rqadri's picture
Upload folder using huggingface_hub
e89b23b verified
Raw History Blame Contribute Delete
49.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.2005736823234134,
"eval_steps": 500,
"global_step": 2797,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007171029042667623,
"grad_norm": 407.18868951296525,
"learning_rate": 1.1061797752808989e-06,
"loss": 1.767791748046875,
"step": 10
},
{
"epoch": 0.0014342058085335247,
"grad_norm": 567.8415841772963,
"learning_rate": 2.2241573033707866e-06,
"loss": 1.2149810791015625,
"step": 20
},
{
"epoch": 0.002151308712800287,
"grad_norm": 28.256766335357522,
"learning_rate": 3.342134831460675e-06,
"loss": 0.7131690979003906,
"step": 30
},
{
"epoch": 0.0028684116170670493,
"grad_norm": 192.44535950007864,
"learning_rate": 4.4601123595505624e-06,
"loss": 0.6729496002197266,
"step": 40
},
{
"epoch": 0.0035855145213338113,
"grad_norm": 63.480210447719905,
"learning_rate": 5.5780898876404495e-06,
"loss": 0.6091592311859131,
"step": 50
},
{
"epoch": 0.004302617425600574,
"grad_norm": 24.91719812284066,
"learning_rate": 6.6960674157303374e-06,
"loss": 0.5691664695739747,
"step": 60
},
{
"epoch": 0.005019720329867336,
"grad_norm": 512.6156292111328,
"learning_rate": 7.814044943820226e-06,
"loss": 0.5604145050048828,
"step": 70
},
{
"epoch": 0.005736823234134099,
"grad_norm": 52.14353148477021,
"learning_rate": 8.932022471910112e-06,
"loss": 0.5461579322814941,
"step": 80
},
{
"epoch": 0.006453926138400861,
"grad_norm": 203.45745330679426,
"learning_rate": 1.005e-05,
"loss": 0.564464521408081,
"step": 90
},
{
"epoch": 0.007171029042667623,
"grad_norm": 24.117935726545646,
"learning_rate": 1.1167977528089888e-05,
"loss": 0.6253236293792724,
"step": 100
},
{
"epoch": 0.007888131946934385,
"grad_norm": 78.69096586985131,
"learning_rate": 1.2285955056179776e-05,
"loss": 0.5682878494262695,
"step": 110
},
{
"epoch": 0.008605234851201148,
"grad_norm": 38.42539201854507,
"learning_rate": 1.3403932584269664e-05,
"loss": 0.6769058227539062,
"step": 120
},
{
"epoch": 0.00932233775546791,
"grad_norm": 53.488347150322625,
"learning_rate": 1.4521910112359552e-05,
"loss": 0.5980616569519043,
"step": 130
},
{
"epoch": 0.010039440659734672,
"grad_norm": 50.040307272011425,
"learning_rate": 1.563988764044944e-05,
"loss": 0.5930599212646485,
"step": 140
},
{
"epoch": 0.010756543564001434,
"grad_norm": 20.307563603035216,
"learning_rate": 1.6757865168539328e-05,
"loss": 0.6094542503356933,
"step": 150
},
{
"epoch": 0.011473646468268197,
"grad_norm": 11.737779939100745,
"learning_rate": 1.7875842696629214e-05,
"loss": 0.5738109588623047,
"step": 160
},
{
"epoch": 0.012190749372534958,
"grad_norm": 6.30430859385885,
"learning_rate": 1.89938202247191e-05,
"loss": 0.579668378829956,
"step": 170
},
{
"epoch": 0.012907852276801721,
"grad_norm": 5.06405076208195,
"learning_rate": 2e-05,
"loss": 0.5459352016448975,
"step": 180
},
{
"epoch": 0.013624955181068482,
"grad_norm": 4.192343622763153,
"learning_rate": 2e-05,
"loss": 0.5992312431335449,
"step": 190
},
{
"epoch": 0.014342058085335245,
"grad_norm": 10.448757223702469,
"learning_rate": 2e-05,
"loss": 0.5643451690673829,
"step": 200
},
{
"epoch": 0.015059160989602008,
"grad_norm": 21.869089910119737,
"learning_rate": 2e-05,
"loss": 0.5753413677215576,
"step": 210
},
{
"epoch": 0.01577626389386877,
"grad_norm": 3.778114364049049,
"learning_rate": 2e-05,
"loss": 0.5592538356781006,
"step": 220
},
{
"epoch": 0.016493366798135532,
"grad_norm": 8.878705688355062,
"learning_rate": 2e-05,
"loss": 0.5649869441986084,
"step": 230
},
{
"epoch": 0.017210469702402295,
"grad_norm": 39.24775031639978,
"learning_rate": 2e-05,
"loss": 0.6837917327880859,
"step": 240
},
{
"epoch": 0.017927572606669058,
"grad_norm": 3.9047301142219935,
"learning_rate": 2e-05,
"loss": 0.623165225982666,
"step": 250
},
{
"epoch": 0.01864467551093582,
"grad_norm": 6.1409826421758265,
"learning_rate": 2e-05,
"loss": 0.5558697700500488,
"step": 260
},
{
"epoch": 0.01936177841520258,
"grad_norm": 2.2935012027209156,
"learning_rate": 2e-05,
"loss": 0.5609950542449951,
"step": 270
},
{
"epoch": 0.020078881319469343,
"grad_norm": 14.443788852978345,
"learning_rate": 2e-05,
"loss": 0.595089340209961,
"step": 280
},
{
"epoch": 0.020795984223736106,
"grad_norm": 3.8329684129578863,
"learning_rate": 2e-05,
"loss": 0.5596456527709961,
"step": 290
},
{
"epoch": 0.02151308712800287,
"grad_norm": 1.3552305408083785,
"learning_rate": 2e-05,
"loss": 0.5825435638427734,
"step": 300
},
{
"epoch": 0.022230190032269632,
"grad_norm": 4.54925707452334,
"learning_rate": 2e-05,
"loss": 0.5733637809753418,
"step": 310
},
{
"epoch": 0.022947292936536395,
"grad_norm": 1.8200874030763512,
"learning_rate": 2e-05,
"loss": 0.5307866096496582,
"step": 320
},
{
"epoch": 0.023664395840803154,
