{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9995203836930455,
  "eval_steps": 500,
  "global_step": 1042,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0009592326139088729,
      "grad_norm": 3.8003539568068034,
      "learning_rate": 1.9047619047619051e-06,
      "loss": 1.3148,
      "step": 1
    },
    {
      "epoch": 0.004796163069544364,
      "grad_norm": 1.3427176668530907,
      "learning_rate": 9.523809523809523e-06,
      "loss": 1.3465,
      "step": 5
    },
    {
      "epoch": 0.009592326139088728,
      "grad_norm": 0.6129061923124927,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 1.3062,
      "step": 10
    },
    {
      "epoch": 0.014388489208633094,
      "grad_norm": 0.5006461190956731,
      "learning_rate": 2.857142857142857e-05,
      "loss": 1.258,
      "step": 15
    },
    {
      "epoch": 0.019184652278177457,
      "grad_norm": 0.3919323663241635,
      "learning_rate": 3.809523809523809e-05,
      "loss": 1.2099,
      "step": 20
    },
    {
      "epoch": 0.023980815347721823,
      "grad_norm": 0.33016517225439684,
      "learning_rate": 4.761904761904762e-05,
      "loss": 1.1774,
      "step": 25
    },
    {
      "epoch": 0.02877697841726619,
      "grad_norm": 0.24241904389090518,
      "learning_rate": 5.714285714285714e-05,
      "loss": 1.1464,
      "step": 30
    },
    {
      "epoch": 0.03357314148681055,
      "grad_norm": 0.21825210432654424,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.1551,
      "step": 35
    },
    {
      "epoch": 0.03836930455635491,
      "grad_norm": 0.2357580892216174,
      "learning_rate": 7.619047619047618e-05,
      "loss": 1.1329,
      "step": 40
    },
    {
      "epoch": 0.04316546762589928,
      "grad_norm": 0.20958683389560634,
      "learning_rate": 8.571428571428571e-05,
      "loss": 1.1255,
      "step": 45
    },
    {
      "epoch": 0.047961630695443645,
      "grad_norm": 0.2313801330004238,
      "learning_rate": 9.523809523809524e-05,
      "loss": 1.126,
      "step": 50
    },
    {
      "epoch": 0.05275779376498801,
      "grad_norm": 0.2081771832083775,
      "learning_rate": 0.00010476190476190477,
      "loss": 1.0935,
      "step": 55
    },
    {
      "epoch": 0.05755395683453238,
      "grad_norm": 0.2039705643041609,
      "learning_rate": 0.00011428571428571428,
      "loss": 1.1394,
      "step": 60
    },
    {
      "epoch": 0.06235011990407674,
      "grad_norm": 0.2148039800819315,
      "learning_rate": 0.0001238095238095238,
      "loss": 1.128,
      "step": 65
    },
    {
      "epoch": 0.0671462829736211,
      "grad_norm": 0.18570989413688033,
      "learning_rate": 0.00013333333333333334,
      "loss": 1.1474,
      "step": 70
    },
    {
      "epoch": 0.07194244604316546,
      "grad_norm": 0.18151401221015465,
      "learning_rate": 0.00014285714285714287,
      "loss": 1.1297,
      "step": 75
    },
    {
      "epoch": 0.07673860911270983,
      "grad_norm": 0.18983827314633334,
      "learning_rate": 0.00015238095238095237,
      "loss": 1.0947,
      "step": 80
    },
    {
      "epoch": 0.0815347721822542,
      "grad_norm": 0.18321896415592545,
      "learning_rate": 0.00016190476190476192,
      "loss": 1.0993,
      "step": 85
    },
    {
      "epoch": 0.08633093525179857,
      "grad_norm": 0.18106666815994735,
      "learning_rate": 0.00017142857142857143,
      "loss": 1.1169,
      "step": 90
    },
    {
      "epoch": 0.09112709832134293,
      "grad_norm": 0.17126918097791768,
      "learning_rate": 0.00018095238095238095,
      "loss": 1.1336,
      "step": 95
    },
    {
      "epoch": 0.09592326139088729,
      "grad_norm": 0.1787566440192351,
      "learning_rate": 0.00019047619047619048,
      "loss": 1.1086,
      "step": 100
    },
    {
      "epoch": 0.10071942446043165,
      "grad_norm": 0.18251300143969484,
      "learning_rate": 0.0002,
      "loss": 1.0947,
      "step": 105
    },
    {
      "epoch": 0.10551558752997602,
      "grad_norm": 0.19966595269166043,
      "learning_rate": 0.00019998594857383755,
      "loss": 1.1381,
      "step": 110
    },
    {
      "epoch": 0.11031175059952038,
      "grad_norm": 0.18103343210950873,
      "learning_rate": 0.0001999437982442017,
      "loss": 1.1246,
      "step": 115
    },
    {
      "epoch": 0.11510791366906475,
      "grad_norm": 0.17391603743469916,
      "learning_rate": 0.00019987356085653736,
      "loss": 1.1109,
      "step": 120
    },
    {
      "epoch": 0.11990407673860912,
      "grad_norm": 0.18378902142023112,
      "learning_rate": 0.00019977525614955387,
      "loss": 1.0974,
      "step": 125
    },
