{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.05434782608695652,
  "eval_steps": 5,
  "global_step": 400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006793478260869565,
      "eval_loss": 4.499350547790527,
      "eval_runtime": 20.1523,
      "eval_samples_per_second": 975.424,
      "eval_steps_per_second": 7.642,
      "step": 5
    },
    {
      "epoch": 0.001358695652173913,
      "eval_loss": 4.498117923736572,
      "eval_runtime": 20.4966,
      "eval_samples_per_second": 959.037,
      "eval_steps_per_second": 7.513,
      "step": 10
    },
    {
      "epoch": 0.0020380434782608695,
      "eval_loss": 4.495996475219727,
      "eval_runtime": 21.0485,
      "eval_samples_per_second": 933.891,
      "eval_steps_per_second": 7.316,
      "step": 15
    },
    {
      "epoch": 0.002717391304347826,
      "eval_loss": 4.492975234985352,
      "eval_runtime": 21.3506,
      "eval_samples_per_second": 920.677,
      "eval_steps_per_second": 7.213,
      "step": 20
    },
    {
      "epoch": 0.0033967391304347825,
      "eval_loss": 4.489028453826904,
      "eval_runtime": 21.5145,
      "eval_samples_per_second": 913.661,
      "eval_steps_per_second": 7.158,
      "step": 25
    },
    {
      "epoch": 0.004076086956521739,
      "eval_loss": 4.484206199645996,
      "eval_runtime": 21.5836,
      "eval_samples_per_second": 910.738,
      "eval_steps_per_second": 7.135,
      "step": 30
    },
    {
      "epoch": 0.004755434782608696,
      "eval_loss": 4.478429317474365,
      "eval_runtime": 21.7892,
      "eval_samples_per_second": 902.143,
      "eval_steps_per_second": 7.068,
      "step": 35
    },
    {
      "epoch": 0.005434782608695652,
      "eval_loss": 4.471565246582031,
      "eval_runtime": 21.9727,
      "eval_samples_per_second": 894.608,
      "eval_steps_per_second": 7.009,
      "step": 40
    },
    {
      "epoch": 0.006114130434782609,
      "eval_loss": 4.463589668273926,
      "eval_runtime": 21.6519,
      "eval_samples_per_second": 907.863,
      "eval_steps_per_second": 7.113,
      "step": 45
    },
    {
      "epoch": 0.006793478260869565,
      "eval_loss": 4.454347610473633,
      "eval_runtime": 21.7784,
      "eval_samples_per_second": 902.593,
      "eval_steps_per_second": 7.071,
      "step": 50
    },
    {
      "epoch": 0.007472826086956522,
      "eval_loss": 4.443789958953857,
      "eval_runtime": 21.9874,
      "eval_samples_per_second": 894.012,
      "eval_steps_per_second": 7.004,
      "step": 55
    },
    {
      "epoch": 0.008152173913043478,
      "eval_loss": 4.432095527648926,
      "eval_runtime": 21.6944,
      "eval_samples_per_second": 906.087,
      "eval_steps_per_second": 7.099,
      "step": 60
    },
    {
      "epoch": 0.008831521739130434,
      "eval_loss": 4.419074535369873,
      "eval_runtime": 21.7766,
      "eval_samples_per_second": 902.668,
      "eval_steps_per_second": 7.072,
      "step": 65
    },
    {
      "epoch": 0.009510869565217392,
      "eval_loss": 4.404222011566162,
      "eval_runtime": 21.9472,
      "eval_samples_per_second": 895.649,
      "eval_steps_per_second": 7.017,
      "step": 70
    },
    {
      "epoch": 0.010190217391304348,
      "eval_loss": 4.387468338012695,
      "eval_runtime": 22.1085,
      "eval_samples_per_second": 889.115,
      "eval_steps_per_second": 6.966,
      "step": 75
    },
    {
      "epoch": 0.010869565217391304,
      "eval_loss": 4.368589878082275,
      "eval_runtime": 21.8154,
      "eval_samples_per_second": 901.063,
      "eval_steps_per_second": 7.059,
      "step": 80
    },
    {
      "epoch": 0.01154891304347826,
