{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 804,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 1.9214146781432249,
      "learning_rate": 4.8780487804878055e-06,
      "loss": 0.857,
      "step": 5
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 1.7625595046451557,
      "learning_rate": 1.0975609756097562e-05,
      "loss": 0.8038,
      "step": 10
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.6948661270341572,
      "learning_rate": 1.707317073170732e-05,
      "loss": 0.7448,
      "step": 15
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.5136824696543728,
      "learning_rate": 2.3170731707317075e-05,
      "loss": 0.7055,
      "step": 20
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.4018338347498522,
      "learning_rate": 2.926829268292683e-05,
      "loss": 0.6712,
      "step": 25
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.28711501314318655,
      "learning_rate": 3.5365853658536584e-05,
      "loss": 0.6599,
      "step": 30
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.2894771986907634,
      "learning_rate": 4.146341463414634e-05,
      "loss": 0.6395,
      "step": 35
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.2821436189327092,
      "learning_rate": 4.75609756097561e-05,
      "loss": 0.627,
      "step": 40
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.3279211667096163,
      "learning_rate": 4.999828351434079e-05,
      "loss": 0.6245,
      "step": 45
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.2537209910093848,
      "learning_rate": 4.998779482816942e-05,
      "loss": 0.6048,
      "step": 50
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.30352094230828874,
      "learning_rate": 4.996777549883426e-05,
      "loss": 0.6056,
      "step": 55
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.35358905732065365,
      "learning_rate": 4.9938234010808136e-05,
      "loss": 0.5995,
      "step": 60
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.5881353923808288,
      "learning_rate": 4.989918288418841e-05,
      "loss": 0.5898,
      "step": 65
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.4365964873532165,
      "learning_rate": 4.9850638669390816e-05,
      "loss": 0.5856,
      "step": 70
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.3310629283404945,
      "learning_rate": 4.97926219401351e-05,
      "loss": 0.583,
      "step": 75
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.39095858353376106,
      "learning_rate": 4.9725157284725665e-05,
      "loss": 0.5818,
      "step": 80
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.4130266981460891,
      "learning_rate": 4.964827329563061e-05,
      "loss": 0.5777,
      "step": 85
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.3251791282337107,
      "learning_rate": 4.956200255736394e-05,
      "loss": 0.5802,
      "step": 90
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.2800299594598508,
      "learning_rate": 4.9466381632675714e-05,
      "loss": 0.5777,
      "step": 95
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.30351899813854083,
      "learning_rate": 4.936145104705629e-05,
      "loss": 0.5732,
      "step": 100
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.3184280675590268,
      "learning_rate": 4.9247255271560994e-05,
      "loss": 0.5693,
      "step": 105
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.35433866607808917,
      "learning_rate": 4.9123842703962754e-05,
      "loss": 0.5762,
      "step": 110
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.28639350303291805,
      "learning_rate": 4.899126564824033e-05,
      "loss": 0.5666,
      "step": 115
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.3274186200975153,
      "learning_rate": 4.884958029241127e-05,
      "loss": 0.5674,
      "step": 120
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.336381141981706,
