grad_norm": 1.5140737295150757,
+      "learning_rate": 2.9957203583434742e-05,
+      "loss": 0.6382,
+      "step": 496
+    },
+    {
+      "epoch": 0.2615789473684211,
+      "grad_norm": 1.3366851806640625,
+      "learning_rate": 2.9957015763835368e-05,
+      "loss": 0.909,
+      "step": 497
+    },
+    {
+      "epoch": 0.26210526315789473,
+      "grad_norm": 1.2181779146194458,
+      "learning_rate": 2.9956827533589592e-05,
+      "loss": 0.8209,
+      "step": 498
+    },
+    {
+      "epoch": 0.26263157894736844,
+      "grad_norm": 5.791037559509277,
+      "learning_rate": 2.995663889270257e-05,
+      "loss": 0.7449,
+      "step": 499
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 8.648612976074219,
+      "learning_rate": 2.9956449841179486e-05,
+      "loss": 1.3843,
+      "step": 500
+    },
+    {
+      "epoch": 0.2636842105263158,
+      "grad_norm": 4.5848307609558105,
+      "learning_rate": 2.995626037902553e-05,
+      "loss": 0.9603,
+      "step": 501
+    },
+    {
+      "epoch": 0.26421052631578945,
+      "grad_norm": 1.4494788646697998,
+      "learning_rate": 2.99560705062459e-05,
+      "loss": 0.7254,
+      "step": 502
+    },
+    {
+      "epoch": 0.26473684210526316,
+      "grad_norm": 16.663463592529297,
+      "learning_rate": 2.9955880222845818e-05,
+      "loss": 0.3266,
+      "step": 503
+    },
+    {
+      "epoch": 0.26526315789473687,
+      "grad_norm": 1.475623607635498,
+      "learning_rate": 2.99556895288305e-05,
+      "loss": 1.3868,
+      "step": 504
+    },
+    {
+      "epoch": 0.2657894736842105,
+      "grad_norm": 1.3773105144500732,
+      "learning_rate": 2.995549842420519e-05,
+      "loss": 0.1299,
+      "step": 505
+    },
+    {
+      "epoch": 0.26631578947368423,
+      "grad_norm": 1.1917214393615723,
+      "learning_rate": 2.9955306908975126e-05,
+      "loss": 1.0726,
+      "step": 506
+    },
+    {
+      "epoch": 0.2668421052631579,
+      "grad_norm": 5.162087917327881,
+      "learning_rate": 2.9955114983145567e-05,
+      "loss": 1.251,
+      "step": 507
+    },
+    {
+      "epoch": 0.2673684210526316,
+      "grad_norm": 1.6344598531723022,
+      "learning_rate": 2.9954922646721785e-05,
+      "loss": 1.1682,
+      "step": 508
+    },
+    {
+      "epoch": 0.26789473684210524,
+      "grad_norm": 4.443221569061279,
+      "learning_rate": 2.9954729899709065e-05,
+      "loss": 1.285,
+      "step": 509
+    },
+    {
+      "epoch": 0.26842105263157895,
+      "grad_norm": 1.5260694026947021,
+      "learning_rate": 2.9954536742112688e-05,
+      "loss": 1.305,
+      "step": 510
+    },
+    {
+      "epoch": 0.26894736842105266,
+      "grad_norm": 1.5212806463241577,
+      "learning_rate": 2.9954343173937968e-05,
+      "loss": 0.8619,
+      "step": 511
+    },
+    {
+      "epoch": 0.2694736842105263,
+      "grad_norm": 46.06574249267578,
+      "learning_rate": 2.9954149195190214e-05,
+      "loss": 1.2454,
+      "step": 512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1406009197235107,
+      "learning_rate": 2.995395480587475e-05,
+      "loss": 0.8901,
+      "step": 513
+    },
+    {
+      "epoch": 0.27052631578947367,
+      "grad_norm": 1.038522720336914,
+      "learning_rate": 2.995376000599692e-05,
+      "loss": 0.5812,
+      "step": 514
+    },
+    {
+      "epoch": 0.2710526315789474,
+      "grad_norm": 4.814447402954102,
+      "learning_rate": 2.9953564795562064e-05,
+      "loss": 0.5695,
+      "step": 515
+    },
+    {
+      "epoch": 0.27157894736842103,
+      "grad_norm": 3.0107245445251465,
+      "learning_rate": 2.995336917457555e-05,
+      "loss": 1.6017,
+      "step": 516
+    },
+    {
+      "epoch": 0.27210526315789474,
+      "grad_norm": 0.9930879473686218,
+      "learning_rate": 2.995317314304274e-05,
+      "loss": 0.9568,
+      "step": 517
+    },
+    {
+      "epoch": 0.27263157894736845,
+      "grad_norm": 24.54469108581543,
+      "learning_rate": 2.995297670096902e-05,
+      "loss": 0.8168,
+      "step": 518
+    },
+    {
+      "epoch": 0.2731578947368421,
+      "grad_norm": 2.448007106781006,
+      "learning_rate": 2.995277984835979e-05,
+      "loss": 1.5648,
+      "step": 519
+    },
+    {
+      "epoch": 0.2736842105263158,
+      "grad_norm": 3.251126527786255,
+      "learning_rate": 2.995258258522044e-05,
+      "loss": 1.4491,
+      "step": 520
+    },
+    {
+      "epoch": 0.27421052631578946,
+      "grad_norm": 1.0893594026565552,
+      "learning_rate": 2.9952384911556394e-05,
+      "loss": 1.3153,
+      "step": 521
+    },
+    {
+      "epoch": 0.27473684210526317,
+      "grad_norm": 4.283177375793457,
+      "learning_rate": 2.9952186827373085e-05,
+      "loss": 0.2124,
+      "step": 522
+    },
+    {
+      "epoch": 0.2752631578947368,
+      "grad_norm": 3.8956921100616455,
+      "learning_rate": 2.995198833267594e-05,
+      "loss": 1.0426,
+      "step": 523
+    },
+    {
+      "epoch": 0.27578947368421053,
+      "grad_norm": 4.345573425292969,
+      "learning_rate": 2.995178942747042e-05,
+      "loss": 0.2201,
+      "step": 524
+    },
+    {
+      "epoch": 0.27631578947368424,
+      "grad_norm": 1.2920668125152588,
+      "learning_rate": 2.9951590111761975e-05,
+      "loss": 1.4519,
+      "step": 525
+    },
+    {
+      "epoch": 0.2768421052631579,
+      "grad_norm": 1.0732390880584717,
+      "learning_rate": 2.9951390385556084e-05,
+      "loss": 0.8767,
+      "step": 526
+    },
+    {
+      "epoch": 0.2773684210526316,
+      "grad_norm": 1.2250876426696777,
+      "learning_rate": 2.995119024885823e-05,
+      "loss": 0.6662,
+      "step": 527
+    },
+    {
+      "epoch": 0.27789473684210525,
+      "grad_norm": 4.883981704711914,
+      "learning_rate": 2.9950989701673906e-05,
+      "loss": 1.5669,
+      "step": 528
+    },
+    {
+      "epoch": 0.27842105263157896,
+      "grad_norm": 1.9038349390029907,
+      "learning_rate": 2.995078874400862e-05,
+      "loss": 1.1141,
+      "step": 529
+    },
+    {
+      "epoch": 0.2789473684210526,
+      "grad_norm": 8.930418968200684,
+      "learning_rate": 2.9950587375867887e-05,
+      "loss": 1.2227,
+      "step": 530
+    },
+    {
+      "epoch": 0.2794736842105263,
+      "grad_norm": 2.2693824768066406,
+      "learning_rate": 2.995038559725724e-05,
+      "loss": 1.0804,
+      "step": 531
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.965463161468506,
+      "learning_rate": 2.995018340818221e-05,
+      "loss": 1.2129,
+      "step": 532
+    },
+    {
+      "epoch": 0.2805263157894737,
+      "grad_norm": 2.060148000717163,
+      "learning_rate": 2.9949980808648357e-05,
+      "loss": 0.6663,
+      "step": 533
+    },
+    {
+      "epoch": 0.2810526315789474,
+      "grad_norm": 1.0076817274093628,
+      "learning_rate": 2.9949777798661236e-05,
+      "loss": 0.7417,
+      "step": 534
+    },
+    {
+      "epoch": 0.28157894736842104,
+      "grad_norm": 2.449284791946411,
+      "learning_rate": 2.9949574378226426e-05,
+      "loss": 1.5616,
+      "step": 535
+    },
+    {
+      "epoch": 0.28210526315789475,
+      "grad_norm": 1.3279973268508911,
+      "learning_rate": 2.994937054734952e-05,
+      "loss": 0.9661,
+      "step": 536
+    },
+    {
+      "epoch": 0.2826315789473684,
+      "grad_norm": 3.356504201889038,
+      "learning_rate": 2.9949166306036095e-05,
+      "loss": 1.584,
+      "step": 537
+    },
+    {
+      "epoch": 0.2831578947368421,
+      "grad_norm": 1.2326231002807617,
+      "learning_rate": 2.994896165429177e-05,
+      "loss": 0.9024,
+      "step": 538
+    },
+    {
+      "epoch": 0.2836842105263158,
+      "grad_norm": 7.02909517288208,
+      "learning_rate": 2.994875659212216e-05,
+      "loss": 1.0278,
+      "step": 539
+    },
+    {
+      "epoch": 0.28421052631578947,
+      "grad_norm": 1.6224465370178223,
+      "learning_rate": 2.9948551119532902e-05,
+      "loss": 1.278,
+      "step": 540
+    },
+    {
+      "epoch": 0.2847368421052632,
+      "grad_norm": 1.3300915956497192,
+      "learning_rate": 2.994834523652963e-05,
+      "loss": 1.1461,
+      "step": 541
+    },
+    {
+      "epoch": 0.28526315789473683,
+      "grad_norm": 1.0725505352020264,
+      "learning_rate": 2.9948138943117996e-05,
+      "loss": 0.9856,
+      "step": 542
+    },
+    {
+      "epoch": 0.28578947368421054,
+      "grad_norm": 1.2995253801345825,
+      "learning_rate": 2.9947932239303673e-05,
+      "loss": 0.922,
+      "step": 543
+    },
+    {
+      "epoch": 0.2863157894736842,
+      "grad_norm": 3.25260066986084,
+      "learning_rate": 2.9947725125092326e-05,
+      "loss": 1.5961,
+      "step": 544
+    },
+    {
+      "epoch": 0.2868421052631579,
+      "grad_norm": 1.5616159439086914,
+      "learning_rate": 2.9947517600489643e-05,
+      "loss": 1.4245,
+      "step": 545
+    },
+    {
+      "epoch": 0.2873684210526316,
+      "grad_norm": 14.093785285949707,
+      "learning_rate": 2.9947309665501327e-05,
+      "loss": 0.9921,
+      "step": 546
+    },
+    {
+      "epoch": 0.28789473684210526,
+      "grad_norm": 1.165595293045044,
+      "learning_rate": 2.9947101320133083e-05,
+      "loss": 1.0,
+      "step": 547
+    },
+    {
+      "epoch": 0.28842105263157897,
+      "grad_norm": 2.5077242851257324,
+      "learning_rate": 2.994689256439063e-05,
+      "loss": 1.5877,
+      "step": 548
+    },
+    {
+      "epoch": 0.2889473684210526,
+      "grad_norm": 1.8317209482192993,
+      "learning_rate": 2.9946683398279706e-05,
+      "loss": 1.0294,
+      "step": 549
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 1.0942083597183228,
+      "learning_rate": 2.9946473821806044e-05,
+      "loss": 0.4813,
+      "step": 550
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6907622814178467,
+      "learning_rate": 2.9946263834975403e-05,
+      "loss": 0.9892,
+      "step": 551
+    },
+    {
+      "epoch": 0.2905263157894737,
+      "grad_norm": 9.270956993103027,
+      "learning_rate": 2.994605343779355e-05,
+      "loss": 0.6519,
+      "step": 552
+    },
+    {
+      "epoch": 0.2910526315789474,
+      "grad_norm": 3.074319362640381,
+      "learning_rate": 2.9945842630266258e-05,
+      "loss": 0.6775,
+      "step": 553
+    },
+    {
+      "epoch": 0.29157894736842105,
+      "grad_norm": 1.4713774919509888,
+      "learning_rate": 2.994563141239932e-05,
+      "loss": 1.1378,
+      "step": 554
+    },
+    {
+      "epoch": 0.29210526315789476,
+      "grad_norm": 3.8262014389038086,
+      "learning_rate": 2.9945419784198524e-05,
+      "loss": 1.2291,
+      "step": 555
+    },
+    {
+      "epoch": 0.2926315789473684,
+      "grad_norm": 3.6956279277801514,
+      "learning_rate": 2.9945207745669695e-05,
+      "loss": 0.2079,
+      "step": 556
+    },
+    {
+      "epoch": 0.2931578947368421,
+      "grad_norm": 1.555780053138733,
+      "learning_rate": 2.9944995296818642e-05,
+      "loss": 0.9855,
+      "step": 557
+    },
+    {
+      "epoch": 0.29368421052631577,
+      "grad_norm": 7.439387321472168,
+      "learning_rate": 2.9944782437651203e-05,
+      "loss": 0.1662,
+      "step": 558
+    },
+    {
+      "epoch": 0.2942105263157895,
+      "grad_norm": 2.846749782562256,
+      "learning_rate": 2.994456916817322e-05,
+      "loss": 0.4499,
+      "step": 559
+    },
+    {
+      "epoch": 0.29473684210526313,
+      "grad_norm": 2.7918450832366943,
+      "learning_rate": 2.9944355488390555e-05,
+      "loss": 0.0985,
+      "step": 560
+    },
+    {
+      "epoch": 0.29526315789473684,
+      "grad_norm": 3.7580766677856445,
+      "learning_rate": 2.9944141398309067e-05,
+      "loss": 0.7624,
+      "step": 561
+    },
+    {
+      "epoch": 0.29578947368421055,
+      "grad_norm": 1.398541808128357,
+      "learning_rate": 2.994392689793464e-05,
+      "loss": 0.7911,
+      "step": 562
+    },
+    {
+      "epoch": 0.2963157894736842,
+      "grad_norm": 7.155591011047363,
+      "learning_rate": 2.9943711987273156e-05,
+      "loss": 0.2768,
+      "step": 563
+    },
+    {
+      "epoch": 0.2968421052631579,
+      "grad_norm": 15.522272109985352,
+      "learning_rate": 2.994349666633052e-05,
+      "loss": 1.0977,
+      "step": 564
+    },
+    {
+      "epoch": 0.29736842105263156,
+      "grad_norm": 2.732008457183838,
+      "learning_rate": 2.9943280935112644e-05,
+      "loss": 0.749,
+      "step": 565
+    },
+    {
+      "epoch": 0.29789473684210527,
+      "grad_norm": 1.2546172142028809,
+      "learning_rate": 2.994306479362545e-05,
+      "loss": 0.823,
+      "step": 566
+    },
+    {
+      "epoch": 0.2984210526315789,
+      "grad_norm": 1.2818124294281006,
+      "learning_rate": 2.994284824187487e-05,
+      "loss": 1.1513,
+      "step": 567
+    },
+    {
+      "epoch": 0.29894736842105263,
+      "grad_norm": 1.5917184352874756,
+      "learning_rate": 2.9942631279866853e-05,
+      "loss": 0.8293,
+      "step": 568
+    },
+    {
+      "epoch": 0.29947368421052634,
+      "grad_norm": 1.2113322019577026,
+      "learning_rate": 2.9942413907607355e-05,
+      "loss": 0.3476,
+      "step": 569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0724539756774902,
+      "learning_rate": 2.994219612510234e-05,
+      "loss": 0.6935,
+      "step": 570
+    },
+    {
+      "epoch": 0.3005263157894737,
+      "grad_norm": 2.444352149963379,
+      "learning_rate": 2.9941977932357794e-05,
+      "loss": 0.1079,
+      "step": 571
+    },
+    {
+      "epoch": 0.30105263157894735,
+      "grad_norm": 4.327816486358643,
+      "learning_rate": 2.99417593293797e-05,
+      "loss": 0.5792,
+      "step": 572
+    },
+    {
+      "epoch": 0.30157894736842106,
+      "grad_norm": 1.2691162824630737,
+      "learning_rate": 2.9941540316174066e-05,
+      "loss": 0.9346,
+      "step": 573
+    },
+    {
+      "epoch": 0.3021052631578947,
+      "grad_norm": 2.0975606441497803,
+      "learning_rate": 2.99413208927469e-05,
+      "loss": 1.1176,
+      "step": 574
+    },
+    {
+      "epoch": 0.3026315789473684,
+      "grad_norm": 1.5340697765350342,
+      "learning_rate": 2.994110105910423e-05,
+      "loss": 0.862,
+      "step": 575
+    },
+    {
+      "epoch": 0.3031578947368421,
+      "grad_norm": 16.70974349975586,
+      "learning_rate": 2.9940880815252097e-05,
+      "loss": 0.6121,
+      "step": 576
+    },
+    {
+      "epoch": 0.3036842105263158,
+      "grad_norm": 3.6483004093170166,
+      "learning_rate": 2.9940660161196536e-05,
+      "loss": 1.0742,
+      "step": 577
+    },
+    {
+      "epoch": 0.3042105263157895,
+      "grad_norm": 1.5293868780136108,
+      "learning_rate": 2.994043909694361e-05,
+      "loss": 0.7131,
+      "step": 578
+    },
+    {
+      "epoch": 0.30473684210526314,
+      "grad_norm": 3.1697440147399902,
+      "learning_rate": 2.994021762249939e-05,
+      "loss": 0.7688,
+      "step": 579
+    },
+    {
+      "epoch": 0.30526315789473685,
+      "grad_norm": 11.020563125610352,
+      "learning_rate": 2.993999573786995e-05,
+      "loss": 2.4059,
+      "step": 580
+    },
+    {
+      "epoch": 0.3057894736842105,
+      "grad_norm": 1.883289098739624,
+      "learning_rate": 2.9939773443061393e-05,
+      "loss": 1.0566,
+      "step": 581
+    },
+    {
+      "epoch": 0.3063157894736842,
+      "grad_norm": 1.52479887008667,
+      "learning_rate": 2.9939550738079814e-05,
+      "loss": 1.0846,
+      "step": 582
+    },
+    {
+      "epoch": 0.3068421052631579,
+      "grad_norm": 4.475525856018066,
+      "learning_rate": 2.9939327622931333e-05,
+      "loss": 0.6157,
+      "step": 583
+    },
+    {
+      "epoch": 0.30736842105263157,
+      "grad_norm": 1.8167206048965454,
+      "learning_rate": 2.993910409762207e-05,
+      "loss": 1.3638,
+      "step": 584
+    },
+    {
+      "epoch": 0.3078947368421053,
+      "grad_norm": 1.2058393955230713,
+      "learning_rate": 2.9938880162158164e-05,
+      "loss": 1.0466,
+      "step": 585
+    },
+    {
+      "epoch": 0.30842105263157893,
+      "grad_norm": 8.277087211608887,
+      "learning_rate": 2.9938655816545764e-05,
+      "loss": 1.4132,
+      "step": 586
+    },
+    {
+      "epoch": 0.30894736842105264,
+      "grad_norm": 2.545767307281494,
+      "learning_rate": 2.993843106079103e-05,
+      "loss": 0.7397,
+      "step": 587
+    },
+    {
+      "epoch": 0.3094736842105263,
+      "grad_norm": 1.162050485610962,
+      "learning_rate": 2.993820589490013e-05,
+      "loss": 1.1346,
+      "step": 588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3114622831344604,
+      "learning_rate": 2.9937980318879247e-05,
+      "loss": 0.9332,
+      "step": 589
+    },
+    {
+      "epoch": 0.3105263157894737,
+      "grad_norm": 2.397101640701294,
+      "learning_rate": 2.9937754332734576e-05,
+      "loss": 0.7573,
+      "step": 590
+    },
+    {
+      "epoch": 0.31105263157894736,
+      "grad_norm": 1.4673283100128174,
+      "learning_rate": 2.993752793647232e-05,
+      "loss": 0.8484,
+      "step": 591
+    },
+    {
+      "epoch": 0.31157894736842107,
+      "grad_norm": 9.11661148071289,
+      "learning_rate": 2.9937301130098696e-05,
+      "loss": 0.4733,
+      "step": 592
+    },
+    {
+      "epoch": 0.3121052631578947,
+      "grad_norm": 1.699437141418457,
+      "learning_rate": 2.9937073913619926e-05,
+      "loss": 0.5332,
+      "step": 593
+    },
+    {
+      "epoch": 0.3126315789473684,
+      "grad_norm": 27.69738006591797,
+      "learning_rate": 2.9936846287042254e-05,
+      "loss": 2.1628,
+      "step": 594
+    },
+    {
+      "epoch": 0.3131578947368421,
+      "grad_norm": 10.93797779083252,
+      "learning_rate": 2.993661825037193e-05,
+      "loss": 1.469,
+      "step": 595
+    },
+    {
+      "epoch": 0.3136842105263158,
+      "grad_norm": 4.551391124725342,
+      "learning_rate": 2.993638980361521e-05,
+      "loss": 0.6163,
+      "step": 596
+    },
+    {
+      "epoch": 0.3142105263157895,
+      "grad_norm": 2.5031051635742188,
+      "learning_rate": 2.9936160946778372e-05,
+      "loss": 1.3955,
+      "step": 597
+    },
+    {
+      "epoch": 0.31473684210526315,
+      "grad_norm": 1.7328745126724243,
+      "learning_rate": 2.993593167986769e-05,
+      "loss": 0.7825,
+      "step": 598
+    },
+    {
+      "epoch": 0.31526315789473686,
+      "grad_norm": 4.045513153076172,
+      "learning_rate": 2.993570200288947e-05,
+      "loss": 0.4318,
+      "step": 599
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 2.73246169090271,
+      "learning_rate": 2.993547191585001e-05,
+      "loss": 1.4336,
+      "step": 600
+    },
+    {
+      "epoch": 0.3163157894736842,
+      "grad_norm": 1.558140754699707,
+      "learning_rate": 2.9935241418755626e-05,
+      "loss": 1.0587,
+      "step": 601
+    },
+    {
+      "epoch": 0.31684210526315787,
+      "grad_norm": 11.362480163574219,
+      "learning_rate": 2.9935010511612655e-05,
+      "loss": 0.4542,
+      "step": 602
+    },
+    {
+      "epoch": 0.3173684210526316,
+      "grad_norm": 1.3051596879959106,
+      "learning_rate": 2.9934779194427427e-05,
+      "loss": 0.7598,
+      "step": 603
+    },
+    {
+      "epoch": 0.3178947368421053,
+      "grad_norm": 3.5925891399383545,
+      "learning_rate": 2.99345474672063e-05,
+      "loss": 1.1406,
+      "step": 604
+    },
+    {
+      "epoch": 0.31842105263157894,
+      "grad_norm": 12.116620063781738,
+      "learning_rate": 2.993431532995563e-05,
+      "loss": 1.519,
+      "step": 605
+    },
+    {
+      "epoch": 0.31894736842105265,
+      "grad_norm": 1.2009161710739136,
+      "learning_rate": 2.9934082782681797e-05,
+      "loss": 0.9572,
+      "step": 606
+    },
+    {
+      "epoch": 0.3194736842105263,
+      "grad_norm": 2.059299945831299,
+      "learning_rate": 2.993384982539118e-05,
+      "loss": 0.6667,
+      "step": 607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.226381778717041,
+      "learning_rate": 2.9933616458090174e-05,
+      "loss": 1.2235,
+      "step": 608
+    },
+    {
+      "epoch": 0.32052631578947366,
+      "grad_norm": 1.5386918783187866,
+      "learning_rate": 2.9933382680785196e-05,
+      "loss": 1.2193,
+      "step": 609
+    },
+    {
+      "epoch": 0.32105263157894737,
+      "grad_norm": 4.095519065856934,
+      "learning_rate": 2.9933148493482653e-05,
+      "loss": 0.2043,
+      "step": 610
+    },
+    {
+      "epoch": 0.3215789473684211,
+      "grad_norm": 16.421064376831055,
+      "learning_rate": 2.9932913896188978e-05,
+      "loss": 1.4538,
+      "step": 611
+    },
+    {
+      "epoch": 0.32210526315789473,
+      "grad_norm": 1.1642028093338013,
+      "learning_rate": 2.9932678888910614e-05,
+      "loss": 0.7069,
+      "step": 612
+    },
+    {
+      "epoch": 0.32263157894736844,
+      "grad_norm": 1.0497969388961792,
+      "learning_rate": 2.9932443471654013e-05,
+      "loss": 1.027,
+      "step": 613
+    },
+    {
+      "epoch": 0.3231578947368421,
+      "grad_norm": 1.70055091381073,
+      "learning_rate": 2.9932207644425635e-05,
+      "loss": 1.0163,
+      "step": 614
+    },
+    {
+      "epoch": 0.3236842105263158,
+      "grad_norm": 0.4913199841976166,
+      "learning_rate": 2.993197140723196e-05,
+      "loss": 0.0092,
+      "step": 615
+    },
+    {
+      "epoch": 0.32421052631578945,
+      "grad_norm": 9.340396881103516,
+      "learning_rate": 2.993173476007947e-05,
+      "loss": 0.7314,
+      "step": 616
+    },
+    {
+      "epoch": 0.32473684210526316,
+      "grad_norm": 3.6009857654571533,
+      "learning_rate": 2.993149770297466e-05,
+      "loss": 0.9824,
+      "step": 617
+    },
+    {
+      "epoch": 0.32526315789473687,
+      "grad_norm": 1.5114206075668335,
+      "learning_rate": 2.9931260235924046e-05,
+      "loss": 0.6525,
+      "step": 618
+    },
+    {
+      "epoch": 0.3257894736842105,
+      "grad_norm": 4.900846004486084,
+      "learning_rate": 2.9931022358934144e-05,
+      "loss": 1.0022,
+      "step": 619
+    },
+    {
+      "epoch": 0.3263157894736842,
+      "grad_norm": 3.2764687538146973,
+      "learning_rate": 2.993078407201148e-05,
+      "loss": 1.571,
+      "step": 620
+    },
+    {
+      "epoch": 0.3268421052631579,
+      "grad_norm": 1.1837842464447021,
+      "learning_rate": 2.9930545375162602e-05,
+      "loss": 1.2526,
+      "step": 621
+    },
+    {
+      "epoch": 0.3273684210526316,
+      "grad_norm": 1.7309764623641968,
+      "learning_rate": 2.993030626839406e-05,
+      "loss": 0.9898,
+      "step": 622
+    },
+    {
+      "epoch": 0.32789473684210524,
+      "grad_norm": 1.766406536102295,
+      "learning_rate": 2.9930066751712427e-05,
+      "loss": 0.6513,
+      "step": 623
+    },
+    {
+      "epoch": 0.32842105263157895,
+      "grad_norm": 1.1759055852890015,
+      "learning_rate": 2.9929826825124268e-05,
+      "loss": 1.0862,
+      "step": 624
+    },
+    {
+      "epoch": 0.32894736842105265,
+      "grad_norm": 46.77162551879883,
+      "learning_rate": 2.9929586488636174e-05,
+      "loss": 3.9448,
+      "step": 625
+    },
+    {
+      "epoch": 0.3294736842105263,
+      "grad_norm": 1.6841140985488892,
+      "learning_rate": 2.992934574225475e-05,
+      "loss": 1.6885,
+      "step": 626
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.449564814567566,
+      "learning_rate": 2.9929104585986594e-05,
+      "loss": 1.354,
+      "step": 627
+    },
+    {
+      "epoch": 0.33052631578947367,
+      "grad_norm": 2.3545868396759033,
+      "learning_rate": 2.992886301983833e-05,
+      "loss": 0.7053,
+      "step": 628
+    },
+    {
+      "epoch": 0.3310526315789474,
+      "grad_norm": 1.053680181503296,
+      "learning_rate": 2.9928621043816602e-05,
+      "loss": 1.3087,
+      "step": 629
+    },
+    {
+      "epoch": 0.33157894736842103,
+      "grad_norm": 7.419710636138916,
+      "learning_rate": 2.9928378657928037e-05,
+      "loss": 0.3903,
+      "step": 630
+    },
+    {
+      "epoch": 0.33210526315789474,
+      "grad_norm": 1.032652497291565,
+      "learning_rate": 2.9928135862179304e-05,
+      "loss": 1.0834,
+      "step": 631
+    },
+    {
+      "epoch": 0.33263157894736844,
+      "grad_norm": 11.945768356323242,
+      "learning_rate": 2.9927892656577057e-05,
+      "loss": 0.5555,
+      "step": 632
+    },
+    {
+      "epoch": 0.3331578947368421,
+      "grad_norm": 3.9728310108184814,
+      "learning_rate": 2.9927649041127978e-05,
+      "loss": 0.2488,
+      "step": 633
+    },
+    {
+      "epoch": 0.3336842105263158,
+      "grad_norm": 1.3373970985412598,
+      "learning_rate": 2.992740501583876e-05,
+      "loss": 1.0203,
+      "step": 634
+    },
+    {
+      "epoch": 0.33421052631578946,
+      "grad_norm": 4.078335762023926,
+      "learning_rate": 2.9927160580716096e-05,
+      "loss": 1.0495,
+      "step": 635
+    },
+    {
+      "epoch": 0.33473684210526317,
+      "grad_norm": 1.9304755926132202,
+      "learning_rate": 2.99269157357667e-05,
+      "loss": 1.4033,
+      "step": 636
+    },
+    {
+      "epoch": 0.3352631578947368,
+      "grad_norm": 2.1133267879486084,
+      "learning_rate": 2.99266704809973e-05,
+      "loss": 1.1063,
+      "step": 637
+    },
+    {
+      "epoch": 0.3357894736842105,
+      "grad_norm": 6.6004180908203125,
+      "learning_rate": 2.9926424816414615e-05,
+      "loss": 1.1123,
+      "step": 638
+    },
+    {
+      "epoch": 0.33631578947368423,
+      "grad_norm": 1.2974992990493774,
+      "learning_rate": 2.9926178742025403e-05,
+      "loss": 2.3209,
+      "step": 639
+    },
+    {
+      "epoch": 0.3368421052631579,
+      "grad_norm": 1.2676901817321777,
+      "learning_rate": 2.992593225783641e-05,
+      "loss": 0.9767,
+      "step": 640
+    },
+    {
+      "epoch": 0.3373684210526316,
+      "grad_norm": 1.001715064048767,
+      "learning_rate": 2.9925685363854413e-05,
+      "loss": 0.8202,
+      "step": 641
+    },
+    {
+      "epoch": 0.33789473684210525,
+      "grad_norm": 3.934710741043091,
+      "learning_rate": 2.9925438060086187e-05,
+      "loss": 1.4305,
+      "step": 642
+    },
+    {
+      "epoch": 0.33842105263157896,
+      "grad_norm": 1.8003560304641724,
+      "learning_rate": 2.992519034653852e-05,
+      "loss": 0.8166,
+      "step": 643
+    },
+    {
+      "epoch": 0.3389473684210526,
+      "grad_norm": 1.138811469078064,
+      "learning_rate": 2.992494222321821e-05,
+      "loss": 0.8121,
+      "step": 644
+    },
+    {
+      "epoch": 0.3394736842105263,
+      "grad_norm": 1.1089503765106201,
+      "learning_rate": 2.992469369013208e-05,
+      "loss": 1.1773,
+      "step": 645
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.685154914855957,
+      "learning_rate": 2.992444474728694e-05,
+      "loss": 1.0016,
+      "step": 646
+    },
+    {
+      "epoch": 0.3405263157894737,
+      "grad_norm": 1.1541014909744263,
+      "learning_rate": 2.9924195394689635e-05,
+      "loss": 1.0952,
+      "step": 647
+    },
+    {
+      "epoch": 0.3410526315789474,
+      "grad_norm": 1.8897780179977417,
+      "learning_rate": 2.9923945632347002e-05,
+      "loss": 1.2069,
+      "step": 648
+    },
+    {
+      "epoch": 0.34157894736842104,
+      "grad_norm": 3.347874879837036,
+      "learning_rate": 2.9923695460265912e-05,
+      "loss": 1.0908,
+      "step": 649
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 12.232775688171387,
+      "learning_rate": 2.992344487845322e-05,
+      "loss": 1.9784,
+      "step": 650
+    },
+    {
+      "epoch": 0.3426315789473684,
+      "grad_norm": 2.2362685203552246,
+      "learning_rate": 2.992319388691581e-05,
+      "loss": 0.3282,
+      "step": 651
+    },
+    {
+      "epoch": 0.3431578947368421,
+      "grad_norm": 1.7842745780944824,
+      "learning_rate": 2.9922942485660577e-05,
+      "loss": 1.34,
+      "step": 652
+    },
+    {
+      "epoch": 0.3436842105263158,
+      "grad_norm": 4.246903419494629,
+      "learning_rate": 2.9922690674694418e-05,
+      "loss": 0.8923,
+      "step": 653
+    },
+    {
+      "epoch": 0.34421052631578947,
+      "grad_norm": 5.96973991394043,
+      "learning_rate": 2.9922438454024246e-05,
+      "loss": 0.6434,
+      "step": 654
+    },
+    {
+      "epoch": 0.3447368421052632,
+      "grad_norm": 3.631617784500122,
+      "learning_rate": 2.992218582365699e-05,
+      "loss": 0.9737,
+      "step": 655
+    },
+    {
+      "epoch": 0.3452631578947368,
+      "grad_norm": 11.352360725402832,
+      "learning_rate": 2.9921932783599585e-05,
+      "loss": 1.8412,
+      "step": 656
+    },
+    {
+      "epoch": 0.34578947368421054,
+      "grad_norm": 1.3203455209732056,
+      "learning_rate": 2.9921679333858976e-05,
+      "loss": 1.0928,
+      "step": 657
+    },
+    {
+      "epoch": 0.3463157894736842,
+      "grad_norm": 1.1191567182540894,
+      "learning_rate": 2.9921425474442127e-05,
+      "loss": 1.4272,
+      "step": 658
+    },
+    {
+      "epoch": 0.3468421052631579,
+      "grad_norm": 2.8861405849456787,
+      "learning_rate": 2.9921171205356e-05,
+      "loss": 0.9221,
+      "step": 659
+    },
+    {
+      "epoch": 0.3473684210526316,
+      "grad_norm": 2.145612955093384,
+      "learning_rate": 2.992091652660758e-05,
+      "loss": 0.3259,
+      "step": 660
+    },
+    {
+      "epoch": 0.34789473684210526,
+      "grad_norm": 3.5325045585632324,
+      "learning_rate": 2.9920661438203862e-05,
+      "loss": 2.93,
+      "step": 661
+    },
+    {
+      "epoch": 0.34842105263157896,
+      "grad_norm": 1.2416492700576782,
+      "learning_rate": 2.9920405940151842e-05,
+      "loss": 1.2741,
+      "step": 662
+    },
+    {
+      "epoch": 0.3489473684210526,
+      "grad_norm": 2.492912530899048,
+      "learning_rate": 2.9920150032458538e-05,
+      "loss": 0.9804,
+      "step": 663
+    },
+    {
+      "epoch": 0.3494736842105263,
+      "grad_norm": 10.124689102172852,
+      "learning_rate": 2.9919893715130983e-05,
+      "loss": 1.3878,
+      "step": 664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.77699613571167,
+      "learning_rate": 2.9919636988176208e-05,
+      "loss": 1.2149,
+      "step": 665
+    },
+    {
+      "epoch": 0.3505263157894737,
+      "grad_norm": 1.3935060501098633,
+      "learning_rate": 2.9919379851601256e-05,
+      "loss": 1.2551,
+      "step": 666
+    },
+    {
+      "epoch": 0.3510526315789474,
+      "grad_norm": 5.249575138092041,
+      "learning_rate": 2.9919122305413196e-05,
+      "loss": 0.7921,
+      "step": 667
+    },
+    {
+      "epoch": 0.35157894736842105,
+      "grad_norm": 3.862658739089966,
+      "learning_rate": 2.9918864349619094e-05,
+      "loss": 0.269,
+      "step": 668
+    },
+    {
+      "epoch": 0.35210526315789475,
+      "grad_norm": 4.229588508605957,
+      "learning_rate": 2.991860598422604e-05,
+      "loss": 0.9079,
+      "step": 669
+    },
+    {
+      "epoch": 0.3526315789473684,
+      "grad_norm": 8.937419891357422,
+      "learning_rate": 2.9918347209241116e-05,
+      "loss": 0.694,
+      "step": 670
+    },
+    {
+      "epoch": 0.3531578947368421,
+      "grad_norm": 9.598597526550293,
+      "learning_rate": 2.9918088024671428e-05,
+      "loss": 0.4956,
+      "step": 671
+    },
+    {
+      "epoch": 0.35368421052631577,
