{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.802653280745787,
  "eval_steps": 500,
  "global_step": 11193,
  "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": 5.885450735495938,
      "learning_rate": 2e-05,
      "loss": 0.4845612943172455,
      "step": 5090
    },
    {
      "epoch": 0.3657224811760488,
      "grad_norm": 2.8709824941403155,
      "learning_rate": 2e-05,
      "loss": 0.5337962627410888,
      "step": 5100
    },
    {
      "epoch": 0.3664395840803155,
      "grad_norm": 1.2945390788637718,
      "learning_rate": 2e-05,
      "loss": 0.5349855422973633,
      "step": 5110
    },
    {
      "epoch": 0.3671566869845823,
      "grad_norm": 2.722858315208877,
      "learning_rate": 2e-05,
      "loss": 0.5012556076049804,
      "step": 5120
    },
    {
      "epoch": 0.36787378988884906,
      "grad_norm": 1.1141667817182908,
      "learning_rate": 2e-05,
      "loss": 0.540011978149414,
      "step": 5130
    },
    {
      "epoch": 0.3685908927931158,
      "grad_norm": 1.8162398991327446,
      "learning_rate": 2e-05,
      "loss": 0.5363057613372803,
      "step": 5140
    },
    {
      "epoch": 0.3693079956973826,
      "grad_norm": 0.8800152198878165,
      "learning_rate": 2e-05,
      "loss": 0.5279021263122559,
      "step": 5150
    },
    {
      "epoch": 0.37002509860164934,
      "grad_norm": 1.2874872917772517,
      "learning_rate": 2e-05,
      "loss": 0.5135499000549316,
      "step": 5160
    },
    {
      "epoch": 0.3707422015059161,
      "grad_norm": 2.0491548354695053,
      "learning_rate": 2e-05,
      "loss": 0.5396522521972656,
      "step": 5170
    },
    {
      "epoch": 0.3714593044101829,
      "grad_norm": 1.5362709278996014,
      "learning_rate": 2e-05,
      "loss": 0.513334083557129,
      "step": 5180
    },
    {
      "epoch": 0.3721764073144496,
      "grad_norm": 2.63079891750155,
      "learning_rate": 2e-05,
      "loss": 0.5350643157958984,
      "step": 5190
    },
    {
      "epoch": 0.37289351021871636,
      "grad_norm": 1.2204085450587876,
      "learning_rate": 2e-05,
      "loss": 0.531666374206543,
      "step": 5200
    },
    {
      "epoch": 0.37361061312298316,
      "grad_norm": 1.4431062179569702,
      "learning_rate": 2e-05,
      "loss": 0.5227872371673584,
      "step": 5210
    },
    {
      "epoch": 0.3743277160272499,
      "grad_norm": 0.8444199443166995,
      "learning_rate": 2e-05,
      "loss": 0.5112868309020996,
      "step": 5220
    },
    {
      "epoch": 0.3750448189315167,
      "grad_norm": 0.8652115544212067,
      "learning_rate": 2e-05,
      "loss": 0.5384913444519043,
      "step": 5230
    },
    {
      "epoch": 0.37576192183578344,
      "grad_norm": 3.9806570351461206,
      "learning_rate": 2e-05,
      "loss": 0.5273669242858887,
      "step": 5240
    },
    {
      "epoch": 0.3764790247400502,
      "grad_norm": 3.3500967640544133,
      "learning_rate": 2e-05,
      "loss": 0.523277473449707,
      "step": 5250
    },
    {
      "epoch": 0.377196127644317,
      "grad_norm": 2.535593160623039,
      "learning_rate": 2e-05,
      "loss": 0.5294794082641602,
      "step": 5260
    },
    {
      "epoch": 0.3779132305485837,
      "grad_norm": 0.935853812197936,
      "learning_rate": 2e-05,
      "loss": 0.5019505500793457,
      "step": 5270
    },
    {
      "epoch": 0.37863033345285046,
      "grad_norm": 0.946231069126987,
      "learning_rate": 2e-05,
      "loss": 0.538211727142334,
      "step": 5280
    },
    {
      "epoch": 0.37934743635711726,
      "grad_norm": 1.7709519124503446,
      "learning_rate": 2e-05,
      "loss": 0.5054364204406738,
      "step": 5290
    },
    {
      "epoch": 0.380064539261384,
      "grad_norm": 1.0338395459282967,
      "learning_rate": 2e-05,
      "loss": 0.5115403652191162,
      "step": 5300
    },
    {
      "epoch": 0.38078164216565075,
      "grad_norm": 1.460091305610164,
      "learning_rate": 2e-05,
      "loss": 0.5257485866546631,
      "step": 5310
    },
    {
      "epoch": 0.38149874506991754,
      "grad_norm": 1.5131654672870716,
      "learning_rate": 2e-05,
      "loss": 0.5165888786315918,
      "step": 5320
    },
    {
      "epoch": 0.3822158479741843,
      "grad_norm": 2.3295466813410988,
      "learning_rate": 2e-05,
      "loss": 0.5405088424682617,
      "step": 5330
    },
    {
      "epoch": 0.3829329508784511,
      "grad_norm": 1.147938808621501,
      "learning_rate": 2e-05,
      "loss": 0.5152339935302734,
      "step": 5340
    },
    {
      "epoch": 0.3836500537827178,
      "grad_norm": 1.1254455696202017,
      "learning_rate": 2e-05,
      "loss": 0.5036863327026367,
      "step": 5350
    },
    {
      "epoch": 0.38436715668698457,
      "grad_norm": 1.3527798701740714,
      "learning_rate": 2e-05,
      "loss": 0.5239635467529297,
      "step": 5360
    },
    {
      "epoch": 0.38508425959125137,
      "grad_norm": 1.0269288445303906,
      "learning_rate": 2e-05,
      "loss": 0.5389392852783204,
      "step": 5370
    },
    {
      "epoch": 0.3858013624955181,
      "grad_norm": 0.8396476945147898,
      "learning_rate": 2e-05,
      "loss": 0.5162075996398926,
      "step": 5380
    },
    {
      "epoch": 0.38651846539978485,
      "grad_norm": 3.851913707129276,
      "learning_rate": 2e-05,
      "loss": 0.5337397575378418,
      "step": 5390
    },
    {
      "epoch": 0.38723556830405165,
      "grad_norm": 1.306037937923041,
      "learning_rate": 2e-05,
      "loss": 0.5159441947937011,
      "step": 5400
    },
    {
      "epoch": 0.3879526712083184,
      "grad_norm": 0.7535694781133777,
      "learning_rate": 2e-05,
      "loss": 0.5087882041931152,
      "step": 5410
    },
    {
      "epoch": 0.38866977411258513,
      "grad_norm": 2.847579342135248,
      "learning_rate": 2e-05,
      "loss": 0.5413926601409912,
      "step": 5420
    },
    {
      "epoch": 0.38938687701685193,
      "grad_norm": 0.8929998251063018,
      "learning_rate": 2e-05,
      "loss": 0.521876335144043,
      "step": 5430
    },
    {
      "epoch": 0.39010397992111867,
      "grad_norm": 2.111633950507903,
      "learning_rate": 2e-05,
      "loss": 0.5384711742401123,
      "step": 5440
    },
    {
      "epoch": 0.39082108282538547,
      "grad_norm": 1.8377363428352038,
      "learning_rate": 2e-05,
      "loss": 0.5314854621887207,
      "step": 5450
    },
    {
      "epoch": 0.3915381857296522,
      "grad_norm": 1.520185113816346,
      "learning_rate": 2e-05,
      "loss": 0.518360185623169,
      "step": 5460
    },
    {
      "epoch": 0.39225528863391895,
      "grad_norm": 0.7271111559102872,
      "learning_rate": 2e-05,
      "loss": 0.5468538284301758,
      "step": 5470
    },
    {
      "epoch": 0.39297239153818575,
      "grad_norm": 4.126351529562747,
      "learning_rate": 2e-05,
      "loss": 0.5073840618133545,
      "step": 5480
    },
    {
      "epoch": 0.3936894944424525,
      "grad_norm": 0.905711620795008,
      "learning_rate": 2e-05,
      "loss": 0.5827337265014648,
      "step": 5490
    },
    {
      "epoch": 0.39440659734671923,
      "grad_norm": 1.8604808055818127,
      "learning_rate": 2e-05,
      "loss": 0.5401986122131348,
      "step": 5500
    },
    {
      "epoch": 0.39512370025098603,
      "grad_norm": 0.9222031241485954,
      "learning_rate": 2e-05,
      "loss": 0.520060920715332,
      "step": 5510
    },
    {
      "epoch": 0.3958408031552528,
      "grad_norm": 1.236797019093231,
      "learning_rate": 2e-05,
      "loss": 0.544208288192749,
      "step": 5520
    },
    {
      "epoch": 0.3965579060595195,
      "grad_norm": 1.119441656402785,
      "learning_rate": 2e-05,
      "loss": 0.5672105312347412,
      "step": 5530
    },
    {
      "epoch": 0.3972750089637863,
      "grad_norm": 1.2632659063605018,
      "learning_rate": 2e-05,
      "loss": 0.5386402130126953,
      "step": 5540
    },
    {
      "epoch": 0.39799211186805306,
      "grad_norm": 1.1259226412199075,
      "learning_rate": 2e-05,
      "loss": 0.5049736976623536,
      "step": 5550
    },
    {
      "epoch": 0.39870921477231985,
      "grad_norm": 1.548639765954057,
      "learning_rate": 2e-05,
      "loss": 0.5358623027801513,
      "step": 5560
    },
    {
      "epoch": 0.3994263176765866,
      "grad_norm": 2.135182575082456,
      "learning_rate": 2e-05,
      "loss": 0.5156712532043457,
      "step": 5570
    },
    {
      "epoch": 0.40014342058085334,
      "grad_norm": 2.0732905341659973,
      "learning_rate": 2e-05,
      "loss": 0.526421070098877,
      "step": 5580
    },
    {
      "epoch": 0.40086052348512013,
      "grad_norm": 1.3458670312000156,
      "learning_rate": 2e-05,
      "loss": 0.5146687984466553,
      "step": 5590
    },
    {
      "epoch": 0.4015776263893869,
      "grad_norm": 1.1059221486100566,
      "learning_rate": 2e-05,
      "loss": 0.5612788677215577,
      "step": 5600
    },
    {
      "epoch": 0.4022947292936536,
      "grad_norm": 1.4462415635323265,
      "learning_rate": 2e-05,
      "loss": 0.5334926605224609,
      "step": 5610
    },
    {
      "epoch": 0.4030118321979204,
      "grad_norm": 1.4004579286019814,
      "learning_rate": 2e-05,
      "loss": 0.5033141136169433,
      "step": 5620
    },
    {
      "epoch": 0.40372893510218716,
      "grad_norm": 1.6534241829176526,
      "learning_rate": 2e-05,
      "loss": 0.497715425491333,
      "step": 5630
    },
    {
      "epoch": 0.4044460380064539,
      "grad_norm": 1.629700540037044,
      "learning_rate": 2e-05,
      "loss": 0.5278496742248535,
      "step": 5640
    },
    {
      "epoch": 0.4051631409107207,
      "grad_norm": 1.0749988283795089,
      "learning_rate": 2e-05,
      "loss": 0.5378868103027343,
      "step": 5650
    },
    {
      "epoch": 0.40588024381498744,
      "grad_norm": 1.025839047811762,
      "learning_rate": 2e-05,
      "loss": 0.553062915802002,
      "step": 5660
    },
    {
      "epoch": 0.40659734671925424,
      "grad_norm": 0.8628646506306391,
      "learning_rate": 2e-05,
      "loss": 0.5129246711730957,
      "step": 5670
    },
    {
      "epoch": 0.407314449623521,
      "grad_norm": 1.1971439831730404,
      "learning_rate": 2e-05,
      "loss": 0.514282512664795,
      "step": 5680
    },
    {
      "epoch": 0.4080315525277877,
      "grad_norm": 1.5831320201010552,
      "learning_rate": 2e-05,
      "loss": 0.540608549118042,
      "step": 5690
    },
    {
      "epoch": 0.4087486554320545,
      "grad_norm": 1.485116958655447,
      "learning_rate": 2e-05,
      "loss": 0.5153422355651855,
      "step": 5700
    },
    {
      "epoch": 0.40946575833632126,
      "grad_norm": 5.272529579314188,
      "learning_rate": 2e-05,
      "loss": 0.5322002410888672,
      "step": 5710
    },
    {
      "epoch": 0.410182861240588,
      "grad_norm": 2.9648760469139623,
      "learning_rate": 2e-05,
      "loss": 0.5251474380493164,
      "step": 5720
    },
    {
      "epoch": 0.4108999641448548,
      "grad_norm": 0.9311292977615292,
      "learning_rate": 2e-05,
      "loss": 0.5479231834411621,
      "step": 5730
    },
    {
      "epoch": 0.41161706704912154,
      "grad_norm": 1.1856414227994456,
      "learning_rate": 2e-05,
      "loss": 0.5043671607971192,
      "step": 5740
    },
    {
      "epoch": 0.4123341699533883,
      "grad_norm": 0.9023142929832717,
      "learning_rate": 2e-05,
      "loss": 0.5138603210449219,
      "step": 5750
    },
    {
      "epoch": 0.4130512728576551,
      "grad_norm": 4.756293365620802,
      "learning_rate": 2e-05,
      "loss": 0.5419458389282227,
      "step": 5760
    },
    {
      "epoch": 0.4137683757619218,
      "grad_norm": 1.6528057094483823,
      "learning_rate": 2e-05,
      "loss": 0.5089199066162109,
      "step": 5770
    },
    {
      "epoch": 0.4144854786661886,
      "grad_norm": 1.2439549320461283,
      "learning_rate": 2e-05,
      "loss": 0.5379460334777832,
      "step": 5780
    },
    {
      "epoch": 0.41520258157045536,
      "grad_norm": 0.8563028871333074,
      "learning_rate": 2e-05,
      "loss": 0.5384929656982422,
      "step": 5790
    },
    {
      "epoch": 0.4159196844747221,
      "grad_norm": 0.8672167110878314,
      "learning_rate": 2e-05,
      "loss": 0.5733089447021484,
      "step": 5800
    },
    {
      "epoch": 0.4166367873789889,
      "grad_norm": 1.0316971900707306,
      "learning_rate": 2e-05,
      "loss": 0.5067746639251709,
      "step": 5810
    },
    {
      "epoch": 0.41735389028325565,
      "grad_norm": 1.3713585113923767,
      "learning_rate": 2e-05,
      "loss": 0.5139177322387696,
      "step": 5820
    },
    {
      "epoch": 0.4180709931875224,
      "grad_norm": 1.006036906650695,
      "learning_rate": 2e-05,
      "loss": 0.544105052947998,
      "step": 5830
    },
    {
      "epoch": 0.4187880960917892,
      "grad_norm": 1.5528121178514331,
      "learning_rate": 2e-05,
      "loss": 0.5179839134216309,
      "step": 5840
    },
    {
      "epoch": 0.41950519899605593,
      "grad_norm": 1.2504063394552074,
      "learning_rate": 2e-05,
      "loss": 0.5363343715667724,
      "step": 5850
    },
    {
      "epoch": 0.42022230190032267,
      "grad_norm": 1.811707081647874,
