{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.07864734299516908,
  "eval_steps": 500,
  "global_step": 2035,
  "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
    }
  ],
  "logging_steps": 10,
  "max_steps": 5935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2035,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 963613493821440.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}