{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2094240,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009550003820001528,
      "grad_norm": 1.1688205003738403,
      "learning_rate": 0.00028647758200830823,
      "loss": 4.7726,
      "mean_token_accuracy": 0.29613289506579993,
      "num_tokens": 716800000.0,
      "step": 20000
    },
    {
      "epoch": 0.019100007640003056,
      "grad_norm": 1.0122355222702026,
      "learning_rate": 0.00029724265264455594,
      "loss": 3.5064,
      "mean_token_accuracy": 0.39106567760258915,
      "num_tokens": 1433600000.0,
      "step": 40000
    },
    {
      "epoch": 0.028650011460004583,
      "grad_norm": 0.9957022070884705,
      "learning_rate": 0.0002943487112555509,
      "loss": 3.4122,
      "mean_token_accuracy": 0.40039802242666483,
      "num_tokens": 2150400000.0,
      "step": 60000
    },
    {
      "epoch": 0.03820001528000611,
      "grad_norm": 1.2025322914123535,
      "learning_rate": 0.00029145476986654586,
      "loss": 3.3766,
      "mean_token_accuracy": 0.404039168266952,
      "num_tokens": 2867200000.0,
      "step": 80000
    },
    {
      "epoch": 0.04775001910000764,
      "grad_norm": 1.0380007028579712,
      "learning_rate": 0.0002885608284775408,
      "loss": 3.3565,
      "mean_token_accuracy": 0.40606449462026356,
      "num_tokens": 3584000000.0,
      "step": 100000
    },
    {
      "epoch": 0.057300022920009165,
      "grad_norm": 1.1203994750976562,
      "learning_rate": 0.0002856668870885358,
      "loss": 3.3421,
      "mean_token_accuracy": 0.40754339938014744,
      "num_tokens": 4300800000.0,
      "step": 120000
    },
    {
      "epoch": 0.0668500267400107,
      "grad_norm": 1.1507455110549927,
      "learning_rate": 0.0002827729456995307,
      "loss": 3.3327,
      "mean_token_accuracy": 0.40846395269036295,
      "num_tokens": 5017600000.0,
      "step": 140000
    },
    {
      "epoch": 0.07640003056001222,
      "grad_norm": 1.2616394758224487,
      "learning_rate": 0.0002798790043105257,
      "loss": 3.3239,
      "mean_token_accuracy": 0.40932240092903377,
      "num_tokens": 5734400000.0,
      "step": 160000
    },
    {
      "epoch": 0.08595003438001375,
      "grad_norm": 1.2600022554397583,
      "learning_rate": 0.00027698506292152063,
      "loss": 3.3172,
      "mean_token_accuracy": 0.41004598908573386,
      "num_tokens": 6451200000.0,
      "step": 180000
    },
    {
      "epoch": 0.09550003820001528,
      "grad_norm": 1.1764442920684814,
      "learning_rate": 0.00027409112153251557,
      "loss": 3.3119,
      "mean_token_accuracy": 0.410561589974165,
      "num_tokens": 7168000000.0,
      "step": 200000
    },
    {
      "epoch": 0.1050500420200168,
      "grad_norm": 1.2408899068832397,
      "learning_rate": 0.00027119718014351055,
      "loss": 3.3061,
      "mean_token_accuracy": 0.411199274918437,
      "num_tokens": 7884800000.0,
      "step": 220000
    },
    {
      "epoch": 0.11460004584001833,
      "grad_norm": 1.2953583002090454,
      "learning_rate": 0.0002683032387545055,
      "loss": 3.3012,
      "mean_token_accuracy": 0.41174386738538743,
      "num_tokens": 8601600000.0,
      "step": 240000
    },
    {
      "epoch": 0.12415004966001987,
      "grad_norm": 1.3780796527862549,
      "learning_rate": 0.0002654092973655004,
      "loss": 3.2974,
      "mean_token_accuracy": 0.41211788419932127,
      "num_tokens": 9318400000.0,
