cap8-frq / trainer_state.json
Aerolandaz's picture
Upload folder using huggingface_hub
1274dcd verified
Raw
History Blame Contribute Delete
30.1 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.220063549704948,
"eval_steps": 500,
"global_step": 1680,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.007262823422605538,
"grad_norm": 0.2615480124950409,
"learning_rate": 8.571428571428573e-06,
"loss": 0.1552,
"step": 10
},
{
"epoch": 0.014525646845211076,
"grad_norm": 0.22226440906524658,
"learning_rate": 1.8095238095238094e-05,
"loss": 0.1592,
"step": 20
},
{
"epoch": 0.021788470267816613,
"grad_norm": 0.20973989367485046,
"learning_rate": 2.7619047619047622e-05,
"loss": 0.1534,
"step": 30
},
{
"epoch": 0.029051293690422152,
"grad_norm": 0.20205365121364594,
"learning_rate": 3.7142857142857143e-05,
"loss": 0.1634,
"step": 40
},
{
"epoch": 0.03631411711302769,
"grad_norm": 0.2007875144481659,
"learning_rate": 4.666666666666667e-05,
"loss": 0.1434,
"step": 50
},
{
"epoch": 0.04357694053563323,
"grad_norm": 0.20036202669143677,
"learning_rate": 5.619047619047619e-05,
"loss": 0.1333,
"step": 60
},
{
"epoch": 0.05083976395823876,
"grad_norm": 0.1990794837474823,
"learning_rate": 6.571428571428571e-05,
"loss": 0.1375,
"step": 70
},
{
"epoch": 0.058102587380844305,
"grad_norm": 0.19063813984394073,
"learning_rate": 7.523809523809524e-05,
"loss": 0.1374,
"step": 80
},
{
"epoch": 0.06536541080344985,
"grad_norm": 0.15606075525283813,
"learning_rate": 8.476190476190477e-05,
"loss": 0.132,
"step": 90
},
{
"epoch": 0.07262823422605538,
"grad_norm": 0.19317622482776642,
"learning_rate": 9.428571428571429e-05,
"loss": 0.1333,
"step": 100
},
{
"epoch": 0.07989105764866092,
"grad_norm": 0.17449212074279785,
"learning_rate": 9.999900808946995e-05,
"loss": 0.1392,
"step": 110
},
{
"epoch": 0.08715388107126645,
"grad_norm": 0.16662831604480743,
"learning_rate": 9.998784954797474e-05,
"loss": 0.1316,
"step": 120
},
{
"epoch": 0.09441670449387199,
"grad_norm": 0.13957402110099792,
"learning_rate": 9.996429535306638e-05,
"loss": 0.1249,
"step": 130
},
{
"epoch": 0.10167952791647752,
"grad_norm": 0.17535850405693054,
"learning_rate": 9.992835134555693e-05,
"loss": 0.1329,
"step": 140
},
{
"epoch": 0.10894235133908307,
"grad_norm": 0.14767180383205414,
"learning_rate": 9.988002643860162e-05,
"loss": 0.1256,
"step": 150
},
{
"epoch": 0.11620517476168861,
"grad_norm": 0.18588787317276,
"learning_rate": 9.981933261548842e-05,
"loss": 0.1453,
"step": 160
},
{
"epoch": 0.12346799818429414,
"grad_norm": 0.15257962048053741,
"learning_rate": 9.974628492666663e-05,
"loss": 0.1329,
"step": 170
},
{
"epoch": 0.1307308216068997,
"grad_norm": 0.1288495510816574,
"learning_rate": 9.966090148601476e-05,
"loss": 0.1331,
"step": 180
},
{
"epoch": 0.13799364502950523,
"grad_norm": 0.17879730463027954,
"learning_rate": 9.956320346634876e-05,
"loss": 0.1358,
"step": 190
},
{
"epoch": 0.14525646845211077,