"grad_norm": 2.459934996745395,
"learning_rate": 2e-05,
"loss": 0.5426007747650147,
"step": 330
},
{
"epoch": 0.024381498745069917,
"grad_norm": 1.6368595236282741,
"learning_rate": 2e-05,
"loss": 0.5503345489501953,
"step": 340
},
{
"epoch": 0.02509860164933668,
"grad_norm": 2.093526925659693,
"learning_rate": 2e-05,
"loss": 0.5389426231384278,
"step": 350
},
{
"epoch": 0.025815704553603443,
"grad_norm": 11.559747422199436,
"learning_rate": 2e-05,
"loss": 0.5869073867797852,
"step": 360
},
{
"epoch": 0.026532807457870206,
"grad_norm": 3.992070335805997,
"learning_rate": 2e-05,
"loss": 0.5316494464874267,
"step": 370
},
{
"epoch": 0.027249910362136965,
"grad_norm": 1.7412673131539824,
"learning_rate": 2e-05,
"loss": 0.5800913810729981,
"step": 380
},
{
"epoch": 0.027967013266403728,
"grad_norm": 1.691156350330339,
"learning_rate": 2e-05,
"loss": 0.5616199493408203,
"step": 390
},
{
"epoch": 0.02868411617067049,
"grad_norm": 2.2546580537446923,
"learning_rate": 2e-05,
"loss": 0.5290282726287842,
"step": 400
},
{
"epoch": 0.029401219074937254,
"grad_norm": 4.956937872768424,
"learning_rate": 2e-05,
"loss": 0.570636510848999,
"step": 410
},
{
"epoch": 0.030118321979204016,
"grad_norm": 4.969590728046793,
"learning_rate": 2e-05,
"loss": 0.549982213973999,
"step": 420
},
{
"epoch": 0.03083542488347078,
"grad_norm": 2.6043811684321203,
"learning_rate": 2e-05,
"loss": 0.5702108383178711,
"step": 430
},
{
"epoch": 0.03155252778773754,
"grad_norm": 2.1043413336167593,
"learning_rate": 2e-05,
"loss": 0.5530909538269043,
"step": 440
},
{
"epoch": 0.0322696306920043,
"grad_norm": 1.8388448759563596,
"learning_rate": 2e-05,
"loss": 0.5460524559020996,
"step": 450
},
{
"epoch": 0.032986733596271064,
"grad_norm": 3.6272533510034233,
"learning_rate": 2e-05,
"loss": 0.5648386955261231,
"step": 460
},
{
"epoch": 0.03370383650053783,
"grad_norm": 3.326768853485267,
"learning_rate": 2e-05,
"loss": 0.5850131034851074,
"step": 470
},
{
"epoch": 0.03442093940480459,
"grad_norm": 3.7884453189533573,
"learning_rate": 2e-05,
"loss": 0.5867877960205078,
"step": 480
},
{
"epoch": 0.03513804230907135,
"grad_norm": 2.8822182693669194,
"learning_rate": 2e-05,
"loss": 0.5637784481048584,
"step": 490
},
{
"epoch": 0.035855145213338116,
"grad_norm": 4.159255713401518,
"learning_rate": 2e-05,
"loss": 0.5402638435363769,
"step": 500
},
{
"epoch": 0.03657224811760488,
"grad_norm": 2.1094818708334646,
"learning_rate": 2e-05,
"loss": 0.5480118751525879,
"step": 510
},
{
"epoch": 0.03728935102187164,
"grad_norm": 2.375624628136267,
"learning_rate": 2e-05,
"loss": 0.5763840675354004,
"step": 520
},
{
"epoch": 0.0380064539261384,
"grad_norm": 3.2639853709166733,
"learning_rate": 2e-05,
"loss": 0.5657975196838378,
"step": 530
},
{
"epoch": 0.03872355683040516,
"grad_norm": 4.02194411766731,
"learning_rate": 2e-05,
"loss": 0.5532924652099609,
"step": 540
},
{
"epoch": 0.03944065973467192,
"grad_norm": 2.5235800313691112,
"learning_rate": 2e-05,
"loss": 0.5330066204071044,
"step": 550
},
{
"epoch": 0.040157762638938686,
"grad_norm": 1.7977002650246399,
"learning_rate": 2e-05,
"loss": 0.5558570861816406,
"step": 560
},
{
"epoch": 0.04087486554320545,
"grad_norm": 1.7670972508928158,
"learning_rate": 2e-05,
"loss": 0.5604548454284668,
"step": 570
},
{
"epoch": 0.04159196844747221,
"grad_norm": 1.7774681997349844,
"learning_rate": 2e-05,
"loss": 0.5438613891601562,
"step": 580
},
{
"epoch": 0.042309071351738975,
"grad_norm": 1.6346238912436575,
"learning_rate": 2e-05,
"loss": 0.5676720619201661,
"step": 590
},
{
"epoch": 0.04302617425600574,
"grad_norm": 1.6463255561065746,
"learning_rate": 2e-05,
"loss": 0.5760777950286865,
"step": 600
},
{
"epoch": 0.0437432771602725,
"grad_norm": 2.0910673930612855,
"learning_rate": 2e-05,
"loss": 0.5535629272460938,
"step": 610
},
{
"epoch": 0.044460380064539264,
"grad_norm": 4.51750941221636,
"learning_rate": 2e-05,
"loss": 0.561228084564209,
"step": 620
},
{
"epoch": 0.045177482968806026,
"grad_norm": 1.9378541587933182,
"learning_rate": 2e-05,
"loss": 0.5568342208862305,
"step": 630
},
{
"epoch": 0.04589458587307279,
"grad_norm": 2.8541213231232163,
"learning_rate": 2e-05,
"loss": 0.5806537628173828,
"step": 640
},
{
"epoch": 0.046611688777339545,
"grad_norm": 1.9777431081857408,
"learning_rate": 2e-05,
"loss": 0.5439037322998047,
"step": 650
},
{
"epoch": 0.04732879168160631,
"grad_norm": 3.5486839244500996,
"learning_rate": 2e-05,
"loss": 0.5219845294952392,
"step": 660
},
{
"epoch": 0.04804589458587307,
"grad_norm": 2.3095301026005415,
"learning_rate": 2e-05,
"loss": 0.5652892112731933,
"step": 670
},
{
"epoch": 0.048762997490139834,
"grad_norm": 1.345294939521322,
"learning_rate": 2e-05,
"loss": 0.5468602657318116,
"step": 680
},
{
"epoch": 0.0494801003944066,
"grad_norm": 1.1952203182785504,