+      "grad_norm": 11.96364974975586,
+      "learning_rate": 2.99178284305241e-05,
+      "loss": 1.865,
+      "step": 672
+    },
+    {
+      "epoch": 0.3542105263157895,
+      "grad_norm": 8.622352600097656,
+      "learning_rate": 2.9917568426806253e-05,
+      "loss": 1.3497,
+      "step": 673
+    },
+    {
+      "epoch": 0.3547368421052632,
+      "grad_norm": 1.6212915182113647,
+      "learning_rate": 2.991730801352503e-05,
+      "loss": 1.0978,
+      "step": 674
+    },
+    {
+      "epoch": 0.35526315789473684,
+      "grad_norm": 1.9762486219406128,
+      "learning_rate": 2.9917047190687578e-05,
+      "loss": 1.0297,
+      "step": 675
+    },
+    {
+      "epoch": 0.35578947368421054,
+      "grad_norm": 1.137174129486084,
+      "learning_rate": 2.991678595830106e-05,
+      "loss": 1.3445,
+      "step": 676
+    },
+    {
+      "epoch": 0.3563157894736842,
+      "grad_norm": 1.2727190256118774,
+      "learning_rate": 2.991652431637264e-05,
+      "loss": 0.6727,
+      "step": 677
+    },
+    {
+      "epoch": 0.3568421052631579,
+      "grad_norm": 3.9113903045654297,
+      "learning_rate": 2.991626226490951e-05,
+      "loss": 1.0311,
+      "step": 678
+    },
+    {
+      "epoch": 0.35736842105263156,
+      "grad_norm": 3.078036069869995,
+      "learning_rate": 2.9915999803918862e-05,
+      "loss": 1.232,
+      "step": 679
+    },
+    {
+      "epoch": 0.35789473684210527,
+      "grad_norm": 2.5189783573150635,
+      "learning_rate": 2.99157369334079e-05,
+      "loss": 1.1612,
+      "step": 680
+    },
+    {
+      "epoch": 0.358421052631579,
+      "grad_norm": 3.2792701721191406,
+      "learning_rate": 2.991547365338385e-05,
+      "loss": 0.1674,
+      "step": 681
+    },
+    {
+      "epoch": 0.3589473684210526,
+      "grad_norm": 2.9844839572906494,
+      "learning_rate": 2.9915209963853928e-05,
+      "loss": 0.2604,
+      "step": 682
+    },
+    {
+      "epoch": 0.35947368421052633,
+      "grad_norm": 1.3294965028762817,
+      "learning_rate": 2.991494586482538e-05,
+      "loss": 1.1524,
+      "step": 683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.3986382484436035,
+      "learning_rate": 2.9914681356305458e-05,
+      "loss": 1.324,
+      "step": 684
+    },
+    {
+      "epoch": 0.3605263157894737,
+      "grad_norm": 2.5278208255767822,
+      "learning_rate": 2.991441643830142e-05,
+      "loss": 0.7384,
+      "step": 685
+    },
+    {
+      "epoch": 0.36105263157894735,
+      "grad_norm": 1.2104181051254272,
+      "learning_rate": 2.991415111082054e-05,
+      "loss": 1.3023,
+      "step": 686
+    },
+    {
+      "epoch": 0.36157894736842106,
+      "grad_norm": 1.2583531141281128,
+      "learning_rate": 2.9913885373870108e-05,
+      "loss": 1.1101,
+      "step": 687
+    },
+    {
+      "epoch": 0.36210526315789476,
+      "grad_norm": 1.007744312286377,
+      "learning_rate": 2.9913619227457413e-05,
+      "loss": 1.1889,
+      "step": 688
+    },
+    {
+      "epoch": 0.3626315789473684,
+      "grad_norm": 6.841166019439697,
+      "learning_rate": 2.991335267158977e-05,
+      "loss": 1.5798,
+      "step": 689
+    },
+    {
+      "epoch": 0.3631578947368421,
+      "grad_norm": 4.756072044372559,
+      "learning_rate": 2.9913085706274485e-05,
+      "loss": 0.4925,
+      "step": 690
+    },
+    {
+      "epoch": 0.3636842105263158,
+      "grad_norm": 2.5629639625549316,
+      "learning_rate": 2.99128183315189e-05,
+      "loss": 0.872,
+      "step": 691
+    },
+    {
+      "epoch": 0.3642105263157895,
+      "grad_norm": 5.531483173370361,
+      "learning_rate": 2.9912550547330348e-05,
+      "loss": 0.8287,
+      "step": 692
+    },
+    {
+      "epoch": 0.36473684210526314,
+      "grad_norm": 1.3215042352676392,
+      "learning_rate": 2.9912282353716184e-05,
+      "loss": 1.0156,
+      "step": 693
+    },
+    {
+      "epoch": 0.36526315789473685,
+      "grad_norm": 2.0720999240875244,
+      "learning_rate": 2.9912013750683773e-05,
+      "loss": 1.2208,
+      "step": 694
+    },
+    {
+      "epoch": 0.36578947368421055,
+      "grad_norm": 1.3862260580062866,
+      "learning_rate": 2.9911744738240487e-05,
+      "loss": 1.0124,
+      "step": 695
+    },
+    {
+      "epoch": 0.3663157894736842,
+      "grad_norm": 9.334665298461914,
+      "learning_rate": 2.991147531639371e-05,
+      "loss": 0.9209,
+      "step": 696
+    },
+    {
+      "epoch": 0.3668421052631579,
+      "grad_norm": 6.73415994644165,
+      "learning_rate": 2.9911205485150846e-05,
+      "loss": 0.4709,
+      "step": 697
+    },
+    {
+      "epoch": 0.36736842105263157,
+      "grad_norm": 3.3749637603759766,
+      "learning_rate": 2.9910935244519294e-05,
+      "loss": 1.026,
+      "step": 698
+    },
+    {
+      "epoch": 0.3678947368421053,
+      "grad_norm": 7.481809139251709,
+      "learning_rate": 2.991066459450648e-05,
+      "loss": 0.6787,
+      "step": 699
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.9085243940353394,
+      "learning_rate": 2.991039353511983e-05,
+      "loss": 1.2883,
+      "step": 700
+    },
+    {
+      "epoch": 0.36894736842105263,
+      "grad_norm": 1.1209938526153564,
+      "learning_rate": 2.991012206636679e-05,
+      "loss": 1.3699,
+      "step": 701
+    },
+    {
+      "epoch": 0.36947368421052634,
+      "grad_norm": 5.468508243560791,
+      "learning_rate": 2.9909850188254814e-05,
+      "loss": 0.5472,
+      "step": 702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.069633960723877,
+      "learning_rate": 2.9909577900791367e-05,
+      "loss": 0.8953,
+      "step": 703
+    },
+    {
+      "epoch": 0.3705263157894737,
+      "grad_norm": 1.5387980937957764,
+      "learning_rate": 2.9909305203983916e-05,
+      "loss": 1.1598,
+      "step": 704
+    },
+    {
+      "epoch": 0.37105263157894736,
+      "grad_norm": 2.226820230484009,
+      "learning_rate": 2.9909032097839958e-05,
+      "loss": 1.1898,
+      "step": 705
+    },
+    {
+      "epoch": 0.37157894736842106,
+      "grad_norm": 1.4225033521652222,
+      "learning_rate": 2.9908758582366985e-05,
+      "loss": 1.3026,
+      "step": 706
+    },
+    {
+      "epoch": 0.3721052631578947,
+      "grad_norm": 5.471367835998535,
+      "learning_rate": 2.9908484657572507e-05,
+      "loss": 0.5267,
+      "step": 707
+    },
+    {
+      "epoch": 0.3726315789473684,
+      "grad_norm": 2.0702991485595703,
+      "learning_rate": 2.9908210323464047e-05,
+      "loss": 1.0605,
+      "step": 708
+    },
+    {
+      "epoch": 0.37315789473684213,
+      "grad_norm": 8.954567909240723,
+      "learning_rate": 2.9907935580049136e-05,
+      "loss": 0.5651,
+      "step": 709
+    },
+    {
+      "epoch": 0.3736842105263158,
+      "grad_norm": 1.6635040044784546,
+      "learning_rate": 2.9907660427335324e-05,
+      "loss": 0.6742,
+      "step": 710
+    },
+    {
+      "epoch": 0.3742105263157895,
+      "grad_norm": 2.7905099391937256,
+      "learning_rate": 2.990738486533015e-05,
+      "loss": 0.1368,
+      "step": 711
+    },
+    {
+      "epoch": 0.37473684210526315,
+      "grad_norm": 2.07755970954895,
+      "learning_rate": 2.990710889404119e-05,
+      "loss": 0.3577,
+      "step": 712
+    },
+    {
+      "epoch": 0.37526315789473685,
+      "grad_norm": 12.278775215148926,
+      "learning_rate": 2.9906832513476022e-05,
+      "loss": 1.7846,
+      "step": 713
+    },
+    {
+      "epoch": 0.3757894736842105,
+      "grad_norm": 2.079127073287964,
+      "learning_rate": 2.990655572364223e-05,
+      "loss": 1.0179,
+      "step": 714
+    },
+    {
+      "epoch": 0.3763157894736842,
+      "grad_norm": 3.017181396484375,
+      "learning_rate": 2.990627852454741e-05,
+      "loss": 1.0674,
+      "step": 715
+    },
+    {
+      "epoch": 0.37684210526315787,
+      "grad_norm": 4.084104537963867,
+      "learning_rate": 2.9906000916199182e-05,
+      "loss": 1.392,
+      "step": 716
+    },
+    {
+      "epoch": 0.3773684210526316,
+      "grad_norm": 2.539130210876465,
+      "learning_rate": 2.9905722898605162e-05,
+      "loss": 1.4383,
+      "step": 717
+    },
+    {
+      "epoch": 0.3778947368421053,
+      "grad_norm": 2.1830291748046875,
+      "learning_rate": 2.9905444471772978e-05,
+      "loss": 2.1873,
+      "step": 718
+    },
+    {
+      "epoch": 0.37842105263157894,
+      "grad_norm": 2.0604703426361084,
+      "learning_rate": 2.9905165635710286e-05,
+      "loss": 1.3176,
+      "step": 719
+    },
+    {
+      "epoch": 0.37894736842105264,
+      "grad_norm": 1.2084317207336426,
+      "learning_rate": 2.990488639042473e-05,
+      "loss": 0.9763,
+      "step": 720
+    },
+    {
+      "epoch": 0.3794736842105263,
+      "grad_norm": 4.228016376495361,
+      "learning_rate": 2.9904606735923988e-05,
+      "loss": 1.2977,
+      "step": 721
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.167438983917236,
+      "learning_rate": 2.990432667221573e-05,
+      "loss": 2.1188,
+      "step": 722
+    },
+    {
+      "epoch": 0.38052631578947366,
+      "grad_norm": 2.2469797134399414,
+      "learning_rate": 2.9904046199307645e-05,
+      "loss": 1.2899,
+      "step": 723
+    },
+    {
+      "epoch": 0.38105263157894737,
+      "grad_norm": 1.2914834022521973,
+      "learning_rate": 2.9903765317207436e-05,
+      "loss": 1.1014,
+      "step": 724
+    },
+    {
+      "epoch": 0.3815789473684211,
+      "grad_norm": 1.4903666973114014,
+      "learning_rate": 2.9903484025922815e-05,
+      "loss": 1.505,
+      "step": 725
+    },
+    {
+      "epoch": 0.3821052631578947,
+      "grad_norm": 7.811427593231201,
+      "learning_rate": 2.9903202325461504e-05,
+      "loss": 1.1364,
+      "step": 726
+    },
+    {
+      "epoch": 0.38263157894736843,
+      "grad_norm": 1.9753170013427734,
+      "learning_rate": 2.9902920215831238e-05,
+      "loss": 1.3579,
+      "step": 727
+    },
+    {
+      "epoch": 0.3831578947368421,
+      "grad_norm": 4.232260227203369,
+      "learning_rate": 2.990263769703976e-05,
+      "loss": 0.3801,
+      "step": 728
+    },
+    {
+      "epoch": 0.3836842105263158,
+      "grad_norm": 4.112211227416992,
+      "learning_rate": 2.9902354769094828e-05,
+      "loss": 1.2913,
+      "step": 729
+    },
+    {
+      "epoch": 0.38421052631578945,
+      "grad_norm": 1.6348377466201782,
+      "learning_rate": 2.990207143200421e-05,
+      "loss": 0.8208,
+      "step": 730
+    },
+    {
+      "epoch": 0.38473684210526315,
+      "grad_norm": 2.1267893314361572,
+      "learning_rate": 2.9901787685775682e-05,
+      "loss": 1.4784,
+      "step": 731
+    },
+    {
+      "epoch": 0.38526315789473686,
+      "grad_norm": 1.5363768339157104,
+      "learning_rate": 2.990150353041704e-05,
+      "loss": 1.1479,
+      "step": 732
+    },
+    {
+      "epoch": 0.3857894736842105,
+      "grad_norm": 1.329637885093689,
+      "learning_rate": 2.9901218965936085e-05,
+      "loss": 1.1127,
+      "step": 733
+    },
+    {
+      "epoch": 0.3863157894736842,
+      "grad_norm": 5.673161029815674,
+      "learning_rate": 2.9900933992340627e-05,
+      "loss": 1.6329,
+      "step": 734
+    },
+    {
+      "epoch": 0.3868421052631579,
+      "grad_norm": 2.2011849880218506,
+      "learning_rate": 2.9900648609638487e-05,
+      "loss": 0.7907,
+      "step": 735
+    },
+    {
+      "epoch": 0.3873684210526316,
+      "grad_norm": 8.786739349365234,
+      "learning_rate": 2.9900362817837506e-05,
+      "loss": 1.9727,
+      "step": 736
+    },
+    {
+      "epoch": 0.38789473684210524,
+      "grad_norm": 1.4037230014801025,
+      "learning_rate": 2.9900076616945527e-05,
+      "loss": 0.0393,
+      "step": 737
+    },
+    {
+      "epoch": 0.38842105263157894,
+      "grad_norm": 1.1987260580062866,
+      "learning_rate": 2.989979000697041e-05,
+      "loss": 0.5779,
+      "step": 738
+    },
+    {
+      "epoch": 0.38894736842105265,
+      "grad_norm": 1.5265027284622192,
+      "learning_rate": 2.989950298792002e-05,
+      "loss": 1.1865,
+      "step": 739
+    },
+    {
+      "epoch": 0.3894736842105263,
+      "grad_norm": 2.8836402893066406,
+      "learning_rate": 2.9899215559802243e-05,
+      "loss": 1.974,
+      "step": 740
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7887353897094727,
+      "learning_rate": 2.9898927722624966e-05,
+      "loss": 0.1404,
+      "step": 741
+    },
+    {
+      "epoch": 0.39052631578947367,
+      "grad_norm": 6.494203567504883,
+      "learning_rate": 2.9898639476396095e-05,
+      "loss": 0.5194,
+      "step": 742
+    },
+    {
+      "epoch": 0.3910526315789474,
+      "grad_norm": 2.408672332763672,
+      "learning_rate": 2.9898350821123536e-05,
+      "loss": 0.6292,
+      "step": 743
+    },
+    {
+      "epoch": 0.391578947368421,
+      "grad_norm": 1.4377915859222412,
+      "learning_rate": 2.989806175681523e-05,
+      "loss": 1.4094,
+      "step": 744
+    },
+    {
+      "epoch": 0.39210526315789473,
+      "grad_norm": 1.4970815181732178,
+      "learning_rate": 2.9897772283479092e-05,
+      "loss": 0.8209,
+      "step": 745
+    },
+    {
+      "epoch": 0.39263157894736844,
+      "grad_norm": 2.360936403274536,
+      "learning_rate": 2.9897482401123088e-05,
+      "loss": 1.5443,
+      "step": 746
+    },
+    {
+      "epoch": 0.3931578947368421,
+      "grad_norm": 5.188052177429199,
+      "learning_rate": 2.9897192109755162e-05,
+      "loss": 1.5809,
+      "step": 747
+    },
+    {
+      "epoch": 0.3936842105263158,
+      "grad_norm": 1.4602450132369995,
+      "learning_rate": 2.9896901409383296e-05,
+      "loss": 0.0463,
+      "step": 748
+    },
+    {
+      "epoch": 0.39421052631578946,
+      "grad_norm": 4.554551601409912,
+      "learning_rate": 2.9896610300015463e-05,
+      "loss": 0.9516,
+      "step": 749
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 1.1301685571670532,
+      "learning_rate": 2.9896318781659662e-05,
+      "loss": 1.0014,
+      "step": 750
+    },
+    {
+      "epoch": 0.3952631578947368,
+      "grad_norm": 1.8906985521316528,
+      "learning_rate": 2.9896026854323896e-05,
+      "loss": 1.1485,
+      "step": 751
+    },
+    {
+      "epoch": 0.3957894736842105,
+      "grad_norm": 9.192191123962402,
+      "learning_rate": 2.9895734518016174e-05,
+      "loss": 0.8871,
+      "step": 752
+    },
+    {
+      "epoch": 0.39631578947368423,
+      "grad_norm": 1.4287827014923096,
+      "learning_rate": 2.9895441772744526e-05,
+      "loss": 1.3338,
+      "step": 753
+    },
+    {
+      "epoch": 0.3968421052631579,
+      "grad_norm": 1.5386351346969604,
+      "learning_rate": 2.989514861851699e-05,
+      "loss": 0.8937,
+      "step": 754
+    },
+    {
+      "epoch": 0.3973684210526316,
+      "grad_norm": 4.3271379470825195,
+      "learning_rate": 2.989485505534161e-05,
+      "loss": 0.9811,
+      "step": 755
+    },
+    {
+      "epoch": 0.39789473684210525,
+      "grad_norm": 13.09150505065918,
+      "learning_rate": 2.9894561083226452e-05,
+      "loss": 1.4184,
+      "step": 756
+    },
+    {
+      "epoch": 0.39842105263157895,
+      "grad_norm": 2.528856039047241,
+      "learning_rate": 2.9894266702179586e-05,
+      "loss": 1.3771,
+      "step": 757
+    },
+    {
+      "epoch": 0.3989473684210526,
+      "grad_norm": 1.3942149877548218,
+      "learning_rate": 2.989397191220909e-05,
+      "loss": 1.0354,
+      "step": 758
+    },
+    {
+      "epoch": 0.3994736842105263,
+      "grad_norm": 1.3833954334259033,
+      "learning_rate": 2.989367671332306e-05,
+      "loss": 1.2427,
+      "step": 759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.9078752994537354,
+      "learning_rate": 2.98933811055296e-05,
+      "loss": 0.2044,
+      "step": 760
+    },
+    {
+      "epoch": 0.4005263157894737,
+      "grad_norm": 4.573373794555664,
+      "learning_rate": 2.9893085088836828e-05,
+      "loss": 1.4173,
+      "step": 761
+    },
+    {
+      "epoch": 0.4010526315789474,
+      "grad_norm": 1.6260179281234741,
+      "learning_rate": 2.989278866325287e-05,
+      "loss": 0.8061,
+      "step": 762
+    },
+    {
+      "epoch": 0.40157894736842104,
+      "grad_norm": 53.73946762084961,
+      "learning_rate": 2.9892491828785866e-05,
+      "loss": 5.7539,
+      "step": 763
+    },
+    {
+      "epoch": 0.40210526315789474,
+      "grad_norm": 1.7073310613632202,
+      "learning_rate": 2.9892194585443964e-05,
+      "loss": 1.3272,
+      "step": 764
+    },
+    {
+      "epoch": 0.4026315789473684,
+      "grad_norm": 1.7463268041610718,
+      "learning_rate": 2.9891896933235324e-05,
+      "loss": 0.6709,
+      "step": 765
+    },
+    {
+      "epoch": 0.4031578947368421,
+      "grad_norm": 1.671372652053833,
+      "learning_rate": 2.9891598872168116e-05,
+      "loss": 1.0062,
+      "step": 766
+    },
+    {
+      "epoch": 0.4036842105263158,
+      "grad_norm": 3.4793472290039062,
+      "learning_rate": 2.989130040225053e-05,
+      "loss": 1.4948,
+      "step": 767
+    },
+    {
+      "epoch": 0.40421052631578946,
+      "grad_norm": 2.485945701599121,
+      "learning_rate": 2.9891001523490754e-05,
+      "loss": 1.3262,
+      "step": 768
+    },
+    {
+      "epoch": 0.4047368421052632,
+      "grad_norm": 1.4725351333618164,
+      "learning_rate": 2.9890702235897e-05,
+      "loss": 1.1616,
+      "step": 769
+    },
+    {
+      "epoch": 0.4052631578947368,
+      "grad_norm": 2.2476260662078857,
+      "learning_rate": 2.9890402539477476e-05,
+      "loss": 1.4187,
+      "step": 770
+    },
+    {
+      "epoch": 0.40578947368421053,
+      "grad_norm": 0.10535408556461334,
+      "learning_rate": 2.9890102434240415e-05,
+      "loss": 0.002,
+      "step": 771
+    },
+    {
+      "epoch": 0.4063157894736842,
+      "grad_norm": 1.7888716459274292,
+      "learning_rate": 2.9889801920194062e-05,
+      "loss": 1.9264,
+      "step": 772
+    },
+    {
+      "epoch": 0.4068421052631579,
+      "grad_norm": 3.374720573425293,
+      "learning_rate": 2.988950099734666e-05,
+      "loss": 0.3194,
+      "step": 773
+    },
+    {
+      "epoch": 0.4073684210526316,
+      "grad_norm": 1.5872454643249512,
+      "learning_rate": 2.9889199665706475e-05,
+      "loss": 0.8289,
+      "step": 774
+    },
+    {
+      "epoch": 0.40789473684210525,
+      "grad_norm": 4.640042304992676,
+      "learning_rate": 2.988889792528178e-05,
+      "loss": 0.8441,
+      "step": 775
+    },
+    {
+      "epoch": 0.40842105263157896,
+      "grad_norm": 5.191117763519287,
+      "learning_rate": 2.9888595776080856e-05,
+      "loss": 2.0781,
+      "step": 776
+    },
+    {
+      "epoch": 0.4089473684210526,
+      "grad_norm": 1.595027208328247,
+      "learning_rate": 2.9888293218111998e-05,
+      "loss": 1.5206,
+      "step": 777
+    },
+    {
+      "epoch": 0.4094736842105263,
+      "grad_norm": 2.689854145050049,
+      "learning_rate": 2.988799025138352e-05,
+      "loss": 0.7231,
+      "step": 778
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 16.25735092163086,
+      "learning_rate": 2.988768687590373e-05,
+      "loss": 1.6409,
+      "step": 779
+    },
+    {
+      "epoch": 0.4105263157894737,
+      "grad_norm": 1.3042892217636108,
+      "learning_rate": 2.9887383091680964e-05,
+      "loss": 1.0157,
+      "step": 780
+    },
+    {
+      "epoch": 0.4110526315789474,
+      "grad_norm": 5.346875190734863,
+      "learning_rate": 2.9887078898723564e-05,
+      "loss": 0.657,
+      "step": 781
+    },
+    {
+      "epoch": 0.41157894736842104,
+      "grad_norm": 1.5279061794281006,
+      "learning_rate": 2.9886774297039878e-05,
+      "loss": 1.4934,
+      "step": 782
+    },
+    {
+      "epoch": 0.41210526315789475,
+      "grad_norm": 12.0784330368042,
+      "learning_rate": 2.9886469286638265e-05,
+      "loss": 3.1849,
+      "step": 783
+    },
+    {
+      "epoch": 0.4126315789473684,
+      "grad_norm": 1.3375122547149658,
+      "learning_rate": 2.9886163867527107e-05,
+      "loss": 1.1898,
+      "step": 784
+    },
+    {
+      "epoch": 0.4131578947368421,
+      "grad_norm": 4.016144275665283,
+      "learning_rate": 2.9885858039714786e-05,
+      "loss": 0.863,
+      "step": 785
+    },
+    {
+      "epoch": 0.41368421052631577,
+      "grad_norm": 2.205333948135376,
+      "learning_rate": 2.98855518032097e-05,
+      "loss": 0.8202,
+      "step": 786
+    },
+    {
+      "epoch": 0.4142105263157895,
+      "grad_norm": 2.130064010620117,
+      "learning_rate": 2.988524515802025e-05,
+      "loss": 1.7152,
+      "step": 787
+    },
+    {
+      "epoch": 0.4147368421052632,
+      "grad_norm": 2.0593249797821045,
+      "learning_rate": 2.9884938104154864e-05,
+      "loss": 1.2797,
+      "step": 788
+    },
+    {
+      "epoch": 0.41526315789473683,
+      "grad_norm": 4.233097553253174,
+      "learning_rate": 2.9884630641621963e-05,
+      "loss": 1.3987,
+      "step": 789
+    },
+    {
+      "epoch": 0.41578947368421054,
+      "grad_norm": 6.048382759094238,
+      "learning_rate": 2.9884322770429996e-05,
+      "loss": 1.7027,
+      "step": 790
+    },
+    {
+      "epoch": 0.4163157894736842,
+      "grad_norm": 4.876536846160889,
+      "learning_rate": 2.9884014490587418e-05,
+      "loss": 1.1787,
+      "step": 791
+    },
+    {
+      "epoch": 0.4168421052631579,
+      "grad_norm": 3.066049337387085,
+      "learning_rate": 2.9883705802102684e-05,
+      "loss": 2.2892,
+      "step": 792
+    },
+    {
+      "epoch": 0.41736842105263156,
+      "grad_norm": 3.7404627799987793,
+      "learning_rate": 2.9883396704984273e-05,
+      "loss": 0.4776,
+      "step": 793
+    },
+    {
+      "epoch": 0.41789473684210526,
+      "grad_norm": 1.4917809963226318,
+      "learning_rate": 2.9883087199240672e-05,
+      "loss": 1.1282,
+      "step": 794
+    },
+    {
+      "epoch": 0.41842105263157897,
+      "grad_norm": 1.7566046714782715,
+      "learning_rate": 2.988277728488038e-05,
+      "loss": 1.3632,
+      "step": 795
+    },
+    {
+      "epoch": 0.4189473684210526,
+      "grad_norm": 3.263943910598755,
+      "learning_rate": 2.98824669619119e-05,
+      "loss": 1.245,
+      "step": 796
+    },
+    {
+      "epoch": 0.41947368421052633,
+      "grad_norm": 1.3323115110397339,
+      "learning_rate": 2.9882156230343755e-05,
+      "loss": 0.8262,
+      "step": 797
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.155140399932861,
+      "learning_rate": 2.988184509018448e-05,
+      "loss": 1.6147,
+      "step": 798
+    },
+    {
+      "epoch": 0.4205263157894737,
+      "grad_norm": 1.3995517492294312,
+      "learning_rate": 2.9881533541442615e-05,
+      "loss": 1.3523,
+      "step": 799
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.254440188407898,
+      "learning_rate": 2.9881221584126716e-05,
+      "loss": 0.7949,
+      "step": 800
+    },
+    {
+      "epoch": 0.42157894736842105,
+      "grad_norm": 1.74281644821167,
+      "learning_rate": 2.9880909218245335e-05,
+      "loss": 1.1625,
+      "step": 801
+    },
+    {
+      "epoch": 0.42210526315789476,
+      "grad_norm": 0.7531611323356628,
+      "learning_rate": 2.9880596443807065e-05,
+      "loss": 0.0144,
+      "step": 802
+    },
+    {
+      "epoch": 0.4226315789473684,
+      "grad_norm": 2.432140827178955,
+      "learning_rate": 2.9880283260820485e-05,
+      "loss": 1.2728,
+      "step": 803
+    },
+    {
+      "epoch": 0.4231578947368421,
+      "grad_norm": 12.019746780395508,
+      "learning_rate": 2.9879969669294193e-05,
+      "loss": 0.3237,
+      "step": 804
+    },
+    {
+      "epoch": 0.4236842105263158,
+      "grad_norm": 9.18172836303711,
+      "learning_rate": 2.98796556692368e-05,
+      "loss": 1.1377,
+      "step": 805
+    },
+    {
+      "epoch": 0.4242105263157895,
+      "grad_norm": 6.744290351867676,
+      "learning_rate": 2.9879341260656926e-05,
+      "loss": 0.8986,
+      "step": 806
+    },
+    {
+      "epoch": 0.42473684210526313,
+      "grad_norm": 6.863618850708008,
+      "learning_rate": 2.9879026443563207e-05,
+      "loss": 0.7325,
+      "step": 807
+    },
+    {
+      "epoch": 0.42526315789473684,
+      "grad_norm": 2.109163761138916,
+      "learning_rate": 2.9878711217964284e-05,
+      "loss": 0.7444,
+      "step": 808
+    },
+    {
+      "epoch": 0.42578947368421055,
+      "grad_norm": 1.5105681419372559,
+      "learning_rate": 2.9878395583868807e-05,
+      "loss": 0.0472,
+      "step": 809
+    },
+    {
+      "epoch": 0.4263157894736842,
+      "grad_norm": 2.5132837295532227,
+      "learning_rate": 2.9878079541285445e-05,
+      "loss": 1.1468,
+      "step": 810
+    },
+    {
+      "epoch": 0.4268421052631579,
+      "grad_norm": 1.0575823783874512,
+      "learning_rate": 2.987776309022288e-05,
+      "loss": 0.7931,
+      "step": 811
+    },
+    {
+      "epoch": 0.42736842105263156,
+      "grad_norm": 2.099186658859253,
+      "learning_rate": 2.9877446230689795e-05,
+      "loss": 0.1261,
+      "step": 812
+    },
+    {
+      "epoch": 0.42789473684210527,
+      "grad_norm": 2.151189088821411,
+      "learning_rate": 2.9877128962694892e-05,
+      "loss": 1.3023,
+      "step": 813
+    },
+    {
+      "epoch": 0.4284210526315789,
+      "grad_norm": 4.0430707931518555,
+      "learning_rate": 2.987681128624688e-05,
+      "loss": 0.4215,
+      "step": 814
+    },
+    {
+      "epoch": 0.42894736842105263,
+      "grad_norm": 2.170595169067383,
+      "learning_rate": 2.9876493201354475e-05,
+      "loss": 1.6202,
+      "step": 815
+    },
+    {
+      "epoch": 0.42947368421052634,
+      "grad_norm": 7.825321674346924,
+      "learning_rate": 2.987617470802642e-05,
+      "loss": 0.9704,
+      "step": 816
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3321274518966675,
+      "learning_rate": 2.9875855806271455e-05,
+      "loss": 0.985,
+      "step": 817
+    },
+    {
+      "epoch": 0.4305263157894737,
+      "grad_norm": 3.4284095764160156,
+      "learning_rate": 2.9875536496098335e-05,
+      "loss": 1.6423,
+      "step": 818
+    },
+    {
+      "epoch": 0.43105263157894735,
+      "grad_norm": 1.4579600095748901,
+      "learning_rate": 2.9875216777515827e-05,
+      "loss": 1.0348,
+      "step": 819
+    },
+    {
+      "epoch": 0.43157894736842106,
+      "grad_norm": 1.253885269165039,
+      "learning_rate": 2.987489665053271e-05,
+      "loss": 0.6495,
+      "step": 820
+    },
+    {
+      "epoch": 0.4321052631578947,
+      "grad_norm": 4.7803144454956055,
+      "learning_rate": 2.9874576115157773e-05,
+      "loss": 0.2175,
+      "step": 821
+    },
+    {
+      "epoch": 0.4326315789473684,
+      "grad_norm": 2.1819753646850586,
+      "learning_rate": 2.987425517139981e-05,
+      "loss": 0.9445,
+      "step": 822
+    },
+    {
+      "epoch": 0.43315789473684213,
+      "grad_norm": 1.4593207836151123,
+      "learning_rate": 2.9873933819267647e-05,
+      "loss": 1.3716,
+      "step": 823
+    },
+    {
+      "epoch": 0.4336842105263158,
+      "grad_norm": 2.473689556121826,
+      "learning_rate": 2.9873612058770094e-05,
+      "loss": 0.5247,
+      "step": 824
+    },
+    {
+      "epoch": 0.4342105263157895,
+      "grad_norm": 2.997178792953491,
+      "learning_rate": 2.9873289889915986e-05,
+      "loss": 0.9902,
+      "step": 825
+    },
+    {
+      "epoch": 0.43473684210526314,
+      "grad_norm": 6.612910270690918,
+      "learning_rate": 2.9872967312714176e-05,
+      "loss": 1.2973,
+      "step": 826
+    },
+    {
+      "epoch": 0.43526315789473685,
+      "grad_norm": 1.0687013864517212,
+      "learning_rate": 2.9872644327173513e-05,
+      "loss": 0.7058,
+      "step": 827
+    },
+    {
+      "epoch": 0.4357894736842105,
+      "grad_norm": 1.554075837135315,
+      "learning_rate": 2.9872320933302867e-05,
+      "loss": 1.4401,
+      "step": 828
+    },
+    {
+      "epoch": 0.4363157894736842,
+      "grad_norm": 4.839378833770752,
+      "learning_rate": 2.9871997131111122e-05,
+      "loss": 1.1889,
+      "step": 829
+    },
+    {
+      "epoch": 0.4368421052631579,
+      "grad_norm": 3.0694851875305176,
+      "learning_rate": 2.9871672920607158e-05,
+      "loss": 1.0211,
+      "step": 830
+    },
+    {
+      "epoch": 0.4373684210526316,
+      "grad_norm": 1.378533124923706,
+      "learning_rate": 2.9871348301799883e-05,
+      "loss": 0.6821,
+      "step": 831
+    },
+    {
+      "epoch": 0.4378947368421053,
+      "grad_norm": 1.238147497177124,
+      "learning_rate": 2.9871023274698204e-05,
+      "loss": 0.0303,
+      "step": 832
+    },
+    {
+      "epoch": 0.43842105263157893,
+      "grad_norm": 1.4806623458862305,
+      "learning_rate": 2.9870697839311053e-05,
+      "loss": 1.2068,
+      "step": 833
+    },
+    {
+      "epoch": 0.43894736842105264,
+      "grad_norm": 1.3188241720199585,
+      "learning_rate": 2.9870371995647353e-05,
+      "loss": 0.6705,
+      "step": 834
+    },
+    {
+      "epoch": 0.4394736842105263,
+      "grad_norm": 4.55914831161499,
+      "learning_rate": 2.9870045743716063e-05,
+      "loss": 0.0898,
+      "step": 835
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.970885694026947,
+      "learning_rate": 2.9869719083526137e-05,
+      "loss": 1.0416,
+      "step": 836
+    },
+    {
+      "epoch": 0.4405263157894737,
+      "grad_norm": 1.664801836013794,
+      "learning_rate": 2.9869392015086538e-05,
+      "loss": 1.3485,
+      "step": 837
+    },
+    {
+      "epoch": 0.44105263157894736,
+      "grad_norm": 2.7517263889312744,
+      "learning_rate": 2.9869064538406247e-05,
+      "loss": 1.0703,
+      "step": 838
+    },
+    {
+      "epoch": 0.44157894736842107,
+      "grad_norm": 0.9083302021026611,
+      "learning_rate": 2.986873665349426e-05,
+      "loss": 0.0277,
+      "step": 839
+    },
+    {
+      "epoch": 0.4421052631578947,
+      "grad_norm": 2.142995834350586,
+      "learning_rate": 2.986840836035957e-05,
+      "loss": 1.0772,
+      "step": 840
+    },
+    {
+      "epoch": 0.44263157894736843,
+      "grad_norm": 0.20032253861427307,
+      "learning_rate": 2.98680796590112e-05,
+      "loss": 0.0051,
+      "step": 841
+    },
+    {
+      "epoch": 0.4431578947368421,
+      "grad_norm": 0.8623015880584717,
+      "learning_rate": 2.9867750549458173e-05,
+      "loss": 0.6315,
+      "step": 842
+    },
+    {
+      "epoch": 0.4436842105263158,
+      "grad_norm": 1.2665982246398926,
+      "learning_rate": 2.9867421031709517e-05,
+      "loss": 1.4777,
+      "step": 843
+    },
+    {
+      "epoch": 0.4442105263157895,
+      "grad_norm": 0.974398672580719,
+      "learning_rate": 2.986709110577429e-05,
+      "loss": 0.9672,
+      "step": 844
+    },
+    {
+      "epoch": 0.44473684210526315,
+      "grad_norm": 0.9700206518173218,
+      "learning_rate": 2.9866760771661544e-05,
+      "loss": 1.0432,
+      "step": 845
+    },
+    {
+      "epoch": 0.44526315789473686,
+      "grad_norm": 2.319387912750244,
+      "learning_rate": 2.9866430029380342e-05,
+      "loss": 1.3111,
+      "step": 846
+    },
+    {
+      "epoch": 0.4457894736842105,
+      "grad_norm": 0.9847918748855591,
+      "learning_rate": 2.9866098878939777e-05,
+      "loss": 1.1562,
