inter-play-sim-assistant-sft / trainer_state.json
jeromeramos's picture
Model save
b4e9ddd verified
raw
history blame
51.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9982692973347178,
"eval_steps": 500,
"global_step": 1444,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0013845621322256837,
"grad_norm": 78.96892547607422,
"learning_rate": 0.0,
"loss": 3.6138,
"step": 1
},
{
"epoch": 0.006922810661128418,
"grad_norm": 26.50972557067871,
"learning_rate": 5.517241379310345e-06,
"loss": 3.3919,
"step": 5
},
{
"epoch": 0.013845621322256836,
"grad_norm": 10.996688842773438,
"learning_rate": 1.2413793103448277e-05,
"loss": 2.8025,
"step": 10
},
{
"epoch": 0.020768431983385256,
"grad_norm": 7.297196388244629,
"learning_rate": 1.9310344827586207e-05,
"loss": 2.5925,
"step": 15
},
{
"epoch": 0.027691242644513673,
"grad_norm": 6.842594146728516,
"learning_rate": 2.620689655172414e-05,
"loss": 2.3888,
"step": 20
},
{
"epoch": 0.034614053305642094,
"grad_norm": 6.920436382293701,
"learning_rate": 3.310344827586207e-05,
"loss": 2.1351,
"step": 25
},
{
"epoch": 0.04153686396677051,
"grad_norm": 6.916154384613037,
"learning_rate": 4e-05,
"loss": 1.7918,
"step": 30
},
{
"epoch": 0.04845967462789893,
"grad_norm": 3.957890033721924,
"learning_rate": 4.689655172413793e-05,
"loss": 1.4435,
"step": 35
},
{
"epoch": 0.055382485289027346,
"grad_norm": 2.863860607147217,
"learning_rate": 5.379310344827586e-05,
"loss": 1.3002,
"step": 40
},
{
"epoch": 0.06230529595015576,
"grad_norm": 15.180214881896973,
"learning_rate": 6.068965517241379e-05,
"loss": 1.3723,
"step": 45
},
{
"epoch": 0.06922810661128419,
"grad_norm": 3.9463565349578857,
"learning_rate": 6.758620689655173e-05,
"loss": 1.3394,
"step": 50
},
{
"epoch": 0.0761509172724126,
"grad_norm": 35.227447509765625,
"learning_rate": 7.448275862068966e-05,
"loss": 1.3752,
"step": 55
},
{
"epoch": 0.08307372793354102,
"grad_norm": 15.025148391723633,
"learning_rate": 8.137931034482759e-05,
"loss": 1.4322,
"step": 60
},
{
"epoch": 0.08999653859466944,
"grad_norm": 45.38770294189453,
"learning_rate": 8.827586206896552e-05,
"loss": 3.0681,
"step": 65
},
{
"epoch": 0.09691934925579786,
"grad_norm": 8.21139907836914,
"learning_rate": 9.517241379310345e-05,
"loss": 1.3441,
"step": 70
},
{
"epoch": 0.10384215991692627,
"grad_norm": 3.560570240020752,
"learning_rate": 0.0001020689655172414,
"loss": 1.2695,
"step": 75
},
{
"epoch": 0.11076497057805469,
"grad_norm": 11.044499397277832,
"learning_rate": 0.00010896551724137931,
"loss": 1.2145,
"step": 80
},
{
"epoch": 0.11768778123918311,
"grad_norm": 3.8201582431793213,
"learning_rate": 0.00011586206896551725,
"loss": 1.1547,
"step": 85
},
{
"epoch": 0.12461059190031153,
"grad_norm": 3.9128692150115967,
"learning_rate": 0.00012275862068965518,
"loss": 1.1285,
"step": 90
},
{
"epoch": 0.13153340256143994,
"grad_norm": 6.907658576965332,
"learning_rate": 0.0001296551724137931,
"loss": 1.1376,
"step": 95
},
{
"epoch": 0.13845621322256838,
"grad_norm": 2.5651073455810547,
"learning_rate": 0.00013655172413793104,
"loss": 1.1429,
"step": 100
},
{
"epoch": 0.14537902388369678,
"grad_norm": 2.3567330837249756,
"learning_rate": 0.00014344827586206896,
"loss": 1.1356,
"step": 105
},
{
"epoch": 0.1523018345448252,
"grad_norm": 47.36102294921875,
"learning_rate": 0.0001503448275862069,
"loss": 1.2263,
"step": 110
},
{
"epoch": 0.1592246452059536,
"grad_norm": 6.215350151062012,
"learning_rate": 0.00015724137931034485,
"loss": 1.4307,
"step": 115
},
{
"epoch": 0.16614745586708204,
"grad_norm": 2.4405245780944824,
"learning_rate": 0.00016413793103448276,
"loss": 1.2431,
"step": 120
},
{
"epoch": 0.17307026652821045,
"grad_norm": 6.445745944976807,
"learning_rate": 0.0001710344827586207,
"loss": 1.4538,
"step": 125
},
{
"epoch": 0.17999307718933888,
"grad_norm": 7.712639808654785,
"learning_rate": 0.00017793103448275862,
"loss": 1.3982,
"step": 130
},
{
"epoch": 0.18691588785046728,
"grad_norm": 8.239908218383789,
"learning_rate": 0.00018482758620689654,
"loss": 1.4976,
"step": 135
},
{
"epoch": 0.19383869851159571,
"grad_norm": 2.3590519428253174,
"learning_rate": 0.0001917241379310345,
"loss": 1.2595,
"step": 140
},
{
"epoch": 0.20076150917272412,
"grad_norm": 2.654240608215332,
"learning_rate": 0.00019862068965517243,
"loss": 1.2232,
"step": 145
},
{
"epoch": 0.20768431983385255,
"grad_norm": 2.5592308044433594,
"learning_rate": 0.0001999953208384105,
"loss": 1.2181,
"step": 150
},
{
"epoch": 0.21460713049498095,
"grad_norm": 1.9601577520370483,
"learning_rate": 0.0001999763124949378,
"loss": 1.1714,
"step": 155
},
{
"epoch": 0.22152994115610938,
"grad_norm": 1.5926353931427002,
"learning_rate": 0.0001999426852993044,
"loss": 1.173,
"step": 160
},
{
"epoch": 0.2284527518172378,
"grad_norm": 1.8037822246551514,
"learning_rate": 0.00019989444416858502,
"loss": 1.1432,
"step": 165
},
{
"epoch": 0.23537556247836622,
"grad_norm": 2.564453601837158,
"learning_rate": 0.0001998315961567502,
"loss": 1.184,
"step": 170
},
{
"epoch": 0.24229837313949462,
"grad_norm": 1.347174882888794,
