{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9152,
  "eval_steps": 500,
  "global_step": 300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0064,
      "grad_norm": 2.0177674293518066,
      "learning_rate": 0.001,
      "loss": 7.5774,
      "mean_token_accuracy": 0.5871803313493729,
      "step": 1
    },
    {
      "epoch": 0.0128,
      "grad_norm": 2.9027979373931885,
      "learning_rate": 0.001,
      "loss": 8.8307,
      "mean_token_accuracy": 0.5205357447266579,
      "step": 2
    },
    {
      "epoch": 0.0192,
      "grad_norm": 20.20247459411621,
      "learning_rate": 0.001,
      "loss": 10.8145,
      "mean_token_accuracy": 0.46558595448732376,
      "step": 3
    },
    {
      "epoch": 0.0256,
      "grad_norm": NaN,
      "learning_rate": 0.001,
      "loss": 12.6533,
      "mean_token_accuracy": 0.4072755202651024,
      "step": 4
    },
    {
      "epoch": 0.032,
      "grad_norm": NaN,
      "learning_rate": 0.001,
      "loss": 13.6561,
      "mean_token_accuracy": 0.37730202078819275,
      "step": 5
    },
    {
      "epoch": 0.0384,
      "grad_norm": 43.733421325683594,
      "learning_rate": 0.001,
      "loss": 12.4744,
      "mean_token_accuracy": 0.40100549906492233,
      "step": 6
    },
    {
      "epoch": 0.0448,
      "grad_norm": 19.141199111938477,
      "learning_rate": 0.001,
      "loss": 9.5344,
      "mean_token_accuracy": 0.5192811861634254,
      "step": 7
    },
    {
      "epoch": 0.0512,
      "grad_norm": 18.41023826599121,
      "learning_rate": 0.001,
      "loss": 14.3608,
      "mean_token_accuracy": 0.43483711034059525,
      "step": 8
    },
    {
      "epoch": 0.0576,
      "grad_norm": 43.658721923828125,
      "learning_rate": 0.001,
      "loss": 13.3078,
      "mean_token_accuracy": 0.41958054155111313,
      "step": 9
    },
    {
      "epoch": 0.064,
      "grad_norm": 8.837448120117188,
      "learning_rate": 0.001,
      "loss": 10.5372,
      "mean_token_accuracy": 0.47571663558483124,
      "step": 10
    },
    {
      "epoch": 0.0704,
      "grad_norm": 10.096052169799805,
      "learning_rate": 0.001,
      "loss": 12.1162,
      "mean_token_accuracy": 0.4255446493625641,
      "step": 11
    },
    {
      "epoch": 0.0768,
      "grad_norm": 5.100619316101074,
      "learning_rate": 0.001,
      "loss": 9.2243,
      "mean_token_accuracy": 0.5292309895157814,
      "step": 12
    },
    {
      "epoch": 0.0832,
      "grad_norm": 4.010665416717529,
      "learning_rate": 0.001,
      "loss": 6.7791,
      "mean_token_accuracy": 0.6034770011901855,
      "step": 13
    },
    {
      "epoch": 0.0896,
      "grad_norm": 7.71261739730835,
      "learning_rate": 0.001,
      "loss": 8.4717,
      "mean_token_accuracy": 0.519000381231308,
      "step": 14
    },
    {
      "epoch": 0.096,
      "grad_norm": 6.087244987487793,
      "learning_rate": 0.001,
      "loss": 8.6032,
      "mean_token_accuracy": 0.5275572314858437,
      "step": 15
    },
    {
      "epoch": 0.1024,
      "grad_norm": 3.9907681941986084,
      "learning_rate": 0.001,
      "loss": 7.6546,
      "mean_token_accuracy": 0.542157307267189,
      "step": 16
    },
    {
      "epoch": 0.1088,
      "grad_norm": 4.168624401092529,
      "learning_rate": 0.001,
      "loss": 7.6228,
      "mean_token_accuracy": 0.5621763542294502,
      "step": 17
    },
    {
      "epoch": 0.1152,
      "grad_norm": 4.9508490562438965,
      "learning_rate": 0.001,
      "loss": 9.808,
      "mean_token_accuracy": 0.49128710478544235,
      "step": 18
    },
    {
      "epoch": 0.1216,
      "grad_norm": 4.520578861236572,
      "learning_rate": 0.001,
      "loss": 8.8476,
      "mean_token_accuracy": 0.4943431541323662,
      "step": 19
    },
    {
      "epoch": 0.128,
      "grad_norm": 6.03663444519043,
      "learning_rate": 0.001,
      "loss": 8.0179,
      "mean_token_accuracy": 0.5464938580989838,
      "step": 20
    },
    {
      "epoch": 0.1344,
      "grad_norm": 4.153735637664795,
      "learning_rate": 0.001,
      "loss": 8.7043,
      "mean_token_accuracy": 0.5475813895463943,
      "step": 21
    },
    {
      "epoch": 0.1408,
      "grad_norm": 3.6078875064849854,
      "learning_rate": 0.001,
      "loss": 8.5597,
      "mean_token_accuracy": 0.5447330102324486,
      "step": 22
    },
    {
      "epoch": 0.1472,
      "grad_norm": 3.8478240966796875,
      "learning_rate": 0.001,
      "loss": 7.2146,
      "mean_token_accuracy": 0.5992539674043655,
      "step": 23
    },
    {
      "epoch": 0.1536,
      "grad_norm": 4.033797264099121,
      "learning_rate": 0.001,
      "loss": 9.3191,
      "mean_token_accuracy": 0.5196451023221016,
      "step": 24
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.2458839416503906,
      "learning_rate": 0.001,
      "loss": 7.4417,
      "mean_token_accuracy": 0.5597032383084297,
      "step": 25
    },
    {
      "epoch": 0.1664,
      "grad_norm": 3.520216226577759,
      "learning_rate": 0.001,
      "loss": 6.7451,
      "mean_token_accuracy": 0.5860115587711334,
      "step": 26
    },
    {
      "epoch": 0.1728,
      "grad_norm": 3.7371273040771484,
      "learning_rate": 0.001,
      "loss": 7.8899,
      "mean_token_accuracy": 0.5652190744876862,
      "step": 27
    },
    {
      "epoch": 0.1792,
      "grad_norm": 4.060925006866455,
      "learning_rate": 0.001,
      "loss": 7.4009,
      "mean_token_accuracy": 0.5925501734018326,
      "step": 28
    },
    {
      "epoch": 0.1856,
      "grad_norm": 3.9062178134918213,
      "learning_rate": 0.001,
      "loss": 7.3145,
      "mean_token_accuracy": 0.5906639099121094,
      "step": 29
    },
    {
      "epoch": 0.192,
      "grad_norm": 3.7855491638183594,
      "learning_rate": 0.001,
      "loss": 7.9822,
      "mean_token_accuracy": 0.5379130840301514,
      "step": 30
    },
    {
      "epoch": 0.1984,
      "grad_norm": 3.340970516204834,
      "learning_rate": 0.001,
      "loss": 7.1766,
      "mean_token_accuracy": 0.5813343971967697,
      "step": 31
    },
    {
      "epoch": 0.2048,
      "grad_norm": 3.692194938659668,
      "learning_rate": 0.001,
      "loss": 7.6622,
      "mean_token_accuracy": 0.5705007463693619,
      "step": 32
    },
    {
      "epoch": 0.2112,
      "grad_norm": 3.3015754222869873,
      "learning_rate": 0.001,
      "loss": 7.1023,
      "mean_token_accuracy": 0.5972073376178741,
      "step": 33
    },
    {
      "epoch": 0.2176,
      "grad_norm": 3.191516160964966,
      "learning_rate": 0.001,
      "loss": 7.8106,
      "mean_token_accuracy": 0.577879324555397,
      "step": 34
    },
    {
      "epoch": 0.224,
      "grad_norm": 3.015953779220581,
      "learning_rate": 0.001,
      "loss": 8.2667,
      "mean_token_accuracy": 0.5295384302735329,
      "step": 35
    },
    {
      "epoch": 0.2304,
      "grad_norm": 3.0506298542022705,
      "learning_rate": 0.001,
      "loss": 8.2168,
      "mean_token_accuracy": 0.5699651688337326,
      "step": 36
    },
    {
      "epoch": 0.2368,
      "grad_norm": 3.11960506439209,
      "learning_rate": 0.001,
      "loss": 7.697,
      "mean_token_accuracy": 0.5628390982747078,
      "step": 37
    },
    {
      "epoch": 0.2432,
      "grad_norm": 3.0201468467712402,