      "eval_loss": 4.347426891326904,
      "eval_runtime": 21.8937,
      "eval_samples_per_second": 897.84,
      "eval_steps_per_second": 7.034,
      "step": 85
    },
    {
      "epoch": 0.012228260869565218,
      "eval_loss": 4.32356595993042,
      "eval_runtime": 21.9031,
      "eval_samples_per_second": 897.453,
      "eval_steps_per_second": 7.031,
      "step": 90
    },
    {
      "epoch": 0.012907608695652174,
      "eval_loss": 4.296773433685303,
      "eval_runtime": 22.045,
      "eval_samples_per_second": 891.674,
      "eval_steps_per_second": 6.986,
      "step": 95
    },
    {
      "epoch": 0.01358695652173913,
      "grad_norm": 11.771966934204102,
      "learning_rate": 4.528985507246377e-07,
      "loss": 4.4995,
      "step": 100
    },
    {
      "epoch": 0.01358695652173913,
      "eval_loss": 4.266582489013672,
      "eval_runtime": 21.7748,
      "eval_samples_per_second": 902.742,
      "eval_steps_per_second": 7.072,
      "step": 100
    },
    {
      "epoch": 0.014266304347826086,
      "eval_loss": 4.232626914978027,
      "eval_runtime": 21.7463,
      "eval_samples_per_second": 903.924,
      "eval_steps_per_second": 7.082,
      "step": 105
    },
    {
      "epoch": 0.014945652173913044,
      "eval_loss": 4.194711208343506,
      "eval_runtime": 22.0767,
      "eval_samples_per_second": 890.396,
      "eval_steps_per_second": 6.976,
      "step": 110
    },
    {
      "epoch": 0.015625,
      "eval_loss": 4.151618957519531,
      "eval_runtime": 21.7284,
      "eval_samples_per_second": 904.667,
      "eval_steps_per_second": 7.087,
      "step": 115
    },
    {
      "epoch": 0.016304347826086956,
      "eval_loss": 4.102933883666992,
      "eval_runtime": 21.8375,
      "eval_samples_per_second": 900.149,
      "eval_steps_per_second": 7.052,
      "step": 120
    },
    {
      "epoch": 0.016983695652173912,
      "eval_loss": 4.047597408294678,
      "eval_runtime": 21.8792,
      "eval_samples_per_second": 898.434,
      "eval_steps_per_second": 7.039,
      "step": 125
    },
    {
      "epoch": 0.017663043478260868,
      "eval_loss": 3.9850316047668457,
      "eval_runtime": 22.0074,
      "eval_samples_per_second": 893.2,
      "eval_steps_per_second": 6.998,
      "step": 130
    },
    {
      "epoch": 0.018342391304347828,
      "eval_loss": 3.9161784648895264,
      "eval_runtime": 21.9382,
      "eval_samples_per_second": 896.016,
      "eval_steps_per_second": 7.02,
      "step": 135
    },
    {
      "epoch": 0.019021739130434784,
      "eval_loss": 3.839725971221924,
      "eval_runtime": 21.8144,
      "eval_samples_per_second": 901.1,
      "eval_steps_per_second": 7.06,
      "step": 140
    },
    {
      "epoch": 0.01970108695652174,
      "eval_loss": 3.7521822452545166,
      "eval_runtime": 21.6841,
      "eval_samples_per_second": 906.516,
      "eval_steps_per_second": 7.102,
      "step": 145
    },
    {
      "epoch": 0.020380434782608696,
      "eval_loss": 3.652061700820923,
      "eval_runtime": 21.8254,
      "eval_samples_per_second": 900.648,
      "eval_steps_per_second": 7.056,
      "step": 150
    },
    {
      "epoch": 0.021059782608695652,
      "eval_loss": 3.5387539863586426,
      "eval_runtime": 21.8927,
      "eval_samples_per_second": 897.878,
      "eval_steps_per_second": 7.034,
      "step": 155
    },
    {
      "epoch": 0.021739130434782608,
      "eval_loss": 3.4114441871643066,
      "eval_runtime": 21.8999,
      "eval_samples_per_second": 897.585,
      "eval_steps_per_second": 7.032,
      "step": 160
    },
    {
      "epoch": 0.022418478260869564,
      "eval_loss": 3.2700562477111816,