      "learning_rate": 4.869884668471853e-05,
      "loss": 0.5641,
      "step": 125
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.4812221145192079,
      "learning_rate": 4.8539128708181276e-05,
      "loss": 0.563,
      "step": 130
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.2985625122934817,
      "learning_rate": 4.8370494053520316e-05,
      "loss": 0.5599,
      "step": 135
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.40904102656943087,
      "learning_rate": 4.8193014190469815e-05,
      "loss": 0.5651,
      "step": 140
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.30462346527690765,
      "learning_rate": 4.800676433748746e-05,
      "loss": 0.5571,
      "step": 145
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.29542772666907074,
      "learning_rate": 4.781182342987577e-05,
      "loss": 0.5548,
      "step": 150
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.3051473053785931,
      "learning_rate": 4.7608274086328275e-05,
      "loss": 0.555,
      "step": 155
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.27345084859980273,
      "learning_rate": 4.739620257391446e-05,
      "loss": 0.5525,
      "step": 160
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.289181458801349,
      "learning_rate": 4.7175698771518656e-05,
      "loss": 0.551,
      "step": 165
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.41307993000424736,
      "learning_rate": 4.6946856131748076e-05,
      "loss": 0.5517,
      "step": 170
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.5340748917517905,
      "learning_rate": 4.6709771641326244e-05,
      "loss": 0.5509,
      "step": 175
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.3291048059480862,
      "learning_rate": 4.6464545779988757e-05,
      "loss": 0.5428,
      "step": 180
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.28986748904908194,
      "learning_rate": 4.621128247789846e-05,
      "loss": 0.5456,
      "step": 185
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.25439473449314687,
      "learning_rate": 4.595008907159847e-05,
      "loss": 0.5485,
      "step": 190
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.22286184197901884,
      "learning_rate": 4.568107625852136e-05,
      "loss": 0.5508,
      "step": 195
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.2642945823026863,
      "learning_rate": 4.5404358050074115e-05,
      "loss": 0.5448,
      "step": 200
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.27448365288562576,
      "learning_rate": 4.512005172331842e-05,
      "loss": 0.5424,
      "step": 205
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.2971424901345131,
      "learning_rate": 4.482827777126706e-05,
      "loss": 0.5471,
      "step": 210
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.35310755203104544,
      "learning_rate": 4.4529159851817255e-05,
      "loss": 0.541,
      "step": 215
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.308625434473984,
      "learning_rate": 4.422282473534271e-05,
      "loss": 0.5405,
      "step": 220
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.31816940842512614,
      "learning_rate": 4.3909402250966534e-05,
      "loss": 0.546,
      "step": 225
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.22827227012944756,
      "learning_rate": 4.358902523153791e-05,
      "loss": 0.542,
      "step": 230
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.24079886565005823,
      "learning_rate": 4.326182945733555e-05,
      "loss": 0.5416,
      "step": 235
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.3638342950483874,
      "learning_rate": 4.292795359852221e-05,
      "loss": 0.5372,
      "step": 240
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.3256070901197167,
      "learning_rate": 4.2587539156374295e-05,
      "loss": 0.5405,
      "step": 245
