1192 lines
27 KiB
JSON
1192 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 4.0,
|
|
"eval_steps": 500,
|
|
"global_step": 1648,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.024301336573511544,
|
|
"grad_norm": 7.96875,
|
|
"learning_rate": 9.945388349514565e-06,
|
|
"loss": 1.9057,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.04860267314702309,
|
|
"grad_norm": 6.875,
|
|
"learning_rate": 9.884708737864079e-06,
|
|
"loss": 1.106,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.07290400972053462,
|
|
"grad_norm": 7.0625,
|
|
"learning_rate": 9.824029126213593e-06,
|
|
"loss": 1.0902,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.09720534629404617,
|
|
"grad_norm": 7.15625,
|
|
"learning_rate": 9.763349514563108e-06,
|
|
"loss": 1.0498,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.12150668286755771,
|
|
"grad_norm": 8.1875,
|
|
"learning_rate": 9.702669902912622e-06,
|
|
"loss": 1.0074,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.14580801944106925,
|
|
"grad_norm": 8.125,
|
|
"learning_rate": 9.641990291262136e-06,
|
|
"loss": 0.9637,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.1701093560145808,
|
|
"grad_norm": 11.5,
|
|
"learning_rate": 9.581310679611652e-06,
|
|
"loss": 0.9106,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.19441069258809235,
|
|
"grad_norm": 12.1875,
|
|
"learning_rate": 9.520631067961167e-06,
|
|
"loss": 0.8787,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.2187120291616039,
|
|
"grad_norm": 10.6875,
|
|
"learning_rate": 9.45995145631068e-06,
|
|
"loss": 0.8187,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.24301336573511542,
|
|
"grad_norm": 9.6875,
|
|
"learning_rate": 9.399271844660195e-06,
|
|
"loss": 0.789,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.267314702308627,
|
|
"grad_norm": 6.59375,
|
|
"learning_rate": 9.33859223300971e-06,
|
|
"loss": 0.7243,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.2916160388821385,
|
|
"grad_norm": 7.21875,
|
|
"learning_rate": 9.277912621359224e-06,
|
|
"loss": 0.7021,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.3159173754556501,
|
|
"grad_norm": 6.125,
|
|
"learning_rate": 9.217233009708738e-06,
|
|
"loss": 0.6839,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.3402187120291616,
|
|
"grad_norm": 6.71875,
|
|
"learning_rate": 9.156553398058252e-06,
|
|
"loss": 0.657,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.3645200486026732,
|
|
"grad_norm": 6.0625,
|
|
"learning_rate": 9.095873786407767e-06,
|
|
"loss": 0.6801,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.3888213851761847,
|
|
"grad_norm": 6.75,
|
|
"learning_rate": 9.035194174757283e-06,
|
|
"loss": 0.6609,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.4131227217496962,
|
|
"grad_norm": 6.46875,
|
|
"learning_rate": 8.974514563106797e-06,
|
|
"loss": 0.6557,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.4374240583232078,
|
|
"grad_norm": 7.09375,
|
|
"learning_rate": 8.913834951456311e-06,
|
|
"loss": 0.6482,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.4617253948967193,
|
|
"grad_norm": 7.03125,
|
|
"learning_rate": 8.853155339805826e-06,
|
|
"loss": 0.6302,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.48602673147023084,
|
|
"grad_norm": 5.9375,
|
|
"learning_rate": 8.79247572815534e-06,
|
|
"loss": 0.6244,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.5103280680437424,
|
|
"grad_norm": 6.5625,
|
|
"learning_rate": 8.731796116504854e-06,
|
|
"loss": 0.6236,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.534629404617254,
|
|
"grad_norm": 5.6875,
|
|
"learning_rate": 8.67111650485437e-06,
|
|
"loss": 0.6089,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.5589307411907655,
|
|
"grad_norm": 5.84375,
|
|
"learning_rate": 8.610436893203885e-06,
|
|
"loss": 0.6201,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.583232077764277,
