u3 / trainer_state.json
Kaipengm2's picture
Upload folder using huggingface_hub
4d81d8c verified
Raw
History Blame Contribute Delete
88.7 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2359,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00211954217888936,
"grad_norm": 1.4533849954605103,
"learning_rate": 8.450704225352112e-06,
"loss": 2.911361312866211,
"step": 5
},
{
"epoch": 0.00423908435777872,
"grad_norm": 0.8717721700668335,
"learning_rate": 1.901408450704225e-05,
"loss": 2.89677734375,
"step": 10
},
{
"epoch": 0.00635862653666808,
"grad_norm": 0.4050150513648987,
"learning_rate": 2.957746478873239e-05,
"loss": 2.620399284362793,
"step": 15
},
{
"epoch": 0.00847816871555744,
"grad_norm": 0.25955942273139954,
"learning_rate": 4.0140845070422526e-05,
"loss": 2.527699661254883,
"step": 20
},
{
"epoch": 0.0105977108944468,
"grad_norm": 0.3238535225391388,
"learning_rate": 5.070422535211267e-05,
"loss": 2.4132320404052736,
"step": 25
},
{
"epoch": 0.01271725307333616,
"grad_norm": 0.2910937964916229,
"learning_rate": 6.126760563380281e-05,
"loss": 2.3594972610473635,
"step": 30
},
{
"epoch": 0.01483679525222552,
"grad_norm": 0.18983197212219238,
"learning_rate": 7.183098591549294e-05,
"loss": 2.244922637939453,
"step": 35
},
{
"epoch": 0.01695633743111488,
"grad_norm": 0.21697290241718292,
"learning_rate": 8.23943661971831e-05,
"loss": 2.1240062713623047,
"step": 40
},
{
"epoch": 0.01907587961000424,
"grad_norm": 0.1837455779314041,
"learning_rate": 9.295774647887322e-05,
"loss": 2.0776336669921873,
"step": 45
},
{
"epoch": 0.0211954217888936,
"grad_norm": 0.2103322446346283,
"learning_rate": 0.00010352112676056336,
"loss": 1.9948764801025392,
"step": 50
},
{
"epoch": 0.02331496396778296,
"grad_norm": 0.18801350891590118,
"learning_rate": 0.00011408450704225351,
"loss": 1.966363525390625,
"step": 55
},
{
"epoch": 0.02543450614667232,
"grad_norm": 0.18234486877918243,
"learning_rate": 0.00012464788732394365,
"loss": 1.9284397125244142,
"step": 60
},
{
"epoch": 0.02755404832556168,
"grad_norm": 0.21645711362361908,
"learning_rate": 0.0001352112676056338,
"loss": 1.8941230773925781,
"step": 65
},
{
"epoch": 0.02967359050445104,
"grad_norm": 0.20508678257465363,
"learning_rate": 0.00014577464788732394,
"loss": 1.8263919830322266,
"step": 70
},
{
"epoch": 0.0317931326833404,
"grad_norm": 0.21742308139801025,
"learning_rate": 0.00014999936370154574,
"loss": 1.8043691635131835,
"step": 75
},
{
"epoch": 0.03391267486222976,
"grad_norm": 0.1825699508190155,
"learning_rate": 0.00014999547525009093,
"loss": 1.771235466003418,
"step": 80
},
{
"epoch": 0.03603221704111912,
"grad_norm": 0.180069237947464,
"learning_rate": 0.00014998805202937523,
"loss": 1.7630111694335937,
"step": 85
},
{
"epoch": 0.03815175922000848,
"grad_norm": 0.23649078607559204,
"learning_rate": 0.00014997709438927803,
"loss": 1.73526611328125,
"step": 90
},
{
"epoch": 0.04027130139889784,
"grad_norm": 0.24478007853031158,
"learning_rate": 0.00014996260284626687,
"loss": 1.702292251586914,
"step": 95
},
{
"epoch": 0.0423908435777872,
"grad_norm": 0.21304433047771454,
"learning_rate": 0.00014994457808337308,
"loss": 1.6731922149658203,
"step": 100
},
{
"epoch": 0.04451038575667656,
"grad_norm": 0.23145578801631927,
"learning_rate": 0.00014992302095015968,
"loss": 1.6981586456298827,
"step": 105
},
{
"epoch": 0.04662992793556592,
"grad_norm": 0.23897255957126617,
"learning_rate": 0.00014989793246268114,
"loss": 1.6977878570556642,
"step": 110
},
{
"epoch": 0.048749470114455275,
"grad_norm": 0.19703088700771332,
"learning_rate": 0.0001498693138034357,
"loss": 1.676886749267578,
"step": 115
},
{
"epoch": 0.05086901229334464,
"grad_norm": 0.2128562331199646,
"learning_rate": 0.00014983716632130942,
"loss": 1.6497159957885743,
"step": 120
},
{
"epoch": 0.052988554472233995,
"grad_norm": 0.21524904668331146,
"learning_rate": 0.0001498014915315129,
"loss": 1.688668441772461,
"step": 125
},
{
"epoch": 0.05510809665112336,
"grad_norm": 0.22113725543022156,
"learning_rate": 0.00014976229111550953,
"loss": 1.6296710968017578,
"step": 130
},
{
"epoch": 0.057227638830012716,
"grad_norm": 0.22267888486385345,
"learning_rate": 0.00014971956692093645,
"loss": 1.607334518432617,
"step": 135
},
{
"epoch": 0.05934718100890208,
"grad_norm": 0.2725050747394562,
"learning_rate": 0.00014967332096151742,
"loss": 1.6353057861328124,
"step": 140
},
{
"epoch": 0.061466723187791436,
"grad_norm": 0.24827592074871063,
"learning_rate": 0.0001496235554169679,
"loss": 1.5972054481506348,
"step": 145
},
{
"epoch": 0.0635862653666808,
"grad_norm": 0.2142190933227539,
"learning_rate": 0.0001495702726328923,
"loss": 1.5774917602539062,
"step": 150
},
{
"epoch": 0.06570580754557016,
"grad_norm": 0.20981644093990326,
"learning_rate": 0.00014951347512067337,
"loss": 1.6079927444458009,
"step": 155
},
{
"epoch": 0.06782534972445951,
"grad_norm": 0.20830883085727692,
"learning_rate": 0.00014945316555735403,
"loss": 1.6010347366333009,
"step": 160
},
{
"epoch": 0.06994489190334888,
"grad_norm": 0.21328674256801605,
"learning_rate": 0.00014938934678551097,
"loss": 1.5738430976867677,
"step": 165
},
{
"epoch": 0.07206443408223824,
"grad_norm": 0.21209032833576202,
"learning_rate": 0.0001493220218131207,
"loss": 1.5949657440185547,
"step": 170
},
{
"epoch": 0.07418397626112759,
"grad_norm": 0.2200002521276474,
"learning_rate": 0.00014925119381341796,
"loss": 1.5288724899291992,
"step": 175
},
{
"epoch": 0.07630351844001695,
"grad_norm": 0.20628951489925385,
"learning_rate": 0.00014917686612474595,
"loss": 1.5350455284118651,
"step": 180
},
{
"epoch": 0.07842306061890632,
"grad_norm": 0.2584723234176636,
"learning_rate": 0.00014909904225039906,
"loss": 1.5907514572143555,
"step": 185
},
{
"epoch": 0.08054260279779568,
"grad_norm": 0.24871385097503662,
"learning_rate": 0.00014901772585845783,
"loss": 1.620928955078125,
"step": 190
},
{
"epoch": 0.08266214497668503,
"grad_norm": 0.2913932204246521,
"learning_rate": 0.00014893292078161586,
"loss": 1.5582874298095704,
"step": 195
},
{
"epoch": 0.0847816871555744,
"grad_norm": 0.20935894548892975,
"learning_rate": 0.00014884463101699937,
"loss": 1.574123001098633,
"step": 200
},
{
"epoch": 0.08690122933446376,
"grad_norm": 0.21228212118148804,
"learning_rate": 0.0001487528607259787,
"loss": 1.5573617935180664,
"step": 205
},
{
"epoch": 0.08902077151335312,
"grad_norm": 0.1916375458240509,
"learning_rate": 0.00014865761423397215,
"loss": 1.5406822204589843,
"step": 210
},
{
"epoch": 0.09114031369224247,
"grad_norm": 0.1980282962322235,
"learning_rate": 0.00014855889603024227,
"loss": 1.5439483642578125,
"step": 215
},
{
"epoch": 0.09325985587113184,
"grad_norm": 0.2618733048439026,
"learning_rate": 0.00014845671076768398,
"loss": 1.5370182037353515,
"step": 220
},
{
"epoch": 0.0953793980500212,
"grad_norm": 0.23130768537521362,
"learning_rate": 0.00014835106326260557,
"loss": 1.480801486968994,
"step": 225
},
{
"epoch": 0.09749894022891055,
"grad_norm": 0.30081960558891296,
"learning_rate": 0.00014824195849450145,
"loss": 1.5637982368469239,
"step": 230
},
{
"epoch": 0.09961848240779991,
"grad_norm": 0.22005034983158112,
"learning_rate": 0.00014812940160581763,
"loss": 1.524923038482666,
"step": 235
},
{
"epoch": 0.10173802458668928,
"grad_norm": 0.20901378989219666,
"learning_rate": 0.00014801339790170923,
"loss": 1.5966582298278809,
"step": 240
},
{
"epoch": 0.10385756676557864,
"grad_norm": 0.22220133244991302,
"learning_rate": 0.00014789395284979049,
"loss": 1.5191898345947266,
"step": 245
},
{
"epoch": 0.10597710894446799,
"grad_norm": 0.21421712636947632,
"learning_rate": 0.000147771072079877,
"loss": 1.5260868072509766,
"step": 250
},
{
"epoch": 0.10809665112335735,
"grad_norm": 0.2327452152967453,
"learning_rate": 0.00014764476138372047,
"loss": 1.5540156364440918,
"step": 255
},
{
"epoch": 0.11021619330224672,
"grad_norm": 0.25371599197387695,
"learning_rate": 0.0001475150267147356,
"loss": 1.5139301300048829,
"step": 260
},
{
"epoch": 0.11233573548113608,
"grad_norm": 0.18664808571338654,
"learning_rate": 0.00014738187418771953,
"loss": 1.564552402496338,
"step": 265
},
{
"epoch": 0.11445527766002543,
"grad_norm": 0.27261000871658325,
"learning_rate": 0.00014724531007856371,
"loss": 1.5058234214782715,
"step": 270
},
{
"epoch": 0.1165748198389148,
"grad_norm": 0.23247990012168884,
"learning_rate": 0.000147105340823958,
"loss": 1.4929523468017578,
"step": 275
},
{
"epoch": 0.11869436201780416,
"grad_norm": 0.22711527347564697,
"learning_rate": 0.0001469619730210873,
"loss": 1.5138790130615234,
"step": 280
},
{
"epoch": 0.12081390419669351,
"grad_norm": 0.16822680830955505,
"learning_rate": 0.00014681521342732067,
"loss": 1.4931448936462401,
"step": 285
},
{
"epoch": 0.12293344637558287,
"grad_norm": 0.19065088033676147,
"learning_rate": 0.00014666506895989276,