+      "step": 847
+    },
+    {
+      "epoch": 0.4463157894736842,
+      "grad_norm": 5.33390474319458,
+      "learning_rate": 2.9865767320348932e-05,
+      "loss": 1.2682,
+      "step": 848
+    },
+    {
+      "epoch": 0.4468421052631579,
+      "grad_norm": 9.808623313903809,
+      "learning_rate": 2.9865435353616915e-05,
+      "loss": 1.3406,
+      "step": 849
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 6.855868816375732,
+      "learning_rate": 2.9865102978752837e-05,
+      "loss": 2.3309,
+      "step": 850
+    },
+    {
+      "epoch": 0.4478947368421053,
+      "grad_norm": 2.986323595046997,
+      "learning_rate": 2.9864770195765828e-05,
+      "loss": 1.596,
+      "step": 851
+    },
+    {
+      "epoch": 0.44842105263157894,
+      "grad_norm": 2.0322799682617188,
+      "learning_rate": 2.9864437004665016e-05,
+      "loss": 1.0168,
+      "step": 852
+    },
+    {
+      "epoch": 0.44894736842105265,
+      "grad_norm": 14.111528396606445,
+      "learning_rate": 2.9864103405459556e-05,
+      "loss": 1.2123,
+      "step": 853
+    },
+    {
+      "epoch": 0.4494736842105263,
+      "grad_norm": 1.4620112180709839,
+      "learning_rate": 2.9863769398158607e-05,
+      "loss": 1.1575,
+      "step": 854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.048269271850586,
+      "learning_rate": 2.9863434982771338e-05,
+      "loss": 1.4347,
+      "step": 855
+    },
+    {
+      "epoch": 0.45052631578947366,
+      "grad_norm": 1.1362407207489014,
+      "learning_rate": 2.9863100159306923e-05,
+      "loss": 0.9232,
+      "step": 856
+    },
+    {
+      "epoch": 0.45105263157894737,
+      "grad_norm": 14.55726432800293,
+      "learning_rate": 2.9862764927774567e-05,
+      "loss": 1.6201,
+      "step": 857
+    },
+    {
+      "epoch": 0.4515789473684211,
+      "grad_norm": 4.592569828033447,
+      "learning_rate": 2.9862429288183468e-05,
+      "loss": 1.5495,
+      "step": 858
+    },
+    {
+      "epoch": 0.45210526315789473,
+      "grad_norm": 0.8918259143829346,
+      "learning_rate": 2.986209324054284e-05,
+      "loss": 0.9341,
+      "step": 859
+    },
+    {
+      "epoch": 0.45263157894736844,
+      "grad_norm": 3.0468339920043945,
+      "learning_rate": 2.986175678486191e-05,
+      "loss": 0.9043,
+      "step": 860
+    },
+    {
+      "epoch": 0.4531578947368421,
+      "grad_norm": 1.4248181581497192,
+      "learning_rate": 2.9861419921149916e-05,
+      "loss": 0.983,
+      "step": 861
+    },
+    {
+      "epoch": 0.4536842105263158,
+      "grad_norm": 2.1727728843688965,
+      "learning_rate": 2.9861082649416107e-05,
+      "loss": 0.3648,
+      "step": 862
+    },
+    {
+      "epoch": 0.45421052631578945,
+      "grad_norm": 1.469250202178955,
+      "learning_rate": 2.9860744969669742e-05,
+      "loss": 0.9197,
+      "step": 863
+    },
+    {
+      "epoch": 0.45473684210526316,
+      "grad_norm": 3.119738817214966,
+      "learning_rate": 2.986040688192009e-05,
+      "loss": 1.2948,
+      "step": 864
+    },
+    {
+      "epoch": 0.45526315789473687,
+      "grad_norm": 7.314899921417236,
+      "learning_rate": 2.9860068386176437e-05,
+      "loss": 1.1322,
+      "step": 865
+    },
+    {
+      "epoch": 0.4557894736842105,
+      "grad_norm": 1.2942012548446655,
+      "learning_rate": 2.9859729482448073e-05,
+      "loss": 1.0436,
+      "step": 866
+    },
+    {
+      "epoch": 0.45631578947368423,
+      "grad_norm": 16.526906967163086,
+      "learning_rate": 2.985939017074431e-05,
+      "loss": 0.7217,
+      "step": 867
+    },
+    {
+      "epoch": 0.4568421052631579,
+      "grad_norm": 0.9518576860427856,
+      "learning_rate": 2.9859050451074453e-05,
+      "loss": 0.0462,
+      "step": 868
+    },
+    {
+      "epoch": 0.4573684210526316,
+      "grad_norm": 1.8784120082855225,
+      "learning_rate": 2.985871032344784e-05,
+      "loss": 1.1546,
+      "step": 869
+    },
+    {
+      "epoch": 0.45789473684210524,
+      "grad_norm": 2.373981475830078,
+      "learning_rate": 2.9858369787873795e-05,
+      "loss": 1.1577,
+      "step": 870
+    },
+    {
+      "epoch": 0.45842105263157895,
+      "grad_norm": 4.2424421310424805,
+      "learning_rate": 2.985802884436168e-05,
+      "loss": 0.8525,
+      "step": 871
+    },
+    {
+      "epoch": 0.4589473684210526,
+      "grad_norm": 1.318802833557129,
+      "learning_rate": 2.9857687492920854e-05,
+      "loss": 0.7319,
+      "step": 872
+    },
+    {
+      "epoch": 0.4594736842105263,
+      "grad_norm": 1.247564435005188,
+      "learning_rate": 2.9857345733560683e-05,
+      "loss": 1.0628,
+      "step": 873
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.433790683746338,
+      "learning_rate": 2.985700356629056e-05,
+      "loss": 1.1417,
+      "step": 874
+    },
+    {
+      "epoch": 0.4605263157894737,
+      "grad_norm": 3.1146788597106934,
+      "learning_rate": 2.9856660991119867e-05,
+      "loss": 1.4214,
+      "step": 875
+    },
+    {
+      "epoch": 0.4610526315789474,
+      "grad_norm": 9.475061416625977,
+      "learning_rate": 2.9856318008058018e-05,
+      "loss": 1.015,
+      "step": 876
+    },
+    {
+      "epoch": 0.46157894736842103,
+      "grad_norm": 1.2461477518081665,
+      "learning_rate": 2.9855974617114425e-05,
+      "loss": 1.0464,
+      "step": 877
+    },
+    {
+      "epoch": 0.46210526315789474,
+      "grad_norm": 1.4924275875091553,
+      "learning_rate": 2.9855630818298518e-05,
+      "loss": 1.7069,
+      "step": 878
+    },
+    {
+      "epoch": 0.4626315789473684,
+      "grad_norm": 1.3920341730117798,
+      "learning_rate": 2.9855286611619733e-05,
+      "loss": 1.5036,
+      "step": 879
+    },
+    {
+      "epoch": 0.4631578947368421,
+      "grad_norm": 2.099785089492798,
+      "learning_rate": 2.985494199708753e-05,
+      "loss": 0.057,
+      "step": 880
+    },
+    {
+      "epoch": 0.4636842105263158,
+      "grad_norm": 2.4705018997192383,
+      "learning_rate": 2.985459697471136e-05,
+      "loss": 1.091,
+      "step": 881
+    },
+    {
+      "epoch": 0.46421052631578946,
+      "grad_norm": 7.607650279998779,
+      "learning_rate": 2.98542515445007e-05,
+      "loss": 0.3252,
+      "step": 882
+    },
+    {
+      "epoch": 0.46473684210526317,
+      "grad_norm": 1.6341148614883423,
+      "learning_rate": 2.985390570646503e-05,
+      "loss": 1.5967,
+      "step": 883
+    },
+    {
+      "epoch": 0.4652631578947368,
+      "grad_norm": 1.6126140356063843,
+      "learning_rate": 2.9853559460613846e-05,
+      "loss": 1.2722,
+      "step": 884
+    },
+    {
+      "epoch": 0.46578947368421053,
+      "grad_norm": 13.708699226379395,
+      "learning_rate": 2.985321280695666e-05,
+      "loss": 1.7056,
+      "step": 885
+    },
+    {
+      "epoch": 0.4663157894736842,
+      "grad_norm": 1.0587328672409058,
+      "learning_rate": 2.9852865745502988e-05,
+      "loss": 0.9254,
+      "step": 886
+    },
+    {
+      "epoch": 0.4668421052631579,
+      "grad_norm": 1.383562445640564,
+      "learning_rate": 2.9852518276262352e-05,
+      "loss": 0.821,
+      "step": 887
+    },
+    {
+      "epoch": 0.4673684210526316,
+      "grad_norm": 5.15762996673584,
+      "learning_rate": 2.9852170399244297e-05,
+      "loss": 1.1727,
+      "step": 888
+    },
+    {
+      "epoch": 0.46789473684210525,
+      "grad_norm": 25.248504638671875,
+      "learning_rate": 2.9851822114458374e-05,
+      "loss": 2.5076,
+      "step": 889
+    },
+    {
+      "epoch": 0.46842105263157896,
+      "grad_norm": 4.159931182861328,
+      "learning_rate": 2.985147342191414e-05,
+      "loss": 0.4708,
+      "step": 890
+    },
+    {
+      "epoch": 0.4689473684210526,
+      "grad_norm": 1.1170576810836792,
+      "learning_rate": 2.9851124321621177e-05,
+      "loss": 0.8062,
+      "step": 891
+    },
+    {
+      "epoch": 0.4694736842105263,
+      "grad_norm": 1.807730793952942,
+      "learning_rate": 2.9850774813589065e-05,
+      "loss": 0.5932,
+      "step": 892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2220526933670044,
+      "learning_rate": 2.98504248978274e-05,
+      "loss": 1.0653,
+      "step": 893
+    },
+    {
+      "epoch": 0.4705263157894737,
+      "grad_norm": 7.340139389038086,
+      "learning_rate": 2.9850074574345787e-05,
+      "loss": 2.1767,
+      "step": 894
+    },
+    {
+      "epoch": 0.4710526315789474,
+      "grad_norm": 2.8421573638916016,
+      "learning_rate": 2.9849723843153847e-05,
+      "loss": 0.619,
+      "step": 895
+    },
+    {
+      "epoch": 0.47157894736842104,
+      "grad_norm": 1.6204451322555542,
+      "learning_rate": 2.9849372704261203e-05,
+      "loss": 0.469,
+      "step": 896
+    },
+    {
+      "epoch": 0.47210526315789475,
+      "grad_norm": 1.9400737285614014,
+      "learning_rate": 2.9849021157677506e-05,
+      "loss": 1.0654,
+      "step": 897
+    },
+    {
+      "epoch": 0.4726315789473684,
+      "grad_norm": 1.4052708148956299,
+      "learning_rate": 2.9848669203412404e-05,
+      "loss": 0.8773,
+      "step": 898
+    },
+    {
+      "epoch": 0.4731578947368421,
+      "grad_norm": 6.795404434204102,
+      "learning_rate": 2.984831684147556e-05,
+      "loss": 2.0904,
+      "step": 899
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 2.273846387863159,
+      "learning_rate": 2.9847964071876642e-05,
+      "loss": 1.3245,
+      "step": 900
+    },
+    {
+      "epoch": 0.47421052631578947,
+      "grad_norm": 1.403640866279602,
+      "learning_rate": 2.9847610894625343e-05,
+      "loss": 0.8352,
+      "step": 901
+    },
+    {
+      "epoch": 0.4747368421052632,
+      "grad_norm": 1.2695461511611938,
+      "learning_rate": 2.9847257309731357e-05,
+      "loss": 1.0238,
+      "step": 902
+    },
+    {
+      "epoch": 0.47526315789473683,
+      "grad_norm": 2.6055543422698975,
+      "learning_rate": 2.9846903317204388e-05,
+      "loss": 0.2152,
+      "step": 903
+    },
+    {
+      "epoch": 0.47578947368421054,
+      "grad_norm": 2.166196584701538,
+      "learning_rate": 2.984654891705416e-05,
+      "loss": 1.2715,
+      "step": 904
+    },
+    {
+      "epoch": 0.4763157894736842,
+      "grad_norm": 7.2398247718811035,
+      "learning_rate": 2.9846194109290404e-05,
+      "loss": 2.0351,
+      "step": 905
+    },
+    {
+      "epoch": 0.4768421052631579,
+      "grad_norm": 4.5485453605651855,
+      "learning_rate": 2.9845838893922854e-05,
+      "loss": 0.6056,
+      "step": 906
+    },
+    {
+      "epoch": 0.47736842105263155,
+      "grad_norm": 2.5836188793182373,
+      "learning_rate": 2.9845483270961267e-05,
+      "loss": 2.0319,
+      "step": 907
+    },
+    {
+      "epoch": 0.47789473684210526,
+      "grad_norm": 1.6816260814666748,
+      "learning_rate": 2.984512724041541e-05,
+      "loss": 0.8913,
+      "step": 908
+    },
+    {
+      "epoch": 0.47842105263157897,
+      "grad_norm": 1.6629785299301147,
+      "learning_rate": 2.9844770802295056e-05,
+      "loss": 1.5281,
+      "step": 909
+    },
+    {
+      "epoch": 0.4789473684210526,
+      "grad_norm": 1.2143621444702148,
+      "learning_rate": 2.9844413956609985e-05,
+      "loss": 1.0262,
+      "step": 910
+    },
+    {
+      "epoch": 0.47947368421052633,
+      "grad_norm": 4.233440399169922,
+      "learning_rate": 2.984405670337e-05,
+      "loss": 1.4231,
+      "step": 911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3030482530593872,
+      "learning_rate": 2.9843699042584908e-05,
+      "loss": 0.6924,
+      "step": 912
+    },
+    {
+      "epoch": 0.4805263157894737,
+      "grad_norm": 5.886407375335693,
+      "learning_rate": 2.9843340974264532e-05,
+      "loss": 2.0718,
+      "step": 913
+    },
+    {
+      "epoch": 0.48105263157894734,
+      "grad_norm": 1.7257221937179565,
+      "learning_rate": 2.98429824984187e-05,
+      "loss": 2.0068,
+      "step": 914
+    },
+    {
+      "epoch": 0.48157894736842105,
+      "grad_norm": 2.2462844848632812,
+      "learning_rate": 2.9842623615057248e-05,
+      "loss": 1.7576,
+      "step": 915
+    },
+    {
+      "epoch": 0.48210526315789476,
+      "grad_norm": 6.659467697143555,
+      "learning_rate": 2.984226432419004e-05,
+      "loss": 0.8356,
+      "step": 916
+    },
+    {
+      "epoch": 0.4826315789473684,
+      "grad_norm": 1.0781395435333252,
+      "learning_rate": 2.9841904625826933e-05,
+      "loss": 1.0799,
+      "step": 917
+    },
+    {
+      "epoch": 0.4831578947368421,
+      "grad_norm": 1.9946705102920532,
+      "learning_rate": 2.9841544519977805e-05,
+      "loss": 1.7794,
+      "step": 918
+    },
+    {
+      "epoch": 0.48368421052631577,
+      "grad_norm": 1.2607777118682861,
+      "learning_rate": 2.9841184006652544e-05,
+      "loss": 1.1695,
+      "step": 919
+    },
+    {
+      "epoch": 0.4842105263157895,
+      "grad_norm": 1.6544026136398315,
+      "learning_rate": 2.9840823085861047e-05,
+      "loss": 1.0608,
+      "step": 920
+    },
+    {
+      "epoch": 0.48473684210526313,
+      "grad_norm": 3.6581201553344727,
+      "learning_rate": 2.9840461757613217e-05,
+      "loss": 0.1158,
+      "step": 921
+    },
+    {
+      "epoch": 0.48526315789473684,
+      "grad_norm": 1.2672442197799683,
+      "learning_rate": 2.9840100021918986e-05,
+      "loss": 1.1773,
+      "step": 922
+    },
+    {
+      "epoch": 0.48578947368421055,
+      "grad_norm": 11.554649353027344,
+      "learning_rate": 2.9839737878788276e-05,
+      "loss": 0.1758,
+      "step": 923
+    },
+    {
+      "epoch": 0.4863157894736842,
+      "grad_norm": 2.514875888824463,
+      "learning_rate": 2.9839375328231033e-05,
+      "loss": 0.6761,
+      "step": 924
+    },
+    {
+      "epoch": 0.4868421052631579,
+      "grad_norm": 1.2187618017196655,
+      "learning_rate": 2.9839012370257213e-05,
+      "loss": 0.8572,
+      "step": 925
+    },
+    {
+      "epoch": 0.48736842105263156,
+      "grad_norm": 11.437232971191406,
+      "learning_rate": 2.983864900487678e-05,
+      "loss": 1.4024,
+      "step": 926
+    },
+    {
+      "epoch": 0.48789473684210527,
+      "grad_norm": 1.1252726316452026,
+      "learning_rate": 2.9838285232099703e-05,
+      "loss": 0.8422,
+      "step": 927
+    },
+    {
+      "epoch": 0.4884210526315789,
+      "grad_norm": 7.090402603149414,
+      "learning_rate": 2.9837921051935983e-05,
+      "loss": 1.7886,
+      "step": 928
+    },
+    {
+      "epoch": 0.48894736842105263,
+      "grad_norm": 0.9771952629089355,
+      "learning_rate": 2.983755646439561e-05,
+      "loss": 0.6664,
+      "step": 929
+    },
+    {
+      "epoch": 0.48947368421052634,
+      "grad_norm": 1.0985069274902344,
+      "learning_rate": 2.983719146948859e-05,
+      "loss": 0.8674,
+      "step": 930
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.855616092681885,
+      "learning_rate": 2.9836826067224953e-05,
+      "loss": 1.08,
+      "step": 931
+    },
+    {
+      "epoch": 0.4905263157894737,
+      "grad_norm": 4.352982044219971,
+      "learning_rate": 2.9836460257614726e-05,
+      "loss": 1.0946,
+      "step": 932
+    },
+    {
+      "epoch": 0.49105263157894735,
+      "grad_norm": 9.891953468322754,
+      "learning_rate": 2.9836094040667953e-05,
+      "loss": 0.5332,
+      "step": 933
+    },
+    {
+      "epoch": 0.49157894736842106,
+      "grad_norm": 4.198849678039551,
+      "learning_rate": 2.9835727416394692e-05,
+      "loss": 0.5912,
+      "step": 934
+    },
+    {
+      "epoch": 0.4921052631578947,
+      "grad_norm": 3.755746841430664,
+      "learning_rate": 2.9835360384805004e-05,
+      "loss": 1.9771,
+      "step": 935
+    },
+    {
+      "epoch": 0.4926315789473684,
+      "grad_norm": 1.8602410554885864,
+      "learning_rate": 2.9834992945908966e-05,
+      "loss": 0.9501,
+      "step": 936
+    },
+    {
+      "epoch": 0.49315789473684213,
+      "grad_norm": 7.063650608062744,
+      "learning_rate": 2.9834625099716668e-05,
+      "loss": 1.1328,
+      "step": 937
+    },
+    {
+      "epoch": 0.4936842105263158,
+      "grad_norm": 48.79571533203125,
+      "learning_rate": 2.983425684623821e-05,
+      "loss": 3.1328,
+      "step": 938
+    },
+    {
+      "epoch": 0.4942105263157895,
+      "grad_norm": 8.612360954284668,
+      "learning_rate": 2.98338881854837e-05,
+      "loss": 0.6469,
+      "step": 939
+    },
+    {
+      "epoch": 0.49473684210526314,
+      "grad_norm": 5.5089898109436035,
+      "learning_rate": 2.9833519117463263e-05,
+      "loss": 2.3209,
+      "step": 940
+    },
+    {
+      "epoch": 0.49526315789473685,
+      "grad_norm": 1.2845944166183472,
+      "learning_rate": 2.9833149642187026e-05,
+      "loss": 1.2498,
+      "step": 941
+    },
+    {
+      "epoch": 0.4957894736842105,
+      "grad_norm": 2.241567373275757,
+      "learning_rate": 2.9832779759665144e-05,
+      "loss": 0.1963,
+      "step": 942
+    },
+    {
+      "epoch": 0.4963157894736842,
+      "grad_norm": 1.6336573362350464,
+      "learning_rate": 2.983240946990776e-05,
+      "loss": 0.6764,
+      "step": 943
+    },
+    {
+      "epoch": 0.4968421052631579,
+      "grad_norm": 2.2285687923431396,
+      "learning_rate": 2.9832038772925044e-05,
+      "loss": 1.1649,
+      "step": 944
+    },
+    {
+      "epoch": 0.49736842105263157,
+      "grad_norm": 1.5614920854568481,
+      "learning_rate": 2.983166766872718e-05,
+      "loss": 1.1911,
+      "step": 945
+    },
+    {
+      "epoch": 0.4978947368421053,
+      "grad_norm": 1.2409616708755493,
+      "learning_rate": 2.9831296157324348e-05,
+      "loss": 0.9644,
+      "step": 946
+    },
+    {
+      "epoch": 0.49842105263157893,
+      "grad_norm": 1.8802260160446167,
+      "learning_rate": 2.983092423872675e-05,
+      "loss": 0.6794,
+      "step": 947
+    },
+    {
+      "epoch": 0.49894736842105264,
+      "grad_norm": 1.1808290481567383,
+      "learning_rate": 2.98305519129446e-05,
+      "loss": 1.2996,
+      "step": 948
+    },
+    {
+      "epoch": 0.4994736842105263,
+      "grad_norm": 1.2290716171264648,
+      "learning_rate": 2.983017917998812e-05,
+      "loss": 1.0257,
+      "step": 949
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3215667009353638,
+      "learning_rate": 2.9829806039867537e-05,
+      "loss": 1.1965,
+      "step": 950
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.0288052558898926,
+      "eval_runtime": 12.7382,
+      "eval_samples_per_second": 7.85,
+      "eval_steps_per_second": 7.85,
+      "step": 950
+    },
+    {
+      "epoch": 0.5005263157894737,
+      "grad_norm": 1.5079114437103271,
+      "learning_rate": 2.9829432492593105e-05,
+      "loss": 1.0393,
+      "step": 951
+    },
+    {
+      "epoch": 0.5010526315789474,
+      "grad_norm": 6.629293441772461,
+      "learning_rate": 2.9829058538175076e-05,
+      "loss": 1.3261,
+      "step": 952
+    },
+    {
+      "epoch": 0.501578947368421,
+      "grad_norm": 6.566483974456787,
+      "learning_rate": 2.9828684176623714e-05,
+      "loss": 1.561,
+      "step": 953
+    },
+    {
+      "epoch": 0.5021052631578947,
+      "grad_norm": 1.9273375272750854,
+      "learning_rate": 2.98283094079493e-05,
+      "loss": 1.0629,
+      "step": 954
+    },
+    {
+      "epoch": 0.5026315789473684,
+      "grad_norm": 2.6924264430999756,
+      "learning_rate": 2.9827934232162128e-05,
+      "loss": 0.967,
+      "step": 955
+    },
+    {
+      "epoch": 0.5031578947368421,
+      "grad_norm": 1.5618788003921509,
+      "learning_rate": 2.982755864927249e-05,
+      "loss": 1.2447,
+      "step": 956
+    },
+    {
+      "epoch": 0.5036842105263157,
+      "grad_norm": 2.599567174911499,
+      "learning_rate": 2.98271826592907e-05,
+      "loss": 0.9259,
+      "step": 957
+    },
+    {
+      "epoch": 0.5042105263157894,
+      "grad_norm": 5.357890605926514,
+      "learning_rate": 2.9826806262227082e-05,
+      "loss": 0.7411,
+      "step": 958
+    },
+    {
+      "epoch": 0.5047368421052632,
+      "grad_norm": 2.8332183361053467,
+      "learning_rate": 2.9826429458091968e-05,
+      "loss": 0.1058,
+      "step": 959
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 1.5336825847625732,
+      "learning_rate": 2.982605224689571e-05,
+      "loss": 0.2564,
+      "step": 960
+    },
+    {
+      "epoch": 0.5057894736842106,
+      "grad_norm": 3.167874336242676,
+      "learning_rate": 2.9825674628648657e-05,
+      "loss": 0.9333,
+      "step": 961
+    },
+    {
+      "epoch": 0.5063157894736842,
+      "grad_norm": 1.4255472421646118,
+      "learning_rate": 2.982529660336118e-05,
+      "loss": 0.0605,
+      "step": 962
+    },
+    {
+      "epoch": 0.5068421052631579,
+      "grad_norm": 5.443603038787842,
+      "learning_rate": 2.9824918171043656e-05,
+      "loss": 0.9617,
+      "step": 963
+    },
+    {
+      "epoch": 0.5073684210526316,
+      "grad_norm": 1.5958141088485718,
+      "learning_rate": 2.9824539331706476e-05,
+      "loss": 1.8712,
+      "step": 964
+    },
+    {
+      "epoch": 0.5078947368421053,
+      "grad_norm": 1.544083595275879,
+      "learning_rate": 2.982416008536004e-05,
+      "loss": 1.9258,
+      "step": 965
+    },
+    {
+      "epoch": 0.508421052631579,
+      "grad_norm": 9.469457626342773,
+      "learning_rate": 2.982378043201476e-05,
+      "loss": 1.3554,
+      "step": 966
+    },
+    {
+      "epoch": 0.5089473684210526,
+      "grad_norm": 1.1175442934036255,
+      "learning_rate": 2.9823400371681062e-05,
+      "loss": 1.5603,
+      "step": 967
+    },
+    {
+      "epoch": 0.5094736842105263,
+      "grad_norm": 5.4420342445373535,
+      "learning_rate": 2.9823019904369377e-05,
+      "loss": 0.5942,
+      "step": 968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5822449922561646,
+      "learning_rate": 2.9822639030090156e-05,
+      "loss": 1.0452,
+      "step": 969
+    },
+    {
+      "epoch": 0.5105263157894737,
+      "grad_norm": 4.8439483642578125,
+      "learning_rate": 2.9822257748853846e-05,
+      "loss": 0.2948,
+      "step": 970
+    },
+    {
+      "epoch": 0.5110526315789473,
+      "grad_norm": 5.272324085235596,
+      "learning_rate": 2.982187606067093e-05,
+      "loss": 1.0793,
+      "step": 971
+    },
+    {
+      "epoch": 0.511578947368421,
+      "grad_norm": 35.564964294433594,
+      "learning_rate": 2.9821493965551877e-05,
+      "loss": 4.9608,
+      "step": 972
+    },
+    {
+      "epoch": 0.5121052631578947,
+      "grad_norm": 8.41565227508545,
+      "learning_rate": 2.9821111463507177e-05,
+      "loss": 0.8292,
+      "step": 973
+    },
+    {
+      "epoch": 0.5126315789473684,
+      "grad_norm": 1.3578308820724487,
+      "learning_rate": 2.9820728554547338e-05,
+      "loss": 0.6588,
+      "step": 974
+    },
+    {
+      "epoch": 0.5131578947368421,
+      "grad_norm": 21.759471893310547,
+      "learning_rate": 2.9820345238682862e-05,
+      "loss": 5.2813,
+      "step": 975
+    },
+    {
+      "epoch": 0.5136842105263157,
+      "grad_norm": 1.089369535446167,
+      "learning_rate": 2.9819961515924288e-05,
+      "loss": 0.7576,
+      "step": 976
+    },
+    {
+      "epoch": 0.5142105263157895,
+      "grad_norm": 1.6790835857391357,
+      "learning_rate": 2.981957738628214e-05,
+      "loss": 1.0218,
+      "step": 977
+    },
+    {
+      "epoch": 0.5147368421052632,
+      "grad_norm": 0.8437327742576599,
+      "learning_rate": 2.9819192849766965e-05,
+      "loss": 0.459,
+      "step": 978
+    },
+    {
+      "epoch": 0.5152631578947369,
+      "grad_norm": 1.327656865119934,
+      "learning_rate": 2.981880790638933e-05,
+      "loss": 1.0783,
+      "step": 979
+    },
+    {
+      "epoch": 0.5157894736842106,
+      "grad_norm": 1.32616126537323,
+      "learning_rate": 2.981842255615979e-05,
+      "loss": 1.0975,
+      "step": 980
+    },
+    {
+      "epoch": 0.5163157894736842,
+      "grad_norm": 1.0787245035171509,
+      "learning_rate": 2.981803679908893e-05,
+      "loss": 1.1274,
+      "step": 981
+    },
+    {
+      "epoch": 0.5168421052631579,
+      "grad_norm": 5.349323272705078,
+      "learning_rate": 2.9817650635187348e-05,
+      "loss": 0.6531,
+      "step": 982
+    },
+    {
+      "epoch": 0.5173684210526316,
+      "grad_norm": 6.324141979217529,
+      "learning_rate": 2.981726406446564e-05,
+      "loss": 1.4738,
+      "step": 983
+    },
+    {
+      "epoch": 0.5178947368421053,
+      "grad_norm": 1.0158735513687134,
+      "learning_rate": 2.9816877086934416e-05,
+      "loss": 0.7424,
+      "step": 984
+    },
+    {
+      "epoch": 0.5184210526315789,
+      "grad_norm": 1.3030285835266113,
+      "learning_rate": 2.9816489702604307e-05,
+      "loss": 1.2574,
+      "step": 985
+    },
+    {
+      "epoch": 0.5189473684210526,
+      "grad_norm": 1.0021758079528809,
+      "learning_rate": 2.9816101911485944e-05,
+      "loss": 0.9003,
+      "step": 986
+    },
+    {
+      "epoch": 0.5194736842105263,
+      "grad_norm": 2.6594042778015137,
+      "learning_rate": 2.981571371358998e-05,
+      "loss": 0.4392,
+      "step": 987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.258882999420166,
+      "learning_rate": 2.981532510892707e-05,
+      "loss": 0.3734,
+      "step": 988
+    },
+    {
+      "epoch": 0.5205263157894737,
+      "grad_norm": 4.136307239532471,
+      "learning_rate": 2.9814936097507878e-05,
+      "loss": 1.9032,
+      "step": 989
+    },
+    {
+      "epoch": 0.5210526315789473,
+      "grad_norm": 4.121081352233887,
+      "learning_rate": 2.981454667934309e-05,
+      "loss": 1.5468,
+      "step": 990
+    },
+    {
+      "epoch": 0.521578947368421,
+      "grad_norm": 6.502449035644531,
+      "learning_rate": 2.9814156854443394e-05,
+      "loss": 1.3444,
+      "step": 991
+    },
+    {
+      "epoch": 0.5221052631578947,
+      "grad_norm": 9.599536895751953,
+      "learning_rate": 2.9813766622819494e-05,
+      "loss": 0.8619,
+      "step": 992
+    },
+    {
+      "epoch": 0.5226315789473684,
+      "grad_norm": 5.463332653045654,
+      "learning_rate": 2.9813375984482108e-05,
+      "loss": 0.8028,
+      "step": 993
+    },
+    {
+      "epoch": 0.5231578947368422,
+      "grad_norm": 2.675305128097534,
+      "learning_rate": 2.9812984939441955e-05,
+      "loss": 0.1877,
+      "step": 994
+    },
+    {
+      "epoch": 0.5236842105263158,
+      "grad_norm": 1.6895674467086792,
+      "learning_rate": 2.9812593487709778e-05,
+      "loss": 1.4556,
+      "step": 995
+    },
+    {
+      "epoch": 0.5242105263157895,
+      "grad_norm": 1.0445210933685303,
+      "learning_rate": 2.9812201629296313e-05,
+      "loss": 0.9134,
+      "step": 996
+    },
+    {
+      "epoch": 0.5247368421052632,
+      "grad_norm": 4.284270286560059,
+      "learning_rate": 2.9811809364212332e-05,
+      "loss": 0.7209,
+      "step": 997
+    },
+    {
+      "epoch": 0.5252631578947369,
+      "grad_norm": 3.461217164993286,
+      "learning_rate": 2.98114166924686e-05,
+      "loss": 0.4954,
+      "step": 998
+    },
+    {
+      "epoch": 0.5257894736842105,
+      "grad_norm": 4.461343288421631,
+      "learning_rate": 2.981102361407589e-05,
+      "loss": 1.5761,
+      "step": 999
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.5852874517440796,
+      "learning_rate": 2.9810630129045003e-05,
+      "loss": 1.1905,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5268421052631579,
+      "grad_norm": 2.101400375366211,
+      "learning_rate": 2.9810236237386736e-05,
+      "loss": 1.3343,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5273684210526316,
+      "grad_norm": 10.774019241333008,
+      "learning_rate": 2.980984193911191e-05,
+      "loss": 1.6976,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5278947368421053,
+      "grad_norm": 1.1593283414840698,
+      "learning_rate": 2.9809447234231347e-05,
+      "loss": 0.9587,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5284210526315789,
+      "grad_norm": 5.7743754386901855,
+      "learning_rate": 2.9809052122755885e-05,
+      "loss": 0.9899,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5289473684210526,
+      "grad_norm": 1.3892626762390137,
+      "learning_rate": 2.9808656604696368e-05,
+      "loss": 1.1057,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5294736842105263,
+      "grad_norm": 3.0294156074523926,
+      "learning_rate": 2.980826068006366e-05,
+      "loss": 0.6647,
+      "step": 1006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.9338483810424805,
+      "learning_rate": 2.9807864348868627e-05,
+      "loss": 1.9628,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5305263157894737,
+      "grad_norm": 4.924473762512207,
+      "learning_rate": 2.980746761112215e-05,
+      "loss": 0.7927,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5310526315789473,
+      "grad_norm": 10.412948608398438,
+      "learning_rate": 2.980707046683513e-05,
+      "loss": 1.1254,
+      "step": 1009
+    },
+    {
+      "epoch": 0.531578947368421,
+      "grad_norm": 9.721678733825684,
+      "learning_rate": 2.980667291601846e-05,
+      "loss": 1.9085,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5321052631578947,
+      "grad_norm": 3.1283535957336426,
+      "learning_rate": 2.980627495868306e-05,
+      "loss": 1.4154,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5326315789473685,
+      "grad_norm": 3.373187303543091,
+      "learning_rate": 2.980587659483985e-05,
+      "loss": 1.3,
+      "step": 1012
+    },
+    {
+      "epoch": 0.533157894736842,
+      "grad_norm": 8.386382102966309,
+      "learning_rate": 2.9805477824499782e-05,
+      "loss": 0.2979,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5336842105263158,
+      "grad_norm": 2.241272211074829,
+      "learning_rate": 2.980507864767379e-05,
+      "loss": 0.2733,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5342105263157895,
+      "grad_norm": 1.518618106842041,
+      "learning_rate": 2.9804679064372836e-05,
+      "loss": 1.8392,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5347368421052632,
+      "grad_norm": 1.0135964155197144,
+      "learning_rate": 2.9804279074607893e-05,
+      "loss": 1.0983,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5352631578947369,
+      "grad_norm": 1.5758641958236694,
+      "learning_rate": 2.9803878678389942e-05,
+      "loss": 1.0834,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5357894736842105,
+      "grad_norm": 3.014568328857422,
+      "learning_rate": 2.9803477875729977e-05,
+      "loss": 0.4795,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5363157894736842,
+      "grad_norm": 1.2805556058883667,
+      "learning_rate": 2.9803076666639e-05,
+      "loss": 1.2692,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5368421052631579,
+      "grad_norm": 5.261938571929932,
+      "learning_rate": 2.9802675051128027e-05,
+      "loss": 0.8414,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5373684210526316,
+      "grad_norm": 3.743300676345825,
+      "learning_rate": 2.9802273029208085e-05,
+      "loss": 1.3266,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5378947368421053,
+      "grad_norm": 4.708719253540039,
+      "learning_rate": 2.9801870600890216e-05,
