yushg's picture
Upload folder using huggingface_hub
7365c8f verified
Raw
History Blame Contribute Delete
47.6 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.10545603142589736,
"eval_steps": 500,
"global_step": 2500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00042182412570358947,
"grad_norm": 0.8065983057022095,
"learning_rate": 2.52808988764045e-07,
"loss": 35.651678466796874,
"step": 10
},
{
"epoch": 0.0008436482514071789,
"grad_norm": 0.9798905849456787,
"learning_rate": 5.337078651685394e-07,
"loss": 35.52773742675781,
"step": 20
},
{
"epoch": 0.0012654723771107683,
"grad_norm": 0.9430276155471802,
"learning_rate": 8.146067415730338e-07,
"loss": 35.94612426757813,
"step": 30
},
{
"epoch": 0.0016872965028143579,
"grad_norm": 0.8668350577354431,
"learning_rate": 1.0955056179775282e-06,
"loss": 35.54988098144531,
"step": 40
},
{
"epoch": 0.0021091206285179472,
"grad_norm": 0.8153179883956909,
"learning_rate": 1.3764044943820225e-06,
"loss": 35.851254272460935,
"step": 50
},
{
"epoch": 0.0025309447542215366,
"grad_norm": 0.9380270838737488,
"learning_rate": 1.6573033707865168e-06,
"loss": 35.082681274414064,
"step": 60
},
{
"epoch": 0.0029527688799251264,
"grad_norm": 0.8418566584587097,
"learning_rate": 1.9382022471910116e-06,
"loss": 35.33572998046875,
"step": 70
},
{
"epoch": 0.0033745930056287158,
"grad_norm": 0.8267859816551208,
"learning_rate": 2.2191011235955057e-06,
"loss": 34.830789184570314,
"step": 80
},
{
"epoch": 0.003796417131332305,
"grad_norm": 0.8174882531166077,
"learning_rate": 2.5e-06,
"loss": 33.9393798828125,
"step": 90
},
{
"epoch": 0.0042182412570358945,
"grad_norm": 0.8325590491294861,
"learning_rate": 2.7808988764044947e-06,
"loss": 33.55337829589844,
"step": 100
},
{
"epoch": 0.004640065382739484,
"grad_norm": 0.7891654372215271,
"learning_rate": 3.0617977528089893e-06,
"loss": 33.115335083007814,
"step": 110
},
{
"epoch": 0.005061889508443073,
"grad_norm": 1.0313451290130615,
"learning_rate": 3.3426966292134834e-06,
"loss": 32.383547973632815,
"step": 120
},
{
"epoch": 0.005483713634146663,
"grad_norm": 0.5737534165382385,
"learning_rate": 3.623595505617978e-06,
"loss": 31.1286865234375,
"step": 130
},
{
"epoch": 0.005905537759850253,
"grad_norm": 0.7010501027107239,
"learning_rate": 3.9044943820224724e-06,
"loss": 29.55970458984375,
"step": 140
},
{
"epoch": 0.006327361885553842,
"grad_norm": 0.5494692325592041,
"learning_rate": 4.185393258426967e-06,
"loss": 29.449929809570314,
"step": 150
},
{
"epoch": 0.0067491860112574315,
"grad_norm": 0.41349244117736816,
"learning_rate": 4.466292134831461e-06,
"loss": 28.477688598632813,
"step": 160
},
{
"epoch": 0.007171010136961021,
"grad_norm": 0.30324503779411316,
"learning_rate": 4.747191011235955e-06,
"loss": 27.354144287109374,
"step": 170
},
{
"epoch": 0.00759283426266461,
"grad_norm": 0.327568918466568,
"learning_rate": 5.02808988764045e-06,
"loss": 26.92130126953125,
"step": 180
},
{
"epoch": 0.0080146583883682,
"grad_norm": 0.29565760493278503,
"learning_rate": 5.308988764044944e-06,
"loss": 26.194525146484374,
"step": 190
},
{
"epoch": 0.008436482514071789,
"grad_norm": 0.2572338283061981,
"learning_rate": 5.589887640449439e-06,
"loss": 25.769549560546874,
"step": 200
},
{
"epoch": 0.008858306639775378,
"grad_norm": 0.2657480239868164,
"learning_rate": 5.870786516853933e-06,
"loss": 25.394000244140624,
"step": 210
},
{
"epoch": 0.009280130765478968,
"grad_norm": 0.20723949372768402,
"learning_rate": 6.151685393258428e-06,
"loss": 24.25175323486328,
"step": 220
},
{
"epoch": 0.009701954891182557,
"grad_norm": 0.19784937798976898,
"learning_rate": 6.4325842696629215e-06,
"loss": 23.95145263671875,
"step": 230
},
{
"epoch": 0.010123779016886146,
"grad_norm": 0.1812913715839386,
"learning_rate": 6.713483146067417e-06,
"loss": 23.77376708984375,
"step": 240
},
{
"epoch": 0.010545603142589736,
"grad_norm": 0.19627007842063904,
"learning_rate": 6.9943820224719105e-06,
"loss": 23.246966552734374,
"step": 250
},
{
"epoch": 0.010967427268293325,
"grad_norm": 0.22846540808677673,
"learning_rate": 7.275280898876404e-06,
"loss": 23.158154296875,
"step": 260
},
{
"epoch": 0.011389251393996916,
"grad_norm": 0.1520307958126068,
"learning_rate": 7.5561797752809e-06,
"loss": 22.498358154296874,
"step": 270
},
{
"epoch": 0.011811075519700506,
"grad_norm": 0.14488539099693298,
"learning_rate": 7.837078651685394e-06,
"loss": 22.60594482421875,
"step": 280
},
{
"epoch": 0.012232899645404095,
"grad_norm": 0.14345483481884003,
"learning_rate": 8.117977528089889e-06,
"loss": 22.301371765136718,
"step": 290
},
{
"epoch": 0.012654723771107684,
"grad_norm": 0.6615362763404846,
"learning_rate": 8.398876404494383e-06,
"loss": 21.755278015136717,
"step": 300
