793 lines
19 KiB
JSON
793 lines
19 KiB
JSON
|
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 2.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 1078,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.01855287569573284,
|
||
|
|
"grad_norm": 60.09148197149517,
|
||
|
|
"learning_rate": 8.333333333333333e-07,
|
||
|
|
"loss": 2.9472,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03710575139146568,
|
||
|
|
"grad_norm": 38.855704459472975,
|
||
|
|
"learning_rate": 1.7592592592592594e-06,
|
||
|
|
"loss": 2.6542,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.055658627087198514,
|
||
|
|
"grad_norm": 5.063334237382181,
|
||
|
|
"learning_rate": 2.6851851851851856e-06,
|
||
|
|
"loss": 1.5155,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07421150278293136,
|
||
|
|
"grad_norm": 2.9672953129926576,
|
||
|
|
"learning_rate": 3.6111111111111115e-06,
|
||
|
|
"loss": 0.9282,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09276437847866419,
|
||
|
|
"grad_norm": 1.8537537673527857,
|
||
|
|
"learning_rate": 4.537037037037038e-06,
|
||
|
|
"loss": 0.8926,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11131725417439703,
|
||
|
|
"grad_norm": 3.428912896471263,
|
||
|
|
"learning_rate": 5.462962962962963e-06,
|
||
|
|
"loss": 0.6386,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12987012987012986,
|
||
|
|
"grad_norm": 5.261034391804638,
|
||
|
|
"learning_rate": 6.3888888888888885e-06,
|
||
|
|
"loss": 0.4684,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14842300556586271,
|
||
|
|
"grad_norm": 1.3915033824673388,
|
||
|
|
"learning_rate": 7.314814814814816e-06,
|
||
|
|
"loss": 0.3764,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16697588126159554,
|
||
|
|
"grad_norm": 1.3167457631005255,
|
||
|
|
"learning_rate": 8.240740740740741e-06,
|
||
|
|
"loss": 0.3314,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18552875695732837,
|
||
|
|
"grad_norm": 1.1616342114119869,
|
||
|
|
"learning_rate": 9.166666666666666e-06,
|
||
|
|
"loss": 0.3162,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20408163265306123,
|
||
|
|
"grad_norm": 1.6904372750211418,
|
||
|
|
"learning_rate": 9.99997377618298e-06,
|
||
|
|
"loss": 0.32,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22263450834879406,
|
||
|
|
"grad_norm": 1.6370524652439618,
|
||
|
|
"learning_rate": 9.99682725096819e-06,
|
||
|
|
"loss": 0.2867,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24118738404452691,
|
||
|
|
"grad_norm": 0.9996394242016245,
|
||
|
|
"learning_rate": 9.988439743984155e-06,
|
||
|
|
"loss": 0.2798,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2597402597402597,
|
||
|
|
"grad_norm": 1.105275657101106,
|
||
|
|
"learning_rate": 9.97482005256746e-06,
|
||
|
|
"loss": 0.285,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2782931354359926,
|
||
|
|
"grad_norm": 1.0656580706685925,
|
||
|
|
"learning_rate": 9.955982461893648e-06,
|
||
|
|
"loss": 0.2687,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29684601113172543,
|
||
|
|
"grad_norm": 1.4055251928327075,
|
||
|
|
"learning_rate": 9.93194672999403e-06,
|
||
|
|
"loss": 0.274,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31539888682745826,
|
||
|
|
"grad_norm": 0.955521719488787,
|
||
|
|
"learning_rate": 9.902738067032254e-06,
|
||
|
|
"loss": 0.2595,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3339517625231911,
|
||
|
|
"grad_norm": 2.7989567654700673,
|
||
|
|
"learning_rate": 9.868387108862307e-06,
|
||
|
|
"loss": 0.2503,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3525046382189239,
|
||
