{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 704,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007106057914372002,
      "grad_norm": 48.15382880127834,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 4.3877,
      "mean_token_accuracy": 0.4402003187686205,
      "num_tokens": 5475610.0,
      "step": 5
    },
    {
      "epoch": 0.014212115828744005,
      "grad_norm": 14.120135940330632,
      "learning_rate": 8.181818181818183e-06,
      "loss": 3.149,
      "mean_token_accuracy": 0.49750813096761703,
      "num_tokens": 10990469.0,
      "step": 10
    },
    {
      "epoch": 0.021318173743116006,
      "grad_norm": 4.9582246488048884,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 1.6042,
      "mean_token_accuracy": 0.6544640183448791,
      "num_tokens": 16509954.0,
      "step": 15
    },
    {
      "epoch": 0.02842423165748801,
      "grad_norm": 2.3061348987678967,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 1.0542,
      "mean_token_accuracy": 0.7447574809193611,
      "num_tokens": 22025438.0,
      "step": 20
    },
    {
      "epoch": 0.03553028957186001,
      "grad_norm": 3.4964571027817337,
      "learning_rate": 1.999961805535155e-05,
      "loss": 0.8969,
      "mean_token_accuracy": 0.7685870260000229,
      "num_tokens": 27537030.0,
      "step": 25
    },
    {
      "epoch": 0.04263634748623201,
      "grad_norm": 3.3071652925849335,
      "learning_rate": 1.9995321550350065e-05,
      "loss": 0.8074,
      "mean_token_accuracy": 0.7807081520557404,
      "num_tokens": 33069854.0,
      "step": 30
    },
    {
      "epoch": 0.04974240540060401,
      "grad_norm": 3.924372429122354,
      "learning_rate": 1.998625339625423e-05,
      "loss": 0.7719,
      "mean_token_accuracy": 0.783011856675148,
      "num_tokens": 38592423.0,
      "step": 35
    },
    {
      "epoch": 0.05684846331497602,
      "grad_norm": 0.9496737705645568,
      "learning_rate": 1.9972418403347817e-05,
      "loss": 0.7245,
      "mean_token_accuracy": 0.7904581762850285,
      "num_tokens": 44102692.0,
      "step": 40
    },
    {
      "epoch": 0.06395452122934801,
      "grad_norm": 0.4437647446141135,
      "learning_rate": 1.9953823910527057e-05,
      "loss": 0.6875,
      "mean_token_accuracy": 0.8015773832798004,
      "num_tokens": 49606158.0,
      "step": 45
    },
    {
      "epoch": 0.07106057914372002,
      "grad_norm": 0.36104953638366394,
      "learning_rate": 1.993047978140764e-05,
      "loss": 0.6681,
      "mean_token_accuracy": 0.8048019059002399,
      "num_tokens": 55124799.0,
      "step": 50
    },
    {
      "epoch": 0.07816663705809203,
      "grad_norm": 0.3451124617137935,
      "learning_rate": 1.9902398399092494e-05,
      "loss": 0.6356,
      "mean_token_accuracy": 0.8128186449408531,
      "num_tokens": 60636117.0,
      "step": 55
    },
    {
      "epoch": 0.08527269497246402,
      "grad_norm": 0.389660261004574,
      "learning_rate": 1.9869594659603032e-05,
      "loss": 0.6398,
      "mean_token_accuracy": 0.8115979641675949,
      "num_tokens": 66154262.0,
      "step": 60
    },
    {
      "epoch": 0.09237875288683603,
      "grad_norm": 0.31317999931398266,
      "learning_rate": 1.9832085963977445e-05,
      "loss": 0.6337,
      "mean_token_accuracy": 0.8128221824765205,
      "num_tokens": 71679661.0,
      "step": 65
    },
    {
      "epoch": 0.09948481080120802,
      "grad_norm": 0.3128254105868865,
      "learning_rate": 1.978989220904016e-05,
      "loss": 0.6227,
      "mean_token_accuracy": 0.8144343480467796,
      "num_tokens": 77204132.0,
      "step": 70
    },
    {
      "epoch": 0.10659086871558003,
      "grad_norm": 0.3000404040639256,
      "learning_rate": 1.9743035776847377e-05,
      "loss": 0.6166,
      "mean_token_accuracy": 0.8157493658363819,
      "num_tokens": 82747984.0,
      "step": 75
    },
    {
      "epoch": 0.11369692662995204,
      "grad_norm": 0.3139627449464346,
      "learning_rate": 1.9691541522814327e-05,
      "loss": 0.5988,
      "mean_token_accuracy": 0.8197014890611172,
      "num_tokens": 88267430.0,
      "step": 80
    },
    {
      "epoch": 0.12080298454432403,
      "grad_norm": 0.3039876079367156,
      "learning_rate": 1.963543676253048e-05,
      "loss": 0.6051,
      "mean_token_accuracy": 0.8184983253479003,
      "num_tokens": 93790180.0,
      "step": 85
    },
    {
      "epoch": 0.12790904245869603,
      "grad_norm": 0.3409273981011121,
      "learning_rate": 1.9574751257269748e-05,
      "loss": 0.5978,
      "mean_token_accuracy": 0.8195111580193043,
      "num_tokens": 99314107.0,
      "step": 90
    },
    {
      "epoch": 0.13501510037306805,
      "grad_norm": 0.2958717785828003,
      "learning_rate": 1.950951719820335e-05,
      "loss": 0.5902,
