{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.32018774277082435,
  "eval_steps": 500,
  "global_step": 5935,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005394907207596029,
      "grad_norm": 369.69301600900053,
      "learning_rate": 2.910112359550562e-06,
      "loss": 1.8730627059936524,
      "step": 10
    },
    {
      "epoch": 0.0010789814415192059,
      "grad_norm": 31.93478853689241,
      "learning_rate": 3.921348314606742e-06,
      "loss": 1.262525463104248,
      "step": 20
    },
    {
      "epoch": 0.0016184721622788088,
      "grad_norm": 141.7556466471986,
      "learning_rate": 4.932584269662922e-06,
      "loss": 1.0669395446777343,
      "step": 30
    },
    {
      "epoch": 0.0021579628830384117,
      "grad_norm": 127.98270782023548,
      "learning_rate": 5.943820224719102e-06,
      "loss": 1.1076365470886231,
      "step": 40
    },
    {
      "epoch": 0.002697453603798015,
      "grad_norm": 101.43748401214339,
      "learning_rate": 6.955056179775282e-06,
      "loss": 1.0109930038452148,
      "step": 50
    },
    {
      "epoch": 0.0032369443245576176,
      "grad_norm": 899.0086783565027,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.9515115737915039,
      "step": 60
    },
    {
      "epoch": 0.0037764350453172208,
      "grad_norm": 15.557933246692217,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.9290966987609863,
      "step": 70
    },
    {
      "epoch": 0.0043159257660768235,
      "grad_norm": 256.064896120912,
      "learning_rate": 9.988764044943822e-06,
      "loss": 1.1323322296142577,
      "step": 80
    },
    {
      "epoch": 0.004855416486836427,
      "grad_norm": 20.848525026966247,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 1.088106918334961,
      "step": 90
    },
    {
      "epoch": 0.00539490720759603,
      "grad_norm": 25.60593783327153,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.9955007553100585,
      "step": 100
    },
    {
      "epoch": 0.005934397928355632,
      "grad_norm": 11.664176937579581,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.942236328125,
      "step": 110
    },
    {
      "epoch": 0.006473888649115235,
      "grad_norm": 7.9248397322900646,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.9238415718078613,
      "step": 120
    },
    {
      "epoch": 0.007013379369874838,
      "grad_norm": 26.682291952362966,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.9361964225769043,
      "step": 130
    },
    {
      "epoch": 0.0075528700906344415,
      "grad_norm": 3.6691592434398466,
      "learning_rate": 1.60561797752809e-05,
      "loss": 0.927226448059082,
      "step": 140
    },
    {
      "epoch": 0.008092360811394044,
      "grad_norm": 15.191132047351614,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.9375665664672852,
      "step": 150
    },
    {
      "epoch": 0.008631851532153647,
      "grad_norm": 7.4209329049996455,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.9221304893493653,
      "step": 160
    },
    {
      "epoch": 0.00917134225291325,
      "grad_norm": 5.050234356436225,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.9327095031738282,
      "step": 170
    },
    {
      "epoch": 0.009710832973672853,
      "grad_norm": 2.532425838265676,
      "learning_rate": 2e-05,
      "loss": 0.90814208984375,
      "step": 180
    },
    {
      "epoch": 0.010250323694432456,
      "grad_norm": 2.8569985447993664,
      "learning_rate": 2e-05,
      "loss": 0.930181884765625,
      "step": 190
    },
    {
      "epoch": 0.01078981441519206,
      "grad_norm": 3.7217733328209746,
      "learning_rate": 2e-05,
      "loss": 0.9137080192565918,
      "step": 200
    },
    {
      "epoch": 0.011329305135951661,
      "grad_norm": 1.6856605125936686,
      "learning_rate": 2e-05,
      "loss": 0.9173044204711914,
      "step": 210
    },
    {
      "epoch": 0.011868795856711264,
      "grad_norm": 4.042279257566151,
      "learning_rate": 2e-05,
      "loss": 0.9044102668762207,
      "step": 220
    },
    {
      "epoch": 0.012408286577470867,
      "grad_norm": 7.693180928481417,
      "learning_rate": 2e-05,
      "loss": 0.9013534545898437,
      "step": 230
    },
    {
      "epoch": 0.01294777729823047,
      "grad_norm": 3.5295771823615647,
      "learning_rate": 2e-05,
      "loss": 0.9148492813110352,
      "step": 240
    },
    {
      "epoch": 0.013487268018990074,
      "grad_norm": 7.839029508109661,
      "learning_rate": 2e-05,
      "loss": 0.9837350845336914,
      "step": 250
    },
    {
      "epoch": 0.014026758739749677,
      "grad_norm": 3.670914514568096,
      "learning_rate": 2e-05,
      "loss": 0.9146605491638183,
      "step": 260
    },
    {
      "epoch": 0.01456624946050928,
      "grad_norm": 3.5577694379027,
      "learning_rate": 2e-05,
      "loss": 0.9401106834411621,
      "step": 270
    },
    {
      "epoch": 0.015105740181268883,
      "grad_norm": 2.453202809896575,
      "learning_rate": 2e-05,
      "loss": 0.9477849960327148,
      "step": 280
    },
    {
      "epoch": 0.015645230902028486,
      "grad_norm": 1.9729075743285769,
      "learning_rate": 2e-05,
      "loss": 0.9066492080688476,
      "step": 290
    },
    {
      "epoch": 0.016184721622788088,
      "grad_norm": 2.8099192390223795,
      "learning_rate": 2e-05,
      "loss": 0.9063838958740235,
      "step": 300
    },
    {
      "epoch": 0.016724212343547692,
      "grad_norm": 4.236623113420927,
      "learning_rate": 2e-05,
      "loss": 0.8637477874755859,
      "step": 310
    },
    {
      "epoch": 0.017263703064307294,
      "grad_norm": 4.167628129961564,
      "learning_rate": 2e-05,
      "loss": 0.9061297416687012,
      "step": 320
    },
    {
      "epoch": 0.017803193785066895,
      "grad_norm": 6.935834245033307,
      "learning_rate": 2e-05,
      "loss": 0.8871530532836914,
      "step": 330
    },
    {
      "epoch": 0.0183426845058265,
      "grad_norm": 3.2423751981314384,
      "learning_rate": 2e-05,
      "loss": 0.9223550796508789,
      "step": 340
    },
    {
      "epoch": 0.0188821752265861,
      "grad_norm": 30.606793216440163,
      "learning_rate": 2e-05,
      "loss": 3.54920654296875,
      "step": 350
    },
    {
      "epoch": 0.019421665947345707,
      "grad_norm": 4.550939608135107,
      "learning_rate": 2e-05,
      "loss": 0.9992364883422852,
      "step": 360
    },
    {
      "epoch": 0.019961156668105308,
      "grad_norm": 10.355788989674835,
      "learning_rate": 2e-05,
      "loss": 0.9463441848754883,
      "step": 370
    },
    {
      "epoch": 0.020500647388864913,
      "grad_norm": 9.470092673566741,
      "learning_rate": 2e-05,
      "loss": 0.9165935516357422,
      "step": 380
    },
    {
      "epoch": 0.021040138109624514,
      "grad_norm": 16.04366173130251,
      "learning_rate": 2e-05,
      "loss": 0.9199161529541016,
      "step": 390
    },
    {
      "epoch": 0.02157962883038412,
      "grad_norm": 16.276111063417574,
      "learning_rate": 2e-05,
      "loss": 0.8900386810302734,
      "step": 400
    },
    {
      "epoch": 0.02211911955114372,
      "grad_norm": 14.653812444903842,
      "learning_rate": 2e-05,
      "loss": 0.8979652404785157,
      "step": 410
    },
    {
      "epoch": 0.022658610271903322,
      "grad_norm": 2.6977548589169094,
      "learning_rate": 2e-05,
      "loss": 0.9128728866577148,
      "step": 420
    },
    {
      "epoch": 0.023198100992662927,
      "grad_norm": 23.582887550971233,
      "learning_rate": 2e-05,
      "loss": 0.8891826629638672,
      "step": 430
    },
    {
      "epoch": 0.023737591713422528,
      "grad_norm": 2.9436146384915443,
      "learning_rate": 2e-05,
      "loss": 0.9091741561889648,
      "step": 440
    },
    {
      "epoch": 0.024277082434182133,
      "grad_norm": 3.8178614247451996,
      "learning_rate": 2e-05,
      "loss": 0.9071731567382812,
      "step": 450
    },
    {
      "epoch": 0.024816573154941735,
      "grad_norm": 9.900017833351434,
      "learning_rate": 2e-05,
      "loss": 0.8803417205810546,
      "step": 460
    },
    {
      "epoch": 0.02535606387570134,
      "grad_norm": 11.377800042408293,
      "learning_rate": 2e-05,
      "loss": 0.9077526092529297,
      "step": 470
    },
    {
      "epoch": 0.02589555459646094,
      "grad_norm": 4.425640907224202,
      "learning_rate": 2e-05,
      "loss": 0.8951313018798828,
      "step": 480
    },
    {
      "epoch": 0.026435045317220542,
      "grad_norm": 5.543696039922662,
      "learning_rate": 2e-05,
      "loss": 0.9170653343200683,
      "step": 490
    },
    {
      "epoch": 0.026974536037980147,
      "grad_norm": 6.637675684831988,
      "learning_rate": 2e-05,
      "loss": 0.9328924179077148,
      "step": 500
    },
    {
      "epoch": 0.02751402675873975,
      "grad_norm": 4.541587002846055,
      "learning_rate": 2e-05,
      "loss": 0.9184404373168945,
      "step": 510
    },
    {
      "epoch": 0.028053517479499353,
      "grad_norm": 4.834955555833834,
      "learning_rate": 2e-05,
      "loss": 0.94840087890625,
      "step": 520
    },
    {
      "epoch": 0.028593008200258955,
      "grad_norm": 2.7476196554801118,
      "learning_rate": 2e-05,
      "loss": 0.9112020492553711,
      "step": 530
    },
    {
      "epoch": 0.02913249892101856,
      "grad_norm": 3.729222437612153,
      "learning_rate": 2e-05,
      "loss": 0.8715265274047852,
      "step": 540
    },
    {
      "epoch": 0.02967198964177816,
      "grad_norm": 1.8986444153489899,
      "learning_rate": 2e-05,
      "loss": 0.8817804336547852,
      "step": 550
    },
    {
      "epoch": 0.030211480362537766,
      "grad_norm": 19.652863197507482,
      "learning_rate": 2e-05,
      "loss": 0.9034578323364257,
      "step": 560
    },
    {
      "epoch": 0.030750971083297367,
      "grad_norm": 4.893935787998432,
      "learning_rate": 2e-05,
      "loss": 0.9201476097106933,
      "step": 570
    },
    {
      "epoch": 0.03129046180405697,
      "grad_norm": 2.121315677688142,
      "learning_rate": 2e-05,
      "loss": 0.9087504386901856,
      "step": 580
    },
    {
      "epoch": 0.031829952524816574,
      "grad_norm": 1.6805303205822455,
      "learning_rate": 2e-05,
      "loss": 0.8938014030456543,
      "step": 590
    },
    {
      "epoch": 0.032369443245576175,
      "grad_norm": 2.091016561123043,
      "learning_rate": 2e-05,
      "loss": 0.8701451301574707,
      "step": 600
    },
    {
      "epoch": 0.03290893396633578,
      "grad_norm": 4.4795444497420345,
      "learning_rate": 2e-05,
      "loss": 0.8988707542419434,
      "step": 610
    },
    {
      "epoch": 0.033448424687095385,
      "grad_norm": 1.40755956780777,
      "learning_rate": 2e-05,
      "loss": 0.8736795425415039,
      "step": 620
    },
    {
      "epoch": 0.033987915407854986,
      "grad_norm": 1.9677540233275401,
      "learning_rate": 2e-05,
      "loss": 0.8994430541992188,
      "step": 630
    },
    {
      "epoch": 0.03452740612861459,
      "grad_norm": 2.8537882081306605,
      "learning_rate": 2e-05,
      "loss": 0.8869138717651367,
      "step": 640
    },
    {
      "epoch": 0.03506689684937419,
      "grad_norm": 2.3406412447640133,
      "learning_rate": 2e-05,
      "loss": 0.8736772537231445,
      "step": 650
    },
    {
      "epoch": 0.03560638757013379,
      "grad_norm": 2.7002297840083647,
      "learning_rate": 2e-05,
      "loss": 0.8878070831298828,
      "step": 660
    },
    {
      "epoch": 0.0361458782908934,
      "grad_norm": 5.478051955992026,
      "learning_rate": 2e-05,
      "loss": 0.9007582664489746,
      "step": 670
    },
    {
      "epoch": 0.036685369011653,
      "grad_norm": 5.987775578117277,
      "learning_rate": 2e-05,
      "loss": 0.877619457244873,
      "step": 680
    },
    {
      "epoch": 0.0372248597324126,
      "grad_norm": 4.062183705280081,
      "learning_rate": 2e-05,
      "loss": 0.9615988731384277,
      "step": 690
    },
    {
      "epoch": 0.0377643504531722,
      "grad_norm": 1.7304116921547956,
      "learning_rate": 2e-05,
      "loss": 0.8999065399169922,
      "step": 700
    },
    {
      "epoch": 0.03830384117393181,
      "grad_norm": 10.224091687714415,
      "learning_rate": 2e-05,
      "loss": 0.8775262832641602,
      "step": 710
    },
    {
      "epoch": 0.03884333189469141,
      "grad_norm": 2.6973024084894885,
      "learning_rate": 2e-05,
      "loss": 0.88863525390625,
      "step": 720
    },
    {
      "epoch": 0.039382822615451014,
      "grad_norm": 3.0097373941657186,
      "learning_rate": 2e-05,
      "loss": 0.8921207427978516,
      "step": 730
    },
    {
      "epoch": 0.039922313336210616,
