{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 867,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005768676088837612,
      "grad_norm": 45.75,
      "learning_rate": 2.222222222222222e-06,
      "loss": 2.812992477416992,
      "step": 5
    },
    {
      "epoch": 0.011537352177675224,
      "grad_norm": 19.625,
      "learning_rate": 4.9999999999999996e-06,
      "loss": 2.6157678604125976,
      "step": 10
    },
    {
      "epoch": 0.017306028266512834,
      "grad_norm": 10.0,
      "learning_rate": 7.777777777777777e-06,
      "loss": 2.550214958190918,
      "step": 15
    },
    {
      "epoch": 0.02307470435535045,
      "grad_norm": 6.5625,
      "learning_rate": 1.0555555555555555e-05,
      "loss": 2.2421979904174805,
      "step": 20
    },
    {
      "epoch": 0.02884338044418806,
      "grad_norm": 12.125,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.3474178314208984,
      "step": 25
    },
    {
      "epoch": 0.03461205653302567,
      "grad_norm": 3.703125,
      "learning_rate": 1.4999790187959524e-05,
      "loss": 2.0521663665771483,
      "step": 30
    },
    {
      "epoch": 0.040380732621863286,
      "grad_norm": 3.09375,
      "learning_rate": 1.4997429937316679e-05,
      "loss": 1.9066032409667968,
      "step": 35
    },
    {
      "epoch": 0.0461494087107009,
      "grad_norm": 3.34375,
      "learning_rate": 1.4992447999059861e-05,
      "loss": 1.8443662643432617,
      "step": 40
    },
    {
      "epoch": 0.05191808479953851,
      "grad_norm": 4.125,
      "learning_rate": 1.498484611526414e-05,
      "loss": 1.8280458450317383,
      "step": 45
    },
    {
      "epoch": 0.05768676088837612,
      "grad_norm": 3.234375,
      "learning_rate": 1.4974626944142366e-05,
      "loss": 1.743063735961914,
      "step": 50
    },
    {
      "epoch": 0.06345543697721373,
      "grad_norm": 4.4375,
      "learning_rate": 1.4961794059115642e-05,
      "loss": 1.6797224044799806,
      "step": 55
    },
    {
      "epoch": 0.06922411306605133,
      "grad_norm": 3.078125,
      "learning_rate": 1.494635194756378e-05,
      "loss": 1.687392807006836,
      "step": 60
    },
    {
      "epoch": 0.07499278915488895,
      "grad_norm": 2.59375,
      "learning_rate": 1.4928306009256156e-05,
      "loss": 1.6969615936279296,
      "step": 65
    },
    {
      "epoch": 0.08076146524372657,
      "grad_norm": 2.453125,
      "learning_rate": 1.4907662554463534e-05,
      "loss": 1.7176189422607422,
      "step": 70
    },
    {
      "epoch": 0.08653014133256418,
      "grad_norm": 4.5625,
      "learning_rate": 1.4884428801751502e-05,
      "loss": 1.5526809692382812,
      "step": 75
    },
    {
      "epoch": 0.0922988174214018,
      "grad_norm": 3.796875,
      "learning_rate": 1.4858612875456295e-05,
      "loss": 1.623830795288086,
      "step": 80
    },
    {
      "epoch": 0.0980674935102394,
      "grad_norm": 2.984375,
      "learning_rate": 1.4830223802843894e-05,
      "loss": 1.6493770599365234,
      "step": 85
    },
    {
      "epoch": 0.10383616959907702,
      "grad_norm": 3.46875,
      "learning_rate": 1.4799271510953386e-05,
      "loss": 1.6132541656494142,
      "step": 90
    },
    {
      "epoch": 0.10960484568791462,
      "grad_norm": 5.0625,
      "learning_rate": 1.4765766823125695e-05,
      "loss": 1.7179641723632812,
      "step": 95
    },
    {
      "epoch": 0.11537352177675224,
      "grad_norm": 2.46875,
      "learning_rate": 1.4729721455218899e-05,
      "loss": 1.5699467658996582,
      "step": 100
    },
    {
      "epoch": 0.12114219786558984,
      "grad_norm": 4.03125,
      "learning_rate": 1.4691148011511447e-05,
      "loss": 1.5707356452941894,
      "step": 105
    },
    {
      "epoch": 0.12691087395442746,
      "grad_norm": 2.671875,