      "learning_rate": 2e-05,
      "loss": 0.5397977352142334,
      "step": 5860
    },
    {
      "epoch": 0.42093940480458947,
      "grad_norm": 0.957503458359509,
      "learning_rate": 2e-05,
      "loss": 0.48824443817138674,
      "step": 5870
    },
    {
      "epoch": 0.4216565077088562,
      "grad_norm": 1.1222891657424572,
      "learning_rate": 2e-05,
      "loss": 0.5444816589355469,
      "step": 5880
    },
    {
      "epoch": 0.422373610613123,
      "grad_norm": 1.5874875475166934,
      "learning_rate": 2e-05,
      "loss": 0.5276690483093261,
      "step": 5890
    },
    {
      "epoch": 0.42309071351738975,
      "grad_norm": 1.241981529440426,
      "learning_rate": 2e-05,
      "loss": 0.539062786102295,
      "step": 5900
    },
    {
      "epoch": 0.4238078164216565,
      "grad_norm": 1.3779266205225706,
      "learning_rate": 2e-05,
      "loss": 0.5208881855010986,
      "step": 5910
    },
    {
      "epoch": 0.4245249193259233,
      "grad_norm": 1.5659252596003148,
      "learning_rate": 2e-05,
      "loss": 0.5127540588378906,
      "step": 5920
    },
    {
      "epoch": 0.42524202223019003,
      "grad_norm": 1.633142473158656,
      "learning_rate": 2e-05,
      "loss": 0.4942172050476074,
      "step": 5930
    },
    {
      "epoch": 0.4259591251344568,
      "grad_norm": 0.7689668742100614,
      "learning_rate": 2e-05,
      "loss": 0.4939394950866699,
      "step": 5940
    },
    {
      "epoch": 0.42667622803872357,
      "grad_norm": 4.450639299335645,
      "learning_rate": 2e-05,
      "loss": 0.5585159301757813,
      "step": 5950
    },
    {
      "epoch": 0.4273933309429903,
      "grad_norm": 0.8638854870275406,
      "learning_rate": 2e-05,
      "loss": 0.5196020126342773,
      "step": 5960
    },
    {
      "epoch": 0.42811043384725705,
      "grad_norm": 1.4583811986570616,
      "learning_rate": 2e-05,
      "loss": 0.544192886352539,
      "step": 5970
    },
    {
      "epoch": 0.42882753675152385,
      "grad_norm": 1.0060547058127647,
      "learning_rate": 2e-05,
      "loss": 0.5164608478546142,
      "step": 5980
    },
    {
      "epoch": 0.4295446396557906,
      "grad_norm": 1.0159918271320811,
      "learning_rate": 2e-05,
      "loss": 0.5187869071960449,
      "step": 5990
    },
    {
      "epoch": 0.4302617425600574,
      "grad_norm": 1.0212473340366888,
      "learning_rate": 2e-05,
      "loss": 0.5110889911651612,
      "step": 6000
    },
    {
      "epoch": 0.43097884546432413,
      "grad_norm": 1.3711818462100223,
      "learning_rate": 2e-05,
      "loss": 0.5473899364471435,
      "step": 6010
    },
    {
      "epoch": 0.4316959483685909,
      "grad_norm": 1.6834034148278143,
      "learning_rate": 2e-05,
      "loss": 0.5210264682769775,
      "step": 6020
    },
    {
      "epoch": 0.4324130512728577,
      "grad_norm": 1.5651385117093723,
      "learning_rate": 2e-05,
      "loss": 0.5073967933654785,
      "step": 6030
    },
    {
      "epoch": 0.4331301541771244,
      "grad_norm": 2.925382449405084,
      "learning_rate": 2e-05,
      "loss": 0.5156419754028321,
      "step": 6040
    },
    {
      "epoch": 0.43384725708139116,
      "grad_norm": 1.3043457060190073,
      "learning_rate": 2e-05,
      "loss": 0.5325917720794677,
      "step": 6050
    },
    {
      "epoch": 0.43456435998565796,
      "grad_norm": 0.8284487167693143,
      "learning_rate": 2e-05,
      "loss": 0.5133628368377685,
      "step": 6060
    },
    {
      "epoch": 0.4352814628899247,
      "grad_norm": 1.016000612638695,
      "learning_rate": 2e-05,
      "loss": 0.4952703475952148,
      "step": 6070
    },
    {
      "epoch": 0.43599856579419144,
      "grad_norm": 1.1576066234572866,
      "learning_rate": 2e-05,
      "loss": 0.5305257320404053,
      "step": 6080
    },
    {
      "epoch": 0.43671566869845824,
      "grad_norm": 0.7549027054886107,
      "learning_rate": 2e-05,
      "loss": 0.5658648490905762,
      "step": 6090
    },
    {
      "epoch": 0.437432771602725,
      "grad_norm": 1.6967520429429384,
      "learning_rate": 2e-05,
      "loss": 0.546066427230835,
      "step": 6100
    },
    {
      "epoch": 0.4381498745069918,
      "grad_norm": 0.7490055292650781,
      "learning_rate": 2e-05,
      "loss": 0.48830885887145997,
      "step": 6110
    },
    {
      "epoch": 0.4388669774112585,
      "grad_norm": 1.2036084902700797,
      "learning_rate": 2e-05,
      "loss": 0.5185158729553223,
      "step": 6120
    },
    {
      "epoch": 0.43958408031552526,
      "grad_norm": 0.812338967965654,
      "learning_rate": 2e-05,
      "loss": 0.5172608375549317,
      "step": 6130
    },
    {
      "epoch": 0.44030118321979206,
      "grad_norm": 0.9447185554446527,
      "learning_rate": 2e-05,
      "loss": 0.5030579090118408,
      "step": 6140
    },
    {
      "epoch": 0.4410182861240588,
      "grad_norm": 1.3083370979880198,
      "learning_rate": 2e-05,
      "loss": 0.5243852138519287,
      "step": 6150
    },
    {
      "epoch": 0.44173538902832554,
      "grad_norm": 1.2402778440939786,
      "learning_rate": 2e-05,
      "loss": 0.5246669769287109,
      "step": 6160
    },
    {
      "epoch": 0.44245249193259234,
      "grad_norm": 1.2939028438047924,
      "learning_rate": 2e-05,
      "loss": 0.5467393398284912,
      "step": 6170
    },
    {
      "epoch": 0.4431695948368591,
      "grad_norm": 1.219015814523352,
      "learning_rate": 2e-05,
      "loss": 0.5271720886230469,
      "step": 6180
    },
    {
      "epoch": 0.4438866977411258,
      "grad_norm": 0.9551394106747191,
      "learning_rate": 2e-05,
      "loss": 0.5271878242492676,
      "step": 6190
    },
    {
      "epoch": 0.4446038006453926,
      "grad_norm": 2.060546910253664,
      "learning_rate": 2e-05,
      "loss": 0.514436674118042,
      "step": 6200
    },
    {
      "epoch": 0.44532090354965936,
      "grad_norm": 0.8955998892775776,
      "learning_rate": 2e-05,
      "loss": 0.5196824073791504,
      "step": 6210
    },
    {
      "epoch": 0.44603800645392616,
      "grad_norm": 0.7043688871400492,
      "learning_rate": 2e-05,
      "loss": 0.5090668678283692,
      "step": 6220
    },
    {
      "epoch": 0.4467551093581929,
      "grad_norm": 0.7308527236638483,
      "learning_rate": 2e-05,
      "loss": 0.5288050174713135,
      "step": 6230
    },
    {
      "epoch": 0.44747221226245965,
      "grad_norm": 0.8096229214790117,
      "learning_rate": 2e-05,
      "loss": 0.5364199638366699,
      "step": 6240
    },
    {
      "epoch": 0.44818931516672644,
      "grad_norm": 0.7461245566320648,
      "learning_rate": 2e-05,
      "loss": 0.5376228332519531,
      "step": 6250
    },
    {
      "epoch": 0.4489064180709932,
      "grad_norm": 1.0638399852337723,
      "learning_rate": 2e-05,
      "loss": 0.5262404918670655,
      "step": 6260
    },
    {
      "epoch": 0.4496235209752599,
      "grad_norm": 0.9156347603277113,
      "learning_rate": 2e-05,
      "loss": 0.5034090042114258,
      "step": 6270
    },
    {
      "epoch": 0.4503406238795267,
      "grad_norm": 0.6977252258145337,
      "learning_rate": 2e-05,
      "loss": 0.534706449508667,
      "step": 6280
    },
    {
      "epoch": 0.45105772678379347,
      "grad_norm": 0.8522760603993185,
      "learning_rate": 2e-05,
      "loss": 0.5163982391357422,
      "step": 6290
    },
    {
      "epoch": 0.45177482968806026,
      "grad_norm": 0.847349773457464,
      "learning_rate": 2e-05,
      "loss": 0.4907248497009277,
      "step": 6300
    },
    {
      "epoch": 0.452491932592327,
      "grad_norm": 1.2355950102075264,
      "learning_rate": 2e-05,
      "loss": 0.5181584358215332,
      "step": 6310
    },
    {
      "epoch": 0.45320903549659375,
      "grad_norm": 0.7190621382120541,
      "learning_rate": 2e-05,
      "loss": 0.5298931121826171,
      "step": 6320
    },
    {
      "epoch": 0.45392613840086055,
      "grad_norm": 0.7202838413097189,
      "learning_rate": 2e-05,
      "loss": 0.5027827739715576,
      "step": 6330
    },
    {
      "epoch": 0.4546432413051273,
      "grad_norm": 0.827724631220609,
      "learning_rate": 2e-05,
      "loss": 0.5020524501800537,
      "step": 6340
    },
    {
      "epoch": 0.45536034420939403,
      "grad_norm": 0.8987462863408229,
      "learning_rate": 2e-05,
      "loss": 0.5029661178588867,
      "step": 6350
    },
    {
      "epoch": 0.4560774471136608,
      "grad_norm": 1.5309767432395616,
      "learning_rate": 2e-05,
      "loss": 0.5420424461364746,
      "step": 6360
    },
    {
      "epoch": 0.45679455001792757,
      "grad_norm": 1.3029276475717717,
      "learning_rate": 2e-05,
      "loss": 0.49623813629150393,
      "step": 6370
    },
    {
      "epoch": 0.4575116529221943,
      "grad_norm": 1.4315776065993124,
      "learning_rate": 2e-05,
      "loss": 0.5240662574768067,
      "step": 6380
    },
    {
      "epoch": 0.4582287558264611,
      "grad_norm": 0.9376597300188894,
      "learning_rate": 2e-05,
      "loss": 0.5171424865722656,
      "step": 6390
    },
    {
      "epoch": 0.45894585873072785,
      "grad_norm": 0.9761630296308905,
      "learning_rate": 2e-05,
      "loss": 0.513507080078125,
      "step": 6400
    },
    {
      "epoch": 0.45966296163499465,
      "grad_norm": 1.5224696127894157,
      "learning_rate": 2e-05,
      "loss": 0.5314605712890625,
      "step": 6410
    },
    {
      "epoch": 0.4603800645392614,
      "grad_norm": 0.7349672638271562,
      "learning_rate": 2e-05,
      "loss": 0.5307876110076905,
      "step": 6420
    },
    {
      "epoch": 0.46109716744352813,
      "grad_norm": 0.702651591343042,
      "learning_rate": 2e-05,
      "loss": 0.5485776424407959,
      "step": 6430
    },
    {
      "epoch": 0.46181427034779493,
      "grad_norm": 1.448660144233961,
      "learning_rate": 2e-05,
      "loss": 0.5036701202392578,
      "step": 6440
    },
    {
      "epoch": 0.4625313732520617,
      "grad_norm": 1.13153535417172,
      "learning_rate": 2e-05,
      "loss": 0.4820553779602051,
      "step": 6450
    },
    {
      "epoch": 0.4632484761563284,
      "grad_norm": 0.9500863949795827,
      "learning_rate": 2e-05,
      "loss": 0.5098736763000489,
      "step": 6460
    },
    {
      "epoch": 0.4639655790605952,
      "grad_norm": 0.8103389161605266,
      "learning_rate": 2e-05,
      "loss": 0.5152860641479492,
      "step": 6470
    },
    {
      "epoch": 0.46468268196486195,
      "grad_norm": 10.086001846347886,
      "learning_rate": 2e-05,
      "loss": 0.5229338645935059,
      "step": 6480
    },
    {
      "epoch": 0.4653997848691287,
      "grad_norm": 1.9490465954764422,
      "learning_rate": 2e-05,
      "loss": 0.5463828086853028,
      "step": 6490
    },
    {
      "epoch": 0.4661168877733955,
      "grad_norm": 1.1592928950854877,
      "learning_rate": 2e-05,
      "loss": 0.5169078826904296,
      "step": 6500
    },
    {
      "epoch": 0.46683399067766224,
      "grad_norm": 1.0973086798841265,
      "learning_rate": 2e-05,
      "loss": 0.5476903915405273,
      "step": 6510
    },
    {
      "epoch": 0.46755109358192903,
      "grad_norm": 0.7529705685099629,
      "learning_rate": 2e-05,
      "loss": 0.5072021961212159,
      "step": 6520
    },
    {
      "epoch": 0.4682681964861958,
      "grad_norm": 1.9469692415841338,
      "learning_rate": 2e-05,
      "loss": 0.5093785285949707,
      "step": 6530
    },
    {
      "epoch": 0.4689852993904625,
      "grad_norm": 1.1453225678913317,
      "learning_rate": 2e-05,
      "loss": 0.5313010215759277,
      "step": 6540
    },
    {
      "epoch": 0.4697024022947293,
      "grad_norm": 3.7568403058601696,
      "learning_rate": 2e-05,
      "loss": 0.5229932308197022,
      "step": 6550
    },
    {
      "epoch": 0.47041950519899606,
      "grad_norm": 0.7904110679620225,
      "learning_rate": 2e-05,
      "loss": 0.504290246963501,
      "step": 6560
    },
    {
      "epoch": 0.4711366081032628,
      "grad_norm": 1.0604907996265862,
      "learning_rate": 2e-05,
      "loss": 0.5312373161315918,
      "step": 6570
    },
    {
      "epoch": 0.4718537110075296,
      "grad_norm": 0.7316604526362106,
      "learning_rate": 2e-05,
      "loss": 0.49577808380126953,
      "step": 6580
    },
    {
      "epoch": 0.47257081391179634,
      "grad_norm": 2.9334037001052162,
      "learning_rate": 2e-05,
      "loss": 0.5149497985839844,
      "step": 6590
    },
    {
      "epoch": 0.4732879168160631,
      "grad_norm": 0.9389301475289864,
      "learning_rate": 2e-05,
      "loss": 0.5393352508544922,
      "step": 6600
    },
    {
      "epoch": 0.4740050197203299,
      "grad_norm": 1.35276376605144,
      "learning_rate": 2e-05,
      "loss": 0.5180294513702393,
      "step": 6610
    },
    {
      "epoch": 0.4747221226245966,
      "grad_norm": 1.6408579377876902,
      "learning_rate": 2e-05,
      "loss": 0.5368182182312011,
      "step": 6620
    },
    {
      "epoch": 0.4754392255288634,
      "grad_norm": 1.0901536725590715,
      "learning_rate": 2e-05,
      "loss": 0.506042194366455,
      "step": 6630
    },
    {
      "epoch": 0.47615632843313016,
      "grad_norm": 0.9102070392005359,