|
|
"grad_norm": 1.1080193292817926,
|
||
|
|
"learning_rate": 9.828929884895753e-06,
|
||
|
|
"loss": 0.2416,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37105751391465674,
|
||
|
|
"grad_norm": 0.8932615001563428,
|
||
|
|
"learning_rate": 9.784407780311845e-06,
|
||
|
|
"loss": 0.2428,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38961038961038963,
|
||
|
|
"grad_norm": 1.2146643079836965,
|
||
|
|
"learning_rate": 9.734867492650187e-06,
|
||
|
|
"loss": 0.2364,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.40816326530612246,
|
||
|
|
"grad_norm": 1.145275092602219,
|
||
|
|
"learning_rate": 9.680360982831467e-06,
|
||
|
|
"loss": 0.2384,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4267161410018553,
|
||
|
|
"grad_norm": 0.7358756050937082,
|
||
|
|
"learning_rate": 9.620945420657625e-06,
|
||
|
|
"loss": 0.2325,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4452690166975881,
|
||
|
|
"grad_norm": 0.9469502291509642,
|
||
|
|
"learning_rate": 9.556683124848624e-06,
|
||
|
|
"loss": 0.2415,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.46382189239332094,
|
||
|
|
"grad_norm": 0.8629958324420408,
|
||
|
|
"learning_rate": 9.487641497678724e-06,
|
||
|
|
"loss": 0.2296,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48237476808905383,
|
||
|
|
"grad_norm": 0.9672085960501411,
|
||
|
|
"learning_rate": 9.413892954280793e-06,
|
||
|
|
"loss": 0.2318,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5009276437847866,
|
||
|
|
"grad_norm": 0.8366109170992587,
|
||
|
|
"learning_rate": 9.335514846692846e-06,
|
||
|
|
"loss": 0.2425,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5194805194805194,
|
||
|
|
"grad_norm": 0.8555696459778686,
|
||
|
|
"learning_rate": 9.252589382726426e-06,
|
||
|
|
"loss": 0.2235,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5380333951762524,
|
||
|
|
"grad_norm": 0.65154561944275,
|
||
|
|
"learning_rate": 9.165203539741976e-06,
|
||
|
|
"loss": 0.2171,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5565862708719852,
|
||
|
|
"grad_norm": 0.8174459194486984,
|
||
|
|
"learning_rate": 9.073448973421581e-06,
|
||
|
|
"loss": 0.2393,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.575139146567718,
|
||
|
|
"grad_norm": 1.179097990857863,
|
||
|
|
"learning_rate": 8.977421921634833e-06,
|
||
|
|
"loss": 0.2276,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5936920222634509,
|
||
|
|
"grad_norm": 1.164135161117637,
|
||
|
|
"learning_rate": 8.877223103498576e-06,
|
||
|
|
"loss": 0.2232,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6122448979591837,
|
||
|
|
"grad_norm": 0.9194559443120759,
|
||
|
|
"learning_rate": 8.772957613736483e-06,
|
||
|
|
"loss": 0.208,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6307977736549165,
|
||
|
|
"grad_norm": 0.8725673809937737,
|
||
|
|
"learning_rate": 8.66473481244918e-06,
|
||
|
|
"loss": 0.2184,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6493506493506493,
|
||
|
|
"grad_norm": 0.7606564604713997,
|
||
|
|
"learning_rate": 8.552668210410624e-06,
|
||
|
|
"loss": 0.2322,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6679035250463822,
|
||
|
|
"grad_norm": 1.7475773499685987,
|
||
|
|
"learning_rate": 8.436875350010958e-06,
|
||
|
|
"loss": 0.2252,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.686456400742115,
|
||
|
|
"grad_norm": 0.8902082858576856,
|
||
|
|
"learning_rate": 8.317477681970786e-06,
|
||
|
|
"loss": 0.2118,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7050092764378478,
|
||
|
|
"grad_norm": 0.6756559134494782,
|
||
|
|