"grad_norm": 0.12357795238494873,
"learning_rate": 9.945321509417169e-05,
"loss": 0.1252,
"step": 200
},
{
"epoch": 0.1525192918747163,
"grad_norm": 0.11386862397193909,
"learning_rate": 9.933096364366626e-05,
"loss": 0.1355,
"step": 210
},
{
"epoch": 0.15978211529732184,
"grad_norm": 0.11719153821468353,
"learning_rate": 9.919647942993148e-05,
"loss": 0.1275,
"step": 220
},
{
"epoch": 0.16704493871992737,
"grad_norm": 0.13855832815170288,
"learning_rate": 9.904979580146542e-05,
"loss": 0.132,
"step": 230
},
{
"epoch": 0.1743077621425329,
"grad_norm": 0.16087959706783295,
"learning_rate": 9.88909491318956e-05,
"loss": 0.1519,
"step": 240
},
{
"epoch": 0.18157058556513844,
"grad_norm": 0.1610720455646515,
"learning_rate": 9.87199788109593e-05,
"loss": 0.1383,
"step": 250
},
{
"epoch": 0.18883340898774398,
"grad_norm": 0.15942955017089844,
"learning_rate": 9.8536927234736e-05,
"loss": 0.1222,
"step": 260
},
{
"epoch": 0.1960962324103495,
"grad_norm": 0.1623896360397339,
"learning_rate": 9.834183979513427e-05,
"loss": 0.1286,
"step": 270
},
{
"epoch": 0.20335905583295505,
"grad_norm": 0.19065435230731964,
"learning_rate": 9.813476486863576e-05,
"loss": 0.1358,
"step": 280
},
{
"epoch": 0.2106218792555606,
"grad_norm": 0.14221009612083435,
"learning_rate": 9.791575380429912e-05,
"loss": 0.1331,
"step": 290
},
{
"epoch": 0.21788470267816615,
"grad_norm": 0.12954957783222198,
"learning_rate": 9.768486091102688e-05,
"loss": 0.1347,
"step": 300
},
{
"epoch": 0.22514752610077168,
"grad_norm": 0.15374745428562164,
"learning_rate": 9.744214344409819e-05,
"loss": 0.1278,
"step": 310
},
{
"epoch": 0.23241034952337722,
"grad_norm": 0.16374695301055908,
"learning_rate": 9.718766159097108e-05,
"loss": 0.1329,
"step": 320
},
{
"epoch": 0.23967317294598275,
"grad_norm": 0.13390475511550903,
"learning_rate": 9.692147845635761e-05,
"loss": 0.1254,
"step": 330
},
{
"epoch": 0.2469359963685883,
"grad_norm": 0.11325599998235703,
"learning_rate": 9.664366004657552e-05,
"loss": 0.1335,
"step": 340
},
{
"epoch": 0.25419881979119385,
"grad_norm": 0.1330733448266983,
"learning_rate": 9.635427525318048e-05,
"loss": 0.1429,
"step": 350
},
{
"epoch": 0.2614616432137994,
"grad_norm": 0.13730360567569733,
"learning_rate": 9.60533958358828e-05,
"loss": 0.125,
"step": 360
},
{
"epoch": 0.2687244666364049,
"grad_norm": 0.12945713102817535,
"learning_rate": 9.574109640475288e-05,
"loss": 0.1321,
"step": 370
},
{
"epoch": 0.27598729005901046,
"grad_norm": 0.11518649756908417,
"learning_rate": 9.541745440172005e-05,
"loss": 0.1308,
"step": 380
},
{
"epoch": 0.283250113481616,
"grad_norm": 0.18112903833389282,
"learning_rate": 9.508255008136885e-05,
"loss": 0.1323,
"step": 390
},
{
"epoch": 0.29051293690422153,
"grad_norm": 0.1332743912935257,
"learning_rate": 9.473646649103818e-05,
"loss": 0.1297,
"step": 400
},
{
"epoch": 0.29777576032682707,
"grad_norm": 0.14073874056339264,
"learning_rate": 9.437928945022771e-05,
