{
  "best_metric": 101.20526885986328,
  "best_model_checkpoint": "./ckpts/tinyllama_v1.1/int3-g128/checkpoint-400",
  "epoch": 4.0,
  "eval_steps": 100,
  "global_step": 400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 0.0,
      "loss": 256.4583,
      "step": 1
    },
    {
      "epoch": 0.02,
      "learning_rate": 2e-05,
      "loss": 355.0836,
      "step": 2
    },
    {
      "epoch": 0.03,
      "learning_rate": 2e-05,
      "loss": 252.4541,
      "step": 3
    },
    {
      "epoch": 0.04,
      "learning_rate": 2e-05,
      "loss": 761.1344,
      "step": 4
    },
    {
      "epoch": 0.05,
      "learning_rate": 2e-05,
      "loss": 465.4658,
      "step": 5
    },
    {
      "epoch": 0.06,
      "learning_rate": 2e-05,
      "loss": 841.6429,
      "step": 6
    },
    {
      "epoch": 0.07,
      "learning_rate": 2e-05,
      "loss": 753.775,
      "step": 7
    },
    {
      "epoch": 0.08,
      "learning_rate": 2e-05,
      "loss": 379.0151,
      "step": 8
    },
    {
      "epoch": 0.09,
      "learning_rate": 2e-05,
      "loss": 402.6746,
      "step": 9
    },
    {
      "epoch": 0.1,
      "learning_rate": 2e-05,
      "loss": 216.2126,
      "step": 10
    },
    {
      "epoch": 0.11,
      "learning_rate": 2e-05,
      "loss": 227.9789,
      "step": 11
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 237.3062,
      "step": 12
    },
    {
      "epoch": 0.13,
      "learning_rate": 2e-05,
      "loss": 299.5192,
      "step": 13
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-05,
      "loss": 186.9158,
      "step": 14
    },
    {
      "epoch": 0.15,
      "learning_rate": 2e-05,
      "loss": 209.0085,
      "step": 15
    },
    {
      "epoch": 0.16,
      "learning_rate": 2e-05,
      "loss": 151.9208,
      "step": 16
    },
    {
      "epoch": 0.17,
      "learning_rate": 2e-05,
      "loss": 177.2832,
      "step": 17
    },
    {
      "epoch": 0.18,
      "learning_rate": 2e-05,
      "loss": 180.6637,
      "step": 18
    },
    {
      "epoch": 0.19,
      "learning_rate": 2e-05,
      "loss": 153.5752,
      "step": 19
    },
    {
      "epoch": 0.2,
      "learning_rate": 2e-05,
      "loss": 241.4604,
      "step": 20
    },
    {
      "epoch": 0.21,
      "learning_rate": 2e-05,
      "loss": 154.7056,
      "step": 21
    },
    {
      "epoch": 0.22,
      "learning_rate": 2e-05,
      "loss": 152.9164,
      "step": 22
    },
    {
      "epoch": 0.23,
      "learning_rate": 2e-05,
      "loss": 117.8652,
      "step": 23
    },
    {
      "epoch": 0.24,
      "learning_rate": 2e-05,
      "loss": 139.5359,
      "step": 24
    },
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 125.2291,
      "step": 25
    },
    {
      "epoch": 0.26,
      "learning_rate": 2e-05,
      "loss": 153.2976,
      "step": 26
    },
    {
      "epoch": 0.27,
      "learning_rate": 2e-05,
      "loss": 138.6838,
      "step": 27
    },
    {
      "epoch": 0.28,
      "learning_rate": 2e-05,
      "loss": 147.1582,
      "step": 28
    },
    {
      "epoch": 0.29,
      "learning_rate": 2e-05,
      "loss": 120.231,
      "step": 29
    },
    {
      "epoch": 0.3,
      "learning_rate": 2e-05,
      "loss": 136.0073,
      "step": 30
    },
    {
      "epoch": 0.31,
      "learning_rate": 2e-05,
      "loss": 125.3789,
      "step": 31
    },
    {
      "epoch": 0.32,
      "learning_rate": 2e-05,
      "loss": 135.6559,
      "step": 32
    },
    {
      "epoch": 0.33,
      "learning_rate": 2e-05,
      "loss": 151.9706,
      "step": 33
    },
    {
      "epoch": 0.34,
      "learning_rate": 2e-05,
      "loss": 126.2539,
      "step": 34
    },
    {
      "epoch": 0.35,
      "learning_rate": 2e-05,
      "loss": 110.8474,
      "step": 35
    },
    {
      "epoch": 0.36,
      "learning_rate": 2e-05,
      "loss": 150.4755,
      "step": 36
    },
    {
      "epoch": 0.37,
      "learning_rate": 2e-05,
      "loss": 133.6725,
      "step": 37
    },
    {
      "epoch": 0.38,
      "learning_rate": 2e-05,
      "loss": 166.5977,
      "step": 38
    },
    {
      "epoch": 0.39,
      "learning_rate": 2e-05,
      "loss": 167.9467,
      "step": 39
    },
    {
      "epoch": 0.4,
      "learning_rate": 2e-05,
      "loss": 125.6363,
      "step": 40
    },
    {
      "epoch": 0.41,
      "learning_rate": 2e-05,
      "loss": 129.4,
      "step": 41
    },
    {
      "epoch": 0.42,
      "learning_rate": 2e-05,
      "loss": 109.1935,
      "step": 42
    },
    {
      "epoch": 0.43,
      "learning_rate": 2e-05,
      "loss": 164.8529,
      "step": 43
    },
    {
      "epoch": 0.44,
      "learning_rate": 2e-05,
      "loss": 127.5447,
      "step": 44
    },
    {
      "epoch": 0.45,
      "learning_rate": 2e-05,
      "loss": 126.8883,
      "step": 45
    },
    {
      "epoch": 0.46,
      "learning_rate": 2e-05,
      "loss": 156.4396,
      "step": 46
    },
    {
      "epoch": 0.47,
      "learning_rate": 2e-05,
      "loss": 90.8995,
      "step": 47
    },
    {
      "epoch": 0.48,
      "learning_rate": 2e-05,
      "loss": 131.5748,
      "step": 48
    },
    {
      "epoch": 0.49,
      "learning_rate": 2e-05,
      "loss": 137.7466,
      "step": 49
    },
    {
      "epoch": 0.5,
      "learning_rate": 2e-05,
      "loss": 105.3827,
      "step": 50
    },
    {
