0113_intents / trainer_state.json
Nevamind-AI's picture
Upload model 2026-01-13 13:12:48
994efbd verified
Raw
History Blame Contribute Delete
21.9 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04069175991861648,
"grad_norm": 89.35326385498047,
"learning_rate": 7.317073170731707e-07,
"loss": 2.6404,
"step": 10
},
{
"epoch": 0.08138351983723296,
"grad_norm": 76.9419937133789,
"learning_rate": 1.5447154471544717e-06,
"loss": 2.4085,
"step": 20
},
{
"epoch": 0.12207527975584945,
"grad_norm": 42.19841003417969,
"learning_rate": 2.3577235772357727e-06,
"loss": 1.312,
"step": 30
},
{
"epoch": 0.16276703967446593,
"grad_norm": 16.992250442504883,
"learning_rate": 3.1707317073170736e-06,
"loss": 0.7585,
"step": 40
},
{
"epoch": 0.2034587995930824,
"grad_norm": 23.777936935424805,
"learning_rate": 3.983739837398374e-06,
"loss": 0.6291,
"step": 50
},
{
"epoch": 0.2441505595116989,
"grad_norm": 33.58175277709961,
"learning_rate": 4.796747967479675e-06,
"loss": 0.7138,
"step": 60
},
{
"epoch": 0.28484231943031535,
"grad_norm": 25.627086639404297,
"learning_rate": 5.609756097560977e-06,
"loss": 0.6864,
"step": 70
},
{
"epoch": 0.32553407934893186,
"grad_norm": 17.266746520996094,
"learning_rate": 6.422764227642278e-06,
"loss": 0.5999,
"step": 80
},
{
"epoch": 0.3662258392675483,
"grad_norm": 19.108999252319336,
"learning_rate": 7.2357723577235786e-06,
"loss": 0.6526,
"step": 90
},
{
"epoch": 0.4069175991861648,
"grad_norm": 17.752649307250977,
"learning_rate": 8.048780487804879e-06,
"loss": 0.5345,
"step": 100
},
{
"epoch": 0.4476093591047813,
"grad_norm": 19.795991897583008,
"learning_rate": 8.86178861788618e-06,
"loss": 0.5766,
"step": 110
},
{
"epoch": 0.4883011190233978,
"grad_norm": 11.328675270080566,
"learning_rate": 9.67479674796748e-06,
"loss": 0.7256,
"step": 120
},
{
"epoch": 0.5289928789420142,
"grad_norm": 19.272499084472656,
"learning_rate": 9.999275169399614e-06,
"loss": 0.6282,
"step": 130
},
{
"epoch": 0.5696846388606307,
"grad_norm": 17.82128143310547,
"learning_rate": 9.994846410084447e-06,
"loss": 0.5881,
"step": 140
},
{
"epoch": 0.6103763987792472,
"grad_norm": 12.624459266662598,
"learning_rate": 9.986395137421607e-06,
"loss": 0.6573,
"step": 150
},
{
"epoch": 0.6510681586978637,
"grad_norm": 15.948984146118164,
"learning_rate": 9.973928157497675e-06,
"loss": 0.6869,
"step": 160
},
{
"epoch": 0.6917599186164801,
"grad_norm": 16.190900802612305,
"learning_rate": 9.957455510379753e-06,
"loss": 0.6831,
"step": 170
},
{
"epoch": 0.7324516785350966,
"grad_norm": 9.356847763061523,
"learning_rate": 9.93699046202989e-06,
"loss": 0.5518,
"step": 180
},
{
"epoch": 0.7731434384537131,
"grad_norm": 19.214969635009766,
"learning_rate": 9.912549493621555e-06,
"loss": 0.6744,
"step": 190
},
{
"epoch": 0.8138351983723296,
"grad_norm": 16.704742431640625,
"learning_rate": 9.88415228826682e-06,
"loss": 0.5619,
"step": 200
},
{
"epoch": 0.854526958290946,
"grad_norm": 19.27890968322754,
"learning_rate": 9.851821715164891e-06,
"loss": 0.5556,
"step": 210
},
{
"epoch": 0.8952187182095626,
"grad_norm": 13.688630104064941,
"learning_rate": 9.815583811184809e-06,
"loss": 0.592,
"step": 220
},
{
"epoch": 0.9359104781281791,
"grad_norm": 12.136804580688477,
"learning_rate": 9.775467759897092e-06,
"loss": 0.5935,
"step": 230
},
{
"epoch": 0.9766022380467956,
"grad_norm": 15.438243865966797,
"learning_rate": 9.731505868071262e-06,