      "grad_norm": 2.4929780841655207,
      "learning_rate": 2e-05,
      "loss": 0.8870817184448242,
      "step": 740
    },
    {
      "epoch": 0.04046180405697022,
      "grad_norm": 2.680211035148592,
      "learning_rate": 2e-05,
      "loss": 0.8718931198120117,
      "step": 750
    },
    {
      "epoch": 0.041001294777729826,
      "grad_norm": 3.314725765838302,
      "learning_rate": 2e-05,
      "loss": 0.8696352005004883,
      "step": 760
    },
    {
      "epoch": 0.04154078549848943,
      "grad_norm": 2.8067410842808824,
      "learning_rate": 2e-05,
      "loss": 0.8808348655700684,
      "step": 770
    },
    {
      "epoch": 0.04208027621924903,
      "grad_norm": 4.568417784852867,
      "learning_rate": 2e-05,
      "loss": 0.8762812614440918,
      "step": 780
    },
    {
      "epoch": 0.04261976694000863,
      "grad_norm": 1.591466243191371,
      "learning_rate": 2e-05,
      "loss": 0.8897010803222656,
      "step": 790
    },
    {
      "epoch": 0.04315925766076824,
      "grad_norm": 1.7378624546635357,
      "learning_rate": 2e-05,
      "loss": 0.8871335983276367,
      "step": 800
    },
    {
      "epoch": 0.04369874838152784,
      "grad_norm": 5.282884941300159,
      "learning_rate": 2e-05,
      "loss": 0.8862749099731445,
      "step": 810
    },
    {
      "epoch": 0.04423823910228744,
      "grad_norm": 1.4282666163813993,
      "learning_rate": 2e-05,
      "loss": 0.9045479774475098,
      "step": 820
    },
    {
      "epoch": 0.04477772982304704,
      "grad_norm": 1.804678546954577,
      "learning_rate": 2e-05,
      "loss": 0.8669574737548829,
      "step": 830
    },
    {
      "epoch": 0.045317220543806644,
      "grad_norm": 2.7003745816690836,
      "learning_rate": 2e-05,
      "loss": 0.8664716720581055,
      "step": 840
    },
    {
      "epoch": 0.04585671126456625,
      "grad_norm": 2.8347718319860626,
      "learning_rate": 2e-05,
      "loss": 0.8740848541259766,
      "step": 850
    },
    {
      "epoch": 0.046396201985325854,
      "grad_norm": 2.3599301199842655,
      "learning_rate": 2e-05,
      "loss": 0.8661260604858398,
      "step": 860
    },
    {
      "epoch": 0.046935692706085455,
      "grad_norm": 2.675130769028597,
      "learning_rate": 2e-05,
      "loss": 0.8940586090087891,
      "step": 870
    },
    {
      "epoch": 0.047475183426845056,
      "grad_norm": 2.13219637159858,
      "learning_rate": 2e-05,
      "loss": 0.8656101226806641,
      "step": 880
    },
    {
      "epoch": 0.04801467414760466,
      "grad_norm": 5.746935578097955,
      "learning_rate": 2e-05,
      "loss": 0.9049406051635742,
      "step": 890
    },
    {
      "epoch": 0.048554164868364266,
      "grad_norm": 3.040442087390683,
      "learning_rate": 2e-05,
      "loss": 0.8673805236816406,
      "step": 900
    },
    {
      "epoch": 0.04909365558912387,
      "grad_norm": 1.8033864790060448,
      "learning_rate": 2e-05,
      "loss": 0.9076686859130859,
      "step": 910
    },
    {
      "epoch": 0.04963314630988347,
      "grad_norm": 3.7270371618478446,
      "learning_rate": 2e-05,
      "loss": 0.8856944084167481,
      "step": 920
    },
    {
      "epoch": 0.05017263703064307,
      "grad_norm": 18.34533608060895,
      "learning_rate": 2e-05,
      "loss": 0.8803435325622558,
      "step": 930
    },
    {
      "epoch": 0.05071212775140268,
      "grad_norm": 7.031326732375769,
      "learning_rate": 2e-05,
      "loss": 0.8922122955322266,
      "step": 940
    },
    {
      "epoch": 0.05125161847216228,
      "grad_norm": 1.271585738790212,
      "learning_rate": 2e-05,
      "loss": 0.8702299118041992,
      "step": 950
    },
    {
      "epoch": 0.05179110919292188,
      "grad_norm": 3.1087560610639984,
      "learning_rate": 2e-05,
      "loss": 0.869837474822998,
      "step": 960
    },
    {
      "epoch": 0.05233059991368148,
      "grad_norm": 1.336885754811761,
      "learning_rate": 2e-05,
      "loss": 0.8768095016479492,
      "step": 970
    },
    {
      "epoch": 0.052870090634441085,
      "grad_norm": 1.2937187665672654,
      "learning_rate": 2e-05,
      "loss": 0.8737648010253907,
      "step": 980
    },
    {
      "epoch": 0.05340958135520069,
      "grad_norm": 6.496337599049736,
      "learning_rate": 2e-05,
      "loss": 0.8852460861206055,
      "step": 990
    },
    {
      "epoch": 0.053949072075960294,
      "grad_norm": 2.314909026977299,
      "learning_rate": 2e-05,
      "loss": 0.9168342590332031,
      "step": 1000
    },
    {
      "epoch": 0.054488562796719896,
      "grad_norm": 1.3797589526126182,
      "learning_rate": 2e-05,
      "loss": 0.9012920379638671,
      "step": 1010
    },
    {
      "epoch": 0.0550280535174795,
      "grad_norm": 1.7456393044082175,
      "learning_rate": 2e-05,
      "loss": 0.8839190483093262,
      "step": 1020
    },
    {
      "epoch": 0.055567544238239105,
      "grad_norm": 2.2368924316262913,
      "learning_rate": 2e-05,
      "loss": 0.8841041564941406,
      "step": 1030
    },
    {
      "epoch": 0.05610703495899871,
      "grad_norm": 2.3060976347623408,
      "learning_rate": 2e-05,
      "loss": 0.9180487632751465,
      "step": 1040
    },
    {
      "epoch": 0.05664652567975831,
      "grad_norm": 2.2473033744197655,
      "learning_rate": 2e-05,
      "loss": 0.8983830451965332,
      "step": 1050
    },
    {
      "epoch": 0.05718601640051791,
      "grad_norm": 3.33794913543599,
      "learning_rate": 2e-05,
      "loss": 0.9058303833007812,
      "step": 1060
    },
    {
      "epoch": 0.05772550712127751,
      "grad_norm": 2.1863704996804874,
      "learning_rate": 2e-05,
      "loss": 0.8500079154968262,
      "step": 1070
    },
    {
      "epoch": 0.05826499784203712,
      "grad_norm": 5.464104262184261,
      "learning_rate": 2e-05,
      "loss": 0.86676025390625,
      "step": 1080
    },
    {
      "epoch": 0.05880448856279672,
      "grad_norm": 1.2580865472126652,
      "learning_rate": 2e-05,
      "loss": 0.8795917510986329,
      "step": 1090
    },
    {
      "epoch": 0.05934397928355632,
      "grad_norm": 1.5392878698728925,
      "learning_rate": 2e-05,
      "loss": 0.855427360534668,
      "step": 1100
    },
    {
      "epoch": 0.059883470004315924,
      "grad_norm": 1.6600456111757242,
      "learning_rate": 2e-05,
      "loss": 0.8580084800720215,
      "step": 1110
    },
    {
      "epoch": 0.06042296072507553,
      "grad_norm": 1.9728666357091431,
      "learning_rate": 2e-05,
      "loss": 0.8592136383056641,
      "step": 1120
    },
    {
      "epoch": 0.060962451445835134,
      "grad_norm": 2.622532245761096,
      "learning_rate": 2e-05,
      "loss": 0.8778785705566406,
      "step": 1130
    },
    {
      "epoch": 0.061501942166594735,
      "grad_norm": 2.436300801764917,
      "learning_rate": 2e-05,
      "loss": 0.8558690071105957,
      "step": 1140
    },
    {
      "epoch": 0.062041432887354336,
      "grad_norm": 2.072486307736441,
      "learning_rate": 2e-05,
      "loss": 0.8554096221923828,
      "step": 1150
    },
    {
      "epoch": 0.06258092360811394,
      "grad_norm": 2.9578754600495105,
      "learning_rate": 2e-05,
      "loss": 0.8688982963562012,
      "step": 1160
    },
    {
      "epoch": 0.06312041432887354,
      "grad_norm": 2.2185001887709106,
      "learning_rate": 2e-05,
      "loss": 0.8960395812988281,
      "step": 1170
    },
    {
      "epoch": 0.06365990504963315,
      "grad_norm": 2.39317741084652,
      "learning_rate": 2e-05,
      "loss": 0.8809322357177735,
      "step": 1180
    },
    {
      "epoch": 0.06419939577039276,
      "grad_norm": 7.053593954669792,
      "learning_rate": 2e-05,
      "loss": 0.8588828086853028,
      "step": 1190
    },
    {
      "epoch": 0.06473888649115235,
      "grad_norm": 2.8316276030407392,
      "learning_rate": 2e-05,
      "loss": 0.863832664489746,
      "step": 1200
    },
    {
      "epoch": 0.06527837721191196,
      "grad_norm": 55.263846840104094,
      "learning_rate": 2e-05,
      "loss": 0.8467828750610351,
      "step": 1210
    },
    {
      "epoch": 0.06581786793267155,
      "grad_norm": 1.891195934618861,
      "learning_rate": 2e-05,
      "loss": 0.8810437202453614,
      "step": 1220
    },
    {
      "epoch": 0.06635735865343116,
      "grad_norm": 4.262649131447584,
      "learning_rate": 2e-05,
      "loss": 0.8573444366455079,
      "step": 1230
    },
    {
      "epoch": 0.06689684937419077,
      "grad_norm": 5.280398859967369,
      "learning_rate": 2e-05,
      "loss": 0.8775759696960449,
      "step": 1240
    },
    {
      "epoch": 0.06743634009495036,
      "grad_norm": 13.161950494331434,
      "learning_rate": 2e-05,
      "loss": 0.8802067756652832,
      "step": 1250
    },
    {
      "epoch": 0.06797583081570997,
      "grad_norm": 2.3840227633253708,
      "learning_rate": 2e-05,
      "loss": 0.8745565414428711,
      "step": 1260
    },
    {
      "epoch": 0.06851532153646957,
      "grad_norm": 2.578881266139901,
      "learning_rate": 2e-05,
      "loss": 0.8929662704467773,
      "step": 1270
    },
    {
      "epoch": 0.06905481225722918,
      "grad_norm": 1.3551093714665399,
      "learning_rate": 2e-05,
      "loss": 0.8856276512145996,
      "step": 1280
    },
    {
      "epoch": 0.06959430297798878,
      "grad_norm": 1.421718866393892,
      "learning_rate": 2e-05,
      "loss": 0.8779384613037109,
      "step": 1290
    },
    {
      "epoch": 0.07013379369874838,
      "grad_norm": 1.3412253586411367,
      "learning_rate": 2e-05,
      "loss": 0.8418590545654296,
      "step": 1300
    },
    {
      "epoch": 0.07067328441950799,
      "grad_norm": 1.3164576570437523,
      "learning_rate": 2e-05,
      "loss": 0.8589597702026367,
      "step": 1310
    },
    {
      "epoch": 0.07121277514026758,
      "grad_norm": 5.63391297375713,
      "learning_rate": 2e-05,
      "loss": 0.8794595718383789,
      "step": 1320
    },
    {
      "epoch": 0.07175226586102719,
      "grad_norm": 1.3610004564150833,
      "learning_rate": 2e-05,
      "loss": 0.8666675567626954,
      "step": 1330
    },
    {
      "epoch": 0.0722917565817868,
      "grad_norm": 1.7200672230946694,
      "learning_rate": 2e-05,
      "loss": 0.8707124710083007,
      "step": 1340
    },
    {
      "epoch": 0.07283124730254639,
      "grad_norm": 1.3160601308789588,
      "learning_rate": 2e-05,
      "loss": 0.8457590103149414,
      "step": 1350
    },
    {
      "epoch": 0.073370738023306,
      "grad_norm": 1.7778775450463142,
      "learning_rate": 2e-05,
      "loss": 0.8669673919677734,
      "step": 1360
    },
    {
      "epoch": 0.0739102287440656,
      "grad_norm": 2.2742353103065485,
      "learning_rate": 2e-05,
      "loss": 0.920014762878418,
      "step": 1370
    },
    {
      "epoch": 0.0744497194648252,
      "grad_norm": 4.185305052537614,
      "learning_rate": 2e-05,
      "loss": 0.9066905021667481,
      "step": 1380
    },
    {
      "epoch": 0.07498921018558481,
      "grad_norm": 7.045183411349042,
      "learning_rate": 2e-05,
      "loss": 0.8720160484313965,
      "step": 1390
    },
    {
      "epoch": 0.0755287009063444,
      "grad_norm": 2.377204425821727,
      "learning_rate": 2e-05,
      "loss": 0.9035792350769043,
      "step": 1400
    },
    {
      "epoch": 0.07606819162710401,
      "grad_norm": 1.5514726466038995,
      "learning_rate": 2e-05,
      "loss": 0.8550067901611328,
      "step": 1410
    },
    {
      "epoch": 0.07660768234786362,
      "grad_norm": 3.2194246326766627,
      "learning_rate": 2e-05,
      "loss": 0.8613847732543946,
      "step": 1420
    },
    {
      "epoch": 0.07714717306862322,
      "grad_norm": 2.191413489389274,
      "learning_rate": 2e-05,
      "loss": 0.8483789443969727,
      "step": 1430
    },
    {
      "epoch": 0.07768666378938283,
      "grad_norm": 2.561083258333701,
      "learning_rate": 2e-05,
      "loss": 0.8561199188232422,
      "step": 1440
    },
    {
      "epoch": 0.07822615451014242,
      "grad_norm": 2.799486266864444,
      "learning_rate": 2e-05,
      "loss": 0.8749614715576172,
      "step": 1450
    },
    {
      "epoch": 0.07876564523090203,
      "grad_norm": 1.6625068228812598,
      "learning_rate": 2e-05,
      "loss": 0.8524051666259765,
      "step": 1460
    },
    {
      "epoch": 0.07930513595166164,
      "grad_norm": 1.5627273498445615,
      "learning_rate": 2e-05,
      "loss": 0.8613437652587891,
      "step": 1470
    },
    {
      "epoch": 0.07984462667242123,
      "grad_norm": 3.1187718040693815,
      "learning_rate": 2e-05,
      "loss": 0.8489472389221191,
      "step": 1480
    },
    {
      "epoch": 0.08038411739318084,
      "grad_norm": 1.999242227430214,