      "epoch": 0.51,
      "learning_rate": 2e-05,
      "loss": 131.8921,
      "step": 51
    },
    {
      "epoch": 0.52,
      "learning_rate": 2e-05,
      "loss": 112.9179,
      "step": 52
    },
    {
      "epoch": 0.53,
      "learning_rate": 2e-05,
      "loss": 143.8214,
      "step": 53
    },
    {
      "epoch": 0.54,
      "learning_rate": 2e-05,
      "loss": 168.2706,
      "step": 54
    },
    {
      "epoch": 0.55,
      "learning_rate": 2e-05,
      "loss": 153.9795,
      "step": 55
    },
    {
      "epoch": 0.56,
      "learning_rate": 2e-05,
      "loss": 115.3438,
      "step": 56
    },
    {
      "epoch": 0.57,
      "learning_rate": 2e-05,
      "loss": 144.034,
      "step": 57
    },
    {
      "epoch": 0.58,
      "learning_rate": 2e-05,
      "loss": 90.717,
      "step": 58
    },
    {
      "epoch": 0.59,
      "learning_rate": 2e-05,
      "loss": 117.4731,
      "step": 59
    },
    {
      "epoch": 0.6,
      "learning_rate": 2e-05,
      "loss": 143.9666,
      "step": 60
    },
    {
      "epoch": 0.61,
      "learning_rate": 2e-05,
      "loss": 127.3566,
      "step": 61
    },
    {
      "epoch": 0.62,
      "learning_rate": 2e-05,
      "loss": 107.2432,
      "step": 62
    },
    {
      "epoch": 0.63,
      "learning_rate": 2e-05,
      "loss": 117.6042,
      "step": 63
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 92.6121,
      "step": 64
    },
    {
      "epoch": 0.65,
      "learning_rate": 2e-05,
      "loss": 107.5126,
      "step": 65
    },
    {
      "epoch": 0.66,
      "learning_rate": 2e-05,
      "loss": 133.897,
      "step": 66
    },
    {
      "epoch": 0.67,
      "learning_rate": 2e-05,
      "loss": 101.0651,
      "step": 67
    },
    {
      "epoch": 0.68,
      "learning_rate": 2e-05,
      "loss": 99.0262,
      "step": 68
    },
    {
      "epoch": 0.69,
      "learning_rate": 2e-05,
      "loss": 156.479,
      "step": 69
    },
    {
      "epoch": 0.7,
      "learning_rate": 2e-05,
      "loss": 172.746,
      "step": 70
    },
    {
      "epoch": 0.71,
      "learning_rate": 2e-05,
      "loss": 104.9099,
      "step": 71
    },
    {
      "epoch": 0.72,
      "learning_rate": 2e-05,
      "loss": 94.8904,
      "step": 72
    },
    {
      "epoch": 0.73,
      "learning_rate": 2e-05,
      "loss": 105.7694,
      "step": 73
    },
    {
      "epoch": 0.74,
      "learning_rate": 2e-05,
      "loss": 92.2346,
      "step": 74
    },
    {
      "epoch": 0.75,
      "learning_rate": 2e-05,
      "loss": 118.597,
      "step": 75
    },
    {
      "epoch": 0.76,
      "learning_rate": 2e-05,
      "loss": 139.2074,
      "step": 76
    },
    {
      "epoch": 0.77,
      "learning_rate": 2e-05,
      "loss": 91.171,
      "step": 77
    },
    {
      "epoch": 0.78,
      "learning_rate": 2e-05,
      "loss": 121.2573,
      "step": 78
    },
    {
      "epoch": 0.79,
      "learning_rate": 2e-05,
      "loss": 91.8224,
      "step": 79
    },
    {
      "epoch": 0.8,
      "learning_rate": 2e-05,
      "loss": 107.0462,
      "step": 80
    },
    {
      "epoch": 0.81,
      "learning_rate": 2e-05,
      "loss": 98.3316,
      "step": 81
    },
    {
      "epoch": 0.82,
      "learning_rate": 2e-05,
      "loss": 111.1321,
      "step": 82
    },
    {
      "epoch": 0.83,
      "learning_rate": 2e-05,
      "loss": 77.9181,
      "step": 83
    },
    {
      "epoch": 0.84,
      "learning_rate": 2e-05,
      "loss": 91.1592,
      "step": 84
    },
    {
      "epoch": 0.85,
      "learning_rate": 2e-05,
      "loss": 125.3041,
      "step": 85
    },
    {
      "epoch": 0.86,
      "learning_rate": 2e-05,
      "loss": 165.3387,
      "step": 86
    },
    {
      "epoch": 0.87,
      "learning_rate": 2e-05,
      "loss": 84.1901,
      "step": 87
    },
    {
      "epoch": 0.88,
      "learning_rate": 2e-05,
      "loss": 120.4,
      "step": 88
    },
    {
      "epoch": 0.89,
      "learning_rate": 2e-05,
      "loss": 116.1603,
      "step": 89
    },
    {
      "epoch": 0.9,
      "learning_rate": 2e-05,
      "loss": 102.2226,
      "step": 90
    },
    {
      "epoch": 0.91,
      "learning_rate": 2e-05,
      "loss": 102.1463,
      "step": 91
    },
    {
      "epoch": 0.92,
      "learning_rate": 2e-05,
      "loss": 156.9727,
      "step": 92
    },
    {
      "epoch": 0.93,
      "learning_rate": 2e-05,
      "loss": 98.1492,
      "step": 93
    },
    {
      "epoch": 0.94,
      "learning_rate": 2e-05,
      "loss": 89.7699,
      "step": 94
    },
    {
      "epoch": 0.95,
      "learning_rate": 2e-05,
      "loss": 106.7468,
      "step": 95
    },
    {
      "epoch": 0.96,
      "learning_rate": 2e-05,
      "loss": 87.7702,
      "step": 96
    },
    {
      "epoch": 0.97,
      "learning_rate": 2e-05,
      "loss": 149.5999,
      "step": 97
    },
    {
      "epoch": 0.98,
      "learning_rate": 2e-05,
      "loss": 105.2152,
      "step": 98
    },
    {
      "epoch": 0.99,
      "learning_rate": 2e-05,
      "loss": 110.4924,
      "step": 99
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 159.247,
      "step": 100
    },
    {
      "epoch": 1.0,
      "eval_loss": 127.25352478027344,
      "eval_runtime": 80.8206,
      "eval_samples_per_second": 19.747,
      "eval_steps_per_second": 1.237,
      "step": 100
    },
    {
