rqadri's picture
Upload folder using huggingface_hub
35c4503 verified
Raw
History Blame Contribute Delete
70.5 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.08537506293002181,
"eval_steps": 500,
"global_step": 4070,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00020976673938580297,
"grad_norm": 236.7567785411443,
"learning_rate": 2.910112359550562e-06,
"loss": 1.5920657157897948,
"step": 10
},
{
"epoch": 0.00041953347877160595,
"grad_norm": 152.92314363522394,
"learning_rate": 3.921348314606742e-06,
"loss": 0.9421598434448242,
"step": 20
},
{
"epoch": 0.000629300218157409,
"grad_norm": 97.19978632147773,
"learning_rate": 4.932584269662922e-06,
"loss": 0.7996736526489258,
"step": 30
},
{
"epoch": 0.0008390669575432119,
"grad_norm": 157.7423675487546,
"learning_rate": 5.943820224719102e-06,
"loss": 0.760670280456543,
"step": 40
},
{
"epoch": 0.001048833696929015,
"grad_norm": 49.50204924369943,
"learning_rate": 6.955056179775282e-06,
"loss": 0.7635975837707519,
"step": 50
},
{
"epoch": 0.001258600436314818,
"grad_norm": 16.48624285559217,
"learning_rate": 7.966292134831462e-06,
"loss": 0.7791356563568115,
"step": 60
},
{
"epoch": 0.0014683671757006209,
"grad_norm": 6.858907596379825,
"learning_rate": 8.977528089887641e-06,
"loss": 0.7495433330535889,
"step": 70
},
{
"epoch": 0.0016781339150864238,
"grad_norm": 26.435928023185784,
"learning_rate": 9.988764044943822e-06,
"loss": 0.7318498134613037,
"step": 80
},
{
"epoch": 0.001887900654472227,
"grad_norm": 214.57004030820568,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.7413080215454102,
"step": 90
},
{
"epoch": 0.00209766739385803,
"grad_norm": 7.939148544393301,
"learning_rate": 1.2011235955056182e-05,
"loss": 0.8061410903930664,
"step": 100
},
{
"epoch": 0.0023074341332438328,
"grad_norm": 14.183647299272815,
"learning_rate": 1.3022471910112362e-05,
"loss": 0.7331802368164062,
"step": 110
},
{
"epoch": 0.002517200872629636,
"grad_norm": 22.423640491145502,
"learning_rate": 1.403370786516854e-05,
"loss": 0.7498276710510254,
"step": 120
},
{
"epoch": 0.002726967612015439,
"grad_norm": 3.136265232048693,
"learning_rate": 1.504494382022472e-05,
"loss": 0.724825668334961,
"step": 130
},
{
"epoch": 0.0029367343514012417,
"grad_norm": 8.214245546263463,
"learning_rate": 1.60561797752809e-05,
"loss": 0.6886782646179199,
"step": 140
},
{
"epoch": 0.003146501090787045,
"grad_norm": 5.619530923664304,
"learning_rate": 1.706741573033708e-05,
"loss": 0.7081296443939209,
"step": 150
},
{
"epoch": 0.0033562678301728476,
"grad_norm": 10.033723007819054,
"learning_rate": 1.8078651685393256e-05,
"loss": 0.7299324035644531,
"step": 160
},
{
"epoch": 0.0035660345695586507,
"grad_norm": 11.134808180954863,
"learning_rate": 1.908988764044944e-05,
"loss": 0.745941162109375,
"step": 170
},
{
"epoch": 0.003775801308944454,
"grad_norm": 2.4304907872424795,
"learning_rate": 2e-05,
"loss": 0.7129830360412598,
"step": 180
},
{
"epoch": 0.0039855680483302566,
"grad_norm": 4.018584134102583,
"learning_rate": 2e-05,
"loss": 0.7092458724975585,
"step": 190
},
{
"epoch": 0.00419533478771606,
"grad_norm": 5.006464475899677,
"learning_rate": 2e-05,
"loss": 0.7565587520599365,
"step": 200
},
{
"epoch": 0.004405101527101863,
"grad_norm": 2.800599112711121,
"learning_rate": 2e-05,
"loss": 0.6959824562072754,
"step": 210
},
{
"epoch": 0.0046148682664876655,
"grad_norm": 6.231621527897506,
"learning_rate": 2e-05,
"loss": 0.6992167472839356,
"step": 220
},
{
"epoch": 0.004824635005873469,
"grad_norm": 10.811727778354081,
"learning_rate": 2e-05,
"loss": 0.6952213287353516,
"step": 230
},
{
"epoch": 0.005034401745259272,
"grad_norm": 11.261067852316236,
"learning_rate": 2e-05,
"loss": 0.7460278987884521,
"step": 240
},
{
"epoch": 0.0052441684846450745,
"grad_norm": 7.649125680617738,
"learning_rate": 2e-05,
"loss": 0.7858521461486816,
"step": 250
},
{
"epoch": 0.005453935224030878,
"grad_norm": 4.490247060675485,
"learning_rate": 2e-05,
"loss": 0.7642570018768311,
"step": 260
},
{
"epoch": 0.005663701963416681,
"grad_norm": 5.320163840820431,
"learning_rate": 2e-05,
"loss": 0.7289691925048828,
"step": 270
},
{
"epoch": 0.0058734687028024835,
"grad_norm": 3.4799647445683717,
"learning_rate": 2e-05,
"loss": 0.6995810508728028,
"step": 280
},
{
"epoch": 0.006083235442188286,
"grad_norm": 34.442893954128465,
"learning_rate": 2e-05,
"loss": 0.7102492809295654,
"step": 290
},
{
"epoch": 0.00629300218157409,
"grad_norm": 4.179341348142214,
"learning_rate": 2e-05,
"loss": 0.7110370635986328,
"step": 300
},
{
"epoch": 0.0065027689209598925,
"grad_norm": 4.270936726323635,
"learning_rate": 2e-05,
"loss": 0.7388000011444091,
"step": 310
},
{
"epoch": 0.006712535660345695,
"grad_norm": 3.52144124715383,
"learning_rate": 2e-05,
"loss": 0.7715152740478516,
"step": 320
},
{
"epoch": 0.006922302399731499,
"grad_norm": 2.1760255438109892,
"learning_rate": 2e-05,
"loss": 0.7474425315856934,
"step": 330
},
{
"epoch": 0.0071320691391173014,
"grad_norm": 1.8391052662069023,
"learning_rate": 2e-05,
"loss": 0.7086240291595459,
"step": 340
},
{
"epoch": 0.007341835878503104,
"grad_norm": 9.85499125185226,
"learning_rate": 2e-05,
"loss": 0.736334228515625,
"step": 350
},
{
"epoch": 0.007551602617888908,
"grad_norm": 7.061981460856605,
"learning_rate": 2e-05,
"loss": 0.7058637619018555,
"step": 360
},
{
"epoch": 0.00776136935727471,
"grad_norm": 2.399773463771947,
"learning_rate": 2e-05,
"loss": 0.7035298824310303,
"step": 370
},
{
"epoch": 0.007971136096660513,
"grad_norm": 1.8715337008198967,
"learning_rate": 2e-05,
"loss": 0.7319884300231934,
"step": 380
},
{
"epoch": 0.008180902836046316,
"grad_norm": 4.402380567892121,
"learning_rate": 2e-05,
"loss": 0.6931228637695312,
"step": 390
},
{
"epoch": 0.00839066957543212,
"grad_norm": 1.997385308499001,
"learning_rate": 2e-05,
"loss": 0.6841664791107178,
"step": 400
},
{
"epoch": 0.008600436314817923,
"grad_norm": 8.77520106541026,
"learning_rate": 2e-05,
"loss": 0.6674545288085938,
"step": 410
},
{
"epoch": 0.008810203054203726,
"grad_norm": 227.00469733968774,
"learning_rate": 2e-05,
"loss": 0.7310820579528808,
"step": 420
},
{
"epoch": 0.009019969793589528,
"grad_norm": 3.283083883742064,
"learning_rate": 2e-05,
"loss": 0.7967383384704589,
"step": 430
},
{
"epoch": 0.009229736532975331,
"grad_norm": 12.336376030486466,
"learning_rate": 2e-05,
"loss": 0.7576329231262207,
"step": 440
},
{
"epoch": 0.009439503272361134,
"grad_norm": 1.8242777860324726,
"learning_rate": 2e-05,
"loss": 0.706461238861084,
"step": 450
},
{
"epoch": 0.009649270011746938,
"grad_norm": 1.8090965567853239,
"learning_rate": 2e-05,
"loss": 0.7032674312591553,
"step": 460
},
{
"epoch": 0.009859036751132741,
"grad_norm": 2.5475991670097953,
"learning_rate": 2e-05,
"loss": 0.7029439449310303,
"step": 470
},
{
"epoch": 0.010068803490518544,
"grad_norm": 1.4692934900457209,