},
{
"epoch": 0.013076547896811274,
"grad_norm": 0.1423025280237198,
"learning_rate": 8.679775280898876e-06,
"loss": 21.386692810058594,
"step": 310
},
{
"epoch": 0.013498372022514863,
"grad_norm": 0.14323103427886963,
"learning_rate": 8.960674157303372e-06,
"loss": 21.5313232421875,
"step": 320
},
{
"epoch": 0.013920196148218452,
"grad_norm": 0.13975073397159576,
"learning_rate": 9.241573033707865e-06,
"loss": 21.04039306640625,
"step": 330
},
{
"epoch": 0.014342020273922042,
"grad_norm": 0.18143784999847412,
"learning_rate": 9.522471910112361e-06,
"loss": 21.435638427734375,
"step": 340
},
{
"epoch": 0.014763844399625631,
"grad_norm": 0.16371376812458038,
"learning_rate": 9.803370786516854e-06,
"loss": 21.240106201171876,
"step": 350
},
{
"epoch": 0.01518566852532922,
"grad_norm": 0.19794277846813202,
"learning_rate": 1.0084269662921349e-05,
"loss": 21.265296936035156,
"step": 360
},
{
"epoch": 0.01560749265103281,
"grad_norm": 0.15830059349536896,
"learning_rate": 1.0365168539325843e-05,
"loss": 20.513796997070312,
"step": 370
},
{
"epoch": 0.0160293167767364,
"grad_norm": 0.15261049568653107,
"learning_rate": 1.064606741573034e-05,
"loss": 20.430271911621094,
"step": 380
},
{
"epoch": 0.01645114090243999,
"grad_norm": 0.17001120746135712,
"learning_rate": 1.0926966292134832e-05,
"loss": 20.2728759765625,
"step": 390
},
{
"epoch": 0.016872965028143578,
"grad_norm": 0.1754172444343567,
"learning_rate": 1.1207865168539327e-05,
"loss": 20.585818481445312,
"step": 400
},
{
"epoch": 0.01729478915384717,
"grad_norm": 0.1494324952363968,
"learning_rate": 1.1488764044943821e-05,
"loss": 20.26520690917969,
"step": 410
},
{
"epoch": 0.017716613279550757,
"grad_norm": 0.2788557708263397,
"learning_rate": 1.1769662921348314e-05,
"loss": 20.216525268554687,
"step": 420
},
{
"epoch": 0.018138437405254348,
"grad_norm": 0.22943347692489624,
"learning_rate": 1.2050561797752809e-05,
"loss": 19.879740905761718,
"step": 430
},
{
"epoch": 0.018560261530957935,
"grad_norm": 0.19310815632343292,
"learning_rate": 1.2331460674157305e-05,
"loss": 20.340792846679687,
"step": 440
},
{
"epoch": 0.018982085656661526,
"grad_norm": 0.17677143216133118,
"learning_rate": 1.26123595505618e-05,
"loss": 20.295317077636717,
"step": 450
},
{
"epoch": 0.019403909782365114,
"grad_norm": 0.1568199247121811,
"learning_rate": 1.2893258426966292e-05,
"loss": 20.30619659423828,
"step": 460
},
{
"epoch": 0.019825733908068705,
"grad_norm": 0.14602652192115784,
"learning_rate": 1.3174157303370787e-05,
"loss": 19.96039581298828,
"step": 470
},
{
"epoch": 0.020247558033772293,
"grad_norm": 0.19724591076374054,
"learning_rate": 1.3455056179775281e-05,
"loss": 19.820477294921876,
"step": 480
},
{
"epoch": 0.020669382159475884,
"grad_norm": 0.14715564250946045,
"learning_rate": 1.3735955056179778e-05,
"loss": 19.374057006835937,
"step": 490
},
{
"epoch": 0.02109120628517947,
"grad_norm": 0.16100069880485535,
"learning_rate": 1.401685393258427e-05,
"loss": 19.71839599609375,
"step": 500
},
{
"epoch": 0.021513030410883063,
"grad_norm": 0.2880576550960541,
"learning_rate": 1.4297752808988765e-05,
"loss": 20.490005493164062,
"step": 510
},
{
"epoch": 0.02193485453658665,
"grad_norm": 0.2523888349533081,
"learning_rate": 1.457865168539326e-05,
"loss": 20.048866271972656,
"step": 520
},
{
"epoch": 0.02235667866229024,
"grad_norm": 0.2273065447807312,
"learning_rate": 1.4859550561797754e-05,
"loss": 19.724508666992186,
"step": 530
},
{
"epoch": 0.022778502787993832,
"grad_norm": 0.2378254532814026,
"learning_rate": 1.5140449438202248e-05,
"loss": 19.543185424804687,
"step": 540
},
{
"epoch": 0.02320032691369742,
"grad_norm": 0.209368035197258,
"learning_rate": 1.5421348314606745e-05,
"loss": 19.98386993408203,
"step": 550
},
{
"epoch": 0.02362215103940101,
"grad_norm": 0.23571042716503143,
"learning_rate": 1.5702247191011238e-05,
"loss": 19.499612426757814,
"step": 560
},
{
"epoch": 0.0240439751651046,
"grad_norm": 0.22823461890220642,
"learning_rate": 1.598314606741573e-05,
"loss": 19.482470703125,
"step": 570
},
{
"epoch": 0.02446579929080819,
"grad_norm": 0.25839850306510925,
"learning_rate": 1.6264044943820227e-05,
"loss": 19.360768127441407,
"step": 580
},
{
"epoch": 0.024887623416511778,
"grad_norm": 0.25028201937675476,
"learning_rate": 1.6544943820224723e-05,
"loss": 19.951307678222655,
"step": 590
},
{
"epoch": 0.02530944754221537,
"grad_norm": 0.22897405922412872,
"learning_rate": 1.6825842696629216e-05,
"loss": 19.387236022949217,
"step": 600
},
{
"epoch": 0.025731271667918956,
"grad_norm": 0.2035793960094452,
"learning_rate": 1.710674157303371e-05,
"loss": 19.838690185546874,
"step": 610
},
{
"epoch": 0.026153095793622547,
"grad_norm": 0.22483032941818237,
