{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.999792917788362,
  "eval_steps": 500,
  "global_step": 1207,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008283288465520812,
      "grad_norm": 24.23784679529398,
      "learning_rate": 8.264462809917357e-08,
      "loss": 1.4304,
      "step": 1
    },
    {
      "epoch": 0.0041416442327604054,
      "grad_norm": 23.234915420644615,
      "learning_rate": 4.132231404958678e-07,
      "loss": 1.4144,
      "step": 5
    },
    {
      "epoch": 0.008283288465520811,
      "grad_norm": 8.222025109804092,
      "learning_rate": 8.264462809917356e-07,
      "loss": 1.307,
      "step": 10
    },
    {
      "epoch": 0.012424932698281217,
      "grad_norm": 10.133789775285283,
      "learning_rate": 1.2396694214876035e-06,
      "loss": 1.1636,
      "step": 15
    },
    {
      "epoch": 0.016566576931041622,
      "grad_norm": 2.9246570997342918,
      "learning_rate": 1.6528925619834712e-06,
      "loss": 1.0151,
      "step": 20
    },
    {
      "epoch": 0.02070822116380203,
      "grad_norm": 2.7388624531591717,
      "learning_rate": 2.066115702479339e-06,
      "loss": 0.9484,
      "step": 25
    },
    {
      "epoch": 0.024849865396562434,
      "grad_norm": 2.365875950285307,
      "learning_rate": 2.479338842975207e-06,
      "loss": 0.9332,
      "step": 30
    },
    {
      "epoch": 0.028991509629322842,
      "grad_norm": 2.374922722027774,
      "learning_rate": 2.8925619834710743e-06,
      "loss": 0.9044,
      "step": 35
    },
    {
      "epoch": 0.033133153862083244,
      "grad_norm": 2.238010214125612,
      "learning_rate": 3.3057851239669424e-06,
      "loss": 0.8913,
      "step": 40
    },
    {
      "epoch": 0.03727479809484365,
      "grad_norm": 2.4490973058865078,
      "learning_rate": 3.71900826446281e-06,
      "loss": 0.8833,
      "step": 45
    },
    {
      "epoch": 0.04141644232760406,
      "grad_norm": 2.469516335084805,
      "learning_rate": 4.132231404958678e-06,
      "loss": 0.8813,
      "step": 50
    },
    {
      "epoch": 0.04555808656036447,
      "grad_norm": 2.2371905595594495,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.8533,
      "step": 55
    },
    {
      "epoch": 0.04969973079312487,
      "grad_norm": 2.409737572391393,
      "learning_rate": 4.958677685950414e-06,
      "loss": 0.8581,
      "step": 60
    },
    {
      "epoch": 0.05384137502588528,
      "grad_norm": 2.2291530452515977,
      "learning_rate": 5.371900826446281e-06,
      "loss": 0.8519,
      "step": 65
    },
    {
      "epoch": 0.057983019258645685,
      "grad_norm": 2.35453266908653,
      "learning_rate": 5.785123966942149e-06,
      "loss": 0.8515,
      "step": 70
    },
    {
      "epoch": 0.062124663491406086,
      "grad_norm": 2.336463988259032,
      "learning_rate": 6.198347107438017e-06,
      "loss": 0.8295,
      "step": 75
    },
    {
      "epoch": 0.06626630772416649,
      "grad_norm": 2.4980329883385624,
      "learning_rate": 6.611570247933885e-06,
      "loss": 0.832,
      "step": 80
    },
    {
      "epoch": 0.0704079519569269,
      "grad_norm": 2.6275854974083384,
      "learning_rate": 7.0247933884297525e-06,
      "loss": 0.816,
      "step": 85
    },
    {
      "epoch": 0.0745495961896873,
      "grad_norm": 2.684536070785031,
      "learning_rate": 7.43801652892562e-06,
      "loss": 0.8162,
      "step": 90
    },
    {
      "epoch": 0.07869124042244771,
      "grad_norm": 2.805145187507444,
      "learning_rate": 7.851239669421489e-06,
      "loss": 0.8309,
      "step": 95
    },
    {
      "epoch": 0.08283288465520812,
      "grad_norm": 2.4685576686167874,
      "learning_rate": 8.264462809917356e-06,
      "loss": 0.8155,
      "step": 100
    },
    {
      "epoch": 0.08697452888796853,
      "grad_norm": 2.540099419306748,
      "learning_rate": 8.677685950413224e-06,
      "loss": 0.8071,
      "step": 105
    },
    {
      "epoch": 0.09111617312072894,
      "grad_norm": 2.562872480103855,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.8029,
      "step": 110
    },
    {
      "epoch": 0.09525781735348933,
      "grad_norm": 2.5970244377170313,
      "learning_rate": 9.50413223140496e-06,
      "loss": 0.7873,
      "step": 115
    },
    {
      "epoch": 0.09939946158624974,
      "grad_norm": 2.4308464508943346,
      "learning_rate": 9.917355371900828e-06,
      "loss": 0.7915,
      "step": 120
    },
    {
      "epoch": 0.10354110581901015,
      "grad_norm": 2.6905207268547158,
      "learning_rate": 9.999665269535307e-06,
      "loss": 0.7993,
      "step": 125
    },
    {
      "epoch": 0.10768275005177055,
      "grad_norm": 34.133455991097065,
      "learning_rate": 9.998305503833872e-06,
      "loss": 0.8009,
      "step": 130
    },
    {
      "epoch": 0.11182439428453096,
      "grad_norm": 2.3166119518346866,
      "learning_rate": 9.995900066492902e-06,
      "loss": 0.785,
      "step": 135
    },
    {
      "epoch": 0.11596603851729137,
      "grad_norm": 2.5556998820575076,
      "learning_rate": 9.992449460742464e-06,
      "loss": 0.8134,
      "step": 140
    },
    {
      "epoch": 0.12010768275005176,
      "grad_norm": 2.615095977165256,
      "learning_rate": 9.98795440846732e-06,
      "loss": 0.7898,
      "step": 145
    },
    {
