{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1022,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009784735812133072,
      "grad_norm": 4.04649150142379,
      "learning_rate": 9.70873786407767e-07,
      "loss": 1.7598,
      "step": 10
    },
    {
      "epoch": 0.019569471624266144,
      "grad_norm": 1.8898281767230303,
      "learning_rate": 1.941747572815534e-06,
      "loss": 1.6964,
      "step": 20
    },
    {
      "epoch": 0.029354207436399216,
      "grad_norm": 1.483970324647816,
      "learning_rate": 2.912621359223301e-06,
      "loss": 1.611,
      "step": 30
    },
    {
      "epoch": 0.03913894324853229,
      "grad_norm": 1.3044067583298566,
      "learning_rate": 3.883495145631068e-06,
      "loss": 1.4946,
      "step": 40
    },
    {
      "epoch": 0.04892367906066536,
      "grad_norm": 1.1307579978177575,
      "learning_rate": 4.854368932038836e-06,
      "loss": 1.4397,
      "step": 50
    },
    {
      "epoch": 0.05870841487279843,
      "grad_norm": 1.1230632491576633,
      "learning_rate": 5.825242718446602e-06,
      "loss": 1.3915,
      "step": 60
    },
    {
      "epoch": 0.0684931506849315,
      "grad_norm": 1.2171487481951635,
      "learning_rate": 6.79611650485437e-06,
      "loss": 1.3412,
      "step": 70
    },
    {
      "epoch": 0.07827788649706457,
      "grad_norm": 1.1602684245071337,
      "learning_rate": 7.766990291262136e-06,
      "loss": 1.3442,
      "step": 80
    },
    {
      "epoch": 0.08806262230919765,
      "grad_norm": 1.1297981589874249,
      "learning_rate": 8.737864077669904e-06,
      "loss": 1.307,
      "step": 90
    },
    {
      "epoch": 0.09784735812133072,
      "grad_norm": 1.1989099168675559,
      "learning_rate": 9.708737864077671e-06,
      "loss": 1.3041,
      "step": 100
    },
    {
      "epoch": 0.10763209393346379,
      "grad_norm": 1.0923068630227757,
      "learning_rate": 9.998568523945896e-06,
      "loss": 1.2911,
      "step": 110
    },
    {
      "epoch": 0.11741682974559686,
      "grad_norm": 1.2597514623600858,
      "learning_rate": 9.991559185756013e-06,
      "loss": 1.2871,
      "step": 120
    },
    {
      "epoch": 0.12720156555772993,
      "grad_norm": 1.137541898001787,
      "learning_rate": 9.978717241383576e-06,
      "loss": 1.261,
      "step": 130
    },
    {
      "epoch": 0.136986301369863,
      "grad_norm": 1.1658887896761427,
      "learning_rate": 9.960057696561538e-06,
      "loss": 1.2554,
      "step": 140
    },
    {
      "epoch": 0.14677103718199608,
      "grad_norm": 1.2229261479256421,
      "learning_rate": 9.935602354853163e-06,
      "loss": 1.2542,
      "step": 150
    },
    {
      "epoch": 0.15655577299412915,
      "grad_norm": 1.1243006143779628,
      "learning_rate": 9.905379792174697e-06,
      "loss": 1.2521,
      "step": 160
    },
    {
      "epoch": 0.16634050880626222,
      "grad_norm": 1.0968643854275186,
      "learning_rate": 9.869425323404593e-06,
      "loss": 1.2677,
      "step": 170
    },
    {
      "epoch": 0.1761252446183953,
      "grad_norm": 1.1045669372583213,
      "learning_rate": 9.82778096111829e-06,
      "loss": 1.2511,
      "step": 180
    },
    {
      "epoch": 0.18590998043052837,
      "grad_norm": 1.1112118778947822,
      "learning_rate": 9.780495366496772e-06,
      "loss": 1.2446,
      "step": 190
    },
    {
      "epoch": 0.19569471624266144,
      "grad_norm": 1.0826613870591981,
      "learning_rate": 9.727623792466285e-06,
      "loss": 1.2314,
      "step": 200
    },
    {
      "epoch": 0.2054794520547945,
      "grad_norm": 1.098249884945961,
