Files
Qwen3-8B-finetuned/trainer_state.json

3817 lines
100 KiB
JSON
Raw Normal View History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 10787,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00185416956380661,
"grad_norm": 3.820171696571671,
"learning_rate": 5.864197530864198e-06,
"loss": 2.011005401611328,
"step": 20
},
{
"epoch": 0.00370833912761322,
"grad_norm": 1.096620777853325,
"learning_rate": 1.2037037037037037e-05,
"loss": 0.5124849319458008,
"step": 40
},
{
"epoch": 0.00556250869141983,
"grad_norm": 0.3680814920079142,
"learning_rate": 1.8209876543209877e-05,
"loss": 0.22042126655578614,
"step": 60
},
{
"epoch": 0.00741667825522644,
"grad_norm": 0.2707298056373599,
"learning_rate": 2.438271604938272e-05,
"loss": 0.14981963634490966,
"step": 80
},
{
"epoch": 0.00927084781903305,
"grad_norm": 0.2659431724016679,
"learning_rate": 3.055555555555556e-05,
"loss": 0.12978166341781616,
"step": 100
},
{
"epoch": 0.01112501738283966,
"grad_norm": 0.22306664143926566,
"learning_rate": 3.67283950617284e-05,
"loss": 0.11404435634613037,
"step": 120
},
{
"epoch": 0.01297918694664627,
"grad_norm": 0.30576307991303725,
"learning_rate": 4.290123456790124e-05,
"loss": 0.1076289176940918,
"step": 140
},
{
"epoch": 0.01483335651045288,
"grad_norm": 0.19329367022139832,
"learning_rate": 4.9074074074074075e-05,
"loss": 0.10596754550933837,
"step": 160
},
{
"epoch": 0.016687526074259492,
"grad_norm": 0.18942231256837483,
"learning_rate": 5.524691358024692e-05,
"loss": 0.100307297706604,
"step": 180
},
{
"epoch": 0.0185416956380661,
"grad_norm": 0.15043901400684026,
"learning_rate": 6.141975308641975e-05,
"loss": 0.0976285457611084,
"step": 200
},
{
"epoch": 0.020395865201872712,
"grad_norm": 0.22845759117348927,
"learning_rate": 6.759259259259259e-05,
"loss": 0.09554426670074463,
"step": 220
},
{
"epoch": 0.02225003476567932,
"grad_norm": 0.19139025279674057,
"learning_rate": 7.376543209876543e-05,
"loss": 0.09421352148056031,
"step": 240
},
{
"epoch": 0.024104204329485932,
"grad_norm": 0.1811447164471412,
"learning_rate": 7.993827160493828e-05,
"loss": 0.0942450761795044,
"step": 260
},
{
"epoch": 0.02595837389329254,
"grad_norm": 0.1650628838740293,
"learning_rate": 8.611111111111112e-05,
"loss": 0.09581634402275085,
"step": 280
},
{
"epoch": 0.027812543457099152,
"grad_norm": 0.15454810302337588,
"learning_rate": 9.228395061728396e-05,
"loss": 0.09313751459121704,
"step": 300
},
{
"epoch": 0.02966671302090576,
"grad_norm": 0.13705135560279838,
"learning_rate": 9.84567901234568e-05,
"loss": 0.09234297275543213,
"step": 320
},
{
"epoch": 0.03152088258471237,
"grad_norm": 0.15487945666116068,
"learning_rate": 9.999949288192827e-05,
"loss": 0.08958485126495361,
"step": 340
},
{
"epoch": 0.033375052148518984,
"grad_norm": 0.12737126215222822,
"learning_rate": 9.99972390445745e-05,
"loss": 0.0907970130443573,
"step": 360
},
{
"epoch": 0.03522922171232559,
"grad_norm": 0.2133019246480929,
"learning_rate": 9.999318222267878e-05,
"loss": 0.08801476955413819,
"step": 380
},
{
"epoch": 0.0370833912761322,
"grad_norm": 0.17216898375465342,
"learning_rate": 9.99873225625369e-05,
"loss": 0.09476873278617859,
"step": 400
},
{
"epoch": 0.03893756083993881,
"grad_norm": 0.1980181417116414,
"learning_rate": 9.997966027545811e-05,
"loss": 0.09878499507904052,
"step": 420
},
{
"epoch": 0.040791730403745424,
"grad_norm": 0.12021105467998261,
"learning_rate": 9.997019563775733e-05,
"loss": 0.0894039511680603,
"step": 440
},
{
"epoch": 0.04264589996755203,
"grad_norm": 0.1506140007132925,
"learning_rate": 9.995892899074535e-05,
"loss": 0.0881020188331604,
"step": 460
},
{
"epoch": 0.04450006953135864,
"grad_norm": 0.09779255157559406,
"learning_rate": 9.994586074071638e-05,
"loss": 0.0859219789505005,
"step": 480
},
{
"epoch": 0.046354239095165256,
"grad_norm": 0.09470959480876269,
"learning_rate": 9.993099135893353e-05,
"loss": 0.08542002439498901,
"step": 500
},
{
"epoch": 0.048208408658971864,
"grad_norm": 0.09917038420264507,
"learning_rate": 9.991432138161169e-05,
"loss": 0.08334171175956726,
"step": 520
},
{
"epoch": 0.05006257822277847,
"grad_norm": 0.10889320362609117,
"learning_rate": 9.989585140989832e-05,
"loss": 0.08618338704109192,
"step": 540
},
{
"epoch": 0.05191674778658508,
"grad_norm": 0.09906113890359036,
"learning_rate": 9.987558210985163e-05,
"loss": 0.08432861566543579,
"step": 560
},
{
"epoch": 0.053770917350391696,
"grad_norm": 0.10940459744489729,
"learning_rate": 9.98535142124167e-05,
"loss": 0.08541926145553588,
"step": 580
},
{
"epoch": 0.055625086914198305,
"grad_norm": 0.09296178818963359,
"learning_rate": 9.982964851339902e-05,
"loss": 0.08227111101150512,
"step": 600
},
{
"epoch": 0.05747925647800491,
"grad_norm": 0.07844261933645442,
"learning_rate": 9.980398587343585e-05,
"loss": 0.08189270496368409,
"step": 620
},
{
"epoch": 0.05933342604181152,
"grad_norm": 0.10575423206665098,
"learning_rate": 9.977652721796516e-05,
"loss": 0.08317902684211731,
"step": 640
},
{
"epoch": 0.061187595605618136,
"grad_norm": 0.09709604090363623,
"learning_rate": 9.974727353719224e-05,
"loss": 0.08244485855102539,
"step": 660
},
{
"epoch": 0.06304176516942474,
"grad_norm": 0.09179391965642997,
"learning_rate": 9.971622588605402e-05,
"loss": 0.08294405937194824,
"step": 680
},
{
"epoch": 0.06489593473323135,
"grad_norm": 0.09556116581662638,
"learning_rate": 9.968338538418107e-05,
"loss": 0.08182405829429626,
"step": 700
},
{
"epoch": 0.06675010429703797,
"grad_norm": 0.08650063066723065,
"learning_rate": 9.96487532158571e-05,
"loss": 0.0813271701335907,
"step": 720
},
{
"epoch": 0.06860427386084457,
"grad_norm": 0.09090183837909592,
"learning_rate": 9.961233062997633e-05,
"loss": 0.08164938688278198,
"step": 740
},
{
"epoch": 0.07045844342465118,
"grad_norm": 0.1015857342614738,
"learning_rate": 9.957411893999851e-05,
"loss": 0.08256611824035645,
"step": 760
},
{
"epoch": 0.0723126129884578,
"grad_norm": 0.09745727469768291,
"learning_rate": 9.953411952390146e-05,
"loss": 0.08092707395553589,
"step": 780
},
{
"epoch": 0.0741667825522644,
"grad_norm": 0.08097614595740198,
"learning_rate": 9.949233382413139e-05,
"loss": 0.0815312385559082,
"step": 800
},
{
"epoch": 0.07602095211607102,
"grad_norm": 0.08685249110755532,
"learning_rate": 9.944876334755093e-05,
"loss": 0.07997688055038452,
"step": 820
},
{
"epoch": 0.07787512167987762,
"grad_norm": 0.07968029411937462,
"learning_rate": 9.940340966538476e-05,
"loss": 0.07891923189163208,
"step": 840
},
{
"epoch": 0.07972929124368423,
"grad_norm": 0.08963184148096366,
"learning_rate": 9.935627441316296e-05,
"loss": 0.08197757005691528,
"step": 860
},
{
"epoch": 0.08158346080749085,
"grad_norm": 0.10326821002778926,
"learning_rate": 9.930735929066198e-05,
"loss": 0.0796918511390686,
"step": 880
},
{
"epoch": 0.08343763037129745,
"grad_norm": 0.07608638835599486,
"learning_rate": 9.925666606184343e-05,
"loss": 0.08107801675796508,
"step": 900
},
{
"epoch": 0.08529179993510407,
"grad_norm": 0.07881338421551327,
"learning_rate": 9.920419655479045e-05,
"loss": 0.08010420799255372,
"step": 920
},
{
"epoch": 0.08714596949891068,
"grad_norm": 0.0888223955745175,
"learning_rate": 9.91499526616417e-05,
"loss": 0.07916066646575928,
"step": 940
},
{
"epoch": 0.08900013906271728,
"grad_norm": 0.08736576160887341,
"learning_rate": 9.909393633852318e-05,
"loss": 0.0791887640953064,
"step": 960
},
{
"epoch": 0.0908543086265239,
"grad_norm": 0.0872645623571638,
"learning_rate": 9.903614960547778e-05,
"loss": 0.07827670574188232,
"step": 980
},
{
"epoch": 0.09270847819033051,
"grad_norm": 0.09683278584901765,
"learning_rate": 9.897659454639227e-05,
"loss": 0.07974874973297119,
"step": 1000
},
{
"epoch": 0.09456264775413711,
"grad_norm": 0.09670970512572408,
"learning_rate": 9.891527330892226e-05,
"loss": 0.07780462503433228,
"step": 1020
},
{
"epoch": 0.09641681731794373,
"grad_norm": 0.07922410697673214,
"learning_rate": 9.885218810441474e-05,
"loss": 0.07922825217247009,
"step": 1040
},
{
"epoch": 0.09827098688175033,
"grad_norm": 0.08764111011974576,
"learning_rate": 9.878734120782831e-05,
"loss": 0.07735201120376586,
"step": 1060
},
{
"epoch": 0.10012515644555695,
"grad_norm": 0.09389190550005704,
"learning_rate": 9.872073495765111e-05,
"loss": 0.07676548957824707,
"step": 1080
},
{
"epoch": 0.10197932600936356,
"grad_norm": 0.08348731011483394,
"learning_rate": 9.865237175581656e-05,
"loss": 0.0770105242729187,
"step": 1100
},
{
"epoch": 0.10383349557317016,
"grad_norm": 0.08265777585705475,
"learning_rate": 9.858225406761677e-05,
"loss": 0.07735653519630432,
"step": 1120
},
{
"epoch": 0.10568766513697678,
"grad_norm": 0.08441811101976025,
"learning_rate": 9.851038442161347e-05,
"loss": 0.07819046974182128,
"step": 1140
},
{
"epoch": 0.10754183470078339,
"grad_norm": 0.07866301603118257,
"learning_rate": 9.843676540954699e-05,
"loss": 0.07668208479881286,
"step": 1160
},
{
"epoch": 0.10939600426459,
"grad_norm": 0.07928327546519687,
"learning_rate": 9.836139968624276e-05,
"loss": 0.07516925930976867,
"step": 1180
},
{
"epoch": 0.11125017382839661,
"grad_norm": 0.08036947844849412,
"learning_rate": 9.828428996951556e-05,
"loss": 0.07837026715278625,
"step": 1200
},
{
"epoch": 0.11310434339220321,
"grad_norm": 0.09143072603694366,
"learning_rate": 9.820543904007143e-05,
"loss": 0.07502893805503845,
"step": 1220
},
{
"epoch": 0.11495851295600983,
"grad_norm": 0.07104320586794924,
"learning_rate": 9.812484974140757e-05,
"loss": 0.07647337913513183,
"step": 1240
},
{
"epoch": 0.11681268251981644,
"grad_norm": 0.07471332199388478,
"learning_rate": 9.804252497970961e-05,
"loss": 0.07538799047470093,
"step": 1260
},
{
"epoch": 0.11866685208362304,
"grad_norm": 0.08695452168038581,
"learning_rate": 9.795846772374698e-05,
"loss": 0.0785320520401001,
"step": 1280
},
{
"epoch": 0.12052102164742966,
"grad_norm": 0.38095845451037036,
"learning_rate": 9.787268100476571e-05,
"loss": 0.14148486852645875,
"step": 1300
},
{
"epoch": 0.12237519121123627,
"grad_norm": 0.23453784446382794,
"learning_rate": 9.778516791637918e-05,
"loss": 0.09106227159500122,
"step": 1320
},
{
"epoch": 0.12422936077504287,
"grad_norm": 0.3078447269608703,
"learning_rate": 9.769593161445658e-05,
"loss": 0.10610485076904297,