      "learning_rate": 1.465005998029472e-05,
      "loss": 1.581485652923584,
      "step": 110
    },
    {
      "epoch": 0.13267955004326507,
      "grad_norm": 4.0,
      "learning_rate": 1.4606471729156468e-05,
      "loss": 1.4936203956604004,
      "step": 115
    },
    {
      "epoch": 0.13844822613210267,
      "grad_norm": 2.640625,
      "learning_rate": 1.4560398499956776e-05,
      "loss": 1.5083925247192382,
      "step": 120
    },
    {
      "epoch": 0.1442169022209403,
      "grad_norm": 2.3125,
      "learning_rate": 1.4511856403498324e-05,
      "loss": 1.4727232933044434,
      "step": 125
    },
    {
      "epoch": 0.1499855783097779,
      "grad_norm": 3.65625,
      "learning_rate": 1.4460862413892792e-05,
      "loss": 1.5390252113342284,
      "step": 130
    },
    {
      "epoch": 0.1557542543986155,
      "grad_norm": 2.671875,
      "learning_rate": 1.4407434362625376e-05,
      "loss": 1.5905132293701172,
      "step": 135
    },
    {
      "epoch": 0.16152293048745314,
      "grad_norm": 3.375,
      "learning_rate": 1.4351590932319506e-05,
      "loss": 1.5457244873046876,
      "step": 140
    },
    {
      "epoch": 0.16729160657629075,
      "grad_norm": 3.4375,
      "learning_rate": 1.4293351650203956e-05,
      "loss": 1.5160816192626954,
      "step": 145
    },
    {
      "epoch": 0.17306028266512835,
      "grad_norm": 2.5625,
      "learning_rate": 1.4232736881284576e-05,
      "loss": 1.5460505485534668,
      "step": 150
    },
    {
      "epoch": 0.17882895875396596,
      "grad_norm": 2.734375,
      "learning_rate": 1.416976782122311e-05,
      "loss": 1.4950276374816895,
      "step": 155
    },
    {
      "epoch": 0.1845976348428036,
      "grad_norm": 4.25,
      "learning_rate": 1.4104466488925536e-05,
      "loss": 1.4827272415161132,
      "step": 160
    },
    {
      "epoch": 0.1903663109316412,
      "grad_norm": 2.859375,
      "learning_rate": 1.4036855718842518e-05,
      "loss": 1.585806179046631,
      "step": 165
    },
    {
      "epoch": 0.1961349870204788,
      "grad_norm": 3.734375,
      "learning_rate": 1.396695915298472e-05,
      "loss": 1.4454821586608886,
      "step": 170
    },
    {
      "epoch": 0.2019036631093164,
      "grad_norm": 2.5,
      "learning_rate": 1.3894801232655692e-05,
      "loss": 1.5184972763061524,
      "step": 175
    },
    {
      "epoch": 0.20767233919815403,
      "grad_norm": 2.5625,
      "learning_rate": 1.3820407189905295e-05,
      "loss": 1.4702855110168458,
      "step": 180
    },
    {
      "epoch": 0.21344101528699164,
      "grad_norm": 4.25,
      "learning_rate": 1.3743803038706587e-05,
      "loss": 1.4560264587402343,
      "step": 185
    },
    {
      "epoch": 0.21920969137582924,
      "grad_norm": 4.71875,
      "learning_rate": 1.3665015565859325e-05,
      "loss": 1.4439519882202148,
      "step": 190
    },
    {
      "epoch": 0.22497836746466685,
      "grad_norm": 3.640625,
      "learning_rate": 1.3584072321623183e-05,
      "loss": 1.3752272605895997,
      "step": 195
    },
    {
      "epoch": 0.23074704355350448,
      "grad_norm": 4.8125,
      "learning_rate": 1.3501001610084048e-05,
      "loss": 1.5652609825134278,
      "step": 200
    },
    {
      "epoch": 0.23651571964234208,
      "grad_norm": 2.859375,
      "learning_rate": 1.3415832479256685e-05,
      "loss": 1.4904936790466308,
      "step": 205
    },
    {
      "epoch": 0.24228439573117969,
      "grad_norm": 3.859375,
      "learning_rate": 1.3328594710927282e-05,
      "loss": 1.4062389373779296,
      "step": 210
    },
    {
      "epoch": 0.24805307182001732,
      "grad_norm": 3.09375,
      "learning_rate": 1.3239318810239423e-05,
      "loss": 1.5080370903015137,
      "step": 215
    },
    {
      "epoch": 0.2538217479088549,
      "grad_norm": 4.0625,
