groot-bs32-split / checkpoint-2000 /trainer_state.json
LucaFrat's picture
Upload folder using huggingface_hub
15f43aa verified
Raw
History Blame Contribute Delete
27.8 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.1,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"grad_norm": 0.16940702497959137,
"learning_rate": 9e-07,
"loss": 1.208,
"step": 10
},
{
"grad_norm": 0.14846493303775787,
"learning_rate": 1.9e-06,
"loss": 1.207,
"step": 20
},
{
"grad_norm": 0.1575060486793518,
"learning_rate": 2.9e-06,
"loss": 1.2031,
"step": 30
},
{
"grad_norm": 0.18646641075611115,
"learning_rate": 3.9e-06,
"loss": 1.19,
"step": 40
},
{
"grad_norm": 0.30270645022392273,
"learning_rate": 4.9000000000000005e-06,
"loss": 1.1729,
"step": 50
},
{
"grad_norm": 0.39245566725730896,
"learning_rate": 5.9e-06,
"loss": 1.1582,
"step": 60
},
{
"grad_norm": 0.44334590435028076,
"learning_rate": 6.900000000000001e-06,
"loss": 1.1293,
"step": 70
},
{
"grad_norm": 0.30264705419540405,
"learning_rate": 7.9e-06,
"loss": 1.1102,
"step": 80
},
{
"grad_norm": 0.24947984516620636,
"learning_rate": 8.9e-06,
"loss": 1.1029,
"step": 90
},
{
"grad_norm": 0.3442651927471161,
"learning_rate": 9.900000000000002e-06,
"loss": 1.0967,
"step": 100
},
{
"grad_norm": 0.2968611717224121,
"learning_rate": 1.09e-05,
"loss": 1.092,
"step": 110
},
{
"grad_norm": 0.8635988235473633,
"learning_rate": 1.19e-05,
"loss": 1.0971,
"step": 120
},
{
"grad_norm": 0.549736738204956,
"learning_rate": 1.29e-05,
"loss": 1.0877,
"step": 130
},
{
"grad_norm": 0.3879706859588623,
"learning_rate": 1.3900000000000002e-05,
"loss": 1.0965,
"step": 140
},
{
"grad_norm": 0.8023251295089722,
"learning_rate": 1.49e-05,
"loss": 1.0902,
"step": 150
},
{
"grad_norm": 0.3089507222175598,
"learning_rate": 1.59e-05,
"loss": 1.0938,
"step": 160
},
{
"grad_norm": 0.2845019996166229,
"learning_rate": 1.69e-05,
"loss": 1.0857,
"step": 170
},
{
"grad_norm": 0.5286039710044861,
"learning_rate": 1.79e-05,
"loss": 1.0811,
"step": 180
},
{
"grad_norm": 0.3651806712150574,
"learning_rate": 1.8900000000000002e-05,
"loss": 1.0711,
"step": 190
},
{
"grad_norm": 0.49551165103912354,
"learning_rate": 1.9900000000000003e-05,
"loss": 1.0568,
"step": 200
},
{
"grad_norm": 0.48400798439979553,
"learning_rate": 2.09e-05,
"loss": 1.0261,
"step": 210
},
{
"grad_norm": 0.5731498599052429,
"learning_rate": 2.19e-05,
"loss": 1.0099,
"step": 220
},
{
"grad_norm": 0.5158259868621826,
"learning_rate": 2.29e-05,
"loss": 0.9947,
"step": 230
},
{
"grad_norm": 0.8482366800308228,
"learning_rate": 2.39e-05,
"loss": 0.9701,
"step": 240
},
{
"grad_norm": 0.668360710144043,
"learning_rate": 2.4900000000000002e-05,
"loss": 0.933,
"step": 250
},
{
"grad_norm": 1.0341784954071045,
"learning_rate": 2.5900000000000003e-05,
"loss": 0.9101,
"step": 260
},
{
"grad_norm": 0.8576267957687378,
"learning_rate": 2.6900000000000003e-05,
"loss": 0.8772,
"step": 270
},
{
"grad_norm": 1.177884578704834,
"learning_rate": 2.7900000000000004e-05,
"loss": 0.8447,
"step": 280