      "learning_rate": 2e-05,
      "loss": 0.8557394027709961,
      "step": 1490
    },
    {
      "epoch": 0.08092360811394043,
      "grad_norm": 2.1662789248263086,
      "learning_rate": 2e-05,
      "loss": 0.8410913467407226,
      "step": 1500
    },
    {
      "epoch": 0.08146309883470004,
      "grad_norm": 1.117593071600269,
      "learning_rate": 2e-05,
      "loss": 0.8719884872436523,
      "step": 1510
    },
    {
      "epoch": 0.08200258955545965,
      "grad_norm": 1.7282463420500256,
      "learning_rate": 2e-05,
      "loss": 0.8429347991943359,
      "step": 1520
    },
    {
      "epoch": 0.08254208027621925,
      "grad_norm": 1.10160222981424,
      "learning_rate": 2e-05,
      "loss": 0.8589319229125977,
      "step": 1530
    },
    {
      "epoch": 0.08308157099697885,
      "grad_norm": 1.657648568922712,
      "learning_rate": 2e-05,
      "loss": 0.8519350051879883,
      "step": 1540
    },
    {
      "epoch": 0.08362106171773845,
      "grad_norm": 1.2061011251185034,
      "learning_rate": 2e-05,
      "loss": 0.8548383712768555,
      "step": 1550
    },
    {
      "epoch": 0.08416055243849806,
      "grad_norm": 1.6503845091419074,
      "learning_rate": 2e-05,
      "loss": 0.8504886627197266,
      "step": 1560
    },
    {
      "epoch": 0.08470004315925767,
      "grad_norm": 1.650450748636652,
      "learning_rate": 2e-05,
      "loss": 0.8272921562194824,
      "step": 1570
    },
    {
      "epoch": 0.08523953388001726,
      "grad_norm": 1.2153538779996496,
      "learning_rate": 2e-05,
      "loss": 0.8586090087890625,
      "step": 1580
    },
    {
      "epoch": 0.08577902460077687,
      "grad_norm": 1.4119654167230689,
      "learning_rate": 2e-05,
      "loss": 0.8150126457214355,
      "step": 1590
    },
    {
      "epoch": 0.08631851532153648,
      "grad_norm": 2.0481812733972435,
      "learning_rate": 2e-05,
      "loss": 0.8605268478393555,
      "step": 1600
    },
    {
      "epoch": 0.08685800604229607,
      "grad_norm": 1.3493621359452488,
      "learning_rate": 2e-05,
      "loss": 0.8761652946472168,
      "step": 1610
    },
    {
      "epoch": 0.08739749676305568,
      "grad_norm": 1.937128776151837,
      "learning_rate": 2e-05,
      "loss": 0.8717695236206054,
      "step": 1620
    },
    {
      "epoch": 0.08793698748381527,
      "grad_norm": 1.2331164884742856,
      "learning_rate": 2e-05,
      "loss": 0.8695238113403321,
      "step": 1630
    },
    {
      "epoch": 0.08847647820457488,
      "grad_norm": 1.0766868462241113,
      "learning_rate": 2e-05,
      "loss": 0.8439886093139648,
      "step": 1640
    },
    {
      "epoch": 0.08901596892533449,
      "grad_norm": 4.825188821823706,
      "learning_rate": 2e-05,
      "loss": 0.8383602142333985,
      "step": 1650
    },
    {
      "epoch": 0.08955545964609408,
      "grad_norm": 1.0473009888263762,
      "learning_rate": 2e-05,
      "loss": 0.8446158409118653,
      "step": 1660
    },
    {
      "epoch": 0.0900949503668537,
      "grad_norm": 2.1550692634060407,
      "learning_rate": 2e-05,
      "loss": 0.8526595115661622,
      "step": 1670
    },
    {
      "epoch": 0.09063444108761329,
      "grad_norm": 1.1805712525240586,
      "learning_rate": 2e-05,
      "loss": 0.8445590019226075,
      "step": 1680
    },
    {
      "epoch": 0.0911739318083729,
      "grad_norm": 1.3154682778410502,
      "learning_rate": 2e-05,
      "loss": 0.8289753913879394,
      "step": 1690
    },
    {
      "epoch": 0.0917134225291325,
      "grad_norm": 1.7425671016352309,
      "learning_rate": 2e-05,
      "loss": 0.8550616264343261,
      "step": 1700
    },
    {
      "epoch": 0.0922529132498921,
      "grad_norm": 1.809326593705475,
      "learning_rate": 2e-05,
      "loss": 0.8592042922973633,
      "step": 1710
    },
    {
      "epoch": 0.09279240397065171,
      "grad_norm": 2.294726987915818,
      "learning_rate": 2e-05,
      "loss": 0.8464397430419922,
      "step": 1720
    },
    {
      "epoch": 0.0933318946914113,
      "grad_norm": 1.3341335707857838,
      "learning_rate": 2e-05,
      "loss": 0.8401545524597168,
      "step": 1730
    },
    {
      "epoch": 0.09387138541217091,
      "grad_norm": 1.7059856939535623,
      "learning_rate": 2e-05,
      "loss": 0.850313663482666,
      "step": 1740
    },
    {
      "epoch": 0.09441087613293052,
      "grad_norm": 1.1913305227905306,
      "learning_rate": 2e-05,
      "loss": 0.8246526718139648,
      "step": 1750
    },
    {
      "epoch": 0.09495036685369011,
      "grad_norm": 2.4712263811656063,
      "learning_rate": 2e-05,
      "loss": 0.8324945449829102,
      "step": 1760
    },
    {
      "epoch": 0.09548985757444972,
      "grad_norm": 1.2460535060628168,
      "learning_rate": 2e-05,
      "loss": 0.83499755859375,
      "step": 1770
    },
    {
      "epoch": 0.09602934829520932,
      "grad_norm": 0.9541283688175325,
      "learning_rate": 2e-05,
      "loss": 0.8438259124755859,
      "step": 1780
    },
    {
      "epoch": 0.09656883901596892,
      "grad_norm": 4.15270429876545,
      "learning_rate": 2e-05,
      "loss": 0.8374526977539063,
      "step": 1790
    },
    {
      "epoch": 0.09710832973672853,
      "grad_norm": 1.2843301784727401,
      "learning_rate": 2e-05,
      "loss": 0.855588722229004,
      "step": 1800
    },
    {
      "epoch": 0.09764782045748813,
      "grad_norm": 1.2852787584009888,
      "learning_rate": 2e-05,
      "loss": 0.8446294784545898,
      "step": 1810
    },
    {
      "epoch": 0.09818731117824774,
      "grad_norm": 1.1732513743084148,
      "learning_rate": 2e-05,
      "loss": 0.8550775527954102,
      "step": 1820
    },
    {
      "epoch": 0.09872680189900734,
      "grad_norm": 1.0484004988454474,
      "learning_rate": 2e-05,
      "loss": 0.8461496353149414,
      "step": 1830
    },
    {
      "epoch": 0.09926629261976694,
      "grad_norm": 1.1796374157652922,
      "learning_rate": 2e-05,
      "loss": 0.8495699882507324,
      "step": 1840
    },
    {
      "epoch": 0.09980578334052655,
      "grad_norm": 1.530009834654909,
      "learning_rate": 2e-05,
      "loss": 0.8529845237731933,
      "step": 1850
    },
    {
      "epoch": 0.10034527406128614,
      "grad_norm": 1.3821515513567662,
      "learning_rate": 2e-05,
      "loss": 0.8383827209472656,
      "step": 1860
    },
    {
      "epoch": 0.10088476478204575,
      "grad_norm": 1.433339075396923,
      "learning_rate": 2e-05,
      "loss": 0.8441427230834961,
      "step": 1870
    },
    {
      "epoch": 0.10142425550280536,
      "grad_norm": 1.4202627926244262,
      "learning_rate": 2e-05,
      "loss": 0.8355476379394531,
      "step": 1880
    },
    {
      "epoch": 0.10196374622356495,
      "grad_norm": 1.275529436709381,
      "learning_rate": 2e-05,
      "loss": 0.8475621223449707,
      "step": 1890
    },
    {
      "epoch": 0.10250323694432456,
      "grad_norm": 4.886608266675648,
      "learning_rate": 2e-05,
      "loss": 0.8222907066345215,
      "step": 1900
    },
    {
      "epoch": 0.10304272766508416,
      "grad_norm": 1.1032859405565207,
      "learning_rate": 2e-05,
      "loss": 0.8717263221740723,
      "step": 1910
    },
    {
      "epoch": 0.10358221838584376,
      "grad_norm": 1.4792708796316567,
      "learning_rate": 2e-05,
      "loss": 0.8422025680541992,
      "step": 1920
    },
    {
      "epoch": 0.10412170910660337,
      "grad_norm": 1.052829235330944,
      "learning_rate": 2e-05,
      "loss": 0.8275224685668945,
      "step": 1930
    },
    {
      "epoch": 0.10466119982736297,
      "grad_norm": 1.5243538008576782,
      "learning_rate": 2e-05,
      "loss": 0.8201045989990234,
      "step": 1940
    },
    {
      "epoch": 0.10520069054812257,
      "grad_norm": 1.8549508993672652,
      "learning_rate": 2e-05,
      "loss": 0.8507599830627441,
      "step": 1950
    },
    {
      "epoch": 0.10574018126888217,
      "grad_norm": 3.2676775863279293,
      "learning_rate": 2e-05,
      "loss": 0.8315736770629882,
      "step": 1960
    },
    {
      "epoch": 0.10627967198964178,
      "grad_norm": 3.170505632566329,
      "learning_rate": 2e-05,
      "loss": 0.8597011566162109,
      "step": 1970
    },
    {
      "epoch": 0.10681916271040139,
      "grad_norm": 1.225518275743383,
      "learning_rate": 2e-05,
      "loss": 0.8449475288391113,
      "step": 1980
    },
    {
      "epoch": 0.10735865343116098,
      "grad_norm": 1.7639666202458668,
      "learning_rate": 2e-05,
      "loss": 0.8267492294311524,
      "step": 1990
    },
    {
      "epoch": 0.10789814415192059,
      "grad_norm": 1.7820697092202835,
      "learning_rate": 2e-05,
      "loss": 0.8312843322753907,
      "step": 2000
    },
    {
      "epoch": 0.10843763487268018,
      "grad_norm": 4.015953726833389,
      "learning_rate": 2e-05,
      "loss": 0.8443691253662109,
      "step": 2010
    },
    {
      "epoch": 0.10897712559343979,
      "grad_norm": 1.999195946930485,
      "learning_rate": 2e-05,
      "loss": 0.8289900779724121,
      "step": 2020
    },
    {
      "epoch": 0.1095166163141994,
      "grad_norm": 1.8651265962731969,
      "learning_rate": 2e-05,
      "loss": 0.840304946899414,
      "step": 2030
    },
    {
      "epoch": 0.110056107034959,
      "grad_norm": 1.7034465477351663,
      "learning_rate": 2e-05,
      "loss": 0.8282983779907227,
      "step": 2040
    },
    {
      "epoch": 0.1105955977557186,
      "grad_norm": 1.292874343679415,
      "learning_rate": 2e-05,
      "loss": 0.8187969207763672,
      "step": 2050
    },
    {
      "epoch": 0.11113508847647821,
      "grad_norm": 9.900066428760711,
      "learning_rate": 2e-05,
      "loss": 0.8275846481323242,
      "step": 2060
    },
    {
      "epoch": 0.1116745791972378,
      "grad_norm": 1.4024064010510686,
      "learning_rate": 2e-05,
      "loss": 0.8306047439575195,
      "step": 2070
    },
    {
      "epoch": 0.11221406991799741,
      "grad_norm": 2.400290796313511,
      "learning_rate": 2e-05,
      "loss": 0.8676346778869629,
      "step": 2080
    },
    {
      "epoch": 0.11275356063875701,
      "grad_norm": 3.514349150867208,
      "learning_rate": 2e-05,
      "loss": 0.8274039268493653,
      "step": 2090
    },
    {
      "epoch": 0.11329305135951662,
      "grad_norm": 5.088709612171823,
      "learning_rate": 2e-05,
      "loss": 0.8288514137268066,
      "step": 2100
    },
    {
      "epoch": 0.11383254208027623,
      "grad_norm": 2.098725303443767,
      "learning_rate": 2e-05,
      "loss": 0.8381128311157227,
      "step": 2110
    },
    {
      "epoch": 0.11437203280103582,
      "grad_norm": 1.3305492466588589,
      "learning_rate": 2e-05,
      "loss": 0.8685014724731446,
      "step": 2120
    },
    {
      "epoch": 0.11491152352179543,
      "grad_norm": 2.7110556064904956,
      "learning_rate": 2e-05,
      "loss": 0.8481921195983887,
      "step": 2130
    },
    {
      "epoch": 0.11545101424255502,
      "grad_norm": 1.10367668589633,
      "learning_rate": 2e-05,
      "loss": 0.8573419570922851,
      "step": 2140
    },
    {
      "epoch": 0.11599050496331463,
      "grad_norm": 2.138417581209623,
      "learning_rate": 2e-05,
      "loss": 0.8302045822143554,
      "step": 2150
    },
    {
      "epoch": 0.11652999568407424,
      "grad_norm": 1.1112744471980331,
      "learning_rate": 2e-05,
      "loss": 0.8130320549011231,
      "step": 2160
    },
    {
      "epoch": 0.11706948640483383,
      "grad_norm": 1.6334189344204764,
      "learning_rate": 2e-05,
      "loss": 0.8478602409362793,
      "step": 2170
    },
    {
      "epoch": 0.11760897712559344,
      "grad_norm": 2.7583960750005816,
      "learning_rate": 2e-05,
      "loss": 0.8246549606323242,
      "step": 2180
    },
    {
      "epoch": 0.11814846784635304,
      "grad_norm": 1.9058948199400891,
      "learning_rate": 2e-05,
      "loss": 0.8250743865966796,
      "step": 2190
    },
    {
      "epoch": 0.11868795856711264,
      "grad_norm": 1.586316165202703,
      "learning_rate": 2e-05,
      "loss": 0.8306868553161622,
      "step": 2200
    },
    {
      "epoch": 0.11922744928787225,
      "grad_norm": 3.6562916703913606,
      "learning_rate": 2e-05,
      "loss": 0.8404013633728027,
      "step": 2210
    },
    {
      "epoch": 0.11976694000863185,
      "grad_norm": 1.914570415791936,
      "learning_rate": 2e-05,
      "loss": 0.8350730895996094,
      "step": 2220
    },
    {
      "epoch": 0.12030643072939146,
      "grad_norm": 1.4049455715505264,
      "learning_rate": 2e-05,
      "loss": 0.8396036148071289,
      "step": 2230
    },
    {
      "epoch": 0.12084592145015106,
      "grad_norm": 1.1836748795523662,
