llama3-8b-summarize-gpt4o-128k / trainer_state.json
chansung's picture
Model save
b0bd756 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 9.954337899543379,
"eval_steps": 500,
"global_step": 1090,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0091324200913242,
"grad_norm": 3.1570751667022705,
"learning_rate": 1.8348623853211011e-06,
"loss": 2.2349,
"step": 1
},
{
"epoch": 0.045662100456621,
"grad_norm": 3.327242374420166,
"learning_rate": 9.174311926605506e-06,
"loss": 2.2947,
"step": 5
},
{
"epoch": 0.091324200913242,
"grad_norm": 3.6639091968536377,
"learning_rate": 1.834862385321101e-05,
"loss": 2.1754,
"step": 10
},
{
"epoch": 0.136986301369863,
"grad_norm": 6.006739616394043,
"learning_rate": 2.7522935779816515e-05,
"loss": 1.8388,
"step": 15
},
{
"epoch": 0.182648401826484,
"grad_norm": 1.5798718929290771,
"learning_rate": 3.669724770642202e-05,
"loss": 1.4696,
"step": 20
},
{
"epoch": 0.228310502283105,
"grad_norm": 1.038404107093811,
"learning_rate": 4.587155963302753e-05,
"loss": 1.2583,
"step": 25
},
{
"epoch": 0.273972602739726,
"grad_norm": 0.5968286395072937,
"learning_rate": 5.504587155963303e-05,
"loss": 1.1721,
"step": 30
},
{
"epoch": 0.319634703196347,
"grad_norm": 0.42477676272392273,
"learning_rate": 6.422018348623854e-05,
"loss": 1.0966,
"step": 35
},
{
"epoch": 0.365296803652968,
"grad_norm": 0.46538373827934265,
"learning_rate": 7.339449541284404e-05,
"loss": 1.0378,
"step": 40
},
{
"epoch": 0.410958904109589,
"grad_norm": 0.32725027203559875,
"learning_rate": 8.256880733944955e-05,
"loss": 0.996,
"step": 45
},
{
"epoch": 0.45662100456621,
"grad_norm": 0.27102047204971313,
"learning_rate": 9.174311926605506e-05,
"loss": 0.9549,
"step": 50
},
{
"epoch": 0.502283105022831,
"grad_norm": 0.20055657625198364,
"learning_rate": 0.00010091743119266055,
"loss": 0.9426,
"step": 55
},
{
"epoch": 0.547945205479452,
"grad_norm": 0.24150966107845306,
"learning_rate": 0.00011009174311926606,
"loss": 0.918,
"step": 60
},
{
"epoch": 0.593607305936073,
"grad_norm": 0.20545370876789093,
"learning_rate": 0.00011926605504587157,
"loss": 0.8995,
"step": 65
},
{
"epoch": 0.639269406392694,
"grad_norm": 0.24165408313274384,
"learning_rate": 0.00012844036697247707,
"loss": 0.8851,
"step": 70
},
{
"epoch": 0.684931506849315,
"grad_norm": 0.22626952826976776,
"learning_rate": 0.00013761467889908258,
"loss": 0.8729,
"step": 75
},
{
"epoch": 0.730593607305936,
"grad_norm": 0.2153974175453186,
"learning_rate": 0.0001467889908256881,
"loss": 0.8588,
"step": 80
},
{
"epoch": 0.776255707762557,
"grad_norm": 0.22552534937858582,
"learning_rate": 0.0001559633027522936,
"loss": 0.8453,
"step": 85
},
{
"epoch": 0.821917808219178,
"grad_norm": 0.2299458086490631,
"learning_rate": 0.0001651376146788991,
"loss": 0.8491,
"step": 90
},
{
"epoch": 0.867579908675799,
"grad_norm": 0.38727596402168274,
"learning_rate": 0.00017431192660550458,
"loss": 0.839,
"step": 95
},
{
"epoch": 0.91324200913242,
"grad_norm": 0.5172630548477173,
"learning_rate": 0.00018348623853211012,
"loss": 0.8359,
"step": 100
},
{
"epoch": 0.958904109589041,
"grad_norm": 0.3504508137702942,
"learning_rate": 0.0001926605504587156,
"loss": 0.8176,
"step": 105
},
{
"epoch": 0.9954337899543378,
"eval_loss": 2.11497163772583,
"eval_runtime": 0.7274,
"eval_samples_per_second": 13.748,
"eval_steps_per_second": 1.375,
"step": 109
},
{
"epoch": 1.004566210045662,
"grad_norm": 0.21483348309993744,
"learning_rate": 0.00019999948721966259,
"loss": 0.812,
"step": 110
},
{
"epoch": 1.0502283105022832,
"grad_norm": 0.25326865911483765,
"learning_rate": 0.00019998154046002822,
"loss": 0.8028,
"step": 115
},
{
"epoch": 1.095890410958904,
"grad_norm": 0.23645517230033875,
"learning_rate": 0.0001999379599421534,
"loss": 0.8003,
"step": 120
},
{
"epoch": 1.1415525114155252,
"grad_norm": 0.22504329681396484,
"learning_rate": 0.00019986875683942535,
"loss": 0.7939,
"step": 125
},
{
"epoch": 1.187214611872146,
"grad_norm": 0.2312326580286026,
"learning_rate": 0.00019977394889447524,
