{
  "best_global_step": 234,
  "best_metric": 0.3496658504009247,
  "best_model_checkpoint": "saves/test/checkpoint-234",
  "epoch": 1.0,
  "eval_steps": 13,
  "global_step": 249,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.020080321285140562,
      "grad_norm": 30.13371467590332,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.8454,
      "num_input_tokens_seen": 832,
      "step": 5
    },
    {
      "epoch": 0.040160642570281124,
      "grad_norm": 20.238059997558594,
      "learning_rate": 1.8e-05,
      "loss": 0.7689,
      "num_input_tokens_seen": 1760,
      "step": 10
    },
    {
      "epoch": 0.05220883534136546,
      "eval_loss": 0.6837943196296692,
      "eval_runtime": 0.9982,
      "eval_samples_per_second": 56.101,
      "eval_steps_per_second": 28.051,
      "num_input_tokens_seen": 2288,
      "step": 13
    },
    {
      "epoch": 0.060240963855421686,
      "grad_norm": 9.776243209838867,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.6956,
      "num_input_tokens_seen": 2608,
      "step": 15
    },
    {
      "epoch": 0.08032128514056225,
      "grad_norm": 17.104515075683594,
      "learning_rate": 3.8e-05,
      "loss": 0.5552,
      "num_input_tokens_seen": 3536,
      "step": 20
    },
    {
      "epoch": 0.10040160642570281,
      "grad_norm": 11.06977653503418,
      "learning_rate": 4.8e-05,
      "loss": 0.6557,
      "num_input_tokens_seen": 4496,
      "step": 25
    },
    {
      "epoch": 0.10441767068273092,
      "eval_loss": 0.46041956543922424,
      "eval_runtime": 0.7224,
      "eval_samples_per_second": 77.516,
      "eval_steps_per_second": 38.758,
      "num_input_tokens_seen": 4656,
      "step": 26
    },
    {
      "epoch": 0.12048192771084337,
      "grad_norm": 8.536067008972168,
      "learning_rate": 4.996067037544542e-05,
      "loss": 0.4952,
      "num_input_tokens_seen": 5424,
      "step": 30
    },
    {
      "epoch": 0.14056224899598393,
      "grad_norm": 7.301750659942627,
      "learning_rate": 4.980110583549062e-05,
      "loss": 0.3647,
      "num_input_tokens_seen": 6304,
      "step": 35
    },
    {
      "epoch": 0.1566265060240964,
      "eval_loss": 0.3834764063358307,
      "eval_runtime": 0.753,
      "eval_samples_per_second": 74.371,
      "eval_steps_per_second": 37.185,
      "num_input_tokens_seen": 6944,
      "step": 39
    },
    {
      "epoch": 0.1606425702811245,
      "grad_norm": 2.0650792121887207,
      "learning_rate": 4.951963201008076e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 7072,
      "step": 40
    },
    {
      "epoch": 0.18072289156626506,
      "grad_norm": 6.442427158355713,
      "learning_rate": 4.91176324775594e-05,
      "loss": 0.3811,
      "num_input_tokens_seen": 7856,
      "step": 45
    },
    {
      "epoch": 0.20080321285140562,
      "grad_norm": 9.193267822265625,
      "learning_rate": 4.8597083257709194e-05,
      "loss": 0.3506,
      "num_input_tokens_seen": 8880,
      "step": 50
    },
    {
      "epoch": 0.20883534136546184,
      "eval_loss": 0.38357535004615784,
      "eval_runtime": 0.8575,
      "eval_samples_per_second": 65.305,
      "eval_steps_per_second": 32.653,
      "num_input_tokens_seen": 9232,
      "step": 52
    },
    {
      "epoch": 0.22088353413654618,
      "grad_norm": 2.6214115619659424,
      "learning_rate": 4.796054309867053e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 9680,
      "step": 55
    },
    {
      "epoch": 0.24096385542168675,
      "grad_norm": 6.665003776550293,
      "learning_rate": 4.721114089947181e-05,
      "loss": 0.3351,
      "num_input_tokens_seen": 10576,
      "step": 60
    },
    {
      "epoch": 0.26104417670682734,
      "grad_norm": 2.606006383895874,
      "learning_rate": 4.6352560329995686e-05,
      "loss": 0.3084,
      "num_input_tokens_seen": 11424,
      "step": 65
    },
    {
      "epoch": 0.26104417670682734,