      "mean_token_accuracy": 0.8219340682029724,
      "num_tokens": 104809146.0,
      "step": 95
    },
    {
      "epoch": 0.14212115828744004,
      "grad_norm": 0.2994112787306848,
      "learning_rate": 1.9439769189323727e-05,
      "loss": 0.5989,
      "mean_token_accuracy": 0.820228873193264,
      "num_tokens": 110323602.0,
      "step": 100
    },
    {
      "epoch": 0.14212115828744004,
      "eval_loss": 0.5710137486457825,
      "eval_mean_token_accuracy": 0.8228120437839574,
      "eval_num_tokens": 110323602.0,
      "eval_runtime": 160.5102,
      "eval_samples_per_second": 22.671,
      "eval_steps_per_second": 0.71,
      "step": 100
    },
    {
      "epoch": 0.14922721620181204,
      "grad_norm": 0.30064641734602476,
      "learning_rate": 1.9365544229088517e-05,
      "loss": 0.5944,
      "mean_token_accuracy": 0.8212312825024128,
      "num_tokens": 115858100.0,
      "step": 105
    },
    {
      "epoch": 0.15633327411618406,
      "grad_norm": 0.2894891012090971,
      "learning_rate": 1.9286881690794425e-05,
      "loss": 0.5832,
      "mean_token_accuracy": 0.8244290247559547,
      "num_tokens": 121393955.0,
      "step": 110
    },
    {
      "epoch": 0.16343933203055605,
      "grad_norm": 0.3177791926774768,
      "learning_rate": 1.9203823301691272e-05,
      "loss": 0.5946,
      "mean_token_accuracy": 0.8195516988635063,
      "num_tokens": 126941522.0,
      "step": 115
    },
    {
      "epoch": 0.17054538994492804,
      "grad_norm": 0.3271845261422823,
      "learning_rate": 1.9116413120847425e-05,
      "loss": 0.5853,
      "mean_token_accuracy": 0.823321682959795,
      "num_tokens": 132467661.0,
      "step": 120
    },
    {
      "epoch": 0.17765144785930007,
      "grad_norm": 0.30578663346709106,
      "learning_rate": 1.902469751577826e-05,
      "loss": 0.5787,
      "mean_token_accuracy": 0.8247792065143585,
      "num_tokens": 137980426.0,
      "step": 125
    },
    {
      "epoch": 0.18475750577367206,
      "grad_norm": 0.28728878535567604,
      "learning_rate": 1.892872513785008e-05,
      "loss": 0.568,
      "mean_token_accuracy": 0.8274045430123806,
      "num_tokens": 143490685.0,
      "step": 130
    },
    {
      "epoch": 0.19186356368804405,
      "grad_norm": 0.3081153281219037,
      "learning_rate": 1.88285468964726e-05,
      "loss": 0.5728,
      "mean_token_accuracy": 0.8262367367744445,
      "num_tokens": 149018034.0,
      "step": 135
    },
    {
      "epoch": 0.19896962160241605,
      "grad_norm": 0.27248779720052707,
      "learning_rate": 1.872421593209355e-05,
      "loss": 0.5679,
      "mean_token_accuracy": 0.8270156674087048,
      "num_tokens": 154549711.0,
      "step": 140
    },
    {
      "epoch": 0.20607567951678807,
      "grad_norm": 0.2876029562240766,
      "learning_rate": 1.861578758800989e-05,
      "loss": 0.5739,
      "mean_token_accuracy": 0.825926473736763,
      "num_tokens": 160056919.0,
      "step": 145
    },
    {
      "epoch": 0.21318173743116006,
      "grad_norm": 0.2592081911074665,
      "learning_rate": 1.8503319381010414e-05,
      "loss": 0.5688,
      "mean_token_accuracy": 0.826562087237835,
      "num_tokens": 165584428.0,
      "step": 150
    },
    {
      "epoch": 0.22028779534553206,
      "grad_norm": 0.2921770757113811,
      "learning_rate": 1.8386870970865488e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8287016794085502,
      "num_tokens": 171107691.0,
      "step": 155
    },
    {
      "epoch": 0.22739385325990408,
      "grad_norm": 0.25717865089543174,
      "learning_rate": 1.8266504128679988e-05,
      "loss": 0.5619,
      "mean_token_accuracy": 0.8289386563003063,
      "num_tokens": 176626158.0,
      "step": 160
    },
    {
      "epoch": 0.23449991117427607,
      "grad_norm": 0.27820799246988237,
      "learning_rate": 1.814228270412624e-05,
      "loss": 0.5772,
      "mean_token_accuracy": 0.8250645868480205,
      "num_tokens": 182151744.0,
      "step": 165
    },
    {
      "epoch": 0.24160596908864806,
      "grad_norm": 0.2821779366904109,
      "learning_rate": 1.8014272591574405e-05,
      "loss": 0.5707,
      "mean_token_accuracy": 0.8270999036729336,
      "num_tokens": 187670198.0,
      "step": 170
    },
    {
      "epoch": 0.2487120270030201,
      "grad_norm": 0.3275959246759839,
      "learning_rate": 1.7882541695138224e-05,
      "loss": 0.5579,
      "mean_token_accuracy": 0.8302620410919189,
      "num_tokens": 193189696.0,
      "step": 175
    },
    {
      "epoch": 0.25581808491739205,
      "grad_norm": 0.29306437619143894,
      "learning_rate": 1.7747159892654646e-05,
      "loss": 0.5564,
      "mean_token_accuracy": 0.8296913146972656,
      "num_tokens": 198706595.0,
      "step": 180
    },
    {
      "epoch": 0.2629241428317641,
