{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9889162927525035,
  "eval_steps": 500,
  "global_step": 5200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0019017621014471222,
      "grad_norm": 5.274278163909912,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 1.3065,
      "step": 10
    },
    {
      "epoch": 0.0038035242028942443,
      "grad_norm": 6.159559726715088,
      "learning_rate": 9.5e-05,
      "loss": 0.5845,
      "step": 20
    },
    {
      "epoch": 0.005705286304341367,
      "grad_norm": 0.032951805740594864,
      "learning_rate": 0.000145,
      "loss": 0.0535,
      "step": 30
    },
    {
      "epoch": 0.007607048405788489,
      "grad_norm": 0.0016232666093856096,
      "learning_rate": 0.00019500000000000002,
      "loss": 0.0011,
      "step": 40
    },
    {
      "epoch": 0.00950881050723561,
      "grad_norm": 0.03492136672139168,
      "learning_rate": 0.000245,
      "loss": 0.0004,
      "step": 50
    },
    {
      "epoch": 0.011410572608682733,
      "grad_norm": 0.03490574657917023,
      "learning_rate": 0.000295,
      "loss": 0.0002,
      "step": 60
    },
    {
      "epoch": 0.013312334710129855,
      "grad_norm": 0.2781233489513397,
      "learning_rate": 0.000345,
      "loss": 0.0028,
      "step": 70
    },
    {
      "epoch": 0.015214096811576977,
      "grad_norm": 0.2495400309562683,
      "learning_rate": 0.000395,
      "loss": 0.0054,
      "step": 80
    },
    {
      "epoch": 0.0171158589130241,
      "grad_norm": 0.016650637611746788,
      "learning_rate": 0.00044500000000000003,
      "loss": 0.0113,
      "step": 90
    },
    {
      "epoch": 0.01901762101447122,
      "grad_norm": 0.006048521026968956,
      "learning_rate": 0.000495,
      "loss": 0.0129,
      "step": 100
    },
    {
      "epoch": 0.020919383115918344,
      "grad_norm": 0.04650586470961571,
      "learning_rate": 0.0004999995932430571,
      "loss": 0.0531,
      "step": 110
    },
    {
      "epoch": 0.022821145217365467,
      "grad_norm": 0.2665582597255707,
      "learning_rate": 0.0004999981871713734,
      "loss": 0.0106,
      "step": 120
    },
    {
      "epoch": 0.024722907318812586,
      "grad_norm": 0.3457753360271454,
      "learning_rate": 0.0004999957767689057,
      "loss": 0.0111,
      "step": 130
    },
    {
      "epoch": 0.02662466942025971,
      "grad_norm": 0.3711691200733185,
      "learning_rate": 0.0004999923620453374,
      "loss": 0.0089,
      "step": 140
    },
    {
      "epoch": 0.028526431521706832,
      "grad_norm": 0.5576126575469971,
      "learning_rate": 0.0004999879430143867,
      "loss": 0.0225,
      "step": 150
    },
    {
      "epoch": 0.030428193623153955,
      "grad_norm": 0.061603959649801254,
      "learning_rate": 0.0004999825196938062,
      "loss": 0.0066,
      "step": 160
    },
    {
      "epoch": 0.032329955724601074,
      "grad_norm": 0.443553626537323,
      "learning_rate": 0.0004999760921053835,
      "loss": 0.0123,
      "step": 170
    },
    {
      "epoch": 0.0342317178260482,
      "grad_norm": 0.24774852395057678,
      "learning_rate": 0.0004999686602749405,
      "loss": 0.0153,
      "step": 180
    },
    {
      "epoch": 0.03613347992749532,
      "grad_norm": 0.07191673666238785,
      "learning_rate": 0.0004999602242323333,
      "loss": 0.0126,
      "step": 190
    },
    {
      "epoch": 0.03803524202894244,
      "grad_norm": 0.48675355315208435,
      "learning_rate": 0.0004999507840114525,
      "loss": 0.0232,
      "step": 200
    },
    {
      "epoch": 0.039937004130389565,
      "grad_norm": 0.37680506706237793,
      "learning_rate": 0.000499940339650223,
      "loss": 0.0113,
      "step": 210
    },
    {
      "epoch": 0.04183876623183669,
      "grad_norm": 0.4189215302467346,
      "learning_rate": 0.0004999288911906033,
      "loss": 0.0111,
      "step": 220
    },
    {
      "epoch": 0.04374052833328381,
      "grad_norm": 0.5190231204032898,
      "learning_rate": 0.0004999164386785859,
      "loss": 0.0181,
      "step": 230
    },
    {
      "epoch": 0.045642290434730934,
      "grad_norm": 0.3827992379665375,
      "learning_rate": 0.0004999029821641969,
      "loss": 0.0254,
      "step": 240
    },
    {
      "epoch": 0.04754405253617805,
      "grad_norm": 0.3790438771247864,
      "learning_rate": 0.0004998885217014959,
      "loss": 0.0135,
      "step": 250
    },
    {
      "epoch": 0.04944581463762517,
      "grad_norm": 0.32959797978401184,
      "learning_rate": 0.0004998730573485757,
      "loss": 0.02,
      "step": 260
    },
    {
      "epoch": 0.051347576739072295,
      "grad_norm": 0.11981680989265442,
      "learning_rate": 0.0004998565891675621,
      "loss": 0.0274,
      "step": 270
    },
    {
      "epoch": 0.05324933884051942,
      "grad_norm": 0.417461097240448,
      "learning_rate": 0.0004998391172246136,
      "loss": 0.0226,
      "step": 280
    },
    {
      "epoch": 0.05515110094196654,
      "grad_norm": 0.5383813381195068,
      "learning_rate": 0.0004998206415899208,
      "loss": 0.0198,
      "step": 290
    },
    {
      "epoch": 0.057052863043413664,
      "grad_norm": 0.49503782391548157,
      "learning_rate": 0.0004998011623377073,
      "loss": 0.0217,
      "step": 300
    },
    {
      "epoch": 0.05895462514486079,
      "grad_norm": 0.5187065005302429,
      "learning_rate": 0.0004997806795462279,
      "loss": 0.0251,
      "step": 310
    },
    {
      "epoch": 0.06085638724630791,
      "grad_norm": 0.27587956190109253,
      "learning_rate": 0.0004997591932977692,
      "loss": 0.0212,
      "step": 320
    },
    {
      "epoch": 0.06275814934775503,
      "grad_norm": 0.19010399281978607,
      "learning_rate": 0.000499736703678649,
      "loss": 0.0186,
      "step": 330
    },
    {
      "epoch": 0.06465991144920215,
      "grad_norm": 0.28244227170944214,
      "learning_rate": 0.0004997132107792161,
      "loss": 0.0281,
      "step": 340
    },
    {
      "epoch": 0.06656167355064928,
      "grad_norm": 0.3918805420398712,
      "learning_rate": 0.0004996887146938497,
      "loss": 0.0349,
      "step": 350
    },
    {
      "epoch": 0.0684634356520964,
      "grad_norm": 0.3364613354206085,
      "learning_rate": 0.0004996632155209592,
      "loss": 0.0192,
      "step": 360
    },
    {
      "epoch": 0.07036519775354352,
      "grad_norm": 0.20451515913009644,
      "learning_rate": 0.0004996367133629837,
      "loss": 0.0341,
      "step": 370
    },
    {
      "epoch": 0.07226695985499064,
      "grad_norm": 0.5754940509796143,
      "learning_rate": 0.0004996092083263919,
      "loss": 0.0327,
      "step": 380
    },
    {
      "epoch": 0.07416872195643776,
      "grad_norm": 0.5575674176216125,
      "learning_rate": 0.000499580700521681,
      "loss": 0.0407,
      "step": 390
    },
    {
      "epoch": 0.07607048405788489,
      "grad_norm": 0.4820100665092468,
      "learning_rate": 0.0004995511900633771,
      "loss": 0.0441,
      "step": 400
    },
    {
      "epoch": 0.077972246159332,
      "grad_norm": 0.49140122532844543,
      "learning_rate": 0.000499520677070034,
      "loss": 0.036,
      "step": 410
    },
    {
      "epoch": 0.07987400826077913,
      "grad_norm": 0.3462704122066498,
      "learning_rate": 0.0004994891616642331,
      "loss": 0.0325,
      "step": 420
    },
    {
      "epoch": 0.08177577036222625,
      "grad_norm": 0.3936806917190552,
      "learning_rate": 0.000499456643972583,
      "loss": 0.0406,
      "step": 430
    },
    {
      "epoch": 0.08367753246367338,
      "grad_norm": 0.3352107107639313,
      "learning_rate": 0.0004994231241257185,
      "loss": 0.0488,
      "step": 440
    },
    {
      "epoch": 0.08557929456512049,
      "grad_norm": 0.4352121353149414,
      "learning_rate": 0.0004993886022583009,
      "loss": 0.0379,
      "step": 450
    },
    {
      "epoch": 0.08748105666656762,
      "grad_norm": 0.4921802282333374,
      "learning_rate": 0.0004993530785090166,
      "loss": 0.1226,
      "step": 460
    },
    {
      "epoch": 0.08938281876801474,
      "grad_norm": 0.5758520364761353,
      "learning_rate": 0.000499316553020577,
      "loss": 0.0551,
      "step": 470
    },
    {
      "epoch": 0.09128458086946187,
      "grad_norm": 0.38870593905448914,
      "learning_rate": 0.0004992790259397178,
      "loss": 0.0497,
      "step": 480
    },
    {
      "epoch": 0.09318634297090898,
      "grad_norm": 0.361654132604599,
      "learning_rate": 0.0004992404974171985,
      "loss": 0.0599,
      "step": 490
    },
    {
      "epoch": 0.0950881050723561,
      "grad_norm": 0.452127069234848,
      "learning_rate": 0.000499200967607802,
      "loss": 0.0611,
      "step": 500
    },
    {
      "epoch": 0.09698986717380323,
      "grad_norm": 0.5591238737106323,
      "learning_rate": 0.0004991604366703332,
      "loss": 0.0586,
      "step": 510
    },
    {
      "epoch": 0.09889162927525035,
      "grad_norm": 0.40963825583457947,
      "learning_rate": 0.0004991189047676192,
      "loss": 0.048,
      "step": 520
    },
    {
      "epoch": 0.10079339137669747,
      "grad_norm": 0.5524506568908691,
      "learning_rate": 0.0004990763720665083,
      "loss": 0.0497,
      "step": 530
    },
    {
      "epoch": 0.10269515347814459,
      "grad_norm": 0.3613679111003876,
      "learning_rate": 0.0004990328387378695,
      "loss": 0.052,
      "step": 540
    },
    {
      "epoch": 0.10459691557959172,
      "grad_norm": 0.5716497898101807,
      "learning_rate": 0.0004989883049565912,
      "loss": 0.0512,
      "step": 550
    },
    {
      "epoch": 0.10649867768103884,
      "grad_norm": 0.48545169830322266,
      "learning_rate": 0.0004989427709015816,
      "loss": 0.066,
      "step": 560
    },
    {
      "epoch": 0.10840043978248597,
      "grad_norm": 0.3815110921859741,
      "learning_rate": 0.0004988962367557668,
      "loss": 0.0629,
      "step": 570
    },
    {
      "epoch": 0.11030220188393308,
      "grad_norm": 0.40193411707878113,
      "learning_rate": 0.000498848702706091,
      "loss": 0.068,
      "step": 580
    },
    {
      "epoch": 0.1122039639853802,
      "grad_norm": 0.42606788873672485,
      "learning_rate": 0.0004988001689435152,
      "loss": 0.0478,
      "step": 590
    },
    {
      "epoch": 0.11410572608682733,
      "grad_norm": 0.7104797959327698,
      "learning_rate": 0.0004987506356630165,
      "loss": 0.0607,
      "step": 600
    },
    {
      "epoch": 0.11600748818827444,
      "grad_norm": 0.3995862305164337,
      "learning_rate": 0.0004987001030635878,
      "loss": 0.0565,
      "step": 610
    },
    {
      "epoch": 0.11790925028972157,
      "grad_norm": 0.6329849362373352,
      "learning_rate": 0.0004986485713482361,
      "loss": 0.0733,
      "step": 620
    },
    {
      "epoch": 0.11981101239116869,
      "grad_norm": 0.5191243886947632,
      "learning_rate": 0.0004985960407239825,
      "loss": 0.0565,
      "step": 630
    },
    {
      "epoch": 0.12171277449261582,
      "grad_norm": 0.4808266758918762,
      "learning_rate": 0.0004985425114018611,
      "loss": 0.075,
      "step": 640
    },
    {
      "epoch": 0.12361453659406293,
      "grad_norm": 0.49313193559646606,
      "learning_rate": 0.000498487983596918,
      "loss": 0.0637,
      "step": 650
    },
    {
