{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.28164322723908214,
  "eval_steps": 500,
  "global_step": 761,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001850481125092524,
      "grad_norm": 12.649392127990723,
      "learning_rate": 6.325018450184502e-07,
      "loss": 1.3603,
      "mean_token_accuracy": 0.7554670929908752,
      "num_tokens": 185821.0,
      "step": 5
    },
    {
      "epoch": 0.003700962250185048,
      "grad_norm": 9.217720985412598,
      "learning_rate": 1.4231291512915131e-06,
      "loss": 1.2892,
      "mean_token_accuracy": 0.7647651076316834,
      "num_tokens": 374500.0,
      "step": 10
    },
    {
      "epoch": 0.0055514433752775726,
      "grad_norm": 4.445683002471924,
      "learning_rate": 2.2137564575645756e-06,
      "loss": 1.1434,
      "mean_token_accuracy": 0.7748205065727234,
      "num_tokens": 563560.0,
      "step": 15
    },
    {
      "epoch": 0.007401924500370096,
      "grad_norm": 3.0582313537597656,
      "learning_rate": 3.004383763837638e-06,
      "loss": 0.9881,
      "mean_token_accuracy": 0.7829701185226441,
      "num_tokens": 746605.0,
      "step": 20
    },
    {
      "epoch": 0.009252405625462621,
      "grad_norm": 2.2693915367126465,
      "learning_rate": 3.795011070110701e-06,
      "loss": 0.8071,
      "mean_token_accuracy": 0.8049428701400757,
      "num_tokens": 932791.0,
      "step": 25
    },
    {
      "epoch": 0.011102886750555145,
      "grad_norm": 2.1072044372558594,
      "learning_rate": 4.585638376383764e-06,
      "loss": 0.6524,
      "mean_token_accuracy": 0.8388792037963867,
      "num_tokens": 1117440.0,
      "step": 30
    },
    {
      "epoch": 0.012953367875647668,
      "grad_norm": 1.808276891708374,
      "learning_rate": 5.376265682656827e-06,
      "loss": 0.4326,
      "mean_token_accuracy": 0.8950777530670166,
      "num_tokens": 1304726.0,
      "step": 35
    },
    {
      "epoch": 0.014803849000740192,
      "grad_norm": 1.2398462295532227,
      "learning_rate": 6.1668929889298895e-06,
      "loss": 0.2749,
      "mean_token_accuracy": 0.9303830146789551,
      "num_tokens": 1495300.0,
      "step": 40
    },
    {
      "epoch": 0.016654330125832718,
      "grad_norm": 0.8731722235679626,
      "learning_rate": 6.957520295202952e-06,
      "loss": 0.1441,
      "mean_token_accuracy": 0.9656038403511047,
      "num_tokens": 1685752.0,
      "step": 45
    },
    {
      "epoch": 0.018504811250925242,
      "grad_norm": 0.5143530964851379,
      "learning_rate": 7.748147601476015e-06,
      "loss": 0.1277,
      "mean_token_accuracy": 0.966814661026001,
      "num_tokens": 1868182.0,
      "step": 50
    },
    {
      "epoch": 0.020355292376017766,
      "grad_norm": 0.4487469792366028,
      "learning_rate": 8.538774907749078e-06,
      "loss": 0.0714,
      "mean_token_accuracy": 0.9793548464775086,
      "num_tokens": 2057318.0,
      "step": 55
    },
    {
      "epoch": 0.02220577350111029,
      "grad_norm": 0.35000377893447876,
      "learning_rate": 9.32940221402214e-06,
      "loss": 0.0824,
      "mean_token_accuracy": 0.9777526021003723,
      "num_tokens": 2245866.0,
      "step": 60
    },
    {
      "epoch": 0.024056254626202814,
      "grad_norm": 0.1533830463886261,
      "learning_rate": 1.0120029520295203e-05,
      "loss": 0.0538,
      "mean_token_accuracy": 0.9833965539932251,
      "num_tokens": 2433677.0,
      "step": 65
    },
    {
      "epoch": 0.025906735751295335,
      "grad_norm": 0.387611985206604,
      "learning_rate": 1.0910656826568265e-05,
      "loss": 0.044,
      "mean_token_accuracy": 0.9864847898483277,
      "num_tokens": 2623909.0,
      "step": 70
    },
    {
      "epoch": 0.02775721687638786,
      "grad_norm": 0.19073283672332764,
      "learning_rate": 1.170128413284133e-05,
      "loss": 0.0295,
      "mean_token_accuracy": 0.9899263978004456,
      "num_tokens": 2802983.0,
      "step": 75
    },
    {
      "epoch": 0.029607698001480384,
      "grad_norm": 0.7775747179985046,
      "learning_rate": 1.2491911439114392e-05,
      "loss": 0.0263,
      "mean_token_accuracy": 0.9908901691436768,
      "num_tokens": 2986382.0,
      "step": 80
    },
    {
      "epoch": 0.03145817912657291,
      "grad_norm": 0.4266139566898346,
      "learning_rate": 1.3282538745387456e-05,
      "loss": 0.0285,
      "mean_token_accuracy": 0.9905906915664673,
      "num_tokens": 3169983.0,
      "step": 85
    },
    {
      "epoch": 0.033308660251665435,
      "grad_norm": 0.10814467072486877,
      "learning_rate": 1.4073166051660517e-05,
      "loss": 0.0204,
      "mean_token_accuracy": 0.9922121882438659,
      "num_tokens": 3357363.0,
      "step": 90
    },
    {
      "epoch": 0.035159141376757956,
      "grad_norm": 0.354408323764801,
      "learning_rate": 1.4863793357933581e-05,
      "loss": 0.0223,
