kurtpayne's picture
Auto fine-tune: 11648 examples (LoRA adapter)
acc7b75 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 2862,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0069881201956673656,
"grad_norm": 14.528067588806152,
"learning_rate": 1.99748427672956e-05,
"loss": 4.6827,
"step": 10
},
{
"epoch": 0.013976240391334731,
"grad_norm": 13.050820350646973,
"learning_rate": 1.994689028651293e-05,
"loss": 4.529,
"step": 20
},
{
"epoch": 0.020964360587002098,
"grad_norm": 2.765507698059082,
"learning_rate": 1.991893780573026e-05,
"loss": 3.4294,
"step": 30
},
{
"epoch": 0.027952480782669462,
"grad_norm": 16.95716667175293,
"learning_rate": 1.989098532494759e-05,
"loss": 3.352,
"step": 40
},
{
"epoch": 0.03494060097833683,
"grad_norm": 5.339346408843994,
"learning_rate": 1.986303284416492e-05,
"loss": 2.2654,
"step": 50
},
{
"epoch": 0.041928721174004195,
"grad_norm": 7.168309688568115,
"learning_rate": 1.983508036338225e-05,
"loss": 2.0076,
"step": 60
},
{
"epoch": 0.04891684136967156,
"grad_norm": 5.793034076690674,
"learning_rate": 1.9807127882599582e-05,
"loss": 1.6779,
"step": 70
},
{
"epoch": 0.055904961565338925,
"grad_norm": 6.093694686889648,
"learning_rate": 1.9779175401816913e-05,
"loss": 1.2564,
"step": 80
},
{
"epoch": 0.06289308176100629,
"grad_norm": 3.321807622909546,
"learning_rate": 1.9751222921034243e-05,
"loss": 1.4143,
"step": 90
},
{
"epoch": 0.06988120195667366,
"grad_norm": 2.9472098350524902,
"learning_rate": 1.9723270440251574e-05,
"loss": 1.0533,
"step": 100
},
{
"epoch": 0.07686932215234102,
"grad_norm": 8.515447616577148,
"learning_rate": 1.9695317959468904e-05,
"loss": 1.2895,
"step": 110
},
{
"epoch": 0.08385744234800839,
"grad_norm": 3.560924530029297,
"learning_rate": 1.9667365478686235e-05,
"loss": 1.097,
"step": 120
},
{
"epoch": 0.09084556254367575,
"grad_norm": 4.986349582672119,
"learning_rate": 1.9639412997903566e-05,
"loss": 0.9556,
"step": 130
},
{
"epoch": 0.09783368273934312,
"grad_norm": 1.27532160282135,
"learning_rate": 1.9611460517120896e-05,
"loss": 0.8306,
"step": 140
},
{
"epoch": 0.10482180293501048,
"grad_norm": 2.6796844005584717,
"learning_rate": 1.9583508036338227e-05,
"loss": 0.7747,
"step": 150
},
{
"epoch": 0.11180992313067785,
"grad_norm": 3.4592413902282715,
"learning_rate": 1.9555555555555557e-05,
"loss": 0.6187,
"step": 160
},
{
"epoch": 0.1187980433263452,
"grad_norm": 4.833241939544678,
"learning_rate": 1.9527603074772888e-05,
"loss": 0.6831,
"step": 170
},
{
"epoch": 0.12578616352201258,
"grad_norm": 1.7203434705734253,
"learning_rate": 1.949965059399022e-05,
"loss": 0.586,
"step": 180
},
{
"epoch": 0.13277428371767994,
"grad_norm": 1.333950161933899,
"learning_rate": 1.947169811320755e-05,
"loss": 0.6214,
"step": 190
},
{
"epoch": 0.13976240391334732,
"grad_norm": 1.5078257322311401,
"learning_rate": 1.944374563242488e-05,
"loss": 0.5652,
"step": 200
},
{
"epoch": 0.14675052410901468,
"grad_norm": 3.7645132541656494,
"learning_rate": 1.941579315164221e-05,
"loss": 0.5229,
"step": 210
},
{
"epoch": 0.15373864430468204,
"grad_norm": 1.2441000938415527,
"learning_rate": 1.938784067085954e-05,
"loss": 0.3669,
"step": 220
},
{
"epoch": 0.1607267645003494,
"grad_norm": 2.2875142097473145,
"learning_rate": 1.935988819007687e-05,
"loss": 0.4162,
"step": 230
},
{
"epoch": 0.16771488469601678,
"grad_norm": 2.1816585063934326,
"learning_rate": 1.9331935709294202e-05,
"loss": 0.3836,
"step": 240
},
{
"epoch": 0.17470300489168414,
"grad_norm": 3.4903249740600586,
"learning_rate": 1.9303983228511532e-05,
"loss": 0.4131,
"step": 250
},
{
"epoch": 0.1816911250873515,
"grad_norm": 2.1905159950256348,
"learning_rate": 1.9276030747728863e-05,
"loss": 0.3725,
"step": 260
},
{
"epoch": 0.18867924528301888,
"grad_norm": 3.6003665924072266,
"learning_rate": 1.9248078266946194e-05,
"loss": 0.4131,
"step": 270
},
{
"epoch": 0.19566736547868624,
"grad_norm": 3.012568235397339,
"learning_rate": 1.9220125786163524e-05,
"loss": 0.3991,
"step": 280
},
{
"epoch": 0.2026554856743536,
"grad_norm": 1.2049503326416016,
"learning_rate": 1.9192173305380855e-05,
"loss": 0.2732,
"step": 290
},
{
"epoch": 0.20964360587002095,
"grad_norm": 2.0943682193756104,
"learning_rate": 1.9164220824598185e-05,
"loss": 0.3052,
"step": 300
},
{
"epoch": 0.21663172606568834,
"grad_norm": 4.1970086097717285,
"learning_rate": 1.9136268343815516e-05,
"loss": 0.2544,
"step": 310
},
{
"epoch": 0.2236198462613557,
"grad_norm": 2.6826064586639404,
"learning_rate": 1.9108315863032846e-05,
"loss": 0.1921,
"step": 320
},
{
"epoch": 0.23060796645702306,
"grad_norm": 6.013642311096191,
"learning_rate": 1.9080363382250177e-05,
"loss": 0.3014,
"step": 330
},
{
"epoch": 0.2375960866526904,
"grad_norm": 4.2075629234313965,
"learning_rate": 1.9052410901467508e-05,
"loss": 0.2569,
"step": 340
},
{
"epoch": 0.2445842068483578,
"grad_norm": 0.5800023078918457,
"learning_rate": 1.9024458420684838e-05,
"loss": 0.1486,
"step": 350
},
