1985 lines
48 KiB
JSON
1985 lines
48 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 2.0,
|
|
"eval_steps": 1000,
|
|
"global_step": 2750,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.007272727272727273,
|
|
"grad_norm": 24.832788467407227,
|
|
"learning_rate": 1.0843373493975905e-06,
|
|
"loss": 1.4729,
|
|
"step": 10
|
|
},
|
|
{
|
|
"epoch": 0.014545454545454545,
|
|
"grad_norm": 3.724433422088623,
|
|
"learning_rate": 2.2891566265060243e-06,
|
|
"loss": 1.314,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.02181818181818182,
|
|
"grad_norm": 1.6110222339630127,
|
|
"learning_rate": 3.4939759036144583e-06,
|
|
"loss": 1.2551,
|
|
"step": 30
|
|
},
|
|
{
|
|
"epoch": 0.02909090909090909,
|
|
"grad_norm": 0.879056990146637,
|
|
"learning_rate": 4.698795180722892e-06,
|
|
"loss": 1.1298,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.03636363636363636,
|
|
"grad_norm": 0.8674604892730713,
|
|
"learning_rate": 5.9036144578313255e-06,
|
|
"loss": 1.0667,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.04363636363636364,
|
|
"grad_norm": 0.8033781051635742,
|
|
"learning_rate": 7.1084337349397595e-06,
|
|
"loss": 1.0391,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.05090909090909091,
|
|
"grad_norm": 0.8189529180526733,
|
|
"learning_rate": 8.313253012048194e-06,
|
|
"loss": 1.0478,
|
|
"step": 70
|
|
},
|
|
{
|
|
"epoch": 0.05818181818181818,
|
|
"grad_norm": 0.8405522704124451,
|
|
"learning_rate": 9.518072289156628e-06,
|
|
"loss": 1.0311,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.06545454545454546,
|
|
"grad_norm": 0.887751579284668,
|
|
"learning_rate": 9.977502812148483e-06,
|
|
"loss": 1.0091,
|
|
"step": 90
|
|
},
|
|
{
|
|
"epoch": 0.07272727272727272,
|
|
"grad_norm": 0.753106415271759,
|
|
"learning_rate": 9.940007499062618e-06,
|
|
"loss": 0.9977,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.08,
|
|
"grad_norm": 0.7858726978302002,
|
|
"learning_rate": 9.902512185976754e-06,
|
|
"loss": 0.9744,
|
|
"step": 110
|
|
},
|
|
{
|
|
"epoch": 0.08727272727272728,
|
|
"grad_norm": 0.6802657842636108,
|
|
"learning_rate": 9.86501687289089e-06,
|
|
"loss": 0.9672,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.09454545454545454,
|
|
"grad_norm": 0.655599057674408,
|
|
"learning_rate": 9.827521559805026e-06,
|
|
"loss": 0.961,
|
|
"step": 130
|
|
},
|
|
{
|
|
"epoch": 0.10181818181818182,
|
|
"grad_norm": 0.6806121468544006,
|
|
"learning_rate": 9.790026246719161e-06,
|
|
"loss": 0.9708,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.10909090909090909,
|
|
"grad_norm": 0.6039880514144897,
|
|
"learning_rate": 9.752530933633297e-06,
|
|
"loss": 0.9565,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.11636363636363636,
|
|
"grad_norm": 0.6666083931922913,
|
|
"learning_rate": 9.715035620547433e-06,
|
|
"loss": 0.944,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.12363636363636364,
|
|
"grad_norm": 0.6580015420913696,
|
|
"learning_rate": 9.677540307461568e-06,
|
|
"loss": 0.9479,
|
|
"step": 170
|
|
},
|
|
{
|
|
"epoch": 0.13090909090909092,
|
|
"grad_norm": 0.6064857244491577,
|
|
"learning_rate": 9.640044994375704e-06,
|
|
"loss": 0.9489,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.13818181818181818,
|
|
"grad_norm": 0.6560870409011841,
|
|
"learning_rate": 9.60254968128984e-06,
|
|
"loss": 0.9319,
|
|
"step": 190
|
|
},
|
|
{
|
|
"epoch": 0.14545454545454545,
|
|
"grad_norm": 0.6755111813545227,
|
|
"learning_rate": 9.565054368203975e-06,
|
|
"loss": 0.9631,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.15272727272727274,
|
|
"grad_norm": 0.6282963156700134,
|
|
"learning_rate": 9.52755905511811e-06,
|
|
"loss": 0.9085,
|
|
"step": 210
|
|
},
|
|
{
|
|
"epoch": 0.16,
|
|
"grad_norm": 0.6093109250068665,
|
|
"learning_rate": 9.490063742032247e-06,
|
|
"loss": 0.9323,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.16727272727272727,
|
|
"grad_norm": 0.6879354119300842,
|
|
"learning_rate": 9.452568428946382e-06,
|
|
"loss": 0.916,
|
|
"step": 230
|
|
},
|
|
{
|
|
"epoch": 0.17454545454545456,
|
|
"grad_norm": 0.6276031136512756,
|
|
"learning_rate": 9.415073115860518e-06,
|
|
"loss": 0.9021,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.18181818181818182,
|
|
"grad_norm": 0.6175471544265747,
|
|
"learning_rate": 9.377577802774654e-06,
|
|
"loss": 0.9,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.1890909090909091,
|
|
"grad_norm": 0.6632671356201172,
|
|
"learning_rate": 9.34008248968879e-06,
|
|
"loss": 0.9258,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.19636363636363635,
|
|
"grad_norm": 0.632043182849884,
|
|
"learning_rate": 9.302587176602925e-06,
|
|
"loss": 0.9246,
|
|
"step": 270
|
|
},
|
|
{
|
|
"epoch": 0.20363636363636364,
|
|
"grad_norm": 0.6637592315673828,
|
|
"learning_rate": 9.265091863517061e-06,
|
|
"loss": 0.9287,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.2109090909090909,
|
|
"grad_norm": 0.6319563984870911,
|
|
"learning_rate": 9.227596550431196e-06,
|
|
"loss": 0.9322,
|
|
"step": 290
|
|
},
|
|
{
|
|
"epoch": 0.21818181818181817,
|
|
"grad_norm": 0.6026636362075806,
|
|
"learning_rate": 9.190101237345333e-06,
|
|
"loss": 0.911,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.22545454545454546,
|
|
"grad_norm": 0.6266274452209473,
|
|
"learning_rate": 9.152605924259468e-06,
|
|
"loss": 0.9288,
|
|
"step": 310
|
|
},
|
|
{
|
|
"epoch": 0.23272727272727273,
|
|
"grad_norm": 0.6260218024253845,
|
|
"learning_rate": 9.115110611173603e-06,
|
|
"loss": 0.89,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.24,
|
|
"grad_norm": 0.619553804397583,
|
|
"learning_rate": 9.07761529808774e-06,
|
|
"loss": 0.9066,
|
|
"step": 330
|
|
},
|
|
{
|
|
"epoch": 0.24727272727272728,
|
|
"grad_norm": 0.6455956101417542,
|
|
"learning_rate": 9.040119985001875e-06,
|
|