      "learning_rate": 1.3148035995027083e-05,
      "loss": 1.403939914703369,
      "step": 220
    },
    {
      "epoch": 0.25959042399769255,
      "grad_norm": 2.453125,
      "learning_rate": 1.3054778184898452e-05,
      "loss": 1.460791778564453,
      "step": 225
    },
    {
      "epoch": 0.26535910008653013,
      "grad_norm": 3.234375,
      "learning_rate": 1.2959577990074333e-05,
      "loss": 1.4795469284057616,
      "step": 230
    },
    {
      "epoch": 0.27112777617536776,
      "grad_norm": 3.8125,
      "learning_rate": 1.2862468699985066e-05,
      "loss": 1.4639040946960449,
      "step": 235
    },
    {
      "epoch": 0.27689645226420534,
      "grad_norm": 2.890625,
      "learning_rate": 1.2763484271629938e-05,
      "loss": 1.4583266258239747,
      "step": 240
    },
    {
      "epoch": 0.28266512835304297,
      "grad_norm": 3.125,
      "learning_rate": 1.2662659317703154e-05,
      "loss": 1.4482952117919923,
      "step": 245
    },
    {
      "epoch": 0.2884338044418806,
      "grad_norm": 3.15625,
      "learning_rate": 1.256002909449056e-05,
      "loss": 1.4816177368164063,
      "step": 250
    },
    {
      "epoch": 0.2942024805307182,
      "grad_norm": 3.09375,
      "learning_rate": 1.2455629489541248e-05,
      "loss": 1.6563011169433595,
      "step": 255
    },
    {
      "epoch": 0.2999711566195558,
      "grad_norm": 3.5,
      "learning_rate": 1.2349497009118497e-05,
      "loss": 1.45164794921875,
      "step": 260
    },
    {
      "epoch": 0.30573983270839344,
      "grad_norm": 2.3125,
      "learning_rate": 1.2241668765434324e-05,
      "loss": 1.5071072578430176,
      "step": 265
    },
    {
      "epoch": 0.311508508797231,
      "grad_norm": 3.0,
      "learning_rate": 1.2132182463672135e-05,
      "loss": 1.4920737266540527,
      "step": 270
    },
    {
      "epoch": 0.31727718488606865,
      "grad_norm": 2.703125,
      "learning_rate": 1.2021076388802062e-05,
      "loss": 1.4359200477600098,
      "step": 275
    },
    {
      "epoch": 0.3230458609749063,
      "grad_norm": 3.96875,
      "learning_rate": 1.1908389392193549e-05,
      "loss": 1.4329303741455077,
      "step": 280
    },
    {
      "epoch": 0.32881453706374386,
      "grad_norm": 3.453125,
      "learning_rate": 1.1794160878029856e-05,
      "loss": 1.3877481460571288,
      "step": 285
    },
    {
      "epoch": 0.3345832131525815,
      "grad_norm": 2.40625,
      "learning_rate": 1.1678430789529292e-05,
      "loss": 1.4761417388916016,
      "step": 290
    },
    {
      "epoch": 0.34035188924141907,
      "grad_norm": 3.46875,
      "learning_rate": 1.1561239594977943e-05,
      "loss": 1.5856515884399414,
      "step": 295
    },
    {
      "epoch": 0.3461205653302567,
      "grad_norm": 3.015625,
      "learning_rate": 1.1442628273578785e-05,
      "loss": 1.3921869277954102,
      "step": 300
    },
    {
      "epoch": 0.35188924141909433,
      "grad_norm": 2.6875,
      "learning_rate": 1.1322638301122164e-05,
      "loss": 1.3925346374511718,
      "step": 305
    },
    {
      "epoch": 0.3576579175079319,
      "grad_norm": 2.96875,
      "learning_rate": 1.1201311635482602e-05,
      "loss": 1.3757241249084473,
      "step": 310
    },
    {
      "epoch": 0.36342659359676954,
      "grad_norm": 3.203125,
      "learning_rate": 1.1078690701947063e-05,
      "loss": 1.4508034706115722,
      "step": 315
    },
    {
      "epoch": 0.3691952696856072,
      "grad_norm": 2.546875,
      "learning_rate": 1.0954818378379743e-05,
      "loss": 1.4826339721679687,
      "step": 320
    },
    {
      "epoch": 0.37496394577444475,
      "grad_norm": 3.0,
      "learning_rate": 1.0829737980228618e-05,
      "loss": 1.41096773147583,
      "step": 325
    },
    {
      "epoch": 0.3807326218632824,