      "learning_rate": 9.669228019135625e-06,
      "loss": 1.2458,
      "step": 210
    },
    {
      "epoch": 0.21526418786692758,
      "grad_norm": 1.1612733633383774,
      "learning_rate": 9.605376281606472e-06,
      "loss": 1.2237,
      "step": 220
    },
    {
      "epoch": 0.22504892367906065,
      "grad_norm": 1.125936251992049,
      "learning_rate": 9.53614319024108e-06,
      "loss": 1.2146,
      "step": 230
    },
    {
      "epoch": 0.23483365949119372,
      "grad_norm": 1.1467535473229693,
      "learning_rate": 9.461609643480541e-06,
      "loss": 1.2138,
      "step": 240
    },
    {
      "epoch": 0.2446183953033268,
      "grad_norm": 1.0542912347811166,
      "learning_rate": 9.381862733315444e-06,
      "loss": 1.2111,
      "step": 250
    },
    {
      "epoch": 0.25440313111545987,
      "grad_norm": 1.149152919711082,
      "learning_rate": 9.296995643519432e-06,
      "loss": 1.2127,
      "step": 260
    },
    {
      "epoch": 0.26418786692759294,
      "grad_norm": 1.0920081819134964,
      "learning_rate": 9.207107540764519e-06,
      "loss": 1.2338,
      "step": 270
    },
    {
      "epoch": 0.273972602739726,
      "grad_norm": 1.1780253225746746,
      "learning_rate": 9.112303458745462e-06,
      "loss": 1.2054,
      "step": 280
    },
    {
      "epoch": 0.2837573385518591,
      "grad_norm": 1.0205557268076189,
      "learning_rate": 9.012694175448526e-06,
      "loss": 1.2223,
      "step": 290
    },
    {
      "epoch": 0.29354207436399216,
      "grad_norm": 1.182238809951801,
      "learning_rate": 8.90839608370809e-06,
      "loss": 1.2137,
      "step": 300
    },
    {
      "epoch": 0.30332681017612523,
      "grad_norm": 1.0623824819302699,
      "learning_rate": 8.799531055202332e-06,
      "loss": 1.2021,
      "step": 310
    },
    {
      "epoch": 0.3131115459882583,
      "grad_norm": 1.103614217675967,
      "learning_rate": 8.686226298046938e-06,
      "loss": 1.2107,
      "step": 320
    },
    {
      "epoch": 0.32289628180039137,
      "grad_norm": 1.0562421868073395,
      "learning_rate": 8.568614208153181e-06,
      "loss": 1.2051,
      "step": 330
    },
    {
      "epoch": 0.33268101761252444,
      "grad_norm": 1.061840954937612,
      "learning_rate": 8.446832214524132e-06,
      "loss": 1.1906,
      "step": 340
    },
    {
      "epoch": 0.3424657534246575,
      "grad_norm": 1.1618629276892396,
      "learning_rate": 8.321022618669706e-06,
      "loss": 1.2189,
      "step": 350
    },
    {
      "epoch": 0.3522504892367906,
      "grad_norm": 0.9952732308566709,
      "learning_rate": 8.191332428328227e-06,
      "loss": 1.2056,
      "step": 360
    },
    {
      "epoch": 0.36203522504892366,
      "grad_norm": 1.08417836514905,
      "learning_rate": 8.057913185688789e-06,
      "loss": 1.2058,
      "step": 370
    },
    {
      "epoch": 0.37181996086105673,
      "grad_norm": 1.0664180337103972,
      "learning_rate": 7.920920790315139e-06,
      "loss": 1.1787,
      "step": 380
    },
    {
      "epoch": 0.3816046966731898,
      "grad_norm": 1.0237212990645388,
      "learning_rate": 7.780515316978003e-06,
      "loss": 1.1868,
      "step": 390
    },
    {
      "epoch": 0.3913894324853229,
      "grad_norm": 1.0196131058576328,
      "learning_rate": 7.6368608286087e-06,
      "loss": 1.1945,
      "step": 400
    },
    {
      "epoch": 0.40117416829745595,
      "grad_norm": 1.095727756369159,
      "learning_rate": 7.490125184592607e-06,
      "loss": 1.1813,
      "step": 410
    },
    {
      "epoch": 0.410958904109589,
      "grad_norm": 1.0260192699123412,