"loss": 0.8969,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.2545454545454545,
|
|
"grad_norm": 0.6297126412391663,
|
|
"learning_rate": 9.002624671916012e-06,
|
|
"loss": 0.9087,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.26181818181818184,
|
|
"grad_norm": 0.6240535378456116,
|
|
"learning_rate": 8.965129358830147e-06,
|
|
"loss": 0.903,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.2690909090909091,
|
|
"grad_norm": 0.6323297023773193,
|
|
"learning_rate": 8.927634045744282e-06,
|
|
"loss": 0.9021,
|
|
"step": 370
|
|
},
|
|
{
|
|
"epoch": 0.27636363636363637,
|
|
"grad_norm": 0.629094660282135,
|
|
"learning_rate": 8.89013873265842e-06,
|
|
"loss": 0.8945,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.28363636363636363,
|
|
"grad_norm": 0.6123955249786377,
|
|
"learning_rate": 8.852643419572554e-06,
|
|
"loss": 0.8997,
|
|
"step": 390
|
|
},
|
|
{
|
|
"epoch": 0.2909090909090909,
|
|
"grad_norm": 0.6139543056488037,
|
|
"learning_rate": 8.81514810648669e-06,
|
|
"loss": 0.907,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.29818181818181816,
|
|
"grad_norm": 0.6207508444786072,
|
|
"learning_rate": 8.777652793400826e-06,
|
|
"loss": 0.9011,
|
|
"step": 410
|
|
},
|
|
{
|
|
"epoch": 0.3054545454545455,
|
|
"grad_norm": 0.6085091233253479,
|
|
"learning_rate": 8.740157480314961e-06,
|
|
"loss": 0.895,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 0.31272727272727274,
|
|
"grad_norm": 0.6231858134269714,
|
|
"learning_rate": 8.702662167229096e-06,
|
|
"loss": 0.9153,
|
|
"step": 430
|
|
},
|
|
{
|
|
"epoch": 0.32,
|
|
"grad_norm": 0.6155027747154236,
|
|
"learning_rate": 8.665166854143233e-06,
|
|
"loss": 0.8809,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 0.32727272727272727,
|
|
"grad_norm": 0.6154032349586487,
|
|
"learning_rate": 8.627671541057368e-06,
|
|
"loss": 0.879,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.33454545454545453,
|
|
"grad_norm": 0.6161757111549377,
|
|
"learning_rate": 8.590176227971503e-06,
|
|
"loss": 0.9212,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 0.3418181818181818,
|
|
"grad_norm": 0.6051761507987976,
|
|
"learning_rate": 8.55268091488564e-06,
|
|
"loss": 0.9021,
|
|
"step": 470
|
|
},
|
|
{
|
|
"epoch": 0.3490909090909091,
|
|
"grad_norm": 0.6489289999008179,
|
|
"learning_rate": 8.515185601799775e-06,
|
|
"loss": 0.8902,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 0.3563636363636364,
|
|
"grad_norm": 0.6450833082199097,
|
|
"learning_rate": 8.47769028871391e-06,
|
|
"loss": 0.9078,
|
|
"step": 490
|
|
},
|
|
{
|
|
"epoch": 0.36363636363636365,
|
|
"grad_norm": 0.5917351841926575,
|
|
"learning_rate": 8.440194975628047e-06,
|
|
"loss": 0.884,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.3709090909090909,
|
|
"grad_norm": 0.6286096572875977,
|
|
"learning_rate": 8.402699662542182e-06,
|
|
"loss": 0.8955,
|
|
"step": 510
|
|
},
|
|
{
|
|
"epoch": 0.3781818181818182,
|
|
"grad_norm": 0.6284880042076111,
|
|
"learning_rate": 8.365204349456319e-06,
|
|
"loss": 0.8909,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 0.38545454545454544,
|
|
"grad_norm": 0.6258820295333862,
|
|
"learning_rate": 8.327709036370454e-06,
|
|
"loss": 0.8852,
|
|
"step": 530
|
|
},
|
|
{
|
|
"epoch": 0.3927272727272727,
|
|
"grad_norm": 0.6000548005104065,
|
|
"learning_rate": 8.290213723284591e-06,
|
|
"loss": 0.865,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 0.4,
|
|
"grad_norm": 0.6457489728927612,
|
|
"learning_rate": 8.252718410198726e-06,
|
|
"loss": 0.8927,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.4072727272727273,
|
|
"grad_norm": 0.6275245547294617,
|
|
"learning_rate": 8.215223097112861e-06,
|
|
"loss": 0.88,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 0.41454545454545455,
|
|
"grad_norm": 0.6320298910140991,
|
|
"learning_rate": 8.177727784026998e-06,
|
|
"loss": 0.8886,
|
|
"step": 570
|
|
},
|
|
{
|
|
"epoch": 0.4218181818181818,
|
|
"grad_norm": 0.6126200556755066,
|
|
"learning_rate": 8.140232470941133e-06,
|
|
"loss": 0.8715,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 0.4290909090909091,
|
|
"grad_norm": 0.5955527424812317,
|
|
"learning_rate": 8.102737157855268e-06,
|
|
"loss": 0.8746,
|
|
"step": 590
|
|
},
|
|
{
|
|
"epoch": 0.43636363636363634,
|
|
"grad_norm": 0.6173805594444275,
|
|
"learning_rate": 8.065241844769405e-06,
|
|
"loss": 0.8798,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.44363636363636366,
|
|
"grad_norm": 0.6467263698577881,
|
|
"learning_rate": 8.02774653168354e-06,
|
|
"loss": 0.8815,
|
|
"step": 610
|
|
},
|
|
{
|
|
"epoch": 0.4509090909090909,
|
|
"grad_norm": 0.6411978602409363,
|
|
"learning_rate": 7.990251218597675e-06,
|
|
"loss": 0.8823,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 0.4581818181818182,
|
|
"grad_norm": 0.6345722079277039,
|
|
"learning_rate": 7.952755905511812e-06,
|
|
"loss": 0.8874,
|
|
"step": 630
|
|
},
|
|
{
|
|
"epoch": 0.46545454545454545,
|
|
"grad_norm": 0.6099885702133179,
|
|
"learning_rate": 7.915260592425947e-06,
|
|
"loss": 0.8809,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 0.4727272727272727,
|
|
"grad_norm": 0.623310923576355,
|
|
"learning_rate": 7.877765279340082e-06,
|
|
"loss": 0.8843,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.48,
|
|
"grad_norm": 0.6418883204460144,
|
|
"learning_rate": 7.840269966254219e-06,
|
|
"loss": 0.8817,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 0.48727272727272725,
|
|
"grad_norm": 0.5989499688148499,
|
|
"learning_rate": 7.802774653168354e-06,
|
|
"loss": 0.8867,
|
|
"step": 670
|
|
},
|
|
{
|
|
"epoch": 0.49454545454545457,
|
|
"grad_norm": 0.6249462366104126,
|
|
"learning_rate": 7.76527934008249e-06,
|
|
"loss": 0.8913,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 0.5018181818181818,
|
|
"grad_norm": 0.6270976662635803,
|
|
"learning_rate": 7.727784026996626e-06,
