{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7527904616945713,
  "eval_steps": 500,
  "global_step": 5935,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012683916793505834,
      "grad_norm": 136.43014181701625,
      "learning_rate": 2.910112359550562e-06,
      "loss": 1.6949037551879882,
      "step": 10
    },
    {
      "epoch": 0.0025367833587011668,
      "grad_norm": 32.859926338602314,
      "learning_rate": 3.921348314606742e-06,
      "loss": 0.9884631156921386,
      "step": 20
    },
    {
      "epoch": 0.00380517503805175,
      "grad_norm": 508.0665769452469,
      "learning_rate": 4.932584269662922e-06,
      "loss": 0.8763895034790039,
      "step": 30
    },
    {
      "epoch": 0.0050735667174023336,
      "grad_norm": 94.60483653595367,
      "learning_rate": 5.943820224719102e-06,
      "loss": 0.7969870567321777,
      "step": 40
    },
    {
      "epoch": 0.006341958396752917,
      "grad_norm": 34.98298797720385,
      "learning_rate": 6.955056179775282e-06,
      "loss": 0.8274876594543457,
      "step": 50
    },
    {
      "epoch": 0.0076103500761035,
      "grad_norm": 47.70887229387195,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.8065635681152343,
      "step": 60
    },
    {
      "epoch": 0.008878741755454084,
      "grad_norm": 24.673891661894817,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.7971744537353516,
      "step": 70
    },
    {
      "epoch": 0.010147133434804667,
      "grad_norm": 11.833491723086226,
      "learning_rate": 9.988764044943822e-06,
      "loss": 0.7927028179168701,
      "step": 80
    },
    {
      "epoch": 0.01141552511415525,
      "grad_norm": 15.898512079979175,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.8186724662780762,
      "step": 90
    },
    {
      "epoch": 0.012683916793505834,
      "grad_norm": 34.741015249504315,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.8292877197265625,
      "step": 100
    },
    {
      "epoch": 0.013952308472856417,
      "grad_norm": 14.238447556079114,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.8440424919128418,
      "step": 110
    },
    {
      "epoch": 0.015220700152207,
      "grad_norm": 13.406711036641335,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.7999032497406006,
      "step": 120
    },
    {
      "epoch": 0.016489091831557584,
      "grad_norm": 35.564834858013306,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.812278938293457,
      "step": 130
    },
    {
      "epoch": 0.017757483510908167,
      "grad_norm": 56.27106274800488,
      "learning_rate": 1.60561797752809e-05,
      "loss": 1.1253692626953125,
      "step": 140
    },
    {
      "epoch": 0.01902587519025875,
      "grad_norm": 13.35128554184124,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.8475324630737304,
      "step": 150
    },
    {
      "epoch": 0.020294266869609334,
      "grad_norm": 48.852659065581165,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.8323675155639648,
      "step": 160
    },
    {
      "epoch": 0.021562658548959918,
      "grad_norm": 37.973058906788545,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.813169002532959,
      "step": 170
    },
    {
      "epoch": 0.0228310502283105,
      "grad_norm": 6.2740268429253945,
      "learning_rate": 2e-05,
      "loss": 0.7670700073242187,
      "step": 180
    },
    {
      "epoch": 0.024099441907661084,
      "grad_norm": 56.74067924769685,
      "learning_rate": 2e-05,
      "loss": 0.7810227870941162,
      "step": 190
    },
    {
      "epoch": 0.025367833587011668,
      "grad_norm": 6.165593180677104,
      "learning_rate": 2e-05,
      "loss": 0.7513003826141358,
      "step": 200
    },
    {