"step": 1340
},
{
"epoch": 0.12608353033884948,
"grad_norm": 0.10392810136594262,
"learning_rate": 9.760497531700905e-05,
"loss": 0.09194117784500122,
"step": 1360
},
{
"epoch": 0.1279376999026561,
"grad_norm": 0.08295602822436413,
"learning_rate": 9.751230230407368e-05,
"loss": 0.08373166918754578,
"step": 1380
},
{
"epoch": 0.1297918694664627,
"grad_norm": 0.08248000439562024,
"learning_rate": 9.741791591759523e-05,
"loss": 0.07780111432075501,
"step": 1400
},
{
"epoch": 0.1316460390302693,
"grad_norm": 0.07283731134539405,
"learning_rate": 9.732181956130555e-05,
"loss": 0.08092001676559449,
"step": 1420
},
{
"epoch": 0.13350020859407594,
"grad_norm": 0.07447210866239493,
"learning_rate": 9.722401670060088e-05,
"loss": 0.07920002937316895,
"step": 1440
},
{
"epoch": 0.13535437815788254,
"grad_norm": 0.09523382561674014,
"learning_rate": 9.712451086241689e-05,
"loss": 0.07667785882949829,
"step": 1460
},
{
"epoch": 0.13720854772168914,
"grad_norm": 0.08215211048712649,
"learning_rate": 9.702330563510153e-05,
"loss": 0.07909265160560608,
"step": 1480
},
{
"epoch": 0.13906271728549577,
"grad_norm": 0.06602271525153754,
"learning_rate": 9.69204046682855e-05,
"loss": 0.07624268531799316,
"step": 1500
},
{
"epoch": 0.14091688684930237,
"grad_norm": 0.06950526744671304,
"learning_rate": 9.681581167275077e-05,
"loss": 0.07816213369369507,
"step": 1520
},
{
"epoch": 0.14277105641310897,
"grad_norm": 0.08286355489711139,
"learning_rate": 9.670953042029669e-05,
"loss": 0.07695078253746032,
"step": 1540
},
{
"epoch": 0.1446252259769156,
"grad_norm": 0.07783014967631326,
"learning_rate": 9.6601564743604e-05,
"loss": 0.07489546537399291,
"step": 1560
},
{
"epoch": 0.1464793955407222,
"grad_norm": 0.7215238075868866,
"learning_rate": 9.649191853609665e-05,
"loss": 0.0800896406173706,
"step": 1580
},
{
"epoch": 0.1483335651045288,
"grad_norm": 0.13862601143690534,
"learning_rate": 9.638059575180127e-05,
"loss": 0.11360986232757568,
"step": 1600
},
{
"epoch": 0.15018773466833543,
"grad_norm": 0.08115822256628541,
"learning_rate": 9.626760040520475e-05,
"loss": 0.08856036067008972,
"step": 1620
},
{
"epoch": 0.15204190423214203,
"grad_norm": 0.07756202053063599,
"learning_rate": 9.615293657110937e-05,
"loss": 0.08102580308914184,
"step": 1640
},
{
"epoch": 0.15389607379594863,
"grad_norm": 0.08278086676458717,
"learning_rate": 9.603660838448584e-05,
"loss": 0.08062328100204467,
"step": 1660
},
{
"epoch": 0.15575024335975524,
"grad_norm": 0.07970365542115601,
"learning_rate": 9.591862004032429e-05,
"loss": 0.08349821567535401,
"step": 1680
},
{
"epoch": 0.15760441292356187,
"grad_norm": 0.06570684068013177,
"learning_rate": 9.579897579348282e-05,
"loss": 0.0809329867362976,
"step": 1700
},
{
"epoch": 0.15945858248736847,
"grad_norm": 0.07584058985918503,
"learning_rate": 9.567767995853429e-05,
"loss": 0.07894476652145385,
"step": 1720
},
{
"epoch": 0.16131275205117507,
"grad_norm": 0.09071650635538053,
"learning_rate": 9.555473690961051e-05,
"loss": 0.07519868612289429,
"step": 1740
},
{
"epoch": 0.1631669216149817,
"grad_norm": 0.06348361705719396,
"learning_rate": 9.543015108024463e-05,
"loss": 0.0785559892654419,
"step": 1760
},
{
"epoch": 0.1650210911787883,
"grad_norm": 0.0647389173291084,
"learning_rate": 9.530392696321126e-05,
"loss": 0.07550161480903625,
"step": 1780
},
{
"epoch": 0.1668752607425949,
"grad_norm": 0.07269711770311187,
"learning_rate": 9.517606911036433e-05,
"loss": 0.07608571052551269,
"step": 1800
},
{
"epoch": 0.16872943030640153,
"grad_norm": 0.07223723741746069,
"learning_rate": 9.504658213247316e-05,
"loss": 0.07413275241851806,
"step": 1820
},
{
"epoch": 0.17058359987020813,
"grad_norm": 0.06738973545359006,
"learning_rate": 9.491547069905593e-05,
"loss": 0.07561180591583253,
"step": 1840
},
{
"epoch": 0.17243776943401473,
"grad_norm": 0.07323238258643082,
"learning_rate": 9.478273953821149e-05,
"loss": 0.07366191148757935,
"step": 1860
},
{
"epoch": 0.17429193899782136,
"grad_norm": 0.07257321862760945,
"learning_rate": 9.464839343644878e-05,
"loss": 0.07775602340698243,
"step": 1880
},
{
"epoch": 0.17614610856162796,
"grad_norm": 0.06847894564081444,
"learning_rate": 9.451243723851421e-05,
"loss": 0.07233434915542603,
"step": 1900
},
{
"epoch": 0.17800027812543456,
"grad_norm": 0.06335817185237871,
"learning_rate": 9.437487584721698e-05,
"loss": 0.07621219754219055,
"step": 1920
},
{
"epoch": 0.1798544476892412,
"grad_norm": 0.0640719189203514,
"learning_rate": 9.423571422325223e-05,
"loss": 0.07463518381118775,
"step": 1940
},
{
"epoch": 0.1817086172530478,
"grad_norm": 0.06515539813387905,
"learning_rate": 9.409495738502226e-05,
"loss": 0.07639362812042236,
"step": 1960
},
{
"epoch": 0.1835627868168544,
"grad_norm": 0.06569879839529959,
"learning_rate": 9.39526104084554e-05,
"loss": 0.07541184425354004,
"step": 1980
},
{
"epoch": 0.18541695638066102,
"grad_norm": 0.0578643398896708,
"learning_rate": 9.380867842682305e-05,
"loss": 0.07277128696441651,
"step": 2000
},
{
"epoch": 0.18727112594446763,
"grad_norm": 0.07226395006206054,
"learning_rate": 9.366316663055462e-05,
"loss": 0.07344133853912353,
"step": 2020
},
{
"epoch": 0.18912529550827423,
"grad_norm": 0.06647566469815315,
"learning_rate": 9.351608026705023e-05,
"loss": 0.0730777382850647,
"step": 2040
},
{
"epoch": 0.19097946507208083,
"grad_norm": 0.06745200883731671,
"learning_rate": 9.336742464049157e-05,
"loss": 0.07285230755805969,
"step": 2060
},
{
"epoch": 0.19283363463588746,
"grad_norm": 0.06455764222430915,
"learning_rate": 9.32172051116506e-05,
"loss": 0.07300975322723388,
"step": 2080
},
{
"epoch": 0.19468780419969406,
"grad_norm": 0.06680772726708664,
"learning_rate": 9.306542709769622e-05,
"loss": 0.07422054409980774,
"step": 2100
},
{
"epoch": 0.19654197376350066,
"grad_norm": 0.08413515246100994,
"learning_rate": 9.291209607199892e-05,
"loss": 0.07437876462936402,
"step": 2120
},
{
"epoch": 0.1983961433273073,
"grad_norm": 0.06339429581775848,
"learning_rate": 9.275721756393339e-05,
"loss": 0.0755193829536438,
"step": 2140
},
{
"epoch": 0.2002503128911139,
"grad_norm": 0.19889530430082372,
"learning_rate": 9.260079715867917e-05,
"loss": 0.07785636186599731,
"step": 2160
},
{
"epoch": 0.2021044824549205,
"grad_norm": 0.07340354751257809,
"learning_rate": 9.244284049701923e-05,
"loss": 0.07767927646636963,
"step": 2180
},
{
"epoch": 0.20395865201872712,
"grad_norm": 0.06404995523476703,
"learning_rate": 9.228335327513644e-05,
"loss": 0.07524172067642212,
"step": 2200
},
{
"epoch": 0.20581282158253372,
"grad_norm": 0.0618291038067499,
"learning_rate": 9.212234124440831e-05,
"loss": 0.07576064467430114,
"step": 2220
},
{
"epoch": 0.20766699114634032,
"grad_norm": 0.07109039366191955,
"learning_rate": 9.195981021119955e-05,
"loss": 0.07323854565620422,
"step": 2240
},
{
"epoch": 0.20952116071014695,
"grad_norm": 0.07029203315945454,
"learning_rate": 9.179576603665258e-05,
"loss": 0.07339696884155274,
"step": 2260
},
{
"epoch": 0.21137533027395355,
"grad_norm": 0.05424922342213343,
"learning_rate": 9.163021463647628e-05,
"loss": 0.07229110598564148,
"step": 2280
},
{
"epoch": 0.21322949983776016,
"grad_norm": 0.06673481099949503,
"learning_rate": 9.146316198073263e-05,
"loss": 0.07312118411064147,
"step": 2300
},
{
"epoch": 0.21508366940156678,
"grad_norm": 0.06063445092674018,
"learning_rate": 9.129461409362141e-05,
"loss": 0.07265024781227111,
"step": 2320
},
{
"epoch": 0.2169378389653734,
"grad_norm": 0.0597139015702256,
"learning_rate": 9.112457705326292e-05,
"loss": 0.07265085577964783,
"step": 2340
},
{
"epoch": 0.21879200852918,
"grad_norm": 0.06807060961713832,
"learning_rate": 9.095305699147889e-05,
"loss": 0.07184367179870606,
"step": 2360
},
{
"epoch": 0.22064617809298662,
"grad_norm": 0.07140570456141754,
"learning_rate": 9.078006009357125e-05,
"loss": 0.07540631890296937,
"step": 2380
},
{
"epoch": 0.22250034765679322,
"grad_norm": 0.06703730660153011,
"learning_rate": 9.060559259809915e-05,
"loss": 0.07418131828308105,
"step": 2400
},
{
"epoch": 0.22435451722059982,
"grad_norm": 0.06548741413963394,
"learning_rate": 9.042966079665397e-05,
"loss": 0.0729554533958435,
"step": 2420
},
{
"epoch": 0.22620868678440642,
"grad_norm": 0.06888642053259762,
"learning_rate": 9.025227103363238e-05,
"loss": 0.07194913625717163,
"step": 2440
},
{
"epoch": 0.22806285634821305,
"grad_norm": 0.06526620320003416,
"learning_rate": 9.007342970600762e-05,
"loss": 0.07399054765701293,
"step": 2460
},
{
"epoch": 0.22991702591201965,
"grad_norm": 0.06599047053204352,
"learning_rate": 8.989314326309885e-05,
"loss": 0.07210756540298462,
"step": 2480
},
{
"epoch": 0.23177119547582625,
"grad_norm": 0.057272141367375115,
"learning_rate": 8.971141820633842e-05,
"loss": 0.07364898920059204,
"step": 2500
},
{
"epoch": 0.23362536503963288,
"grad_norm": 0.07141711922984666,
"learning_rate": 8.95282610890376e-05,
"loss": 0.07184793949127197,
"step": 2520
},
{
"epoch": 0.23547953460343948,
"grad_norm": 0.05980379022872412,
"learning_rate": 8.934367851615009e-05,
"loss": 0.07045998573303222,
"step": 2540
},
{
"epoch": 0.23733370416724608,
"grad_norm": 0.05868034732418876,
"learning_rate": 8.9157677144034e-05,
"loss": 0.07100774049758911,
"step": 2560
},
{
"epoch": 0.2391878737310527,
"grad_norm": 0.057496908915986564,
"learning_rate": 8.897026368021167e-05,
"loss": 0.07279749512672425,
"step": 2580
},
{
"epoch": 0.24104204329485932,
"grad_norm": 0.067832866197957,
"learning_rate": 8.878144488312786e-05,
"loss": 0.07153629064559937,
"step": 2600
},
{
"epoch": 0.24289621285866592,
"grad_norm": 0.056524705132541014,
"learning_rate": 8.859122756190601e-05,
"loss": 0.07133421301841736,
"step": 2620
},
{
"epoch": 0.24475038242247255,
"grad_norm": 0.05764463896204705,
"learning_rate": 8.83996185761027e-05,
"loss": 0.07275062203407287,
"step": 2640
},
{
"epoch": 0.24660455198627915,
"grad_norm": 0.06490036766892582,
"learning_rate": 8.820662483546027e-05,
"loss": 0.07059406042098999,
"step": 2660
},
{
"epoch": 0.24845872155008575,
"grad_norm": 0.0585184759576509,
"learning_rate": 8.801225329965764e-05,
"loss": 0.07081823348999024,
"step": 2680
},
{
"epoch": 0.2503128911138924,
"grad_norm": 0.06661292390742107,
"learning_rate": 8.781651097805936e-05,