      "mean_token_accuracy": 0.9915817618370056,
      "num_tokens": 3546512.0,
      "step": 95
    },
    {
      "epoch": 0.037009622501850484,
      "grad_norm": 0.1480749547481537,
      "learning_rate": 1.5654420664206644e-05,
      "loss": 0.0185,
      "mean_token_accuracy": 0.992724347114563,
      "num_tokens": 3735563.0,
      "step": 100
    },
    {
      "epoch": 0.038860103626943004,
      "grad_norm": 0.1472485065460205,
      "learning_rate": 1.6445047970479706e-05,
      "loss": 0.0183,
      "mean_token_accuracy": 0.992663311958313,
      "num_tokens": 3922268.0,
      "step": 105
    },
    {
      "epoch": 0.04071058475203553,
      "grad_norm": 0.10689185559749603,
      "learning_rate": 1.723567527675277e-05,
      "loss": 0.0188,
      "mean_token_accuracy": 0.9926952004432679,
      "num_tokens": 4110402.0,
      "step": 110
    },
    {
      "epoch": 0.04256106587712805,
      "grad_norm": 0.1283423900604248,
      "learning_rate": 1.802630258302583e-05,
      "loss": 0.0154,
      "mean_token_accuracy": 0.9934558510780335,
      "num_tokens": 4301143.0,
      "step": 115
    },
    {
      "epoch": 0.04441154700222058,
      "grad_norm": 0.5970118641853333,
      "learning_rate": 1.8816929889298894e-05,
      "loss": 0.0169,
      "mean_token_accuracy": 0.9931488752365112,
      "num_tokens": 4485870.0,
      "step": 120
    },
    {
      "epoch": 0.0462620281273131,
      "grad_norm": 0.1334739327430725,
      "learning_rate": 1.9607557195571956e-05,
      "loss": 0.0158,
      "mean_token_accuracy": 0.9934451222419739,
      "num_tokens": 4671825.0,
      "step": 125
    },
    {
      "epoch": 0.04811250925240563,
      "grad_norm": 0.1656082421541214,
      "learning_rate": 2.039818450184502e-05,
      "loss": 0.0146,
      "mean_token_accuracy": 0.9936710596084595,
      "num_tokens": 4854954.0,
      "step": 130
    },
    {
      "epoch": 0.04996299037749815,
      "grad_norm": 0.10422646999359131,
      "learning_rate": 2.1188811808118085e-05,
      "loss": 0.0147,
      "mean_token_accuracy": 0.9935644507408142,
      "num_tokens": 5042206.0,
      "step": 135
    },
    {
      "epoch": 0.05181347150259067,
      "grad_norm": 0.1137189045548439,
      "learning_rate": 2.1979439114391147e-05,
      "loss": 0.0147,
      "mean_token_accuracy": 0.9937207102775574,
      "num_tokens": 5231792.0,
      "step": 140
    },
    {
      "epoch": 0.0536639526276832,
      "grad_norm": 0.1359642595052719,
      "learning_rate": 2.277006642066421e-05,
      "loss": 0.0142,
      "mean_token_accuracy": 0.993699848651886,
      "num_tokens": 5412297.0,
      "step": 145
    },
    {
      "epoch": 0.05551443375277572,
      "grad_norm": 0.17001201212406158,
      "learning_rate": 2.356069372693727e-05,
      "loss": 0.0138,
      "mean_token_accuracy": 0.993944239616394,
      "num_tokens": 5599717.0,
      "step": 150
    },
    {
      "epoch": 0.057364914877868246,
      "grad_norm": 0.11590602993965149,
      "learning_rate": 2.435132103321033e-05,
      "loss": 0.0136,
      "mean_token_accuracy": 0.9938363671302796,
      "num_tokens": 5786537.0,
      "step": 155
    },
    {
      "epoch": 0.05921539600296077,
      "grad_norm": 0.11727731674909592,
      "learning_rate": 2.5141948339483397e-05,
      "loss": 0.0143,
      "mean_token_accuracy": 0.993740177154541,
      "num_tokens": 5969585.0,
      "step": 160
    },
    {
      "epoch": 0.061065877128053295,
      "grad_norm": 0.08245962113142014,
      "learning_rate": 2.593257564575646e-05,
      "loss": 0.0141,
      "mean_token_accuracy": 0.9940990090370179,
      "num_tokens": 6156724.0,
      "step": 165
    },
    {
      "epoch": 0.06291635825314582,
      "grad_norm": 0.07970687001943588,
      "learning_rate": 2.6723202952029522e-05,
      "loss": 0.013,
      "mean_token_accuracy": 0.9941760420799255,
      "num_tokens": 6345575.0,
      "step": 170
    },
    {
      "epoch": 0.06476683937823834,
      "grad_norm": 0.20664170384407043,
      "learning_rate": 2.751383025830258e-05,
      "loss": 0.0128,
      "mean_token_accuracy": 0.9942710280418396,
      "num_tokens": 6531385.0,
      "step": 175
    },
    {
      "epoch": 0.06661732050333087,
      "grad_norm": 0.2238393872976303,
      "learning_rate": 2.830445756457565e-05,
      "loss": 0.0143,
      "mean_token_accuracy": 0.9939396977424622,
      "num_tokens": 6715295.0,
      "step": 180
    },
    {
      "epoch": 0.06846780162842339,
      "grad_norm": 0.16475221514701843,
      "learning_rate": 2.909508487084871e-05,
      "loss": 0.0139,
      "mean_token_accuracy": 0.9937853693962098,
      "num_tokens": 6901640.0,
      "step": 185
    },
    {
      "epoch": 0.07031828275351591,
      "grad_norm": 0.11899226903915405,
      "learning_rate": 2.9885712177121772e-05,
      "loss": 0.0139,
      "mean_token_accuracy": 0.9938597559928894,