"loss": 1.4882875442504884,
"step": 290
},
{
"epoch": 0.12505298855447222,
"grad_norm": 0.2147723287343979,
"learning_rate": 0.00014651154669557783,
"loss": 1.4989696502685548,
"step": 295
},
{
"epoch": 0.1271725307333616,
"grad_norm": 0.2251991331577301,
"learning_rate": 0.00014635465387035615,
"loss": 1.4763580322265626,
"step": 300
},
{
"epoch": 0.12929207291225095,
"grad_norm": 0.2073274552822113,
"learning_rate": 0.00014619439787907302,
"loss": 1.4306302070617676,
"step": 305
},
{
"epoch": 0.13141161509114033,
"grad_norm": 0.1931733638048172,
"learning_rate": 0.00014603078627509016,
"loss": 1.4966580390930175,
"step": 310
},
{
"epoch": 0.13353115727002968,
"grad_norm": 0.21649642288684845,
"learning_rate": 0.0001458638267699297,
"loss": 1.5137117385864258,
"step": 315
},
{
"epoch": 0.13565069944891903,
"grad_norm": 0.2783515453338623,
"learning_rate": 0.00014569352723291077,
"loss": 1.4853612899780273,
"step": 320
},
{
"epoch": 0.1377702416278084,
"grad_norm": 0.23952129483222961,
"learning_rate": 0.00014551989569077856,
"loss": 1.4821554183959962,
"step": 325
},
{
"epoch": 0.13988978380669775,
"grad_norm": 0.23521435260772705,
"learning_rate": 0.0001453429403273259,
"loss": 1.4857416152954102,
"step": 330
},
{
"epoch": 0.1420093259855871,
"grad_norm": 0.21960721909999847,
"learning_rate": 0.00014516266948300774,
"loss": 1.5065743446350097,
"step": 335
},
{
"epoch": 0.14412886816447648,
"grad_norm": 0.21124182641506195,
"learning_rate": 0.00014497909165454788,
"loss": 1.4274757385253907,
"step": 340
},
{
"epoch": 0.14624841034336583,
"grad_norm": 0.22351916134357452,
"learning_rate": 0.0001447922154945385,
"loss": 1.4317169189453125,
"step": 345
},
{
"epoch": 0.14836795252225518,
"grad_norm": 0.21286199986934662,
"learning_rate": 0.00014460204981103244,
"loss": 1.4460826873779298,
"step": 350
},
{
"epoch": 0.15048749470114456,
"grad_norm": 0.22857090830802917,
"learning_rate": 0.00014440860356712792,
"loss": 1.4680397033691406,
"step": 355
},
{
"epoch": 0.1526070368800339,
"grad_norm": 0.24696066975593567,
"learning_rate": 0.00014421188588054618,
"loss": 1.4768699645996093,
"step": 360
},
{
"epoch": 0.15472657905892329,
"grad_norm": 0.19838055968284607,
"learning_rate": 0.00014401190602320168,
"loss": 1.4469011306762696,
"step": 365
},
{
"epoch": 0.15684612123781264,
"grad_norm": 0.2080274522304535,
"learning_rate": 0.0001438086734207651,
"loss": 1.4298072814941407,
"step": 370
},
{
"epoch": 0.15896566341670199,
"grad_norm": 0.2138231247663498,
"learning_rate": 0.00014360219765221905,
"loss": 1.4198074340820312,
"step": 375
},
{
"epoch": 0.16108520559559136,
"grad_norm": 0.21890293061733246,
"learning_rate": 0.00014339248844940676,
"loss": 1.4695507049560548,
"step": 380
},
{
"epoch": 0.1632047477744807,
"grad_norm": 0.21057920157909393,
"learning_rate": 0.000143179555696573,
"loss": 1.4966022491455078,
"step": 385
},
{
"epoch": 0.16532428995337006,
"grad_norm": 0.2032330334186554,
"learning_rate": 0.00014296340942989864,
"loss": 1.459632968902588,
"step": 390
},
{
"epoch": 0.16744383213225944,
"grad_norm": 0.20372611284255981,
"learning_rate": 0.00014274405983702738,
"loss": 1.405313205718994,
"step": 395
},
{
"epoch": 0.1695633743111488,
"grad_norm": 0.23118329048156738,
"learning_rate": 0.00014252151725658555,
"loss": 1.4490900039672852,
"step": 400
},
{
"epoch": 0.17168291649003814,
"grad_norm": 0.20457163453102112,
"learning_rate": 0.00014229579217769493,
"loss": 1.4749466896057128,
"step": 405
},
{
"epoch": 0.17380245866892752,
"grad_norm": 0.20578953623771667,
"learning_rate": 0.00014206689523947832,
"loss": 1.4004918098449708,
"step": 410
},
{
"epoch": 0.17592200084781687,
"grad_norm": 0.2327735424041748,
"learning_rate": 0.0001418348372305581,
"loss": 1.465897560119629,
"step": 415
},
{
"epoch": 0.17804154302670624,
"grad_norm": 0.21945743262767792,
"learning_rate": 0.0001415996290885477,
"loss": 1.3908300399780273,
"step": 420
},
{
"epoch": 0.1801610852055956,
"grad_norm": 0.2154647409915924,
"learning_rate": 0.00014136128189953605,
"loss": 1.4920503616333007,
"step": 425
},
{
"epoch": 0.18228062738448494,
"grad_norm": 0.2337629646062851,
"learning_rate": 0.00014111980689756514,
"loss": 1.4768125534057617,
"step": 430
},
{
"epoch": 0.18440016956337432,
"grad_norm": 0.22426795959472656,
"learning_rate": 0.00014087521546410044,
"loss": 1.4597837448120117,
"step": 435
},
{
"epoch": 0.18651971174226367,
"grad_norm": 0.2243419736623764,
"learning_rate": 0.00014062751912749457,
"loss": 1.4114381790161132,
"step": 440
},
{
"epoch": 0.18863925392115302,
"grad_norm": 0.1989261358976364,
"learning_rate": 0.00014037672956244382,
"loss": 1.4108015060424806,
"step": 445
},
{
"epoch": 0.1907587961000424,
"grad_norm": 0.19225820899009705,
"learning_rate": 0.0001401228585894379,
"loss": 1.4406494140625,
"step": 450
},
{
"epoch": 0.19287833827893175,
"grad_norm": 0.20971807837486267,
"learning_rate": 0.00013986591817420288,
"loss": 1.448514175415039,
"step": 455
},
{
"epoch": 0.1949978804578211,
"grad_norm": 0.20189028978347778,
"learning_rate": 0.0001396059204271371,
"loss": 1.4122689247131348,
"step": 460
},
{
"epoch": 0.19711742263671048,
"grad_norm": 0.19423413276672363,
"learning_rate": 0.00013934287760274051,
"loss": 1.4285457611083985,
"step": 465
},
{
"epoch": 0.19923696481559983,
"grad_norm": 0.20719105005264282,
"learning_rate": 0.00013907680209903695,
"loss": 1.388277244567871,
"step": 470
},
{
"epoch": 0.2013565069944892,
"grad_norm": 0.19392803311347961,
"learning_rate": 0.00013880770645698982,
"loss": 1.4308454513549804,
"step": 475
},
{
"epoch": 0.20347604917337855,
"grad_norm": 0.2348402440547943,
"learning_rate": 0.00013853560335991098,
"loss": 1.4246279716491699,
"step": 480
},
{
"epoch": 0.2055955913522679,
"grad_norm": 0.21878980100154877,
"learning_rate": 0.00013826050563286307,
"loss": 1.433729076385498,
"step": 485
},
{
"epoch": 0.20771513353115728,
"grad_norm": 0.19843058288097382,
"learning_rate": 0.00013798242624205484,
"loss": 1.4005677223205566,
"step": 490
},
{
"epoch": 0.20983467571004663,
"grad_norm": 0.2311900556087494,
"learning_rate": 0.00013770137829423012,
"loss": 1.4019628524780274,
"step": 495
},
{
"epoch": 0.21195421788893598,
"grad_norm": 0.2418936938047409,
"learning_rate": 0.00013741737503605004,
"loss": 1.426509189605713,
"step": 500
},
{
"epoch": 0.21407376006782536,
"grad_norm": 0.2732597589492798,
"learning_rate": 0.0001371304298534687,
"loss": 1.4062681198120117,
"step": 505
},
{
"epoch": 0.2161933022467147,
"grad_norm": 0.2327418476343155,
"learning_rate": 0.00013684055627110215,
"loss": 1.4135140419006347,
"step": 510
},
{
"epoch": 0.21831284442560406,
"grad_norm": 0.20132744312286377,
"learning_rate": 0.00013654776795159104,
"loss": 1.3860734939575194,
"step": 515
},
{
"epoch": 0.22043238660449344,
"grad_norm": 0.2001914381980896,
"learning_rate": 0.00013625207869495662,
"loss": 1.4492733001708984,
"step": 520
},
{
"epoch": 0.22255192878338279,
"grad_norm": 0.26076483726501465,
"learning_rate": 0.0001359535024379503,
"loss": 1.4178794860839843,
"step": 525
},
{
"epoch": 0.22467147096227216,
"grad_norm": 0.2128058224916458,
"learning_rate": 0.00013565205325339678,
"loss": 1.394708251953125,
"step": 530
},
{
"epoch": 0.2267910131411615,
"grad_norm": 0.17924794554710388,
"learning_rate": 0.00013534774534953067,
"loss": 1.3627028465270996,
"step": 535
},
{
"epoch": 0.22891055532005086,
"grad_norm": 0.23135896027088165,
"learning_rate": 0.00013504059306932703,
"loss": 1.394637393951416,
"step": 540
},
{
"epoch": 0.23103009749894024,
"grad_norm": 0.18716266751289368,
"learning_rate": 0.00013473061088982503,
"loss": 1.4000116348266602,
"step": 545
},
{
"epoch": 0.2331496396778296,
"grad_norm": 0.24864238500595093,
"learning_rate": 0.00013441781342144585,
"loss": 1.3741465568542481,
"step": 550
},
{
"epoch": 0.23526918185671894,
"grad_norm": 0.22332683205604553,
"learning_rate": 0.000134102215407304,
"loss": 1.3871079444885255,
"step": 555
},
{
"epoch": 0.23738872403560832,
"grad_norm": 0.23110778629779816,
"learning_rate": 0.00013378383172251232,
"loss": 1.3991371154785157,
"step": 560
},
{
"epoch": 0.23950826621449767,
"grad_norm": 0.17871485650539398,
"learning_rate": 0.000133462677373481,
"loss": 1.4029035568237305,
"step": 565
},
{
"epoch": 0.24162780839338702,
"grad_norm": 0.19950291514396667,
"learning_rate": 0.0001331387674972102,
"loss": 1.3823355674743651,
"step": 570
},
{
"epoch": 0.2437473505722764,
"grad_norm": 0.22258326411247253,
"learning_rate": 0.00013281211736057655,
"loss": 1.4099298477172852,
"step": 575
},
{
"epoch": 0.24586689275116574,
"grad_norm": 0.20416663587093353,
"learning_rate": 0.00013248274235961384,
"loss": 1.4341743469238282,
"step": 580
},
{
"epoch": 0.24798643493005512,
"grad_norm": 0.1936974674463272,
"learning_rate": 0.00013215065801878696,