"learning_rate": 2e-05,
"loss": 0.6630709648132325,
"step": 480
},
{
"epoch": 0.010278570229904346,
"grad_norm": 2.1331924790853534,
"learning_rate": 2e-05,
"loss": 0.7047128677368164,
"step": 490
},
{
"epoch": 0.010488336969290149,
"grad_norm": 2.4224465959516164,
"learning_rate": 2e-05,
"loss": 0.7062530517578125,
"step": 500
},
{
"epoch": 0.010698103708675952,
"grad_norm": 1.9672184607046945,
"learning_rate": 2e-05,
"loss": 0.6654829978942871,
"step": 510
},
{
"epoch": 0.010907870448061756,
"grad_norm": 1.8632555104139144,
"learning_rate": 2e-05,
"loss": 0.6872918128967285,
"step": 520
},
{
"epoch": 0.011117637187447559,
"grad_norm": 2.696765708839419,
"learning_rate": 2e-05,
"loss": 0.712721586227417,
"step": 530
},
{
"epoch": 0.011327403926833362,
"grad_norm": 3.3114397630142145,
"learning_rate": 2e-05,
"loss": 0.665414571762085,
"step": 540
},
{
"epoch": 0.011537170666219164,
"grad_norm": 1.7287592058057928,
"learning_rate": 2e-05,
"loss": 0.7328654289245605,
"step": 550
},
{
"epoch": 0.011746937405604967,
"grad_norm": 1.7622716775547056,
"learning_rate": 2e-05,
"loss": 0.692367696762085,
"step": 560
},
{
"epoch": 0.01195670414499077,
"grad_norm": 2.3708233226098807,
"learning_rate": 2e-05,
"loss": 0.6588814735412598,
"step": 570
},
{
"epoch": 0.012166470884376572,
"grad_norm": 4.863667938523831,
"learning_rate": 2e-05,
"loss": 0.7014576911926269,
"step": 580
},
{
"epoch": 0.012376237623762377,
"grad_norm": 2.1632337576331073,
"learning_rate": 2e-05,
"loss": 0.7060187816619873,
"step": 590
},
{
"epoch": 0.01258600436314818,
"grad_norm": 2.1446479504758584,
"learning_rate": 2e-05,
"loss": 0.7127746105194092,
"step": 600
},
{
"epoch": 0.012795771102533982,
"grad_norm": 2.103648073321234,
"learning_rate": 2e-05,
"loss": 0.7010409355163574,
"step": 610
},
{
"epoch": 0.013005537841919785,
"grad_norm": 2.300281953023738,
"learning_rate": 2e-05,
"loss": 0.7145485877990723,
"step": 620
},
{
"epoch": 0.013215304581305588,
"grad_norm": 2.1745603402017855,
"learning_rate": 2e-05,
"loss": 0.6661243438720703,
"step": 630
},
{
"epoch": 0.01342507132069139,
"grad_norm": 3.3911386633188085,
"learning_rate": 2e-05,
"loss": 0.7290327072143554,
"step": 640
},
{
"epoch": 0.013634838060077195,
"grad_norm": 2.838234667045874,
"learning_rate": 2e-05,
"loss": 0.7416918754577637,
"step": 650
},
{
"epoch": 0.013844604799462997,
"grad_norm": 2.6030763255310494,
"learning_rate": 2e-05,
"loss": 0.6962085247039795,
"step": 660
},
{
"epoch": 0.0140543715388488,
"grad_norm": 3.9879154447494263,
"learning_rate": 2e-05,
"loss": 0.7320832729339599,
"step": 670
},
{
"epoch": 0.014264138278234603,
"grad_norm": 5.147234411879652,
"learning_rate": 2e-05,
"loss": 0.690255880355835,
"step": 680
},
{
"epoch": 0.014473905017620406,
"grad_norm": 4.41792994494731,
"learning_rate": 2e-05,
"loss": 0.7107495784759521,
"step": 690
},
{
"epoch": 0.014683671757006208,
"grad_norm": 4.0092748239949465,
"learning_rate": 2e-05,
"loss": 0.7144205093383789,
"step": 700
},
{
"epoch": 0.014893438496392013,
"grad_norm": 3.663722199190995,
"learning_rate": 2e-05,
"loss": 0.678815221786499,
"step": 710
},
{
"epoch": 0.015103205235777815,
"grad_norm": 374.6324936776676,
"learning_rate": 2e-05,
"loss": 0.7061020851135253,
"step": 720
},
{
"epoch": 0.015312971975163618,
"grad_norm": 2.830711027072347,
"learning_rate": 2e-05,
"loss": 0.716461181640625,
"step": 730
},
{
"epoch": 0.01552273871454942,
"grad_norm": 17.687806831387658,
"learning_rate": 2e-05,
"loss": 0.6958938598632812,
"step": 740
},
{
"epoch": 0.015732505453935224,
"grad_norm": 67.22665881659304,
"learning_rate": 2e-05,
"loss": 0.6571521759033203,
"step": 750
},
{
"epoch": 0.015942272193321026,
"grad_norm": 3.2787186518009888,
"learning_rate": 2e-05,
"loss": 0.737470006942749,
"step": 760
},
{
"epoch": 0.01615203893270683,
"grad_norm": 5.764665619355419,
"learning_rate": 2e-05,
"loss": 0.6784905433654785,
"step": 770
},
{
"epoch": 0.01636180567209263,
"grad_norm": 2.8187442313274156,
"learning_rate": 2e-05,
"loss": 0.698319387435913,
"step": 780
},
{
"epoch": 0.016571572411478434,
"grad_norm": 11.052681670850987,
"learning_rate": 2e-05,
"loss": 0.7136563301086426,
"step": 790
},
{
"epoch": 0.01678133915086424,
"grad_norm": 2.738088534726021,
"learning_rate": 2e-05,
"loss": 0.7217898368835449,
"step": 800
},
{
"epoch": 0.016991105890250043,
"grad_norm": 22.99130603074463,
"learning_rate": 2e-05,
"loss": 0.6949929237365723,
"step": 810
},
{
"epoch": 0.017200872629635846,
"grad_norm": 2.2673846343594293,
"learning_rate": 2e-05,
"loss": 0.6527021408081055,
"step": 820
},
{
"epoch": 0.01741063936902165,
"grad_norm": 4.040669613703935,
"learning_rate": 2e-05,
"loss": 0.6478307723999024,
"step": 830
},
{
"epoch": 0.01762040610840745,
"grad_norm": 1.7450269280537714,
"learning_rate": 2e-05,
"loss": 0.7058829307556153,
"step": 840
},
{
"epoch": 0.017830172847793254,
"grad_norm": 1.3888768876108615,
"learning_rate": 2e-05,
"loss": 0.6917304039001465,
"step": 850
},
{
"epoch": 0.018039939587179057,
"grad_norm": 2.497192781134032,
"learning_rate": 2e-05,
"loss": 0.6497130393981934,
"step": 860
},
{
"epoch": 0.01824970632656486,
"grad_norm": 2.096909255558357,
"learning_rate": 2e-05,
"loss": 0.6941553115844726,
"step": 870
},
{
"epoch": 0.018459473065950662,
"grad_norm": 1.7631454307114556,
"learning_rate": 2e-05,
"loss": 0.7313367366790772,
"step": 880
},
{
"epoch": 0.018669239805336465,
"grad_norm": 1.2137215799580678,
"learning_rate": 2e-05,
"loss": 0.734074592590332,
"step": 890
},
{
"epoch": 0.018879006544722268,
"grad_norm": 8.527470156498259,
"learning_rate": 2e-05,
"loss": 0.7139863967895508,
"step": 900
},
{
"epoch": 0.01908877328410807,
"grad_norm": 1.150419868331103,
"learning_rate": 2e-05,
"loss": 0.7071406364440918,
"step": 910
},
{
"epoch": 0.019298540023493876,
"grad_norm": 2.4598251273735596,
"learning_rate": 2e-05,
"loss": 0.7090801239013672,
"step": 920
},
{
"epoch": 0.01950830676287968,
"grad_norm": 2.0085089193495707,
"learning_rate": 2e-05,
"loss": 0.6835993766784668,
"step": 930
},
{
"epoch": 0.019718073502265482,
"grad_norm": 3.0778601053919536,
"learning_rate": 2e-05,
"loss": 0.7152371883392334,
"step": 940
},
{
"epoch": 0.019927840241651285,
"grad_norm": 6.504260026804486,
"learning_rate": 2e-05,
"loss": 0.6743127346038819,
"step": 950
},
{
"epoch": 0.020137606981037087,
"grad_norm": 1.7969857643634526,
"learning_rate": 2e-05,
"loss": 0.6778482437133789,
"step": 960
},
{
"epoch": 0.02034737372042289,
"grad_norm": 29.433979140864412,
"learning_rate": 2e-05,
"loss": 0.6873385429382324,
"step": 970
},
{
"epoch": 0.020557140459808693,
"grad_norm": 2.035543156408046,
"learning_rate": 2e-05,
"loss": 0.6699578762054443,
"step": 980
},
{
"epoch": 0.020766907199194495,
"grad_norm": 1.1357536583710182,
"learning_rate": 2e-05,
"loss": 0.695745849609375,
"step": 990
},
{
"epoch": 0.020976673938580298,
