my_Emotion_DF_Image_ViT_V1 / trainer_state.json
thenewsupercell's picture
thenewsupercell/Louis_EMOTION_Image_VIT
9713a2c verified
{
"best_metric": 0.7127333519086096,
"best_model_checkpoint": "Louis_Emotion_DF_Image_VIT_V1/checkpoint-5385",
"epoch": 3.0,
"eval_steps": 500,
"global_step": 5385,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.005571030640668524,
"grad_norm": 1.948201060295105,
"learning_rate": 4.990714948932219e-05,
"loss": 1.8845,
"step": 10
},
{
"epoch": 0.011142061281337047,
"grad_norm": 2.1474852561950684,
"learning_rate": 4.981429897864439e-05,
"loss": 1.7726,
"step": 20
},
{
"epoch": 0.016713091922005572,
"grad_norm": 1.7504295110702515,
"learning_rate": 4.972144846796657e-05,
"loss": 1.7291,
"step": 30
},
{
"epoch": 0.022284122562674095,
"grad_norm": 1.9806510210037231,
"learning_rate": 4.962859795728877e-05,
"loss": 1.631,
"step": 40
},
{
"epoch": 0.027855153203342618,
"grad_norm": 2.3543717861175537,
"learning_rate": 4.953574744661096e-05,
"loss": 1.5524,
"step": 50
},
{
"epoch": 0.033426183844011144,
"grad_norm": 2.5557103157043457,
"learning_rate": 4.9442896935933144e-05,
"loss": 1.5061,
"step": 60
},
{
"epoch": 0.03899721448467967,
"grad_norm": 2.1124281883239746,
"learning_rate": 4.9350046425255343e-05,
"loss": 1.5135,
"step": 70
},
{
"epoch": 0.04456824512534819,
"grad_norm": 2.636732578277588,
"learning_rate": 4.925719591457753e-05,
"loss": 1.4545,
"step": 80
},
{
"epoch": 0.05013927576601671,
"grad_norm": 2.1990904808044434,
"learning_rate": 4.916434540389973e-05,
"loss": 1.3375,
"step": 90
},
{
"epoch": 0.055710306406685235,
"grad_norm": 2.593843698501587,
"learning_rate": 4.9071494893221914e-05,
"loss": 1.3125,
"step": 100
},
{
"epoch": 0.06128133704735376,
"grad_norm": 2.3658862113952637,
"learning_rate": 4.897864438254411e-05,
"loss": 1.2438,
"step": 110
},
{
"epoch": 0.06685236768802229,
"grad_norm": 3.614569902420044,
"learning_rate": 4.88857938718663e-05,
"loss": 1.352,
"step": 120
},
{
"epoch": 0.07242339832869081,
"grad_norm": 3.075983762741089,
"learning_rate": 4.8792943361188485e-05,
"loss": 1.3506,
"step": 130
},
{
"epoch": 0.07799442896935933,
"grad_norm": 2.9833717346191406,
"learning_rate": 4.8700092850510685e-05,
"loss": 1.3486,
"step": 140
},
{
"epoch": 0.08356545961002786,
"grad_norm": 4.222170829772949,
"learning_rate": 4.860724233983287e-05,
"loss": 1.3921,
"step": 150
},
{
"epoch": 0.08913649025069638,
"grad_norm": 2.5788440704345703,
"learning_rate": 4.851439182915506e-05,
"loss": 1.3421,
"step": 160
},
{
"epoch": 0.0947075208913649,
"grad_norm": 2.7408132553100586,
"learning_rate": 4.8421541318477255e-05,
"loss": 1.1812,
"step": 170
},
{
"epoch": 0.10027855153203342,
"grad_norm": 2.91229510307312,
"learning_rate": 4.832869080779944e-05,
"loss": 1.1881,
"step": 180
},
{
"epoch": 0.10584958217270195,
"grad_norm": 2.4319872856140137,
"learning_rate": 4.823584029712164e-05,
"loss": 1.3427,
"step": 190
},
{
"epoch": 0.11142061281337047,
"grad_norm": 3.0504567623138428,
"learning_rate": 4.8142989786443826e-05,
"loss": 1.2334,
"step": 200
},
{
"epoch": 0.116991643454039,
"grad_norm": 2.9363996982574463,
"learning_rate": 4.805013927576602e-05,
"loss": 1.1809,
"step": 210
},
{
"epoch": 0.12256267409470752,
"grad_norm": 2.2370383739471436,
"learning_rate": 4.795728876508821e-05,
"loss": 1.1973,
"step": 220
},
{
"epoch": 0.12813370473537605,
"grad_norm": 2.7246251106262207,
"learning_rate": 4.7864438254410404e-05,
"loss": 1.2224,
"step": 230
},
{
"epoch": 0.13370473537604458,
"grad_norm": 3.5851166248321533,
"learning_rate": 4.7771587743732597e-05,
"loss": 1.0445,
"step": 240
},
{
"epoch": 0.1392757660167131,
"grad_norm": 1.9959110021591187,
"learning_rate": 4.767873723305478e-05,
"loss": 1.1164,
"step": 250
},
{
"epoch": 0.14484679665738162,
"grad_norm": 2.933548927307129,
"learning_rate": 4.7585886722376975e-05,
"loss": 1.1229,
"step": 260
},
{
"epoch": 0.15041782729805014,
"grad_norm": 4.168135643005371,
"learning_rate": 4.749303621169917e-05,
"loss": 1.25,
"step": 270
},
{
"epoch": 0.15598885793871867,
"grad_norm": 3.2790322303771973,
"learning_rate": 4.740018570102136e-05,
"loss": 1.177,
"step": 280
},
{
"epoch": 0.1615598885793872,
"grad_norm": 4.025652885437012,
"learning_rate": 4.7307335190343546e-05,
"loss": 1.1185,
"step": 290
},
{
"epoch": 0.1671309192200557,
"grad_norm": 2.921226978302002,
"learning_rate": 4.721448467966574e-05,
"loss": 1.1773,
"step": 300
},
{
"epoch": 0.17270194986072424,
"grad_norm": 3.095017671585083,
"learning_rate": 4.712163416898793e-05,
"loss": 1.2818,
"step": 310
},
{
"epoch": 0.17827298050139276,
"grad_norm": 2.596940279006958,
"learning_rate": 4.702878365831012e-05,
"loss": 1.0857,
"step": 320
},
{
"epoch": 0.18384401114206128,
"grad_norm": 3.0986833572387695,
"learning_rate": 4.6935933147632316e-05,
"loss": 1.161,
"step": 330
},
{
"epoch": 0.1894150417827298,
"grad_norm": 3.9067625999450684,
"learning_rate": 4.68430826369545e-05,
"loss": 1.126,
"step": 340
},
{
"epoch": 0.19498607242339833,
"grad_norm": 2.4455080032348633,
"learning_rate": 4.6750232126276694e-05,
"loss": 1.0944,
"step": 350
},
{
"epoch": 0.20055710306406685,
"grad_norm": 3.0159671306610107,
"learning_rate": 4.665738161559889e-05,
"loss": 1.0525,
"step": 360
},
{
"epoch": 0.20612813370473537,
"grad_norm": 3.257310152053833,
"learning_rate": 4.656453110492108e-05,
"loss": 1.1423,
"step": 370
},
{
"epoch": 0.2116991643454039,
"grad_norm": 2.324500322341919,
"learning_rate": 4.647168059424327e-05,
"loss": 1.0441,
"step": 380
},
{
"epoch": 0.21727019498607242,
"grad_norm": 2.893946409225464,
"learning_rate": 4.637883008356546e-05,
"loss": 1.0956,
"step": 390
},
{
"epoch": 0.22284122562674094,
"grad_norm": 3.661740779876709,
"learning_rate": 4.628597957288766e-05,
"loss": 1.0648,
"step": 400
},
{
"epoch": 0.22841225626740946,
"grad_norm": 3.601794719696045,
"learning_rate": 4.619312906220984e-05,
"loss": 1.1556,
"step": 410
},
{
"epoch": 0.233983286908078,
"grad_norm": 2.6910550594329834,
"learning_rate": 4.6100278551532035e-05,
"loss": 0.9603,
"step": 420
},
{
"epoch": 0.2395543175487465,
"grad_norm": 2.952540636062622,
"learning_rate": 4.600742804085423e-05,
"loss": 1.1827,
"step": 430
},
{
"epoch": 0.24512534818941503,
"grad_norm": 4.099498271942139,
"learning_rate": 4.5914577530176414e-05,
"loss": 1.0573,
"step": 440
},
{
"epoch": 0.25069637883008355,
"grad_norm": 3.884427070617676,
"learning_rate": 4.582172701949861e-05,
"loss": 1.1742,
"step": 450
},
{
"epoch": 0.2562674094707521,
"grad_norm": 3.0782787799835205,
"learning_rate": 4.57288765088208e-05,
"loss": 1.051,
"step": 460
},
{
"epoch": 0.2618384401114206,
"grad_norm": 4.079422473907471,
"learning_rate": 4.563602599814299e-05,
"loss": 0.9732,
"step": 470
},
{
"epoch": 0.26740947075208915,
"grad_norm": 3.0462193489074707,
"learning_rate": 4.5543175487465184e-05,
"loss": 1.0204,
"step": 480
},
{
"epoch": 0.27298050139275765,
"grad_norm": 3.944453716278076,
"learning_rate": 4.545032497678737e-05,
"loss": 1.2336,
"step": 490
},
{
"epoch": 0.2785515320334262,
"grad_norm": 3.276047945022583,
"learning_rate": 4.535747446610957e-05,
"loss": 1.1331,
"step": 500
},
{
"epoch": 0.2841225626740947,
"grad_norm": 2.8910679817199707,
"learning_rate": 4.5264623955431755e-05,
"loss": 1.0437,
"step": 510
},
{
"epoch": 0.28969359331476324,
"grad_norm": 2.332932472229004,
"learning_rate": 4.5171773444753954e-05,
"loss": 1.264,
"step": 520
},
{
"epoch": 0.29526462395543174,
"grad_norm": 3.207519292831421,
"learning_rate": 4.507892293407614e-05,
"loss": 1.0362,
"step": 530
},
{
"epoch": 0.3008356545961003,
"grad_norm": 3.792386770248413,
"learning_rate": 4.4986072423398326e-05,
"loss": 1.0799,
"step": 540
},
{
"epoch": 0.3064066852367688,
"grad_norm": 3.6370913982391357,
"learning_rate": 4.4893221912720525e-05,
"loss": 1.0702,
"step": 550
},
{
"epoch": 0.31197771587743733,
"grad_norm": 7.727144241333008,
"learning_rate": 4.480037140204271e-05,
"loss": 1.0313,
"step": 560
},
{
"epoch": 0.31754874651810583,
"grad_norm": 3.0483336448669434,
"learning_rate": 4.470752089136491e-05,
"loss": 1.0816,
"step": 570
},
{
"epoch": 0.3231197771587744,
"grad_norm": 4.045107841491699,
"learning_rate": 4.4614670380687096e-05,
"loss": 1.1036,
"step": 580
},
{
"epoch": 0.3286908077994429,
"grad_norm": 2.4204320907592773,
"learning_rate": 4.452181987000928e-05,
"loss": 0.9899,
"step": 590
},
{
"epoch": 0.3342618384401114,
"grad_norm": 1.9364632368087769,
"learning_rate": 4.442896935933148e-05,
"loss": 0.9319,
"step": 600
},
{
"epoch": 0.3398328690807799,
"grad_norm": 2.769692897796631,
"learning_rate": 4.433611884865367e-05,
"loss": 0.9948,
"step": 610
},
{
"epoch": 0.34540389972144847,
"grad_norm": 3.5920908451080322,
"learning_rate": 4.4243268337975866e-05,
"loss": 0.9915,
"step": 620
},
{
"epoch": 0.35097493036211697,
"grad_norm": 3.9949071407318115,
"learning_rate": 4.415041782729805e-05,
"loss": 1.0652,
"step": 630
},
{
"epoch": 0.3565459610027855,
"grad_norm": 3.3296191692352295,
"learning_rate": 4.4057567316620244e-05,
"loss": 1.0703,
"step": 640
},
{
"epoch": 0.362116991643454,
"grad_norm": 5.59266996383667,
"learning_rate": 4.396471680594244e-05,
"loss": 1.0078,
"step": 650
},
{
"epoch": 0.36768802228412256,
"grad_norm": 3.6050076484680176,
"learning_rate": 4.387186629526462e-05,
"loss": 0.9541,
