MultimodalAgent's picture
Upload ensemble_student_6/checkpoint-3600/trainer_state.json with huggingface_hub
d47e8ce verified
Raw
History Blame Contribute Delete
63.4 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.7094017094017095,
"eval_steps": 500,
"global_step": 3600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004748338081671415,
"grad_norm": 17.63759727702213,
"learning_rate": 8.547008547008548e-07,
"loss": 8.2288,
"step": 10
},
{
"epoch": 0.00949667616334283,
"grad_norm": 16.55331628816789,
"learning_rate": 1.8043684710351376e-06,
"loss": 8.1201,
"step": 20
},
{
"epoch": 0.014245014245014245,
"grad_norm": 8.640410254153872,
"learning_rate": 2.7540360873694207e-06,
"loss": 7.5412,
"step": 30
},
{
"epoch": 0.01899335232668566,
"grad_norm": 4.566505177802538,
"learning_rate": 3.7037037037037037e-06,
"loss": 6.8629,
"step": 40
},
{
"epoch": 0.023741690408357077,
"grad_norm": 2.864446696398661,
"learning_rate": 4.653371320037987e-06,
"loss": 6.4326,
"step": 50
},
{
"epoch": 0.02849002849002849,
"grad_norm": 2.5971601795227213,
"learning_rate": 5.60303893637227e-06,
"loss": 6.1193,
"step": 60
},
{
"epoch": 0.03323836657169991,
"grad_norm": 3.4992046160114754,
"learning_rate": 6.5527065527065525e-06,
"loss": 5.7872,
"step": 70
},
{
"epoch": 0.03798670465337132,
"grad_norm": 4.463387222422498,
"learning_rate": 7.5023741690408365e-06,
"loss": 5.2738,
"step": 80
},
{
"epoch": 0.042735042735042736,
"grad_norm": 4.248262221605644,
"learning_rate": 8.452041785375119e-06,
"loss": 4.7131,
"step": 90
},
{
"epoch": 0.04748338081671415,
"grad_norm": 5.191970326833423,
"learning_rate": 9.401709401709402e-06,
"loss": 4.2233,
"step": 100
},
{
"epoch": 0.052231718898385564,
"grad_norm": 5.432360531003347,
"learning_rate": 1.0351377018043685e-05,
"loss": 3.8663,
"step": 110
},
{
"epoch": 0.05698005698005698,
"grad_norm": 4.978861964528436,
"learning_rate": 1.1301044634377968e-05,
"loss": 3.6028,
"step": 120
},
{
"epoch": 0.06172839506172839,
"grad_norm": 5.599559449586108,
"learning_rate": 1.2250712250712251e-05,
"loss": 3.3907,
"step": 130
},
{
"epoch": 0.06647673314339982,
"grad_norm": 7.446137447102387,
"learning_rate": 1.3200379867046534e-05,
"loss": 3.2088,
"step": 140
},
{
"epoch": 0.07122507122507123,
"grad_norm": 6.509696264905773,
"learning_rate": 1.4150047483380819e-05,
"loss": 3.0656,
"step": 150
},
{
"epoch": 0.07597340930674264,
"grad_norm": 7.829914077248805,
"learning_rate": 1.50997150997151e-05,
"loss": 2.9331,
"step": 160
},
{
"epoch": 0.08072174738841406,
"grad_norm": 7.285383947823254,
"learning_rate": 1.604938271604938e-05,
"loss": 2.838,
"step": 170
},
{
"epoch": 0.08547008547008547,
"grad_norm": 10.067970984061127,
"learning_rate": 1.6999050332383668e-05,
"loss": 2.7511,
"step": 180
},
{
"epoch": 0.09021842355175688,
"grad_norm": 8.064377047624383,
"learning_rate": 1.794871794871795e-05,
"loss": 2.6917,
"step": 190
},
{
"epoch": 0.0949667616334283,
"grad_norm": 6.019477719424393,
"learning_rate": 1.889838556505223e-05,
"loss": 2.6274,
"step": 200
},
{
"epoch": 0.09971509971509972,
"grad_norm": 17.329313918397528,
"learning_rate": 1.9848053181386517e-05,
"loss": 2.5887,
"step": 210
},
{
"epoch": 0.10446343779677113,
"grad_norm": 18.525182928882433,
"learning_rate": 2.07977207977208e-05,
"loss": 2.5259,
"step": 220
},
{
"epoch": 0.10921177587844255,
"grad_norm": 12.065734481434616,
"learning_rate": 2.174738841405508e-05,
"loss": 2.4817,
"step": 230
},
{
"epoch": 0.11396011396011396,
"grad_norm": 11.52698789657618,
"learning_rate": 2.2697056030389367e-05,
"loss": 2.4295,
"step": 240
},
{
"epoch": 0.11870845204178537,
"grad_norm": 7.988095459399686,
"learning_rate": 2.364672364672365e-05,
"loss": 2.3606,
"step": 250
},
{
"epoch": 0.12345679012345678,
"grad_norm": 15.262786387358695,
"learning_rate": 2.459639126305793e-05,
"loss": 2.2793,
"step": 260
},
{
"epoch": 0.1282051282051282,
"grad_norm": 18.38370103563499,
"learning_rate": 2.5546058879392216e-05,
"loss": 2.2197,
"step": 270
},
{
"epoch": 0.13295346628679963,
"grad_norm": 17.766932385864916,
"learning_rate": 2.64957264957265e-05,
"loss": 2.1368,
"step": 280
},
{
"epoch": 0.13770180436847104,
"grad_norm": 11.834696406394011,
"learning_rate": 2.744539411206078e-05,
"loss": 2.0475,
"step": 290
},
{
"epoch": 0.14245014245014245,
"grad_norm": 9.49347564334006,
"learning_rate": 2.839506172839506e-05,
"loss": 1.9491,
"step": 300
},
{
"epoch": 0.14719848053181386,
"grad_norm": 8.951641858579654,
"learning_rate": 2.9344729344729345e-05,
"loss": 1.8526,
"step": 310
},
{
"epoch": 0.15194681861348527,
"grad_norm": 11.26441744664593,
"learning_rate": 3.029439696106363e-05,
"loss": 1.7429,
"step": 320
},
{
"epoch": 0.15669515669515668,
"grad_norm": 13.666989025258017,
"learning_rate": 3.1244064577397914e-05,
"loss": 1.6404,
"step": 330
},
{
"epoch": 0.16144349477682812,
"grad_norm": 8.966939896866544,
"learning_rate": 3.2193732193732194e-05,
"loss": 1.5652,
"step": 340
},
{
"epoch": 0.16619183285849953,
"grad_norm": 9.032487935353275,
"learning_rate": 3.314339981006648e-05,
"loss": 1.484,
"step": 350
},
{
"epoch": 0.17094017094017094,
"grad_norm": 8.611413445684526,
"learning_rate": 3.409306742640076e-05,
"loss": 1.4051,
"step": 360
},
{
"epoch": 0.17568850902184235,
"grad_norm": 6.481296319433045,
"learning_rate": 3.504273504273504e-05,
"loss": 1.3608,
"step": 370
},
{
"epoch": 0.18043684710351376,
"grad_norm": 8.883017454929599,
"learning_rate": 3.599240265906933e-05,
"loss": 1.3003,
"step": 380
},
{
"epoch": 0.18518518518518517,
"grad_norm": 7.101109204455062,
"learning_rate": 3.694207027540361e-05,
"loss": 1.2457,
"step": 390
},
{
"epoch": 0.1899335232668566,
"grad_norm": 7.269008071103493,
"learning_rate": 3.789173789173789e-05,
"loss": 1.1995,
"step": 400
},
{
"epoch": 0.19468186134852802,
"grad_norm": 6.401062623759157,
"learning_rate": 3.884140550807218e-05,
"loss": 1.1731,
"step": 410
},
{
"epoch": 0.19943019943019943,
"grad_norm": 6.037506539024576,
"learning_rate": 3.979107312440646e-05,
"loss": 1.1356,
"step": 420
},
{
"epoch": 0.20417853751187084,
"grad_norm": 4.350981374666263,
"learning_rate": 4.074074074074074e-05,
"loss": 1.0801,
"step": 430
},
{
"epoch": 0.20892687559354226,
"grad_norm": 5.533196437793109,
