{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 21573,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0009271062695561478,
      "grad_norm": 4.991297361547488,
      "learning_rate": 2.932098765432099e-06,
      "loss": 2.260660934448242,
      "step": 20
    },
    {
      "epoch": 0.0018542125391122957,
      "grad_norm": 2.352637442527221,
      "learning_rate": 6.0185185185185185e-06,
      "loss": 1.0406965255737304,
      "step": 40
    },
    {
      "epoch": 0.0027813188086684435,
      "grad_norm": 2.8776893794040106,
      "learning_rate": 9.104938271604939e-06,
      "loss": 0.34813120365142824,
      "step": 60
    },
    {
      "epoch": 0.0037084250782245913,
      "grad_norm": 1.1107183306206494,
      "learning_rate": 1.219135802469136e-05,
      "loss": 0.22572717666625977,
      "step": 80
    },
    {
      "epoch": 0.004635531347780739,
      "grad_norm": 0.4633165478210156,
      "learning_rate": 1.527777777777778e-05,
      "loss": 0.17446967363357543,
      "step": 100
    },
    {
      "epoch": 0.005562637617336887,
      "grad_norm": 0.36384984473545345,
      "learning_rate": 1.83641975308642e-05,
      "loss": 0.14924054145812987,
      "step": 120
    },
    {
      "epoch": 0.006489743886893035,
      "grad_norm": 0.3010553165258027,
      "learning_rate": 2.145061728395062e-05,
      "loss": 0.13816986083984376,
      "step": 140
    },
    {
      "epoch": 0.007416850156449183,
      "grad_norm": 0.28050609761922285,
      "learning_rate": 2.4537037037037038e-05,
      "loss": 0.12528257369995116,
      "step": 160
    },
    {
      "epoch": 0.008343956426005331,
      "grad_norm": 0.47515129627015734,
      "learning_rate": 2.762345679012346e-05,
      "loss": 0.12326407432556152,
      "step": 180
    },
    {
      "epoch": 0.009271062695561478,
      "grad_norm": 0.22663463929430802,
      "learning_rate": 3.0709876543209876e-05,
      "loss": 0.12256077527999878,
      "step": 200
    },
    {
      "epoch": 0.010198168965117627,
      "grad_norm": 0.2396858861700947,
      "learning_rate": 3.3796296296296295e-05,
      "loss": 0.11252110004425049,
      "step": 220
    },
    {
      "epoch": 0.011125275234673774,
      "grad_norm": 0.3071785416952425,
      "learning_rate": 3.6882716049382714e-05,
      "loss": 0.1089066743850708,
      "step": 240
    },
    {
      "epoch": 0.012052381504229923,
      "grad_norm": 0.26390914225555007,
      "learning_rate": 3.996913580246914e-05,
      "loss": 0.1067718505859375,
      "step": 260
    },
    {
      "epoch": 0.01297948777378607,
      "grad_norm": 0.2016508195484771,
      "learning_rate": 4.305555555555556e-05,
      "loss": 0.10530509948730468,
      "step": 280
    },
    {
      "epoch": 0.013906594043342218,
      "grad_norm": 0.21057359575368,
      "learning_rate": 4.614197530864198e-05,
      "loss": 0.10589008331298828,
      "step": 300
    },
    {
      "epoch": 0.014833700312898365,
      "grad_norm": 0.18201832216488453,
      "learning_rate": 4.92283950617284e-05,
      "loss": 0.10339449644088745,
      "step": 320
    },
    {
      "epoch": 0.015760806582454514,
      "grad_norm": 0.24714221571312664,
      "learning_rate": 5.231481481481482e-05,
      "loss": 0.09798303246498108,
      "step": 340
    },
    {
      "epoch": 0.016687912852010663,
      "grad_norm": 0.3412129084606895,
      "learning_rate": 5.5401234567901236e-05,
      "loss": 0.10233824253082276,
      "step": 360
    },
    {
      "epoch": 0.01761501912156681,
      "grad_norm": 0.5978287640690901,
      "learning_rate": 5.848765432098766e-05,
      "loss": 0.09728931188583374,
      "step": 380
    },
    {
      "epoch": 0.018542125391122957,
      "grad_norm": 0.21837629455342994,
      "learning_rate": 6.157407407407407e-05,
      "loss": 0.09973480701446533,
      "step": 400
    },
    {
      "epoch": 0.019469231660679105,
      "grad_norm": 0.2525544686566133,
      "learning_rate": 6.46604938271605e-05,
      "loss": 0.10057148933410645,
      "step": 420
    },
    {
      "epoch": 0.020396337930235254,
      "grad_norm": 0.14090852658032088,
      "learning_rate": 6.774691358024692e-05,
      "loss": 0.09503650069236755,
      "step": 440
    },
    {
      "epoch": 0.021323444199791403,
      "grad_norm": 0.15049601224466577,
      "learning_rate": 7.083333333333334e-05,
      "loss": 0.09429141283035278,
      "step": 460
    },
    {
      "epoch": 0.022250550469347548,
      "grad_norm": 0.17388161127312235,
      "learning_rate": 7.391975308641976e-05,
      "loss": 0.09546964168548584,
      "step": 480
    },
    {
      "epoch": 0.023177656738903697,
      "grad_norm": 0.23224028868503246,
      "learning_rate": 7.700617283950618e-05,
      "loss": 0.09512162208557129,
      "step": 500
    },
    {
      "epoch": 0.024104763008459845,
      "grad_norm": 0.1508971705279842,
      "learning_rate": 8.00925925925926e-05,
      "loss": 0.09294140338897705,
      "step": 520
    },
    {
      "epoch": 0.025031869278015994,
      "grad_norm": 0.26918141136332707,
      "learning_rate": 8.317901234567902e-05,
      "loss": 0.0949424147605896,
      "step": 540
    },
    {
      "epoch": 0.02595897554757214,
      "grad_norm": 0.14158375451620514,
      "learning_rate": 8.626543209876543e-05,
      "loss": 0.09304938912391662,
      "step": 560
    },
    {
      "epoch": 0.026886081817128288,
      "grad_norm": 0.13836526913050481,
      "learning_rate": 8.935185185185185e-05,
      "loss": 0.09392238855361938,
      "step": 580
    },
    {
      "epoch": 0.027813188086684437,
      "grad_norm": 0.5031966633620734,
      "learning_rate": 9.243827160493827e-05,
      "loss": 0.10051620006561279,
      "step": 600
    },
    {
      "epoch": 0.028740294356240585,
      "grad_norm": 0.141771048256938,
      "learning_rate": 9.552469135802469e-05,
      "loss": 0.09560077786445617,
      "step": 620
    },
    {
      "epoch": 0.02966740062579673,
      "grad_norm": 0.1316908800173742,
      "learning_rate": 9.861111111111112e-05,
      "loss": 0.09526835680007935,
      "step": 640
    },
    {
      "epoch": 0.03059450689535288,
      "grad_norm": 0.15584758133292023,
      "learning_rate": 9.999993181417613e-05,
      "loss": 0.09366775155067444,
      "step": 660
    },
    {
      "epoch": 0.03152161316490903,
      "grad_norm": 0.13404482456056072,
      "learning_rate": 9.999945845889795e-05,
      "loss": 0.09144880771636962,
      "step": 680
    },
    {
      "epoch": 0.03244871943446517,
      "grad_norm": 0.12139147683678879,
      "learning_rate": 9.999853429337427e-05,
      "loss": 0.09430141448974609,
      "step": 700
    },
    {
      "epoch": 0.033375825704021325,
      "grad_norm": 0.12588274851910022,
      "learning_rate": 9.999715932593767e-05,
      "loss": 0.09396094679832459,
      "step": 720
    },
    {
      "epoch": 0.03430293197357747,
      "grad_norm": 0.1287876580154401,
      "learning_rate": 9.999533356898527e-05,
      "loss": 0.0914988398551941,
      "step": 740
    },
    {
      "epoch": 0.03523003824313362,
      "grad_norm": 0.10958861274350834,
      "learning_rate": 9.999305703897863e-05,
      "loss": 0.09083436131477356,
      "step": 760
    },
    {
      "epoch": 0.03615714451268977,
      "grad_norm": 0.12698035634957164,
      "learning_rate": 9.999032975644365e-05,
      "loss": 0.09051967859268188,
      "step": 780
    },
    {
      "epoch": 0.03708425078224591,
      "grad_norm": 0.11435138036624376,
      "learning_rate": 9.998715174597033e-05,
      "loss": 0.09183573722839355,
      "step": 800
    },
    {
      "epoch": 0.038011357051802065,
      "grad_norm": 0.1551353764518608,
      "learning_rate": 9.998352303621257e-05,
      "loss": 0.09200174808502197,
      "step": 820
    },
    {
      "epoch": 0.03893846332135821,
      "grad_norm": 0.11945682844036895,
      "learning_rate": 9.997944365988793e-05,
      "loss": 0.09626489281654357,
      "step": 840
    },
    {
      "epoch": 0.039865569590914356,
      "grad_norm": 0.12606763385582406,
      "learning_rate": 9.99749136537773e-05,
      "loss": 0.0895212709903717,
      "step": 860
    },
    {
      "epoch": 0.04079267586047051,
      "grad_norm": 0.12004916426853117,
      "learning_rate": 9.996993305872455e-05,
      "loss": 0.08929449319839478,
      "step": 880
    },
    {
      "epoch": 0.04171978213002665,
      "grad_norm": 0.10046057364494865,
      "learning_rate": 9.996450191963627e-05,
      "loss": 0.09043698310852051,
      "step": 900
    },
    {
      "epoch": 0.042646888399582805,
      "grad_norm": 0.13298671791306557,
      "learning_rate": 9.995862028548124e-05,
      "loss": 0.08920669555664062,
      "step": 920
    },
    {
      "epoch": 0.04357399466913895,
      "grad_norm": 0.10795187784102302,
      "learning_rate": 9.995228820929006e-05,
      "loss": 0.08987456560134888,
      "step": 940
    },
    {
      "epoch": 0.044501100938695096,
      "grad_norm": 0.12615058550959093,
      "learning_rate": 9.994550574815462e-05,
      "loss": 0.086093270778656,
      "step": 960
    },
    {
      "epoch": 0.04542820720825125,
      "grad_norm": 0.10752289689788681,
      "learning_rate": 9.993827296322768e-05,
      "loss": 0.09071444272994995,
      "step": 980
    },
    {
      "epoch": 0.04635531347780739,
      "grad_norm": 0.12798598817723866,
      "learning_rate": 9.993058991972219e-05,
      "loss": 0.08705731630325317,
      "step": 1000
    },
    {
      "epoch": 0.04728241974736354,
      "grad_norm": 0.1015276405122553,
      "learning_rate": 9.99224566869108e-05,
      "loss": 0.08551009893417358,
      "step": 1020
    },
    {
      "epoch": 0.04820952601691969,
      "grad_norm": 0.11133475229328385,
      "learning_rate": 9.991387333812521e-05,
      "loss": 0.08708484172821045,
      "step": 1040
    },
    {
      "epoch": 0.049136632286475836,
      "grad_norm": 0.0998945635242702,
      "learning_rate": 9.990483995075549e-05,
      "loss": 0.09118589758872986,
      "step": 1060
    },
    {
      "epoch": 0.05006373855603199,
      "grad_norm": 0.09542267254857545,
      "learning_rate": 9.989535660624936e-05,
      "loss": 0.08221766352653503,
      "step": 1080
    },
    {
      "epoch": 0.05099084482558813,
      "grad_norm": 0.12480145608973266,
      "learning_rate": 9.988542339011155e-05,
      "loss": 0.08833218812942505,
      "step": 1100
    },
    {
      "epoch": 0.05191795109514428,
      "grad_norm": 0.0905050742461164,
      "learning_rate": 9.987504039190296e-05,
      "loss": 0.08436912298202515,
      "step": 1120
    },
    {
      "epoch": 0.05284505736470043,
      "grad_norm": 0.11264219744501416,
      "learning_rate": 9.986420770523981e-05,
      "loss": 0.08801413774490356,
      "step": 1140
    },
    {
      "epoch": 0.053772163634256576,
      "grad_norm": 0.10377161153557686,
      "learning_rate": 9.98529254277929e-05,
      "loss": 0.08621397018432617,
      "step": 1160
    },
    {
      "epoch": 0.05469926990381273,
      "grad_norm": 0.08601951847545118,
      "learning_rate": 9.984119366128664e-05,
      "loss": 0.08271821737289428,
      "step": 1180
    },
    {
      "epoch": 0.05562637617336887,
      "grad_norm": 0.1429049655611849,
      "learning_rate": 9.982901251149821e-05,
      "loss": 0.08757910132408142,
      "step": 1200
    },
    {
      "epoch": 0.05655348244292502,
      "grad_norm": 0.3349772615388689,
      "learning_rate": 9.981638208825652e-05,
      "loss": 0.09688454866409302,
      "step": 1220
    },
    {
      "epoch": 0.05748058871248117,
      "grad_norm": 0.1727147616810692,
      "learning_rate": 9.98033025054413e-05,
      "loss": 0.11212072372436524,
      "step": 1240
    },
    {
      "epoch": 0.058407694982037316,
      "grad_norm": 0.12302567746500513,
      "learning_rate": 9.9789773880982e-05,
      "loss": 0.09487801194190978,
      "step": 1260
    },
    {
      "epoch": 0.05933480125159346,
      "grad_norm": 0.1322347317978799,
      "learning_rate": 9.977579633685685e-05,
      "loss": 0.0908069670200348,
      "step": 1280
    },
    {
      "epoch": 0.06026190752114961,
      "grad_norm": 0.18952666229327833,
      "learning_rate": 9.976136999909156e-05,
      "loss": 0.08480611443519592,
      "step": 1300
    },
    {
      "epoch": 0.06118901379070576,
      "grad_norm": 0.11615488275136567,
      "learning_rate": 9.974649499775839e-05,
      "loss": 0.09090152978897095,
      "step": 1320
    },
    {
      "epoch": 0.06211612006026191,
      "grad_norm": 0.1428542767348437,
      "learning_rate": 9.973117146697484e-05,
      "loss": 0.09052110314369202,
      "step": 1340
    },
    {
      "epoch": 0.06304322632981806,
      "grad_norm": 0.11265644710529298,
      "learning_rate": 9.971539954490248e-05,
      "loss": 0.09151997566223144,
      "step": 1360
    },
    {
      "epoch": 0.06397033259937421,
      "grad_norm": 0.09293038700782541,
      "learning_rate": 9.969917937374576e-05,
      "loss": 0.08710743188858032,
      "step": 1380
    },
    {
      "epoch": 0.06489743886893035,
      "grad_norm": 0.09378020871119054,
      "learning_rate": 9.968251109975066e-05,
      "loss": 0.08760959506034852,
      "step": 1400
    },
    {
      "epoch": 0.0658245451384865,
      "grad_norm": 0.10137517599900807,
      "learning_rate": 9.966539487320334e-05,
      "loss": 0.08796795606613159,
      "step": 1420
    },
    {
      "epoch": 0.06675165140804265,
      "grad_norm": 0.09323018549025336,
      "learning_rate": 9.964783084842892e-05,
      "loss": 0.08488919734954833,
      "step": 1440
    },
    {
      "epoch": 0.06767875767759879,
      "grad_norm": 0.0828570858134194,
      "learning_rate": 9.962981918378993e-05,
      "loss": 0.08710559606552123,
      "step": 1460
    },
    {
      "epoch": 0.06860586394715494,
      "grad_norm": 0.10274064360001671,
      "learning_rate": 9.9611360041685e-05,
      "loss": 0.08583176136016846,
      "step": 1480
    },
    {
      "epoch": 0.06953297021671109,
      "grad_norm": 0.08354109663262504,
      "learning_rate": 9.959245358854734e-05,
      "loss": 0.08574545979499817,
      "step": 1500
    },
    {
      "epoch": 0.07046007648626725,
      "grad_norm": 0.11086009318614183,
      "learning_rate": 9.957309999484326e-05,
      "loss": 0.08827937841415405,
      "step": 1520
    },
    {
      "epoch": 0.07138718275582338,
      "grad_norm": 0.09985542593963441,
      "learning_rate": 9.95532994350706e-05,
      "loss": 0.0827792763710022,
      "step": 1540
    },
    {
      "epoch": 0.07231428902537954,
      "grad_norm": 0.09691323900853271,
      "learning_rate": 9.95330520877572e-05,
      "loss": 0.08816390037536621,
      "step": 1560
    },
    {
      "epoch": 0.07324139529493569,
      "grad_norm": 0.09936151167350331,
      "learning_rate": 9.951235813545931e-05,
      "loss": 0.08926963806152344,
      "step": 1580
    },
    {
      "epoch": 0.07416850156449183,
      "grad_norm": 0.0890093713054036,
      "learning_rate": 9.94912177647598e-05,
      "loss": 0.08309009075164794,
      "step": 1600
    },
    {
      "epoch": 0.07509560783404798,
      "grad_norm": 0.10763302566847548,
      "learning_rate": 9.946963116626669e-05,
      "loss": 0.0866729736328125,
      "step": 1620
    },
    {
      "epoch": 0.07602271410360413,
      "grad_norm": 0.09922685977197299,
      "learning_rate": 9.944759853461128e-05,
      "loss": 0.0814972162246704,
      "step": 1640
    },
    {
      "epoch": 0.07694982037316027,
      "grad_norm": 0.09298329422756285,
      "learning_rate": 9.942512006844643e-05,
      "loss": 0.08366335034370423,
      "step": 1660
    },
    {
      "epoch": 0.07787692664271642,
      "grad_norm": 0.07731686501039184,
      "learning_rate": 9.94021959704448e-05,
      "loss": 0.08161032795906067,
      "step": 1680
    },
    {
      "epoch": 0.07880403291227257,
      "grad_norm": 0.09600820926826618,
      "learning_rate": 9.937882644729699e-05,
      "loss": 0.08618324398994445,
      "step": 1700
    },
    {
      "epoch": 0.07973113918182871,
      "grad_norm": 0.1115867721615848,
      "learning_rate": 9.935501170970969e-05,
      "loss": 0.08547018766403199,
      "step": 1720
    },
    {
      "epoch": 0.08065824545138486,
      "grad_norm": 0.09048716745753649,
      "learning_rate": 9.93307519724038e-05,
      "loss": 0.08442913293838501,
      "step": 1740
    },
    {
      "epoch": 0.08158535172094102,
      "grad_norm": 0.10993168109341972,
      "learning_rate": 9.930604745411243e-05,
      "loss": 0.08224459886550903,
      "step": 1760
    },
    {
      "epoch": 0.08251245799049715,
      "grad_norm": 0.08799649567562472,
      "learning_rate": 9.928089837757898e-05,
      "loss": 0.08340892791748047,
      "step": 1780
    },
    {
      "epoch": 0.0834395642600533,
      "grad_norm": 0.08539257358636532,
      "learning_rate": 9.925530496955517e-05,
      "loss": 0.08413429856300354,
      "step": 1800
    },
    {
      "epoch": 0.08436667052960946,
      "grad_norm": 0.12200912016016094,
      "learning_rate": 9.922926746079888e-05,
      "loss": 0.08397374153137208,
      "step": 1820
    },
    {
      "epoch": 0.08529377679916561,
      "grad_norm": 0.0820332863468393,
      "learning_rate": 9.920278608607217e-05,
      "loss": 0.08265299201011658,
      "step": 1840
    },
    {
      "epoch": 0.08622088306872175,
      "grad_norm": 0.08490660097222258,
      "learning_rate": 9.917586108413914e-05,
      "loss": 0.08029210567474365,
      "step": 1860
    },
    {
      "epoch": 0.0871479893382779,
      "grad_norm": 0.1135245217185241,
      "learning_rate": 9.914849269776377e-05,
      "loss": 0.08532413840293884,
      "step": 1880
    },
    {
      "epoch": 0.08807509560783405,
      "grad_norm": 0.09270304070733293,
      "learning_rate": 9.912068117370765e-05,
      "loss": 0.08346261978149414,
      "step": 1900
    },
    {
      "epoch": 0.08900220187739019,
      "grad_norm": 0.08573189255502627,
      "learning_rate": 9.909242676272796e-05,
      "loss": 0.08159434795379639,
      "step": 1920
    },
    {
      "epoch": 0.08992930814694634,
      "grad_norm": 0.08138259592111408,
      "learning_rate": 9.9063729719575e-05,
      "loss": 0.08114290237426758,
      "step": 1940
    },
    {
      "epoch": 0.0908564144165025,
      "grad_norm": 0.09061365730503652,
      "learning_rate": 9.903459030298999e-05,
      "loss": 0.08165414929389954,
      "step": 1960
    },
    {
      "epoch": 0.09178352068605863,
      "grad_norm": 0.11575481037257399,
      "learning_rate": 9.90050087757027e-05,
      "loss": 0.08309119939804077,
      "step": 1980
    },
    {
      "epoch": 0.09271062695561479,
      "grad_norm": 0.08559139467453314,
      "learning_rate": 9.897498540442917e-05,
      "loss": 0.0838346004486084,
      "step": 2000
    },
    {
      "epoch": 0.09363773322517094,
      "grad_norm": 0.09586034507382546,
      "learning_rate": 9.894452045986918e-05,
      "loss": 0.0836217761039734,
      "step": 2020
    },
    {
      "epoch": 0.09456483949472708,
      "grad_norm": 0.14596289964106565,
      "learning_rate": 9.89136142167039e-05,
      "loss": 0.08335698843002319,
      "step": 2040
    },
    {
      "epoch": 0.09549194576428323,
      "grad_norm": 0.1123338873616271,
      "learning_rate": 9.888226695359335e-05,
      "loss": 0.09116023182868957,
      "step": 2060
    },
    {
      "epoch": 0.09641905203383938,
      "grad_norm": 0.0918577941948434,
      "learning_rate": 9.885047895317396e-05,
      "loss": 0.08133417963981629,
      "step": 2080
    },
    {
      "epoch": 0.09734615830339553,
      "grad_norm": 0.0965760556962887,
      "learning_rate": 9.881825050205595e-05,
      "loss": 0.08540960550308227,
      "step": 2100
    },
    {
      "epoch": 0.09827326457295167,
      "grad_norm": 0.09069287000367536,
      "learning_rate": 9.878558189082076e-05,
      "loss": 0.07978609800338746,
      "step": 2120
    },
    {
      "epoch": 0.09920037084250782,
      "grad_norm": 0.0873476734367481,
      "learning_rate": 9.875247341401854e-05,
      "loss": 0.08357367515563965,
      "step": 2140
    },
    {
      "epoch": 0.10012747711206398,
      "grad_norm": 0.08324928646092765,
      "learning_rate": 9.871892537016527e-05,
      "loss": 0.07909860014915467,
      "step": 2160
    },
    {
      "epoch": 0.10105458338162011,
      "grad_norm": 0.08821572545479298,
      "learning_rate": 9.868493806174028e-05,
      "loss": 0.08077600002288818,
      "step": 2180
    },
    {
      "epoch": 0.10198168965117627,
      "grad_norm": 0.07774542589427139,
      "learning_rate": 9.865051179518345e-05,
      "loss": 0.08276307582855225,
      "step": 2200
    },
    {
      "epoch": 0.10290879592073242,
      "grad_norm": 0.19706749161556286,
      "learning_rate": 9.86156468808924e-05,
      "loss": 0.08591691255569459,
      "step": 2220
    },
    {
      "epoch": 0.10383590219028856,
      "grad_norm": 0.3996229964577432,
      "learning_rate": 9.858034363321975e-05,
      "loss": 0.08299864530563354,
      "step": 2240
    },
    {
      "epoch": 0.10476300845984471,
      "grad_norm": 0.15984540846768522,
      "learning_rate": 9.854460237047023e-05,
      "loss": 0.09130228757858276,
      "step": 2260
    },
    {
      "epoch": 0.10569011472940086,
      "grad_norm": 0.0872471133772497,
      "learning_rate": 9.850842341489792e-05,
      "loss": 0.08695286512374878,
      "step": 2280
    },
    {
      "epoch": 0.106617220998957,
      "grad_norm": 0.08888920296251855,
      "learning_rate": 9.847180709270323e-05,
      "loss": 0.08585473299026489,
      "step": 2300
    },
    {
      "epoch": 0.10754432726851315,
      "grad_norm": 0.08911168098207818,
      "learning_rate": 9.843475373402997e-05,
      "loss": 0.0813271462917328,
      "step": 2320
    },
    {
      "epoch": 0.1084714335380693,
      "grad_norm": 0.0832446710983605,
      "learning_rate": 9.839726367296247e-05,
      "loss": 0.07933728694915772,
      "step": 2340
    },
    {
      "epoch": 0.10939853980762546,
      "grad_norm": 0.131908221615062,
      "learning_rate": 9.835933724752241e-05,
      "loss": 0.09006819725036622,
      "step": 2360
    },
    {
      "epoch": 0.1103256460771816,
      "grad_norm": 0.12949973739641,
      "learning_rate": 9.832097479966598e-05,
      "loss": 0.08978718519210815,
      "step": 2380
    },
    {
      "epoch": 0.11125275234673775,
      "grad_norm": 0.09038571526515603,
      "learning_rate": 9.828217667528057e-05,
      "loss": 0.08505611419677735,
      "step": 2400
    },
    {
      "epoch": 0.1121798586162939,
      "grad_norm": 0.07947632297304015,
      "learning_rate": 9.824294322418181e-05,
      "loss": 0.08193202018737793,
      "step": 2420
    },
    {
      "epoch": 0.11310696488585004,
      "grad_norm": 0.10883368117527431,
      "learning_rate": 9.82032748001104e-05,
      "loss": 0.08009602427482605,
      "step": 2440
    },
    {
      "epoch": 0.11403407115540619,
      "grad_norm": 0.09463191476521107,
      "learning_rate": 9.816317176072885e-05,
      "loss": 0.0807394802570343,
      "step": 2460
    },
    {
      "epoch": 0.11496117742496234,
      "grad_norm": 0.08739867136372058,
      "learning_rate": 9.812263446761825e-05,
      "loss": 0.08212369680404663,
      "step": 2480
    },
    {
      "epoch": 0.11588828369451848,
      "grad_norm": 0.09852898671246539,
      "learning_rate": 9.808166328627515e-05,
      "loss": 0.08372091054916382,
      "step": 2500
    },
    {
      "epoch": 0.11681538996407463,
      "grad_norm": 0.10135558893016475,
      "learning_rate": 9.804025858610808e-05,
      "loss": 0.08037556409835815,
      "step": 2520
    },
    {
      "epoch": 0.11774249623363078,
      "grad_norm": 0.09879613176114369,
      "learning_rate": 9.799842074043439e-05,
      "loss": 0.08481465578079224,
      "step": 2540
    },
    {
      "epoch": 0.11866960250318692,
      "grad_norm": 0.15343164644336998,
      "learning_rate": 9.79561501264767e-05,
      "loss": 0.08005721569061279,
      "step": 2560
    },
    {
      "epoch": 0.11959670877274307,
      "grad_norm": 0.09219291403994632,
      "learning_rate": 9.791344712535971e-05,
      "loss": 0.08477810621261597,
      "step": 2580
    },
    {
      "epoch": 0.12052381504229923,
      "grad_norm": 0.1034136558182803,
      "learning_rate": 9.787031212210653e-05,
      "loss": 0.08597326278686523,
      "step": 2600
    },
    {
      "epoch": 0.12145092131185538,
      "grad_norm": 0.0996013154868888,
      "learning_rate": 9.782674550563548e-05,
      "loss": 0.07919917106628419,
      "step": 2620
    },
    {
      "epoch": 0.12237802758141152,
      "grad_norm": 0.11681772406397624,
      "learning_rate": 9.77827476687563e-05,
      "loss": 0.08028674721717835,
      "step": 2640
    },
    {
      "epoch": 0.12330513385096767,
      "grad_norm": 0.07627237279663689,
      "learning_rate": 9.773831900816683e-05,
      "loss": 0.07904940247535705,
      "step": 2660
    },
    {
      "epoch": 0.12423224012052382,
      "grad_norm": 0.08471655431457124,
      "learning_rate": 9.769345992444929e-05,
      "loss": 0.07652558088302612,
      "step": 2680
    },
    {
      "epoch": 0.12515934639007997,
      "grad_norm": 0.10052349905212829,
      "learning_rate": 9.764817082206676e-05,
      "loss": 0.08205742835998535,
      "step": 2700
    },
    {
      "epoch": 0.1260864526596361,