"step": 660
},
{
"epoch": 0.3732590529247911,
"grad_norm": 2.4195094108581543,
"learning_rate": 4.377901578458682e-05,
"loss": 0.8932,
"step": 670
},
{
"epoch": 0.3788300835654596,
"grad_norm": 3.135007619857788,
"learning_rate": 4.368616527390901e-05,
"loss": 1.13,
"step": 680
},
{
"epoch": 0.38440111420612816,
"grad_norm": 3.0735456943511963,
"learning_rate": 4.35933147632312e-05,
"loss": 0.9413,
"step": 690
},
{
"epoch": 0.38997214484679665,
"grad_norm": 4.281991481781006,
"learning_rate": 4.350046425255339e-05,
"loss": 0.9636,
"step": 700
},
{
"epoch": 0.3955431754874652,
"grad_norm": 3.6378397941589355,
"learning_rate": 4.340761374187558e-05,
"loss": 1.0549,
"step": 710
},
{
"epoch": 0.4011142061281337,
"grad_norm": 2.570322036743164,
"learning_rate": 4.331476323119778e-05,
"loss": 1.03,
"step": 720
},
{
"epoch": 0.40668523676880225,
"grad_norm": 3.7980899810791016,
"learning_rate": 4.3221912720519964e-05,
"loss": 0.9204,
"step": 730
},
{
"epoch": 0.41225626740947074,
"grad_norm": 2.180097818374634,
"learning_rate": 4.3129062209842156e-05,
"loss": 0.9956,
"step": 740
},
{
"epoch": 0.4178272980501393,
"grad_norm": 3.4004600048065186,
"learning_rate": 4.303621169916435e-05,
"loss": 1.0413,
"step": 750
},
{
"epoch": 0.4233983286908078,
"grad_norm": 2.9130539894104004,
"learning_rate": 4.2943361188486535e-05,
"loss": 0.9712,
"step": 760
},
{
"epoch": 0.42896935933147634,
"grad_norm": 2.5885844230651855,
"learning_rate": 4.2850510677808734e-05,
"loss": 1.0015,
"step": 770
},
{
"epoch": 0.43454038997214484,
"grad_norm": 2.890307664871216,
"learning_rate": 4.275766016713092e-05,
"loss": 0.9238,
"step": 780
},
{
"epoch": 0.4401114206128134,
"grad_norm": 2.586287498474121,
"learning_rate": 4.266480965645311e-05,
"loss": 1.1793,
"step": 790
},
{
"epoch": 0.4456824512534819,
"grad_norm": 3.8789122104644775,
"learning_rate": 4.2571959145775305e-05,
"loss": 0.9671,
"step": 800
},
{
"epoch": 0.45125348189415043,
"grad_norm": 2.513338804244995,
"learning_rate": 4.24791086350975e-05,
"loss": 1.1206,
"step": 810
},
{
"epoch": 0.4568245125348189,
"grad_norm": 4.040161609649658,
"learning_rate": 4.238625812441969e-05,
"loss": 0.9741,
"step": 820
},
{
"epoch": 0.4623955431754875,
"grad_norm": 3.625004768371582,
"learning_rate": 4.2293407613741876e-05,
"loss": 1.0511,
"step": 830
},
{
"epoch": 0.467966573816156,
"grad_norm": 2.6117324829101562,
"learning_rate": 4.220055710306407e-05,
"loss": 0.973,
"step": 840
},
{
"epoch": 0.4735376044568245,
"grad_norm": 1.8209770917892456,
"learning_rate": 4.210770659238626e-05,
"loss": 0.9672,
"step": 850
},
{
"epoch": 0.479108635097493,
"grad_norm": 2.686767339706421,
"learning_rate": 4.201485608170845e-05,
"loss": 0.993,
"step": 860
},
{
"epoch": 0.48467966573816157,
"grad_norm": 2.2788872718811035,
"learning_rate": 4.192200557103064e-05,
"loss": 0.9141,
"step": 870
},
{
"epoch": 0.49025069637883006,
"grad_norm": 3.2874503135681152,
"learning_rate": 4.182915506035283e-05,
"loss": 1.0949,
"step": 880
},
{
"epoch": 0.4958217270194986,
"grad_norm": 3.21272349357605,
"learning_rate": 4.1736304549675024e-05,
"loss": 0.9701,
"step": 890
},
{
"epoch": 0.5013927576601671,
"grad_norm": 4.310153007507324,
"learning_rate": 4.164345403899722e-05,
"loss": 0.9595,
"step": 900
},
{
"epoch": 0.5069637883008357,
"grad_norm": 2.943368434906006,
"learning_rate": 4.155060352831941e-05,
"loss": 1.0342,
"step": 910
},
{
"epoch": 0.5125348189415042,
"grad_norm": 3.337778329849243,
"learning_rate": 4.1457753017641595e-05,
"loss": 0.8848,
"step": 920
},
{
"epoch": 0.5181058495821727,
"grad_norm": 3.8520374298095703,
"learning_rate": 4.1364902506963794e-05,
"loss": 0.955,
"step": 930
},
{
"epoch": 0.5236768802228412,
"grad_norm": 3.1856610774993896,
"learning_rate": 4.127205199628598e-05,
"loss": 0.9171,
"step": 940
},
{
"epoch": 0.5292479108635098,
"grad_norm": 2.813004493713379,
"learning_rate": 4.117920148560817e-05,
"loss": 0.8912,
"step": 950
},
{
"epoch": 0.5348189415041783,
"grad_norm": 3.73842716217041,
"learning_rate": 4.1086350974930365e-05,
"loss": 0.87,
"step": 960
},
{
"epoch": 0.5403899721448467,
"grad_norm": 4.500996112823486,
"learning_rate": 4.099350046425255e-05,
"loss": 1.0216,
"step": 970
},
{
"epoch": 0.5459610027855153,
"grad_norm": 2.80609393119812,
"learning_rate": 4.090064995357475e-05,
"loss": 0.9803,
"step": 980
},
{
"epoch": 0.5515320334261838,
"grad_norm": 3.340052604675293,
"learning_rate": 4.0807799442896936e-05,
"loss": 0.8977,
"step": 990
},
{
"epoch": 0.5571030640668524,
"grad_norm": 3.450836420059204,
"learning_rate": 4.071494893221913e-05,
"loss": 0.9845,
"step": 1000
},
{
"epoch": 0.5626740947075209,
"grad_norm": 2.639638662338257,
"learning_rate": 4.062209842154132e-05,
"loss": 0.9415,
"step": 1010
},
{
"epoch": 0.5682451253481894,
"grad_norm": 3.1470372676849365,
"learning_rate": 4.052924791086351e-05,
"loss": 1.0006,
"step": 1020
},
{
"epoch": 0.5738161559888579,
"grad_norm": 3.556126832962036,
"learning_rate": 4.0436397400185706e-05,
"loss": 0.9859,
"step": 1030
},
{
"epoch": 0.5793871866295265,
"grad_norm": 4.580634117126465,
"learning_rate": 4.034354688950789e-05,
"loss": 0.8674,
"step": 1040
},
{
"epoch": 0.584958217270195,
"grad_norm": 3.796309232711792,
"learning_rate": 4.0250696378830085e-05,
"loss": 0.9327,
"step": 1050
},
{
"epoch": 0.5905292479108635,
"grad_norm": 2.9451568126678467,
"learning_rate": 4.015784586815228e-05,
"loss": 0.9889,
"step": 1060
},
{
"epoch": 0.596100278551532,
"grad_norm": 2.4490549564361572,
"learning_rate": 4.006499535747446e-05,
"loss": 0.9661,
"step": 1070
},
{
"epoch": 0.6016713091922006,
"grad_norm": 3.1423044204711914,
"learning_rate": 3.997214484679666e-05,
"loss": 0.8991,
"step": 1080
},
{
"epoch": 0.6072423398328691,
"grad_norm": 2.6744227409362793,
"learning_rate": 3.987929433611885e-05,
"loss": 0.9822,
"step": 1090
},
{
"epoch": 0.6128133704735376,
"grad_norm": 2.410128593444824,
"learning_rate": 3.978644382544105e-05,
"loss": 0.87,
"step": 1100
},
{
"epoch": 0.6183844011142061,
"grad_norm": 3.4056427478790283,
"learning_rate": 3.969359331476323e-05,
"loss": 0.9011,
"step": 1110
},
{
"epoch": 0.6239554317548747,
"grad_norm": 3.5832958221435547,
"learning_rate": 3.960074280408542e-05,
"loss": 0.9179,
"step": 1120
},
{
"epoch": 0.6295264623955432,
"grad_norm": 4.678562164306641,
"learning_rate": 3.950789229340762e-05,
"loss": 0.9475,
"step": 1130
},
{
"epoch": 0.6350974930362117,
"grad_norm": 3.114361047744751,
"learning_rate": 3.9415041782729804e-05,
"loss": 0.95,
"step": 1140
},
{
"epoch": 0.6406685236768802,
"grad_norm": 2.6210415363311768,
"learning_rate": 3.9322191272052003e-05,
"loss": 0.8933,
"step": 1150
},
{
"epoch": 0.6462395543175488,
"grad_norm": 2.759856939315796,
"learning_rate": 3.922934076137419e-05,
"loss": 0.9336,
"step": 1160
},
{
"epoch": 0.6518105849582173,
"grad_norm": 3.449307680130005,
"learning_rate": 3.9136490250696375e-05,
"loss": 0.9851,
"step": 1170
},
{
"epoch": 0.6573816155988857,
"grad_norm": 4.323282718658447,
"learning_rate": 3.9043639740018574e-05,
"loss": 0.9735,
"step": 1180
},
{
"epoch": 0.6629526462395543,
"grad_norm": 2.79763126373291,
"learning_rate": 3.895078922934076e-05,
"loss": 0.8509,
"step": 1190
},
{
"epoch": 0.6685236768802229,
"grad_norm": 5.2140045166015625,
"learning_rate": 3.885793871866296e-05,
"loss": 1.0342,
"step": 1200
},
{
"epoch": 0.6740947075208914,
"grad_norm": 3.2437567710876465,
"learning_rate": 3.8765088207985145e-05,
"loss": 0.9919,
"step": 1210
},
{
"epoch": 0.6796657381615598,
"grad_norm": 3.9220008850097656,
"learning_rate": 3.867223769730734e-05,
"loss": 0.9809,
"step": 1220
},
{
"epoch": 0.6852367688022284,
"grad_norm": 1.8723160028457642,
"learning_rate": 3.857938718662953e-05,
"loss": 0.7315,
"step": 1230
},
{
"epoch": 0.6908077994428969,
"grad_norm": 5.3923211097717285,
"learning_rate": 3.8486536675951716e-05,
"loss": 0.9709,
"step": 1240
},
{
"epoch": 0.6963788300835655,
"grad_norm": 3.8164379596710205,
"learning_rate": 3.8393686165273915e-05,
"loss": 1.0331,
"step": 1250
},
{
"epoch": 0.7019498607242339,
"grad_norm": 4.3825836181640625,
"learning_rate": 3.83008356545961e-05,
"loss": 1.0042,
"step": 1260
},
{
"epoch": 0.7075208913649025,
"grad_norm": 2.7108051776885986,
"learning_rate": 3.8207985143918294e-05,
"loss": 0.974,
"step": 1270
},
{
"epoch": 0.713091922005571,
"grad_norm": 2.64860463142395,
"learning_rate": 3.8115134633240486e-05,
"loss": 0.8971,
"step": 1280
},
{
"epoch": 0.7186629526462396,
"grad_norm": 3.869168758392334,
"learning_rate": 3.802228412256267e-05,
"loss": 0.8225,
"step": 1290
},
{
"epoch": 0.724233983286908,
"grad_norm": 3.02199125289917,
"learning_rate": 3.792943361188487e-05,
"loss": 0.9492,
"step": 1300
},
{
"epoch": 0.7298050139275766,
"grad_norm": 5.071811199188232,
"learning_rate": 3.783658310120706e-05,
"loss": 1.018,
"step": 1310
},
{
"epoch": 0.7353760445682451,
"grad_norm": 3.2059361934661865,
"learning_rate": 3.774373259052925e-05,
"loss": 0.9163,
"step": 1320
},
{
"epoch": 0.7409470752089137,
"grad_norm": 4.003905296325684,
"learning_rate": 3.765088207985144e-05,
"loss": 0.8783,
"step": 1330
},
{
"epoch": 0.7465181058495822,
"grad_norm": 3.159921646118164,
"learning_rate": 3.755803156917363e-05,
"loss": 0.7841,
"step": 1340
},
{
"epoch": 0.7520891364902507,
"grad_norm": 2.852421760559082,
