MultimodalAgent's picture
Upload ensemble_student_6/trainer_state.json with huggingface_hub
f005d61 verified
Raw
History Blame Contribute Delete
184 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 5.0,
"eval_steps": 500,
"global_step": 10530,
"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
},
{
"epoch": 1.7141500474833808,
"grad_norm": 0.11996633668866984,
"learning_rate": 8.310026708033447e-05,
"loss": 0.0375,
"step": 3610
},
{
"epoch": 1.7188983855650521,
"grad_norm": 0.11908177213133414,
"learning_rate": 8.297585725874743e-05,
"loss": 0.0381,
"step": 3620
},
{
"epoch": 1.7236467236467237,
"grad_norm": 0.10984963900518893,
"learning_rate": 8.285108506595471e-05,
"loss": 0.039,
"step": 3630
},
{
"epoch": 1.7283950617283952,
"grad_norm": 0.11552120343595351,
"learning_rate": 8.272595187307612e-05,
"loss": 0.0386,
"step": 3640
},
{
"epoch": 1.7331433998100665,
"grad_norm": 0.12119155876912908,
"learning_rate": 8.260045905519842e-05,
"loss": 0.0398,
"step": 3650
},
{
"epoch": 1.7378917378917378,
"grad_norm": 0.1268497973623033,
"learning_rate": 8.247460799136032e-05,
"loss": 0.0381,
"step": 3660
},
{
"epoch": 1.7426400759734093,
"grad_norm": 0.11245476654603592,
"learning_rate": 8.234840006453734e-05,
"loss": 0.0369,
"step": 3670
},
{
"epoch": 1.7473884140550808,
"grad_norm": 0.1142435347880095,
"learning_rate": 8.222183666162647e-05,
"loss": 0.0381,
"step": 3680
},
{
"epoch": 1.7521367521367521,
"grad_norm": 0.10605581380310948,
"learning_rate": 8.209491917343104e-05,
"loss": 0.0374,
"step": 3690
},
{
"epoch": 1.7568850902184234,
"grad_norm": 0.10689347248100933,
"learning_rate": 8.196764899464551e-05,
"loss": 0.0385,
"step": 3700
},
{
"epoch": 1.761633428300095,
"grad_norm": 0.11117045887105995,
"learning_rate": 8.184002752383993e-05,
"loss": 0.0381,
"step": 3710
},
{
"epoch": 1.7663817663817665,
"grad_norm": 0.12160356170245705,
"learning_rate": 8.171205616344478e-05,
"loss": 0.0385,
"step": 3720
},
{
"epoch": 1.7711301044634378,
"grad_norm": 0.12343589149887738,
"learning_rate": 8.158373631973542e-05,
"loss": 0.0378,
"step": 3730
},
{
"epoch": 1.775878442545109,
"grad_norm": 0.11698246266599253,
"learning_rate": 8.14550694028167e-05,
"loss": 0.038,
"step": 3740
},
{
"epoch": 1.7806267806267806,
"grad_norm": 0.1159715277970788,
"learning_rate": 8.13260568266075e-05,
"loss": 0.0376,
"step": 3750
},
{
"epoch": 1.7853751187084521,
"grad_norm": 0.1491223321009647,
"learning_rate": 8.119670000882506e-05,
"loss": 0.039,
"step": 3760
},
{
"epoch": 1.7901234567901234,
"grad_norm": 0.11800350698542214,
"learning_rate": 8.106700037096955e-05,
"loss": 0.0381,
"step": 3770
},
{
"epoch": 1.7948717948717947,
"grad_norm": 0.10435090457724086,
"learning_rate": 8.093695933830838e-05,
"loss": 0.0376,
"step": 3780
},
{
"epoch": 1.7996201329534662,
"grad_norm": 0.11120453621314899,
"learning_rate": 8.080657833986047e-05,
"loss": 0.0371,
"step": 3790
},
{
"epoch": 1.8043684710351378,
"grad_norm": 0.11078807415780473,
"learning_rate": 8.067585880838073e-05,
"loss": 0.0387,
"step": 3800
},
{
"epoch": 1.809116809116809,
"grad_norm": 0.0987290066310742,
"learning_rate": 8.054480218034413e-05,
"loss": 0.0378,
"step": 3810
},
{
"epoch": 1.8138651471984806,
"grad_norm": 0.10647185447746484,
"learning_rate": 8.041340989593003e-05,
"loss": 0.0376,
"step": 3820
},
{
"epoch": 1.818613485280152,
"grad_norm": 0.10453000908812413,
"learning_rate": 8.028168339900625e-05,
"loss": 0.0368,
"step": 3830
},
{
"epoch": 1.8233618233618234,
"grad_norm": 0.12268184512057981,
"learning_rate": 8.014962413711334e-05,
"loss": 0.0373,
"step": 3840
},
{
"epoch": 1.8281101614434947,
"grad_norm": 0.1025601404234581,
"learning_rate": 8.001723356144855e-05,
"loss": 0.0373,
"step": 3850
},
{
"epoch": 1.8328584995251662,
"grad_norm": 0.1117681866279554,
"learning_rate": 7.988451312684997e-05,
"loss": 0.037,
"step": 3860
},
{
"epoch": 1.8376068376068377,
"grad_norm": 0.11537840010508768,
"learning_rate": 7.975146429178047e-05,
"loss": 0.0384,
"step": 3870
},
{
"epoch": 1.842355175688509,
"grad_norm": 0.11246206860428429,
"learning_rate": 7.961808851831172e-05,
"loss": 0.0372,
"step": 3880
},
{
"epoch": 1.8471035137701803,
"grad_norm": 0.12374644243198947,
"learning_rate": 7.948438727210814e-05,
"loss": 0.0368,
"step": 3890
},
{
"epoch": 1.8518518518518519,
"grad_norm": 0.10321284687124542,
"learning_rate": 7.935036202241071e-05,
"loss": 0.0369,
"step": 3900
},
{
"epoch": 1.8566001899335234,
"grad_norm": 0.10499421079179089,
"learning_rate": 7.921601424202091e-05,
"loss": 0.0372,
"step": 3910
},
{
"epoch": 1.8613485280151947,
"grad_norm": 0.12177770167544143,
"learning_rate": 7.908134540728452e-05,
"loss": 0.0381,
"step": 3920
},
{
"epoch": 1.866096866096866,
"grad_norm": 0.1036235864939653,
"learning_rate": 7.894635699807533e-05,
"loss": 0.0385,
"step": 3930
},
{
"epoch": 1.8708452041785375,
"grad_norm": 0.10474580024238409,
"learning_rate": 7.881105049777901e-05,
"loss": 0.0363,
"step": 3940
},
{
"epoch": 1.875593542260209,
"grad_norm": 0.1038424143983712,
"learning_rate": 7.867542739327667e-05,
"loss": 0.036,
"step": 3950
},
{
"epoch": 1.8803418803418803,
"grad_norm": 0.1028952940875255,
"learning_rate": 7.853948917492857e-05,
"loss": 0.037,
"step": 3960
},
{
"epoch": 1.8850902184235516,
"grad_norm": 0.11085500954253853,
"learning_rate": 7.840323733655778e-05,
"loss": 0.0374,
"step": 3970
},
{
"epoch": 1.8898385565052231,
"grad_norm": 0.10665992145604658,
"learning_rate": 7.826667337543377e-05,
"loss": 0.0377,
"step": 3980
},
{
"epoch": 1.8945868945868947,
"grad_norm": 0.10345709687168737,
"learning_rate": 7.812979879225581e-05,
"loss": 0.0389,
"step": 3990
},
{
"epoch": 1.899335232668566,
"grad_norm": 0.10263625641754884,
"learning_rate": 7.799261509113671e-05,
"loss": 0.0359,
"step": 4000
},
{
"epoch": 1.9040835707502373,
"grad_norm": 0.11182345391247611,
"learning_rate": 7.78551237795861e-05,
"loss": 0.0364,
"step": 4010
},
{
"epoch": 1.9088319088319088,
"grad_norm": 0.10368991540522884,
"learning_rate": 7.771732636849399e-05,
"loss": 0.0377,
"step": 4020
},
{
"epoch": 1.9135802469135803,
"grad_norm": 0.10012895727039359,
"learning_rate": 7.757922437211405e-05,
"loss": 0.0358,
"step": 4030
},
{
"epoch": 1.9183285849952516,
"grad_norm": 0.09582579402506071,
"learning_rate": 7.744081930804707e-05,
"loss": 0.0364,
"step": 4040
},
{
"epoch": 1.9230769230769231,
"grad_norm": 0.25226447034194854,
"learning_rate": 7.730211269722427e-05,
"loss": 0.0371,
"step": 4050
},
{
"epoch": 1.9278252611585946,
"grad_norm": 0.11216657736385678,
"learning_rate": 7.716310606389052e-05,
"loss": 0.0382,
"step": 4060
},
{
"epoch": 1.932573599240266,
"grad_norm": 0.12427747638208547,
"learning_rate": 7.702380093558765e-05,
"loss": 0.0368,
"step": 4070
},
{
"epoch": 1.9373219373219372,
"grad_norm": 0.10905238178097779,
"learning_rate": 7.688419884313764e-05,
"loss": 0.0373,
"step": 4080
},
{
"epoch": 1.9420702754036088,
"grad_norm": 0.11017336578064867,
"learning_rate": 7.674430132062585e-05,
"loss": 0.037,
"step": 4090
},
{
"epoch": 1.9468186134852803,
"grad_norm": 0.10804077622778656,
"learning_rate": 7.660410990538403e-05,
"loss": 0.0364,
"step": 4100
},
{
"epoch": 1.9515669515669516,
"grad_norm": 0.11539015778626042,
"learning_rate": 7.646362613797356e-05,
"loss": 0.0371,
"step": 4110
},
{
"epoch": 1.956315289648623,
"grad_norm": 0.13775150227845018,
"learning_rate": 7.63228515621685e-05,
"loss": 0.0367,
"step": 4120
},
{
"epoch": 1.9610636277302944,
"grad_norm": 0.1027344900726605,
"learning_rate": 7.618178772493856e-05,
"loss": 0.0363,
"step": 4130
},
{
"epoch": 1.965811965811966,
"grad_norm": 0.1355918872467284,
"learning_rate": 7.604043617643215e-05,
"loss": 0.0365,
"step": 4140
},
{
"epoch": 1.9705603038936372,
"grad_norm": 0.10684503155283133,
"learning_rate": 7.589879846995934e-05,
"loss": 0.0369,
"step": 4150
},
{
"epoch": 1.9753086419753085,
"grad_norm": 0.11182215944890649,
"learning_rate": 7.575687616197477e-05,
"loss": 0.0375,
"step": 4160
},
{
"epoch": 1.98005698005698,
"grad_norm": 0.10155835926621704,
"learning_rate": 7.561467081206061e-05,
"loss": 0.0366,
"step": 4170
},
{
"epoch": 1.9848053181386516,
"grad_norm": 0.09739652028620334,
"learning_rate": 7.547218398290932e-05,
"loss": 0.0361,
"step": 4180
},
{
"epoch": 1.9895536562203229,
"grad_norm": 0.10312974122855112,
"learning_rate": 7.532941724030654e-05,
"loss": 0.0365,
"step": 4190
},
{
"epoch": 1.9943019943019942,
"grad_norm": 0.10343697863574187,
"learning_rate": 7.518637215311388e-05,
"loss": 0.0355,
"step": 4200
},
{
"epoch": 1.9990503323836657,
"grad_norm": 0.10107869023785024,
"learning_rate": 7.504305029325172e-05,
"loss": 0.036,
"step": 4210
},
{
"epoch": 2.003798670465337,
"grad_norm": 0.1048290499094816,
"learning_rate": 7.489945323568178e-05,
"loss": 0.0354,
"step": 4220
},
{
"epoch": 2.0085470085470085,
"grad_norm": 0.09937099335912418,
"learning_rate": 7.475558255839006e-05,
"loss": 0.0354,
"step": 4230
},
{
"epoch": 2.01329534662868,
"grad_norm": 0.08516538471288625,
"learning_rate": 7.461143984236924e-05,
"loss": 0.0337,
"step": 4240
},
{
"epoch": 2.0180436847103516,
"grad_norm": 0.09491626011942883,
"learning_rate": 7.446702667160154e-05,
"loss": 0.0351,
"step": 4250
},
{
"epoch": 2.022792022792023,
"grad_norm": 0.09383313020483353,
"learning_rate": 7.432234463304111e-05,
"loss": 0.0347,
"step": 4260
},
{
"epoch": 2.027540360873694,
"grad_norm": 0.08379252020027697,
"learning_rate": 7.417739531659674e-05,
"loss": 0.0332,
"step": 4270
},
{
"epoch": 2.0322886989553655,
"grad_norm": 0.0885642048794524,
"learning_rate": 7.403218031511428e-05,
"loss": 0.0336,
"step": 4280
},
{
"epoch": 2.037037037037037,
"grad_norm": 0.09204820873488205,
"learning_rate": 7.388670122435928e-05,
"loss": 0.0343,
"step": 4290
},
{
"epoch": 2.0417853751187085,
"grad_norm": 0.09695733221959003,
"learning_rate": 7.374095964299928e-05,
"loss": 0.035,
"step": 4300
},
{
"epoch": 2.04653371320038,
"grad_norm": 0.09760809550499844,
"learning_rate": 7.359495717258636e-05,
"loss": 0.033,
"step": 4310
},
{
"epoch": 2.051282051282051,
"grad_norm": 0.0924620931414907,
"learning_rate": 7.34486954175395e-05,
"loss": 0.0334,
"step": 4320
},
{
"epoch": 2.056030389363723,
"grad_norm": 0.09092222678680734,
"learning_rate": 7.330217598512695e-05,
"loss": 0.0336,
"step": 4330
},
{
"epoch": 2.060778727445394,
"grad_norm": 0.15841505668151176,
"learning_rate": 7.315540048544857e-05,
"loss": 0.0343,
"step": 4340
},
{
"epoch": 2.0655270655270654,
"grad_norm": 0.09709842766289198,
"learning_rate": 7.300837053141815e-05,
"loss": 0.0343,
"step": 4350
},
{
"epoch": 2.0702754036087367,
"grad_norm": 0.08697336038504488,
"learning_rate": 7.286108773874566e-05,
"loss": 0.0344,
"step": 4360
},
{
"epoch": 2.0750237416904085,
"grad_norm": 0.08354192261344456,
"learning_rate": 7.271355372591956e-05,
"loss": 0.0341,
"step": 4370
},
{
"epoch": 2.07977207977208,
"grad_norm": 0.08985348685004625,
"learning_rate": 7.256577011418888e-05,
"loss": 0.0343,
"step": 4380
},
{
"epoch": 2.084520417853751,
"grad_norm": 0.09631761182230314,
"learning_rate": 7.241773852754556e-05,
"loss": 0.0334,
"step": 4390
},
{
"epoch": 2.0892687559354224,
"grad_norm": 0.12523187957970167,
"learning_rate": 7.226946059270651e-05,
"loss": 0.0339,
"step": 4400
},
{
"epoch": 2.094017094017094,
"grad_norm": 0.10359793698125809,
"learning_rate": 7.212093793909573e-05,
"loss": 0.0333,
"step": 4410
},
{
"epoch": 2.0987654320987654,
"grad_norm": 0.09753529835523764,
"learning_rate": 7.197217219882647e-05,
"loss": 0.0344,
"step": 4420
},
{
"epoch": 2.1035137701804367,
"grad_norm": 0.09468508667099515,
"learning_rate": 7.182316500668326e-05,
"loss": 0.0329,
"step": 4430
},
{
"epoch": 2.1082621082621085,
"grad_norm": 0.09267335762303072,
"learning_rate": 7.167391800010389e-05,
"loss": 0.033,
"step": 4440
},
{
"epoch": 2.1130104463437798,
"grad_norm": 0.08929136500763987,
"learning_rate": 7.152443281916151e-05,
"loss": 0.0335,
"step": 4450
},
{
