{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.95,
  "eval_steps": 500,
  "global_step": 316,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0625,
      "grad_norm": 2.886897563934326,
      "learning_rate": 1.3636363636363635e-05,
      "loss": 1.3691,
      "mean_token_accuracy": 0.6758252620697022,
      "num_tokens": 967205.0,
      "step": 5
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.7570586800575256,
      "learning_rate": 3.068181818181818e-05,
      "loss": 1.269,
      "mean_token_accuracy": 0.6787961244583129,
      "num_tokens": 1921321.0,
      "step": 10
    },
    {
      "epoch": 0.1875,
      "grad_norm": 0.5072895288467407,
      "learning_rate": 4.7727272727272724e-05,
      "loss": 1.1276,
      "mean_token_accuracy": 0.7031140804290772,
      "num_tokens": 2869536.0,
      "step": 15
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.43829378485679626,
      "learning_rate": 6.477272727272727e-05,
      "loss": 1.0321,
      "mean_token_accuracy": 0.7193317890167237,
      "num_tokens": 3858500.0,
      "step": 20
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.4101383090019226,
      "learning_rate": 7.499374865230175e-05,
      "loss": 0.9679,
      "mean_token_accuracy": 0.734985601902008,
      "num_tokens": 4789698.0,
      "step": 25
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.4972609281539917,
      "learning_rate": 7.492345290159602e-05,
      "loss": 0.8265,
      "mean_token_accuracy": 0.7632619500160217,
      "num_tokens": 5778900.0,
      "step": 30
    },
    {
      "epoch": 0.4375,
      "grad_norm": 0.3812519609928131,
      "learning_rate": 7.477524313699733e-05,
      "loss": 0.777,
      "mean_token_accuracy": 0.780980110168457,
      "num_tokens": 6747243.0,
      "step": 35
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.41406524181365967,
      "learning_rate": 7.454953106098704e-05,
      "loss": 0.745,
      "mean_token_accuracy": 0.7827757239341736,
      "num_tokens": 7718219.0,
      "step": 40
    },
    {
      "epoch": 0.5625,
      "grad_norm": 0.4349099099636078,
      "learning_rate": 7.42469436647895e-05,
      "loss": 0.7068,
      "mean_token_accuracy": 0.7929372310638427,
      "num_tokens": 8655461.0,
      "step": 45
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.4021725356578827,
      "learning_rate": 7.386832148669304e-05,
      "loss": 0.7122,
      "mean_token_accuracy": 0.7939839363098145,
      "num_tokens": 9604900.0,
      "step": 50
    },
    {
      "epoch": 0.6875,
      "grad_norm": 0.4050525426864624,
      "learning_rate": 7.341471627717199e-05,
      "loss": 0.6396,
      "mean_token_accuracy": 0.813041603565216,
      "num_tokens": 10597014.0,
      "step": 55
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4006189703941345,
      "learning_rate": 7.288738807729602e-05,
      "loss": 0.6521,
      "mean_token_accuracy": 0.8042582392692565,
      "num_tokens": 11601956.0,
      "step": 60
    },
    {
      "epoch": 0.8125,
      "grad_norm": 0.41143348813056946,
      "learning_rate": 7.228780171854215e-05,
      "loss": 0.638,
      "mean_token_accuracy": 0.8092883229255676,
      "num_tokens": 12547547.0,
      "step": 65
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.534580409526825,
      "learning_rate": 7.161762275373244e-05,
      "loss": 0.5964,
      "mean_token_accuracy": 0.8182763457298279,
      "num_tokens": 13518585.0,
      "step": 70
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.5157734751701355,
      "learning_rate": 7.087871283040086e-05,
      "loss": 0.6566,
      "mean_token_accuracy": 0.7993115663528443,
      "num_tokens": 14487359.0,
      "step": 75
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.47575053572654724,
      "learning_rate": 7.007312451944087e-05,
      "loss": 0.5802,