"loss": 0.6359,
"step": 240
},
{
"epoch": 1.0162767039674465,
"grad_norm": 16.588224411010742,
"learning_rate": 9.68373353965814e-06,
"loss": 0.4375,
"step": 250
},
{
"epoch": 1.056968463886063,
"grad_norm": 11.66283130645752,
"learning_rate": 9.632189247277885e-06,
"loss": 0.2664,
"step": 260
},
{
"epoch": 1.0976602238046795,
"grad_norm": 15.249889373779297,
"learning_rate": 9.576914501236734e-06,
"loss": 0.2865,
"step": 270
},
{
"epoch": 1.138351983723296,
"grad_norm": 10.971893310546875,
"learning_rate": 9.517953816097396e-06,
"loss": 0.2501,
"step": 280
},
{
"epoch": 1.1790437436419126,
"grad_norm": 9.27628231048584,
"learning_rate": 9.455354674830016e-06,
"loss": 0.3166,
"step": 290
},
{
"epoch": 1.219735503560529,
"grad_norm": 13.709596633911133,
"learning_rate": 9.38916749057258e-06,
"loss": 0.2858,
"step": 300
},
{
"epoch": 1.2604272634791456,
"grad_norm": 15.793980598449707,
"learning_rate": 9.31944556603157e-06,
"loss": 0.2958,
"step": 310
},
{
"epoch": 1.3011190233977619,
"grad_norm": 18.49693489074707,
"learning_rate": 9.24624505055555e-06,
"loss": 0.2226,
"step": 320
},
{
"epoch": 1.3418107833163784,
"grad_norm": 15.116625785827637,
"learning_rate": 9.169624894916252e-06,
"loss": 0.408,
"step": 330
},
{
"epoch": 1.3825025432349949,
"grad_norm": 12.354613304138184,
"learning_rate": 9.089646803833589e-06,
"loss": 0.275,
"step": 340
},
{
"epoch": 1.4231943031536114,
"grad_norm": 15.7548189163208,
"learning_rate": 9.006375186282832e-06,
"loss": 0.3446,
"step": 350
},
{
"epoch": 1.463886063072228,
"grad_norm": 8.44507122039795,
"learning_rate": 8.919877103623949e-06,
"loss": 0.2696,
"step": 360
},
{
"epoch": 1.5045778229908442,
"grad_norm": 21.99254608154297,
"learning_rate": 8.83022221559489e-06,
"loss": 0.3169,
"step": 370
},
{
"epoch": 1.545269582909461,
"grad_norm": 11.84708023071289,
"learning_rate": 8.737482724212331e-06,
"loss": 0.3177,
"step": 380
},
{
"epoch": 1.5859613428280772,
"grad_norm": 8.813089370727539,
"learning_rate": 8.641733315625014e-06,
"loss": 0.39,
"step": 390
},
{
"epoch": 1.626653102746694,
"grad_norm": 13.93511962890625,
"learning_rate": 8.543051099966558e-06,
"loss": 0.3782,
"step": 400
},
{
"epoch": 1.6673448626653102,
"grad_norm": 11.278246879577637,
"learning_rate": 8.441515549256134e-06,
"loss": 0.3698,
"step": 410
},
{
"epoch": 1.7080366225839267,
"grad_norm": 15.365361213684082,
"learning_rate": 8.337208433397051e-06,
"loss": 0.3354,
"step": 420
},
{
"epoch": 1.7487283825025433,
"grad_norm": 14.246538162231445,
"learning_rate": 8.230213754324773e-06,
"loss": 0.3034,
"step": 430
},
{
"epoch": 1.7894201424211598,
"grad_norm": 23.70111656188965,
"learning_rate": 8.120617678357415e-06,
"loss": 0.3206,
"step": 440
},
{
"epoch": 1.8301119023397763,
"grad_norm": 14.0983247756958,
"learning_rate": 8.00850846680318e-06,
"loss": 0.307,
"step": 450
},
{
"epoch": 1.8708036622583926,
"grad_norm": 13.082509994506836,
"learning_rate": 7.893976404880643e-06,
"loss": 0.3137,
"step": 460
},
{
"epoch": 1.9114954221770093,
"grad_norm": 7.890928268432617,
"learning_rate": 7.7771137290091e-06,
"loss": 0.3006,
"step": 470
},
{
"epoch": 1.9521871820956256,
"grad_norm": 12.845849990844727,
"learning_rate": 7.658014552527572e-06,
"loss": 0.2692,
"step": 480
},
{
"epoch": 1.992878942014242,
"grad_norm": 14.034470558166504,
"learning_rate": 7.536774789902246e-06,
"loss": 0.347,
"step": 490
},
{