"loss": 0.7847,
"step": 130
},
{
"epoch": 1.2328767123287672,
"grad_norm": 0.21880114078521729,
"learning_rate": 0.00019965356041462955,
"loss": 0.7856,
"step": 135
},
{
"epoch": 1.278538812785388,
"grad_norm": 0.19915428757667542,
"learning_rate": 0.00019950762226567781,
"loss": 0.7772,
"step": 140
},
{
"epoch": 1.3242009132420092,
"grad_norm": 0.21661226451396942,
"learning_rate": 0.00019933617186395917,
"loss": 0.7698,
"step": 145
},
{
"epoch": 1.36986301369863,
"grad_norm": 0.23418781161308289,
"learning_rate": 0.00019913925316676945,
"loss": 0.771,
"step": 150
},
{
"epoch": 1.4155251141552512,
"grad_norm": 0.2712844908237457,
"learning_rate": 0.00019891691666109113,
"loss": 0.7754,
"step": 155
},
{
"epoch": 1.461187214611872,
"grad_norm": 0.20658765733242035,
"learning_rate": 0.00019866921935064906,
"loss": 0.772,
"step": 160
},
{
"epoch": 1.5068493150684932,
"grad_norm": 0.2858370542526245,
"learning_rate": 0.00019839622474129596,
"loss": 0.7648,
"step": 165
},
{
"epoch": 1.5525114155251143,
"grad_norm": 0.2993509769439697,
"learning_rate": 0.00019809800282473013,
"loss": 0.7553,
"step": 170
},
{
"epoch": 1.5981735159817352,
"grad_norm": 0.19971656799316406,
"learning_rate": 0.0001977746300605507,
"loss": 0.7591,
"step": 175
},
{
"epoch": 1.643835616438356,
"grad_norm": 0.2134828269481659,
"learning_rate": 0.00019742618935665476,
"loss": 0.7549,
"step": 180
},
{
"epoch": 1.6894977168949772,
"grad_norm": 0.18528909981250763,
"learning_rate": 0.00019705277004798073,
"loss": 0.7511,
"step": 185
},
{
"epoch": 1.7351598173515983,
"grad_norm": 0.22641536593437195,
"learning_rate": 0.0001966544678736044,
"loss": 0.754,
"step": 190
},
{
"epoch": 1.7808219178082192,
"grad_norm": 0.23213805258274078,
"learning_rate": 0.00019623138495219292,
"loss": 0.7476,
"step": 195
},
{
"epoch": 1.82648401826484,
"grad_norm": 0.28454649448394775,
"learning_rate": 0.00019578362975582292,
"loss": 0.7535,
"step": 200
},
{
"epoch": 1.8721461187214612,
"grad_norm": 0.3238324224948883,
"learning_rate": 0.00019531131708217005,
"loss": 0.7489,
"step": 205
},
{
"epoch": 1.9178082191780823,
"grad_norm": 0.23858831822872162,
"learning_rate": 0.0001948145680250766,
"loss": 0.7514,
"step": 210
},
{
"epoch": 1.9634703196347032,
"grad_norm": 0.21854574978351593,
"learning_rate": 0.00019429350994350483,
"loss": 0.7464,
"step": 215
},
{
"epoch": 2.0,
"eval_loss": 2.1313483715057373,
"eval_runtime": 0.7223,
"eval_samples_per_second": 13.845,
"eval_steps_per_second": 1.384,
"step": 219
},
{
"epoch": 2.009132420091324,
"grad_norm": 0.24195519089698792,
"learning_rate": 0.00019374827642888398,
"loss": 0.7367,
"step": 220
},
{
"epoch": 2.0547945205479454,
"grad_norm": 0.2783088684082031,
"learning_rate": 0.0001931790072708596,
"loss": 0.7283,
"step": 225
},
{
"epoch": 2.1004566210045663,
"grad_norm": 0.25543445348739624,
"learning_rate": 0.00019258584842145343,
"loss": 0.7298,
"step": 230
},
{
"epoch": 2.146118721461187,
"grad_norm": 0.23951423168182373,
"learning_rate": 0.00019196895195764362,
"loss": 0.7296,
"step": 235
},
{
"epoch": 2.191780821917808,
"grad_norm": 0.2517695426940918,
"learning_rate": 0.0001913284760423745,
"loss": 0.719,
"step": 240
},
{
"epoch": 2.237442922374429,
"grad_norm": 0.22162802517414093,
"learning_rate": 0.00019066458488400584,
"loss": 0.7252,
"step": 245
},
{
"epoch": 2.2831050228310503,
"grad_norm": 0.2749063968658447,
"learning_rate": 0.00018997744869421246,
"loss": 0.7253,
"step": 250
},
{
"epoch": 2.328767123287671,
"grad_norm": 0.2638462483882904,
"learning_rate": 0.00018926724364434446,
"loss": 0.718,
"step": 255
},
{
"epoch": 2.374429223744292,
"grad_norm": 0.23074951767921448,
"learning_rate": 0.0001885341518202595,
"loss": 0.7203,
"step": 260
},
{
"epoch": 2.4200913242009134,
"grad_norm": 0.18681403994560242,
"learning_rate": 0.00018777836117563892,
"loss": 0.7253,
"step": 265
},
{
"epoch": 2.4657534246575343,
"grad_norm": 0.19247153401374817,
"learning_rate": 0.00018700006548379898,
"loss": 0.7175,