      "epoch": 1.01,
      "learning_rate": 2e-05,
      "loss": 117.8373,
      "step": 101
    },
    {
      "epoch": 1.02,
      "learning_rate": 2e-05,
      "loss": 81.0658,
      "step": 102
    },
    {
      "epoch": 1.03,
      "learning_rate": 2e-05,
      "loss": 79.7859,
      "step": 103
    },
    {
      "epoch": 1.04,
      "learning_rate": 2e-05,
      "loss": 168.3286,
      "step": 104
    },
    {
      "epoch": 1.05,
      "learning_rate": 2e-05,
      "loss": 172.5758,
      "step": 105
    },
    {
      "epoch": 1.06,
      "learning_rate": 2e-05,
      "loss": 83.6487,
      "step": 106
    },
    {
      "epoch": 1.07,
      "learning_rate": 2e-05,
      "loss": 140.8672,
      "step": 107
    },
    {
      "epoch": 1.08,
      "learning_rate": 2e-05,
      "loss": 118.9715,
      "step": 108
    },
    {
      "epoch": 1.09,
      "learning_rate": 2e-05,
      "loss": 102.4753,
      "step": 109
    },
    {
      "epoch": 1.1,
      "learning_rate": 2e-05,
      "loss": 90.8443,
      "step": 110
    },
    {
      "epoch": 1.11,
      "learning_rate": 2e-05,
      "loss": 108.1487,
      "step": 111
    },
    {
      "epoch": 1.12,
      "learning_rate": 2e-05,
      "loss": 80.9609,
      "step": 112
    },
    {
      "epoch": 1.13,
      "learning_rate": 2e-05,
      "loss": 99.0464,
      "step": 113
    },
    {
      "epoch": 1.14,
      "learning_rate": 2e-05,
      "loss": 100.1197,
      "step": 114
    },
    {
      "epoch": 1.15,
      "learning_rate": 2e-05,
      "loss": 110.2522,
      "step": 115
    },
    {
      "epoch": 1.16,
      "learning_rate": 2e-05,
      "loss": 75.1395,
      "step": 116
    },
    {
      "epoch": 1.17,
      "learning_rate": 2e-05,
      "loss": 170.2633,
      "step": 117
    },
    {
      "epoch": 1.18,
      "learning_rate": 2e-05,
      "loss": 96.691,
      "step": 118
    },
    {
      "epoch": 1.19,
      "learning_rate": 2e-05,
      "loss": 126.8882,
      "step": 119
    },
    {
      "epoch": 1.2,
      "learning_rate": 2e-05,
      "loss": 127.7979,
      "step": 120
    },
    {
      "epoch": 1.21,
      "learning_rate": 2e-05,
      "loss": 163.4338,
      "step": 121
    },
    {
      "epoch": 1.22,
      "learning_rate": 2e-05,
      "loss": 175.9516,
      "step": 122
    },
    {
      "epoch": 1.23,
      "learning_rate": 2e-05,
      "loss": 88.2406,
      "step": 123
    },
    {
      "epoch": 1.24,
      "learning_rate": 2e-05,
      "loss": 146.5514,
      "step": 124
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 98.5611,
      "step": 125
    },
    {
      "epoch": 1.26,
      "learning_rate": 2e-05,
      "loss": 101.5959,
      "step": 126
    },
    {
      "epoch": 1.27,
      "learning_rate": 2e-05,
      "loss": 105.368,
      "step": 127
    },
    {
      "epoch": 1.28,
      "learning_rate": 2e-05,
      "loss": 130.4881,
      "step": 128
    },
    {
      "epoch": 1.29,
      "learning_rate": 2e-05,
      "loss": 111.5857,
      "step": 129
    },
    {
      "epoch": 1.3,
      "learning_rate": 2e-05,
      "loss": 102.0928,
      "step": 130
    },
    {
      "epoch": 1.31,
      "learning_rate": 2e-05,
      "loss": 99.7146,
      "step": 131
    },
    {
      "epoch": 1.32,
      "learning_rate": 2e-05,
      "loss": 100.4634,
      "step": 132
    },
    {
      "epoch": 1.33,
      "learning_rate": 2e-05,
      "loss": 95.0521,
      "step": 133
    },
    {
      "epoch": 1.34,
      "learning_rate": 2e-05,
      "loss": 163.5093,
      "step": 134
    },
    {
      "epoch": 1.35,
      "learning_rate": 2e-05,
      "loss": 164.4761,
      "step": 135
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 108.3719,
      "step": 136
    },
    {
      "epoch": 1.37,
      "learning_rate": 2e-05,
      "loss": 164.0315,
      "step": 137
    },
    {
      "epoch": 1.38,
      "learning_rate": 2e-05,
      "loss": 80.1498,
      "step": 138
    },
    {
      "epoch": 1.39,
      "learning_rate": 2e-05,
      "loss": 82.825,
      "step": 139
    },
    {
      "epoch": 1.4,
      "learning_rate": 2e-05,
      "loss": 119.8586,
      "step": 140
    },
    {
      "epoch": 1.41,
      "learning_rate": 2e-05,
      "loss": 118.1383,
      "step": 141
    },
    {
      "epoch": 1.42,
      "learning_rate": 2e-05,
      "loss": 145.6537,
      "step": 142
    },
    {
      "epoch": 1.43,
      "learning_rate": 2e-05,
      "loss": 131.2617,
      "step": 143
    },
    {
      "epoch": 1.44,
      "learning_rate": 2e-05,
      "loss": 122.246,
      "step": 144
    },
    {
      "epoch": 1.45,
      "learning_rate": 2e-05,
      "loss": 98.2258,
      "step": 145
    },
    {
      "epoch": 1.46,
      "learning_rate": 2e-05,
      "loss": 166.0939,
      "step": 146
    },
    {
      "epoch": 1.47,
      "learning_rate": 2e-05,
      "loss": 112.9639,
      "step": 147
    },
    {
      "epoch": 1.48,
      "learning_rate": 2e-05,
      "loss": 122.0515,
      "step": 148
    },
    {
      "epoch": 1.49,
      "learning_rate": 2e-05,
      "loss": 140.3237,
      "step": 149
    },
    {
      "epoch": 1.5,
      "learning_rate": 2e-05,
      "loss": 101.4403,
      "step": 150
    },
    {
      "epoch": 1.51,
      "learning_rate": 2e-05,
      "loss": 95.5035,
      "step": 151
    },
    {
      "epoch": 1.52,
      "learning_rate": 2e-05,