      "step": 260000
    },
    {
      "epoch": 0.1337000534800214,
      "grad_norm": 1.2816623449325562,
      "learning_rate": 0.0002625153559764954,
      "loss": 3.2952,
      "mean_token_accuracy": 0.4122900270193815,
      "num_tokens": 10035200000.0,
      "step": 280000
    },
    {
      "epoch": 0.1432500573000229,
      "grad_norm": 1.462896466255188,
      "learning_rate": 0.00025962141458749034,
      "loss": 3.2898,
      "mean_token_accuracy": 0.4128769779801369,
      "num_tokens": 10752000000.0,
      "step": 300000
    },
    {
      "epoch": 0.15280006112002445,
      "grad_norm": 1.4050343036651611,
      "learning_rate": 0.0002567274731984853,
      "loss": 3.2866,
      "mean_token_accuracy": 0.4132985157236457,
      "num_tokens": 11468800000.0,
      "step": 320000
    },
    {
      "epoch": 0.162350064940026,
      "grad_norm": 1.3888232707977295,
      "learning_rate": 0.00025383353180948026,
      "loss": 3.2839,
      "mean_token_accuracy": 0.41356196113973853,
      "num_tokens": 12185600000.0,
      "step": 340000
    },
    {
      "epoch": 0.1719000687600275,
      "grad_norm": 1.3985786437988281,
      "learning_rate": 0.0002509395904204752,
      "loss": 3.2824,
      "mean_token_accuracy": 0.4136822006031871,
      "num_tokens": 12902400000.0,
      "step": 360000
    },
    {
      "epoch": 0.18145007258002904,
      "grad_norm": 1.5853444337844849,
      "learning_rate": 0.00024804564903147013,
      "loss": 3.2796,
      "mean_token_accuracy": 0.41403463195711376,
      "num_tokens": 13619200000.0,
      "step": 380000
    },
    {
      "epoch": 0.19100007640003056,
      "grad_norm": 1.5639410018920898,
      "learning_rate": 0.00024515170764246506,
      "loss": 3.2773,
      "mean_token_accuracy": 0.4142179542243481,
      "num_tokens": 14336000000.0,
      "step": 400000
    },
    {
      "epoch": 0.2005500802200321,
      "grad_norm": 1.4351433515548706,
      "learning_rate": 0.00024225776625346005,
      "loss": 3.2752,
      "mean_token_accuracy": 0.41444926087111233,
      "num_tokens": 15052800000.0,
      "step": 420000
    },
    {
      "epoch": 0.2101000840400336,
      "grad_norm": 1.3414918184280396,
      "learning_rate": 0.00023936382486445498,
      "loss": 3.2725,
      "mean_token_accuracy": 0.41476476649940014,
      "num_tokens": 15769600000.0,
      "step": 440000
    },
    {
      "epoch": 0.21965008786003515,
      "grad_norm": 1.3302001953125,
      "learning_rate": 0.00023646988347544994,
      "loss": 3.2696,
      "mean_token_accuracy": 0.4150806698143482,
      "num_tokens": 16486400000.0,
      "step": 460000
    },
    {
      "epoch": 0.22920009168003666,
      "grad_norm": 1.3756541013717651,
      "learning_rate": 0.00023357594208644488,
      "loss": 3.2683,
      "mean_token_accuracy": 0.4152164765149355,
      "num_tokens": 17203200000.0,
      "step": 480000
    },
    {
      "epoch": 0.2387500955000382,
      "grad_norm": 1.330269694328308,
      "learning_rate": 0.00023068200069743984,
      "loss": 3.2665,
      "mean_token_accuracy": 0.4154003126785159,
      "num_tokens": 17920000000.0,
      "step": 500000
    },
    {
      "epoch": 0.24830009932003974,
      "grad_norm": 1.7756025791168213,
      "learning_rate": 0.00022778805930843483,
      "loss": 3.2648,
      "mean_token_accuracy": 0.4155686768323183,
      "num_tokens": 18636800000.0,
      "step": 520000
    },