      "learning_rate": 2e-05,
      "loss": 0.49120941162109377,
      "step": 6640
    },
    {
      "epoch": 0.4768734313373969,
      "grad_norm": 3.5159876505157484,
      "learning_rate": 2e-05,
      "loss": 0.510521650314331,
      "step": 6650
    },
    {
      "epoch": 0.4775905342416637,
      "grad_norm": 1.099465001397764,
      "learning_rate": 2e-05,
      "loss": 0.4974991798400879,
      "step": 6660
    },
    {
      "epoch": 0.47830763714593044,
      "grad_norm": 3.8837940916308695,
      "learning_rate": 2e-05,
      "loss": 0.5043962478637696,
      "step": 6670
    },
    {
      "epoch": 0.4790247400501972,
      "grad_norm": 0.9365675674640453,
      "learning_rate": 2e-05,
      "loss": 0.5328217506408691,
      "step": 6680
    },
    {
      "epoch": 0.479741842954464,
      "grad_norm": 1.5713741100850178,
      "learning_rate": 2e-05,
      "loss": 0.4920331001281738,
      "step": 6690
    },
    {
      "epoch": 0.4804589458587307,
      "grad_norm": 3.7734094660523345,
      "learning_rate": 2e-05,
      "loss": 0.5161496639251709,
      "step": 6700
    },
    {
      "epoch": 0.48117604876299747,
      "grad_norm": 1.0680495208184426,
      "learning_rate": 2e-05,
      "loss": 0.5300738334655761,
      "step": 6710
    },
    {
      "epoch": 0.48189315166726426,
      "grad_norm": 1.685070713720968,
      "learning_rate": 2e-05,
      "loss": 0.5296891212463379,
      "step": 6720
    },
    {
      "epoch": 0.482610254571531,
      "grad_norm": 2.651438883287199,
      "learning_rate": 2e-05,
      "loss": 0.5121171474456787,
      "step": 6730
    },
    {
      "epoch": 0.4833273574757978,
      "grad_norm": 14.430830627718997,
      "learning_rate": 2e-05,
      "loss": 0.49791483879089354,
      "step": 6740
    },
    {
      "epoch": 0.48404446038006455,
      "grad_norm": 1.130871435175107,
      "learning_rate": 2e-05,
      "loss": 0.5328787803649903,
      "step": 6750
    },
    {
      "epoch": 0.4847615632843313,
      "grad_norm": 2.3560679712043116,
      "learning_rate": 2e-05,
      "loss": 0.5369107723236084,
      "step": 6760
    },
    {
      "epoch": 0.4854786661885981,
      "grad_norm": 1.6543310680975258,
      "learning_rate": 2e-05,
      "loss": 0.5069418907165527,
      "step": 6770
    },
    {
      "epoch": 0.4861957690928648,
      "grad_norm": 2.073017984336786,
      "learning_rate": 2e-05,
      "loss": 0.5080055713653564,
      "step": 6780
    },
    {
      "epoch": 0.48691287199713157,
      "grad_norm": 1.1216206322260165,
      "learning_rate": 2e-05,
      "loss": 0.5288147926330566,
      "step": 6790
    },
    {
      "epoch": 0.48762997490139837,
      "grad_norm": 1.1723918516186254,
      "learning_rate": 2e-05,
      "loss": 0.5277268409729003,
      "step": 6800
    },
    {
      "epoch": 0.4883470778056651,
      "grad_norm": 1.0821409435477256,
      "learning_rate": 2e-05,
      "loss": 0.5201609134674072,
      "step": 6810
    },
    {
      "epoch": 0.48906418070993185,
      "grad_norm": 31.31871743740438,
      "learning_rate": 2e-05,
      "loss": 0.5345252513885498,
      "step": 6820
    },
    {
      "epoch": 0.48978128361419865,
      "grad_norm": 2.070130402305151,
      "learning_rate": 2e-05,
      "loss": 0.49734721183776853,
      "step": 6830
    },
    {
      "epoch": 0.4904983865184654,
      "grad_norm": 1.3014554736759636,
      "learning_rate": 2e-05,
      "loss": 0.5233652114868164,
      "step": 6840
    },
    {
      "epoch": 0.4912154894227322,
      "grad_norm": 2.7348128324406638,
      "learning_rate": 2e-05,
      "loss": 0.5209245681762695,
      "step": 6850
    },
    {
      "epoch": 0.49193259232699893,
      "grad_norm": 1.991919081728195,
      "learning_rate": 2e-05,
      "loss": 0.5143784523010254,
      "step": 6860
    },
    {
      "epoch": 0.49264969523126567,
      "grad_norm": 6.08590749014203,
      "learning_rate": 2e-05,
      "loss": 0.5389420986175537,
      "step": 6870
    },
    {
      "epoch": 0.49336679813553247,
      "grad_norm": 0.9737614482005105,
      "learning_rate": 2e-05,
      "loss": 0.5167324066162109,
      "step": 6880
    },
    {
      "epoch": 0.4940839010397992,
      "grad_norm": 2.6511467972081184,
      "learning_rate": 2e-05,
      "loss": 0.5411920070648193,
      "step": 6890
    },
    {
      "epoch": 0.49480100394406595,
      "grad_norm": 0.8390446731459984,
      "learning_rate": 2e-05,
      "loss": 0.5223187446594239,
      "step": 6900
    },
    {
      "epoch": 0.49551810684833275,
      "grad_norm": 1.0302641431828217,
      "learning_rate": 2e-05,
      "loss": 0.5067597389221191,
      "step": 6910
    },
    {
      "epoch": 0.4962352097525995,
      "grad_norm": 1.749785472297488,
      "learning_rate": 2e-05,
      "loss": 0.5289133071899415,
      "step": 6920
    },
    {
      "epoch": 0.49695231265686624,
      "grad_norm": 1.830297190589475,
      "learning_rate": 2e-05,
      "loss": 0.5032759666442871,
      "step": 6930
    },
    {
      "epoch": 0.49766941556113303,
      "grad_norm": 1.5751779351894304,
      "learning_rate": 2e-05,
      "loss": 0.5416972160339355,
      "step": 6940
    },
    {
      "epoch": 0.4983865184653998,
      "grad_norm": 1.0169878495405602,
      "learning_rate": 2e-05,
      "loss": 0.5245810508728027,
      "step": 6950
    },
    {
      "epoch": 0.4991036213696666,
      "grad_norm": 0.8528500796412259,
      "learning_rate": 2e-05,
      "loss": 0.49609527587890623,
      "step": 6960
    },
    {
      "epoch": 0.4998207242739333,
      "grad_norm": 2.39915238888399,
      "learning_rate": 2e-05,
      "loss": 0.5181064128875732,
      "step": 6970
    },
    {
      "epoch": 0.5005378271782001,
      "grad_norm": 1.0105821920017966,
      "learning_rate": 2e-05,
      "loss": 0.5338099479675293,
      "step": 6980
    },
    {
      "epoch": 0.5012549300824668,
      "grad_norm": 0.8596890659133923,
      "learning_rate": 2e-05,
      "loss": 0.524176025390625,
      "step": 6990
    },
    {
      "epoch": 0.5019720329867337,
      "grad_norm": 1.6774920335224235,
      "learning_rate": 2e-05,
      "loss": 0.5651663780212403,
      "step": 7000
    },
    {
      "epoch": 0.5026891358910004,
      "grad_norm": 1.2824813433677364,
      "learning_rate": 2e-05,
      "loss": 0.5304973125457764,
      "step": 7010
    },
    {
      "epoch": 0.5034062387952671,
      "grad_norm": 1.0912480050976416,
      "learning_rate": 2e-05,
      "loss": 0.523806095123291,
      "step": 7020
    },
    {
      "epoch": 0.5041233416995339,
      "grad_norm": 0.9753795562400791,
      "learning_rate": 2e-05,
      "loss": 0.5432420253753663,
      "step": 7030
    },
    {
      "epoch": 0.5048404446038006,
      "grad_norm": 1.0784199323083734,
      "learning_rate": 2e-05,
      "loss": 0.5509329795837402,
      "step": 7040
    },
    {
      "epoch": 0.5055575475080674,
      "grad_norm": 1.1411974122801951,
      "learning_rate": 2e-05,
      "loss": 0.5119345188140869,
      "step": 7050
    },
    {
      "epoch": 0.5062746504123342,
      "grad_norm": 3.109548174352217,
      "learning_rate": 2e-05,
      "loss": 0.5285029888153077,
      "step": 7060
    },
    {
      "epoch": 0.506991753316601,
      "grad_norm": 1.5279588325457774,
      "learning_rate": 2e-05,
      "loss": 0.5225870132446289,
      "step": 7070
    },
    {
      "epoch": 0.5077088562208677,
      "grad_norm": 12.464259260115439,
      "learning_rate": 2e-05,
      "loss": 0.5130463123321534,
      "step": 7080
    },
    {
      "epoch": 0.5084259591251344,
      "grad_norm": 0.9026112732356519,
      "learning_rate": 2e-05,
      "loss": 0.528814697265625,
      "step": 7090
    },
    {
      "epoch": 0.5091430620294012,
      "grad_norm": 0.9441876939692447,
      "learning_rate": 2e-05,
      "loss": 0.5187839508056641,
      "step": 7100
    },
    {
      "epoch": 0.509860164933668,
      "grad_norm": 0.9788585002323438,
      "learning_rate": 2e-05,
      "loss": 0.5331783294677734,
      "step": 7110
    },
    {
      "epoch": 0.5105772678379348,
      "grad_norm": 0.9053775995475958,
      "learning_rate": 2e-05,
      "loss": 0.49416437149047854,
      "step": 7120
    },
    {
      "epoch": 0.5112943707422015,
      "grad_norm": 0.9608153572729928,
      "learning_rate": 2e-05,
      "loss": 0.49881610870361326,
      "step": 7130
    },
    {
      "epoch": 0.5120114736464683,
      "grad_norm": 0.7365041971680599,
      "learning_rate": 2e-05,
      "loss": 0.5098479747772217,
      "step": 7140
    },
    {
      "epoch": 0.512728576550735,
      "grad_norm": 0.8553553202561308,
      "learning_rate": 2e-05,
      "loss": 0.509791898727417,
      "step": 7150
    },
    {
      "epoch": 0.5134456794550017,
      "grad_norm": 1.1360258630429632,
      "learning_rate": 2e-05,
      "loss": 0.5043924808502197,
      "step": 7160
    },
    {
      "epoch": 0.5141627823592686,
      "grad_norm": 1.0338361109004426,
      "learning_rate": 2e-05,
      "loss": 0.49025697708129884,
      "step": 7170
    },
    {
      "epoch": 0.5148798852635353,
      "grad_norm": 1.2702192422384913,
      "learning_rate": 2e-05,
      "loss": 0.495805549621582,
      "step": 7180
    },
    {
      "epoch": 0.5155969881678021,
      "grad_norm": 1.6352897077341237,
      "learning_rate": 2e-05,
      "loss": 0.5438719749450683,
      "step": 7190
    },
    {
      "epoch": 0.5163140910720688,
      "grad_norm": 3.299169994081323,
      "learning_rate": 2e-05,
      "loss": 0.5200569152832031,
      "step": 7200
    },
    {
      "epoch": 0.5170311939763356,
      "grad_norm": 2.0270945479269242,
      "learning_rate": 2e-05,
      "loss": 0.5344393730163575,
      "step": 7210
    },
    {
      "epoch": 0.5177482968806024,
      "grad_norm": 1.3410388310325445,
      "learning_rate": 2e-05,
      "loss": 0.5024619579315186,
      "step": 7220
    },
    {
      "epoch": 0.5184653997848692,
      "grad_norm": 1.2782127006082338,
      "learning_rate": 2e-05,
      "loss": 0.5013829231262207,
      "step": 7230
    },
    {
      "epoch": 0.5191825026891359,
      "grad_norm": 0.8748896535153174,
      "learning_rate": 2e-05,
      "loss": 0.5218202114105225,
      "step": 7240
    },
    {
      "epoch": 0.5198996055934026,
      "grad_norm": 0.8295091521903556,
      "learning_rate": 2e-05,
      "loss": 0.5159372806549072,
      "step": 7250
    },
    {
      "epoch": 0.5206167084976694,
      "grad_norm": 0.6950104964679813,
      "learning_rate": 2e-05,
      "loss": 0.5316265106201172,
      "step": 7260
    },
    {
      "epoch": 0.5213338114019361,
      "grad_norm": 0.9046564739420331,
      "learning_rate": 2e-05,
      "loss": 0.47831058502197266,
      "step": 7270
    },
    {
      "epoch": 0.522050914306203,
      "grad_norm": 1.3257547515093593,
      "learning_rate": 2e-05,
      "loss": 0.49481678009033203,
      "step": 7280
    },
    {
      "epoch": 0.5227680172104697,
      "grad_norm": 1.9039208293313872,
      "learning_rate": 2e-05,
      "loss": 0.5271610260009766,
      "step": 7290
    },
    {
      "epoch": 0.5234851201147365,
      "grad_norm": 1.3322296957009778,
      "learning_rate": 2e-05,
      "loss": 0.5051034927368164,
      "step": 7300
    },
    {
      "epoch": 0.5242022230190032,
      "grad_norm": 0.9506793367220124,
      "learning_rate": 2e-05,
      "loss": 0.5123958110809326,
      "step": 7310
    },
    {
      "epoch": 0.52491932592327,
      "grad_norm": 0.9011085657169259,
      "learning_rate": 2e-05,
      "loss": 0.5068558692932129,
      "step": 7320
    },
    {
      "epoch": 0.5256364288275368,
      "grad_norm": 1.8466839353177948,
      "learning_rate": 2e-05,
      "loss": 0.5306342124938965,
      "step": 7330
    },
    {
      "epoch": 0.5263535317318035,
      "grad_norm": 1.257246948391457,
      "learning_rate": 2e-05,
      "loss": 0.5284415245056152,
      "step": 7340
    },
    {
      "epoch": 0.5270706346360703,
      "grad_norm": 0.696834134437849,
      "learning_rate": 2e-05,
      "loss": 0.48610992431640626,
      "step": 7350
    },
    {
      "epoch": 0.527787737540337,
      "grad_norm": 1.5171278980866398,
      "learning_rate": 2e-05,
      "loss": 0.5055984497070313,
      "step": 7360
    },
    {
      "epoch": 0.5285048404446038,
      "grad_norm": 0.6974250806201853,
      "learning_rate": 2e-05,
      "loss": 0.5224874496459961,
      "step": 7370
    },
    {
      "epoch": 0.5292219433488705,
      "grad_norm": 1.5851366369317252,
      "learning_rate": 2e-05,
      "loss": 0.4935168743133545,
      "step": 7380
    },
    {
      "epoch": 0.5299390462531374,
      "grad_norm": 0.6118433942956579,
      "learning_rate": 2e-05,
      "loss": 0.5112913131713868,
      "step": 7390
    },
    {
      "epoch": 0.5306561491574041,
      "grad_norm": 1.3762564102166295,
      "learning_rate": 2e-05,
      "loss": 0.49964213371276855,
      "step": 7400
    },
    {
      "epoch": 0.5313732520616709,
      "grad_norm": 0.7357629887079652,
      "learning_rate": 2e-05,
      "loss": 0.5217706203460694,
      "step": 7410
    },
    {
      "epoch": 0.5320903549659376,
      "grad_norm": 0.8916506257922856,