      "epoch": 0.12551629869551006,
      "grad_norm": 0.6945317387580872,
      "learning_rate": 0.0004984324575282107,
      "loss": 0.0778,
      "step": 660
    },
    {
      "epoch": 0.1274180607969572,
      "grad_norm": 0.3684280514717102,
      "learning_rate": 0.0004983759334188068,
      "loss": 0.0577,
      "step": 670
    },
    {
      "epoch": 0.1293198228984043,
      "grad_norm": 0.5096045136451721,
      "learning_rate": 0.0004983184114957836,
      "loss": 0.073,
      "step": 680
    },
    {
      "epoch": 0.13122158499985143,
      "grad_norm": 0.24991731345653534,
      "learning_rate": 0.000498259891990227,
      "loss": 0.079,
      "step": 690
    },
    {
      "epoch": 0.13312334710129856,
      "grad_norm": 0.4371449649333954,
      "learning_rate": 0.0004982003751372306,
      "loss": 0.0814,
      "step": 700
    },
    {
      "epoch": 0.13502510920274566,
      "grad_norm": 0.7052175402641296,
      "learning_rate": 0.0004981398611758942,
      "loss": 0.0813,
      "step": 710
    },
    {
      "epoch": 0.1369268713041928,
      "grad_norm": 0.3389616310596466,
      "learning_rate": 0.0004980783503493241,
      "loss": 0.0806,
      "step": 720
    },
    {
      "epoch": 0.13882863340563992,
      "grad_norm": 0.5502128601074219,
      "learning_rate": 0.0004980158429046306,
      "loss": 0.0623,
      "step": 730
    },
    {
      "epoch": 0.14073039550708705,
      "grad_norm": 0.6836037635803223,
      "learning_rate": 0.0004979523390929285,
      "loss": 0.0701,
      "step": 740
    },
    {
      "epoch": 0.14263215760853415,
      "grad_norm": 0.7100756168365479,
      "learning_rate": 0.0004978878391693346,
      "loss": 0.0812,
      "step": 750
    },
    {
      "epoch": 0.14453391970998128,
      "grad_norm": 0.5102308392524719,
      "learning_rate": 0.000497822343392968,
      "loss": 0.06,
      "step": 760
    },
    {
      "epoch": 0.1464356818114284,
      "grad_norm": 0.4712526202201843,
      "learning_rate": 0.0004977558520269484,
      "loss": 0.0793,
      "step": 770
    },
    {
      "epoch": 0.1483374439128755,
      "grad_norm": 0.7232615947723389,
      "learning_rate": 0.0004976883653383948,
      "loss": 0.0674,
      "step": 780
    },
    {
      "epoch": 0.15023920601432264,
      "grad_norm": 0.5397039651870728,
      "learning_rate": 0.0004976198835984253,
      "loss": 0.0613,
      "step": 790
    },
    {
      "epoch": 0.15214096811576977,
      "grad_norm": 0.5102890133857727,
      "learning_rate": 0.0004975504070821548,
      "loss": 0.0823,
      "step": 800
    },
    {
      "epoch": 0.1540427302172169,
      "grad_norm": 0.5335078239440918,
      "learning_rate": 0.0004974799360686952,
      "loss": 0.0833,
      "step": 810
    },
    {
      "epoch": 0.155944492318664,
      "grad_norm": 0.5386427640914917,
      "learning_rate": 0.0004974084708411535,
      "loss": 0.0736,
      "step": 820
    },
    {
      "epoch": 0.15784625442011113,
      "grad_norm": 0.5704069137573242,
      "learning_rate": 0.0004973360116866303,
      "loss": 0.0788,
      "step": 830
    },
    {
      "epoch": 0.15974801652155826,
      "grad_norm": 0.6669997572898865,
      "learning_rate": 0.0004972625588962199,
      "loss": 0.0713,
      "step": 840
    },
    {
      "epoch": 0.1616497786230054,
      "grad_norm": 0.4686433970928192,
      "learning_rate": 0.000497188112765008,
      "loss": 0.0768,
      "step": 850
    },
    {
      "epoch": 0.1635515407244525,
      "grad_norm": 0.622735321521759,
      "learning_rate": 0.0004971126735920707,
      "loss": 0.0857,
      "step": 860
    },
    {
      "epoch": 0.16545330282589962,
      "grad_norm": 0.4435874819755554,
      "learning_rate": 0.0004970362416804739,
      "loss": 0.108,
      "step": 870
    },
    {
      "epoch": 0.16735506492734675,
      "grad_norm": 0.6319471001625061,
      "learning_rate": 0.0004969588173372716,
      "loss": 0.1079,
      "step": 880
    },
    {
      "epoch": 0.16925682702879385,
      "grad_norm": 0.730019748210907,
      "learning_rate": 0.0004968804008735044,
      "loss": 0.1124,
      "step": 890
    },
    {
      "epoch": 0.17115858913024098,
      "grad_norm": 0.32919904589653015,
      "learning_rate": 0.0004968009926041991,
      "loss": 0.0877,
      "step": 900
    },
    {
      "epoch": 0.17306035123168811,
      "grad_norm": 0.5234276652336121,
      "learning_rate": 0.0004967205928483666,
      "loss": 0.081,
      "step": 910
    },
    {
      "epoch": 0.17496211333313524,
      "grad_norm": 0.5863314270973206,
      "learning_rate": 0.000496639201929001,
      "loss": 0.0709,
      "step": 920
    },
    {
      "epoch": 0.17686387543458235,
      "grad_norm": 0.42288297414779663,
      "learning_rate": 0.0004965568201730783,
      "loss": 0.0923,
      "step": 930
    },
    {
      "epoch": 0.17876563753602948,
      "grad_norm": 0.6558719873428345,
      "learning_rate": 0.0004964734479115552,
      "loss": 0.0787,
      "step": 940
    },
    {
      "epoch": 0.1806673996374766,
      "grad_norm": 0.42140552401542664,
      "learning_rate": 0.0004963890854793673,
      "loss": 0.0836,
      "step": 950
    },
    {
      "epoch": 0.18256916173892374,
      "grad_norm": 0.8134212493896484,
      "learning_rate": 0.0004963037332154281,
      "loss": 0.0949,
      "step": 960
    },
    {
      "epoch": 0.18447092384037084,
      "grad_norm": 0.6248766183853149,
      "learning_rate": 0.0004962173914626279,
      "loss": 0.1023,
      "step": 970
    },
    {
      "epoch": 0.18637268594181797,
      "grad_norm": 0.7062785029411316,
      "learning_rate": 0.0004961300605678318,
      "loss": 0.0872,
      "step": 980
    },
    {
      "epoch": 0.1882744480432651,
      "grad_norm": 0.5721538662910461,
      "learning_rate": 0.000496041740881879,
      "loss": 0.104,
      "step": 990
    },
    {
      "epoch": 0.1901762101447122,
      "grad_norm": 0.6057381629943848,
      "learning_rate": 0.0004959524327595805,
      "loss": 0.071,
      "step": 1000
    },
    {
      "epoch": 0.19207797224615933,
      "grad_norm": 0.5917540192604065,
      "learning_rate": 0.0004958621365597186,
      "loss": 0.0931,
      "step": 1010
    },
    {
      "epoch": 0.19397973434760646,
      "grad_norm": 0.8036534190177917,
      "learning_rate": 0.000495770852645045,
      "loss": 0.1035,
      "step": 1020
    },
    {
      "epoch": 0.1958814964490536,
      "grad_norm": 0.5942894816398621,
      "learning_rate": 0.0004956785813822794,
      "loss": 0.1072,
      "step": 1030
    },
    {
      "epoch": 0.1977832585505007,
      "grad_norm": 0.7127736210823059,
      "learning_rate": 0.0004955853231421077,
      "loss": 0.091,
      "step": 1040
    },
    {
      "epoch": 0.19968502065194782,
      "grad_norm": 0.507114827632904,
      "learning_rate": 0.0004954910782991814,
      "loss": 0.0889,
      "step": 1050
    },
    {
      "epoch": 0.20158678275339495,
      "grad_norm": 0.5299419164657593,
      "learning_rate": 0.000495395847232115,
      "loss": 0.1045,
      "step": 1060
    },
    {
      "epoch": 0.20348854485484205,
      "grad_norm": 0.4970870912075043,
      "learning_rate": 0.0004952996303234854,
      "loss": 0.0869,
      "step": 1070
    },
    {
      "epoch": 0.20539030695628918,
      "grad_norm": 0.7103607654571533,
      "learning_rate": 0.0004952024279598298,
      "loss": 0.0953,
      "step": 1080
    },
    {
      "epoch": 0.2072920690577363,
      "grad_norm": 0.38354578614234924,
      "learning_rate": 0.0004951042405316443,
      "loss": 0.109,
      "step": 1090
    },
    {
      "epoch": 0.20919383115918344,
      "grad_norm": 0.7918898463249207,
      "learning_rate": 0.0004950050684333823,
      "loss": 0.0931,
      "step": 1100
    },
    {
      "epoch": 0.21109559326063054,
      "grad_norm": 0.37249505519866943,
      "learning_rate": 0.0004949049120634532,
      "loss": 0.112,
      "step": 1110
    },
    {
      "epoch": 0.21299735536207767,
      "grad_norm": 0.4019123613834381,
      "learning_rate": 0.0004948037718242204,
      "loss": 0.1038,
      "step": 1120
    },
    {
      "epoch": 0.2148991174635248,
      "grad_norm": 0.5207439661026001,
      "learning_rate": 0.0004947016481219997,
      "loss": 0.1193,
      "step": 1130
    },
    {
      "epoch": 0.21680087956497193,
      "grad_norm": 0.41174229979515076,
      "learning_rate": 0.0004945985413670581,
      "loss": 0.0942,
      "step": 1140
    },
    {
      "epoch": 0.21870264166641903,
      "grad_norm": 0.6560219526290894,
      "learning_rate": 0.000494494451973612,
      "loss": 0.0643,
      "step": 1150
    },
    {
      "epoch": 0.22060440376786616,
      "grad_norm": 0.4203394651412964,
      "learning_rate": 0.0004943893803598247,
      "loss": 0.1187,
      "step": 1160
    },
    {
      "epoch": 0.2225061658693133,
      "grad_norm": 0.6167390942573547,
      "learning_rate": 0.0004942833269478063,
      "loss": 0.0941,
      "step": 1170
    },
    {
      "epoch": 0.2244079279707604,
      "grad_norm": 0.5587171912193298,
      "learning_rate": 0.0004941762921636104,
      "loss": 0.0665,
      "step": 1180
    },
    {
      "epoch": 0.22630969007220753,
      "grad_norm": 0.5269094109535217,
      "learning_rate": 0.0004940682764372336,
      "loss": 0.1143,
      "step": 1190
    },
    {
      "epoch": 0.22821145217365466,
      "grad_norm": 0.5605872273445129,
      "learning_rate": 0.0004939592802026132,
      "loss": 0.0913,
      "step": 1200
    },
    {
      "epoch": 0.23011321427510179,
      "grad_norm": 0.49498802423477173,
      "learning_rate": 0.0004938493038976251,
      "loss": 0.1008,
      "step": 1210
    },
    {
      "epoch": 0.2320149763765489,
      "grad_norm": 0.5630708932876587,
      "learning_rate": 0.0004937383479640834,
      "loss": 0.0885,
      "step": 1220
    },
    {
      "epoch": 0.23391673847799602,
      "grad_norm": 0.5097892880439758,
      "learning_rate": 0.0004936264128477368,
      "loss": 0.0914,
      "step": 1230
    },
    {
      "epoch": 0.23581850057944315,
      "grad_norm": 0.3842105567455292,
      "learning_rate": 0.0004935134989982682,
      "loss": 0.0792,
      "step": 1240
    },
    {
      "epoch": 0.23772026268089025,
      "grad_norm": 0.5472444891929626,
      "learning_rate": 0.0004933996068692922,
      "loss": 0.1069,
      "step": 1250
    },
    {
      "epoch": 0.23962202478233738,
      "grad_norm": 0.32953816652297974,
      "learning_rate": 0.0004932847369183538,
      "loss": 0.0745,
      "step": 1260
    },
    {
      "epoch": 0.2415237868837845,
      "grad_norm": 0.5384835600852966,
      "learning_rate": 0.0004931688896069258,
      "loss": 0.1093,
      "step": 1270
    },
    {
      "epoch": 0.24342554898523164,
      "grad_norm": 0.5735477209091187,
      "learning_rate": 0.000493052065400408,
      "loss": 0.0804,
      "step": 1280
    },
    {
      "epoch": 0.24532731108667874,
      "grad_norm": 0.47209662199020386,
      "learning_rate": 0.000492934264768124,
      "loss": 0.0848,
      "step": 1290
    },
    {
      "epoch": 0.24722907318812587,
      "grad_norm": 0.7762870192527771,
      "learning_rate": 0.0004928154881833208,
      "loss": 0.1005,
      "step": 1300
    },
    {
      "epoch": 0.249130835289573,
      "grad_norm": 0.5226859450340271,