"loss": 1.389725112915039,
"step": 585
},
{
"epoch": 0.25010597710894444,
"grad_norm": 0.21888478100299835,
"learning_rate": 0.00013181587999026048,
"loss": 1.3756709098815918,
"step": 590
},
{
"epoch": 0.2522255192878338,
"grad_norm": 0.24368789792060852,
"learning_rate": 0.00013147842405316084,
"loss": 1.3930557250976563,
"step": 595
},
{
"epoch": 0.2543450614667232,
"grad_norm": 0.25530099868774414,
"learning_rate": 0.00013113830611283258,
"loss": 1.4278820037841797,
"step": 600
},
{
"epoch": 0.2564646036456125,
"grad_norm": 0.2082163542509079,
"learning_rate": 0.0001307955422000887,
"loss": 1.3634056091308593,
"step": 605
},
{
"epoch": 0.2585841458245019,
"grad_norm": 0.21113799512386322,
"learning_rate": 0.00013045014847045511,
"loss": 1.3915785789489745,
"step": 610
},
{
"epoch": 0.2607036880033913,
"grad_norm": 0.19068703055381775,
"learning_rate": 0.00013010214120340912,
"loss": 1.3578933715820312,
"step": 615
},
{
"epoch": 0.26282323018228065,
"grad_norm": 0.21349555253982544,
"learning_rate": 0.0001297515368016122,
"loss": 1.3700779914855956,
"step": 620
},
{
"epoch": 0.26494277236117,
"grad_norm": 0.239017516374588,
"learning_rate": 0.00012939835179013676,
"loss": 1.2932103157043457,
"step": 625
},
{
"epoch": 0.26706231454005935,
"grad_norm": 0.2203160524368286,
"learning_rate": 0.0001290426028156874,
"loss": 1.3808835983276366,
"step": 630
},
{
"epoch": 0.26918185671894873,
"grad_norm": 0.2005006968975067,
"learning_rate": 0.00012868430664581625,
"loss": 1.3864688873291016,
"step": 635
},
{
"epoch": 0.27130139889783805,
"grad_norm": 0.21884654462337494,
"learning_rate": 0.00012832348016813262,
"loss": 1.3849655151367188,
"step": 640
},
{
"epoch": 0.27342094107672743,
"grad_norm": 0.21828524768352509,
"learning_rate": 0.00012796014038950714,
"loss": 1.337683868408203,
"step": 645
},
{
"epoch": 0.2755404832556168,
"grad_norm": 0.23754437267780304,
"learning_rate": 0.00012759430443527,
"loss": 1.3622608184814453,
"step": 650
},
{
"epoch": 0.27766002543450613,
"grad_norm": 0.26920199394226074,
"learning_rate": 0.00012722598954840395,
"loss": 1.3621455192565919,
"step": 655
},
{
"epoch": 0.2797795676133955,
"grad_norm": 0.20296823978424072,
"learning_rate": 0.00012685521308873157,
"loss": 1.3798117637634277,
"step": 660
},
{
"epoch": 0.2818991097922849,
"grad_norm": 0.2063087821006775,
"learning_rate": 0.00012648199253209686,
"loss": 1.3982159614562988,
"step": 665
},
{
"epoch": 0.2840186519711742,
"grad_norm": 0.24138645827770233,
"learning_rate": 0.00012610634546954182,
"loss": 1.4047200202941894,
"step": 670
},
{
"epoch": 0.2861381941500636,
"grad_norm": 0.22688671946525574,
"learning_rate": 0.00012572828960647715,
"loss": 1.3494701385498047,
"step": 675
},
{
"epoch": 0.28825773632895296,
"grad_norm": 0.21970339119434357,
"learning_rate": 0.00012534784276184777,
"loss": 1.3665117263793944,
"step": 680
},
{
"epoch": 0.2903772785078423,
"grad_norm": 0.2116941660642624,
"learning_rate": 0.000124965022867293,
"loss": 1.3794803619384766,
"step": 685
},
{
"epoch": 0.29249682068673166,
"grad_norm": 0.18924342095851898,
"learning_rate": 0.0001245798479663013,
"loss": 1.3910794258117676,
"step": 690
},
{
"epoch": 0.29461636286562104,
"grad_norm": 0.18736907839775085,
"learning_rate": 0.00012419233621335996,
"loss": 1.4316083908081054,
"step": 695
},
{
"epoch": 0.29673590504451036,
"grad_norm": 0.26405763626098633,
"learning_rate": 0.00012380250587309933,
"loss": 1.3403692245483398,
"step": 700
},
{
"epoch": 0.29885544722339974,
"grad_norm": 0.21924880146980286,
"learning_rate": 0.00012341037531943196,
"loss": 1.373030948638916,
"step": 705
},
{
"epoch": 0.3009749894022891,
"grad_norm": 0.20239311456680298,
"learning_rate": 0.00012301596303468658,
"loss": 1.361757183074951,
"step": 710
},
{
"epoch": 0.30309453158117844,
"grad_norm": 0.22322991490364075,
"learning_rate": 0.00012261928760873698,
"loss": 1.3591099739074708,
"step": 715
},
{
"epoch": 0.3052140737600678,
"grad_norm": 0.24470502138137817,
"learning_rate": 0.00012222036773812584,
"loss": 1.3124917984008788,
"step": 720
},
{
"epoch": 0.3073336159389572,
"grad_norm": 0.21314536035060883,
"learning_rate": 0.00012181922222518343,
"loss": 1.3787457466125488,
"step": 725
},
{
"epoch": 0.30945315811784657,
"grad_norm": 0.20746316015720367,
"learning_rate": 0.00012141586997714146,
"loss": 1.3208590507507325,
"step": 730
},
{
"epoch": 0.3115727002967359,
"grad_norm": 0.20587608218193054,
"learning_rate": 0.00012101033000524195,
"loss": 1.3260491371154786,
"step": 735
},
{
"epoch": 0.31369224247562527,
"grad_norm": 0.23164571821689606,
"learning_rate": 0.00012060262142384106,
"loss": 1.3296494483947754,
"step": 740
},
{
"epoch": 0.31581178465451465,
"grad_norm": 0.24576550722122192,
"learning_rate": 0.00012019276344950829,
"loss": 1.3644144058227539,
"step": 745
},
{
"epoch": 0.31793132683340397,
"grad_norm": 0.24007348716259003,
"learning_rate": 0.00011978077540012067,
"loss": 1.3857242584228515,
"step": 750
},
{
"epoch": 0.32005086901229335,
"grad_norm": 0.23115232586860657,
"learning_rate": 0.00011936667669395227,
"loss": 1.3476673126220704,
"step": 755
},
{
"epoch": 0.3221704111911827,
"grad_norm": 0.2215014100074768,
"learning_rate": 0.00011895048684875896,
"loss": 1.293463134765625,
"step": 760
},
{
"epoch": 0.32428995337007205,
"grad_norm": 0.18449951708316803,
"learning_rate": 0.0001185322254808585,
"loss": 1.370039176940918,
"step": 765
},
{
"epoch": 0.3264094955489614,
"grad_norm": 0.22450090944766998,
"learning_rate": 0.00011811191230420588,
"loss": 1.3402008056640624,
"step": 770
},
{
"epoch": 0.3285290377278508,
"grad_norm": 0.20768851041793823,
"learning_rate": 0.00011768956712946427,
"loss": 1.3732899665832519,
"step": 775
},
{
"epoch": 0.3306485799067401,
"grad_norm": 0.21607705950737,
"learning_rate": 0.00011726520986307116,
"loss": 1.4038126945495606,
"step": 780
},
{
"epoch": 0.3327681220856295,
"grad_norm": 0.19296714663505554,
"learning_rate": 0.00011683886050630015,
"loss": 1.3047401428222656,
"step": 785
},
{
"epoch": 0.3348876642645189,
"grad_norm": 0.24172437191009521,
"learning_rate": 0.00011641053915431831,
"loss": 1.3391592979431153,
"step": 790
},
{
"epoch": 0.3370072064434082,
"grad_norm": 0.2072303295135498,
"learning_rate": 0.00011598026599523885,
"loss": 1.3352015495300293,
"step": 795
},
{
"epoch": 0.3391267486222976,
"grad_norm": 0.21765024960041046,
"learning_rate": 0.00011554806130916982,
"loss": 1.3156769752502442,
"step": 800
},
{
"epoch": 0.34124629080118696,
"grad_norm": 0.2462266981601715,
"learning_rate": 0.00011511394546725801,
"loss": 1.3615104675292968,
"step": 805
},
{
"epoch": 0.3433658329800763,
"grad_norm": 0.2640639543533325,
"learning_rate": 0.00011467793893072894,
"loss": 1.354515838623047,
"step": 810
},
{
"epoch": 0.34548537515896566,
"grad_norm": 0.1990615427494049,
"learning_rate": 0.00011424006224992253,
"loss": 1.3528491020202638,
"step": 815
},
{
"epoch": 0.34760491733785503,
"grad_norm": 0.2158903330564499,
"learning_rate": 0.00011380033606332425,
"loss": 1.3205657005310059,
"step": 820
},
{
"epoch": 0.34972445951674436,
"grad_norm": 0.18695850670337677,
"learning_rate": 0.0001133587810965926,
"loss": 1.367375659942627,
"step": 825
},
{
"epoch": 0.35184400169563373,
"grad_norm": 0.26189661026000977,
"learning_rate": 0.00011291541816158217,
"loss": 1.3225163459777831,
"step": 830
},
{
"epoch": 0.3539635438745231,
"grad_norm": 0.2998012900352478,
"learning_rate": 0.0001124702681553626,
"loss": 1.2807294845581054,
"step": 835
},
{
"epoch": 0.3560830860534125,
"grad_norm": 0.18943734467029572,
"learning_rate": 0.00011202335205923388,
"loss": 1.343043327331543,
"step": 840
},
{
"epoch": 0.3582026282323018,
"grad_norm": 0.24869421124458313,
"learning_rate": 0.00011157469093773716,
"loss": 1.325399112701416,
"step": 845
},
{
"epoch": 0.3603221704111912,
"grad_norm": 0.2146313637495041,
"learning_rate": 0.00011112430593766214,
"loss": 1.304315185546875,
"step": 850
},
{
"epoch": 0.36244171259008057,
"grad_norm": 0.24832850694656372,
"learning_rate": 0.0001106722182870502,
"loss": 1.309903907775879,
"step": 855
},
{
"epoch": 0.3645612547689699,
"grad_norm": 0.18670570850372314,
"learning_rate": 0.00011021844929419397,
"loss": 1.314599609375,
"step": 860
},
{
"epoch": 0.36668079694785927,
"grad_norm": 0.21524663269519806,
"learning_rate": 0.00010976302034663289,
"loss": 1.3505349159240723,
"step": 865
},
{
"epoch": 0.36880033912674864,
"grad_norm": 0.21804650127887726,
"learning_rate": 0.0001093059529101453,
"loss": 1.3111852645874023,
"step": 870
},
{
"epoch": 0.37091988130563797,
"grad_norm": 0.20508825778961182,
"learning_rate": 0.00010884726852773652,
"loss": 1.3346372604370118,
"step": 875
},
{
"epoch": 0.37303942348452734,
"grad_norm": 0.20978358387947083,
"learning_rate": 0.00010838698881862356,
"loss": 1.369366455078125,
"step": 880