      "learning_rate": 2e-05,
      "loss": 0.49975385665893557,
      "step": 7420
    },
    {
      "epoch": 0.5328074578702043,
      "grad_norm": 1.3928714632514048,
      "learning_rate": 2e-05,
      "loss": 0.5170638561248779,
      "step": 7430
    },
    {
      "epoch": 0.5335245607744712,
      "grad_norm": 5.899482335256363,
      "learning_rate": 2e-05,
      "loss": 0.5324769973754883,
      "step": 7440
    },
    {
      "epoch": 0.5342416636787379,
      "grad_norm": 2.034992476722796,
      "learning_rate": 2e-05,
      "loss": 0.5238707065582275,
      "step": 7450
    },
    {
      "epoch": 0.5349587665830047,
      "grad_norm": 0.8635145795734963,
      "learning_rate": 2e-05,
      "loss": 0.48725261688232424,
      "step": 7460
    },
    {
      "epoch": 0.5356758694872714,
      "grad_norm": 0.8808201419595689,
      "learning_rate": 2e-05,
      "loss": 0.5027401924133301,
      "step": 7470
    },
    {
      "epoch": 0.5363929723915382,
      "grad_norm": 4.651219232309251,
      "learning_rate": 2e-05,
      "loss": 0.5375476837158203,
      "step": 7480
    },
    {
      "epoch": 0.5371100752958049,
      "grad_norm": 1.0974645040762774,
      "learning_rate": 2e-05,
      "loss": 0.5213850975036621,
      "step": 7490
    },
    {
      "epoch": 0.5378271782000718,
      "grad_norm": 5.290516209289302,
      "learning_rate": 2e-05,
      "loss": 0.5264504909515381,
      "step": 7500
    },
    {
      "epoch": 0.5385442811043385,
      "grad_norm": 1.4023091759603943,
      "learning_rate": 2e-05,
      "loss": 0.5041522026062012,
      "step": 7510
    },
    {
      "epoch": 0.5392613840086052,
      "grad_norm": 0.9457804864658033,
      "learning_rate": 2e-05,
      "loss": 0.5341561317443848,
      "step": 7520
    },
    {
      "epoch": 0.539978486912872,
      "grad_norm": 0.9309558207152469,
      "learning_rate": 2e-05,
      "loss": 0.5168865203857422,
      "step": 7530
    },
    {
      "epoch": 0.5406955898171387,
      "grad_norm": 0.6832902365018415,
      "learning_rate": 2e-05,
      "loss": 0.5349706172943115,
      "step": 7540
    },
    {
      "epoch": 0.5414126927214056,
      "grad_norm": 0.9937595185326706,
      "learning_rate": 2e-05,
      "loss": 0.5065614700317382,
      "step": 7550
    },
    {
      "epoch": 0.5421297956256723,
      "grad_norm": 0.9485435096584648,
      "learning_rate": 2e-05,
      "loss": 0.49179911613464355,
      "step": 7560
    },
    {
      "epoch": 0.5428468985299391,
      "grad_norm": 1.0016215273466604,
      "learning_rate": 2e-05,
      "loss": 0.522728157043457,
      "step": 7570
    },
    {
      "epoch": 0.5435640014342058,
      "grad_norm": 0.9004053741262238,
      "learning_rate": 2e-05,
      "loss": 0.5229884147644043,
      "step": 7580
    },
    {
      "epoch": 0.5442811043384725,
      "grad_norm": 0.8252419253657323,
      "learning_rate": 2e-05,
      "loss": 0.4827853202819824,
      "step": 7590
    },
    {
      "epoch": 0.5449982072427393,
      "grad_norm": 1.9477813320744883,
      "learning_rate": 2e-05,
      "loss": 0.5148584365844726,
      "step": 7600
    },
    {
      "epoch": 0.5457153101470061,
      "grad_norm": 1.3892071610479437,
      "learning_rate": 2e-05,
      "loss": 0.49146389961242676,
      "step": 7610
    },
    {
      "epoch": 0.5464324130512729,
      "grad_norm": 4.564234292006207,
      "learning_rate": 2e-05,
      "loss": 0.522843885421753,
      "step": 7620
    },
    {
      "epoch": 0.5471495159555396,
      "grad_norm": 1.1440021558644677,
      "learning_rate": 2e-05,
      "loss": 0.4963046073913574,
      "step": 7630
    },
    {
      "epoch": 0.5478666188598064,
      "grad_norm": 1.1732762086558663,
      "learning_rate": 2e-05,
      "loss": 0.521145761013031,
      "step": 7640
    },
    {
      "epoch": 0.5485837217640731,
      "grad_norm": 0.9998304727322452,
      "learning_rate": 2e-05,
      "loss": 0.5430165767669678,
      "step": 7650
    },
    {
      "epoch": 0.54930082466834,
      "grad_norm": 1.2728917475618957,
      "learning_rate": 2e-05,
      "loss": 0.5263105392456054,
      "step": 7660
    },
    {
      "epoch": 0.5500179275726067,
      "grad_norm": 0.972708224095732,
      "learning_rate": 2e-05,
      "loss": 0.486739444732666,
      "step": 7670
    },
    {
      "epoch": 0.5507350304768734,
      "grad_norm": 0.7721107884876419,
      "learning_rate": 2e-05,
      "loss": 0.48564634323120115,
      "step": 7680
    },
    {
      "epoch": 0.5514521333811402,
      "grad_norm": 1.065996559010077,
      "learning_rate": 2e-05,
      "loss": 0.5282321453094483,
      "step": 7690
    },
    {
      "epoch": 0.5521692362854069,
      "grad_norm": 0.6827021352688556,
      "learning_rate": 2e-05,
      "loss": 0.5041511535644532,
      "step": 7700
    },
    {
      "epoch": 0.5528863391896737,
      "grad_norm": 0.7369540350311007,
      "learning_rate": 2e-05,
      "loss": 0.5134336471557617,
      "step": 7710
    },
    {
      "epoch": 0.5536034420939405,
      "grad_norm": 3.260313028521785,
      "learning_rate": 2e-05,
      "loss": 0.5203777313232422,
      "step": 7720
    },
    {
      "epoch": 0.5543205449982073,
      "grad_norm": 0.9115660708156116,
      "learning_rate": 2e-05,
      "loss": 0.5566596984863281,
      "step": 7730
    },
    {
      "epoch": 0.555037647902474,
      "grad_norm": 1.6982028492697052,
      "learning_rate": 2e-05,
      "loss": 0.5098933219909668,
      "step": 7740
    },
    {
      "epoch": 0.5557547508067407,
      "grad_norm": 4.583590238770378,
      "learning_rate": 2e-05,
      "loss": 0.5083906650543213,
      "step": 7750
    },
    {
      "epoch": 0.5564718537110075,
      "grad_norm": 1.0388504023854306,
      "learning_rate": 2e-05,
      "loss": 0.5206740379333497,
      "step": 7760
    },
    {
      "epoch": 0.5571889566152743,
      "grad_norm": 0.7622699484127513,
      "learning_rate": 2e-05,
      "loss": 0.5139105796813965,
      "step": 7770
    },
    {
      "epoch": 0.5579060595195411,
      "grad_norm": 0.8765494414270234,
      "learning_rate": 2e-05,
      "loss": 0.5242534160614014,
      "step": 7780
    },
    {
      "epoch": 0.5586231624238078,
      "grad_norm": 0.8255786077671922,
      "learning_rate": 2e-05,
      "loss": 0.5042855739593506,
      "step": 7790
    },
    {
      "epoch": 0.5593402653280746,
      "grad_norm": 1.0829905417642174,
      "learning_rate": 2e-05,
      "loss": 0.5081568241119385,
      "step": 7800
    },
    {
      "epoch": 0.5600573682323413,
      "grad_norm": 4.406054011058295,
      "learning_rate": 2e-05,
      "loss": 0.5220559120178223,
      "step": 7810
    },
    {
      "epoch": 0.560774471136608,
      "grad_norm": 0.8796950447499373,
      "learning_rate": 2e-05,
      "loss": 0.5137046813964844,
      "step": 7820
    },
    {
      "epoch": 0.5614915740408749,
      "grad_norm": 0.7312179311030306,
      "learning_rate": 2e-05,
      "loss": 0.5103032112121582,
      "step": 7830
    },
    {
      "epoch": 0.5622086769451416,
      "grad_norm": 0.9949512607134006,
      "learning_rate": 2e-05,
      "loss": 0.5025225639343261,
      "step": 7840
    },
    {
      "epoch": 0.5629257798494084,
      "grad_norm": 1.0118562207393023,
      "learning_rate": 2e-05,
      "loss": 0.5089486122131348,
      "step": 7850
    },
    {
      "epoch": 0.5636428827536751,
      "grad_norm": 2.3279170275209573,
      "learning_rate": 2e-05,
      "loss": 0.5237641334533691,
      "step": 7860
    },
    {
      "epoch": 0.5643599856579419,
      "grad_norm": 12.188185950632674,
      "learning_rate": 2e-05,
      "loss": 0.5093648433685303,
      "step": 7870
    },
    {
      "epoch": 0.5650770885622087,
      "grad_norm": 1.4227492481092305,
      "learning_rate": 2e-05,
      "loss": 0.6008241176605225,
      "step": 7880
    },
    {
      "epoch": 0.5657941914664755,
      "grad_norm": 2.31391296099499,
      "learning_rate": 2e-05,
      "loss": 0.5128987312316895,
      "step": 7890
    },
    {
      "epoch": 0.5665112943707422,
      "grad_norm": 2.3168952220806545,
      "learning_rate": 2e-05,
      "loss": 0.5019460201263428,
      "step": 7900
    },
    {
      "epoch": 0.567228397275009,
      "grad_norm": 0.8229746394855947,
      "learning_rate": 2e-05,
      "loss": 0.5198930263519287,
      "step": 7910
    },
    {
      "epoch": 0.5679455001792757,
      "grad_norm": 0.9090512949174614,
      "learning_rate": 2e-05,
      "loss": 0.5094467163085937,
      "step": 7920
    },
    {
      "epoch": 0.5686626030835424,
      "grad_norm": 0.786127998758279,
      "learning_rate": 2e-05,
      "loss": 0.5158935546875,
      "step": 7930
    },
    {
      "epoch": 0.5693797059878093,
      "grad_norm": 1.2962892213633685,
      "learning_rate": 2e-05,
      "loss": 0.4970674514770508,
      "step": 7940
    },
    {
      "epoch": 0.570096808892076,
      "grad_norm": 0.9676061436759387,
      "learning_rate": 2e-05,
      "loss": 0.5276697635650635,
      "step": 7950
    },
    {
      "epoch": 0.5708139117963428,
      "grad_norm": 1.3575948900488068,
      "learning_rate": 2e-05,
      "loss": 0.49666848182678225,
      "step": 7960
    },
    {
      "epoch": 0.5715310147006095,
      "grad_norm": 0.9657952790027661,
      "learning_rate": 2e-05,
      "loss": 0.48691678047180176,
      "step": 7970
    },
    {
      "epoch": 0.5722481176048763,
      "grad_norm": 4.210234353644016,
      "learning_rate": 2e-05,
      "loss": 0.5144421577453613,
      "step": 7980
    },
    {
      "epoch": 0.5729652205091431,
      "grad_norm": 0.7343804632585349,
      "learning_rate": 2e-05,
      "loss": 0.5325693130493164,
      "step": 7990
    },
    {
      "epoch": 0.5736823234134099,
      "grad_norm": 0.7332095660219121,
      "learning_rate": 2e-05,
      "loss": 0.5109401702880859,
      "step": 8000
    },
    {
      "epoch": 0.5743994263176766,
      "grad_norm": 1.3533666110708917,
      "learning_rate": 2e-05,
      "loss": 0.5244138717651368,
      "step": 8010
    },
    {
      "epoch": 0.5751165292219433,
      "grad_norm": 0.7393689343841967,
      "learning_rate": 2e-05,
      "loss": 0.48618316650390625,
      "step": 8020
    },
    {
      "epoch": 0.5758336321262101,
      "grad_norm": 1.2841745076140476,
      "learning_rate": 2e-05,
      "loss": 0.5222994804382324,
      "step": 8030
    },
    {
      "epoch": 0.5765507350304768,
      "grad_norm": 0.7242415179361174,
      "learning_rate": 2e-05,
      "loss": 0.5313247203826904,
      "step": 8040
    },
    {
      "epoch": 0.5772678379347437,
      "grad_norm": 1.0849267468002497,
      "learning_rate": 2e-05,
      "loss": 0.5093178749084473,
      "step": 8050
    },
    {
      "epoch": 0.5779849408390104,
      "grad_norm": 0.8170432193359219,
      "learning_rate": 2e-05,
      "loss": 0.4812446594238281,
      "step": 8060
    },
    {
      "epoch": 0.5787020437432772,
      "grad_norm": 0.7518239661063011,
      "learning_rate": 2e-05,
      "loss": 0.504496955871582,
      "step": 8070
    },
    {
      "epoch": 0.5794191466475439,
      "grad_norm": 0.8246612356647478,
      "learning_rate": 2e-05,
      "loss": 0.5147820472717285,
      "step": 8080
    },
    {
      "epoch": 0.5801362495518106,
      "grad_norm": 0.6998674170544454,
      "learning_rate": 2e-05,
      "loss": 0.49538631439208985,
      "step": 8090
    },
    {
      "epoch": 0.5808533524560775,
      "grad_norm": 0.8868975935814853,
      "learning_rate": 2e-05,
      "loss": 0.5237655639648438,
      "step": 8100
    },
    {
      "epoch": 0.5815704553603442,
      "grad_norm": 1.9692278277923991,
      "learning_rate": 2e-05,
      "loss": 0.5237738132476807,
      "step": 8110
    },
    {
      "epoch": 0.582287558264611,
      "grad_norm": 0.9980411033785309,
      "learning_rate": 2e-05,
      "loss": 0.5062626838684082,
      "step": 8120
    },
    {
      "epoch": 0.5830046611688777,
      "grad_norm": 1.0317872553088179,
      "learning_rate": 2e-05,
      "loss": 0.50491943359375,
      "step": 8130
    },
    {
      "epoch": 0.5837217640731445,
      "grad_norm": 1.106101669140827,
      "learning_rate": 2e-05,
      "loss": 0.48985910415649414,
      "step": 8140
    },
    {
      "epoch": 0.5844388669774112,
      "grad_norm": 0.6963655013373172,
      "learning_rate": 2e-05,
      "loss": 0.5428201198577881,
      "step": 8150
    },
    {
      "epoch": 0.5851559698816781,
      "grad_norm": 0.751484341754584,
      "learning_rate": 2e-05,
      "loss": 0.5123166561126709,
      "step": 8160
    },
    {
      "epoch": 0.5858730727859448,
      "grad_norm": 0.7706317154840989,
      "learning_rate": 2e-05,
      "loss": 0.5125321388244629,
      "step": 8170
    },
    {
      "epoch": 0.5865901756902115,
      "grad_norm": 0.9442566772430051,
      "learning_rate": 2e-05,
      "loss": 0.5043001174926758,
      "step": 8180
    },
    {
      "epoch": 0.5873072785944783,
      "grad_norm": 0.8929874454646707,
      "learning_rate": 2e-05,
      "loss": 0.49517574310302737,
      "step": 8190
    },
    {
      "epoch": 0.588024381498745,
      "grad_norm": 0.8657883133458075,
      "learning_rate": 2e-05,