      "learning_rate": 0.001,
      "loss": 7.6835,
      "mean_token_accuracy": 0.5989912003278732,
      "step": 38
    },
    {
      "epoch": 0.2496,
      "grad_norm": 3.462604522705078,
      "learning_rate": 0.001,
      "loss": 7.8899,
      "mean_token_accuracy": 0.5595831423997879,
      "step": 39
    },
    {
      "epoch": 0.256,
      "grad_norm": 3.401752233505249,
      "learning_rate": 0.001,
      "loss": 7.9503,
      "mean_token_accuracy": 0.5780320316553116,
      "step": 40
    },
    {
      "epoch": 0.2624,
      "grad_norm": 4.185473442077637,
      "learning_rate": 0.001,
      "loss": 7.799,
      "mean_token_accuracy": 0.5548681393265724,
      "step": 41
    },
    {
      "epoch": 0.2688,
      "grad_norm": 3.9332938194274902,
      "learning_rate": 0.001,
      "loss": 7.7152,
      "mean_token_accuracy": 0.5636641383171082,
      "step": 42
    },
    {
      "epoch": 0.2752,
      "grad_norm": 3.2568907737731934,
      "learning_rate": 0.001,
      "loss": 7.4339,
      "mean_token_accuracy": 0.5839549452066422,
      "step": 43
    },
    {
      "epoch": 0.2816,
      "grad_norm": 4.672369956970215,
      "learning_rate": 0.001,
      "loss": 9.5802,
      "mean_token_accuracy": 0.5100178346037865,
      "step": 44
    },
    {
      "epoch": 0.288,
      "grad_norm": 2.7107110023498535,
      "learning_rate": 0.001,
      "loss": 6.8152,
      "mean_token_accuracy": 0.6055551916360855,
      "step": 45
    },
    {
      "epoch": 0.2944,
      "grad_norm": 4.762430667877197,
      "learning_rate": 0.001,
      "loss": 9.5548,
      "mean_token_accuracy": 0.5057197883725166,
      "step": 46
    },
    {
      "epoch": 0.3008,
      "grad_norm": 3.7424633502960205,
      "learning_rate": 0.001,
      "loss": 8.5479,
      "mean_token_accuracy": 0.5115512683987617,
      "step": 47
    },
    {
      "epoch": 0.3072,
      "grad_norm": 3.2961721420288086,
      "learning_rate": 0.001,
      "loss": 8.2113,
      "mean_token_accuracy": 0.557342454791069,
      "step": 48
    },
    {
      "epoch": 0.3136,
      "grad_norm": 3.239579677581787,
      "learning_rate": 0.001,
      "loss": 8.8304,
      "mean_token_accuracy": 0.5229588970541954,
      "step": 49
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.273914337158203,
      "learning_rate": 0.001,
      "loss": 7.7483,
      "mean_token_accuracy": 0.5654094517230988,
      "step": 50
    },
    {
      "epoch": 0.3264,
      "grad_norm": 3.6419663429260254,
      "learning_rate": 0.001,
      "loss": 7.3461,
      "mean_token_accuracy": 0.6018975228071213,
      "step": 51
    },
    {
      "epoch": 0.3328,
      "grad_norm": 3.4327375888824463,
      "learning_rate": 0.001,
      "loss": 7.6107,
      "mean_token_accuracy": 0.5776806026697159,
      "step": 52
    },
    {
      "epoch": 0.3392,
      "grad_norm": 2.8938372135162354,
      "learning_rate": 0.001,
      "loss": 7.7927,
      "mean_token_accuracy": 0.5748834684491158,
      "step": 53
    },
    {
      "epoch": 0.3456,
      "grad_norm": 3.780987024307251,
      "learning_rate": 0.001,
      "loss": 7.7148,
      "mean_token_accuracy": 0.5800818204879761,
      "step": 54
    },
    {
      "epoch": 0.352,
      "grad_norm": 7.018247604370117,
      "learning_rate": 0.001,
      "loss": 7.9333,
      "mean_token_accuracy": 0.557623416185379,
      "step": 55
    },
    {
      "epoch": 0.3584,
      "grad_norm": 4.366345405578613,
      "learning_rate": 0.001,
      "loss": 8.7168,
      "mean_token_accuracy": 0.5106043368577957,
      "step": 56
    },
    {
      "epoch": 0.3648,
      "grad_norm": 3.5044593811035156,
      "learning_rate": 0.001,
      "loss": 7.3359,
      "mean_token_accuracy": 0.5744027197360992,
      "step": 57
    },
    {
      "epoch": 0.3712,
      "grad_norm": 3.710293769836426,
      "learning_rate": 0.001,
      "loss": 7.1611,
      "mean_token_accuracy": 0.6022080183029175,
      "step": 58
    },
    {
      "epoch": 0.3776,
      "grad_norm": 3.784938097000122,
      "learning_rate": 0.001,
      "loss": 7.0631,
      "mean_token_accuracy": 0.5821704566478729,
      "step": 59
    },
    {
      "epoch": 0.384,
      "grad_norm": 3.79630708694458,
      "learning_rate": 0.001,
      "loss": 6.6392,
      "mean_token_accuracy": 0.6132841110229492,
      "step": 60
    },
    {
      "epoch": 0.3904,
      "grad_norm": 3.004554510116577,
      "learning_rate": 0.001,
      "loss": 7.1997,
      "mean_token_accuracy": 0.5954362899065018,
      "step": 61
    },
    {
      "epoch": 0.3968,
      "grad_norm": 4.015055179595947,
      "learning_rate": 0.001,
      "loss": 7.4768,
      "mean_token_accuracy": 0.5736093372106552,
      "step": 62
    },
    {
      "epoch": 0.4032,
      "grad_norm": 3.654245615005493,
      "learning_rate": 0.001,
      "loss": 7.8212,
      "mean_token_accuracy": 0.5659709572792053,
      "step": 63
    },
    {
      "epoch": 0.4096,
      "grad_norm": 4.0627641677856445,
      "learning_rate": 0.001,
      "loss": 9.2678,
      "mean_token_accuracy": 0.5201347693800926,
      "step": 64
    },
    {
      "epoch": 0.416,
      "grad_norm": 3.4690985679626465,
      "learning_rate": 0.001,
      "loss": 8.8517,
      "mean_token_accuracy": 0.5343466401100159,
      "step": 65
    },
    {
      "epoch": 0.4224,
      "grad_norm": 3.752857208251953,
      "learning_rate": 0.001,
      "loss": 8.8329,
      "mean_token_accuracy": 0.5228310152888298,
      "step": 66
    },
    {
      "epoch": 0.4288,
      "grad_norm": 2.968993902206421,
      "learning_rate": 0.001,
      "loss": 7.6058,
      "mean_token_accuracy": 0.5344163477420807,
      "step": 67
    },
    {
      "epoch": 0.4352,
      "grad_norm": 3.72929310798645,
      "learning_rate": 0.001,
      "loss": 9.3222,
      "mean_token_accuracy": 0.516074076294899,
      "step": 68
    },
    {
      "epoch": 0.4416,
      "grad_norm": 3.6030168533325195,
      "learning_rate": 0.001,
      "loss": 8.5513,
      "mean_token_accuracy": 0.5421805381774902,
      "step": 69
    },
    {
      "epoch": 0.448,
      "grad_norm": 3.1739273071289062,
      "learning_rate": 0.001,
      "loss": 7.0802,
      "mean_token_accuracy": 0.5998114049434662,
      "step": 70
    },
    {
      "epoch": 0.4544,
      "grad_norm": 2.880340099334717,
      "learning_rate": 0.001,
      "loss": 6.7046,
      "mean_token_accuracy": 0.6100719571113586,
      "step": 71
    },
    {
      "epoch": 0.4608,
      "grad_norm": 3.2981016635894775,
      "learning_rate": 0.001,
      "loss": 7.9998,
      "mean_token_accuracy": 0.5657886415719986,
      "step": 72
    },
    {
      "epoch": 0.4672,
      "grad_norm": 4.456881999969482,
      "learning_rate": 0.001,
      "loss": 7.1571,
      "mean_token_accuracy": 0.5802148729562759,
      "step": 73
    },
    {
      "epoch": 0.4736,
      "grad_norm": 4.069066047668457,
      "learning_rate": 0.001,
      "loss": 8.8054,
      "mean_token_accuracy": 0.5348180085420609,
      "step": 74
    },
    {
      "epoch": 0.48,
      "grad_norm": 9.803370475769043,
      "learning_rate": 0.001,
      "loss": 8.5662,
      "mean_token_accuracy": 0.5325330048799515,
      "step": 75
    },
    {
      "epoch": 0.4864,
      "grad_norm": 4.328041076660156,