"step": 270
},
{
"epoch": 2.5114155251141552,
"grad_norm": 0.18738825619220734,
"learning_rate": 0.0001861994642880105,
"loss": 0.7206,
"step": 275
},
{
"epoch": 2.557077625570776,
"grad_norm": 0.2461051046848297,
"learning_rate": 0.00018537676285033887,
"loss": 0.7134,
"step": 280
},
{
"epoch": 2.602739726027397,
"grad_norm": 0.31902387738227844,
"learning_rate": 0.0001845321720990181,
"loss": 0.7135,
"step": 285
},
{
"epoch": 2.6484018264840183,
"grad_norm": 0.19439548254013062,
"learning_rate": 0.00018366590857437184,
"loss": 0.7194,
"step": 290
},
{
"epoch": 2.6940639269406392,
"grad_norm": 0.18397387862205505,
"learning_rate": 0.00018277819437329576,
"loss": 0.7172,
"step": 295
},
{
"epoch": 2.73972602739726,
"grad_norm": 0.1935635656118393,
"learning_rate": 0.00018186925709231532,
"loss": 0.7188,
"step": 300
},
{
"epoch": 2.7853881278538815,
"grad_norm": 0.20888376235961914,
"learning_rate": 0.0001809393297692334,
"loss": 0.71,
"step": 305
},
{
"epoch": 2.8310502283105023,
"grad_norm": 0.20157787203788757,
"learning_rate": 0.0001799886508233829,
"loss": 0.7152,
"step": 310
},
{
"epoch": 2.8767123287671232,
"grad_norm": 0.201010599732399,
"learning_rate": 0.0001790174639944997,
"loss": 0.7147,
"step": 315
},
{
"epoch": 2.922374429223744,
"grad_norm": 0.20877932012081146,
"learning_rate": 0.00017802601828023138,
"loss": 0.7101,
"step": 320
},
{
"epoch": 2.968036529680365,
"grad_norm": 0.19995956122875214,
"learning_rate": 0.00017701456787229804,
"loss": 0.7128,
"step": 325
},
{
"epoch": 2.9954337899543377,
"eval_loss": 2.144402027130127,
"eval_runtime": 0.7303,
"eval_samples_per_second": 13.694,
"eval_steps_per_second": 1.369,
"step": 328
},
{
"epoch": 3.0136986301369864,
"grad_norm": 0.18146736919879913,
"learning_rate": 0.0001759833720913214,
"loss": 0.7071,
"step": 330
},
{
"epoch": 3.0593607305936072,
"grad_norm": 0.20878320932388306,
"learning_rate": 0.00017493269532033883,
"loss": 0.6943,
"step": 335
},
{
"epoch": 3.105022831050228,
"grad_norm": 0.24599188566207886,
"learning_rate": 0.0001738628069370195,
"loss": 0.6968,
"step": 340
},
{
"epoch": 3.1506849315068495,
"grad_norm": 0.22413116693496704,
"learning_rate": 0.00017277398124460023,
"loss": 0.6949,
"step": 345
},
{
"epoch": 3.1963470319634704,
"grad_norm": 0.22311843931674957,
"learning_rate": 0.000171666497401558,
"loss": 0.6988,
"step": 350
},
{
"epoch": 3.2420091324200913,
"grad_norm": 0.19590473175048828,
"learning_rate": 0.0001705406393500381,
"loss": 0.692,
"step": 355
},
{
"epoch": 3.287671232876712,
"grad_norm": 0.18890812993049622,
"learning_rate": 0.00016939669574305566,
"loss": 0.6923,
"step": 360
},
{
"epoch": 3.3333333333333335,
"grad_norm": 0.19763173162937164,
"learning_rate": 0.0001682349598704892,
"loss": 0.6899,
"step": 365
},
{
"epoch": 3.3789954337899544,
"grad_norm": 0.21059083938598633,
"learning_rate": 0.00016705572958388576,
"loss": 0.689,
"step": 370
},
{
"epoch": 3.4246575342465753,
"grad_norm": 0.2090144008398056,
"learning_rate": 0.00016585930722009601,
"loss": 0.6909,
"step": 375
},
{
"epoch": 3.470319634703196,
"grad_norm": 0.1937251091003418,
"learning_rate": 0.00016464599952375998,
"loss": 0.6943,
"step": 380
},
{
"epoch": 3.5159817351598175,
"grad_norm": 0.19087150692939758,
"learning_rate": 0.000163416117568662,
"loss": 0.6936,
"step": 385
},
{
"epoch": 3.5616438356164384,
"grad_norm": 0.21907366812229156,
"learning_rate": 0.0001621699766779763,
"loss": 0.6928,
"step": 390
},
{
"epoch": 3.6073059360730593,
"grad_norm": 0.1987326741218567,
"learning_rate": 0.00016090789634342278,
"loss": 0.6913,
"step": 395
},
{
"epoch": 3.65296803652968,
"grad_norm": 0.19414757192134857,
"learning_rate": 0.00015963020014335438,
"loss": 0.6873,
"step": 400
},
{
"epoch": 3.6986301369863015,
"grad_norm": 0.19000251591205597,
"learning_rate": 0.0001583372156597961,
"loss": 0.6895,
"step": 405
},
{
"epoch": 3.7442922374429224,
"grad_norm": 0.17563720047473907,
"learning_rate": 0.00015702927439445826,
"loss": 0.6905,
"step": 410