      "learning_rate": 2e-05,
      "loss": 0.8288493156433105,
      "step": 2240
    },
    {
      "epoch": 0.12138541217091066,
      "grad_norm": 2.5165436971018735,
      "learning_rate": 2e-05,
      "loss": 0.8394759178161622,
      "step": 2250
    },
    {
      "epoch": 0.12192490289167027,
      "grad_norm": 1.8023654123652022,
      "learning_rate": 2e-05,
      "loss": 0.8561839103698731,
      "step": 2260
    },
    {
      "epoch": 0.12246439361242986,
      "grad_norm": 2.197758047068318,
      "learning_rate": 2e-05,
      "loss": 0.8342147827148437,
      "step": 2270
    },
    {
      "epoch": 0.12300388433318947,
      "grad_norm": 1.3512470884395074,
      "learning_rate": 2e-05,
      "loss": 0.8416089057922364,
      "step": 2280
    },
    {
      "epoch": 0.12354337505394908,
      "grad_norm": 1.158221169858105,
      "learning_rate": 2e-05,
      "loss": 0.815174674987793,
      "step": 2290
    },
    {
      "epoch": 0.12408286577470867,
      "grad_norm": 2.1931315632737336,
      "learning_rate": 2e-05,
      "loss": 0.8198381423950195,
      "step": 2300
    },
    {
      "epoch": 0.12462235649546828,
      "grad_norm": 1.2020698197548285,
      "learning_rate": 2e-05,
      "loss": 0.8212334632873535,
      "step": 2310
    },
    {
      "epoch": 0.1251618472162279,
      "grad_norm": 1.2630839020259081,
      "learning_rate": 2e-05,
      "loss": 0.8234346389770508,
      "step": 2320
    },
    {
      "epoch": 0.12570133793698748,
      "grad_norm": 1.05835478339789,
      "learning_rate": 2e-05,
      "loss": 0.8266730308532715,
      "step": 2330
    },
    {
      "epoch": 0.12624082865774708,
      "grad_norm": 1.153807197212215,
      "learning_rate": 2e-05,
      "loss": 0.8297053337097168,
      "step": 2340
    },
    {
      "epoch": 0.1267803193785067,
      "grad_norm": 1.081701471063699,
      "learning_rate": 2e-05,
      "loss": 0.850483512878418,
      "step": 2350
    },
    {
      "epoch": 0.1273198100992663,
      "grad_norm": 1.1032276472226967,
      "learning_rate": 2e-05,
      "loss": 0.8351642608642578,
      "step": 2360
    },
    {
      "epoch": 0.1278593008200259,
      "grad_norm": 0.9373186180851497,
      "learning_rate": 2e-05,
      "loss": 0.850758171081543,
      "step": 2370
    },
    {
      "epoch": 0.1283987915407855,
      "grad_norm": 0.9185447727528806,
      "learning_rate": 2e-05,
      "loss": 0.8138019561767578,
      "step": 2380
    },
    {
      "epoch": 0.1289382822615451,
      "grad_norm": 1.4547050261067835,
      "learning_rate": 2e-05,
      "loss": 0.8109475135803222,
      "step": 2390
    },
    {
      "epoch": 0.1294777729823047,
      "grad_norm": 0.8962680388570466,
      "learning_rate": 2e-05,
      "loss": 0.8358553886413574,
      "step": 2400
    },
    {
      "epoch": 0.1300172637030643,
      "grad_norm": 0.9967276931315883,
      "learning_rate": 2e-05,
      "loss": 0.8311929702758789,
      "step": 2410
    },
    {
      "epoch": 0.13055675442382392,
      "grad_norm": 1.171890528912994,
      "learning_rate": 2e-05,
      "loss": 0.8504758834838867,
      "step": 2420
    },
    {
      "epoch": 0.1310962451445835,
      "grad_norm": 0.9664067450616619,
      "learning_rate": 2e-05,
      "loss": 0.8218025207519531,
      "step": 2430
    },
    {
      "epoch": 0.1316357358653431,
      "grad_norm": 1.0236162825206507,
      "learning_rate": 2e-05,
      "loss": 0.8196083068847656,
      "step": 2440
    },
    {
      "epoch": 0.13217522658610273,
      "grad_norm": 1.9012584301013733,
      "learning_rate": 2e-05,
      "loss": 0.823670768737793,
      "step": 2450
    },
    {
      "epoch": 0.13271471730686232,
      "grad_norm": 1.1023811611206806,
      "learning_rate": 2e-05,
      "loss": 0.8224227905273438,
      "step": 2460
    },
    {
      "epoch": 0.13325420802762192,
      "grad_norm": 9.134961938530486,
      "learning_rate": 2e-05,
      "loss": 0.8160662651062012,
      "step": 2470
    },
    {
      "epoch": 0.13379369874838154,
      "grad_norm": 4.065113504605792,
      "learning_rate": 2e-05,
      "loss": 0.842041015625,
      "step": 2480
    },
    {
      "epoch": 0.13433318946914113,
      "grad_norm": 6.789485151507618,
      "learning_rate": 2e-05,
      "loss": 0.8334197998046875,
      "step": 2490
    },
    {
      "epoch": 0.13487268018990073,
      "grad_norm": 3.755408787739404,
      "learning_rate": 2e-05,
      "loss": 0.8373711585998536,
      "step": 2500
    },
    {
      "epoch": 0.13541217091066032,
      "grad_norm": 4.1766718888496115,
      "learning_rate": 2e-05,
      "loss": 0.826263427734375,
      "step": 2510
    },
    {
      "epoch": 0.13595166163141995,
      "grad_norm": 1.4503676530546985,
      "learning_rate": 2e-05,
      "loss": 0.8411909103393554,
      "step": 2520
    },
    {
      "epoch": 0.13649115235217954,
      "grad_norm": 2.9419252466356443,
      "learning_rate": 2e-05,
      "loss": 0.8234809875488281,
      "step": 2530
    },
    {
      "epoch": 0.13703064307293913,
      "grad_norm": 3.2828905803244144,
      "learning_rate": 2e-05,
      "loss": 0.8695075035095214,
      "step": 2540
    },
    {
      "epoch": 0.13757013379369876,
      "grad_norm": 1.096460041177136,
      "learning_rate": 2e-05,
      "loss": 0.8468042373657226,
      "step": 2550
    },
    {
      "epoch": 0.13810962451445835,
      "grad_norm": 1.0597785888179652,
      "learning_rate": 2e-05,
      "loss": 0.8341262817382813,
      "step": 2560
    },
    {
      "epoch": 0.13864911523521795,
      "grad_norm": 9.087223552942303,
      "learning_rate": 2e-05,
      "loss": 0.8169255256652832,
      "step": 2570
    },
    {
      "epoch": 0.13918860595597757,
      "grad_norm": 7.72894258287119,
      "learning_rate": 2e-05,
      "loss": 0.8361001014709473,
      "step": 2580
    },
    {
      "epoch": 0.13972809667673716,
      "grad_norm": 1.84749896259604,
      "learning_rate": 2e-05,
      "loss": 0.8118861198425293,
      "step": 2590
    },
    {
      "epoch": 0.14026758739749676,
      "grad_norm": 1.499923768048581,
      "learning_rate": 2e-05,
      "loss": 0.8059223175048829,
      "step": 2600
    },
    {
      "epoch": 0.14080707811825638,
      "grad_norm": 2.469774702348604,
      "learning_rate": 2e-05,
      "loss": 0.8464885711669922,
      "step": 2610
    },
    {
      "epoch": 0.14134656883901597,
      "grad_norm": 1.4562619868061146,
      "learning_rate": 2e-05,
      "loss": 0.855031681060791,
      "step": 2620
    },
    {
      "epoch": 0.14188605955977557,
      "grad_norm": 1.1580412885044675,
      "learning_rate": 2e-05,
      "loss": 0.8250855445861817,
      "step": 2630
    },
    {
      "epoch": 0.14242555028053516,
      "grad_norm": 5.85834983160985,
      "learning_rate": 2e-05,
      "loss": 0.8170978546142578,
      "step": 2640
    },
    {
      "epoch": 0.14296504100129478,
      "grad_norm": 2.1118214513392153,
      "learning_rate": 2e-05,
      "loss": 0.8098548889160156,
      "step": 2650
    },
    {
      "epoch": 0.14350453172205438,
      "grad_norm": 14.992058801195402,
      "learning_rate": 2e-05,
      "loss": 0.8217266082763672,
      "step": 2660
    },
    {
      "epoch": 0.14404402244281397,
      "grad_norm": 2.3085553913502035,
      "learning_rate": 2e-05,
      "loss": 0.8170116424560547,
      "step": 2670
    },
    {
      "epoch": 0.1445835131635736,
      "grad_norm": 17.707266771089483,
      "learning_rate": 2e-05,
      "loss": 0.8280685424804688,
      "step": 2680
    },
    {
      "epoch": 0.1451230038843332,
      "grad_norm": 3.1378615826812766,
      "learning_rate": 2e-05,
      "loss": 0.8434804916381836,
      "step": 2690
    },
    {
      "epoch": 0.14566249460509278,
      "grad_norm": 2.5515862288748354,
      "learning_rate": 2e-05,
      "loss": 0.8259258270263672,
      "step": 2700
    },
    {
      "epoch": 0.1462019853258524,
      "grad_norm": 3.9230866849646424,
      "learning_rate": 2e-05,
      "loss": 0.8185581207275391,
      "step": 2710
    },
    {
      "epoch": 0.146741476046612,
      "grad_norm": 3.2217451837996975,
      "learning_rate": 2e-05,
      "loss": 0.8379020690917969,
      "step": 2720
    },
    {
      "epoch": 0.1472809667673716,
      "grad_norm": 1.557738526634509,
      "learning_rate": 2e-05,
      "loss": 0.8445243835449219,
      "step": 2730
    },
    {
      "epoch": 0.1478204574881312,
      "grad_norm": 1.5481180648756823,
      "learning_rate": 2e-05,
      "loss": 0.8328718185424805,
      "step": 2740
    },
    {
      "epoch": 0.1483599482088908,
      "grad_norm": 1.4388832631294857,
      "learning_rate": 2e-05,
      "loss": 0.8253545761108398,
      "step": 2750
    },
    {
      "epoch": 0.1488994389296504,
      "grad_norm": 1.0539255503501914,
      "learning_rate": 2e-05,
      "loss": 0.8445590972900391,
      "step": 2760
    },
    {
      "epoch": 0.14943892965041,
      "grad_norm": 4.95376827487241,
      "learning_rate": 2e-05,
      "loss": 0.8173467636108398,
      "step": 2770
    },
    {
      "epoch": 0.14997842037116962,
      "grad_norm": 4.600755881351036,
      "learning_rate": 2e-05,
      "loss": 0.8473451614379883,
      "step": 2780
    },
    {
      "epoch": 0.15051791109192922,
      "grad_norm": 1.3528051911908148,
      "learning_rate": 2e-05,
      "loss": 0.8311235427856445,
      "step": 2790
    },
    {
      "epoch": 0.1510574018126888,
      "grad_norm": 2.5244296910788786,
      "learning_rate": 2e-05,
      "loss": 0.8462516784667968,
      "step": 2800
    },
    {
      "epoch": 0.15159689253344844,
      "grad_norm": 1.3547145890263894,
      "learning_rate": 2e-05,
      "loss": 0.8145876884460449,
      "step": 2810
    },
    {
      "epoch": 0.15213638325420803,
      "grad_norm": 4.904666690029179,
      "learning_rate": 2e-05,
      "loss": 0.8427519798278809,
      "step": 2820
    },
    {
      "epoch": 0.15267587397496762,
      "grad_norm": 1.609216240729035,
      "learning_rate": 2e-05,
      "loss": 0.8435686111450196,
      "step": 2830
    },
    {
      "epoch": 0.15321536469572725,
      "grad_norm": 1.4281030100315293,
      "learning_rate": 2e-05,
      "loss": 0.8191211700439454,
      "step": 2840
    },
    {
      "epoch": 0.15375485541648684,
      "grad_norm": 1.7798898314789362,
      "learning_rate": 2e-05,
      "loss": 0.8122947692871094,
      "step": 2850
    },
    {
      "epoch": 0.15429434613724644,
      "grad_norm": 2.830774437932546,
      "learning_rate": 2e-05,
      "loss": 0.8318849563598633,
      "step": 2860
    },
    {
      "epoch": 0.15483383685800603,
      "grad_norm": 5.397744742205086,
      "learning_rate": 2e-05,
      "loss": 0.8305788993835449,
      "step": 2870
    },
    {
      "epoch": 0.15537332757876565,
      "grad_norm": 2.898600936491205,
      "learning_rate": 2e-05,
      "loss": 0.8408110618591309,
      "step": 2880
    },
    {
      "epoch": 0.15591281829952525,
      "grad_norm": 8.773413677216482,
      "learning_rate": 2e-05,
      "loss": 0.8318193435668946,
      "step": 2890
    },
    {
      "epoch": 0.15645230902028484,
      "grad_norm": 2.3331580039530095,
      "learning_rate": 2e-05,
      "loss": 0.8389785766601563,
      "step": 2900
    },
    {
      "epoch": 0.15699179974104446,
      "grad_norm": 2.568113164269533,
      "learning_rate": 2e-05,
      "loss": 0.8040918350219727,
      "step": 2910
    },
    {
      "epoch": 0.15753129046180406,
      "grad_norm": 1.4668399503612397,
      "learning_rate": 2e-05,
      "loss": 0.8159090042114258,
      "step": 2920
    },
    {
      "epoch": 0.15807078118256365,
      "grad_norm": 4.911971870615249,
      "learning_rate": 2e-05,
      "loss": 0.8376889228820801,
      "step": 2930
    },
    {
      "epoch": 0.15861027190332327,
      "grad_norm": 2.981636311233402,
      "learning_rate": 2e-05,
      "loss": 0.8562882423400879,
      "step": 2940
    },
    {
      "epoch": 0.15914976262408287,
      "grad_norm": 7.625355317159627,
      "learning_rate": 2e-05,
      "loss": 0.8109912872314453,
      "step": 2950
    },
    {
      "epoch": 0.15968925334484246,
      "grad_norm": 2.1473747341723555,
      "learning_rate": 2e-05,
      "loss": 0.852902603149414,
      "step": 2960
    },
    {
      "epoch": 0.16022874406560206,
      "grad_norm": 4.307544898725762,
      "learning_rate": 2e-05,
      "loss": 0.8281949043273926,
      "step": 2970
    },
    {
      "epoch": 0.16076823478636168,
      "grad_norm": 5.656390266767483,
      "learning_rate": 2e-05,
      "loss": 0.7931592464447021,
      "step": 2980
    },
    {
      "epoch": 0.16130772550712127,
      "grad_norm": 2.595677289608645,