"grad_norm": 3.7046794359871447,
"learning_rate": 2e-05,
"loss": 0.6789576530456543,
"step": 1000
},
{
"epoch": 0.0211864406779661,
"grad_norm": 3.130957938855976,
"learning_rate": 2e-05,
"loss": 0.6930006980895996,
"step": 1010
},
{
"epoch": 0.021396207417351903,
"grad_norm": 11.276914781871405,
"learning_rate": 2e-05,
"loss": 0.6952384948730469,
"step": 1020
},
{
"epoch": 0.021605974156737706,
"grad_norm": 6.337861678743805,
"learning_rate": 2e-05,
"loss": 0.6785932064056397,
"step": 1030
},
{
"epoch": 0.021815740896123512,
"grad_norm": 4.784777304261159,
"learning_rate": 2e-05,
"loss": 0.7035085678100585,
"step": 1040
},
{
"epoch": 0.022025507635509315,
"grad_norm": 5.679658714316126,
"learning_rate": 2e-05,
"loss": 0.6616763114929199,
"step": 1050
},
{
"epoch": 0.022235274374895118,
"grad_norm": 3.478649303767567,
"learning_rate": 2e-05,
"loss": 0.6685798645019532,
"step": 1060
},
{
"epoch": 0.02244504111428092,
"grad_norm": 5.634898233949741,
"learning_rate": 2e-05,
"loss": 0.7499291896820068,
"step": 1070
},
{
"epoch": 0.022654807853666723,
"grad_norm": 5.773543600338411,
"learning_rate": 2e-05,
"loss": 0.6558557033538819,
"step": 1080
},
{
"epoch": 0.022864574593052526,
"grad_norm": 39.700232901357666,
"learning_rate": 2e-05,
"loss": 0.6942555904388428,
"step": 1090
},
{
"epoch": 0.02307434133243833,
"grad_norm": 6.2820558818505265,
"learning_rate": 2e-05,
"loss": 0.6771759033203125,
"step": 1100
},
{
"epoch": 0.02328410807182413,
"grad_norm": 2.9316401080763823,
"learning_rate": 2e-05,
"loss": 0.6969615936279296,
"step": 1110
},
{
"epoch": 0.023493874811209934,
"grad_norm": 2.8974904226392466,
"learning_rate": 2e-05,
"loss": 0.6931879997253418,
"step": 1120
},
{
"epoch": 0.023703641550595737,
"grad_norm": 4.189761025368283,
"learning_rate": 2e-05,
"loss": 0.6718255043029785,
"step": 1130
},
{
"epoch": 0.02391340828998154,
"grad_norm": 2.417352201543177,
"learning_rate": 2e-05,
"loss": 0.7471983909606934,
"step": 1140
},
{
"epoch": 0.024123175029367342,
"grad_norm": 2.839365825197699,
"learning_rate": 2e-05,
"loss": 0.6793410301208496,
"step": 1150
},
{
"epoch": 0.024332941768753145,
"grad_norm": 3.141367299942584,
"learning_rate": 2e-05,
"loss": 0.7086249351501465,
"step": 1160
},
{
"epoch": 0.02454270850813895,
"grad_norm": 2.060642183486379,
"learning_rate": 2e-05,
"loss": 0.6504176139831543,
"step": 1170
},
{
"epoch": 0.024752475247524754,
"grad_norm": 23.95719687060173,
"learning_rate": 2e-05,
"loss": 0.6888862609863281,
"step": 1180
},
{
"epoch": 0.024962241986910556,
"grad_norm": 5.209642741753617,
"learning_rate": 2e-05,
"loss": 0.6663257598876953,
"step": 1190
},
{
"epoch": 0.02517200872629636,
"grad_norm": 4.668122000612013,
"learning_rate": 2e-05,
"loss": 0.6789710521697998,
"step": 1200
},
{
"epoch": 0.025381775465682162,
"grad_norm": 2.21317699920472,
"learning_rate": 2e-05,
"loss": 0.6676199913024903,
"step": 1210
},
{
"epoch": 0.025591542205067964,
"grad_norm": 4.368293636030707,
"learning_rate": 2e-05,
"loss": 0.690043592453003,
"step": 1220
},
{
"epoch": 0.025801308944453767,
"grad_norm": 3.8199119576336176,
"learning_rate": 2e-05,
"loss": 0.7009891033172607,
"step": 1230
},
{
"epoch": 0.02601107568383957,
"grad_norm": 2.3977516666695142,
"learning_rate": 2e-05,
"loss": 0.6778505325317383,
"step": 1240
},
{
"epoch": 0.026220842423225373,
"grad_norm": 12.32747378491102,
"learning_rate": 2e-05,
"loss": 0.7028230190277099,
"step": 1250
},
{
"epoch": 0.026430609162611175,
"grad_norm": 5.6463852018519844,
"learning_rate": 2e-05,
"loss": 0.6816394329071045,
"step": 1260
},
{
"epoch": 0.026640375901996978,
"grad_norm": 3.0036641576788443,
"learning_rate": 2e-05,
"loss": 0.6653374671936035,
"step": 1270
},
{
"epoch": 0.02685014264138278,
"grad_norm": 3.963136105803816,
"learning_rate": 2e-05,
"loss": 0.70212721824646,
"step": 1280
},
{
"epoch": 0.027059909380768587,
"grad_norm": 2.3922579167108338,
"learning_rate": 2e-05,
"loss": 0.6652139186859131,
"step": 1290
},
{
"epoch": 0.02726967612015439,
"grad_norm": 1.430307897978201,
"learning_rate": 2e-05,
"loss": 0.649704122543335,
"step": 1300
},
{
"epoch": 0.027479442859540192,
"grad_norm": 12.579409728745457,
"learning_rate": 2e-05,
"loss": 0.7093433380126953,
"step": 1310
},
{
"epoch": 0.027689209598925995,
"grad_norm": 1.8768014730990727,
"learning_rate": 2e-05,
"loss": 0.6947019100189209,
"step": 1320
},
{
"epoch": 0.027898976338311798,
"grad_norm": 4.087020619174566,
"learning_rate": 2e-05,
"loss": 0.6580727577209473,
"step": 1330
},
{
"epoch": 0.0281087430776976,
"grad_norm": 2.2104247401520505,
"learning_rate": 2e-05,
"loss": 0.6761454582214356,
"step": 1340
},
{
"epoch": 0.028318509817083403,
"grad_norm": 1.54359381136085,
"learning_rate": 2e-05,
"loss": 0.7064968109130859,
"step": 1350
},
{
"epoch": 0.028528276556469206,
"grad_norm": 9.445161276396444,
"learning_rate": 2e-05,
"loss": 0.669188404083252,
"step": 1360
},
{
"epoch": 0.02873804329585501,
"grad_norm": 2.7373325593304836,
"learning_rate": 2e-05,
"loss": 0.6705168724060059,
"step": 1370
},
{
"epoch": 0.02894781003524081,
"grad_norm": 2.276476538972258,
"learning_rate": 2e-05,
"loss": 0.7053794860839844,
"step": 1380
},
{
"epoch": 0.029157576774626614,
"grad_norm": 9.277889149036925,
"learning_rate": 2e-05,
"loss": 0.6553175926208497,
"step": 1390
},
{
"epoch": 0.029367343514012417,
"grad_norm": 3.368267525079364,
"learning_rate": 2e-05,
"loss": 0.6869246959686279,
"step": 1400
},
{
"epoch": 0.029577110253398223,
"grad_norm": 3.2226378523733645,
"learning_rate": 2e-05,
"loss": 0.6814833641052246,
"step": 1410
},
{
"epoch": 0.029786876992784025,
"grad_norm": 1.930571790233891,
"learning_rate": 2e-05,
"loss": 0.6552035331726074,
"step": 1420
},
{
"epoch": 0.029996643732169828,
"grad_norm": 3.5201808130997003,
"learning_rate": 2e-05,
"loss": 0.6717945098876953,
"step": 1430
},
{
"epoch": 0.03020641047155563,
"grad_norm": 660.3628631506493,
"learning_rate": 2e-05,
"loss": 0.6539801597595215,
"step": 1440
},
{
"epoch": 0.030416177210941434,
"grad_norm": 5.251919957664849,
"learning_rate": 2e-05,
"loss": 0.6830947875976563,
"step": 1450
},
{
"epoch": 0.030625943950327236,
"grad_norm": 5.249421083313879,
"learning_rate": 2e-05,
"loss": 0.681186580657959,
"step": 1460
},
{
"epoch": 0.03083571068971304,
"grad_norm": 2.0509657202306752,
"learning_rate": 2e-05,
"loss": 0.6720983028411865,
"step": 1470
},
{
"epoch": 0.03104547742909884,
"grad_norm": 2.2734683613081064,
"learning_rate": 2e-05,
"loss": 0.67911696434021,
"step": 1480
},
{
"epoch": 0.031255244168484644,
"grad_norm": 1.4795523059161764,
"learning_rate": 2e-05,
"loss": 0.6568414211273194,
"step": 1490
},
{
"epoch": 0.03146501090787045,
"grad_norm": 1.2453772205405704,
"learning_rate": 2e-05,
"loss": 0.6762202262878418,
"step": 1500
},
{
"epoch": 0.03167477764725625,
"grad_norm": 2.0010667543841123,
"learning_rate": 2e-05,
"loss": 0.6813722133636475,