},
{
"grad_norm": 1.2616709470748901,
"learning_rate": 2.8899999999999998e-05,
"loss": 0.8277,
"step": 290
},
{
"grad_norm": 0.9310820698738098,
"learning_rate": 2.9900000000000002e-05,
"loss": 0.8179,
"step": 300
},
{
"grad_norm": 0.9291635751724243,
"learning_rate": 3.09e-05,
"loss": 0.7966,
"step": 310
},
{
"grad_norm": 0.9610940217971802,
"learning_rate": 3.19e-05,
"loss": 0.7595,
"step": 320
},
{
"grad_norm": 1.082502841949463,
"learning_rate": 3.29e-05,
"loss": 0.7442,
"step": 330
},
{
"grad_norm": 1.0246247053146362,
"learning_rate": 3.3900000000000004e-05,
"loss": 0.7153,
"step": 340
},
{
"grad_norm": 1.1535388231277466,
"learning_rate": 3.49e-05,
"loss": 0.7008,
"step": 350
},
{
"grad_norm": 1.1344460248947144,
"learning_rate": 3.59e-05,
"loss": 0.6781,
"step": 360
},
{
"grad_norm": 1.0874427556991577,
"learning_rate": 3.69e-05,
"loss": 0.6773,
"step": 370
},
{
"grad_norm": 1.1591013669967651,
"learning_rate": 3.79e-05,
"loss": 0.6567,
"step": 380
},
{
"grad_norm": 1.2492725849151611,
"learning_rate": 3.8900000000000004e-05,
"loss": 0.6557,
"step": 390
},
{
"grad_norm": 1.333844542503357,
"learning_rate": 3.99e-05,
"loss": 0.6557,
"step": 400
},
{
"grad_norm": 1.157532811164856,
"learning_rate": 4.09e-05,
"loss": 0.6327,
"step": 410
},
{
"grad_norm": 1.0811901092529297,
"learning_rate": 4.19e-05,
"loss": 0.6181,
"step": 420
},
{
"grad_norm": 1.211959958076477,
"learning_rate": 4.29e-05,
"loss": 0.6222,
"step": 430
},
{
"grad_norm": 1.1791653633117676,
"learning_rate": 4.39e-05,
"loss": 0.6018,
"step": 440
},
{
"grad_norm": 1.483304500579834,
"learning_rate": 4.49e-05,
"loss": 0.5883,
"step": 450
},
{
"grad_norm": 1.136581540107727,
"learning_rate": 4.5900000000000004e-05,
"loss": 0.5781,
"step": 460
},
{
"grad_norm": 1.0122281312942505,
"learning_rate": 4.69e-05,
"loss": 0.5833,
"step": 470
},
{
"grad_norm": 1.294203519821167,
"learning_rate": 4.79e-05,
"loss": 0.5811,
"step": 480
},
{
"grad_norm": 1.036759614944458,
"learning_rate": 4.89e-05,
"loss": 0.5826,
"step": 490
},
{
"eval/loss": 0.540949667096138,
"step": 500
},
{
"grad_norm": 1.4752445220947266,
"learning_rate": 4.99e-05,
"loss": 0.5589,
"step": 500
},
{
"grad_norm": 0.9996066093444824,
"learning_rate": 5.0900000000000004e-05,
"loss": 0.5726,
"step": 510
},
{
"grad_norm": 1.1314283609390259,
"learning_rate": 5.19e-05,
"loss": 0.5489,
"step": 520
},
{
"grad_norm": 1.0463645458221436,
"learning_rate": 5.2900000000000005e-05,
"loss": 0.537,
"step": 530
},
{
"grad_norm": 1.1870821714401245,
"learning_rate": 5.390000000000001e-05,
"loss": 0.5419,
"step": 540
},
{
"grad_norm": 1.0127766132354736,
"learning_rate": 5.4900000000000006e-05,
"loss": 0.5383,
"step": 550
},
{
"grad_norm": 1.1530522108078003,
"learning_rate": 5.590000000000001e-05,
"loss": 0.5255,
"step": 560
},
{
"grad_norm": 1.6963386535644531,
"learning_rate": 5.69e-05,
"loss": 0.5248,
"step": 570
},
{
"grad_norm": 1.5842453241348267,
"learning_rate": 5.79e-05,
"loss": 0.5243,
"step": 580
},
{
"grad_norm": 1.3649457693099976,