    {
      "epoch": 0.25785010314004125,
      "grad_norm": 1.6679662466049194,
      "learning_rate": 0.00022489411791942976,
      "loss": 3.2617,
      "mean_token_accuracy": 0.41595828180462124,
      "num_tokens": 19353600000.0,
      "step": 540000
    },
    {
      "epoch": 0.2674001069600428,
      "grad_norm": 1.5160603523254395,
      "learning_rate": 0.00022200017653042472,
      "loss": 3.2601,
      "mean_token_accuracy": 0.4160811348050833,
      "num_tokens": 716800000.0,
      "step": 560000
    },
    {
      "epoch": 0.27695011078004433,
      "grad_norm": 2.128943681716919,
      "learning_rate": 0.00021910623514141968,
      "loss": 3.259,
      "mean_token_accuracy": 0.41622272021621465,
      "num_tokens": 1433600000.0,
      "step": 580000
    },
    {
      "epoch": 0.2865001146000458,
      "grad_norm": 1.550946593284607,
      "learning_rate": 0.00021621229375241461,
      "loss": 3.2555,
      "mean_token_accuracy": 0.4166563056409359,
      "num_tokens": 2150400000.0,
      "step": 600000
    },
    {
      "epoch": 0.29605011842004736,
      "grad_norm": 1.5084576606750488,
      "learning_rate": 0.00021331835236340957,
      "loss": 3.2543,
      "mean_token_accuracy": 0.4167668910384178,
      "num_tokens": 2867200000.0,
      "step": 620000
    },
    {
      "epoch": 0.3056001222400489,
      "grad_norm": 1.7308917045593262,
      "learning_rate": 0.00021042441097440454,
      "loss": 3.2531,
      "mean_token_accuracy": 0.416923209066689,
      "num_tokens": 3584000000.0,
      "step": 640000
    },
    {
      "epoch": 0.31515012606005044,
      "grad_norm": 1.6496219635009766,
      "learning_rate": 0.00020753046958539947,
      "loss": 3.2503,
      "mean_token_accuracy": 0.4171900423392653,
      "num_tokens": 4300800000.0,
      "step": 660000
    },
    {
      "epoch": 0.324700129880052,
      "grad_norm": 1.911421537399292,
      "learning_rate": 0.00020463652819639443,
      "loss": 3.2494,
      "mean_token_accuracy": 0.4172973266944289,
      "num_tokens": 5017600000.0,
      "step": 680000
    },
    {
      "epoch": 0.33425013370005346,
      "grad_norm": 1.625441074371338,
      "learning_rate": 0.00020174258680738936,
      "loss": 3.2485,
      "mean_token_accuracy": 0.4173892762258649,
      "num_tokens": 5734400000.0,
      "step": 700000
    },
    {
      "epoch": 0.343800137520055,
      "grad_norm": 1.570096731185913,
      "learning_rate": 0.00019884864541838432,
      "loss": 3.2467,
      "mean_token_accuracy": 0.4176161937117577,
      "num_tokens": 6451200000.0,
      "step": 720000
    },
    {
      "epoch": 0.35335014134005654,
      "grad_norm": 1.5831801891326904,
      "learning_rate": 0.00019595470402937928,
      "loss": 3.2441,
      "mean_token_accuracy": 0.41793804090470077,
      "num_tokens": 7168000000.0,
      "step": 740000
    },
    {
      "epoch": 0.3629001451600581,
      "grad_norm": 1.6359102725982666,
      "learning_rate": 0.00019306076264037422,
      "loss": 3.2439,
      "mean_token_accuracy": 0.41790355779081584,
      "num_tokens": 7884800000.0,
      "step": 760000
    },
    {
      "epoch": 0.37245014898005957,
      "grad_norm": 1.9125442504882812,
      "learning_rate": 0.00019016682125136918,
      "loss": 3.2416,
      "mean_token_accuracy": 0.4181554499194026,
      "num_tokens": 8601600000.0,
      "step": 780000
    },
    {
      "epoch": 0.3820001528000611,