"epoch": 2.117758784425451,
"grad_norm": 0.09296989964896214,
"learning_rate": 7.137471110654656e-05,
"loss": 0.0326,
"step": 4460
},
{
"epoch": 2.1225071225071224,
"grad_norm": 0.10051192810430776,
"learning_rate": 7.122475450754869e-05,
"loss": 0.0336,
"step": 4470
},
{
"epoch": 2.127255460588794,
"grad_norm": 0.09381238773738447,
"learning_rate": 7.107456467003877e-05,
"loss": 0.0336,
"step": 4480
},
{
"epoch": 2.1320037986704654,
"grad_norm": 0.08555915211623592,
"learning_rate": 7.092414324445068e-05,
"loss": 0.0343,
"step": 4490
},
{
"epoch": 2.1367521367521367,
"grad_norm": 0.09418166439249125,
"learning_rate": 7.077349188376325e-05,
"loss": 0.0335,
"step": 4500
},
{
"epoch": 2.141500474833808,
"grad_norm": 0.12833271049833236,
"learning_rate": 7.062261224348203e-05,
"loss": 0.0343,
"step": 4510
},
{
"epoch": 2.1462488129154798,
"grad_norm": 0.5517729341991755,
"learning_rate": 7.047150598162114e-05,
"loss": 0.0367,
"step": 4520
},
{
"epoch": 2.150997150997151,
"grad_norm": 0.11804434221523738,
"learning_rate": 7.032017475868505e-05,
"loss": 0.0359,
"step": 4530
},
{
"epoch": 2.1557454890788224,
"grad_norm": 0.10332470532670134,
"learning_rate": 7.016862023765033e-05,
"loss": 0.0346,
"step": 4540
},
{
"epoch": 2.1604938271604937,
"grad_norm": 0.09608821311984503,
"learning_rate": 7.001684408394735e-05,
"loss": 0.0358,
"step": 4550
},
{
"epoch": 2.1652421652421654,
"grad_norm": 0.08641991425090413,
"learning_rate": 6.986484796544203e-05,
"loss": 0.0336,
"step": 4560
},
{
"epoch": 2.1699905033238367,
"grad_norm": 0.10059625511623149,
"learning_rate": 6.971263355241742e-05,
"loss": 0.0346,
"step": 4570
},
{
"epoch": 2.174738841405508,
"grad_norm": 0.09438011734088783,
"learning_rate": 6.956020251755548e-05,
"loss": 0.0347,
"step": 4580
},
{
"epoch": 2.1794871794871793,
"grad_norm": 0.10702517414735709,
"learning_rate": 6.940755653591858e-05,
"loss": 0.0327,
"step": 4590
},
{
"epoch": 2.184235517568851,
"grad_norm": 0.08138272216249738,
"learning_rate": 6.925469728493114e-05,
"loss": 0.0338,
"step": 4600
},
{
"epoch": 2.1889838556505223,
"grad_norm": 0.09732033415377356,
"learning_rate": 6.91016264443612e-05,
"loss": 0.0326,
"step": 4610
},
{
"epoch": 2.1937321937321936,
"grad_norm": 0.09455730370883421,
"learning_rate": 6.894834569630195e-05,
"loss": 0.0331,
"step": 4620
},
{
"epoch": 2.198480531813865,
"grad_norm": 0.10393841811283785,
"learning_rate": 6.879485672515327e-05,
"loss": 0.0347,
"step": 4630
},
{
"epoch": 2.2032288698955367,
"grad_norm": 0.1006910552481963,
"learning_rate": 6.864116121760316e-05,
"loss": 0.0339,
"step": 4640
},
{
"epoch": 2.207977207977208,
"grad_norm": 0.09117286532348516,
"learning_rate": 6.848726086260929e-05,
"loss": 0.0335,
"step": 4650
},
{
"epoch": 2.2127255460588793,
"grad_norm": 0.10704354598309285,
"learning_rate": 6.833315735138035e-05,
"loss": 0.0332,
"step": 4660
},
{
"epoch": 2.217473884140551,
"grad_norm": 0.09526125442652099,
"learning_rate": 6.817885237735758e-05,
"loss": 0.0337,
"step": 4670
},
{
"epoch": 2.2222222222222223,
"grad_norm": 0.09173918628932443,
"learning_rate": 6.802434763619599e-05,
"loss": 0.0339,
"step": 4680
},
{
"epoch": 2.2269705603038936,
"grad_norm": 0.08741156538634473,
"learning_rate": 6.786964482574594e-05,
"loss": 0.0334,
"step": 4690
},
{
"epoch": 2.231718898385565,
"grad_norm": 0.10011252655871726,
"learning_rate": 6.771474564603428e-05,
"loss": 0.0336,
"step": 4700
},
{
"epoch": 2.2364672364672367,
"grad_norm": 0.09295504016636089,
"learning_rate": 6.755965179924582e-05,
"loss": 0.0342,
"step": 4710
},
{
"epoch": 2.241215574548908,
"grad_norm": 0.10686823375667763,
"learning_rate": 6.740436498970452e-05,
"loss": 0.0327,
"step": 4720
},
{
"epoch": 2.2459639126305793,
"grad_norm": 0.0950136464832783,
"learning_rate": 6.724888692385485e-05,
"loss": 0.0332,
"step": 4730
},
{
"epoch": 2.2507122507122506,
"grad_norm": 0.07622884958826211,
"learning_rate": 6.709321931024293e-05,
"loss": 0.0335,
"step": 4740
},
{
"epoch": 2.2554605887939223,
"grad_norm": 0.0840550886867304,
"learning_rate": 6.693736385949788e-05,
"loss": 0.0337,
"step": 4750
},
{
"epoch": 2.2602089268755936,
"grad_norm": 0.09329525393454677,
"learning_rate": 6.678132228431294e-05,
"loss": 0.0338,
"step": 4760
},
{
"epoch": 2.264957264957265,
"grad_norm": 0.08912481894539537,
"learning_rate": 6.662509629942661e-05,
"loss": 0.0343,
"step": 4770
},
{
"epoch": 2.269705603038936,
"grad_norm": 0.0826855504859376,
"learning_rate": 6.646868762160399e-05,
"loss": 0.033,
"step": 4780
},
{
"epoch": 2.274453941120608,
"grad_norm": 0.08739257556285092,
"learning_rate": 6.631209796961767e-05,
"loss": 0.0338,
"step": 4790
},
{
"epoch": 2.2792022792022792,
"grad_norm": 0.09721411374831243,
"learning_rate": 6.615532906422902e-05,
"loss": 0.0333,
"step": 4800
},
{
"epoch": 2.2839506172839505,
"grad_norm": 0.07670182235836218,
"learning_rate": 6.599838262816922e-05,
"loss": 0.0324,
"step": 4810
},
{
"epoch": 2.288698955365622,
"grad_norm": 0.09604039197176296,
"learning_rate": 6.584126038612033e-05,
"loss": 0.0328,
"step": 4820
},
{
"epoch": 2.2934472934472936,
"grad_norm": 0.08057224561648799,
"learning_rate": 6.568396406469632e-05,
"loss": 0.0345,
"step": 4830
},
{
"epoch": 2.298195631528965,
"grad_norm": 0.08599691745568716,
"learning_rate": 6.552649539242414e-05,
"loss": 0.0327,
"step": 4840
},
{
"epoch": 2.302943969610636,
"grad_norm": 0.09350979569292024,
"learning_rate": 6.536885609972467e-05,
"loss": 0.0323,
"step": 4850
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.10394272808691775,
"learning_rate": 6.521104791889381e-05,
"loss": 0.0328,
"step": 4860
},
{
"epoch": 2.3124406457739792,
"grad_norm": 0.08913481965194521,
"learning_rate": 6.505307258408325e-05,
"loss": 0.0319,
"step": 4870
},
{
"epoch": 2.3171889838556505,
"grad_norm": 0.09045268174195037,
"learning_rate": 6.489493183128163e-05,
"loss": 0.034,
"step": 4880
},
{
"epoch": 2.321937321937322,
"grad_norm": 0.09840156046097087,
"learning_rate": 6.473662739829533e-05,
"loss": 0.0344,
"step": 4890
},
{
"epoch": 2.3266856600189936,
"grad_norm": 0.0947479472372267,
"learning_rate": 6.457816102472942e-05,
"loss": 0.0337,
"step": 4900
},
{
"epoch": 2.331433998100665,
"grad_norm": 0.09456940592918854,
"learning_rate": 6.44195344519685e-05,
"loss": 0.0331,
"step": 4910
},
{
"epoch": 2.336182336182336,
"grad_norm": 0.08158030513555407,
"learning_rate": 6.426074942315766e-05,
"loss": 0.0333,
"step": 4920
},
{
"epoch": 2.3409306742640075,
"grad_norm": 0.08686944956661297,
"learning_rate": 6.410180768318321e-05,
"loss": 0.033,
"step": 4930
},
{
"epoch": 2.3456790123456788,
"grad_norm": 0.09422974802710382,
"learning_rate": 6.394271097865355e-05,
"loss": 0.0334,
"step": 4940
},
{
"epoch": 2.3504273504273505,
"grad_norm": 0.09969735861512831,
"learning_rate": 6.378346105788002e-05,
"loss": 0.0332,
"step": 4950
},
{
"epoch": 2.355175688509022,
"grad_norm": 0.09334873152573418,
"learning_rate": 6.362405967085765e-05,
"loss": 0.0339,
"step": 4960
},
{
"epoch": 2.359924026590693,
"grad_norm": 0.08684182627091865,
"learning_rate": 6.346450856924588e-05,
"loss": 0.0336,
"step": 4970
},
{
"epoch": 2.364672364672365,
"grad_norm": 0.09232062691438885,
"learning_rate": 6.330480950634942e-05,
"loss": 0.0346,
"step": 4980
},
{
"epoch": 2.369420702754036,
"grad_norm": 0.08548034785062326,
"learning_rate": 6.314496423709887e-05,
"loss": 0.0338,
"step": 4990
},
{
"epoch": 2.3741690408357075,
"grad_norm": 0.0881499904476215,
"learning_rate": 6.298497451803157e-05,
"loss": 0.0324,
"step": 5000
},
{
"epoch": 2.3789173789173788,
"grad_norm": 0.08694560974385165,
"learning_rate": 6.282484210727213e-05,
"loss": 0.0324,
"step": 5010
},
{
"epoch": 2.3836657169990505,
"grad_norm": 0.09553614314690372,
"learning_rate": 6.266456876451324e-05,
"loss": 0.0326,
"step": 5020
},
{
"epoch": 2.388414055080722,
"grad_norm": 0.07784439475880009,
"learning_rate": 6.25041562509963e-05,
"loss": 0.0327,
"step": 5030
},
{
"epoch": 2.393162393162393,
"grad_norm": 0.08886665892266726,
"learning_rate": 6.234360632949201e-05,
"loss": 0.0328,
"step": 5040
},
{
"epoch": 2.3979107312440644,
"grad_norm": 0.09303084598004047,
"learning_rate": 6.21829207642811e-05,
"loss": 0.0325,
"step": 5050
},
{
"epoch": 2.402659069325736,
"grad_norm": 0.09079599170415052,
"learning_rate": 6.202210132113484e-05,
"loss": 0.0333,
"step": 5060
},
{
"epoch": 2.4074074074074074,
"grad_norm": 0.08696633551333814,
"learning_rate": 6.186114976729573e-05,
"loss": 0.0324,
"step": 5070
},
{
"epoch": 2.4121557454890787,
"grad_norm": 0.09679153433267164,
"learning_rate": 6.170006787145796e-05,
"loss": 0.0328,
"step": 5080
},
{
"epoch": 2.41690408357075,
"grad_norm": 0.07684368348320036,
"learning_rate": 6.153885740374813e-05,
"loss": 0.0326,
"step": 5090
},
{
"epoch": 2.421652421652422,
"grad_norm": 0.09060234321511978,
"learning_rate": 6.137752013570566e-05,
"loss": 0.0325,
"step": 5100
},
{
"epoch": 2.426400759734093,
"grad_norm": 0.07564822074429885,
"learning_rate": 6.121605784026339e-05,
"loss": 0.0319,
"step": 5110
},
{
"epoch": 2.4311490978157644,
"grad_norm": 0.08416481259519201,
"learning_rate": 6.10544722917281e-05,
"loss": 0.0323,
"step": 5120
},
{
"epoch": 2.435897435897436,
"grad_norm": 0.09105989427576541,
"learning_rate": 6.089276526576094e-05,
"loss": 0.0329,
"step": 5130
},
{
"epoch": 2.4406457739791074,
"grad_norm": 0.08777770931969169,
"learning_rate": 6.0730938539358084e-05,
"loss": 0.0321,
"step": 5140
},
{
"epoch": 2.4453941120607787,
"grad_norm": 0.0895279993033145,
"learning_rate": 6.056899389083097e-05,
"loss": 0.0334,
"step": 5150
},
{
"epoch": 2.45014245014245,
"grad_norm": 0.08590705799962003,
"learning_rate": 6.0406933099786976e-05,
"loss": 0.0329,
"step": 5160
},
{
"epoch": 2.4548907882241213,
"grad_norm": 0.08574765382060986,
"learning_rate": 6.02447579471097e-05,
"loss": 0.0331,
"step": 5170
},
{
"epoch": 2.459639126305793,
"grad_norm": 0.09975553197840999,
"learning_rate": 6.0082470214939514e-05,
"loss": 0.0335,
"step": 5180
},
{
"epoch": 2.4643874643874644,
"grad_norm": 0.08108696256008983,
"learning_rate": 5.9920071686653855e-05,
"loss": 0.0326,
"step": 5190
},
{
"epoch": 2.4691358024691357,
"grad_norm": 0.08333423733105016,
"learning_rate": 5.9757564146847764e-05,
"loss": 0.0323,
"step": 5200
},
{
"epoch": 2.4738841405508074,
"grad_norm": 0.07624902008305147,
"learning_rate": 5.959494938131418e-05,
"loss": 0.0324,
"step": 5210
},
{
"epoch": 2.4786324786324787,
"grad_norm": 0.08783019727470423,
"learning_rate": 5.9432229177024346e-05,
"loss": 0.0334,
"step": 5220
},
{
"epoch": 2.48338081671415,
"grad_norm": 0.08364426657137947,
"learning_rate": 5.926940532210817e-05,
"loss": 0.0315,
"step": 5230
},
{
"epoch": 2.4881291547958213,
"grad_norm": 0.08418245017597606,
"learning_rate": 5.910647960583457e-05,
"loss": 0.0329,
"step": 5240
},
{
"epoch": 2.492877492877493,
"grad_norm": 0.08279672093384424,
"learning_rate": 5.8943453818591845e-05,
"loss": 0.0318,
"step": 5250
},
{
"epoch": 2.4976258309591644,
"grad_norm": 0.08335470529633829,
"learning_rate": 5.878032975186791e-05,
"loss": 0.0326,
"step": 5260
},
{
"epoch": 2.5023741690408356,
"grad_norm": 0.08841783611203549,
"learning_rate": 5.8617109198230744e-05,
"loss": 0.0331,
"step": 5270
},
{
"epoch": 2.5071225071225074,
"grad_norm": 0.07536448612660623,
"learning_rate": 5.845379395130859e-05,
"loss": 0.0326,
"step": 5280
},
{
"epoch": 2.5118708452041787,
"grad_norm": 0.07800881277719429,
"learning_rate": 5.829038580577025e-05,
"loss": 0.0338,
"step": 5290
},
{
"epoch": 2.51661918328585,
"grad_norm": 0.08064948001059684,
"learning_rate": 5.812688655730542e-05,
"loss": 0.0339,
"step": 5300
},
{
"epoch": 2.5213675213675213,
"grad_norm": 0.0854646647024056,
"learning_rate": 5.796329800260489e-05,
"loss": 0.0326,
"step": 5310
},
{
"epoch": 2.5261158594491926,
"grad_norm": 0.07099847166838609,
"learning_rate": 5.7799621939340896e-05,
"loss": 0.0335,
"step": 5320
},
{
"epoch": 2.5308641975308643,
"grad_norm": 0.0919571649966859,