      "epoch": 0.12424932698281217,
      "grad_norm": 2.729536225853297,
      "learning_rate": 9.982415850055902e-06,
      "loss": 0.7681,
      "step": 150
    },
    {
      "epoch": 0.12839097121557258,
      "grad_norm": 3.895339838479258,
      "learning_rate": 9.975834944203581e-06,
      "loss": 0.7503,
      "step": 155
    },
    {
      "epoch": 0.13253261544833297,
      "grad_norm": 2.4875917416718187,
      "learning_rate": 9.968213067670265e-06,
      "loss": 0.7632,
      "step": 160
    },
    {
      "epoch": 0.1366742596810934,
      "grad_norm": 2.5249141397756847,
      "learning_rate": 9.959551814992364e-06,
      "loss": 0.7444,
      "step": 165
    },
    {
      "epoch": 0.1408159039138538,
      "grad_norm": 2.6613194516011656,
      "learning_rate": 9.949852998149217e-06,
      "loss": 0.7592,
      "step": 170
    },
    {
      "epoch": 0.1449575481466142,
      "grad_norm": 2.535010612771426,
      "learning_rate": 9.939118646184007e-06,
      "loss": 0.7625,
      "step": 175
    },
    {
      "epoch": 0.1490991923793746,
      "grad_norm": 2.658211254916015,
      "learning_rate": 9.927351004779275e-06,
      "loss": 0.7405,
      "step": 180
    },
    {
      "epoch": 0.15324083661213503,
      "grad_norm": 2.343186435782198,
      "learning_rate": 9.914552535787122e-06,
      "loss": 0.7285,
      "step": 185
    },
    {
      "epoch": 0.15738248084489542,
      "grad_norm": 2.4661024566837653,
      "learning_rate": 9.900725916714157e-06,
      "loss": 0.7405,
      "step": 190
    },
    {
      "epoch": 0.16152412507765582,
      "grad_norm": 2.7732706905434497,
      "learning_rate": 9.885874040161373e-06,
      "loss": 0.7373,
      "step": 195
    },
    {
      "epoch": 0.16566576931041624,
      "grad_norm": 2.523332136572105,
      "learning_rate": 9.87000001321898e-06,
      "loss": 0.7261,
      "step": 200
    },
    {
      "epoch": 0.16980741354317663,
      "grad_norm": 2.2311284634557937,
      "learning_rate": 9.853107156816393e-06,
      "loss": 0.7184,
      "step": 205
    },
    {
      "epoch": 0.17394905777593705,
      "grad_norm": 2.379531410344561,
      "learning_rate": 9.835199005027477e-06,
      "loss": 0.7144,
      "step": 210
    },
    {
      "epoch": 0.17809070200869745,
      "grad_norm": 2.3058139982021215,
      "learning_rate": 9.816279304331202e-06,
      "loss": 0.7252,
      "step": 215
    },
    {
      "epoch": 0.18223234624145787,
      "grad_norm": 2.2361882677362974,
      "learning_rate": 9.79635201282785e-06,
      "loss": 0.7151,
      "step": 220
    },
    {
      "epoch": 0.18637399047421827,
      "grad_norm": 2.7702057401479006,
      "learning_rate": 9.775421299410977e-06,
      "loss": 0.7045,
      "step": 225
    },
    {
      "epoch": 0.19051563470697866,
      "grad_norm": 2.5524695305725404,
      "learning_rate": 9.753491542895237e-06,
      "loss": 0.715,
      "step": 230
    },
    {
      "epoch": 0.19465727893973908,
      "grad_norm": 2.4487445215481305,
      "learning_rate": 9.730567331100333e-06,
      "loss": 0.7082,
      "step": 235
    },
    {
      "epoch": 0.19879892317249948,
      "grad_norm": 2.2482641063346116,
      "learning_rate": 9.706653459891207e-06,
      "loss": 0.699,
      "step": 240
    },
    {
      "epoch": 0.2029405674052599,
      "grad_norm": 2.451295868720965,
      "learning_rate": 9.681754932174719e-06,
      "loss": 0.6913,
      "step": 245
    },
    {
      "epoch": 0.2070822116380203,
      "grad_norm": 2.56626439770444,
      "learning_rate": 9.655876956853025e-06,
      "loss": 0.6819,
      "step": 250
    },
    {
      "epoch": 0.2112238558707807,
      "grad_norm": 2.490881018154125,
      "learning_rate": 9.629024947733836e-06,
      "loss": 0.6843,
      "step": 255
    },
    {
      "epoch": 0.2153655001035411,
      "grad_norm": 2.4655094247796283,
      "learning_rate": 9.601204522397826e-06,
      "loss": 0.6874,
      "step": 260
    },
    {
      "epoch": 0.2195071443363015,
      "grad_norm": 2.6029112443690625,
      "learning_rate": 9.572421501023403e-06,
      "loss": 0.6773,
      "step": 265
    },
    {
      "epoch": 0.22364878856906192,
      "grad_norm": 2.303292362769761,
      "learning_rate": 9.5426819051691e-06,
      "loss": 0.6669,
      "step": 270
    },
    {
      "epoch": 0.22779043280182232,
      "grad_norm": 2.2947735744081568,
      "learning_rate": 9.511991956513828e-06,
      "loss": 0.6636,
      "step": 275
    },
    {
      "epoch": 0.23193207703458274,
      "grad_norm": 2.3561731246407045,
      "learning_rate": 9.480358075555278e-06,
      "loss": 0.6671,
      "step": 280
    },
    {
      "epoch": 0.23607372126734313,
      "grad_norm": 2.4666706522590442,
      "learning_rate": 9.447786880266706e-06,
      "loss": 0.6618,
      "step": 285
    },
    {
      "epoch": 0.24021536550010353,
      "grad_norm": 2.186824941231579,
      "learning_rate": 9.414285184712432e-06,
      "loss": 0.6619,
      "step": 290
    },
    {
      "epoch": 0.24435700973286395,
      "grad_norm": 2.1988625035941047,
      "learning_rate": 9.37985999762229e-06,
      "loss": 0.6473,
      "step": 295
    },
    {
      "epoch": 0.24849865396562434,
      "grad_norm": 2.43262245251786,
      "learning_rate": 9.344518520925377e-06,
      "loss": 0.6534,