"loss": 0.07319986224174499,
"step": 2700
},
{
"epoch": 0.25216706067769895,
"grad_norm": 0.058350353721374024,
"learning_rate": 8.76194049294628e-05,
"loss": 0.07124414443969726,
"step": 2720
},
{
"epoch": 0.2540212302415056,
"grad_norm": 0.08957973296693779,
"learning_rate": 8.742094226184365e-05,
"loss": 0.07674325704574585,
"step": 2740
},
{
"epoch": 0.2558753998053122,
"grad_norm": 0.071281796664039,
"learning_rate": 8.722113013209954e-05,
"loss": 0.07280395030975342,
"step": 2760
},
{
"epoch": 0.2577295693691188,
"grad_norm": 0.07898835447697226,
"learning_rate": 8.701997574579196e-05,
"loss": 0.07572759985923767,
"step": 2780
},
{
"epoch": 0.2595837389329254,
"grad_norm": 0.06414597933057603,
"learning_rate": 8.681748635688645e-05,
"loss": 0.07128249406814575,
"step": 2800
},
{
"epoch": 0.26143790849673204,
"grad_norm": 0.06678113600827112,
"learning_rate": 8.661366926749101e-05,
"loss": 0.07258655428886414,
"step": 2820
},
{
"epoch": 0.2632920780605386,
"grad_norm": 0.06377380307680854,
"learning_rate": 8.640853182759274e-05,
"loss": 0.07168116569519042,
"step": 2840
},
{
"epoch": 0.26514624762434524,
"grad_norm": 0.06714205421571565,
"learning_rate": 8.620208143479276e-05,
"loss": 0.07134959697723389,
"step": 2860
},
{
"epoch": 0.2670004171881519,
"grad_norm": 0.0586694563117733,
"learning_rate": 8.59943255340395e-05,
"loss": 0.07181867361068725,
"step": 2880
},
{
"epoch": 0.26885458675195845,
"grad_norm": 0.06080732058251141,
"learning_rate": 8.578527161736027e-05,
"loss": 0.07058979272842407,
"step": 2900
},
{
"epoch": 0.2707087563157651,
"grad_norm": 0.054343947451525515,
"learning_rate": 8.557492722359091e-05,
"loss": 0.06852529644966125,
"step": 2920
},
{
"epoch": 0.2725629258795717,
"grad_norm": 0.06408410445364573,
"learning_rate": 8.53632999381041e-05,
"loss": 0.07001596689224243,
"step": 2940
},
{
"epoch": 0.2744170954433783,
"grad_norm": 0.06594523188673412,
"learning_rate": 8.515039739253576e-05,
"loss": 0.07448661923408509,
"step": 2960
},
{
"epoch": 0.2762712650071849,
"grad_norm": 0.0661265872233322,
"learning_rate": 8.493622726450981e-05,
"loss": 0.07278544902801513,
"step": 2980
},
{
"epoch": 0.27812543457099154,
"grad_norm": 0.05787193677356709,
"learning_rate": 8.472079727736128e-05,
"loss": 0.06984232664108277,
"step": 3000
},
{
"epoch": 0.2799796041347981,
"grad_norm": 0.05692169605401866,
"learning_rate": 8.450411519985794e-05,
"loss": 0.07119113206863403,
"step": 3020
},
{
"epoch": 0.28183377369860474,
"grad_norm": 0.05714621162711715,
"learning_rate": 8.428618884592001e-05,
"loss": 0.0721335768699646,
"step": 3040
},
{
"epoch": 0.28368794326241137,
"grad_norm": 0.056044521516826984,
"learning_rate": 8.406702607433833e-05,
"loss": 0.07185147404670715,
"step": 3060
},
{
"epoch": 0.28554211282621794,
"grad_norm": 0.05866908411820298,
"learning_rate": 8.384663478849116e-05,
"loss": 0.06987054347991943,
"step": 3080
},
{
"epoch": 0.28739628239002457,
"grad_norm": 0.05304897286452575,
"learning_rate": 8.362502293605898e-05,
"loss": 0.07024177312850952,
"step": 3100
},
{
"epoch": 0.2892504519538312,
"grad_norm": 0.0665427531160301,
"learning_rate": 8.340219850873794e-05,
"loss": 0.07095595598220825,
"step": 3120
},
{
"epoch": 0.2911046215176378,
"grad_norm": 0.06428509102988876,
"learning_rate": 8.317816954195175e-05,
"loss": 0.07078310251235961,
"step": 3140
},
{
"epoch": 0.2929587910814444,
"grad_norm": 0.05450559848407936,
"learning_rate": 8.295294411456176e-05,
"loss": 0.07231824398040772,
"step": 3160
},
{
"epoch": 0.29481296064525103,
"grad_norm": 0.05831466107721256,
"learning_rate": 8.272653034857577e-05,
"loss": 0.07037442922592163,
"step": 3180
},
{
"epoch": 0.2966671302090576,
"grad_norm": 0.10743868781185825,
"learning_rate": 8.249893640885505e-05,
"loss": 0.07160881757736207,
"step": 3200
},
{
"epoch": 0.29852129977286423,
"grad_norm": 0.06273025352399365,
"learning_rate": 8.227017050281987e-05,
"loss": 0.07143862247467041,
"step": 3220
},
{
"epoch": 0.30037546933667086,
"grad_norm": 0.057489325657076214,
"learning_rate": 8.204024088015363e-05,
"loss": 0.07265503406524658,
"step": 3240
},
{
"epoch": 0.30222963890047744,
"grad_norm": 0.06875218626973734,
"learning_rate": 8.180915583250532e-05,
"loss": 0.07018079757690429,
"step": 3260
},
{
"epoch": 0.30408380846428407,
"grad_norm": 0.05917589418608448,
"learning_rate": 8.157692369319045e-05,
"loss": 0.07140896916389465,
"step": 3280
},
{
"epoch": 0.3059379780280907,
"grad_norm": 0.06423179153616093,
"learning_rate": 8.134355283689063e-05,
"loss": 0.07040358781814575,
"step": 3300
},
{
"epoch": 0.30779214759189727,
"grad_norm": 0.058824570513165975,
"learning_rate": 8.110905167935147e-05,
"loss": 0.07193192243576049,
"step": 3320
},
{
"epoch": 0.3096463171557039,
"grad_norm": 0.06801954697849306,
"learning_rate": 8.087342867707918e-05,
"loss": 0.07015389204025269,
"step": 3340
},
{
"epoch": 0.31150048671951047,
"grad_norm": 0.05871440603852085,
"learning_rate": 8.063669232703559e-05,
"loss": 0.06960692405700683,
"step": 3360
},
{
"epoch": 0.3133546562833171,
"grad_norm": 0.05730341489672219,
"learning_rate": 8.03988511663317e-05,
"loss": 0.06944797039031983,
"step": 3380
},
{
"epoch": 0.31520882584712373,
"grad_norm": 0.0568291514334782,
"learning_rate": 8.015991377191984e-05,
"loss": 0.06993094086647034,
"step": 3400
},
{
"epoch": 0.3170629954109303,
"grad_norm": 0.0549687290382305,
"learning_rate": 7.991988876028437e-05,
"loss": 0.06836211681365967,
"step": 3420
},
{
"epoch": 0.31891716497473693,
"grad_norm": 0.06274212662593626,
"learning_rate": 7.967878478713097e-05,
"loss": 0.0685504674911499,
"step": 3440
},
{
"epoch": 0.32077133453854356,
"grad_norm": 0.05313533630838134,
"learning_rate": 7.943661054707448e-05,
"loss": 0.06650092601776122,
"step": 3460
},
{
"epoch": 0.32262550410235014,
"grad_norm": 0.052743862770069115,
"learning_rate": 7.919337477332536e-05,
"loss": 0.0690037727355957,
"step": 3480
},
{
"epoch": 0.32447967366615676,
"grad_norm": 0.05567356208095886,
"learning_rate": 7.894908623737478e-05,
"loss": 0.06971213817596436,
"step": 3500
},
{
"epoch": 0.3263338432299634,
"grad_norm": 0.06645957761141873,
"learning_rate": 7.870375374867827e-05,
"loss": 0.06895116567611695,
"step": 3520
},
{
"epoch": 0.32818801279376997,
"grad_norm": 0.05552776725124249,
"learning_rate": 7.845738615433804e-05,
"loss": 0.07218914031982422,
"step": 3540
},
{
"epoch": 0.3300421823575766,
"grad_norm": 0.05876603194289993,
"learning_rate": 7.820999233878398e-05,
"loss": 0.06853553056716918,
"step": 3560
},
{
"epoch": 0.3318963519213832,
"grad_norm": 0.05786796219083153,
"learning_rate": 7.796158122345322e-05,
"loss": 0.06862747669219971,
"step": 3580
},
{
"epoch": 0.3337505214851898,
"grad_norm": 0.05626094635104044,
"learning_rate": 7.771216176646844e-05,
"loss": 0.07211055755615234,
"step": 3600
},
{
"epoch": 0.33560469104899643,
"grad_norm": 0.05358929639409373,
"learning_rate": 7.746174296231482e-05,
"loss": 0.0689316987991333,
"step": 3620
},
{
"epoch": 0.33745886061280306,
"grad_norm": 0.06834932968026425,
"learning_rate": 7.721033384151568e-05,
"loss": 0.06952942609786987,
"step": 3640
},
{
"epoch": 0.33931303017660963,
"grad_norm": 0.053843769856845705,
"learning_rate": 7.69579434703068e-05,
"loss": 0.0695588767528534,
"step": 3660
},
{
"epoch": 0.34116719974041626,
"grad_norm": 0.0537965169492173,
"learning_rate": 7.670458095030955e-05,
"loss": 0.07019612789154053,
"step": 3680
},
{
"epoch": 0.3430213693042229,
"grad_norm": 0.053970780996903334,
"learning_rate": 7.645025541820254e-05,
"loss": 0.07074885964393615,
"step": 3700
},
{
"epoch": 0.34487553886802946,
"grad_norm": 0.05500723418032885,
"learning_rate": 7.619497604539232e-05,
"loss": 0.06914512515068054,
"step": 3720
},
{
"epoch": 0.3467297084318361,
"grad_norm": 0.05286890609596364,
"learning_rate": 7.593875203768247e-05,
"loss": 0.06664897799491883,
"step": 3740
},
{
"epoch": 0.3485838779956427,
"grad_norm": 0.05218547592002417,
"learning_rate": 7.568159263494171e-05,
"loss": 0.06811505556106567,
"step": 3760
},
{
"epoch": 0.3504380475594493,
"grad_norm": 0.053639225720427654,
"learning_rate": 7.542350711077068e-05,
"loss": 0.06916459798812866,
"step": 3780
},
{
"epoch": 0.3522922171232559,
"grad_norm": 0.05160367111654179,
"learning_rate": 7.516450477216749e-05,
"loss": 0.06800092458724975,
"step": 3800
},
{
"epoch": 0.35414638668706255,
"grad_norm": 0.05341978285742432,
"learning_rate": 7.49045949591922e-05,
"loss": 0.06905245780944824,
"step": 3820
},
{
"epoch": 0.3560005562508691,
"grad_norm": 0.05616921682628937,
"learning_rate": 7.464378704462984e-05,
"loss": 0.0688902497291565,
"step": 3840
},
{
"epoch": 0.35785472581467576,
"grad_norm": 0.05279023341750433,
"learning_rate": 7.438209043365249e-05,
"loss": 0.06842745542526245,
"step": 3860
},
{
"epoch": 0.3597088953784824,
"grad_norm": 0.04750647210947389,
"learning_rate": 7.41195145634802e-05,
"loss": 0.06902355551719666,
"step": 3880
},
{
"epoch": 0.36156306494228896,
"grad_norm": 0.051454390959194005,
"learning_rate": 7.385606890304052e-05,
"loss": 0.0672761082649231,
"step": 3900
},
{
"epoch": 0.3634172345060956,
"grad_norm": 0.05547366674280859,
"learning_rate": 7.359176295262711e-05,
"loss": 0.06806289553642272,
"step": 3920
},
{
"epoch": 0.3652714040699022,
"grad_norm": 0.05543290058096399,
"learning_rate": 7.332660624355715e-05,
"loss": 0.06858486533164979,
"step": 3940
},
{
"epoch": 0.3671255736337088,
"grad_norm": 0.05355034043300937,
"learning_rate": 7.306060833782759e-05,
"loss": 0.06711235046386718,
"step": 3960
},
{
"epoch": 0.3689797431975154,
"grad_norm": 0.05307257946024117,
"learning_rate": 7.279377882777036e-05,
"loss": 0.07175731658935547,
"step": 3980
},
{
"epoch": 0.37083391276132205,
"grad_norm": 0.052004343105162644,
"learning_rate": 7.252612733570642e-05,
"loss": 0.06841775178909301,
"step": 4000
},
{
"epoch": 0.3726880823251286,
"grad_norm": 0.052420044317239005,
"learning_rate": 7.225766351359881e-05,
"loss": 0.06980134844779969,
"step": 4020
},
{
"epoch": 0.37454225188893525,
"grad_norm": 0.05076973367976567,
"learning_rate": 7.198839704270454e-05,
"loss": 0.06585143804550171,
"step": 4040
},
{
"epoch": 0.3763964214527419,
"grad_norm": 0.04933338155179922,
"learning_rate": 7.171833763322553e-05,