      "epoch": 0.02663622526636225,
      "grad_norm": 11.754348794858206,
      "learning_rate": 2e-05,
      "loss": 0.7863423824310303,
      "step": 210
    },
    {
      "epoch": 0.027904616945712835,
      "grad_norm": 4.717858822403969,
      "learning_rate": 2e-05,
      "loss": 0.7572269439697266,
      "step": 220
    },
    {
      "epoch": 0.029173008625063418,
      "grad_norm": 4.802552009597004,
      "learning_rate": 2e-05,
      "loss": 0.7903682231903076,
      "step": 230
    },
    {
      "epoch": 0.030441400304414,
      "grad_norm": 4.397246302198522,
      "learning_rate": 2e-05,
      "loss": 0.7518584251403808,
      "step": 240
    },
    {
      "epoch": 0.031709791983764585,
      "grad_norm": 22.17509345885548,
      "learning_rate": 2e-05,
      "loss": 0.8089160919189453,
      "step": 250
    },
    {
      "epoch": 0.03297818366311517,
      "grad_norm": 3.3116184268025286,
      "learning_rate": 2e-05,
      "loss": 0.8135837554931641,
      "step": 260
    },
    {
      "epoch": 0.03424657534246575,
      "grad_norm": 52.70496260695796,
      "learning_rate": 2e-05,
      "loss": 0.7688494682312011,
      "step": 270
    },
    {
      "epoch": 0.035514967021816335,
      "grad_norm": 9.399212137798266,
      "learning_rate": 2e-05,
      "loss": 0.7652685165405273,
      "step": 280
    },
    {
      "epoch": 0.03678335870116692,
      "grad_norm": 7.991715889148137,
      "learning_rate": 2e-05,
      "loss": 0.8129077911376953,
      "step": 290
    },
    {
      "epoch": 0.0380517503805175,
      "grad_norm": 8.01583439481696,
      "learning_rate": 2e-05,
      "loss": 0.805167293548584,
      "step": 300
    },
    {
      "epoch": 0.039320142059868085,
      "grad_norm": 22.17306932017089,
      "learning_rate": 2e-05,
      "loss": 0.7667516708374024,
      "step": 310
    },
    {
      "epoch": 0.04058853373921867,
      "grad_norm": 23.892861194320176,
      "learning_rate": 2e-05,
      "loss": 0.7735576152801513,
      "step": 320
    },
    {
      "epoch": 0.04185692541856925,
      "grad_norm": 15.075078732304078,
      "learning_rate": 2e-05,
      "loss": 0.8448783874511718,
      "step": 330
    },
    {
      "epoch": 0.043125317097919835,
      "grad_norm": 10.77618719509466,
      "learning_rate": 2e-05,
      "loss": 0.8245027542114258,
      "step": 340
    },
    {
      "epoch": 0.04439370877727042,
      "grad_norm": 8.020802979289998,
      "learning_rate": 2e-05,
      "loss": 0.8667722702026367,
      "step": 350
    },
    {
      "epoch": 0.045662100456621,
      "grad_norm": 47.2224243902643,
      "learning_rate": 2e-05,
      "loss": 0.8066923141479492,
      "step": 360
    },
    {
      "epoch": 0.046930492135971585,
      "grad_norm": 10.288491190749616,
      "learning_rate": 2e-05,
      "loss": 0.8301499366760254,
      "step": 370
    },
    {
      "epoch": 0.04819888381532217,
      "grad_norm": 56.86663226171858,
      "learning_rate": 2e-05,
      "loss": 0.7861690044403076,
      "step": 380
    },
    {
      "epoch": 0.04946727549467275,
      "grad_norm": 18.160386146267793,
      "learning_rate": 2e-05,
      "loss": 0.7772885322570801,
      "step": 390
    },
    {
      "epoch": 0.050735667174023336,
      "grad_norm": 6.241693375860191,
      "learning_rate": 2e-05,
      "loss": 0.8417564392089844,
      "step": 400
    },
    {
      "epoch": 0.05200405885337392,
      "grad_norm": 4.64428003403802,
      "learning_rate": 2e-05,
      "loss": 0.8099604606628418,
      "step": 410
    },
    {
      "epoch": 0.0532724505327245,
      "grad_norm": 3.2136193770299086,
      "learning_rate": 2e-05,
      "loss": 0.777653980255127,
      "step": 420
    },
    {