"learning_rate": 1.7387640449438205e-05,
"loss": 19.445848083496095,
"step": 620
},
{
"epoch": 0.026574919919326135,
"grad_norm": 0.2724807560443878,
"learning_rate": 1.7668539325842698e-05,
"loss": 19.43202362060547,
"step": 630
},
{
"epoch": 0.026996744045029726,
"grad_norm": 0.3173777759075165,
"learning_rate": 1.7949438202247194e-05,
"loss": 19.13489685058594,
"step": 640
},
{
"epoch": 0.027418568170733314,
"grad_norm": 0.22422824800014496,
"learning_rate": 1.8230337078651687e-05,
"loss": 19.380812072753905,
"step": 650
},
{
"epoch": 0.027840392296436905,
"grad_norm": 0.24893997609615326,
"learning_rate": 1.8511235955056183e-05,
"loss": 18.949267578125,
"step": 660
},
{
"epoch": 0.028262216422140492,
"grad_norm": 0.3204539716243744,
"learning_rate": 1.8792134831460676e-05,
"loss": 18.789942932128906,
"step": 670
},
{
"epoch": 0.028684040547844084,
"grad_norm": 0.2648240327835083,
"learning_rate": 1.907303370786517e-05,
"loss": 18.86900939941406,
"step": 680
},
{
"epoch": 0.02910586467354767,
"grad_norm": 0.25324350595474243,
"learning_rate": 1.9353932584269665e-05,
"loss": 19.13300323486328,
"step": 690
},
{
"epoch": 0.029527688799251262,
"grad_norm": 0.2934200167655945,
"learning_rate": 1.963483146067416e-05,
"loss": 19.383299255371092,
"step": 700
},
{
"epoch": 0.029949512924954853,
"grad_norm": 0.24456587433815002,
"learning_rate": 1.9915730337078654e-05,
"loss": 19.293736267089844,
"step": 710
},
{
"epoch": 0.03037133705065844,
"grad_norm": 0.23921170830726624,
"learning_rate": 1.999999542702351e-05,
"loss": 18.000759887695313,
"step": 720
},
{
"epoch": 0.030793161176362032,
"grad_norm": 0.23188889026641846,
"learning_rate": 1.9999973028781374e-05,
"loss": 18.787120056152343,
"step": 730
},
{
"epoch": 0.03121498530206562,
"grad_norm": 0.33560246229171753,
"learning_rate": 1.9999931965380888e-05,
"loss": 19.33695068359375,
"step": 740
},
{
"epoch": 0.03163680942776921,
"grad_norm": 0.28916871547698975,
"learning_rate": 1.9999872236898698e-05,
"loss": 19.302365112304688,
"step": 750
},
{
"epoch": 0.0320586335534728,
"grad_norm": 0.33663684129714966,
"learning_rate": 1.9999793843446293e-05,
"loss": 19.0060546875,
"step": 760
},
{
"epoch": 0.032480457679176386,
"grad_norm": 0.3129170536994934,
"learning_rate": 1.9999696785169994e-05,
"loss": 18.472871398925783,
"step": 770
},
{
"epoch": 0.03290228180487998,
"grad_norm": 0.26207199692726135,
"learning_rate": 1.999958106225096e-05,
"loss": 18.929963684082033,
"step": 780
},
{
"epoch": 0.03332410593058357,
"grad_norm": 0.2932567596435547,
"learning_rate": 1.999944667490519e-05,
"loss": 18.533164978027344,
"step": 790
},
{
"epoch": 0.033745930056287156,
"grad_norm": 0.2569630444049835,
"learning_rate": 1.999929362338352e-05,
"loss": 18.535040283203124,
"step": 800
},
{
"epoch": 0.034167754181990743,
"grad_norm": 0.2609207034111023,
"learning_rate": 1.9999121907971632e-05,
"loss": 18.693728637695312,
"step": 810
},
{
"epoch": 0.03458957830769434,
"grad_norm": 0.5386088490486145,
"learning_rate": 1.9998931528990027e-05,
"loss": 18.692576599121093,
"step": 820
},
{
"epoch": 0.035011402433397926,
"grad_norm": 0.26829224824905396,
"learning_rate": 1.9998722486794055e-05,
"loss": 18.050027465820314,
"step": 830
},
{
"epoch": 0.03543322655910151,
"grad_norm": 0.3113354444503784,
"learning_rate": 1.99984947817739e-05,
"loss": 18.855671691894532,
"step": 840
},
{
"epoch": 0.0358550506848051,
"grad_norm": 0.29941871762275696,
"learning_rate": 1.9998248414354575e-05,
"loss": 18.62144775390625,
"step": 850
},
{
"epoch": 0.036276874810508696,
"grad_norm": 0.2699904441833496,
"learning_rate": 1.9997983384995927e-05,
"loss": 18.160052490234374,
"step": 860
},
{
"epoch": 0.03669869893621228,
"grad_norm": 0.2435958981513977,
"learning_rate": 1.9997699694192646e-05,
"loss": 18.752867126464842,
"step": 870
},
{
"epoch": 0.03712052306191587,
"grad_norm": 0.2814241647720337,
"learning_rate": 1.9997397342474243e-05,
"loss": 18.577685546875,
"step": 880
},
{
"epoch": 0.037542347187619465,
"grad_norm": 0.29350870847702026,
"learning_rate": 1.9997076330405062e-05,
"loss": 18.84843292236328,
"step": 890
},
{
"epoch": 0.03796417131332305,
"grad_norm": 0.2537597715854645,
"learning_rate": 1.999673665858428e-05,
"loss": 18.444070434570314,
"step": 900
},
{
"epoch": 0.03838599543902664,
"grad_norm": 0.26214396953582764,
"learning_rate": 1.9996378327645907e-05,
"loss": 18.22960205078125,
"step": 910
},
{
"epoch": 0.03880781956473023,
"grad_norm": 0.2742043435573578,
"learning_rate": 1.9996001338258768e-05,
"loss": 18.233761596679688,
"step": 920
},
{
"epoch": 0.03922964369043382,
"grad_norm": 0.250119149684906,
"learning_rate": 1.9995605691126524e-05,
"loss": 18.234817504882812,
"step": 930
},
{