      "learning_rate": 0.0004926957361231655,
      "loss": 0.0874,
      "step": 1310
    },
    {
      "epoch": 0.25103259739102013,
      "grad_norm": 0.6678869128227234,
      "learning_rate": 0.0004925750090687445,
      "loss": 0.0833,
      "step": 1320
    },
    {
      "epoch": 0.25293435949246723,
      "grad_norm": 0.5953888893127441,
      "learning_rate": 0.0004924533075050609,
      "loss": 0.0894,
      "step": 1330
    },
    {
      "epoch": 0.2548361215939144,
      "grad_norm": 0.5249069333076477,
      "learning_rate": 0.0004923306319210327,
      "loss": 0.1101,
      "step": 1340
    },
    {
      "epoch": 0.2567378836953615,
      "grad_norm": 0.5314027667045593,
      "learning_rate": 0.0004922069828094908,
      "loss": 0.0916,
      "step": 1350
    },
    {
      "epoch": 0.2586396457968086,
      "grad_norm": 0.4231284558773041,
      "learning_rate": 0.0004920823606671774,
      "loss": 0.0772,
      "step": 1360
    },
    {
      "epoch": 0.26054140789825575,
      "grad_norm": 0.5623658895492554,
      "learning_rate": 0.0004919567659947435,
      "loss": 0.0908,
      "step": 1370
    },
    {
      "epoch": 0.26244316999970285,
      "grad_norm": 0.5130127668380737,
      "learning_rate": 0.0004918301992967472,
      "loss": 0.0994,
      "step": 1380
    },
    {
      "epoch": 0.26434493210114995,
      "grad_norm": 0.33209264278411865,
      "learning_rate": 0.0004917026610816516,
      "loss": 0.0935,
      "step": 1390
    },
    {
      "epoch": 0.2662466942025971,
      "grad_norm": 0.39504528045654297,
      "learning_rate": 0.0004915741518618222,
      "loss": 0.0931,
      "step": 1400
    },
    {
      "epoch": 0.2681484563040442,
      "grad_norm": 0.48092228174209595,
      "learning_rate": 0.0004914446721535263,
      "loss": 0.1158,
      "step": 1410
    },
    {
      "epoch": 0.2700502184054913,
      "grad_norm": 0.5582765936851501,
      "learning_rate": 0.0004913142224769292,
      "loss": 0.0828,
      "step": 1420
    },
    {
      "epoch": 0.2719519805069385,
      "grad_norm": 0.3209240138530731,
      "learning_rate": 0.0004911828033560934,
      "loss": 0.0905,
      "step": 1430
    },
    {
      "epoch": 0.2738537426083856,
      "grad_norm": 1.9139939546585083,
      "learning_rate": 0.0004910504153189758,
      "loss": 0.1058,
      "step": 1440
    },
    {
      "epoch": 0.2757555047098327,
      "grad_norm": 0.5591135025024414,
      "learning_rate": 0.0004909170588974256,
      "loss": 0.4107,
      "step": 1450
    },
    {
      "epoch": 0.27765726681127983,
      "grad_norm": 0.3873596787452698,
      "learning_rate": 0.0004907827346271826,
      "loss": 0.1475,
      "step": 1460
    },
    {
      "epoch": 0.27955902891272694,
      "grad_norm": 0.5181601047515869,
      "learning_rate": 0.0004906474430478746,
      "loss": 0.113,
      "step": 1470
    },
    {
      "epoch": 0.2814607910141741,
      "grad_norm": 0.3383086025714874,
      "learning_rate": 0.0004905111847030159,
      "loss": 0.5679,
      "step": 1480
    },
    {
      "epoch": 0.2833625531156212,
      "grad_norm": 0.49057555198669434,
      "learning_rate": 0.0004903739601400039,
      "loss": 0.1181,
      "step": 1490
    },
    {
      "epoch": 0.2852643152170683,
      "grad_norm": 0.572108268737793,
      "learning_rate": 0.0004902357699101182,
      "loss": 0.1126,
      "step": 1500
    },
    {
      "epoch": 0.28716607731851546,
      "grad_norm": 0.40325993299484253,
      "learning_rate": 0.0004900966145685176,
      "loss": 0.0911,
      "step": 1510
    },
    {
      "epoch": 0.28906783941996256,
      "grad_norm": 0.40683841705322266,
      "learning_rate": 0.000489956494674238,
      "loss": 0.1002,
      "step": 1520
    },
    {
      "epoch": 0.29096960152140966,
      "grad_norm": 0.5523852705955505,
      "learning_rate": 0.0004898154107901905,
      "loss": 0.1002,
      "step": 1530
    },
    {
      "epoch": 0.2928713636228568,
      "grad_norm": 0.5049010515213013,
      "learning_rate": 0.0004896733634831589,
      "loss": 0.1015,
      "step": 1540
    },
    {
      "epoch": 0.2947731257243039,
      "grad_norm": 0.3494808077812195,
      "learning_rate": 0.0004895303533237969,
      "loss": 0.2635,
      "step": 1550
    },
    {
      "epoch": 0.296674887825751,
      "grad_norm": 0.5694501399993896,
      "learning_rate": 0.000489386380886627,
      "loss": 0.1048,
      "step": 1560
    },
    {
      "epoch": 0.2985766499271982,
      "grad_norm": 0.4427148997783661,
      "learning_rate": 0.0004892414467500371,
      "loss": 0.1254,
      "step": 1570
    },
    {
      "epoch": 0.3004784120286453,
      "grad_norm": 0.28379982709884644,
      "learning_rate": 0.0004890955514962786,
      "loss": 0.2777,
      "step": 1580
    },
    {
      "epoch": 0.30238017413009244,
      "grad_norm": 0.3206646740436554,
      "learning_rate": 0.0004889486957114642,
      "loss": 0.1164,
      "step": 1590
    },
    {
      "epoch": 0.30428193623153954,
      "grad_norm": 0.5560258030891418,
      "learning_rate": 0.0004888008799855655,
      "loss": 0.0796,
      "step": 1600
    },
    {
      "epoch": 0.30618369833298664,
      "grad_norm": 0.43756505846977234,
      "learning_rate": 0.00048865210491241,
      "loss": 0.1102,
      "step": 1610
    },
    {
      "epoch": 0.3080854604344338,
      "grad_norm": 0.4390459954738617,
      "learning_rate": 0.0004885023710896799,
      "loss": 0.1831,
      "step": 1620
    },
    {
      "epoch": 0.3099872225358809,
      "grad_norm": 0.3506847620010376,
      "learning_rate": 0.0004883516791189084,
      "loss": 0.1157,
      "step": 1630
    },
    {
      "epoch": 0.311888984637328,
      "grad_norm": 0.4262866675853729,
      "learning_rate": 0.00048820002960547844,
      "loss": 0.1143,
      "step": 1640
    },
    {
      "epoch": 0.31379074673877516,
      "grad_norm": 0.5669584274291992,
      "learning_rate": 0.0004880474231586195,
      "loss": 0.1037,
      "step": 1650
    },
    {
      "epoch": 0.31569250884022226,
      "grad_norm": 0.5108714699745178,
      "learning_rate": 0.00048789386039140535,
      "loss": 1.2545,
      "step": 1660
    },
    {
      "epoch": 0.31759427094166937,
      "grad_norm": 0.5437518358230591,
      "learning_rate": 0.00048773934192075186,
      "loss": 0.097,
      "step": 1670
    },
    {
      "epoch": 0.3194960330431165,
      "grad_norm": 0.7191082835197449,
      "learning_rate": 0.0004875838683674141,
      "loss": 0.0965,
      "step": 1680
    },
    {
      "epoch": 0.3213977951445636,
      "grad_norm": 0.5136532783508301,
      "learning_rate": 0.00048742744035598407,
      "loss": 0.0818,
      "step": 1690
    },
    {
      "epoch": 0.3232995572460108,
      "grad_norm": 0.5209280848503113,
      "learning_rate": 0.0004872700585148882,
      "loss": 0.0967,
      "step": 1700
    },
    {
      "epoch": 0.3252013193474579,
      "grad_norm": 0.36031627655029297,
      "learning_rate": 0.00048711172347638484,
      "loss": 0.113,
      "step": 1710
    },
    {
      "epoch": 0.327103081448905,
      "grad_norm": 0.39448150992393494,
      "learning_rate": 0.0004869524358765616,
      "loss": 0.0733,
      "step": 1720
    },
    {
      "epoch": 0.32900484355035214,
      "grad_norm": 0.4772893786430359,
      "learning_rate": 0.00048679219635533276,
      "loss": 0.0937,
      "step": 1730
    },
    {
      "epoch": 0.33090660565179925,
      "grad_norm": 0.4700233042240143,
      "learning_rate": 0.0004866310055564371,
      "loss": 0.0874,
      "step": 1740
    },
    {
      "epoch": 0.33280836775324635,
      "grad_norm": 0.444872111082077,
      "learning_rate": 0.00048646886412743475,
      "loss": 0.0994,
      "step": 1750
    },
    {
      "epoch": 0.3347101298546935,
      "grad_norm": 0.28628623485565186,
      "learning_rate": 0.0004863057727197049,
      "loss": 0.097,
      "step": 1760
    },
    {
      "epoch": 0.3366118919561406,
      "grad_norm": 0.3799450695514679,
      "learning_rate": 0.0004861417319884434,
      "loss": 0.0807,
      "step": 1770
    },
    {
      "epoch": 0.3385136540575877,
      "grad_norm": 0.34396281838417053,
      "learning_rate": 0.00048597674259265934,
      "loss": 0.0895,
      "step": 1780
    },
    {
      "epoch": 0.34041541615903487,
      "grad_norm": 0.3908982276916504,
      "learning_rate": 0.0004858108051951735,
      "loss": 0.0872,
      "step": 1790
    },
    {
      "epoch": 0.34231717826048197,
      "grad_norm": 0.35899925231933594,
      "learning_rate": 0.0004856439204626147,
      "loss": 0.0699,
      "step": 1800
    },
    {
      "epoch": 0.3442189403619291,
      "grad_norm": 0.23835837841033936,
      "learning_rate": 0.00048547608906541784,
      "loss": 0.0763,
      "step": 1810
    },
    {
      "epoch": 0.34612070246337623,
      "grad_norm": 0.3790901303291321,
      "learning_rate": 0.0004853073116778207,
      "loss": 0.1081,
      "step": 1820
    },
    {
      "epoch": 0.34802246456482333,
      "grad_norm": 0.4319717288017273,
      "learning_rate": 0.0004851375889778614,
      "loss": 0.0859,
      "step": 1830
    },
    {
      "epoch": 0.3499242266662705,
      "grad_norm": 0.3756701648235321,
      "learning_rate": 0.00048496692164737596,
      "loss": 0.1024,
      "step": 1840
    },
    {
      "epoch": 0.3518259887677176,
      "grad_norm": 0.1727105975151062,
      "learning_rate": 0.0004847953103719951,
      "loss": 0.094,
      "step": 1850
    },
    {
      "epoch": 0.3537277508691647,
      "grad_norm": 0.49954915046691895,
      "learning_rate": 0.0004846227558411417,
      "loss": 0.0923,
      "step": 1860
    },
    {
      "epoch": 0.35562951297061185,
      "grad_norm": 0.3634874224662781,
      "learning_rate": 0.0004844492587480283,
      "loss": 0.0742,
      "step": 1870
    },
    {
      "epoch": 0.35753127507205895,
      "grad_norm": 0.3322870433330536,
      "learning_rate": 0.0004842748197896537,
      "loss": 0.0939,
      "step": 1880
    },
    {
      "epoch": 0.35943303717350605,
      "grad_norm": 0.22122009098529816,
      "learning_rate": 0.00048409943966680057,
      "loss": 0.0669,
      "step": 1890
    },
    {
      "epoch": 0.3613347992749532,
      "grad_norm": 0.5171657204627991,
      "learning_rate": 0.00048392311908403276,
      "loss": 0.0649,
      "step": 1900
    },
    {
      "epoch": 0.3632365613764003,
      "grad_norm": 0.45435401797294617,
      "learning_rate": 0.0004837458587496921,
      "loss": 0.0841,
      "step": 1910
    },
    {
      "epoch": 0.36513832347784747,
      "grad_norm": 0.35136786103248596,
      "learning_rate": 0.00048356765937589597,
      "loss": 0.0915,
      "step": 1920
    },
    {
      "epoch": 0.3670400855792946,
      "grad_norm": 0.5352773666381836,
      "learning_rate": 0.0004833885216785338,
      "loss": 0.0622,
      "step": 1930
    },
    {
      "epoch": 0.3689418476807417,
      "grad_norm": 0.3906407654285431,
      "learning_rate": 0.0004832084463772649,
      "loss": 0.0739,
      "step": 1940
    },
    {
      "epoch": 0.37084360978218883,
      "grad_norm": 0.23543521761894226,
      "learning_rate": 0.0004830274341955152,
      "loss": 0.0766,
      "step": 1950
    },
    {
      "epoch": 0.37274537188363593,
      "grad_norm": 0.31237101554870605,
      "learning_rate": 0.0004828454858604744,
