{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2293719806763285,
  "eval_steps": 500,
  "global_step": 5935,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003864734299516908,
      "grad_norm": 67.77150678538916,
      "learning_rate": 2.910112359550562e-06,
      "loss": 1.7119647979736328,
      "step": 10
    },
    {
      "epoch": 0.0007729468599033816,
      "grad_norm": 112.20087349554393,
      "learning_rate": 3.921348314606742e-06,
      "loss": 1.0294610977172851,
      "step": 20
    },
    {
      "epoch": 0.0011594202898550724,
      "grad_norm": 17.998285122767445,
      "learning_rate": 4.932584269662922e-06,
      "loss": 0.9339380264282227,
      "step": 30
    },
    {
      "epoch": 0.0015458937198067632,
      "grad_norm": 44.13244966887987,
      "learning_rate": 5.943820224719102e-06,
      "loss": 0.8280322074890136,
      "step": 40
    },
    {
      "epoch": 0.001932367149758454,
      "grad_norm": 35.342167635210814,
      "learning_rate": 6.955056179775282e-06,
      "loss": 0.8310356140136719,
      "step": 50
    },
    {
      "epoch": 0.002318840579710145,
      "grad_norm": 13.961543042810325,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.8424642562866211,
      "step": 60
    },
    {
      "epoch": 0.002705314009661836,
      "grad_norm": 16.09056628824688,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.8591781616210937,
      "step": 70
    },
    {
      "epoch": 0.0030917874396135265,
      "grad_norm": 54.47103241018983,
      "learning_rate": 9.988764044943822e-06,
      "loss": 0.8048875808715821,
      "step": 80
    },
    {
      "epoch": 0.0034782608695652175,
      "grad_norm": 288.41808740691243,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.7839107513427734,
      "step": 90
    },
    {
      "epoch": 0.003864734299516908,
      "grad_norm": 90.30364936720191,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.8227674484252929,
      "step": 100
    },
    {
      "epoch": 0.004251207729468599,
      "grad_norm": 7.031443917093176,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.8219305038452148,
      "step": 110
    },
    {
      "epoch": 0.00463768115942029,
      "grad_norm": 64.81590934958837,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.8873563766479492,
      "step": 120
    },
    {
      "epoch": 0.00502415458937198,
      "grad_norm": 7.4010759674397075,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.8595630645751953,
      "step": 130
    },
    {
      "epoch": 0.005410628019323672,
      "grad_norm": 19.21019328586394,
      "learning_rate": 1.60561797752809e-05,
      "loss": 0.8214578628540039,
      "step": 140
    },
    {
      "epoch": 0.005797101449275362,
      "grad_norm": 12.640756734713527,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.8313806533813477,
      "step": 150
    },
    {
      "epoch": 0.006183574879227053,
      "grad_norm": 30.722029453858426,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.8253059387207031,
      "step": 160
    },
    {
      "epoch": 0.006570048309178744,
      "grad_norm": 9.610182761607462,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.8571659088134765,
      "step": 170
    },
    {
      "epoch": 0.006956521739130435,
      "grad_norm": 66.94480959155315,
      "learning_rate": 2e-05,
      "loss": 0.825499439239502,
      "step": 180
    },
    {
      "epoch": 0.007342995169082126,
      "grad_norm": 18.927988413767896,
      "learning_rate": 2e-05,
      "loss": 0.8307562828063965,
      "step": 190
    },
    {
      "epoch": 0.007729468599033816,
      "grad_norm": 12.259205440850302,
      "learning_rate": 2e-05,
      "loss": 0.8559392929077149,
      "step": 200
    },
    {
      "epoch": 0.008115942028985508,
      "grad_norm": 4.705925662053298,
      "learning_rate": 2e-05,
      "loss": 0.8608473777770996,
      "step": 210
    },
    {
      "epoch": 0.008502415458937198,
      "grad_norm": 2.301827055644639,
      "learning_rate": 2e-05,
      "loss": 0.8078425407409668,
      "step": 220
    },
    {
      "epoch": 0.008888888888888889,
      "grad_norm": 7.309011200551571,
      "learning_rate": 2e-05,
      "loss": 0.8398617744445801,
      "step": 230
    },
    {
      "epoch": 0.00927536231884058,
      "grad_norm": 6.173820350919054,
      "learning_rate": 2e-05,
      "loss": 0.8554814338684082,
      "step": 240
    },
    {
      "epoch": 0.00966183574879227,
      "grad_norm": 6.755722543096072,
      "learning_rate": 2e-05,
      "loss": 0.8236814498901367,
      "step": 250
    },
    {
      "epoch": 0.01004830917874396,
      "grad_norm": 8.495921320038692,
      "learning_rate": 2e-05,
      "loss": 0.8807731628417969,
      "step": 260
    },
    {
      "epoch": 0.010434782608695653,
      "grad_norm": 6.836865467270201,
      "learning_rate": 2e-05,
      "loss": 0.8989522933959961,
      "step": 270
    },
    {
      "epoch": 0.010821256038647344,
      "grad_norm": 14.122874054097522,
      "learning_rate": 2e-05,
      "loss": 0.8458233833312988,
      "step": 280
    },
    {
      "epoch": 0.011207729468599034,
      "grad_norm": 16.985124793104738,
      "learning_rate": 2e-05,
      "loss": 0.8226848602294922,
      "step": 290
    },
    {
      "epoch": 0.011594202898550725,
      "grad_norm": 41.65329788058193,
      "learning_rate": 2e-05,
      "loss": 0.8596689224243164,
      "step": 300
    },
    {
      "epoch": 0.011980676328502415,
      "grad_norm": 4.572131278336222,
      "learning_rate": 2e-05,
      "loss": 0.8586993217468262,
      "step": 310
    },
    {
      "epoch": 0.012367149758454106,
      "grad_norm": 15.52553197362288,
      "learning_rate": 2e-05,
      "loss": 0.8715810775756836,
      "step": 320
    },
    {
      "epoch": 0.012753623188405797,
      "grad_norm": 2.099292672310561,
      "learning_rate": 2e-05,
      "loss": 0.859681510925293,
      "step": 330
    },
    {
      "epoch": 0.013140096618357487,
      "grad_norm": 2.9884282294816904,
      "learning_rate": 2e-05,
      "loss": 0.8609855651855469,
      "step": 340
    },
    {
      "epoch": 0.01352657004830918,
      "grad_norm": 2.13249355879509,
      "learning_rate": 2e-05,
      "loss": 0.8378034591674804,
      "step": 350
    },
    {
      "epoch": 0.01391304347826087,
      "grad_norm": 1.6770141008927584,
      "learning_rate": 2e-05,
      "loss": 0.8262413024902344,
      "step": 360
    },
    {
      "epoch": 0.01429951690821256,
      "grad_norm": 4.682041458440538,
      "learning_rate": 2e-05,
      "loss": 0.8071828842163086,
      "step": 370
    },
    {
      "epoch": 0.014685990338164251,
      "grad_norm": 1.8513588195141903,
      "learning_rate": 2e-05,
      "loss": 0.7924742698669434,
      "step": 380
    },
    {
      "epoch": 0.015072463768115942,
      "grad_norm": 2.4641373360496774,
      "learning_rate": 2e-05,
      "loss": 0.8004615783691407,
      "step": 390
    },
    {
      "epoch": 0.015458937198067632,
      "grad_norm": 2.3004772602625914,
      "learning_rate": 2e-05,
      "loss": 0.7985801219940185,
      "step": 400
    },
    {
      "epoch": 0.015845410628019325,
      "grad_norm": 2.0543279811562636,
      "learning_rate": 2e-05,
      "loss": 0.8145970344543457,
      "step": 410
    },
    {
      "epoch": 0.016231884057971015,
      "grad_norm": 11.830783545012162,
      "learning_rate": 2e-05,
      "loss": 0.8268651962280273,
      "step": 420
    },
    {
      "epoch": 0.016618357487922706,
      "grad_norm": 23.315575489389328,
      "learning_rate": 2e-05,
      "loss": 0.8241431236267089,
      "step": 430
    },
    {
      "epoch": 0.017004830917874397,
      "grad_norm": 21.42157001872646,
      "learning_rate": 2e-05,
      "loss": 0.8197898864746094,
      "step": 440
    },
    {
      "epoch": 0.017391304347826087,
      "grad_norm": 6.089448482941218,
      "learning_rate": 2e-05,
      "loss": 0.799440574645996,
      "step": 450
    },
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 2.0297328988872207,
      "learning_rate": 2e-05,
      "loss": 0.7915650844573975,
      "step": 460
    },
    {
      "epoch": 0.01816425120772947,
      "grad_norm": 2.5804977642175406,
      "learning_rate": 2e-05,
      "loss": 0.8292923927307129,
      "step": 470
    },
    {
      "epoch": 0.01855072463768116,
      "grad_norm": 2.038221241470096,
      "learning_rate": 2e-05,
      "loss": 0.8466601371765137,
      "step": 480
    },
    {
      "epoch": 0.01893719806763285,
      "grad_norm": 2.6279678065101404,
      "learning_rate": 2e-05,
      "loss": 0.7829755306243896,
      "step": 490
    },
    {
      "epoch": 0.01932367149758454,
      "grad_norm": 3.6287666464295207,
      "learning_rate": 2e-05,
      "loss": 0.8295025825500488,
      "step": 500
    },
    {
      "epoch": 0.01971014492753623,
      "grad_norm": 5.056996216526932,
      "learning_rate": 2e-05,
      "loss": 0.8213679313659668,
      "step": 510
    },
    {
      "epoch": 0.02009661835748792,
      "grad_norm": 3.2319172805464556,
      "learning_rate": 2e-05,
      "loss": 0.8206055641174317,
      "step": 520
    },
    {
      "epoch": 0.020483091787439612,
      "grad_norm": 8.57107033721087,
      "learning_rate": 2e-05,
      "loss": 0.8187108039855957,
      "step": 530
    },
    {
      "epoch": 0.020869565217391306,
      "grad_norm": 2.3249597398057866,
      "learning_rate": 2e-05,
      "loss": 0.8607352256774903,
      "step": 540
    },
    {
      "epoch": 0.021256038647342997,
      "grad_norm": 4.207073762673164,
      "learning_rate": 2e-05,
      "loss": 0.803462028503418,
      "step": 550
    },
    {
      "epoch": 0.021642512077294687,
      "grad_norm": 2.16004233284329,
      "learning_rate": 2e-05,
      "loss": 0.8118793487548828,
      "step": 560
    },
    {
      "epoch": 0.022028985507246378,
      "grad_norm": 5.922676764922052,
      "learning_rate": 2e-05,
      "loss": 0.8660179138183594,
      "step": 570
    },
    {
      "epoch": 0.02241545893719807,
      "grad_norm": 2.8910298488294783,
      "learning_rate": 2e-05,
      "loss": 0.8383306503295899,
      "step": 580
    },
    {
      "epoch": 0.02280193236714976,
      "grad_norm": 10.435855529697001,
      "learning_rate": 2e-05,
      "loss": 0.8274753570556641,
      "step": 590
    },
    {
      "epoch": 0.02318840579710145,
      "grad_norm": 5.716952677601055,
      "learning_rate": 2e-05,
      "loss": 0.8341848373413085,
      "step": 600
    },
    {
      "epoch": 0.02357487922705314,
      "grad_norm": 1.7990410820641871,
      "learning_rate": 2e-05,
      "loss": 0.8170646667480469,
      "step": 610
    },
    {
      "epoch": 0.02396135265700483,
      "grad_norm": 2.5141846970054984,
      "learning_rate": 2e-05,
      "loss": 0.8089756965637207,
      "step": 620
    },
    {
      "epoch": 0.02434782608695652,
      "grad_norm": 1.644941789727098,
      "learning_rate": 2e-05,
      "loss": 0.7915188789367675,
      "step": 630
    },
    {
      "epoch": 0.024734299516908212,
      "grad_norm": 4.69436608310746,
      "learning_rate": 2e-05,
      "loss": 0.8204584121704102,
      "step": 640
    },
    {
      "epoch": 0.025120772946859903,
      "grad_norm": 14.985321654813715,
      "learning_rate": 2e-05,
      "loss": 0.8408443450927734,
      "step": 650
    },
    {
      "epoch": 0.025507246376811593,
      "grad_norm": 3.606103990135419,
      "learning_rate": 2e-05,
      "loss": 0.8335097312927247,
      "step": 660
    },
    {
      "epoch": 0.025893719806763284,
      "grad_norm": 302.2907667494139,
      "learning_rate": 2e-05,
      "loss": 0.8303115844726563,
      "step": 670
    },
    {
      "epoch": 0.026280193236714974,
      "grad_norm": 5.094641064796761,
      "learning_rate": 2e-05,
      "loss": 0.8183450698852539,
      "step": 680
    },
    {
      "epoch": 0.02666666666666667,
      "grad_norm": 7.71531360876406,
      "learning_rate": 2e-05,
      "loss": 0.8168371200561524,
      "step": 690
    },
    {
      "epoch": 0.02705314009661836,
      "grad_norm": 27.557957472594705,
      "learning_rate": 2e-05,
      "loss": 0.8344195365905762,
      "step": 700
    },
    {
      "epoch": 0.02743961352657005,
      "grad_norm": 6.340736949473045,
      "learning_rate": 2e-05,
      "loss": 0.8478621482849121,
      "step": 710
    },
    {
      "epoch": 0.02782608695652174,
      "grad_norm": 84.49603950900422,
      "learning_rate": 2e-05,
      "loss": 0.8494409561157227,
      "step": 720
    },
    {
      "epoch": 0.02821256038647343,
      "grad_norm": 5.239994048119582,
      "learning_rate": 2e-05,
      "loss": 0.8430055618286133,
      "step": 730
    },
    {
      "epoch": 0.02859903381642512,