"epoch": 0.03965146781613741,
"grad_norm": 0.2798559069633484,
"learning_rate": 1.999519138698766e-05,
"loss": 18.693147277832033,
"step": 940
},
{
"epoch": 0.040073291941841,
"grad_norm": 0.3584083318710327,
"learning_rate": 1.9994758426615478e-05,
"loss": 18.757966613769533,
"step": 950
},
{
"epoch": 0.040495116067544586,
"grad_norm": 0.28565338253974915,
"learning_rate": 1.9994306810818114e-05,
"loss": 18.272166442871093,
"step": 960
},
{
"epoch": 0.04091694019324818,
"grad_norm": 0.26797354221343994,
"learning_rate": 1.9993836540438516e-05,
"loss": 18.52594909667969,
"step": 970
},
{
"epoch": 0.04133876431895177,
"grad_norm": 0.3331480920314789,
"learning_rate": 1.9993347616354453e-05,
"loss": 18.326254272460936,
"step": 980
},
{
"epoch": 0.041760588444655355,
"grad_norm": 0.28409120440483093,
"learning_rate": 1.9992840039478514e-05,
"loss": 17.864889526367186,
"step": 990
},
{
"epoch": 0.04218241257035894,
"grad_norm": 0.2808006703853607,
"learning_rate": 1.9992313810758098e-05,
"loss": 18.465538024902344,
"step": 1000
},
{
"epoch": 0.04260423669606254,
"grad_norm": 0.33005738258361816,
"learning_rate": 1.9991768931175425e-05,
"loss": 18.249317932128907,
"step": 1010
},
{
"epoch": 0.043026060821766125,
"grad_norm": 0.27972036600112915,
"learning_rate": 1.9991205401747523e-05,
"loss": 18.170388793945314,
"step": 1020
},
{
"epoch": 0.04344788494746971,
"grad_norm": 0.4121862053871155,
"learning_rate": 1.999062322352623e-05,
"loss": 17.821492004394532,
"step": 1030
},
{
"epoch": 0.0438697090731733,
"grad_norm": 0.32552066445350647,
"learning_rate": 1.9990022397598197e-05,
"loss": 18.105780029296874,
"step": 1040
},
{
"epoch": 0.044291533198876895,
"grad_norm": 0.31183332204818726,
"learning_rate": 1.9989402925084876e-05,
"loss": 18.235502624511717,
"step": 1050
},
{
"epoch": 0.04471335732458048,
"grad_norm": 0.3100496828556061,
"learning_rate": 1.9988764807142526e-05,
"loss": 18.1237548828125,
"step": 1060
},
{
"epoch": 0.04513518145028407,
"grad_norm": 0.3076971769332886,
"learning_rate": 1.9988108044962207e-05,
"loss": 18.27461395263672,
"step": 1070
},
{
"epoch": 0.045557005575987665,
"grad_norm": 0.3156045377254486,
"learning_rate": 1.998743263976978e-05,
"loss": 17.71241455078125,
"step": 1080
},
{
"epoch": 0.04597882970169125,
"grad_norm": 0.29673850536346436,
"learning_rate": 1.9986738592825905e-05,
"loss": 17.840818786621092,
"step": 1090
},
{
"epoch": 0.04640065382739484,
"grad_norm": 0.28813785314559937,
"learning_rate": 1.998602590542603e-05,
"loss": 18.591908264160157,
"step": 1100
},
{
"epoch": 0.04682247795309843,
"grad_norm": 0.28837504982948303,
"learning_rate": 1.9985294578900406e-05,
"loss": 18.081942749023437,
"step": 1110
},
{
"epoch": 0.04724430207880202,
"grad_norm": 0.3204016089439392,
"learning_rate": 1.9984544614614064e-05,
"loss": 17.493878173828126,
"step": 1120
},
{
"epoch": 0.04766612620450561,
"grad_norm": 0.269089013338089,
"learning_rate": 1.9983776013966832e-05,
"loss": 17.79596405029297,
"step": 1130
},
{
"epoch": 0.0480879503302092,
"grad_norm": 0.28135010600090027,
"learning_rate": 1.9982988778393316e-05,
"loss": 17.925094604492188,
"step": 1140
},
{
"epoch": 0.048509774455912785,
"grad_norm": 0.30634036660194397,
"learning_rate": 1.9982182909362914e-05,
"loss": 17.815794372558592,
"step": 1150
},
{
"epoch": 0.04893159858161638,
"grad_norm": 0.3512091636657715,
"learning_rate": 1.9981358408379788e-05,
"loss": 18.366172790527344,
"step": 1160
},
{
"epoch": 0.04935342270731997,
"grad_norm": 0.32550451159477234,
"learning_rate": 1.9980515276982895e-05,
"loss": 17.663365173339844,
"step": 1170
},
{
"epoch": 0.049775246833023555,
"grad_norm": 0.3501587510108948,
"learning_rate": 1.997965351674595e-05,
"loss": 17.88150634765625,
"step": 1180
},
{
"epoch": 0.05019707095872714,
"grad_norm": 0.35774561762809753,
"learning_rate": 1.9978773129277457e-05,
"loss": 17.622511291503905,
"step": 1190
},
{
"epoch": 0.05061889508443074,
"grad_norm": 0.3506692945957184,
"learning_rate": 1.9977874116220665e-05,
"loss": 17.542771911621095,
"step": 1200
},
{
"epoch": 0.051040719210134325,
"grad_norm": 0.5348653793334961,
"learning_rate": 1.9976956479253605e-05,
"loss": 17.32509002685547,
"step": 1210
},
{
"epoch": 0.05146254333583791,
"grad_norm": 0.3314587473869324,
"learning_rate": 1.9976020220089068e-05,
"loss": 18.14122772216797,
"step": 1220
},
{
"epoch": 0.0518843674615415,
"grad_norm": 0.3455180525779724,
"learning_rate": 1.9975065340474602e-05,
"loss": 17.583253479003908,
"step": 1230
},
{
"epoch": 0.052306191587245095,
"grad_norm": 0.3618074357509613,
"learning_rate": 1.9974091842192507e-05,
"loss": 17.98968505859375,
"step": 1240
},
{
"epoch": 0.05272801571294868,
"grad_norm": 0.4234869182109833,