    {
      "epoch": 0.12470023980815348,
      "grad_norm": 0.18844867723630976,
      "learning_rate": 0.00019964891174967784,
      "loss": 1.0954,
      "step": 130
    },
    {
      "epoch": 0.12949640287769784,
      "grad_norm": 0.17258678109282075,
      "learning_rate": 0.0001994945631632894,
      "loss": 1.1045,
      "step": 135
    },
    {
      "epoch": 0.1342925659472422,
      "grad_norm": 0.19501601610290073,
      "learning_rate": 0.00019931225376674388,
      "loss": 1.1151,
      "step": 140
    },
    {
      "epoch": 0.13908872901678657,
      "grad_norm": 0.16557441103633752,
      "learning_rate": 0.00019910203479418172,
      "loss": 1.0931,
      "step": 145
    },
    {
      "epoch": 0.14388489208633093,
      "grad_norm": 0.1808422208750127,
      "learning_rate": 0.00019886396532313032,
      "loss": 1.1439,
      "step": 150
    },
    {
      "epoch": 0.1486810551558753,
      "grad_norm": 0.17153212310391777,
      "learning_rate": 0.00019859811225790162,
      "loss": 1.1238,
      "step": 155
    },
    {
      "epoch": 0.15347721822541965,
      "grad_norm": 0.1725494178552086,
      "learning_rate": 0.00019830455031078992,
      "loss": 1.0819,
      "step": 160
    },
    {
      "epoch": 0.15827338129496402,
      "grad_norm": 0.17821716654615946,
      "learning_rate": 0.00019798336198107567,
      "loss": 1.106,
      "step": 165
    },
    {
      "epoch": 0.1630695443645084,
      "grad_norm": 0.17454115891711316,
      "learning_rate": 0.0001976346375318409,
      "loss": 1.1176,
      "step": 170
    },
    {
      "epoch": 0.16786570743405277,
      "grad_norm": 0.17165723260861443,
      "learning_rate": 0.00019725847496460257,
      "loss": 1.104,
      "step": 175
    },
    {
      "epoch": 0.17266187050359713,
      "grad_norm": 0.16959903827879758,
      "learning_rate": 0.00019685497999177146,
      "loss": 1.1032,
      "step": 180
    },
    {
      "epoch": 0.1774580335731415,
      "grad_norm": 0.16370292319935667,
      "learning_rate": 0.00019642426600694396,
      "loss": 1.1058,
      "step": 185
    },
    {
      "epoch": 0.18225419664268586,
      "grad_norm": 0.17362596041949555,
      "learning_rate": 0.00019596645405303508,
      "loss": 1.0897,
      "step": 190
    },
    {
      "epoch": 0.18705035971223022,
      "grad_norm": 0.1705391917273308,
      "learning_rate": 0.00019548167278826223,
      "loss": 1.0793,
      "step": 195
    },
    {
      "epoch": 0.19184652278177458,
      "grad_norm": 0.16987370700655263,
      "learning_rate": 0.00019497005844998835,
      "loss": 1.0825,
      "step": 200
    },
    {
      "epoch": 0.19664268585131894,
      "grad_norm": 0.1741514081542761,
      "learning_rate": 0.00019443175481643533,
      "loss": 1.0971,
      "step": 205
    },
    {
      "epoch": 0.2014388489208633,
      "grad_norm": 0.1697720073525996,
      "learning_rate": 0.00019386691316627846,
      "loss": 1.1004,
      "step": 210
    },
    {
      "epoch": 0.20623501199040767,
      "grad_norm": 0.16941098489834122,
      "learning_rate": 0.00019327569223613248,
      "loss": 1.1078,
      "step": 215
    },
    {
      "epoch": 0.21103117505995203,
      "grad_norm": 0.16231625897366397,
      "learning_rate": 0.0001926582581759423,
      "loss": 1.0941,
      "step": 220
    },
    {
      "epoch": 0.2158273381294964,
      "grad_norm": 0.16520366738739287,
      "learning_rate": 0.00019201478450229012,
      "loss": 1.0868,
      "step": 225
    },
    {
      "epoch": 0.22062350119904076,
      "grad_norm": 0.16835416401080175,
      "learning_rate": 0.00019134545204963212,
      "loss": 1.1037,
      "step": 230
    },
    {
      "epoch": 0.22541966426858512,
      "grad_norm": 0.16657613469605134,
      "learning_rate": 0.0001906504489194791,
      "loss": 1.093,
      "step": 235
    },
    {
      "epoch": 0.2302158273381295,
      "grad_norm": 0.1746269506944523,
      "learning_rate": 0.00018992997042753434,
      "loss": 1.1124,
      "step": 240
    },
    {
      "epoch": 0.23501199040767387,
      "grad_norm": 0.16885168676581888,
      "learning_rate": 0.0001891842190488045,
      "loss": 1.0972,
      "step": 245
    },
    {
      "epoch": 0.23980815347721823,
      "grad_norm": 0.16263325756495298,
      "learning_rate": 0.00018841340436069826,
      "loss": 1.0725,
      "step": 250
    },
    {
      "epoch": 0.2446043165467626,
      "grad_norm": 0.16758205657577824,
      "learning_rate": 0.00018761774298412903,
      "loss": 1.1012,
      "step": 255
    },
    {
      "epoch": 0.24940047961630696,
      "grad_norm": 0.1592358008735748,
