{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4013624955181068,
  "eval_steps": 500,
  "global_step": 5597,
  "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": 0.7970165091830483,
      "learning_rate": 2e-05,
      "loss": 0.5419019971575055,
      "step": 2550
    },
    {
      "epoch": 0.18357834349229116,
      "grad_norm": 1.212872451737735,
      "learning_rate": 2e-05,
      "loss": 0.5441554546356201,
      "step": 2560
    },
    {
      "epoch": 0.1842954463965579,
      "grad_norm": 0.8005626868556313,
      "learning_rate": 2e-05,
      "loss": 0.5173744201660156,
      "step": 2570
    },
    {
      "epoch": 0.18501254930082467,
      "grad_norm": 2.9379310099657925,
      "learning_rate": 2e-05,
      "loss": 0.5431180953979492,
      "step": 2580
    },
    {
      "epoch": 0.18572965220509144,
      "grad_norm": 0.949445723217109,
      "learning_rate": 2e-05,
      "loss": 0.527243185043335,
      "step": 2590
    },
    {
      "epoch": 0.18644675510935818,
      "grad_norm": 3.7147371453843046,
      "learning_rate": 2e-05,
      "loss": 0.5601873397827148,
      "step": 2600
    },
    {
      "epoch": 0.18716385801362495,
      "grad_norm": 1.915538807542611,
      "learning_rate": 2e-05,
      "loss": 0.5382901668548584,
      "step": 2610
    },
    {
      "epoch": 0.18788096091789172,
      "grad_norm": 2.6077714685736306,
      "learning_rate": 2e-05,
      "loss": 0.5370364665985108,
      "step": 2620
    },
    {
      "epoch": 0.1885980638221585,
      "grad_norm": 3.530042179716205,
      "learning_rate": 2e-05,
      "loss": 0.5232605457305908,
      "step": 2630
    },
    {
      "epoch": 0.18931516672642523,
      "grad_norm": 1.065202023541238,
      "learning_rate": 2e-05,
      "loss": 0.5182710647583008,
      "step": 2640
    },
    {
      "epoch": 0.190032269630692,
      "grad_norm": 2.813129356499845,
      "learning_rate": 2e-05,
      "loss": 0.5102169036865234,
      "step": 2650
    },
    {
      "epoch": 0.19074937253495877,
      "grad_norm": 2.2728120863920447,
      "learning_rate": 2e-05,
      "loss": 0.5203429698944092,
      "step": 2660
    },
    {
      "epoch": 0.19146647543922554,
      "grad_norm": 3.825459217860836,
      "learning_rate": 2e-05,
      "loss": 0.5492259502410889,
      "step": 2670
    },
    {
      "epoch": 0.19218357834349228,
      "grad_norm": 22.950291112353018,
      "learning_rate": 2e-05,
      "loss": 0.5592641353607177,
      "step": 2680
    },
    {
      "epoch": 0.19290068124775905,
      "grad_norm": 1.6610535958144417,
      "learning_rate": 2e-05,
      "loss": 0.5212313652038574,
      "step": 2690
    },
    {
      "epoch": 0.19361778415202582,
      "grad_norm": 1.3943686694644843,
      "learning_rate": 2e-05,
      "loss": 0.540802001953125,
      "step": 2700
    },
    {
      "epoch": 0.19433488705629257,
      "grad_norm": 1.3589842470045335,
      "learning_rate": 2e-05,
      "loss": 0.5652813911437988,
      "step": 2710
    },
    {
      "epoch": 0.19505198996055934,
      "grad_norm": 0.8826561412653801,
      "learning_rate": 2e-05,
      "loss": 0.5242929458618164,
      "step": 2720
    },
    {
      "epoch": 0.1957690928648261,
      "grad_norm": 0.9315280057335483,
      "learning_rate": 2e-05,
      "loss": 0.5131863117218017,
      "step": 2730
    },
    {
      "epoch": 0.19648619576909288,
      "grad_norm": 1.0886980020319452,
      "learning_rate": 2e-05,
      "loss": 0.5389771461486816,
      "step": 2740
    },
    {
      "epoch": 0.19720329867335962,
      "grad_norm": 1.3318612312597833,
      "learning_rate": 2e-05,
      "loss": 0.5457143306732177,
      "step": 2750
    },
    {
      "epoch": 0.1979204015776264,
      "grad_norm": 0.7786238367400904,
      "learning_rate": 2e-05,
      "loss": 0.5433188438415527,
      "step": 2760
    },
    {
      "epoch": 0.19863750448189316,
      "grad_norm": 1.0604607878751526,
      "learning_rate": 2e-05,
      "loss": 0.5557928085327148,
      "step": 2770
    },
    {
      "epoch": 0.19935460738615993,
      "grad_norm": 1.217409849968991,
      "learning_rate": 2e-05,
      "loss": 0.5180927276611328,
      "step": 2780
    },
    {
      "epoch": 0.20007171029042667,
      "grad_norm": 1.6265598412259648,
      "learning_rate": 2e-05,
      "loss": 0.5455813407897949,
      "step": 2790
    },
    {
      "epoch": 0.20078881319469344,
      "grad_norm": 9.857717007267182,
      "learning_rate": 2e-05,
      "loss": 0.5268736839294433,
      "step": 2800
    },
    {
      "epoch": 0.2015059160989602,
      "grad_norm": 1.164019528477356,
      "learning_rate": 2e-05,
      "loss": 0.5568728923797608,
      "step": 2810
    },
    {
      "epoch": 0.20222301900322695,
      "grad_norm": 1.0961229710968308,
      "learning_rate": 2e-05,
      "loss": 0.5290776252746582,
      "step": 2820
    },
    {
      "epoch": 0.20294012190749372,
      "grad_norm": 2.0912219642452583,
      "learning_rate": 2e-05,
      "loss": 0.5580238342285156,
      "step": 2830
    },
    {
      "epoch": 0.2036572248117605,
      "grad_norm": 4.873018987508456,
      "learning_rate": 2e-05,
      "loss": 0.5235507965087891,
      "step": 2840
    },
    {
      "epoch": 0.20437432771602726,
      "grad_norm": 0.9957804107946927,
      "learning_rate": 2e-05,
      "loss": 0.5344703674316407,
      "step": 2850
    },
    {
      "epoch": 0.205091430620294,
      "grad_norm": 1.097152646320635,
      "learning_rate": 2e-05,
      "loss": 0.5440820217132568,
      "step": 2860
    },
    {
      "epoch": 0.20580853352456077,
      "grad_norm": 1.402455427320016,
      "learning_rate": 2e-05,
      "loss": 0.5374917030334473,
      "step": 2870
    },
    {
      "epoch": 0.20652563642882754,
      "grad_norm": 1.3061484721695487,
      "learning_rate": 2e-05,
      "loss": 0.559415626525879,
      "step": 2880
    },
    {
      "epoch": 0.2072427393330943,
      "grad_norm": 1.757205153769175,
      "learning_rate": 2e-05,
      "loss": 0.5418906688690186,
      "step": 2890
    },
    {
      "epoch": 0.20795984223736105,
      "grad_norm": 1.8864659503779855,
      "learning_rate": 2e-05,
      "loss": 0.5330214977264405,
      "step": 2900
    },
    {
      "epoch": 0.20867694514162782,
      "grad_norm": 1.3378572555996144,
      "learning_rate": 2e-05,
      "loss": 0.5348206520080566,
      "step": 2910
    },
    {
      "epoch": 0.2093940480458946,
      "grad_norm": 17.085726007279217,
      "learning_rate": 2e-05,
      "loss": 0.5154592990875244,
      "step": 2920
    },
    {
      "epoch": 0.21011115095016134,
      "grad_norm": 1.3195399295288364,