|
|
"grad_norm": 7.0,
|
|
"learning_rate": 8.549757281553399e-06,
|
|
"loss": 0.5984,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.6075334143377886,
|
|
"grad_norm": 5.4375,
|
|
"learning_rate": 8.489077669902913e-06,
|
|
"loss": 0.6151,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.6318347509113001,
|
|
"grad_norm": 6.34375,
|
|
"learning_rate": 8.428398058252428e-06,
|
|
"loss": 0.6182,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.6561360874848117,
|
|
"grad_norm": 5.9375,
|
|
"learning_rate": 8.367718446601942e-06,
|
|
"loss": 0.5751,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.6804374240583232,
|
|
"grad_norm": 5.375,
|
|
"learning_rate": 8.307038834951458e-06,
|
|
"loss": 0.5789,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.7047387606318347,
|
|
"grad_norm": 6.1875,
|
|
"learning_rate": 8.246359223300972e-06,
|
|
"loss": 0.5861,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 0.7290400972053463,
|
|
"grad_norm": 6.21875,
|
|
"learning_rate": 8.185679611650487e-06,
|
|
"loss": 0.6002,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.7533414337788579,
|
|
"grad_norm": 6.21875,
|
|
"learning_rate": 8.125000000000001e-06,
|
|
"loss": 0.5614,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 0.7776427703523694,
|
|
"grad_norm": 5.5,
|
|
"learning_rate": 8.064320388349515e-06,
|
|
"loss": 0.6124,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.8019441069258809,
|
|
"grad_norm": 6.09375,
|
|
"learning_rate": 8.00364077669903e-06,
|
|
"loss": 0.5762,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 0.8262454434993924,
|
|
"grad_norm": 5.65625,
|
|
"learning_rate": 7.942961165048544e-06,
|
|
"loss": 0.558,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.850546780072904,
|
|
"grad_norm": 6.3125,
|
|
"learning_rate": 7.882281553398058e-06,
|
|
"loss": 0.5925,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.8748481166464156,
|
|
"grad_norm": 5.71875,
|
|
"learning_rate": 7.821601941747573e-06,
|
|
"loss": 0.5542,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.8991494532199271,
|
|
"grad_norm": 5.28125,
|
|
"learning_rate": 7.760922330097089e-06,
|
|
"loss": 0.5518,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 0.9234507897934386,
|
|
"grad_norm": 6.84375,
|
|
"learning_rate": 7.700242718446603e-06,
|
|
"loss": 0.5648,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.9477521263669502,
|
|
"grad_norm": 6.5625,
|
|
"learning_rate": 7.639563106796117e-06,
|
|
"loss": 0.5779,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 0.9720534629404617,
|
|
"grad_norm": 6.71875,
|
|
"learning_rate": 7.5788834951456316e-06,
|
|
"loss": 0.532,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.9963547995139733,
|
|
"grad_norm": 6.65625,
|
|
"learning_rate": 7.518203883495147e-06,
|
|
"loss": 0.5498,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 1.0194410692588092,
|
|
"grad_norm": 6.0625,
|
|
"learning_rate": 7.45752427184466e-06,
|
|
"loss": 0.5358,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 1.0437424058323208,
|
|
"grad_norm": 6.125,
|
|
"learning_rate": 7.396844660194175e-06,
|
|
"loss": 0.5068,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 1.0680437424058322,
|
|
"grad_norm": 7.5,
|
|
"learning_rate": 7.33616504854369e-06,
|
|
"loss": 0.5367,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 1.0923450789793439,
|
|
"grad_norm": 6.46875,
|
|
"learning_rate": 7.275485436893205e-06,
|
|
"loss": 0.5168,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 1.1166464155528555,
|
|
"grad_norm": 6.34375,
|
|
"learning_rate": 7.214805825242719e-06,
|
|
"loss": 0.5349,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 1.140947752126367,
|
|
"grad_norm": 6.34375,
|
|
"learning_rate": 7.154126213592234e-06,
|
|
"loss": 0.5245,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 1.1652490886998785,
|
|
"grad_norm": 6.15625,