      "loss": 0.5130547523498535,
      "step": 8200
    },
    {
      "epoch": 0.5887414844030119,
      "grad_norm": 0.7378473504983424,
      "learning_rate": 2e-05,
      "loss": 0.507685375213623,
      "step": 8210
    },
    {
      "epoch": 0.5894585873072786,
      "grad_norm": 0.7747328622279064,
      "learning_rate": 2e-05,
      "loss": 0.49251570701599123,
      "step": 8220
    },
    {
      "epoch": 0.5901756902115454,
      "grad_norm": 0.7886318420007136,
      "learning_rate": 2e-05,
      "loss": 0.49441065788269045,
      "step": 8230
    },
    {
      "epoch": 0.5908927931158121,
      "grad_norm": 1.4126474087698304,
      "learning_rate": 2e-05,
      "loss": 0.5141373634338379,
      "step": 8240
    },
    {
      "epoch": 0.5916098960200789,
      "grad_norm": 1.2368553589621254,
      "learning_rate": 2e-05,
      "loss": 0.5007548332214355,
      "step": 8250
    },
    {
      "epoch": 0.5923269989243456,
      "grad_norm": 38.17068528048916,
      "learning_rate": 2e-05,
      "loss": 0.5102075099945068,
      "step": 8260
    },
    {
      "epoch": 0.5930441018286124,
      "grad_norm": 1.1181208478685152,
      "learning_rate": 2e-05,
      "loss": 0.5245469570159912,
      "step": 8270
    },
    {
      "epoch": 0.5937612047328792,
      "grad_norm": 2.036809421558534,
      "learning_rate": 2e-05,
      "loss": 0.5250093460083007,
      "step": 8280
    },
    {
      "epoch": 0.5944783076371459,
      "grad_norm": 1.613448301954448,
      "learning_rate": 2e-05,
      "loss": 0.5226828575134277,
      "step": 8290
    },
    {
      "epoch": 0.5951954105414127,
      "grad_norm": 0.9282201908460104,
      "learning_rate": 2e-05,
      "loss": 0.5120925903320312,
      "step": 8300
    },
    {
      "epoch": 0.5959125134456794,
      "grad_norm": 0.9455462073016642,
      "learning_rate": 2e-05,
      "loss": 0.5144097328186035,
      "step": 8310
    },
    {
      "epoch": 0.5966296163499463,
      "grad_norm": 0.8538515695707133,
      "learning_rate": 2e-05,
      "loss": 0.537442684173584,
      "step": 8320
    },
    {
      "epoch": 0.597346719254213,
      "grad_norm": 0.8110814199967968,
      "learning_rate": 2e-05,
      "loss": 0.48949995040893557,
      "step": 8330
    },
    {
      "epoch": 0.5980638221584798,
      "grad_norm": 0.7679714994025922,
      "learning_rate": 2e-05,
      "loss": 0.530856704711914,
      "step": 8340
    },
    {
      "epoch": 0.5987809250627465,
      "grad_norm": 0.8177380459932198,
      "learning_rate": 2e-05,
      "loss": 0.5251891136169433,
      "step": 8350
    },
    {
      "epoch": 0.5994980279670132,
      "grad_norm": 0.7213702596901129,
      "learning_rate": 2e-05,
      "loss": 0.5145583629608155,
      "step": 8360
    },
    {
      "epoch": 0.60021513087128,
      "grad_norm": 0.7702148189304847,
      "learning_rate": 2e-05,
      "loss": 0.4903076648712158,
      "step": 8370
    },
    {
      "epoch": 0.6009322337755468,
      "grad_norm": 0.8855187208083997,
      "learning_rate": 2e-05,
      "loss": 0.5117664337158203,
      "step": 8380
    },
    {
      "epoch": 0.6016493366798136,
      "grad_norm": 0.8006800293507436,
      "learning_rate": 2e-05,
      "loss": 0.47857189178466797,
      "step": 8390
    },
    {
      "epoch": 0.6023664395840803,
      "grad_norm": 0.6492403252874012,
      "learning_rate": 2e-05,
      "loss": 0.5202676773071289,
      "step": 8400
    },
    {
      "epoch": 0.6030835424883471,
      "grad_norm": 1.0115255588774974,
      "learning_rate": 2e-05,
      "loss": 0.508674430847168,
      "step": 8410
    },
    {
      "epoch": 0.6038006453926138,
      "grad_norm": 0.9955751471927611,
      "learning_rate": 2e-05,
      "loss": 0.49298815727233886,
      "step": 8420
    },
    {
      "epoch": 0.6045177482968807,
      "grad_norm": 0.7689178601884348,
      "learning_rate": 2e-05,
      "loss": 0.5167996883392334,
      "step": 8430
    },
    {
      "epoch": 0.6052348512011474,
      "grad_norm": 0.9920047763227099,
      "learning_rate": 2e-05,
      "loss": 0.46744389533996583,
      "step": 8440
    },
    {
      "epoch": 0.6059519541054141,
      "grad_norm": 0.728197986857094,
      "learning_rate": 2e-05,
      "loss": 0.5188340187072754,
      "step": 8450
    },
    {
      "epoch": 0.6066690570096809,
      "grad_norm": 1.1620010411662005,
      "learning_rate": 2e-05,
      "loss": 0.49208831787109375,
      "step": 8460
    },
    {
      "epoch": 0.6073861599139476,
      "grad_norm": 1.573354268222793,
      "learning_rate": 2e-05,
      "loss": 0.5036312103271484,
      "step": 8470
    },
    {
      "epoch": 0.6081032628182144,
      "grad_norm": 0.885346236659806,
      "learning_rate": 2e-05,
      "loss": 0.5107982635498047,
      "step": 8480
    },
    {
      "epoch": 0.6088203657224812,
      "grad_norm": 2.4011662117375177,
      "learning_rate": 2e-05,
      "loss": 0.49164018630981443,
      "step": 8490
    },
    {
      "epoch": 0.609537468626748,
      "grad_norm": 0.9966987655034003,
      "learning_rate": 2e-05,
      "loss": 0.5047776699066162,
      "step": 8500
    },
    {
      "epoch": 0.6102545715310147,
      "grad_norm": 1.9535926758163875,
      "learning_rate": 2e-05,
      "loss": 0.48383293151855467,
      "step": 8510
    },
    {
      "epoch": 0.6109716744352814,
      "grad_norm": 0.6839896419636821,
      "learning_rate": 2e-05,
      "loss": 0.49138822555541994,
      "step": 8520
    },
    {
      "epoch": 0.6116887773395482,
      "grad_norm": 3.491092950754823,
      "learning_rate": 2e-05,
      "loss": 0.5055727005004883,
      "step": 8530
    },
    {
      "epoch": 0.612405880243815,
      "grad_norm": 0.9043228262180714,
      "learning_rate": 2e-05,
      "loss": 0.5192901611328125,
      "step": 8540
    },
    {
      "epoch": 0.6131229831480818,
      "grad_norm": 0.8683413289315278,
      "learning_rate": 2e-05,
      "loss": 0.5148396968841553,
      "step": 8550
    },
    {
      "epoch": 0.6138400860523485,
      "grad_norm": 5.275204291625261,
      "learning_rate": 2e-05,
      "loss": 0.49720091819763185,
      "step": 8560
    },
    {
      "epoch": 0.6145571889566153,
      "grad_norm": 1.2444801175747138,
      "learning_rate": 2e-05,
      "loss": 0.5502876281738281,
      "step": 8570
    },
    {
      "epoch": 0.615274291860882,
      "grad_norm": 0.8301310809348096,
      "learning_rate": 2e-05,
      "loss": 0.5137484550476075,
      "step": 8580
    },
    {
      "epoch": 0.6159913947651487,
      "grad_norm": 1.5708660325923756,
      "learning_rate": 2e-05,
      "loss": 0.526588773727417,
      "step": 8590
    },
    {
      "epoch": 0.6167084976694156,
      "grad_norm": 1.280391061409431,
      "learning_rate": 2e-05,
      "loss": 0.5236782550811767,
      "step": 8600
    },
    {
      "epoch": 0.6174256005736823,
      "grad_norm": 1.3287421697435169,
      "learning_rate": 2e-05,
      "loss": 0.5265653133392334,
      "step": 8610
    },
    {
      "epoch": 0.6181427034779491,
      "grad_norm": 0.8878236149736098,
      "learning_rate": 2e-05,
      "loss": 0.518764591217041,
      "step": 8620
    },
    {
      "epoch": 0.6188598063822158,
      "grad_norm": 1.0725410785913,
      "learning_rate": 2e-05,
      "loss": 0.5056464672088623,
      "step": 8630
    },
    {
      "epoch": 0.6195769092864826,
      "grad_norm": 0.8964222903178919,
      "learning_rate": 2e-05,
      "loss": 0.5224818706512451,
      "step": 8640
    },
    {
      "epoch": 0.6202940121907494,
      "grad_norm": 1.450807706266201,
      "learning_rate": 2e-05,
      "loss": 0.48691754341125487,
      "step": 8650
    },
    {
      "epoch": 0.6210111150950162,
      "grad_norm": 1.3247263832686376,
      "learning_rate": 2e-05,
      "loss": 0.5345084190368652,
      "step": 8660
    },
    {
      "epoch": 0.6217282179992829,
      "grad_norm": 0.7312597321259491,
      "learning_rate": 2e-05,
      "loss": 0.5219978332519531,
      "step": 8670
    },
    {
      "epoch": 0.6224453209035496,
      "grad_norm": 0.9111732271067247,
      "learning_rate": 2e-05,
      "loss": 0.5102910995483398,
      "step": 8680
    },
    {
      "epoch": 0.6231624238078164,
      "grad_norm": 4.542120201418321,
      "learning_rate": 2e-05,
      "loss": 0.534779167175293,
      "step": 8690
    },
    {
      "epoch": 0.6238795267120831,
      "grad_norm": 0.9421954051297252,
      "learning_rate": 2e-05,
      "loss": 0.4998478889465332,
      "step": 8700
    },
    {
      "epoch": 0.62459662961635,
      "grad_norm": 0.8702701433281207,
      "learning_rate": 2e-05,
      "loss": 0.5007111549377441,
      "step": 8710
    },
    {
      "epoch": 0.6253137325206167,
      "grad_norm": 1.276237543411455,
      "learning_rate": 2e-05,
      "loss": 0.5254611968994141,
      "step": 8720
    },
    {
      "epoch": 0.6260308354248835,
      "grad_norm": 1.0656149597839066,
      "learning_rate": 2e-05,
      "loss": 0.5306049346923828,
      "step": 8730
    },
    {
      "epoch": 0.6267479383291502,
      "grad_norm": 1.048401973851732,
      "learning_rate": 2e-05,
      "loss": 0.5383347988128662,
      "step": 8740
    },
    {
      "epoch": 0.627465041233417,
      "grad_norm": 1.2613197712314528,
      "learning_rate": 2e-05,
      "loss": 0.49928922653198243,
      "step": 8750
    },
    {
      "epoch": 0.6281821441376838,
      "grad_norm": 0.7471145764540091,
      "learning_rate": 2e-05,
      "loss": 0.5252915382385254,
      "step": 8760
    },
    {
      "epoch": 0.6288992470419505,
      "grad_norm": 0.7821314241966657,
      "learning_rate": 2e-05,
      "loss": 0.521507453918457,
      "step": 8770
    },
    {
      "epoch": 0.6296163499462173,
      "grad_norm": 0.8006770351480943,
      "learning_rate": 2e-05,
      "loss": 0.513083028793335,
      "step": 8780
    },
    {
      "epoch": 0.630333452850484,
      "grad_norm": 0.8898485218566742,
      "learning_rate": 2e-05,
      "loss": 0.5292594909667969,
      "step": 8790
    },
    {
      "epoch": 0.6310505557547508,
      "grad_norm": 0.7447903002406171,
      "learning_rate": 2e-05,
      "loss": 0.4957902908325195,
      "step": 8800
    },
    {
      "epoch": 0.6317676586590175,
      "grad_norm": 1.117382877056016,
      "learning_rate": 2e-05,
      "loss": 0.5322527408599853,
      "step": 8810
    },
    {
      "epoch": 0.6324847615632844,
      "grad_norm": 0.941094638279192,
      "learning_rate": 2e-05,
      "loss": 0.49531707763671873,
      "step": 8820
    },
    {
      "epoch": 0.6332018644675511,
      "grad_norm": 4.391983986172127,
      "learning_rate": 2e-05,
      "loss": 0.5224556922912598,
      "step": 8830
    },
    {
      "epoch": 0.6339189673718179,
      "grad_norm": 1.4298723795589916,
      "learning_rate": 2e-05,
      "loss": 0.4993555545806885,
      "step": 8840
    },
    {
      "epoch": 0.6346360702760846,
      "grad_norm": 0.8433548093220169,
      "learning_rate": 2e-05,
      "loss": 0.5471954345703125,
      "step": 8850
    },
    {
      "epoch": 0.6353531731803513,
      "grad_norm": 0.7442636681141717,
      "learning_rate": 2e-05,
      "loss": 0.5041863918304443,
      "step": 8860
    },
    {
      "epoch": 0.6360702760846182,
      "grad_norm": 0.7320458874748506,
      "learning_rate": 2e-05,
      "loss": 0.5256603240966797,
      "step": 8870
    },
    {
      "epoch": 0.6367873789888849,
      "grad_norm": 2.598450594216475,
      "learning_rate": 2e-05,
      "loss": 0.5149674415588379,
      "step": 8880
    },
    {
      "epoch": 0.6375044818931517,
      "grad_norm": 1.08512566734015,
      "learning_rate": 2e-05,
      "loss": 0.49846601486206055,
      "step": 8890
    },
    {
      "epoch": 0.6382215847974184,
      "grad_norm": 1.2388498929085596,
      "learning_rate": 2e-05,
      "loss": 0.4951361656188965,
      "step": 8900
    },
    {
      "epoch": 0.6389386877016852,
      "grad_norm": 2.4333193835385725,
      "learning_rate": 2e-05,
      "loss": 0.4926475524902344,
      "step": 8910
    },
    {
      "epoch": 0.6396557906059519,
      "grad_norm": 0.8392175218474921,
      "learning_rate": 2e-05,
      "loss": 0.5210039138793945,
      "step": 8920
    },
    {
      "epoch": 0.6403728935102188,
      "grad_norm": 0.7436305144160047,
      "learning_rate": 2e-05,
      "loss": 0.530491304397583,
      "step": 8930
    },
    {
      "epoch": 0.6410899964144855,
      "grad_norm": 0.8401235373266993,
      "learning_rate": 2e-05,
      "loss": 0.5219184398651123,
      "step": 8940
    },
    {
      "epoch": 0.6418070993187522,
      "grad_norm": 0.9116912218359972,
      "learning_rate": 2e-05,
      "loss": 0.5039981365203857,
      "step": 8950
    },
    {
      "epoch": 0.642524202223019,
      "grad_norm": 0.7997906078438934,
      "learning_rate": 2e-05,
      "loss": 0.510408592224121,
      "step": 8960
    },
    {
      "epoch": 0.6432413051272857,
      "grad_norm": 0.6851477901786045,
      "learning_rate": 2e-05,
      "loss": 0.4914252281188965,
      "step": 8970
    },
    {
      "epoch": 0.6439584080315526,
      "grad_norm": 0.7323151231509668,
      "learning_rate": 2e-05,
      "loss": 0.49104900360107423,
      "step": 8980