      "learning_rate": 0.00018679745852263858,
      "loss": 1.1163,
      "step": 260
    },
    {
      "epoch": 0.2541966426858513,
      "grad_norm": 0.18307373045740924,
      "learning_rate": 0.0001859527814995577,
      "loss": 1.0883,
      "step": 265
    },
    {
      "epoch": 0.2589928057553957,
      "grad_norm": 0.1603879751504812,
      "learning_rate": 0.00018508394929322286,
      "loss": 1.0837,
      "step": 270
    },
    {
      "epoch": 0.2637889688249401,
      "grad_norm": 0.16169604116219924,
      "learning_rate": 0.0001841912060702659,
      "loss": 1.1059,
      "step": 275
    },
    {
      "epoch": 0.2685851318944844,
      "grad_norm": 0.21284544973156141,
      "learning_rate": 0.00018327480271699645,
      "loss": 1.1197,
      "step": 280
    },
    {
      "epoch": 0.2733812949640288,
      "grad_norm": 0.16556421777414207,
      "learning_rate": 0.00018233499676889556,
      "loss": 1.0857,
      "step": 285
    },
    {
      "epoch": 0.27817745803357313,
      "grad_norm": 0.17311293794732555,
      "learning_rate": 0.00018137205233824098,
      "loss": 1.1215,
      "step": 290
    },
    {
      "epoch": 0.2829736211031175,
      "grad_norm": 0.16045436773084543,
      "learning_rate": 0.00018038624003988404,
      "loss": 1.1102,
      "step": 295
    },
    {
      "epoch": 0.28776978417266186,
      "grad_norm": 0.16385963917512852,
      "learning_rate": 0.0001793778369151991,
      "loss": 1.1237,
      "step": 300
    },
    {
      "epoch": 0.29256594724220625,
      "grad_norm": 0.16336813988660637,
      "learning_rate": 0.00017834712635422716,
      "loss": 1.07,
      "step": 305
    },
    {
      "epoch": 0.2973621103117506,
      "grad_norm": 0.17112565194779492,
      "learning_rate": 0.0001772943980160351,
      "loss": 1.1009,
      "step": 310
    },
    {
      "epoch": 0.302158273381295,
      "grad_norm": 0.16697464006830123,
      "learning_rate": 0.0001762199477473131,
      "loss": 1.0972,
      "step": 315
    },
    {
      "epoch": 0.3069544364508393,
      "grad_norm": 0.16453710745761058,
      "learning_rate": 0.0001751240774992336,
      "loss": 1.074,
      "step": 320
    },
    {
      "epoch": 0.3117505995203837,
      "grad_norm": 0.17019017682628554,
      "learning_rate": 0.000174007095242594,
      "loss": 1.0605,
      "step": 325
    },
    {
      "epoch": 0.31654676258992803,
      "grad_norm": 0.17847410199181396,
      "learning_rate": 0.00017286931488126839,
      "loss": 1.1109,
      "step": 330
    },
    {
      "epoch": 0.3213429256594724,
      "grad_norm": 0.1628727862316133,
      "learning_rate": 0.00017171105616399152,
      "loss": 1.1024,
      "step": 335
    },
    {
      "epoch": 0.3261390887290168,
      "grad_norm": 0.16567336419547818,
      "learning_rate": 0.0001705326445945002,
      "loss": 1.072,
      "step": 340
    },
    {
      "epoch": 0.33093525179856115,
      "grad_norm": 0.16722439948774648,
      "learning_rate": 0.0001693344113400577,
      "loss": 1.0873,
      "step": 345
    },
    {
      "epoch": 0.33573141486810554,
      "grad_norm": 0.16122286760160678,
      "learning_rate": 0.0001681166931383859,
      "loss": 1.0938,
      "step": 350
    },
    {
      "epoch": 0.3405275779376499,
      "grad_norm": 0.15969633584876602,
      "learning_rate": 0.00016687983220303282,
      "loss": 1.107,
      "step": 355
    },
    {
      "epoch": 0.34532374100719426,
      "grad_norm": 0.20032537034543332,
      "learning_rate": 0.00016562417612720054,
      "loss": 1.0925,
      "step": 360
    },
    {
      "epoch": 0.3501199040767386,
      "grad_norm": 0.1636161155593432,
      "learning_rate": 0.00016435007778606178,
      "loss": 1.0925,
      "step": 365
    },
    {
      "epoch": 0.354916067146283,
      "grad_norm": 0.17509644211474343,
      "learning_rate": 0.00016305789523759186,
      "loss": 1.0853,
      "step": 370
    },
    {
      "epoch": 0.3597122302158273,
      "grad_norm": 0.15845793640793487,
      "learning_rate": 0.00016174799162194407,
      "loss": 1.074,
      "step": 375
    },
    {
      "epoch": 0.3645083932853717,
      "grad_norm": 0.1643115007590844,
      "learning_rate": 0.00016042073505939718,
      "loss": 1.0904,
      "step": 380
    },
    {
      "epoch": 0.36930455635491605,
      "grad_norm": 0.159834353020724,
      "learning_rate": 0.0001590764985469029,
      "loss": 1.0913,
      "step": 385
    },
    {
      "epoch": 0.37410071942446044,
      "grad_norm": 0.15671259478172928,
      "learning_rate": 0.00015771565985326323,
      "loss": 1.0949,
      "step": 390
    },
    {
      "epoch": 0.37889688249400477,