"epoch": 2.032553407934893,
"grad_norm": 6.240048885345459,
"learning_rate": 7.413492079483405e-06,
"loss": 0.1513,
"step": 500
},
{
"epoch": 2.0732451678535098,
"grad_norm": 9.91797924041748,
"learning_rate": 7.288265704874089e-06,
"loss": 0.097,
"step": 510
},
{
"epoch": 2.113936927772126,
"grad_norm": 11.786038398742676,
"learning_rate": 7.161196514973735e-06,
"loss": 0.1396,
"step": 520
},
{
"epoch": 2.154628687690743,
"grad_norm": 16.4578914642334,
"learning_rate": 7.032386842761282e-06,
"loss": 0.099,
"step": 530
},
{
"epoch": 2.195320447609359,
"grad_norm": 8.25195598602295,
"learning_rate": 6.9019404228830465e-06,
"loss": 0.1554,
"step": 540
},
{
"epoch": 2.236012207527976,
"grad_norm": 16.354429244995117,
"learning_rate": 6.769962308111839e-06,
"loss": 0.1177,
"step": 550
},
{
"epoch": 2.276703967446592,
"grad_norm": 5.283965110778809,
"learning_rate": 6.636558784744507e-06,
"loss": 0.0596,
"step": 560
},
{
"epoch": 2.3173957273652084,
"grad_norm": 11.216028213500977,
"learning_rate": 6.501837287006112e-06,
"loss": 0.1157,
"step": 570
},
{
"epoch": 2.358087487283825,
"grad_norm": 14.208075523376465,
"learning_rate": 6.365906310529631e-06,
"loss": 0.1156,
"step": 580
},
{
"epoch": 2.3987792472024414,
"grad_norm": 9.671122550964355,
"learning_rate": 6.228875324980862e-06,
"loss": 0.1089,
"step": 590
},
{
"epoch": 2.439471007121058,
"grad_norm": 9.490513801574707,
"learning_rate": 6.090854685898928e-06,
"loss": 0.1043,
"step": 600
},
{
"epoch": 2.4801627670396744,
"grad_norm": 20.91253089904785,
"learning_rate": 5.951955545823342e-06,
"loss": 0.1569,
"step": 610
},
{
"epoch": 2.520854526958291,
"grad_norm": 12.481582641601562,
"learning_rate": 5.812289764779232e-06,
"loss": 0.0881,
"step": 620
},
{
"epoch": 2.5615462868769074,
"grad_norm": 10.88056755065918,
"learning_rate": 5.671969820192794e-06,
"loss": 0.1032,
"step": 630
},
{
"epoch": 2.6022380467955237,
"grad_norm": 11.25527286529541,
"learning_rate": 5.5311087163095475e-06,
"loss": 0.0861,
"step": 640
},
{
"epoch": 2.6429298067141405,
"grad_norm": 15.59957504272461,
"learning_rate": 5.389819893188304e-06,
"loss": 0.1308,
"step": 650
},
{
"epoch": 2.6836215666327567,
"grad_norm": 9.788556098937988,
"learning_rate": 5.248217135344191e-06,
"loss": 0.0892,
"step": 660
},
{
"epoch": 2.7243133265513735,
"grad_norm": 17.310951232910156,
"learning_rate": 5.106414480114238e-06,
"loss": 0.1437,
"step": 670
},
{
"epoch": 2.7650050864699898,
"grad_norm": 13.309972763061523,
"learning_rate": 4.96452612581939e-06,
"loss": 0.079,
"step": 680
},
{
"epoch": 2.8056968463886065,
"grad_norm": 10.911243438720703,
"learning_rate": 4.82266633979685e-06,
"loss": 0.1153,
"step": 690
},
{
"epoch": 2.846388606307223,
"grad_norm": 11.69054126739502,
"learning_rate": 4.680949366376858e-06,
"loss": 0.1396,
"step": 700
},
{
"epoch": 2.887080366225839,
"grad_norm": 16.399673461914062,
"learning_rate": 4.539489334877992e-06,
"loss": 0.0961,
"step": 710
},
{
"epoch": 2.927772126144456,
"grad_norm": 10.901628494262695,
"learning_rate": 4.3984001676950875e-06,
"loss": 0.1238,
"step": 720
},
{
"epoch": 2.968463886063072,
"grad_norm": 10.190411567687988,
"learning_rate": 4.2577954885537985e-06,
"loss": 0.1376,
"step": 730
},
{
"epoch": 3.0081383519837233,
"grad_norm": 5.727451801300049,
"learning_rate": 4.1177885310057045e-06,
"loss": 0.0798,
"step": 740
},
{
"epoch": 3.04883011190234,
