{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 457,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01095290251916758,
      "grad_norm": 1.3321008682250977,
      "learning_rate": 1.0434782608695653e-06,
      "loss": 1.3073,
      "step": 5
    },
    {
      "epoch": 0.02190580503833516,
      "grad_norm": 1.1412581205368042,
      "learning_rate": 2.347826086956522e-06,
      "loss": 1.3358,
      "step": 10
    },
    {
      "epoch": 0.03285870755750274,
      "grad_norm": 0.9217167496681213,
      "learning_rate": 3.6521739130434787e-06,
      "loss": 1.2979,
      "step": 15
    },
    {
      "epoch": 0.04381161007667032,
      "grad_norm": 0.6123777627944946,
      "learning_rate": 4.956521739130435e-06,
      "loss": 1.2163,
      "step": 20
    },
    {
      "epoch": 0.054764512595837894,
      "grad_norm": 0.9250121116638184,
      "learning_rate": 6.260869565217391e-06,
      "loss": 1.264,
      "step": 25
    },
    {
      "epoch": 0.06571741511500548,
      "grad_norm": 0.5497902035713196,
      "learning_rate": 7.5652173913043475e-06,
      "loss": 1.225,
      "step": 30
    },
    {
      "epoch": 0.07667031763417305,
      "grad_norm": 0.602593183517456,
      "learning_rate": 8.869565217391306e-06,
      "loss": 1.1959,
      "step": 35
    },
    {
      "epoch": 0.08762322015334063,
      "grad_norm": 0.5248430371284485,
      "learning_rate": 1.017391304347826e-05,
      "loss": 1.2186,
      "step": 40
    },
    {
      "epoch": 0.09857612267250822,
      "grad_norm": 0.5515073537826538,
      "learning_rate": 1.1478260869565218e-05,
      "loss": 1.1472,
      "step": 45
    },
    {
      "epoch": 0.10952902519167579,
      "grad_norm": 1.2510318756103516,
      "learning_rate": 1.2782608695652173e-05,
      "loss": 1.1666,
      "step": 50
    },
    {
      "epoch": 0.12048192771084337,
      "grad_norm": 0.49298274517059326,
      "learning_rate": 1.4086956521739131e-05,
      "loss": 1.2019,
      "step": 55
    },
    {
      "epoch": 0.13143483023001096,
      "grad_norm": 0.4494762122631073,
      "learning_rate": 1.5391304347826088e-05,
      "loss": 1.173,
      "step": 60
    },
    {
      "epoch": 0.14238773274917854,
      "grad_norm": 0.6092376112937927,
      "learning_rate": 1.6695652173913044e-05,
      "loss": 1.1323,
      "step": 65
    },
    {
      "epoch": 0.1533406352683461,
      "grad_norm": 0.4575347602367401,
      "learning_rate": 1.8e-05,
      "loss": 1.1164,
      "step": 70
    },
    {
      "epoch": 0.16429353778751368,
      "grad_norm": 1.1347095966339111,
      "learning_rate": 1.9304347826086957e-05,
      "loss": 1.193,
      "step": 75
    },
    {
      "epoch": 0.17524644030668127,
      "grad_norm": 0.45952585339546204,
      "learning_rate": 2.0608695652173913e-05,
      "loss": 1.1403,
      "step": 80
    },
    {
      "epoch": 0.18619934282584885,
      "grad_norm": 0.49114152789115906,
      "learning_rate": 2.191304347826087e-05,
      "loss": 1.1535,
      "step": 85
    },
    {
      "epoch": 0.19715224534501644,
      "grad_norm": 0.5684127807617188,
      "learning_rate": 2.3217391304347826e-05,
      "loss": 1.1398,
      "step": 90
    },
    {
      "epoch": 0.20810514786418402,
      "grad_norm": 0.5454962849617004,
      "learning_rate": 2.4521739130434786e-05,
      "loss": 1.0928,
      "step": 95
    },
    {
      "epoch": 0.21905805038335158,
      "grad_norm": 0.491813600063324,
      "learning_rate": 2.582608695652174e-05,
      "loss": 1.0858,
      "step": 100
    },
    {
      "epoch": 0.23001095290251916,
      "grad_norm": 0.5732094645500183,
      "learning_rate": 2.7130434782608695e-05,
      "loss": 1.0765,
      "step": 105
    },
    {
      "epoch": 0.24096385542168675,
      "grad_norm": 0.5950366258621216,
      "learning_rate": 2.8434782608695652e-05,
      "loss": 1.149,
      "step": 110
    },
    {
      "epoch": 0.25191675794085433,
      "grad_norm": 0.536044180393219,
      "learning_rate": 2.973913043478261e-05,