},
{
"epoch": 0.3751589656634167,
"grad_norm": 0.2045287787914276,
"learning_rate": 0.00010792513547721618,
"loss": 1.3833175659179688,
"step": 885
},
{
"epoch": 0.37727850784230604,
"grad_norm": 0.21166202425956726,
"learning_rate": 0.00010746173027209421,
"loss": 1.3038890838623047,
"step": 890
},
{
"epoch": 0.3793980500211954,
"grad_norm": 0.24340365827083588,
"learning_rate": 0.00010699679504498176,
"loss": 1.2533443450927735,
"step": 895
},
{
"epoch": 0.3815175922000848,
"grad_norm": 0.225352481007576,
"learning_rate": 0.0001065303517097175,
"loss": 1.2584978103637696,
"step": 900
},
{
"epoch": 0.3836371343789741,
"grad_norm": 0.25294095277786255,
"learning_rate": 0.00010606242225122197,
"loss": 1.328216552734375,
"step": 905
},
{
"epoch": 0.3857566765578635,
"grad_norm": 0.19258682429790497,
"learning_rate": 0.00010559302872446128,
"loss": 1.3038272857666016,
"step": 910
},
{
"epoch": 0.3878762187367529,
"grad_norm": 0.2542521059513092,
"learning_rate": 0.00010512219325340766,
"loss": 1.2768152236938477,
"step": 915
},
{
"epoch": 0.3899957609156422,
"grad_norm": 0.23314662277698517,
"learning_rate": 0.0001046499380299966,
"loss": 1.3441388130187988,
"step": 920
},
{
"epoch": 0.3921153030945316,
"grad_norm": 0.24483489990234375,
"learning_rate": 0.00010417628531308094,
"loss": 1.3174713134765625,
"step": 925
},
{
"epoch": 0.39423484527342095,
"grad_norm": 0.24751359224319458,
"learning_rate": 0.00010370125742738173,
"loss": 1.3388001441955566,
"step": 930
},
{
"epoch": 0.3963543874523103,
"grad_norm": 0.253334105014801,
"learning_rate": 0.00010322487676243594,
"loss": 1.349887752532959,
"step": 935
},
{
"epoch": 0.39847392963119965,
"grad_norm": 0.20523115992546082,
"learning_rate": 0.00010274716577154131,
"loss": 1.291508674621582,
"step": 940
},
{
"epoch": 0.40059347181008903,
"grad_norm": 0.23577848076820374,
"learning_rate": 0.00010226814697069786,
"loss": 1.2507284164428711,
"step": 945
},
{
"epoch": 0.4027130139889784,
"grad_norm": 0.2180488258600235,
"learning_rate": 0.00010178784293754685,
"loss": 1.2895357131958007,
"step": 950
},
{
"epoch": 0.40483255616786773,
"grad_norm": 0.23442329466342926,
"learning_rate": 0.00010130627631030643,
"loss": 1.351835346221924,
"step": 955
},
{
"epoch": 0.4069520983467571,
"grad_norm": 0.23377206921577454,
"learning_rate": 0.00010082346978670478,
"loss": 1.3373378753662108,
"step": 960
},
{
"epoch": 0.4090716405256465,
"grad_norm": 0.21928517520427704,
"learning_rate": 0.00010033944612291028,
"loss": 1.346639633178711,
"step": 965
},
{
"epoch": 0.4111911827045358,
"grad_norm": 0.2067331075668335,
"learning_rate": 9.985422813245875e-05,
"loss": 1.3276933670043944,
"step": 970
},
{
"epoch": 0.4133107248834252,
"grad_norm": 0.21390070021152496,
"learning_rate": 9.936783868517849e-05,
"loss": 1.3124217987060547,
"step": 975
},
{
"epoch": 0.41543026706231456,
"grad_norm": 0.19038116931915283,
"learning_rate": 9.888030070611215e-05,
"loss": 1.284699058532715,
"step": 980
},
{
"epoch": 0.4175498092412039,
"grad_norm": 0.3529278635978699,
"learning_rate": 9.839163717443617e-05,
"loss": 1.3291648864746093,
"step": 985
},
{
"epoch": 0.41966935142009326,
"grad_norm": 0.30006375908851624,
"learning_rate": 9.790187112237786e-05,
"loss": 1.3090578079223634,
"step": 990
},
{
"epoch": 0.42178889359898264,
"grad_norm": 0.23081862926483154,
"learning_rate": 9.741102563412969e-05,
"loss": 1.3270970344543458,
"step": 995
},
{
"epoch": 0.42390843577787196,
"grad_norm": 0.21247537434101105,
"learning_rate": 9.691912384476128e-05,
"loss": 1.3049468040466308,
"step": 1000
},
{
"epoch": 0.42602797795676134,
"grad_norm": 0.25940096378326416,
"learning_rate": 9.64261889391291e-05,
"loss": 1.3142683029174804,
"step": 1005
},
{
"epoch": 0.4281475201356507,
"grad_norm": 0.21281564235687256,
"learning_rate": 9.593224415078348e-05,
"loss": 1.29989652633667,
"step": 1010
},
{
"epoch": 0.43026706231454004,
"grad_norm": 0.26989611983299255,
"learning_rate": 9.54373127608737e-05,
"loss": 1.3559316635131835,
"step": 1015
},
{
"epoch": 0.4323866044934294,
"grad_norm": 0.20252250134944916,
"learning_rate": 9.494141809705063e-05,
"loss": 1.3052260398864746,
"step": 1020
},
{
"epoch": 0.4345061466723188,
"grad_norm": 0.22775596380233765,
"learning_rate": 9.444458353236725e-05,
"loss": 1.339052391052246,
"step": 1025
},
{
"epoch": 0.4366256888512081,
"grad_norm": 0.2322123795747757,
"learning_rate": 9.394683248417692e-05,
"loss": 1.3342934608459474,
"step": 1030
},
{
"epoch": 0.4387452310300975,
"grad_norm": 0.19912663102149963,
"learning_rate": 9.344818841302977e-05,
"loss": 1.295450782775879,
"step": 1035
},
{
"epoch": 0.44086477320898687,
"grad_norm": 0.2255859524011612,
"learning_rate": 9.294867482156682e-05,
"loss": 1.35633544921875,
"step": 1040
},
{
"epoch": 0.4429843153878762,
"grad_norm": 0.23680174350738525,
"learning_rate": 9.244831525341233e-05,
"loss": 1.27927827835083,
"step": 1045
},
{
"epoch": 0.44510385756676557,
"grad_norm": 0.23993147909641266,
"learning_rate": 9.194713329206404e-05,
"loss": 1.3515398025512695,
"step": 1050
},
{
"epoch": 0.44722339974565495,
"grad_norm": 0.23341205716133118,
"learning_rate": 9.144515255978158e-05,
"loss": 1.323683738708496,
"step": 1055
},
{
"epoch": 0.4493429419245443,
"grad_norm": 0.23291119933128357,
"learning_rate": 9.094239671647324e-05,
"loss": 1.3159423828125,
"step": 1060
},
{
"epoch": 0.45146248410343365,
"grad_norm": 0.2228938341140747,
"learning_rate": 9.043888945858062e-05,
"loss": 1.2429805755615235,
"step": 1065
},
{
"epoch": 0.453582026282323,
"grad_norm": 0.23924370110034943,
"learning_rate": 8.99346545179618e-05,
"loss": 1.2914630889892578,
"step": 1070
},
{
"epoch": 0.4557015684612124,
"grad_norm": 0.23762305080890656,
"learning_rate": 8.942971566077287e-05,
"loss": 1.3357767105102538,
"step": 1075
},
{
"epoch": 0.4578211106401017,
"grad_norm": 0.226181760430336,
"learning_rate": 8.892409668634764e-05,
"loss": 1.2719341278076173,
"step": 1080
},
{
"epoch": 0.4599406528189911,
"grad_norm": 0.2674066722393036,
"learning_rate": 8.841782142607598e-05,
"loss": 1.2833306312561035,
"step": 1085
},
{
"epoch": 0.4620601949978805,
"grad_norm": 0.22831888496875763,
"learning_rate": 8.791091374228057e-05,
"loss": 1.2862982749938965,
"step": 1090
},
{
"epoch": 0.4641797371767698,
"grad_norm": 0.2320777326822281,
"learning_rate": 8.740339752709209e-05,
"loss": 1.2587108612060547,
"step": 1095
},
{
"epoch": 0.4662992793556592,
"grad_norm": 0.19897054135799408,
"learning_rate": 8.689529670132325e-05,
"loss": 1.3042229652404784,
"step": 1100
},
{
"epoch": 0.46841882153454856,
"grad_norm": 0.2363954484462738,
"learning_rate": 8.638663521334127e-05,
"loss": 1.2335227012634278,
"step": 1105
},
{
"epoch": 0.4705383637134379,
"grad_norm": 0.2121724784374237,
"learning_rate": 8.587743703793913e-05,
"loss": 1.2948514938354492,
"step": 1110
},
{
"epoch": 0.47265790589232726,
"grad_norm": 0.22796151041984558,
"learning_rate": 8.536772617520548e-05,
"loss": 1.3376188278198242,
"step": 1115
},
{
"epoch": 0.47477744807121663,
"grad_norm": 0.24525348842144012,
"learning_rate": 8.485752664939361e-05,
"loss": 1.2620004653930663,
"step": 1120
},
{
"epoch": 0.47689699025010596,
"grad_norm": 0.20538678765296936,
"learning_rate": 8.434686250778894e-05,
"loss": 1.297200870513916,
"step": 1125
},
{
"epoch": 0.47901653242899533,
"grad_norm": 0.21144822239875793,
"learning_rate": 8.383575781957571e-05,
"loss": 1.2759347915649415,
"step": 1130
},
{
"epoch": 0.4811360746078847,
"grad_norm": 0.2555171549320221,
"learning_rate": 8.33242366747025e-05,
"loss": 1.2726491928100585,
"step": 1135
},
{
"epoch": 0.48325561678677403,
"grad_norm": 0.27514195442199707,
"learning_rate": 8.281232318274674e-05,
"loss": 1.2626585960388184,
"step": 1140
},
{
"epoch": 0.4853751589656634,
"grad_norm": 0.23342320322990417,
"learning_rate": 8.230004147177844e-05,
"loss": 1.3402240753173829,
"step": 1145
},
{
"epoch": 0.4874947011445528,
"grad_norm": 0.22106461226940155,
"learning_rate": 8.178741568722284e-05,
"loss": 1.211595344543457,
"step": 1150
},
{
"epoch": 0.4896142433234421,
"grad_norm": 0.250161737203598,
"learning_rate": 8.127446999072254e-05,
"loss": 1.2594094276428223,
"step": 1155
},
{
"epoch": 0.4917337855023315,
"grad_norm": 0.2475946843624115,
"learning_rate": 8.07612285589985e-05,
"loss": 1.2840205192565919,
"step": 1160
},
{
"epoch": 0.49385332768122087,
"grad_norm": 0.21844692528247833,
"learning_rate": 8.024771558271066e-05,
"loss": 1.2698797225952148,
"step": 1165
},
{
"epoch": 0.49597286986011024,
"grad_norm": 0.2111423760652542,
"learning_rate": 7.973395526531762e-05,
"loss": 1.2886319160461426,
"step": 1170
},
{
"epoch": 0.49809241203899957,
"grad_norm": 0.215132936835289,
"learning_rate": 7.921997182193601e-05,
"loss": 1.2682478904724122,
"step": 1175
},
{
"epoch": 0.5002119542178889,