{
"epoch": 0.25157232704402516,
"grad_norm": 8.123499870300293,
"learning_rate": 1.899650593990217e-05,
"loss": 0.2215,
"step": 360
},
{
"epoch": 0.2585604472396925,
"grad_norm": 4.439255237579346,
"learning_rate": 1.89685534591195e-05,
"loss": 0.1692,
"step": 370
},
{
"epoch": 0.2655485674353599,
"grad_norm": 4.177631855010986,
"learning_rate": 1.894060097833683e-05,
"loss": 0.1176,
"step": 380
},
{
"epoch": 0.27253668763102723,
"grad_norm": 0.5754140019416809,
"learning_rate": 1.891264849755416e-05,
"loss": 0.1538,
"step": 390
},
{
"epoch": 0.27952480782669464,
"grad_norm": 0.5612320303916931,
"learning_rate": 1.888469601677149e-05,
"loss": 0.1054,
"step": 400
},
{
"epoch": 0.286512928022362,
"grad_norm": 1.949729561805725,
"learning_rate": 1.8856743535988818e-05,
"loss": 0.0824,
"step": 410
},
{
"epoch": 0.29350104821802936,
"grad_norm": 3.153822898864746,
"learning_rate": 1.8828791055206152e-05,
"loss": 0.1174,
"step": 420
},
{
"epoch": 0.3004891684136967,
"grad_norm": 0.6545250415802002,
"learning_rate": 1.8800838574423483e-05,
"loss": 0.1305,
"step": 430
},
{
"epoch": 0.3074772886093641,
"grad_norm": 1.2621614933013916,
"learning_rate": 1.8772886093640813e-05,
"loss": 0.0684,
"step": 440
},
{
"epoch": 0.31446540880503143,
"grad_norm": 0.3028416335582733,
"learning_rate": 1.8744933612858144e-05,
"loss": 0.0863,
"step": 450
},
{
"epoch": 0.3214535290006988,
"grad_norm": 0.4452091455459595,
"learning_rate": 1.8716981132075474e-05,
"loss": 0.1016,
"step": 460
},
{
"epoch": 0.3284416491963662,
"grad_norm": 0.4046468436717987,
"learning_rate": 1.8689028651292805e-05,
"loss": 0.2317,
"step": 470
},
{
"epoch": 0.33542976939203356,
"grad_norm": 0.5121694803237915,
"learning_rate": 1.8661076170510135e-05,
"loss": 0.1965,
"step": 480
},
{
"epoch": 0.3424178895877009,
"grad_norm": 2.7892794609069824,
"learning_rate": 1.8633123689727466e-05,
"loss": 0.2082,
"step": 490
},
{
"epoch": 0.3494060097833683,
"grad_norm": 0.29012227058410645,
"learning_rate": 1.8605171208944793e-05,
"loss": 0.1393,
"step": 500
},
{
"epoch": 0.35639412997903563,
"grad_norm": 0.14969919621944427,
"learning_rate": 1.8577218728162124e-05,
"loss": 0.1917,
"step": 510
},
{
"epoch": 0.363382250174703,
"grad_norm": 2.5301618576049805,
"learning_rate": 1.8549266247379458e-05,
"loss": 0.0939,
"step": 520
},
{
"epoch": 0.37037037037037035,
"grad_norm": 0.5171930193901062,
"learning_rate": 1.8521313766596788e-05,
"loss": 0.0643,
"step": 530
},
{
"epoch": 0.37735849056603776,
"grad_norm": 0.23746764659881592,
"learning_rate": 1.849336128581412e-05,
"loss": 0.2094,
"step": 540
},
{
"epoch": 0.3843466107617051,
"grad_norm": 1.2422730922698975,
"learning_rate": 1.846540880503145e-05,
"loss": 0.109,
"step": 550
},
{
"epoch": 0.3913347309573725,
"grad_norm": 0.8466539978981018,
"learning_rate": 1.843745632424878e-05,
"loss": 0.1514,
"step": 560
},
{
"epoch": 0.39832285115303984,
"grad_norm": 0.201154425740242,
"learning_rate": 1.840950384346611e-05,
"loss": 0.1597,
"step": 570
},
{
"epoch": 0.4053109713487072,
"grad_norm": 0.19054199755191803,
"learning_rate": 1.838155136268344e-05,
"loss": 0.0955,
"step": 580
},
{
"epoch": 0.41229909154437455,
"grad_norm": 1.8036576509475708,
"learning_rate": 1.8353598881900768e-05,
"loss": 0.1369,
"step": 590
},
{
"epoch": 0.4192872117400419,
"grad_norm": 0.3122437596321106,
"learning_rate": 1.83256464011181e-05,
"loss": 0.1439,
"step": 600
},
{
"epoch": 0.42627533193570927,
"grad_norm": 2.6353561878204346,
"learning_rate": 1.829769392033543e-05,
"loss": 0.1566,
"step": 610
},
{
"epoch": 0.4332634521313767,
"grad_norm": 3.645707368850708,
"learning_rate": 1.8269741439552763e-05,
"loss": 0.2505,
"step": 620
},
{
"epoch": 0.44025157232704404,
"grad_norm": 0.20553793013095856,
"learning_rate": 1.8241788958770094e-05,
"loss": 0.0509,
"step": 630
},
{
"epoch": 0.4472396925227114,
"grad_norm": 3.8460299968719482,
"learning_rate": 1.8213836477987425e-05,
"loss": 0.118,
"step": 640
},
{
"epoch": 0.45422781271837875,
"grad_norm": 0.4825759828090668,
"learning_rate": 1.8185883997204755e-05,
"loss": 0.0671,
"step": 650
},
{
"epoch": 0.4612159329140461,
"grad_norm": 2.5147883892059326,
"learning_rate": 1.8157931516422086e-05,
"loss": 0.1381,
"step": 660
},
{
"epoch": 0.46820405310971347,
"grad_norm": 4.72428035736084,
"learning_rate": 1.8129979035639413e-05,
"loss": 0.1093,
"step": 670
},
{
"epoch": 0.4751921733053808,
"grad_norm": 3.091625928878784,
"learning_rate": 1.8102026554856743e-05,
"loss": 0.2363,
"step": 680
},
{
"epoch": 0.48218029350104824,
"grad_norm": 0.10304142534732819,
"learning_rate": 1.8074074074074074e-05,
"loss": 0.1347,
"step": 690
},
{
"epoch": 0.4891684136967156,
"grad_norm": 0.4455307424068451,
"learning_rate": 1.8046121593291405e-05,
"loss": 0.1064,
"step": 700
},
{
"epoch": 0.49615653389238296,
"grad_norm": 0.11594414710998535,
"learning_rate": 1.8018169112508735e-05,
"loss": 0.123,
"step": 710
},
{
"epoch": 0.5031446540880503,