},
{
"epoch": 3.7899543378995433,
"grad_norm": 0.22648802399635315,
"learning_rate": 0.00015570671168374438,
"loss": 0.685,
"step": 415
},
{
"epoch": 3.8356164383561646,
"grad_norm": 0.1822802722454071,
"learning_rate": 0.00015436986661277577,
"loss": 0.6897,
"step": 420
},
{
"epoch": 3.8812785388127855,
"grad_norm": 0.18406091630458832,
"learning_rate": 0.0001530190819284555,
"loss": 0.6849,
"step": 425
},
{
"epoch": 3.9269406392694064,
"grad_norm": 0.1968332827091217,
"learning_rate": 0.00015165470395159313,
"loss": 0.6841,
"step": 430
},
{
"epoch": 3.9726027397260273,
"grad_norm": 0.2520104646682739,
"learning_rate": 0.0001502770824881133,
"loss": 0.6924,
"step": 435
},
{
"epoch": 4.0,
"eval_loss": 2.1630699634552,
"eval_runtime": 0.7164,
"eval_samples_per_second": 13.958,
"eval_steps_per_second": 1.396,
"step": 438
},
{
"epoch": 4.018264840182648,
"grad_norm": 0.20841118693351746,
"learning_rate": 0.00014888657073937076,
"loss": 0.6786,
"step": 440
},
{
"epoch": 4.063926940639269,
"grad_norm": 0.2335735559463501,
"learning_rate": 0.00014748352521159493,
"loss": 0.6687,
"step": 445
},
{
"epoch": 4.109589041095891,
"grad_norm": 0.23169922828674316,
"learning_rate": 0.0001460683056244869,
"loss": 0.6701,
"step": 450
},
{
"epoch": 4.155251141552512,
"grad_norm": 0.20742633938789368,
"learning_rate": 0.00014464127481899312,
"loss": 0.6678,
"step": 455
},
{
"epoch": 4.200913242009133,
"grad_norm": 0.233662411570549,
"learning_rate": 0.00014320279866427796,
"loss": 0.669,
"step": 460
},
{
"epoch": 4.2465753424657535,
"grad_norm": 0.22496378421783447,
"learning_rate": 0.00014175324596392075,
"loss": 0.6695,
"step": 465
},
{
"epoch": 4.292237442922374,
"grad_norm": 0.1961798071861267,
"learning_rate": 0.00014029298836135988,
"loss": 0.6694,
"step": 470
},
{
"epoch": 4.337899543378995,
"grad_norm": 0.21112968027591705,
"learning_rate": 0.00013882240024460927,
"loss": 0.6758,
"step": 475
},
{
"epoch": 4.383561643835616,
"grad_norm": 0.21480615437030792,
"learning_rate": 0.0001373418586502706,
"loss": 0.6685,
"step": 480
},
{
"epoch": 4.429223744292237,
"grad_norm": 0.24116870760917664,
"learning_rate": 0.0001358517431668672,
"loss": 0.6686,
"step": 485
},
{
"epoch": 4.474885844748858,
"grad_norm": 0.20268571376800537,
"learning_rate": 0.00013435243583752294,
"loss": 0.6642,
"step": 490
},
{
"epoch": 4.52054794520548,
"grad_norm": 0.19283969700336456,
"learning_rate": 0.00013284432106201233,
"loss": 0.6729,
"step": 495
},
{
"epoch": 4.566210045662101,
"grad_norm": 0.18919415771961212,
"learning_rate": 0.00013132778549820618,
"loss": 0.6713,
"step": 500
},
{
"epoch": 4.6118721461187215,
"grad_norm": 0.19388696551322937,
"learning_rate": 0.00012980321796293836,
"loss": 0.6698,
"step": 505
},
{
"epoch": 4.657534246575342,
"grad_norm": 0.19509977102279663,
"learning_rate": 0.00012827100933231905,
"loss": 0.6732,
"step": 510
},
{
"epoch": 4.703196347031963,
"grad_norm": 0.19472207129001617,
"learning_rate": 0.00012673155244151985,
"loss": 0.6689,
"step": 515
},
{
"epoch": 4.748858447488584,
"grad_norm": 0.1974049210548401,
"learning_rate": 0.000125185241984057,
"loss": 0.6661,
"step": 520
},
{
"epoch": 4.794520547945205,
"grad_norm": 0.18362760543823242,
"learning_rate": 0.00012363247441059776,
"loss": 0.6705,
"step": 525
},
{
"epoch": 4.840182648401827,
"grad_norm": 0.1863856315612793,
"learning_rate": 0.00012207364782731655,
"loss": 0.6663,
"step": 530
},
{
"epoch": 4.885844748858448,
"grad_norm": 0.19237390160560608,
"learning_rate": 0.00012050916189382646,
"loss": 0.6701,
"step": 535
},
{
"epoch": 4.931506849315069,
"grad_norm": 0.1915617734193802,
"learning_rate": 0.00011893941772071249,
"loss": 0.6781,
"step": 540
},
{
"epoch": 4.9771689497716896,
"grad_norm": 0.18157874047756195,
"learning_rate": 0.00011736481776669306,
"loss": 0.6777,
"step": 545
},
{
"epoch": 4.995433789954338,
"eval_loss": 2.182265520095825,
"eval_runtime": 0.7269,
"eval_samples_per_second": 13.757,
"eval_steps_per_second": 1.376,
"step": 547
},