"learning_rate": 1.997309972705984e-05,
"loss": 18.163101196289062,
"step": 1250
},
{
"epoch": 0.05314983983865227,
"grad_norm": 0.5593043565750122,
"learning_rate": 1.99720889969284e-05,
"loss": 17.82501220703125,
"step": 1260
},
{
"epoch": 0.053571663964355865,
"grad_norm": 0.2861349582672119,
"learning_rate": 1.997105965368474e-05,
"loss": 17.73271484375,
"step": 1270
},
{
"epoch": 0.05399348809005945,
"grad_norm": 0.3276013433933258,
"learning_rate": 1.997001169925015e-05,
"loss": 17.142120361328125,
"step": 1280
},
{
"epoch": 0.05441531221576304,
"grad_norm": 0.3348940908908844,
"learning_rate": 1.9968945135580662e-05,
"loss": 17.70765075683594,
"step": 1290
},
{
"epoch": 0.05483713634146663,
"grad_norm": 0.3119870722293854,
"learning_rate": 1.9967859964667033e-05,
"loss": 17.554115295410156,
"step": 1300
},
{
"epoch": 0.05525896046717022,
"grad_norm": 0.295555055141449,
"learning_rate": 1.9966756188534758e-05,
"loss": 17.459405517578126,
"step": 1310
},
{
"epoch": 0.05568078459287381,
"grad_norm": 0.30782192945480347,
"learning_rate": 1.9965633809244064e-05,
"loss": 17.691316223144533,
"step": 1320
},
{
"epoch": 0.0561026087185774,
"grad_norm": 0.34530341625213623,
"learning_rate": 1.996449282888989e-05,
"loss": 17.750396728515625,
"step": 1330
},
{
"epoch": 0.056524432844280985,
"grad_norm": 0.5588768720626831,
"learning_rate": 1.9963333249601903e-05,
"loss": 17.97089538574219,
"step": 1340
},
{
"epoch": 0.05694625696998458,
"grad_norm": 0.3901899456977844,
"learning_rate": 1.996215507354448e-05,
"loss": 17.23460998535156,
"step": 1350
},
{
"epoch": 0.05736808109568817,
"grad_norm": 0.4388797879219055,
"learning_rate": 1.996095830291671e-05,
"loss": 18.03730010986328,
"step": 1360
},
{
"epoch": 0.057789905221391755,
"grad_norm": 0.31590718030929565,
"learning_rate": 1.9959742939952393e-05,
"loss": 18.033831787109374,
"step": 1370
},
{
"epoch": 0.05821172934709534,
"grad_norm": 0.3215778172016144,
"learning_rate": 1.9958508986920026e-05,
"loss": 17.679611206054688,
"step": 1380
},
{
"epoch": 0.05863355347279894,
"grad_norm": 0.31076571345329285,
"learning_rate": 1.9957256446122813e-05,
"loss": 18.017640686035158,
"step": 1390
},
{
"epoch": 0.059055377598502524,
"grad_norm": 0.3545955717563629,
"learning_rate": 1.995598531989864e-05,
"loss": 17.61700439453125,
"step": 1400
},
{
"epoch": 0.05947720172420611,
"grad_norm": 0.33153286576271057,
"learning_rate": 1.99546956106201e-05,
"loss": 18.042115783691408,
"step": 1410
},
{
"epoch": 0.05989902584990971,
"grad_norm": 0.34877341985702515,
"learning_rate": 1.995338732069446e-05,
"loss": 17.477146911621094,
"step": 1420
},
{
"epoch": 0.060320849975613294,
"grad_norm": 0.3148050308227539,
"learning_rate": 1.995206045256366e-05,
"loss": 17.641142272949217,
"step": 1430
},
{
"epoch": 0.06074267410131688,
"grad_norm": 0.3306654095649719,
"learning_rate": 1.995071500870434e-05,
"loss": 17.359457397460936,
"step": 1440
},
{
"epoch": 0.06116449822702047,
"grad_norm": 0.34123876690864563,
"learning_rate": 1.9949350991627796e-05,
"loss": 18.056146240234376,
"step": 1450
},
{
"epoch": 0.061586322352724064,
"grad_norm": 0.37769562005996704,
"learning_rate": 1.994796840387999e-05,
"loss": 17.03343811035156,
"step": 1460
},
{
"epoch": 0.06200814647842765,
"grad_norm": 0.3330378234386444,
"learning_rate": 1.9946567248041557e-05,
"loss": 17.69776153564453,
"step": 1470
},
{
"epoch": 0.06242997060413124,
"grad_norm": 0.37690991163253784,
"learning_rate": 1.9945147526727775e-05,
"loss": 17.852751159667967,
"step": 1480
},
{
"epoch": 0.06285179472983483,
"grad_norm": 0.34739160537719727,
"learning_rate": 1.9943709242588588e-05,
"loss": 17.656039428710937,
"step": 1490
},
{
"epoch": 0.06327361885553842,
"grad_norm": 0.3611721992492676,
"learning_rate": 1.994225239830859e-05,
"loss": 17.2119140625,
"step": 1500
},
{
"epoch": 0.06369544298124201,
"grad_norm": 0.3363858163356781,
"learning_rate": 1.9940776996607e-05,
"loss": 17.814456176757812,
"step": 1510
},
{
"epoch": 0.0641172671069456,
"grad_norm": 0.5571886301040649,
"learning_rate": 1.9939283040237695e-05,
"loss": 17.853152465820312,
"step": 1520
},
{
"epoch": 0.06453909123264918,
"grad_norm": 0.36795610189437866,
"learning_rate": 1.993777053198917e-05,
"loss": 17.811891174316408,
"step": 1530
},
{
"epoch": 0.06496091535835277,
"grad_norm": 0.3559716045856476,
"learning_rate": 1.9936239474684552e-05,
"loss": 17.460633850097658,
"step": 1540
},
{
"epoch": 0.06538273948405636,
"grad_norm": 0.3193165063858032,
"learning_rate": 1.99346898711816e-05,
"loss": 17.222584533691407,
"step": 1550
},
{
"epoch": 0.06580456360975996,
"grad_norm": 0.3640180826187134,
"learning_rate": 1.993312172437267e-05,
"loss": 16.96032257080078,
"step": 1560