"grad_norm": 0.1224653497338295,
"learning_rate": 1.799021663172607e-05,
"loss": 0.1263,
"step": 720
},
{
"epoch": 0.5101327742837177,
"grad_norm": 4.951889991760254,
"learning_rate": 1.79622641509434e-05,
"loss": 0.2111,
"step": 730
},
{
"epoch": 0.517120894479385,
"grad_norm": 0.16316205263137817,
"learning_rate": 1.793431167016073e-05,
"loss": 0.0366,
"step": 740
},
{
"epoch": 0.5241090146750524,
"grad_norm": 0.1896398663520813,
"learning_rate": 1.790635918937806e-05,
"loss": 0.1656,
"step": 750
},
{
"epoch": 0.5310971348707197,
"grad_norm": 0.3920069932937622,
"learning_rate": 1.7878406708595388e-05,
"loss": 0.2804,
"step": 760
},
{
"epoch": 0.5380852550663872,
"grad_norm": 0.12546010315418243,
"learning_rate": 1.785045422781272e-05,
"loss": 0.1404,
"step": 770
},
{
"epoch": 0.5450733752620545,
"grad_norm": 5.382510185241699,
"learning_rate": 1.782250174703005e-05,
"loss": 0.1194,
"step": 780
},
{
"epoch": 0.5520614954577219,
"grad_norm": 0.3077394366264343,
"learning_rate": 1.779454926624738e-05,
"loss": 0.1744,
"step": 790
},
{
"epoch": 0.5590496156533893,
"grad_norm": 0.41757112741470337,
"learning_rate": 1.776659678546471e-05,
"loss": 0.0729,
"step": 800
},
{
"epoch": 0.5660377358490566,
"grad_norm": 0.17300930619239807,
"learning_rate": 1.773864430468204e-05,
"loss": 0.1443,
"step": 810
},
{
"epoch": 0.573025856044724,
"grad_norm": 0.09419082850217819,
"learning_rate": 1.7710691823899375e-05,
"loss": 0.1587,
"step": 820
},
{
"epoch": 0.5800139762403913,
"grad_norm": 0.14113612473011017,
"learning_rate": 1.7682739343116705e-05,
"loss": 0.1196,
"step": 830
},
{
"epoch": 0.5870020964360587,
"grad_norm": 0.6690120697021484,
"learning_rate": 1.7654786862334036e-05,
"loss": 0.0619,
"step": 840
},
{
"epoch": 0.593990216631726,
"grad_norm": 0.5076020956039429,
"learning_rate": 1.7626834381551363e-05,
"loss": 0.2922,
"step": 850
},
{
"epoch": 0.6009783368273934,
"grad_norm": 0.12461690604686737,
"learning_rate": 1.7598881900768694e-05,
"loss": 0.0351,
"step": 860
},
{
"epoch": 0.6079664570230608,
"grad_norm": 5.236762523651123,
"learning_rate": 1.7570929419986024e-05,
"loss": 0.1571,
"step": 870
},
{
"epoch": 0.6149545772187281,
"grad_norm": 0.12950463593006134,
"learning_rate": 1.7542976939203355e-05,
"loss": 0.0204,
"step": 880
},
{
"epoch": 0.6219426974143956,
"grad_norm": 0.08026222884654999,
"learning_rate": 1.7515024458420685e-05,
"loss": 0.1185,
"step": 890
},
{
"epoch": 0.6289308176100629,
"grad_norm": 4.237085342407227,
"learning_rate": 1.7487071977638016e-05,
"loss": 0.0876,
"step": 900
},
{
"epoch": 0.6359189378057303,
"grad_norm": 0.916748046875,
"learning_rate": 1.7459119496855346e-05,
"loss": 0.0619,
"step": 910
},
{
"epoch": 0.6429070580013976,
"grad_norm": 4.51674747467041,
"learning_rate": 1.743116701607268e-05,
"loss": 0.1116,
"step": 920
},
{
"epoch": 0.649895178197065,
"grad_norm": 1.1469999551773071,
"learning_rate": 1.740321453529001e-05,
"loss": 0.0147,
"step": 930
},
{
"epoch": 0.6568832983927324,
"grad_norm": 2.3627333641052246,
"learning_rate": 1.7375262054507338e-05,
"loss": 0.1485,
"step": 940
},
{
"epoch": 0.6638714185883997,
"grad_norm": 8.216973304748535,
"learning_rate": 1.734730957372467e-05,
"loss": 0.0472,
"step": 950
},
{
"epoch": 0.6708595387840671,
"grad_norm": 3.5901684761047363,
"learning_rate": 1.7319357092942e-05,
"loss": 0.0979,
"step": 960
},
{
"epoch": 0.6778476589797344,
"grad_norm": 7.015125274658203,
"learning_rate": 1.729140461215933e-05,
"loss": 0.0537,
"step": 970
},
{
"epoch": 0.6848357791754018,
"grad_norm": 4.372157096862793,
"learning_rate": 1.726345213137666e-05,
"loss": 0.1152,
"step": 980
},
{
"epoch": 0.6918238993710691,
"grad_norm": 4.862729072570801,
"learning_rate": 1.723549965059399e-05,
"loss": 0.1131,
"step": 990
},
{
"epoch": 0.6988120195667366,
"grad_norm": 0.07732150703668594,
"learning_rate": 1.720754716981132e-05,
"loss": 0.0707,
"step": 1000
},
{
"epoch": 0.705800139762404,
"grad_norm": 0.19970574975013733,
"learning_rate": 1.7179594689028652e-05,
"loss": 0.0738,
"step": 1010
},
{
"epoch": 0.7127882599580713,
"grad_norm": 0.21677608788013458,
"learning_rate": 1.7151642208245983e-05,
"loss": 0.0992,
"step": 1020
},
{
"epoch": 0.7197763801537387,
"grad_norm": 0.10738188028335571,
"learning_rate": 1.7123689727463313e-05,
"loss": 0.0103,
"step": 1030
},
{
"epoch": 0.726764500349406,
"grad_norm": 3.7014143466949463,
"learning_rate": 1.7095737246680644e-05,
"loss": 0.0803,
"step": 1040
},
{
"epoch": 0.7337526205450734,
"grad_norm": 0.04619055986404419,
"learning_rate": 1.7067784765897974e-05,
"loss": 0.049,
"step": 1050
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.2753339409828186,
"learning_rate": 1.7039832285115305e-05,
"loss": 0.144,
"step": 1060
},
{
"epoch": 0.7477288609364081,
"grad_norm": 2.8835108280181885,
"learning_rate": 1.7011879804332635e-05,
"loss": 0.0885,
"step": 1070
},
{
"epoch": 0.7547169811320755,
"grad_norm": 0.22654394805431366,