      "learning_rate": 2e-05,
      "loss": 0.8464061737060546,
      "step": 2990
    },
    {
      "epoch": 0.16184721622788087,
      "grad_norm": 2.5286497672098376,
      "learning_rate": 2e-05,
      "loss": 0.8217041015625,
      "step": 3000
    },
    {
      "epoch": 0.1623867069486405,
      "grad_norm": 1.6815165067247406,
      "learning_rate": 2e-05,
      "loss": 0.8131906509399414,
      "step": 3010
    },
    {
      "epoch": 0.16292619766940009,
      "grad_norm": 4.470085181162879,
      "learning_rate": 2e-05,
      "loss": 0.8476617813110352,
      "step": 3020
    },
    {
      "epoch": 0.16346568839015968,
      "grad_norm": 57.757712059800895,
      "learning_rate": 2e-05,
      "loss": 0.8235321044921875,
      "step": 3030
    },
    {
      "epoch": 0.1640051791109193,
      "grad_norm": 1.3196900653412662,
      "learning_rate": 2e-05,
      "loss": 0.8411979675292969,
      "step": 3040
    },
    {
      "epoch": 0.1645446698316789,
      "grad_norm": 1.3565437997414518,
      "learning_rate": 2e-05,
      "loss": 0.8321975708007813,
      "step": 3050
    },
    {
      "epoch": 0.1650841605524385,
      "grad_norm": 1.9250529621406256,
      "learning_rate": 2e-05,
      "loss": 0.8183302879333496,
      "step": 3060
    },
    {
      "epoch": 0.1656236512731981,
      "grad_norm": 1.6665163841705182,
      "learning_rate": 2e-05,
      "loss": 0.8212745666503907,
      "step": 3070
    },
    {
      "epoch": 0.1661631419939577,
      "grad_norm": 0.9624218525439471,
      "learning_rate": 2e-05,
      "loss": 0.8360950469970703,
      "step": 3080
    },
    {
      "epoch": 0.1667026327147173,
      "grad_norm": 1.6154798724669914,
      "learning_rate": 2e-05,
      "loss": 0.8246768951416016,
      "step": 3090
    },
    {
      "epoch": 0.1672421234354769,
      "grad_norm": 1.4309300679224686,
      "learning_rate": 2e-05,
      "loss": 0.8143646240234375,
      "step": 3100
    },
    {
      "epoch": 0.16778161415623652,
      "grad_norm": 1.1954152653837478,
      "learning_rate": 2e-05,
      "loss": 0.8255207061767578,
      "step": 3110
    },
    {
      "epoch": 0.1683211048769961,
      "grad_norm": 3.874535720645589,
      "learning_rate": 2e-05,
      "loss": 0.8268468856811524,
      "step": 3120
    },
    {
      "epoch": 0.1688605955977557,
      "grad_norm": 1.666585578869163,
      "learning_rate": 2e-05,
      "loss": 0.8370718955993652,
      "step": 3130
    },
    {
      "epoch": 0.16940008631851533,
      "grad_norm": 2.1122716764285214,
      "learning_rate": 2e-05,
      "loss": 0.8222600936889648,
      "step": 3140
    },
    {
      "epoch": 0.16993957703927492,
      "grad_norm": 1.626502422759658,
      "learning_rate": 2e-05,
      "loss": 0.8307378768920899,
      "step": 3150
    },
    {
      "epoch": 0.17047906776003452,
      "grad_norm": 1.249709631021305,
      "learning_rate": 2e-05,
      "loss": 0.8274057388305665,
      "step": 3160
    },
    {
      "epoch": 0.17101855848079414,
      "grad_norm": 1.2755622837169793,
      "learning_rate": 2e-05,
      "loss": 0.8418668746948242,
      "step": 3170
    },
    {
      "epoch": 0.17155804920155374,
      "grad_norm": 1.1471931685322956,
      "learning_rate": 2e-05,
      "loss": 0.8037077903747558,
      "step": 3180
    },
    {
      "epoch": 0.17209753992231333,
      "grad_norm": 2.731067427188893,
      "learning_rate": 2e-05,
      "loss": 0.7964674949645996,
      "step": 3190
    },
    {
      "epoch": 0.17263703064307295,
      "grad_norm": 1.4738003450064086,
      "learning_rate": 2e-05,
      "loss": 0.8007509231567382,
      "step": 3200
    },
    {
      "epoch": 0.17317652136383255,
      "grad_norm": 1.310820117633026,
      "learning_rate": 2e-05,
      "loss": 0.8201332092285156,
      "step": 3210
    },
    {
      "epoch": 0.17371601208459214,
      "grad_norm": 1.3355379129188545,
      "learning_rate": 2e-05,
      "loss": 0.8006715774536133,
      "step": 3220
    },
    {
      "epoch": 0.17425550280535174,
      "grad_norm": 2.5523015565267078,
      "learning_rate": 2e-05,
      "loss": 0.8170543670654297,
      "step": 3230
    },
    {
      "epoch": 0.17479499352611136,
      "grad_norm": 7.648994384982638,
      "learning_rate": 2e-05,
      "loss": 0.7966285705566406,
      "step": 3240
    },
    {
      "epoch": 0.17533448424687095,
      "grad_norm": 2.2291376008212502,
      "learning_rate": 2e-05,
      "loss": 0.8066202163696289,
      "step": 3250
    },
    {
      "epoch": 0.17587397496763055,
      "grad_norm": 1.9013943798842645,
      "learning_rate": 2e-05,
      "loss": 0.8028726577758789,
      "step": 3260
    },
    {
      "epoch": 0.17641346568839017,
      "grad_norm": 1.2094542057997177,
      "learning_rate": 2e-05,
      "loss": 0.8179764747619629,
      "step": 3270
    },
    {
      "epoch": 0.17695295640914976,
      "grad_norm": 2.7367851475598,
      "learning_rate": 2e-05,
      "loss": 0.8103551864624023,
      "step": 3280
    },
    {
      "epoch": 0.17749244712990936,
      "grad_norm": 2.9058412897859656,
      "learning_rate": 2e-05,
      "loss": 0.8264286041259765,
      "step": 3290
    },
    {
      "epoch": 0.17803193785066898,
      "grad_norm": 3.226004030131601,
      "learning_rate": 2e-05,
      "loss": 0.8190610885620118,
      "step": 3300
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 1.8721988330858943,
      "learning_rate": 2e-05,
      "loss": 0.8306256294250488,
      "step": 3310
    },
    {
      "epoch": 0.17911091929218817,
      "grad_norm": 1.3899444238570022,
      "learning_rate": 2e-05,
      "loss": 0.8446199417114257,
      "step": 3320
    },
    {
      "epoch": 0.17965041001294776,
      "grad_norm": 1.0929088079648228,
      "learning_rate": 2e-05,
      "loss": 0.7976521968841552,
      "step": 3330
    },
    {
      "epoch": 0.1801899007337074,
      "grad_norm": 1.1793561049774377,
      "learning_rate": 2e-05,
      "loss": 0.8168546676635742,
      "step": 3340
    },
    {
      "epoch": 0.18072939145446698,
      "grad_norm": 15.176731004709248,
      "learning_rate": 2e-05,
      "loss": 0.8124425888061524,
      "step": 3350
    },
    {
      "epoch": 0.18126888217522658,
      "grad_norm": 1.0594125222809312,
      "learning_rate": 2e-05,
      "loss": 0.806308364868164,
      "step": 3360
    },
    {
      "epoch": 0.1818083728959862,
      "grad_norm": 1.9453105377863469,
      "learning_rate": 2e-05,
      "loss": 0.8045181274414063,
      "step": 3370
    },
    {
      "epoch": 0.1823478636167458,
      "grad_norm": 0.9815060321000388,
      "learning_rate": 2e-05,
      "loss": 0.8289067268371582,
      "step": 3380
    },
    {
      "epoch": 0.1828873543375054,
      "grad_norm": 1.107626410337324,
      "learning_rate": 2e-05,
      "loss": 0.8187108993530273,
      "step": 3390
    },
    {
      "epoch": 0.183426845058265,
      "grad_norm": 1.3713782627751003,
      "learning_rate": 2e-05,
      "loss": 0.8212367057800293,
      "step": 3400
    },
    {
      "epoch": 0.1839663357790246,
      "grad_norm": 1.1934434763982462,
      "learning_rate": 2e-05,
      "loss": 0.8170612335205079,
      "step": 3410
    },
    {
      "epoch": 0.1845058264997842,
      "grad_norm": 1.0151425442795299,
      "learning_rate": 2e-05,
      "loss": 0.8014205932617188,
      "step": 3420
    },
    {
      "epoch": 0.18504531722054382,
      "grad_norm": 0.9339054423212217,
      "learning_rate": 2e-05,
      "loss": 0.8339118003845215,
      "step": 3430
    },
    {
      "epoch": 0.18558480794130341,
      "grad_norm": 0.967368753176617,
      "learning_rate": 2e-05,
      "loss": 0.8132877349853516,
      "step": 3440
    },
    {
      "epoch": 0.186124298662063,
      "grad_norm": 2.162294345081955,
      "learning_rate": 2e-05,
      "loss": 0.8348134994506836,
      "step": 3450
    },
    {
      "epoch": 0.1866637893828226,
      "grad_norm": 2.687504589015093,
      "learning_rate": 2e-05,
      "loss": 0.8514577865600585,
      "step": 3460
    },
    {
      "epoch": 0.18720328010358223,
      "grad_norm": 1.3616314045359637,
      "learning_rate": 2e-05,
      "loss": 0.8360185623168945,
      "step": 3470
    },
    {
      "epoch": 0.18774277082434182,
      "grad_norm": 0.7945412849733318,
      "learning_rate": 2e-05,
      "loss": 0.8060104370117187,
      "step": 3480
    },
    {
      "epoch": 0.18828226154510141,
      "grad_norm": 1.2865284566965982,
      "learning_rate": 2e-05,
      "loss": 0.8192827224731445,
      "step": 3490
    },
    {
      "epoch": 0.18882175226586104,
      "grad_norm": 1.0470430689868928,
      "learning_rate": 2e-05,
      "loss": 0.8104745864868164,
      "step": 3500
    },
    {
      "epoch": 0.18936124298662063,
      "grad_norm": 1.0033888590154632,
      "learning_rate": 2e-05,
      "loss": 0.8070037841796875,
      "step": 3510
    },
    {
      "epoch": 0.18990073370738023,
      "grad_norm": 1.4237596335672749,
      "learning_rate": 2e-05,
      "loss": 0.7981664657592773,
      "step": 3520
    },
    {
      "epoch": 0.19044022442813985,
      "grad_norm": 0.9746666528478075,
      "learning_rate": 2e-05,
      "loss": 0.8370217323303223,
      "step": 3530
    },
    {
      "epoch": 0.19097971514889944,
      "grad_norm": 2.370032426413775,
      "learning_rate": 2e-05,
      "loss": 0.8089248657226562,
      "step": 3540
    },
    {
      "epoch": 0.19151920586965904,
      "grad_norm": 1.4096737349425104,
      "learning_rate": 2e-05,
      "loss": 0.8232223510742187,
      "step": 3550
    },
    {
      "epoch": 0.19205869659041863,
      "grad_norm": 1.5464845999590675,
      "learning_rate": 2e-05,
      "loss": 0.8215899467468262,
      "step": 3560
    },
    {
      "epoch": 0.19259818731117825,
      "grad_norm": 1.8902013609436816,
      "learning_rate": 2e-05,
      "loss": 0.8071002960205078,
      "step": 3570
    },
    {
      "epoch": 0.19313767803193785,
      "grad_norm": 1.0834961800168135,
      "learning_rate": 2e-05,
      "loss": 0.818499755859375,
      "step": 3580
    },
    {
      "epoch": 0.19367716875269744,
      "grad_norm": 1.469175436322324,
      "learning_rate": 2e-05,
      "loss": 0.8098075866699219,
      "step": 3590
    },
    {
      "epoch": 0.19421665947345707,
      "grad_norm": 1.4457828669467623,
      "learning_rate": 2e-05,
      "loss": 0.8190594673156738,
      "step": 3600
    },
    {
      "epoch": 0.19475615019421666,
      "grad_norm": 1.2339738565051197,
      "learning_rate": 2e-05,
      "loss": 0.8001772880554199,
      "step": 3610
    },
    {
      "epoch": 0.19529564091497625,
      "grad_norm": 1.0129969586529555,
      "learning_rate": 2e-05,
      "loss": 0.8169227600097656,
      "step": 3620
    },
    {
      "epoch": 0.19583513163573588,
      "grad_norm": 1.8405886663060382,
      "learning_rate": 2e-05,
      "loss": 0.8161848068237305,
      "step": 3630
    },
    {
      "epoch": 0.19637462235649547,
      "grad_norm": 0.8630501803188849,
      "learning_rate": 2e-05,
      "loss": 0.7952068328857422,
      "step": 3640
    },
    {
      "epoch": 0.19691411307725507,
      "grad_norm": 0.9252474483323625,
      "learning_rate": 2e-05,
      "loss": 0.836590576171875,
      "step": 3650
    },
    {
      "epoch": 0.1974536037980147,
      "grad_norm": 0.9338702892336649,
      "learning_rate": 2e-05,
      "loss": 0.8275209426879883,
      "step": 3660
    },
    {
      "epoch": 0.19799309451877428,
      "grad_norm": 0.9675019978413074,
      "learning_rate": 2e-05,
      "loss": 0.8117290496826172,
      "step": 3670
    },
    {
      "epoch": 0.19853258523953388,
      "grad_norm": 0.8692908353972926,
      "learning_rate": 2e-05,
      "loss": 0.8203290939331055,
      "step": 3680
    },
    {
      "epoch": 0.19907207596029347,
      "grad_norm": 0.9432619722907122,
      "learning_rate": 2e-05,
      "loss": 0.8062999725341797,
      "step": 3690
    },
    {
      "epoch": 0.1996115666810531,
      "grad_norm": 0.9465547041865168,
      "learning_rate": 2e-05,
      "loss": 0.8359698295593262,
      "step": 3700
    },
    {
      "epoch": 0.2001510574018127,
      "grad_norm": 0.8544475554924795,
      "learning_rate": 2e-05,
      "loss": 0.8113945960998535,
      "step": 3710
    },
    {
      "epoch": 0.20069054812257228,
      "grad_norm": 1.0261458270031636,
      "learning_rate": 2e-05,
      "loss": 0.8012717247009278,
      "step": 3720
    },
    {
      "epoch": 0.2012300388433319,
      "grad_norm": 2.3510637262715686,
      "learning_rate": 2e-05,
      "loss": 0.7971127510070801,
      "step": 3730
    },
    {
      "epoch": 0.2017695295640915,
      "grad_norm": 1.0401878649419207,
      "learning_rate": 2e-05,