"learning_rate": 5.89e-05,
"loss": 0.5147,
"step": 590
},
{
"grad_norm": 1.018904447555542,
"learning_rate": 5.99e-05,
"loss": 0.5042,
"step": 600
},
{
"grad_norm": 1.252278208732605,
"learning_rate": 6.09e-05,
"loss": 0.5213,
"step": 610
},
{
"grad_norm": 1.2415512800216675,
"learning_rate": 6.19e-05,
"loss": 0.4769,
"step": 620
},
{
"grad_norm": 1.3829114437103271,
"learning_rate": 6.29e-05,
"loss": 0.4806,
"step": 630
},
{
"grad_norm": 1.2860313653945923,
"learning_rate": 6.390000000000001e-05,
"loss": 0.4687,
"step": 640
},
{
"grad_norm": 1.1453088521957397,
"learning_rate": 6.49e-05,
"loss": 0.477,
"step": 650
},
{
"grad_norm": 1.2535901069641113,
"learning_rate": 6.59e-05,
"loss": 0.4541,
"step": 660
},
{
"grad_norm": 1.2619575262069702,
"learning_rate": 6.690000000000001e-05,
"loss": 0.4565,
"step": 670
},
{
"grad_norm": 1.1378668546676636,
"learning_rate": 6.790000000000001e-05,
"loss": 0.4395,
"step": 680
},
{
"grad_norm": 1.0631095170974731,
"learning_rate": 6.89e-05,
"loss": 0.4185,
"step": 690
},
{
"grad_norm": 1.1509623527526855,
"learning_rate": 6.99e-05,
"loss": 0.437,
"step": 700
},
{
"grad_norm": 1.249911904335022,
"learning_rate": 7.09e-05,
"loss": 0.4273,
"step": 710
},
{
"grad_norm": 1.1548298597335815,
"learning_rate": 7.19e-05,
"loss": 0.4296,
"step": 720
},
{
"grad_norm": 1.0660429000854492,
"learning_rate": 7.29e-05,
"loss": 0.4438,
"step": 730
},
{
"grad_norm": 1.2336221933364868,
"learning_rate": 7.390000000000001e-05,
"loss": 0.407,
"step": 740
},
{
"grad_norm": 1.073397159576416,
"learning_rate": 7.49e-05,
"loss": 0.3936,
"step": 750
},
{
"grad_norm": 1.2548182010650635,
"learning_rate": 7.59e-05,
"loss": 0.3991,
"step": 760
},
{
"grad_norm": 1.4380117654800415,
"learning_rate": 7.69e-05,
"loss": 0.3957,
"step": 770
},
{
"grad_norm": 1.2932844161987305,
"learning_rate": 7.790000000000001e-05,
"loss": 0.4081,
"step": 780
},
{
"grad_norm": 1.1372441053390503,
"learning_rate": 7.890000000000001e-05,
"loss": 0.3812,
"step": 790
},
{
"grad_norm": 1.1620570421218872,
"learning_rate": 7.99e-05,
"loss": 0.3952,
"step": 800
},
{
"grad_norm": 1.1965490579605103,
"learning_rate": 8.090000000000001e-05,
"loss": 0.3677,
"step": 810
},
{
"grad_norm": 1.176527738571167,
"learning_rate": 8.19e-05,
"loss": 0.3798,
"step": 820
},
{
"grad_norm": 1.153993010520935,
"learning_rate": 8.29e-05,
"loss": 0.3629,
"step": 830
},
{
"grad_norm": 1.3327205181121826,
"learning_rate": 8.39e-05,
"loss": 0.3578,
"step": 840
},
{
"grad_norm": 1.1645392179489136,
"learning_rate": 8.49e-05,
"loss": 0.3542,
"step": 850
},
{
"grad_norm": 1.1183959245681763,
"learning_rate": 8.59e-05,
"loss": 0.3452,
"step": 860
},
{
"grad_norm": 1.4171571731567383,
"learning_rate": 8.69e-05,
"loss": 0.328,
"step": 870
},
{
"grad_norm": 1.2265501022338867,
"learning_rate": 8.790000000000001e-05,
"loss": 0.3427,
"step": 880
},
{
"grad_norm": 1.3434756994247437,
"learning_rate": 8.89e-05,
"loss": 0.3333,
"step": 890
},
{
"grad_norm": 1.3676091432571411,
"learning_rate": 8.99e-05,