      "learning_rate": 2e-05,
      "loss": 0.5361568450927734,
      "step": 2930
    },
    {
      "epoch": 0.2108282538544281,
      "grad_norm": 1.110272289945594,
      "learning_rate": 2e-05,
      "loss": 0.5523487091064453,
      "step": 2940
    },
    {
      "epoch": 0.21154535675869487,
      "grad_norm": 0.8720572510769875,
      "learning_rate": 2e-05,
      "loss": 0.5503731727600097,
      "step": 2950
    },
    {
      "epoch": 0.21226245966296164,
      "grad_norm": 0.8775629662923128,
      "learning_rate": 2e-05,
      "loss": 0.52593994140625,
      "step": 2960
    },
    {
      "epoch": 0.2129795625672284,
      "grad_norm": 1.4833496960689059,
      "learning_rate": 2e-05,
      "loss": 0.5133781909942627,
      "step": 2970
    },
    {
      "epoch": 0.21369666547149516,
      "grad_norm": 1.2914369801636472,
      "learning_rate": 2e-05,
      "loss": 0.5397880554199219,
      "step": 2980
    },
    {
      "epoch": 0.21441376837576193,
      "grad_norm": 0.9292086289799888,
      "learning_rate": 2e-05,
      "loss": 0.5184877872467041,
      "step": 2990
    },
    {
      "epoch": 0.2151308712800287,
      "grad_norm": 0.9986029194493474,
      "learning_rate": 2e-05,
      "loss": 0.535045051574707,
      "step": 3000
    },
    {
      "epoch": 0.21584797418429544,
      "grad_norm": 1.0052305722500934,
      "learning_rate": 2e-05,
      "loss": 0.5459346771240234,
      "step": 3010
    },
    {
      "epoch": 0.2165650770885622,
      "grad_norm": 0.9306599340606266,
      "learning_rate": 2e-05,
      "loss": 0.5160114288330078,
      "step": 3020
    },
    {
      "epoch": 0.21728217999282898,
      "grad_norm": 0.795239365095,
      "learning_rate": 2e-05,
      "loss": 0.5201301097869873,
      "step": 3030
    },
    {
      "epoch": 0.21799928289709572,
      "grad_norm": 2.5853271907958666,
      "learning_rate": 2e-05,
      "loss": 0.5145270347595214,
      "step": 3040
    },
    {
      "epoch": 0.2187163858013625,
      "grad_norm": 1.8897441923162515,
      "learning_rate": 2e-05,
      "loss": 0.5408259391784668,
      "step": 3050
    },
    {
      "epoch": 0.21943348870562926,
      "grad_norm": 1.3454937351660057,
      "learning_rate": 2e-05,
      "loss": 0.525616455078125,
      "step": 3060
    },
    {
      "epoch": 0.22015059160989603,
      "grad_norm": 1.4698501069950678,
      "learning_rate": 2e-05,
      "loss": 0.5587374687194824,
      "step": 3070
    },
    {
      "epoch": 0.22086769451416277,
      "grad_norm": 1.2731317005401546,
      "learning_rate": 2e-05,
      "loss": 0.5245851516723633,
      "step": 3080
    },
    {
      "epoch": 0.22158479741842954,
      "grad_norm": 1.157866180564051,
      "learning_rate": 2e-05,
      "loss": 0.5394145965576171,
      "step": 3090
    },
    {
      "epoch": 0.2223019003226963,
      "grad_norm": 1.2826957631680433,
      "learning_rate": 2e-05,
      "loss": 0.5228326797485352,
      "step": 3100
    },
    {
      "epoch": 0.22301900322696308,
      "grad_norm": 1.3341564827255854,
      "learning_rate": 2e-05,
      "loss": 0.525169563293457,
      "step": 3110
    },
    {
      "epoch": 0.22373610613122982,
      "grad_norm": 0.9939192465447331,
      "learning_rate": 2e-05,
      "loss": 0.5138655185699463,
      "step": 3120
    },
    {
      "epoch": 0.2244532090354966,
      "grad_norm": 1.0447934785155872,
      "learning_rate": 2e-05,
      "loss": 0.5518413066864014,
      "step": 3130
    },
    {
      "epoch": 0.22517031193976336,
      "grad_norm": 1.0570053350373516,
      "learning_rate": 2e-05,
      "loss": 0.5387558937072754,
      "step": 3140
    },
    {
      "epoch": 0.22588741484403013,
      "grad_norm": 1.5372579580078292,
      "learning_rate": 2e-05,
      "loss": 0.5300348281860352,
      "step": 3150
    },
    {
      "epoch": 0.22660451774829687,
      "grad_norm": 1.6943786493138226,
      "learning_rate": 2e-05,
      "loss": 0.5494170188903809,
      "step": 3160
    },
    {
      "epoch": 0.22732162065256364,
      "grad_norm": 4.286841026732124,
      "learning_rate": 2e-05,
      "loss": 0.5373885154724121,
      "step": 3170
    },
    {
      "epoch": 0.2280387235568304,
      "grad_norm": 1.1383322747986921,
      "learning_rate": 2e-05,
      "loss": 0.5370725154876709,
      "step": 3180
    },
    {
      "epoch": 0.22875582646109716,
      "grad_norm": 0.8159529616287994,
      "learning_rate": 2e-05,
      "loss": 0.5347590446472168,
      "step": 3190
    },
    {
      "epoch": 0.22947292936536393,
      "grad_norm": 2.7114842944149466,
      "learning_rate": 2e-05,
      "loss": 0.5403803825378418,
      "step": 3200
    },
    {
      "epoch": 0.2301900322696307,
      "grad_norm": 0.8655339392013862,
      "learning_rate": 2e-05,
      "loss": 0.5468146324157714,
      "step": 3210
    },
    {
      "epoch": 0.23090713517389747,
      "grad_norm": 0.9020065519506806,
      "learning_rate": 2e-05,
      "loss": 0.5497066497802734,
      "step": 3220
    },
    {
      "epoch": 0.2316242380781642,
      "grad_norm": 1.5889125547725025,
      "learning_rate": 2e-05,
      "loss": 0.5209928512573242,
      "step": 3230
    },
    {
      "epoch": 0.23234134098243098,
      "grad_norm": 1.5067058275169936,
      "learning_rate": 2e-05,
      "loss": 0.5649847984313965,
      "step": 3240
    },
    {
      "epoch": 0.23305844388669775,
      "grad_norm": 1.222359656954987,
      "learning_rate": 2e-05,
      "loss": 0.5382471084594727,
      "step": 3250
    },
    {
      "epoch": 0.23377554679096452,
      "grad_norm": 0.8189745553358863,
      "learning_rate": 2e-05,
      "loss": 0.5513941764831543,
      "step": 3260
    },
    {
      "epoch": 0.23449264969523126,
      "grad_norm": 1.8934644678944872,
      "learning_rate": 2e-05,
      "loss": 0.5290876865386963,
      "step": 3270
    },
    {
      "epoch": 0.23520975259949803,
      "grad_norm": 1.3550867245782727,
      "learning_rate": 2e-05,
      "loss": 0.5335357666015625,
      "step": 3280
    },
    {
      "epoch": 0.2359268555037648,
      "grad_norm": 1.841390318543767,
      "learning_rate": 2e-05,
      "loss": 0.5329647541046143,
      "step": 3290
    },
    {
      "epoch": 0.23664395840803154,
      "grad_norm": 1.0726025261274934,
      "learning_rate": 2e-05,
      "loss": 0.4872093200683594,
      "step": 3300
    },
    {
      "epoch": 0.2373610613122983,
      "grad_norm": 1.1264303467860137,
      "learning_rate": 2e-05,
      "loss": 0.5237213611602783,
      "step": 3310
    },
    {
      "epoch": 0.23807816421656508,
      "grad_norm": 0.9459372638697248,
