tttw1004mmm / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
3819615 verified
Raw
History Blame Contribute Delete
98.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2622,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0019074868860276585,
"grad_norm": 1.4878780841827393,
"learning_rate": 1.0126582278481012e-05,
"loss": 2.7075782775878907,
"step": 5
},
{
"epoch": 0.003814973772055317,
"grad_norm": 0.7339136600494385,
"learning_rate": 2.278481012658228e-05,
"loss": 2.760773849487305,
"step": 10
},
{
"epoch": 0.005722460658082976,
"grad_norm": 0.3320401608943939,
"learning_rate": 3.5443037974683544e-05,
"loss": 2.5718051910400392,
"step": 15
},
{
"epoch": 0.007629947544110634,
"grad_norm": 0.34751737117767334,
"learning_rate": 4.810126582278481e-05,
"loss": 2.4293994903564453,
"step": 20
},
{
"epoch": 0.009537434430138292,
"grad_norm": 0.3222845494747162,
"learning_rate": 6.0759493670886084e-05,
"loss": 2.3764602661132814,
"step": 25
},
{
"epoch": 0.011444921316165951,
"grad_norm": 0.2202848643064499,
"learning_rate": 7.341772151898734e-05,
"loss": 2.2593830108642576,
"step": 30
},
{
"epoch": 0.01335240820219361,
"grad_norm": 0.19390147924423218,
"learning_rate": 8.607594936708861e-05,
"loss": 2.18019962310791,
"step": 35
},
{
"epoch": 0.015259895088221268,
"grad_norm": 0.20631256699562073,
"learning_rate": 9.873417721518988e-05,
"loss": 2.12917423248291,
"step": 40
},
{
"epoch": 0.017167381974248927,
"grad_norm": 0.19954322278499603,
"learning_rate": 0.00011139240506329114,
"loss": 1.9549896240234375,
"step": 45
},
{
"epoch": 0.019074868860276584,
"grad_norm": 0.2255832999944687,
"learning_rate": 0.0001240506329113924,
"loss": 1.9424320220947267,
"step": 50
},
{
"epoch": 0.020982355746304245,
"grad_norm": 0.18627505004405975,
"learning_rate": 0.00013670886075949366,
"loss": 1.9573286056518555,
"step": 55
},
{
"epoch": 0.022889842632331903,
"grad_norm": 0.18573220074176788,
"learning_rate": 0.00014936708860759494,
"loss": 1.8692535400390624,
"step": 60
},
{
"epoch": 0.02479732951835956,
"grad_norm": 0.17530076205730438,
"learning_rate": 0.0001620253164556962,
"loss": 1.8167079925537108,
"step": 65
},
{
"epoch": 0.02670481640438722,
"grad_norm": 0.21299882233142853,
"learning_rate": 0.00017468354430379748,
"loss": 1.7701532363891601,
"step": 70
},
{
"epoch": 0.02861230329041488,
"grad_norm": 0.22173331677913666,
"learning_rate": 0.00018734177215189873,
"loss": 1.699182891845703,
"step": 75
},
{
"epoch": 0.030519790176442536,
"grad_norm": 0.189641535282135,
"learning_rate": 0.0002,
"loss": 1.7398967742919922,
"step": 80
},
{
"epoch": 0.03242727706247019,
"grad_norm": 0.20088158547878265,
"learning_rate": 0.0001999980922755,
"loss": 1.7099655151367188,
"step": 85
},
{
"epoch": 0.034334763948497854,
"grad_norm": 0.22996209561824799,
"learning_rate": 0.00019999236917478817,
"loss": 1.6587194442749023,
"step": 90
},
{
"epoch": 0.036242250834525515,
"grad_norm": 0.21480797231197357,
"learning_rate": 0.00019998283091622653,
"loss": 1.6763959884643556,
"step": 95
},
{
"epoch": 0.03814973772055317,
"grad_norm": 0.1857302039861679,
"learning_rate": 0.00019996947786374248,
"loss": 1.668678092956543,
"step": 100
},
{
"epoch": 0.04005722460658083,
"grad_norm": 0.1957833468914032,
"learning_rate": 0.0001999523105268149,
"loss": 1.643313217163086,
"step": 105
},
{
"epoch": 0.04196471149260849,
"grad_norm": 0.22404824197292328,
"learning_rate": 0.0001999313295604548,
"loss": 1.643670654296875,
"step": 110
},
{
"epoch": 0.043872198378636144,
"grad_norm": 0.251034140586853,
"learning_rate": 0.0001999065357651803,
"loss": 1.6665712356567384,
"step": 115
},
{
"epoch": 0.045779685264663805,
"grad_norm": 0.19587400555610657,
"learning_rate": 0.0001998779300869859,
"loss": 1.6421092987060546,
"step": 120
},
{
"epoch": 0.047687172150691466,
"grad_norm": 0.2188933938741684,
"learning_rate": 0.00019984551361730673,
"loss": 1.6790538787841798,
"step": 125
},
{
"epoch": 0.04959465903671912,
"grad_norm": 0.20079247653484344,
"learning_rate": 0.00019980928759297663,
"loss": 1.6375768661499024,
"step": 130
},
{
"epoch": 0.05150214592274678,
"grad_norm": 0.20902550220489502,
"learning_rate": 0.00019976925339618116,
"loss": 1.608419418334961,
"step": 135
},
{
"epoch": 0.05340963280877444,
"grad_norm": 0.23899352550506592,
"learning_rate": 0.00019972541255440458,
"loss": 1.5500898361206055,
"step": 140
},
{
"epoch": 0.055317119694802096,
"grad_norm": 0.22073692083358765,
"learning_rate": 0.0001996777667403719,
"loss": 1.547786808013916,
"step": 145
},
{
"epoch": 0.05722460658082976,
"grad_norm": 0.2127048671245575,
"learning_rate": 0.00019962631777198486,
"loss": 1.6028972625732423,
"step": 150
},
{
"epoch": 0.05913209346685742,
"grad_norm": 0.19889818131923676,
"learning_rate": 0.0001995710676122526,
"loss": 1.5462740898132323,
"step": 155
},
{
"epoch": 0.06103958035288507,
"grad_norm": 0.1874799281358719,
"learning_rate": 0.0001995120183692168,
"loss": 1.5087604522705078,
"step": 160
},
{
"epoch": 0.06294706723891273,
"grad_norm": 0.17689307034015656,
"learning_rate": 0.00019944917229587115,
"loss": 1.5921195030212403,
"step": 165
},
{
"epoch": 0.06485455412494039,
"grad_norm": 0.18311065435409546,
"learning_rate": 0.00019938253179007564,
"loss": 1.5631231307983398,
"step": 170
},
{
"epoch": 0.06676204101096805,
"grad_norm": 0.19345863163471222,
"learning_rate": 0.00019931209939446466,
"loss": 1.5347042083740234,
"step": 175
},
{
"epoch": 0.06866952789699571,
"grad_norm": 0.18271172046661377,
"learning_rate": 0.00019923787779635042,
"loss": 1.561203384399414,
"step": 180
},
{
"epoch": 0.07057701478302336,
"grad_norm": 0.20191366970539093,
"learning_rate": 0.0001991598698276201,
"loss": 1.5551844596862794,
"step": 185
},
{
"epoch": 0.07248450166905103,
"grad_norm": 0.18917542695999146,
"learning_rate": 0.00019907807846462804,
"loss": 1.532535457611084,
"step": 190
},
{
"epoch": 0.07439198855507868,
"grad_norm": 0.2253742516040802,
"learning_rate": 0.00019899250682808189,
"loss": 1.535140323638916,
"step": 195
},
{
"epoch": 0.07629947544110634,
"grad_norm": 0.19846348464488983,
"learning_rate": 0.00019890315818292388,
"loss": 1.4907356262207032,
"step": 200
},
{
"epoch": 0.078206962327134,
"grad_norm": 0.20741984248161316,
"learning_rate": 0.00019881003593820592,
"loss": 1.494680118560791,
"step": 205
},
{
"epoch": 0.08011444921316166,
"grad_norm": 0.1832393854856491,
"learning_rate": 0.0001987131436469598,
"loss": 1.4730798721313476,
"step": 210
},
{
"epoch": 0.08202193609918931,
"grad_norm": 0.19742344319820404,
"learning_rate": 0.00019861248500606152,
"loss": 1.4961782455444337,
"step": 215
},
{
"epoch": 0.08392942298521698,
"grad_norm": 0.19493134319782257,
"learning_rate": 0.0001985080638560901,
"loss": 1.5330699920654296,
"step": 220
},
{
"epoch": 0.08583690987124463,
"grad_norm": 0.19508495926856995,
"learning_rate": 0.00019839988418118135,
"loss": 1.4591824531555175,
"step": 225
},
{
"epoch": 0.08774439675727229,
"grad_norm": 0.19378654658794403,
"learning_rate": 0.00019828795010887553,
"loss": 1.4908330917358399,
"step": 230
},
{
"epoch": 0.08965188364329996,
"grad_norm": 0.2403799593448639,
"learning_rate": 0.0001981722659099601,
"loss": 1.440755844116211,
"step": 235
},
{
"epoch": 0.09155937052932761,
"grad_norm": 0.21802683174610138,
"learning_rate": 0.0001980528359983067,
"loss": 1.5115406036376953,
"step": 240
},
{
"epoch": 0.09346685741535526,
"grad_norm": 0.1773674637079239,
"learning_rate": 0.00019792966493070268,
"loss": 1.5292924880981444,
"step": 245
},
{
"epoch": 0.09537434430138293,
"grad_norm": 0.18615812063217163,
"learning_rate": 0.00019780275740667732,
"loss": 1.4494241714477538,
"step": 250
},
{
"epoch": 0.09728183118741059,
"grad_norm": 0.1957283616065979,
"learning_rate": 0.0001976721182683224,
"loss": 1.5168323516845703,
"step": 255
},
{
"epoch": 0.09918931807343824,
"grad_norm": 0.19212238490581512,
"learning_rate": 0.00019753775250010773,
"loss": 1.4644780158996582,
"step": 260
},
{
"epoch": 0.10109680495946591,
"grad_norm": 0.16360795497894287,
"learning_rate": 0.0001973996652286906,
"loss": 1.4810004234313965,
"step": 265
},
{
"epoch": 0.10300429184549356,
"grad_norm": 0.25217700004577637,
"learning_rate": 0.0001972578617227205,
"loss": 1.4786994934082032,
"step": 270
},
{
"epoch": 0.10491177873152122,
"grad_norm": 0.1815384477376938,
"learning_rate": 0.0001971123473926378,
"loss": 1.435354232788086,
"step": 275
},
{
"epoch": 0.10681926561754888,
"grad_norm": 0.1852998435497284,
"learning_rate": 0.00019696312779046753,
"loss": 1.4307600021362306,
"step": 280
},
{
"epoch": 0.10872675250357654,
"grad_norm": 0.19455985724925995,
"learning_rate": 0.0001968102086096076,
"loss": 1.4663272857666017,
"step": 285
},
{
"epoch": 0.11063423938960419,
"grad_norm": 0.20260846614837646,
"learning_rate": 0.0001966535956846113,
"loss": 1.4134601593017577,
"step": 290
},
{
"epoch": 0.11254172627563186,
"grad_norm": 0.19934098422527313,
"learning_rate": 0.00019649329499096495,
"loss": 1.496086883544922,
"step": 295
},
{
"epoch": 0.11444921316165951,
"grad_norm": 0.18789365887641907,
"learning_rate": 0.00019632931264485976,
"loss": 1.513808822631836,
"step": 300
},
{
"epoch": 0.11635670004768717,
"grad_norm": 0.20016062259674072,
"learning_rate": 0.00019616165490295848,
"loss": 1.4521299362182618,
"step": 305
},
{
"epoch": 0.11826418693371483,
"grad_norm": 0.22385726869106293,
"learning_rate": 0.0001959903281621568,
"loss": 1.4325057983398437,
"step": 310
},
{
"epoch": 0.12017167381974249,
"grad_norm": 0.19611406326293945,
"learning_rate": 0.00019581533895933907,
"loss": 1.5017765045166016,
"step": 315
},
{
"epoch": 0.12207916070577014,
"grad_norm": 0.20493510365486145,
"learning_rate": 0.00019563669397112916,
"loss": 1.3843585014343263,
"step": 320
},
{
"epoch": 0.12398664759179781,
"grad_norm": 0.19239553809165955,