      "epoch": 0.054540842212075086,
      "grad_norm": 6.141028366995221,
      "learning_rate": 2e-05,
      "loss": 0.7952105522155761,
      "step": 430
    },
    {
      "epoch": 0.05580923389142567,
      "grad_norm": 3.8503648233948393,
      "learning_rate": 2e-05,
      "loss": 0.7994479656219482,
      "step": 440
    },
    {
      "epoch": 0.05707762557077625,
      "grad_norm": 44.09309047364388,
      "learning_rate": 2e-05,
      "loss": 0.7930583477020263,
      "step": 450
    },
    {
      "epoch": 0.058346017250126836,
      "grad_norm": 4.398752803135007,
      "learning_rate": 2e-05,
      "loss": 0.7821313858032226,
      "step": 460
    },
    {
      "epoch": 0.05961440892947742,
      "grad_norm": 24.213080461855856,
      "learning_rate": 2e-05,
      "loss": 0.8682317733764648,
      "step": 470
    },
    {
      "epoch": 0.060882800608828,
      "grad_norm": 5.772589336535053,
      "learning_rate": 2e-05,
      "loss": 0.772674560546875,
      "step": 480
    },
    {
      "epoch": 0.062151192288178586,
      "grad_norm": 13.468763231926662,
      "learning_rate": 2e-05,
      "loss": 0.7964553356170654,
      "step": 490
    },
    {
      "epoch": 0.06341958396752917,
      "grad_norm": 28.844825657288197,
      "learning_rate": 2e-05,
      "loss": 0.7964128494262696,
      "step": 500
    },
    {
      "epoch": 0.06468797564687975,
      "grad_norm": 5.410307335791166,
      "learning_rate": 2e-05,
      "loss": 0.7609683990478515,
      "step": 510
    },
    {
      "epoch": 0.06595636732623034,
      "grad_norm": 14.013807121502579,
      "learning_rate": 2e-05,
      "loss": 0.7820713996887207,
      "step": 520
    },
    {
      "epoch": 0.06722475900558092,
      "grad_norm": 3.6452579726563163,
      "learning_rate": 2e-05,
      "loss": 0.7552390098571777,
      "step": 530
    },
    {
      "epoch": 0.0684931506849315,
      "grad_norm": 8.037156418454089,
      "learning_rate": 2e-05,
      "loss": 0.8040719985961914,
      "step": 540
    },
    {
      "epoch": 0.06976154236428209,
      "grad_norm": 7.030243426782184,
      "learning_rate": 2e-05,
      "loss": 0.9523216247558594,
      "step": 550
    },
    {
      "epoch": 0.07102993404363267,
      "grad_norm": 3.93355288552293,
      "learning_rate": 2e-05,
      "loss": 0.8417517662048339,
      "step": 560
    },
    {
      "epoch": 0.07229832572298325,
      "grad_norm": 5.1801882052511505,
      "learning_rate": 2e-05,
      "loss": 0.7942523956298828,
      "step": 570
    },
    {
      "epoch": 0.07356671740233384,
      "grad_norm": 6.255003571586162,
      "learning_rate": 2e-05,
      "loss": 0.8063060760498046,
      "step": 580
    },
    {
      "epoch": 0.07483510908168442,
      "grad_norm": 17.288766083597647,
      "learning_rate": 2e-05,
      "loss": 0.7871021270751953,
      "step": 590
    },
    {
      "epoch": 0.076103500761035,
      "grad_norm": 2.574066301419686,
      "learning_rate": 2e-05,
      "loss": 0.7339196681976319,
      "step": 600
    },
    {
      "epoch": 0.07737189244038559,
      "grad_norm": 3.4850913902929954,
      "learning_rate": 2e-05,
      "loss": 0.7536786556243896,
      "step": 610
    },
    {
      "epoch": 0.07864028411973617,
      "grad_norm": 7.6742829245515685,
      "learning_rate": 2e-05,
      "loss": 0.7633294582366943,
      "step": 620
    },
    {
      "epoch": 0.07990867579908675,
      "grad_norm": 5.3653813857439285,
      "learning_rate": 2e-05,
      "loss": 0.7759026527404785,
      "step": 630
    },
    {
      "epoch": 0.08117706747843734,
      "grad_norm": 21.430282724935108,
      "learning_rate": 2e-05,
      "loss": 0.7921567440032959,
      "step": 640
    },
    {