      "loss": 0.0668,
      "step": 1960
    },
    {
      "epoch": 0.37464713398508304,
      "grad_norm": 0.4033367931842804,
      "learning_rate": 0.0004826626021030931,
      "loss": 0.0822,
      "step": 1970
    },
    {
      "epoch": 0.3765488960865302,
      "grad_norm": 0.3153851628303528,
      "learning_rate": 0.0004824787836580797,
      "loss": 0.0506,
      "step": 1980
    },
    {
      "epoch": 0.3784506581879773,
      "grad_norm": 0.37774601578712463,
      "learning_rate": 0.0004822940312638977,
      "loss": 0.0885,
      "step": 1990
    },
    {
      "epoch": 0.3803524202894244,
      "grad_norm": 0.3738980293273926,
      "learning_rate": 0.0004821083456627625,
      "loss": 0.0705,
      "step": 2000
    },
    {
      "epoch": 0.38225418239087156,
      "grad_norm": 0.3323149085044861,
      "learning_rate": 0.00048192172760063866,
      "loss": 0.0765,
      "step": 2010
    },
    {
      "epoch": 0.38415594449231866,
      "grad_norm": 0.38756170868873596,
      "learning_rate": 0.0004817341778272366,
      "loss": 0.0826,
      "step": 2020
    },
    {
      "epoch": 0.38605770659376576,
      "grad_norm": 0.27625781297683716,
      "learning_rate": 0.0004815456970960098,
      "loss": 0.0931,
      "step": 2030
    },
    {
      "epoch": 0.3879594686952129,
      "grad_norm": 0.28700143098831177,
      "learning_rate": 0.00048135628616415184,
      "loss": 0.0782,
      "step": 2040
    },
    {
      "epoch": 0.38986123079666,
      "grad_norm": 0.3487882912158966,
      "learning_rate": 0.0004811659457925931,
      "loss": 0.0834,
      "step": 2050
    },
    {
      "epoch": 0.3917629928981072,
      "grad_norm": 0.3156881630420685,
      "learning_rate": 0.00048097467674599795,
      "loss": 0.076,
      "step": 2060
    },
    {
      "epoch": 0.3936647549995543,
      "grad_norm": 0.28736022114753723,
      "learning_rate": 0.0004807824797927615,
      "loss": 0.0656,
      "step": 2070
    },
    {
      "epoch": 0.3955665171010014,
      "grad_norm": 0.370560884475708,
      "learning_rate": 0.0004805893557050065,
      "loss": 0.0685,
      "step": 2080
    },
    {
      "epoch": 0.39746827920244854,
      "grad_norm": 0.32262367010116577,
      "learning_rate": 0.00048039530525858067,
      "loss": 0.0669,
      "step": 2090
    },
    {
      "epoch": 0.39937004130389564,
      "grad_norm": 0.21956555545330048,
      "learning_rate": 0.00048020032923305284,
      "loss": 0.0827,
      "step": 2100
    },
    {
      "epoch": 0.40127180340534274,
      "grad_norm": 0.295251727104187,
      "learning_rate": 0.0004800044284117104,
      "loss": 0.0557,
      "step": 2110
    },
    {
      "epoch": 0.4031735655067899,
      "grad_norm": 1.090646505355835,
      "learning_rate": 0.00047980760358155616,
      "loss": 0.0861,
      "step": 2120
    },
    {
      "epoch": 0.405075327608237,
      "grad_norm": 0.44668304920196533,
      "learning_rate": 0.0004796098555333045,
      "loss": 0.0726,
      "step": 2130
    },
    {
      "epoch": 0.4069770897096841,
      "grad_norm": 0.3137907385826111,
      "learning_rate": 0.00047941118506137915,
      "loss": 0.0894,
      "step": 2140
    },
    {
      "epoch": 0.40887885181113126,
      "grad_norm": 0.36221790313720703,
      "learning_rate": 0.0004792115929639094,
      "loss": 0.0716,
      "step": 2150
    },
    {
      "epoch": 0.41078061391257836,
      "grad_norm": 0.53058922290802,
      "learning_rate": 0.000479011080042727,
      "loss": 0.0684,
      "step": 2160
    },
    {
      "epoch": 0.4126823760140255,
      "grad_norm": 0.23870202898979187,
      "learning_rate": 0.0004788096471033629,
      "loss": 0.0834,
      "step": 2170
    },
    {
      "epoch": 0.4145841381154726,
      "grad_norm": 0.27887409925460815,
      "learning_rate": 0.0004786072949550443,
      "loss": 0.0614,
      "step": 2180
    },
    {
      "epoch": 0.4164859002169197,
      "grad_norm": 0.5123117566108704,
      "learning_rate": 0.0004784040244106909,
      "loss": 0.0667,
      "step": 2190
    },
    {
      "epoch": 0.4183876623183669,
      "grad_norm": 0.2506055533885956,
      "learning_rate": 0.0004781998362869123,
      "loss": 0.0651,
      "step": 2200
    },
    {
      "epoch": 0.420289424419814,
      "grad_norm": 0.48434486985206604,
      "learning_rate": 0.0004779947314040039,
      "loss": 0.0663,
      "step": 2210
    },
    {
      "epoch": 0.4221911865212611,
      "grad_norm": 0.3052046597003937,
      "learning_rate": 0.0004777887105859444,
      "loss": 0.0866,
      "step": 2220
    },
    {
      "epoch": 0.42409294862270824,
      "grad_norm": 0.4667761027812958,
      "learning_rate": 0.00047758177466039197,
      "loss": 0.1044,
      "step": 2230
    },
    {
      "epoch": 0.42599471072415535,
      "grad_norm": 0.25950753688812256,
      "learning_rate": 0.0004773739244586812,
      "loss": 0.0998,
      "step": 2240
    },
    {
      "epoch": 0.42789647282560245,
      "grad_norm": 0.38734811544418335,
      "learning_rate": 0.0004771651608158194,
      "loss": 0.0791,
      "step": 2250
    },
    {
      "epoch": 0.4297982349270496,
      "grad_norm": 0.5071670413017273,
      "learning_rate": 0.0004769554845704838,
      "loss": 0.072,
      "step": 2260
    },
    {
      "epoch": 0.4316999970284967,
      "grad_norm": 0.32346102595329285,
      "learning_rate": 0.00047674489656501773,
      "loss": 0.0563,
      "step": 2270
    },
    {
      "epoch": 0.43360175912994386,
      "grad_norm": 0.20895633101463318,
      "learning_rate": 0.0004765333976454273,
      "loss": 0.0642,
      "step": 2280
    },
    {
      "epoch": 0.43550352123139097,
      "grad_norm": 0.3482305407524109,
      "learning_rate": 0.00047632098866137826,
      "loss": 0.0716,
      "step": 2290
    },
    {
      "epoch": 0.43740528333283807,
      "grad_norm": 0.2852407991886139,
      "learning_rate": 0.00047610767046619225,
      "loss": 0.0745,
      "step": 2300
    },
    {
      "epoch": 0.4393070454342852,
      "grad_norm": 0.45966169238090515,
      "learning_rate": 0.0004758934439168436,
      "loss": 0.0741,
      "step": 2310
    },
    {
      "epoch": 0.44120880753573233,
      "grad_norm": 0.21411827206611633,
      "learning_rate": 0.00047567830987395597,
      "loss": 0.0598,
      "step": 2320
    },
    {
      "epoch": 0.44311056963717943,
      "grad_norm": 0.36762118339538574,
      "learning_rate": 0.0004754622692017985,
      "loss": 0.0837,
      "step": 2330
    },
    {
      "epoch": 0.4450123317386266,
      "grad_norm": 0.2668420076370239,
      "learning_rate": 0.0004752453227682827,
      "loss": 0.0745,
      "step": 2340
    },
    {
      "epoch": 0.4469140938400737,
      "grad_norm": 0.3845922350883484,
      "learning_rate": 0.00047502747144495875,
      "loss": 0.0747,
      "step": 2350
    },
    {
      "epoch": 0.4488158559415208,
      "grad_norm": 0.38041210174560547,
      "learning_rate": 0.00047480871610701213,
      "loss": 0.0743,
      "step": 2360
    },
    {
      "epoch": 0.45071761804296795,
      "grad_norm": 0.25837719440460205,
      "learning_rate": 0.00047458905763326023,
      "loss": 0.0689,
      "step": 2370
    },
    {
      "epoch": 0.45261938014441505,
      "grad_norm": 0.32403454184532166,
      "learning_rate": 0.0004743684969061484,
      "loss": 0.0669,
      "step": 2380
    },
    {
      "epoch": 0.4545211422458622,
      "grad_norm": 0.4562000334262848,
      "learning_rate": 0.00047414703481174687,
      "loss": 0.0888,
      "step": 2390
    },
    {
      "epoch": 0.4564229043473093,
      "grad_norm": 0.22524714469909668,
      "learning_rate": 0.00047392467223974686,
      "loss": 0.0748,
      "step": 2400
    },
    {
      "epoch": 0.4583246664487564,
      "grad_norm": 0.33924999833106995,
      "learning_rate": 0.0004737014100834571,
      "loss": 0.0757,
      "step": 2410
    },
    {
      "epoch": 0.46022642855020357,
      "grad_norm": 0.26549142599105835,
      "learning_rate": 0.0004734772492398003,
      "loss": 0.0695,
      "step": 2420
    },
    {
      "epoch": 0.46212819065165067,
      "grad_norm": 0.4339611530303955,
      "learning_rate": 0.0004732521906093097,
      "loss": 0.0868,
      "step": 2430
    },
    {
      "epoch": 0.4640299527530978,
      "grad_norm": 0.3820763826370239,
      "learning_rate": 0.00047302623509612484,
      "loss": 0.0687,
      "step": 2440
    },
    {
      "epoch": 0.46593171485454493,
      "grad_norm": 0.2838119864463806,
      "learning_rate": 0.00047279938360798884,
      "loss": 0.0636,
      "step": 2450
    },
    {
      "epoch": 0.46783347695599203,
      "grad_norm": 0.26066961884498596,
      "learning_rate": 0.00047257163705624394,
      "loss": 0.0555,
      "step": 2460
    },
    {
      "epoch": 0.46973523905743914,
      "grad_norm": 0.23981156945228577,
      "learning_rate": 0.00047234299635582835,
      "loss": 0.0606,
      "step": 2470
    },
    {
      "epoch": 0.4716370011588863,
      "grad_norm": 0.3982088267803192,
      "learning_rate": 0.0004721134624252722,
      "loss": 0.0631,
      "step": 2480
    },
    {
      "epoch": 0.4735387632603334,
      "grad_norm": 0.24350067973136902,
      "learning_rate": 0.00047188303618669414,
      "loss": 0.0664,
      "step": 2490
    },
    {
      "epoch": 0.4754405253617805,
      "grad_norm": 0.2303943932056427,
      "learning_rate": 0.0004716517185657977,
      "loss": 0.0665,
      "step": 2500
    },
    {
      "epoch": 0.47734228746322765,
      "grad_norm": 0.23348209261894226,
      "learning_rate": 0.00047141951049186703,
      "loss": 0.0572,
      "step": 2510
    },
    {
      "epoch": 0.47924404956467476,
      "grad_norm": 0.24423716962337494,
      "learning_rate": 0.00047118641289776395,
      "loss": 0.099,
      "step": 2520
    },
    {
      "epoch": 0.4811458116661219,
      "grad_norm": 0.371750146150589,
      "learning_rate": 0.00047095242671992346,
      "loss": 0.064,
      "step": 2530
    },
    {
      "epoch": 0.483047573767569,
      "grad_norm": 0.45275604724884033,
      "learning_rate": 0.0004707175528983506,
      "loss": 0.0745,
      "step": 2540
    },
    {
      "epoch": 0.4849493358690161,
      "grad_norm": 0.23620209097862244,
      "learning_rate": 0.00047048179237661617,
      "loss": 0.0594,
      "step": 2550
    },
    {
      "epoch": 0.4868510979704633,
      "grad_norm": 0.13869664072990417,
      "learning_rate": 0.00047024514610185316,
      "loss": 0.0575,
      "step": 2560
    },
    {
      "epoch": 0.4887528600719104,
      "grad_norm": 0.17631912231445312,
      "learning_rate": 0.00047000761502475317,
      "loss": 0.0434,
      "step": 2570
    },
    {
      "epoch": 0.4906546221733575,
      "grad_norm": 0.30484968423843384,
      "learning_rate": 0.0004697692000995621,
      "loss": 0.0746,
      "step": 2580
    },
    {
      "epoch": 0.49255638427480464,
      "grad_norm": 0.24606972932815552,
      "learning_rate": 0.0004695299022840768,
      "loss": 0.0639,
      "step": 2590
    },
    {
      "epoch": 0.49445814637625174,
      "grad_norm": 0.36455589532852173,
      "learning_rate": 0.00046928972253964087,
      "loss": 0.0677,
      "step": 2600
    },
    {
      "epoch": 0.49635990847769884,
      "grad_norm": 0.20996150374412537,