"learning_rate": 1.6983927323549966e-05,
"loss": 0.0512,
"step": 1080
},
{
"epoch": 0.7617051013277428,
"grad_norm": 2.421400547027588,
"learning_rate": 1.6955974842767297e-05,
"loss": 0.1544,
"step": 1090
},
{
"epoch": 0.7686932215234102,
"grad_norm": 0.03806561604142189,
"learning_rate": 1.6928022361984627e-05,
"loss": 0.0225,
"step": 1100
},
{
"epoch": 0.7756813417190775,
"grad_norm": 0.08347579091787338,
"learning_rate": 1.6900069881201958e-05,
"loss": 0.1386,
"step": 1110
},
{
"epoch": 0.782669461914745,
"grad_norm": 0.2059401273727417,
"learning_rate": 1.687211740041929e-05,
"loss": 0.2406,
"step": 1120
},
{
"epoch": 0.7896575821104123,
"grad_norm": 0.060870036482810974,
"learning_rate": 1.684416491963662e-05,
"loss": 0.1158,
"step": 1130
},
{
"epoch": 0.7966457023060797,
"grad_norm": 0.14614495635032654,
"learning_rate": 1.681621243885395e-05,
"loss": 0.1938,
"step": 1140
},
{
"epoch": 0.803633822501747,
"grad_norm": 0.09593985229730606,
"learning_rate": 1.678825995807128e-05,
"loss": 0.0627,
"step": 1150
},
{
"epoch": 0.8106219426974144,
"grad_norm": 9.543107032775879,
"learning_rate": 1.676030747728861e-05,
"loss": 0.1365,
"step": 1160
},
{
"epoch": 0.8176100628930818,
"grad_norm": 0.09678485244512558,
"learning_rate": 1.673235499650594e-05,
"loss": 0.1438,
"step": 1170
},
{
"epoch": 0.8245981830887491,
"grad_norm": 0.24556872248649597,
"learning_rate": 1.6704402515723272e-05,
"loss": 0.2233,
"step": 1180
},
{
"epoch": 0.8315863032844165,
"grad_norm": 0.06892254948616028,
"learning_rate": 1.6676450034940602e-05,
"loss": 0.0601,
"step": 1190
},
{
"epoch": 0.8385744234800838,
"grad_norm": 0.6429911255836487,
"learning_rate": 1.6648497554157933e-05,
"loss": 0.132,
"step": 1200
},
{
"epoch": 0.8455625436757512,
"grad_norm": 0.03480253368616104,
"learning_rate": 1.6620545073375263e-05,
"loss": 0.0472,
"step": 1210
},
{
"epoch": 0.8525506638714185,
"grad_norm": 0.25101321935653687,
"learning_rate": 1.6592592592592594e-05,
"loss": 0.0485,
"step": 1220
},
{
"epoch": 0.859538784067086,
"grad_norm": 0.0525231771171093,
"learning_rate": 1.6564640111809925e-05,
"loss": 0.1824,
"step": 1230
},
{
"epoch": 0.8665269042627534,
"grad_norm": 4.694910526275635,
"learning_rate": 1.6536687631027255e-05,
"loss": 0.1666,
"step": 1240
},
{
"epoch": 0.8735150244584207,
"grad_norm": 0.09604217112064362,
"learning_rate": 1.6508735150244586e-05,
"loss": 0.0638,
"step": 1250
},
{
"epoch": 0.8805031446540881,
"grad_norm": 9.125068664550781,
"learning_rate": 1.6480782669461916e-05,
"loss": 0.0744,
"step": 1260
},
{
"epoch": 0.8874912648497554,
"grad_norm": 0.5258679986000061,
"learning_rate": 1.6452830188679247e-05,
"loss": 0.019,
"step": 1270
},
{
"epoch": 0.8944793850454228,
"grad_norm": 3.6146976947784424,
"learning_rate": 1.6424877707896577e-05,
"loss": 0.0141,
"step": 1280
},
{
"epoch": 0.9014675052410901,
"grad_norm": 0.05512338504195213,
"learning_rate": 1.6396925227113908e-05,
"loss": 0.0138,
"step": 1290
},
{
"epoch": 0.9084556254367575,
"grad_norm": 0.028468603268265724,
"learning_rate": 1.636897274633124e-05,
"loss": 0.0815,
"step": 1300
},
{
"epoch": 0.9154437456324249,
"grad_norm": 0.7173348665237427,
"learning_rate": 1.634102026554857e-05,
"loss": 0.0919,
"step": 1310
},
{
"epoch": 0.9224318658280922,
"grad_norm": 0.06138516217470169,
"learning_rate": 1.63130677847659e-05,
"loss": 0.1383,
"step": 1320
},
{
"epoch": 0.9294199860237596,
"grad_norm": 10.367135047912598,
"learning_rate": 1.628511530398323e-05,
"loss": 0.1265,
"step": 1330
},
{
"epoch": 0.9364081062194269,
"grad_norm": 0.3321629762649536,
"learning_rate": 1.625716282320056e-05,
"loss": 0.1149,
"step": 1340
},
{
"epoch": 0.9433962264150944,
"grad_norm": 0.21935391426086426,
"learning_rate": 1.622921034241789e-05,
"loss": 0.1147,
"step": 1350
},
{
"epoch": 0.9503843466107617,
"grad_norm": 4.630390644073486,
"learning_rate": 1.6201257861635222e-05,
"loss": 0.1159,
"step": 1360
},
{
"epoch": 0.9573724668064291,
"grad_norm": 0.03748277202248573,
"learning_rate": 1.6173305380852552e-05,
"loss": 0.0029,
"step": 1370
},
{
"epoch": 0.9643605870020965,
"grad_norm": 0.09677435457706451,
"learning_rate": 1.6145352900069883e-05,
"loss": 0.179,
"step": 1380
},
{
"epoch": 0.9713487071977638,
"grad_norm": 0.022979794070124626,
"learning_rate": 1.6117400419287214e-05,
"loss": 0.0034,
"step": 1390
},
{
"epoch": 0.9783368273934312,
"grad_norm": 3.3972229957580566,
"learning_rate": 1.6089447938504544e-05,
"loss": 0.1633,
"step": 1400
},
{
"epoch": 0.9853249475890985,
"grad_norm": 0.0410487987101078,
"learning_rate": 1.6061495457721875e-05,
"loss": 0.0758,
"step": 1410
},
{
"epoch": 0.9923130677847659,
"grad_norm": 0.1519143134355545,
"learning_rate": 1.6033542976939205e-05,
"loss": 0.0967,
"step": 1420
},
{
"epoch": 0.9993011879804332,
"grad_norm": 0.11000066250562668,
"learning_rate": 1.6005590496156536e-05,
"loss": 0.1054,
"step": 1430
},
{
"epoch": 1.0062893081761006,
"grad_norm": 9.595348358154297,