"step": 1510
},
{
"epoch": 0.03188454438664205,
"grad_norm": 39.56264214911426,
"learning_rate": 2e-05,
"loss": 0.7298669815063477,
"step": 1520
},
{
"epoch": 0.032094311126027855,
"grad_norm": 1.9507720129351338,
"learning_rate": 2e-05,
"loss": 0.6247622489929199,
"step": 1530
},
{
"epoch": 0.03230407786541366,
"grad_norm": 1.484907363192865,
"learning_rate": 2e-05,
"loss": 0.6759968757629394,
"step": 1540
},
{
"epoch": 0.03251384460479946,
"grad_norm": 1.9147431404888071,
"learning_rate": 2e-05,
"loss": 0.6800020217895508,
"step": 1550
},
{
"epoch": 0.03272361134418526,
"grad_norm": 3.5428313404092,
"learning_rate": 2e-05,
"loss": 0.6901631355285645,
"step": 1560
},
{
"epoch": 0.032933378083571066,
"grad_norm": 4.7250773473958105,
"learning_rate": 2e-05,
"loss": 0.6709762573242187,
"step": 1570
},
{
"epoch": 0.03314314482295687,
"grad_norm": 5.898810367933252,
"learning_rate": 2e-05,
"loss": 0.66232271194458,
"step": 1580
},
{
"epoch": 0.03335291156234268,
"grad_norm": 1.9369336440014253,
"learning_rate": 2e-05,
"loss": 0.7334044933319092,
"step": 1590
},
{
"epoch": 0.03356267830172848,
"grad_norm": 8.737915524161552,
"learning_rate": 2e-05,
"loss": 0.6792474269866944,
"step": 1600
},
{
"epoch": 0.033772445041114284,
"grad_norm": 1.9836799556383249,
"learning_rate": 2e-05,
"loss": 0.7033929824829102,
"step": 1610
},
{
"epoch": 0.033982211780500086,
"grad_norm": 1.1854978830042706,
"learning_rate": 2e-05,
"loss": 0.6635163307189942,
"step": 1620
},
{
"epoch": 0.03419197851988589,
"grad_norm": 1.2197328294566572,
"learning_rate": 2e-05,
"loss": 0.6634726524353027,
"step": 1630
},
{
"epoch": 0.03440174525927169,
"grad_norm": 1.4709834426885486,
"learning_rate": 2e-05,
"loss": 0.6794222831726074,
"step": 1640
},
{
"epoch": 0.034611511998657495,
"grad_norm": 1.5366826867181045,
"learning_rate": 2e-05,
"loss": 0.6605133056640625,
"step": 1650
},
{
"epoch": 0.0348212787380433,
"grad_norm": 2.9147800494231335,
"learning_rate": 2e-05,
"loss": 0.6418501853942871,
"step": 1660
},
{
"epoch": 0.0350310454774291,
"grad_norm": 1.202686476162776,
"learning_rate": 2e-05,
"loss": 0.633223295211792,
"step": 1670
},
{
"epoch": 0.0352408122168149,
"grad_norm": 0.9631875057504979,
"learning_rate": 2e-05,
"loss": 0.666954517364502,
"step": 1680
},
{
"epoch": 0.035450578956200705,
"grad_norm": 1.4192365422706907,
"learning_rate": 2e-05,
"loss": 0.6560389518737793,
"step": 1690
},
{
"epoch": 0.03566034569558651,
"grad_norm": 1.1713801817552898,
"learning_rate": 2e-05,
"loss": 0.641888427734375,
"step": 1700
},
{
"epoch": 0.03587011243497231,
"grad_norm": 1.4022563568499706,
"learning_rate": 2e-05,
"loss": 0.7102468013763428,
"step": 1710
},
{
"epoch": 0.036079879174358113,
"grad_norm": 2.0986091486389826,
"learning_rate": 2e-05,
"loss": 0.6136519908905029,
"step": 1720
},
{
"epoch": 0.036289645913743916,
"grad_norm": 1.1634525834102774,
"learning_rate": 2e-05,
"loss": 0.6924193382263184,
"step": 1730
},
{
"epoch": 0.03649941265312972,
"grad_norm": 2.081748039936143,
"learning_rate": 2e-05,
"loss": 0.6236453533172608,
"step": 1740
},
{
"epoch": 0.03670917939251552,
"grad_norm": 2.484079612019191,
"learning_rate": 2e-05,
"loss": 0.6446088314056396,
"step": 1750
},
{
"epoch": 0.036918946131901324,
"grad_norm": 2.142979943466157,
"learning_rate": 2e-05,
"loss": 0.6566426753997803,
"step": 1760
},
{
"epoch": 0.03712871287128713,
"grad_norm": 2.727971395956995,
"learning_rate": 2e-05,
"loss": 0.6372243404388428,
"step": 1770
},
{
"epoch": 0.03733847961067293,
"grad_norm": 74.34931845644519,
"learning_rate": 2e-05,
"loss": 0.6198902130126953,
"step": 1780
},
{
"epoch": 0.03754824635005873,
"grad_norm": 2.066092794611252,
"learning_rate": 2e-05,
"loss": 0.6555921554565429,
"step": 1790
},
{
"epoch": 0.037758013089444535,
"grad_norm": 3.0945696621385244,
"learning_rate": 2e-05,
"loss": 0.6903239250183105,
"step": 1800
},
{
"epoch": 0.03796777982883034,
"grad_norm": 2.0553070965106786,
"learning_rate": 2e-05,
"loss": 0.6789304733276367,
"step": 1810
},
{
"epoch": 0.03817754656821614,
"grad_norm": 1.9236355838089434,
"learning_rate": 2e-05,
"loss": 0.6835772514343261,
"step": 1820
},
{
"epoch": 0.03838731330760194,
"grad_norm": 1.1647199736335656,
"learning_rate": 2e-05,
"loss": 0.6774552822113037,
"step": 1830
},
{
"epoch": 0.03859708004698775,
"grad_norm": 5.471968133112629,
"learning_rate": 2e-05,
"loss": 0.6373446464538575,
"step": 1840
},
{
"epoch": 0.038806846786373556,
"grad_norm": 1.4710591021325636,
"learning_rate": 2e-05,
"loss": 0.6392849922180176,
"step": 1850
},
{
"epoch": 0.03901661352575936,
"grad_norm": 2.6999041311336756,
"learning_rate": 2e-05,
"loss": 0.6585689067840577,
"step": 1860
},
{
"epoch": 0.03922638026514516,
"grad_norm": 1.1333104342309857,
"learning_rate": 2e-05,
"loss": 0.6750880241394043,
"step": 1870
},
{
"epoch": 0.039436147004530964,
"grad_norm": 1.8072514966793143,
"learning_rate": 2e-05,
"loss": 0.6438388824462891,
"step": 1880
},
{
"epoch": 0.039645913743916766,
"grad_norm": 1.0781542511331554,
"learning_rate": 2e-05,
"loss": 0.6553819179534912,
"step": 1890
},
{
"epoch": 0.03985568048330257,
"grad_norm": 3.0719828705316616,
"learning_rate": 2e-05,
"loss": 0.6419807434082031,
"step": 1900
},
{
"epoch": 0.04006544722268837,
"grad_norm": 1.5436345848872006,
"learning_rate": 2e-05,
"loss": 0.6960256099700928,
"step": 1910
},
{
"epoch": 0.040275213962074174,
"grad_norm": 1.0011408997937439,
"learning_rate": 2e-05,
"loss": 0.6509086608886718,
"step": 1920
},
{
"epoch": 0.04048498070145998,
"grad_norm": 1.294803313028635,
"learning_rate": 2e-05,
"loss": 0.6446637153625489,
"step": 1930
},
{
"epoch": 0.04069474744084578,
"grad_norm": 1.4561867936361599,
"learning_rate": 2e-05,
"loss": 0.6471066474914551,
"step": 1940
},
{
"epoch": 0.04090451418023158,
"grad_norm": 1.420758879294619,
"learning_rate": 2e-05,
"loss": 0.6668454647064209,
"step": 1950
},
{
"epoch": 0.041114280919617385,
"grad_norm": 1.339731233422897,
"learning_rate": 2e-05,
"loss": 0.6452267646789551,
"step": 1960
},
{
"epoch": 0.04132404765900319,
"grad_norm": 3.1201962896167976,
"learning_rate": 2e-05,
"loss": 0.643893814086914,
"step": 1970
},
{
"epoch": 0.04153381439838899,
"grad_norm": 1.724067302763631,
"learning_rate": 2e-05,
"loss": 0.6555222988128662,
"step": 1980
},
{
"epoch": 0.04174358113777479,
"grad_norm": 2.166362359535913,
"learning_rate": 2e-05,
"loss": 0.6665771484375,
"step": 1990
},
{
"epoch": 0.041953347877160596,
"grad_norm": 2.744806451285816,
"learning_rate": 2e-05,
"loss": 0.6609567165374756,
"step": 2000
},
{
"epoch": 0.0421631146165464,
"grad_norm": 1.1718498443702758,
"learning_rate": 2e-05,
"loss": 0.6552879333496093,
"step": 2010
},
{
"epoch": 0.0423728813559322,
"grad_norm": 20.838352988266458,
"learning_rate": 2e-05,
"loss": 0.6685161590576172,
"step": 2020
},
{
"epoch": 0.042582648095318004,