      "mean_token_accuracy": 0.8218901753425598,
      "num_tokens": 15456737.0,
      "step": 80
    },
    {
      "epoch": 1.0625,
      "grad_norm": 0.5268984436988831,
      "learning_rate": 6.92030956133992e-05,
      "loss": 0.5475,
      "mean_token_accuracy": 0.8278343200683593,
      "num_tokens": 16398357.0,
      "step": 85
    },
    {
      "epoch": 1.125,
      "grad_norm": 0.4960674047470093,
      "learning_rate": 6.827104291025441e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.828590726852417,
      "num_tokens": 17358500.0,
      "step": 90
    },
    {
      "epoch": 1.1875,
      "grad_norm": 0.5320209860801697,
      "learning_rate": 6.727955549994725e-05,
      "loss": 0.5207,
      "mean_token_accuracy": 0.8370408773422241,
      "num_tokens": 18342498.0,
      "step": 95
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.579660177230835,
      "learning_rate": 6.623138757231244e-05,
      "loss": 0.5369,
      "mean_token_accuracy": 0.835267984867096,
      "num_tokens": 19327440.0,
      "step": 100
    },
    {
      "epoch": 1.3125,
      "grad_norm": 0.551972508430481,
      "learning_rate": 6.512945076638986e-05,
      "loss": 0.5268,
      "mean_token_accuracy": 0.8328914523124695,
      "num_tokens": 20296574.0,
      "step": 105
    },
    {
      "epoch": 1.375,
      "grad_norm": 0.47652754187583923,
      "learning_rate": 6.397680608236752e-05,
      "loss": 0.5298,
      "mean_token_accuracy": 0.8306650876998901,
      "num_tokens": 21246957.0,
      "step": 110
    },
    {
      "epoch": 1.4375,
      "grad_norm": 0.5899170637130737,
      "learning_rate": 6.277665537862383e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8339821338653565,
      "num_tokens": 22189101.0,
      "step": 115
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.5031387805938721,
      "learning_rate": 6.153233247748881e-05,
      "loss": 0.4929,
      "mean_token_accuracy": 0.8448952674865723,
      "num_tokens": 23178128.0,
      "step": 120
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.5996953248977661,
      "learning_rate": 6.0247293904430945e-05,
      "loss": 0.4807,
      "mean_token_accuracy": 0.8454113602638245,
      "num_tokens": 24150209.0,
      "step": 125
    },
    {
      "epoch": 1.625,
      "grad_norm": 0.5869215726852417,
      "learning_rate": 5.892510928639492e-05,
      "loss": 0.4759,
      "mean_token_accuracy": 0.8487537622451782,
      "num_tokens": 25114253.0,
      "step": 130
    },
    {
      "epoch": 1.6875,
      "grad_norm": 0.6149603128433228,
      "learning_rate": 5.756945143596188e-05,
      "loss": 0.4714,
      "mean_token_accuracy": 0.8475272536277771,
      "num_tokens": 26051613.0,
      "step": 135
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.6760532259941101,
      "learning_rate": 5.6184086148876894e-05,
      "loss": 0.5068,
      "mean_token_accuracy": 0.8413119912147522,
      "num_tokens": 26990255.0,
      "step": 140
    },
    {
      "epoch": 1.8125,
      "grad_norm": 0.6503332853317261,
      "learning_rate": 5.477286174328442e-05,
      "loss": 0.4402,
      "mean_token_accuracy": 0.8568204998970032,
      "num_tokens": 27999730.0,
      "step": 145
    },
    {
      "epoch": 1.875,
      "grad_norm": 0.611965000629425,
      "learning_rate": 5.333969836972997e-05,
      "loss": 0.4814,
      "mean_token_accuracy": 0.8444833040237427,
      "num_tokens": 28967267.0,
      "step": 150
    },
    {
      "epoch": 1.9375,
      "grad_norm": 0.6395599842071533,
      "learning_rate": 5.188857712162322e-05,
      "loss": 0.4613,
      "mean_token_accuracy": 0.8521966099739074,
      "num_tokens": 29916924.0,
      "step": 155
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.670397937297821,
      "learning_rate": 5.042352897641175e-05,
      "loss": 0.4742,
      "mean_token_accuracy": 0.8465383648872375,