"learning_rate": 0.00019545440001363543,
"loss": 1.5033424377441407,
"step": 325
},
{
"epoch": 0.12589413447782546,
"grad_norm": 0.17423652112483978,
"learning_rate": 0.00019526846404219086,
"loss": 1.4363468170166016,
"step": 330
},
{
"epoch": 0.12780162136385312,
"grad_norm": 0.1988988071680069,
"learning_rate": 0.00019507889315108763,
"loss": 1.4650516510009766,
"step": 335
},
{
"epoch": 0.12970910824988077,
"grad_norm": 0.18917162716388702,
"learning_rate": 0.0001948856945733064,
"loss": 1.4346914291381836,
"step": 340
},
{
"epoch": 0.13161659513590845,
"grad_norm": 0.19002121686935425,
"learning_rate": 0.0001946888756802404,
"loss": 1.4026443481445312,
"step": 345
},
{
"epoch": 0.1335240820219361,
"grad_norm": 0.18209987878799438,
"learning_rate": 0.0001944884439814141,
"loss": 1.4519460678100586,
"step": 350
},
{
"epoch": 0.13543156890796376,
"grad_norm": 0.23619088530540466,
"learning_rate": 0.00019428440712419669,
"loss": 1.4020172119140626,
"step": 355
},
{
"epoch": 0.13733905579399142,
"grad_norm": 0.17148365080356598,
"learning_rate": 0.0001940767728935105,
"loss": 1.3774951934814452,
"step": 360
},
{
"epoch": 0.13924654268001907,
"grad_norm": 0.20871533453464508,
"learning_rate": 0.00019386554921153363,
"loss": 1.409653091430664,
"step": 365
},
{
"epoch": 0.14115402956604672,
"grad_norm": 0.19056005775928497,
"learning_rate": 0.000193650744137398,
"loss": 1.4494594573974608,
"step": 370
},
{
"epoch": 0.1430615164520744,
"grad_norm": 0.2262170910835266,
"learning_rate": 0.00019343236586688166,
"loss": 1.3713117599487306,
"step": 375
},
{
"epoch": 0.14496900333810206,
"grad_norm": 0.20798605680465698,
"learning_rate": 0.00019321042273209607,
"loss": 1.4061501502990723,
"step": 380
},
{
"epoch": 0.1468764902241297,
"grad_norm": 0.19938936829566956,
"learning_rate": 0.00019298492320116842,
"loss": 1.4196322441101075,
"step": 385
},
{
"epoch": 0.14878397711015737,
"grad_norm": 0.18144837021827698,
"learning_rate": 0.0001927558758779183,
"loss": 1.4253101348876953,
"step": 390
},
{
"epoch": 0.15069146399618502,
"grad_norm": 0.22917026281356812,
"learning_rate": 0.00019252328950152952,
"loss": 1.3228367805480956,
"step": 395
},
{
"epoch": 0.15259895088221268,
"grad_norm": 0.18507860600948334,
"learning_rate": 0.00019228717294621662,
"loss": 1.4104506492614746,
"step": 400
},
{
"epoch": 0.15450643776824036,
"grad_norm": 0.17714622616767883,
"learning_rate": 0.00019204753522088637,
"loss": 1.4143267631530763,
"step": 405
},
{
"epoch": 0.156413924654268,
"grad_norm": 0.2012583315372467,
"learning_rate": 0.00019180438546879396,
"loss": 1.4302897453308105,
"step": 410
},
{
"epoch": 0.15832141154029566,
"grad_norm": 0.22738420963287354,
"learning_rate": 0.00019155773296719418,
"loss": 1.4935453414916993,
"step": 415
},
{
"epoch": 0.16022889842632332,
"grad_norm": 0.18231648206710815,
"learning_rate": 0.00019130758712698743,
"loss": 1.3929641723632813,
"step": 420
},
{
"epoch": 0.16213638531235097,
"grad_norm": 0.18094943463802338,
"learning_rate": 0.00019105395749236066,
"loss": 1.3849353790283203,
"step": 425
},
{
"epoch": 0.16404387219837863,
"grad_norm": 0.18865585327148438,
"learning_rate": 0.00019079685374042324,
"loss": 1.4427093505859374,
"step": 430
},
{
"epoch": 0.1659513590844063,
"grad_norm": 0.22992666065692902,
"learning_rate": 0.0001905362856808377,
"loss": 1.384532356262207,
"step": 435
},
{
"epoch": 0.16785884597043396,
"grad_norm": 0.19635477662086487,
"learning_rate": 0.00019027226325544548,
"loss": 1.3813672065734863,
"step": 440
},
{
"epoch": 0.16976633285646162,
"grad_norm": 0.18714486062526703,
"learning_rate": 0.00019000479653788753,
"loss": 1.4129705429077148,
"step": 445
},
{
"epoch": 0.17167381974248927,
"grad_norm": 0.19622617959976196,
"learning_rate": 0.00018973389573322008,
"loss": 1.3845558166503906,
"step": 450
},
{
"epoch": 0.17358130662851692,
"grad_norm": 0.1831693947315216,
"learning_rate": 0.0001894595711775252,
"loss": 1.4312222480773926,
"step": 455
},
{
"epoch": 0.17548879351454458,
"grad_norm": 0.1820661872625351,
"learning_rate": 0.00018918183333751633,
"loss": 1.3978130340576171,
"step": 460
},
{
"epoch": 0.17739628040057226,
"grad_norm": 0.19805996119976044,
"learning_rate": 0.0001889006928101392,
"loss": 1.3540464401245118,
"step": 465
},
{
"epoch": 0.1793037672865999,
"grad_norm": 0.16892942786216736,
"learning_rate": 0.00018861616032216715,
"loss": 1.37994441986084,
"step": 470
},
{
"epoch": 0.18121125417262757,
"grad_norm": 0.18616019189357758,
"learning_rate": 0.00018832824672979225,
"loss": 1.3682477951049805,
"step": 475
},
{
"epoch": 0.18311874105865522,
"grad_norm": 0.1713143289089203,
"learning_rate": 0.0001880369630182107,
"loss": 1.4018159866333009,
"step": 480
},
{
"epoch": 0.18502622794468288,
"grad_norm": 0.20789170265197754,
"learning_rate": 0.000187742320301204,
"loss": 1.4139459609985352,
"step": 485
},
{
"epoch": 0.18693371483071053,
"grad_norm": 0.1693965643644333,
"learning_rate": 0.00018744432982071478,
"loss": 1.3742374420166015,
"step": 490
},
{
"epoch": 0.1888412017167382,
"grad_norm": 0.20864346623420715,
"learning_rate": 0.0001871430029464178,
"loss": 1.3525377273559571,
"step": 495
},
{
"epoch": 0.19074868860276586,
"grad_norm": 0.18742972612380981,
"learning_rate": 0.00018683835117528624,
"loss": 1.3390338897705079,
"step": 500
},
{
"epoch": 0.19265617548879352,
"grad_norm": 0.17989704012870789,
"learning_rate": 0.00018653038613115314,
"loss": 1.3564807891845703,
"step": 505
},
{
"epoch": 0.19456366237482117,
"grad_norm": 0.17508839070796967,
"learning_rate": 0.00018621911956426764,
"loss": 1.3876054763793946,
"step": 510
},
{
"epoch": 0.19647114926084883,
"grad_norm": 0.18916957080364227,
"learning_rate": 0.00018590456335084688,
"loss": 1.3788718223571776,
"step": 515
},
{
"epoch": 0.19837863614687648,
"grad_norm": 0.1722261607646942,
"learning_rate": 0.00018558672949262276,
"loss": 1.3902565002441407,
"step": 520
},
{
"epoch": 0.20028612303290416,
"grad_norm": 0.17852212488651276,
"learning_rate": 0.00018526563011638398,
"loss": 1.35484037399292,
"step": 525
},
{
"epoch": 0.20219360991893182,
"grad_norm": 0.21537213027477264,
"learning_rate": 0.00018494127747351358,
"loss": 1.3869443893432618,
"step": 530
},
{
"epoch": 0.20410109680495947,
"grad_norm": 0.21420900523662567,
"learning_rate": 0.00018461368393952118,
"loss": 1.3468827247619628,
"step": 535
},
{
"epoch": 0.20600858369098712,
"grad_norm": 0.1926850527524948,
"learning_rate": 0.00018428286201357098,
"loss": 1.352040672302246,
"step": 540
},
{
"epoch": 0.20791607057701478,
"grad_norm": 0.20665980875492096,
"learning_rate": 0.0001839488243180049,
"loss": 1.3783277511596679,
"step": 545
},
{
"epoch": 0.20982355746304243,
"grad_norm": 0.1789952665567398,
"learning_rate": 0.00018361158359786082,
"loss": 1.369606113433838,
"step": 550
},
{
"epoch": 0.2117310443490701,
"grad_norm": 0.17610864341259003,
"learning_rate": 0.0001832711527203864,
"loss": 1.3921396255493164,
"step": 555
},
{
"epoch": 0.21363853123509777,
"grad_norm": 0.23415693640708923,
"learning_rate": 0.0001829275446745482,
"loss": 1.365463638305664,
"step": 560
},
{
"epoch": 0.21554601812112542,
"grad_norm": 0.17977698147296906,
"learning_rate": 0.00018258077257053595,
"loss": 1.3199995040893555,
"step": 565
},
{
"epoch": 0.21745350500715308,
"grad_norm": 0.21059787273406982,
"learning_rate": 0.0001822308496392624,
"loss": 1.3279150009155274,
"step": 570
},
{
"epoch": 0.21936099189318073,
"grad_norm": 0.19721657037734985,
"learning_rate": 0.00018187778923185857,
"loss": 1.3629688262939452,
"step": 575
},
{
"epoch": 0.22126847877920838,
"grad_norm": 0.20456600189208984,
"learning_rate": 0.0001815216048191642,
"loss": 1.360971450805664,
"step": 580
},
{
"epoch": 0.22317596566523606,
"grad_norm": 0.23791781067848206,
"learning_rate": 0.00018116230999121394,
"loss": 1.394046688079834,
"step": 585
},
{
"epoch": 0.22508345255126372,
"grad_norm": 0.21695254743099213,
"learning_rate": 0.0001807999184567187,
"loss": 1.3695282936096191,
"step": 590
},
{
"epoch": 0.22699093943729137,
"grad_norm": 0.19964495301246643,
"learning_rate": 0.00018043444404254269,
"loss": 1.3753707885742188,
"step": 595
},
{
"epoch": 0.22889842632331903,
"grad_norm": 0.1965213567018509,
"learning_rate": 0.00018006590069317573,
"loss": 1.3258387565612793,
"step": 600
},
{
"epoch": 0.23080591320934668,
"grad_norm": 0.18316830694675446,
"learning_rate": 0.0001796943024702014,
"loss": 1.3746368408203125,
"step": 605
},
{
"epoch": 0.23271340009537433,
"grad_norm": 0.18314281105995178,
"learning_rate": 0.00017931966355176035,
"loss": 1.3545222282409668,
"step": 610
},
{
"epoch": 0.23462088698140202,
"grad_norm": 0.22398529946804047,
"learning_rate": 0.00017894199823200948,
"loss": 1.3660289764404296,
"step": 615
},
{
"epoch": 0.23652837386742967,
"grad_norm": 0.187253937125206,
"learning_rate": 0.00017856132092057645,
"loss": 1.3202295303344727,
"step": 620
},
{
"epoch": 0.23843586075345732,
"grad_norm": 0.16926705837249756,
"learning_rate": 0.00017817764614200994,
"loss": 1.3768438339233398,
"step": 625
},
{
"epoch": 0.24034334763948498,
"grad_norm": 0.17729657888412476,
"learning_rate": 0.0001777909885352254,
"loss": 1.3506011962890625,
"step": 630
},
{
"epoch": 0.24225083452551263,
"grad_norm": 0.20710691809654236,
"learning_rate": 0.0001774013628529467,
"loss": 1.3145681381225587,
"step": 635
},
{
"epoch": 0.24415832141154029,
"grad_norm": 0.1813536286354065,
"learning_rate": 0.000177008783961143,
"loss": 1.3291179656982421,
"step": 640
},
{
"epoch": 0.24606580829756797,
"grad_norm": 0.1942223459482193,
"learning_rate": 0.0001766132668384617,
"loss": 1.2867470741271974,
"step": 645
},
{
"epoch": 0.24797329518359562,
"grad_norm": 0.2119131237268448,
"learning_rate": 0.00017621482657565686,
"loss": 1.3372028350830079,
"step": 650
},
{