"learning_rate": 5.7635860166147246e-05,
"loss": 0.0326,
"step": 5330
},
{
"epoch": 2.5356125356125356,
"grad_norm": 0.08136925452343646,
"learning_rate": 5.747201448259964e-05,
"loss": 0.033,
"step": 5340
},
{
"epoch": 2.540360873694207,
"grad_norm": 0.07768121682236279,
"learning_rate": 5.730808668919584e-05,
"loss": 0.0322,
"step": 5350
},
{
"epoch": 2.5451092117758787,
"grad_norm": 0.0832113041183157,
"learning_rate": 5.714407858733596e-05,
"loss": 0.0326,
"step": 5360
},
{
"epoch": 2.54985754985755,
"grad_norm": 0.07916835447229251,
"learning_rate": 5.6979991979302585e-05,
"loss": 0.0318,
"step": 5370
},
{
"epoch": 2.5546058879392213,
"grad_norm": 0.08583357145960785,
"learning_rate": 5.6815828668241e-05,
"loss": 0.0329,
"step": 5380
},
{
"epoch": 2.5593542260208926,
"grad_norm": 0.08633840869426782,
"learning_rate": 5.6651590458139424e-05,
"loss": 0.0337,
"step": 5390
},
{
"epoch": 2.564102564102564,
"grad_norm": 0.08750158018722907,
"learning_rate": 5.6487279153809094e-05,
"loss": 0.034,
"step": 5400
},
{
"epoch": 2.5688509021842356,
"grad_norm": 0.08428686527861566,
"learning_rate": 5.632289656086448e-05,
"loss": 0.0324,
"step": 5410
},
{
"epoch": 2.573599240265907,
"grad_norm": 0.07806249817855916,
"learning_rate": 5.6158444485703465e-05,
"loss": 0.0321,
"step": 5420
},
{
"epoch": 2.578347578347578,
"grad_norm": 0.08604945437408525,
"learning_rate": 5.5993924735487455e-05,
"loss": 0.0334,
"step": 5430
},
{
"epoch": 2.58309591642925,
"grad_norm": 0.08073065007282099,
"learning_rate": 5.582933911812155e-05,
"loss": 0.032,
"step": 5440
},
{
"epoch": 2.5878442545109213,
"grad_norm": 0.08589683927065168,
"learning_rate": 5.5664689442234654e-05,
"loss": 0.0315,
"step": 5450
},
{
"epoch": 2.5925925925925926,
"grad_norm": 0.07671918728308563,
"learning_rate": 5.549997751715961e-05,
"loss": 0.0334,
"step": 5460
},
{
"epoch": 2.597340930674264,
"grad_norm": 0.08212888489357918,
"learning_rate": 5.5335205152913305e-05,
"loss": 0.0322,
"step": 5470
},
{
"epoch": 2.602089268755935,
"grad_norm": 0.08851254070965249,
"learning_rate": 5.5170374160176815e-05,
"loss": 0.0317,
"step": 5480
},
{
"epoch": 2.606837606837607,
"grad_norm": 0.07885333909754141,
"learning_rate": 5.500548635027547e-05,
"loss": 0.0323,
"step": 5490
},
{
"epoch": 2.611585944919278,
"grad_norm": 0.08410229173254444,
"learning_rate": 5.484054353515896e-05,
"loss": 0.032,
"step": 5500
},
{
"epoch": 2.61633428300095,
"grad_norm": 0.08276510446854092,
"learning_rate": 5.467554752738143e-05,
"loss": 0.0322,
"step": 5510
},
{
"epoch": 2.6210826210826212,
"grad_norm": 0.0943926498548148,
"learning_rate": 5.451050014008157e-05,
"loss": 0.0314,
"step": 5520
},
{
"epoch": 2.6258309591642925,
"grad_norm": 0.09342391459236635,
"learning_rate": 5.434540318696265e-05,
"loss": 0.0329,
"step": 5530
},
{
"epoch": 2.630579297245964,
"grad_norm": 0.08325118233498842,
"learning_rate": 5.418025848227265e-05,
"loss": 0.0329,
"step": 5540
},
{
"epoch": 2.635327635327635,
"grad_norm": 0.08220911325584158,
"learning_rate": 5.401506784078427e-05,
"loss": 0.0321,
"step": 5550
},
{
"epoch": 2.640075973409307,
"grad_norm": 0.09171902271989865,
"learning_rate": 5.384983307777502e-05,
"loss": 0.0326,
"step": 5560
},
{
"epoch": 2.644824311490978,
"grad_norm": 0.08498951901622376,
"learning_rate": 5.368455600900725e-05,
"loss": 0.0321,
"step": 5570
},
{
"epoch": 2.6495726495726495,
"grad_norm": 0.07807089159363576,
"learning_rate": 5.351923845070822e-05,
"loss": 0.0322,
"step": 5580
},
{
"epoch": 2.6543209876543212,
"grad_norm": 0.09292093721102322,
"learning_rate": 5.335388221955012e-05,
"loss": 0.0331,
"step": 5590
},
{
"epoch": 2.6590693257359925,
"grad_norm": 0.07769770831547565,
"learning_rate": 5.3188489132630124e-05,
"loss": 0.0326,
"step": 5600
},
{
"epoch": 2.663817663817664,
"grad_norm": 0.0856602219430577,
"learning_rate": 5.3023061007450404e-05,
"loss": 0.0317,
"step": 5610
},
{
"epoch": 2.668566001899335,
"grad_norm": 0.08656430394955437,
"learning_rate": 5.285759966189817e-05,
"loss": 0.0326,
"step": 5620
},
{
"epoch": 2.6733143399810064,
"grad_norm": 0.0777689790147013,
"learning_rate": 5.269210691422569e-05,
"loss": 0.0318,
"step": 5630
},
{
"epoch": 2.678062678062678,
"grad_norm": 0.08804471522147092,
"learning_rate": 5.252658458303032e-05,
"loss": 0.0316,
"step": 5640
},
{
"epoch": 2.6828110161443495,
"grad_norm": 0.07938227834463157,
"learning_rate": 5.236103448723448e-05,
"loss": 0.0318,
"step": 5650
},
{
"epoch": 2.6875593542260208,
"grad_norm": 0.10317236107818047,
"learning_rate": 5.219545844606573e-05,
"loss": 0.0326,
"step": 5660
},
{
"epoch": 2.6923076923076925,
"grad_norm": 0.11010219209713488,
"learning_rate": 5.202985827903672e-05,
"loss": 0.0325,
"step": 5670
},
{
"epoch": 2.697056030389364,
"grad_norm": 0.08392775183699804,
"learning_rate": 5.186423580592523e-05,
"loss": 0.0322,
"step": 5680
},
{
"epoch": 2.701804368471035,
"grad_norm": 0.07934521672500533,
"learning_rate": 5.1698592846754156e-05,
"loss": 0.0313,
"step": 5690
},
{
"epoch": 2.7065527065527064,
"grad_norm": 0.07786545955483862,
"learning_rate": 5.1532931221771515e-05,
"loss": 0.0307,
"step": 5700
},
{
"epoch": 2.7113010446343777,
"grad_norm": 0.07398335204267062,
"learning_rate": 5.136725275143045e-05,
"loss": 0.032,
"step": 5710
},
{
"epoch": 2.7160493827160495,
"grad_norm": 0.07993306108423914,
"learning_rate": 5.1201559256369215e-05,
"loss": 0.032,
"step": 5720
},
{
"epoch": 2.7207977207977208,
"grad_norm": 0.07752182162787653,
"learning_rate": 5.103585255739113e-05,
"loss": 0.0332,
"step": 5730
},
{
"epoch": 2.725546058879392,
"grad_norm": 0.07332760631930846,
"learning_rate": 5.087013447544469e-05,
"loss": 0.0322,
"step": 5740
},
{
"epoch": 2.730294396961064,
"grad_norm": 0.09815231050152398,
"learning_rate": 5.0704406831603414e-05,
"loss": 0.0327,
"step": 5750
},
{
"epoch": 2.735042735042735,
"grad_norm": 0.08531312219775723,
"learning_rate": 5.0538671447045936e-05,
"loss": 0.0324,
"step": 5760
},
{
"epoch": 2.7397910731244064,
"grad_norm": 0.08179347888702546,
"learning_rate": 5.037293014303592e-05,
"loss": 0.0326,
"step": 5770
},
{
"epoch": 2.7445394112060777,
"grad_norm": 0.07379964853442134,
"learning_rate": 5.020718474090212e-05,
"loss": 0.0313,
"step": 5780
},
{
"epoch": 2.7492877492877494,
"grad_norm": 0.07389548243729202,
"learning_rate": 5.004143706201826e-05,
"loss": 0.0314,
"step": 5790
},
{
"epoch": 2.7540360873694207,
"grad_norm": 0.08637255139238713,
"learning_rate": 4.987568892778314e-05,
"loss": 0.0321,
"step": 5800
},
{
"epoch": 2.758784425451092,
"grad_norm": 0.0825583527833555,
"learning_rate": 4.970994215960056e-05,
"loss": 0.0325,
"step": 5810
},
{
"epoch": 2.763532763532764,
"grad_norm": 0.08021207801056433,
"learning_rate": 4.954419857885926e-05,
"loss": 0.0329,
"step": 5820
},
{
"epoch": 2.768281101614435,
"grad_norm": 0.0768666173835066,
"learning_rate": 4.9378460006913035e-05,
"loss": 0.031,
"step": 5830
},
{
"epoch": 2.7730294396961064,
"grad_norm": 0.071709967334629,
"learning_rate": 4.921272826506054e-05,
"loss": 0.0319,
"step": 5840
},
{
"epoch": 2.7777777777777777,
"grad_norm": 0.09741012607605265,
"learning_rate": 4.9047005174525454e-05,
"loss": 0.0324,
"step": 5850
},
{
"epoch": 2.782526115859449,
"grad_norm": 0.0839874401917823,
"learning_rate": 4.8881292556436354e-05,
"loss": 0.0318,
"step": 5860
},
{
"epoch": 2.7872744539411207,
"grad_norm": 0.0835724307560738,
"learning_rate": 4.871559223180673e-05,
"loss": 0.0312,
"step": 5870
},
{
"epoch": 2.792022792022792,
"grad_norm": 0.0965132006919279,
"learning_rate": 4.854990602151502e-05,
"loss": 0.0322,
"step": 5880
},
{
"epoch": 2.7967711301044633,
"grad_norm": 0.07851476810574484,
"learning_rate": 4.8384235746284476e-05,
"loss": 0.0327,
"step": 5890
},
{
"epoch": 2.801519468186135,
"grad_norm": 0.08361712818415301,
"learning_rate": 4.821858322666333e-05,
"loss": 0.0315,
"step": 5900
},
{
"epoch": 2.8062678062678064,
"grad_norm": 0.07728202920751281,
"learning_rate": 4.8052950283004637e-05,
"loss": 0.0315,
"step": 5910
},
{
"epoch": 2.8110161443494777,
"grad_norm": 0.08208798307594348,
"learning_rate": 4.7887338735446394e-05,
"loss": 0.0313,
"step": 5920
},
{
"epoch": 2.815764482431149,
"grad_norm": 0.07699487433929669,
"learning_rate": 4.772175040389138e-05,
"loss": 0.0309,
"step": 5930
},
{
"epoch": 2.8205128205128203,
"grad_norm": 0.07938229301685117,
"learning_rate": 4.7556187107987365e-05,
"loss": 0.0316,
"step": 5940
},
{
"epoch": 2.825261158594492,
"grad_norm": 0.07544358342284178,
"learning_rate": 4.73906506671069e-05,
"loss": 0.0304,
"step": 5950
},
{
"epoch": 2.8300094966761633,
"grad_norm": 0.07837176856411082,
"learning_rate": 4.722514290032753e-05,
"loss": 0.0323,
"step": 5960
},
{
"epoch": 2.8347578347578346,
"grad_norm": 0.07860111691921295,
"learning_rate": 4.70596656264116e-05,
"loss": 0.0308,
"step": 5970
},
{
"epoch": 2.8395061728395063,
"grad_norm": 0.09363628024729989,
"learning_rate": 4.689422066378644e-05,
"loss": 0.0316,
"step": 5980
},
{
"epoch": 2.8442545109211776,
"grad_norm": 0.07649608849610263,
"learning_rate": 4.6728809830524285e-05,
"loss": 0.0317,
"step": 5990
},
{
"epoch": 2.849002849002849,
"grad_norm": 0.0837700867967604,
"learning_rate": 4.65634349443223e-05,
"loss": 0.0318,
"step": 6000
},
{
"epoch": 2.8537511870845202,
"grad_norm": 0.08663729819194713,
"learning_rate": 4.63980978224827e-05,
"loss": 0.033,
"step": 6010
},
{
"epoch": 2.858499525166192,
"grad_norm": 0.08209160278456785,
"learning_rate": 4.62328002818926e-05,
"loss": 0.0303,
"step": 6020
},
{
"epoch": 2.8632478632478633,
"grad_norm": 0.07427636776758134,
"learning_rate": 4.606754413900428e-05,
"loss": 0.0315,
"step": 6030
},
{
"epoch": 2.8679962013295346,
"grad_norm": 0.07877608448218568,
"learning_rate": 4.5902331209814995e-05,
"loss": 0.0325,
"step": 6040
},
{
"epoch": 2.8727445394112063,
"grad_norm": 0.08152658980207392,
"learning_rate": 4.5737163309847205e-05,
"loss": 0.0318,
"step": 6050
},
{
"epoch": 2.8774928774928776,
"grad_norm": 0.07681299449738352,
"learning_rate": 4.557204225412848e-05,
"loss": 0.0311,
"step": 6060
},
{
"epoch": 2.882241215574549,
"grad_norm": 0.08112902573299945,
"learning_rate": 4.540696985717168e-05,
"loss": 0.0316,
"step": 6070
},
{
"epoch": 2.8869895536562202,
"grad_norm": 0.07638478530656273,
"learning_rate": 4.524194793295488e-05,
"loss": 0.032,
"step": 6080
},
{
"epoch": 2.8917378917378915,
"grad_norm": 0.07219966509710556,
"learning_rate": 4.5076978294901604e-05,
"loss": 0.0319,
"step": 6090
},
{
"epoch": 2.8964862298195633,
"grad_norm": 0.07190520015487549,
"learning_rate": 4.491206275586072e-05,
"loss": 0.0321,
"step": 6100
},
{
"epoch": 2.9012345679012346,
"grad_norm": 0.08477785809339851,
"learning_rate": 4.4747203128086634e-05,
"loss": 0.032,
"step": 6110
},
{
"epoch": 2.905982905982906,
"grad_norm": 0.07664484950606382,
"learning_rate": 4.458240122321939e-05,
"loss": 0.0308,
"step": 6120
},
{
"epoch": 2.9107312440645776,
"grad_norm": 0.07414811863723912,
"learning_rate": 4.4417658852264614e-05,
"loss": 0.031,
"step": 6130
},
{
"epoch": 2.915479582146249,
"grad_norm": 0.08238423731188826,
"learning_rate": 4.425297782557383e-05,
"loss": 0.0306,
"step": 6140
},
{
"epoch": 2.92022792022792,
"grad_norm": 0.07990574238185166,
"learning_rate": 4.4088359952824344e-05,
"loss": 0.0308,
"step": 6150
},
{
"epoch": 2.9249762583095915,
"grad_norm": 0.08044455289388773,
"learning_rate": 4.392380704299955e-05,
"loss": 0.0318,
"step": 6160
},
{
"epoch": 2.929724596391263,
"grad_norm": 0.07091349989488455,
"learning_rate": 4.3759320904368913e-05,
"loss": 0.0306,
"step": 6170
},
{
"epoch": 2.9344729344729346,
"grad_norm": 0.0741794501526838,
"learning_rate": 4.359490334446819e-05,
"loss": 0.031,
"step": 6180
},
{
"epoch": 2.939221272554606,
"grad_norm": 0.08024416184300108,
"learning_rate": 4.343055617007947e-05,
"loss": 0.0321,
"step": 6190
},
{
"epoch": 2.943969610636277,
"grad_norm": 0.0727594143050465,