      "grad_norm": 6.81571454851921,
      "learning_rate": 2e-05,
      "loss": 0.8379552841186524,
      "step": 740
    },
    {
      "epoch": 0.028985507246376812,
      "grad_norm": 13.230544669517375,
      "learning_rate": 2e-05,
      "loss": 0.8137977600097657,
      "step": 750
    },
    {
      "epoch": 0.029371980676328503,
      "grad_norm": 10.83917272801397,
      "learning_rate": 2e-05,
      "loss": 0.8010368347167969,
      "step": 760
    },
    {
      "epoch": 0.029758454106280193,
      "grad_norm": 4.561007311907072,
      "learning_rate": 2e-05,
      "loss": 0.8132765769958497,
      "step": 770
    },
    {
      "epoch": 0.030144927536231884,
      "grad_norm": 5.052478275742948,
      "learning_rate": 2e-05,
      "loss": 0.8024440765380859,
      "step": 780
    },
    {
      "epoch": 0.030531400966183574,
      "grad_norm": 39.20779782791551,
      "learning_rate": 2e-05,
      "loss": 0.8082746505737305,
      "step": 790
    },
    {
      "epoch": 0.030917874396135265,
      "grad_norm": 6.722990812199062,
      "learning_rate": 2e-05,
      "loss": 0.7964837551116943,
      "step": 800
    },
    {
      "epoch": 0.03130434782608696,
      "grad_norm": 6.263608751717215,
      "learning_rate": 2e-05,
      "loss": 0.8117972373962402,
      "step": 810
    },
    {
      "epoch": 0.03169082125603865,
      "grad_norm": 6.362861042949148,
      "learning_rate": 2e-05,
      "loss": 0.8188319206237793,
      "step": 820
    },
    {
      "epoch": 0.03207729468599034,
      "grad_norm": 6.379235603582932,
      "learning_rate": 2e-05,
      "loss": 0.7999361991882324,
      "step": 830
    },
    {
      "epoch": 0.03246376811594203,
      "grad_norm": 7.771520162515829,
      "learning_rate": 2e-05,
      "loss": 0.8278095245361328,
      "step": 840
    },
    {
      "epoch": 0.03285024154589372,
      "grad_norm": 3.6283672901509307,
      "learning_rate": 2e-05,
      "loss": 0.8165772438049317,
      "step": 850
    },
    {
      "epoch": 0.03323671497584541,
      "grad_norm": 24.13531059310475,
      "learning_rate": 2e-05,
      "loss": 0.8282254219055176,
      "step": 860
    },
    {
      "epoch": 0.0336231884057971,
      "grad_norm": 6.38211185754252,
      "learning_rate": 2e-05,
      "loss": 0.857562255859375,
      "step": 870
    },
    {
      "epoch": 0.03400966183574879,
      "grad_norm": 8.151353292941916,
      "learning_rate": 2e-05,
      "loss": 0.8778319358825684,
      "step": 880
    },
    {
      "epoch": 0.034396135265700484,
      "grad_norm": 2.6425732312042793,
      "learning_rate": 2e-05,
      "loss": 0.8224431991577148,
      "step": 890
    },
    {
      "epoch": 0.034782608695652174,
      "grad_norm": 3.5769751025573835,
      "learning_rate": 2e-05,
      "loss": 0.8090167999267578,
      "step": 900
    },
    {
      "epoch": 0.035169082125603865,
      "grad_norm": 4.710064964796984,
      "learning_rate": 2e-05,
      "loss": 0.7755887508392334,
      "step": 910
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 2.85171049842582,
      "learning_rate": 2e-05,
      "loss": 0.80325927734375,
      "step": 920
    },
    {
      "epoch": 0.035942028985507246,
      "grad_norm": 2.7025040213443625,
      "learning_rate": 2e-05,
      "loss": 0.8224540710449219,
      "step": 930
    },
    {
      "epoch": 0.03632850241545894,
      "grad_norm": 4.365935909605305,
      "learning_rate": 2e-05,
      "loss": 0.8001300811767578,
      "step": 940
    },
    {
      "epoch": 0.03671497584541063,
      "grad_norm": 4.205991130813683,
      "learning_rate": 2e-05,
      "loss": 0.8388648986816406,
      "step": 950
    },
    {
      "epoch": 0.03710144927536232,
      "grad_norm": 9.511180584595577,
      "learning_rate": 2e-05,
      "loss": 0.8006820678710938,
      "step": 960
    },
    {
      "epoch": 0.03748792270531401,
      "grad_norm": 1.8991371909927346,
      "learning_rate": 2e-05,
      "loss": 0.7927989959716797,
      "step": 970
    },
    {
      "epoch": 0.0378743961352657,
      "grad_norm": 1.7204963535857005,
      "learning_rate": 2e-05,
      "loss": 0.7866291046142578,
      "step": 980
    },
    {
      "epoch": 0.03826086956521739,
      "grad_norm": 4.982975467165141,
      "learning_rate": 2e-05,
      "loss": 0.8541565895080566,
      "step": 990
    },
    {
      "epoch": 0.03864734299516908,
      "grad_norm": 1.4823804814569383,
      "learning_rate": 2e-05,
      "loss": 0.7768624782562256,
      "step": 1000
    },
    {
      "epoch": 0.03903381642512077,
      "grad_norm": 1.6741186213295625,
      "learning_rate": 2e-05,
      "loss": 0.7901049613952636,
      "step": 1010
    },
    {
      "epoch": 0.03942028985507246,
      "grad_norm": 2.4322679920118455,
      "learning_rate": 2e-05,
      "loss": 0.7638983726501465,
      "step": 1020
    },
    {
      "epoch": 0.03980676328502415,
      "grad_norm": 3.3701720746584383,
      "learning_rate": 2e-05,
      "loss": 0.7828890800476074,
      "step": 1030
    },
    {
      "epoch": 0.04019323671497584,
      "grad_norm": 2.0742844849026607,
      "learning_rate": 2e-05,
      "loss": 0.8018475532531738,
      "step": 1040
    },
    {
      "epoch": 0.04057971014492753,
      "grad_norm": 2.1569898099072904,
      "learning_rate": 2e-05,
      "loss": 0.7848191738128663,
      "step": 1050
    },
    {
      "epoch": 0.040966183574879224,
      "grad_norm": 1.3882337646082599,
      "learning_rate": 2e-05,
      "loss": 0.7792089462280274,
      "step": 1060
    },
    {
      "epoch": 0.041352657004830914,
      "grad_norm": 2.662094444417472,
      "learning_rate": 2e-05,
      "loss": 0.793889045715332,
      "step": 1070
    },
    {
      "epoch": 0.04173913043478261,
      "grad_norm": 2.587246330505362,
      "learning_rate": 2e-05,
      "loss": 0.7864011287689209,
      "step": 1080
    },
    {
      "epoch": 0.0421256038647343,
      "grad_norm": 1.0896303438627253,
      "learning_rate": 2e-05,
      "loss": 0.7887033462524414,
      "step": 1090
    },
    {
      "epoch": 0.04251207729468599,
      "grad_norm": 1.6180241575772059,
      "learning_rate": 2e-05,
      "loss": 0.7555316925048828,
      "step": 1100
    },
    {
      "epoch": 0.042898550724637684,
      "grad_norm": 5.15524054228135,
      "learning_rate": 2e-05,
      "loss": 0.7596688747406006,
      "step": 1110
    },
    {
      "epoch": 0.043285024154589374,
      "grad_norm": 3.791811827235934,
      "learning_rate": 2e-05,
      "loss": 0.7762317180633544,
      "step": 1120
    },
    {
      "epoch": 0.043671497584541065,
      "grad_norm": 3.8840676386572754,
      "learning_rate": 2e-05,
      "loss": 0.7993937492370605,
      "step": 1130
    },
    {
      "epoch": 0.044057971014492756,
      "grad_norm": 1.7203756023551415,
      "learning_rate": 2e-05,
      "loss": 0.8044867515563965,
      "step": 1140
    },
    {
      "epoch": 0.044444444444444446,
      "grad_norm": 1.8455815673442082,
      "learning_rate": 2e-05,
      "loss": 0.7740165710449218,
      "step": 1150
    },
    {
      "epoch": 0.04483091787439614,
      "grad_norm": 1.2497425235654502,
      "learning_rate": 2e-05,
      "loss": 0.7792739868164062,
      "step": 1160
    },
    {
      "epoch": 0.04521739130434783,
      "grad_norm": 3.582519109266027,
      "learning_rate": 2e-05,
      "loss": 0.7844886302947998,
      "step": 1170
    },
    {
      "epoch": 0.04560386473429952,
      "grad_norm": 1.7447476237477404,
      "learning_rate": 2e-05,
      "loss": 0.7678474426269531,
      "step": 1180
    },
    {
      "epoch": 0.04599033816425121,
      "grad_norm": 2.3052344816348285,
      "learning_rate": 2e-05,
      "loss": 0.7841787338256836,
      "step": 1190
    },
    {
      "epoch": 0.0463768115942029,
      "grad_norm": 6.171873978979098,
      "learning_rate": 2e-05,
      "loss": 0.7667727470397949,
      "step": 1200
    },
    {
      "epoch": 0.04676328502415459,
      "grad_norm": 2.5238752863132166,
      "learning_rate": 2e-05,
      "loss": 0.7614020824432373,
      "step": 1210
    },
    {
      "epoch": 0.04714975845410628,
      "grad_norm": 3.3006024749410345,
      "learning_rate": 2e-05,
      "loss": 0.7787034988403321,
      "step": 1220
    },
    {
      "epoch": 0.04753623188405797,
      "grad_norm": 1.7800834333389988,
      "learning_rate": 2e-05,
      "loss": 0.781245756149292,
      "step": 1230
    },
    {
      "epoch": 0.04792270531400966,
      "grad_norm": 4.086703804045758,
      "learning_rate": 2e-05,
      "loss": 0.7704942226409912,
      "step": 1240
    },
    {
      "epoch": 0.04830917874396135,
      "grad_norm": 1.8232412748458093,
      "learning_rate": 2e-05,
      "loss": 0.830727481842041,
      "step": 1250
    },
    {
      "epoch": 0.04869565217391304,
      "grad_norm": 1.438606233505821,
      "learning_rate": 2e-05,
      "loss": 0.7969734191894531,
      "step": 1260
    },
    {
      "epoch": 0.04908212560386473,
      "grad_norm": 14.874207481468607,
      "learning_rate": 2e-05,
      "loss": 0.7737374305725098,
      "step": 1270
    },
    {
      "epoch": 0.049468599033816424,
      "grad_norm": 1.8577472378752176,
      "learning_rate": 2e-05,
      "loss": 0.7808979034423829,
      "step": 1280
    },
    {
      "epoch": 0.049855072463768114,
      "grad_norm": 1.1647916231280007,
      "learning_rate": 2e-05,
      "loss": 0.7798149108886718,
      "step": 1290
    },
    {
      "epoch": 0.050241545893719805,
      "grad_norm": 1.8949891297132515,
      "learning_rate": 2e-05,
      "loss": 0.7773360252380371,
      "step": 1300
    },
    {
      "epoch": 0.050628019323671496,
      "grad_norm": 1.8051946173023403,
      "learning_rate": 2e-05,
      "loss": 0.7962361812591553,
      "step": 1310
    },
    {
      "epoch": 0.051014492753623186,
      "grad_norm": 8.517857032168914,
      "learning_rate": 2e-05,
      "loss": 0.7579530715942383,
      "step": 1320
    },
    {
      "epoch": 0.05140096618357488,
      "grad_norm": 1.2056710567353992,
      "learning_rate": 2e-05,
      "loss": 0.7956658840179444,
      "step": 1330
    },
    {
      "epoch": 0.05178743961352657,
      "grad_norm": 1.5620096177153822,
      "learning_rate": 2e-05,
      "loss": 0.7909909248352051,
      "step": 1340
    },
    {
      "epoch": 0.05217391304347826,
      "grad_norm": 1.2051599267987412,
      "learning_rate": 2e-05,
      "loss": 0.802952766418457,
      "step": 1350
    },
    {
      "epoch": 0.05256038647342995,
      "grad_norm": 1.442909003316718,
      "learning_rate": 2e-05,
      "loss": 0.792842435836792,
      "step": 1360
    },
    {
      "epoch": 0.05294685990338164,
      "grad_norm": 1.8033837673657727,
      "learning_rate": 2e-05,
      "loss": 0.8045886039733887,
      "step": 1370
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 1.7301184849803795,
      "learning_rate": 2e-05,
      "loss": 0.7730770111083984,
      "step": 1380
    },
    {
      "epoch": 0.05371980676328503,
      "grad_norm": 2.593314064685021,
      "learning_rate": 2e-05,
      "loss": 0.7898385524749756,
      "step": 1390
    },
    {
      "epoch": 0.05410628019323672,
      "grad_norm": 2.6612126149781123,
      "learning_rate": 2e-05,
      "loss": 0.7972373008728028,
      "step": 1400
    },
    {
      "epoch": 0.05449275362318841,
      "grad_norm": 4.014591492016486,
      "learning_rate": 2e-05,
      "loss": 0.7940009593963623,
      "step": 1410
    },
    {
      "epoch": 0.0548792270531401,
      "grad_norm": 1.88666817691601,
      "learning_rate": 2e-05,
      "loss": 0.7976803779602051,
      "step": 1420
    },
    {
      "epoch": 0.05526570048309179,
      "grad_norm": 1.6808638117085843,
      "learning_rate": 2e-05,
      "loss": 0.7751292228698731,
      "step": 1430
    },
    {
      "epoch": 0.05565217391304348,
      "grad_norm": 2.8607658596215573,
      "learning_rate": 2e-05,
      "loss": 0.7797608375549316,
      "step": 1440
    },
    {
      "epoch": 0.05603864734299517,
      "grad_norm": 1.743259772518007,
      "learning_rate": 2e-05,
      "loss": 0.7893115520477295,
      "step": 1450
    },
    {
      "epoch": 0.05642512077294686,
      "grad_norm": 4.094105063418816,
      "learning_rate": 2e-05,
      "loss": 0.7623452186584473,
      "step": 1460
    },
    {
      "epoch": 0.05681159420289855,
      "grad_norm": 1.0326374629324264,
      "learning_rate": 2e-05,
      "loss": 0.8003641128540039,
      "step": 1470
    },
    {
      "epoch": 0.05719806763285024,
      "grad_norm": 1.5470981584627885,
      "learning_rate": 2e-05,
      "loss": 0.7471415519714355,
      "step": 1480
    },
    {
      "epoch": 0.05758454106280193,
      "grad_norm": 4.233012132760419,
      "learning_rate": 2e-05,