      "grad_norm": 1.6356027126312256,
      "learning_rate": 0.0001872728798623641,
      "loss": 3.2392,
      "mean_token_accuracy": 0.41842390780746935,
      "num_tokens": 9318400000.0,
      "step": 800000
    },
    {
      "epoch": 0.39155015662006265,
      "grad_norm": 1.69579017162323,
      "learning_rate": 0.00018437893847335907,
      "loss": 3.2381,
      "mean_token_accuracy": 0.41857809690237047,
      "num_tokens": 10035200000.0,
      "step": 820000
    },
    {
      "epoch": 0.4011001604400642,
      "grad_norm": 1.7878586053848267,
      "learning_rate": 0.00018148499708435403,
      "loss": 3.2375,
      "mean_token_accuracy": 0.41865148201435803,
      "num_tokens": 10752000000.0,
      "step": 840000
    },
    {
      "epoch": 0.41065016426006573,
      "grad_norm": 1.7767938375473022,
      "learning_rate": 0.00017859105569534897,
      "loss": 3.2353,
      "mean_token_accuracy": 0.41889262205660344,
      "num_tokens": 11468800000.0,
      "step": 860000
    },
    {
      "epoch": 0.4202001680800672,
      "grad_norm": 1.6678400039672852,
      "learning_rate": 0.00017569711430634393,
      "loss": 3.2337,
      "mean_token_accuracy": 0.41908056329786775,
      "num_tokens": 12185600000.0,
      "step": 880000
    },
    {
      "epoch": 0.42975017190006876,
      "grad_norm": 1.862349033355713,
      "learning_rate": 0.00017280317291733891,
      "loss": 3.2329,
      "mean_token_accuracy": 0.4191736038953066,
      "num_tokens": 12902400000.0,
      "step": 900000
    },
    {
      "epoch": 0.4393001757200703,
      "grad_norm": 1.8526560068130493,
      "learning_rate": 0.00016990923152833385,
      "loss": 3.232,
      "mean_token_accuracy": 0.4192189232364297,
      "num_tokens": 13619200000.0,
      "step": 920000
    },
    {
      "epoch": 0.44885017954007184,
      "grad_norm": 1.7189236879348755,
      "learning_rate": 0.0001670152901393288,
      "loss": 3.2302,
      "mean_token_accuracy": 0.4194715451017022,
      "num_tokens": 14336000000.0,
      "step": 940000
    },
    {
      "epoch": 0.4584001833600733,
      "grad_norm": 2.0065693855285645,
      "learning_rate": 0.00016412134875032377,
      "loss": 3.2277,
      "mean_token_accuracy": 0.419769259378314,
      "num_tokens": 15052800000.0,
      "step": 960000
    },
    {
      "epoch": 0.46795018718007486,
      "grad_norm": 1.647645354270935,
      "learning_rate": 0.0001612274073613187,
      "loss": 3.2272,
      "mean_token_accuracy": 0.41980642325282097,
      "num_tokens": 15769600000.0,
      "step": 980000
    },
    {
      "epoch": 0.4775001910000764,
      "grad_norm": 1.8431377410888672,
      "learning_rate": 0.00015833346597231366,
      "loss": 3.2244,
      "mean_token_accuracy": 0.42018424404114485,
      "num_tokens": 16486400000.0,
      "step": 1000000
    },
    {
      "epoch": 0.48705019482007794,
      "grad_norm": 1.8152481317520142,
      "learning_rate": 0.0001554395245833086,
      "loss": 3.2239,
      "mean_token_accuracy": 0.42016951968967914,
      "num_tokens": 17203200000.0,
      "step": 1020000
    },
    {
      "epoch": 0.4966001986400795,
      "grad_norm": 1.979134440422058,
      "learning_rate": 0.00015254558319430356,
      "loss": 3.2226,
      "mean_token_accuracy": 0.42036232096105813,
      "num_tokens": 17920000000.0,
      "step": 1040000
    },
    {
      "epoch": 0.506150202460081,
      "grad_norm": 2.2059521675109863,