      "grad_norm": 0.2977278839411146,
      "learning_rate": 1.7608198998616533e-05,
      "loss": 0.5635,
      "mean_token_accuracy": 0.8275744579732418,
      "num_tokens": 204263565.0,
      "step": 185
    },
    {
      "epoch": 0.2700302007461361,
      "grad_norm": 0.3317727367369204,
      "learning_rate": 1.7465732726077993e-05,
      "loss": 0.559,
      "mean_token_accuracy": 0.8284266702830791,
      "num_tokens": 209753822.0,
      "step": 190
    },
    {
      "epoch": 0.27713625866050806,
      "grad_norm": 0.27234470597159066,
      "learning_rate": 1.731983664755264e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8291354134678841,
      "num_tokens": 215275472.0,
      "step": 195
    },
    {
      "epoch": 0.2842423165748801,
      "grad_norm": 0.26544343779539864,
      "learning_rate": 1.717058815492548e-05,
      "loss": 0.5613,
      "mean_token_accuracy": 0.8280082412064076,
      "num_tokens": 220789450.0,
      "step": 200
    },
    {
      "epoch": 0.2842423165748801,
      "eval_loss": 0.5400242805480957,
      "eval_mean_token_accuracy": 0.8298156413069943,
      "eval_num_tokens": 220789450.0,
      "eval_runtime": 161.1614,
      "eval_samples_per_second": 22.58,
      "eval_steps_per_second": 0.707,
      "step": 200
    },
    {
      "epoch": 0.2913483744892521,
      "grad_norm": 0.2813762656625426,
      "learning_rate": 1.701806641839967e-05,
      "loss": 0.5615,
      "mean_token_accuracy": 0.8296020865440369,
      "num_tokens": 226318335.0,
      "step": 205
    },
    {
      "epoch": 0.29845443240362407,
      "grad_norm": 0.2738481182825542,
      "learning_rate": 1.6862352344500004e-05,
      "loss": 0.5604,
      "mean_token_accuracy": 0.8292979046702385,
      "num_tokens": 231830330.0,
      "step": 210
    },
    {
      "epoch": 0.3055604903179961,
      "grad_norm": 0.2724828799767463,
      "learning_rate": 1.6703528533155283e-05,
      "loss": 0.5555,
      "mean_token_accuracy": 0.8297731988132,
      "num_tokens": 237344871.0,
      "step": 215
    },
    {
      "epoch": 0.3126665482323681,
      "grad_norm": 0.26306657580185255,
      "learning_rate": 1.6541679233882477e-05,
      "loss": 0.5516,
      "mean_token_accuracy": 0.8313047230243683,
      "num_tokens": 242868508.0,
      "step": 220
    },
    {
      "epoch": 0.3197726061467401,
      "grad_norm": 0.2918451339288788,
      "learning_rate": 1.63768903010958e-05,
      "loss": 0.5549,
      "mean_token_accuracy": 0.8298390731215477,
      "num_tokens": 248407721.0,
      "step": 225
    },
    {
      "epoch": 0.3268786640611121,
      "grad_norm": 0.2778992598258234,
      "learning_rate": 1.6209249148564437e-05,
      "loss": 0.5503,
      "mean_token_accuracy": 0.831806804984808,
      "num_tokens": 253911671.0,
      "step": 230
    },
    {
      "epoch": 0.3339847219754841,
      "grad_norm": 0.29816424123268476,
      "learning_rate": 1.603884470304318e-05,
      "loss": 0.5625,
      "mean_token_accuracy": 0.8286240585148335,
      "num_tokens": 259442838.0,
      "step": 235
    },
    {
      "epoch": 0.3410907798898561,
      "grad_norm": 0.3434968878817897,
      "learning_rate": 1.5865767357100383e-05,
      "loss": 0.5454,
      "mean_token_accuracy": 0.8329082369804383,
      "num_tokens": 264975290.0,
      "step": 240
    },
    {
      "epoch": 0.3481968378042281,
      "grad_norm": 0.28342056849119923,
      "learning_rate": 1.5690108921168428e-05,
      "loss": 0.5509,
      "mean_token_accuracy": 0.831651521474123,
      "num_tokens": 270498217.0,
      "step": 245
    },
    {
      "epoch": 0.35530289571860013,
      "grad_norm": 0.2801391925223262,
      "learning_rate": 1.5511962574842073e-05,
      "loss": 0.5506,
      "mean_token_accuracy": 0.8316259779036045,
      "num_tokens": 276014342.0,
      "step": 250
    },
    {
      "epoch": 0.3624089536329721,
      "grad_norm": 0.275450732541475,
      "learning_rate": 1.5331422817450485e-05,
      "loss": 0.5533,
      "mean_token_accuracy": 0.830129113048315,
      "num_tokens": 281549490.0,
      "step": 255
    },
    {
      "epoch": 0.3695150115473441,
      "grad_norm": 0.30780371017473934,
      "learning_rate": 1.5148585417929212e-05,
      "loss": 0.5497,
      "mean_token_accuracy": 0.8321508087217808,
      "num_tokens": 287067865.0,
      "step": 260
    },
    {
      "epoch": 0.37662106946171614,
      "grad_norm": 0.32141139168463606,
      "learning_rate": 1.4963547364018711e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8327535085380078,
      "num_tokens": 292575698.0,
      "step": 265
    },
    {
      "epoch": 0.3837271273760881,
      "grad_norm": 0.2941090649445171,
      "learning_rate": 1.477640681081632e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8316998913884163,