    },
    {
      "epoch": 0.6446755109358193,
      "grad_norm": 0.7704240687540154,
      "learning_rate": 2e-05,
      "loss": 0.5074891090393067,
      "step": 8990
    },
    {
      "epoch": 0.6453926138400861,
      "grad_norm": 0.7891282501354971,
      "learning_rate": 2e-05,
      "loss": 0.5147742748260498,
      "step": 9000
    },
    {
      "epoch": 0.6461097167443528,
      "grad_norm": 0.6296907294128203,
      "learning_rate": 2e-05,
      "loss": 0.49088010787963865,
      "step": 9010
    },
    {
      "epoch": 0.6468268196486195,
      "grad_norm": 0.7181654788564282,
      "learning_rate": 2e-05,
      "loss": 0.4962336540222168,
      "step": 9020
    },
    {
      "epoch": 0.6475439225528863,
      "grad_norm": 1.0570940064602183,
      "learning_rate": 2e-05,
      "loss": 0.5051019191741943,
      "step": 9030
    },
    {
      "epoch": 0.6482610254571531,
      "grad_norm": 4.38953069893088,
      "learning_rate": 2e-05,
      "loss": 0.47617206573486326,
      "step": 9040
    },
    {
      "epoch": 0.6489781283614199,
      "grad_norm": 0.7689885788636343,
      "learning_rate": 2e-05,
      "loss": 0.5094996452331543,
      "step": 9050
    },
    {
      "epoch": 0.6496952312656866,
      "grad_norm": 0.8400529105505269,
      "learning_rate": 2e-05,
      "loss": 0.509449815750122,
      "step": 9060
    },
    {
      "epoch": 0.6504123341699534,
      "grad_norm": 3.2720212577160375,
      "learning_rate": 2e-05,
      "loss": 0.5192841529846192,
      "step": 9070
    },
    {
      "epoch": 0.6511294370742201,
      "grad_norm": 1.8280562420883482,
      "learning_rate": 2e-05,
      "loss": 0.4834700584411621,
      "step": 9080
    },
    {
      "epoch": 0.651846539978487,
      "grad_norm": 0.9879149418813111,
      "learning_rate": 2e-05,
      "loss": 0.49909520149230957,
      "step": 9090
    },
    {
      "epoch": 0.6525636428827537,
      "grad_norm": 1.2966257490752022,
      "learning_rate": 2e-05,
      "loss": 0.5031177043914795,
      "step": 9100
    },
    {
      "epoch": 0.6532807457870204,
      "grad_norm": 1.0227719422986221,
      "learning_rate": 2e-05,
      "loss": 0.5242321968078614,
      "step": 9110
    },
    {
      "epoch": 0.6539978486912872,
      "grad_norm": 1.2783167948537857,
      "learning_rate": 2e-05,
      "loss": 0.5188937664031983,
      "step": 9120
    },
    {
      "epoch": 0.6547149515955539,
      "grad_norm": 0.8013983098822107,
      "learning_rate": 2e-05,
      "loss": 0.48288631439208984,
      "step": 9130
    },
    {
      "epoch": 0.6554320544998207,
      "grad_norm": 0.8379215331809486,
      "learning_rate": 2e-05,
      "loss": 0.5002352714538574,
      "step": 9140
    },
    {
      "epoch": 0.6561491574040875,
      "grad_norm": 1.268743000713838,
      "learning_rate": 2e-05,
      "loss": 0.5516677856445312,
      "step": 9150
    },
    {
      "epoch": 0.6568662603083543,
      "grad_norm": 0.8131322635643912,
      "learning_rate": 2e-05,
      "loss": 0.5267547607421875,
      "step": 9160
    },
    {
      "epoch": 0.657583363212621,
      "grad_norm": 1.7829761093686383,
      "learning_rate": 2e-05,
      "loss": 0.5247080802917481,
      "step": 9170
    },
    {
      "epoch": 0.6583004661168878,
      "grad_norm": 1.5817074940445985,
      "learning_rate": 2e-05,
      "loss": 0.5125644207000732,
      "step": 9180
    },
    {
      "epoch": 0.6590175690211545,
      "grad_norm": 0.7726797130561259,
      "learning_rate": 2e-05,
      "loss": 0.5072769165039063,
      "step": 9190
    },
    {
      "epoch": 0.6597346719254213,
      "grad_norm": 1.330601346544019,
      "learning_rate": 2e-05,
      "loss": 0.4813969612121582,
      "step": 9200
    },
    {
      "epoch": 0.6604517748296881,
      "grad_norm": 0.8031744188181617,
      "learning_rate": 2e-05,
      "loss": 0.49367594718933105,
      "step": 9210
    },
    {
      "epoch": 0.6611688777339548,
      "grad_norm": 0.7008284371263963,
      "learning_rate": 2e-05,
      "loss": 0.500235652923584,
      "step": 9220
    },
    {
      "epoch": 0.6618859806382216,
      "grad_norm": 0.9617366839254453,
      "learning_rate": 2e-05,
      "loss": 0.5263199806213379,
      "step": 9230
    },
    {
      "epoch": 0.6626030835424883,
      "grad_norm": 1.3917781262223823,
      "learning_rate": 2e-05,
      "loss": 0.5174361228942871,
      "step": 9240
    },
    {
      "epoch": 0.6633201864467551,
      "grad_norm": 0.6900809671072387,
      "learning_rate": 2e-05,
      "loss": 0.5136399269104004,
      "step": 9250
    },
    {
      "epoch": 0.6640372893510219,
      "grad_norm": 0.7888374335029795,
      "learning_rate": 2e-05,
      "loss": 0.471953010559082,
      "step": 9260
    },
    {
      "epoch": 0.6647543922552887,
      "grad_norm": 0.8675635899477714,
      "learning_rate": 2e-05,
      "loss": 0.47684154510498045,
      "step": 9270
    },
    {
      "epoch": 0.6654714951595554,
      "grad_norm": 1.6450126245905488,
      "learning_rate": 2e-05,
      "loss": 0.5219486236572266,
      "step": 9280
    },
    {
      "epoch": 0.6661885980638221,
      "grad_norm": 1.3442468874810043,
      "learning_rate": 2e-05,
      "loss": 0.49803647994995115,
      "step": 9290
    },
    {
      "epoch": 0.6669057009680889,
      "grad_norm": 0.9419905372382547,
      "learning_rate": 2e-05,
      "loss": 0.5147303581237793,
      "step": 9300
    },
    {
      "epoch": 0.6676228038723557,
      "grad_norm": 0.9387067742039735,
      "learning_rate": 2e-05,
      "loss": 0.5163489818572998,
      "step": 9310
    },
    {
      "epoch": 0.6683399067766225,
      "grad_norm": 1.7804405104276397,
      "learning_rate": 2e-05,
      "loss": 0.4944204807281494,
      "step": 9320
    },
    {
      "epoch": 0.6690570096808892,
      "grad_norm": 2.2918212975807277,
      "learning_rate": 2e-05,
      "loss": 0.5250197410583496,
      "step": 9330
    },
    {
      "epoch": 0.669774112585156,
      "grad_norm": 0.7982191477776637,
      "learning_rate": 2e-05,
      "loss": 0.480912971496582,
      "step": 9340
    },
    {
      "epoch": 0.6704912154894227,
      "grad_norm": 0.9146017714280487,
      "learning_rate": 2e-05,
      "loss": 0.5006585597991944,
      "step": 9350
    },
    {
      "epoch": 0.6712083183936894,
      "grad_norm": 5.61716351893917,
      "learning_rate": 2e-05,
      "loss": 0.5077700614929199,
      "step": 9360
    },
    {
      "epoch": 0.6719254212979563,
      "grad_norm": 1.0036447886212758,
      "learning_rate": 2e-05,
      "loss": 0.4942521572113037,
      "step": 9370
    },
    {
      "epoch": 0.672642524202223,
      "grad_norm": 1.3172322022369227,
      "learning_rate": 2e-05,
      "loss": 0.49320554733276367,
      "step": 9380
    },
    {
      "epoch": 0.6733596271064898,
      "grad_norm": 0.8309365986385783,
      "learning_rate": 2e-05,
      "loss": 0.4751739978790283,
      "step": 9390
    },
    {
      "epoch": 0.6740767300107565,
      "grad_norm": 1.0688856196492815,
      "learning_rate": 2e-05,
      "loss": 0.5327031135559082,
      "step": 9400
    },
    {
      "epoch": 0.6747938329150233,
      "grad_norm": 0.8733595347212629,
      "learning_rate": 2e-05,
      "loss": 0.5149003982543945,
      "step": 9410
    },
    {
      "epoch": 0.6755109358192901,
      "grad_norm": 0.7086327612347922,
      "learning_rate": 2e-05,
      "loss": 0.5080751895904541,
      "step": 9420
    },
    {
      "epoch": 0.6762280387235569,
      "grad_norm": 0.721019986236191,
      "learning_rate": 2e-05,
      "loss": 0.5218631744384765,
      "step": 9430
    },
    {
      "epoch": 0.6769451416278236,
      "grad_norm": 0.7246104625322608,
      "learning_rate": 2e-05,
      "loss": 0.5058404922485351,
      "step": 9440
    },
    {
      "epoch": 0.6776622445320903,
      "grad_norm": 0.9538959579322058,
      "learning_rate": 2e-05,
      "loss": 0.524578857421875,
      "step": 9450
    },
    {
      "epoch": 0.6783793474363571,
      "grad_norm": 0.9323949260470956,
      "learning_rate": 2e-05,
      "loss": 0.47898058891296386,
      "step": 9460
    },
    {
      "epoch": 0.6790964503406238,
      "grad_norm": 1.1274810682100944,
      "learning_rate": 2e-05,
      "loss": 0.5220388412475586,
      "step": 9470
    },
    {
      "epoch": 0.6798135532448907,
      "grad_norm": 0.7906610910356318,
      "learning_rate": 2e-05,
      "loss": 0.49958114624023436,
      "step": 9480
    },
    {
      "epoch": 0.6805306561491574,
      "grad_norm": 0.8096702879602589,
      "learning_rate": 2e-05,
      "loss": 0.5081011295318604,
      "step": 9490
    },
    {
      "epoch": 0.6812477590534242,
      "grad_norm": 0.7945484830217577,
      "learning_rate": 2e-05,
      "loss": 0.49727277755737304,
      "step": 9500
    },
    {
      "epoch": 0.6819648619576909,
      "grad_norm": 0.9066549915794057,
      "learning_rate": 2e-05,
      "loss": 0.49504451751708983,
      "step": 9510
    },
    {
      "epoch": 0.6826819648619576,
      "grad_norm": 0.8175533456903449,
      "learning_rate": 2e-05,
      "loss": 0.4801572322845459,
      "step": 9520
    },
    {
      "epoch": 0.6833990677662245,
      "grad_norm": 0.7110541617637068,
      "learning_rate": 2e-05,
      "loss": 0.4962407112121582,
      "step": 9530
    },
    {
      "epoch": 0.6841161706704912,
      "grad_norm": 1.3210516359015299,
      "learning_rate": 2e-05,
      "loss": 0.49747066497802733,
      "step": 9540
    },
    {
      "epoch": 0.684833273574758,
      "grad_norm": 1.1189498784664558,
      "learning_rate": 2e-05,
      "loss": 0.49578227996826174,
      "step": 9550
    },
    {
      "epoch": 0.6855503764790247,
      "grad_norm": 1.019563166195094,
      "learning_rate": 2e-05,
      "loss": 0.4858851432800293,
      "step": 9560
    },
    {
      "epoch": 0.6862674793832915,
      "grad_norm": 0.9492045938824456,
      "learning_rate": 2e-05,
      "loss": 0.4732502937316895,
      "step": 9570
    },
    {
      "epoch": 0.6869845822875582,
      "grad_norm": 2.1739789552393685,
      "learning_rate": 2e-05,
      "loss": 0.4906927108764648,
      "step": 9580
    },
    {
      "epoch": 0.6877016851918251,
      "grad_norm": 22.73273366473134,
      "learning_rate": 2e-05,
      "loss": 0.5165609359741211,
      "step": 9590
    },
    {
      "epoch": 0.6884187880960918,
      "grad_norm": 1.9436289820321626,
      "learning_rate": 2e-05,
      "loss": 0.5112183570861817,
      "step": 9600
    },
    {
      "epoch": 0.6891358910003585,
      "grad_norm": 0.9637763413484648,
      "learning_rate": 2e-05,
      "loss": 0.4905545711517334,
      "step": 9610
    },
    {
      "epoch": 0.6898529939046253,
      "grad_norm": 0.8177911574278787,
      "learning_rate": 2e-05,
      "loss": 0.48284101486206055,
      "step": 9620
    },
    {
      "epoch": 0.690570096808892,
      "grad_norm": 5.908457026893531,
      "learning_rate": 2e-05,
      "loss": 0.5252858161926269,
      "step": 9630
    },
    {
      "epoch": 0.6912871997131589,
      "grad_norm": 1.0056211705429612,
      "learning_rate": 2e-05,
      "loss": 0.5047056198120117,
      "step": 9640
    },
    {
      "epoch": 0.6920043026174256,
      "grad_norm": 2.799559518957004,
      "learning_rate": 2e-05,
      "loss": 0.5058962345123291,
      "step": 9650
    },
    {
      "epoch": 0.6927214055216924,
      "grad_norm": 1.8232381558015078,
      "learning_rate": 2e-05,
      "loss": 0.5243574142456054,
      "step": 9660
    },
    {
      "epoch": 0.6934385084259591,
      "grad_norm": 0.8400299974740638,
      "learning_rate": 2e-05,
      "loss": 0.519249677658081,
      "step": 9670
    },
    {
      "epoch": 0.6941556113302259,
      "grad_norm": 2.464641928050492,
      "learning_rate": 2e-05,
      "loss": 0.500908088684082,
      "step": 9680
    },
    {
      "epoch": 0.6948727142344926,
      "grad_norm": 1.251952345365196,
      "learning_rate": 2e-05,
      "loss": 0.51309814453125,
      "step": 9690
    },
    {
      "epoch": 0.6955898171387594,
      "grad_norm": 0.9455295794416523,
      "learning_rate": 2e-05,
      "loss": 0.516147232055664,
      "step": 9700
    },
    {
      "epoch": 0.6963069200430262,
      "grad_norm": 0.9572632759478101,
      "learning_rate": 2e-05,
      "loss": 0.50735502243042,
      "step": 9710
    },
    {
      "epoch": 0.6970240229472929,
      "grad_norm": 1.6287207883351822,
      "learning_rate": 2e-05,
      "loss": 0.5208641052246094,
      "step": 9720
    },
    {
      "epoch": 0.6977411258515597,
      "grad_norm": 1.899614263528001,
      "learning_rate": 2e-05,
      "loss": 0.47897872924804685,
      "step": 9730
    },
    {
      "epoch": 0.6984582287558264,
      "grad_norm": 0.8814283574209666,
      "learning_rate": 2e-05,
      "loss": 0.5037140846252441,
      "step": 9740
    },
    {
      "epoch": 0.6991753316600933,
      "grad_norm": 1.0387290767804602,
      "learning_rate": 2e-05,
      "loss": 0.49989981651306153,
      "step": 9750
    },
    {
      "epoch": 0.69989243456436,
      "grad_norm": 1.4374840583410184,
      "learning_rate": 2e-05,
      "loss": 0.5341364860534668,
      "step": 9760
    },
    {