+      "loss": 0.132,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5384210526315789,
+      "grad_norm": 6.205244064331055,
+      "learning_rate": 2.9801467766185457e-05,
+      "loss": 0.2055,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5389473684210526,
+      "grad_norm": 11.566554069519043,
+      "learning_rate": 2.980106452510488e-05,
+      "loss": 0.1921,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5394736842105263,
+      "grad_norm": 2.550504446029663,
+      "learning_rate": 2.9800660877659546e-05,
+      "loss": 0.6581,
+      "step": 1025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.586308002471924,
+      "learning_rate": 2.9800256823860548e-05,
+      "loss": 0.2069,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5405263157894736,
+      "grad_norm": 2.6851885318756104,
+      "learning_rate": 2.9799852363718968e-05,
+      "loss": 1.5168,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5410526315789473,
+      "grad_norm": 1.1070444583892822,
+      "learning_rate": 2.979944749724592e-05,
+      "loss": 0.8637,
+      "step": 1028
+    },
+    {
+      "epoch": 0.541578947368421,
+      "grad_norm": 0.5393061637878418,
+      "learning_rate": 2.979904222445252e-05,
+      "loss": 0.0123,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5421052631578948,
+      "grad_norm": 4.805749416351318,
+      "learning_rate": 2.979863654534988e-05,
+      "loss": 0.6532,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5426315789473685,
+      "grad_norm": 1.2212395668029785,
+      "learning_rate": 2.9798230459949154e-05,
+      "loss": 1.0349,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5431578947368421,
+      "grad_norm": 1.0666829347610474,
+      "learning_rate": 2.9797823968261483e-05,
+      "loss": 0.816,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5436842105263158,
+      "grad_norm": 1.6240037679672241,
+      "learning_rate": 2.9797417070298033e-05,
+      "loss": 1.1863,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5442105263157895,
+      "grad_norm": 80.0791244506836,
+      "learning_rate": 2.979700976606997e-05,
+      "loss": 3.8047,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5447368421052632,
+      "grad_norm": 7.121600151062012,
+      "learning_rate": 2.979660205558848e-05,
+      "loss": 1.8238,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5452631578947369,
+      "grad_norm": 5.182654857635498,
+      "learning_rate": 2.9796193938864758e-05,
+      "loss": 1.7294,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5457894736842105,
+      "grad_norm": 1.5659658908843994,
+      "learning_rate": 2.9795785415910002e-05,
+      "loss": 1.3367,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5463157894736842,
+      "grad_norm": 2.371896266937256,
+      "learning_rate": 2.979537648673544e-05,
+      "loss": 0.6634,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5468421052631579,
+      "grad_norm": 6.189696311950684,
+      "learning_rate": 2.979496715135228e-05,
+      "loss": 1.1687,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5473684210526316,
+      "grad_norm": 3.078950881958008,
+      "learning_rate": 2.9794557409771775e-05,
+      "loss": 0.2819,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5478947368421052,
+      "grad_norm": 1.354801893234253,
+      "learning_rate": 2.9794147262005182e-05,
+      "loss": 1.0736,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5484210526315789,
+      "grad_norm": 7.661222457885742,
+      "learning_rate": 2.9793736708063735e-05,
+      "loss": 0.757,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5489473684210526,
+      "grad_norm": 1.1365411281585693,
+      "learning_rate": 2.9793325747958736e-05,
+      "loss": 0.9885,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5494736842105263,
+      "grad_norm": 2.5599515438079834,
+      "learning_rate": 2.9792914381701448e-05,
+      "loss": 1.5999,
+      "step": 1044
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.250511169433594,
+      "learning_rate": 2.979250260930317e-05,
+      "loss": 0.8261,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5505263157894736,
+      "grad_norm": 5.77577543258667,
+      "learning_rate": 2.9792090430775212e-05,
+      "loss": 1.4922,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5510526315789473,
+      "grad_norm": 10.645989418029785,
+      "learning_rate": 2.979167784612888e-05,
+      "loss": 1.5039,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5515789473684211,
+      "grad_norm": 22.560731887817383,
+      "learning_rate": 2.9791264855375516e-05,
+      "loss": 1.1108,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5521052631578948,
+      "grad_norm": 0.9451678395271301,
+      "learning_rate": 2.9790851458526445e-05,
+      "loss": 1.0083,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 1.7276222705841064,
+      "learning_rate": 2.9790437655593025e-05,
+      "loss": 1.2542,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5531578947368421,
+      "grad_norm": 12.997016906738281,
+      "learning_rate": 2.9790023446586616e-05,
+      "loss": 1.5738,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5536842105263158,
+      "grad_norm": 5.94761848449707,
+      "learning_rate": 2.9789608831518585e-05,
+      "loss": 0.5001,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5542105263157895,
+      "grad_norm": 3.714102029800415,
+      "learning_rate": 2.9789193810400324e-05,
+      "loss": 0.4654,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5547368421052632,
+      "grad_norm": 1.2676738500595093,
+      "learning_rate": 2.978877838324322e-05,
+      "loss": 0.9719,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5552631578947368,
+      "grad_norm": 4.261148452758789,
+      "learning_rate": 2.9788362550058683e-05,
+      "loss": 0.481,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5557894736842105,
+      "grad_norm": 2.568772077560425,
+      "learning_rate": 2.9787946310858126e-05,
+      "loss": 0.2567,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5563157894736842,
+      "grad_norm": 17.453506469726562,
+      "learning_rate": 2.9787529665652983e-05,
+      "loss": 1.1513,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5568421052631579,
+      "grad_norm": 4.880800724029541,
+      "learning_rate": 2.9787112614454682e-05,
+      "loss": 0.7086,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5573684210526316,
+      "grad_norm": 2.3661293983459473,
+      "learning_rate": 2.9786695157274686e-05,
+      "loss": 0.3181,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5578947368421052,
+      "grad_norm": 1.5389021635055542,
+      "learning_rate": 2.9786277294124446e-05,
+      "loss": 1.1855,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5584210526315789,
+      "grad_norm": 1.2657150030136108,
+      "learning_rate": 2.978585902501544e-05,
+      "loss": 1.0923,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5589473684210526,
+      "grad_norm": 1.9125850200653076,
+      "learning_rate": 2.9785440349959154e-05,
+      "loss": 1.5275,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5594736842105263,
+      "grad_norm": 1.2199132442474365,
+      "learning_rate": 2.978502126896708e-05,
+      "loss": 0.9496,
+      "step": 1063
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.9119816422462463,
+      "learning_rate": 2.9784601782050716e-05,
+      "loss": 1.1481,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5605263157894737,
+      "grad_norm": 1.264197587966919,
+      "learning_rate": 2.9784181889221588e-05,
+      "loss": 1.0066,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5610526315789474,
+      "grad_norm": 1.3065340518951416,
+      "learning_rate": 2.978376159049123e-05,
+      "loss": 1.5019,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5615789473684211,
+      "grad_norm": 5.354816913604736,
+      "learning_rate": 2.978334088587117e-05,
+      "loss": 1.2623,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5621052631578948,
+      "grad_norm": 1.1491811275482178,
+      "learning_rate": 2.9782919775372958e-05,
+      "loss": 0.9462,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5626315789473684,
+      "grad_norm": 1.9805035591125488,
+      "learning_rate": 2.9782498259008163e-05,
+      "loss": 0.8934,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5631578947368421,
+      "grad_norm": 6.078151702880859,
+      "learning_rate": 2.9782076336788355e-05,
+      "loss": 2.088,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5636842105263158,
+      "grad_norm": 1.2258517742156982,
+      "learning_rate": 2.9781654008725118e-05,
+      "loss": 1.2418,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5642105263157895,
+      "grad_norm": 4.132420539855957,
+      "learning_rate": 2.9781231274830045e-05,
+      "loss": 1.0597,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5647368421052632,
+      "grad_norm": 2.6240031719207764,
+      "learning_rate": 2.9780808135114745e-05,
+      "loss": 1.8114,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5652631578947368,
+      "grad_norm": 5.4064154624938965,
+      "learning_rate": 2.9780384589590832e-05,
+      "loss": 1.7053,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5657894736842105,
+      "grad_norm": 1.50705885887146,
+      "learning_rate": 2.9779960638269944e-05,
+      "loss": 1.2326,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5663157894736842,
+      "grad_norm": 4.527035713195801,
+      "learning_rate": 2.9779536281163707e-05,
+      "loss": 1.0151,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5668421052631579,
+      "grad_norm": 2.740678071975708,
+      "learning_rate": 2.9779111518283778e-05,
+      "loss": 1.3425,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5673684210526316,
+      "grad_norm": 1.8033603429794312,
+      "learning_rate": 2.977868634964182e-05,
+      "loss": 1.4094,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5678947368421052,
+      "grad_norm": 2.819211483001709,
+      "learning_rate": 2.9778260775249507e-05,
+      "loss": 0.5859,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5684210526315789,
+      "grad_norm": 2.2031707763671875,
+      "learning_rate": 2.9777834795118516e-05,
+      "loss": 1.6467,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5689473684210526,
+      "grad_norm": 1.0411590337753296,
+      "learning_rate": 2.9777408409260556e-05,
+      "loss": 1.1856,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5694736842105264,
+      "grad_norm": 1.240462303161621,
+      "learning_rate": 2.977698161768732e-05,
+      "loss": 0.9601,
+      "step": 1082
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4794399738311768,
+      "learning_rate": 2.977655442041053e-05,
+      "loss": 0.6428,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5705263157894737,
+      "grad_norm": 3.423283576965332,
+      "learning_rate": 2.9776126817441918e-05,
+      "loss": 0.5233,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5710526315789474,
+      "grad_norm": 1.0337262153625488,
+      "learning_rate": 2.9775698808793216e-05,
+      "loss": 1.2001,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5715789473684211,
+      "grad_norm": 2.4375438690185547,
+      "learning_rate": 2.9775270394476187e-05,
+      "loss": 0.8193,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5721052631578948,
+      "grad_norm": 1.048431158065796,
+      "learning_rate": 2.9774841574502584e-05,
+      "loss": 1.0908,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5726315789473684,
+      "grad_norm": 1.6780253648757935,
+      "learning_rate": 2.9774412348884184e-05,
+      "loss": 1.7247,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5731578947368421,
+      "grad_norm": 3.6449801921844482,
+      "learning_rate": 2.9773982717632768e-05,
+      "loss": 1.8992,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5736842105263158,
+      "grad_norm": 2.392915725708008,
+      "learning_rate": 2.9773552680760133e-05,
+      "loss": 0.5871,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5742105263157895,
+      "grad_norm": 25.228097915649414,
+      "learning_rate": 2.9773122238278088e-05,
+      "loss": 1.6074,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5747368421052632,
+      "grad_norm": 11.282571792602539,
+      "learning_rate": 2.9772691390198452e-05,
+      "loss": 1.8768,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5752631578947368,
+      "grad_norm": 6.598527908325195,
+      "learning_rate": 2.9772260136533048e-05,
+      "loss": 1.2115,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5757894736842105,
+      "grad_norm": 0.990164577960968,
+      "learning_rate": 2.977182847729372e-05,
+      "loss": 1.0003,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5763157894736842,
+      "grad_norm": 2.4671967029571533,
+      "learning_rate": 2.9771396412492316e-05,
+      "loss": 0.0873,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5768421052631579,
+      "grad_norm": 2.4675517082214355,
+      "learning_rate": 2.977096394214071e-05,
+      "loss": 1.2669,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5773684210526315,
+      "grad_norm": 1.001388430595398,
+      "learning_rate": 2.9770531066250754e-05,
+      "loss": 1.3442,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5778947368421052,
+      "grad_norm": 2.501490831375122,
+      "learning_rate": 2.9770097784834352e-05,
+      "loss": 0.8013,
+      "step": 1098
+    },
+    {
+      "epoch": 0.578421052631579,
+      "grad_norm": 1.80197274684906,
+      "learning_rate": 2.9769664097903394e-05,
+      "loss": 1.3928,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 1.4630690813064575,
+      "learning_rate": 2.9769230005469785e-05,
+      "loss": 1.5564,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5794736842105264,
+      "grad_norm": 2.549288034439087,
+      "learning_rate": 2.9768795507545444e-05,
+      "loss": 0.8852,
+      "step": 1101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0548112392425537,
+      "learning_rate": 2.97683606041423e-05,
+      "loss": 1.2653,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5805263157894737,
+      "grad_norm": 2.3279740810394287,
+      "learning_rate": 2.9767925295272292e-05,
+      "loss": 1.3083,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5810526315789474,
+      "grad_norm": 1.0182585716247559,
+      "learning_rate": 2.9767489580947375e-05,
+      "loss": 0.0161,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5815789473684211,
+      "grad_norm": 2.9629054069519043,
+      "learning_rate": 2.976705346117951e-05,
+      "loss": 1.6623,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5821052631578948,
+      "grad_norm": 5.767422199249268,
+      "learning_rate": 2.9766616935980668e-05,
+      "loss": 0.9786,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5826315789473684,
+      "grad_norm": 4.677647590637207,
+      "learning_rate": 2.9766180005362835e-05,
+      "loss": 0.3723,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5831578947368421,
+      "grad_norm": 1.2974311113357544,
+      "learning_rate": 2.9765742669338012e-05,
+      "loss": 0.793,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5836842105263158,
+      "grad_norm": 1.323523759841919,
+      "learning_rate": 2.97653049279182e-05,
+      "loss": 1.2581,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5842105263157895,
+      "grad_norm": 2.9345805644989014,
+      "learning_rate": 2.9764866781115417e-05,
+      "loss": 0.7428,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5847368421052631,
+      "grad_norm": 8.754313468933105,
+      "learning_rate": 2.9764428228941702e-05,
+      "loss": 1.5404,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5852631578947368,
+      "grad_norm": 5.4265875816345215,
+      "learning_rate": 2.976398927140908e-05,
+      "loss": 2.3934,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5857894736842105,
+      "grad_norm": 3.4678874015808105,
+      "learning_rate": 2.9763549908529614e-05,
+      "loss": 1.079,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5863157894736842,
+      "grad_norm": 9.305590629577637,
+      "learning_rate": 2.9763110140315365e-05,
+      "loss": 1.6112,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5868421052631579,
+      "grad_norm": 2.989252805709839,
+      "learning_rate": 2.9762669966778406e-05,
+      "loss": 0.6553,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5873684210526315,
+      "grad_norm": 1.2564352750778198,
+      "learning_rate": 2.9762229387930817e-05,
+      "loss": 1.0478,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5878947368421052,
+      "grad_norm": 2.9239699840545654,
+      "learning_rate": 2.9761788403784703e-05,
+      "loss": 1.4082,
+      "step": 1117
+    },
+    {
+      "epoch": 0.588421052631579,
+      "grad_norm": 2.38325834274292,
+      "learning_rate": 2.9761347014352168e-05,
+      "loss": 1.1565,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5889473684210527,
+      "grad_norm": 4.686885356903076,
+      "learning_rate": 2.9760905219645325e-05,
+      "loss": 1.3522,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5894736842105263,
+      "grad_norm": 1.1427439451217651,
+      "learning_rate": 2.976046301967631e-05,
+      "loss": 0.7014,
+      "step": 1120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.191329002380371,
+      "learning_rate": 2.976002041445726e-05,
+      "loss": 1.1254,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5905263157894737,
+      "grad_norm": 1.5960139036178589,
+      "learning_rate": 2.9759577404000332e-05,
+      "loss": 1.5535,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5910526315789474,
+      "grad_norm": 1.104758858680725,
+      "learning_rate": 2.975913398831768e-05,
+      "loss": 1.2053,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5915789473684211,
+      "grad_norm": 8.011280059814453,
+      "learning_rate": 2.9758690167421486e-05,
+      "loss": 1.5689,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5921052631578947,
+      "grad_norm": 2.1890242099761963,
+      "learning_rate": 2.9758245941323932e-05,
+      "loss": 1.3737,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5926315789473684,
+      "grad_norm": 5.8035502433776855,
+      "learning_rate": 2.9757801310037218e-05,
+      "loss": 1.4497,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5931578947368421,
+      "grad_norm": 3.329540967941284,
+      "learning_rate": 2.9757356273573543e-05,
+      "loss": 0.3897,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5936842105263158,
+      "grad_norm": 3.944429874420166,
+      "learning_rate": 2.9756910831945136e-05,
+      "loss": 1.0921,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5942105263157895,
+      "grad_norm": 3.396190643310547,
+      "learning_rate": 2.9756464985164214e-05,
+      "loss": 0.4456,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5947368421052631,
+      "grad_norm": 1.586104393005371,
+      "learning_rate": 2.9756018733243033e-05,
+      "loss": 1.401,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5952631578947368,
+      "grad_norm": 1.2480674982070923,
+      "learning_rate": 2.9755572076193833e-05,
+      "loss": 1.3656,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5957894736842105,
+      "grad_norm": 1.0409289598464966,
+      "learning_rate": 2.9755125014028876e-05,
+      "loss": 1.1957,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5963157894736842,
+      "grad_norm": 1.0716015100479126,
+      "learning_rate": 2.9754677546760444e-05,
+      "loss": 0.9709,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5968421052631578,
+      "grad_norm": 1.9948137998580933,
+      "learning_rate": 2.9754229674400822e-05,
+      "loss": 1.1416,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5973684210526315,
+      "grad_norm": 1.064031720161438,
+      "learning_rate": 2.9753781396962303e-05,
+      "loss": 0.9325,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5978947368421053,
+      "grad_norm": 2.7508461475372314,
+      "learning_rate": 2.9753332714457193e-05,
+      "loss": 0.1521,
+      "step": 1136
+    },
+    {
+      "epoch": 0.598421052631579,
+      "grad_norm": 2.104701519012451,
+      "learning_rate": 2.9752883626897813e-05,
+      "loss": 1.3733,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5989473684210527,
+      "grad_norm": 1.180884838104248,
+      "learning_rate": 2.9752434134296494e-05,
+      "loss": 1.0614,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5994736842105263,
+      "grad_norm": 1.3136661052703857,
+      "learning_rate": 2.9751984236665578e-05,
+      "loss": 0.761,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.025672435760498,
+      "learning_rate": 2.975153393401741e-05,
+      "loss": 0.8013,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6005263157894737,
+      "grad_norm": 0.5844347476959229,
+      "learning_rate": 2.975108322636436e-05,
+      "loss": 0.0122,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6010526315789474,
+      "grad_norm": 1.406011700630188,
+      "learning_rate": 2.9750632113718795e-05,
+      "loss": 1.2908,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6015789473684211,
+      "grad_norm": 51.026851654052734,
+      "learning_rate": 2.9750180596093108e-05,
+      "loss": 2.5605,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6021052631578947,
+      "grad_norm": 1.4514580965042114,
+      "learning_rate": 2.9749728673499695e-05,
+      "loss": 1.0439,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6026315789473684,
+      "grad_norm": 1.7348302602767944,
+      "learning_rate": 2.9749276345950957e-05,
+      "loss": 1.8712,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6031578947368421,
+      "grad_norm": 2.1238996982574463,
+      "learning_rate": 2.974882361345932e-05,
+      "loss": 1.0086,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6036842105263158,
+      "grad_norm": 3.6767418384552,
+      "learning_rate": 2.974837047603721e-05,
+      "loss": 0.2294,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6042105263157894,
+      "grad_norm": 7.874570369720459,
+      "learning_rate": 2.9747916933697066e-05,
+      "loss": 2.2632,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6047368421052631,
+      "grad_norm": 8.193496704101562,
+      "learning_rate": 2.9747462986451347e-05,
+      "loss": 1.793,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 1.6370347738265991,
+      "learning_rate": 2.9747008634312508e-05,
+      "loss": 1.4331,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6057894736842105,
+      "grad_norm": 2.3346762657165527,
+      "learning_rate": 2.9746553877293025e-05,
+      "loss": 1.4356,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6063157894736843,
+      "grad_norm": 3.7243125438690186,
+      "learning_rate": 2.9746098715405394e-05,
+      "loss": 1.8991,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6068421052631578,
+      "grad_norm": 4.2178168296813965,
+      "learning_rate": 2.9745643148662095e-05,
+      "loss": 1.3015,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6073684210526316,
+      "grad_norm": 2.0063540935516357,
+      "learning_rate": 2.9745187177075645e-05,
+      "loss": 1.0399,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6078947368421053,
+      "grad_norm": 3.9274532794952393,
+      "learning_rate": 2.9744730800658563e-05,
+      "loss": 0.9394,
+      "step": 1155
+    },
+    {
+      "epoch": 0.608421052631579,
+      "grad_norm": 0.9116756319999695,
+      "learning_rate": 2.9744274019423375e-05,
+      "loss": 0.7879,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6089473684210527,
+      "grad_norm": 1.725326418876648,
+      "learning_rate": 2.9743816833382627e-05,
+      "loss": 1.2359,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6094736842105263,
+      "grad_norm": 2.4110519886016846,
+      "learning_rate": 2.9743359242548865e-05,
+      "loss": 1.0438,
+      "step": 1158
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.515199661254883,
+      "learning_rate": 2.9742901246934657e-05,
+      "loss": 1.0046,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6105263157894737,
+      "grad_norm": 3.636535167694092,
+      "learning_rate": 2.9742442846552578e-05,
+      "loss": 1.0394,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6110526315789474,
+      "grad_norm": 8.082115173339844,
+      "learning_rate": 2.974198404141521e-05,
+      "loss": 0.8219,
+      "step": 1161
+    },
+    {
+      "epoch": 0.611578947368421,
+      "grad_norm": 1.6529840230941772,
+      "learning_rate": 2.9741524831535147e-05,
+      "loss": 0.0603,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6121052631578947,
+      "grad_norm": 1.2058558464050293,
+      "learning_rate": 2.9741065216925002e-05,
+      "loss": 0.8499,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6126315789473684,
+      "grad_norm": 1.5454597473144531,
+      "learning_rate": 2.97406051975974e-05,
+      "loss": 0.7998,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6131578947368421,
+      "grad_norm": 1.1322457790374756,
+      "learning_rate": 2.9740144773564952e-05,
+      "loss": 1.1103,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6136842105263158,
+      "grad_norm": 1.2677356004714966,
+      "learning_rate": 2.9739683944840315e-05,
+      "loss": 0.7271,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6142105263157894,
+      "grad_norm": 29.934062957763672,
+      "learning_rate": 2.9739222711436137e-05,
+      "loss": 2.5771,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6147368421052631,
+      "grad_norm": 1.108261227607727,
+      "learning_rate": 2.973876107336508e-05,
+      "loss": 1.017,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6152631578947368,
+      "grad_norm": 4.538265705108643,
+      "learning_rate": 2.9738299030639818e-05,
+      "loss": 1.3243,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6157894736842106,
+      "grad_norm": 1.448798418045044,
+      "learning_rate": 2.9737836583273037e-05,
+      "loss": 1.1748,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6163157894736843,
+      "grad_norm": 1.834809422492981,
+      "learning_rate": 2.9737373731277432e-05,
+      "loss": 1.1335,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6168421052631579,
+      "grad_norm": 1.8159127235412598,
+      "learning_rate": 2.9736910474665714e-05,
+      "loss": 1.3667,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6173684210526316,
+      "grad_norm": 1.2512363195419312,
+      "learning_rate": 2.9736446813450603e-05,
+      "loss": 1.1666,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6178947368421053,
+      "grad_norm": 13.280810356140137,
+      "learning_rate": 2.9735982747644817e-05,
+      "loss": 1.3808,
+      "step": 1174
+    },
+    {
+      "epoch": 0.618421052631579,
+      "grad_norm": 10.25016975402832,
+      "learning_rate": 2.9735518277261113e-05,
+      "loss": 0.7421,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6189473684210526,
+      "grad_norm": 2.516862630844116,
+      "learning_rate": 2.9735053402312234e-05,
+      "loss": 1.9148,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6194736842105263,
+      "grad_norm": 1.1309010982513428,
+      "learning_rate": 2.973458812281095e-05,
+      "loss": 1.2431,
+      "step": 1177
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.400505542755127,
+      "learning_rate": 2.9734122438770023e-05,
+      "loss": 1.3809,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6205263157894737,
+      "grad_norm": 1.2216683626174927,
+      "learning_rate": 2.9733656350202248e-05,
+      "loss": 0.7622,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6210526315789474,
+      "grad_norm": 1.942233920097351,
+      "learning_rate": 2.973318985712042e-05,
+      "loss": 1.2894,
+      "step": 1180
+    },
+    {
+      "epoch": 0.621578947368421,
+      "grad_norm": 3.7196786403656006,
+      "learning_rate": 2.9732722959537345e-05,
+      "loss": 0.9568,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6221052631578947,
+      "grad_norm": 1.3381714820861816,
+      "learning_rate": 2.973225565746585e-05,
+      "loss": 1.1391,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6226315789473684,
+      "grad_norm": 3.379469871520996,
+      "learning_rate": 2.973178795091875e-05,
+      "loss": 1.4507,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6231578947368421,
+      "grad_norm": 4.584420680999756,
+      "learning_rate": 2.9731319839908895e-05,
+      "loss": 0.5189,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6236842105263158,
+      "grad_norm": 1.164870023727417,
+      "learning_rate": 2.9730851324449133e-05,
+      "loss": 1.2698,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6242105263157894,
+      "grad_norm": 2.1792705059051514,
+      "learning_rate": 2.9730382404552334e-05,
+      "loss": 0.9649,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6247368421052631,
+      "grad_norm": 10.801166534423828,
+      "learning_rate": 2.9729913080231364e-05,
+      "loss": 1.006,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6252631578947369,
+      "grad_norm": 6.744489669799805,
+      "learning_rate": 2.9729443351499118e-05,
+      "loss": 1.5334,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6257894736842106,
+      "grad_norm": 0.9813660979270935,
+      "learning_rate": 2.9728973218368477e-05,
+      "loss": 0.8807,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6263157894736842,
+      "grad_norm": 1.4452024698257446,
+      "learning_rate": 2.9728502680852368e-05,
+      "loss": 1.2235,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6268421052631579,
+      "grad_norm": 1.5050314664840698,
+      "learning_rate": 2.9728031738963695e-05,
+      "loss": 0.9459,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6273684210526316,
+      "grad_norm": 1.1202926635742188,
+      "learning_rate": 2.9727560392715395e-05,
+      "loss": 0.7596,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6278947368421053,
+      "grad_norm": 0.803450345993042,
+      "learning_rate": 2.9727088642120406e-05,
+      "loss": 0.3567,
+      "step": 1193
+    },
+    {
+      "epoch": 0.628421052631579,
+      "grad_norm": 1.9323444366455078,
+      "learning_rate": 2.9726616487191675e-05,
+      "loss": 1.1573,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6289473684210526,
+      "grad_norm": 1.0270202159881592,
+      "learning_rate": 2.9726143927942176e-05,
+      "loss": 1.09,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6294736842105263,
+      "grad_norm": 12.028121948242188,
+      "learning_rate": 2.9725670964384876e-05,
+      "loss": 0.8462,
+      "step": 1196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.259978771209717,
+      "learning_rate": 2.9725197596532768e-05,
+      "loss": 1.0018,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6305263157894737,
+      "grad_norm": 1.1468654870986938,
+      "learning_rate": 2.9724723824398838e-05,
+      "loss": 1.1135,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6310526315789474,