      "step": 300
    },
    {
      "epoch": 0.25264029819838474,
      "grad_norm": 2.3140467166469527,
      "learning_rate": 9.308268148243355e-06,
      "loss": 0.637,
      "step": 305
    },
    {
      "epoch": 0.25678194243114516,
      "grad_norm": 2.5057247056088996,
      "learning_rate": 9.271116463343692e-06,
      "loss": 0.6417,
      "step": 310
    },
    {
      "epoch": 0.2609235866639056,
      "grad_norm": 2.288973054378257,
      "learning_rate": 9.23307123855307e-06,
      "loss": 0.663,
      "step": 315
    },
    {
      "epoch": 0.26506523089666595,
      "grad_norm": 2.416778922999839,
      "learning_rate": 9.194140433131397e-06,
      "loss": 0.6552,
      "step": 320
    },
    {
      "epoch": 0.26920687512942637,
      "grad_norm": 2.2651292744956244,
      "learning_rate": 9.154332191606671e-06,
      "loss": 0.6267,
      "step": 325
    },
    {
      "epoch": 0.2733485193621868,
      "grad_norm": 2.3801866420145954,
      "learning_rate": 9.113654842071114e-06,
      "loss": 0.6306,
      "step": 330
    },
    {
      "epoch": 0.2774901635949472,
      "grad_norm": 2.5824252886751964,
      "learning_rate": 9.072116894438885e-06,
      "loss": 0.6369,
      "step": 335
    },
    {
      "epoch": 0.2816318078277076,
      "grad_norm": 2.3287880261931875,
      "learning_rate": 9.029727038665765e-06,
      "loss": 0.6252,
      "step": 340
    },
    {
      "epoch": 0.285773452060468,
      "grad_norm": 2.3253013276198358,
      "learning_rate": 8.986494142931168e-06,
      "loss": 0.6165,
      "step": 345
    },
    {
      "epoch": 0.2899150962932284,
      "grad_norm": 2.3141576363210086,
      "learning_rate": 8.94242725178288e-06,
      "loss": 0.6003,
      "step": 350
    },
    {
      "epoch": 0.2940567405259888,
      "grad_norm": 2.2453531818296595,
      "learning_rate": 8.89753558424488e-06,
      "loss": 0.6,
      "step": 355
    },
    {
      "epoch": 0.2981983847587492,
      "grad_norm": 2.224520930970933,
      "learning_rate": 8.851828531888692e-06,
      "loss": 0.6117,
      "step": 360
    },
    {
      "epoch": 0.30234002899150964,
      "grad_norm": 2.312078602771557,
      "learning_rate": 8.805315656868587e-06,
      "loss": 0.6067,
      "step": 365
    },
    {
      "epoch": 0.30648167322427006,
      "grad_norm": 2.287152946702887,
      "learning_rate": 8.75800668992117e-06,
      "loss": 0.5979,
      "step": 370
    },
    {
      "epoch": 0.3106233174570304,
      "grad_norm": 2.0335651194660596,
      "learning_rate": 8.709911528329623e-06,
      "loss": 0.5911,
      "step": 375
    },
    {
      "epoch": 0.31476496168979085,
      "grad_norm": 2.190338892871162,
      "learning_rate": 8.661040233853166e-06,
      "loss": 0.5884,
      "step": 380
    },
    {
      "epoch": 0.31890660592255127,
      "grad_norm": 2.198737460879657,
      "learning_rate": 8.611403030622074e-06,
      "loss": 0.578,
      "step": 385
    },
    {
      "epoch": 0.32304825015531163,
      "grad_norm": 2.320069378855735,
      "learning_rate": 8.561010302998734e-06,
      "loss": 0.5809,
      "step": 390
    },
    {
      "epoch": 0.32718989438807206,
      "grad_norm": 2.1718618035260793,
      "learning_rate": 8.509872593405189e-06,
      "loss": 0.5834,
      "step": 395
    },
    {
      "epoch": 0.3313315386208325,
      "grad_norm": 2.306249404711425,
      "learning_rate": 8.458000600117604e-06,
      "loss": 0.5817,
      "step": 400
    },
    {
      "epoch": 0.3354731828535929,
      "grad_norm": 2.2320669776692377,
      "learning_rate": 8.40540517502813e-06,
      "loss": 0.579,
      "step": 405
    },
    {
      "epoch": 0.33961482708635327,
      "grad_norm": 2.380541232391175,
      "learning_rate": 8.35209732137463e-06,
      "loss": 0.5657,
      "step": 410
    },
    {
      "epoch": 0.3437564713191137,
      "grad_norm": 2.270445952132619,
      "learning_rate": 8.298088191438753e-06,
      "loss": 0.5569,
      "step": 415
    },
    {
      "epoch": 0.3478981155518741,
      "grad_norm": 2.353449916475927,
      "learning_rate": 8.243389084212808e-06,
      "loss": 0.5642,
      "step": 420
    },
    {
      "epoch": 0.3520397597846345,
      "grad_norm": 2.321086875711968,
      "learning_rate": 8.188011443035962e-06,
      "loss": 0.5519,
      "step": 425
    },
    {
      "epoch": 0.3561814040173949,
      "grad_norm": 2.112886796528374,
      "learning_rate": 8.131966853200226e-06,
      "loss": 0.561,
      "step": 430
    },
    {
      "epoch": 0.3603230482501553,
      "grad_norm": 2.354896774745211,
      "learning_rate": 8.075267039526764e-06,
      "loss": 0.5586,
      "step": 435
    },
    {
      "epoch": 0.36446469248291574,
      "grad_norm": 2.291143610369071,
      "learning_rate": 8.017923863912989e-06,
      "loss": 0.5491,
      "step": 440
    },
    {
      "epoch": 0.3686063367156761,
      "grad_norm": 2.2300308493171115,
      "learning_rate": 7.959949322850994e-06,
      "loss": 0.5562,
      "step": 445
    },
    {
      "epoch": 0.37274798094843653,
      "grad_norm": 2.28359859552775,
      "learning_rate": 7.901355544917827e-06,
      "loss": 0.5494,
      "step": 450
    },
    {
      "epoch": 0.37688962518119695,
      "grad_norm": 2.251157158354555,
      "learning_rate": 7.842154788238124e-06,