      "num_tokens": 30895018.0,
      "step": 160
    },
    {
      "epoch": 2.0625,
      "grad_norm": 0.6061509847640991,
      "learning_rate": 4.894862359818504e-05,
      "loss": 0.4264,
      "mean_token_accuracy": 0.8613315701484681,
      "num_tokens": 31832051.0,
      "step": 165
    },
    {
      "epoch": 2.125,
      "grad_norm": 0.6346914768218994,
      "learning_rate": 4.746795803281343e-05,
      "loss": 0.4153,
      "mean_token_accuracy": 0.8654327511787414,
      "num_tokens": 32782084.0,
      "step": 170
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.6239711046218872,
      "learning_rate": 4.59856453270249e-05,
      "loss": 0.4147,
      "mean_token_accuracy": 0.8632163763046264,
      "num_tokens": 33740572.0,
      "step": 175
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.5688380002975464,
      "learning_rate": 4.450580310303404e-05,
      "loss": 0.4035,
      "mean_token_accuracy": 0.864958119392395,
      "num_tokens": 34760832.0,
      "step": 180
    },
    {
      "epoch": 2.3125,
      "grad_norm": 0.6970889568328857,
      "learning_rate": 4.303254212046112e-05,
      "loss": 0.4278,
      "mean_token_accuracy": 0.8623119592666626,
      "num_tokens": 35675387.0,
      "step": 185
    },
    {
      "epoch": 2.375,
      "grad_norm": 0.6932367086410522,
      "learning_rate": 4.156995485731436e-05,
      "loss": 0.4167,
      "mean_token_accuracy": 0.8627482533454895,
      "num_tokens": 36654691.0,
      "step": 190
    },
    {
      "epoch": 2.4375,
      "grad_norm": 0.74919193983078,
      "learning_rate": 4.0122104141755454e-05,
      "loss": 0.4041,
      "mean_token_accuracy": 0.8695870995521545,
      "num_tokens": 37623031.0,
      "step": 195
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.6283291578292847,
      "learning_rate": 3.869301186622754e-05,
      "loss": 0.3863,
      "mean_token_accuracy": 0.8726918458938598,
      "num_tokens": 38620801.0,
      "step": 200
    },
    {
      "epoch": 2.5625,
      "grad_norm": 0.5696513652801514,
      "learning_rate": 3.728664781529596e-05,
      "loss": 0.3979,
      "mean_token_accuracy": 0.8664975047111512,
      "num_tokens": 39619709.0,
      "step": 205
    },
    {
      "epoch": 2.625,
      "grad_norm": 0.804267168045044,
      "learning_rate": 3.590691863823604e-05,
      "loss": 0.4252,
      "mean_token_accuracy": 0.8591230988502503,
      "num_tokens": 40603368.0,
      "step": 210
    },
    {
      "epoch": 2.6875,
      "grad_norm": 0.8184316158294678,
      "learning_rate": 3.4557656997000574e-05,
      "loss": 0.4003,
      "mean_token_accuracy": 0.8659304857254029,
      "num_tokens": 41548896.0,
      "step": 215
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.8127630352973938,
      "learning_rate": 3.324261091971178e-05,
      "loss": 0.3871,
      "mean_token_accuracy": 0.8709235548973083,
      "num_tokens": 42524953.0,
      "step": 220
    },
    {
      "epoch": 2.8125,
      "grad_norm": 0.7622901797294617,
      "learning_rate": 3.1965433389252365e-05,
      "loss": 0.3845,
      "mean_token_accuracy": 0.8702649950981141,
      "num_tokens": 43484492.0,
      "step": 225
    },
    {
      "epoch": 2.875,
      "grad_norm": 0.6691771149635315,
      "learning_rate": 3.0729672195876615e-05,
      "loss": 0.3907,
      "mean_token_accuracy": 0.8697134971618652,
      "num_tokens": 44459752.0,
      "step": 230
    },
    {
      "epoch": 2.9375,
      "grad_norm": 0.8623871207237244,
      "learning_rate": 2.953876008202943e-05,
      "loss": 0.4045,
      "mean_token_accuracy": 0.8645989656448364,
      "num_tokens": 45374136.0,
      "step": 235
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6397231817245483,
      "learning_rate": 2.8396005206749356e-05,
      "loss": 0.3915,
      "mean_token_accuracy": 0.8694759726524353,