      "loss": 0.8017045974731445,
      "step": 1490
    },
    {
      "epoch": 0.057971014492753624,
      "grad_norm": 1.5670919578440439,
      "learning_rate": 2e-05,
      "loss": 0.7809749603271484,
      "step": 1500
    },
    {
      "epoch": 0.058357487922705314,
      "grad_norm": 1.1146358925152047,
      "learning_rate": 2e-05,
      "loss": 0.8092234611511231,
      "step": 1510
    },
    {
      "epoch": 0.058743961352657005,
      "grad_norm": 1.8519058887339446,
      "learning_rate": 2e-05,
      "loss": 0.7762881755828858,
      "step": 1520
    },
    {
      "epoch": 0.059130434782608696,
      "grad_norm": 1.842244630499373,
      "learning_rate": 2e-05,
      "loss": 0.7717338085174561,
      "step": 1530
    },
    {
      "epoch": 0.059516908212560386,
      "grad_norm": 1.8659924669555197,
      "learning_rate": 2e-05,
      "loss": 0.8045551300048828,
      "step": 1540
    },
    {
      "epoch": 0.05990338164251208,
      "grad_norm": 1.789268515534286,
      "learning_rate": 2e-05,
      "loss": 0.802666187286377,
      "step": 1550
    },
    {
      "epoch": 0.06028985507246377,
      "grad_norm": 3.049390694136811,
      "learning_rate": 2e-05,
      "loss": 0.7977540493011475,
      "step": 1560
    },
    {
      "epoch": 0.06067632850241546,
      "grad_norm": 1.1550233137997092,
      "learning_rate": 2e-05,
      "loss": 0.759410810470581,
      "step": 1570
    },
    {
      "epoch": 0.06106280193236715,
      "grad_norm": 1.4840972042589198,
      "learning_rate": 2e-05,
      "loss": 0.7568341255187988,
      "step": 1580
    },
    {
      "epoch": 0.06144927536231884,
      "grad_norm": 2.4324918941876517,
      "learning_rate": 2e-05,
      "loss": 0.7721347808837891,
      "step": 1590
    },
    {
      "epoch": 0.06183574879227053,
      "grad_norm": 1.3236007075392429,
      "learning_rate": 2e-05,
      "loss": 0.7291332244873047,
      "step": 1600
    },
    {
      "epoch": 0.06222222222222222,
      "grad_norm": 1.8825475500130031,
      "learning_rate": 2e-05,
      "loss": 0.7765389919281006,
      "step": 1610
    },
    {
      "epoch": 0.06260869565217392,
      "grad_norm": 1.5021722283739194,
      "learning_rate": 2e-05,
      "loss": 0.7942019939422608,
      "step": 1620
    },
    {
      "epoch": 0.0629951690821256,
      "grad_norm": 1.3488892169893916,
      "learning_rate": 2e-05,
      "loss": 0.7698382377624512,
      "step": 1630
    },
    {
      "epoch": 0.0633816425120773,
      "grad_norm": 2.412898234110765,
      "learning_rate": 2e-05,
      "loss": 0.7915114402770996,
      "step": 1640
    },
    {
      "epoch": 0.06376811594202898,
      "grad_norm": 4.032177721449734,
      "learning_rate": 2e-05,
      "loss": 0.767926549911499,
      "step": 1650
    },
    {
      "epoch": 0.06415458937198068,
      "grad_norm": 3.3886389099102976,
      "learning_rate": 2e-05,
      "loss": 0.7934833526611328,
      "step": 1660
    },
    {
      "epoch": 0.06454106280193236,
      "grad_norm": 1.8196514648355098,
      "learning_rate": 2e-05,
      "loss": 0.8287544250488281,
      "step": 1670
    },
    {
      "epoch": 0.06492753623188406,
      "grad_norm": 1.933296233701157,
      "learning_rate": 2e-05,
      "loss": 0.7792496681213379,
      "step": 1680
    },
    {
      "epoch": 0.06531400966183575,
      "grad_norm": 5.5160092569417944,
      "learning_rate": 2e-05,
      "loss": 0.7874814987182617,
      "step": 1690
    },
    {
      "epoch": 0.06570048309178744,
      "grad_norm": 2.6740319109507054,
      "learning_rate": 2e-05,
      "loss": 0.7673942565917968,
      "step": 1700
    },
    {
      "epoch": 0.06608695652173913,
      "grad_norm": 4.46192092842804,
      "learning_rate": 2e-05,
      "loss": 0.7866016387939453,
      "step": 1710
    },
    {
      "epoch": 0.06647342995169082,
      "grad_norm": 9.393669028550853,
      "learning_rate": 2e-05,
      "loss": 0.7943299770355224,
      "step": 1720
    },
    {
      "epoch": 0.06685990338164251,
      "grad_norm": 2.805958882681691,
      "learning_rate": 2e-05,
      "loss": 0.7814671993255615,
      "step": 1730
    },
    {
      "epoch": 0.0672463768115942,
      "grad_norm": 8.944621163352808,
      "learning_rate": 2e-05,
      "loss": 0.7775825500488281,
      "step": 1740
    },
    {
      "epoch": 0.06763285024154589,
      "grad_norm": 1.3932999773436152,
      "learning_rate": 2e-05,
      "loss": 0.813438606262207,
      "step": 1750
    },
    {
      "epoch": 0.06801932367149759,
      "grad_norm": 1.1571499874919873,
      "learning_rate": 2e-05,
      "loss": 0.7704301834106445,
      "step": 1760
    },
    {
      "epoch": 0.06840579710144927,
      "grad_norm": 16.767188751455745,
      "learning_rate": 2e-05,
      "loss": 0.778882360458374,
      "step": 1770
    },
    {
      "epoch": 0.06879227053140097,
      "grad_norm": 1.8040160227751316,
      "learning_rate": 2e-05,
      "loss": 0.7672492980957031,
      "step": 1780
    },
    {
      "epoch": 0.06917874396135265,
      "grad_norm": 1.9003065034750897,
      "learning_rate": 2e-05,
      "loss": 0.7559861183166504,
      "step": 1790
    },
    {
      "epoch": 0.06956521739130435,
      "grad_norm": 4.328779850300501,
      "learning_rate": 2e-05,
      "loss": 0.7670548439025879,
      "step": 1800
    },
    {
      "epoch": 0.06995169082125603,
      "grad_norm": 1.5009521124497522,
      "learning_rate": 2e-05,
      "loss": 0.7466334342956543,
      "step": 1810
    },
    {
      "epoch": 0.07033816425120773,
      "grad_norm": 0.9373529421029959,
      "learning_rate": 2e-05,
      "loss": 0.7706023216247558,
      "step": 1820
    },
    {
      "epoch": 0.07072463768115941,
      "grad_norm": 2.5673019538259636,
      "learning_rate": 2e-05,
      "loss": 0.7813019752502441,
      "step": 1830
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 10.595792073325587,
      "learning_rate": 2e-05,
      "loss": 0.7712044715881348,
      "step": 1840
    },
    {
      "epoch": 0.07149758454106281,
      "grad_norm": 1.4549261759502392,
      "learning_rate": 2e-05,
      "loss": 0.7399177551269531,
      "step": 1850
    },
    {
      "epoch": 0.07188405797101449,
      "grad_norm": 1.3679794209107425,
      "learning_rate": 2e-05,
      "loss": 0.7697343349456787,
      "step": 1860
    },
    {
      "epoch": 0.07227053140096619,
      "grad_norm": 25.30685684908231,
      "learning_rate": 2e-05,
      "loss": 0.7722678661346436,
      "step": 1870
    },
    {
      "epoch": 0.07265700483091787,
      "grad_norm": 5.813362887543927,
      "learning_rate": 2e-05,
      "loss": 0.8079950332641601,
      "step": 1880
    },
    {
      "epoch": 0.07304347826086957,
      "grad_norm": 7.55091986778727,
      "learning_rate": 2e-05,
      "loss": 0.7749129295349121,
      "step": 1890
    },
    {
      "epoch": 0.07342995169082125,
      "grad_norm": 2.576691425426098,
      "learning_rate": 2e-05,
      "loss": 0.7901420593261719,
      "step": 1900
    },
    {
      "epoch": 0.07381642512077295,
      "grad_norm": 2.850989381932755,
      "learning_rate": 2e-05,
      "loss": 0.7870397567749023,
      "step": 1910
    },
    {
      "epoch": 0.07420289855072464,
      "grad_norm": 2.8388031664275593,
      "learning_rate": 2e-05,
      "loss": 0.798320722579956,
      "step": 1920
    },
    {
      "epoch": 0.07458937198067633,
      "grad_norm": 1.2660870766825572,
      "learning_rate": 2e-05,
      "loss": 0.7990159034729004,
      "step": 1930
    },
    {
      "epoch": 0.07497584541062802,
      "grad_norm": 7.521731822457599,
      "learning_rate": 2e-05,
      "loss": 0.8066888809204101,
      "step": 1940
    },
    {
      "epoch": 0.07536231884057971,
      "grad_norm": 2.2847479728174886,
      "learning_rate": 2e-05,
      "loss": 0.7807460784912109,
      "step": 1950
    },
    {
      "epoch": 0.0757487922705314,
      "grad_norm": 36.52969282187383,
      "learning_rate": 2e-05,
      "loss": 0.7917587280273437,
      "step": 1960
    },
    {
      "epoch": 0.0761352657004831,
      "grad_norm": 3.301236759706295,
      "learning_rate": 2e-05,
      "loss": 0.7987814903259277,
      "step": 1970
    },
    {
      "epoch": 0.07652173913043478,
      "grad_norm": 2.9129761845971176,
      "learning_rate": 2e-05,
      "loss": 0.7681584358215332,
      "step": 1980
    },
    {
      "epoch": 0.07690821256038648,
      "grad_norm": 2.124239464203845,
      "learning_rate": 2e-05,
      "loss": 0.7853168487548828,
      "step": 1990
    },
    {
      "epoch": 0.07729468599033816,
      "grad_norm": 6.0767545983301225,
      "learning_rate": 2e-05,
      "loss": 0.7794623374938965,
      "step": 2000
    },
    {
      "epoch": 0.07768115942028986,
      "grad_norm": 1.3709998398850736,
      "learning_rate": 2e-05,
      "loss": 0.7739840984344483,
      "step": 2010
    },
    {
      "epoch": 0.07806763285024154,
      "grad_norm": 1.5709783521988716,
      "learning_rate": 2e-05,
      "loss": 0.7280101776123047,
      "step": 2020
    },
    {
      "epoch": 0.07845410628019324,
      "grad_norm": 2.0426823491051156,
      "learning_rate": 2e-05,
      "loss": 0.7502191066741943,
      "step": 2030
    },
    {
      "epoch": 0.07884057971014492,
      "grad_norm": 3.1670032654507825,
      "learning_rate": 2e-05,
      "loss": 0.7697875022888183,
      "step": 2040
    },
    {
      "epoch": 0.07922705314009662,
      "grad_norm": 1.519486508417274,
      "learning_rate": 2e-05,
      "loss": 0.7478765487670899,
      "step": 2050
    },
    {
      "epoch": 0.0796135265700483,
      "grad_norm": 2.2518000319069476,
      "learning_rate": 2e-05,
      "loss": 0.7812594413757324,
      "step": 2060
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5740210899218194,
      "learning_rate": 2e-05,
      "loss": 0.7684998989105225,
      "step": 2070
    },
    {
      "epoch": 0.08038647342995169,
      "grad_norm": 2.0349709099256845,
      "learning_rate": 2e-05,
      "loss": 0.7811843872070312,
      "step": 2080
    },
    {
      "epoch": 0.08077294685990338,
      "grad_norm": 2.2636959526681157,
      "learning_rate": 2e-05,
      "loss": 0.7650601387023925,
      "step": 2090
    },
    {
      "epoch": 0.08115942028985507,
      "grad_norm": 3.742795215033163,
      "learning_rate": 2e-05,
      "loss": 0.7615516185760498,
      "step": 2100
    },
    {
      "epoch": 0.08154589371980676,
      "grad_norm": 1.138738386234341,
      "learning_rate": 2e-05,
      "loss": 0.7689103126525879,
      "step": 2110
    },
    {
      "epoch": 0.08193236714975845,
      "grad_norm": 1.6534859784091112,
      "learning_rate": 2e-05,
      "loss": 0.7665469646453857,
      "step": 2120
    },
    {
      "epoch": 0.08231884057971015,
      "grad_norm": 1.3318959611026695,
      "learning_rate": 2e-05,
      "loss": 0.769937801361084,
      "step": 2130
    },
    {
      "epoch": 0.08270531400966183,
      "grad_norm": 1.3547014303403682,
      "learning_rate": 2e-05,
      "loss": 0.7504668235778809,
      "step": 2140
    },
    {
      "epoch": 0.08309178743961353,
      "grad_norm": 17.52193810186475,
      "learning_rate": 2e-05,
      "loss": 0.7507330894470214,
      "step": 2150
    },
    {
      "epoch": 0.08347826086956522,
      "grad_norm": 1.8162401489013498,
      "learning_rate": 2e-05,
      "loss": 0.7667189598083496,
      "step": 2160
    },
    {
      "epoch": 0.08386473429951691,
      "grad_norm": 2.4747448006663,
      "learning_rate": 2e-05,
      "loss": 0.7687741279602051,
      "step": 2170
    },
    {
      "epoch": 0.0842512077294686,
      "grad_norm": 1.8861120110910479,
      "learning_rate": 2e-05,
      "loss": 0.7431299209594726,
      "step": 2180
    },
    {
      "epoch": 0.08463768115942029,
      "grad_norm": 1.598592624418475,
      "learning_rate": 2e-05,
      "loss": 0.7552878379821777,
      "step": 2190
    },
    {
      "epoch": 0.08502415458937199,
      "grad_norm": 2.431095132727302,
      "learning_rate": 2e-05,
      "loss": 0.734284496307373,
      "step": 2200
    },
    {
      "epoch": 0.08541062801932367,
      "grad_norm": 2.7269516195691077,
      "learning_rate": 2e-05,
      "loss": 0.7517167091369629,
      "step": 2210
    },
    {
      "epoch": 0.08579710144927537,
      "grad_norm": 1.661492853949561,
      "learning_rate": 2e-05,
      "loss": 0.7628636360168457,
      "step": 2220
    },
    {
      "epoch": 0.08618357487922705,
      "grad_norm": 2.855274848798419,
      "learning_rate": 2e-05,
      "loss": 0.7595593929290771,
      "step": 2230
    },
    {
      "epoch": 0.08657004830917875,
      "grad_norm": 1.285434436992812,
      "learning_rate": 2e-05,
      "loss": 0.7832144737243653,