      "loss": 0.5424,
      "step": 455
    },
    {
      "epoch": 0.3810312694139573,
      "grad_norm": 2.285389132836112,
      "learning_rate": 7.782359437919644e-06,
      "loss": 0.5423,
      "step": 460
    },
    {
      "epoch": 0.38517291364671774,
      "grad_norm": 2.578954308129232,
      "learning_rate": 7.721982003462255e-06,
      "loss": 0.5335,
      "step": 465
    },
    {
      "epoch": 0.38931455787947816,
      "grad_norm": 2.1334475270465774,
      "learning_rate": 7.661035116140856e-06,
      "loss": 0.5342,
      "step": 470
    },
    {
      "epoch": 0.3934562021122386,
      "grad_norm": 2.3233034598991797,
      "learning_rate": 7.599531526362873e-06,
      "loss": 0.5358,
      "step": 475
    },
    {
      "epoch": 0.39759784634499895,
      "grad_norm": 2.2758511904313785,
      "learning_rate": 7.537484101000787e-06,
      "loss": 0.532,
      "step": 480
    },
    {
      "epoch": 0.4017394905777594,
      "grad_norm": 2.151297761784936,
      "learning_rate": 7.474905820700334e-06,
      "loss": 0.5135,
      "step": 485
    },
    {
      "epoch": 0.4058811348105198,
      "grad_norm": 2.3747508627507066,
      "learning_rate": 7.411809777164873e-06,
      "loss": 0.5306,
      "step": 490
    },
    {
      "epoch": 0.41002277904328016,
      "grad_norm": 2.202301523886571,
      "learning_rate": 7.3482091704165405e-06,
      "loss": 0.5247,
      "step": 495
    },
    {
      "epoch": 0.4141644232760406,
      "grad_norm": 2.153444078468182,
      "learning_rate": 7.284117306034733e-06,
      "loss": 0.5243,
      "step": 500
    },
    {
      "epoch": 0.418306067508801,
      "grad_norm": 2.171064236273241,
      "learning_rate": 7.219547592372512e-06,
      "loss": 0.5187,
      "step": 505
    },
    {
      "epoch": 0.4224477117415614,
      "grad_norm": 2.2667901238143586,
      "learning_rate": 7.15451353775151e-06,
      "loss": 0.5126,
      "step": 510
    },
    {
      "epoch": 0.4265893559743218,
      "grad_norm": 2.4217664685492997,
      "learning_rate": 7.089028747635908e-06,
      "loss": 0.5166,
      "step": 515
    },
    {
      "epoch": 0.4307310002070822,
      "grad_norm": 2.269198948091921,
      "learning_rate": 7.023106921786118e-06,
      "loss": 0.5102,
      "step": 520
    },
    {
      "epoch": 0.43487264443984264,
      "grad_norm": 2.3389291631103717,
      "learning_rate": 6.956761851392706e-06,
      "loss": 0.5147,
      "step": 525
    },
    {
      "epoch": 0.439014288672603,
      "grad_norm": 2.130307708383959,
      "learning_rate": 6.890007416191209e-06,
      "loss": 0.5,
      "step": 530
    },
    {
      "epoch": 0.4431559329053634,
      "grad_norm": 2.2319755819000133,
      "learning_rate": 6.822857581558423e-06,
      "loss": 0.5031,
      "step": 535
    },
    {
      "epoch": 0.44729757713812385,
      "grad_norm": 2.201034276249066,
      "learning_rate": 6.7553263955907755e-06,
      "loss": 0.5003,
      "step": 540
    },
    {
      "epoch": 0.4514392213708842,
      "grad_norm": 2.1906710526713877,
      "learning_rate": 6.687427986165379e-06,
      "loss": 0.498,
      "step": 545
    },
    {
      "epoch": 0.45558086560364464,
      "grad_norm": 2.421545219253794,
      "learning_rate": 6.6191765579844205e-06,
      "loss": 0.4996,
      "step": 550
    },
    {
      "epoch": 0.45972250983640506,
      "grad_norm": 2.2909926333329977,
      "learning_rate": 6.550586389603451e-06,
      "loss": 0.4969,
      "step": 555
    },
    {
      "epoch": 0.4638641540691655,
      "grad_norm": 2.3444032170833116,
      "learning_rate": 6.481671830444243e-06,
      "loss": 0.4945,
      "step": 560
    },
    {
      "epoch": 0.46800579830192585,
      "grad_norm": 2.1725157981177916,
      "learning_rate": 6.412447297792818e-06,
      "loss": 0.4863,
      "step": 565
    },
    {
      "epoch": 0.47214744253468627,
      "grad_norm": 2.1573058335782753,
      "learning_rate": 6.3429272737832726e-06,
      "loss": 0.4891,
      "step": 570
    },
    {
      "epoch": 0.4762890867674467,
      "grad_norm": 2.2084194407586373,
      "learning_rate": 6.273126302368037e-06,
      "loss": 0.487,
      "step": 575
    },
    {
      "epoch": 0.48043073100020706,
      "grad_norm": 2.21344464284995,
      "learning_rate": 6.203058986275207e-06,
      "loss": 0.4857,
      "step": 580
    },
    {
      "epoch": 0.4845723752329675,
      "grad_norm": 2.2203608524225666,
      "learning_rate": 6.132739983953579e-06,
      "loss": 0.4828,
      "step": 585
    },
    {
      "epoch": 0.4887140194657279,
      "grad_norm": 2.113371175419138,
      "learning_rate": 6.062184006506027e-06,
      "loss": 0.4826,
      "step": 590
    },
    {
      "epoch": 0.4928556636984883,
      "grad_norm": 2.3454680885393464,
      "learning_rate": 5.991405814611855e-06,
      "loss": 0.4676,
      "step": 595
    },
    {
      "epoch": 0.4969973079312487,
      "grad_norm": 2.321964160860843,
      "learning_rate": 5.920420215438794e-06,
      "loss": 0.4737,
      "step": 600
    },
    {
      "epoch": 0.5011389521640092,
      "grad_norm": 2.1991634408721943,
      "learning_rate": 5.849242059545259e-06,
      "loss": 0.465,
      "step": 605
    },
    {
      "epoch": 0.5052805963967695,
      "grad_norm": 2.165800122945675,