"learning_rate": 3.746518105849583e-05,
"loss": 0.9545,
"step": 1350
},
{
"epoch": 0.7576601671309192,
"grad_norm": 4.6871490478515625,
"learning_rate": 3.737233054781801e-05,
"loss": 0.9276,
"step": 1360
},
{
"epoch": 0.7632311977715878,
"grad_norm": 1.9484212398529053,
"learning_rate": 3.7279480037140206e-05,
"loss": 0.9676,
"step": 1370
},
{
"epoch": 0.7688022284122563,
"grad_norm": 4.423326015472412,
"learning_rate": 3.71866295264624e-05,
"loss": 1.0121,
"step": 1380
},
{
"epoch": 0.7743732590529248,
"grad_norm": 3.569786310195923,
"learning_rate": 3.709377901578459e-05,
"loss": 0.9413,
"step": 1390
},
{
"epoch": 0.7799442896935933,
"grad_norm": 1.8504449129104614,
"learning_rate": 3.700092850510678e-05,
"loss": 0.9478,
"step": 1400
},
{
"epoch": 0.7855153203342619,
"grad_norm": 3.4679548740386963,
"learning_rate": 3.690807799442897e-05,
"loss": 0.88,
"step": 1410
},
{
"epoch": 0.7910863509749304,
"grad_norm": 6.313425064086914,
"learning_rate": 3.681522748375116e-05,
"loss": 1.0565,
"step": 1420
},
{
"epoch": 0.7966573816155988,
"grad_norm": 3.7939631938934326,
"learning_rate": 3.6722376973073354e-05,
"loss": 1.0129,
"step": 1430
},
{
"epoch": 0.8022284122562674,
"grad_norm": 4.260021209716797,
"learning_rate": 3.662952646239555e-05,
"loss": 0.8527,
"step": 1440
},
{
"epoch": 0.807799442896936,
"grad_norm": 3.4048142433166504,
"learning_rate": 3.653667595171773e-05,
"loss": 1.0715,
"step": 1450
},
{
"epoch": 0.8133704735376045,
"grad_norm": 4.1536736488342285,
"learning_rate": 3.6443825441039925e-05,
"loss": 0.9155,
"step": 1460
},
{
"epoch": 0.8189415041782729,
"grad_norm": 3.329630136489868,
"learning_rate": 3.635097493036212e-05,
"loss": 1.0557,
"step": 1470
},
{
"epoch": 0.8245125348189415,
"grad_norm": 2.2894420623779297,
"learning_rate": 3.625812441968431e-05,
"loss": 0.9423,
"step": 1480
},
{
"epoch": 0.83008356545961,
"grad_norm": 2.885784864425659,
"learning_rate": 3.61652739090065e-05,
"loss": 0.9669,
"step": 1490
},
{
"epoch": 0.8356545961002786,
"grad_norm": 3.7429628372192383,
"learning_rate": 3.607242339832869e-05,
"loss": 0.9009,
"step": 1500
},
{
"epoch": 0.841225626740947,
"grad_norm": 2.1827738285064697,
"learning_rate": 3.597957288765089e-05,
"loss": 0.9713,
"step": 1510
},
{
"epoch": 0.8467966573816156,
"grad_norm": 1.2572903633117676,
"learning_rate": 3.5886722376973074e-05,
"loss": 0.9426,
"step": 1520
},
{
"epoch": 0.8523676880222841,
"grad_norm": 4.154526710510254,
"learning_rate": 3.5793871866295266e-05,
"loss": 0.9636,
"step": 1530
},
{
"epoch": 0.8579387186629527,
"grad_norm": 3.4728987216949463,
"learning_rate": 3.570102135561746e-05,
"loss": 1.1009,
"step": 1540
},
{
"epoch": 0.8635097493036211,
"grad_norm": 1.9633491039276123,
"learning_rate": 3.5608170844939645e-05,
"loss": 0.9549,
"step": 1550
},
{
"epoch": 0.8690807799442897,
"grad_norm": 3.696641206741333,
"learning_rate": 3.5515320334261844e-05,
"loss": 0.9211,
"step": 1560
},
{
"epoch": 0.8746518105849582,
"grad_norm": 3.0027382373809814,
"learning_rate": 3.542246982358403e-05,
"loss": 0.8844,
"step": 1570
},
{
"epoch": 0.8802228412256268,
"grad_norm": 1.705255150794983,
"learning_rate": 3.532961931290622e-05,
"loss": 0.8733,
"step": 1580
},
{
"epoch": 0.8857938718662952,
"grad_norm": 4.757926940917969,
"learning_rate": 3.5236768802228415e-05,
"loss": 0.9165,
"step": 1590
},
{
"epoch": 0.8913649025069638,
"grad_norm": 3.9741427898406982,
"learning_rate": 3.51439182915506e-05,
"loss": 0.8411,
"step": 1600
},
{
"epoch": 0.8969359331476323,
"grad_norm": 3.1221024990081787,
"learning_rate": 3.50510677808728e-05,
"loss": 0.8529,
"step": 1610
},
{
"epoch": 0.9025069637883009,
"grad_norm": 5.116636753082275,
"learning_rate": 3.4958217270194986e-05,
"loss": 1.0049,
"step": 1620
},
{
"epoch": 0.9080779944289693,
"grad_norm": 2.5207858085632324,
"learning_rate": 3.4865366759517185e-05,
"loss": 0.8374,
"step": 1630
},
{
"epoch": 0.9136490250696379,
"grad_norm": 4.013001918792725,
"learning_rate": 3.477251624883937e-05,
"loss": 0.8902,
"step": 1640
},
{
"epoch": 0.9192200557103064,
"grad_norm": 5.2283935546875,
"learning_rate": 3.4679665738161556e-05,
"loss": 0.8788,
"step": 1650
},
{
"epoch": 0.924791086350975,
"grad_norm": 2.6664016246795654,
"learning_rate": 3.4586815227483756e-05,
"loss": 0.9204,
"step": 1660
},
{
"epoch": 0.9303621169916435,
"grad_norm": 4.3207292556762695,
"learning_rate": 3.449396471680594e-05,
"loss": 0.9589,
"step": 1670
},
{
"epoch": 0.935933147632312,
"grad_norm": 5.634909152984619,
"learning_rate": 3.440111420612814e-05,
"loss": 0.8883,
"step": 1680
},
{
"epoch": 0.9415041782729805,
"grad_norm": 4.487209320068359,
"learning_rate": 3.430826369545033e-05,
"loss": 0.9111,
"step": 1690
},
{
"epoch": 0.947075208913649,
"grad_norm": 3.9250285625457764,
"learning_rate": 3.421541318477251e-05,
"loss": 0.8759,
"step": 1700
},
{
"epoch": 0.9526462395543176,
"grad_norm": 4.271919250488281,
"learning_rate": 3.412256267409471e-05,
"loss": 1.0486,
"step": 1710
},
{
"epoch": 0.958217270194986,
"grad_norm": 3.3821752071380615,
"learning_rate": 3.40297121634169e-05,
"loss": 0.8707,
"step": 1720
},
{
"epoch": 0.9637883008356546,
"grad_norm": 2.2104811668395996,
"learning_rate": 3.39368616527391e-05,
"loss": 0.8422,
"step": 1730
},
{
"epoch": 0.9693593314763231,
"grad_norm": 2.4628074169158936,
"learning_rate": 3.384401114206128e-05,
"loss": 0.8869,
"step": 1740
},
{
"epoch": 0.9749303621169917,
"grad_norm": 4.211429595947266,
"learning_rate": 3.375116063138347e-05,
"loss": 0.8037,
"step": 1750
},
{
"epoch": 0.9805013927576601,
"grad_norm": 2.9044151306152344,
"learning_rate": 3.365831012070567e-05,
"loss": 0.9166,
"step": 1760
},
{
"epoch": 0.9860724233983287,
"grad_norm": 3.072744846343994,
"learning_rate": 3.3565459610027854e-05,
"loss": 0.9653,
"step": 1770
},
{
"epoch": 0.9916434540389972,
"grad_norm": 3.8640897274017334,
"learning_rate": 3.347260909935005e-05,
"loss": 0.8946,
"step": 1780
},
{
"epoch": 0.9972144846796658,
"grad_norm": 3.481308698654175,
"learning_rate": 3.337975858867224e-05,
"loss": 0.9613,
"step": 1790
},
{
"epoch": 1.0,
"eval_accuracy": 0.6692672053496795,
"eval_loss": 0.9019954800605774,
"eval_runtime": 53.0161,
"eval_samples_per_second": 67.696,
"eval_steps_per_second": 4.244,
"step": 1795
},
{
"epoch": 1.0027855153203342,
"grad_norm": 2.660780191421509,
"learning_rate": 3.328690807799443e-05,
"loss": 0.8447,
"step": 1800
},
{
"epoch": 1.0083565459610029,
"grad_norm": 5.112563133239746,
"learning_rate": 3.3194057567316624e-05,
"loss": 0.7709,
"step": 1810
},
{
"epoch": 1.0139275766016713,
"grad_norm": 3.217790126800537,
"learning_rate": 3.310120705663881e-05,
"loss": 0.6957,
"step": 1820
},
{
"epoch": 1.0194986072423398,
"grad_norm": 2.386460065841675,
"learning_rate": 3.300835654596101e-05,
"loss": 0.5801,
"step": 1830
},
{
"epoch": 1.0250696378830084,
"grad_norm": 3.742408514022827,
"learning_rate": 3.2915506035283195e-05,
"loss": 0.7494,
"step": 1840
},
{
"epoch": 1.0306406685236769,
"grad_norm": 5.288496017456055,
"learning_rate": 3.282265552460539e-05,
"loss": 0.7163,
"step": 1850
},
{
"epoch": 1.0362116991643453,
"grad_norm": 3.0935630798339844,
"learning_rate": 3.272980501392758e-05,
"loss": 0.8476,
"step": 1860
},
{
"epoch": 1.041782729805014,
"grad_norm": 3.274470329284668,
"learning_rate": 3.2636954503249766e-05,
"loss": 0.7659,
"step": 1870
},
{
"epoch": 1.0473537604456824,
"grad_norm": 4.337296485900879,
"learning_rate": 3.2544103992571965e-05,
"loss": 0.8371,
"step": 1880
},
{
"epoch": 1.052924791086351,
"grad_norm": 3.3266661167144775,
"learning_rate": 3.245125348189415e-05,
"loss": 0.7201,
"step": 1890
},
{
"epoch": 1.0584958217270195,
"grad_norm": 2.803584337234497,
"learning_rate": 3.235840297121634e-05,
"loss": 0.7094,
"step": 1900
},
{
"epoch": 1.064066852367688,
"grad_norm": 2.7187163829803467,
"learning_rate": 3.2265552460538536e-05,
"loss": 0.7677,
"step": 1910
},
{
"epoch": 1.0696378830083566,
"grad_norm": 3.310220956802368,
"learning_rate": 3.217270194986073e-05,
"loss": 0.8175,
"step": 1920
},
{
"epoch": 1.075208913649025,
"grad_norm": 3.0058999061584473,
"learning_rate": 3.207985143918292e-05,
"loss": 0.6781,
"step": 1930
},
{
"epoch": 1.0807799442896937,
"grad_norm": 5.092315673828125,
"learning_rate": 3.1987000928505107e-05,
"loss": 0.8581,
"step": 1940
},
{
"epoch": 1.0863509749303621,
"grad_norm": 3.070526599884033,
"learning_rate": 3.18941504178273e-05,
"loss": 0.7415,
"step": 1950
},
{
"epoch": 1.0919220055710306,
"grad_norm": 3.610179901123047,
"learning_rate": 3.180129990714949e-05,
"loss": 0.7457,
"step": 1960
},
{
"epoch": 1.0974930362116992,
"grad_norm": 5.436527729034424,
"learning_rate": 3.1708449396471684e-05,
"loss": 0.8023,
"step": 1970
},
{
"epoch": 1.1030640668523677,
"grad_norm": 2.607541084289551,
"learning_rate": 3.161559888579388e-05,
"loss": 0.7311,
"step": 1980
},
{
"epoch": 1.1086350974930361,
"grad_norm": 3.960480213165283,
"learning_rate": 3.152274837511606e-05,
"loss": 0.825,
"step": 1990
},
{
"epoch": 1.1142061281337048,
"grad_norm": 2.607386589050293,
"learning_rate": 3.1429897864438255e-05,
"loss": 0.7247,
"step": 2000
},
{
"epoch": 1.1197771587743732,
"grad_norm": 4.60785436630249,
"learning_rate": 3.133704735376045e-05,
"loss": 0.6815,
"step": 2010
},
{
"epoch": 1.1253481894150417,
"grad_norm": 2.857929229736328,