      "step": 2240
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 1.2309831181727102,
      "learning_rate": 2e-05,
      "loss": 0.7642119407653809,
      "step": 2250
    },
    {
      "epoch": 0.08734299516908213,
      "grad_norm": 1.3727116026773452,
      "learning_rate": 2e-05,
      "loss": 0.7471759796142579,
      "step": 2260
    },
    {
      "epoch": 0.08772946859903381,
      "grad_norm": 1.3248395463889682,
      "learning_rate": 2e-05,
      "loss": 0.7310284614562989,
      "step": 2270
    },
    {
      "epoch": 0.08811594202898551,
      "grad_norm": 7.664609758100217,
      "learning_rate": 2e-05,
      "loss": 0.7528136253356934,
      "step": 2280
    },
    {
      "epoch": 0.0885024154589372,
      "grad_norm": 2.339105261431996,
      "learning_rate": 2e-05,
      "loss": 0.7621146202087402,
      "step": 2290
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 1.9665131401980414,
      "learning_rate": 2e-05,
      "loss": 0.7420401573181152,
      "step": 2300
    },
    {
      "epoch": 0.08927536231884058,
      "grad_norm": 1.2677168413530253,
      "learning_rate": 2e-05,
      "loss": 0.7691300868988037,
      "step": 2310
    },
    {
      "epoch": 0.08966183574879227,
      "grad_norm": 2.3372668170559647,
      "learning_rate": 2e-05,
      "loss": 0.7166595458984375,
      "step": 2320
    },
    {
      "epoch": 0.09004830917874396,
      "grad_norm": 4.592447961406041,
      "learning_rate": 2e-05,
      "loss": 0.7677947998046875,
      "step": 2330
    },
    {
      "epoch": 0.09043478260869565,
      "grad_norm": 2.431963843797448,
      "learning_rate": 2e-05,
      "loss": 0.7791108608245849,
      "step": 2340
    },
    {
      "epoch": 0.09082125603864734,
      "grad_norm": 1.5196869951311995,
      "learning_rate": 2e-05,
      "loss": 0.7593348503112793,
      "step": 2350
    },
    {
      "epoch": 0.09120772946859904,
      "grad_norm": 3.1796994866171824,
      "learning_rate": 2e-05,
      "loss": 0.7442672252655029,
      "step": 2360
    },
    {
      "epoch": 0.09159420289855072,
      "grad_norm": 4.225399402363425,
      "learning_rate": 2e-05,
      "loss": 0.7718966960906982,
      "step": 2370
    },
    {
      "epoch": 0.09198067632850242,
      "grad_norm": 1.7013977773889133,
      "learning_rate": 2e-05,
      "loss": 0.7620273113250733,
      "step": 2380
    },
    {
      "epoch": 0.0923671497584541,
      "grad_norm": 2.5304881183937313,
      "learning_rate": 2e-05,
      "loss": 0.7750027656555176,
      "step": 2390
    },
    {
      "epoch": 0.0927536231884058,
      "grad_norm": 1.5583085149697413,
      "learning_rate": 2e-05,
      "loss": 0.7332376003265381,
      "step": 2400
    },
    {
      "epoch": 0.09314009661835748,
      "grad_norm": 2.10232463079107,
      "learning_rate": 2e-05,
      "loss": 0.7827877044677735,
      "step": 2410
    },
    {
      "epoch": 0.09352657004830918,
      "grad_norm": 2.598171395066132,
      "learning_rate": 2e-05,
      "loss": 0.7785562038421631,
      "step": 2420
    },
    {
      "epoch": 0.09391304347826086,
      "grad_norm": 1.1757264697207033,
      "learning_rate": 2e-05,
      "loss": 0.7309589385986328,
      "step": 2430
    },
    {
      "epoch": 0.09429951690821256,
      "grad_norm": 1.6443213162705979,
      "learning_rate": 2e-05,
      "loss": 0.7487944602966309,
      "step": 2440
    },
    {
      "epoch": 0.09468599033816426,
      "grad_norm": 1.2468919443245798,
      "learning_rate": 2e-05,
      "loss": 0.7694966793060303,
      "step": 2450
    },
    {
      "epoch": 0.09507246376811594,
      "grad_norm": 1.464565055146109,
      "learning_rate": 2e-05,
      "loss": 0.7803127288818359,
      "step": 2460
    },
    {
      "epoch": 0.09545893719806764,
      "grad_norm": 3.8074836748750904,
      "learning_rate": 2e-05,
      "loss": 0.7282808303833008,
      "step": 2470
    },
    {
      "epoch": 0.09584541062801932,
      "grad_norm": 6.896151811673643,
      "learning_rate": 2e-05,
      "loss": 0.7671591758728027,
      "step": 2480
    },
    {
      "epoch": 0.09623188405797102,
      "grad_norm": 2.2239543445775425,
      "learning_rate": 2e-05,
      "loss": 0.7999413967132568,
      "step": 2490
    },
    {
      "epoch": 0.0966183574879227,
      "grad_norm": 5.616878725097995,
      "learning_rate": 2e-05,
      "loss": 0.7862523078918457,
      "step": 2500
    },
    {
      "epoch": 0.0970048309178744,
      "grad_norm": 1.627578685293566,
      "learning_rate": 2e-05,
      "loss": 0.7508011817932129,
      "step": 2510
    },
    {
      "epoch": 0.09739130434782609,
      "grad_norm": 0.9906826307911483,
      "learning_rate": 2e-05,
      "loss": 0.7621040344238281,
      "step": 2520
    },
    {
      "epoch": 0.09777777777777778,
      "grad_norm": 3.35224143508381,
      "learning_rate": 2e-05,
      "loss": 0.74591064453125,
      "step": 2530
    },
    {
      "epoch": 0.09816425120772947,
      "grad_norm": 2.2671864034926354,
      "learning_rate": 2e-05,
      "loss": 0.7632014274597168,
      "step": 2540
    },
    {
      "epoch": 0.09855072463768116,
      "grad_norm": 1.267319283410017,
      "learning_rate": 2e-05,
      "loss": 0.7705909729003906,
      "step": 2550
    },
    {
      "epoch": 0.09893719806763285,
      "grad_norm": 2.8927069345813985,
      "learning_rate": 2e-05,
      "loss": 0.7812301635742187,
      "step": 2560
    },
    {
      "epoch": 0.09932367149758455,
      "grad_norm": 1.4493285796125683,
      "learning_rate": 2e-05,
      "loss": 0.7883975028991699,
      "step": 2570
    },
    {
      "epoch": 0.09971014492753623,
      "grad_norm": 1.2446418817300966,
      "learning_rate": 2e-05,
      "loss": 0.7746667385101318,
      "step": 2580
    },
    {
      "epoch": 0.10009661835748793,
      "grad_norm": 3.217368686501776,
      "learning_rate": 2e-05,
      "loss": 0.7654871940612793,
      "step": 2590
    },
    {
      "epoch": 0.10048309178743961,
      "grad_norm": 1.1883893155114615,
      "learning_rate": 2e-05,
      "loss": 0.7502057075500488,
      "step": 2600
    },
    {
      "epoch": 0.10086956521739131,
      "grad_norm": 1.695556949452006,
      "learning_rate": 2e-05,
      "loss": 0.7216050148010253,
      "step": 2610
    },
    {
      "epoch": 0.10125603864734299,
      "grad_norm": 3.473686297073087,
      "learning_rate": 2e-05,
      "loss": 0.7521714210510254,
      "step": 2620
    },
    {
      "epoch": 0.10164251207729469,
      "grad_norm": 5.9788011582249805,
      "learning_rate": 2e-05,
      "loss": 0.7292653560638428,
      "step": 2630
    },
    {
      "epoch": 0.10202898550724637,
      "grad_norm": 7.808440837834781,
      "learning_rate": 2e-05,
      "loss": 0.7407905101776123,
      "step": 2640
    },
    {
      "epoch": 0.10241545893719807,
      "grad_norm": 2.5686061602885637,
      "learning_rate": 2e-05,
      "loss": 0.7778966426849365,
      "step": 2650
    },
    {
      "epoch": 0.10280193236714975,
      "grad_norm": 3.154626905551962,
      "learning_rate": 2e-05,
      "loss": 0.7876219749450684,
      "step": 2660
    },
    {
      "epoch": 0.10318840579710145,
      "grad_norm": 3.468843923168503,
      "learning_rate": 2e-05,
      "loss": 0.7734580039978027,
      "step": 2670
    },
    {
      "epoch": 0.10357487922705313,
      "grad_norm": 8.009584748577065,
      "learning_rate": 2e-05,
      "loss": 0.7431020259857177,
      "step": 2680
    },
    {
      "epoch": 0.10396135265700483,
      "grad_norm": 2.6545478840796384,
      "learning_rate": 2e-05,
      "loss": 0.7354369163513184,
      "step": 2690
    },
    {
      "epoch": 0.10434782608695652,
      "grad_norm": 3.013460214296052,
      "learning_rate": 2e-05,
      "loss": 0.7398846626281739,
      "step": 2700
    },
    {
      "epoch": 0.10473429951690821,
      "grad_norm": 2.8697969948962103,
      "learning_rate": 2e-05,
      "loss": 0.774082088470459,
      "step": 2710
    },
    {
      "epoch": 0.1051207729468599,
      "grad_norm": 4.130301132781164,
      "learning_rate": 2e-05,
      "loss": 0.7540542125701905,
      "step": 2720
    },
    {
      "epoch": 0.1055072463768116,
      "grad_norm": 3.24272431858663,
      "learning_rate": 2e-05,
      "loss": 0.7488170146942139,
      "step": 2730
    },
    {
      "epoch": 0.10589371980676328,
      "grad_norm": 5.974999494911175,
      "learning_rate": 2e-05,
      "loss": 0.7478822231292724,
      "step": 2740
    },
    {
      "epoch": 0.10628019323671498,
      "grad_norm": 10.37771215379445,
      "learning_rate": 2e-05,
      "loss": 0.7524117469787598,
      "step": 2750
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 2.3430180737894437,
      "learning_rate": 2e-05,
      "loss": 0.7408059597015381,
      "step": 2760
    },
    {
      "epoch": 0.10705314009661836,
      "grad_norm": 2.518159773041028,
      "learning_rate": 2e-05,
      "loss": 0.7606561183929443,
      "step": 2770
    },
    {
      "epoch": 0.10743961352657005,
      "grad_norm": 4.228058376115772,
      "learning_rate": 2e-05,
      "loss": 0.7550743103027344,
      "step": 2780
    },
    {
      "epoch": 0.10782608695652174,
      "grad_norm": 2.0014204387088883,
      "learning_rate": 2e-05,
      "loss": 0.7893956184387207,
      "step": 2790
    },
    {
      "epoch": 0.10821256038647344,
      "grad_norm": 1.4580798360101905,
      "learning_rate": 2e-05,
      "loss": 0.7935238838195801,
      "step": 2800
    },
    {
      "epoch": 0.10859903381642512,
      "grad_norm": 1.440341687449925,
      "learning_rate": 2e-05,
      "loss": 0.7824203491210937,
      "step": 2810
    },
    {
      "epoch": 0.10898550724637682,
      "grad_norm": 1.4960713444728275,
      "learning_rate": 2e-05,
      "loss": 0.7783860683441162,
      "step": 2820
    },
    {
      "epoch": 0.1093719806763285,
      "grad_norm": 4.912105643942074,
      "learning_rate": 2e-05,
      "loss": 0.7596704483032226,
      "step": 2830
    },
    {
      "epoch": 0.1097584541062802,
      "grad_norm": 4.039027525263609,
      "learning_rate": 2e-05,
      "loss": 0.7591814994812012,
      "step": 2840
    },
    {
      "epoch": 0.11014492753623188,
      "grad_norm": 1.5391399654261342,
      "learning_rate": 2e-05,
      "loss": 0.7540448665618896,
      "step": 2850
    },
    {
      "epoch": 0.11053140096618358,
      "grad_norm": 1.3907629383677877,
      "learning_rate": 2e-05,
      "loss": 0.7578011989593506,
      "step": 2860
    },
    {
      "epoch": 0.11091787439613526,
      "grad_norm": 2.418756458605362,
      "learning_rate": 2e-05,
      "loss": 0.7361119747161865,
      "step": 2870
    },
    {
      "epoch": 0.11130434782608696,
      "grad_norm": 3.739348851910294,
      "learning_rate": 2e-05,
      "loss": 0.7639998912811279,
      "step": 2880
    },
    {
      "epoch": 0.11169082125603864,
      "grad_norm": 1.670125865609373,
      "learning_rate": 2e-05,
      "loss": 0.7774595260620117,
      "step": 2890
    },
    {
      "epoch": 0.11207729468599034,
      "grad_norm": 1.7170327776104877,
      "learning_rate": 2e-05,
      "loss": 0.6949478149414062,
      "step": 2900
    },
    {
      "epoch": 0.11246376811594203,
      "grad_norm": 2.4782554453410786,
      "learning_rate": 2e-05,
      "loss": 0.771298885345459,
      "step": 2910
    },
    {
      "epoch": 0.11285024154589372,
      "grad_norm": 2.721787201809589,
      "learning_rate": 2e-05,
      "loss": 0.7795491695404053,
      "step": 2920
    },
    {
      "epoch": 0.1132367149758454,
      "grad_norm": 1.8882417313191724,
      "learning_rate": 2e-05,
      "loss": 0.7629229545593261,
      "step": 2930
    },
    {
      "epoch": 0.1136231884057971,
      "grad_norm": 1.020404408117238,
      "learning_rate": 2e-05,
      "loss": 0.7520963191986084,
      "step": 2940
    },
    {
      "epoch": 0.11400966183574879,
      "grad_norm": 1.2953592621672139,
      "learning_rate": 2e-05,
      "loss": 0.7835359096527099,
      "step": 2950
    },
    {
      "epoch": 0.11439613526570049,
      "grad_norm": 1.402126121933966,
      "learning_rate": 2e-05,
      "loss": 0.7469770431518554,
      "step": 2960
    },
    {
      "epoch": 0.11478260869565217,
      "grad_norm": 1.0565102242948623,
      "learning_rate": 2e-05,
      "loss": 0.7419232845306396,
      "step": 2970
    },
    {
      "epoch": 0.11516908212560387,
      "grad_norm": 1.2672419412998372,
      "learning_rate": 2e-05,
      "loss": 0.7683830261230469,
      "step": 2980
    },
    {
      "epoch": 0.11555555555555555,
      "grad_norm": 2.401558201984435,
      "learning_rate": 2e-05,
      "loss": 0.7525971889495849,
      "step": 2990
    },
    {