      "grad_norm": 0.15892776737127523,
      "learning_rate": 0.0001563386014129667,
      "loss": 1.0882,
      "step": 395
    },
    {
      "epoch": 0.38369304556354916,
      "grad_norm": 0.1554951930431914,
      "learning_rate": 0.00015494571021871308,
      "loss": 1.0849,
      "step": 400
    },
    {
      "epoch": 0.38848920863309355,
      "grad_norm": 0.15596140521064988,
      "learning_rate": 0.00015353737771265787,
      "loss": 1.0892,
      "step": 405
    },
    {
      "epoch": 0.3932853717026379,
      "grad_norm": 0.16030122057012702,
      "learning_rate": 0.00015211399967640537,
      "loss": 1.073,
      "step": 410
    },
    {
      "epoch": 0.3980815347721823,
      "grad_norm": 0.15896611447165424,
      "learning_rate": 0.00015067597611978327,
      "loss": 1.1113,
      "step": 415
    },
    {
      "epoch": 0.4028776978417266,
      "grad_norm": 0.1532650010931279,
      "learning_rate": 0.000149223711168428,
      "loss": 1.0881,
      "step": 420
    },
    {
      "epoch": 0.407673860911271,
      "grad_norm": 0.16086089606757772,
      "learning_rate": 0.00014775761295021417,
      "loss": 1.0887,
      "step": 425
    },
    {
      "epoch": 0.41247002398081534,
      "grad_norm": 0.15347896448970985,
      "learning_rate": 0.00014627809348055908,
      "loss": 1.0838,
      "step": 430
    },
    {
      "epoch": 0.4172661870503597,
      "grad_norm": 0.16005243796029608,
      "learning_rate": 0.00014478556854663434,
      "loss": 1.1036,
      "step": 435
    },
    {
      "epoch": 0.42206235011990406,
      "grad_norm": 0.23613805183753908,
      "learning_rate": 0.00014328045759051805,
      "loss": 1.0886,
      "step": 440
    },
    {
      "epoch": 0.42685851318944845,
      "grad_norm": 0.15961403055430387,
      "learning_rate": 0.00014176318359131955,
      "loss": 1.0807,
      "step": 445
    },
    {
      "epoch": 0.4316546762589928,
      "grad_norm": 0.15677042464881533,
      "learning_rate": 0.00014023417294631017,
      "loss": 1.1133,
      "step": 450
    },
    {
      "epoch": 0.4364508393285372,
      "grad_norm": 0.14974169551675579,
      "learning_rate": 0.0001386938553510936,
      "loss": 1.0799,
      "step": 455
    },
    {
      "epoch": 0.4412470023980815,
      "grad_norm": 0.15803754347469678,
      "learning_rate": 0.00013714266367884884,
      "loss": 1.0735,
      "step": 460
    },
    {
      "epoch": 0.4460431654676259,
      "grad_norm": 0.1534708691023063,
      "learning_rate": 0.00013558103385868085,
      "loss": 1.0941,
      "step": 465
    },
    {
      "epoch": 0.45083932853717024,
      "grad_norm": 0.15911138328699584,
      "learning_rate": 0.00013400940475311192,
      "loss": 1.1036,
      "step": 470
    },
    {
      "epoch": 0.4556354916067146,
      "grad_norm": 0.1535927522203036,
      "learning_rate": 0.0001324282180347486,
      "loss": 1.077,
      "step": 475
    },
    {
      "epoch": 0.460431654676259,
      "grad_norm": 0.15031744415995577,
      "learning_rate": 0.00013083791806215938,
      "loss": 1.1175,
      "step": 480
    },
    {
      "epoch": 0.46522781774580335,
      "grad_norm": 0.15368131899023496,
      "learning_rate": 0.0001292389517549971,
      "loss": 1.0858,
      "step": 485
    },
    {
      "epoch": 0.47002398081534774,
      "grad_norm": 0.15817599988761602,
      "learning_rate": 0.0001276317684684017,
      "loss": 1.1105,
      "step": 490
    },
    {
      "epoch": 0.4748201438848921,
      "grad_norm": 0.1544524844304045,
      "learning_rate": 0.0001260168198667189,
      "loss": 1.0675,
      "step": 495
    },
    {
      "epoch": 0.47961630695443647,
      "grad_norm": 0.153601743822322,
      "learning_rate": 0.00012439455979656932,
      "loss": 1.0792,
      "step": 500
    },
    {
      "epoch": 0.4844124700239808,
      "grad_norm": 0.15310739937604212,
      "learning_rate": 0.00012276544415930476,
      "loss": 1.1037,
      "step": 505
    },
    {
      "epoch": 0.4892086330935252,
      "grad_norm": 0.15940815066320235,
      "learning_rate": 0.00012112993078288702,
      "loss": 1.0893,
      "step": 510
    },
    {
      "epoch": 0.4940047961630695,
      "grad_norm": 0.15593480397437104,
      "learning_rate": 0.00011948847929322497,
      "loss": 1.0736,
      "step": 515
    },
    {
      "epoch": 0.4988009592326139,
      "grad_norm": 0.15351285356914193,
      "learning_rate": 0.00011784155098500682,
      "loss": 1.0734,
      "step": 520
    },
    {
      "epoch": 0.5035971223021583,
      "grad_norm": 0.1648474397335208,
      "learning_rate": 0.00011618960869206285,