"loss": 0.06848374009132385,
"step": 4060
},
{
"epoch": 0.37825059101654845,
"grad_norm": 0.056698449655288474,
"learning_rate": 7.144749502395832e-05,
"loss": 0.06790245175361634,
"step": 4080
},
{
"epoch": 0.3801047605803551,
"grad_norm": 0.05659410847003245,
"learning_rate": 7.1175878981943e-05,
"loss": 0.06954320073127747,
"step": 4100
},
{
"epoch": 0.38195893014416166,
"grad_norm": 0.055205533803030456,
"learning_rate": 7.090349930211094e-05,
"loss": 0.06869494915008545,
"step": 4120
},
{
"epoch": 0.3838130997079683,
"grad_norm": 0.04801055885664534,
"learning_rate": 7.063036580693159e-05,
"loss": 0.06774266958236694,
"step": 4140
},
{
"epoch": 0.3856672692717749,
"grad_norm": 0.05817794340467535,
"learning_rate": 7.035648834605821e-05,
"loss": 0.06900098323822021,
"step": 4160
},
{
"epoch": 0.3875214388355815,
"grad_norm": 0.055261341198151466,
"learning_rate": 7.008187679597276e-05,
"loss": 0.06982098817825318,
"step": 4180
},
{
"epoch": 0.3893756083993881,
"grad_norm": 0.046136016119472184,
"learning_rate": 6.980654105962965e-05,
"loss": 0.06795721054077149,
"step": 4200
},
{
"epoch": 0.39122977796319475,
"grad_norm": 0.04428693135314919,
"learning_rate": 6.95304910660987e-05,
"loss": 0.07044395208358764,
"step": 4220
},
{
"epoch": 0.3930839475270013,
"grad_norm": 0.05867699879205952,
"learning_rate": 6.9253736770207e-05,
"loss": 0.06826889514923096,
"step": 4240
},
{
"epoch": 0.39493811709080795,
"grad_norm": 0.05524625679458913,
"learning_rate": 6.897628815218004e-05,
"loss": 0.06704115867614746,
"step": 4260
},
{
"epoch": 0.3967922866546146,
"grad_norm": 0.04784807226686665,
"learning_rate": 6.869815521728167e-05,
"loss": 0.06963608860969543,
"step": 4280
},
{
"epoch": 0.39864645621842115,
"grad_norm": 0.05626967952229251,
"learning_rate": 6.841934799545337e-05,
"loss": 0.07049011588096618,
"step": 4300
},
{
"epoch": 0.4005006257822278,
"grad_norm": 0.049080750288660435,
"learning_rate": 6.813987654095258e-05,
"loss": 0.06666951179504395,
"step": 4320
},
{
"epoch": 0.4023547953460344,
"grad_norm": 0.05428166421605746,
"learning_rate": 6.785975093199001e-05,
"loss": 0.06822763085365295,
"step": 4340
},
{
"epoch": 0.404208964909841,
"grad_norm": 0.05187808062037162,
"learning_rate": 6.757898127036638e-05,
"loss": 0.06750587821006775,
"step": 4360
},
{
"epoch": 0.4060631344736476,
"grad_norm": 0.04994621419507147,
"learning_rate": 6.729757768110792e-05,
"loss": 0.06976407766342163,
"step": 4380
},
{
"epoch": 0.40791730403745424,
"grad_norm": 0.0530031470064577,
"learning_rate": 6.701555031210148e-05,
"loss": 0.06885499358177186,
"step": 4400
},
{
"epoch": 0.4097714736012608,
"grad_norm": 0.04836802069144157,
"learning_rate": 6.673290933372835e-05,
"loss": 0.06657370328903198,
"step": 4420
},
{
"epoch": 0.41162564316506745,
"grad_norm": 0.052208575398455824,
"learning_rate": 6.644966493849768e-05,
"loss": 0.06978561878204345,
"step": 4440
},
{
"epoch": 0.4134798127288741,
"grad_norm": 0.04896375874971665,
"learning_rate": 6.61658273406788e-05,
"loss": 0.06789119243621826,
"step": 4460
},
{
"epoch": 0.41533398229268065,
"grad_norm": 0.05207583875443149,
"learning_rate": 6.588140677593293e-05,
"loss": 0.06801660060882568,
"step": 4480
},
{
"epoch": 0.4171881518564873,
"grad_norm": 0.052321701855526835,
"learning_rate": 6.55964135009441e-05,
"loss": 0.0674366295337677,
"step": 4500
},
{
"epoch": 0.4190423214202939,
"grad_norm": 0.052460056969454404,
"learning_rate": 6.531085779304918e-05,
"loss": 0.06703794598579407,
"step": 4520
},
{
"epoch": 0.4208964909841005,
"grad_norm": 0.05984149730805949,
"learning_rate": 6.502474994986734e-05,
"loss": 0.06836413145065308,
"step": 4540
},
{
"epoch": 0.4227506605479071,
"grad_norm": 0.051043482645265106,
"learning_rate": 6.473810028892865e-05,
"loss": 0.06563597917556763,
"step": 4560
},
{
"epoch": 0.42460483011171374,
"grad_norm": 0.05440966274797062,
"learning_rate": 6.445091914730213e-05,
"loss": 0.06778264045715332,
"step": 4580
},
{
"epoch": 0.4264589996755203,
"grad_norm": 0.05744240077417072,
"learning_rate": 6.416321688122278e-05,
"loss": 0.06670353412628174,
"step": 4600
},
{
"epoch": 0.42831316923932694,
"grad_norm": 0.0563410864459583,
"learning_rate": 6.387500386571827e-05,
"loss": 0.06842938661575318,
"step": 4620
},
{
"epoch": 0.43016733880313357,
"grad_norm": 0.05677881581473309,
"learning_rate": 6.358629049423477e-05,
"loss": 0.06564785242080688,
"step": 4640
},
{
"epoch": 0.43202150836694014,
"grad_norm": 0.051386096259373924,
"learning_rate": 6.329708717826209e-05,
"loss": 0.06799172163009644,
"step": 4660
},
{
"epoch": 0.4338756779307468,
"grad_norm": 0.054688365900863255,
"learning_rate": 6.30074043469583e-05,
"loss": 0.06615757346153259,
"step": 4680
},
{
"epoch": 0.4357298474945534,
"grad_norm": 0.05221690983924291,
"learning_rate": 6.271725244677357e-05,
"loss": 0.06634211540222168,
"step": 4700
},
{
"epoch": 0.43758401705836,
"grad_norm": 0.05447789674208455,
"learning_rate": 6.24266419410735e-05,
"loss": 0.06808164119720458,
"step": 4720
},
{
"epoch": 0.4394381866221666,
"grad_norm": 0.045986624762272,
"learning_rate": 6.213558330976177e-05,
"loss": 0.06660630106925965,
"step": 4740
},
{
"epoch": 0.44129235618597323,
"grad_norm": 0.06057721119805469,
"learning_rate": 6.184408704890221e-05,
"loss": 0.06852835416793823,
"step": 4760
},
{
"epoch": 0.4431465257497798,
"grad_norm": 0.046200947505290135,
"learning_rate": 6.155216367034035e-05,
"loss": 0.0657484233379364,
"step": 4780
},
{
"epoch": 0.44500069531358644,
"grad_norm": 0.051188517402448765,
"learning_rate": 6.125982370132428e-05,
"loss": 0.06651111841201782,
"step": 4800
},
{
"epoch": 0.446854864877393,
"grad_norm": 0.04434185112458488,
"learning_rate": 6.0967077684124996e-05,
"loss": 0.06592975258827209,
"step": 4820
},
{
"epoch": 0.44870903444119964,
"grad_norm": 0.052132687845858554,
"learning_rate": 6.067393617565633e-05,
"loss": 0.06635326147079468,
"step": 4840
},
{
"epoch": 0.45056320400500627,
"grad_norm": 0.045119727419172674,
"learning_rate": 6.03804097470942e-05,
"loss": 0.06596879959106446,
"step": 4860
},
{
"epoch": 0.45241737356881284,
"grad_norm": 0.05246974817161853,
"learning_rate": 6.008650898349536e-05,
"loss": 0.06706082224845886,
"step": 4880
},
{
"epoch": 0.45427154313261947,
"grad_norm": 0.04974037668743926,
"learning_rate": 5.9792244483415717e-05,
"loss": 0.06731083393096923,
"step": 4900
},
{
"epoch": 0.4561257126964261,
"grad_norm": 0.049176896941555956,
"learning_rate": 5.9497626858528156e-05,
"loss": 0.06804193258285522,
"step": 4920
},
{
"epoch": 0.4579798822602327,
"grad_norm": 0.04754516758165675,
"learning_rate": 5.92026667332398e-05,
"loss": 0.06767497062683106,
"step": 4940
},
{
"epoch": 0.4598340518240393,
"grad_norm": 0.050756677328236376,
"learning_rate": 5.890737474430892e-05,
"loss": 0.06706864833831787,
"step": 4960
},
{
"epoch": 0.46168822138784593,
"grad_norm": 0.05135634722815149,
"learning_rate": 5.8611761540461386e-05,
"loss": 0.06532019376754761,
"step": 4980
},
{
"epoch": 0.4635423909516525,
"grad_norm": 0.04709592553591549,
"learning_rate": 5.831583778200656e-05,
"loss": 0.0650861382484436,
"step": 5000
},
{
"epoch": 0.46539656051545913,
"grad_norm": 0.05007547630131113,
"learning_rate": 5.801961414045299e-05,
"loss": 0.06561164855957032,
"step": 5020
},
{
"epoch": 0.46725073007926576,
"grad_norm": 0.05050759770015865,
"learning_rate": 5.772310129812345e-05,
"loss": 0.06795594096183777,
"step": 5040
},
{
"epoch": 0.46910489964307234,
"grad_norm": 0.05062220466713996,
"learning_rate": 5.742630994776982e-05,
"loss": 0.06659960150718688,
"step": 5060
},
{
"epoch": 0.47095906920687897,
"grad_norm": 0.04765243391149529,
"learning_rate": 5.712925079218741e-05,
"loss": 0.06613020896911621,
"step": 5080
},
{
"epoch": 0.4728132387706856,
"grad_norm": 0.04663876275325524,
"learning_rate": 5.683193454382909e-05,
"loss": 0.06645833253860474,
"step": 5100
},
{
"epoch": 0.47466740833449217,
"grad_norm": 0.05430827572057466,
"learning_rate": 5.6534371924418913e-05,
"loss": 0.06557634472846985,
"step": 5120
},
{
"epoch": 0.4765215778982988,
"grad_norm": 0.05355476380062587,
"learning_rate": 5.623657366456545e-05,
"loss": 0.06667708158493042,
"step": 5140
},
{
"epoch": 0.4783757474621054,
"grad_norm": 0.0539010131124388,
"learning_rate": 5.59385505033749e-05,
"loss": 0.06924290657043457,
"step": 5160
},
{
"epoch": 0.480229917025912,
"grad_norm": 0.053887588219191936,
"learning_rate": 5.564031318806376e-05,
"loss": 0.06771377325057984,
"step": 5180
},
{
"epoch": 0.48208408658971863,
"grad_norm": 0.05200039551143676,
"learning_rate": 5.534187247357131e-05,
"loss": 0.06644275188446044,
"step": 5200
},
{
"epoch": 0.48393825615352526,
"grad_norm": 0.04411301193034954,
"learning_rate": 5.504323912217171e-05,
"loss": 0.06886463165283203,
"step": 5220
},
{
"epoch": 0.48579242571733183,
"grad_norm": 0.04945757926239084,
"learning_rate": 5.474442390308598e-05,
"loss": 0.06803133487701415,
"step": 5240
},
{
"epoch": 0.48764659528113846,
"grad_norm": 0.049760885213020335,
"learning_rate": 5.444543759209354e-05,
"loss": 0.06640726327896118,
"step": 5260
},
{
"epoch": 0.4895007648449451,
"grad_norm": 0.05729644020155762,
"learning_rate": 5.414629097114371e-05,
"loss": 0.06845622062683106,
"step": 5280
},
{
"epoch": 0.49135493440875166,
"grad_norm": 0.04458570133378101,
"learning_rate": 5.384699482796686e-05,
"loss": 0.06699745655059815,
"step": 5300
},
{
"epoch": 0.4932091039725583,
"grad_norm": 0.06376727879348462,
"learning_rate": 5.354755995568534e-05,
"loss": 0.06719425320625305,
"step": 5320
},
{
"epoch": 0.4950632735363649,
"grad_norm": 0.05295423655645316,
"learning_rate": 5.3247997152424366e-05,
"loss": 0.06554273366928101,
"step": 5340
},
{
"epoch": 0.4969174431001715,
"grad_norm": 0.0491963165583997,
"learning_rate": 5.2948317220922506e-05,
"loss": 0.0648796796798706,
"step": 5360
},
{
"epoch": 0.4987716126639781,
"grad_norm": 0.047344994455222826,
"learning_rate": 5.26485309681422e-05,
"loss": 0.06562695503234864,
"step": 5380
},
{
"epoch": 0.5006257822277848,
"grad_norm": 0.0488658569397386,
"learning_rate": 5.2348649204879993e-05,
"loss": 0.06633292436599732,
"step": 5400
},
{
"epoch": 0.5024799517915913,
"grad_norm": 0.047741841923988455,
