GR00T-N1.5-3B-LIBERO-OBJECT-8K / trainer_state.json
Tacoin's picture
Upload libero_object_8k checkpoint
78eabf3 verified
Raw
History Blame Contribute Delete
140 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.910675901600191,
"eval_steps": 500,
"global_step": 8000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0023883448770002386,
"grad_norm": 6.263468265533447,
"learning_rate": 2.25e-06,
"loss": 0.9113,
"step": 10
},
{
"epoch": 0.004776689754000477,
"grad_norm": 3.527216672897339,
"learning_rate": 4.75e-06,
"loss": 0.7654,
"step": 20
},
{
"epoch": 0.007165034631000717,
"grad_norm": 2.6621556282043457,
"learning_rate": 7.25e-06,
"loss": 0.4555,
"step": 30
},
{
"epoch": 0.009553379508000955,
"grad_norm": 1.9681612253189087,
"learning_rate": 9.750000000000002e-06,
"loss": 0.3406,
"step": 40
},
{
"epoch": 0.011941724385001195,
"grad_norm": 1.6213421821594238,
"learning_rate": 1.225e-05,
"loss": 0.2842,
"step": 50
},
{
"epoch": 0.014330069262001434,
"grad_norm": 1.9234356880187988,
"learning_rate": 1.475e-05,
"loss": 0.2738,
"step": 60
},
{
"epoch": 0.016718414139001672,
"grad_norm": 1.6524076461791992,
"learning_rate": 1.725e-05,
"loss": 0.2066,
"step": 70
},
{
"epoch": 0.01910675901600191,
"grad_norm": 3.277963399887085,
"learning_rate": 1.9750000000000002e-05,
"loss": 0.2439,
"step": 80
},
{
"epoch": 0.02149510389300215,
"grad_norm": 1.1675255298614502,
"learning_rate": 2.2250000000000002e-05,
"loss": 0.2131,
"step": 90
},
{
"epoch": 0.02388344877000239,
"grad_norm": 1.6110434532165527,
"learning_rate": 2.4750000000000002e-05,
"loss": 0.2038,
"step": 100
},
{
"epoch": 0.026271793647002627,
"grad_norm": 1.467454195022583,
"learning_rate": 2.725e-05,
"loss": 0.1974,
"step": 110
},
{
"epoch": 0.028660138524002867,
"grad_norm": 1.2697089910507202,
"learning_rate": 2.975e-05,
"loss": 0.1825,
"step": 120
},
{
"epoch": 0.031048483401003104,
"grad_norm": 1.1076679229736328,
"learning_rate": 3.2250000000000005e-05,
"loss": 0.1377,
"step": 130
},
{
"epoch": 0.033436828278003344,
"grad_norm": 0.9517481327056885,
"learning_rate": 3.475e-05,
"loss": 0.1494,
"step": 140
},
{
"epoch": 0.03582517315500358,
"grad_norm": 0.8394818305969238,
"learning_rate": 3.7250000000000004e-05,
"loss": 0.1478,
"step": 150
},
{
"epoch": 0.03821351803200382,
"grad_norm": 1.0298035144805908,
"learning_rate": 3.9750000000000004e-05,
"loss": 0.1485,
"step": 160
},
{
"epoch": 0.04060186290900406,
"grad_norm": 0.8090702891349792,
"learning_rate": 4.2250000000000004e-05,
"loss": 0.1623,
"step": 170
},
{
"epoch": 0.0429902077860043,
"grad_norm": 0.9527009725570679,
"learning_rate": 4.4750000000000004e-05,
"loss": 0.1336,
"step": 180
},
{
"epoch": 0.045378552663004536,
"grad_norm": 0.9270368218421936,
"learning_rate": 4.7249999999999997e-05,
"loss": 0.1446,
"step": 190
},
{
"epoch": 0.04776689754000478,
"grad_norm": 1.063560962677002,
"learning_rate": 4.975e-05,
"loss": 0.1453,
"step": 200
},
{
"epoch": 0.050155242417005017,
"grad_norm": 0.9871803522109985,
"learning_rate": 5.2249999999999996e-05,
"loss": 0.1458,
"step": 210
},
{
"epoch": 0.05254358729400525,
"grad_norm": 1.0081197023391724,
"learning_rate": 5.475e-05,
"loss": 0.1321,
"step": 220
},
{
"epoch": 0.05493193217100549,
"grad_norm": 1.0061386823654175,
"learning_rate": 5.725e-05,
"loss": 0.1391,
"step": 230
},
{
"epoch": 0.057320277048005734,
"grad_norm": 1.5330615043640137,
"learning_rate": 5.975000000000001e-05,
"loss": 0.1503,
"step": 240
},
{
"epoch": 0.05970862192500597,
"grad_norm": 1.42104172706604,
"learning_rate": 6.225000000000001e-05,
"loss": 0.1606,
"step": 250
},
{
"epoch": 0.06209696680200621,
"grad_norm": 0.785142183303833,
"learning_rate": 6.475e-05,
"loss": 0.1467,
"step": 260
},
{
"epoch": 0.06448531167900645,
"grad_norm": 1.0392279624938965,
"learning_rate": 6.725000000000001e-05,
"loss": 0.134,
"step": 270
},
{
"epoch": 0.06687365655600669,
"grad_norm": 1.149312138557434,
"learning_rate": 6.975e-05,
"loss": 0.145,
"step": 280
},
{
"epoch": 0.06926200143300693,
"grad_norm": 1.2992204427719116,
"learning_rate": 7.225000000000001e-05,
"loss": 0.1364,
"step": 290
},
{
"epoch": 0.07165034631000716,
"grad_norm": 0.9730273485183716,
"learning_rate": 7.475000000000001e-05,
"loss": 0.1544,
"step": 300
},
{
"epoch": 0.0740386911870074,
"grad_norm": 0.9651837944984436,
"learning_rate": 7.725e-05,
"loss": 0.138,
"step": 310
},
{
"epoch": 0.07642703606400764,
"grad_norm": 1.523910403251648,
"learning_rate": 7.975e-05,
"loss": 0.1338,
"step": 320
},
{
"epoch": 0.07881538094100789,
"grad_norm": 1.4649964570999146,
"learning_rate": 8.225000000000001e-05,
"loss": 0.1397,
"step": 330
},
{
"epoch": 0.08120372581800812,
"grad_norm": 0.976882815361023,
"learning_rate": 8.475000000000001e-05,
"loss": 0.1195,
"step": 340
},
{
"epoch": 0.08359207069500836,
"grad_norm": 1.1676348447799683,
"learning_rate": 8.725e-05,
"loss": 0.138,
"step": 350
},
{
"epoch": 0.0859804155720086,
"grad_norm": 0.9947729706764221,
"learning_rate": 8.975e-05,
"loss": 0.1365,
"step": 360
},
{
"epoch": 0.08836876044900883,
"grad_norm": 0.96565181016922,
"learning_rate": 9.225e-05,
"loss": 0.1202,
"step": 370
},
{
"epoch": 0.09075710532600907,
"grad_norm": 0.5443023443222046,
"learning_rate": 9.475e-05,
"loss": 0.1222,
"step": 380
},
{
"epoch": 0.09314545020300931,
"grad_norm": 0.8974901437759399,
"learning_rate": 9.725e-05,
"loss": 0.1438,
"step": 390
},
{
"epoch": 0.09553379508000956,
"grad_norm": 1.0482345819473267,
"learning_rate": 9.975000000000001e-05,
"loss": 0.1143,
"step": 400
},
{
"epoch": 0.0979221399570098,
"grad_norm": 1.1348724365234375,
"learning_rate": 9.999965398327804e-05,
"loss": 0.1256,
"step": 410
},
{
"epoch": 0.10031048483401003,
"grad_norm": 1.030310869216919,
"learning_rate": 9.999845788223949e-05,
"loss": 0.1372,
"step": 420
},
{
"epoch": 0.10269882971101027,
"grad_norm": 0.8888779878616333,
"learning_rate": 9.999640744550616e-05,
"loss": 0.1327,
"step": 430
},
{
"epoch": 0.1050871745880105,
"grad_norm": 1.2149293422698975,
"learning_rate": 9.999350270811438e-05,
"loss": 0.1455,
"step": 440
},
{
"epoch": 0.10747551946501074,
"grad_norm": 1.0437705516815186,
"learning_rate": 9.99897437196981e-05,
"loss": 0.1533,
"step": 450
},
{
"epoch": 0.10986386434201098,
"grad_norm": 1.137521505355835,
"learning_rate": 9.998513054448802e-05,
"loss": 0.1277,
"step": 460
},
{
"epoch": 0.11225220921901123,
"grad_norm": 1.001363754272461,
"learning_rate": 9.99796632613106e-05,
"loss": 0.1378,
"step": 470
},
{
"epoch": 0.11464055409601147,
"grad_norm": 0.9264760613441467,
"learning_rate": 9.997334196358664e-05,
"loss": 0.1089,
"step": 480
},
{
"epoch": 0.1170288989730117,
"grad_norm": 0.867767870426178,
"learning_rate": 9.996616675932966e-05,
"loss": 0.1372,
"step": 490
},
{
"epoch": 0.11941724385001194,
"grad_norm": 1.0822848081588745,
"learning_rate": 9.995813777114411e-05,
"loss": 0.1092,
"step": 500
},
{
"epoch": 0.12180558872701218,
"grad_norm": 0.8188408613204956,
"learning_rate": 9.994925513622324e-05,
"loss": 0.1351,
"step": 510
},
{
"epoch": 0.12419393360401242,
"grad_norm": 0.6109073758125305,
"learning_rate": 9.993951900634678e-05,
"loss": 0.1235,
"step": 520
},
{
"epoch": 0.12658227848101267,
"grad_norm": 0.5777956247329712,
"learning_rate": 9.992892954787831e-05,
"loss": 0.129,
"step": 530
},
{
"epoch": 0.1289706233580129,
"grad_norm": 0.8576014041900635,
"learning_rate": 9.991748694176246e-05,
"loss": 0.1373,
"step": 540
},
{
"epoch": 0.13135896823501314,
"grad_norm": 0.586820125579834,
"learning_rate": 9.990519138352184e-05,
"loss": 0.1126,
"step": 550
},
{
"epoch": 0.13374731311201338,
"grad_norm": 0.5999759435653687,
"learning_rate": 9.989204308325356e-05,
"loss": 0.1256,
"step": 560
},
{
"epoch": 0.13613565798901361,
"grad_norm": 0.9213500022888184,
"learning_rate": 9.987804226562582e-05,
"loss": 0.1419,
"step": 570
},
{
"epoch": 0.13852400286601385,
"grad_norm": 0.6835944652557373,
"learning_rate": 9.986318916987395e-05,
"loss": 0.1325,
"step": 580
},
{
"epoch": 0.1409123477430141,
"grad_norm": 0.9460831880569458,
"learning_rate": 9.984748404979642e-05,
"loss": 0.1296,
"step": 590
},
{
"epoch": 0.14330069262001432,
"grad_norm": 1.101009488105774,
"learning_rate": 9.983092717375033e-05,
"loss": 0.1255,
"step": 600
},
{
"epoch": 0.14568903749701456,
"grad_norm": 0.7150104641914368,
"learning_rate": 9.981351882464706e-05,
"loss": 0.1466,
"step": 610
},
{
"epoch": 0.1480773823740148,
"grad_norm": 1.1312141418457031,
"learning_rate": 9.979525929994727e-05,
"loss": 0.1245,
"step": 620
},
{
"epoch": 0.15046572725101504,
"grad_norm": 0.8617872595787048,
"learning_rate": 9.977614891165581e-05,
"loss": 0.1159,
"step": 630
},
{
"epoch": 0.15285407212801527,
"grad_norm": 0.8925164937973022,
"learning_rate": 9.975618798631653e-05,
"loss": 0.108,
"step": 640
},
{
"epoch": 0.15524241700501554,
"grad_norm": 0.5199300646781921,
"learning_rate": 9.973537686500656e-05,
"loss": 0.115,
"step": 650
},
{
"epoch": 0.15763076188201577,
"grad_norm": 0.7184763550758362,
"learning_rate": 9.971371590333053e-05,
"loss": 0.1377,
"step": 660
},
{
"epoch": 0.160019106759016,
"grad_norm": 0.5908390283584595,
"learning_rate": 9.969120547141453e-05,
"loss": 0.1047,
"step": 670
},
{
"epoch": 0.16240745163601625,
"grad_norm": 0.8077511787414551,
"learning_rate": 9.966784595389971e-05,
"loss": 0.1169,
"step": 680
},
{
"epoch": 0.16479579651301648,
"grad_norm": 0.6751367449760437,
"learning_rate": 9.96436377499358e-05,
"loss": 0.1079,
"step": 690
},
{
"epoch": 0.16718414139001672,
"grad_norm": 0.9484745860099792,
"learning_rate": 9.961858127317419e-05,
"loss": 0.1314,
"step": 700
},
{
"epoch": 0.16957248626701696,
"grad_norm": 0.8461595773696899,
"learning_rate": 9.959267695176096e-05,
"loss": 0.1232,
"step": 710
},
{
"epoch": 0.1719608311440172,
"grad_norm": 0.6955356597900391,
"learning_rate": 9.956592522832946e-05,
"loss": 0.1065,
"step": 720
},
{
"epoch": 0.17434917602101743,
"grad_norm": 0.5954784154891968,
"learning_rate": 9.953832655999284e-05,
"loss": 0.1255,
"step": 730
},
{
"epoch": 0.17673752089801767,
"grad_norm": 1.1098157167434692,
"learning_rate": 9.950988141833621e-05,
"loss": 0.1023,
"step": 740
},
{
"epoch": 0.1791258657750179,
"grad_norm": 1.1819376945495605,
"learning_rate": 9.948059028940857e-05,
"loss": 0.1191,
"step": 750
},
{
"epoch": 0.18151421065201814,
"grad_norm": 0.5757019519805908,
"learning_rate": 9.945045367371446e-05,
"loss": 0.1192,
"step": 760
},
{
"epoch": 0.18390255552901838,
"grad_norm": 0.8608731627464294,
"learning_rate": 9.941947208620551e-05,
"loss": 0.1136,
"step": 770
},
{
"epoch": 0.18629090040601862,
"grad_norm": 0.8185464143753052,
"learning_rate": 9.938764605627158e-05,
"loss": 0.1238,
"step": 780
},
{
"epoch": 0.18867924528301888,
"grad_norm": 0.6100537776947021,
"learning_rate": 9.935497612773167e-05,
"loss": 0.1109,
"step": 790
},
{
"epoch": 0.19106759016001912,
"grad_norm": 0.5345736742019653,
"learning_rate": 9.932146285882477e-05,
"loss": 0.124,
"step": 800
},
{
"epoch": 0.19345593503701936,
"grad_norm": 0.7329362034797668,
"learning_rate": 9.928710682220017e-05,
"loss": 0.1076,
"step": 810
},
{
"epoch": 0.1958442799140196,
"grad_norm": 0.7284510135650635,
"learning_rate": 9.925190860490772e-05,
"loss": 0.136,
"step": 820
},
{
"epoch": 0.19823262479101983,
"grad_norm": 0.7314885258674622,
"learning_rate": 9.921586880838784e-05,
"loss": 0.1078,
"step": 830
},
{
"epoch": 0.20062096966802007,
"grad_norm": 0.7253186106681824,
"learning_rate": 9.917898804846124e-05,
"loss": 0.106,
"step": 840
},
{
"epoch": 0.2030093145450203,
"grad_norm": 0.6058793067932129,
"learning_rate": 9.914126695531833e-05,
"loss": 0.1085,
"step": 850
},
{
"epoch": 0.20539765942202054,
"grad_norm": 0.773330569267273,
"learning_rate": 9.910270617350853e-05,
"loss": 0.1156,
"step": 860
},
{
"epoch": 0.20778600429902078,
"grad_norm": 0.4878668189048767,
"learning_rate": 9.90633063619292e-05,
"loss": 0.0994,
"step": 870
},
{
"epoch": 0.210174349176021,
"grad_norm": 0.5970842838287354,
"learning_rate": 9.90230681938144e-05,
"loss": 0.1124,
"step": 880
},
{
"epoch": 0.21256269405302125,
"grad_norm": 0.740959107875824,
"learning_rate": 9.898199235672341e-05,
"loss": 0.1052,
"step": 890
},
{
"epoch": 0.2149510389300215,
"grad_norm": 0.8148726224899292,
"learning_rate": 9.894007955252898e-05,
"loss": 0.1027,
"step": 900
},
{
"epoch": 0.21733938380702172,
"grad_norm": 0.9836689829826355,
"learning_rate": 9.889733049740527e-05,
"loss": 0.1096,
"step": 910
},
{
"epoch": 0.21972772868402196,
"grad_norm": 0.6567099690437317,
"learning_rate": 9.885374592181567e-05,
"loss": 0.1142,
"step": 920
},
{
"epoch": 0.2221160735610222,
"grad_norm": 0.7335542440414429,
"learning_rate": 9.880932657050033e-05,
"loss": 0.1278,
"step": 930
},
{
"epoch": 0.22450441843802246,
"grad_norm": 0.746810793876648,
"learning_rate": 9.876407320246342e-05,
"loss": 0.0989,
"step": 940
},
{
"epoch": 0.2268927633150227,
"grad_norm": 0.6894835829734802,
"learning_rate": 9.871798659096014e-05,
"loss": 0.1054,
"step": 950
},
{
"epoch": 0.22928110819202294,
"grad_norm": 0.651219367980957,
"learning_rate": 9.867106752348351e-05,
"loss": 0.1185,
"step": 960
},
{
"epoch": 0.23166945306902317,
"grad_norm": 0.6215059757232666,
"learning_rate": 9.862331680175096e-05,
"loss": 0.1131,
"step": 970
},
{
"epoch": 0.2340577979460234,
"grad_norm": 0.5786034464836121,
"learning_rate": 9.857473524169056e-05,
"loss": 0.1068,
"step": 980
},
{
"epoch": 0.23644614282302365,
"grad_norm": 0.8351436853408813,
"learning_rate": 9.852532367342713e-05,
"loss": 0.1117,
"step": 990
},
{
"epoch": 0.23883448770002388,
"grad_norm": 0.6387096643447876,
"learning_rate": 9.847508294126804e-05,
"loss": 0.1137,
"step": 1000
},
{
"epoch": 0.24122283257702412,