"learning_rate": 8.19460043795614e-06,
|
||
|
|
"loss": 0.2286,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7235621521335807,
|
||
|
|
"grad_norm": 0.7753939520244841,
|
||
|
|
"learning_rate": 8.068372499227738e-06,
|
||
|
|
"loss": 0.226,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7421150278293135,
|
||
|
|
"grad_norm": 0.8189644955934055,
|
||
|
|
"learning_rate": 7.938926261462366e-06,
|
||
|
|
"loss": 0.2172,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7606679035250464,
|
||
|
|
"grad_norm": 1.2227576197844174,
|
||
|
|
"learning_rate": 7.806397495888074e-06,
|
||
|
|
"loss": 0.2282,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7792207792207793,
|
||
|
|
"grad_norm": 0.7133788769555174,
|
||
|
|
"learning_rate": 7.670925206878917e-06,
|
||
|
|
"loss": 0.2174,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7977736549165121,
|
||
|
|
"grad_norm": 0.767835676921408,
|
||
|
|
"learning_rate": 7.532651486158554e-06,
|
||
|
|
"loss": 0.2107,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8163265306122449,
|
||
|
|
"grad_norm": 2.48600742003163,
|
||
|
|
"learning_rate": 7.391721363765664e-06,
|
||
|
|
"loss": 0.2123,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8348794063079777,
|
||
|
|
"grad_norm": 0.6665575100282729,
|
||
|
|
"learning_rate": 7.248282655937451e-06,
|
||
|
|
"loss": 0.2159,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8534322820037106,
|
||
|
|
"grad_norm": 1.2649819297037939,
|
||
|
|
"learning_rate": 7.102485810070824e-06,
|
||
|
|
"loss": 0.2213,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8719851576994434,
|
||
|
|
"grad_norm": 1.237103957840642,
|
||
|
|
"learning_rate": 6.954483746923865e-06,
|
||
|
|
"loss": 0.2143,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8905380333951762,
|
||
|
|
"grad_norm": 1.1721776707126947,
|
||
|
|
"learning_rate": 6.804431700223057e-06,
|
||
|
|
"loss": 0.2156,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9090909090909091,
|
||
|
|
"grad_norm": 1.0611712427213547,
|
||
|
|
"learning_rate": 6.652487053844544e-06,
|
||
|
|
"loss": 0.2191,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9276437847866419,
|
||
|
|
"grad_norm": 0.7509649989043815,
|
||
|
|
"learning_rate": 6.49880917674019e-06,
|
||
|
|
"loss": 0.208,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9461966604823747,
|
||
|
|
"grad_norm": 1.9679575531610882,
|
||
|
|
"learning_rate": 6.343559255781538e-06,
|
||
|
|
"loss": 0.2045,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9647495361781077,
|
||
|
|
"grad_norm": 0.9410395170594658,
|
||
|
|
"learning_rate": 6.18690012669705e-06,
|
||
|
|
"loss": 0.1979,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9833024118738405,
|
||
|
|
"grad_norm": 0.7001206344622368,
|
||
|
|
"learning_rate": 6.028996103279918e-06,
|
||
|
|
"loss": 0.2027,
|
||
|
|
"step": 530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0018552875695732,
|
||
|
|
"grad_norm": 1.0171159957483642,
|
||
|
|
"learning_rate": 5.87001280504558e-06,
|
||
|
|
"loss": 0.213,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0204081632653061,
|
||
|
|
"grad_norm": 1.0453391912762757,
|
||
|
|
"learning_rate": 5.7101169835197115e-06,
|
||
|
|
"loss": 0.14,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0389610389610389,
|
||
|
|
"grad_norm": 0.8392060229822733,
|
||
|
|
"learning_rate": 5.549476347338915e-06,
|
||
|
|
"loss": 0.1305,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0575139146567718,
|
||
|
|
"grad_norm": 0.8088617971203363,
|
||
|
|
"learning_rate": 5.388259386347518e-06,
|
||
|
|
"loss": 0.132,