      "learning_rate": 0.0004690486618311408,
      "loss": 0.0648,
      "step": 2610
    },
    {
      "epoch": 0.498261670579146,
      "grad_norm": 0.12294875085353851,
      "learning_rate": 0.0004688067211270025,
      "loss": 0.0561,
      "step": 2620
    },
    {
      "epoch": 0.5001634326805932,
      "grad_norm": 0.3342492878437042,
      "learning_rate": 0.00046856390139918703,
      "loss": 0.0627,
      "step": 2630
    },
    {
      "epoch": 0.5020651947820403,
      "grad_norm": 0.28927573561668396,
      "learning_rate": 0.00046832020362318677,
      "loss": 0.0947,
      "step": 2640
    },
    {
      "epoch": 0.5039669568834874,
      "grad_norm": 0.2924613952636719,
      "learning_rate": 0.00046807562877802144,
      "loss": 0.036,
      "step": 2650
    },
    {
      "epoch": 0.5058687189849345,
      "grad_norm": 0.22599062323570251,
      "learning_rate": 0.0004678301778462344,
      "loss": 0.0671,
      "step": 2660
    },
    {
      "epoch": 0.5077704810863816,
      "grad_norm": 0.40285536646842957,
      "learning_rate": 0.0004675838518138885,
      "loss": 0.0557,
      "step": 2670
    },
    {
      "epoch": 0.5096722431878288,
      "grad_norm": 0.35334137082099915,
      "learning_rate": 0.0004673366516705623,
      "loss": 0.0639,
      "step": 2680
    },
    {
      "epoch": 0.5115740052892759,
      "grad_norm": 0.1980566680431366,
      "learning_rate": 0.00046708857840934564,
      "loss": 0.0539,
      "step": 2690
    },
    {
      "epoch": 0.513475767390723,
      "grad_norm": 0.23285962641239166,
      "learning_rate": 0.0004668396330268364,
      "loss": 0.0646,
      "step": 2700
    },
    {
      "epoch": 0.5153775294921701,
      "grad_norm": 0.23495745658874512,
      "learning_rate": 0.00046658981652313573,
      "loss": 0.0585,
      "step": 2710
    },
    {
      "epoch": 0.5172792915936172,
      "grad_norm": 0.18733032047748566,
      "learning_rate": 0.00046633912990184457,
      "loss": 0.0551,
      "step": 2720
    },
    {
      "epoch": 0.5191810536950643,
      "grad_norm": 0.23610015213489532,
      "learning_rate": 0.00046608757417005944,
      "loss": 0.0543,
      "step": 2730
    },
    {
      "epoch": 0.5210828157965115,
      "grad_norm": 0.35595059394836426,
      "learning_rate": 0.00046583515033836833,
      "loss": 0.0633,
      "step": 2740
    },
    {
      "epoch": 0.5229845778979586,
      "grad_norm": 0.23900267481803894,
      "learning_rate": 0.00046558185942084657,
      "loss": 0.0483,
      "step": 2750
    },
    {
      "epoch": 0.5248863399994057,
      "grad_norm": 0.362981915473938,
      "learning_rate": 0.00046532770243505313,
      "loss": 0.0462,
      "step": 2760
    },
    {
      "epoch": 0.5267881021008528,
      "grad_norm": 0.5131031274795532,
      "learning_rate": 0.0004650726804020261,
      "loss": 0.055,
      "step": 2770
    },
    {
      "epoch": 0.5286898642022999,
      "grad_norm": 0.22358974814414978,
      "learning_rate": 0.00046481679434627886,
      "loss": 0.0505,
      "step": 2780
    },
    {
      "epoch": 0.530591626303747,
      "grad_norm": 0.4018985331058502,
      "learning_rate": 0.00046456004529579574,
      "loss": 0.065,
      "step": 2790
    },
    {
      "epoch": 0.5324933884051942,
      "grad_norm": 0.14427751302719116,
      "learning_rate": 0.00046430243428202824,
      "loss": 0.0566,
      "step": 2800
    },
    {
      "epoch": 0.5343951505066413,
      "grad_norm": 0.2335805743932724,
      "learning_rate": 0.00046404396233989053,
      "loss": 0.0582,
      "step": 2810
    },
    {
      "epoch": 0.5362969126080884,
      "grad_norm": 0.20551510155200958,
      "learning_rate": 0.00046378463050775534,
      "loss": 0.069,
      "step": 2820
    },
    {
      "epoch": 0.5381986747095355,
      "grad_norm": 0.2913135588169098,
      "learning_rate": 0.0004635244398274501,
      "loss": 0.0787,
      "step": 2830
    },
    {
      "epoch": 0.5401004368109826,
      "grad_norm": 0.25064730644226074,
      "learning_rate": 0.0004632633913442524,
      "loss": 0.0473,
      "step": 2840
    },
    {
      "epoch": 0.5420021989124298,
      "grad_norm": 0.24504418671131134,
      "learning_rate": 0.0004630014861068861,
      "loss": 0.0724,
      "step": 2850
    },
    {
      "epoch": 0.543903961013877,
      "grad_norm": 0.14822843670845032,
      "learning_rate": 0.00046273872516751645,
      "loss": 0.057,
      "step": 2860
    },
    {
      "epoch": 0.545805723115324,
      "grad_norm": 0.3440416753292084,
      "learning_rate": 0.0004624751095817471,
      "loss": 0.0518,
      "step": 2870
    },
    {
      "epoch": 0.5477074852167712,
      "grad_norm": 0.3021232485771179,
      "learning_rate": 0.0004622106404086144,
      "loss": 0.0353,
      "step": 2880
    },
    {
      "epoch": 0.5496092473182183,
      "grad_norm": 0.27382078766822815,
      "learning_rate": 0.00046194531871058435,
      "loss": 0.0517,
      "step": 2890
    },
    {
      "epoch": 0.5515110094196654,
      "grad_norm": 0.2883986830711365,
      "learning_rate": 0.0004616791455535477,
      "loss": 0.0526,
      "step": 2900
    },
    {
      "epoch": 0.5534127715211126,
      "grad_norm": 0.1981579065322876,
      "learning_rate": 0.0004614121220068157,
      "loss": 0.0441,
      "step": 2910
    },
    {
      "epoch": 0.5553145336225597,
      "grad_norm": 0.133266881108284,
      "learning_rate": 0.000461144249143116,
      "loss": 0.0529,
      "step": 2920
    },
    {
      "epoch": 0.5572162957240068,
      "grad_norm": 0.2734558880329132,
      "learning_rate": 0.0004608755280385883,
      "loss": 0.0482,
      "step": 2930
    },
    {
      "epoch": 0.5591180578254539,
      "grad_norm": 0.1929953396320343,
      "learning_rate": 0.00046060595977277997,
      "loss": 0.0545,
      "step": 2940
    },
    {
      "epoch": 0.561019819926901,
      "grad_norm": 0.34874358773231506,
      "learning_rate": 0.00046033554542864157,
      "loss": 0.0563,
      "step": 2950
    },
    {
      "epoch": 0.5629215820283482,
      "grad_norm": 0.2735145390033722,
      "learning_rate": 0.00046006428609252293,
      "loss": 0.0534,
      "step": 2960
    },
    {
      "epoch": 0.5648233441297953,
      "grad_norm": 0.1838793009519577,
      "learning_rate": 0.0004597921828541682,
      "loss": 0.04,
      "step": 2970
    },
    {
      "epoch": 0.5667251062312424,
      "grad_norm": 0.2556362748146057,
      "learning_rate": 0.00045951923680671213,
      "loss": 0.0659,
      "step": 2980
    },
    {
      "epoch": 0.5686268683326895,
      "grad_norm": 0.2070973515510559,
      "learning_rate": 0.000459245449046675,
      "loss": 0.0449,
      "step": 2990
    },
    {
      "epoch": 0.5705286304341366,
      "grad_norm": 0.31435173749923706,
      "learning_rate": 0.0004589708206739587,
      "loss": 0.0668,
      "step": 3000
    },
    {
      "epoch": 0.5724303925355837,
      "grad_norm": 0.2508755922317505,
      "learning_rate": 0.0004586953527918422,
      "loss": 0.0538,
      "step": 3010
    },
    {
      "epoch": 0.5743321546370309,
      "grad_norm": 0.11883655935525894,
      "learning_rate": 0.000458419046506977,
      "loss": 0.0414,
      "step": 3020
    },
    {
      "epoch": 0.576233916738478,
      "grad_norm": 0.08662758767604828,
      "learning_rate": 0.0004581419029293828,
      "loss": 0.0317,
      "step": 3030
    },
    {
      "epoch": 0.5781356788399251,
      "grad_norm": 0.22283309698104858,
      "learning_rate": 0.0004578639231724429,
      "loss": 0.0392,
      "step": 3040
    },
    {
      "epoch": 0.5800374409413722,
      "grad_norm": 0.25671476125717163,
      "learning_rate": 0.0004575851083529,
      "loss": 0.0375,
      "step": 3050
    },
    {
      "epoch": 0.5819392030428193,
      "grad_norm": 0.2501511871814728,
      "learning_rate": 0.0004573054595908514,
      "loss": 0.0367,
      "step": 3060
    },
    {
      "epoch": 0.5838409651442665,
      "grad_norm": 0.196889728307724,
      "learning_rate": 0.00045702497800974474,
      "loss": 0.0887,
      "step": 3070
    },
    {
      "epoch": 0.5857427272457136,
      "grad_norm": 0.28852781653404236,
      "learning_rate": 0.00045674366473637317,
      "loss": 0.0511,
      "step": 3080
    },
    {
      "epoch": 0.5876444893471607,
      "grad_norm": 0.20641787350177765,
      "learning_rate": 0.00045646152090087145,
      "loss": 0.0594,
      "step": 3090
    },
    {
      "epoch": 0.5895462514486078,
      "grad_norm": 0.1833396703004837,
      "learning_rate": 0.0004561785476367106,
      "loss": 0.0476,
      "step": 3100
    },
    {
      "epoch": 0.5914480135500549,
      "grad_norm": 0.3527635335922241,
      "learning_rate": 0.0004558947460806939,
      "loss": 0.0429,
      "step": 3110
    },
    {
      "epoch": 0.593349775651502,
      "grad_norm": 0.2580285668373108,
      "learning_rate": 0.00045561011737295235,
      "loss": 0.0307,
      "step": 3120
    },
    {
      "epoch": 0.5952515377529493,
      "grad_norm": 0.21790006756782532,
      "learning_rate": 0.0004553246626569395,
      "loss": 0.0314,
      "step": 3130
    },
    {
      "epoch": 0.5971532998543964,
      "grad_norm": 0.30504387617111206,
      "learning_rate": 0.00045503838307942756,
      "loss": 0.0398,
      "step": 3140
    },
    {
      "epoch": 0.5990550619558435,
      "grad_norm": 0.24918900430202484,
      "learning_rate": 0.00045475127979050254,
      "loss": 0.0473,
      "step": 3150
    },
    {
      "epoch": 0.6009568240572906,
      "grad_norm": 0.30492204427719116,
      "learning_rate": 0.00045446335394355947,
      "loss": 0.0536,
      "step": 3160
    },
    {
      "epoch": 0.6028585861587377,
      "grad_norm": 0.23205405473709106,
      "learning_rate": 0.0004541746066952978,
      "loss": 0.0219,
      "step": 3170
    },
    {
      "epoch": 0.6047603482601849,
      "grad_norm": 0.25788795948028564,
      "learning_rate": 0.000453885039205717,
      "loss": 0.0458,
      "step": 3180
    },
    {
      "epoch": 0.606662110361632,
      "grad_norm": 0.14208689332008362,
      "learning_rate": 0.0004535946526381117,
      "loss": 0.0583,
      "step": 3190
    },
    {
      "epoch": 0.6085638724630791,
      "grad_norm": 0.3792324364185333,
      "learning_rate": 0.0004533034481590671,
      "loss": 0.0543,
      "step": 3200
    },
    {
      "epoch": 0.6104656345645262,
      "grad_norm": 0.11220109462738037,
      "learning_rate": 0.00045301142693845406,
      "loss": 0.06,
      "step": 3210
    },
    {
      "epoch": 0.6123673966659733,
      "grad_norm": 0.37494245171546936,
      "learning_rate": 0.0004527185901494247,
      "loss": 0.0333,
      "step": 3220
    },
    {
      "epoch": 0.6142691587674204,
      "grad_norm": 0.181362122297287,
      "learning_rate": 0.0004524249389684075,
      "loss": 0.0413,
      "step": 3230
    },
    {
      "epoch": 0.6161709208688676,
      "grad_norm": 0.1204942986369133,
      "learning_rate": 0.0004521304745751029,
      "loss": 0.0606,
      "step": 3240
    },
    {
      "epoch": 0.6180726829703147,
      "grad_norm": 0.28631624579429626,
      "learning_rate": 0.00045183519815247803,
      "loss": 0.0511,
      "step": 3250
    },
    {
      "epoch": 0.6199744450717618,
      "grad_norm": 0.3006780743598938,
      "learning_rate": 0.00045153911088676216,