      "num_tokens": 7088485.0,
      "step": 190
    },
    {
      "epoch": 0.07216876387860843,
      "grad_norm": 0.05497577413916588,
      "learning_rate": 3.0676339483394835e-05,
      "loss": 0.0127,
      "mean_token_accuracy": 0.9940391182899475,
      "num_tokens": 7275336.0,
      "step": 195
    },
    {
      "epoch": 0.07401924500370097,
      "grad_norm": 0.0667397677898407,
      "learning_rate": 3.14669667896679e-05,
      "loss": 0.0136,
      "mean_token_accuracy": 0.994173550605774,
      "num_tokens": 7462575.0,
      "step": 200
    },
    {
      "epoch": 0.07586972612879349,
      "grad_norm": 0.13350225985050201,
      "learning_rate": 3.225759409594096e-05,
      "loss": 0.0142,
      "mean_token_accuracy": 0.9938544273376465,
      "num_tokens": 7650192.0,
      "step": 205
    },
    {
      "epoch": 0.07772020725388601,
      "grad_norm": 0.12862999737262726,
      "learning_rate": 3.3048221402214026e-05,
      "loss": 0.0129,
      "mean_token_accuracy": 0.9943220376968384,
      "num_tokens": 7841316.0,
      "step": 210
    },
    {
      "epoch": 0.07957068837897853,
      "grad_norm": 0.08196930587291718,
      "learning_rate": 3.3838848708487085e-05,
      "loss": 0.0135,
      "mean_token_accuracy": 0.9939719796180725,
      "num_tokens": 8022577.0,
      "step": 215
    },
    {
      "epoch": 0.08142116950407106,
      "grad_norm": 0.11810460686683655,
      "learning_rate": 3.462947601476015e-05,
      "loss": 0.0123,
      "mean_token_accuracy": 0.9945261120796204,
      "num_tokens": 8211139.0,
      "step": 220
    },
    {
      "epoch": 0.08327165062916358,
      "grad_norm": 0.06291542947292328,
      "learning_rate": 3.542010332103321e-05,
      "loss": 0.0126,
      "mean_token_accuracy": 0.9944621562957764,
      "num_tokens": 8397245.0,
      "step": 225
    },
    {
      "epoch": 0.0851221317542561,
      "grad_norm": 0.06278537958860397,
      "learning_rate": 3.6210730627306276e-05,
      "loss": 0.014,
      "mean_token_accuracy": 0.9939502358436585,
      "num_tokens": 8580660.0,
      "step": 230
    },
    {
      "epoch": 0.08697261287934863,
      "grad_norm": 0.051632992923259735,
      "learning_rate": 3.7001357933579335e-05,
      "loss": 0.0122,
      "mean_token_accuracy": 0.9946012616157531,
      "num_tokens": 8768383.0,
      "step": 235
    },
    {
      "epoch": 0.08882309400444116,
      "grad_norm": 0.06995566934347153,
      "learning_rate": 3.77919852398524e-05,
      "loss": 0.0128,
      "mean_token_accuracy": 0.9943013906478881,
      "num_tokens": 8953368.0,
      "step": 240
    },
    {
      "epoch": 0.09067357512953368,
      "grad_norm": 0.1363731175661087,
      "learning_rate": 3.858261254612547e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9944965839385986,
      "num_tokens": 9138830.0,
      "step": 245
    },
    {
      "epoch": 0.0925240562546262,
      "grad_norm": 0.05171045660972595,
      "learning_rate": 3.9373239852398526e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.9946675181388855,
      "num_tokens": 9328907.0,
      "step": 250
    },
    {
      "epoch": 0.09437453737971872,
      "grad_norm": 0.05888763815164566,
      "learning_rate": 4.0163867158671585e-05,
      "loss": 0.0124,
      "mean_token_accuracy": 0.9943785905838013,
      "num_tokens": 9514714.0,
      "step": 255
    },
    {
      "epoch": 0.09622501850481126,
      "grad_norm": 0.0596054270863533,
      "learning_rate": 4.095449446494465e-05,
      "loss": 0.0134,
      "mean_token_accuracy": 0.9940859913825989,
      "num_tokens": 9698135.0,
      "step": 260
    },
    {
      "epoch": 0.09807549962990378,
      "grad_norm": 0.08483681082725525,
      "learning_rate": 4.174512177121772e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9941369533538819,
      "num_tokens": 9882979.0,
      "step": 265
    },
    {
      "epoch": 0.0999259807549963,
      "grad_norm": 0.07195551693439484,
      "learning_rate": 4.2535749077490776e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.994370698928833,
      "num_tokens": 10069134.0,
      "step": 270
    },
    {
      "epoch": 0.10177646188008882,
      "grad_norm": 0.040128063410520554,
      "learning_rate": 4.285197291230987e-05,
      "loss": 0.0127,
      "mean_token_accuracy": 0.9942069530487061,
      "num_tokens": 10257393.0,
      "step": 275
    },
    {
      "epoch": 0.10362694300518134,
      "grad_norm": 0.03847283869981766,
      "learning_rate": 4.285180737675649e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9942315220832825,
      "num_tokens": 10446564.0,
      "step": 280
    },
    {
      "epoch": 0.10547742413027388,
      "grad_norm": 0.052056606858968735,
      "learning_rate": 4.285149135591477e-05,
      "loss": 0.0128,
      "mean_token_accuracy": 0.9941303849220275,
      "num_tokens": 10635225.0,