"grad_norm": 1.0046844547391474,
"learning_rate": 2e-05,
"loss": 0.6807273864746094,
"step": 2030
},
{
"epoch": 0.04279241483470381,
"grad_norm": 1.0422462861274395,
"learning_rate": 2e-05,
"loss": 0.6735074996948243,
"step": 2040
},
{
"epoch": 0.04300218157408961,
"grad_norm": 2.3271030192407123,
"learning_rate": 2e-05,
"loss": 0.7008447170257568,
"step": 2050
},
{
"epoch": 0.04321194831347541,
"grad_norm": 1.0466437770509067,
"learning_rate": 2e-05,
"loss": 0.6856519222259522,
"step": 2060
},
{
"epoch": 0.043421715052861215,
"grad_norm": 1.8010026654547595,
"learning_rate": 2e-05,
"loss": 0.6528799057006835,
"step": 2070
},
{
"epoch": 0.043631481792247025,
"grad_norm": 1.3505658857593894,
"learning_rate": 2e-05,
"loss": 0.6574287414550781,
"step": 2080
},
{
"epoch": 0.04384124853163283,
"grad_norm": 1.4621149054171736,
"learning_rate": 2e-05,
"loss": 0.6371768951416016,
"step": 2090
},
{
"epoch": 0.04405101527101863,
"grad_norm": 1.063808823913143,
"learning_rate": 2e-05,
"loss": 0.6493932247161865,
"step": 2100
},
{
"epoch": 0.04426078201040443,
"grad_norm": 1.0941475684679507,
"learning_rate": 2e-05,
"loss": 0.6469996452331543,
"step": 2110
},
{
"epoch": 0.044470548749790235,
"grad_norm": 2.9013958178163155,
"learning_rate": 2e-05,
"loss": 0.6639802932739258,
"step": 2120
},
{
"epoch": 0.04468031548917604,
"grad_norm": 1.3274950977190392,
"learning_rate": 2e-05,
"loss": 0.6418691158294678,
"step": 2130
},
{
"epoch": 0.04489008222856184,
"grad_norm": 0.9289121119554267,
"learning_rate": 2e-05,
"loss": 0.6633244514465332,
"step": 2140
},
{
"epoch": 0.045099848967947644,
"grad_norm": 1.3315812946093735,
"learning_rate": 2e-05,
"loss": 0.6491587638854981,
"step": 2150
},
{
"epoch": 0.045309615707333446,
"grad_norm": 1.4176253250948148,
"learning_rate": 2e-05,
"loss": 0.6659342765808105,
"step": 2160
},
{
"epoch": 0.04551938244671925,
"grad_norm": 1.417591137202183,
"learning_rate": 2e-05,
"loss": 0.6725269317626953,
"step": 2170
},
{
"epoch": 0.04572914918610505,
"grad_norm": 1.13378390472357,
"learning_rate": 2e-05,
"loss": 0.6531981468200684,
"step": 2180
},
{
"epoch": 0.045938915925490854,
"grad_norm": 1.4942397260353406,
"learning_rate": 2e-05,
"loss": 0.6550272941589356,
"step": 2190
},
{
"epoch": 0.04614868266487666,
"grad_norm": 7.502507584526351,
"learning_rate": 2e-05,
"loss": 0.6380607604980468,
"step": 2200
},
{
"epoch": 0.04635844940426246,
"grad_norm": 2.0666055878713925,
"learning_rate": 2e-05,
"loss": 0.6765716552734375,
"step": 2210
},
{
"epoch": 0.04656821614364826,
"grad_norm": 3.7526650968480615,
"learning_rate": 2e-05,
"loss": 0.6970747947692871,
"step": 2220
},
{
"epoch": 0.046777982883034065,
"grad_norm": 1.0839625187073827,
"learning_rate": 2e-05,
"loss": 0.6326667785644531,
"step": 2230
},
{
"epoch": 0.04698774962241987,
"grad_norm": 1.4444230099278503,
"learning_rate": 2e-05,
"loss": 0.6718984127044678,
"step": 2240
},
{
"epoch": 0.04719751636180567,
"grad_norm": 5.2812918783484015,
"learning_rate": 2e-05,
"loss": 0.6714650630950928,
"step": 2250
},
{
"epoch": 0.04740728310119147,
"grad_norm": 1.3513967222907057,
"learning_rate": 2e-05,
"loss": 0.623006534576416,
"step": 2260
},
{
"epoch": 0.047617049840577276,
"grad_norm": 1.2616632620184862,
"learning_rate": 2e-05,
"loss": 0.6513717174530029,
"step": 2270
},
{
"epoch": 0.04782681657996308,
"grad_norm": 0.9944118591830172,
"learning_rate": 2e-05,
"loss": 0.6423852443695068,
"step": 2280
},
{
"epoch": 0.04803658331934888,
"grad_norm": 2.5602239897084993,
"learning_rate": 2e-05,
"loss": 0.6237055778503418,
"step": 2290
},
{
"epoch": 0.048246350058734684,
"grad_norm": 5.098047996535586,
"learning_rate": 2e-05,
"loss": 0.6563852310180665,
"step": 2300
},
{
"epoch": 0.04845611679812049,
"grad_norm": 2.4908638567535495,
"learning_rate": 2e-05,
"loss": 0.6325714111328125,
"step": 2310
},
{
"epoch": 0.04866588353750629,
"grad_norm": 7.314280111917294,
"learning_rate": 2e-05,
"loss": 0.6511282920837402,
"step": 2320
},
{
"epoch": 0.0488756502768921,
"grad_norm": 3.9514336521890736,
"learning_rate": 2e-05,
"loss": 0.6881688594818115,
"step": 2330
},
{
"epoch": 0.0490854170162779,
"grad_norm": 5.593476626022951,
"learning_rate": 2e-05,
"loss": 0.678310775756836,
"step": 2340
},
{
"epoch": 0.049295183755663705,
"grad_norm": 1.09495038750096,
"learning_rate": 2e-05,
"loss": 0.6575408935546875,
"step": 2350
},
{
"epoch": 0.04950495049504951,
"grad_norm": 1.7774709237343953,
"learning_rate": 2e-05,
"loss": 0.631895637512207,
"step": 2360
},
{
"epoch": 0.04971471723443531,
"grad_norm": 6.054243872558155,
"learning_rate": 2e-05,
"loss": 0.6792542934417725,
"step": 2370
},
{
"epoch": 0.04992448397382111,
"grad_norm": 10.448245720311721,
"learning_rate": 2e-05,
"loss": 0.6161588668823242,
"step": 2380
},
{
"epoch": 0.050134250713206915,
"grad_norm": 4.349668514735628,
"learning_rate": 2e-05,
"loss": 0.6573549270629883,
"step": 2390
},
{
"epoch": 0.05034401745259272,
"grad_norm": 1.0492871670169195,
"learning_rate": 2e-05,
"loss": 0.6595412254333496,
"step": 2400
},
{
"epoch": 0.05055378419197852,
"grad_norm": 4.208402496962076,
"learning_rate": 2e-05,
"loss": 0.6440282821655273,
"step": 2410
},
{
"epoch": 0.050763550931364324,
"grad_norm": 1.6377864001891818,
"learning_rate": 2e-05,
"loss": 0.6259332180023194,
"step": 2420
},
{
"epoch": 0.050973317670750126,
"grad_norm": 3.5201211960751166,
"learning_rate": 2e-05,
"loss": 0.6841580867767334,
"step": 2430
},
{
"epoch": 0.05118308441013593,
"grad_norm": 3.1099904753941923,
"learning_rate": 2e-05,
"loss": 0.6848432064056397,
"step": 2440
},
{
"epoch": 0.05139285114952173,
"grad_norm": 5.180282203706244,
"learning_rate": 2e-05,
"loss": 0.6627151489257812,
"step": 2450
},
{
"epoch": 0.051602617888907534,
"grad_norm": 4.740946754901601,
"learning_rate": 2e-05,
"loss": 0.6182768821716309,
"step": 2460
},
{
"epoch": 0.05181238462829334,
"grad_norm": 4.622580865957454,
"learning_rate": 2e-05,
"loss": 0.6499597549438476,
"step": 2470
},
{
"epoch": 0.05202215136767914,
"grad_norm": 5.26590300113533,
"learning_rate": 2e-05,
"loss": 0.6509732723236084,
"step": 2480
},
{
"epoch": 0.05223191810706494,
"grad_norm": 19.25978934862213,
"learning_rate": 2e-05,
"loss": 0.6599420070648193,
"step": 2490
},
{
"epoch": 0.052441684846450745,
"grad_norm": 6.960936314568276,
"learning_rate": 2e-05,
"loss": 0.6727987289428711,
"step": 2500
},
{
"epoch": 0.05265145158583655,
"grad_norm": 1.414075477389571,
"learning_rate": 2e-05,
"loss": 0.635031795501709,
"step": 2510
},
{
"epoch": 0.05286121832522235,
"grad_norm": 4.066649172039289,
"learning_rate": 2e-05,
"loss": 0.6597941398620606,
"step": 2520
},
{
"epoch": 0.05307098506460815,
"grad_norm": 2.364634340031721,
"learning_rate": 2e-05,
"loss": 0.6823457241058349,
"step": 2530
},
{
"epoch": 0.053280751803993956,
"grad_norm": 4.716601083750949,
"learning_rate": 2e-05,
"loss": 0.6498630523681641,