      "grad_norm": 0.09581354866514151,
      "learning_rate": 9.76024521093595e-05,
      "loss": 0.07989401817321777,
      "step": 2720
    },
    {
      "epoch": 0.12701355892919225,
      "grad_norm": 0.09058365680742418,
      "learning_rate": 9.755630419854126e-05,
      "loss": 0.08182575702667236,
      "step": 2740
    },
    {
      "epoch": 0.12794066519874842,
      "grad_norm": 0.19478889052852388,
      "learning_rate": 9.750972750569556e-05,
      "loss": 0.07974591255187988,
      "step": 2760
    },
    {
      "epoch": 0.12886777146830455,
      "grad_norm": 0.08021590272131429,
      "learning_rate": 9.746272245077193e-05,
      "loss": 0.07683500051498413,
      "step": 2780
    },
    {
      "epoch": 0.1297948777378607,
      "grad_norm": 0.07605792098586034,
      "learning_rate": 9.74152894575822e-05,
      "loss": 0.0776015043258667,
      "step": 2800
    },
    {
      "epoch": 0.13072198400741686,
      "grad_norm": 0.07767043690580354,
      "learning_rate": 9.736742895379659e-05,
      "loss": 0.08104702234268188,
      "step": 2820
    },
    {
      "epoch": 0.131649090276973,
      "grad_norm": 0.06903523422549838,
      "learning_rate": 9.731914137093986e-05,
      "loss": 0.08125472068786621,
      "step": 2840
    },
    {
      "epoch": 0.13257619654652913,
      "grad_norm": 0.08604031348409921,
      "learning_rate": 9.727042714438748e-05,
      "loss": 0.0798558235168457,
      "step": 2860
    },
    {
      "epoch": 0.1335033028160853,
      "grad_norm": 0.07161322282068992,
      "learning_rate": 9.722128671336165e-05,
      "loss": 0.07996355295181275,
      "step": 2880
    },
    {
      "epoch": 0.13443040908564144,
      "grad_norm": 0.08061278541010523,
      "learning_rate": 9.717172052092738e-05,
      "loss": 0.07677702903747559,
      "step": 2900
    },
    {
      "epoch": 0.13535751535519758,
      "grad_norm": 0.08713179902508664,
      "learning_rate": 9.712172901398844e-05,
      "loss": 0.07869374752044678,
      "step": 2920
    },
    {
      "epoch": 0.13628462162475374,
      "grad_norm": 0.08455366284058966,
      "learning_rate": 9.707131264328339e-05,
      "loss": 0.07974661588668823,
      "step": 2940
    },
    {
      "epoch": 0.13721172789430988,
      "grad_norm": 0.07975300229866615,
      "learning_rate": 9.702047186338147e-05,
      "loss": 0.08084556460380554,
      "step": 2960
    },
    {
      "epoch": 0.13813883416386602,
      "grad_norm": 0.07561080474694325,
      "learning_rate": 9.696920713267856e-05,
      "loss": 0.07706475257873535,
      "step": 2980
    },
    {
      "epoch": 0.13906594043342219,
      "grad_norm": 0.0807219576298629,
      "learning_rate": 9.691751891339296e-05,
      "loss": 0.07801978588104248,
      "step": 3000
    },
    {
      "epoch": 0.13999304670297832,
      "grad_norm": 0.08095171909797877,
      "learning_rate": 9.68654076715613e-05,
      "loss": 0.08146966099739075,
      "step": 3020
    },
    {
      "epoch": 0.1409201529725345,
      "grad_norm": 0.0909704944794839,
      "learning_rate": 9.68128738770343e-05,
      "loss": 0.0785717248916626,
      "step": 3040
    },
    {
      "epoch": 0.14184725924209063,
      "grad_norm": 0.07347086062368631,
      "learning_rate": 9.675991800347257e-05,
      "loss": 0.08078370690345764,
      "step": 3060
    },
    {
      "epoch": 0.14277436551164677,
      "grad_norm": 0.08383088577329738,
      "learning_rate": 9.670654052834227e-05,
      "loss": 0.07747275829315185,
      "step": 3080
    },
    {
      "epoch": 0.14370147178120293,
      "grad_norm": 0.06347603238247887,
      "learning_rate": 9.665274193291086e-05,
      "loss": 0.07575545310974122,
      "step": 3100
    },
    {
      "epoch": 0.14462857805075907,
      "grad_norm": 0.11754747897141529,
      "learning_rate": 9.659852270224275e-05,
      "loss": 0.07863978147506714,
      "step": 3120
    },
    {
      "epoch": 0.1455556843203152,
      "grad_norm": 0.09176540874278705,
      "learning_rate": 9.654388332519493e-05,
      "loss": 0.08438264131546021,
      "step": 3140
    },
    {
      "epoch": 0.14648279058987138,
      "grad_norm": 0.08807828504089057,
      "learning_rate": 9.648882429441257e-05,
      "loss": 0.08201777338981628,
      "step": 3160
    },
    {
      "epoch": 0.1474098968594275,
      "grad_norm": 0.08237053592732457,
      "learning_rate": 9.643334610632455e-05,
      "loss": 0.08363336324691772,
      "step": 3180
    },
    {
      "epoch": 0.14833700312898365,
      "grad_norm": 0.11331259935648921,
      "learning_rate": 9.6377449261139e-05,
      "loss": 0.08310807943344116,
      "step": 3200
    },
    {
      "epoch": 0.14926410939853982,
      "grad_norm": 0.08562841670813581,
      "learning_rate": 9.632113426283878e-05,
      "loss": 0.07947392463684082,
      "step": 3220
    },
    {
      "epoch": 0.15019121566809596,
      "grad_norm": 0.07553618181201636,
      "learning_rate": 9.626440161917693e-05,
      "loss": 0.08322449326515198,
      "step": 3240
    },
    {
      "epoch": 0.1511183219376521,
      "grad_norm": 0.08684370774728792,
      "learning_rate": 9.620725184167219e-05,
      "loss": 0.07906761765480042,
      "step": 3260
    },
    {
      "epoch": 0.15204542820720826,
      "grad_norm": 0.08121449707979174,
      "learning_rate": 9.61496854456042e-05,
      "loss": 0.07632812857627869,
      "step": 3280
    },
    {
      "epoch": 0.1529725344767644,
      "grad_norm": 0.07506817465235674,
      "learning_rate": 9.6091702950009e-05,
      "loss": 0.07769007682800293,
      "step": 3300
    },
    {
      "epoch": 0.15389964074632054,
      "grad_norm": 0.12682690603611876,
      "learning_rate": 9.603330487767434e-05,
      "loss": 0.07713000774383545,
      "step": 3320
    },
    {
      "epoch": 0.1548267470158767,
      "grad_norm": 0.09629402268257964,
      "learning_rate": 9.59744917551349e-05,
      "loss": 0.08252835273742676,
      "step": 3340
    },
    {
      "epoch": 0.15575385328543284,
      "grad_norm": 0.08606367570400894,
      "learning_rate": 9.59152641126676e-05,
      "loss": 0.08223342895507812,
      "step": 3360
    },
    {
      "epoch": 0.15668095955498898,
      "grad_norm": 0.08520820732088029,
      "learning_rate": 9.585562248428677e-05,
      "loss": 0.08223506808280945,
      "step": 3380
    },
    {
      "epoch": 0.15760806582454515,
      "grad_norm": 0.07872066608735398,
      "learning_rate": 9.579556740773936e-05,
      "loss": 0.0811029314994812,
      "step": 3400
    },
    {
      "epoch": 0.15853517209410128,
      "grad_norm": 0.07538783106917514,
      "learning_rate": 9.573509942450014e-05,
      "loss": 0.08072139024734497,
      "step": 3420
    },
    {
      "epoch": 0.15946227836365742,
      "grad_norm": 0.07804933490140105,
      "learning_rate": 9.567421907976673e-05,
      "loss": 0.07947325706481934,
      "step": 3440
    },
    {
      "epoch": 0.1603893846332136,
      "grad_norm": 0.06882246723805557,
      "learning_rate": 9.561292692245473e-05,
      "loss": 0.07520513534545899,
      "step": 3460
    },
    {
      "epoch": 0.16131649090276973,
      "grad_norm": 0.10034022208141886,
      "learning_rate": 9.55512235051928e-05,
      "loss": 0.07770430445671081,
      "step": 3480
    },
    {
      "epoch": 0.16224359717232587,
      "grad_norm": 0.0705202675314821,
      "learning_rate": 9.548910938431758e-05,
      "loss": 0.08016504645347595,
      "step": 3500
    },
    {
      "epoch": 0.16317070344188203,
      "grad_norm": 0.0836438111690838,
      "learning_rate": 9.542658511986882e-05,
      "loss": 0.08068887591361999,
      "step": 3520
    },
    {
      "epoch": 0.16409780971143817,
      "grad_norm": 0.06482868647682753,
      "learning_rate": 9.536365127558418e-05,
      "loss": 0.07675939798355103,
      "step": 3540
    },
    {
      "epoch": 0.1650249159809943,
      "grad_norm": 0.06940944794965902,
      "learning_rate": 9.530030841889424e-05,
      "loss": 0.07810875177383422,
      "step": 3560
    },
    {
      "epoch": 0.16595202225055047,
      "grad_norm": 0.06517067430027278,
      "learning_rate": 9.523655712091739e-05,
      "loss": 0.0769991397857666,
      "step": 3580
    },
    {
      "epoch": 0.1668791285201066,
      "grad_norm": 0.0753625049223485,
      "learning_rate": 9.517239795645459e-05,
      "loss": 0.07939453721046448,
      "step": 3600
    },
    {
      "epoch": 0.16780623478966278,
      "grad_norm": 0.0693605109640091,
      "learning_rate": 9.510783150398427e-05,
      "loss": 0.07608392834663391,
      "step": 3620
    },
    {
      "epoch": 0.16873334105921892,
      "grad_norm": 0.07868514095525836,
      "learning_rate": 9.504285834565714e-05,
      "loss": 0.07615221738815307,
      "step": 3640
    },
    {
      "epoch": 0.16966044732877505,
      "grad_norm": 0.0843564098445614,
      "learning_rate": 9.497747906729084e-05,
      "loss": 0.07579313516616822,
      "step": 3660
    },
    {
      "epoch": 0.17058755359833122,
      "grad_norm": 0.07341552879511899,
      "learning_rate": 9.49116942583647e-05,
      "loss": 0.08014138340950012,
      "step": 3680
    },
    {
      "epoch": 0.17151465986788736,
      "grad_norm": 0.06932061107577686,
      "learning_rate": 9.48455045120145e-05,
      "loss": 0.07681684494018555,
      "step": 3700
    },
    {
      "epoch": 0.1724417661374435,
      "grad_norm": 0.0674592126306665,
      "learning_rate": 9.4778910425027e-05,
      "loss": 0.07509281635284423,
      "step": 3720
    },
    {
      "epoch": 0.17336887240699966,
      "grad_norm": 0.07334045985431246,
      "learning_rate": 9.471191259783463e-05,
      "loss": 0.0804215431213379,
      "step": 3740
    },
    {
      "epoch": 0.1742959786765558,
      "grad_norm": 0.07455241578043256,
      "learning_rate": 9.464451163451008e-05,
      "loss": 0.08022613525390625,
      "step": 3760
    },
    {
      "epoch": 0.17522308494611194,
      "grad_norm": 0.07211273430853943,
      "learning_rate": 9.457670814276083e-05,
      "loss": 0.07562567591667176,
      "step": 3780
    },
    {
      "epoch": 0.1761501912156681,
      "grad_norm": 0.07846667099291334,
      "learning_rate": 9.450850273392366e-05,
      "loss": 0.07435235977172852,
      "step": 3800
    },
    {
      "epoch": 0.17707729748522424,
      "grad_norm": 0.8382026375913739,
      "learning_rate": 9.443989602295922e-05,
      "loss": 0.0773086428642273,
      "step": 3820
    },
    {
      "epoch": 0.17800440375478038,
      "grad_norm": 0.06948867021170023,
      "learning_rate": 9.437088862844632e-05,
      "loss": 0.08087101578712463,
      "step": 3840
    },
    {
      "epoch": 0.17893151002433655,
      "grad_norm": 0.0741776472023575,
      "learning_rate": 9.430148117257652e-05,
      "loss": 0.07564412355422974,
      "step": 3860
    },
    {
      "epoch": 0.1798586162938927,
      "grad_norm": 0.06690393440270216,
      "learning_rate": 9.423167428114845e-05,
      "loss": 0.07901864647865295,
      "step": 3880
    },
    {
      "epoch": 0.18078572256344883,
      "grad_norm": 0.07175686803549533,
      "learning_rate": 9.416146858356218e-05,
      "loss": 0.07960241436958312,
      "step": 3900
    },
    {
      "epoch": 0.181712828833005,
      "grad_norm": 0.060590512602924494,
      "learning_rate": 9.40908647128135e-05,
      "loss": 0.07939324378967286,
      "step": 3920
    },
    {
      "epoch": 0.18263993510256113,
      "grad_norm": 0.06760839916401633,
      "learning_rate": 9.401986330548829e-05,
      "loss": 0.07885625958442688,
      "step": 3940
    },
    {
      "epoch": 0.18356704137211727,
      "grad_norm": 0.06442658490113141,
      "learning_rate": 9.394846500175669e-05,
      "loss": 0.07809650897979736,
      "step": 3960
    },
    {
      "epoch": 0.18449414764167343,
      "grad_norm": 0.07176785743799488,
      "learning_rate": 9.387667044536743e-05,
      "loss": 0.07659803032875061,
      "step": 3980
    },
    {
      "epoch": 0.18542125391122957,
      "grad_norm": 0.06668326961900273,
      "learning_rate": 9.380448028364194e-05,
      "loss": 0.07449266314506531,
      "step": 4000
    },
    {
      "epoch": 0.1863483601807857,
      "grad_norm": 0.06428609244955287,
      "learning_rate": 9.373189516746858e-05,
      "loss": 0.07572462558746337,
      "step": 4020
    },
    {
      "epoch": 0.18727546645034188,
      "grad_norm": 0.07243980339700581,
      "learning_rate": 9.365891575129672e-05,
      "loss": 0.07724069356918335,
      "step": 4040
    },
    {
      "epoch": 0.18820257271989801,
      "grad_norm": 0.0876605223667969,
      "learning_rate": 9.358554269313085e-05,
      "loss": 0.07875486612319946,
      "step": 4060
    },
    {
      "epoch": 0.18912967898945415,
      "grad_norm": 0.09016586672179783,
      "learning_rate": 9.351177665452463e-05,
      "loss": 0.07332387566566467,
      "step": 4080
    },
    {
      "epoch": 0.19005678525901032,
      "grad_norm": 0.07451303729904303,
      "learning_rate": 9.343761830057503e-05,
      "loss": 0.07739952802658082,
      "step": 4100
    },
    {
      "epoch": 0.19098389152856646,
      "grad_norm": 0.06972059418605771,
      "learning_rate": 9.336306829991619e-05,
      "loss": 0.07446897029876709,
      "step": 4120
    },
    {
      "epoch": 0.19191099779812262,
      "grad_norm": 0.07316104977904392,
      "learning_rate": 9.328812732471345e-05,
      "loss": 0.07650957107543946,
      "step": 4140
    },
    {
      "epoch": 0.19283810406767876,
      "grad_norm": 0.0697248450236681,
      "learning_rate": 9.321279605065736e-05,
      "loss": 0.0750017523765564,
      "step": 4160
    },
    {
      "epoch": 0.1937652103372349,
      "grad_norm": 0.08324276277730187,
      "learning_rate": 9.313707515695744e-05,
      "loss": 0.0768101692199707,
      "step": 4180
    },
    {
      "epoch": 0.19469231660679107,
      "grad_norm": 0.06702789271794142,
      "learning_rate": 9.306096532633621e-05,
      "loss": 0.07705305814743042,
      "step": 4200
    },
    {
      "epoch": 0.1956194228763472,
      "grad_norm": 0.06347835137910703,
      "learning_rate": 9.298446724502292e-05,
      "loss": 0.07845529317855834,
      "step": 4220
    },
    {
      "epoch": 0.19654652914590334,
      "grad_norm": 0.07677916552725536,
      "learning_rate": 9.290758160274739e-05,
      "loss": 0.07541437149047851,
      "step": 4240
    },
    {
      "epoch": 0.1974736354154595,
      "grad_norm": 0.09112473459480099,
      "learning_rate": 9.283030909273387e-05,
      "loss": 0.07931250333786011,
      "step": 4260
    },
    {
      "epoch": 0.19840074168501565,
      "grad_norm": 0.07608949062168033,
      "learning_rate": 9.275265041169467e-05,
      "loss": 0.07697776556015015,
      "step": 4280
    },
    {
      "epoch": 0.19932784795457179,
      "grad_norm": 0.06982238574468308,
      "learning_rate": 9.267460625982397e-05,
      "loss": 0.07524523735046387,
      "step": 4300
    },
    {
      "epoch": 0.20025495422412795,
      "grad_norm": 0.08165905227939095,
      "learning_rate": 9.259617734079144e-05,
      "loss": 0.07687885761260986,
      "step": 4320
    },
    {
      "epoch": 0.2011820604936841,
      "grad_norm": 0.07801536195877896,
      "learning_rate": 9.251736436173594e-05,
      "loss": 0.07631038427352906,
      "step": 4340
    },
    {
      "epoch": 0.20210916676324023,
      "grad_norm": 0.07310313697126014,
      "learning_rate": 9.243816803325918e-05,
      "loss": 0.08152819275856019,
      "step": 4360
    },
    {
      "epoch": 0.2030362730327964,
      "grad_norm": 0.0713833053243852,
      "learning_rate": 9.235858906941915e-05,
      "loss": 0.07985727787017823,
      "step": 4380
    },
    {
      "epoch": 0.20396337930235253,
      "grad_norm": 0.07042655997545465,
      "learning_rate": 9.227862818772393e-05,
      "loss": 0.07732853889465333,
      "step": 4400
    },
    {
      "epoch": 0.20489048557190867,
      "grad_norm": 0.072299015569342,
      "learning_rate": 9.219828610912497e-05,
      "loss": 0.07840472459793091,
      "step": 4420
    },
    {
      "epoch": 0.20581759184146484,
      "grad_norm": 0.06936854656071505,
      "learning_rate": 9.21175635580108e-05,
      "loss": 0.0798102855682373,
      "step": 4440
    },
    {
      "epoch": 0.20674469811102097,
      "grad_norm": 0.05742922092975488,
      "learning_rate": 9.203646126220034e-05,
      "loss": 0.0744663953781128,
      "step": 4460
    },
    {
      "epoch": 0.2076718043805771,
      "grad_norm": 0.07103354735152609,
      "learning_rate": 9.195497995293643e-05,
      "loss": 0.0785027027130127,
      "step": 4480
    },
    {
      "epoch": 0.20859891065013328,
      "grad_norm": 0.06449930132333753,
      "learning_rate": 9.187312036487922e-05,
      "loss": 0.07595589756965637,
      "step": 4500
    },
    {
      "epoch": 0.20952601691968942,
      "grad_norm": 0.0666648958514048,
      "learning_rate": 9.179088323609952e-05,
      "loss": 0.07669690847396851,
      "step": 4520
    },
    {
      "epoch": 0.21045312318924556,
      "grad_norm": 0.06882096559597742,
      "learning_rate": 9.170826930807215e-05,
      "loss": 0.07462993264198303,
      "step": 4540
    },
    {
      "epoch": 0.21138022945880172,
      "grad_norm": 0.0612856131398295,
      "learning_rate": 9.16252793256693e-05,
      "loss": 0.07590610980987549,
      "step": 4560
    },
    {
      "epoch": 0.21230733572835786,
      "grad_norm": 0.07616943974361534,
      "learning_rate": 9.154191403715376e-05,
      "loss": 0.0753056287765503,
      "step": 4580
    },
    {
      "epoch": 0.213234441997914,
      "grad_norm": 0.06889043986893169,
      "learning_rate": 9.145817419417218e-05,
      "loss": 0.0770835280418396,
      "step": 4600
    },
    {
      "epoch": 0.21416154826747016,
      "grad_norm": 0.08049032482504447,
      "learning_rate": 9.137406055174833e-05,
      "loss": 0.07717642188072205,
      "step": 4620
    },
    {
      "epoch": 0.2150886545370263,
      "grad_norm": 0.06674403738189989,
      "learning_rate": 9.128957386827626e-05,
      "loss": 0.07454350590705872,
      "step": 4640
    },
    {
      "epoch": 0.21601576080658244,
      "grad_norm": 0.0728002798220127,
      "learning_rate": 9.120471490551348e-05,
      "loss": 0.0749647855758667,
      "step": 4660
    },
    {
      "epoch": 0.2169428670761386,
      "grad_norm": 0.06998431538406782,
      "learning_rate": 9.111948442857406e-05,
      "loss": 0.07710611820220947,
      "step": 4680
    },
    {
      "epoch": 0.21786997334569475,
      "grad_norm": 0.07478761818840474,
      "learning_rate": 9.103388320592176e-05,
      "loss": 0.07427996397018433,
      "step": 4700
    },
    {
      "epoch": 0.2187970796152509,
      "grad_norm": 0.07445900430244656,
      "learning_rate": 9.094791200936313e-05,
      "loss": 0.07640702724456787,
      "step": 4720
    },
    {
      "epoch": 0.21972418588480705,
      "grad_norm": 0.06434892088216404,
      "learning_rate": 9.086157161404045e-05,
      "loss": 0.07828493118286133,
      "step": 4740
    },
    {
      "epoch": 0.2206512921543632,
      "grad_norm": 0.0723454111171107,
      "learning_rate": 9.077486279842489e-05,
      "loss": 0.07613458037376404,
      "step": 4760
    },
    {
      "epoch": 0.22157839842391935,
      "grad_norm": 0.07123704820867308,
      "learning_rate": 9.068778634430935e-05,
      "loss": 0.0765914797782898,
      "step": 4780
    },
    {
      "epoch": 0.2225055046934755,
      "grad_norm": 0.057283300802926554,
      "learning_rate": 9.060034303680149e-05,
      "loss": 0.07773427963256836,
      "step": 4800
    },
    {
      "epoch": 0.22343261096303163,
      "grad_norm": 0.06354369165051069,
      "learning_rate": 9.05125336643166e-05,
      "loss": 0.07798044681549073,
      "step": 4820
    },
    {
      "epoch": 0.2243597172325878,
      "grad_norm": 0.07413696583836642,
      "learning_rate": 9.04243590185706e-05,
      "loss": 0.07399480938911437,
      "step": 4840
    },
    {
      "epoch": 0.22528682350214393,
      "grad_norm": 0.07031783494850803,
      "learning_rate": 9.033581989457272e-05,
      "loss": 0.07377501130104065,
      "step": 4860
    },
    {
      "epoch": 0.22621392977170007,
      "grad_norm": 0.08130797201409347,
      "learning_rate": 9.024691709061854e-05,
      "loss": 0.0765376329421997,
      "step": 4880
    },
    {
      "epoch": 0.22714103604125624,
      "grad_norm": 0.06340386823570394,
      "learning_rate": 9.01576514082826e-05,
      "loss": 0.07940353751182556,
      "step": 4900
    },
    {
      "epoch": 0.22806814231081238,
      "grad_norm": 0.07717986544316227,
      "learning_rate": 9.006802365241133e-05,
      "loss": 0.07578405737876892,
      "step": 4920
    },
    {
      "epoch": 0.22899524858036852,
      "grad_norm": 0.08079090687234158,
      "learning_rate": 8.99780346311157e-05,
      "loss": 0.07634626626968384,
      "step": 4940
    },
    {
      "epoch": 0.22992235484992468,
      "grad_norm": 0.06829665790519572,
      "learning_rate": 8.988768515576391e-05,
      "loss": 0.07581467032432557,
      "step": 4960
    },
    {
      "epoch": 0.23084946111948082,
      "grad_norm": 0.0803026812002199,
      "learning_rate": 8.97969760409742e-05,
      "loss": 0.07663472294807434,
      "step": 4980
    },
    {
      "epoch": 0.23177656738903696,
      "grad_norm": 0.06478527974313332,
      "learning_rate": 8.970590810460739e-05,
      "loss": 0.07858390808105468,
      "step": 5000
    },
    {
      "epoch": 0.23270367365859312,
      "grad_norm": 0.07585711573409425,
      "learning_rate": 8.961448216775954e-05,
      "loss": 0.07818331122398377,
      "step": 5020
    },
    {
      "epoch": 0.23363077992814926,
      "grad_norm": 0.06470619849490004,
      "learning_rate": 8.952269905475457e-05,
      "loss": 0.07375747561454774,
      "step": 5040
    },
    {
      "epoch": 0.2345578861977054,
      "grad_norm": 0.07023565588467741,
      "learning_rate": 8.943055959313678e-05,
      "loss": 0.07337473630905152,
      "step": 5060
    },
    {
      "epoch": 0.23548499246726157,
      "grad_norm": 0.06340218843036498,
      "learning_rate": 8.933806461366345e-05,
      "loss": 0.07409183979034424,
      "step": 5080
    },
    {
      "epoch": 0.2364120987368177,
      "grad_norm": 0.1064703708076755,
      "learning_rate": 8.92452149502973e-05,
      "loss": 0.0759962558746338,
      "step": 5100
    },
    {
      "epoch": 0.23733920500637384,
      "grad_norm": 0.06607674764813727,
      "learning_rate": 8.915201144019898e-05,
      "loss": 0.07317070960998535,
      "step": 5120
    },
    {
      "epoch": 0.23826631127593,
      "grad_norm": 0.07014566369261731,
      "learning_rate": 8.905845492371953e-05,
      "loss": 0.07603597640991211,
      "step": 5140
    },
    {
      "epoch": 0.23919341754548615,
      "grad_norm": 0.07152552598107388,
      "learning_rate": 8.896454624439283e-05,
      "loss": 0.07629239559173584,
      "step": 5160
    },
    {
      "epoch": 0.24012052381504229,
      "grad_norm": 0.06596241363750086,
      "learning_rate": 8.887028624892794e-05,
      "loss": 0.07434371709823609,
      "step": 5180
    },
    {
      "epoch": 0.24104763008459845,
      "grad_norm": 0.0646394013723455,
      "learning_rate": 8.87756757872015e-05,
      "loss": 0.07509366869926452,
      "step": 5200
    },
    {
      "epoch": 0.2419747363541546,
      "grad_norm": 0.07131047600508243,
      "learning_rate": 8.868071571225004e-05,
      "loss": 0.07466341257095337,
      "step": 5220
    },
    {
      "epoch": 0.24290184262371076,
      "grad_norm": 0.0633401567348906,
      "learning_rate": 8.858540688026234e-05,
      "loss": 0.0750316858291626,
      "step": 5240
    },
    {
      "epoch": 0.2438289488932669,
      "grad_norm": 0.06563412467579705,
      "learning_rate": 8.848975015057168e-05,
      "loss": 0.07504474520683288,
      "step": 5260
    },
    {
      "epoch": 0.24475605516282303,
      "grad_norm": 0.06684301980857908,
      "learning_rate": 8.839374638564806e-05,
      "loss": 0.07970876693725586,
      "step": 5280
    },
    {
      "epoch": 0.2456831614323792,
      "grad_norm": 0.058939526303855384,
      "learning_rate": 8.829739645109046e-05,
      "loss": 0.07780908346176148,
      "step": 5300
    },
    {
      "epoch": 0.24661026770193534,
      "grad_norm": 0.05901091707187526,
      "learning_rate": 8.820070121561906e-05,
      "loss": 0.07149989604949951,
      "step": 5320
    },
    {
      "epoch": 0.24753737397149148,
      "grad_norm": 0.07345185192875342,
      "learning_rate": 8.810366155106738e-05,
      "loss": 0.07438422441482544,
      "step": 5340
    },
    {
      "epoch": 0.24846448024104764,
      "grad_norm": 0.0708440227750237,
      "learning_rate": 8.800627833237434e-05,
      "loss": 0.07411037683486939,
      "step": 5360
    },
    {
      "epoch": 0.24939158651060378,
      "grad_norm": 0.07507459975551163,
      "learning_rate": 8.790855243757656e-05,
      "loss": 0.07510088682174683,
      "step": 5380
    },
    {
      "epoch": 0.25031869278015995,
      "grad_norm": 0.0922192865215547,
      "learning_rate": 8.781048474780025e-05,
      "loss": 0.07834984064102173,
      "step": 5400
    },
    {
      "epoch": 0.25124579904971606,
      "grad_norm": 0.0880141189972007,
      "learning_rate": 8.771207614725337e-05,