"learning_rate": 1.5977638015373866e-05,
"loss": 0.0367,
"step": 1440
},
{
"epoch": 1.013277428371768,
"grad_norm": 0.044866789132356644,
"learning_rate": 1.5949685534591197e-05,
"loss": 0.145,
"step": 1450
},
{
"epoch": 1.0202655485674355,
"grad_norm": 6.3100361824035645,
"learning_rate": 1.5921733053808524e-05,
"loss": 0.2702,
"step": 1460
},
{
"epoch": 1.0272536687631026,
"grad_norm": 1.0666788816452026,
"learning_rate": 1.5893780573025858e-05,
"loss": 0.1257,
"step": 1470
},
{
"epoch": 1.03424178895877,
"grad_norm": 15.676182746887207,
"learning_rate": 1.586582809224319e-05,
"loss": 0.0373,
"step": 1480
},
{
"epoch": 1.0412299091544375,
"grad_norm": 13.464042663574219,
"learning_rate": 1.583787561146052e-05,
"loss": 0.1107,
"step": 1490
},
{
"epoch": 1.0482180293501049,
"grad_norm": 0.029935991391539574,
"learning_rate": 1.580992313067785e-05,
"loss": 0.0778,
"step": 1500
},
{
"epoch": 1.0552061495457723,
"grad_norm": 1.0842170715332031,
"learning_rate": 1.578197064989518e-05,
"loss": 0.2003,
"step": 1510
},
{
"epoch": 1.0621942697414395,
"grad_norm": 0.8753694891929626,
"learning_rate": 1.575401816911251e-05,
"loss": 0.0869,
"step": 1520
},
{
"epoch": 1.069182389937107,
"grad_norm": 0.03639127314090729,
"learning_rate": 1.572606568832984e-05,
"loss": 0.0035,
"step": 1530
},
{
"epoch": 1.0761705101327743,
"grad_norm": 0.10639658570289612,
"learning_rate": 1.5698113207547172e-05,
"loss": 0.0709,
"step": 1540
},
{
"epoch": 1.0831586303284417,
"grad_norm": 0.09013110399246216,
"learning_rate": 1.56701607267645e-05,
"loss": 0.0041,
"step": 1550
},
{
"epoch": 1.090146750524109,
"grad_norm": 0.1398763209581375,
"learning_rate": 1.564220824598183e-05,
"loss": 0.0039,
"step": 1560
},
{
"epoch": 1.0971348707197763,
"grad_norm": 0.13840065896511078,
"learning_rate": 1.5614255765199164e-05,
"loss": 0.0288,
"step": 1570
},
{
"epoch": 1.1041229909154437,
"grad_norm": 0.03229377046227455,
"learning_rate": 1.5586303284416494e-05,
"loss": 0.023,
"step": 1580
},
{
"epoch": 1.1111111111111112,
"grad_norm": 0.09332744032144547,
"learning_rate": 1.5558350803633825e-05,
"loss": 0.148,
"step": 1590
},
{
"epoch": 1.1180992313067786,
"grad_norm": 0.24537114799022675,
"learning_rate": 1.5530398322851156e-05,
"loss": 0.0143,
"step": 1600
},
{
"epoch": 1.1250873515024458,
"grad_norm": 0.02455071546137333,
"learning_rate": 1.5502445842068486e-05,
"loss": 0.0378,
"step": 1610
},
{
"epoch": 1.1320754716981132,
"grad_norm": 0.027023492380976677,
"learning_rate": 1.5474493361285817e-05,
"loss": 0.0558,
"step": 1620
},
{
"epoch": 1.1390635918937806,
"grad_norm": 0.048673246055841446,
"learning_rate": 1.5446540880503147e-05,
"loss": 0.0519,
"step": 1630
},
{
"epoch": 1.146051712089448,
"grad_norm": 8.062767028808594,
"learning_rate": 1.5418588399720474e-05,
"loss": 0.0404,
"step": 1640
},
{
"epoch": 1.1530398322851152,
"grad_norm": 8.358656883239746,
"learning_rate": 1.5390635918937805e-05,
"loss": 0.1054,
"step": 1650
},
{
"epoch": 1.1600279524807826,
"grad_norm": 1.3057444095611572,
"learning_rate": 1.5362683438155136e-05,
"loss": 0.1049,
"step": 1660
},
{
"epoch": 1.16701607267645,
"grad_norm": 0.018700115382671356,
"learning_rate": 1.533473095737247e-05,
"loss": 0.0106,
"step": 1670
},
{
"epoch": 1.1740041928721174,
"grad_norm": 8.731624603271484,
"learning_rate": 1.53067784765898e-05,
"loss": 0.0541,
"step": 1680
},
{
"epoch": 1.1809923130677848,
"grad_norm": 1.358834147453308,
"learning_rate": 1.527882599580713e-05,
"loss": 0.1515,
"step": 1690
},
{
"epoch": 1.187980433263452,
"grad_norm": 0.05055036395788193,
"learning_rate": 1.525087351502446e-05,
"loss": 0.1181,
"step": 1700
},
{
"epoch": 1.1949685534591195,
"grad_norm": 0.19085977971553802,
"learning_rate": 1.5222921034241792e-05,
"loss": 0.0624,
"step": 1710
},
{
"epoch": 1.2019566736547869,
"grad_norm": 1.282493233680725,
"learning_rate": 1.5194968553459122e-05,
"loss": 0.016,
"step": 1720
},
{
"epoch": 1.2089447938504543,
"grad_norm": 0.03596067801117897,
"learning_rate": 1.5167016072676451e-05,
"loss": 0.0211,
"step": 1730
},
{
"epoch": 1.2159329140461215,
"grad_norm": 0.033778220415115356,
"learning_rate": 1.5139063591893782e-05,
"loss": 0.0105,
"step": 1740
},
{
"epoch": 1.2229210342417889,
"grad_norm": 0.022970452904701233,
"learning_rate": 1.5111111111111112e-05,
"loss": 0.1634,
"step": 1750
},
{
"epoch": 1.2299091544374563,
"grad_norm": 0.08840584754943848,
"learning_rate": 1.5083158630328443e-05,
"loss": 0.0062,
"step": 1760
},
{
"epoch": 1.2368972746331237,
"grad_norm": 0.0703321248292923,
"learning_rate": 1.5055206149545773e-05,
"loss": 0.0554,
"step": 1770
},
{
"epoch": 1.2438853948287911,
"grad_norm": 2.026752471923828,
"learning_rate": 1.5027253668763104e-05,
"loss": 0.03,
"step": 1780
},
{
"epoch": 1.2508735150244585,
"grad_norm": 0.08599984645843506,
"learning_rate": 1.4999301187980435e-05,
"loss": 0.0303,
"step": 1790
},
{
"epoch": 1.2578616352201257,
"grad_norm": 0.0998782068490982,