      "num_tokens": 298104461.0,
      "step": 270
    },
    {
      "epoch": 0.39083318529046013,
      "grad_norm": 0.3156113844180264,
      "learning_rate": 1.4587263028709013e-05,
      "loss": 0.546,
      "mean_token_accuracy": 0.8325068384408951,
      "num_tokens": 303615295.0,
      "step": 275
    },
    {
      "epoch": 0.3979392432048321,
      "grad_norm": 0.27848176976199285,
      "learning_rate": 1.4396216350714512e-05,
      "loss": 0.5473,
      "mean_token_accuracy": 0.8326819702982903,
      "num_tokens": 309131939.0,
      "step": 280
    },
    {
      "epoch": 0.4050453011192041,
      "grad_norm": 0.28930577025792253,
      "learning_rate": 1.4203368119258759e-05,
      "loss": 0.5433,
      "mean_token_accuracy": 0.8328328810632228,
      "num_tokens": 314661318.0,
      "step": 285
    },
    {
      "epoch": 0.41215135903357614,
      "grad_norm": 0.3660686775923689,
      "learning_rate": 1.4008820632417906e-05,
      "loss": 0.5401,
      "mean_token_accuracy": 0.8341327331960201,
      "num_tokens": 320185466.0,
      "step": 290
    },
    {
      "epoch": 0.4192574169479481,
      "grad_norm": 0.30653539266099966,
      "learning_rate": 1.381267708965339e-05,
      "loss": 0.5437,
      "mean_token_accuracy": 0.8333369344472885,
      "num_tokens": 325709519.0,
      "step": 295
    },
    {
      "epoch": 0.4263634748623201,
      "grad_norm": 0.2716446207789635,
      "learning_rate": 1.3615041537068831e-05,
      "loss": 0.5515,
      "mean_token_accuracy": 0.8306186564266682,
      "num_tokens": 331248121.0,
      "step": 300
    },
    {
      "epoch": 0.4263634748623201,
      "eval_loss": 0.5254271030426025,
      "eval_mean_token_accuracy": 0.8335580465040708,
      "eval_num_tokens": 331248121.0,
      "eval_runtime": 160.8696,
      "eval_samples_per_second": 22.621,
      "eval_steps_per_second": 0.709,
      "step": 300
    },
    {
      "epoch": 0.43346953277669215,
      "grad_norm": 0.27817622720791757,
      "learning_rate": 1.3416018812217866e-05,
      "loss": 0.5495,
      "mean_token_accuracy": 0.8312446601688862,
      "num_tokens": 336772706.0,
      "step": 305
    },
    {
      "epoch": 0.4405755906910641,
      "grad_norm": 0.2770013355984933,
      "learning_rate": 1.3215714488492121e-05,
      "loss": 0.535,
      "mean_token_accuracy": 0.8355906143784523,
      "num_tokens": 342302933.0,
      "step": 310
    },
    {
      "epoch": 0.44768164860543613,
      "grad_norm": 0.2774944414170579,
      "learning_rate": 1.3014234819118846e-05,
      "loss": 0.533,
      "mean_token_accuracy": 0.8360061995685101,
      "num_tokens": 347832122.0,
      "step": 315
    },
    {
      "epoch": 0.45478770651980815,
      "grad_norm": 0.2822504430152847,
      "learning_rate": 1.2811686680797942e-05,
      "loss": 0.5464,
      "mean_token_accuracy": 0.8318669557571411,
      "num_tokens": 353357495.0,
      "step": 320
    },
    {
      "epoch": 0.4618937644341801,
      "grad_norm": 0.25912651097009326,
      "learning_rate": 1.2608177517008268e-05,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8345673441886902,
      "num_tokens": 358876427.0,
      "step": 325
    },
    {
      "epoch": 0.46899982234855214,
      "grad_norm": 0.2899391004734366,
      "learning_rate": 1.240381528101327e-05,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8358384124934674,
      "num_tokens": 364395480.0,
      "step": 330
    },
    {
      "epoch": 0.47610588026292416,
      "grad_norm": 0.2885199318810423,
      "learning_rate": 1.2198708378596198e-05,
      "loss": 0.5276,
      "mean_token_accuracy": 0.8363832160830498,
      "num_tokens": 369904006.0,
      "step": 335
    },
    {
      "epoch": 0.48321193817729613,
      "grad_norm": 0.2803340014077296,
      "learning_rate": 1.19929656105553e-05,
      "loss": 0.5313,
      "mean_token_accuracy": 0.8354136534035206,
      "num_tokens": 375415817.0,
      "step": 340
    },
    {
      "epoch": 0.49031799609166815,
      "grad_norm": 0.2841127075165474,
      "learning_rate": 1.1786696114989455e-05,
      "loss": 0.5324,
      "mean_token_accuracy": 0.8360840916633606,
      "num_tokens": 380931196.0,
      "step": 345
    },
    {
      "epoch": 0.4974240540060402,
      "grad_norm": 0.27070169414461726,
      "learning_rate": 1.1580009309404887e-05,
      "loss": 0.5339,
      "mean_token_accuracy": 0.8355580635368824,
      "num_tokens": 386462024.0,
      "step": 350
    },
    {
      "epoch": 0.5045301119204122,
      "grad_norm": 0.27595636054491235,
      "learning_rate": 1.1373014832673661e-05,
      "loss": 0.5363,
      "mean_token_accuracy": 0.8354052670300007,
      "num_tokens": 391971484.0,
      "step": 355
    },
    {
      "epoch": 0.5116361698347841,
      "grad_norm": 0.2526139742391355,