      "loss": 1.0859,
      "step": 115
    },
    {
      "epoch": 0.2628696604600219,
      "grad_norm": 0.5769652724266052,
      "learning_rate": 2.999974848707377e-05,
      "loss": 1.0433,
      "step": 120
    },
    {
      "epoch": 0.2738225629791895,
      "grad_norm": 0.5500301718711853,
      "learning_rate": 2.9998726730266534e-05,
      "loss": 1.051,
      "step": 125
    },
    {
      "epoch": 0.2847754654983571,
      "grad_norm": 0.5901659727096558,
      "learning_rate": 2.9996919063517692e-05,
      "loss": 1.0166,
      "step": 130
    },
    {
      "epoch": 0.29572836801752467,
      "grad_norm": 0.6402804851531982,
      "learning_rate": 2.999432558154617e-05,
      "loss": 1.078,
      "step": 135
    },
    {
      "epoch": 0.3066812705366922,
      "grad_norm": 0.5208508372306824,
      "learning_rate": 2.9990946420246395e-05,
      "loss": 1.0167,
      "step": 140
    },
    {
      "epoch": 0.3176341730558598,
      "grad_norm": 0.6010607481002808,
      "learning_rate": 2.9986781756681146e-05,
      "loss": 1.0342,
      "step": 145
    },
    {
      "epoch": 0.32858707557502737,
      "grad_norm": 0.5925593376159668,
      "learning_rate": 2.9981831809072318e-05,
      "loss": 0.9813,
      "step": 150
    },
    {
      "epoch": 0.33953997809419495,
      "grad_norm": 0.7121065855026245,
      "learning_rate": 2.997609683678947e-05,
      "loss": 0.9768,
      "step": 155
    },
    {
      "epoch": 0.35049288061336253,
      "grad_norm": 0.5926089882850647,
      "learning_rate": 2.996957714033622e-05,
      "loss": 1.0256,
      "step": 160
    },
    {
      "epoch": 0.3614457831325301,
      "grad_norm": 0.6746991276741028,
      "learning_rate": 2.9962273061334515e-05,
      "loss": 0.9901,
      "step": 165
    },
    {
      "epoch": 0.3723986856516977,
      "grad_norm": 0.6542229056358337,
      "learning_rate": 2.9954184982506734e-05,
      "loss": 0.9707,
      "step": 170
    },
    {
      "epoch": 0.3833515881708653,
      "grad_norm": 0.7646006345748901,
      "learning_rate": 2.9945313327655622e-05,
      "loss": 0.9826,
      "step": 175
    },
    {
      "epoch": 0.39430449069003287,
      "grad_norm": 0.6529242396354675,
      "learning_rate": 2.9935658561642083e-05,
      "loss": 0.9659,
      "step": 180
    },
    {
      "epoch": 0.40525739320920046,
      "grad_norm": 0.6744837760925293,
      "learning_rate": 2.9925221190360836e-05,
      "loss": 0.9517,
      "step": 185
    },
    {
      "epoch": 0.41621029572836804,
      "grad_norm": 0.6999198198318481,
      "learning_rate": 2.9914001760713898e-05,
      "loss": 1.0202,
      "step": 190
    },
    {
      "epoch": 0.42716319824753557,
      "grad_norm": 0.7155759334564209,
      "learning_rate": 2.9902000860581916e-05,
      "loss": 0.9947,
      "step": 195
    },
    {
      "epoch": 0.43811610076670315,
      "grad_norm": 0.6805759072303772,
      "learning_rate": 2.9889219118793393e-05,
      "loss": 0.9751,
      "step": 200
    },
    {
      "epoch": 0.44906900328587074,
      "grad_norm": 0.7391874194145203,
      "learning_rate": 2.9875657205091703e-05,
      "loss": 1.0113,
      "step": 205
    },
    {
      "epoch": 0.4600219058050383,
      "grad_norm": 0.7946105003356934,
      "learning_rate": 2.9861315830100023e-05,
      "loss": 0.952,
      "step": 210
    },
    {
      "epoch": 0.4709748083242059,
      "grad_norm": 0.6928731203079224,
      "learning_rate": 2.984619574528408e-05,
      "loss": 0.9207,
      "step": 215
    },
    {
      "epoch": 0.4819277108433735,
      "grad_norm": 0.7348776459693909,
      "learning_rate": 2.9830297742912798e-05,
      "loss": 0.933,
      "step": 220
    },
    {
      "epoch": 0.4928806133625411,
      "grad_norm": 0.7400885820388794,
      "learning_rate": 2.981362265601676e-05,
      "loss": 0.8837,
      "step": 225
    },
    {
      "epoch": 0.5038335158817087,
      "grad_norm": 0.79625004529953,
      "learning_rate": 2.9796171358344578e-05,