"loss": 0.1171,
"step": 410
},
{
"epoch": 0.3050385837494326,
"grad_norm": 0.11776047945022583,
"learning_rate": 9.401110752931694e-05,
"loss": 0.129,
"step": 420
},
{
"epoch": 0.31230140717203814,
"grad_norm": 0.13818472623825073,
"learning_rate": 9.363201202760213e-05,
"loss": 0.1298,
"step": 430
},
{
"epoch": 0.3195642305946437,
"grad_norm": 0.14196033775806427,
"learning_rate": 9.324209695065645e-05,
"loss": 0.1184,
"step": 440
},
{
"epoch": 0.3268270540172492,
"grad_norm": 0.11807554960250854,
"learning_rate": 9.284145898701921e-05,
"loss": 0.1286,
"step": 450
},
{
"epoch": 0.33408987743985474,
"grad_norm": 0.13083140552043915,
"learning_rate": 9.243019748421955e-05,
"loss": 0.1475,
"step": 460
},
{
"epoch": 0.3413527008624603,
"grad_norm": 0.11743377894163132,
"learning_rate": 9.200841442414106e-05,
"loss": 0.127,
"step": 470
},
{
"epoch": 0.3486155242850658,
"grad_norm": 0.1455356925725937,
"learning_rate": 9.157621439773278e-05,
"loss": 0.1242,
"step": 480
},
{
"epoch": 0.35587834770767135,
"grad_norm": 0.13083097338676453,
"learning_rate": 9.113370457907368e-05,
"loss": 0.1246,
"step": 490
},
{
"epoch": 0.3631411711302769,
"grad_norm": 0.19750870764255524,
"learning_rate": 9.068099469879618e-05,
"loss": 0.1302,
"step": 500
},
{
"epoch": 0.3704039945528824,
"grad_norm": 0.1284891664981842,
"learning_rate": 9.02181970168759e-05,
"loss": 0.1233,
"step": 510
},
{
"epoch": 0.37766681797548796,
"grad_norm": 0.13317878544330597,
"learning_rate": 8.974542629479425e-05,
"loss": 0.1221,
"step": 520
},
{
"epoch": 0.3849296413980935,
"grad_norm": 0.11880248039960861,
"learning_rate": 8.926279976708055e-05,
"loss": 0.1244,
"step": 530
},
{
"epoch": 0.392192464820699,
"grad_norm": 0.13789477944374084,
"learning_rate": 8.877043711224108e-05,
"loss": 0.1263,
"step": 540
},
{
"epoch": 0.39945528824330456,
"grad_norm": 0.14878354966640472,
"learning_rate": 8.826846042308196e-05,
"loss": 0.1209,
"step": 550
},
{
"epoch": 0.4067181116659101,
"grad_norm": 0.18324986100196838,
"learning_rate": 8.775699417643338e-05,
"loss": 0.125,
"step": 560
},
{
"epoch": 0.41398093508851563,
"grad_norm": 0.13719946146011353,
"learning_rate": 8.723616520228271e-05,
"loss": 0.1376,
"step": 570
},
{
"epoch": 0.4212437585111212,
"grad_norm": 0.12926670908927917,
"learning_rate": 8.670610265232397e-05,
"loss": 0.1366,
"step": 580
},
{
"epoch": 0.42850658193372676,
"grad_norm": 0.11852458119392395,
"learning_rate": 8.616693796793179e-05,
"loss": 0.1378,
"step": 590
},
{
"epoch": 0.4357694053563323,
"grad_norm": 0.12678755819797516,
"learning_rate": 8.561880484756725e-05,
"loss": 0.1352,
"step": 600
},
{
"epoch": 0.44303222877893783,
"grad_norm": 0.14523130655288696,
"learning_rate": 8.506183921362443e-05,
"loss": 0.1332,
"step": 610
},
{
"epoch": 0.45029505220154337,
"grad_norm": 0.11939208209514618,
"learning_rate": 8.44961791787252e-05,
"loss": 0.133,
"step": 620
},
{
"epoch": 0.4575578756241489,