"epoch": 0.24988078206962328,
"grad_norm": 0.17588655650615692,
"learning_rate": 0.00017581347837501357,
"loss": 1.3778643608093262,
"step": 655
},
{
"epoch": 0.25178826895565093,
"grad_norm": 0.20653095841407776,
"learning_rate": 0.0001754092375497677,
"loss": 1.3122962951660155,
"step": 660
},
{
"epoch": 0.2536957558416786,
"grad_norm": 0.1691495180130005,
"learning_rate": 0.00017500211952352183,
"loss": 1.318179225921631,
"step": 665
},
{
"epoch": 0.25560324272770624,
"grad_norm": 0.16325566172599792,
"learning_rate": 0.00017459213982965656,
"loss": 1.3097293853759766,
"step": 670
},
{
"epoch": 0.2575107296137339,
"grad_norm": 0.19829092919826508,
"learning_rate": 0.00017417931411073804,
"loss": 1.271612548828125,
"step": 675
},
{
"epoch": 0.25941821649976154,
"grad_norm": 0.1984972357749939,
"learning_rate": 0.00017376365811792104,
"loss": 1.3652052879333496,
"step": 680
},
{
"epoch": 0.2613257033857892,
"grad_norm": 0.19141259789466858,
"learning_rate": 0.00017334518771034799,
"loss": 1.3176380157470704,
"step": 685
},
{
"epoch": 0.2632331902718169,
"grad_norm": 0.22068865597248077,
"learning_rate": 0.00017292391885454381,
"loss": 1.2949045181274415,
"step": 690
},
{
"epoch": 0.26514067715784456,
"grad_norm": 0.20108754932880402,
"learning_rate": 0.00017249986762380695,
"loss": 1.278764247894287,
"step": 695
},
{
"epoch": 0.2670481640438722,
"grad_norm": 0.17690886557102203,
"learning_rate": 0.00017207305019759577,
"loss": 1.3423615455627442,
"step": 700
},
{
"epoch": 0.26895565092989987,
"grad_norm": 0.16012556850910187,
"learning_rate": 0.00017164348286091152,
"loss": 1.30777587890625,
"step": 705
},
{
"epoch": 0.2708631378159275,
"grad_norm": 0.17252199351787567,
"learning_rate": 0.00017121118200367683,
"loss": 1.3042643547058106,
"step": 710
},
{
"epoch": 0.2727706247019552,
"grad_norm": 0.18792898952960968,
"learning_rate": 0.0001707761641201105,
"loss": 1.2652257919311523,
"step": 715
},
{
"epoch": 0.27467811158798283,
"grad_norm": 0.19583508372306824,
"learning_rate": 0.00017033844580809792,
"loss": 1.351115608215332,
"step": 720
},
{
"epoch": 0.2765855984740105,
"grad_norm": 0.19298936426639557,
"learning_rate": 0.00016989804376855815,
"loss": 1.355845832824707,
"step": 725
},
{
"epoch": 0.27849308536003814,
"grad_norm": 0.18880149722099304,
"learning_rate": 0.0001694549748048063,
"loss": 1.3094839096069335,
"step": 730
},
{
"epoch": 0.2804005722460658,
"grad_norm": 0.17870928347110748,
"learning_rate": 0.00016900925582191283,
"loss": 1.3336545944213867,
"step": 735
},
{
"epoch": 0.28230805913209345,
"grad_norm": 0.16946914792060852,
"learning_rate": 0.00016856090382605813,
"loss": 1.3054865837097167,
"step": 740
},
{
"epoch": 0.2842155460181211,
"grad_norm": 0.1863696426153183,
"learning_rate": 0.00016810993592388397,
"loss": 1.2880358695983887,
"step": 745
},
{
"epoch": 0.2861230329041488,
"grad_norm": 0.19960841536521912,
"learning_rate": 0.00016765636932184066,
"loss": 1.3031925201416015,
"step": 750
},
{
"epoch": 0.28803051979017646,
"grad_norm": 0.1861932873725891,
"learning_rate": 0.0001672002213255306,
"loss": 1.2319182395935058,
"step": 755
},
{
"epoch": 0.2899380066762041,
"grad_norm": 0.1846766322851181,
"learning_rate": 0.00016674150933904794,
"loss": 1.3236438751220703,
"step": 760
},
{
"epoch": 0.2918454935622318,
"grad_norm": 0.2234894335269928,
"learning_rate": 0.00016628025086431458,
"loss": 1.3154956817626953,
"step": 765
},
{
"epoch": 0.2937529804482594,
"grad_norm": 0.18873247504234314,
"learning_rate": 0.00016581646350041237,
"loss": 1.267470932006836,
"step": 770
},
{
"epoch": 0.2956604673342871,
"grad_norm": 0.1765766590833664,
"learning_rate": 0.0001653501649429117,
"loss": 1.2449243545532227,
"step": 775
},
{
"epoch": 0.29756795422031473,
"grad_norm": 0.21779510378837585,
"learning_rate": 0.00016488137298319616,
"loss": 1.3252256393432618,
"step": 780
},
{
"epoch": 0.2994754411063424,
"grad_norm": 0.18757256865501404,
"learning_rate": 0.0001644101055077839,
"loss": 1.3002010345458985,
"step": 785
},
{
"epoch": 0.30138292799237004,
"grad_norm": 0.19776006042957306,
"learning_rate": 0.00016393638049764513,
"loss": 1.2941418647766114,
"step": 790
},
{
"epoch": 0.3032904148783977,
"grad_norm": 0.17863471806049347,
"learning_rate": 0.000163460216027516,
"loss": 1.3113096237182618,
"step": 795
},
{
"epoch": 0.30519790176442535,
"grad_norm": 0.20542913675308228,
"learning_rate": 0.00016298163026520898,
"loss": 1.2751188278198242,
"step": 800
},
{
"epoch": 0.307105388650453,
"grad_norm": 0.20658336579799652,
"learning_rate": 0.0001625006414709198,
"loss": 1.2631260871887207,
"step": 805
},
{
"epoch": 0.3090128755364807,
"grad_norm": 0.18959608674049377,
"learning_rate": 0.0001620172679965306,
"loss": 1.24639949798584,
"step": 810
},
{
"epoch": 0.31092036242250837,
"grad_norm": 0.18943072855472565,
"learning_rate": 0.0001615315282849098,
"loss": 1.3100577354431153,
"step": 815
},
{
"epoch": 0.312827849308536,
"grad_norm": 0.1695513129234314,
"learning_rate": 0.00016104344086920826,
"loss": 1.2780745506286622,
"step": 820
},
{
"epoch": 0.3147353361945637,
"grad_norm": 0.1825125515460968,
"learning_rate": 0.0001605530243721525,
"loss": 1.307393455505371,
"step": 825
},
{
"epoch": 0.31664282308059133,
"grad_norm": 0.1687004119157791,
"learning_rate": 0.00016006029750533388,
"loss": 1.3595869064331054,
"step": 830
},
{
"epoch": 0.318550309966619,
"grad_norm": 0.21706590056419373,
"learning_rate": 0.00015956527906849463,
"loss": 1.2534672737121582,
"step": 835
},
{
"epoch": 0.32045779685264664,
"grad_norm": 0.1913917362689972,
"learning_rate": 0.0001590679879488108,
"loss": 1.3069406509399415,
"step": 840
},
{
"epoch": 0.3223652837386743,
"grad_norm": 0.2003803551197052,
"learning_rate": 0.00015856844312017138,
"loss": 1.3091538429260254,
"step": 845
},
{
"epoch": 0.32427277062470194,
"grad_norm": 0.21809028089046478,
"learning_rate": 0.00015806666364245463,
"loss": 1.257524299621582,
"step": 850
},
{
"epoch": 0.3261802575107296,
"grad_norm": 0.18535156548023224,
"learning_rate": 0.00015756266866080054,
"loss": 1.3084287643432617,
"step": 855
},
{
"epoch": 0.32808774439675725,
"grad_norm": 0.18393605947494507,
"learning_rate": 0.00015705647740488066,
"loss": 1.2479835510253907,
"step": 860
},
{
"epoch": 0.3299952312827849,
"grad_norm": 0.2007489949464798,
"learning_rate": 0.00015654810918816416,
"loss": 1.3321989059448243,
"step": 865
},
{
"epoch": 0.3319027181688126,
"grad_norm": 0.19368581473827362,
"learning_rate": 0.00015603758340718106,
"loss": 1.3160361289978026,
"step": 870
},
{
"epoch": 0.33381020505484027,
"grad_norm": 0.23067910969257355,
"learning_rate": 0.00015552491954078223,
"loss": 1.2804203033447266,
"step": 875
},
{
"epoch": 0.3357176919408679,
"grad_norm": 0.19344988465309143,
"learning_rate": 0.00015501013714939597,
"loss": 1.248469638824463,
"step": 880
},
{
"epoch": 0.3376251788268956,
"grad_norm": 0.28040802478790283,
"learning_rate": 0.00015449325587428193,
"loss": 1.3148443222045898,
"step": 885
},
{
"epoch": 0.33953266571292323,
"grad_norm": 0.16799642145633698,
"learning_rate": 0.00015397429543678154,
"loss": 1.2800113677978515,
"step": 890
},
{
"epoch": 0.3414401525989509,
"grad_norm": 0.21939077973365784,
"learning_rate": 0.00015345327563756565,
"loss": 1.2717226028442383,
"step": 895
},
{
"epoch": 0.34334763948497854,
"grad_norm": 0.19597913324832916,
"learning_rate": 0.00015293021635587892,
"loss": 1.232340145111084,
"step": 900
},
{
"epoch": 0.3452551263710062,
"grad_norm": 0.19841314852237701,
"learning_rate": 0.00015240513754878152,
"loss": 1.281689167022705,
"step": 905
},
{
"epoch": 0.34716261325703385,
"grad_norm": 0.24963513016700745,
"learning_rate": 0.00015187805925038749,
"loss": 1.278480339050293,
"step": 910
},
{
"epoch": 0.3490701001430615,
"grad_norm": 0.18897737562656403,
"learning_rate": 0.0001513490015711006,
"loss": 1.2911778450012208,
"step": 915
},
{
"epoch": 0.35097758702908916,
"grad_norm": 0.17709249258041382,
"learning_rate": 0.00015081798469684674,
"loss": 1.276332187652588,
"step": 920
},
{
"epoch": 0.3528850739151168,
"grad_norm": 0.19648757576942444,
"learning_rate": 0.0001502850288883039,
"loss": 1.3356638908386231,
"step": 925
},
{
"epoch": 0.3547925608011445,
"grad_norm": 0.20174409449100494,
"learning_rate": 0.00014975015448012914,
"loss": 1.288701629638672,
"step": 930
},
{
"epoch": 0.3567000476871722,
"grad_norm": 0.17610758543014526,
"learning_rate": 0.00014921338188018276,
"loss": 1.334892463684082,
"step": 935
},
{
"epoch": 0.3586075345731998,
"grad_norm": 0.1958654373884201,
"learning_rate": 0.00014867473156874957,
"loss": 1.2658604621887206,
"step": 940
},
{
"epoch": 0.3605150214592275,
"grad_norm": 0.17997604608535767,
"learning_rate": 0.0001481342240977574,
"loss": 1.2249727249145508,
"step": 945
},
{
"epoch": 0.36242250834525513,
"grad_norm": 0.18638291954994202,
"learning_rate": 0.00014759188008999324,
"loss": 1.2768918991088867,
"step": 950
},
{
"epoch": 0.3643299952312828,
"grad_norm": 0.19101013243198395,
"learning_rate": 0.00014704772023831607,
"loss": 1.2294296264648437,
"step": 955
},
{
"epoch": 0.36623748211731044,
"grad_norm": 0.23659354448318481,
"learning_rate": 0.00014650176530486752,
"loss": 1.2219261169433593,
"step": 960
},
{
"epoch": 0.3681449690033381,
"grad_norm": 0.19072403013706207,
"learning_rate": 0.00014595403612027968,
"loss": 1.2200937271118164,
"step": 965
},
{
"epoch": 0.37005245588936575,
"grad_norm": 0.18234267830848694,
"learning_rate": 0.00014540455358288016,
"loss": 1.3056456565856933,
"step": 970
},
{
"epoch": 0.3719599427753934,
"grad_norm": 0.19649538397789001,
"learning_rate": 0.000144853338657895,
"loss": 1.2643993377685547,
"step": 975
},
{
"epoch": 0.37386742966142106,
"grad_norm": 0.16372917592525482,
"learning_rate": 0.00014430041237664852,
"loss": 1.232365608215332,
"step": 980