      "learning_rate": 2e-05,
      "loss": 0.5413484096527099,
      "step": 3320
    },
    {
      "epoch": 0.23879526712083185,
      "grad_norm": 0.8668638766433358,
      "learning_rate": 2e-05,
      "loss": 0.5182044506072998,
      "step": 3330
    },
    {
      "epoch": 0.2395123700250986,
      "grad_norm": 4.667712875971681,
      "learning_rate": 2e-05,
      "loss": 0.5280697822570801,
      "step": 3340
    },
    {
      "epoch": 0.24022947292936536,
      "grad_norm": 2.241383387057955,
      "learning_rate": 2e-05,
      "loss": 0.5417804718017578,
      "step": 3350
    },
    {
      "epoch": 0.24094657583363213,
      "grad_norm": 0.9192454513461864,
      "learning_rate": 2e-05,
      "loss": 0.5354562282562256,
      "step": 3360
    },
    {
      "epoch": 0.2416636787378989,
      "grad_norm": 1.7669829080992607,
      "learning_rate": 2e-05,
      "loss": 0.5265363216400146,
      "step": 3370
    },
    {
      "epoch": 0.24238078164216564,
      "grad_norm": 1.1498872599122203,
      "learning_rate": 2e-05,
      "loss": 0.5269899845123291,
      "step": 3380
    },
    {
      "epoch": 0.2430978845464324,
      "grad_norm": 1.221951222563851,
      "learning_rate": 2e-05,
      "loss": 0.5453744411468506,
      "step": 3390
    },
    {
      "epoch": 0.24381498745069918,
      "grad_norm": 1.5269409721956062,
      "learning_rate": 2e-05,
      "loss": 0.510442066192627,
      "step": 3400
    },
    {
      "epoch": 0.24453209035496593,
      "grad_norm": 1.0029207814802068,
      "learning_rate": 2e-05,
      "loss": 0.5701600551605225,
      "step": 3410
    },
    {
      "epoch": 0.2452491932592327,
      "grad_norm": 0.8084316716061816,
      "learning_rate": 2e-05,
      "loss": 0.5356520652770996,
      "step": 3420
    },
    {
      "epoch": 0.24596629616349946,
      "grad_norm": 0.9444985856972721,
      "learning_rate": 2e-05,
      "loss": 0.5171973705291748,
      "step": 3430
    },
    {
      "epoch": 0.24668339906776623,
      "grad_norm": 1.239231030557441,
      "learning_rate": 2e-05,
      "loss": 0.5146626472473145,
      "step": 3440
    },
    {
      "epoch": 0.24740050197203298,
      "grad_norm": 0.8860347217971428,
      "learning_rate": 2e-05,
      "loss": 0.5645951271057129,
      "step": 3450
    },
    {
      "epoch": 0.24811760487629975,
      "grad_norm": 0.9524017468707715,
      "learning_rate": 2e-05,
      "loss": 0.5330866813659668,
      "step": 3460
    },
    {
      "epoch": 0.24883470778056652,
      "grad_norm": 2.4021819187714897,
      "learning_rate": 2e-05,
      "loss": 0.5168228149414062,
      "step": 3470
    },
    {
      "epoch": 0.2495518106848333,
      "grad_norm": 0.9192021490084639,
      "learning_rate": 2e-05,
      "loss": 0.5290431976318359,
      "step": 3480
    },
    {
      "epoch": 0.25026891358910003,
      "grad_norm": 5.274174125099747,
      "learning_rate": 2e-05,
      "loss": 0.5303543090820313,
      "step": 3490
    },
    {
      "epoch": 0.2509860164933668,
      "grad_norm": 0.8344525861907959,
      "learning_rate": 2e-05,
      "loss": 0.5316099166870117,
      "step": 3500
    },
    {
      "epoch": 0.25170311939763357,
      "grad_norm": 2.532316574091822,
      "learning_rate": 2e-05,
      "loss": 0.536371374130249,
      "step": 3510
    },
    {
      "epoch": 0.2524202223019003,
      "grad_norm": 1.014915926059095,
      "learning_rate": 2e-05,
      "loss": 0.5119599342346192,
      "step": 3520
    },
    {
      "epoch": 0.2531373252061671,
      "grad_norm": 1.0726759257721148,
      "learning_rate": 2e-05,
      "loss": 0.5139832496643066,
      "step": 3530
    },
    {
      "epoch": 0.25385442811043385,
      "grad_norm": 2.0756616365943032,
      "learning_rate": 2e-05,
      "loss": 0.539209508895874,
      "step": 3540
    },
    {
      "epoch": 0.2545715310147006,
      "grad_norm": 1.4781916753433906,
      "learning_rate": 2e-05,
      "loss": 0.5556020259857177,
      "step": 3550
    },
    {
      "epoch": 0.2552886339189674,
      "grad_norm": 1.65751674075069,
      "learning_rate": 2e-05,
      "loss": 0.5389930725097656,
      "step": 3560
    },
    {
      "epoch": 0.25600573682323413,
      "grad_norm": 3.311024605587316,
      "learning_rate": 2e-05,
      "loss": 0.5277709007263184,
      "step": 3570
    },
    {
      "epoch": 0.2567228397275009,
      "grad_norm": 2.1532638970995874,
      "learning_rate": 2e-05,
      "loss": 0.5385482788085938,
      "step": 3580
    },
    {
      "epoch": 0.25743994263176767,
      "grad_norm": 1.1805518941031328,
      "learning_rate": 2e-05,
      "loss": 0.5248273849487305,
      "step": 3590
    },
    {
      "epoch": 0.2581570455360344,
      "grad_norm": 29.354213445015297,
      "learning_rate": 2e-05,
      "loss": 0.5063570976257324,
      "step": 3600
    },
    {
      "epoch": 0.2588741484403012,
      "grad_norm": 1.665322524196583,
      "learning_rate": 2e-05,
      "loss": 0.5178260803222656,
      "step": 3610
    },
    {
      "epoch": 0.25959125134456795,
      "grad_norm": 0.8407935012783503,
      "learning_rate": 2e-05,
      "loss": 0.5238101005554199,
      "step": 3620
    },
    {
      "epoch": 0.2603083542488347,
      "grad_norm": 2.5302057816202312,
      "learning_rate": 2e-05,
      "loss": 0.54418044090271,
      "step": 3630
    },
    {
      "epoch": 0.2610254571531015,
      "grad_norm": 1.552121980779559,
      "learning_rate": 2e-05,
      "loss": 0.5144917488098144,
      "step": 3640
    },
    {
      "epoch": 0.26174256005736823,
      "grad_norm": 1.225435775677276,
      "learning_rate": 2e-05,
      "loss": 0.562162208557129,
      "step": 3650
    },
    {
      "epoch": 0.262459662961635,
      "grad_norm": 0.8797288554528758,
      "learning_rate": 2e-05,
      "loss": 0.5646907329559326,
      "step": 3660
    },
    {
      "epoch": 0.2631767658659018,
      "grad_norm": 1.347051134805898,
      "learning_rate": 2e-05,
      "loss": 0.5711068153381348,
      "step": 3670
    },
    {
      "epoch": 0.2638938687701685,
      "grad_norm": 1.333551450324836,
      "learning_rate": 2e-05,
      "loss": 0.5483678817749024,
      "step": 3680
    },
    {
      "epoch": 0.26461097167443526,
      "grad_norm": 2.195866516016493,
      "learning_rate": 2e-05,
      "loss": 0.5161405563354492,
      "step": 3690
    },
    {
      "epoch": 0.26532807457870206,
      "grad_norm": 1.6014945613608733,
      "learning_rate": 2e-05,
      "loss": 0.532387113571167,
      "step": 3700
    },
    {
      "epoch": 0.2660451774829688,
      "grad_norm": 0.8520925187454512,
      "learning_rate": 2e-05,