},
{
"epoch": 0.06622638773546355,
"grad_norm": 0.3581082224845886,
"learning_rate": 1.993153503718475e-05,
"loss": 17.633184814453124,
"step": 1570
},
{
"epoch": 0.06664821186116714,
"grad_norm": 0.3320285677909851,
"learning_rate": 1.992992981257942e-05,
"loss": 16.92774353027344,
"step": 1580
},
{
"epoch": 0.06707003598687072,
"grad_norm": 0.33919626474380493,
"learning_rate": 1.9928306053552872e-05,
"loss": 17.283872985839842,
"step": 1590
},
{
"epoch": 0.06749186011257431,
"grad_norm": 0.41237595677375793,
"learning_rate": 1.9926663763135874e-05,
"loss": 17.487303161621092,
"step": 1600
},
{
"epoch": 0.0679136842382779,
"grad_norm": 0.3715861141681671,
"learning_rate": 1.9925002944393806e-05,
"loss": 17.45185089111328,
"step": 1610
},
{
"epoch": 0.06833550836398149,
"grad_norm": 0.5211480855941772,
"learning_rate": 1.9923323600426615e-05,
"loss": 17.200485229492188,
"step": 1620
},
{
"epoch": 0.06875733248968509,
"grad_norm": 0.3421076536178589,
"learning_rate": 1.9921625734368835e-05,
"loss": 17.648558044433592,
"step": 1630
},
{
"epoch": 0.06917915661538868,
"grad_norm": 0.3373667299747467,
"learning_rate": 1.9919909349389564e-05,
"loss": 17.757981872558595,
"step": 1640
},
{
"epoch": 0.06960098074109226,
"grad_norm": 0.44610220193862915,
"learning_rate": 1.9918174448692475e-05,
"loss": 17.162387084960937,
"step": 1650
},
{
"epoch": 0.07002280486679585,
"grad_norm": 0.31521204113960266,
"learning_rate": 1.9916421035515795e-05,
"loss": 16.34774169921875,
"step": 1660
},
{
"epoch": 0.07044462899249944,
"grad_norm": 0.48609769344329834,
"learning_rate": 1.9914649113132304e-05,
"loss": 17.159957885742188,
"step": 1670
},
{
"epoch": 0.07086645311820303,
"grad_norm": 0.5699203610420227,
"learning_rate": 1.9912858684849335e-05,
"loss": 17.107530212402345,
"step": 1680
},
{
"epoch": 0.07128827724390661,
"grad_norm": 0.5126078128814697,
"learning_rate": 1.991104975400876e-05,
"loss": 16.917098999023438,
"step": 1690
},
{
"epoch": 0.0717101013696102,
"grad_norm": 0.34865614771842957,
"learning_rate": 1.9909222323986984e-05,
"loss": 16.796833801269532,
"step": 1700
},
{
"epoch": 0.0721319254953138,
"grad_norm": 0.4295679032802582,
"learning_rate": 1.990737639819494e-05,
"loss": 17.01697235107422,
"step": 1710
},
{
"epoch": 0.07255374962101739,
"grad_norm": 0.541847825050354,
"learning_rate": 1.9905511980078097e-05,
"loss": 17.542742919921874,
"step": 1720
},
{
"epoch": 0.07297557374672098,
"grad_norm": 0.3423381447792053,
"learning_rate": 1.9903629073116422e-05,
"loss": 16.936869812011718,
"step": 1730
},
{
"epoch": 0.07339739787242457,
"grad_norm": 0.3373686671257019,
"learning_rate": 1.9901727680824405e-05,
"loss": 17.406088256835936,
"step": 1740
},
{
"epoch": 0.07381922199812815,
"grad_norm": 0.30602356791496277,
"learning_rate": 1.989980780675103e-05,
"loss": 17.589431762695312,
"step": 1750
},
{
"epoch": 0.07424104612383174,
"grad_norm": 0.36831897497177124,
"learning_rate": 1.9897869454479787e-05,
"loss": 17.372021484375,
"step": 1760
},
{
"epoch": 0.07466287024953533,
"grad_norm": 0.3464960753917694,
"learning_rate": 1.9895912627628646e-05,
"loss": 17.509791564941406,
"step": 1770
},
{
"epoch": 0.07508469437523893,
"grad_norm": 0.383433997631073,
"learning_rate": 1.9893937329850063e-05,
"loss": 17.448422241210938,
"step": 1780
},
{
"epoch": 0.07550651850094252,
"grad_norm": 0.3892641067504883,
"learning_rate": 1.989194356483098e-05,
"loss": 17.055230712890626,
"step": 1790
},
{
"epoch": 0.0759283426266461,
"grad_norm": 0.34991690516471863,
"learning_rate": 1.98899313362928e-05,
"loss": 17.407589721679688,
"step": 1800
},
{
"epoch": 0.0763501667523497,
"grad_norm": 0.3640735447406769,
"learning_rate": 1.9887900647991385e-05,
"loss": 17.507382202148438,
"step": 1810
},
{
"epoch": 0.07677199087805328,
"grad_norm": 0.416438490152359,
"learning_rate": 1.988585150371706e-05,
"loss": 17.261448669433594,
"step": 1820
},
{
"epoch": 0.07719381500375687,
"grad_norm": 0.4024409055709839,
"learning_rate": 1.9883783907294592e-05,
"loss": 17.188780212402342,
"step": 1830
},
{
"epoch": 0.07761563912946046,
"grad_norm": 0.3822477161884308,
"learning_rate": 1.98816978625832e-05,
"loss": 17.296994018554688,
"step": 1840
},
{
"epoch": 0.07803746325516404,
"grad_norm": 0.3287546932697296,
"learning_rate": 1.9879593373476524e-05,
"loss": 16.1343994140625,
"step": 1850
},
{
"epoch": 0.07845928738086765,
"grad_norm": 0.3958519399166107,
"learning_rate": 1.9877470443902638e-05,
"loss": 17.35173645019531,
"step": 1860
},
{
"epoch": 0.07888111150657123,
"grad_norm": 0.37541836500167847,
"learning_rate": 1.9875329077824037e-05,
"loss": 17.28033905029297,
"step": 1870
},
{
"epoch": 0.07930293563227482,
"grad_norm": 0.34392815828323364,