"learning_rate": 5.204868274537673e-05,
"loss": 0.06479262113571167,
"step": 5420
},
{
"epoch": 0.5043341213553979,
"grad_norm": 0.05667018797109963,
"learning_rate": 5.1748642406927506e-05,
"loss": 0.0654104232788086,
"step": 5440
},
{
"epoch": 0.5061882909192046,
"grad_norm": 0.04754587562985411,
"learning_rate": 5.1448539009491635e-05,
"loss": 0.06632285118103028,
"step": 5460
},
{
"epoch": 0.5080424604830112,
"grad_norm": 0.05609620041404095,
"learning_rate": 5.114838337530244e-05,
"loss": 0.06753811836242676,
"step": 5480
},
{
"epoch": 0.5098966300468177,
"grad_norm": 0.04656503893277827,
"learning_rate": 5.084818632847701e-05,
"loss": 0.06613025069236755,
"step": 5500
},
{
"epoch": 0.5117507996106244,
"grad_norm": 0.04736067371428652,
"learning_rate": 5.054795869462581e-05,
"loss": 0.06611216068267822,
"step": 5520
},
{
"epoch": 0.513604969174431,
"grad_norm": 0.049346478479543515,
"learning_rate": 5.024771130046234e-05,
"loss": 0.06571578383445739,
"step": 5540
},
{
"epoch": 0.5154591387382376,
"grad_norm": 0.04626565533371505,
"learning_rate": 4.994745497341269e-05,
"loss": 0.06566253900527955,
"step": 5560
},
{
"epoch": 0.5173133083020443,
"grad_norm": 0.04565379716480066,
"learning_rate": 4.964720054122509e-05,
"loss": 0.06430123448371887,
"step": 5580
},
{
"epoch": 0.5191674778658508,
"grad_norm": 0.04594537471780315,
"learning_rate": 4.934695883157943e-05,
"loss": 0.06554259061813354,
"step": 5600
},
{
"epoch": 0.5210216474296574,
"grad_norm": 0.052013576795831996,
"learning_rate": 4.9046740671696804e-05,
"loss": 0.06661000251770019,
"step": 5620
},
{
"epoch": 0.5228758169934641,
"grad_norm": 0.051090964226342454,
"learning_rate": 4.874655688794908e-05,
"loss": 0.0648884654045105,
"step": 5640
},
{
"epoch": 0.5247299865572707,
"grad_norm": 0.04724305963653279,
"learning_rate": 4.844641830546842e-05,
"loss": 0.06492191553115845,
"step": 5660
},
{
"epoch": 0.5265841561210772,
"grad_norm": 0.04902816710436297,
"learning_rate": 4.8146335747757e-05,
"loss": 0.06522760391235352,
"step": 5680
},
{
"epoch": 0.5284383256848839,
"grad_norm": 0.04941475339442223,
"learning_rate": 4.784632003629662e-05,
"loss": 0.06443318128585815,
"step": 5700
},
{
"epoch": 0.5302924952486905,
"grad_norm": 0.04733271095021244,
"learning_rate": 4.754638199015851e-05,
"loss": 0.06498017311096191,
"step": 5720
},
{
"epoch": 0.5321466648124971,
"grad_norm": 0.04798333935806849,
"learning_rate": 4.724653242561314e-05,
"loss": 0.0655935525894165,
"step": 5740
},
{
"epoch": 0.5340008343763037,
"grad_norm": 0.04822235889333689,
"learning_rate": 4.694678215574023e-05,
"loss": 0.06374808549880981,
"step": 5760
},
{
"epoch": 0.5358550039401103,
"grad_norm": 0.05227364492353613,
"learning_rate": 4.664714199003867e-05,
"loss": 0.06397829055786133,
"step": 5780
},
{
"epoch": 0.5377091735039169,
"grad_norm": 0.04952429568599143,
"learning_rate": 4.634762273403692e-05,
"loss": 0.06379647254943847,
"step": 5800
},
{
"epoch": 0.5395633430677236,
"grad_norm": 0.04485795030378189,
"learning_rate": 4.604823518890315e-05,
"loss": 0.06524479389190674,
"step": 5820
},
{
"epoch": 0.5414175126315302,
"grad_norm": 0.04725077082162728,
"learning_rate": 4.5748990151055876e-05,
"loss": 0.063930743932724,
"step": 5840
},
{
"epoch": 0.5432716821953367,
"grad_norm": 0.045571947917625574,
"learning_rate": 4.5449898411774504e-05,
"loss": 0.06565963625907897,
"step": 5860
},
{
"epoch": 0.5451258517591434,
"grad_norm": 0.04512121708064776,
"learning_rate": 4.515097075681029e-05,
"loss": 0.06600631475448608,
"step": 5880
},
{
"epoch": 0.54698002132295,
"grad_norm": 0.04796594600522924,
"learning_rate": 4.485221796599729e-05,
"loss": 0.06486299037933349,
"step": 5900
},
{
"epoch": 0.5488341908867566,
"grad_norm": 0.05035116256547558,
"learning_rate": 4.455365081286371e-05,
"loss": 0.06194999217987061,
"step": 5920
},
{
"epoch": 0.5506883604505632,
"grad_norm": 0.047597307893845496,
"learning_rate": 4.425528006424329e-05,
"loss": 0.06475620269775391,
"step": 5940
},
{
"epoch": 0.5525425300143698,
"grad_norm": 0.04960910114042581,
"learning_rate": 4.395711647988714e-05,
"loss": 0.06375322341918946,
"step": 5960
},
{
"epoch": 0.5543966995781764,
"grad_norm": 0.049130127096257234,
"learning_rate": 4.365917081207566e-05,
"loss": 0.0641134262084961,
"step": 5980
},
{
"epoch": 0.5562508691419831,
"grad_norm": 0.049172771141996544,
"learning_rate": 4.3361453805230814e-05,
"loss": 0.06496943831443787,
"step": 6000
},
{
"epoch": 0.5581050387057896,
"grad_norm": 0.04697420126877478,
"learning_rate": 4.3063976195528666e-05,
"loss": 0.06265571117401122,
"step": 6020
},
{
"epoch": 0.5599592082695962,
"grad_norm": 0.051790011897020614,
"learning_rate": 4.2766748710512225e-05,
"loss": 0.06494598984718322,
"step": 6040
},
{
"epoch": 0.5618133778334029,
"grad_norm": 0.046873609269863045,
"learning_rate": 4.2469782068704566e-05,
"loss": 0.06517527103424073,
"step": 6060
},
{
"epoch": 0.5636675473972095,
"grad_norm": 0.052474352319326345,
"learning_rate": 4.217308697922236e-05,
"loss": 0.06466827392578126,
"step": 6080
},
{
"epoch": 0.565521716961016,
"grad_norm": 0.049136762781101115,
"learning_rate": 4.187667414138958e-05,
"loss": 0.06504812240600585,
"step": 6100
},
{
"epoch": 0.5673758865248227,
"grad_norm": 0.04877058356279773,
"learning_rate": 4.158055424435177e-05,
"loss": 0.06338067054748535,
"step": 6120
},
{
"epoch": 0.5692300560886293,
"grad_norm": 0.05077888378135565,
"learning_rate": 4.128473796669054e-05,
"loss": 0.06424614191055297,
"step": 6140
},
{
"epoch": 0.5710842256524359,
"grad_norm": 0.04686343405918838,
"learning_rate": 4.0989235976038463e-05,
"loss": 0.06370105147361756,
"step": 6160
},
{
"epoch": 0.5729383952162426,
"grad_norm": 0.04552201782807027,
"learning_rate": 4.069405892869439e-05,
"loss": 0.06322758197784424,
"step": 6180
},
{
"epoch": 0.5747925647800491,
"grad_norm": 0.05142990846690097,
"learning_rate": 4.0399217469239174e-05,
"loss": 0.06390318274497986,
"step": 6200
},
{
"epoch": 0.5766467343438557,
"grad_norm": 0.05355258710451629,
"learning_rate": 4.010472223015182e-05,
"loss": 0.06472771167755127,
"step": 6220
},
{
"epoch": 0.5785009039076624,
"grad_norm": 0.048165961612589056,
"learning_rate": 3.981058383142602e-05,
"loss": 0.061773234605789186,
"step": 6240
},
{
"epoch": 0.580355073471469,
"grad_norm": 0.05429088591442539,
"learning_rate": 3.9516812880187227e-05,
"loss": 0.06225782036781311,
"step": 6260
},
{
"epoch": 0.5822092430352755,
"grad_norm": 0.0604603829071715,
"learning_rate": 3.922341997031007e-05,
"loss": 0.06542969346046448,
"step": 6280
},
{
"epoch": 0.5840634125990822,
"grad_norm": 0.04602032275751061,
"learning_rate": 3.8930415682036435e-05,
"loss": 0.06415734887123108,
"step": 6300
},
{
"epoch": 0.5859175821628888,
"grad_norm": 0.047392659065524145,
"learning_rate": 3.863781058159381e-05,
"loss": 0.06317670941352845,
"step": 6320
},
{
"epoch": 0.5877717517266954,
"grad_norm": 0.046575981405470104,
"learning_rate": 3.834561522081434e-05,
"loss": 0.06259400844573974,
"step": 6340
},
{
"epoch": 0.5896259212905021,
"grad_norm": 0.04333812916706854,
"learning_rate": 3.805384013675423e-05,
"loss": 0.06343435049057007,
"step": 6360
},
{
"epoch": 0.5914800908543086,
"grad_norm": 0.04556101087924968,
"learning_rate": 3.776249585131382e-05,
"loss": 0.0633241891860962,
"step": 6380
},
{
"epoch": 0.5933342604181152,
"grad_norm": 0.04884870911304169,
"learning_rate": 3.747159287085812e-05,
"loss": 0.0616495668888092,
"step": 6400
},
{
"epoch": 0.5951884299819219,
"grad_norm": 0.04401211899298281,
"learning_rate": 3.7181141685837943e-05,
"loss": 0.06143069267272949,
"step": 6420
},
{
"epoch": 0.5970425995457285,
"grad_norm": 0.05094195053050948,
"learning_rate": 3.68911527704116e-05,
"loss": 0.0653111219406128,
"step": 6440
},
{
"epoch": 0.598896769109535,
"grad_norm": 0.04487752972178212,
"learning_rate": 3.660163658206718e-05,
"loss": 0.0612871527671814,
"step": 6460
},
{
"epoch": 0.6007509386733417,
"grad_norm": 0.04708924884783961,
"learning_rate": 3.631260356124545e-05,
"loss": 0.06316884756088256,
"step": 6480
},
{
"epoch": 0.6026051082371483,
"grad_norm": 0.049087841455752984,
"learning_rate": 3.6024064130963316e-05,
"loss": 0.06191706657409668,
"step": 6500
},
{
"epoch": 0.6044592778009549,
"grad_norm": 0.05066544046233841,
"learning_rate": 3.573602869643801e-05,
"loss": 0.06327232122421264,
"step": 6520
},
{
"epoch": 0.6063134473647616,
"grad_norm": 0.044158995584808736,
"learning_rate": 3.5448507644711806e-05,
"loss": 0.060802686214447024,
"step": 6540
},
{
"epoch": 0.6081676169285681,
"grad_norm": 0.043851940149196716,
"learning_rate": 3.516151134427745e-05,
"loss": 0.06530459523200989,
"step": 6560
},
{
"epoch": 0.6100217864923747,
"grad_norm": 0.04152111869549671,
"learning_rate": 3.487505014470437e-05,
"loss": 0.06244962215423584,
"step": 6580
},
{
"epoch": 0.6118759560561814,
"grad_norm": 0.048155508768233726,
"learning_rate": 3.458913437626524e-05,
"loss": 0.06283918023109436,
"step": 6600
},
{
"epoch": 0.613730125619988,
"grad_norm": 0.04499639289333585,
"learning_rate": 3.4303774349563623e-05,
"loss": 0.06382766962051392,
"step": 6620
},
{
"epoch": 0.6155842951837945,
"grad_norm": 0.04317627130891494,
"learning_rate": 3.4018980355162114e-05,
"loss": 0.06351471543312073,
"step": 6640
},
{
"epoch": 0.6174384647476012,
"grad_norm": 0.05614677644073403,
"learning_rate": 3.373476266321118e-05,
"loss": 0.06272737979888916,
"step": 6660
},
{
"epoch": 0.6192926343114078,
"grad_norm": 0.042284697745079196,
"learning_rate": 3.3451131523078915e-05,
"loss": 0.062467378377914426,
"step": 6680
},
{
"epoch": 0.6211468038752144,
"grad_norm": 0.044517009371772165,
"learning_rate": 3.316809716298134e-05,
"loss": 0.06361099481582641,
"step": 6700
},
{
"epoch": 0.6230009734390209,
"grad_norm": 0.04892616862217898,
"learning_rate": 3.288566978961356e-05,
"loss": 0.06314346790313721,
"step": 6720
},
{
"epoch": 0.6248551430028276,
"grad_norm": 0.047445153141174416,
"learning_rate": 3.260385958778174e-05,
"loss": 0.06231023669242859,
"step": 6740
},
{
"epoch": 0.6267093125666342,
"grad_norm": 0.049105883450664495,
"learning_rate": 3.23226767200358e-05,
"loss": 0.06305838823318481,
"step": 6760
},
{
"epoch": 0.6285634821304408,
"grad_norm": 0.050572022069277335,