"learning_rate": 0.00019975415045363467,
"loss": 1.1879,
"step": 175
},
{
"epoch": 0.24922118380062305,
"grad_norm": 1.8346368074417114,
"learning_rate": 0.00019966211838359378,
"loss": 1.1895,
"step": 180
},
{
"epoch": 0.25614399446175146,
"grad_norm": 1.5952632427215576,
"learning_rate": 0.00019955551340384743,
"loss": 1.1717,
"step": 185
},
{
"epoch": 0.2630668051228799,
"grad_norm": 10.082040786743164,
"learning_rate": 0.00019943435110251248,
"loss": 1.2318,
"step": 190
},
{
"epoch": 0.2699896157840083,
"grad_norm": 9.962459564208984,
"learning_rate": 0.0001992986491963232,
"loss": 1.2411,
"step": 195
},
{
"epoch": 0.27691242644513675,
"grad_norm": 1.3912514448165894,
"learning_rate": 0.00019914842752804103,
"loss": 1.2747,
"step": 200
},
{
"epoch": 0.2838352371062651,
"grad_norm": 2.4310293197631836,
"learning_rate": 0.00019898370806355263,
"loss": 1.1451,
"step": 205
},
{
"epoch": 0.29075804776739356,
"grad_norm": 2.632899761199951,
"learning_rate": 0.00019880451488865836,
"loss": 1.1757,
"step": 210
},
{
"epoch": 0.297680858428522,
"grad_norm": 2.9560701847076416,
"learning_rate": 0.00019861087420555018,
"loss": 1.2112,
"step": 215
},
{
"epoch": 0.3046036690896504,
"grad_norm": 1.2769361734390259,
"learning_rate": 0.0001984028143289803,
"loss": 1.1494,
"step": 220
},
{
"epoch": 0.3115264797507788,
"grad_norm": 0.9540227055549622,
"learning_rate": 0.00019818036568212106,
"loss": 1.1274,
"step": 225
},
{
"epoch": 0.3184492904119072,
"grad_norm": 1.050113320350647,
"learning_rate": 0.00019794356079211604,
"loss": 1.1254,
"step": 230
},
{
"epoch": 0.32537210107303566,
"grad_norm": 1.335493564605713,
"learning_rate": 0.00019769243428532422,
"loss": 1.1688,
"step": 235
},
{
"epoch": 0.3322949117341641,
"grad_norm": 1.0514490604400635,
"learning_rate": 0.00019742702288225652,
"loss": 1.153,
"step": 240
},
{
"epoch": 0.33921772239529246,
"grad_norm": 1.2041540145874023,
"learning_rate": 0.00019714736539220648,
"loss": 1.1369,
"step": 245
},
{
"epoch": 0.3461405330564209,
"grad_norm": 1.1564425230026245,
"learning_rate": 0.00019685350270757555,
"loss": 1.1567,
"step": 250
},
{
"epoch": 0.3530633437175493,
"grad_norm": 1.2910791635513306,
"learning_rate": 0.0001965454777978936,
"loss": 1.1633,
"step": 255
},
{
"epoch": 0.35998615437867776,
"grad_norm": 1.6193257570266724,
"learning_rate": 0.00019622333570353567,
"loss": 1.1348,
"step": 260
},
{
"epoch": 0.36690896503980613,
"grad_norm": 0.8881533741950989,
"learning_rate": 0.00019588712352913625,
"loss": 1.1118,
"step": 265
},
{
"epoch": 0.37383177570093457,
"grad_norm": 1.275267243385315,
"learning_rate": 0.00019553689043670127,
"loss": 1.1278,
"step": 270
},
{
"epoch": 0.380754586362063,
"grad_norm": 1.1025657653808594,
"learning_rate": 0.00019517268763841962,
"loss": 1.0943,
"step": 275
},
{
"epoch": 0.38767739702319143,
"grad_norm": 0.9336686730384827,
"learning_rate": 0.00019479456838917473,
"loss": 1.1002,
"step": 280
},
{
"epoch": 0.39460020768431986,
"grad_norm": 0.8029949069023132,
"learning_rate": 0.0001944025879787574,
"loss": 1.089,
"step": 285
},
{
"epoch": 0.40152301834544823,
"grad_norm": 1.0376158952713013,
"learning_rate": 0.0001939968037237812,
"loss": 1.1215,
"step": 290
},
{
"epoch": 0.40844582900657667,
"grad_norm": 0.8140847682952881,
"learning_rate": 0.00019357727495930146,
"loss": 1.0881,
"step": 295
},
{
"epoch": 0.4153686396677051,
"grad_norm": 0.8856680393218994,
"learning_rate": 0.00019314406303013904,
"loss": 1.0907,
"step": 300
},
{
"epoch": 0.42229145032883353,
"grad_norm": 0.880487859249115,
"learning_rate": 0.00019269723128191048,
"loss": 1.0793,
"step": 305
},
{
"epoch": 0.4292142609899619,
"grad_norm": 0.8125194907188416,
"learning_rate": 0.00019223684505176517,
"loss": 1.0859,
"step": 310
},
{
"epoch": 0.43613707165109034,
"grad_norm": 1.1872682571411133,
"learning_rate": 0.00019176297165883165,
"loss": 1.0933,
"step": 315
},
{
"epoch": 0.44305988231221877,
"grad_norm": 0.8477077484130859,
"learning_rate": 0.00019127568039437406,
"loss": 1.0726,
"step": 320
},
{
"epoch": 0.4499826929733472,
"grad_norm": 0.8214722275733948,
"learning_rate": 0.00019077504251166006,
"loss": 1.0685,
"step": 325
},
{
"epoch": 0.4569055036344756,
"grad_norm": 0.8920738101005554,
"learning_rate": 0.00019026113121554204,
"loss": 1.0697,
"step": 330
},
{
"epoch": 0.463828314295604,
"grad_norm": 0.748255729675293,
"learning_rate": 0.00018973402165175268,
"loss": 1.0703,
"step": 335
},
{
"epoch": 0.47075112495673244,
"grad_norm": 0.7890346050262451,
"learning_rate": 0.0001891937908959172,
"loss": 1.0471,
"step": 340
},
{
"epoch": 0.47767393561786087,
"grad_norm": 0.7698908448219299,
"learning_rate": 0.00018864051794228282,
"loss": 1.0529,
"step": 345
},
{
"epoch": 0.48459674627898924,
"grad_norm": 0.7833818197250366,
"learning_rate": 0.00018807428369216822,
"loss": 1.0187,
"step": 350
},
{
"epoch": 0.4915195569401177,
"grad_norm": 0.7760437726974487,
"learning_rate": 0.00018749517094213375,
"loss": 1.0155,
"step": 355
},
{
"epoch": 0.4984423676012461,
"grad_norm": 0.7187851071357727,
