{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 1070,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 2.701171636581421,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.7223,
      "step": 10
    },
    {
      "epoch": 0.18691588785046728,
      "grad_norm": 1.4309473037719727,
      "learning_rate": 3.518518518518519e-05,
      "loss": 0.313,
      "step": 20
    },
    {
      "epoch": 0.2803738317757009,
      "grad_norm": 1.99528968334198,
      "learning_rate": 5.370370370370371e-05,
      "loss": 0.2547,
      "step": 30
    },
    {
      "epoch": 0.37383177570093457,
      "grad_norm": 1.5343451499938965,
      "learning_rate": 7.222222222222222e-05,
      "loss": 0.2301,
      "step": 40
    },
    {
      "epoch": 0.4672897196261682,
      "grad_norm": 1.0875062942504883,
      "learning_rate": 9.074074074074075e-05,
      "loss": 0.1949,
      "step": 50
    },
    {
      "epoch": 0.5607476635514018,
      "grad_norm": 1.1911485195159912,
      "learning_rate": 9.999402437003975e-05,
      "loss": 0.1649,
      "step": 60
    },
    {
      "epoch": 0.6542056074766355,
      "grad_norm": 0.9195408821105957,
      "learning_rate": 9.99462278999732e-05,
      "loss": 0.1468,
      "step": 70
    },
    {
      "epoch": 0.7476635514018691,
      "grad_norm": 1.2133328914642334,
      "learning_rate": 9.985068065535225e-05,
      "loss": 0.1285,
      "step": 80
    },
    {
      "epoch": 0.8411214953271028,
      "grad_norm": 0.8248424530029297,
      "learning_rate": 9.970747398351445e-05,
      "loss": 0.109,
      "step": 90
    },
    {
      "epoch": 0.9345794392523364,
      "grad_norm": 0.8372947573661804,
      "learning_rate": 9.951674479629056e-05,
      "loss": 0.1036,
      "step": 100
    },
    {
      "epoch": 1.02803738317757,
      "grad_norm": 1.1082693338394165,
      "learning_rate": 9.927867543911091e-05,
      "loss": 0.0895,
      "step": 110
    },
    {
      "epoch": 1.1214953271028036,
      "grad_norm": 0.9761236310005188,
      "learning_rate": 9.899349351667522e-05,
      "loss": 0.0844,
      "step": 120
    },
    {
      "epoch": 1.2149532710280373,
      "grad_norm": 0.6534666419029236,
      "learning_rate": 9.866147167535254e-05,
      "loss": 0.0749,
      "step": 130
    },
    {
      "epoch": 1.308411214953271,
      "grad_norm": 0.946822464466095,
      "learning_rate": 9.828292734251944e-05,
      "loss": 0.0713,
      "step": 140
    },
    {
      "epoch": 1.4018691588785046,
      "grad_norm": 0.8314803838729858,
      "learning_rate": 9.785822242308562e-05,
      "loss": 0.0674,
      "step": 150
    },
    {
      "epoch": 1.4953271028037383,
      "grad_norm": 0.6225787997245789,
      "learning_rate": 9.738776295349687e-05,
      "loss": 0.0654,
      "step": 160
    },
    {
      "epoch": 1.588785046728972,
      "grad_norm": 0.6679856777191162,
      "learning_rate": 9.687199871354669e-05,
      "loss": 0.0666,
      "step": 170
    },
    {
      "epoch": 1.6822429906542056,
      "grad_norm": 0.4407249391078949,
      "learning_rate": 9.631142279636706e-05,
      "loss": 0.056,
      "step": 180
    },
    {
      "epoch": 1.7757009345794392,
      "grad_norm": 0.5206877589225769,
      "learning_rate": 9.570657113700985e-05,
      "loss": 0.0595,
      "step": 190
    },
    {
      "epoch": 1.8691588785046729,
      "grad_norm": 0.7172874212265015,
      "learning_rate": 9.50580220000696e-05,
      "loss": 0.0611,
      "step": 200
    },
    {
      "epoch": 1.9626168224299065,
      "grad_norm": 0.5936960577964783,
      "learning_rate": 9.436639542683727e-05,
      "loss": 0.0538,
      "step": 210
    },
    {
      "epoch": 2.05607476635514,
      "grad_norm": 0.4505106806755066,
      "learning_rate": 9.363235264251369e-05,