      "epoch": 0.11594202898550725,
      "grad_norm": 1.526434742827244,
      "learning_rate": 2e-05,
      "loss": 0.7671350002288818,
      "step": 3000
    },
    {
      "epoch": 0.11632850241545893,
      "grad_norm": 1.3426982308084237,
      "learning_rate": 2e-05,
      "loss": 0.7217348098754883,
      "step": 3010
    },
    {
      "epoch": 0.11671497584541063,
      "grad_norm": 2.5859563913304275,
      "learning_rate": 2e-05,
      "loss": 0.724910306930542,
      "step": 3020
    },
    {
      "epoch": 0.11710144927536231,
      "grad_norm": 1.6037857614752977,
      "learning_rate": 2e-05,
      "loss": 0.7487847805023193,
      "step": 3030
    },
    {
      "epoch": 0.11748792270531401,
      "grad_norm": 2.6902562084555504,
      "learning_rate": 2e-05,
      "loss": 0.7263116359710693,
      "step": 3040
    },
    {
      "epoch": 0.1178743961352657,
      "grad_norm": 2.2878601990703658,
      "learning_rate": 2e-05,
      "loss": 0.7548857688903808,
      "step": 3050
    },
    {
      "epoch": 0.11826086956521739,
      "grad_norm": 2.0090920961910994,
      "learning_rate": 2e-05,
      "loss": 0.7582263469696044,
      "step": 3060
    },
    {
      "epoch": 0.11864734299516909,
      "grad_norm": 4.636914226969478,
      "learning_rate": 2e-05,
      "loss": 0.7098802089691162,
      "step": 3070
    },
    {
      "epoch": 0.11903381642512077,
      "grad_norm": 10.632169657924953,
      "learning_rate": 2e-05,
      "loss": 0.7436557292938233,
      "step": 3080
    },
    {
      "epoch": 0.11942028985507247,
      "grad_norm": 4.092776992680432,
      "learning_rate": 2e-05,
      "loss": 0.7685037136077881,
      "step": 3090
    },
    {
      "epoch": 0.11980676328502415,
      "grad_norm": 2.2184443588512455,
      "learning_rate": 2e-05,
      "loss": 0.7629373073577881,
      "step": 3100
    },
    {
      "epoch": 0.12019323671497585,
      "grad_norm": 5.981987414163901,
      "learning_rate": 2e-05,
      "loss": 0.7763461112976074,
      "step": 3110
    },
    {
      "epoch": 0.12057971014492753,
      "grad_norm": 1.583113847170495,
      "learning_rate": 2e-05,
      "loss": 0.7681999206542969,
      "step": 3120
    },
    {
      "epoch": 0.12096618357487923,
      "grad_norm": 1.481558880113795,
      "learning_rate": 2e-05,
      "loss": 0.7741986751556397,
      "step": 3130
    },
    {
      "epoch": 0.12135265700483092,
      "grad_norm": 1.9151056127493966,
      "learning_rate": 2e-05,
      "loss": 0.7652425289154052,
      "step": 3140
    },
    {
      "epoch": 0.12173913043478261,
      "grad_norm": 2.387560581863014,
      "learning_rate": 2e-05,
      "loss": 0.7649560451507569,
      "step": 3150
    },
    {
      "epoch": 0.1221256038647343,
      "grad_norm": 1.1173760546950033,
      "learning_rate": 2e-05,
      "loss": 0.7181439876556397,
      "step": 3160
    },
    {
      "epoch": 0.122512077294686,
      "grad_norm": 2.083761932159444,
      "learning_rate": 2e-05,
      "loss": 0.7495922088623047,
      "step": 3170
    },
    {
      "epoch": 0.12289855072463768,
      "grad_norm": 1.1652939980593255,
      "learning_rate": 2e-05,
      "loss": 0.7515378475189209,
      "step": 3180
    },
    {
      "epoch": 0.12328502415458938,
      "grad_norm": 1.4918368808564153,
      "learning_rate": 2e-05,
      "loss": 0.7546276092529297,
      "step": 3190
    },
    {
      "epoch": 0.12367149758454106,
      "grad_norm": 1.0293176663426336,
      "learning_rate": 2e-05,
      "loss": 0.7552709579467773,
      "step": 3200
    },
    {
      "epoch": 0.12405797101449276,
      "grad_norm": 1.1057665645933867,
      "learning_rate": 2e-05,
      "loss": 0.741522741317749,
      "step": 3210
    },
    {
      "epoch": 0.12444444444444444,
      "grad_norm": 0.9363596854554345,
      "learning_rate": 2e-05,
      "loss": 0.7667788982391357,
      "step": 3220
    },
    {
      "epoch": 0.12483091787439614,
      "grad_norm": 1.04985033716726,
      "learning_rate": 2e-05,
      "loss": 0.7525820732116699,
      "step": 3230
    },
    {
      "epoch": 0.12521739130434784,
      "grad_norm": 7.721167519610794,
      "learning_rate": 2e-05,
      "loss": 0.7389856338500976,
      "step": 3240
    },
    {
      "epoch": 0.12560386473429952,
      "grad_norm": 1.1895781786282957,
      "learning_rate": 2e-05,
      "loss": 0.7608234405517578,
      "step": 3250
    },
    {
      "epoch": 0.1259903381642512,
      "grad_norm": 0.9818564982551652,
      "learning_rate": 2e-05,
      "loss": 0.7519477844238281,
      "step": 3260
    },
    {
      "epoch": 0.1263768115942029,
      "grad_norm": 1.0761562982208566,
      "learning_rate": 2e-05,
      "loss": 0.7019442558288574,
      "step": 3270
    },
    {
      "epoch": 0.1267632850241546,
      "grad_norm": 1.373293148249932,
      "learning_rate": 2e-05,
      "loss": 0.7445249557495117,
      "step": 3280
    },
    {
      "epoch": 0.12714975845410628,
      "grad_norm": 1.1099005212509954,
      "learning_rate": 2e-05,
      "loss": 0.7563081741333008,
      "step": 3290
    },
    {
      "epoch": 0.12753623188405797,
      "grad_norm": 1.046722701612444,
      "learning_rate": 2e-05,
      "loss": 0.7404123306274414,
      "step": 3300
    },
    {
      "epoch": 0.12792270531400965,
      "grad_norm": 0.9518798965383707,
      "learning_rate": 2e-05,
      "loss": 0.740294361114502,
      "step": 3310
    },
    {
      "epoch": 0.12830917874396136,
      "grad_norm": 0.9266301254714255,
      "learning_rate": 2e-05,
      "loss": 0.7460558891296387,
      "step": 3320
    },
    {
      "epoch": 0.12869565217391304,
      "grad_norm": 1.4702716849340487,
      "learning_rate": 2e-05,
      "loss": 0.7604526519775391,
      "step": 3330
    },
    {
      "epoch": 0.12908212560386473,
      "grad_norm": 0.9102929590923553,
      "learning_rate": 2e-05,
      "loss": 0.7236586570739746,
      "step": 3340
    },
    {
      "epoch": 0.1294685990338164,
      "grad_norm": 1.6391901805267906,
      "learning_rate": 2e-05,
      "loss": 0.748450756072998,
      "step": 3350
    },
    {
      "epoch": 0.12985507246376812,
      "grad_norm": 1.6075723054223627,
      "learning_rate": 2e-05,
      "loss": 0.7403520584106446,
      "step": 3360
    },
    {
      "epoch": 0.1302415458937198,
      "grad_norm": 1.5094379479330864,
      "learning_rate": 2e-05,
      "loss": 0.7653777122497558,
      "step": 3370
    },
    {
      "epoch": 0.1306280193236715,
      "grad_norm": 7.0703944204633125,
      "learning_rate": 2e-05,
      "loss": 0.7486815452575684,
      "step": 3380
    },
    {
      "epoch": 0.1310144927536232,
      "grad_norm": 1.1641301126393144,
      "learning_rate": 2e-05,
      "loss": 0.7190661907196045,
      "step": 3390
    },
    {
      "epoch": 0.13140096618357489,
      "grad_norm": 3.2432498754052586,
      "learning_rate": 2e-05,
      "loss": 0.7289729595184327,
      "step": 3400
    },
    {
      "epoch": 0.13178743961352657,
      "grad_norm": 1.7875108272154512,
      "learning_rate": 2e-05,
      "loss": 0.7148953437805176,
      "step": 3410
    },
    {
      "epoch": 0.13217391304347825,
      "grad_norm": 2.3978591933447944,
      "learning_rate": 2e-05,
      "loss": 0.7307319641113281,
      "step": 3420
    },
    {
      "epoch": 0.13256038647342996,
      "grad_norm": 1.090192637666955,
      "learning_rate": 2e-05,
      "loss": 0.719948673248291,
      "step": 3430
    },
    {
      "epoch": 0.13294685990338165,
      "grad_norm": 2.38275417744872,
      "learning_rate": 2e-05,
      "loss": 0.7476545333862304,
      "step": 3440
    },
    {
      "epoch": 0.13333333333333333,
      "grad_norm": 1.3431211574818809,
      "learning_rate": 2e-05,
      "loss": 0.756253719329834,
      "step": 3450
    },
    {
      "epoch": 0.13371980676328502,
      "grad_norm": 3.805683855046064,
      "learning_rate": 2e-05,
      "loss": 0.7468399047851563,
      "step": 3460
    },
    {
      "epoch": 0.13410628019323673,
      "grad_norm": 2.2843861478547156,
      "learning_rate": 2e-05,
      "loss": 0.7557847499847412,
      "step": 3470
    },
    {
      "epoch": 0.1344927536231884,
      "grad_norm": 1.2152156459163166,
      "learning_rate": 2e-05,
      "loss": 0.7322972774505615,
      "step": 3480
    },
    {
      "epoch": 0.1348792270531401,
      "grad_norm": 1.1054501561944003,
      "learning_rate": 2e-05,
      "loss": 0.7517436027526856,
      "step": 3490
    },
    {
      "epoch": 0.13526570048309178,
      "grad_norm": 1.091680834103932,
      "learning_rate": 2e-05,
      "loss": 0.7240757942199707,
      "step": 3500
    },
    {
      "epoch": 0.1356521739130435,
      "grad_norm": 1.3237009823355679,
      "learning_rate": 2e-05,
      "loss": 0.7183950424194336,
      "step": 3510
    },
    {
      "epoch": 0.13603864734299517,
      "grad_norm": 1.370967554841046,
      "learning_rate": 2e-05,
      "loss": 0.7314060211181641,
      "step": 3520
    },
    {
      "epoch": 0.13642512077294686,
      "grad_norm": 2.723221150680608,
      "learning_rate": 2e-05,
      "loss": 0.7527001857757568,
      "step": 3530
    },
    {
      "epoch": 0.13681159420289854,
      "grad_norm": 1.643858518959685,
      "learning_rate": 2e-05,
      "loss": 0.7089616775512695,
      "step": 3540
    },
    {
      "epoch": 0.13719806763285025,
      "grad_norm": 1.8011121015120353,
      "learning_rate": 2e-05,
      "loss": 0.7594071865081787,
      "step": 3550
    },
    {
      "epoch": 0.13758454106280193,
      "grad_norm": 1.4392224952583492,
      "learning_rate": 2e-05,
      "loss": 0.7581882476806641,
      "step": 3560
    },
    {
      "epoch": 0.13797101449275362,
      "grad_norm": 1.865868005098526,
      "learning_rate": 2e-05,
      "loss": 0.7558856964111328,
      "step": 3570
    },
    {
      "epoch": 0.1383574879227053,
      "grad_norm": 1.6923737573552522,
      "learning_rate": 2e-05,
      "loss": 0.7306047439575195,
      "step": 3580
    },
    {
      "epoch": 0.138743961352657,
      "grad_norm": 1.88507424299738,
      "learning_rate": 2e-05,
      "loss": 0.7101869106292724,
      "step": 3590
    },
    {
      "epoch": 0.1391304347826087,
      "grad_norm": 0.9924120914192339,
      "learning_rate": 2e-05,
      "loss": 0.7175912380218505,
      "step": 3600
    },
    {
      "epoch": 0.13951690821256038,
      "grad_norm": 0.8927761483995479,
      "learning_rate": 2e-05,
      "loss": 0.7569256782531738,
      "step": 3610
    },
    {
      "epoch": 0.13990338164251206,
      "grad_norm": 1.3692755664777077,
      "learning_rate": 2e-05,
      "loss": 0.7137792110443115,
      "step": 3620
    },
    {
      "epoch": 0.14028985507246378,
      "grad_norm": 1.0524944418131104,
      "learning_rate": 2e-05,
      "loss": 0.744718360900879,
      "step": 3630
    },
    {
      "epoch": 0.14067632850241546,
      "grad_norm": 2.0607924335998846,
      "learning_rate": 2e-05,
      "loss": 0.7238254547119141,
      "step": 3640
    },
    {
      "epoch": 0.14106280193236714,
      "grad_norm": 0.9979369127767396,
      "learning_rate": 2e-05,
      "loss": 0.7257954597473144,
      "step": 3650
    },
    {
      "epoch": 0.14144927536231883,
      "grad_norm": 1.055922473652503,
      "learning_rate": 2e-05,
      "loss": 0.7422584533691406,
      "step": 3660
    },
    {
      "epoch": 0.14183574879227054,
      "grad_norm": 1.5689729751838162,
      "learning_rate": 2e-05,
      "loss": 0.7416232585906982,
      "step": 3670
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 1.8868421803920619,
      "learning_rate": 2e-05,
      "loss": 0.7324557304382324,
      "step": 3680
    },
    {
      "epoch": 0.1426086956521739,
      "grad_norm": 12.329809011991264,
      "learning_rate": 2e-05,
      "loss": 0.7361815929412842,
      "step": 3690
    },
    {
      "epoch": 0.14299516908212562,
      "grad_norm": 1.0839203997009181,
      "learning_rate": 2e-05,
      "loss": 0.7092767238616944,
      "step": 3700
    },
    {
      "epoch": 0.1433816425120773,
      "grad_norm": 0.9551779050924308,
      "learning_rate": 2e-05,
      "loss": 0.7078923225402832,
      "step": 3710
    },
    {
      "epoch": 0.14376811594202898,
      "grad_norm": 2.91617148391284,
      "learning_rate": 2e-05,
      "loss": 0.6721621513366699,
      "step": 3720
    },
    {
      "epoch": 0.14415458937198067,
      "grad_norm": 0.8751905547586575,
      "learning_rate": 2e-05,
      "loss": 0.7346085548400879,
      "step": 3730
    },
    {
      "epoch": 0.14454106280193238,
      "grad_norm": 2.454493767312037,
      "learning_rate": 2e-05,
      "loss": 0.7111119747161865,
      "step": 3740
    },
    {
      "epoch": 0.14492753623188406,