"learning_rate": 4.326628118721144e-05,
"loss": 0.0318,
"step": 6200
},
{
"epoch": 2.948717948717949,
"grad_norm": 0.07684102044239445,
"learning_rate": 4.3102080201079435e-05,
"loss": 0.0311,
"step": 6210
},
{
"epoch": 2.95346628679962,
"grad_norm": 0.0782864814968308,
"learning_rate": 4.293795501608565e-05,
"loss": 0.0331,
"step": 6220
},
{
"epoch": 2.9582146248812915,
"grad_norm": 0.0766096109368061,
"learning_rate": 4.277390743579932e-05,
"loss": 0.0318,
"step": 6230
},
{
"epoch": 2.962962962962963,
"grad_norm": 0.07416026844620732,
"learning_rate": 4.2609939262936835e-05,
"loss": 0.0317,
"step": 6240
},
{
"epoch": 2.9677113010446345,
"grad_norm": 0.08325936907518372,
"learning_rate": 4.244605229934206e-05,
"loss": 0.0311,
"step": 6250
},
{
"epoch": 2.972459639126306,
"grad_norm": 0.08401039843862111,
"learning_rate": 4.228224834596636e-05,
"loss": 0.0313,
"step": 6260
},
{
"epoch": 2.977207977207977,
"grad_norm": 0.0759425734680834,
"learning_rate": 4.211852920284899e-05,
"loss": 0.0305,
"step": 6270
},
{
"epoch": 2.981956315289649,
"grad_norm": 0.0697543226420682,
"learning_rate": 4.195489666909713e-05,
"loss": 0.0311,
"step": 6280
},
{
"epoch": 2.98670465337132,
"grad_norm": 0.07098780613087384,
"learning_rate": 4.179135254286631e-05,
"loss": 0.0307,
"step": 6290
},
{
"epoch": 2.9914529914529915,
"grad_norm": 0.08394290373586102,
"learning_rate": 4.162789862134047e-05,
"loss": 0.0311,
"step": 6300
},
{
"epoch": 2.996201329534663,
"grad_norm": 0.0787230043292495,
"learning_rate": 4.146453670071236e-05,
"loss": 0.0315,
"step": 6310
},
{
"epoch": 3.000949667616334,
"grad_norm": 0.06998234391022959,
"learning_rate": 4.1301268576163665e-05,
"loss": 0.0313,
"step": 6320
},
{
"epoch": 3.005698005698006,
"grad_norm": 0.06827381695763439,
"learning_rate": 4.11380960418454e-05,
"loss": 0.0285,
"step": 6330
},
{
"epoch": 3.010446343779677,
"grad_norm": 0.06479682212758979,
"learning_rate": 4.0975020890858115e-05,
"loss": 0.0294,
"step": 6340
},
{
"epoch": 3.0151946818613484,
"grad_norm": 0.07058002412643832,
"learning_rate": 4.081204491523222e-05,
"loss": 0.0297,
"step": 6350
},
{
"epoch": 3.0199430199430197,
"grad_norm": 0.06956790722741975,
"learning_rate": 4.06491699059083e-05,
"loss": 0.0287,
"step": 6360
},
{
"epoch": 3.0246913580246915,
"grad_norm": 0.06875364527032,
"learning_rate": 4.04863976527174e-05,
"loss": 0.0288,
"step": 6370
},
{
"epoch": 3.0294396961063628,
"grad_norm": 0.06814500732217052,
"learning_rate": 4.032372994436142e-05,
"loss": 0.0296,
"step": 6380
},
{
"epoch": 3.034188034188034,
"grad_norm": 0.07526481165906473,
"learning_rate": 4.016116856839335e-05,
"loss": 0.0285,
"step": 6390
},
{
"epoch": 3.038936372269706,
"grad_norm": 0.07137339889681192,
"learning_rate": 3.9998715311197785e-05,
"loss": 0.0293,
"step": 6400
},
{
"epoch": 3.043684710351377,
"grad_norm": 0.07794609266964109,
"learning_rate": 3.9836371957971106e-05,
"loss": 0.0294,
"step": 6410
},
{
"epoch": 3.0484330484330484,
"grad_norm": 0.06211880851439585,
"learning_rate": 3.967414029270206e-05,
"loss": 0.0282,
"step": 6420
},
{
"epoch": 3.0531813865147197,
"grad_norm": 0.07517702764052028,
"learning_rate": 3.9512022098151966e-05,
"loss": 0.0282,
"step": 6430
},
{
"epoch": 3.0579297245963915,
"grad_norm": 0.06972627450779884,
"learning_rate": 3.935001915583526e-05,
"loss": 0.0279,
"step": 6440
},
{
"epoch": 3.0626780626780628,
"grad_norm": 0.06277975345984187,
"learning_rate": 3.918813324599989e-05,
"loss": 0.0287,
"step": 6450
},
{
"epoch": 3.067426400759734,
"grad_norm": 0.07037344046837897,
"learning_rate": 3.902636614760771e-05,
"loss": 0.0286,
"step": 6460
},
{
"epoch": 3.0721747388414054,
"grad_norm": 0.06501191846812195,
"learning_rate": 3.8864719638314955e-05,
"loss": 0.0287,
"step": 6470
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.06267342227677175,
"learning_rate": 3.87031954944527e-05,
"loss": 0.0286,
"step": 6480
},
{
"epoch": 3.0816714150047484,
"grad_norm": 0.06408766218450679,
"learning_rate": 3.854179549100741e-05,
"loss": 0.0292,
"step": 6490
},
{
"epoch": 3.0864197530864197,
"grad_norm": 0.07146272820664924,
"learning_rate": 3.838052140160128e-05,
"loss": 0.028,
"step": 6500
},
{
"epoch": 3.091168091168091,
"grad_norm": 0.07314342057777799,
"learning_rate": 3.821937499847291e-05,
"loss": 0.0285,
"step": 6510
},
{
"epoch": 3.0959164292497627,
"grad_norm": 0.07364411434751586,
"learning_rate": 3.805835805245772e-05,
"loss": 0.0286,
"step": 6520
},
{
"epoch": 3.100664767331434,
"grad_norm": 0.07312046230738507,
"learning_rate": 3.789747233296852e-05,
"loss": 0.0298,
"step": 6530
},
{
"epoch": 3.1054131054131053,
"grad_norm": 0.06907683288125434,
"learning_rate": 3.7736719607976125e-05,
"loss": 0.0287,
"step": 6540
},
{
"epoch": 3.1101614434947766,
"grad_norm": 0.0707198424316592,
"learning_rate": 3.7576101643989816e-05,
"loss": 0.0284,
"step": 6550
},
{
"epoch": 3.1149097815764484,
"grad_norm": 0.07069373385448174,
"learning_rate": 3.7415620206038036e-05,
"loss": 0.0287,
"step": 6560
},
{
"epoch": 3.1196581196581197,
"grad_norm": 0.06568790790080523,
"learning_rate": 3.7255277057648906e-05,
"loss": 0.0283,
"step": 6570
},
{
"epoch": 3.124406457739791,
"grad_norm": 0.07315556329110613,
"learning_rate": 3.7095073960830905e-05,
"loss": 0.0288,
"step": 6580
},
{
"epoch": 3.1291547958214627,
"grad_norm": 0.06352040479044523,
"learning_rate": 3.69350126760535e-05,
"loss": 0.0278,
"step": 6590
},
{
"epoch": 3.133903133903134,
"grad_norm": 0.06908192304519066,
"learning_rate": 3.677509496222776e-05,
"loss": 0.029,
"step": 6600
},
{
"epoch": 3.1386514719848053,
"grad_norm": 0.06625671331481194,
"learning_rate": 3.661532257668705e-05,
"loss": 0.0293,
"step": 6610
},
{
"epoch": 3.1433998100664766,
"grad_norm": 0.07433751823757842,
"learning_rate": 3.645569727516779e-05,
"loss": 0.029,
"step": 6620
},
{
"epoch": 3.148148148148148,
"grad_norm": 0.0680925141191956,
"learning_rate": 3.6296220811790015e-05,
"loss": 0.028,
"step": 6630
},
{
"epoch": 3.1528964862298197,
"grad_norm": 0.059969769548992316,
"learning_rate": 3.613689493903819e-05,
"loss": 0.0291,
"step": 6640
},
{
"epoch": 3.157644824311491,
"grad_norm": 0.07213357276148355,
"learning_rate": 3.597772140774201e-05,
"loss": 0.0284,
"step": 6650
},
{
"epoch": 3.1623931623931623,
"grad_norm": 0.0665895655580276,
"learning_rate": 3.5818701967057014e-05,
"loss": 0.0291,
"step": 6660
},
{
"epoch": 3.167141500474834,
"grad_norm": 0.0690967643977019,
"learning_rate": 3.5659838364445505e-05,
"loss": 0.0295,
"step": 6670
},
{
"epoch": 3.1718898385565053,
"grad_norm": 0.06823320854629436,
"learning_rate": 3.5501132345657204e-05,
"loss": 0.0287,
"step": 6680
},
{
"epoch": 3.1766381766381766,
"grad_norm": 0.07682870343646064,
"learning_rate": 3.5342585654710256e-05,
"loss": 0.0292,
"step": 6690
},
{
"epoch": 3.181386514719848,
"grad_norm": 0.07502990820872905,
"learning_rate": 3.518420003387183e-05,
"loss": 0.0293,
"step": 6700
},
{
"epoch": 3.1861348528015196,
"grad_norm": 0.06762873937237043,
"learning_rate": 3.502597722363921e-05,
"loss": 0.0283,
"step": 6710
},
{
"epoch": 3.190883190883191,
"grad_norm": 0.07060505716151702,
"learning_rate": 3.486791896272049e-05,
"loss": 0.0296,
"step": 6720
},
{
"epoch": 3.1956315289648622,
"grad_norm": 0.06769346265174028,
"learning_rate": 3.471002698801555e-05,
"loss": 0.0277,
"step": 6730
},
{
"epoch": 3.2003798670465335,
"grad_norm": 0.06863924588187657,
"learning_rate": 3.4552303034596965e-05,
"loss": 0.0281,
"step": 6740
},
{
"epoch": 3.2051282051282053,
"grad_norm": 0.0930695117363435,
"learning_rate": 3.4394748835690894e-05,
"loss": 0.0285,
"step": 6750
},
{
"epoch": 3.2098765432098766,
"grad_norm": 0.08549730359996649,
"learning_rate": 3.423736612265812e-05,
"loss": 0.029,
"step": 6760
},
{
"epoch": 3.214624881291548,
"grad_norm": 0.07586890822777904,
"learning_rate": 3.4080156624974904e-05,
"loss": 0.0285,
"step": 6770
},
{
"epoch": 3.219373219373219,
"grad_norm": 0.06593331522200611,
"learning_rate": 3.3923122070214126e-05,
"loss": 0.029,
"step": 6780
},
{
"epoch": 3.224121557454891,
"grad_norm": 0.06410862537033282,
"learning_rate": 3.376626418402612e-05,
"loss": 0.028,
"step": 6790
},
{
"epoch": 3.2288698955365622,
"grad_norm": 0.06897626853457038,
"learning_rate": 3.3609584690119925e-05,
"loss": 0.0287,
"step": 6800
},
{
"epoch": 3.2336182336182335,
"grad_norm": 0.06792335795611693,
"learning_rate": 3.345308531024412e-05,
"loss": 0.0292,
"step": 6810
},
{
"epoch": 3.238366571699905,
"grad_norm": 0.07533629103155168,
"learning_rate": 3.329676776416811e-05,
"loss": 0.0293,
"step": 6820
},
{
"epoch": 3.2431149097815766,
"grad_norm": 0.07439117501028221,
"learning_rate": 3.3140633769663057e-05,
"loss": 0.0293,
"step": 6830
},
{
"epoch": 3.247863247863248,
"grad_norm": 0.0699482080053469,
"learning_rate": 3.298468504248313e-05,
"loss": 0.0286,
"step": 6840
},
{
"epoch": 3.252611585944919,
"grad_norm": 0.07588251948889628,
"learning_rate": 3.282892329634657e-05,
"loss": 0.03,
"step": 6850
},
{
"epoch": 3.2573599240265905,
"grad_norm": 0.06473398330875815,
"learning_rate": 3.267335024291687e-05,
"loss": 0.0278,
"step": 6860
},
{
"epoch": 3.262108262108262,
"grad_norm": 0.08103919968131498,
"learning_rate": 3.251796759178403e-05,
"loss": 0.0301,
"step": 6870
},
{
"epoch": 3.2668566001899335,
"grad_norm": 0.07123464747823359,
"learning_rate": 3.236277705044568e-05,
"loss": 0.0295,
"step": 6880
},
{
"epoch": 3.271604938271605,
"grad_norm": 0.084595861476684,
"learning_rate": 3.220778032428839e-05,
"loss": 0.0291,
"step": 6890
},
{
"epoch": 3.2763532763532766,
"grad_norm": 0.07144619155374488,
"learning_rate": 3.205297911656886e-05,
"loss": 0.0291,
"step": 6900
},
{
"epoch": 3.281101614434948,
"grad_norm": 0.07271009416489914,
"learning_rate": 3.18983751283953e-05,
"loss": 0.0293,
"step": 6910
},
{
"epoch": 3.285849952516619,
"grad_norm": 0.08098521629270833,
"learning_rate": 3.174397005870858e-05,
"loss": 0.0291,
"step": 6920
},
{
"epoch": 3.2905982905982905,
"grad_norm": 0.06549472832811996,
"learning_rate": 3.158976560426379e-05,
"loss": 0.0286,
"step": 6930
},
{
"epoch": 3.295346628679962,
"grad_norm": 0.07588007921827748,
"learning_rate": 3.143576345961132e-05,
"loss": 0.0292,
"step": 6940
},
{
"epoch": 3.3000949667616335,
"grad_norm": 0.07173191311299712,
"learning_rate": 3.128196531707849e-05,
"loss": 0.0286,
"step": 6950
},
{
"epoch": 3.304843304843305,
"grad_norm": 0.083627070861235,
"learning_rate": 3.112837286675082e-05,
"loss": 0.029,
"step": 6960
},
{
"epoch": 3.309591642924976,
"grad_norm": 0.07407127547716451,
"learning_rate": 3.097498779645341e-05,
"loss": 0.0283,
"step": 6970
},
{
"epoch": 3.314339981006648,
"grad_norm": 0.0686781441193216,
"learning_rate": 3.082181179173257e-05,
"loss": 0.0283,
"step": 6980
},
{
"epoch": 3.319088319088319,
"grad_norm": 0.06834929491325734,
"learning_rate": 3.0668846535837114e-05,
"loss": 0.0289,
"step": 6990
},
{
"epoch": 3.3238366571699904,
"grad_norm": 0.06631924235594419,
"learning_rate": 3.051609370969999e-05,
"loss": 0.0295,
"step": 7000
},
{
"epoch": 3.3285849952516617,
"grad_norm": 0.07336301080819957,
"learning_rate": 3.03635549919197e-05,
"loss": 0.0289,
"step": 7010
},
{
"epoch": 3.3333333333333335,
"grad_norm": 0.07636860889869526,
"learning_rate": 3.021123205874199e-05,
"loss": 0.0287,
"step": 7020
},
{
"epoch": 3.338081671415005,
"grad_norm": 0.07310857545889464,
"learning_rate": 3.0059126584041275e-05,
"loss": 0.0282,
"step": 7030
},
{
"epoch": 3.342830009496676,
"grad_norm": 0.06773150643411327,
"learning_rate": 2.9907240239302393e-05,
"loss": 0.0285,
"step": 7040
},
{
"epoch": 3.347578347578348,
"grad_norm": 0.07367136478814061,
"learning_rate": 2.9755574693602083e-05,
"loss": 0.0285,
"step": 7050
},
{
"epoch": 3.352326685660019,
"grad_norm": 0.06757882669330503,
"learning_rate": 2.9604131613590824e-05,
"loss": 0.0289,
"step": 7060
},
{
"epoch": 3.3570750237416904,
"grad_norm": 0.06950395706438342,
"learning_rate": 2.9452912663474348e-05,
"loss": 0.0289,