      "loss": 145.9391,
      "step": 152
    },
    {
      "epoch": 1.53,
      "learning_rate": 2e-05,
      "loss": 105.1173,
      "step": 153
    },
    {
      "epoch": 1.54,
      "learning_rate": 2e-05,
      "loss": 105.8428,
      "step": 154
    },
    {
      "epoch": 1.55,
      "learning_rate": 2e-05,
      "loss": 85.6933,
      "step": 155
    },
    {
      "epoch": 1.56,
      "learning_rate": 2e-05,
      "loss": 118.8751,
      "step": 156
    },
    {
      "epoch": 1.57,
      "learning_rate": 2e-05,
      "loss": 101.1875,
      "step": 157
    },
    {
      "epoch": 1.58,
      "learning_rate": 2e-05,
      "loss": 105.5553,
      "step": 158
    },
    {
      "epoch": 1.59,
      "learning_rate": 2e-05,
      "loss": 112.3205,
      "step": 159
    },
    {
      "epoch": 1.6,
      "learning_rate": 2e-05,
      "loss": 143.0757,
      "step": 160
    },
    {
      "epoch": 1.61,
      "learning_rate": 2e-05,
      "loss": 112.1303,
      "step": 161
    },
    {
      "epoch": 1.62,
      "learning_rate": 2e-05,
      "loss": 95.4685,
      "step": 162
    },
    {
      "epoch": 1.63,
      "learning_rate": 2e-05,
      "loss": 111.5353,
      "step": 163
    },
    {
      "epoch": 1.64,
      "learning_rate": 2e-05,
      "loss": 121.3872,
      "step": 164
    },
    {
      "epoch": 1.65,
      "learning_rate": 2e-05,
      "loss": 90.6964,
      "step": 165
    },
    {
      "epoch": 1.66,
      "learning_rate": 2e-05,
      "loss": 96.5425,
      "step": 166
    },
    {
      "epoch": 1.67,
      "learning_rate": 2e-05,
      "loss": 112.4037,
      "step": 167
    },
    {
      "epoch": 1.68,
      "learning_rate": 2e-05,
      "loss": 143.2002,
      "step": 168
    },
    {
      "epoch": 1.69,
      "learning_rate": 2e-05,
      "loss": 114.4325,
      "step": 169
    },
    {
      "epoch": 1.7,
      "learning_rate": 2e-05,
      "loss": 116.9987,
      "step": 170
    },
    {
      "epoch": 1.71,
      "learning_rate": 2e-05,
      "loss": 148.152,
      "step": 171
    },
    {
      "epoch": 1.72,
      "learning_rate": 2e-05,
      "loss": 82.8326,
      "step": 172
    },
    {
      "epoch": 1.73,
      "learning_rate": 2e-05,
      "loss": 86.1253,
      "step": 173
    },
    {
      "epoch": 1.74,
      "learning_rate": 2e-05,
      "loss": 96.1883,
      "step": 174
    },
    {
      "epoch": 1.75,
      "learning_rate": 2e-05,
      "loss": 192.9743,
      "step": 175
    },
    {
      "epoch": 1.76,
      "learning_rate": 2e-05,
      "loss": 98.351,
      "step": 176
    },
    {
      "epoch": 1.77,
      "learning_rate": 2e-05,
      "loss": 361.8705,
      "step": 177
    },
    {
      "epoch": 1.78,
      "learning_rate": 2e-05,
      "loss": 102.5058,
      "step": 178
    },
    {
      "epoch": 1.79,
      "learning_rate": 2e-05,
      "loss": 154.9456,
      "step": 179
    },
    {
      "epoch": 1.8,
      "learning_rate": 2e-05,
      "loss": 86.3787,
      "step": 180
    },
    {
      "epoch": 1.81,
      "learning_rate": 2e-05,
      "loss": 139.9809,
      "step": 181
    },
    {
      "epoch": 1.82,
      "learning_rate": 2e-05,
      "loss": 112.6604,
      "step": 182
    },
    {
      "epoch": 1.83,
      "learning_rate": 2e-05,
      "loss": 238.6009,
      "step": 183
    },
    {
      "epoch": 1.84,
      "learning_rate": 2e-05,
      "loss": 102.7639,
      "step": 184
    },
    {
      "epoch": 1.85,
      "learning_rate": 2e-05,
      "loss": 146.489,
      "step": 185
    },
    {
      "epoch": 1.86,
      "learning_rate": 2e-05,
      "loss": 76.5674,
      "step": 186
    },
    {
      "epoch": 1.87,
      "learning_rate": 2e-05,
      "loss": 105.2609,
      "step": 187
    },
    {
      "epoch": 1.88,
      "learning_rate": 2e-05,
      "loss": 92.5275,
      "step": 188
    },
    {
      "epoch": 1.89,
      "learning_rate": 2e-05,
      "loss": 120.9601,
      "step": 189
    },
    {
      "epoch": 1.9,
      "learning_rate": 2e-05,
      "loss": 139.8529,
      "step": 190
    },
    {
      "epoch": 1.91,
      "learning_rate": 2e-05,
      "loss": 86.0494,
      "step": 191
    },
    {
      "epoch": 1.92,
      "learning_rate": 2e-05,
      "loss": 104.6433,
      "step": 192
    },
    {
      "epoch": 1.93,
      "learning_rate": 2e-05,
      "loss": 82.0844,
      "step": 193
    },
    {
      "epoch": 1.94,
      "learning_rate": 2e-05,
      "loss": 98.7083,
      "step": 194
    },
    {
      "epoch": 1.95,
      "learning_rate": 2e-05,
      "loss": 103.9359,
      "step": 195
    },
    {
      "epoch": 1.96,
      "learning_rate": 2e-05,
      "loss": 138.1158,
      "step": 196
    },
    {
      "epoch": 1.97,
      "learning_rate": 2e-05,
      "loss": 122.3703,
      "step": 197
    },
    {
      "epoch": 1.98,
      "learning_rate": 2e-05,
      "loss": 92.0145,
      "step": 198
    },
    {
      "epoch": 1.99,
      "learning_rate": 2e-05,
      "loss": 86.6408,
      "step": 199
    },
    {
      "epoch": 2.0,
      "learning_rate": 2e-05,
      "loss": 118.3698,
      "step": 200
    },
    {
      "epoch": 2.0,
      "eval_loss": 123.61127471923828,
      "eval_runtime": 80.8471,
      "eval_samples_per_second": 19.741,
      "eval_steps_per_second": 1.237,
      "step": 200
    },
    {
      "epoch": 2.01,
      "learning_rate": 2e-05,