      "grad_norm": 2.78125,
      "learning_rate": 1.0703493245378976e-05,
      "loss": 1.502821445465088,
      "step": 330
    },
    {
      "epoch": 0.38650129795212,
      "grad_norm": 2.8125,
      "learning_rate": 1.0576128318859246e-05,
      "loss": 1.3927604675292968,
      "step": 335
    },
    {
      "epoch": 0.3922699740409576,
      "grad_norm": 3.625,
      "learning_rate": 1.0447687737404428e-05,
      "loss": 1.3787881851196289,
      "step": 340
    },
    {
      "epoch": 0.3980386501297952,
      "grad_norm": 3.1875,
      "learning_rate": 1.0318216413882576e-05,
      "loss": 1.3673920631408691,
      "step": 345
    },
    {
      "epoch": 0.4038073262186328,
      "grad_norm": 4.25,
      "learning_rate": 1.018775962158975e-05,
      "loss": 1.3431027412414551,
      "step": 350
    },
    {
      "epoch": 0.40957600230747043,
      "grad_norm": 2.1875,
      "learning_rate": 1.0056362978418937e-05,
      "loss": 1.483182430267334,
      "step": 355
    },
    {
      "epoch": 0.41534467839630806,
      "grad_norm": 4.25,
      "learning_rate": 9.924072430908451e-06,
      "loss": 1.4253923416137695,
      "step": 360
    },
    {
      "epoch": 0.42111335448514564,
      "grad_norm": 2.65625,
      "learning_rate": 9.79093423817545e-06,
      "loss": 1.436702823638916,
      "step": 365
    },
    {
      "epoch": 0.4268820305739833,
      "grad_norm": 2.640625,
      "learning_rate": 9.656994955740129e-06,
      "loss": 1.4487831115722656,
      "step": 370
    },
    {
      "epoch": 0.4326507066628209,
      "grad_norm": 2.4375,
      "learning_rate": 9.522301419246267e-06,
      "loss": 1.3315213203430176,
      "step": 375
    },
    {
      "epoch": 0.4384193827516585,
      "grad_norm": 2.546875,
      "learning_rate": 9.386900728083816e-06,
      "loss": 1.464303684234619,
      "step": 380
    },
    {
      "epoch": 0.4441880588404961,
      "grad_norm": 4.03125,
      "learning_rate": 9.250840228919291e-06,
      "loss": 1.3498079299926757,
      "step": 385
    },
    {
      "epoch": 0.4499567349293337,
      "grad_norm": 2.25,
      "learning_rate": 9.114167499139666e-06,
      "loss": 1.444554042816162,
      "step": 390
    },
    {
      "epoch": 0.4557254110181713,
      "grad_norm": 4.21875,
      "learning_rate": 8.976930330215592e-06,
      "loss": 1.311547088623047,
      "step": 395
    },
    {
      "epoch": 0.46149408710700895,
      "grad_norm": 2.84375,
      "learning_rate": 8.839176710989775e-06,
      "loss": 1.423337173461914,
      "step": 400
    },
    {
      "epoch": 0.46726276319584653,
      "grad_norm": 2.828125,
      "learning_rate": 8.700954810896315e-06,
      "loss": 1.4692988395690918,
      "step": 405
    },
    {
      "epoch": 0.47303143928468416,
      "grad_norm": 4.28125,
      "learning_rate": 8.562312963116918e-06,
      "loss": 1.416618251800537,
      "step": 410
    },
    {
      "epoch": 0.4788001153735218,
      "grad_norm": 2.828125,
      "learning_rate": 8.423299647679823e-06,
      "loss": 1.3839403152465821,
      "step": 415
    },
    {
      "epoch": 0.48456879146235937,
      "grad_norm": 3.609375,
      "learning_rate": 8.283963474507402e-06,
      "loss": 1.4335898399353026,
      "step": 420
    },
    {
      "epoch": 0.490337467551197,
      "grad_norm": 2.53125,
      "learning_rate": 8.144353166418336e-06,
      "loss": 1.3607335090637207,
      "step": 425
    },
    {
      "epoch": 0.49610614364003464,
      "grad_norm": 2.90625,
      "learning_rate": 8.004517542090298e-06,
      "loss": 1.3808890342712403,
      "step": 430
    },
    {
      "epoch": 0.5018748197288723,
      "grad_norm": 4.25,
      "learning_rate": 7.864505498989136e-06,
      "loss": 1.339999294281006,
      "step": 435
    },
    {
      "epoch": 0.5076434958177098,
      "grad_norm": 2.484375,