"step": 7070
},
{
"epoch": 3.3618233618233617,
"grad_norm": 0.0781068084845991,
"learning_rate": 2.9301919504995436e-05,
"loss": 0.029,
"step": 7080
},
{
"epoch": 3.366571699905033,
"grad_norm": 0.06963412687212238,
"learning_rate": 2.915115379741571e-05,
"loss": 0.0289,
"step": 7090
},
{
"epoch": 3.3713200379867048,
"grad_norm": 0.06360027367575186,
"learning_rate": 2.900061719749724e-05,
"loss": 0.0284,
"step": 7100
},
{
"epoch": 3.376068376068376,
"grad_norm": 0.0602006658470827,
"learning_rate": 2.8850311359484527e-05,
"loss": 0.028,
"step": 7110
},
{
"epoch": 3.3808167141500474,
"grad_norm": 0.08130802880209016,
"learning_rate": 2.870023793508616e-05,
"loss": 0.0294,
"step": 7120
},
{
"epoch": 3.385565052231719,
"grad_norm": 0.06599396875058026,
"learning_rate": 2.855039857345681e-05,
"loss": 0.029,
"step": 7130
},
{
"epoch": 3.3903133903133904,
"grad_norm": 0.06890220770066616,
"learning_rate": 2.8400794921178953e-05,
"loss": 0.0277,
"step": 7140
},
{
"epoch": 3.3950617283950617,
"grad_norm": 0.06974147369254255,
"learning_rate": 2.825142862224494e-05,
"loss": 0.0288,
"step": 7150
},
{
"epoch": 3.399810066476733,
"grad_norm": 0.0695599655022747,
"learning_rate": 2.810230131803878e-05,
"loss": 0.0277,
"step": 7160
},
{
"epoch": 3.4045584045584047,
"grad_norm": 0.0657252843620543,
"learning_rate": 2.7953414647318232e-05,
"loss": 0.0292,
"step": 7170
},
{
"epoch": 3.409306742640076,
"grad_norm": 0.06525351642838627,
"learning_rate": 2.78047702461967e-05,
"loss": 0.029,
"step": 7180
},
{
"epoch": 3.4140550807217473,
"grad_norm": 0.067700769695352,
"learning_rate": 2.7656369748125276e-05,
"loss": 0.0288,
"step": 7190
},
{
"epoch": 3.4188034188034186,
"grad_norm": 0.0653215094943802,
"learning_rate": 2.7508214783874874e-05,
"loss": 0.0275,
"step": 7200
},
{
"epoch": 3.4235517568850904,
"grad_norm": 0.0686804257591295,
"learning_rate": 2.7360306981518146e-05,
"loss": 0.0274,
"step": 7210
},
{
"epoch": 3.4283000949667617,
"grad_norm": 0.070487101572742,
"learning_rate": 2.721264796641178e-05,
"loss": 0.0286,
"step": 7220
},
{
"epoch": 3.433048433048433,
"grad_norm": 0.06226687242183555,
"learning_rate": 2.7065239361178464e-05,
"loss": 0.0279,
"step": 7230
},
{
"epoch": 3.4377967711301043,
"grad_norm": 0.0769724996531241,
"learning_rate": 2.6918082785689208e-05,
"loss": 0.0288,
"step": 7240
},
{
"epoch": 3.442545109211776,
"grad_norm": 0.07037687702507503,
"learning_rate": 2.6771179857045396e-05,
"loss": 0.0288,
"step": 7250
},
{
"epoch": 3.4472934472934473,
"grad_norm": 0.06423949138505133,
"learning_rate": 2.6624532189561145e-05,
"loss": 0.0296,
"step": 7260
},
{
"epoch": 3.4520417853751186,
"grad_norm": 0.0756689459458659,
"learning_rate": 2.6478141394745524e-05,
"loss": 0.0286,
"step": 7270
},
{
"epoch": 3.45679012345679,
"grad_norm": 0.082529343905233,
"learning_rate": 2.6332009081284754e-05,
"loss": 0.0293,
"step": 7280
},
{
"epoch": 3.4615384615384617,
"grad_norm": 0.07818120833185002,
"learning_rate": 2.61861368550247e-05,
"loss": 0.0286,
"step": 7290
},
{
"epoch": 3.466286799620133,
"grad_norm": 0.0693967985314235,
"learning_rate": 2.6040526318953052e-05,
"loss": 0.0285,
"step": 7300
},
{
"epoch": 3.4710351377018043,
"grad_norm": 0.06277007348076565,
"learning_rate": 2.5895179073181863e-05,
"loss": 0.0278,
"step": 7310
},
{
"epoch": 3.4757834757834756,
"grad_norm": 0.0643168681153013,
"learning_rate": 2.575009671492983e-05,
"loss": 0.0282,
"step": 7320
},
{
"epoch": 3.4805318138651473,
"grad_norm": 0.06743501040952948,
"learning_rate": 2.5605280838504863e-05,
"loss": 0.0292,
"step": 7330
},
{
"epoch": 3.4852801519468186,
"grad_norm": 0.06748321952789942,
"learning_rate": 2.5460733035286432e-05,
"loss": 0.0284,
"step": 7340
},
{
"epoch": 3.49002849002849,
"grad_norm": 0.07182334020934222,
"learning_rate": 2.5316454893708252e-05,
"loss": 0.0284,
"step": 7350
},
{
"epoch": 3.4947768281101617,
"grad_norm": 0.07088771078578752,
"learning_rate": 2.517244799924064e-05,
"loss": 0.0279,
"step": 7360
},
{
"epoch": 3.499525166191833,
"grad_norm": 0.07012506603136587,
"learning_rate": 2.502871393437325e-05,
"loss": 0.0279,
"step": 7370
},
{
"epoch": 3.5042735042735043,
"grad_norm": 0.07054209656579866,
"learning_rate": 2.4885254278597564e-05,
"loss": 0.0291,
"step": 7380
},
{
"epoch": 3.5090218423551756,
"grad_norm": 0.07450759831833657,
"learning_rate": 2.47420706083896e-05,
"loss": 0.0288,
"step": 7390
},
{
"epoch": 3.513770180436847,
"grad_norm": 0.06853352319838514,
"learning_rate": 2.4599164497192617e-05,
"loss": 0.0288,
"step": 7400
},
{
"epoch": 3.5185185185185186,
"grad_norm": 0.062305179610995014,
"learning_rate": 2.4456537515399707e-05,
"loss": 0.0276,
"step": 7410
},
{
"epoch": 3.52326685660019,
"grad_norm": 0.07142472155291188,
"learning_rate": 2.4314191230336702e-05,
"loss": 0.0289,
"step": 7420
},
{
"epoch": 3.528015194681861,
"grad_norm": 0.06329265888527177,
"learning_rate": 2.4172127206244766e-05,
"loss": 0.0279,
"step": 7430
},
{
"epoch": 3.532763532763533,
"grad_norm": 0.06755685850489059,
"learning_rate": 2.4030347004263392e-05,
"loss": 0.0286,
"step": 7440
},
{
"epoch": 3.5375118708452042,
"grad_norm": 0.07663038843664749,
"learning_rate": 2.3888852182413085e-05,
"loss": 0.0288,
"step": 7450
},
{
"epoch": 3.5422602089268755,
"grad_norm": 0.07318447358072278,
"learning_rate": 2.3747644295578364e-05,
"loss": 0.0277,
"step": 7460
},
{
"epoch": 3.547008547008547,
"grad_norm": 0.0685023932099314,
"learning_rate": 2.3606724895490568e-05,
"loss": 0.0279,
"step": 7470
},
{
"epoch": 3.5517568850902186,
"grad_norm": 0.06795699320174689,
"learning_rate": 2.346609553071093e-05,
"loss": 0.0278,
"step": 7480
},
{
"epoch": 3.55650522317189,
"grad_norm": 0.0679934180119207,
"learning_rate": 2.332575774661342e-05,
"loss": 0.0278,
"step": 7490
},
{
"epoch": 3.561253561253561,
"grad_norm": 0.06362170615194565,
"learning_rate": 2.3185713085367866e-05,
"loss": 0.0279,
"step": 7500
},
{
"epoch": 3.566001899335233,
"grad_norm": 0.06657040484201165,
"learning_rate": 2.3045963085923007e-05,
"loss": 0.0284,
"step": 7510
},
{
"epoch": 3.5707502374169042,
"grad_norm": 0.07647567124362589,
"learning_rate": 2.2906509283989462e-05,
"loss": 0.0293,
"step": 7520
},
{
"epoch": 3.5754985754985755,
"grad_norm": 0.06340327420722991,
"learning_rate": 2.276735321202303e-05,
"loss": 0.0292,
"step": 7530
},
{
"epoch": 3.580246913580247,
"grad_norm": 0.07144707493084852,
"learning_rate": 2.2628496399207683e-05,
"loss": 0.0277,
"step": 7540
},
{
"epoch": 3.584995251661918,
"grad_norm": 0.069413616247519,
"learning_rate": 2.2489940371438912e-05,
"loss": 0.029,
"step": 7550
},
{
"epoch": 3.58974358974359,
"grad_norm": 0.06129128098704001,
"learning_rate": 2.2351686651306815e-05,
"loss": 0.0282,
"step": 7560
},
{
"epoch": 3.594491927825261,
"grad_norm": 0.06936169856525251,
"learning_rate": 2.2213736758079494e-05,
"loss": 0.0283,
"step": 7570
},
{
"epoch": 3.5992402659069325,
"grad_norm": 0.06776060040148728,
"learning_rate": 2.207609220768625e-05,
"loss": 0.0285,
"step": 7580
},
{
"epoch": 3.603988603988604,
"grad_norm": 0.06353062294125177,
"learning_rate": 2.1938754512701042e-05,
"loss": 0.0288,
"step": 7590
},
{
"epoch": 3.6087369420702755,
"grad_norm": 0.07127535174113613,
"learning_rate": 2.1801725182325733e-05,
"loss": 0.0278,
"step": 7600
},
{
"epoch": 3.613485280151947,
"grad_norm": 0.07072299157682889,
"learning_rate": 2.1665005722373588e-05,
"loss": 0.0279,
"step": 7610
},
{
"epoch": 3.618233618233618,
"grad_norm": 0.06702139335951722,
"learning_rate": 2.1528597635252756e-05,
"loss": 0.0283,
"step": 7620
},
{
"epoch": 3.6229819563152894,
"grad_norm": 0.06640411110744364,
"learning_rate": 2.139250241994964e-05,
"loss": 0.028,
"step": 7630
},
{
"epoch": 3.627730294396961,
"grad_norm": 0.0727752436814663,
"learning_rate": 2.1256721572012584e-05,
"loss": 0.0282,
"step": 7640
},
{
"epoch": 3.6324786324786325,
"grad_norm": 0.07630520943435848,
"learning_rate": 2.1121256583535272e-05,
"loss": 0.0282,
"step": 7650
},
{
"epoch": 3.6372269705603038,
"grad_norm": 0.07058472891004565,
"learning_rate": 2.098610894314048e-05,
"loss": 0.0278,
"step": 7660
},
{
"epoch": 3.6419753086419755,
"grad_norm": 0.07320687845008718,
"learning_rate": 2.085128013596359e-05,
"loss": 0.028,
"step": 7670
},
{
"epoch": 3.646723646723647,
"grad_norm": 0.07038074182481655,
"learning_rate": 2.07167716436364e-05,
"loss": 0.029,
"step": 7680
},
{
"epoch": 3.651471984805318,
"grad_norm": 0.06807332199144848,
"learning_rate": 2.0582584944270694e-05,
"loss": 0.0276,
"step": 7690
},
{
"epoch": 3.6562203228869894,
"grad_norm": 0.06703244660139986,
"learning_rate": 2.0448721512442153e-05,
"loss": 0.0274,
"step": 7700
},
{
"epoch": 3.6609686609686607,
"grad_norm": 0.0732565429503293,
"learning_rate": 2.031518281917403e-05,
"loss": 0.0286,
"step": 7710
},
{
"epoch": 3.6657169990503324,
"grad_norm": 0.0677186816537058,
"learning_rate": 2.018197033192102e-05,
"loss": 0.0284,
"step": 7720
},
{
"epoch": 3.6704653371320037,
"grad_norm": 0.07112561837948107,
"learning_rate": 2.0049085514553183e-05,
"loss": 0.0288,
"step": 7730
},
{
"epoch": 3.6752136752136755,
"grad_norm": 0.0637292200453614,
"learning_rate": 1.9916529827339825e-05,
"loss": 0.028,
"step": 7740
},
{
"epoch": 3.679962013295347,
"grad_norm": 0.06996170025127464,
"learning_rate": 1.9784304726933383e-05,
"loss": 0.0276,
"step": 7750
},
{
"epoch": 3.684710351377018,
"grad_norm": 0.0691704082096398,
"learning_rate": 1.9652411666353547e-05,
"loss": 0.0284,
"step": 7760
},
{
"epoch": 3.6894586894586894,
"grad_norm": 0.07108695362225634,
"learning_rate": 1.9520852094971215e-05,
"loss": 0.029,
"step": 7770
},
{
"epoch": 3.6942070275403607,
"grad_norm": 0.06983215533239741,
"learning_rate": 1.9389627458492526e-05,
"loss": 0.0286,
"step": 7780
},
{
"epoch": 3.6989553656220324,
"grad_norm": 0.061157748084917585,
"learning_rate": 1.9258739198943105e-05,
"loss": 0.0291,
"step": 7790
},
{
"epoch": 3.7037037037037037,
"grad_norm": 0.07398867200257035,
"learning_rate": 1.9128188754652042e-05,
"loss": 0.0284,
"step": 7800
},
{
"epoch": 3.708452041785375,
"grad_norm": 0.06287539192821646,
"learning_rate": 1.8997977560236276e-05,
"loss": 0.0273,
"step": 7810
},
{
"epoch": 3.7132003798670468,
"grad_norm": 0.06643311931322488,
"learning_rate": 1.886810704658467e-05,
"loss": 0.0278,
"step": 7820
},
{
"epoch": 3.717948717948718,
"grad_norm": 0.07735124691433044,
"learning_rate": 1.8738578640842353e-05,
"loss": 0.0288,
"step": 7830
},
{
"epoch": 3.7226970560303894,
"grad_norm": 0.06263747964901435,
"learning_rate": 1.8609393766395085e-05,
"loss": 0.0276,
"step": 7840
},
{
"epoch": 3.7274453941120607,
"grad_norm": 0.06867389407576382,
"learning_rate": 1.8480553842853492e-05,
"loss": 0.0276,
"step": 7850
},
{
"epoch": 3.732193732193732,
"grad_norm": 0.06969105508854904,
"learning_rate": 1.8352060286037626e-05,
"loss": 0.028,
"step": 7860
},
{
"epoch": 3.7369420702754037,
"grad_norm": 0.06753559178598251,
"learning_rate": 1.8223914507961232e-05,
"loss": 0.0279,
"step": 7870
},
{
"epoch": 3.741690408357075,
"grad_norm": 0.06524685682555084,
"learning_rate": 1.80961179168164e-05,
"loss": 0.0278,
"step": 7880
},
{
"epoch": 3.7464387464387463,
"grad_norm": 0.06649372477216164,
"learning_rate": 1.796867191695794e-05,
"loss": 0.0284,
"step": 7890
},
{
"epoch": 3.751187084520418,
"grad_norm": 0.0678327062351489,
"learning_rate": 1.784157790888808e-05,
"loss": 0.0287,
"step": 7900
},
{
"epoch": 3.7559354226020893,
"grad_norm": 0.06411530187757249,
"learning_rate": 1.7714837289240958e-05,
"loss": 0.028,
"step": 7910
},
{
"epoch": 3.7606837606837606,
"grad_norm": 0.07044659309262939,
"learning_rate": 1.7588451450767396e-05,
"loss": 0.0279,
"step": 7920
},
{
"epoch": 3.765432098765432,
"grad_norm": 0.0676233423823099,
"learning_rate": 1.7462421782319472e-05,
"loss": 0.0292,
"step": 7930
},
{
"epoch": 3.7701804368471032,
"grad_norm": 0.0614035592449899,
"learning_rate": 1.733674966883533e-05,
"loss": 0.028,
"step": 7940