+      "grad_norm": 1.0924807786941528,
+      "learning_rate": 2.9724249647996095e-05,
+      "loss": 1.2861,
+      "step": 1199
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.762115478515625,
+      "learning_rate": 2.972377506733756e-05,
+      "loss": 1.2612,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6321052631578947,
+      "grad_norm": 1.1973563432693481,
+      "learning_rate": 2.9723300082436266e-05,
+      "loss": 1.042,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6326315789473684,
+      "grad_norm": 1.2056132555007935,
+      "learning_rate": 2.972282469330525e-05,
+      "loss": 0.8789,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6331578947368421,
+      "grad_norm": 0.9773816466331482,
+      "learning_rate": 2.9722348899957564e-05,
+      "loss": 0.9174,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6336842105263157,
+      "grad_norm": 2.236300468444824,
+      "learning_rate": 2.972187270240627e-05,
+      "loss": 1.404,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6342105263157894,
+      "grad_norm": 1.2097399234771729,
+      "learning_rate": 2.9721396100664447e-05,
+      "loss": 1.2159,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6347368421052632,
+      "grad_norm": 1.800818681716919,
+      "learning_rate": 2.9720919094745176e-05,
+      "loss": 0.8479,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6352631578947369,
+      "grad_norm": 2.954768180847168,
+      "learning_rate": 2.9720441684661552e-05,
+      "loss": 1.3383,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6357894736842106,
+      "grad_norm": 2.48077130317688,
+      "learning_rate": 2.9719963870426685e-05,
+      "loss": 1.2594,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6363157894736842,
+      "grad_norm": 4.164140701293945,
+      "learning_rate": 2.9719485652053696e-05,
+      "loss": 1.3007,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6368421052631579,
+      "grad_norm": 4.807854175567627,
+      "learning_rate": 2.971900702955571e-05,
+      "loss": 0.2724,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6373684210526316,
+      "grad_norm": 2.658689498901367,
+      "learning_rate": 2.971852800294587e-05,
+      "loss": 1.4455,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6378947368421053,
+      "grad_norm": 5.018266677856445,
+      "learning_rate": 2.9718048572237323e-05,
+      "loss": 1.5204,
+      "step": 1212
+    },
+    {
+      "epoch": 0.638421052631579,
+      "grad_norm": 7.115269184112549,
+      "learning_rate": 2.9717568737443236e-05,
+      "loss": 1.763,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6389473684210526,
+      "grad_norm": 1.1763184070587158,
+      "learning_rate": 2.9717088498576787e-05,
+      "loss": 0.7555,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6394736842105263,
+      "grad_norm": 0.8004067540168762,
+      "learning_rate": 2.9716607855651154e-05,
+      "loss": 0.5879,
+      "step": 1215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.9092245697975159,
+      "learning_rate": 2.971612680867953e-05,
+      "loss": 0.9113,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6405263157894737,
+      "grad_norm": 1.03469717502594,
+      "learning_rate": 2.9715645357675133e-05,
+      "loss": 0.8598,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6410526315789473,
+      "grad_norm": 13.371722221374512,
+      "learning_rate": 2.9715163502651176e-05,
+      "loss": 0.6834,
+      "step": 1218
+    },
+    {
+      "epoch": 0.641578947368421,
+      "grad_norm": 1.745549201965332,
+      "learning_rate": 2.9714681243620885e-05,
+      "loss": 1.4798,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6421052631578947,
+      "grad_norm": 2.9852797985076904,
+      "learning_rate": 2.971419858059751e-05,
+      "loss": 1.6255,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6426315789473684,
+      "grad_norm": 1.1600061655044556,
+      "learning_rate": 2.971371551359429e-05,
+      "loss": 0.9992,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6431578947368422,
+      "grad_norm": 1.1885265111923218,
+      "learning_rate": 2.97132320426245e-05,
+      "loss": 0.9063,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6436842105263157,
+      "grad_norm": 1.2887687683105469,
+      "learning_rate": 2.9712748167701406e-05,
+      "loss": 1.0904,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6442105263157895,
+      "grad_norm": 3.543919324874878,
+      "learning_rate": 2.9712263888838295e-05,
+      "loss": 0.6287,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6447368421052632,
+      "grad_norm": 1.4697569608688354,
+      "learning_rate": 2.9711779206048457e-05,
+      "loss": 1.6434,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6452631578947369,
+      "grad_norm": 1.5276075601577759,
+      "learning_rate": 2.9711294119345212e-05,
+      "loss": 0.9159,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6457894736842106,
+      "grad_norm": 1.8441684246063232,
+      "learning_rate": 2.9710808628741866e-05,
+      "loss": 1.2164,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6463157894736842,
+      "grad_norm": 5.690790176391602,
+      "learning_rate": 2.9710322734251753e-05,
+      "loss": 0.818,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6468421052631579,
+      "grad_norm": 2.064948081970215,
+      "learning_rate": 2.9709836435888213e-05,
+      "loss": 1.5381,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6473684210526316,
+      "grad_norm": 1.3582369089126587,
+      "learning_rate": 2.9709349733664602e-05,
+      "loss": 0.4742,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6478947368421053,
+      "grad_norm": 1.6712125539779663,
+      "learning_rate": 2.9708862627594275e-05,
+      "loss": 0.9667,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6484210526315789,
+      "grad_norm": 5.520822048187256,
+      "learning_rate": 2.970837511769061e-05,
+      "loss": 0.7173,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6489473684210526,
+      "grad_norm": 3.7994682788848877,
+      "learning_rate": 2.9707887203966986e-05,
+      "loss": 0.8053,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6494736842105263,
+      "grad_norm": 1.3790806531906128,
+      "learning_rate": 2.970739888643681e-05,
+      "loss": 1.2264,
+      "step": 1234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.789325475692749,
+      "learning_rate": 2.9706910165113477e-05,
+      "loss": 0.6273,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6505263157894737,
+      "grad_norm": 2.6303861141204834,
+      "learning_rate": 2.970642104001041e-05,
+      "loss": 0.0705,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6510526315789473,
+      "grad_norm": 6.236796855926514,
+      "learning_rate": 2.9705931511141037e-05,
+      "loss": 0.147,
+      "step": 1237
+    },
+    {
+      "epoch": 0.651578947368421,
+      "grad_norm": 7.691266059875488,
+      "learning_rate": 2.9705441578518798e-05,
+      "loss": 0.6646,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6521052631578947,
+      "grad_norm": 1.5175530910491943,
+      "learning_rate": 2.9704951242157145e-05,
+      "loss": 1.33,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6526315789473685,
+      "grad_norm": 1.3474842309951782,
+      "learning_rate": 2.9704460502069544e-05,
+      "loss": 0.8195,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6531578947368422,
+      "grad_norm": 4.029512882232666,
+      "learning_rate": 2.9703969358269462e-05,
+      "loss": 0.3568,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6536842105263158,
+      "grad_norm": 1.3466565608978271,
+      "learning_rate": 2.9703477810770384e-05,
+      "loss": 1.259,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6542105263157895,
+      "grad_norm": 2.5386829376220703,
+      "learning_rate": 2.970298585958581e-05,
+      "loss": 1.0745,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6547368421052632,
+      "grad_norm": 2.428055763244629,
+      "learning_rate": 2.9702493504729244e-05,
+      "loss": 0.8593,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6552631578947369,
+      "grad_norm": 1.3094781637191772,
+      "learning_rate": 2.97020007462142e-05,
+      "loss": 0.8212,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6557894736842105,
+      "grad_norm": 1.4779995679855347,
+      "learning_rate": 2.970150758405421e-05,
+      "loss": 1.0186,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6563157894736842,
+      "grad_norm": 1.1934393644332886,
+      "learning_rate": 2.9701014018262817e-05,
+      "loss": 1.1701,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6568421052631579,
+      "grad_norm": 0.7196000218391418,
+      "learning_rate": 2.9700520048853566e-05,
+      "loss": 0.0174,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6573684210526316,
+      "grad_norm": 0.9140834808349609,
+      "learning_rate": 2.9700025675840028e-05,
+      "loss": 0.9861,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 4.055169582366943,
+      "learning_rate": 2.969953089923577e-05,
+      "loss": 0.4162,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6584210526315789,
+      "grad_norm": 0.5134720802307129,
+      "learning_rate": 2.9699035719054367e-05,
+      "loss": 0.0143,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6589473684210526,
+      "grad_norm": 3.9289088249206543,
+      "learning_rate": 2.9698540135309434e-05,
+      "loss": 0.5238,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6594736842105263,
+      "grad_norm": 6.439306735992432,
+      "learning_rate": 2.969804414801456e-05,
+      "loss": 0.549,
+      "step": 1253
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.025859832763672,
+      "learning_rate": 2.969754775718337e-05,
+      "loss": 0.9216,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6605263157894737,
+      "grad_norm": 1.8443411588668823,
+      "learning_rate": 2.969705096282949e-05,
+      "loss": 0.1711,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6610526315789473,
+      "grad_norm": 1.5247849225997925,
+      "learning_rate": 2.9696553764966562e-05,
+      "loss": 1.1315,
+      "step": 1256
+    },
+    {
+      "epoch": 0.661578947368421,
+      "grad_norm": 3.6713950634002686,
+      "learning_rate": 2.9696056163608237e-05,
+      "loss": 0.467,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6621052631578948,
+      "grad_norm": 0.790451169013977,
+      "learning_rate": 2.969555815876818e-05,
+      "loss": 0.0082,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6626315789473685,
+      "grad_norm": 1.4589107036590576,
+      "learning_rate": 2.9695059750460052e-05,
+      "loss": 0.6006,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6631578947368421,
+      "grad_norm": 1.3655271530151367,
+      "learning_rate": 2.9694560938697548e-05,
+      "loss": 1.2306,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6636842105263158,
+      "grad_norm": 1.3729549646377563,
+      "learning_rate": 2.9694061723494358e-05,
+      "loss": 0.7385,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6642105263157895,
+      "grad_norm": 4.8193864822387695,
+      "learning_rate": 2.9693562104864182e-05,
+      "loss": 0.7079,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6647368421052632,
+      "grad_norm": 1.563333511352539,
+      "learning_rate": 2.9693062082820752e-05,
+      "loss": 0.5117,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6652631578947369,
+      "grad_norm": 1.3271199464797974,
+      "learning_rate": 2.969256165737779e-05,
+      "loss": 1.0954,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6657894736842105,
+      "grad_norm": 4.284012794494629,
+      "learning_rate": 2.9692060828549025e-05,
+      "loss": 1.9864,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6663157894736842,
+      "grad_norm": 3.3707847595214844,
+      "learning_rate": 2.969155959634822e-05,
+      "loss": 1.0695,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6668421052631579,
+      "grad_norm": 2.1350784301757812,
+      "learning_rate": 2.9691057960789133e-05,
+      "loss": 1.2387,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6673684210526316,
+      "grad_norm": 1.350180745124817,
+      "learning_rate": 2.9690555921885534e-05,
+      "loss": 1.1326,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6678947368421052,
+      "grad_norm": 1.3490583896636963,
+      "learning_rate": 2.9690053479651207e-05,
+      "loss": 0.7916,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6684210526315789,
+      "grad_norm": 5.74185848236084,
+      "learning_rate": 2.968955063409995e-05,
+      "loss": 0.9312,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6689473684210526,
+      "grad_norm": 1.1950428485870361,
+      "learning_rate": 2.968904738524557e-05,
+      "loss": 0.8421,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6694736842105263,
+      "grad_norm": 1.2086502313613892,
+      "learning_rate": 2.9688543733101876e-05,
+      "loss": 1.205,
+      "step": 1272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.652189254760742,
+      "learning_rate": 2.96880396776827e-05,
+      "loss": 1.0971,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6705263157894736,
+      "grad_norm": 1.5372319221496582,
+      "learning_rate": 2.968753521900188e-05,
+      "loss": 1.4361,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6710526315789473,
+      "grad_norm": 3.254009485244751,
+      "learning_rate": 2.9687030357073265e-05,
+      "loss": 1.5015,
+      "step": 1275
+    },
+    {
+      "epoch": 0.671578947368421,
+      "grad_norm": 1.3365062475204468,
+      "learning_rate": 2.968652509191072e-05,
+      "loss": 1.2392,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6721052631578948,
+      "grad_norm": 8.001225471496582,
+      "learning_rate": 2.9686019423528117e-05,
+      "loss": 1.1131,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6726315789473685,
+      "grad_norm": 4.288140296936035,
+      "learning_rate": 2.9685513351939335e-05,
+      "loss": 1.2558,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6731578947368421,
+      "grad_norm": 0.9908925294876099,
+      "learning_rate": 2.968500687715827e-05,
+      "loss": 1.0251,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6736842105263158,
+      "grad_norm": 1.2662602663040161,
+      "learning_rate": 2.968449999919883e-05,
+      "loss": 1.0004,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6742105263157895,
+      "grad_norm": 1.4708739519119263,
+      "learning_rate": 2.9683992718074926e-05,
+      "loss": 1.3322,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6747368421052632,
+      "grad_norm": 1.1707433462142944,
+      "learning_rate": 2.9683485033800494e-05,
+      "loss": 1.3583,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6752631578947368,
+      "grad_norm": 0.060910388827323914,
+      "learning_rate": 2.9682976946389463e-05,
+      "loss": 0.0018,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6757894736842105,
+      "grad_norm": 1.6072977781295776,
+      "learning_rate": 2.9682468455855783e-05,
+      "loss": 0.0838,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6763157894736842,
+      "grad_norm": 1.179344892501831,
+      "learning_rate": 2.9681959562213422e-05,
+      "loss": 1.3671,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6768421052631579,
+      "grad_norm": 6.977228164672852,
+      "learning_rate": 2.9681450265476345e-05,
+      "loss": 0.5456,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6773684210526316,
+      "grad_norm": 1.475023865699768,
+      "learning_rate": 2.968094056565854e-05,
+      "loss": 1.0063,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6778947368421052,
+      "grad_norm": 2.039761781692505,
+      "learning_rate": 2.9680430462774e-05,
+      "loss": 0.3243,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6784210526315789,
+      "grad_norm": 1.093194603919983,
+      "learning_rate": 2.967991995683673e-05,
+      "loss": 1.2607,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6789473684210526,
+      "grad_norm": 24.999378204345703,
+      "learning_rate": 2.9679409047860743e-05,
+      "loss": 1.5342,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6794736842105263,
+      "grad_norm": 1.6621330976486206,
+      "learning_rate": 2.9678897735860066e-05,
+      "loss": 1.8376,
+      "step": 1291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.57814621925354,
+      "learning_rate": 2.9678386020848742e-05,
+      "loss": 1.4534,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6805263157894736,
+      "grad_norm": 1.6566349267959595,
+      "learning_rate": 2.9677873902840813e-05,
+      "loss": 0.049,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6810526315789474,
+      "grad_norm": 1.090785264968872,
+      "learning_rate": 2.9677361381850345e-05,
+      "loss": 1.0962,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6815789473684211,
+      "grad_norm": 1.6356092691421509,
+      "learning_rate": 2.9676848457891407e-05,
+      "loss": 0.6644,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6821052631578948,
+      "grad_norm": 1.247987151145935,
+      "learning_rate": 2.9676335130978082e-05,
+      "loss": 0.8771,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6826315789473684,
+      "grad_norm": 1.4328747987747192,
+      "learning_rate": 2.9675821401124465e-05,
+      "loss": 1.0978,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6831578947368421,
+      "grad_norm": 6.744469165802002,
+      "learning_rate": 2.967530726834466e-05,
+      "loss": 0.8294,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6836842105263158,
+      "grad_norm": 1.233546495437622,
+      "learning_rate": 2.9674792732652773e-05,
+      "loss": 0.9119,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 1.5304890871047974,
+      "learning_rate": 2.967427779406295e-05,
+      "loss": 1.2472,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6847368421052632,
+      "grad_norm": 1.6759600639343262,
+      "learning_rate": 2.9673762452589307e-05,
+      "loss": 1.0797,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6852631578947368,
+      "grad_norm": 2.450378894805908,
+      "learning_rate": 2.967324670824601e-05,
+      "loss": 0.7401,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6857894736842105,
+      "grad_norm": 0.631327748298645,
+      "learning_rate": 2.9672730561047214e-05,
+      "loss": 0.0149,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6863157894736842,
+      "grad_norm": 2.2037277221679688,
+      "learning_rate": 2.9672214011007087e-05,
+      "loss": 0.9195,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6868421052631579,
+      "grad_norm": 20.760221481323242,
+      "learning_rate": 2.967169705813981e-05,
+      "loss": 2.7392,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6873684210526316,
+      "grad_norm": 1.0082573890686035,
+      "learning_rate": 2.9671179702459576e-05,
+      "loss": 1.0737,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6878947368421052,
+      "grad_norm": 2.9095327854156494,
+      "learning_rate": 2.9670661943980595e-05,
+      "loss": 1.4494,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6884210526315789,
+      "grad_norm": 0.9664977788925171,
+      "learning_rate": 2.9670143782717075e-05,
+      "loss": 1.1525,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6889473684210526,
+      "grad_norm": 0.6551389098167419,
+      "learning_rate": 2.966962521868325e-05,
+      "loss": 0.0284,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6894736842105263,
+      "grad_norm": 6.767817497253418,
+      "learning_rate": 2.966910625189335e-05,
+      "loss": 0.253,
+      "step": 1310
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.445407867431641,
+      "learning_rate": 2.9668586882361625e-05,
+      "loss": 1.5918,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6905263157894737,
+      "grad_norm": 3.130711078643799,
+      "learning_rate": 2.9668067110102338e-05,
+      "loss": 1.0437,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6910526315789474,
+      "grad_norm": 1.7217767238616943,
+      "learning_rate": 2.9667546935129757e-05,
+      "loss": 0.5301,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6915789473684211,
+      "grad_norm": 6.565402984619141,
+      "learning_rate": 2.966702635745816e-05,
+      "loss": 1.4053,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6921052631578948,
+      "grad_norm": 1.4495912790298462,
+      "learning_rate": 2.9666505377101845e-05,
+      "loss": 0.8034,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6926315789473684,
+      "grad_norm": 1.2847636938095093,
+      "learning_rate": 2.9665983994075113e-05,
+      "loss": 1.0825,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6931578947368421,
+      "grad_norm": 1.0292222499847412,
+      "learning_rate": 2.966546220839228e-05,
+      "loss": 1.1681,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6936842105263158,
+      "grad_norm": 1.537844181060791,
+      "learning_rate": 2.966494002006767e-05,
+      "loss": 1.6044,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6942105263157895,
+      "grad_norm": 4.573411464691162,
+      "learning_rate": 2.966441742911562e-05,
+      "loss": 0.7895,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6947368421052632,
+      "grad_norm": 1.3798898458480835,
+      "learning_rate": 2.9663894435550477e-05,
+      "loss": 0.5207,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6952631578947368,
+      "grad_norm": 1.8828356266021729,
+      "learning_rate": 2.96633710393866e-05,
+      "loss": 0.4363,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6957894736842105,
+      "grad_norm": 4.514081954956055,
+      "learning_rate": 2.966284724063836e-05,
+      "loss": 0.9505,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6963157894736842,
+      "grad_norm": 6.411378860473633,
+      "learning_rate": 2.966232303932014e-05,
+      "loss": 0.2739,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6968421052631579,
+      "grad_norm": 3.237457036972046,
+      "learning_rate": 2.9661798435446325e-05,
+      "loss": 0.2196,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6973684210526315,
+      "grad_norm": 2.3053643703460693,
+      "learning_rate": 2.966127342903133e-05,
+      "loss": 1.4729,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6978947368421052,
+      "grad_norm": 1.2993172407150269,
+      "learning_rate": 2.9660748020089555e-05,
+      "loss": 0.8165,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6984210526315789,
+      "grad_norm": 1.4997568130493164,
+      "learning_rate": 2.9660222208635438e-05,
+      "loss": 1.0576,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6989473684210527,
+      "grad_norm": 6.542171001434326,
+      "learning_rate": 2.9659695994683404e-05,
+      "loss": 0.7472,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6994736842105264,
+      "grad_norm": 2.04111647605896,
+      "learning_rate": 2.965916937824791e-05,
+      "loss": 1.0342,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6200339794158936,
+      "learning_rate": 2.965864235934341e-05,
+      "loss": 0.9009,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7005263157894737,
+      "grad_norm": 3.2301101684570312,
+      "learning_rate": 2.965811493798437e-05,
+      "loss": 1.1608,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7010526315789474,
+      "grad_norm": 13.88673210144043,
+      "learning_rate": 2.9657587114185272e-05,
+      "loss": 2.5288,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7015789473684211,
+      "grad_norm": 3.9736039638519287,
+      "learning_rate": 2.9657058887960613e-05,
+      "loss": 1.0553,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7021052631578948,
+      "grad_norm": 24.85307502746582,
+      "learning_rate": 2.965653025932489e-05,
+      "loss": 0.916,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7026315789473684,
+      "grad_norm": 4.859387397766113,
+      "learning_rate": 2.965600122829262e-05,
+      "loss": 0.9876,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7031578947368421,
+      "grad_norm": 2.863323211669922,
+      "learning_rate": 2.9655471794878325e-05,
+      "loss": 0.8016,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7036842105263158,
+      "grad_norm": 2.4096951484680176,
+      "learning_rate": 2.9654941959096543e-05,
+      "loss": 0.6945,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7042105263157895,
+      "grad_norm": 1.333808183670044,
+      "learning_rate": 2.9654411720961816e-05,
+      "loss": 1.1654,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7047368421052631,
+      "grad_norm": 6.172205448150635,
+      "learning_rate": 2.9653881080488705e-05,
+      "loss": 0.5594,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7052631578947368,
+      "grad_norm": 5.650049686431885,
+      "learning_rate": 2.9653350037691777e-05,
+      "loss": 0.8009,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7057894736842105,
+      "grad_norm": 1.504169225692749,
+      "learning_rate": 2.9652818592585616e-05,
+      "loss": 1.1546,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7063157894736842,
+      "grad_norm": 2.2842626571655273,
+      "learning_rate": 2.965228674518481e-05,
+      "loss": 0.9654,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7068421052631579,
+      "grad_norm": 1.429457187652588,
+      "learning_rate": 2.965175449550396e-05,
+      "loss": 1.3076,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7073684210526315,
+      "grad_norm": 1.1388136148452759,
+      "learning_rate": 2.9651221843557682e-05,
+      "loss": 0.9347,
+      "step": 1344
+    },
+    {
+      "epoch": 0.7078947368421052,
+      "grad_norm": 2.501084566116333,
+      "learning_rate": 2.96506887893606e-05,
+      "loss": 1.11,
+      "step": 1345
+    },
+    {
+      "epoch": 0.708421052631579,
+      "grad_norm": 27.422794342041016,
+      "learning_rate": 2.9650155332927343e-05,
+      "loss": 7.5874,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7089473684210527,
+      "grad_norm": 5.256887912750244,
+      "learning_rate": 2.9649621474272564e-05,
+      "loss": 1.4157,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7094736842105264,
+      "grad_norm": 1.0665907859802246,
+      "learning_rate": 2.9649087213410918e-05,
+      "loss": 0.8372,
+      "step": 1348
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.498046398162842,
+      "learning_rate": 2.964855255035707e-05,
+      "loss": 1.2181,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 2.614759683609009,
+      "learning_rate": 2.9648017485125708e-05,
+      "loss": 0.1237,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7110526315789474,
+      "grad_norm": 1.4356591701507568,
+      "learning_rate": 2.9647482017731508e-05,
+      "loss": 1.2685,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7115789473684211,
+      "grad_norm": 1.245197057723999,
+      "learning_rate": 2.964694614818918e-05,
+      "loss": 1.1176,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7121052631578947,
+      "grad_norm": 1.7546439170837402,
+      "learning_rate": 2.9646409876513444e-05,
+      "loss": 1.8027,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7126315789473684,
+      "grad_norm": 6.851596355438232,
+      "learning_rate": 2.9645873202719013e-05,
+      "loss": 0.9774,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7131578947368421,
+      "grad_norm": 2.6459946632385254,
+      "learning_rate": 2.9645336126820616e-05,
+      "loss": 1.5187,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7136842105263158,
+      "grad_norm": 11.114506721496582,
+      "learning_rate": 2.9644798648833013e-05,
+      "loss": 1.5982,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7142105263157895,
+      "grad_norm": 1.1266565322875977,
+      "learning_rate": 2.9644260768770953e-05,
+      "loss": 1.043,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7147368421052631,
+      "grad_norm": 1.3332453966140747,
+      "learning_rate": 2.9643722486649203e-05,
+      "loss": 0.9823,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7152631578947368,
+      "grad_norm": 7.991135120391846,
+      "learning_rate": 2.9643183802482542e-05,
+      "loss": 1.8691,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7157894736842105,
+      "grad_norm": 1.4576728343963623,
+      "learning_rate": 2.9642644716285765e-05,
+      "loss": 1.2554,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7163157894736842,
+      "grad_norm": 24.029621124267578,
+      "learning_rate": 2.9642105228073662e-05,
+      "loss": 1.4993,
+      "step": 1361
+    },
+    {
+      "epoch": 0.716842105263158,
+      "grad_norm": 1.7569326162338257,
+      "learning_rate": 2.9641565337861055e-05,
+      "loss": 0.6511,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7173684210526315,
+      "grad_norm": 0.8566625118255615,
+      "learning_rate": 2.9641025045662765e-05,
+      "loss": 0.0214,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7178947368421053,
+      "grad_norm": 1.8023531436920166,
+      "learning_rate": 2.9640484351493616e-05,
+      "loss": 1.3085,
+      "step": 1364
+    },
+    {
+      "epoch": 0.718421052631579,
+      "grad_norm": 1.1481828689575195,
+      "learning_rate": 2.9639943255368468e-05,
+      "loss": 1.3198,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7189473684210527,
+      "grad_norm": 2.6472365856170654,
+      "learning_rate": 2.9639401757302162e-05,
+      "loss": 0.2092,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7194736842105263,
+      "grad_norm": 1.8456323146820068,
+      "learning_rate": 2.9638859857309574e-05,
+      "loss": 1.0272,
+      "step": 1367
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3874828815460205,
+      "learning_rate": 2.9638317555405582e-05,
+      "loss": 0.9814,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7205263157894737,
+      "grad_norm": 1.234138011932373,
+      "learning_rate": 2.9637774851605073e-05,
+      "loss": 0.8918,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7210526315789474,
+      "grad_norm": 2.1231067180633545,
+      "learning_rate": 2.9637231745922942e-05,
+      "loss": 0.8428,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7215789473684211,
+      "grad_norm": 2.287672281265259,
+      "learning_rate": 2.963668823837411e-05,
+      "loss": 1.5556,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7221052631578947,
+      "grad_norm": 1.185556173324585,
+      "learning_rate": 2.9636144328973495e-05,
+      "loss": 0.8633,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7226315789473684,
+      "grad_norm": 5.523141860961914,
+      "learning_rate": 2.9635600017736024e-05,
+      "loss": 0.7142,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7231578947368421,
+      "grad_norm": 12.063275337219238,
+      "learning_rate": 2.9635055304676647e-05,
+      "loss": 0.3757,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7236842105263158,
+      "grad_norm": 4.580484867095947,
+      "learning_rate": 2.963451018981032e-05,