{
"epoch": 5.0228310502283104,
"grad_norm": 0.19536516070365906,
"learning_rate": 0.0001157857657354354,
"loss": 0.6657,
"step": 550
},
{
"epoch": 5.068493150684931,
"grad_norm": 0.22191356122493744,
"learning_rate": 0.00011420266647205231,
"loss": 0.6547,
"step": 555
},
{
"epoch": 5.114155251141552,
"grad_norm": 0.2473655790090561,
"learning_rate": 0.00011261592585930576,
"loss": 0.649,
"step": 560
},
{
"epoch": 5.159817351598173,
"grad_norm": 0.21013621985912323,
"learning_rate": 0.00011102595071354472,
"loss": 0.6493,
"step": 565
},
{
"epoch": 5.205479452054795,
"grad_norm": 0.2081068456172943,
"learning_rate": 0.00010943314868040364,
"loss": 0.647,
"step": 570
},
{
"epoch": 5.251141552511416,
"grad_norm": 0.23985308408737183,
"learning_rate": 0.00010783792813028827,
"loss": 0.6505,
"step": 575
},
{
"epoch": 5.296803652968037,
"grad_norm": 0.19800134003162384,
"learning_rate": 0.00010624069805367559,
"loss": 0.6537,
"step": 580
},
{
"epoch": 5.342465753424658,
"grad_norm": 0.21391679346561432,
"learning_rate": 0.00010464186795625482,
"loss": 0.6558,
"step": 585
},
{
"epoch": 5.3881278538812785,
"grad_norm": 0.205993190407753,
"learning_rate": 0.00010304184775393642,
"loss": 0.6578,
"step": 590
},
{
"epoch": 5.433789954337899,
"grad_norm": 0.20510774850845337,
"learning_rate": 0.00010144104766775572,
"loss": 0.6565,
"step": 595
},
{
"epoch": 5.47945205479452,
"grad_norm": 0.1970670372247696,
"learning_rate": 9.983987811869862e-05,
"loss": 0.6515,
"step": 600
},
{
"epoch": 5.525114155251142,
"grad_norm": 0.18836742639541626,
"learning_rate": 9.823874962247564e-05,
"loss": 0.6557,
"step": 605
},
{
"epoch": 5.570776255707763,
"grad_norm": 0.20027703046798706,
"learning_rate": 9.663807268427198e-05,
"loss": 0.6522,
"step": 610
},
{
"epoch": 5.616438356164384,
"grad_norm": 0.2243603765964508,
"learning_rate": 9.503825769350017e-05,
"loss": 0.6496,
"step": 615
},
{
"epoch": 5.662100456621005,
"grad_norm": 0.18932883441448212,
"learning_rate": 9.343971481858246e-05,
"loss": 0.6532,
"step": 620
},
{
"epoch": 5.707762557077626,
"grad_norm": 0.19952723383903503,
"learning_rate": 9.184285390178978e-05,
"loss": 0.6511,
"step": 625
},
{
"epoch": 5.7534246575342465,
"grad_norm": 0.20302745699882507,
"learning_rate": 9.024808435416434e-05,
"loss": 0.6524,
"step": 630
},
{
"epoch": 5.799086757990867,
"grad_norm": 0.1958286166191101,
"learning_rate": 8.865581505055291e-05,
"loss": 0.6543,
"step": 635
},
{
"epoch": 5.844748858447488,
"grad_norm": 0.189736470580101,
"learning_rate": 8.706645422477739e-05,
"loss": 0.6529,
"step": 640
},
{
"epoch": 5.890410958904109,
"grad_norm": 0.2069847285747528,
"learning_rate": 8.548040936496989e-05,
"loss": 0.6542,
"step": 645
},
{
"epoch": 5.936073059360731,
"grad_norm": 0.2031773030757904,
"learning_rate": 8.389808710909881e-05,
"loss": 0.6515,
"step": 650
},
{
"epoch": 5.981735159817352,
"grad_norm": 0.20669801533222198,
"learning_rate": 8.231989314071317e-05,
"loss": 0.6526,
"step": 655
},
{
"epoch": 6.0,
"eval_loss": 2.207833766937256,
"eval_runtime": 0.7173,
"eval_samples_per_second": 13.941,
"eval_steps_per_second": 1.394,
"step": 657
},
{
"epoch": 6.027397260273973,
"grad_norm": 0.2104310542345047,
"learning_rate": 8.07462320849313e-05,
"loss": 0.6444,
"step": 660
},
{
"epoch": 6.073059360730594,
"grad_norm": 0.18349581956863403,
"learning_rate": 7.917750740470117e-05,
"loss": 0.6397,
"step": 665
},
{
"epoch": 6.1187214611872145,
"grad_norm": 0.20204049348831177,
"learning_rate": 7.761412129735852e-05,
"loss": 0.6439,
"step": 670
},
{
"epoch": 6.164383561643835,
"grad_norm": 0.20316743850708008,
"learning_rate": 7.605647459150961e-05,
"loss": 0.6392,
"step": 675
},
{
"epoch": 6.210045662100456,
"grad_norm": 0.2150379866361618,
"learning_rate": 7.450496664426477e-05,
"loss": 0.634,
"step": 680
},
{
"epoch": 6.255707762557078,
"grad_norm": 0.19887958467006683,
"learning_rate": 7.295999523884921e-05,
"loss": 0.6442,
"step": 685
},
{
"epoch": 6.301369863013699,
"grad_norm": 0.1974460333585739,