      "epoch": 0.08244545915778792,
      "grad_norm": 4.2250838369835995,
      "learning_rate": 2e-05,
      "loss": 0.7983972072601319,
      "step": 650
    },
    {
      "epoch": 0.0837138508371385,
      "grad_norm": 3.400756338744929,
      "learning_rate": 2e-05,
      "loss": 0.7732720851898194,
      "step": 660
    },
    {
      "epoch": 0.08498224251648909,
      "grad_norm": 3.378273815388625,
      "learning_rate": 2e-05,
      "loss": 0.7511517524719238,
      "step": 670
    },
    {
      "epoch": 0.08625063419583967,
      "grad_norm": 7.977471259521566,
      "learning_rate": 2e-05,
      "loss": 0.7838554382324219,
      "step": 680
    },
    {
      "epoch": 0.08751902587519025,
      "grad_norm": 12.791982347134267,
      "learning_rate": 2e-05,
      "loss": 0.7150951385498047,
      "step": 690
    },
    {
      "epoch": 0.08878741755454084,
      "grad_norm": 4.89344325930809,
      "learning_rate": 2e-05,
      "loss": 0.7147266387939453,
      "step": 700
    },
    {
      "epoch": 0.09005580923389142,
      "grad_norm": 10.031854241822552,
      "learning_rate": 2e-05,
      "loss": 0.7478272914886475,
      "step": 710
    },
    {
      "epoch": 0.091324200913242,
      "grad_norm": 7.312592163945764,
      "learning_rate": 2e-05,
      "loss": 0.7561559200286865,
      "step": 720
    },
    {
      "epoch": 0.09259259259259259,
      "grad_norm": 11.09817990507854,
      "learning_rate": 2e-05,
      "loss": 0.7951640605926513,
      "step": 730
    },
    {
      "epoch": 0.09386098427194317,
      "grad_norm": 27.769822647988097,
      "learning_rate": 2e-05,
      "loss": 0.7679110527038574,
      "step": 740
    },
    {
      "epoch": 0.09512937595129375,
      "grad_norm": 21.450872584188705,
      "learning_rate": 2e-05,
      "loss": 0.7979055404663086,
      "step": 750
    },
    {
      "epoch": 0.09639776763064434,
      "grad_norm": 4.765278860263813,
      "learning_rate": 2e-05,
      "loss": 0.8046473503112793,
      "step": 760
    },
    {
      "epoch": 0.09766615930999492,
      "grad_norm": 2.04007320738131,
      "learning_rate": 2e-05,
      "loss": 0.8117101669311524,
      "step": 770
    },
    {
      "epoch": 0.0989345509893455,
      "grad_norm": 2.7051506989482275,
      "learning_rate": 2e-05,
      "loss": 0.8098690986633301,
      "step": 780
    },
    {
      "epoch": 0.10020294266869609,
      "grad_norm": 2.281127749493329,
      "learning_rate": 2e-05,
      "loss": 0.7546245098114014,
      "step": 790
    },
    {
      "epoch": 0.10147133434804667,
      "grad_norm": 9.68965910889911,
      "learning_rate": 2e-05,
      "loss": 0.7671361446380616,
      "step": 800
    },
    {
      "epoch": 0.10273972602739725,
      "grad_norm": 2.7141801026370884,
      "learning_rate": 2e-05,
      "loss": 0.7550729274749756,
      "step": 810
    },
    {
      "epoch": 0.10400811770674784,
      "grad_norm": 4.703177691438535,
      "learning_rate": 2e-05,
      "loss": 0.7441450119018554,
      "step": 820
    },
    {
      "epoch": 0.10527650938609842,
      "grad_norm": 5.497045164078779,
      "learning_rate": 2e-05,
      "loss": 0.7636425971984864,
      "step": 830
    },
    {
      "epoch": 0.106544901065449,
      "grad_norm": 8.505864828865908,
      "learning_rate": 2e-05,
      "loss": 0.7817797660827637,
      "step": 840
    },
    {
      "epoch": 0.10781329274479959,
      "grad_norm": 1.6752582092320205,
      "learning_rate": 2e-05,
      "loss": 0.7469396114349365,
      "step": 850
    },
    {
      "epoch": 0.10908168442415017,
      "grad_norm": 1.9988229757535019,
      "learning_rate": 2e-05,
      "loss": 0.7521775722503662,
      "step": 860
    },
    {
      "epoch": 0.11035007610350075,