"learning_rate": 4.169040835707503e-05,
"loss": 1.0484,
"step": 440
},
{
"epoch": 0.21367521367521367,
"grad_norm": 4.648569253166192,
"learning_rate": 4.264007597340931e-05,
"loss": 1.0124,
"step": 450
},
{
"epoch": 0.2184235517568851,
"grad_norm": 4.919584486678508,
"learning_rate": 4.358974358974359e-05,
"loss": 0.9834,
"step": 460
},
{
"epoch": 0.22317188983855651,
"grad_norm": 4.729852263611172,
"learning_rate": 4.453941120607788e-05,
"loss": 0.9619,
"step": 470
},
{
"epoch": 0.22792022792022792,
"grad_norm": 3.5279772277937833,
"learning_rate": 4.548907882241216e-05,
"loss": 0.9369,
"step": 480
},
{
"epoch": 0.23266856600189934,
"grad_norm": 4.113530945603486,
"learning_rate": 4.643874643874644e-05,
"loss": 0.9034,
"step": 490
},
{
"epoch": 0.23741690408357075,
"grad_norm": 6.40208264671417,
"learning_rate": 4.738841405508073e-05,
"loss": 0.8878,
"step": 500
},
{
"epoch": 0.24216524216524216,
"grad_norm": 3.4822280397629584,
"learning_rate": 4.833808167141501e-05,
"loss": 0.8636,
"step": 510
},
{
"epoch": 0.24691358024691357,
"grad_norm": 4.59120452944987,
"learning_rate": 4.928774928774929e-05,
"loss": 0.8405,
"step": 520
},
{
"epoch": 0.251661918328585,
"grad_norm": 3.635159997832406,
"learning_rate": 5.0237416904083576e-05,
"loss": 0.8205,
"step": 530
},
{
"epoch": 0.2564102564102564,
"grad_norm": 3.6917616564179836,
"learning_rate": 5.118708452041786e-05,
"loss": 0.7915,
"step": 540
},
{
"epoch": 0.2611585944919278,
"grad_norm": 3.959116512724086,
"learning_rate": 5.213675213675214e-05,
"loss": 0.7723,
"step": 550
},
{
"epoch": 0.26590693257359926,
"grad_norm": 4.861766032436544,
"learning_rate": 5.308641975308642e-05,
"loss": 0.7502,
"step": 560
},
{
"epoch": 0.2706552706552707,
"grad_norm": 3.7737799895240514,
"learning_rate": 5.403608736942071e-05,
"loss": 0.7323,
"step": 570
},
{
"epoch": 0.2754036087369421,
"grad_norm": 6.459534619897314,
"learning_rate": 5.498575498575499e-05,
"loss": 0.7176,
"step": 580
},
{
"epoch": 0.2801519468186135,
"grad_norm": 4.487403313410193,
"learning_rate": 5.5935422602089274e-05,
"loss": 0.6986,
"step": 590
},
{
"epoch": 0.2849002849002849,
"grad_norm": 4.856897608888932,
"learning_rate": 5.6885090218423554e-05,
"loss": 0.6707,
"step": 600
},
{
"epoch": 0.2896486229819563,
"grad_norm": 4.625463065333766,
"learning_rate": 5.7834757834757834e-05,
"loss": 0.657,
"step": 610
},
{
"epoch": 0.2943969610636277,
"grad_norm": 3.732435852993538,
"learning_rate": 5.878442545109212e-05,
"loss": 0.6231,
"step": 620
},
{
"epoch": 0.29914529914529914,
"grad_norm": 2.9176115431301985,
"learning_rate": 5.97340930674264e-05,
"loss": 0.6139,
"step": 630
},
{
"epoch": 0.30389363722697055,
"grad_norm": 3.808564166407003,
"learning_rate": 6.068376068376068e-05,
"loss": 0.5897,
"step": 640
},
{
"epoch": 0.30864197530864196,
"grad_norm": 3.470737056485456,
"learning_rate": 6.163342830009497e-05,
"loss": 0.5821,
"step": 650
},
{
"epoch": 0.31339031339031337,
"grad_norm": 2.6054168172531553,
"learning_rate": 6.258309591642925e-05,
"loss": 0.5551,
"step": 660
},
{
"epoch": 0.3181386514719848,
"grad_norm": 2.9337690818827085,
"learning_rate": 6.353276353276355e-05,
"loss": 0.5403,
"step": 670
},
{
"epoch": 0.32288698955365625,
"grad_norm": 3.2883499653828054,
"learning_rate": 6.448243114909783e-05,
"loss": 0.5175,
"step": 680
},
{
"epoch": 0.32763532763532766,
"grad_norm": 2.4638098712125833,
"learning_rate": 6.54320987654321e-05,
"loss": 0.5078,
"step": 690
},
{
"epoch": 0.33238366571699907,
"grad_norm": 2.9399706476560024,
"learning_rate": 6.638176638176638e-05,
"loss": 0.4911,
"step": 700
},
{
"epoch": 0.3371320037986705,
"grad_norm": 3.936910224438004,
"learning_rate": 6.733143399810066e-05,
"loss": 0.4807,
"step": 710
},
{
"epoch": 0.3418803418803419,
"grad_norm": 2.770685544684776,
"learning_rate": 6.828110161443494e-05,
"loss": 0.4636,
"step": 720
},
{
"epoch": 0.3466286799620133,
"grad_norm": 2.3416686629884125,
"learning_rate": 6.923076923076924e-05,
"loss": 0.4501,
"step": 730
},
{
"epoch": 0.3513770180436847,
"grad_norm": 2.503600609224435,
"learning_rate": 7.018043684710352e-05,
"loss": 0.433,
"step": 740
},
{
"epoch": 0.3561253561253561,
"grad_norm": 3.092052162474194,
"learning_rate": 7.11301044634378e-05,
"loss": 0.4217,
"step": 750
},
{
"epoch": 0.36087369420702753,
"grad_norm": 2.5808584963735757,
"learning_rate": 7.207977207977208e-05,
"loss": 0.4058,
"step": 760
},
{
"epoch": 0.36562203228869894,
"grad_norm": 2.3040299943298765,
"learning_rate": 7.302943969610636e-05,
"loss": 0.3953,
"step": 770
},
{
"epoch": 0.37037037037037035,
"grad_norm": 2.4247285440193114,
"learning_rate": 7.397910731244065e-05,
"loss": 0.3885,
"step": 780
},
{
"epoch": 0.37511870845204176,
"grad_norm": 2.299460271661061,
"learning_rate": 7.492877492877494e-05,
"loss": 0.3856,
"step": 790
},
{
"epoch": 0.3798670465337132,
"grad_norm": 2.3728777171514643,
"learning_rate": 7.587844254510922e-05,
"loss": 0.3671,
"step": 800
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.5858525108449084,
"learning_rate": 7.68281101614435e-05,
"loss": 0.3548,
"step": 810
},
{
"epoch": 0.38936372269705605,
"grad_norm": 2.221079942306085,
"learning_rate": 7.777777777777778e-05,
"loss": 0.3413,
"step": 820
},
{
"epoch": 0.39411206077872746,
"grad_norm": 1.4927718796013116,
"learning_rate": 7.872744539411206e-05,
"loss": 0.3335,
"step": 830
},
{
"epoch": 0.39886039886039887,
"grad_norm": 1.8478935006184494,
"learning_rate": 7.967711301044634e-05,
"loss": 0.3141,
"step": 840
},
{
"epoch": 0.4036087369420703,
"grad_norm": 1.5283818988709323,
"learning_rate": 8.062678062678063e-05,
"loss": 0.3068,
"step": 850
},
{
"epoch": 0.4083570750237417,
"grad_norm": 1.5319669328773278,
"learning_rate": 8.157644824311491e-05,
"loss": 0.3002,
"step": 860
},
{
"epoch": 0.4131054131054131,
"grad_norm": 1.750633828307664,
"learning_rate": 8.25261158594492e-05,
"loss": 0.2937,
"step": 870
},
{
"epoch": 0.4178537511870845,
"grad_norm": 1.3097472277186086,
"learning_rate": 8.347578347578347e-05,
"loss": 0.2779,
"step": 880
},
{
"epoch": 0.4226020892687559,
"grad_norm": 1.431925867592132,
"learning_rate": 8.442545109211775e-05,
"loss": 0.27,
"step": 890
},
{
"epoch": 0.42735042735042733,