"grad_norm": 0.2918669879436493,
"learning_rate": 7.870578947819903e-05,
"loss": 1.2975797653198242,
"step": 1180
},
{
"epoch": 0.5023314963967783,
"grad_norm": 0.23630967736244202,
"learning_rate": 7.81914324691147e-05,
"loss": 1.3017580032348632,
"step": 1185
},
{
"epoch": 0.5044510385756676,
"grad_norm": 0.20302191376686096,
"learning_rate": 7.767692503792353e-05,
"loss": 1.2496055603027343,
"step": 1190
},
{
"epoch": 0.506570580754557,
"grad_norm": 0.2616707384586334,
"learning_rate": 7.71622914349559e-05,
"loss": 1.31884822845459,
"step": 1195
},
{
"epoch": 0.5086901229334464,
"grad_norm": 0.20567265152931213,
"learning_rate": 7.66475559164891e-05,
"loss": 1.29827880859375,
"step": 1200
},
{
"epoch": 0.5108096651123357,
"grad_norm": 0.24866405129432678,
"learning_rate": 7.613274274360393e-05,
"loss": 1.3497620582580567,
"step": 1205
},
{
"epoch": 0.512929207291225,
"grad_norm": 0.2534573972225189,
"learning_rate": 7.561787618104135e-05,
"loss": 1.2164529800415038,
"step": 1210
},
{
"epoch": 0.5150487494701145,
"grad_norm": 0.22314591705799103,
"learning_rate": 7.510298049605876e-05,
"loss": 1.2759552955627442,
"step": 1215
},
{
"epoch": 0.5171682916490038,
"grad_norm": 0.24239255487918854,
"learning_rate": 7.458807995728608e-05,
"loss": 1.2748052597045898,
"step": 1220
},
{
"epoch": 0.5192878338278932,
"grad_norm": 0.24678169190883636,
"learning_rate": 7.407319883358213e-05,
"loss": 1.2845211029052734,
"step": 1225
},
{
"epoch": 0.5214073760067826,
"grad_norm": 0.27573883533477783,
"learning_rate": 7.355836139289055e-05,
"loss": 1.2725817680358886,
"step": 1230
},
{
"epoch": 0.5235269181856719,
"grad_norm": 0.26353344321250916,
"learning_rate": 7.304359190109611e-05,
"loss": 1.1792613029479981,
"step": 1235
},
{
"epoch": 0.5256464603645613,
"grad_norm": 0.25233304500579834,
"learning_rate": 7.252891462088094e-05,
"loss": 1.2287206649780273,
"step": 1240
},
{
"epoch": 0.5277660025434506,
"grad_norm": 0.2386220246553421,
"learning_rate": 7.201435381058091e-05,
"loss": 1.2566810607910157,
"step": 1245
},
{
"epoch": 0.52988554472234,
"grad_norm": 0.2219053953886032,
"learning_rate": 7.149993372304234e-05,
"loss": 1.3123289108276368,
"step": 1250
},
{
"epoch": 0.5320050869012294,
"grad_norm": 0.24438749253749847,
"learning_rate": 7.098567860447886e-05,
"loss": 1.2317842483520507,
"step": 1255
},
{
"epoch": 0.5341246290801187,
"grad_norm": 0.22169619798660278,
"learning_rate": 7.047161269332854e-05,
"loss": 1.2540470123291017,
"step": 1260
},
{
"epoch": 0.536244171259008,
"grad_norm": 0.22319285571575165,
"learning_rate": 6.995776021911159e-05,
"loss": 1.2298927307128906,
"step": 1265
},
{
"epoch": 0.5383637134378975,
"grad_norm": 0.2211482673883438,
"learning_rate": 6.944414540128821e-05,
"loss": 1.3062271118164062,
"step": 1270
},
{
"epoch": 0.5404832556167868,
"grad_norm": 0.23572184145450592,
"learning_rate": 6.893079244811718e-05,
"loss": 1.244802761077881,
"step": 1275
},
{
"epoch": 0.5426027977956761,
"grad_norm": 0.21840710937976837,
"learning_rate": 6.841772555551478e-05,
"loss": 1.2223194122314454,
"step": 1280
},
{
"epoch": 0.5447223399745655,
"grad_norm": 0.227216899394989,
"learning_rate": 6.790496890591429e-05,
"loss": 1.2200170516967774,
"step": 1285
},
{
"epoch": 0.5468418821534549,
"grad_norm": 0.2458992302417755,
"learning_rate": 6.739254666712638e-05,
"loss": 1.2554248809814452,
"step": 1290
},
{
"epoch": 0.5489614243323442,
"grad_norm": 0.26667773723602295,
"learning_rate": 6.688048299119982e-05,
"loss": 1.2833917617797852,
"step": 1295
},
{
"epoch": 0.5510809665112336,
"grad_norm": 0.18585461378097534,
"learning_rate": 6.636880201328324e-05,
"loss": 1.2331452369689941,
"step": 1300
},
{
"epoch": 0.5532005086901229,
"grad_norm": 0.23805899918079376,
"learning_rate": 6.585752785048753e-05,
"loss": 1.2405157089233398,
"step": 1305
},
{
"epoch": 0.5553200508690123,
"grad_norm": 0.232192724943161,
"learning_rate": 6.53466846007491e-05,
"loss": 1.2438085556030274,
"step": 1310
},
{
"epoch": 0.5574395930479017,
"grad_norm": 0.24143503606319427,
"learning_rate": 6.483629634169413e-05,
"loss": 1.2390453338623046,
"step": 1315
},
{
"epoch": 0.559559135226791,
"grad_norm": 0.23756113648414612,
"learning_rate": 6.432638712950362e-05,
"loss": 1.3378015518188477,
"step": 1320
},
{
"epoch": 0.5616786774056803,
"grad_norm": 0.2451499104499817,
"learning_rate": 6.381698099777967e-05,
"loss": 1.2714644432067872,
"step": 1325
},
{
"epoch": 0.5637982195845698,
"grad_norm": 0.1943550854921341,
"learning_rate": 6.330810195641258e-05,
"loss": 1.227932834625244,
"step": 1330
},
{
"epoch": 0.5659177617634591,
"grad_norm": 0.24250562489032745,
"learning_rate": 6.279977399044927e-05,
"loss": 1.2669570922851563,
"step": 1335
},
{
"epoch": 0.5680373039423484,
"grad_norm": 0.24428966641426086,
"learning_rate": 6.229202105896282e-05,
"loss": 1.2109387397766114,
"step": 1340
},
{
"epoch": 0.5701568461212378,
"grad_norm": 0.2314421534538269,
"learning_rate": 6.178486709392305e-05,
"loss": 1.2569910049438477,
"step": 1345
},
{
"epoch": 0.5722763883001272,
"grad_norm": 0.25165829062461853,
"learning_rate": 6.12783359990687e-05,
"loss": 1.3107463836669921,
"step": 1350
},
{
"epoch": 0.5743959304790165,
"grad_norm": 0.2383921891450882,
"learning_rate": 6.077245164878072e-05,
"loss": 1.2175171852111817,
"step": 1355
},
{
"epoch": 0.5765154726579059,
"grad_norm": 0.19886526465415955,
"learning_rate": 6.0267237886956916e-05,
"loss": 1.239117431640625,
"step": 1360
},
{
"epoch": 0.5786350148367952,
"grad_norm": 0.22061046957969666,
"learning_rate": 5.9762718525888285e-05,
"loss": 1.2453453063964843,
"step": 1365
},
{
"epoch": 0.5807545570156846,
"grad_norm": 0.21772538125514984,
"learning_rate": 5.925891734513644e-05,
"loss": 1.269737434387207,
"step": 1370
},
{
"epoch": 0.582874099194574,
"grad_norm": 0.2243538647890091,
"learning_rate": 5.875585809041301e-05,
"loss": 1.259224033355713,
"step": 1375
},
{
"epoch": 0.5849936413734633,
"grad_norm": 0.2140427827835083,
"learning_rate": 5.825356447246037e-05,
"loss": 1.2117403030395508,
"step": 1380
},
{
"epoch": 0.5871131835523526,
"grad_norm": 0.2444586306810379,
"learning_rate": 5.775206016593396e-05,
"loss": 1.2429166793823243,
"step": 1385
},
{
"epoch": 0.5892327257312421,
"grad_norm": 0.26761749386787415,
"learning_rate": 5.7251368808286616e-05,
"loss": 1.2779346466064454,
"step": 1390
},
{
"epoch": 0.5913522679101314,
"grad_norm": 0.25677070021629333,
"learning_rate": 5.675151399865437e-05,
"loss": 1.2512057304382325,
"step": 1395
},
{
"epoch": 0.5934718100890207,
"grad_norm": 0.20167134702205658,
"learning_rate": 5.625251929674407e-05,
"loss": 1.2714438438415527,
"step": 1400
},
{
"epoch": 0.5955913522679102,
"grad_norm": 0.2455359548330307,
"learning_rate": 5.575440822172314e-05,
"loss": 1.1900412559509277,
"step": 1405
},
{
"epoch": 0.5977108944467995,
"grad_norm": 0.23965312540531158,
"learning_rate": 5.525720425111079e-05,
"loss": 1.2655703544616699,
"step": 1410
},
{
"epoch": 0.5998304366256888,
"grad_norm": 0.237847700715065,
"learning_rate": 5.476093081967172e-05,
"loss": 1.2071823120117187,
"step": 1415
},
{
"epoch": 0.6019499788045782,
"grad_norm": 0.23325993120670319,
"learning_rate": 5.426561131831137e-05,
"loss": 1.2358476638793945,
"step": 1420
},
{
"epoch": 0.6040695209834676,
"grad_norm": 0.21446572244167328,
"learning_rate": 5.3771269092973474e-05,
"loss": 1.297337532043457,
"step": 1425
},
{
"epoch": 0.6061890631623569,
"grad_norm": 0.22668008506298065,
"learning_rate": 5.3277927443539764e-05,
"loss": 1.2532476425170898,
"step": 1430
},
{
"epoch": 0.6083086053412463,
"grad_norm": 0.2606396973133087,
"learning_rate": 5.278560962273172e-05,
"loss": 1.2136953353881836,
"step": 1435
},
{
"epoch": 0.6104281475201356,
"grad_norm": 0.24887678027153015,
"learning_rate": 5.229433883501459e-05,
"loss": 1.2502172470092774,
"step": 1440
},
{
"epoch": 0.612547689699025,
"grad_norm": 0.2320316582918167,
"learning_rate": 5.180413823550372e-05,
"loss": 1.2568342208862304,
"step": 1445
},
{
"epoch": 0.6146672318779144,
"grad_norm": 0.2184932827949524,
"learning_rate": 5.1315030928873136e-05,
"loss": 1.261972713470459,
"step": 1450
},
{
"epoch": 0.6167867740568037,
"grad_norm": 0.23082026839256287,
"learning_rate": 5.0827039968266604e-05,
"loss": 1.2761316299438477,
"step": 1455
},
{
"epoch": 0.6189063162356931,
"grad_norm": 0.24504710733890533,
"learning_rate": 5.0340188354211076e-05,
"loss": 1.2408381462097169,
"step": 1460
},
{
"epoch": 0.6210258584145825,
"grad_norm": 0.24866129457950592,
"learning_rate": 4.98544990335325e-05,
"loss": 1.2138835906982421,
"step": 1465
},
{
"epoch": 0.6231454005934718,
"grad_norm": 0.2694995701313019,
"learning_rate": 4.936999489827438e-05,
"loss": 1.265713882446289,
"step": 1470
},
{
"epoch": 0.6252649427723612,
"grad_norm": 0.20004554092884064,
"learning_rate": 4.888669878461875e-05,