"loss": 0.3142,
"step": 900
},
{
"grad_norm": 1.0545670986175537,
"learning_rate": 9.090000000000001e-05,
"loss": 0.3242,
"step": 910
},
{
"grad_norm": 1.1802937984466553,
"learning_rate": 9.190000000000001e-05,
"loss": 0.3419,
"step": 920
},
{
"grad_norm": 1.2357131242752075,
"learning_rate": 9.290000000000001e-05,
"loss": 0.2918,
"step": 930
},
{
"grad_norm": 1.2467869520187378,
"learning_rate": 9.39e-05,
"loss": 0.2812,
"step": 940
},
{
"grad_norm": 1.2177903652191162,
"learning_rate": 9.49e-05,
"loss": 0.2786,
"step": 950
},
{
"grad_norm": 1.2031254768371582,
"learning_rate": 9.59e-05,
"loss": 0.2801,
"step": 960
},
{
"grad_norm": 1.27996826171875,
"learning_rate": 9.69e-05,
"loss": 0.2944,
"step": 970
},
{
"grad_norm": 1.4937174320220947,
"learning_rate": 9.790000000000001e-05,
"loss": 0.259,
"step": 980
},
{
"grad_norm": 1.2263216972351074,
"learning_rate": 9.89e-05,
"loss": 0.269,
"step": 990
},
{
"eval/loss": 0.2459017077088356,
"step": 1000
},
{
"grad_norm": 1.1868503093719482,
"learning_rate": 9.99e-05,
"loss": 0.2673,
"step": 1000
},
{
"grad_norm": 1.2517995834350586,
"learning_rate": 9.999994463727085e-05,
"loss": 0.2957,
"step": 1010
},
{
"grad_norm": 1.1621079444885254,
"learning_rate": 9.999975326009292e-05,
"loss": 0.2406,
"step": 1020
},
{
"grad_norm": 1.2248700857162476,
"learning_rate": 9.999942518549879e-05,
"loss": 0.2588,
"step": 1030
},
{
"grad_norm": 1.1486198902130127,
"learning_rate": 9.999896041438544e-05,
"loss": 0.2869,
"step": 1040
},
{
"grad_norm": 1.1869938373565674,
"learning_rate": 9.999835894802353e-05,
"loss": 0.2613,
"step": 1050
},
{
"grad_norm": 1.2058380842208862,
"learning_rate": 9.999762078805743e-05,
"loss": 0.2367,
"step": 1060
},
{
"grad_norm": 1.2073358297348022,
"learning_rate": 9.999674593650526e-05,
"loss": 0.2343,
"step": 1070
},
{
"grad_norm": 1.3462257385253906,
"learning_rate": 9.99957343957588e-05,
"loss": 0.2043,
"step": 1080
},
{
"grad_norm": 1.21333646774292,
"learning_rate": 9.99945861685836e-05,
"loss": 0.22,
"step": 1090
},
{
"grad_norm": 1.172276496887207,
"learning_rate": 9.999330125811884e-05,
"loss": 0.2268,
"step": 1100
},
{
"grad_norm": 1.5802624225616455,
"learning_rate": 9.999187966787744e-05,
"loss": 0.2389,
"step": 1110
},
{
"grad_norm": 1.0722038745880127,
"learning_rate": 9.999032140174595e-05,
"loss": 0.2069,
"step": 1120
},
{
"grad_norm": 1.2428017854690552,
"learning_rate": 9.998862646398464e-05,
"loss": 0.2105,
"step": 1130
},
{
"grad_norm": 1.109406590461731,
"learning_rate": 9.998679485922739e-05,
"loss": 0.204,
"step": 1140
},
{
"grad_norm": 1.133062720298767,
"learning_rate": 9.998482659248174e-05,
"loss": 0.1862,
"step": 1150
},
{
"grad_norm": 1.185992956161499,
"learning_rate": 9.998272166912883e-05,
"loss": 0.1944,
"step": 1160
},
{
"grad_norm": 1.0539828538894653,
"learning_rate": 9.998048009492347e-05,
"loss": 0.1603,
"step": 1170
},
{
"grad_norm": 1.2745673656463623,
"learning_rate": 9.997810187599403e-05,
"loss": 0.1815,
"step": 1180
},
{