"grad_norm": 1.2153305608248626,
"learning_rate": 8.537511870845205e-05,
"loss": 0.2628,
"step": 900
},
{
"epoch": 0.43209876543209874,
"grad_norm": 1.5144247743213561,
"learning_rate": 8.632478632478634e-05,
"loss": 0.2534,
"step": 910
},
{
"epoch": 0.4368471035137702,
"grad_norm": 1.396573710314859,
"learning_rate": 8.727445394112062e-05,
"loss": 0.2417,
"step": 920
},
{
"epoch": 0.4415954415954416,
"grad_norm": 1.6569382378591764,
"learning_rate": 8.82241215574549e-05,
"loss": 0.2398,
"step": 930
},
{
"epoch": 0.44634377967711303,
"grad_norm": 0.9462961666890293,
"learning_rate": 8.917378917378918e-05,
"loss": 0.229,
"step": 940
},
{
"epoch": 0.45109211775878444,
"grad_norm": 1.2696416429272817,
"learning_rate": 9.012345679012346e-05,
"loss": 0.2234,
"step": 950
},
{
"epoch": 0.45584045584045585,
"grad_norm": 1.2709884797532915,
"learning_rate": 9.107312440645774e-05,
"loss": 0.2196,
"step": 960
},
{
"epoch": 0.46058879392212726,
"grad_norm": 1.2498733670203395,
"learning_rate": 9.202279202279203e-05,
"loss": 0.2128,
"step": 970
},
{
"epoch": 0.46533713200379867,
"grad_norm": 0.9528039566238911,
"learning_rate": 9.297245963912631e-05,
"loss": 0.198,
"step": 980
},
{
"epoch": 0.4700854700854701,
"grad_norm": 1.3142562475777297,
"learning_rate": 9.392212725546059e-05,
"loss": 0.2016,
"step": 990
},
{
"epoch": 0.4748338081671415,
"grad_norm": 1.3811285187473328,
"learning_rate": 9.487179487179487e-05,
"loss": 0.1929,
"step": 1000
},
{
"epoch": 0.4795821462488129,
"grad_norm": 1.2410934628823824,
"learning_rate": 9.582146248812915e-05,
"loss": 0.1869,
"step": 1010
},
{
"epoch": 0.4843304843304843,
"grad_norm": 0.9115010937849052,
"learning_rate": 9.677113010446344e-05,
"loss": 0.1791,
"step": 1020
},
{
"epoch": 0.4890788224121557,
"grad_norm": 0.7639624650030499,
"learning_rate": 9.772079772079774e-05,
"loss": 0.1739,
"step": 1030
},
{
"epoch": 0.49382716049382713,
"grad_norm": 0.8168718163889598,
"learning_rate": 9.867046533713202e-05,
"loss": 0.1704,
"step": 1040
},
{
"epoch": 0.4985754985754986,
"grad_norm": 0.8029665722621272,
"learning_rate": 9.96201329534663e-05,
"loss": 0.1668,
"step": 1050
},
{
"epoch": 0.50332383665717,
"grad_norm": 0.8989761598172166,
"learning_rate": 9.999990109907572e-05,
"loss": 0.1626,
"step": 1060
},
{
"epoch": 0.5080721747388414,
"grad_norm": 0.935687615820874,
"learning_rate": 9.999929670595522e-05,
"loss": 0.157,
"step": 1070
},
{
"epoch": 0.5128205128205128,
"grad_norm": 0.6547919652872191,
"learning_rate": 9.99981428713058e-05,
"loss": 0.1528,
"step": 1080
},
{
"epoch": 0.5175688509021842,
"grad_norm": 0.9362644232120221,
"learning_rate": 9.999643960780693e-05,
"loss": 0.1521,
"step": 1090
},
{
"epoch": 0.5223171889838556,
"grad_norm": 0.6337819742683569,
"learning_rate": 9.999418693417573e-05,
"loss": 0.1438,
"step": 1100
},
{
"epoch": 0.5270655270655271,
"grad_norm": 0.7212325246537258,
"learning_rate": 9.999138487516681e-05,
"loss": 0.1386,
"step": 1110
},
{
"epoch": 0.5318138651471985,
"grad_norm": 0.61949338190157,
"learning_rate": 9.998803346157195e-05,
"loss": 0.1354,
"step": 1120
},
{
"epoch": 0.5365622032288699,
"grad_norm": 1.0431614499800197,
"learning_rate": 9.99841327302198e-05,
"loss": 0.137,
"step": 1130
},
{
"epoch": 0.5413105413105413,
"grad_norm": 0.6605352823286541,
"learning_rate": 9.997968272397542e-05,
"loss": 0.1314,
"step": 1140
},
{
"epoch": 0.5460588793922128,
"grad_norm": 0.5549351637682506,
"learning_rate": 9.997468349173985e-05,
"loss": 0.1293,
"step": 1150
},
{
"epoch": 0.5508072174738842,
"grad_norm": 0.899789973194271,
"learning_rate": 9.99691350884496e-05,
"loss": 0.1265,
"step": 1160
},
{
"epoch": 0.5555555555555556,
"grad_norm": 0.5183845915341633,
"learning_rate": 9.996303757507602e-05,
"loss": 0.1252,
"step": 1170
},
{
"epoch": 0.560303893637227,
"grad_norm": 0.5042036783549486,
"learning_rate": 9.995639101862453e-05,
"loss": 0.1232,
"step": 1180
},
{
"epoch": 0.5650522317188984,
"grad_norm": 0.4866674490443575,
"learning_rate": 9.994919549213408e-05,
"loss": 0.1215,
"step": 1190
},
{
"epoch": 0.5698005698005698,
"grad_norm": 1.3326692291107858,
"learning_rate": 9.99414510746762e-05,
"loss": 0.1204,
"step": 1200
},
{
"epoch": 0.5745489078822412,
"grad_norm": 0.5373611233934212,
"learning_rate": 9.993315785135416e-05,
"loss": 0.1201,
"step": 1210
},
{
"epoch": 0.5792972459639126,
"grad_norm": 0.5800188355485815,
"learning_rate": 9.99243159133021e-05,
"loss": 0.1193,
"step": 1220
},
{
"epoch": 0.584045584045584,
"grad_norm": 0.4625068911720861,
"learning_rate": 9.991492535768392e-05,
"loss": 0.1151,
"step": 1230
},
{
"epoch": 0.5887939221272555,
"grad_norm": 0.5124385169771172,
"learning_rate": 9.99049862876923e-05,
"loss": 0.1135,
"step": 1240
},
{
"epoch": 0.5935422602089269,
"grad_norm": 0.6051916207047685,
"learning_rate": 9.989449881254754e-05,
"loss": 0.1136,
"step": 1250
},
{
"epoch": 0.5982905982905983,
"grad_norm": 0.5365607455610492,
"learning_rate": 9.988346304749635e-05,
"loss": 0.1113,
"step": 1260
},
{
"epoch": 0.6030389363722697,
"grad_norm": 0.43372022580677017,
"learning_rate": 9.987187911381059e-05,
"loss": 0.1098,
"step": 1270
},
{
"epoch": 0.6077872744539411,
"grad_norm": 0.4737216171942805,
"learning_rate": 9.985974713878592e-05,
"loss": 0.1089,
"step": 1280
},
{
"epoch": 0.6125356125356125,
"grad_norm": 0.6374648190448973,
"learning_rate": 9.984706725574044e-05,
"loss": 0.1061,
"step": 1290
},
{
"epoch": 0.6172839506172839,
"grad_norm": 0.4868246558825449,
"learning_rate": 9.983383960401319e-05,
"loss": 0.1082,
"step": 1300
},
{
"epoch": 0.6220322886989553,
"grad_norm": 0.5879097726603935,
"learning_rate": 9.982006432896266e-05,
"loss": 0.1056,
"step": 1310
},
{
"epoch": 0.6267806267806267,
"grad_norm": 0.5978141469311284,
"learning_rate": 9.980574158196512e-05,
"loss": 0.1053,
"step": 1320
},
{
"epoch": 0.6315289648622981,
"grad_norm": 0.4840689426719305,
"learning_rate": 9.979087152041304e-05,
"loss": 0.1048,
"step": 1330
},
{
"epoch": 0.6362773029439696,
"grad_norm": 0.44372916176342936,
"learning_rate": 9.977545430771332e-05,
"loss": 0.1042,
"step": 1340
},
{
"epoch": 0.6410256410256411,
"grad_norm": 0.3983590518107677,
"learning_rate": 9.975949011328542e-05,
"loss": 0.103,