},
{
"epoch": 0.3757749165474487,
"grad_norm": 0.1872534155845642,
"learning_rate": 0.000143745795835761,
"loss": 1.2186118125915528,
"step": 985
},
{
"epoch": 0.3776824034334764,
"grad_norm": 0.19309327006340027,
"learning_rate": 0.00014318951019634367,
"loss": 1.2885890007019043,
"step": 990
},
{
"epoch": 0.3795898903195041,
"grad_norm": 0.19395118951797485,
"learning_rate": 0.00014263157668319146,
"loss": 1.2730616569519042,
"step": 995
},
{
"epoch": 0.38149737720553173,
"grad_norm": 0.2152969241142273,
"learning_rate": 0.00014207201658397297,
"loss": 1.263326644897461,
"step": 1000
},
{
"epoch": 0.3834048640915594,
"grad_norm": 0.18065625429153442,
"learning_rate": 0.00014151085124841845,
"loss": 1.2817727088928224,
"step": 1005
},
{
"epoch": 0.38531235097758704,
"grad_norm": 0.17509140074253082,
"learning_rate": 0.00014094810208750504,
"loss": 1.260600471496582,
"step": 1010
},
{
"epoch": 0.3872198378636147,
"grad_norm": 0.18434332311153412,
"learning_rate": 0.00014038379057263996,
"loss": 1.2359878540039062,
"step": 1015
},
{
"epoch": 0.38912732474964234,
"grad_norm": 0.1682165563106537,
"learning_rate": 0.00013981793823484136,
"loss": 1.2533650398254395,
"step": 1020
},
{
"epoch": 0.39103481163567,
"grad_norm": 0.23548661172389984,
"learning_rate": 0.00013925056666391647,
"loss": 1.2457769393920899,
"step": 1025
},
{
"epoch": 0.39294229852169765,
"grad_norm": 0.1865137219429016,
"learning_rate": 0.0001386816975076383,
"loss": 1.2864510536193847,
"step": 1030
},
{
"epoch": 0.3948497854077253,
"grad_norm": 0.19426438212394714,
"learning_rate": 0.00013811135247091943,
"loss": 1.2255964279174805,
"step": 1035
},
{
"epoch": 0.39675727229375296,
"grad_norm": 0.19577838480472565,
"learning_rate": 0.00013753955331498373,
"loss": 1.2952438354492188,
"step": 1040
},
{
"epoch": 0.3986647591797806,
"grad_norm": 0.19571255147457123,
"learning_rate": 0.00013696632185653646,
"loss": 1.266249179840088,
"step": 1045
},
{
"epoch": 0.4005722460658083,
"grad_norm": 0.2026452124118805,
"learning_rate": 0.0001363916799669316,
"loss": 1.2452361106872558,
"step": 1050
},
{
"epoch": 0.402479732951836,
"grad_norm": 0.2117244303226471,
"learning_rate": 0.00013581564957133732,
"loss": 1.2749139785766601,
"step": 1055
},
{
"epoch": 0.40438721983786363,
"grad_norm": 0.19827903807163239,
"learning_rate": 0.0001352382526478996,
"loss": 1.2434818267822265,
"step": 1060
},
{
"epoch": 0.4062947067238913,
"grad_norm": 0.1747330278158188,
"learning_rate": 0.00013465951122690356,
"loss": 1.2268753051757812,
"step": 1065
},
{
"epoch": 0.40820219360991894,
"grad_norm": 0.22783830761909485,
"learning_rate": 0.00013407944738993304,
"loss": 1.2166577339172364,
"step": 1070
},
{
"epoch": 0.4101096804959466,
"grad_norm": 0.20430943369865417,
"learning_rate": 0.0001334980832690278,
"loss": 1.2934347152709962,
"step": 1075
},
{
"epoch": 0.41201716738197425,
"grad_norm": 0.17204000055789948,
"learning_rate": 0.0001329154410458395,
"loss": 1.251021957397461,
"step": 1080
},
{
"epoch": 0.4139246542680019,
"grad_norm": 0.22515660524368286,
"learning_rate": 0.00013233154295078494,
"loss": 1.1626641273498535,
"step": 1085
},
{
"epoch": 0.41583214115402956,
"grad_norm": 0.20006167888641357,
"learning_rate": 0.00013174641126219815,
"loss": 1.2613157272338866,
"step": 1090
},
{
"epoch": 0.4177396280400572,
"grad_norm": 0.20634442567825317,
"learning_rate": 0.0001311600683054803,
"loss": 1.2518367767333984,
"step": 1095
},
{
"epoch": 0.41964711492608486,
"grad_norm": 0.1790790855884552,
"learning_rate": 0.00013057253645224788,
"loss": 1.2314929962158203,
"step": 1100
},
{
"epoch": 0.4215546018121125,
"grad_norm": 0.24240708351135254,
"learning_rate": 0.00012998383811947908,
"loss": 1.272528076171875,
"step": 1105
},
{
"epoch": 0.4234620886981402,
"grad_norm": 0.24455605447292328,
"learning_rate": 0.00012939399576865858,
"loss": 1.2026556015014649,
"step": 1110
},
{
"epoch": 0.4253695755841679,
"grad_norm": 0.1865748018026352,
"learning_rate": 0.00012880303190492044,
"loss": 1.2480610847473144,
"step": 1115
},
{
"epoch": 0.42727706247019553,
"grad_norm": 0.17938241362571716,
"learning_rate": 0.00012821096907618954,
"loss": 1.2315555572509767,
"step": 1120
},
{
"epoch": 0.4291845493562232,
"grad_norm": 0.21218787133693695,
"learning_rate": 0.0001276178298723211,
"loss": 1.2253150939941406,
"step": 1125
},
{
"epoch": 0.43109203624225084,
"grad_norm": 0.1933385133743286,
"learning_rate": 0.00012702363692423894,
"loss": 1.195798110961914,
"step": 1130
},
{
"epoch": 0.4329995231282785,
"grad_norm": 0.2011338770389557,
"learning_rate": 0.000126428412903072,
"loss": 1.2208805084228516,
"step": 1135
},
{
"epoch": 0.43490701001430615,
"grad_norm": 0.19478750228881836,
"learning_rate": 0.00012583218051928917,
"loss": 1.2897579193115234,
"step": 1140
},
{
"epoch": 0.4368144969003338,
"grad_norm": 0.1855502873659134,
"learning_rate": 0.00012523496252183307,
"loss": 1.257915210723877,
"step": 1145
},
{
"epoch": 0.43872198378636146,
"grad_norm": 0.24520094692707062,
"learning_rate": 0.00012463678169725176,
"loss": 1.24306640625,
"step": 1150
},
{
"epoch": 0.4406294706723891,
"grad_norm": 0.19714589416980743,
"learning_rate": 0.00012403766086882954,
"loss": 1.2973305702209472,
"step": 1155
},
{
"epoch": 0.44253695755841677,
"grad_norm": 0.20193324983119965,
"learning_rate": 0.00012343762289571605,
"loss": 1.2306537628173828,
"step": 1160
},
{
"epoch": 0.4444444444444444,
"grad_norm": 0.20026658475399017,
"learning_rate": 0.0001228366906720542,
"loss": 1.1923845291137696,
"step": 1165
},
{
"epoch": 0.44635193133047213,
"grad_norm": 0.16747166216373444,
"learning_rate": 0.0001222348871261064,
"loss": 1.2517187118530273,
"step": 1170
},
{
"epoch": 0.4482594182164998,
"grad_norm": 0.20404314994812012,
"learning_rate": 0.0001216322352193801,
"loss": 1.254922580718994,
"step": 1175
},
{
"epoch": 0.45016690510252744,
"grad_norm": 0.18388152122497559,
"learning_rate": 0.00012102875794575142,
"loss": 1.2511006355285645,
"step": 1180
},
{
"epoch": 0.4520743919885551,
"grad_norm": 0.25624749064445496,
"learning_rate": 0.00012042447833058799,
"loss": 1.1854961395263672,
"step": 1185
},
{
"epoch": 0.45398187887458274,
"grad_norm": 0.2018139660358429,
"learning_rate": 0.00011981941942987034,
"loss": 1.2183271408081056,
"step": 1190
},
{
"epoch": 0.4558893657606104,
"grad_norm": 0.1980033665895462,
"learning_rate": 0.00011921360432931222,
"loss": 1.2349556922912597,
"step": 1195
},
{
"epoch": 0.45779685264663805,
"grad_norm": 0.1895417869091034,
"learning_rate": 0.00011860705614347985,
"loss": 1.194553756713867,
"step": 1200
},
{
"epoch": 0.4597043395326657,
"grad_norm": 0.21431392431259155,
"learning_rate": 0.00011799979801490992,
"loss": 1.1978261947631836,
"step": 1205
},
{
"epoch": 0.46161182641869336,
"grad_norm": 0.20577406883239746,
"learning_rate": 0.0001173918531132266,
"loss": 1.182002639770508,
"step": 1210
},
{
"epoch": 0.463519313304721,
"grad_norm": 0.1944798082113266,
"learning_rate": 0.00011678324463425758,
"loss": 1.1809057235717773,
"step": 1215
},
{
"epoch": 0.46542680019074867,
"grad_norm": 0.18515107035636902,
"learning_rate": 0.000116173995799149,
"loss": 1.2245164871215821,
"step": 1220
},
{
"epoch": 0.4673342870767763,
"grad_norm": 0.2592999041080475,
"learning_rate": 0.00011556412985347942,
"loss": 1.2729585647583008,
"step": 1225
},
{
"epoch": 0.46924177396280403,
"grad_norm": 0.22398388385772705,
"learning_rate": 0.00011495367006637299,
"loss": 1.2588771820068358,
"step": 1230
},
{
"epoch": 0.4711492608488317,
"grad_norm": 0.23689639568328857,
"learning_rate": 0.00011434263972961154,
"loss": 1.209442710876465,
"step": 1235
},
{
"epoch": 0.47305674773485934,
"grad_norm": 0.21210502088069916,
"learning_rate": 0.00011373106215674596,
"loss": 1.253824234008789,
"step": 1240
},
{
"epoch": 0.474964234620887,
"grad_norm": 0.23728817701339722,
"learning_rate": 0.00011311896068220659,
"loss": 1.1961438179016113,
"step": 1245
},
{
"epoch": 0.47687172150691465,
"grad_norm": 0.22245438396930695,
"learning_rate": 0.00011250635866041308,
"loss": 1.1674911499023437,
"step": 1250
},
{
"epoch": 0.4787792083929423,
"grad_norm": 0.1887146681547165,
"learning_rate": 0.0001118932794648831,
"loss": 1.2453363418579102,
"step": 1255
},
{
"epoch": 0.48068669527896996,
"grad_norm": 0.2031584084033966,
"learning_rate": 0.0001112797464873407,
"loss": 1.2116168022155762,
"step": 1260
},
{
"epoch": 0.4825941821649976,
"grad_norm": 0.21245276927947998,
"learning_rate": 0.00011066578313682377,
"loss": 1.2114702224731446,
"step": 1265
},
{
"epoch": 0.48450166905102526,
"grad_norm": 0.200450599193573,
"learning_rate": 0.00011005141283879076,
"loss": 1.2835848808288575,
"step": 1270
},
{
"epoch": 0.4864091559370529,
"grad_norm": 0.19158755242824554,
"learning_rate": 0.00010943665903422713,
"loss": 1.214493751525879,
"step": 1275
},
{
"epoch": 0.48831664282308057,
"grad_norm": 0.20944878458976746,
"learning_rate": 0.00010882154517875074,
"loss": 1.225049877166748,
"step": 1280
},
{
"epoch": 0.4902241297091082,
"grad_norm": 0.17850066721439362,
"learning_rate": 0.00010820609474171704,
"loss": 1.2514646530151368,
"step": 1285
},
{
"epoch": 0.49213161659513593,
"grad_norm": 0.21679556369781494,
"learning_rate": 0.00010759033120532356,
"loss": 1.2022608757019042,
"step": 1290
},
{
"epoch": 0.4940391034811636,
"grad_norm": 0.21753506362438202,
"learning_rate": 0.000106974278063714,
"loss": 1.1957876205444335,
"step": 1295
},
{
"epoch": 0.49594659036719124,
"grad_norm": 0.20135703682899475,
"learning_rate": 0.00010635795882208179,
"loss": 1.2079596519470215,
"step": 1300
},
{
"epoch": 0.4978540772532189,
"grad_norm": 0.2133401483297348,
"learning_rate": 0.00010574139699577332,
"loss": 1.2517601013183595,
"step": 1305
},
{
"epoch": 0.49976156413924655,
"grad_norm": 0.19723205268383026,