"learning_rate": 1.4971348707197765e-05,
"loss": 0.1389,
"step": 1800
},
{
"epoch": 1.2648497554157931,
"grad_norm": 1.7642574310302734,
"learning_rate": 1.4943396226415094e-05,
"loss": 0.0556,
"step": 1810
},
{
"epoch": 1.2718378756114606,
"grad_norm": 0.012091516517102718,
"learning_rate": 1.4915443745632425e-05,
"loss": 0.1116,
"step": 1820
},
{
"epoch": 1.2788259958071277,
"grad_norm": 0.030048305168747902,
"learning_rate": 1.4887491264849757e-05,
"loss": 0.0433,
"step": 1830
},
{
"epoch": 1.2858141160027952,
"grad_norm": 0.025840098038315773,
"learning_rate": 1.4859538784067087e-05,
"loss": 0.038,
"step": 1840
},
{
"epoch": 1.2928022361984626,
"grad_norm": 0.4607846736907959,
"learning_rate": 1.4831586303284418e-05,
"loss": 0.0815,
"step": 1850
},
{
"epoch": 1.29979035639413,
"grad_norm": 3.925401210784912,
"learning_rate": 1.4803633822501749e-05,
"loss": 0.0708,
"step": 1860
},
{
"epoch": 1.3067784765897974,
"grad_norm": 5.653714656829834,
"learning_rate": 1.4775681341719079e-05,
"loss": 0.0166,
"step": 1870
},
{
"epoch": 1.3137665967854648,
"grad_norm": 0.02135496586561203,
"learning_rate": 1.474772886093641e-05,
"loss": 0.0782,
"step": 1880
},
{
"epoch": 1.320754716981132,
"grad_norm": 0.020742664113640785,
"learning_rate": 1.471977638015374e-05,
"loss": 0.0048,
"step": 1890
},
{
"epoch": 1.3277428371767994,
"grad_norm": 0.02374901808798313,
"learning_rate": 1.4691823899371069e-05,
"loss": 0.081,
"step": 1900
},
{
"epoch": 1.3347309573724668,
"grad_norm": 0.03886565566062927,
"learning_rate": 1.46638714185884e-05,
"loss": 0.005,
"step": 1910
},
{
"epoch": 1.3417190775681342,
"grad_norm": 0.16429579257965088,
"learning_rate": 1.463591893780573e-05,
"loss": 0.0154,
"step": 1920
},
{
"epoch": 1.3487071977638014,
"grad_norm": 0.15556152164936066,
"learning_rate": 1.4607966457023063e-05,
"loss": 0.0659,
"step": 1930
},
{
"epoch": 1.3556953179594688,
"grad_norm": 0.022907955572009087,
"learning_rate": 1.4580013976240393e-05,
"loss": 0.068,
"step": 1940
},
{
"epoch": 1.3626834381551363,
"grad_norm": 0.028623567894101143,
"learning_rate": 1.4552061495457724e-05,
"loss": 0.0023,
"step": 1950
},
{
"epoch": 1.3696715583508037,
"grad_norm": 7.010847091674805,
"learning_rate": 1.4524109014675054e-05,
"loss": 0.147,
"step": 1960
},
{
"epoch": 1.376659678546471,
"grad_norm": 0.03581179678440094,
"learning_rate": 1.4496156533892385e-05,
"loss": 0.1209,
"step": 1970
},
{
"epoch": 1.3836477987421385,
"grad_norm": 0.019916867837309837,
"learning_rate": 1.4468204053109715e-05,
"loss": 0.0233,
"step": 1980
},
{
"epoch": 1.3906359189378057,
"grad_norm": 10.045747756958008,
"learning_rate": 1.4440251572327044e-05,
"loss": 0.0416,
"step": 1990
},
{
"epoch": 1.397624039133473,
"grad_norm": 0.04035484418272972,
"learning_rate": 1.4412299091544375e-05,
"loss": 0.0265,
"step": 2000
},
{
"epoch": 1.4046121593291405,
"grad_norm": 0.7727859616279602,
"learning_rate": 1.4384346610761705e-05,
"loss": 0.0331,
"step": 2010
},
{
"epoch": 1.4116002795248077,
"grad_norm": 0.018749171867966652,
"learning_rate": 1.4356394129979036e-05,
"loss": 0.0749,
"step": 2020
},
{
"epoch": 1.4185883997204751,
"grad_norm": 5.582619667053223,
"learning_rate": 1.4328441649196368e-05,
"loss": 0.0622,
"step": 2030
},
{
"epoch": 1.4255765199161425,
"grad_norm": 0.03347029909491539,
"learning_rate": 1.4300489168413699e-05,
"loss": 0.0021,
"step": 2040
},
{
"epoch": 1.43256464011181,
"grad_norm": 0.1696143001317978,
"learning_rate": 1.427253668763103e-05,
"loss": 0.0021,
"step": 2050
},
{
"epoch": 1.4395527603074774,
"grad_norm": 0.021968642249703407,
"learning_rate": 1.424458420684836e-05,
"loss": 0.0025,
"step": 2060
},
{
"epoch": 1.4465408805031448,
"grad_norm": 0.007907007820904255,
"learning_rate": 1.421663172606569e-05,
"loss": 0.0478,
"step": 2070
},
{
"epoch": 1.453529000698812,
"grad_norm": 0.035767070949077606,
"learning_rate": 1.418867924528302e-05,
"loss": 0.0012,
"step": 2080
},
{
"epoch": 1.4605171208944794,
"grad_norm": 0.011390830390155315,
"learning_rate": 1.416072676450035e-05,
"loss": 0.1533,
"step": 2090
},
{
"epoch": 1.4675052410901468,
"grad_norm": 7.848016738891602,
"learning_rate": 1.413277428371768e-05,
"loss": 0.0675,
"step": 2100
},
{
"epoch": 1.474493361285814,
"grad_norm": 0.011802544817328453,
"learning_rate": 1.4104821802935011e-05,
"loss": 0.0261,
"step": 2110
},
{
"epoch": 1.4814814814814814,
"grad_norm": 2.9148340225219727,
"learning_rate": 1.4076869322152342e-05,
"loss": 0.2831,
"step": 2120
},
{
"epoch": 1.4884696016771488,
"grad_norm": 0.08199399709701538,
"learning_rate": 1.4048916841369674e-05,
"loss": 0.0088,
"step": 2130
},
{
"epoch": 1.4954577218728162,
"grad_norm": 0.01199902594089508,
"learning_rate": 1.4020964360587004e-05,
"loss": 0.0693,
"step": 2140
},
{
"epoch": 1.5024458420684836,
"grad_norm": 0.07146008312702179,
"learning_rate": 1.3993011879804335e-05,
"loss": 0.0746,
"step": 2150
},
{
"epoch": 1.509433962264151,
"grad_norm": 0.02118564024567604,