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.3165450063227312,
      "learning_rate": 4.2240730403311586e-05,
      "loss": 0.5364,
      "step": 250
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.30669208285819877,
      "learning_rate": 4.188767432175263e-05,
      "loss": 0.5322,
      "step": 255
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.257272912960744,
      "learning_rate": 4.1528520541821506e-05,
      "loss": 0.5353,
      "step": 260
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.30370874605896786,
      "learning_rate": 4.116342127793245e-05,
      "loss": 0.5385,
      "step": 265
    },
    {
      "epoch": 1.007462686567164,
      "grad_norm": 0.23367419347190874,
      "learning_rate": 4.0792531264279285e-05,
      "loss": 0.5246,
      "step": 270
    },
    {
      "epoch": 1.0261194029850746,
      "grad_norm": 0.28960025979101867,
      "learning_rate": 4.041600768925687e-05,
      "loss": 0.5175,
      "step": 275
    },
    {
      "epoch": 1.044776119402985,
      "grad_norm": 0.3544154485769114,
      "learning_rate": 4.0034010128842484e-05,
      "loss": 0.5183,
      "step": 280
    },
    {
      "epoch": 1.0634328358208955,
      "grad_norm": 0.2830781283593527,
      "learning_rate": 3.964670047896525e-05,
      "loss": 0.5181,
      "step": 285
    },
    {
      "epoch": 1.0820895522388059,
      "grad_norm": 0.26140038642658264,
      "learning_rate": 3.925424288689239e-05,
      "loss": 0.5087,
      "step": 290
    },
    {
      "epoch": 1.1007462686567164,
      "grad_norm": 0.2617191990469776,
      "learning_rate": 3.8856803681661296e-05,
      "loss": 0.516,
      "step": 295
    },
    {
      "epoch": 1.1194029850746268,
      "grad_norm": 0.26589479989678716,
      "learning_rate": 3.8454551303586964e-05,
      "loss": 0.5111,
      "step": 300
    },
    {
      "epoch": 1.1380597014925373,
      "grad_norm": 0.2406368600470144,
      "learning_rate": 3.8047656232874624e-05,
      "loss": 0.5139,
      "step": 305
    },
    {
      "epoch": 1.1567164179104479,
      "grad_norm": 0.25755826798857223,
      "learning_rate": 3.763629091736781e-05,
      "loss": 0.5116,
      "step": 310
    },
    {
      "epoch": 1.1753731343283582,
      "grad_norm": 0.33892425739707405,
      "learning_rate": 3.722062969946254e-05,
      "loss": 0.5118,
      "step": 315
    },
    {
      "epoch": 1.1940298507462686,
      "grad_norm": 0.23443314308723368,
      "learning_rate": 3.6800848742218644e-05,
      "loss": 0.5133,
      "step": 320
    },
    {
      "epoch": 1.212686567164179,
      "grad_norm": 0.23884530654624148,
      "learning_rate": 3.6377125954699254e-05,
      "loss": 0.5113,
      "step": 325
    },
    {
      "epoch": 1.2313432835820897,
      "grad_norm": 0.2090085180372967,
      "learning_rate": 3.5949640916570566e-05,
      "loss": 0.5085,
      "step": 330
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.23291376449474052,
      "learning_rate": 3.551857480199336e-05,
      "loss": 0.5135,
      "step": 335
    },
    {
      "epoch": 1.2686567164179103,
      "grad_norm": 0.24011296587040026,
      "learning_rate": 3.5084110302838916e-05,
      "loss": 0.505,
      "step": 340
    },
    {
      "epoch": 1.287313432835821,
      "grad_norm": 0.21935005312866615,
      "learning_rate": 3.464643155126162e-05,
      "loss": 0.5104,
      "step": 345
    },
    {
      "epoch": 1.3059701492537314,
      "grad_norm": 0.2261621862735514,
      "learning_rate": 3.4205724041661135e-05,
      "loss": 0.5081,
      "step": 350
    },
    {
      "epoch": 1.3246268656716418,
      "grad_norm": 0.2574439184051825,
      "learning_rate": 3.376217455206732e-05,
      "loss": 0.5115,
      "step": 355
    },
    {
      "epoch": 1.3432835820895521,
      "grad_norm": 0.2129876722579311,
      "learning_rate": 3.3315971064981025e-05,
      "loss": 0.506,