"learning_rate": 0.00010512461610939055,
"loss": 1.187291717529297,
"step": 1310
},
{
"epoch": 0.5016690510252741,
"grad_norm": 0.20361660420894623,
"learning_rate": 0.0001045076396958937,
"loss": 1.2871199607849122,
"step": 1315
},
{
"epoch": 0.5035765379113019,
"grad_norm": 0.2710402309894562,
"learning_rate": 0.00010389049129570315,
"loss": 1.246526336669922,
"step": 1320
},
{
"epoch": 0.5054840247973296,
"grad_norm": 0.23143447935581207,
"learning_rate": 0.00010327319445580133,
"loss": 1.2055854797363281,
"step": 1325
},
{
"epoch": 0.5073915116833572,
"grad_norm": 0.20099104940891266,
"learning_rate": 0.0001026557727288344,
"loss": 1.2153850555419923,
"step": 1330
},
{
"epoch": 0.5092989985693849,
"grad_norm": 0.22185488045215607,
"learning_rate": 0.00010203824967221342,
"loss": 1.1709200859069824,
"step": 1335
},
{
"epoch": 0.5112064854554125,
"grad_norm": 0.24854551255702972,
"learning_rate": 0.00010142064884721572,
"loss": 1.2193925857543946,
"step": 1340
},
{
"epoch": 0.5131139723414402,
"grad_norm": 0.18007396161556244,
"learning_rate": 0.00010080299381808578,
"loss": 1.2481110572814942,
"step": 1345
},
{
"epoch": 0.5150214592274678,
"grad_norm": 0.26078617572784424,
"learning_rate": 0.00010018530815113624,
"loss": 1.191064453125,
"step": 1350
},
{
"epoch": 0.5169289461134955,
"grad_norm": 0.1919834464788437,
"learning_rate": 9.956761541384874e-05,
"loss": 1.1143486976623536,
"step": 1355
},
{
"epoch": 0.5188364329995231,
"grad_norm": 0.20681452751159668,
"learning_rate": 9.894993917397456e-05,
"loss": 1.230989646911621,
"step": 1360
},
{
"epoch": 0.5207439198855508,
"grad_norm": 0.1893768161535263,
"learning_rate": 9.833230299863567e-05,
"loss": 1.2528681755065918,
"step": 1365
},
{
"epoch": 0.5226514067715784,
"grad_norm": 0.20343783497810364,
"learning_rate": 9.771473045342535e-05,
"loss": 1.192424201965332,
"step": 1370
},
{
"epoch": 0.5245588936576061,
"grad_norm": 0.22195684909820557,
"learning_rate": 9.709724510150907e-05,
"loss": 1.2481208801269532,
"step": 1375
},
{
"epoch": 0.5264663805436338,
"grad_norm": 0.22458794713020325,
"learning_rate": 9.647987050272555e-05,
"loss": 1.2139442443847657,
"step": 1380
},
{
"epoch": 0.5283738674296614,
"grad_norm": 0.1819061040878296,
"learning_rate": 9.586263021268779e-05,
"loss": 1.1616003036499023,
"step": 1385
},
{
"epoch": 0.5302813543156891,
"grad_norm": 0.19910086691379547,
"learning_rate": 9.524554778188416e-05,
"loss": 1.1937817573547362,
"step": 1390
},
{
"epoch": 0.5321888412017167,
"grad_norm": 0.2040385603904724,
"learning_rate": 9.462864675478015e-05,
"loss": 1.193544578552246,
"step": 1395
},
{
"epoch": 0.5340963280877444,
"grad_norm": 0.23836196959018707,
"learning_rate": 9.401195066891981e-05,
"loss": 1.2564398765563964,
"step": 1400
},
{
"epoch": 0.536003814973772,
"grad_norm": 0.19225947558879852,
"learning_rate": 9.33954830540278e-05,
"loss": 1.2032994270324706,
"step": 1405
},
{
"epoch": 0.5379113018597997,
"grad_norm": 0.21631194651126862,
"learning_rate": 9.277926743111157e-05,
"loss": 1.236619758605957,
"step": 1410
},
{
"epoch": 0.5398187887458273,
"grad_norm": 0.2309265434741974,
"learning_rate": 9.216332731156395e-05,
"loss": 1.185368537902832,
"step": 1415
},
{
"epoch": 0.541726275631855,
"grad_norm": 0.19440126419067383,
"learning_rate": 9.154768619626605e-05,
"loss": 1.2355932235717773,
"step": 1420
},
{
"epoch": 0.5436337625178826,
"grad_norm": 0.16589856147766113,
"learning_rate": 9.093236757469065e-05,
"loss": 1.1948314666748048,
"step": 1425
},
{
"epoch": 0.5455412494039104,
"grad_norm": 0.21594128012657166,
"learning_rate": 9.031739492400594e-05,
"loss": 1.1655755996704102,
"step": 1430
},
{
"epoch": 0.547448736289938,
"grad_norm": 0.24031832814216614,
"learning_rate": 8.970279170817978e-05,
"loss": 1.2017078399658203,
"step": 1435
},
{
"epoch": 0.5493562231759657,
"grad_norm": 0.20147471129894257,
"learning_rate": 8.908858137708444e-05,
"loss": 1.178144931793213,
"step": 1440
},
{
"epoch": 0.5512637100619934,
"grad_norm": 0.21785195171833038,
"learning_rate": 8.847478736560187e-05,
"loss": 1.1745618820190429,
"step": 1445
},
{
"epoch": 0.553171196948021,
"grad_norm": 0.23431022465229034,
"learning_rate": 8.786143309272946e-05,
"loss": 1.153176498413086,
"step": 1450
},
{
"epoch": 0.5550786838340487,
"grad_norm": 0.26836276054382324,
"learning_rate": 8.724854196068675e-05,
"loss": 1.227570152282715,
"step": 1455
},
{
"epoch": 0.5569861707200763,
"grad_norm": 0.21933206915855408,
"learning_rate": 8.66361373540223e-05,
"loss": 1.2512958526611329,
"step": 1460
},
{
"epoch": 0.558893657606104,
"grad_norm": 0.22832703590393066,
"learning_rate": 8.602424263872154e-05,
"loss": 1.2528444290161134,
"step": 1465
},
{
"epoch": 0.5608011444921316,
"grad_norm": 0.23598788678646088,
"learning_rate": 8.54128811613153e-05,
"loss": 1.2174627304077148,
"step": 1470
},
{
"epoch": 0.5627086313781593,
"grad_norm": 0.1953684538602829,
"learning_rate": 8.48020762479889e-05,
"loss": 1.1286799430847168,
"step": 1475
},
{
"epoch": 0.5646161182641869,
"grad_norm": 0.19513633847236633,
"learning_rate": 8.419185120369232e-05,
"loss": 1.1735822677612304,
"step": 1480
},
{
"epoch": 0.5665236051502146,
"grad_norm": 0.20188085734844208,
"learning_rate": 8.358222931125091e-05,
"loss": 1.1577712059020997,
"step": 1485
},
{
"epoch": 0.5684310920362422,
"grad_norm": 0.19986091554164886,
"learning_rate": 8.297323383047708e-05,
"loss": 1.2358123779296875,
"step": 1490
},
{
"epoch": 0.5703385789222699,
"grad_norm": 0.19604338705539703,
"learning_rate": 8.236488799728279e-05,
"loss": 1.220244789123535,
"step": 1495
},
{
"epoch": 0.5722460658082976,
"grad_norm": 0.18865710496902466,
"learning_rate": 8.175721502279313e-05,
"loss": 1.1471840858459472,
"step": 1500
},
{
"epoch": 0.5741535526943252,
"grad_norm": 0.27022239565849304,
"learning_rate": 8.115023809246041e-05,
"loss": 1.1590518951416016,
"step": 1505
},
{
"epoch": 0.5760610395803529,
"grad_norm": 0.2555239796638489,
"learning_rate": 8.054398036517991e-05,
"loss": 1.2529569625854493,
"step": 1510
},
{
"epoch": 0.5779685264663805,
"grad_norm": 0.19677112996578217,
"learning_rate": 7.993846497240603e-05,
"loss": 1.1390647888183594,
"step": 1515
},
{
"epoch": 0.5798760133524082,
"grad_norm": 0.22353534400463104,
"learning_rate": 7.933371501726975e-05,
"loss": 1.142059898376465,
"step": 1520
},
{
"epoch": 0.5817835002384358,
"grad_norm": 0.24535755813121796,
"learning_rate": 7.872975357369719e-05,
"loss": 1.2249109268188476,
"step": 1525
},
{
"epoch": 0.5836909871244635,
"grad_norm": 0.20363497734069824,
"learning_rate": 7.812660368552922e-05,
"loss": 1.186641311645508,
"step": 1530
},
{
"epoch": 0.5855984740104911,
"grad_norm": 0.22370173037052155,
"learning_rate": 7.752428836564221e-05,
"loss": 1.1996564865112305,
"step": 1535
},
{
"epoch": 0.5875059608965189,
"grad_norm": 0.21880581974983215,
"learning_rate": 7.692283059506998e-05,
"loss": 1.1891421318054198,
"step": 1540
},
{
"epoch": 0.5894134477825465,
"grad_norm": 0.22140686213970184,
"learning_rate": 7.632225332212707e-05,
"loss": 1.1702460289001464,
"step": 1545
},
{
"epoch": 0.5913209346685742,
"grad_norm": 0.1957419216632843,
"learning_rate": 7.5722579461533e-05,
"loss": 1.108344268798828,
"step": 1550
},
{
"epoch": 0.5932284215546018,
"grad_norm": 0.22176416218280792,
"learning_rate": 7.512383189353817e-05,
"loss": 1.1665712356567384,
"step": 1555
},
{
"epoch": 0.5951359084406295,
"grad_norm": 0.22830532491207123,
"learning_rate": 7.452603346305054e-05,
"loss": 1.1701374053955078,
"step": 1560
},
{
"epoch": 0.5970433953266572,
"grad_norm": 0.26137056946754456,
"learning_rate": 7.392920697876443e-05,
"loss": 1.1804938316345215,
"step": 1565
},
{
"epoch": 0.5989508822126848,
"grad_norm": 0.21527014672756195,
"learning_rate": 7.333337521228996e-05,
"loss": 1.14210205078125,
"step": 1570
},
{
"epoch": 0.6008583690987125,
"grad_norm": 0.2155296504497528,
"learning_rate": 7.273856089728429e-05,
"loss": 1.2072952270507813,
"step": 1575
},
{
"epoch": 0.6027658559847401,
"grad_norm": 0.20168620347976685,
"learning_rate": 7.214478672858428e-05,
"loss": 1.1606990814208984,
"step": 1580
},
{
"epoch": 0.6046733428707678,
"grad_norm": 0.20272770524024963,
"learning_rate": 7.155207536134054e-05,
"loss": 1.1442641258239745,
"step": 1585
},
{
"epoch": 0.6065808297567954,
"grad_norm": 0.22614730894565582,
"learning_rate": 7.09604494101529e-05,
"loss": 1.1707661628723145,
"step": 1590
},
{
"epoch": 0.6084883166428231,
"grad_norm": 0.22346791625022888,
"learning_rate": 7.036993144820787e-05,
"loss": 1.1845563888549804,
"step": 1595
},
{
"epoch": 0.6103958035288507,
"grad_norm": 0.24826298654079437,
"learning_rate": 6.978054400641713e-05,
"loss": 1.1342114448547362,
"step": 1600
},
{
"epoch": 0.6123032904148784,
"grad_norm": 0.21613281965255737,
"learning_rate": 6.919230957255791e-05,
"loss": 1.1967657089233399,
"step": 1605
},
{
"epoch": 0.614210777300906,
"grad_norm": 0.21833890676498413,
"learning_rate": 6.860525059041501e-05,
"loss": 1.1509429931640625,
"step": 1610
},
{
"epoch": 0.6161182641869337,
"grad_norm": 0.2539721727371216,
"learning_rate": 6.80193894589246e-05,
"loss": 1.15333890914917,
"step": 1615
},
{
"epoch": 0.6180257510729614,
"grad_norm": 0.22569319605827332,
"learning_rate": 6.743474853131925e-05,
"loss": 1.1677268981933593,
"step": 1620
},
{
"epoch": 0.619933237958989,
"grad_norm": 0.2163088172674179,
"learning_rate": 6.685135011427541e-05,
"loss": 1.1319167137145996,
"step": 1625
},
{
"epoch": 0.6218407248450167,
"grad_norm": 0.24095430970191956,
"learning_rate": 6.626921646706218e-05,
"loss": 1.1875012397766114,
"step": 1630
},
{
"epoch": 0.6237482117310443,
"grad_norm": 0.21890592575073242,
"learning_rate": 6.568836980069198e-05,
"loss": 1.0996828079223633,
"step": 1635
},
{
"epoch": 0.625655698617072,
"grad_norm": 0.20873771607875824,