"grad_norm": 0.11850901693105698,
"learning_rate": 8.392196501147092e-05,
"loss": 0.1296,
"step": 630
},
{
"epoch": 0.46482069904675444,
"grad_norm": 0.18242543935775757,
"learning_rate": 8.333933910165963e-05,
"loss": 0.1294,
"step": 640
},
{
"epoch": 0.47208352246936,
"grad_norm": 0.11849816143512726,
"learning_rate": 8.274844592497719e-05,
"loss": 0.1272,
"step": 650
},
{
"epoch": 0.4793463458919655,
"grad_norm": 0.1463852971792221,
"learning_rate": 8.214943200717114e-05,
"loss": 0.1292,
"step": 660
},
{
"epoch": 0.48660916931457104,
"grad_norm": 0.11856704205274582,
"learning_rate": 8.154244588771621e-05,
"loss": 0.1251,
"step": 670
},
{
"epoch": 0.4938719927371766,
"grad_norm": 0.1263067126274109,
"learning_rate": 8.092763808298048e-05,
"loss": 0.1236,
"step": 680
},
{
"epoch": 0.5011348161597822,
"grad_norm": 0.14322146773338318,
"learning_rate": 8.03051610489014e-05,
"loss": 0.1244,
"step": 690
},
{
"epoch": 0.5083976395823877,
"grad_norm": 0.1455278843641281,
"learning_rate": 7.967516914318074e-05,
"loss": 0.1424,
"step": 700
},
{
"epoch": 0.5156604630049932,
"grad_norm": 0.13176371157169342,
"learning_rate": 7.903781858700799e-05,
"loss": 0.1298,
"step": 710
},
{
"epoch": 0.5229232864275988,
"grad_norm": 0.1257869303226471,
"learning_rate": 7.839326742632167e-05,
"loss": 0.1242,
"step": 720
},
{
"epoch": 0.5301861098502043,
"grad_norm": 0.11117416620254517,
"learning_rate": 7.774167549261817e-05,
"loss": 0.1182,
"step": 730
},
{
"epoch": 0.5374489332728098,
"grad_norm": 0.11336909979581833,
"learning_rate": 7.708320436331782e-05,
"loss": 0.1238,
"step": 740
},
{
"epoch": 0.5447117566954154,
"grad_norm": 0.12919354438781738,
"learning_rate": 7.641801732169795e-05,
"loss": 0.1328,
"step": 750
},
{
"epoch": 0.5519745801180209,
"grad_norm": 0.1134372279047966,
"learning_rate": 7.574627931640303e-05,
"loss": 0.1267,
"step": 760
},
{
"epoch": 0.5592374035406265,
"grad_norm": 0.14933618903160095,
"learning_rate": 7.50681569205418e-05,
"loss": 0.1274,
"step": 770
},
{
"epoch": 0.566500226963232,
"grad_norm": 0.16035373508930206,
"learning_rate": 7.438381829038156e-05,
"loss": 0.1238,
"step": 780
},
{
"epoch": 0.5737630503858375,
"grad_norm": 0.14108073711395264,
"learning_rate": 7.369343312364993e-05,
"loss": 0.1402,
"step": 790
},
{
"epoch": 0.5810258738084431,
"grad_norm": 0.12364397943019867,
"learning_rate": 7.299717261745434e-05,
"loss": 0.1256,
"step": 800
},
{
"epoch": 0.5882886972310486,
"grad_norm": 0.11245425045490265,
"learning_rate": 7.229520942582965e-05,
"loss": 0.1205,
"step": 810
},
{
"epoch": 0.5955515206536541,
"grad_norm": 0.15711691975593567,
"learning_rate": 7.158771761692464e-05,
"loss": 0.1268,
"step": 820
},
{
"epoch": 0.6028143440762597,
"grad_norm": 0.09989681094884872,
"learning_rate": 7.087487262983776e-05,
"loss": 0.1284,
"step": 830
},
{
"epoch": 0.6100771674988652,
"grad_norm": 0.11307062953710556,
"learning_rate": 7.015685123111276e-05,
"loss": 0.1284,