"learning_rate": 0.00018690326437187472,
"loss": 1.0361,
"step": 360
},
{
"epoch": 0.5053651782623745,
"grad_norm": 0.7711403965950012,
"learning_rate": 0.00018629865053183911,
"loss": 1.0248,
"step": 365
},
{
"epoch": 0.5122879889235029,
"grad_norm": 0.8515886068344116,
"learning_rate": 0.0001856814178305722,
"loss": 1.0077,
"step": 370
},
{
"epoch": 0.5192107995846313,
"grad_norm": 0.8935133218765259,
"learning_rate": 0.00018505165652178893,
"loss": 0.9975,
"step": 375
},
{
"epoch": 0.5261336102457598,
"grad_norm": 0.7678277492523193,
"learning_rate": 0.00018440945869117675,
"loss": 1.0241,
"step": 380
},
{
"epoch": 0.5330564209068882,
"grad_norm": 0.8560886383056641,
"learning_rate": 0.00018375491824293067,
"loss": 1.0204,
"step": 385
},
{
"epoch": 0.5399792315680166,
"grad_norm": 0.973434329032898,
"learning_rate": 0.00018308813088602227,
"loss": 1.0232,
"step": 390
},
{
"epoch": 0.5469020422291451,
"grad_norm": 0.7952532172203064,
"learning_rate": 0.00018240919412020466,
"loss": 0.9945,
"step": 395
},
{
"epoch": 0.5538248528902735,
"grad_norm": 0.7281926870346069,
"learning_rate": 0.00018171820722175604,
"loss": 1.0214,
"step": 400
},
{
"epoch": 0.5607476635514018,
"grad_norm": 0.6621217131614685,
"learning_rate": 0.00018101527122896293,
"loss": 1.0157,
"step": 405
},
{
"epoch": 0.5676704742125303,
"grad_norm": 0.6945295929908752,
"learning_rate": 0.0001803004889273463,
"loss": 0.9936,
"step": 410
},
{
"epoch": 0.5745932848736587,
"grad_norm": 0.6977776885032654,
"learning_rate": 0.0001795739648346317,
"loss": 0.9647,
"step": 415
},
{
"epoch": 0.5815160955347871,
"grad_norm": 0.7883012890815735,
"learning_rate": 0.00017883580518546647,
"loss": 0.9637,
"step": 420
},
{
"epoch": 0.5884389061959155,
"grad_norm": 0.6024463176727295,
"learning_rate": 0.00017808611791588584,
"loss": 0.9791,
"step": 425
},
{
"epoch": 0.595361716857044,
"grad_norm": 0.8087362051010132,
"learning_rate": 0.00017732501264752996,
"loss": 0.9714,
"step": 430
},
{
"epoch": 0.6022845275181724,
"grad_norm": 0.6511737108230591,
"learning_rate": 0.00017655260067161503,
"loss": 0.9914,
"step": 435
},
{
"epoch": 0.6092073381793008,
"grad_norm": 0.6464141607284546,
"learning_rate": 0.00017576899493265954,
"loss": 0.9766,
"step": 440
},
{
"epoch": 0.6161301488404292,
"grad_norm": 0.6870825290679932,
"learning_rate": 0.00017497431001196943,
"loss": 0.9775,
"step": 445
},
{
"epoch": 0.6230529595015576,
"grad_norm": 0.7621681094169617,
"learning_rate": 0.00017416866211088366,
"loss": 0.9823,
"step": 450
},
{
"epoch": 0.629975770162686,
"grad_norm": 0.7556800842285156,
"learning_rate": 0.00017335216903378267,
"loss": 0.9605,
"step": 455
},
{
"epoch": 0.6368985808238145,
"grad_norm": 0.715800404548645,
"learning_rate": 0.00017252495017086296,
"loss": 0.9465,
"step": 460
},
{
"epoch": 0.6438213914849429,
"grad_norm": 0.6417484283447266,
"learning_rate": 0.0001716871264806791,
"loss": 0.9631,
"step": 465
},
{
"epoch": 0.6507442021460713,
"grad_norm": 0.7063789367675781,
"learning_rate": 0.0001708388204724572,
"loss": 0.9628,
"step": 470
},
{
"epoch": 0.6576670128071997,
"grad_norm": 0.741763710975647,
"learning_rate": 0.00016998015618818087,
"loss": 0.9786,
"step": 475
},
{
"epoch": 0.6645898234683282,
"grad_norm": 0.5884097814559937,
"learning_rate": 0.00016911125918445363,
"loss": 0.9563,
"step": 480
},
{
"epoch": 0.6715126341294566,
"grad_norm": 0.5698336362838745,
"learning_rate": 0.00016823225651413953,
"loss": 0.9456,
"step": 485
},
{
"epoch": 0.6784354447905849,
"grad_norm": 0.5873739719390869,
"learning_rate": 0.00016734327670778508,
"loss": 0.9356,
"step": 490
},
{
"epoch": 0.6853582554517134,
"grad_norm": 0.6806433200836182,
"learning_rate": 0.00016644444975482523,
"loss": 0.9311,
"step": 495
},
{
"epoch": 0.6922810661128418,
"grad_norm": 0.595560610294342,
"learning_rate": 0.0001655359070845757,
"loss": 0.9354,
"step": 500
},
{
"epoch": 0.6992038767739702,
"grad_norm": 0.593410849571228,
"learning_rate": 0.0001646177815470151,
"loss": 0.9353,
"step": 505
},
{
"epoch": 0.7061266874350987,
"grad_norm": 0.5478509068489075,
"learning_rate": 0.00016369020739335924,
"loss": 0.9195,
"step": 510
},
{
"epoch": 0.7130494980962271,
"grad_norm": 0.5979152321815491,
"learning_rate": 0.00016275332025643028,
"loss": 0.9331,
"step": 515
},
{
"epoch": 0.7199723087573555,
"grad_norm": 0.6606078147888184,
"learning_rate": 0.00016180725713082434,
"loss": 0.9369,
"step": 520
},
{
"epoch": 0.726895119418484,
"grad_norm": 0.5818477869033813,
"learning_rate": 0.0001608521563528795,
"loss": 0.9204,
"step": 525
},
{
"epoch": 0.7338179300796123,
"grad_norm": 0.6692835688591003,
"learning_rate": 0.00015988815758044792,
"loss": 0.9115,
"step": 530
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.731097936630249,
"learning_rate": 0.00015891540177247467,
"loss": 0.9457,
"step": 535
},
{
"epoch": 0.7476635514018691,
"grad_norm": 0.7018098831176758,
"learning_rate": 0.00015793403116838613,
"loss": 0.9256,
"step": 540
},
{
"epoch": 0.7545863620629976,
"grad_norm": 0.6631230711936951,