"grad_norm": 0.8003013730049133,
"learning_rate": 9.842401390368879e-05,
"loss": 0.1207,
"step": 1010
},
{
"epoch": 0.24361117745402436,
"grad_norm": 0.7528656125068665,
"learning_rate": 9.837211743331833e-05,
"loss": 0.1106,
"step": 1020
},
{
"epoch": 0.2459995223310246,
"grad_norm": 0.5455472469329834,
"learning_rate": 9.831939441692415e-05,
"loss": 0.109,
"step": 1030
},
{
"epoch": 0.24838786720802483,
"grad_norm": 0.5621523857116699,
"learning_rate": 9.826584575539712e-05,
"loss": 0.12,
"step": 1040
},
{
"epoch": 0.2507762120850251,
"grad_norm": 0.8602862358093262,
"learning_rate": 9.821147236373614e-05,
"loss": 0.1275,
"step": 1050
},
{
"epoch": 0.25316455696202533,
"grad_norm": 0.7243831753730774,
"learning_rate": 9.815627517103243e-05,
"loss": 0.1111,
"step": 1060
},
{
"epoch": 0.25555290183902557,
"grad_norm": 0.6485002636909485,
"learning_rate": 9.810025512045373e-05,
"loss": 0.119,
"step": 1070
},
{
"epoch": 0.2579412467160258,
"grad_norm": 0.6710385680198669,
"learning_rate": 9.804341316922814e-05,
"loss": 0.1328,
"step": 1080
},
{
"epoch": 0.26032959159302604,
"grad_norm": 0.5449047684669495,
"learning_rate": 9.798575028862778e-05,
"loss": 0.1268,
"step": 1090
},
{
"epoch": 0.2627179364700263,
"grad_norm": 0.9255445003509521,
"learning_rate": 9.792726746395218e-05,
"loss": 0.1171,
"step": 1100
},
{
"epoch": 0.2651062813470265,
"grad_norm": 0.777152955532074,
"learning_rate": 9.786796569451145e-05,
"loss": 0.1029,
"step": 1110
},
{
"epoch": 0.26749462622402675,
"grad_norm": 0.9189852476119995,
"learning_rate": 9.78078459936092e-05,
"loss": 0.1095,
"step": 1120
},
{
"epoch": 0.269882971101027,
"grad_norm": 0.5503485798835754,
"learning_rate": 9.774690938852525e-05,
"loss": 0.1046,
"step": 1130
},
{
"epoch": 0.27227131597802723,
"grad_norm": 0.8353752493858337,
"learning_rate": 9.768515692049806e-05,
"loss": 0.102,
"step": 1140
},
{
"epoch": 0.27465966085502747,
"grad_norm": 0.6734784245491028,
"learning_rate": 9.762258964470694e-05,
"loss": 0.1184,
"step": 1150
},
{
"epoch": 0.2770480057320277,
"grad_norm": 0.5363646745681763,
"learning_rate": 9.755920863025397e-05,
"loss": 0.1002,
"step": 1160
},
{
"epoch": 0.27943635060902794,
"grad_norm": 0.5017086863517761,
"learning_rate": 9.749501496014583e-05,
"loss": 0.111,
"step": 1170
},
{
"epoch": 0.2818246954860282,
"grad_norm": 0.6441454291343689,
"learning_rate": 9.743000973127523e-05,
"loss": 0.1078,
"step": 1180
},
{
"epoch": 0.2842130403630284,
"grad_norm": 0.7968981266021729,
"learning_rate": 9.736419405440218e-05,
"loss": 0.114,
"step": 1190
},
{
"epoch": 0.28660138524002865,
"grad_norm": 0.6817290186882019,
"learning_rate": 9.729756905413499e-05,
"loss": 0.1152,
"step": 1200
},
{
"epoch": 0.2889897301170289,
"grad_norm": 0.5880157351493835,
"learning_rate": 9.723013586891108e-05,
"loss": 0.1256,
"step": 1210
},
{
"epoch": 0.2913780749940291,
"grad_norm": 0.5446798801422119,
"learning_rate": 9.716189565097758e-05,
"loss": 0.1063,
"step": 1220
},
{
"epoch": 0.29376641987102936,
"grad_norm": 0.4995010495185852,
"learning_rate": 9.709284956637147e-05,
"loss": 0.092,
"step": 1230
},
{
"epoch": 0.2961547647480296,
"grad_norm": 0.4224517047405243,
"learning_rate": 9.702299879489989e-05,
"loss": 0.1058,
"step": 1240
},
{
"epoch": 0.29854310962502983,
"grad_norm": 0.6987228989601135,
"learning_rate": 9.695234453011977e-05,
"loss": 0.0951,
"step": 1250
},
{
"epoch": 0.30093145450203007,
"grad_norm": 0.5517745614051819,
"learning_rate": 9.688088797931754e-05,
"loss": 0.1003,
"step": 1260
},
{
"epoch": 0.3033197993790303,
"grad_norm": 0.6665986180305481,
"learning_rate": 9.680863036348851e-05,
"loss": 0.0997,
"step": 1270
},
{
"epoch": 0.30570814425603055,
"grad_norm": 0.4847840964794159,
"learning_rate": 9.673557291731598e-05,
"loss": 0.0965,
"step": 1280
},
{
"epoch": 0.3080964891330308,
"grad_norm": 0.5157939195632935,
"learning_rate": 9.666171688915014e-05,
"loss": 0.0995,
"step": 1290
},
{
"epoch": 0.3104848340100311,
"grad_norm": 0.6290860176086426,
"learning_rate": 9.65870635409867e-05,
"loss": 0.1102,
"step": 1300
},
{
"epoch": 0.3128731788870313,
"grad_norm": 0.4248766005039215,
"learning_rate": 9.651161414844544e-05,
"loss": 0.094,
"step": 1310
},
{
"epoch": 0.31526152376403155,
"grad_norm": 0.48888951539993286,
"learning_rate": 9.64353700007483e-05,
"loss": 0.1203,
"step": 1320
},
{
"epoch": 0.3176498686410318,
"grad_norm": 0.7862128019332886,
"learning_rate": 9.635833240069735e-05,
"loss": 0.1135,
"step": 1330
},
{
"epoch": 0.320038213518032,
"grad_norm": 0.8491302728652954,
"learning_rate": 9.62805026646527e-05,
"loss": 0.1176,
"step": 1340
},
{
"epoch": 0.32242655839503226,
"grad_norm": 0.6367400288581848,
"learning_rate": 9.620188212250976e-05,
"loss": 0.1026,
"step": 1350
},
{
"epoch": 0.3248149032720325,
"grad_norm": 0.6905094385147095,
"learning_rate": 9.612247211767667e-05,
"loss": 0.1126,
"step": 1360
},
{
"epoch": 0.32720324814903273,
"grad_norm": 0.8616007566452026,
"learning_rate": 9.604227400705133e-05,
"loss": 0.1083,
"step": 1370
},
{
"epoch": 0.32959159302603297,
"grad_norm": 0.7048349380493164,
"learning_rate": 9.596128916099814e-05,
"loss": 0.1159,
"step": 1380
},
{
"epoch": 0.3319799379030332,
"grad_norm": 0.47000810503959656,
"learning_rate": 9.587951896332471e-05,
"loss": 0.1151,
"step": 1390
},
{
"epoch": 0.33436828278003344,
"grad_norm": 0.48382118344306946,
"learning_rate": 9.579696481125811e-05,
"loss": 0.1197,
"step": 1400
},
{
"epoch": 0.3367566276570337,
"grad_norm": 0.4684751629829407,
"learning_rate": 9.5713628115421e-05,
"loss": 0.1319,
"step": 1410
},
{
"epoch": 0.3391449725340339,
"grad_norm": 0.4609277844429016,
"learning_rate": 9.562951029980758e-05,
"loss": 0.1151,
"step": 1420
},
{
"epoch": 0.34153331741103415,
"grad_norm": 0.44542503356933594,
"learning_rate": 9.554461280175924e-05,
"loss": 0.1217,
"step": 1430
},
{
"epoch": 0.3439216622880344,
"grad_norm": 0.6385927200317383,
"learning_rate": 9.545893707193997e-05,
"loss": 0.1032,
"step": 1440
},
{
"epoch": 0.34631000716503463,
"grad_norm": 0.3624906837940216,
"learning_rate": 9.537248457431165e-05,
"loss": 0.1131,
"step": 1450
},
{
"epoch": 0.34869835204203486,
"grad_norm": 0.6893216967582703,
"learning_rate": 9.528525678610886e-05,
"loss": 0.0999,
"step": 1460
},
{
"epoch": 0.3510866969190351,
"grad_norm": 0.7240498661994934,
"learning_rate": 9.519725519781391e-05,
"loss": 0.079,
"step": 1470
},
{
"epoch": 0.35347504179603534,
"grad_norm": 0.5612568259239197,
"learning_rate": 9.510848131313106e-05,
"loss": 0.0882,
"step": 1480
},
{
"epoch": 0.3558633866730356,
"grad_norm": 0.4929567873477936,
"learning_rate": 9.501893664896113e-05,
"loss": 0.0808,
"step": 1490
},
{
"epoch": 0.3582517315500358,
"grad_norm": 1.1698434352874756,
"learning_rate": 9.492862273537533e-05,
"loss": 0.1036,
"step": 1500
},
{
"epoch": 0.36064007642703605,
"grad_norm": 0.3677363395690918,
"learning_rate": 9.483754111558927e-05,
"loss": 0.0801,
"step": 1510
},
{
"epoch": 0.3630284213040363,
"grad_norm": 0.5670194029808044,
"learning_rate": 9.474569334593652e-05,
"loss": 0.0965,
"step": 1520
},
{
"epoch": 0.3654167661810365,
"grad_norm": 0.4842435121536255,
"learning_rate": 9.465308099584208e-05,
"loss": 0.1005,
"step": 1530
},
{
"epoch": 0.36780511105803676,
"grad_norm": 0.41923508048057556,
"learning_rate": 9.455970564779543e-05,
"loss": 0.1081,
"step": 1540
},
{
"epoch": 0.370193455935037,
"grad_norm": 0.6009585857391357,
"learning_rate": 9.446556889732368e-05,
"loss": 0.1026,
"step": 1550
},
{
"epoch": 0.37258180081203723,
"grad_norm": 0.8315684795379639,
"learning_rate": 9.437067235296418e-05,
"loss": 0.1006,
"step": 1560
},
{
"epoch": 0.37497014568903747,
"grad_norm": 0.5452921986579895,
"learning_rate": 9.427501763623698e-05,
"loss": 0.0927,
"step": 1570
},
{
"epoch": 0.37735849056603776,
"grad_norm": 0.6011863946914673,
"learning_rate": 9.417860638161733e-05,
"loss": 0.0906,
"step": 1580
},
{
"epoch": 0.379746835443038,
"grad_norm": 0.4528885781764984,
"learning_rate": 9.408144023650754e-05,
"loss": 0.0967,
"step": 1590
},
{
"epoch": 0.38213518032003824,
"grad_norm": 0.4767697751522064,
"learning_rate": 9.398352086120891e-05,
"loss": 0.095,
"step": 1600
},
{
"epoch": 0.3845235251970385,
"grad_norm": 0.6745051741600037,
"learning_rate": 9.388484992889346e-05,
"loss": 0.1127,
"step": 1610
},
{
"epoch": 0.3869118700740387,
"grad_norm": 0.6268698573112488,
"learning_rate": 9.378542912557513e-05,
"loss": 0.1145,
"step": 1620
},
{
"epoch": 0.38930021495103895,
"grad_norm": 0.9256007671356201,
"learning_rate": 9.368526015008116e-05,
"loss": 0.0968,
"step": 1630
},
{
"epoch": 0.3916885598280392,
"grad_norm": 0.7772823572158813,
"learning_rate": 9.358434471402296e-05,
"loss": 0.101,
"step": 1640
},
{
"epoch": 0.3940769047050394,
"grad_norm": 0.5281357169151306,
"learning_rate": 9.348268454176687e-05,
"loss": 0.1063,
"step": 1650
},
{
"epoch": 0.39646524958203966,
"grad_norm": 0.6283060312271118,
"learning_rate": 9.33802813704048e-05,
"loss": 0.1125,
"step": 1660
},
{
"epoch": 0.3988535944590399,
"grad_norm": 0.41661298274993896,
"learning_rate": 9.327713694972436e-05,
"loss": 0.0921,
"step": 1670
},
{
"epoch": 0.40124193933604013,
"grad_norm": 0.7662723660469055,
"learning_rate": 9.317325304217914e-05,
"loss": 0.0988,
"step": 1680
},
{
"epoch": 0.40363028421304037,
"grad_norm": 0.5023656487464905,
"learning_rate": 9.306863142285848e-05,
"loss": 0.1111,
"step": 1690
},
{
"epoch": 0.4060186290900406,
"grad_norm": 0.7543429732322693,
"learning_rate": 9.29632738794572e-05,
"loss": 0.0932,
"step": 1700
},
{
"epoch": 0.40840697396704084,
"grad_norm": 0.6077080368995667,
"learning_rate": 9.285718221224504e-05,
"loss": 0.0944,
"step": 1710
},
{
"epoch": 0.4107953188440411,
"grad_norm": 0.7034826278686523,
"learning_rate": 9.275035823403584e-05,
"loss": 0.0936,
"step": 1720
},
{
"epoch": 0.4131836637210413,
"grad_norm": 1.0592331886291504,
"learning_rate": 9.264280377015669e-05,
"loss": 0.0973,
"step": 1730
},
{
"epoch": 0.41557200859804155,
"grad_norm": 0.5010759234428406,
"learning_rate": 9.25345206584166e-05,
"loss": 0.1056,
"step": 1740
},
{
"epoch": 0.4179603534750418,
"grad_norm": 0.6567232608795166,
"learning_rate": 9.242551074907519e-05,
"loss": 0.0851,
"step": 1750
},
{
"epoch": 0.420348698352042,
"grad_norm": 0.5500295758247375,
"learning_rate": 9.231577590481104e-05,
"loss": 0.1087,
"step": 1760
},
{
"epoch": 0.42273704322904226,
"grad_norm": 0.9526474475860596,
"learning_rate": 9.220531800068988e-05,
"loss": 0.1047,
"step": 1770
},
{
"epoch": 0.4251253881060425,
"grad_norm": 0.33729642629623413,
"learning_rate": 9.209413892413252e-05,
"loss": 0.0775,
"step": 1780
},
{
"epoch": 0.42751373298304274,
"grad_norm": 0.5979863405227661,
"learning_rate": 9.198224057488262e-05,
"loss": 0.1006,
"step": 1790
},
{
"epoch": 0.429902077860043,
"grad_norm": 0.5933387279510498,
"learning_rate": 9.186962486497425e-05,
"loss": 0.1082,
"step": 1800
},
{
"epoch": 0.4322904227370432,
"grad_norm": 0.5830422639846802,
"learning_rate": 9.175629371869918e-05,
"loss": 0.0992,
"step": 1810
},
{
"epoch": 0.43467876761404345,
"grad_norm": 0.5181281566619873,
"learning_rate": 9.164224907257395e-05,
"loss": 0.088,
"step": 1820
},
{
"epoch": 0.4370671124910437,
"grad_norm": 0.5915122032165527,
"learning_rate": 9.152749287530698e-05,
"loss": 0.1167,
"step": 1830
},
{
"epoch": 0.4394554573680439,
"grad_norm": 0.6373990178108215,
"learning_rate": 9.1412027087765e-05,
"loss": 0.1097,
"step": 1840
},
{
"epoch": 0.44184380224504416,
"grad_norm": 0.564217746257782,
"learning_rate": 9.129585368293975e-05,
"loss": 0.0922,
"step": 1850
},
{
"epoch": 0.4442321471220444,
"grad_norm": 0.6150906682014465,
"learning_rate": 9.117897464591423e-05,
"loss": 0.1098,
"step": 1860
},
{
"epoch": 0.4466204919990447,
"grad_norm": 0.4399804174900055,
"learning_rate": 9.106139197382864e-05,
"loss": 0.0886,
"step": 1870
},
{
"epoch": 0.4490088368760449,
"grad_norm": 0.7823533415794373,
"learning_rate": 9.094310767584646e-05,
"loss": 0.101,
"step": 1880
},
{
"epoch": 0.45139718175304516,
"grad_norm": 0.5411809086799622,
"learning_rate": 9.082412377311998e-05,
"loss": 0.1146,
"step": 1890
},
{
"epoch": 0.4537855266300454,
"grad_norm": 0.4220235049724579,
"learning_rate": 9.070444229875581e-05,
"loss": 0.098,
"step": 1900
},
{
"epoch": 0.45617387150704564,
"grad_norm": 0.5104357004165649,
"learning_rate": 9.058406529778012e-05,
"loss": 0.0872,
"step": 1910
},
{
"epoch": 0.4585622163840459,
"grad_norm": 0.4071063995361328,
"learning_rate": 9.046299482710374e-05,
"loss": 0.0819,
"step": 1920
},
{
"epoch": 0.4609505612610461,
"grad_norm": 0.5750535130500793,
"learning_rate": 9.034123295548695e-05,
"loss": 0.1018,
"step": 1930
},
{
"epoch": 0.46333890613804635,
"grad_norm": 0.39777764678001404,
"learning_rate": 9.021878176350423e-05,
"loss": 0.0981,
"step": 1940
},
{
"epoch": 0.4657272510150466,
"grad_norm": 0.6830609440803528,
"learning_rate": 9.009564334350855e-05,
"loss": 0.1107,
"step": 1950
},
{
"epoch": 0.4681155958920468,
"grad_norm": 0.8429521918296814,
"learning_rate": 8.997181979959577e-05,
"loss": 0.1229,
"step": 1960
},
{
"epoch": 0.47050394076904706,
"grad_norm": 0.6539687514305115,
"learning_rate": 8.984731324756863e-05,
"loss": 0.1061,
"step": 1970
},
{
"epoch": 0.4728922856460473,
"grad_norm": 0.28856325149536133,
"learning_rate": 8.972212581490056e-05,
"loss": 0.1201,
"step": 1980
},
{
"epoch": 0.47528063052304753,
"grad_norm": 0.5003455877304077,
"learning_rate": 8.959625964069938e-05,
"loss": 0.0992,
"step": 1990
},
{
"epoch": 0.47766897540004777,
"grad_norm": 0.692571759223938,
"learning_rate": 8.946971687567074e-05,
"loss": 0.1063,
"step": 2000
},
{
"epoch": 0.480057320277048,