      "learning_rate": 0.001,
      "loss": 8.1187,
      "mean_token_accuracy": 0.5573177188634872,
      "step": 76
    },
    {
      "epoch": 0.4928,
      "grad_norm": 5.3780975341796875,
      "learning_rate": 0.001,
      "loss": 8.1414,
      "mean_token_accuracy": 0.5307755619287491,
      "step": 77
    },
    {
      "epoch": 0.4992,
      "grad_norm": 4.601280212402344,
      "learning_rate": 0.001,
      "loss": 8.1343,
      "mean_token_accuracy": 0.5730244666337967,
      "step": 78
    },
    {
      "epoch": 0.5056,
      "grad_norm": 5.185049533843994,
      "learning_rate": 0.001,
      "loss": 10.3809,
      "mean_token_accuracy": 0.45638662576675415,
      "step": 79
    },
    {
      "epoch": 0.512,
      "grad_norm": 4.1511054039001465,
      "learning_rate": 0.001,
      "loss": 7.903,
      "mean_token_accuracy": 0.5588131695985794,
      "step": 80
    },
    {
      "epoch": 0.5184,
      "grad_norm": 3.8643414974212646,
      "learning_rate": 0.001,
      "loss": 9.1992,
      "mean_token_accuracy": 0.526216022670269,
      "step": 81
    },
    {
      "epoch": 0.5248,
      "grad_norm": 3.9921276569366455,
      "learning_rate": 0.001,
      "loss": 7.9791,
      "mean_token_accuracy": 0.5335483700037003,
      "step": 82
    },
    {
      "epoch": 0.5312,
      "grad_norm": 3.2328104972839355,
      "learning_rate": 0.001,
      "loss": 8.3164,
      "mean_token_accuracy": 0.5265945047140121,
      "step": 83
    },
    {
      "epoch": 0.5376,
      "grad_norm": 4.729416370391846,
      "learning_rate": 0.001,
      "loss": 8.9703,
      "mean_token_accuracy": 0.5294319912791252,
      "step": 84
    },
    {
      "epoch": 0.544,
      "grad_norm": 3.562079906463623,
      "learning_rate": 0.001,
      "loss": 8.2047,
      "mean_token_accuracy": 0.557565338909626,
      "step": 85
    },
    {
      "epoch": 0.5504,
      "grad_norm": 3.8010246753692627,
      "learning_rate": 0.001,
      "loss": 8.0639,
      "mean_token_accuracy": 0.5428211987018585,
      "step": 86
    },
    {
      "epoch": 0.5568,
      "grad_norm": 3.3087170124053955,
      "learning_rate": 0.001,
      "loss": 8.145,
      "mean_token_accuracy": 0.5458494797348976,
      "step": 87
    },
    {
      "epoch": 0.5632,
      "grad_norm": 3.956712245941162,
      "learning_rate": 0.001,
      "loss": 8.6958,
      "mean_token_accuracy": 0.5472606420516968,
      "step": 88
    },
    {
      "epoch": 0.5696,
      "grad_norm": 3.548285722732544,
      "learning_rate": 0.001,
      "loss": 7.0706,
      "mean_token_accuracy": 0.605229914188385,
      "step": 89
    },
    {
      "epoch": 0.576,
      "grad_norm": 3.6893436908721924,
      "learning_rate": 0.001,
      "loss": 6.2735,
      "mean_token_accuracy": 0.6206476241350174,
      "step": 90
    },
    {
      "epoch": 0.5824,
      "grad_norm": 8.876976013183594,
      "learning_rate": 0.001,
      "loss": 9.1893,
      "mean_token_accuracy": 0.5234974399209023,
      "step": 91
    },
    {
      "epoch": 0.5888,
      "grad_norm": 4.515777587890625,
      "learning_rate": 0.001,
      "loss": 7.1549,
      "mean_token_accuracy": 0.5581575930118561,
      "step": 92
    },
    {
      "epoch": 0.5952,
      "grad_norm": 3.814516067504883,
      "learning_rate": 0.001,
      "loss": 7.7558,
      "mean_token_accuracy": 0.572398416697979,
      "step": 93
    },
    {
      "epoch": 0.6016,
      "grad_norm": 4.637458801269531,
      "learning_rate": 0.001,
      "loss": 8.2918,
      "mean_token_accuracy": 0.5480805337429047,
      "step": 94
    },
    {
      "epoch": 0.608,
      "grad_norm": 3.827664613723755,
      "learning_rate": 0.001,
      "loss": 7.6153,
      "mean_token_accuracy": 0.5540450513362885,
      "step": 95
    },
    {
      "epoch": 0.6144,
      "grad_norm": 3.596336841583252,
      "learning_rate": 0.001,
      "loss": 8.1713,
      "mean_token_accuracy": 0.5356313809752464,
      "step": 96
    },
    {
      "epoch": 0.6208,
      "grad_norm": 8.448469161987305,
      "learning_rate": 0.001,
      "loss": 8.668,
      "mean_token_accuracy": 0.5483140274882317,
      "step": 97
    },
    {
      "epoch": 0.6272,
      "grad_norm": 5.3135576248168945,
      "learning_rate": 0.001,
      "loss": 7.7838,
      "mean_token_accuracy": 0.5676964446902275,
      "step": 98
    },
    {
      "epoch": 0.6336,
      "grad_norm": 4.343587398529053,
      "learning_rate": 0.001,
      "loss": 7.2509,
      "mean_token_accuracy": 0.5730399489402771,
      "step": 99
    },
    {
      "epoch": 0.64,
      "grad_norm": 3.066798686981201,
      "learning_rate": 0.001,
      "loss": 7.0591,
      "mean_token_accuracy": 0.6044407039880753,
      "step": 100
    },
    {
      "epoch": 0.6464,
      "grad_norm": 4.594414234161377,
      "learning_rate": 0.001,
      "loss": 7.8146,
      "mean_token_accuracy": 0.5625345483422279,
      "step": 101
    },
    {
      "epoch": 0.6528,
      "grad_norm": 7.372500419616699,
      "learning_rate": 0.001,
      "loss": 7.9078,
      "mean_token_accuracy": 0.5716247260570526,
      "step": 102
    },
    {
      "epoch": 0.6592,
      "grad_norm": 4.589946269989014,
      "learning_rate": 0.001,
      "loss": 7.0762,
      "mean_token_accuracy": 0.5866946280002594,
      "step": 103
    },
    {
      "epoch": 0.6656,
      "grad_norm": 8.939701080322266,
      "learning_rate": 0.001,
      "loss": 8.9255,
      "mean_token_accuracy": 0.528026632964611,
      "step": 104
    },
    {
      "epoch": 0.672,
      "grad_norm": 4.3290534019470215,
      "learning_rate": 0.001,
      "loss": 8.1217,
      "mean_token_accuracy": 0.5670952945947647,
      "step": 105
    },
    {
      "epoch": 0.6784,
      "grad_norm": 3.7743752002716064,
      "learning_rate": 0.001,
      "loss": 7.4548,
      "mean_token_accuracy": 0.5619177967309952,
      "step": 106
    },
    {
      "epoch": 0.6848,
      "grad_norm": 4.942656993865967,
      "learning_rate": 0.001,
      "loss": 8.1344,
      "mean_token_accuracy": 0.565968431532383,
      "step": 107
    },
    {
      "epoch": 0.6912,
      "grad_norm": 4.2500715255737305,
      "learning_rate": 0.001,
      "loss": 7.5658,
      "mean_token_accuracy": 0.5492838993668556,
      "step": 108
    },
    {
      "epoch": 0.6976,
      "grad_norm": 3.6989309787750244,
      "learning_rate": 0.001,
      "loss": 7.7848,
      "mean_token_accuracy": 0.5549194514751434,
      "step": 109
    },
    {
      "epoch": 0.704,
      "grad_norm": 4.687801837921143,
      "learning_rate": 0.001,
      "loss": 8.1714,
      "mean_token_accuracy": 0.5541109219193459,
      "step": 110
    },
    {
      "epoch": 0.7104,
      "grad_norm": 3.4812569618225098,
      "learning_rate": 0.001,
      "loss": 9.1957,
      "mean_token_accuracy": 0.5032462030649185,
      "step": 111
    },
    {
      "epoch": 0.7168,
      "grad_norm": 4.627752780914307,
      "learning_rate": 0.001,
      "loss": 9.3127,
      "mean_token_accuracy": 0.5073471590876579,
      "step": 112
    },
    {
      "epoch": 0.7232,
      "grad_norm": 3.9376790523529053,
      "learning_rate": 0.001,
      "loss": 8.562,
      "mean_token_accuracy": 0.5392563939094543,
      "step": 113
    },
    {
      "epoch": 0.7296,