      "learning_rate": 1.1165822486874773e-05,
      "loss": 0.5288,
      "mean_token_accuracy": 0.8361794017255306,
      "num_tokens": 397503191.0,
      "step": 360
    },
    {
      "epoch": 0.5187422277491561,
      "grad_norm": 0.28153871306210954,
      "learning_rate": 1.0958542179048637e-05,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8368273265659809,
      "num_tokens": 403001146.0,
      "step": 365
    },
    {
      "epoch": 0.5258482856635281,
      "grad_norm": 0.3145297412780038,
      "learning_rate": 1.0751283862895914e-05,
      "loss": 0.5422,
      "mean_token_accuracy": 0.8336515955626964,
      "num_tokens": 408526858.0,
      "step": 370
    },
    {
      "epoch": 0.5329543435779002,
      "grad_norm": 0.28928430258415916,
      "learning_rate": 1.0544157480451586e-05,
      "loss": 0.5402,
      "mean_token_accuracy": 0.833558625727892,
      "num_tokens": 414051639.0,
      "step": 375
    },
    {
      "epoch": 0.5400604014922722,
      "grad_norm": 0.26142431075193084,
      "learning_rate": 1.033727290376522e-05,
      "loss": 0.542,
      "mean_token_accuracy": 0.8336416870355606,
      "num_tokens": 419571564.0,
      "step": 380
    },
    {
      "epoch": 0.5471664594066442,
      "grad_norm": 0.31125146975357,
      "learning_rate": 1.013073987661834e-05,
      "loss": 0.54,
      "mean_token_accuracy": 0.8338716469705105,
      "num_tokens": 425096738.0,
      "step": 385
    },
    {
      "epoch": 0.5542725173210161,
      "grad_norm": 0.2668746904615073,
      "learning_rate": 9.924667956309862e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.836452516913414,
      "num_tokens": 430620174.0,
      "step": 390
    },
    {
      "epoch": 0.5613785752353881,
      "grad_norm": 0.40741674797587885,
      "learning_rate": 9.719166455540437e-06,
      "loss": 0.5374,
      "mean_token_accuracy": 0.8344319149851799,
      "num_tokens": 436140181.0,
      "step": 395
    },
    {
      "epoch": 0.5684846331497602,
      "grad_norm": 0.27348406881195864,
      "learning_rate": 9.51434438442655e-06,
      "loss": 0.5348,
      "mean_token_accuracy": 0.8360598988831043,
      "num_tokens": 441662675.0,
      "step": 400
    },
    {
      "epoch": 0.5684846331497602,
      "eval_loss": 0.5157341361045837,
      "eval_mean_token_accuracy": 0.8361675456950539,
      "eval_num_tokens": 441662675.0,
      "eval_runtime": 161.4058,
      "eval_samples_per_second": 22.546,
      "eval_steps_per_second": 0.706,
      "step": 400
    },
    {
      "epoch": 0.5755906910641322,
      "grad_norm": 0.27295040196213805,
      "learning_rate": 9.310310392675132e-06,
      "loss": 0.5232,
      "mean_token_accuracy": 0.8382402293384075,
      "num_tokens": 447154493.0,
      "step": 405
    },
    {
      "epoch": 0.5826967489785042,
      "grad_norm": 0.26519161043459566,
      "learning_rate": 9.107172711949324e-06,
      "loss": 0.5382,
      "mean_token_accuracy": 0.8342290692031383,
      "num_tokens": 452684121.0,
      "step": 410
    },
    {
      "epoch": 0.5898028068928762,
      "grad_norm": 0.2629255148134165,
      "learning_rate": 8.905039098456049e-06,
      "loss": 0.53,
      "mean_token_accuracy": 0.835949394851923,
      "num_tokens": 458209573.0,
      "step": 415
    },
    {
      "epoch": 0.5969088648072481,
      "grad_norm": 0.2762404944964492,
      "learning_rate": 8.704016775785742e-06,
      "loss": 0.5345,
      "mean_token_accuracy": 0.8355151884257793,
      "num_tokens": 463743007.0,
      "step": 420
    },
    {
      "epoch": 0.6040149227216202,
      "grad_norm": 0.3862082589618725,
      "learning_rate": 8.50421237803464e-06,
      "loss": 0.529,
      "mean_token_accuracy": 0.8362281493842602,
      "num_tokens": 469259672.0,
      "step": 425
    },
    {
      "epoch": 0.6111209806359922,
      "grad_norm": 0.29912793555607287,
      "learning_rate": 8.30573189323978e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.8389238156378269,
      "num_tokens": 474761309.0,
      "step": 430
    },
    {
      "epoch": 0.6182270385503642,
      "grad_norm": 0.28697574980998686,
      "learning_rate": 8.108680607156669e-06,
      "loss": 0.5373,
      "mean_token_accuracy": 0.8346886426210404,
      "num_tokens": 480283239.0,
      "step": 435
    },
    {
      "epoch": 0.6253330964647362,
      "grad_norm": 0.2981386592415112,
      "learning_rate": 7.913163047409533e-06,
      "loss": 0.5299,
      "mean_token_accuracy": 0.8359711997210979,
      "num_tokens": 485803053.0,
      "step": 440
    },
    {
      "epoch": 0.6324391543791081,
      "grad_norm": 0.2811128507955913,
      "learning_rate": 7.719282928043688e-06,
      "loss": 0.5312,
      "mean_token_accuracy": 0.8358238264918327,
      "num_tokens": 491312112.0,
      "step": 445
    },
    {