      "learning_rate": 0.00014965164180529852,
      "loss": 3.2199,
      "mean_token_accuracy": 0.4206514175161719,
      "num_tokens": 18636800000.0,
      "step": 1060000
    },
    {
      "epoch": 0.5157002062800825,
      "grad_norm": 2.0843093395233154,
      "learning_rate": 0.00014675770041629345,
      "loss": 3.2192,
      "mean_token_accuracy": 0.42072975924313066,
      "num_tokens": 19353600000.0,
      "step": 1080000
    },
    {
      "epoch": 0.525250210100084,
      "grad_norm": 1.7924293279647827,
      "learning_rate": 0.0001438637590272884,
      "loss": 3.2178,
      "mean_token_accuracy": 0.4209353769227862,
      "num_tokens": 20070400000.0,
      "step": 1100000
    },
    {
      "epoch": 0.5348002139200856,
      "grad_norm": 1.8186842203140259,
      "learning_rate": 0.00014096981763828334,
      "loss": 3.2166,
      "mean_token_accuracy": 0.42104669906646014,
      "num_tokens": 20787200000.0,
      "step": 1120000
    },
    {
      "epoch": 0.5443502177400871,
      "grad_norm": 2.0781736373901367,
      "learning_rate": 0.0001380758762492783,
      "loss": 3.2147,
      "mean_token_accuracy": 0.42130602815449236,
      "num_tokens": 21504000000.0,
      "step": 1140000
    },
    {
      "epoch": 0.5539002215600887,
      "grad_norm": 2.1371679306030273,
      "learning_rate": 0.00013518193486027327,
      "loss": 3.2135,
      "mean_token_accuracy": 0.421437505723536,
      "num_tokens": 22220800000.0,
      "step": 1160000
    },
    {
      "epoch": 0.5634502253800902,
      "grad_norm": 1.9272387027740479,
      "learning_rate": 0.00013228799347126823,
      "loss": 3.2129,
      "mean_token_accuracy": 0.42148565844893454,
      "num_tokens": 22937600000.0,
      "step": 1180000
    },
    {
      "epoch": 0.5730002292000916,
      "grad_norm": 1.975583553314209,
      "learning_rate": 0.00012939405208226316,
      "loss": 3.2101,
      "mean_token_accuracy": 0.42179442739486694,
      "num_tokens": 23654400000.0,
      "step": 1200000
    },
    {
      "epoch": 0.5825502330200932,
      "grad_norm": 1.8845446109771729,
      "learning_rate": 0.00012650011069325812,
      "loss": 3.209,
      "mean_token_accuracy": 0.42191650020480154,
      "num_tokens": 24371200000.0,
      "step": 1220000
    },
    {
      "epoch": 0.5921002368400947,
      "grad_norm": 1.8266512155532837,
      "learning_rate": 0.00012360616930425308,
      "loss": 3.2075,
      "mean_token_accuracy": 0.4220881850525737,
      "num_tokens": 25088000000.0,
      "step": 1240000
    },
    {
      "epoch": 0.6016502406600963,
      "grad_norm": 1.9238523244857788,
      "learning_rate": 0.00012071222791524803,
      "loss": 3.2051,
      "mean_token_accuracy": 0.42236345537304876,
      "num_tokens": 25804800000.0,
      "step": 1260000
    },
    {
      "epoch": 0.6112002444800978,
      "grad_norm": 1.8870298862457275,
      "learning_rate": 0.00011781828652624297,
      "loss": 3.2044,
      "mean_token_accuracy": 0.4224274314776063,
      "num_tokens": 26521600000.0,
      "step": 1280000
    },
    {
      "epoch": 0.6207502483000993,
      "grad_norm": 2.10017991065979,
      "learning_rate": 0.00011492434513723792,
      "loss": 3.2038,
      "mean_token_accuracy": 0.4225288334354758,
      "num_tokens": 27238400000.0,
      "step": 1300000
    },
    {
      "epoch": 0.6303002521201009,
      "grad_norm": 2.2103898525238037,
      "learning_rate": 0.00011203040374823287,