"learning_rate": 0.00015694418926729146,
"loss": 0.912,
"step": 545
},
{
"epoch": 0.761509172724126,
"grad_norm": 0.6354436874389648,
"learning_rate": 0.00015594602080699956,
"loss": 0.9141,
"step": 550
},
{
"epoch": 0.7684319833852544,
"grad_norm": 0.7462336421012878,
"learning_rate": 0.00015493967174285512,
"loss": 0.9094,
"step": 555
},
{
"epoch": 0.7753547940463829,
"grad_norm": 0.5586750507354736,
"learning_rate": 0.00015392528922639662,
"loss": 0.9188,
"step": 560
},
{
"epoch": 0.7822776047075113,
"grad_norm": 0.5627267956733704,
"learning_rate": 0.00015290302158383923,
"loss": 0.9077,
"step": 565
},
{
"epoch": 0.7892004153686397,
"grad_norm": 0.603247880935669,
"learning_rate": 0.0001518730182943862,
"loss": 0.9043,
"step": 570
},
{
"epoch": 0.796123226029768,
"grad_norm": 0.5933719277381897,
"learning_rate": 0.0001508354299683715,
"loss": 0.9011,
"step": 575
},
{
"epoch": 0.8030460366908965,
"grad_norm": 0.5786928534507751,
"learning_rate": 0.0001497904083252371,
"loss": 0.9074,
"step": 580
},
{
"epoch": 0.8099688473520249,
"grad_norm": 0.591249942779541,
"learning_rate": 0.00014873810617134806,
"loss": 0.9001,
"step": 585
},
{
"epoch": 0.8168916580131533,
"grad_norm": 0.5570294857025146,
"learning_rate": 0.00014767867737764876,
"loss": 0.8942,
"step": 590
},
{
"epoch": 0.8238144686742818,
"grad_norm": 0.6055375337600708,
"learning_rate": 0.0001466122768571633,
"loss": 0.8992,
"step": 595
},
{
"epoch": 0.8307372793354102,
"grad_norm": 0.5531699657440186,
"learning_rate": 0.0001455390605423437,
"loss": 0.9092,
"step": 600
},
{
"epoch": 0.8376600899965386,
"grad_norm": 5.621354103088379,
"learning_rate": 0.00014445918536226904,
"loss": 0.8969,
"step": 605
},
{
"epoch": 0.8445829006576671,
"grad_norm": 0.6026459336280823,
"learning_rate": 0.00014337280921969858,
"loss": 0.8898,
"step": 610
},
{
"epoch": 0.8515057113187954,
"grad_norm": 0.5794016718864441,
"learning_rate": 0.0001422800909679829,
"loss": 0.8964,
"step": 615
},
{
"epoch": 0.8584285219799238,
"grad_norm": 0.5771220922470093,
"learning_rate": 0.00014118119038783588,
"loss": 0.8905,
"step": 620
},
{
"epoch": 0.8653513326410522,
"grad_norm": 0.6543847918510437,
"learning_rate": 0.00014007626816397087,
"loss": 0.9124,
"step": 625
},
{
"epoch": 0.8722741433021807,
"grad_norm": 0.5902742147445679,
"learning_rate": 0.00013896548586160503,
"loss": 0.8901,
"step": 630
},
{
"epoch": 0.8791969539633091,
"grad_norm": 0.6009337902069092,
"learning_rate": 0.00013784900590283473,
"loss": 0.8904,
"step": 635
},
{
"epoch": 0.8861197646244375,
"grad_norm": 0.5576380491256714,
"learning_rate": 0.00013672699154288569,
"loss": 0.8768,
"step": 640
},
{
"epoch": 0.893042575285566,
"grad_norm": 0.6333903074264526,
"learning_rate": 0.00013559960684624112,
"loss": 0.8993,
"step": 645
},
{
"epoch": 0.8999653859466944,
"grad_norm": 0.6231611371040344,
"learning_rate": 0.0001344670166626519,
"loss": 0.8701,
"step": 650
},
{
"epoch": 0.9068881966078228,
"grad_norm": 0.5712286829948425,
"learning_rate": 0.0001333293866030317,
"loss": 0.8729,
"step": 655
},
{
"epoch": 0.9138110072689511,
"grad_norm": 0.5796388983726501,
"learning_rate": 0.0001321868830152407,
"loss": 0.8521,
"step": 660
},
{
"epoch": 0.9207338179300796,
"grad_norm": 0.5513543486595154,
"learning_rate": 0.00013103967295976179,
"loss": 0.8777,
"step": 665
},
{
"epoch": 0.927656628591208,
"grad_norm": 0.5434843897819519,
"learning_rate": 0.0001298879241852724,
"loss": 0.8689,
"step": 670
},
{
"epoch": 0.9345794392523364,
"grad_norm": 0.5376566648483276,
"learning_rate": 0.0001287318051041159,
"loss": 0.8776,
"step": 675
},
{
"epoch": 0.9415022499134649,
"grad_norm": 0.5433745384216309,
"learning_rate": 0.00012757148476767553,
"loss": 0.8643,
"step": 680
},
{
"epoch": 0.9484250605745933,
"grad_norm": 0.6016582250595093,
"learning_rate": 0.0001264071328416555,
"loss": 0.8659,
"step": 685
},
{
"epoch": 0.9553478712357217,
"grad_norm": 0.5252034664154053,
"learning_rate": 0.00012523891958127168,
"loss": 0.8675,
"step": 690
},
{
"epoch": 0.9622706818968502,
"grad_norm": 0.5640143752098083,
"learning_rate": 0.0001240670158063565,
"loss": 0.8582,
"step": 695
},
{
"epoch": 0.9691934925579785,
"grad_norm": 0.5769084095954895,
"learning_rate": 0.00012289159287638105,
"loss": 0.8686,
"step": 700
},
{
"epoch": 0.9761163032191069,
"grad_norm": 0.5791013240814209,
"learning_rate": 0.00012171282266539845,
"loss": 0.8651,
"step": 705
},
{
"epoch": 0.9830391138802353,
"grad_norm": 0.5205267071723938,
"learning_rate": 0.00012053087753691172,
"loss": 0.8487,
"step": 710
},
{
"epoch": 0.9899619245413638,
"grad_norm": 0.6154927611351013,
"learning_rate": 0.00011934593031867038,
"loss": 0.8658,
"step": 715
},
{
"epoch": 0.9968847352024922,
"grad_norm": 0.5034484267234802,
"learning_rate": 0.00011815815427739911,
"loss": 0.8655,
"step": 720
},
{
"epoch": 1.0,
"eval_loss": 0.8625231385231018,
"eval_runtime": 54.6603,
"eval_samples_per_second": 37.852,
"eval_steps_per_second": 2.378,
"step": 723
},
{
"epoch": 1.0027691242644514,
"grad_norm": 0.5045917630195618,