      "loss": 1.0902,
      "step": 525
    },
    {
      "epoch": 0.5083932853717026,
      "grad_norm": 0.15294486843196387,
      "learning_rate": 0.00011453311665729618,
      "loss": 1.0963,
      "step": 530
    },
    {
      "epoch": 0.513189448441247,
      "grad_norm": 0.148639402509462,
      "learning_rate": 0.0001128725404022171,
      "loss": 1.0796,
      "step": 535
    },
    {
      "epoch": 0.5179856115107914,
      "grad_norm": 0.15562736350879572,
      "learning_rate": 0.00011120834659611831,
      "loss": 1.0906,
      "step": 540
    },
    {
      "epoch": 0.5227817745803357,
      "grad_norm": 0.1502074864032036,
      "learning_rate": 0.00010954100292492757,
      "loss": 1.0711,
      "step": 545
    },
    {
      "epoch": 0.5275779376498801,
      "grad_norm": 0.157682125372101,
      "learning_rate": 0.00010787097795977448,
      "loss": 1.0728,
      "step": 550
    },
    {
      "epoch": 0.5323741007194245,
      "grad_norm": 0.1516246607062371,
      "learning_rate": 0.00010619874102530885,
      "loss": 1.0833,
      "step": 555
    },
    {
      "epoch": 0.5371702637889688,
      "grad_norm": 0.14890743598378445,
      "learning_rate": 0.00010452476206780685,
      "loss": 1.1037,
      "step": 560
    },
    {
      "epoch": 0.5419664268585132,
      "grad_norm": 0.1544250948669629,
      "learning_rate": 0.00010284951152310292,
      "loss": 1.0882,
      "step": 565
    },
    {
      "epoch": 0.5467625899280576,
      "grad_norm": 0.15154065813090323,
      "learning_rate": 0.00010117346018438367,
      "loss": 1.0601,
      "step": 570
    },
    {
      "epoch": 0.5515587529976019,
      "grad_norm": 0.1541132913924603,
      "learning_rate": 9.949707906988165e-05,
      "loss": 1.0825,
      "step": 575
    },
    {
      "epoch": 0.5563549160671463,
      "grad_norm": 0.15035087453783752,
      "learning_rate": 9.7820839290506e-05,
      "loss": 1.0893,
      "step": 580
    },
    {
      "epoch": 0.5611510791366906,
      "grad_norm": 0.1566481016467638,
      "learning_rate": 9.614521191744644e-05,
      "loss": 1.0717,
      "step": 585
    },
    {
      "epoch": 0.565947242206235,
      "grad_norm": 0.14690623261948202,
      "learning_rate": 9.447066784978914e-05,
      "loss": 1.0843,
      "step": 590
    },
    {
      "epoch": 0.5707434052757794,
      "grad_norm": 0.14805971637155899,
      "learning_rate": 9.279767768218057e-05,
      "loss": 1.0639,
      "step": 595
    },
    {
      "epoch": 0.5755395683453237,
      "grad_norm": 0.15165832191817782,
      "learning_rate": 9.112671157257698e-05,
      "loss": 1.0762,
      "step": 600
    },
    {
      "epoch": 0.580335731414868,
      "grad_norm": 0.16254973894493494,
      "learning_rate": 8.945823911011648e-05,
      "loss": 1.0627,
      "step": 605
    },
    {
      "epoch": 0.5851318944844125,
      "grad_norm": 0.15231765571558437,
      "learning_rate": 8.779272918315134e-05,
      "loss": 1.1191,
      "step": 610
    },
    {
      "epoch": 0.5899280575539568,
      "grad_norm": 0.15662286211747672,
      "learning_rate": 8.613064984747672e-05,
      "loss": 1.0814,
      "step": 615
    },
    {
      "epoch": 0.5947242206235012,
      "grad_norm": 0.15604291232878043,
      "learning_rate": 8.44724681947939e-05,
      "loss": 1.062,
      "step": 620
    },
    {
      "epoch": 0.5995203836930456,
      "grad_norm": 0.17656148460505006,
      "learning_rate": 8.281865022144402e-05,
      "loss": 1.0877,
      "step": 625
    },
    {
      "epoch": 0.60431654676259,
      "grad_norm": 0.15407547139459882,
      "learning_rate": 8.116966069744987e-05,
      "loss": 1.0883,
      "step": 630
    },
    {
      "epoch": 0.6091127098321343,
      "grad_norm": 0.15686792104917918,
      "learning_rate": 7.952596303590214e-05,
      "loss": 1.0726,
      "step": 635
    },
    {
      "epoch": 0.6139088729016786,
      "grad_norm": 0.14900379281026824,
      "learning_rate": 7.788801916272739e-05,
      "loss": 1.0873,
      "step": 640
    },
    {
      "epoch": 0.6187050359712231,
      "grad_norm": 0.15399136244929024,
      "learning_rate": 7.625628938687348e-05,
      "loss": 1.0673,
      "step": 645
    },
    {
      "epoch": 0.6235011990407674,
      "grad_norm": 0.1454561721402151,
      "learning_rate": 7.463123227094961e-05,
      "loss": 1.0648,
      "step": 650
    },
    {
      "epoch": 0.6282973621103117,
      "grad_norm": 0.15049453308757393,
      "learning_rate": 7.301330450235733e-05,
      "loss": 1.0952,
      "step": 655
    },
    {
      "epoch": 0.6330935251798561,
      "grad_norm": 0.14976244376291453,