"learning_rate": 3.124419684308264e-05,
"loss": 0.8937,
"step": 2020
},
{
"epoch": 1.1309192200557103,
"grad_norm": 3.8009235858917236,
"learning_rate": 3.1151346332404826e-05,
"loss": 0.663,
"step": 2030
},
{
"epoch": 1.1364902506963788,
"grad_norm": 2.8590617179870605,
"learning_rate": 3.105849582172702e-05,
"loss": 0.7552,
"step": 2040
},
{
"epoch": 1.1420612813370474,
"grad_norm": 2.73160719871521,
"learning_rate": 3.096564531104921e-05,
"loss": 0.6979,
"step": 2050
},
{
"epoch": 1.1476323119777159,
"grad_norm": 3.1918087005615234,
"learning_rate": 3.0872794800371404e-05,
"loss": 0.7999,
"step": 2060
},
{
"epoch": 1.1532033426183843,
"grad_norm": 1.336449384689331,
"learning_rate": 3.0779944289693596e-05,
"loss": 0.5635,
"step": 2070
},
{
"epoch": 1.158774373259053,
"grad_norm": 3.852604627609253,
"learning_rate": 3.068709377901578e-05,
"loss": 0.7839,
"step": 2080
},
{
"epoch": 1.1643454038997214,
"grad_norm": 3.373872995376587,
"learning_rate": 3.059424326833798e-05,
"loss": 0.6604,
"step": 2090
},
{
"epoch": 1.16991643454039,
"grad_norm": 5.2556586265563965,
"learning_rate": 3.050139275766017e-05,
"loss": 0.7676,
"step": 2100
},
{
"epoch": 1.1754874651810585,
"grad_norm": 4.4580912590026855,
"learning_rate": 3.0408542246982356e-05,
"loss": 0.7738,
"step": 2110
},
{
"epoch": 1.181058495821727,
"grad_norm": 2.804713726043701,
"learning_rate": 3.0315691736304552e-05,
"loss": 0.6995,
"step": 2120
},
{
"epoch": 1.1866295264623956,
"grad_norm": 3.6438372135162354,
"learning_rate": 3.022284122562674e-05,
"loss": 0.6511,
"step": 2130
},
{
"epoch": 1.192200557103064,
"grad_norm": 2.212045907974243,
"learning_rate": 3.0129990714948937e-05,
"loss": 0.603,
"step": 2140
},
{
"epoch": 1.1977715877437327,
"grad_norm": 3.458812713623047,
"learning_rate": 3.0037140204271123e-05,
"loss": 0.7317,
"step": 2150
},
{
"epoch": 1.2033426183844012,
"grad_norm": 2.6540637016296387,
"learning_rate": 2.9944289693593312e-05,
"loss": 0.7921,
"step": 2160
},
{
"epoch": 1.2089136490250696,
"grad_norm": 5.65538215637207,
"learning_rate": 2.9851439182915508e-05,
"loss": 0.8803,
"step": 2170
},
{
"epoch": 1.2144846796657383,
"grad_norm": 3.8509607315063477,
"learning_rate": 2.9758588672237697e-05,
"loss": 0.7686,
"step": 2180
},
{
"epoch": 1.2200557103064067,
"grad_norm": 4.367794990539551,
"learning_rate": 2.9665738161559893e-05,
"loss": 0.7243,
"step": 2190
},
{
"epoch": 1.2256267409470751,
"grad_norm": 3.6232688426971436,
"learning_rate": 2.957288765088208e-05,
"loss": 0.6483,
"step": 2200
},
{
"epoch": 1.2311977715877438,
"grad_norm": 3.4824819564819336,
"learning_rate": 2.9480037140204275e-05,
"loss": 0.8249,
"step": 2210
},
{
"epoch": 1.2367688022284122,
"grad_norm": 5.086853981018066,
"learning_rate": 2.9387186629526464e-05,
"loss": 0.7303,
"step": 2220
},
{
"epoch": 1.2423398328690807,
"grad_norm": 3.1858561038970947,
"learning_rate": 2.9294336118848653e-05,
"loss": 0.6699,
"step": 2230
},
{
"epoch": 1.2479108635097493,
"grad_norm": 5.4035139083862305,
"learning_rate": 2.920148560817085e-05,
"loss": 0.8184,
"step": 2240
},
{
"epoch": 1.2534818941504178,
"grad_norm": 4.2386345863342285,
"learning_rate": 2.9108635097493035e-05,
"loss": 0.783,
"step": 2250
},
{
"epoch": 1.2590529247910864,
"grad_norm": 5.696313858032227,
"learning_rate": 2.901578458681523e-05,
"loss": 0.7901,
"step": 2260
},
{
"epoch": 1.2646239554317549,
"grad_norm": 2.38380765914917,
"learning_rate": 2.892293407613742e-05,
"loss": 0.7271,
"step": 2270
},
{
"epoch": 1.2701949860724233,
"grad_norm": 2.800536632537842,
"learning_rate": 2.883008356545961e-05,
"loss": 0.7253,
"step": 2280
},
{
"epoch": 1.275766016713092,
"grad_norm": 5.40772008895874,
"learning_rate": 2.8737233054781805e-05,
"loss": 0.8461,
"step": 2290
},
{
"epoch": 1.2813370473537604,
"grad_norm": 2.7099337577819824,
"learning_rate": 2.864438254410399e-05,
"loss": 0.7545,
"step": 2300
},
{
"epoch": 1.286908077994429,
"grad_norm": 5.34475040435791,
"learning_rate": 2.8551532033426187e-05,
"loss": 0.6853,
"step": 2310
},
{
"epoch": 1.2924791086350975,
"grad_norm": 3.6724138259887695,
"learning_rate": 2.8458681522748376e-05,
"loss": 0.6719,
"step": 2320
},
{
"epoch": 1.298050139275766,
"grad_norm": 6.54721212387085,
"learning_rate": 2.8365831012070565e-05,
"loss": 0.6367,
"step": 2330
},
{
"epoch": 1.3036211699164346,
"grad_norm": 4.257522106170654,
"learning_rate": 2.827298050139276e-05,
"loss": 0.8234,
"step": 2340
},
{
"epoch": 1.309192200557103,
"grad_norm": 4.256605625152588,
"learning_rate": 2.8180129990714947e-05,
"loss": 0.8149,
"step": 2350
},
{
"epoch": 1.3147632311977717,
"grad_norm": 3.181926965713501,
"learning_rate": 2.8087279480037143e-05,
"loss": 0.7522,
"step": 2360
},
{
"epoch": 1.3203342618384402,
"grad_norm": 2.8374698162078857,
"learning_rate": 2.7994428969359332e-05,
"loss": 0.6871,
"step": 2370
},
{
"epoch": 1.3259052924791086,
"grad_norm": 7.006384372711182,
"learning_rate": 2.7901578458681528e-05,
"loss": 0.7822,
"step": 2380
},
{
"epoch": 1.331476323119777,
"grad_norm": 5.170847415924072,
"learning_rate": 2.7808727948003717e-05,
"loss": 0.8323,
"step": 2390
},
{
"epoch": 1.3370473537604457,
"grad_norm": 4.078238010406494,
"learning_rate": 2.7715877437325903e-05,
"loss": 0.6984,
"step": 2400
},
{
"epoch": 1.3426183844011141,
"grad_norm": 3.4335429668426514,
"learning_rate": 2.76230269266481e-05,
"loss": 0.6943,
"step": 2410
},
{
"epoch": 1.3481894150417828,
"grad_norm": 3.247312545776367,
"learning_rate": 2.7530176415970288e-05,
"loss": 0.7341,
"step": 2420
},
{
"epoch": 1.3537604456824512,
"grad_norm": 5.7520318031311035,
"learning_rate": 2.7437325905292484e-05,
"loss": 0.7103,
"step": 2430
},
{
"epoch": 1.3593314763231197,
"grad_norm": 1.9237465858459473,
"learning_rate": 2.735376044568245e-05,
"loss": 0.7064,
"step": 2440
},
{
"epoch": 1.3649025069637883,
"grad_norm": 3.3437535762786865,
"learning_rate": 2.726090993500464e-05,
"loss": 0.7222,
"step": 2450
},
{
"epoch": 1.3704735376044568,
"grad_norm": 3.8634493350982666,
"learning_rate": 2.7168059424326836e-05,
"loss": 0.5842,
"step": 2460
},
{
"epoch": 1.3760445682451254,
"grad_norm": 4.395915508270264,
"learning_rate": 2.7075208913649025e-05,
"loss": 0.8458,
"step": 2470
},
{
"epoch": 1.3816155988857939,
"grad_norm": 4.998650074005127,
"learning_rate": 2.6982358402971218e-05,
"loss": 0.7941,
"step": 2480
},
{
"epoch": 1.3871866295264623,
"grad_norm": 5.347882270812988,
"learning_rate": 2.6889507892293407e-05,
"loss": 0.728,
"step": 2490
},
{
"epoch": 1.392757660167131,
"grad_norm": 4.277406692504883,
"learning_rate": 2.6796657381615596e-05,
"loss": 0.7203,
"step": 2500
},
{
"epoch": 1.3983286908077994,
"grad_norm": 3.2961020469665527,
"learning_rate": 2.6703806870937792e-05,
"loss": 0.675,
"step": 2510
},
{
"epoch": 1.403899721448468,
"grad_norm": 3.227165460586548,
"learning_rate": 2.661095636025998e-05,
"loss": 0.8199,
"step": 2520
},
{
"epoch": 1.4094707520891365,
"grad_norm": 2.727691650390625,
"learning_rate": 2.6518105849582174e-05,
"loss": 0.6727,
"step": 2530
},
{
"epoch": 1.415041782729805,
"grad_norm": 3.660712957382202,
"learning_rate": 2.6425255338904363e-05,
"loss": 0.7168,
"step": 2540
},
{
"epoch": 1.4206128133704734,
"grad_norm": 4.9975266456604,
"learning_rate": 2.633240482822656e-05,
"loss": 0.7329,
"step": 2550
},
{
"epoch": 1.426183844011142,
"grad_norm": 4.956869602203369,
"learning_rate": 2.6239554317548748e-05,
"loss": 0.7576,
"step": 2560
},
{
"epoch": 1.4317548746518105,
"grad_norm": 3.2841885089874268,
"learning_rate": 2.6146703806870937e-05,
"loss": 0.687,
"step": 2570
},
{
"epoch": 1.4373259052924792,
"grad_norm": 5.767863750457764,
"learning_rate": 2.605385329619313e-05,
"loss": 0.7346,
"step": 2580
},
{
"epoch": 1.4428969359331476,
"grad_norm": 2.8011934757232666,
"learning_rate": 2.596100278551532e-05,
"loss": 0.7072,
"step": 2590
},
{
"epoch": 1.448467966573816,
"grad_norm": 2.4750354290008545,
"learning_rate": 2.5868152274837515e-05,
"loss": 0.5924,
"step": 2600
},
{
"epoch": 1.4540389972144847,
"grad_norm": 2.4504594802856445,
"learning_rate": 2.5775301764159704e-05,
"loss": 0.6229,
"step": 2610
},
{
"epoch": 1.4596100278551531,
"grad_norm": 6.350492000579834,
"learning_rate": 2.5682451253481893e-05,
"loss": 0.7909,
"step": 2620
},
{
"epoch": 1.4651810584958218,
"grad_norm": 3.0052127838134766,
"learning_rate": 2.5589600742804086e-05,
"loss": 0.6516,
"step": 2630
},
{
"epoch": 1.4707520891364902,
"grad_norm": 3.0360071659088135,
"learning_rate": 2.5496750232126275e-05,
"loss": 0.7202,
"step": 2640
},
{
"epoch": 1.4763231197771587,
"grad_norm": 5.270174026489258,
"learning_rate": 2.540389972144847e-05,
"loss": 0.7491,
"step": 2650
},
{
"epoch": 1.4818941504178273,
"grad_norm": 6.691749572753906,
"learning_rate": 2.531104921077066e-05,
"loss": 0.6879,
"step": 2660
},
{
"epoch": 1.4874651810584958,
"grad_norm": 3.4790680408477783,
"learning_rate": 2.5218198700092853e-05,
"loss": 0.7869,
"step": 2670
},
{
"epoch": 1.4930362116991645,
"grad_norm": 4.1881914138793945,
"learning_rate": 2.5125348189415042e-05,
"loss": 0.7038,
"step": 2680
},
{
"epoch": 1.498607242339833,
"grad_norm": 3.5954504013061523,
"learning_rate": 2.503249767873723e-05,
"loss": 0.8112,
"step": 2690
},
{
"epoch": 1.5041782729805013,
"grad_norm": 5.3881354331970215,