|
|
"loss": 0.8735,
|
|
"step": 690
|
|
},
|
|
{
|
|
"epoch": 0.509090909090909,
|
|
"grad_norm": 0.6204495429992676,
|
|
"learning_rate": 7.690288713910761e-06,
|
|
"loss": 0.8723,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.5163636363636364,
|
|
"grad_norm": 0.6200954914093018,
|
|
"learning_rate": 7.652793400824898e-06,
|
|
"loss": 0.8791,
|
|
"step": 710
|
|
},
|
|
{
|
|
"epoch": 0.5236363636363637,
|
|
"grad_norm": 0.6141085624694824,
|
|
"learning_rate": 7.615298087739034e-06,
|
|
"loss": 0.8731,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 0.5309090909090909,
|
|
"grad_norm": 0.657077431678772,
|
|
"learning_rate": 7.577802774653169e-06,
|
|
"loss": 0.8877,
|
|
"step": 730
|
|
},
|
|
{
|
|
"epoch": 0.5381818181818182,
|
|
"grad_norm": 0.6287103891372681,
|
|
"learning_rate": 7.540307461567304e-06,
|
|
"loss": 0.8688,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 0.5454545454545454,
|
|
"grad_norm": 0.6096076369285583,
|
|
"learning_rate": 7.502812148481441e-06,
|
|
"loss": 0.8837,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.5527272727272727,
|
|
"grad_norm": 0.6509519815444946,
|
|
"learning_rate": 7.465316835395576e-06,
|
|
"loss": 0.8833,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 0.56,
|
|
"grad_norm": 0.6012381315231323,
|
|
"learning_rate": 7.427821522309712e-06,
|
|
"loss": 0.8587,
|
|
"step": 770
|
|
},
|
|
{
|
|
"epoch": 0.5672727272727273,
|
|
"grad_norm": 0.6379227638244629,
|
|
"learning_rate": 7.390326209223848e-06,
|
|
"loss": 0.8659,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 0.5745454545454546,
|
|
"grad_norm": 0.5909913778305054,
|
|
"learning_rate": 7.352830896137984e-06,
|
|
"loss": 0.8822,
|
|
"step": 790
|
|
},
|
|
{
|
|
"epoch": 0.5818181818181818,
|
|
"grad_norm": 0.600817084312439,
|
|
"learning_rate": 7.315335583052119e-06,
|
|
"loss": 0.8781,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.5890909090909091,
|
|
"grad_norm": 0.603955090045929,
|
|
"learning_rate": 7.277840269966256e-06,
|
|
"loss": 0.8715,
|
|
"step": 810
|
|
},
|
|
{
|
|
"epoch": 0.5963636363636363,
|
|
"grad_norm": 0.5938793420791626,
|
|
"learning_rate": 7.240344956880391e-06,
|
|
"loss": 0.8792,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 0.6036363636363636,
|
|
"grad_norm": 0.6032233834266663,
|
|
"learning_rate": 7.202849643794526e-06,
|
|
"loss": 0.8911,
|
|
"step": 830
|
|
},
|
|
{
|
|
"epoch": 0.610909090909091,
|
|
"grad_norm": 0.6159318089485168,
|
|
"learning_rate": 7.165354330708662e-06,
|
|
"loss": 0.8579,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 0.6181818181818182,
|
|
"grad_norm": 0.6361089944839478,
|
|
"learning_rate": 7.127859017622798e-06,
|
|
"loss": 0.8528,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.6254545454545455,
|
|
"grad_norm": 0.6163432598114014,
|
|
"learning_rate": 7.090363704536934e-06,
|
|
"loss": 0.8685,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 0.6327272727272727,
|
|
"grad_norm": 0.5784986615180969,
|
|
"learning_rate": 7.052868391451069e-06,
|
|
"loss": 0.8654,
|
|
"step": 870
|
|
},
|
|
{
|
|
"epoch": 0.64,
|
|
"grad_norm": 0.6399606466293335,
|
|
"learning_rate": 7.015373078365206e-06,
|
|
"loss": 0.8681,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 0.6472727272727272,
|
|
"grad_norm": 0.6454589366912842,
|
|
"learning_rate": 6.977877765279341e-06,
|
|
"loss": 0.8755,
|
|
"step": 890
|
|
},
|
|
{
|
|
"epoch": 0.6545454545454545,
|
|
"grad_norm": 0.6353079080581665,
|
|
"learning_rate": 6.940382452193476e-06,
|
|
"loss": 0.8773,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.6618181818181819,
|
|
"grad_norm": 0.5652336478233337,
|
|
"learning_rate": 6.902887139107613e-06,
|
|
"loss": 0.8494,
|
|
"step": 910
|
|
},
|
|
{
|
|
"epoch": 0.6690909090909091,
|
|
"grad_norm": 0.5925925374031067,
|
|
"learning_rate": 6.865391826021748e-06,
|
|
"loss": 0.8566,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 0.6763636363636364,
|
|
"grad_norm": 0.6264874339103699,
|
|
"learning_rate": 6.827896512935883e-06,
|
|
"loss": 0.8755,
|
|
"step": 930
|
|
},
|
|
{
|
|
"epoch": 0.6836363636363636,
|
|
"grad_norm": 0.6029939651489258,
|
|
"learning_rate": 6.79040119985002e-06,
|
|
"loss": 0.8815,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 0.6909090909090909,
|
|
"grad_norm": 0.6007083654403687,
|
|
"learning_rate": 6.752905886764155e-06,
|
|
"loss": 0.8688,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.6981818181818182,
|
|
"grad_norm": 0.5800495147705078,
|
|
"learning_rate": 6.715410573678291e-06,
|
|
"loss": 0.8732,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 0.7054545454545454,
|
|
"grad_norm": 0.5926985740661621,
|
|
"learning_rate": 6.677915260592427e-06,
|
|
"loss": 0.8719,
|
|
"step": 970
|
|
},
|
|
{
|
|
"epoch": 0.7127272727272728,
|
|
"grad_norm": 0.6431898474693298,
|
|
"learning_rate": 6.6404199475065626e-06,
|
|
"loss": 0.8808,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 0.72,
|
|
"grad_norm": 0.5893067717552185,
|
|
"learning_rate": 6.602924634420698e-06,
|
|
"loss": 0.8571,
|
|
"step": 990
|
|
},
|
|
{
|
|
"epoch": 0.7272727272727273,
|
|
"grad_norm": 0.592795729637146,
|
|
"learning_rate": 6.5654293213348345e-06,
|
|
"loss": 0.8672,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.7272727272727273,
|
|
"eval_loss": 0.919316291809082,
|
|
"eval_runtime": 2141.287,
|
|
"eval_samples_per_second": 55.842,
|
|
"eval_steps_per_second": 6.98,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.7345454545454545,
|
|
"grad_norm": 0.6309456825256348,
|
|
"learning_rate": 6.5279340082489696e-06,
|
|
"loss": 0.8528,
|
|
"step": 1010
|
|
},
|
|
{
|
|
"epoch": 0.7418181818181818,
|
|
"grad_norm": 0.6349218487739563,
|
|
"learning_rate": 6.490438695163105e-06,
|
|
"loss": 0.8513,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 0.7490909090909091,
|
|