      "loss": 0.8233554840087891,
      "step": 3740
    },
    {
      "epoch": 0.2023090202848511,
      "grad_norm": 1.6871599880351058,
      "learning_rate": 2e-05,
      "loss": 0.8121973037719726,
      "step": 3750
    },
    {
      "epoch": 0.20284851100561072,
      "grad_norm": 0.9834662903562297,
      "learning_rate": 2e-05,
      "loss": 0.799620532989502,
      "step": 3760
    },
    {
      "epoch": 0.2033880017263703,
      "grad_norm": 13.856106429547337,
      "learning_rate": 2e-05,
      "loss": 0.8100821495056152,
      "step": 3770
    },
    {
      "epoch": 0.2039274924471299,
      "grad_norm": 1.7971124760415595,
      "learning_rate": 2e-05,
      "loss": 0.822230339050293,
      "step": 3780
    },
    {
      "epoch": 0.2044669831678895,
      "grad_norm": 1.898826077210772,
      "learning_rate": 2e-05,
      "loss": 0.8280195236206055,
      "step": 3790
    },
    {
      "epoch": 0.20500647388864912,
      "grad_norm": 1.5690048084284152,
      "learning_rate": 2e-05,
      "loss": 0.7768063545227051,
      "step": 3800
    },
    {
      "epoch": 0.20554596460940872,
      "grad_norm": 1.2924271342168896,
      "learning_rate": 2e-05,
      "loss": 0.8139206886291503,
      "step": 3810
    },
    {
      "epoch": 0.2060854553301683,
      "grad_norm": 1.713906617227459,
      "learning_rate": 2e-05,
      "loss": 0.8099102020263672,
      "step": 3820
    },
    {
      "epoch": 0.20662494605092793,
      "grad_norm": 1.0000768599712642,
      "learning_rate": 2e-05,
      "loss": 0.8023160934448242,
      "step": 3830
    },
    {
      "epoch": 0.20716443677168753,
      "grad_norm": 1.8923446682002336,
      "learning_rate": 2e-05,
      "loss": 0.8316279411315918,
      "step": 3840
    },
    {
      "epoch": 0.20770392749244712,
      "grad_norm": 0.9410427812888499,
      "learning_rate": 2e-05,
      "loss": 0.8073904037475585,
      "step": 3850
    },
    {
      "epoch": 0.20824341821320674,
      "grad_norm": 1.8310165996432775,
      "learning_rate": 2e-05,
      "loss": 0.8441598892211915,
      "step": 3860
    },
    {
      "epoch": 0.20878290893396634,
      "grad_norm": 3.2539518984867444,
      "learning_rate": 2e-05,
      "loss": 0.8111666679382324,
      "step": 3870
    },
    {
      "epoch": 0.20932239965472593,
      "grad_norm": 1.0747584520027538,
      "learning_rate": 2e-05,
      "loss": 0.8289793968200684,
      "step": 3880
    },
    {
      "epoch": 0.20986189037548555,
      "grad_norm": 1.1928486588388492,
      "learning_rate": 2e-05,
      "loss": 0.7862348556518555,
      "step": 3890
    },
    {
      "epoch": 0.21040138109624515,
      "grad_norm": 1.7434197055178486,
      "learning_rate": 2e-05,
      "loss": 0.8227115631103515,
      "step": 3900
    },
    {
      "epoch": 0.21094087181700474,
      "grad_norm": 1.0267098850458918,
      "learning_rate": 2e-05,
      "loss": 0.8255733489990235,
      "step": 3910
    },
    {
      "epoch": 0.21148036253776434,
      "grad_norm": 1.2265343976727108,
      "learning_rate": 2e-05,
      "loss": 0.7926362514495849,
      "step": 3920
    },
    {
      "epoch": 0.21201985325852396,
      "grad_norm": 0.9373764058253546,
      "learning_rate": 2e-05,
      "loss": 0.7910985469818115,
      "step": 3930
    },
    {
      "epoch": 0.21255934397928355,
      "grad_norm": 1.194071164302341,
      "learning_rate": 2e-05,
      "loss": 0.8127195358276367,
      "step": 3940
    },
    {
      "epoch": 0.21309883470004315,
      "grad_norm": 1.260475572232143,
      "learning_rate": 2e-05,
      "loss": 0.8285701751708985,
      "step": 3950
    },
    {
      "epoch": 0.21363832542080277,
      "grad_norm": 0.9898466932738494,
      "learning_rate": 2e-05,
      "loss": 0.816091251373291,
      "step": 3960
    },
    {
      "epoch": 0.21417781614156237,
      "grad_norm": 1.0090063218881065,
      "learning_rate": 2e-05,
      "loss": 0.7883634090423584,
      "step": 3970
    },
    {
      "epoch": 0.21471730686232196,
      "grad_norm": 1.6131569146168403,
      "learning_rate": 2e-05,
      "loss": 0.8127913475036621,
      "step": 3980
    },
    {
      "epoch": 0.21525679758308158,
      "grad_norm": 1.1595573594793847,
      "learning_rate": 2e-05,
      "loss": 0.8038475036621093,
      "step": 3990
    },
    {
      "epoch": 0.21579628830384118,
      "grad_norm": 1.9569762958682257,
      "learning_rate": 2e-05,
      "loss": 0.8159636497497559,
      "step": 4000
    },
    {
      "epoch": 0.21633577902460077,
      "grad_norm": 2.335005979386267,
      "learning_rate": 2e-05,
      "loss": 0.8046688079833985,
      "step": 4010
    },
    {
      "epoch": 0.21687526974536037,
      "grad_norm": 1.0399747803957953,
      "learning_rate": 2e-05,
      "loss": 0.8062277793884277,
      "step": 4020
    },
    {
      "epoch": 0.21741476046612,
      "grad_norm": 2.0436577055413636,
      "learning_rate": 2e-05,
      "loss": 0.7829396724700928,
      "step": 4030
    },
    {
      "epoch": 0.21795425118687958,
      "grad_norm": 3.7324962194464963,
      "learning_rate": 2e-05,
      "loss": 0.779371976852417,
      "step": 4040
    },
    {
      "epoch": 0.21849374190763918,
      "grad_norm": 1.1382565851810948,
      "learning_rate": 2e-05,
      "loss": 0.7965741634368897,
      "step": 4050
    },
    {
      "epoch": 0.2190332326283988,
      "grad_norm": 1.8123028132941512,
      "learning_rate": 2e-05,
      "loss": 0.7883219718933105,
      "step": 4060
    },
    {
      "epoch": 0.2195727233491584,
      "grad_norm": 1.0506410293373154,
      "learning_rate": 2e-05,
      "loss": 0.797592306137085,
      "step": 4070
    },
    {
      "epoch": 0.220112214069918,
      "grad_norm": 1.4692142673014317,
      "learning_rate": 2e-05,
      "loss": 0.8219276428222656,
      "step": 4080
    },
    {
      "epoch": 0.2206517047906776,
      "grad_norm": 0.8906168312090097,
      "learning_rate": 2e-05,
      "loss": 0.8257383346557617,
      "step": 4090
    },
    {
      "epoch": 0.2211911955114372,
      "grad_norm": 1.2342718782532682,
      "learning_rate": 2e-05,
      "loss": 0.809934139251709,
      "step": 4100
    },
    {
      "epoch": 0.2217306862321968,
      "grad_norm": 0.9694554865448979,
      "learning_rate": 2e-05,
      "loss": 0.8013545989990234,
      "step": 4110
    },
    {
      "epoch": 0.22227017695295642,
      "grad_norm": 1.137884850761222,
      "learning_rate": 2e-05,
      "loss": 0.8090425491333008,
      "step": 4120
    },
    {
      "epoch": 0.22280966767371602,
      "grad_norm": 1.3482482834337026,
      "learning_rate": 2e-05,
      "loss": 0.8125991821289062,
      "step": 4130
    },
    {
      "epoch": 0.2233491583944756,
      "grad_norm": 1.2578930714985972,
      "learning_rate": 2e-05,
      "loss": 0.8139532089233399,
      "step": 4140
    },
    {
      "epoch": 0.2238886491152352,
      "grad_norm": 1.8012059953899737,
      "learning_rate": 2e-05,
      "loss": 0.8208517074584961,
      "step": 4150
    },
    {
      "epoch": 0.22442813983599483,
      "grad_norm": 1.008618597600832,
      "learning_rate": 2e-05,
      "loss": 0.8184005737304687,
      "step": 4160
    },
    {
      "epoch": 0.22496763055675442,
      "grad_norm": 1.2342516361348905,
      "learning_rate": 2e-05,
      "loss": 0.7810928344726562,
      "step": 4170
    },
    {
      "epoch": 0.22550712127751402,
      "grad_norm": 1.1183469001633872,
      "learning_rate": 2e-05,
      "loss": 0.7920833587646484,
      "step": 4180
    },
    {
      "epoch": 0.22604661199827364,
      "grad_norm": 2.278840527298114,
      "learning_rate": 2e-05,
      "loss": 0.8206521987915039,
      "step": 4190
    },
    {
      "epoch": 0.22658610271903323,
      "grad_norm": 2.4185004922027704,
      "learning_rate": 2e-05,
      "loss": 0.8077364921569824,
      "step": 4200
    },
    {
      "epoch": 0.22712559343979283,
      "grad_norm": 1.7614476008204176,
      "learning_rate": 2e-05,
      "loss": 0.7963101863861084,
      "step": 4210
    },
    {
      "epoch": 0.22766508416055245,
      "grad_norm": 1.3805130673932433,
      "learning_rate": 2e-05,
      "loss": 0.8123518943786621,
      "step": 4220
    },
    {
      "epoch": 0.22820457488131204,
      "grad_norm": 0.8757337643002813,
      "learning_rate": 2e-05,
      "loss": 0.817231559753418,
      "step": 4230
    },
    {
      "epoch": 0.22874406560207164,
      "grad_norm": 0.9189083746356497,
      "learning_rate": 2e-05,
      "loss": 0.812701416015625,
      "step": 4240
    },
    {
      "epoch": 0.22928355632283123,
      "grad_norm": 1.128056082076602,
      "learning_rate": 2e-05,
      "loss": 0.7835369110107422,
      "step": 4250
    },
    {
      "epoch": 0.22982304704359086,
      "grad_norm": 1.185230657692006,
      "learning_rate": 2e-05,
      "loss": 0.8072416305541992,
      "step": 4260
    },
    {
      "epoch": 0.23036253776435045,
      "grad_norm": 1.0656311997626688,
      "learning_rate": 2e-05,
      "loss": 0.7766363143920898,
      "step": 4270
    },
    {
      "epoch": 0.23090202848511004,
      "grad_norm": 2.673193204464597,
      "learning_rate": 2e-05,
      "loss": 0.8131553649902343,
      "step": 4280
    },
    {
      "epoch": 0.23144151920586967,
      "grad_norm": 1.9420639834564013,
      "learning_rate": 2e-05,
      "loss": 0.801763916015625,
      "step": 4290
    },
    {
      "epoch": 0.23198100992662926,
      "grad_norm": 2.19924220420487,
      "learning_rate": 2e-05,
      "loss": 0.7914301872253418,
      "step": 4300
    },
    {
      "epoch": 0.23252050064738886,
      "grad_norm": 1.0573366517424367,
      "learning_rate": 2e-05,
      "loss": 0.7900327682495117,
      "step": 4310
    },
    {
      "epoch": 0.23305999136814848,
      "grad_norm": 2.939736739352607,
      "learning_rate": 2e-05,
      "loss": 0.798931884765625,
      "step": 4320
    },
    {
      "epoch": 0.23359948208890807,
      "grad_norm": 2.458564521712793,
      "learning_rate": 2e-05,
      "loss": 0.8206014633178711,
      "step": 4330
    },
    {
      "epoch": 0.23413897280966767,
      "grad_norm": 5.885090977120574,
      "learning_rate": 2e-05,
      "loss": 0.8013025283813476,
      "step": 4340
    },
    {
      "epoch": 0.2346784635304273,
      "grad_norm": 1.2223506625327372,
      "learning_rate": 2e-05,
      "loss": 0.8057741165161133,
      "step": 4350
    },
    {
      "epoch": 0.23521795425118688,
      "grad_norm": 1.2671465051267325,
      "learning_rate": 2e-05,
      "loss": 0.8048929214477539,
      "step": 4360
    },
    {
      "epoch": 0.23575744497194648,
      "grad_norm": 20.36348071831498,
      "learning_rate": 2e-05,
      "loss": 0.7987167358398437,
      "step": 4370
    },
    {
      "epoch": 0.23629693569270607,
      "grad_norm": 1.8460057138384762,
      "learning_rate": 2e-05,
      "loss": 0.7847419261932373,
      "step": 4380
    },
    {
      "epoch": 0.2368364264134657,
      "grad_norm": 1.2294517053170533,
      "learning_rate": 2e-05,
      "loss": 0.8108327865600586,
      "step": 4390
    },
    {
      "epoch": 0.2373759171342253,
      "grad_norm": 1.816846078059956,
      "learning_rate": 2e-05,
      "loss": 0.822684383392334,
      "step": 4400
    },
    {
      "epoch": 0.23791540785498488,
      "grad_norm": 2.654537912239173,
      "learning_rate": 2e-05,
      "loss": 0.7995866298675537,
      "step": 4410
    },
    {
      "epoch": 0.2384548985757445,
      "grad_norm": 1.237212727457312,
      "learning_rate": 2e-05,
      "loss": 0.8010119438171387,
      "step": 4420
    },
    {
      "epoch": 0.2389943892965041,
      "grad_norm": 3.1325525009622006,
      "learning_rate": 2e-05,
      "loss": 0.8015875816345215,
      "step": 4430
    },
    {
      "epoch": 0.2395338800172637,
      "grad_norm": 1.523928142639394,
      "learning_rate": 2e-05,
      "loss": 0.8150830268859863,
      "step": 4440
    },
    {
      "epoch": 0.24007337073802332,
      "grad_norm": 1.2277911739000757,
      "learning_rate": 2e-05,
      "loss": 0.8062484741210938,
      "step": 4450
    },
    {
      "epoch": 0.2406128614587829,
      "grad_norm": 1.0545115565413794,
      "learning_rate": 2e-05,
      "loss": 0.7971771240234375,
      "step": 4460
    },
    {
      "epoch": 0.2411523521795425,
      "grad_norm": 1.2496456841536927,
      "learning_rate": 2e-05,
      "loss": 0.8115372657775879,
      "step": 4470
    },
    {
      "epoch": 0.24169184290030213,
      "grad_norm": 1.8854175836161469,
      "learning_rate": 2e-05,
      "loss": 0.7749072074890136,
      "step": 4480
    },
    {
      "epoch": 0.24223133362106172,
      "grad_norm": 0.9093539816139048,
      "learning_rate": 2e-05,
