{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.16,
  "eval_steps": 1000,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008,
      "grad_norm": 0.0,
      "learning_rate": 0.0005,
      "loss": 9.9524,
      "step": 1
    },
    {
      "epoch": 0.0016,
      "grad_norm": 1.437297100892465e-06,
      "learning_rate": 0.0005,
      "loss": 9.9048,
      "step": 2
    },
    {
      "epoch": 0.0024,
      "grad_norm": 1.1676747798919678,
      "learning_rate": 0.0005,
      "loss": 9.8558,
      "step": 3
    },
    {
      "epoch": 0.0032,
      "grad_norm": 11.5746431350708,
      "learning_rate": 0.0005,
      "loss": 9.5539,
      "step": 4
    },
    {
      "epoch": 0.004,
      "grad_norm": 3.968721628189087,
      "learning_rate": 0.0005,
      "loss": 9.2858,
      "step": 5
    },
    {
      "epoch": 0.0048,
      "grad_norm": 6.475513935089111,
      "learning_rate": 0.0005,
      "loss": 8.3592,
      "step": 6
    },
    {
      "epoch": 0.0056,
      "grad_norm": 11.020626068115234,
      "learning_rate": 0.0005,
      "loss": 7.1505,
      "step": 7
    },
    {
      "epoch": 0.0064,
      "grad_norm": 16.665830612182617,
      "learning_rate": 0.0005,
      "loss": 3.6912,
      "step": 8
    },
    {
      "epoch": 0.0072,
      "grad_norm": 5.093105316162109,
      "learning_rate": 0.0005,
      "loss": 1.7986,
      "step": 9
    },
    {
      "epoch": 0.008,
      "grad_norm": 5.546494007110596,
      "learning_rate": 0.0005,
      "loss": 2.6467,
      "step": 10
    },
    {
      "epoch": 0.0088,
      "grad_norm": 20.05790138244629,
      "learning_rate": 0.0005,
      "loss": 0.9881,
      "step": 11
    },
    {
      "epoch": 0.0096,
      "grad_norm": 5.210952281951904,
      "learning_rate": 0.0005,
      "loss": 1.7098,
      "step": 12
    },
    {
      "epoch": 0.0104,
      "grad_norm": 7.218165397644043,
      "learning_rate": 0.0005,
      "loss": 1.7179,
      "step": 13
    },
    {
      "epoch": 0.0112,
      "grad_norm": 3.7357559204101562,
      "learning_rate": 0.0005,
      "loss": 0.7633,
      "step": 14
    },
    {
      "epoch": 0.012,
      "grad_norm": 5.467613220214844,
      "learning_rate": 0.0005,
      "loss": 1.2452,
      "step": 15
    },
    {
      "epoch": 0.0128,
      "grad_norm": 13.217657089233398,
      "learning_rate": 0.0005,
      "loss": 2.4923,
      "step": 16
    },
    {
      "epoch": 0.0136,
      "grad_norm": 1.690468430519104,
      "learning_rate": 0.0005,
      "loss": 0.9401,
      "step": 17
    },
    {
      "epoch": 0.0144,
      "grad_norm": 1.6662808656692505,
      "learning_rate": 0.0005,
      "loss": 1.7828,
      "step": 18
    },
    {
      "epoch": 0.0152,
      "grad_norm": 1.3934170007705688,
      "learning_rate": 0.0005,
      "loss": 1.5044,
      "step": 19
    },
    {
      "epoch": 0.016,
      "grad_norm": 1.0231975317001343,
      "learning_rate": 0.0005,
      "loss": 1.2216,
      "step": 20
    },
    {
      "epoch": 0.0168,
      "grad_norm": 1.9770835638046265,
      "learning_rate": 0.0005,
      "loss": 2.5877,
      "step": 21
    },
    {
      "epoch": 0.0176,
      "grad_norm": 1.619942307472229,
      "learning_rate": 0.0005,
      "loss": 1.9136,
      "step": 22
    },
    {
      "epoch": 0.0184,
      "grad_norm": 3.4306373596191406,
      "learning_rate": 0.0005,
      "loss": 1.7088,
      "step": 23
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.9691258072853088,
      "learning_rate": 0.0005,
      "loss": 1.2861,
      "step": 24
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.9319524765014648,
      "learning_rate": 0.0005,