      "loss": 0.5006357669830322,
      "step": 3710
    },
    {
      "epoch": 0.2667622803872356,
      "grad_norm": 1.0232420993118372,
      "learning_rate": 2e-05,
      "loss": 0.5466555595397949,
      "step": 3720
    },
    {
      "epoch": 0.26747938329150234,
      "grad_norm": 0.9063998789417105,
      "learning_rate": 2e-05,
      "loss": 0.5533414840698242,
      "step": 3730
    },
    {
      "epoch": 0.2681964861957691,
      "grad_norm": 0.9546873245254411,
      "learning_rate": 2e-05,
      "loss": 0.5072546005249023,
      "step": 3740
    },
    {
      "epoch": 0.2689135891000359,
      "grad_norm": 1.5263898065195027,
      "learning_rate": 2e-05,
      "loss": 0.5570329666137696,
      "step": 3750
    },
    {
      "epoch": 0.2696306920043026,
      "grad_norm": 0.8443160777695132,
      "learning_rate": 2e-05,
      "loss": 0.5478463649749756,
      "step": 3760
    },
    {
      "epoch": 0.27034779490856936,
      "grad_norm": 2.1174151477553647,
      "learning_rate": 2e-05,
      "loss": 0.5399794578552246,
      "step": 3770
    },
    {
      "epoch": 0.27106489781283616,
      "grad_norm": 1.950903072184313,
      "learning_rate": 2e-05,
      "loss": 0.5461333274841309,
      "step": 3780
    },
    {
      "epoch": 0.2717820007171029,
      "grad_norm": 0.9136340620181176,
      "learning_rate": 2e-05,
      "loss": 0.5399363517761231,
      "step": 3790
    },
    {
      "epoch": 0.27249910362136964,
      "grad_norm": 0.8861245330145846,
      "learning_rate": 2e-05,
      "loss": 0.5201873302459716,
      "step": 3800
    },
    {
      "epoch": 0.27321620652563644,
      "grad_norm": 1.180221150603258,
      "learning_rate": 2e-05,
      "loss": 0.5267035484313964,
      "step": 3810
    },
    {
      "epoch": 0.2739333094299032,
      "grad_norm": 0.9389187060838954,
      "learning_rate": 2e-05,
      "loss": 0.5209812521934509,
      "step": 3820
    },
    {
      "epoch": 0.27465041233417,
      "grad_norm": 1.3220100280712899,
      "learning_rate": 2e-05,
      "loss": 0.5314033031463623,
      "step": 3830
    },
    {
      "epoch": 0.2753675152384367,
      "grad_norm": 0.9608258726988983,
      "learning_rate": 2e-05,
      "loss": 0.5270042419433594,
      "step": 3840
    },
    {
      "epoch": 0.27608461814270346,
      "grad_norm": 1.016477631154093,
      "learning_rate": 2e-05,
      "loss": 0.5294764518737793,
      "step": 3850
    },
    {
      "epoch": 0.27680172104697026,
      "grad_norm": 0.9996171693502619,
      "learning_rate": 2e-05,
      "loss": 0.530165433883667,
      "step": 3860
    },
    {
      "epoch": 0.277518823951237,
      "grad_norm": 1.3573493573744488,
      "learning_rate": 2e-05,
      "loss": 0.5282859325408935,
      "step": 3870
    },
    {
      "epoch": 0.27823592685550375,
      "grad_norm": 1.0847483668887548,
      "learning_rate": 2e-05,
      "loss": 0.49652738571166993,
      "step": 3880
    },
    {
      "epoch": 0.27895302975977054,
      "grad_norm": 0.7712798214604878,
      "learning_rate": 2e-05,
      "loss": 0.5158729553222656,
      "step": 3890
    },
    {
      "epoch": 0.2796701326640373,
      "grad_norm": 1.106602119960059,
      "learning_rate": 2e-05,
      "loss": 0.5592980861663819,
      "step": 3900
    },
    {
      "epoch": 0.280387235568304,
      "grad_norm": 1.4724735547610919,
      "learning_rate": 2e-05,
      "loss": 0.5308915615081787,
      "step": 3910
    },
    {
      "epoch": 0.2811043384725708,
      "grad_norm": 0.9403666728628767,
      "learning_rate": 2e-05,
      "loss": 0.5160143375396729,
      "step": 3920
    },
    {
      "epoch": 0.28182144137683757,
      "grad_norm": 1.694656937624822,
      "learning_rate": 2e-05,
      "loss": 0.5392077922821045,
      "step": 3930
    },
    {
      "epoch": 0.28253854428110436,
      "grad_norm": 1.2452076016180897,
      "learning_rate": 2e-05,
      "loss": 0.5507253170013428,
      "step": 3940
    },
    {
      "epoch": 0.2832556471853711,
      "grad_norm": 2.974873619913777,
      "learning_rate": 2e-05,
      "loss": 0.5465413570404053,
      "step": 3950
    },
    {
      "epoch": 0.28397275008963785,
      "grad_norm": 1.1534235526219585,
      "learning_rate": 2e-05,
      "loss": 0.5165030479431152,
      "step": 3960
    },
    {
      "epoch": 0.28468985299390465,
      "grad_norm": 2.9502152994146624,
      "learning_rate": 2e-05,
      "loss": 0.5044128894805908,
      "step": 3970
    },
    {
      "epoch": 0.2854069558981714,
      "grad_norm": 0.9184504001729815,
      "learning_rate": 2e-05,
      "loss": 0.5352005958557129,
      "step": 3980
    },
    {
      "epoch": 0.28612405880243813,
      "grad_norm": 1.0964621945520778,
      "learning_rate": 2e-05,
      "loss": 0.5197076320648193,
      "step": 3990
    },
    {
      "epoch": 0.28684116170670493,
      "grad_norm": 1.4099469907213347,
      "learning_rate": 2e-05,
      "loss": 0.5218649864196777,
      "step": 4000
    },
    {
      "epoch": 0.28755826461097167,
      "grad_norm": 1.1386574741495505,
      "learning_rate": 2e-05,
      "loss": 0.5544497489929199,
      "step": 4010
    },
    {
      "epoch": 0.2882753675152384,
      "grad_norm": 1.0553293170974636,
      "learning_rate": 2e-05,
      "loss": 0.5686587810516357,
      "step": 4020
    },
    {
      "epoch": 0.2889924704195052,
      "grad_norm": 2.642830531991087,
      "learning_rate": 2e-05,
      "loss": 0.5437402725219727,
      "step": 4030
    },
    {
      "epoch": 0.28970957332377195,
      "grad_norm": 1.0242995774868573,
      "learning_rate": 2e-05,
      "loss": 0.5438858032226562,
      "step": 4040
    },
    {
      "epoch": 0.29042667622803875,
      "grad_norm": 0.7190137841709671,
      "learning_rate": 2e-05,
      "loss": 0.5411813259124756,
      "step": 4050
    },
    {
      "epoch": 0.2911437791323055,
      "grad_norm": 14.614135045516889,
      "learning_rate": 2e-05,
      "loss": 0.5463586330413819,
      "step": 4060
    },
    {
      "epoch": 0.29186088203657223,
      "grad_norm": 1.113213654004875,
      "learning_rate": 2e-05,
      "loss": 0.53390531539917,
      "step": 4070
    },
    {
      "epoch": 0.29257798494083903,
      "grad_norm": 1.0015230929868442,
      "learning_rate": 2e-05,
      "loss": 0.5596723079681396,
      "step": 4080
    },
    {
      "epoch": 0.2932950878451058,
      "grad_norm": 1.218465853073684,
      "learning_rate": 2e-05,
      "loss": 0.5296479225158691,
      "step": 4090
    },
    {
      "epoch": 0.2940121907493725,
      "grad_norm": 0.7039612452404016,
      "learning_rate": 2e-05,
      "loss": 0.5507210731506348,