      "step": 285
    },
    {
      "epoch": 0.1073279052553664,
      "grad_norm": 0.11386790126562119,
      "learning_rate": 4.285102485274419e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9943561673164367,
      "num_tokens": 10824251.0,
      "step": 290
    },
    {
      "epoch": 0.10917838638045892,
      "grad_norm": 0.08637174218893051,
      "learning_rate": 4.285040787161343e-05,
      "loss": 0.0122,
      "mean_token_accuracy": 0.9942921042442322,
      "num_tokens": 11007473.0,
      "step": 295
    },
    {
      "epoch": 0.11102886750555144,
      "grad_norm": 0.08053375780582428,
      "learning_rate": 4.284964041830038e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9943666338920594,
      "num_tokens": 11195505.0,
      "step": 300
    },
    {
      "epoch": 0.11287934863064397,
      "grad_norm": 0.053617142140865326,
      "learning_rate": 4.284872249999205e-05,
      "loss": 0.0122,
      "mean_token_accuracy": 0.9943767070770264,
      "num_tokens": 11383359.0,
      "step": 305
    },
    {
      "epoch": 0.11472982975573649,
      "grad_norm": 0.236342191696167,
      "learning_rate": 4.284765412528454e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9943820238113403,
      "num_tokens": 11570508.0,
      "step": 310
    },
    {
      "epoch": 0.11658031088082901,
      "grad_norm": 0.0628127008676529,
      "learning_rate": 4.2846435304182934e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.9945466637611389,
      "num_tokens": 11760407.0,
      "step": 315
    },
    {
      "epoch": 0.11843079200592153,
      "grad_norm": 0.07259124517440796,
      "learning_rate": 4.284506604810118e-05,
      "loss": 0.0117,
      "mean_token_accuracy": 0.994793963432312,
      "num_tokens": 11946394.0,
      "step": 320
    },
    {
      "epoch": 0.12028127313101407,
      "grad_norm": 0.12081380188465118,
      "learning_rate": 4.284354636986206e-05,
      "loss": 0.0131,
      "mean_token_accuracy": 0.9942123293876648,
      "num_tokens": 12135829.0,
      "step": 325
    },
    {
      "epoch": 0.12213175425610659,
      "grad_norm": 0.032616619020700455,
      "learning_rate": 4.284187628369701e-05,
      "loss": 0.0123,
      "mean_token_accuracy": 0.9943543672561646,
      "num_tokens": 12326292.0,
      "step": 330
    },
    {
      "epoch": 0.12398223538119911,
      "grad_norm": 0.0659327581524849,
      "learning_rate": 4.284005580524598e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.9946148633956909,
      "num_tokens": 12512244.0,
      "step": 335
    },
    {
      "epoch": 0.12583271650629163,
      "grad_norm": 0.0538957342505455,
      "learning_rate": 4.283808495155733e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9943965435028076,
      "num_tokens": 12698345.0,
      "step": 340
    },
    {
      "epoch": 0.12768319763138417,
      "grad_norm": 0.026756659150123596,
      "learning_rate": 4.283596374108763e-05,
      "loss": 0.0117,
      "mean_token_accuracy": 0.9946674346923828,
      "num_tokens": 12886124.0,
      "step": 345
    },
    {
      "epoch": 0.12953367875647667,
      "grad_norm": 0.03937589004635811,
      "learning_rate": 4.2833692193701535e-05,
      "loss": 0.0124,
      "mean_token_accuracy": 0.9941404342651368,
      "num_tokens": 13072522.0,
      "step": 350
    },
    {
      "epoch": 0.1313841598815692,
      "grad_norm": 0.11673159897327423,
      "learning_rate": 4.283127033067154e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.994466495513916,
      "num_tokens": 13257550.0,
      "step": 355
    },
    {
      "epoch": 0.13323464100666174,
      "grad_norm": 0.04208122193813324,
      "learning_rate": 4.282869817467783e-05,
      "loss": 0.0127,
      "mean_token_accuracy": 0.9941461205482482,
      "num_tokens": 13443219.0,
      "step": 360
    },
    {
      "epoch": 0.13508512213175425,
      "grad_norm": 0.0433068685233593,
      "learning_rate": 4.282597574980804e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.9947281122207642,
      "num_tokens": 13630868.0,
      "step": 365
    },
    {
      "epoch": 0.13693560325684678,
      "grad_norm": 0.11256636679172516,
      "learning_rate": 4.2823103081557036e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.994697916507721,
      "num_tokens": 13819698.0,
      "step": 370
    },
    {
      "epoch": 0.13878608438193932,
      "grad_norm": 0.05056292563676834,
      "learning_rate": 4.282008019682668e-05,
      "loss": 0.0122,
      "mean_token_accuracy": 0.9944531917572021,
      "num_tokens": 14006605.0,
      "step": 375
    },
    {
      "epoch": 0.14063656550703182,
      "grad_norm": 0.04138778895139694,
      "learning_rate": 4.2816907123925595e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.9946407318115235,
      "num_tokens": 14194590.0,
      "step": 380
    },
    {
      "epoch": 0.14248704663212436,