"learning_rate": 2e-05,
"loss": 0.5759331703186035,
"step": 690
},
{
"epoch": 0.05019720329867336,
"grad_norm": 1.8539812501492536,
"learning_rate": 2e-05,
"loss": 0.5355586528778076,
"step": 700
},
{
"epoch": 0.05091430620294012,
"grad_norm": 6.172053481724467,
"learning_rate": 2e-05,
"loss": 0.5475461483001709,
"step": 710
},
{
"epoch": 0.051631409107206885,
"grad_norm": 1.1318198070091172,
"learning_rate": 2e-05,
"loss": 0.5526755332946778,
"step": 720
},
{
"epoch": 0.05234851201147365,
"grad_norm": 1.5054898615573356,
"learning_rate": 2e-05,
"loss": 0.5575639247894287,
"step": 730
},
{
"epoch": 0.05306561491574041,
"grad_norm": 1.8153166775343457,
"learning_rate": 2e-05,
"loss": 0.5713375091552735,
"step": 740
},
{
"epoch": 0.053782717820007174,
"grad_norm": 2.2659827396189467,
"learning_rate": 2e-05,
"loss": 0.5738441467285156,
"step": 750
},
{
"epoch": 0.05449982072427393,
"grad_norm": 2.252953864325539,
"learning_rate": 2e-05,
"loss": 0.5843099594116211,
"step": 760
},
{
"epoch": 0.05521692362854069,
"grad_norm": 2.2665119662895803,
"learning_rate": 2e-05,
"loss": 0.580151891708374,
"step": 770
},
{
"epoch": 0.055934026532807456,
"grad_norm": 1.585050610847029,
"learning_rate": 2e-05,
"loss": 0.5856458187103272,
"step": 780
},
{
"epoch": 0.05665112943707422,
"grad_norm": 1.2394601998691548,
"learning_rate": 2e-05,
"loss": 0.5632565975189209,
"step": 790
},
{
"epoch": 0.05736823234134098,
"grad_norm": 1.756235559775501,
"learning_rate": 2e-05,
"loss": 0.5605405330657959,
"step": 800
},
{
"epoch": 0.058085335245607744,
"grad_norm": 1.3423732665501396,
"learning_rate": 2e-05,
"loss": 0.5436521530151367,
"step": 810
},
{
"epoch": 0.05880243814987451,
"grad_norm": 9.423239856770886,
"learning_rate": 2e-05,
"loss": 0.5691229820251464,
"step": 820
},
{
"epoch": 0.05951954105414127,
"grad_norm": 1.1270115260912026,
"learning_rate": 2e-05,
"loss": 0.5732899665832519,
"step": 830
},
{
"epoch": 0.06023664395840803,
"grad_norm": 1.7447207447502826,
"learning_rate": 2e-05,
"loss": 0.5396980285644531,
"step": 840
},
{
"epoch": 0.060953746862674796,
"grad_norm": 6.67003895134874,
"learning_rate": 2e-05,
"loss": 0.5526090145111084,
"step": 850
},
{
"epoch": 0.06167084976694156,
"grad_norm": 1.550512453091403,
"learning_rate": 2e-05,
"loss": 0.5934078216552734,
"step": 860
},
{
"epoch": 0.06238795267120832,
"grad_norm": 1.4948941851618258,
"learning_rate": 2e-05,
"loss": 0.5663764953613282,
"step": 870
},
{
"epoch": 0.06310505557547508,
"grad_norm": 1.6244038346449619,
"learning_rate": 2e-05,
"loss": 0.5600580215454102,
"step": 880
},
{
"epoch": 0.06382215847974185,
"grad_norm": 2.112611569627416,
"learning_rate": 2e-05,
"loss": 0.5366658210754395,
"step": 890
},
{
"epoch": 0.0645392613840086,
"grad_norm": 6.753457683010765,
"learning_rate": 2e-05,
"loss": 0.5443595886230469,
"step": 900
},
{
"epoch": 0.06525636428827537,
"grad_norm": 10.430596168156878,
"learning_rate": 2e-05,
"loss": 0.5880915641784668,
"step": 910
},
{
"epoch": 0.06597346719254213,
"grad_norm": 5.695197824545816,
"learning_rate": 2e-05,
"loss": 0.5542924880981446,
"step": 920
},
{
"epoch": 0.0666905700968089,
"grad_norm": 39.208408511129576,
"learning_rate": 2e-05,
"loss": 0.5712676525115967,
"step": 930
},
{
"epoch": 0.06740767300107565,
"grad_norm": 1.2893484486598836,
"learning_rate": 2e-05,
"loss": 0.5429130077362061,
"step": 940
},
{
"epoch": 0.06812477590534241,
"grad_norm": 2.4226393977358027,
"learning_rate": 2e-05,
"loss": 0.5704053401947021,
"step": 950
},
{
"epoch": 0.06884187880960918,
"grad_norm": 1.5751855857626986,
"learning_rate": 2e-05,
"loss": 0.5415806770324707,
"step": 960
},
{
"epoch": 0.06955898171387594,
"grad_norm": 2.5631690778468936,
"learning_rate": 2e-05,
"loss": 0.557930850982666,
"step": 970
},
{
"epoch": 0.0702760846181427,
"grad_norm": 3.3081150567736466,
"learning_rate": 2e-05,
"loss": 0.568875503540039,
"step": 980
},
{
"epoch": 0.07099318752240946,
"grad_norm": 6.34001063563305,
"learning_rate": 2e-05,
"loss": 0.5582491874694824,
"step": 990
},
{
"epoch": 0.07171029042667623,
"grad_norm": 4.874954359164638,
"learning_rate": 2e-05,
"loss": 0.5891461849212647,
"step": 1000
},
{
"epoch": 0.07242739333094299,
"grad_norm": 2.845478581515535,
"learning_rate": 2e-05,
"loss": 0.5877192497253418,
"step": 1010
},
{
"epoch": 0.07314449623520976,
"grad_norm": 6.6832113770801245,
"learning_rate": 2e-05,
"loss": 0.5936095237731933,
"step": 1020
},
{
"epoch": 0.07386159913947651,
"grad_norm": 3.1831812810869176,
"learning_rate": 2e-05,
"loss": 0.5392140865325927,
"step": 1030
},
{
"epoch": 0.07457870204374328,
"grad_norm": 2.0917237719957327,
"learning_rate": 2e-05,
"loss": 0.5553333282470703,
"step": 1040
},
{
"epoch": 0.07529580494801004,
"grad_norm": 24.130295532697943,
"learning_rate": 2e-05,