      "step": 4100
    },
    {
      "epoch": 0.2947292936536393,
      "grad_norm": 1.0123835593741664,
      "learning_rate": 2e-05,
      "loss": 0.5401140213012695,
      "step": 4110
    },
    {
      "epoch": 0.29544639655790605,
      "grad_norm": 1.2212902328196638,
      "learning_rate": 2e-05,
      "loss": 0.5185504913330078,
      "step": 4120
    },
    {
      "epoch": 0.2961634994621728,
      "grad_norm": 1.8693332764017534,
      "learning_rate": 2e-05,
      "loss": 0.5265801906585693,
      "step": 4130
    },
    {
      "epoch": 0.2968806023664396,
      "grad_norm": 1.1857902550373702,
      "learning_rate": 2e-05,
      "loss": 0.5151444911956787,
      "step": 4140
    },
    {
      "epoch": 0.29759770527070634,
      "grad_norm": 1.7785456338767356,
      "learning_rate": 2e-05,
      "loss": 0.5372767448425293,
      "step": 4150
    },
    {
      "epoch": 0.29831480817497313,
      "grad_norm": 1.363346178823168,
      "learning_rate": 2e-05,
      "loss": 0.5341384887695313,
      "step": 4160
    },
    {
      "epoch": 0.2990319110792399,
      "grad_norm": 2.524915505766991,
      "learning_rate": 2e-05,
      "loss": 0.5464763641357422,
      "step": 4170
    },
    {
      "epoch": 0.2997490139835066,
      "grad_norm": 1.4082213101116239,
      "learning_rate": 2e-05,
      "loss": 0.5652936458587646,
      "step": 4180
    },
    {
      "epoch": 0.3004661168877734,
      "grad_norm": 0.8530093265865526,
      "learning_rate": 2e-05,
      "loss": 0.5270353794097901,
      "step": 4190
    },
    {
      "epoch": 0.30118321979204016,
      "grad_norm": 1.38664713014284,
      "learning_rate": 2e-05,
      "loss": 0.5116106986999511,
      "step": 4200
    },
    {
      "epoch": 0.3019003226963069,
      "grad_norm": 1.392138235272807,
      "learning_rate": 2e-05,
      "loss": 0.5565159797668457,
      "step": 4210
    },
    {
      "epoch": 0.3026174256005737,
      "grad_norm": 1.550581001539546,
      "learning_rate": 2e-05,
      "loss": 0.5518420219421387,
      "step": 4220
    },
    {
      "epoch": 0.30333452850484044,
      "grad_norm": 1.2208588043056474,
      "learning_rate": 2e-05,
      "loss": 0.5366368293762207,
      "step": 4230
    },
    {
      "epoch": 0.3040516314091072,
      "grad_norm": 3.7825770886550214,
      "learning_rate": 2e-05,
      "loss": 0.5275087356567383,
      "step": 4240
    },
    {
      "epoch": 0.304768734313374,
      "grad_norm": 1.0212443117449186,
      "learning_rate": 2e-05,
      "loss": 0.5116767406463623,
      "step": 4250
    },
    {
      "epoch": 0.3054858372176407,
      "grad_norm": 2.8731481864687507,
      "learning_rate": 2e-05,
      "loss": 0.5333958625793457,
      "step": 4260
    },
    {
      "epoch": 0.3062029401219075,
      "grad_norm": 1.9745399240757586,
      "learning_rate": 2e-05,
      "loss": 0.5407708168029786,
      "step": 4270
    },
    {
      "epoch": 0.30692004302617426,
      "grad_norm": 1.2889965271129373,
      "learning_rate": 2e-05,
      "loss": 0.5515640258789063,
      "step": 4280
    },
    {
      "epoch": 0.307637145930441,
      "grad_norm": 0.983464229245475,
      "learning_rate": 2e-05,
      "loss": 0.5141390800476074,
      "step": 4290
    },
    {
      "epoch": 0.3083542488347078,
      "grad_norm": 8.515348719332833,
      "learning_rate": 2e-05,
      "loss": 0.5346375942230225,
      "step": 4300
    },
    {
      "epoch": 0.30907135173897454,
      "grad_norm": 4.951509439930093,
      "learning_rate": 2e-05,
      "loss": 0.5025247097015381,
      "step": 4310
    },
    {
      "epoch": 0.3097884546432413,
      "grad_norm": 2.466353920240643,
      "learning_rate": 2e-05,
      "loss": 0.519663143157959,
      "step": 4320
    },
    {
      "epoch": 0.3105055575475081,
      "grad_norm": 6.555134202015613,
      "learning_rate": 2e-05,
      "loss": 0.5303162574768067,
      "step": 4330
    },
    {
      "epoch": 0.3112226604517748,
      "grad_norm": 1.7222927500272274,
      "learning_rate": 2e-05,
      "loss": 0.5187395095825196,
      "step": 4340
    },
    {
      "epoch": 0.31193976335604157,
      "grad_norm": 2.313892767433708,
      "learning_rate": 2e-05,
      "loss": 0.533453893661499,
      "step": 4350
    },
    {
      "epoch": 0.31265686626030836,
      "grad_norm": 2.7627030619045936,
      "learning_rate": 2e-05,
      "loss": 0.5287415504455566,
      "step": 4360
    },
    {
      "epoch": 0.3133739691645751,
      "grad_norm": 1.3962793407820708,
      "learning_rate": 2e-05,
      "loss": 0.5264087677001953,
      "step": 4370
    },
    {
      "epoch": 0.3140910720688419,
      "grad_norm": 1.4640850710111042,
      "learning_rate": 2e-05,
      "loss": 0.5092585563659668,
      "step": 4380
    },
    {
      "epoch": 0.31480817497310865,
      "grad_norm": 2.6786212148275745,
      "learning_rate": 2e-05,
      "loss": 0.5113193035125733,
      "step": 4390
    },
    {
      "epoch": 0.3155252778773754,
      "grad_norm": 1.1800758915920027,
      "learning_rate": 2e-05,
      "loss": 0.5511657238006592,
      "step": 4400
    },
    {
      "epoch": 0.3162423807816422,
      "grad_norm": 1.12627421631863,
      "learning_rate": 2e-05,
      "loss": 0.5638663291931152,
      "step": 4410
    },
    {
      "epoch": 0.3169594836859089,
      "grad_norm": 2.366781634213916,
      "learning_rate": 2e-05,
      "loss": 0.5290012359619141,
      "step": 4420
    },
    {
      "epoch": 0.31767658659017567,
      "grad_norm": 3.3950310351101756,
      "learning_rate": 2e-05,
      "loss": 0.5063624858856202,
      "step": 4430
    },
    {
      "epoch": 0.31839368949444247,
      "grad_norm": 1.0834323735768503,
      "learning_rate": 2e-05,
      "loss": 0.5289322376251221,
      "step": 4440
    },
    {
      "epoch": 0.3191107923987092,
      "grad_norm": 1.5137404555368836,
      "learning_rate": 2e-05,
      "loss": 0.5425243377685547,
      "step": 4450
    },
    {
      "epoch": 0.31982789530297595,
      "grad_norm": 1.1508927274229868,
      "learning_rate": 2e-05,
      "loss": 0.5203255653381348,
      "step": 4460
    },
    {
      "epoch": 0.32054499820724275,
      "grad_norm": 1.8683452570591337,
      "learning_rate": 2e-05,
      "loss": 0.5171571254730225,
      "step": 4470
    },
    {
      "epoch": 0.3212621011115095,
      "grad_norm": 1.1643514612025672,
      "learning_rate": 2e-05,
      "loss": 0.5488979339599609,
      "step": 4480
    },
    {
      "epoch": 0.3219792040157763,
      "grad_norm": 0.9864118527933871,
      "learning_rate": 2e-05,
      "loss": 0.5390284538269043,
      "step": 4490
    },