      "eval_loss": 0.36912769079208374,
      "eval_runtime": 0.9476,
      "eval_samples_per_second": 59.094,
      "eval_steps_per_second": 29.547,
      "num_input_tokens_seen": 11424,
      "step": 65
    },
    {
      "epoch": 0.28112449799196787,
      "grad_norm": 2.763232946395874,
      "learning_rate": 4.538902172398151e-05,
      "loss": 0.374,
      "num_input_tokens_seen": 12224,
      "step": 70
    },
    {
      "epoch": 0.30120481927710846,
      "grad_norm": 1.5414751768112183,
      "learning_rate": 4.4325261334068426e-05,
      "loss": 0.3649,
      "num_input_tokens_seen": 13168,
      "step": 75
    },
    {
      "epoch": 0.3132530120481928,
      "eval_loss": 0.3668709397315979,
      "eval_runtime": 0.8423,
      "eval_samples_per_second": 66.484,
      "eval_steps_per_second": 33.242,
      "num_input_tokens_seen": 13760,
      "step": 78
    },
    {
      "epoch": 0.321285140562249,
      "grad_norm": 7.335175514221191,
      "learning_rate": 4.316650805085068e-05,
      "loss": 0.3613,
      "num_input_tokens_seen": 14080,
      "step": 80
    },
    {
      "epoch": 0.3413654618473896,
      "grad_norm": 6.502185344696045,
      "learning_rate": 4.1918457700381855e-05,
      "loss": 0.3601,
      "num_input_tokens_seen": 15056,
      "step": 85
    },
    {
      "epoch": 0.3614457831325301,
      "grad_norm": 6.903041362762451,
      "learning_rate": 4.058724504646834e-05,
      "loss": 0.3612,
      "num_input_tokens_seen": 15904,
      "step": 90
    },
    {
      "epoch": 0.3654618473895582,
      "eval_loss": 0.36205965280532837,
      "eval_runtime": 0.7854,
      "eval_samples_per_second": 71.302,
      "eval_steps_per_second": 35.651,
      "num_input_tokens_seen": 16048,
      "step": 91
    },
    {
      "epoch": 0.3815261044176707,
      "grad_norm": 6.267019748687744,
      "learning_rate": 3.9179413635373897e-05,
      "loss": 0.3307,
      "num_input_tokens_seen": 16688,
      "step": 95
    },
    {
      "epoch": 0.40160642570281124,
      "grad_norm": 6.203659534454346,
      "learning_rate": 3.770188363116324e-05,
      "loss": 0.2896,
      "num_input_tokens_seen": 17552,
      "step": 100
    },
    {
      "epoch": 0.41767068273092367,
      "eval_loss": 0.37523433566093445,
      "eval_runtime": 0.7387,
      "eval_samples_per_second": 75.808,
      "eval_steps_per_second": 37.904,
      "num_input_tokens_seen": 18272,
      "step": 104
    },
    {
      "epoch": 0.42168674698795183,
      "grad_norm": 2.0911366939544678,
      "learning_rate": 3.616191779978907e-05,
      "loss": 0.3154,
      "num_input_tokens_seen": 18400,
      "step": 105
    },
    {
      "epoch": 0.44176706827309237,
      "grad_norm": 5.106855392456055,
      "learning_rate": 3.456708580912725e-05,
      "loss": 0.3694,
      "num_input_tokens_seen": 19456,
      "step": 110
    },
    {
      "epoch": 0.46184738955823296,
      "grad_norm": 8.49660587310791,
      "learning_rate": 3.292522702044221e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 20288,
      "step": 115
    },
    {
      "epoch": 0.46987951807228917,
      "eval_loss": 0.3691355884075165,
      "eval_runtime": 0.7901,
      "eval_samples_per_second": 70.875,
      "eval_steps_per_second": 35.437,
      "num_input_tokens_seen": 20656,
      "step": 117
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 2.1965019702911377,
      "learning_rate": 3.1244411954180676e-05,
      "loss": 0.3821,
      "num_input_tokens_seen": 21328,
      "step": 120
    },
    {
      "epoch": 0.5020080321285141,
      "grad_norm": 1.8442065715789795,
      "learning_rate": 2.9532902619507462e-05,
      "loss": 0.3707,
      "num_input_tokens_seen": 22304,
      "step": 125
    },
    {
      "epoch": 0.5220883534136547,
      "grad_norm": 6.632619857788086,
      "learning_rate": 2.7799111902582696e-05,
      "loss": 0.3591,
      "num_input_tokens_seen": 23056,
      "step": 130
    },