"learning_rate": 1.3965059399021666e-05,
"loss": 0.003,
"step": 2160
},
{
"epoch": 1.5164220824598185,
"grad_norm": 0.016972247511148453,
"learning_rate": 1.3937106918238994e-05,
"loss": 0.0106,
"step": 2170
},
{
"epoch": 1.5234102026554857,
"grad_norm": 0.023427631705999374,
"learning_rate": 1.3909154437456325e-05,
"loss": 0.0444,
"step": 2180
},
{
"epoch": 1.530398322851153,
"grad_norm": 0.03280281275510788,
"learning_rate": 1.3881201956673656e-05,
"loss": 0.0026,
"step": 2190
},
{
"epoch": 1.5373864430468203,
"grad_norm": 0.06073632091283798,
"learning_rate": 1.3853249475890986e-05,
"loss": 0.194,
"step": 2200
},
{
"epoch": 1.5443745632424877,
"grad_norm": 0.008114098571240902,
"learning_rate": 1.3825296995108317e-05,
"loss": 0.0016,
"step": 2210
},
{
"epoch": 1.551362683438155,
"grad_norm": 0.0467991828918457,
"learning_rate": 1.3797344514325647e-05,
"loss": 0.1305,
"step": 2220
},
{
"epoch": 1.5583508036338225,
"grad_norm": 0.06984475255012512,
"learning_rate": 1.376939203354298e-05,
"loss": 0.1386,
"step": 2230
},
{
"epoch": 1.56533892382949,
"grad_norm": 0.014983494766056538,
"learning_rate": 1.374143955276031e-05,
"loss": 0.1399,
"step": 2240
},
{
"epoch": 1.5723270440251573,
"grad_norm": 0.01568700559437275,
"learning_rate": 1.3713487071977637e-05,
"loss": 0.1084,
"step": 2250
},
{
"epoch": 1.5793151642208247,
"grad_norm": 10.549098014831543,
"learning_rate": 1.368553459119497e-05,
"loss": 0.0082,
"step": 2260
},
{
"epoch": 1.586303284416492,
"grad_norm": 0.02495586685836315,
"learning_rate": 1.36575821104123e-05,
"loss": 0.0019,
"step": 2270
},
{
"epoch": 1.5932914046121593,
"grad_norm": 0.01876898854970932,
"learning_rate": 1.362962962962963e-05,
"loss": 0.1002,
"step": 2280
},
{
"epoch": 1.6002795248078265,
"grad_norm": 0.044464971870183945,
"learning_rate": 1.3601677148846961e-05,
"loss": 0.0019,
"step": 2290
},
{
"epoch": 1.607267645003494,
"grad_norm": 0.04108593240380287,
"learning_rate": 1.3573724668064292e-05,
"loss": 0.0753,
"step": 2300
},
{
"epoch": 1.6142557651991614,
"grad_norm": 7.812976837158203,
"learning_rate": 1.3545772187281622e-05,
"loss": 0.2875,
"step": 2310
},
{
"epoch": 1.6212438853948288,
"grad_norm": 4.261682987213135,
"learning_rate": 1.3517819706498953e-05,
"loss": 0.1059,
"step": 2320
},
{
"epoch": 1.6282320055904962,
"grad_norm": 0.03309008479118347,
"learning_rate": 1.3489867225716285e-05,
"loss": 0.0051,
"step": 2330
},
{
"epoch": 1.6352201257861636,
"grad_norm": 9.983598709106445,
"learning_rate": 1.3461914744933612e-05,
"loss": 0.0249,
"step": 2340
},
{
"epoch": 1.642208245981831,
"grad_norm": 4.973055839538574,
"learning_rate": 1.3433962264150943e-05,
"loss": 0.1441,
"step": 2350
},
{
"epoch": 1.6491963661774982,
"grad_norm": 9.258291244506836,
"learning_rate": 1.3406009783368275e-05,
"loss": 0.1151,
"step": 2360
},
{
"epoch": 1.6561844863731656,
"grad_norm": 0.015135078690946102,
"learning_rate": 1.3378057302585606e-05,
"loss": 0.0458,
"step": 2370
},
{
"epoch": 1.663172606568833,
"grad_norm": 0.027922656387090683,
"learning_rate": 1.3350104821802936e-05,
"loss": 0.012,
"step": 2380
},
{
"epoch": 1.6701607267645002,
"grad_norm": 0.043312493711709976,
"learning_rate": 1.3322152341020267e-05,
"loss": 0.1028,
"step": 2390
},
{
"epoch": 1.6771488469601676,
"grad_norm": 0.012087292037904263,
"learning_rate": 1.3294199860237597e-05,
"loss": 0.0475,
"step": 2400
},
{
"epoch": 1.684136967155835,
"grad_norm": 0.5556979775428772,
"learning_rate": 1.3266247379454928e-05,
"loss": 0.0142,
"step": 2410
},
{
"epoch": 1.6911250873515025,
"grad_norm": 5.196042537689209,
"learning_rate": 1.3238294898672259e-05,
"loss": 0.1002,
"step": 2420
},
{
"epoch": 1.6981132075471699,
"grad_norm": 0.010861317627131939,
"learning_rate": 1.3210342417889587e-05,
"loss": 0.0953,
"step": 2430
},
{
"epoch": 1.7051013277428373,
"grad_norm": 0.017955880612134933,
"learning_rate": 1.3182389937106918e-05,
"loss": 0.0642,
"step": 2440
},
{
"epoch": 1.7120894479385047,
"grad_norm": 0.028966449201107025,
"learning_rate": 1.3154437456324249e-05,
"loss": 0.1619,
"step": 2450
},
{
"epoch": 1.719077568134172,
"grad_norm": 0.8712167739868164,
"learning_rate": 1.312648497554158e-05,
"loss": 0.0324,
"step": 2460
},
{
"epoch": 1.7260656883298393,
"grad_norm": 0.2874550521373749,
"learning_rate": 1.3098532494758911e-05,
"loss": 0.026,
"step": 2470
},
{
"epoch": 1.7330538085255065,
"grad_norm": 0.010579444468021393,
"learning_rate": 1.3070580013976242e-05,
"loss": 0.0048,
"step": 2480
},
{
"epoch": 1.740041928721174,
"grad_norm": 0.15773366391658783,
"learning_rate": 1.3042627533193573e-05,
"loss": 0.108,
"step": 2490
},
{
"epoch": 1.7470300489168413,
"grad_norm": 0.10269400477409363,
"learning_rate": 1.3014675052410903e-05,
"loss": 0.0884,
"step": 2500
},
{
"epoch": 1.7540181691125087,
"grad_norm": 0.012210669927299023,
"learning_rate": 1.2986722571628234e-05,
"loss": 0.0025,
"step": 2510
},
{
"epoch": 1.7610062893081762,
"grad_norm": 0.046242229640483856,