      "learning_rate": 7.140296076494809e-05,
      "loss": 1.096,
      "step": 660
    },
    {
      "epoch": 0.6378896882494005,
      "grad_norm": 0.15150823609919292,
      "learning_rate": 6.980065361124437e-05,
      "loss": 1.0973,
      "step": 665
    },
    {
      "epoch": 0.6426858513189448,
      "grad_norm": 0.1471164569431601,
      "learning_rate": 6.820683333525942e-05,
      "loss": 1.0886,
      "step": 670
    },
    {
      "epoch": 0.6474820143884892,
      "grad_norm": 0.1478399832545343,
      "learning_rate": 6.662194784595164e-05,
      "loss": 1.0778,
      "step": 675
    },
    {
      "epoch": 0.6522781774580336,
      "grad_norm": 0.14733730171420292,
      "learning_rate": 6.504644254134969e-05,
      "loss": 1.0622,
      "step": 680
    },
    {
      "epoch": 0.657074340527578,
      "grad_norm": 0.15089732572854292,
      "learning_rate": 6.34807601833826e-05,
      "loss": 1.054,
      "step": 685
    },
    {
      "epoch": 0.6618705035971223,
      "grad_norm": 0.15257943560973228,
      "learning_rate": 6.19253407734514e-05,
      "loss": 1.0653,
      "step": 690
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.1555839364213723,
      "learning_rate": 6.038062142877583e-05,
      "loss": 1.0743,
      "step": 695
    },
    {
      "epoch": 0.6714628297362111,
      "grad_norm": 0.15847863555269973,
      "learning_rate": 5.884703625955219e-05,
      "loss": 1.0714,
      "step": 700
    },
    {
      "epoch": 0.6762589928057554,
      "grad_norm": 0.14915505074346633,
      "learning_rate": 5.73250162469559e-05,
      "loss": 1.0777,
      "step": 705
    },
    {
      "epoch": 0.6810551558752997,
      "grad_norm": 0.1558291126854066,
      "learning_rate": 5.581498912202339e-05,
      "loss": 1.0946,
      "step": 710
    },
    {
      "epoch": 0.6858513189448441,
      "grad_norm": 0.15124530700511218,
      "learning_rate": 5.431737924544763e-05,
      "loss": 1.0802,
      "step": 715
    },
    {
      "epoch": 0.6906474820143885,
      "grad_norm": 0.151312999497745,
      "learning_rate": 5.283260748832072e-05,
      "loss": 1.0809,
      "step": 720
    },
    {
      "epoch": 0.6954436450839329,
      "grad_norm": 0.14330547504302646,
      "learning_rate": 5.1361091113856875e-05,
      "loss": 1.0801,
      "step": 725
    },
    {
      "epoch": 0.7002398081534772,
      "grad_norm": 0.1480298490267667,
      "learning_rate": 4.990324366012977e-05,
      "loss": 1.0553,
      "step": 730
    },
    {
      "epoch": 0.7050359712230215,
      "grad_norm": 0.153500392941282,
      "learning_rate": 4.845947482385645e-05,
      "loss": 1.0694,
      "step": 735
    },
    {
      "epoch": 0.709832134292566,
      "grad_norm": 0.1446376081330043,
      "learning_rate": 4.7030190345260816e-05,
      "loss": 1.0684,
      "step": 740
    },
    {
      "epoch": 0.7146282973621103,
      "grad_norm": 0.15009181859267848,
      "learning_rate": 4.5615791894049286e-05,
      "loss": 1.0812,
      "step": 745
    },
    {
      "epoch": 0.7194244604316546,
      "grad_norm": 0.1489429463837929,
      "learning_rate": 4.4216676956529866e-05,
      "loss": 1.1022,
      "step": 750
    },
    {
      "epoch": 0.7242206235011991,
      "grad_norm": 0.14571898395542737,
      "learning_rate": 4.2833238723907275e-05,
      "loss": 1.0446,
      "step": 755
    },
    {
      "epoch": 0.7290167865707434,
      "grad_norm": 0.1613600673322617,
      "learning_rate": 4.146586598178506e-05,
      "loss": 1.0452,
      "step": 760
    },
    {
      "epoch": 0.7338129496402878,
      "grad_norm": 0.14371687625758103,
      "learning_rate": 4.011494300090565e-05,
      "loss": 1.0848,
      "step": 765
    },
    {
      "epoch": 0.7386091127098321,
      "grad_norm": 0.1540465353357514,
      "learning_rate": 3.878084942915936e-05,
      "loss": 1.0866,
      "step": 770
    },
    {
      "epoch": 0.7434052757793765,
      "grad_norm": 0.14845273692539365,
      "learning_rate": 3.746396018489261e-05,
      "loss": 1.064,
      "step": 775
    },
    {
      "epoch": 0.7482014388489209,
      "grad_norm": 0.1510032414720602,
      "learning_rate": 3.616464535154496e-05,
      "loss": 1.0775,
      "step": 780
    },
    {
      "epoch": 0.7529976019184652,
      "grad_norm": 0.14414705671457814,
      "learning_rate": 3.488327007364525e-05,
      "loss": 1.0664,
      "step": 785
    },
    {
      "epoch": 0.7577937649880095,
      "grad_norm": 0.14609563936342954,
      "learning_rate": 3.3620194454195564e-05,
      "loss": 1.0732,
      "step": 790
    },
    {
      "epoch": 0.762589928057554,
      "grad_norm": 0.15124924277730611,