|
|
"learning_rate": 7.093446601941748e-06,
|
|
"loss": 0.5297,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 1.18955042527339,
|
|
"grad_norm": 6.59375,
|
|
"learning_rate": 7.032766990291263e-06,
|
|
"loss": 0.5229,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 1.2138517618469016,
|
|
"grad_norm": 6.53125,
|
|
"learning_rate": 6.972087378640777e-06,
|
|
"loss": 0.5061,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 1.2381530984204132,
|
|
"grad_norm": 6.8125,
|
|
"learning_rate": 6.9114077669902925e-06,
|
|
"loss": 0.4853,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 1.2624544349939246,
|
|
"grad_norm": 6.96875,
|
|
"learning_rate": 6.850728155339806e-06,
|
|
"loss": 0.5192,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 1.2867557715674363,
|
|
"grad_norm": 7.40625,
|
|
"learning_rate": 6.79004854368932e-06,
|
|
"loss": 0.5169,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 1.3110571081409477,
|
|
"grad_norm": 6.40625,
|
|
"learning_rate": 6.7293689320388355e-06,
|
|
"loss": 0.5183,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 1.3353584447144593,
|
|
"grad_norm": 6.1875,
|
|
"learning_rate": 6.66868932038835e-06,
|
|
"loss": 0.5067,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 1.359659781287971,
|
|
"grad_norm": 7.09375,
|
|
"learning_rate": 6.608009708737865e-06,
|
|
"loss": 0.5228,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 1.3839611178614823,
|
|
"grad_norm": 6.75,
|
|
"learning_rate": 6.5473300970873784e-06,
|
|
"loss": 0.5169,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 1.408262454434994,
|
|
"grad_norm": 6.09375,
|
|
"learning_rate": 6.486650485436894e-06,
|
|
"loss": 0.5042,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 1.4325637910085054,
|
|
"grad_norm": 7.09375,
|
|
"learning_rate": 6.425970873786408e-06,
|
|
"loss": 0.5032,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 1.456865127582017,
|
|
"grad_norm": 6.625,
|
|
"learning_rate": 6.365291262135923e-06,
|
|
"loss": 0.5046,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 1.4811664641555287,
|
|
"grad_norm": 6.53125,
|
|
"learning_rate": 6.304611650485437e-06,
|
|
"loss": 0.5237,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 1.50546780072904,
|
|
"grad_norm": 6.03125,
|
|
"learning_rate": 6.243932038834953e-06,
|
|
"loss": 0.5108,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 1.5297691373025515,
|
|
"grad_norm": 5.90625,
|
|
"learning_rate": 6.183252427184466e-06,
|
|
"loss": 0.486,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 1.5540704738760631,
|
|
"grad_norm": 6.46875,
|
|
"learning_rate": 6.122572815533981e-06,
|
|
"loss": 0.513,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 1.5783718104495748,
|
|
"grad_norm": 6.375,
|
|
"learning_rate": 6.0618932038834955e-06,
|
|
"loss": 0.4996,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 1.6026731470230864,
|
|
"grad_norm": 5.96875,
|
|
"learning_rate": 6.001213592233011e-06,
|
|
"loss": 0.4988,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 1.6269744835965978,
|
|
"grad_norm": 7.46875,
|
|
"learning_rate": 5.940533980582525e-06,
|
|
"loss": 0.5018,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 1.6512758201701092,
|
|
"grad_norm": 7.25,
|
|
"learning_rate": 5.879854368932039e-06,
|
|
"loss": 0.498,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 1.6755771567436208,
|
|
"grad_norm": 6.5,
|
|
"learning_rate": 5.819174757281554e-06,
|
|
"loss": 0.493,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 1.6998784933171325,
|
|
"grad_norm": 7.0,
|
|
"learning_rate": 5.758495145631069e-06,
|
|
"loss": 0.5027,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 1.7241798298906441,
|
|
"grad_norm": 6.5,
|
|
"learning_rate": 5.697815533980583e-06,
|
|
"loss": 0.478,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 1.7484811664641555,
|
|
"grad_norm": 6.28125,