"loss": 1.2647174835205077,
"step": 1475
},
{
"epoch": 0.6273844849512505,
"grad_norm": 0.2345903515815735,
"learning_rate": 4.840463347180982e-05,
"loss": 1.27838077545166,
"step": 1480
},
{
"epoch": 0.6295040271301399,
"grad_norm": 0.20837561786174774,
"learning_rate": 4.7923821681080376e-05,
"loss": 1.2347232818603515,
"step": 1485
},
{
"epoch": 0.6316235693090293,
"grad_norm": 0.2348533421754837,
"learning_rate": 4.7444286074580774e-05,
"loss": 1.28297119140625,
"step": 1490
},
{
"epoch": 0.6337431114879186,
"grad_norm": 0.2925850450992584,
"learning_rate": 4.6966049254310886e-05,
"loss": 1.193382453918457,
"step": 1495
},
{
"epoch": 0.6358626536668079,
"grad_norm": 0.25511837005615234,
"learning_rate": 4.648913376105478e-05,
"loss": 1.2335779190063476,
"step": 1500
},
{
"epoch": 0.6379821958456974,
"grad_norm": 0.2385036200284958,
"learning_rate": 4.601356207331822e-05,
"loss": 1.222652053833008,
"step": 1505
},
{
"epoch": 0.6401017380245867,
"grad_norm": 0.22773680090904236,
"learning_rate": 4.553935660626932e-05,
"loss": 1.2264936447143555,
"step": 1510
},
{
"epoch": 0.642221280203476,
"grad_norm": 0.251341849565506,
"learning_rate": 4.506653971068193e-05,
"loss": 1.2267975807189941,
"step": 1515
},
{
"epoch": 0.6443408223823655,
"grad_norm": 0.24920202791690826,
"learning_rate": 4.4595133671882234e-05,
"loss": 1.2328675270080567,
"step": 1520
},
{
"epoch": 0.6464603645612548,
"grad_norm": 0.2338518351316452,
"learning_rate": 4.412516070869837e-05,
"loss": 1.184193229675293,
"step": 1525
},
{
"epoch": 0.6485799067401441,
"grad_norm": 0.23110844194889069,
"learning_rate": 4.365664297241323e-05,
"loss": 1.2358314514160156,
"step": 1530
},
{
"epoch": 0.6506994489190335,
"grad_norm": 0.2423774003982544,
"learning_rate": 4.318960254572017e-05,
"loss": 1.2614693641662598,
"step": 1535
},
{
"epoch": 0.6528189910979229,
"grad_norm": 0.21875004470348358,
"learning_rate": 4.2724061441682606e-05,
"loss": 1.1697239875793457,
"step": 1540
},
{
"epoch": 0.6549385332768122,
"grad_norm": 0.3008062243461609,
"learning_rate": 4.226004160269605e-05,
"loss": 1.2755146026611328,
"step": 1545
},
{
"epoch": 0.6570580754557016,
"grad_norm": 0.22835373878479004,
"learning_rate": 4.179756489945413e-05,
"loss": 1.2458907127380372,
"step": 1550
},
{
"epoch": 0.6591776176345909,
"grad_norm": 0.23178282380104065,
"learning_rate": 4.1336653129917776e-05,
"loss": 1.222042942047119,
"step": 1555
},
{
"epoch": 0.6612971598134803,
"grad_norm": 0.23778164386749268,
"learning_rate": 4.087732801828764e-05,
"loss": 1.2079696655273438,
"step": 1560
},
{
"epoch": 0.6634167019923697,
"grad_norm": 0.23278649151325226,
"learning_rate": 4.041961121398029e-05,
"loss": 1.276837158203125,
"step": 1565
},
{
"epoch": 0.665536244171259,
"grad_norm": 0.2167709767818451,
"learning_rate": 3.996352429060777e-05,
"loss": 1.2689397811889649,
"step": 1570
},
{
"epoch": 0.6676557863501483,
"grad_norm": 0.23578795790672302,
"learning_rate": 3.950908874496088e-05,
"loss": 1.257314109802246,
"step": 1575
},
{
"epoch": 0.6697753285290378,
"grad_norm": 0.19978851079940796,
"learning_rate": 3.905632599599576e-05,
"loss": 1.273353672027588,
"step": 1580
},
{
"epoch": 0.6718948707079271,
"grad_norm": 0.23701950907707214,
"learning_rate": 3.860525738382446e-05,
"loss": 1.2617359161376953,
"step": 1585
},
{
"epoch": 0.6740144128868164,
"grad_norm": 0.26524561643600464,
"learning_rate": 3.815590416870921e-05,
"loss": 1.244351863861084,
"step": 1590
},
{
"epoch": 0.6761339550657058,
"grad_norm": 0.26031601428985596,
"learning_rate": 3.770828753006016e-05,
"loss": 1.231771469116211,
"step": 1595
},
{
"epoch": 0.6782534972445952,
"grad_norm": 0.27454307675361633,
"learning_rate": 3.726242856543721e-05,
"loss": 1.2163453102111816,
"step": 1600
},
{
"epoch": 0.6803730394234845,
"grad_norm": 0.24329088628292084,
"learning_rate": 3.6818348289555764e-05,
"loss": 1.219782829284668,
"step": 1605
},
{
"epoch": 0.6824925816023739,
"grad_norm": 0.2623275816440582,
"learning_rate": 3.6376067633295996e-05,
"loss": 1.3031278610229493,
"step": 1610
},
{
"epoch": 0.6846121237812632,
"grad_norm": 0.2366313487291336,
"learning_rate": 3.5935607442716435e-05,
"loss": 1.1511903762817384,
"step": 1615
},
{
"epoch": 0.6867316659601526,
"grad_norm": 0.2202633023262024,
"learning_rate": 3.54969884780715e-05,
"loss": 1.2090656280517578,
"step": 1620
},
{
"epoch": 0.688851208139042,
"grad_norm": 0.22059793770313263,
"learning_rate": 3.506023141283283e-05,
"loss": 1.2075151443481444,
"step": 1625
},
{
"epoch": 0.6909707503179313,
"grad_norm": 0.25567343831062317,
"learning_rate": 3.462535683271498e-05,
"loss": 1.2180883407592773,
"step": 1630
},
{
"epoch": 0.6930902924968206,
"grad_norm": 0.23150235414505005,
"learning_rate": 3.4192385234705254e-05,
"loss": 1.2282455444335938,
"step": 1635
},
{
"epoch": 0.6952098346757101,
"grad_norm": 0.235373854637146,
"learning_rate": 3.376133702609739e-05,
"loss": 1.2259226799011231,
"step": 1640
},
{
"epoch": 0.6973293768545994,
"grad_norm": 0.2610265016555786,
"learning_rate": 3.333223252352985e-05,
"loss": 1.172861099243164,
"step": 1645
},
{
"epoch": 0.6994489190334887,
"grad_norm": 0.2560023367404938,
"learning_rate": 3.2905091952028204e-05,
"loss": 1.2130403518676758,
"step": 1650
},
{
"epoch": 0.7015684612123781,
"grad_norm": 0.2423076331615448,
"learning_rate": 3.2479935444051916e-05,
"loss": 1.1915353775024413,
"step": 1655
},
{
"epoch": 0.7036880033912675,
"grad_norm": 0.22340212762355804,
"learning_rate": 3.2056783038545327e-05,
"loss": 1.2548277854919434,
"step": 1660
},
{
"epoch": 0.7058075455701568,
"grad_norm": 0.23918607831001282,
"learning_rate": 3.1635654679993185e-05,
"loss": 1.1839475631713867,
"step": 1665
},
{
"epoch": 0.7079270877490462,
"grad_norm": 0.2531600594520569,
"learning_rate": 3.1216570217480744e-05,
"loss": 1.245924472808838,
"step": 1670
},
{
"epoch": 0.7100466299279355,
"grad_norm": 0.2571106553077698,
"learning_rate": 3.0799549403758e-05,
"loss": 1.2773694038391112,
"step": 1675
},
{
"epoch": 0.712166172106825,
"grad_norm": 0.23295314610004425,
"learning_rate": 3.0384611894308797e-05,
"loss": 1.286403274536133,
"step": 1680
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.20814260840415955,
"learning_rate": 2.997177724642444e-05,
"loss": 1.2674052238464355,
"step": 1685
},
{
"epoch": 0.7164052564646036,
"grad_norm": 0.23304776847362518,
"learning_rate": 2.9561064918281803e-05,
"loss": 1.2546541213989257,
"step": 1690
},
{
"epoch": 0.7185247986434931,
"grad_norm": 0.21517516672611237,
"learning_rate": 2.9152494268026245e-05,
"loss": 1.1234692573547362,
"step": 1695
},
{
"epoch": 0.7206443408223824,
"grad_norm": 0.21827828884124756,
"learning_rate": 2.8746084552859267e-05,
"loss": 1.1751296997070313,
"step": 1700
},
{
"epoch": 0.7227638830012717,
"grad_norm": 0.2263326197862625,
"learning_rate": 2.8341854928130736e-05,
"loss": 1.227060031890869,
"step": 1705
},
{
"epoch": 0.7248834251801611,
"grad_norm": 0.252080500125885,
"learning_rate": 2.793982444643609e-05,
"loss": 1.1479147911071776,
"step": 1710
},
{
"epoch": 0.7270029673590505,
"grad_norm": 0.20595385134220123,
"learning_rate": 2.754001205671842e-05,
"loss": 1.1766535758972168,
"step": 1715
},
{
"epoch": 0.7291225095379398,
"grad_norm": 0.24391108751296997,
"learning_rate": 2.714243660337519e-05,
"loss": 1.2480141639709472,
"step": 1720
},
{
"epoch": 0.7312420517168292,
"grad_norm": 0.2905571162700653,
"learning_rate": 2.6747116825370114e-05,
"loss": 1.2400187492370605,
"step": 1725
},
{
"epoch": 0.7333615938957185,
"grad_norm": 0.215235635638237,
"learning_rate": 2.6354071355350006e-05,
"loss": 1.18489408493042,
"step": 1730
},
{
"epoch": 0.7354811360746079,
"grad_norm": 0.21497538685798645,
"learning_rate": 2.5963318718766428e-05,
"loss": 1.2627497673034669,
"step": 1735
},
{
"epoch": 0.7376006782534973,
"grad_norm": 0.24850307404994965,
"learning_rate": 2.557487733300261e-05,
"loss": 1.194486713409424,
"step": 1740
},
{
"epoch": 0.7397202204323866,
"grad_norm": 0.2618556320667267,
"learning_rate": 2.5188765506505343e-05,
"loss": 1.1494468688964843,
"step": 1745
},
{
"epoch": 0.7418397626112759,
"grad_norm": 0.2557094991207123,
"learning_rate": 2.4805001437922146e-05,
"loss": 1.3124690055847168,
"step": 1750
},
{
"epoch": 0.7439593047901654,
"grad_norm": 0.22282132506370544,
"learning_rate": 2.442360321524335e-05,
"loss": 1.2190415382385253,
"step": 1755
},
{
"epoch": 0.7460788469690547,
"grad_norm": 0.24310137331485748,
"learning_rate": 2.4044588814949623e-05,
"loss": 1.2235212326049805,
"step": 1760
},
{
"epoch": 0.748198389147944,
"grad_norm": 0.25061145424842834,
"learning_rate": 2.3667976101164816e-05,
"loss": 1.2448448181152343,
"step": 1765
},
{
"epoch": 0.7503179313268334,
"grad_norm": 0.24184198677539825,
"learning_rate": 2.3293782824813648e-05,
"loss": 1.2871383666992187,