"learning_rate": 2.4939647168059427e-05,
"loss": 0.8329,
"step": 2700
},
{
"epoch": 1.5097493036211698,
"grad_norm": 2.5170273780822754,
"learning_rate": 2.4846796657381616e-05,
"loss": 0.5086,
"step": 2710
},
{
"epoch": 1.5153203342618384,
"grad_norm": 2.370330333709717,
"learning_rate": 2.475394614670381e-05,
"loss": 0.6625,
"step": 2720
},
{
"epoch": 1.520891364902507,
"grad_norm": 2.416132926940918,
"learning_rate": 2.4661095636025998e-05,
"loss": 0.7835,
"step": 2730
},
{
"epoch": 1.5264623955431755,
"grad_norm": 4.824071407318115,
"learning_rate": 2.456824512534819e-05,
"loss": 0.6605,
"step": 2740
},
{
"epoch": 1.532033426183844,
"grad_norm": 5.555083274841309,
"learning_rate": 2.4475394614670383e-05,
"loss": 0.675,
"step": 2750
},
{
"epoch": 1.5376044568245124,
"grad_norm": 5.953351020812988,
"learning_rate": 2.4382544103992576e-05,
"loss": 0.6572,
"step": 2760
},
{
"epoch": 1.543175487465181,
"grad_norm": 3.702897787094116,
"learning_rate": 2.4289693593314765e-05,
"loss": 0.7674,
"step": 2770
},
{
"epoch": 1.5487465181058497,
"grad_norm": 3.995495080947876,
"learning_rate": 2.4196843082636954e-05,
"loss": 0.7461,
"step": 2780
},
{
"epoch": 1.5543175487465182,
"grad_norm": 4.654627323150635,
"learning_rate": 2.4103992571959146e-05,
"loss": 0.7709,
"step": 2790
},
{
"epoch": 1.5598885793871866,
"grad_norm": 4.607250690460205,
"learning_rate": 2.401114206128134e-05,
"loss": 0.8664,
"step": 2800
},
{
"epoch": 1.565459610027855,
"grad_norm": 4.200578212738037,
"learning_rate": 2.391829155060353e-05,
"loss": 0.7295,
"step": 2810
},
{
"epoch": 1.5710306406685237,
"grad_norm": 3.5812277793884277,
"learning_rate": 2.382544103992572e-05,
"loss": 0.7653,
"step": 2820
},
{
"epoch": 1.5766016713091922,
"grad_norm": 3.6298224925994873,
"learning_rate": 2.373259052924791e-05,
"loss": 0.6548,
"step": 2830
},
{
"epoch": 1.5821727019498608,
"grad_norm": 3.4680569171905518,
"learning_rate": 2.3639740018570102e-05,
"loss": 0.699,
"step": 2840
},
{
"epoch": 1.5877437325905293,
"grad_norm": 3.139227867126465,
"learning_rate": 2.3546889507892295e-05,
"loss": 0.6505,
"step": 2850
},
{
"epoch": 1.5933147632311977,
"grad_norm": 4.027279853820801,
"learning_rate": 2.3454038997214488e-05,
"loss": 0.7388,
"step": 2860
},
{
"epoch": 1.5988857938718661,
"grad_norm": 3.547569990158081,
"learning_rate": 2.3361188486536677e-05,
"loss": 0.642,
"step": 2870
},
{
"epoch": 1.6044568245125348,
"grad_norm": 5.281991004943848,
"learning_rate": 2.3268337975858866e-05,
"loss": 0.6009,
"step": 2880
},
{
"epoch": 1.6100278551532035,
"grad_norm": 3.9563615322113037,
"learning_rate": 2.317548746518106e-05,
"loss": 0.695,
"step": 2890
},
{
"epoch": 1.615598885793872,
"grad_norm": 5.508429050445557,
"learning_rate": 2.308263695450325e-05,
"loss": 0.7169,
"step": 2900
},
{
"epoch": 1.6211699164345403,
"grad_norm": 4.158377647399902,
"learning_rate": 2.2989786443825444e-05,
"loss": 0.6337,
"step": 2910
},
{
"epoch": 1.6267409470752088,
"grad_norm": 4.293817520141602,
"learning_rate": 2.2896935933147633e-05,
"loss": 0.6769,
"step": 2920
},
{
"epoch": 1.6323119777158774,
"grad_norm": 3.0530028343200684,
"learning_rate": 2.2804085422469825e-05,
"loss": 0.6436,
"step": 2930
},
{
"epoch": 1.637883008356546,
"grad_norm": 3.3768982887268066,
"learning_rate": 2.2711234911792014e-05,
"loss": 0.6443,
"step": 2940
},
{
"epoch": 1.6434540389972145,
"grad_norm": 6.057602882385254,
"learning_rate": 2.2618384401114207e-05,
"loss": 0.6208,
"step": 2950
},
{
"epoch": 1.649025069637883,
"grad_norm": 7.647638320922852,
"learning_rate": 2.25255338904364e-05,
"loss": 0.8125,
"step": 2960
},
{
"epoch": 1.6545961002785514,
"grad_norm": 6.915018081665039,
"learning_rate": 2.243268337975859e-05,
"loss": 0.8279,
"step": 2970
},
{
"epoch": 1.66016713091922,
"grad_norm": 3.073479413986206,
"learning_rate": 2.233983286908078e-05,
"loss": 0.5983,
"step": 2980
},
{
"epoch": 1.6657381615598887,
"grad_norm": 6.27379035949707,
"learning_rate": 2.2246982358402974e-05,
"loss": 0.7413,
"step": 2990
},
{
"epoch": 1.6713091922005572,
"grad_norm": 2.8790829181671143,
"learning_rate": 2.2154131847725163e-05,
"loss": 0.5785,
"step": 3000
},
{
"epoch": 1.6768802228412256,
"grad_norm": 3.810025691986084,
"learning_rate": 2.2061281337047355e-05,
"loss": 0.6187,
"step": 3010
},
{
"epoch": 1.682451253481894,
"grad_norm": 4.597643852233887,
"learning_rate": 2.1968430826369545e-05,
"loss": 0.7871,
"step": 3020
},
{
"epoch": 1.6880222841225627,
"grad_norm": 4.470458507537842,
"learning_rate": 2.1875580315691737e-05,
"loss": 0.7397,
"step": 3030
},
{
"epoch": 1.6935933147632312,
"grad_norm": 3.957707643508911,
"learning_rate": 2.178272980501393e-05,
"loss": 0.6893,
"step": 3040
},
{
"epoch": 1.6991643454038998,
"grad_norm": 4.973650932312012,
"learning_rate": 2.1689879294336122e-05,
"loss": 0.601,
"step": 3050
},
{
"epoch": 1.7047353760445683,
"grad_norm": 3.1551012992858887,
"learning_rate": 2.159702878365831e-05,
"loss": 0.6792,
"step": 3060
},
{
"epoch": 1.7103064066852367,
"grad_norm": 21.842958450317383,
"learning_rate": 2.15041782729805e-05,
"loss": 0.7582,
"step": 3070
},
{
"epoch": 1.7158774373259051,
"grad_norm": 6.199334144592285,
"learning_rate": 2.1411327762302693e-05,
"loss": 0.635,
"step": 3080
},
{
"epoch": 1.7214484679665738,
"grad_norm": 5.763338088989258,
"learning_rate": 2.1318477251624886e-05,
"loss": 0.7636,
"step": 3090
},
{
"epoch": 1.7270194986072425,
"grad_norm": 3.1481924057006836,
"learning_rate": 2.1225626740947078e-05,
"loss": 0.7613,
"step": 3100
},
{
"epoch": 1.732590529247911,
"grad_norm": 2.0901215076446533,
"learning_rate": 2.1132776230269267e-05,
"loss": 0.6326,
"step": 3110
},
{
"epoch": 1.7381615598885793,
"grad_norm": 3.8175764083862305,
"learning_rate": 2.1039925719591457e-05,
"loss": 0.668,
"step": 3120
},
{
"epoch": 1.7437325905292478,
"grad_norm": 5.669551849365234,
"learning_rate": 2.094707520891365e-05,
"loss": 0.6897,
"step": 3130
},
{
"epoch": 1.7493036211699164,
"grad_norm": 2.3246147632598877,
"learning_rate": 2.085422469823584e-05,
"loss": 0.9081,
"step": 3140
},
{
"epoch": 1.754874651810585,
"grad_norm": 4.618974685668945,
"learning_rate": 2.0761374187558034e-05,
"loss": 0.7226,
"step": 3150
},
{
"epoch": 1.7604456824512535,
"grad_norm": 3.3900840282440186,
"learning_rate": 2.0668523676880223e-05,
"loss": 0.719,
"step": 3160
},
{
"epoch": 1.766016713091922,
"grad_norm": 6.443446159362793,
"learning_rate": 2.0575673166202413e-05,
"loss": 0.6676,
"step": 3170
},
{
"epoch": 1.7715877437325904,
"grad_norm": 3.93987774848938,
"learning_rate": 2.0482822655524605e-05,
"loss": 0.6198,
"step": 3180
},
{
"epoch": 1.777158774373259,
"grad_norm": 3.811434030532837,
"learning_rate": 2.0389972144846798e-05,
"loss": 0.7962,
"step": 3190
},
{
"epoch": 1.7827298050139275,
"grad_norm": 6.011105537414551,
"learning_rate": 2.029712163416899e-05,
"loss": 0.6282,
"step": 3200
},
{
"epoch": 1.7883008356545962,
"grad_norm": 4.521935939788818,
"learning_rate": 2.020427112349118e-05,
"loss": 0.6502,
"step": 3210
},
{
"epoch": 1.7938718662952646,
"grad_norm": 6.527938365936279,
"learning_rate": 2.0111420612813372e-05,
"loss": 0.731,
"step": 3220
},
{
"epoch": 1.799442896935933,
"grad_norm": 3.292177438735962,
"learning_rate": 2.001857010213556e-05,
"loss": 0.6674,
"step": 3230
},
{
"epoch": 1.8050139275766015,
"grad_norm": 4.2430548667907715,
"learning_rate": 1.9925719591457754e-05,
"loss": 0.6192,
"step": 3240
},
{
"epoch": 1.8105849582172702,
"grad_norm": 2.8606340885162354,
"learning_rate": 1.9832869080779946e-05,
"loss": 0.724,
"step": 3250
},
{
"epoch": 1.8161559888579388,
"grad_norm": 5.849732398986816,
"learning_rate": 1.9740018570102135e-05,
"loss": 0.6134,
"step": 3260
},
{
"epoch": 1.8217270194986073,
"grad_norm": 2.940549850463867,
"learning_rate": 1.9647168059424328e-05,
"loss": 0.5696,
"step": 3270
},
{
"epoch": 1.8272980501392757,
"grad_norm": 5.734061241149902,
"learning_rate": 1.955431754874652e-05,
"loss": 0.6265,
"step": 3280
},
{
"epoch": 1.8328690807799441,
"grad_norm": 4.678475379943848,
"learning_rate": 1.946146703806871e-05,
"loss": 0.543,
"step": 3290
},
{
"epoch": 1.8384401114206128,
"grad_norm": 5.988973617553711,
"learning_rate": 1.9368616527390902e-05,
"loss": 0.5986,
"step": 3300
},
{
"epoch": 1.8440111420612815,
"grad_norm": 5.648441314697266,
"learning_rate": 1.927576601671309e-05,
"loss": 0.7231,
"step": 3310
},
{
"epoch": 1.84958217270195,
"grad_norm": 3.596468448638916,
"learning_rate": 1.9182915506035284e-05,
"loss": 0.7182,
"step": 3320
},
{
"epoch": 1.8551532033426184,
"grad_norm": 3.6702165603637695,
"learning_rate": 1.9090064995357476e-05,
"loss": 0.6835,
"step": 3330
},
{
"epoch": 1.8607242339832868,
"grad_norm": 5.897074222564697,
"learning_rate": 1.899721448467967e-05,
"loss": 0.6812,
"step": 3340
},
{
"epoch": 1.8662952646239555,
"grad_norm": 5.111124515533447,
"learning_rate": 1.8904363974001858e-05,
"loss": 0.7924,
"step": 3350
},
{
"epoch": 1.8718662952646241,
"grad_norm": 5.206398010253906,
"learning_rate": 1.8811513463324047e-05,
"loss": 0.6493,
"step": 3360
},
{
"epoch": 1.8774373259052926,
"grad_norm": 3.4764597415924072,
"learning_rate": 1.871866295264624e-05,
"loss": 0.6969,
"step": 3370
},
{
"epoch": 1.883008356545961,
"grad_norm": 3.6830670833587646,
"learning_rate": 1.8625812441968432e-05,
"loss": 0.6874,
