{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "global_step": 70000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.5576,
      "step": 500
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.154,
      "step": 1000
    },
    {
      "epoch": 0.04,
      "learning_rate": 6e-06,
      "loss": 0.1118,
      "step": 1500
    },
    {
      "epoch": 0.06,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.112,
      "step": 2000
    },
    {
      "epoch": 0.07,
      "learning_rate": 1e-05,
      "loss": 0.113,
      "step": 2500
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.2e-05,
      "loss": 0.1134,
      "step": 3000
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.4e-05,
      "loss": 0.1183,
      "step": 3500
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.1133,
      "step": 4000
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.8e-05,
      "loss": 0.1154,
      "step": 4500
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-05,
      "loss": 0.1111,
      "step": 5000
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.998561151079137e-05,
      "loss": 0.125,
      "step": 5500
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9971223021582734e-05,
      "loss": 0.1178,
      "step": 6000
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.9956834532374104e-05,
      "loss": 0.1077,
      "step": 6500
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.994244604316547e-05,
      "loss": 0.109,
      "step": 7000
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.9928057553956836e-05,
      "loss": 0.1001,
      "step": 7500
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.9913669064748202e-05,
      "loss": 0.1065,
      "step": 8000
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.989928057553957e-05,
      "loss": 0.1077,
      "step": 8500
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.9884892086330934e-05,
      "loss": 0.1036,
      "step": 9000
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.9870503597122304e-05,
      "loss": 0.1102,
      "step": 9500
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.985611510791367e-05,
      "loss": 0.0962,
      "step": 10000
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.9841726618705036e-05,
      "loss": 0.1148,
      "step": 10500
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.9827338129496406e-05,
      "loss": 0.1088,
      "step": 11000
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.9812949640287772e-05,
      "loss": 0.1042,
      "step": 11500
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.9798561151079138e-05,
      "loss": 0.0956,
      "step": 12000
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.9784172661870504e-05,
      "loss": 0.1015,
      "step": 12500
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.9769784172661874e-05,
      "loss": 0.1113,
      "step": 13000
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.975539568345324e-05,
      "loss": 0.1032,
      "step": 13500
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.9741007194244606e-05,
      "loss": 0.0984,
      "step": 14000
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.9726618705035972e-05,
      "loss": 0.1059,
      "step": 14500
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.971223021582734e-05,
      "loss": 0.1041,
      "step": 15000
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.9697841726618708e-05,
      "loss": 0.1035,
      "step": 15500
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.9683453237410074e-05,
      "loss": 0.0931,
      "step": 16000
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.966906474820144e-05,
      "loss": 0.1025,
      "step": 16500
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.9654676258992807e-05,
      "loss": 0.0967,
      "step": 17000
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.9640287769784173e-05,
      "loss": 0.0936,
      "step": 17500
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.9625899280575542e-05,
      "loss": 0.0947,
      "step": 18000
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.961151079136691e-05,
      "loss": 0.0959,
      "step": 18500
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.9597122302158275e-05,
      "loss": 0.093,
      "step": 19000
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.958273381294964e-05,
      "loss": 0.0976,
      "step": 19500
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.956834532374101e-05,
      "loss": 0.1014,
      "step": 20000
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.9553956834532377e-05,
      "loss": 0.0935,
      "step": 20500
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.9539568345323743e-05,
      "loss": 0.1038,
      "step": 21000
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.952517985611511e-05,
      "loss": 0.1059,
      "step": 21500
    },
    {
      "epoch": 0.63,
      "learning_rate": 1.9510791366906475e-05,
      "loss": 0.0937,
      "step": 22000
    },
    {
      "epoch": 0.64,
      "learning_rate": 1.9496402877697845e-05,
      "loss": 0.0965,
      "step": 22500
    },
    {
      "epoch": 0.66,
      "learning_rate": 1.948201438848921e-05,
      "loss": 0.1054,
      "step": 23000
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.9467625899280577e-05,
      "loss": 0.0966,
      "step": 23500
    },
    {
      "epoch": 0.69,
      "learning_rate": 1.9453237410071943e-05,
      "loss": 0.0952,
      "step": 24000
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.9438848920863313e-05,
      "loss": 0.0954,
      "step": 24500
    },
    {
      "epoch": 0.71,
      "learning_rate": 1.9424460431654675e-05,