"loss": 0.5648217678070069,
"step": 1050
},
{
"epoch": 0.0760129078522768,
"grad_norm": 16.060468283541823,
"learning_rate": 2e-05,
"loss": 0.5578883171081543,
"step": 1060
},
{
"epoch": 0.07673001075654357,
"grad_norm": 3.4059299289427973,
"learning_rate": 2e-05,
"loss": 0.5588356018066406,
"step": 1070
},
{
"epoch": 0.07744711366081032,
"grad_norm": 3.392575917695726,
"learning_rate": 2e-05,
"loss": 0.5971034049987793,
"step": 1080
},
{
"epoch": 0.07816421656507709,
"grad_norm": 2.75964264791242,
"learning_rate": 2e-05,
"loss": 0.5971877098083496,
"step": 1090
},
{
"epoch": 0.07888131946934385,
"grad_norm": 3.8086463493365152,
"learning_rate": 2e-05,
"loss": 0.5783935070037842,
"step": 1100
},
{
"epoch": 0.07959842237361062,
"grad_norm": 2.58367503325779,
"learning_rate": 2e-05,
"loss": 0.5899093151092529,
"step": 1110
},
{
"epoch": 0.08031552527787737,
"grad_norm": 1.1621693693121284,
"learning_rate": 2e-05,
"loss": 0.5835798263549805,
"step": 1120
},
{
"epoch": 0.08103262818214414,
"grad_norm": 1.508549335071549,
"learning_rate": 2e-05,
"loss": 0.5874055385589599,
"step": 1130
},
{
"epoch": 0.0817497310864109,
"grad_norm": 1.8012151576180655,
"learning_rate": 2e-05,
"loss": 0.5666553497314453,
"step": 1140
},
{
"epoch": 0.08246683399067767,
"grad_norm": 1.381814981155295,
"learning_rate": 2e-05,
"loss": 0.5351155281066895,
"step": 1150
},
{
"epoch": 0.08318393689494442,
"grad_norm": 3.3933223742777403,
"learning_rate": 2e-05,
"loss": 0.5747641563415528,
"step": 1160
},
{
"epoch": 0.08390103979921118,
"grad_norm": 3.4224544409368485,
"learning_rate": 2e-05,
"loss": 0.5581981658935546,
"step": 1170
},
{
"epoch": 0.08461814270347795,
"grad_norm": 1.978909956599825,
"learning_rate": 2e-05,
"loss": 0.5532891273498535,
"step": 1180
},
{
"epoch": 0.0853352456077447,
"grad_norm": 19.438157203071867,
"learning_rate": 2e-05,
"loss": 0.5519776821136475,
"step": 1190
},
{
"epoch": 0.08605234851201148,
"grad_norm": 2.5565977324257427,
"learning_rate": 2e-05,
"loss": 0.5592182159423829,
"step": 1200
},
{
"epoch": 0.08676945141627823,
"grad_norm": 1.8412057701311577,
"learning_rate": 2e-05,
"loss": 0.5500537872314453,
"step": 1210
},
{
"epoch": 0.087486554320545,
"grad_norm": 3.3351345984826577,
"learning_rate": 2e-05,
"loss": 0.5454347133636475,
"step": 1220
},
{
"epoch": 0.08820365722481176,
"grad_norm": 1.673409013117643,
"learning_rate": 2e-05,
"loss": 0.5704385757446289,
"step": 1230
},
{
"epoch": 0.08892076012907853,
"grad_norm": 2.6996798661736188,
"learning_rate": 2e-05,
"loss": 0.5756550788879394,
"step": 1240
},
{
"epoch": 0.08963786303334528,
"grad_norm": 1.5092839008364487,
"learning_rate": 2e-05,
"loss": 0.5831204891204834,
"step": 1250
},
{
"epoch": 0.09035496593761205,
"grad_norm": 1.8182439348881136,
"learning_rate": 2e-05,
"loss": 0.5437450408935547,
"step": 1260
},
{
"epoch": 0.09107206884187881,
"grad_norm": 1.9633401279250697,
"learning_rate": 2e-05,
"loss": 0.5590100765228272,
"step": 1270
},
{
"epoch": 0.09178917174614558,
"grad_norm": 2.0421680478109603,
"learning_rate": 2e-05,
"loss": 0.5691956996917724,
"step": 1280
},
{
"epoch": 0.09250627465041233,
"grad_norm": 4.552583885012217,
"learning_rate": 2e-05,
"loss": 0.5702752590179443,
"step": 1290
},
{
"epoch": 0.09322337755467909,
"grad_norm": 2.355805404071415,
"learning_rate": 2e-05,
"loss": 0.578900146484375,
"step": 1300
},
{
"epoch": 0.09394048045894586,
"grad_norm": 1.9752619410318872,
"learning_rate": 2e-05,
"loss": 0.5673203945159913,
"step": 1310
},
{
"epoch": 0.09465758336321262,
"grad_norm": 1.7290046269631103,
"learning_rate": 2e-05,
"loss": 0.5562186717987061,
"step": 1320
},
{
"epoch": 0.09537468626747939,
"grad_norm": 2.0888670764258968,
"learning_rate": 2e-05,
"loss": 0.5605551719665527,
"step": 1330
},
{
"epoch": 0.09609178917174614,
"grad_norm": 2.479093099558694,
"learning_rate": 2e-05,
"loss": 0.5779216766357422,
"step": 1340
},
{
"epoch": 0.09680889207601291,
"grad_norm": 23.340334088434542,
"learning_rate": 2e-05,
"loss": 0.5345910072326661,
"step": 1350
},
{
"epoch": 0.09752599498027967,
"grad_norm": 4.261898702807557,
"learning_rate": 2e-05,
"loss": 0.617152738571167,
"step": 1360
},
{
"epoch": 0.09824309788454644,
"grad_norm": 2.541660870456259,
"learning_rate": 2e-05,
"loss": 0.5389508724212646,
"step": 1370
},
{
"epoch": 0.0989602007888132,
"grad_norm": 2.185679172731964,
"learning_rate": 2e-05,
"loss": 0.5475359916687011,
"step": 1380
},
{
"epoch": 0.09967730369307996,
"grad_norm": 1.5807688223209044,
"learning_rate": 2e-05,
"loss": 0.545291519165039,
"step": 1390
},
{
"epoch": 0.10039440659734672,
"grad_norm": 1.421834241690961,
"learning_rate": 2e-05,
"loss": 0.5282414436340332,
"step": 1400
},
{
"epoch": 0.10111150950161348,
"grad_norm": 1.2162273515936992,
"learning_rate": 2e-05,
"loss": 0.5563668727874755,