"grad_norm": 0.5786005258560181,
|
|
"learning_rate": 6.4529433820772415e-06,
|
|
"loss": 0.8856,
|
|
"step": 1030
|
|
},
|
|
{
|
|
"epoch": 0.7563636363636363,
|
|
"grad_norm": 0.5970295667648315,
|
|
"learning_rate": 6.415448068991377e-06,
|
|
"loss": 0.8567,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 0.7636363636363637,
|
|
"grad_norm": 0.5964470505714417,
|
|
"learning_rate": 6.3779527559055125e-06,
|
|
"loss": 0.8656,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.7709090909090909,
|
|
"grad_norm": 0.6192710995674133,
|
|
"learning_rate": 6.3404574428196485e-06,
|
|
"loss": 0.8496,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 0.7781818181818182,
|
|
"grad_norm": 0.5974547863006592,
|
|
"learning_rate": 6.302962129733784e-06,
|
|
"loss": 0.8534,
|
|
"step": 1070
|
|
},
|
|
{
|
|
"epoch": 0.7854545454545454,
|
|
"grad_norm": 0.5947730541229248,
|
|
"learning_rate": 6.2654668166479195e-06,
|
|
"loss": 0.8524,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 0.7927272727272727,
|
|
"grad_norm": 0.6238484978675842,
|
|
"learning_rate": 6.227971503562055e-06,
|
|
"loss": 0.8605,
|
|
"step": 1090
|
|
},
|
|
{
|
|
"epoch": 0.8,
|
|
"grad_norm": 0.6511428952217102,
|
|
"learning_rate": 6.1904761904761914e-06,
|
|
"loss": 0.867,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.8072727272727273,
|
|
"grad_norm": 0.6051254868507385,
|
|
"learning_rate": 6.1529808773903265e-06,
|
|
"loss": 0.8759,
|
|
"step": 1110
|
|
},
|
|
{
|
|
"epoch": 0.8145454545454546,
|
|
"grad_norm": 0.6452696323394775,
|
|
"learning_rate": 6.115485564304462e-06,
|
|
"loss": 0.8839,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 0.8218181818181818,
|
|
"grad_norm": 0.6039929389953613,
|
|
"learning_rate": 6.0779902512185984e-06,
|
|
"loss": 0.8465,
|
|
"step": 1130
|
|
},
|
|
{
|
|
"epoch": 0.8290909090909091,
|
|
"grad_norm": 0.6003919243812561,
|
|
"learning_rate": 6.0404949381327335e-06,
|
|
"loss": 0.8308,
|
|
"step": 1140
|
|
},
|
|
{
|
|
"epoch": 0.8363636363636363,
|
|
"grad_norm": 0.6162687540054321,
|
|
"learning_rate": 6.0029996250468695e-06,
|
|
"loss": 0.8835,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 0.8436363636363636,
|
|
"grad_norm": 0.6076496243476868,
|
|
"learning_rate": 5.9655043119610054e-06,
|
|
"loss": 0.8677,
|
|
"step": 1160
|
|
},
|
|
{
|
|
"epoch": 0.850909090909091,
|
|
"grad_norm": 0.6274608969688416,
|
|
"learning_rate": 5.928008998875141e-06,
|
|
"loss": 0.8705,
|
|
"step": 1170
|
|
},
|
|
{
|
|
"epoch": 0.8581818181818182,
|
|
"grad_norm": 0.6353155970573425,
|
|
"learning_rate": 5.8905136857892765e-06,
|
|
"loss": 0.8519,
|
|
"step": 1180
|
|
},
|
|
{
|
|
"epoch": 0.8654545454545455,
|
|
"grad_norm": 0.6303426623344421,
|
|
"learning_rate": 5.853018372703413e-06,
|
|
"loss": 0.8536,
|
|
"step": 1190
|
|
},
|
|
{
|
|
"epoch": 0.8727272727272727,
|
|
"grad_norm": 0.607300877571106,
|
|
"learning_rate": 5.815523059617548e-06,
|
|
"loss": 0.8695,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.88,
|
|
"grad_norm": 0.6049074530601501,
|
|
"learning_rate": 5.7780277465316835e-06,
|
|
"loss": 0.8722,
|
|
"step": 1210
|
|
},
|
|
{
|
|
"epoch": 0.8872727272727273,
|
|
"grad_norm": 0.5959789156913757,
|
|
"learning_rate": 5.74053243344582e-06,
|
|
"loss": 0.8507,
|
|
"step": 1220
|
|
},
|
|
{
|
|
"epoch": 0.8945454545454545,
|
|
"grad_norm": 0.5832094550132751,
|
|
"learning_rate": 5.703037120359955e-06,
|
|
"loss": 0.858,
|
|
"step": 1230
|
|
},
|
|
{
|
|
"epoch": 0.9018181818181819,
|
|
"grad_norm": 0.5952485799789429,
|
|
"learning_rate": 5.6655418072740905e-06,
|
|
"loss": 0.8521,
|
|
"step": 1240
|
|
},
|
|
{
|
|
"epoch": 0.9090909090909091,
|
|
"grad_norm": 0.5914123058319092,
|
|
"learning_rate": 5.628046494188227e-06,
|
|
"loss": 0.8564,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 0.9163636363636364,
|
|
"grad_norm": 0.5735720992088318,
|
|
"learning_rate": 5.590551181102362e-06,
|
|
"loss": 0.8721,
|
|
"step": 1260
|
|
},
|
|
{
|
|
"epoch": 0.9236363636363636,
|
|
"grad_norm": 0.6073433756828308,
|
|
"learning_rate": 5.553055868016498e-06,
|
|
"loss": 0.8653,
|
|
"step": 1270
|
|
},
|
|
{
|
|
"epoch": 0.9309090909090909,
|
|
"grad_norm": 0.5949496626853943,
|
|
"learning_rate": 5.515560554930634e-06,
|
|
"loss": 0.8665,
|
|
"step": 1280
|
|
},
|
|
{
|
|
"epoch": 0.9381818181818182,
|
|
"grad_norm": 0.6578984260559082,
|
|
"learning_rate": 5.47806524184477e-06,
|
|
"loss": 0.8544,
|
|
"step": 1290
|
|
},
|
|
{
|
|
"epoch": 0.9454545454545454,
|
|
"grad_norm": 0.6195460557937622,
|
|
"learning_rate": 5.440569928758905e-06,
|
|
"loss": 0.8561,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.9527272727272728,
|
|
"grad_norm": 0.5933972597122192,
|
|
"learning_rate": 5.403074615673042e-06,
|
|
"loss": 0.863,
|
|
"step": 1310
|
|
},
|
|
{
|
|
"epoch": 0.96,
|
|
"grad_norm": 0.5625194311141968,
|
|
"learning_rate": 5.365579302587177e-06,
|
|
"loss": 0.8698,
|
|
"step": 1320
|
|
},
|
|
{
|
|
"epoch": 0.9672727272727273,
|
|
"grad_norm": 0.5966600775718689,
|
|
"learning_rate": 5.328083989501312e-06,
|
|
"loss": 0.8412,
|
|
"step": 1330
|
|
},
|
|
{
|
|
"epoch": 0.9745454545454545,
|
|
"grad_norm": 0.6089814901351929,
|
|
"learning_rate": 5.290588676415449e-06,
|
|
"loss": 0.853,
|
|
"step": 1340
|
|
},
|
|
{
|
|
"epoch": 0.9818181818181818,
|
|
"grad_norm": 0.5774534344673157,
|
|
"learning_rate": 5.253093363329584e-06,
|
|
"loss": 0.8682,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 0.9890909090909091,
|
|
"grad_norm": 0.6104735732078552,
|
|
"learning_rate": 5.21559805024372e-06,
|
|
"loss": 0.8553,
|
|
"step": 1360
|
|
},
|
|
{
|
|
"epoch": 0.9963636363636363,
|
|
"grad_norm": 0.6288988590240479,
|
|
"learning_rate": 5.178102737157855e-06,
|
|
"loss": 0.8257,
|
|
"step": 1370
|
|
},
|
|
{
|
|
"epoch": 1.0036363636363637,
|
|
"grad_norm": 0.6185275316238403,
|
|