      "learning_rate": 7.724365996270489e-06,
      "loss": 1.4851154327392577,
      "step": 440
    },
    {
      "epoch": 0.5134121719065474,
      "grad_norm": 2.921875,
      "learning_rate": 7.584148037659845e-06,
      "loss": 1.4813553810119628,
      "step": 445
    },
    {
      "epoch": 0.5191808479953851,
      "grad_norm": 2.90625,
      "learning_rate": 7.443900654317003e-06,
      "loss": 1.5473807334899903,
      "step": 450
    },
    {
      "epoch": 0.5249495240842227,
      "grad_norm": 3.046875,
      "learning_rate": 7.303672887690953e-06,
      "loss": 1.4310360908508302,
      "step": 455
    },
    {
      "epoch": 0.5307182001730603,
      "grad_norm": 3.6875,
      "learning_rate": 7.163513772371141e-06,
      "loss": 1.4008095741271973,
      "step": 460
    },
    {
      "epoch": 0.536486876261898,
      "grad_norm": 2.265625,
      "learning_rate": 7.02347231894115e-06,
      "loss": 1.476534080505371,
      "step": 465
    },
    {
      "epoch": 0.5422555523507355,
      "grad_norm": 2.671875,
      "learning_rate": 6.883597496840774e-06,
      "loss": 1.4958248138427734,
      "step": 470
    },
    {
      "epoch": 0.5480242284395731,
      "grad_norm": 2.640625,
      "learning_rate": 6.7439382172424665e-06,
      "loss": 1.4489681243896484,
      "step": 475
    },
    {
      "epoch": 0.5537929045284107,
      "grad_norm": 2.625,
      "learning_rate": 6.604543315948169e-06,
      "loss": 1.3916007995605468,
      "step": 480
    },
    {
      "epoch": 0.5595615806172484,
      "grad_norm": 2.796875,
      "learning_rate": 6.465461536312499e-06,
      "loss": 1.40925931930542,
      "step": 485
    },
    {
      "epoch": 0.5653302567060859,
      "grad_norm": 4.15625,
      "learning_rate": 6.326741512198267e-06,
      "loss": 1.4209882736206054,
      "step": 490
    },
    {
      "epoch": 0.5710989327949235,
      "grad_norm": 2.859375,
      "learning_rate": 6.188431750970279e-06,
      "loss": 1.4472535133361817,
      "step": 495
    },
    {
      "epoch": 0.5768676088837612,
      "grad_norm": 2.71875,
      "learning_rate": 6.050580616533352e-06,
      "loss": 1.4107693672180175,
      "step": 500
    },
    {
      "epoch": 0.5826362849725988,
      "grad_norm": 2.640625,
      "learning_rate": 5.913236312420529e-06,
      "loss": 1.3599486351013184,
      "step": 505
    },
    {
      "epoch": 0.5884049610614364,
      "grad_norm": 2.9375,
      "learning_rate": 5.7764468649373555e-06,
      "loss": 1.4726941108703613,
      "step": 510
    },
    {
      "epoch": 0.594173637150274,
      "grad_norm": 3.90625,
      "learning_rate": 5.640260106368114e-06,
      "loss": 1.3863279342651367,
      "step": 515
    },
    {
      "epoch": 0.5999423132391116,
      "grad_norm": 2.375,
      "learning_rate": 5.5047236582499375e-06,
      "loss": 1.4163952827453614,
      "step": 520
    },
    {
      "epoch": 0.6057109893279492,
      "grad_norm": 2.296875,
      "learning_rate": 5.36988491472058e-06,
      "loss": 1.37623929977417,
      "step": 525
    },
    {
      "epoch": 0.6114796654167869,
      "grad_norm": 2.671875,
      "learning_rate": 5.235791025945735e-06,
      "loss": 1.5111130714416503,
      "step": 530
    },
    {
      "epoch": 0.6172483415056245,
      "grad_norm": 4.0625,
      "learning_rate": 5.10248888163164e-06,
      "loss": 1.3977880477905273,
      "step": 535
    },
    {
      "epoch": 0.623017017594462,
      "grad_norm": 2.96875,
      "learning_rate": 4.970025094628776e-06,
      "loss": 1.4750987052917481,
      "step": 540
    },
    {
      "epoch": 0.6287856936832997,
      "grad_norm": 3.515625,
      "learning_rate": 4.8384459846323695e-06,
      "loss": 1.3118095397949219,
      "step": 545
    },
    {
      "epoch": 0.6345543697721373,
      "grad_norm": 2.578125,