"step": 1350
},
{
"epoch": 0.6457739791073125,
"grad_norm": 0.48826981218174903,
"learning_rate": 9.97429791125597e-05,
"loss": 0.1003,
"step": 1360
},
{
"epoch": 0.6505223171889839,
"grad_norm": 0.38446333868718363,
"learning_rate": 9.972592148697528e-05,
"loss": 0.099,
"step": 1370
},
{
"epoch": 0.6552706552706553,
"grad_norm": 0.37335016451291064,
"learning_rate": 9.970831742397815e-05,
"loss": 0.1004,
"step": 1380
},
{
"epoch": 0.6600189933523267,
"grad_norm": 0.6247245585479205,
"learning_rate": 9.969016711701908e-05,
"loss": 0.099,
"step": 1390
},
{
"epoch": 0.6647673314339981,
"grad_norm": 0.5290448685371406,
"learning_rate": 9.967147076555157e-05,
"loss": 0.0986,
"step": 1400
},
{
"epoch": 0.6695156695156695,
"grad_norm": 0.3845466087849717,
"learning_rate": 9.965222857502951e-05,
"loss": 0.0977,
"step": 1410
},
{
"epoch": 0.674264007597341,
"grad_norm": 0.3970051520716528,
"learning_rate": 9.963244075690505e-05,
"loss": 0.0978,
"step": 1420
},
{
"epoch": 0.6790123456790124,
"grad_norm": 0.36705191170454815,
"learning_rate": 9.961210752862626e-05,
"loss": 0.0962,
"step": 1430
},
{
"epoch": 0.6837606837606838,
"grad_norm": 0.3467899794541735,
"learning_rate": 9.959122911363463e-05,
"loss": 0.0925,
"step": 1440
},
{
"epoch": 0.6885090218423552,
"grad_norm": 0.3946164858813157,
"learning_rate": 9.956980574136281e-05,
"loss": 0.094,
"step": 1450
},
{
"epoch": 0.6932573599240266,
"grad_norm": 0.37975090885201573,
"learning_rate": 9.954783764723188e-05,
"loss": 0.0901,
"step": 1460
},
{
"epoch": 0.698005698005698,
"grad_norm": 0.3396030408747898,
"learning_rate": 9.952532507264891e-05,
"loss": 0.0911,
"step": 1470
},
{
"epoch": 0.7027540360873694,
"grad_norm": 0.3054480674620304,
"learning_rate": 9.950226826500426e-05,
"loss": 0.0921,
"step": 1480
},
{
"epoch": 0.7075023741690408,
"grad_norm": 0.34630288479504856,
"learning_rate": 9.947866747766882e-05,
"loss": 0.0905,
"step": 1490
},
{
"epoch": 0.7122507122507122,
"grad_norm": 0.33464835523042297,
"learning_rate": 9.945452296999131e-05,
"loss": 0.09,
"step": 1500
},
{
"epoch": 0.7169990503323836,
"grad_norm": 0.38657639169391506,
"learning_rate": 9.942983500729536e-05,
"loss": 0.0888,
"step": 1510
},
{
"epoch": 0.7217473884140551,
"grad_norm": 0.35838146423892914,
"learning_rate": 9.940460386087667e-05,
"loss": 0.0875,
"step": 1520
},
{
"epoch": 0.7264957264957265,
"grad_norm": 0.37006832089600883,
"learning_rate": 9.937882980799986e-05,
"loss": 0.0874,
"step": 1530
},
{
"epoch": 0.7312440645773979,
"grad_norm": 0.32006810855437906,
"learning_rate": 9.935251313189564e-05,
"loss": 0.0887,
"step": 1540
},
{
"epoch": 0.7359924026590693,
"grad_norm": 0.30211240966376957,
"learning_rate": 9.93256541217576e-05,
"loss": 0.0887,
"step": 1550
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.33346488094391713,
"learning_rate": 9.929825307273898e-05,
"loss": 0.0854,
"step": 1560
},
{
"epoch": 0.7454890788224121,
"grad_norm": 0.3372048468233677,
"learning_rate": 9.927031028594951e-05,
"loss": 0.0863,
"step": 1570
},
{
"epoch": 0.7502374169040835,
"grad_norm": 0.33238326917217514,
"learning_rate": 9.924182606845204e-05,
"loss": 0.0835,
"step": 1580
},
{
"epoch": 0.7549857549857549,
"grad_norm": 0.31206255033046837,
"learning_rate": 9.921280073325922e-05,
"loss": 0.0831,
"step": 1590
},
{
"epoch": 0.7597340930674265,
"grad_norm": 0.3355517639004823,
"learning_rate": 9.918323459933005e-05,
"loss": 0.0817,
"step": 1600
},
{
"epoch": 0.7644824311490979,
"grad_norm": 0.3266312200524982,
"learning_rate": 9.915312799156632e-05,
"loss": 0.0811,
"step": 1610
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.330019004852785,
"learning_rate": 9.91224812408091e-05,
"loss": 0.0823,
"step": 1620
},
{
"epoch": 0.7739791073124407,
"grad_norm": 0.29046890730095737,
"learning_rate": 9.909129468383506e-05,
"loss": 0.083,
"step": 1630
},
{
"epoch": 0.7787274453941121,
"grad_norm": 0.34534651261515237,
"learning_rate": 9.905956866335286e-05,
"loss": 0.0817,
"step": 1640
},
{
"epoch": 0.7834757834757835,
"grad_norm": 0.29222249408318923,
"learning_rate": 9.90273035279992e-05,
"loss": 0.0805,
"step": 1650
},
{
"epoch": 0.7882241215574549,
"grad_norm": 0.28002849857725115,
"learning_rate": 9.899449963233523e-05,
"loss": 0.0784,
"step": 1660
},
{
"epoch": 0.7929724596391263,
"grad_norm": 0.2846412677880468,
"learning_rate": 9.896115733684244e-05,
"loss": 0.0782,
"step": 1670
},
{
"epoch": 0.7977207977207977,
"grad_norm": 0.3109998434771394,
"learning_rate": 9.892727700791881e-05,
"loss": 0.0771,
"step": 1680
},
{
"epoch": 0.8024691358024691,
"grad_norm": 0.3230796159493278,
"learning_rate": 9.889285901787482e-05,
"loss": 0.0793,
"step": 1690
},
{
"epoch": 0.8072174738841406,
"grad_norm": 0.29436142766859763,
"learning_rate": 9.885790374492921e-05,
"loss": 0.0792,
"step": 1700
},
{
"epoch": 0.811965811965812,
"grad_norm": 0.3569498671792048,
"learning_rate": 9.882241157320497e-05,
"loss": 0.0767,
"step": 1710
},
{
"epoch": 0.8167141500474834,
"grad_norm": 0.2833550559041438,
"learning_rate": 9.878638289272507e-05,
"loss": 0.0745,
"step": 1720
},
{
"epoch": 0.8214624881291548,
"grad_norm": 0.33829454925263097,
"learning_rate": 9.874981809940808e-05,
"loss": 0.073,
"step": 1730
},
{
"epoch": 0.8262108262108262,
"grad_norm": 0.28499602258280904,
"learning_rate": 9.871271759506404e-05,
"loss": 0.0753,
"step": 1740
},
{
"epoch": 0.8309591642924976,
"grad_norm": 0.32902378177594305,
"learning_rate": 9.86750817873898e-05,
"loss": 0.0751,
"step": 1750
},
{
"epoch": 0.835707502374169,
"grad_norm": 0.3140573755458756,
"learning_rate": 9.863691108996471e-05,
"loss": 0.072,
"step": 1760
},
{
"epoch": 0.8404558404558404,
"grad_norm": 0.27462153823258645,
"learning_rate": 9.859820592224599e-05,
"loss": 0.0722,
"step": 1770
},
{
"epoch": 0.8452041785375118,
"grad_norm": 0.3050765661268052,
"learning_rate": 9.855896670956415e-05,
"loss": 0.0735,
"step": 1780
},
{
"epoch": 0.8499525166191833,
"grad_norm": 0.3544579042112188,
"learning_rate": 9.851919388311832e-05,
"loss": 0.0718,
"step": 1790
},
{
"epoch": 0.8547008547008547,
"grad_norm": 0.30779596114918906,
"learning_rate": 9.84788878799715e-05,
"loss": 0.0702,
"step": 1800
},
{
"epoch": 0.8594491927825261,