    {
      "epoch": 0.32269630692004303,
      "grad_norm": 1.6318540638937824,
      "learning_rate": 2e-05,
      "loss": 0.532197380065918,
      "step": 4500
    },
    {
      "epoch": 0.32341340982430977,
      "grad_norm": 1.5072901835830799,
      "learning_rate": 2e-05,
      "loss": 0.5034425735473633,
      "step": 4510
    },
    {
      "epoch": 0.32413051272857657,
      "grad_norm": 0.8753447833806923,
      "learning_rate": 2e-05,
      "loss": 0.5351991176605224,
      "step": 4520
    },
    {
      "epoch": 0.3248476156328433,
      "grad_norm": 1.52815700375204,
      "learning_rate": 2e-05,
      "loss": 0.5048668861389161,
      "step": 4530
    },
    {
      "epoch": 0.32556471853711005,
      "grad_norm": 1.901855364681695,
      "learning_rate": 2e-05,
      "loss": 0.5330635070800781,
      "step": 4540
    },
    {
      "epoch": 0.32628182144137685,
      "grad_norm": 1.757007968540639,
      "learning_rate": 2e-05,
      "loss": 0.5253935813903808,
      "step": 4550
    },
    {
      "epoch": 0.3269989243456436,
      "grad_norm": 1.00566028061641,
      "learning_rate": 2e-05,
      "loss": 0.5167190551757812,
      "step": 4560
    },
    {
      "epoch": 0.32771602724991034,
      "grad_norm": 1.107327077762691,
      "learning_rate": 2e-05,
      "loss": 0.5395001888275146,
      "step": 4570
    },
    {
      "epoch": 0.32843313015417713,
      "grad_norm": 1.8501148291422014,
      "learning_rate": 2e-05,
      "loss": 0.5178787231445312,
      "step": 4580
    },
    {
      "epoch": 0.3291502330584439,
      "grad_norm": 2.799641520445622,
      "learning_rate": 2e-05,
      "loss": 0.5427615642547607,
      "step": 4590
    },
    {
      "epoch": 0.3298673359627107,
      "grad_norm": 2.711284491704616,
      "learning_rate": 2e-05,
      "loss": 0.534752607345581,
      "step": 4600
    },
    {
      "epoch": 0.3305844388669774,
      "grad_norm": 1.2533239046028486,
      "learning_rate": 2e-05,
      "loss": 0.5291178703308106,
      "step": 4610
    },
    {
      "epoch": 0.33130154177124416,
      "grad_norm": 0.9915195655896355,
      "learning_rate": 2e-05,
      "loss": 0.5130389213562012,
      "step": 4620
    },
    {
      "epoch": 0.33201864467551095,
      "grad_norm": 1.2227054501571062,
      "learning_rate": 2e-05,
      "loss": 0.5204622268676757,
      "step": 4630
    },
    {
      "epoch": 0.3327357475797777,
      "grad_norm": 2.3802400022667562,
      "learning_rate": 2e-05,
      "loss": 0.5109448432922363,
      "step": 4640
    },
    {
      "epoch": 0.33345285048404444,
      "grad_norm": 1.0456704938477226,
      "learning_rate": 2e-05,
      "loss": 0.5191596984863281,
      "step": 4650
    },
    {
      "epoch": 0.33416995338831124,
      "grad_norm": 2.65731513943049,
      "learning_rate": 2e-05,
      "loss": 0.5229734420776367,
      "step": 4660
    },
    {
      "epoch": 0.334887056292578,
      "grad_norm": 2.0036119142051914,
      "learning_rate": 2e-05,
      "loss": 0.5267868041992188,
      "step": 4670
    },
    {
      "epoch": 0.3356041591968447,
      "grad_norm": 1.035263199734906,
      "learning_rate": 2e-05,
      "loss": 0.4989652156829834,
      "step": 4680
    },
    {
      "epoch": 0.3363212621011115,
      "grad_norm": 1.0163646754379652,
      "learning_rate": 2e-05,
      "loss": 0.5221956253051758,
      "step": 4690
    },
    {
      "epoch": 0.33703836500537826,
      "grad_norm": 1.1074936638302706,
      "learning_rate": 2e-05,
      "loss": 0.5124061107635498,
      "step": 4700
    },
    {
      "epoch": 0.33775546790964506,
      "grad_norm": 3.888852140890692,
      "learning_rate": 2e-05,
      "loss": 0.5347827434539795,
      "step": 4710
    },
    {
      "epoch": 0.3384725708139118,
      "grad_norm": 0.90629220837514,
      "learning_rate": 2e-05,
      "loss": 0.5260302543640136,
      "step": 4720
    },
    {
      "epoch": 0.33918967371817854,
      "grad_norm": 3.126440824180423,
      "learning_rate": 2e-05,
      "loss": 0.5285738945007324,
      "step": 4730
    },
    {
      "epoch": 0.33990677662244534,
      "grad_norm": 1.0586499350100411,
      "learning_rate": 2e-05,
      "loss": 0.531917142868042,
      "step": 4740
    },
    {
      "epoch": 0.3406238795267121,
      "grad_norm": 3.1764112878282975,
      "learning_rate": 2e-05,
      "loss": 0.5203787803649902,
      "step": 4750
    },
    {
      "epoch": 0.3413409824309788,
      "grad_norm": 1.1331256005626882,
      "learning_rate": 2e-05,
      "loss": 0.5360140800476074,
      "step": 4760
    },
    {
      "epoch": 0.3420580853352456,
      "grad_norm": 2.0215060580892685,
      "learning_rate": 2e-05,
      "loss": 0.526363754272461,
      "step": 4770
    },
    {
      "epoch": 0.34277518823951236,
      "grad_norm": 4.533042501599692,
      "learning_rate": 2e-05,
      "loss": 0.5227174758911133,
      "step": 4780
    },
    {
      "epoch": 0.3434922911437791,
      "grad_norm": 1.2325856800697683,
      "learning_rate": 2e-05,
      "loss": 0.5319549560546875,
      "step": 4790
    },
    {
      "epoch": 0.3442093940480459,
      "grad_norm": 1.340920686152608,
      "learning_rate": 2e-05,
      "loss": 0.5372313499450684,
      "step": 4800
    },
    {
      "epoch": 0.34492649695231264,
      "grad_norm": 1.3173933192544798,
      "learning_rate": 2e-05,
      "loss": 0.5511189937591553,
      "step": 4810
    },
    {
      "epoch": 0.34564359985657944,
      "grad_norm": 0.9596099801534932,
      "learning_rate": 2e-05,
      "loss": 0.5339679718017578,
      "step": 4820
    },
    {
      "epoch": 0.3463607027608462,
      "grad_norm": 1.0040086330382332,
      "learning_rate": 2e-05,
      "loss": 0.50563063621521,
      "step": 4830
    },
    {
      "epoch": 0.3470778056651129,
      "grad_norm": 1.0598091678012862,
      "learning_rate": 2e-05,
      "loss": 0.5031904697418212,
      "step": 4840
    },
    {
      "epoch": 0.3477949085693797,
      "grad_norm": 1.5244838558070852,
      "learning_rate": 2e-05,
      "loss": 0.5157588958740235,
      "step": 4850
    },
    {
      "epoch": 0.34851201147364647,
      "grad_norm": 0.9026251934815781,
      "learning_rate": 2e-05,
      "loss": 0.5442389965057373,
      "step": 4860
    },
    {
      "epoch": 0.3492291143779132,
      "grad_norm": 0.9478540582915709,
      "learning_rate": 2e-05,
      "loss": 0.540727949142456,
      "step": 4870
    },
    {
      "epoch": 0.34994621728218,
      "grad_norm": 1.0421560990213585,
      "learning_rate": 2e-05,
      "loss": 0.5248729705810546,
      "step": 4880
    },
    {
      "epoch": 0.35066332018644675,