|
||
|
|
"step": 570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0760667903525047,
|
||
|
|
"grad_norm": 1.443734395132653,
|
||
|
|
"learning_rate": 5.226635194874978e-06,
|
||
|
|
"loss": 0.1371,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0946196660482375,
|
||
|
|
"grad_norm": 0.8920531372659211,
|
||
|
|
"learning_rate": 5.064773294379302e-06,
|
||
|
|
"loss": 0.1292,
|
||
|
|
"step": 590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1131725417439704,
|
||
|
|
"grad_norm": 1.0600375600094234,
|
||
|
|
"learning_rate": 4.9028434556424335e-06,
|
||
|
|
"loss": 0.1366,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1317254174397031,
|
||
|
|
"grad_norm": 0.6174457523927003,
|
||
|
|
"learning_rate": 4.741015520704148e-06,
|
||
|
|
"loss": 0.1461,
|
||
|
|
"step": 610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.150278293135436,
|
||
|
|
"grad_norm": 0.7413719867170724,
|
||
|
|
"learning_rate": 4.579459224721212e-06,
|
||
|
|
"loss": 0.125,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1688311688311688,
|
||
|
|
"grad_norm": 0.7428769835248965,
|
||
|
|
"learning_rate": 4.418344017938634e-06,
|
||
|
|
"loss": 0.1308,
|
||
|
|
"step": 630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1873840445269017,
|
||
|
|
"grad_norm": 1.1554971703426302,
|
||
|
|
"learning_rate": 4.257838887959764e-06,
|
||
|
|
"loss": 0.1405,
|
||
|
|
"step": 640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2059369202226344,
|
||
|
|
"grad_norm": 0.9187379299508138,
|
||
|
|
"learning_rate": 4.098112182501633e-06,
|
||
|
|
"loss": 0.1306,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2244897959183674,
|
||
|
|
"grad_norm": 0.7989292570187497,
|
||
|
|
"learning_rate": 3.93933143282144e-06,
|
||
|
|
"loss": 0.1319,
|
||
|
|
"step": 660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2430426716141,
|
||
|
|
"grad_norm": 0.8037568461641328,
|
||
|
|
"learning_rate": 3.7816631779994018e-06,
|
||
|
|
"loss": 0.1367,
|
||
|
|
"step": 670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.261595547309833,
|
||
|
|
"grad_norm": 0.7347468570406386,
|
||
|
|
"learning_rate": 3.6252727902622575e-06,
|
||
|
|
"loss": 0.1227,
|
||
|
|
"step": 680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2801484230055657,
|
||
|
|
"grad_norm": 1.144050562175369,
|
||
|
|
"learning_rate": 3.4703243015306314e-06,
|
||
|
|
"loss": 0.1382,
|
||
|
|
"step": 690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2987012987012987,
|
||
|
|
"grad_norm": 0.7989428953338974,
|
||
|
|
"learning_rate": 3.3169802313722076e-06,
|
||
|
|
"loss": 0.1444,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3172541743970316,
|
||
|
|
"grad_norm": 0.616665850742756,
|
||
|
|
"learning_rate": 3.16540141654114e-06,
|
||
|
|
"loss": 0.1301,
|
||
|
|
"step": 710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3358070500927643,
|
||
|
|
"grad_norm": 0.6664134211210629,
|
||
|
|
"learning_rate": 3.0157468422825148e-06,
|
||
|
|
"loss": 0.1283,
|
||
|
|
"step": 720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3543599257884973,
|
||
|
|
"grad_norm": 0.5929267472787093,
|
||
|
|
"learning_rate": 2.868173475578772e-06,
|
||
|
|
"loss": 0.1223,
|
||
|
|
"step": 730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.37291280148423,
|
||
|
|
"grad_norm": 1.1909639489199855,
|
||
|
|
"learning_rate": 2.722836100513027e-06,
|
||
|
|
"loss": 0.1471,
|
||
|
|
"step": 740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.391465677179963,
|
||
|
|
"grad_norm": 0.7917032964998825,
|
||
|
|
"learning_rate": 2.5798871559219362e-06,