"step": 1410
},
{
"epoch": 0.10182861240588024,
"grad_norm": 1.839759708750941,
"learning_rate": 2e-05,
"loss": 0.5481462001800537,
"step": 1420
},
{
"epoch": 0.102545715310147,
"grad_norm": 1.5675680272657555,
"learning_rate": 2e-05,
"loss": 0.5708026885986328,
"step": 1430
},
{
"epoch": 0.10326281821441377,
"grad_norm": 1.927651602988223,
"learning_rate": 2e-05,
"loss": 0.5257652282714844,
"step": 1440
},
{
"epoch": 0.10397992111868053,
"grad_norm": 3.350732435480856,
"learning_rate": 2e-05,
"loss": 0.5064706325531005,
"step": 1450
},
{
"epoch": 0.1046970240229473,
"grad_norm": 2.8354881463808526,
"learning_rate": 2e-05,
"loss": 0.5483434200286865,
"step": 1460
},
{
"epoch": 0.10541412692721405,
"grad_norm": 1.3180552517112605,
"learning_rate": 2e-05,
"loss": 0.5558376312255859,
"step": 1470
},
{
"epoch": 0.10613122983148082,
"grad_norm": 1.453787737702332,
"learning_rate": 2e-05,
"loss": 0.5345685958862305,
"step": 1480
},
{
"epoch": 0.10684833273574758,
"grad_norm": 1.6143142430339519,
"learning_rate": 2e-05,
"loss": 0.5461667060852051,
"step": 1490
},
{
"epoch": 0.10756543564001435,
"grad_norm": 1.0541879511974963,
"learning_rate": 2e-05,
"loss": 0.543915843963623,
"step": 1500
},
{
"epoch": 0.1082825385442811,
"grad_norm": 1.1840358132227575,
"learning_rate": 2e-05,
"loss": 0.5387839317321778,
"step": 1510
},
{
"epoch": 0.10899964144854786,
"grad_norm": 2.1425725141208876,
"learning_rate": 2e-05,
"loss": 0.5789884090423584,
"step": 1520
},
{
"epoch": 0.10971674435281463,
"grad_norm": 1.7058534011016113,
"learning_rate": 2e-05,
"loss": 0.5545422554016113,
"step": 1530
},
{
"epoch": 0.11043384725708139,
"grad_norm": 1.01371886276308,
"learning_rate": 2e-05,
"loss": 0.5442469596862793,
"step": 1540
},
{
"epoch": 0.11115095016134816,
"grad_norm": 2.3602814474145593,
"learning_rate": 2e-05,
"loss": 0.5491166591644288,
"step": 1550
},
{
"epoch": 0.11186805306561491,
"grad_norm": 2.808118905252683,
"learning_rate": 2e-05,
"loss": 0.5387770652770996,
"step": 1560
},
{
"epoch": 0.11258515596988168,
"grad_norm": 1.201202663255044,
"learning_rate": 2e-05,
"loss": 0.5430009841918946,
"step": 1570
},
{
"epoch": 0.11330225887414844,
"grad_norm": 4.700262983965676,
"learning_rate": 2e-05,
"loss": 0.5618073463439941,
"step": 1580
},
{
"epoch": 0.1140193617784152,
"grad_norm": 1.3533825932255663,
"learning_rate": 2e-05,
"loss": 0.5579128265380859,
"step": 1590
},
{
"epoch": 0.11473646468268196,
"grad_norm": 1.5857379391510027,
"learning_rate": 2e-05,
"loss": 0.5602166652679443,
"step": 1600
},
{
"epoch": 0.11545356758694873,
"grad_norm": 4.6484526042797185,
"learning_rate": 2e-05,
"loss": 0.5370266914367676,
"step": 1610
},
{
"epoch": 0.11617067049121549,
"grad_norm": 1.3186459053370545,
"learning_rate": 2e-05,
"loss": 0.587346076965332,
"step": 1620
},
{
"epoch": 0.11688777339548226,
"grad_norm": 2.3078508253099326,
"learning_rate": 2e-05,
"loss": 0.56480712890625,
"step": 1630
},
{
"epoch": 0.11760487629974901,
"grad_norm": 1.5069099846449054,
"learning_rate": 2e-05,
"loss": 0.5705337524414062,
"step": 1640
},
{
"epoch": 0.11832197920401577,
"grad_norm": 4.018520435411039,
"learning_rate": 2e-05,
"loss": 0.5952683925628662,
"step": 1650
},
{
"epoch": 0.11903908210828254,
"grad_norm": 1.9898833529179372,
"learning_rate": 2e-05,
"loss": 0.5610923767089844,
"step": 1660
},
{
"epoch": 0.1197561850125493,
"grad_norm": 2.6978818352707448,
"learning_rate": 2e-05,
"loss": 0.564918327331543,
"step": 1670
},
{
"epoch": 0.12047328791681607,
"grad_norm": 26.62718759167103,
"learning_rate": 2e-05,
"loss": 0.5717484474182128,
"step": 1680
},
{
"epoch": 0.12119039082108282,
"grad_norm": 6.128201486001608,
"learning_rate": 2e-05,
"loss": 0.5359064102172851,
"step": 1690
},
{
"epoch": 0.12190749372534959,
"grad_norm": 2.4136180588009926,
"learning_rate": 2e-05,
"loss": 0.5571248531341553,
"step": 1700
},
{
"epoch": 0.12262459662961635,
"grad_norm": 23.57181706006007,
"learning_rate": 2e-05,
"loss": 0.5414628505706787,
"step": 1710
},
{
"epoch": 0.12334169953388312,
"grad_norm": 2.061702966147612,
"learning_rate": 2e-05,
"loss": 0.5444112777709961,
"step": 1720
},
{
"epoch": 0.12405880243814987,
"grad_norm": 2.3085794839169136,
"learning_rate": 2e-05,
"loss": 0.576922607421875,
"step": 1730
},
{
"epoch": 0.12477590534241664,
"grad_norm": 5.710972200205366,
"learning_rate": 2e-05,
"loss": 0.5441991329193115,
"step": 1740
},
{
"epoch": 0.1254930082466834,
"grad_norm": 1.7864396003339693,
"learning_rate": 2e-05,
"loss": 0.5717742919921875,
"step": 1750
},
{
"epoch": 0.12621011115095015,
"grad_norm": 0.9014723079637288,
"learning_rate": 2e-05,
"loss": 0.5551410675048828,
"step": 1760
},
{
"epoch": 0.12692721405521692,
"grad_norm": 0.9370087293733383,
"learning_rate": 2e-05,
"loss": 0.5544151306152344,
"step": 1770
},