      "loss": 0.07598562240600586,
      "step": 5420
    },
    {
      "epoch": 0.2521729053192722,
      "grad_norm": 0.07414647830917258,
      "learning_rate": 8.761332752321764e-05,
      "loss": 0.07403573393821716,
      "step": 5440
    },
    {
      "epoch": 0.2531000115888284,
      "grad_norm": 0.06018487374270652,
      "learning_rate": 8.751423976604056e-05,
      "loss": 0.07665326595306396,
      "step": 5460
    },
    {
      "epoch": 0.2540271178583845,
      "grad_norm": 0.08835966021744765,
      "learning_rate": 8.741481376912728e-05,
      "loss": 0.076629376411438,
      "step": 5480
    },
    {
      "epoch": 0.25495422412794067,
      "grad_norm": 0.06181037104984363,
      "learning_rate": 8.73150504289327e-05,
      "loss": 0.07047075033187866,
      "step": 5500
    },
    {
      "epoch": 0.25588133039749683,
      "grad_norm": 0.0673286263271081,
      "learning_rate": 8.721495064495328e-05,
      "loss": 0.07512806057929992,
      "step": 5520
    },
    {
      "epoch": 0.25680843666705294,
      "grad_norm": 0.07579533136467483,
      "learning_rate": 8.711451531971896e-05,
      "loss": 0.07796473503112793,
      "step": 5540
    },
    {
      "epoch": 0.2577355429366091,
      "grad_norm": 0.08636163897944657,
      "learning_rate": 8.701374535878502e-05,
      "loss": 0.07579964399337769,
      "step": 5560
    },
    {
      "epoch": 0.2586626492061653,
      "grad_norm": 0.13074761375370783,
      "learning_rate": 8.691264167072388e-05,
      "loss": 0.07326544523239135,
      "step": 5580
    },
    {
      "epoch": 0.2595897554757214,
      "grad_norm": 0.07256946508850105,
      "learning_rate": 8.681120516711704e-05,
      "loss": 0.07315768003463745,
      "step": 5600
    },
    {
      "epoch": 0.26051686174527755,
      "grad_norm": 0.062026891325574283,
      "learning_rate": 8.670943676254667e-05,
      "loss": 0.0761407196521759,
      "step": 5620
    },
    {
      "epoch": 0.2614439680148337,
      "grad_norm": 0.0718775571131328,
      "learning_rate": 8.660733737458752e-05,
      "loss": 0.07447476983070374,
      "step": 5640
    },
    {
      "epoch": 0.2623710742843898,
      "grad_norm": 0.06888753398085139,
      "learning_rate": 8.650490792379855e-05,
      "loss": 0.07509008646011353,
      "step": 5660
    },
    {
      "epoch": 0.263298180553946,
      "grad_norm": 0.07844487057737232,
      "learning_rate": 8.640214933371467e-05,
      "loss": 0.07488833665847779,
      "step": 5680
    },
    {
      "epoch": 0.26422528682350216,
      "grad_norm": 0.06363080472572674,
      "learning_rate": 8.629906253083844e-05,
      "loss": 0.07349667549133301,
      "step": 5700
    },
    {
      "epoch": 0.26515239309305827,
      "grad_norm": 0.08131756681839972,
      "learning_rate": 8.619564844463163e-05,
      "loss": 0.07553036212921142,
      "step": 5720
    },
    {
      "epoch": 0.26607949936261444,
      "grad_norm": 0.06022387030716428,
      "learning_rate": 8.609190800750699e-05,
      "loss": 0.07443491220474244,
      "step": 5740
    },
    {
      "epoch": 0.2670066056321706,
      "grad_norm": 0.06388953778061979,
      "learning_rate": 8.598784215481965e-05,
      "loss": 0.07595537900924683,
      "step": 5760
    },
    {
      "epoch": 0.2679337119017267,
      "grad_norm": 0.06439249000510272,
      "learning_rate": 8.588345182485882e-05,
      "loss": 0.07346284985542298,
      "step": 5780
    },
    {
      "epoch": 0.2688608181712829,
      "grad_norm": 0.06856011210638639,
      "learning_rate": 8.57787379588393e-05,
      "loss": 0.07406741380691528,
      "step": 5800
    },
    {
      "epoch": 0.26978792444083904,
      "grad_norm": 0.06467507391760796,
      "learning_rate": 8.567370150089302e-05,
      "loss": 0.07141650915145874,
      "step": 5820
    },
    {
      "epoch": 0.27071503071039515,
      "grad_norm": 0.07256490355879096,
      "learning_rate": 8.556834339806044e-05,
      "loss": 0.07287147045135497,
      "step": 5840
    },
    {
      "epoch": 0.2716421369799513,
      "grad_norm": 0.05949001161043578,
      "learning_rate": 8.546266460028209e-05,
      "loss": 0.07385816574096679,
      "step": 5860
    },
    {
      "epoch": 0.2725692432495075,
      "grad_norm": 0.07535740176772583,
      "learning_rate": 8.535666606038999e-05,
      "loss": 0.07265498638153076,
      "step": 5880
    },
    {
      "epoch": 0.2734963495190636,
      "grad_norm": 0.06849112755471702,
      "learning_rate": 8.525034873409905e-05,
      "loss": 0.07564167976379395,
      "step": 5900
    },
    {
      "epoch": 0.27442345578861976,
      "grad_norm": 0.06761433555082014,
      "learning_rate": 8.514371357999846e-05,
      "loss": 0.08036400079727173,
      "step": 5920
    },
    {
      "epoch": 0.27535056205817593,
      "grad_norm": 0.08906062826741618,
      "learning_rate": 8.5036761559543e-05,
      "loss": 0.0771210789680481,
      "step": 5940
    },
    {
      "epoch": 0.27627766832773204,
      "grad_norm": 0.06761653380905454,
      "learning_rate": 8.492949363704448e-05,
      "loss": 0.07450598478317261,
      "step": 5960
    },
    {
      "epoch": 0.2772047745972882,
      "grad_norm": 0.04941191166028973,
      "learning_rate": 8.482191077966293e-05,
      "loss": 0.07542569637298584,
      "step": 5980
    },
    {
      "epoch": 0.27813188086684437,
      "grad_norm": 0.05834689367885952,
      "learning_rate": 8.471401395739796e-05,
      "loss": 0.07543001770973205,
      "step": 6000
    },
    {
      "epoch": 0.2790589871364005,
      "grad_norm": 0.059889295917629225,
      "learning_rate": 8.460580414307994e-05,
      "loss": 0.07370642423629761,
      "step": 6020
    },
    {
      "epoch": 0.27998609340595665,
      "grad_norm": 0.06313189260470523,
      "learning_rate": 8.449728231236131e-05,
      "loss": 0.0749735414981842,
      "step": 6040
    },
    {
      "epoch": 0.2809131996755128,
      "grad_norm": 0.05856457324704203,
      "learning_rate": 8.438844944370773e-05,
      "loss": 0.07263882160186767,
      "step": 6060
    },
    {
      "epoch": 0.281840305945069,
      "grad_norm": 0.06589910701504983,
      "learning_rate": 8.427930651838927e-05,
      "loss": 0.07773587703704835,
      "step": 6080
    },
    {
      "epoch": 0.2827674122146251,
      "grad_norm": 0.05981821489512958,
      "learning_rate": 8.416985452047157e-05,
      "loss": 0.07497872114181518,
      "step": 6100
    },
    {
      "epoch": 0.28369451848418126,
      "grad_norm": 0.09696785492962093,
      "learning_rate": 8.406009443680696e-05,
      "loss": 0.07509639263153076,
      "step": 6120
    },
    {
      "epoch": 0.2846216247537374,
      "grad_norm": 0.05991409043262136,
      "learning_rate": 8.395002725702557e-05,
      "loss": 0.07105750441551209,
      "step": 6140
    },
    {
      "epoch": 0.28554873102329353,
      "grad_norm": 0.07058247656822032,
      "learning_rate": 8.383965397352637e-05,
      "loss": 0.07501793503761292,
      "step": 6160
    },
    {
      "epoch": 0.2864758372928497,
      "grad_norm": 0.06476787955115126,
      "learning_rate": 8.372897558146831e-05,
      "loss": 0.07241331338882447,
      "step": 6180
    },
    {
      "epoch": 0.28740294356240587,
      "grad_norm": 0.061758733782570976,
      "learning_rate": 8.361799307876124e-05,
      "loss": 0.07378792762756348,
      "step": 6200
    },
    {
      "epoch": 0.288330049831962,
      "grad_norm": 0.12343557280884002,
      "learning_rate": 8.350670746605698e-05,
      "loss": 0.07377973794937134,
      "step": 6220
    },
    {
      "epoch": 0.28925715610151814,
      "grad_norm": 0.061745955915717056,
      "learning_rate": 8.33951197467403e-05,
      "loss": 0.0737960159778595,
      "step": 6240
    },
    {
      "epoch": 0.2901842623710743,
      "grad_norm": 0.06157997227659551,
      "learning_rate": 8.328323092691985e-05,
      "loss": 0.07219173312187195,
      "step": 6260
    },
    {
      "epoch": 0.2911113686406304,
      "grad_norm": 0.07378954323731418,
      "learning_rate": 8.317104201541904e-05,
      "loss": 0.07557147741317749,
      "step": 6280
    },
    {
      "epoch": 0.2920384749101866,
      "grad_norm": 0.05957251191001676,
      "learning_rate": 8.305855402376709e-05,
      "loss": 0.074476557970047,
      "step": 6300
    },
    {
      "epoch": 0.29296558117974275,
      "grad_norm": 0.06723486388580951,
      "learning_rate": 8.294576796618973e-05,
      "loss": 0.07679529786109925,
      "step": 6320
    },
    {
      "epoch": 0.29389268744929886,
      "grad_norm": 0.06393647282434042,
      "learning_rate": 8.28326848596002e-05,
      "loss": 0.07275311350822448,
      "step": 6340
    },
    {
      "epoch": 0.294819793718855,
      "grad_norm": 0.06421683428554714,
      "learning_rate": 8.271930572359002e-05,
      "loss": 0.074833083152771,
      "step": 6360
    },
    {
      "epoch": 0.2957468999884112,
      "grad_norm": 0.05761110451805251,
      "learning_rate": 8.260563158041974e-05,
      "loss": 0.0731091022491455,
      "step": 6380
    },
    {
      "epoch": 0.2966740062579673,
      "grad_norm": 0.06813400616795082,
      "learning_rate": 8.249166345500985e-05,
      "loss": 0.07645623683929444,
      "step": 6400
    },
    {
      "epoch": 0.29760111252752347,
      "grad_norm": 0.08754152582865571,
      "learning_rate": 8.237740237493148e-05,
      "loss": 0.07599068880081176,
      "step": 6420
    },
    {
      "epoch": 0.29852821879707964,
      "grad_norm": 0.08874677998290044,
      "learning_rate": 8.226284937039706e-05,
      "loss": 0.07406764030456543,
      "step": 6440
    },
    {
      "epoch": 0.29945532506663575,
      "grad_norm": 0.07135009328610928,
      "learning_rate": 8.214800547425114e-05,
      "loss": 0.07538463473320008,
      "step": 6460
    },
    {
      "epoch": 0.3003824313361919,
      "grad_norm": 0.07508827324097464,
      "learning_rate": 8.203287172196106e-05,
      "loss": 0.08206729292869568,
      "step": 6480
    },
    {
      "epoch": 0.3013095376057481,
      "grad_norm": 0.06773065275871519,
      "learning_rate": 8.191744915160754e-05,
      "loss": 0.07659539580345154,
      "step": 6500
    },
    {
      "epoch": 0.3022366438753042,
      "grad_norm": 0.08666234771926429,
      "learning_rate": 8.180173880387542e-05,
      "loss": 0.07337627410888672,
      "step": 6520
    },
    {
      "epoch": 0.30316375014486036,
      "grad_norm": 0.05788354926375118,
      "learning_rate": 8.168574172204418e-05,
      "loss": 0.07501276731491088,
      "step": 6540
    },
    {
      "epoch": 0.3040908564144165,
      "grad_norm": 0.06119211233538267,
      "learning_rate": 8.156945895197864e-05,
      "loss": 0.07404306530952454,
      "step": 6560
    },
    {
      "epoch": 0.30501796268397263,
      "grad_norm": 0.07283309684573482,
      "learning_rate": 8.145289154211942e-05,
      "loss": 0.0733154833316803,
      "step": 6580
    },
    {
      "epoch": 0.3059450689535288,
      "grad_norm": 0.07424607988131313,
      "learning_rate": 8.133604054347355e-05,
      "loss": 0.07469432353973389,
      "step": 6600
    },
    {
      "epoch": 0.30687217522308496,
      "grad_norm": 0.05577433265908877,
      "learning_rate": 8.121890700960498e-05,
      "loss": 0.07555993795394897,
      "step": 6620
    },
    {
      "epoch": 0.3077992814926411,
      "grad_norm": 0.059344917590590684,
      "learning_rate": 8.110149199662512e-05,
      "loss": 0.07502079010009766,
      "step": 6640
    },
    {
      "epoch": 0.30872638776219724,
      "grad_norm": 0.06488827664076578,
      "learning_rate": 8.098379656318322e-05,
      "loss": 0.07209377884864807,
      "step": 6660
    },
    {
      "epoch": 0.3096534940317534,
      "grad_norm": 0.05640397875804107,
      "learning_rate": 8.086582177045696e-05,
      "loss": 0.07356261014938355,
      "step": 6680
    },
    {
      "epoch": 0.3105806003013095,
      "grad_norm": 0.05996438788382959,
      "learning_rate": 8.074756868214272e-05,
      "loss": 0.07091883420944214,
      "step": 6700
    },
    {
      "epoch": 0.3115077065708657,
      "grad_norm": 0.06383010495369079,
      "learning_rate": 8.062903836444616e-05,
      "loss": 0.07409636974334717,
      "step": 6720
    },
    {
      "epoch": 0.31243481284042185,
      "grad_norm": 0.06489401591336032,
      "learning_rate": 8.051023188607248e-05,
      "loss": 0.07241814136505127,
      "step": 6740
    },
    {
      "epoch": 0.31336191910997796,
      "grad_norm": 0.06750469321148132,
      "learning_rate": 8.039115031821683e-05,
      "loss": 0.07176982164382935,
      "step": 6760
    },
    {
      "epoch": 0.3142890253795341,
      "grad_norm": 0.6202793269168291,
      "learning_rate": 8.027179473455463e-05,
      "loss": 0.07424108982086182,
      "step": 6780
    },
    {
      "epoch": 0.3152161316490903,
      "grad_norm": 0.07751779724646034,
      "learning_rate": 8.015216621123197e-05,
      "loss": 0.0765438735485077,
      "step": 6800
    },
    {
      "epoch": 0.3161432379186464,
      "grad_norm": 0.0667607229232211,
      "learning_rate": 8.003226582685582e-05,
      "loss": 0.0713448166847229,
      "step": 6820
    },
    {
      "epoch": 0.31707034418820257,
      "grad_norm": 0.08641777120387521,
      "learning_rate": 7.991209466248428e-05,
      "loss": 0.08027853965759277,
      "step": 6840
    },
    {
      "epoch": 0.31799745045775873,
      "grad_norm": 0.11507143351815359,
      "learning_rate": 7.979165380161695e-05,
      "loss": 0.07583537697792053,
      "step": 6860
    },
    {
      "epoch": 0.31892455672731485,
      "grad_norm": 0.07948055128085395,
      "learning_rate": 7.967094433018508e-05,
      "loss": 0.0709994912147522,
      "step": 6880
    },
    {
      "epoch": 0.319851662996871,
      "grad_norm": 0.053100329547294475,
      "learning_rate": 7.954996733654178e-05,
      "loss": 0.0740058958530426,
      "step": 6900
    },
    {
      "epoch": 0.3207787692664272,
      "grad_norm": 0.06124290886212812,
      "learning_rate": 7.942872391145221e-05,
      "loss": 0.06868508458137512,
      "step": 6920
    },
    {
      "epoch": 0.3217058755359833,
      "grad_norm": 0.0673254214330765,
      "learning_rate": 7.930721514808383e-05,
      "loss": 0.07134757041931153,
      "step": 6940
    },
    {
      "epoch": 0.32263298180553945,
      "grad_norm": 0.06254964952254342,
      "learning_rate": 7.918544214199636e-05,
      "loss": 0.07370802760124207,
      "step": 6960
    },
    {
      "epoch": 0.3235600880750956,
      "grad_norm": 0.06722065374334649,
      "learning_rate": 7.906340599113208e-05,
      "loss": 0.07434462308883667,
      "step": 6980
    },
    {
      "epoch": 0.32448719434465173,
      "grad_norm": 0.0708645746431618,
      "learning_rate": 7.894110779580585e-05,
      "loss": 0.07249525785446168,
      "step": 7000
    },
    {
      "epoch": 0.3254143006142079,
      "grad_norm": 0.06231997836717292,
      "learning_rate": 7.881854865869522e-05,
      "loss": 0.07076464891433716,
      "step": 7020
    },
    {
      "epoch": 0.32634140688376406,
      "grad_norm": 0.06622291272389368,
      "learning_rate": 7.869572968483043e-05,
      "loss": 0.07391721606254578,
      "step": 7040
    },
    {
      "epoch": 0.3272685131533202,
      "grad_norm": 0.057064678892770795,
      "learning_rate": 7.857265198158452e-05,
      "loss": 0.07716140747070313,
      "step": 7060
    },
    {
      "epoch": 0.32819561942287634,
      "grad_norm": 0.06409469363905317,
      "learning_rate": 7.84493166586633e-05,
      "loss": 0.07295069694519044,
      "step": 7080
    },
    {
      "epoch": 0.3291227256924325,
      "grad_norm": 0.058378821987493314,
      "learning_rate": 7.832572482809535e-05,
      "loss": 0.06818488240242004,
      "step": 7100
    },
    {
      "epoch": 0.3300498319619886,
      "grad_norm": 0.061108725531433455,
      "learning_rate": 7.820187760422202e-05,
      "loss": 0.07450971603393555,
      "step": 7120
    },
    {
      "epoch": 0.3309769382315448,
      "grad_norm": 0.0550301515634861,
      "learning_rate": 7.807777610368734e-05,
      "loss": 0.07034422159194946,
      "step": 7140
    },
    {
      "epoch": 0.33190404450110095,
      "grad_norm": 0.057763999093711216,
      "learning_rate": 7.795342144542801e-05,
      "loss": 0.07236452102661133,
      "step": 7160
    },
    {
      "epoch": 0.3328311507706571,
      "grad_norm": 0.05506320193225926,
      "learning_rate": 7.782881475066323e-05,
      "loss": 0.07370430827140809,
      "step": 7180
    },
    {
      "epoch": 0.3337582570402132,
      "grad_norm": 0.06050669777945925,
      "learning_rate": 7.770395714288468e-05,
      "loss": 0.07665219902992249,
      "step": 7200
    },
    {
      "epoch": 0.3346853633097694,
      "grad_norm": 0.15569569440042622,
      "learning_rate": 7.757884974784631e-05,
      "loss": 0.07170350551605224,
      "step": 7220
    },
    {
      "epoch": 0.33561246957932556,
      "grad_norm": 0.06461803892645793,
      "learning_rate": 7.745349369355426e-05,
      "loss": 0.0720145046710968,
      "step": 7240
    },
    {
      "epoch": 0.33653957584888167,
      "grad_norm": 0.06255603288344817,
      "learning_rate": 7.732789011025665e-05,
      "loss": 0.07081273198127747,
      "step": 7260
    },
    {
      "epoch": 0.33746668211843783,
      "grad_norm": 0.07451431070120196,
      "learning_rate": 7.720204013043336e-05,
      "loss": 0.07302408814430236,
      "step": 7280
    },
    {
      "epoch": 0.338393788387994,
      "grad_norm": 0.05922958757628343,
      "learning_rate": 7.707594488878591e-05,
      "loss": 0.07100723385810852,
      "step": 7300
    },
    {
      "epoch": 0.3393208946575501,
      "grad_norm": 0.07288997675355016,
      "learning_rate": 7.694960552222715e-05,
      "loss": 0.07284167408943176,
      "step": 7320
    },
    {
      "epoch": 0.3402480009271063,
      "grad_norm": 0.05604672420499752,
      "learning_rate": 7.682302316987104e-05,
      "loss": 0.07077852487564087,
      "step": 7340
    },
    {
      "epoch": 0.34117510719666244,
      "grad_norm": 0.05696633005978958,
      "learning_rate": 7.669619897302234e-05,
      "loss": 0.07376844882965088,
      "step": 7360
    },
    {
      "epoch": 0.34210221346621855,
      "grad_norm": 0.07291613992813219,
      "learning_rate": 7.656913407516642e-05,
      "loss": 0.07229496240615844,
      "step": 7380
    },
    {
      "epoch": 0.3430293197357747,
      "grad_norm": 0.058369997478408764,
      "learning_rate": 7.644182962195881e-05,
      "loss": 0.07333348989486695,
      "step": 7400
    },
    {
      "epoch": 0.3439564260053309,
      "grad_norm": 0.061826736402982495,
      "learning_rate": 7.631428676121501e-05,
      "loss": 0.0713516891002655,
      "step": 7420
    },
    {
      "epoch": 0.344883532274887,
      "grad_norm": 0.058855092492190134,
      "learning_rate": 7.618650664290002e-05,
      "loss": 0.07172548770904541,
      "step": 7440
    },
    {
      "epoch": 0.34581063854444316,
      "grad_norm": 0.06405938678101386,
      "learning_rate": 7.605849041911804e-05,
      "loss": 0.06828615069389343,
      "step": 7460
    },
    {
      "epoch": 0.3467377448139993,
      "grad_norm": 0.05145715487198017,
      "learning_rate": 7.593023924410208e-05,
      "loss": 0.06908615231513977,
      "step": 7480
    },
    {
      "epoch": 0.34766485108355544,
      "grad_norm": 0.06652763639415586,
      "learning_rate": 7.580175427420358e-05,
      "loss": 0.07108408808708191,
      "step": 7500
    },
    {
      "epoch": 0.3485919573531116,
      "grad_norm": 0.06798045747095019,
      "learning_rate": 7.567303666788184e-05,
      "loss": 0.07113984227180481,
      "step": 7520
    },
    {
      "epoch": 0.34951906362266777,
      "grad_norm": 0.08109711888615893,
      "learning_rate": 7.55440875856938e-05,
      "loss": 0.07682186961174012,
      "step": 7540
    },
    {
      "epoch": 0.3504461698922239,
      "grad_norm": 0.06137241394535925,
      "learning_rate": 7.541490819028336e-05,
      "loss": 0.07754748463630676,
      "step": 7560
    },
    {
      "epoch": 0.35137327616178005,
      "grad_norm": 0.06743259378293887,
      "learning_rate": 7.528549964637107e-05,
      "loss": 0.07150028944015503,
      "step": 7580
    },
    {
      "epoch": 0.3523003824313362,
      "grad_norm": 0.06369397870650284,
      "learning_rate": 7.515586312074348e-05,
      "loss": 0.07170009016990661,
      "step": 7600
    },
    {
      "epoch": 0.3532274887008923,
      "grad_norm": 0.05778914619164209,
      "learning_rate": 7.502599978224273e-05,
      "loss": 0.07141064405441284,
      "step": 7620
    },
    {
      "epoch": 0.3541545949704485,
      "grad_norm": 0.061098326682360295,
      "learning_rate": 7.4895910801756e-05,
      "loss": 0.07194435596466064,
      "step": 7640
    },
    {
      "epoch": 0.35508170124000465,
      "grad_norm": 0.06363431373867703,
      "learning_rate": 7.476559735220488e-05,
      "loss": 0.07232903242111206,
      "step": 7660
    },
    {
      "epoch": 0.35600880750956077,
      "grad_norm": 0.06049988487252925,
      "learning_rate": 7.463506060853487e-05,
      "loss": 0.0708425521850586,
      "step": 7680
    },
    {
      "epoch": 0.35693591377911693,
      "grad_norm": 0.0658565202284551,
      "learning_rate": 7.450430174770473e-05,
      "loss": 0.07119995355606079,
      "step": 7700
    },
    {
      "epoch": 0.3578630200486731,
      "grad_norm": 0.05831499648624062,
      "learning_rate": 7.437332194867594e-05,
      "loss": 0.0702292799949646,
      "step": 7720
    },
    {
      "epoch": 0.3587901263182292,
      "grad_norm": 0.05457668221592735,
      "learning_rate": 7.424212239240197e-05,
      "loss": 0.06930825114250183,
      "step": 7740
    },
    {
      "epoch": 0.3597172325877854,
      "grad_norm": 0.05928841182112665,
      "learning_rate": 7.411070426181773e-05,
      "loss": 0.07295409440994263,
      "step": 7760
    },
    {
      "epoch": 0.36064433885734154,
      "grad_norm": 0.08190037564809158,
      "learning_rate": 7.397906874182883e-05,
      "loss": 0.07227935194969178,
      "step": 7780
    },
    {
      "epoch": 0.36157144512689765,
      "grad_norm": 0.05589858357091087,
      "learning_rate": 7.384721701930098e-05,
      "loss": 0.06653584837913513,
      "step": 7800
    },
    {
      "epoch": 0.3624985513964538,
      "grad_norm": 0.06250741667715043,
      "learning_rate": 7.371515028304916e-05,
      "loss": 0.06951590776443481,
      "step": 7820
    },
    {
      "epoch": 0.36342565766601,
      "grad_norm": 0.05517052893791215,
      "learning_rate": 7.358286972382707e-05,
      "loss": 0.07067946195602418,
      "step": 7840
    },
    {
      "epoch": 0.3643527639355661,
      "grad_norm": 0.061756213301981555,
      "learning_rate": 7.345037653431621e-05,
      "loss": 0.07223882675170898,
      "step": 7860
    },
    {
      "epoch": 0.36527987020512226,
      "grad_norm": 0.05421920470477643,
      "learning_rate": 7.331767190911528e-05,
      "loss": 0.06952041387557983,
      "step": 7880
    },
    {
      "epoch": 0.3662069764746784,
      "grad_norm": 0.06793019695228361,
      "learning_rate": 7.318475704472936e-05,
      "loss": 0.06834464073181153,
      "step": 7900
    },
    {
      "epoch": 0.36713408274423454,
      "grad_norm": 0.05179279406718639,
      "learning_rate": 7.305163313955905e-05,
      "loss": 0.06984994411468506,
      "step": 7920
    },
    {
      "epoch": 0.3680611890137907,
      "grad_norm": 0.06437040148088448,
      "learning_rate": 7.291830139388978e-05,
      "loss": 0.07342504262924195,
      "step": 7940
    },
    {
      "epoch": 0.36898829528334687,
      "grad_norm": 0.05000389335786004,
      "learning_rate": 7.278476300988088e-05,
      "loss": 0.07468822598457336,
      "step": 7960
    },
    {
      "epoch": 0.369915401552903,
      "grad_norm": 0.058798915110267,
      "learning_rate": 7.265101919155485e-05,
      "loss": 0.071200430393219,
      "step": 7980
    },
    {
      "epoch": 0.37084250782245914,
      "grad_norm": 0.05683027976814328,
      "learning_rate": 7.25170711447864e-05,
      "loss": 0.06964697837829589,
      "step": 8000
    },
    {
      "epoch": 0.3717696140920153,
      "grad_norm": 0.07328666251809984,
      "learning_rate": 7.23829200772917e-05,
      "loss": 0.07012683153152466,
      "step": 8020
    },
    {
      "epoch": 0.3726967203615714,
      "grad_norm": 0.06077711405558852,
      "learning_rate": 7.224856719861727e-05,
      "loss": 0.07310692071914673,
      "step": 8040
    },
    {
      "epoch": 0.3736238266311276,
      "grad_norm": 0.17436557367260727,
      "learning_rate": 7.211401372012941e-05,
      "loss": 0.07378835678100586,
      "step": 8060
    },
    {
      "epoch": 0.37455093290068375,
      "grad_norm": 0.06588927901245954,
      "learning_rate": 7.197926085500295e-05,
      "loss": 0.07211959958076478,
      "step": 8080
    },
    {
      "epoch": 0.37547803917023986,
      "grad_norm": 0.06457999673432532,
      "learning_rate": 7.184430981821049e-05,
      "loss": 0.07012794017791749,
      "step": 8100
    },
    {
      "epoch": 0.37640514543979603,
      "grad_norm": 0.061872521697988195,
      "learning_rate": 7.170916182651141e-05,
      "loss": 0.07442641258239746,
      "step": 8120
    },
    {
      "epoch": 0.3773322517093522,
      "grad_norm": 0.06519802723496167,