      "loss": 0.0603,
      "step": 3260
    },
    {
      "epoch": 0.6218762071732089,
      "grad_norm": 0.20401684939861298,
      "learning_rate": 0.00045124221396744226,
      "loss": 0.0462,
      "step": 3270
    },
    {
      "epoch": 0.623777969274656,
      "grad_norm": 0.12001892924308777,
      "learning_rate": 0.00045094450858725775,
      "loss": 0.0624,
      "step": 3280
    },
    {
      "epoch": 0.6256797313761032,
      "grad_norm": 0.15589171648025513,
      "learning_rate": 0.0004506459959421962,
      "loss": 0.0518,
      "step": 3290
    },
    {
      "epoch": 0.6275814934775503,
      "grad_norm": 0.24381360411643982,
      "learning_rate": 0.0004503466772314878,
      "loss": 0.0516,
      "step": 3300
    },
    {
      "epoch": 0.6294832555789974,
      "grad_norm": 0.24374118447303772,
      "learning_rate": 0.0004500465536576015,
      "loss": 0.0428,
      "step": 3310
    },
    {
      "epoch": 0.6313850176804445,
      "grad_norm": 0.25958311557769775,
      "learning_rate": 0.00044974562642623926,
      "loss": 0.0385,
      "step": 3320
    },
    {
      "epoch": 0.6332867797818916,
      "grad_norm": 0.20062792301177979,
      "learning_rate": 0.0004494438967463318,
      "loss": 0.0327,
      "step": 3330
    },
    {
      "epoch": 0.6351885418833387,
      "grad_norm": 0.21975015103816986,
      "learning_rate": 0.0004491413658300336,
      "loss": 0.0476,
      "step": 3340
    },
    {
      "epoch": 0.6370903039847859,
      "grad_norm": 0.2386341094970703,
      "learning_rate": 0.00044883803489271785,
      "loss": 0.052,
      "step": 3350
    },
    {
      "epoch": 0.638992066086233,
      "grad_norm": 0.24763132631778717,
      "learning_rate": 0.00044853390515297176,
      "loss": 0.0552,
      "step": 3360
    },
    {
      "epoch": 0.6408938281876801,
      "grad_norm": 0.24076853692531586,
      "learning_rate": 0.00044822897783259176,
      "loss": 0.0456,
      "step": 3370
    },
    {
      "epoch": 0.6427955902891272,
      "grad_norm": 0.22434252500534058,
      "learning_rate": 0.0004479232541565782,
      "loss": 0.0467,
      "step": 3380
    },
    {
      "epoch": 0.6446973523905744,
      "grad_norm": 0.12866199016571045,
      "learning_rate": 0.00044761673535313084,
      "loss": 0.0454,
      "step": 3390
    },
    {
      "epoch": 0.6465991144920216,
      "grad_norm": 0.2736774682998657,
      "learning_rate": 0.0004473094226536436,
      "loss": 0.0451,
      "step": 3400
    },
    {
      "epoch": 0.6485008765934687,
      "grad_norm": 0.29224392771720886,
      "learning_rate": 0.0004470013172926999,
      "loss": 0.0508,
      "step": 3410
    },
    {
      "epoch": 0.6504026386949158,
      "grad_norm": 0.23447951674461365,
      "learning_rate": 0.00044669242050806766,
      "loss": 0.0479,
      "step": 3420
    },
    {
      "epoch": 0.6523044007963629,
      "grad_norm": 0.2928425967693329,
      "learning_rate": 0.0004463827335406938,
      "loss": 0.0266,
      "step": 3430
    },
    {
      "epoch": 0.65420616289781,
      "grad_norm": 0.282113641500473,
      "learning_rate": 0.0004460722576347002,
      "loss": 0.0442,
      "step": 3440
    },
    {
      "epoch": 0.6561079249992571,
      "grad_norm": 0.24653756618499756,
      "learning_rate": 0.0004457609940373777,
      "loss": 0.0392,
      "step": 3450
    },
    {
      "epoch": 0.6580096871007043,
      "grad_norm": 0.2997114658355713,
      "learning_rate": 0.00044544894399918214,
      "loss": 0.0477,
      "step": 3460
    },
    {
      "epoch": 0.6599114492021514,
      "grad_norm": 0.08997475355863571,
      "learning_rate": 0.00044513610877372814,
      "loss": 0.0424,
      "step": 3470
    },
    {
      "epoch": 0.6618132113035985,
      "grad_norm": 0.18397915363311768,
      "learning_rate": 0.00044482248961778516,
      "loss": 0.0339,
      "step": 3480
    },
    {
      "epoch": 0.6637149734050456,
      "grad_norm": 0.18064412474632263,
      "learning_rate": 0.00044450808779127185,
      "loss": 0.0437,
      "step": 3490
    },
    {
      "epoch": 0.6656167355064927,
      "grad_norm": 0.18558929860591888,
      "learning_rate": 0.00044419290455725103,
      "loss": 0.0554,
      "step": 3500
    },
    {
      "epoch": 0.6675184976079399,
      "grad_norm": 0.3115510940551758,
      "learning_rate": 0.00044387694118192477,
      "loss": 0.0458,
      "step": 3510
    },
    {
      "epoch": 0.669420259709387,
      "grad_norm": 0.4627574384212494,
      "learning_rate": 0.0004435601989346293,
      "loss": 0.0463,
      "step": 3520
    },
    {
      "epoch": 0.6713220218108341,
      "grad_norm": 0.17065434157848358,
      "learning_rate": 0.0004432426790878298,
      "loss": 0.0457,
      "step": 3530
    },
    {
      "epoch": 0.6732237839122812,
      "grad_norm": 0.3017977178096771,
      "learning_rate": 0.00044292438291711536,
      "loss": 0.0452,
      "step": 3540
    },
    {
      "epoch": 0.6751255460137283,
      "grad_norm": 0.21006427705287933,
      "learning_rate": 0.00044260531170119377,
      "loss": 0.041,
      "step": 3550
    },
    {
      "epoch": 0.6770273081151754,
      "grad_norm": 0.14174698293209076,
      "learning_rate": 0.0004422854667218865,
      "loss": 0.0421,
      "step": 3560
    },
    {
      "epoch": 0.6789290702166226,
      "grad_norm": 0.21689032018184662,
      "learning_rate": 0.00044196484926412364,
      "loss": 0.0426,
      "step": 3570
    },
    {
      "epoch": 0.6808308323180697,
      "grad_norm": 0.25217998027801514,
      "learning_rate": 0.00044164346061593827,
      "loss": 0.0616,
      "step": 3580
    },
    {
      "epoch": 0.6827325944195168,
      "grad_norm": 0.07392685115337372,
      "learning_rate": 0.0004413213020684619,
      "loss": 0.0366,
      "step": 3590
    },
    {
      "epoch": 0.6846343565209639,
      "grad_norm": 0.2156752347946167,
      "learning_rate": 0.0004409983749159189,
      "loss": 0.0352,
      "step": 3600
    },
    {
      "epoch": 0.686536118622411,
      "grad_norm": 0.007870941422879696,
      "learning_rate": 0.0004406746804556214,
      "loss": 0.0342,
      "step": 3610
    },
    {
      "epoch": 0.6884378807238583,
      "grad_norm": 0.18450793623924255,
      "learning_rate": 0.000440350219987964,
      "loss": 0.0292,
      "step": 3620
    },
    {
      "epoch": 0.6903396428253054,
      "grad_norm": 0.3035867214202881,
      "learning_rate": 0.00044002499481641876,
      "loss": 0.0439,
      "step": 3630
    },
    {
      "epoch": 0.6922414049267525,
      "grad_norm": 0.16944675147533417,
      "learning_rate": 0.0004396990062475299,
      "loss": 0.0368,
      "step": 3640
    },
    {
      "epoch": 0.6941431670281996,
      "grad_norm": 0.17465433478355408,
      "learning_rate": 0.0004393722555909081,
      "loss": 0.0512,
      "step": 3650
    },
    {
      "epoch": 0.6960449291296467,
      "grad_norm": 0.4197070002555847,
      "learning_rate": 0.0004390447441592259,
      "loss": 0.0494,
      "step": 3660
    },
    {
      "epoch": 0.6979466912310938,
      "grad_norm": 0.21269617974758148,
      "learning_rate": 0.0004387164732682122,
      "loss": 0.0432,
      "step": 3670
    },
    {
      "epoch": 0.699848453332541,
      "grad_norm": 0.3487708270549774,
      "learning_rate": 0.0004383874442366468,
      "loss": 0.0405,
      "step": 3680
    },
    {
      "epoch": 0.7017502154339881,
      "grad_norm": 0.3449212908744812,
      "learning_rate": 0.0004380576583863551,
      "loss": 0.0604,
      "step": 3690
    },
    {
      "epoch": 0.7036519775354352,
      "grad_norm": 0.1568797379732132,
      "learning_rate": 0.0004377271170422031,
      "loss": 0.0493,
      "step": 3700
    },
    {
      "epoch": 0.7055537396368823,
      "grad_norm": 0.18511539697647095,
      "learning_rate": 0.0004373958215320918,
      "loss": 0.0313,
      "step": 3710
    },
    {
      "epoch": 0.7074555017383294,
      "grad_norm": 0.12269023805856705,
      "learning_rate": 0.0004370637731869521,
      "loss": 0.0343,
      "step": 3720
    },
    {
      "epoch": 0.7093572638397766,
      "grad_norm": 0.1673561930656433,
      "learning_rate": 0.000436730973340739,
      "loss": 0.033,
      "step": 3730
    },
    {
      "epoch": 0.7112590259412237,
      "grad_norm": 0.3525213301181793,
      "learning_rate": 0.00043639742333042675,
      "loss": 0.1303,
      "step": 3740
    },
    {
      "epoch": 0.7131607880426708,
      "grad_norm": 0.33365005254745483,
      "learning_rate": 0.00043606312449600334,
      "loss": 0.1418,
      "step": 3750
    },
    {
      "epoch": 0.7150625501441179,
      "grad_norm": 0.3005249500274658,
      "learning_rate": 0.00043572807818046484,
      "loss": 0.0664,
      "step": 3760
    },
    {
      "epoch": 0.716964312245565,
      "grad_norm": 0.4880094826221466,
      "learning_rate": 0.00043539228572981036,
      "loss": 0.0788,
      "step": 3770
    },
    {
      "epoch": 0.7188660743470121,
      "grad_norm": 0.7306382656097412,
      "learning_rate": 0.00043505574849303654,
      "loss": 0.1543,
      "step": 3780
    },
    {
      "epoch": 0.7207678364484593,
      "grad_norm": 0.3662378489971161,
      "learning_rate": 0.00043471846782213184,
      "loss": 0.0682,
      "step": 3790
    },
    {
      "epoch": 0.7226695985499064,
      "grad_norm": 0.4383717477321625,
      "learning_rate": 0.0004343804450720717,
      "loss": 0.4356,
      "step": 3800
    },
    {
      "epoch": 0.7245713606513535,
      "grad_norm": 0.24195566773414612,
      "learning_rate": 0.0004340416816008125,
      "loss": 0.058,
      "step": 3810
    },
    {
      "epoch": 0.7264731227528006,
      "grad_norm": 0.746701717376709,
      "learning_rate": 0.00043370217876928643,
      "loss": 0.137,
      "step": 3820
    },
    {
      "epoch": 0.7283748848542477,
      "grad_norm": 0.33316388726234436,
      "learning_rate": 0.000433361937941396,
      "loss": 0.0669,
      "step": 3830
    },
    {
      "epoch": 0.7302766469556949,
      "grad_norm": 0.5440863370895386,
      "learning_rate": 0.00043302096048400846,
      "loss": 0.0573,
      "step": 3840
    },
    {
      "epoch": 0.732178409057142,
      "grad_norm": 0.20268595218658447,
      "learning_rate": 0.00043267924776695034,
      "loss": 0.0647,
      "step": 3850
    },
    {
      "epoch": 0.7340801711585891,
      "grad_norm": 0.12712806463241577,
      "learning_rate": 0.0004323368011630021,
      "loss": 0.0656,
      "step": 3860
    },
    {
      "epoch": 0.7359819332600362,
      "grad_norm": 0.25622451305389404,
      "learning_rate": 0.00043199362204789224,
      "loss": 0.0369,
      "step": 3870
    },
    {
      "epoch": 0.7378836953614833,
      "grad_norm": 0.48822298645973206,
      "learning_rate": 0.0004316497118002922,
      "loss": 0.0612,
      "step": 3880
    },
    {
      "epoch": 0.7397854574629305,
      "grad_norm": 0.6518691778182983,
      "learning_rate": 0.0004313050718018107,
      "loss": 0.0536,
      "step": 3890
    },
    {
      "epoch": 0.7416872195643777,
      "grad_norm": 0.4019169509410858,
      "learning_rate": 0.0004309597034369878,
      "loss": 0.0511,
      "step": 3900
    },
    {
      "epoch": 0.7435889816658248,
      "grad_norm": 0.3208947479724884,
      "learning_rate": 0.00043061360809329007,
      "loss": 0.0612,