"step": 2540
},
{
"epoch": 0.05349051854337976,
"grad_norm": 5.285387148880005,
"learning_rate": 2e-05,
"loss": 0.639859962463379,
"step": 2550
},
{
"epoch": 0.05370028528276556,
"grad_norm": 1.5808950827421306,
"learning_rate": 2e-05,
"loss": 0.6604787349700928,
"step": 2560
},
{
"epoch": 0.05391005202215137,
"grad_norm": 1.5822303036177494,
"learning_rate": 2e-05,
"loss": 0.6555892944335937,
"step": 2570
},
{
"epoch": 0.054119818761537174,
"grad_norm": 2.9431810050217284,
"learning_rate": 2e-05,
"loss": 0.6334521293640136,
"step": 2580
},
{
"epoch": 0.054329585500922976,
"grad_norm": 1.7788715048102133,
"learning_rate": 2e-05,
"loss": 0.6453732967376709,
"step": 2590
},
{
"epoch": 0.05453935224030878,
"grad_norm": 2.50270616666147,
"learning_rate": 2e-05,
"loss": 0.6774460792541503,
"step": 2600
},
{
"epoch": 0.05474911897969458,
"grad_norm": 10.707562798836856,
"learning_rate": 2e-05,
"loss": 0.6564057350158692,
"step": 2610
},
{
"epoch": 0.054958885719080385,
"grad_norm": 7.655005229658258,
"learning_rate": 2e-05,
"loss": 0.6521840572357178,
"step": 2620
},
{
"epoch": 0.05516865245846619,
"grad_norm": 3.2841435118348516,
"learning_rate": 2e-05,
"loss": 0.656426191329956,
"step": 2630
},
{
"epoch": 0.05537841919785199,
"grad_norm": 1.4775967192316672,
"learning_rate": 2e-05,
"loss": 0.6167282104492188,
"step": 2640
},
{
"epoch": 0.05558818593723779,
"grad_norm": 2.1146359434093567,
"learning_rate": 2e-05,
"loss": 0.6516013145446777,
"step": 2650
},
{
"epoch": 0.055797952676623595,
"grad_norm": 3.922483673727723,
"learning_rate": 2e-05,
"loss": 0.6396924018859863,
"step": 2660
},
{
"epoch": 0.0560077194160094,
"grad_norm": 3.5454952032437412,
"learning_rate": 2e-05,
"loss": 0.6680498123168945,
"step": 2670
},
{
"epoch": 0.0562174861553952,
"grad_norm": 1.3429385611798919,
"learning_rate": 2e-05,
"loss": 0.6435057640075683,
"step": 2680
},
{
"epoch": 0.056427252894781,
"grad_norm": 3.0538281935853266,
"learning_rate": 2e-05,
"loss": 0.6330676078796387,
"step": 2690
},
{
"epoch": 0.056637019634166806,
"grad_norm": 16.829414730202203,
"learning_rate": 2e-05,
"loss": 0.6667710781097412,
"step": 2700
},
{
"epoch": 0.05684678637355261,
"grad_norm": 4.39087328540998,
"learning_rate": 2e-05,
"loss": 0.633282995223999,
"step": 2710
},
{
"epoch": 0.05705655311293841,
"grad_norm": 5.287462256642327,
"learning_rate": 2e-05,
"loss": 0.6073061943054199,
"step": 2720
},
{
"epoch": 0.057266319852324214,
"grad_norm": 1.0838576240977882,
"learning_rate": 2e-05,
"loss": 0.6258288383483886,
"step": 2730
},
{
"epoch": 0.05747608659171002,
"grad_norm": 0.9900910697266448,
"learning_rate": 2e-05,
"loss": 0.6326725006103515,
"step": 2740
},
{
"epoch": 0.05768585333109582,
"grad_norm": 3.1076601428219357,
"learning_rate": 2e-05,
"loss": 0.6563446044921875,
"step": 2750
},
{
"epoch": 0.05789562007048162,
"grad_norm": 1.0038588549026581,
"learning_rate": 2e-05,
"loss": 0.6112082481384278,
"step": 2760
},
{
"epoch": 0.058105386809867425,
"grad_norm": 1.3442378974062874,
"learning_rate": 2e-05,
"loss": 0.6177424907684326,
"step": 2770
},
{
"epoch": 0.05831515354925323,
"grad_norm": 2.8108346453547948,
"learning_rate": 2e-05,
"loss": 0.6193189144134521,
"step": 2780
},
{
"epoch": 0.05852492028863903,
"grad_norm": 1.905205374029777,
"learning_rate": 2e-05,
"loss": 0.648875379562378,
"step": 2790
},
{
"epoch": 0.05873468702802483,
"grad_norm": 3.74823423744243,
"learning_rate": 2e-05,
"loss": 0.6043904781341553,
"step": 2800
},
{
"epoch": 0.05894445376741064,
"grad_norm": 18.168630622234243,
"learning_rate": 2e-05,
"loss": 0.6348827362060547,
"step": 2810
},
{
"epoch": 0.059154220506796446,
"grad_norm": 4.41375082760251,
"learning_rate": 2e-05,
"loss": 0.6650748252868652,
"step": 2820
},
{
"epoch": 0.05936398724618225,
"grad_norm": 4.666265939789063,
"learning_rate": 2e-05,
"loss": 0.646610164642334,
"step": 2830
},
{
"epoch": 0.05957375398556805,
"grad_norm": 7.434065397758597,
"learning_rate": 2e-05,
"loss": 0.6720587730407714,
"step": 2840
},
{
"epoch": 0.059783520724953854,
"grad_norm": 14.577190757849174,
"learning_rate": 2e-05,
"loss": 0.6914210319519043,
"step": 2850
},
{
"epoch": 0.059993287464339656,
"grad_norm": 7.690469349544798,
"learning_rate": 2e-05,
"loss": 0.6587874889373779,
"step": 2860
},
{
"epoch": 0.06020305420372546,
"grad_norm": 4.5317526555397265,
"learning_rate": 2e-05,
"loss": 0.648529052734375,
"step": 2870
},
{
"epoch": 0.06041282094311126,
"grad_norm": 2.4738472787325754,
"learning_rate": 2e-05,
"loss": 0.6423982143402099,
"step": 2880
},
{
"epoch": 0.060622587682497064,
"grad_norm": 9.389047282481764,
"learning_rate": 2e-05,
"loss": 0.6508796215057373,
"step": 2890
},
{
"epoch": 0.06083235442188287,
"grad_norm": 1.319738088613341,
"learning_rate": 2e-05,
"loss": 0.6483406066894531,
"step": 2900
},
{
"epoch": 0.06104212116126867,
"grad_norm": 1.3986494540414636,
"learning_rate": 2e-05,
"loss": 0.6722330093383789,
"step": 2910
},
{
"epoch": 0.06125188790065447,
"grad_norm": 1.3248596727395392,
"learning_rate": 2e-05,
"loss": 0.6521361827850342,
"step": 2920
},
{
"epoch": 0.061461654640040275,
"grad_norm": 8.01789018839688,
"learning_rate": 2e-05,
"loss": 0.6545214653015137,
"step": 2930
},
{
"epoch": 0.06167142137942608,
"grad_norm": 1.867876365359386,
"learning_rate": 2e-05,
"loss": 0.6807507514953614,
"step": 2940
},
{
"epoch": 0.06188118811881188,
"grad_norm": 1.9579123185059468,
"learning_rate": 2e-05,
"loss": 0.6590943813323975,
"step": 2950
},
{
"epoch": 0.06209095485819768,
"grad_norm": 4.666099434423608,
"learning_rate": 2e-05,
"loss": 0.6239662170410156,
"step": 2960
},
{
"epoch": 0.062300721597583486,
"grad_norm": 6.629107363464002,
"learning_rate": 2e-05,
"loss": 0.654464864730835,
"step": 2970
},
{
"epoch": 0.06251048833696929,
"grad_norm": 1.2941207879585257,
"learning_rate": 2e-05,
"loss": 0.6336467742919922,
"step": 2980
},
{
"epoch": 0.06272025507635509,
"grad_norm": 2.7567451703280867,
"learning_rate": 2e-05,
"loss": 0.6776866912841797,
"step": 2990
},
{
"epoch": 0.0629300218157409,
"grad_norm": 1.8713254083285533,
"learning_rate": 2e-05,
"loss": 0.6494972229003906,
"step": 3000
},
{
"epoch": 0.0631397885551267,
"grad_norm": 4.968645545352801,
"learning_rate": 2e-05,
"loss": 0.6557379722595215,
"step": 3010
},
{
"epoch": 0.0633495552945125,
"grad_norm": 5.29522254711333,
"learning_rate": 2e-05,
"loss": 0.6359572887420655,
"step": 3020
},
{
"epoch": 0.0635593220338983,
"grad_norm": 2.273924269870783,
"learning_rate": 2e-05,
"loss": 0.6379573822021485,
"step": 3030
},
{
"epoch": 0.0637690887732841,
"grad_norm": 1.1328251395273925,
"learning_rate": 2e-05,
"loss": 0.6374843120574951,
"step": 3040
},
{
"epoch": 0.06397885551266991,
"grad_norm": 15.368295822345088,
"learning_rate": 2e-05,
"loss": 0.6400310516357421,
"step": 3050
},
{
"epoch": 0.06418862225205571,
"grad_norm": 1.4607094526795188,