|
|
"learning_rate": 5.637135922330098e-06,
|
|
"loss": 0.5045,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 1.772782503037667,
|
|
"grad_norm": 6.25,
|
|
"learning_rate": 5.576456310679612e-06,
|
|
"loss": 0.4913,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 1.7970838396111786,
|
|
"grad_norm": 6.90625,
|
|
"learning_rate": 5.515776699029126e-06,
|
|
"loss": 0.5046,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 1.8213851761846902,
|
|
"grad_norm": 6.59375,
|
|
"learning_rate": 5.455097087378641e-06,
|
|
"loss": 0.5207,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 1.8456865127582018,
|
|
"grad_norm": 7.8125,
|
|
"learning_rate": 5.394417475728156e-06,
|
|
"loss": 0.5014,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 1.8699878493317132,
|
|
"grad_norm": 6.375,
|
|
"learning_rate": 5.333737864077671e-06,
|
|
"loss": 0.5051,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 1.8942891859052247,
|
|
"grad_norm": 6.34375,
|
|
"learning_rate": 5.273058252427184e-06,
|
|
"loss": 0.5025,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 1.9185905224787363,
|
|
"grad_norm": 6.96875,
|
|
"learning_rate": 5.2123786407766994e-06,
|
|
"loss": 0.4872,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 1.942891859052248,
|
|
"grad_norm": 6.65625,
|
|
"learning_rate": 5.151699029126214e-06,
|
|
"loss": 0.5187,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 1.9671931956257596,
|
|
"grad_norm": 6.46875,
|
|
"learning_rate": 5.091019417475729e-06,
|
|
"loss": 0.5003,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 1.991494532199271,
|
|
"grad_norm": 6.125,
|
|
"learning_rate": 5.030339805825243e-06,
|
|
"loss": 0.4733,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 2.0145808019441067,
|
|
"grad_norm": 6.65625,
|
|
"learning_rate": 4.9696601941747576e-06,
|
|
"loss": 0.4885,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 2.0388821385176183,
|
|
"grad_norm": 6.96875,
|
|
"learning_rate": 4.908980582524272e-06,
|
|
"loss": 0.4626,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 2.06318347509113,
|
|
"grad_norm": 6.96875,
|
|
"learning_rate": 4.848300970873786e-06,
|
|
"loss": 0.4682,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 2.0874848116646416,
|
|
"grad_norm": 6.96875,
|
|
"learning_rate": 4.787621359223301e-06,
|
|
"loss": 0.4632,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 2.1117861482381532,
|
|
"grad_norm": 7.5,
|
|
"learning_rate": 4.726941747572816e-06,
|
|
"loss": 0.4871,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 2.1360874848116644,
|
|
"grad_norm": 6.9375,
|
|
"learning_rate": 4.66626213592233e-06,
|
|
"loss": 0.4764,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 2.160388821385176,
|
|
"grad_norm": 6.0625,
|
|
"learning_rate": 4.605582524271845e-06,
|
|
"loss": 0.4664,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 2.1846901579586877,
|
|
"grad_norm": 7.4375,
|
|
"learning_rate": 4.5449029126213595e-06,
|
|
"loss": 0.4913,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 2.2089914945321993,
|
|
"grad_norm": 6.15625,
|
|
"learning_rate": 4.484223300970874e-06,
|
|
"loss": 0.4768,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 2.233292831105711,
|
|
"grad_norm": 6.71875,
|
|
"learning_rate": 4.423543689320389e-06,
|
|
"loss": 0.4615,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 2.257594167679222,
|
|
"grad_norm": 6.53125,
|
|
"learning_rate": 4.362864077669903e-06,
|
|
"loss": 0.4681,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 2.281895504252734,
|
|
"grad_norm": 6.21875,
|
|
"learning_rate": 4.302184466019418e-06,
|
|
"loss": 0.4585,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 2.3061968408262454,
|
|
"grad_norm": 7.8125,
|
|
"learning_rate": 4.241504854368933e-06,
|
|
"loss": 0.467,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 2.330498177399757,
|
|
"grad_norm": 8.125,