"learning_rate": 7.142195648261747e-05,
"loss": 0.6396,
"step": 690
},
{
"epoch": 6.34703196347032,
"grad_norm": 0.21475858986377716,
"learning_rate": 6.989124470549745e-05,
"loss": 0.6354,
"step": 695
},
{
"epoch": 6.392694063926941,
"grad_norm": 0.19873230159282684,
"learning_rate": 6.83682523588902e-05,
"loss": 0.64,
"step": 700
},
{
"epoch": 6.438356164383562,
"grad_norm": 0.20658674836158752,
"learning_rate": 6.685336991505122e-05,
"loss": 0.6367,
"step": 705
},
{
"epoch": 6.4840182648401825,
"grad_norm": 0.2072938233613968,
"learning_rate": 6.534698576697939e-05,
"loss": 0.6388,
"step": 710
},
{
"epoch": 6.529680365296803,
"grad_norm": 0.20482133328914642,
"learning_rate": 6.384948612883873e-05,
"loss": 0.6397,
"step": 715
},
{
"epoch": 6.575342465753424,
"grad_norm": 0.20856335759162903,
"learning_rate": 6.2361254936939e-05,
"loss": 0.6407,
"step": 720
},
{
"epoch": 6.621004566210045,
"grad_norm": 0.21489538252353668,
"learning_rate": 6.088267375130023e-05,
"loss": 0.6414,
"step": 725
},
{
"epoch": 6.666666666666667,
"grad_norm": 0.19792255759239197,
"learning_rate": 5.941412165782645e-05,
"loss": 0.634,
"step": 730
},
{
"epoch": 6.712328767123288,
"grad_norm": 0.19678843021392822,
"learning_rate": 5.79559751711138e-05,
"loss": 0.6407,
"step": 735
},
{
"epoch": 6.757990867579909,
"grad_norm": 0.18998436629772186,
"learning_rate": 5.650860813791785e-05,
"loss": 0.6346,
"step": 740
},
{
"epoch": 6.80365296803653,
"grad_norm": 0.19847027957439423,
"learning_rate": 5.507239164130501e-05,
"loss": 0.6409,
"step": 745
},
{
"epoch": 6.8493150684931505,
"grad_norm": 0.205197274684906,
"learning_rate": 5.364769390551225e-05,
"loss": 0.6404,
"step": 750
},
{
"epoch": 6.894977168949771,
"grad_norm": 0.19521455466747284,
"learning_rate": 5.2234880201540284e-05,
"loss": 0.6386,
"step": 755
},
{
"epoch": 6.940639269406392,
"grad_norm": 0.1955854445695877,
"learning_rate": 5.0834312753503124e-05,
"loss": 0.6349,
"step": 760
},
{
"epoch": 6.986301369863014,
"grad_norm": 0.1964850276708603,
"learning_rate": 4.9446350645759885e-05,
"loss": 0.6326,
"step": 765
},
{
"epoch": 6.995433789954338,
"eval_loss": 2.229551315307617,
"eval_runtime": 0.7351,
"eval_samples_per_second": 13.604,
"eval_steps_per_second": 1.36,
"step": 766
},
{
"epoch": 7.031963470319635,
"grad_norm": 0.1950986385345459,
"learning_rate": 4.807134973085036e-05,
"loss": 0.626,
"step": 770
},
{
"epoch": 7.077625570776256,
"grad_norm": 0.20411163568496704,
"learning_rate": 4.6709662538260267e-05,
"loss": 0.6293,
"step": 775
},
{
"epoch": 7.123287671232877,
"grad_norm": 0.21672751009464264,
"learning_rate": 4.53616381840377e-05,
"loss": 0.6255,
"step": 780
},
{
"epoch": 7.168949771689498,
"grad_norm": 0.20474384725093842,
"learning_rate": 4.402762228128531e-05,
"loss": 0.6317,
"step": 785
},
{
"epoch": 7.2146118721461185,
"grad_norm": 0.20330122113227844,
"learning_rate": 4.2707956851550016e-05,
"loss": 0.6297,
"step": 790
},
{
"epoch": 7.260273972602739,
"grad_norm": 0.21286101639270782,
"learning_rate": 4.140298023713416e-05,
"loss": 0.6278,
"step": 795
},
{
"epoch": 7.30593607305936,
"grad_norm": 0.20241223275661469,
"learning_rate": 4.011302701434937e-05,
"loss": 0.6223,
"step": 800
},
{
"epoch": 7.351598173515982,
"grad_norm": 0.20820164680480957,
"learning_rate": 3.8838427907736476e-05,
"loss": 0.631,
"step": 805
},
{
"epoch": 7.397260273972603,
"grad_norm": 0.2071073353290558,
"learning_rate": 3.757950970527249e-05,
"loss": 0.627,
"step": 810
},
{
"epoch": 7.442922374429224,
"grad_norm": 0.20754040777683258,
"learning_rate": 3.633659517458736e-05,
"loss": 0.6284,
"step": 815
},
{
"epoch": 7.488584474885845,
"grad_norm": 0.204667329788208,
"learning_rate": 3.5110002980210975e-05,
"loss": 0.6214,
"step": 820
},
{
"epoch": 7.534246575342466,
"grad_norm": 0.204596146941185,
"learning_rate": 3.3900047601872596e-05,
"loss": 0.6241,
"step": 825
},
{
"epoch": 7.579908675799087,
"grad_norm": 0.2120332568883896,
"learning_rate": 3.270703925387279e-05,