      "grad_norm": 1.098230072601708,
      "learning_rate": 2e-05,
      "loss": 0.7394096374511718,
      "step": 3750
    },
    {
      "epoch": 0.14531400966183575,
      "grad_norm": 0.847309564400027,
      "learning_rate": 2e-05,
      "loss": 0.745591688156128,
      "step": 3760
    },
    {
      "epoch": 0.14570048309178743,
      "grad_norm": 0.8994061014918723,
      "learning_rate": 2e-05,
      "loss": 0.7250127792358398,
      "step": 3770
    },
    {
      "epoch": 0.14608695652173914,
      "grad_norm": 1.3998490193646915,
      "learning_rate": 2e-05,
      "loss": 0.735257625579834,
      "step": 3780
    },
    {
      "epoch": 0.14647342995169083,
      "grad_norm": 0.9361364759050644,
      "learning_rate": 2e-05,
      "loss": 0.7239673614501954,
      "step": 3790
    },
    {
      "epoch": 0.1468599033816425,
      "grad_norm": 1.2674676404999363,
      "learning_rate": 2e-05,
      "loss": 0.7421964645385742,
      "step": 3800
    },
    {
      "epoch": 0.1472463768115942,
      "grad_norm": 1.0867546663958008,
      "learning_rate": 2e-05,
      "loss": 0.7409372329711914,
      "step": 3810
    },
    {
      "epoch": 0.1476328502415459,
      "grad_norm": 0.9949696992154828,
      "learning_rate": 2e-05,
      "loss": 0.7259463787078857,
      "step": 3820
    },
    {
      "epoch": 0.1480193236714976,
      "grad_norm": 0.883986580274618,
      "learning_rate": 2e-05,
      "loss": 0.7099449157714843,
      "step": 3830
    },
    {
      "epoch": 0.14840579710144927,
      "grad_norm": 2.2514274137395276,
      "learning_rate": 2e-05,
      "loss": 0.7430593490600585,
      "step": 3840
    },
    {
      "epoch": 0.14879227053140096,
      "grad_norm": 1.0936205656684879,
      "learning_rate": 2e-05,
      "loss": 0.7251758098602294,
      "step": 3850
    },
    {
      "epoch": 0.14917874396135267,
      "grad_norm": 1.383671343879474,
      "learning_rate": 2e-05,
      "loss": 0.7615509986877441,
      "step": 3860
    },
    {
      "epoch": 0.14956521739130435,
      "grad_norm": 1.443464485447786,
      "learning_rate": 2e-05,
      "loss": 0.7056533813476562,
      "step": 3870
    },
    {
      "epoch": 0.14995169082125603,
      "grad_norm": 1.064261115918148,
      "learning_rate": 2e-05,
      "loss": 0.71234130859375,
      "step": 3880
    },
    {
      "epoch": 0.15033816425120772,
      "grad_norm": 1.0827805440023188,
      "learning_rate": 2e-05,
      "loss": 0.7419578552246093,
      "step": 3890
    },
    {
      "epoch": 0.15072463768115943,
      "grad_norm": 1.6136927301190087,
      "learning_rate": 2e-05,
      "loss": 0.707129955291748,
      "step": 3900
    },
    {
      "epoch": 0.1511111111111111,
      "grad_norm": 1.249721503946207,
      "learning_rate": 2e-05,
      "loss": 0.731418228149414,
      "step": 3910
    },
    {
      "epoch": 0.1514975845410628,
      "grad_norm": 3.687096122389948,
      "learning_rate": 2e-05,
      "loss": 0.7409755229949951,
      "step": 3920
    },
    {
      "epoch": 0.15188405797101448,
      "grad_norm": 1.5686049606063899,
      "learning_rate": 2e-05,
      "loss": 0.7458938121795654,
      "step": 3930
    },
    {
      "epoch": 0.1522705314009662,
      "grad_norm": 3.820752682189002,
      "learning_rate": 2e-05,
      "loss": 0.720269775390625,
      "step": 3940
    },
    {
      "epoch": 0.15265700483091788,
      "grad_norm": 1.1513505198950094,
      "learning_rate": 2e-05,
      "loss": 0.6879904747009278,
      "step": 3950
    },
    {
      "epoch": 0.15304347826086956,
      "grad_norm": 9.485762994064826,
      "learning_rate": 2e-05,
      "loss": 0.7602463722229004,
      "step": 3960
    },
    {
      "epoch": 0.15342995169082124,
      "grad_norm": 1.8772387698538013,
      "learning_rate": 2e-05,
      "loss": 0.7675003051757813,
      "step": 3970
    },
    {
      "epoch": 0.15381642512077295,
      "grad_norm": 1.355858239389418,
      "learning_rate": 2e-05,
      "loss": 0.7349637031555176,
      "step": 3980
    },
    {
      "epoch": 0.15420289855072464,
      "grad_norm": 1.8300137058507315,
      "learning_rate": 2e-05,
      "loss": 0.741055679321289,
      "step": 3990
    },
    {
      "epoch": 0.15458937198067632,
      "grad_norm": 1.411446637818369,
      "learning_rate": 2e-05,
      "loss": 0.7304685115814209,
      "step": 4000
    },
    {
      "epoch": 0.15497584541062803,
      "grad_norm": 1.3254261860847476,
      "learning_rate": 2e-05,
      "loss": 0.7354779720306397,
      "step": 4010
    },
    {
      "epoch": 0.15536231884057972,
      "grad_norm": 1.9083967216165392,
      "learning_rate": 2e-05,
      "loss": 0.7637022972106934,
      "step": 4020
    },
    {
      "epoch": 0.1557487922705314,
      "grad_norm": 1.1440526920208254,
      "learning_rate": 2e-05,
      "loss": 0.7484094619750976,
      "step": 4030
    },
    {
      "epoch": 0.15613526570048308,
      "grad_norm": 0.8919705614319361,
      "learning_rate": 2e-05,
      "loss": 0.7086593627929687,
      "step": 4040
    },
    {
      "epoch": 0.1565217391304348,
      "grad_norm": 1.6998920732383438,
      "learning_rate": 2e-05,
      "loss": 0.7250052928924561,
      "step": 4050
    },
    {
      "epoch": 0.15690821256038648,
      "grad_norm": 1.3826627270101395,
      "learning_rate": 2e-05,
      "loss": 0.726510763168335,
      "step": 4060
    },
    {
      "epoch": 0.15729468599033816,
      "grad_norm": 1.6908756019041942,
      "learning_rate": 2e-05,
      "loss": 0.7307948112487793,
      "step": 4070
    },
    {
      "epoch": 0.15768115942028985,
      "grad_norm": 1.2834347576356087,
      "learning_rate": 2e-05,
      "loss": 0.7643456935882569,
      "step": 4080
    },
    {
      "epoch": 0.15806763285024156,
      "grad_norm": 1.7335837582526523,
      "learning_rate": 2e-05,
      "loss": 0.734253978729248,
      "step": 4090
    },
    {
      "epoch": 0.15845410628019324,
      "grad_norm": 1.264881816969034,
      "learning_rate": 2e-05,
      "loss": 0.7155088424682617,
      "step": 4100
    },
    {
      "epoch": 0.15884057971014492,
      "grad_norm": 5.3282018922333805,
      "learning_rate": 2e-05,
      "loss": 0.7114031791687012,
      "step": 4110
    },
    {
      "epoch": 0.1592270531400966,
      "grad_norm": 10.84993287289751,
      "learning_rate": 2e-05,
      "loss": 0.7341934204101562,
      "step": 4120
    },
    {
      "epoch": 0.15961352657004832,
      "grad_norm": 1.2320004587082407,
      "learning_rate": 2e-05,
      "loss": 0.740689468383789,
      "step": 4130
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.0614425761794013,
      "learning_rate": 2e-05,
      "loss": 0.7513533592224121,
      "step": 4140
    },
    {
      "epoch": 0.1603864734299517,
      "grad_norm": 1.634821742258226,
      "learning_rate": 2e-05,
      "loss": 0.7651058197021484,
      "step": 4150
    },
    {
      "epoch": 0.16077294685990337,
      "grad_norm": 1.716758350261656,
      "learning_rate": 2e-05,
      "loss": 0.7156450271606445,
      "step": 4160
    },
    {
      "epoch": 0.16115942028985508,
      "grad_norm": 3.176573221044488,
      "learning_rate": 2e-05,
      "loss": 0.7619912624359131,
      "step": 4170
    },
    {
      "epoch": 0.16154589371980677,
      "grad_norm": 2.1083134407282533,
      "learning_rate": 2e-05,
      "loss": 0.7010386943817138,
      "step": 4180
    },
    {
      "epoch": 0.16193236714975845,
      "grad_norm": 1.3313041883595207,
      "learning_rate": 2e-05,
      "loss": 0.7185277462005615,
      "step": 4190
    },
    {
      "epoch": 0.16231884057971013,
      "grad_norm": 2.3828731831726957,
      "learning_rate": 2e-05,
      "loss": 0.7398797512054444,
      "step": 4200
    },
    {
      "epoch": 0.16270531400966184,
      "grad_norm": 1.3948487740846882,
      "learning_rate": 2e-05,
      "loss": 0.7273225307464599,
      "step": 4210
    },
    {
      "epoch": 0.16309178743961353,
      "grad_norm": 1.9157802821526446,
      "learning_rate": 2e-05,
      "loss": 0.7060066223144531,
      "step": 4220
    },
    {
      "epoch": 0.1634782608695652,
      "grad_norm": 1.4825425014513973,
      "learning_rate": 2e-05,
      "loss": 0.7114836692810058,
      "step": 4230
    },
    {
      "epoch": 0.1638647342995169,
      "grad_norm": 6.531364799275646,
      "learning_rate": 2e-05,
      "loss": 0.7266668319702149,
      "step": 4240
    },
    {
      "epoch": 0.1642512077294686,
      "grad_norm": 1.2356440846729588,
      "learning_rate": 2e-05,
      "loss": 0.7722157955169677,
      "step": 4250
    },
    {
      "epoch": 0.1646376811594203,
      "grad_norm": 1.1531435236410732,
      "learning_rate": 2e-05,
      "loss": 0.7130059719085693,
      "step": 4260
    },
    {
      "epoch": 0.16502415458937197,
      "grad_norm": 1.0915337652863544,
      "learning_rate": 2e-05,
      "loss": 0.738848876953125,
      "step": 4270
    },
    {
      "epoch": 0.16541062801932366,
      "grad_norm": 0.9786081405607182,
      "learning_rate": 2e-05,
      "loss": 0.7572171211242675,
      "step": 4280
    },
    {
      "epoch": 0.16579710144927537,
      "grad_norm": 1.177926092754115,
      "learning_rate": 2e-05,
      "loss": 0.7173291683197022,
      "step": 4290
    },
    {
      "epoch": 0.16618357487922705,
      "grad_norm": 1.9449074464967013,
      "learning_rate": 2e-05,
      "loss": 0.6851347923278809,
      "step": 4300
    },
    {
      "epoch": 0.16657004830917874,
      "grad_norm": 1.051251341590575,
      "learning_rate": 2e-05,
      "loss": 0.717801284790039,
      "step": 4310
    },
    {
      "epoch": 0.16695652173913045,
      "grad_norm": 1.2373613011626008,
      "learning_rate": 2e-05,
      "loss": 0.7180459499359131,
      "step": 4320
    },
    {
      "epoch": 0.16734299516908213,
      "grad_norm": 11.000647180261412,
      "learning_rate": 2e-05,
      "loss": 0.75267333984375,
      "step": 4330
    },
    {
      "epoch": 0.16772946859903382,
      "grad_norm": 0.9034961884432832,
      "learning_rate": 2e-05,
      "loss": 0.7531850337982178,
      "step": 4340
    },
    {
      "epoch": 0.1681159420289855,
      "grad_norm": 1.0858520889607721,
      "learning_rate": 2e-05,
      "loss": 0.7531901359558105,
      "step": 4350
    },
    {
      "epoch": 0.1685024154589372,
      "grad_norm": 1.3661820052772764,
      "learning_rate": 2e-05,
      "loss": 0.7079280853271485,
      "step": 4360
    },
    {
      "epoch": 0.1688888888888889,
      "grad_norm": 2.1839682885818497,
      "learning_rate": 2e-05,
      "loss": 0.7075569152832031,
      "step": 4370
    },
    {
      "epoch": 0.16927536231884058,
      "grad_norm": 1.6034838183681954,
      "learning_rate": 2e-05,
      "loss": 0.7059199333190918,
      "step": 4380
    },
    {
      "epoch": 0.16966183574879226,
      "grad_norm": 1.290033562922369,
      "learning_rate": 2e-05,
      "loss": 0.7122645378112793,
      "step": 4390
    },
    {
      "epoch": 0.17004830917874397,
      "grad_norm": 1.1139843286033466,
      "learning_rate": 2e-05,
      "loss": 0.7275619506835938,
      "step": 4400
    },
    {
      "epoch": 0.17043478260869566,
      "grad_norm": 1.1874723438321881,
      "learning_rate": 2e-05,
      "loss": 0.7264988899230957,
      "step": 4410
    },
    {
      "epoch": 0.17082125603864734,
      "grad_norm": 1.9677653132666741,
      "learning_rate": 2e-05,
      "loss": 0.7497766494750977,
      "step": 4420
    },
    {
      "epoch": 0.17120772946859902,
      "grad_norm": 1.431470883981004,
      "learning_rate": 2e-05,
      "loss": 0.7160844326019287,
      "step": 4430
    },
    {
      "epoch": 0.17159420289855074,
      "grad_norm": 1.7335117472099906,
      "learning_rate": 2e-05,
      "loss": 0.7455331325531006,
      "step": 4440
    },
    {
      "epoch": 0.17198067632850242,
      "grad_norm": 2.0719741530181466,
      "learning_rate": 2e-05,
      "loss": 0.7123561382293702,
      "step": 4450
    },
    {
      "epoch": 0.1723671497584541,
      "grad_norm": 6.604289347169797,
      "learning_rate": 2e-05,
      "loss": 0.7266652107238769,
      "step": 4460
    },
    {
      "epoch": 0.17275362318840579,
      "grad_norm": 3.0995561279455703,
      "learning_rate": 2e-05,
      "loss": 0.7668704509735107,
      "step": 4470
    },
    {
      "epoch": 0.1731400966183575,
      "grad_norm": 1.1410025101367869,
      "learning_rate": 2e-05,
      "loss": 0.7429126262664795,
      "step": 4480
    },
    {
      "epoch": 0.17352657004830918,
      "grad_norm": 1.1987029482898006,
      "learning_rate": 2e-05,
      "loss": 0.718827199935913,
      "step": 4490
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 1.3344093474662273,