      "step": 360
    },
    {
      "epoch": 1.3619402985074627,
      "grad_norm": 0.20753843021299664,
      "learning_rate": 3.286730268770452e-05,
      "loss": 0.5038,
      "step": 365
    },
    {
      "epoch": 1.3805970149253732,
      "grad_norm": 0.3217141035376458,
      "learning_rate": 3.2416359572195155e-05,
      "loss": 0.5132,
      "step": 370
    },
    {
      "epoch": 1.3992537313432836,
      "grad_norm": 0.21539693831207604,
      "learning_rate": 3.1963332834476247e-05,
      "loss": 0.5081,
      "step": 375
    },
    {
      "epoch": 1.417910447761194,
      "grad_norm": 0.27467469680387524,
      "learning_rate": 3.150841447363948e-05,
      "loss": 0.5067,
      "step": 380
    },
    {
      "epoch": 1.4365671641791045,
      "grad_norm": 0.23691311945378055,
      "learning_rate": 3.1051797290472966e-05,
      "loss": 0.5018,
      "step": 385
    },
    {
      "epoch": 1.455223880597015,
      "grad_norm": 0.30787370613511683,
      "learning_rate": 3.059367480574958e-05,
      "loss": 0.5035,
      "step": 390
    },
    {
      "epoch": 1.4738805970149254,
      "grad_norm": 0.21214521272783723,
      "learning_rate": 3.0134241178210103e-05,
      "loss": 0.5004,
      "step": 395
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 0.2567066053575565,
      "learning_rate": 2.9673691122276086e-05,
      "loss": 0.5041,
      "step": 400
    },
    {
      "epoch": 1.5111940298507462,
      "grad_norm": 0.23880722667167875,
      "learning_rate": 2.9212219825527075e-05,
      "loss": 0.4981,
      "step": 405
    },
    {
      "epoch": 1.5298507462686568,
      "grad_norm": 0.19930339353323281,
      "learning_rate": 2.8750022865977443e-05,
      "loss": 0.502,
      "step": 410
    },
    {
      "epoch": 1.5485074626865671,
      "grad_norm": 0.24784475664597616,
      "learning_rate": 2.82872961291876e-05,
      "loss": 0.5068,
      "step": 415
    },
    {
      "epoch": 1.5671641791044775,
      "grad_norm": 0.2637097189189239,
      "learning_rate": 2.7824235725245042e-05,
      "loss": 0.5031,
      "step": 420
    },
    {
      "epoch": 1.585820895522388,
      "grad_norm": 0.20097295957360145,
      "learning_rate": 2.7361037905650032e-05,
      "loss": 0.5051,
      "step": 425
    },
    {
      "epoch": 1.6044776119402986,
      "grad_norm": 0.1855097808926682,
      "learning_rate": 2.689789898014155e-05,
      "loss": 0.5043,
      "step": 430
    },
    {
      "epoch": 1.623134328358209,
      "grad_norm": 0.19387268984078115,
      "learning_rate": 2.6435015233498443e-05,
      "loss": 0.503,
      "step": 435
    },
    {
      "epoch": 1.6417910447761193,
      "grad_norm": 0.1959979856590692,
      "learning_rate": 2.5972582842351156e-05,
      "loss": 0.5031,
      "step": 440
    },
    {
      "epoch": 1.6604477611940298,
      "grad_norm": 0.2130968015995863,
      "learning_rate": 2.551079779203932e-05,
      "loss": 0.5056,
      "step": 445
    },
    {
      "epoch": 1.6791044776119404,
      "grad_norm": 0.19562804614219742,
      "learning_rate": 2.504985579355047e-05,
      "loss": 0.4971,
      "step": 450
    },
    {
      "epoch": 1.6977611940298507,
      "grad_norm": 0.20132877928156498,
      "learning_rate": 2.458995220057491e-05,
      "loss": 0.5048,
      "step": 455
    },
    {
      "epoch": 1.716417910447761,
      "grad_norm": 0.19111009428414005,
      "learning_rate": 2.4131281926712146e-05,
      "loss": 0.5051,
      "step": 460
    },
    {
      "epoch": 1.7350746268656716,
      "grad_norm": 0.19713565236023642,
      "learning_rate": 2.3674039362863687e-05,
      "loss": 0.5052,
      "step": 465
    },
    {
      "epoch": 1.7537313432835822,
      "grad_norm": 0.2278886953059889,
      "learning_rate": 2.3218418294847517e-05,
      "loss": 0.503,
      "step": 470
    },
    {
      "epoch": 1.7723880597014925,
      "grad_norm": 0.21076902236368567,