      "epoch": 0.6395452122934802,
      "grad_norm": 0.2538506418412949,
      "learning_rate": 7.527143094509492e-06,
      "loss": 0.5292,
      "mean_token_accuracy": 0.8370420016348362,
      "num_tokens": 496828932.0,
      "step": 450
    },
    {
      "epoch": 0.6466512702078522,
      "grad_norm": 0.26664943478343317,
      "learning_rate": 7.336845469107061e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8362418174743652,
      "num_tokens": 502329483.0,
      "step": 455
    },
    {
      "epoch": 0.6537573281222242,
      "grad_norm": 0.2822285091829132,
      "learning_rate": 7.148490996920661e-06,
      "loss": 0.5315,
      "mean_token_accuracy": 0.8359686724841595,
      "num_tokens": 507853922.0,
      "step": 460
    },
    {
      "epoch": 0.6608633860365962,
      "grad_norm": 0.26891344059858413,
      "learning_rate": 6.9621795922714805e-06,
      "loss": 0.5283,
      "mean_token_accuracy": 0.837210227549076,
      "num_tokens": 513366033.0,
      "step": 465
    },
    {
      "epoch": 0.6679694439509682,
      "grad_norm": 0.27887938299077314,
      "learning_rate": 6.778010085717202e-06,
      "loss": 0.5275,
      "mean_token_accuracy": 0.8373380437493324,
      "num_tokens": 518888798.0,
      "step": 470
    },
    {
      "epoch": 0.6750755018653402,
      "grad_norm": 0.25953435317326623,
      "learning_rate": 6.596080171626409e-06,
      "loss": 0.5307,
      "mean_token_accuracy": 0.8357150256633759,
      "num_tokens": 524392285.0,
      "step": 475
    },
    {
      "epoch": 0.6821815597797122,
      "grad_norm": 0.2665111343594617,
      "learning_rate": 6.416486356355769e-06,
      "loss": 0.5365,
      "mean_token_accuracy": 0.8344089902937413,
      "num_tokens": 529904641.0,
      "step": 480
    },
    {
      "epoch": 0.6892876176940842,
      "grad_norm": 0.26605568148767894,
      "learning_rate": 6.239323907057342e-06,
      "loss": 0.5334,
      "mean_token_accuracy": 0.8357222154736519,
      "num_tokens": 535417362.0,
      "step": 485
    },
    {
      "epoch": 0.6963936756084562,
      "grad_norm": 0.2636227512736411,
      "learning_rate": 6.064686801143271e-06,
      "loss": 0.5166,
      "mean_token_accuracy": 0.8401257589459419,
      "num_tokens": 540908836.0,
      "step": 490
    },
    {
      "epoch": 0.7034997335228282,
      "grad_norm": 0.26826679232229517,
      "learning_rate": 5.892667676434633e-06,
      "loss": 0.5238,
      "mean_token_accuracy": 0.8382793001830577,
      "num_tokens": 546444796.0,
      "step": 495
    },
    {
      "epoch": 0.7106057914372003,
      "grad_norm": 0.3599829471194239,
      "learning_rate": 5.723357782020867e-06,
      "loss": 0.5225,
      "mean_token_accuracy": 0.838120236247778,
      "num_tokens": 551952947.0,
      "step": 500
    },
    {
      "epoch": 0.7106057914372003,
      "eval_loss": 0.5095566511154175,
      "eval_mean_token_accuracy": 0.8375082057819032,
      "eval_num_tokens": 551952947.0,
      "eval_runtime": 161.2359,
      "eval_samples_per_second": 22.569,
      "eval_steps_per_second": 0.707,
      "step": 500
    },
    {
      "epoch": 0.7177118493515722,
      "grad_norm": 0.26704698034936375,
      "learning_rate": 5.556846929855857e-06,
      "loss": 0.5203,
      "mean_token_accuracy": 0.8389924250543117,
      "num_tokens": 557467246.0,
      "step": 505
    },
    {
      "epoch": 0.7248179072659442,
      "grad_norm": 0.26539658794683185,
      "learning_rate": 5.393223447116409e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8358133606612682,
      "num_tokens": 562988077.0,
      "step": 510
    },
    {
      "epoch": 0.7319239651803162,
      "grad_norm": 0.24829603150024088,
      "learning_rate": 5.232574129348278e-06,
      "loss": 0.5235,
      "mean_token_accuracy": 0.8386510953307151,
      "num_tokens": 568506583.0,
      "step": 515
    },
    {
      "epoch": 0.7390300230946882,
      "grad_norm": 0.2605719227696348,
      "learning_rate": 5.0749841944247e-06,
      "loss": 0.5338,
      "mean_token_accuracy": 0.83480354398489,
      "num_tokens": 574039672.0,
      "step": 520
    },
    {
      "epoch": 0.7461360810090603,
      "grad_norm": 0.26778891488350126,
      "learning_rate": 4.92053723734182e-06,
      "loss": 0.5314,
      "mean_token_accuracy": 0.8356986902654171,
      "num_tokens": 579544417.0,
      "step": 525
    },
    {
      "epoch": 0.7532421389234323,
      "grad_norm": 0.42601785989143315,
      "learning_rate": 4.769315185874951e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8374913208186626,
      "num_tokens": 585066703.0,
      "step": 530
    },
    {
      "epoch": 0.7603481968378042,
      "grad_norm": 0.2541554106960908,
      "learning_rate": 4.621398257119266e-06,
      "loss": 0.5262,
      "mean_token_accuracy": 0.837667242437601,