      "loss": 1.9669,
      "step": 25
    },
    {
      "epoch": 0.0208,
      "grad_norm": 4.606762409210205,
      "learning_rate": 0.0005,
      "loss": 2.1804,
      "step": 26
    },
    {
      "epoch": 0.0216,
      "grad_norm": 0.6729273200035095,
      "learning_rate": 0.0005,
      "loss": 0.5727,
      "step": 27
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.9792982339859009,
      "learning_rate": 0.0005,
      "loss": 0.7054,
      "step": 28
    },
    {
      "epoch": 0.0232,
      "grad_norm": 3.6252601146698,
      "learning_rate": 0.0005,
      "loss": 2.201,
      "step": 29
    },
    {
      "epoch": 0.024,
      "grad_norm": 1.40730619430542,
      "learning_rate": 0.0005,
      "loss": 1.6824,
      "step": 30
    },
    {
      "epoch": 0.0248,
      "grad_norm": 10.60210132598877,
      "learning_rate": 0.0005,
      "loss": 2.8354,
      "step": 31
    },
    {
      "epoch": 0.0256,
      "grad_norm": 4.060218811035156,
      "learning_rate": 0.0005,
      "loss": 3.0198,
      "step": 32
    },
    {
      "epoch": 0.0264,
      "grad_norm": 2.5272302627563477,
      "learning_rate": 0.0005,
      "loss": 1.8971,
      "step": 33
    },
    {
      "epoch": 0.0272,
      "grad_norm": 3.3175830841064453,
      "learning_rate": 0.0005,
      "loss": 2.2358,
      "step": 34
    },
    {
      "epoch": 0.028,
      "grad_norm": 13.316032409667969,
      "learning_rate": 0.0005,
      "loss": 0.9997,
      "step": 35
    },
    {
      "epoch": 0.0288,
      "grad_norm": 7.873894691467285,
      "learning_rate": 0.0005,
      "loss": 4.109,
      "step": 36
    },
    {
      "epoch": 0.0296,
      "grad_norm": 12.632922172546387,
      "learning_rate": 0.0005,
      "loss": 4.6703,
      "step": 37
    },
    {
      "epoch": 0.0304,
      "grad_norm": 5.1721110343933105,
      "learning_rate": 0.0005,
      "loss": 0.2371,
      "step": 38
    },
    {
      "epoch": 0.0312,
      "grad_norm": 4.0092854499816895,
      "learning_rate": 0.0005,
      "loss": 1.6763,
      "step": 39
    },
    {
      "epoch": 0.032,
      "grad_norm": 13.548662185668945,
      "learning_rate": 0.0005,
      "loss": 3.3977,
      "step": 40
    },
    {
      "epoch": 0.0328,
      "grad_norm": 10.380196571350098,
      "learning_rate": 0.0005,
      "loss": 2.8333,
      "step": 41
    },
    {
      "epoch": 0.0336,
      "grad_norm": 10.124567031860352,
      "learning_rate": 0.0005,
      "loss": 1.8736,
      "step": 42
    },
    {
      "epoch": 0.0344,
      "grad_norm": 33.22747039794922,
      "learning_rate": 0.0005,
      "loss": 4.0325,
      "step": 43
    },
    {
      "epoch": 0.0352,
      "grad_norm": 23.466447830200195,
      "learning_rate": 0.0005,
      "loss": 3.209,
      "step": 44
    },
    {
      "epoch": 0.036,
      "grad_norm": 97.76129150390625,
      "learning_rate": 0.0005,
      "loss": 3.4155,
      "step": 45
    },
    {
      "epoch": 0.0368,
      "grad_norm": 12.2396879196167,
      "learning_rate": 0.0005,
      "loss": 1.5068,
      "step": 46
    },
    {
      "epoch": 0.0376,
      "grad_norm": 18.19890785217285,
      "learning_rate": 0.0005,
      "loss": 1.5872,
      "step": 47
    },
    {
      "epoch": 0.0384,
      "grad_norm": 33.10652160644531,
      "learning_rate": 0.0005,
      "loss": 0.4156,
      "step": 48
    },
    {
      "epoch": 0.0392,
      "grad_norm": 3.613457202911377,
      "learning_rate": 0.0005,
      "loss": 0.8332,
      "step": 49
    },
    {
      "epoch": 0.04,
      "grad_norm": 23.454383850097656,
      "learning_rate": 0.0005,
      "loss": 2.4502,
      "step": 50