      "grad_norm": 2.9771762529902226,
      "learning_rate": 2e-05,
      "loss": 0.7262326717376709,
      "step": 870
    },
    {
      "epoch": 0.11161846778285134,
      "grad_norm": 4.359971674060599,
      "learning_rate": 2e-05,
      "loss": 0.7141817092895508,
      "step": 880
    },
    {
      "epoch": 0.11288685946220192,
      "grad_norm": 2.7133251609907787,
      "learning_rate": 2e-05,
      "loss": 0.7290772438049317,
      "step": 890
    },
    {
      "epoch": 0.1141552511415525,
      "grad_norm": 6.106170728411568,
      "learning_rate": 2e-05,
      "loss": 0.7642468929290771,
      "step": 900
    },
    {
      "epoch": 0.11542364282090309,
      "grad_norm": 3.3301287722292923,
      "learning_rate": 2e-05,
      "loss": 0.7298625946044922,
      "step": 910
    },
    {
      "epoch": 0.11669203450025367,
      "grad_norm": 39.04355754651565,
      "learning_rate": 2e-05,
      "loss": 0.7515697002410888,
      "step": 920
    },
    {
      "epoch": 0.11796042617960426,
      "grad_norm": 1.6743927853780125,
      "learning_rate": 2e-05,
      "loss": 0.7205712795257568,
      "step": 930
    },
    {
      "epoch": 0.11922881785895484,
      "grad_norm": 2.11526914939453,
      "learning_rate": 2e-05,
      "loss": 0.7325911521911621,
      "step": 940
    },
    {
      "epoch": 0.12049720953830542,
      "grad_norm": 1.5910009685089883,
      "learning_rate": 2e-05,
      "loss": 0.7398830413818359,
      "step": 950
    },
    {
      "epoch": 0.121765601217656,
      "grad_norm": 1.3266787607239245,
      "learning_rate": 2e-05,
      "loss": 0.7558593273162841,
      "step": 960
    },
    {
      "epoch": 0.12303399289700659,
      "grad_norm": 1.3480281216204522,
      "learning_rate": 2e-05,
      "loss": 0.7310697078704834,
      "step": 970
    },
    {
      "epoch": 0.12430238457635717,
      "grad_norm": 2.926412513267041,
      "learning_rate": 2e-05,
      "loss": 0.7443419456481933,
      "step": 980
    },
    {
      "epoch": 0.12557077625570776,
      "grad_norm": 3.7779872148241593,
      "learning_rate": 2e-05,
      "loss": 0.7498525619506836,
      "step": 990
    },
    {
      "epoch": 0.12683916793505834,
      "grad_norm": 2.0522172619812333,
      "learning_rate": 2e-05,
      "loss": 0.735056209564209,
      "step": 1000
    },
    {
      "epoch": 0.12810755961440892,
      "grad_norm": 10.869060832827572,
      "learning_rate": 2e-05,
      "loss": 0.7341260433197021,
      "step": 1010
    },
    {
      "epoch": 0.1293759512937595,
      "grad_norm": 1.6798605890669833,
      "learning_rate": 2e-05,
      "loss": 0.7389917373657227,
      "step": 1020
    },
    {
      "epoch": 0.1306443429731101,
      "grad_norm": 2.8685953405858613,
      "learning_rate": 2e-05,
      "loss": 0.7330316066741943,
      "step": 1030
    },
    {
      "epoch": 0.13191273465246067,
      "grad_norm": 4.115814313787979,
      "learning_rate": 2e-05,
      "loss": 0.7543672084808349,
      "step": 1040
    },
    {
      "epoch": 0.13318112633181126,
      "grad_norm": 2.198966228015639,
      "learning_rate": 2e-05,
      "loss": 0.7534980773925781,
      "step": 1050
    },
    {
      "epoch": 0.13444951801116184,
      "grad_norm": 2.0978505874507167,
      "learning_rate": 2e-05,
      "loss": 0.7538117408752442,
      "step": 1060
    },
    {
      "epoch": 0.13571790969051242,
      "grad_norm": 2.611270635882454,
      "learning_rate": 2e-05,
      "loss": 0.7457163810729981,
      "step": 1070
    },
    {
      "epoch": 0.136986301369863,
      "grad_norm": 3.450646794729338,
      "learning_rate": 2e-05,
      "loss": 0.7410408020019531,
      "step": 1080
    },
    {
      "epoch": 0.1382546930492136,