      "learning_rate": 4.707797561985411e-06,
      "loss": 1.3975916862487794,
      "step": 550
    },
    {
      "epoch": 0.6403230458609749,
      "grad_norm": 3.0625,
      "learning_rate": 4.5781255115898316e-06,
      "loss": 1.3796093940734864,
      "step": 555
    },
    {
      "epoch": 0.6460917219498126,
      "grad_norm": 2.75,
      "learning_rate": 4.4494751769315e-06,
      "loss": 1.4118720054626466,
      "step": 560
    },
    {
      "epoch": 0.6518603980386501,
      "grad_norm": 3.328125,
      "learning_rate": 4.321891544224585e-06,
      "loss": 1.4323857307434082,
      "step": 565
    },
    {
      "epoch": 0.6576290741274877,
      "grad_norm": 3.234375,
      "learning_rate": 4.195419226680871e-06,
      "loss": 1.3628763198852538,
      "step": 570
    },
    {
      "epoch": 0.6633977502163253,
      "grad_norm": 3.21875,
      "learning_rate": 4.0701024489094845e-06,
      "loss": 1.4056241989135743,
      "step": 575
    },
    {
      "epoch": 0.669166426305163,
      "grad_norm": 3.578125,
      "learning_rate": 3.94598503145251e-06,
      "loss": 1.436029052734375,
      "step": 580
    },
    {
      "epoch": 0.6749351023940006,
      "grad_norm": 4.21875,
      "learning_rate": 3.823110375461909e-06,
      "loss": 1.4853976249694825,
      "step": 585
    },
    {
      "epoch": 0.6807037784828381,
      "grad_norm": 4.15625,
      "learning_rate": 3.7015214475230806e-06,
      "loss": 1.4030592918395997,
      "step": 590
    },
    {
      "epoch": 0.6864724545716758,
      "grad_norm": 3.09375,
      "learning_rate": 3.5812607646303843e-06,
      "loss": 1.3297285079956054,
      "step": 595
    },
    {
      "epoch": 0.6922411306605134,
      "grad_norm": 3.4375,
      "learning_rate": 3.462370379319884e-06,
      "loss": 1.3419939994812011,
      "step": 600
    },
    {
      "epoch": 0.698009806749351,
      "grad_norm": 4.875,
      "learning_rate": 3.3448918649644763e-06,
      "loss": 1.3997164726257325,
      "step": 605
    },
    {
      "epoch": 0.7037784828381887,
      "grad_norm": 2.078125,
      "learning_rate": 3.2288663012366104e-06,
      "loss": 1.417185115814209,
      "step": 610
    },
    {
      "epoch": 0.7095471589270262,
      "grad_norm": 2.078125,
      "learning_rate": 3.114334259743604e-06,
      "loss": 1.4784677505493165,
      "step": 615
    },
    {
      "epoch": 0.7153158350158638,
      "grad_norm": 2.15625,
      "learning_rate": 3.0013357898406473e-06,
      "loss": 1.5040478706359863,
      "step": 620
    },
    {
      "epoch": 0.7210845111047015,
      "grad_norm": 2.859375,
      "learning_rate": 2.889910404626421e-06,
      "loss": 1.4505383491516113,
      "step": 625
    },
    {
      "epoch": 0.7268531871935391,
      "grad_norm": 4.46875,
      "learning_rate": 2.7800970671262205e-06,
      "loss": 1.398913860321045,
      "step": 630
    },
    {
      "epoch": 0.7326218632823767,
      "grad_norm": 1.9765625,
      "learning_rate": 2.6719341766674354e-06,
      "loss": 1.4610247611999512,
      "step": 635
    },
    {
      "epoch": 0.7383905393712143,
      "grad_norm": 3.46875,
      "learning_rate": 2.5654595554521557e-06,
      "loss": 1.4679314613342285,
      "step": 640
    },
    {
      "epoch": 0.7441592154600519,
      "grad_norm": 2.828125,
      "learning_rate": 2.4607104353315556e-06,
      "loss": 1.4416119575500488,
      "step": 645
    },
    {
      "epoch": 0.7499278915488895,
      "grad_norm": 2.484375,
      "learning_rate": 2.3577234447867435e-06,
      "loss": 1.4032243728637694,
      "step": 650
    },
    {
      "epoch": 0.7556965676377272,
      "grad_norm": 2.90625,
      "learning_rate": 2.2565345961205697e-06,
      "loss": 1.4307050704956055,
      "step": 655
    },
    {
      "epoch": 0.7614652437265648,
      "grad_norm": 2.40625,