      "learning_rate": 7.340479844626516e-06,
      "loss": 1.1996,
      "step": 420
    },
    {
      "epoch": 0.4207436399217221,
      "grad_norm": 1.0159524740262653,
      "learning_rate": 7.188099668369015e-06,
      "loss": 1.193,
      "step": 430
    },
    {
      "epoch": 0.43052837573385516,
      "grad_norm": 1.1094365729981088,
      "learning_rate": 7.033162711118077e-06,
      "loss": 1.182,
      "step": 440
    },
    {
      "epoch": 0.44031311154598823,
      "grad_norm": 1.0230643234668957,
      "learning_rate": 6.8758500157545185e-06,
      "loss": 1.1781,
      "step": 450
    },
    {
      "epoch": 0.4500978473581213,
      "grad_norm": 1.0282046095119801,
      "learning_rate": 6.716345401194519e-06,
      "loss": 1.1613,
      "step": 460
    },
    {
      "epoch": 0.4598825831702544,
      "grad_norm": 1.0383674282070607,
      "learning_rate": 6.554835247598347e-06,
      "loss": 1.1798,
      "step": 470
    },
    {
      "epoch": 0.46966731898238745,
      "grad_norm": 0.9837704553579479,
      "learning_rate": 6.391508278586287e-06,
      "loss": 1.181,
      "step": 480
    },
    {
      "epoch": 0.4794520547945205,
      "grad_norm": 1.0580246864370755,
      "learning_rate": 6.226555340716258e-06,
      "loss": 1.1887,
      "step": 490
    },
    {
      "epoch": 0.4892367906066536,
      "grad_norm": 1.017013165650618,
      "learning_rate": 6.060169180480776e-06,
      "loss": 1.1714,
      "step": 500
    },
    {
      "epoch": 0.49902152641878667,
      "grad_norm": 1.0072748086969616,
      "learning_rate": 5.8925442190838735e-06,
      "loss": 1.1818,
      "step": 510
    },
    {
      "epoch": 0.5088062622309197,
      "grad_norm": 1.0791337064960191,
      "learning_rate": 5.723876325261109e-06,
      "loss": 1.1766,
      "step": 520
    },
    {
      "epoch": 0.5185909980430529,
      "grad_norm": 1.0401219276204419,
      "learning_rate": 5.554362586408164e-06,
      "loss": 1.188,
      "step": 530
    },
    {
      "epoch": 0.5283757338551859,
      "grad_norm": 1.04026227937617,
      "learning_rate": 5.3842010782854306e-06,
      "loss": 1.1793,
      "step": 540
    },
    {
      "epoch": 0.538160469667319,
      "grad_norm": 1.17417294090351,
      "learning_rate": 5.2135906335677075e-06,
      "loss": 1.1734,
      "step": 550
    },
    {
      "epoch": 0.547945205479452,
      "grad_norm": 1.014702815772571,
      "learning_rate": 5.0427306095094365e-06,
      "loss": 1.169,
      "step": 560
    },
    {
      "epoch": 0.5577299412915852,
      "grad_norm": 1.0275982848917589,
      "learning_rate": 4.871820654996985e-06,
      "loss": 1.1856,
      "step": 570
    },
    {
      "epoch": 0.5675146771037182,
      "grad_norm": 1.0530785296526888,
      "learning_rate": 4.701060477260152e-06,
      "loss": 1.1662,
      "step": 580
    },
    {
      "epoch": 0.5772994129158513,
      "grad_norm": 1.023986668412493,
      "learning_rate": 4.5306496085154905e-06,
      "loss": 1.1786,
      "step": 590
    },
    {
      "epoch": 0.5870841487279843,
      "grad_norm": 1.0639406354816834,
      "learning_rate": 4.360787172814148e-06,
      "loss": 1.1516,
      "step": 600
    },
    {
      "epoch": 0.5968688845401174,
      "grad_norm": 0.9500520632946152,
      "learning_rate": 4.191671653366638e-06,
      "loss": 1.175,
      "step": 610
    },
    {
      "epoch": 0.6066536203522505,
      "grad_norm": 0.984391446028831,
      "learning_rate": 4.02350066061643e-06,
      "loss": 1.1656,
      "step": 620
    },
    {
      "epoch": 0.6164383561643836,
      "grad_norm": 1.005147372886877,