"grad_norm": 2.0749783515930176,
"learning_rate": 3.978492047237608e-06,
"loss": 0.0299,
"step": 750
},
{
"epoch": 3.0895218718209563,
"grad_norm": 1.896183967590332,
"learning_rate": 3.840018217268527e-06,
"loss": 0.0293,
"step": 760
},
{
"epoch": 3.1302136317395726,
"grad_norm": 10.008612632751465,
"learning_rate": 3.702478558607429e-06,
"loss": 0.0258,
"step": 770
},
{
"epoch": 3.1709053916581893,
"grad_norm": 8.061917304992676,
"learning_rate": 3.5659838364445505e-06,
"loss": 0.0365,
"step": 780
},
{
"epoch": 3.2115971515768056,
"grad_norm": 6.632195472717285,
"learning_rate": 3.4306439744485453e-06,
"loss": 0.0421,
"step": 790
},
{
"epoch": 3.2522889114954223,
"grad_norm": 10.166711807250977,
"learning_rate": 3.2965679662413772e-06,
"loss": 0.0307,
"step": 800
},
{
"epoch": 3.2929806714140386,
"grad_norm": 1.6602095365524292,
"learning_rate": 3.163863787622162e-06,
"loss": 0.0221,
"step": 810
},
{
"epoch": 3.333672431332655,
"grad_norm": 13.902002334594727,
"learning_rate": 3.0326383096107424e-06,
"loss": 0.0419,
"step": 820
},
{
"epoch": 3.3743641912512716,
"grad_norm": 15.942489624023438,
"learning_rate": 2.9029972123809425e-06,
"loss": 0.0611,
"step": 830
},
{
"epoch": 3.415055951169888,
"grad_norm": 2.8235344886779785,
"learning_rate": 2.775044900152873e-06,
"loss": 0.0604,
"step": 840
},
{
"epoch": 3.4557477110885046,
"grad_norm": 8.490610122680664,
"learning_rate": 2.6488844171127903e-06,
"loss": 0.034,
"step": 850
},
{
"epoch": 3.496439471007121,
"grad_norm": 1.7496405839920044,
"learning_rate": 2.52461736442824e-06,
"loss": 0.0472,
"step": 860
},
{
"epoch": 3.5371312309257377,
"grad_norm": 2.676968812942505,
"learning_rate": 2.4023438184253115e-06,
"loss": 0.0322,
"step": 870
},
{
"epoch": 3.577822990844354,
"grad_norm": 17.634803771972656,
"learning_rate": 2.282162249993895e-06,
"loss": 0.0435,
"step": 880
},
{
"epoch": 3.6185147507629702,
"grad_norm": 3.301680326461792,
"learning_rate": 2.1641694452858486e-06,
"loss": 0.0459,
"step": 890
},
{
"epoch": 3.659206510681587,
"grad_norm": 2.321408987045288,
"learning_rate": 2.0484604277699437e-06,
"loss": 0.025,
"step": 900
},
{
"epoch": 3.6998982706002037,
"grad_norm": 12.765607833862305,
"learning_rate": 1.935128381706355e-06,
"loss": 0.035,
"step": 910
},
{
"epoch": 3.74059003051882,
"grad_norm": 7.213560581207275,
"learning_rate": 1.8242645771023205e-06,
"loss": 0.0336,
"step": 920
},
{
"epoch": 3.7812817904374363,
"grad_norm": 7.408237934112549,
"learning_rate": 1.7159582962094224e-06,
"loss": 0.0285,
"step": 930
},
{
"epoch": 3.821973550356053,
"grad_norm": 5.603072643280029,
"learning_rate": 1.610296761621662e-06,
"loss": 0.0217,
"step": 940
},
{
"epoch": 3.8626653102746693,
"grad_norm": 3.787916898727417,
"learning_rate": 1.5073650660322509e-06,
"loss": 0.0512,
"step": 950
},
{
"epoch": 3.9033570701932856,
"grad_norm": 4.275238037109375,
"learning_rate": 1.4072461037056806e-06,
"loss": 0.0358,
"step": 960
},
{
"epoch": 3.9440488301119023,
"grad_norm": 3.2695820331573486,
"learning_rate": 1.310020503720254e-06,
"loss": 0.0461,
"step": 970
},
{
"epoch": 3.984740590030519,
"grad_norm": 8.396214485168457,
"learning_rate": 1.2157665650348516e-06,
"loss": 0.0245,
"step": 980
},
{
"epoch": 4.02441505595117,
"grad_norm": 4.22248649597168,
"learning_rate": 1.1245601934322148e-06,
"loss": 0.033,
"step": 990
},
{
"epoch": 4.065106815869786,
"grad_norm": 3.4119231700897217,