      "loss": 3.2006,
      "mean_token_accuracy": 0.42296220604628326,
      "num_tokens": 27955200000.0,
      "step": 1320000
    },
    {
      "epoch": 0.6398502559401024,
      "grad_norm": 2.1420845985412598,
      "learning_rate": 0.00010913646235922783,
      "loss": 3.1995,
      "mean_token_accuracy": 0.42305554908663034,
      "num_tokens": 28672000000.0,
      "step": 1340000
    },
    {
      "epoch": 0.649400259760104,
      "grad_norm": 2.468029737472534,
      "learning_rate": 0.00010624252097022279,
      "loss": 3.1994,
      "mean_token_accuracy": 0.4230162718191743,
      "num_tokens": 29388800000.0,
      "step": 1360000
    },
    {
      "epoch": 0.6589502635801054,
      "grad_norm": 2.0718960762023926,
      "learning_rate": 0.00010334857958121774,
      "loss": 3.1964,
      "mean_token_accuracy": 0.4234230771496892,
      "num_tokens": 30105600000.0,
      "step": 1380000
    },
    {
      "epoch": 0.6685002674001069,
      "grad_norm": 2.7991926670074463,
      "learning_rate": 0.00010045463819221268,
      "loss": 3.1943,
      "mean_token_accuracy": 0.4237000042423606,
      "num_tokens": 30822400000.0,
      "step": 1400000
    },
    {
      "epoch": 0.6780502712201085,
      "grad_norm": 2.4362847805023193,
      "learning_rate": 9.756069680320764e-05,
      "loss": 3.1922,
      "mean_token_accuracy": 0.4239600421443582,
      "num_tokens": 31539200000.0,
      "step": 1420000
    },
    {
      "epoch": 0.68760027504011,
      "grad_norm": 2.591127872467041,
      "learning_rate": 9.466675541420259e-05,
      "loss": 3.191,
      "mean_token_accuracy": 0.4240850882053375,
      "num_tokens": 32256000000.0,
      "step": 1440000
    },
    {
      "epoch": 0.6971502788601115,
      "grad_norm": 2.234511375427246,
      "learning_rate": 9.177281402519754e-05,
      "loss": 3.1897,
      "mean_token_accuracy": 0.4242406051442027,
      "num_tokens": 32972800000.0,
      "step": 1460000
    },
    {
      "epoch": 0.7067002826801131,
      "grad_norm": 2.165247917175293,
      "learning_rate": 8.887887263619248e-05,
      "loss": 3.1876,
      "mean_token_accuracy": 0.42447070305049417,
      "num_tokens": 33689600000.0,
      "step": 1480000
    },
    {
      "epoch": 0.7162502865001146,
      "grad_norm": 2.373894453048706,
      "learning_rate": 8.598493124718745e-05,
      "loss": 3.1858,
      "mean_token_accuracy": 0.4246869832932949,
      "num_tokens": 34406400000.0,
      "step": 1500000
    },
    {
      "epoch": 0.7258002903201162,
      "grad_norm": 2.232844352722168,
      "learning_rate": 8.309098985818239e-05,
      "loss": 3.1842,
      "mean_token_accuracy": 0.424876312276721,
      "num_tokens": 35123200000.0,
      "step": 1520000
    },
    {
      "epoch": 0.7353502941401177,
      "grad_norm": 2.4994523525238037,
      "learning_rate": 8.019704846917734e-05,
      "loss": 3.1827,
      "mean_token_accuracy": 0.42511233622431754,
      "num_tokens": 35840000000.0,
      "step": 1540000
    },
    {
      "epoch": 0.7449002979601191,
      "grad_norm": 2.4633235931396484,
      "learning_rate": 7.73031070801723e-05,
      "loss": 3.1798,
      "mean_token_accuracy": 0.4254420336738229,
      "num_tokens": 36556800000.0,
      "step": 1560000
    },
    {
      "epoch": 0.7544503017801207,
      "grad_norm": 2.386373281478882,
      "learning_rate": 7.440916569116725e-05,
      "loss": 3.1786,
      "mean_token_accuracy": 0.4255216441363096,