      "loss": 92.5948,
      "step": 201
    },
    {
      "epoch": 2.02,
      "learning_rate": 2e-05,
      "loss": 106.3745,
      "step": 202
    },
    {
      "epoch": 2.03,
      "learning_rate": 2e-05,
      "loss": 90.8155,
      "step": 203
    },
    {
      "epoch": 2.04,
      "learning_rate": 2e-05,
      "loss": 121.7541,
      "step": 204
    },
    {
      "epoch": 2.05,
      "learning_rate": 2e-05,
      "loss": 69.567,
      "step": 205
    },
    {
      "epoch": 2.06,
      "learning_rate": 2e-05,
      "loss": 112.1705,
      "step": 206
    },
    {
      "epoch": 2.07,
      "learning_rate": 2e-05,
      "loss": 99.2473,
      "step": 207
    },
    {
      "epoch": 2.08,
      "learning_rate": 2e-05,
      "loss": 109.7618,
      "step": 208
    },
    {
      "epoch": 2.09,
      "learning_rate": 2e-05,
      "loss": 89.1843,
      "step": 209
    },
    {
      "epoch": 2.1,
      "learning_rate": 2e-05,
      "loss": 133.0722,
      "step": 210
    },
    {
      "epoch": 2.11,
      "learning_rate": 2e-05,
      "loss": 70.0389,
      "step": 211
    },
    {
      "epoch": 2.12,
      "learning_rate": 2e-05,
      "loss": 127.7498,
      "step": 212
    },
    {
      "epoch": 2.13,
      "learning_rate": 2e-05,
      "loss": 175.2826,
      "step": 213
    },
    {
      "epoch": 2.14,
      "learning_rate": 2e-05,
      "loss": 98.1167,
      "step": 214
    },
    {
      "epoch": 2.15,
      "learning_rate": 2e-05,
      "loss": 131.4668,
      "step": 215
    },
    {
      "epoch": 2.16,
      "learning_rate": 2e-05,
      "loss": 163.0928,
      "step": 216
    },
    {
      "epoch": 2.17,
      "learning_rate": 2e-05,
      "loss": 139.924,
      "step": 217
    },
    {
      "epoch": 2.18,
      "learning_rate": 2e-05,
      "loss": 100.2071,
      "step": 218
    },
    {
      "epoch": 2.19,
      "learning_rate": 2e-05,
      "loss": 100.4305,
      "step": 219
    },
    {
      "epoch": 2.2,
      "learning_rate": 2e-05,
      "loss": 112.9075,
      "step": 220
    },
    {
      "epoch": 2.21,
      "learning_rate": 2e-05,
      "loss": 119.6214,
      "step": 221
    },
    {
      "epoch": 2.22,
      "learning_rate": 2e-05,
      "loss": 88.9599,
      "step": 222
    },
    {
      "epoch": 2.23,
      "learning_rate": 2e-05,
      "loss": 95.6708,
      "step": 223
    },
    {
      "epoch": 2.24,
      "learning_rate": 2e-05,
      "loss": 88.6898,
      "step": 224
    },
    {
      "epoch": 2.25,
      "learning_rate": 2e-05,
      "loss": 93.868,
      "step": 225
    },
    {
      "epoch": 2.26,
      "learning_rate": 2e-05,
      "loss": 108.3749,
      "step": 226
    },
    {
      "epoch": 2.27,
      "learning_rate": 2e-05,
      "loss": 94.327,
      "step": 227
    },
    {
      "epoch": 2.28,
      "learning_rate": 2e-05,
      "loss": 81.9183,
      "step": 228
    },
    {
      "epoch": 2.29,
      "learning_rate": 2e-05,
      "loss": 134.5717,
      "step": 229
    },
    {
      "epoch": 2.3,
      "learning_rate": 2e-05,
      "loss": 92.9158,
      "step": 230
    },
    {
      "epoch": 2.31,
      "learning_rate": 2e-05,
      "loss": 118.63,
      "step": 231
    },
    {
      "epoch": 2.32,
      "learning_rate": 2e-05,
      "loss": 90.0151,
      "step": 232
    },
    {
      "epoch": 2.33,
      "learning_rate": 2e-05,
      "loss": 128.4179,
      "step": 233
    },
    {
      "epoch": 2.34,
      "learning_rate": 2e-05,
      "loss": 132.9601,
      "step": 234
    },
    {
      "epoch": 2.35,
      "learning_rate": 2e-05,
      "loss": 110.689,
      "step": 235
    },
    {
      "epoch": 2.36,
      "learning_rate": 2e-05,
      "loss": 302.4031,
      "step": 236
    },
    {
      "epoch": 2.37,
      "learning_rate": 2e-05,
      "loss": 128.6338,
      "step": 237
    },
    {
      "epoch": 2.38,
      "learning_rate": 2e-05,
      "loss": 357.8953,
      "step": 238
    },
    {
      "epoch": 2.39,
      "learning_rate": 2e-05,
      "loss": 73.8348,
      "step": 239
    },
    {
      "epoch": 2.4,
      "learning_rate": 2e-05,
      "loss": 99.6502,
      "step": 240
    },
    {
      "epoch": 2.41,
      "learning_rate": 2e-05,
      "loss": 91.3185,
      "step": 241
    },
    {
      "epoch": 2.42,
      "learning_rate": 2e-05,
      "loss": 83.6264,
      "step": 242
    },
    {
      "epoch": 2.43,
      "learning_rate": 2e-05,
      "loss": 99.7315,
      "step": 243
    },
    {
      "epoch": 2.44,
      "learning_rate": 2e-05,
      "loss": 74.5451,
      "step": 244
    },
    {
      "epoch": 2.45,
      "learning_rate": 2e-05,
      "loss": 86.7471,
      "step": 245
    },
    {
      "epoch": 2.46,
      "learning_rate": 2e-05,
      "loss": 92.1487,
      "step": 246
    },
    {
      "epoch": 2.47,
      "learning_rate": 2e-05,
      "loss": 71.9365,
      "step": 247
    },
    {
      "epoch": 2.48,
      "learning_rate": 2e-05,
      "loss": 85.449,
      "step": 248
    },
    {
      "epoch": 2.49,
      "learning_rate": 2e-05,
      "loss": 74.9259,
      "step": 249
    },
    {
      "epoch": 2.5,
      "learning_rate": 2e-05,
      "loss": 212.1755,
      "step": 250
    },
    {
      "epoch": 2.51,
      "learning_rate": 2e-05,
      "loss": 141.1738,
      "step": 251
    },
    {
      "epoch": 2.52,
      "learning_rate": 2e-05,
      "loss": 68.8899,