"learning_rate": 5.140607424071991e-06,
|
|
"loss": 0.8182,
|
|
"step": 1380
|
|
},
|
|
{
|
|
"epoch": 1.010909090909091,
|
|
"grad_norm": 0.6155534386634827,
|
|
"learning_rate": 5.103112110986127e-06,
|
|
"loss": 0.8235,
|
|
"step": 1390
|
|
},
|
|
{
|
|
"epoch": 1.018181818181818,
|
|
"grad_norm": 0.6087826490402222,
|
|
"learning_rate": 5.065616797900262e-06,
|
|
"loss": 0.8182,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 1.0254545454545454,
|
|
"grad_norm": 0.5803642272949219,
|
|
"learning_rate": 5.028121484814399e-06,
|
|
"loss": 0.8218,
|
|
"step": 1410
|
|
},
|
|
{
|
|
"epoch": 1.0327272727272727,
|
|
"grad_norm": 0.6309434771537781,
|
|
"learning_rate": 4.990626171728534e-06,
|
|
"loss": 0.8125,
|
|
"step": 1420
|
|
},
|
|
{
|
|
"epoch": 1.04,
|
|
"grad_norm": 0.6362752914428711,
|
|
"learning_rate": 4.95313085864267e-06,
|
|
"loss": 0.8066,
|
|
"step": 1430
|
|
},
|
|
{
|
|
"epoch": 1.0472727272727274,
|
|
"grad_norm": 0.5848464369773865,
|
|
"learning_rate": 4.915635545556805e-06,
|
|
"loss": 0.8156,
|
|
"step": 1440
|
|
},
|
|
{
|
|
"epoch": 1.0545454545454545,
|
|
"grad_norm": 0.619362473487854,
|
|
"learning_rate": 4.878140232470941e-06,
|
|
"loss": 0.8158,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 1.0618181818181818,
|
|
"grad_norm": 0.5853692889213562,
|
|
"learning_rate": 4.840644919385077e-06,
|
|
"loss": 0.8319,
|
|
"step": 1460
|
|
},
|
|
{
|
|
"epoch": 1.069090909090909,
|
|
"grad_norm": 0.607729434967041,
|
|
"learning_rate": 4.803149606299213e-06,
|
|
"loss": 0.8145,
|
|
"step": 1470
|
|
},
|
|
{
|
|
"epoch": 1.0763636363636364,
|
|
"grad_norm": 0.5888605117797852,
|
|
"learning_rate": 4.765654293213349e-06,
|
|
"loss": 0.8114,
|
|
"step": 1480
|
|
},
|
|
{
|
|
"epoch": 1.0836363636363637,
|
|
"grad_norm": 0.6180461049079895,
|
|
"learning_rate": 4.728158980127485e-06,
|
|
"loss": 0.8234,
|
|
"step": 1490
|
|
},
|
|
{
|
|
"epoch": 1.0909090909090908,
|
|
"grad_norm": 0.5963523387908936,
|
|
"learning_rate": 4.69066366704162e-06,
|
|
"loss": 0.8159,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 1.0981818181818181,
|
|
"grad_norm": 0.6214097738265991,
|
|
"learning_rate": 4.653168353955756e-06,
|
|
"loss": 0.8256,
|
|
"step": 1510
|
|
},
|
|
{
|
|
"epoch": 1.1054545454545455,
|
|
"grad_norm": 0.5866185426712036,
|
|
"learning_rate": 4.615673040869892e-06,
|
|
"loss": 0.8431,
|
|
"step": 1520
|
|
},
|
|
{
|
|
"epoch": 1.1127272727272728,
|
|
"grad_norm": 0.5828819870948792,
|
|
"learning_rate": 4.578177727784027e-06,
|
|
"loss": 0.8064,
|
|
"step": 1530
|
|
},
|
|
{
|
|
"epoch": 1.12,
|
|
"grad_norm": 0.5965102910995483,
|
|
"learning_rate": 4.540682414698163e-06,
|
|
"loss": 0.8281,
|
|
"step": 1540
|
|
},
|
|
{
|
|
"epoch": 1.1272727272727272,
|
|
"grad_norm": 0.5992522239685059,
|
|
"learning_rate": 4.503187101612299e-06,
|
|
"loss": 0.8134,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 1.1345454545454545,
|
|
"grad_norm": 0.6310180425643921,
|
|
"learning_rate": 4.465691788526434e-06,
|
|
"loss": 0.819,
|
|
"step": 1560
|
|
},
|
|
{
|
|
"epoch": 1.1418181818181818,
|
|
"grad_norm": 0.588926374912262,
|
|
"learning_rate": 4.42819647544057e-06,
|
|
"loss": 0.8333,
|
|
"step": 1570
|
|
},
|
|
{
|
|
"epoch": 1.1490909090909092,
|
|
"grad_norm": 0.5981424450874329,
|
|
"learning_rate": 4.390701162354706e-06,
|
|
"loss": 0.8386,
|
|
"step": 1580
|
|
},
|
|
{
|
|
"epoch": 1.1563636363636363,
|
|
"grad_norm": 0.634250283241272,
|
|
"learning_rate": 4.353205849268842e-06,
|
|
"loss": 0.8034,
|
|
"step": 1590
|
|
},
|
|
{
|
|
"epoch": 1.1636363636363636,
|
|
"grad_norm": 0.5827885866165161,
|
|
"learning_rate": 4.315710536182978e-06,
|
|
"loss": 0.8288,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 1.170909090909091,
|
|
"grad_norm": 0.6020681858062744,
|
|
"learning_rate": 4.278215223097113e-06,
|
|
"loss": 0.8181,
|
|
"step": 1610
|
|
},
|
|
{
|
|
"epoch": 1.1781818181818182,
|
|
"grad_norm": 0.5973985195159912,
|
|
"learning_rate": 4.240719910011249e-06,
|
|
"loss": 0.8206,
|
|
"step": 1620
|
|
},
|
|
{
|
|
"epoch": 1.1854545454545455,
|
|
"grad_norm": 0.5879353880882263,
|
|
"learning_rate": 4.203224596925385e-06,
|
|
"loss": 0.8129,
|
|
"step": 1630
|
|
},
|
|
{
|
|
"epoch": 1.1927272727272726,
|
|
"grad_norm": 0.6243763566017151,
|
|
"learning_rate": 4.16572928383952e-06,
|
|
"loss": 0.8067,
|
|
"step": 1640
|
|
},
|
|
{
|
|
"epoch": 1.2,
|
|
"grad_norm": 0.583442211151123,
|
|
"learning_rate": 4.128233970753656e-06,
|
|
"loss": 0.8279,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 1.2072727272727273,
|
|
"grad_norm": 0.6032597422599792,
|
|
"learning_rate": 4.090738657667792e-06,
|
|
"loss": 0.8115,
|
|
"step": 1660
|
|
},
|
|
{
|
|
"epoch": 1.2145454545454546,
|
|
"grad_norm": 0.6115988492965698,
|
|
"learning_rate": 4.053243344581928e-06,
|
|
"loss": 0.8221,
|
|
"step": 1670
|
|
},
|
|
{
|
|
"epoch": 1.221818181818182,
|
|
"grad_norm": 0.600167453289032,
|
|
"learning_rate": 4.015748031496064e-06,
|
|
"loss": 0.8008,
|
|
"step": 1680
|
|
},
|
|
{
|
|
"epoch": 1.229090909090909,
|
|
"grad_norm": 0.6090269088745117,
|
|
"learning_rate": 3.978252718410199e-06,
|
|
"loss": 0.8196,
|
|
"step": 1690
|
|
},
|
|
{
|
|
"epoch": 1.2363636363636363,
|
|
"grad_norm": 0.598509669303894,
|
|
"learning_rate": 3.940757405324335e-06,
|
|
"loss": 0.8168,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 1.2436363636363637,
|
|
"grad_norm": 0.5862186551094055,
|
|
"learning_rate": 3.903262092238471e-06,
|
|
"loss": 0.819,
|
|
"step": 1710
|
|
},
|
|
{
|
|
"epoch": 1.250909090909091,
|
|
"grad_norm": 0.6207866072654724,
|
|
"learning_rate": 3.865766779152606e-06,
|
|
"loss": 0.8074,
|
|
"step": 1720
|
|
},
|
|
{
|
|
"epoch": 1.2581818181818183,
|
|
"grad_norm": 0.5965321063995361,
|
|
"learning_rate": 3.828271466066742e-06,
|
|
"loss": 0.8149,
|
|
"step": 1730
|
|
},
|
|
{
|
|
"epoch": 1.2654545454545454,