"grad_norm": 0.33717756695679624,
"learning_rate": 9.843804914304577e-05,
"loss": 0.0715,
"step": 1810
},
{
"epoch": 0.8641975308641975,
"grad_norm": 0.2889442501586868,
"learning_rate": 9.839667812111739e-05,
"loss": 0.0683,
"step": 1820
},
{
"epoch": 0.8689458689458689,
"grad_norm": 0.24733007224448939,
"learning_rate": 9.835477526881189e-05,
"loss": 0.069,
"step": 1830
},
{
"epoch": 0.8736942070275404,
"grad_norm": 0.2553292846440847,
"learning_rate": 9.831234104659913e-05,
"loss": 0.0676,
"step": 1840
},
{
"epoch": 0.8784425451092118,
"grad_norm": 0.27150619713634844,
"learning_rate": 9.826937592078811e-05,
"loss": 0.0675,
"step": 1850
},
{
"epoch": 0.8831908831908832,
"grad_norm": 0.2682914038948634,
"learning_rate": 9.822588036352201e-05,
"loss": 0.0668,
"step": 1860
},
{
"epoch": 0.8879392212725546,
"grad_norm": 0.24296274436724274,
"learning_rate": 9.81818548527728e-05,
"loss": 0.066,
"step": 1870
},
{
"epoch": 0.8926875593542261,
"grad_norm": 0.2383012029491128,
"learning_rate": 9.81372998723362e-05,
"loss": 0.0665,
"step": 1880
},
{
"epoch": 0.8974358974358975,
"grad_norm": 0.3087223611188802,
"learning_rate": 9.809221591182623e-05,
"loss": 0.0656,
"step": 1890
},
{
"epoch": 0.9021842355175689,
"grad_norm": 0.2280733016964919,
"learning_rate": 9.804660346666984e-05,
"loss": 0.063,
"step": 1900
},
{
"epoch": 0.9069325735992403,
"grad_norm": 0.22712490669852592,
"learning_rate": 9.800046303810154e-05,
"loss": 0.0637,
"step": 1910
},
{
"epoch": 0.9116809116809117,
"grad_norm": 0.24259417294432375,
"learning_rate": 9.795379513315779e-05,
"loss": 0.0637,
"step": 1920
},
{
"epoch": 0.9164292497625831,
"grad_norm": 0.2317635099498286,
"learning_rate": 9.790660026467153e-05,
"loss": 0.0622,
"step": 1930
},
{
"epoch": 0.9211775878442545,
"grad_norm": 0.2353728305939909,
"learning_rate": 9.785887895126644e-05,
"loss": 0.0621,
"step": 1940
},
{
"epoch": 0.9259259259259259,
"grad_norm": 0.2410155300853678,
"learning_rate": 9.781063171735134e-05,
"loss": 0.0596,
"step": 1950
},
{
"epoch": 0.9306742640075973,
"grad_norm": 0.21946425391866467,
"learning_rate": 9.776185909311437e-05,
"loss": 0.0596,
"step": 1960
},
{
"epoch": 0.9354226020892688,
"grad_norm": 0.21958290933241775,
"learning_rate": 9.771256161451716e-05,
"loss": 0.0611,
"step": 1970
},
{
"epoch": 0.9401709401709402,
"grad_norm": 0.24813950559016779,
"learning_rate": 9.766273982328899e-05,
"loss": 0.059,
"step": 1980
},
{
"epoch": 0.9449192782526116,
"grad_norm": 0.21962092496193408,
"learning_rate": 9.761239426692077e-05,
"loss": 0.0583,
"step": 1990
},
{
"epoch": 0.949667616334283,
"grad_norm": 0.2623061013383692,
"learning_rate": 9.756152549865907e-05,
"loss": 0.0584,
"step": 2000
},
{
"epoch": 0.9544159544159544,
"grad_norm": 0.22957179268037364,
"learning_rate": 9.751013407750004e-05,
"loss": 0.0582,
"step": 2010
},
{
"epoch": 0.9591642924976258,
"grad_norm": 0.22871805734899975,
"learning_rate": 9.745822056818324e-05,
"loss": 0.058,
"step": 2020
},
{
"epoch": 0.9639126305792972,
"grad_norm": 0.19172396362719954,
"learning_rate": 9.740578554118545e-05,
"loss": 0.0561,
"step": 2030
},
{
"epoch": 0.9686609686609686,
"grad_norm": 0.22777955575125425,
"learning_rate": 9.735282957271438e-05,
"loss": 0.0553,
"step": 2040
},
{
"epoch": 0.97340930674264,
"grad_norm": 0.21497973610543095,
"learning_rate": 9.729935324470243e-05,
"loss": 0.0562,
"step": 2050
},
{
"epoch": 0.9781576448243114,
"grad_norm": 0.21598201771798728,
"learning_rate": 9.724535714480012e-05,
"loss": 0.0551,
"step": 2060
},
{
"epoch": 0.9829059829059829,
"grad_norm": 0.20229612217180784,
"learning_rate": 9.719084186636984e-05,
"loss": 0.0565,
"step": 2070
},
{
"epoch": 0.9876543209876543,
"grad_norm": 0.20724699505935157,
"learning_rate": 9.713580800847916e-05,
"loss": 0.0549,
"step": 2080
},
{
"epoch": 0.9924026590693258,
"grad_norm": 0.6767775831486229,
"learning_rate": 9.708025617589431e-05,
"loss": 0.0568,
"step": 2090
},
{
"epoch": 0.9971509971509972,
"grad_norm": 0.21686450756810657,
"learning_rate": 9.702418697907357e-05,
"loss": 0.0552,
"step": 2100
},
{
"epoch": 1.0018993352326686,
"grad_norm": 0.18424532132589896,
"learning_rate": 9.696760103416049e-05,
"loss": 0.0537,
"step": 2110
},
{
"epoch": 1.00664767331434,
"grad_norm": 0.19727671472990363,
"learning_rate": 9.69104989629772e-05,
"loss": 0.0538,
"step": 2120
},
{
"epoch": 1.0113960113960114,
"grad_norm": 0.1822771329871061,
"learning_rate": 9.685288139301749e-05,
"loss": 0.0514,
"step": 2130
},
{
"epoch": 1.0161443494776827,
"grad_norm": 0.18787745076355336,
"learning_rate": 9.679474895744002e-05,
"loss": 0.0508,
"step": 2140
},
{
"epoch": 1.0208926875593543,
"grad_norm": 0.18701764681951918,
"learning_rate": 9.673610229506123e-05,
"loss": 0.0511,
"step": 2150
},
{
"epoch": 1.0256410256410255,
"grad_norm": 0.1963026735387144,
"learning_rate": 9.667694205034842e-05,
"loss": 0.0506,
"step": 2160
},
{
"epoch": 1.030389363722697,
"grad_norm": 0.19068637753183512,
"learning_rate": 9.661726887341264e-05,
"loss": 0.0517,
"step": 2170
},
{
"epoch": 1.0351377018043684,
"grad_norm": 0.16444742484151312,
"learning_rate": 9.655708342000158e-05,
"loss": 0.049,
"step": 2180
},
{
"epoch": 1.03988603988604,
"grad_norm": 0.186883071491633,
"learning_rate": 9.649638635149225e-05,
"loss": 0.051,
"step": 2190
},
{
"epoch": 1.0446343779677112,
"grad_norm": 0.19252551114960448,
"learning_rate": 9.643517833488387e-05,
"loss": 0.0498,
"step": 2200
},
{
"epoch": 1.0493827160493827,
"grad_norm": 0.16023931110665707,
"learning_rate": 9.637346004279038e-05,
"loss": 0.0499,
"step": 2210
},
{
"epoch": 1.0541310541310542,
"grad_norm": 0.19082405791879545,
"learning_rate": 9.631123215343323e-05,
"loss": 0.0512,
"step": 2220
},
{
"epoch": 1.0588793922127255,
"grad_norm": 0.16410936327340184,
"learning_rate": 9.624849535063373e-05,
"loss": 0.0509,
"step": 2230
},
{
"epoch": 1.063627730294397,
"grad_norm": 0.1737068085608732,
"learning_rate": 9.61852503238057e-05,
"loss": 0.0502,
"step": 2240
},
{
"epoch": 1.0683760683760684,
"grad_norm": 0.15806510408346716,
"learning_rate": 9.612149776794776e-05,
"loss": 0.0499,
"step": 2250
},
{
"epoch": 1.0731244064577399,
"grad_norm": 0.16559771954661884,
"learning_rate": 9.605723838363586e-05,