|
||
|
|
"loss": 0.1287,
|
||
|
|
"step": 750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4100185528756957,
|
||
|
|
"grad_norm": 0.7775412493202811,
|
||
|
|
"learning_rate": 2.439476575508408e-06,
|
||
|
|
"loss": 0.1341,
|
||
|
|
"step": 760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4285714285714286,
|
||
|
|
"grad_norm": 0.6871081961127913,
|
||
|
|
"learning_rate": 2.301751630581855e-06,
|
||
|
|
"loss": 0.1256,
|
||
|
|
"step": 770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4471243042671613,
|
||
|
|
"grad_norm": 3.819276268656922,
|
||
|
|
"learning_rate": 2.1668567755909257e-06,
|
||
|
|
"loss": 0.1245,
|
||
|
|
"step": 780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4656771799628943,
|
||
|
|
"grad_norm": 1.5879388915114736,
|
||
|
|
"learning_rate": 2.0349334966107363e-06,
|
||
|
|
"loss": 0.1451,
|
||
|
|
"step": 790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4842300556586272,
|
||
|
|
"grad_norm": 1.0732264147107038,
|
||
|
|
"learning_rate": 1.906120162943515e-06,
|
||
|
|
"loss": 0.1339,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.50278293135436,
|
||
|
|
"grad_norm": 0.9759155260672688,
|
||
|
|
"learning_rate": 1.7805518819883134e-06,
|
||
|
|
"loss": 0.1271,
|
||
|
|
"step": 810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5213358070500926,
|
||
|
|
"grad_norm": 0.8313050751146777,
|
||
|
|
"learning_rate": 1.6583603575320002e-06,
|
||
|
|
"loss": 0.1262,
|
||
|
|
"step": 820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5398886827458256,
|
||
|
|
"grad_norm": 0.6978608117394565,
|
||
|
|
"learning_rate": 1.5396737516101757e-06,
|
||
|
|
"loss": 0.1244,
|
||
|
|
"step": 830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5584415584415585,
|
||
|
|
"grad_norm": 1.5257058397579806,
|
||
|
|
"learning_rate": 1.4246165500828974e-06,
|
||
|
|
"loss": 0.1266,
|
||
|
|
"step": 840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5769944341372912,
|
||
|
|
"grad_norm": 0.7687999727178765,
|
||
|
|
"learning_rate": 1.3133094320662e-06,
|
||
|
|
"loss": 0.1298,
|
||
|
|
"step": 850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.595547309833024,
|
||
|
|
"grad_norm": 0.9443564999213552,
|
||
|
|
"learning_rate": 1.2058691433563675e-06,
|
||
|
|
"loss": 0.13,
|
||
|
|
"step": 860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.614100185528757,
|
||
|
|
"grad_norm": 1.455717508617582,
|
||
|
|
"learning_rate": 1.102408373979717e-06,
|
||
|
|
"loss": 0.1228,
|
||
|
|
"step": 870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6326530612244898,
|
||
|
|
"grad_norm": 0.8425010133335726,
|
||
|
|
"learning_rate": 1.0030356399963204e-06,
|
||
|
|
"loss": 0.1214,
|
||
|
|
"step": 880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6512059369202228,
|
||
|
|
"grad_norm": 1.4449024332670837,
|
||
|
|
"learning_rate": 9.078551696816434e-07,
|
||
|
|
"loss": 0.1287,
|
||
|
|
"step": 890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6697588126159555,
|
||
|
|
"grad_norm": 0.8028839534623334,
|
||
|
|
"learning_rate": 8.16966794205476e-07,
|
||
|
|
"loss": 0.1276,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6883116883116882,
|
||
|
|
"grad_norm": 0.6958879285582347,
|
||
|
|
"learning_rate": 7.304658429228245e-07,
|
||
|
|
"loss": 0.1314,
|
||
|
|
"step": 910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7068645640074211,
|
||
|
|
"grad_norm": 0.7638017132226432,
|
||
|
|
"learning_rate": 6.484430433865785e-07,
|
||
|
|
"loss": 0.1267,
|
||
|
|
"step": 920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.725417439703154,
|