      "learning_rate": 3.85647070133336e-06,
      "loss": 1.1655,
      "step": 630
    },
    {
      "epoch": 0.6262230919765166,
      "grad_norm": 1.015642402945713,
      "learning_rate": 3.69077694899668e-06,
      "loss": 1.1572,
      "step": 640
    },
    {
      "epoch": 0.6360078277886497,
      "grad_norm": 1.0201007215429072,
      "learning_rate": 3.5266130157360557e-06,
      "loss": 1.1707,
      "step": 650
    },
    {
      "epoch": 0.6457925636007827,
      "grad_norm": 0.9956935095703923,
      "learning_rate": 3.3641707260969724e-06,
      "loss": 1.1619,
      "step": 660
    },
    {
      "epoch": 0.6555772994129159,
      "grad_norm": 0.9644023257276766,
      "learning_rate": 3.2036398928949475e-06,
      "loss": 1.1681,
      "step": 670
    },
    {
      "epoch": 0.6653620352250489,
      "grad_norm": 1.029502516764066,
      "learning_rate": 3.0452080954204133e-06,
      "loss": 1.1578,
      "step": 680
    },
    {
      "epoch": 0.675146771037182,
      "grad_norm": 0.9452739847782414,
      "learning_rate": 2.8890604602534777e-06,
      "loss": 1.166,
      "step": 690
    },
    {
      "epoch": 0.684931506849315,
      "grad_norm": 0.9549493201312858,
      "learning_rate": 2.735379444944655e-06,
      "loss": 1.156,
      "step": 700
    },
    {
      "epoch": 0.6947162426614482,
      "grad_norm": 0.9808822395078565,
      "learning_rate": 2.5843446248143308e-06,
      "loss": 1.1498,
      "step": 710
    },
    {
      "epoch": 0.7045009784735812,
      "grad_norm": 0.9725284687123298,
      "learning_rate": 2.436132483120121e-06,
      "loss": 1.1414,
      "step": 720
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 1.0146025684804074,
      "learning_rate": 2.290916204837262e-06,
      "loss": 1.1494,
      "step": 730
    },
    {
      "epoch": 0.7240704500978473,
      "grad_norm": 0.9942038199138818,
      "learning_rate": 2.148865474293036e-06,
      "loss": 1.1532,
      "step": 740
    },
    {
      "epoch": 0.7338551859099804,
      "grad_norm": 0.9634937304590753,
      "learning_rate": 2.010146276891675e-06,
      "loss": 1.1524,
      "step": 750
    },
    {
      "epoch": 0.7436399217221135,
      "grad_norm": 0.9646041486253625,
      "learning_rate": 1.8749207051614448e-06,
      "loss": 1.1553,
      "step": 760
    },
    {
      "epoch": 0.7534246575342466,
      "grad_norm": 0.9645946065696461,
      "learning_rate": 1.743346769350519e-06,
      "loss": 1.1493,
      "step": 770
    },
    {
      "epoch": 0.7632093933463796,
      "grad_norm": 0.9987538300542498,
      "learning_rate": 1.6155782127929787e-06,
      "loss": 1.1544,
      "step": 780
    },
    {
      "epoch": 0.7729941291585127,
      "grad_norm": 0.9706627230186383,
      "learning_rate": 1.4917643322606668e-06,
      "loss": 1.169,
      "step": 790
    },
    {
      "epoch": 0.7827788649706457,
      "grad_norm": 0.995763457209775,
      "learning_rate": 1.3720498035108393e-06,
      "loss": 1.1406,
      "step": 800
    },
    {
      "epoch": 0.7925636007827789,
      "grad_norm": 0.9780490169698375,
      "learning_rate": 1.256574512233426e-06,
      "loss": 1.1464,
      "step": 810
    },
    {
      "epoch": 0.8023483365949119,
      "grad_norm": 1.058131773284045,
      "learning_rate": 1.1454733905954679e-06,
      "loss": 1.1682,
      "step": 820
    },
    {
      "epoch": 0.812133072407045,
      "grad_norm": 0.9663337521771117,
      "learning_rate": 1.0388762595737012e-06,
      "loss": 1.1514,
      "step": 830
    },
    {
      "epoch": 0.821917808219178,
      "grad_norm": 1.0060251032305096,