      "learning_rate": 2.157179272864915e-06,
      "loss": 1.3724449157714844,
      "step": 660
    },
    {
      "epoch": 0.7672339198154023,
      "grad_norm": 2.25,
      "learning_rate": 2.0596922174078447e-06,
      "loss": 1.4334271430969239,
      "step": 665
    },
    {
      "epoch": 0.77300259590424,
      "grad_norm": 3.828125,
      "learning_rate": 1.9641075188449458e-06,
      "loss": 1.3669556617736816,
      "step": 670
    },
    {
      "epoch": 0.7787712719930776,
      "grad_norm": 2.90625,
      "learning_rate": 1.8704586010591252e-06,
      "loss": 1.4459367752075196,
      "step": 675
    },
    {
      "epoch": 0.7845399480819152,
      "grad_norm": 2.796875,
      "learning_rate": 1.7787782110330011e-06,
      "loss": 1.4034392356872558,
      "step": 680
    },
    {
      "epoch": 0.7903086241707528,
      "grad_norm": 2.203125,
      "learning_rate": 1.6890984073979934e-06,
      "loss": 1.3815957069396974,
      "step": 685
    },
    {
      "epoch": 0.7960773002595904,
      "grad_norm": 2.3125,
      "learning_rate": 1.601450549224127e-06,
      "loss": 1.3128249168395996,
      "step": 690
    },
    {
      "epoch": 0.801845976348428,
      "grad_norm": 2.890625,
      "learning_rate": 1.5158652850544483e-06,
      "loss": 1.3994318962097168,
      "step": 695
    },
    {
      "epoch": 0.8076146524372656,
      "grad_norm": 2.28125,
      "learning_rate": 1.432372542187895e-06,
      "loss": 1.4310274124145508,
      "step": 700
    },
    {
      "epoch": 0.8133833285261033,
      "grad_norm": 2.625,
      "learning_rate": 1.35100151621438e-06,
      "loss": 1.3345772743225097,
      "step": 705
    },
    {
      "epoch": 0.8191520046149409,
      "grad_norm": 2.25,
      "learning_rate": 1.2717806608057217e-06,
      "loss": 1.4705657958984375,
      "step": 710
    },
    {
      "epoch": 0.8249206807037784,
      "grad_norm": 1.921875,
      "learning_rate": 1.1947376777660244e-06,
      "loss": 1.4911749839782715,
      "step": 715
    },
    {
      "epoch": 0.8306893567926161,
      "grad_norm": 3.6875,
      "learning_rate": 1.1198995073449542e-06,
      "loss": 1.3002596855163575,
      "step": 720
    },
    {
      "epoch": 0.8364580328814537,
      "grad_norm": 2.421875,
      "learning_rate": 1.0472923188173266e-06,
      "loss": 1.381908130645752,
      "step": 725
    },
    {
      "epoch": 0.8422267089702913,
      "grad_norm": 2.546875,
      "learning_rate": 9.769415013322849e-07,
      "loss": 1.4264190673828125,
      "step": 730
    },
    {
      "epoch": 0.847995385059129,
      "grad_norm": 2.515625,
      "learning_rate": 9.088716550352605e-07,
      "loss": 1.3962379455566407,
      "step": 735
    },
    {
      "epoch": 0.8537640611479665,
      "grad_norm": 2.6875,
      "learning_rate": 8.431065824658443e-07,
      "loss": 1.4328977584838867,
      "step": 740
    },
    {
      "epoch": 0.8595327372368041,
      "grad_norm": 2.84375,
      "learning_rate": 7.796692802345615e-07,
      "loss": 1.4179375648498536,
      "step": 745
    },
    {
      "epoch": 0.8653014133256418,
      "grad_norm": 3.0,
      "learning_rate": 7.185819309814492e-07,
      "loss": 1.3624920845031738,
      "step": 750
    },
    {
      "epoch": 0.8710700894144794,
      "grad_norm": 3.0625,
      "learning_rate": 6.598658956192829e-07,
      "loss": 1.3818011283874512,
      "step": 755
    },
    {
      "epoch": 0.876838765503317,
      "grad_norm": 2.796875,
      "learning_rate": 6.035417058641207e-07,
      "loss": 1.334589385986328,
      "step": 760
    },
    {
      "epoch": 0.8826074415921547,
      "grad_norm": 3.5,
      "learning_rate": 5.496290570558157e-07,
      "loss": 1.4661783218383788,
      "step": 765
    },
    {
      "epoch": 0.8883761176809922,
      "grad_norm": 2.453125,