      "learning_rate": 3.237577345347196e-05,
      "loss": 1.0796,
      "step": 795
    },
    {
      "epoch": 0.7673860911270983,
      "grad_norm": 0.14663622293812884,
      "learning_rate": 3.115035678927063e-05,
      "loss": 1.0789,
      "step": 800
    },
    {
      "epoch": 0.7721822541966427,
      "grad_norm": 0.14712216022696742,
      "learning_rate": 2.9944288838627054e-05,
      "loss": 1.0812,
      "step": 805
    },
    {
      "epoch": 0.7769784172661871,
      "grad_norm": 0.1466540809367841,
      "learning_rate": 2.875790854103634e-05,
      "loss": 1.0371,
      "step": 810
    },
    {
      "epoch": 0.7817745803357314,
      "grad_norm": 0.15209652176203295,
      "learning_rate": 2.7591549303201514e-05,
      "loss": 1.0666,
      "step": 815
    },
    {
      "epoch": 0.7865707434052758,
      "grad_norm": 0.15265531709332175,
      "learning_rate": 2.6445538905336763e-05,
      "loss": 1.0716,
      "step": 820
    },
    {
      "epoch": 0.7913669064748201,
      "grad_norm": 0.1472763184905413,
      "learning_rate": 2.532019940905186e-05,
      "loss": 1.0774,
      "step": 825
    },
    {
      "epoch": 0.7961630695443646,
      "grad_norm": 0.14917213471816188,
      "learning_rate": 2.421584706684359e-05,
      "loss": 1.0585,
      "step": 830
    },
    {
      "epoch": 0.8009592326139089,
      "grad_norm": 0.1487175153721826,
      "learning_rate": 2.3132792233219813e-05,
      "loss": 1.0845,
      "step": 835
    },
    {
      "epoch": 0.8057553956834532,
      "grad_norm": 0.14774839089244932,
      "learning_rate": 2.207133927748104e-05,
      "loss": 1.0676,
      "step": 840
    },
    {
      "epoch": 0.8105515587529976,
      "grad_norm": 0.14579576803524566,
      "learning_rate": 2.103178649818387e-05,
      "loss": 1.0618,
      "step": 845
    },
    {
      "epoch": 0.815347721822542,
      "grad_norm": 0.15161345454257397,
      "learning_rate": 2.0014426039310786e-05,
      "loss": 1.0647,
      "step": 850
    },
    {
      "epoch": 0.8201438848920863,
      "grad_norm": 0.15041577767342024,
      "learning_rate": 1.9019543808169115e-05,
      "loss": 1.0634,
      "step": 855
    },
    {
      "epoch": 0.8249400479616307,
      "grad_norm": 0.14832965984424024,
      "learning_rate": 1.8047419395043086e-05,
      "loss": 1.078,
      "step": 860
    },
    {
      "epoch": 0.829736211031175,
      "grad_norm": 0.14605268717157927,
      "learning_rate": 1.7098325994620934e-05,
      "loss": 1.0763,
      "step": 865
    },
    {
      "epoch": 0.8345323741007195,
      "grad_norm": 0.14667368288338628,
      "learning_rate": 1.6172530329219416e-05,
      "loss": 1.0792,
      "step": 870
    },
    {
      "epoch": 0.8393285371702638,
      "grad_norm": 0.1468054846510282,
      "learning_rate": 1.5270292573827173e-05,
      "loss": 1.0643,
      "step": 875
    },
    {
      "epoch": 0.8441247002398081,
      "grad_norm": 0.14221973189330656,
      "learning_rate": 1.4391866282988265e-05,
      "loss": 1.0448,
      "step": 880
    },
    {
      "epoch": 0.8489208633093526,
      "grad_norm": 0.14260459844091208,
      "learning_rate": 1.3537498319545983e-05,
      "loss": 1.0729,
      "step": 885
    },
    {
      "epoch": 0.8537170263788969,
      "grad_norm": 0.15011630591055675,
      "learning_rate": 1.2707428785267394e-05,
      "loss": 1.0737,
      "step": 890
    },
    {
      "epoch": 0.8585131894484412,
      "grad_norm": 0.1406245046991364,
      "learning_rate": 1.1901890953367911e-05,
      "loss": 1.0542,
      "step": 895
    },
    {
      "epoch": 0.8633093525179856,
      "grad_norm": 0.1452217620527171,
      "learning_rate": 1.1121111202954836e-05,
      "loss": 1.0404,
      "step": 900
    },
    {
      "epoch": 0.86810551558753,
      "grad_norm": 0.14762449970761843,
      "learning_rate": 1.0365308955408459e-05,
      "loss": 1.0668,
      "step": 905
    },
    {
      "epoch": 0.8729016786570744,
      "grad_norm": 0.14280442405170934,
      "learning_rate": 9.634696612718242e-06,
      "loss": 1.0589,
      "step": 910
    },
    {
      "epoch": 0.8776978417266187,
      "grad_norm": 0.14727617951380836,
      "learning_rate": 8.929479497791926e-06,
      "loss": 1.0521,
      "step": 915
    },
    {
      "epoch": 0.882494004796163,
      "grad_norm": 0.14230938752904093,
      "learning_rate": 8.24985579675388e-06,
      "loss": 1.0615,
      "step": 920
    },
    {
      "epoch": 0.8872901678657075,
      "grad_norm": 0.14834529182832706,
      "learning_rate": 7.59601650324917e-06,
      "loss": 1.0596,
      "step": 925
    },
    {