      "loss": 0.0586,
      "step": 220
    },
    {
      "epoch": 2.149532710280374,
      "grad_norm": 0.45083609223365784,
      "learning_rate": 9.285659542404941e-05,
      "loss": 0.0571,
      "step": 230
    },
    {
      "epoch": 2.2429906542056073,
      "grad_norm": 0.5857895612716675,
      "learning_rate": 9.203986542921532e-05,
      "loss": 0.0496,
      "step": 240
    },
    {
      "epoch": 2.336448598130841,
      "grad_norm": 0.600140392780304,
      "learning_rate": 9.11829434875454e-05,
      "loss": 0.056,
      "step": 250
    },
    {
      "epoch": 2.4299065420560746,
      "grad_norm": 0.46349039673805237,
      "learning_rate": 9.02866488538296e-05,
      "loss": 0.0482,
      "step": 260
    },
    {
      "epoch": 2.5233644859813085,
      "grad_norm": 0.597632884979248,
      "learning_rate": 8.93518384248705e-05,
      "loss": 0.0498,
      "step": 270
    },
    {
      "epoch": 2.616822429906542,
      "grad_norm": 0.6381852030754089,
      "learning_rate": 8.837940592025257e-05,
      "loss": 0.0521,
      "step": 280
    },
    {
      "epoch": 2.710280373831776,
      "grad_norm": 0.5228458642959595,
      "learning_rate": 8.737028102790723e-05,
      "loss": 0.0523,
      "step": 290
    },
    {
      "epoch": 2.803738317757009,
      "grad_norm": 0.6592262983322144,
      "learning_rate": 8.632542851529051e-05,
      "loss": 0.0481,
      "step": 300
    },
    {
      "epoch": 2.897196261682243,
      "grad_norm": 0.4157472848892212,
      "learning_rate": 8.524584730702339e-05,
      "loss": 0.0459,
      "step": 310
    },
    {
      "epoch": 2.9906542056074765,
      "grad_norm": 0.2829182744026184,
      "learning_rate": 8.413256952987611e-05,
      "loss": 0.0479,
      "step": 320
    },
    {
      "epoch": 3.0841121495327104,
      "grad_norm": 0.6042788624763489,
      "learning_rate": 8.298665952600999e-05,
      "loss": 0.0401,
      "step": 330
    },
    {
      "epoch": 3.177570093457944,
      "grad_norm": 0.31885308027267456,
      "learning_rate": 8.180921283541986e-05,
      "loss": 0.0441,
      "step": 340
    },
    {
      "epoch": 3.2710280373831777,
      "grad_norm": 0.3092094659805298,
      "learning_rate": 8.060135514854994e-05,
      "loss": 0.041,
      "step": 350
    },
    {
      "epoch": 3.364485981308411,
      "grad_norm": 0.39221957325935364,
      "learning_rate": 7.936424123008464e-05,
      "loss": 0.0378,
      "step": 360
    },
    {
      "epoch": 3.457943925233645,
      "grad_norm": 0.5164565443992615,
      "learning_rate": 7.809905381494316e-05,
      "loss": 0.0436,
      "step": 370
    },
    {
      "epoch": 3.5514018691588785,
      "grad_norm": 0.4761320650577545,
      "learning_rate": 7.68070024775332e-05,
      "loss": 0.0393,
      "step": 380
    },
    {
      "epoch": 3.6448598130841123,
      "grad_norm": 0.24083180725574493,
      "learning_rate": 7.548932247534506e-05,
      "loss": 0.0432,
      "step": 390
    },
    {
      "epoch": 3.7383177570093458,
      "grad_norm": 0.32576093077659607,
      "learning_rate": 7.414727356799154e-05,
      "loss": 0.04,
      "step": 400
    },
    {
      "epoch": 3.831775700934579,
      "grad_norm": 0.4475277066230774,
      "learning_rate": 7.27821388128227e-05,
      "loss": 0.0423,
      "step": 410
    },
    {
      "epoch": 3.925233644859813,
      "grad_norm": 0.47092771530151367,
      "learning_rate": 7.139522333826707e-05,
      "loss": 0.0408,
      "step": 420
    },
    {
      "epoch": 4.018691588785047,
      "grad_norm": 0.37364593148231506,
      "learning_rate": 6.99878530960719e-05,
      "loss": 0.0361,
      "step": 430
    },
    {
      "epoch": 4.11214953271028,
      "grad_norm": 0.3749428391456604,