      "learning_rate": 2.2764611821268918e-05,
      "loss": 0.5004,
      "step": 475
    },
    {
      "epoch": 1.7910447761194028,
      "grad_norm": 0.2018936557220544,
      "learning_rate": 2.231281227168257e-05,
      "loss": 0.4972,
      "step": 480
    },
    {
      "epoch": 1.8097014925373134,
      "grad_norm": 0.1900377713824044,
      "learning_rate": 2.18632111250806e-05,
      "loss": 0.4987,
      "step": 485
    },
    {
      "epoch": 1.828358208955224,
      "grad_norm": 0.1818444743971112,
      "learning_rate": 2.141599892874107e-05,
      "loss": 0.4988,
      "step": 490
    },
    {
      "epoch": 1.8470149253731343,
      "grad_norm": 0.1843911681060789,
      "learning_rate": 2.09713652174714e-05,
      "loss": 0.5045,
      "step": 495
    },
    {
      "epoch": 1.8656716417910446,
      "grad_norm": 0.1756356146626833,
      "learning_rate": 2.0529498433280807e-05,
      "loss": 0.5062,
      "step": 500
    },
    {
      "epoch": 1.8843283582089554,
      "grad_norm": 0.17919268852808168,
      "learning_rate": 2.0090585845516012e-05,
      "loss": 0.5011,
      "step": 505
    },
    {
      "epoch": 1.9029850746268657,
      "grad_norm": 0.1861488051548517,
      "learning_rate": 1.965481347149376e-05,
      "loss": 0.4964,
      "step": 510
    },
    {
      "epoch": 1.921641791044776,
      "grad_norm": 0.1808039014665401,
      "learning_rate": 1.9222365997664165e-05,
      "loss": 0.5038,
      "step": 515
    },
    {
      "epoch": 1.9402985074626866,
      "grad_norm": 0.18185075285557611,
      "learning_rate": 1.8793426701337947e-05,
      "loss": 0.5033,
      "step": 520
    },
    {
      "epoch": 1.9589552238805972,
      "grad_norm": 0.1802698151327418,
      "learning_rate": 1.8368177373010954e-05,
      "loss": 0.498,
      "step": 525
    },
    {
      "epoch": 1.9776119402985075,
      "grad_norm": 0.1839298648566157,
      "learning_rate": 1.7946798239318775e-05,
      "loss": 0.5033,
      "step": 530
    },
    {
      "epoch": 1.9962686567164178,
      "grad_norm": 0.1787192449198445,
      "learning_rate": 1.75294678866542e-05,
      "loss": 0.4937,
      "step": 535
    },
    {
      "epoch": 2.014925373134328,
      "grad_norm": 0.19908015949978794,
      "learning_rate": 1.7116363185479754e-05,
      "loss": 0.4861,
      "step": 540
    },
    {
      "epoch": 2.033582089552239,
      "grad_norm": 0.17330047242909324,
      "learning_rate": 1.670765921536755e-05,
      "loss": 0.4807,
      "step": 545
    },
    {
      "epoch": 2.0522388059701493,
      "grad_norm": 0.1887761738499672,
      "learning_rate": 1.6303529190798088e-05,
      "loss": 0.4807,
      "step": 550
    },
    {
      "epoch": 2.0708955223880596,
      "grad_norm": 0.18086274895322357,
      "learning_rate": 1.590414438774954e-05,
      "loss": 0.478,
      "step": 555
    },
    {
      "epoch": 2.08955223880597,
      "grad_norm": 0.1666312790289245,
      "learning_rate": 1.550967407110856e-05,
      "loss": 0.4801,
      "step": 560
    },
    {
      "epoch": 2.1082089552238807,
      "grad_norm": 0.16804347810729678,
      "learning_rate": 1.5120285422933478e-05,
      "loss": 0.4814,
      "step": 565
    },
    {
      "epoch": 2.126865671641791,
      "grad_norm": 0.1741311542160614,
      "learning_rate": 1.4736143471600173e-05,
      "loss": 0.4841,
      "step": 570
    },
    {
      "epoch": 2.1455223880597014,
      "grad_norm": 0.16178175453073618,
      "learning_rate": 1.4357411021860773e-05,
      "loss": 0.4768,
      "step": 575
    },
    {
      "epoch": 2.1641791044776117,
      "grad_norm": 0.16621312000057425,
      "learning_rate": 1.3984248585844645e-05,
      "loss": 0.4804,
      "step": 580
    },
    {
      "epoch": 2.1828358208955225,
      "grad_norm": 0.16578503722378288,
      "learning_rate": 1.3616814315031146e-05,
      "loss": 0.4806,
      "step": 585
    },