"learning_rate": 0.00011696772309346182,
"loss": 0.7816,
"step": 725
},
{
"epoch": 1.0096919349255797,
"grad_norm": 0.6295811533927917,
"learning_rate": 0.0001157748108354659,
"loss": 0.6759,
"step": 730
},
{
"epoch": 1.0166147455867083,
"grad_norm": 0.6809110045433044,
"learning_rate": 0.00011457959193480925,
"loss": 0.6842,
"step": 735
},
{
"epoch": 1.0235375562478366,
"grad_norm": 0.5183382630348206,
"learning_rate": 0.00011338224116017423,
"loss": 0.6853,
"step": 740
},
{
"epoch": 1.0304603669089651,
"grad_norm": 0.4994449317455292,
"learning_rate": 0.00011218293359197267,
"loss": 0.6822,
"step": 745
},
{
"epoch": 1.0373831775700935,
"grad_norm": 0.5263209342956543,
"learning_rate": 0.00011098184459674484,
"loss": 0.6475,
"step": 750
},
{
"epoch": 1.0443059882312218,
"grad_norm": 0.5542846322059631,
"learning_rate": 0.00010977914980151691,
"loss": 0.6848,
"step": 755
},
{
"epoch": 1.0512287988923503,
"grad_norm": 0.539220929145813,
"learning_rate": 0.00010857502506812029,
"loss": 0.6583,
"step": 760
},
{
"epoch": 1.0581516095534786,
"grad_norm": 0.5188484191894531,
"learning_rate": 0.00010736964646747629,
"loss": 0.6609,
"step": 765
},
{
"epoch": 1.0650744202146072,
"grad_norm": 0.5194864869117737,
"learning_rate": 0.00010616319025385089,
"loss": 0.6779,
"step": 770
},
{
"epoch": 1.0719972308757355,
"grad_norm": 0.532119870185852,
"learning_rate": 0.00010495583283908177,
"loss": 0.6696,
"step": 775
},
{
"epoch": 1.078920041536864,
"grad_norm": 0.5120254755020142,
"learning_rate": 0.00010374775076678332,
"loss": 0.6471,
"step": 780
},
{
"epoch": 1.0858428521979924,
"grad_norm": 0.5310303568840027,
"learning_rate": 0.00010253912068653146,
"loss": 0.6754,
"step": 785
},
{
"epoch": 1.092765662859121,
"grad_norm": 0.4805733859539032,
"learning_rate": 0.00010133011932803378,
"loss": 0.6555,
"step": 790
},
{
"epoch": 1.0996884735202492,
"grad_norm": 0.5308163166046143,
"learning_rate": 0.00010012092347528705,
"loss": 0.6753,
"step": 795
},
{
"epoch": 1.1066112841813776,
"grad_norm": 0.5158370137214661,
"learning_rate": 9.891170994072793e-05,
"loss": 0.6559,
"step": 800
},
{
"epoch": 1.113534094842506,
"grad_norm": 0.5924445390701294,
"learning_rate": 9.770265553937831e-05,
"loss": 0.6539,
"step": 805
},
{
"epoch": 1.1204569055036344,
"grad_norm": 0.5329247713088989,
"learning_rate": 9.649393706299127e-05,
"loss": 0.6444,
"step": 810
},
{
"epoch": 1.127379716164763,
"grad_norm": 0.544358491897583,
"learning_rate": 9.52857312541998e-05,
"loss": 0.6558,
"step": 815
},
{
"epoch": 1.1343025268258913,
"grad_norm": 0.4904942214488983,
"learning_rate": 9.407821478067304e-05,
"loss": 0.6471,
"step": 820
},
{
"epoch": 1.1412253374870198,
"grad_norm": 0.5689169764518738,
"learning_rate": 9.287156420928342e-05,
"loss": 0.6513,
"step": 825
},
{
"epoch": 1.1481481481481481,
"grad_norm": 0.5149237513542175,
"learning_rate": 9.166595598028832e-05,
"loss": 0.6663,
"step": 830
},
{
"epoch": 1.1550709588092767,
"grad_norm": 0.5315957069396973,
"learning_rate": 9.046156638153056e-05,
"loss": 0.6504,
"step": 835
},
{
"epoch": 1.161993769470405,
"grad_norm": 0.4693305492401123,
"learning_rate": 8.925857152266103e-05,
"loss": 0.6503,
"step": 840
},
{
"epoch": 1.1689165801315333,
"grad_norm": 0.5185849666595459,
"learning_rate": 8.805714730938728e-05,
"loss": 0.6598,
"step": 845
},
{
"epoch": 1.1758393907926619,
"grad_norm": 0.5028591156005859,
"learning_rate": 8.685746941775219e-05,
"loss": 0.6461,
"step": 850
},
{
"epoch": 1.1827622014537902,
"grad_norm": 0.5199161767959595,
"learning_rate": 8.565971326844584e-05,
"loss": 0.6468,
"step": 855
},
{
"epoch": 1.1896850121149187,
"grad_norm": 0.5406040549278259,
"learning_rate": 8.44640540011553e-05,
"loss": 0.6497,
"step": 860
},
{
"epoch": 1.196607822776047,
"grad_norm": 0.4970882833003998,
"learning_rate": 8.327066644895469e-05,
"loss": 0.6542,
"step": 865
},
{
"epoch": 1.2035306334371756,
"grad_norm": 0.5521699786186218,
"learning_rate": 8.207972511274095e-05,
"loss": 0.6413,
"step": 870
},
{
"epoch": 1.210453444098304,
"grad_norm": 0.473796010017395,
"learning_rate": 8.089140413571747e-05,
"loss": 0.6411,
"step": 875
},
{
"epoch": 1.2173762547594325,
"grad_norm": 0.4899975061416626,
"learning_rate": 7.970587727793051e-05,
"loss": 0.637,
"step": 880
},
{
"epoch": 1.2242990654205608,
"grad_norm": 0.4936094284057617,
"learning_rate": 7.852331789086135e-05,
"loss": 0.6401,
"step": 885
},
{
"epoch": 1.231221876081689,
"grad_norm": 0.5000462532043457,
"learning_rate": 7.73438988920784e-05,
"loss": 0.6482,
"step": 890
},
{
"epoch": 1.2381446867428176,
"grad_norm": 0.5123869180679321,
"learning_rate": 7.616779273995252e-05,
"loss": 0.6492,
"step": 895
},
{
"epoch": 1.245067497403946,
"grad_norm": 0.48594337701797485,
"learning_rate": 7.499517140843968e-05,
"loss": 0.6448,
"step": 900
},
{
"epoch": 1.2519903080650745,
"grad_norm": 0.4969227910041809,
"learning_rate": 7.382620636193438e-05,
"loss": 0.636,
"step": 905
},
{
"epoch": 1.2589131187262028,
"grad_norm": 0.4929391145706177,
"learning_rate": 7.266106853019759e-05,