"grad_norm": 0.5598740577697754,
"learning_rate": 8.934249968208134e-05,
"loss": 0.0933,
"step": 2010
},
{
"epoch": 0.48244566515404824,
"grad_norm": 0.5776444673538208,
"learning_rate": 8.921461023372204e-05,
"loss": 0.0873,
"step": 2020
},
{
"epoch": 0.4848340100310485,
"grad_norm": 0.7545134425163269,
"learning_rate": 8.908605071587064e-05,
"loss": 0.1052,
"step": 2030
},
{
"epoch": 0.4872223549080487,
"grad_norm": 0.5354888439178467,
"learning_rate": 8.895682332525456e-05,
"loss": 0.0912,
"step": 2040
},
{
"epoch": 0.48961069978504895,
"grad_norm": 0.7095310091972351,
"learning_rate": 8.882693027001341e-05,
"loss": 0.0988,
"step": 2050
},
{
"epoch": 0.4919990446620492,
"grad_norm": 0.5006904602050781,
"learning_rate": 8.869637376966109e-05,
"loss": 0.0966,
"step": 2060
},
{
"epoch": 0.4943873895390494,
"grad_norm": 0.6439457535743713,
"learning_rate": 8.856515605504796e-05,
"loss": 0.1012,
"step": 2070
},
{
"epoch": 0.49677573441604966,
"grad_norm": 0.7942578196525574,
"learning_rate": 8.843327936832274e-05,
"loss": 0.0994,
"step": 2080
},
{
"epoch": 0.4991640792930499,
"grad_norm": 0.4847765564918518,
"learning_rate": 8.830074596289414e-05,
"loss": 0.0943,
"step": 2090
},
{
"epoch": 0.5015524241700502,
"grad_norm": 0.7488170862197876,
"learning_rate": 8.816755810339238e-05,
"loss": 0.1044,
"step": 2100
},
{
"epoch": 0.5039407690470504,
"grad_norm": 0.49015161395072937,
"learning_rate": 8.803371806563052e-05,
"loss": 0.099,
"step": 2110
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.5630750060081482,
"learning_rate": 8.789922813656551e-05,
"loss": 0.118,
"step": 2120
},
{
"epoch": 0.5087174588010509,
"grad_norm": 0.5637487173080444,
"learning_rate": 8.776409061425919e-05,
"loss": 0.1093,
"step": 2130
},
{
"epoch": 0.5111058036780511,
"grad_norm": 0.5367555618286133,
"learning_rate": 8.762830780783893e-05,
"loss": 0.0991,
"step": 2140
},
{
"epoch": 0.5134941485550514,
"grad_norm": 0.3627997934818268,
"learning_rate": 8.749188203745826e-05,
"loss": 0.0929,
"step": 2150
},
{
"epoch": 0.5158824934320516,
"grad_norm": 0.5912765264511108,
"learning_rate": 8.73548156342572e-05,
"loss": 0.1013,
"step": 2160
},
{
"epoch": 0.5182708383090519,
"grad_norm": 0.4376015067100525,
"learning_rate": 8.721711094032239e-05,
"loss": 0.0886,
"step": 2170
},
{
"epoch": 0.5206591831860521,
"grad_norm": 0.7576462626457214,
"learning_rate": 8.707877030864712e-05,
"loss": 0.087,
"step": 2180
},
{
"epoch": 0.5230475280630523,
"grad_norm": 0.47268545627593994,
"learning_rate": 8.693979610309106e-05,
"loss": 0.0973,
"step": 2190
},
{
"epoch": 0.5254358729400526,
"grad_norm": 0.5312506556510925,
"learning_rate": 8.680019069834e-05,
"loss": 0.092,
"step": 2200
},
{
"epoch": 0.5278242178170528,
"grad_norm": 0.5842658281326294,
"learning_rate": 8.665995647986506e-05,
"loss": 0.0863,
"step": 2210
},
{
"epoch": 0.530212562694053,
"grad_norm": 0.6047576665878296,
"learning_rate": 8.651909584388214e-05,
"loss": 0.0994,
"step": 2220
},
{
"epoch": 0.5326009075710533,
"grad_norm": 0.5044541358947754,
"learning_rate": 8.637761119731086e-05,
"loss": 0.0888,
"step": 2230
},
{
"epoch": 0.5349892524480535,
"grad_norm": 0.43564099073410034,
"learning_rate": 8.623550495773347e-05,
"loss": 0.0908,
"step": 2240
},
{
"epoch": 0.5373775973250537,
"grad_norm": 0.4529109597206116,
"learning_rate": 8.609277955335351e-05,
"loss": 0.0936,
"step": 2250
},
{
"epoch": 0.539765942202054,
"grad_norm": 0.7701318264007568,
"learning_rate": 8.594943742295436e-05,
"loss": 0.1065,
"step": 2260
},
{
"epoch": 0.5421542870790542,
"grad_norm": 0.4490266740322113,
"learning_rate": 8.58054810158575e-05,
"loss": 0.1045,
"step": 2270
},
{
"epoch": 0.5445426319560545,
"grad_norm": 0.5808409452438354,
"learning_rate": 8.566091279188078e-05,
"loss": 0.0924,
"step": 2280
},
{
"epoch": 0.5469309768330547,
"grad_norm": 0.6063486933708191,
"learning_rate": 8.551573522129626e-05,
"loss": 0.1001,
"step": 2290
},
{
"epoch": 0.5493193217100549,
"grad_norm": 0.5852213501930237,
"learning_rate": 8.536995078478805e-05,
"loss": 0.0972,
"step": 2300
},
{
"epoch": 0.5517076665870552,
"grad_norm": 0.6312397718429565,
"learning_rate": 8.522356197340993e-05,
"loss": 0.0882,
"step": 2310
},
{
"epoch": 0.5540960114640554,
"grad_norm": 0.5887095928192139,
"learning_rate": 8.50765712885428e-05,
"loss": 0.1107,
"step": 2320
},
{
"epoch": 0.5564843563410556,
"grad_norm": 0.7400378584861755,
"learning_rate": 8.492898124185189e-05,
"loss": 0.0984,
"step": 2330
},
{
"epoch": 0.5588727012180559,
"grad_norm": 0.5219861268997192,
"learning_rate": 8.478079435524391e-05,
"loss": 0.1034,
"step": 2340
},
{
"epoch": 0.5612610460950561,
"grad_norm": 0.5666329264640808,
"learning_rate": 8.463201316082387e-05,
"loss": 0.0943,
"step": 2350
},
{
"epoch": 0.5636493909720564,
"grad_norm": 0.4448127746582031,
"learning_rate": 8.44826402008519e-05,
"loss": 0.0892,
"step": 2360
},
{
"epoch": 0.5660377358490566,
"grad_norm": 0.2999361455440521,
"learning_rate": 8.433267802769973e-05,
"loss": 0.0928,
"step": 2370
},
{
"epoch": 0.5684260807260568,
"grad_norm": 0.5117233991622925,
"learning_rate": 8.418212920380715e-05,
"loss": 0.0882,
"step": 2380
},
{
"epoch": 0.5708144256030571,
"grad_norm": 0.3534957766532898,
"learning_rate": 8.403099630163817e-05,
"loss": 0.1116,
"step": 2390
},
{
"epoch": 0.5732027704800573,
"grad_norm": 0.5342962145805359,
"learning_rate": 8.387928190363711e-05,
"loss": 0.0881,
"step": 2400
},
{
"epoch": 0.5755911153570575,
"grad_norm": 0.3818695843219757,
"learning_rate": 8.372698860218444e-05,
"loss": 0.0963,
"step": 2410
},
{
"epoch": 0.5779794602340578,
"grad_norm": 0.782770037651062,
"learning_rate": 8.35741189995525e-05,
"loss": 0.0953,
"step": 2420
},
{
"epoch": 0.580367805111058,
"grad_norm": 0.5747333765029907,
"learning_rate": 8.342067570786099e-05,
"loss": 0.0894,
"step": 2430
},
{
"epoch": 0.5827561499880582,
"grad_norm": 0.5100058317184448,
"learning_rate": 8.326666134903241e-05,
"loss": 0.0895,
"step": 2440
},
{
"epoch": 0.5851444948650585,
"grad_norm": 0.43753543496131897,
"learning_rate": 8.31120785547472e-05,
"loss": 0.0913,
"step": 2450
},
{
"epoch": 0.5875328397420587,
"grad_norm": 0.7520521879196167,
"learning_rate": 8.295692996639886e-05,
"loss": 0.1117,
"step": 2460
},
{
"epoch": 0.589921184619059,
"grad_norm": 0.5778405666351318,
"learning_rate": 8.280121823504863e-05,
"loss": 0.0991,
"step": 2470
},
{
"epoch": 0.5923095294960592,
"grad_norm": 0.5665181875228882,
"learning_rate": 8.264494602138042e-05,
"loss": 0.0946,
"step": 2480
},
{
"epoch": 0.5946978743730594,
"grad_norm": 0.5809256434440613,
"learning_rate": 8.248811599565515e-05,
"loss": 0.0922,
"step": 2490
},
{
"epoch": 0.5970862192500597,
"grad_norm": 0.6336116194725037,
"learning_rate": 8.233073083766531e-05,
"loss": 0.1028,
"step": 2500
},
{
"epoch": 0.5994745641270599,
"grad_norm": 0.4314132034778595,
"learning_rate": 8.217279323668895e-05,
"loss": 0.0906,
"step": 2510
},
{
"epoch": 0.6018629090040601,
"grad_norm": 0.5118670463562012,
"learning_rate": 8.201430589144395e-05,
"loss": 0.087,
"step": 2520
},
{
"epoch": 0.6042512538810604,
"grad_norm": 0.45639073848724365,
"learning_rate": 8.185527151004173e-05,
"loss": 0.0891,
"step": 2530
},
{
"epoch": 0.6066395987580606,
"grad_norm": 0.5048205852508545,
"learning_rate": 8.169569280994107e-05,
"loss": 0.0958,
"step": 2540
},
{
"epoch": 0.6090279436350609,
"grad_norm": 0.48785558342933655,
"learning_rate": 8.153557251790171e-05,
"loss": 0.0832,
"step": 2550
},
{
"epoch": 0.6114162885120611,
"grad_norm": 0.49757617712020874,
"learning_rate": 8.137491336993761e-05,
"loss": 0.0923,
"step": 2560
},
{
"epoch": 0.6138046333890613,
"grad_norm": 0.512619137763977,
"learning_rate": 8.121371811127034e-05,
"loss": 0.0984,
"step": 2570
},
{
"epoch": 0.6161929782660616,
"grad_norm": 0.5055660009384155,
"learning_rate": 8.105198949628215e-05,
"loss": 0.0918,
"step": 2580
},
{
"epoch": 0.6185813231430619,
"grad_norm": 0.49760058522224426,
"learning_rate": 8.08897302884688e-05,
"loss": 0.1024,
"step": 2590
},
{
"epoch": 0.6209696680200621,
"grad_norm": 0.48902204632759094,
"learning_rate": 8.072694326039248e-05,
"loss": 0.0989,
"step": 2600
},
{
"epoch": 0.6233580128970624,
"grad_norm": 0.5674299597740173,
"learning_rate": 8.056363119363437e-05,
"loss": 0.0936,
"step": 2610
},
{
"epoch": 0.6257463577740626,
"grad_norm": 0.695819079875946,
"learning_rate": 8.03997968787471e-05,
"loss": 0.0877,
"step": 2620
},
{
"epoch": 0.6281347026510629,
"grad_norm": 0.6024104952812195,
"learning_rate": 8.023544311520705e-05,
"loss": 0.0942,
"step": 2630
},
{
"epoch": 0.6305230475280631,
"grad_norm": 0.6605867147445679,
"learning_rate": 8.00705727113666e-05,
"loss": 0.0879,
"step": 2640
},
{
"epoch": 0.6329113924050633,
"grad_norm": 0.5188372731208801,
"learning_rate": 7.990518848440604e-05,
"loss": 0.0921,
"step": 2650
},
{
"epoch": 0.6352997372820636,
"grad_norm": 0.48158928751945496,
"learning_rate": 7.973929326028551e-05,
"loss": 0.0831,
"step": 2660
},
{
"epoch": 0.6376880821590638,
"grad_norm": 0.31054624915122986,
"learning_rate": 7.957288987369664e-05,
"loss": 0.0954,
"step": 2670
},
{
"epoch": 0.640076427036064,
"grad_norm": 0.5120779871940613,
"learning_rate": 7.940598116801423e-05,
"loss": 0.0938,
"step": 2680
},
{
"epoch": 0.6424647719130643,
"grad_norm": 0.9595939517021179,
"learning_rate": 7.923856999524751e-05,
"loss": 0.0925,
"step": 2690
},
{
"epoch": 0.6448531167900645,
"grad_norm": 0.46394315361976624,
"learning_rate": 7.907065921599154e-05,
"loss": 0.0973,
"step": 2700
},
{
"epoch": 0.6472414616670648,
"grad_norm": 0.5199283361434937,
"learning_rate": 7.890225169937824e-05,
"loss": 0.0875,
"step": 2710
},
{
"epoch": 0.649629806544065,
"grad_norm": 0.5251132249832153,
"learning_rate": 7.873335032302742e-05,
"loss": 0.0936,
"step": 2720
},
{
"epoch": 0.6520181514210652,
"grad_norm": 0.4188412129878998,
"learning_rate": 7.856395797299759e-05,
"loss": 0.0924,
"step": 2730
},
{
"epoch": 0.6544064962980655,
"grad_norm": 0.4517621099948883,
"learning_rate": 7.839407754373665e-05,
"loss": 0.0907,
"step": 2740
},
{
"epoch": 0.6567948411750657,
"grad_norm": 0.5232812762260437,
"learning_rate": 7.82237119380324e-05,
"loss": 0.0936,
"step": 2750
},
{
"epoch": 0.6591831860520659,
"grad_norm": 0.4168354272842407,
"learning_rate": 7.805286406696302e-05,
"loss": 0.089,
"step": 2760
},
{
"epoch": 0.6615715309290662,
"grad_norm": 0.83026123046875,
"learning_rate": 7.78815368498472e-05,
"loss": 0.0871,
"step": 2770
},
{
"epoch": 0.6639598758060664,
"grad_norm": 0.7121137380599976,
"learning_rate": 7.77097332141944e-05,
"loss": 0.0948,
"step": 2780
},
{
"epoch": 0.6663482206830666,
"grad_norm": 0.37756726145744324,
"learning_rate": 7.753745609565472e-05,
"loss": 0.0855,
"step": 2790
},
{
"epoch": 0.6687365655600669,
"grad_norm": 0.48371440172195435,
"learning_rate": 7.73647084379688e-05,
"loss": 0.0806,
"step": 2800
},
{
"epoch": 0.6711249104370671,
"grad_norm": 0.38711991906166077,
"learning_rate": 7.719149319291747e-05,
"loss": 0.0858,
"step": 2810
},
{
"epoch": 0.6735132553140674,
"grad_norm": 0.5841792821884155,
"learning_rate": 7.701781332027136e-05,
"loss": 0.0835,
"step": 2820
},
{
"epoch": 0.6759016001910676,
"grad_norm": 0.8211458921432495,
"learning_rate": 7.684367178774029e-05,
"loss": 0.1028,
"step": 2830
},
{
"epoch": 0.6782899450680678,
"grad_norm": 0.5430380702018738,
"learning_rate": 7.666907157092256e-05,
"loss": 0.0945,
"step": 2840
},
{
"epoch": 0.6806782899450681,
"grad_norm": 0.46453964710235596,
"learning_rate": 7.649401565325418e-05,
"loss": 0.0795,
"step": 2850
},
{
"epoch": 0.6830666348220683,
"grad_norm": 0.391396701335907,
"learning_rate": 7.631850702595776e-05,
"loss": 0.0937,
"step": 2860
},
{
"epoch": 0.6854549796990685,
"grad_norm": 0.49795594811439514,
"learning_rate": 7.614254868799146e-05,
"loss": 0.0949,
"step": 2870
},
{
"epoch": 0.6878433245760688,
"grad_norm": 0.38805457949638367,
"learning_rate": 7.596614364599787e-05,
"loss": 0.0863,
"step": 2880
},
{
"epoch": 0.690231669453069,
"grad_norm": 0.5313634276390076,
"learning_rate": 7.578929491425238e-05,
"loss": 0.087,
"step": 2890
},
{
"epoch": 0.6926200143300693,
"grad_norm": 0.5416390299797058,
"learning_rate": 7.56120055146119e-05,
"loss": 0.0976,
"step": 2900
},
{
"epoch": 0.6950083592070695,
"grad_norm": 0.4653552770614624,
"learning_rate": 7.543427847646309e-05,
"loss": 0.0813,
"step": 2910
},
{
"epoch": 0.6973967040840697,
"grad_norm": 0.4295271933078766,
"learning_rate": 7.525611683667068e-05,
"loss": 0.0841,
"step": 2920
},
{
"epoch": 0.69978504896107,
"grad_norm": 0.3986028730869293,
"learning_rate": 7.507752363952554e-05,
"loss": 0.0862,
"step": 2930
},
{
"epoch": 0.7021733938380702,
"grad_norm": 0.6848331093788147,
"learning_rate": 7.489850193669264e-05,
"loss": 0.0906,
"step": 2940
},
{
"epoch": 0.7045617387150704,
"grad_norm": 0.5594953894615173,
"learning_rate": 7.471905478715893e-05,
"loss": 0.1067,
"step": 2950
},
{
"epoch": 0.7069500835920707,
"grad_norm": 0.5157433748245239,
"learning_rate": 7.453918525718111e-05,
"loss": 0.0856,
"step": 2960
},
{
"epoch": 0.7093384284690709,
"grad_norm": 0.6397213935852051,
"learning_rate": 7.435889642023316e-05,
"loss": 0.0889,
"step": 2970
},
{
"epoch": 0.7117267733460712,
"grad_norm": 0.534459114074707,
"learning_rate": 7.417819135695388e-05,
"loss": 0.0925,
"step": 2980
},
{
"epoch": 0.7141151182230714,
"grad_norm": 0.624421238899231,
"learning_rate": 7.399707315509418e-05,
"loss": 0.0916,
"step": 2990
},
{
"epoch": 0.7165034631000716,
"grad_norm": 0.5104103684425354,
"learning_rate": 7.381554490946446e-05,
"loss": 0.1034,
"step": 3000
},
{
"epoch": 0.7188918079770719,
"grad_norm": 0.4029245674610138,