      "eval_runtime": 22.0501,
      "eval_samples_per_second": 891.469,
      "eval_steps_per_second": 6.984,
      "step": 165
    },
    {
      "epoch": 0.02309782608695652,
      "eval_loss": 3.114734172821045,
      "eval_runtime": 21.7106,
      "eval_samples_per_second": 905.412,
      "eval_steps_per_second": 7.093,
      "step": 170
    },
    {
      "epoch": 0.02377717391304348,
      "eval_loss": 2.9471371173858643,
      "eval_runtime": 21.8419,
      "eval_samples_per_second": 899.966,
      "eval_steps_per_second": 7.051,
      "step": 175
    },
    {
      "epoch": 0.024456521739130436,
      "eval_loss": 2.7709779739379883,
      "eval_runtime": 21.8693,
      "eval_samples_per_second": 898.839,
      "eval_steps_per_second": 7.042,
      "step": 180
    },
    {
      "epoch": 0.025135869565217392,
      "eval_loss": 2.590932607650757,
      "eval_runtime": 22.0818,
      "eval_samples_per_second": 890.189,
      "eval_steps_per_second": 6.974,
      "step": 185
    },
    {
      "epoch": 0.025815217391304348,
      "eval_loss": 2.4126803874969482,
      "eval_runtime": 21.8508,
      "eval_samples_per_second": 899.601,
      "eval_steps_per_second": 7.048,
      "step": 190
    },
    {
      "epoch": 0.026494565217391304,
      "eval_loss": 2.243886947631836,
      "eval_runtime": 21.9268,
      "eval_samples_per_second": 896.483,
      "eval_steps_per_second": 7.023,
      "step": 195
    },
    {
      "epoch": 0.02717391304347826,
      "grad_norm": 25.208101272583008,
      "learning_rate": 9.057971014492754e-07,
      "loss": 3.6918,
      "step": 200
    },
    {
      "epoch": 0.02717391304347826,
      "eval_loss": 2.0869383811950684,
      "eval_runtime": 21.9994,
      "eval_samples_per_second": 893.524,
      "eval_steps_per_second": 7.0,
      "step": 200
    },
    {
      "epoch": 0.027853260869565216,
      "eval_loss": 1.9476724863052368,
      "eval_runtime": 21.8137,
      "eval_samples_per_second": 901.132,
      "eval_steps_per_second": 7.06,
      "step": 205
    },
    {
      "epoch": 0.028532608695652172,
      "eval_loss": 1.8273799419403076,
      "eval_runtime": 21.8665,
      "eval_samples_per_second": 898.954,
      "eval_steps_per_second": 7.043,
      "step": 210
    },
    {
      "epoch": 0.029211956521739132,
      "eval_loss": 1.71555495262146,
      "eval_runtime": 21.8011,
      "eval_samples_per_second": 901.651,
      "eval_steps_per_second": 7.064,
      "step": 215
    },
    {
      "epoch": 0.029891304347826088,
      "eval_loss": 1.621083378791809,
      "eval_runtime": 21.9545,
      "eval_samples_per_second": 895.354,
      "eval_steps_per_second": 7.015,
      "step": 220
    },
    {
      "epoch": 0.030570652173913044,
      "eval_loss": 1.5416371822357178,
      "eval_runtime": 21.6636,
      "eval_samples_per_second": 907.377,
      "eval_steps_per_second": 7.109,
      "step": 225
    },
    {
      "epoch": 0.03125,
      "eval_loss": 1.4732381105422974,
      "eval_runtime": 21.6731,
      "eval_samples_per_second": 906.976,
      "eval_steps_per_second": 7.106,
      "step": 230
    },
    {
      "epoch": 0.03192934782608696,
      "eval_loss": 1.417628288269043,
      "eval_runtime": 21.9177,
      "eval_samples_per_second": 896.857,
      "eval_steps_per_second": 7.026,
      "step": 235
    },
    {
      "epoch": 0.03260869565217391,
      "eval_loss": 1.3701567649841309,
      "eval_runtime": 22.0593,
      "eval_samples_per_second": 891.097,
      "eval_steps_per_second": 6.981,
      "step": 240
    },
    {
      "epoch": 0.03328804347826087,
      "eval_loss": 1.326924204826355,
      "eval_runtime": 21.7518,