"grad_norm": 1.2294188737869263,
"learning_rate": 9.997558701884249e-05,
"loss": 0.1774,
"step": 1190
},
{
"grad_norm": 1.6289048194885254,
"learning_rate": 9.997293553034433e-05,
"loss": 0.171,
"step": 1200
},
{
"grad_norm": 1.2011067867279053,
"learning_rate": 9.997014741774866e-05,
"loss": 0.1657,
"step": 1210
},
{
"grad_norm": 1.1529210805892944,
"learning_rate": 9.996722268867803e-05,
"loss": 0.1642,
"step": 1220
},
{
"grad_norm": 0.8735513091087341,
"learning_rate": 9.996416135112858e-05,
"loss": 0.1393,
"step": 1230
},
{
"grad_norm": 1.3112437725067139,
"learning_rate": 9.996096341346988e-05,
"loss": 0.1392,
"step": 1240
},
{
"grad_norm": 1.2347687482833862,
"learning_rate": 9.995762888444495e-05,
"loss": 0.1674,
"step": 1250
},
{
"grad_norm": 1.539437174797058,
"learning_rate": 9.995415777317027e-05,
"loss": 0.1625,
"step": 1260
},
{
"grad_norm": 1.2333788871765137,
"learning_rate": 9.995055008913574e-05,
"loss": 0.1328,
"step": 1270
},
{
"grad_norm": 1.1541303396224976,
"learning_rate": 9.994680584220463e-05,
"loss": 0.1294,
"step": 1280
},
{
"grad_norm": 1.0528708696365356,
"learning_rate": 9.994292504261355e-05,
"loss": 0.1441,
"step": 1290
},
{
"grad_norm": 1.0454338788986206,
"learning_rate": 9.993890770097247e-05,
"loss": 0.1266,
"step": 1300
},
{
"grad_norm": 1.1280555725097656,
"learning_rate": 9.993475382826467e-05,
"loss": 0.1426,
"step": 1310
},
{
"grad_norm": 1.187239170074463,
"learning_rate": 9.993046343584664e-05,
"loss": 0.1422,
"step": 1320
},
{
"grad_norm": 1.0262149572372437,
"learning_rate": 9.992603653544816e-05,
"loss": 0.1161,
"step": 1330
},
{
"grad_norm": 1.1586066484451294,
"learning_rate": 9.992147313917222e-05,
"loss": 0.1408,
"step": 1340
},
{
"grad_norm": 0.9765651226043701,
"learning_rate": 9.991677325949497e-05,
"loss": 0.1611,
"step": 1350
},
{
"grad_norm": 0.9763075709342957,
"learning_rate": 9.991193690926568e-05,
"loss": 0.1464,
"step": 1360
},
{
"grad_norm": 1.2092801332473755,
"learning_rate": 9.990696410170678e-05,
"loss": 0.1466,
"step": 1370
},
{
"grad_norm": 1.0392274856567383,
"learning_rate": 9.990185485041371e-05,
"loss": 0.1263,
"step": 1380
},
{
"grad_norm": 1.0358021259307861,
"learning_rate": 9.989660916935498e-05,
"loss": 0.1282,
"step": 1390
},
{
"grad_norm": 1.0262398719787598,
"learning_rate": 9.989122707287208e-05,
"loss": 0.1391,
"step": 1400
},
{
"grad_norm": 1.1978421211242676,
"learning_rate": 9.988570857567945e-05,
"loss": 0.1218,
"step": 1410
},
{
"grad_norm": 0.9296699166297913,
"learning_rate": 9.988005369286446e-05,
"loss": 0.1331,
"step": 1420
},
{
"grad_norm": 1.0004020929336548,
"learning_rate": 9.987426243988734e-05,
"loss": 0.1372,
"step": 1430
},
{
"grad_norm": 1.0646557807922363,
"learning_rate": 9.986833483258114e-05,
"loss": 0.1334,
"step": 1440
},
{
"grad_norm": 0.9959461688995361,
"learning_rate": 9.986227088715173e-05,
"loss": 0.1187,
"step": 1450
},
{
"grad_norm": 1.0928994417190552,
"learning_rate": 9.98560706201777e-05,
"loss": 0.1455,
"step": 1460
},
{