{
"epoch": 0.1276443169594837,
"grad_norm": 2.674839389932199,
"learning_rate": 2e-05,
"loss": 0.5430533409118652,
"step": 1780
},
{
"epoch": 0.12836141986375044,
"grad_norm": 2.273289919159861,
"learning_rate": 2e-05,
"loss": 0.5444197654724121,
"step": 1790
},
{
"epoch": 0.1290785227680172,
"grad_norm": 3.6023559077364498,
"learning_rate": 2e-05,
"loss": 0.5236852169036865,
"step": 1800
},
{
"epoch": 0.12979562567228398,
"grad_norm": 2.3065219860590096,
"learning_rate": 2e-05,
"loss": 0.5680782318115234,
"step": 1810
},
{
"epoch": 0.13051272857655075,
"grad_norm": 2.084022964738366,
"learning_rate": 2e-05,
"loss": 0.5856912136077881,
"step": 1820
},
{
"epoch": 0.1312298314808175,
"grad_norm": 2.1598047949045434,
"learning_rate": 2e-05,
"loss": 0.5665178298950195,
"step": 1830
},
{
"epoch": 0.13194693438508426,
"grad_norm": 3.4942841572353713,
"learning_rate": 2e-05,
"loss": 0.5104799270629883,
"step": 1840
},
{
"epoch": 0.13266403728935103,
"grad_norm": 2.6508913591970003,
"learning_rate": 2e-05,
"loss": 0.5309865951538086,
"step": 1850
},
{
"epoch": 0.1333811401936178,
"grad_norm": 2.3166121873122494,
"learning_rate": 2e-05,
"loss": 0.5639969825744628,
"step": 1860
},
{
"epoch": 0.13409824309788454,
"grad_norm": 0.9244294142195673,
"learning_rate": 2e-05,
"loss": 0.5573611259460449,
"step": 1870
},
{
"epoch": 0.1348153460021513,
"grad_norm": 1.727174236209999,
"learning_rate": 2e-05,
"loss": 0.5519887447357178,
"step": 1880
},
{
"epoch": 0.13553244890641808,
"grad_norm": 1.6186172320504204,
"learning_rate": 2e-05,
"loss": 0.5558316707611084,
"step": 1890
},
{
"epoch": 0.13624955181068482,
"grad_norm": 5.797337940362527,
"learning_rate": 2e-05,
"loss": 0.5600574493408204,
"step": 1900
},
{
"epoch": 0.1369666547149516,
"grad_norm": 1.4320979482604466,
"learning_rate": 2e-05,
"loss": 0.5349112987518311,
"step": 1910
},
{
"epoch": 0.13768375761921836,
"grad_norm": 2.128945502552639,
"learning_rate": 2e-05,
"loss": 0.5613338470458984,
"step": 1920
},
{
"epoch": 0.13840086052348513,
"grad_norm": 2.8191870400195307,
"learning_rate": 2e-05,
"loss": 0.5300923824310303,
"step": 1930
},
{
"epoch": 0.13911796342775187,
"grad_norm": 2.0673663166478216,
"learning_rate": 2e-05,
"loss": 0.538027286529541,
"step": 1940
},
{
"epoch": 0.13983506633201864,
"grad_norm": 1.8236578957860592,
"learning_rate": 2e-05,
"loss": 0.5343324184417725,
"step": 1950
},
{
"epoch": 0.1405521692362854,
"grad_norm": 3.3977884211025335,
"learning_rate": 2e-05,
"loss": 0.5443202972412109,
"step": 1960
},
{
"epoch": 0.14126927214055218,
"grad_norm": 3.8609988917549347,
"learning_rate": 2e-05,
"loss": 0.5392850399017334,
"step": 1970
},
{
"epoch": 0.14198637504481892,
"grad_norm": 1.437206263243472,
"learning_rate": 2e-05,
"loss": 0.5444744110107422,
"step": 1980
},
{
"epoch": 0.1427034779490857,
"grad_norm": 1.5343678694917298,
"learning_rate": 2e-05,
"loss": 0.5262813568115234,
"step": 1990
},
{
"epoch": 0.14342058085335246,
"grad_norm": 0.8756623603379825,
"learning_rate": 2e-05,
"loss": 0.5296908378601074,
"step": 2000
},
{
"epoch": 0.1441376837576192,
"grad_norm": 3.4802550609605976,
"learning_rate": 2e-05,
"loss": 0.5779629707336426,
"step": 2010
},
{
"epoch": 0.14485478666188598,
"grad_norm": 2.934219579458467,
"learning_rate": 2e-05,
"loss": 0.5296000957489013,
"step": 2020
},
{
"epoch": 0.14557188956615275,
"grad_norm": 1.3040073087327337,
"learning_rate": 2e-05,
"loss": 0.5569531440734863,
"step": 2030
},
{
"epoch": 0.14628899247041952,
"grad_norm": 4.113376666837119,
"learning_rate": 2e-05,
"loss": 0.5557994842529297,
"step": 2040
},
{
"epoch": 0.14700609537468626,
"grad_norm": 2.8352481929695244,
"learning_rate": 2e-05,
"loss": 0.5364577293395996,
"step": 2050
},
{
"epoch": 0.14772319827895303,
"grad_norm": 1.745221137869923,
"learning_rate": 2e-05,
"loss": 0.5374924659729003,
"step": 2060
},
{
"epoch": 0.1484403011832198,
"grad_norm": 1.9733965751347848,
"learning_rate": 2e-05,
"loss": 0.5620547294616699,
"step": 2070
},
{
"epoch": 0.14915740408748657,
"grad_norm": 2.5401198474019697,
"learning_rate": 2e-05,
"loss": 0.5819472789764404,
"step": 2080
},
{
"epoch": 0.1498745069917533,
"grad_norm": 1.6508987995216626,
"learning_rate": 2e-05,
"loss": 0.5497961044311523,
"step": 2090
},
{
"epoch": 0.15059160989602008,
"grad_norm": 4.028829550318998,
"learning_rate": 2e-05,
"loss": 0.5571159362792969,
"step": 2100
},
{
"epoch": 0.15130871280028685,
"grad_norm": 2.3975822216230247,
"learning_rate": 2e-05,
"loss": 0.5467790126800537,
"step": 2110
},
{
"epoch": 0.1520258157045536,
"grad_norm": 1.4480780443079728,
"learning_rate": 2e-05,
"loss": 0.5441234588623047,
"step": 2120
},
{
"epoch": 0.15274291860882036,
"grad_norm": 1.4790891678648428,
"learning_rate": 2e-05,
"loss": 0.5478582382202148,
"step": 2130
},
{