      "grad_norm": 0.04915003105998039,
      "learning_rate": 4.2813583892568835e-05,
      "loss": 0.0115,
      "mean_token_accuracy": 0.9946125507354736,
      "num_tokens": 14382780.0,
      "step": 385
    },
    {
      "epoch": 0.14433752775721687,
      "grad_norm": 0.0317862406373024,
      "learning_rate": 4.28101105338777e-05,
      "loss": 0.0127,
      "mean_token_accuracy": 0.9941527843475342,
      "num_tokens": 14570415.0,
      "step": 390
    },
    {
      "epoch": 0.1461880088823094,
      "grad_norm": 0.04320110008120537,
      "learning_rate": 4.280648708037937e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.9945583343505859,
      "num_tokens": 14757326.0,
      "step": 395
    },
    {
      "epoch": 0.14803849000740193,
      "grad_norm": 0.0487859845161438,
      "learning_rate": 4.280271356600664e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9947203397750854,
      "num_tokens": 14939541.0,
      "step": 400
    },
    {
      "epoch": 0.14988897113249444,
      "grad_norm": 0.019251326099038124,
      "learning_rate": 4.2798790026097594e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.994676387310028,
      "num_tokens": 15126309.0,
      "step": 405
    },
    {
      "epoch": 0.15173945225758698,
      "grad_norm": 0.06465357542037964,
      "learning_rate": 4.2794716497395245e-05,
      "loss": 0.0129,
      "mean_token_accuracy": 0.9943367958068847,
      "num_tokens": 15310070.0,
      "step": 410
    },
    {
      "epoch": 0.15358993338267948,
      "grad_norm": 0.05526921898126602,
      "learning_rate": 4.2790493018047233e-05,
      "loss": 0.0122,
      "mean_token_accuracy": 0.9944483518600464,
      "num_tokens": 15495949.0,
      "step": 415
    },
    {
      "epoch": 0.15544041450777202,
      "grad_norm": 0.034128863364458084,
      "learning_rate": 4.278611962760546e-05,
      "loss": 0.0123,
      "mean_token_accuracy": 0.9941602945327759,
      "num_tokens": 15683861.0,
      "step": 420
    },
    {
      "epoch": 0.15729089563286455,
      "grad_norm": 0.040949802845716476,
      "learning_rate": 4.2781596367025705e-05,
      "loss": 0.0124,
      "mean_token_accuracy": 0.9942546725273133,
      "num_tokens": 15870718.0,
      "step": 425
    },
    {
      "epoch": 0.15914137675795706,
      "grad_norm": 0.035726647824048996,
      "learning_rate": 4.277692327866723e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9944644689559936,
      "num_tokens": 16061189.0,
      "step": 430
    },
    {
      "epoch": 0.1609918578830496,
      "grad_norm": 0.04440707713365555,
      "learning_rate": 4.27721004062924e-05,
      "loss": 0.0131,
      "mean_token_accuracy": 0.9941565513610839,
      "num_tokens": 16244331.0,
      "step": 435
    },
    {
      "epoch": 0.16284233900814213,
      "grad_norm": 0.04277065768837929,
      "learning_rate": 4.27671277950663e-05,
      "loss": 0.0128,
      "mean_token_accuracy": 0.9941763043403625,
      "num_tokens": 16428979.0,
      "step": 440
    },
    {
      "epoch": 0.16469282013323464,
      "grad_norm": 0.05249185860157013,
      "learning_rate": 4.276200549155626e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9945122838020325,
      "num_tokens": 16616412.0,
      "step": 445
    },
    {
      "epoch": 0.16654330125832717,
      "grad_norm": 0.05636998265981674,
      "learning_rate": 4.2756733543731456e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.9945295214653015,
      "num_tokens": 16803515.0,
      "step": 450
    },
    {
      "epoch": 0.16839378238341968,
      "grad_norm": 0.022253001108765602,
      "learning_rate": 4.275131200096242e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.994416344165802,
      "num_tokens": 16989032.0,
      "step": 455
    },
    {
      "epoch": 0.1702442635085122,
      "grad_norm": 0.032556477934122086,
      "learning_rate": 4.2745740914020645e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.9945114493370056,
      "num_tokens": 17178481.0,
      "step": 460
    },
    {
      "epoch": 0.17209474463360475,
      "grad_norm": 0.057244472205638885,
      "learning_rate": 4.2740020335078044e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.9945224165916443,
      "num_tokens": 17364736.0,
      "step": 465
    },
    {
      "epoch": 0.17394522575869725,
      "grad_norm": 0.052654314786195755,
      "learning_rate": 4.27341503177065e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9944727659225464,
      "num_tokens": 17546812.0,
      "step": 470
    },
    {
      "epoch": 0.1757957068837898,
      "grad_norm": 0.06236141547560692,
      "learning_rate": 4.2728130916877336e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9944144725799561,
      "num_tokens": 17732195.0,
      "step": 475
    },
    {
      "epoch": 0.17764618800888232,
      "grad_norm": 0.04595118761062622,