"step": 840
},
{
"epoch": 0.6173399909214707,
"grad_norm": 0.11042868345975876,
"learning_rate": 6.943383147090552e-05,
"loss": 0.1253,
"step": 850
},
{
"epoch": 0.6246028143440763,
"grad_norm": 0.12835900485515594,
"learning_rate": 6.870599263883219e-05,
"loss": 0.1187,
"step": 860
},
{
"epoch": 0.6318656377666818,
"grad_norm": 0.11985421180725098,
"learning_rate": 6.797351521951021e-05,
"loss": 0.1253,
"step": 870
},
{
"epoch": 0.6391284611892873,
"grad_norm": 0.13332399725914001,
"learning_rate": 6.723658084780297e-05,
"loss": 0.1147,
"step": 880
},
{
"epoch": 0.6463912846118929,
"grad_norm": 0.1253022998571396,
"learning_rate": 6.649537226377915e-05,
"loss": 0.1366,
"step": 890
},
{
"epoch": 0.6536541080344984,
"grad_norm": 0.1309010088443756,
"learning_rate": 6.575007326739809e-05,
"loss": 0.1249,
"step": 900
},
{
"epoch": 0.660916931457104,
"grad_norm": 0.1471216380596161,
"learning_rate": 6.50008686729323e-05,
"loss": 0.1252,
"step": 910
},
{
"epoch": 0.6681797548797095,
"grad_norm": 0.1386883407831192,
"learning_rate": 6.424794426313845e-05,
"loss": 0.1291,
"step": 920
},
{
"epoch": 0.675442578302315,
"grad_norm": 0.11485538631677628,
"learning_rate": 6.349148674318816e-05,
"loss": 0.1291,
"step": 930
},
{
"epoch": 0.6827054017249206,
"grad_norm": 0.13339963555335999,
"learning_rate": 6.273168369437018e-05,
"loss": 0.1257,
"step": 940
},
{
"epoch": 0.6899682251475261,
"grad_norm": 0.15178564190864563,
"learning_rate": 6.196872352757516e-05,
"loss": 0.1146,
"step": 950
},
{
"epoch": 0.6972310485701316,
"grad_norm": 0.1355406641960144,
"learning_rate": 6.12027954365748e-05,
"loss": 0.1339,
"step": 960
},
{
"epoch": 0.7044938719927372,
"grad_norm": 0.15000687539577484,
"learning_rate": 6.043408935110688e-05,
"loss": 0.1229,
"step": 970
},
{
"epoch": 0.7117566954153427,
"grad_norm": 0.12488489598035812,
"learning_rate": 5.9662795889777666e-05,
"loss": 0.1262,
"step": 980
},
{
"epoch": 0.7190195188379482,
"grad_norm": 0.11761627346277237,
"learning_rate": 5.888910631279365e-05,
"loss": 0.1216,
"step": 990
},
{
"epoch": 0.7262823422605538,
"grad_norm": 0.1294727474451065,
"learning_rate": 5.81132124745341e-05,
"loss": 0.1376,
"step": 1000
},
{
"epoch": 0.7335451656831593,
"grad_norm": 0.14842727780342102,
"learning_rate": 5.733530677597627e-05,
"loss": 0.1165,
"step": 1010
},
{
"epoch": 0.7408079891057648,
"grad_norm": 0.14372870326042175,
"learning_rate": 5.655558211698513e-05,
"loss": 0.1255,
"step": 1020
},
{
"epoch": 0.7480708125283704,
"grad_norm": 0.14958761632442474,
"learning_rate": 5.577423184847932e-05,
"loss": 0.1244,
"step": 1030
},
{
"epoch": 0.7553336359509759,
"grad_norm": 0.1290394514799118,
"learning_rate": 5.499144972448525e-05,
"loss": 0.1238,
"step": 1040
},
{
"epoch": 0.7625964593735814,
"grad_norm": 0.14663836359977722,
"learning_rate": 5.420742985409132e-05,
"loss": 0.1114,
"step": 1050
},
{
"epoch": 0.769859282796187,
"grad_norm": 0.11597703397274017,