      "learning_rate": 6.856137359363533e-05,
      "loss": 0.0347,
      "step": 440
    },
    {
      "epoch": 4.205607476635514,
      "grad_norm": 0.28443631529808044,
      "learning_rate": 6.711714860764266e-05,
      "loss": 0.0378,
      "step": 450
    },
    {
      "epoch": 4.299065420560748,
      "grad_norm": 0.31837236881256104,
      "learning_rate": 6.565655888023618e-05,
      "loss": 0.0363,
      "step": 460
    },
    {
      "epoch": 4.392523364485982,
      "grad_norm": 0.3948926329612732,
      "learning_rate": 6.418100079896556e-05,
      "loss": 0.0388,
      "step": 470
    },
    {
      "epoch": 4.485981308411215,
      "grad_norm": 0.5089420080184937,
      "learning_rate": 6.269188506178019e-05,
      "loss": 0.0377,
      "step": 480
    },
    {
      "epoch": 4.579439252336448,
      "grad_norm": 0.266495943069458,
      "learning_rate": 6.11906353283405e-05,
      "loss": 0.0367,
      "step": 490
    },
    {
      "epoch": 4.672897196261682,
      "grad_norm": 0.33666715025901794,
      "learning_rate": 5.967868685893715e-05,
      "loss": 0.0372,
      "step": 500
    },
    {
      "epoch": 4.766355140186916,
      "grad_norm": 0.37420037388801575,
      "learning_rate": 5.815748514231944e-05,
      "loss": 0.032,
      "step": 510
    },
    {
      "epoch": 4.859813084112149,
      "grad_norm": 0.33954936265945435,
      "learning_rate": 5.6628484513745e-05,
      "loss": 0.0299,
      "step": 520
    },
    {
      "epoch": 4.953271028037383,
      "grad_norm": 0.2828126847743988,
      "learning_rate": 5.5093146764571866e-05,
      "loss": 0.0343,
      "step": 530
    },
    {
      "epoch": 5.046728971962617,
      "grad_norm": 0.48174425959587097,
      "learning_rate": 5.355293974472197e-05,
      "loss": 0.0326,
      "step": 540
    },
    {
      "epoch": 5.140186915887851,
      "grad_norm": 0.42725300788879395,
      "learning_rate": 5.2009335959352666e-05,
      "loss": 0.0338,
      "step": 550
    },
    {
      "epoch": 5.233644859813084,
      "grad_norm": 0.34893345832824707,
      "learning_rate": 5.046381116107742e-05,
      "loss": 0.0322,
      "step": 560
    },
    {
      "epoch": 5.327102803738318,
      "grad_norm": 0.3411373198032379,
      "learning_rate": 4.891784293908192e-05,
      "loss": 0.0312,
      "step": 570
    },
    {
      "epoch": 5.420560747663552,
      "grad_norm": 0.35624727606773376,
      "learning_rate": 4.7372909306484276e-05,
      "loss": 0.0353,
      "step": 580
    },
    {
      "epoch": 5.5140186915887845,
      "grad_norm": 0.3242332935333252,
      "learning_rate": 4.5830487287289966e-05,
      "loss": 0.0288,
      "step": 590
    },
    {
      "epoch": 5.607476635514018,
      "grad_norm": 0.4070276618003845,
      "learning_rate": 4.429205150429241e-05,
      "loss": 0.0322,
      "step": 600
    },
    {
      "epoch": 5.700934579439252,
      "grad_norm": 0.2987990975379944,
      "learning_rate": 4.275907276926918e-05,
      "loss": 0.0293,
      "step": 610
    },
    {
      "epoch": 5.794392523364486,
      "grad_norm": 0.36411428451538086,
      "learning_rate": 4.123301667682171e-05,
      "loss": 0.0296,
      "step": 620
    },
    {
      "epoch": 5.88785046728972,
      "grad_norm": 0.30948546528816223,
      "learning_rate": 3.971534220320291e-05,
      "loss": 0.0274,
      "step": 630
    },
    {
      "epoch": 5.981308411214953,
      "grad_norm": 0.4233947694301605,
      "learning_rate": 3.820750031147211e-05,
      "loss": 0.0316,
      "step": 640
    },
    {
      "epoch": 6.074766355140187,
      "grad_norm": 0.47963348031044006,
      "learning_rate": 3.67109325643111e-05,
      "loss": 0.029,
      "step": 650
    },
    {
      "epoch": 6.168224299065421,
      "grad_norm": 0.33454999327659607,