      "grad_norm": 2.4573365052298346,
      "learning_rate": 2e-05,
      "loss": 0.7419426918029786,
      "step": 1090
    },
    {
      "epoch": 0.13952308472856417,
      "grad_norm": 1.748725505992879,
      "learning_rate": 2e-05,
      "loss": 0.7437388896942139,
      "step": 1100
    },
    {
      "epoch": 0.14079147640791476,
      "grad_norm": 1.924832474483585,
      "learning_rate": 2e-05,
      "loss": 0.7412117004394532,
      "step": 1110
    },
    {
      "epoch": 0.14205986808726534,
      "grad_norm": 3.215512120480945,
      "learning_rate": 2e-05,
      "loss": 0.7271268367767334,
      "step": 1120
    },
    {
      "epoch": 0.14332825976661592,
      "grad_norm": 1.5670129420356125,
      "learning_rate": 2e-05,
      "loss": 0.7251224040985107,
      "step": 1130
    },
    {
      "epoch": 0.1445966514459665,
      "grad_norm": 16.612579774844967,
      "learning_rate": 2e-05,
      "loss": 0.7283561706542969,
      "step": 1140
    },
    {
      "epoch": 0.1458650431253171,
      "grad_norm": 5.8423278328944095,
      "learning_rate": 2e-05,
      "loss": 0.7479527473449707,
      "step": 1150
    },
    {
      "epoch": 0.14713343480466767,
      "grad_norm": 1.9006248134347878,
      "learning_rate": 2e-05,
      "loss": 0.7520655632019043,
      "step": 1160
    },
    {
      "epoch": 0.14840182648401826,
      "grad_norm": 1.7166375828020946,
      "learning_rate": 2e-05,
      "loss": 0.7638471126556396,
      "step": 1170
    },
    {
      "epoch": 0.14967021816336884,
      "grad_norm": 2.3480339420357725,
      "learning_rate": 2e-05,
      "loss": 0.722047758102417,
      "step": 1180
    },
    {
      "epoch": 0.15093860984271942,
      "grad_norm": 1.3207278692310833,
      "learning_rate": 2e-05,
      "loss": 0.7205617427825928,
      "step": 1190
    },
    {
      "epoch": 0.15220700152207,
      "grad_norm": 2.9950437796687046,
      "learning_rate": 2e-05,
      "loss": 0.7470908164978027,
      "step": 1200
    },
    {
      "epoch": 0.1534753932014206,
      "grad_norm": 6.571774732431327,
      "learning_rate": 2e-05,
      "loss": 0.7376546859741211,
      "step": 1210
    },
    {
      "epoch": 0.15474378488077117,
      "grad_norm": 5.904648441953073,
      "learning_rate": 2e-05,
      "loss": 0.77837233543396,
      "step": 1220
    },
    {
      "epoch": 0.15601217656012176,
      "grad_norm": 3.2711212827912597,
      "learning_rate": 2e-05,
      "loss": 0.7514628410339356,
      "step": 1230
    },
    {
      "epoch": 0.15728056823947234,
      "grad_norm": 1.9508676742710445,
      "learning_rate": 2e-05,
      "loss": 0.7363358974456787,
      "step": 1240
    },
    {
      "epoch": 0.15854895991882292,
      "grad_norm": 13.220540634786495,
      "learning_rate": 2e-05,
      "loss": 0.7491374969482422,
      "step": 1250
    },
    {
      "epoch": 0.1598173515981735,
      "grad_norm": 3.5098678059958566,
      "learning_rate": 2e-05,
      "loss": 0.7424610137939454,
      "step": 1260
    },
    {
      "epoch": 0.1610857432775241,
      "grad_norm": 1.6389079515026512,
      "learning_rate": 2e-05,
      "loss": 0.7745975017547607,
      "step": 1270
    },
    {
      "epoch": 0.16235413495687467,
      "grad_norm": 13.97225969589928,
      "learning_rate": 2e-05,
      "loss": 0.7480620384216309,
      "step": 1280
    },
    {
      "epoch": 0.16362252663622526,
      "grad_norm": 2.1924141970599487,
      "learning_rate": 2e-05,
      "loss": 0.7510156154632568,
      "step": 1290
    },
    {
      "epoch": 0.16489091831557584,
      "grad_norm": 1.3452790613599073,
      "learning_rate": 2e-05,
      "loss": 0.7685614585876465,
      "step": 1300
    },
    {