      "grad_norm": 3.162787675857544,
      "learning_rate": 0.001,
      "loss": 7.3301,
      "mean_token_accuracy": 0.5914620906114578,
      "step": 114
    },
    {
      "epoch": 0.736,
      "grad_norm": 4.065870761871338,
      "learning_rate": 0.001,
      "loss": 7.881,
      "mean_token_accuracy": 0.5708122700452805,
      "step": 115
    },
    {
      "epoch": 0.7424,
      "grad_norm": 4.9694318771362305,
      "learning_rate": 0.001,
      "loss": 7.6188,
      "mean_token_accuracy": 0.5728907287120819,
      "step": 116
    },
    {
      "epoch": 0.7488,
      "grad_norm": 3.600388526916504,
      "learning_rate": 0.001,
      "loss": 6.7958,
      "mean_token_accuracy": 0.6140304803848267,
      "step": 117
    },
    {
      "epoch": 0.7552,
      "grad_norm": 4.772489070892334,
      "learning_rate": 0.001,
      "loss": 7.8477,
      "mean_token_accuracy": 0.5720326602458954,
      "step": 118
    },
    {
      "epoch": 0.7616,
      "grad_norm": 4.109669208526611,
      "learning_rate": 0.001,
      "loss": 7.8857,
      "mean_token_accuracy": 0.5725124627351761,
      "step": 119
    },
    {
      "epoch": 0.768,
      "grad_norm": 3.9959371089935303,
      "learning_rate": 0.001,
      "loss": 7.8924,
      "mean_token_accuracy": 0.5729866176843643,
      "step": 120
    },
    {
      "epoch": 0.7744,
      "grad_norm": 4.218033790588379,
      "learning_rate": 0.001,
      "loss": 8.5045,
      "mean_token_accuracy": 0.5448136478662491,
      "step": 121
    },
    {
      "epoch": 0.7808,
      "grad_norm": 10.680689811706543,
      "learning_rate": 0.001,
      "loss": 8.5677,
      "mean_token_accuracy": 0.5437836945056915,
      "step": 122
    },
    {
      "epoch": 0.7872,
      "grad_norm": 6.1264729499816895,
      "learning_rate": 0.001,
      "loss": 10.1013,
      "mean_token_accuracy": 0.4813794791698456,
      "step": 123
    },
    {
      "epoch": 0.7936,
      "grad_norm": 3.4104678630828857,
      "learning_rate": 0.001,
      "loss": 7.5739,
      "mean_token_accuracy": 0.5576579570770264,
      "step": 124
    },
    {
      "epoch": 0.8,
      "grad_norm": 4.8629255294799805,
      "learning_rate": 0.001,
      "loss": 8.2406,
      "mean_token_accuracy": 0.5386597216129303,
      "step": 125
    },
    {
      "epoch": 0.8064,
      "grad_norm": 4.2899627685546875,
      "learning_rate": 0.001,
      "loss": 8.4591,
      "mean_token_accuracy": 0.5415358692407608,
      "step": 126
    },
    {
      "epoch": 0.8128,
      "grad_norm": 3.7481234073638916,
      "learning_rate": 0.001,
      "loss": 7.0793,
      "mean_token_accuracy": 0.6038873642683029,
      "step": 127
    },
    {
      "epoch": 0.8192,
      "grad_norm": 3.3412671089172363,
      "learning_rate": 0.001,
      "loss": 6.6192,
      "mean_token_accuracy": 0.6217593550682068,
      "step": 128
    },
    {
      "epoch": 0.8256,
      "grad_norm": 4.920543670654297,
      "learning_rate": 0.001,
      "loss": 8.4534,
      "mean_token_accuracy": 0.5469236895442009,
      "step": 129
    },
    {
      "epoch": 0.832,
      "grad_norm": 3.746335029602051,
      "learning_rate": 0.001,
      "loss": 8.2707,
      "mean_token_accuracy": 0.5559498444199562,
      "step": 130
    },
    {
      "epoch": 0.8384,
      "grad_norm": 3.816923141479492,
      "learning_rate": 0.001,
      "loss": 7.3012,
      "mean_token_accuracy": 0.5551306903362274,
      "step": 131
    },
    {
      "epoch": 0.8448,
      "grad_norm": 5.7253265380859375,
      "learning_rate": 0.001,
      "loss": 7.7264,
      "mean_token_accuracy": 0.5539910942316055,
      "step": 132
    },
    {
      "epoch": 0.8512,
      "grad_norm": 9.863517761230469,
      "learning_rate": 0.001,
      "loss": 8.4424,
      "mean_token_accuracy": 0.551963597536087,
      "step": 133
    },
    {
      "epoch": 0.8576,
      "grad_norm": 5.7852582931518555,
      "learning_rate": 0.001,
      "loss": 8.8394,
      "mean_token_accuracy": 0.503598153591156,
      "step": 134
    },
    {
      "epoch": 0.864,
      "grad_norm": 4.214358806610107,
      "learning_rate": 0.001,
      "loss": 8.0501,
      "mean_token_accuracy": 0.578537181019783,
      "step": 135
    },
    {
      "epoch": 0.8704,
      "grad_norm": 4.677027225494385,
      "learning_rate": 0.001,
      "loss": 8.2782,
      "mean_token_accuracy": 0.556526243686676,
      "step": 136
    },
    {
      "epoch": 0.8768,
      "grad_norm": 4.3732590675354,
      "learning_rate": 0.001,
      "loss": 7.1854,
      "mean_token_accuracy": 0.5627769976854324,
      "step": 137
    },
    {
      "epoch": 0.8832,
      "grad_norm": 4.285269260406494,
      "learning_rate": 0.001,
      "loss": 7.5983,
      "mean_token_accuracy": 0.5761594474315643,
      "step": 138
    },
    {
      "epoch": 0.8896,
      "grad_norm": 4.189344882965088,
      "learning_rate": 0.001,
      "loss": 7.796,
      "mean_token_accuracy": 0.5710103362798691,
      "step": 139
    },
    {
      "epoch": 0.896,
      "grad_norm": 5.359838008880615,
      "learning_rate": 0.001,
      "loss": 9.3741,
      "mean_token_accuracy": 0.522984579205513,
      "step": 140
    },
    {
      "epoch": 0.9024,
      "grad_norm": 4.133533000946045,
      "learning_rate": 0.001,
      "loss": 8.7714,
      "mean_token_accuracy": 0.49868279695510864,
      "step": 141
    },
    {
      "epoch": 0.9088,
      "grad_norm": 3.432337522506714,
      "learning_rate": 0.001,
      "loss": 7.1933,
      "mean_token_accuracy": 0.5911879688501358,
      "step": 142
    },
    {
      "epoch": 0.9152,
      "grad_norm": 3.4508094787597656,
      "learning_rate": 0.001,
      "loss": 8.4234,
      "mean_token_accuracy": 0.544566810131073,
      "step": 143
    },
    {
      "epoch": 0.9216,
      "grad_norm": 4.863613605499268,
      "learning_rate": 0.001,
      "loss": 8.2328,
      "mean_token_accuracy": 0.5418634340167046,
      "step": 144
    },
    {
      "epoch": 0.928,
      "grad_norm": 3.8848912715911865,
      "learning_rate": 0.001,
      "loss": 8.2236,
      "mean_token_accuracy": 0.5698046088218689,
      "step": 145
    },
    {
      "epoch": 0.9344,
      "grad_norm": NaN,
      "learning_rate": 0.001,
      "loss": 8.5748,
      "mean_token_accuracy": 0.5264320075511932,
      "step": 146
    },
    {
      "epoch": 0.9408,
      "grad_norm": 4.209794998168945,
      "learning_rate": 0.001,
      "loss": 8.6393,
      "mean_token_accuracy": 0.5443025454878807,
      "step": 147
    },
    {
      "epoch": 0.9472,
      "grad_norm": 5.81450891494751,
      "learning_rate": 0.001,
      "loss": 8.0271,
      "mean_token_accuracy": 0.5531649142503738,
      "step": 148
    },
    {
      "epoch": 0.9536,
      "grad_norm": 3.847665786743164,
      "learning_rate": 0.001,
      "loss": 7.4199,
      "mean_token_accuracy": 0.5528692901134491,
      "step": 149
    },
    {
      "epoch": 0.96,
      "grad_norm": 4.067963123321533,
      "learning_rate": 0.001,
      "loss": 6.8948,
      "mean_token_accuracy": 0.6041955649852753,
      "step": 150
    },
    {
      "epoch": 0.9664,
      "grad_norm": 3.6745011806488037,
      "learning_rate": 0.001,
      "loss": 7.916,
      "mean_token_accuracy": 0.5488733500242233,
      "step": 151
    },
    {