      "learning_rate": 3.522706974582717e-05,
      "loss": 0.0301,
      "step": 660
    },
    {
      "epoch": 6.261682242990654,
      "grad_norm": 0.22388812899589539,
      "learning_rate": 3.375733049366115e-05,
      "loss": 0.0273,
      "step": 670
    },
    {
      "epoch": 6.355140186915888,
      "grad_norm": 0.3955516219139099,
      "learning_rate": 3.2303119942707796e-05,
      "loss": 0.0259,
      "step": 680
    },
    {
      "epoch": 6.4485981308411215,
      "grad_norm": 0.33848804235458374,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.029,
      "step": 690
    },
    {
      "epoch": 6.542056074766355,
      "grad_norm": 0.6161491274833679,
      "learning_rate": 2.944682992425959e-05,
      "loss": 0.0273,
      "step": 700
    },
    {
      "epoch": 6.635514018691588,
      "grad_norm": 0.3765656352043152,
      "learning_rate": 2.804748119472969e-05,
      "loss": 0.027,
      "step": 710
    },
    {
      "epoch": 6.728971962616822,
      "grad_norm": 0.22724078595638275,
      "learning_rate": 2.6669120031637663e-05,
      "loss": 0.0259,
      "step": 720
    },
    {
      "epoch": 6.822429906542056,
      "grad_norm": 0.39581674337387085,
      "learning_rate": 2.5313064208435423e-05,
      "loss": 0.0282,
      "step": 730
    },
    {
      "epoch": 6.91588785046729,
      "grad_norm": 0.4597732126712799,
      "learning_rate": 2.3980610173696255e-05,
      "loss": 0.0269,
      "step": 740
    },
    {
      "epoch": 7.009345794392523,
      "grad_norm": 0.39131224155426025,
      "learning_rate": 2.2673031811653034e-05,
      "loss": 0.0309,
      "step": 750
    },
    {
      "epoch": 7.102803738317757,
      "grad_norm": 0.33914846181869507,
      "learning_rate": 2.139157922430956e-05,
      "loss": 0.0249,
      "step": 760
    },
    {
      "epoch": 7.196261682242991,
      "grad_norm": 0.40823277831077576,
      "learning_rate": 2.01374775362883e-05,
      "loss": 0.026,
      "step": 770
    },
    {
      "epoch": 7.289719626168225,
      "grad_norm": 0.20548699796199799,
      "learning_rate": 1.8911925723557806e-05,
      "loss": 0.0214,
      "step": 780
    },
    {
      "epoch": 7.383177570093458,
      "grad_norm": 0.28237590193748474,
      "learning_rate": 1.7716095467159393e-05,
      "loss": 0.0262,
      "step": 790
    },
    {
      "epoch": 7.4766355140186915,
      "grad_norm": 0.29693859815597534,
      "learning_rate": 1.6551130033028827e-05,
      "loss": 0.0233,
      "step": 800
    },
    {
      "epoch": 7.570093457943925,
      "grad_norm": 0.3963673710823059,
      "learning_rate": 1.541814317898425e-05,
      "loss": 0.0247,
      "step": 810
    },
    {
      "epoch": 7.663551401869158,
      "grad_norm": 0.22003985941410065,
      "learning_rate": 1.4318218089924962e-05,
      "loss": 0.0249,
      "step": 820
    },
    {
      "epoch": 7.757009345794392,
      "grad_norm": 0.2355249971151352,
      "learning_rate": 1.3252406342259527e-05,
      "loss": 0.0245,
      "step": 830
    },
    {
      "epoch": 7.850467289719626,
      "grad_norm": 0.45212939381599426,
      "learning_rate": 1.2221726898552665e-05,
      "loss": 0.0228,
      "step": 840
    },
    {
      "epoch": 7.94392523364486,
      "grad_norm": 0.4631796181201935,
      "learning_rate": 1.122716513335262e-05,
      "loss": 0.0215,
      "step": 850
    },
    {
      "epoch": 8.037383177570094,
      "grad_norm": 0.24740639328956604,
      "learning_rate": 1.0269671891130123e-05,
      "loss": 0.0267,
      "step": 860
    },
    {
      "epoch": 8.130841121495328,
      "grad_norm": 0.21961306035518646,
      "learning_rate": 9.350162577229432e-06,
      "loss": 0.0222,
      "step": 870
    },
    {
      "epoch": 8.22429906542056,