      "num_tokens": 46347054.0,
      "step": 240
    },
    {
      "epoch": 3.0625,
      "grad_norm": 0.6764381527900696,
      "learning_rate": 2.7304581956143917e-05,
      "loss": 0.3568,
      "mean_token_accuracy": 0.8802210927009583,
      "num_tokens": 47320044.0,
      "step": 245
    },
    {
      "epoch": 3.125,
      "grad_norm": 0.7345840334892273,
      "learning_rate": 2.626752212546434e-05,
      "loss": 0.3549,
      "mean_token_accuracy": 0.878583550453186,
      "num_tokens": 48309769.0,
      "step": 250
    },
    {
      "epoch": 3.1875,
      "grad_norm": 0.8476727604866028,
      "learning_rate": 2.528770649727422e-05,
      "loss": 0.3341,
      "mean_token_accuracy": 0.8889845728874206,
      "num_tokens": 49276272.0,
      "step": 255
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.825459361076355,
      "learning_rate": 2.4367856839106895e-05,
      "loss": 0.3821,
      "mean_token_accuracy": 0.8741106748580932,
      "num_tokens": 50223204.0,
      "step": 260
    },
    {
      "epoch": 3.3125,
      "grad_norm": 0.7627800703048706,
      "learning_rate": 2.3510528342840405e-05,
      "loss": 0.3394,
      "mean_token_accuracy": 0.8844169974327087,
      "num_tokens": 51217236.0,
      "step": 265
    },
    {
      "epoch": 3.375,
      "grad_norm": 0.7601186037063599,
      "learning_rate": 2.2718102526792474e-05,
      "loss": 0.334,
      "mean_token_accuracy": 0.8848990678787232,
      "num_tokens": 52214250.0,
      "step": 270
    },
    {
      "epoch": 3.4375,
      "grad_norm": 1.0226414203643799,
      "learning_rate": 2.1992780620252184e-05,
      "loss": 0.3371,
      "mean_token_accuracy": 0.8848458886146545,
      "num_tokens": 53147457.0,
      "step": 275
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.6799741387367249,
      "learning_rate": 2.1336577448825122e-05,
      "loss": 0.3391,
      "mean_token_accuracy": 0.884827721118927,
      "num_tokens": 54144048.0,
      "step": 280
    },
    {
      "epoch": 3.5625,
      "grad_norm": 0.7272220849990845,
      "learning_rate": 2.075131583757742e-05,
      "loss": 0.3392,
      "mean_token_accuracy": 0.8837083220481873,
      "num_tokens": 55118118.0,
      "step": 285
    },
    {
      "epoch": 3.625,
      "grad_norm": 0.7519774436950684,
      "learning_rate": 2.0238621547525902e-05,
      "loss": 0.3381,
      "mean_token_accuracy": 0.8849397540092468,
      "num_tokens": 56034052.0,
      "step": 290
    },
    {
      "epoch": 3.6875,
      "grad_norm": 0.917468249797821,
      "learning_rate": 1.979991875953994e-05,
      "loss": 0.3109,
      "mean_token_accuracy": 0.8913192272186279,
      "num_tokens": 56994081.0,
      "step": 295
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.7441599369049072,
      "learning_rate": 1.943642611819996e-05,
      "loss": 0.3364,
      "mean_token_accuracy": 0.8865782380104065,
      "num_tokens": 57997199.0,
      "step": 300
    },
    {
      "epoch": 3.8125,
      "grad_norm": 0.9184499382972717,
      "learning_rate": 1.9149153346602176e-05,
      "loss": 0.3488,
      "mean_token_accuracy": 0.8831375479698181,
      "num_tokens": 58922928.0,
      "step": 305
    },
    {
      "epoch": 3.875,
      "grad_norm": 0.8689393997192383,
      "learning_rate": 1.893889844151295e-05,
      "loss": 0.3406,
      "mean_token_accuracy": 0.884752631187439,
      "num_tokens": 59846729.0,
      "step": 310
    },
    {
      "epoch": 3.9375,
      "grad_norm": 0.822659432888031,
      "learning_rate": 1.8806245456664362e-05,
      "loss": 0.336,
      "mean_token_accuracy": 0.8861027598381043,
      "num_tokens": 60832044.0,
      "step": 315
    }
  ],
  "logging_steps": 5,
  "max_steps": 320,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.0379419406848492e+18,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": null
}