    },
    {
      "epoch": 0.0408,
      "grad_norm": 5.318363189697266,
      "learning_rate": 0.0005,
      "loss": 1.3799,
      "step": 51
    },
    {
      "epoch": 0.0416,
      "grad_norm": 6.815697193145752,
      "learning_rate": 0.0005,
      "loss": 2.1557,
      "step": 52
    },
    {
      "epoch": 0.0424,
      "grad_norm": 11.527610778808594,
      "learning_rate": 0.0005,
      "loss": 1.298,
      "step": 53
    },
    {
      "epoch": 0.0432,
      "grad_norm": 4.768402099609375,
      "learning_rate": 0.0005,
      "loss": 1.107,
      "step": 54
    },
    {
      "epoch": 0.044,
      "grad_norm": 1.4602067470550537,
      "learning_rate": 0.0005,
      "loss": 0.3143,
      "step": 55
    },
    {
      "epoch": 0.0448,
      "grad_norm": 9.193857192993164,
      "learning_rate": 0.0005,
      "loss": 1.1428,
      "step": 56
    },
    {
      "epoch": 0.0456,
      "grad_norm": 11.116622924804688,
      "learning_rate": 0.0005,
      "loss": 1.788,
      "step": 57
    },
    {
      "epoch": 0.0464,
      "grad_norm": 4.629782676696777,
      "learning_rate": 0.0005,
      "loss": 1.0826,
      "step": 58
    },
    {
      "epoch": 0.0472,
      "grad_norm": 12.43929672241211,
      "learning_rate": 0.0005,
      "loss": 1.8785,
      "step": 59
    },
    {
      "epoch": 0.048,
      "grad_norm": 36.3989372253418,
      "learning_rate": 0.0005,
      "loss": 2.4311,
      "step": 60
    },
    {
      "epoch": 0.0488,
      "grad_norm": 22.378089904785156,
      "learning_rate": 0.0005,
      "loss": 2.036,
      "step": 61
    },
    {
      "epoch": 0.0496,
      "grad_norm": 37.49986267089844,
      "learning_rate": 0.0005,
      "loss": 2.8407,
      "step": 62
    },
    {
      "epoch": 0.0504,
      "grad_norm": 27.25046730041504,
      "learning_rate": 0.0005,
      "loss": 1.4435,
      "step": 63
    },
    {
      "epoch": 0.0512,
      "grad_norm": 10.643092155456543,
      "learning_rate": 0.0005,
      "loss": 0.8146,
      "step": 64
    },
    {
      "epoch": 0.052,
      "grad_norm": 25.00066566467285,
      "learning_rate": 0.0005,
      "loss": 1.4657,
      "step": 65
    },
    {
      "epoch": 0.0528,
      "grad_norm": 45.32073974609375,
      "learning_rate": 0.0005,
      "loss": 1.1125,
      "step": 66
    },
    {
      "epoch": 0.0536,
      "grad_norm": 19.11292266845703,
      "learning_rate": 0.0005,
      "loss": 1.5612,
      "step": 67
    },
    {
      "epoch": 0.0544,
      "grad_norm": 16.346227645874023,
      "learning_rate": 0.0005,
      "loss": 1.7633,
      "step": 68
    },
    {
      "epoch": 0.0552,
      "grad_norm": 27.169462203979492,
      "learning_rate": 0.0005,
      "loss": 1.3909,
      "step": 69
    },
    {
      "epoch": 0.056,
      "grad_norm": 6.169379711151123,
      "learning_rate": 0.0005,
      "loss": 0.6378,
      "step": 70
    },
    {
      "epoch": 0.0568,
      "grad_norm": 20.114524841308594,
      "learning_rate": 0.0005,
      "loss": 1.0569,
      "step": 71
    },
    {
      "epoch": 0.0576,
      "grad_norm": 29.240468978881836,
      "learning_rate": 0.0005,
      "loss": 1.118,
      "step": 72
    },
    {
      "epoch": 0.0584,
      "grad_norm": 49.98060989379883,
      "learning_rate": 0.0005,
      "loss": 2.3733,
      "step": 73
    },
    {
      "epoch": 0.0592,
      "grad_norm": 7.8704118728637695,
      "learning_rate": 0.0005,
      "loss": 0.9115,
      "step": 74
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.570856094360352,
      "learning_rate": 0.0005,
      "loss": 1.5051,
      "step": 75
    },
    {
      "epoch": 0.0608,