      "num_tokens": 590596017.0,
      "step": 535
    },
    {
      "epoch": 0.7674542547521762,
      "grad_norm": 0.2695567911282263,
      "learning_rate": 4.476864914937923e-06,
      "loss": 0.5195,
      "mean_token_accuracy": 0.8392265714704991,
      "num_tokens": 596113106.0,
      "step": 540
    },
    {
      "epoch": 0.7745603126665482,
      "grad_norm": 0.25052432871031344,
      "learning_rate": 4.335791828340183e-06,
      "loss": 0.5296,
      "mean_token_accuracy": 0.837313498556614,
      "num_tokens": 601659151.0,
      "step": 545
    },
    {
      "epoch": 0.7816663705809203,
      "grad_norm": 0.26020461836934194,
      "learning_rate": 4.1982538308116775e-06,
      "loss": 0.5251,
      "mean_token_accuracy": 0.8361369468271732,
      "num_tokens": 607175134.0,
      "step": 550
    },
    {
      "epoch": 0.7887724284952923,
      "grad_norm": 0.24684211829794916,
      "learning_rate": 4.064323880618279e-06,
      "loss": 0.5274,
      "mean_token_accuracy": 0.8380102440714836,
      "num_tokens": 612699036.0,
      "step": 555
    },
    {
      "epoch": 0.7958784864096642,
      "grad_norm": 0.2745271550137196,
      "learning_rate": 3.934073022104759e-06,
      "loss": 0.5233,
      "mean_token_accuracy": 0.8380540162324905,
      "num_tokens": 618232472.0,
      "step": 560
    },
    {
      "epoch": 0.8029845443240362,
      "grad_norm": 0.2695298031436579,
      "learning_rate": 3.807570348008672e-06,
      "loss": 0.5243,
      "mean_token_accuracy": 0.8377192810177803,
      "num_tokens": 623766598.0,
      "step": 565
    },
    {
      "epoch": 0.8100906022384082,
      "grad_norm": 0.2646862071287686,
      "learning_rate": 3.684882962809484e-06,
      "loss": 0.5304,
      "mean_token_accuracy": 0.8363631062209607,
      "num_tokens": 629297697.0,
      "step": 570
    },
    {
      "epoch": 0.8171966601527803,
      "grad_norm": 0.23795317969691135,
      "learning_rate": 3.5660759471324037e-06,
      "loss": 0.5287,
      "mean_token_accuracy": 0.836943382024765,
      "num_tokens": 634808630.0,
      "step": 575
    },
    {
      "epoch": 0.8243027180671523,
      "grad_norm": 0.25754712205145064,
      "learning_rate": 3.451212323225786e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8387916676700116,
      "num_tokens": 640315396.0,
      "step": 580
    },
    {
      "epoch": 0.8314087759815243,
      "grad_norm": 0.24240514105833283,
      "learning_rate": 3.340353021530409e-06,
      "loss": 0.5214,
      "mean_token_accuracy": 0.8374654315412045,
      "num_tokens": 645841425.0,
      "step": 585
    },
    {
      "epoch": 0.8385148338958962,
      "grad_norm": 0.26461850588888575,
      "learning_rate": 3.2335568483583708e-06,
      "loss": 0.516,
      "mean_token_accuracy": 0.8414874106645585,
      "num_tokens": 651349089.0,
      "step": 590
    },
    {
      "epoch": 0.8456208918102682,
      "grad_norm": 0.25434414665474275,
      "learning_rate": 3.1308804546987615e-06,
      "loss": 0.5305,
      "mean_token_accuracy": 0.8366403698921203,
      "num_tokens": 656881938.0,
      "step": 595
    },
    {
      "epoch": 0.8527269497246402,
      "grad_norm": 0.25710193414257565,
      "learning_rate": 3.0323783061666307e-06,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8387804657220841,
      "num_tokens": 662399294.0,
      "step": 600
    },
    {
      "epoch": 0.8527269497246402,
      "eval_loss": 0.5056362748146057,
      "eval_mean_token_accuracy": 0.8385671649062842,
      "eval_num_tokens": 662399294.0,
      "eval_runtime": 161.5872,
      "eval_samples_per_second": 22.52,
      "eval_steps_per_second": 0.706,
      "step": 600
    },
    {
      "epoch": 0.8598330076390123,
      "grad_norm": 0.24729771911584983,
      "learning_rate": 2.9381026541112145e-06,
      "loss": 0.5237,
      "mean_token_accuracy": 0.8380155004560947,
      "num_tokens": 667931430.0,
      "step": 605
    },
    {
      "epoch": 0.8669390655533843,
      "grad_norm": 0.25660228503672733,
      "learning_rate": 2.848103507898745e-06,
      "loss": 0.5269,
      "mean_token_accuracy": 0.8368470750749111,
      "num_tokens": 673461286.0,
      "step": 610
    },
    {
      "epoch": 0.8740451234677563,
      "grad_norm": 0.2856740804187853,
      "learning_rate": 2.7624286083845187e-06,
      "loss": 0.5218,
      "mean_token_accuracy": 0.8375120624899864,
      "num_tokens": 678984176.0,
      "step": 615
    },
    {
      "epoch": 0.8811511813821282,
      "grad_norm": 0.2678064141268326,
      "learning_rate": 2.6811234025883457e-06,
      "loss": 0.5172,
      "mean_token_accuracy": 0.839438085258007,
      "num_tokens": 684519341.0,
      "step": 620
    },
    {
      "epoch": 0.8882572392965002,
      "grad_norm": 0.26981981962318713,