      "epoch": 0.9728,
      "grad_norm": 3.931896686553955,
      "learning_rate": 0.001,
      "loss": 7.7169,
      "mean_token_accuracy": 0.5718288719654083,
      "step": 152
    },
    {
      "epoch": 0.9792,
      "grad_norm": 4.551197052001953,
      "learning_rate": 0.001,
      "loss": 8.1617,
      "mean_token_accuracy": 0.5373415797948837,
      "step": 153
    },
    {
      "epoch": 0.9856,
      "grad_norm": 4.994777679443359,
      "learning_rate": 0.001,
      "loss": 6.9665,
      "mean_token_accuracy": 0.5668836236000061,
      "step": 154
    },
    {
      "epoch": 0.992,
      "grad_norm": 4.934599876403809,
      "learning_rate": 0.001,
      "loss": 8.9109,
      "mean_token_accuracy": 0.5285608246922493,
      "step": 155
    },
    {
      "epoch": 0.9984,
      "grad_norm": 4.362490653991699,
      "learning_rate": 0.001,
      "loss": 7.9886,
      "mean_token_accuracy": 0.5652178972959518,
      "step": 156
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.0191662311553955,
      "learning_rate": 0.001,
      "loss": 2.2094,
      "mean_token_accuracy": 0.49253731966018677,
      "step": 157
    },
    {
      "epoch": 1.0064,
      "grad_norm": 4.179709434509277,
      "learning_rate": 0.001,
      "loss": 6.0564,
      "mean_token_accuracy": 0.6441845297813416,
      "step": 158
    },
    {
      "epoch": 1.0128,
      "grad_norm": 4.1790971755981445,
      "learning_rate": 0.001,
      "loss": 7.223,
      "mean_token_accuracy": 0.5850173979997635,
      "step": 159
    },
    {
      "epoch": 1.0192,
      "grad_norm": 3.393885374069214,
      "learning_rate": 0.001,
      "loss": 5.8742,
      "mean_token_accuracy": 0.6555072665214539,
      "step": 160
    },
    {
      "epoch": 1.0256,
      "grad_norm": 4.170093536376953,
      "learning_rate": 0.001,
      "loss": 6.4492,
      "mean_token_accuracy": 0.6223380863666534,
      "step": 161
    },
    {
      "epoch": 1.032,
      "grad_norm": 4.816079616546631,
      "learning_rate": 0.001,
      "loss": 8.458,
      "mean_token_accuracy": 0.5561687871813774,
      "step": 162
    },
    {
      "epoch": 1.0384,
      "grad_norm": 7.0680108070373535,
      "learning_rate": 0.001,
      "loss": 6.4478,
      "mean_token_accuracy": 0.6458811461925507,
      "step": 163
    },
    {
      "epoch": 1.0448,
      "grad_norm": 5.175418376922607,
      "learning_rate": 0.001,
      "loss": 8.6144,
      "mean_token_accuracy": 0.5237443894147873,
      "step": 164
    },
    {
      "epoch": 1.0512,
      "grad_norm": 3.455251693725586,
      "learning_rate": 0.001,
      "loss": 6.6271,
      "mean_token_accuracy": 0.6214667707681656,
      "step": 165
    },
    {
      "epoch": 1.0576,
      "grad_norm": 4.4586992263793945,
      "learning_rate": 0.001,
      "loss": 5.9971,
      "mean_token_accuracy": 0.654425784945488,
      "step": 166
    },
    {
      "epoch": 1.064,
      "grad_norm": 4.797828197479248,
      "learning_rate": 0.001,
      "loss": 7.2682,
      "mean_token_accuracy": 0.5919231176376343,
      "step": 167
    },
    {
      "epoch": 1.0704,
      "grad_norm": 4.61244010925293,
      "learning_rate": 0.001,
      "loss": 6.1862,
      "mean_token_accuracy": 0.6452524065971375,
      "step": 168
    },
    {
      "epoch": 1.0768,
      "grad_norm": 5.201780796051025,
      "learning_rate": 0.001,
      "loss": 9.1169,
      "mean_token_accuracy": 0.5475455224514008,
      "step": 169
    },
    {
      "epoch": 1.0832,
      "grad_norm": 5.166975975036621,
      "learning_rate": 0.001,
      "loss": 6.9593,
      "mean_token_accuracy": 0.5929352939128876,
      "step": 170
    },
    {
      "epoch": 1.0896,
      "grad_norm": 5.824632167816162,
      "learning_rate": 0.001,
      "loss": 8.294,
      "mean_token_accuracy": 0.5294871926307678,
      "step": 171
    },
    {
      "epoch": 1.096,
      "grad_norm": 9.257010459899902,
      "learning_rate": 0.001,
      "loss": 7.0368,
      "mean_token_accuracy": 0.6035382002592087,
      "step": 172
    },
    {
      "epoch": 1.1024,
      "grad_norm": 5.524269104003906,
      "learning_rate": 0.001,
      "loss": 5.6529,
      "mean_token_accuracy": 0.6525093466043472,
      "step": 173
    },
    {
      "epoch": 1.1088,
      "grad_norm": 5.32748556137085,
      "learning_rate": 0.001,
      "loss": 6.7001,
      "mean_token_accuracy": 0.6103013008832932,
      "step": 174
    },
    {
      "epoch": 1.1152,
      "grad_norm": 5.178861141204834,
      "learning_rate": 0.001,
      "loss": 6.7637,
      "mean_token_accuracy": 0.6111182868480682,
      "step": 175
    },
    {
      "epoch": 1.1216,
      "grad_norm": 3.9681689739227295,
      "learning_rate": 0.001,
      "loss": 6.0235,
      "mean_token_accuracy": 0.6353351175785065,
      "step": 176
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 4.23559045791626,
      "learning_rate": 0.001,
      "loss": 6.2243,
      "mean_token_accuracy": 0.636825680732727,
      "step": 177
    },
    {
      "epoch": 1.1344,
      "grad_norm": 9.25437068939209,
      "learning_rate": 0.001,
      "loss": 7.3467,
      "mean_token_accuracy": 0.6091018617153168,
      "step": 178
    },
    {
      "epoch": 1.1408,
      "grad_norm": 6.518167018890381,
      "learning_rate": 0.001,
      "loss": 7.4647,
      "mean_token_accuracy": 0.5769879817962646,
      "step": 179
    },
    {
      "epoch": 1.1472,
      "grad_norm": 8.98256778717041,
      "learning_rate": 0.001,
      "loss": 8.0621,
      "mean_token_accuracy": 0.5615440681576729,
      "step": 180
    },
    {
      "epoch": 1.1536,
      "grad_norm": 6.962597370147705,
      "learning_rate": 0.001,
      "loss": 7.0736,
      "mean_token_accuracy": 0.6006975322961807,
      "step": 181
    },
    {
      "epoch": 1.16,
      "grad_norm": 15.468770027160645,
      "learning_rate": 0.001,
      "loss": 7.8873,
      "mean_token_accuracy": 0.5628447085618973,
      "step": 182
    },
    {
      "epoch": 1.1663999999999999,
      "grad_norm": 6.4656596183776855,
      "learning_rate": 0.001,
      "loss": 7.7006,
      "mean_token_accuracy": 0.5857141762971878,
      "step": 183
    },
    {
      "epoch": 1.1728,
      "grad_norm": 6.264251708984375,
      "learning_rate": 0.001,
      "loss": 8.5055,
      "mean_token_accuracy": 0.5526725947856903,
      "step": 184
    },
    {
      "epoch": 1.1792,
      "grad_norm": 9.173776626586914,
      "learning_rate": 0.001,
      "loss": 8.3511,
      "mean_token_accuracy": 0.5329839885234833,
      "step": 185
    },
    {
      "epoch": 1.1856,
      "grad_norm": 6.497045993804932,
      "learning_rate": 0.001,
      "loss": 8.5874,
      "mean_token_accuracy": 0.5194432288408279,
      "step": 186
    },
    {
      "epoch": 1.192,
      "grad_norm": 6.55042839050293,
      "learning_rate": 0.001,
      "loss": 7.2464,
      "mean_token_accuracy": 0.5852644890546799,
      "step": 187
    },
    {
      "epoch": 1.1984,
      "grad_norm": 4.928357124328613,
      "learning_rate": 0.001,
      "loss": 7.421,
      "mean_token_accuracy": 0.575689509510994,
      "step": 188
    },
    {
      "epoch": 1.2048,
      "grad_norm": 6.555416107177734,
      "learning_rate": 0.001,
      "loss": 7.1958,
      "mean_token_accuracy": 0.6004299223423004,