"learning_rate": 3.204213132630296e-05,
"loss": 0.06089724898338318,
"step": 6780
},
{
"epoch": 0.6304176516942475,
"grad_norm": 0.04206886641894423,
"learning_rate": 3.176223352352201e-05,
"loss": 0.062238240242004396,
"step": 6800
},
{
"epoch": 0.632271821258054,
"grad_norm": 0.0418493494393944,
"learning_rate": 3.148299340527857e-05,
"loss": 0.060497474670410153,
"step": 6820
},
{
"epoch": 0.6341259908218606,
"grad_norm": 0.0490312711310362,
"learning_rate": 3.120442104144099e-05,
"loss": 0.060766249895095825,
"step": 6840
},
{
"epoch": 0.6359801603856673,
"grad_norm": 0.04354326975358762,
"learning_rate": 3.092652647779733e-05,
"loss": 0.062382560968399045,
"step": 6860
},
{
"epoch": 0.6378343299494739,
"grad_norm": 0.04925031587484033,
"learning_rate": 3.0649319735692984e-05,
"loss": 0.06159923672676086,
"step": 6880
},
{
"epoch": 0.6396884995132804,
"grad_norm": 0.04435824396894195,
"learning_rate": 3.037281081166938e-05,
"loss": 0.06012410521507263,
"step": 6900
},
{
"epoch": 0.6415426690770871,
"grad_norm": 0.04636097546402966,
"learning_rate": 3.0097009677103472e-05,
"loss": 0.062421870231628415,
"step": 6920
},
{
"epoch": 0.6433968386408937,
"grad_norm": 0.04510143941986499,
"learning_rate": 2.9821926277848082e-05,
"loss": 0.062098950147628784,
"step": 6940
},
{
"epoch": 0.6452510082047003,
"grad_norm": 0.051468922824027076,
"learning_rate": 2.9547570533873316e-05,
"loss": 0.06206930875778198,
"step": 6960
},
{
"epoch": 0.647105177768507,
"grad_norm": 0.04447352421134208,
"learning_rate": 2.9273952338908805e-05,
"loss": 0.06005668640136719,
"step": 6980
},
{
"epoch": 0.6489593473323135,
"grad_norm": 0.04544630265467154,
"learning_rate": 2.900108156008693e-05,
"loss": 0.06390679478645325,
"step": 7000
},
{
"epoch": 0.6508135168961201,
"grad_norm": 0.0440316629640448,
"learning_rate": 2.8728968037586966e-05,
"loss": 0.061896789073944095,
"step": 7020
},
{
"epoch": 0.6526676864599268,
"grad_norm": 0.06308094532142822,
"learning_rate": 2.8457621584280282e-05,
"loss": 0.06296419501304626,
"step": 7040
},
{
"epoch": 0.6545218560237334,
"grad_norm": 0.04910828003002323,
"learning_rate": 2.8187051985376412e-05,
"loss": 0.06295238733291626,
"step": 7060
},
{
"epoch": 0.6563760255875399,
"grad_norm": 0.04902831553853863,
"learning_rate": 2.7917268998070245e-05,
"loss": 0.06277818083763123,
"step": 7080
},
{
"epoch": 0.6582301951513466,
"grad_norm": 0.04474910712199461,
"learning_rate": 2.764828235119008e-05,
"loss": 0.06296697854995728,
"step": 7100
},
{
"epoch": 0.6600843647151532,
"grad_norm": 0.045427696600719006,
"learning_rate": 2.7380101744846898e-05,
"loss": 0.06177424788475037,
"step": 7120
},
{
"epoch": 0.6619385342789598,
"grad_norm": 0.043537695583022555,
"learning_rate": 2.7112736850084476e-05,
"loss": 0.06221094131469727,
"step": 7140
},
{
"epoch": 0.6637927038427665,
"grad_norm": 0.047252055160959865,
"learning_rate": 2.6846197308530662e-05,
"loss": 0.06288429498672485,
"step": 7160
},
{
"epoch": 0.665646873406573,
"grad_norm": 0.04093048002300296,
"learning_rate": 2.6580492732049677e-05,
"loss": 0.061210185289382935,
"step": 7180
},
{
"epoch": 0.6675010429703796,
"grad_norm": 0.044138439067128124,
"learning_rate": 2.631563270239551e-05,
"loss": 0.060793828964233396,
"step": 7200
},
{
"epoch": 0.6693552125341863,
"grad_norm": 0.048270908164784763,
"learning_rate": 2.6051626770866343e-05,
"loss": 0.06118553876876831,
"step": 7220
},
{
"epoch": 0.6712093820979929,
"grad_norm": 0.04192358707110994,
"learning_rate": 2.5788484457960155e-05,
"loss": 0.06138489246368408,
"step": 7240
},
{
"epoch": 0.6730635516617994,
"grad_norm": 0.047264151685931725,
"learning_rate": 2.552621525303144e-05,
"loss": 0.06212824583053589,
"step": 7260
},
{
"epoch": 0.6749177212256061,
"grad_norm": 0.043126828895902566,
"learning_rate": 2.5264828613948842e-05,
"loss": 0.059517890214920044,
"step": 7280
},
{
"epoch": 0.6767718907894127,
"grad_norm": 0.04502796682666258,
"learning_rate": 2.500433396675428e-05,
"loss": 0.05914331078529358,
"step": 7300
},
{
"epoch": 0.6786260603532193,
"grad_norm": 0.046555462215773086,
"learning_rate": 2.4744740705322912e-05,
"loss": 0.06060423851013184,
"step": 7320
},
{
"epoch": 0.680480229917026,
"grad_norm": 0.047731159871982634,
"learning_rate": 2.448605819102442e-05,
"loss": 0.059140574932098386,
"step": 7340
},
{
"epoch": 0.6823343994808325,
"grad_norm": 0.04434178582787534,
"learning_rate": 2.42282957523854e-05,
"loss": 0.060806494951248166,
"step": 7360
},
{
"epoch": 0.6841885690446391,
"grad_norm": 0.05200993355357385,
"learning_rate": 2.3971462684752983e-05,
"loss": 0.06241018772125244,
"step": 7380
},
{
"epoch": 0.6860427386084458,
"grad_norm": 0.04451607128480399,
"learning_rate": 2.371556824995962e-05,
"loss": 0.06107102632522583,
"step": 7400
},
{
"epoch": 0.6878969081722524,
"grad_norm": 0.04672584936907622,
"learning_rate": 2.3460621675989075e-05,
"loss": 0.061403405666351316,
"step": 7420
},
{
"epoch": 0.6897510777360589,
"grad_norm": 0.043451300395321676,
"learning_rate": 2.320663215664366e-05,
"loss": 0.059847354888916016,
"step": 7440
},
{
"epoch": 0.6916052472998656,
"grad_norm": 0.042078989419009484,
"learning_rate": 2.2953608851212687e-05,
"loss": 0.06029091477394104,
"step": 7460
},
{
"epoch": 0.6934594168636722,
"grad_norm": 0.05215931582044261,
"learning_rate": 2.270156088414217e-05,
"loss": 0.06045289039611816,
"step": 7480
},
{
"epoch": 0.6953135864274788,
"grad_norm": 0.04304544077869624,
"learning_rate": 2.2450497344705794e-05,
"loss": 0.06151350736618042,
"step": 7500
},
{
"epoch": 0.6971677559912854,
"grad_norm": 0.042418084422899426,
"learning_rate": 2.2200427286677077e-05,
"loss": 0.05916010141372681,
"step": 7520
},
{
"epoch": 0.699021925555092,
"grad_norm": 0.044102051228497634,
"learning_rate": 2.195135972800299e-05,
"loss": 0.06019682884216308,
"step": 7540
},
{
"epoch": 0.7008760951188986,
"grad_norm": 0.03989097633004308,
"learning_rate": 2.170330365047865e-05,
"loss": 0.060142922401428225,
"step": 7560
},
{
"epoch": 0.7027302646827053,
"grad_norm": 0.04396154557734182,
"learning_rate": 2.1456267999423468e-05,
"loss": 0.06119601726531983,
"step": 7580
},
{
"epoch": 0.7045844342465118,
"grad_norm": 0.046452591818073735,
"learning_rate": 2.1210261683358595e-05,
"loss": 0.05998234748840332,
"step": 7600
},
{
"epoch": 0.7064386038103184,
"grad_norm": 0.041849405904192685,
"learning_rate": 2.0965293573685595e-05,
"loss": 0.059972834587097165,
"step": 7620
},
{
"epoch": 0.7082927733741251,
"grad_norm": 0.04648194004262054,
"learning_rate": 2.072137250436657e-05,
"loss": 0.061202061176300046,
"step": 7640
},
{
"epoch": 0.7101469429379317,
"grad_norm": 0.04203816172547841,
"learning_rate": 2.0478507271605583e-05,
"loss": 0.05759580135345459,
"step": 7660
},
{
"epoch": 0.7120011125017383,
"grad_norm": 0.04159246437703577,
"learning_rate": 2.023670663353146e-05,
"loss": 0.05981243848800659,
"step": 7680
},
{
"epoch": 0.7138552820655449,
"grad_norm": 0.04558860752127258,
"learning_rate": 1.9995979309881975e-05,
"loss": 0.05943760871887207,
"step": 7700
},
{
"epoch": 0.7157094516293515,
"grad_norm": 0.04168710404499,
"learning_rate": 1.9756333981689308e-05,
"loss": 0.05771678686141968,
"step": 7720
},
{
"epoch": 0.7175636211931581,
"grad_norm": 0.04112165344234754,
"learning_rate": 1.951777929096713e-05,
"loss": 0.060441124439239505,
"step": 7740
},
{
"epoch": 0.7194177907569648,
"grad_norm": 0.04588808450921567,
"learning_rate": 1.9280323840398872e-05,
"loss": 0.060588645935058597,
"step": 7760
},
{
"epoch": 0.7212719603207713,
"grad_norm": 0.0423833586200973,
"learning_rate": 1.904397619302752e-05,
"loss": 0.059179884195327756,
"step": 7780
},
{
"epoch": 0.7231261298845779,
"grad_norm": 0.04360990379593918,
"learning_rate": 1.8808744871946794e-05,
"loss": 0.060727423429489134,
"step": 7800
},
{
"epoch": 0.7249802994483846,
"grad_norm": 0.044329689258979815,
"learning_rate": 1.8574638359993825e-05,
"loss": 0.05868118405342102,
"step": 7820
},
{
"epoch": 0.7268344690121912,
"grad_norm": 0.05481220787679917,
"learning_rate": 1.8341665099443235e-05,
"loss": 0.061199039220809937,
"step": 7840
},
{
"epoch": 0.7286886385759977,
"grad_norm": 0.04355173821302689,
"learning_rate": 1.8109833491702695e-05,
"loss": 0.06120600700378418,
"step": 7860
},
{
"epoch": 0.7305428081398044,
"grad_norm": 0.047542762018884,
"learning_rate": 1.7879151897009943e-05,
"loss": 0.05963542461395264,
"step": 7880
},
{
"epoch": 0.732396977703611,
"grad_norm": 0.04859362398408984,
"learning_rate": 1.7649628634131326e-05,
"loss": 0.05993167161941528,
"step": 7900
},
{
"epoch": 0.7342511472674176,
"grad_norm": 0.07696156343679936,
"learning_rate": 1.74212719800618e-05,
"loss": 0.05901572108268738,
"step": 7920
},
{
"epoch": 0.7361053168312243,
"grad_norm": 0.04571216891813866,
"learning_rate": 1.7194090169726417e-05,
"loss": 0.06098179817199707,
"step": 7940
},
{
"epoch": 0.7379594863950308,
"grad_norm": 0.042788171803501564,
"learning_rate": 1.6968091395683428e-05,
"loss": 0.058015406131744385,
"step": 7960
},
{
"epoch": 0.7398136559588374,
"grad_norm": 0.060228689880550704,
"learning_rate": 1.6743283807828776e-05,
"loss": 0.059897661209106445,
"step": 7980
},
{
"epoch": 0.7416678255226441,
"grad_norm": 0.045260107901039764,
"learning_rate": 1.6519675513102228e-05,
"loss": 0.06034721732139588,
"step": 8000
},
{
"epoch": 0.7435219950864507,
"grad_norm": 0.042569996313028956,
"learning_rate": 1.629727457519505e-05,
"loss": 0.05882605314254761,
"step": 8020
},
{
"epoch": 0.7453761646502572,
"grad_norm": 0.04035259949630576,
"learning_rate": 1.607608901425915e-05,
"loss": 0.05778524875640869,
"step": 8040
},
{
"epoch": 0.7472303342140639,
"grad_norm": 0.04381818680813006,
"learning_rate": 1.5856126806617932e-05,
"loss": 0.05960347652435303,
"step": 8060
},
{
"epoch": 0.7490845037778705,
"grad_norm": 0.041521924269364706,
"learning_rate": 1.5637395884478607e-05,
"loss": 0.05910018682479858,
"step": 8080
},
{
"epoch": 0.7509386733416771,
"grad_norm": 0.04483661643227132,
"learning_rate": 1.5419904135646156e-05,
"loss": 0.05863996744155884,
"step": 8100
},
{
"epoch": 0.7527928429054838,
"grad_norm": 0.04170345267394013,
"learning_rate": 1.5203659403238918e-05,
"loss": 0.06005077362060547,
"step": 8120
},
{