      "grad_norm": 8.354825973510742,
      "learning_rate": 0.0005,
      "loss": 0.6045,
      "step": 76
    },
    {
      "epoch": 0.0616,
      "grad_norm": 6.444540977478027,
      "learning_rate": 0.0005,
      "loss": 1.5749,
      "step": 77
    },
    {
      "epoch": 0.0624,
      "grad_norm": 20.538169860839844,
      "learning_rate": 0.0005,
      "loss": 1.6162,
      "step": 78
    },
    {
      "epoch": 0.0632,
      "grad_norm": 29.843812942504883,
      "learning_rate": 0.0005,
      "loss": 2.057,
      "step": 79
    },
    {
      "epoch": 0.064,
      "grad_norm": 12.291327476501465,
      "learning_rate": 0.0005,
      "loss": 1.8096,
      "step": 80
    },
    {
      "epoch": 0.0648,
      "grad_norm": 4.892437934875488,
      "learning_rate": 0.0005,
      "loss": 0.4261,
      "step": 81
    },
    {
      "epoch": 0.0656,
      "grad_norm": 8.39865779876709,
      "learning_rate": 0.0005,
      "loss": 0.9653,
      "step": 82
    },
    {
      "epoch": 0.0664,
      "grad_norm": 33.09463882446289,
      "learning_rate": 0.0005,
      "loss": 2.1087,
      "step": 83
    },
    {
      "epoch": 0.0672,
      "grad_norm": 12.01011848449707,
      "learning_rate": 0.0005,
      "loss": 0.9802,
      "step": 84
    },
    {
      "epoch": 0.068,
      "grad_norm": 4.892343044281006,
      "learning_rate": 0.0005,
      "loss": 1.0565,
      "step": 85
    },
    {
      "epoch": 0.0688,
      "grad_norm": 29.355051040649414,
      "learning_rate": 0.0005,
      "loss": 3.1704,
      "step": 86
    },
    {
      "epoch": 0.0696,
      "grad_norm": 12.124067306518555,
      "learning_rate": 0.0005,
      "loss": 3.0637,
      "step": 87
    },
    {
      "epoch": 0.0704,
      "grad_norm": 7.9273762702941895,
      "learning_rate": 0.0005,
      "loss": 1.93,
      "step": 88
    },
    {
      "epoch": 0.0712,
      "grad_norm": 14.156713485717773,
      "learning_rate": 0.0005,
      "loss": 2.8417,
      "step": 89
    },
    {
      "epoch": 0.072,
      "grad_norm": 8.770395278930664,
      "learning_rate": 0.0005,
      "loss": 2.9201,
      "step": 90
    },
    {
      "epoch": 0.0728,
      "grad_norm": 8.601066589355469,
      "learning_rate": 0.0005,
      "loss": 2.7391,
      "step": 91
    },
    {
      "epoch": 0.0736,
      "grad_norm": 11.746129989624023,
      "learning_rate": 0.0005,
      "loss": 1.79,
      "step": 92
    },
    {
      "epoch": 0.0744,
      "grad_norm": 83.34435272216797,
      "learning_rate": 0.0005,
      "loss": 3.4365,
      "step": 93
    },
    {
      "epoch": 0.0752,
      "grad_norm": 52.08599090576172,
      "learning_rate": 0.0005,
      "loss": 3.0587,
      "step": 94
    },
    {
      "epoch": 0.076,
      "grad_norm": 9.474879264831543,
      "learning_rate": 0.0005,
      "loss": 1.8807,
      "step": 95
    },
    {
      "epoch": 0.0768,
      "grad_norm": 4.777618408203125,
      "learning_rate": 0.0005,
      "loss": 1.1716,
      "step": 96
    },
    {
      "epoch": 0.0776,
      "grad_norm": 8.437329292297363,
      "learning_rate": 0.0005,
      "loss": 1.897,
      "step": 97
    },
    {
      "epoch": 0.0784,
      "grad_norm": 9.909850120544434,
      "learning_rate": 0.0005,
      "loss": 1.026,
      "step": 98
    },
    {
      "epoch": 0.0792,
      "grad_norm": 18.70866584777832,
      "learning_rate": 0.0005,
      "loss": 2.8539,
      "step": 99
    },
    {
      "epoch": 0.08,
      "grad_norm": 16.791696548461914,
      "learning_rate": 0.0005,
      "loss": 1.5087,
      "step": 100
    },
    {
      "epoch": 0.0808,
      "grad_norm": 10.638335227966309,