      "num_tokens": 37273600000.0,
      "step": 1580000
    },
    {
      "epoch": 0.7640003056001222,
      "grad_norm": 2.4472737312316895,
      "learning_rate": 7.151522430216221e-05,
      "loss": 3.1767,
      "mean_token_accuracy": 0.4258053430899978,
      "num_tokens": 37990400000.0,
      "step": 1600000
    },
    {
      "epoch": 0.7735503094201238,
      "grad_norm": 2.5622825622558594,
      "learning_rate": 6.862128291315715e-05,
      "loss": 3.1746,
      "mean_token_accuracy": 0.4260344804123044,
      "num_tokens": 38707200000.0,
      "step": 1620000
    },
    {
      "epoch": 0.7831003132401253,
      "grad_norm": 2.4821906089782715,
      "learning_rate": 6.57273415241521e-05,
      "loss": 3.1723,
      "mean_token_accuracy": 0.4263139396473765,
      "num_tokens": 39424000000.0,
      "step": 1640000
    },
    {
      "epoch": 0.7926503170601268,
      "grad_norm": 2.5140678882598877,
      "learning_rate": 6.283340013514706e-05,
      "loss": 3.1701,
      "mean_token_accuracy": 0.42658785569369795,
      "num_tokens": 40140800000.0,
      "step": 1660000
    },
    {
      "epoch": 0.8022003208801284,
      "grad_norm": 2.634791851043701,
      "learning_rate": 5.9939458746142e-05,
      "loss": 3.1678,
      "mean_token_accuracy": 0.42689967503100634,
      "num_tokens": 40857600000.0,
      "step": 1680000
    },
    {
      "epoch": 0.8117503247001299,
      "grad_norm": 3.0459752082824707,
      "learning_rate": 5.704551735713696e-05,
      "loss": 3.1672,
      "mean_token_accuracy": 0.42699285422861577,
      "num_tokens": 41574400000.0,
      "step": 1700000
    },
    {
      "epoch": 0.8213003285201315,
      "grad_norm": 2.686730146408081,
      "learning_rate": 5.4151575968131916e-05,
      "loss": 3.1642,
      "mean_token_accuracy": 0.42733070581257343,
      "num_tokens": 42291200000.0,
      "step": 1720000
    },
    {
      "epoch": 0.830850332340133,
      "grad_norm": 2.967567205429077,
      "learning_rate": 5.125763457912686e-05,
      "loss": 3.1618,
      "mean_token_accuracy": 0.42769208399355413,
      "num_tokens": 43008000000.0,
      "step": 1740000
    },
    {
      "epoch": 0.8404003361601344,
      "grad_norm": 2.7259535789489746,
      "learning_rate": 4.836369319012181e-05,
      "loss": 3.1596,
      "mean_token_accuracy": 0.42790325200855733,
      "num_tokens": 43724800000.0,
      "step": 1760000
    },
    {
      "epoch": 0.849950339980136,
      "grad_norm": 2.709784746170044,
      "learning_rate": 4.546975180111677e-05,
      "loss": 3.1567,
      "mean_token_accuracy": 0.42829815539866684,
      "num_tokens": 44441600000.0,
      "step": 1780000
    },
    {
      "epoch": 0.8595003438001375,
      "grad_norm": 3.054241418838501,
      "learning_rate": 4.2575810412111724e-05,
      "loss": 3.1551,
      "mean_token_accuracy": 0.42849865667670967,
      "num_tokens": 45158400000.0,
      "step": 1800000
    },
    {
      "epoch": 0.869050347620139,
      "grad_norm": 2.976240873336792,
      "learning_rate": 3.968186902310667e-05,
      "loss": 3.1519,
      "mean_token_accuracy": 0.42887963571995497,
      "num_tokens": 45875200000.0,
      "step": 1820000
    },
    {
      "epoch": 0.8786003514401406,
      "grad_norm": 2.723175525665283,
      "learning_rate": 3.678792763410162e-05,
      "loss": 3.1498,
      "mean_token_accuracy": 0.42917205161750316,
      "num_tokens": 46592000000.0,