      "step": 252
    },
    {
      "epoch": 2.53,
      "learning_rate": 2e-05,
      "loss": 89.9649,
      "step": 253
    },
    {
      "epoch": 2.54,
      "learning_rate": 2e-05,
      "loss": 85.3244,
      "step": 254
    },
    {
      "epoch": 2.55,
      "learning_rate": 2e-05,
      "loss": 102.01,
      "step": 255
    },
    {
      "epoch": 2.56,
      "learning_rate": 2e-05,
      "loss": 87.7782,
      "step": 256
    },
    {
      "epoch": 2.57,
      "learning_rate": 2e-05,
      "loss": 98.1192,
      "step": 257
    },
    {
      "epoch": 2.58,
      "learning_rate": 2e-05,
      "loss": 96.4169,
      "step": 258
    },
    {
      "epoch": 2.59,
      "learning_rate": 2e-05,
      "loss": 79.6971,
      "step": 259
    },
    {
      "epoch": 2.6,
      "learning_rate": 2e-05,
      "loss": 102.9386,
      "step": 260
    },
    {
      "epoch": 2.61,
      "learning_rate": 2e-05,
      "loss": 140.3832,
      "step": 261
    },
    {
      "epoch": 2.62,
      "learning_rate": 2e-05,
      "loss": 85.5271,
      "step": 262
    },
    {
      "epoch": 2.63,
      "learning_rate": 2e-05,
      "loss": 95.7376,
      "step": 263
    },
    {
      "epoch": 2.64,
      "learning_rate": 2e-05,
      "loss": 75.8506,
      "step": 264
    },
    {
      "epoch": 2.65,
      "learning_rate": 2e-05,
      "loss": 72.1343,
      "step": 265
    },
    {
      "epoch": 2.66,
      "learning_rate": 2e-05,
      "loss": 86.7606,
      "step": 266
    },
    {
      "epoch": 2.67,
      "learning_rate": 2e-05,
      "loss": 219.3665,
      "step": 267
    },
    {
      "epoch": 2.68,
      "learning_rate": 2e-05,
      "loss": 72.1039,
      "step": 268
    },
    {
      "epoch": 2.69,
      "learning_rate": 2e-05,
      "loss": 93.7625,
      "step": 269
    },
    {
      "epoch": 2.7,
      "learning_rate": 2e-05,
      "loss": 72.9779,
      "step": 270
    },
    {
      "epoch": 2.71,
      "learning_rate": 2e-05,
      "loss": 76.5516,
      "step": 271
    },
    {
      "epoch": 2.72,
      "learning_rate": 2e-05,
      "loss": 96.9595,
      "step": 272
    },
    {
      "epoch": 2.73,
      "learning_rate": 2e-05,
      "loss": 86.1967,
      "step": 273
    },
    {
      "epoch": 2.74,
      "learning_rate": 2e-05,
      "loss": 151.0705,
      "step": 274
    },
    {
      "epoch": 2.75,
      "learning_rate": 2e-05,
      "loss": 110.2601,
      "step": 275
    },
    {
      "epoch": 2.76,
      "learning_rate": 2e-05,
      "loss": 76.5183,
      "step": 276
    },
    {
      "epoch": 2.77,
      "learning_rate": 2e-05,
      "loss": 82.5286,
      "step": 277
    },
    {
      "epoch": 2.78,
      "learning_rate": 2e-05,
      "loss": 77.1184,
      "step": 278
    },
    {
      "epoch": 2.79,
      "learning_rate": 2e-05,
      "loss": 71.968,
      "step": 279
    },
    {
      "epoch": 2.8,
      "learning_rate": 2e-05,
      "loss": 84.9687,
      "step": 280
    },
    {
      "epoch": 2.81,
      "learning_rate": 2e-05,
      "loss": 82.9308,
      "step": 281
    },
    {
      "epoch": 2.82,
      "learning_rate": 2e-05,
      "loss": 75.31,
      "step": 282
    },
    {
      "epoch": 2.83,
      "learning_rate": 2e-05,
      "loss": 92.4162,
      "step": 283
    },
    {
      "epoch": 2.84,
      "learning_rate": 2e-05,
      "loss": 147.4037,
      "step": 284
    },
    {
      "epoch": 2.85,
      "learning_rate": 2e-05,
      "loss": 158.5738,
      "step": 285
    },
    {
      "epoch": 2.86,
      "learning_rate": 2e-05,
      "loss": 74.0794,
      "step": 286
    },
    {
      "epoch": 2.87,
      "learning_rate": 2e-05,
      "loss": 76.1675,
      "step": 287
    },
    {
      "epoch": 2.88,
      "learning_rate": 2e-05,
      "loss": 119.9409,
      "step": 288
    },
    {
      "epoch": 2.89,
      "learning_rate": 2e-05,
      "loss": 83.3971,
      "step": 289
    },
    {
      "epoch": 2.9,
      "learning_rate": 2e-05,
      "loss": 123.4943,
      "step": 290
    },
    {
      "epoch": 2.91,
      "learning_rate": 2e-05,
      "loss": 163.733,
      "step": 291
    },
    {
      "epoch": 2.92,
      "learning_rate": 2e-05,
      "loss": 84.0254,
      "step": 292
    },
    {
      "epoch": 2.93,
      "learning_rate": 2e-05,
      "loss": 100.9542,
      "step": 293
    },
    {
      "epoch": 2.94,
      "learning_rate": 2e-05,
      "loss": 79.933,
      "step": 294
    },
    {
      "epoch": 2.95,
      "learning_rate": 2e-05,
      "loss": 88.4858,
      "step": 295
    },
    {
      "epoch": 2.96,
      "learning_rate": 2e-05,
      "loss": 94.7497,
      "step": 296
    },
    {
      "epoch": 2.97,
      "learning_rate": 2e-05,
      "loss": 119.0287,
      "step": 297
    },
    {
      "epoch": 2.98,
      "learning_rate": 2e-05,
      "loss": 127.9053,
      "step": 298
    },
    {
      "epoch": 2.99,
      "learning_rate": 2e-05,
      "loss": 88.5058,
      "step": 299
    },
    {
      "epoch": 3.0,
      "learning_rate": 2e-05,
      "loss": 86.8415,
      "step": 300
    },
    {
      "epoch": 3.0,
      "eval_loss": 119.90803527832031,
      "eval_runtime": 80.9383,
      "eval_samples_per_second": 19.719,
      "eval_steps_per_second": 1.236,
      "step": 300
    },
    {
      "epoch": 3.01,
      "learning_rate": 2e-05,
      "loss": 103.3673,
      "step": 301