      "loss": 0.9132,
      "step": 230
    },
    {
      "epoch": 0.5147864184008762,
      "grad_norm": 0.8105800747871399,
      "learning_rate": 2.9777944764317082e-05,
      "loss": 0.9753,
      "step": 235
    },
    {
      "epoch": 0.5257393209200438,
      "grad_norm": 0.8556815981864929,
      "learning_rate": 2.9758943828979444e-05,
      "loss": 0.9439,
      "step": 240
    },
    {
      "epoch": 0.5366922234392114,
      "grad_norm": 0.8349775671958923,
      "learning_rate": 2.9739169547951116e-05,
      "loss": 0.897,
      "step": 245
    },
    {
      "epoch": 0.547645125958379,
      "grad_norm": 0.6672224402427673,
      "learning_rate": 2.971862295737364e-05,
      "loss": 0.8832,
      "step": 250
    },
    {
      "epoch": 0.5585980284775466,
      "grad_norm": 0.7600496411323547,
      "learning_rate": 2.9697305133856394e-05,
      "loss": 0.8834,
      "step": 255
    },
    {
      "epoch": 0.5695509309967142,
      "grad_norm": 0.8782379031181335,
      "learning_rate": 2.9675217194420174e-05,
      "loss": 0.906,
      "step": 260
    },
    {
      "epoch": 0.5805038335158818,
      "grad_norm": 0.7369331121444702,
      "learning_rate": 2.9652360296438612e-05,
      "loss": 0.8909,
      "step": 265
    },
    {
      "epoch": 0.5914567360350493,
      "grad_norm": 0.7942291498184204,
      "learning_rate": 2.9628735637577612e-05,
      "loss": 0.8707,
      "step": 270
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 0.782035231590271,
      "learning_rate": 2.960434445573253e-05,
      "loss": 0.8819,
      "step": 275
    },
    {
      "epoch": 0.6133625410733844,
      "grad_norm": 0.9603850841522217,
      "learning_rate": 2.957918802896334e-05,
      "loss": 0.8689,
      "step": 280
    },
    {
      "epoch": 0.624315443592552,
      "grad_norm": 0.7741917967796326,
      "learning_rate": 2.9553267675427667e-05,
      "loss": 0.8543,
      "step": 285
    },
    {
      "epoch": 0.6352683461117196,
      "grad_norm": 0.7536712884902954,
      "learning_rate": 2.952658475331169e-05,
      "loss": 0.8515,
      "step": 290
    },
    {
      "epoch": 0.6462212486308871,
      "grad_norm": 0.8260005712509155,
      "learning_rate": 2.9499140660759018e-05,
      "loss": 0.8181,
      "step": 295
    },
    {
      "epoch": 0.6571741511500547,
      "grad_norm": 0.8287461400032043,
      "learning_rate": 2.9470936835797392e-05,
      "loss": 0.8527,
      "step": 300
    },
    {
      "epoch": 0.6681270536692223,
      "grad_norm": 0.8062161207199097,
      "learning_rate": 2.944197475626335e-05,
      "loss": 0.8389,
      "step": 305
    },
    {
      "epoch": 0.6790799561883899,
      "grad_norm": 0.8538839817047119,
      "learning_rate": 2.94122559397248e-05,
      "loss": 0.8177,
      "step": 310
    },
    {
      "epoch": 0.6900328587075575,
      "grad_norm": 0.7861185073852539,
      "learning_rate": 2.9381781943401477e-05,
      "loss": 0.839,
      "step": 315
    },
    {
      "epoch": 0.7009857612267251,
      "grad_norm": 0.9167773723602295,
      "learning_rate": 2.935055436408337e-05,
      "loss": 0.7825,
      "step": 320
    },
    {
      "epoch": 0.7119386637458927,
      "grad_norm": 0.807101309299469,
      "learning_rate": 2.9318574838047045e-05,
      "loss": 0.7862,
      "step": 325
    },
    {
      "epoch": 0.7228915662650602,
      "grad_norm": 0.8477482199668884,
      "learning_rate": 2.9285845040969906e-05,
      "loss": 0.8001,
      "step": 330
    },
    {
      "epoch": 0.7338444687842278,
      "grad_norm": 1.5396567583084106,
      "learning_rate": 2.9252366687842383e-05,
      "loss": 0.7666,
      "step": 335
    },
    {
      "epoch": 0.7447973713033954,
      "grad_norm": 0.8536133170127869,
      "learning_rate": 2.921814153287809e-05,
      "loss": 0.767,
      "step": 340
    },
    {
      "epoch": 0.755750273822563,
      "grad_norm": 0.8579494953155518,
      "learning_rate": 2.918317136942189e-05,
      "loss": 0.8055,