      "learning_rate": 4.272196218896084e-05,
      "loss": 0.0127,
      "mean_token_accuracy": 0.9942219495773316,
      "num_tokens": 17915325.0,
      "step": 480
    },
    {
      "epoch": 0.17949666913397483,
      "grad_norm": 0.023451240733265877,
      "learning_rate": 4.2715644191725686e-05,
      "loss": 0.0129,
      "mean_token_accuracy": 0.9942497849464417,
      "num_tokens": 18101917.0,
      "step": 485
    },
    {
      "epoch": 0.18134715025906736,
      "grad_norm": 0.06435070186853409,
      "learning_rate": 4.270917698433845e-05,
      "loss": 0.0123,
      "mean_token_accuracy": 0.9943255424499512,
      "num_tokens": 18286995.0,
      "step": 490
    },
    {
      "epoch": 0.18319763138415987,
      "grad_norm": 0.020520444959402084,
      "learning_rate": 4.270256062736299e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.994460117816925,
      "num_tokens": 18473747.0,
      "step": 495
    },
    {
      "epoch": 0.1850481125092524,
      "grad_norm": 0.060110948979854584,
      "learning_rate": 4.269579518275996e-05,
      "loss": 0.0117,
      "mean_token_accuracy": 0.9946680665016174,
      "num_tokens": 18661411.0,
      "step": 500
    },
    {
      "epoch": 0.18689859363434494,
      "grad_norm": 0.025209549814462662,
      "learning_rate": 4.268888071388614e-05,
      "loss": 0.0117,
      "mean_token_accuracy": 0.9945947647094726,
      "num_tokens": 18850199.0,
      "step": 505
    },
    {
      "epoch": 0.18874907475943745,
      "grad_norm": 0.018915211781859398,
      "learning_rate": 4.268181728549391e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9943953394889832,
      "num_tokens": 19035469.0,
      "step": 510
    },
    {
      "epoch": 0.19059955588452998,
      "grad_norm": 0.02561143785715103,
      "learning_rate": 4.267460496373061e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9942945837974548,
      "num_tokens": 19217948.0,
      "step": 515
    },
    {
      "epoch": 0.19245003700962252,
      "grad_norm": 0.0442853718996048,
      "learning_rate": 4.266724381613796e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9944578647613526,
      "num_tokens": 19404069.0,
      "step": 520
    },
    {
      "epoch": 0.19430051813471502,
      "grad_norm": 0.04662763699889183,
      "learning_rate": 4.265973391165136e-05,
      "loss": 0.0129,
      "mean_token_accuracy": 0.9942412853240967,
      "num_tokens": 19585254.0,
      "step": 525
    },
    {
      "epoch": 0.19615099925980756,
      "grad_norm": 0.0530804768204689,
      "learning_rate": 4.2652075320599303e-05,
      "loss": 0.0117,
      "mean_token_accuracy": 0.9946515679359436,
      "num_tokens": 19775480.0,
      "step": 530
    },
    {
      "epoch": 0.19800148038490006,
      "grad_norm": 0.03273262456059456,
      "learning_rate": 4.264426811470271e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9944777727127075,
      "num_tokens": 19959932.0,
      "step": 535
    },
    {
      "epoch": 0.1998519615099926,
      "grad_norm": 0.04370114952325821,
      "learning_rate": 4.263631236707421e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.9945811629295349,
      "num_tokens": 20146524.0,
      "step": 540
    },
    {
      "epoch": 0.20170244263508513,
      "grad_norm": 0.023784466087818146,
      "learning_rate": 4.262820815221752e-05,
      "loss": 0.0123,
      "mean_token_accuracy": 0.9943390250205993,
      "num_tokens": 20332234.0,
      "step": 545
    },
    {
      "epoch": 0.20355292376017764,
      "grad_norm": 0.8499415516853333,
      "learning_rate": 4.261995554602671e-05,
      "loss": 0.0353,
      "mean_token_accuracy": 0.9902504801750183,
      "num_tokens": 20514499.0,
      "step": 550
    },
    {
      "epoch": 0.20540340488527017,
      "grad_norm": 1.1610054969787598,
      "learning_rate": 4.261155462578549e-05,
      "loss": 0.0206,
      "mean_token_accuracy": 0.9920717954635621,
      "num_tokens": 20700594.0,
      "step": 555
    },
    {
      "epoch": 0.20725388601036268,
      "grad_norm": 0.13676072657108307,
      "learning_rate": 4.260300547016651e-05,
      "loss": 0.0163,
      "mean_token_accuracy": 0.9935830473899842,
      "num_tokens": 20890167.0,
      "step": 560
    },
    {
      "epoch": 0.20910436713545522,
      "grad_norm": 0.15704846382141113,
      "learning_rate": 4.259430815923061e-05,
      "loss": 0.0147,
      "mean_token_accuracy": 0.9940526485443115,
      "num_tokens": 21076844.0,
      "step": 565
    },
    {
      "epoch": 0.21095484826054775,
      "grad_norm": 0.11786659061908722,
      "learning_rate": 4.258546277442605e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9944088220596313,
      "num_tokens": 21262899.0,
      "step": 570
    },
    {
      "epoch": 0.21280532938564026,
      "grad_norm": 0.11840520799160004,
      "learning_rate": 4.2576469398587796e-05,