"step": 3380
},
{
"epoch": 1.8885793871866294,
"grad_norm": 5.803642272949219,
"learning_rate": 1.8532961931290625e-05,
"loss": 0.5756,
"step": 3390
},
{
"epoch": 1.894150417827298,
"grad_norm": 5.948879718780518,
"learning_rate": 1.8440111420612814e-05,
"loss": 0.741,
"step": 3400
},
{
"epoch": 1.8997214484679665,
"grad_norm": 5.844179630279541,
"learning_rate": 1.8347260909935003e-05,
"loss": 0.5393,
"step": 3410
},
{
"epoch": 1.9052924791086352,
"grad_norm": 4.613275051116943,
"learning_rate": 1.8254410399257196e-05,
"loss": 0.8167,
"step": 3420
},
{
"epoch": 1.9108635097493036,
"grad_norm": 4.484188079833984,
"learning_rate": 1.816155988857939e-05,
"loss": 0.6047,
"step": 3430
},
{
"epoch": 1.916434540389972,
"grad_norm": 5.882885932922363,
"learning_rate": 1.806870937790158e-05,
"loss": 0.5126,
"step": 3440
},
{
"epoch": 1.9220055710306405,
"grad_norm": 4.281798839569092,
"learning_rate": 1.797585886722377e-05,
"loss": 0.6512,
"step": 3450
},
{
"epoch": 1.9275766016713092,
"grad_norm": 7.347241401672363,
"learning_rate": 1.788300835654596e-05,
"loss": 0.7092,
"step": 3460
},
{
"epoch": 1.9331476323119778,
"grad_norm": 5.490599155426025,
"learning_rate": 1.7790157845868152e-05,
"loss": 0.7738,
"step": 3470
},
{
"epoch": 1.9387186629526463,
"grad_norm": 4.461576461791992,
"learning_rate": 1.7697307335190344e-05,
"loss": 0.7116,
"step": 3480
},
{
"epoch": 1.9442896935933147,
"grad_norm": 3.0554678440093994,
"learning_rate": 1.7604456824512537e-05,
"loss": 0.7397,
"step": 3490
},
{
"epoch": 1.9498607242339832,
"grad_norm": 3.393239736557007,
"learning_rate": 1.7511606313834726e-05,
"loss": 0.7141,
"step": 3500
},
{
"epoch": 1.9554317548746518,
"grad_norm": 3.1929807662963867,
"learning_rate": 1.74280408542247e-05,
"loss": 0.6955,
"step": 3510
},
{
"epoch": 1.9610027855153205,
"grad_norm": 5.299395561218262,
"learning_rate": 1.733519034354689e-05,
"loss": 0.6491,
"step": 3520
},
{
"epoch": 1.966573816155989,
"grad_norm": 3.9555628299713135,
"learning_rate": 1.7242339832869082e-05,
"loss": 0.6156,
"step": 3530
},
{
"epoch": 1.9721448467966574,
"grad_norm": 4.251757621765137,
"learning_rate": 1.7149489322191274e-05,
"loss": 0.6155,
"step": 3540
},
{
"epoch": 1.9777158774373258,
"grad_norm": 7.542901515960693,
"learning_rate": 1.7056638811513463e-05,
"loss": 0.7173,
"step": 3550
},
{
"epoch": 1.9832869080779945,
"grad_norm": 4.958390712738037,
"learning_rate": 1.6963788300835656e-05,
"loss": 0.5184,
"step": 3560
},
{
"epoch": 1.988857938718663,
"grad_norm": 4.884737491607666,
"learning_rate": 1.687093779015785e-05,
"loss": 0.6729,
"step": 3570
},
{
"epoch": 1.9944289693593316,
"grad_norm": 4.648261547088623,
"learning_rate": 1.6778087279480038e-05,
"loss": 0.6181,
"step": 3580
},
{
"epoch": 2.0,
"grad_norm": 6.863659381866455,
"learning_rate": 1.668523676880223e-05,
"loss": 0.5198,
"step": 3590
},
{
"epoch": 2.0,
"eval_accuracy": 0.7071607690164391,
"eval_loss": 0.817263662815094,
"eval_runtime": 37.2879,
"eval_samples_per_second": 96.251,
"eval_steps_per_second": 6.034,
"step": 3590
},
{
"epoch": 2.0055710306406684,
"grad_norm": 4.615173816680908,
"learning_rate": 1.659238625812442e-05,
"loss": 0.5817,
"step": 3600
},
{
"epoch": 2.011142061281337,
"grad_norm": 2.830014944076538,
"learning_rate": 1.6499535747446612e-05,
"loss": 0.4535,
"step": 3610
},
{
"epoch": 2.0167130919220058,
"grad_norm": 4.435030460357666,
"learning_rate": 1.6406685236768805e-05,
"loss": 0.4874,
"step": 3620
},
{
"epoch": 2.022284122562674,
"grad_norm": 6.543429374694824,
"learning_rate": 1.6313834726090994e-05,
"loss": 0.5775,
"step": 3630
},
{
"epoch": 2.0278551532033426,
"grad_norm": 4.525574684143066,
"learning_rate": 1.6220984215413186e-05,
"loss": 0.6027,
"step": 3640
},
{
"epoch": 2.033426183844011,
"grad_norm": 10.019755363464355,
"learning_rate": 1.6128133704735375e-05,
"loss": 0.4922,
"step": 3650
},
{
"epoch": 2.0389972144846795,
"grad_norm": 4.950735569000244,
"learning_rate": 1.6035283194057568e-05,
"loss": 0.4555,
"step": 3660
},
{
"epoch": 2.0445682451253484,
"grad_norm": 2.9464316368103027,
"learning_rate": 1.594243268337976e-05,
"loss": 0.454,
"step": 3670
},
{
"epoch": 2.050139275766017,
"grad_norm": 10.776598930358887,
"learning_rate": 1.5849582172701953e-05,
"loss": 0.4013,
"step": 3680
},
{
"epoch": 2.0557103064066853,
"grad_norm": 4.557794094085693,
"learning_rate": 1.5756731662024142e-05,
"loss": 0.4761,
"step": 3690
},
{
"epoch": 2.0612813370473537,
"grad_norm": 4.7235107421875,
"learning_rate": 1.566388115134633e-05,
"loss": 0.4825,
"step": 3700
},
{
"epoch": 2.066852367688022,
"grad_norm": 3.367194414138794,
"learning_rate": 1.5571030640668524e-05,
"loss": 0.5587,
"step": 3710
},
{
"epoch": 2.0724233983286906,
"grad_norm": 4.634751319885254,
"learning_rate": 1.5478180129990717e-05,
"loss": 0.5894,
"step": 3720
},
{
"epoch": 2.0779944289693595,
"grad_norm": 3.5962271690368652,
"learning_rate": 1.538532961931291e-05,
"loss": 0.5022,
"step": 3730
},
{
"epoch": 2.083565459610028,
"grad_norm": 2.4621849060058594,
"learning_rate": 1.5292479108635098e-05,
"loss": 0.3771,
"step": 3740
},
{
"epoch": 2.0891364902506964,
"grad_norm": 0.884714663028717,
"learning_rate": 1.5199628597957289e-05,
"loss": 0.4954,
"step": 3750
},
{
"epoch": 2.094707520891365,
"grad_norm": 3.478480577468872,
"learning_rate": 1.510677808727948e-05,
"loss": 0.504,
"step": 3760
},
{
"epoch": 2.1002785515320332,
"grad_norm": 4.723543643951416,
"learning_rate": 1.5013927576601673e-05,
"loss": 0.4238,
"step": 3770
},
{
"epoch": 2.105849582172702,
"grad_norm": 6.865334510803223,
"learning_rate": 1.4921077065923863e-05,
"loss": 0.5369,
"step": 3780
},
{
"epoch": 2.1114206128133706,
"grad_norm": 3.348172664642334,
"learning_rate": 1.4828226555246056e-05,
"loss": 0.4709,
"step": 3790
},
{
"epoch": 2.116991643454039,
"grad_norm": 3.8873541355133057,
"learning_rate": 1.4735376044568247e-05,
"loss": 0.3504,
"step": 3800
},
{
"epoch": 2.1225626740947074,
"grad_norm": 6.476009845733643,
"learning_rate": 1.4642525533890436e-05,
"loss": 0.378,
"step": 3810
},
{
"epoch": 2.128133704735376,
"grad_norm": 4.0402445793151855,
"learning_rate": 1.4549675023212628e-05,
"loss": 0.4167,
"step": 3820
},
{
"epoch": 2.1337047353760448,
"grad_norm": 5.197554588317871,
"learning_rate": 1.445682451253482e-05,
"loss": 0.5053,
"step": 3830
},
{
"epoch": 2.139275766016713,
"grad_norm": 3.450927495956421,
"learning_rate": 1.4363974001857012e-05,
"loss": 0.4532,
"step": 3840
},
{
"epoch": 2.1448467966573816,
"grad_norm": 4.431534767150879,
"learning_rate": 1.4271123491179203e-05,
"loss": 0.5658,
"step": 3850
},
{
"epoch": 2.15041782729805,
"grad_norm": 3.2116923332214355,
"learning_rate": 1.4178272980501395e-05,
"loss": 0.3628,
"step": 3860
},
{
"epoch": 2.1559888579387185,
"grad_norm": 3.2377572059631348,
"learning_rate": 1.4085422469823584e-05,
"loss": 0.3513,
"step": 3870
},
{
"epoch": 2.1615598885793874,
"grad_norm": 6.522520065307617,
"learning_rate": 1.3992571959145775e-05,
"loss": 0.4474,
"step": 3880
},
{
"epoch": 2.167130919220056,
"grad_norm": 4.152089595794678,
"learning_rate": 1.3899721448467968e-05,
"loss": 0.4717,
"step": 3890
},
{
"epoch": 2.1727019498607243,
"grad_norm": 3.9194087982177734,
"learning_rate": 1.3806870937790159e-05,
"loss": 0.4681,
"step": 3900
},
{
"epoch": 2.1782729805013927,
"grad_norm": 3.955049991607666,
"learning_rate": 1.3714020427112351e-05,
"loss": 0.4825,
"step": 3910
},
{
"epoch": 2.183844011142061,
"grad_norm": 7.73003625869751,
"learning_rate": 1.362116991643454e-05,
"loss": 0.4505,
"step": 3920
},
{
"epoch": 2.1894150417827296,
"grad_norm": 7.201216220855713,
"learning_rate": 1.3528319405756731e-05,
"loss": 0.5176,
"step": 3930
},
{
"epoch": 2.1949860724233985,
"grad_norm": 4.639294624328613,
"learning_rate": 1.3435468895078924e-05,
"loss": 0.4535,
"step": 3940
},
{
"epoch": 2.200557103064067,
"grad_norm": 4.728760242462158,
"learning_rate": 1.3342618384401115e-05,
"loss": 0.5659,
"step": 3950
},
{
"epoch": 2.2061281337047354,
"grad_norm": 2.1953437328338623,
"learning_rate": 1.3249767873723307e-05,
"loss": 0.3706,
"step": 3960
},
{
"epoch": 2.211699164345404,
"grad_norm": 2.9534456729888916,
"learning_rate": 1.3156917363045498e-05,
"loss": 0.5308,
"step": 3970
},
{
"epoch": 2.2172701949860723,
"grad_norm": 6.660586833953857,
"learning_rate": 1.3064066852367687e-05,
"loss": 0.4657,
"step": 3980
},
{
"epoch": 2.222841225626741,
"grad_norm": 3.0846714973449707,
"learning_rate": 1.297121634168988e-05,
"loss": 0.4446,
"step": 3990
},
{
"epoch": 2.2284122562674096,
"grad_norm": 9.60133171081543,
"learning_rate": 1.287836583101207e-05,
"loss": 0.5949,
"step": 4000
},
{
"epoch": 2.233983286908078,
"grad_norm": 2.2764580249786377,
"learning_rate": 1.2785515320334263e-05,
"loss": 0.4514,
"step": 4010
},
{
"epoch": 2.2395543175487465,
"grad_norm": 2.598289728164673,
"learning_rate": 1.2692664809656454e-05,
"loss": 0.4929,
"step": 4020
},
{
"epoch": 2.245125348189415,
"grad_norm": 2.7438085079193115,
"learning_rate": 1.2599814298978647e-05,
"loss": 0.5142,
"step": 4030
},
{
"epoch": 2.2506963788300833,
"grad_norm": 2.9122653007507324,
"learning_rate": 1.2506963788300836e-05,
"loss": 0.4944,
"step": 4040
},
{
"epoch": 2.256267409470752,
"grad_norm": 4.093932151794434,
"learning_rate": 1.2414113277623028e-05,