      "step": 189
    },
    {
      "epoch": 1.2112,
      "grad_norm": 6.944830417633057,
      "learning_rate": 0.001,
      "loss": 6.8997,
      "mean_token_accuracy": 0.5822109580039978,
      "step": 190
    },
    {
      "epoch": 1.2176,
      "grad_norm": 8.618417739868164,
      "learning_rate": 0.001,
      "loss": 8.1635,
      "mean_token_accuracy": 0.5582073777914047,
      "step": 191
    },
    {
      "epoch": 1.224,
      "grad_norm": 8.295172691345215,
      "learning_rate": 0.001,
      "loss": 7.5801,
      "mean_token_accuracy": 0.5919029116630554,
      "step": 192
    },
    {
      "epoch": 1.2304,
      "grad_norm": 12.239678382873535,
      "learning_rate": 0.001,
      "loss": 7.3438,
      "mean_token_accuracy": 0.5995124354958534,
      "step": 193
    },
    {
      "epoch": 1.2368000000000001,
      "grad_norm": 7.0297441482543945,
      "learning_rate": 0.001,
      "loss": 8.2068,
      "mean_token_accuracy": 0.537700891494751,
      "step": 194
    },
    {
      "epoch": 1.2432,
      "grad_norm": 5.0657639503479,
      "learning_rate": 0.001,
      "loss": 7.3465,
      "mean_token_accuracy": 0.587388165295124,
      "step": 195
    },
    {
      "epoch": 1.2496,
      "grad_norm": 5.155961990356445,
      "learning_rate": 0.001,
      "loss": 7.4988,
      "mean_token_accuracy": 0.6019035428762436,
      "step": 196
    },
    {
      "epoch": 1.256,
      "grad_norm": 3.893186569213867,
      "learning_rate": 0.001,
      "loss": 5.796,
      "mean_token_accuracy": 0.624050185084343,
      "step": 197
    },
    {
      "epoch": 1.2624,
      "grad_norm": 7.488790512084961,
      "learning_rate": 0.001,
      "loss": 8.4914,
      "mean_token_accuracy": 0.5403816178441048,
      "step": 198
    },
    {
      "epoch": 1.2688,
      "grad_norm": 10.15825080871582,
      "learning_rate": 0.001,
      "loss": 6.9975,
      "mean_token_accuracy": 0.5812895894050598,
      "step": 199
    },
    {
      "epoch": 1.2752,
      "grad_norm": 5.484641075134277,
      "learning_rate": 0.001,
      "loss": 6.5434,
      "mean_token_accuracy": 0.6408251523971558,
      "step": 200
    },
    {
      "epoch": 1.2816,
      "grad_norm": 6.395635604858398,
      "learning_rate": 0.001,
      "loss": 7.7083,
      "mean_token_accuracy": 0.5760865062475204,
      "step": 201
    },
    {
      "epoch": 1.288,
      "grad_norm": 7.382782936096191,
      "learning_rate": 0.001,
      "loss": 7.147,
      "mean_token_accuracy": 0.5934395641088486,
      "step": 202
    },
    {
      "epoch": 1.2944,
      "grad_norm": 6.834714412689209,
      "learning_rate": 0.001,
      "loss": 7.6222,
      "mean_token_accuracy": 0.5739501044154167,
      "step": 203
    },
    {
      "epoch": 1.3008,
      "grad_norm": 5.695088863372803,
      "learning_rate": 0.001,
      "loss": 7.475,
      "mean_token_accuracy": 0.5647561475634575,
      "step": 204
    },
    {
      "epoch": 1.3072,
      "grad_norm": 5.658081531524658,
      "learning_rate": 0.001,
      "loss": 6.6691,
      "mean_token_accuracy": 0.6017946600914001,
      "step": 205
    },
    {
      "epoch": 1.3136,
      "grad_norm": 9.058781623840332,
      "learning_rate": 0.001,
      "loss": 6.3451,
      "mean_token_accuracy": 0.6354028433561325,
      "step": 206
    },
    {
      "epoch": 1.32,
      "grad_norm": 6.336489200592041,
      "learning_rate": 0.001,
      "loss": 7.9945,
      "mean_token_accuracy": 0.5321162715554237,
      "step": 207
    },
    {
      "epoch": 1.3264,
      "grad_norm": 14.819890022277832,
      "learning_rate": 0.001,
      "loss": 9.4337,
      "mean_token_accuracy": 0.5192963108420372,
      "step": 208
    },
    {
      "epoch": 1.3328,
      "grad_norm": 6.626960754394531,
      "learning_rate": 0.001,
      "loss": 7.0263,
      "mean_token_accuracy": 0.6068531572818756,
      "step": 209
    },
    {
      "epoch": 1.3392,
      "grad_norm": 7.5550384521484375,
      "learning_rate": 0.001,
      "loss": 7.5964,
      "mean_token_accuracy": 0.5845440551638603,
      "step": 210
    },
    {
      "epoch": 1.3456000000000001,
      "grad_norm": 9.043644905090332,
      "learning_rate": 0.001,
      "loss": 8.9667,
      "mean_token_accuracy": 0.5143468156456947,
      "step": 211
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 15.724257469177246,
      "learning_rate": 0.001,
      "loss": 7.97,
      "mean_token_accuracy": 0.5721839368343353,
      "step": 212
    },
    {
      "epoch": 1.3584,
      "grad_norm": 12.023893356323242,
      "learning_rate": 0.001,
      "loss": 7.454,
      "mean_token_accuracy": 0.599329836666584,
      "step": 213
    },
    {
      "epoch": 1.3648,
      "grad_norm": 6.79151725769043,
      "learning_rate": 0.001,
      "loss": 7.7166,
      "mean_token_accuracy": 0.5553706511855125,
      "step": 214
    },
    {
      "epoch": 1.3712,
      "grad_norm": 6.837470054626465,
      "learning_rate": 0.001,
      "loss": 10.8317,
      "mean_token_accuracy": 0.4465588107705116,
      "step": 215
    },
    {
      "epoch": 1.3776,
      "grad_norm": 9.545637130737305,
      "learning_rate": 0.001,
      "loss": 6.7084,
      "mean_token_accuracy": 0.593764916062355,
      "step": 216
    },
    {
      "epoch": 1.384,
      "grad_norm": 5.631826877593994,
      "learning_rate": 0.001,
      "loss": 6.8109,
      "mean_token_accuracy": 0.5985722541809082,
      "step": 217
    },
    {
      "epoch": 1.3904,
      "grad_norm": 7.997573375701904,
      "learning_rate": 0.001,
      "loss": 7.6082,
      "mean_token_accuracy": 0.5720451474189758,
      "step": 218
    },
    {
      "epoch": 1.3968,
      "grad_norm": 10.023303985595703,
      "learning_rate": 0.001,
      "loss": 8.4064,
      "mean_token_accuracy": 0.5365717485547066,
      "step": 219
    },
    {
      "epoch": 1.4032,
      "grad_norm": 7.888483047485352,
      "learning_rate": 0.001,
      "loss": 8.3661,
      "mean_token_accuracy": 0.5367622226476669,
      "step": 220
    },
    {
      "epoch": 1.4096,
      "grad_norm": 7.387235164642334,
      "learning_rate": 0.001,
      "loss": 7.0045,
      "mean_token_accuracy": 0.5931309461593628,
      "step": 221
    },
    {
      "epoch": 1.416,
      "grad_norm": 4.908820629119873,
      "learning_rate": 0.001,
      "loss": 7.457,
      "mean_token_accuracy": 0.5676573291420937,
      "step": 222
    },
    {
      "epoch": 1.4224,
      "grad_norm": 7.541878700256348,
      "learning_rate": 0.001,
      "loss": 8.5576,
      "mean_token_accuracy": 0.5507887750864029,
      "step": 223
    },
    {
      "epoch": 1.4288,
      "grad_norm": 6.35562801361084,
      "learning_rate": 0.001,
      "loss": 7.1046,
      "mean_token_accuracy": 0.5813699215650558,
      "step": 224
    },
    {
      "epoch": 1.4352,
      "grad_norm": 5.955767631530762,
      "learning_rate": 0.001,
      "loss": 7.5295,
      "mean_token_accuracy": 0.5491400882601738,
      "step": 225
    },
    {
      "epoch": 1.4416,
      "grad_norm": 10.310933113098145,
      "learning_rate": 0.001,
      "loss": 7.4046,
      "mean_token_accuracy": 0.5788294672966003,
      "step": 226
    },
    {
      "epoch": 1.448,
      "grad_norm": 8.590178489685059,
      "learning_rate": 0.001,