      "loss": 0.0144,
      "mean_token_accuracy": 0.9937516927719117,
      "num_tokens": 21448343.0,
      "step": 575
    },
    {
      "epoch": 0.2146558105107328,
      "grad_norm": 0.08956237137317657,
      "learning_rate": 4.2567328115936684e-05,
      "loss": 0.0136,
      "mean_token_accuracy": 0.9941320896148682,
      "num_tokens": 21634019.0,
      "step": 580
    },
    {
      "epoch": 0.21650629163582533,
      "grad_norm": 0.02889622002840042,
      "learning_rate": 4.255803901207869e-05,
      "loss": 0.0139,
      "mean_token_accuracy": 0.9940349817276001,
      "num_tokens": 21816536.0,
      "step": 585
    },
    {
      "epoch": 0.21835677276091783,
      "grad_norm": 0.026375873014330864,
      "learning_rate": 4.254860217400409e-05,
      "loss": 0.0123,
      "mean_token_accuracy": 0.9942944407463074,
      "num_tokens": 22002114.0,
      "step": 590
    },
    {
      "epoch": 0.22020725388601037,
      "grad_norm": 0.023582832887768745,
      "learning_rate": 4.253901769008664e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9944105982780457,
      "num_tokens": 22186004.0,
      "step": 595
    },
    {
      "epoch": 0.22205773501110287,
      "grad_norm": 0.029931364580988884,
      "learning_rate": 4.25292856500828e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.9946899533271789,
      "num_tokens": 22369126.0,
      "step": 600
    },
    {
      "epoch": 0.2239082161361954,
      "grad_norm": 0.03375266492366791,
      "learning_rate": 4.251940614513083e-05,
      "loss": 0.0126,
      "mean_token_accuracy": 0.9943894147872925,
      "num_tokens": 22557508.0,
      "step": 605
    },
    {
      "epoch": 0.22575869726128794,
      "grad_norm": 0.037536900490522385,
      "learning_rate": 4.2509379267749967e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9944296598434448,
      "num_tokens": 22740164.0,
      "step": 610
    },
    {
      "epoch": 0.22760917838638045,
      "grad_norm": 0.02084357477724552,
      "learning_rate": 4.249920511183957e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.994500744342804,
      "num_tokens": 22926105.0,
      "step": 615
    },
    {
      "epoch": 0.22945965951147299,
      "grad_norm": 0.018535546958446503,
      "learning_rate": 4.248888377267821e-05,
      "loss": 0.0122,
      "mean_token_accuracy": 0.9944629311561585,
      "num_tokens": 23109821.0,
      "step": 620
    },
    {
      "epoch": 0.23131014063656552,
      "grad_norm": 0.03311913460493088,
      "learning_rate": 4.247841534692281e-05,
      "loss": 0.0126,
      "mean_token_accuracy": 0.9942508697509765,
      "num_tokens": 23298500.0,
      "step": 625
    },
    {
      "epoch": 0.23316062176165803,
      "grad_norm": 0.03697752580046654,
      "learning_rate": 4.246779993260773e-05,
      "loss": 0.0125,
      "mean_token_accuracy": 0.9944444537162781,
      "num_tokens": 23484827.0,
      "step": 630
    },
    {
      "epoch": 0.23501110288675056,
      "grad_norm": 0.04822184517979622,
      "learning_rate": 4.24570376291438e-05,
      "loss": 0.013,
      "mean_token_accuracy": 0.9942883372306823,
      "num_tokens": 23672624.0,
      "step": 635
    },
    {
      "epoch": 0.23686158401184307,
      "grad_norm": 0.015962179750204086,
      "learning_rate": 4.2446128537317495e-05,
      "loss": 0.0126,
      "mean_token_accuracy": 0.9940625190734863,
      "num_tokens": 23855305.0,
      "step": 640
    },
    {
      "epoch": 0.2387120651369356,
      "grad_norm": 0.023596221581101418,
      "learning_rate": 4.243507275928988e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9944760918617248,
      "num_tokens": 24039098.0,
      "step": 645
    },
    {
      "epoch": 0.24056254626202814,
      "grad_norm": 0.04977529123425484,
      "learning_rate": 4.2423870398595724e-05,
      "loss": 0.0131,
      "mean_token_accuracy": 0.9939121842384339,
      "num_tokens": 24224255.0,
      "step": 650
    },
    {
      "epoch": 0.24241302738712064,
      "grad_norm": 0.02917158044874668,
      "learning_rate": 4.24125215601425e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9942941784858703,
      "num_tokens": 24408183.0,
      "step": 655
    },
    {
      "epoch": 0.24426350851221318,
      "grad_norm": 0.023342030122876167,
      "learning_rate": 4.240102635020941e-05,
      "loss": 0.0115,
      "mean_token_accuracy": 0.9947278022766113,
      "num_tokens": 24596912.0,
      "step": 660
    },
    {
      "epoch": 0.24611398963730569,
      "grad_norm": 0.019149361178278923,
      "learning_rate": 4.238938487644642e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9943392276763916,
      "num_tokens": 24783474.0,
      "step": 665
    },
    {
      "epoch": 0.24796447076239822,
      "grad_norm": 0.0178193561732769,
      "learning_rate": 4.237759724787315e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9944109082221985,