      "learning_rate": 7.157381809844091e-05,
      "loss": 0.07120200395584106,
      "step": 8140
    },
    {
      "epoch": 0.3782593579789083,
      "grad_norm": 0.06339687253020734,
      "learning_rate": 7.143827985429897e-05,
      "loss": 0.0700366735458374,
      "step": 8160
    },
    {
      "epoch": 0.37918646424846447,
      "grad_norm": 0.06681205359126911,
      "learning_rate": 7.130254831613943e-05,
      "loss": 0.07206454277038574,
      "step": 8180
    },
    {
      "epoch": 0.38011357051802064,
      "grad_norm": 0.055908417899135074,
      "learning_rate": 7.116662470775889e-05,
      "loss": 0.07202800512313842,
      "step": 8200
    },
    {
      "epoch": 0.38104067678757675,
      "grad_norm": 0.06409176693698779,
      "learning_rate": 7.103051025468574e-05,
      "loss": 0.06932897567749023,
      "step": 8220
    },
    {
      "epoch": 0.3819677830571329,
      "grad_norm": 0.056795766963674076,
      "learning_rate": 7.089420618416905e-05,
      "loss": 0.07277878522872924,
      "step": 8240
    },
    {
      "epoch": 0.3828948893266891,
      "grad_norm": 0.057152880142219024,
      "learning_rate": 7.075771372516759e-05,
      "loss": 0.07027421593666076,
      "step": 8260
    },
    {
      "epoch": 0.38382199559624525,
      "grad_norm": 0.051666913812318256,
      "learning_rate": 7.062103410833866e-05,
      "loss": 0.06899116039276124,
      "step": 8280
    },
    {
      "epoch": 0.38474910186580136,
      "grad_norm": 0.06349348626288141,
      "learning_rate": 7.048416856602702e-05,
      "loss": 0.06868364810943603,
      "step": 8300
    },
    {
      "epoch": 0.3856762081353575,
      "grad_norm": 0.07003885459773933,
      "learning_rate": 7.034711833225384e-05,
      "loss": 0.07406530380249024,
      "step": 8320
    },
    {
      "epoch": 0.3866033144049137,
      "grad_norm": 0.06158091511813546,
      "learning_rate": 7.020988464270546e-05,
      "loss": 0.0713687002658844,
      "step": 8340
    },
    {
      "epoch": 0.3875304206744698,
      "grad_norm": 0.060532678334162,
      "learning_rate": 7.007246873472237e-05,
      "loss": 0.07255741357803344,
      "step": 8360
    },
    {
      "epoch": 0.38845752694402597,
      "grad_norm": 0.058219006894719004,
      "learning_rate": 6.993487184728798e-05,
      "loss": 0.07127118110656738,
      "step": 8380
    },
    {
      "epoch": 0.38938463321358213,
      "grad_norm": 0.0548943116658426,
      "learning_rate": 6.979709522101744e-05,
      "loss": 0.06869758367538452,
      "step": 8400
    },
    {
      "epoch": 0.39031173948313824,
      "grad_norm": 0.056120822949716855,
      "learning_rate": 6.965914009814651e-05,
      "loss": 0.07157270908355713,
      "step": 8420
    },
    {
      "epoch": 0.3912388457526944,
      "grad_norm": 0.0493489866976689,
      "learning_rate": 6.952100772252031e-05,
      "loss": 0.07348506450653076,
      "step": 8440
    },
    {
      "epoch": 0.3921659520222506,
      "grad_norm": 0.055528731770065044,
      "learning_rate": 6.938269933958213e-05,
      "loss": 0.06816398501396179,
      "step": 8460
    },
    {
      "epoch": 0.3930930582918067,
      "grad_norm": 0.0621921626601533,
      "learning_rate": 6.924421619636221e-05,
      "loss": 0.07218122482299805,
      "step": 8480
    },
    {
      "epoch": 0.39402016456136285,
      "grad_norm": 0.06063536308655912,
      "learning_rate": 6.910555954146647e-05,
      "loss": 0.07273417115211486,
      "step": 8500
    },
    {
      "epoch": 0.394947270830919,
      "grad_norm": 0.04720623129924152,
      "learning_rate": 6.896673062506524e-05,
      "loss": 0.06545677185058593,
      "step": 8520
    },
    {
      "epoch": 0.3958743771004751,
      "grad_norm": 0.051744171391392924,
      "learning_rate": 6.882773069888208e-05,
      "loss": 0.07149837613105774,
      "step": 8540
    },
    {
      "epoch": 0.3968014833700313,
      "grad_norm": 0.056166318531532596,
      "learning_rate": 6.868856101618233e-05,
      "loss": 0.07216413021087646,
      "step": 8560
    },
    {
      "epoch": 0.39772858963958746,
      "grad_norm": 0.05936381490265723,
      "learning_rate": 6.854922283176198e-05,
      "loss": 0.07196494340896606,
      "step": 8580
    },
    {
      "epoch": 0.39865569590914357,
      "grad_norm": 0.0682228656637518,
      "learning_rate": 6.840971740193627e-05,
      "loss": 0.07325062155723572,
      "step": 8600
    },
    {
      "epoch": 0.39958280217869974,
      "grad_norm": 0.05546642724987347,
      "learning_rate": 6.827004598452833e-05,
      "loss": 0.0680731177330017,
      "step": 8620
    },
    {
      "epoch": 0.4005099084482559,
      "grad_norm": 0.05564747581452941,
      "learning_rate": 6.813020983885796e-05,
      "loss": 0.06899272203445435,
      "step": 8640
    },
    {
      "epoch": 0.401437014717812,
      "grad_norm": 1.2949908906806882,
      "learning_rate": 6.799021022573014e-05,
      "loss": 0.07849831581115722,
      "step": 8660
    },
    {
      "epoch": 0.4023641209873682,
      "grad_norm": 0.07335413294996637,
      "learning_rate": 6.785004840742374e-05,
      "loss": 0.07621760964393616,
      "step": 8680
    },
    {
      "epoch": 0.40329122725692435,
      "grad_norm": 0.06597642261510535,
      "learning_rate": 6.770972564768013e-05,
      "loss": 0.07188231945037842,
      "step": 8700
    },
    {
      "epoch": 0.40421833352648046,
      "grad_norm": 0.06054969909859594,
      "learning_rate": 6.756924321169174e-05,
      "loss": 0.07226096987724304,
      "step": 8720
    },
    {
      "epoch": 0.4051454397960366,
      "grad_norm": 0.06476750763866697,
      "learning_rate": 6.742860236609077e-05,
      "loss": 0.07668436765670776,
      "step": 8740
    },
    {
      "epoch": 0.4060725460655928,
      "grad_norm": 0.06570224859129872,
      "learning_rate": 6.728780437893758e-05,
      "loss": 0.0732303261756897,
      "step": 8760
    },
    {
      "epoch": 0.4069996523351489,
      "grad_norm": 0.17606852583759058,
      "learning_rate": 6.714685051970947e-05,
      "loss": 0.07211008071899414,
      "step": 8780
    },
    {
      "epoch": 0.40792675860470506,
      "grad_norm": 0.06741211242743908,
      "learning_rate": 6.700574205928904e-05,
      "loss": 0.0737863302230835,
      "step": 8800
    },
    {
      "epoch": 0.40885386487426123,
      "grad_norm": 0.10734441315597312,
      "learning_rate": 6.686448026995289e-05,
      "loss": 0.0698961079120636,
      "step": 8820
    },
    {
      "epoch": 0.40978097114381734,
      "grad_norm": 0.05796649469094367,
      "learning_rate": 6.672306642536006e-05,
      "loss": 0.06932593584060669,
      "step": 8840
    },
    {
      "epoch": 0.4107080774133735,
      "grad_norm": 0.06988486673932746,
      "learning_rate": 6.658150180054054e-05,
      "loss": 0.07208383083343506,
      "step": 8860
    },
    {
      "epoch": 0.4116351836829297,
      "grad_norm": 0.06171679349676563,
      "learning_rate": 6.643978767188382e-05,
      "loss": 0.07484859824180604,
      "step": 8880
    },
    {
      "epoch": 0.4125622899524858,
      "grad_norm": 0.06200243094075337,
      "learning_rate": 6.629792531712737e-05,
      "loss": 0.06866522431373596,
      "step": 8900
    },
    {
      "epoch": 0.41348939622204195,
      "grad_norm": 0.054055272643492025,
      "learning_rate": 6.615591601534513e-05,
      "loss": 0.07249740362167359,
      "step": 8920
    },
    {
      "epoch": 0.4144165024915981,
      "grad_norm": 0.05167398871468604,
      "learning_rate": 6.601376104693586e-05,
      "loss": 0.07154306769371033,
      "step": 8940
    },
    {
      "epoch": 0.4153436087611542,
      "grad_norm": 0.058860566879181016,
      "learning_rate": 6.587146169361183e-05,
      "loss": 0.0688104510307312,
      "step": 8960
    },
    {
      "epoch": 0.4162707150307104,
      "grad_norm": 0.05876534091601072,
      "learning_rate": 6.572901923838704e-05,
      "loss": 0.06802005767822265,
      "step": 8980
    },
    {
      "epoch": 0.41719782130026656,
      "grad_norm": 0.0664856049115416,
      "learning_rate": 6.558643496556575e-05,
      "loss": 0.07041721343994141,
      "step": 9000
    },
    {
      "epoch": 0.41812492756982267,
      "grad_norm": 0.059538986754152305,
      "learning_rate": 6.544371016073094e-05,
      "loss": 0.06777328252792358,
      "step": 9020
    },
    {
      "epoch": 0.41905203383937883,
      "grad_norm": 0.05604144071917939,
      "learning_rate": 6.530084611073263e-05,
      "loss": 0.06975816488265991,
      "step": 9040
    },
    {
      "epoch": 0.419979140108935,
      "grad_norm": 0.07693169082929725,
      "learning_rate": 6.515784410367629e-05,
      "loss": 0.07223490476608277,
      "step": 9060
    },
    {
      "epoch": 0.4209062463784911,
      "grad_norm": 0.048894009845641705,
      "learning_rate": 6.50147054289113e-05,
      "loss": 0.06806935667991638,
      "step": 9080
    },
    {
      "epoch": 0.4218333526480473,
      "grad_norm": 0.0682401850594772,
      "learning_rate": 6.487143137701925e-05,
      "loss": 0.06676151156425476,
      "step": 9100
    },
    {
      "epoch": 0.42276045891760344,
      "grad_norm": 0.06531126640079608,
      "learning_rate": 6.472802323980237e-05,
      "loss": 0.0679443895816803,
      "step": 9120
    },
    {
      "epoch": 0.42368756518715955,
      "grad_norm": 0.05718211547065105,
      "learning_rate": 6.458448231027179e-05,
      "loss": 0.06895342469215393,
      "step": 9140
    },
    {
      "epoch": 0.4246146714567157,
      "grad_norm": 0.051432666547099354,
      "learning_rate": 6.444080988263597e-05,
      "loss": 0.07011189460754394,
      "step": 9160
    },
    {
      "epoch": 0.4255417777262719,
      "grad_norm": 0.0572939459915378,
      "learning_rate": 6.429700725228896e-05,
      "loss": 0.06691262722015381,
      "step": 9180
    },
    {
      "epoch": 0.426468883995828,
      "grad_norm": 0.06186676312860538,
      "learning_rate": 6.415307571579883e-05,
      "loss": 0.06950963139533997,
      "step": 9200
    },
    {
      "epoch": 0.42739599026538416,
      "grad_norm": 0.06881535899999948,
      "learning_rate": 6.400901657089584e-05,
      "loss": 0.07066357135772705,
      "step": 9220
    },
    {
      "epoch": 0.42832309653494033,
      "grad_norm": 0.056869801967248765,
      "learning_rate": 6.386483111646082e-05,
      "loss": 0.06954029798507691,
      "step": 9240
    },
    {
      "epoch": 0.42925020280449644,
      "grad_norm": 0.05923794372081229,
      "learning_rate": 6.372052065251346e-05,
      "loss": 0.06746643781661987,
      "step": 9260
    },
    {
      "epoch": 0.4301773090740526,
      "grad_norm": 0.06167052205820658,
      "learning_rate": 6.357608648020057e-05,
      "loss": 0.06677300930023193,
      "step": 9280
    },
    {
      "epoch": 0.43110441534360877,
      "grad_norm": 0.060839659939225466,
      "learning_rate": 6.343152990178435e-05,
      "loss": 0.06932535767555237,
      "step": 9300
    },
    {
      "epoch": 0.4320315216131649,
      "grad_norm": 0.05487105279897788,
      "learning_rate": 6.328685222063063e-05,
      "loss": 0.06984179615974426,
      "step": 9320
    },
    {
      "epoch": 0.43295862788272105,
      "grad_norm": 0.05648488259068478,
      "learning_rate": 6.314205474119718e-05,
      "loss": 0.06659715175628662,
      "step": 9340
    },
    {
      "epoch": 0.4338857341522772,
      "grad_norm": 0.9903433461688329,
      "learning_rate": 6.299713876902187e-05,
      "loss": 0.07139703035354614,
      "step": 9360
    },
    {
      "epoch": 0.4348128404218334,
      "grad_norm": 0.057132294455422956,
      "learning_rate": 6.285210561071097e-05,
      "loss": 0.07028393745422364,
      "step": 9380
    },
    {
      "epoch": 0.4357399466913895,
      "grad_norm": 0.05857614496427812,
      "learning_rate": 6.270695657392729e-05,
      "loss": 0.0702647626399994,
      "step": 9400
    },
    {
      "epoch": 0.43666705296094566,
      "grad_norm": 0.056661560519639706,
      "learning_rate": 6.256169296737848e-05,
      "loss": 0.06969016790390015,
      "step": 9420
    },
    {
      "epoch": 0.4375941592305018,
      "grad_norm": 0.49177246720542006,
      "learning_rate": 6.241631610080518e-05,
      "loss": 0.07438609600067139,
      "step": 9440
    },
    {
      "epoch": 0.43852126550005793,
      "grad_norm": 2.431655313212214,
      "learning_rate": 6.227082728496915e-05,
      "loss": 0.08695851564407349,
      "step": 9460
    },
    {
      "epoch": 0.4394483717696141,
      "grad_norm": 0.07659818739561748,
      "learning_rate": 6.212522783164165e-05,
      "loss": 0.08092879056930542,
      "step": 9480
    },
    {
      "epoch": 0.44037547803917027,
      "grad_norm": 0.059975666347355495,
      "learning_rate": 6.197951905359135e-05,
      "loss": 0.07423359751701356,
      "step": 9500
    },
    {
      "epoch": 0.4413025843087264,
      "grad_norm": 0.08370475715268878,
      "learning_rate": 6.183370226457271e-05,
      "loss": 0.07136415243148804,
      "step": 9520
    },
    {
      "epoch": 0.44222969057828254,
      "grad_norm": 0.054777010621705036,
      "learning_rate": 6.1687778779314e-05,
      "loss": 0.06574440598487855,
      "step": 9540
    },
    {
      "epoch": 0.4431567968478387,
      "grad_norm": 0.05580460468968765,
      "learning_rate": 6.15417499135055e-05,
      "loss": 0.07091556787490845,
      "step": 9560
    },
    {
      "epoch": 0.4440839031173948,
      "grad_norm": 0.06406101930139682,
      "learning_rate": 6.139561698378767e-05,
      "loss": 0.07055695652961731,
      "step": 9580
    },
    {
      "epoch": 0.445011009386951,
      "grad_norm": 0.058953741063551085,
      "learning_rate": 6.124938130773924e-05,
      "loss": 0.06751711368560791,
      "step": 9600
    },
    {
      "epoch": 0.44593811565650715,
      "grad_norm": 0.05338656233981756,
      "learning_rate": 6.110304420386527e-05,
      "loss": 0.06980046033859252,
      "step": 9620
    },
    {
      "epoch": 0.44686522192606326,
      "grad_norm": 0.05063214746006459,
      "learning_rate": 6.095660699158538e-05,
      "loss": 0.06688464283943177,
      "step": 9640
    },
    {
      "epoch": 0.4477923281956194,
      "grad_norm": 0.06333427274798475,
      "learning_rate": 6.081007099122178e-05,
      "loss": 0.0685849666595459,
      "step": 9660
    },
    {
      "epoch": 0.4487194344651756,
      "grad_norm": 0.058378938623546935,
      "learning_rate": 6.0663437523987396e-05,
      "loss": 0.0672727346420288,
      "step": 9680
    },
    {
      "epoch": 0.4496465407347317,
      "grad_norm": 0.04971631507606275,
      "learning_rate": 6.051670791197392e-05,
      "loss": 0.0664828896522522,
      "step": 9700
    },
    {
      "epoch": 0.45057364700428787,
      "grad_norm": 0.05161097533161306,
      "learning_rate": 6.036988347813994e-05,
      "loss": 0.07142003178596497,
      "step": 9720
    },
    {
      "epoch": 0.45150075327384404,
      "grad_norm": 0.04923131947397141,
      "learning_rate": 6.022296554629898e-05,
      "loss": 0.0693086862564087,
      "step": 9740
    },
    {
      "epoch": 0.45242785954340015,
      "grad_norm": 0.058661885026131615,
      "learning_rate": 6.0075955441107555e-05,
      "loss": 0.06907542943954467,
      "step": 9760
    },
    {
      "epoch": 0.4533549658129563,
      "grad_norm": 0.05368519821103752,
      "learning_rate": 5.9928854488053244e-05,
      "loss": 0.07158472537994384,
      "step": 9780
    },
    {
      "epoch": 0.4542820720825125,
      "grad_norm": 0.1663674248877333,
      "learning_rate": 5.9781664013442764e-05,
      "loss": 0.06934982538223267,
      "step": 9800
    },
    {
      "epoch": 0.4552091783520686,
      "grad_norm": 0.05497490906186732,
      "learning_rate": 5.963438534438997e-05,
      "loss": 0.0683956801891327,
      "step": 9820
    },
    {
      "epoch": 0.45613628462162475,
      "grad_norm": 0.05190712437348263,
      "learning_rate": 5.948701980880388e-05,
      "loss": 0.07192785739898681,
      "step": 9840
    },
    {
      "epoch": 0.4570633908911809,
      "grad_norm": 0.058053203801108205,
      "learning_rate": 5.933956873537677e-05,
      "loss": 0.06807593107223511,
      "step": 9860
    },
    {
      "epoch": 0.45799049716073703,
      "grad_norm": 0.05263219187239627,
      "learning_rate": 5.919203345357214e-05,
      "loss": 0.07042191028594971,
      "step": 9880
    },
    {
      "epoch": 0.4589176034302932,
      "grad_norm": 0.08964227753647068,
      "learning_rate": 5.904441529361271e-05,
      "loss": 0.06729475259780884,
      "step": 9900
    },
    {
      "epoch": 0.45984470969984936,
      "grad_norm": 0.053198066270887204,
      "learning_rate": 5.889671558646848e-05,
      "loss": 0.0692562460899353,
      "step": 9920
    },
    {
      "epoch": 0.4607718159694055,
      "grad_norm": 0.05872038104678272,
      "learning_rate": 5.87489356638447e-05,
      "loss": 0.07206550240516663,
      "step": 9940
    },
    {
      "epoch": 0.46169892223896164,
      "grad_norm": 0.059633965063798124,
      "learning_rate": 5.8601076858169856e-05,
      "loss": 0.06614031791687011,
      "step": 9960
    },
    {
      "epoch": 0.4626260285085178,
      "grad_norm": 0.05716560277383062,
      "learning_rate": 5.84531405025837e-05,
      "loss": 0.06638523340225219,
      "step": 9980
    },
    {
      "epoch": 0.4635531347780739,
      "grad_norm": 0.05359247761990254,
      "learning_rate": 5.830512793092516e-05,
      "loss": 0.06717640161514282,
      "step": 10000
    },
    {
      "epoch": 0.4644802410476301,
      "grad_norm": 0.05403999393689296,
      "learning_rate": 5.815704047772037e-05,
      "loss": 0.06788140535354614,
      "step": 10020
    },
    {
      "epoch": 0.46540734731718625,
      "grad_norm": 0.05311465843493346,
      "learning_rate": 5.80088794781706e-05,
      "loss": 0.06636354923248292,
      "step": 10040
    },
    {
      "epoch": 0.46633445358674236,
      "grad_norm": 0.050376362602807485,
      "learning_rate": 5.786064626814026e-05,
      "loss": 0.07039104700088501,
      "step": 10060
    },
    {
      "epoch": 0.4672615598562985,
      "grad_norm": 0.05034803848797613,
      "learning_rate": 5.771234218414482e-05,
      "loss": 0.07016848921775817,
      "step": 10080
    },
    {
      "epoch": 0.4681886661258547,
      "grad_norm": 0.05277104901816341,
      "learning_rate": 5.7563968563338776e-05,
      "loss": 0.06645816564559937,
      "step": 10100
    },
    {
      "epoch": 0.4691157723954108,
      "grad_norm": 0.057917775279076426,
      "learning_rate": 5.741552674350359e-05,
      "loss": 0.0701315939426422,
      "step": 10120
    },
    {
      "epoch": 0.47004287866496697,
      "grad_norm": 0.05925733296312407,
      "learning_rate": 5.7267018063035604e-05,
      "loss": 0.06787375807762146,
      "step": 10140
    },
    {
      "epoch": 0.47096998493452313,
      "grad_norm": 0.04761028131615447,
      "learning_rate": 5.711844386093402e-05,
      "loss": 0.06677426099777221,
      "step": 10160
    },
    {
      "epoch": 0.47189709120407924,
      "grad_norm": 0.05249637835375034,
      "learning_rate": 5.696980547678881e-05,
      "loss": 0.06861708760261535,
      "step": 10180
    },
    {
      "epoch": 0.4728241974736354,
      "grad_norm": 0.05149740032105075,
      "learning_rate": 5.682110425076862e-05,
      "loss": 0.0658948540687561,
      "step": 10200
    },
    {
      "epoch": 0.4737513037431916,
      "grad_norm": 0.06264528738329353,
      "learning_rate": 5.667234152360867e-05,
      "loss": 0.06772146821022033,
      "step": 10220
    },
    {
      "epoch": 0.4746784100127477,
      "grad_norm": 0.1976627391132156,
      "learning_rate": 5.652351863659876e-05,
      "loss": 0.06549715995788574,
      "step": 10240
    },
    {
      "epoch": 0.47560551628230385,
      "grad_norm": 0.05451382423683908,
      "learning_rate": 5.637463693157102e-05,
      "loss": 0.06611493825912476,
      "step": 10260
    },
    {
      "epoch": 0.47653262255186,
      "grad_norm": 0.058264134874698545,
      "learning_rate": 5.6225697750888e-05,
      "loss": 0.0714371919631958,
      "step": 10280
    },
    {
      "epoch": 0.47745972882141613,
      "grad_norm": 0.05335102330542413,
      "learning_rate": 5.607670243743036e-05,
      "loss": 0.06993522644042968,
      "step": 10300
    },
    {
      "epoch": 0.4783868350909723,
      "grad_norm": 0.06184291763398293,
      "learning_rate": 5.5927652334584965e-05,
      "loss": 0.07093091011047363,
      "step": 10320
    },
    {
      "epoch": 0.47931394136052846,
      "grad_norm": 0.05321181478734491,
      "learning_rate": 5.577854878623263e-05,
      "loss": 0.06893997192382813,
      "step": 10340
    },
    {
      "epoch": 0.48024104763008457,
      "grad_norm": 0.058539063565010464,
      "learning_rate": 5.5629393136736055e-05,
      "loss": 0.06837838292121887,
      "step": 10360
    },
    {
      "epoch": 0.48116815389964074,
      "grad_norm": 0.04996568856410487,
      "learning_rate": 5.548018673092772e-05,
      "loss": 0.06891497373580932,
      "step": 10380
    },
    {
      "epoch": 0.4820952601691969,
      "grad_norm": 0.056018382056293656,
      "learning_rate": 5.53309309140977e-05,
      "loss": 0.06571992039680481,
      "step": 10400
    },
    {
      "epoch": 0.483022366438753,
      "grad_norm": 0.051684552528505276,
      "learning_rate": 5.518162703198162e-05,
      "loss": 0.07190435528755187,
      "step": 10420
    },
    {
      "epoch": 0.4839494727083092,
      "grad_norm": 0.05512715067080913,
      "learning_rate": 5.503227643074844e-05,
      "loss": 0.06946102976799011,
      "step": 10440
    },
    {
      "epoch": 0.48487657897786535,
      "grad_norm": 0.049556007367337174,
      "learning_rate": 5.48828804569884e-05,
      "loss": 0.07101499438285827,
      "step": 10460
    },
    {
      "epoch": 0.4858036852474215,
      "grad_norm": 0.05629932700378504,
      "learning_rate": 5.473344045770075e-05,
      "loss": 0.06669636368751526,
      "step": 10480
    },
    {
      "epoch": 0.4867307915169776,
      "grad_norm": 0.06209023341037895,
      "learning_rate": 5.458395778028178e-05,
      "loss": 0.06712688803672791,
      "step": 10500
    },
    {
      "epoch": 0.4876578977865338,
      "grad_norm": 0.05273016441704952,
      "learning_rate": 5.443443377251252e-05,
      "loss": 0.06710394620895385,
      "step": 10520
    },
    {
      "epoch": 0.48858500405608996,
      "grad_norm": 0.0540427264064011,
      "learning_rate": 5.428486978254666e-05,
      "loss": 0.06847825646400452,
      "step": 10540
    },
    {
      "epoch": 0.48951211032564607,
      "grad_norm": 0.06840437885213248,
      "learning_rate": 5.413526715889841e-05,
      "loss": 0.06999188065528869,
      "step": 10560
    },
    {
      "epoch": 0.49043921659520223,
      "grad_norm": 0.050435145636937614,
      "learning_rate": 5.398562725043027e-05,
      "loss": 0.06642810106277466,
      "step": 10580
    },
    {
      "epoch": 0.4913663228647584,
      "grad_norm": 0.055102485921551723,
      "learning_rate": 5.383595140634093e-05,
      "loss": 0.06909053325653076,
      "step": 10600
    },
    {
      "epoch": 0.4922934291343145,
      "grad_norm": 0.05074835267327844,
      "learning_rate": 5.36862409761531e-05,
      "loss": 0.06797543168067932,
      "step": 10620
    },
    {
      "epoch": 0.4932205354038707,
      "grad_norm": 0.0540049358738909,
      "learning_rate": 5.35364973097013e-05,
      "loss": 0.06855064630508423,
      "step": 10640
    },
    {
      "epoch": 0.49414764167342684,
      "grad_norm": 0.054127269384924126,
      "learning_rate": 5.338672175711975e-05,
      "loss": 0.06768226623535156,
      "step": 10660
    },
    {
      "epoch": 0.49507474794298295,
      "grad_norm": 0.057025127996070325,
      "learning_rate": 5.3236915668830123e-05,
      "loss": 0.06507717370986939,
      "step": 10680
    },
    {
      "epoch": 0.4960018542125391,
      "grad_norm": 0.05199872229811869,
      "learning_rate": 5.308708039552946e-05,
      "loss": 0.06520705223083496,
      "step": 10700
    },
    {
      "epoch": 0.4969289604820953,
      "grad_norm": 0.049759609224732174,
      "learning_rate": 5.293721728817791e-05,
      "loss": 0.06770432591438294,
      "step": 10720
    },
    {
      "epoch": 0.4978560667516514,
      "grad_norm": 0.051826842264391874,
      "learning_rate": 5.27873276979866e-05,
      "loss": 0.06897132396697998,
      "step": 10740
    },
    {
      "epoch": 0.49878317302120756,
      "grad_norm": 0.058461179945935596,
      "learning_rate": 5.263741297640542e-05,
      "loss": 0.06364284753799439,
      "step": 10760
    },
    {
      "epoch": 0.4997102792907637,
      "grad_norm": 0.05900068621105399,
      "learning_rate": 5.2487474475110854e-05,
      "loss": 0.06831877827644348,
      "step": 10780
    },
    {
      "epoch": 0.5006373855603199,
      "grad_norm": 0.05274733772525755,
      "learning_rate": 5.233751354599381e-05,
      "loss": 0.06658394932746887,
      "step": 10800
    },
    {
      "epoch": 0.501564491829876,
      "grad_norm": 0.06036227581341777,
      "learning_rate": 5.218753154114738e-05,
      "loss": 0.06891183257102966,
      "step": 10820
    },
    {
      "epoch": 0.5024915980994321,
      "grad_norm": 0.049312918784907116,
      "learning_rate": 5.203752981285471e-05,