"grad_norm": 1.1495130062103271,
"learning_rate": 9.984973404861036e-05,
"loss": 0.1098,
"step": 1470
},
{
"grad_norm": 1.3037567138671875,
"learning_rate": 9.984326118977361e-05,
"loss": 0.1255,
"step": 1480
},
{
"grad_norm": 0.923818051815033,
"learning_rate": 9.983665206136406e-05,
"loss": 0.1439,
"step": 1490
},
{
"eval/loss": 0.12131376132369041,
"step": 1500
},
{
"grad_norm": 0.9496746063232422,
"learning_rate": 9.982990668145075e-05,
"loss": 0.123,
"step": 1500
},
{
"grad_norm": 1.0832738876342773,
"learning_rate": 9.982302506847534e-05,
"loss": 0.1408,
"step": 1510
},
{
"grad_norm": 1.033626914024353,
"learning_rate": 9.981600724125189e-05,
"loss": 0.1108,
"step": 1520
},
{
"grad_norm": 1.0005450248718262,
"learning_rate": 9.980885321896685e-05,
"loss": 0.1274,
"step": 1530
},
{
"grad_norm": 1.0442663431167603,
"learning_rate": 9.980156302117905e-05,
"loss": 0.1238,
"step": 1540
},
{
"grad_norm": 0.8260471820831299,
"learning_rate": 9.979413666781963e-05,
"loss": 0.1231,
"step": 1550
},
{
"grad_norm": 0.884735107421875,
"learning_rate": 9.978657417919193e-05,
"loss": 0.1495,
"step": 1560
},
{
"grad_norm": 0.9236319661140442,
"learning_rate": 9.977887557597153e-05,
"loss": 0.1327,
"step": 1570
},
{
"grad_norm": 0.9460572004318237,
"learning_rate": 9.97710408792061e-05,
"loss": 0.1263,
"step": 1580
},
{
"grad_norm": 0.9749200344085693,
"learning_rate": 9.976307011031542e-05,
"loss": 0.12,
"step": 1590
},
{
"grad_norm": 1.1136820316314697,
"learning_rate": 9.975496329109126e-05,
"loss": 0.1323,
"step": 1600
},
{
"grad_norm": 0.8567096590995789,
"learning_rate": 9.974672044369732e-05,
"loss": 0.125,
"step": 1610
},
{
"grad_norm": 1.0884920358657837,
"learning_rate": 9.97383415906693e-05,
"loss": 0.115,
"step": 1620
},
{
"grad_norm": 1.0339338779449463,
"learning_rate": 9.97298267549146e-05,
"loss": 0.1386,
"step": 1630
},
{
"grad_norm": 0.9121850728988647,
"learning_rate": 9.972117595971249e-05,
"loss": 0.1249,
"step": 1640
},
{
"grad_norm": 0.9620202779769897,
"learning_rate": 9.971238922871391e-05,
"loss": 0.1322,
"step": 1650
},
{
"grad_norm": 1.0946760177612305,
"learning_rate": 9.970346658594142e-05,
"loss": 0.1123,
"step": 1660
},
{
"grad_norm": 0.9470517635345459,
"learning_rate": 9.969440805578923e-05,
"loss": 0.1308,
"step": 1670
},
{
"grad_norm": 0.7607911229133606,
"learning_rate": 9.968521366302298e-05,
"loss": 0.1212,
"step": 1680
},
{
"grad_norm": 1.0321109294891357,
"learning_rate": 9.967588343277981e-05,
"loss": 0.1351,
"step": 1690
},
{
"grad_norm": 1.2910419702529907,
"learning_rate": 9.966641739056818e-05,
"loss": 0.1594,
"step": 1700
},
{
"grad_norm": 0.7687619924545288,
"learning_rate": 9.965681556226793e-05,
"loss": 0.1442,
"step": 1710
},
{
"grad_norm": 0.882544755935669,
"learning_rate": 9.964707797413006e-05,
"loss": 0.1131,
"step": 1720
},
{
"grad_norm": 1.0736138820648193,
"learning_rate": 9.963720465277679e-05,
"loss": 0.0997,
"step": 1730
},
{
"grad_norm": 1.0465947389602661,
"learning_rate": 9.96271956252014e-05,
"loss": 0.1114,