      "learning_rate": 5.777886237773542e-06,
      "loss": 0.4623,
      "step": 610
    },
    {
      "epoch": 0.5094222406295299,
      "grad_norm": 2.1403595970614133,
      "learning_rate": 5.706367678134562e-06,
      "loss": 0.4767,
      "step": 615
    },
    {
      "epoch": 0.5135638848622903,
      "grad_norm": 2.1427662247598986,
      "learning_rate": 5.634701342684852e-06,
      "loss": 0.4607,
      "step": 620
    },
    {
      "epoch": 0.5177055290950507,
      "grad_norm": 2.0626908104196673,
      "learning_rate": 5.562902224396416e-06,
      "loss": 0.4617,
      "step": 625
    },
    {
      "epoch": 0.5218471733278112,
      "grad_norm": 2.0098208614700326,
      "learning_rate": 5.49098534402012e-06,
      "loss": 0.4618,
      "step": 630
    },
    {
      "epoch": 0.5259888175605716,
      "grad_norm": 2.0735743487283607,
      "learning_rate": 5.418965746943281e-06,
      "loss": 0.459,
      "step": 635
    },
    {
      "epoch": 0.5301304617933319,
      "grad_norm": 2.024087007627894,
      "learning_rate": 5.34685850004208e-06,
      "loss": 0.4539,
      "step": 640
    },
    {
      "epoch": 0.5342721060260923,
      "grad_norm": 2.130234594696483,
      "learning_rate": 5.2746786885295034e-06,
      "loss": 0.453,
      "step": 645
    },
    {
      "epoch": 0.5384137502588527,
      "grad_norm": 2.0986269383387706,
      "learning_rate": 5.2024414127994325e-06,
      "loss": 0.4538,
      "step": 650
    },
    {
      "epoch": 0.5425553944916132,
      "grad_norm": 2.209414073449144,
      "learning_rate": 5.13016178526756e-06,
      "loss": 0.4448,
      "step": 655
    },
    {
      "epoch": 0.5466970387243736,
      "grad_norm": 2.15118663280229,
      "learning_rate": 5.057854927209804e-06,
      "loss": 0.4446,
      "step": 660
    },
    {
      "epoch": 0.550838682957134,
      "grad_norm": 2.033860154274491,
      "learning_rate": 4.985535965598843e-06,
      "loss": 0.4447,
      "step": 665
    },
    {
      "epoch": 0.5549803271898944,
      "grad_norm": 2.129135189760016,
      "learning_rate": 4.913220029939491e-06,
      "loss": 0.4512,
      "step": 670
    },
    {
      "epoch": 0.5591219714226547,
      "grad_norm": 2.090833400754654,
      "learning_rate": 4.840922249103506e-06,
      "loss": 0.4467,
      "step": 675
    },
    {
      "epoch": 0.5632636156554152,
      "grad_norm": 2.0680502896348836,
      "learning_rate": 4.7686577481645745e-06,
      "loss": 0.4316,
      "step": 680
    },
    {
      "epoch": 0.5674052598881756,
      "grad_norm": 2.07449578249477,
      "learning_rate": 4.696441645234042e-06,
      "loss": 0.4421,
      "step": 685
    },
    {
      "epoch": 0.571546904120936,
      "grad_norm": 2.2932786860060235,
      "learning_rate": 4.624289048298147e-06,
      "loss": 0.4433,
      "step": 690
    },
    {
      "epoch": 0.5756885483536964,
      "grad_norm": 2.058591317854592,
      "learning_rate": 4.55221505205734e-06,
      "loss": 0.4298,
      "step": 695
    },
    {
      "epoch": 0.5798301925864568,
      "grad_norm": 2.3196625048015704,
      "learning_rate": 4.480234734768393e-06,
      "loss": 0.4326,
      "step": 700
    },
    {
      "epoch": 0.5839718368192173,
      "grad_norm": 2.073473068971661,
      "learning_rate": 4.408363155089952e-06,
      "loss": 0.4335,
      "step": 705
    },
    {
      "epoch": 0.5881134810519776,
      "grad_norm": 2.029766952429309,
      "learning_rate": 4.3366153489321855e-06,
      "loss": 0.4273,
      "step": 710
    },
    {
      "epoch": 0.592255125284738,
      "grad_norm": 2.1078664280012007,
      "learning_rate": 4.265006326311199e-06,
      "loss": 0.415,
      "step": 715
    },
    {
      "epoch": 0.5963967695174984,
      "grad_norm": 2.0038145759132915,
      "learning_rate": 4.1935510682088545e-06,
      "loss": 0.4244,
      "step": 720
    },
    {
      "epoch": 0.6005384137502588,
      "grad_norm": 2.0468827081169976,
      "learning_rate": 4.122264523438668e-06,
      "loss": 0.4226,
      "step": 725
    },
    {
      "epoch": 0.6046800579830193,
      "grad_norm": 2.0660110564277923,
      "learning_rate": 4.051161605518453e-06,
      "loss": 0.4222,
      "step": 730
    },
    {
      "epoch": 0.6088217022157797,
      "grad_norm": 2.088150514221739,
      "learning_rate": 3.980257189550316e-06,
      "loss": 0.433,
      "step": 735
    },
    {
      "epoch": 0.6129633464485401,
      "grad_norm": 2.1225320313052447,
      "learning_rate": 3.909566109108727e-06,
      "loss": 0.4161,
      "step": 740
    },
    {
      "epoch": 0.6171049906813004,
      "grad_norm": 2.0990451349998227,
      "learning_rate": 3.839103153137247e-06,
      "loss": 0.417,
      "step": 745
    },
    {
      "epoch": 0.6212466349140608,
      "grad_norm": 2.104242942119707,
      "learning_rate": 3.768883062854598e-06,
      "loss": 0.4081,
      "step": 750
    },
    {
      "epoch": 0.6253882791468213,
      "grad_norm": 2.051137757693821,
      "learning_rate": 3.6989205286707398e-06,
      "loss": 0.4108,
      "step": 755
    },
    {
      "epoch": 0.6295299233795817,
      "grad_norm": 2.099437836109555,
      "learning_rate": 3.6292301871135425e-06,
      "loss": 0.411,
      "step": 760
    },
    {
      "epoch": 0.6336715676123421,
      "grad_norm": 2.056296390597689,