      "learning_rate": 2e-05,
      "loss": 0.7187135219573975,
      "step": 4500
    },
    {
      "epoch": 0.17429951690821255,
      "grad_norm": 1.201136766802463,
      "learning_rate": 2e-05,
      "loss": 0.7574004650115966,
      "step": 4510
    },
    {
      "epoch": 0.17468599033816426,
      "grad_norm": 1.7920218239874124,
      "learning_rate": 2e-05,
      "loss": 0.7154958248138428,
      "step": 4520
    },
    {
      "epoch": 0.17507246376811594,
      "grad_norm": 1.73269416875713,
      "learning_rate": 2e-05,
      "loss": 0.7384397029876709,
      "step": 4530
    },
    {
      "epoch": 0.17545893719806763,
      "grad_norm": 1.3075742301957145,
      "learning_rate": 2e-05,
      "loss": 0.7336397171020508,
      "step": 4540
    },
    {
      "epoch": 0.1758454106280193,
      "grad_norm": 2.0182688206782977,
      "learning_rate": 2e-05,
      "loss": 0.735529088973999,
      "step": 4550
    },
    {
      "epoch": 0.17623188405797102,
      "grad_norm": 0.9650475634951186,
      "learning_rate": 2e-05,
      "loss": 0.718187665939331,
      "step": 4560
    },
    {
      "epoch": 0.1766183574879227,
      "grad_norm": 1.0748790541423243,
      "learning_rate": 2e-05,
      "loss": 0.6952043056488038,
      "step": 4570
    },
    {
      "epoch": 0.1770048309178744,
      "grad_norm": 1.1714690255662652,
      "learning_rate": 2e-05,
      "loss": 0.747919750213623,
      "step": 4580
    },
    {
      "epoch": 0.17739130434782607,
      "grad_norm": 1.128321791436026,
      "learning_rate": 2e-05,
      "loss": 0.6863909721374511,
      "step": 4590
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 1.1424962804301904,
      "learning_rate": 2e-05,
      "loss": 0.7397925376892089,
      "step": 4600
    },
    {
      "epoch": 0.17816425120772947,
      "grad_norm": 1.5327096020421922,
      "learning_rate": 2e-05,
      "loss": 0.7165988445281982,
      "step": 4610
    },
    {
      "epoch": 0.17855072463768115,
      "grad_norm": 1.119197988652412,
      "learning_rate": 2e-05,
      "loss": 0.7382492065429688,
      "step": 4620
    },
    {
      "epoch": 0.17893719806763286,
      "grad_norm": 1.0581285556064415,
      "learning_rate": 2e-05,
      "loss": 0.7172673225402832,
      "step": 4630
    },
    {
      "epoch": 0.17932367149758455,
      "grad_norm": 1.0048615777698786,
      "learning_rate": 2e-05,
      "loss": 0.6933206081390381,
      "step": 4640
    },
    {
      "epoch": 0.17971014492753623,
      "grad_norm": 1.0258247876962883,
      "learning_rate": 2e-05,
      "loss": 0.70902419090271,
      "step": 4650
    },
    {
      "epoch": 0.18009661835748791,
      "grad_norm": 2.2361732863979062,
      "learning_rate": 2e-05,
      "loss": 0.7442170143127441,
      "step": 4660
    },
    {
      "epoch": 0.18048309178743963,
      "grad_norm": 0.9749262103607496,
      "learning_rate": 2e-05,
      "loss": 0.707926893234253,
      "step": 4670
    },
    {
      "epoch": 0.1808695652173913,
      "grad_norm": 1.0520602879050982,
      "learning_rate": 2e-05,
      "loss": 0.69857177734375,
      "step": 4680
    },
    {
      "epoch": 0.181256038647343,
      "grad_norm": 2.826079413320698,
      "learning_rate": 2e-05,
      "loss": 0.7357229232788086,
      "step": 4690
    },
    {
      "epoch": 0.18164251207729468,
      "grad_norm": 1.1920760544399023,
      "learning_rate": 2e-05,
      "loss": 0.7186788558959961,
      "step": 4700
    },
    {
      "epoch": 0.1820289855072464,
      "grad_norm": 0.9277355409615122,
      "learning_rate": 2e-05,
      "loss": 0.6909773826599122,
      "step": 4710
    },
    {
      "epoch": 0.18241545893719807,
      "grad_norm": 9.747934033167812,
      "learning_rate": 2e-05,
      "loss": 0.7203973293304443,
      "step": 4720
    },
    {
      "epoch": 0.18280193236714976,
      "grad_norm": 0.987698346422706,
      "learning_rate": 2e-05,
      "loss": 0.7233858108520508,
      "step": 4730
    },
    {
      "epoch": 0.18318840579710144,
      "grad_norm": 0.8641741759054851,
      "learning_rate": 2e-05,
      "loss": 0.7096927642822266,
      "step": 4740
    },
    {
      "epoch": 0.18357487922705315,
      "grad_norm": 1.0229213121410448,
      "learning_rate": 2e-05,
      "loss": 0.7027717590332031,
      "step": 4750
    },
    {
      "epoch": 0.18396135265700483,
      "grad_norm": 1.2770946633174016,
      "learning_rate": 2e-05,
      "loss": 0.73540358543396,
      "step": 4760
    },
    {
      "epoch": 0.18434782608695652,
      "grad_norm": 0.8892390813488997,
      "learning_rate": 2e-05,
      "loss": 0.7259696960449219,
      "step": 4770
    },
    {
      "epoch": 0.1847342995169082,
      "grad_norm": 1.018339775192088,
      "learning_rate": 2e-05,
      "loss": 0.6697995662689209,
      "step": 4780
    },
    {
      "epoch": 0.1851207729468599,
      "grad_norm": 1.2823428575488793,
      "learning_rate": 2e-05,
      "loss": 0.7263753890991211,
      "step": 4790
    },
    {
      "epoch": 0.1855072463768116,
      "grad_norm": 0.7684280600995894,
      "learning_rate": 2e-05,
      "loss": 0.7320945739746094,
      "step": 4800
    },
    {
      "epoch": 0.18589371980676328,
      "grad_norm": 1.021315435265701,
      "learning_rate": 2e-05,
      "loss": 0.7274471282958984,
      "step": 4810
    },
    {
      "epoch": 0.18628019323671496,
      "grad_norm": 0.9622820098348405,
      "learning_rate": 2e-05,
      "loss": 0.7500141143798829,
      "step": 4820
    },
    {
      "epoch": 0.18666666666666668,
      "grad_norm": 1.4050355259854743,
      "learning_rate": 2e-05,
      "loss": 0.6895289421081543,
      "step": 4830
    },
    {
      "epoch": 0.18705314009661836,
      "grad_norm": 1.7625498580922958,
      "learning_rate": 2e-05,
      "loss": 0.7406788825988769,
      "step": 4840
    },
    {
      "epoch": 0.18743961352657004,
      "grad_norm": 1.067712703050187,
      "learning_rate": 2e-05,
      "loss": 0.7196520805358887,
      "step": 4850
    },
    {
      "epoch": 0.18782608695652173,
      "grad_norm": 1.302810427591836,
      "learning_rate": 2e-05,
      "loss": 0.7243830680847168,
      "step": 4860
    },
    {
      "epoch": 0.18821256038647344,
      "grad_norm": 1.4605248330061051,
      "learning_rate": 2e-05,
      "loss": 0.6998039245605469,
      "step": 4870
    },
    {
      "epoch": 0.18859903381642512,
      "grad_norm": 0.8825316291754807,
      "learning_rate": 2e-05,
      "loss": 0.7116286754608154,
      "step": 4880
    },
    {
      "epoch": 0.1889855072463768,
      "grad_norm": 1.1422059123621557,
      "learning_rate": 2e-05,
      "loss": 0.7193767070770264,
      "step": 4890
    },
    {
      "epoch": 0.18937198067632852,
      "grad_norm": 1.0410358854068464,
      "learning_rate": 2e-05,
      "loss": 0.720572566986084,
      "step": 4900
    },
    {
      "epoch": 0.1897584541062802,
      "grad_norm": 1.0903340311245089,
      "learning_rate": 2e-05,
      "loss": 0.6995194911956787,
      "step": 4910
    },
    {
      "epoch": 0.19014492753623188,
      "grad_norm": 1.0202576000113759,
      "learning_rate": 2e-05,
      "loss": 0.6987897396087647,
      "step": 4920
    },
    {
      "epoch": 0.19053140096618357,
      "grad_norm": 1.091349034679939,
      "learning_rate": 2e-05,
      "loss": 0.732855224609375,
      "step": 4930
    },
    {
      "epoch": 0.19091787439613528,
      "grad_norm": 1.2341808282613096,
      "learning_rate": 2e-05,
      "loss": 0.7200730800628662,
      "step": 4940
    },
    {
      "epoch": 0.19130434782608696,
      "grad_norm": 0.9224582820642155,
      "learning_rate": 2e-05,
      "loss": 0.7242010116577149,
      "step": 4950
    },
    {
      "epoch": 0.19169082125603865,
      "grad_norm": 1.1735102587601551,
      "learning_rate": 2e-05,
      "loss": 0.7138439178466797,
      "step": 4960
    },
    {
      "epoch": 0.19207729468599033,
      "grad_norm": 1.710727312230191,
      "learning_rate": 2e-05,
      "loss": 0.7399075984954834,
      "step": 4970
    },
    {
      "epoch": 0.19246376811594204,
      "grad_norm": 1.2230589650110089,
      "learning_rate": 2e-05,
      "loss": 0.7408647537231445,
      "step": 4980
    },
    {
      "epoch": 0.19285024154589372,
      "grad_norm": 1.2983537095986883,
      "learning_rate": 2e-05,
      "loss": 0.6835604667663574,
      "step": 4990
    },
    {
      "epoch": 0.1932367149758454,
      "grad_norm": 1.4788440790795512,
      "learning_rate": 2e-05,
      "loss": 0.715909767150879,
      "step": 5000
    },
    {
      "epoch": 0.1936231884057971,
      "grad_norm": 1.4264990896680747,
      "learning_rate": 2e-05,
      "loss": 0.6949624061584473,
      "step": 5010
    },
    {
      "epoch": 0.1940096618357488,
      "grad_norm": 1.1527109499106305,
      "learning_rate": 2e-05,
      "loss": 0.7330007553100586,
      "step": 5020
    },
    {
      "epoch": 0.1943961352657005,
      "grad_norm": 0.9440416804730175,
      "learning_rate": 2e-05,
      "loss": 0.7006573677062988,
      "step": 5030
    },
    {
      "epoch": 0.19478260869565217,
      "grad_norm": 1.0713149502943946,
      "learning_rate": 2e-05,
      "loss": 0.7057698249816895,
      "step": 5040
    },
    {
      "epoch": 0.19516908212560385,
      "grad_norm": 0.963217782760897,
      "learning_rate": 2e-05,
      "loss": 0.71975736618042,
      "step": 5050
    },
    {
      "epoch": 0.19555555555555557,
      "grad_norm": 0.9965379672773939,
      "learning_rate": 2e-05,
      "loss": 0.7374117851257325,
      "step": 5060
    },
    {
      "epoch": 0.19594202898550725,
      "grad_norm": 2.0427592648086024,
      "learning_rate": 2e-05,
      "loss": 0.6966311454772949,
      "step": 5070
    },
    {
      "epoch": 0.19632850241545893,
      "grad_norm": 1.254913989327177,
      "learning_rate": 2e-05,
      "loss": 0.7121417045593261,
      "step": 5080
    },
    {
      "epoch": 0.19671497584541062,
      "grad_norm": 1.017324520978876,
      "learning_rate": 2e-05,
      "loss": 0.7623320579528808,
      "step": 5090
    },
    {
      "epoch": 0.19710144927536233,
      "grad_norm": 1.3446622621912172,
      "learning_rate": 2e-05,
      "loss": 0.7287818908691406,
      "step": 5100
    },
    {
      "epoch": 0.197487922705314,
      "grad_norm": 5.283141359326179,
      "learning_rate": 2e-05,
      "loss": 0.7268217086791993,
      "step": 5110
    },
    {
      "epoch": 0.1978743961352657,
      "grad_norm": 2.013323137216242,
      "learning_rate": 2e-05,
      "loss": 0.713960838317871,
      "step": 5120
    },
    {
      "epoch": 0.19826086956521738,
      "grad_norm": 0.8486362400788193,
      "learning_rate": 2e-05,
      "loss": 0.7176273345947266,
      "step": 5130
    },
    {
      "epoch": 0.1986473429951691,
      "grad_norm": 1.253405007894602,
      "learning_rate": 2e-05,
      "loss": 0.7150094032287597,
      "step": 5140
    },
    {
      "epoch": 0.19903381642512077,
      "grad_norm": 3.091410015827302,
      "learning_rate": 2e-05,
      "loss": 0.7038010597229004,
      "step": 5150
    },
    {
      "epoch": 0.19942028985507246,
      "grad_norm": 1.1567106778924343,
      "learning_rate": 2e-05,
      "loss": 0.7240169048309326,
      "step": 5160
    },
    {
      "epoch": 0.19980676328502414,
      "grad_norm": 1.3568132227619873,
      "learning_rate": 2e-05,
      "loss": 0.702213191986084,
      "step": 5170
    },
    {
      "epoch": 0.20019323671497585,
      "grad_norm": 1.5758093601702232,
      "learning_rate": 2e-05,
      "loss": 0.7454169273376465,
      "step": 5180
    },
    {
      "epoch": 0.20057971014492754,
      "grad_norm": 0.9917043343538454,
      "learning_rate": 2e-05,
      "loss": 0.6948194026947021,
      "step": 5190
    },
    {
      "epoch": 0.20096618357487922,
      "grad_norm": 0.9382588892689538,
      "learning_rate": 2e-05,
      "loss": 0.7076561927795411,
      "step": 5200
    },
    {
      "epoch": 0.20135265700483093,
      "grad_norm": 1.3378387928205202,
      "learning_rate": 2e-05,
      "loss": 0.7205132961273193,
      "step": 5210
    },
    {
      "epoch": 0.20173913043478262,
      "grad_norm": 1.4013897572931409,
      "learning_rate": 2e-05,
      "loss": 0.704619312286377,
      "step": 5220
    },
    {
      "epoch": 0.2021256038647343,
      "grad_norm": 1.0574529723704258,
      "learning_rate": 2e-05,
      "loss": 0.6813540935516358,
      "step": 5230
    },
    {
      "epoch": 0.20251207729468598,
      "grad_norm": 1.2386285783865574,
      "learning_rate": 2e-05,
      "loss": 0.6983018398284913,
      "step": 5240
    },
    {
      "epoch": 0.2028985507246377,
      "grad_norm": 0.9847217376243098,
      "learning_rate": 2e-05,