"loss": 0.0494,
"step": 2260
},
{
"epoch": 1.0778727445394112,
"grad_norm": 0.17257140757782607,
"learning_rate": 9.599247287701539e-05,
"loss": 0.0502,
"step": 2270
},
{
"epoch": 1.0826210826210827,
"grad_norm": 0.17556574658197802,
"learning_rate": 9.592720195979353e-05,
"loss": 0.0507,
"step": 2280
},
{
"epoch": 1.087369420702754,
"grad_norm": 0.1724266236945596,
"learning_rate": 9.586142634923141e-05,
"loss": 0.0491,
"step": 2290
},
{
"epoch": 1.0921177587844255,
"grad_norm": 0.1608694370104181,
"learning_rate": 9.579514676813627e-05,
"loss": 0.0486,
"step": 2300
},
{
"epoch": 1.0968660968660968,
"grad_norm": 0.1751204527565512,
"learning_rate": 9.572836394485345e-05,
"loss": 0.0484,
"step": 2310
},
{
"epoch": 1.1016144349477683,
"grad_norm": 0.16683804563231602,
"learning_rate": 9.566107861325836e-05,
"loss": 0.049,
"step": 2320
},
{
"epoch": 1.1063627730294396,
"grad_norm": 0.17107220109180088,
"learning_rate": 9.559329151274854e-05,
"loss": 0.0476,
"step": 2330
},
{
"epoch": 1.1111111111111112,
"grad_norm": 0.17475586438187513,
"learning_rate": 9.552500338823543e-05,
"loss": 0.0486,
"step": 2340
},
{
"epoch": 1.1158594491927825,
"grad_norm": 0.1753327898998664,
"learning_rate": 9.54562149901362e-05,
"loss": 0.0479,
"step": 2350
},
{
"epoch": 1.120607787274454,
"grad_norm": 0.16428291924601113,
"learning_rate": 9.538692707436554e-05,
"loss": 0.0481,
"step": 2360
},
{
"epoch": 1.1253561253561253,
"grad_norm": 0.2313022547113298,
"learning_rate": 9.531714040232733e-05,
"loss": 0.0474,
"step": 2370
},
{
"epoch": 1.1301044634377968,
"grad_norm": 0.34311176244533276,
"learning_rate": 9.524685574090627e-05,
"loss": 0.0486,
"step": 2380
},
{
"epoch": 1.134852801519468,
"grad_norm": 0.47480966268475455,
"learning_rate": 9.517607386245948e-05,
"loss": 0.0573,
"step": 2390
},
{
"epoch": 1.1396011396011396,
"grad_norm": 0.21414254549794035,
"learning_rate": 9.510479554480793e-05,
"loss": 0.0492,
"step": 2400
},
{
"epoch": 1.144349477682811,
"grad_norm": 0.16899420694724102,
"learning_rate": 9.503302157122804e-05,
"loss": 0.0481,
"step": 2410
},
{
"epoch": 1.1490978157644824,
"grad_norm": 0.17715352917570737,
"learning_rate": 9.496075273044293e-05,
"loss": 0.0481,
"step": 2420
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.15868465146826358,
"learning_rate": 9.488798981661379e-05,
"loss": 0.0476,
"step": 2430
},
{
"epoch": 1.1585944919278253,
"grad_norm": 0.16031610656726267,
"learning_rate": 9.481473362933122e-05,
"loss": 0.0464,
"step": 2440
},
{
"epoch": 1.1633428300094968,
"grad_norm": 0.28020559262037614,
"learning_rate": 9.474098497360637e-05,
"loss": 0.0488,
"step": 2450
},
{
"epoch": 1.168091168091168,
"grad_norm": 0.17211870369113508,
"learning_rate": 9.466674465986211e-05,
"loss": 0.0469,
"step": 2460
},
{
"epoch": 1.1728395061728394,
"grad_norm": 0.18477475254247877,
"learning_rate": 9.459201350392418e-05,
"loss": 0.0474,
"step": 2470
},
{
"epoch": 1.177587844254511,
"grad_norm": 0.16165908485554836,
"learning_rate": 9.451679232701212e-05,
"loss": 0.0467,
"step": 2480
},
{
"epoch": 1.1823361823361824,
"grad_norm": 0.15100176044362995,
"learning_rate": 9.444108195573031e-05,
"loss": 0.0468,
"step": 2490
},
{
"epoch": 1.1870845204178537,
"grad_norm": 0.154194460230295,
"learning_rate": 9.436488322205892e-05,
"loss": 0.0471,
"step": 2500
},
{
"epoch": 1.1918328584995252,
"grad_norm": 0.19026130416229434,
"learning_rate": 9.42881969633447e-05,
"loss": 0.0466,
"step": 2510
},
{
"epoch": 1.1965811965811965,
"grad_norm": 0.15084451107490007,
"learning_rate": 9.42110240222918e-05,
"loss": 0.0457,
"step": 2520
},
{
"epoch": 1.201329534662868,
"grad_norm": 0.15834890789392483,
"learning_rate": 9.413336524695253e-05,
"loss": 0.0462,
"step": 2530
},
{
"epoch": 1.2060778727445394,
"grad_norm": 0.14908511792461682,
"learning_rate": 9.405522149071801e-05,
"loss": 0.0453,
"step": 2540
},
{
"epoch": 1.210826210826211,
"grad_norm": 0.1597318119838342,
"learning_rate": 9.397659361230883e-05,
"loss": 0.0454,
"step": 2550
},
{
"epoch": 1.2155745489078822,
"grad_norm": 0.15019295679489722,
"learning_rate": 9.389748247576557e-05,
"loss": 0.0439,
"step": 2560
},
{
"epoch": 1.2203228869895537,
"grad_norm": 0.15261097407460317,
"learning_rate": 9.381788895043933e-05,
"loss": 0.0461,
"step": 2570
},
{
"epoch": 1.225071225071225,
"grad_norm": 0.15893916466820565,
"learning_rate": 9.373781391098218e-05,
"loss": 0.0438,
"step": 2580
},
{
"epoch": 1.2298195631528965,
"grad_norm": 0.14962284104862464,
"learning_rate": 9.365725823733753e-05,
"loss": 0.0449,
"step": 2590
},
{
"epoch": 1.2345679012345678,
"grad_norm": 0.17548926124525194,
"learning_rate": 9.35762228147305e-05,
"loss": 0.0448,
"step": 2600
},
{
"epoch": 1.2393162393162394,
"grad_norm": 0.1313270663168934,
"learning_rate": 9.349470853365815e-05,
"loss": 0.0433,
"step": 2610
},
{
"epoch": 1.2440645773979107,
"grad_norm": 0.14306097883598046,
"learning_rate": 9.341271628987968e-05,
"loss": 0.0448,
"step": 2620
},
{
"epoch": 1.2488129154795822,
"grad_norm": 0.1536280576738689,
"learning_rate": 9.333024698440667e-05,
"loss": 0.0451,
"step": 2630
},
{
"epoch": 1.2535612535612537,
"grad_norm": 0.1853127006706868,
"learning_rate": 9.324730152349306e-05,
"loss": 0.045,
"step": 2640
},
{
"epoch": 1.258309591642925,
"grad_norm": 0.15385856853358393,
"learning_rate": 9.31638808186253e-05,
"loss": 0.0437,
"step": 2650
},
{
"epoch": 1.2630579297245963,
"grad_norm": 0.1795815172833766,
"learning_rate": 9.307998578651228e-05,
"loss": 0.0449,
"step": 2660
},
{
"epoch": 1.2678062678062678,
"grad_norm": 0.15671626347483983,
"learning_rate": 9.299561734907528e-05,
"loss": 0.045,
"step": 2670
},
{
"epoch": 1.2725546058879393,
"grad_norm": 0.1486057863863386,
"learning_rate": 9.291077643343776e-05,
"loss": 0.0456,
"step": 2680
},
{
"epoch": 1.2773029439696106,
"grad_norm": 0.15986711921073898,
"learning_rate": 9.282546397191531e-05,
"loss": 0.0443,
"step": 2690
},
{
"epoch": 1.282051282051282,
"grad_norm": 0.15225612959552576,
"learning_rate": 9.273968090200533e-05,
"loss": 0.0436,
"step": 2700
},
{
"epoch": 1.2867996201329535,
"grad_norm": 0.15335400483443493,
"learning_rate": 9.265342816637666e-05,
"loss": 0.0438,
"step": 2710
},