      "learning_rate": 3.55982661776676e-06,
      "loss": 0.4047,
      "step": 765
    },
    {
      "epoch": 0.6378132118451025,
      "grad_norm": 2.0454981613636503,
      "learning_rate": 3.4907243402199013e-06,
      "loss": 0.4044,
      "step": 770
    },
    {
      "epoch": 0.641954856077863,
      "grad_norm": 2.0625058430327092,
      "learning_rate": 3.4219378110306523e-06,
      "loss": 0.4103,
      "step": 775
    },
    {
      "epoch": 0.6460965003106233,
      "grad_norm": 2.0640184519207123,
      "learning_rate": 3.353481420700495e-06,
      "loss": 0.4109,
      "step": 780
    },
    {
      "epoch": 0.6502381445433837,
      "grad_norm": 2.096435391565135,
      "learning_rate": 3.285369490664133e-06,
      "loss": 0.4103,
      "step": 785
    },
    {
      "epoch": 0.6543797887761441,
      "grad_norm": 2.0511959776297983,
      "learning_rate": 3.2176162702933816e-06,
      "loss": 0.3991,
      "step": 790
    },
    {
      "epoch": 0.6585214330089045,
      "grad_norm": 2.004915794597741,
      "learning_rate": 3.150235933916115e-06,
      "loss": 0.401,
      "step": 795
    },
    {
      "epoch": 0.662663077241665,
      "grad_norm": 1.989931694202458,
      "learning_rate": 3.0832425778509235e-06,
      "loss": 0.4015,
      "step": 800
    },
    {
      "epoch": 0.6668047214744254,
      "grad_norm": 2.0927029677784383,
      "learning_rate": 3.0166502174581012e-06,
      "loss": 0.3904,
      "step": 805
    },
    {
      "epoch": 0.6709463657071858,
      "grad_norm": 2.05569943435716,
      "learning_rate": 2.950472784207544e-06,
      "loss": 0.3976,
      "step": 810
    },
    {
      "epoch": 0.6750880099399461,
      "grad_norm": 2.0070990032518456,
      "learning_rate": 2.8847241227642255e-06,
      "loss": 0.3855,
      "step": 815
    },
    {
      "epoch": 0.6792296541727065,
      "grad_norm": 1.9791880977464777,
      "learning_rate": 2.819417988091814e-06,
      "loss": 0.3831,
      "step": 820
    },
    {
      "epoch": 0.683371298405467,
      "grad_norm": 1.9874217780709027,
      "learning_rate": 2.754568042575061e-06,
      "loss": 0.3928,
      "step": 825
    },
    {
      "epoch": 0.6875129426382274,
      "grad_norm": 1.9444889966862584,
      "learning_rate": 2.6901878531615677e-06,
      "loss": 0.3967,
      "step": 830
    },
    {
      "epoch": 0.6916545868709878,
      "grad_norm": 2.190695001941064,
      "learning_rate": 2.6262908885235046e-06,
      "loss": 0.384,
      "step": 835
    },
    {
      "epoch": 0.6957962311037482,
      "grad_norm": 1.991552664612379,
      "learning_rate": 2.5628905162398797e-06,
      "loss": 0.3831,
      "step": 840
    },
    {
      "epoch": 0.6999378753365086,
      "grad_norm": 2.017973650880143,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.3851,
      "step": 845
    },
    {
      "epoch": 0.704079519569269,
      "grad_norm": 2.138390804530181,
      "learning_rate": 2.4376324968286154e-06,
      "loss": 0.3777,
      "step": 850
    },
    {
      "epoch": 0.7082211638020294,
      "grad_norm": 2.0021049336250814,
      "learning_rate": 2.375801054333409e-06,
      "loss": 0.3891,
      "step": 855
    },
    {
      "epoch": 0.7123628080347898,
      "grad_norm": 2.1027439407928505,
      "learning_rate": 2.3145186079753685e-06,
      "loss": 0.381,
      "step": 860
    },
    {
      "epoch": 0.7165044522675502,
      "grad_norm": 2.0200252919367823,
      "learning_rate": 2.253797978362617e-06,
      "loss": 0.3754,
      "step": 865
    },
    {
      "epoch": 0.7206460965003106,
      "grad_norm": 2.022944794755911,
      "learning_rate": 2.193651868568285e-06,
      "loss": 0.3719,
      "step": 870
    },
    {
      "epoch": 0.7247877407330711,
      "grad_norm": 2.011383665562108,
      "learning_rate": 2.1340928614729445e-06,
      "loss": 0.3716,
      "step": 875
    },
    {
      "epoch": 0.7289293849658315,
      "grad_norm": 2.1000401745759767,
      "learning_rate": 2.075133417132223e-06,
      "loss": 0.3773,
      "step": 880
    },
    {
      "epoch": 0.7330710291985918,
      "grad_norm": 2.0322415551222277,
      "learning_rate": 2.016785870170079e-06,
      "loss": 0.3755,
      "step": 885
    },
    {
      "epoch": 0.7372126734313522,
      "grad_norm": 2.0586335139183327,
      "learning_rate": 1.9590624271983406e-06,
      "loss": 0.3749,
      "step": 890
    },
    {
      "epoch": 0.7413543176641126,
      "grad_norm": 2.057098686991852,
      "learning_rate": 1.9019751642630252e-06,
      "loss": 0.3733,
      "step": 895
    },
    {
      "epoch": 0.7454959618968731,
      "grad_norm": 1.9856940958346814,
      "learning_rate": 1.8455360243179537e-06,
      "loss": 0.3737,
      "step": 900
    },
    {
      "epoch": 0.7496376061296335,
      "grad_norm": 2.049797779671496,
      "learning_rate": 1.7897568147262323e-06,
      "loss": 0.3678,
      "step": 905
    },
    {
      "epoch": 0.7537792503623939,
      "grad_norm": 2.1869162271482083,
      "learning_rate": 1.7346492047900897e-06,
      "loss": 0.3769,
      "step": 910
    },
    {
      "epoch": 0.7579208945951543,
      "grad_norm": 2.0166177063427444,
      "learning_rate": 1.6802247233095914e-06,
      "loss": 0.3722,
      "step": 915
    },
    {
      "epoch": 0.7620625388279146,