      "epoch": 0.7006095374686268,
      "grad_norm": 0.8278699260748927,
      "learning_rate": 2e-05,
      "loss": 0.5062505722045898,
      "step": 9770
    },
    {
      "epoch": 0.7013266403728935,
      "grad_norm": 0.8241051602256564,
      "learning_rate": 2e-05,
      "loss": 0.4958690643310547,
      "step": 9780
    },
    {
      "epoch": 0.7020437432771602,
      "grad_norm": 0.927042937250422,
      "learning_rate": 2e-05,
      "loss": 0.49940900802612304,
      "step": 9790
    },
    {
      "epoch": 0.7027608461814271,
      "grad_norm": 2.0505470607431175,
      "learning_rate": 2e-05,
      "loss": 0.48992338180541994,
      "step": 9800
    },
    {
      "epoch": 0.7034779490856938,
      "grad_norm": 1.4216193776360082,
      "learning_rate": 2e-05,
      "loss": 0.48383846282958987,
      "step": 9810
    },
    {
      "epoch": 0.7041950519899606,
      "grad_norm": 0.8507999265636841,
      "learning_rate": 2e-05,
      "loss": 0.49851169586181643,
      "step": 9820
    },
    {
      "epoch": 0.7049121548942273,
      "grad_norm": 1.139729583130427,
      "learning_rate": 2e-05,
      "loss": 0.48622560501098633,
      "step": 9830
    },
    {
      "epoch": 0.7056292577984941,
      "grad_norm": 1.6858231761013758,
      "learning_rate": 2e-05,
      "loss": 0.4862226486206055,
      "step": 9840
    },
    {
      "epoch": 0.7063463607027608,
      "grad_norm": 21.770625902518155,
      "learning_rate": 2e-05,
      "loss": 0.5033634185791016,
      "step": 9850
    },
    {
      "epoch": 0.7070634636070277,
      "grad_norm": 0.7385565153788971,
      "learning_rate": 2e-05,
      "loss": 0.5113471031188965,
      "step": 9860
    },
    {
      "epoch": 0.7077805665112944,
      "grad_norm": 1.1275003593901156,
      "learning_rate": 2e-05,
      "loss": 0.5080653190612793,
      "step": 9870
    },
    {
      "epoch": 0.7084976694155611,
      "grad_norm": 2.1413356213719146,
      "learning_rate": 2e-05,
      "loss": 0.5055551528930664,
      "step": 9880
    },
    {
      "epoch": 0.7092147723198279,
      "grad_norm": 0.6734044958250573,
      "learning_rate": 2e-05,
      "loss": 0.5081550121307373,
      "step": 9890
    },
    {
      "epoch": 0.7099318752240946,
      "grad_norm": 0.8576355713009912,
      "learning_rate": 2e-05,
      "loss": 0.4975290298461914,
      "step": 9900
    },
    {
      "epoch": 0.7106489781283615,
      "grad_norm": 0.9688323326860521,
      "learning_rate": 2e-05,
      "loss": 0.48853163719177245,
      "step": 9910
    },
    {
      "epoch": 0.7113660810326282,
      "grad_norm": 0.7602399668467928,
      "learning_rate": 2e-05,
      "loss": 0.46695337295532224,
      "step": 9920
    },
    {
      "epoch": 0.712083183936895,
      "grad_norm": 1.2159636174434698,
      "learning_rate": 2e-05,
      "loss": 0.5177065849304199,
      "step": 9930
    },
    {
      "epoch": 0.7128002868411617,
      "grad_norm": 1.0270642810820807,
      "learning_rate": 2e-05,
      "loss": 0.4954838752746582,
      "step": 9940
    },
    {
      "epoch": 0.7135173897454284,
      "grad_norm": 1.2710166798780178,
      "learning_rate": 2e-05,
      "loss": 0.5169979095458984,
      "step": 9950
    },
    {
      "epoch": 0.7142344926496952,
      "grad_norm": 0.815046064885696,
      "learning_rate": 2e-05,
      "loss": 0.4995615005493164,
      "step": 9960
    },
    {
      "epoch": 0.714951595553962,
      "grad_norm": 0.6731772694767381,
      "learning_rate": 2e-05,
      "loss": 0.4867525100708008,
      "step": 9970
    },
    {
      "epoch": 0.7156686984582288,
      "grad_norm": 0.6835984407702023,
      "learning_rate": 2e-05,
      "loss": 0.49335269927978515,
      "step": 9980
    },
    {
      "epoch": 0.7163858013624955,
      "grad_norm": 2.429808575619327,
      "learning_rate": 2e-05,
      "loss": 0.49614701271057127,
      "step": 9990
    },
    {
      "epoch": 0.7171029042667623,
      "grad_norm": 0.8543675933848818,
      "learning_rate": 2e-05,
      "loss": 0.4928924560546875,
      "step": 10000
    },
    {
      "epoch": 0.717820007171029,
      "grad_norm": 3.606502395770394,
      "learning_rate": 2e-05,
      "loss": 0.4957789421081543,
      "step": 10010
    },
    {
      "epoch": 0.7185371100752959,
      "grad_norm": 1.3188860428926166,
      "learning_rate": 2e-05,
      "loss": 0.4851681709289551,
      "step": 10020
    },
    {
      "epoch": 0.7192542129795626,
      "grad_norm": 0.918274631364225,
      "learning_rate": 2e-05,
      "loss": 0.5144068717956543,
      "step": 10030
    },
    {
      "epoch": 0.7199713158838293,
      "grad_norm": 0.9050917700615675,
      "learning_rate": 2e-05,
      "loss": 0.5138280868530274,
      "step": 10040
    },
    {
      "epoch": 0.7206884187880961,
      "grad_norm": 1.016396682981612,
      "learning_rate": 2e-05,
      "loss": 0.5060923099517822,
      "step": 10050
    },
    {
      "epoch": 0.7214055216923628,
      "grad_norm": 0.7883480067764334,
      "learning_rate": 2e-05,
      "loss": 0.5374709129333496,
      "step": 10060
    },
    {
      "epoch": 0.7221226245966296,
      "grad_norm": 1.035207148454411,
      "learning_rate": 2e-05,
      "loss": 0.49805312156677245,
      "step": 10070
    },
    {
      "epoch": 0.7228397275008964,
      "grad_norm": 0.7634611263285472,
      "learning_rate": 2e-05,
      "loss": 0.5084001541137695,
      "step": 10080
    },
    {
      "epoch": 0.7235568304051632,
      "grad_norm": 1.1470185471512249,
      "learning_rate": 2e-05,
      "loss": 0.5125796318054199,
      "step": 10090
    },
    {
      "epoch": 0.7242739333094299,
      "grad_norm": 1.2938223718698396,
      "learning_rate": 2e-05,
      "loss": 0.5057493209838867,
      "step": 10100
    },
    {
      "epoch": 0.7249910362136966,
      "grad_norm": 1.0301311488028368,
      "learning_rate": 2e-05,
      "loss": 0.5119041442871094,
      "step": 10110
    },
    {
      "epoch": 0.7257081391179634,
      "grad_norm": 1.043959345812144,
      "learning_rate": 2e-05,
      "loss": 0.5235150337219239,
      "step": 10120
    },
    {
      "epoch": 0.7264252420222302,
      "grad_norm": 1.5451024042631125,
      "learning_rate": 2e-05,
      "loss": 0.48797006607055665,
      "step": 10130
    },
    {
      "epoch": 0.727142344926497,
      "grad_norm": 2.0015011724655087,
      "learning_rate": 2e-05,
      "loss": 0.49192066192626954,
      "step": 10140
    },
    {
      "epoch": 0.7278594478307637,
      "grad_norm": 0.9924212986282381,
      "learning_rate": 2e-05,
      "loss": 0.49672822952270507,
      "step": 10150
    },
    {
      "epoch": 0.7285765507350305,
      "grad_norm": 0.7343226744655088,
      "learning_rate": 2e-05,
      "loss": 0.4967620849609375,
      "step": 10160
    },
    {
      "epoch": 0.7292936536392972,
      "grad_norm": 0.7113337204673753,
      "learning_rate": 2e-05,
      "loss": 0.5088300704956055,
      "step": 10170
    },
    {
      "epoch": 0.730010756543564,
      "grad_norm": 1.3125319954501786,
      "learning_rate": 1.999957274187931e-05,
      "loss": 0.49834975600242615,
      "step": 10180
    },
    {
      "epoch": 0.7307278594478308,
      "grad_norm": 1.2571534833220983,
      "learning_rate": 1.9991978062682906e-05,
      "loss": 0.5141496181488037,
      "step": 10190
    },
    {
      "epoch": 0.7314449623520975,
      "grad_norm": 0.6685863033456172,
      "learning_rate": 1.997489709987378e-05,
      "loss": 0.4754651546478271,
      "step": 10200
    },
    {
      "epoch": 0.7321620652563643,
      "grad_norm": 0.9238524045077421,
      "learning_rate": 1.9948346151502135e-05,
      "loss": 0.5090078353881836,
      "step": 10210
    },
    {
      "epoch": 0.732879168160631,
      "grad_norm": 1.5656551426452576,
      "learning_rate": 1.991235055154325e-05,
      "loss": 0.4722578525543213,
      "step": 10220
    },
    {
      "epoch": 0.7335962710648978,
      "grad_norm": 0.8433525197524905,
      "learning_rate": 1.9866944645724693e-05,
      "loss": 0.48906526565551756,
      "step": 10230
    },
    {
      "epoch": 0.7343133739691646,
      "grad_norm": 1.1297185403747954,
      "learning_rate": 1.9812171758754842e-05,
      "loss": 0.4846484184265137,
      "step": 10240
    },
    {
      "epoch": 0.7350304768734314,
      "grad_norm": 0.7286155369380147,
      "learning_rate": 1.9748084152983987e-05,
      "loss": 0.47603626251220704,
      "step": 10250
    },
    {
      "epoch": 0.7357475797776981,
      "grad_norm": 1.0855584325391128,
      "learning_rate": 1.967474297853744e-05,
      "loss": 0.502126407623291,
      "step": 10260
    },
    {
      "epoch": 0.7364646826819649,
      "grad_norm": 0.8459181476800953,
      "learning_rate": 1.9592218214968256e-05,
      "loss": 0.4876286506652832,
      "step": 10270
    },
    {
      "epoch": 0.7371817855862316,
      "grad_norm": 1.8025587341569034,
      "learning_rate": 1.9500588604485213e-05,
      "loss": 0.479932975769043,
      "step": 10280
    },
    {
      "epoch": 0.7378988884904983,
      "grad_norm": 0.9587774444243159,
      "learning_rate": 1.939994157681979e-05,
      "loss": 0.5157277107238769,
      "step": 10290
    },
    {
      "epoch": 0.7386159913947652,
      "grad_norm": 0.8055060409693898,
      "learning_rate": 1.9290373165803803e-05,
      "loss": 0.47760915756225586,
      "step": 10300
    },
    {
      "epoch": 0.7393330942990319,
      "grad_norm": 1.2535861575398903,
      "learning_rate": 1.9171987917737328e-05,
      "loss": 0.5134068489074707,
      "step": 10310
    },
    {
      "epoch": 0.7400501972032987,
      "grad_norm": 0.9282262128352239,
      "learning_rate": 1.9044898791634322e-05,
      "loss": 0.5137625694274902,
      "step": 10320
    },
    {
      "epoch": 0.7407673001075654,
      "grad_norm": 1.9095027140298455,
      "learning_rate": 1.8909227051441126e-05,
      "loss": 0.5047832489013672,
      "step": 10330
    },
    {
      "epoch": 0.7414844030118322,
      "grad_norm": 0.9913913402411693,
      "learning_rate": 1.8765102150330692e-05,
      "loss": 0.50838623046875,
      "step": 10340
    },
    {
      "epoch": 0.742201505916099,
      "grad_norm": 0.7024024978522814,
      "learning_rate": 1.861266160718297e-05,
      "loss": 0.49452695846557615,
      "step": 10350
    },
    {
      "epoch": 0.7429186088203658,
      "grad_norm": 1.6647790724422173,
      "learning_rate": 1.845205087536923e-05,
      "loss": 0.47962632179260256,
      "step": 10360
    },
    {
      "epoch": 0.7436357117246325,
      "grad_norm": 1.1343187761211675,
      "learning_rate": 1.8283423203965633e-05,
      "loss": 0.49178309440612794,
      "step": 10370
    },
    {
      "epoch": 0.7443528146288992,
      "grad_norm": 3.8431550816183027,
      "learning_rate": 1.810693949152835e-05,
      "loss": 0.5093101501464844,
      "step": 10380
    },
    {
      "epoch": 0.745069917533166,
      "grad_norm": 0.7858323670647077,
      "learning_rate": 1.7922768132569902e-05,
      "loss": 0.48663840293884275,
      "step": 10390
    },
    {
      "epoch": 0.7457870204374327,
      "grad_norm": 0.834262746782999,
      "learning_rate": 1.773108485688302e-05,
      "loss": 0.5181979179382324,
      "step": 10400
    },
    {
      "epoch": 0.7465041233416996,
      "grad_norm": 1.8821925709733929,
      "learning_rate": 1.7532072561865537e-05,
      "loss": 0.4863869667053223,
      "step": 10410
    },
    {
      "epoch": 0.7472212262459663,
      "grad_norm": 0.7302790789761268,
      "learning_rate": 1.732592113800616e-05,
      "loss": 0.49062314033508303,
      "step": 10420
    },
    {
      "epoch": 0.7479383291502331,
      "grad_norm": 0.7571399482380871,
      "learning_rate": 1.7112827287697668e-05,
      "loss": 0.5217619895935058,
      "step": 10430
    },
    {
      "epoch": 0.7486554320544998,
      "grad_norm": 0.8930653441818344,
      "learning_rate": 1.6892994337550452e-05,
      "loss": 0.5065752029418945,
      "step": 10440
    },
    {
      "epoch": 0.7493725349587665,
      "grad_norm": 0.9211905623295054,
      "learning_rate": 1.6666632044385487e-05,
      "loss": 0.48835458755493166,
      "step": 10450
    },
    {
      "epoch": 0.7500896378630334,
      "grad_norm": 0.8928607458086855,
      "learning_rate": 1.6433956395091757e-05,
      "loss": 0.5194371223449707,
      "step": 10460
    },
    {
      "epoch": 0.7508067407673001,
      "grad_norm": 0.7848688827205408,
      "learning_rate": 1.6195189400539238e-05,
      "loss": 0.5096420288085938,
      "step": 10470
    },
    {
      "epoch": 0.7515238436715669,
      "grad_norm": 0.7801897207479728,
      "learning_rate": 1.595055888374398e-05,
      "loss": 0.48164167404174807,
      "step": 10480
    },
    {
      "epoch": 0.7522409465758336,
      "grad_norm": 0.7533098064756493,
      "learning_rate": 1.5700298262487434e-05,
      "loss": 0.45863637924194334,
      "step": 10490
    },
    {
      "epoch": 0.7529580494801004,
      "grad_norm": 0.8227291548970657,
      "learning_rate": 1.5444646326597467e-05,
      "loss": 0.48957386016845705,
      "step": 10500
    },
    {
      "epoch": 0.7536751523843671,
      "grad_norm": 3.00783965213211,
      "learning_rate": 1.5183847010103586e-05,