"learning_rate": 7.363360972188156e-05,
"loss": 0.0817,
"step": 3010
},
{
"epoch": 0.7212801528540721,
"grad_norm": 0.5501995086669922,
"learning_rate": 7.345127070111586e-05,
"loss": 0.0818,
"step": 3020
},
{
"epoch": 0.7236684977310723,
"grad_norm": 0.548947274684906,
"learning_rate": 7.326853096283813e-05,
"loss": 0.0872,
"step": 3030
},
{
"epoch": 0.7260568426080726,
"grad_norm": 0.4799128770828247,
"learning_rate": 7.308539362956632e-05,
"loss": 0.0855,
"step": 3040
},
{
"epoch": 0.7284451874850728,
"grad_norm": 0.4064118564128876,
"learning_rate": 7.290186183061212e-05,
"loss": 0.0886,
"step": 3050
},
{
"epoch": 0.730833532362073,
"grad_norm": 0.5113388299942017,
"learning_rate": 7.271793870202765e-05,
"loss": 0.0803,
"step": 3060
},
{
"epoch": 0.7332218772390733,
"grad_norm": 0.4490203559398651,
"learning_rate": 7.253362738655168e-05,
"loss": 0.0784,
"step": 3070
},
{
"epoch": 0.7356102221160735,
"grad_norm": 0.5459372997283936,
"learning_rate": 7.234893103355607e-05,
"loss": 0.0905,
"step": 3080
},
{
"epoch": 0.7379985669930738,
"grad_norm": 0.5430892705917358,
"learning_rate": 7.216385279899189e-05,
"loss": 0.0912,
"step": 3090
},
{
"epoch": 0.740386911870074,
"grad_norm": 0.6683062314987183,
"learning_rate": 7.197839584533551e-05,
"loss": 0.1031,
"step": 3100
},
{
"epoch": 0.7427752567470742,
"grad_norm": 0.5698011517524719,
"learning_rate": 7.179256334153461e-05,
"loss": 0.0885,
"step": 3110
},
{
"epoch": 0.7451636016240745,
"grad_norm": 0.5895237922668457,
"learning_rate": 7.160635846295392e-05,
"loss": 0.1029,
"step": 3120
},
{
"epoch": 0.7475519465010747,
"grad_norm": 0.42501479387283325,
"learning_rate": 7.141978439132108e-05,
"loss": 0.082,
"step": 3130
},
{
"epoch": 0.7499402913780749,
"grad_norm": 0.45501509308815,
"learning_rate": 7.123284431467217e-05,
"loss": 0.0827,
"step": 3140
},
{
"epoch": 0.7523286362550752,
"grad_norm": 0.49989697337150574,
"learning_rate": 7.104554142729736e-05,
"loss": 0.0906,
"step": 3150
},
{
"epoch": 0.7547169811320755,
"grad_norm": 0.5568277835845947,
"learning_rate": 7.085787892968616e-05,
"loss": 0.0699,
"step": 3160
},
{
"epoch": 0.7571053260090758,
"grad_norm": 0.4961911141872406,
"learning_rate": 7.06698600284729e-05,
"loss": 0.0815,
"step": 3170
},
{
"epoch": 0.759493670886076,
"grad_norm": 0.4467281699180603,
"learning_rate": 7.048148793638184e-05,
"loss": 0.0788,
"step": 3180
},
{
"epoch": 0.7618820157630762,
"grad_norm": 0.5697978734970093,
"learning_rate": 7.029276587217229e-05,
"loss": 0.0947,
"step": 3190
},
{
"epoch": 0.7642703606400765,
"grad_norm": 0.5323075652122498,
"learning_rate": 7.010369706058365e-05,
"loss": 0.0911,
"step": 3200
},
{
"epoch": 0.7666587055170767,
"grad_norm": 0.35793957114219666,
"learning_rate": 6.991428473228022e-05,
"loss": 0.083,
"step": 3210
},
{
"epoch": 0.769047050394077,
"grad_norm": 0.34435173869132996,
"learning_rate": 6.972453212379609e-05,
"loss": 0.0894,
"step": 3220
},
{
"epoch": 0.7714353952710772,
"grad_norm": 0.3402794301509857,
"learning_rate": 6.953444247747982e-05,
"loss": 0.0813,
"step": 3230
},
{
"epoch": 0.7738237401480774,
"grad_norm": 0.5104530453681946,
"learning_rate": 6.934401904143895e-05,
"loss": 0.0974,
"step": 3240
},
{
"epoch": 0.7762120850250777,
"grad_norm": 0.3650020956993103,
"learning_rate": 6.915326506948462e-05,
"loss": 0.0941,
"step": 3250
},
{
"epoch": 0.7786004299020779,
"grad_norm": 0.5072294473648071,
"learning_rate": 6.896218382107592e-05,
"loss": 0.0823,
"step": 3260
},
{
"epoch": 0.7809887747790781,
"grad_norm": 0.4305017292499542,
"learning_rate": 6.877077856126416e-05,
"loss": 0.0927,
"step": 3270
},
{
"epoch": 0.7833771196560784,
"grad_norm": 0.6784612536430359,
"learning_rate": 6.857905256063713e-05,
"loss": 0.1062,
"step": 3280
},
{
"epoch": 0.7857654645330786,
"grad_norm": 0.41373685002326965,
"learning_rate": 6.838700909526319e-05,
"loss": 0.0967,
"step": 3290
},
{
"epoch": 0.7881538094100788,
"grad_norm": 0.45850133895874023,
"learning_rate": 6.819465144663529e-05,
"loss": 0.0955,
"step": 3300
},
{
"epoch": 0.7905421542870791,
"grad_norm": 0.34656307101249695,
"learning_rate": 6.800198290161492e-05,
"loss": 0.0886,
"step": 3310
},
{
"epoch": 0.7929304991640793,
"grad_norm": 0.42274972796440125,
"learning_rate": 6.780900675237592e-05,
"loss": 0.0879,
"step": 3320
},
{
"epoch": 0.7953188440410796,
"grad_norm": 0.435006707906723,
"learning_rate": 6.761572629634824e-05,
"loss": 0.0844,
"step": 3330
},
{
"epoch": 0.7977071889180798,
"grad_norm": 0.755415141582489,
"learning_rate": 6.742214483616162e-05,
"loss": 0.0925,
"step": 3340
},
{
"epoch": 0.80009553379508,
"grad_norm": 0.5019872784614563,
"learning_rate": 6.722826567958907e-05,
"loss": 0.0879,
"step": 3350
},
{
"epoch": 0.8024838786720803,
"grad_norm": 0.37853872776031494,
"learning_rate": 6.703409213949045e-05,
"loss": 0.093,
"step": 3360
},
{
"epoch": 0.8048722235490805,
"grad_norm": 0.5818341970443726,
"learning_rate": 6.683962753375585e-05,
"loss": 0.1046,
"step": 3370
},
{
"epoch": 0.8072605684260807,
"grad_norm": 0.42232608795166016,
"learning_rate": 6.664487518524878e-05,
"loss": 0.0796,
"step": 3380
},
{
"epoch": 0.809648913303081,
"grad_norm": 0.42375126481056213,
"learning_rate": 6.644983842174956e-05,
"loss": 0.082,
"step": 3390
},
{
"epoch": 0.8120372581800812,
"grad_norm": 0.36204779148101807,
"learning_rate": 6.62545205758984e-05,
"loss": 0.0921,
"step": 3400
},
{
"epoch": 0.8144256030570814,
"grad_norm": 0.6738161444664001,
"learning_rate": 6.605892498513831e-05,
"loss": 0.0899,
"step": 3410
},
{
"epoch": 0.8168139479340817,
"grad_norm": 0.5603547096252441,
"learning_rate": 6.586305499165832e-05,
"loss": 0.0919,
"step": 3420
},
{
"epoch": 0.8192022928110819,
"grad_norm": 0.5394812822341919,
"learning_rate": 6.566691394233619e-05,
"loss": 0.0851,
"step": 3430
},
{
"epoch": 0.8215906376880822,
"grad_norm": 0.5879244208335876,
"learning_rate": 6.547050518868126e-05,
"loss": 0.1056,
"step": 3440
},
{
"epoch": 0.8239789825650824,
"grad_norm": 0.4407655894756317,
"learning_rate": 6.527383208677723e-05,
"loss": 0.0856,
"step": 3450
},
{
"epoch": 0.8263673274420826,
"grad_norm": 0.478908896446228,
"learning_rate": 6.507689799722478e-05,
"loss": 0.1052,
"step": 3460
},
{
"epoch": 0.8287556723190829,
"grad_norm": 0.35551053285598755,
"learning_rate": 6.487970628508414e-05,
"loss": 0.0952,
"step": 3470
},
{
"epoch": 0.8311440171960831,
"grad_norm": 0.4705964922904968,
"learning_rate": 6.468226031981758e-05,
"loss": 0.0805,
"step": 3480
},
{
"epoch": 0.8335323620730833,
"grad_norm": 0.49963054060935974,
"learning_rate": 6.44845634752319e-05,
"loss": 0.0929,
"step": 3490
},
{
"epoch": 0.8359207069500836,
"grad_norm": 0.3886867165565491,
"learning_rate": 6.428661912942069e-05,
"loss": 0.0855,
"step": 3500
},
{
"epoch": 0.8383090518270838,
"grad_norm": 0.3998318314552307,
"learning_rate": 6.408843066470668e-05,
"loss": 0.0828,
"step": 3510
},
{
"epoch": 0.840697396704084,
"grad_norm": 0.3639577627182007,
"learning_rate": 6.389000146758392e-05,
"loss": 0.0936,
"step": 3520
},
{
"epoch": 0.8430857415810843,
"grad_norm": 0.3037131726741791,
"learning_rate": 6.369133492865989e-05,
"loss": 0.082,
"step": 3530
},
{
"epoch": 0.8454740864580845,
"grad_norm": 0.47817716002464294,
"learning_rate": 6.349243444259763e-05,
"loss": 0.0944,
"step": 3540
},
{
"epoch": 0.8478624313350848,
"grad_norm": 0.5974783301353455,
"learning_rate": 6.329330340805761e-05,
"loss": 0.0736,
"step": 3550
},
{
"epoch": 0.850250776212085,
"grad_norm": 0.5753908753395081,
"learning_rate": 6.309394522763983e-05,
"loss": 0.0904,
"step": 3560
},
{
"epoch": 0.8526391210890852,
"grad_norm": 0.3667502999305725,
"learning_rate": 6.289436330782553e-05,
"loss": 0.0836,
"step": 3570
},
{
"epoch": 0.8550274659660855,
"grad_norm": 0.7074549794197083,
"learning_rate": 6.269456105891906e-05,
"loss": 0.0945,
"step": 3580
},
{
"epoch": 0.8574158108430857,
"grad_norm": 0.5719613432884216,
"learning_rate": 6.249454189498958e-05,
"loss": 0.0795,
"step": 3590
},
{
"epoch": 0.859804155720086,
"grad_norm": 0.5231560468673706,
"learning_rate": 6.229430923381272e-05,
"loss": 0.0849,
"step": 3600
},
{
"epoch": 0.8621925005970862,
"grad_norm": 0.35364142060279846,
"learning_rate": 6.209386649681223e-05,
"loss": 0.0844,
"step": 3610
},
{
"epoch": 0.8645808454740864,
"grad_norm": 0.45364946126937866,
"learning_rate": 6.18932171090014e-05,
"loss": 0.0785,
"step": 3620
},
{
"epoch": 0.8669691903510867,
"grad_norm": 0.4877249300479889,
"learning_rate": 6.169236449892469e-05,
"loss": 0.088,
"step": 3630
},
{
"epoch": 0.8693575352280869,
"grad_norm": 0.4699724614620209,
"learning_rate": 6.149131209859902e-05,
"loss": 0.0718,
"step": 3640
},
{
"epoch": 0.8717458801050871,
"grad_norm": 0.46347159147262573,
"learning_rate": 6.12900633434552e-05,
"loss": 0.0958,
"step": 3650
},
{
"epoch": 0.8741342249820874,
"grad_norm": 0.4624701142311096,
"learning_rate": 6.108862167227916e-05,
"loss": 0.0754,
"step": 3660
},
{
"epoch": 0.8765225698590876,
"grad_norm": 0.8754489421844482,
"learning_rate": 6.088699052715332e-05,
"loss": 0.0938,
"step": 3670
},
{
"epoch": 0.8789109147360878,
"grad_norm": 0.5448271036148071,
"learning_rate": 6.068517335339758e-05,
"loss": 0.0771,
"step": 3680
},
{
"epoch": 0.8812992596130881,
"grad_norm": 0.5465822815895081,
"learning_rate": 6.048317359951063e-05,
"loss": 0.0904,
"step": 3690
},
{
"epoch": 0.8836876044900883,
"grad_norm": 0.6568249464035034,
"learning_rate": 6.028099471711094e-05,
"loss": 0.0922,
"step": 3700
},
{
"epoch": 0.8860759493670886,
"grad_norm": 0.6030942797660828,
"learning_rate": 6.0078640160877754e-05,
"loss": 0.0875,
"step": 3710
},
{
"epoch": 0.8884642942440888,
"grad_norm": 0.406304270029068,
"learning_rate": 5.9876113388492125e-05,
"loss": 0.0975,
"step": 3720
},
{
"epoch": 0.8908526391210891,
"grad_norm": 0.46345117688179016,
"learning_rate": 5.967341786057782e-05,
"loss": 0.0893,
"step": 3730
},
{
"epoch": 0.8932409839980894,
"grad_norm": 0.39527013897895813,
"learning_rate": 5.947055704064215e-05,
"loss": 0.0984,
"step": 3740
},
{
"epoch": 0.8956293288750896,
"grad_norm": 0.44176241755485535,
"learning_rate": 5.92675343950168e-05,
"loss": 0.0849,
"step": 3750
},
{
"epoch": 0.8980176737520899,
"grad_norm": 0.4174538552761078,
"learning_rate": 5.906435339279864e-05,
"loss": 0.0829,
"step": 3760
},
{
"epoch": 0.9004060186290901,
"grad_norm": 0.21322879195213318,
"learning_rate": 5.886101750579041e-05,
"loss": 0.0723,
"step": 3770
},
{
"epoch": 0.9027943635060903,
"grad_norm": 0.32623594999313354,
"learning_rate": 5.865753020844136e-05,
"loss": 0.0835,
"step": 3780
},
{
"epoch": 0.9051827083830906,
"grad_norm": 0.5261030793190002,
"learning_rate": 5.8453894977788003e-05,
"loss": 0.0821,
"step": 3790
},
{
"epoch": 0.9075710532600908,
"grad_norm": 0.4309581518173218,
"learning_rate": 5.8250115293394555e-05,
"loss": 0.0879,
"step": 3800
},
{
"epoch": 0.909959398137091,
"grad_norm": 0.667290210723877,
"learning_rate": 5.8046194637293593e-05,
"loss": 0.0899,
"step": 3810
},
{
"epoch": 0.9123477430140913,
"grad_norm": 0.386171817779541,
"learning_rate": 5.784213649392649e-05,
"loss": 0.0865,
"step": 3820
},
{
"epoch": 0.9147360878910915,
"grad_norm": 0.2908395230770111,
"learning_rate": 5.7637944350083894e-05,
"loss": 0.0756,
"step": 3830
},
{
"epoch": 0.9171244327680917,
"grad_norm": 0.3564385175704956,
"learning_rate": 5.743362169484616e-05,
"loss": 0.0816,
"step": 3840
},
{
"epoch": 0.919512777645092,
"grad_norm": 0.519212007522583,
"learning_rate": 5.722917201952371e-05,
"loss": 0.0768,
"step": 3850
},
{
"epoch": 0.9219011225220922,
"grad_norm": 0.47325587272644043,
"learning_rate": 5.702459881759743e-05,
"loss": 0.0952,
"step": 3860
},
{
"epoch": 0.9242894673990925,
"grad_norm": 0.3965644836425781,
"learning_rate": 5.681990558465887e-05,
"loss": 0.089,
"step": 3870
},
{
"epoch": 0.9266778122760927,
"grad_norm": 0.4902348518371582,
"learning_rate": 5.661509581835063e-05,
"loss": 0.0844,
"step": 3880
},
{
"epoch": 0.9290661571530929,
"grad_norm": 0.5155518651008606,
"learning_rate": 5.6410173018306534e-05,
"loss": 0.0914,
"step": 3890
},
{
"epoch": 0.9314545020300932,
"grad_norm": 0.4669664800167084,
"learning_rate": 5.620514068609182e-05,
"loss": 0.0735,
"step": 3900
},
{
"epoch": 0.9338428469070934,
"grad_norm": 0.5426563620567322,
"learning_rate": 5.600000232514333e-05,
"loss": 0.0874,
"step": 3910
},
{
"epoch": 0.9362311917840936,
"grad_norm": 0.5035495162010193,
"learning_rate": 5.57947614407097e-05,
"loss": 0.0811,
"step": 3920
},
{
"epoch": 0.9386195366610939,
"grad_norm": 0.5714557766914368,
"learning_rate": 5.558942153979132e-05,
"loss": 0.0847,
"step": 3930
},
{
"epoch": 0.9410078815380941,
"grad_norm": 0.5045674443244934,
"learning_rate": 5.538398613108057e-05,
"loss": 0.0994,
"step": 3940
},
{
"epoch": 0.9433962264150944,
"grad_norm": 0.7179984450340271,
"learning_rate": 5.517845872490176e-05,
"loss": 0.0907,
"step": 3950
},
{
"epoch": 0.9457845712920946,
"grad_norm": 0.37904471158981323,
"learning_rate": 5.4972842833151193e-05,
"loss": 0.0962,
"step": 3960
},
{
"epoch": 0.9481729161690948,
"grad_norm": 0.39448070526123047,
"learning_rate": 5.4767141969237154e-05,
"loss": 0.0823,
"step": 3970
},
{
"epoch": 0.9505612610460951,
"grad_norm": 0.5061440467834473,
"learning_rate": 5.456135964801985e-05,
"loss": 0.0845,
"step": 3980
},
{
"epoch": 0.9529496059230953,
"grad_norm": 0.49993205070495605,
"learning_rate": 5.435549938575138e-05,
"loss": 0.0788,
"step": 3990
},
{
"epoch": 0.9553379508000955,
"grad_norm": 0.46544066071510315,
"learning_rate": 5.414956470001564e-05,
"loss": 0.0802,
"step": 4000
},
{
"epoch": 0.9577262956770958,
"grad_norm": 0.40973523259162903,
"learning_rate": 5.394355910966822e-05,