},
{
"epoch": 3.774928774928775,
"grad_norm": 0.072489946966981,
"learning_rate": 1.721143649132399e-05,
"loss": 0.0281,
"step": 7950
},
{
"epoch": 3.7796771130104463,
"grad_norm": 0.06717578754177722,
"learning_rate": 1.7086483626850063e-05,
"loss": 0.0274,
"step": 7960
},
{
"epoch": 3.784425451092118,
"grad_norm": 0.062451226540089354,
"learning_rate": 1.696189244851874e-05,
"loss": 0.0281,
"step": 7970
},
{
"epoch": 3.7891737891737893,
"grad_norm": 0.06075628070442532,
"learning_rate": 1.683766432546065e-05,
"loss": 0.0275,
"step": 7980
},
{
"epoch": 3.7939221272554606,
"grad_norm": 0.06427983365268919,
"learning_rate": 1.671380062281675e-05,
"loss": 0.0288,
"step": 7990
},
{
"epoch": 3.798670465337132,
"grad_norm": 0.06412117484713672,
"learning_rate": 1.6590302701723454e-05,
"loss": 0.0271,
"step": 8000
},
{
"epoch": 3.8034188034188032,
"grad_norm": 0.06964313045517292,
"learning_rate": 1.6467171919297598e-05,
"loss": 0.0277,
"step": 8010
},
{
"epoch": 3.808167141500475,
"grad_norm": 0.06405013701879049,
"learning_rate": 1.6344409628621484e-05,
"loss": 0.0272,
"step": 8020
},
{
"epoch": 3.8129154795821463,
"grad_norm": 0.056343860638701046,
"learning_rate": 1.6222017178728134e-05,
"loss": 0.0278,
"step": 8030
},
{
"epoch": 3.8176638176638176,
"grad_norm": 0.06580988107395615,
"learning_rate": 1.6099995914586347e-05,
"loss": 0.0279,
"step": 8040
},
{
"epoch": 3.8224121557454893,
"grad_norm": 0.06740022999813228,
"learning_rate": 1.5978347177085967e-05,
"loss": 0.0273,
"step": 8050
},
{
"epoch": 3.8271604938271606,
"grad_norm": 0.05918621928025624,
"learning_rate": 1.5857072303023195e-05,
"loss": 0.0284,
"step": 8060
},
{
"epoch": 3.831908831908832,
"grad_norm": 0.06805614255820627,
"learning_rate": 1.5736172625085792e-05,
"loss": 0.0292,
"step": 8070
},
{
"epoch": 3.836657169990503,
"grad_norm": 0.062307796598217224,
"learning_rate": 1.5615649471838555e-05,
"loss": 0.0274,
"step": 8080
},
{
"epoch": 3.8414055080721745,
"grad_norm": 0.06198007192832983,
"learning_rate": 1.5495504167708592e-05,
"loss": 0.0284,
"step": 8090
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.06689727398069602,
"learning_rate": 1.5375738032970893e-05,
"loss": 0.0286,
"step": 8100
},
{
"epoch": 3.8509021842355176,
"grad_norm": 0.07306996711614937,
"learning_rate": 1.5256352383733707e-05,
"loss": 0.0286,
"step": 8110
},
{
"epoch": 3.855650522317189,
"grad_norm": 0.07009952670673278,
"learning_rate": 1.513734853192419e-05,
"loss": 0.0284,
"step": 8120
},
{
"epoch": 3.8603988603988606,
"grad_norm": 0.077919566734238,
"learning_rate": 1.5018727785273872e-05,
"loss": 0.028,
"step": 8130
},
{
"epoch": 3.865147198480532,
"grad_norm": 0.06414890868804238,
"learning_rate": 1.4900491447304409e-05,
"loss": 0.0277,
"step": 8140
},
{
"epoch": 3.869895536562203,
"grad_norm": 0.0602955068054816,
"learning_rate": 1.4782640817313149e-05,
"loss": 0.0279,
"step": 8150
},
{
"epoch": 3.8746438746438745,
"grad_norm": 0.06623613920024082,
"learning_rate": 1.4665177190358886e-05,
"loss": 0.0278,
"step": 8160
},
{
"epoch": 3.879392212725546,
"grad_norm": 0.0634871781311035,
"learning_rate": 1.454810185724772e-05,
"loss": 0.0267,
"step": 8170
},
{
"epoch": 3.8841405508072175,
"grad_norm": 0.06956923646878543,
"learning_rate": 1.4431416104518708e-05,
"loss": 0.0284,
"step": 8180
},
{
"epoch": 3.888888888888889,
"grad_norm": 0.06845788940134993,
"learning_rate": 1.4315121214429882e-05,
"loss": 0.0268,
"step": 8190
},
{
"epoch": 3.89363722697056,
"grad_norm": 0.06994409080553063,
"learning_rate": 1.4199218464944042e-05,
"loss": 0.0285,
"step": 8200
},
{
"epoch": 3.898385565052232,
"grad_norm": 0.06994044804983268,
"learning_rate": 1.40837091297148e-05,
"loss": 0.0276,
"step": 8210
},
{
"epoch": 3.903133903133903,
"grad_norm": 0.06102833866717814,
"learning_rate": 1.39685944780725e-05,
"loss": 0.0278,
"step": 8220
},
{
"epoch": 3.9078822412155745,
"grad_norm": 0.07102581671154552,
"learning_rate": 1.3853875775010355e-05,
"loss": 0.0287,
"step": 8230
},
{
"epoch": 3.912630579297246,
"grad_norm": 0.06769416330797612,
"learning_rate": 1.3739554281170509e-05,
"loss": 0.0276,
"step": 8240
},
{
"epoch": 3.9173789173789175,
"grad_norm": 0.06694300660402334,
"learning_rate": 1.3625631252830118e-05,
"loss": 0.0279,
"step": 8250
},
{
"epoch": 3.922127255460589,
"grad_norm": 0.0675326753406088,
"learning_rate": 1.3512107941887686e-05,
"loss": 0.0274,
"step": 8260
},
{
"epoch": 3.92687559354226,
"grad_norm": 0.0664704902474989,
"learning_rate": 1.3398985595849156e-05,
"loss": 0.0272,
"step": 8270
},
{
"epoch": 3.931623931623932,
"grad_norm": 0.06998266159789487,
"learning_rate": 1.3286265457814345e-05,
"loss": 0.0278,
"step": 8280
},
{
"epoch": 3.936372269705603,
"grad_norm": 0.07563362295605276,
"learning_rate": 1.3173948766463145e-05,
"loss": 0.0279,
"step": 8290
},
{
"epoch": 3.9411206077872745,
"grad_norm": 0.0607151926275586,
"learning_rate": 1.3062036756042034e-05,
"loss": 0.0287,
"step": 8300
},
{
"epoch": 3.9458689458689458,
"grad_norm": 0.07195747798254198,
"learning_rate": 1.2950530656350413e-05,
"loss": 0.0279,
"step": 8310
},
{
"epoch": 3.950617283950617,
"grad_norm": 0.06944762356243943,
"learning_rate": 1.2839431692727172e-05,
"loss": 0.0272,
"step": 8320
},
{
"epoch": 3.955365622032289,
"grad_norm": 0.08041387285717223,
"learning_rate": 1.2728741086037149e-05,
"loss": 0.0293,
"step": 8330
},
{
"epoch": 3.96011396011396,
"grad_norm": 0.06484707262961723,
"learning_rate": 1.2618460052657804e-05,
"loss": 0.0272,
"step": 8340
},
{
"epoch": 3.9648622981956314,
"grad_norm": 0.07180205989946953,
"learning_rate": 1.2508589804465765e-05,
"loss": 0.0285,
"step": 8350
},
{
"epoch": 3.969610636277303,
"grad_norm": 0.0627713125644083,
"learning_rate": 1.2399131548823539e-05,
"loss": 0.0279,
"step": 8360
},
{
"epoch": 3.9743589743589745,
"grad_norm": 0.059587435412033415,
"learning_rate": 1.2290086488566305e-05,
"loss": 0.0273,
"step": 8370
},
{
"epoch": 3.9791073124406457,
"grad_norm": 0.0658702053954759,
"learning_rate": 1.2181455821988574e-05,
"loss": 0.0273,
"step": 8380
},
{
"epoch": 3.983855650522317,
"grad_norm": 0.06586150645638703,
"learning_rate": 1.2073240742831165e-05,
"loss": 0.0278,
"step": 8390
},
{
"epoch": 3.9886039886039883,
"grad_norm": 0.07286519229980391,
"learning_rate": 1.1965442440267943e-05,
"loss": 0.0288,
"step": 8400
},
{
"epoch": 3.99335232668566,
"grad_norm": 0.06341460333334668,
"learning_rate": 1.185806209889288e-05,
"loss": 0.0281,
"step": 8410
},
{
"epoch": 3.9981006647673314,
"grad_norm": 0.059913195391255517,
"learning_rate": 1.1751100898706929e-05,
"loss": 0.0279,
"step": 8420
},
{
"epoch": 4.002849002849003,
"grad_norm": 0.06841128914194776,
"learning_rate": 1.1644560015105155e-05,
"loss": 0.0267,
"step": 8430
},
{
"epoch": 4.007597340930674,
"grad_norm": 0.08208997466977529,
"learning_rate": 1.1538440618863722e-05,
"loss": 0.0251,
"step": 8440
},
{
"epoch": 4.012345679012346,
"grad_norm": 0.06210086067972121,
"learning_rate": 1.1432743876127122e-05,
"loss": 0.0254,
"step": 8450
},
{
"epoch": 4.017094017094017,
"grad_norm": 0.06648965472773795,
"learning_rate": 1.1327470948395263e-05,
"loss": 0.0256,
"step": 8460
},
{
"epoch": 4.021842355175688,
"grad_norm": 0.06344851944190863,
"learning_rate": 1.1222622992510807e-05,
"loss": 0.0257,
"step": 8470
},
{
"epoch": 4.02659069325736,
"grad_norm": 0.06741732261641768,
"learning_rate": 1.1118201160646386e-05,
"loss": 0.0248,
"step": 8480
},
{
"epoch": 4.031339031339031,
"grad_norm": 0.06332912481015883,
"learning_rate": 1.1014206600291938e-05,
"loss": 0.0254,
"step": 8490
},
{
"epoch": 4.036087369420703,
"grad_norm": 0.06221965101945773,
"learning_rate": 1.0910640454242172e-05,
"loss": 0.0251,
"step": 8500
},
{
"epoch": 4.040835707502374,
"grad_norm": 0.06374866576514718,
"learning_rate": 1.0807503860583895e-05,
"loss": 0.0258,
"step": 8510
},
{
"epoch": 4.045584045584046,
"grad_norm": 0.06392881075096923,
"learning_rate": 1.0704797952683638e-05,
"loss": 0.0254,
"step": 8520
},
{
"epoch": 4.050332383665717,
"grad_norm": 0.06034602609742372,
"learning_rate": 1.0602523859175061e-05,
"loss": 0.0257,
"step": 8530
},
{
"epoch": 4.055080721747388,
"grad_norm": 0.065664429527023,
"learning_rate": 1.0500682703946695e-05,
"loss": 0.0254,
"step": 8540
},
{
"epoch": 4.05982905982906,
"grad_norm": 0.06582325613940465,
"learning_rate": 1.0399275606129449e-05,
"loss": 0.0261,
"step": 8550
},
{
"epoch": 4.064577397910731,
"grad_norm": 0.0634794517684744,
"learning_rate": 1.0298303680084448e-05,
"loss": 0.0252,
"step": 8560
},
{
"epoch": 4.069325735992402,
"grad_norm": 0.061541254952135874,
"learning_rate": 1.0197768035390665e-05,
"loss": 0.0254,
"step": 8570
},
{
"epoch": 4.074074074074074,
"grad_norm": 0.06101154979644648,
"learning_rate": 1.0097669776832801e-05,
"loss": 0.0252,
"step": 8580
},
{
"epoch": 4.078822412155746,
"grad_norm": 0.06572844377676408,
"learning_rate": 9.998010004389158e-06,
"loss": 0.026,
"step": 8590
},
{
"epoch": 4.083570750237417,
"grad_norm": 0.06830451908826547,
"learning_rate": 9.898789813219467e-06,
"loss": 0.0247,
"step": 8600
},
{
"epoch": 4.088319088319088,
"grad_norm": 0.06442396777181708,
"learning_rate": 9.800010293652962e-06,
"loss": 0.0249,
"step": 8610
},
{
"epoch": 4.09306742640076,
"grad_norm": 0.06613770257220306,
"learning_rate": 9.701672531176286e-06,
"loss": 0.0259,
"step": 8620
},
{
"epoch": 4.097815764482431,
"grad_norm": 0.06951304130764932,
"learning_rate": 9.603777606421682e-06,
"loss": 0.0254,
"step": 8630
},
{
"epoch": 4.102564102564102,
"grad_norm": 0.06634918180055019,
"learning_rate": 9.506326595155002e-06,
"loss": 0.0248,
"step": 8640
},
{
"epoch": 4.107312440645774,
"grad_norm": 0.058667178244782556,
"learning_rate": 9.40932056826398e-06,
"loss": 0.0248,
"step": 8650
},
{
"epoch": 4.112060778727446,
"grad_norm": 0.07364579006677409,
"learning_rate": 9.312760591746377e-06,
"loss": 0.0257,
"step": 8660
},
{
"epoch": 4.116809116809117,
"grad_norm": 0.06817532446844374,
"learning_rate": 9.216647726698374e-06,
"loss": 0.0255,
"step": 8670
},
{
"epoch": 4.121557454890788,
"grad_norm": 0.06660575670950503,
"learning_rate": 9.120983029302805e-06,
"loss": 0.0254,
"step": 8680
},
{
"epoch": 4.12630579297246,
"grad_norm": 0.07265358708985209,
"learning_rate": 9.025767550817588e-06,
"loss": 0.026,
"step": 8690
},
{
"epoch": 4.131054131054131,
"grad_norm": 0.06894787969474898,
"learning_rate": 8.93100233756422e-06,
"loss": 0.0269,
"step": 8700
},
{
"epoch": 4.135802469135802,
"grad_norm": 0.0627728598788623,
"learning_rate": 8.83668843091624e-06,
"loss": 0.0255,
"step": 8710
},
{
"epoch": 4.1405508072174735,
"grad_norm": 0.06312069608462036,
"learning_rate": 8.74282686728773e-06,
"loss": 0.0251,
"step": 8720
},
{
"epoch": 4.145299145299146,
"grad_norm": 0.06565965904188746,
"learning_rate": 8.64941867812204e-06,
"loss": 0.0256,
"step": 8730
},
{
"epoch": 4.150047483380817,
"grad_norm": 0.06630717041023335,
"learning_rate": 8.556464889880383e-06,
"loss": 0.0254,
"step": 8740
},
{
"epoch": 4.154795821462488,
"grad_norm": 0.07049231654507018,
"learning_rate": 8.46396652403053e-06,
"loss": 0.0247,
"step": 8750
},
{
"epoch": 4.15954415954416,
"grad_norm": 0.07060945271667128,
"learning_rate": 8.371924597035685e-06,
"loss": 0.0255,
"step": 8760
},
{
"epoch": 4.164292497625831,
"grad_norm": 0.06079428604107492,
"learning_rate": 8.280340120343182e-06,
"loss": 0.0249,
"step": 8770
},
{
"epoch": 4.169040835707502,
"grad_norm": 0.0629223478067205,
"learning_rate": 8.189214100373499e-06,
"loss": 0.0254,
"step": 8780
},
{
"epoch": 4.1737891737891735,
"grad_norm": 0.06418850823028961,
"learning_rate": 8.09854753850911e-06,
"loss": 0.0257,
"step": 8790
},
{
"epoch": 4.178537511870845,
"grad_norm": 0.06905664001807565,
"learning_rate": 8.0083414310835e-06,
"loss": 0.0251,
"step": 8800
},
{
"epoch": 4.183285849952517,
"grad_norm": 0.06497332822023333,
"learning_rate": 7.918596769370273e-06,
"loss": 0.0249,
"step": 8810
},
{
"epoch": 4.188034188034188,