"learning_rate": 6.510883227707312e-05,
"loss": 1.1962814331054688,
"step": 1640
},
{
"epoch": 0.6275631855030996,
"grad_norm": 0.21107596158981323,
"learning_rate": 6.453062600816426e-05,
"loss": 1.13582124710083,
"step": 1645
},
{
"epoch": 0.6294706723891274,
"grad_norm": 0.22814150154590607,
"learning_rate": 6.395377305513067e-05,
"loss": 1.1759732246398926,
"step": 1650
},
{
"epoch": 0.631378159275155,
"grad_norm": 0.22088365256786346,
"learning_rate": 6.337829542750263e-05,
"loss": 1.1874792098999023,
"step": 1655
},
{
"epoch": 0.6332856461611827,
"grad_norm": 0.2006022334098816,
"learning_rate": 6.280421508233548e-05,
"loss": 1.1828422546386719,
"step": 1660
},
{
"epoch": 0.6351931330472103,
"grad_norm": 0.19948694109916687,
"learning_rate": 6.223155392337204e-05,
"loss": 1.166551113128662,
"step": 1665
},
{
"epoch": 0.637100619933238,
"grad_norm": 0.2671700716018677,
"learning_rate": 6.16603338002068e-05,
"loss": 1.1790518760681152,
"step": 1670
},
{
"epoch": 0.6390081068192656,
"grad_norm": 0.2384413778781891,
"learning_rate": 6.109057650745216e-05,
"loss": 1.159766960144043,
"step": 1675
},
{
"epoch": 0.6409155937052933,
"grad_norm": 0.2287103533744812,
"learning_rate": 6.0522303783907095e-05,
"loss": 1.1437957763671875,
"step": 1680
},
{
"epoch": 0.642823080591321,
"grad_norm": 0.21127451956272125,
"learning_rate": 5.995553731172756e-05,
"loss": 1.2146809577941895,
"step": 1685
},
{
"epoch": 0.6447305674773486,
"grad_norm": 0.2115635722875595,
"learning_rate": 5.9390298715599245e-05,
"loss": 1.166009521484375,
"step": 1690
},
{
"epoch": 0.6466380543633763,
"grad_norm": 0.2387341558933258,
"learning_rate": 5.8826609561912504e-05,
"loss": 1.1473962783813476,
"step": 1695
},
{
"epoch": 0.6485455412494039,
"grad_norm": 0.2259090095758438,
"learning_rate": 5.826449135793957e-05,
"loss": 1.1924949645996095,
"step": 1700
},
{
"epoch": 0.6504530281354316,
"grad_norm": 0.1968328058719635,
"learning_rate": 5.7703965551013736e-05,
"loss": 1.1942377090454102,
"step": 1705
},
{
"epoch": 0.6523605150214592,
"grad_norm": 0.29217374324798584,
"learning_rate": 5.714505352771136e-05,
"loss": 1.1619529724121094,
"step": 1710
},
{
"epoch": 0.6542680019074869,
"grad_norm": 0.21156348288059235,
"learning_rate": 5.6587776613035625e-05,
"loss": 1.1163355827331543,
"step": 1715
},
{
"epoch": 0.6561754887935145,
"grad_norm": 0.23455148935317993,
"learning_rate": 5.6032156069603016e-05,
"loss": 1.1744614601135255,
"step": 1720
},
{
"epoch": 0.6580829756795422,
"grad_norm": 0.21962733566761017,
"learning_rate": 5.547821309683199e-05,
"loss": 1.217427635192871,
"step": 1725
},
{
"epoch": 0.6599904625655698,
"grad_norm": 0.2295837551355362,
"learning_rate": 5.49259688301342e-05,
"loss": 1.178595733642578,
"step": 1730
},
{
"epoch": 0.6618979494515975,
"grad_norm": 0.20390628278255463,
"learning_rate": 5.4375444340107894e-05,
"loss": 1.1660377502441406,
"step": 1735
},
{
"epoch": 0.6638054363376252,
"grad_norm": 0.26694226264953613,
"learning_rate": 5.382666063173434e-05,
"loss": 1.1281030654907227,
"step": 1740
},
{
"epoch": 0.6657129232236528,
"grad_norm": 0.2512981593608856,
"learning_rate": 5.3279638643575994e-05,
"loss": 1.1735931396484376,
"step": 1745
},
{
"epoch": 0.6676204101096805,
"grad_norm": 0.24102783203125,
"learning_rate": 5.273439924697783e-05,
"loss": 1.1657618522644042,
"step": 1750
},
{
"epoch": 0.6695278969957081,
"grad_norm": 0.26361793279647827,
"learning_rate": 5.219096324527099e-05,
"loss": 1.1728330612182618,
"step": 1755
},
{
"epoch": 0.6714353838817358,
"grad_norm": 0.21043244004249573,
"learning_rate": 5.164935137297897e-05,
"loss": 1.1257946968078614,
"step": 1760
},
{
"epoch": 0.6733428707677634,
"grad_norm": 0.22392340004444122,
"learning_rate": 5.110958429502647e-05,
"loss": 1.1028033256530763,
"step": 1765
},
{
"epoch": 0.6752503576537912,
"grad_norm": 0.2338101863861084,
"learning_rate": 5.057168260595111e-05,
"loss": 1.2020700454711915,
"step": 1770
},
{
"epoch": 0.6771578445398188,
"grad_norm": 0.2419530749320984,
"learning_rate": 5.003566682911751e-05,
"loss": 1.1887866973876953,
"step": 1775
},
{
"epoch": 0.6790653314258465,
"grad_norm": 0.20150014758110046,
"learning_rate": 4.950155741593425e-05,
"loss": 1.1544455528259276,
"step": 1780
},
{
"epoch": 0.6809728183118741,
"grad_norm": 0.19747178256511688,
"learning_rate": 4.896937474507366e-05,
"loss": 1.1654293060302734,
"step": 1785
},
{
"epoch": 0.6828803051979018,
"grad_norm": 0.2741665542125702,
"learning_rate": 4.843913912169403e-05,
"loss": 1.1489416122436524,
"step": 1790
},
{
"epoch": 0.6847877920839294,
"grad_norm": 0.22320930659770966,
"learning_rate": 4.7910870776665204e-05,
"loss": 1.1555864334106445,
"step": 1795
},
{
"epoch": 0.6866952789699571,
"grad_norm": 0.2183150053024292,
"learning_rate": 4.738458986579647e-05,
"loss": 1.1381507873535157,
"step": 1800
},
{
"epoch": 0.6886027658559848,
"grad_norm": 0.21290890872478485,
"learning_rate": 4.686031646906757e-05,
"loss": 1.1559734344482422,
"step": 1805
},
{
"epoch": 0.6905102527420124,
"grad_norm": 0.18775756657123566,
"learning_rate": 4.633807058986258e-05,
"loss": 1.1064885139465332,
"step": 1810
},
{
"epoch": 0.6924177396280401,
"grad_norm": 0.1854240745306015,
"learning_rate": 4.581787215420669e-05,
"loss": 1.2117759704589843,
"step": 1815
},
{
"epoch": 0.6943252265140677,
"grad_norm": 0.21344280242919922,
"learning_rate": 4.529974101000589e-05,
"loss": 1.1865493774414062,
"step": 1820
},
{
"epoch": 0.6962327134000954,
"grad_norm": 0.21992447972297668,
"learning_rate": 4.478369692628975e-05,
"loss": 1.1341306686401367,
"step": 1825
},
{
"epoch": 0.698140200286123,
"grad_norm": 0.25210610032081604,
"learning_rate": 4.4269759592457116e-05,
"loss": 1.180692481994629,
"step": 1830
},
{
"epoch": 0.7000476871721507,
"grad_norm": 0.2453182190656662,
"learning_rate": 4.3757948617524825e-05,
"loss": 1.185411262512207,
"step": 1835
},
{
"epoch": 0.7019551740581783,
"grad_norm": 0.2249467670917511,
"learning_rate": 4.324828352937962e-05,
"loss": 1.1835633277893067,
"step": 1840
},
{
"epoch": 0.703862660944206,
"grad_norm": 0.21584834158420563,
"learning_rate": 4.274078377403304e-05,
"loss": 1.1303343772888184,
"step": 1845
},
{
"epoch": 0.7057701478302336,
"grad_norm": 0.21426378190517426,
"learning_rate": 4.2235468714879366e-05,
"loss": 1.0878255844116211,
"step": 1850
},
{
"epoch": 0.7076776347162613,
"grad_norm": 0.22179460525512695,
"learning_rate": 4.173235763195699e-05,
"loss": 1.2107229232788086,
"step": 1855
},
{
"epoch": 0.709585121602289,
"grad_norm": 0.18949757516384125,
"learning_rate": 4.12314697212127e-05,
"loss": 1.1272903442382813,
"step": 1860
},
{
"epoch": 0.7114926084883166,
"grad_norm": 0.23181162774562836,
"learning_rate": 4.0732824093769274e-05,
"loss": 1.1631418228149415,
"step": 1865
},
{
"epoch": 0.7134000953743443,
"grad_norm": 0.23257216811180115,
"learning_rate": 4.023643977519637e-05,
"loss": 1.1605160713195801,
"step": 1870
},
{
"epoch": 0.7153075822603719,
"grad_norm": 0.21041543781757355,
"learning_rate": 3.974233570478442e-05,
"loss": 1.1026935577392578,
"step": 1875
},
{
"epoch": 0.7172150691463997,
"grad_norm": 0.20009055733680725,
"learning_rate": 3.9250530734822266e-05,
"loss": 1.1321779251098634,
"step": 1880
},
{
"epoch": 0.7191225560324273,
"grad_norm": 0.20863768458366394,
"learning_rate": 3.8761043629877736e-05,
"loss": 1.1488612174987793,
"step": 1885
},
{
"epoch": 0.721030042918455,
"grad_norm": 0.22087125480175018,
"learning_rate": 3.8273893066081676e-05,
"loss": 1.1405168533325196,
"step": 1890
},
{
"epoch": 0.7229375298044826,
"grad_norm": 0.21345745027065277,
"learning_rate": 3.7789097630415396e-05,
"loss": 1.0503695487976075,
"step": 1895
},
{
"epoch": 0.7248450166905103,
"grad_norm": 0.20958393812179565,
"learning_rate": 3.73066758200015e-05,
"loss": 1.0739940643310546,
"step": 1900
},
{
"epoch": 0.7267525035765379,
"grad_norm": 0.20313677191734314,
"learning_rate": 3.682664604139813e-05,
"loss": 1.1517399787902831,
"step": 1905
},
{
"epoch": 0.7286599904625656,
"grad_norm": 0.2925751507282257,
"learning_rate": 3.634902660989667e-05,
"loss": 1.136181640625,
"step": 1910
},
{
"epoch": 0.7305674773485932,
"grad_norm": 0.2331283688545227,
"learning_rate": 3.587383574882296e-05,
"loss": 1.1208656311035157,
"step": 1915
},
{
"epoch": 0.7324749642346209,
"grad_norm": 0.23703224956989288,
"learning_rate": 3.540109158884193e-05,
"loss": 1.1330846786499023,
"step": 1920
},
{
"epoch": 0.7343824511206486,
"grad_norm": 0.1794946938753128,
"learning_rate": 3.493081216726593e-05,
"loss": 1.1613752365112304,
"step": 1925
},
{
"epoch": 0.7362899380066762,
"grad_norm": 0.21329441666603088,
"learning_rate": 3.446301542736646e-05,
"loss": 1.1352320671081544,
"step": 1930
},
{
"epoch": 0.7381974248927039,
"grad_norm": 0.22618280351161957,
"learning_rate": 3.399771921768951e-05,
"loss": 1.1310253143310547,
"step": 1935
},
{
"epoch": 0.7401049117787315,
"grad_norm": 0.18707576394081116,
"learning_rate": 3.353494129137468e-05,
"loss": 1.1597959518432617,
"step": 1940
},
{
"epoch": 0.7420123986647592,
"grad_norm": 0.2145681381225586,
"learning_rate": 3.307469930547772e-05,
"loss": 1.1611350059509278,
"step": 1945
},
{
"epoch": 0.7439198855507868,
"grad_norm": 0.2475588023662567,
"learning_rate": 3.26170108202969e-05,
"loss": 1.1909062385559082,
"step": 1950
},
{
"epoch": 0.7458273724368145,
"grad_norm": 0.21284571290016174,
"learning_rate": 3.216189329870295e-05,
"loss": 1.1203042030334474,
"step": 1955
},
{
"epoch": 0.7477348593228421,
"grad_norm": 0.21369685232639313,
"learning_rate": 3.1709364105472814e-05,
"loss": 1.0986623764038086,
"step": 1960
},
{
"epoch": 0.7496423462088698,
"grad_norm": 0.25102755427360535,
"learning_rate": 3.125944050662701e-05,
"loss": 1.148338508605957,
"step": 1965
},
{
"epoch": 0.7515498330948974,
"grad_norm": 0.22295311093330383,