"loss": 0.636,
"step": 910
},
{
"epoch": 1.2658359293873311,
"grad_norm": 0.48114389181137085,
"learning_rate": 7.149992828336274e-05,
"loss": 0.6399,
"step": 915
},
{
"epoch": 1.2727587400484597,
"grad_norm": 0.5209516286849976,
"learning_rate": 7.034295540702397e-05,
"loss": 0.6329,
"step": 920
},
{
"epoch": 1.2796815507095882,
"grad_norm": 0.4901733994483948,
"learning_rate": 6.919031907740922e-05,
"loss": 0.6543,
"step": 925
},
{
"epoch": 1.2866043613707165,
"grad_norm": 0.5102095603942871,
"learning_rate": 6.804218783664288e-05,
"loss": 0.6462,
"step": 930
},
{
"epoch": 1.2935271720318449,
"grad_norm": 0.46668604016304016,
"learning_rate": 6.689872956810102e-05,
"loss": 0.6346,
"step": 935
},
{
"epoch": 1.3004499826929734,
"grad_norm": 0.48058241605758667,
"learning_rate": 6.5760111471863e-05,
"loss": 0.6302,
"step": 940
},
{
"epoch": 1.3073727933541017,
"grad_norm": 0.5134970545768738,
"learning_rate": 6.46265000402628e-05,
"loss": 0.6389,
"step": 945
},
{
"epoch": 1.3142956040152303,
"grad_norm": 0.49927303194999695,
"learning_rate": 6.349806103354417e-05,
"loss": 0.6265,
"step": 950
},
{
"epoch": 1.3212184146763586,
"grad_norm": 0.5016152858734131,
"learning_rate": 6.23749594556226e-05,
"loss": 0.6255,
"step": 955
},
{
"epoch": 1.328141225337487,
"grad_norm": 0.4818898141384125,
"learning_rate": 6.125735952995775e-05,
"loss": 0.6299,
"step": 960
},
{
"epoch": 1.3350640359986155,
"grad_norm": 0.5072540044784546,
"learning_rate": 6.0145424675540394e-05,
"loss": 0.6256,
"step": 965
},
{
"epoch": 1.341986846659744,
"grad_norm": 0.47364550828933716,
"learning_rate": 5.9039317482996735e-05,
"loss": 0.6369,
"step": 970
},
{
"epoch": 1.3489096573208723,
"grad_norm": 0.4561605453491211,
"learning_rate": 5.793919969081374e-05,
"loss": 0.6136,
"step": 975
},
{
"epoch": 1.3558324679820006,
"grad_norm": 0.5107749700546265,
"learning_rate": 5.68452321616894e-05,
"loss": 0.6104,
"step": 980
},
{
"epoch": 1.3627552786431292,
"grad_norm": 0.4962555170059204,
"learning_rate": 5.5757574859010886e-05,
"loss": 0.6367,
"step": 985
},
{
"epoch": 1.3696780893042575,
"grad_norm": 0.5515092015266418,
"learning_rate": 5.467638682346403e-05,
"loss": 0.6087,
"step": 990
},
{
"epoch": 1.376600899965386,
"grad_norm": 0.5041881799697876,
"learning_rate": 5.3601826149777966e-05,
"loss": 0.6424,
"step": 995
},
{
"epoch": 1.3835237106265144,
"grad_norm": 0.4918469488620758,
"learning_rate": 5.25340499636082e-05,
"loss": 0.6283,
"step": 1000
},
{
"epoch": 1.3904465212876427,
"grad_norm": 0.5199057459831238,
"learning_rate": 5.147321439856091e-05,
"loss": 0.6248,
"step": 1005
},
{
"epoch": 1.3973693319487712,
"grad_norm": 0.46590614318847656,
"learning_rate": 5.041947457336274e-05,
"loss": 0.6309,
"step": 1010
},
{
"epoch": 1.4042921426098995,
"grad_norm": 0.45809245109558105,
"learning_rate": 4.937298456917889e-05,
"loss": 0.6095,
"step": 1015
},
{
"epoch": 1.411214953271028,
"grad_norm": 0.5258966088294983,
"learning_rate": 4.8333897407082896e-05,
"loss": 0.6133,
"step": 1020
},
{
"epoch": 1.4181377639321564,
"grad_norm": 0.4614286720752716,
"learning_rate": 4.7302365025681206e-05,
"loss": 0.623,
"step": 1025
},
{
"epoch": 1.425060574593285,
"grad_norm": 0.493114173412323,
"learning_rate": 4.627853825889642e-05,
"loss": 0.6017,
"step": 1030
},
{
"epoch": 1.4319833852544133,
"grad_norm": 0.481393039226532,
"learning_rate": 4.526256681391192e-05,
"loss": 0.616,
"step": 1035
},
{
"epoch": 1.4389061959155418,
"grad_norm": 0.5155964493751526,
"learning_rate": 4.4254599249281016e-05,
"loss": 0.5972,
"step": 1040
},
{
"epoch": 1.4458290065766701,
"grad_norm": 0.5064235925674438,
"learning_rate": 4.3254782953204375e-05,
"loss": 0.6093,
"step": 1045
},
{
"epoch": 1.4527518172377984,
"grad_norm": 0.4666582942008972,
"learning_rate": 4.2263264121978505e-05,
"loss": 0.6126,
"step": 1050
},
{
"epoch": 1.459674627898927,
"grad_norm": 0.5063011646270752,
"learning_rate": 4.12801877386183e-05,
"loss": 0.6026,
"step": 1055
},
{
"epoch": 1.4665974385600553,
"grad_norm": 0.47389891743659973,
"learning_rate": 4.030569755165736e-05,
"loss": 0.6116,
"step": 1060
},
{
"epoch": 1.4735202492211839,
"grad_norm": 0.46991026401519775,
"learning_rate": 3.9339936054128466e-05,
"loss": 0.6106,
"step": 1065
},
{
"epoch": 1.4804430598823122,
"grad_norm": 0.49021977186203003,
"learning_rate": 3.8383044462727826e-05,
"loss": 0.5915,
"step": 1070
},
{
"epoch": 1.4873658705434407,
"grad_norm": 0.5076444149017334,
"learning_rate": 3.7435162697166107e-05,
"loss": 0.6187,
"step": 1075
},
{
"epoch": 1.494288681204569,
"grad_norm": 0.4995920956134796,
"learning_rate": 3.649642935970859e-05,
"loss": 0.6204,
"step": 1080
},
{
"epoch": 1.5012114918656976,
"grad_norm": 0.4842764735221863,
"learning_rate": 3.556698171490871e-05,
"loss": 0.6013,
"step": 1085
},
{
"epoch": 1.508134302526826,
"grad_norm": 0.48687830567359924,
"learning_rate": 3.464695566953644e-05,
"loss": 0.5962,
"step": 1090
},
{
"epoch": 1.5150571131879542,
"grad_norm": 0.5043684840202332,
"learning_rate": 3.373648575270576e-05,
"loss": 0.6021,