      "grad_norm": 2.0284630775550094,
      "learning_rate": 1.626494756170765e-06,
      "loss": 0.3562,
      "step": 920
    },
    {
      "epoch": 0.7662041830606751,
      "grad_norm": 1.9844703515401159,
      "learning_rate": 1.5734705439636017e-06,
      "loss": 0.3641,
      "step": 925
    },
    {
      "epoch": 0.7703458272934355,
      "grad_norm": 2.026880974919187,
      "learning_rate": 1.5211631796304721e-06,
      "loss": 0.3671,
      "step": 930
    },
    {
      "epoch": 0.7744874715261959,
      "grad_norm": 2.2261033018640775,
      "learning_rate": 1.46958360614543e-06,
      "loss": 0.3677,
      "step": 935
    },
    {
      "epoch": 0.7786291157589563,
      "grad_norm": 2.0419572811826527,
      "learning_rate": 1.4187426142248723e-06,
      "loss": 0.3567,
      "step": 940
    },
    {
      "epoch": 0.7827707599917167,
      "grad_norm": 1.9436995231419443,
      "learning_rate": 1.3686508400700787e-06,
      "loss": 0.3659,
      "step": 945
    },
    {
      "epoch": 0.7869124042244772,
      "grad_norm": 2.118970998919544,
      "learning_rate": 1.3193187631420462e-06,
      "loss": 0.3621,
      "step": 950
    },
    {
      "epoch": 0.7910540484572375,
      "grad_norm": 2.059747456229496,
      "learning_rate": 1.2707567039691505e-06,
      "loss": 0.3565,
      "step": 955
    },
    {
      "epoch": 0.7951956926899979,
      "grad_norm": 2.090689117796637,
      "learning_rate": 1.222974821988024e-06,
      "loss": 0.3583,
      "step": 960
    },
    {
      "epoch": 0.7993373369227583,
      "grad_norm": 2.059076495278081,
      "learning_rate": 1.1759831134181504e-06,
      "loss": 0.3622,
      "step": 965
    },
    {
      "epoch": 0.8034789811555187,
      "grad_norm": 1.9884532847703864,
      "learning_rate": 1.1297914091706086e-06,
      "loss": 0.3541,
      "step": 970
    },
    {
      "epoch": 0.8076206253882792,
      "grad_norm": 1.8905068597211632,
      "learning_rate": 1.0844093727913868e-06,
      "loss": 0.3578,
      "step": 975
    },
    {
      "epoch": 0.8117622696210396,
      "grad_norm": 2.0536970678158206,
      "learning_rate": 1.039846498439727e-06,
      "loss": 0.353,
      "step": 980
    },
    {
      "epoch": 0.8159039138538,
      "grad_norm": 1.9394493842379006,
      "learning_rate": 9.961121089018933e-07,
      "loss": 0.3552,
      "step": 985
    },
    {
      "epoch": 0.8200455580865603,
      "grad_norm": 1.9161234213176144,
      "learning_rate": 9.532153536407923e-07,
      "loss": 0.3572,
      "step": 990
    },
    {
      "epoch": 0.8241872023193207,
      "grad_norm": 2.0679533219870394,
      "learning_rate": 9.111652068818621e-07,
      "loss": 0.3499,
      "step": 995
    },
    {
      "epoch": 0.8283288465520812,
      "grad_norm": 2.007021263318756,
      "learning_rate": 8.699704657356195e-07,
      "loss": 0.3503,
      "step": 1000
    },
    {
      "epoch": 0.8324704907848416,
      "grad_norm": 1.940869076922602,
      "learning_rate": 8.296397483572515e-07,
      "loss": 0.3588,
      "step": 1005
    },
    {
      "epoch": 0.836612135017602,
      "grad_norm": 2.0578809556631774,
      "learning_rate": 7.901814921436624e-07,
      "loss": 0.3497,
      "step": 1010
    },
    {
      "epoch": 0.8407537792503624,
      "grad_norm": 1.9963820836617243,
      "learning_rate": 7.516039519683105e-07,
      "loss": 0.3459,
      "step": 1015
    },
    {
      "epoch": 0.8448954234831229,
      "grad_norm": 2.1312384802093707,
      "learning_rate": 7.139151984542636e-07,
      "loss": 0.3515,
      "step": 1020
    },
    {
      "epoch": 0.8490370677158832,
      "grad_norm": 2.051969582922354,
      "learning_rate": 6.771231162857722e-07,
      "loss": 0.3497,
      "step": 1025
    },
    {
      "epoch": 0.8531787119486436,
      "grad_norm": 1.944237335752014,
      "learning_rate": 6.412354025587509e-07,
      "loss": 0.3454,
      "step": 1030
    },
    {
      "epoch": 0.857320356181404,
      "grad_norm": 2.0013927970500824,
      "learning_rate": 6.062595651705111e-07,
      "loss": 0.3484,
      "step": 1035
    },
    {
      "epoch": 0.8614620004141644,
      "grad_norm": 1.9953241481292785,
      "learning_rate": 5.722029212490666e-07,
      "loss": 0.3467,
      "step": 1040
    },
    {
      "epoch": 0.8656036446469249,
      "grad_norm": 1.8562753423514788,
      "learning_rate": 5.390725956223531e-07,
      "loss": 0.3439,
      "step": 1045
    },
    {
      "epoch": 0.8697452888796853,
      "grad_norm": 2.014412721333329,
      "learning_rate": 5.068755193276798e-07,
      "loss": 0.3475,
      "step": 1050
    },
    {
      "epoch": 0.8738869331124457,
      "grad_norm": 2.0883606449971013,
      "learning_rate": 4.756184281617121e-07,
      "loss": 0.3442,
      "step": 1055
    },
    {
      "epoch": 0.878028577345206,
      "grad_norm": 1.9479502623498623,
      "learning_rate": 4.4530786127131575e-07,
      "loss": 0.3516,
      "step": 1060
    },
    {
      "epoch": 0.8821702215779664,
      "grad_norm": 1.9878832197736611,
      "learning_rate": 4.159501597855287e-07,
      "loss": 0.3468,
      "step": 1065
    },
    {
      "epoch": 0.8863118658107269,
      "grad_norm": 1.981044415868457,
      "learning_rate": 3.8755146548896784e-07,
      "loss": 0.3442,
      "step": 1070
    },
    {