+      "loss": 1.5171,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7242105263157895,
+      "grad_norm": 1.9087570905685425,
+      "learning_rate": 2.9633964673152004e-05,
+      "loss": 1.3982,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7247368421052631,
+      "grad_norm": 2.3344743251800537,
+      "learning_rate": 2.9633418754716682e-05,
+      "loss": 1.0389,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7252631578947368,
+      "grad_norm": 4.389303207397461,
+      "learning_rate": 2.9632872434519342e-05,
+      "loss": 0.3592,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7257894736842105,
+      "grad_norm": 4.647209644317627,
+      "learning_rate": 2.963232571257498e-05,
+      "loss": 2.2252,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7263157894736842,
+      "grad_norm": 2.98408579826355,
+      "learning_rate": 2.9631778588898606e-05,
+      "loss": 1.3005,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7268421052631578,
+      "grad_norm": 10.173641204833984,
+      "learning_rate": 2.9631231063505245e-05,
+      "loss": 0.4029,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7273684210526316,
+      "grad_norm": 2.366199254989624,
+      "learning_rate": 2.963068313640992e-05,
+      "loss": 1.408,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7278947368421053,
+      "grad_norm": 0.9890458583831787,
+      "learning_rate": 2.963013480762769e-05,
+      "loss": 0.941,
+      "step": 1383
+    },
+    {
+      "epoch": 0.728421052631579,
+      "grad_norm": 2.4400086402893066,
+      "learning_rate": 2.96295860771736e-05,
+      "loss": 0.303,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7289473684210527,
+      "grad_norm": 7.326104640960693,
+      "learning_rate": 2.9629036945062715e-05,
+      "loss": 0.1054,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7294736842105263,
+      "grad_norm": 0.08032506704330444,
+      "learning_rate": 2.9628487411310113e-05,
+      "loss": 0.0038,
+      "step": 1386
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.0595088005065918,
+      "learning_rate": 2.962793747593088e-05,
+      "loss": 1.1041,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7305263157894737,
+      "grad_norm": 0.06474526226520538,
+      "learning_rate": 2.9627387138940117e-05,
+      "loss": 0.0022,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7310526315789474,
+      "grad_norm": 1.7989740371704102,
+      "learning_rate": 2.9626836400352932e-05,
+      "loss": 0.8054,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7315789473684211,
+      "grad_norm": 0.8924505710601807,
+      "learning_rate": 2.962628526018445e-05,
+      "loss": 1.0706,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7321052631578947,
+      "grad_norm": 1.269217848777771,
+      "learning_rate": 2.9625733718449792e-05,
+      "loss": 0.7377,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7326315789473684,
+      "grad_norm": 2.0392255783081055,
+      "learning_rate": 2.962518177516411e-05,
+      "loss": 0.5988,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7331578947368421,
+      "grad_norm": 9.548563003540039,
+      "learning_rate": 2.9624629430342557e-05,
+      "loss": 0.437,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7336842105263158,
+      "grad_norm": 2.005530595779419,
+      "learning_rate": 2.9624076684000292e-05,
+      "loss": 1.4294,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7342105263157894,
+      "grad_norm": 8.551248550415039,
+      "learning_rate": 2.96235235361525e-05,
+      "loss": 1.0252,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7347368421052631,
+      "grad_norm": 2.2077114582061768,
+      "learning_rate": 2.962296998681436e-05,
+      "loss": 0.749,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7352631578947368,
+      "grad_norm": 1.9976532459259033,
+      "learning_rate": 2.962241603600107e-05,
+      "loss": 0.6527,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7357894736842105,
+      "grad_norm": 1.972212314605713,
+      "learning_rate": 2.962186168372784e-05,
+      "loss": 1.1974,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7363157894736843,
+      "grad_norm": 0.9594762921333313,
+      "learning_rate": 2.962130693000989e-05,
+      "loss": 1.5027,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 1.018154501914978,
+      "learning_rate": 2.9620751774862456e-05,
+      "loss": 1.1123,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7373684210526316,
+      "grad_norm": 1.4608852863311768,
+      "learning_rate": 2.962019621830077e-05,
+      "loss": 1.2949,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7378947368421053,
+      "grad_norm": 19.72736167907715,
+      "learning_rate": 2.9619640260340092e-05,
+      "loss": 1.1407,
+      "step": 1402
+    },
+    {
+      "epoch": 0.738421052631579,
+      "grad_norm": 2.0531411170959473,
+      "learning_rate": 2.9619083900995684e-05,
+      "loss": 1.2994,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7389473684210527,
+      "grad_norm": 1.4088972806930542,
+      "learning_rate": 2.961852714028282e-05,
+      "loss": 1.0896,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7394736842105263,
+      "grad_norm": 0.31528720259666443,
+      "learning_rate": 2.961796997821679e-05,
+      "loss": 0.0114,
+      "step": 1405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.371922969818115,
+      "learning_rate": 2.9617412414812883e-05,
+      "loss": 0.6344,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7405263157894737,
+      "grad_norm": 2.7792177200317383,
+      "learning_rate": 2.9616854450086415e-05,
+      "loss": 0.0651,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7410526315789474,
+      "grad_norm": 1.3516690731048584,
+      "learning_rate": 2.9616296084052698e-05,
+      "loss": 0.9898,
+      "step": 1408
+    },
+    {
+      "epoch": 0.741578947368421,
+      "grad_norm": 5.8592448234558105,
+      "learning_rate": 2.961573731672707e-05,
+      "loss": 1.8225,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7421052631578947,
+      "grad_norm": 1.515937089920044,
+      "learning_rate": 2.9615178148124867e-05,
+      "loss": 0.8826,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7426315789473684,
+      "grad_norm": 1.1607184410095215,
+      "learning_rate": 2.9614618578261436e-05,
+      "loss": 0.9627,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7431578947368421,
+      "grad_norm": 1.7677443027496338,
+      "learning_rate": 2.9614058607152153e-05,
+      "loss": 1.2879,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7436842105263158,
+      "grad_norm": 1.718802809715271,
+      "learning_rate": 2.9613498234812378e-05,
+      "loss": 0.9766,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7442105263157894,
+      "grad_norm": 1.4020949602127075,
+      "learning_rate": 2.9612937461257504e-05,
+      "loss": 1.319,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7447368421052631,
+      "grad_norm": 1.2930610179901123,
+      "learning_rate": 2.9612376286502934e-05,
+      "loss": 1.0887,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7452631578947368,
+      "grad_norm": 1.6021417379379272,
+      "learning_rate": 2.961181471056406e-05,
+      "loss": 0.861,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7457894736842106,
+      "grad_norm": 1.3957011699676514,
+      "learning_rate": 2.9611252733456306e-05,
+      "loss": 0.8134,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7463157894736843,
+      "grad_norm": 2.1977591514587402,
+      "learning_rate": 2.9610690355195108e-05,
+      "loss": 1.5508,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7468421052631579,
+      "grad_norm": 2.2483537197113037,
+      "learning_rate": 2.9610127575795894e-05,
+      "loss": 0.7797,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7473684210526316,
+      "grad_norm": 1.5289298295974731,
+      "learning_rate": 2.9609564395274125e-05,
+      "loss": 1.4425,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7478947368421053,
+      "grad_norm": 1.3084803819656372,
+      "learning_rate": 2.9609000813645257e-05,
+      "loss": 1.117,
+      "step": 1421
+    },
+    {
+      "epoch": 0.748421052631579,
+      "grad_norm": 2.422757863998413,
+      "learning_rate": 2.960843683092477e-05,
+      "loss": 1.7103,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7489473684210526,
+      "grad_norm": 1.4796887636184692,
+      "learning_rate": 2.9607872447128142e-05,
+      "loss": 1.0487,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7494736842105263,
+      "grad_norm": 1.0981144905090332,
+      "learning_rate": 2.960730766227087e-05,
+      "loss": 0.5913,
+      "step": 1424
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8127541542053223,
+      "learning_rate": 2.9606742476368464e-05,
+      "loss": 0.8062,
+      "step": 1425
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.0191138982772827,
+      "eval_runtime": 12.9816,
+      "eval_samples_per_second": 7.703,
+      "eval_steps_per_second": 7.703,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7505263157894737,
+      "grad_norm": 2.0109336376190186,
+      "learning_rate": 2.9606176889436435e-05,
+      "loss": 1.5174,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7510526315789474,
+      "grad_norm": 1.7353535890579224,
+      "learning_rate": 2.9605610901490312e-05,
+      "loss": 1.0043,
+      "step": 1427
+    },
+    {
+      "epoch": 0.751578947368421,
+      "grad_norm": 2.5789828300476074,
+      "learning_rate": 2.960504451254564e-05,
+      "loss": 1.056,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7521052631578947,
+      "grad_norm": 10.612245559692383,
+      "learning_rate": 2.960447772261796e-05,
+      "loss": 1.3747,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7526315789473684,
+      "grad_norm": 1.5246334075927734,
+      "learning_rate": 2.960391053172285e-05,
+      "loss": 1.2934,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7531578947368421,
+      "grad_norm": 1.2563579082489014,
+      "learning_rate": 2.9603342939875863e-05,
+      "loss": 1.1843,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7536842105263157,
+      "grad_norm": 2.347818613052368,
+      "learning_rate": 2.960277494709259e-05,
+      "loss": 0.9651,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7542105263157894,
+      "grad_norm": 1.2420134544372559,
+      "learning_rate": 2.960220655338863e-05,
+      "loss": 0.8367,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7547368421052632,
+      "grad_norm": 1.5213626623153687,
+      "learning_rate": 2.9601637758779577e-05,
+      "loss": 1.2834,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7552631578947369,
+      "grad_norm": 3.097379684448242,
+      "learning_rate": 2.9601068563281063e-05,
+      "loss": 1.1477,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7557894736842106,
+      "grad_norm": 1.0866178274154663,
+      "learning_rate": 2.9600498966908702e-05,
+      "loss": 1.1422,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7563157894736842,
+      "grad_norm": 1.3330771923065186,
+      "learning_rate": 2.959992896967814e-05,
+      "loss": 1.2408,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7568421052631579,
+      "grad_norm": 2.1840741634368896,
+      "learning_rate": 2.959935857160502e-05,
+      "loss": 0.7088,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7573684210526316,
+      "grad_norm": 1.8869200944900513,
+      "learning_rate": 2.9598787772705006e-05,
+      "loss": 1.0891,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7578947368421053,
+      "grad_norm": 1.4808701276779175,
+      "learning_rate": 2.959821657299377e-05,
+      "loss": 1.7629,
+      "step": 1440
+    },
+    {
+      "epoch": 0.758421052631579,
+      "grad_norm": 1.3917648792266846,
+      "learning_rate": 2.959764497248699e-05,
+      "loss": 1.3897,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7589473684210526,
+      "grad_norm": 1.1825001239776611,
+      "learning_rate": 2.9597072971200366e-05,
+      "loss": 1.1223,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7594736842105263,
+      "grad_norm": 1.76896071434021,
+      "learning_rate": 2.9596500569149603e-05,
+      "loss": 1.6206,
+      "step": 1443
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2272676229476929,
+      "learning_rate": 2.9595927766350406e-05,
+      "loss": 1.1395,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7605263157894737,
+      "grad_norm": 4.527194499969482,
+      "learning_rate": 2.959535456281851e-05,
+      "loss": 0.4346,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7610526315789473,
+      "grad_norm": 1.3327759504318237,
+      "learning_rate": 2.959478095856965e-05,
+      "loss": 0.9762,
+      "step": 1446
+    },
+    {
+      "epoch": 0.761578947368421,
+      "grad_norm": 2.397428512573242,
+      "learning_rate": 2.959420695361958e-05,
+      "loss": 0.7909,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7621052631578947,
+      "grad_norm": 1.3565646409988403,
+      "learning_rate": 2.9593632547984047e-05,
+      "loss": 0.7872,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7626315789473684,
+      "grad_norm": 0.9641793370246887,
+      "learning_rate": 2.9593057741678832e-05,
+      "loss": 0.8944,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 2.6557819843292236,
+      "learning_rate": 2.959248253471971e-05,
+      "loss": 0.8311,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7636842105263157,
+      "grad_norm": 4.478829860687256,
+      "learning_rate": 2.9591906927122477e-05,
+      "loss": 0.4445,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7642105263157895,
+      "grad_norm": 2.393716335296631,
+      "learning_rate": 2.9591330918902935e-05,
+      "loss": 1.7126,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7647368421052632,
+      "grad_norm": 1.553025722503662,
+      "learning_rate": 2.95907545100769e-05,
+      "loss": 1.1763,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7652631578947369,
+      "grad_norm": 1.0929442644119263,
+      "learning_rate": 2.9590177700660193e-05,
+      "loss": 1.14,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7657894736842106,
+      "grad_norm": 7.241302013397217,
+      "learning_rate": 2.9589600490668655e-05,
+      "loss": 1.111,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7663157894736842,
+      "grad_norm": 0.972644567489624,
+      "learning_rate": 2.9589022880118133e-05,
+      "loss": 1.1578,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7668421052631579,
+      "grad_norm": 7.614607810974121,
+      "learning_rate": 2.9588444869024484e-05,
+      "loss": 1.0112,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7673684210526316,
+      "grad_norm": 1.8197392225265503,
+      "learning_rate": 2.9587866457403577e-05,
+      "loss": 1.2878,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7678947368421053,
+      "grad_norm": 1.1428899765014648,
+      "learning_rate": 2.9587287645271293e-05,
+      "loss": 0.7216,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7684210526315789,
+      "grad_norm": 1.5418213605880737,
+      "learning_rate": 2.958670843264353e-05,
+      "loss": 0.9176,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7689473684210526,
+      "grad_norm": 6.577853679656982,
+      "learning_rate": 2.9586128819536172e-05,
+      "loss": 1.3075,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7694736842105263,
+      "grad_norm": 1.7671222686767578,
+      "learning_rate": 2.958554880596515e-05,
+      "loss": 1.0772,
+      "step": 1462
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2378417253494263,
+      "learning_rate": 2.9584968391946378e-05,
+      "loss": 1.048,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7705263157894737,
+      "grad_norm": 1.9807356595993042,
+      "learning_rate": 2.9584387577495803e-05,
+      "loss": 1.6231,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7710526315789473,
+      "grad_norm": 1.2167612314224243,
+      "learning_rate": 2.958380636262936e-05,
+      "loss": 1.5038,
+      "step": 1465
+    },
+    {
+      "epoch": 0.771578947368421,
+      "grad_norm": 1.2333711385726929,
+      "learning_rate": 2.9583224747363008e-05,
+      "loss": 0.6763,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7721052631578947,
+      "grad_norm": 1.9440076351165771,
+      "learning_rate": 2.958264273171272e-05,
+      "loss": 1.8108,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7726315789473684,
+      "grad_norm": 1.7781847715377808,
+      "learning_rate": 2.958206031569447e-05,
+      "loss": 1.5469,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7731578947368422,
+      "grad_norm": 1.5457514524459839,
+      "learning_rate": 2.9581477499324254e-05,
+      "loss": 1.1067,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7736842105263158,
+      "grad_norm": 6.0052103996276855,
+      "learning_rate": 2.9580894282618073e-05,
+      "loss": 0.5737,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7742105263157895,
+      "grad_norm": 1.827090859413147,
+      "learning_rate": 2.9580310665591933e-05,
+      "loss": 1.1473,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7747368421052632,
+      "grad_norm": 5.207357883453369,
+      "learning_rate": 2.9579726648261862e-05,
+      "loss": 2.123,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7752631578947369,
+      "grad_norm": 4.914318084716797,
+      "learning_rate": 2.957914223064389e-05,
+      "loss": 1.4473,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7757894736842105,
+      "grad_norm": 4.295709609985352,
+      "learning_rate": 2.9578557412754067e-05,
+      "loss": 0.1821,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7763157894736842,
+      "grad_norm": 14.05141830444336,
+      "learning_rate": 2.9577972194608453e-05,
+      "loss": 1.7767,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7768421052631579,
+      "grad_norm": 3.5937631130218506,
+      "learning_rate": 2.9577386576223105e-05,
+      "loss": 0.8665,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7773684210526316,
+      "grad_norm": 1.2602375745773315,
+      "learning_rate": 2.9576800557614103e-05,
+      "loss": 1.6036,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7778947368421053,
+      "grad_norm": 6.486451148986816,
+      "learning_rate": 2.9576214138797544e-05,
+      "loss": 0.8866,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7784210526315789,
+      "grad_norm": 2.9001505374908447,
+      "learning_rate": 2.9575627319789523e-05,
+      "loss": 1.085,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7789473684210526,
+      "grad_norm": 1.1088868379592896,
+      "learning_rate": 2.957504010060615e-05,
+      "loss": 1.0509,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7794736842105263,
+      "grad_norm": 2.7946763038635254,
+      "learning_rate": 2.9574452481263553e-05,
+      "loss": 0.1451,
+      "step": 1481
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2279927730560303,
+      "learning_rate": 2.9573864461777856e-05,
+      "loss": 0.9693,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7805263157894737,
+      "grad_norm": 9.112462997436523,
+      "learning_rate": 2.957327604216521e-05,
+      "loss": 0.4725,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7810526315789473,
+      "grad_norm": 4.834194183349609,
+      "learning_rate": 2.957268722244177e-05,
+      "loss": 0.7559,
+      "step": 1484
+    },
+    {
+      "epoch": 0.781578947368421,
+      "grad_norm": 1.3210901021957397,
+      "learning_rate": 2.9572098002623697e-05,
+      "loss": 0.9653,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7821052631578947,
+      "grad_norm": 6.407845973968506,
+      "learning_rate": 2.9571508382727173e-05,
+      "loss": 2.3573,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7826315789473685,
+      "grad_norm": 1.6584391593933105,
+      "learning_rate": 2.9570918362768386e-05,
+      "loss": 1.6321,
+      "step": 1487
+    },
+    {
+      "epoch": 0.783157894736842,
+      "grad_norm": 1.5938304662704468,
+      "learning_rate": 2.9570327942763535e-05,
+      "loss": 1.1299,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7836842105263158,
+      "grad_norm": 0.9389225840568542,
+      "learning_rate": 2.9569737122728823e-05,
+      "loss": 0.9162,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7842105263157895,
+      "grad_norm": 2.557969331741333,
+      "learning_rate": 2.956914590268048e-05,
+      "loss": 1.1852,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7847368421052632,
+      "grad_norm": 1.1017169952392578,
+      "learning_rate": 2.9568554282634733e-05,
+      "loss": 1.3037,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7852631578947369,
+      "grad_norm": 1.8029855489730835,
+      "learning_rate": 2.956796226260783e-05,
+      "loss": 1.9646,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7857894736842105,
+      "grad_norm": 10.63049030303955,
+      "learning_rate": 2.9567369842616015e-05,
+      "loss": 0.8632,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7863157894736842,
+      "grad_norm": 6.249732494354248,
+      "learning_rate": 2.9566777022675563e-05,
+      "loss": 0.5931,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7868421052631579,
+      "grad_norm": 3.90535306930542,
+      "learning_rate": 2.956618380280275e-05,
+      "loss": 1.2675,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7873684210526316,
+      "grad_norm": 5.921347618103027,
+      "learning_rate": 2.9565590183013855e-05,
+      "loss": 0.2349,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7878947368421053,
+      "grad_norm": 0.9792536497116089,
+      "learning_rate": 2.9564996163325186e-05,
+      "loss": 1.1103,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7884210526315789,
+      "grad_norm": 3.9059622287750244,
+      "learning_rate": 2.956440174375304e-05,
+      "loss": 0.0963,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7889473684210526,
+      "grad_norm": 2.304387092590332,
+      "learning_rate": 2.9563806924313746e-05,
+      "loss": 1.1423,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 6.4439263343811035,
+      "learning_rate": 2.956321170502363e-05,
+      "loss": 1.7772,
+      "step": 1500
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.323302745819092,
+      "learning_rate": 2.956261608589904e-05,
+      "loss": 0.9714,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7905263157894736,
+      "grad_norm": 2.1680712699890137,
+      "learning_rate": 2.956202006695632e-05,
+      "loss": 1.3609,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7910526315789473,
+      "grad_norm": 1.4925416707992554,
+      "learning_rate": 2.9561423648211842e-05,
+      "loss": 1.1193,
+      "step": 1503
+    },
+    {
+      "epoch": 0.791578947368421,
+      "grad_norm": 2.1561765670776367,
+      "learning_rate": 2.956082682968197e-05,
+      "loss": 1.3798,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7921052631578948,
+      "grad_norm": 2.159775495529175,
+      "learning_rate": 2.9560229611383104e-05,
+      "loss": 1.2499,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7926315789473685,
+      "grad_norm": 2.9001150131225586,
+      "learning_rate": 2.955963199333163e-05,
+      "loss": 1.2439,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7931578947368421,
+      "grad_norm": 20.27045440673828,
+      "learning_rate": 2.9559033975543962e-05,
+      "loss": 1.3729,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7936842105263158,
+      "grad_norm": 1.0026558637619019,
+      "learning_rate": 2.9558435558036507e-05,
+      "loss": 0.8089,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7942105263157895,
+      "grad_norm": 2.685729503631592,
+      "learning_rate": 2.955783674082571e-05,
+      "loss": 0.9878,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7947368421052632,
+      "grad_norm": 15.882140159606934,
+      "learning_rate": 2.9557237523928005e-05,
+      "loss": 2.2313,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7952631578947369,
+      "grad_norm": 1.5289243459701538,
+      "learning_rate": 2.9556637907359845e-05,
+      "loss": 0.6709,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7957894736842105,
+      "grad_norm": 1.3777986764907837,
+      "learning_rate": 2.9556037891137686e-05,
+      "loss": 0.8706,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7963157894736842,
+      "grad_norm": 3.52500319480896,
+      "learning_rate": 2.9555437475278013e-05,
+      "loss": 2.0162,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7968421052631579,
+      "grad_norm": 1.4890594482421875,
+      "learning_rate": 2.95548366597973e-05,
+      "loss": 0.8976,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7973684210526316,
+      "grad_norm": 3.3992414474487305,
+      "learning_rate": 2.9554235444712045e-05,
+      "loss": 0.5914,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7978947368421052,
+      "grad_norm": 4.6464152336120605,
+      "learning_rate": 2.9553633830038757e-05,
+      "loss": 2.0883,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7984210526315789,
+      "grad_norm": 2.2460274696350098,
+      "learning_rate": 2.955303181579395e-05,
+      "loss": 1.3262,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7989473684210526,
+      "grad_norm": 1.4003204107284546,
+      "learning_rate": 2.955242940199416e-05,
+      "loss": 1.3034,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7994736842105263,
+      "grad_norm": 1.23296058177948,
+      "learning_rate": 2.955182658865592e-05,
+      "loss": 1.2704,
+      "step": 1519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.1843321323394775,
+      "learning_rate": 2.9551223375795778e-05,
+      "loss": 1.0394,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8005263157894736,
+      "grad_norm": 1.3480000495910645,
+      "learning_rate": 2.95506197634303e-05,
+      "loss": 1.1087,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8010526315789473,
+      "grad_norm": 1.0556117296218872,
+      "learning_rate": 2.9550015751576055e-05,
+      "loss": 0.9642,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8015789473684211,
+      "grad_norm": 5.54053258895874,
+      "learning_rate": 2.9549411340249627e-05,
+      "loss": 1.3033,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8021052631578948,
+      "grad_norm": 3.8652968406677246,
+      "learning_rate": 2.9548806529467617e-05,
+      "loss": 1.0023,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8026315789473685,
+      "grad_norm": 3.5290725231170654,
+      "learning_rate": 2.954820131924662e-05,
+      "loss": 0.5537,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8031578947368421,
+      "grad_norm": 0.9234074950218201,
+      "learning_rate": 2.9547595709603255e-05,
+      "loss": 0.9543,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8036842105263158,
+      "grad_norm": 13.975532531738281,
+      "learning_rate": 2.9546989700554154e-05,
+      "loss": 1.2983,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8042105263157895,
+      "grad_norm": 1.407110333442688,
+      "learning_rate": 2.9546383292115947e-05,
+      "loss": 1.2283,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8047368421052632,
+      "grad_norm": 1.1215810775756836,
+      "learning_rate": 2.9545776484305293e-05,
+      "loss": 0.9641,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8052631578947368,
+      "grad_norm": 2.0094728469848633,
+      "learning_rate": 2.9545169277138845e-05,
+      "loss": 1.0771,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8057894736842105,
+      "grad_norm": 3.301666021347046,
+      "learning_rate": 2.9544561670633272e-05,
+      "loss": 0.5507,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8063157894736842,
+      "grad_norm": 2.8523879051208496,
+      "learning_rate": 2.954395366480526e-05,
+      "loss": 0.414,
+      "step": 1532
+    },
+    {
+      "epoch": 0.8068421052631579,
+      "grad_norm": 5.4080986976623535,
+      "learning_rate": 2.9543345259671505e-05,
+      "loss": 0.9862,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8073684210526316,
+      "grad_norm": 3.0157690048217773,
+      "learning_rate": 2.9542736455248702e-05,
+      "loss": 1.1812,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8078947368421052,
+      "grad_norm": 2.6960086822509766,
+      "learning_rate": 2.9542127251553573e-05,
+      "loss": 0.8758,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8084210526315789,
+      "grad_norm": 1.4676748514175415,
+      "learning_rate": 2.954151764860284e-05,
+      "loss": 1.2065,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8089473684210526,
+      "grad_norm": 1.893027424812317,
+      "learning_rate": 2.9540907646413247e-05,
+      "loss": 1.7415,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8094736842105263,
+      "grad_norm": 2.425906181335449,
+      "learning_rate": 2.954029724500153e-05,
+      "loss": 1.0721,
+      "step": 1538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.737662315368652,
+      "learning_rate": 2.953968644438445e-05,
+      "loss": 1.0819,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8105263157894737,
+      "grad_norm": 2.0641164779663086,
+      "learning_rate": 2.9539075244578793e-05,
+      "loss": 1.6513,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8110526315789474,
+      "grad_norm": 1.7806847095489502,
+      "learning_rate": 2.9538463645601317e-05,
+      "loss": 1.0087,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8115789473684211,
+      "grad_norm": 1.2224940061569214,
+      "learning_rate": 2.9537851647468827e-05,
+      "loss": 1.0176,
+      "step": 1542
+    },
+    {
+      "epoch": 0.8121052631578948,
+      "grad_norm": 3.8275363445281982,
+      "learning_rate": 2.953723925019812e-05,
+      "loss": 0.2529,
+      "step": 1543
+    },
+    {
+      "epoch": 0.8126315789473684,
+      "grad_norm": 4.72785758972168,
+      "learning_rate": 2.9536626453806008e-05,
+      "loss": 1.2509,
+      "step": 1544
+    },
+    {
+      "epoch": 0.8131578947368421,
+      "grad_norm": 1.0455762147903442,
+      "learning_rate": 2.9536013258309323e-05,
+      "loss": 0.766,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8136842105263158,
+      "grad_norm": 1.5553256273269653,
+      "learning_rate": 2.9535399663724893e-05,
+      "loss": 1.1175,
+      "step": 1546
+    },
+    {
+      "epoch": 0.8142105263157895,
+      "grad_norm": 2.9578685760498047,
+      "learning_rate": 2.953478567006957e-05,
+      "loss": 0.6569,
+      "step": 1547
+    },
+    {
+      "epoch": 0.8147368421052632,
+      "grad_norm": 1.4119994640350342,
+      "learning_rate": 2.9534171277360207e-05,
+      "loss": 1.2938,
+      "step": 1548
+    },
+    {
+      "epoch": 0.8152631578947368,
+      "grad_norm": 2.524916887283325,
+      "learning_rate": 2.9533556485613673e-05,
+      "loss": 0.4687,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 2.7491307258605957,
+      "learning_rate": 2.9532941294846848e-05,