{
"epoch": 1.291547958214625,
"grad_norm": 0.14783268691936724,
"learning_rate": 9.256670671285934e-05,
"loss": 0.0426,
"step": 2720
},
{
"epoch": 1.2962962962962963,
"grad_norm": 0.13749880062830538,
"learning_rate": 9.247951749443413e-05,
"loss": 0.0428,
"step": 2730
},
{
"epoch": 1.3010446343779676,
"grad_norm": 0.142974339939963,
"learning_rate": 9.239186146922208e-05,
"loss": 0.0426,
"step": 2740
},
{
"epoch": 1.305792972459639,
"grad_norm": 0.1558390700312084,
"learning_rate": 9.230373960047394e-05,
"loss": 0.0439,
"step": 2750
},
{
"epoch": 1.3105413105413106,
"grad_norm": 0.14074567305868227,
"learning_rate": 9.22151528565596e-05,
"loss": 0.0425,
"step": 2760
},
{
"epoch": 1.315289648622982,
"grad_norm": 0.14048888956507766,
"learning_rate": 9.212610221095748e-05,
"loss": 0.0439,
"step": 2770
},
{
"epoch": 1.3200379867046534,
"grad_norm": 0.14077566870112865,
"learning_rate": 9.203658864224382e-05,
"loss": 0.0432,
"step": 2780
},
{
"epoch": 1.3247863247863247,
"grad_norm": 0.14802349774572246,
"learning_rate": 9.194661313408187e-05,
"loss": 0.0435,
"step": 2790
},
{
"epoch": 1.3295346628679963,
"grad_norm": 0.14683256983293036,
"learning_rate": 9.185617667521117e-05,
"loss": 0.0422,
"step": 2800
},
{
"epoch": 1.3342830009496676,
"grad_norm": 0.14259478466503417,
"learning_rate": 9.176528025943663e-05,
"loss": 0.0428,
"step": 2810
},
{
"epoch": 1.339031339031339,
"grad_norm": 0.139804063457447,
"learning_rate": 9.167392488561758e-05,
"loss": 0.0423,
"step": 2820
},
{
"epoch": 1.3437796771130104,
"grad_norm": 0.14131105727541277,
"learning_rate": 9.158211155765692e-05,
"loss": 0.0422,
"step": 2830
},
{
"epoch": 1.348528015194682,
"grad_norm": 0.13421386725946377,
"learning_rate": 9.14898412844899e-05,
"loss": 0.0422,
"step": 2840
},
{
"epoch": 1.3532763532763532,
"grad_norm": 0.13452410037650797,
"learning_rate": 9.13971150800732e-05,
"loss": 0.0418,
"step": 2850
},
{
"epoch": 1.3580246913580247,
"grad_norm": 0.1328596789760814,
"learning_rate": 9.130393396337368e-05,
"loss": 0.0434,
"step": 2860
},
{
"epoch": 1.362773029439696,
"grad_norm": 0.12745313152844992,
"learning_rate": 9.121029895835731e-05,
"loss": 0.042,
"step": 2870
},
{
"epoch": 1.3675213675213675,
"grad_norm": 0.15031747384794045,
"learning_rate": 9.111621109397769e-05,
"loss": 0.0431,
"step": 2880
},
{
"epoch": 1.3722697056030388,
"grad_norm": 0.12834130517579306,
"learning_rate": 9.102167140416503e-05,
"loss": 0.0404,
"step": 2890
},
{
"epoch": 1.3770180436847104,
"grad_norm": 0.17211306829632797,
"learning_rate": 9.092668092781454e-05,
"loss": 0.0422,
"step": 2900
},
{
"epoch": 1.381766381766382,
"grad_norm": 0.15446206976262547,
"learning_rate": 9.083124070877516e-05,
"loss": 0.0418,
"step": 2910
},
{
"epoch": 1.3865147198480532,
"grad_norm": 0.13007692638701887,
"learning_rate": 9.073535179583806e-05,
"loss": 0.0421,
"step": 2920
},
{
"epoch": 1.3912630579297245,
"grad_norm": 0.14163306664079398,
"learning_rate": 9.063901524272505e-05,
"loss": 0.0412,
"step": 2930
},
{
"epoch": 1.396011396011396,
"grad_norm": 0.1362673410954657,
"learning_rate": 9.054223210807709e-05,
"loss": 0.0417,
"step": 2940
},
{
"epoch": 1.4007597340930675,
"grad_norm": 0.13133944911035692,
"learning_rate": 9.044500345544262e-05,
"loss": 0.0413,
"step": 2950
},
{
"epoch": 1.4055080721747388,
"grad_norm": 0.15026287597856436,
"learning_rate": 9.034733035326584e-05,
"loss": 0.0417,
"step": 2960
},
{
"epoch": 1.4102564102564101,
"grad_norm": 0.12434037545267754,
"learning_rate": 9.024921387487504e-05,
"loss": 0.0419,
"step": 2970
},
{
"epoch": 1.4150047483380817,
"grad_norm": 0.1337108813287562,
"learning_rate": 9.015065509847075e-05,
"loss": 0.0405,
"step": 2980
},
{
"epoch": 1.4197530864197532,
"grad_norm": 0.16780084314616397,
"learning_rate": 9.005165510711387e-05,
"loss": 0.042,
"step": 2990
},
{
"epoch": 1.4245014245014245,
"grad_norm": 0.12943128836561044,
"learning_rate": 8.995221498871387e-05,
"loss": 0.0415,
"step": 3000
},
{
"epoch": 1.429249762583096,
"grad_norm": 0.12056336517753806,
"learning_rate": 8.98523358360167e-05,
"loss": 0.0414,
"step": 3010
},
{
"epoch": 1.4339981006647673,
"grad_norm": 0.11452447941427116,
"learning_rate": 8.975201874659292e-05,
"loss": 0.0407,
"step": 3020
},
{
"epoch": 1.4387464387464388,
"grad_norm": 0.14313287963976554,
"learning_rate": 8.965126482282549e-05,
"loss": 0.0405,
"step": 3030
},
{
"epoch": 1.4434947768281101,
"grad_norm": 0.147901578052728,
"learning_rate": 8.955007517189784e-05,
"loss": 0.0405,
"step": 3040
},
{
"epoch": 1.4482431149097816,
"grad_norm": 0.13754056095708528,
"learning_rate": 8.944845090578149e-05,
"loss": 0.0423,
"step": 3050
},
{
"epoch": 1.452991452991453,
"grad_norm": 0.1416260971349369,
"learning_rate": 8.934639314122401e-05,
"loss": 0.0428,
"step": 3060
},
{
"epoch": 1.4577397910731245,
"grad_norm": 0.13748552740676037,
"learning_rate": 8.924390299973666e-05,
"loss": 0.0406,
"step": 3070
},
{
"epoch": 1.4624881291547958,
"grad_norm": 0.14858094548530582,
"learning_rate": 8.91409816075821e-05,
"loss": 0.0419,
"step": 3080
},
{
"epoch": 1.4672364672364673,
"grad_norm": 0.2078682391747803,
"learning_rate": 8.903763009576195e-05,
"loss": 0.0408,
"step": 3090
},
{
"epoch": 1.4719848053181386,
"grad_norm": 0.17586181038691126,
"learning_rate": 8.893384960000445e-05,
"loss": 0.0409,
"step": 3100
},
{
"epoch": 1.47673314339981,
"grad_norm": 0.1384691069553198,
"learning_rate": 8.882964126075192e-05,
"loss": 0.041,
"step": 3110
},
{
"epoch": 1.4814814814814814,
"grad_norm": 0.12687085286060168,
"learning_rate": 8.872500622314829e-05,
"loss": 0.0424,
"step": 3120
},
{
"epoch": 1.486229819563153,
"grad_norm": 0.12678358905787643,
"learning_rate": 8.861994563702638e-05,
"loss": 0.0415,
"step": 3130
},
{
"epoch": 1.4909781576448244,
"grad_norm": 0.13797982217172405,
"learning_rate": 8.851446065689544e-05,
"loss": 0.0403,
"step": 3140
},
{
"epoch": 1.4957264957264957,
"grad_norm": 0.11933174077174406,
"learning_rate": 8.840855244192836e-05,
"loss": 0.0392,
"step": 3150
},
{
"epoch": 1.500474833808167,
"grad_norm": 0.14729441556965195,
"learning_rate": 8.83022221559489e-05,
"loss": 0.0413,
"step": 3160
},
{
"epoch": 1.5052231718898386,
"grad_norm": 0.12661651711344382,