      "step": 1840000
    },
    {
      "epoch": 0.8881503552601421,
      "grad_norm": 2.917259693145752,
      "learning_rate": 3.389398624509658e-05,
      "loss": 3.146,
      "mean_token_accuracy": 0.42966016797572376,
      "num_tokens": 47308800000.0,
      "step": 1860000
    },
    {
      "epoch": 0.8977003590801437,
      "grad_norm": 3.254523515701294,
      "learning_rate": 3.1000044856091526e-05,
      "loss": 3.1442,
      "mean_token_accuracy": 0.42988122535943984,
      "num_tokens": 48025600000.0,
      "step": 1880000
    },
    {
      "epoch": 0.9072503629001452,
      "grad_norm": 2.9410157203674316,
      "learning_rate": 2.810610346708648e-05,
      "loss": 3.1408,
      "mean_token_accuracy": 0.43033207250982525,
      "num_tokens": 48742400000.0,
      "step": 1900000
    },
    {
      "epoch": 0.9168003667201466,
      "grad_norm": 3.0534310340881348,
      "learning_rate": 2.521216207808143e-05,
      "loss": 3.1383,
      "mean_token_accuracy": 0.43062722102552653,
      "num_tokens": 49459200000.0,
      "step": 1920000
    },
    {
      "epoch": 0.9263503705401482,
      "grad_norm": 3.0121352672576904,
      "learning_rate": 2.2318220689076384e-05,
      "loss": 3.1352,
      "mean_token_accuracy": 0.43105802639722823,
      "num_tokens": 50176000000.0,
      "step": 1940000
    },
    {
      "epoch": 0.9359003743601497,
      "grad_norm": 3.4966187477111816,
      "learning_rate": 1.9424279300071334e-05,
      "loss": 3.1317,
      "mean_token_accuracy": 0.43149784843176603,
      "num_tokens": 50892800000.0,
      "step": 1960000
    },
    {
      "epoch": 0.9454503781801513,
      "grad_norm": 3.3833205699920654,
      "learning_rate": 1.6530337911066284e-05,
      "loss": 3.1285,
      "mean_token_accuracy": 0.4319396187588572,
      "num_tokens": 51609600000.0,
      "step": 1980000
    },
    {
      "epoch": 0.9550003820001528,
      "grad_norm": 3.30391001701355,
      "learning_rate": 1.3636396522061238e-05,
      "loss": 3.1248,
      "mean_token_accuracy": 0.43242690176963805,
      "num_tokens": 52326400000.0,
      "step": 2000000
    },
    {
      "epoch": 0.9645503858201543,
      "grad_norm": 3.326011896133423,
      "learning_rate": 1.074245513305619e-05,
      "loss": 3.1213,
      "mean_token_accuracy": 0.4328598498493433,
      "num_tokens": 53043200000.0,
      "step": 2020000
    },
    {
      "epoch": 0.9741003896401559,
      "grad_norm": 3.2196035385131836,
      "learning_rate": 7.84851374405114e-06,
      "loss": 3.1176,
      "mean_token_accuracy": 0.4334187435388565,
      "num_tokens": 53760000000.0,
      "step": 2040000
    },
    {
      "epoch": 0.9836503934601574,
      "grad_norm": 3.748178720474243,
      "learning_rate": 4.9545723550460936e-06,
      "loss": 3.1139,
      "mean_token_accuracy": 0.43390719720572235,
      "num_tokens": 54476800000.0,
      "step": 2060000
    },
    {
      "epoch": 0.993200397280159,
      "grad_norm": 3.283731460571289,
      "learning_rate": 2.0606309660410444e-06,
      "loss": 3.1103,
      "mean_token_accuracy": 0.4344122279688716,
      "num_tokens": 55193600000.0,
      "step": 2080000
    }
  ],
  "logging_steps": 20000,
  "max_steps": 2094240,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 20000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1815907720024556e+18,
  "train_batch_size": 140,
  "trial_name": null,
  "trial_params": null
}