|
|
"learning_rate": 4.180825242718447e-06,
|
|
"loss": 0.4896,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 2.3547995139732687,
|
|
"grad_norm": 7.5,
|
|
"learning_rate": 4.1201456310679615e-06,
|
|
"loss": 0.4739,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 2.37910085054678,
|
|
"grad_norm": 7.09375,
|
|
"learning_rate": 4.059466019417476e-06,
|
|
"loss": 0.4703,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 2.4034021871202915,
|
|
"grad_norm": 6.15625,
|
|
"learning_rate": 3.998786407766991e-06,
|
|
"loss": 0.4713,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 2.427703523693803,
|
|
"grad_norm": 6.8125,
|
|
"learning_rate": 3.938106796116505e-06,
|
|
"loss": 0.4826,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 2.452004860267315,
|
|
"grad_norm": 6.09375,
|
|
"learning_rate": 3.87742718446602e-06,
|
|
"loss": 0.4745,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 2.4763061968408264,
|
|
"grad_norm": 6.375,
|
|
"learning_rate": 3.816747572815535e-06,
|
|
"loss": 0.4424,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 2.500607533414338,
|
|
"grad_norm": 7.375,
|
|
"learning_rate": 3.756067961165049e-06,
|
|
"loss": 0.4921,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 2.5249088699878492,
|
|
"grad_norm": 7.46875,
|
|
"learning_rate": 3.695388349514563e-06,
|
|
"loss": 0.4617,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 2.549210206561361,
|
|
"grad_norm": 6.6875,
|
|
"learning_rate": 3.6347087378640777e-06,
|
|
"loss": 0.4811,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 2.5735115431348725,
|
|
"grad_norm": 7.9375,
|
|
"learning_rate": 3.5740291262135925e-06,
|
|
"loss": 0.484,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 2.5978128797083837,
|
|
"grad_norm": 9.1875,
|
|
"learning_rate": 3.513349514563107e-06,
|
|
"loss": 0.4753,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 2.6221142162818953,
|
|
"grad_norm": 7.4375,
|
|
"learning_rate": 3.4526699029126215e-06,
|
|
"loss": 0.4641,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 2.646415552855407,
|
|
"grad_norm": 6.75,
|
|
"learning_rate": 3.3919902912621363e-06,
|
|
"loss": 0.4693,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 2.6707168894289186,
|
|
"grad_norm": 7.53125,
|
|
"learning_rate": 3.3313106796116506e-06,
|
|
"loss": 0.4641,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 2.6950182260024302,
|
|
"grad_norm": 7.09375,
|
|
"learning_rate": 3.2706310679611654e-06,
|
|
"loss": 0.4789,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 2.719319562575942,
|
|
"grad_norm": 6.5625,
|
|
"learning_rate": 3.2099514563106797e-06,
|
|
"loss": 0.487,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 2.743620899149453,
|
|
"grad_norm": 7.65625,
|
|
"learning_rate": 3.1492718446601944e-06,
|
|
"loss": 0.4825,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 2.7679222357229647,
|
|
"grad_norm": 7.28125,
|
|
"learning_rate": 3.088592233009709e-06,
|
|
"loss": 0.4761,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 2.7922235722964763,
|
|
"grad_norm": 8.0625,
|
|
"learning_rate": 3.0279126213592235e-06,
|
|
"loss": 0.4768,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 2.816524908869988,
|
|
"grad_norm": 6.90625,
|
|
"learning_rate": 2.9672330097087382e-06,
|
|
"loss": 0.468,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 2.840826245443499,
|
|
"grad_norm": 7.0,
|
|
"learning_rate": 2.9065533980582526e-06,
|
|
"loss": 0.4767,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 2.865127582017011,
|
|
"grad_norm": 6.09375,
|
|
"learning_rate": 2.8458737864077673e-06,
|
|
"loss": 0.4723,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 2.8894289185905224,
|
|
"grad_norm": 6.3125,
|
|
"learning_rate": 2.785194174757282e-06,
|
|
"loss": 0.4542,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 2.913730255164034,
|
|
"grad_norm": 7.28125,