      "learning_rate": 2.604231019586761e-06,
      "loss": 0.5207,
      "mean_token_accuracy": 0.839027612656355,
      "num_tokens": 690037943.0,
      "step": 625
    },
    {
      "epoch": 0.8953632972108723,
      "grad_norm": 0.2653800927885472,
      "learning_rate": 2.5317922476348194e-06,
      "loss": 0.523,
      "mean_token_accuracy": 0.8382897555828095,
      "num_tokens": 695563755.0,
      "step": 630
    },
    {
      "epoch": 0.9024693551252443,
      "grad_norm": 0.2909130920485432,
      "learning_rate": 2.4638455125296043e-06,
      "loss": 0.5246,
      "mean_token_accuracy": 0.837975486367941,
      "num_tokens": 701088593.0,
      "step": 635
    },
    {
      "epoch": 0.9095754130396163,
      "grad_norm": 0.2538310488730525,
      "learning_rate": 2.400426857226914e-06,
      "loss": 0.5181,
      "mean_token_accuracy": 0.8393405571579933,
      "num_tokens": 706621161.0,
      "step": 640
    },
    {
      "epoch": 0.9166814709539882,
      "grad_norm": 0.2569434114852769,
      "learning_rate": 2.3415699227219517e-06,
      "loss": 0.5295,
      "mean_token_accuracy": 0.8363286212086678,
      "num_tokens": 712135507.0,
      "step": 645
    },
    {
      "epoch": 0.9237875288683602,
      "grad_norm": 0.26172097069920536,
      "learning_rate": 2.2873059302041627e-06,
      "loss": 0.52,
      "mean_token_accuracy": 0.8396071724593639,
      "num_tokens": 717654276.0,
      "step": 650
    },
    {
      "epoch": 0.9308935867827323,
      "grad_norm": 0.26549742257601705,
      "learning_rate": 2.2376636644956656e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8391963638365268,
      "num_tokens": 723171964.0,
      "step": 655
    },
    {
      "epoch": 0.9379996446971043,
      "grad_norm": 0.2556259295868791,
      "learning_rate": 2.192669458782096e-06,
      "loss": 0.5255,
      "mean_token_accuracy": 0.8376853354275227,
      "num_tokens": 728678520.0,
      "step": 660
    },
    {
      "epoch": 0.9451057026114763,
      "grad_norm": 0.2511199627055941,
      "learning_rate": 2.1523471806439205e-06,
      "loss": 0.5344,
      "mean_token_accuracy": 0.8351106189191342,
      "num_tokens": 734210332.0,
      "step": 665
    },
    {
      "epoch": 0.9522117605258483,
      "grad_norm": 0.2744720323508251,
      "learning_rate": 2.1167182193956738e-06,
      "loss": 0.5156,
      "mean_token_accuracy": 0.8407034426927567,
      "num_tokens": 739702586.0,
      "step": 670
    },
    {
      "epoch": 0.9593178184402202,
      "grad_norm": 0.2996101632719976,
      "learning_rate": 2.0858014747397952e-06,
      "loss": 0.5245,
      "mean_token_accuracy": 0.8386490792036057,
      "num_tokens": 745219812.0,
      "step": 675
    },
    {
      "epoch": 0.9664238763545923,
      "grad_norm": 0.2615861429363619,
      "learning_rate": 2.0596133467411213e-06,
      "loss": 0.5175,
      "mean_token_accuracy": 0.8394487954676151,
      "num_tokens": 750717984.0,
      "step": 680
    },
    {
      "epoch": 0.9735299342689643,
      "grad_norm": 0.2670223147061545,
      "learning_rate": 2.0381677271273177e-06,
      "loss": 0.5209,
      "mean_token_accuracy": 0.8379806369543076,
      "num_tokens": 756252357.0,
      "step": 685
    },
    {
      "epoch": 0.9806359921833363,
      "grad_norm": 0.260083835300812,
      "learning_rate": 2.0214759919198904e-06,
      "loss": 0.5149,
      "mean_token_accuracy": 0.839554088562727,
      "num_tokens": 761750059.0,
      "step": 690
    },
    {
      "epoch": 0.9877420500977083,
      "grad_norm": 0.3230971032122484,
      "learning_rate": 2.0095469953996724e-06,
      "loss": 0.5244,
      "mean_token_accuracy": 0.8375246554613114,
      "num_tokens": 767275567.0,
      "step": 695
    },
    {
      "epoch": 0.9948481080120803,
      "grad_norm": 0.26793724105887723,
      "learning_rate": 2.002387065409989e-06,
      "loss": 0.527,
      "mean_token_accuracy": 0.8372870542109012,
      "num_tokens": 772815373.0,
      "step": 700
    },
    {
      "epoch": 0.9948481080120803,
      "eval_loss": 0.5037879943847656,
      "eval_mean_token_accuracy": 0.8392795601434875,
      "eval_num_tokens": 772815373.0,
      "eval_runtime": 162.0902,
      "eval_samples_per_second": 22.45,
      "eval_steps_per_second": 0.703,
      "step": 700
    },
    {
      "epoch": 1.0,
      "mean_token_accuracy": 0.8380202034424091,
      "num_tokens": 776817701.0,
      "step": 704,
      "total_flos": 6035679775555584.0,
      "train_loss": 0.6108774590221319,
      "train_runtime": 28363.9285,
      "train_samples_per_second": 12.701,
      "train_steps_per_second": 0.025
    }
  ],
  "logging_steps": 5,
  "max_steps": 704,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6035679775555584.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}