      "loss": 6.6688,
      "mean_token_accuracy": 0.5879671573638916,
      "step": 227
    },
    {
      "epoch": 1.4544000000000001,
      "grad_norm": 7.05693244934082,
      "learning_rate": 0.001,
      "loss": 9.785,
      "mean_token_accuracy": 0.4796893522143364,
      "step": 228
    },
    {
      "epoch": 1.4607999999999999,
      "grad_norm": 7.907512187957764,
      "learning_rate": 0.001,
      "loss": 6.369,
      "mean_token_accuracy": 0.6156620383262634,
      "step": 229
    },
    {
      "epoch": 1.4672,
      "grad_norm": 5.780457973480225,
      "learning_rate": 0.001,
      "loss": 7.7141,
      "mean_token_accuracy": 0.5557262748479843,
      "step": 230
    },
    {
      "epoch": 1.4736,
      "grad_norm": 12.134108543395996,
      "learning_rate": 0.001,
      "loss": 8.9121,
      "mean_token_accuracy": 0.5156212523579597,
      "step": 231
    },
    {
      "epoch": 1.48,
      "grad_norm": 6.636194229125977,
      "learning_rate": 0.001,
      "loss": 7.12,
      "mean_token_accuracy": 0.6036765873432159,
      "step": 232
    },
    {
      "epoch": 1.4864,
      "grad_norm": 5.020055770874023,
      "learning_rate": 0.001,
      "loss": 6.3509,
      "mean_token_accuracy": 0.6170566231012344,
      "step": 233
    },
    {
      "epoch": 1.4928,
      "grad_norm": 10.214864730834961,
      "learning_rate": 0.001,
      "loss": 6.5362,
      "mean_token_accuracy": 0.6308254599571228,
      "step": 234
    },
    {
      "epoch": 1.4992,
      "grad_norm": 11.291118621826172,
      "learning_rate": 0.001,
      "loss": 8.2135,
      "mean_token_accuracy": 0.5482836216688156,
      "step": 235
    },
    {
      "epoch": 1.5056,
      "grad_norm": 10.337629318237305,
      "learning_rate": 0.001,
      "loss": 7.4203,
      "mean_token_accuracy": 0.5811362043023109,
      "step": 236
    },
    {
      "epoch": 1.512,
      "grad_norm": 8.213200569152832,
      "learning_rate": 0.001,
      "loss": 8.0874,
      "mean_token_accuracy": 0.5553788542747498,
      "step": 237
    },
    {
      "epoch": 1.5184,
      "grad_norm": 35.60185623168945,
      "learning_rate": 0.001,
      "loss": 8.3935,
      "mean_token_accuracy": 0.5098748654127121,
      "step": 238
    },
    {
      "epoch": 1.5248,
      "grad_norm": 10.800209045410156,
      "learning_rate": 0.001,
      "loss": 8.3645,
      "mean_token_accuracy": 0.5591541230678558,
      "step": 239
    },
    {
      "epoch": 1.5312000000000001,
      "grad_norm": 10.682641983032227,
      "learning_rate": 0.001,
      "loss": 8.0174,
      "mean_token_accuracy": 0.5395123660564423,
      "step": 240
    },
    {
      "epoch": 1.5375999999999999,
      "grad_norm": 16.261472702026367,
      "learning_rate": 0.001,
      "loss": 8.4602,
      "mean_token_accuracy": 0.5196909606456757,
      "step": 241
    },
    {
      "epoch": 1.544,
      "grad_norm": 7.711510181427002,
      "learning_rate": 0.001,
      "loss": 7.8776,
      "mean_token_accuracy": 0.5560385212302208,
      "step": 242
    },
    {
      "epoch": 1.5504,
      "grad_norm": 10.960959434509277,
      "learning_rate": 0.001,
      "loss": 8.9444,
      "mean_token_accuracy": 0.5182236731052399,
      "step": 243
    },
    {
      "epoch": 1.5568,
      "grad_norm": 7.446345806121826,
      "learning_rate": 0.001,
      "loss": 6.9821,
      "mean_token_accuracy": 0.6031645685434341,
      "step": 244
    },
    {
      "epoch": 1.5632000000000001,
      "grad_norm": 8.444716453552246,
      "learning_rate": 0.001,
      "loss": 7.3512,
      "mean_token_accuracy": 0.5782665312290192,
      "step": 245
    },
    {
      "epoch": 1.5695999999999999,
      "grad_norm": 11.063419342041016,
      "learning_rate": 0.001,
      "loss": 8.5629,
      "mean_token_accuracy": 0.5185030549764633,
      "step": 246
    },
    {
      "epoch": 1.576,
      "grad_norm": 8.18127727508545,
      "learning_rate": 0.001,
      "loss": 8.4838,
      "mean_token_accuracy": 0.5051439926028252,
      "step": 247
    },
    {
      "epoch": 1.5824,
      "grad_norm": 9.148998260498047,
      "learning_rate": 0.001,
      "loss": 8.8174,
      "mean_token_accuracy": 0.515631377696991,
      "step": 248
    },
    {
      "epoch": 1.5888,
      "grad_norm": 7.322795867919922,
      "learning_rate": 0.001,
      "loss": 7.4007,
      "mean_token_accuracy": 0.5654035359621048,
      "step": 249
    },
    {
      "epoch": 1.5952,
      "grad_norm": 9.227678298950195,
      "learning_rate": 0.001,
      "loss": 9.3753,
      "mean_token_accuracy": 0.5103882774710655,
      "step": 250
    },
    {
      "epoch": 1.6016,
      "grad_norm": 13.938738822937012,
      "learning_rate": 0.001,
      "loss": 7.1796,
      "mean_token_accuracy": 0.5777264684438705,
      "step": 251
    },
    {
      "epoch": 1.608,
      "grad_norm": 23.90799903869629,
      "learning_rate": 0.001,
      "loss": 7.985,
      "mean_token_accuracy": 0.5455044433474541,
      "step": 252
    },
    {
      "epoch": 1.6143999999999998,
      "grad_norm": 10.978887557983398,
      "learning_rate": 0.001,
      "loss": 7.6049,
      "mean_token_accuracy": 0.5702199786901474,
      "step": 253
    },
    {
      "epoch": 1.6208,
      "grad_norm": 8.08303165435791,
      "learning_rate": 0.001,
      "loss": 7.5714,
      "mean_token_accuracy": 0.5959209054708481,
      "step": 254
    },
    {
      "epoch": 1.6272,
      "grad_norm": 9.555181503295898,
      "learning_rate": 0.001,
      "loss": 7.6343,
      "mean_token_accuracy": 0.5681372657418251,
      "step": 255
    },
    {
      "epoch": 1.6336,
      "grad_norm": 8.243929862976074,
      "learning_rate": 0.001,
      "loss": 8.2715,
      "mean_token_accuracy": 0.5482152029871941,
      "step": 256
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 8.662081718444824,
      "learning_rate": 0.001,
      "loss": 7.8241,
      "mean_token_accuracy": 0.5627938807010651,
      "step": 257
    },
    {
      "epoch": 1.6463999999999999,
      "grad_norm": 8.32907772064209,
      "learning_rate": 0.001,
      "loss": 7.335,
      "mean_token_accuracy": 0.5363176316022873,
      "step": 258
    },
    {
      "epoch": 1.6528,
      "grad_norm": 11.057466506958008,
      "learning_rate": 0.001,
      "loss": 9.5491,
      "mean_token_accuracy": 0.5060033053159714,
      "step": 259
    },
    {
      "epoch": 1.6592,
      "grad_norm": 83.32975006103516,
      "learning_rate": 0.001,
      "loss": 9.0589,
      "mean_token_accuracy": 0.50930255651474,
      "step": 260
    },
    {
      "epoch": 1.6656,
      "grad_norm": 8.165995597839355,
      "learning_rate": 0.001,
      "loss": 7.9928,
      "mean_token_accuracy": 0.563197672367096,
      "step": 261
    },
    {
      "epoch": 1.6720000000000002,
      "grad_norm": 16.992876052856445,
      "learning_rate": 0.001,
      "loss": 9.8182,
      "mean_token_accuracy": 0.48126567155122757,
      "step": 262
    },
    {
      "epoch": 1.6784,
      "grad_norm": 8.095207214355469,
      "learning_rate": 0.001,
      "loss": 9.419,
      "mean_token_accuracy": 0.500243678689003,
      "step": 263
    },
    {
      "epoch": 1.6848,
      "grad_norm": 16.054994583129883,
      "learning_rate": 0.001,
      "loss": 8.7803,