      "epoch": 0.8920863309352518,
      "grad_norm": 0.15376739416260338,
      "learning_rate": 6.96814536476893e-06,
      "loss": 1.0596,
      "step": 930
    },
    {
      "epoch": 0.8968824940047961,
      "grad_norm": 0.15726124144994705,
      "learning_rate": 6.366418831011956e-06,
      "loss": 1.0639,
      "step": 935
    },
    {
      "epoch": 0.9016786570743405,
      "grad_norm": 0.14356439831596873,
      "learning_rate": 5.79100600429745e-06,
      "loss": 1.09,
      "step": 940
    },
    {
      "epoch": 0.9064748201438849,
      "grad_norm": 0.1418205242753867,
      "learning_rate": 5.242068592042349e-06,
      "loss": 1.0747,
      "step": 945
    },
    {
      "epoch": 0.9112709832134293,
      "grad_norm": 0.14390931236295634,
      "learning_rate": 4.7197608613169685e-06,
      "loss": 1.0718,
      "step": 950
    },
    {
      "epoch": 0.9160671462829736,
      "grad_norm": 0.1538309152736152,
      "learning_rate": 4.224229595491591e-06,
      "loss": 1.0717,
      "step": 955
    },
    {
      "epoch": 0.920863309352518,
      "grad_norm": 0.14425154981452512,
      "learning_rate": 3.7556140529860563e-06,
      "loss": 1.1077,
      "step": 960
    },
    {
      "epoch": 0.9256594724220624,
      "grad_norm": 0.14548690711150744,
      "learning_rate": 3.314045928134224e-06,
      "loss": 1.0682,
      "step": 965
    },
    {
      "epoch": 0.9304556354916067,
      "grad_norm": 0.14192393205318143,
      "learning_rate": 2.8996493141741687e-06,
      "loss": 1.0754,
      "step": 970
    },
    {
      "epoch": 0.935251798561151,
      "grad_norm": 0.1483650093466018,
      "learning_rate": 2.5125406683743414e-06,
      "loss": 1.0496,
      "step": 975
    },
    {
      "epoch": 0.9400479616306955,
      "grad_norm": 0.14456885419782456,
      "learning_rate": 2.152828779305793e-06,
      "loss": 1.0789,
      "step": 980
    },
    {
      "epoch": 0.9448441247002398,
      "grad_norm": 0.14797777926339026,
      "learning_rate": 1.8206147362695213e-06,
      "loss": 1.0694,
      "step": 985
    },
    {
      "epoch": 0.9496402877697842,
      "grad_norm": 0.1533259331869082,
      "learning_rate": 1.5159919008874369e-06,
      "loss": 1.0581,
      "step": 990
    },
    {
      "epoch": 0.9544364508393285,
      "grad_norm": 0.14794187907454043,
      "learning_rate": 1.2390458808651083e-06,
      "loss": 1.0805,
      "step": 995
    },
    {
      "epoch": 0.9592326139088729,
      "grad_norm": 0.14668802282059784,
      "learning_rate": 9.898545059335852e-07,
      "loss": 1.056,
      "step": 1000
    },
    {
      "epoch": 0.9640287769784173,
      "grad_norm": 0.14996161626389323,
      "learning_rate": 7.684878059769363e-07,
      "loss": 1.0528,
      "step": 1005
    },
    {
      "epoch": 0.9688249400479616,
      "grad_norm": 0.14311589381161047,
      "learning_rate": 5.750079913519835e-07,
      "loss": 1.0712,
      "step": 1010
    },
    {
      "epoch": 0.973621103117506,
      "grad_norm": 0.14129817073562634,
      "learning_rate": 4.094694354052742e-07,
      "loss": 1.0547,
      "step": 1015
    },
    {
      "epoch": 0.9784172661870504,
      "grad_norm": 0.14555078065518706,
      "learning_rate": 2.7191865919276026e-07,
      "loss": 1.053,
      "step": 1020
    },
    {
      "epoch": 0.9832134292565947,
      "grad_norm": 0.1507162059547356,
      "learning_rate": 1.623943184059229e-07,
      "loss": 1.0469,
      "step": 1025
    },
    {
      "epoch": 0.988009592326139,
      "grad_norm": 0.14248345329368778,
      "learning_rate": 8.092719250853975e-08,
      "loss": 1.0614,
      "step": 1030
    },
    {
      "epoch": 0.9928057553956835,
      "grad_norm": 0.1444089044813381,
      "learning_rate": 2.7540176086671144e-08,
      "loss": 1.081,
      "step": 1035
    },
    {
      "epoch": 0.9976019184652278,
      "grad_norm": 0.14756744866469018,
      "learning_rate": 2.2482724147177005e-09,
      "loss": 1.0966,
      "step": 1040
    },
    {
      "epoch": 0.9995203836930455,
      "eval_loss": 1.0771065950393677,
      "eval_runtime": 1906.9852,
      "eval_samples_per_second": 3.517,
      "eval_steps_per_second": 0.879,
      "step": 1042
    },
    {
      "epoch": 0.9995203836930455,
      "step": 1042,
      "total_flos": 2151475014795264.0,
      "train_loss": 1.0904040080343236,
      "train_runtime": 22004.2311,
      "train_samples_per_second": 3.032,
      "train_steps_per_second": 0.047
    }
  ],
  "logging_steps": 5,
  "max_steps": 1042,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2151475014795264.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}