"loss": 0.6295,
"step": 830
},
{
"epoch": 7.6255707762557075,
"grad_norm": 0.20006106793880463,
"learning_rate": 3.153128380554941e-05,
"loss": 0.6259,
"step": 835
},
{
"epoch": 7.671232876712329,
"grad_norm": 0.19608016312122345,
"learning_rate": 3.037308270285709e-05,
"loss": 0.6234,
"step": 840
},
{
"epoch": 7.71689497716895,
"grad_norm": 0.20698235929012299,
"learning_rate": 2.923273289108115e-05,
"loss": 0.6312,
"step": 845
},
{
"epoch": 7.762557077625571,
"grad_norm": 0.20038080215454102,
"learning_rate": 2.8110526738705344e-05,
"loss": 0.6266,
"step": 850
},
{
"epoch": 7.808219178082192,
"grad_norm": 0.2022303193807602,
"learning_rate": 2.7006751962452882e-05,
"loss": 0.631,
"step": 855
},
{
"epoch": 7.853881278538813,
"grad_norm": 0.20605237782001495,
"learning_rate": 2.592169155352031e-05,
"loss": 0.6331,
"step": 860
},
{
"epoch": 7.899543378995434,
"grad_norm": 0.2080409824848175,
"learning_rate": 2.485562370502279e-05,
"loss": 0.6313,
"step": 865
},
{
"epoch": 7.945205479452055,
"grad_norm": 0.2000366896390915,
"learning_rate": 2.3808821740669606e-05,
"loss": 0.6285,
"step": 870
},
{
"epoch": 7.9908675799086755,
"grad_norm": 0.20418736338615417,
"learning_rate": 2.2781554044688015e-05,
"loss": 0.6311,
"step": 875
},
{
"epoch": 8.0,
"eval_loss": 2.2484524250030518,
"eval_runtime": 0.7257,
"eval_samples_per_second": 13.779,
"eval_steps_per_second": 1.378,
"step": 876
},
{
"epoch": 8.036529680365296,
"grad_norm": 0.2047678381204605,
"learning_rate": 2.1774083993013718e-05,
"loss": 0.6168,
"step": 880
},
{
"epoch": 8.082191780821917,
"grad_norm": 0.20055288076400757,
"learning_rate": 2.078666988576504e-05,
"loss": 0.6163,
"step": 885
},
{
"epoch": 8.127853881278538,
"grad_norm": 0.2002391219139099,
"learning_rate": 1.9819564881018983e-05,
"loss": 0.6216,
"step": 890
},
{
"epoch": 8.173515981735159,
"grad_norm": 0.19824841618537903,
"learning_rate": 1.887301692990494e-05,
"loss": 0.6215,
"step": 895
},
{
"epoch": 8.219178082191782,
"grad_norm": 0.20471996068954468,
"learning_rate": 1.7947268713034127e-05,
"loss": 0.6168,
"step": 900
},
{
"epoch": 8.264840182648403,
"grad_norm": 0.20544229447841644,
"learning_rate": 1.7042557578279626e-05,
"loss": 0.6188,
"step": 905
},
{
"epoch": 8.310502283105023,
"grad_norm": 0.208919957280159,
"learning_rate": 1.6159115479924257e-05,
"loss": 0.6221,
"step": 910
},
{
"epoch": 8.356164383561644,
"grad_norm": 0.20433735847473145,
"learning_rate": 1.529716891919074e-05,
"loss": 0.6187,
"step": 915
},
{
"epoch": 8.401826484018265,
"grad_norm": 0.20233039557933807,
"learning_rate": 1.4456938886170412e-05,
"loss": 0.6224,
"step": 920
},
{
"epoch": 8.447488584474886,
"grad_norm": 0.2014380395412445,
"learning_rate": 1.3638640803164516e-05,
"loss": 0.6186,
"step": 925
},
{
"epoch": 8.493150684931507,
"grad_norm": 0.20106054842472076,
"learning_rate": 1.2842484469453365e-05,
"loss": 0.6229,
"step": 930
},
{
"epoch": 8.538812785388128,
"grad_norm": 0.20424200594425201,
"learning_rate": 1.2068674007506786e-05,
"loss": 0.6225,
"step": 935
},
{
"epoch": 8.584474885844749,
"grad_norm": 0.19919848442077637,
"learning_rate": 1.1317407810650372e-05,
"loss": 0.621,
"step": 940
},
{
"epoch": 8.63013698630137,
"grad_norm": 0.21022921800613403,
"learning_rate": 1.058887849220026e-05,
"loss": 0.6208,
"step": 945
},
{
"epoch": 8.67579908675799,
"grad_norm": 0.2036212980747223,
"learning_rate": 9.883272836080116e-06,
"loss": 0.6184,
"step": 950
},
{
"epoch": 8.721461187214611,
"grad_norm": 0.19858643412590027,
"learning_rate": 9.200771748932513e-06,
"loss": 0.6224,
"step": 955
},
{
"epoch": 8.767123287671232,
"grad_norm": 0.2013389766216278,
"learning_rate": 8.541550213737171e-06,
"loss": 0.6175,
"step": 960
},
{
"epoch": 8.812785388127853,
"grad_norm": 0.20145344734191895,
"learning_rate": 7.905777244947954e-06,
"loss": 0.6204,
"step": 965
},
{
"epoch": 8.858447488584474,
"grad_norm": 0.20189128816127777,
"learning_rate": 7.293615845160196e-06,
"loss": 0.6225,
"step": 970
},
{
"epoch": 8.904109589041095,