      "grad_norm": 0.2656007409095764,
      "learning_rate": 8.46951628270098e-06,
      "loss": 0.0234,
      "step": 880
    },
    {
      "epoch": 8.317757009345794,
      "grad_norm": 0.1633329689502716,
      "learning_rate": 7.628574943851852e-06,
      "loss": 0.0189,
      "step": 890
    },
    {
      "epoch": 8.411214953271028,
      "grad_norm": 0.15491314232349396,
      "learning_rate": 6.82814253731801e-06,
      "loss": 0.0199,
      "step": 900
    },
    {
      "epoch": 8.504672897196262,
      "grad_norm": 0.464375376701355,
      "learning_rate": 6.06898431142745e-06,
      "loss": 0.0216,
      "step": 910
    },
    {
      "epoch": 8.598130841121495,
      "grad_norm": 0.21681295335292816,
      "learning_rate": 5.351826054589393e-06,
      "loss": 0.0203,
      "step": 920
    },
    {
      "epoch": 8.69158878504673,
      "grad_norm": 0.18899723887443542,
      "learning_rate": 4.677353401408974e-06,
      "loss": 0.0184,
      "step": 930
    },
    {
      "epoch": 8.785046728971963,
      "grad_norm": 0.28756183385849,
      "learning_rate": 4.04621117719049e-06,
      "loss": 0.0257,
      "step": 940
    },
    {
      "epoch": 8.878504672897197,
      "grad_norm": 0.27610504627227783,
      "learning_rate": 3.459002781456344e-06,
      "loss": 0.0208,
      "step": 950
    },
    {
      "epoch": 8.97196261682243,
      "grad_norm": 0.21105413138866425,
      "learning_rate": 2.9162896110707163e-06,
      "loss": 0.0219,
      "step": 960
    },
    {
      "epoch": 9.065420560747663,
      "grad_norm": 0.23299843072891235,
      "learning_rate": 2.418590523519687e-06,
      "loss": 0.0217,
      "step": 970
    },
    {
      "epoch": 9.158878504672897,
      "grad_norm": 0.3126954436302185,
      "learning_rate": 1.9663813408607845e-06,
      "loss": 0.0188,
      "step": 980
    },
    {
      "epoch": 9.25233644859813,
      "grad_norm": 0.32796964049339294,
      "learning_rate": 1.5600943948163527e-06,
      "loss": 0.0227,
      "step": 990
    },
    {
      "epoch": 9.345794392523365,
      "grad_norm": 0.2622697949409485,
      "learning_rate": 1.2001181134455475e-06,
      "loss": 0.0205,
      "step": 1000
    },
    {
      "epoch": 9.439252336448599,
      "grad_norm": 0.23692955076694489,
      "learning_rate": 8.867966497901282e-07,
      "loss": 0.0216,
      "step": 1010
    },
    {
      "epoch": 9.532710280373832,
      "grad_norm": 0.23177599906921387,
      "learning_rate": 6.204295528491555e-07,
      "loss": 0.0197,
      "step": 1020
    },
    {
      "epoch": 9.626168224299064,
      "grad_norm": 0.18032234907150269,
      "learning_rate": 4.012714811970464e-07,
      "loss": 0.0223,
      "step": 1030
    },
    {
      "epoch": 9.719626168224298,
      "grad_norm": 0.4851064682006836,
      "learning_rate": 2.295319595188805e-07,
      "loss": 0.0209,
      "step": 1040
    },
    {
      "epoch": 9.813084112149532,
      "grad_norm": 0.2691369950771332,
      "learning_rate": 1.0537517829562472e-07,
      "loss": 0.0195,
      "step": 1050
    },
    {
      "epoch": 9.906542056074766,
      "grad_norm": 0.19837787747383118,
      "learning_rate": 2.8919836830887392e-08,
      "loss": 0.0186,
      "step": 1060
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.2632281482219696,
      "learning_rate": 2.3902976920009423e-10,
      "loss": 0.0186,
      "step": 1070
    },
    {
      "epoch": 10.0,
      "step": 1070,
      "total_flos": 0.0,
      "train_loss": 0.0546069160939377,
      "train_runtime": 1371.0334,
      "train_samples_per_second": 77.708,
      "train_steps_per_second": 0.78
    }
  ],
  "logging_steps": 10,
  "max_steps": 1070,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 100,
  "trial_name": null,
  "trial_params": null
}