      "learning_rate": 0.0005,
      "loss": 1.9842,
      "step": 101
    },
    {
      "epoch": 0.0816,
      "grad_norm": 8.751673698425293,
      "learning_rate": 0.0005,
      "loss": 1.8033,
      "step": 102
    },
    {
      "epoch": 0.0824,
      "grad_norm": 10.246282577514648,
      "learning_rate": 0.0005,
      "loss": 2.2316,
      "step": 103
    },
    {
      "epoch": 0.0832,
      "grad_norm": 15.839383125305176,
      "learning_rate": 0.0005,
      "loss": 1.7817,
      "step": 104
    },
    {
      "epoch": 0.084,
      "grad_norm": 8.457541465759277,
      "learning_rate": 0.0005,
      "loss": 1.7869,
      "step": 105
    },
    {
      "epoch": 0.0848,
      "grad_norm": 6.870094299316406,
      "learning_rate": 0.0005,
      "loss": 1.8344,
      "step": 106
    },
    {
      "epoch": 0.0856,
      "grad_norm": 10.195223808288574,
      "learning_rate": 0.0005,
      "loss": 2.8778,
      "step": 107
    },
    {
      "epoch": 0.0864,
      "grad_norm": 12.052081108093262,
      "learning_rate": 0.0005,
      "loss": 2.1871,
      "step": 108
    },
    {
      "epoch": 0.0872,
      "grad_norm": 9.184932708740234,
      "learning_rate": 0.0005,
      "loss": 1.5251,
      "step": 109
    },
    {
      "epoch": 0.088,
      "grad_norm": 11.652626991271973,
      "learning_rate": 0.0005,
      "loss": 2.5694,
      "step": 110
    },
    {
      "epoch": 0.0888,
      "grad_norm": 17.4293212890625,
      "learning_rate": 0.0005,
      "loss": 2.753,
      "step": 111
    },
    {
      "epoch": 0.0896,
      "grad_norm": 16.894798278808594,
      "learning_rate": 0.0005,
      "loss": 2.4887,
      "step": 112
    },
    {
      "epoch": 0.0904,
      "grad_norm": 8.354668617248535,
      "learning_rate": 0.0005,
      "loss": 1.9092,
      "step": 113
    },
    {
      "epoch": 0.0912,
      "grad_norm": 7.03963565826416,
      "learning_rate": 0.0005,
      "loss": 1.364,
      "step": 114
    },
    {
      "epoch": 0.092,
      "grad_norm": 9.681304931640625,
      "learning_rate": 0.0005,
      "loss": 2.539,
      "step": 115
    },
    {
      "epoch": 0.0928,
      "grad_norm": 7.012979030609131,
      "learning_rate": 0.0005,
      "loss": 3.0951,
      "step": 116
    },
    {
      "epoch": 0.0936,
      "grad_norm": 6.343447208404541,
      "learning_rate": 0.0005,
      "loss": 1.1429,
      "step": 117
    },
    {
      "epoch": 0.0944,
      "grad_norm": 8.117640495300293,
      "learning_rate": 0.0005,
      "loss": 1.6064,
      "step": 118
    },
    {
      "epoch": 0.0952,
      "grad_norm": 7.953670501708984,
      "learning_rate": 0.0005,
      "loss": 2.1902,
      "step": 119
    },
    {
      "epoch": 0.096,
      "grad_norm": 21.180335998535156,
      "learning_rate": 0.0005,
      "loss": 3.1061,
      "step": 120
    },
    {
      "epoch": 0.0968,
      "grad_norm": 5.133480548858643,
      "learning_rate": 0.0005,
      "loss": 0.5414,
      "step": 121
    },
    {
      "epoch": 0.0976,
      "grad_norm": 5.382786750793457,
      "learning_rate": 0.0005,
      "loss": 1.3347,
      "step": 122
    },
    {
      "epoch": 0.0984,
      "grad_norm": 10.311848640441895,
      "learning_rate": 0.0005,
      "loss": 1.9284,
      "step": 123
    },
    {
      "epoch": 0.0992,
      "grad_norm": 7.540424823760986,
      "learning_rate": 0.0005,
      "loss": 1.8246,
      "step": 124
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.958598613739014,
      "learning_rate": 0.0005,
      "loss": 0.954,
      "step": 125
    },
    {
      "epoch": 0.1008,
      "grad_norm": 5.362274646759033,
      "learning_rate": 0.0005,