"loss": 0.0803,
"step": 4010
},
{
"epoch": 0.960114640554096,
"grad_norm": 0.3154264986515045,
"learning_rate": 5.373748613477627e-05,
"loss": 0.0736,
"step": 4020
},
{
"epoch": 0.9625029854310962,
"grad_norm": 0.35841289162635803,
"learning_rate": 5.3531349296558345e-05,
"loss": 0.0761,
"step": 4030
},
{
"epoch": 0.9648913303080965,
"grad_norm": 0.44680920243263245,
"learning_rate": 5.332515211732426e-05,
"loss": 0.0891,
"step": 4040
},
{
"epoch": 0.9672796751850967,
"grad_norm": 0.6256673336029053,
"learning_rate": 5.311889812041491e-05,
"loss": 0.0877,
"step": 4050
},
{
"epoch": 0.969668020062097,
"grad_norm": 0.45096686482429504,
"learning_rate": 5.2912590830142e-05,
"loss": 0.0897,
"step": 4060
},
{
"epoch": 0.9720563649390972,
"grad_norm": 0.395324170589447,
"learning_rate": 5.270623377172792e-05,
"loss": 0.0887,
"step": 4070
},
{
"epoch": 0.9744447098160974,
"grad_norm": 0.5654628276824951,
"learning_rate": 5.249983047124543e-05,
"loss": 0.091,
"step": 4080
},
{
"epoch": 0.9768330546930977,
"grad_norm": 0.41339489817619324,
"learning_rate": 5.229338445555746e-05,
"loss": 0.0881,
"step": 4090
},
{
"epoch": 0.9792213995700979,
"grad_norm": 0.6517373323440552,
"learning_rate": 5.20868992522568e-05,
"loss": 0.0766,
"step": 4100
},
{
"epoch": 0.9816097444470981,
"grad_norm": 0.5136675238609314,
"learning_rate": 5.1880378389605874e-05,
"loss": 0.0669,
"step": 4110
},
{
"epoch": 0.9839980893240984,
"grad_norm": 0.319338858127594,
"learning_rate": 5.167382539647642e-05,
"loss": 0.0785,
"step": 4120
},
{
"epoch": 0.9863864342010986,
"grad_norm": 0.5537667870521545,
"learning_rate": 5.146724380228917e-05,
"loss": 0.0819,
"step": 4130
},
{
"epoch": 0.9887747790780989,
"grad_norm": 0.7587404251098633,
"learning_rate": 5.1260637136953596e-05,
"loss": 0.0862,
"step": 4140
},
{
"epoch": 0.9911631239550991,
"grad_norm": 0.3112753629684448,
"learning_rate": 5.105400893080759e-05,
"loss": 0.0751,
"step": 4150
},
{
"epoch": 0.9935514688320993,
"grad_norm": 0.45201295614242554,
"learning_rate": 5.0847362714557055e-05,
"loss": 0.065,
"step": 4160
},
{
"epoch": 0.9959398137090996,
"grad_norm": 0.27288514375686646,
"learning_rate": 5.064070201921568e-05,
"loss": 0.0662,
"step": 4170
},
{
"epoch": 0.9983281585860998,
"grad_norm": 0.47334882616996765,
"learning_rate": 5.043403037604455e-05,
"loss": 0.0765,
"step": 4180
},
{
"epoch": 1.0007165034631,
"grad_norm": 0.337606817483902,
"learning_rate": 5.022735131649182e-05,
"loss": 0.0924,
"step": 4190
},
{
"epoch": 1.0031048483401004,
"grad_norm": 0.4341793358325958,
"learning_rate": 5.0020668372132385e-05,
"loss": 0.0727,
"step": 4200
},
{
"epoch": 1.0054931932171005,
"grad_norm": 0.438784658908844,
"learning_rate": 4.9813985074607496e-05,
"loss": 0.0739,
"step": 4210
},
{
"epoch": 1.0078815380941009,
"grad_norm": 0.40135127305984497,
"learning_rate": 4.960730495556446e-05,
"loss": 0.0701,
"step": 4220
},
{
"epoch": 1.010269882971101,
"grad_norm": 0.38818687200546265,
"learning_rate": 4.9400631546596216e-05,
"loss": 0.0682,
"step": 4230
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.28933069109916687,
"learning_rate": 4.919396837918116e-05,
"loss": 0.0736,
"step": 4240
},
{
"epoch": 1.0150465727251015,
"grad_norm": 0.40420788526535034,
"learning_rate": 4.898731898462257e-05,
"loss": 0.0831,
"step": 4250
},
{
"epoch": 1.0174349176021018,
"grad_norm": 0.393276572227478,
"learning_rate": 4.8780686893988456e-05,
"loss": 0.0669,
"step": 4260
},
{
"epoch": 1.019823262479102,
"grad_norm": 0.4949725568294525,
"learning_rate": 4.857407563805111e-05,
"loss": 0.0813,
"step": 4270
},
{
"epoch": 1.0222116073561023,
"grad_norm": 0.6876609921455383,
"learning_rate": 4.836748874722688e-05,
"loss": 0.0803,
"step": 4280
},
{
"epoch": 1.0245999522331024,
"grad_norm": 0.486724853515625,
"learning_rate": 4.816092975151568e-05,
"loss": 0.0954,
"step": 4290
},
{
"epoch": 1.0269882971101028,
"grad_norm": 0.524359941482544,
"learning_rate": 4.795440218044088e-05,
"loss": 0.0773,
"step": 4300
},
{
"epoch": 1.0293766419871029,
"grad_norm": 0.4168134927749634,
"learning_rate": 4.77479095629888e-05,
"loss": 0.0871,
"step": 4310
},
{
"epoch": 1.0317649868641032,
"grad_norm": 0.5238415598869324,
"learning_rate": 4.754145542754859e-05,
"loss": 0.0796,
"step": 4320
},
{
"epoch": 1.0341533317411034,
"grad_norm": 0.5205885767936707,
"learning_rate": 4.7335043301851735e-05,
"loss": 0.0718,
"step": 4330
},
{
"epoch": 1.0365416766181037,
"grad_norm": 0.6914841532707214,
"learning_rate": 4.7128676712912e-05,
"loss": 0.0761,
"step": 4340
},
{
"epoch": 1.0389300214951038,
"grad_norm": 0.30167150497436523,
"learning_rate": 4.6922359186964984e-05,
"loss": 0.066,
"step": 4350
},
{
"epoch": 1.0413183663721042,
"grad_norm": 0.4861651360988617,
"learning_rate": 4.671609424940797e-05,
"loss": 0.0838,
"step": 4360
},
{
"epoch": 1.0437067112491043,
"grad_norm": 0.4792875647544861,
"learning_rate": 4.650988542473964e-05,
"loss": 0.0796,
"step": 4370
},
{
"epoch": 1.0460950561261046,
"grad_norm": 0.36251622438430786,
"learning_rate": 4.6303736236499826e-05,
"loss": 0.0698,
"step": 4380
},
{
"epoch": 1.0484834010031048,
"grad_norm": 0.5330509543418884,
"learning_rate": 4.609765020720941e-05,
"loss": 0.0873,
"step": 4390
},
{
"epoch": 1.0508717458801051,
"grad_norm": 0.4140687882900238,
"learning_rate": 4.5891630858309986e-05,
"loss": 0.0675,
"step": 4400
},
{
"epoch": 1.0532600907571052,
"grad_norm": 0.448095440864563,
"learning_rate": 4.568568171010384e-05,
"loss": 0.0674,
"step": 4410
},
{
"epoch": 1.0556484356341056,
"grad_norm": 0.463425874710083,
"learning_rate": 4.5479806281693635e-05,
"loss": 0.0811,
"step": 4420
},
{
"epoch": 1.0580367805111057,
"grad_norm": 0.38941511511802673,
"learning_rate": 4.527400809092246e-05,
"loss": 0.0737,
"step": 4430
},
{
"epoch": 1.060425125388106,
"grad_norm": 0.42491796612739563,
"learning_rate": 4.506829065431353e-05,
"loss": 0.0785,
"step": 4440
},
{
"epoch": 1.0628134702651062,
"grad_norm": 0.3731493651866913,
"learning_rate": 4.4862657487010296e-05,
"loss": 0.0773,
"step": 4450
},
{
"epoch": 1.0652018151421065,
"grad_norm": 0.3927256464958191,
"learning_rate": 4.465711210271617e-05,
"loss": 0.0747,
"step": 4460
},
{
"epoch": 1.0675901600191067,
"grad_norm": 0.38888612389564514,
"learning_rate": 4.445165801363471e-05,
"loss": 0.0859,
"step": 4470
},
{
"epoch": 1.069978504896107,
"grad_norm": 0.4089820981025696,
"learning_rate": 4.424629873040937e-05,
"loss": 0.0695,
"step": 4480
},
{
"epoch": 1.0723668497731071,
"grad_norm": 0.36741238832473755,
"learning_rate": 4.4041037762063754e-05,
"loss": 0.0742,
"step": 4490
},
{
"epoch": 1.0747551946501075,
"grad_norm": 0.37752893567085266,
"learning_rate": 4.3835878615941446e-05,
"loss": 0.0923,
"step": 4500
},
{
"epoch": 1.0771435395271076,
"grad_norm": 0.3443499207496643,
"learning_rate": 4.363082479764618e-05,
"loss": 0.0656,
"step": 4510
},
{
"epoch": 1.079531884404108,
"grad_norm": 0.6358171105384827,
"learning_rate": 4.342587981098199e-05,
"loss": 0.0848,
"step": 4520
},
{
"epoch": 1.081920229281108,
"grad_norm": 0.477409303188324,
"learning_rate": 4.3221047157893196e-05,
"loss": 0.0869,
"step": 4530
},
{
"epoch": 1.0843085741581084,
"grad_norm": 0.3898080587387085,
"learning_rate": 4.301633033840473e-05,
"loss": 0.0711,
"step": 4540
},
{
"epoch": 1.0866969190351086,
"grad_norm": 0.45350539684295654,
"learning_rate": 4.281173285056215e-05,
"loss": 0.0716,
"step": 4550
},
{
"epoch": 1.089085263912109,
"grad_norm": 0.3930779695510864,
"learning_rate": 4.2607258190372084e-05,
"loss": 0.0897,
"step": 4560
},
{
"epoch": 1.091473608789109,
"grad_norm": 0.36369219422340393,
"learning_rate": 4.2402909851742245e-05,
"loss": 0.0751,
"step": 4570
},
{
"epoch": 1.0938619536661094,
"grad_norm": 0.5176813006401062,
"learning_rate": 4.219869132642198e-05,
"loss": 0.0756,
"step": 4580
},
{
"epoch": 1.0962502985431097,
"grad_norm": 0.5521029233932495,
"learning_rate": 4.1994606103942383e-05,
"loss": 0.0757,
"step": 4590
},
{
"epoch": 1.0986386434201099,
"grad_norm": 0.47750312089920044,
"learning_rate": 4.179065767155686e-05,
"loss": 0.0645,
"step": 4600
},
{
"epoch": 1.10102698829711,
"grad_norm": 0.3444133400917053,
"learning_rate": 4.1586849514181356e-05,
"loss": 0.0921,
"step": 4610
},
{
"epoch": 1.1034153331741103,
"grad_norm": 0.45913535356521606,
"learning_rate": 4.138318511433501e-05,
"loss": 0.0714,
"step": 4620
},
{
"epoch": 1.1058036780511107,
"grad_norm": 0.31060707569122314,
"learning_rate": 4.117966795208046e-05,
"loss": 0.0678,
"step": 4630
},
{
"epoch": 1.1081920229281108,
"grad_norm": 0.5543773174285889,
"learning_rate": 4.097630150496446e-05,
"loss": 0.0832,
"step": 4640
},
{
"epoch": 1.1105803678051112,
"grad_norm": 0.7926197648048401,
"learning_rate": 4.077308924795853e-05,
"loss": 0.0919,
"step": 4650
},
{
"epoch": 1.1129687126821113,
"grad_norm": 0.3399224281311035,
"learning_rate": 4.057003465339942e-05,
"loss": 0.0776,
"step": 4660
},
{
"epoch": 1.1153570575591116,
"grad_norm": 0.3181399405002594,
"learning_rate": 4.036714119092995e-05,
"loss": 0.0826,
"step": 4670
},
{
"epoch": 1.1177454024361118,
"grad_norm": 0.47512686252593994,
"learning_rate": 4.016441232743953e-05,
"loss": 0.077,
"step": 4680
},
{
"epoch": 1.120133747313112,
"grad_norm": 0.4550682306289673,
"learning_rate": 3.996185152700516e-05,
"loss": 0.0818,
"step": 4690
},
{
"epoch": 1.1225220921901122,
"grad_norm": 0.42231303453445435,
"learning_rate": 3.975946225083197e-05,
"loss": 0.0899,
"step": 4700
},
{
"epoch": 1.1249104370671126,
"grad_norm": 0.38459691405296326,
"learning_rate": 3.9557247957194335e-05,
"loss": 0.0819,
"step": 4710
},
{
"epoch": 1.1272987819441127,
"grad_norm": 0.4329834282398224,
"learning_rate": 3.9355212101376554e-05,
"loss": 0.0684,
"step": 4720
},
{
"epoch": 1.129687126821113,
"grad_norm": 0.3824212849140167,
"learning_rate": 3.9153358135614016e-05,
"loss": 0.0732,
"step": 4730
},
{
"epoch": 1.1320754716981132,
"grad_norm": 0.291305810213089,
"learning_rate": 3.895168950903401e-05,
"loss": 0.0801,
"step": 4740
},
{
"epoch": 1.1344638165751135,
"grad_norm": 0.34754371643066406,
"learning_rate": 3.875020966759699e-05,
"loss": 0.0601,
"step": 4750
},
{
"epoch": 1.1368521614521137,
"grad_norm": 0.37193042039871216,
"learning_rate": 3.854892205403752e-05,
"loss": 0.0699,
"step": 4760
},
{
"epoch": 1.139240506329114,
"grad_norm": 0.38301366567611694,
"learning_rate": 3.8347830107805543e-05,
"loss": 0.0674,
"step": 4770
},
{
"epoch": 1.1416288512061141,
"grad_norm": 0.48410362005233765,
"learning_rate": 3.81469372650076e-05,
"loss": 0.09,
"step": 4780
},
{
"epoch": 1.1440171960831145,
"grad_norm": 0.4602961242198944,
"learning_rate": 3.794624695834808e-05,
"loss": 0.0779,
"step": 4790
},
{
"epoch": 1.1464055409601146,
"grad_norm": 0.42936572432518005,
"learning_rate": 3.77457626170706e-05,
"loss": 0.0716,
"step": 4800
},
{
"epoch": 1.148793885837115,
"grad_norm": 0.3211369216442108,
"learning_rate": 3.754548766689941e-05,
"loss": 0.0629,
"step": 4810
},
{
"epoch": 1.151182230714115,
"grad_norm": 0.4017835259437561,
"learning_rate": 3.734542552998082e-05,
"loss": 0.0635,
"step": 4820
},
{
"epoch": 1.1535705755911154,
"grad_norm": 0.32386353611946106,
"learning_rate": 3.7145579624824765e-05,
"loss": 0.0879,
"step": 4830
},
{
"epoch": 1.1559589204681155,
"grad_norm": 0.2893616259098053,
"learning_rate": 3.694595336624638e-05,
"loss": 0.0712,
"step": 4840
},
{
"epoch": 1.158347265345116,
"grad_norm": 0.3600180745124817,
"learning_rate": 3.6746550165307615e-05,
"loss": 0.0681,
"step": 4850
},
{
"epoch": 1.160735610222116,
"grad_norm": 0.4287347197532654,
"learning_rate": 3.654737342925901e-05,
"loss": 0.0747,
"step": 4860
},
{
"epoch": 1.1631239550991164,
"grad_norm": 0.39198926091194153,
"learning_rate": 3.634842656148144e-05,
"loss": 0.0785,
"step": 4870
},
{
"epoch": 1.1655122999761165,
"grad_norm": 0.5814246535301208,
"learning_rate": 3.6149712961427976e-05,
"loss": 0.0688,
"step": 4880
},
{
"epoch": 1.1679006448531168,
"grad_norm": 0.30154722929000854,
"learning_rate": 3.5951236024565745e-05,
"loss": 0.0764,
"step": 4890
},
{
"epoch": 1.170288989730117,
"grad_norm": 0.24106436967849731,
"learning_rate": 3.5752999142318e-05,
"loss": 0.0738,
"step": 4900
},
{
"epoch": 1.1726773346071173,
"grad_norm": 0.32780545949935913,
"learning_rate": 3.555500570200611e-05,
"loss": 0.0772,
"step": 4910
},
{
"epoch": 1.1750656794841174,
"grad_norm": 0.4114924669265747,
"learning_rate": 3.535725908679165e-05,
"loss": 0.0879,
"step": 4920
},
{
"epoch": 1.1774540243611178,
"grad_norm": 0.537018358707428,
"learning_rate": 3.515976267561873e-05,
"loss": 0.0623,
"step": 4930
},
{
"epoch": 1.179842369238118,
"grad_norm": 0.541746199131012,
"learning_rate": 3.496251984315606e-05,
"loss": 0.0793,
"step": 4940
},
{
"epoch": 1.1822307141151183,
"grad_norm": 0.35712435841560364,
"learning_rate": 3.476553395973946e-05,
"loss": 0.0708,
"step": 4950
},
{
"epoch": 1.1846190589921184,
"grad_norm": 0.4787076711654663,
"learning_rate": 3.456880839131415e-05,
"loss": 0.0674,
"step": 4960
},
{
"epoch": 1.1870074038691187,
"grad_norm": 0.40953099727630615,
"learning_rate": 3.437234649937735e-05,
"loss": 0.0732,
"step": 4970
},
{
"epoch": 1.1893957487461189,
"grad_norm": 0.4978320896625519,
"learning_rate": 3.417615164092069e-05,
"loss": 0.0592,
"step": 4980
},
{
"epoch": 1.1917840936231192,
"grad_norm": 0.5183027386665344,
"learning_rate": 3.3980227168373024e-05,
"loss": 0.0696,
"step": 4990
},
{
"epoch": 1.1941724385001193,
"grad_norm": 0.3381482660770416,
"learning_rate": 3.3784576429542974e-05,
"loss": 0.0754,
"step": 5000
},
{
"epoch": 1.1965607833771197,
"grad_norm": 0.3869054317474365,
"learning_rate": 3.358920276756189e-05,
"loss": 0.0759,