"epoch": 0.7546470124692903,
"grad_norm": 0.04314173300757244,
"learning_rate": 1.4988669485405654e-05,
"loss": 0.06016584038734436,
"step": 8140
},
{
"epoch": 0.7565011820330969,
"grad_norm": 0.042521117092671874,
"learning_rate": 1.4774942135044467e-05,
"loss": 0.06094483137130737,
"step": 8160
},
{
"epoch": 0.7583553515969035,
"grad_norm": 0.03935765032771935,
"learning_rate": 1.4562485059523123e-05,
"loss": 0.05919557809829712,
"step": 8180
},
{
"epoch": 0.7602095211607102,
"grad_norm": 0.0465010858441232,
"learning_rate": 1.4351305920401148e-05,
"loss": 0.05924775004386902,
"step": 8200
},
{
"epoch": 0.7620636907245167,
"grad_norm": 0.041847767906551384,
"learning_rate": 1.4141412333153531e-05,
"loss": 0.05741736888885498,
"step": 8220
},
{
"epoch": 0.7639178602883233,
"grad_norm": 0.04556519610235416,
"learning_rate": 1.39328118668961e-05,
"loss": 0.06202120184898376,
"step": 8240
},
{
"epoch": 0.76577202985213,
"grad_norm": 0.03944873638533435,
"learning_rate": 1.3725512044112555e-05,
"loss": 0.05924309492111206,
"step": 8260
},
{
"epoch": 0.7676261994159366,
"grad_norm": 0.043579723817803145,
"learning_rate": 1.3519520340383202e-05,
"loss": 0.05931833982467651,
"step": 8280
},
{
"epoch": 0.7694803689797431,
"grad_norm": 0.04250241195591931,
"learning_rate": 1.3314844184115405e-05,
"loss": 0.05921122431755066,
"step": 8300
},
{
"epoch": 0.7713345385435498,
"grad_norm": 0.040768383521444185,
"learning_rate": 1.3111490956275663e-05,
"loss": 0.05935220122337341,
"step": 8320
},
{
"epoch": 0.7731887081073564,
"grad_norm": 0.040466842442575784,
"learning_rate": 1.2909467990123402e-05,
"loss": 0.0584735095500946,
"step": 8340
},
{
"epoch": 0.775042877671163,
"grad_norm": 0.04578923242449133,
"learning_rate": 1.270878257094663e-05,
"loss": 0.05790427327156067,
"step": 8360
},
{
"epoch": 0.7768970472349697,
"grad_norm": 0.04015645828975025,
"learning_rate": 1.2509441935799139e-05,
"loss": 0.05840156078338623,
"step": 8380
},
{
"epoch": 0.7787512167987762,
"grad_norm": 0.04505252691393977,
"learning_rate": 1.2311453273239548e-05,
"loss": 0.0588322639465332,
"step": 8400
},
{
"epoch": 0.7806053863625828,
"grad_norm": 0.04616645906001549,
"learning_rate": 1.2114823723072077e-05,
"loss": 0.05915929079055786,
"step": 8420
},
{
"epoch": 0.7824595559263895,
"grad_norm": 0.038235671796898386,
"learning_rate": 1.1919560376089062e-05,
"loss": 0.05861332416534424,
"step": 8440
},
{
"epoch": 0.7843137254901961,
"grad_norm": 0.04185333629619464,
"learning_rate": 1.1725670273815248e-05,
"loss": 0.05743967890739441,
"step": 8460
},
{
"epoch": 0.7861678950540026,
"grad_norm": 0.0452666487111135,
"learning_rate": 1.1533160408253879e-05,
"loss": 0.056289541721344,
"step": 8480
},
{
"epoch": 0.7880220646178093,
"grad_norm": 0.041946528912211886,
"learning_rate": 1.134203772163453e-05,
"loss": 0.05721364617347717,
"step": 8500
},
{
"epoch": 0.7898762341816159,
"grad_norm": 0.04391208195397317,
"learning_rate": 1.1152309106162777e-05,
"loss": 0.058383059501647946,
"step": 8520
},
{
"epoch": 0.7917304037454225,
"grad_norm": 0.04316038206952691,
"learning_rate": 1.096398140377164e-05,
"loss": 0.059176528453826906,
"step": 8540
},
{
"epoch": 0.7935845733092292,
"grad_norm": 0.047489102105126565,
"learning_rate": 1.0777061405874872e-05,
"loss": 0.05802996158599853,
"step": 8560
},
{
"epoch": 0.7954387428730357,
"grad_norm": 0.04540449359760401,
"learning_rate": 1.0591555853121998e-05,
"loss": 0.05956486463546753,
"step": 8580
},
{
"epoch": 0.7972929124368423,
"grad_norm": 0.03887464619210056,
"learning_rate": 1.0407471435155313e-05,
"loss": 0.05859973430633545,
"step": 8600
},
{
"epoch": 0.799147082000649,
"grad_norm": 0.0423623187098013,
"learning_rate": 1.0224814790368558e-05,
"loss": 0.05868813991546631,
"step": 8620
},
{
"epoch": 0.8010012515644556,
"grad_norm": 0.04291959975527767,
"learning_rate": 1.0043592505667632e-05,
"loss": 0.057529407739639285,
"step": 8640
},
{
"epoch": 0.8028554211282621,
"grad_norm": 0.04161053163954707,
"learning_rate": 9.863811116232946e-06,
"loss": 0.057827317714691163,
"step": 8660
},
{
"epoch": 0.8047095906920688,
"grad_norm": 0.039652467368728676,
"learning_rate": 9.685477105283825e-06,
"loss": 0.05748317241668701,
"step": 8680
},
{
"epoch": 0.8065637602558754,
"grad_norm": 0.0463013238541496,
"learning_rate": 9.508596903844668e-06,
"loss": 0.05913045406341553,
"step": 8700
},
{
"epoch": 0.808417929819682,
"grad_norm": 0.039819163372741787,
"learning_rate": 9.333176890513095e-06,
"loss": 0.05958133339881897,
"step": 8720
},
{
"epoch": 0.8102720993834887,
"grad_norm": 0.041806269641331766,
"learning_rate": 9.15922339122986e-06,
"loss": 0.05868912935256958,
"step": 8740
},
{
"epoch": 0.8121262689472952,
"grad_norm": 0.03828066923207382,
"learning_rate": 8.986742679050764e-06,
"loss": 0.05822650790214538,
"step": 8760
},
{
"epoch": 0.8139804385111018,
"grad_norm": 0.04169883461469084,
"learning_rate": 8.815740973920411e-06,
"loss": 0.05828083753585815,
"step": 8780
},
{
"epoch": 0.8158346080749085,
"grad_norm": 0.040155940979838246,
"learning_rate": 8.646224442447937e-06,
"loss": 0.05864385366439819,
"step": 8800
},
{
"epoch": 0.8176887776387151,
"grad_norm": 0.04793987505855137,
"learning_rate": 8.478199197684627e-06,
"loss": 0.060294324159622194,
"step": 8820
},
{
"epoch": 0.8195429472025216,
"grad_norm": 0.043479469746193086,
"learning_rate": 8.311671298903439e-06,
"loss": 0.058967453241348264,
"step": 8840
},
{
"epoch": 0.8213971167663283,
"grad_norm": 0.0453819372959514,
"learning_rate": 8.146646751380532e-06,
"loss": 0.057421672344207766,
"step": 8860
},
{
"epoch": 0.8232512863301349,
"grad_norm": 0.04145958828741325,
"learning_rate": 7.983131506178676e-06,
"loss": 0.05738742351531982,
"step": 8880
},
{
"epoch": 0.8251054558939415,
"grad_norm": 0.04270621605428974,
"learning_rate": 7.821131459932673e-06,
"loss": 0.05822542905807495,
"step": 8900
},
{
"epoch": 0.8269596254577481,
"grad_norm": 0.03988428126639982,
"learning_rate": 7.660652454636696e-06,
"loss": 0.055188226699829104,
"step": 8920
},
{
"epoch": 0.8288137950215547,
"grad_norm": 0.0444594845643032,
"learning_rate": 7.501700277433621e-06,
"loss": 0.059385663270950316,
"step": 8940
},
{
"epoch": 0.8306679645853613,
"grad_norm": 0.04895959613153919,
"learning_rate": 7.344280660406333e-06,
"loss": 0.058476173877716066,
"step": 8960
},
{
"epoch": 0.832522134149168,
"grad_norm": 0.046447821589595005,
"learning_rate": 7.188399280371033e-06,
"loss": 0.05890741944313049,
"step": 8980
},
{
"epoch": 0.8343763037129746,
"grad_norm": 0.040919231212288,
"learning_rate": 7.034061758672489e-06,
"loss": 0.057273268699645996,
"step": 9000
},
{
"epoch": 0.8362304732767811,
"grad_norm": 0.04341624851269591,
"learning_rate": 6.881273660981352e-06,
"loss": 0.06076174974441528,
"step": 9020
},
{
"epoch": 0.8380846428405878,
"grad_norm": 0.04407832512819701,
"learning_rate": 6.730040497093437e-06,
"loss": 0.05773304700851441,
"step": 9040
},
{
"epoch": 0.8399388124043944,
"grad_norm": 0.045206674958552885,
"learning_rate": 6.580367720731029e-06,
"loss": 0.057690989971160886,
"step": 9060
},
{
"epoch": 0.841792981968201,
"grad_norm": 0.04184821375620571,
"learning_rate": 6.4322607293462065e-06,
"loss": 0.05713052749633789,
"step": 9080
},
{
"epoch": 0.8436471515320076,
"grad_norm": 0.04260664949525322,
"learning_rate": 6.285724863926224e-06,
"loss": 0.05859750509262085,
"step": 9100
},
{
"epoch": 0.8455013210958142,
"grad_norm": 0.040353521145304376,
"learning_rate": 6.140765408800869e-06,
"loss": 0.05815015435218811,
"step": 9120
},
{
"epoch": 0.8473554906596208,
"grad_norm": 0.04308098450102359,
"learning_rate": 5.997387591451947e-06,
"loss": 0.05699597597122193,
"step": 9140
},
{
"epoch": 0.8492096602234275,
"grad_norm": 0.04208165862880175,
"learning_rate": 5.855596582324724e-06,
"loss": 0.05847395658493042,
"step": 9160
},
{
"epoch": 0.851063829787234,
"grad_norm": 0.04278651313519997,
"learning_rate": 5.7153974946415165e-06,
"loss": 0.05857234001159668,
"step": 9180
},
{
"epoch": 0.8529179993510406,
"grad_norm": 0.04300820482667514,
"learning_rate": 5.576795384217232e-06,
"loss": 0.057513868808746337,
"step": 9200
},
{
"epoch": 0.8547721689148473,
"grad_norm": 0.04184979151941222,
"learning_rate": 5.439795249277141e-06,
"loss": 0.05662814378738403,
"step": 9220
},
{
"epoch": 0.8566263384786539,
"grad_norm": 0.04238638559267024,
"learning_rate": 5.304402030276551e-06,
"loss": 0.05687806010246277,
"step": 9240
},
{
"epoch": 0.8584805080424605,
"grad_norm": 0.04160902207940939,
"learning_rate": 5.1706206097226925e-06,
"loss": 0.05730761289596557,
"step": 9260
},
{
"epoch": 0.8603346776062671,
"grad_norm": 0.05031467787868118,
"learning_rate": 5.038455811998627e-06,
"loss": 0.05795542001724243,
"step": 9280
},
{
"epoch": 0.8621888471700737,
"grad_norm": 0.041455506138344166,
"learning_rate": 4.907912403189285e-06,
"loss": 0.05497853755950928,
"step": 9300
},
{
"epoch": 0.8640430167338803,
"grad_norm": 0.04730451151432445,
"learning_rate": 4.77899509090956e-06,
"loss": 0.05861245393753052,
"step": 9320
},
{
"epoch": 0.865897186297687,
"grad_norm": 0.042890295818311636,
"learning_rate": 4.651708524134624e-06,
"loss": 0.05678141713142395,
"step": 9340
},
{
"epoch": 0.8677513558614935,
"grad_norm": 0.04472612219626013,
"learning_rate": 4.52605729303216e-06,
"loss": 0.0577523946762085,
"step": 9360
},
{
"epoch": 0.8696055254253001,
"grad_norm": 0.04405663125228961,
"learning_rate": 4.402045928796944e-06,
"loss": 0.058051574230194095,
"step": 9380
},
{
"epoch": 0.8714596949891068,
"grad_norm": 0.03734709383772662,
"learning_rate": 4.279678903487355e-06,
"loss": 0.056524813175201416,
"step": 9400
},
{
"epoch": 0.8733138645529134,
"grad_norm": 0.04367894462048559,
"learning_rate": 4.158960629864167e-06,
"loss": 0.05766661167144775,
"step": 9420
},
{
"epoch": 0.87516803411672,
"grad_norm": 0.043940010716802015,
"learning_rate": 4.039895461231396e-06,
"loss": 0.05792294144630432,
"step": 9440
},
{
"epoch": 0.8770222036805266,
"grad_norm": 0.04157334593197274,
"learning_rate": 3.922487691279309e-06,
"loss": 0.05697953701019287,
"step": 9460
},
{
"epoch": 0.8788763732443332,
"grad_norm": 0.04280963353466024,
"learning_rate": 3.806741553929577e-06,
"loss": 0.05770295858383179,