"learning_rate": 3.081213966877101e-05,
"loss": 1.1612619400024413,
"step": 1970
},
{
"epoch": 0.7534573199809251,
"grad_norm": 0.22581028938293457,
"learning_rate": 3.0367478658440198e-05,
"loss": 1.165948486328125,
"step": 1975
},
{
"epoch": 0.7553648068669528,
"grad_norm": 0.19900238513946533,
"learning_rate": 2.992547444144863e-05,
"loss": 1.1082708358764648,
"step": 1980
},
{
"epoch": 0.7572722937529804,
"grad_norm": 0.20476290583610535,
"learning_rate": 2.9486143882241803e-05,
"loss": 1.1167781829833985,
"step": 1985
},
{
"epoch": 0.7591797806390082,
"grad_norm": 0.25547683238983154,
"learning_rate": 2.9049503743253093e-05,
"loss": 1.1323695182800293,
"step": 1990
},
{
"epoch": 0.7610872675250357,
"grad_norm": 0.2212236076593399,
"learning_rate": 2.8615570684264347e-05,
"loss": 1.160726547241211,
"step": 1995
},
{
"epoch": 0.7629947544110635,
"grad_norm": 0.25291791558265686,
"learning_rate": 2.818436126177012e-05,
"loss": 1.2163751602172852,
"step": 2000
},
{
"epoch": 0.7649022412970911,
"grad_norm": 0.21982431411743164,
"learning_rate": 2.775589192834601e-05,
"loss": 1.1515751838684083,
"step": 2005
},
{
"epoch": 0.7668097281831188,
"grad_norm": 0.24241040647029877,
"learning_rate": 2.733017903202092e-05,
"loss": 1.1150344848632812,
"step": 2010
},
{
"epoch": 0.7687172150691464,
"grad_norm": 0.2245641052722931,
"learning_rate": 2.6907238815653324e-05,
"loss": 1.1793850898742675,
"step": 2015
},
{
"epoch": 0.7706247019551741,
"grad_norm": 0.22021226584911346,
"learning_rate": 2.648708741631142e-05,
"loss": 1.0567512512207031,
"step": 2020
},
{
"epoch": 0.7725321888412017,
"grad_norm": 0.20851437747478485,
"learning_rate": 2.6069740864657587e-05,
"loss": 1.155423641204834,
"step": 2025
},
{
"epoch": 0.7744396757272294,
"grad_norm": 0.23057712614536285,
"learning_rate": 2.565521508433668e-05,
"loss": 1.1357767105102539,
"step": 2030
},
{
"epoch": 0.776347162613257,
"grad_norm": 0.22620734572410583,
"learning_rate": 2.5243525891368435e-05,
"loss": 1.137866497039795,
"step": 2035
},
{
"epoch": 0.7782546494992847,
"grad_norm": 0.23022662103176117,
"learning_rate": 2.483468899354404e-05,
"loss": 1.1148431777954102,
"step": 2040
},
{
"epoch": 0.7801621363853124,
"grad_norm": 0.20083515346050262,
"learning_rate": 2.4428719989826833e-05,
"loss": 1.094567108154297,
"step": 2045
},
{
"epoch": 0.78206962327134,
"grad_norm": 0.24217429757118225,
"learning_rate": 2.4025634369757087e-05,
"loss": 1.1915393829345704,
"step": 2050
},
{
"epoch": 0.7839771101573677,
"grad_norm": 0.2980445623397827,
"learning_rate": 2.3625447512861086e-05,
"loss": 1.1448716163635253,
"step": 2055
},
{
"epoch": 0.7858845970433953,
"grad_norm": 0.25267890095710754,
"learning_rate": 2.322817468806425e-05,
"loss": 1.0882713317871093,
"step": 2060
},
{
"epoch": 0.787792083929423,
"grad_norm": 0.2755645513534546,
"learning_rate": 2.2833831053108613e-05,
"loss": 1.1582210540771485,
"step": 2065
},
{
"epoch": 0.7896995708154506,
"grad_norm": 0.24219442903995514,
"learning_rate": 2.244243165397443e-05,
"loss": 1.1399827957153321,
"step": 2070
},
{
"epoch": 0.7916070577014783,
"grad_norm": 0.21595719456672668,
"learning_rate": 2.2053991424306198e-05,
"loss": 1.1543403625488282,
"step": 2075
},
{
"epoch": 0.7935145445875059,
"grad_norm": 0.22069472074508667,
"learning_rate": 2.1668525184842714e-05,
"loss": 1.1744007110595702,
"step": 2080
},
{
"epoch": 0.7954220314735336,
"grad_norm": 0.2289930284023285,
"learning_rate": 2.1286047642851804e-05,
"loss": 1.1891432762145997,
"step": 2085
},
{
"epoch": 0.7973295183595612,
"grad_norm": 0.2365955412387848,
"learning_rate": 2.0906573391569007e-05,
"loss": 1.1245343208312988,
"step": 2090
},
{
"epoch": 0.7992370052455889,
"grad_norm": 0.25190407037734985,
"learning_rate": 2.053011690964084e-05,
"loss": 1.1232264518737793,
"step": 2095
},
{
"epoch": 0.8011444921316166,
"grad_norm": 0.22288674116134644,
"learning_rate": 2.015669256057242e-05,
"loss": 1.154467487335205,
"step": 2100
},
{
"epoch": 0.8030519790176442,
"grad_norm": 0.20601993799209595,
"learning_rate": 1.978631459217929e-05,
"loss": 1.102585506439209,
"step": 2105
},
{
"epoch": 0.804959465903672,
"grad_norm": 0.25253918766975403,
"learning_rate": 1.9418997136043947e-05,
"loss": 1.1514110565185547,
"step": 2110
},
{
"epoch": 0.8068669527896996,
"grad_norm": 0.21954931318759918,
"learning_rate": 1.905475420697661e-05,
"loss": 1.158360195159912,
"step": 2115
},
{
"epoch": 0.8087744396757273,
"grad_norm": 0.19882351160049438,
"learning_rate": 1.869359970248048e-05,
"loss": 1.1656829833984375,
"step": 2120
},
{
"epoch": 0.8106819265617549,
"grad_norm": 0.238766148686409,
"learning_rate": 1.8335547402221485e-05,
"loss": 1.176176166534424,
"step": 2125
},
{
"epoch": 0.8125894134477826,
"grad_norm": 0.20012234151363373,
"learning_rate": 1.798061096750252e-05,
"loss": 1.0627614974975585,
"step": 2130
},
{
"epoch": 0.8144969003338102,
"grad_norm": 0.23309870064258575,
"learning_rate": 1.7628803940742257e-05,
"loss": 1.1396465301513672,
"step": 2135
},
{
"epoch": 0.8164043872198379,
"grad_norm": 0.2519659698009491,
"learning_rate": 1.7280139744958356e-05,
"loss": 1.1535280227661133,
"step": 2140
},
{
"epoch": 0.8183118741058655,
"grad_norm": 0.19456864893436432,
"learning_rate": 1.693463168325542e-05,
"loss": 1.1232088088989258,
"step": 2145
},
{
"epoch": 0.8202193609918932,
"grad_norm": 0.21354719996452332,
"learning_rate": 1.65922929383173e-05,
"loss": 1.1516935348510742,
"step": 2150
},
{
"epoch": 0.8221268478779208,
"grad_norm": 0.20758101344108582,
"learning_rate": 1.6253136571904236e-05,
"loss": 1.1620672225952149,
"step": 2155
},
{
"epoch": 0.8240343347639485,
"grad_norm": 0.19862081110477448,
"learning_rate": 1.5917175524354444e-05,
"loss": 1.0714295387268067,
"step": 2160
},
{
"epoch": 0.8259418216499762,
"grad_norm": 0.21623754501342773,
"learning_rate": 1.55844226140903e-05,
"loss": 1.1215152740478516,
"step": 2165
},
{
"epoch": 0.8278493085360038,
"grad_norm": 0.25673818588256836,
"learning_rate": 1.5254890537129417e-05,
"loss": 1.1280359268188476,
"step": 2170
},
{
"epoch": 0.8297567954220315,
"grad_norm": 0.2876787483692169,
"learning_rate": 1.4928591866600128e-05,
"loss": 1.0949363708496094,
"step": 2175
},
{
"epoch": 0.8316642823080591,
"grad_norm": 0.216287299990654,
"learning_rate": 1.4605539052261797e-05,
"loss": 1.041240119934082,
"step": 2180
},
{
"epoch": 0.8335717691940868,
"grad_norm": 0.21410544216632843,
"learning_rate": 1.4285744420029811e-05,
"loss": 1.135991096496582,
"step": 2185
},
{
"epoch": 0.8354792560801144,
"grad_norm": 0.2441902458667755,
"learning_rate": 1.3969220171505249e-05,
"loss": 1.098809814453125,
"step": 2190
},
{
"epoch": 0.8373867429661421,
"grad_norm": 0.2370656281709671,
"learning_rate": 1.3655978383509382e-05,
"loss": 1.0981058120727538,
"step": 2195
},
{
"epoch": 0.8392942298521697,
"grad_norm": 0.2267371267080307,
"learning_rate": 1.3346031007622862e-05,
"loss": 1.1900452613830566,
"step": 2200
},
{
"epoch": 0.8412017167381974,
"grad_norm": 0.2336042821407318,
"learning_rate": 1.303938986972979e-05,
"loss": 1.0903155326843261,
"step": 2205
},
{
"epoch": 0.843109203624225,
"grad_norm": 0.23731689155101776,
"learning_rate": 1.2736066669566382e-05,
"loss": 1.0552220344543457,
"step": 2210
},
{
"epoch": 0.8450166905102527,
"grad_norm": 0.2219412922859192,
"learning_rate": 1.243607298027466e-05,
"loss": 1.0662223815917968,
"step": 2215
},
{
"epoch": 0.8469241773962805,
"grad_norm": 0.2387835681438446,
"learning_rate": 1.2139420247960798e-05,
"loss": 1.1238874435424804,
"step": 2220
},
{
"epoch": 0.848831664282308,
"grad_norm": 0.2534177899360657,
"learning_rate": 1.1846119791258515e-05,
"loss": 1.097713088989258,
"step": 2225
},
{
"epoch": 0.8507391511683358,
"grad_norm": 0.23023399710655212,
"learning_rate": 1.1556182800897175e-05,
"loss": 1.1359582901000977,
"step": 2230
},
{
"epoch": 0.8526466380543634,
"grad_norm": 0.2506740689277649,
"learning_rate": 1.1269620339274767e-05,
"loss": 1.1142709732055665,
"step": 2235
},
{
"epoch": 0.8545541249403911,
"grad_norm": 0.21922656893730164,
"learning_rate": 1.0986443340035867e-05,
"loss": 1.0912172317504882,
"step": 2240
},
{
"epoch": 0.8564616118264187,
"grad_norm": 0.25788071751594543,
"learning_rate": 1.0706662607654484e-05,
"loss": 1.1749443054199218,
"step": 2245
},
{
"epoch": 0.8583690987124464,
"grad_norm": 0.23726382851600647,
"learning_rate": 1.0430288817021728e-05,
"loss": 1.0870184898376465,
"step": 2250
},
{
"epoch": 0.860276585598474,
"grad_norm": 0.22818847000598907,
"learning_rate": 1.0157332513038653e-05,
"loss": 1.1289539337158203,
"step": 2255
},
{
"epoch": 0.8621840724845017,
"grad_norm": 0.23203957080841064,
"learning_rate": 9.887804110213828e-06,
"loss": 1.1057024955749513,
"step": 2260
},
{
"epoch": 0.8640915593705293,
"grad_norm": 0.22027398645877838,
"learning_rate": 9.621713892266016e-06,
"loss": 1.0933302879333495,
"step": 2265
},
{
"epoch": 0.865999046256557,
"grad_norm": 0.2558651864528656,
"learning_rate": 9.359072011731762e-06,
"loss": 1.247685432434082,
"step": 2270
},
{
"epoch": 0.8679065331425846,
"grad_norm": 0.22376199066638947,
"learning_rate": 9.099888489578079e-06,
"loss": 1.1251508712768554,
"step": 2275
},
{
"epoch": 0.8698140200286123,
"grad_norm": 0.20759578049182892,
"learning_rate": 8.844173214820061e-06,
"loss": 1.1414949417114257,
"step": 2280
},
{
"epoch": 0.87172150691464,
"grad_norm": 0.2659814655780792,
"learning_rate": 8.591935944143615e-06,
"loss": 1.1263145446777343,
"step": 2285
},
{
"epoch": 0.8736289938006676,
"grad_norm": 0.2113100290298462,
"learning_rate": 8.343186301533157e-06,
"loss": 1.133929443359375,
"step": 2290
},
{
"epoch": 0.8755364806866953,
"grad_norm": 0.2152625471353531,
"learning_rate": 8.097933777904431e-06,
"loss": 1.1101900100708009,
"step": 2295
},
{
"epoch": 0.8774439675727229,