      "loss": 0.06287815570831298,
      "step": 10840
    },
    {
      "epoch": 0.5034187043689883,
      "grad_norm": 0.056650313251153286,
      "learning_rate": 5.1887509713576766e-05,
      "loss": 0.06659935712814331,
      "step": 10860
    },
    {
      "epoch": 0.5043458106385444,
      "grad_norm": 0.0539589465426644,
      "learning_rate": 5.173747259594016e-05,
      "loss": 0.06665378212928771,
      "step": 10880
    },
    {
      "epoch": 0.5052729169081006,
      "grad_norm": 0.05685820151662816,
      "learning_rate": 5.158741981272493e-05,
      "loss": 0.06755189895629883,
      "step": 10900
    },
    {
      "epoch": 0.5062000231776568,
      "grad_norm": 0.05864271348869779,
      "learning_rate": 5.143735271685235e-05,
      "loss": 0.06776425242424011,
      "step": 10920
    },
    {
      "epoch": 0.5071271294472128,
      "grad_norm": 0.05675119679348412,
      "learning_rate": 5.12872726613728e-05,
      "loss": 0.06746330261230468,
      "step": 10940
    },
    {
      "epoch": 0.508054235716769,
      "grad_norm": 0.054339224793516355,
      "learning_rate": 5.113718099945346e-05,
      "loss": 0.06903930902481079,
      "step": 10960
    },
    {
      "epoch": 0.5089813419863252,
      "grad_norm": 0.05008097771151558,
      "learning_rate": 5.098707908436615e-05,
      "loss": 0.06670130491256714,
      "step": 10980
    },
    {
      "epoch": 0.5099084482558813,
      "grad_norm": 0.05504853464209096,
      "learning_rate": 5.083696826947517e-05,
      "loss": 0.0668412983417511,
      "step": 11000
    },
    {
      "epoch": 0.5108355545254375,
      "grad_norm": 0.05266694995604862,
      "learning_rate": 5.0686849908225045e-05,
      "loss": 0.06841858625411987,
      "step": 11020
    },
    {
      "epoch": 0.5117626607949937,
      "grad_norm": 0.06149237662462454,
      "learning_rate": 5.053672535412831e-05,
      "loss": 0.06596924662590027,
      "step": 11040
    },
    {
      "epoch": 0.5126897670645498,
      "grad_norm": 0.05222277698316076,
      "learning_rate": 5.038659596075341e-05,
      "loss": 0.06674530506134033,
      "step": 11060
    },
    {
      "epoch": 0.5136168733341059,
      "grad_norm": 0.056114794605305864,
      "learning_rate": 5.023646308171234e-05,
      "loss": 0.06750274896621704,
      "step": 11080
    },
    {
      "epoch": 0.514543979603662,
      "grad_norm": 0.053523643458517275,
      "learning_rate": 5.008632807064859e-05,
      "loss": 0.066013765335083,
      "step": 11100
    },
    {
      "epoch": 0.5154710858732182,
      "grad_norm": 0.05196810517246805,
      "learning_rate": 4.993619228122482e-05,
      "loss": 0.068993079662323,
      "step": 11120
    },
    {
      "epoch": 0.5163981921427744,
      "grad_norm": 0.05685300904914473,
      "learning_rate": 4.978605706711076e-05,
      "loss": 0.0662363350391388,
      "step": 11140
    },
    {
      "epoch": 0.5173252984123305,
      "grad_norm": 0.053580147887131516,
      "learning_rate": 4.963592378197091e-05,
      "loss": 0.06478111147880554,
      "step": 11160
    },
    {
      "epoch": 0.5182524046818867,
      "grad_norm": 0.04581969675577541,
      "learning_rate": 4.948579377945241e-05,
      "loss": 0.06617172956466674,
      "step": 11180
    },
    {
      "epoch": 0.5191795109514428,
      "grad_norm": 0.054856376597437584,
      "learning_rate": 4.9335668413172767e-05,
      "loss": 0.06699142456054688,
      "step": 11200
    },
    {
      "epoch": 0.5201066172209989,
      "grad_norm": 0.05340832148427101,
      "learning_rate": 4.918554903670771e-05,
      "loss": 0.0696499764919281,
      "step": 11220
    },
    {
      "epoch": 0.5210337234905551,
      "grad_norm": 0.04924685276202388,
      "learning_rate": 4.903543700357897e-05,
      "loss": 0.06533169746398926,
      "step": 11240
    },
    {
      "epoch": 0.5219608297601113,
      "grad_norm": 0.051275638233785815,
      "learning_rate": 4.888533366724204e-05,
      "loss": 0.0666303277015686,
      "step": 11260
    },
    {
      "epoch": 0.5228879360296674,
      "grad_norm": 0.05141946486146715,
      "learning_rate": 4.873524038107402e-05,
      "loss": 0.06471054553985596,
      "step": 11280
    },
    {
      "epoch": 0.5238150422992236,
      "grad_norm": 0.046924908433075424,
      "learning_rate": 4.8585158498361405e-05,
      "loss": 0.06533972024917603,
      "step": 11300
    },
    {
      "epoch": 0.5247421485687797,
      "grad_norm": 0.05109932478273038,
      "learning_rate": 4.843508937228784e-05,
      "loss": 0.06648097038269044,
      "step": 11320
    },
    {
      "epoch": 0.5256692548383358,
      "grad_norm": 0.05580726425855662,
      "learning_rate": 4.8285034355921986e-05,
      "loss": 0.06593974232673645,
      "step": 11340
    },
    {
      "epoch": 0.526596361107892,
      "grad_norm": 0.05663305655301178,
      "learning_rate": 4.8134994802205256e-05,
      "loss": 0.06652325391769409,
      "step": 11360
    },
    {
      "epoch": 0.5275234673774482,
      "grad_norm": 0.04993353906670951,
      "learning_rate": 4.798497206393969e-05,
      "loss": 0.06556081771850586,
      "step": 11380
    },
    {
      "epoch": 0.5284505736470043,
      "grad_norm": 0.05658470731678929,
      "learning_rate": 4.783496749377564e-05,
      "loss": 0.06470659971237183,
      "step": 11400
    },
    {
      "epoch": 0.5293776799165605,
      "grad_norm": 0.07057079096583119,
      "learning_rate": 4.7684982444199737e-05,
      "loss": 0.06594622135162354,
      "step": 11420
    },
    {
      "epoch": 0.5303047861861165,
      "grad_norm": 0.055589204182857545,
      "learning_rate": 4.753501826752258e-05,
      "loss": 0.06553026437759399,
      "step": 11440
    },
    {
      "epoch": 0.5312318924556727,
      "grad_norm": 0.048263553811536776,
      "learning_rate": 4.73850763158665e-05,
      "loss": 0.06758419275283814,
      "step": 11460
    },
    {
      "epoch": 0.5321589987252289,
      "grad_norm": 0.05574430133400558,
      "learning_rate": 4.7235157941153575e-05,
      "loss": 0.06501986980438232,
      "step": 11480
    },
    {
      "epoch": 0.533086104994785,
      "grad_norm": 0.05382876578276598,
      "learning_rate": 4.7085264495093166e-05,
      "loss": 0.06543768048286439,
      "step": 11500
    },
    {
      "epoch": 0.5340132112643412,
      "grad_norm": 0.04905062534578719,
      "learning_rate": 4.693539732916998e-05,
      "loss": 0.06361998319625854,
      "step": 11520
    },
    {
      "epoch": 0.5349403175338974,
      "grad_norm": 0.058583373766727447,
      "learning_rate": 4.67855577946317e-05,
      "loss": 0.06510854363441468,
      "step": 11540
    },
    {
      "epoch": 0.5358674238034534,
      "grad_norm": 0.06271909462208268,
      "learning_rate": 4.663574724247692e-05,
      "loss": 0.06480238437652588,
      "step": 11560
    },
    {
      "epoch": 0.5367945300730096,
      "grad_norm": 0.052705991066855765,
      "learning_rate": 4.648596702344291e-05,
      "loss": 0.06369982361793518,
      "step": 11580
    },
    {
      "epoch": 0.5377216363425658,
      "grad_norm": 0.05926976832855315,
      "learning_rate": 4.633621848799343e-05,
      "loss": 0.06559765338897705,
      "step": 11600
    },
    {
      "epoch": 0.5386487426121219,
      "grad_norm": 0.05751658876680402,
      "learning_rate": 4.618650298630658e-05,
      "loss": 0.06657824516296387,
      "step": 11620
    },
    {
      "epoch": 0.5395758488816781,
      "grad_norm": 0.05052051106909937,
      "learning_rate": 4.6036821868262646e-05,
      "loss": 0.06554304361343384,
      "step": 11640
    },
    {
      "epoch": 0.5405029551512343,
      "grad_norm": 0.056450630804972625,
      "learning_rate": 4.588717648343185e-05,
      "loss": 0.06605019569396972,
      "step": 11660
    },
    {
      "epoch": 0.5414300614207903,
      "grad_norm": 0.04853403732479463,
      "learning_rate": 4.573756818106228e-05,
      "loss": 0.06343193650245667,
      "step": 11680
    },
    {
      "epoch": 0.5423571676903465,
      "grad_norm": 0.0551831978788852,
      "learning_rate": 4.558799831006764e-05,
      "loss": 0.0670366644859314,
      "step": 11700
    },
    {
      "epoch": 0.5432842739599026,
      "grad_norm": 0.04961887773327719,
      "learning_rate": 4.543846821901515e-05,
      "loss": 0.066197270154953,
      "step": 11720
    },
    {
      "epoch": 0.5442113802294588,
      "grad_norm": 0.05673463945275406,
      "learning_rate": 4.528897925611334e-05,
      "loss": 0.06722276210784912,
      "step": 11740
    },
    {
      "epoch": 0.545138486499015,
      "grad_norm": 0.05466206496671141,
      "learning_rate": 4.513953276919996e-05,
      "loss": 0.06649364233016967,
      "step": 11760
    },
    {
      "epoch": 0.5460655927685711,
      "grad_norm": 0.05373389712004471,
      "learning_rate": 4.4990130105729726e-05,
      "loss": 0.06585872769355774,
      "step": 11780
    },
    {
      "epoch": 0.5469926990381272,
      "grad_norm": 0.04888287430472321,
      "learning_rate": 4.484077261276224e-05,
      "loss": 0.06578683853149414,
      "step": 11800
    },
    {
      "epoch": 0.5479198053076834,
      "grad_norm": 0.048702759078127496,
      "learning_rate": 4.4691461636949895e-05,
      "loss": 0.0591300904750824,
      "step": 11820
    },
    {
      "epoch": 0.5488469115772395,
      "grad_norm": 0.051510479979236246,
      "learning_rate": 4.4542198524525596e-05,
      "loss": 0.0662675142288208,
      "step": 11840
    },
    {
      "epoch": 0.5497740178467957,
      "grad_norm": 0.05229633938398748,
      "learning_rate": 4.439298462129074e-05,
      "loss": 0.06687729358673096,
      "step": 11860
    },
    {
      "epoch": 0.5507011241163519,
      "grad_norm": 0.05080546222514308,
      "learning_rate": 4.424382127260301e-05,
      "loss": 0.06698527336120605,
      "step": 11880
    },
    {
      "epoch": 0.551628230385908,
      "grad_norm": 0.04776755277858745,
      "learning_rate": 4.409470982336431e-05,
      "loss": 0.06411564946174622,
      "step": 11900
    },
    {
      "epoch": 0.5525553366554641,
      "grad_norm": 0.06578837368317896,
      "learning_rate": 4.394565161800856e-05,
      "loss": 0.06554464101791382,
      "step": 11920
    },
    {
      "epoch": 0.5534824429250202,
      "grad_norm": 0.057995931106989626,
      "learning_rate": 4.379664800048964e-05,
      "loss": 0.06582286357879638,
      "step": 11940
    },
    {
      "epoch": 0.5544095491945764,
      "grad_norm": 0.05440657271599997,
      "learning_rate": 4.3647700314269266e-05,
      "loss": 0.06445524096488953,
      "step": 11960
    },
    {
      "epoch": 0.5553366554641326,
      "grad_norm": 0.05505789201059762,
      "learning_rate": 4.3498809902304824e-05,
      "loss": 0.0664166808128357,
      "step": 11980
    },
    {
      "epoch": 0.5562637617336887,
      "grad_norm": 0.0531897586265914,
      "learning_rate": 4.334997810703734e-05,
      "loss": 0.06556921005249024,
      "step": 12000
    },
    {
      "epoch": 0.5571908680032449,
      "grad_norm": 0.05973797112116463,
      "learning_rate": 4.3201206270379285e-05,
      "loss": 0.06338551640510559,
      "step": 12020
    },
    {
      "epoch": 0.558117974272801,
      "grad_norm": 0.048839628696251663,
      "learning_rate": 4.3052495733702576e-05,
      "loss": 0.06362003087997437,
      "step": 12040
    },
    {
      "epoch": 0.5590450805423571,
      "grad_norm": 0.0506996390654391,
      "learning_rate": 4.2903847837826386e-05,
      "loss": 0.06662671566009522,
      "step": 12060
    },
    {
      "epoch": 0.5599721868119133,
      "grad_norm": 0.05293381280714898,
      "learning_rate": 4.275526392300513e-05,
      "loss": 0.06514765620231629,
      "step": 12080
    },
    {
      "epoch": 0.5608992930814695,
      "grad_norm": 0.06574430841643815,
      "learning_rate": 4.2606745328916334e-05,
      "loss": 0.06714925765991211,
      "step": 12100
    },
    {
      "epoch": 0.5618263993510256,
      "grad_norm": 0.05623303888087702,
      "learning_rate": 4.245829339464859e-05,
      "loss": 0.06525035500526429,
      "step": 12120
    },
    {
      "epoch": 0.5627535056205818,
      "grad_norm": 0.05368953397440512,
      "learning_rate": 4.230990945868943e-05,
      "loss": 0.06534684896469116,
      "step": 12140
    },
    {
      "epoch": 0.563680611890138,
      "grad_norm": 0.06922013058545824,
      "learning_rate": 4.216159485891335e-05,
      "loss": 0.06529357433319091,
      "step": 12160
    },
    {
      "epoch": 0.564607718159694,
      "grad_norm": 0.052271901882029945,
      "learning_rate": 4.201335093256962e-05,
      "loss": 0.06600250005722046,
      "step": 12180
    },
    {
      "epoch": 0.5655348244292502,
      "grad_norm": 0.04995854230025834,
      "learning_rate": 4.1865179016270354e-05,
      "loss": 0.06561869382858276,
      "step": 12200
    },
    {
      "epoch": 0.5664619306988063,
      "grad_norm": 0.04808531970975858,
      "learning_rate": 4.17170804459784e-05,
      "loss": 0.06454001665115357,
      "step": 12220
    },
    {
      "epoch": 0.5673890369683625,
      "grad_norm": 0.05366099437836096,
      "learning_rate": 4.1569056556995233e-05,
      "loss": 0.06399540305137634,
      "step": 12240
    },
    {
      "epoch": 0.5683161432379187,
      "grad_norm": 0.0516715787791668,
      "learning_rate": 4.142110868394906e-05,
      "loss": 0.06460065841674804,
      "step": 12260
    },
    {
      "epoch": 0.5692432495074748,
      "grad_norm": 0.05532438685908628,
      "learning_rate": 4.127323816078264e-05,
      "loss": 0.06570006012916565,
      "step": 12280
    },
    {
      "epoch": 0.5701703557770309,
      "grad_norm": 0.06187647095341613,
      "learning_rate": 4.112544632074138e-05,
      "loss": 0.06358541250228882,
      "step": 12300
    },
    {
      "epoch": 0.5710974620465871,
      "grad_norm": 0.05361618506760777,
      "learning_rate": 4.09777344963612e-05,
      "loss": 0.06562268733978271,
      "step": 12320
    },
    {
      "epoch": 0.5720245683161432,
      "grad_norm": 0.04795690962775267,
      "learning_rate": 4.0830104019456616e-05,
      "loss": 0.0638882040977478,
      "step": 12340
    },
    {
      "epoch": 0.5729516745856994,
      "grad_norm": 0.050681893936996174,
      "learning_rate": 4.0682556221108656e-05,
      "loss": 0.06368978023529052,
      "step": 12360
    },
    {
      "epoch": 0.5738787808552556,
      "grad_norm": 0.05533054251174027,
      "learning_rate": 4.0535092431652946e-05,
      "loss": 0.06392117738723754,
      "step": 12380
    },
    {
      "epoch": 0.5748058871248117,
      "grad_norm": 0.05367648995587571,
      "learning_rate": 4.038771398066759e-05,
      "loss": 0.06591871380805969,
      "step": 12400
    },
    {
      "epoch": 0.5757329933943678,
      "grad_norm": 0.049861208100661886,
      "learning_rate": 4.0240422196961325e-05,
      "loss": 0.06433215737342834,
      "step": 12420
    },
    {
      "epoch": 0.576660099663924,
      "grad_norm": 0.06101855801657709,
      "learning_rate": 4.0093218408561416e-05,
      "loss": 0.06634079813957214,
      "step": 12440
    },
    {
      "epoch": 0.5775872059334801,
      "grad_norm": 0.052059434666904415,
      "learning_rate": 3.9946103942701777e-05,
      "loss": 0.06267635822296143,
      "step": 12460
    },
    {
      "epoch": 0.5785143122030363,
      "grad_norm": 0.05065802177679029,
      "learning_rate": 3.979908012581093e-05,
      "loss": 0.062290889024734494,
      "step": 12480
    },
    {
      "epoch": 0.5794414184725925,
      "grad_norm": 0.05241776667351534,
      "learning_rate": 3.9652148283500095e-05,
      "loss": 0.06085524559020996,
      "step": 12500
    },
    {
      "epoch": 0.5803685247421486,
      "grad_norm": 0.056135438882992786,
      "learning_rate": 3.9505309740551216e-05,
      "loss": 0.06510010957717896,
      "step": 12520
    },
    {
      "epoch": 0.5812956310117047,
      "grad_norm": 0.057472575894767654,
      "learning_rate": 3.9358565820905034e-05,
      "loss": 0.06703221201896667,
      "step": 12540
    },
    {
      "epoch": 0.5822227372812608,
      "grad_norm": 0.04645668025881864,
      "learning_rate": 3.921191784764911e-05,
      "loss": 0.065163254737854,
      "step": 12560
    },
    {
      "epoch": 0.583149843550817,
      "grad_norm": 0.06099767631099625,
      "learning_rate": 3.906536714300595e-05,
      "loss": 0.0646634042263031,
      "step": 12580
    },
    {
      "epoch": 0.5840769498203732,
      "grad_norm": 0.05286615510450817,
      "learning_rate": 3.8918915028321074e-05,
      "loss": 0.06526516675949097,
      "step": 12600
    },
    {
      "epoch": 0.5850040560899293,
      "grad_norm": 0.05282362636127958,
      "learning_rate": 3.877256282405102e-05,
      "loss": 0.06202825903892517,
      "step": 12620
    },
    {
      "epoch": 0.5859311623594855,
      "grad_norm": 0.06182156159568968,
      "learning_rate": 3.8626311849751574e-05,
      "loss": 0.06597340106964111,
      "step": 12640
    },
    {
      "epoch": 0.5868582686290416,
      "grad_norm": 0.04633738134367921,
      "learning_rate": 3.8480163424065734e-05,
      "loss": 0.06565766930580139,
      "step": 12660
    },
    {
      "epoch": 0.5877853748985977,
      "grad_norm": 0.05106255106308873,
      "learning_rate": 3.833411886471197e-05,
      "loss": 0.06096769571304321,
      "step": 12680
    },
    {
      "epoch": 0.5887124811681539,
      "grad_norm": 0.05246893915986374,
      "learning_rate": 3.818817948847217e-05,
      "loss": 0.06392142176628113,
      "step": 12700
    },
    {
      "epoch": 0.58963958743771,
      "grad_norm": 0.04546892705130382,
      "learning_rate": 3.804234661117995e-05,
      "loss": 0.06440478563308716,
      "step": 12720
    },
    {
      "epoch": 0.5905666937072662,
      "grad_norm": 0.04462132045960336,
      "learning_rate": 3.789662154770862e-05,
      "loss": 0.062234115600585935,
      "step": 12740
    },
    {
      "epoch": 0.5914937999768224,
      "grad_norm": 0.04420493079742613,
      "learning_rate": 3.775100561195946e-05,
      "loss": 0.06604772210121154,
      "step": 12760
    },
    {
      "epoch": 0.5924209062463784,
      "grad_norm": 0.05007519064408673,
      "learning_rate": 3.76055001168498e-05,
      "loss": 0.06394462585449219,
      "step": 12780
    },
    {
      "epoch": 0.5933480125159346,
      "grad_norm": 0.05063377838479155,
      "learning_rate": 3.7460106374301224e-05,
      "loss": 0.06103905439376831,
      "step": 12800
    },
    {
      "epoch": 0.5942751187854908,
      "grad_norm": 0.04534595551803043,
      "learning_rate": 3.731482569522768e-05,
      "loss": 0.06254583597183228,
      "step": 12820
    },
    {
      "epoch": 0.5952022250550469,
      "grad_norm": 0.05239922337744047,
      "learning_rate": 3.716965938952375e-05,
      "loss": 0.06174516677856445,
      "step": 12840
    },
    {
      "epoch": 0.5961293313246031,
      "grad_norm": 0.04852296362470289,
      "learning_rate": 3.7024608766052755e-05,
      "loss": 0.06775041818618774,
      "step": 12860
    },
    {
      "epoch": 0.5970564375941593,
      "grad_norm": 0.08132327786867243,
      "learning_rate": 3.687967513263501e-05,
      "loss": 0.0644269585609436,
      "step": 12880
    },
    {
      "epoch": 0.5979835438637153,
      "grad_norm": 0.05046642633348292,
      "learning_rate": 3.6734859796035995e-05,
      "loss": 0.06376562118530274,
      "step": 12900
    },
    {
      "epoch": 0.5989106501332715,
      "grad_norm": 0.045854939265676777,
      "learning_rate": 3.6590164061954615e-05,
      "loss": 0.060318171977996826,
      "step": 12920
    },
    {
      "epoch": 0.5998377564028277,
      "grad_norm": 0.053627580928548156,
      "learning_rate": 3.6445589235011363e-05,
      "loss": 0.06378052234649659,
      "step": 12940
    },
    {
      "epoch": 0.6007648626723838,
      "grad_norm": 0.051461469726852206,
      "learning_rate": 3.630113661873665e-05,
      "loss": 0.06393607258796692,
      "step": 12960
    },
    {
      "epoch": 0.60169196894194,
      "grad_norm": 0.049529344120311146,
      "learning_rate": 3.615680751555897e-05,
      "loss": 0.06163822412490845,
      "step": 12980
    },
    {
      "epoch": 0.6026190752114962,
      "grad_norm": 0.05993860914074786,
      "learning_rate": 3.601260322679317e-05,
      "loss": 0.06363263130187988,
      "step": 13000
    },
    {
      "epoch": 0.6035461814810522,
      "grad_norm": 0.05937918197526256,
      "learning_rate": 3.586852505262876e-05,
      "loss": 0.06614599227905274,
      "step": 13020
    },
    {
      "epoch": 0.6044732877506084,
      "grad_norm": 0.05526258317544042,
      "learning_rate": 3.572457429211814e-05,
      "loss": 0.06187690496444702,
      "step": 13040
    },
    {
      "epoch": 0.6054003940201645,
      "grad_norm": 0.05115156580489509,
      "learning_rate": 3.558075224316493e-05,
      "loss": 0.06315644979476928,
      "step": 13060
    },
    {
      "epoch": 0.6063275002897207,
      "grad_norm": 0.04999591192932574,
      "learning_rate": 3.5437060202512226e-05,
      "loss": 0.05984244346618652,
      "step": 13080
    },
    {
      "epoch": 0.6072546065592769,
      "grad_norm": 0.050399539079609416,
      "learning_rate": 3.529349946573097e-05,
      "loss": 0.06891499161720276,
      "step": 13100
    },
    {
      "epoch": 0.608181712828833,
      "grad_norm": 0.04589559285845564,
      "learning_rate": 3.515007132720817e-05,
      "loss": 0.06319842338562012,
      "step": 13120
    },
    {
      "epoch": 0.6091088190983891,
      "grad_norm": 0.04951783695801608,
      "learning_rate": 3.500677708013534e-05,
      "loss": 0.0644094169139862,
      "step": 13140
    },
    {
      "epoch": 0.6100359253679453,
      "grad_norm": 0.048125613826557855,
      "learning_rate": 3.486361801649676e-05,
      "loss": 0.06219049692153931,
      "step": 13160
    },
    {
      "epoch": 0.6109630316375014,
      "grad_norm": 0.04908658056836134,
      "learning_rate": 3.472059542705786e-05,
      "loss": 0.06271125078201294,
      "step": 13180
    },
    {
      "epoch": 0.6118901379070576,
      "grad_norm": 0.05921919990863233,
      "learning_rate": 3.4577710601353575e-05,
      "loss": 0.06433905959129334,
      "step": 13200
    },
    {
      "epoch": 0.6128172441766138,
      "grad_norm": 0.056357403489593415,
      "learning_rate": 3.443496482767674e-05,
      "loss": 0.06682388782501221,
      "step": 13220
    },
    {
      "epoch": 0.6137443504461699,
      "grad_norm": 0.05108546158018032,
      "learning_rate": 3.4292359393066424e-05,
      "loss": 0.06270695924758911,
      "step": 13240
    },
    {
      "epoch": 0.6146714567157261,
      "grad_norm": 0.048764479208541565,
      "learning_rate": 3.4149895583296394e-05,
      "loss": 0.06609930992126464,
      "step": 13260
    },
    {
      "epoch": 0.6155985629852821,
      "grad_norm": 0.04965389164449825,
      "learning_rate": 3.400757468286344e-05,
      "loss": 0.062494289875030515,
      "step": 13280
    },
    {
      "epoch": 0.6165256692548383,
      "grad_norm": 0.05122916772727866,
      "learning_rate": 3.386539797497588e-05,
      "loss": 0.06150481700897217,
      "step": 13300
    },
    {
      "epoch": 0.6174527755243945,
      "grad_norm": 0.04752553749719427,
      "learning_rate": 3.3723366741541895e-05,
      "loss": 0.0653360664844513,
      "step": 13320
    },
    {
      "epoch": 0.6183798817939506,
      "grad_norm": 0.04854177362739221,
      "learning_rate": 3.35814822631581e-05,
      "loss": 0.06191425919532776,
      "step": 13340
    },
    {
      "epoch": 0.6193069880635068,
      "grad_norm": 0.050752488455509046,
      "learning_rate": 3.343974581909788e-05,
      "loss": 0.06433194875717163,
      "step": 13360
    },
    {
      "epoch": 0.620234094333063,
      "grad_norm": 0.05212398651811507,
      "learning_rate": 3.329815868729987e-05,
      "loss": 0.06522014141082763,
      "step": 13380
    },
    {
      "epoch": 0.621161200602619,
      "grad_norm": 0.04874961925186329,
      "learning_rate": 3.3156722144356526e-05,
      "loss": 0.06363330483436584,
      "step": 13400
    },
    {
      "epoch": 0.6220883068721752,
      "grad_norm": 0.052179552048181144,
      "learning_rate": 3.301543746550248e-05,
      "loss": 0.0650145947933197,
      "step": 13420
    },
    {
      "epoch": 0.6230154131417314,
      "grad_norm": 0.052992480052277274,
      "learning_rate": 3.287430592460315e-05,
      "loss": 0.0635890543460846,
      "step": 13440
    },
    {
      "epoch": 0.6239425194112875,
      "grad_norm": 0.13557269320402637,
      "learning_rate": 3.2733328794143224e-05,
      "loss": 0.06465160846710205,
      "step": 13460
    },
    {
      "epoch": 0.6248696256808437,
      "grad_norm": 0.054803340766929436,
      "learning_rate": 3.2592507345215165e-05,
      "loss": 0.06395940780639649,
      "step": 13480
    },
    {
      "epoch": 0.6257967319503999,
      "grad_norm": 0.05166632027345761,
      "learning_rate": 3.2451842847507734e-05,
      "loss": 0.06677281856536865,
      "step": 13500
    },
    {
      "epoch": 0.6267238382199559,
      "grad_norm": 0.05200591123673862,
      "learning_rate": 3.231133656929464e-05,
      "loss": 0.061854970455169675,
      "step": 13520
    },
    {
      "epoch": 0.6276509444895121,
      "grad_norm": 0.05647641815345,