      "grad_norm": 5.25101085370611,
      "learning_rate": 2e-05,
      "loss": 0.5678351879119873,
      "step": 4890
    },
    {
      "epoch": 0.35138042309071355,
      "grad_norm": 1.389189851582616,
      "learning_rate": 2e-05,
      "loss": 0.5246406555175781,
      "step": 4900
    },
    {
      "epoch": 0.3520975259949803,
      "grad_norm": 1.0951700002698306,
      "learning_rate": 2e-05,
      "loss": 0.5252338409423828,
      "step": 4910
    },
    {
      "epoch": 0.35281462889924703,
      "grad_norm": 0.9122396027782109,
      "learning_rate": 2e-05,
      "loss": 0.5341548919677734,
      "step": 4920
    },
    {
      "epoch": 0.3535317318035138,
      "grad_norm": 1.0076267328096273,
      "learning_rate": 2e-05,
      "loss": 0.49324932098388674,
      "step": 4930
    },
    {
      "epoch": 0.35424883470778057,
      "grad_norm": 0.8047253829881544,
      "learning_rate": 2e-05,
      "loss": 0.5239223003387451,
      "step": 4940
    },
    {
      "epoch": 0.3549659376120473,
      "grad_norm": 1.0098798188999853,
      "learning_rate": 2e-05,
      "loss": 0.5509189605712891,
      "step": 4950
    },
    {
      "epoch": 0.3556830405163141,
      "grad_norm": 8.60583974006901,
      "learning_rate": 2e-05,
      "loss": 0.5055221557617188,
      "step": 4960
    },
    {
      "epoch": 0.35640014342058085,
      "grad_norm": 1.1607639145470612,
      "learning_rate": 2e-05,
      "loss": 0.5417621612548829,
      "step": 4970
    },
    {
      "epoch": 0.3571172463248476,
      "grad_norm": 3.751013345669409,
      "learning_rate": 2e-05,
      "loss": 0.5157419681549072,
      "step": 4980
    },
    {
      "epoch": 0.3578343492291144,
      "grad_norm": 1.4857148503055762,
      "learning_rate": 2e-05,
      "loss": 0.5178676128387452,
      "step": 4990
    },
    {
      "epoch": 0.35855145213338113,
      "grad_norm": 7.915702626598026,
      "learning_rate": 2e-05,
      "loss": 0.531338882446289,
      "step": 5000
    },
    {
      "epoch": 0.35926855503764793,
      "grad_norm": 1.9769264204888977,
      "learning_rate": 2e-05,
      "loss": 0.5668124198913574,
      "step": 5010
    },
    {
      "epoch": 0.35998565794191467,
      "grad_norm": 10.838044522511575,
      "learning_rate": 2e-05,
      "loss": 0.5197597026824952,
      "step": 5020
    },
    {
      "epoch": 0.3607027608461814,
      "grad_norm": 1.8319351960877779,
      "learning_rate": 2e-05,
      "loss": 0.5356554985046387,
      "step": 5030
    },
    {
      "epoch": 0.3614198637504482,
      "grad_norm": 3.455049659550564,
      "learning_rate": 2e-05,
      "loss": 0.5165534019470215,
      "step": 5040
    },
    {
      "epoch": 0.36213696665471495,
      "grad_norm": 9.466819552958066,
      "learning_rate": 2e-05,
      "loss": 0.5224682807922363,
      "step": 5050
    },
    {
      "epoch": 0.3628540695589817,
      "grad_norm": 5.21432942555595,
      "learning_rate": 2e-05,
      "loss": 0.4877131462097168,
      "step": 5060
    },
    {
      "epoch": 0.3635711724632485,
      "grad_norm": 1.0673995253463284,
      "learning_rate": 2e-05,
      "loss": 0.5266811370849609,
      "step": 5070
    },
    {
      "epoch": 0.36428827536751524,
      "grad_norm": 6.129977722955653,
      "learning_rate": 2e-05,
      "loss": 0.5531869888305664,
      "step": 5080
    },
    {
      "epoch": 0.365005378271782,
      "grad_norm": 5.885450735489292,
      "learning_rate": 1.9999810479633703e-05,
      "loss": 0.4845612943172455,
      "step": 5090
    },
    {
      "epoch": 0.3657224811760488,
      "grad_norm": 0.8578374432488265,
      "learning_rate": 1.997707677020786e-05,
      "loss": 0.5335223197937011,
      "step": 5100
    },
    {
      "epoch": 0.3664395840803155,
      "grad_norm": 0.8468743019783441,
      "learning_rate": 1.9916538196124637e-05,
      "loss": 0.5337824821472168,
      "step": 5110
    },
    {
      "epoch": 0.3671566869845823,
      "grad_norm": 0.9107340589976517,
      "learning_rate": 1.9818425303863402e-05,
      "loss": 0.5015262126922607,
      "step": 5120
    },
    {
      "epoch": 0.36787378988884906,
      "grad_norm": 3.198994812480517,
      "learning_rate": 1.9683111732582887e-05,
      "loss": 0.5401925086975098,
      "step": 5130
    },
    {
      "epoch": 0.3685908927931158,
      "grad_norm": 1.1797806454942736,
      "learning_rate": 1.951111279120706e-05,
      "loss": 0.5354208946228027,
      "step": 5140
    },
    {
      "epoch": 0.3693079956973826,
      "grad_norm": 4.008021567932825,
      "learning_rate": 1.9303083495996028e-05,
      "loss": 0.5266697883605957,
      "step": 5150
    },
    {
      "epoch": 0.37002509860164934,
      "grad_norm": 1.0541632757348802,
      "learning_rate": 1.9059816076075482e-05,
      "loss": 0.5131072044372559,
      "step": 5160
    },
    {
      "epoch": 0.3707422015059161,
      "grad_norm": 1.1645188029493285,
      "learning_rate": 1.8782236956424142e-05,
      "loss": 0.540519905090332,
      "step": 5170
    },
    {
      "epoch": 0.3714593044101829,
      "grad_norm": 0.8826591303441744,
      "learning_rate": 1.8471403229808897e-05,
      "loss": 0.5153768539428711,
      "step": 5180
    },
    {
      "epoch": 0.3721764073144496,
      "grad_norm": 1.8263623841214607,
      "learning_rate": 1.8128498631103333e-05,
      "loss": 0.5364732265472412,
      "step": 5190
    },
    {
      "epoch": 0.37289351021871636,
      "grad_norm": 1.173800821007004,
      "learning_rate": 1.7754829029320507e-05,
      "loss": 0.5266950130462646,
      "step": 5200
    },
    {
      "epoch": 0.37361061312298316,
      "grad_norm": 1.077467289001195,
      "learning_rate": 1.7351817454527473e-05,
      "loss": 0.5172656059265137,
      "step": 5210
    },
    {
      "epoch": 0.3743277160272499,
      "grad_norm": 1.947216500100435,
      "learning_rate": 1.6920998678580345e-05,
      "loss": 0.5077250003814697,
      "step": 5220
    },
    {
      "epoch": 0.3750448189315167,
      "grad_norm": 1.5624160181522377,
      "learning_rate": 1.6464013370317854e-05,
      "loss": 0.5360892295837403,
      "step": 5230
    },
    {
      "epoch": 0.37576192183578344,
      "grad_norm": 2.2006449362365585,
      "learning_rate": 1.5982601847471915e-05,
      "loss": 0.522796630859375,
      "step": 5240
    },
    {
      "epoch": 0.3764790247400502,
      "grad_norm": 1.0225361683690688,
      "learning_rate": 1.5478597449089553e-05,
      "loss": 0.5150138854980468,
      "step": 5250
    },
    {
      "epoch": 0.377196127644317,
      "grad_norm": 6.8584735404611905,
      "learning_rate": 1.4953919553705725e-05,