"grad_norm": 0.0676601789262791,
"learning_rate": 7.829314539572152e-06,
"loss": 0.0262,
"step": 8820
},
{
"epoch": 4.1927825261158596,
"grad_norm": 0.060325423826822565,
"learning_rate": 7.740495722810271e-06,
"loss": 0.0251,
"step": 8830
},
{
"epoch": 4.197530864197531,
"grad_norm": 0.07444900442367944,
"learning_rate": 7.652141295113246e-06,
"loss": 0.0255,
"step": 8840
},
{
"epoch": 4.202279202279202,
"grad_norm": 0.07342840461397124,
"learning_rate": 7.56425222740661e-06,
"loss": 0.0255,
"step": 8850
},
{
"epoch": 4.2070275403608735,
"grad_norm": 0.06819313559756843,
"learning_rate": 7.476829485501996e-06,
"loss": 0.0251,
"step": 8860
},
{
"epoch": 4.211775878442545,
"grad_norm": 0.0643645546296177,
"learning_rate": 7.389874030086636e-06,
"loss": 0.0246,
"step": 8870
},
{
"epoch": 4.216524216524217,
"grad_norm": 0.07092471831391034,
"learning_rate": 7.303386816712715e-06,
"loss": 0.0254,
"step": 8880
},
{
"epoch": 4.221272554605888,
"grad_norm": 0.060971436894407094,
"learning_rate": 7.217368795786961e-06,
"loss": 0.0249,
"step": 8890
},
{
"epoch": 4.2260208926875595,
"grad_norm": 0.06409181439047049,
"learning_rate": 7.131820912560116e-06,
"loss": 0.0248,
"step": 8900
},
{
"epoch": 4.230769230769231,
"grad_norm": 0.06520444809087475,
"learning_rate": 7.046744107116587e-06,
"loss": 0.025,
"step": 8910
},
{
"epoch": 4.235517568850902,
"grad_norm": 0.06712232516044428,
"learning_rate": 6.962139314364152e-06,
"loss": 0.0253,
"step": 8920
},
{
"epoch": 4.240265906932573,
"grad_norm": 0.07386855383331557,
"learning_rate": 6.878007464023595e-06,
"loss": 0.0255,
"step": 8930
},
{
"epoch": 4.245014245014245,
"grad_norm": 0.0605452724781097,
"learning_rate": 6.794349480618583e-06,
"loss": 0.0248,
"step": 8940
},
{
"epoch": 4.249762583095916,
"grad_norm": 0.06604900407713182,
"learning_rate": 6.711166283465464e-06,
"loss": 0.0256,
"step": 8950
},
{
"epoch": 4.254510921177588,
"grad_norm": 0.07298172396694141,
"learning_rate": 6.628458786663133e-06,
"loss": 0.0255,
"step": 8960
},
{
"epoch": 4.2592592592592595,
"grad_norm": 0.07169247874879303,
"learning_rate": 6.546227899083062e-06,
"loss": 0.0248,
"step": 8970
},
{
"epoch": 4.264007597340931,
"grad_norm": 0.0688338909583261,
"learning_rate": 6.464474524359243e-06,
"loss": 0.0248,
"step": 8980
},
{
"epoch": 4.268755935422602,
"grad_norm": 0.0680145825576697,
"learning_rate": 6.3831995608783006e-06,
"loss": 0.0255,
"step": 8990
},
{
"epoch": 4.273504273504273,
"grad_norm": 0.06589679259526002,
"learning_rate": 6.30240390176961e-06,
"loss": 0.0261,
"step": 9000
},
{
"epoch": 4.278252611585945,
"grad_norm": 0.06559203905306465,
"learning_rate": 6.222088434895462e-06,
"loss": 0.0245,
"step": 9010
},
{
"epoch": 4.283000949667616,
"grad_norm": 0.06912173907104972,
"learning_rate": 6.142254042841317e-06,
"loss": 0.0256,
"step": 9020
},
{
"epoch": 4.287749287749287,
"grad_norm": 0.0615555586473594,
"learning_rate": 6.062901602906146e-06,
"loss": 0.0252,
"step": 9030
},
{
"epoch": 4.2924976258309595,
"grad_norm": 0.06265905701941747,
"learning_rate": 5.9840319870927085e-06,
"loss": 0.0243,
"step": 9040
},
{
"epoch": 4.297245963912631,
"grad_norm": 0.0706183903870341,
"learning_rate": 5.905646062098069e-06,
"loss": 0.0252,
"step": 9050
},
{
"epoch": 4.301994301994302,
"grad_norm": 0.07465478161883318,
"learning_rate": 5.827744689303965e-06,
"loss": 0.025,
"step": 9060
},
{
"epoch": 4.306742640075973,
"grad_norm": 0.06507708145406323,
"learning_rate": 5.750328724767451e-06,
"loss": 0.0251,
"step": 9070
},
{
"epoch": 4.311490978157645,
"grad_norm": 0.0665017862203972,
"learning_rate": 5.6733990192113895e-06,
"loss": 0.0253,
"step": 9080
},
{
"epoch": 4.316239316239316,
"grad_norm": 0.06937938866456321,
"learning_rate": 5.5969564180151865e-06,
"loss": 0.0247,
"step": 9090
},
{
"epoch": 4.320987654320987,
"grad_norm": 0.06556942554157015,
"learning_rate": 5.521001761205441e-06,
"loss": 0.0255,
"step": 9100
},
{
"epoch": 4.3257359924026595,
"grad_norm": 0.06706832057544795,
"learning_rate": 5.445535883446767e-06,
"loss": 0.025,
"step": 9110
},
{
"epoch": 4.330484330484331,
"grad_norm": 0.07079713109662138,
"learning_rate": 5.370559614032561e-06,
"loss": 0.0254,
"step": 9120
},
{
"epoch": 4.335232668566002,
"grad_norm": 0.06626148188639107,
"learning_rate": 5.296073776875921e-06,
"loss": 0.0255,
"step": 9130
},
{
"epoch": 4.339981006647673,
"grad_norm": 0.07230503183967142,
"learning_rate": 5.222079190500634e-06,
"loss": 0.0253,
"step": 9140
},
{
"epoch": 4.344729344729345,
"grad_norm": 0.06769998918991922,
"learning_rate": 5.1485766680320845e-06,
"loss": 0.0253,
"step": 9150
},
{
"epoch": 4.349477682811016,
"grad_norm": 0.07613019893817513,
"learning_rate": 5.075567017188421e-06,
"loss": 0.0254,
"step": 9160
},
{
"epoch": 4.354226020892687,
"grad_norm": 0.06027294626875827,
"learning_rate": 5.003051040271589e-06,
"loss": 0.0254,
"step": 9170
},
{
"epoch": 4.358974358974359,
"grad_norm": 0.06362331722080705,
"learning_rate": 4.93102953415861e-06,
"loss": 0.0251,
"step": 9180
},
{
"epoch": 4.363722697056031,
"grad_norm": 0.06444979103090218,
"learning_rate": 4.8595032902927216e-06,
"loss": 0.0249,
"step": 9190
},
{
"epoch": 4.368471035137702,
"grad_norm": 0.06406091220140214,
"learning_rate": 4.788473094674762e-06,
"loss": 0.0253,
"step": 9200
},
{
"epoch": 4.373219373219373,
"grad_norm": 0.0740321644775331,
"learning_rate": 4.717939727854515e-06,
"loss": 0.0245,
"step": 9210
},
{
"epoch": 4.377967711301045,
"grad_norm": 0.07288741793201965,
"learning_rate": 4.647903964922068e-06,
"loss": 0.0256,
"step": 9220
},
{
"epoch": 4.382716049382716,
"grad_norm": 0.06395030606373239,
"learning_rate": 4.57836657549941e-06,
"loss": 0.0254,
"step": 9230
},
{
"epoch": 4.387464387464387,
"grad_norm": 0.07111991418406609,
"learning_rate": 4.509328323731854e-06,
"loss": 0.0252,
"step": 9240
},
{
"epoch": 4.392212725546059,
"grad_norm": 0.06406139978870073,
"learning_rate": 4.440789968279746e-06,
"loss": 0.0254,
"step": 9250
},
{
"epoch": 4.39696106362773,
"grad_norm": 0.06216422684666234,
"learning_rate": 4.372752262310037e-06,
"loss": 0.0243,
"step": 9260
},
{
"epoch": 4.401709401709402,
"grad_norm": 0.06927548111919477,
"learning_rate": 4.30521595348809e-06,
"loss": 0.0254,
"step": 9270
},
{
"epoch": 4.406457739791073,
"grad_norm": 0.06125582301243834,
"learning_rate": 4.238181783969375e-06,
"loss": 0.0246,
"step": 9280
},
{
"epoch": 4.411206077872745,
"grad_norm": 0.07139707832718938,
"learning_rate": 4.171650490391405e-06,
"loss": 0.0248,
"step": 9290
},
{
"epoch": 4.415954415954416,
"grad_norm": 0.06281406071472967,
"learning_rate": 4.105622803865561e-06,
"loss": 0.0253,
"step": 9300
},
{
"epoch": 4.420702754036087,
"grad_norm": 0.0616492176985963,
"learning_rate": 4.040099449969126e-06,
"loss": 0.0255,
"step": 9310
},
{
"epoch": 4.4254510921177586,
"grad_norm": 0.07520886403940373,
"learning_rate": 3.975081148737253e-06,
"loss": 0.0264,
"step": 9320
},
{
"epoch": 4.43019943019943,
"grad_norm": 0.06572296276442756,
"learning_rate": 3.910568614655086e-06,
"loss": 0.0254,
"step": 9330
},
{
"epoch": 4.434947768281102,
"grad_norm": 0.07216469767021776,
"learning_rate": 3.846562556649913e-06,
"loss": 0.0249,
"step": 9340
},
{
"epoch": 4.439696106362773,
"grad_norm": 0.06830053062879306,
"learning_rate": 3.7830636780833375e-06,
"loss": 0.0245,
"step": 9350
},
{
"epoch": 4.444444444444445,
"grad_norm": 0.06614195498127927,
"learning_rate": 3.720072676743619e-06,
"loss": 0.0249,
"step": 9360
},
{
"epoch": 4.449192782526116,
"grad_norm": 0.0785052790319247,
"learning_rate": 3.6575902448379105e-06,
"loss": 0.0259,
"step": 9370
},
{
"epoch": 4.453941120607787,
"grad_norm": 0.0826769854229507,
"learning_rate": 3.5956170689847525e-06,
"loss": 0.0259,
"step": 9380
},
{
"epoch": 4.4586894586894585,
"grad_norm": 0.07326375814030694,
"learning_rate": 3.534153830206427e-06,
"loss": 0.0252,
"step": 9390
},
{
"epoch": 4.46343779677113,
"grad_norm": 0.05969482566103077,
"learning_rate": 3.4732012039215776e-06,
"loss": 0.0257,
"step": 9400
},
{
"epoch": 4.468186134852801,
"grad_norm": 0.07431570524019311,
"learning_rate": 3.4127598599376866e-06,
"loss": 0.0257,
"step": 9410
},
{
"epoch": 4.472934472934473,
"grad_norm": 0.07248472038162904,
"learning_rate": 3.352830462443807e-06,
"loss": 0.0249,
"step": 9420
},
{
"epoch": 4.477682811016145,
"grad_norm": 0.06743436137533572,
"learning_rate": 3.2934136700031814e-06,
"loss": 0.0256,
"step": 9430
},
{
"epoch": 4.482431149097816,
"grad_norm": 0.0691948815425552,
"learning_rate": 3.2345101355460704e-06,
"loss": 0.0247,
"step": 9440
},
{
"epoch": 4.487179487179487,
"grad_norm": 0.07663872722493385,
"learning_rate": 3.176120506362529e-06,
"loss": 0.0249,
"step": 9450
},
{
"epoch": 4.4919278252611585,
"grad_norm": 0.07591452668931536,
"learning_rate": 3.1182454240953286e-06,
"loss": 0.0248,
"step": 9460
},
{
"epoch": 4.49667616334283,
"grad_norm": 0.06486615246601676,
"learning_rate": 3.060885524732904e-06,
"loss": 0.0244,
"step": 9470
},
{
"epoch": 4.501424501424501,
"grad_norm": 0.06637554413995865,
"learning_rate": 3.0040414386023117e-06,
"loss": 0.0248,
"step": 9480
},
{
"epoch": 4.506172839506172,
"grad_norm": 0.06631861254315682,
"learning_rate": 2.9477137903623887e-06,
"loss": 0.0257,
"step": 9490
},
{
"epoch": 4.510921177587845,
"grad_norm": 0.06546079767628887,
"learning_rate": 2.891903198996815e-06,
"loss": 0.0254,
"step": 9500
},
{
"epoch": 4.515669515669516,
"grad_norm": 0.06947977216858249,
"learning_rate": 2.8366102778073554e-06,
"loss": 0.0249,
"step": 9510
},
{
"epoch": 4.520417853751187,
"grad_norm": 0.06310235803905616,
"learning_rate": 2.7818356344070985e-06,
"loss": 0.025,
"step": 9520
},
{
"epoch": 4.5251661918328585,
"grad_norm": 0.07480882845364888,
"learning_rate": 2.7275798707137866e-06,
"loss": 0.0261,
"step": 9530
},
{
"epoch": 4.52991452991453,
"grad_norm": 0.07298897316289958,
"learning_rate": 2.6738435829432085e-06,
"loss": 0.0244,
"step": 9540
},
{
"epoch": 4.534662867996201,
"grad_norm": 0.07074944779859509,
"learning_rate": 2.6206273616026123e-06,
"loss": 0.0246,
"step": 9550
},
{
"epoch": 4.539411206077872,
"grad_norm": 0.06556331854118135,
"learning_rate": 2.567931791484285e-06,
"loss": 0.0249,
"step": 9560
},
{
"epoch": 4.544159544159545,
"grad_norm": 0.06132795606913437,
"learning_rate": 2.515757451659051e-06,
"loss": 0.0246,
"step": 9570
},
{
"epoch": 4.548907882241216,
"grad_norm": 0.06378200673195784,
"learning_rate": 2.464104915469967e-06,
"loss": 0.0247,
"step": 9580
},
{
"epoch": 4.553656220322887,
"grad_norm": 0.07195327186244141,
"learning_rate": 2.4129747505259703e-06,
"loss": 0.0254,
"step": 9590
},
{
"epoch": 4.5584045584045585,
"grad_norm": 0.0691279506149155,
"learning_rate": 2.3623675186957073e-06,
"loss": 0.0253,
"step": 9600
},
{
"epoch": 4.56315289648623,
"grad_norm": 0.06470035416495518,
"learning_rate": 2.3122837761012707e-06,
"loss": 0.0243,
"step": 9610
},
{
"epoch": 4.567901234567901,
"grad_norm": 0.06882618352608165,
"learning_rate": 2.2627240731121822e-06,
"loss": 0.0257,
"step": 9620
},
{
"epoch": 4.572649572649572,
"grad_norm": 0.06654857924250075,
"learning_rate": 2.2136889543392715e-06,
"loss": 0.0253,
"step": 9630
},
{
"epoch": 4.577397910731244,
"grad_norm": 0.06649542845630631,
"learning_rate": 2.1651789586287442e-06,
"loss": 0.0254,
"step": 9640
},
{
"epoch": 4.582146248812915,
"grad_norm": 0.07118158360798947,
"learning_rate": 2.117194619056212e-06,
"loss": 0.0252,
"step": 9650
},
{
"epoch": 4.586894586894587,
"grad_norm": 0.07112510977428659,
"learning_rate": 2.0697364629208626e-06,
"loss": 0.0253,
"step": 9660
},
{
"epoch": 4.5916429249762585,
"grad_norm": 0.06592780711488014,
"learning_rate": 2.0228050117396813e-06,
"loss": 0.025,
"step": 9670
},
{
"epoch": 4.59639126305793,
"grad_norm": 0.06706525628405442,
"learning_rate": 1.976400781241694e-06,
"loss": 0.0255,
"step": 9680
},
{
"epoch": 4.601139601139601,
"grad_norm": 0.06863212451055198,
"learning_rate": 1.930524281362278e-06,