|
|
"learning_rate": 2.7245145631067964e-06,
|
|
"loss": 0.4746,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 2.9380315917375457,
|
|
"grad_norm": 6.625,
|
|
"learning_rate": 2.663834951456311e-06,
|
|
"loss": 0.4442,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 2.9623329283110573,
|
|
"grad_norm": 8.0625,
|
|
"learning_rate": 2.603155339805826e-06,
|
|
"loss": 0.4723,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 2.9866342648845685,
|
|
"grad_norm": 5.96875,
|
|
"learning_rate": 2.54247572815534e-06,
|
|
"loss": 0.4631,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 3.0097205346294045,
|
|
"grad_norm": 7.34375,
|
|
"learning_rate": 2.4817961165048545e-06,
|
|
"loss": 0.4666,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 3.034021871202916,
|
|
"grad_norm": 6.8125,
|
|
"learning_rate": 2.4211165048543692e-06,
|
|
"loss": 0.4572,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 3.0583232077764277,
|
|
"grad_norm": 7.09375,
|
|
"learning_rate": 2.3604368932038836e-06,
|
|
"loss": 0.4494,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 3.0826245443499394,
|
|
"grad_norm": 6.9375,
|
|
"learning_rate": 2.2997572815533983e-06,
|
|
"loss": 0.4711,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 3.106925880923451,
|
|
"grad_norm": 6.8125,
|
|
"learning_rate": 2.239077669902913e-06,
|
|
"loss": 0.4466,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 3.131227217496962,
|
|
"grad_norm": 6.8125,
|
|
"learning_rate": 2.1783980582524274e-06,
|
|
"loss": 0.4525,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 3.155528554070474,
|
|
"grad_norm": 6.71875,
|
|
"learning_rate": 2.1177184466019417e-06,
|
|
"loss": 0.4539,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 3.1798298906439855,
|
|
"grad_norm": 6.71875,
|
|
"learning_rate": 2.0570388349514564e-06,
|
|
"loss": 0.4579,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 3.204131227217497,
|
|
"grad_norm": 7.46875,
|
|
"learning_rate": 1.996359223300971e-06,
|
|
"loss": 0.465,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 3.2284325637910083,
|
|
"grad_norm": 6.96875,
|
|
"learning_rate": 1.9356796116504855e-06,
|
|
"loss": 0.4608,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 3.25273390036452,
|
|
"grad_norm": 7.3125,
|
|
"learning_rate": 1.8750000000000003e-06,
|
|
"loss": 0.4659,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 3.2770352369380316,
|
|
"grad_norm": 6.5,
|
|
"learning_rate": 1.8143203883495148e-06,
|
|
"loss": 0.4545,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 3.301336573511543,
|
|
"grad_norm": 6.34375,
|
|
"learning_rate": 1.7536407766990293e-06,
|
|
"loss": 0.4623,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 3.325637910085055,
|
|
"grad_norm": 6.59375,
|
|
"learning_rate": 1.6929611650485439e-06,
|
|
"loss": 0.4733,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 3.3499392466585665,
|
|
"grad_norm": 7.1875,
|
|
"learning_rate": 1.6322815533980584e-06,
|
|
"loss": 0.4384,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 3.3742405832320777,
|
|
"grad_norm": 6.90625,
|
|
"learning_rate": 1.5716019417475731e-06,
|
|
"loss": 0.4747,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 3.3985419198055893,
|
|
"grad_norm": 7.375,
|
|
"learning_rate": 1.5109223300970875e-06,
|
|
"loss": 0.4614,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 3.422843256379101,
|
|
"grad_norm": 7.0625,
|
|
"learning_rate": 1.450242718446602e-06,
|
|
"loss": 0.4576,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"epoch": 3.4471445929526126,
|
|
"grad_norm": 7.625,
|
|
"learning_rate": 1.3895631067961165e-06,
|
|
"loss": 0.4652,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 3.4714459295261237,
|
|
"grad_norm": 6.75,
|
|
"learning_rate": 1.328883495145631e-06,
|
|
"loss": 0.4437,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"epoch": 3.4957472660996354,