      "loss": 0.1044,
      "step": 25000
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.9410071942446045e-05,
      "loss": 0.1026,
      "step": 25500
    },
    {
      "epoch": 0.74,
      "learning_rate": 1.939568345323741e-05,
      "loss": 0.0891,
      "step": 26000
    },
    {
      "epoch": 0.76,
      "learning_rate": 1.9381294964028777e-05,
      "loss": 0.092,
      "step": 26500
    },
    {
      "epoch": 0.77,
      "learning_rate": 1.9366906474820147e-05,
      "loss": 0.0913,
      "step": 27000
    },
    {
      "epoch": 0.79,
      "learning_rate": 1.9352517985611513e-05,
      "loss": 0.0901,
      "step": 27500
    },
    {
      "epoch": 0.8,
      "learning_rate": 1.933812949640288e-05,
      "loss": 0.1018,
      "step": 28000
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.9323741007194245e-05,
      "loss": 0.0927,
      "step": 28500
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.9309352517985615e-05,
      "loss": 0.096,
      "step": 29000
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.9294964028776978e-05,
      "loss": 0.0945,
      "step": 29500
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.9280575539568347e-05,
      "loss": 0.0918,
      "step": 30000
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.9266187050359713e-05,
      "loss": 0.0929,
      "step": 30500
    },
    {
      "epoch": 0.89,
      "learning_rate": 1.925179856115108e-05,
      "loss": 0.0975,
      "step": 31000
    },
    {
      "epoch": 0.9,
      "learning_rate": 1.923741007194245e-05,
      "loss": 0.0944,
      "step": 31500
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.9223021582733815e-05,
      "loss": 0.0911,
      "step": 32000
    },
    {
      "epoch": 0.93,
      "learning_rate": 1.920863309352518e-05,
      "loss": 0.0974,
      "step": 32500
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.9194244604316548e-05,
      "loss": 0.0914,
      "step": 33000
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.9179856115107914e-05,
      "loss": 0.0915,
      "step": 33500
    },
    {
      "epoch": 0.97,
      "learning_rate": 1.916546762589928e-05,
      "loss": 0.092,
      "step": 34000
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.915107913669065e-05,
      "loss": 0.0924,
      "step": 34500
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.9136690647482016e-05,
      "loss": 0.0884,
      "step": 35000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9749736842105263,
      "eval_loss": 0.09605808556079865,
      "eval_runtime": 105.081,
      "eval_samples_per_second": 361.626,
      "eval_steps_per_second": 22.602,
      "step": 35000
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.9122302158273382e-05,
      "loss": 0.064,
      "step": 35500
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.910791366906475e-05,
      "loss": 0.0677,
      "step": 36000
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.9093525179856118e-05,
      "loss": 0.0711,
      "step": 36500
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.9079136690647484e-05,
      "loss": 0.0651,
      "step": 37000
    },
    {
      "epoch": 1.07,
      "learning_rate": 1.906474820143885e-05,
      "loss": 0.0649,
      "step": 37500
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.9050359712230216e-05,
      "loss": 0.0691,
      "step": 38000
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.9035971223021586e-05,
      "loss": 0.0715,
      "step": 38500
    },
    {
      "epoch": 1.11,
      "learning_rate": 1.9021582733812952e-05,
      "loss": 0.0709,
      "step": 39000
    },
    {
      "epoch": 1.13,
      "learning_rate": 1.9007194244604318e-05,
      "loss": 0.0717,
      "step": 39500
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.8992805755395684e-05,
      "loss": 0.0728,
      "step": 40000
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.8978417266187054e-05,
      "loss": 0.0717,
      "step": 40500
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.8964028776978416e-05,
      "loss": 0.0761,
      "step": 41000
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.8949640287769786e-05,
      "loss": 0.0611,
      "step": 41500
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.8935251798561152e-05,
      "loss": 0.0755,
      "step": 42000
    },
    {
      "epoch": 1.21,
      "learning_rate": 1.892086330935252e-05,
      "loss": 0.0711,
      "step": 42500
    },
    {
      "epoch": 1.23,
      "learning_rate": 1.8906474820143888e-05,
      "loss": 0.0729,
      "step": 43000
    },
    {
      "epoch": 1.24,
      "learning_rate": 1.8892086330935254e-05,
      "loss": 0.0677,
      "step": 43500
    },
    {
      "epoch": 1.26,
      "learning_rate": 1.887769784172662e-05,
      "loss": 0.0782,
      "step": 44000
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.8863309352517986e-05,
      "loss": 0.0688,
      "step": 44500
    },
    {
      "epoch": 1.29,
      "learning_rate": 1.8848920863309356e-05,
      "loss": 0.0651,
      "step": 45000
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.883453237410072e-05,
      "loss": 0.071,
      "step": 45500
    },
    {
      "epoch": 1.31,
      "learning_rate": 1.8820143884892088e-05,
      "loss": 0.0703,
      "step": 46000
    },
    {
      "epoch": 1.33,
      "learning_rate": 1.8805755395683454e-05,
      "loss": 0.0727,
      "step": 46500
    },
    {
      "epoch": 1.34,
      "learning_rate": 1.879136690647482e-05,
      "loss": 0.0709,
      "step": 47000
    },
    {
      "epoch": 1.36,
      "learning_rate": 1.877697841726619e-05,
      "loss": 0.0751,
      "step": 47500
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.8762589928057556e-05,