    },
    {
      "epoch": 3.02,
      "learning_rate": 2e-05,
      "loss": 107.5574,
      "step": 302
    },
    {
      "epoch": 3.03,
      "learning_rate": 2e-05,
      "loss": 95.606,
      "step": 303
    },
    {
      "epoch": 3.04,
      "learning_rate": 2e-05,
      "loss": 78.7149,
      "step": 304
    },
    {
      "epoch": 3.05,
      "learning_rate": 2e-05,
      "loss": 72.7483,
      "step": 305
    },
    {
      "epoch": 3.06,
      "learning_rate": 2e-05,
      "loss": 337.7869,
      "step": 306
    },
    {
      "epoch": 3.07,
      "learning_rate": 2e-05,
      "loss": 81.3133,
      "step": 307
    },
    {
      "epoch": 3.08,
      "learning_rate": 2e-05,
      "loss": 95.4,
      "step": 308
    },
    {
      "epoch": 3.09,
      "learning_rate": 2e-05,
      "loss": 117.039,
      "step": 309
    },
    {
      "epoch": 3.1,
      "learning_rate": 2e-05,
      "loss": 83.7925,
      "step": 310
    },
    {
      "epoch": 3.11,
      "learning_rate": 2e-05,
      "loss": 80.7962,
      "step": 311
    },
    {
      "epoch": 3.12,
      "learning_rate": 2e-05,
      "loss": 163.7045,
      "step": 312
    },
    {
      "epoch": 3.13,
      "learning_rate": 2e-05,
      "loss": 90.9779,
      "step": 313
    },
    {
      "epoch": 3.14,
      "learning_rate": 2e-05,
      "loss": 79.9667,
      "step": 314
    },
    {
      "epoch": 3.15,
      "learning_rate": 2e-05,
      "loss": 91.632,
      "step": 315
    },
    {
      "epoch": 3.16,
      "learning_rate": 2e-05,
      "loss": 69.9252,
      "step": 316
    },
    {
      "epoch": 3.17,
      "learning_rate": 2e-05,
      "loss": 94.818,
      "step": 317
    },
    {
      "epoch": 3.18,
      "learning_rate": 2e-05,
      "loss": 68.6146,
      "step": 318
    },
    {
      "epoch": 3.19,
      "learning_rate": 2e-05,
      "loss": 76.8548,
      "step": 319
    },
    {
      "epoch": 3.2,
      "learning_rate": 2e-05,
      "loss": 186.7993,
      "step": 320
    },
    {
      "epoch": 3.21,
      "learning_rate": 2e-05,
      "loss": 79.4541,
      "step": 321
    },
    {
      "epoch": 3.22,
      "learning_rate": 2e-05,
      "loss": 91.6424,
      "step": 322
    },
    {
      "epoch": 3.23,
      "learning_rate": 2e-05,
      "loss": 100.9761,
      "step": 323
    },
    {
      "epoch": 3.24,
      "learning_rate": 2e-05,
      "loss": 117.9019,
      "step": 324
    },
    {
      "epoch": 3.25,
      "learning_rate": 2e-05,
      "loss": 83.9062,
      "step": 325
    },
    {
      "epoch": 3.26,
      "learning_rate": 2e-05,
      "loss": 77.0838,
      "step": 326
    },
    {
      "epoch": 3.27,
      "learning_rate": 2e-05,
      "loss": 108.4668,
      "step": 327
    },
    {
      "epoch": 3.28,
      "learning_rate": 2e-05,
      "loss": 137.5493,
      "step": 328
    },
    {
      "epoch": 3.29,
      "learning_rate": 2e-05,
      "loss": 86.924,
      "step": 329
    },
    {
      "epoch": 3.3,
      "learning_rate": 2e-05,
      "loss": 101.1962,
      "step": 330
    },
    {
      "epoch": 3.31,
      "learning_rate": 2e-05,
      "loss": 81.7145,
      "step": 331
    },
    {
      "epoch": 3.32,
      "learning_rate": 2e-05,
      "loss": 106.9531,
      "step": 332
    },
    {
      "epoch": 3.33,
      "learning_rate": 2e-05,
      "loss": 89.0009,
      "step": 333
    },
    {
      "epoch": 3.34,
      "learning_rate": 2e-05,
      "loss": 111.5955,
      "step": 334
    },
    {
      "epoch": 3.35,
      "learning_rate": 2e-05,
      "loss": 108.5236,
      "step": 335
    },
    {
      "epoch": 3.36,
      "learning_rate": 2e-05,
      "loss": 78.6253,
      "step": 336
    },
    {
      "epoch": 3.37,
      "learning_rate": 2e-05,
      "loss": 92.5898,
      "step": 337
    },
    {
      "epoch": 3.38,
      "learning_rate": 2e-05,
      "loss": 73.9657,
      "step": 338
    },
    {
      "epoch": 3.39,
      "learning_rate": 2e-05,
      "loss": 76.5264,
      "step": 339
    },
    {
      "epoch": 3.4,
      "learning_rate": 2e-05,
      "loss": 76.9044,
      "step": 340
    },
    {
      "epoch": 3.41,
      "learning_rate": 2e-05,
      "loss": 73.2403,
      "step": 341
    },
    {
      "epoch": 3.42,
      "learning_rate": 2e-05,
      "loss": 94.6415,
      "step": 342
    },
    {
      "epoch": 3.43,
      "learning_rate": 2e-05,
      "loss": 76.2357,
      "step": 343
    },
    {
      "epoch": 3.44,
      "learning_rate": 2e-05,
      "loss": 76.6514,
      "step": 344
    },
    {
      "epoch": 3.45,
      "learning_rate": 2e-05,
      "loss": 65.6005,
      "step": 345
    },
    {
      "epoch": 3.46,
      "learning_rate": 2e-05,
      "loss": 108.3323,
      "step": 346
    },
    {
      "epoch": 3.47,
      "learning_rate": 2e-05,
      "loss": 93.3479,
      "step": 347
    },
    {
      "epoch": 3.48,
      "learning_rate": 2e-05,
      "loss": 90.2143,
      "step": 348
    },
    {
      "epoch": 3.49,
      "learning_rate": 2e-05,
      "loss": 127.5598,
      "step": 349
    },
    {
      "epoch": 3.5,
      "learning_rate": 2e-05,
      "loss": 86.6066,
      "step": 350
    },
    {
      "epoch": 3.51,
      "learning_rate": 2e-05,
      "loss": 73.5306,
      "step": 351
    },
    {
      "epoch": 3.52,
      "learning_rate": 2e-05,
      "loss": 60.8274,
      "step": 352
    },
    {