|
|
"grad_norm": 7.90625,
|
|
"learning_rate": 1.2682038834951458e-06,
|
|
"loss": 0.4655,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 3.520048602673147,
|
|
"grad_norm": 7.03125,
|
|
"learning_rate": 1.2075242718446603e-06,
|
|
"loss": 0.4702,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 3.5443499392466586,
|
|
"grad_norm": 8.125,
|
|
"learning_rate": 1.1468446601941749e-06,
|
|
"loss": 0.4731,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 3.5686512758201703,
|
|
"grad_norm": 7.71875,
|
|
"learning_rate": 1.0861650485436894e-06,
|
|
"loss": 0.4593,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"epoch": 3.592952612393682,
|
|
"grad_norm": 7.5,
|
|
"learning_rate": 1.025485436893204e-06,
|
|
"loss": 0.4865,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 3.617253948967193,
|
|
"grad_norm": 6.375,
|
|
"learning_rate": 9.648058252427185e-07,
|
|
"loss": 0.4573,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"epoch": 3.6415552855407047,
|
|
"grad_norm": 6.59375,
|
|
"learning_rate": 9.041262135922331e-07,
|
|
"loss": 0.4581,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 3.6658566221142164,
|
|
"grad_norm": 6.46875,
|
|
"learning_rate": 8.434466019417476e-07,
|
|
"loss": 0.4351,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"epoch": 3.6901579586877276,
|
|
"grad_norm": 7.28125,
|
|
"learning_rate": 7.827669902912623e-07,
|
|
"loss": 0.4787,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 3.714459295261239,
|
|
"grad_norm": 7.1875,
|
|
"learning_rate": 7.220873786407767e-07,
|
|
"loss": 0.4692,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"epoch": 3.738760631834751,
|
|
"grad_norm": 6.5625,
|
|
"learning_rate": 6.614077669902914e-07,
|
|
"loss": 0.4702,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 3.7630619684082625,
|
|
"grad_norm": 7.21875,
|
|
"learning_rate": 6.007281553398059e-07,
|
|
"loss": 0.4626,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 3.787363304981774,
|
|
"grad_norm": 6.90625,
|
|
"learning_rate": 5.400485436893204e-07,
|
|
"loss": 0.4487,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 3.8116646415552857,
|
|
"grad_norm": 7.125,
|
|
"learning_rate": 4.79368932038835e-07,
|
|
"loss": 0.4539,
|
|
"step": 1570
|
|
},
|
|
{
|
|
"epoch": 3.8359659781287974,
|
|
"grad_norm": 6.84375,
|
|
"learning_rate": 4.186893203883496e-07,
|
|
"loss": 0.4538,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"epoch": 3.8602673147023085,
|
|
"grad_norm": 7.875,
|
|
"learning_rate": 3.5800970873786407e-07,
|
|
"loss": 0.4669,
|
|
"step": 1590
|
|
},
|
|
{
|
|
"epoch": 3.88456865127582,
|
|
"grad_norm": 7.34375,
|
|
"learning_rate": 2.9733009708737866e-07,
|
|
"loss": 0.4802,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 3.908869987849332,
|
|
"grad_norm": 6.71875,
|
|
"learning_rate": 2.3665048543689322e-07,
|
|
"loss": 0.4445,
|
|
"step": 1610
|
|
},
|
|
{
|
|
"epoch": 3.933171324422843,
|
|
"grad_norm": 6.53125,
|
|
"learning_rate": 1.7597087378640778e-07,
|
|
"loss": 0.4476,
|
|
"step": 1620
|
|
},
|
|
{
|
|
"epoch": 3.9574726609963546,
|
|
"grad_norm": 6.1875,
|
|
"learning_rate": 1.1529126213592234e-07,
|
|
"loss": 0.4614,
|
|
"step": 1630
|
|
},
|
|
{
|
|
"epoch": 3.9817739975698663,
|
|
"grad_norm": 7.53125,
|
|
"learning_rate": 5.46116504854369e-08,
|
|
"loss": 0.457,
|
|
"step": 1640
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"step": 1648,
|
|
"total_flos": 7.87315998893015e+16,
|
|
"train_loss": 0.5393084028681505,
|
|
"train_runtime": 981.8974,
|
|
"train_samples_per_second": 26.834,
|
|
"train_steps_per_second": 1.678
|
|
}
|
|
],
|
|
"logging_steps": 10,
|
|
"max_steps": 1648,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 4,
|
|
"save_steps": 411,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 7.87315998893015e+16,
|
|
"train_batch_size": 8,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|