"learning_rate": 2e-05,
"loss": 0.6543275356292725,
"step": 3060
},
{
"epoch": 0.06439838899144151,
"grad_norm": 0.961287901503162,
"learning_rate": 2e-05,
"loss": 0.6664879798889161,
"step": 3070
},
{
"epoch": 0.06460815573082732,
"grad_norm": 2.191551250716189,
"learning_rate": 2e-05,
"loss": 0.6127400398254395,
"step": 3080
},
{
"epoch": 0.06481792247021312,
"grad_norm": 1.2296826571837989,
"learning_rate": 2e-05,
"loss": 0.6447425365447998,
"step": 3090
},
{
"epoch": 0.06502768920959892,
"grad_norm": 1.4143474486735077,
"learning_rate": 2e-05,
"loss": 0.6479999542236328,
"step": 3100
},
{
"epoch": 0.06523745594898472,
"grad_norm": 2.5009608440595854,
"learning_rate": 2e-05,
"loss": 0.6653982162475586,
"step": 3110
},
{
"epoch": 0.06544722268837053,
"grad_norm": 1.2857554395308046,
"learning_rate": 2e-05,
"loss": 0.6534938812255859,
"step": 3120
},
{
"epoch": 0.06565698942775633,
"grad_norm": 1.083653801140928,
"learning_rate": 2e-05,
"loss": 0.6390992641448975,
"step": 3130
},
{
"epoch": 0.06586675616714213,
"grad_norm": 1.8441477694886703,
"learning_rate": 2e-05,
"loss": 0.6085839748382569,
"step": 3140
},
{
"epoch": 0.06607652290652793,
"grad_norm": 2.533533734029522,
"learning_rate": 2e-05,
"loss": 0.6358766555786133,
"step": 3150
},
{
"epoch": 0.06628628964591374,
"grad_norm": 1.141802239759352,
"learning_rate": 2e-05,
"loss": 0.6275746822357178,
"step": 3160
},
{
"epoch": 0.06649605638529954,
"grad_norm": 1.4525638657170827,
"learning_rate": 2e-05,
"loss": 0.6507892608642578,
"step": 3170
},
{
"epoch": 0.06670582312468536,
"grad_norm": 1.1588031097230553,
"learning_rate": 2e-05,
"loss": 0.6174406051635742,
"step": 3180
},
{
"epoch": 0.06691558986407116,
"grad_norm": 2.5106093744433733,
"learning_rate": 2e-05,
"loss": 0.6564050674438476,
"step": 3190
},
{
"epoch": 0.06712535660345696,
"grad_norm": 1.5797241142582132,
"learning_rate": 2e-05,
"loss": 0.6210603713989258,
"step": 3200
},
{
"epoch": 0.06733512334284276,
"grad_norm": 1.1249954656400734,
"learning_rate": 2e-05,
"loss": 0.649308729171753,
"step": 3210
},
{
"epoch": 0.06754489008222857,
"grad_norm": 2.20992490602421,
"learning_rate": 2e-05,
"loss": 0.6323455810546875,
"step": 3220
},
{
"epoch": 0.06775465682161437,
"grad_norm": 4.469544409281363,
"learning_rate": 2e-05,
"loss": 0.6276360988616944,
"step": 3230
},
{
"epoch": 0.06796442356100017,
"grad_norm": 1.5415451708593162,
"learning_rate": 2e-05,
"loss": 0.6344407558441162,
"step": 3240
},
{
"epoch": 0.06817419030038598,
"grad_norm": 1.5949984481189907,
"learning_rate": 2e-05,
"loss": 0.6543967247009277,
"step": 3250
},
{
"epoch": 0.06838395703977178,
"grad_norm": 4.337420652054659,
"learning_rate": 2e-05,
"loss": 0.6937810897827148,
"step": 3260
},
{
"epoch": 0.06859372377915758,
"grad_norm": 5.830490852479801,
"learning_rate": 2e-05,
"loss": 0.6038520812988282,
"step": 3270
},
{
"epoch": 0.06880349051854338,
"grad_norm": 4.631265483483824,
"learning_rate": 2e-05,
"loss": 0.5950284004211426,
"step": 3280
},
{
"epoch": 0.06901325725792919,
"grad_norm": 1.3176387751671785,
"learning_rate": 2e-05,
"loss": 0.6093755245208741,
"step": 3290
},
{
"epoch": 0.06922302399731499,
"grad_norm": 1.4717375829615742,
"learning_rate": 2e-05,
"loss": 0.6016129970550537,
"step": 3300
},
{
"epoch": 0.06943279073670079,
"grad_norm": 1.030735076162183,
"learning_rate": 2e-05,
"loss": 0.6150879859924316,
"step": 3310
},
{
"epoch": 0.0696425574760866,
"grad_norm": 1.0121046050263844,
"learning_rate": 2e-05,
"loss": 0.6329584121704102,
"step": 3320
},
{
"epoch": 0.0698523242154724,
"grad_norm": 1.0842693289830008,
"learning_rate": 2e-05,
"loss": 0.6599737167358398,
"step": 3330
},
{
"epoch": 0.0700620909548582,
"grad_norm": 0.9212902872267731,
"learning_rate": 2e-05,
"loss": 0.5934088706970215,
"step": 3340
},
{
"epoch": 0.070271857694244,
"grad_norm": 1.4550776446428906,
"learning_rate": 2e-05,
"loss": 0.6075980186462402,
"step": 3350
},
{
"epoch": 0.0704816244336298,
"grad_norm": 1.5913770936750142,
"learning_rate": 2e-05,
"loss": 0.6325076580047607,
"step": 3360
},
{
"epoch": 0.07069139117301561,
"grad_norm": 1.7644895084736516,
"learning_rate": 2e-05,
"loss": 0.6227485656738281,
"step": 3370
},
{
"epoch": 0.07090115791240141,
"grad_norm": 1.6486442155005259,
"learning_rate": 2e-05,
"loss": 0.6122469902038574,
"step": 3380
},
{
"epoch": 0.07111092465178721,
"grad_norm": 2.1016110075472705,
"learning_rate": 2e-05,
"loss": 0.6439970970153809,
"step": 3390
},
{
"epoch": 0.07132069139117302,
"grad_norm": 1.1474703819153493,
"learning_rate": 2e-05,
"loss": 0.6407495021820069,
"step": 3400
},
{
"epoch": 0.07153045813055882,
"grad_norm": 1.6594212119031326,
"learning_rate": 2e-05,
"loss": 0.6275229454040527,
"step": 3410
},
{
"epoch": 0.07174022486994462,
"grad_norm": 1.5566863992273559,
"learning_rate": 2e-05,
"loss": 0.6257375717163086,
"step": 3420
},
{
"epoch": 0.07194999160933042,
"grad_norm": 1.3482677008573798,
"learning_rate": 2e-05,
"loss": 0.600681209564209,
"step": 3430
},
{
"epoch": 0.07215975834871623,
"grad_norm": 1.1085116434458875,
"learning_rate": 2e-05,
"loss": 0.6113192558288574,
"step": 3440
},
{
"epoch": 0.07236952508810203,
"grad_norm": 1.0470200412519313,
"learning_rate": 2e-05,
"loss": 0.6392256736755371,
"step": 3450
},
{
"epoch": 0.07257929182748783,
"grad_norm": 1.208091709460663,
"learning_rate": 2e-05,
"loss": 0.6172694206237793,
"step": 3460
},
{
"epoch": 0.07278905856687364,
"grad_norm": 1.391808307290741,
"learning_rate": 2e-05,
"loss": 0.6202544212341309,
"step": 3470
},
{
"epoch": 0.07299882530625944,
"grad_norm": 1.3324061305312844,
"learning_rate": 2e-05,
"loss": 0.6044974327087402,
"step": 3480
},
{
"epoch": 0.07320859204564524,
"grad_norm": 1.603768952138755,
"learning_rate": 2e-05,
"loss": 0.6273996353149414,
"step": 3490
},
{
"epoch": 0.07341835878503104,
"grad_norm": 1.1515236435171623,
"learning_rate": 2e-05,
"loss": 0.6415897369384765,
"step": 3500
},
{
"epoch": 0.07362812552441685,
"grad_norm": 1.2465427409602832,
"learning_rate": 2e-05,
"loss": 0.6239260673522949,
"step": 3510
},
{
"epoch": 0.07383789226380265,
"grad_norm": 3.195547073882619,
"learning_rate": 2e-05,
"loss": 0.6058053493499755,
"step": 3520
},
{
"epoch": 0.07404765900318845,
"grad_norm": 1.2612335270585324,
"learning_rate": 2e-05,
"loss": 0.6198372840881348,
"step": 3530
},
{
"epoch": 0.07425742574257425,
"grad_norm": 0.9419472600007744,
"learning_rate": 2e-05,
"loss": 0.6054319381713867,
"step": 3540
},
{
"epoch": 0.07446719248196006,
"grad_norm": 1.089531938815818,
"learning_rate": 2e-05,
"loss": 0.6167111873626709,
"step": 3550
},
{
"epoch": 0.07467695922134586,
"grad_norm": 1.3477436055009677,
"learning_rate": 2e-05,
"loss": 0.6194707870483398,
"step": 3560
},
{
"epoch": 0.07488672596073166,
"grad_norm": 1.4774128762036365,
"learning_rate": 2e-05,
"loss": 0.6233315944671631,
"step": 3570
},
{