      "step": 3910
    },
    {
      "epoch": 0.7454907437672719,
      "grad_norm": 0.6742276549339294,
      "learning_rate": 0.0004302667871611045,
      "loss": 0.05,
      "step": 3920
    },
    {
      "epoch": 0.747392505868719,
      "grad_norm": 0.32614463567733765,
      "learning_rate": 0.0004299192420337326,
      "loss": 0.0508,
      "step": 3930
    },
    {
      "epoch": 0.7492942679701661,
      "grad_norm": 0.407650887966156,
      "learning_rate": 0.0004295709741073859,
      "loss": 0.0459,
      "step": 3940
    },
    {
      "epoch": 0.7511960300716132,
      "grad_norm": 0.11781582981348038,
      "learning_rate": 0.00042922198478117927,
      "loss": 0.0452,
      "step": 3950
    },
    {
      "epoch": 0.7530977921730604,
      "grad_norm": 0.10125182569026947,
      "learning_rate": 0.0004288722754571257,
      "loss": 0.0391,
      "step": 3960
    },
    {
      "epoch": 0.7549995542745075,
      "grad_norm": 0.19000868499279022,
      "learning_rate": 0.00042852184754013075,
      "loss": 0.0345,
      "step": 3970
    },
    {
      "epoch": 0.7569013163759546,
      "grad_norm": 0.00830161478370428,
      "learning_rate": 0.00042817070243798686,
      "loss": 0.0446,
      "step": 3980
    },
    {
      "epoch": 0.7588030784774017,
      "grad_norm": 0.37715670466423035,
      "learning_rate": 0.0004278188415613675,
      "loss": 0.0533,
      "step": 3990
    },
    {
      "epoch": 0.7607048405788488,
      "grad_norm": 0.02961985021829605,
      "learning_rate": 0.000427466266323822,
      "loss": 0.0274,
      "step": 4000
    },
    {
      "epoch": 0.762606602680296,
      "grad_norm": 0.3095683753490448,
      "learning_rate": 0.00042711297814176897,
      "loss": 0.0333,
      "step": 4010
    },
    {
      "epoch": 0.7645083647817431,
      "grad_norm": 0.17599205672740936,
      "learning_rate": 0.0004267589784344917,
      "loss": 0.0354,
      "step": 4020
    },
    {
      "epoch": 0.7664101268831902,
      "grad_norm": 0.29849332571029663,
      "learning_rate": 0.00042640426862413164,
      "loss": 0.0388,
      "step": 4030
    },
    {
      "epoch": 0.7683118889846373,
      "grad_norm": 0.2587386667728424,
      "learning_rate": 0.00042604885013568304,
      "loss": 0.0535,
      "step": 4040
    },
    {
      "epoch": 0.7702136510860844,
      "grad_norm": 0.1985943764448166,
      "learning_rate": 0.00042569272439698725,
      "loss": 0.0357,
      "step": 4050
    },
    {
      "epoch": 0.7721154131875315,
      "grad_norm": 0.28795111179351807,
      "learning_rate": 0.00042533589283872677,
      "loss": 0.0437,
      "step": 4060
    },
    {
      "epoch": 0.7740171752889787,
      "grad_norm": 0.15408733487129211,
      "learning_rate": 0.0004249783568944197,
      "loss": 0.051,
      "step": 4070
    },
    {
      "epoch": 0.7759189373904258,
      "grad_norm": 0.35579803586006165,
      "learning_rate": 0.00042462011800041376,
      "loss": 0.0445,
      "step": 4080
    },
    {
      "epoch": 0.7778206994918729,
      "grad_norm": 0.1227014809846878,
      "learning_rate": 0.0004242611775958809,
      "loss": 0.0333,
      "step": 4090
    },
    {
      "epoch": 0.77972246159332,
      "grad_norm": 0.3193877339363098,
      "learning_rate": 0.0004239015371228111,
      "loss": 0.0273,
      "step": 4100
    },
    {
      "epoch": 0.7816242236947671,
      "grad_norm": 0.09365034103393555,
      "learning_rate": 0.0004235411980260069,
      "loss": 0.0252,
      "step": 4110
    },
    {
      "epoch": 0.7835259857962144,
      "grad_norm": 0.28869110345840454,
      "learning_rate": 0.0004231801617530773,
      "loss": 0.035,
      "step": 4120
    },
    {
      "epoch": 0.7854277478976615,
      "grad_norm": 0.2648963928222656,
      "learning_rate": 0.0004228184297544323,
      "loss": 0.0579,
      "step": 4130
    },
    {
      "epoch": 0.7873295099991086,
      "grad_norm": 0.24735620617866516,
      "learning_rate": 0.00042245600348327664,
      "loss": 0.0321,
      "step": 4140
    },
    {
      "epoch": 0.7892312721005557,
      "grad_norm": 0.14888009428977966,
      "learning_rate": 0.00042209288439560444,
      "loss": 0.0347,
      "step": 4150
    },
    {
      "epoch": 0.7911330342020028,
      "grad_norm": 0.11169561743736267,
      "learning_rate": 0.0004217290739501929,
      "loss": 0.0409,
      "step": 4160
    },
    {
      "epoch": 0.7930347963034499,
      "grad_norm": 0.1095479279756546,
      "learning_rate": 0.0004213645736085968,
      "loss": 0.0323,
      "step": 4170
    },
    {
      "epoch": 0.7949365584048971,
      "grad_norm": 0.21435153484344482,
      "learning_rate": 0.00042099938483514235,
      "loss": 0.0418,
      "step": 4180
    },
    {
      "epoch": 0.7968383205063442,
      "grad_norm": 0.1274380087852478,
      "learning_rate": 0.0004206335090969215,
      "loss": 0.04,
      "step": 4190
    },
    {
      "epoch": 0.7987400826077913,
      "grad_norm": 0.10879164189100266,
      "learning_rate": 0.00042026694786378603,
      "loss": 0.0443,
      "step": 4200
    },
    {
      "epoch": 0.8006418447092384,
      "grad_norm": 0.21737106144428253,
      "learning_rate": 0.00041989970260834157,
      "loss": 0.0333,
      "step": 4210
    },
    {
      "epoch": 0.8025436068106855,
      "grad_norm": 0.3080846071243286,
      "learning_rate": 0.00041953177480594163,
      "loss": 0.0521,
      "step": 4220
    },
    {
      "epoch": 0.8044453689121327,
      "grad_norm": 0.23476336896419525,
      "learning_rate": 0.0004191631659346818,
      "loss": 0.0312,
      "step": 4230
    },
    {
      "epoch": 0.8063471310135798,
      "grad_norm": 0.311959832906723,
      "learning_rate": 0.00041879387747539376,
      "loss": 0.0351,
      "step": 4240
    },
    {
      "epoch": 0.8082488931150269,
      "grad_norm": 0.24322810769081116,
      "learning_rate": 0.0004184239109116393,
      "loss": 0.0426,
      "step": 4250
    },
    {
      "epoch": 0.810150655216474,
      "grad_norm": 0.2684605121612549,
      "learning_rate": 0.00041805326772970455,
      "loss": 0.0313,
      "step": 4260
    },
    {
      "epoch": 0.8120524173179211,
      "grad_norm": 0.2264401763677597,
      "learning_rate": 0.0004176819494185936,
      "loss": 0.0499,
      "step": 4270
    },
    {
      "epoch": 0.8139541794193682,
      "grad_norm": 0.12556996941566467,
      "learning_rate": 0.00041730995747002296,
      "loss": 0.0269,
      "step": 4280
    },
    {
      "epoch": 0.8158559415208154,
      "grad_norm": 0.22863665223121643,
      "learning_rate": 0.00041693729337841546,
      "loss": 0.0404,
      "step": 4290
    },
    {
      "epoch": 0.8177577036222625,
      "grad_norm": 0.007670534774661064,
      "learning_rate": 0.00041656395864089383,
      "loss": 0.0318,
      "step": 4300
    },
    {
      "epoch": 0.8196594657237096,
      "grad_norm": 0.2585577368736267,
      "learning_rate": 0.0004161899547572753,
      "loss": 0.0293,
      "step": 4310
    },
    {
      "epoch": 0.8215612278251567,
      "grad_norm": 0.36932700872421265,
      "learning_rate": 0.00041581528323006526,
      "loss": 0.0379,
      "step": 4320
    },
    {
      "epoch": 0.8234629899266038,
      "grad_norm": 0.2459120750427246,
      "learning_rate": 0.0004154399455644512,
      "loss": 0.0281,
      "step": 4330
    },
    {
      "epoch": 0.825364752028051,
      "grad_norm": 0.03090517409145832,
      "learning_rate": 0.0004150639432682967,
      "loss": 0.0401,
      "step": 4340
    },
    {
      "epoch": 0.8272665141294981,
      "grad_norm": 0.11881538480520248,
      "learning_rate": 0.0004146872778521355,
      "loss": 0.0376,
      "step": 4350
    },
    {
      "epoch": 0.8291682762309452,
      "grad_norm": 0.2136705368757248,
      "learning_rate": 0.0004143099508291652,
      "loss": 0.0247,
      "step": 4360
    },
    {
      "epoch": 0.8310700383323923,
      "grad_norm": 0.22370411455631256,
      "learning_rate": 0.00041393196371524143,
      "loss": 0.0287,
      "step": 4370
    },
    {
      "epoch": 0.8329718004338394,
      "grad_norm": 0.26691102981567383,
      "learning_rate": 0.00041355331802887156,
      "loss": 0.0329,
      "step": 4380
    },
    {
      "epoch": 0.8348735625352866,
      "grad_norm": 0.16438519954681396,
      "learning_rate": 0.00041317401529120866,
      "loss": 0.0281,
      "step": 4390
    },
    {
      "epoch": 0.8367753246367338,
      "grad_norm": 0.1785949021577835,
      "learning_rate": 0.0004127940570260456,
      "loss": 0.0313,
      "step": 4400
    },
    {
      "epoch": 0.8386770867381809,
      "grad_norm": 0.11144755035638809,
      "learning_rate": 0.00041241344475980823,
      "loss": 0.0283,
      "step": 4410
    },
    {
      "epoch": 0.840578848839628,
      "grad_norm": 0.3532226085662842,
      "learning_rate": 0.00041203218002155046,
      "loss": 0.027,
      "step": 4420
    },
    {
      "epoch": 0.8424806109410751,
      "grad_norm": 0.15130820870399475,
      "learning_rate": 0.0004116502643429469,
      "loss": 0.0283,
      "step": 4430
    },
    {
      "epoch": 0.8443823730425222,
      "grad_norm": 0.10683543235063553,
      "learning_rate": 0.00041126769925828733,
      "loss": 0.0256,
      "step": 4440
    },
    {
      "epoch": 0.8462841351439694,
      "grad_norm": 0.12012962251901627,
      "learning_rate": 0.0004108844863044706,
      "loss": 0.0244,
      "step": 4450
    },
    {
      "epoch": 0.8481858972454165,
      "grad_norm": 0.209054633975029,
      "learning_rate": 0.00041050062702099795,
      "loss": 0.0371,
      "step": 4460
    },
    {
      "epoch": 0.8500876593468636,
      "grad_norm": 0.27558189630508423,
      "learning_rate": 0.00041011612294996746,
      "loss": 0.0241,
      "step": 4470
    },
    {
      "epoch": 0.8519894214483107,
      "grad_norm": 0.2076425403356552,
      "learning_rate": 0.0004097309756360674,
      "loss": 0.0376,
      "step": 4480
    },
    {
      "epoch": 0.8538911835497578,
      "grad_norm": 0.1804133802652359,
      "learning_rate": 0.00040934518662657035,
      "loss": 0.0412,
      "step": 4490
    },
    {
      "epoch": 0.8557929456512049,
      "grad_norm": 0.30146369338035583,
      "learning_rate": 0.0004089587574713264,
      "loss": 0.0271,
      "step": 4500
    },
    {
      "epoch": 0.8576947077526521,
      "grad_norm": 0.24618133902549744,
      "learning_rate": 0.00040857168972275785,
      "loss": 0.0348,
      "step": 4510
    },
    {
      "epoch": 0.8595964698540992,
      "grad_norm": 0.1946696937084198,
      "learning_rate": 0.00040818398493585185,
      "loss": 0.0217,
      "step": 4520
    },
    {
      "epoch": 0.8614982319555463,
      "grad_norm": 0.14857546985149384,
      "learning_rate": 0.0004077956446681554,
      "loss": 0.0429,
      "step": 4530
    },
    {
      "epoch": 0.8633999940569934,
      "grad_norm": 0.1880268156528473,
      "learning_rate": 0.0004074066704797682,
      "loss": 0.0476,
      "step": 4540
    },
    {
      "epoch": 0.8653017561584405,
      "grad_norm": 0.13485951721668243,
      "learning_rate": 0.00040701706393333635,
      "loss": 0.0369,
      "step": 4550
    },
    {
      "epoch": 0.8672035182598877,
      "grad_norm": 0.022607989609241486,
      "learning_rate": 0.00040662682659404684,
      "loss": 0.017,
      "step": 4560