"step": 1095
},
{
"epoch": 1.5219799238490828,
"grad_norm": 0.4804685711860657,
"learning_rate": 3.283570509620344e-05,
"loss": 0.606,
"step": 1100
},
{
"epoch": 1.5289027345102113,
"grad_norm": 0.4706571400165558,
"learning_rate": 3.1944745415021915e-05,
"loss": 0.607,
"step": 1105
},
{
"epoch": 1.5358255451713396,
"grad_norm": 0.43756231665611267,
"learning_rate": 3.106373698809958e-05,
"loss": 0.6149,
"step": 1110
},
{
"epoch": 1.542748355832468,
"grad_norm": 0.47284871339797974,
"learning_rate": 3.0192808639271065e-05,
"loss": 0.6173,
"step": 1115
},
{
"epoch": 1.5496711664935963,
"grad_norm": 0.46921491622924805,
"learning_rate": 2.933208771843018e-05,
"loss": 0.6036,
"step": 1120
},
{
"epoch": 1.5565939771547248,
"grad_norm": 0.4850091338157654,
"learning_rate": 2.8481700082908268e-05,
"loss": 0.599,
"step": 1125
},
{
"epoch": 1.5635167878158533,
"grad_norm": 0.48241791129112244,
"learning_rate": 2.7641770079071127e-05,
"loss": 0.6,
"step": 1130
},
{
"epoch": 1.5704395984769817,
"grad_norm": 0.4461960792541504,
"learning_rate": 2.681242052413656e-05,
"loss": 0.5818,
"step": 1135
},
{
"epoch": 1.57736240913811,
"grad_norm": 0.49775242805480957,
"learning_rate": 2.59937726882158e-05,
"loss": 0.5829,
"step": 1140
},
{
"epoch": 1.5842852197992385,
"grad_norm": 0.5127538442611694,
"learning_rate": 2.5185946276580918e-05,
"loss": 0.6164,
"step": 1145
},
{
"epoch": 1.591208030460367,
"grad_norm": 0.47250741720199585,
"learning_rate": 2.4389059412161087e-05,
"loss": 0.5978,
"step": 1150
},
{
"epoch": 1.5981308411214954,
"grad_norm": 0.4654916524887085,
"learning_rate": 2.3603228618270356e-05,
"loss": 0.5952,
"step": 1155
},
{
"epoch": 1.6050536517826237,
"grad_norm": 0.45718511939048767,
"learning_rate": 2.2828568801569283e-05,
"loss": 0.5719,
"step": 1160
},
{
"epoch": 1.611976462443752,
"grad_norm": 0.443266361951828,
"learning_rate": 2.2065193235262782e-05,
"loss": 0.5734,
"step": 1165
},
{
"epoch": 1.6188992731048806,
"grad_norm": 0.46032777428627014,
"learning_rate": 2.1313213542537058e-05,
"loss": 0.5949,
"step": 1170
},
{
"epoch": 1.6258220837660091,
"grad_norm": 0.4900462329387665,
"learning_rate": 2.0572739680237717e-05,
"loss": 0.6047,
"step": 1175
},
{
"epoch": 1.6327448944271374,
"grad_norm": 0.46509256958961487,
"learning_rate": 1.984387992279153e-05,
"loss": 0.6182,
"step": 1180
},
{
"epoch": 1.6396677050882658,
"grad_norm": 0.4681301712989807,
"learning_rate": 1.912674084637409e-05,
"loss": 0.5983,
"step": 1185
},
{
"epoch": 1.6465905157493943,
"grad_norm": 0.4606594145298004,
"learning_rate": 1.8421427313326046e-05,
"loss": 0.5796,
"step": 1190
},
{
"epoch": 1.6535133264105226,
"grad_norm": 0.47154656052589417,
"learning_rate": 1.7728042456819704e-05,
"loss": 0.5992,
"step": 1195
},
{
"epoch": 1.6604361370716512,
"grad_norm": 0.4498773515224457,
"learning_rate": 1.7046687665778717e-05,
"loss": 0.5811,
"step": 1200
},
{
"epoch": 1.6673589477327795,
"grad_norm": 0.45348694920539856,
"learning_rate": 1.6377462570052438e-05,
"loss": 0.5784,
"step": 1205
},
{
"epoch": 1.6742817583939078,
"grad_norm": 0.536761999130249,
"learning_rate": 1.5720465025847987e-05,
"loss": 0.5861,
"step": 1210
},
{
"epoch": 1.6812045690550363,
"grad_norm": 0.4612574577331543,
"learning_rate": 1.5075791101421166e-05,
"loss": 0.5904,
"step": 1215
},
{
"epoch": 1.6881273797161649,
"grad_norm": 0.47577938437461853,
"learning_rate": 1.4443535063029279e-05,
"loss": 0.593,
"step": 1220
},
{
"epoch": 1.6950501903772932,
"grad_norm": 0.4678874611854553,
"learning_rate": 1.3823789361147044e-05,
"loss": 0.5761,
"step": 1225
},
{
"epoch": 1.7019730010384215,
"grad_norm": 0.4748591482639313,
"learning_rate": 1.3216644616948282e-05,
"loss": 0.5878,
"step": 1230
},
{
"epoch": 1.70889581169955,
"grad_norm": 0.46718382835388184,
"learning_rate": 1.2622189609054979e-05,
"loss": 0.5748,
"step": 1235
},
{
"epoch": 1.7158186223606784,
"grad_norm": 0.49095532298088074,
"learning_rate": 1.2040511260555831e-05,
"loss": 0.5797,
"step": 1240
},
{
"epoch": 1.722741433021807,
"grad_norm": 0.4614841341972351,
"learning_rate": 1.1471694626296049e-05,
"loss": 0.5787,
"step": 1245
},
{
"epoch": 1.7296642436829353,
"grad_norm": 0.45303580164909363,
"learning_rate": 1.0915822880440308e-05,
"loss": 0.5796,
"step": 1250
},
{
"epoch": 1.7365870543440636,
"grad_norm": 0.46378132700920105,
"learning_rate": 1.037297730431085e-05,
"loss": 0.573,
"step": 1255
},
{
"epoch": 1.7435098650051921,
"grad_norm": 0.45119667053222656,
"learning_rate": 9.843237274502282e-06,
"loss": 0.5825,
"step": 1260
},
{
"epoch": 1.7504326756663207,
"grad_norm": 0.47857195138931274,
"learning_rate": 9.326680251274777e-06,
"loss": 0.5904,
"step": 1265
},
{
"epoch": 1.757355486327449,
"grad_norm": 0.4850415885448456,
"learning_rate": 8.823381767227667e-06,
"loss": 0.5877,
"step": 1270
},
{
"epoch": 1.7642782969885773,
"grad_norm": 0.4598202407360077,
"learning_rate": 8.333415416254831e-06,
"loss": 0.5749,
"step": 1275
},
{
"epoch": 1.7712011076497056,
"grad_norm": 0.4723171293735504,
"learning_rate": 7.856852842783547e-06,