"epoch": 0.07509649270011746,
"grad_norm": 2.180364295402137,
"learning_rate": 2e-05,
"loss": 0.6441976070404053,
"step": 3580
},
{
"epoch": 0.07530625943950327,
"grad_norm": 2.348014561727009,
"learning_rate": 2e-05,
"loss": 0.622609281539917,
"step": 3590
},
{
"epoch": 0.07551602617888907,
"grad_norm": 2.0546033771693244,
"learning_rate": 2e-05,
"loss": 0.6218820571899414,
"step": 3600
},
{
"epoch": 0.07572579291827487,
"grad_norm": 3.948630055653689,
"learning_rate": 2e-05,
"loss": 0.6077123165130616,
"step": 3610
},
{
"epoch": 0.07593555965766068,
"grad_norm": 1.425460243175804,
"learning_rate": 2e-05,
"loss": 0.6164697170257568,
"step": 3620
},
{
"epoch": 0.07614532639704648,
"grad_norm": 2.8560957332391355,
"learning_rate": 2e-05,
"loss": 0.6065490245819092,
"step": 3630
},
{
"epoch": 0.07635509313643228,
"grad_norm": 1.4382360748720746,
"learning_rate": 2e-05,
"loss": 0.631071949005127,
"step": 3640
},
{
"epoch": 0.07656485987581808,
"grad_norm": 2.4612191373604095,
"learning_rate": 2e-05,
"loss": 0.6297952175140381,
"step": 3650
},
{
"epoch": 0.07677462661520389,
"grad_norm": 0.9581431067857364,
"learning_rate": 2e-05,
"loss": 0.6396968841552735,
"step": 3660
},
{
"epoch": 0.0769843933545897,
"grad_norm": 0.84258061503981,
"learning_rate": 2e-05,
"loss": 0.6055449962615966,
"step": 3670
},
{
"epoch": 0.0771941600939755,
"grad_norm": 3.58877286495232,
"learning_rate": 2e-05,
"loss": 0.6315743923187256,
"step": 3680
},
{
"epoch": 0.07740392683336131,
"grad_norm": 1.293421173281621,
"learning_rate": 2e-05,
"loss": 0.6112199783325195,
"step": 3690
},
{
"epoch": 0.07761369357274711,
"grad_norm": 1.5684550738154965,
"learning_rate": 2e-05,
"loss": 0.6496015071868897,
"step": 3700
},
{
"epoch": 0.07782346031213291,
"grad_norm": 1.2200393519933037,
"learning_rate": 2e-05,
"loss": 0.5959799766540528,
"step": 3710
},
{
"epoch": 0.07803322705151872,
"grad_norm": 1.0163116397855736,
"learning_rate": 2e-05,
"loss": 0.6234962463378906,
"step": 3720
},
{
"epoch": 0.07824299379090452,
"grad_norm": 1.717517892283433,
"learning_rate": 2e-05,
"loss": 0.616788673400879,
"step": 3730
},
{
"epoch": 0.07845276053029032,
"grad_norm": 1.2080101695844014,
"learning_rate": 2e-05,
"loss": 0.6270909786224366,
"step": 3740
},
{
"epoch": 0.07866252726967612,
"grad_norm": 1.2642428576235951,
"learning_rate": 2e-05,
"loss": 0.634865665435791,
"step": 3750
},
{
"epoch": 0.07887229400906193,
"grad_norm": 1.7755472921599644,
"learning_rate": 2e-05,
"loss": 0.6161201477050782,
"step": 3760
},
{
"epoch": 0.07908206074844773,
"grad_norm": 1.1611024568844615,
"learning_rate": 2e-05,
"loss": 0.6064630508422851,
"step": 3770
},
{
"epoch": 0.07929182748783353,
"grad_norm": 1.3367012479877545,
"learning_rate": 2e-05,
"loss": 0.6257253170013428,
"step": 3780
},
{
"epoch": 0.07950159422721934,
"grad_norm": 1.0410919406167345,
"learning_rate": 2e-05,
"loss": 0.6141583442687988,
"step": 3790
},
{
"epoch": 0.07971136096660514,
"grad_norm": 0.9751600085611505,
"learning_rate": 2e-05,
"loss": 0.6254030227661133,
"step": 3800
},
{
"epoch": 0.07992112770599094,
"grad_norm": 1.4908385976253524,
"learning_rate": 2e-05,
"loss": 0.6212417602539062,
"step": 3810
},
{
"epoch": 0.08013089444537674,
"grad_norm": 1.3485953686507417,
"learning_rate": 2e-05,
"loss": 0.6176995754241943,
"step": 3820
},
{
"epoch": 0.08034066118476255,
"grad_norm": 1.2206130924099254,
"learning_rate": 2e-05,
"loss": 0.5810882568359375,
"step": 3830
},
{
"epoch": 0.08055042792414835,
"grad_norm": 1.6334570441244893,
"learning_rate": 2e-05,
"loss": 0.6379716873168946,
"step": 3840
},
{
"epoch": 0.08076019466353415,
"grad_norm": 1.3369545468363697,
"learning_rate": 2e-05,
"loss": 0.6377719879150391,
"step": 3850
},
{
"epoch": 0.08096996140291995,
"grad_norm": 1.8818167536690795,
"learning_rate": 2e-05,
"loss": 0.6293519496917724,
"step": 3860
},
{
"epoch": 0.08117972814230576,
"grad_norm": 1.4023282025205475,
"learning_rate": 2e-05,
"loss": 0.6685831069946289,
"step": 3870
},
{
"epoch": 0.08138949488169156,
"grad_norm": 1.2005685243156285,
"learning_rate": 2e-05,
"loss": 0.6329050064086914,
"step": 3880
},
{
"epoch": 0.08159926162107736,
"grad_norm": 1.0700487697578784,
"learning_rate": 2e-05,
"loss": 0.6079745292663574,
"step": 3890
},
{
"epoch": 0.08180902836046317,
"grad_norm": 0.9160942375676819,
"learning_rate": 2e-05,
"loss": 0.6010311126708985,
"step": 3900
},
{
"epoch": 0.08201879509984897,
"grad_norm": 0.9783151957554443,
"learning_rate": 2e-05,
"loss": 0.5906445503234863,
"step": 3910
},
{
"epoch": 0.08222856183923477,
"grad_norm": 2.1241581007326023,
"learning_rate": 2e-05,
"loss": 0.6245768547058106,
"step": 3920
},
{
"epoch": 0.08243832857862057,
"grad_norm": 1.1142880534951982,
"learning_rate": 2e-05,
"loss": 0.6391409873962403,
"step": 3930
},
{
"epoch": 0.08264809531800638,
"grad_norm": 5.801643529836836,
"learning_rate": 2e-05,
"loss": 0.5939390182495117,
"step": 3940
},
{
"epoch": 0.08285786205739218,
"grad_norm": 1.1381215245542802,
"learning_rate": 2e-05,
"loss": 0.6113842964172364,
"step": 3950
},
{
"epoch": 0.08306762879677798,
"grad_norm": 1.0450250941065826,
"learning_rate": 2e-05,
"loss": 0.640468978881836,
"step": 3960
},
{
"epoch": 0.08327739553616378,
"grad_norm": 0.9218406597772941,
"learning_rate": 2e-05,
"loss": 0.5912976264953613,
"step": 3970
},
{
"epoch": 0.08348716227554959,
"grad_norm": 0.9837293096882525,
"learning_rate": 2e-05,
"loss": 0.6223003387451171,
"step": 3980
},
{
"epoch": 0.08369692901493539,
"grad_norm": 0.990643501319215,
"learning_rate": 2e-05,
"loss": 0.6180656433105469,
"step": 3990
},
{
"epoch": 0.08390669575432119,
"grad_norm": 0.9212243888272387,
"learning_rate": 2e-05,
"loss": 0.6109961986541748,
"step": 4000
},
{
"epoch": 0.084116462493707,
"grad_norm": 1.0851483978977774,
"learning_rate": 2e-05,
"loss": 0.6065635681152344,
"step": 4010
},
{
"epoch": 0.0843262292330928,
"grad_norm": 1.1576097032428736,
"learning_rate": 2e-05,
"loss": 0.5832537651062012,
"step": 4020
},
{
"epoch": 0.0845359959724786,
"grad_norm": 1.8856590779704239,
"learning_rate": 2e-05,
"loss": 0.6252013683319092,
"step": 4030
},
{
"epoch": 0.0847457627118644,
"grad_norm": 0.9200345861822318,
"learning_rate": 2e-05,
"loss": 0.5908494472503663,
"step": 4040
},
{
"epoch": 0.0849555294512502,
"grad_norm": 0.9332338359441046,
"learning_rate": 2e-05,
"loss": 0.5820940971374512,
"step": 4050
},
{
"epoch": 0.08516529619063601,
"grad_norm": 2.7597827875708423,
"learning_rate": 2e-05,
"loss": 0.6051089286804199,
"step": 4060
},
{
"epoch": 0.08537506293002181,
"grad_norm": 1.2095388396803786,
"learning_rate": 2e-05,
"loss": 0.6083034038543701,
"step": 4070
}
],
"logging_steps": 10,
"max_steps": 5935,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2035,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1927226987642880.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}