    {
      "epoch": 0.5220883534136547,
      "eval_loss": 0.3583144545555115,
      "eval_runtime": 0.9332,
      "eval_samples_per_second": 60.01,
      "eval_steps_per_second": 30.005,
      "num_input_tokens_seen": 23056,
      "step": 130
    },
    {
      "epoch": 0.5421686746987951,
      "grad_norm": 7.6188459396362305,
      "learning_rate": 2.6051562213206632e-05,
      "loss": 0.3438,
      "num_input_tokens_seen": 23840,
      "step": 135
    },
    {
      "epoch": 0.5622489959839357,
      "grad_norm": 1.9679665565490723,
      "learning_rate": 2.429884359310328e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 24832,
      "step": 140
    },
    {
      "epoch": 0.5742971887550201,
      "eval_loss": 0.3531239330768585,
      "eval_runtime": 0.8476,
      "eval_samples_per_second": 66.071,
      "eval_steps_per_second": 33.036,
      "num_input_tokens_seen": 25312,
      "step": 143
    },
    {
      "epoch": 0.5823293172690763,
      "grad_norm": 1.664465069770813,
      "learning_rate": 2.2549571491760986e-05,
      "loss": 0.3791,
      "num_input_tokens_seen": 25648,
      "step": 145
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 7.226130962371826,
      "learning_rate": 2.0812344417381595e-05,
      "loss": 0.3411,
      "num_input_tokens_seen": 26496,
      "step": 150
    },
    {
      "epoch": 0.6224899598393574,
      "grad_norm": 7.071829319000244,
      "learning_rate": 1.909570167110415e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 27392,
      "step": 155
    },
    {
      "epoch": 0.6265060240963856,
      "eval_loss": 0.35204586386680603,
      "eval_runtime": 0.805,
      "eval_samples_per_second": 69.565,
      "eval_steps_per_second": 34.782,
      "num_input_tokens_seen": 27552,
      "step": 156
    },
    {
      "epoch": 0.642570281124498,
      "grad_norm": 1.6194074153900146,
      "learning_rate": 1.7408081372259632e-05,
      "loss": 0.3653,
      "num_input_tokens_seen": 28272,
      "step": 160
    },
    {
      "epoch": 0.6626506024096386,
      "grad_norm": 5.642275810241699,
      "learning_rate": 1.5757778980982626e-05,
      "loss": 0.3318,
      "num_input_tokens_seen": 29184,
      "step": 165
    },
    {
      "epoch": 0.678714859437751,
      "eval_loss": 0.35015299916267395,
      "eval_runtime": 0.7303,
      "eval_samples_per_second": 76.682,
      "eval_steps_per_second": 38.341,
      "num_input_tokens_seen": 29984,
      "step": 169
    },
    {
      "epoch": 0.6827309236947792,
      "grad_norm": 1.980563759803772,
      "learning_rate": 1.4152906522061048e-05,
      "loss": 0.337,
      "num_input_tokens_seen": 30128,
      "step": 170
    },
    {
      "epoch": 0.7028112449799196,
      "grad_norm": 2.214860200881958,
      "learning_rate": 1.2601352710458313e-05,
      "loss": 0.3303,
      "num_input_tokens_seen": 30976,
      "step": 175
    },
    {
      "epoch": 0.7228915662650602,
      "grad_norm": 5.042296886444092,
      "learning_rate": 1.1110744174509952e-05,
      "loss": 0.3655,
      "num_input_tokens_seen": 31776,
      "step": 180
    },
    {
      "epoch": 0.7309236947791165,
      "eval_loss": 0.3543336093425751,
      "eval_runtime": 0.8576,
      "eval_samples_per_second": 65.301,
      "eval_steps_per_second": 32.65,
      "num_input_tokens_seen": 32080,
      "step": 182
    },
    {
      "epoch": 0.7429718875502008,
      "grad_norm": 7.1150078773498535,
      "learning_rate": 9.688407967401248e-06,
      "loss": 0.3755,
      "num_input_tokens_seen": 32608,
      "step": 185
    },
    {
      "epoch": 0.7630522088353414,
      "grad_norm": 6.07709264755249,
      "learning_rate": 8.341335551199902e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 33360,
      "step": 190
    },
    {
      "epoch": 0.7831325301204819,
      "grad_norm": 7.137906074523926,
      "learning_rate": 7.076148430479321e-06,
      "loss": 0.3703,