"learning_rate": 1.0364748403895368e-06,
"loss": 0.0136,
"step": 1000
},
{
"epoch": 4.105798575788403,
"grad_norm": 2.9896421432495117,
"learning_rate": 9.51581443925576e-07,
"loss": 0.0073,
"step": 1010
},
{
"epoch": 4.1464903357070195,
"grad_norm": 4.508012294769287,
"learning_rate": 8.699483714719547e-07,
"loss": 0.0084,
"step": 1020
},
{
"epoch": 4.187182095625635,
"grad_norm": 0.24614429473876953,
"learning_rate": 7.916413648146282e-07,
"loss": 0.0055,
"step": 1030
},
{
"epoch": 4.227873855544252,
"grad_norm": 1.215620517730713,
"learning_rate": 7.167234871498646e-07,
"loss": 0.0067,
"step": 1040
},
{
"epoch": 4.268565615462869,
"grad_norm": 0.1597607284784317,
"learning_rate": 6.452550722973927e-07,
"loss": 0.0138,
"step": 1050
},
{
"epoch": 4.309257375381486,
"grad_norm": 0.5498334169387817,
"learning_rate": 5.772936761116027e-07,
"loss": 0.0032,
"step": 1060
},
{
"epoch": 4.349949135300101,
"grad_norm": 5.39498233795166,
"learning_rate": 5.128940301299334e-07,
"loss": 0.0075,
"step": 1070
},
{
"epoch": 4.390640895218718,
"grad_norm": 1.214426040649414,
"learning_rate": 4.5210799749576815e-07,
"loss": 0.0076,
"step": 1080
},
{
"epoch": 4.431332655137335,
"grad_norm": 3.13959002494812,
"learning_rate": 3.9498453119134917e-07,
"loss": 0.0081,
"step": 1090
},
{
"epoch": 4.472024415055952,
"grad_norm": 2.9278972148895264,
"learning_rate": 3.4156963461434156e-07,
"loss": 0.0068,
"step": 1100
},
{
"epoch": 4.5127161749745675,
"grad_norm": 3.3142828941345215,
"learning_rate": 2.9190632452978706e-07,
"loss": 0.0106,
"step": 1110
},
{
"epoch": 4.553407934893184,
"grad_norm": 1.1654607057571411,
"learning_rate": 2.4603459642729867e-07,
"loss": 0.003,
"step": 1120
},
{
"epoch": 4.594099694811801,
"grad_norm": 1.5315704345703125,
"learning_rate": 2.0399139231137731e-07,
"loss": 0.0074,
"step": 1130
},
{
"epoch": 4.634791454730417,
"grad_norm": 0.38135647773742676,
"learning_rate": 1.6581057095081288e-07,
"loss": 0.0077,
"step": 1140
},
{
"epoch": 4.6754832146490335,
"grad_norm": 2.646099090576172,
"learning_rate": 1.3152288061110518e-07,
"loss": 0.0065,
"step": 1150
},
{
"epoch": 4.71617497456765,
"grad_norm": 0.751757025718689,
"learning_rate": 1.0115593429187942e-07,
"loss": 0.0075,
"step": 1160
},
{
"epoch": 4.756866734486266,
"grad_norm": 1.1848853826522827,
"learning_rate": 7.473418748923545e-08,
"loss": 0.0022,
"step": 1170
},
{
"epoch": 4.797558494404883,
"grad_norm": 1.9046436548233032,
"learning_rate": 5.227891850093314e-08,
"loss": 0.004,
"step": 1180
},
{
"epoch": 4.8382502543234995,
"grad_norm": 3.036072015762329,
"learning_rate": 3.3808211290284886e-08,
"loss": 0.0063,
"step": 1190
},
{
"epoch": 4.878942014242116,
"grad_norm": 9.75121784210205,
"learning_rate": 1.933694092254801e-08,
"loss": 0.0164,
"step": 1200
},
{
"epoch": 4.919633774160732,
"grad_norm": 1.7381020784378052,
"learning_rate": 8.87676158554507e-09,
"loss": 0.0035,
"step": 1210
},
{
"epoch": 4.960325534079349,
"grad_norm": 0.9791738390922546,
"learning_rate": 2.436097204167043e-09,
"loss": 0.0025,
"step": 1220
},
{
"epoch": 5.0,
"grad_norm": 1.383083701133728,
"learning_rate": 2.0134656311676658e-11,
"loss": 0.008,
"step": 1230
}
],
"logging_steps": 10,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"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": 4673958743900160.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}