"step": 1770
},
{
"epoch": 0.7524374735057228,
"grad_norm": 0.25554177165031433,
"learning_rate": 2.292202662278546e-05,
"loss": 1.1916116714477538,
"step": 1775
},
{
"epoch": 0.7545570156846121,
"grad_norm": 0.2480154186487198,
"learning_rate": 2.2552725017102588e-05,
"loss": 1.2548465728759766,
"step": 1780
},
{
"epoch": 0.7566765578635015,
"grad_norm": 0.2535206079483032,
"learning_rate": 2.2185895414094672e-05,
"loss": 1.2987776756286622,
"step": 1785
},
{
"epoch": 0.7587961000423908,
"grad_norm": 0.24425561726093292,
"learning_rate": 2.1821555103578185e-05,
"loss": 1.2256616592407226,
"step": 1790
},
{
"epoch": 0.7609156422212802,
"grad_norm": 0.27984052896499634,
"learning_rate": 2.1459721258041602e-05,
"loss": 1.2090517997741699,
"step": 1795
},
{
"epoch": 0.7630351844001696,
"grad_norm": 0.22827006876468658,
"learning_rate": 2.1100410931835832e-05,
"loss": 1.198941707611084,
"step": 1800
},
{
"epoch": 0.7651547265790589,
"grad_norm": 0.2688623368740082,
"learning_rate": 2.0743641060370526e-05,
"loss": 1.2268586158752441,
"step": 1805
},
{
"epoch": 0.7672742687579482,
"grad_norm": 0.2658863961696625,
"learning_rate": 2.038942845931588e-05,
"loss": 1.2375852584838867,
"step": 1810
},
{
"epoch": 0.7693938109368377,
"grad_norm": 0.2301843911409378,
"learning_rate": 2.003778982380994e-05,
"loss": 1.2316068649291991,
"step": 1815
},
{
"epoch": 0.771513353115727,
"grad_norm": 0.24156822264194489,
"learning_rate": 1.968874172767179e-05,
"loss": 1.2411081314086914,
"step": 1820
},
{
"epoch": 0.7736328952946163,
"grad_norm": 0.26218628883361816,
"learning_rate": 1.934230062262041e-05,
"loss": 1.2409934043884276,
"step": 1825
},
{
"epoch": 0.7757524374735058,
"grad_norm": 0.28127264976501465,
"learning_rate": 1.899848283749915e-05,
"loss": 1.2560542106628418,
"step": 1830
},
{
"epoch": 0.7778719796523951,
"grad_norm": 0.27439236640930176,
"learning_rate": 1.8657304577506155e-05,
"loss": 1.1889332771301269,
"step": 1835
},
{
"epoch": 0.7799915218312844,
"grad_norm": 0.2547718286514282,
"learning_rate": 1.8318781923430636e-05,
"loss": 1.1872411727905274,
"step": 1840
},
{
"epoch": 0.7821110640101738,
"grad_norm": 0.26395854353904724,
"learning_rate": 1.7982930830894793e-05,
"loss": 1.2034486770629882,
"step": 1845
},
{
"epoch": 0.7842306061890632,
"grad_norm": 0.2656222879886627,
"learning_rate": 1.7649767129601863e-05,
"loss": 1.2152717590332032,
"step": 1850
},
{
"epoch": 0.7863501483679525,
"grad_norm": 0.27528712153434753,
"learning_rate": 1.7319306522590043e-05,
"loss": 1.2389745712280273,
"step": 1855
},
{
"epoch": 0.7884696905468419,
"grad_norm": 0.20311610400676727,
"learning_rate": 1.6991564585492273e-05,
"loss": 1.246045970916748,
"step": 1860
},
{
"epoch": 0.7905892327257312,
"grad_norm": 0.2335929125547409,
"learning_rate": 1.666655676580214e-05,
"loss": 1.240429973602295,
"step": 1865
},
{
"epoch": 0.7927087749046206,
"grad_norm": 0.2188822329044342,
"learning_rate": 1.6344298382145866e-05,
"loss": 1.1916460037231444,
"step": 1870
},
{
"epoch": 0.79482831708351,
"grad_norm": 0.24300524592399597,
"learning_rate": 1.6024804623560164e-05,
"loss": 1.2375968933105468,
"step": 1875
},
{
"epoch": 0.7969478592623993,
"grad_norm": 0.22874438762664795,
"learning_rate": 1.5708090548776423e-05,
"loss": 1.2663753509521485,
"step": 1880
},
{
"epoch": 0.7990674014412886,
"grad_norm": 0.2334616631269455,
"learning_rate": 1.5394171085510896e-05,
"loss": 1.168552780151367,
"step": 1885
},
{
"epoch": 0.8011869436201781,
"grad_norm": 0.3264888823032379,
"learning_rate": 1.5083061029761176e-05,
"loss": 1.214361572265625,
"step": 1890
},
{
"epoch": 0.8033064857990674,
"grad_norm": 0.31234920024871826,
"learning_rate": 1.4774775045108739e-05,
"loss": 1.2208282470703125,
"step": 1895
},
{
"epoch": 0.8054260279779568,
"grad_norm": 0.2120732218027115,
"learning_rate": 1.44693276620278e-05,
"loss": 1.1900595664978026,
"step": 1900
},
{
"epoch": 0.8075455701568461,
"grad_norm": 0.22563967108726501,
"learning_rate": 1.4166733277200563e-05,
"loss": 1.1866695404052734,
"step": 1905
},
{
"epoch": 0.8096651123357355,
"grad_norm": 0.2642882764339447,
"learning_rate": 1.3867006152838514e-05,
"loss": 1.1998025894165039,
"step": 1910
},
{
"epoch": 0.8117846545146249,
"grad_norm": 0.222847118973732,
"learning_rate": 1.357016041601027e-05,
"loss": 1.197617530822754,
"step": 1915
},
{
"epoch": 0.8139041966935142,
"grad_norm": 0.2953515350818634,
"learning_rate": 1.3276210057975772e-05,
"loss": 1.1943370819091796,
"step": 1920
},
{
"epoch": 0.8160237388724035,
"grad_norm": 0.23096834123134613,
"learning_rate": 1.2985168933526719e-05,
"loss": 1.1839122772216797,
"step": 1925
},
{
"epoch": 0.818143281051293,
"grad_norm": 0.25712159276008606,
"learning_rate": 1.2697050760333615e-05,
"loss": 1.2583272933959961,
"step": 1930
},
{
"epoch": 0.8202628232301823,
"grad_norm": 0.269237220287323,
"learning_rate": 1.2411869118299275e-05,
"loss": 1.2126113891601562,
"step": 1935
},
{
"epoch": 0.8223823654090716,
"grad_norm": 0.2036949247121811,
"learning_rate": 1.212963744891863e-05,
"loss": 1.171532154083252,
"step": 1940
},
{
"epoch": 0.824501907587961,
"grad_norm": 0.19737450778484344,
"learning_rate": 1.1850369054645253e-05,
"loss": 1.2296975135803223,
"step": 1945
},
{
"epoch": 0.8266214497668504,
"grad_norm": 0.28607380390167236,
"learning_rate": 1.157407709826444e-05,
"loss": 1.2230495452880858,
"step": 1950
},
{
"epoch": 0.8287409919457397,
"grad_norm": 0.27250775694847107,
"learning_rate": 1.1300774602272678e-05,
"loss": 1.193952751159668,
"step": 1955
},
{
"epoch": 0.8308605341246291,
"grad_norm": 0.25226935744285583,
"learning_rate": 1.1030474448263935e-05,
"loss": 1.2080106735229492,
"step": 1960
},
{
"epoch": 0.8329800763035184,
"grad_norm": 0.24940216541290283,
"learning_rate": 1.0763189376322482e-05,
"loss": 1.2480884552001954,
"step": 1965
},
{
"epoch": 0.8350996184824078,
"grad_norm": 0.2572363018989563,
"learning_rate": 1.0498931984422457e-05,
"loss": 1.1861747741699218,
"step": 1970
},
{
"epoch": 0.8372191606612972,
"grad_norm": 0.21477046608924866,
"learning_rate": 1.0237714727834034e-05,
"loss": 1.2388809204101563,
"step": 1975
},
{
"epoch": 0.8393387028401865,
"grad_norm": 0.231699138879776,
"learning_rate": 9.979549918536359e-06,
"loss": 1.2256881713867187,
"step": 1980
},
{
"epoch": 0.8414582450190758,
"grad_norm": 0.29070281982421875,
"learning_rate": 9.724449724637317e-06,
"loss": 1.1785457611083985,
"step": 1985
},
{
"epoch": 0.8435777871979653,
"grad_norm": 0.28162720799446106,
"learning_rate": 9.472426169799952e-06,
"loss": 1.2014272689819336,
"step": 1990
},
{
"epoch": 0.8456973293768546,
"grad_norm": 0.29717785120010376,
"learning_rate": 9.223491132675723e-06,
"loss": 1.2185420989990234,
"step": 1995
},
{
"epoch": 0.8478168715557439,
"grad_norm": 0.2569153308868408,
"learning_rate": 8.977656346344772e-06,
"loss": 1.1993830680847168,
"step": 2000
},
{
"epoch": 0.8499364137346334,
"grad_norm": 0.20953696966171265,
"learning_rate": 8.73493339776272e-06,
"loss": 1.2431264877319337,
"step": 2005
},
{
"epoch": 0.8520559559135227,
"grad_norm": 0.23639428615570068,
"learning_rate": 8.495333727214675e-06,
"loss": 1.1554552078247071,
"step": 2010
},
{
"epoch": 0.854175498092412,
"grad_norm": 0.22558613121509552,
"learning_rate": 8.258868627775942e-06,
"loss": 1.2167752265930176,
"step": 2015
},
{
"epoch": 0.8562950402713014,
"grad_norm": 0.2367500513792038,
"learning_rate": 8.025549244779781e-06,
"loss": 1.196829605102539,
"step": 2020
},
{
"epoch": 0.8584145824501908,
"grad_norm": 0.25534623861312866,
"learning_rate": 7.795386575292084e-06,
"loss": 1.2867817878723145,
"step": 2025
},
{
"epoch": 0.8605341246290801,
"grad_norm": 0.24774640798568726,
"learning_rate": 7.56839146759306e-06,
"loss": 1.2028039932250976,
"step": 2030
},
{
"epoch": 0.8626536668079695,
"grad_norm": 0.25155386328697205,
"learning_rate": 7.3445746206659e-06,
"loss": 1.2109484672546387,
"step": 2035
},
{
"epoch": 0.8647732089868588,
"grad_norm": 0.28053009510040283,
"learning_rate": 7.123946583692489e-06,
"loss": 1.1832473754882813,
"step": 2040
},
{
"epoch": 0.8668927511657482,
"grad_norm": 0.27603018283843994,
"learning_rate": 6.906517755556274e-06,
"loss": 1.236221694946289,
"step": 2045
},
{
"epoch": 0.8690122933446376,
"grad_norm": 0.23505373299121857,
"learning_rate": 6.692298384351993e-06,
"loss": 1.2576598167419433,
"step": 2050
},
{
"epoch": 0.8711318355235269,
"grad_norm": 0.21782788634300232,
"learning_rate": 6.481298566902776e-06,
"loss": 1.166575813293457,
"step": 2055
},
{
"epoch": 0.8732513777024162,
"grad_norm": 0.21667101979255676,
"learning_rate": 6.273528248284209e-06,
"loss": 1.2306554794311524,
"step": 2060
},
{
"epoch": 0.8753709198813057,
"grad_norm": 0.2405225783586502,
"learning_rate": 6.068997221355567e-06,
"loss": 1.167166805267334,
"step": 2065
},