      "loss": 1.6635,
      "step": 126
    },
    {
      "epoch": 0.1016,
      "grad_norm": 15.276226997375488,
      "learning_rate": 0.0005,
      "loss": 4.6421,
      "step": 127
    },
    {
      "epoch": 0.1024,
      "grad_norm": 10.148447036743164,
      "learning_rate": 0.0005,
      "loss": 1.8489,
      "step": 128
    },
    {
      "epoch": 0.1032,
      "grad_norm": 10.659564018249512,
      "learning_rate": 0.0005,
      "loss": 1.9508,
      "step": 129
    },
    {
      "epoch": 0.104,
      "grad_norm": 5.988990306854248,
      "learning_rate": 0.0005,
      "loss": 1.1783,
      "step": 130
    },
    {
      "epoch": 0.1048,
      "grad_norm": 7.680914402008057,
      "learning_rate": 0.0005,
      "loss": 1.6839,
      "step": 131
    },
    {
      "epoch": 0.1056,
      "grad_norm": 6.716479301452637,
      "learning_rate": 0.0005,
      "loss": 1.6122,
      "step": 132
    },
    {
      "epoch": 0.1064,
      "grad_norm": 14.102774620056152,
      "learning_rate": 0.0005,
      "loss": 2.3155,
      "step": 133
    },
    {
      "epoch": 0.1072,
      "grad_norm": 9.768166542053223,
      "learning_rate": 0.0005,
      "loss": 2.0646,
      "step": 134
    },
    {
      "epoch": 0.108,
      "grad_norm": 6.7697834968566895,
      "learning_rate": 0.0005,
      "loss": 1.1913,
      "step": 135
    },
    {
      "epoch": 0.1088,
      "grad_norm": 6.375694274902344,
      "learning_rate": 0.0005,
      "loss": 2.243,
      "step": 136
    },
    {
      "epoch": 0.1096,
      "grad_norm": 12.372509002685547,
      "learning_rate": 0.0005,
      "loss": 2.4898,
      "step": 137
    },
    {
      "epoch": 0.1104,
      "grad_norm": 14.37572193145752,
      "learning_rate": 0.0005,
      "loss": 3.1866,
      "step": 138
    },
    {
      "epoch": 0.1112,
      "grad_norm": 19.542118072509766,
      "learning_rate": 0.0005,
      "loss": 3.4329,
      "step": 139
    },
    {
      "epoch": 0.112,
      "grad_norm": 25.670761108398438,
      "learning_rate": 0.0005,
      "loss": 4.2204,
      "step": 140
    },
    {
      "epoch": 0.1128,
      "grad_norm": 3.7592780590057373,
      "learning_rate": 0.0005,
      "loss": 0.6329,
      "step": 141
    },
    {
      "epoch": 0.1136,
      "grad_norm": 13.57589340209961,
      "learning_rate": 0.0005,
      "loss": 3.4199,
      "step": 142
    },
    {
      "epoch": 0.1144,
      "grad_norm": 15.773186683654785,
      "learning_rate": 0.0005,
      "loss": 1.6622,
      "step": 143
    },
    {
      "epoch": 0.1152,
      "grad_norm": 21.95498275756836,
      "learning_rate": 0.0005,
      "loss": 2.502,
      "step": 144
    },
    {
      "epoch": 0.116,
      "grad_norm": 14.796022415161133,
      "learning_rate": 0.0005,
      "loss": 3.3742,
      "step": 145
    },
    {
      "epoch": 0.1168,
      "grad_norm": 11.765939712524414,
      "learning_rate": 0.0005,
      "loss": 3.1213,
      "step": 146
    },
    {
      "epoch": 0.1176,
      "grad_norm": 14.750462532043457,
      "learning_rate": 0.0005,
      "loss": 4.411,
      "step": 147
    },
    {
      "epoch": 0.1184,
      "grad_norm": 13.234830856323242,
      "learning_rate": 0.0005,
      "loss": 3.3631,
      "step": 148
    },
    {
      "epoch": 0.1192,
      "grad_norm": 6.553050518035889,
      "learning_rate": 0.0005,
      "loss": 1.8951,
      "step": 149
    },
    {
      "epoch": 0.12,
      "grad_norm": 14.43946647644043,
      "learning_rate": 0.0005,
      "loss": 1.8703,
      "step": 150
    },
    {
      "epoch": 0.1208,
      "grad_norm": 9.733112335205078,
      "learning_rate": 0.0005,