"learning_rate": 1.987316927923762e-05,
"loss": 17.26871337890625,
"step": 1880
},
{
"epoch": 0.07972475975797841,
"grad_norm": 0.37766674160957336,
"learning_rate": 1.9870991052174704e-05,
"loss": 17.64359893798828,
"step": 1890
},
{
"epoch": 0.080146583883682,
"grad_norm": 0.3722328543663025,
"learning_rate": 1.986879440070099e-05,
"loss": 17.241679382324218,
"step": 1900
},
{
"epoch": 0.08056840800938558,
"grad_norm": 0.41544944047927856,
"learning_rate": 1.986657932891657e-05,
"loss": 17.12104187011719,
"step": 1910
},
{
"epoch": 0.08099023213508917,
"grad_norm": 0.39178556203842163,
"learning_rate": 1.9864345840955932e-05,
"loss": 17.20569610595703,
"step": 1920
},
{
"epoch": 0.08141205626079276,
"grad_norm": 0.32549089193344116,
"learning_rate": 1.986209394098792e-05,
"loss": 16.99742889404297,
"step": 1930
},
{
"epoch": 0.08183388038649636,
"grad_norm": 0.3502710163593292,
"learning_rate": 1.9859823633215755e-05,
"loss": 17.434646606445312,
"step": 1940
},
{
"epoch": 0.08225570451219995,
"grad_norm": 0.37979257106781006,
"learning_rate": 1.9857534921877013e-05,
"loss": 16.822279357910155,
"step": 1950
},
{
"epoch": 0.08267752863790354,
"grad_norm": 0.36643335223197937,
"learning_rate": 1.9855227811243623e-05,
"loss": 16.503253173828124,
"step": 1960
},
{
"epoch": 0.08309935276360712,
"grad_norm": 0.41441091895103455,
"learning_rate": 1.985290230562185e-05,
"loss": 17.45409240722656,
"step": 1970
},
{
"epoch": 0.08352117688931071,
"grad_norm": 0.3626113831996918,
"learning_rate": 1.9850558409352305e-05,
"loss": 16.81719512939453,
"step": 1980
},
{
"epoch": 0.0839430010150143,
"grad_norm": 0.39800775051116943,
"learning_rate": 1.984819612680992e-05,
"loss": 17.30420684814453,
"step": 1990
},
{
"epoch": 0.08436482514071789,
"grad_norm": 0.508959174156189,
"learning_rate": 1.984581546240394e-05,
"loss": 17.380404663085937,
"step": 2000
},
{
"epoch": 0.08478664926642149,
"grad_norm": 0.40276896953582764,
"learning_rate": 1.9843416420577924e-05,
"loss": 16.7919677734375,
"step": 2010
},
{
"epoch": 0.08520847339212508,
"grad_norm": 0.362185001373291,
"learning_rate": 1.984099900580974e-05,
"loss": 16.68143310546875,
"step": 2020
},
{
"epoch": 0.08563029751782866,
"grad_norm": 0.5219398140907288,
"learning_rate": 1.9838563222611536e-05,
"loss": 16.81324157714844,
"step": 2030
},
{
"epoch": 0.08605212164353225,
"grad_norm": 0.6212902665138245,
"learning_rate": 1.9836109075529758e-05,
"loss": 17.86339874267578,
"step": 2040
},
{
"epoch": 0.08647394576923584,
"grad_norm": 0.4826348125934601,
"learning_rate": 1.9833636569145118e-05,
"loss": 17.522291564941405,
"step": 2050
},
{
"epoch": 0.08689576989493943,
"grad_norm": 0.4099210202693939,
"learning_rate": 1.98311457080726e-05,
"loss": 17.420944213867188,
"step": 2060
},
{
"epoch": 0.08731759402064301,
"grad_norm": 0.4783196747303009,
"learning_rate": 1.982863649696146e-05,
"loss": 16.711087036132813,
"step": 2070
},
{
"epoch": 0.0877394181463466,
"grad_norm": 0.4891475737094879,
"learning_rate": 1.9826108940495178e-05,
"loss": 16.979551696777342,
"step": 2080
},
{
"epoch": 0.0881612422720502,
"grad_norm": 0.4523952007293701,
"learning_rate": 1.98235630433915e-05,
"loss": 16.44328155517578,
"step": 2090
},
{
"epoch": 0.08858306639775379,
"grad_norm": 0.5939908027648926,
"learning_rate": 1.982099881040239e-05,
"loss": 17.6498046875,
"step": 2100
},
{
"epoch": 0.08900489052345738,
"grad_norm": 0.40050771832466125,
"learning_rate": 1.9818416246314047e-05,
"loss": 17.505574035644532,
"step": 2110
},
{
"epoch": 0.08942671464916097,
"grad_norm": 0.3944551646709442,
"learning_rate": 1.981581535594688e-05,
"loss": 17.52928009033203,
"step": 2120
},
{
"epoch": 0.08984853877486455,
"grad_norm": 0.5643343329429626,
"learning_rate": 1.98131961441555e-05,
"loss": 17.12488708496094,
"step": 2130
},
{
"epoch": 0.09027036290056814,
"grad_norm": 0.37489837408065796,
"learning_rate": 1.981055861582873e-05,
"loss": 16.972441101074217,
"step": 2140
},
{
"epoch": 0.09069218702627173,
"grad_norm": 0.37068799138069153,
"learning_rate": 1.9807902775889565e-05,
"loss": 17.224105834960938,
"step": 2150
},
{
"epoch": 0.09111401115197533,
"grad_norm": 0.4233551025390625,
"learning_rate": 1.9805228629295184e-05,
"loss": 16.891741943359374,
"step": 2160
},
{
"epoch": 0.09153583527767892,
"grad_norm": 0.6357532739639282,
"learning_rate": 1.9802536181036945e-05,
"loss": 17.310926818847655,
"step": 2170
},
{
"epoch": 0.0919576594033825,
"grad_norm": 0.4123924672603607,
"learning_rate": 1.9799825436140353e-05,
"loss": 16.99944610595703,
"step": 2180
},
{
"epoch": 0.09237948352908609,
"grad_norm": 0.612068235874176,
"learning_rate": 1.9797096399665072e-05,
"loss": 16.675027465820314,
"step": 2190
},
{