"learning_rate": 5.342236665331407e-05,
"loss": 0.1282,
"step": 1060
},
{
"epoch": 0.7771221062187925,
"grad_norm": 0.1349128633737564,
"learning_rate": 5.2636454796888066e-05,
"loss": 0.1225,
"step": 1070
},
{
"epoch": 0.784384929641398,
"grad_norm": 0.11372572183609009,
"learning_rate": 5.1849889169991906e-05,
"loss": 0.1186,
"step": 1080
},
{
"epoch": 0.7916477530640036,
"grad_norm": 0.122071273624897,
"learning_rate": 5.1062864819921784e-05,
"loss": 0.1221,
"step": 1090
},
{
"epoch": 0.7989105764866091,
"grad_norm": 0.08935081958770752,
"learning_rate": 5.027557690772503e-05,
"loss": 0.1256,
"step": 1100
},
{
"epoch": 0.8061733999092147,
"grad_norm": 0.11327587813138962,
"learning_rate": 4.9488220659805326e-05,
"loss": 0.1239,
"step": 1110
},
{
"epoch": 0.8134362233318202,
"grad_norm": 0.10857807844877243,
"learning_rate": 4.870099131951185e-05,
"loss": 0.1238,
"step": 1120
},
{
"epoch": 0.8206990467544257,
"grad_norm": 0.12304297089576721,
"learning_rate": 4.791408409872398e-05,
"loss": 0.1183,
"step": 1130
},
{
"epoch": 0.8279618701770313,
"grad_norm": 0.1079617366194725,
"learning_rate": 4.7127694129444134e-05,
"loss": 0.124,
"step": 1140
},
{
"epoch": 0.8352246935996369,
"grad_norm": 0.12092699110507965,
"learning_rate": 4.634201641541012e-05,
"loss": 0.127,
"step": 1150
},
{
"epoch": 0.8424875170222424,
"grad_norm": 0.12485072761774063,
"learning_rate": 4.555724578373942e-05,
"loss": 0.1194,
"step": 1160
},
{
"epoch": 0.849750340444848,
"grad_norm": 0.12260372936725616,
"learning_rate": 4.477357683661734e-05,
"loss": 0.1279,
"step": 1170
},
{
"epoch": 0.8570131638674535,
"grad_norm": 0.1227007731795311,
"learning_rate": 4.399120390304072e-05,
"loss": 0.1223,
"step": 1180
},
{
"epoch": 0.8642759872900591,
"grad_norm": 0.11986847966909409,
"learning_rate": 4.3210320990629694e-05,
"loss": 0.1273,
"step": 1190
},
{
"epoch": 0.8715388107126646,
"grad_norm": 0.11769075691699982,
"learning_rate": 4.243112173751882e-05,
"loss": 0.1234,
"step": 1200
},
{
"epoch": 0.8788016341352701,
"grad_norm": 0.1076694205403328,
"learning_rate": 4.165379936434011e-05,
"loss": 0.124,
"step": 1210
},
{
"epoch": 0.8860644575578757,
"grad_norm": 0.13070237636566162,
"learning_rate": 4.087854662630937e-05,
"loss": 0.1178,
"step": 1220
},
{
"epoch": 0.8933272809804812,
"grad_norm": 0.09483371675014496,
"learning_rate": 4.0105555765428114e-05,
"loss": 0.1295,
"step": 1230
},
{
"epoch": 0.9005901044030867,
"grad_norm": 0.11234448850154877,
"learning_rate": 3.933501846281267e-05,
"loss": 0.121,
"step": 1240
},
{
"epoch": 0.9078529278256923,
"grad_norm": 0.14361174404621124,
"learning_rate": 3.856712579116229e-05,
"loss": 0.1158,
"step": 1250
},
{
"epoch": 0.9151157512482978,
"grad_norm": 0.10222570598125458,
"learning_rate": 3.780206816737837e-05,
"loss": 0.1152,
"step": 1260
},
{
"epoch": 0.9223785746709033,
"grad_norm": 0.12171214818954468,
"learning_rate": 3.704003530534597e-05,
"loss": 0.1319,