"loss": 0.4334,
"step": 4050
},
{
"epoch": 2.2618384401114207,
"grad_norm": 5.076513290405273,
"learning_rate": 1.232126276694522e-05,
"loss": 0.4213,
"step": 4060
},
{
"epoch": 2.267409470752089,
"grad_norm": 3.2467005252838135,
"learning_rate": 1.222841225626741e-05,
"loss": 0.6094,
"step": 4070
},
{
"epoch": 2.2729805013927575,
"grad_norm": 2.402123212814331,
"learning_rate": 1.2135561745589603e-05,
"loss": 0.4469,
"step": 4080
},
{
"epoch": 2.2785515320334264,
"grad_norm": 4.679557800292969,
"learning_rate": 1.2042711234911792e-05,
"loss": 0.4029,
"step": 4090
},
{
"epoch": 2.284122562674095,
"grad_norm": 1.4909316301345825,
"learning_rate": 1.1949860724233984e-05,
"loss": 0.3351,
"step": 4100
},
{
"epoch": 2.2896935933147633,
"grad_norm": 6.870358467102051,
"learning_rate": 1.1857010213556175e-05,
"loss": 0.5222,
"step": 4110
},
{
"epoch": 2.2952646239554317,
"grad_norm": 1.7165311574935913,
"learning_rate": 1.1764159702878366e-05,
"loss": 0.326,
"step": 4120
},
{
"epoch": 2.3008356545961,
"grad_norm": 5.359385967254639,
"learning_rate": 1.1671309192200559e-05,
"loss": 0.36,
"step": 4130
},
{
"epoch": 2.3064066852367686,
"grad_norm": 4.74082612991333,
"learning_rate": 1.1578458681522748e-05,
"loss": 0.5068,
"step": 4140
},
{
"epoch": 2.3119777158774375,
"grad_norm": 2.6061348915100098,
"learning_rate": 1.148560817084494e-05,
"loss": 0.4599,
"step": 4150
},
{
"epoch": 2.317548746518106,
"grad_norm": 6.639983177185059,
"learning_rate": 1.1392757660167131e-05,
"loss": 0.5885,
"step": 4160
},
{
"epoch": 2.3231197771587744,
"grad_norm": 3.9153246879577637,
"learning_rate": 1.1299907149489322e-05,
"loss": 0.418,
"step": 4170
},
{
"epoch": 2.328690807799443,
"grad_norm": 7.114092826843262,
"learning_rate": 1.1207056638811515e-05,
"loss": 0.4293,
"step": 4180
},
{
"epoch": 2.3342618384401113,
"grad_norm": 3.978114366531372,
"learning_rate": 1.1114206128133705e-05,
"loss": 0.4194,
"step": 4190
},
{
"epoch": 2.33983286908078,
"grad_norm": 3.607412099838257,
"learning_rate": 1.1021355617455896e-05,
"loss": 0.4783,
"step": 4200
},
{
"epoch": 2.3454038997214486,
"grad_norm": 6.030869960784912,
"learning_rate": 1.0928505106778087e-05,
"loss": 0.421,
"step": 4210
},
{
"epoch": 2.350974930362117,
"grad_norm": 2.8264944553375244,
"learning_rate": 1.083565459610028e-05,
"loss": 0.4424,
"step": 4220
},
{
"epoch": 2.3565459610027855,
"grad_norm": 3.267326593399048,
"learning_rate": 1.074280408542247e-05,
"loss": 0.4451,
"step": 4230
},
{
"epoch": 2.362116991643454,
"grad_norm": 3.6425423622131348,
"learning_rate": 1.0649953574744661e-05,
"loss": 0.5231,
"step": 4240
},
{
"epoch": 2.3676880222841223,
"grad_norm": 3.3133020401000977,
"learning_rate": 1.0557103064066854e-05,
"loss": 0.5509,
"step": 4250
},
{
"epoch": 2.3732590529247912,
"grad_norm": 3.444788694381714,
"learning_rate": 1.0464252553389043e-05,
"loss": 0.2895,
"step": 4260
},
{
"epoch": 2.3788300835654597,
"grad_norm": 7.917379856109619,
"learning_rate": 1.0371402042711236e-05,
"loss": 0.5842,
"step": 4270
},
{
"epoch": 2.384401114206128,
"grad_norm": 4.689010143280029,
"learning_rate": 1.0278551532033427e-05,
"loss": 0.467,
"step": 4280
},
{
"epoch": 2.3899721448467965,
"grad_norm": 4.501764297485352,
"learning_rate": 1.0185701021355617e-05,
"loss": 0.4238,
"step": 4290
},
{
"epoch": 2.3955431754874654,
"grad_norm": 4.01366662979126,
"learning_rate": 1.009285051067781e-05,
"loss": 0.4781,
"step": 4300
},
{
"epoch": 2.401114206128134,
"grad_norm": 11.674251556396484,
"learning_rate": 1e-05,
"loss": 0.403,
"step": 4310
},
{
"epoch": 2.4066852367688023,
"grad_norm": 5.943558692932129,
"learning_rate": 9.907149489322192e-06,
"loss": 0.3621,
"step": 4320
},
{
"epoch": 2.4122562674094707,
"grad_norm": 1.9158827066421509,
"learning_rate": 9.814298978644383e-06,
"loss": 0.3801,
"step": 4330
},
{
"epoch": 2.417827298050139,
"grad_norm": 3.426053047180176,
"learning_rate": 9.721448467966575e-06,
"loss": 0.419,
"step": 4340
},
{
"epoch": 2.4233983286908076,
"grad_norm": 3.975104570388794,
"learning_rate": 9.628597957288766e-06,
"loss": 0.414,
"step": 4350
},
{
"epoch": 2.4289693593314765,
"grad_norm": 2.1982619762420654,
"learning_rate": 9.535747446610957e-06,
"loss": 0.4647,
"step": 4360
},
{
"epoch": 2.434540389972145,
"grad_norm": 4.273013591766357,
"learning_rate": 9.44289693593315e-06,
"loss": 0.5589,
"step": 4370
},
{
"epoch": 2.4401114206128134,
"grad_norm": 3.543038845062256,
"learning_rate": 9.350046425255339e-06,
"loss": 0.4927,
"step": 4380
},
{
"epoch": 2.445682451253482,
"grad_norm": 3.844630718231201,
"learning_rate": 9.257195914577531e-06,
"loss": 0.3665,
"step": 4390
},
{
"epoch": 2.4512534818941503,
"grad_norm": 5.011111736297607,
"learning_rate": 9.164345403899722e-06,
"loss": 0.4216,
"step": 4400
},
{
"epoch": 2.456824512534819,
"grad_norm": 5.382537841796875,
"learning_rate": 9.071494893221913e-06,
"loss": 0.4904,
"step": 4410
},
{
"epoch": 2.4623955431754876,
"grad_norm": 2.391098976135254,
"learning_rate": 8.978644382544105e-06,
"loss": 0.4759,
"step": 4420
},
{
"epoch": 2.467966573816156,
"grad_norm": 3.50219464302063,
"learning_rate": 8.885793871866294e-06,
"loss": 0.4777,
"step": 4430
},
{
"epoch": 2.4735376044568245,
"grad_norm": 4.964571475982666,
"learning_rate": 8.792943361188487e-06,
"loss": 0.3474,
"step": 4440
},
{
"epoch": 2.479108635097493,
"grad_norm": 5.0892510414123535,
"learning_rate": 8.700092850510678e-06,
"loss": 0.3845,
"step": 4450
},
{
"epoch": 2.4846796657381613,
"grad_norm": 6.2226362228393555,
"learning_rate": 8.607242339832869e-06,
"loss": 0.5149,
"step": 4460
},
{
"epoch": 2.4902506963788302,
"grad_norm": 5.676084995269775,
"learning_rate": 8.514391829155061e-06,
"loss": 0.3087,
"step": 4470
},
{
"epoch": 2.4958217270194987,
"grad_norm": 5.1531982421875,
"learning_rate": 8.421541318477252e-06,
"loss": 0.5312,
"step": 4480
},
{
"epoch": 2.501392757660167,
"grad_norm": 4.618714332580566,
"learning_rate": 8.328690807799443e-06,
"loss": 0.4738,
"step": 4490
},
{
"epoch": 2.5069637883008355,
"grad_norm": 3.6886181831359863,
"learning_rate": 8.235840297121634e-06,
"loss": 0.5054,
"step": 4500
},
{
"epoch": 2.5125348189415044,
"grad_norm": 6.665517330169678,
"learning_rate": 8.142989786443826e-06,
"loss": 0.4265,
"step": 4510
},
{
"epoch": 2.518105849582173,
"grad_norm": 4.075329303741455,
"learning_rate": 8.050139275766017e-06,
"loss": 0.414,
"step": 4520
},
{
"epoch": 2.5236768802228413,
"grad_norm": 4.333959579467773,
"learning_rate": 7.957288765088208e-06,
"loss": 0.4542,
"step": 4530
},
{
"epoch": 2.5292479108635098,
"grad_norm": 4.7477827072143555,
"learning_rate": 7.8644382544104e-06,
"loss": 0.4223,
"step": 4540
},
{
"epoch": 2.534818941504178,
"grad_norm": 4.310934066772461,
"learning_rate": 7.77158774373259e-06,
"loss": 0.342,
"step": 4550
},
{
"epoch": 2.5403899721448466,
"grad_norm": 3.5395619869232178,
"learning_rate": 7.678737233054782e-06,
"loss": 0.4755,
"step": 4560
},
{
"epoch": 2.545961002785515,
"grad_norm": 4.999756813049316,
"learning_rate": 7.585886722376974e-06,
"loss": 0.4619,
"step": 4570
},
{
"epoch": 2.551532033426184,
"grad_norm": 5.0525383949279785,
"learning_rate": 7.493036211699164e-06,
"loss": 0.4846,
"step": 4580
},
{
"epoch": 2.5571030640668524,
"grad_norm": 8.564767837524414,
"learning_rate": 7.400185701021356e-06,
"loss": 0.4261,
"step": 4590
},
{
"epoch": 2.562674094707521,
"grad_norm": 4.90235710144043,
"learning_rate": 7.3073351903435475e-06,
"loss": 0.435,
"step": 4600
},
{
"epoch": 2.5682451253481893,
"grad_norm": 5.1071672439575195,
"learning_rate": 7.214484679665738e-06,
"loss": 0.4378,
"step": 4610
},
{
"epoch": 2.573816155988858,
"grad_norm": 4.659866809844971,
"learning_rate": 7.12163416898793e-06,
"loss": 0.4755,
"step": 4620
},
{
"epoch": 2.5793871866295266,
"grad_norm": 6.740070343017578,
"learning_rate": 7.028783658310122e-06,
"loss": 0.5263,
"step": 4630
},
{
"epoch": 2.584958217270195,
"grad_norm": 7.245567321777344,
"learning_rate": 6.935933147632312e-06,
"loss": 0.4753,
"step": 4640
},
{
"epoch": 2.5905292479108635,
"grad_norm": 3.481055736541748,
"learning_rate": 6.8430826369545035e-06,
"loss": 0.334,
"step": 4650
},
{
"epoch": 2.596100278551532,
"grad_norm": 9.103506088256836,
"learning_rate": 6.750232126276695e-06,
"loss": 0.5612,
"step": 4660
},
{
"epoch": 2.6016713091922004,
"grad_norm": 2.4863336086273193,
"learning_rate": 6.657381615598886e-06,
"loss": 0.4572,
"step": 4670
},
{
"epoch": 2.6072423398328692,
"grad_norm": 3.1609058380126953,
"learning_rate": 6.564531104921078e-06,
"loss": 0.4082,
"step": 4680
},
{
"epoch": 2.6128133704735377,
"grad_norm": 4.247745037078857,
"learning_rate": 6.4716805942432695e-06,
"loss": 0.5278,
"step": 4690
},
{
"epoch": 2.618384401114206,
"grad_norm": 5.853703022003174,
"learning_rate": 6.3788300835654595e-06,
"loss": 0.4127,
"step": 4700
},
{
"epoch": 2.6239554317548746,
"grad_norm": 4.09820032119751,
"learning_rate": 6.285979572887651e-06,
"loss": 0.399,
"step": 4710
},
{
"epoch": 2.6295264623955434,
"grad_norm": 4.088954925537109,
"learning_rate": 6.193129062209842e-06,
"loss": 0.3394,
"step": 4720
},
{
"epoch": 2.635097493036212,
"grad_norm": 5.091104984283447,
"learning_rate": 6.100278551532034e-06,
"loss": 0.4218,
"step": 4730
},
{
"epoch": 2.6406685236768803,
"grad_norm": 4.649893760681152,