      "loss": 0.7246550559997559,
      "step": 5250
    },
    {
      "epoch": 0.20328502415458938,
      "grad_norm": 0.9452774456943679,
      "learning_rate": 2e-05,
      "loss": 0.7033842086791993,
      "step": 5260
    },
    {
      "epoch": 0.20367149758454106,
      "grad_norm": 1.7926492014311681,
      "learning_rate": 2e-05,
      "loss": 0.7108603477478027,
      "step": 5270
    },
    {
      "epoch": 0.20405797101449274,
      "grad_norm": 0.7998274316890636,
      "learning_rate": 2e-05,
      "loss": 0.680066442489624,
      "step": 5280
    },
    {
      "epoch": 0.20444444444444446,
      "grad_norm": 1.5112646734973114,
      "learning_rate": 2e-05,
      "loss": 0.7166651725769043,
      "step": 5290
    },
    {
      "epoch": 0.20483091787439614,
      "grad_norm": 1.0483389229557751,
      "learning_rate": 2e-05,
      "loss": 0.7424324989318848,
      "step": 5300
    },
    {
      "epoch": 0.20521739130434782,
      "grad_norm": 1.0232927161449363,
      "learning_rate": 2e-05,
      "loss": 0.6954901695251465,
      "step": 5310
    },
    {
      "epoch": 0.2056038647342995,
      "grad_norm": 1.3660267905893448,
      "learning_rate": 2e-05,
      "loss": 0.6616646766662597,
      "step": 5320
    },
    {
      "epoch": 0.20599033816425122,
      "grad_norm": 1.286450214994112,
      "learning_rate": 2e-05,
      "loss": 0.6984244346618652,
      "step": 5330
    },
    {
      "epoch": 0.2063768115942029,
      "grad_norm": 1.0454357280325042,
      "learning_rate": 2e-05,
      "loss": 0.740637731552124,
      "step": 5340
    },
    {
      "epoch": 0.20676328502415459,
      "grad_norm": 1.1242828999960899,
      "learning_rate": 1.999194520388485e-05,
      "loss": 0.6997871398925781,
      "step": 5350
    },
    {
      "epoch": 0.20714975845410627,
      "grad_norm": 1.0513083874669853,
      "learning_rate": 1.9959247303157763e-05,
      "loss": 0.7108020782470703,
      "step": 5360
    },
    {
      "epoch": 0.20753623188405798,
      "grad_norm": 1.2883390088535822,
      "learning_rate": 1.990149424459817e-05,
      "loss": 0.7122337341308593,
      "step": 5370
    },
    {
      "epoch": 0.20792270531400966,
      "grad_norm": 1.053137537640253,
      "learning_rate": 1.9818847538648688e-05,
      "loss": 0.7302716732025146,
      "step": 5380
    },
    {
      "epoch": 0.20830917874396135,
      "grad_norm": 1.9562766048016125,
      "learning_rate": 1.971153831256721e-05,
      "loss": 0.6772193908691406,
      "step": 5390
    },
    {
      "epoch": 0.20869565217391303,
      "grad_norm": 0.8280239799941486,
      "learning_rate": 1.9579866664063435e-05,
      "loss": 0.6907012462615967,
      "step": 5400
    },
    {
      "epoch": 0.20908212560386474,
      "grad_norm": 1.6419872315999635,
      "learning_rate": 1.9424200822054782e-05,
      "loss": 0.692955493927002,
      "step": 5410
    },
    {
      "epoch": 0.20946859903381643,
      "grad_norm": 1.1057659058145781,
      "learning_rate": 1.9244976116888626e-05,
      "loss": 0.6924299240112305,
      "step": 5420
    },
    {
      "epoch": 0.2098550724637681,
      "grad_norm": 1.1380365420683456,
      "learning_rate": 1.904269376291071e-05,
      "loss": 0.688160514831543,
      "step": 5430
    },
    {
      "epoch": 0.2102415458937198,
      "grad_norm": 1.0687909480880293,
      "learning_rate": 1.88179194567844e-05,
      "loss": 0.6906280517578125,
      "step": 5440
    },
    {
      "epoch": 0.2106280193236715,
      "grad_norm": 0.9654618467578738,
      "learning_rate": 1.8571281795480632e-05,
      "loss": 0.6934199810028077,
      "step": 5450
    },
    {
      "epoch": 0.2110144927536232,
      "grad_norm": 0.9012618792174377,
      "learning_rate": 1.8303470518362808e-05,
      "loss": 0.7280242919921875,
      "step": 5460
    },
    {
      "epoch": 0.21140096618357487,
      "grad_norm": 0.882599237027553,
      "learning_rate": 1.801523457828271e-05,
      "loss": 0.725037670135498,
      "step": 5470
    },
    {
      "epoch": 0.21178743961352656,
      "grad_norm": 1.550151756423383,
      "learning_rate": 1.7707380047081796e-05,
      "loss": 0.7048633575439454,
      "step": 5480
    },
    {
      "epoch": 0.21217391304347827,
      "grad_norm": 0.9611755404154515,
      "learning_rate": 1.7380767861355252e-05,
      "loss": 0.685286283493042,
      "step": 5490
    },
    {
      "epoch": 0.21256038647342995,
      "grad_norm": 0.9215723505968825,
      "learning_rate": 1.703631141478299e-05,
      "loss": 0.6968507289886474,
      "step": 5500
    },
    {
      "epoch": 0.21294685990338164,
      "grad_norm": 0.9365594631180093,
      "learning_rate": 1.6674974003760705e-05,
      "loss": 0.6745092391967773,
      "step": 5510
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 5.1254852605440915,
      "learning_rate": 1.6297766133474556e-05,
      "loss": 0.7054040431976318,
      "step": 5520
    },
    {
      "epoch": 0.21371980676328503,
      "grad_norm": 1.1791759130118333,
      "learning_rate": 1.5905742691953186e-05,
      "loss": 0.6666687965393067,
      "step": 5530
    },
    {
      "epoch": 0.21410628019323671,
      "grad_norm": 0.798661331928038,
      "learning_rate": 1.55e-05,
      "loss": 0.7174147605895996,
      "step": 5540
    },
    {
      "epoch": 0.2144927536231884,
      "grad_norm": 1.1820540387037355,
      "learning_rate": 1.5081672745255862e-05,
      "loss": 0.7021834850311279,
      "step": 5550
    },
    {
      "epoch": 0.2148792270531401,
      "grad_norm": 0.9296311495644336,
      "learning_rate": 1.4651930808966206e-05,
      "loss": 0.706732702255249,
      "step": 5560
    },
    {
      "epoch": 0.2152657004830918,
      "grad_norm": 0.8802908352641984,
      "learning_rate": 1.4211975994326848e-05,
      "loss": 0.7034182548522949,
      "step": 5570
    },
    {
      "epoch": 0.21565217391304348,
      "grad_norm": 2.8225849122866387,
      "learning_rate": 1.376303866555779e-05,
      "loss": 0.683725643157959,
      "step": 5580
    },
    {
      "epoch": 0.21603864734299516,
      "grad_norm": 1.1163184572031415,
      "learning_rate": 1.3306374307104164e-05,
      "loss": 0.699821949005127,
      "step": 5590
    },
    {
      "epoch": 0.21642512077294687,
      "grad_norm": 1.1900156459787095,
      "learning_rate": 1.2843260012586718e-05,
      "loss": 0.7381949424743652,
      "step": 5600
    },
    {
      "epoch": 0.21681159420289856,
      "grad_norm": 3.812840871651647,
      "learning_rate": 1.2374990913320726e-05,
      "loss": 0.7069476127624512,
      "step": 5610
    },
    {
      "epoch": 0.21719806763285024,
      "grad_norm": 9.80490003053943,
      "learning_rate": 1.1902876556391247e-05,
      "loss": 0.6832869529724122,
      "step": 5620
    },
    {
      "epoch": 0.21758454106280192,
      "grad_norm": 2.1165471642443,
      "learning_rate": 1.1428237242413685e-05,
      "loss": 0.672675609588623,
      "step": 5630
    },
    {
      "epoch": 0.21797101449275363,
      "grad_norm": 0.9771105467489954,
      "learning_rate": 1.095240033322132e-05,
      "loss": 0.6635546684265137,
      "step": 5640
    },
    {
      "epoch": 0.21835748792270532,
      "grad_norm": 0.9151059453390374,
      "learning_rate": 1.047669653980572e-05,
      "loss": 0.6838153839111328,
      "step": 5650
    },
    {
      "epoch": 0.218743961352657,
      "grad_norm": 1.7926948985964104,
      "learning_rate": 1.0002456200890902e-05,
      "loss": 0.7035691261291503,
      "step": 5660
    },
    {
      "epoch": 0.21913043478260869,
      "grad_norm": 1.8301790651654564,
      "learning_rate": 9.531005562548668e-06,
      "loss": 0.6558338165283203,
      "step": 5670
    },
    {
      "epoch": 0.2195169082125604,
      "grad_norm": 1.1364005199233806,
      "learning_rate": 9.06366306925925e-06,
      "loss": 0.6784430503845215,
      "step": 5680
    },
    {
      "epoch": 0.21990338164251208,
      "grad_norm": 7.497910377622659,
      "learning_rate": 8.601735676789686e-06,
      "loss": 0.705353832244873,
      "step": 5690
    },
    {
      "epoch": 0.22028985507246376,
      "grad_norm": 0.9840272476969066,
      "learning_rate": 8.14651519720118e-06,
      "loss": 0.6876489639282226,
      "step": 5700
    },
    {
      "epoch": 0.22067632850241545,
      "grad_norm": 0.8734133817279891,
      "learning_rate": 7.699274686206866e-06,
      "loss": 0.6783504009246826,
      "step": 5710
    },
    {
      "epoch": 0.22106280193236716,
      "grad_norm": 0.969273969855585,
      "learning_rate": 7.261264882983024e-06,
      "loss": 0.6895233154296875,
      "step": 5720
    },
    {
      "epoch": 0.22144927536231884,
      "grad_norm": 1.0407036500980014,
      "learning_rate": 6.8337107123900386e-06,
      "loss": 0.7053444862365723,
      "step": 5730
    },
    {
      "epoch": 0.22183574879227053,
      "grad_norm": 0.8878500602329745,
      "learning_rate": 6.417807859384881e-06,
      "loss": 0.6848884582519531,
      "step": 5740
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 1.408137437049562,
      "learning_rate": 6.014719425205009e-06,
      "loss": 0.6462078094482422,
      "step": 5750
    },
    {
      "epoch": 0.22260869565217392,
      "grad_norm": 1.7180308876565746,
      "learning_rate": 5.6255726746749275e-06,
      "loss": 0.7042626857757568,
      "step": 5760
    },
    {
      "epoch": 0.2229951690821256,
      "grad_norm": 1.945756395753943,
      "learning_rate": 5.251455883731728e-06,
      "loss": 0.654316520690918,
      "step": 5770
    },
    {
      "epoch": 0.2233816425120773,
      "grad_norm": 0.9186902303890861,
      "learning_rate": 4.893415295985812e-06,
      "loss": 0.7044451713562012,
      "step": 5780
    },
    {
      "epoch": 0.22376811594202897,
      "grad_norm": 2.200643609565289,
      "learning_rate": 4.552452196827885e-06,
      "loss": 0.67486572265625,
      "step": 5790
    },
    {
      "epoch": 0.22415458937198068,
      "grad_norm": 1.695355334817592,
      "learning_rate": 4.229520113264785e-06,
      "loss": 0.6731067657470703,
      "step": 5800
    },
    {
      "epoch": 0.22454106280193237,
      "grad_norm": 1.0131027322622268,
      "learning_rate": 3.925522147314915e-06,
      "loss": 0.7048968315124512,
      "step": 5810
    },
    {
      "epoch": 0.22492753623188405,
      "grad_norm": 1.0015409899364882,
      "learning_rate": 3.641308450420675e-06,
      "loss": 0.6881979942321778,
      "step": 5820
    },
    {
      "epoch": 0.22531400966183576,
      "grad_norm": 1.1868102576347859,
      "learning_rate": 3.37767384594087e-06,
      "loss": 0.6695881843566894,
      "step": 5830
    },
    {
      "epoch": 0.22570048309178745,
      "grad_norm": 1.5593203885658466,
      "learning_rate": 3.1353556063719357e-06,
      "loss": 0.6629329681396484,
      "step": 5840
    },
    {
      "epoch": 0.22608695652173913,
      "grad_norm": 0.8041837313973521,
      "learning_rate": 2.9150313915141474e-06,
      "loss": 0.6779134273529053,
      "step": 5850
    },
    {
      "epoch": 0.2264734299516908,
      "grad_norm": 1.0169914300968392,
      "learning_rate": 2.717317353348891e-06,
      "loss": 0.6744910717010498,
      "step": 5860
    },
    {
      "epoch": 0.22685990338164252,
      "grad_norm": 0.831993287032017,
      "learning_rate": 2.5427664129268253e-06,
      "loss": 0.6612636566162109,
      "step": 5870
    },
    {
      "epoch": 0.2272463768115942,
      "grad_norm": 0.9953168064114183,
      "learning_rate": 2.391866714085717e-06,
      "loss": 0.6742310523986816,
      "step": 5880
    },
    {
      "epoch": 0.2276328502415459,
      "grad_norm": 1.6258557967570229,
      "learning_rate": 2.2650402583222753e-06,
      "loss": 0.654944372177124,
      "step": 5890
    },
    {
      "epoch": 0.22801932367149758,
      "grad_norm": 1.2987688236432129,
      "learning_rate": 2.1626417246356398e-06,
      "loss": 0.6731691360473633,
      "step": 5900
    },
    {
      "epoch": 0.2284057971014493,
      "grad_norm": 1.1786490763216573,
      "learning_rate": 2.084957477642894e-06,
      "loss": 0.6719546794891358,
      "step": 5910
    },
    {
      "epoch": 0.22879227053140097,
      "grad_norm": 1.3952980716896994,
      "learning_rate": 2.0322047667405147e-06,
      "loss": 0.6761349678039551,
      "step": 5920
    },
    {
      "epoch": 0.22917874396135265,
      "grad_norm": 1.1438704239365953,
      "learning_rate": 2.0045311185513342e-06,
      "loss": 0.657864761352539,
      "step": 5930
    }
  ],
  "logging_steps": 10,
  "max_steps": 5935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2035,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2810342056919040.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}