"epoch": 0.15346002151308713,
"grad_norm": 1.2546481239787297,
"learning_rate": 2e-05,
"loss": 0.5563353538513184,
"step": 2140
},
{
"epoch": 0.1541771244173539,
"grad_norm": 0.9819256656146061,
"learning_rate": 2e-05,
"loss": 0.5564832210540771,
"step": 2150
},
{
"epoch": 0.15489422732162064,
"grad_norm": 2.566754230339073,
"learning_rate": 2e-05,
"loss": 0.5165680408477783,
"step": 2160
},
{
"epoch": 0.1556113302258874,
"grad_norm": 0.8311248159633984,
"learning_rate": 2e-05,
"loss": 0.5284764289855957,
"step": 2170
},
{
"epoch": 0.15632843313015418,
"grad_norm": 1.8410382032231034,
"learning_rate": 2e-05,
"loss": 0.5319634437561035,
"step": 2180
},
{
"epoch": 0.15704553603442095,
"grad_norm": 0.9088140586053679,
"learning_rate": 2e-05,
"loss": 0.542931318283081,
"step": 2190
},
{
"epoch": 0.1577626389386877,
"grad_norm": 1.1599160205724215,
"learning_rate": 2e-05,
"loss": 0.552515697479248,
"step": 2200
},
{
"epoch": 0.15847974184295446,
"grad_norm": 0.8853220986303392,
"learning_rate": 2e-05,
"loss": 0.5501457214355469,
"step": 2210
},
{
"epoch": 0.15919684474722123,
"grad_norm": 2.0740258891068835,
"learning_rate": 2e-05,
"loss": 0.543214225769043,
"step": 2220
},
{
"epoch": 0.15991394765148798,
"grad_norm": 3.638417995818362,
"learning_rate": 2e-05,
"loss": 0.5436822891235351,
"step": 2230
},
{
"epoch": 0.16063105055575475,
"grad_norm": 0.8144603584145417,
"learning_rate": 2e-05,
"loss": 0.5217988014221191,
"step": 2240
},
{
"epoch": 0.16134815346002152,
"grad_norm": 1.0452638192846837,
"learning_rate": 2e-05,
"loss": 0.5355788230895996,
"step": 2250
},
{
"epoch": 0.16206525636428828,
"grad_norm": 1.3003265396721033,
"learning_rate": 2e-05,
"loss": 0.5429640769958496,
"step": 2260
},
{
"epoch": 0.16278235926855503,
"grad_norm": 1.2224235191550799,
"learning_rate": 2e-05,
"loss": 0.5476203441619873,
"step": 2270
},
{
"epoch": 0.1634994621728218,
"grad_norm": 0.9447309520879374,
"learning_rate": 2e-05,
"loss": 0.5285428524017334,
"step": 2280
},
{
"epoch": 0.16421656507708857,
"grad_norm": 0.976975105833434,
"learning_rate": 2e-05,
"loss": 0.5368287563323975,
"step": 2290
},
{
"epoch": 0.16493366798135534,
"grad_norm": 1.7020593967778337,
"learning_rate": 2e-05,
"loss": 0.5516303539276123,
"step": 2300
},
{
"epoch": 0.16565077088562208,
"grad_norm": 1.1134720838830279,
"learning_rate": 2e-05,
"loss": 0.5425692558288574,
"step": 2310
},
{
"epoch": 0.16636787378988885,
"grad_norm": 1.5392906735570862,
"learning_rate": 2e-05,
"loss": 0.5487800121307373,
"step": 2320
},
{
"epoch": 0.16708497669415562,
"grad_norm": 1.266543281789761,
"learning_rate": 2e-05,
"loss": 0.5495414733886719,
"step": 2330
},
{
"epoch": 0.16780207959842236,
"grad_norm": 0.8428229926488787,
"learning_rate": 2e-05,
"loss": 0.5339969635009766,
"step": 2340
},
{
"epoch": 0.16851918250268913,
"grad_norm": 1.248525897850751,
"learning_rate": 2e-05,
"loss": 0.5505522727966309,
"step": 2350
},
{
"epoch": 0.1692362854069559,
"grad_norm": 0.9057046779507363,
"learning_rate": 2e-05,
"loss": 0.5328867912292481,
"step": 2360
},
{
"epoch": 0.16995338831122267,
"grad_norm": 2.3723677611050062,
"learning_rate": 2e-05,
"loss": 0.5437640190124512,
"step": 2370
},
{
"epoch": 0.1706704912154894,
"grad_norm": 2.042150750776014,
"learning_rate": 2e-05,
"loss": 0.5396258354187011,
"step": 2380
},
{
"epoch": 0.17138759411975618,
"grad_norm": 3.838167099539153,
"learning_rate": 2e-05,
"loss": 0.5363749980926513,
"step": 2390
},
{
"epoch": 0.17210469702402295,
"grad_norm": 6.1848557321177795,
"learning_rate": 2e-05,
"loss": 0.5548676967620849,
"step": 2400
},
{
"epoch": 0.17282179992828972,
"grad_norm": 6.579070478414869,
"learning_rate": 2e-05,
"loss": 0.5562749385833741,
"step": 2410
},
{
"epoch": 0.17353890283255646,
"grad_norm": 5.185362047776957,
"learning_rate": 2e-05,
"loss": 0.564643383026123,
"step": 2420
},
{
"epoch": 0.17425600573682323,
"grad_norm": 5.910383156548448,
"learning_rate": 2e-05,
"loss": 0.5481171607971191,
"step": 2430
},
{
"epoch": 0.17497310864109,
"grad_norm": 1.763722226804498,
"learning_rate": 2e-05,
"loss": 0.5655164241790771,
"step": 2440
},
{
"epoch": 0.17569021154535677,
"grad_norm": 3.376109338691145,
"learning_rate": 2e-05,
"loss": 0.5555837154388428,
"step": 2450
},
{
"epoch": 0.17640731444962351,
"grad_norm": 0.9568745205231425,
"learning_rate": 2e-05,
"loss": 0.5491615295410156,
"step": 2460
},
{
"epoch": 0.17712441735389028,
"grad_norm": 16.059766905754536,
"learning_rate": 2e-05,
"loss": 0.5338088512420655,
"step": 2470
},
{
"epoch": 0.17784152025815705,
"grad_norm": 2.975321349891421,
"learning_rate": 2e-05,
"loss": 0.5540153503417968,
"step": 2480
},
{
"epoch": 0.1785586231624238,
"grad_norm": 6.47424146966675,
"learning_rate": 2e-05,
"loss": 0.5426904678344726,
"step": 2490
},
{
"epoch": 0.17927572606669057,