      "learning_rate": 4.981468012709877e-07,
      "loss": 1.320871353149414,
      "step": 770
    },
    {
      "epoch": 0.8941447937698298,
      "grad_norm": 3.109375,
      "learning_rate": 4.4911294073085724e-07,
      "loss": 1.4195501327514648,
      "step": 775
    },
    {
      "epoch": 0.8999134698586674,
      "grad_norm": 2.9375,
      "learning_rate": 4.025446215062681e-07,
      "loss": 1.388649845123291,
      "step": 780
    },
    {
      "epoch": 0.9056821459475051,
      "grad_norm": 2.359375,
      "learning_rate": 3.584581275220772e-07,
      "loss": 1.4097150802612304,
      "step": 785
    },
    {
      "epoch": 0.9114508220363426,
      "grad_norm": 3.828125,
      "learning_rate": 3.1686887486302164e-07,
      "loss": 1.3754265785217286,
      "step": 790
    },
    {
      "epoch": 0.9172194981251802,
      "grad_norm": 2.453125,
      "learning_rate": 2.777914063830492e-07,
      "loss": 1.381131935119629,
      "step": 795
    },
    {
      "epoch": 0.9229881742140179,
      "grad_norm": 2.453125,
      "learning_rate": 2.4123938662000186e-07,
      "loss": 1.471685028076172,
      "step": 800
    },
    {
      "epoch": 0.9287568503028555,
      "grad_norm": 2.65625,
      "learning_rate": 2.0722559701742582e-07,
      "loss": 1.4830981254577638,
      "step": 805
    },
    {
      "epoch": 0.9345255263916931,
      "grad_norm": 2.59375,
      "learning_rate": 1.7576193145517782e-07,
      "loss": 1.4231231689453125,
      "step": 810
    },
    {
      "epoch": 0.9402942024805307,
      "grad_norm": 2.65625,
      "learning_rate": 1.4685939209039324e-07,
      "loss": 1.3563322067260741,
      "step": 815
    },
    {
      "epoch": 0.9460628785693683,
      "grad_norm": 3.09375,
      "learning_rate": 1.205280855102775e-07,
      "loss": 1.4409172058105468,
      "step": 820
    },
    {
      "epoch": 0.9518315546582059,
      "grad_norm": 3.109375,
      "learning_rate": 9.677721919804816e-08,
      "loss": 1.4546229362487793,
      "step": 825
    },
    {
      "epoch": 0.9576002307470436,
      "grad_norm": 2.8125,
      "learning_rate": 7.561509831327779e-08,
      "loss": 1.4547360420227051,
      "step": 830
    },
    {
      "epoch": 0.9633689068358812,
      "grad_norm": 2.625,
      "learning_rate": 5.704912278776952e-08,
      "loss": 1.3702272415161132,
      "step": 835
    },
    {
      "epoch": 0.9691375829247187,
      "grad_norm": 4.4375,
      "learning_rate": 4.108578473795033e-08,
      "loss": 1.2763612747192383,
      "step": 840
    },
    {
      "epoch": 0.9749062590135564,
      "grad_norm": 3.890625,
      "learning_rate": 2.773066619472936e-08,
      "loss": 1.4001253128051758,
      "step": 845
    },
    {
      "epoch": 0.980674935102394,
      "grad_norm": 2.28125,
      "learning_rate": 1.6988437151579385e-08,
      "loss": 1.4111320495605468,
      "step": 850
    },
    {
      "epoch": 0.9864436111912316,
      "grad_norm": 4.90625,
      "learning_rate": 8.862853931542181e-09,
      "loss": 1.367732334136963,
      "step": 855
    },
    {
      "epoch": 0.9922122872800693,
      "grad_norm": 4.4375,
      "learning_rate": 3.3567578737250802e-09,
      "loss": 1.3816195487976075,
      "step": 860
    },
    {
      "epoch": 0.9979809633689068,
      "grad_norm": 2.46875,
      "learning_rate": 4.720743397457205e-10,
      "loss": 1.3967268943786622,
      "step": 865
    },
    {
      "epoch": 1.0,
      "step": 867,
      "total_flos": 4.2033452439454925e+17,
      "train_loss": 1.4887716360433154,
      "train_runtime": 7587.3157,
      "train_samples_per_second": 0.457,
      "train_steps_per_second": 0.114
    }
  ],
  "logging_steps": 5,
  "max_steps": 867,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.2033452439454925e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}