      "learning_rate": 3.217098977742299e-05,
      "loss": 0.06253820061683654,
      "step": 13540
    },
    {
      "epoch": 0.6285780507590683,
      "grad_norm": 0.0610099528317983,
      "learning_rate": 3.2030803737301896e-05,
      "loss": 0.061298006772994997,
      "step": 13560
    },
    {
      "epoch": 0.6295051570286244,
      "grad_norm": 0.048569687029945054,
      "learning_rate": 3.189077971289115e-05,
      "loss": 0.06349403262138367,
      "step": 13580
    },
    {
      "epoch": 0.6304322632981806,
      "grad_norm": 0.04978410455569841,
      "learning_rate": 3.175091896668969e-05,
      "loss": 0.06262670755386353,
      "step": 13600
    },
    {
      "epoch": 0.6313593695677368,
      "grad_norm": 0.04921363211429435,
      "learning_rate": 3.161122275972435e-05,
      "loss": 0.06025437116622925,
      "step": 13620
    },
    {
      "epoch": 0.6322864758372928,
      "grad_norm": 0.046915791964208746,
      "learning_rate": 3.147169235153837e-05,
      "loss": 0.06226421594619751,
      "step": 13640
    },
    {
      "epoch": 0.633213582106849,
      "grad_norm": 0.046560361880960736,
      "learning_rate": 3.133232900018015e-05,
      "loss": 0.06082407236099243,
      "step": 13660
    },
    {
      "epoch": 0.6341406883764051,
      "grad_norm": 0.0508242536623233,
      "learning_rate": 3.1193133962191815e-05,
      "loss": 0.06218297481536865,
      "step": 13680
    },
    {
      "epoch": 0.6350677946459613,
      "grad_norm": 0.05083848962595208,
      "learning_rate": 3.105410849259795e-05,
      "loss": 0.06175011396408081,
      "step": 13700
    },
    {
      "epoch": 0.6359949009155175,
      "grad_norm": 0.04648707087031881,
      "learning_rate": 3.091525384489424e-05,
      "loss": 0.06474980115890502,
      "step": 13720
    },
    {
      "epoch": 0.6369220071850736,
      "grad_norm": 0.05147998581000941,
      "learning_rate": 3.077657127103624e-05,
      "loss": 0.06363303661346435,
      "step": 13740
    },
    {
      "epoch": 0.6378491134546297,
      "grad_norm": 0.05656118791091504,
      "learning_rate": 3.063806202142796e-05,
      "loss": 0.06301457285881043,
      "step": 13760
    },
    {
      "epoch": 0.6387762197241859,
      "grad_norm": 0.057005466313639265,
      "learning_rate": 3.049972734491071e-05,
      "loss": 0.05961810350418091,
      "step": 13780
    },
    {
      "epoch": 0.639703325993742,
      "grad_norm": 0.04810738011960062,
      "learning_rate": 3.03615684887518e-05,
      "loss": 0.06208142042160034,
      "step": 13800
    },
    {
      "epoch": 0.6406304322632982,
      "grad_norm": 0.05039613162597633,
      "learning_rate": 3.0223586698633268e-05,
      "loss": 0.06209646463394165,
      "step": 13820
    },
    {
      "epoch": 0.6415575385328544,
      "grad_norm": 0.055154310694890865,
      "learning_rate": 3.008578321864069e-05,
      "loss": 0.06454386711120605,
      "step": 13840
    },
    {
      "epoch": 0.6424846448024105,
      "grad_norm": 0.048540480596161,
      "learning_rate": 2.994815929125192e-05,
      "loss": 0.06207352876663208,
      "step": 13860
    },
    {
      "epoch": 0.6434117510719666,
      "grad_norm": 0.049077791621891874,
      "learning_rate": 2.9810716157325948e-05,
      "loss": 0.06354498863220215,
      "step": 13880
    },
    {
      "epoch": 0.6443388573415227,
      "grad_norm": 0.05000300570946731,
      "learning_rate": 2.9673455056091636e-05,
      "loss": 0.06338932514190673,
      "step": 13900
    },
    {
      "epoch": 0.6452659636110789,
      "grad_norm": 0.050672234437650114,
      "learning_rate": 2.953637722513662e-05,
      "loss": 0.06386594772338867,
      "step": 13920
    },
    {
      "epoch": 0.6461930698806351,
      "grad_norm": 0.04689645034094619,
      "learning_rate": 2.939948390039608e-05,
      "loss": 0.060531580448150636,
      "step": 13940
    },
    {
      "epoch": 0.6471201761501912,
      "grad_norm": 0.04826610298419959,
      "learning_rate": 2.9262776316141673e-05,
      "loss": 0.061343008279800416,
      "step": 13960
    },
    {
      "epoch": 0.6480472824197474,
      "grad_norm": 0.06042496774601238,
      "learning_rate": 2.912625570497032e-05,
      "loss": 0.06625347733497619,
      "step": 13980
    },
    {
      "epoch": 0.6489743886893035,
      "grad_norm": 0.04893986385362358,
      "learning_rate": 2.8989923297793164e-05,
      "loss": 0.06337761878967285,
      "step": 14000
    },
    {
      "epoch": 0.6499014949588596,
      "grad_norm": 0.05199338151979727,
      "learning_rate": 2.885378032382446e-05,
      "loss": 0.06317566037178039,
      "step": 14020
    },
    {
      "epoch": 0.6508286012284158,
      "grad_norm": 0.05040754237069569,
      "learning_rate": 2.8717828010570413e-05,
      "loss": 0.062115275859832765,
      "step": 14040
    },
    {
      "epoch": 0.651755707497972,
      "grad_norm": 0.05112238064631296,
      "learning_rate": 2.858206758381824e-05,
      "loss": 0.06314098834991455,
      "step": 14060
    },
    {
      "epoch": 0.6526828137675281,
      "grad_norm": 0.04656253072378448,
      "learning_rate": 2.844650026762501e-05,
      "loss": 0.0645134687423706,
      "step": 14080
    },
    {
      "epoch": 0.6536099200370843,
      "grad_norm": 0.0458909973168415,
      "learning_rate": 2.831112728430668e-05,
      "loss": 0.06152822971343994,
      "step": 14100
    },
    {
      "epoch": 0.6545370263066403,
      "grad_norm": 0.04821912686102437,
      "learning_rate": 2.8175949854427013e-05,
      "loss": 0.06609168052673339,
      "step": 14120
    },
    {
      "epoch": 0.6554641325761965,
      "grad_norm": 0.04409770616498031,
      "learning_rate": 2.8040969196786628e-05,
      "loss": 0.06383690834045411,
      "step": 14140
    },
    {
      "epoch": 0.6563912388457527,
      "grad_norm": 0.054486454541445774,
      "learning_rate": 2.7906186528411982e-05,
      "loss": 0.06307547092437744,
      "step": 14160
    },
    {
      "epoch": 0.6573183451153088,
      "grad_norm": 0.060521370867158114,
      "learning_rate": 2.777160306454437e-05,
      "loss": 0.06182036995887756,
      "step": 14180
    },
    {
      "epoch": 0.658245451384865,
      "grad_norm": 0.05216377464954535,
      "learning_rate": 2.7637220018629028e-05,
      "loss": 0.06543076634407044,
      "step": 14200
    },
    {
      "epoch": 0.6591725576544212,
      "grad_norm": 0.05111229153576258,
      "learning_rate": 2.750303860230416e-05,
      "loss": 0.060222935676574704,
      "step": 14220
    },
    {
      "epoch": 0.6600996639239772,
      "grad_norm": 0.04437292464889052,
      "learning_rate": 2.7369060025390035e-05,
      "loss": 0.06446950435638428,
      "step": 14240
    },
    {
      "epoch": 0.6610267701935334,
      "grad_norm": 0.04870905676212439,
      "learning_rate": 2.7235285495878006e-05,
      "loss": 0.0638662576675415,
      "step": 14260
    },
    {
      "epoch": 0.6619538764630896,
      "grad_norm": 0.050354345501078,
      "learning_rate": 2.7101716219919717e-05,
      "loss": 0.06206049919128418,
      "step": 14280
    },
    {
      "epoch": 0.6628809827326457,
      "grad_norm": 0.053044302804342615,
      "learning_rate": 2.6968353401816183e-05,
      "loss": 0.0631550133228302,
      "step": 14300
    },
    {
      "epoch": 0.6638080890022019,
      "grad_norm": 0.05127531709253182,
      "learning_rate": 2.6835198244006927e-05,
      "loss": 0.06432077884674073,
      "step": 14320
    },
    {
      "epoch": 0.6647351952717581,
      "grad_norm": 0.04857753198455333,
      "learning_rate": 2.6702251947059115e-05,
      "loss": 0.06265321373939514,
      "step": 14340
    },
    {
      "epoch": 0.6656623015413142,
      "grad_norm": 0.04362244940939983,
      "learning_rate": 2.65695157096568e-05,
      "loss": 0.06121989488601685,
      "step": 14360
    },
    {
      "epoch": 0.6665894078108703,
      "grad_norm": 0.0481585486989303,
      "learning_rate": 2.643699072859006e-05,
      "loss": 0.06126395463943481,
      "step": 14380
    },
    {
      "epoch": 0.6675165140804264,
      "grad_norm": 0.05685711931627772,
      "learning_rate": 2.630467819874424e-05,
      "loss": 0.06136850118637085,
      "step": 14400
    },
    {
      "epoch": 0.6684436203499826,
      "grad_norm": 0.04334010720904258,
      "learning_rate": 2.6172579313089103e-05,
      "loss": 0.06024292707443237,
      "step": 14420
    },
    {
      "epoch": 0.6693707266195388,
      "grad_norm": 0.05300639365689139,
      "learning_rate": 2.6040695262668192e-05,
      "loss": 0.06384731531143188,
      "step": 14440
    },
    {
      "epoch": 0.670297832889095,
      "grad_norm": 0.051447825698258565,
      "learning_rate": 2.5909027236588e-05,
      "loss": 0.06182212233543396,
      "step": 14460
    },
    {
      "epoch": 0.6712249391586511,
      "grad_norm": 0.04764092153633654,
      "learning_rate": 2.5777576422007298e-05,
      "loss": 0.062230199575424194,
      "step": 14480
    },
    {
      "epoch": 0.6721520454282072,
      "grad_norm": 0.051194208793662634,
      "learning_rate": 2.5646344004126405e-05,
      "loss": 0.06158599853515625,
      "step": 14500
    },
    {
      "epoch": 0.6730791516977633,
      "grad_norm": 0.0504512340499311,
      "learning_rate": 2.5515331166176498e-05,
      "loss": 0.06429290175437927,
      "step": 14520
    },
    {
      "epoch": 0.6740062579673195,
      "grad_norm": 0.04853397531643219,
      "learning_rate": 2.5384539089408987e-05,
      "loss": 0.060837453603744505,
      "step": 14540
    },
    {
      "epoch": 0.6749333642368757,
      "grad_norm": 0.045948191007637335,
      "learning_rate": 2.5253968953084774e-05,
      "loss": 0.05951639413833618,
      "step": 14560
    },
    {
      "epoch": 0.6758604705064318,
      "grad_norm": 0.046213105926445244,
      "learning_rate": 2.512362193446372e-05,
      "loss": 0.05889805555343628,
      "step": 14580
    },
    {
      "epoch": 0.676787576775988,
      "grad_norm": 0.05427422312194556,
      "learning_rate": 2.4993499208793995e-05,
      "loss": 0.0607657790184021,
      "step": 14600
    },
    {
      "epoch": 0.677714683045544,
      "grad_norm": 0.05137345064288375,
      "learning_rate": 2.4863601949301486e-05,
      "loss": 0.06040535569190979,
      "step": 14620
    },
    {
      "epoch": 0.6786417893151002,
      "grad_norm": 0.05494705513028713,
      "learning_rate": 2.473393132717914e-05,
      "loss": 0.06201455593109131,
      "step": 14640
    },
    {
      "epoch": 0.6795688955846564,
      "grad_norm": 0.05086985287739992,
      "learning_rate": 2.460448851157654e-05,
      "loss": 0.059794497489929196,
      "step": 14660
    },
    {
      "epoch": 0.6804960018542126,
      "grad_norm": 0.04786650898084123,
      "learning_rate": 2.4475274669589277e-05,
      "loss": 0.05950331687927246,
      "step": 14680
    },
    {
      "epoch": 0.6814231081237687,
      "grad_norm": 0.05400017332993512,
      "learning_rate": 2.434629096624847e-05,
      "loss": 0.06126256585121155,
      "step": 14700
    },
    {
      "epoch": 0.6823502143933249,
      "grad_norm": 0.0551004998205847,
      "learning_rate": 2.4217538564510167e-05,
      "loss": 0.06148295402526856,
      "step": 14720
    },
    {
      "epoch": 0.6832773206628809,
      "grad_norm": 0.05178638073090484,
      "learning_rate": 2.408901862524501e-05,
      "loss": 0.06360439658164978,
      "step": 14740
    },
    {
      "epoch": 0.6842044269324371,
      "grad_norm": 0.05667379578346301,
      "learning_rate": 2.396073230722764e-05,
      "loss": 0.06254050731658936,
      "step": 14760
    },
    {
      "epoch": 0.6851315332019933,
      "grad_norm": 0.05672686099374685,
      "learning_rate": 2.3832680767126348e-05,
      "loss": 0.06115987300872803,
      "step": 14780
    },
    {
      "epoch": 0.6860586394715494,
      "grad_norm": 0.049092277264012316,
      "learning_rate": 2.370486515949251e-05,
      "loss": 0.06217692494392395,
      "step": 14800
    },
    {
      "epoch": 0.6869857457411056,
      "grad_norm": 0.04624765560948606,
      "learning_rate": 2.3577286636750357e-05,
      "loss": 0.06284316182136536,
      "step": 14820
    },
    {
      "epoch": 0.6879128520106618,
      "grad_norm": 0.04779809736891506,
      "learning_rate": 2.3449946349186435e-05,
      "loss": 0.061065506935119626,
      "step": 14840
    },
    {
      "epoch": 0.6888399582802178,
      "grad_norm": 0.05498381182472391,
      "learning_rate": 2.3322845444939312e-05,
      "loss": 0.06122352480888367,
      "step": 14860
    },
    {
      "epoch": 0.689767064549774,
      "grad_norm": 0.05076724143090053,
      "learning_rate": 2.319598506998919e-05,
      "loss": 0.060102933645248414,
      "step": 14880
    },
    {
      "epoch": 0.6906941708193302,
      "grad_norm": 0.06833467953191609,
      "learning_rate": 2.30693663681476e-05,
      "loss": 0.06075221300125122,
      "step": 14900
    },
    {
      "epoch": 0.6916212770888863,
      "grad_norm": 0.05169383028973068,
      "learning_rate": 2.2942990481047083e-05,
      "loss": 0.06117861270904541,
      "step": 14920
    },
    {
      "epoch": 0.6925483833584425,
      "grad_norm": 0.05709565644765517,
      "learning_rate": 2.2816858548130836e-05,
      "loss": 0.06132649183273316,
      "step": 14940
    },
    {
      "epoch": 0.6934754896279987,
      "grad_norm": 0.05824467675898879,
      "learning_rate": 2.269097170664255e-05,
      "loss": 0.06097906827926636,
      "step": 14960
    },
    {
      "epoch": 0.6944025958975547,
      "grad_norm": 0.051323216147163155,
      "learning_rate": 2.2565331091616077e-05,
      "loss": 0.06300780177116394,
      "step": 14980
    },
    {
      "epoch": 0.6953297021671109,
      "grad_norm": 0.05384201770989914,
      "learning_rate": 2.243993783586523e-05,
      "loss": 0.06162100434303284,
      "step": 15000
    },
    {
      "epoch": 0.696256808436667,
      "grad_norm": 0.09122931328484767,
      "learning_rate": 2.2314793069973517e-05,
      "loss": 0.05884939432144165,
      "step": 15020
    },
    {
      "epoch": 0.6971839147062232,
      "grad_norm": 0.04867893723669023,
      "learning_rate": 2.2189897922284016e-05,
      "loss": 0.060624444484710695,
      "step": 15040
    },
    {
      "epoch": 0.6981110209757794,
      "grad_norm": 0.046539939838735486,
      "learning_rate": 2.2065253518889173e-05,
      "loss": 0.061492109298706056,
      "step": 15060
    },
    {
      "epoch": 0.6990381272453355,
      "grad_norm": 0.050454496552781815,
      "learning_rate": 2.1940860983620665e-05,
      "loss": 0.06003788113594055,
      "step": 15080
    },
    {
      "epoch": 0.6999652335148916,
      "grad_norm": 0.04300212432191378,
      "learning_rate": 2.1816721438039177e-05,
      "loss": 0.06057480573654175,
      "step": 15100
    },
    {
      "epoch": 0.7008923397844478,
      "grad_norm": 0.05768894936765751,
      "learning_rate": 2.169283600142445e-05,
      "loss": 0.06077735424041748,
      "step": 15120
    },
    {
      "epoch": 0.7018194460540039,
      "grad_norm": 0.06993838140242448,
      "learning_rate": 2.1569205790765047e-05,
      "loss": 0.06323583722114563,
      "step": 15140
    },
    {
      "epoch": 0.7027465523235601,
      "grad_norm": 0.04947149315118965,
      "learning_rate": 2.144583192074837e-05,
      "loss": 0.06037414073944092,
      "step": 15160
    },
    {
      "epoch": 0.7036736585931163,
      "grad_norm": 0.05039590728245431,
      "learning_rate": 2.132271550375054e-05,
      "loss": 0.06047593355178833,
      "step": 15180
    },
    {
      "epoch": 0.7046007648626724,
      "grad_norm": 0.053764200339381465,
      "learning_rate": 2.1199857649826427e-05,
      "loss": 0.06059522032737732,
      "step": 15200
    },
    {
      "epoch": 0.7055278711322285,
      "grad_norm": 0.04874180225681996,
      "learning_rate": 2.107725946669964e-05,
      "loss": 0.06181482076644897,
      "step": 15220
    },
    {
      "epoch": 0.7064549774017846,
      "grad_norm": 0.05338100057265876,
      "learning_rate": 2.0954922059752437e-05,
      "loss": 0.05934479236602783,
      "step": 15240
    },
    {
      "epoch": 0.7073820836713408,
      "grad_norm": 0.043917745464175616,
      "learning_rate": 2.0832846532015937e-05,
      "loss": 0.06192747950553894,
      "step": 15260
    },
    {
      "epoch": 0.708309189940897,
      "grad_norm": 0.05544018939158692,
      "learning_rate": 2.071103398416003e-05,
      "loss": 0.06174721717834473,
      "step": 15280
    },
    {
      "epoch": 0.7092362962104531,
      "grad_norm": 0.04761845202116307,
      "learning_rate": 2.058948551448352e-05,
      "loss": 0.05761594176292419,
      "step": 15300
    },
    {
      "epoch": 0.7101634024800093,
      "grad_norm": 0.04750360786029068,
      "learning_rate": 2.0468202218904136e-05,
      "loss": 0.058632105588912964,
      "step": 15320
    },
    {
      "epoch": 0.7110905087495654,
      "grad_norm": 0.05040042868784111,
      "learning_rate": 2.0347185190948802e-05,
      "loss": 0.06146861314773559,
      "step": 15340
    },
    {
      "epoch": 0.7120176150191215,
      "grad_norm": 0.04459556343548379,
      "learning_rate": 2.0226435521743665e-05,
      "loss": 0.05942659974098206,
      "step": 15360
    },
    {
      "epoch": 0.7129447212886777,
      "grad_norm": 0.052289058810451726,
      "learning_rate": 2.010595430000424e-05,
      "loss": 0.061128878593444826,
      "step": 15380
    },
    {
      "epoch": 0.7138718275582339,
      "grad_norm": 0.05179041004651462,
      "learning_rate": 1.9985742612025705e-05,
      "loss": 0.05865001678466797,
      "step": 15400
    },
    {
      "epoch": 0.71479893382779,
      "grad_norm": 0.037568986770586295,
      "learning_rate": 1.9865801541673e-05,
      "loss": 0.05941717624664307,
      "step": 15420
    },
    {
      "epoch": 0.7157260400973462,
      "grad_norm": 0.0459286011758259,
      "learning_rate": 1.9746132170371124e-05,
      "loss": 0.05735700130462647,
      "step": 15440
    },
    {
      "epoch": 0.7166531463669024,
      "grad_norm": 0.04894741729317627,
      "learning_rate": 1.9626735577095302e-05,
      "loss": 0.06146210432052612,
      "step": 15460
    },
    {
      "epoch": 0.7175802526364584,
      "grad_norm": 0.04943434722439843,
      "learning_rate": 1.9507612838361365e-05,
      "loss": 0.06059056520462036,
      "step": 15480
    },
    {
      "epoch": 0.7185073589060146,
      "grad_norm": 0.05344400661573999,
      "learning_rate": 1.9388765028215966e-05,
      "loss": 0.06110994815826416,
      "step": 15500
    },
    {
      "epoch": 0.7194344651755707,
      "grad_norm": 0.05519783113277079,
      "learning_rate": 1.9270193218226934e-05,
      "loss": 0.06149757504463196,
      "step": 15520
    },
    {
      "epoch": 0.7203615714451269,
      "grad_norm": 0.04644592771943375,
      "learning_rate": 1.915189847747355e-05,
      "loss": 0.061004418134689334,
      "step": 15540
    },
    {
      "epoch": 0.7212886777146831,
      "grad_norm": 0.044390207498056236,
      "learning_rate": 1.9033881872537006e-05,
      "loss": 0.058844202756881715,
      "step": 15560
    },
    {
      "epoch": 0.7222157839842392,
      "grad_norm": 0.05100601349549141,
      "learning_rate": 1.8916144467490725e-05,
      "loss": 0.060409164428710936,
      "step": 15580
    },
    {
      "epoch": 0.7231428902537953,
      "grad_norm": 0.053819637866503366,
      "learning_rate": 1.8798687323890784e-05,
      "loss": 0.06232894659042358,
      "step": 15600
    },
    {
      "epoch": 0.7240699965233515,
      "grad_norm": 0.04832078643778615,
      "learning_rate": 1.8681511500766296e-05,
      "loss": 0.06120274066925049,
      "step": 15620
    },
    {
      "epoch": 0.7249971027929076,
      "grad_norm": 0.0544387144984958,
      "learning_rate": 1.8564618054609933e-05,
      "loss": 0.057281243801116946,
      "step": 15640
    },
    {
      "epoch": 0.7259242090624638,
      "grad_norm": 0.05245934643791841,
      "learning_rate": 1.844800803936843e-05,
      "loss": 0.0594766914844513,
      "step": 15660
    },
    {
      "epoch": 0.72685131533202,
      "grad_norm": 0.07821888764355657,
      "learning_rate": 1.833168250643292e-05,
      "loss": 0.06403745412826538,
      "step": 15680
    },
    {
      "epoch": 0.7277784216015761,
      "grad_norm": 0.04919325477431574,
      "learning_rate": 1.8215642504629603e-05,
      "loss": 0.06194213628768921,
      "step": 15700
    },
    {
      "epoch": 0.7287055278711322,
      "grad_norm": 0.05204434383280467,
      "learning_rate": 1.809988908021026e-05,
      "loss": 0.061628663539886476,
      "step": 15720
    },
    {
      "epoch": 0.7296326341406884,
      "grad_norm": 0.04573790022786436,
      "learning_rate": 1.7984423276842795e-05,
      "loss": 0.05932833552360535,
      "step": 15740
    },
    {
      "epoch": 0.7305597404102445,
      "grad_norm": 0.05232458466977827,
      "learning_rate": 1.7869246135601803e-05,
      "loss": 0.06094588041305542,
      "step": 15760
    },
    {
      "epoch": 0.7314868466798007,
      "grad_norm": 0.049329576784068664,
      "learning_rate": 1.775435869495925e-05,
      "loss": 0.06023161411285401,
      "step": 15780
    },
    {
      "epoch": 0.7324139529493569,
      "grad_norm": 0.04924151989335634,
      "learning_rate": 1.7639761990775055e-05,
      "loss": 0.06049606204032898,
      "step": 15800
    },
    {
      "epoch": 0.733341059218913,
      "grad_norm": 0.04712241356202147,
      "learning_rate": 1.752545705628781e-05,
      "loss": 0.0589922308921814,
      "step": 15820
    },
    {
      "epoch": 0.7342681654884691,
      "grad_norm": 0.20355048699938086,
      "learning_rate": 1.7411444922105336e-05,
      "loss": 0.060230028629302976,
      "step": 15840
    },
    {
      "epoch": 0.7351952717580252,
      "grad_norm": 0.057929183576447586,
      "learning_rate": 1.7297726616195557e-05,
      "loss": 0.06148542165756225,
      "step": 15860
    },
    {
      "epoch": 0.7361223780275814,
      "grad_norm": 0.04128511447185095,
      "learning_rate": 1.718430316387713e-05,
      "loss": 0.061528468132019044,
      "step": 15880
    },
    {
      "epoch": 0.7370494842971376,
      "grad_norm": 0.047852780528877814,
      "learning_rate": 1.7071175587810228e-05,
      "loss": 0.05776445865631104,
      "step": 15900
    },
    {
      "epoch": 0.7379765905666937,
      "grad_norm": 0.04529351291686641,
      "learning_rate": 1.6958344907987278e-05,
      "loss": 0.0594045877456665,
      "step": 15920
    },
    {
      "epoch": 0.7389036968362499,
      "grad_norm": 0.04494651134582276,
      "learning_rate": 1.684581214172384e-05,
      "loss": 0.058308207988739015,
      "step": 15940
    },
    {
      "epoch": 0.739830803105806,
      "grad_norm": 0.04701400689251207,
      "learning_rate": 1.6733578303649377e-05,
      "loss": 0.06261276006698609,
      "step": 15960
    },
    {
      "epoch": 0.7407579093753621,
      "grad_norm": 0.05088963441186104,
      "learning_rate": 1.6621644405698157e-05,
      "loss": 0.059018731117248535,
      "step": 15980
    },
    {
      "epoch": 0.7416850156449183,
      "grad_norm": 0.047745921008014565,
      "learning_rate": 1.651001145710004e-05,
      "loss": 0.06264204978942871,
      "step": 16000
    },
    {
      "epoch": 0.7426121219144745,
      "grad_norm": 0.050973838025907645,
      "learning_rate": 1.639868046437146e-05,
      "loss": 0.06120291948318481,
      "step": 16020
    },
    {
      "epoch": 0.7435392281840306,
      "grad_norm": 0.04854467333800795,
      "learning_rate": 1.6287652431306377e-05,
      "loss": 0.05769929885864258,
      "step": 16040
    },
    {
      "epoch": 0.7444663344535868,
      "grad_norm": 0.05041492096280157,
      "learning_rate": 1.61769283589671e-05,
      "loss": 0.05729357004165649,
      "step": 16060
    },
    {
      "epoch": 0.7453934407231428,
      "grad_norm": 0.048528513756649745,
      "learning_rate": 1.606650924567536e-05,
      "loss": 0.05918492674827576,
      "step": 16080
    },
    {
      "epoch": 0.746320546992699,
      "grad_norm": 0.04891126836907767,
      "learning_rate": 1.5956396087003302e-05,
      "loss": 0.06132298707962036,
      "step": 16100
    },
    {
      "epoch": 0.7472476532622552,
      "grad_norm": 0.055623514522799365,
      "learning_rate": 1.58465898757645e-05,
      "loss": 0.05870202779769897,
      "step": 16120
    },
    {
      "epoch": 0.7481747595318113,
      "grad_norm": 0.04964629368796431,
      "learning_rate": 1.5737091602004938e-05,
      "loss": 0.05943009853363037,
      "step": 16140
    },
    {
      "epoch": 0.7491018658013675,
      "grad_norm": 0.0481465037386536,
      "learning_rate": 1.562790225299421e-05,
      "loss": 0.05991309881210327,
      "step": 16160
    },
    {
      "epoch": 0.7500289720709237,
      "grad_norm": 0.05020029078869949,
      "learning_rate": 1.551902281321651e-05,
      "loss": 0.05931590795516968,
      "step": 16180
    },
    {
      "epoch": 0.7509560783404797,
      "grad_norm": 0.0448463631327419,
      "learning_rate": 1.5410454264361827e-05,
      "loss": 0.05894432663917541,
      "step": 16200
    },
    {
      "epoch": 0.7518831846100359,
      "grad_norm": 0.04794231742239611,
      "learning_rate": 1.5302197585317028e-05,
      "loss": 0.06392773389816284,
      "step": 16220
    },
    {