      "num_input_tokens_seen": 34176,
      "step": 195
    },
    {
      "epoch": 0.7831325301204819,
      "eval_loss": 0.35257503390312195,
      "eval_runtime": 0.7915,
      "eval_samples_per_second": 70.751,
      "eval_steps_per_second": 35.376,
      "num_input_tokens_seen": 34176,
      "step": 195
    },
    {
      "epoch": 0.8032128514056225,
      "grad_norm": 1.6914492845535278,
      "learning_rate": 5.899065604459814e-06,
      "loss": 0.355,
      "num_input_tokens_seen": 34992,
      "step": 200
    },
    {
      "epoch": 0.8232931726907631,
      "grad_norm": 1.7085988521575928,
      "learning_rate": 4.81587299765594e-06,
      "loss": 0.3585,
      "num_input_tokens_seen": 35888,
      "step": 205
    },
    {
      "epoch": 0.8353413654618473,
      "eval_loss": 0.3534963130950928,
      "eval_runtime": 0.8554,
      "eval_samples_per_second": 65.464,
      "eval_steps_per_second": 32.732,
      "num_input_tokens_seen": 36512,
      "step": 208
    },
    {
      "epoch": 0.8433734939759037,
      "grad_norm": 1.570613145828247,
      "learning_rate": 3.831895019292897e-06,
      "loss": 0.3715,
      "num_input_tokens_seen": 36848,
      "step": 210
    },
    {
      "epoch": 0.8634538152610441,
      "grad_norm": 2.1737060546875,
      "learning_rate": 2.9519683912911266e-06,
      "loss": 0.3719,
      "num_input_tokens_seen": 37888,
      "step": 215
    },
    {
      "epoch": 0.8835341365461847,
      "grad_norm": 1.8593710660934448,
      "learning_rate": 2.1804183734670277e-06,
      "loss": 0.3626,
      "num_input_tokens_seen": 38768,
      "step": 220
    },
    {
      "epoch": 0.8875502008032129,
      "eval_loss": 0.35169586539268494,
      "eval_runtime": 0.8105,
      "eval_samples_per_second": 69.096,
      "eval_steps_per_second": 34.548,
      "num_input_tokens_seen": 38912,
      "step": 221
    },
    {
      "epoch": 0.9036144578313253,
      "grad_norm": 1.1120434999465942,
      "learning_rate": 1.5210375028143097e-06,
      "loss": 0.3549,
      "num_input_tokens_seen": 39488,
      "step": 225
    },
    {
      "epoch": 0.9236947791164659,
      "grad_norm": 1.7197054624557495,
      "learning_rate": 9.770669513725128e-07,
      "loss": 0.3419,
      "num_input_tokens_seen": 40336,
      "step": 230
    },
    {
      "epoch": 0.9397590361445783,
      "eval_loss": 0.3496658504009247,
      "eval_runtime": 0.8315,
      "eval_samples_per_second": 67.352,
      "eval_steps_per_second": 33.676,
      "num_input_tokens_seen": 41120,
      "step": 234
    },
    {
      "epoch": 0.9437751004016064,
      "grad_norm": 1.5268943309783936,
      "learning_rate": 5.5118059431781e-07,
      "loss": 0.3635,
      "num_input_tokens_seen": 41328,
      "step": 235
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 1.8964738845825195,
      "learning_rate": 2.454718665888589e-07,
      "loss": 0.3463,
      "num_input_tokens_seen": 42176,
      "step": 240
    },
    {
      "epoch": 0.9839357429718876,
      "grad_norm": 1.2437193393707275,
      "learning_rate": 6.14434726538493e-08,
      "loss": 0.3311,
      "num_input_tokens_seen": 43312,
      "step": 245
    },
    {
      "epoch": 0.9919678714859438,
      "eval_loss": 0.3558541238307953,
      "eval_runtime": 0.7677,
      "eval_samples_per_second": 72.948,
      "eval_steps_per_second": 36.474,
      "num_input_tokens_seen": 43600,
      "step": 247
    },
    {
      "epoch": 1.0,
      "num_input_tokens_seen": 43904,
      "step": 249,
      "total_flos": 278458437992448.0,
      "train_loss": 0.3966633373474979,
      "train_runtime": 65.6097,
      "train_samples_per_second": 7.59,
      "train_steps_per_second": 3.795
    }
  ],
  "logging_steps": 5,
  "max_steps": 249,
  "num_input_tokens_seen": 43904,
  "num_train_epochs": 1,
  "save_steps": 13,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 278458437992448.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}