"step": 1740
},
{
"grad_norm": 0.8517502546310425,
"learning_rate": 9.961705091876816e-05,
"loss": 0.1054,
"step": 1750
},
{
"grad_norm": 0.8722822666168213,
"learning_rate": 9.960677056121235e-05,
"loss": 0.1409,
"step": 1760
},
{
"grad_norm": 1.3352707624435425,
"learning_rate": 9.959635458064005e-05,
"loss": 0.1207,
"step": 1770
},
{
"grad_norm": 1.0375741720199585,
"learning_rate": 9.958580300552815e-05,
"loss": 0.1018,
"step": 1780
},
{
"grad_norm": 0.9663418531417847,
"learning_rate": 9.957511586472426e-05,
"loss": 0.1131,
"step": 1790
},
{
"grad_norm": 0.9925614595413208,
"learning_rate": 9.956429318744662e-05,
"loss": 0.0994,
"step": 1800
},
{
"grad_norm": 0.9052272439002991,
"learning_rate": 9.955333500328404e-05,
"loss": 0.1439,
"step": 1810
},
{
"grad_norm": 1.1061209440231323,
"learning_rate": 9.95422413421957e-05,
"loss": 0.099,
"step": 1820
},
{
"grad_norm": 0.9718905687332153,
"learning_rate": 9.953101223451133e-05,
"loss": 0.1334,
"step": 1830
},
{
"grad_norm": 1.041150450706482,
"learning_rate": 9.951964771093085e-05,
"loss": 0.102,
"step": 1840
},
{
"grad_norm": 0.7296974658966064,
"learning_rate": 9.950814780252442e-05,
"loss": 0.1216,
"step": 1850
},
{
"grad_norm": 0.956504225730896,
"learning_rate": 9.949651254073236e-05,
"loss": 0.1504,
"step": 1860
},
{
"grad_norm": 0.9494929909706116,
"learning_rate": 9.948474195736504e-05,
"loss": 0.1176,
"step": 1870
},
{
"grad_norm": 1.073912501335144,
"learning_rate": 9.947283608460277e-05,
"loss": 0.0968,
"step": 1880
},
{
"grad_norm": 0.9738394618034363,
"learning_rate": 9.946079495499577e-05,
"loss": 0.1157,
"step": 1890
},
{
"grad_norm": 0.9605513215065002,
"learning_rate": 9.944861860146401e-05,
"loss": 0.122,
"step": 1900
},
{
"grad_norm": 0.8335412740707397,
"learning_rate": 9.943630705729719e-05,
"loss": 0.0999,
"step": 1910
},
{
"grad_norm": 0.881373405456543,
"learning_rate": 9.942386035615459e-05,
"loss": 0.1014,
"step": 1920
},
{
"grad_norm": 0.7942837476730347,
"learning_rate": 9.941127853206503e-05,
"loss": 0.1095,
"step": 1930
},
{
"grad_norm": 1.0058091878890991,
"learning_rate": 9.939856161942673e-05,
"loss": 0.118,
"step": 1940
},
{
"grad_norm": 0.8951269388198853,
"learning_rate": 9.938570965300724e-05,
"loss": 0.1093,
"step": 1950
},
{
"grad_norm": 0.8100780844688416,
"learning_rate": 9.937272266794335e-05,
"loss": 0.1344,
"step": 1960
},
{
"grad_norm": 0.8485944867134094,
"learning_rate": 9.935960069974096e-05,
"loss": 0.1062,
"step": 1970
},
{
"grad_norm": 1.0462373495101929,
"learning_rate": 9.934634378427506e-05,
"loss": 0.1073,
"step": 1980
},
{
"grad_norm": 0.7918229103088379,
"learning_rate": 9.933295195778954e-05,
"loss": 0.0865,
"step": 1990
},
{
"eval/loss": 0.10448359854519368,
"step": 2000
},
{
"grad_norm": 1.0324554443359375,
"learning_rate": 9.931942525689715e-05,
"loss": 0.1235,
"step": 2000
}
],
"logging_steps": 10,
"max_steps": 20000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 2000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}