+      "loss": 0.9021,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8163157894736842,
+      "grad_norm": 3.837688446044922,
+      "learning_rate": 2.9532325705076622e-05,
+      "loss": 0.7552,
+      "step": 1551
+    },
+    {
+      "epoch": 0.8168421052631579,
+      "grad_norm": 1.9643511772155762,
+      "learning_rate": 2.9531709716319895e-05,
+      "loss": 1.2493,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8173684210526316,
+      "grad_norm": 2.230646848678589,
+      "learning_rate": 2.953109332859358e-05,
+      "loss": 0.1658,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8178947368421052,
+      "grad_norm": 4.156702995300293,
+      "learning_rate": 2.9530476541914602e-05,
+      "loss": 0.4678,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8184210526315789,
+      "grad_norm": 10.191547393798828,
+      "learning_rate": 2.9529859356299894e-05,
+      "loss": 0.6664,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8189473684210526,
+      "grad_norm": 1.6715644598007202,
+      "learning_rate": 2.9529241771766396e-05,
+      "loss": 0.929,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8194736842105264,
+      "grad_norm": 1.5630298852920532,
+      "learning_rate": 2.9528623788331067e-05,
+      "loss": 1.5926,
+      "step": 1557
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.9629634618759155,
+      "learning_rate": 2.9528005406010877e-05,
+      "loss": 0.7064,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8205263157894737,
+      "grad_norm": 2.5397849082946777,
+      "learning_rate": 2.9527386624822805e-05,
+      "loss": 1.0266,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8210526315789474,
+      "grad_norm": 0.8805807828903198,
+      "learning_rate": 2.952676744478383e-05,
+      "loss": 0.1851,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8215789473684211,
+      "grad_norm": 1.243837594985962,
+      "learning_rate": 2.952614786591096e-05,
+      "loss": 1.4598,
+      "step": 1561
+    },
+    {
+      "epoch": 0.8221052631578948,
+      "grad_norm": 2.9784929752349854,
+      "learning_rate": 2.9525527888221203e-05,
+      "loss": 1.4596,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8226315789473684,
+      "grad_norm": 1.0616602897644043,
+      "learning_rate": 2.9524907511731582e-05,
+      "loss": 1.151,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8231578947368421,
+      "grad_norm": 3.9415078163146973,
+      "learning_rate": 2.9524286736459125e-05,
+      "loss": 0.7022,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8236842105263158,
+      "grad_norm": 4.667412757873535,
+      "learning_rate": 2.9523665562420882e-05,
+      "loss": 1.614,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8242105263157895,
+      "grad_norm": 2.3108322620391846,
+      "learning_rate": 2.95230439896339e-05,
+      "loss": 0.8213,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8247368421052632,
+      "grad_norm": 1.3090465068817139,
+      "learning_rate": 2.9522422018115254e-05,
+      "loss": 0.9995,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8252631578947368,
+      "grad_norm": 2.5572314262390137,
+      "learning_rate": 2.952179964788201e-05,
+      "loss": 0.3968,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8257894736842105,
+      "grad_norm": 4.847593784332275,
+      "learning_rate": 2.9521176878951262e-05,
+      "loss": 1.5573,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8263157894736842,
+      "grad_norm": 3.584441661834717,
+      "learning_rate": 2.9520553711340107e-05,
+      "loss": 1.6182,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8268421052631579,
+      "grad_norm": 0.06894690543413162,
+      "learning_rate": 2.951993014506565e-05,
+      "loss": 0.0021,
+      "step": 1571
+    },
+    {
+      "epoch": 0.8273684210526315,
+      "grad_norm": 1.0614222288131714,
+      "learning_rate": 2.951930618014502e-05,
+      "loss": 1.1364,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8278947368421052,
+      "grad_norm": 1.2423781156539917,
+      "learning_rate": 2.9518681816595337e-05,
+      "loss": 0.7434,
+      "step": 1573
+    },
+    {
+      "epoch": 0.828421052631579,
+      "grad_norm": 1.0514869689941406,
+      "learning_rate": 2.9518057054433753e-05,
+      "loss": 0.9144,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8289473684210527,
+      "grad_norm": 6.573333263397217,
+      "learning_rate": 2.951743189367741e-05,
+      "loss": 0.6118,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8294736842105264,
+      "grad_norm": 6.8987016677856445,
+      "learning_rate": 2.9516806334343482e-05,
+      "loss": 0.7301,
+      "step": 1576
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.210350513458252,
+      "learning_rate": 2.951618037644914e-05,
+      "loss": 0.9962,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8305263157894737,
+      "grad_norm": 4.0031328201293945,
+      "learning_rate": 2.9515554020011567e-05,
+      "loss": 1.1273,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8310526315789474,
+      "grad_norm": 6.773041725158691,
+      "learning_rate": 2.9514927265047968e-05,
+      "loss": 2.3973,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8315789473684211,
+      "grad_norm": 2.425973892211914,
+      "learning_rate": 2.951430011157554e-05,
+      "loss": 0.6053,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8321052631578948,
+      "grad_norm": 6.728580951690674,
+      "learning_rate": 2.951367255961151e-05,
+      "loss": 0.9388,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8326315789473684,
+      "grad_norm": 1.9585034847259521,
+      "learning_rate": 2.95130446091731e-05,
+      "loss": 1.0964,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8331578947368421,
+      "grad_norm": 1.406274676322937,
+      "learning_rate": 2.9512416260277554e-05,
+      "loss": 1.4249,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8336842105263158,
+      "grad_norm": 20.80072593688965,
+      "learning_rate": 2.9511787512942125e-05,
+      "loss": 0.8256,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8342105263157895,
+      "grad_norm": 1.4816763401031494,
+      "learning_rate": 2.9511158367184078e-05,
+      "loss": 0.6075,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8347368421052631,
+      "grad_norm": 6.253325939178467,
+      "learning_rate": 2.9510528823020683e-05,
+      "loss": 0.124,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8352631578947368,
+      "grad_norm": 1.8239108324050903,
+      "learning_rate": 2.9509898880469215e-05,
+      "loss": 0.7881,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8357894736842105,
+      "grad_norm": 9.914002418518066,
+      "learning_rate": 2.9509268539546985e-05,
+      "loss": 0.3637,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8363157894736842,
+      "grad_norm": 2.900563955307007,
+      "learning_rate": 2.9508637800271293e-05,
+      "loss": 0.9361,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8368421052631579,
+      "grad_norm": 1.4834281206130981,
+      "learning_rate": 2.9508006662659447e-05,
+      "loss": 0.4473,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8373684210526315,
+      "grad_norm": 2.494849443435669,
+      "learning_rate": 2.9507375126728787e-05,
+      "loss": 1.2612,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8378947368421052,
+      "grad_norm": 1.2127383947372437,
+      "learning_rate": 2.950674319249665e-05,
+      "loss": 0.8488,
+      "step": 1592
+    },
+    {
+      "epoch": 0.838421052631579,
+      "grad_norm": 1.7087959051132202,
+      "learning_rate": 2.950611085998038e-05,
+      "loss": 0.1118,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8389473684210527,
+      "grad_norm": 1.0232149362564087,
+      "learning_rate": 2.9505478129197343e-05,
+      "loss": 0.0376,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8394736842105263,
+      "grad_norm": 1.3383780717849731,
+      "learning_rate": 2.950484500016491e-05,
+      "loss": 1.1654,
+      "step": 1595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3777598142623901,
+      "learning_rate": 2.950421147290046e-05,
+      "loss": 0.8772,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8405263157894737,
+      "grad_norm": 3.801234722137451,
+      "learning_rate": 2.950357754742139e-05,
+      "loss": 1.8463,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8410526315789474,
+      "grad_norm": 1.922048568725586,
+      "learning_rate": 2.9502943223745104e-05,
+      "loss": 1.6611,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8415789473684211,
+      "grad_norm": 2.133612632751465,
+      "learning_rate": 2.9502308501889016e-05,
+      "loss": 2.0384,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 3.9198570251464844,
+      "learning_rate": 2.950167338187056e-05,
+      "loss": 1.6621,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8426315789473684,
+      "grad_norm": 2.9979448318481445,
+      "learning_rate": 2.950103786370716e-05,
+      "loss": 0.2946,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8431578947368421,
+      "grad_norm": 1.4680919647216797,
+      "learning_rate": 2.950040194741627e-05,
+      "loss": 1.1285,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8436842105263158,
+      "grad_norm": 2.139376640319824,
+      "learning_rate": 2.9499765633015354e-05,
+      "loss": 0.8158,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8442105263157895,
+      "grad_norm": 2.2096030712127686,
+      "learning_rate": 2.9499128920521875e-05,
+      "loss": 0.0758,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8447368421052631,
+      "grad_norm": 2.995516300201416,
+      "learning_rate": 2.949849180995332e-05,
+      "loss": 0.8989,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8452631578947368,
+      "grad_norm": 3.639786958694458,
+      "learning_rate": 2.9497854301327175e-05,
+      "loss": 1.5573,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8457894736842105,
+      "grad_norm": 9.909292221069336,
+      "learning_rate": 2.9497216394660948e-05,
+      "loss": 0.7076,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8463157894736842,
+      "grad_norm": 8.76042366027832,
+      "learning_rate": 2.949657808997215e-05,
+      "loss": 1.8939,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8468421052631578,
+      "grad_norm": 8.321390151977539,
+      "learning_rate": 2.9495939387278303e-05,
+      "loss": 0.3237,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8473684210526315,
+      "grad_norm": 2.2840054035186768,
+      "learning_rate": 2.949530028659695e-05,
+      "loss": 1.3387,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8478947368421053,
+      "grad_norm": 1.5071499347686768,
+      "learning_rate": 2.9494660787945634e-05,
+      "loss": 1.1994,
+      "step": 1611
+    },
+    {
+      "epoch": 0.848421052631579,
+      "grad_norm": 8.956307411193848,
+      "learning_rate": 2.9494020891341912e-05,
+      "loss": 1.5923,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8489473684210527,
+      "grad_norm": 1.6805635690689087,
+      "learning_rate": 2.949338059680335e-05,
+      "loss": 1.6323,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8494736842105263,
+      "grad_norm": 0.6971753835678101,
+      "learning_rate": 2.9492739904347533e-05,
+      "loss": 0.0236,
+      "step": 1614
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6537336111068726,
+      "learning_rate": 2.9492098813992045e-05,
+      "loss": 1.2005,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8505263157894737,
+      "grad_norm": 1.5291582345962524,
+      "learning_rate": 2.9491457325754495e-05,
+      "loss": 1.1125,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8510526315789474,
+      "grad_norm": 2.4934237003326416,
+      "learning_rate": 2.949081543965249e-05,
+      "loss": 1.3744,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8515789473684211,
+      "grad_norm": 2.8750171661376953,
+      "learning_rate": 2.949017315570365e-05,
+      "loss": 1.6325,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8521052631578947,
+      "grad_norm": 0.9199529886245728,
+      "learning_rate": 2.9489530473925615e-05,
+      "loss": 1.0891,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8526315789473684,
+      "grad_norm": 4.175845146179199,
+      "learning_rate": 2.9488887394336025e-05,
+      "loss": 0.306,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8531578947368421,
+      "grad_norm": 2.5200769901275635,
+      "learning_rate": 2.948824391695254e-05,
+      "loss": 0.2277,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8536842105263158,
+      "grad_norm": 1.024670124053955,
+      "learning_rate": 2.9487600041792825e-05,
+      "loss": 1.073,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8542105263157894,
+      "grad_norm": 3.6856977939605713,
+      "learning_rate": 2.9486955768874555e-05,
+      "loss": 0.9715,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8547368421052631,
+      "grad_norm": 2.900958776473999,
+      "learning_rate": 2.9486311098215425e-05,
+      "loss": 1.142,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8552631578947368,
+      "grad_norm": 1.7244210243225098,
+      "learning_rate": 2.948566602983313e-05,
+      "loss": 1.1344,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8557894736842105,
+      "grad_norm": 5.9757232666015625,
+      "learning_rate": 2.948502056374538e-05,
+      "loss": 0.8719,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8563157894736843,
+      "grad_norm": 3.8142871856689453,
+      "learning_rate": 2.94843746999699e-05,
+      "loss": 0.8588,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8568421052631578,
+      "grad_norm": 5.312744617462158,
+      "learning_rate": 2.9483728438524417e-05,
+      "loss": 0.9706,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8573684210526316,
+      "grad_norm": 0.9603210687637329,
+      "learning_rate": 2.9483081779426678e-05,
+      "loss": 0.8257,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8578947368421053,
+      "grad_norm": 12.436172485351562,
+      "learning_rate": 2.9482434722694434e-05,
+      "loss": 1.303,
+      "step": 1630
+    },
+    {
+      "epoch": 0.858421052631579,
+      "grad_norm": 2.7977070808410645,
+      "learning_rate": 2.9481787268345456e-05,
+      "loss": 1.4104,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8589473684210527,
+      "grad_norm": 1.3737499713897705,
+      "learning_rate": 2.9481139416397512e-05,
+      "loss": 1.7754,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8594736842105263,
+      "grad_norm": 4.523530960083008,
+      "learning_rate": 2.9480491166868396e-05,
+      "loss": 1.9328,
+      "step": 1633
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 14.998103141784668,
+      "learning_rate": 2.9479842519775903e-05,
+      "loss": 1.3404,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8605263157894737,
+      "grad_norm": 3.4863827228546143,
+      "learning_rate": 2.9479193475137834e-05,
+      "loss": 1.7372,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8610526315789474,
+      "grad_norm": 3.7734429836273193,
+      "learning_rate": 2.9478544032972024e-05,
+      "loss": 2.3958,
+      "step": 1636
+    },
+    {
+      "epoch": 0.861578947368421,
+      "grad_norm": 1.744239091873169,
+      "learning_rate": 2.9477894193296295e-05,
+      "loss": 0.9227,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8621052631578947,
+      "grad_norm": 2.1991264820098877,
+      "learning_rate": 2.9477243956128484e-05,
+      "loss": 1.5509,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8626315789473684,
+      "grad_norm": 13.812848091125488,
+      "learning_rate": 2.9476593321486455e-05,
+      "loss": 1.1571,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8631578947368421,
+      "grad_norm": 5.264194488525391,
+      "learning_rate": 2.9475942289388056e-05,
+      "loss": 1.2813,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8636842105263158,
+      "grad_norm": 1.047844648361206,
+      "learning_rate": 2.9475290859851173e-05,
+      "loss": 0.9731,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8642105263157894,
+      "grad_norm": 6.97605562210083,
+      "learning_rate": 2.9474639032893685e-05,
+      "loss": 1.1533,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8647368421052631,
+      "grad_norm": 1.112858772277832,
+      "learning_rate": 2.9473986808533495e-05,
+      "loss": 1.062,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8652631578947368,
+      "grad_norm": 1.1823147535324097,
+      "learning_rate": 2.9473334186788503e-05,
+      "loss": 1.3267,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8657894736842106,
+      "grad_norm": 1.9988354444503784,
+      "learning_rate": 2.947268116767663e-05,
+      "loss": 0.8981,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8663157894736843,
+      "grad_norm": 10.701164245605469,
+      "learning_rate": 2.9472027751215803e-05,
+      "loss": 1.2634,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8668421052631579,
+      "grad_norm": 8.62368106842041,
+      "learning_rate": 2.9471373937423963e-05,
+      "loss": 0.6866,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8673684210526316,
+      "grad_norm": 7.085811138153076,
+      "learning_rate": 2.947071972631906e-05,
+      "loss": 0.4653,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8678947368421053,
+      "grad_norm": 0.7952293753623962,
+      "learning_rate": 2.9470065117919057e-05,
+      "loss": 0.688,
+      "step": 1649
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 0.9955103993415833,
+      "learning_rate": 2.9469410112241925e-05,
+      "loss": 0.8748,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8689473684210526,
+      "grad_norm": 1.0168907642364502,
+      "learning_rate": 2.9468754709305643e-05,
+      "loss": 1.3661,
+      "step": 1651
+    },
+    {
+      "epoch": 0.8694736842105263,
+      "grad_norm": 5.310661315917969,
+      "learning_rate": 2.9468098909128212e-05,
+      "loss": 0.8655,
+      "step": 1652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.678101539611816,
+      "learning_rate": 2.9467442711727637e-05,
+      "loss": 0.914,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8705263157894737,
+      "grad_norm": 7.018301963806152,
+      "learning_rate": 2.9466786117121928e-05,
+      "loss": 1.3955,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8710526315789474,
+      "grad_norm": 1.5973879098892212,
+      "learning_rate": 2.9466129125329114e-05,
+      "loss": 1.1749,
+      "step": 1655
+    },
+    {
+      "epoch": 0.871578947368421,
+      "grad_norm": 1.043853521347046,
+      "learning_rate": 2.9465471736367234e-05,
+      "loss": 0.0401,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8721052631578947,
+      "grad_norm": 1.4394428730010986,
+      "learning_rate": 2.9464813950254336e-05,
+      "loss": 0.0699,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8726315789473684,
+      "grad_norm": 1.1438318490982056,
+      "learning_rate": 2.9464155767008485e-05,
+      "loss": 1.5743,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8731578947368421,
+      "grad_norm": 7.7294392585754395,
+      "learning_rate": 2.9463497186647747e-05,
+      "loss": 0.961,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8736842105263158,
+      "grad_norm": 2.75779390335083,
+      "learning_rate": 2.9462838209190198e-05,
+      "loss": 0.7825,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8742105263157894,
+      "grad_norm": 1.5052375793457031,
+      "learning_rate": 2.9462178834653937e-05,
+      "loss": 1.0688,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8747368421052631,
+      "grad_norm": 4.883263111114502,
+      "learning_rate": 2.9461519063057064e-05,
+      "loss": 0.4308,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8752631578947369,
+      "grad_norm": 2.8982460498809814,
+      "learning_rate": 2.9460858894417694e-05,
+      "loss": 0.4369,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8757894736842106,
+      "grad_norm": 38.408145904541016,
+      "learning_rate": 2.9460198328753955e-05,
+      "loss": 2.1035,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8763157894736842,
+      "grad_norm": 1.7316091060638428,
+      "learning_rate": 2.9459537366083983e-05,
+      "loss": 0.9198,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8768421052631579,
+      "grad_norm": 0.7877925634384155,
+      "learning_rate": 2.945887600642592e-05,
+      "loss": 0.5887,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8773684210526316,
+      "grad_norm": 1.3625367879867554,
+      "learning_rate": 2.9458214249797924e-05,
+      "loss": 1.1236,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8778947368421053,
+      "grad_norm": 1.776649832725525,
+      "learning_rate": 2.9457552096218168e-05,
+      "loss": 1.2667,
+      "step": 1668
+    },
+    {
+      "epoch": 0.878421052631579,
+      "grad_norm": 5.151728630065918,
+      "learning_rate": 2.945688954570483e-05,
+      "loss": 1.5045,
+      "step": 1669
+    },
+    {
+      "epoch": 0.8789473684210526,
+      "grad_norm": 1.2965806722640991,
+      "learning_rate": 2.9456226598276097e-05,
+      "loss": 1.1484,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8794736842105263,
+      "grad_norm": 1.1582598686218262,
+      "learning_rate": 2.9455563253950176e-05,
+      "loss": 1.1762,
+      "step": 1671
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.36055588722229,
+      "learning_rate": 2.945489951274527e-05,
+      "loss": 1.2704,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8805263157894737,
+      "grad_norm": 5.386001110076904,
+      "learning_rate": 2.9454235374679612e-05,
+      "loss": 1.0296,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8810526315789474,
+      "grad_norm": 4.373623371124268,
+      "learning_rate": 2.9453570839771432e-05,
+      "loss": 1.3767,
+      "step": 1674
+    },
+    {
+      "epoch": 0.881578947368421,
+      "grad_norm": 1.0014320611953735,
+      "learning_rate": 2.9452905908038975e-05,
+      "loss": 1.3597,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8821052631578947,
+      "grad_norm": 4.0518412590026855,
+      "learning_rate": 2.9452240579500496e-05,
+      "loss": 0.7262,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8826315789473684,
+      "grad_norm": 2.685443878173828,
+      "learning_rate": 2.9451574854174265e-05,
+      "loss": 1.2386,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8831578947368421,
+      "grad_norm": 2.57368540763855,
+      "learning_rate": 2.9450908732078553e-05,
+      "loss": 1.0309,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8836842105263157,
+      "grad_norm": 1.2405550479888916,
+      "learning_rate": 2.9450242213231654e-05,
+      "loss": 1.1075,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8842105263157894,
+      "grad_norm": 1.6623353958129883,
+      "learning_rate": 2.9449575297651865e-05,
+      "loss": 0.9361,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8847368421052632,
+      "grad_norm": 14.208858489990234,
+      "learning_rate": 2.9448907985357498e-05,
+      "loss": 2.1596,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8852631578947369,
+      "grad_norm": 1.5413142442703247,
+      "learning_rate": 2.944824027636687e-05,
+      "loss": 1.4468,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8857894736842106,
+      "grad_norm": 0.9851499795913696,
+      "learning_rate": 2.9447572170698324e-05,
+      "loss": 0.9406,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8863157894736842,
+      "grad_norm": 0.8356818556785583,
+      "learning_rate": 2.9446903668370188e-05,
+      "loss": 0.0297,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8868421052631579,
+      "grad_norm": 2.6945416927337646,
+      "learning_rate": 2.944623476940082e-05,
+      "loss": 1.2829,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8873684210526316,
+      "grad_norm": 33.185028076171875,
+      "learning_rate": 2.9445565473808593e-05,
+      "loss": 0.9756,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8878947368421053,
+      "grad_norm": 1.7505784034729004,
+      "learning_rate": 2.9444895781611876e-05,
+      "loss": 1.2576,
+      "step": 1687
+    },
+    {
+      "epoch": 0.888421052631579,
+      "grad_norm": 4.850620269775391,
+      "learning_rate": 2.944422569282906e-05,
+      "loss": 0.1032,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8889473684210526,
+      "grad_norm": 1.0093870162963867,
+      "learning_rate": 2.9443555207478536e-05,
+      "loss": 0.9847,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8894736842105263,
+      "grad_norm": 2.135535955429077,
+      "learning_rate": 2.9442884325578714e-05,
+      "loss": 1.3696,
+      "step": 1690
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.9265260696411133,
+      "learning_rate": 2.9442213047148012e-05,
+      "loss": 0.9913,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8905263157894737,
+      "grad_norm": 4.452930927276611,
+      "learning_rate": 2.944154137220487e-05,
+      "loss": 0.7837,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8910526315789473,
+      "grad_norm": 2.796504259109497,
+      "learning_rate": 2.944086930076771e-05,
+      "loss": 0.442,
+      "step": 1693
+    },
+    {
+      "epoch": 0.891578947368421,
+      "grad_norm": 2.876086950302124,
+      "learning_rate": 2.9440196832855004e-05,
+      "loss": 1.2997,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8921052631578947,
+      "grad_norm": 1.2075906991958618,
+      "learning_rate": 2.94395239684852e-05,
+      "loss": 0.8753,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8926315789473684,
+      "grad_norm": 7.354285717010498,
+      "learning_rate": 2.9438850707676786e-05,
+      "loss": 0.9609,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8931578947368422,
+      "grad_norm": 5.216385364532471,
+      "learning_rate": 2.943817705044823e-05,
+      "loss": 1.3972,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8936842105263157,
+      "grad_norm": 1.4113494157791138,
+      "learning_rate": 2.9437502996818035e-05,
+      "loss": 1.4202,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8942105263157895,
+      "grad_norm": 6.46826696395874,
+      "learning_rate": 2.9436828546804707e-05,
+      "loss": 0.6637,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 1.4000974893569946,
+      "learning_rate": 2.943615370042677e-05,
+      "loss": 0.0493,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8952631578947369,
+      "grad_norm": 1.1557705402374268,
+      "learning_rate": 2.943547845770274e-05,
+      "loss": 0.4349,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8957894736842106,
+      "grad_norm": 8.672079086303711,
+      "learning_rate": 2.943480281865116e-05,
+      "loss": 1.1025,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8963157894736842,
+      "grad_norm": 1.1264904737472534,
+      "learning_rate": 2.943412678329058e-05,
+      "loss": 1.0514,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8968421052631579,
+      "grad_norm": 0.8493797183036804,
+      "learning_rate": 2.9433450351639567e-05,
+      "loss": 0.7548,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8973684210526316,
+      "grad_norm": 3.7704265117645264,
+      "learning_rate": 2.9432773523716683e-05,
+      "loss": 1.2445,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8978947368421053,
+      "grad_norm": 1.2486798763275146,
+      "learning_rate": 2.9432096299540518e-05,
+      "loss": 1.1246,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8984210526315789,
+      "grad_norm": 1.134177803993225,
+      "learning_rate": 2.9431418679129655e-05,
+      "loss": 0.8451,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8989473684210526,
+      "grad_norm": 1.9805290699005127,
+      "learning_rate": 2.9430740662502712e-05,
+      "loss": 0.4611,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8994736842105263,
+      "grad_norm": 2.323180675506592,
+      "learning_rate": 2.9430062249678297e-05,
+      "loss": 1.2539,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.190372347831726,
+      "learning_rate": 2.942938344067503e-05,
+      "loss": 0.992,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9005263157894737,
+      "grad_norm": 2.458120822906494,
+      "learning_rate": 2.9428704235511557e-05,
+      "loss": 1.1641,
+      "step": 1711
+    },
+    {
+      "epoch": 0.9010526315789473,
+      "grad_norm": 1.1921024322509766,
+      "learning_rate": 2.942802463420652e-05,
+      "loss": 0.9713,
+      "step": 1712
+    },
+    {
+      "epoch": 0.901578947368421,
+      "grad_norm": 1.0940788984298706,
+      "learning_rate": 2.942734463677858e-05,
+      "loss": 0.8756,
+      "step": 1713
+    },
+    {
+      "epoch": 0.9021052631578947,
+      "grad_norm": 1.3359793424606323,
+      "learning_rate": 2.9426664243246404e-05,
+      "loss": 1.3792,
+      "step": 1714
+    },
+    {
+      "epoch": 0.9026315789473685,
+      "grad_norm": 2.6136605739593506,
+      "learning_rate": 2.9425983453628677e-05,
+      "loss": 0.954,
+      "step": 1715
+    },
+    {
+      "epoch": 0.9031578947368422,
+      "grad_norm": 1.5081348419189453,
+      "learning_rate": 2.942530226794409e-05,
+      "loss": 1.3498,
+      "step": 1716
+    },
+    {
+      "epoch": 0.9036842105263158,
+      "grad_norm": 3.4692232608795166,
+      "learning_rate": 2.942462068621134e-05,
+      "loss": 1.0917,
+      "step": 1717
+    },
+    {
+      "epoch": 0.9042105263157895,
+      "grad_norm": 4.044422149658203,
+      "learning_rate": 2.942393870844914e-05,
+      "loss": 1.806,
+      "step": 1718
+    },
+    {
+      "epoch": 0.9047368421052632,
+      "grad_norm": 1.875767707824707,
+      "learning_rate": 2.9423256334676215e-05,
+      "loss": 1.1397,
+      "step": 1719
+    },
+    {
+      "epoch": 0.9052631578947369,
+      "grad_norm": 2.141014337539673,
+      "learning_rate": 2.9422573564911305e-05,
+      "loss": 1.2411,
+      "step": 1720
+    },
+    {
+      "epoch": 0.9057894736842105,
+      "grad_norm": 1.6898494958877563,
+      "learning_rate": 2.9421890399173153e-05,
+      "loss": 1.4801,
+      "step": 1721
+    },
+    {
+      "epoch": 0.9063157894736842,
+      "grad_norm": 2.024906635284424,
+      "learning_rate": 2.942120683748051e-05,
+      "loss": 0.317,
+      "step": 1722
+    },
+    {
+      "epoch": 0.9068421052631579,
+      "grad_norm": 18.252716064453125,
+      "learning_rate": 2.9420522879852148e-05,
+      "loss": 0.7255,
+      "step": 1723
+    },
+    {
+      "epoch": 0.9073684210526316,
+      "grad_norm": 2.799224853515625,
+      "learning_rate": 2.9419838526306845e-05,
+      "loss": 0.7641,
+      "step": 1724
+    },
+    {
+      "epoch": 0.9078947368421053,
+      "grad_norm": 1.1127315759658813,
+      "learning_rate": 2.941915377686339e-05,
+      "loss": 1.0762,
+      "step": 1725
+    },
+    {
+      "epoch": 0.9084210526315789,
+      "grad_norm": 2.6971869468688965,
+      "learning_rate": 2.9418468631540578e-05,
+      "loss": 1.15,
+      "step": 1726
+    },
+    {