"grad_norm": 0.2583655118942261,
"learning_rate": 7.856187730742438e-06,
"loss": 1.1228748321533204,
"step": 2300
},
{
"epoch": 0.8793514544587506,
"grad_norm": 0.20810461044311523,
"learning_rate": 7.617957383744245e-06,
"loss": 1.1034461975097656,
"step": 2305
},
{
"epoch": 0.8812589413447782,
"grad_norm": 0.21229910850524902,
"learning_rate": 7.383251826467297e-06,
"loss": 1.1300959587097168,
"step": 2310
},
{
"epoch": 0.8831664282308059,
"grad_norm": 0.26863279938697815,
"learning_rate": 7.152080013982421e-06,
"loss": 1.1783254623413086,
"step": 2315
},
{
"epoch": 0.8850739151168335,
"grad_norm": 0.19874270260334015,
"learning_rate": 6.924450766532209e-06,
"loss": 1.0705080032348633,
"step": 2320
},
{
"epoch": 0.8869814020028612,
"grad_norm": 0.21158960461616516,
"learning_rate": 6.70037276919453e-06,
"loss": 1.1780382156372071,
"step": 2325
},
{
"epoch": 0.8888888888888888,
"grad_norm": 0.20735962688922882,
"learning_rate": 6.4798545715510916e-06,
"loss": 1.1040769577026368,
"step": 2330
},
{
"epoch": 0.8907963757749165,
"grad_norm": 0.18355117738246918,
"learning_rate": 6.2629045873612405e-06,
"loss": 1.0807653427124024,
"step": 2335
},
{
"epoch": 0.8927038626609443,
"grad_norm": 0.23413380980491638,
"learning_rate": 6.04953109424099e-06,
"loss": 1.1116905212402344,
"step": 2340
},
{
"epoch": 0.8946113495469719,
"grad_norm": 0.2297702431678772,
"learning_rate": 5.839742233347156e-06,
"loss": 1.1403563499450684,
"step": 2345
},
{
"epoch": 0.8965188364329996,
"grad_norm": 0.22440911829471588,
"learning_rate": 5.633546009066715e-06,
"loss": 1.1003211975097655,
"step": 2350
},
{
"epoch": 0.8984263233190272,
"grad_norm": 0.27772337198257446,
"learning_rate": 5.430950288711467e-06,
"loss": 1.148040199279785,
"step": 2355
},
{
"epoch": 0.9003338102050549,
"grad_norm": 0.21390412747859955,
"learning_rate": 5.231962802217783e-06,
"loss": 1.1419692993164063,
"step": 2360
},
{
"epoch": 0.9022412970910825,
"grad_norm": 0.21749059855937958,
"learning_rate": 5.036591141851732e-06,
"loss": 1.1392088890075684,
"step": 2365
},
{
"epoch": 0.9041487839771102,
"grad_norm": 0.2439197450876236,
"learning_rate": 4.844842761919388e-06,
"loss": 1.1405391693115234,
"step": 2370
},
{
"epoch": 0.9060562708631378,
"grad_norm": 0.2308351844549179,
"learning_rate": 4.656724978482374e-06,
"loss": 1.1562992095947267,
"step": 2375
},
{
"epoch": 0.9079637577491655,
"grad_norm": 0.23048758506774902,
"learning_rate": 4.472244969078787e-06,
"loss": 1.1268766403198243,
"step": 2380
},
{
"epoch": 0.9098712446351931,
"grad_norm": 0.27241912484169006,
"learning_rate": 4.2914097724492955e-06,
"loss": 1.157505989074707,
"step": 2385
},
{
"epoch": 0.9117787315212208,
"grad_norm": 0.23174434900283813,
"learning_rate": 4.11422628826863e-06,
"loss": 1.1393831253051758,
"step": 2390
},
{
"epoch": 0.9136862184072484,
"grad_norm": 0.22239695489406586,
"learning_rate": 3.940701276882231e-06,
"loss": 1.1609222412109375,
"step": 2395
},
{
"epoch": 0.9155937052932761,
"grad_norm": 0.26512646675109863,
"learning_rate": 3.770841359048416e-06,
"loss": 1.1378469467163086,
"step": 2400
},
{
"epoch": 0.9175011921793038,
"grad_norm": 0.21327486634254456,
"learning_rate": 3.6046530156857218e-06,
"loss": 1.1494600296020507,
"step": 2405
},
{
"epoch": 0.9194086790653314,
"grad_norm": 0.22115978598594666,
"learning_rate": 3.4421425876256406e-06,
"loss": 1.0902729034423828,
"step": 2410
},
{
"epoch": 0.9213161659513591,
"grad_norm": 0.22342635691165924,
"learning_rate": 3.2833162753706957e-06,
"loss": 1.1097753524780274,
"step": 2415
},
{
"epoch": 0.9232236528373867,
"grad_norm": 0.2765384018421173,
"learning_rate": 3.1281801388577836e-06,
"loss": 1.1319202423095702,
"step": 2420
},
{
"epoch": 0.9251311397234144,
"grad_norm": 0.2197486013174057,
"learning_rate": 2.9767400972271042e-06,
"loss": 1.176529884338379,
"step": 2425
},
{
"epoch": 0.927038626609442,
"grad_norm": 0.23287062346935272,
"learning_rate": 2.829001928596209e-06,
"loss": 1.1135154724121095,
"step": 2430
},
{
"epoch": 0.9289461134954697,
"grad_norm": 0.266107976436615,
"learning_rate": 2.6849712698395536e-06,
"loss": 1.0989619255065919,
"step": 2435
},
{
"epoch": 0.9308536003814973,
"grad_norm": 0.2772996723651886,
"learning_rate": 2.5446536163734935e-06,
"loss": 1.2017412185668945,
"step": 2440
},
{
"epoch": 0.932761087267525,
"grad_norm": 0.1954774707555771,
"learning_rate": 2.4080543219465403e-06,
"loss": 1.0754703521728515,
"step": 2445
},
{
"epoch": 0.9346685741535526,
"grad_norm": 0.20997653901576996,
"learning_rate": 2.2751785984350792e-06,
"loss": 1.1754121780395508,
"step": 2450
},
{
"epoch": 0.9365760610395804,
"grad_norm": 0.23438486456871033,
"learning_rate": 2.1460315156445975e-06,
"loss": 1.1126022338867188,
"step": 2455
},
{
"epoch": 0.9384835479256081,
"grad_norm": 0.21886298060417175,
"learning_rate": 2.020618001116159e-06,
"loss": 1.1078985214233399,
"step": 2460
},
{
"epoch": 0.9403910348116357,
"grad_norm": 0.20734335482120514,
"learning_rate": 1.8989428399384558e-06,
"loss": 1.130316162109375,
"step": 2465
},
{
"epoch": 0.9422985216976634,
"grad_norm": 0.2098223716020584,
"learning_rate": 1.781010674565209e-06,
"loss": 1.065493392944336,
"step": 2470
},
{
"epoch": 0.944206008583691,
"grad_norm": 0.19588448107242584,
"learning_rate": 1.6668260046380447e-06,
"loss": 1.1139305114746094,
"step": 2475
},
{
"epoch": 0.9461134954697187,
"grad_norm": 0.2644004225730896,
"learning_rate": 1.5563931868148085e-06,
"loss": 1.109051513671875,
"step": 2480
},
{
"epoch": 0.9480209823557463,
"grad_norm": 0.24326738715171814,
"learning_rate": 1.4497164346033432e-06,
"loss": 1.139406108856201,
"step": 2485
},
{
"epoch": 0.949928469241774,
"grad_norm": 0.3140820860862732,
"learning_rate": 1.3467998182007169e-06,
"loss": 1.1468605995178223,
"step": 2490
},
{
"epoch": 0.9518359561278016,
"grad_norm": 0.23210427165031433,
"learning_rate": 1.2476472643379478e-06,
"loss": 1.1381516456604004,
"step": 2495
},
{
"epoch": 0.9537434430138293,
"grad_norm": 0.24174664914608002,
"learning_rate": 1.152262556130168e-06,
"loss": 1.1345512390136718,
"step": 2500
},
{
"epoch": 0.9556509298998569,
"grad_norm": 0.20972232520580292,
"learning_rate": 1.060649332932262e-06,
"loss": 1.1098050117492675,
"step": 2505
},
{
"epoch": 0.9575584167858846,
"grad_norm": 0.18688908219337463,
"learning_rate": 9.728110902000432e-07,
"loss": 1.1572980880737305,
"step": 2510
},
{
"epoch": 0.9594659036719122,
"grad_norm": 0.22631202638149261,
"learning_rate": 8.887511793568614e-07,
"loss": 1.1380080223083495,
"step": 2515
},
{
"epoch": 0.9613733905579399,
"grad_norm": 0.25536733865737915,
"learning_rate": 8.084728076657499e-07,
"loss": 1.130523109436035,
"step": 2520
},
{
"epoch": 0.9632808774439676,
"grad_norm": 0.23546051979064941,
"learning_rate": 7.319790381070446e-07,
"loss": 1.18359375,
"step": 2525
},
{
"epoch": 0.9651883643299952,
"grad_norm": 0.2104456126689911,
"learning_rate": 6.592727892615003e-07,
"loss": 1.1894597053527831,
"step": 2530
},
{
"epoch": 0.9670958512160229,
"grad_norm": 0.20208291709423065,
"learning_rate": 5.903568351989575e-07,
"loss": 1.1283025741577148,
"step": 2535
},
{
"epoch": 0.9690033381020505,
"grad_norm": 0.1970140039920807,
"learning_rate": 5.252338053725047e-07,
"loss": 1.1401033401489258,
"step": 2540
},
{
"epoch": 0.9709108249880782,
"grad_norm": 0.23345957696437836,
"learning_rate": 4.639061845181258e-07,
"loss": 1.1129070281982423,
"step": 2545
},
{
"epoch": 0.9728183118741058,
"grad_norm": 0.2700229585170746,
"learning_rate": 4.063763125599307e-07,
"loss": 1.1588699340820312,
"step": 2550
},
{
"epoch": 0.9747257987601335,
"grad_norm": 0.21467843651771545,
"learning_rate": 3.526463845208272e-07,
"loss": 1.1579412460327148,
"step": 2555
},
{
"epoch": 0.9766332856461611,
"grad_norm": 0.2097923904657364,
"learning_rate": 3.027184504388325e-07,
"loss": 1.0812199592590332,
"step": 2560
},
{
"epoch": 0.9785407725321889,
"grad_norm": 0.21349556744098663,
"learning_rate": 2.5659441528880226e-07,
"loss": 1.124245834350586,
"step": 2565
},
{
"epoch": 0.9804482594182165,
"grad_norm": 0.29865762591362,
"learning_rate": 2.1427603890976644e-07,
"loss": 1.194851016998291,
"step": 2570
},
{
"epoch": 0.9823557463042442,
"grad_norm": 0.23294523358345032,
"learning_rate": 1.757649359378055e-07,
"loss": 1.0922676086425782,
"step": 2575
},
{
"epoch": 0.9842632331902719,
"grad_norm": 0.22059480845928192,
"learning_rate": 1.410625757444106e-07,
"loss": 1.1443885803222655,
"step": 2580
},
{
"epoch": 0.9861707200762995,
"grad_norm": 0.2602921724319458,
"learning_rate": 1.1017028238042848e-07,
"loss": 1.1158178329467774,
"step": 2585
},
{
"epoch": 0.9880782069623272,
"grad_norm": 0.2522714138031006,
"learning_rate": 8.30892345255574e-08,
"loss": 1.1615970611572266,
"step": 2590
},
{
"epoch": 0.9899856938483548,
"grad_norm": 0.2601059079170227,
"learning_rate": 5.982046544338315e-08,
"loss": 1.1701422691345216,
"step": 2595
},
{
"epoch": 0.9918931807343825,
"grad_norm": 0.26516997814178467,
"learning_rate": 4.036486294191066e-08,
"loss": 1.1095975875854491,
"step": 2600
},
{
"epoch": 0.9938006676204101,
"grad_norm": 0.2678978145122528,
"learning_rate": 2.4723169339724295e-08,
"loss": 1.1590506553649902,
"step": 2605
},
{
"epoch": 0.9957081545064378,
"grad_norm": 0.26719844341278076,
"learning_rate": 1.2895981437677229e-08,
"loss": 1.1161251068115234,
"step": 2610
},
{
"epoch": 0.9976156413924654,
"grad_norm": 0.22992399334907532,
"learning_rate": 4.883750496098571e-09,
"loss": 1.119638442993164,
"step": 2615
},
{
"epoch": 0.9995231282784931,
"grad_norm": 0.22952629625797272,
"learning_rate": 6.867822175515848e-10,
"loss": 1.0825986862182617,
"step": 2620
}
],
"logging_steps": 5,
"max_steps": 2622,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.529579479129129e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}