      "loss": 0.0729,
      "step": 48000
    },
    {
      "epoch": 1.39,
      "learning_rate": 1.8748201438848923e-05,
      "loss": 0.0715,
      "step": 48500
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.873381294964029e-05,
      "loss": 0.0721,
      "step": 49000
    },
    {
      "epoch": 1.41,
      "learning_rate": 1.8719424460431658e-05,
      "loss": 0.0725,
      "step": 49500
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.870503597122302e-05,
      "loss": 0.0785,
      "step": 50000
    },
    {
      "epoch": 1.44,
      "learning_rate": 1.869064748201439e-05,
      "loss": 0.0795,
      "step": 50500
    },
    {
      "epoch": 1.46,
      "learning_rate": 1.8676258992805757e-05,
      "loss": 0.0659,
      "step": 51000
    },
    {
      "epoch": 1.47,
      "learning_rate": 1.8661870503597123e-05,
      "loss": 0.0718,
      "step": 51500
    },
    {
      "epoch": 1.49,
      "learning_rate": 1.8647482014388492e-05,
      "loss": 0.073,
      "step": 52000
    },
    {
      "epoch": 1.5,
      "learning_rate": 1.863309352517986e-05,
      "loss": 0.0777,
      "step": 52500
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.8618705035971225e-05,
      "loss": 0.0763,
      "step": 53000
    },
    {
      "epoch": 1.53,
      "learning_rate": 1.860431654676259e-05,
      "loss": 0.0711,
      "step": 53500
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.8589928057553957e-05,
      "loss": 0.0789,
      "step": 54000
    },
    {
      "epoch": 1.56,
      "learning_rate": 1.8575539568345323e-05,
      "loss": 0.0638,
      "step": 54500
    },
    {
      "epoch": 1.57,
      "learning_rate": 1.8561151079136693e-05,
      "loss": 0.0718,
      "step": 55000
    },
    {
      "epoch": 1.59,
      "learning_rate": 1.854676258992806e-05,
      "loss": 0.0811,
      "step": 55500
    },
    {
      "epoch": 1.6,
      "learning_rate": 1.8532374100719425e-05,
      "loss": 0.075,
      "step": 56000
    },
    {
      "epoch": 1.61,
      "learning_rate": 1.8517985611510795e-05,
      "loss": 0.0817,
      "step": 56500
    },
    {
      "epoch": 1.63,
      "learning_rate": 1.850359712230216e-05,
      "loss": 0.0712,
      "step": 57000
    },
    {
      "epoch": 1.64,
      "learning_rate": 1.8489208633093527e-05,
      "loss": 0.0727,
      "step": 57500
    },
    {
      "epoch": 1.66,
      "learning_rate": 1.8474820143884893e-05,
      "loss": 0.0774,
      "step": 58000
    },
    {
      "epoch": 1.67,
      "learning_rate": 1.846043165467626e-05,
      "loss": 0.079,
      "step": 58500
    },
    {
      "epoch": 1.69,
      "learning_rate": 1.8446043165467626e-05,
      "loss": 0.0712,
      "step": 59000
    },
    {
      "epoch": 1.7,
      "learning_rate": 1.8431654676258995e-05,
      "loss": 0.0782,
      "step": 59500
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.841726618705036e-05,
      "loss": 0.0678,
      "step": 60000
    },
    {
      "epoch": 1.73,
      "learning_rate": 1.8402877697841727e-05,
      "loss": 0.0792,
      "step": 60500
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.8388489208633097e-05,
      "loss": 0.0761,
      "step": 61000
    },
    {
      "epoch": 1.76,
      "learning_rate": 1.837410071942446e-05,
      "loss": 0.0775,
      "step": 61500
    },
    {
      "epoch": 1.77,
      "learning_rate": 1.835971223021583e-05,
      "loss": 0.069,
      "step": 62000
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.8345323741007196e-05,
      "loss": 0.0823,
      "step": 62500
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.833093525179856e-05,
      "loss": 0.0728,
      "step": 63000
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.8316546762589928e-05,
      "loss": 0.07,
      "step": 63500
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.8302158273381297e-05,
      "loss": 0.0747,
      "step": 64000
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.8287769784172664e-05,
      "loss": 0.073,
      "step": 64500
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.827338129496403e-05,
      "loss": 0.0711,
      "step": 65000
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.82589928057554e-05,
      "loss": 0.0723,
      "step": 65500
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.8244604316546762e-05,
      "loss": 0.0672,
      "step": 66000
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.823021582733813e-05,
      "loss": 0.076,
      "step": 66500
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.8215827338129498e-05,
      "loss": 0.0765,
      "step": 67000
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.8201438848920864e-05,
      "loss": 0.0735,
      "step": 67500
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.8187050359712234e-05,
      "loss": 0.079,
      "step": 68000
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.81726618705036e-05,
      "loss": 0.0699,
      "step": 68500
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.8158273381294966e-05,
      "loss": 0.0748,
      "step": 69000
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.8143884892086332e-05,
      "loss": 0.0779,
      "step": 69500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.8129496402877698e-05,
      "loss": 0.0701,
      "step": 70000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.975421052631579,
      "eval_loss": 0.11228523403406143,
      "eval_runtime": 104.8411,
      "eval_samples_per_second": 362.453,
      "eval_steps_per_second": 22.653,
      "step": 70000
    }
  ],
  "max_steps": 700000,
  "num_train_epochs": 20,
  "total_flos": 1.473421910016e+17,
  "trial_name": null,
  "trial_params": null
}