      "loss": 2.8055,
      "step": 151
    },
    {
      "epoch": 0.1216,
      "grad_norm": 10.737160682678223,
      "learning_rate": 0.0005,
      "loss": 2.9521,
      "step": 152
    },
    {
      "epoch": 0.1224,
      "grad_norm": 12.203970909118652,
      "learning_rate": 0.0005,
      "loss": 2.9785,
      "step": 153
    },
    {
      "epoch": 0.1232,
      "grad_norm": 10.63853645324707,
      "learning_rate": 0.0005,
      "loss": 2.6857,
      "step": 154
    },
    {
      "epoch": 0.124,
      "grad_norm": 10.457780838012695,
      "learning_rate": 0.0005,
      "loss": 1.743,
      "step": 155
    },
    {
      "epoch": 0.1248,
      "grad_norm": 12.22805404663086,
      "learning_rate": 0.0005,
      "loss": 3.9353,
      "step": 156
    },
    {
      "epoch": 0.1256,
      "grad_norm": 7.321228981018066,
      "learning_rate": 0.0005,
      "loss": 2.2439,
      "step": 157
    },
    {
      "epoch": 0.1264,
      "grad_norm": 8.463542938232422,
      "learning_rate": 0.0005,
      "loss": 3.9458,
      "step": 158
    },
    {
      "epoch": 0.1272,
      "grad_norm": 8.441742897033691,
      "learning_rate": 0.0005,
      "loss": 3.0918,
      "step": 159
    },
    {
      "epoch": 0.128,
      "grad_norm": 15.034650802612305,
      "learning_rate": 0.0005,
      "loss": 4.1888,
      "step": 160
    },
    {
      "epoch": 0.1288,
      "grad_norm": 21.658571243286133,
      "learning_rate": 0.0005,
      "loss": 5.6277,
      "step": 161
    },
    {
      "epoch": 0.1296,
      "grad_norm": 9.453361511230469,
      "learning_rate": 0.0005,
      "loss": 3.1392,
      "step": 162
    },
    {
      "epoch": 0.1304,
      "grad_norm": 17.79288101196289,
      "learning_rate": 0.0005,
      "loss": 3.6066,
      "step": 163
    },
    {
      "epoch": 0.1312,
      "grad_norm": 14.434952735900879,
      "learning_rate": 0.0005,
      "loss": 4.2082,
      "step": 164
    },
    {
      "epoch": 0.132,
      "grad_norm": 7.758333206176758,
      "learning_rate": 0.0005,
      "loss": 1.6918,
      "step": 165
    },
    {
      "epoch": 0.1328,
      "grad_norm": 11.610477447509766,
      "learning_rate": 0.0005,
      "loss": 3.4442,
      "step": 166
    },
    {
      "epoch": 0.1336,
      "grad_norm": 17.194080352783203,
      "learning_rate": 0.0005,
      "loss": 4.6679,
      "step": 167
    },
    {
      "epoch": 0.1344,
      "grad_norm": 11.869499206542969,
      "learning_rate": 0.0005,
      "loss": 4.1471,
      "step": 168
    },
    {
      "epoch": 0.1352,
      "grad_norm": 8.291399955749512,
      "learning_rate": 0.0005,
      "loss": 2.4761,
      "step": 169
    },
    {
      "epoch": 0.136,
      "grad_norm": 10.868124008178711,
      "learning_rate": 0.0005,
      "loss": 3.1574,
      "step": 170
    },
    {
      "epoch": 0.1368,
      "grad_norm": 28.47736167907715,
      "learning_rate": 0.0005,
      "loss": 3.4398,
      "step": 171
    },
    {
      "epoch": 0.1376,
      "grad_norm": 7.985611438751221,
      "learning_rate": 0.0005,
      "loss": 1.7169,
      "step": 172
    },
    {
      "epoch": 0.1384,
      "grad_norm": 7.796197414398193,
      "learning_rate": 0.0005,
      "loss": 1.6151,
      "step": 173
    },
    {
      "epoch": 0.1392,
      "grad_norm": 31.824750900268555,
      "learning_rate": 0.0005,
      "loss": 5.5771,
      "step": 174
    },
    {
      "epoch": 0.14,
      "grad_norm": 11.322871208190918,
      "learning_rate": 0.0005,
      "loss": 1.7649,
      "step": 175
    },
    {
      "epoch": 0.1408,
      "grad_norm": 12.81942081451416,
      "learning_rate": 0.0005,