      "loss": 0.4901371955871582,
      "step": 10510
    },
    {
      "epoch": 0.754392255288634,
      "grad_norm": 0.7842110269316798,
      "learning_rate": 1.4918149158483686e-05,
      "loss": 0.48599929809570314,
      "step": 10520
    },
    {
      "epoch": 0.7551093581929007,
      "grad_norm": 0.8151813295282105,
      "learning_rate": 1.464780629122452e-05,
      "loss": 0.5000704765319824,
      "step": 10530
    },
    {
      "epoch": 0.7558264610971674,
      "grad_norm": 2.387883038423856,
      "learning_rate": 1.4373076359922374e-05,
      "loss": 0.5148778438568116,
      "step": 10540
    },
    {
      "epoch": 0.7565435640014342,
      "grad_norm": 0.8383616826936007,
      "learning_rate": 1.4094221502154774e-05,
      "loss": 0.516185712814331,
      "step": 10550
    },
    {
      "epoch": 0.7572606669057009,
      "grad_norm": 0.6377037017012589,
      "learning_rate": 1.3811507791358113e-05,
      "loss": 0.46940174102783205,
      "step": 10560
    },
    {
      "epoch": 0.7579777698099678,
      "grad_norm": 3.7742987148986717,
      "learning_rate": 1.3525204982949753e-05,
      "loss": 0.4904289245605469,
      "step": 10570
    },
    {
      "epoch": 0.7586948727142345,
      "grad_norm": 0.7844847034043596,
      "learning_rate": 1.3235586256936996e-05,
      "loss": 0.47499990463256836,
      "step": 10580
    },
    {
      "epoch": 0.7594119756185013,
      "grad_norm": 0.7506174914948536,
      "learning_rate": 1.294292795725836e-05,
      "loss": 0.49221296310424806,
      "step": 10590
    },
    {
      "epoch": 0.760129078522768,
      "grad_norm": 0.7103006182825843,
      "learning_rate": 1.2647509328105978e-05,
      "loss": 0.5475531578063965,
      "step": 10600
    },
    {
      "epoch": 0.7608461814270348,
      "grad_norm": 0.6557080055293819,
      "learning_rate": 1.2349612247480686e-05,
      "loss": 0.5034311771392822,
      "step": 10610
    },
    {
      "epoch": 0.7615632843313015,
      "grad_norm": 1.6844202789563303,
      "learning_rate": 1.2049520958234e-05,
      "loss": 0.5117573738098145,
      "step": 10620
    },
    {
      "epoch": 0.7622803872355683,
      "grad_norm": 2.1914416536876695,
      "learning_rate": 1.1747521796853638e-05,
      "loss": 0.49009037017822266,
      "step": 10630
    },
    {
      "epoch": 0.7629974901398351,
      "grad_norm": 0.7213350789867643,
      "learning_rate": 1.1443902920251395e-05,
      "loss": 0.4660952568054199,
      "step": 10640
    },
    {
      "epoch": 0.7637145930441018,
      "grad_norm": 0.8756120767405681,
      "learning_rate": 1.1138954030814009e-05,
      "loss": 0.5008755207061768,
      "step": 10650
    },
    {
      "epoch": 0.7644316959483686,
      "grad_norm": 0.9627162065153606,
      "learning_rate": 1.0832966099979388e-05,
      "loss": 0.47268214225769045,
      "step": 10660
    },
    {
      "epoch": 0.7651487988526353,
      "grad_norm": 0.6853794759261299,
      "learning_rate": 1.0526231090601969e-05,
      "loss": 0.4888431549072266,
      "step": 10670
    },
    {
      "epoch": 0.7658659017569022,
      "grad_norm": 0.8992674203441011,
      "learning_rate": 1.0219041678372106e-05,
      "loss": 0.5097752571105957,
      "step": 10680
    },
    {
      "epoch": 0.7665830046611689,
      "grad_norm": 1.106296185406063,
      "learning_rate": 9.911690972555268e-06,
      "loss": 0.46671457290649415,
      "step": 10690
    },
    {
      "epoch": 0.7673001075654357,
      "grad_norm": 0.7505696652454542,
      "learning_rate": 9.604472236317541e-06,
      "loss": 0.46439523696899415,
      "step": 10700
    },
    {
      "epoch": 0.7680172104697024,
      "grad_norm": 0.7887839789713776,
      "learning_rate": 9.297678606904335e-06,
      "loss": 0.47592673301696775,
      "step": 10710
    },
    {
      "epoch": 0.7687343133739691,
      "grad_norm": 3.226306432745247,
      "learning_rate": 8.991602815939155e-06,
      "loss": 0.4807309627532959,
      "step": 10720
    },
    {
      "epoch": 0.7694514162782359,
      "grad_norm": 0.7714259028756096,
      "learning_rate": 8.686536910109464e-06,
      "loss": 0.5007771968841552,
      "step": 10730
    },
    {
      "epoch": 0.7701685191825027,
      "grad_norm": 1.1473005435189936,
      "learning_rate": 8.382771972506084e-06,
      "loss": 0.5231281280517578,
      "step": 10740
    },
    {
      "epoch": 0.7708856220867695,
      "grad_norm": 1.0971619862497648,
      "learning_rate": 8.08059784488198e-06,
      "loss": 0.4763030529022217,
      "step": 10750
    },
    {
      "epoch": 0.7716027249910362,
      "grad_norm": 0.7963427467004226,
      "learning_rate": 7.780302851095544e-06,
      "loss": 0.4660953521728516,
      "step": 10760
    },
    {
      "epoch": 0.772319827895303,
      "grad_norm": 0.9907785052253164,
      "learning_rate": 7.482173522002166e-06,
      "loss": 0.47362842559814455,
      "step": 10770
    },
    {
      "epoch": 0.7730369307995697,
      "grad_norm": 1.1822724934482711,
      "learning_rate": 7.186494322056624e-06,
      "loss": 0.4902331352233887,
      "step": 10780
    },
    {
      "epoch": 0.7737540337038366,
      "grad_norm": 0.727948856287179,
      "learning_rate": 6.893547377887183e-06,
      "loss": 0.47796926498413084,
      "step": 10790
    },
    {
      "epoch": 0.7744711366081033,
      "grad_norm": 0.7706091229037056,
      "learning_rate": 6.6036122091003365e-06,
      "loss": 0.4482910633087158,
      "step": 10800
    },
    {
      "epoch": 0.77518823951237,
      "grad_norm": 0.7219156368057554,
      "learning_rate": 6.316965461573097e-06,
      "loss": 0.4931656837463379,
      "step": 10810
    },
    {
      "epoch": 0.7759053424166368,
      "grad_norm": 0.7838502318484459,
      "learning_rate": 6.0338806434872965e-06,
      "loss": 0.4703873634338379,
      "step": 10820
    },
    {
      "epoch": 0.7766224453209035,
      "grad_norm": 0.8338781385381151,
      "learning_rate": 5.754627864357749e-06,
      "loss": 0.45143928527832033,
      "step": 10830
    },
    {
      "epoch": 0.7773395482251703,
      "grad_norm": 0.8029139102759046,
      "learning_rate": 5.479473577303319e-06,
      "loss": 0.48016843795776365,
      "step": 10840
    },
    {
      "epoch": 0.7780566511294371,
      "grad_norm": 0.8086334189467614,
      "learning_rate": 5.208680324806778e-06,
      "loss": 0.4473318099975586,
      "step": 10850
    },
    {
      "epoch": 0.7787737540337039,
      "grad_norm": 0.7610576304216898,
      "learning_rate": 4.94250648820606e-06,
      "loss": 0.48781986236572267,
      "step": 10860
    },
    {
      "epoch": 0.7794908569379706,
      "grad_norm": 1.0510158482728826,
      "learning_rate": 4.681206041155951e-06,
      "loss": 0.4429471969604492,
      "step": 10870
    },
    {
      "epoch": 0.7802079598422373,
      "grad_norm": 0.7772570182629249,
      "learning_rate": 4.42502830729538e-06,
      "loss": 0.4819638729095459,
      "step": 10880
    },
    {
      "epoch": 0.7809250627465041,
      "grad_norm": 0.7938992172159133,
      "learning_rate": 4.174217722351669e-06,
      "loss": 0.4389060974121094,
      "step": 10890
    },
    {
      "epoch": 0.7816421656507709,
      "grad_norm": 0.8593972742216387,
      "learning_rate": 3.929013600908575e-06,
      "loss": 0.49288244247436525,
      "step": 10900
    },
    {
      "epoch": 0.7823592685550377,
      "grad_norm": 1.085771312925711,
      "learning_rate": 3.689649908060807e-06,
      "loss": 0.4854423522949219,
      "step": 10910
    },
    {
      "epoch": 0.7830763714593044,
      "grad_norm": 0.7694177105313834,
      "learning_rate": 3.4563550361728028e-06,
      "loss": 0.4779941558837891,
      "step": 10920
    },
    {
      "epoch": 0.7837934743635712,
      "grad_norm": 1.202177465581903,
      "learning_rate": 3.2293515869548216e-06,
      "loss": 0.4765610694885254,
      "step": 10930
    },
    {
      "epoch": 0.7845105772678379,
      "grad_norm": 0.7394515876822204,
      "learning_rate": 3.0088561590642746e-06,
      "loss": 0.44635539054870604,
      "step": 10940
    },
    {
      "epoch": 0.7852276801721046,
      "grad_norm": 1.8117876671275102,
      "learning_rate": 2.795079141434941e-06,
      "loss": 0.47303009033203125,
      "step": 10950
    },
    {
      "epoch": 0.7859447830763715,
      "grad_norm": 0.8414862656054083,
      "learning_rate": 2.588224512531299e-06,
      "loss": 0.45820136070251466,
      "step": 10960
    },
    {
      "epoch": 0.7866618859806382,
      "grad_norm": 1.5240934442415712,
      "learning_rate": 2.388489645719495e-06,
      "loss": 0.498581600189209,
      "step": 10970
    },
    {
      "epoch": 0.787378988884905,
      "grad_norm": 0.9095248824825837,
      "learning_rate": 2.196065120940663e-06,
      "loss": 0.47748546600341796,
      "step": 10980
    },
    {
      "epoch": 0.7880960917891717,
      "grad_norm": 0.7266097197710967,
      "learning_rate": 2.0111345428662808e-06,
      "loss": 0.47298598289489746,
      "step": 10990
    },
    {
      "epoch": 0.7888131946934385,
      "grad_norm": 1.7916289852758844,
      "learning_rate": 1.8338743657091102e-06,
      "loss": 0.511493730545044,
      "step": 11000
    },
    {
      "epoch": 0.7895302975977053,
      "grad_norm": 0.7320680286320005,
      "learning_rate": 1.6644537248568113e-06,
      "loss": 0.4647524833679199,
      "step": 11010
    },
    {
      "epoch": 0.7902474005019721,
      "grad_norm": 0.815136831398101,
      "learning_rate": 1.503034275488951e-06,
      "loss": 0.4940489292144775,
      "step": 11020
    },
    {
      "epoch": 0.7909645034062388,
      "grad_norm": 0.6319952166368067,
      "learning_rate": 1.3497700383313495e-06,
      "loss": 0.4621113777160645,
      "step": 11030
    },
    {
      "epoch": 0.7916816063105055,
      "grad_norm": 0.7567692950870308,
      "learning_rate": 1.2048072526949562e-06,
      "loss": 0.4783414363861084,
      "step": 11040
    },
    {
      "epoch": 0.7923987092147723,
      "grad_norm": 0.8632756555057272,
      "learning_rate": 1.0682842369394756e-06,
      "loss": 0.491253662109375,
      "step": 11050
    },
    {
      "epoch": 0.793115812119039,
      "grad_norm": 0.7644062440848356,
      "learning_rate": 9.403312564948792e-07,
      "loss": 0.4396210670471191,
      "step": 11060
    },
    {
      "epoch": 0.7938329150233059,
      "grad_norm": 0.8551921381519436,
      "learning_rate": 8.210703995667471e-07,
      "loss": 0.47462072372436526,
      "step": 11070
    },
    {
      "epoch": 0.7945500179275726,
      "grad_norm": 0.8474452527235217,
      "learning_rate": 7.106154606440245e-07,
      "loss": 0.47446374893188475,
      "step": 11080
    },
    {
      "epoch": 0.7952671208318394,
      "grad_norm": 0.8204696742450667,
      "learning_rate": 6.090718319203393e-07,
      "loss": 0.4776448726654053,
      "step": 11090
    },
    {
      "epoch": 0.7959842237361061,
      "grad_norm": 1.0100310931177854,
      "learning_rate": 5.165364027325089e-07,
      "loss": 0.4678969860076904,
      "step": 11100
    },
    {
      "epoch": 0.7967013266403729,
      "grad_norm": 0.7708214599415789,
      "learning_rate": 4.330974671121585e-07,
      "loss": 0.48924694061279295,
      "step": 11110
    },
    {
      "epoch": 0.7974184295446397,
      "grad_norm": 1.2188357808924604,
      "learning_rate": 3.588346395386846e-07,
      "loss": 0.46663618087768555,
      "step": 11120
    },
    {
      "epoch": 0.7981355324489064,
      "grad_norm": 0.7822584002385463,
      "learning_rate": 2.938187789739395e-07,
      "loss": 0.4769263744354248,
      "step": 11130
    },
    {
      "epoch": 0.7988526353531732,
      "grad_norm": 0.6825344854451846,
      "learning_rate": 2.3811192125110724e-07,
      "loss": 0.4802879333496094,
      "step": 11140
    },
    {
      "epoch": 0.7995697382574399,
      "grad_norm": 0.7594935224629794,
      "learning_rate": 1.917672198823177e-07,
      "loss": 0.48874454498291015,
      "step": 11150
    },
    {
      "epoch": 0.8002868411617067,
      "grad_norm": 0.7206491787419553,
      "learning_rate": 1.5482889534143795e-07,
      "loss": 0.4680596351623535,
      "step": 11160
    },
    {
      "epoch": 0.8010039440659734,
      "grad_norm": 0.9316074024560024,
      "learning_rate": 1.273321928704694e-07,
      "loss": 0.4766650676727295,
      "step": 11170
    },
    {
      "epoch": 0.8017210469702403,
      "grad_norm": 0.6965755633763265,
      "learning_rate": 1.0930334884977866e-07,
      "loss": 0.4605852603912354,
      "step": 11180
    },
    {
      "epoch": 0.802438149874507,
      "grad_norm": 0.851009841668216,
      "learning_rate": 1.0075956576427487e-07,
      "loss": 0.4739173412322998,
      "step": 11190
    },
    {
      "epoch": 0.802653280745787,
      "step": 11193,
      "total_flos": 5300110976090112.0,
      "train_loss": 0.04410107749445471,
      "train_runtime": 14625.8681,
      "train_samples_per_second": 9.183,
      "train_steps_per_second": 0.765
    }
  ],
  "logging_steps": 10,
  "max_steps": 11193,
  "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": 5300110976090112.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}