      "num_tokens": 24969623.0,
      "step": 670
    },
    {
      "epoch": 0.24981495188749075,
      "grad_norm": 0.05727710947394371,
      "learning_rate": 4.2365663574878014e-05,
      "loss": 0.012,
      "mean_token_accuracy": 0.9943978309631347,
      "num_tokens": 25158466.0,
      "step": 675
    },
    {
      "epoch": 0.25166543301258326,
      "grad_norm": 0.030884942039847374,
      "learning_rate": 4.235358396921704e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.9944790363311767,
      "num_tokens": 25346048.0,
      "step": 680
    },
    {
      "epoch": 0.25351591413767577,
      "grad_norm": 0.026175323873758316,
      "learning_rate": 4.234135854401289e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.9945067286491394,
      "num_tokens": 25532502.0,
      "step": 685
    },
    {
      "epoch": 0.25536639526276833,
      "grad_norm": 0.023625630885362625,
      "learning_rate": 4.232898741375382e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9944512009620666,
      "num_tokens": 25717818.0,
      "step": 690
    },
    {
      "epoch": 0.25721687638786084,
      "grad_norm": 0.023306002840399742,
      "learning_rate": 4.2316470694292536e-05,
      "loss": 0.0115,
      "mean_token_accuracy": 0.9947312355041504,
      "num_tokens": 25905758.0,
      "step": 695
    },
    {
      "epoch": 0.25906735751295334,
      "grad_norm": 0.04023386910557747,
      "learning_rate": 4.230380850284518e-05,
      "loss": 0.0116,
      "mean_token_accuracy": 0.9947368025779724,
      "num_tokens": 26096154.0,
      "step": 700
    },
    {
      "epoch": 0.2609178386380459,
      "grad_norm": 0.054880719631910324,
      "learning_rate": 4.229100095799022e-05,
      "loss": 0.0127,
      "mean_token_accuracy": 0.9944626927375794,
      "num_tokens": 26282576.0,
      "step": 705
    },
    {
      "epoch": 0.2627683197631384,
      "grad_norm": 0.05358368158340454,
      "learning_rate": 4.227804817966728e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.994605278968811,
      "num_tokens": 26466539.0,
      "step": 710
    },
    {
      "epoch": 0.2646188008882309,
      "grad_norm": 0.049045782536268234,
      "learning_rate": 4.226495028917609e-05,
      "loss": 0.0128,
      "mean_token_accuracy": 0.9942863821983338,
      "num_tokens": 26656251.0,
      "step": 715
    },
    {
      "epoch": 0.2664692820133235,
      "grad_norm": 0.0411219596862793,
      "learning_rate": 4.225170740917533e-05,
      "loss": 0.0121,
      "mean_token_accuracy": 0.9943994283676147,
      "num_tokens": 26838459.0,
      "step": 720
    },
    {
      "epoch": 0.268319763138416,
      "grad_norm": 0.030834250152111053,
      "learning_rate": 4.223831966368145e-05,
      "loss": 0.0122,
      "mean_token_accuracy": 0.9942387580871582,
      "num_tokens": 27019549.0,
      "step": 725
    },
    {
      "epoch": 0.2701702442635085,
      "grad_norm": 0.07497315853834152,
      "learning_rate": 4.222478717806755e-05,
      "loss": 0.0127,
      "mean_token_accuracy": 0.9943762421607971,
      "num_tokens": 27205313.0,
      "step": 730
    },
    {
      "epoch": 0.27202072538860106,
      "grad_norm": 0.05946696922183037,
      "learning_rate": 4.2211110079062164e-05,
      "loss": 0.0127,
      "mean_token_accuracy": 0.9942971229553222,
      "num_tokens": 27391816.0,
      "step": 735
    },
    {
      "epoch": 0.27387120651369357,
      "grad_norm": 0.03350157290697098,
      "learning_rate": 4.219728849474812e-05,
      "loss": 0.0119,
      "mean_token_accuracy": 0.9944700241088867,
      "num_tokens": 27582143.0,
      "step": 740
    },
    {
      "epoch": 0.2757216876387861,
      "grad_norm": 0.034784168004989624,
      "learning_rate": 4.218332255456131e-05,
      "loss": 0.0128,
      "mean_token_accuracy": 0.994387972354889,
      "num_tokens": 27769962.0,
      "step": 745
    },
    {
      "epoch": 0.27757216876387864,
      "grad_norm": 0.02214915119111538,
      "learning_rate": 4.2169212389289474e-05,
      "loss": 0.0118,
      "mean_token_accuracy": 0.994516122341156,
      "num_tokens": 27954489.0,
      "step": 750
    },
    {
      "epoch": 0.27942264988897114,
      "grad_norm": 0.019403742626309395,
      "learning_rate": 4.215495813107101e-05,
      "loss": 0.0117,
      "mean_token_accuracy": 0.9947906136512756,
      "num_tokens": 28146292.0,
      "step": 755
    },
    {
      "epoch": 0.28127313101406365,
      "grad_norm": 0.02483655884861946,
      "learning_rate": 4.214055991339368e-05,
      "loss": 0.0122,
      "mean_token_accuracy": 0.9943630814552307,
      "num_tokens": 28330741.0,
      "step": 760
    }
  ],
  "logging_steps": 5,
  "max_steps": 5404,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.634430263784899e+17,
  "train_batch_size": 20,
  "trial_name": null,
  "trial_params": null
}