"learning_rate": 8.819547096741904e-05,
"loss": 0.0408,
"step": 3170
},
{
"epoch": 1.50997150997151,
"grad_norm": 0.13220919794222843,
"learning_rate": 8.8088300049426e-05,
"loss": 0.0409,
"step": 3180
},
{
"epoch": 1.5147198480531814,
"grad_norm": 0.11201350262678006,
"learning_rate": 8.798071057966935e-05,
"loss": 0.039,
"step": 3190
},
{
"epoch": 1.5194681861348527,
"grad_norm": 0.11971895630564829,
"learning_rate": 8.787270374044824e-05,
"loss": 0.0402,
"step": 3200
},
{
"epoch": 1.5242165242165242,
"grad_norm": 0.13843777811955457,
"learning_rate": 8.776428071864819e-05,
"loss": 0.0408,
"step": 3210
},
{
"epoch": 1.5289648622981957,
"grad_norm": 0.12909148907023082,
"learning_rate": 8.76554427057282e-05,
"loss": 0.0393,
"step": 3220
},
{
"epoch": 1.533713200379867,
"grad_norm": 0.1063666674737681,
"learning_rate": 8.754619089770755e-05,
"loss": 0.0393,
"step": 3230
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.12848957062946695,
"learning_rate": 8.743652649515275e-05,
"loss": 0.0403,
"step": 3240
},
{
"epoch": 1.5432098765432098,
"grad_norm": 0.14111602954280902,
"learning_rate": 8.73264507031643e-05,
"loss": 0.0404,
"step": 3250
},
{
"epoch": 1.5479582146248814,
"grad_norm": 0.13459595368717894,
"learning_rate": 8.721596473136342e-05,
"loss": 0.0409,
"step": 3260
},
{
"epoch": 1.5527065527065527,
"grad_norm": 0.11289198865999905,
"learning_rate": 8.710506979387882e-05,
"loss": 0.0404,
"step": 3270
},
{
"epoch": 1.557454890788224,
"grad_norm": 0.12660606317440712,
"learning_rate": 8.699376710933336e-05,
"loss": 0.0409,
"step": 3280
},
{
"epoch": 1.5622032288698955,
"grad_norm": 0.1319331902644405,
"learning_rate": 8.688205790083053e-05,
"loss": 0.04,
"step": 3290
},
{
"epoch": 1.566951566951567,
"grad_norm": 0.11629334096382375,
"learning_rate": 8.676994339594118e-05,
"loss": 0.0399,
"step": 3300
},
{
"epoch": 1.5716999050332383,
"grad_norm": 0.1304610460953442,
"learning_rate": 8.665742482668996e-05,
"loss": 0.0407,
"step": 3310
},
{
"epoch": 1.5764482431149098,
"grad_norm": 0.13095650553618451,
"learning_rate": 8.654450342954175e-05,
"loss": 0.0398,
"step": 3320
},
{
"epoch": 1.5811965811965814,
"grad_norm": 0.12150689642437001,
"learning_rate": 8.643118044538808e-05,
"loss": 0.04,
"step": 3330
},
{
"epoch": 1.5859449192782527,
"grad_norm": 0.12939681786704482,
"learning_rate": 8.631745711953357e-05,
"loss": 0.0399,
"step": 3340
},
{
"epoch": 1.590693257359924,
"grad_norm": 0.1308032630903187,
"learning_rate": 8.620333470168214e-05,
"loss": 0.0399,
"step": 3350
},
{
"epoch": 1.5954415954415955,
"grad_norm": 0.11188863215539778,
"learning_rate": 8.608881444592337e-05,
"loss": 0.0389,
"step": 3360
},
{
"epoch": 1.600189933523267,
"grad_norm": 0.1131891510168041,
"learning_rate": 8.597389761071862e-05,
"loss": 0.0384,
"step": 3370
},
{
"epoch": 1.6049382716049383,
"grad_norm": 0.12773065390668614,
"learning_rate": 8.58585854588873e-05,
"loss": 0.0392,
"step": 3380
},
{
"epoch": 1.6096866096866096,
"grad_norm": 0.11544216543657933,
"learning_rate": 8.574287925759294e-05,
"loss": 0.038,
"step": 3390
},
{
"epoch": 1.6144349477682811,
"grad_norm": 0.11723752131285087,
"learning_rate": 8.562678027832925e-05,
"loss": 0.0388,
"step": 3400
},
{
"epoch": 1.6191832858499526,
"grad_norm": 0.11814605057456463,
"learning_rate": 8.551028979690621e-05,
"loss": 0.0388,
"step": 3410
},
{
"epoch": 1.623931623931624,
"grad_norm": 0.11528518071912393,
"learning_rate": 8.539340909343597e-05,
"loss": 0.039,
"step": 3420
},
{
"epoch": 1.6286799620132952,
"grad_norm": 0.11445104203868725,
"learning_rate": 8.527613945231885e-05,
"loss": 0.0388,
"step": 3430
},
{
"epoch": 1.6334283000949668,
"grad_norm": 0.11465099700540989,
"learning_rate": 8.51584821622292e-05,
"loss": 0.039,
"step": 3440
},
{
"epoch": 1.6381766381766383,
"grad_norm": 0.12131533819918404,
"learning_rate": 8.504043851610122e-05,
"loss": 0.0395,
"step": 3450
},
{
"epoch": 1.6429249762583096,
"grad_norm": 0.11109113049598257,
"learning_rate": 8.492200981111482e-05,
"loss": 0.0383,
"step": 3460
},
{
"epoch": 1.6476733143399809,
"grad_norm": 0.12358144612849818,
"learning_rate": 8.480319734868124e-05,
"loss": 0.0391,
"step": 3470
},
{
"epoch": 1.6524216524216524,
"grad_norm": 0.11465357658645761,
"learning_rate": 8.468400243442888e-05,
"loss": 0.0388,
"step": 3480
},
{
"epoch": 1.657169990503324,
"grad_norm": 0.11121754628942188,
"learning_rate": 8.456442637818889e-05,
"loss": 0.0376,
"step": 3490
},
{
"epoch": 1.6619183285849952,
"grad_norm": 0.11099942262003033,
"learning_rate": 8.444447049398076e-05,
"loss": 0.0386,
"step": 3500
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.10992807898019825,
"learning_rate": 8.432413609999794e-05,
"loss": 0.0384,
"step": 3510
},
{
"epoch": 1.671415004748338,
"grad_norm": 0.11376652817952089,
"learning_rate": 8.420342451859329e-05,
"loss": 0.0384,
"step": 3520
},
{
"epoch": 1.6761633428300096,
"grad_norm": 0.1169343499459042,
"learning_rate": 8.408233707626458e-05,
"loss": 0.0393,
"step": 3530
},
{
"epoch": 1.6809116809116809,
"grad_norm": 0.11540096113503476,
"learning_rate": 8.39608751036399e-05,
"loss": 0.0387,
"step": 3540
},
{
"epoch": 1.6856600189933522,
"grad_norm": 0.1245899502856239,
"learning_rate": 8.383903993546311e-05,
"loss": 0.0394,
"step": 3550
},
{
"epoch": 1.690408357075024,
"grad_norm": 0.10990746395253856,
"learning_rate": 8.371683291057902e-05,
"loss": 0.0382,
"step": 3560
},
{
"epoch": 1.6951566951566952,
"grad_norm": 0.12435245024774425,
"learning_rate": 8.359425537191883e-05,
"loss": 0.0386,
"step": 3570
},
{
"epoch": 1.6999050332383665,
"grad_norm": 0.1204397044160426,
"learning_rate": 8.347130866648528e-05,
"loss": 0.0387,
"step": 3580
},
{
"epoch": 1.704653371320038,
"grad_norm": 0.12961047088204164,
"learning_rate": 8.334799414533791e-05,
"loss": 0.0389,
"step": 3590
},
{
"epoch": 1.7094017094017095,
"grad_norm": 0.2717406976809826,
"learning_rate": 8.322431316357815e-05,
"loss": 0.0368,
"step": 3600
}
],
"logging_steps": 10,
"max_steps": 10530,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 1800,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 325906930335744.0,
"train_batch_size": 128,
"trial_name": null,
"trial_params": null
}