      "eval_samples_per_second": 903.694,
      "eval_steps_per_second": 7.08,
      "step": 245
    },
    {
      "epoch": 0.033967391304347824,
      "eval_loss": 1.2892364263534546,
      "eval_runtime": 21.807,
      "eval_samples_per_second": 901.406,
      "eval_steps_per_second": 7.062,
      "step": 250
    },
    {
      "epoch": 0.034646739130434784,
      "eval_loss": 1.256325125694275,
      "eval_runtime": 21.977,
      "eval_samples_per_second": 894.436,
      "eval_steps_per_second": 7.007,
      "step": 255
    },
    {
      "epoch": 0.035326086956521736,
      "eval_loss": 1.2281286716461182,
      "eval_runtime": 21.9123,
      "eval_samples_per_second": 897.075,
      "eval_steps_per_second": 7.028,
      "step": 260
    },
    {
      "epoch": 0.036005434782608696,
      "eval_loss": 1.2023590803146362,
      "eval_runtime": 21.7217,
      "eval_samples_per_second": 904.947,
      "eval_steps_per_second": 7.09,
      "step": 265
    },
    {
      "epoch": 0.036684782608695655,
      "eval_loss": 1.1796302795410156,
      "eval_runtime": 21.8625,
      "eval_samples_per_second": 899.119,
      "eval_steps_per_second": 7.044,
      "step": 270
    },
    {
      "epoch": 0.03736413043478261,
      "eval_loss": 1.1600725650787354,
      "eval_runtime": 22.083,
      "eval_samples_per_second": 890.141,
      "eval_steps_per_second": 6.974,
      "step": 275
    },
    {
      "epoch": 0.03804347826086957,
      "eval_loss": 1.1428364515304565,
      "eval_runtime": 21.958,
      "eval_samples_per_second": 895.209,
      "eval_steps_per_second": 7.013,
      "step": 280
    },
    {
      "epoch": 0.03872282608695652,
      "eval_loss": 1.1271179914474487,
      "eval_runtime": 21.9261,
      "eval_samples_per_second": 896.511,
      "eval_steps_per_second": 7.024,
      "step": 285
    },
    {
      "epoch": 0.03940217391304348,
      "eval_loss": 1.1129034757614136,
      "eval_runtime": 21.8882,
      "eval_samples_per_second": 898.065,
      "eval_steps_per_second": 7.036,
      "step": 290
    },
    {
      "epoch": 0.04008152173913043,
      "eval_loss": 1.100168228149414,
      "eval_runtime": 21.8776,
      "eval_samples_per_second": 898.498,
      "eval_steps_per_second": 7.039,
      "step": 295
    },
    {
      "epoch": 0.04076086956521739,
      "grad_norm": 14.775556564331055,
      "learning_rate": 1.3586956521739131e-06,
      "loss": 1.7071,
      "step": 300
    },
    {
      "epoch": 0.04076086956521739,
      "eval_loss": 1.087642788887024,
      "eval_runtime": 21.8501,
      "eval_samples_per_second": 899.629,
      "eval_steps_per_second": 7.048,
      "step": 300
    },
    {
      "epoch": 0.041440217391304345,
      "eval_loss": 1.076054334640503,
      "eval_runtime": 21.9626,
      "eval_samples_per_second": 895.023,
      "eval_steps_per_second": 7.012,
      "step": 305
    },
    {
      "epoch": 0.042119565217391304,
      "eval_loss": 1.06575608253479,
      "eval_runtime": 22.0431,
      "eval_samples_per_second": 891.751,
      "eval_steps_per_second": 6.986,
      "step": 310
    },
    {
      "epoch": 0.042798913043478264,
      "eval_loss": 1.055428385734558,
      "eval_runtime": 21.9208,
      "eval_samples_per_second": 896.727,
      "eval_steps_per_second": 7.025,
      "step": 315
    },
    {
      "epoch": 0.043478260869565216,
      "eval_loss": 1.0457570552825928,
      "eval_runtime": 21.8629,
      "eval_samples_per_second": 899.103,
      "eval_steps_per_second": 7.044,
      "step": 320
    },
    {
      "epoch": 0.044157608695652176,
      "eval_loss": 1.0364974737167358,
      "eval_runtime": 21.8717,
      "eval_samples_per_second": 898.741,