      "epoch": 0.8904535100434873,
      "grad_norm": 2.078622119500157,
      "learning_rate": 3.6011771953693044e-07,
      "loss": 0.3414,
      "step": 1075
    },
    {
      "epoch": 0.8945951542762477,
      "grad_norm": 1.9832065578238605,
      "learning_rate": 3.336546612124758e-07,
      "loss": 0.3462,
      "step": 1080
    },
    {
      "epoch": 0.8987367985090081,
      "grad_norm": 2.077188737775822,
      "learning_rate": 3.081678267257404e-07,
      "loss": 0.3445,
      "step": 1085
    },
    {
      "epoch": 0.9028784427417684,
      "grad_norm": 1.9680080545892447,
      "learning_rate": 2.836625480557265e-07,
      "loss": 0.3433,
      "step": 1090
    },
    {
      "epoch": 0.9070200869745288,
      "grad_norm": 2.0554539318334344,
      "learning_rate": 2.601439518348331e-07,
      "loss": 0.3411,
      "step": 1095
    },
    {
      "epoch": 0.9111617312072893,
      "grad_norm": 1.9738460247856697,
      "learning_rate": 2.376169582763288e-07,
      "loss": 0.3423,
      "step": 1100
    },
    {
      "epoch": 0.9153033754400497,
      "grad_norm": 2.100671455290073,
      "learning_rate": 2.1608628014502364e-07,
      "loss": 0.3412,
      "step": 1105
    },
    {
      "epoch": 0.9194450196728101,
      "grad_norm": 2.0790276531527865,
      "learning_rate": 1.955564217713335e-07,
      "loss": 0.3376,
      "step": 1110
    },
    {
      "epoch": 0.9235866639055705,
      "grad_norm": 2.039950191776352,
      "learning_rate": 1.7603167810894662e-07,
      "loss": 0.3406,
      "step": 1115
    },
    {
      "epoch": 0.927728308138331,
      "grad_norm": 2.0162487373459435,
      "learning_rate": 1.5751613383630128e-07,
      "loss": 0.3465,
      "step": 1120
    },
    {
      "epoch": 0.9318699523710913,
      "grad_norm": 2.092197022372298,
      "learning_rate": 1.4001366250204762e-07,
      "loss": 0.337,
      "step": 1125
    },
    {
      "epoch": 0.9360115966038517,
      "grad_norm": 1.944362051868436,
      "learning_rate": 1.235279257146804e-07,
      "loss": 0.3378,
      "step": 1130
    },
    {
      "epoch": 0.9401532408366121,
      "grad_norm": 2.06431956869675,
      "learning_rate": 1.080623723765134e-07,
      "loss": 0.3352,
      "step": 1135
    },
    {
      "epoch": 0.9442948850693725,
      "grad_norm": 2.067267199918096,
      "learning_rate": 9.362023796215036e-08,
      "loss": 0.3385,
      "step": 1140
    },
    {
      "epoch": 0.948436529302133,
      "grad_norm": 2.050374938357445,
      "learning_rate": 8.020454384160437e-08,
      "loss": 0.345,
      "step": 1145
    },
    {
      "epoch": 0.9525781735348934,
      "grad_norm": 2.058842049841172,
      "learning_rate": 6.78180966482156e-08,
      "loss": 0.3431,
      "step": 1150
    },
    {
      "epoch": 0.9567198177676538,
      "grad_norm": 2.1228820426864825,
      "learning_rate": 5.646348769148491e-08,
      "loss": 0.3415,
      "step": 1155
    },
    {
      "epoch": 0.9608614620004141,
      "grad_norm": 1.9574417210663868,
      "learning_rate": 4.6143092414961396e-08,
      "loss": 0.3346,
      "step": 1160
    },
    {
      "epoch": 0.9650031062331745,
      "grad_norm": 2.0568028289295293,
      "learning_rate": 3.685906989928656e-08,
      "loss": 0.3404,
      "step": 1165
    },
    {
      "epoch": 0.969144750465935,
      "grad_norm": 2.187483815397574,
      "learning_rate": 2.861336241050061e-08,
      "loss": 0.3366,
      "step": 1170
    },
    {
      "epoch": 0.9732863946986954,
      "grad_norm": 2.0602914938549626,
      "learning_rate": 2.1407694993714755e-08,
      "loss": 0.3419,
      "step": 1175
    },
    {
      "epoch": 0.9774280389314558,
      "grad_norm": 2.0097264769694174,
      "learning_rate": 1.5243575112218744e-08,
      "loss": 0.3391,
      "step": 1180
    },
    {
      "epoch": 0.9815696831642162,
      "grad_norm": 2.1300950765468456,
      "learning_rate": 1.0122292332114814e-08,
      "loss": 0.3479,
      "step": 1185
    },
    {
      "epoch": 0.9857113273969766,
      "grad_norm": 1.9814636182776308,
      "learning_rate": 6.044918052531268e-09,
      "loss": 0.3359,
      "step": 1190
    },
    {
      "epoch": 0.989852971629737,
      "grad_norm": 2.0905112637682053,
      "learning_rate": 3.0123052814812203e-09,
      "loss": 0.3383,
      "step": 1195
    },
    {
      "epoch": 0.9939946158624974,
      "grad_norm": 1.966590567136286,
      "learning_rate": 1.025088457409229e-09,
      "loss": 0.3325,
      "step": 1200
    },
    {
      "epoch": 0.9981362600952578,
      "grad_norm": 2.035603528671296,
      "learning_rate": 8.368331646302353e-11,
      "loss": 0.3416,
      "step": 1205
    },
    {
      "epoch": 0.999792917788362,
      "eval_loss": 0.3488326072692871,
      "eval_runtime": 0.9501,
      "eval_samples_per_second": 3.158,
      "eval_steps_per_second": 1.053,
      "step": 1207
    },
    {
      "epoch": 0.999792917788362,
      "step": 1207,
      "total_flos": 252668899491840.0,
      "train_loss": 0.5272446889569172,
      "train_runtime": 29415.576,
      "train_samples_per_second": 1.313,
      "train_steps_per_second": 0.041
    }
  ],
  "logging_steps": 5,
  "max_steps": 1207,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 252668899491840.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}