"step": 5010
},
{
"epoch": 1.1989491282541198,
"grad_norm": 0.3766176700592041,
"learning_rate": 3.339410952082656e-05,
"loss": 0.084,
"step": 5020
},
{
"epoch": 1.2013374731311202,
"grad_norm": 0.3348686695098877,
"learning_rate": 3.319930002294232e-05,
"loss": 0.0708,
"step": 5030
},
{
"epoch": 1.2037258180081203,
"grad_norm": 0.48407381772994995,
"learning_rate": 3.3004777602665985e-05,
"loss": 0.0689,
"step": 5040
},
{
"epoch": 1.2061141628851206,
"grad_norm": 0.4639897048473358,
"learning_rate": 3.2810545583848986e-05,
"loss": 0.0693,
"step": 5050
},
{
"epoch": 1.2085025077621208,
"grad_norm": 0.49152323603630066,
"learning_rate": 3.261660728538066e-05,
"loss": 0.068,
"step": 5060
},
{
"epoch": 1.210890852639121,
"grad_norm": 0.35651206970214844,
"learning_rate": 3.2422966021131397e-05,
"loss": 0.0763,
"step": 5070
},
{
"epoch": 1.2132791975161212,
"grad_norm": 0.3952496647834778,
"learning_rate": 3.222962509989616e-05,
"loss": 0.078,
"step": 5080
},
{
"epoch": 1.2156675423931216,
"grad_norm": 0.4173029661178589,
"learning_rate": 3.203658782533781e-05,
"loss": 0.0805,
"step": 5090
},
{
"epoch": 1.218055887270122,
"grad_norm": 0.3484851121902466,
"learning_rate": 3.184385749593082e-05,
"loss": 0.0802,
"step": 5100
},
{
"epoch": 1.220444232147122,
"grad_norm": 0.36740779876708984,
"learning_rate": 3.1651437404904694e-05,
"loss": 0.0724,
"step": 5110
},
{
"epoch": 1.2228325770241222,
"grad_norm": 0.4029271602630615,
"learning_rate": 3.145933084018794e-05,
"loss": 0.0695,
"step": 5120
},
{
"epoch": 1.2252209219011225,
"grad_norm": 0.26390138268470764,
"learning_rate": 3.126754108435165e-05,
"loss": 0.0657,
"step": 5130
},
{
"epoch": 1.2276092667781229,
"grad_norm": 0.34551748633384705,
"learning_rate": 3.1076071414553645e-05,
"loss": 0.078,
"step": 5140
},
{
"epoch": 1.229997611655123,
"grad_norm": 0.3767643868923187,
"learning_rate": 3.088492510248224e-05,
"loss": 0.073,
"step": 5150
},
{
"epoch": 1.2323859565321231,
"grad_norm": 0.4803922772407532,
"learning_rate": 3.069410541430058e-05,
"loss": 0.0713,
"step": 5160
},
{
"epoch": 1.2347743014091235,
"grad_norm": 0.5446774363517761,
"learning_rate": 3.0503615610590603e-05,
"loss": 0.0744,
"step": 5170
},
{
"epoch": 1.2371626462861238,
"grad_norm": 0.4428233504295349,
"learning_rate": 3.031345894629747e-05,
"loss": 0.0779,
"step": 5180
},
{
"epoch": 1.239550991163124,
"grad_norm": 0.40515047311782837,
"learning_rate": 3.0123638670673947e-05,
"loss": 0.0781,
"step": 5190
},
{
"epoch": 1.241939336040124,
"grad_norm": 0.4483841359615326,
"learning_rate": 2.9934158027224785e-05,
"loss": 0.0701,
"step": 5200
},
{
"epoch": 1.2443276809171244,
"grad_norm": 0.4725356698036194,
"learning_rate": 2.974502025365141e-05,
"loss": 0.0777,
"step": 5210
},
{
"epoch": 1.2467160257941248,
"grad_norm": 0.44211339950561523,
"learning_rate": 2.9556228581796492e-05,
"loss": 0.0631,
"step": 5220
},
{
"epoch": 1.249104370671125,
"grad_norm": 0.42972704768180847,
"learning_rate": 2.9367786237588847e-05,
"loss": 0.078,
"step": 5230
},
{
"epoch": 1.251492715548125,
"grad_norm": 0.4280390739440918,
"learning_rate": 2.9179696440988158e-05,
"loss": 0.0714,
"step": 5240
},
{
"epoch": 1.2538810604251254,
"grad_norm": 0.39674246311187744,
"learning_rate": 2.8991962405930144e-05,
"loss": 0.0653,
"step": 5250
},
{
"epoch": 1.2562694053021257,
"grad_norm": 0.4163968861103058,
"learning_rate": 2.880458734027145e-05,
"loss": 0.0756,
"step": 5260
},
{
"epoch": 1.2586577501791258,
"grad_norm": 0.3326818645000458,
"learning_rate": 2.8617574445734996e-05,
"loss": 0.0641,
"step": 5270
},
{
"epoch": 1.261046095056126,
"grad_norm": 0.4225046932697296,
"learning_rate": 2.8430926917855127e-05,
"loss": 0.0716,
"step": 5280
},
{
"epoch": 1.2634344399331263,
"grad_norm": 0.21198680996894836,
"learning_rate": 2.8244647945923152e-05,
"loss": 0.0621,
"step": 5290
},
{
"epoch": 1.2658227848101267,
"grad_norm": 0.30797064304351807,
"learning_rate": 2.8058740712932706e-05,
"loss": 0.0737,
"step": 5300
},
{
"epoch": 1.2682111296871268,
"grad_norm": 0.275452584028244,
"learning_rate": 2.787320839552551e-05,
"loss": 0.0778,
"step": 5310
},
{
"epoch": 1.2705994745641271,
"grad_norm": 0.5148422718048096,
"learning_rate": 2.7688054163936926e-05,
"loss": 0.0799,
"step": 5320
},
{
"epoch": 1.2729878194411273,
"grad_norm": 0.3826289474964142,
"learning_rate": 2.7503281181941914e-05,
"loss": 0.0694,
"step": 5330
},
{
"epoch": 1.2753761643181276,
"grad_norm": 0.2690454125404358,
"learning_rate": 2.731889260680096e-05,
"loss": 0.0724,
"step": 5340
},
{
"epoch": 1.2777645091951277,
"grad_norm": 0.3726663589477539,
"learning_rate": 2.7134891589206034e-05,
"loss": 0.0787,
"step": 5350
},
{
"epoch": 1.280152854072128,
"grad_norm": 0.37403494119644165,
"learning_rate": 2.695128127322689e-05,
"loss": 0.0824,
"step": 5360
},
{
"epoch": 1.2825411989491282,
"grad_norm": 0.4260822534561157,
"learning_rate": 2.6768064796257185e-05,
"loss": 0.0641,
"step": 5370
},
{
"epoch": 1.2849295438261286,
"grad_norm": 0.3740355372428894,
"learning_rate": 2.658524528896105e-05,
"loss": 0.0799,
"step": 5380
},
{
"epoch": 1.2873178887031287,
"grad_norm": 0.3999808430671692,
"learning_rate": 2.640282587521939e-05,
"loss": 0.0765,
"step": 5390
},
{
"epoch": 1.289706233580129,
"grad_norm": 0.5166588425636292,
"learning_rate": 2.622080967207673e-05,
"loss": 0.0795,
"step": 5400
},
{
"epoch": 1.2920945784571292,
"grad_norm": 0.24830542504787445,
"learning_rate": 2.603919978968773e-05,
"loss": 0.08,
"step": 5410
},
{
"epoch": 1.2944829233341295,
"grad_norm": 0.2548556625843048,
"learning_rate": 2.5857999331264248e-05,
"loss": 0.0704,
"step": 5420
},
{
"epoch": 1.2968712682111296,
"grad_norm": 0.4048694372177124,
"learning_rate": 2.5677211393022138e-05,
"loss": 0.0566,
"step": 5430
},
{
"epoch": 1.29925961308813,
"grad_norm": 0.4749496579170227,
"learning_rate": 2.5496839064128474e-05,
"loss": 0.0797,
"step": 5440
},
{
"epoch": 1.30164795796513,
"grad_norm": 0.3442319333553314,
"learning_rate": 2.5316885426648697e-05,
"loss": 0.0765,
"step": 5450
},
{
"epoch": 1.3040363028421305,
"grad_norm": 0.39732515811920166,
"learning_rate": 2.5137353555493913e-05,
"loss": 0.0688,
"step": 5460
},
{
"epoch": 1.3064246477191306,
"grad_norm": 0.5200353860855103,
"learning_rate": 2.495824651836852e-05,
"loss": 0.0736,
"step": 5470
},
{
"epoch": 1.308812992596131,
"grad_norm": 0.466342031955719,
"learning_rate": 2.4779567375717556e-05,
"loss": 0.0717,
"step": 5480
},
{
"epoch": 1.311201337473131,
"grad_norm": 0.33231985569000244,
"learning_rate": 2.460131918067462e-05,
"loss": 0.071,
"step": 5490
},
{
"epoch": 1.3135896823501314,
"grad_norm": 0.6182300448417664,
"learning_rate": 2.4423504979009525e-05,
"loss": 0.0749,
"step": 5500
},
{
"epoch": 1.3159780272271315,
"grad_norm": 0.326463520526886,
"learning_rate": 2.4246127809076407e-05,
"loss": 0.074,
"step": 5510
},
{
"epoch": 1.3183663721041319,
"grad_norm": 0.3448762595653534,
"learning_rate": 2.4069190701761663e-05,
"loss": 0.0754,
"step": 5520
},
{
"epoch": 1.320754716981132,
"grad_norm": 0.5516566634178162,
"learning_rate": 2.38926966804323e-05,
"loss": 0.0753,
"step": 5530
},
{
"epoch": 1.3231430618581324,
"grad_norm": 0.2736930251121521,
"learning_rate": 2.3716648760884157e-05,
"loss": 0.0766,
"step": 5540
},
{
"epoch": 1.3255314067351325,
"grad_norm": 0.34717538952827454,
"learning_rate": 2.3541049951290476e-05,
"loss": 0.0735,
"step": 5550
},
{
"epoch": 1.3279197516121328,
"grad_norm": 0.5675294399261475,
"learning_rate": 2.336590325215038e-05,
"loss": 0.0823,
"step": 5560
},
{
"epoch": 1.330308096489133,
"grad_norm": 0.3436681628227234,
"learning_rate": 2.319121165623775e-05,
"loss": 0.0717,
"step": 5570
},
{
"epoch": 1.3326964413661333,
"grad_norm": 0.30631816387176514,
"learning_rate": 2.3016978148549955e-05,
"loss": 0.0492,
"step": 5580
},
{
"epoch": 1.3350847862431334,
"grad_norm": 0.369118869304657,
"learning_rate": 2.2843205706256887e-05,
"loss": 0.0635,
"step": 5590
},
{
"epoch": 1.3374731311201338,
"grad_norm": 0.5049929022789001,
"learning_rate": 2.2669897298650176e-05,
"loss": 0.0789,
"step": 5600
},
{
"epoch": 1.3398614759971341,
"grad_norm": 0.4120734930038452,
"learning_rate": 2.24970558870923e-05,
"loss": 0.0772,
"step": 5610
},
{
"epoch": 1.3422498208741342,
"grad_norm": 0.34382253885269165,
"learning_rate": 2.232468442496612e-05,
"loss": 0.0772,
"step": 5620
},
{
"epoch": 1.3446381657511344,
"grad_norm": 0.3989104628562927,
"learning_rate": 2.215278585762431e-05,
"loss": 0.0511,
"step": 5630
},
{
"epoch": 1.3470265106281347,
"grad_norm": 0.4745878279209137,
"learning_rate": 2.1981363122339128e-05,
"loss": 0.0739,
"step": 5640
},
{
"epoch": 1.349414855505135,
"grad_norm": 0.2686302065849304,
"learning_rate": 2.181041914825212e-05,
"loss": 0.0674,
"step": 5650
},
{
"epoch": 1.3518032003821352,
"grad_norm": 0.24377219378948212,
"learning_rate": 2.163995685632419e-05,
"loss": 0.0657,
"step": 5660
},
{
"epoch": 1.3541915452591353,
"grad_norm": 0.32141876220703125,
"learning_rate": 2.1469979159285547e-05,
"loss": 0.0545,
"step": 5670
},
{
"epoch": 1.3565798901361357,
"grad_norm": 0.3667854070663452,
"learning_rate": 2.130048896158608e-05,
"loss": 0.0822,
"step": 5680
},
{
"epoch": 1.358968235013136,
"grad_norm": 0.5508107542991638,
"learning_rate": 2.1131489159345602e-05,
"loss": 0.0594,
"step": 5690
},
{
"epoch": 1.3613565798901361,
"grad_norm": 0.43387120962142944,
"learning_rate": 2.0962982640304473e-05,
"loss": 0.0776,
"step": 5700
},
{
"epoch": 1.3637449247671363,
"grad_norm": 0.29643702507019043,
"learning_rate": 2.0794972283774167e-05,
"loss": 0.0686,
"step": 5710
},
{
"epoch": 1.3661332696441366,
"grad_norm": 0.3128342926502228,
"learning_rate": 2.06274609605881e-05,
"loss": 0.0632,
"step": 5720
},
{
"epoch": 1.368521614521137,
"grad_norm": 0.36617615818977356,
"learning_rate": 2.0460451533052644e-05,
"loss": 0.0743,
"step": 5730
},
{
"epoch": 1.370909959398137,
"grad_norm": 0.33750998973846436,
"learning_rate": 2.0293946854898077e-05,
"loss": 0.0715,
"step": 5740
},
{
"epoch": 1.3732983042751372,
"grad_norm": 0.38130059838294983,
"learning_rate": 2.012794977123e-05,
"loss": 0.0788,
"step": 5750
},
{
"epoch": 1.3756866491521376,
"grad_norm": 0.2846888601779938,
"learning_rate": 1.9962463118480512e-05,
"loss": 0.0628,
"step": 5760
},
{
"epoch": 1.378074994029138,
"grad_norm": 0.48023509979248047,
"learning_rate": 1.9797489724359976e-05,
"loss": 0.0556,
"step": 5770
},
{
"epoch": 1.380463338906138,
"grad_norm": 0.41425007581710815,
"learning_rate": 1.9633032407808462e-05,
"loss": 0.0681,
"step": 5780
},
{
"epoch": 1.3828516837831382,
"grad_norm": 0.35490521788597107,
"learning_rate": 1.9469093978947806e-05,
"loss": 0.0763,
"step": 5790
},
{
"epoch": 1.3852400286601385,
"grad_norm": 0.4281676113605499,
"learning_rate": 1.9305677239033393e-05,
"loss": 0.0639,
"step": 5800
},
{
"epoch": 1.3876283735371389,
"grad_norm": 0.36021822690963745,
"learning_rate": 1.914278498040647e-05,
"loss": 0.0861,
"step": 5810
},
{
"epoch": 1.390016718414139,
"grad_norm": 0.3075598180294037,
"learning_rate": 1.8980419986446263e-05,
"loss": 0.0752,
"step": 5820
},
{
"epoch": 1.3924050632911391,
"grad_norm": 0.34387725591659546,
"learning_rate": 1.8818585031522567e-05,
"loss": 0.0724,
"step": 5830
},
{
"epoch": 1.3947934081681395,
"grad_norm": 0.2783162295818329,
"learning_rate": 1.8657282880948217e-05,
"loss": 0.0731,
"step": 5840
},
{
"epoch": 1.3971817530451398,
"grad_norm": 0.26851415634155273,
"learning_rate": 1.8496516290931953e-05,
"loss": 0.0655,
"step": 5850
},
{
"epoch": 1.39957009792214,
"grad_norm": 0.4195689260959625,
"learning_rate": 1.8336288008531204e-05,
"loss": 0.0615,
"step": 5860
},
{
"epoch": 1.40195844279914,
"grad_norm": 0.39157816767692566,
"learning_rate": 1.8176600771605217e-05,
"loss": 0.0653,
"step": 5870
},
{
"epoch": 1.4043467876761404,
"grad_norm": 0.3706957995891571,
"learning_rate": 1.801745730876832e-05,
"loss": 0.0689,
"step": 5880
},
{
"epoch": 1.4067351325531408,
"grad_norm": 0.24662348628044128,
"learning_rate": 1.7858860339343157e-05,
"loss": 0.0657,
"step": 5890
},
{
"epoch": 1.4091234774301409,
"grad_norm": 0.5005415678024292,
"learning_rate": 1.7700812573314397e-05,
"loss": 0.0657,
"step": 5900
},
{
"epoch": 1.4115118223071412,
"grad_norm": 0.3899209499359131,
"learning_rate": 1.754331671128225e-05,
"loss": 0.0755,
"step": 5910
},
{
"epoch": 1.4139001671841414,
"grad_norm": 0.41633015871047974,
"learning_rate": 1.7386375444416485e-05,
"loss": 0.0665,
"step": 5920
},
{
"epoch": 1.4162885120611417,
"grad_norm": 0.4023812413215637,
"learning_rate": 1.722999145441031e-05,
"loss": 0.0717,
"step": 5930
},
{
"epoch": 1.4186768569381418,
"grad_norm": 0.39149409532546997,
"learning_rate": 1.7074167413434667e-05,
"loss": 0.0797,
"step": 5940
},
{
"epoch": 1.4210652018151422,
"grad_norm": 0.346272736787796,
"learning_rate": 1.6918905984092448e-05,
"loss": 0.0568,
"step": 5950
},
{
"epoch": 1.4234535466921423,
"grad_norm": 0.32263851165771484,
"learning_rate": 1.6764209819373133e-05,
"loss": 0.066,
"step": 5960
},
{
"epoch": 1.4258418915691426,
"grad_norm": 0.38288140296936035,
"learning_rate": 1.661008156260732e-05,
"loss": 0.0584,
"step": 5970
},
{
"epoch": 1.4282302364461428,
"grad_norm": 0.5386690497398376,
"learning_rate": 1.6456523847421705e-05,
"loss": 0.0643,
"step": 5980
},
{
"epoch": 1.4306185813231431,
"grad_norm": 0.3323499262332916,
"learning_rate": 1.630353929769392e-05,
"loss": 0.057,
"step": 5990
},
{
"epoch": 1.4330069262001432,
"grad_norm": 0.2828684151172638,
"learning_rate": 1.6151130527507858e-05,
"loss": 0.0734,
"step": 6000
},
{
"epoch": 1.4353952710771436,
"grad_norm": 0.40432727336883545,
"learning_rate": 1.599930014110886e-05,