      "mean_token_accuracy": 0.5126448944211006,
      "step": 264
    },
    {
      "epoch": 1.6912,
      "grad_norm": 16.552730560302734,
      "learning_rate": 0.001,
      "loss": 8.9847,
      "mean_token_accuracy": 0.5172990635037422,
      "step": 265
    },
    {
      "epoch": 1.6976,
      "grad_norm": 13.06109619140625,
      "learning_rate": 0.001,
      "loss": 10.1881,
      "mean_token_accuracy": 0.49380549043416977,
      "step": 266
    },
    {
      "epoch": 1.704,
      "grad_norm": 10.187677383422852,
      "learning_rate": 0.001,
      "loss": 8.4531,
      "mean_token_accuracy": 0.5305923670530319,
      "step": 267
    },
    {
      "epoch": 1.7104,
      "grad_norm": 8.00294017791748,
      "learning_rate": 0.001,
      "loss": 8.55,
      "mean_token_accuracy": 0.5253442898392677,
      "step": 268
    },
    {
      "epoch": 1.7168,
      "grad_norm": 8.169543266296387,
      "learning_rate": 0.001,
      "loss": 10.1152,
      "mean_token_accuracy": 0.4511877968907356,
      "step": 269
    },
    {
      "epoch": 1.7231999999999998,
      "grad_norm": 8.316620826721191,
      "learning_rate": 0.001,
      "loss": 7.8197,
      "mean_token_accuracy": 0.5667182505130768,
      "step": 270
    },
    {
      "epoch": 1.7296,
      "grad_norm": 7.95322847366333,
      "learning_rate": 0.001,
      "loss": 8.585,
      "mean_token_accuracy": 0.5186630189418793,
      "step": 271
    },
    {
      "epoch": 1.736,
      "grad_norm": 10.643982887268066,
      "learning_rate": 0.001,
      "loss": 8.1634,
      "mean_token_accuracy": 0.5534348115324974,
      "step": 272
    },
    {
      "epoch": 1.7424,
      "grad_norm": 11.432665824890137,
      "learning_rate": 0.001,
      "loss": 9.9191,
      "mean_token_accuracy": 0.4892067015171051,
      "step": 273
    },
    {
      "epoch": 1.7488000000000001,
      "grad_norm": 12.11987590789795,
      "learning_rate": 0.001,
      "loss": 7.9962,
      "mean_token_accuracy": 0.5544887632131577,
      "step": 274
    },
    {
      "epoch": 1.7551999999999999,
      "grad_norm": 16.141345977783203,
      "learning_rate": 0.001,
      "loss": 9.2237,
      "mean_token_accuracy": 0.5020880922675133,
      "step": 275
    },
    {
      "epoch": 1.7616,
      "grad_norm": 11.97499942779541,
      "learning_rate": 0.001,
      "loss": 9.4026,
      "mean_token_accuracy": 0.4923320710659027,
      "step": 276
    },
    {
      "epoch": 1.768,
      "grad_norm": 20.49795913696289,
      "learning_rate": 0.001,
      "loss": 8.6141,
      "mean_token_accuracy": 0.5215461701154709,
      "step": 277
    },
    {
      "epoch": 1.7744,
      "grad_norm": 20.775779724121094,
      "learning_rate": 0.001,
      "loss": 11.6113,
      "mean_token_accuracy": 0.4201728478074074,
      "step": 278
    },
    {
      "epoch": 1.7808000000000002,
      "grad_norm": 8.977204322814941,
      "learning_rate": 0.001,
      "loss": 8.8081,
      "mean_token_accuracy": 0.49830809235572815,
      "step": 279
    },
    {
      "epoch": 1.7872,
      "grad_norm": 11.948820114135742,
      "learning_rate": 0.001,
      "loss": 9.8821,
      "mean_token_accuracy": 0.49480871111154556,
      "step": 280
    },
    {
      "epoch": 1.7936,
      "grad_norm": 10.269512176513672,
      "learning_rate": 0.001,
      "loss": 8.2235,
      "mean_token_accuracy": 0.5259484350681305,
      "step": 281
    },
    {
      "epoch": 1.8,
      "grad_norm": 5.860231876373291,
      "learning_rate": 0.001,
      "loss": 8.3539,
      "mean_token_accuracy": 0.5288261473178864,
      "step": 282
    },
    {
      "epoch": 1.8064,
      "grad_norm": 9.967249870300293,
      "learning_rate": 0.001,
      "loss": 9.6113,
      "mean_token_accuracy": 0.45666269958019257,
      "step": 283
    },
    {
      "epoch": 1.8128,
      "grad_norm": 10.237951278686523,
      "learning_rate": 0.001,
      "loss": 7.8902,
      "mean_token_accuracy": 0.5173894092440605,
      "step": 284
    },
    {
      "epoch": 1.8192,
      "grad_norm": 14.716999053955078,
      "learning_rate": 0.001,
      "loss": 9.5343,
      "mean_token_accuracy": 0.5037772506475449,
      "step": 285
    },
    {
      "epoch": 1.8256000000000001,
      "grad_norm": 10.649497032165527,
      "learning_rate": 0.001,
      "loss": 9.8439,
      "mean_token_accuracy": 0.48058757930994034,
      "step": 286
    },
    {
      "epoch": 1.8319999999999999,
      "grad_norm": 15.107233047485352,
      "learning_rate": 0.001,
      "loss": 8.1109,
      "mean_token_accuracy": 0.5341241806745529,
      "step": 287
    },
    {
      "epoch": 1.8384,
      "grad_norm": 12.771748542785645,
      "learning_rate": 0.001,
      "loss": 9.2594,
      "mean_token_accuracy": 0.48890865594148636,
      "step": 288
    },
    {
      "epoch": 1.8448,
      "grad_norm": 9.82298469543457,
      "learning_rate": 0.001,
      "loss": 8.4437,
      "mean_token_accuracy": 0.5308887958526611,
      "step": 289
    },
    {
      "epoch": 1.8512,
      "grad_norm": 22.638954162597656,
      "learning_rate": 0.001,
      "loss": 8.8215,
      "mean_token_accuracy": 0.5280868858098984,
      "step": 290
    },
    {
      "epoch": 1.8576000000000001,
      "grad_norm": 10.145191192626953,
      "learning_rate": 0.001,
      "loss": 9.1644,
      "mean_token_accuracy": 0.5151188969612122,
      "step": 291
    },
    {
      "epoch": 1.8639999999999999,
      "grad_norm": 7.520384311676025,
      "learning_rate": 0.001,
      "loss": 8.3904,
      "mean_token_accuracy": 0.5109719932079315,
      "step": 292
    },
    {
      "epoch": 1.8704,
      "grad_norm": 16.988628387451172,
      "learning_rate": 0.001,
      "loss": 8.5715,
      "mean_token_accuracy": 0.5133798718452454,
      "step": 293
    },
    {
      "epoch": 1.8768,
      "grad_norm": 10.056252479553223,
      "learning_rate": 0.001,
      "loss": 8.0868,
      "mean_token_accuracy": 0.544953741133213,
      "step": 294
    },
    {
      "epoch": 1.8832,
      "grad_norm": 27.852039337158203,
      "learning_rate": 0.001,
      "loss": 6.7302,
      "mean_token_accuracy": 0.594126746058464,
      "step": 295
    },
    {
      "epoch": 1.8896,
      "grad_norm": 12.71786880493164,
      "learning_rate": 0.001,
      "loss": 9.0788,
      "mean_token_accuracy": 0.5247168093919754,
      "step": 296
    },
    {
      "epoch": 1.896,
      "grad_norm": 15.346375465393066,
      "learning_rate": 0.001,
      "loss": 8.3889,
      "mean_token_accuracy": 0.5260215997695923,
      "step": 297
    },
    {
      "epoch": 1.9024,
      "grad_norm": 23.141864776611328,
      "learning_rate": 0.001,
      "loss": 8.985,
      "mean_token_accuracy": 0.5044504255056381,
      "step": 298
    },
    {
      "epoch": 1.9088,
      "grad_norm": 13.86075210571289,
      "learning_rate": 0.001,
      "loss": 9.9779,
      "mean_token_accuracy": 0.4863489866256714,
      "step": 299
    },
    {
      "epoch": 1.9152,
      "grad_norm": 12.302156448364258,
      "learning_rate": 0.001,
      "loss": 10.3654,
      "mean_token_accuracy": 0.4447909966111183,
      "step": 300
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 20,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.069164560285696e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}