+      "epoch": 0.9089473684210526,
+      "grad_norm": 1.2686923742294312,
+      "learning_rate": 2.9417783090357224e-05,
+      "loss": 0.9524,
+      "step": 1727
+    },
+    {
+      "epoch": 0.9094736842105263,
+      "grad_norm": 3.2218449115753174,
+      "learning_rate": 2.9417097153332152e-05,
+      "loss": 1.7397,
+      "step": 1728
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8627907037734985,
+      "learning_rate": 2.941641082048419e-05,
+      "loss": 1.2889,
+      "step": 1729
+    },
+    {
+      "epoch": 0.9105263157894737,
+      "grad_norm": 3.6902270317077637,
+      "learning_rate": 2.9415724091832184e-05,
+      "loss": 1.9215,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9110526315789473,
+      "grad_norm": 1.4535077810287476,
+      "learning_rate": 2.9415036967394988e-05,
+      "loss": 1.0473,
+      "step": 1731
+    },
+    {
+      "epoch": 0.911578947368421,
+      "grad_norm": 1.3268829584121704,
+      "learning_rate": 2.9414349447191466e-05,
+      "loss": 1.0177,
+      "step": 1732
+    },
+    {
+      "epoch": 0.9121052631578948,
+      "grad_norm": 1.7916074991226196,
+      "learning_rate": 2.9413661531240493e-05,
+      "loss": 1.0315,
+      "step": 1733
+    },
+    {
+      "epoch": 0.9126315789473685,
+      "grad_norm": 1.504307508468628,
+      "learning_rate": 2.941297321956096e-05,
+      "loss": 1.1304,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9131578947368421,
+      "grad_norm": 4.602481365203857,
+      "learning_rate": 2.9412284512171756e-05,
+      "loss": 1.52,
+      "step": 1735
+    },
+    {
+      "epoch": 0.9136842105263158,
+      "grad_norm": 4.820444107055664,
+      "learning_rate": 2.94115954090918e-05,
+      "loss": 1.8141,
+      "step": 1736
+    },
+    {
+      "epoch": 0.9142105263157895,
+      "grad_norm": 1.313570261001587,
+      "learning_rate": 2.9410905910340004e-05,
+      "loss": 0.741,
+      "step": 1737
+    },
+    {
+      "epoch": 0.9147368421052632,
+      "grad_norm": 1.3603005409240723,
+      "learning_rate": 2.9410216015935304e-05,
+      "loss": 1.313,
+      "step": 1738
+    },
+    {
+      "epoch": 0.9152631578947369,
+      "grad_norm": 4.094880104064941,
+      "learning_rate": 2.9409525725896636e-05,
+      "loss": 1.0756,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9157894736842105,
+      "grad_norm": 1.1733410358428955,
+      "learning_rate": 2.9408835040242953e-05,
+      "loss": 0.9995,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9163157894736842,
+      "grad_norm": 3.4846761226654053,
+      "learning_rate": 2.9408143958993218e-05,
+      "loss": 0.9578,
+      "step": 1741
+    },
+    {
+      "epoch": 0.9168421052631579,
+      "grad_norm": 1.4159212112426758,
+      "learning_rate": 2.9407452482166412e-05,
+      "loss": 1.1337,
+      "step": 1742
+    },
+    {
+      "epoch": 0.9173684210526316,
+      "grad_norm": 7.647956371307373,
+      "learning_rate": 2.9406760609781507e-05,
+      "loss": 1.4257,
+      "step": 1743
+    },
+    {
+      "epoch": 0.9178947368421052,
+      "grad_norm": 1.0013185739517212,
+      "learning_rate": 2.9406068341857505e-05,
+      "loss": 1.2189,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9184210526315789,
+      "grad_norm": 0.9043682813644409,
+      "learning_rate": 2.9405375678413417e-05,
+      "loss": 0.6736,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9189473684210526,
+      "grad_norm": 0.8028733134269714,
+      "learning_rate": 2.940468261946825e-05,
+      "loss": 0.9183,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9194736842105263,
+      "grad_norm": 11.042693138122559,
+      "learning_rate": 2.9403989165041044e-05,
+      "loss": 1.6583,
+      "step": 1747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.432318687438965,
+      "learning_rate": 2.940329531515082e-05,
+      "loss": 0.7686,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9205263157894736,
+      "grad_norm": 6.241179466247559,
+      "learning_rate": 2.9402601069816645e-05,
+      "loss": 0.654,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 4.325395584106445,
+      "learning_rate": 2.9401906429057574e-05,
+      "loss": 0.3556,
+      "step": 1750
+    },
+    {
+      "epoch": 0.921578947368421,
+      "grad_norm": 1.5290063619613647,
+      "learning_rate": 2.9401211392892677e-05,
+      "loss": 1.601,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9221052631578948,
+      "grad_norm": 1.4590153694152832,
+      "learning_rate": 2.9400515961341035e-05,
+      "loss": 1.0342,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9226315789473685,
+      "grad_norm": 2.1654052734375,
+      "learning_rate": 2.939982013442174e-05,
+      "loss": 1.6038,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9231578947368421,
+      "grad_norm": 11.959127426147461,
+      "learning_rate": 2.9399123912153908e-05,
+      "loss": 0.6332,
+      "step": 1754
+    },
+    {
+      "epoch": 0.9236842105263158,
+      "grad_norm": 2.1623222827911377,
+      "learning_rate": 2.939842729455664e-05,
+      "loss": 0.1993,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9242105263157895,
+      "grad_norm": 2.1751246452331543,
+      "learning_rate": 2.9397730281649067e-05,
+      "loss": 1.1215,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9247368421052632,
+      "grad_norm": 3.4143033027648926,
+      "learning_rate": 2.9397032873450323e-05,
+      "loss": 0.9459,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9252631578947368,
+      "grad_norm": 1.3410007953643799,
+      "learning_rate": 2.9396335069979562e-05,
+      "loss": 0.9964,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9257894736842105,
+      "grad_norm": 1.164724588394165,
+      "learning_rate": 2.9395636871255933e-05,
+      "loss": 0.3303,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9263157894736842,
+      "grad_norm": 1.904927372932434,
+      "learning_rate": 2.9394938277298613e-05,
+      "loss": 0.7966,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9268421052631579,
+      "grad_norm": 2.1621692180633545,
+      "learning_rate": 2.9394239288126782e-05,
+      "loss": 1.0655,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9273684210526316,
+      "grad_norm": 1.9396727085113525,
+      "learning_rate": 2.939353990375962e-05,
+      "loss": 1.1009,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9278947368421052,
+      "grad_norm": 2.1531522274017334,
+      "learning_rate": 2.9392840124216342e-05,
+      "loss": 0.9206,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9284210526315789,
+      "grad_norm": 1.473966360092163,
+      "learning_rate": 2.9392139949516154e-05,
+      "loss": 1.0062,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9289473684210526,
+      "grad_norm": 5.68750524520874,
+      "learning_rate": 2.939143937967828e-05,
+      "loss": 1.3941,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9294736842105263,
+      "grad_norm": 1.1354111433029175,
+      "learning_rate": 2.9390738414721954e-05,
+      "loss": 1.3447,
+      "step": 1766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.276787757873535,
+      "learning_rate": 2.939003705466642e-05,
+      "loss": 1.534,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9305263157894736,
+      "grad_norm": 1.2840847969055176,
+      "learning_rate": 2.938933529953094e-05,
+      "loss": 0.7145,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9310526315789474,
+      "grad_norm": 1.7286664247512817,
+      "learning_rate": 2.9388633149334777e-05,
+      "loss": 0.1145,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9315789473684211,
+      "grad_norm": 6.341940879821777,
+      "learning_rate": 2.9387930604097205e-05,
+      "loss": 2.5274,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9321052631578948,
+      "grad_norm": 1.1878658533096313,
+      "learning_rate": 2.938722766383751e-05,
+      "loss": 0.5424,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9326315789473684,
+      "grad_norm": 3.3078019618988037,
+      "learning_rate": 2.9386524328575003e-05,
+      "loss": 0.2485,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9331578947368421,
+      "grad_norm": 1.6980602741241455,
+      "learning_rate": 2.938582059832899e-05,
+      "loss": 1.1616,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9336842105263158,
+      "grad_norm": 3.3054215908050537,
+      "learning_rate": 2.9385116473118785e-05,
+      "loss": 1.3687,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9342105263157895,
+      "grad_norm": 3.2444324493408203,
+      "learning_rate": 2.9384411952963724e-05,
+      "loss": 1.7401,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9347368421052632,
+      "grad_norm": 1.1763970851898193,
+      "learning_rate": 2.9383707037883153e-05,
+      "loss": 1.25,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9352631578947368,
+      "grad_norm": 1.8481565713882446,
+      "learning_rate": 2.938300172789642e-05,
+      "loss": 0.904,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9357894736842105,
+      "grad_norm": 1.5394039154052734,
+      "learning_rate": 2.9382296023022895e-05,
+      "loss": 1.2871,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9363157894736842,
+      "grad_norm": 2.9562089443206787,
+      "learning_rate": 2.9381589923281952e-05,
+      "loss": 0.9019,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9368421052631579,
+      "grad_norm": 2.452955961227417,
+      "learning_rate": 2.9380883428692972e-05,
+      "loss": 0.6759,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9373684210526316,
+      "grad_norm": 5.483946800231934,
+      "learning_rate": 2.9380176539275355e-05,
+      "loss": 1.4204,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9378947368421052,
+      "grad_norm": 9.330098152160645,
+      "learning_rate": 2.9379469255048513e-05,
+      "loss": 2.1166,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9384210526315789,
+      "grad_norm": 5.44893741607666,
+      "learning_rate": 2.9378761576031858e-05,
+      "loss": 1.667,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9389473684210526,
+      "grad_norm": 21.2198543548584,
+      "learning_rate": 2.937805350224482e-05,
+      "loss": 1.3647,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9394736842105263,
+      "grad_norm": 1.1891385316848755,
+      "learning_rate": 2.9377345033706843e-05,
+      "loss": 1.1004,
+      "step": 1785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.092072010040283,
+      "learning_rate": 2.937663617043738e-05,
+      "loss": 0.3973,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9405263157894737,
+      "grad_norm": 4.912240505218506,
+      "learning_rate": 2.9375926912455884e-05,
+      "loss": 0.4901,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9410526315789474,
+      "grad_norm": 1.5781091451644897,
+      "learning_rate": 2.9375217259781833e-05,
+      "loss": 1.2677,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9415789473684211,
+      "grad_norm": 16.19452667236328,
+      "learning_rate": 2.937450721243471e-05,
+      "loss": 1.1125,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9421052631578948,
+      "grad_norm": 1.818876028060913,
+      "learning_rate": 2.9373796770434015e-05,
+      "loss": 1.052,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9426315789473684,
+      "grad_norm": 1.8112890720367432,
+      "learning_rate": 2.9373085933799242e-05,
+      "loss": 0.8413,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9431578947368421,
+      "grad_norm": 0.8109291195869446,
+      "learning_rate": 2.937237470254992e-05,
+      "loss": 0.4724,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9436842105263158,
+      "grad_norm": 1.2814196348190308,
+      "learning_rate": 2.9371663076705566e-05,
+      "loss": 1.2765,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9442105263157895,
+      "grad_norm": 47.906097412109375,
+      "learning_rate": 2.937095105628572e-05,
+      "loss": 2.2085,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9447368421052632,
+      "grad_norm": 1.119391918182373,
+      "learning_rate": 2.937023864130993e-05,
+      "loss": 1.2151,
+      "step": 1795
+    },
+    {
+      "epoch": 0.9452631578947368,
+      "grad_norm": 2.7573742866516113,
+      "learning_rate": 2.936952583179776e-05,
+      "loss": 1.4973,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9457894736842105,
+      "grad_norm": 1.0621460676193237,
+      "learning_rate": 2.9368812627768777e-05,
+      "loss": 1.7011,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9463157894736842,
+      "grad_norm": 1.82633638381958,
+      "learning_rate": 2.9368099029242564e-05,
+      "loss": 0.6421,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9468421052631579,
+      "grad_norm": 2.238363742828369,
+      "learning_rate": 2.9367385036238707e-05,
+      "loss": 1.4712,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 1.1244910955429077,
+      "learning_rate": 2.9366670648776818e-05,
+      "loss": 0.7498,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9478947368421052,
+      "grad_norm": 2.550694704055786,
+      "learning_rate": 2.9365955866876503e-05,
+      "loss": 0.8432,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9484210526315789,
+      "grad_norm": 0.9619981050491333,
+      "learning_rate": 2.9365240690557387e-05,
+      "loss": 0.9428,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9489473684210527,
+      "grad_norm": 3.0906801223754883,
+      "learning_rate": 2.9364525119839107e-05,
+      "loss": 0.254,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9494736842105264,
+      "grad_norm": 17.610000610351562,
+      "learning_rate": 2.936380915474131e-05,
+      "loss": 2.4892,
+      "step": 1804
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.0261893272399902,
+      "learning_rate": 2.9363092795283654e-05,
+      "loss": 1.0611,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9505263157894737,
+      "grad_norm": 1.3554835319519043,
+      "learning_rate": 2.9362376041485807e-05,
+      "loss": 0.9824,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9510526315789474,
+      "grad_norm": 1.1118865013122559,
+      "learning_rate": 2.936165889336744e-05,
+      "loss": 0.9515,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9515789473684211,
+      "grad_norm": 9.150864601135254,
+      "learning_rate": 2.936094135094825e-05,
+      "loss": 0.5341,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9521052631578948,
+      "grad_norm": 3.7893803119659424,
+      "learning_rate": 2.936022341424794e-05,
+      "loss": 1.7552,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9526315789473684,
+      "grad_norm": 1.512068271636963,
+      "learning_rate": 2.9359505083286215e-05,
+      "loss": 1.0817,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9531578947368421,
+      "grad_norm": 2.110664129257202,
+      "learning_rate": 2.935878635808279e-05,
+      "loss": 1.1328,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9536842105263158,
+      "grad_norm": 6.667463779449463,
+      "learning_rate": 2.9358067238657414e-05,
+      "loss": 1.9689,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9542105263157895,
+      "grad_norm": 8.632956504821777,
+      "learning_rate": 2.9357347725029814e-05,
+      "loss": 0.961,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9547368421052631,
+      "grad_norm": 1.0220603942871094,
+      "learning_rate": 2.935662781721976e-05,
+      "loss": 0.7937,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9552631578947368,
+      "grad_norm": 1.1829009056091309,
+      "learning_rate": 2.9355907515247008e-05,
+      "loss": 0.5874,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9557894736842105,
+      "grad_norm": 1.6473325490951538,
+      "learning_rate": 2.9355186819131334e-05,
+      "loss": 1.3057,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9563157894736842,
+      "grad_norm": 1.158306360244751,
+      "learning_rate": 2.9354465728892528e-05,
+      "loss": 0.0314,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9568421052631579,
+      "grad_norm": 10.786005973815918,
+      "learning_rate": 2.9353744244550382e-05,
+      "loss": 2.2784,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9573684210526315,
+      "grad_norm": 7.915785312652588,
+      "learning_rate": 2.935302236612471e-05,
+      "loss": 0.8294,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9578947368421052,
+      "grad_norm": 1.1970584392547607,
+      "learning_rate": 2.9352300093635335e-05,
+      "loss": 1.3563,
+      "step": 1820
+    },
+    {
+      "epoch": 0.958421052631579,
+      "grad_norm": 1.5558509826660156,
+      "learning_rate": 2.9351577427102075e-05,
+      "loss": 1.0525,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9589473684210527,
+      "grad_norm": 0.9184319972991943,
+      "learning_rate": 2.935085436654478e-05,
+      "loss": 0.668,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9594736842105264,
+      "grad_norm": 1.1146199703216553,
+      "learning_rate": 2.93501309119833e-05,
+      "loss": 1.2854,
+      "step": 1823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9025349617004395,
+      "learning_rate": 2.9349407063437496e-05,
+      "loss": 0.8864,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9605263157894737,
+      "grad_norm": 6.680036544799805,
+      "learning_rate": 2.934868282092724e-05,
+      "loss": 1.3019,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9610526315789474,
+      "grad_norm": 2.815171241760254,
+      "learning_rate": 2.934795818447242e-05,
+      "loss": 0.3377,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9615789473684211,
+      "grad_norm": 1.282981276512146,
+      "learning_rate": 2.934723315409293e-05,
+      "loss": 1.0778,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9621052631578947,
+      "grad_norm": 6.144766330718994,
+      "learning_rate": 2.9346507729808676e-05,
+      "loss": 0.1954,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9626315789473684,
+      "grad_norm": 1.2909563779830933,
+      "learning_rate": 2.9345781911639576e-05,
+      "loss": 1.141,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9631578947368421,
+      "grad_norm": 10.239936828613281,
+      "learning_rate": 2.9345055699605546e-05,
+      "loss": 3.1448,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9636842105263158,
+      "grad_norm": 3.629612922668457,
+      "learning_rate": 2.9344329093726542e-05,
+      "loss": 0.5131,
+      "step": 1831
+    },
+    {
+      "epoch": 0.9642105263157895,
+      "grad_norm": 1.5074138641357422,
+      "learning_rate": 2.93436020940225e-05,
+      "loss": 1.7405,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9647368421052631,
+      "grad_norm": 2.027520179748535,
+      "learning_rate": 2.9342874700513387e-05,
+      "loss": 0.6478,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9652631578947368,
+      "grad_norm": 2.1978070735931396,
+      "learning_rate": 2.934214691321917e-05,
+      "loss": 0.7294,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9657894736842105,
+      "grad_norm": 2.9730889797210693,
+      "learning_rate": 2.9341418732159826e-05,
+      "loss": 0.7626,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9663157894736842,
+      "grad_norm": 1.2914584875106812,
+      "learning_rate": 2.9340690157355358e-05,
+      "loss": 1.3947,
+      "step": 1836
+    },
+    {
+      "epoch": 0.966842105263158,
+      "grad_norm": 7.379958629608154,
+      "learning_rate": 2.9339961188825765e-05,
+      "loss": 2.4647,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9673684210526315,
+      "grad_norm": 2.926400899887085,
+      "learning_rate": 2.9339231826591057e-05,
+      "loss": 0.5296,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9678947368421053,
+      "grad_norm": 1.7127336263656616,
+      "learning_rate": 2.9338502070671258e-05,
+      "loss": 0.8867,
+      "step": 1839
+    },
+    {
+      "epoch": 0.968421052631579,
+      "grad_norm": 5.599946975708008,
+      "learning_rate": 2.933777192108641e-05,
+      "loss": 1.0288,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9689473684210527,
+      "grad_norm": 1.7110023498535156,
+      "learning_rate": 2.9337041377856562e-05,
+      "loss": 1.5123,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9694736842105263,
+      "grad_norm": 6.067218780517578,
+      "learning_rate": 2.9336310441001757e-05,
+      "loss": 1.374,
+      "step": 1842
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 13.339165687561035,
+      "learning_rate": 2.9335579110542075e-05,
+      "loss": 2.0621,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9705263157894737,
+      "grad_norm": 1.5139007568359375,
+      "learning_rate": 2.9334847386497587e-05,
+      "loss": 1.1468,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9710526315789474,
+      "grad_norm": 2.751978874206543,
+      "learning_rate": 2.9334115268888392e-05,
+      "loss": 1.4863,
+      "step": 1845
+    },
+    {
+      "epoch": 0.9715789473684211,
+      "grad_norm": 0.912716805934906,
+      "learning_rate": 2.933338275773458e-05,
+      "loss": 0.575,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9721052631578947,
+      "grad_norm": 6.437097072601318,
+      "learning_rate": 2.933264985305627e-05,
+      "loss": 1.4573,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9726315789473684,
+      "grad_norm": 2.406679391860962,
+      "learning_rate": 2.933191655487358e-05,
+      "loss": 1.426,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9731578947368421,
+      "grad_norm": 1.209294319152832,
+      "learning_rate": 2.9331182863206647e-05,
+      "loss": 1.1083,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 1.4055339097976685,
+      "learning_rate": 2.933044877807561e-05,
+      "loss": 1.1797,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9742105263157895,
+      "grad_norm": 0.443554550409317,
+      "learning_rate": 2.9329714299500624e-05,
+      "loss": 0.0177,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9747368421052631,
+      "grad_norm": 1.1881070137023926,
+      "learning_rate": 2.9328979427501854e-05,
+      "loss": 1.2222,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9752631578947368,
+      "grad_norm": 1.7724149227142334,
+      "learning_rate": 2.9328244162099475e-05,
+      "loss": 1.0751,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9757894736842105,
+      "grad_norm": 3.7436909675598145,
+      "learning_rate": 2.932750850331368e-05,
+      "loss": 0.5386,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9763157894736842,
+      "grad_norm": 1.2736313343048096,
+      "learning_rate": 2.932677245116466e-05,
+      "loss": 1.151,
+      "step": 1855
+    },
+    {
+      "epoch": 0.9768421052631578,
+      "grad_norm": 10.034126281738281,
+      "learning_rate": 2.932603600567263e-05,
+      "loss": 0.5635,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9773684210526316,
+      "grad_norm": 1.5949655771255493,
+      "learning_rate": 2.9325299166857802e-05,
+      "loss": 1.3225,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9778947368421053,
+      "grad_norm": 5.0193376541137695,
+      "learning_rate": 2.9324561934740407e-05,
+      "loss": 0.7483,
+      "step": 1858
+    },
+    {
+      "epoch": 0.978421052631579,
+      "grad_norm": 1.0509535074234009,
+      "learning_rate": 2.932382430934069e-05,
+      "loss": 0.9718,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9789473684210527,
+      "grad_norm": 2.3995654582977295,
+      "learning_rate": 2.9323086290678897e-05,
+      "loss": 0.7383,
+      "step": 1860
+    },
+    {
+      "epoch": 0.9794736842105263,
+      "grad_norm": 1.2402573823928833,
+      "learning_rate": 2.9322347878775294e-05,
+      "loss": 1.2607,
+      "step": 1861
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.9786499738693237,
+      "learning_rate": 2.9321609073650157e-05,
+      "loss": 1.065,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9805263157894737,
+      "grad_norm": 6.080226898193359,
+      "learning_rate": 2.9320869875323767e-05,
+      "loss": 0.3911,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9810526315789474,
+      "grad_norm": 5.234240531921387,
+      "learning_rate": 2.9320130283816417e-05,
+      "loss": 0.958,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9815789473684211,
+      "grad_norm": 48.88657760620117,
+      "learning_rate": 2.9319390299148417e-05,
+      "loss": 2.8281,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9821052631578947,
+      "grad_norm": 1.2519406080245972,
+      "learning_rate": 2.9318649921340076e-05,
+      "loss": 0.5246,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9826315789473684,
+      "grad_norm": 2.0961368083953857,
+      "learning_rate": 2.931790915041173e-05,
+      "loss": 1.0672,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9831578947368421,
+      "grad_norm": 3.2622575759887695,
+      "learning_rate": 2.9317167986383705e-05,
+      "loss": 0.9251,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9836842105263158,
+      "grad_norm": 1.264971375465393,
+      "learning_rate": 2.9316426429276366e-05,
+      "loss": 0.2829,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9842105263157894,
+      "grad_norm": 16.512882232666016,
+      "learning_rate": 2.9315684479110062e-05,
+      "loss": 1.0379,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9847368421052631,
+      "grad_norm": 3.8302979469299316,
+      "learning_rate": 2.931494213590516e-05,
+      "loss": 0.7015,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9852631578947368,
+      "grad_norm": 2.289673089981079,
+      "learning_rate": 2.9314199399682053e-05,
+      "loss": 1.1638,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9857894736842105,
+      "grad_norm": 1.4307857751846313,
+      "learning_rate": 2.9313456270461123e-05,
+      "loss": 0.0371,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9863157894736843,
+      "grad_norm": 4.602456569671631,
+      "learning_rate": 2.9312712748262774e-05,
+      "loss": 2.1234,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9868421052631579,
+      "grad_norm": 2.3993990421295166,
+      "learning_rate": 2.9311968833107423e-05,
+      "loss": 0.5891,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9873684210526316,
+      "grad_norm": 31.689926147460938,
+      "learning_rate": 2.9311224525015494e-05,
+      "loss": 1.496,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9878947368421053,
+      "grad_norm": 2.8490824699401855,
+      "learning_rate": 2.9310479824007416e-05,
+      "loss": 0.3687,
+      "step": 1877
+    },
+    {
+      "epoch": 0.988421052631579,
+      "grad_norm": 1.2507219314575195,
+      "learning_rate": 2.930973473010365e-05,
+      "loss": 0.9543,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9889473684210527,
+      "grad_norm": 2.8183679580688477,
+      "learning_rate": 2.930898924332463e-05,
+      "loss": 0.1816,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9894736842105263,
+      "grad_norm": 1.2869808673858643,
+      "learning_rate": 2.9308243363690844e-05,
+      "loss": 0.5637,
+      "step": 1880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.482398986816406,
+      "learning_rate": 2.9307497091222753e-05,
+      "loss": 0.2045,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9905263157894737,
+      "grad_norm": 4.65263557434082,
+      "learning_rate": 2.930675042594086e-05,
+      "loss": 1.2288,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9910526315789474,
+      "grad_norm": 1.563451886177063,
+      "learning_rate": 2.9306003367865662e-05,
+      "loss": 1.599,
+      "step": 1883
+    },
+    {
+      "epoch": 0.991578947368421,
+      "grad_norm": 1.2689063549041748,
+      "learning_rate": 2.9305255917017665e-05,
+      "loss": 1.0688,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9921052631578947,
+      "grad_norm": 1.954530119895935,
+      "learning_rate": 2.930450807341739e-05,
+      "loss": 0.2562,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9926315789473684,
+      "grad_norm": 9.376235008239746,
+      "learning_rate": 2.9303759837085375e-05,
+      "loss": 0.9329,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9931578947368421,
+      "grad_norm": 4.2806477546691895,
+      "learning_rate": 2.9303011208042158e-05,
+      "loss": 1.4794,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9936842105263158,
+      "grad_norm": 3.1993582248687744,
+      "learning_rate": 2.9302262186308297e-05,
+      "loss": 0.9944,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9942105263157894,
+      "grad_norm": 1.6474570035934448,
+      "learning_rate": 2.9301512771904347e-05,
+      "loss": 1.0777,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9947368421052631,
+      "grad_norm": 1.2916871309280396,
+      "learning_rate": 2.9300762964850895e-05,
+      "loss": 1.3304,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9952631578947368,
+      "grad_norm": 1.1107717752456665,
+      "learning_rate": 2.930001276516852e-05,
+      "loss": 1.1093,
+      "step": 1891
+    },
+    {
+      "epoch": 0.9957894736842106,
+      "grad_norm": 1.7554007768630981,
+      "learning_rate": 2.9299262172877817e-05,
+      "loss": 1.1471,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9963157894736843,
+      "grad_norm": 4.788026332855225,
+      "learning_rate": 2.9298511187999404e-05,
+      "loss": 0.49,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9968421052631579,
+      "grad_norm": 5.148394584655762,
+      "learning_rate": 2.929775981055389e-05,
+      "loss": 0.2753,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9973684210526316,
+      "grad_norm": 1.212622046470642,
+      "learning_rate": 2.9297008040561907e-05,
+      "loss": 1.0885,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9978947368421053,
+      "grad_norm": 1.1759798526763916,
+      "learning_rate": 2.9296255878044094e-05,
+      "loss": 0.8806,
+      "step": 1896
+    },
+    {
+      "epoch": 0.998421052631579,
+      "grad_norm": 1.2078953981399536,
+      "learning_rate": 2.9295503323021103e-05,
+      "loss": 1.0733,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9989473684210526,
+      "grad_norm": 5.538965225219727,
+      "learning_rate": 2.9294750375513598e-05,
+      "loss": 2.357,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9994736842105263,
+      "grad_norm": 0.9785202741622925,
+      "learning_rate": 2.9293997035542244e-05,
+      "loss": 0.7298,
+      "step": 1899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.185262680053711,
+      "learning_rate": 2.929324330312773e-05,
+      "loss": 1.5887,
+      "step": 1900
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.0015404224395752,
+      "eval_runtime": 12.876,
+      "eval_samples_per_second": 7.766,
+      "eval_steps_per_second": 7.766,
+      "step": 1900
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 19000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 1900,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6932541497704448e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}