"loss": 0.0612,
"step": 6010
},
{
"epoch": 1.4377836159541437,
"grad_norm": 0.25748762488365173,
"learning_rate": 1.5848050732859338e-05,
"loss": 0.0667,
"step": 6020
},
{
"epoch": 1.440171960831144,
"grad_norm": 0.247622549533844,
"learning_rate": 1.569738488719434e-05,
"loss": 0.0668,
"step": 6030
},
{
"epoch": 1.4425603057081442,
"grad_norm": 0.3375413417816162,
"learning_rate": 1.5547305178577503e-05,
"loss": 0.0724,
"step": 6040
},
{
"epoch": 1.4449486505851445,
"grad_norm": 0.4188424348831177,
"learning_rate": 1.5397814171456933e-05,
"loss": 0.0625,
"step": 6050
},
{
"epoch": 1.4473369954621447,
"grad_norm": 0.29855257272720337,
"learning_rate": 1.5248914420221506e-05,
"loss": 0.0679,
"step": 6060
},
{
"epoch": 1.449725340339145,
"grad_norm": 0.2769811153411865,
"learning_rate": 1.5100608469157119e-05,
"loss": 0.0609,
"step": 6070
},
{
"epoch": 1.4521136852161451,
"grad_norm": 0.4417251646518707,
"learning_rate": 1.4952898852403307e-05,
"loss": 0.0639,
"step": 6080
},
{
"epoch": 1.4545020300931455,
"grad_norm": 0.37149304151535034,
"learning_rate": 1.4805788093909855e-05,
"loss": 0.0665,
"step": 6090
},
{
"epoch": 1.4568903749701456,
"grad_norm": 0.48007506132125854,
"learning_rate": 1.4659278707393742e-05,
"loss": 0.0737,
"step": 6100
},
{
"epoch": 1.459278719847146,
"grad_norm": 0.3375127911567688,
"learning_rate": 1.451337319629617e-05,
"loss": 0.0822,
"step": 6110
},
{
"epoch": 1.461667064724146,
"grad_norm": 0.28468725085258484,
"learning_rate": 1.4368074053739734e-05,
"loss": 0.0687,
"step": 6120
},
{
"epoch": 1.4640554096011464,
"grad_norm": 0.6998528242111206,
"learning_rate": 1.4223383762485881e-05,
"loss": 0.0793,
"step": 6130
},
{
"epoch": 1.4664437544781466,
"grad_norm": 0.35145285725593567,
"learning_rate": 1.4079304794892501e-05,
"loss": 0.0604,
"step": 6140
},
{
"epoch": 1.468832099355147,
"grad_norm": 0.33022791147232056,
"learning_rate": 1.3935839612871588e-05,
"loss": 0.0665,
"step": 6150
},
{
"epoch": 1.471220444232147,
"grad_norm": 0.376506507396698,
"learning_rate": 1.3792990667847299e-05,
"loss": 0.0662,
"step": 6160
},
{
"epoch": 1.4736087891091474,
"grad_norm": 0.41991305351257324,
"learning_rate": 1.3650760400713952e-05,
"loss": 0.0662,
"step": 6170
},
{
"epoch": 1.4759971339861475,
"grad_norm": 0.34855782985687256,
"learning_rate": 1.3509151241794405e-05,
"loss": 0.0732,
"step": 6180
},
{
"epoch": 1.4783854788631479,
"grad_norm": 0.24274849891662598,
"learning_rate": 1.3368165610798444e-05,
"loss": 0.0608,
"step": 6190
},
{
"epoch": 1.4807738237401482,
"grad_norm": 0.48918581008911133,
"learning_rate": 1.3227805916781537e-05,
"loss": 0.0794,
"step": 6200
},
{
"epoch": 1.4831621686171483,
"grad_norm": 0.6149623990058899,
"learning_rate": 1.3088074558103563e-05,
"loss": 0.0788,
"step": 6210
},
{
"epoch": 1.4855505134941485,
"grad_norm": 0.3137202560901642,
"learning_rate": 1.2948973922387941e-05,
"loss": 0.0767,
"step": 6220
},
{
"epoch": 1.4879388583711488,
"grad_norm": 0.4283827543258667,
"learning_rate": 1.2810506386480714e-05,
"loss": 0.0684,
"step": 6230
},
{
"epoch": 1.4903272032481492,
"grad_norm": 0.2887345552444458,
"learning_rate": 1.2672674316410049e-05,
"loss": 0.0654,
"step": 6240
},
{
"epoch": 1.4927155481251493,
"grad_norm": 0.28072062134742737,
"learning_rate": 1.2535480067345712e-05,
"loss": 0.0716,
"step": 6250
},
{
"epoch": 1.4951038930021494,
"grad_norm": 0.42105478048324585,
"learning_rate": 1.2398925983558906e-05,
"loss": 0.0786,
"step": 6260
},
{
"epoch": 1.4974922378791498,
"grad_norm": 0.44790470600128174,
"learning_rate": 1.226301439838215e-05,
"loss": 0.0751,
"step": 6270
},
{
"epoch": 1.49988058275615,
"grad_norm": 0.23907090723514557,
"learning_rate": 1.2127747634169428e-05,
"loss": 0.0606,
"step": 6280
},
{
"epoch": 1.5022689276331502,
"grad_norm": 0.3297266662120819,
"learning_rate": 1.1993128002256554e-05,
"loss": 0.0756,
"step": 6290
},
{
"epoch": 1.5046572725101504,
"grad_norm": 0.3529978394508362,
"learning_rate": 1.1859157802921578e-05,
"loss": 0.0665,
"step": 6300
},
{
"epoch": 1.5070456173871507,
"grad_norm": 0.27625182271003723,
"learning_rate": 1.17258393253456e-05,
"loss": 0.0608,
"step": 6310
},
{
"epoch": 1.509433962264151,
"grad_norm": 0.31399810314178467,
"learning_rate": 1.1593174847573535e-05,
"loss": 0.0661,
"step": 6320
},
{
"epoch": 1.5118223071411512,
"grad_norm": 0.6096377372741699,
"learning_rate": 1.1461166636475302e-05,
"loss": 0.0645,
"step": 6330
},
{
"epoch": 1.5142106520181513,
"grad_norm": 0.4462642967700958,
"learning_rate": 1.1329816947706984e-05,
"loss": 0.0722,
"step": 6340
},
{
"epoch": 1.5165989968951517,
"grad_norm": 0.4132387638092041,
"learning_rate": 1.1199128025672379e-05,
"loss": 0.0574,
"step": 6350
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.3920154571533203,
"learning_rate": 1.1069102103484547e-05,
"loss": 0.0627,
"step": 6360
},
{
"epoch": 1.5213756866491521,
"grad_norm": 0.5844139456748962,
"learning_rate": 1.0939741402927783e-05,
"loss": 0.0694,
"step": 6370
},
{
"epoch": 1.5237640315261523,
"grad_norm": 0.29464778304100037,
"learning_rate": 1.081104813441951e-05,
"loss": 0.0674,
"step": 6380
},
{
"epoch": 1.5261523764031526,
"grad_norm": 0.28407859802246094,
"learning_rate": 1.0683024496972638e-05,
"loss": 0.0796,
"step": 6390
},
{
"epoch": 1.528540721280153,
"grad_norm": 0.3596804141998291,
"learning_rate": 1.0555672678157886e-05,
"loss": 0.0747,
"step": 6400
},
{
"epoch": 1.530929066157153,
"grad_norm": 0.32208487391471863,
"learning_rate": 1.0428994854066448e-05,
"loss": 0.0682,
"step": 6410
},
{
"epoch": 1.5333174110341532,
"grad_norm": 0.4079877436161041,
"learning_rate": 1.0302993189272853e-05,
"loss": 0.0648,
"step": 6420
},
{
"epoch": 1.5357057559111535,
"grad_norm": 0.2763961851596832,
"learning_rate": 1.0177669836797865e-05,
"loss": 0.066,
"step": 6430
},
{
"epoch": 1.538094100788154,
"grad_norm": 0.3884141445159912,
"learning_rate": 1.0053026938071841e-05,
"loss": 0.0615,
"step": 6440
},
{
"epoch": 1.540482445665154,
"grad_norm": 0.38694334030151367,
"learning_rate": 9.929066622897982e-06,
"loss": 0.0742,
"step": 6450
},
{
"epoch": 1.5428707905421541,
"grad_norm": 0.3485255539417267,
"learning_rate": 9.805791009416081e-06,
"loss": 0.0553,
"step": 6460
},
{
"epoch": 1.5452591354191545,
"grad_norm": 0.38298311829566956,
"learning_rate": 9.683202204066222e-06,
"loss": 0.0581,
"step": 6470
},
{
"epoch": 1.5476474802961548,
"grad_norm": 0.45903831720352173,
"learning_rate": 9.561302301552877e-06,
"loss": 0.0667,
"step": 6480
},
{
"epoch": 1.550035825173155,
"grad_norm": 0.21236850321292877,
"learning_rate": 9.440093384809028e-06,
"loss": 0.0678,
"step": 6490
},
{
"epoch": 1.552424170050155,
"grad_norm": 0.3039874732494354,
"learning_rate": 9.319577524960655e-06,
"loss": 0.0575,
"step": 6500
},
{
"epoch": 1.5548125149271554,
"grad_norm": 0.4687940776348114,
"learning_rate": 9.199756781291264e-06,
"loss": 0.0794,
"step": 6510
},
{
"epoch": 1.5572008598041558,
"grad_norm": 0.4368513524532318,
"learning_rate": 9.080633201206789e-06,
"loss": 0.0621,
"step": 6520
},
{
"epoch": 1.559589204681156,
"grad_norm": 0.44934943318367004,
"learning_rate": 8.962208820200535e-06,
"loss": 0.0547,
"step": 6530
},
{
"epoch": 1.561977549558156,
"grad_norm": 0.33911368250846863,
"learning_rate": 8.844485661818408e-06,
"loss": 0.069,
"step": 6540
},
{
"epoch": 1.5643658944351564,
"grad_norm": 0.28564295172691345,
"learning_rate": 8.727465737624414e-06,
"loss": 0.067,
"step": 6550
},
{
"epoch": 1.5667542393121567,
"grad_norm": 0.3013613522052765,
"learning_rate": 8.61115104716615e-06,
"loss": 0.0535,
"step": 6560
},
{
"epoch": 1.5691425841891569,
"grad_norm": 0.2956767678260803,
"learning_rate": 8.495543577940784e-06,
"loss": 0.0676,
"step": 6570
},
{
"epoch": 1.571530929066157,
"grad_norm": 0.22847913205623627,
"learning_rate": 8.380645305360968e-06,
"loss": 0.0628,
"step": 6580
},
{
"epoch": 1.5739192739431573,
"grad_norm": 0.388623982667923,
"learning_rate": 8.266458192721193e-06,
"loss": 0.062,
"step": 6590
},
{
"epoch": 1.5763076188201577,
"grad_norm": 0.31546780467033386,
"learning_rate": 8.152984191164131e-06,
"loss": 0.056,
"step": 6600
},
{
"epoch": 1.5786959636971578,
"grad_norm": 0.35583969950675964,
"learning_rate": 8.040225239647408e-06,
"loss": 0.0637,
"step": 6610
},
{
"epoch": 1.5810843085741582,
"grad_norm": 0.4171866178512573,
"learning_rate": 7.92818326491037e-06,
"loss": 0.0676,
"step": 6620
},
{
"epoch": 1.5834726534511585,
"grad_norm": 0.37667766213417053,
"learning_rate": 7.816860181441254e-06,
"loss": 0.0614,
"step": 6630
},
{
"epoch": 1.5858609983281586,
"grad_norm": 0.3363400399684906,
"learning_rate": 7.70625789144439e-06,
"loss": 0.0576,
"step": 6640
},
{
"epoch": 1.5882493432051588,
"grad_norm": 0.358175665140152,
"learning_rate": 7.596378284807765e-06,
"loss": 0.0573,
"step": 6650
},
{
"epoch": 1.590637688082159,
"grad_norm": 0.41003715991973877,
"learning_rate": 7.487223239070678e-06,
"loss": 0.0728,
"step": 6660
},
{
"epoch": 1.5930260329591595,
"grad_norm": 0.3572424650192261,
"learning_rate": 7.378794619391682e-06,
"loss": 0.0733,
"step": 6670
},
{
"epoch": 1.5954143778361596,
"grad_norm": 0.47185713052749634,
"learning_rate": 7.271094278516749e-06,
"loss": 0.0645,
"step": 6680
},
{
"epoch": 1.5978027227131597,
"grad_norm": 0.2734459340572357,
"learning_rate": 7.164124056747523e-06,
"loss": 0.054,
"step": 6690
},
{
"epoch": 1.60019106759016,
"grad_norm": 0.3619970679283142,
"learning_rate": 7.057885781909973e-06,
"loss": 0.0805,
"step": 6700
},
{
"epoch": 1.6025794124671604,
"grad_norm": 0.22864112257957458,
"learning_rate": 6.952381269323077e-06,
"loss": 0.0664,
"step": 6710
},
{
"epoch": 1.6049677573441605,
"grad_norm": 0.3344721794128418,
"learning_rate": 6.847612321767882e-06,
"loss": 0.0647,
"step": 6720
},
{
"epoch": 1.6073561022211607,
"grad_norm": 0.3794196844100952,
"learning_rate": 6.743580729456606e-06,
"loss": 0.0844,
"step": 6730
},
{
"epoch": 1.609744447098161,
"grad_norm": 0.40472012758255005,
"learning_rate": 6.640288270002154e-06,
"loss": 0.0733,
"step": 6740
},
{
"epoch": 1.6121327919751613,
"grad_norm": 0.36237967014312744,
"learning_rate": 6.537736708387638e-06,
"loss": 0.0664,
"step": 6750
},
{
"epoch": 1.6145211368521615,
"grad_norm": 0.376791775226593,
"learning_rate": 6.435927796936331e-06,
"loss": 0.076,
"step": 6760
},
{
"epoch": 1.6169094817291616,
"grad_norm": 0.35781851410865784,
"learning_rate": 6.3348632752815935e-06,
"loss": 0.0672,
"step": 6770
},
{
"epoch": 1.619297826606162,
"grad_norm": 0.34404456615448,
"learning_rate": 6.234544870337283e-06,
"loss": 0.0652,
"step": 6780
},
{
"epoch": 1.6216861714831623,
"grad_norm": 0.4969412684440613,
"learning_rate": 6.134974296268125e-06,
"loss": 0.0557,
"step": 6790
},
{
"epoch": 1.6240745163601624,
"grad_norm": 0.28869327902793884,
"learning_rate": 6.036153254460536e-06,
"loss": 0.0834,
"step": 6800
},
{
"epoch": 1.6264628612371625,
"grad_norm": 0.4533311724662781,
"learning_rate": 5.938083433493446e-06,
"loss": 0.0655,
"step": 6810
},
{
"epoch": 1.628851206114163,
"grad_norm": 0.40970295667648315,
"learning_rate": 5.840766509109507e-06,
"loss": 0.0552,
"step": 6820
},
{
"epoch": 1.6312395509911632,
"grad_norm": 0.3142627477645874,
"learning_rate": 5.744204144186466e-06,
"loss": 0.07,
"step": 6830
},
{
"epoch": 1.6336278958681634,
"grad_norm": 0.37267547845840454,
"learning_rate": 5.648397988708698e-06,
"loss": 0.0658,
"step": 6840
},
{
"epoch": 1.6360162407451635,
"grad_norm": 0.4293549060821533,
"learning_rate": 5.553349679739078e-06,
"loss": 0.0569,
"step": 6850
},
{
"epoch": 1.6384045856221638,
"grad_norm": 0.5119047164916992,
"learning_rate": 5.459060841390939e-06,
"loss": 0.065,
"step": 6860
},
{
"epoch": 1.6407929304991642,
"grad_norm": 0.312116414308548,
"learning_rate": 5.3655330848004e-06,
"loss": 0.0636,
"step": 6870
},
{
"epoch": 1.6431812753761643,
"grad_norm": 0.49624183773994446,
"learning_rate": 5.272768008098749e-06,
"loss": 0.067,
"step": 6880
},
{
"epoch": 1.6455696202531644,
"grad_norm": 0.4097059369087219,
"learning_rate": 5.180767196385217e-06,
"loss": 0.0568,
"step": 6890
},
{
"epoch": 1.6479579651301648,
"grad_norm": 0.4349365532398224,
"learning_rate": 5.089532221699822e-06,
"loss": 0.068,
"step": 6900
},
{
"epoch": 1.6503463100071651,
"grad_norm": 0.41091611981391907,
"learning_rate": 4.99906464299657e-06,
"loss": 0.0646,
"step": 6910
},
{
"epoch": 1.6527346548841653,
"grad_norm": 0.32905831933021545,
"learning_rate": 4.9093660061167555e-06,
"loss": 0.0731,
"step": 6920
},
{
"epoch": 1.6551229997611654,
"grad_norm": 0.4123990535736084,
"learning_rate": 4.82043784376261e-06,
"loss": 0.069,
"step": 6930
},
{
"epoch": 1.6575113446381657,
"grad_norm": 0.3229024112224579,
"learning_rate": 4.732281675471056e-06,
"loss": 0.0611,
"step": 6940
},
{
"epoch": 1.659899689515166,
"grad_norm": 0.3599684238433838,
"learning_rate": 4.644899007587794e-06,
"loss": 0.058,
"step": 6950
},
{
"epoch": 1.6622880343921662,
"grad_norm": 0.3050236701965332,
"learning_rate": 4.558291333241504e-06,
"loss": 0.0668,
"step": 6960
},
{
"epoch": 1.6646763792691663,
"grad_norm": 0.30929887294769287,
"learning_rate": 4.472460132318396e-06,
"loss": 0.0644,
"step": 6970
},
{
"epoch": 1.6670647241461667,
"grad_norm": 0.42303964495658875,
"learning_rate": 4.387406871436861e-06,
"loss": 0.0665,
"step": 6980
},
{
"epoch": 1.669453069023167,
"grad_norm": 0.5272735357284546,
"learning_rate": 4.303133003922477e-06,
"loss": 0.0755,
"step": 6990
},
{
"epoch": 1.6718414139001672,
"grad_norm": 0.4046907126903534,
"learning_rate": 4.219639969783107e-06,
"loss": 0.0682,
"step": 7000
},
{
"epoch": 1.6742297587771673,
"grad_norm": 0.32538893818855286,
"learning_rate": 4.136929195684347e-06,