"step": 1270
},
{
"epoch": 0.9296413980935089,
"grad_norm": 0.11967138200998306,
"learning_rate": 3.628121616888999e-05,
"loss": 0.1173,
"step": 1280
},
{
"epoch": 0.9369042215161144,
"grad_norm": 0.12233515828847885,
"learning_rate": 3.5525798924917036e-05,
"loss": 0.1225,
"step": 1290
},
{
"epoch": 0.94416704493872,
"grad_norm": 0.13005629181861877,
"learning_rate": 3.4773970896755164e-05,
"loss": 0.135,
"step": 1300
},
{
"epoch": 0.9514298683613255,
"grad_norm": 0.12671716511249542,
"learning_rate": 3.40259185177026e-05,
"loss": 0.124,
"step": 1310
},
{
"epoch": 0.958692691783931,
"grad_norm": 0.1453058123588562,
"learning_rate": 3.3281827284797316e-05,
"loss": 0.1143,
"step": 1320
},
{
"epoch": 0.9659555152065366,
"grad_norm": 0.13116319477558136,
"learning_rate": 3.254188171281871e-05,
"loss": 0.1115,
"step": 1330
},
{
"epoch": 0.9732183386291421,
"grad_norm": 0.1564348042011261,
"learning_rate": 3.18062652885328e-05,
"loss": 0.1276,
"step": 1340
},
{
"epoch": 0.9804811620517476,
"grad_norm": 0.1428423672914505,
"learning_rate": 3.107516042519248e-05,
"loss": 0.1203,
"step": 1350
},
{
"epoch": 0.9877439854743532,
"grad_norm": 0.11770793795585632,
"learning_rate": 3.0348748417303823e-05,
"loss": 0.1234,
"step": 1360
},
{
"epoch": 0.9950068088969587,
"grad_norm": 0.13844764232635498,
"learning_rate": 2.9627209395669975e-05,
"loss": 0.1293,
"step": 1370
},
{
"epoch": 1.0021788470267816,
"grad_norm": 0.09861095994710922,
"learning_rate": 2.89107222827234e-05,
"loss": 0.1065,
"step": 1380
},
{
"epoch": 1.0094416704493872,
"grad_norm": 0.12454795092344284,
"learning_rate": 2.8199464748157983e-05,
"loss": 0.0996,
"step": 1390
},
{
"epoch": 1.0167044938719927,
"grad_norm": 0.13660405576229095,
"learning_rate": 2.749361316487168e-05,
"loss": 0.0999,
"step": 1400
},
{
"epoch": 1.0239673172945982,
"grad_norm": 0.14286108314990997,
"learning_rate": 2.6793342565230674e-05,
"loss": 0.1105,
"step": 1410
},
{
"epoch": 1.0312301407172038,
"grad_norm": 0.08573127537965775,
"learning_rate": 2.6098826597666047e-05,
"loss": 0.0958,
"step": 1420
},
{
"epoch": 1.0384929641398093,
"grad_norm": 0.12062367796897888,
"learning_rate": 2.5410237483613685e-05,
"loss": 0.0934,
"step": 1430
},
{
"epoch": 1.0457557875624148,
"grad_norm": 0.15126581490039825,
"learning_rate": 2.472774597480783e-05,
"loss": 0.1058,
"step": 1440
},
{
"epoch": 1.0530186109850204,
"grad_norm": 0.1533040553331375,
"learning_rate": 2.405152131093926e-05,
"loss": 0.0923,
"step": 1450
},
{
"epoch": 1.060281434407626,
"grad_norm": 0.13080339133739471,
"learning_rate": 2.3381731177688344e-05,
"loss": 0.1039,
"step": 1460
},
{
"epoch": 1.0675442578302314,
"grad_norm": 0.13630296289920807,
"learning_rate": 2.2718541665143543e-05,
"loss": 0.1039,
"step": 1470
},
{
"epoch": 1.074807081252837,
"grad_norm": 0.16583819687366486,
"learning_rate": 2.2062117226615374e-05,
"loss": 0.0922,
"step": 1480
},
{
"epoch": 1.0820699046754425,