      "epoch": 3.53,
      "learning_rate": 2e-05,
      "loss": 75.1176,
      "step": 353
    },
    {
      "epoch": 3.54,
      "learning_rate": 2e-05,
      "loss": 108.5266,
      "step": 354
    },
    {
      "epoch": 3.55,
      "learning_rate": 2e-05,
      "loss": 69.0899,
      "step": 355
    },
    {
      "epoch": 3.56,
      "learning_rate": 2e-05,
      "loss": 92.8721,
      "step": 356
    },
    {
      "epoch": 3.57,
      "learning_rate": 2e-05,
      "loss": 90.7292,
      "step": 357
    },
    {
      "epoch": 3.58,
      "learning_rate": 2e-05,
      "loss": 55.9025,
      "step": 358
    },
    {
      "epoch": 3.59,
      "learning_rate": 2e-05,
      "loss": 80.1136,
      "step": 359
    },
    {
      "epoch": 3.6,
      "learning_rate": 2e-05,
      "loss": 129.3794,
      "step": 360
    },
    {
      "epoch": 3.61,
      "learning_rate": 2e-05,
      "loss": 65.8789,
      "step": 361
    },
    {
      "epoch": 3.62,
      "learning_rate": 2e-05,
      "loss": 89.3556,
      "step": 362
    },
    {
      "epoch": 3.63,
      "learning_rate": 2e-05,
      "loss": 104.0831,
      "step": 363
    },
    {
      "epoch": 3.64,
      "learning_rate": 2e-05,
      "loss": 90.5861,
      "step": 364
    },
    {
      "epoch": 3.65,
      "learning_rate": 2e-05,
      "loss": 70.7352,
      "step": 365
    },
    {
      "epoch": 3.66,
      "learning_rate": 2e-05,
      "loss": 113.6718,
      "step": 366
    },
    {
      "epoch": 3.67,
      "learning_rate": 2e-05,
      "loss": 81.7421,
      "step": 367
    },
    {
      "epoch": 3.68,
      "learning_rate": 2e-05,
      "loss": 59.3722,
      "step": 368
    },
    {
      "epoch": 3.69,
      "learning_rate": 2e-05,
      "loss": 74.7749,
      "step": 369
    },
    {
      "epoch": 3.7,
      "learning_rate": 2e-05,
      "loss": 98.0698,
      "step": 370
    },
    {
      "epoch": 3.71,
      "learning_rate": 2e-05,
      "loss": 81.5395,
      "step": 371
    },
    {
      "epoch": 3.72,
      "learning_rate": 2e-05,
      "loss": 70.3278,
      "step": 372
    },
    {
      "epoch": 3.73,
      "learning_rate": 2e-05,
      "loss": 77.4071,
      "step": 373
    },
    {
      "epoch": 3.74,
      "learning_rate": 2e-05,
      "loss": 203.0746,
      "step": 374
    },
    {
      "epoch": 3.75,
      "learning_rate": 2e-05,
      "loss": 248.7437,
      "step": 375
    },
    {
      "epoch": 3.76,
      "learning_rate": 2e-05,
      "loss": 73.0192,
      "step": 376
    },
    {
      "epoch": 3.77,
      "learning_rate": 2e-05,
      "loss": 72.3248,
      "step": 377
    },
    {
      "epoch": 3.78,
      "learning_rate": 2e-05,
      "loss": 77.9258,
      "step": 378
    },
    {
      "epoch": 3.79,
      "learning_rate": 2e-05,
      "loss": 72.0043,
      "step": 379
    },
    {
      "epoch": 3.8,
      "learning_rate": 2e-05,
      "loss": 112.4846,
      "step": 380
    },
    {
      "epoch": 3.81,
      "learning_rate": 2e-05,
      "loss": 69.6211,
      "step": 381
    },
    {
      "epoch": 3.82,
      "learning_rate": 2e-05,
      "loss": 89.8621,
      "step": 382
    },
    {
      "epoch": 3.83,
      "learning_rate": 2e-05,
      "loss": 75.9898,
      "step": 383
    },
    {
      "epoch": 3.84,
      "learning_rate": 2e-05,
      "loss": 122.9664,
      "step": 384
    },
    {
      "epoch": 3.85,
      "learning_rate": 2e-05,
      "loss": 76.6402,
      "step": 385
    },
    {
      "epoch": 3.86,
      "learning_rate": 2e-05,
      "loss": 88.9508,
      "step": 386
    },
    {
      "epoch": 3.87,
      "learning_rate": 2e-05,
      "loss": 89.0049,
      "step": 387
    },
    {
      "epoch": 3.88,
      "learning_rate": 2e-05,
      "loss": 81.3426,
      "step": 388
    },
    {
      "epoch": 3.89,
      "learning_rate": 2e-05,
      "loss": 70.0686,
      "step": 389
    },
    {
      "epoch": 3.9,
      "learning_rate": 2e-05,
      "loss": 87.9824,
      "step": 390
    },
    {
      "epoch": 3.91,
      "learning_rate": 2e-05,
      "loss": 82.3193,
      "step": 391
    },
    {
      "epoch": 3.92,
      "learning_rate": 2e-05,
      "loss": 89.7547,
      "step": 392
    },
    {
      "epoch": 3.93,
      "learning_rate": 2e-05,
      "loss": 61.2356,
      "step": 393
    },
    {
      "epoch": 3.94,
      "learning_rate": 2e-05,
      "loss": 76.2497,
      "step": 394
    },
    {
      "epoch": 3.95,
      "learning_rate": 2e-05,
      "loss": 90.9001,
      "step": 395
    },
    {
      "epoch": 3.96,
      "learning_rate": 2e-05,
      "loss": 74.1517,
      "step": 396
    },
    {
      "epoch": 3.97,
      "learning_rate": 2e-05,
      "loss": 74.2053,
      "step": 397
    },
    {
      "epoch": 3.98,
      "learning_rate": 2e-05,
      "loss": 63.4142,
      "step": 398
    },
    {
      "epoch": 3.99,
      "learning_rate": 2e-05,
      "loss": 99.6599,
      "step": 399
    },
    {
      "epoch": 4.0,
      "learning_rate": 2e-05,
      "loss": 80.2025,
      "step": 400
    },
    {
      "epoch": 4.0,
      "eval_loss": 101.20526885986328,
      "eval_runtime": 80.9879,
      "eval_samples_per_second": 19.707,
      "eval_steps_per_second": 1.235,
      "step": 400
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 20,
  "total_flos": 1.624099600371548e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}