{
"epoch": 0.877490462060195,
"grad_norm": 0.24216975271701813,
"learning_rate": 5.867715126298237e-06,
"loss": 1.1691235542297362,
"step": 2070
},
{
"epoch": 0.8796100042390843,
"grad_norm": 0.25538262724876404,
"learning_rate": 5.669691450161437e-06,
"loss": 1.2023805618286132,
"step": 2075
},
{
"epoch": 0.8817295464179737,
"grad_norm": 0.32346659898757935,
"learning_rate": 5.474935526414967e-06,
"loss": 1.2079744338989258,
"step": 2080
},
{
"epoch": 0.8838490885968631,
"grad_norm": 0.2660439610481262,
"learning_rate": 5.2834565345092925e-06,
"loss": 1.2419456481933593,
"step": 2085
},
{
"epoch": 0.8859686307757524,
"grad_norm": 0.24010635912418365,
"learning_rate": 5.095263499442986e-06,
"loss": 1.1505148887634278,
"step": 2090
},
{
"epoch": 0.8880881729546418,
"grad_norm": 0.23079979419708252,
"learning_rate": 4.910365291337254e-06,
"loss": 1.20126314163208,
"step": 2095
},
{
"epoch": 0.8902077151335311,
"grad_norm": 0.27824434638023376,
"learning_rate": 4.72877062501788e-06,
"loss": 1.218956184387207,
"step": 2100
},
{
"epoch": 0.8923272573124205,
"grad_norm": 0.25365084409713745,
"learning_rate": 4.5504880596044875e-06,
"loss": 1.2244325637817384,
"step": 2105
},
{
"epoch": 0.8944467994913099,
"grad_norm": 0.26706644892692566,
"learning_rate": 4.375525998107146e-06,
"loss": 1.2031432151794434,
"step": 2110
},
{
"epoch": 0.8965663416701992,
"grad_norm": 0.2815406620502472,
"learning_rate": 4.203892687030244e-06,
"loss": 1.206797218322754,
"step": 2115
},
{
"epoch": 0.8986858838490887,
"grad_norm": 0.2703615427017212,
"learning_rate": 4.035596215983855e-06,
"loss": 1.1795376777648925,
"step": 2120
},
{
"epoch": 0.900805426027978,
"grad_norm": 0.21370485424995422,
"learning_rate": 3.870644517302451e-06,
"loss": 1.188817596435547,
"step": 2125
},
{
"epoch": 0.9029249682068673,
"grad_norm": 0.2289111614227295,
"learning_rate": 3.7090453656709955e-06,
"loss": 1.1960336685180664,
"step": 2130
},
{
"epoch": 0.9050445103857567,
"grad_norm": 0.2908213138580322,
"learning_rate": 3.5508063777584914e-06,
"loss": 1.2193121910095215,
"step": 2135
},
{
"epoch": 0.907164052564646,
"grad_norm": 0.3031107485294342,
"learning_rate": 3.39593501185908e-06,
"loss": 1.2259504318237304,
"step": 2140
},
{
"epoch": 0.9092835947435354,
"grad_norm": 0.24484696984291077,
"learning_rate": 3.2444385675403613e-06,
"loss": 1.1841787338256835,
"step": 2145
},
{
"epoch": 0.9114031369224248,
"grad_norm": 0.19316215813159943,
"learning_rate": 3.0963241852994535e-06,
"loss": 1.283358097076416,
"step": 2150
},
{
"epoch": 0.9135226791013141,
"grad_norm": 0.20729324221611023,
"learning_rate": 2.9515988462264133e-06,
"loss": 1.229339027404785,
"step": 2155
},
{
"epoch": 0.9156422212802034,
"grad_norm": 0.25223687291145325,
"learning_rate": 2.8102693716751555e-06,
"loss": 1.2121631622314453,
"step": 2160
},
{
"epoch": 0.9177617634590929,
"grad_norm": 0.23386164009571075,
"learning_rate": 2.6723424229419783e-06,
"loss": 1.2156064987182618,
"step": 2165
},
{
"epoch": 0.9198813056379822,
"grad_norm": 0.22168220579624176,
"learning_rate": 2.5378245009516297e-06,
"loss": 1.1857851982116698,
"step": 2170
},
{
"epoch": 0.9220008478168715,
"grad_norm": 0.2494061440229416,
"learning_rate": 2.406721945950796e-06,
"loss": 1.1707450866699218,
"step": 2175
},
{
"epoch": 0.924120389995761,
"grad_norm": 0.22970493137836456,
"learning_rate": 2.279040937209373e-06,
"loss": 1.2123796463012695,
"step": 2180
},
{
"epoch": 0.9262399321746503,
"grad_norm": 0.24052271246910095,
"learning_rate": 2.1547874927291485e-06,
"loss": 1.1799467086791993,
"step": 2185
},
{
"epoch": 0.9283594743535396,
"grad_norm": 0.23852086067199707,
"learning_rate": 2.0339674689601987e-06,
"loss": 1.1694490432739257,
"step": 2190
},
{
"epoch": 0.930479016532429,
"grad_norm": 0.24721400439739227,
"learning_rate": 1.91658656052483e-06,
"loss": 1.1695253372192382,
"step": 2195
},
{
"epoch": 0.9325985587113184,
"grad_norm": 0.2726411521434784,
"learning_rate": 1.8026502999491626e-06,
"loss": 1.2767223358154296,
"step": 2200
},
{
"epoch": 0.9347181008902077,
"grad_norm": 0.2515442967414856,
"learning_rate": 1.6921640574024315e-06,
"loss": 1.1741837501525878,
"step": 2205
},
{
"epoch": 0.9368376430690971,
"grad_norm": 0.2630488872528076,
"learning_rate": 1.5851330404437617e-06,
"loss": 1.2584412574768067,
"step": 2210
},
{
"epoch": 0.9389571852479864,
"grad_norm": 0.24313470721244812,
"learning_rate": 1.4815622937768334e-06,
"loss": 1.203806209564209,
"step": 2215
},
{
"epoch": 0.9410767274268758,
"grad_norm": 0.2748136818408966,
"learning_rate": 1.3814566990120374e-06,
"loss": 1.2514909744262694,
"step": 2220
},
{
"epoch": 0.9431962696057652,
"grad_norm": 0.23232340812683105,
"learning_rate": 1.2848209744364263e-06,
"loss": 1.2144655227661132,
"step": 2225
},
{
"epoch": 0.9453158117846545,
"grad_norm": 0.27784907817840576,
"learning_rate": 1.1916596747912927e-06,
"loss": 1.1628128051757813,
"step": 2230
},
{
"epoch": 0.9474353539635438,
"grad_norm": 0.23810848593711853,
"learning_rate": 1.1019771910575236e-06,
"loss": 1.130190372467041,
"step": 2235
},
{
"epoch": 0.9495548961424333,
"grad_norm": 0.22645658254623413,
"learning_rate": 1.0157777502486337e-06,
"loss": 1.2297857284545899,
"step": 2240
},
{
"epoch": 0.9516744383213226,
"grad_norm": 0.21455734968185425,
"learning_rate": 9.330654152115064e-07,
"loss": 1.2333463668823241,
"step": 2245
},
{
"epoch": 0.9537939805002119,
"grad_norm": 0.23416118323802948,
"learning_rate": 8.538440844349403e-07,
"loss": 1.2018747329711914,
"step": 2250
},
{
"epoch": 0.9559135226791013,
"grad_norm": 0.2789277732372284,
"learning_rate": 7.78117491865854e-07,
"loss": 1.2901772499084472,
"step": 2255
},
{
"epoch": 0.9580330648579907,
"grad_norm": 0.24100664258003235,
"learning_rate": 7.058892067333427e-07,
"loss": 1.2156628608703612,
"step": 2260
},
{
"epoch": 0.96015260703688,
"grad_norm": 0.25092175602912903,
"learning_rate": 6.371626333804141e-07,
"loss": 1.218484115600586,
"step": 2265
},
{
"epoch": 0.9622721492157694,
"grad_norm": 0.23545299470424652,
"learning_rate": 5.719410111035577e-07,
"loss": 1.1988481521606444,
"step": 2270
},
{
"epoch": 0.9643916913946587,
"grad_norm": 0.2572426497936249,
"learning_rate": 5.102274140000334e-07,
"loss": 1.2472898483276367,
"step": 2275
},
{
"epoch": 0.9665112335735481,
"grad_norm": 0.2869701087474823,
"learning_rate": 4.520247508230046e-07,
"loss": 1.2034643173217774,
"step": 2280
},
{
"epoch": 0.9686307757524375,
"grad_norm": 0.2547682523727417,
"learning_rate": 3.973357648444392e-07,
"loss": 1.2677287101745605,
"step": 2285
},
{
"epoch": 0.9707503179313268,
"grad_norm": 0.2538366913795471,
"learning_rate": 3.4616303372579645e-07,
"loss": 1.236396026611328,
"step": 2290
},
{
"epoch": 0.9728698601102161,
"grad_norm": 0.2516900599002838,
"learning_rate": 2.9850896939655754e-07,
"loss": 1.2048575401306152,
"step": 2295
},
{
"epoch": 0.9749894022891056,
"grad_norm": 0.2662227153778076,
"learning_rate": 2.543758179405164e-07,
"loss": 1.1859526634216309,
"step": 2300
},
{
"epoch": 0.9771089444679949,
"grad_norm": 0.2448529601097107,
"learning_rate": 2.1376565948991453e-07,
"loss": 1.1678646087646485,
"step": 2305
},
{
"epoch": 0.9792284866468842,
"grad_norm": 0.25061237812042236,
"learning_rate": 1.7668040812744422e-07,
"loss": 1.2213254928588868,
"step": 2310
},
{
"epoch": 0.9813480288257737,
"grad_norm": 0.26765909790992737,
"learning_rate": 1.4312181179597092e-07,
"loss": 1.2179094314575196,
"step": 2315
},
{
"epoch": 0.983467571004663,
"grad_norm": 0.24919472634792328,
"learning_rate": 1.1309145221618244e-07,
"loss": 1.2002128601074218,
"step": 2320
},
{
"epoch": 0.9855871131835523,
"grad_norm": 0.25498664379119873,
"learning_rate": 8.659074481203177e-08,
"loss": 1.2240735054016114,
"step": 2325
},
{
"epoch": 0.9877066553624417,
"grad_norm": 0.2464495748281479,
"learning_rate": 6.362093864402395e-08,
"loss": 1.2199889183044434,
"step": 2330
},
{
"epoch": 0.989826197541331,
"grad_norm": 0.25545111298561096,
"learning_rate": 4.4183116350329786e-08,
"loss": 1.2511631011962892,
"step": 2335
},
{
"epoch": 0.9919457397202205,
"grad_norm": 0.28013259172439575,
"learning_rate": 2.8278194095784913e-08,
"loss": 1.1898666381835938,
"step": 2340
},
{
"epoch": 0.9940652818991098,
"grad_norm": 0.22563761472702026,
"learning_rate": 1.5906921528691108e-08,
"loss": 1.2562253952026368,
"step": 2345
},
{
"epoch": 0.9961848240779991,
"grad_norm": 0.29211103916168213,
"learning_rate": 7.069881745469519e-09,
"loss": 1.1563979148864747,
"step": 2350
},
{
"epoch": 0.9983043662568886,
"grad_norm": 0.2630368173122406,
"learning_rate": 1.7674912631993054e-09,
"loss": 1.1870248794555665,
"step": 2355
}
],
"logging_steps": 5,
"max_steps": 2359,
"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": 2.744030824717353e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}