      "step": 345
    },
    {
      "epoch": 0.7667031763417306,
      "grad_norm": 0.8801035284996033,
      "learning_rate": 2.9147458029855933e-05,
      "loss": 0.7404,
      "step": 350
    },
    {
      "epoch": 0.7776560788608982,
      "grad_norm": 0.8523620963096619,
      "learning_rate": 2.911100338550364e-05,
      "loss": 0.7868,
      "step": 355
    },
    {
      "epoch": 0.7886089813800657,
      "grad_norm": 0.8762844800949097,
      "learning_rate": 2.9073809346531645e-05,
      "loss": 0.786,
      "step": 360
    },
    {
      "epoch": 0.7995618838992333,
      "grad_norm": 0.8992785215377808,
      "learning_rate": 2.903587786184972e-05,
      "loss": 0.7726,
      "step": 365
    },
    {
      "epoch": 0.8105147864184009,
      "grad_norm": 0.8587476015090942,
      "learning_rate": 2.899721091900863e-05,
      "loss": 0.7611,
      "step": 370
    },
    {
      "epoch": 0.8214676889375685,
      "grad_norm": 0.9412994980812073,
      "learning_rate": 2.895781054409602e-05,
      "loss": 0.7366,
      "step": 375
    },
    {
      "epoch": 0.8324205914567361,
      "grad_norm": 0.8805333375930786,
      "learning_rate": 2.891767880163022e-05,
      "loss": 0.7364,
      "step": 380
    },
    {
      "epoch": 0.8433734939759037,
      "grad_norm": 0.944273054599762,
      "learning_rate": 2.8876817794452087e-05,
      "loss": 0.7449,
      "step": 385
    },
    {
      "epoch": 0.8543263964950711,
      "grad_norm": 0.8578236103057861,
      "learning_rate": 2.883522966361482e-05,
      "loss": 0.7288,
      "step": 390
    },
    {
      "epoch": 0.8652792990142387,
      "grad_norm": 0.8459666967391968,
      "learning_rate": 2.8792916588271762e-05,
      "loss": 0.7062,
      "step": 395
    },
    {
      "epoch": 0.8762322015334063,
      "grad_norm": 0.9263506531715393,
      "learning_rate": 2.874988078556221e-05,
      "loss": 0.7203,
      "step": 400
    },
    {
      "epoch": 0.8871851040525739,
      "grad_norm": 0.9023187756538391,
      "learning_rate": 2.8706124510495263e-05,
      "loss": 0.6927,
      "step": 405
    },
    {
      "epoch": 0.8981380065717415,
      "grad_norm": 0.9095348119735718,
      "learning_rate": 2.8661650055831645e-05,
      "loss": 0.7178,
      "step": 410
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.9606519937515259,
      "learning_rate": 2.8616459751963583e-05,
      "loss": 0.7349,
      "step": 415
    },
    {
      "epoch": 0.9200438116100766,
      "grad_norm": 0.9363222122192383,
      "learning_rate": 2.8570555966792677e-05,
      "loss": 0.6886,
      "step": 420
    },
    {
      "epoch": 0.9309967141292442,
      "grad_norm": 0.8668004870414734,
      "learning_rate": 2.852394110560583e-05,
      "loss": 0.689,
      "step": 425
    },
    {
      "epoch": 0.9419496166484118,
      "grad_norm": 0.8470218181610107,
      "learning_rate": 2.8476617610949246e-05,
      "loss": 0.7109,
      "step": 430
    },
    {
      "epoch": 0.9529025191675794,
      "grad_norm": 1.0149718523025513,
      "learning_rate": 2.84285879625004e-05,
      "loss": 0.6739,
      "step": 435
    },
    {
      "epoch": 0.963855421686747,
      "grad_norm": 0.8750612139701843,
      "learning_rate": 2.8379854676938137e-05,
      "loss": 0.6981,
      "step": 440
    },
    {
      "epoch": 0.9748083242059146,
      "grad_norm": 0.9368709325790405,
      "learning_rate": 2.8330420307810784e-05,
      "loss": 0.6762,
      "step": 445
    },
    {
      "epoch": 0.9857612267250822,
      "grad_norm": 1.0392800569534302,
      "learning_rate": 2.828028744540236e-05,
      "loss": 0.6816,
      "step": 450
    },
    {
      "epoch": 0.9967141292442497,
      "grad_norm": 1.062735915184021,
      "learning_rate": 2.822945871659685e-05,
      "loss": 0.6951,
      "step": 455
    }
  ],
  "logging_steps": 5,
  "max_steps": 2285,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.804863634510971e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}