      "loss": 0.8023496627807617,
      "step": 4490
    },
    {
      "epoch": 0.24277082434182132,
      "grad_norm": 4.5883558811477165,
      "learning_rate": 2e-05,
      "loss": 0.7956307888031006,
      "step": 4500
    },
    {
      "epoch": 0.2433103150625809,
      "grad_norm": 2.317336751963953,
      "learning_rate": 2e-05,
      "loss": 0.8047412872314453,
      "step": 4510
    },
    {
      "epoch": 0.24384980578334053,
      "grad_norm": 1.007260229258514,
      "learning_rate": 2e-05,
      "loss": 0.8072935104370117,
      "step": 4520
    },
    {
      "epoch": 0.24438929650410013,
      "grad_norm": 0.9337093677097955,
      "learning_rate": 2e-05,
      "loss": 0.8341692924499512,
      "step": 4530
    },
    {
      "epoch": 0.24492878722485972,
      "grad_norm": 1.0849010132472683,
      "learning_rate": 2e-05,
      "loss": 0.8139280319213867,
      "step": 4540
    },
    {
      "epoch": 0.24546827794561935,
      "grad_norm": 3.8085843346374673,
      "learning_rate": 2e-05,
      "loss": 0.7983991622924804,
      "step": 4550
    },
    {
      "epoch": 0.24600776866637894,
      "grad_norm": 1.1069593900436312,
      "learning_rate": 2e-05,
      "loss": 0.8211133003234863,
      "step": 4560
    },
    {
      "epoch": 0.24654725938713853,
      "grad_norm": 1.7679431808317294,
      "learning_rate": 2e-05,
      "loss": 0.7952757835388183,
      "step": 4570
    },
    {
      "epoch": 0.24708675010789816,
      "grad_norm": 1.204211968460817,
      "learning_rate": 2e-05,
      "loss": 0.8002294540405274,
      "step": 4580
    },
    {
      "epoch": 0.24762624082865775,
      "grad_norm": 1.0739381921841171,
      "learning_rate": 2e-05,
      "loss": 0.8141761779785156,
      "step": 4590
    },
    {
      "epoch": 0.24816573154941735,
      "grad_norm": 1.0025779827603045,
      "learning_rate": 2e-05,
      "loss": 0.7826519012451172,
      "step": 4600
    },
    {
      "epoch": 0.24870522227017694,
      "grad_norm": 0.9466750748826841,
      "learning_rate": 2e-05,
      "loss": 0.8070455551147461,
      "step": 4610
    },
    {
      "epoch": 0.24924471299093656,
      "grad_norm": 1.2131970956423264,
      "learning_rate": 2e-05,
      "loss": 0.7810938835144043,
      "step": 4620
    },
    {
      "epoch": 0.24978420371169616,
      "grad_norm": 1.0981184283688172,
      "learning_rate": 2e-05,
      "loss": 0.7963405609130859,
      "step": 4630
    },
    {
      "epoch": 0.2503236944324558,
      "grad_norm": 1.0138800653866957,
      "learning_rate": 2e-05,
      "loss": 0.7861242294311523,
      "step": 4640
    },
    {
      "epoch": 0.2508631851532154,
      "grad_norm": 4.097239243333837,
      "learning_rate": 2e-05,
      "loss": 0.8028505325317383,
      "step": 4650
    },
    {
      "epoch": 0.25140267587397497,
      "grad_norm": 1.1310700561897664,
      "learning_rate": 2e-05,
      "loss": 0.8024848937988281,
      "step": 4660
    },
    {
      "epoch": 0.25194216659473456,
      "grad_norm": 3.2303094028183543,
      "learning_rate": 2e-05,
      "loss": 0.8002676010131836,
      "step": 4670
    },
    {
      "epoch": 0.25248165731549416,
      "grad_norm": 4.334179652818097,
      "learning_rate": 2e-05,
      "loss": 0.7879270553588867,
      "step": 4680
    },
    {
      "epoch": 0.25302114803625375,
      "grad_norm": 1.3799811734001441,
      "learning_rate": 2e-05,
      "loss": 0.7918018341064453,
      "step": 4690
    },
    {
      "epoch": 0.2535606387570134,
      "grad_norm": 1.0535386647114935,
      "learning_rate": 2e-05,
      "loss": 0.787726354598999,
      "step": 4700
    },
    {
      "epoch": 0.254100129477773,
      "grad_norm": 1.0780233294027906,
      "learning_rate": 2e-05,
      "loss": 0.8174970626831055,
      "step": 4710
    },
    {
      "epoch": 0.2546396201985326,
      "grad_norm": 1.9463043994121851,
      "learning_rate": 2e-05,
      "loss": 0.7929119110107422,
      "step": 4720
    },
    {
      "epoch": 0.2551791109192922,
      "grad_norm": 0.9918265526819622,
      "learning_rate": 2e-05,
      "loss": 0.7915786266326904,
      "step": 4730
    },
    {
      "epoch": 0.2557186016400518,
      "grad_norm": 2.1785632580927716,
      "learning_rate": 2e-05,
      "loss": 0.7979188442230225,
      "step": 4740
    },
    {
      "epoch": 0.2562580923608114,
      "grad_norm": 1.0273805631017123,
      "learning_rate": 2e-05,
      "loss": 0.7942420959472656,
      "step": 4750
    },
    {
      "epoch": 0.256797583081571,
      "grad_norm": 0.9176975985223724,
      "learning_rate": 2e-05,
      "loss": 0.7860095977783204,
      "step": 4760
    },
    {
      "epoch": 0.2573370738023306,
      "grad_norm": 2.132463793482237,
      "learning_rate": 2e-05,
      "loss": 0.7917288780212403,
      "step": 4770
    },
    {
      "epoch": 0.2578765645230902,
      "grad_norm": 1.0501050240952219,
      "learning_rate": 2e-05,
      "loss": 0.799459981918335,
      "step": 4780
    },
    {
      "epoch": 0.2584160552438498,
      "grad_norm": 0.8546120634684826,
      "learning_rate": 2e-05,
      "loss": 0.7682689189910888,
      "step": 4790
    },
    {
      "epoch": 0.2589555459646094,
      "grad_norm": 1.9845543830476717,
      "learning_rate": 2e-05,
      "loss": 0.7785700798034668,
      "step": 4800
    },
    {
      "epoch": 0.259495036685369,
      "grad_norm": 1.1674563454485274,
      "learning_rate": 2e-05,
      "loss": 0.7985703945159912,
      "step": 4810
    },
    {
      "epoch": 0.2600345274061286,
      "grad_norm": 0.8710562418288261,
      "learning_rate": 2e-05,
      "loss": 0.7821848869323731,
      "step": 4820
    },
    {
      "epoch": 0.26057401812688824,
      "grad_norm": 0.9756368089831883,
      "learning_rate": 2e-05,
      "loss": 0.767934513092041,
      "step": 4830
    },
    {
      "epoch": 0.26111350884764783,
      "grad_norm": 1.1180921433718192,
      "learning_rate": 2e-05,
      "loss": 0.8063995361328125,
      "step": 4840
    },
    {
      "epoch": 0.26165299956840743,
      "grad_norm": 1.2530512910301759,
      "learning_rate": 2e-05,
      "loss": 0.7923648834228516,
      "step": 4850
    },
    {
      "epoch": 0.262192490289167,
      "grad_norm": 1.090169882565525,
      "learning_rate": 2e-05,
      "loss": 0.8143489837646485,
      "step": 4860
    },
    {
      "epoch": 0.2627319810099266,
      "grad_norm": 1.2374972544253948,
      "learning_rate": 2e-05,
      "loss": 0.7990952968597412,
      "step": 4870
    },
    {
      "epoch": 0.2632714717306862,
      "grad_norm": 2.0802207218753748,
      "learning_rate": 2e-05,
      "loss": 0.7822625160217285,
      "step": 4880
    },
    {
      "epoch": 0.26381096245144586,
      "grad_norm": 1.328298916749292,
      "learning_rate": 2e-05,
      "loss": 0.7958024501800537,
      "step": 4890
    },
    {
      "epoch": 0.26435045317220546,
      "grad_norm": 0.9139641954979899,
      "learning_rate": 2e-05,
      "loss": 0.7784433364868164,
      "step": 4900
    },
    {
      "epoch": 0.26488994389296505,
      "grad_norm": 0.9275711811200684,
      "learning_rate": 2e-05,
      "loss": 0.7815862655639648,
      "step": 4910
    },
    {
      "epoch": 0.26542943461372465,
      "grad_norm": 0.9898072142266595,
      "learning_rate": 2e-05,
      "loss": 0.7841447353363037,
      "step": 4920
    },
    {
      "epoch": 0.26596892533448424,
      "grad_norm": 1.1994309553695597,
      "learning_rate": 2e-05,
      "loss": 0.8109785079956054,
      "step": 4930
    },
    {
      "epoch": 0.26650841605524384,
      "grad_norm": 2.8794105154816596,
      "learning_rate": 2e-05,
      "loss": 0.7862544536590577,
      "step": 4940
    },
    {
      "epoch": 0.26704790677600343,
      "grad_norm": 1.1975772787075112,
      "learning_rate": 2e-05,
      "loss": 0.7912877559661865,
      "step": 4950
    },
    {
      "epoch": 0.2675873974967631,
      "grad_norm": 2.7457990313458382,
      "learning_rate": 2e-05,
      "loss": 0.804524040222168,
      "step": 4960
    },
    {
      "epoch": 0.2681268882175227,
      "grad_norm": 2.1126991984169754,
      "learning_rate": 2e-05,
      "loss": 0.8053647994995117,
      "step": 4970
    },
    {
      "epoch": 0.26866637893828227,
      "grad_norm": 1.0177170910944318,
      "learning_rate": 2e-05,
      "loss": 0.7716758728027344,
      "step": 4980
    },
    {
      "epoch": 0.26920586965904186,
      "grad_norm": 0.8759620830564809,
      "learning_rate": 2e-05,
      "loss": 0.798637580871582,
      "step": 4990
    },
    {
      "epoch": 0.26974536037980146,
      "grad_norm": 1.0309218660430153,
      "learning_rate": 2e-05,
      "loss": 0.7905425071716309,
      "step": 5000
    },
    {
      "epoch": 0.27028485110056105,
      "grad_norm": 1.1400869894163015,
      "learning_rate": 2e-05,
      "loss": 0.7887660980224609,
      "step": 5010
    },
    {
      "epoch": 0.27082434182132065,
      "grad_norm": 0.8201715945801922,
      "learning_rate": 2e-05,
      "loss": 0.7985705375671387,
      "step": 5020
    },
    {
      "epoch": 0.2713638325420803,
      "grad_norm": 0.8359412161026315,
      "learning_rate": 2e-05,
      "loss": 0.8124397277832032,
      "step": 5030
    },
    {
      "epoch": 0.2719033232628399,
      "grad_norm": 2.9571467114845515,
      "learning_rate": 2e-05,
      "loss": 0.804383373260498,
      "step": 5040
    },
    {
      "epoch": 0.2724428139835995,
      "grad_norm": 1.1158117745551008,
      "learning_rate": 2e-05,
      "loss": 0.7748370170593262,
      "step": 5050
    },
    {
      "epoch": 0.2729823047043591,
      "grad_norm": 3.3092129214257593,
      "learning_rate": 2e-05,
      "loss": 0.777961254119873,
      "step": 5060
    },
    {
      "epoch": 0.2735217954251187,
      "grad_norm": 1.014611756530086,
      "learning_rate": 2e-05,
      "loss": 0.8041595458984375,
      "step": 5070
    },
    {
      "epoch": 0.27406128614587827,
      "grad_norm": 1.6500649364592126,
      "learning_rate": 2e-05,
      "loss": 0.7854903697967529,
      "step": 5080
    },
    {
      "epoch": 0.2746007768666379,
      "grad_norm": 1.1824947563310242,
      "learning_rate": 2e-05,
      "loss": 0.7920733451843261,
      "step": 5090
    },
    {
      "epoch": 0.2751402675873975,
      "grad_norm": 1.0187542343320222,
      "learning_rate": 2e-05,
      "loss": 0.7976733207702636,
      "step": 5100
    },
    {
      "epoch": 0.2756797583081571,
      "grad_norm": 0.9470765385988701,
      "learning_rate": 2e-05,
      "loss": 0.8028879165649414,
      "step": 5110
    },
    {
      "epoch": 0.2762192490289167,
      "grad_norm": 0.9457213110850333,
      "learning_rate": 2e-05,
      "loss": 0.8163872718811035,
      "step": 5120
    },
    {
      "epoch": 0.2767587397496763,
      "grad_norm": 1.6731491811765,
      "learning_rate": 2e-05,
      "loss": 0.813693618774414,
      "step": 5130
    },
    {
      "epoch": 0.2772982304704359,
      "grad_norm": 0.8984113618744548,
      "learning_rate": 2e-05,
      "loss": 0.7750575542449951,
      "step": 5140
    },
    {
      "epoch": 0.2778377211911955,
      "grad_norm": 0.926808106221653,
      "learning_rate": 2e-05,
      "loss": 0.820768165588379,
      "step": 5150
    },
    {
      "epoch": 0.27837721191195514,
      "grad_norm": 1.5063721861099815,
      "learning_rate": 2e-05,
      "loss": 0.7994228363037109,
      "step": 5160
    },
    {
      "epoch": 0.27891670263271473,
      "grad_norm": 0.9078508768384885,
      "learning_rate": 2e-05,
      "loss": 0.8026659965515137,
      "step": 5170
    },
    {
      "epoch": 0.2794561933534743,
      "grad_norm": 1.544450567545405,
      "learning_rate": 2e-05,
      "loss": 0.7981088161468506,
      "step": 5180
    },
    {
      "epoch": 0.2799956840742339,
      "grad_norm": 0.8883697099363166,
      "learning_rate": 2e-05,
      "loss": 0.7581671714782715,
      "step": 5190
    },
    {
      "epoch": 0.2805351747949935,
      "grad_norm": 1.0563585088261542,
      "learning_rate": 2e-05,
      "loss": 0.7537007331848145,
      "step": 5200
    },
    {
      "epoch": 0.2810746655157531,
      "grad_norm": 53.920692592795916,
      "learning_rate": 2e-05,
      "loss": 0.8015555381774903,
      "step": 5210
    },
    {
      "epoch": 0.28161415623651276,
      "grad_norm": 1.2004807362960381,
      "learning_rate": 2e-05,
      "loss": 0.804405403137207,
      "step": 5220
    },
    {
      "epoch": 0.28215364695727235,
      "grad_norm": 0.8131496439861595,
      "learning_rate": 2e-05,
      "loss": 0.7767237663269043,
      "step": 5230
    },
    {
      "epoch": 0.28269313767803195,
      "grad_norm": 0.9455776486909698,
      "learning_rate": 2e-05,