    },
    {
      "epoch": 0.8691052803613348,
      "grad_norm": 0.4116274416446686,
      "learning_rate": 0.00040623596002962027,
      "loss": 0.0333,
      "step": 4570
    },
    {
      "epoch": 0.8710070424627819,
      "grad_norm": 0.1159391701221466,
      "learning_rate": 0.0004058444658103055,
      "loss": 0.0206,
      "step": 4580
    },
    {
      "epoch": 0.872908804564229,
      "grad_norm": 0.22013840079307556,
      "learning_rate": 0.00040545234550887264,
      "loss": 0.0394,
      "step": 4590
    },
    {
      "epoch": 0.8748105666656761,
      "grad_norm": 0.1728079915046692,
      "learning_rate": 0.0004050596007006072,
      "loss": 0.0435,
      "step": 4600
    },
    {
      "epoch": 0.8767123287671232,
      "grad_norm": 0.14915741980075836,
      "learning_rate": 0.0004046662329633032,
      "loss": 0.0402,
      "step": 4610
    },
    {
      "epoch": 0.8786140908685705,
      "grad_norm": 0.20652537047863007,
      "learning_rate": 0.0004042722438772576,
      "loss": 0.037,
      "step": 4620
    },
    {
      "epoch": 0.8805158529700176,
      "grad_norm": 0.2016768902540207,
      "learning_rate": 0.00040387763502526325,
      "loss": 0.0247,
      "step": 4630
    },
    {
      "epoch": 0.8824176150714647,
      "grad_norm": 0.07532606273889542,
      "learning_rate": 0.00040348240799260296,
      "loss": 0.0302,
      "step": 4640
    },
    {
      "epoch": 0.8843193771729118,
      "grad_norm": 0.16426606476306915,
      "learning_rate": 0.00040308656436704294,
      "loss": 0.0219,
      "step": 4650
    },
    {
      "epoch": 0.8862211392743589,
      "grad_norm": 0.2340356707572937,
      "learning_rate": 0.0004026901057388265,
      "loss": 0.0347,
      "step": 4660
    },
    {
      "epoch": 0.8881229013758061,
      "grad_norm": 0.019909365102648735,
      "learning_rate": 0.0004022930337006676,
      "loss": 0.0411,
      "step": 4670
    },
    {
      "epoch": 0.8900246634772532,
      "grad_norm": 0.17835919559001923,
      "learning_rate": 0.0004018953498477444,
      "loss": 0.0393,
      "step": 4680
    },
    {
      "epoch": 0.8919264255787003,
      "grad_norm": 0.13876459002494812,
      "learning_rate": 0.00040149705577769313,
      "loss": 0.0401,
      "step": 4690
    },
    {
      "epoch": 0.8938281876801474,
      "grad_norm": 0.16416268050670624,
      "learning_rate": 0.00040109815309060135,
      "loss": 0.0396,
      "step": 4700
    },
    {
      "epoch": 0.8957299497815945,
      "grad_norm": 0.09078884869813919,
      "learning_rate": 0.0004006986433890017,
      "loss": 0.039,
      "step": 4710
    },
    {
      "epoch": 0.8976317118830416,
      "grad_norm": 0.08101935684680939,
      "learning_rate": 0.00040029852827786535,
      "loss": 0.0269,
      "step": 4720
    },
    {
      "epoch": 0.8995334739844888,
      "grad_norm": 0.12669484317302704,
      "learning_rate": 0.00039989780936459566,
      "loss": 0.0392,
      "step": 4730
    },
    {
      "epoch": 0.9014352360859359,
      "grad_norm": 0.11005831509828568,
      "learning_rate": 0.00039949648825902165,
      "loss": 0.0378,
      "step": 4740
    },
    {
      "epoch": 0.903336998187383,
      "grad_norm": 0.21079300343990326,
      "learning_rate": 0.0003990945665733916,
      "loss": 0.0334,
      "step": 4750
    },
    {
      "epoch": 0.9052387602888301,
      "grad_norm": 0.07260707765817642,
      "learning_rate": 0.0003986920459223665,
      "loss": 0.0287,
      "step": 4760
    },
    {
      "epoch": 0.9071405223902772,
      "grad_norm": 0.01923411153256893,
      "learning_rate": 0.0003982889279230135,
      "loss": 0.0454,
      "step": 4770
    },
    {
      "epoch": 0.9090422844917244,
      "grad_norm": 0.13614831864833832,
      "learning_rate": 0.0003978852141947998,
      "loss": 0.0407,
      "step": 4780
    },
    {
      "epoch": 0.9109440465931715,
      "grad_norm": 0.1618584245443344,
      "learning_rate": 0.00039748090635958555,
      "loss": 0.0389,
      "step": 4790
    },
    {
      "epoch": 0.9128458086946186,
      "grad_norm": 0.16158536076545715,
      "learning_rate": 0.00039707600604161773,
      "loss": 0.0343,
      "step": 4800
    },
    {
      "epoch": 0.9147475707960657,
      "grad_norm": 0.1736435741186142,
      "learning_rate": 0.00039667051486752357,
      "loss": 0.0204,
      "step": 4810
    },
    {
      "epoch": 0.9166493328975128,
      "grad_norm": 0.15715183317661285,
      "learning_rate": 0.000396264434466304,
      "loss": 0.0423,
      "step": 4820
    },
    {
      "epoch": 0.9185510949989599,
      "grad_norm": 0.1514957845211029,
      "learning_rate": 0.00039585776646932703,
      "loss": 0.0329,
      "step": 4830
    },
    {
      "epoch": 0.9204528571004071,
      "grad_norm": 0.0464215911924839,
      "learning_rate": 0.0003954505125103212,
      "loss": 0.0281,
      "step": 4840
    },
    {
      "epoch": 0.9223546192018542,
      "grad_norm": 0.24331432580947876,
      "learning_rate": 0.0003950426742253692,
      "loss": 0.0385,
      "step": 4850
    },
    {
      "epoch": 0.9242563813033013,
      "grad_norm": 0.006427076645195484,
      "learning_rate": 0.00039463425325290095,
      "loss": 0.0312,
      "step": 4860
    },
    {
      "epoch": 0.9261581434047484,
      "grad_norm": 0.3959774971008301,
      "learning_rate": 0.00039422525123368755,
      "loss": 0.0168,
      "step": 4870
    },
    {
      "epoch": 0.9280599055061955,
      "grad_norm": 0.14587469398975372,
      "learning_rate": 0.0003938156698108342,
      "loss": 0.0319,
      "step": 4880
    },
    {
      "epoch": 0.9299616676076428,
      "grad_norm": 0.1308862864971161,
      "learning_rate": 0.0003934055106297735,
      "loss": 0.0247,
      "step": 4890
    },
    {
      "epoch": 0.9318634297090899,
      "grad_norm": 0.03495126590132713,
      "learning_rate": 0.0003929947753382596,
      "loss": 0.0199,
      "step": 4900
    },
    {
      "epoch": 0.933765191810537,
      "grad_norm": 0.1539415717124939,
      "learning_rate": 0.0003925834655863608,
      "loss": 0.0375,
      "step": 4910
    },
    {
      "epoch": 0.9356669539119841,
      "grad_norm": 0.1579003781080246,
      "learning_rate": 0.00039217158302645326,
      "loss": 0.0295,
      "step": 4920
    },
    {
      "epoch": 0.9375687160134312,
      "grad_norm": 0.007031635381281376,
      "learning_rate": 0.00039175912931321426,
      "loss": 0.0258,
      "step": 4930
    },
    {
      "epoch": 0.9394704781148783,
      "grad_norm": 0.3409572243690491,
      "learning_rate": 0.00039134610610361574,
      "loss": 0.0351,
      "step": 4940
    },
    {
      "epoch": 0.9413722402163255,
      "grad_norm": 0.27617499232292175,
      "learning_rate": 0.00039093251505691745,
      "loss": 0.0224,
      "step": 4950
    },
    {
      "epoch": 0.9432740023177726,
      "grad_norm": 0.18574129045009613,
      "learning_rate": 0.00039051835783466025,
      "loss": 0.0258,
      "step": 4960
    },
    {
      "epoch": 0.9451757644192197,
      "grad_norm": 0.006668297573924065,
      "learning_rate": 0.0003901036361006596,
      "loss": 0.0241,
      "step": 4970
    },
    {
      "epoch": 0.9470775265206668,
      "grad_norm": 0.14266005158424377,
      "learning_rate": 0.00038968835152099874,
      "loss": 0.0262,
      "step": 4980
    },
    {
      "epoch": 0.9489792886221139,
      "grad_norm": 0.18967001140117645,
      "learning_rate": 0.00038927250576402227,
      "loss": 0.0233,
      "step": 4990
    },
    {
      "epoch": 0.950881050723561,
      "grad_norm": 0.1983349472284317,
      "learning_rate": 0.00038885610050032896,
      "loss": 0.0364,
      "step": 5000
    },
    {
      "epoch": 0.9527828128250082,
      "grad_norm": 0.1807214319705963,
      "learning_rate": 0.00038843913740276546,
      "loss": 0.0345,
      "step": 5010
    },
    {
      "epoch": 0.9546845749264553,
      "grad_norm": 0.336378812789917,
      "learning_rate": 0.0003880216181464195,
      "loss": 0.021,
      "step": 5020
    },
    {
      "epoch": 0.9565863370279024,
      "grad_norm": 0.2094445377588272,
      "learning_rate": 0.0003876035444086129,
      "loss": 0.0453,
      "step": 5030
    },
    {
      "epoch": 0.9584880991293495,
      "grad_norm": 0.1133272647857666,
      "learning_rate": 0.0003871849178688952,
      "loss": 0.0339,
      "step": 5040
    },
    {
      "epoch": 0.9603898612307966,
      "grad_norm": 0.21298140287399292,
      "learning_rate": 0.00038676574020903666,
      "loss": 0.0417,
      "step": 5050
    },
    {
      "epoch": 0.9622916233322438,
      "grad_norm": 0.2184475064277649,
      "learning_rate": 0.00038634601311302166,
      "loss": 0.02,
      "step": 5060
    },
    {
      "epoch": 0.9641933854336909,
      "grad_norm": 0.13946540653705597,
      "learning_rate": 0.0003859257382670417,
      "loss": 0.0258,
      "step": 5070
    },
    {
      "epoch": 0.966095147535138,
      "grad_norm": 0.09975899010896683,
      "learning_rate": 0.000385504917359489,
      "loss": 0.0261,
      "step": 5080
    },
    {
      "epoch": 0.9679969096365851,
      "grad_norm": 0.10357426851987839,
      "learning_rate": 0.00038508355208094925,
      "loss": 0.0367,
      "step": 5090
    },
    {
      "epoch": 0.9698986717380322,
      "grad_norm": 0.2898187041282654,
      "learning_rate": 0.0003846616441241953,
      "loss": 0.0398,
      "step": 5100
    },
    {
      "epoch": 0.9718004338394793,
      "grad_norm": 0.21133796870708466,
      "learning_rate": 0.00038423919518418,
      "loss": 0.0311,
      "step": 5110
    },
    {
      "epoch": 0.9737021959409266,
      "grad_norm": 0.20074529945850372,
      "learning_rate": 0.0003838162069580296,
      "loss": 0.0319,
      "step": 5120
    },
    {
      "epoch": 0.9756039580423737,
      "grad_norm": 0.10800763219594955,
      "learning_rate": 0.0003833926811450368,
      "loss": 0.0331,
      "step": 5130
    },
    {
      "epoch": 0.9775057201438208,
      "grad_norm": 0.17509090900421143,
      "learning_rate": 0.0003829686194466539,
      "loss": 0.0176,
      "step": 5140
    },
    {
      "epoch": 0.9794074822452679,
      "grad_norm": 0.16533738374710083,
      "learning_rate": 0.0003825440235664863,
      "loss": 0.0182,
      "step": 5150
    },
    {
      "epoch": 0.981309244346715,
      "grad_norm": 0.15933182835578918,
      "learning_rate": 0.000382118895210285,
      "loss": 0.0398,
      "step": 5160
    },
    {
      "epoch": 0.9832110064481622,
      "grad_norm": 0.2111063301563263,
      "learning_rate": 0.0003816932360859405,
      "loss": 0.0247,
      "step": 5170
    },
    {
      "epoch": 0.9851127685496093,
      "grad_norm": 0.18295446038246155,
      "learning_rate": 0.0003812670479034754,
      "loss": 0.041,
      "step": 5180
    },
    {
      "epoch": 0.9870145306510564,
      "grad_norm": 0.17066188156604767,
      "learning_rate": 0.0003808403323750379,
      "loss": 0.032,
      "step": 5190
    },
    {
      "epoch": 0.9889162927525035,
      "grad_norm": 0.17090731859207153,
      "learning_rate": 0.00038041309121489443,
      "loss": 0.0313,
      "step": 5200
    }
  ],
  "logging_steps": 10,
  "max_steps": 15774,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.3194238878926234e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}