      "eval_steps_per_second": 7.041,
      "step": 325
    },
    {
      "epoch": 0.04483695652173913,
      "eval_loss": 1.0275565385818481,
      "eval_runtime": 22.0891,
      "eval_samples_per_second": 889.895,
      "eval_steps_per_second": 6.972,
      "step": 330
    },
    {
      "epoch": 0.04551630434782609,
      "eval_loss": 1.0180351734161377,
      "eval_runtime": 21.8038,
      "eval_samples_per_second": 901.541,
      "eval_steps_per_second": 7.063,
      "step": 335
    },
    {
      "epoch": 0.04619565217391304,
      "eval_loss": 1.0085766315460205,
      "eval_runtime": 21.9023,
      "eval_samples_per_second": 897.484,
      "eval_steps_per_second": 7.031,
      "step": 340
    },
    {
      "epoch": 0.046875,
      "eval_loss": 0.9996482729911804,
      "eval_runtime": 21.6674,
      "eval_samples_per_second": 907.216,
      "eval_steps_per_second": 7.107,
      "step": 345
    },
    {
      "epoch": 0.04755434782608696,
      "eval_loss": 0.9920084476470947,
      "eval_runtime": 21.9122,
      "eval_samples_per_second": 897.081,
      "eval_steps_per_second": 7.028,
      "step": 350
    },
    {
      "epoch": 0.04823369565217391,
      "eval_loss": 0.9846301674842834,
      "eval_runtime": 21.8368,
      "eval_samples_per_second": 900.179,
      "eval_steps_per_second": 7.052,
      "step": 355
    },
    {
      "epoch": 0.04891304347826087,
      "eval_loss": 0.9782416820526123,
      "eval_runtime": 21.8836,
      "eval_samples_per_second": 898.252,
      "eval_steps_per_second": 7.037,
      "step": 360
    },
    {
      "epoch": 0.049592391304347824,
      "eval_loss": 0.9715093374252319,
      "eval_runtime": 22.0024,
      "eval_samples_per_second": 893.402,
      "eval_steps_per_second": 6.999,
      "step": 365
    },
    {
      "epoch": 0.050271739130434784,
      "eval_loss": 0.9640997648239136,
      "eval_runtime": 21.7358,
      "eval_samples_per_second": 904.359,
      "eval_steps_per_second": 7.085,
      "step": 370
    },
    {
      "epoch": 0.050951086956521736,
      "eval_loss": 0.9571502804756165,
      "eval_runtime": 21.7367,
      "eval_samples_per_second": 904.322,
      "eval_steps_per_second": 7.085,
      "step": 375
    },
    {
      "epoch": 0.051630434782608696,
      "eval_loss": 0.9503456950187683,
      "eval_runtime": 21.899,
      "eval_samples_per_second": 897.623,
      "eval_steps_per_second": 7.032,
      "step": 380
    },
    {
      "epoch": 0.052309782608695655,
      "eval_loss": 0.9444239139556885,
      "eval_runtime": 22.041,
      "eval_samples_per_second": 891.838,
      "eval_steps_per_second": 6.987,
      "step": 385
    },
    {
      "epoch": 0.05298913043478261,
      "eval_loss": 0.938381552696228,
      "eval_runtime": 21.867,
      "eval_samples_per_second": 898.934,
      "eval_steps_per_second": 7.043,
      "step": 390
    },
    {
      "epoch": 0.05366847826086957,
      "eval_loss": 0.932941734790802,
      "eval_runtime": 21.9262,
      "eval_samples_per_second": 896.508,
      "eval_steps_per_second": 7.024,
      "step": 395
    },
    {
      "epoch": 0.05434782608695652,
      "grad_norm": 12.524120330810547,
      "learning_rate": 1.8115942028985508e-06,
      "loss": 1.2083,
      "step": 400
    },
    {
      "epoch": 0.05434782608695652,
      "eval_loss": 0.9275586009025574,
      "eval_runtime": 21.8696,
      "eval_samples_per_second": 898.826,
      "eval_steps_per_second": 7.042,
      "step": 400
    }
  ],
  "logging_steps": 100,
  "max_steps": 22080,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 5,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}