    {
      "epoch": 2.201492537313433,
      "grad_norm": 0.17421634398822453,
      "learning_rate": 1.3255263933222833e-05,
      "loss": 0.477,
      "step": 590
    },
    {
      "epoch": 2.220149253731343,
      "grad_norm": 0.1584162460344385,
      "learning_rate": 1.2899750670547473e-05,
      "loss": 0.4761,
      "step": 595
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 0.1669406945768043,
      "learning_rate": 1.2550425198516973e-05,
      "loss": 0.4807,
      "step": 600
    },
    {
      "epoch": 2.2574626865671643,
      "grad_norm": 0.1755571617029938,
      "learning_rate": 1.2207435566170722e-05,
      "loss": 0.4836,
      "step": 605
    },
    {
      "epoch": 2.2761194029850746,
      "grad_norm": 0.16049791652334905,
      "learning_rate": 1.1870927137330267e-05,
      "loss": 0.48,
      "step": 610
    },
    {
      "epoch": 2.294776119402985,
      "grad_norm": 0.14841988134789108,
      "learning_rate": 1.1541042528992152e-05,
      "loss": 0.4797,
      "step": 615
    },
    {
      "epoch": 2.3134328358208958,
      "grad_norm": 0.14669420324350554,
      "learning_rate": 1.1217921550884774e-05,
      "loss": 0.4781,
      "step": 620
    },
    {
      "epoch": 2.332089552238806,
      "grad_norm": 0.16275970644355697,
      "learning_rate": 1.0901701146215085e-05,
      "loss": 0.4677,
      "step": 625
    },
    {
      "epoch": 2.3507462686567164,
      "grad_norm": 0.15639224942974775,
      "learning_rate": 1.0592515333630128e-05,
      "loss": 0.4795,
      "step": 630
    },
    {
      "epoch": 2.3694029850746268,
      "grad_norm": 0.17837315275471302,
      "learning_rate": 1.029049515041808e-05,
      "loss": 0.4757,
      "step": 635
    },
    {
      "epoch": 2.388059701492537,
      "grad_norm": 0.15999076690352537,
      "learning_rate": 9.99576859697277e-06,
      "loss": 0.483,
      "step": 640
    },
    {
      "epoch": 2.406716417910448,
      "grad_norm": 0.14623430655091096,
      "learning_rate": 9.708460582545337e-06,
      "loss": 0.479,
      "step": 645
    },
    {
      "epoch": 2.425373134328358,
      "grad_norm": 0.14446705361316536,
      "learning_rate": 9.428692872305925e-06,
      "loss": 0.4782,
      "step": 650
    },
    {
      "epoch": 2.4440298507462686,
      "grad_norm": 0.1499556136205738,
      "learning_rate": 9.15658403573792e-06,
      "loss": 0.4771,
      "step": 655
    },
    {
      "epoch": 2.4626865671641793,
      "grad_norm": 0.13688966492338145,
      "learning_rate": 8.892249396386513e-06,
      "loss": 0.477,
      "step": 660
    },
    {
      "epoch": 2.4813432835820897,
      "grad_norm": 0.1405824099115518,
      "learning_rate": 8.635800982982958e-06,
      "loss": 0.4801,
      "step": 665
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.1556802805035451,
      "learning_rate": 8.387347481965244e-06,
      "loss": 0.4767,
      "step": 670
    },
    {
      "epoch": 2.5186567164179103,
      "grad_norm": 0.14957588217337456,
      "learning_rate": 8.14699419141525e-06,
      "loss": 0.4796,
      "step": 675
    },
    {
      "epoch": 2.5373134328358207,
      "grad_norm": 0.15266928757717518,
      "learning_rate": 7.914842976431932e-06,
      "loss": 0.4752,
      "step": 680
    },
    {
      "epoch": 2.5559701492537314,
      "grad_norm": 0.15337919631830368,
      "learning_rate": 7.690992225959465e-06,
      "loss": 0.4805,
      "step": 685
    },
    {
      "epoch": 2.574626865671642,
      "grad_norm": 0.14478875983801728,
      "learning_rate": 7.4755368110886366e-06,
      "loss": 0.4768,
      "step": 690
    },
    {
      "epoch": 2.593283582089552,
      "grad_norm": 0.15398254303144288,
      "learning_rate": 7.268568044849132e-06,
      "loss": 0.4765,
      "step": 695
    },
    {
      "epoch": 2.611940298507463,
      "grad_norm": 0.15439199576307303,
      "learning_rate": 7.0701736435098155e-06,