      "epoch": 0.7528102908795921,
      "grad_norm": 0.04735838626799169,
      "learning_rate": 1.519425375215709e-05,
      "loss": 0.057221269607543944,
      "step": 16240
    },
    {
      "epoch": 0.7537373971491482,
      "grad_norm": 0.051523459713987925,
      "learning_rate": 1.5086623738136274e-05,
      "loss": 0.06156526207923889,
      "step": 16260
    },
    {
      "epoch": 0.7546645034187044,
      "grad_norm": 0.04339718118496949,
      "learning_rate": 1.4979308513679374e-05,
      "loss": 0.05997901558876038,
      "step": 16280
    },
    {
      "epoch": 0.7555916096882606,
      "grad_norm": 0.05137973736167374,
      "learning_rate": 1.4872309046372885e-05,
      "loss": 0.0629388928413391,
      "step": 16300
    },
    {
      "epoch": 0.7565187159578166,
      "grad_norm": 0.047672576023041854,
      "learning_rate": 1.4765626300956392e-05,
      "loss": 0.06029292345046997,
      "step": 16320
    },
    {
      "epoch": 0.7574458222273728,
      "grad_norm": 0.04937090540374253,
      "learning_rate": 1.465926123931381e-05,
      "loss": 0.062035870552062986,
      "step": 16340
    },
    {
      "epoch": 0.7583729284969289,
      "grad_norm": 0.044208624329598725,
      "learning_rate": 1.4553214820464734e-05,
      "loss": 0.0572911262512207,
      "step": 16360
    },
    {
      "epoch": 0.7593000347664851,
      "grad_norm": 0.0500859586785516,
      "learning_rate": 1.444748800055573e-05,
      "loss": 0.05843917727470398,
      "step": 16380
    },
    {
      "epoch": 0.7602271410360413,
      "grad_norm": 0.050562553754695816,
      "learning_rate": 1.4342081732851792e-05,
      "loss": 0.06119612455368042,
      "step": 16400
    },
    {
      "epoch": 0.7611542473055974,
      "grad_norm": 0.05476121156783501,
      "learning_rate": 1.4236996967727778e-05,
      "loss": 0.058240222930908206,
      "step": 16420
    },
    {
      "epoch": 0.7620813535751535,
      "grad_norm": 0.05050328759244789,
      "learning_rate": 1.4132234652659665e-05,
      "loss": 0.05758323669433594,
      "step": 16440
    },
    {
      "epoch": 0.7630084598447097,
      "grad_norm": 0.05087112515393608,
      "learning_rate": 1.402779573221621e-05,
      "loss": 0.0637636661529541,
      "step": 16460
    },
    {
      "epoch": 0.7639355661142658,
      "grad_norm": 0.04918400119773233,
      "learning_rate": 1.3923681148050305e-05,
      "loss": 0.06149212121963501,
      "step": 16480
    },
    {
      "epoch": 0.764862672383822,
      "grad_norm": 0.047211087387714804,
      "learning_rate": 1.3819891838890575e-05,
      "loss": 0.06012523174285889,
      "step": 16500
    },
    {
      "epoch": 0.7657897786533782,
      "grad_norm": 0.046080873669770935,
      "learning_rate": 1.3716428740532794e-05,
      "loss": 0.05955320596694946,
      "step": 16520
    },
    {
      "epoch": 0.7667168849229343,
      "grad_norm": 0.05165227256839973,
      "learning_rate": 1.3613292785831578e-05,
      "loss": 0.05964708924293518,
      "step": 16540
    },
    {
      "epoch": 0.7676439911924905,
      "grad_norm": 0.044172493938736235,
      "learning_rate": 1.3510484904691895e-05,
      "loss": 0.06006022691726685,
      "step": 16560
    },
    {
      "epoch": 0.7685710974620465,
      "grad_norm": 0.043338036596873686,
      "learning_rate": 1.3408006024060732e-05,
      "loss": 0.06014246344566345,
      "step": 16580
    },
    {
      "epoch": 0.7694982037316027,
      "grad_norm": 0.05036039272040791,
      "learning_rate": 1.3305857067918647e-05,
      "loss": 0.0594563364982605,
      "step": 16600
    },
    {
      "epoch": 0.7704253100011589,
      "grad_norm": 0.050417961468972684,
      "learning_rate": 1.3204038957271548e-05,
      "loss": 0.0606998085975647,
      "step": 16620
    },
    {
      "epoch": 0.771352416270715,
      "grad_norm": 0.04526808345623893,
      "learning_rate": 1.3102552610142332e-05,
      "loss": 0.05902277231216431,
      "step": 16640
    },
    {
      "epoch": 0.7722795225402712,
      "grad_norm": 0.048497961197404384,
      "learning_rate": 1.300139894156262e-05,
      "loss": 0.0569302499294281,
      "step": 16660
    },
    {
      "epoch": 0.7732066288098274,
      "grad_norm": 0.0437456719250575,
      "learning_rate": 1.2900578863564472e-05,
      "loss": 0.06099953651428223,
      "step": 16680
    },
    {
      "epoch": 0.7741337350793834,
      "grad_norm": 0.04406880844053459,
      "learning_rate": 1.2800093285172226e-05,
      "loss": 0.056679391860961915,
      "step": 16700
    },
    {
      "epoch": 0.7750608413489396,
      "grad_norm": 0.05232279535503188,
      "learning_rate": 1.2699943112394258e-05,
      "loss": 0.06059268116950989,
      "step": 16720
    },
    {
      "epoch": 0.7759879476184958,
      "grad_norm": 0.05278327725259602,
      "learning_rate": 1.2600129248214842e-05,
      "loss": 0.061165344715118405,
      "step": 16740
    },
    {
      "epoch": 0.7769150538880519,
      "grad_norm": 0.04439282701311901,
      "learning_rate": 1.2500652592585955e-05,
      "loss": 0.05630176067352295,
      "step": 16760
    },
    {
      "epoch": 0.7778421601576081,
      "grad_norm": 0.054539167148899746,
      "learning_rate": 1.2401514042419215e-05,
      "loss": 0.06083977222442627,
      "step": 16780
    },
    {
      "epoch": 0.7787692664271643,
      "grad_norm": 0.050225957508062406,
      "learning_rate": 1.2302714491577832e-05,
      "loss": 0.05797850489616394,
      "step": 16800
    },
    {
      "epoch": 0.7796963726967203,
      "grad_norm": 0.05861522684956706,
      "learning_rate": 1.2204254830868406e-05,
      "loss": 0.05906668901443481,
      "step": 16820
    },
    {
      "epoch": 0.7806234789662765,
      "grad_norm": 0.05498270871179051,
      "learning_rate": 1.2106135948033049e-05,
      "loss": 0.06047302484512329,
      "step": 16840
    },
    {
      "epoch": 0.7815505852358327,
      "grad_norm": 0.04342348370921852,
      "learning_rate": 1.2008358727741281e-05,
      "loss": 0.061077332496643065,
      "step": 16860
    },
    {
      "epoch": 0.7824776915053888,
      "grad_norm": 0.04444908192620191,
      "learning_rate": 1.1910924051582128e-05,
      "loss": 0.05731930732727051,
      "step": 16880
    },
    {
      "epoch": 0.783404797774945,
      "grad_norm": 0.04617583916924598,
      "learning_rate": 1.1813832798056085e-05,
      "loss": 0.057947248220443726,
      "step": 16900
    },
    {
      "epoch": 0.7843319040445011,
      "grad_norm": 0.04514635222499561,
      "learning_rate": 1.1717085842567282e-05,
      "loss": 0.05813829898834229,
      "step": 16920
    },
    {
      "epoch": 0.7852590103140572,
      "grad_norm": 0.05137263466281721,
      "learning_rate": 1.1620684057415537e-05,
      "loss": 0.05704503059387207,
      "step": 16940
    },
    {
      "epoch": 0.7861861165836134,
      "grad_norm": 0.04595090095037721,
      "learning_rate": 1.1524628311788526e-05,
      "loss": 0.0565543532371521,
      "step": 16960
    },
    {
      "epoch": 0.7871132228531695,
      "grad_norm": 0.04482003350516348,
      "learning_rate": 1.1428919471753896e-05,
      "loss": 0.05513795614242554,
      "step": 16980
    },
    {
      "epoch": 0.7880403291227257,
      "grad_norm": 0.048002917861008416,
      "learning_rate": 1.1333558400251516e-05,
      "loss": 0.060343867540359496,
      "step": 17000
    },
    {
      "epoch": 0.7889674353922819,
      "grad_norm": 0.05374305306919219,
      "learning_rate": 1.1238545957085667e-05,
      "loss": 0.05681766271591186,
      "step": 17020
    },
    {
      "epoch": 0.789894541661838,
      "grad_norm": 0.04496747980799546,
      "learning_rate": 1.1143882998917265e-05,
      "loss": 0.060854780673980716,
      "step": 17040
    },
    {
      "epoch": 0.7908216479313941,
      "grad_norm": 0.05273678294287813,
      "learning_rate": 1.1049570379256191e-05,
      "loss": 0.05980669260025025,
      "step": 17060
    },
    {
      "epoch": 0.7917487542009503,
      "grad_norm": 0.04813268776893703,
      "learning_rate": 1.0955608948453561e-05,
      "loss": 0.05949196219444275,
      "step": 17080
    },
    {
      "epoch": 0.7926758604705064,
      "grad_norm": 0.04990723247976874,
      "learning_rate": 1.086199955369408e-05,
      "loss": 0.057298320531845096,
      "step": 17100
    },
    {
      "epoch": 0.7936029667400626,
      "grad_norm": 0.05098141398354817,
      "learning_rate": 1.0768743038988344e-05,
      "loss": 0.059625649452209474,
      "step": 17120
    },
    {
      "epoch": 0.7945300730096188,
      "grad_norm": 0.0569575352303254,
      "learning_rate": 1.067584024516532e-05,
      "loss": 0.05985774397850037,
      "step": 17140
    },
    {
      "epoch": 0.7954571792791749,
      "grad_norm": 0.05211468013489324,
      "learning_rate": 1.0583292009864709e-05,
      "loss": 0.060726594924926755,
      "step": 17160
    },
    {
      "epoch": 0.796384285548731,
      "grad_norm": 0.05664402921078555,
      "learning_rate": 1.0491099167529394e-05,
      "loss": 0.061726498603820804,
      "step": 17180
    },
    {
      "epoch": 0.7973113918182871,
      "grad_norm": 0.042788865294761384,
      "learning_rate": 1.0399262549397948e-05,
      "loss": 0.05645383596420288,
      "step": 17200
    },
    {
      "epoch": 0.7982384980878433,
      "grad_norm": 0.0455888804451592,
      "learning_rate": 1.0307782983497094e-05,
      "loss": 0.06045401692390442,
      "step": 17220
    },
    {
      "epoch": 0.7991656043573995,
      "grad_norm": 0.04561254833756392,
      "learning_rate": 1.02166612946343e-05,
      "loss": 0.05789419412612915,
      "step": 17240
    },
    {
      "epoch": 0.8000927106269556,
      "grad_norm": 0.049103957674379375,
      "learning_rate": 1.0125898304390246e-05,
      "loss": 0.0576973557472229,
      "step": 17260
    },
    {
      "epoch": 0.8010198168965118,
      "grad_norm": 0.044611668154624716,
      "learning_rate": 1.0035494831111536e-05,
      "loss": 0.05838904976844787,
      "step": 17280
    },
    {
      "epoch": 0.8019469231660679,
      "grad_norm": 0.05503462630872144,
      "learning_rate": 9.94545168990324e-06,
      "loss": 0.056771910190582274,
      "step": 17300
    },
    {
      "epoch": 0.802874029435624,
      "grad_norm": 0.04703353051766481,
      "learning_rate": 9.855769692621563e-06,
      "loss": 0.05993037223815918,
      "step": 17320
    },
    {
      "epoch": 0.8038011357051802,
      "grad_norm": 0.047878001113748375,
      "learning_rate": 9.76644964786651e-06,
      "loss": 0.05929996967315674,
      "step": 17340
    },
    {
      "epoch": 0.8047282419747364,
      "grad_norm": 0.04544490427833938,
      "learning_rate": 9.677492360974644e-06,
      "loss": 0.0570427417755127,
      "step": 17360
    },
    {
      "epoch": 0.8056553482442925,
      "grad_norm": 0.051765804121816666,
      "learning_rate": 9.588898634011773e-06,
      "loss": 0.057377088069915774,
      "step": 17380
    },
    {
      "epoch": 0.8065824545138487,
      "grad_norm": 0.057829942821774304,
      "learning_rate": 9.500669265765766e-06,
      "loss": 0.06165332794189453,
      "step": 17400
    },
    {
      "epoch": 0.8075095607834047,
      "grad_norm": 0.048746413737398846,
      "learning_rate": 9.412805051739265e-06,
      "loss": 0.060013145208358765,
      "step": 17420
    },
    {
      "epoch": 0.8084366670529609,
      "grad_norm": 0.05261160749625894,
      "learning_rate": 9.32530678414263e-06,
      "loss": 0.060103428363800046,
      "step": 17440
    },
    {
      "epoch": 0.8093637733225171,
      "grad_norm": 0.04420823726137285,
      "learning_rate": 9.23817525188671e-06,
      "loss": 0.06115436553955078,
      "step": 17460
    },
    {
      "epoch": 0.8102908795920732,
      "grad_norm": 0.04432143471528275,
      "learning_rate": 9.151411240575775e-06,
      "loss": 0.05716307163238525,
      "step": 17480
    },
    {
      "epoch": 0.8112179858616294,
      "grad_norm": 0.04649324107606914,
      "learning_rate": 9.065015532500388e-06,
      "loss": 0.058395767211914064,
      "step": 17500
    },
    {
      "epoch": 0.8121450921311856,
      "grad_norm": 0.04713335801422334,
      "learning_rate": 8.978988906630393e-06,
      "loss": 0.05896450281143188,
      "step": 17520
    },
    {
      "epoch": 0.8130721984007416,
      "grad_norm": 0.04327655995492146,
      "learning_rate": 8.89333213860788e-06,
      "loss": 0.059333956241607665,
      "step": 17540
    },
    {
      "epoch": 0.8139993046702978,
      "grad_norm": 0.050813175996178885,
      "learning_rate": 8.808046000740188e-06,
      "loss": 0.05825115442276001,
      "step": 17560
    },
    {
      "epoch": 0.814926410939854,
      "grad_norm": 0.05138510271301179,
      "learning_rate": 8.723131261992945e-06,
      "loss": 0.06070665121078491,
      "step": 17580
    },
    {
      "epoch": 0.8158535172094101,
      "grad_norm": 0.05141789618116487,
      "learning_rate": 8.638588687983117e-06,
      "loss": 0.05743167400360107,
      "step": 17600
    },
    {
      "epoch": 0.8167806234789663,
      "grad_norm": 0.05339506959936526,
      "learning_rate": 8.554419040972155e-06,
      "loss": 0.05785367488861084,
      "step": 17620
    },
    {
      "epoch": 0.8177077297485225,
      "grad_norm": 0.04757403496600389,
      "learning_rate": 8.47062307985903e-06,
      "loss": 0.06386864185333252,
      "step": 17640
    },
    {
      "epoch": 0.8186348360180786,
      "grad_norm": 0.04616531673760392,
      "learning_rate": 8.387201560173502e-06,
      "loss": 0.056666576862335206,
      "step": 17660
    },
    {
      "epoch": 0.8195619422876347,
      "grad_norm": 0.054320631298186886,
      "learning_rate": 8.304155234069221e-06,
      "loss": 0.06218389272689819,
      "step": 17680
    },
    {
      "epoch": 0.8204890485571908,
      "grad_norm": 0.04841570285281037,
      "learning_rate": 8.221484850317013e-06,
      "loss": 0.058232420682907106,
      "step": 17700
    },
    {
      "epoch": 0.821416154826747,
      "grad_norm": 0.051326390254033495,
      "learning_rate": 8.139191154298038e-06,
      "loss": 0.05738552212715149,
      "step": 17720
    },
    {
      "epoch": 0.8223432610963032,
      "grad_norm": 0.055533046268653616,
      "learning_rate": 8.057274887997174e-06,
      "loss": 0.05868532657623291,
      "step": 17740
    },
    {
      "epoch": 0.8232703673658593,
      "grad_norm": 0.046616486618029494,
      "learning_rate": 7.975736789996263e-06,
      "loss": 0.05714106559753418,
      "step": 17760
    },
    {
      "epoch": 0.8241974736354155,
      "grad_norm": 0.047464130181968045,
      "learning_rate": 7.894577595467483e-06,
      "loss": 0.05900055170059204,
      "step": 17780
    },
    {
      "epoch": 0.8251245799049716,
      "grad_norm": 0.051955630579307584,
      "learning_rate": 7.813798036166658e-06,
      "loss": 0.05795750021934509,
      "step": 17800
    },
    {
      "epoch": 0.8260516861745277,
      "grad_norm": 0.04596839313793552,
      "learning_rate": 7.733398840426748e-06,
      "loss": 0.05429661273956299,
      "step": 17820
    },
    {
      "epoch": 0.8269787924440839,
      "grad_norm": 0.04460169394638108,
      "learning_rate": 7.653380733151228e-06,
      "loss": 0.056835669279098514,
      "step": 17840
    },
    {
      "epoch": 0.8279058987136401,
      "grad_norm": 0.05345012477480762,
      "learning_rate": 7.573744435807584e-06,
      "loss": 0.06000061631202698,
      "step": 17860
    },
    {
      "epoch": 0.8288330049831962,
      "grad_norm": 0.048363666072192196,
      "learning_rate": 7.494490666420728e-06,
      "loss": 0.05979467630386352,
      "step": 17880
    },
    {
      "epoch": 0.8297601112527524,
      "grad_norm": 0.050171340169734284,
      "learning_rate": 7.415620139566653e-06,
      "loss": 0.0598601222038269,
      "step": 17900
    },
    {
      "epoch": 0.8306872175223085,
      "grad_norm": 0.054460879619936564,
      "learning_rate": 7.337133566365878e-06,
      "loss": 0.05797959566116333,
      "step": 17920
    },
    {
      "epoch": 0.8316143237918646,
      "grad_norm": 0.051711281647288644,
      "learning_rate": 7.259031654477105e-06,
      "loss": 0.05768338441848755,
      "step": 17940
    },
    {
      "epoch": 0.8325414300614208,
      "grad_norm": 0.052805751180748095,
      "learning_rate": 7.181315108090797e-06,
      "loss": 0.06080697774887085,
      "step": 17960
    },
    {
      "epoch": 0.833468536330977,
      "grad_norm": 0.055014917652565054,
      "learning_rate": 7.1039846279228495e-06,
      "loss": 0.05583765506744385,
      "step": 17980
    },
    {
      "epoch": 0.8343956426005331,
      "grad_norm": 0.0463304459584256,
      "learning_rate": 7.0270409112082825e-06,
      "loss": 0.05950177311897278,
      "step": 18000
    },
    {
      "epoch": 0.8353227488700893,
      "grad_norm": 0.049646411995576095,
      "learning_rate": 6.950484651694905e-06,
      "loss": 0.06469150781631469,
      "step": 18020
    },
    {
      "epoch": 0.8362498551396453,
      "grad_norm": 0.04646000526108706,
      "learning_rate": 6.874316539637127e-06,
      "loss": 0.05806264877319336,
      "step": 18040
    },
    {
      "epoch": 0.8371769614092015,
      "grad_norm": 0.056761905056810044,
      "learning_rate": 6.798537261789689e-06,
      "loss": 0.05710660219192505,
      "step": 18060
    },
    {
      "epoch": 0.8381040676787577,
      "grad_norm": 0.050417686744321163,
      "learning_rate": 6.7231475014015e-06,
      "loss": 0.059623163938522336,
      "step": 18080
    },
    {
      "epoch": 0.8390311739483138,
      "grad_norm": 0.046556597348368485,
      "learning_rate": 6.648147938209431e-06,
      "loss": 0.05778909921646118,
      "step": 18100
    },
    {
      "epoch": 0.83995828021787,
      "grad_norm": 0.05245396899535419,
      "learning_rate": 6.573539248432248e-06,
      "loss": 0.0584197998046875,
      "step": 18120
    },
    {
      "epoch": 0.8408853864874262,
      "grad_norm": 0.05401788882918738,
      "learning_rate": 6.499322104764466e-06,
      "loss": 0.05711585283279419,
      "step": 18140
    },
    {
      "epoch": 0.8418124927569822,
      "grad_norm": 0.04672129129062841,
      "learning_rate": 6.425497176370321e-06,
      "loss": 0.05773715972900391,
      "step": 18160
    },
    {
      "epoch": 0.8427395990265384,
      "grad_norm": 0.05060791421577387,
      "learning_rate": 6.352065128877688e-06,
      "loss": 0.05938397645950318,
      "step": 18180
    },
    {
      "epoch": 0.8436667052960946,
      "grad_norm": 0.0446610543336781,
      "learning_rate": 6.279026624372125e-06,
      "loss": 0.058554458618164065,
      "step": 18200
    },
    {
      "epoch": 0.8445938115656507,
      "grad_norm": 0.04521400085691856,
      "learning_rate": 6.2063823213908914e-06,
      "loss": 0.05861642360687256,
      "step": 18220
    },
    {
      "epoch": 0.8455209178352069,
      "grad_norm": 0.04859904209557068,
      "learning_rate": 6.13413287491702e-06,
      "loss": 0.05845460891723633,
      "step": 18240
    },
    {
      "epoch": 0.846448024104763,
      "grad_norm": 0.05534078747724061,
      "learning_rate": 6.062278936373355e-06,
      "loss": 0.05860517621040344,
      "step": 18260
    },
    {
      "epoch": 0.8473751303743191,
      "grad_norm": 0.04366307998596083,
      "learning_rate": 5.990821153616755e-06,
      "loss": 0.05628583431243896,
      "step": 18280
    },
    {
      "epoch": 0.8483022366438753,
      "grad_norm": 0.05063562654666035,
      "learning_rate": 5.9197601709322095e-06,
      "loss": 0.060969024896621704,
      "step": 18300
    },
    {
      "epoch": 0.8492293429134314,
      "grad_norm": 0.04868847439224681,
      "learning_rate": 5.84909662902704e-06,
      "loss": 0.056896209716796875,
      "step": 18320
    },
    {
      "epoch": 0.8501564491829876,
      "grad_norm": 0.049327726338164166,
      "learning_rate": 5.778831165025123e-06,
      "loss": 0.05948683023452759,
      "step": 18340
    },
    {
      "epoch": 0.8510835554525438,
      "grad_norm": 0.04948181638362051,
      "learning_rate": 5.708964412461137e-06,
      "loss": 0.05829896926879883,
      "step": 18360
    },
    {
      "epoch": 0.8520106617220999,
      "grad_norm": 0.05220738589957395,
      "learning_rate": 5.6394970012748795e-06,
      "loss": 0.05889915227890015,
      "step": 18380
    },
    {
      "epoch": 0.852937767991656,
      "grad_norm": 0.050899718409974336,
      "learning_rate": 5.570429557805518e-06,
      "loss": 0.056837928295135495,
      "step": 18400
    },
    {
      "epoch": 0.8538648742612122,
      "grad_norm": 0.05098101381150459,
      "learning_rate": 5.50176270478604e-06,
      "loss": 0.05628340840339661,
      "step": 18420
    },
    {
      "epoch": 0.8547919805307683,
      "grad_norm": 0.05251519133814445,
      "learning_rate": 5.433497061337561e-06,
      "loss": 0.05755155682563782,
      "step": 18440
    },
    {
      "epoch": 0.8557190868003245,
      "grad_norm": 0.051769712193200555,
      "learning_rate": 5.365633242963791e-06,
      "loss": 0.05773601531982422,
      "step": 18460
    },
    {
      "epoch": 0.8566461930698807,
      "grad_norm": 0.05004282331139003,
      "learning_rate": 5.2981718615454314e-06,
      "loss": 0.056486237049102786,
      "step": 18480
    },
    {
      "epoch": 0.8575732993394368,
      "grad_norm": 0.054530868941876996,
      "learning_rate": 5.231113525334713e-06,
      "loss": 0.05716860294342041,
      "step": 18500
    },
    {
      "epoch": 0.8585004056089929,
      "grad_norm": 0.049209144459847386,
      "learning_rate": 5.164458838949887e-06,
      "loss": 0.05820910930633545,
      "step": 18520
    },
    {
      "epoch": 0.859427511878549,
      "grad_norm": 0.046912699425193764,
      "learning_rate": 5.0982084033697865e-06,
      "loss": 0.060020279884338376,
      "step": 18540
    },
    {
      "epoch": 0.8603546181481052,
      "grad_norm": 0.06180815300166195,
      "learning_rate": 5.03236281592836e-06,
      "loss": 0.05633742809295654,
      "step": 18560
    },
    {
      "epoch": 0.8612817244176614,
      "grad_norm": 0.042067145894395426,
      "learning_rate": 4.966922670309354e-06,
      "loss": 0.05567646026611328,
      "step": 18580
    },
    {
      "epoch": 0.8622088306872175,
      "grad_norm": 0.04902954595417262,
      "learning_rate": 4.901888556540929e-06,
      "loss": 0.05492345094680786,
      "step": 18600
    },
    {
      "epoch": 0.8631359369567737,
      "grad_norm": 0.05775991144786959,
      "learning_rate": 4.837261060990328e-06,
      "loss": 0.05936078429222107,
      "step": 18620
    },
    {
      "epoch": 0.8640630432263298,
      "grad_norm": 0.05136868890329805,
      "learning_rate": 4.7730407663586e-06,
      "loss": 0.05857958197593689,
      "step": 18640
    },
    {
      "epoch": 0.8649901494958859,
      "grad_norm": 0.04594310399818881,
      "learning_rate": 4.709228251675357e-06,
      "loss": 0.05697960257530212,
      "step": 18660
    },
    {
      "epoch": 0.8659172557654421,
      "grad_norm": 0.04794046468006459,
      "learning_rate": 4.645824092293544e-06,
      "loss": 0.05718956589698791,
      "step": 18680
    },
    {
      "epoch": 0.8668443620349983,
      "grad_norm": 0.05176376086432388,
      "learning_rate": 4.5828288598842196e-06,
      "loss": 0.05958237648010254,
      "step": 18700
    },
    {
      "epoch": 0.8677714683045544,
      "grad_norm": 0.0538716086901053,
      "learning_rate": 4.520243122431489e-06,
      "loss": 0.05664359927177429,
      "step": 18720
    },
    {
      "epoch": 0.8686985745741106,
      "grad_norm": 0.06202857992382528,
      "learning_rate": 4.458067444227293e-06,
      "loss": 0.05787636041641235,
      "step": 18740
    },
    {
      "epoch": 0.8696256808436668,
      "grad_norm": 0.048148482241143764,
      "learning_rate": 4.396302385866352e-06,
      "loss": 0.05948035717010498,
      "step": 18760
    },
    {
      "epoch": 0.8705527871132228,
      "grad_norm": 0.044278455134471165,
      "learning_rate": 4.334948504241121e-06,
      "loss": 0.05834465026855469,
      "step": 18780
    },
    {
      "epoch": 0.871479893382779,
      "grad_norm": 0.044649748882328984,
      "learning_rate": 4.274006352536752e-06,
      "loss": 0.05538003444671631,
      "step": 18800
    },
    {
      "epoch": 0.8724069996523351,
      "grad_norm": 0.04436813297119467,
      "learning_rate": 4.2134764802261415e-06,
      "loss": 0.058388859033584595,
      "step": 18820
    },
    {
      "epoch": 0.8733341059218913,
      "grad_norm": 0.054175862334321694,
      "learning_rate": 4.153359433064919e-06,
      "loss": 0.05769250392913818,
      "step": 18840
    },
    {
      "epoch": 0.8742612121914475,
      "grad_norm": 0.052549228311279124,
      "learning_rate": 4.093655753086567e-06,
      "loss": 0.05964196920394897,
      "step": 18860
    },
    {
      "epoch": 0.8751883184610036,
      "grad_norm": 0.05094657747751493,
      "learning_rate": 4.034365978597543e-06,
      "loss": 0.05703572034835815,
      "step": 18880
    },
    {
      "epoch": 0.8761154247305597,
      "grad_norm": 0.050524925681744805,
      "learning_rate": 3.975490644172386e-06,
      "loss": 0.05787246823310852,
      "step": 18900
    },
    {
      "epoch": 0.8770425310001159,
      "grad_norm": 0.05039264346789014,
      "learning_rate": 3.91703028064892e-06,