"epoch": 0.09280130765478968,
"grad_norm": 0.42969682812690735,
"learning_rate": 1.9794349076704903e-05,
"loss": 17.50474395751953,
"step": 2200
},
{
"epoch": 0.09322313178049327,
"grad_norm": 0.3635287582874298,
"learning_rate": 1.9791583472387783e-05,
"loss": 17.455885314941405,
"step": 2210
},
{
"epoch": 0.09364495590619686,
"grad_norm": 0.392044335603714,
"learning_rate": 1.978879959187577e-05,
"loss": 17.156585693359375,
"step": 2220
},
{
"epoch": 0.09406678003190044,
"grad_norm": 0.48159560561180115,
"learning_rate": 1.978599744036504e-05,
"loss": 17.075947570800782,
"step": 2230
},
{
"epoch": 0.09448860415760404,
"grad_norm": 0.3881451189517975,
"learning_rate": 1.9783177023085864e-05,
"loss": 17.03486022949219,
"step": 2240
},
{
"epoch": 0.09491042828330763,
"grad_norm": 0.5363468527793884,
"learning_rate": 1.9780338345302605e-05,
"loss": 16.983058166503906,
"step": 2250
},
{
"epoch": 0.09533225240901122,
"grad_norm": 0.45004919171333313,
"learning_rate": 1.9777481412313717e-05,
"loss": 16.96190185546875,
"step": 2260
},
{
"epoch": 0.09575407653471481,
"grad_norm": 0.62456214427948,
"learning_rate": 1.9774606229451727e-05,
"loss": 16.844752502441406,
"step": 2270
},
{
"epoch": 0.0961759006604184,
"grad_norm": 0.43555748462677,
"learning_rate": 1.977171280208322e-05,
"loss": 16.13968505859375,
"step": 2280
},
{
"epoch": 0.09659772478612198,
"grad_norm": 0.45719555020332336,
"learning_rate": 1.976880113560885e-05,
"loss": 17.10240936279297,
"step": 2290
},
{
"epoch": 0.09701954891182557,
"grad_norm": 0.36745473742485046,
"learning_rate": 1.9765871235463288e-05,
"loss": 16.881980895996094,
"step": 2300
},
{
"epoch": 0.09744137303752917,
"grad_norm": 0.3965582847595215,
"learning_rate": 1.9762923107115264e-05,
"loss": 16.904936218261717,
"step": 2310
},
{
"epoch": 0.09786319716323276,
"grad_norm": 0.3897360861301422,
"learning_rate": 1.9759956756067514e-05,
"loss": 17.281739807128908,
"step": 2320
},
{
"epoch": 0.09828502128893635,
"grad_norm": 0.422806054353714,
"learning_rate": 1.97569721878568e-05,
"loss": 16.82386016845703,
"step": 2330
},
{
"epoch": 0.09870684541463993,
"grad_norm": 0.5344931483268738,
"learning_rate": 1.975396940805388e-05,
"loss": 16.420701599121095,
"step": 2340
},
{
"epoch": 0.09912866954034352,
"grad_norm": 0.4889315068721771,
"learning_rate": 1.97509484222635e-05,
"loss": 16.85874786376953,
"step": 2350
},
{
"epoch": 0.09955049366604711,
"grad_norm": 0.4195805490016937,
"learning_rate": 1.97479092361244e-05,
"loss": 16.6565673828125,
"step": 2360
},
{
"epoch": 0.0999723177917507,
"grad_norm": 0.4987180531024933,
"learning_rate": 1.9744851855309276e-05,
"loss": 16.764617919921875,
"step": 2370
},
{
"epoch": 0.10039414191745429,
"grad_norm": 0.5001860857009888,
"learning_rate": 1.9741776285524804e-05,
"loss": 16.95183563232422,
"step": 2380
},
{
"epoch": 0.10081596604315789,
"grad_norm": 0.36856579780578613,
"learning_rate": 1.973868253251159e-05,
"loss": 17.013652038574218,
"step": 2390
},
{
"epoch": 0.10123779016886147,
"grad_norm": 0.42868202924728394,
"learning_rate": 1.9735570602044194e-05,
"loss": 17.014599609375,
"step": 2400
},
{
"epoch": 0.10165961429456506,
"grad_norm": 0.5270141959190369,
"learning_rate": 1.9732440499931097e-05,
"loss": 16.600096130371092,
"step": 2410
},
{
"epoch": 0.10208143842026865,
"grad_norm": 0.46056610345840454,
"learning_rate": 1.97292922320147e-05,
"loss": 16.69386444091797,
"step": 2420
},
{
"epoch": 0.10250326254597224,
"grad_norm": 0.5768168568611145,
"learning_rate": 1.972612580417131e-05,
"loss": 17.025120544433594,
"step": 2430
},
{
"epoch": 0.10292508667167583,
"grad_norm": 0.42351484298706055,
"learning_rate": 1.972294122231113e-05,
"loss": 17.150408935546874,
"step": 2440
},
{
"epoch": 0.10334691079737941,
"grad_norm": 0.4627014398574829,
"learning_rate": 1.9719738492378254e-05,
"loss": 16.846009826660158,
"step": 2450
},
{
"epoch": 0.103768734923083,
"grad_norm": 0.49214082956314087,
"learning_rate": 1.9716517620350637e-05,
"loss": 17.20526123046875,
"step": 2460
},
{
"epoch": 0.1041905590487866,
"grad_norm": 0.49011021852493286,
"learning_rate": 1.971327861224011e-05,
"loss": 16.686956787109374,
"step": 2470
},
{
"epoch": 0.10461238317449019,
"grad_norm": 0.4025215804576874,
"learning_rate": 1.9710021474092343e-05,
"loss": 16.620086669921875,
"step": 2480
},
{
"epoch": 0.10503420730019378,
"grad_norm": 0.40574750304222107,
"learning_rate": 1.970674621198686e-05,
"loss": 16.245057678222658,
"step": 2490
},
{
"epoch": 0.10545603142589736,
"grad_norm": 0.38037168979644775,
"learning_rate": 1.9703452832036995e-05,
"loss": 16.545228576660158,
"step": 2500
}
],
"logging_steps": 10,
"max_steps": 23707,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}