|
|
"grad_norm": 0.6035894751548767,
|
|
"learning_rate": 3.790776152980878e-06,
|
|
"loss": 0.8347,
|
|
"step": 1740
|
|
},
|
|
{
|
|
"epoch": 1.2727272727272727,
|
|
"grad_norm": 0.5681039094924927,
|
|
"learning_rate": 3.7532808398950133e-06,
|
|
"loss": 0.8038,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 1.28,
|
|
"grad_norm": 0.6101064085960388,
|
|
"learning_rate": 3.7157855268091493e-06,
|
|
"loss": 0.7951,
|
|
"step": 1760
|
|
},
|
|
{
|
|
"epoch": 1.2872727272727273,
|
|
"grad_norm": 0.5739514827728271,
|
|
"learning_rate": 3.6782902137232852e-06,
|
|
"loss": 0.8189,
|
|
"step": 1770
|
|
},
|
|
{
|
|
"epoch": 1.2945454545454544,
|
|
"grad_norm": 0.5720605254173279,
|
|
"learning_rate": 3.6407949006374203e-06,
|
|
"loss": 0.8021,
|
|
"step": 1780
|
|
},
|
|
{
|
|
"epoch": 1.3018181818181818,
|
|
"grad_norm": 0.5951290726661682,
|
|
"learning_rate": 3.6032995875515563e-06,
|
|
"loss": 0.7991,
|
|
"step": 1790
|
|
},
|
|
{
|
|
"epoch": 1.309090909090909,
|
|
"grad_norm": 0.5760191679000854,
|
|
"learning_rate": 3.5658042744656922e-06,
|
|
"loss": 0.8131,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 1.3163636363636364,
|
|
"grad_norm": 0.6369950771331787,
|
|
"learning_rate": 3.5283089613798277e-06,
|
|
"loss": 0.8319,
|
|
"step": 1810
|
|
},
|
|
{
|
|
"epoch": 1.3236363636363637,
|
|
"grad_norm": 0.6073457598686218,
|
|
"learning_rate": 3.4908136482939637e-06,
|
|
"loss": 0.8173,
|
|
"step": 1820
|
|
},
|
|
{
|
|
"epoch": 1.330909090909091,
|
|
"grad_norm": 0.6086942553520203,
|
|
"learning_rate": 3.453318335208099e-06,
|
|
"loss": 0.8218,
|
|
"step": 1830
|
|
},
|
|
{
|
|
"epoch": 1.3381818181818181,
|
|
"grad_norm": 0.587218165397644,
|
|
"learning_rate": 3.4158230221222347e-06,
|
|
"loss": 0.8167,
|
|
"step": 1840
|
|
},
|
|
{
|
|
"epoch": 1.3454545454545455,
|
|
"grad_norm": 0.5894433856010437,
|
|
"learning_rate": 3.3783277090363707e-06,
|
|
"loss": 0.8253,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 1.3527272727272728,
|
|
"grad_norm": 0.5601096749305725,
|
|
"learning_rate": 3.3408323959505062e-06,
|
|
"loss": 0.801,
|
|
"step": 1860
|
|
},
|
|
{
|
|
"epoch": 1.3599999999999999,
|
|
"grad_norm": 0.6079325079917908,
|
|
"learning_rate": 3.303337082864642e-06,
|
|
"loss": 0.8117,
|
|
"step": 1870
|
|
},
|
|
{
|
|
"epoch": 1.3672727272727272,
|
|
"grad_norm": 0.581007182598114,
|
|
"learning_rate": 3.265841769778778e-06,
|
|
"loss": 0.8241,
|
|
"step": 1880
|
|
},
|
|
{
|
|
"epoch": 1.3745454545454545,
|
|
"grad_norm": 0.5827407836914062,
|
|
"learning_rate": 3.2283464566929136e-06,
|
|
"loss": 0.8193,
|
|
"step": 1890
|
|
},
|
|
{
|
|
"epoch": 1.3818181818181818,
|
|
"grad_norm": 0.5792385935783386,
|
|
"learning_rate": 3.1908511436070496e-06,
|
|
"loss": 0.8107,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 1.3890909090909092,
|
|
"grad_norm": 0.5755916833877563,
|
|
"learning_rate": 3.153355830521185e-06,
|
|
"loss": 0.8083,
|
|
"step": 1910
|
|
},
|
|
{
|
|
"epoch": 1.3963636363636365,
|
|
"grad_norm": 0.6043480634689331,
|
|
"learning_rate": 3.1158605174353206e-06,
|
|
"loss": 0.8047,
|
|
"step": 1920
|
|
},
|
|
{
|
|
"epoch": 1.4036363636363636,
|
|
"grad_norm": 0.5923509001731873,
|
|
"learning_rate": 3.0783652043494566e-06,
|
|
"loss": 0.8264,
|
|
"step": 1930
|
|
},
|
|
{
|
|
"epoch": 1.410909090909091,
|
|
"grad_norm": 0.5509864091873169,
|
|
"learning_rate": 3.0408698912635925e-06,
|
|
"loss": 0.8219,
|
|
"step": 1940
|
|
},
|
|
{
|
|
"epoch": 1.4181818181818182,
|
|
"grad_norm": 0.6133068799972534,
|
|
"learning_rate": 3.003374578177728e-06,
|
|
"loss": 0.8303,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 1.4254545454545455,
|
|
"grad_norm": 0.5911954045295715,
|
|
"learning_rate": 2.965879265091864e-06,
|
|
"loss": 0.8245,
|
|
"step": 1960
|
|
},
|
|
{
|
|
"epoch": 1.4327272727272726,
|
|
"grad_norm": 0.609562337398529,
|
|
"learning_rate": 2.928383952005999e-06,
|
|
"loss": 0.7991,
|
|
"step": 1970
|
|
},
|
|
{
|
|
"epoch": 1.44,
|
|
"grad_norm": 0.5898092985153198,
|
|
"learning_rate": 2.890888638920135e-06,
|
|
"loss": 0.8264,
|
|
"step": 1980
|
|
},
|
|
{
|
|
"epoch": 1.4472727272727273,
|
|
"grad_norm": 0.6039696931838989,
|
|
"learning_rate": 2.853393325834271e-06,
|
|
"loss": 0.8176,
|
|
"step": 1990
|
|
},
|
|
{
|
|
"epoch": 1.4545454545454546,
|
|
"grad_norm": 0.6092890501022339,
|
|
"learning_rate": 2.8158980127484065e-06,
|
|
"loss": 0.8101,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 1.4545454545454546,
|
|
"eval_loss": 0.8971814513206482,
|
|
"eval_runtime": 2118.2926,
|
|
"eval_samples_per_second": 56.448,
|
|
"eval_steps_per_second": 7.056,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 1.461818181818182,
|
|
"grad_norm": 0.6092020273208618,
|
|
"learning_rate": 2.7784026996625425e-06,
|
|
"loss": 0.8359,
|
|
"step": 2010
|
|
},
|
|
{
|
|
"epoch": 1.4690909090909092,
|
|
"grad_norm": 0.6030200719833374,
|
|
"learning_rate": 2.7409073865766784e-06,
|
|
"loss": 0.7967,
|
|
"step": 2020
|
|
},
|
|
{
|
|
"epoch": 1.4763636363636363,
|
|
"grad_norm": 0.5874107480049133,
|
|
"learning_rate": 2.7034120734908135e-06,
|
|
"loss": 0.8141,
|
|
"step": 2030
|
|
},
|
|
{
|
|
"epoch": 1.4836363636363636,
|
|
"grad_norm": 0.6113495826721191,
|
|
"learning_rate": 2.6659167604049495e-06,
|
|
"loss": 0.8291,
|
|
"step": 2040
|
|
},
|
|
{
|
|
"epoch": 1.490909090909091,
|
|
"grad_norm": 0.5739171504974365,
|
|
"learning_rate": 2.6284214473190854e-06,
|
|
"loss": 0.8159,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 1.498181818181818,
|
|
"grad_norm": 0.5858970880508423,
|
|
"learning_rate": 2.590926134233221e-06,
|
|
"loss": 0.822,
|
|
"step": 2060
|
|
},
|
|
{
|
|
"epoch": 1.5054545454545454,
|
|
"grad_norm": 0.5980799794197083,
|
|
"learning_rate": 2.553430821147357e-06,
|
|
"loss": 0.8052,
|
|
"step": 2070
|
|
},
|
|
{
|
|
"epoch": 1.5127272727272727,