"loss": 0.0251,
"step": 9690
},
{
"epoch": 4.605887939221272,
"grad_norm": 0.0720673293205912,
"learning_rate": 1.8851760162376275e-06,
"loss": 0.025,
"step": 9700
},
{
"epoch": 4.610636277302944,
"grad_norm": 0.07015700102631829,
"learning_rate": 1.8403564841991515e-06,
"loss": 0.025,
"step": 9710
},
{
"epoch": 4.615384615384615,
"grad_norm": 0.06642487287083568,
"learning_rate": 1.7960661777680188e-06,
"loss": 0.0255,
"step": 9720
},
{
"epoch": 4.620132953466287,
"grad_norm": 0.06442386180455718,
"learning_rate": 1.752305583649755e-06,
"loss": 0.0258,
"step": 9730
},
{
"epoch": 4.6248812915479585,
"grad_norm": 0.0614154986085382,
"learning_rate": 1.7090751827288643e-06,
"loss": 0.025,
"step": 9740
},
{
"epoch": 4.62962962962963,
"grad_norm": 0.06491000908719136,
"learning_rate": 1.6663754500635954e-06,
"loss": 0.0245,
"step": 9750
},
{
"epoch": 4.634377967711301,
"grad_norm": 0.06963988650052566,
"learning_rate": 1.6242068548806777e-06,
"loss": 0.0253,
"step": 9760
},
{
"epoch": 4.639126305792972,
"grad_norm": 0.0661456143798181,
"learning_rate": 1.58256986057016e-06,
"loss": 0.0243,
"step": 9770
},
{
"epoch": 4.643874643874644,
"grad_norm": 0.06863027293205566,
"learning_rate": 1.5414649246803693e-06,
"loss": 0.0244,
"step": 9780
},
{
"epoch": 4.648622981956315,
"grad_norm": 0.06685999125664335,
"learning_rate": 1.5008924989128258e-06,
"loss": 0.0254,
"step": 9790
},
{
"epoch": 4.653371320037987,
"grad_norm": 0.06983933617582107,
"learning_rate": 1.4608530291173317e-06,
"loss": 0.0251,
"step": 9800
},
{
"epoch": 4.6581196581196584,
"grad_norm": 0.07005814727869818,
"learning_rate": 1.4213469552870006e-06,
"loss": 0.0247,
"step": 9810
},
{
"epoch": 4.66286799620133,
"grad_norm": 0.07483289933592781,
"learning_rate": 1.3823747115535135e-06,
"loss": 0.0253,
"step": 9820
},
{
"epoch": 4.667616334283001,
"grad_norm": 0.06641323714838579,
"learning_rate": 1.3439367261822599e-06,
"loss": 0.0263,
"step": 9830
},
{
"epoch": 4.672364672364672,
"grad_norm": 0.06414615479281954,
"learning_rate": 1.306033421567704e-06,
"loss": 0.0247,
"step": 9840
},
{
"epoch": 4.677113010446344,
"grad_norm": 0.06633480968620521,
"learning_rate": 1.2686652142286926e-06,
"loss": 0.0254,
"step": 9850
},
{
"epoch": 4.681861348528015,
"grad_norm": 0.07236639977675421,
"learning_rate": 1.2318325148038989e-06,
"loss": 0.0248,
"step": 9860
},
{
"epoch": 4.686609686609686,
"grad_norm": 0.07305524827886428,
"learning_rate": 1.1955357280473145e-06,
"loss": 0.0259,
"step": 9870
},
{
"epoch": 4.6913580246913575,
"grad_norm": 0.06716743698418563,
"learning_rate": 1.159775252823786e-06,
"loss": 0.0246,
"step": 9880
},
{
"epoch": 4.69610636277303,
"grad_norm": 0.0611337308522803,
"learning_rate": 1.1245514821046467e-06,
"loss": 0.0251,
"step": 9890
},
{
"epoch": 4.700854700854701,
"grad_norm": 0.07029192609969008,
"learning_rate": 1.0898648029633807e-06,
"loss": 0.0249,
"step": 9900
},
{
"epoch": 4.705603038936372,
"grad_norm": 0.07133294495027631,
"learning_rate": 1.055715596571405e-06,
"loss": 0.0257,
"step": 9910
},
{
"epoch": 4.710351377018044,
"grad_norm": 0.06801034430456666,
"learning_rate": 1.0221042381938327e-06,
"loss": 0.0253,
"step": 9920
},
{
"epoch": 4.715099715099715,
"grad_norm": 0.0685219635926125,
"learning_rate": 9.890310971853833e-07,
"loss": 0.0251,
"step": 9930
},
{
"epoch": 4.719848053181386,
"grad_norm": 0.06676652511236601,
"learning_rate": 9.56496536986312e-07,
"loss": 0.0247,
"step": 9940
},
{
"epoch": 4.7245963912630575,
"grad_norm": 0.06985463790341669,
"learning_rate": 9.245009151184203e-07,
"loss": 0.0251,
"step": 9950
},
{
"epoch": 4.72934472934473,
"grad_norm": 0.06583483033576405,
"learning_rate": 8.930445831811129e-07,
"loss": 0.0248,
"step": 9960
},
{
"epoch": 4.734093067426401,
"grad_norm": 0.06300399725563778,
"learning_rate": 8.62127886847569e-07,
"loss": 0.0247,
"step": 9970
},
{
"epoch": 4.738841405508072,
"grad_norm": 0.06129466967588919,
"learning_rate": 8.317511658609e-07,
"loss": 0.0256,
"step": 9980
},
{
"epoch": 4.743589743589744,
"grad_norm": 0.0695501320615958,
"learning_rate": 8.019147540304306e-07,
"loss": 0.0248,
"step": 9990
},
{
"epoch": 4.748338081671415,
"grad_norm": 0.07850085898476744,
"learning_rate": 7.726189792280625e-07,
"loss": 0.0252,
"step": 10000
},
{
"epoch": 4.753086419753086,
"grad_norm": 0.06541283243733857,
"learning_rate": 7.438641633846166e-07,
"loss": 0.0248,
"step": 10010
},
{
"epoch": 4.7578347578347575,
"grad_norm": 0.061703586177595665,
"learning_rate": 7.156506224863468e-07,
"loss": 0.0246,
"step": 10020
},
{
"epoch": 4.76258309591643,
"grad_norm": 0.07226318210726161,
"learning_rate": 6.879786665714316e-07,
"loss": 0.0247,
"step": 10030
},
{
"epoch": 4.767331433998101,
"grad_norm": 0.07256474290079169,
"learning_rate": 6.608485997265878e-07,
"loss": 0.025,
"step": 10040
},
{
"epoch": 4.772079772079772,
"grad_norm": 0.061195816896114356,
"learning_rate": 6.34260720083707e-07,
"loss": 0.0252,
"step": 10050
},
{
"epoch": 4.776828110161444,
"grad_norm": 0.06208984075735179,
"learning_rate": 6.08215319816613e-07,
"loss": 0.0249,
"step": 10060
},
{
"epoch": 4.781576448243115,
"grad_norm": 0.07424512779926115,
"learning_rate": 5.827126851378206e-07,
"loss": 0.0252,
"step": 10070
},
{
"epoch": 4.786324786324786,
"grad_norm": 0.06330004490025186,
"learning_rate": 5.577530962953992e-07,
"loss": 0.0243,
"step": 10080
},
{
"epoch": 4.7910731244064575,
"grad_norm": 0.06620938421180654,
"learning_rate": 5.33336827569908e-07,
"loss": 0.0245,
"step": 10090
},
{
"epoch": 4.795821462488129,
"grad_norm": 0.06646949803341141,
"learning_rate": 5.09464147271349e-07,
"loss": 0.0249,
"step": 10100
},
{
"epoch": 4.8005698005698,
"grad_norm": 0.06625100249133577,
"learning_rate": 4.861353177362638e-07,
"loss": 0.025,
"step": 10110
},
{
"epoch": 4.805318138651472,
"grad_norm": 0.06370682342909428,
"learning_rate": 4.633505953248074e-07,
"loss": 0.0245,
"step": 10120
},
{
"epoch": 4.810066476733144,
"grad_norm": 0.07526857914918711,
"learning_rate": 4.411102304179571e-07,
"loss": 0.0253,
"step": 10130
},
{
"epoch": 4.814814814814815,
"grad_norm": 0.06575856269641163,
"learning_rate": 4.1941446741475267e-07,
"loss": 0.025,
"step": 10140
},
{
"epoch": 4.819563152896486,
"grad_norm": 0.06676437858912253,
"learning_rate": 3.9826354472961016e-07,
"loss": 0.0257,
"step": 10150
},
{
"epoch": 4.8243114909781575,
"grad_norm": 0.06873607311211577,
"learning_rate": 3.7765769478970704e-07,
"loss": 0.0249,
"step": 10160
},
{
"epoch": 4.829059829059829,
"grad_norm": 0.0670367040088164,
"learning_rate": 3.575971440324233e-07,
"loss": 0.0247,
"step": 10170
},
{
"epoch": 4.8338081671415,
"grad_norm": 0.06571127424646639,
"learning_rate": 3.380821129028489e-07,
"loss": 0.0246,
"step": 10180
},
{
"epoch": 4.838556505223172,
"grad_norm": 0.06894855679198632,
"learning_rate": 3.191128158513745e-07,
"loss": 0.0247,
"step": 10190
},
{
"epoch": 4.843304843304844,
"grad_norm": 0.07429488832441244,
"learning_rate": 3.0068946133132713e-07,
"loss": 0.0251,
"step": 10200
},
{
"epoch": 4.848053181386515,
"grad_norm": 0.07368100694821154,
"learning_rate": 2.828122517966658e-07,
"loss": 0.0251,
"step": 10210
},
{
"epoch": 4.852801519468186,
"grad_norm": 0.06042038526796139,
"learning_rate": 2.6548138369978935e-07,
"loss": 0.0248,
"step": 10220
},
{
"epoch": 4.8575498575498575,
"grad_norm": 0.06937803164487256,
"learning_rate": 2.4869704748934907e-07,
"loss": 0.0251,
"step": 10230
},
{
"epoch": 4.862298195631529,
"grad_norm": 0.07026043753015684,
"learning_rate": 2.3245942760816152e-07,
"loss": 0.0252,
"step": 10240
},
{
"epoch": 4.8670465337132,
"grad_norm": 0.06979416991043594,
"learning_rate": 2.167687024911935e-07,
"loss": 0.0257,
"step": 10250
},
{
"epoch": 4.871794871794872,
"grad_norm": 0.06881357873016393,
"learning_rate": 2.0162504456359143e-07,
"loss": 0.0247,
"step": 10260
},
{
"epoch": 4.8765432098765435,
"grad_norm": 0.07137191891679427,
"learning_rate": 1.8702862023878277e-07,
"loss": 0.0253,
"step": 10270
},
{
"epoch": 4.881291547958215,
"grad_norm": 0.0668030589232134,
"learning_rate": 1.7297958991666085e-07,
"loss": 0.0243,
"step": 10280
},
{
"epoch": 4.886039886039886,
"grad_norm": 0.07250978144589652,
"learning_rate": 1.5947810798181416e-07,
"loss": 0.0252,
"step": 10290
},
{
"epoch": 4.8907882241215574,
"grad_norm": 0.06578547148131111,
"learning_rate": 1.4652432280182203e-07,
"loss": 0.0251,
"step": 10300
},
{
"epoch": 4.895536562203229,
"grad_norm": 0.06768094358930553,
"learning_rate": 1.3411837672564487e-07,
"loss": 0.0262,
"step": 10310
},
{
"epoch": 4.9002849002849,
"grad_norm": 0.07893341104302869,
"learning_rate": 1.2226040608202537e-07,
"loss": 0.0252,
"step": 10320
},
{
"epoch": 4.905033238366571,
"grad_norm": 0.07340129111869041,
"learning_rate": 1.1095054117804537e-07,
"loss": 0.0252,
"step": 10330
},
{
"epoch": 4.909781576448243,
"grad_norm": 0.06957248648020418,
"learning_rate": 1.0018890629762689e-07,
"loss": 0.025,
"step": 10340
},
{
"epoch": 4.914529914529915,
"grad_norm": 0.06381627515141826,
"learning_rate": 8.997561970022217e-08,
"loss": 0.0254,
"step": 10350
},
{
"epoch": 4.919278252611586,
"grad_norm": 0.06476968126956692,
"learning_rate": 8.03107936194869e-08,
"loss": 0.0256,
"step": 10360
},
{
"epoch": 4.924026590693257,
"grad_norm": 0.06502816273954719,
"learning_rate": 7.119453426205347e-08,
"loss": 0.0245,
"step": 10370
},
{
"epoch": 4.928774928774929,
"grad_norm": 0.0727298488842256,
"learning_rate": 6.262694180634853e-08,
"loss": 0.0252,
"step": 10380
},
{
"epoch": 4.9335232668566,
"grad_norm": 0.06787347670413804,
"learning_rate": 5.4608110401532796e-08,
"loss": 0.0249,
"step": 10390
},
{
"epoch": 4.938271604938271,
"grad_norm": 0.07123053523960597,
"learning_rate": 4.713812816641849e-08,
"loss": 0.0256,
"step": 10400
},
{
"epoch": 4.943019943019943,
"grad_norm": 0.06945052824184504,
"learning_rate": 4.0217077188536845e-08,
"loss": 0.0248,
"step": 10410
},
{
"epoch": 4.947768281101615,
"grad_norm": 0.0636373503665069,
"learning_rate": 3.38450335231999e-08,
"loss": 0.0257,
"step": 10420
},
{
"epoch": 4.952516619183286,
"grad_norm": 0.0628183241460227,
"learning_rate": 2.8022067192712277e-08,
"loss": 0.0239,
"step": 10430
},
{
"epoch": 4.957264957264957,
"grad_norm": 0.06392354768650385,
"learning_rate": 2.274824218556071e-08,
"loss": 0.0245,
"step": 10440
},
{
"epoch": 4.962013295346629,
"grad_norm": 0.06981839897945984,
"learning_rate": 1.802361645573125e-08,
"loss": 0.0253,
"step": 10450
},
{
"epoch": 4.9667616334283,
"grad_norm": 0.061768621790248696,
"learning_rate": 1.3848241922059802e-08,
"loss": 0.0246,
"step": 10460
},
{
"epoch": 4.971509971509971,
"grad_norm": 0.062142857097907275,
"learning_rate": 1.0222164467688089e-08,
"loss": 0.025,
"step": 10470
},
{
"epoch": 4.976258309591643,
"grad_norm": 0.06597776035895064,
"learning_rate": 7.1454239395141175e-09,
"loss": 0.0252,
"step": 10480
},
{
"epoch": 4.981006647673315,
"grad_norm": 0.06957781773263208,
"learning_rate": 4.618054147803585e-09,
"loss": 0.0248,
"step": 10490
},
{
"epoch": 4.985754985754986,
"grad_norm": 0.06922411981752029,
"learning_rate": 2.6400828657791033e-09,
"loss": 0.0241,
"step": 10500
},
{
"epoch": 4.990503323836657,
"grad_norm": 0.06652673970121814,
"learning_rate": 1.211531829337087e-09,
"loss": 0.0255,
"step": 10510
},
{
"epoch": 4.995251661918329,
"grad_norm": 0.06648387513848081,
"learning_rate": 3.324167368090603e-10,
"loss": 0.0249,
"step": 10520
},
{
"epoch": 5.0,
"grad_norm": 0.06578801160989119,
"learning_rate": 2.747248778467437e-12,
"loss": 0.0252,
"step": 10530
},
{
"epoch": 5.0,
"step": 10530,
"total_flos": 952962456748032.0,
"train_loss": 0.192076750309096,
"train_runtime": 28895.8962,
"train_samples_per_second": 186.566,
"train_steps_per_second": 0.364
}
],
"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": true
},
"attributes": {}
}
},
"total_flos": 952962456748032.0,
"train_batch_size": 128,
"trial_name": null,
"trial_params": null
}