"loss": 0.0675,
"step": 7010
},
{
"epoch": 1.6766181036541676,
"grad_norm": 0.2995414733886719,
"learning_rate": 4.0550020949251085e-06,
"loss": 0.0699,
"step": 7020
},
{
"epoch": 1.679006448531168,
"grad_norm": 0.33142927289009094,
"learning_rate": 3.97386006741351e-06,
"loss": 0.0614,
"step": 7030
},
{
"epoch": 1.681394793408168,
"grad_norm": 0.49354231357574463,
"learning_rate": 3.893504499642914e-06,
"loss": 0.0654,
"step": 7040
},
{
"epoch": 1.6837831382851682,
"grad_norm": 0.33486661314964294,
"learning_rate": 3.8139367646682742e-06,
"loss": 0.0747,
"step": 7050
},
{
"epoch": 1.6861714831621686,
"grad_norm": 0.43350932002067566,
"learning_rate": 3.7351582220826464e-06,
"loss": 0.0667,
"step": 7060
},
{
"epoch": 1.688559828039169,
"grad_norm": 0.31854572892189026,
"learning_rate": 3.6571702179939603e-06,
"loss": 0.0605,
"step": 7070
},
{
"epoch": 1.690948172916169,
"grad_norm": 0.386692076921463,
"learning_rate": 3.5799740850020303e-06,
"loss": 0.062,
"step": 7080
},
{
"epoch": 1.6933365177931692,
"grad_norm": 0.38695356249809265,
"learning_rate": 3.503571142175782e-06,
"loss": 0.0599,
"step": 7090
},
{
"epoch": 1.6957248626701695,
"grad_norm": 0.338236927986145,
"learning_rate": 3.427962695030695e-06,
"loss": 0.0738,
"step": 7100
},
{
"epoch": 1.6981132075471699,
"grad_norm": 0.254668265581131,
"learning_rate": 3.353150035506536e-06,
"loss": 0.0714,
"step": 7110
},
{
"epoch": 1.70050155242417,
"grad_norm": 0.3446265459060669,
"learning_rate": 3.2791344419452252e-06,
"loss": 0.0599,
"step": 7120
},
{
"epoch": 1.7028898973011701,
"grad_norm": 0.3183712959289551,
"learning_rate": 3.2059171790690567e-06,
"loss": 0.0618,
"step": 7130
},
{
"epoch": 1.7052782421781705,
"grad_norm": 0.41936060786247253,
"learning_rate": 3.1334994979590195e-06,
"loss": 0.0776,
"step": 7140
},
{
"epoch": 1.7076665870551708,
"grad_norm": 0.41712450981140137,
"learning_rate": 3.061882636033497e-06,
"loss": 0.0703,
"step": 7150
},
{
"epoch": 1.710054931932171,
"grad_norm": 0.2925345003604889,
"learning_rate": 2.991067817027049e-06,
"loss": 0.0697,
"step": 7160
},
{
"epoch": 1.7124432768091713,
"grad_norm": 0.43560653924942017,
"learning_rate": 2.921056250969556e-06,
"loss": 0.0654,
"step": 7170
},
{
"epoch": 1.7148316216861716,
"grad_norm": 0.30326196551322937,
"learning_rate": 2.8518491341655074e-06,
"loss": 0.0637,
"step": 7180
},
{
"epoch": 1.7172199665631718,
"grad_norm": 0.44300079345703125,
"learning_rate": 2.783447649173582e-06,
"loss": 0.0605,
"step": 7190
},
{
"epoch": 1.719608311440172,
"grad_norm": 0.35670390725135803,
"learning_rate": 2.7158529647864305e-06,
"loss": 0.0637,
"step": 7200
},
{
"epoch": 1.7219966563171722,
"grad_norm": 0.38413602113723755,
"learning_rate": 2.6490662360107168e-06,
"loss": 0.0582,
"step": 7210
},
{
"epoch": 1.7243850011941726,
"grad_norm": 0.4491639733314514,
"learning_rate": 2.583088604047357e-06,
"loss": 0.0665,
"step": 7220
},
{
"epoch": 1.7267733460711727,
"grad_norm": 0.34005463123321533,
"learning_rate": 2.5179211962720385e-06,
"loss": 0.0628,
"step": 7230
},
{
"epoch": 1.7291616909481728,
"grad_norm": 0.36418479681015015,
"learning_rate": 2.453565126215973e-06,
"loss": 0.0678,
"step": 7240
},
{
"epoch": 1.7315500358251732,
"grad_norm": 0.3217214345932007,
"learning_rate": 2.390021493546818e-06,
"loss": 0.0738,
"step": 7250
},
{
"epoch": 1.7339383807021735,
"grad_norm": 0.2504056990146637,
"learning_rate": 2.32729138404994e-06,
"loss": 0.0582,
"step": 7260
},
{
"epoch": 1.7363267255791737,
"grad_norm": 0.342367947101593,
"learning_rate": 2.265375869609826e-06,
"loss": 0.0706,
"step": 7270
},
{
"epoch": 1.7387150704561738,
"grad_norm": 0.5572667717933655,
"learning_rate": 2.204276008191802e-06,
"loss": 0.0665,
"step": 7280
},
{
"epoch": 1.7411034153331741,
"grad_norm": 0.3529682755470276,
"learning_rate": 2.1439928438238997e-06,
"loss": 0.0645,
"step": 7290
},
{
"epoch": 1.7434917602101745,
"grad_norm": 0.34254124760627747,
"learning_rate": 2.0845274065790845e-06,
"loss": 0.0725,
"step": 7300
},
{
"epoch": 1.7458801050871746,
"grad_norm": 0.3985847234725952,
"learning_rate": 2.025880712557604e-06,
"loss": 0.0576,
"step": 7310
},
{
"epoch": 1.7482684499641747,
"grad_norm": 0.3425021469593048,
"learning_rate": 1.9680537638696482e-06,
"loss": 0.0602,
"step": 7320
},
{
"epoch": 1.750656794841175,
"grad_norm": 0.34825029969215393,
"learning_rate": 1.9110475486182224e-06,
"loss": 0.0588,
"step": 7330
},
{
"epoch": 1.7530451397181754,
"grad_norm": 0.47028207778930664,
"learning_rate": 1.8548630408822543e-06,
"loss": 0.0671,
"step": 7340
},
{
"epoch": 1.7554334845951756,
"grad_norm": 0.2827095091342926,
"learning_rate": 1.799501200699971e-06,
"loss": 0.0643,
"step": 7350
},
{
"epoch": 1.7578218294721757,
"grad_norm": 0.2834967076778412,
"learning_rate": 1.7449629740524554e-06,
"loss": 0.0684,
"step": 7360
},
{
"epoch": 1.760210174349176,
"grad_norm": 0.30589234828948975,
"learning_rate": 1.6912492928475465e-06,
"loss": 0.0715,
"step": 7370
},
{
"epoch": 1.7625985192261764,
"grad_norm": 0.37452325224876404,
"learning_rate": 1.6383610749038435e-06,
"loss": 0.064,
"step": 7380
},
{
"epoch": 1.7649868641031765,
"grad_norm": 0.3154360055923462,
"learning_rate": 1.5862992239350672e-06,
"loss": 0.0589,
"step": 7390
},
{
"epoch": 1.7673752089801766,
"grad_norm": 0.29880574345588684,
"learning_rate": 1.5350646295346104e-06,
"loss": 0.0528,
"step": 7400
},
{
"epoch": 1.769763553857177,
"grad_norm": 0.32569319009780884,
"learning_rate": 1.48465816716033e-06,
"loss": 0.0677,
"step": 7410
},
{
"epoch": 1.7721518987341773,
"grad_norm": 0.33918216824531555,
"learning_rate": 1.4350806981195897e-06,
"loss": 0.0873,
"step": 7420
},
{
"epoch": 1.7745402436111775,
"grad_norm": 0.5083154439926147,
"learning_rate": 1.3863330695545518e-06,
"loss": 0.0687,
"step": 7430
},
{
"epoch": 1.7769285884881776,
"grad_norm": 0.320112407207489,
"learning_rate": 1.3384161144276764e-06,
"loss": 0.064,
"step": 7440
},
{
"epoch": 1.779316933365178,
"grad_norm": 0.41778478026390076,
"learning_rate": 1.291330651507533e-06,
"loss": 0.0622,
"step": 7450
},
{
"epoch": 1.7817052782421783,
"grad_norm": 0.32113003730773926,
"learning_rate": 1.245077485354762e-06,
"loss": 0.0502,
"step": 7460
},
{
"epoch": 1.7840936231191784,
"grad_norm": 0.29655033349990845,
"learning_rate": 1.1996574063083687e-06,
"loss": 0.07,
"step": 7470
},
{
"epoch": 1.7864819679961785,
"grad_norm": 0.39924848079681396,
"learning_rate": 1.1550711904721678e-06,
"loss": 0.0621,
"step": 7480
},
{
"epoch": 1.7888703128731789,
"grad_norm": 0.2844867408275604,
"learning_rate": 1.1113195997015767e-06,
"loss": 0.0601,
"step": 7490
},
{
"epoch": 1.7912586577501792,
"grad_norm": 0.23533496260643005,
"learning_rate": 1.0684033815905714e-06,
"loss": 0.0706,
"step": 7500
},
{
"epoch": 1.7936470026271794,
"grad_norm": 0.29888856410980225,
"learning_rate": 1.0263232694589064e-06,
"loss": 0.0606,
"step": 7510
},
{
"epoch": 1.7960353475041795,
"grad_norm": 0.2885245084762573,
"learning_rate": 9.850799823395985e-07,
"loss": 0.0655,
"step": 7520
},
{
"epoch": 1.7984236923811798,
"grad_norm": 0.3960576057434082,
"learning_rate": 9.446742249666241e-07,
"loss": 0.0689,
"step": 7530
},
{
"epoch": 1.8008120372581802,
"grad_norm": 0.3243632912635803,
"learning_rate": 9.051066877628967e-07,
"loss": 0.0605,
"step": 7540
},
{
"epoch": 1.8032003821351803,
"grad_norm": 0.21384885907173157,
"learning_rate": 8.663780468284532e-07,
"loss": 0.069,
"step": 7550
},
{
"epoch": 1.8055887270121804,
"grad_norm": 0.28082549571990967,
"learning_rate": 8.28488963928925e-07,
"loss": 0.0654,
"step": 7560
},
{
"epoch": 1.8079770718891808,
"grad_norm": 0.8143417239189148,
"learning_rate": 7.914400864841798e-07,
"loss": 0.0634,
"step": 7570
},
{
"epoch": 1.8103654167661811,
"grad_norm": 0.6125345826148987,
"learning_rate": 7.552320475573249e-07,
"loss": 0.072,
"step": 7580
},
{
"epoch": 1.8127537616431812,
"grad_norm": 0.33566775918006897,
"learning_rate": 7.198654658438386e-07,
"loss": 0.0693,
"step": 7590
},
{
"epoch": 1.8151421065201814,
"grad_norm": 0.3584268093109131,
"learning_rate": 6.853409456610227e-07,
"loss": 0.0682,
"step": 7600
},
{
"epoch": 1.8175304513971817,
"grad_norm": 0.3100437521934509,
"learning_rate": 6.516590769376774e-07,
"loss": 0.0584,
"step": 7610
},
{
"epoch": 1.819918796274182,
"grad_norm": 0.3123069405555725,
"learning_rate": 6.188204352039984e-07,
"loss": 0.0555,
"step": 7620
},
{
"epoch": 1.8223071411511822,
"grad_norm": 0.2791978120803833,
"learning_rate": 5.868255815817735e-07,
"loss": 0.0672,
"step": 7630
},
{
"epoch": 1.8246954860281823,
"grad_norm": 0.22785606980323792,
"learning_rate": 5.556750627747743e-07,
"loss": 0.07,
"step": 7640
},
{
"epoch": 1.8270838309051827,
"grad_norm": 0.3727695941925049,
"learning_rate": 5.253694110594176e-07,
"loss": 0.0688,
"step": 7650
},
{
"epoch": 1.829472175782183,
"grad_norm": 0.3077755272388458,
"learning_rate": 4.959091442756747e-07,
"loss": 0.0594,
"step": 7660
},
{
"epoch": 1.8318605206591831,
"grad_norm": 0.33861351013183594,
"learning_rate": 4.6729476581822697e-07,
"loss": 0.0693,
"step": 7670
},
{
"epoch": 1.8342488655361833,
"grad_norm": 0.3712623417377472,
"learning_rate": 4.3952676462784e-07,
"loss": 0.0612,
"step": 7680
},
{
"epoch": 1.8366372104131836,
"grad_norm": 0.2320345789194107,
"learning_rate": 4.126056151830593e-07,
"loss": 0.056,
"step": 7690
},
{
"epoch": 1.839025555290184,
"grad_norm": 0.233370840549469,
"learning_rate": 3.8653177749203296e-07,
"loss": 0.062,
"step": 7700
},
{
"epoch": 1.841413900167184,
"grad_norm": 0.33854004740715027,
"learning_rate": 3.6130569708471283e-07,
"loss": 0.0604,
"step": 7710
},
{
"epoch": 1.8438022450441842,
"grad_norm": 0.4767954647541046,
"learning_rate": 3.3692780500521605e-07,
"loss": 0.0761,
"step": 7720
},
{
"epoch": 1.8461905899211846,
"grad_norm": 0.3321741819381714,
"learning_rate": 3.133985178044529e-07,
"loss": 0.0627,
"step": 7730
},
{
"epoch": 1.848578934798185,
"grad_norm": 0.516524612903595,
"learning_rate": 2.9071823753302196e-07,
"loss": 0.0575,
"step": 7740
},
{
"epoch": 1.850967279675185,
"grad_norm": 0.4143713414669037,
"learning_rate": 2.688873517343371e-07,
"loss": 0.0623,
"step": 7750
},
{
"epoch": 1.8533556245521854,
"grad_norm": 0.333374947309494,
"learning_rate": 2.479062334380111e-07,
"loss": 0.0593,
"step": 7760
},
{
"epoch": 1.8557439694291857,
"grad_norm": 0.2602499723434448,
"learning_rate": 2.277752411534495e-07,
"loss": 0.0552,
"step": 7770
},
{
"epoch": 1.8581323143061859,
"grad_norm": 0.3373238146305084,
"learning_rate": 2.0849471886378313e-07,
"loss": 0.0727,
"step": 7780
},
{
"epoch": 1.860520659183186,
"grad_norm": 0.4108473062515259,
"learning_rate": 1.9006499601992855e-07,
"loss": 0.0711,
"step": 7790
},
{
"epoch": 1.8629090040601863,
"grad_norm": 0.32196056842803955,
"learning_rate": 1.7248638753500912e-07,
"loss": 0.0567,
"step": 7800
},
{
"epoch": 1.8652973489371867,
"grad_norm": 0.3691728711128235,
"learning_rate": 1.557591937789371e-07,
"loss": 0.0661,
"step": 7810
},
{
"epoch": 1.8676856938141868,
"grad_norm": 0.2999248206615448,
"learning_rate": 1.3988370057331225e-07,
"loss": 0.0619,
"step": 7820
},
{
"epoch": 1.870074038691187,
"grad_norm": 0.5292859673500061,
"learning_rate": 1.2486017918649784e-07,
"loss": 0.0705,
"step": 7830
},
{
"epoch": 1.8724623835681873,
"grad_norm": 0.3337211608886719,
"learning_rate": 1.1068888632903007e-07,
"loss": 0.0606,
"step": 7840
},
{
"epoch": 1.8748507284451876,
"grad_norm": 0.3953368067741394,
"learning_rate": 9.737006414919925e-08,
"loss": 0.0643,
"step": 7850
},
{
"epoch": 1.8772390733221878,
"grad_norm": 0.5659822225570679,
"learning_rate": 8.490394022893645e-08,
"loss": 0.0653,
"step": 7860
},
{
"epoch": 1.8796274181991879,
"grad_norm": 0.24949753284454346,
"learning_rate": 7.329072757989996e-08,
"loss": 0.0663,
"step": 7870
},
{
"epoch": 1.8820157630761882,
"grad_norm": 0.3791586458683014,
"learning_rate": 6.25306246398505e-08,
"loss": 0.0705,
"step": 7880
},
{
"epoch": 1.8844041079531886,
"grad_norm": 0.3127134442329407,
"learning_rate": 5.26238152692593e-08,
"loss": 0.0639,
"step": 7890
},
{
"epoch": 1.8867924528301887,
"grad_norm": 0.2775420844554901,
"learning_rate": 4.357046874817194e-08,
"loss": 0.0675,
"step": 7900
},
{
"epoch": 1.8891807977071888,
"grad_norm": 0.3876698613166809,
"learning_rate": 3.5370739773288306e-08,
"loss": 0.0682,
"step": 7910
},
{
"epoch": 1.8915691425841892,
"grad_norm": 0.29213425517082214,
"learning_rate": 2.8024768455364724e-08,
"loss": 0.0671,
"step": 7920
},
{
"epoch": 1.8939574874611895,
"grad_norm": 0.34360945224761963,
"learning_rate": 2.153268031677702e-08,
"loss": 0.0644,
"step": 7930
},
{
"epoch": 1.8963458323381897,
"grad_norm": 0.36947551369667053,
"learning_rate": 1.5894586289399995e-08,
"loss": 0.0653,
"step": 7940
},
{
"epoch": 1.8987341772151898,
"grad_norm": 0.47420355677604675,
"learning_rate": 1.1110582712703377e-08,
"loss": 0.0631,
"step": 7950
},
{
"epoch": 1.9011225220921901,
"grad_norm": 0.2983955442905426,
"learning_rate": 7.180751332114266e-09,
"loss": 0.0737,
"step": 7960
},
{
"epoch": 1.9035108669691905,
"grad_norm": 0.4033881425857544,
"learning_rate": 4.1051592976015796e-09,
"loss": 0.0649,
"step": 7970
},
{
"epoch": 1.9058992118461906,
"grad_norm": 0.30320605635643005,
"learning_rate": 1.8838591625491843e-09,
"loss": 0.0628,
"step": 7980
},
{
"epoch": 1.9082875567231907,
"grad_norm": 0.2856868803501129,
"learning_rate": 5.168888828455121e-10,
"loss": 0.0613,
"step": 7990
},
{
"epoch": 1.910675901600191,
"grad_norm": 0.32585567235946655,
"learning_rate": 4.271816250733806e-12,
"loss": 0.0728,
"step": 8000
}
],
"logging_steps": 10,
"max_steps": 8000,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}