"loss": 0.5721,
"step": 1280
},
{
"epoch": 1.7781239183108342,
"grad_norm": 0.4513912498950958,
"learning_rate": 7.393763731298387e-06,
"loss": 0.5689,
"step": 1285
},
{
"epoch": 1.7850467289719627,
"grad_norm": 0.4459558427333832,
"learning_rate": 6.944215796151765e-06,
"loss": 0.5726,
"step": 1290
},
{
"epoch": 1.791969539633091,
"grad_norm": 0.4863142669200897,
"learning_rate": 6.5082747716625255e-06,
"loss": 0.5853,
"step": 1295
},
{
"epoch": 1.7988923502942193,
"grad_norm": 0.44126665592193604,
"learning_rate": 6.086004402504098e-06,
"loss": 0.5599,
"step": 1300
},
{
"epoch": 1.8058151609553479,
"grad_norm": 0.46068254113197327,
"learning_rate": 5.6774664343834495e-06,
"loss": 0.5939,
"step": 1305
},
{
"epoch": 1.8127379716164764,
"grad_norm": 0.46407076716423035,
"learning_rate": 5.2827206050125585e-06,
"loss": 0.5782,
"step": 1310
},
{
"epoch": 1.8196607822776047,
"grad_norm": 0.4844149649143219,
"learning_rate": 4.9018246353732755e-06,
"loss": 0.5958,
"step": 1315
},
{
"epoch": 1.826583592938733,
"grad_norm": 0.44412821531295776,
"learning_rate": 4.534834221277262e-06,
"loss": 0.568,
"step": 1320
},
{
"epoch": 1.8335064035998614,
"grad_norm": 0.4620952308177948,
"learning_rate": 4.181803025221898e-06,
"loss": 0.5797,
"step": 1325
},
{
"epoch": 1.84042921426099,
"grad_norm": 0.46902209520339966,
"learning_rate": 3.842782668543599e-06,
"loss": 0.5784,
"step": 1330
},
{
"epoch": 1.8473520249221185,
"grad_norm": 0.4106810390949249,
"learning_rate": 3.5178227238696347e-06,
"loss": 0.5638,
"step": 1335
},
{
"epoch": 1.8542748355832468,
"grad_norm": 0.47497299313545227,
"learning_rate": 3.2069707078694057e-06,
"loss": 0.5582,
"step": 1340
},
{
"epoch": 1.8611976462443751,
"grad_norm": 0.44151052832603455,
"learning_rate": 2.9102720743064527e-06,
"loss": 0.5876,
"step": 1345
},
{
"epoch": 1.8681204569055037,
"grad_norm": 0.4456900656223297,
"learning_rate": 2.6277702073919997e-06,
"loss": 0.5659,
"step": 1350
},
{
"epoch": 1.8750432675666322,
"grad_norm": 0.44609665870666504,
"learning_rate": 2.3595064154412374e-06,
"loss": 0.5683,
"step": 1355
},
{
"epoch": 1.8819660782277605,
"grad_norm": 0.4577263593673706,
"learning_rate": 2.1055199248330546e-06,
"loss": 0.5716,
"step": 1360
},
{
"epoch": 1.8888888888888888,
"grad_norm": 0.4626077711582184,
"learning_rate": 1.8658478742742625e-06,
"loss": 0.5688,
"step": 1365
},
{
"epoch": 1.8958116995500172,
"grad_norm": 0.4604962170124054,
"learning_rate": 1.6405253093690343e-06,
"loss": 0.5702,
"step": 1370
},
{
"epoch": 1.9027345102111457,
"grad_norm": 0.46187230944633484,
"learning_rate": 1.429585177494419e-06,
"loss": 0.5671,
"step": 1375
},
{
"epoch": 1.9096573208722742,
"grad_norm": 0.44326135516166687,
"learning_rate": 1.2330583229827053e-06,
"loss": 0.5812,
"step": 1380
},
{
"epoch": 1.9165801315334026,
"grad_norm": 0.46138623356819153,
"learning_rate": 1.0509734826112394e-06,
"loss": 0.5651,
"step": 1385
},
{
"epoch": 1.9235029421945309,
"grad_norm": 0.46161016821861267,
"learning_rate": 8.833572814004321e-07,
"loss": 0.5634,
"step": 1390
},
{
"epoch": 1.9304257528556594,
"grad_norm": 0.4505715072154999,
"learning_rate": 7.302342287205943e-07,
"loss": 0.5688,
"step": 1395
},
{
"epoch": 1.937348563516788,
"grad_norm": 0.4546772837638855,
"learning_rate": 5.916267147080934e-07,
"loss": 0.5727,
"step": 1400
},
{
"epoch": 1.9442713741779163,
"grad_norm": 0.4859260022640228,
"learning_rate": 4.6755500699138253e-07,
"loss": 0.5614,
"step": 1405
},
{
"epoch": 1.9511941848390446,
"grad_norm": 0.483917236328125,
"learning_rate": 3.5803724772740523e-07,
"loss": 0.5635,
"step": 1410
},
{
"epoch": 1.958116995500173,
"grad_norm": 0.45442262291908264,
"learning_rate": 2.630894509488058e-07,
"loss": 0.5754,
"step": 1415
},
{
"epoch": 1.9650398061613015,
"grad_norm": 0.45076045393943787,
"learning_rate": 1.8272550022230272e-07,
"loss": 0.5665,
"step": 1420
},
{
"epoch": 1.97196261682243,
"grad_norm": 0.4759106934070587,
"learning_rate": 1.1695714661856816e-07,
"loss": 0.5583,
"step": 1425
},
{
"epoch": 1.9788854274835583,
"grad_norm": 0.48755067586898804,
"learning_rate": 6.579400699397998e-08,
"loss": 0.57,
"step": 1430
},
{
"epoch": 1.9858082381446867,
"grad_norm": 0.4441988170146942,
"learning_rate": 2.9243562584402305e-08,
"loss": 0.5784,
"step": 1435
},
{
"epoch": 1.9927310488058152,
"grad_norm": 0.44851863384246826,
"learning_rate": 7.311157911249478e-09,
"loss": 0.5849,
"step": 1440
},
{
"epoch": 1.9982692973347178,
"eval_loss": 0.7809522747993469,
"eval_runtime": 54.6607,
"eval_samples_per_second": 37.852,
"eval_steps_per_second": 2.378,
"step": 1444
},
{
"epoch": 1.9982692973347178,
"step": 1444,
"total_flos": 4.75579504107769e+18,
"train_loss": 0.8681613599824773,
"train_runtime": 10264.6857,
"train_samples_per_second": 9.005,
"train_steps_per_second": 0.141
}
],
"logging_steps": 5,
"max_steps": 1444,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": false,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.75579504107769e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}