"step": 9480
},
{
"epoch": 0.8807305428081398,
"grad_norm": 0.04360315147910498,
"learning_rate": 3.6926612231826197e-06,
"loss": 0.056513524055480956,
"step": 9500
},
{
"epoch": 0.8825847123719465,
"grad_norm": 0.0433451861162174,
"learning_rate": 3.5802508129670564e-06,
"loss": 0.05862005949020386,
"step": 9520
},
{
"epoch": 0.884438881935753,
"grad_norm": 0.04073468808567981,
"learning_rate": 3.4695143769913784e-06,
"loss": 0.05859295129776001,
"step": 9540
},
{
"epoch": 0.8862930514995596,
"grad_norm": 0.04239051801792883,
"learning_rate": 3.3604559085977317e-06,
"loss": 0.057581615447998044,
"step": 9560
},
{
"epoch": 0.8881472210633663,
"grad_norm": 0.041622529510343655,
"learning_rate": 3.253079340617937e-06,
"loss": 0.0586451530456543,
"step": 9580
},
{
"epoch": 0.8900013906271729,
"grad_norm": 0.04111746214218652,
"learning_rate": 3.1473885452316665e-06,
"loss": 0.05632882118225098,
"step": 9600
},
{
"epoch": 0.8918555601909794,
"grad_norm": 0.04096285048135897,
"learning_rate": 3.0433873338267915e-06,
"loss": 0.058455199003219604,
"step": 9620
},
{
"epoch": 0.893709729754786,
"grad_norm": 0.042381302291485984,
"learning_rate": 2.941079456861917e-06,
"loss": 0.05593242645263672,
"step": 9640
},
{
"epoch": 0.8955638993185927,
"grad_norm": 0.044748109156357865,
"learning_rate": 2.840468603731189e-06,
"loss": 0.05731314420700073,
"step": 9660
},
{
"epoch": 0.8974180688823993,
"grad_norm": 0.04319157638334523,
"learning_rate": 2.7415584026312223e-06,
"loss": 0.058256596326828,
"step": 9680
},
{
"epoch": 0.8992722384462059,
"grad_norm": 0.04388308005862721,
"learning_rate": 2.644352420430246e-06,
"loss": 0.05611896514892578,
"step": 9700
},
{
"epoch": 0.9011264080100125,
"grad_norm": 0.04519528134021101,
"learning_rate": 2.5488541625394933e-06,
"loss": 0.05943126082420349,
"step": 9720
},
{
"epoch": 0.9029805775738191,
"grad_norm": 0.04251086029166456,
"learning_rate": 2.4550670727867763e-06,
"loss": 0.05685185194015503,
"step": 9740
},
{
"epoch": 0.9048347471376257,
"grad_norm": 0.03944944934977318,
"learning_rate": 2.362994533292323e-06,
"loss": 0.05601794123649597,
"step": 9760
},
{
"epoch": 0.9066889167014324,
"grad_norm": 0.044177980988593026,
"learning_rate": 2.272639864346776e-06,
"loss": 0.05724194049835205,
"step": 9780
},
{
"epoch": 0.9085430862652389,
"grad_norm": 0.04189831748981764,
"learning_rate": 2.1840063242915045e-06,
"loss": 0.056692618131637576,
"step": 9800
},
{
"epoch": 0.9103972558290455,
"grad_norm": 0.043673059482412624,
"learning_rate": 2.0970971094010305e-06,
"loss": 0.05840623378753662,
"step": 9820
},
{
"epoch": 0.9122514253928522,
"grad_norm": 0.045504438871193945,
"learning_rate": 2.0119153537678513e-06,
"loss": 0.05786421298980713,
"step": 9840
},
{
"epoch": 0.9141055949566588,
"grad_norm": 0.04557063826056399,
"learning_rate": 1.9284641291893615e-06,
"loss": 0.057442504167556765,
"step": 9860
},
{
"epoch": 0.9159597645204653,
"grad_norm": 0.04545043254971616,
"learning_rate": 1.8467464450570903e-06,
"loss": 0.05783541202545166,
"step": 9880
},
{
"epoch": 0.917813934084272,
"grad_norm": 0.04107832026051421,
"learning_rate": 1.7667652482481922e-06,
"loss": 0.058336269855499265,
"step": 9900
},
{
"epoch": 0.9196681036480786,
"grad_norm": 0.041143537773437405,
"learning_rate": 1.6885234230191615e-06,
"loss": 0.05814253091812134,
"step": 9920
},
{
"epoch": 0.9215222732118852,
"grad_norm": 0.04132684264242279,
"learning_rate": 1.6120237909018287e-06,
"loss": 0.055842363834381105,
"step": 9940
},
{
"epoch": 0.9233764427756919,
"grad_norm": 0.043378747724282775,
"learning_rate": 1.5372691106016135e-06,
"loss": 0.05697543621063232,
"step": 9960
},
{
"epoch": 0.9252306123394984,
"grad_norm": 0.043723662998569265,
"learning_rate": 1.464262077898032e-06,
"loss": 0.05675085783004761,
"step": 9980
},
{
"epoch": 0.927084781903305,
"grad_norm": 0.0413347860452099,
"learning_rate": 1.3930053255474863e-06,
"loss": 0.05656332969665527,
"step": 10000
},
{
"epoch": 0.9289389514671117,
"grad_norm": 0.042641886326804886,
"learning_rate": 1.3235014231883391e-06,
"loss": 0.05754779577255249,
"step": 10020
},
{
"epoch": 0.9307931210309183,
"grad_norm": 0.04240135883945068,
"learning_rate": 1.2557528772482175e-06,
"loss": 0.05811010599136353,
"step": 10040
},
{
"epoch": 0.9326472905947248,
"grad_norm": 0.04366464083419553,
"learning_rate": 1.1897621308536556e-06,
"loss": 0.057920646667480466,
"step": 10060
},
{
"epoch": 0.9345014601585315,
"grad_norm": 0.04251203177049702,
"learning_rate": 1.125531563741966e-06,
"loss": 0.05935220718383789,
"step": 10080
},
{
"epoch": 0.9363556297223381,
"grad_norm": 0.043375762175896594,
"learning_rate": 1.0630634921754478e-06,
"loss": 0.05946972370147705,
"step": 10100
},
{
"epoch": 0.9382097992861447,
"grad_norm": 0.038546399018732955,
"learning_rate": 1.0023601688578467e-06,
"loss": 0.05542746782302856,
"step": 10120
},
{
"epoch": 0.9400639688499514,
"grad_norm": 0.046981608590331715,
"learning_rate": 9.434237828531034e-07,
"loss": 0.058707261085510255,
"step": 10140
},
{
"epoch": 0.9419181384137579,
"grad_norm": 0.04263830532013562,
"learning_rate": 8.862564595064349e-07,
"loss": 0.05690791606903076,
"step": 10160
},
{
"epoch": 0.9437723079775645,
"grad_norm": 0.04582874833814285,
"learning_rate": 8.30860260367694e-07,
"loss": 0.054658901691436765,
"step": 10180
},
{
"epoch": 0.9456264775413712,
"grad_norm": 0.048305071305060066,
"learning_rate": 7.772371831170022e-07,
"loss": 0.05803183317184448,
"step": 10200
},
{
"epoch": 0.9474806471051778,
"grad_norm": 0.0424560803178508,
"learning_rate": 7.253891614927177e-07,
"loss": 0.05667737126350403,
"step": 10220
},
{
"epoch": 0.9493348166689843,
"grad_norm": 0.04137248823550589,
"learning_rate": 6.753180652217306e-07,
"loss": 0.05827021598815918,
"step": 10240
},
{
"epoch": 0.951188986232791,
"grad_norm": 0.04384442951859904,
"learning_rate": 6.270256999519885e-07,
"loss": 0.05811399817466736,
"step": 10260
},
{
"epoch": 0.9530431557965976,
"grad_norm": 0.04076128530415887,
"learning_rate": 5.805138071874161e-07,
"loss": 0.05737486481666565,
"step": 10280
},
{
"epoch": 0.9548973253604042,
"grad_norm": 0.04109366675767673,
"learning_rate": 5.357840642251089e-07,
"loss": 0.05651825666427612,
"step": 10300
},
{
"epoch": 0.9567514949242109,
"grad_norm": 0.042700389903543114,
"learning_rate": 4.928380840948377e-07,
"loss": 0.05737701654434204,
"step": 10320
},
{
"epoch": 0.9586056644880174,
"grad_norm": 0.038933652888396515,
"learning_rate": 4.5167741550088406e-07,
"loss": 0.05547755360603333,
"step": 10340
},
{
"epoch": 0.960459834051824,
"grad_norm": 0.04499435022161428,
"learning_rate": 4.12303542766207e-07,
"loss": 0.056265252828598025,
"step": 10360
},
{
"epoch": 0.9623140036156307,
"grad_norm": 0.052695850731408304,
"learning_rate": 3.747178857788802e-07,
"loss": 0.05831778049468994,
"step": 10380
},
{
"epoch": 0.9641681731794373,
"grad_norm": 0.041355706582808766,
"learning_rate": 3.389217999409222e-07,
"loss": 0.05711444616317749,
"step": 10400
},
{
"epoch": 0.9660223427432438,
"grad_norm": 0.04541694220798916,
"learning_rate": 3.0491657611939615e-07,
"loss": 0.05954468250274658,
"step": 10420
},
{
"epoch": 0.9678765123070505,
"grad_norm": 0.037048819417347126,
"learning_rate": 2.7270344059988074e-07,
"loss": 0.058045947551727296,
"step": 10440
},
{
"epoch": 0.9697306818708571,
"grad_norm": 0.04118746781043515,
"learning_rate": 2.4228355504222756e-07,
"loss": 0.05838167667388916,
"step": 10460
},
{
"epoch": 0.9715848514346637,
"grad_norm": 0.04512917614986576,
"learning_rate": 2.1365801643868366e-07,
"loss": 0.05804737210273743,
"step": 10480
},
{
"epoch": 0.9734390209984704,
"grad_norm": 0.04157722297753857,
"learning_rate": 1.8682785707431205e-07,
"loss": 0.05664463639259339,
"step": 10500
},
{
"epoch": 0.9752931905622769,
"grad_norm": 0.042394494386684016,
"learning_rate": 1.6179404448980474e-07,
"loss": 0.056443202495574954,
"step": 10520
},
{
"epoch": 0.9771473601260835,
"grad_norm": 0.04209941914717253,
"learning_rate": 1.38557481446544e-07,
"loss": 0.0574646532535553,
"step": 10540
},
{
"epoch": 0.9790015296898902,
"grad_norm": 0.0368439030248634,
"learning_rate": 1.1711900589408409e-07,
"loss": 0.05568208694458008,
"step": 10560
},
{
"epoch": 0.9808556992536968,
"grad_norm": 0.043675112327923045,
"learning_rate": 9.74793909399141e-08,
"loss": 0.05898926854133606,
"step": 10580
},
{
"epoch": 0.9827098688175033,
"grad_norm": 0.043385754999838926,
"learning_rate": 7.963934482158597e-08,
"loss": 0.05666929483413696,
"step": 10600
},
{
"epoch": 0.98456403838131,
"grad_norm": 0.042106162195492784,
"learning_rate": 6.359951088117933e-08,
"loss": 0.05665904879570007,
"step": 10620
},
{
"epoch": 0.9864182079451166,
"grad_norm": 0.04469787281700635,
"learning_rate": 4.936046754207557e-08,
"loss": 0.05817822813987732,
"step": 10640
},
{
"epoch": 0.9882723775089232,
"grad_norm": 0.04979760711070391,
"learning_rate": 3.69227282881357e-08,
"loss": 0.057973289489746095,
"step": 10660
},
{
"epoch": 0.9901265470727298,
"grad_norm": 0.043219446213679145,
"learning_rate": 2.628674164515954e-08,
"loss": 0.057922327518463136,
"step": 10680
},
{
"epoch": 0.9919807166365364,
"grad_norm": 0.03814284539819102,
"learning_rate": 1.745289116471538e-08,
"loss": 0.05686279535293579,
"step": 10700
},
{
"epoch": 0.993834886200343,
"grad_norm": 0.04435709286094493,
"learning_rate": 1.042149541031212e-08,
"loss": 0.05729717016220093,
"step": 10720
},
{
"epoch": 0.9956890557641497,
"grad_norm": 0.043708263904856166,
"learning_rate": 5.192807945902934e-09,
"loss": 0.05722014904022217,
"step": 10740
},
{
"epoch": 0.9975432253279563,
"grad_norm": 0.04332449630622139,
"learning_rate": 1.767017326764764e-09,
"loss": 0.059349268674850464,
"step": 10760
},
{
"epoch": 0.9993973948917628,
"grad_norm": 0.0389811994209693,
"learning_rate": 1.4424709265936642e-10,
"loss": 0.056857752799987796,
"step": 10780
},
{
"epoch": 1.0,
"step": 10787,
"total_flos": 5228992760119296.0,
"train_loss": 0.07239722475082117,
"train_runtime": 152754.4145,
"train_samples_per_second": 1.13,
"train_steps_per_second": 0.071
}
],
"logging_steps": 20,
"max_steps": 10787,
"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": 2,
"trial_name": null,
"trial_params": null
}