      "loss": 0.520006799697876,
      "step": 5260
    },
    {
      "epoch": 0.3779132305485837,
      "grad_norm": 1.6169849878866462,
      "learning_rate": 1.441056626985559e-05,
      "loss": 0.4943543434143066,
      "step": 5270
    },
    {
      "epoch": 0.37863033345285046,
      "grad_norm": 1.1952284542762845,
      "learning_rate": 1.3850606826762668e-05,
      "loss": 0.5314961433410644,
      "step": 5280
    },
    {
      "epoch": 0.37934743635711726,
      "grad_norm": 1.1706034124327218,
      "learning_rate": 1.3276173694181128e-05,
      "loss": 0.500275707244873,
      "step": 5290
    },
    {
      "epoch": 0.380064539261384,
      "grad_norm": 1.8061689319638707,
      "learning_rate": 1.2689454461401822e-05,
      "loss": 0.5084277629852295,
      "step": 5300
    },
    {
      "epoch": 0.38078164216565075,
      "grad_norm": 2.940853362522607,
      "learning_rate": 1.2092683506348939e-05,
      "loss": 0.5183092594146729,
      "step": 5310
    },
    {
      "epoch": 0.38149874506991754,
      "grad_norm": 0.9959893222401127,
      "learning_rate": 1.14881334864935e-05,
      "loss": 0.5086060523986816,
      "step": 5320
    },
    {
      "epoch": 0.3822158479741843,
      "grad_norm": 1.3819621629867098,
      "learning_rate": 1.0878106683988481e-05,
      "loss": 0.533271598815918,
      "step": 5330
    },
    {
      "epoch": 0.3829329508784511,
      "grad_norm": 1.0298740669444277,
      "learning_rate": 1.026492623798551e-05,
      "loss": 0.5067360401153564,
      "step": 5340
    },
    {
      "epoch": 0.3836500537827178,
      "grad_norm": 0.9430567448493488,
      "learning_rate": 9.650927297522743e-06,
      "loss": 0.49420967102050783,
      "step": 5350
    },
    {
      "epoch": 0.38436715668698457,
      "grad_norm": 4.36703762075774,
      "learning_rate": 9.038448128676012e-06,
      "loss": 0.5140265464782715,
      "step": 5360
    },
    {
      "epoch": 0.38508425959125137,
      "grad_norm": 1.0258157561705894,
      "learning_rate": 8.42982120983949e-06,
      "loss": 0.5288449287414551,
      "step": 5370
    },
    {
      "epoch": 0.3858013624955181,
      "grad_norm": 0.8309707443339012,
      "learning_rate": 7.827364349047303e-06,
      "loss": 0.507795000076294,
      "step": 5380
    },
    {
      "epoch": 0.38651846539978485,
      "grad_norm": 0.8888894835458627,
      "learning_rate": 7.233371857163706e-06,
      "loss": 0.5214560985565185,
      "step": 5390
    },
    {
      "epoch": 0.38723556830405165,
      "grad_norm": 1.154523457832301,
      "learning_rate": 6.650105810556492e-06,
      "loss": 0.5035552024841309,
      "step": 5400
    },
    {
      "epoch": 0.3879526712083184,
      "grad_norm": 1.7315878931056325,
      "learning_rate": 6.079787436527681e-06,
      "loss": 0.4948132514953613,
      "step": 5410
    },
    {
      "epoch": 0.38866977411258513,
      "grad_norm": 1.1458467054743722,
      "learning_rate": 5.524588654307961e-06,
      "loss": 0.5273829460144043,
      "step": 5420
    },
    {
      "epoch": 0.38938687701685193,
      "grad_norm": 1.2326530849712252,
      "learning_rate": 4.986623803828985e-06,
      "loss": 0.5072244167327881,
      "step": 5430
    },
    {
      "epoch": 0.39010397992111867,
      "grad_norm": 1.7109409474705501,
      "learning_rate": 4.467941593772423e-06,
      "loss": 0.5228410243988038,
      "step": 5440
    },
    {
      "epoch": 0.39082108282538547,
      "grad_norm": 1.4688711772604919,
      "learning_rate": 3.9705172995597606e-06,
      "loss": 0.5167178630828857,
      "step": 5450
    },
    {
      "epoch": 0.3915381857296522,
      "grad_norm": 0.9293744894100463,
      "learning_rate": 3.4962452409948346e-06,
      "loss": 0.5037213325500488,
      "step": 5460
    },
    {
      "epoch": 0.39225528863391895,
      "grad_norm": 0.8179235689894482,
      "learning_rate": 3.046931568206211e-06,
      "loss": 0.5308568000793457,
      "step": 5470
    },
    {
      "epoch": 0.39297239153818575,
      "grad_norm": 0.7888398518081081,
      "learning_rate": 2.624287383362404e-06,
      "loss": 0.49132394790649414,
      "step": 5480
    },
    {
      "epoch": 0.3936894944424525,
      "grad_norm": 0.728709433610097,
      "learning_rate": 2.2299222243541845e-06,
      "loss": 0.5655451774597168,
      "step": 5490
    },
    {
      "epoch": 0.39440659734671923,
      "grad_norm": 0.9718226253818731,
      "learning_rate": 1.8653379352598246e-06,
      "loss": 0.5242514610290527,
      "step": 5500
    },
    {
      "epoch": 0.39512370025098603,
      "grad_norm": 0.9029968921109245,
      "learning_rate": 1.5319229469361075e-06,
      "loss": 0.5026238441467286,
      "step": 5510
    },
    {
      "epoch": 0.3958408031552528,
      "grad_norm": 0.984976085431073,
      "learning_rate": 1.2309469895160917e-06,
      "loss": 0.5218173027038574,
      "step": 5520
    },
    {
      "epoch": 0.3965579060595195,
      "grad_norm": 0.7823317973581628,
      "learning_rate": 9.635562569497974e-07,
      "loss": 0.5378507137298584,
      "step": 5530
    },
    {
      "epoch": 0.3972750089637863,
      "grad_norm": 1.4500016561259723,
      "learning_rate": 7.307690420024867e-07,
      "loss": 0.5203337669372559,
      "step": 5540
    },
    {
      "epoch": 0.39799211186805306,
      "grad_norm": 0.8584226900297329,
      "learning_rate": 5.334718583335942e-07,
      "loss": 0.48284616470336916,
      "step": 5550
    },
    {
      "epoch": 0.39870921477231985,
      "grad_norm": 1.679756589559498,
      "learning_rate": 3.7241606442442763e-07,
      "loss": 0.513909101486206,
      "step": 5560
    },
    {
      "epoch": 0.3994263176765866,
      "grad_norm": 1.726778100976104,
      "learning_rate": 2.482150022115879e-07,
      "loss": 0.4923055171966553,
      "step": 5570
    },
    {
      "epoch": 0.40014342058085334,
      "grad_norm": 0.9152960832999824,
      "learning_rate": 1.6134166132293174e-07,
      "loss": 0.5052491188049316,
      "step": 5580
    },
    {
      "epoch": 0.40086052348512013,
      "grad_norm": 0.8113705761614092,
      "learning_rate": 1.1212687781127357e-07,
      "loss": 0.49610447883605957,
      "step": 5590
    },
    {
      "epoch": 0.4013624955181068,
      "step": 5597,
      "total_flos": 2650292248117248.0,
      "train_loss": 0.04694117999447442,
      "train_runtime": 7327.2293,
      "train_samples_per_second": 9.166,
      "train_steps_per_second": 0.764
    }
  ],
  "logging_steps": 10,
  "max_steps": 5597,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2650292248117248.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}