      "loss": 4.2761,
      "step": 176
    },
    {
      "epoch": 0.1416,
      "grad_norm": 14.586856842041016,
      "learning_rate": 0.0005,
      "loss": 3.7712,
      "step": 177
    },
    {
      "epoch": 0.1424,
      "grad_norm": 8.232019424438477,
      "learning_rate": 0.0005,
      "loss": 3.6163,
      "step": 178
    },
    {
      "epoch": 0.1432,
      "grad_norm": 12.344880104064941,
      "learning_rate": 0.0005,
      "loss": 2.9607,
      "step": 179
    },
    {
      "epoch": 0.144,
      "grad_norm": 15.326787948608398,
      "learning_rate": 0.0005,
      "loss": 3.7187,
      "step": 180
    },
    {
      "epoch": 0.1448,
      "grad_norm": 8.811168670654297,
      "learning_rate": 0.0005,
      "loss": 1.5032,
      "step": 181
    },
    {
      "epoch": 0.1456,
      "grad_norm": 26.123863220214844,
      "learning_rate": 0.0005,
      "loss": 2.9496,
      "step": 182
    },
    {
      "epoch": 0.1464,
      "grad_norm": 7.073259353637695,
      "learning_rate": 0.0005,
      "loss": 2.0154,
      "step": 183
    },
    {
      "epoch": 0.1472,
      "grad_norm": 10.833226203918457,
      "learning_rate": 0.0005,
      "loss": 3.3223,
      "step": 184
    },
    {
      "epoch": 0.148,
      "grad_norm": 12.570619583129883,
      "learning_rate": 0.0005,
      "loss": 3.9599,
      "step": 185
    },
    {
      "epoch": 0.1488,
      "grad_norm": 10.237825393676758,
      "learning_rate": 0.0005,
      "loss": 1.4809,
      "step": 186
    },
    {
      "epoch": 0.1496,
      "grad_norm": 9.36881160736084,
      "learning_rate": 0.0005,
      "loss": 2.7919,
      "step": 187
    },
    {
      "epoch": 0.1504,
      "grad_norm": 7.162893772125244,
      "learning_rate": 0.0005,
      "loss": 1.7637,
      "step": 188
    },
    {
      "epoch": 0.1512,
      "grad_norm": 13.637219429016113,
      "learning_rate": 0.0005,
      "loss": 1.7047,
      "step": 189
    },
    {
      "epoch": 0.152,
      "grad_norm": 13.284880638122559,
      "learning_rate": 0.0005,
      "loss": 3.9825,
      "step": 190
    },
    {
      "epoch": 0.1528,
      "grad_norm": 17.694873809814453,
      "learning_rate": 0.0005,
      "loss": 4.8066,
      "step": 191
    },
    {
      "epoch": 0.1536,
      "grad_norm": 20.54854965209961,
      "learning_rate": 0.0005,
      "loss": 6.1623,
      "step": 192
    },
    {
      "epoch": 0.1544,
      "grad_norm": 8.24622917175293,
      "learning_rate": 0.0005,
      "loss": 2.0798,
      "step": 193
    },
    {
      "epoch": 0.1552,
      "grad_norm": 7.209560871124268,
      "learning_rate": 0.0005,
      "loss": 1.9318,
      "step": 194
    },
    {
      "epoch": 0.156,
      "grad_norm": 12.79217529296875,
      "learning_rate": 0.0005,
      "loss": 2.8844,
      "step": 195
    },
    {
      "epoch": 0.1568,
      "grad_norm": 5.256202697753906,
      "learning_rate": 0.0005,
      "loss": 1.4686,
      "step": 196
    },
    {
      "epoch": 0.1576,
      "grad_norm": 12.968066215515137,
      "learning_rate": 0.0005,
      "loss": 2.5109,
      "step": 197
    },
    {
      "epoch": 0.1584,
      "grad_norm": 11.350977897644043,
      "learning_rate": 0.0005,
      "loss": 4.5141,
      "step": 198
    },
    {
      "epoch": 0.1592,
      "grad_norm": 9.951147079467773,
      "learning_rate": 0.0005,
      "loss": 3.2528,
      "step": 199
    },
    {
      "epoch": 0.16,
      "grad_norm": 20.016956329345703,
      "learning_rate": 0.0005,
      "loss": 5.7353,
      "step": 200
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.02021397118976e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}