|
|
"grad_norm": 0.5883092284202576,
|
|
"learning_rate": 2.515935508061493e-06,
|
|
"loss": 0.8125,
|
|
"step": 2080
|
|
},
|
|
{
|
|
"epoch": 1.52,
|
|
"grad_norm": 0.5935123562812805,
|
|
"learning_rate": 2.478440194975628e-06,
|
|
"loss": 0.7978,
|
|
"step": 2090
|
|
},
|
|
{
|
|
"epoch": 1.5272727272727273,
|
|
"grad_norm": 0.5738603472709656,
|
|
"learning_rate": 2.440944881889764e-06,
|
|
"loss": 0.8352,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 1.5345454545454547,
|
|
"grad_norm": 0.5987142324447632,
|
|
"learning_rate": 2.4034495688039e-06,
|
|
"loss": 0.8151,
|
|
"step": 2110
|
|
},
|
|
{
|
|
"epoch": 1.541818181818182,
|
|
"grad_norm": 0.5858705639839172,
|
|
"learning_rate": 2.3659542557180354e-06,
|
|
"loss": 0.8258,
|
|
"step": 2120
|
|
},
|
|
{
|
|
"epoch": 1.549090909090909,
|
|
"grad_norm": 0.5583072900772095,
|
|
"learning_rate": 2.328458942632171e-06,
|
|
"loss": 0.8126,
|
|
"step": 2130
|
|
},
|
|
{
|
|
"epoch": 1.5563636363636364,
|
|
"grad_norm": 0.5778517723083496,
|
|
"learning_rate": 2.290963629546307e-06,
|
|
"loss": 0.7926,
|
|
"step": 2140
|
|
},
|
|
{
|
|
"epoch": 1.5636363636363635,
|
|
"grad_norm": 0.5871937274932861,
|
|
"learning_rate": 2.253468316460443e-06,
|
|
"loss": 0.8216,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 1.5709090909090908,
|
|
"grad_norm": 0.581562876701355,
|
|
"learning_rate": 2.2159730033745783e-06,
|
|
"loss": 0.8096,
|
|
"step": 2160
|
|
},
|
|
{
|
|
"epoch": 1.5781818181818181,
|
|
"grad_norm": 0.5970920324325562,
|
|
"learning_rate": 2.1784776902887143e-06,
|
|
"loss": 0.8306,
|
|
"step": 2170
|
|
},
|
|
{
|
|
"epoch": 1.5854545454545454,
|
|
"grad_norm": 0.5595991015434265,
|
|
"learning_rate": 2.14098237720285e-06,
|
|
"loss": 0.8053,
|
|
"step": 2180
|
|
},
|
|
{
|
|
"epoch": 1.5927272727272728,
|
|
"grad_norm": 0.592958927154541,
|
|
"learning_rate": 2.1034870641169853e-06,
|
|
"loss": 0.7871,
|
|
"step": 2190
|
|
},
|
|
{
|
|
"epoch": 1.6,
|
|
"grad_norm": 0.5841421484947205,
|
|
"learning_rate": 2.0659917510311213e-06,
|
|
"loss": 0.8051,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 1.6072727272727274,
|
|
"grad_norm": 0.577006995677948,
|
|
"learning_rate": 2.0284964379452572e-06,
|
|
"loss": 0.8123,
|
|
"step": 2210
|
|
},
|
|
{
|
|
"epoch": 1.6145454545454545,
|
|
"grad_norm": 0.5763828754425049,
|
|
"learning_rate": 1.9910011248593928e-06,
|
|
"loss": 0.7833,
|
|
"step": 2220
|
|
},
|
|
{
|
|
"epoch": 1.6218181818181818,
|
|
"grad_norm": 0.582817792892456,
|
|
"learning_rate": 1.9535058117735283e-06,
|
|
"loss": 0.8126,
|
|
"step": 2230
|
|
},
|
|
{
|
|
"epoch": 1.6290909090909091,
|
|
"grad_norm": 0.5699591040611267,
|
|
"learning_rate": 1.9160104986876642e-06,
|
|
"loss": 0.8128,
|
|
"step": 2240
|
|
},
|
|
{
|
|
"epoch": 1.6363636363636362,
|
|
"grad_norm": 0.5779479742050171,
|
|
"learning_rate": 1.8785151856018e-06,
|
|
"loss": 0.8121,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 1.6436363636363636,
|
|
"grad_norm": 0.5711218118667603,
|
|
"learning_rate": 1.8410198725159355e-06,
|
|
"loss": 0.8032,
|
|
"step": 2260
|
|
},
|
|
{
|
|
"epoch": 1.6509090909090909,
|
|
"grad_norm": 0.6018957495689392,
|
|
"learning_rate": 1.8035245594300713e-06,
|
|
"loss": 0.8125,
|
|
"step": 2270
|
|
},
|
|
{
|
|
"epoch": 1.6581818181818182,
|
|
"grad_norm": 0.570153534412384,
|
|
"learning_rate": 1.7660292463442072e-06,
|
|
"loss": 0.8143,
|
|
"step": 2280
|
|
},
|
|
{
|
|
"epoch": 1.6654545454545455,
|
|
"grad_norm": 0.5871094465255737,
|
|
"learning_rate": 1.728533933258343e-06,
|
|
"loss": 0.8179,
|
|
"step": 2290
|
|
},
|
|
{
|
|
"epoch": 1.6727272727272728,
|
|
"grad_norm": 0.5894734859466553,
|
|
"learning_rate": 1.6910386201724785e-06,
|
|
"loss": 0.8015,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 1.6800000000000002,
|
|
"grad_norm": 0.6005243062973022,
|
|
"learning_rate": 1.6535433070866144e-06,
|
|
"loss": 0.8009,
|
|
"step": 2310
|
|
},
|
|
{
|
|
"epoch": 1.6872727272727273,
|
|
"grad_norm": 0.5786961317062378,
|
|
"learning_rate": 1.6160479940007502e-06,
|
|
"loss": 0.82,
|
|
"step": 2320
|
|
},
|
|
{
|
|
"epoch": 1.6945454545454546,
|
|
"grad_norm": 0.5606324672698975,
|
|
"learning_rate": 1.5785526809148857e-06,
|
|
"loss": 0.8074,
|
|
"step": 2330
|
|
},
|
|
{
|
|
"epoch": 1.7018181818181817,
|
|
"grad_norm": 0.5757468938827515,
|
|
"learning_rate": 1.5410573678290214e-06,
|
|
"loss": 0.7997,
|
|
"step": 2340
|
|
},
|
|
{
|
|
"epoch": 1.709090909090909,
|
|
"grad_norm": 0.5801776051521301,
|
|
"learning_rate": 1.5035620547431574e-06,
|
|
"loss": 0.8159,
|
|
"step": 2350
|
|
},
|
|
{
|
|
"epoch": 1.7163636363636363,
|
|
"grad_norm": 0.561937689781189,
|
|
"learning_rate": 1.4660667416572929e-06,
|
|
"loss": 0.8088,
|
|
"step": 2360
|
|
},
|
|
{
|
|
"epoch": 1.7236363636363636,
|
|
"grad_norm": 0.581520140171051,
|
|
"learning_rate": 1.4285714285714286e-06,
|
|
"loss": 0.8137,
|
|
"step": 2370
|
|
},
|
|
{
|
|
"epoch": 1.730909090909091,
|
|
"grad_norm": 0.5766869783401489,
|
|
"learning_rate": 1.3910761154855646e-06,
|
|
"loss": 0.8115,
|
|
"step": 2380
|
|
},
|
|
{
|
|
"epoch": 1.7381818181818183,
|
|
"grad_norm": 0.5791675448417664,
|
|
"learning_rate": 1.3535808023997e-06,
|
|
"loss": 0.7978,
|
|
"step": 2390
|
|
},
|
|
{
|
|
"epoch": 1.7454545454545456,
|
|
"grad_norm": 0.5874686241149902,
|
|
"learning_rate": 1.3160854893138358e-06,
|
|
"loss": 0.8226,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 1.7527272727272727,
|
|
"grad_norm": 0.5620043277740479,
|
|
"learning_rate": 1.2785901762279716e-06,
|
|
"loss": 0.7917,
|
|
"step": 2410
|
|
},
|
|
{
|
|
"epoch": 1.76,
|
|
"grad_norm": 0.563328206539154,
|
|
"learning_rate": 1.2410948631421073e-06,
|
|
"loss": 0.8061,
|
|
"step": 2420
|
|
},
|
|
{
|
|
"epoch": 1.767272727272727,
|
|
"grad_norm": 0.5629045963287354,
|
|
"learning_rate": 1.203599550056243e-06,
|
|