"learning_rate": 6.0074280408542255e-06,
"loss": 0.4801,
"step": 4740
},
{
"epoch": 2.6462395543175488,
"grad_norm": 6.470804691314697,
"learning_rate": 5.914577530176416e-06,
"loss": 0.5307,
"step": 4750
},
{
"epoch": 2.651810584958217,
"grad_norm": 3.143728017807007,
"learning_rate": 5.821727019498607e-06,
"loss": 0.3806,
"step": 4760
},
{
"epoch": 2.6573816155988856,
"grad_norm": 3.0394294261932373,
"learning_rate": 5.728876508820799e-06,
"loss": 0.4565,
"step": 4770
},
{
"epoch": 2.662952646239554,
"grad_norm": 4.775937557220459,
"learning_rate": 5.63602599814299e-06,
"loss": 0.3728,
"step": 4780
},
{
"epoch": 2.668523676880223,
"grad_norm": 3.5294461250305176,
"learning_rate": 5.5431754874651814e-06,
"loss": 0.4864,
"step": 4790
},
{
"epoch": 2.6740947075208914,
"grad_norm": 6.9017486572265625,
"learning_rate": 5.450324976787373e-06,
"loss": 0.4679,
"step": 4800
},
{
"epoch": 2.67966573816156,
"grad_norm": 2.8387668132781982,
"learning_rate": 5.357474466109564e-06,
"loss": 0.2643,
"step": 4810
},
{
"epoch": 2.6852367688022283,
"grad_norm": 5.770737171173096,
"learning_rate": 5.264623955431755e-06,
"loss": 0.4973,
"step": 4820
},
{
"epoch": 2.690807799442897,
"grad_norm": 2.5987281799316406,
"learning_rate": 5.171773444753946e-06,
"loss": 0.4445,
"step": 4830
},
{
"epoch": 2.6963788300835656,
"grad_norm": 5.727529048919678,
"learning_rate": 5.078922934076137e-06,
"loss": 0.4761,
"step": 4840
},
{
"epoch": 2.701949860724234,
"grad_norm": 3.2433037757873535,
"learning_rate": 4.986072423398329e-06,
"loss": 0.5198,
"step": 4850
},
{
"epoch": 2.7075208913649025,
"grad_norm": 2.264634847640991,
"learning_rate": 4.89322191272052e-06,
"loss": 0.3812,
"step": 4860
},
{
"epoch": 2.713091922005571,
"grad_norm": 3.4057681560516357,
"learning_rate": 4.800371402042712e-06,
"loss": 0.5236,
"step": 4870
},
{
"epoch": 2.7186629526462394,
"grad_norm": 5.159498691558838,
"learning_rate": 4.7075208913649025e-06,
"loss": 0.4401,
"step": 4880
},
{
"epoch": 2.724233983286908,
"grad_norm": 6.718446731567383,
"learning_rate": 4.614670380687093e-06,
"loss": 0.4825,
"step": 4890
},
{
"epoch": 2.7298050139275767,
"grad_norm": 4.028508186340332,
"learning_rate": 4.521819870009285e-06,
"loss": 0.3836,
"step": 4900
},
{
"epoch": 2.735376044568245,
"grad_norm": 2.027120351791382,
"learning_rate": 4.428969359331477e-06,
"loss": 0.3456,
"step": 4910
},
{
"epoch": 2.7409470752089136,
"grad_norm": 3.1550886631011963,
"learning_rate": 4.336118848653668e-06,
"loss": 0.3802,
"step": 4920
},
{
"epoch": 2.7465181058495824,
"grad_norm": 5.188984394073486,
"learning_rate": 4.243268337975859e-06,
"loss": 0.383,
"step": 4930
},
{
"epoch": 2.752089136490251,
"grad_norm": 6.997900009155273,
"learning_rate": 4.15041782729805e-06,
"loss": 0.4606,
"step": 4940
},
{
"epoch": 2.7576601671309193,
"grad_norm": 4.508044719696045,
"learning_rate": 4.057567316620241e-06,
"loss": 0.4017,
"step": 4950
},
{
"epoch": 2.7632311977715878,
"grad_norm": 6.631628036499023,
"learning_rate": 3.964716805942433e-06,
"loss": 0.398,
"step": 4960
},
{
"epoch": 2.768802228412256,
"grad_norm": 7.107889175415039,
"learning_rate": 3.8718662952646245e-06,
"loss": 0.4823,
"step": 4970
},
{
"epoch": 2.7743732590529246,
"grad_norm": 4.978562355041504,
"learning_rate": 3.7790157845868153e-06,
"loss": 0.3995,
"step": 4980
},
{
"epoch": 2.779944289693593,
"grad_norm": 5.335329055786133,
"learning_rate": 3.686165273909006e-06,
"loss": 0.39,
"step": 4990
},
{
"epoch": 2.785515320334262,
"grad_norm": 3.1053812503814697,
"learning_rate": 3.593314763231198e-06,
"loss": 0.3362,
"step": 5000
},
{
"epoch": 2.7910863509749304,
"grad_norm": 4.562804222106934,
"learning_rate": 3.500464252553389e-06,
"loss": 0.3663,
"step": 5010
},
{
"epoch": 2.796657381615599,
"grad_norm": 2.6694986820220947,
"learning_rate": 3.40761374187558e-06,
"loss": 0.3584,
"step": 5020
},
{
"epoch": 2.8022284122562673,
"grad_norm": 5.513758659362793,
"learning_rate": 3.3147632311977717e-06,
"loss": 0.4915,
"step": 5030
},
{
"epoch": 2.807799442896936,
"grad_norm": 4.651453971862793,
"learning_rate": 3.221912720519963e-06,
"loss": 0.364,
"step": 5040
},
{
"epoch": 2.8133704735376046,
"grad_norm": 4.250921249389648,
"learning_rate": 3.129062209842154e-06,
"loss": 0.3958,
"step": 5050
},
{
"epoch": 2.818941504178273,
"grad_norm": 3.3700807094573975,
"learning_rate": 3.0362116991643456e-06,
"loss": 0.2812,
"step": 5060
},
{
"epoch": 2.8245125348189415,
"grad_norm": 5.31024694442749,
"learning_rate": 2.943361188486537e-06,
"loss": 0.4218,
"step": 5070
},
{
"epoch": 2.83008356545961,
"grad_norm": 2.610002279281616,
"learning_rate": 2.850510677808728e-06,
"loss": 0.3653,
"step": 5080
},
{
"epoch": 2.8356545961002784,
"grad_norm": 5.629008769989014,
"learning_rate": 2.7576601671309194e-06,
"loss": 0.418,
"step": 5090
},
{
"epoch": 2.841225626740947,
"grad_norm": 3.8689396381378174,
"learning_rate": 2.6648096564531107e-06,
"loss": 0.3105,
"step": 5100
},
{
"epoch": 2.8467966573816157,
"grad_norm": 3.4457309246063232,
"learning_rate": 2.5719591457753016e-06,
"loss": 0.4024,
"step": 5110
},
{
"epoch": 2.852367688022284,
"grad_norm": 4.193653583526611,
"learning_rate": 2.4791086350974933e-06,
"loss": 0.4946,
"step": 5120
},
{
"epoch": 2.8579387186629526,
"grad_norm": 8.699234962463379,
"learning_rate": 2.3862581244196846e-06,
"loss": 0.4511,
"step": 5130
},
{
"epoch": 2.863509749303621,
"grad_norm": 4.196930408477783,
"learning_rate": 2.2934076137418754e-06,
"loss": 0.4671,
"step": 5140
},
{
"epoch": 2.86908077994429,
"grad_norm": 5.245868682861328,
"learning_rate": 2.200557103064067e-06,
"loss": 0.4111,
"step": 5150
},
{
"epoch": 2.8746518105849583,
"grad_norm": 4.189416885375977,
"learning_rate": 2.1077065923862584e-06,
"loss": 0.4043,
"step": 5160
},
{
"epoch": 2.8802228412256268,
"grad_norm": 2.8182320594787598,
"learning_rate": 2.0148560817084493e-06,
"loss": 0.3343,
"step": 5170
},
{
"epoch": 2.885793871866295,
"grad_norm": 7.292479038238525,
"learning_rate": 1.922005571030641e-06,
"loss": 0.3844,
"step": 5180
},
{
"epoch": 2.8913649025069637,
"grad_norm": 2.767341375350952,
"learning_rate": 1.8291550603528318e-06,
"loss": 0.3948,
"step": 5190
},
{
"epoch": 2.896935933147632,
"grad_norm": 5.170936107635498,
"learning_rate": 1.7363045496750233e-06,
"loss": 0.4235,
"step": 5200
},
{
"epoch": 2.902506963788301,
"grad_norm": 7.14333438873291,
"learning_rate": 1.6434540389972146e-06,
"loss": 0.3677,
"step": 5210
},
{
"epoch": 2.9080779944289694,
"grad_norm": 6.775882244110107,
"learning_rate": 1.5506035283194059e-06,
"loss": 0.486,
"step": 5220
},
{
"epoch": 2.913649025069638,
"grad_norm": 4.250217914581299,
"learning_rate": 1.4577530176415971e-06,
"loss": 0.5274,
"step": 5230
},
{
"epoch": 2.9192200557103063,
"grad_norm": 4.6204633712768555,
"learning_rate": 1.3649025069637884e-06,
"loss": 0.5229,
"step": 5240
},
{
"epoch": 2.924791086350975,
"grad_norm": 3.8870279788970947,
"learning_rate": 1.2720519962859795e-06,
"loss": 0.535,
"step": 5250
},
{
"epoch": 2.9303621169916436,
"grad_norm": 9.080047607421875,
"learning_rate": 1.179201485608171e-06,
"loss": 0.507,
"step": 5260
},
{
"epoch": 2.935933147632312,
"grad_norm": 10.549107551574707,
"learning_rate": 1.0863509749303623e-06,
"loss": 0.4738,
"step": 5270
},
{
"epoch": 2.9415041782729805,
"grad_norm": 3.7926011085510254,
"learning_rate": 9.935004642525533e-07,
"loss": 0.4498,
"step": 5280
},
{
"epoch": 2.947075208913649,
"grad_norm": 2.45975399017334,
"learning_rate": 9.006499535747446e-07,
"loss": 0.3481,
"step": 5290
},
{
"epoch": 2.9526462395543174,
"grad_norm": 4.364011764526367,
"learning_rate": 8.07799442896936e-07,
"loss": 0.4118,
"step": 5300
},
{
"epoch": 2.958217270194986,
"grad_norm": 5.1891255378723145,
"learning_rate": 7.149489322191273e-07,
"loss": 0.5408,
"step": 5310
},
{
"epoch": 2.9637883008356547,
"grad_norm": 5.055240631103516,
"learning_rate": 6.220984215413185e-07,
"loss": 0.3917,
"step": 5320
},
{
"epoch": 2.969359331476323,
"grad_norm": 4.434938907623291,
"learning_rate": 5.292479108635097e-07,
"loss": 0.3526,
"step": 5330
},
{
"epoch": 2.9749303621169916,
"grad_norm": 5.2575860023498535,
"learning_rate": 4.36397400185701e-07,
"loss": 0.3464,
"step": 5340
},
{
"epoch": 2.98050139275766,
"grad_norm": 8.497797966003418,
"learning_rate": 3.435468895078923e-07,
"loss": 0.5106,
"step": 5350
},
{
"epoch": 2.986072423398329,
"grad_norm": 4.38588285446167,
"learning_rate": 2.506963788300836e-07,
"loss": 0.3509,
"step": 5360
},
{
"epoch": 2.9916434540389973,
"grad_norm": 7.261599540710449,
"learning_rate": 1.5784586815227484e-07,
"loss": 0.5137,
"step": 5370
},
{
"epoch": 2.997214484679666,
"grad_norm": 7.3228936195373535,
"learning_rate": 6.499535747446612e-08,
"loss": 0.4838,
"step": 5380
},
{
"epoch": 3.0,
"eval_accuracy": 0.7127333519086096,
"eval_loss": 0.8501355051994324,
"eval_runtime": 37.2414,
"eval_samples_per_second": 96.371,
"eval_steps_per_second": 6.042,
"step": 5385
},
{
"epoch": 3.0,
"step": 5385,
"total_flos": 6.674451681129007e+18,
"train_loss": 0.7317446124697573,
"train_runtime": 2231.3604,
"train_samples_per_second": 38.598,
"train_steps_per_second": 2.413
}
],
"logging_steps": 10,
"max_steps": 5385,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6.674451681129007e+18,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}