      "loss": 0.7976688385009766,
      "step": 5240
    },
    {
      "epoch": 0.28323262839879154,
      "grad_norm": 0.8159205542775527,
      "learning_rate": 2e-05,
      "loss": 0.7810776710510254,
      "step": 5250
    },
    {
      "epoch": 0.28377211911955114,
      "grad_norm": 0.8035968179285776,
      "learning_rate": 2e-05,
      "loss": 0.7829414367675781,
      "step": 5260
    },
    {
      "epoch": 0.28431160984031073,
      "grad_norm": 0.9759491725844237,
      "learning_rate": 2e-05,
      "loss": 0.8078337669372558,
      "step": 5270
    },
    {
      "epoch": 0.2848511005610703,
      "grad_norm": 0.9226690067778774,
      "learning_rate": 2e-05,
      "loss": 0.80551176071167,
      "step": 5280
    },
    {
      "epoch": 0.28539059128183,
      "grad_norm": 1.8505416279508915,
      "learning_rate": 2e-05,
      "loss": 0.8313540458679199,
      "step": 5290
    },
    {
      "epoch": 0.28593008200258957,
      "grad_norm": 0.8739876378991774,
      "learning_rate": 2e-05,
      "loss": 0.7921354293823242,
      "step": 5300
    },
    {
      "epoch": 0.28646957272334916,
      "grad_norm": 3.2051380925373962,
      "learning_rate": 2e-05,
      "loss": 0.7898638725280762,
      "step": 5310
    },
    {
      "epoch": 0.28700906344410876,
      "grad_norm": 1.3993463213470974,
      "learning_rate": 2e-05,
      "loss": 0.7983866691589355,
      "step": 5320
    },
    {
      "epoch": 0.28754855416486835,
      "grad_norm": 0.9041613682694449,
      "learning_rate": 2e-05,
      "loss": 0.7882991790771484,
      "step": 5330
    },
    {
      "epoch": 0.28808804488562795,
      "grad_norm": 3.38263456052886,
      "learning_rate": 2e-05,
      "loss": 0.7734818935394288,
      "step": 5340
    },
    {
      "epoch": 0.2886275356063876,
      "grad_norm": 0.868973658809523,
      "learning_rate": 1.999194520388485e-05,
      "loss": 0.8212265014648438,
      "step": 5350
    },
    {
      "epoch": 0.2891670263271472,
      "grad_norm": 1.4576397236374967,
      "learning_rate": 1.9959247303157763e-05,
      "loss": 0.7948694229125977,
      "step": 5360
    },
    {
      "epoch": 0.2897065170479068,
      "grad_norm": 1.0866801265234474,
      "learning_rate": 1.990149424459817e-05,
      "loss": 0.777640438079834,
      "step": 5370
    },
    {
      "epoch": 0.2902460077686664,
      "grad_norm": 1.28125266327225,
      "learning_rate": 1.9818847538648688e-05,
      "loss": 0.8106527328491211,
      "step": 5380
    },
    {
      "epoch": 0.290785498489426,
      "grad_norm": 1.0814153494413632,
      "learning_rate": 1.971153831256721e-05,
      "loss": 0.7647212982177735,
      "step": 5390
    },
    {
      "epoch": 0.29132498921018557,
      "grad_norm": 1.1013440312592218,
      "learning_rate": 1.9579866664063435e-05,
      "loss": 0.7767222404479981,
      "step": 5400
    },
    {
      "epoch": 0.29186447993094516,
      "grad_norm": 0.9477812584232342,
      "learning_rate": 1.9424200822054782e-05,
      "loss": 0.7967702865600585,
      "step": 5410
    },
    {
      "epoch": 0.2924039706517048,
      "grad_norm": 1.1185239278834358,
      "learning_rate": 1.9244976116888626e-05,
      "loss": 0.7856125831604004,
      "step": 5420
    },
    {
      "epoch": 0.2929434613724644,
      "grad_norm": 0.9699922870194968,
      "learning_rate": 1.904269376291071e-05,
      "loss": 0.7780052185058594,
      "step": 5430
    },
    {
      "epoch": 0.293482952093224,
      "grad_norm": 1.4925604100521892,
      "learning_rate": 1.88179194567844e-05,
      "loss": 0.7879739284515381,
      "step": 5440
    },
    {
      "epoch": 0.2940224428139836,
      "grad_norm": 0.9585906385064894,
      "learning_rate": 1.8571281795480632e-05,
      "loss": 0.8060611724853516,
      "step": 5450
    },
    {
      "epoch": 0.2945619335347432,
      "grad_norm": 0.8310186563189833,
      "learning_rate": 1.8303470518362808e-05,
      "loss": 0.820162582397461,
      "step": 5460
    },
    {
      "epoch": 0.2951014242555028,
      "grad_norm": 0.8771935822537728,
      "learning_rate": 1.801523457828271e-05,
      "loss": 0.779834508895874,
      "step": 5470
    },
    {
      "epoch": 0.2956409149762624,
      "grad_norm": 1.6167090546670861,
      "learning_rate": 1.7707380047081796e-05,
      "loss": 0.8007966995239257,
      "step": 5480
    },
    {
      "epoch": 0.29618040569702203,
      "grad_norm": 0.9474641169650851,
      "learning_rate": 1.7380767861355252e-05,
      "loss": 0.7679849147796631,
      "step": 5490
    },
    {
      "epoch": 0.2967198964177816,
      "grad_norm": 1.0453751801055784,
      "learning_rate": 1.703631141478299e-05,
      "loss": 0.8166125297546387,
      "step": 5500
    },
    {
      "epoch": 0.2972593871385412,
      "grad_norm": 3.0491607158490064,
      "learning_rate": 1.6674974003760705e-05,
      "loss": 0.7915538787841797,
      "step": 5510
    },
    {
      "epoch": 0.2977988778593008,
      "grad_norm": 1.0629982062606982,
      "learning_rate": 1.6297766133474556e-05,
      "loss": 0.7773507595062256,
      "step": 5520
    },
    {
      "epoch": 0.2983383685800604,
      "grad_norm": 0.8585373329621083,
      "learning_rate": 1.5905742691953186e-05,
      "loss": 0.8239789962768554,
      "step": 5530
    },
    {
      "epoch": 0.29887785930082,
      "grad_norm": 0.9771737100883071,
      "learning_rate": 1.55e-05,
      "loss": 0.7988348007202148,
      "step": 5540
    },
    {
      "epoch": 0.29941735002157965,
      "grad_norm": 0.884594842107015,
      "learning_rate": 1.5081672745255862e-05,
      "loss": 0.7885036468505859,
      "step": 5550
    },
    {
      "epoch": 0.29995684074233925,
      "grad_norm": 0.7850214225230454,
      "learning_rate": 1.4651930808966206e-05,
      "loss": 0.7711037635803223,
      "step": 5560
    },
    {
      "epoch": 0.30049633146309884,
      "grad_norm": 0.8545082227628484,
      "learning_rate": 1.4211975994326848e-05,
      "loss": 0.7680245876312256,
      "step": 5570
    },
    {
      "epoch": 0.30103582218385844,
      "grad_norm": 0.8046795828654701,
      "learning_rate": 1.376303866555779e-05,
      "loss": 0.7869452476501465,
      "step": 5580
    },
    {
      "epoch": 0.30157531290461803,
      "grad_norm": 0.7956415381512352,
      "learning_rate": 1.3306374307104164e-05,
      "loss": 0.7833169937133789,
      "step": 5590
    },
    {
      "epoch": 0.3021148036253776,
      "grad_norm": 0.9971718562173801,
      "learning_rate": 1.2843260012586718e-05,
      "loss": 0.7874101161956787,
      "step": 5600
    },
    {
      "epoch": 0.3026542943461372,
      "grad_norm": 0.8029750472782808,
      "learning_rate": 1.2374990913320726e-05,
      "loss": 0.7451698303222656,
      "step": 5610
    },
    {
      "epoch": 0.30319378506689687,
      "grad_norm": 1.2688642978402311,
      "learning_rate": 1.1902876556391247e-05,
      "loss": 0.7717950344085693,
      "step": 5620
    },
    {
      "epoch": 0.30373327578765646,
      "grad_norm": 1.222547055625709,
      "learning_rate": 1.1428237242413685e-05,
      "loss": 0.755951976776123,
      "step": 5630
    },
    {
      "epoch": 0.30427276650841606,
      "grad_norm": 7.479813239527541,
      "learning_rate": 1.095240033322132e-05,
      "loss": 0.7752552986145019,
      "step": 5640
    },
    {
      "epoch": 0.30481225722917565,
      "grad_norm": 0.8737490835807221,
      "learning_rate": 1.047669653980572e-05,
      "loss": 0.7807745933532715,
      "step": 5650
    },
    {
      "epoch": 0.30535174794993525,
      "grad_norm": 1.0740558183532354,
      "learning_rate": 1.0002456200890902e-05,
      "loss": 0.7601753234863281,
      "step": 5660
    },
    {
      "epoch": 0.30589123867069484,
      "grad_norm": 0.8446106358110976,
      "learning_rate": 9.531005562548668e-06,
      "loss": 0.7519152641296387,
      "step": 5670
    },
    {
      "epoch": 0.3064307293914545,
      "grad_norm": 0.8519918203614549,
      "learning_rate": 9.06366306925925e-06,
      "loss": 0.7516709804534912,
      "step": 5680
    },
    {
      "epoch": 0.3069702201122141,
      "grad_norm": 0.9555561937296932,
      "learning_rate": 8.601735676789686e-06,
      "loss": 0.7596233367919922,
      "step": 5690
    },
    {
      "epoch": 0.3075097108329737,
      "grad_norm": 0.8865126586294536,
      "learning_rate": 8.14651519720118e-06,
      "loss": 0.7628667831420899,
      "step": 5700
    },
    {
      "epoch": 0.3080492015537333,
      "grad_norm": 0.8726084354145968,
      "learning_rate": 7.699274686206866e-06,
      "loss": 0.7802728176116943,
      "step": 5710
    },
    {
      "epoch": 0.30858869227449287,
      "grad_norm": 0.8405076929336869,
      "learning_rate": 7.261264882983024e-06,
      "loss": 0.8020237922668457,
      "step": 5720
    },
    {
      "epoch": 0.30912818299525247,
      "grad_norm": 0.7868919851826516,
      "learning_rate": 6.8337107123900386e-06,
      "loss": 0.7673155784606933,
      "step": 5730
    },
    {
      "epoch": 0.30966767371601206,
      "grad_norm": 0.8746962243569558,
      "learning_rate": 6.417807859384881e-06,
      "loss": 0.7523904800415039,
      "step": 5740
    },
    {
      "epoch": 0.3102071644367717,
      "grad_norm": 1.1364878631345965,
      "learning_rate": 6.014719425205009e-06,
      "loss": 0.7582972526550293,
      "step": 5750
    },
    {
      "epoch": 0.3107466551575313,
      "grad_norm": 1.1170859821908525,
      "learning_rate": 5.6255726746749275e-06,
      "loss": 0.7664484024047852,
      "step": 5760
    },
    {
      "epoch": 0.3112861458782909,
      "grad_norm": 1.0000619161748914,
      "learning_rate": 5.251455883731728e-06,
      "loss": 0.7634631156921386,
      "step": 5770
    },
    {
      "epoch": 0.3118256365990505,
      "grad_norm": 0.8398693090076838,
      "learning_rate": 4.893415295985812e-06,
      "loss": 0.7638420104980469,
      "step": 5780
    },
    {
      "epoch": 0.3123651273198101,
      "grad_norm": 0.9222432680965927,
      "learning_rate": 4.552452196827885e-06,
      "loss": 0.7642083168029785,
      "step": 5790
    },
    {
      "epoch": 0.3129046180405697,
      "grad_norm": 1.0660091332296988,
      "learning_rate": 4.229520113264785e-06,
      "loss": 0.7636802196502686,
      "step": 5800
    },
    {
      "epoch": 0.31344410876132933,
      "grad_norm": 1.2020556163096852,
      "learning_rate": 3.925522147314915e-06,
      "loss": 0.7888552665710449,
      "step": 5810
    },
    {
      "epoch": 0.3139835994820889,
      "grad_norm": 0.9789808192361225,
      "learning_rate": 3.641308450420675e-06,
      "loss": 0.7717351913452148,
      "step": 5820
    },
    {
      "epoch": 0.3145230902028485,
      "grad_norm": 0.7983800089757431,
      "learning_rate": 3.37767384594087e-06,
      "loss": 0.7382369995117187,
      "step": 5830
    },
    {
      "epoch": 0.3150625809236081,
      "grad_norm": 1.1576469479947942,
      "learning_rate": 3.1353556063719357e-06,
      "loss": 0.7756743431091309,
      "step": 5840
    },
    {
      "epoch": 0.3156020716443677,
      "grad_norm": 0.9441431432702874,
      "learning_rate": 2.9150313915141474e-06,
      "loss": 0.7549386024475098,
      "step": 5850
    },
    {
      "epoch": 0.3161415623651273,
      "grad_norm": 0.7986921787886277,
      "learning_rate": 2.717317353348891e-06,
      "loss": 0.7606843471527099,
      "step": 5860
    },
    {
      "epoch": 0.3166810530858869,
      "grad_norm": 1.6239976322206668,
      "learning_rate": 2.5427664129268253e-06,
      "loss": 0.7230315208435059,
      "step": 5870
    },
    {
      "epoch": 0.31722054380664655,
      "grad_norm": 1.6482415089362032,
      "learning_rate": 2.391866714085717e-06,
      "loss": 0.7719806671142578,
      "step": 5880
    },
    {
      "epoch": 0.31776003452740614,
      "grad_norm": 0.8608428183789886,
      "learning_rate": 2.2650402583222753e-06,
      "loss": 0.752475357055664,
      "step": 5890
    },
    {
      "epoch": 0.31829952524816574,
      "grad_norm": 0.8012060739196416,
      "learning_rate": 2.1626417246356398e-06,
      "loss": 0.7699872016906738,
      "step": 5900
    },
    {
      "epoch": 0.31883901596892533,
      "grad_norm": 0.9622288112833073,
      "learning_rate": 2.084957477642894e-06,
      "loss": 0.7457541942596435,
      "step": 5910
    },
    {
      "epoch": 0.3193785066896849,
      "grad_norm": 1.2736502772909783,
      "learning_rate": 2.0322047667405147e-06,
      "loss": 0.757701301574707,
      "step": 5920
    },
    {
      "epoch": 0.3199179974104445,
      "grad_norm": 0.8091212730887813,
      "learning_rate": 2.0045311185513342e-06,
      "loss": 0.7599719047546387,
      "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
}