||
|
|
"grad_norm": 0.5637674831928584,
|
||
|
|
"learning_rate": 5.709844261868381e-07,
|
||
|
|
"loss": 0.1251,
|
||
|
|
"step": 930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7439703153988868,
|
||
|
|
"grad_norm": 0.8835722824389809,
|
||
|
|
"learning_rate": 4.981712347167061e-07,
|
||
|
|
"loss": 0.1263,
|
||
|
|
"step": 940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7625231910946195,
|
||
|
|
"grad_norm": 1.0327912071957213,
|
||
|
|
"learning_rate": 4.30079839959186e-07,
|
||
|
|
"loss": 0.1267,
|
||
|
|
"step": 950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7810760667903525,
|
||
|
|
"grad_norm": 0.760831900859538,
|
||
|
|
"learning_rate": 3.667816603845681e-07,
|
||
|
|
"loss": 0.1233,
|
||
|
|
"step": 960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7996289424860854,
|
||
|
|
"grad_norm": 0.7886002974401526,
|
||
|
|
"learning_rate": 3.0834308704231485e-07,
|
||
|
|
"loss": 0.1196,
|
||
|
|
"step": 970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8181818181818183,
|
||
|
|
"grad_norm": 0.7691394163831202,
|
||
|
|
"learning_rate": 2.5482541392601924e-07,
|
||
|
|
"loss": 0.1147,
|
||
|
|
"step": 980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.836734693877551,
|
||
|
|
"grad_norm": 1.3655213319729242,
|
||
|
|
"learning_rate": 2.062847736844703e-07,
|
||
|
|
"loss": 0.1311,
|
||
|
|
"step": 990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8552875695732838,
|
||
|
|
"grad_norm": 0.9020502820778742,
|
||
|
|
"learning_rate": 1.6277207874625444e-07,
|
||
|
|
"loss": 0.1326,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8738404452690167,
|
||
|
|
"grad_norm": 1.1891039699760202,
|
||
|
|
"learning_rate": 1.2433296791964754e-07,
|
||
|
|
"loss": 0.1203,
|
||
|
|
"step": 1010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8923933209647497,
|
||
|
|
"grad_norm": 0.6855058516961975,
|
||
|
|
"learning_rate": 9.100775852381227e-08,
|
||
|
|
"loss": 0.1162,
|
||
|
|
"step": 1020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9109461966604824,
|
||
|
|
"grad_norm": 0.9669643922429069,
|
||
|
|
"learning_rate": 6.283140410149213e-08,
|
||
|
|
"loss": 0.1174,
|
||
|
|
"step": 1030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.929499072356215,
|
||
|
|
"grad_norm": 0.94328699549889,
|
||
|
|
"learning_rate": 3.9833457757572636e-08,
|
||
|
|
"loss": 0.1362,
|
||
|
|
"step": 1040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.948051948051948,
|
||
|
|
"grad_norm": 0.884695843246741,
|
||
|
|
"learning_rate": 2.2038041161960288e-08,
|
||
|
|
"loss": 0.1215,
|
||
|
|
"step": 1050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.966604823747681,
|
||
|
|
"grad_norm": 0.5985819936855686,
|
||
|
|
"learning_rate": 9.463819249275751e-09,
|
||
|
|
"loss": 0.1181,
|
||
|
|
"step": 1060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9851576994434137,
|
||
|
|
"grad_norm": 0.6657018316360679,
|
||
|
|
"learning_rate": 2.1239806419176556e-09,
|
||
|
|
"loss": 0.1224,
|
||
|
|
"step": 1070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"step": 1078,
|
||
|
|
"total_flos": 456031784665088.0,
|
||
|
|
"train_loss": 0.26182658547373117,
|
||
|
|
"train_runtime": 16760.0523,
|
||
|
|
"train_samples_per_second": 0.386,
|
||
|
|
"train_steps_per_second": 0.064
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 10,
|
||
|
|
"max_steps": 1078,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 2,
|
||
|
|
"save_steps": 10000,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 456031784665088.0,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|