"grad_norm": 1.3423883094161901,
"learning_rate": 2e-05,
"loss": 0.5475606441497802,
"step": 2500
},
{
"epoch": 0.17999282897095734,
"grad_norm": 2.4973587366137995,
"learning_rate": 2e-05,
"loss": 0.5557956218719482,
"step": 2510
},
{
"epoch": 0.1807099318752241,
"grad_norm": 1.0026541588256375,
"learning_rate": 2e-05,
"loss": 0.5389398574829102,
"step": 2520
},
{
"epoch": 0.18142703477949085,
"grad_norm": 0.7595318143310303,
"learning_rate": 2e-05,
"loss": 0.5713970184326171,
"step": 2530
},
{
"epoch": 0.18214413768375762,
"grad_norm": 1.016156362531317,
"learning_rate": 2e-05,
"loss": 0.5476889610290527,
"step": 2540
},
{
"epoch": 0.1828612405880244,
"grad_norm": 1.1104838610457881,
"learning_rate": 1.997261400182901e-05,
"loss": 0.5419221605573382,
"step": 2550
},
{
"epoch": 0.18357834349229116,
"grad_norm": 1.0948314978981517,
"learning_rate": 1.9805800746810423e-05,
"loss": 0.5443652629852295,
"step": 2560
},
{
"epoch": 0.1842954463965579,
"grad_norm": 0.942909129642652,
"learning_rate": 1.9489934220110274e-05,
"loss": 0.5187523365020752,
"step": 2570
},
{
"epoch": 0.18501254930082467,
"grad_norm": 2.498357530841941,
"learning_rate": 1.9029840364820392e-05,
"loss": 0.54703688621521,
"step": 2580
},
{
"epoch": 0.18572965220509144,
"grad_norm": 1.050915878037599,
"learning_rate": 1.8432548690479908e-05,
"loss": 0.5294360160827637,
"step": 2590
},
{
"epoch": 0.18644675510935818,
"grad_norm": 1.9942088565115943,
"learning_rate": 1.7707184873239772e-05,
"loss": 0.5605529308319092,
"step": 2600
},
{
"epoch": 0.18716385801362495,
"grad_norm": 0.961342633630974,
"learning_rate": 1.6864831329902565e-05,
"loss": 0.5377580642700195,
"step": 2610
},
{
"epoch": 0.18788096091789172,
"grad_norm": 2.876057206556651,
"learning_rate": 1.591835789604667e-05,
"loss": 0.5342798233032227,
"step": 2620
},
{
"epoch": 0.1885980638221585,
"grad_norm": 1.0021327204981894,
"learning_rate": 1.4882225195206459e-05,
"loss": 0.5192233085632324,
"step": 2630
},
{
"epoch": 0.18931516672642523,
"grad_norm": 3.1066883574587174,
"learning_rate": 1.3772263703317695e-05,
"loss": 0.5124124526977539,
"step": 2640
},
{
"epoch": 0.190032269630692,
"grad_norm": 1.6890114478162948,
"learning_rate": 1.2605431883975547e-05,
"loss": 0.5044251441955566,
"step": 2650
},
{
"epoch": 0.19074937253495877,
"grad_norm": 6.522787376641639,
"learning_rate": 1.1399557089817574e-05,
"loss": 0.5150196552276611,
"step": 2660
},
{
"epoch": 0.19146647543922554,
"grad_norm": 1.4824535355197326,
"learning_rate": 1.0173063188650463e-05,
"loss": 0.5414235115051269,
"step": 2670
},
{
"epoch": 0.19218357834349228,
"grad_norm": 0.8690612230958721,
"learning_rate": 8.944689075764396e-06,
"loss": 0.5502989768981934,
"step": 2680
},
{
"epoch": 0.19290068124775905,
"grad_norm": 3.7840717869359426,
"learning_rate": 7.73320237312362e-06,
"loss": 0.5113142967224121,
"step": 2690
},
{
"epoch": 0.19361778415202582,
"grad_norm": 2.3025950044090995,
"learning_rate": 6.557112689658629e-06,
"loss": 0.5296006202697754,
"step": 2700
},
{
"epoch": 0.19433488705629257,
"grad_norm": 1.0694654085420516,
"learning_rate": 5.434388823591213e-06,
"loss": 0.5536742210388184,
"step": 2710
},
{
"epoch": 0.19505198996055934,
"grad_norm": 1.1124214182448702,
"learning_rate": 4.382184227495424e-06,
"loss": 0.5129819869995117,
"step": 2720
},
{
"epoch": 0.1957690928648261,
"grad_norm": 0.9916864960655524,
"learning_rate": 3.4165749305560203e-06,
"loss": 0.498732328414917,
"step": 2730
},
{
"epoch": 0.19648619576909288,
"grad_norm": 1.6432467789542329,
"learning_rate": 2.5523139221595556e-06,
"loss": 0.5243377208709716,
"step": 2740
},
{
"epoch": 0.19720329867335962,
"grad_norm": 1.6923412472424093,
"learning_rate": 1.8026057494499821e-06,
"loss": 0.531996488571167,
"step": 2750
},
{
"epoch": 0.1979204015776264,
"grad_norm": 1.8684286091471634,
"learning_rate": 1.178904772643223e-06,
"loss": 0.5293401718139649,
"step": 2760
},
{
"epoch": 0.19863750448189316,
"grad_norm": 1.3510552879003406,
"learning_rate": 6.90740160441754e-07,
"loss": 0.5419890403747558,
"step": 2770
},
{
"epoch": 0.19935460738615993,
"grad_norm": 6.011413037566841,
"learning_rate": 3.45570299343632e-07,
"loss": 0.5053925037384033,
"step": 2780
},
{
"epoch": 0.20007171029042667,
"grad_norm": 1.014709725454001,
"learning_rate": 1.486688412425152e-07,
"loss": 0.5323020935058593,
"step": 2790
},
{
"epoch": 0.2005736823234134,
"step": 2797,
"total_flos": 1324435843842048.0,
"train_loss": 0.04788554878289417,
"train_runtime": 3648.7564,
"train_samples_per_second": 9.199,
"train_steps_per_second": 0.767
}
],
"logging_steps": 10,
"max_steps": 2797,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2543,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1324435843842048.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}