"grad_norm": 0.09012618660926819,
"learning_rate": 2.1412620637856447e-05,
"loss": 0.1116,
"step": 1490
},
{
"epoch": 1.089332728098048,
"grad_norm": 0.11278939992189407,
"learning_rate": 2.077021295669743e-05,
"loss": 0.107,
"step": 1500
},
{
"epoch": 1.0965955515206536,
"grad_norm": 0.12911301851272583,
"learning_rate": 2.0135053483108973e-05,
"loss": 0.0943,
"step": 1510
},
{
"epoch": 1.1038583749432591,
"grad_norm": 0.1219969317317009,
"learning_rate": 1.950729971969955e-05,
"loss": 0.1022,
"step": 1520
},
{
"epoch": 1.1111211983658646,
"grad_norm": 0.13249413669109344,
"learning_rate": 1.8887107332659047e-05,
"loss": 0.097,
"step": 1530
},
{
"epoch": 1.1183840217884702,
"grad_norm": 0.12937743961811066,
"learning_rate": 1.8274630113157725e-05,
"loss": 0.0963,
"step": 1540
},
{
"epoch": 1.1256468452110757,
"grad_norm": 0.14856944978237152,
"learning_rate": 1.7670019939210024e-05,
"loss": 0.0973,
"step": 1550
},
{
"epoch": 1.1329096686336813,
"grad_norm": 0.11450286209583282,
"learning_rate": 1.707342673801294e-05,
"loss": 0.0978,
"step": 1560
},
{
"epoch": 1.1401724920562868,
"grad_norm": 0.1289551854133606,
"learning_rate": 1.648499844876802e-05,
"loss": 0.1036,
"step": 1570
},
{
"epoch": 1.1474353154788923,
"grad_norm": 0.12307775765657425,
"learning_rate": 1.59048809859965e-05,
"loss": 0.0973,
"step": 1580
},
{
"epoch": 1.1546981389014979,
"grad_norm": 0.13354001939296722,
"learning_rate": 1.5333218203356243e-05,
"loss": 0.1044,
"step": 1590
},
{
"epoch": 1.1619609623241036,
"grad_norm": 0.11185191571712494,
"learning_rate": 1.4770151857969949e-05,
"loss": 0.104,
"step": 1600
},
{
"epoch": 1.1692237857467092,
"grad_norm": 0.1212216317653656,
"learning_rate": 1.421582157527322e-05,
"loss": 0.113,
"step": 1610
},
{
"epoch": 1.1764866091693147,
"grad_norm": 0.11214473098516464,
"learning_rate": 1.3670364814391062e-05,
"loss": 0.097,
"step": 1620
},
{
"epoch": 1.1837494325919202,
"grad_norm": 0.09202374517917633,
"learning_rate": 1.3133916834051768e-05,
"loss": 0.0991,
"step": 1630
},
{
"epoch": 1.1910122560145258,
"grad_norm": 0.11675283312797546,
"learning_rate": 1.2606610659046314e-05,
"loss": 0.105,
"step": 1640
},
{
"epoch": 1.1982750794371313,
"grad_norm": 0.1229914203286171,
"learning_rate": 1.2088577047241833e-05,
"loss": 0.1,
"step": 1650
},
{
"epoch": 1.2055379028597368,
"grad_norm": 0.09079571068286896,
"learning_rate": 1.157994445715706e-05,
"loss": 0.1026,
"step": 1660
},
{
"epoch": 1.2128007262823424,
"grad_norm": 0.12256491184234619,
"learning_rate": 1.1080839016108086e-05,
"loss": 0.0934,
"step": 1670
},
{
"epoch": 1.220063549704948,
"grad_norm": 0.11560764908790588,
"learning_rate": 1.0591384488932187e-05,
"loss": 0.1008,
"step": 1680
}
],
"logging_steps": 10,
"max_steps": 2100,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 420,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.5179406469592402e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}