      "loss": 0.05687538981437683,
      "step": 18920
    },
    {
      "epoch": 0.877969637269672,
      "grad_norm": 0.05363058526593667,
      "learning_rate": 3.858985415123473e-06,
      "loss": 0.0576495885848999,
      "step": 18940
    },
    {
      "epoch": 0.8788967435392282,
      "grad_norm": 0.054828145356224794,
      "learning_rate": 3.8013565709461184e-06,
      "loss": 0.05844680070877075,
      "step": 18960
    },
    {
      "epoch": 0.8798238498087844,
      "grad_norm": 0.056634076953621874,
      "learning_rate": 3.74414426771596e-06,
      "loss": 0.05621675848960876,
      "step": 18980
    },
    {
      "epoch": 0.8807509560783405,
      "grad_norm": 0.04850784142685619,
      "learning_rate": 3.6873490212764094e-06,
      "loss": 0.057649075984954834,
      "step": 19000
    },
    {
      "epoch": 0.8816780623478966,
      "grad_norm": 0.04878080327837706,
      "learning_rate": 3.6309713437106098e-06,
      "loss": 0.05968189835548401,
      "step": 19020
    },
    {
      "epoch": 0.8826051686174528,
      "grad_norm": 0.04967767894412179,
      "learning_rate": 3.575011743336759e-06,
      "loss": 0.05831393003463745,
      "step": 19040
    },
    {
      "epoch": 0.8835322748870089,
      "grad_norm": 0.047459417097776437,
      "learning_rate": 3.51947072470355e-06,
      "loss": 0.058126252889633176,
      "step": 19060
    },
    {
      "epoch": 0.8844593811565651,
      "grad_norm": 0.04447957512371265,
      "learning_rate": 3.4643487885855865e-06,
      "loss": 0.05986871719360352,
      "step": 19080
    },
    {
      "epoch": 0.8853864874261212,
      "grad_norm": 0.05332460849243031,
      "learning_rate": 3.4096464319789523e-06,
      "loss": 0.05962523818016052,
      "step": 19100
    },
    {
      "epoch": 0.8863135936956774,
      "grad_norm": 0.04839037998778892,
      "learning_rate": 3.35536414809664e-06,
      "loss": 0.05625807046890259,
      "step": 19120
    },
    {
      "epoch": 0.8872406999652335,
      "grad_norm": 0.048637654569926775,
      "learning_rate": 3.3015024263641337e-06,
      "loss": 0.05935103893280029,
      "step": 19140
    },
    {
      "epoch": 0.8881678062347896,
      "grad_norm": 0.050052326829702246,
      "learning_rate": 3.2480617524150157e-06,
      "loss": 0.05888447165489197,
      "step": 19160
    },
    {
      "epoch": 0.8890949125043458,
      "grad_norm": 0.059140081976628206,
      "learning_rate": 3.1950426080865794e-06,
      "loss": 0.0564881443977356,
      "step": 19180
    },
    {
      "epoch": 0.890022018773902,
      "grad_norm": 0.049210949556210644,
      "learning_rate": 3.142445471415467e-06,
      "loss": 0.057204604148864746,
      "step": 19200
    },
    {
      "epoch": 0.8909491250434581,
      "grad_norm": 0.05366723409442657,
      "learning_rate": 3.0902708166333726e-06,
      "loss": 0.06004486083984375,
      "step": 19220
    },
    {
      "epoch": 0.8918762313130143,
      "grad_norm": 0.044556516085901096,
      "learning_rate": 3.0385191141627746e-06,
      "loss": 0.0576244592666626,
      "step": 19240
    },
    {
      "epoch": 0.8928033375825704,
      "grad_norm": 0.04960996296862988,
      "learning_rate": 2.9871908306126796e-06,
      "loss": 0.05561840534210205,
      "step": 19260
    },
    {
      "epoch": 0.8937304438521265,
      "grad_norm": 0.053763824787078716,
      "learning_rate": 2.936286428774426e-06,
      "loss": 0.05704617500305176,
      "step": 19280
    },
    {
      "epoch": 0.8946575501216827,
      "grad_norm": 0.047984194446115826,
      "learning_rate": 2.885806367617494e-06,
      "loss": 0.05541048049926758,
      "step": 19300
    },
    {
      "epoch": 0.8955846563912389,
      "grad_norm": 0.05025681349535523,
      "learning_rate": 2.8357511022853945e-06,
      "loss": 0.060337722301483154,
      "step": 19320
    },
    {
      "epoch": 0.896511762660795,
      "grad_norm": 0.06024302457841739,
      "learning_rate": 2.786121084091542e-06,
      "loss": 0.05997539758682251,
      "step": 19340
    },
    {
      "epoch": 0.8974388689303512,
      "grad_norm": 0.04914718847360551,
      "learning_rate": 2.7369167605152068e-06,
      "loss": 0.05711795091629028,
      "step": 19360
    },
    {
      "epoch": 0.8983659751999072,
      "grad_norm": 0.04496466826210933,
      "learning_rate": 2.688138575197441e-06,
      "loss": 0.05466601848602295,
      "step": 19380
    },
    {
      "epoch": 0.8992930814694634,
      "grad_norm": 0.051759577304832854,
      "learning_rate": 2.63978696793713e-06,
      "loss": 0.05825486779212952,
      "step": 19400
    },
    {
      "epoch": 0.9002201877390196,
      "grad_norm": 0.046022224568802124,
      "learning_rate": 2.591862374686993e-06,
      "loss": 0.05798335075378418,
      "step": 19420
    },
    {
      "epoch": 0.9011472940085757,
      "grad_norm": 0.05362139777498272,
      "learning_rate": 2.544365227549672e-06,
      "loss": 0.06275264620780945,
      "step": 19440
    },
    {
      "epoch": 0.9020744002781319,
      "grad_norm": 0.04575597414143818,
      "learning_rate": 2.497295954773804e-06,
      "loss": 0.057760345935821536,
      "step": 19460
    },
    {
      "epoch": 0.9030015065476881,
      "grad_norm": 0.04722223458444896,
      "learning_rate": 2.450654980750189e-06,
      "loss": 0.05673818588256836,
      "step": 19480
    },
    {
      "epoch": 0.9039286128172441,
      "grad_norm": 0.04694719949871601,
      "learning_rate": 2.404442726007983e-06,
      "loss": 0.056690484285354614,
      "step": 19500
    },
    {
      "epoch": 0.9048557190868003,
      "grad_norm": 0.04811316270625525,
      "learning_rate": 2.3586596072108393e-06,
      "loss": 0.05581275224685669,
      "step": 19520
    },
    {
      "epoch": 0.9057828253563565,
      "grad_norm": 0.046451144979732226,
      "learning_rate": 2.313306037153218e-06,
      "loss": 0.059000533819198606,
      "step": 19540
    },
    {
      "epoch": 0.9067099316259126,
      "grad_norm": 0.05668763084521468,
      "learning_rate": 2.268382424756621e-06,
      "loss": 0.0562749981880188,
      "step": 19560
    },
    {
      "epoch": 0.9076370378954688,
      "grad_norm": 0.04793035932899476,
      "learning_rate": 2.223889175065941e-06,
      "loss": 0.05623648166656494,
      "step": 19580
    },
    {
      "epoch": 0.908564144165025,
      "grad_norm": 0.048360543569423806,
      "learning_rate": 2.179826689245762e-06,
      "loss": 0.05779556035995483,
      "step": 19600
    },
    {
      "epoch": 0.909491250434581,
      "grad_norm": 0.05170316788058635,
      "learning_rate": 2.136195364576804e-06,
      "loss": 0.059479916095733644,
      "step": 19620
    },
    {
      "epoch": 0.9104183567041372,
      "grad_norm": 0.05169560930869728,
      "learning_rate": 2.092995594452285e-06,
      "loss": 0.05806620121002197,
      "step": 19640
    },
    {
      "epoch": 0.9113454629736933,
      "grad_norm": 0.0511601774909589,
      "learning_rate": 2.0502277683744074e-06,
      "loss": 0.05798548460006714,
      "step": 19660
    },
    {
      "epoch": 0.9122725692432495,
      "grad_norm": 0.053785362511476247,
      "learning_rate": 2.007892271950823e-06,
      "loss": 0.058463799953460696,
      "step": 19680
    },
    {
      "epoch": 0.9131996755128057,
      "grad_norm": 0.05067300160085607,
      "learning_rate": 1.965989486891184e-06,
      "loss": 0.05823978185653687,
      "step": 19700
    },
    {
      "epoch": 0.9141267817823618,
      "grad_norm": 0.053445008750700594,
      "learning_rate": 1.9245197910036705e-06,
      "loss": 0.05742321014404297,
      "step": 19720
    },
    {
      "epoch": 0.9150538880519179,
      "grad_norm": 0.05337561730416442,
      "learning_rate": 1.8834835581916233e-06,
      "loss": 0.05843052864074707,
      "step": 19740
    },
    {
      "epoch": 0.9159809943214741,
      "grad_norm": 0.05532941015515067,
      "learning_rate": 1.842881158450116e-06,
      "loss": 0.05790799856185913,
      "step": 19760
    },
    {
      "epoch": 0.9169081005910302,
      "grad_norm": 0.04450320227512486,
      "learning_rate": 1.802712957862679e-06,
      "loss": 0.058239877223968506,
      "step": 19780
    },
    {
      "epoch": 0.9178352068605864,
      "grad_norm": 0.04692587222399893,
      "learning_rate": 1.762979318597957e-06,
      "loss": 0.05915763974189758,
      "step": 19800
    },
    {
      "epoch": 0.9187623131301426,
      "grad_norm": 0.524356361396534,
      "learning_rate": 1.7236805989064742e-06,
      "loss": 0.0603798508644104,
      "step": 19820
    },
    {
      "epoch": 0.9196894193996987,
      "grad_norm": 0.04990506653278282,
      "learning_rate": 1.6848171531173685e-06,
      "loss": 0.05706942081451416,
      "step": 19840
    },
    {
      "epoch": 0.9206165256692549,
      "grad_norm": 0.045037930646277306,
      "learning_rate": 1.6463893316352175e-06,
      "loss": 0.055731916427612306,
      "step": 19860
    },
    {
      "epoch": 0.921543631938811,
      "grad_norm": 0.04560174448532109,
      "learning_rate": 1.6083974809369073e-06,
      "loss": 0.0566458523273468,
      "step": 19880
    },
    {
      "epoch": 0.9224707382083671,
      "grad_norm": 0.04763326572516237,
      "learning_rate": 1.5708419435684462e-06,
      "loss": 0.05709692239761353,
      "step": 19900
    },
    {
      "epoch": 0.9233978444779233,
      "grad_norm": 0.048057534639948996,
      "learning_rate": 1.5337230581419227e-06,
      "loss": 0.057702839374542236,
      "step": 19920
    },
    {
      "epoch": 0.9243249507474794,
      "grad_norm": 0.0489870345901107,
      "learning_rate": 1.4970411593324352e-06,
      "loss": 0.058676493167877194,
      "step": 19940
    },
    {
      "epoch": 0.9252520570170356,
      "grad_norm": 0.050544582584891695,
      "learning_rate": 1.4607965778750954e-06,
      "loss": 0.05552069544792175,
      "step": 19960
    },
    {
      "epoch": 0.9261791632865918,
      "grad_norm": 0.06280076984578158,
      "learning_rate": 1.4249896405620079e-06,
      "loss": 0.057135379314422606,
      "step": 19980
    },
    {
      "epoch": 0.9271062695561478,
      "grad_norm": 0.04511659688250614,
      "learning_rate": 1.3896206702393444e-06,
      "loss": 0.056685060262680054,
      "step": 20000
    },
    {
      "epoch": 0.928033375825704,
      "grad_norm": 0.04924866814994787,
      "learning_rate": 1.3546899858044582e-06,
      "loss": 0.05704180002212524,
      "step": 20020
    },
    {
      "epoch": 0.9289604820952602,
      "grad_norm": 0.046740965807465625,
      "learning_rate": 1.3201979022029632e-06,
      "loss": 0.05864759683609009,
      "step": 20040
    },
    {
      "epoch": 0.9298875883648163,
      "grad_norm": 0.051059769253919474,
      "learning_rate": 1.2861447304259145e-06,
      "loss": 0.059252357482910155,
      "step": 20060
    },
    {
      "epoch": 0.9308146946343725,
      "grad_norm": 0.04787181284278493,
      "learning_rate": 1.2525307775070162e-06,
      "loss": 0.05766612887382507,
      "step": 20080
    },
    {
      "epoch": 0.9317418009039287,
      "grad_norm": 0.05614007141568827,
      "learning_rate": 1.21935634651984e-06,
      "loss": 0.05731832981109619,
      "step": 20100
    },
    {
      "epoch": 0.9326689071734847,
      "grad_norm": 0.05662767281777056,
      "learning_rate": 1.1866217365750999e-06,
      "loss": 0.05924726128578186,
      "step": 20120
    },
    {
      "epoch": 0.9335960134430409,
      "grad_norm": 0.049425289427901496,
      "learning_rate": 1.1543272428179264e-06,
      "loss": 0.061063098907470706,
      "step": 20140
    },
    {
      "epoch": 0.934523119712597,
      "grad_norm": 0.04915845295225296,
      "learning_rate": 1.1224731564252634e-06,
      "loss": 0.05853087306022644,
      "step": 20160
    },
    {
      "epoch": 0.9354502259821532,
      "grad_norm": 0.046166047765715514,
      "learning_rate": 1.091059764603175e-06,
      "loss": 0.060241210460662845,
      "step": 20180
    },
    {
      "epoch": 0.9363773322517094,
      "grad_norm": 0.05160898671044604,
      "learning_rate": 1.0600873505843268e-06,
      "loss": 0.0596448540687561,
      "step": 20200
    },
    {
      "epoch": 0.9373044385212655,
      "grad_norm": 0.057652325493545896,
      "learning_rate": 1.0295561936253584e-06,
      "loss": 0.055621218681335446,
      "step": 20220
    },
    {
      "epoch": 0.9382315447908216,
      "grad_norm": 0.04417766928484997,
      "learning_rate": 9.994665690044203e-07,
      "loss": 0.056035745143890384,
      "step": 20240
    },
    {
      "epoch": 0.9391586510603778,
      "grad_norm": 0.06414042802956324,
      "learning_rate": 9.698187480186805e-07,
      "loss": 0.058949506282806395,
      "step": 20260
    },
    {
      "epoch": 0.9400857573299339,
      "grad_norm": 0.051106447949205845,
      "learning_rate": 9.406129979818545e-07,
      "loss": 0.05908852815628052,
      "step": 20280
    },
    {
      "epoch": 0.9410128635994901,
      "grad_norm": 0.054143563522887926,
      "learning_rate": 9.118495822218299e-07,
      "loss": 0.057335054874420165,
      "step": 20300
    },
    {
      "epoch": 0.9419399698690463,
      "grad_norm": 0.04646566265813083,
      "learning_rate": 8.83528760078256e-07,
      "loss": 0.057302170991897584,
      "step": 20320
    },
    {
      "epoch": 0.9428670761386024,
      "grad_norm": 0.05205774343851159,
      "learning_rate": 8.556507869002473e-07,
      "loss": 0.05532693862915039,
      "step": 20340
    },
    {
      "epoch": 0.9437941824081585,
      "grad_norm": 0.0482434812896233,
      "learning_rate": 8.282159140440337e-07,
      "loss": 0.05476514101028442,
      "step": 20360
    },
    {
      "epoch": 0.9447212886777147,
      "grad_norm": 0.053510480368897835,
      "learning_rate": 8.012243888707305e-07,
      "loss": 0.05795966386795044,
      "step": 20380
    },
    {
      "epoch": 0.9456483949472708,
      "grad_norm": 0.049367763840699457,
      "learning_rate": 7.746764547440999e-07,
      "loss": 0.058704215288162234,
      "step": 20400
    },
    {
      "epoch": 0.946575501216827,
      "grad_norm": 0.051546708033184965,
      "learning_rate": 7.485723510283427e-07,
      "loss": 0.057125824689865115,
      "step": 20420
    },
    {
      "epoch": 0.9475026074863832,
      "grad_norm": 0.053427823749844935,
      "learning_rate": 7.229123130859605e-07,
      "loss": 0.056904202699661253,
      "step": 20440
    },
    {
      "epoch": 0.9484297137559393,
      "grad_norm": 0.07389253511825052,
      "learning_rate": 6.97696572275619e-07,
      "loss": 0.05820508599281311,
      "step": 20460
    },
    {
      "epoch": 0.9493568200254954,
      "grad_norm": 0.05099979982480766,
      "learning_rate": 6.729253559500715e-07,
      "loss": 0.059120839834213255,
      "step": 20480
    },
    {
      "epoch": 0.9502839262950515,
      "grad_norm": 0.0492619116493846,
      "learning_rate": 6.485988874540882e-07,
      "loss": 0.05862786173820496,
      "step": 20500
    },
    {
      "epoch": 0.9512110325646077,
      "grad_norm": 0.04843225967811867,
      "learning_rate": 6.247173861224753e-07,
      "loss": 0.05851049423217773,
      "step": 20520
    },
    {
      "epoch": 0.9521381388341639,
      "grad_norm": 0.05195184367081577,
      "learning_rate": 6.0128106727807e-07,
      "loss": 0.05784143209457397,
      "step": 20540
    },
    {
      "epoch": 0.95306524510372,
      "grad_norm": 0.046762958997749626,
      "learning_rate": 5.782901422298259e-07,
      "loss": 0.05785082578659058,
      "step": 20560
    },
    {
      "epoch": 0.9539923513732762,
      "grad_norm": 0.04471809218941821,
      "learning_rate": 5.557448182708702e-07,
      "loss": 0.05716483592987061,
      "step": 20580
    },
    {
      "epoch": 0.9549194576428323,
      "grad_norm": 0.04591262151763236,
      "learning_rate": 5.33645298676666e-07,
      "loss": 0.05687132477760315,
      "step": 20600
    },
    {
      "epoch": 0.9558465639123884,
      "grad_norm": 0.05055885971457279,
      "learning_rate": 5.11991782703164e-07,
      "loss": 0.05676789283752441,
      "step": 20620
    },
    {
      "epoch": 0.9567736701819446,
      "grad_norm": 0.049288967689513824,
      "learning_rate": 4.907844655850258e-07,
      "loss": 0.058823162317276,
      "step": 20640
    },
    {
      "epoch": 0.9577007764515008,
      "grad_norm": 0.05200183700196877,
      "learning_rate": 4.700235385338314e-07,
      "loss": 0.056910812854766846,
      "step": 20660
    },
    {
      "epoch": 0.9586278827210569,
      "grad_norm": 0.04574561427226024,
      "learning_rate": 4.497091887363858e-07,
      "loss": 0.05462713241577148,
      "step": 20680
    },
    {
      "epoch": 0.9595549889906131,
      "grad_norm": 0.04604087172975439,
      "learning_rate": 4.2984159935300915e-07,
      "loss": 0.05526896715164185,
      "step": 20700
    },
    {
      "epoch": 0.9604820952601691,
      "grad_norm": 0.05781409963329426,
      "learning_rate": 4.104209495158995e-07,
      "loss": 0.05807181000709534,
      "step": 20720
    },
    {
      "epoch": 0.9614092015297253,
      "grad_norm": 0.046091020708219466,
      "learning_rate": 3.9144741432751155e-07,
      "loss": 0.05884546637535095,
      "step": 20740
    },
    {
      "epoch": 0.9623363077992815,
      "grad_norm": 0.06370360981463187,
      "learning_rate": 3.7292116485898587e-07,
      "loss": 0.05833401679992676,
      "step": 20760
    },
    {
      "epoch": 0.9632634140688376,
      "grad_norm": 0.046962899311715066,
      "learning_rate": 3.5484236814858354e-07,
      "loss": 0.058104634284973145,
      "step": 20780
    },
    {
      "epoch": 0.9641905203383938,
      "grad_norm": 0.05311520253713704,
      "learning_rate": 3.3721118720020926e-07,
      "loss": 0.05689337253570557,
      "step": 20800
    },
    {
      "epoch": 0.96511762660795,
      "grad_norm": 0.05783239239469601,
      "learning_rate": 3.200277809819241e-07,
      "loss": 0.06119933128356934,
      "step": 20820
    },
    {
      "epoch": 0.966044732877506,
      "grad_norm": 0.05250832849633904,
      "learning_rate": 3.032923044245184e-07,
      "loss": 0.058338123559951785,
      "step": 20840
    },
    {
      "epoch": 0.9669718391470622,
      "grad_norm": 0.04611594589845792,
      "learning_rate": 2.8700490842011874e-07,
      "loss": 0.05823901891708374,
      "step": 20860
    },
    {
      "epoch": 0.9678989454166184,
      "grad_norm": 0.04228589152275358,
      "learning_rate": 2.711657398208001e-07,
      "loss": 0.058671307563781736,
      "step": 20880
    },
    {
      "epoch": 0.9688260516861745,
      "grad_norm": 0.052698185183678546,
      "learning_rate": 2.5577494143731453e-07,
      "loss": 0.05769298672676086,
      "step": 20900
    },
    {
      "epoch": 0.9697531579557307,
      "grad_norm": 0.04555971442292579,
      "learning_rate": 2.40832652037748e-07,
      "loss": 0.059823358058929445,
      "step": 20920
    },
    {
      "epoch": 0.9706802642252869,
      "grad_norm": 0.055117827828624504,
      "learning_rate": 2.2633900634630444e-07,
      "loss": 0.059781765937805174,
      "step": 20940
    },
    {
      "epoch": 0.971607370494843,
      "grad_norm": 0.056510040444828186,
      "learning_rate": 2.122941350420793e-07,
      "loss": 0.05672365427017212,
      "step": 20960
    },
    {
      "epoch": 0.9725344767643991,
      "grad_norm": 0.050053709891701666,
      "learning_rate": 1.986981647578823e-07,
      "loss": 0.05719013214111328,
      "step": 20980
    },
    {
      "epoch": 0.9734615830339552,
      "grad_norm": 0.04965772871141253,
      "learning_rate": 1.855512180790997e-07,
      "loss": 0.05678032040596008,
      "step": 21000
    },
    {
      "epoch": 0.9743886893035114,
      "grad_norm": 0.0530831130789604,
      "learning_rate": 1.7285341354256747e-07,
      "loss": 0.056834304332733156,
      "step": 21020
    },
    {
      "epoch": 0.9753157955730676,
      "grad_norm": 0.04707389612146195,
      "learning_rate": 1.606048656355441e-07,
      "loss": 0.05674393773078919,
      "step": 21040
    },
    {
      "epoch": 0.9762429018426237,
      "grad_norm": 0.0488705116944185,
      "learning_rate": 1.4880568479464508e-07,
      "loss": 0.05772377252578735,
      "step": 21060
    },
    {
      "epoch": 0.9771700081121799,
      "grad_norm": 0.051743590162587255,
      "learning_rate": 1.3745597740484896e-07,
      "loss": 0.057591086626052855,
      "step": 21080
    },
    {
      "epoch": 0.978097114381736,
      "grad_norm": 0.051324633897750487,
      "learning_rate": 1.2655584579854828e-07,
      "loss": 0.05727114081382752,
      "step": 21100
    },
    {
      "epoch": 0.9790242206512921,
      "grad_norm": 0.0450598917967124,
      "learning_rate": 1.1610538825463924e-07,
      "loss": 0.05462503433227539,
      "step": 21120
    },
    {
      "epoch": 0.9799513269208483,
      "grad_norm": 0.049920588691037986,
      "learning_rate": 1.061046989976e-07,
      "loss": 0.0616515040397644,
      "step": 21140
    },
    {
      "epoch": 0.9808784331904045,
      "grad_norm": 0.05427503958031083,
      "learning_rate": 9.655386819666934e-08,
      "loss": 0.05698297619819641,
      "step": 21160
    },
    {
      "epoch": 0.9818055394599606,
      "grad_norm": 0.046367681412642345,
      "learning_rate": 8.745298196503604e-08,
      "loss": 0.05826429724693298,
      "step": 21180
    },
    {
      "epoch": 0.9827326457295168,
      "grad_norm": 0.05079747160339065,
      "learning_rate": 7.880212235903406e-08,
      "loss": 0.055833280086517334,
      "step": 21200
    },
    {
      "epoch": 0.9836597519990729,
      "grad_norm": 0.05496941855259286,
      "learning_rate": 7.06013673774375e-08,
      "loss": 0.05561456084251404,
      "step": 21220
    },
    {
      "epoch": 0.984586858268629,
      "grad_norm": 0.05056966458273638,
      "learning_rate": 6.285079096073898e-08,
      "loss": 0.05830684304237366,
      "step": 21240
    },
    {
      "epoch": 0.9855139645381852,
      "grad_norm": 0.05006057927676643,
      "learning_rate": 5.5550462990466846e-08,
      "loss": 0.0554352879524231,
      "step": 21260
    },
    {
      "epoch": 0.9864410708077413,
      "grad_norm": 0.0644471510386673,
      "learning_rate": 4.870044928859674e-08,
      "loss": 0.06168003678321839,
      "step": 21280
    },
    {
      "epoch": 0.9873681770772975,
      "grad_norm": 0.05153007305303215,
      "learning_rate": 4.230081161692434e-08,
      "loss": 0.05833532810211182,
      "step": 21300
    },
    {
      "epoch": 0.9882952833468537,
      "grad_norm": 0.052481728179304875,
      "learning_rate": 3.6351607676515796e-08,
      "loss": 0.05807930827140808,
      "step": 21320
    },
    {
      "epoch": 0.9892223896164097,
      "grad_norm": 0.050221631530631174,
      "learning_rate": 3.085289110719702e-08,
      "loss": 0.05771827697753906,
      "step": 21340
    },
    {
      "epoch": 0.9901494958859659,
      "grad_norm": 0.05205457120253359,
      "learning_rate": 2.580471148705965e-08,
      "loss": 0.05922415852546692,
      "step": 21360
    },
    {
      "epoch": 0.9910766021555221,
      "grad_norm": 0.04934046554962467,
      "learning_rate": 2.1207114332016943e-08,
      "loss": 0.059090471267700194,
      "step": 21380
    },
    {
      "epoch": 0.9920037084250782,
      "grad_norm": 0.043651455909973144,
      "learning_rate": 1.7060141095393e-08,
      "loss": 0.055290830135345456,
      "step": 21400
    },
    {
      "epoch": 0.9929308146946344,
      "grad_norm": 0.051414602356183454,
      "learning_rate": 1.3363829167561958e-08,
      "loss": 0.05687727928161621,
      "step": 21420
    },
    {
      "epoch": 0.9938579209641906,
      "grad_norm": 0.04915157018855412,
      "learning_rate": 1.0118211875581595e-08,
      "loss": 0.05864974856376648,
      "step": 21440
    },
    {
      "epoch": 0.9947850272337466,
      "grad_norm": 0.046368563674900345,
      "learning_rate": 7.323318482926889e-09,
      "loss": 0.05641416311264038,
      "step": 21460
    },
    {
      "epoch": 0.9957121335033028,
      "grad_norm": 0.052243937275788514,
      "learning_rate": 4.979174189195801e-09,
      "loss": 0.05897241830825806,
      "step": 21480
    },
    {
      "epoch": 0.996639239772859,
      "grad_norm": 0.05505317821499118,
      "learning_rate": 3.085800129909444e-09,
      "loss": 0.058701437711715695,
      "step": 21500
    },
    {
      "epoch": 0.9975663460424151,
      "grad_norm": 0.052336038499922466,
      "learning_rate": 1.6432133762900314e-09,
      "loss": 0.06072026491165161,
      "step": 21520
    },
    {
      "epoch": 0.9984934523119713,
      "grad_norm": 0.04668286755815645,
      "learning_rate": 6.51426935149857e-10,
      "loss": 0.05787107944488525,
      "step": 21540
    },
    {
      "epoch": 0.9994205585815275,
      "grad_norm": 0.04099258647254872,
      "learning_rate": 1.1044974873031244e-10,
      "loss": 0.0565392017364502,
      "step": 21560
    },
    {
      "epoch": 1.0,
      "step": 21573,
      "total_flos": 5228992760119296.0,
      "train_loss": 0.07257944115598637,
      "train_runtime": 156672.9322,
      "train_samples_per_second": 1.102,
      "train_steps_per_second": 0.138
    }
  ],
  "logging_steps": 20,
  "max_steps": 21573,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5228992760119296.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}