"loss": 0.8104,
|
|
"step": 2430
|
|
},
|
|
{
|
|
"epoch": 1.7745454545454544,
|
|
"grad_norm": 0.5813615918159485,
|
|
"learning_rate": 1.1661042369703788e-06,
|
|
"loss": 0.8108,
|
|
"step": 2440
|
|
},
|
|
{
|
|
"epoch": 1.7818181818181817,
|
|
"grad_norm": 0.5680747628211975,
|
|
"learning_rate": 1.1286089238845145e-06,
|
|
"loss": 0.8117,
|
|
"step": 2450
|
|
},
|
|
{
|
|
"epoch": 1.789090909090909,
|
|
"grad_norm": 0.5629953742027283,
|
|
"learning_rate": 1.0911136107986503e-06,
|
|
"loss": 0.7866,
|
|
"step": 2460
|
|
},
|
|
{
|
|
"epoch": 1.7963636363636364,
|
|
"grad_norm": 0.5827091932296753,
|
|
"learning_rate": 1.053618297712786e-06,
|
|
"loss": 0.7988,
|
|
"step": 2470
|
|
},
|
|
{
|
|
"epoch": 1.8036363636363637,
|
|
"grad_norm": 0.5806558728218079,
|
|
"learning_rate": 1.0161229846269217e-06,
|
|
"loss": 0.8169,
|
|
"step": 2480
|
|
},
|
|
{
|
|
"epoch": 1.810909090909091,
|
|
"grad_norm": 0.5832482576370239,
|
|
"learning_rate": 9.786276715410575e-07,
|
|
"loss": 0.8205,
|
|
"step": 2490
|
|
},
|
|
{
|
|
"epoch": 1.8181818181818183,
|
|
"grad_norm": 0.5685613751411438,
|
|
"learning_rate": 9.411323584551932e-07,
|
|
"loss": 0.8071,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 1.8254545454545454,
|
|
"grad_norm": 0.6105993390083313,
|
|
"learning_rate": 9.03637045369329e-07,
|
|
"loss": 0.8097,
|
|
"step": 2510
|
|
},
|
|
{
|
|
"epoch": 1.8327272727272728,
|
|
"grad_norm": 0.5787394046783447,
|
|
"learning_rate": 8.661417322834646e-07,
|
|
"loss": 0.8085,
|
|
"step": 2520
|
|
},
|
|
{
|
|
"epoch": 1.8399999999999999,
|
|
"grad_norm": 0.5817368030548096,
|
|
"learning_rate": 8.286464191976004e-07,
|
|
"loss": 0.7993,
|
|
"step": 2530
|
|
},
|
|
{
|
|
"epoch": 1.8472727272727272,
|
|
"grad_norm": 0.5391492247581482,
|
|
"learning_rate": 7.911511061117361e-07,
|
|
"loss": 0.8174,
|
|
"step": 2540
|
|
},
|
|
{
|
|
"epoch": 1.8545454545454545,
|
|
"grad_norm": 0.5950977206230164,
|
|
"learning_rate": 7.536557930258719e-07,
|
|
"loss": 0.8269,
|
|
"step": 2550
|
|
},
|
|
{
|
|
"epoch": 1.8618181818181818,
|
|
"grad_norm": 0.5649157166481018,
|
|
"learning_rate": 7.161604799400075e-07,
|
|
"loss": 0.8101,
|
|
"step": 2560
|
|
},
|
|
{
|
|
"epoch": 1.8690909090909091,
|
|
"grad_norm": 0.5596181154251099,
|
|
"learning_rate": 6.786651668541433e-07,
|
|
"loss": 0.8058,
|
|
"step": 2570
|
|
},
|
|
{
|
|
"epoch": 1.8763636363636365,
|
|
"grad_norm": 0.5372329354286194,
|
|
"learning_rate": 6.411698537682789e-07,
|
|
"loss": 0.8121,
|
|
"step": 2580
|
|
},
|
|
{
|
|
"epoch": 1.8836363636363638,
|
|
"grad_norm": 0.5739951133728027,
|
|
"learning_rate": 6.036745406824148e-07,
|
|
"loss": 0.8098,
|
|
"step": 2590
|
|
},
|
|
{
|
|
"epoch": 1.8909090909090909,
|
|
"grad_norm": 0.5760049223899841,
|
|
"learning_rate": 5.661792275965505e-07,
|
|
"loss": 0.813,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 1.8981818181818182,
|
|
"grad_norm": 0.5617313981056213,
|
|
"learning_rate": 5.286839145106862e-07,
|
|
"loss": 0.8019,
|
|
"step": 2610
|
|
},
|
|
{
|
|
"epoch": 1.9054545454545453,
|
|
"grad_norm": 0.5732679963111877,
|
|
"learning_rate": 4.91188601424822e-07,
|
|
"loss": 0.8049,
|
|
"step": 2620
|
|
},
|
|
{
|
|
"epoch": 1.9127272727272726,
|
|
"grad_norm": 0.5599384307861328,
|
|
"learning_rate": 4.536932883389577e-07,
|
|
"loss": 0.7894,
|
|
"step": 2630
|
|
},
|
|
{
|
|
"epoch": 1.92,
|
|
"grad_norm": 0.5590060353279114,
|
|
"learning_rate": 4.161979752530934e-07,
|
|
"loss": 0.8021,
|
|
"step": 2640
|
|
},
|
|
{
|
|
"epoch": 1.9272727272727272,
|
|
"grad_norm": 0.5524166822433472,
|
|
"learning_rate": 3.787026621672291e-07,
|
|
"loss": 0.7941,
|
|
"step": 2650
|
|
},
|
|
{
|
|
"epoch": 1.9345454545454546,
|
|
"grad_norm": 0.5761271715164185,
|
|
"learning_rate": 3.4120734908136486e-07,
|
|
"loss": 0.8243,
|
|
"step": 2660
|
|
},
|
|
{
|
|
"epoch": 1.9418181818181819,
|
|
"grad_norm": 0.5809080600738525,
|
|
"learning_rate": 3.037120359955006e-07,
|
|
"loss": 0.804,
|
|
"step": 2670
|
|
},
|
|
{
|
|
"epoch": 1.9490909090909092,
|
|
"grad_norm": 0.5772132873535156,
|
|
"learning_rate": 2.6621672290963634e-07,
|
|
"loss": 0.8194,
|
|
"step": 2680
|
|
},
|
|
{
|
|
"epoch": 1.9563636363636365,
|
|
"grad_norm": 0.5759870409965515,
|
|
"learning_rate": 2.2872140982377205e-07,
|
|
"loss": 0.822,
|
|
"step": 2690
|
|
},
|
|
{
|
|
"epoch": 1.9636363636363636,
|
|
"grad_norm": 0.5843390226364136,
|
|
"learning_rate": 1.9122609673790776e-07,
|
|
"loss": 0.8049,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 1.970909090909091,
|
|
"grad_norm": 0.5843693614006042,
|
|
"learning_rate": 1.537307836520435e-07,
|
|
"loss": 0.8169,
|
|
"step": 2710
|
|
},
|
|
{
|
|
"epoch": 1.978181818181818,
|
|
"grad_norm": 0.5962398052215576,
|
|
"learning_rate": 1.1623547056617925e-07,
|
|
"loss": 0.8102,
|
|
"step": 2720
|
|
},
|
|
{
|
|
"epoch": 1.9854545454545454,
|
|
"grad_norm": 0.5839393734931946,
|
|
"learning_rate": 7.874015748031497e-08,
|
|
"loss": 0.7998,
|
|
"step": 2730
|
|
},
|
|
{
|
|
"epoch": 1.9927272727272727,
|
|
"grad_norm": 0.5290048122406006,
|
|
"learning_rate": 4.1244844394450697e-08,
|
|
"loss": 0.7923,
|
|
"step": 2740
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"grad_norm": 0.581894040107727,
|
|
"learning_rate": 3.7495313085864276e-09,
|
|
"loss": 0.8008,
|
|
"step": 2750
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"step": 2750,
|
|
"total_flos": 8.418445674676224e+18,
|
|
"train_loss": 0.8579012258703058,
|
|
"train_runtime": 27047.1098,
|
|
"train_samples_per_second": 13.014,
|
|
"train_steps_per_second": 0.102
|
|
}
|
|
],
|
|
"logging_steps": 10,
|
|
"max_steps": 2750,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 2,
|
|
"save_steps": 1000,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 8.418445674676224e+18,
|
|
"train_batch_size": 8,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|