      "learning_rate": 9.36907677259562e-07,
      "loss": 1.1381,
      "step": 840
    },
    {
      "epoch": 0.8317025440313112,
      "grad_norm": 1.0006702241602894,
      "learning_rate": 8.396867933138031e-07,
      "loss": 1.1544,
      "step": 850
    },
    {
      "epoch": 0.8414872798434442,
      "grad_norm": 0.9655200991402652,
      "learning_rate": 7.473272097408657e-07,
      "loss": 1.1572,
      "step": 860
    },
    {
      "epoch": 0.8512720156555773,
      "grad_norm": 1.0010147904810007,
      "learning_rate": 6.599368481456237e-07,
      "loss": 1.1645,
      "step": 870
    },
    {
      "epoch": 0.8610567514677103,
      "grad_norm": 0.9615849194888773,
      "learning_rate": 5.776178236276642e-07,
      "loss": 1.1402,
      "step": 880
    },
    {
      "epoch": 0.8708414872798435,
      "grad_norm": 0.9899952335825283,
      "learning_rate": 5.004663254604125e-07,
      "loss": 1.1434,
      "step": 890
    },
    {
      "epoch": 0.8806262230919765,
      "grad_norm": 0.9522480445427419,
      "learning_rate": 4.28572504694566e-07,
      "loss": 1.1404,
      "step": 900
    },
    {
      "epoch": 0.8904109589041096,
      "grad_norm": 0.974551459961046,
      "learning_rate": 3.620203688171381e-07,
      "loss": 1.1577,
      "step": 910
    },
    {
      "epoch": 0.9001956947162426,
      "grad_norm": 0.9668235111962397,
      "learning_rate": 3.0088768358923545e-07,
      "loss": 1.1427,
      "step": 920
    },
    {
      "epoch": 0.9099804305283757,
      "grad_norm": 1.0319353813022967,
      "learning_rate": 2.4524588217723746e-07,
      "loss": 1.1373,
      "step": 930
    },
    {
      "epoch": 0.9197651663405088,
      "grad_norm": 1.029861953702673,
      "learning_rate": 1.9515998168358809e-07,
      "loss": 1.1565,
      "step": 940
    },
    {
      "epoch": 0.9295499021526419,
      "grad_norm": 0.950517130129004,
      "learning_rate": 1.506885071747033e-07,
      "loss": 1.1409,
      "step": 950
    },
    {
      "epoch": 0.9393346379647749,
      "grad_norm": 0.9918632320485316,
      "learning_rate": 1.1188342329479962e-07,
      "loss": 1.1592,
      "step": 960
    },
    {
      "epoch": 0.949119373776908,
      "grad_norm": 0.9513450607234911,
      "learning_rate": 7.879007354552271e-08,
      "loss": 1.1331,
      "step": 970
    },
    {
      "epoch": 0.958904109589041,
      "grad_norm": 0.9860328317263265,
      "learning_rate": 5.14471273023448e-08,
      "loss": 1.1588,
      "step": 980
    },
    {
      "epoch": 0.9686888454011742,
      "grad_norm": 0.9557954337884952,
      "learning_rate": 2.9886534629638975e-08,
      "loss": 1.1482,
      "step": 990
    },
    {
      "epoch": 0.9784735812133072,
      "grad_norm": 0.982073690453174,
      "learning_rate": 1.4133488947225548e-08,
      "loss": 1.17,
      "step": 1000
    },
    {
      "epoch": 0.9882583170254403,
      "grad_norm": 0.9769182862134684,
      "learning_rate": 4.2063975920153545e-09,
      "loss": 1.155,
      "step": 1010
    },
    {
      "epoch": 0.9980430528375733,
      "grad_norm": 0.9371852604717477,
      "learning_rate": 1.168603091489251e-10,
      "loss": 1.1449,
      "step": 1020
    },
    {
      "epoch": 1.0,
      "step": 1022,
      "total_flos": 1.3956152400113828e+18,
      "train_loss": 1.210920991029758,
      "train_runtime": 4090.7445,
      "train_samples_per_second": 31.964,
      "train_steps_per_second": 0.25
    }
  ],
  "logging_steps": 10,
  "max_steps": 1022,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.3956152400113828e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}