"grad_norm": 0.2002427577972412,
"learning_rate": 6.705222963319191e-06,
"loss": 0.6197,
"step": 975
},
{
"epoch": 8.949771689497716,
"grad_norm": 0.21808940172195435,
"learning_rate": 6.140749454480932e-06,
"loss": 0.6228,
"step": 980
},
{
"epoch": 8.995433789954339,
"grad_norm": 0.20092381536960602,
"learning_rate": 5.6003400411351325e-06,
"loss": 0.6233,
"step": 985
},
{
"epoch": 8.995433789954339,
"eval_loss": 2.258711099624634,
"eval_runtime": 0.7286,
"eval_samples_per_second": 13.726,
"eval_steps_per_second": 1.373,
"step": 985
},
{
"epoch": 9.04109589041096,
"grad_norm": 0.19878605008125305,
"learning_rate": 5.0841332761005e-06,
"loss": 0.6193,
"step": 990
},
{
"epoch": 9.08675799086758,
"grad_norm": 0.20306305587291718,
"learning_rate": 4.592261507001993e-06,
"loss": 0.6177,
"step": 995
},
{
"epoch": 9.132420091324201,
"grad_norm": 0.2019263654947281,
"learning_rate": 4.124850842338779e-06,
"loss": 0.6133,
"step": 1000
},
{
"epoch": 9.178082191780822,
"grad_norm": 0.2005142867565155,
"learning_rate": 3.6820211191520125e-06,
"loss": 0.6141,
"step": 1005
},
{
"epoch": 9.223744292237443,
"grad_norm": 0.20048119127750397,
"learning_rate": 3.263885872300343e-06,
"loss": 0.6224,
"step": 1010
},
{
"epoch": 9.269406392694064,
"grad_norm": 0.20118780434131622,
"learning_rate": 2.8705523053513816e-06,
"loss": 0.6158,
"step": 1015
},
{
"epoch": 9.315068493150685,
"grad_norm": 0.19994989037513733,
"learning_rate": 2.502121263096224e-06,
"loss": 0.6133,
"step": 1020
},
{
"epoch": 9.360730593607306,
"grad_norm": 0.19647841155529022,
"learning_rate": 2.1586872056944428e-06,
"loss": 0.6172,
"step": 1025
},
{
"epoch": 9.406392694063927,
"grad_norm": 0.20165683329105377,
"learning_rate": 1.840338184455881e-06,
"loss": 0.6193,
"step": 1030
},
{
"epoch": 9.452054794520548,
"grad_norm": 0.20188592374324799,
"learning_rate": 1.5471558192656777e-06,
"loss": 0.6216,
"step": 1035
},
{
"epoch": 9.497716894977168,
"grad_norm": 0.20157510042190552,
"learning_rate": 1.2792152776580968e-06,
"loss": 0.616,
"step": 1040
},
{
"epoch": 9.54337899543379,
"grad_norm": 0.20351751148700714,
"learning_rate": 1.036585255544764e-06,
"loss": 0.6149,
"step": 1045
},
{
"epoch": 9.58904109589041,
"grad_norm": 0.1994985044002533,
"learning_rate": 8.193279596020121e-07,
"loss": 0.6184,
"step": 1050
},
{
"epoch": 9.634703196347033,
"grad_norm": 0.2011667788028717,
"learning_rate": 6.274990913221035e-07,
"loss": 0.6149,
"step": 1055
},
{
"epoch": 9.680365296803654,
"grad_norm": 0.2020387500524521,
"learning_rate": 4.6114783273213393e-07,
"loss": 0.619,
"step": 1060
},
{
"epoch": 9.726027397260275,
"grad_norm": 0.19661925733089447,
"learning_rate": 3.203168337845508e-07,
"loss": 0.6193,
"step": 1065
},
{
"epoch": 9.771689497716896,
"grad_norm": 0.20111018419265747,
"learning_rate": 2.05042201422323e-07,
"loss": 0.615,
"step": 1070
},
{
"epoch": 9.817351598173516,
"grad_norm": 0.20135089755058289,
"learning_rate": 1.1535349032167908e-07,
"loss": 0.6158,
"step": 1075
},
{
"epoch": 9.863013698630137,
"grad_norm": 0.20196588337421417,
"learning_rate": 5.127369531473525e-08,
"loss": 0.614,
"step": 1080
},
{
"epoch": 9.908675799086758,
"grad_norm": 0.19875915348529816,
"learning_rate": 1.2819245493955744e-08,
"loss": 0.6174,
"step": 1085
},
{
"epoch": 9.954337899543379,
"grad_norm": 0.19983670115470886,
"learning_rate": 0.0,
"loss": 0.6194,
"step": 1090
},
{
"epoch": 9.954337899543379,
"eval_loss": 2.260620594024658,
"eval_runtime": 0.6934,
"eval_samples_per_second": 14.422,
"eval_steps_per_second": 1.442,
"step": 1090
},
{
"epoch": 9.954337899543379,
"step": 1090,
"total_flos": 6.456679991336763e+18,
"train_loss": 0.7130365929472338,
"train_runtime": 9668.9725,
"train_samples_per_second": 14.452,
"train_steps_per_second": 0.113
}
],
"logging_steps": 5,
"max_steps": 1090,
"num_input_tokens_seen": 0,
"num_train_epochs": 10,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.456679991336763e+18,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}