"learning_rate": 1.2958770090845563e-05,
"loss": 0.0124,
"step": 2520
},
{
"epoch": 1.7679944095038436,
"grad_norm": 12.532265663146973,
"learning_rate": 1.2930817610062893e-05,
"loss": 0.1502,
"step": 2530
},
{
"epoch": 1.774982529699511,
"grad_norm": 0.018337048590183258,
"learning_rate": 1.2902865129280224e-05,
"loss": 0.0402,
"step": 2540
},
{
"epoch": 1.7819706498951782,
"grad_norm": 0.9082455039024353,
"learning_rate": 1.2874912648497554e-05,
"loss": 0.0719,
"step": 2550
},
{
"epoch": 1.7889587700908456,
"grad_norm": 0.6127817034721375,
"learning_rate": 1.2846960167714887e-05,
"loss": 0.011,
"step": 2560
},
{
"epoch": 1.7959468902865128,
"grad_norm": 0.02741195261478424,
"learning_rate": 1.2819007686932217e-05,
"loss": 0.0436,
"step": 2570
},
{
"epoch": 1.8029350104821802,
"grad_norm": 0.012224216014146805,
"learning_rate": 1.2791055206149548e-05,
"loss": 0.0026,
"step": 2580
},
{
"epoch": 1.8099231306778476,
"grad_norm": 0.01689975894987583,
"learning_rate": 1.2763102725366878e-05,
"loss": 0.0067,
"step": 2590
},
{
"epoch": 1.816911250873515,
"grad_norm": 0.010701341554522514,
"learning_rate": 1.2735150244584207e-05,
"loss": 0.008,
"step": 2600
},
{
"epoch": 1.8238993710691824,
"grad_norm": 0.12163686007261276,
"learning_rate": 1.2707197763801538e-05,
"loss": 0.0767,
"step": 2610
},
{
"epoch": 1.8308874912648498,
"grad_norm": 0.07165365666151047,
"learning_rate": 1.2679245283018868e-05,
"loss": 0.0028,
"step": 2620
},
{
"epoch": 1.8378756114605173,
"grad_norm": 0.007849736139178276,
"learning_rate": 1.2651292802236199e-05,
"loss": 0.0488,
"step": 2630
},
{
"epoch": 1.8448637316561844,
"grad_norm": 0.01707429252564907,
"learning_rate": 1.262334032145353e-05,
"loss": 0.0022,
"step": 2640
},
{
"epoch": 1.8518518518518519,
"grad_norm": 6.567415237426758,
"learning_rate": 1.259538784067086e-05,
"loss": 0.0677,
"step": 2650
},
{
"epoch": 1.858839972047519,
"grad_norm": 0.005324068479239941,
"learning_rate": 1.2567435359888192e-05,
"loss": 0.0057,
"step": 2660
},
{
"epoch": 1.8658280922431865,
"grad_norm": 11.791135787963867,
"learning_rate": 1.2539482879105523e-05,
"loss": 0.0764,
"step": 2670
},
{
"epoch": 1.8728162124388539,
"grad_norm": 0.005892631132155657,
"learning_rate": 1.2511530398322853e-05,
"loss": 0.0736,
"step": 2680
},
{
"epoch": 1.8798043326345213,
"grad_norm": 0.04005246236920357,
"learning_rate": 1.2483577917540182e-05,
"loss": 0.0617,
"step": 2690
},
{
"epoch": 1.8867924528301887,
"grad_norm": 0.010624396614730358,
"learning_rate": 1.2455625436757513e-05,
"loss": 0.0011,
"step": 2700
},
{
"epoch": 1.8937805730258561,
"grad_norm": 0.01761520840227604,
"learning_rate": 1.2427672955974843e-05,
"loss": 0.1574,
"step": 2710
},
{
"epoch": 1.9007686932215235,
"grad_norm": 0.05212152749300003,
"learning_rate": 1.2399720475192174e-05,
"loss": 0.0618,
"step": 2720
},
{
"epoch": 1.9077568134171907,
"grad_norm": 0.012297810055315495,
"learning_rate": 1.2371767994409504e-05,
"loss": 0.0802,
"step": 2730
},
{
"epoch": 1.9147449336128581,
"grad_norm": 1.4671796560287476,
"learning_rate": 1.2343815513626835e-05,
"loss": 0.0351,
"step": 2740
},
{
"epoch": 1.9217330538085255,
"grad_norm": 1.28997802734375,
"learning_rate": 1.2315863032844166e-05,
"loss": 0.0069,
"step": 2750
},
{
"epoch": 1.9287211740041927,
"grad_norm": 0.15810908377170563,
"learning_rate": 1.2287910552061498e-05,
"loss": 0.0037,
"step": 2760
},
{
"epoch": 1.9357092941998602,
"grad_norm": 0.01939568482339382,
"learning_rate": 1.2259958071278828e-05,
"loss": 0.1375,
"step": 2770
},
{
"epoch": 1.9426974143955276,
"grad_norm": 0.023233506828546524,
"learning_rate": 1.2232005590496157e-05,
"loss": 0.0016,
"step": 2780
},
{
"epoch": 1.949685534591195,
"grad_norm": 13.808353424072266,
"learning_rate": 1.2204053109713488e-05,
"loss": 0.1251,
"step": 2790
},
{
"epoch": 1.9566736547868624,
"grad_norm": 0.007430393248796463,
"learning_rate": 1.2176100628930818e-05,
"loss": 0.0957,
"step": 2800
},
{
"epoch": 1.9636617749825298,
"grad_norm": 0.015328769572079182,
"learning_rate": 1.2148148148148149e-05,
"loss": 0.0019,
"step": 2810
},
{
"epoch": 1.9706498951781972,
"grad_norm": 0.05328751727938652,
"learning_rate": 1.212019566736548e-05,
"loss": 0.2193,
"step": 2820
},
{
"epoch": 1.9776380153738644,
"grad_norm": 0.05544736981391907,
"learning_rate": 1.209224318658281e-05,
"loss": 0.011,
"step": 2830
},
{
"epoch": 1.9846261355695318,
"grad_norm": 0.006596289575099945,
"learning_rate": 1.206429070580014e-05,
"loss": 0.0454,
"step": 2840
},
{
"epoch": 1.991614255765199,
"grad_norm": 7.7824015617370605,
"learning_rate": 1.2036338225017471e-05,
"loss": 0.024,
"step": 2850
},
{
"epoch": 1.9986023759608664,
"grad_norm": 0.024772530421614647,
"learning_rate": 1.2008385744234804e-05,
"loss": 0.0109,
"step": 2860
}
],
"logging_steps": 10,
"max_steps": 7155,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3093823953408000.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}