      "loss": 0.4782,
      "step": 700
    },
    {
      "epoch": 2.6305970149253732,
      "grad_norm": 0.14978173171132997,
      "learning_rate": 6.880437689403316e-06,
      "loss": 0.4741,
      "step": 705
    },
    {
      "epoch": 2.6492537313432836,
      "grad_norm": 0.1412522116274614,
      "learning_rate": 6.699440595290754e-06,
      "loss": 0.4778,
      "step": 710
    },
    {
      "epoch": 2.667910447761194,
      "grad_norm": 0.15039092893602207,
      "learning_rate": 6.527259070281722e-06,
      "loss": 0.4816,
      "step": 715
    },
    {
      "epoch": 2.6865671641791042,
      "grad_norm": 0.14048749084616627,
      "learning_rate": 6.363966087323844e-06,
      "loss": 0.48,
      "step": 720
    },
    {
      "epoch": 2.705223880597015,
      "grad_norm": 0.14837195306929193,
      "learning_rate": 6.209630852275836e-06,
      "loss": 0.4715,
      "step": 725
    },
    {
      "epoch": 2.7238805970149254,
      "grad_norm": 0.14480369337511617,
      "learning_rate": 6.06431877457709e-06,
      "loss": 0.4752,
      "step": 730
    },
    {
      "epoch": 2.7425373134328357,
      "grad_norm": 0.14464453043475206,
      "learning_rate": 5.928091439526226e-06,
      "loss": 0.4758,
      "step": 735
    },
    {
      "epoch": 2.7611940298507465,
      "grad_norm": 0.14318518863987956,
      "learning_rate": 5.801006582180398e-06,
      "loss": 0.4795,
      "step": 740
    },
    {
      "epoch": 2.779850746268657,
      "grad_norm": 0.1378858672961454,
      "learning_rate": 5.683118062886346e-06,
      "loss": 0.478,
      "step": 745
    },
    {
      "epoch": 2.798507462686567,
      "grad_norm": 0.1473158474351985,
      "learning_rate": 5.574475844453634e-06,
      "loss": 0.477,
      "step": 750
    },
    {
      "epoch": 2.8171641791044775,
      "grad_norm": 0.13519605107478774,
      "learning_rate": 5.475125970979702e-06,
      "loss": 0.4777,
      "step": 755
    },
    {
      "epoch": 2.835820895522388,
      "grad_norm": 0.1469374974433458,
      "learning_rate": 5.385110548335753e-06,
      "loss": 0.4746,
      "step": 760
    },
    {
      "epoch": 2.8544776119402986,
      "grad_norm": 0.13682187728455042,
      "learning_rate": 5.30446772632166e-06,
      "loss": 0.4778,
      "step": 765
    },
    {
      "epoch": 2.873134328358209,
      "grad_norm": 0.1468534460053792,
      "learning_rate": 5.233231682497572e-06,
      "loss": 0.4791,
      "step": 770
    },
    {
      "epoch": 2.8917910447761193,
      "grad_norm": 0.1381366037912041,
      "learning_rate": 5.171432607698975e-06,
      "loss": 0.4737,
      "step": 775
    },
    {
      "epoch": 2.91044776119403,
      "grad_norm": 0.13873604296833256,
      "learning_rate": 5.119096693241395e-06,
      "loss": 0.4785,
      "step": 780
    },
    {
      "epoch": 2.9291044776119404,
      "grad_norm": 0.1354800953615682,
      "learning_rate": 5.07624611982014e-06,
      "loss": 0.4719,
      "step": 785
    },
    {
      "epoch": 2.9477611940298507,
      "grad_norm": 0.1412434132198824,
      "learning_rate": 5.0428990481098275e-06,
      "loss": 0.481,
      "step": 790
    },
    {
      "epoch": 2.966417910447761,
      "grad_norm": 0.14472205562980991,
      "learning_rate": 5.01906961106762e-06,
      "loss": 0.4742,
      "step": 795
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 0.1389514632202056,
      "learning_rate": 5.004767907943488e-06,
      "loss": 0.4824,
      "step": 800
    },
    {
      "epoch": 3.0,
      "step": 804,
      "total_flos": 1465863748190208.0,
      "train_loss": 0.52242068924121,
      "train_runtime": 15658.3478,
      "train_samples_per_second": 6.572,
      "train_steps_per_second": 0.051
    }
  ],
  "logging_steps": 5,
  "max_steps": 804,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1465863748190208.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}