      "epoch": 0.16615930999492642,
      "grad_norm": 2.071107770735763,
      "learning_rate": 2e-05,
      "loss": 0.7374220848083496,
      "step": 1310
    },
    {
      "epoch": 0.167427701674277,
      "grad_norm": 2.3317358334951868,
      "learning_rate": 2e-05,
      "loss": 0.7329941749572754,
      "step": 1320
    },
    {
      "epoch": 0.1686960933536276,
      "grad_norm": 2.049534758856822,
      "learning_rate": 2e-05,
      "loss": 0.7270294189453125,
      "step": 1330
    },
    {
      "epoch": 0.16996448503297817,
      "grad_norm": 1.2605992715180732,
      "learning_rate": 2e-05,
      "loss": 0.7454747676849365,
      "step": 1340
    },
    {
      "epoch": 0.17123287671232876,
      "grad_norm": 1.7131700085587056,
      "learning_rate": 2e-05,
      "loss": 0.7461020469665527,
      "step": 1350
    },
    {
      "epoch": 0.17250126839167934,
      "grad_norm": 2.849784492937859,
      "learning_rate": 2e-05,
      "loss": 0.8431632995605469,
      "step": 1360
    },
    {
      "epoch": 0.17376966007102992,
      "grad_norm": 3.0236082194089984,
      "learning_rate": 2e-05,
      "loss": 0.726828670501709,
      "step": 1370
    },
    {
      "epoch": 0.1750380517503805,
      "grad_norm": 1.166381172185534,
      "learning_rate": 2e-05,
      "loss": 0.7194171905517578,
      "step": 1380
    },
    {
      "epoch": 0.1763064434297311,
      "grad_norm": 2.004785022977192,
      "learning_rate": 2e-05,
      "loss": 0.7266772270202637,
      "step": 1390
    },
    {
      "epoch": 0.17757483510908167,
      "grad_norm": 2.7640622828407233,
      "learning_rate": 2e-05,
      "loss": 0.7030804634094239,
      "step": 1400
    },
    {
      "epoch": 0.17884322678843226,
      "grad_norm": 0.9942969879152285,
      "learning_rate": 2e-05,
      "loss": 0.7589815139770508,
      "step": 1410
    },
    {
      "epoch": 0.18011161846778284,
      "grad_norm": 1.1575392845460777,
      "learning_rate": 2e-05,
      "loss": 0.7294116020202637,
      "step": 1420
    },
    {
      "epoch": 0.18138001014713342,
      "grad_norm": 1.2694035310775686,
      "learning_rate": 2e-05,
      "loss": 0.7456828117370605,
      "step": 1430
    },
    {
      "epoch": 0.182648401826484,
      "grad_norm": 1.1155648418704232,
      "learning_rate": 2e-05,
      "loss": 0.6931851863861084,
      "step": 1440
    },
    {
      "epoch": 0.1839167935058346,
      "grad_norm": 1.121651564064941,
      "learning_rate": 2e-05,
      "loss": 0.7264601707458496,
      "step": 1450
    },
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 1.7083170624115234,
      "learning_rate": 2e-05,
      "loss": 0.737404203414917,
      "step": 1460
    },
    {
      "epoch": 0.18645357686453576,
      "grad_norm": 1.5317494768904325,
      "learning_rate": 2e-05,
      "loss": 0.7606875419616699,
      "step": 1470
    },
    {
      "epoch": 0.18772196854388634,
      "grad_norm": 1.3676356906499576,
      "learning_rate": 2e-05,
      "loss": 0.7529849052429199,
      "step": 1480
    },
    {
      "epoch": 0.18899036022323693,
      "grad_norm": 2.7476674910958465,
      "learning_rate": 2e-05,
      "loss": 0.7227797985076905,
      "step": 1490
    },
    {
      "epoch": 0.1902587519025875,
      "grad_norm": 5.647437085895668,
      "learning_rate": 2e-05,
      "loss": 0.7442465782165527,
      "step": 1500
    },
    {
      "epoch": 0.1915271435819381,
      "grad_norm": 1.9297518793351933,
      "learning_rate": 2e-05,
      "loss": 0.7151328086853027,
      "step": 1510
    },
    {
      "epoch": 0.19279553526128868,
      "grad_norm": 1.4791642571910928,
      "learning_rate": 2e-05,
      "loss": 0.7258416652679444,
      "step": 1520
    },
    {
      "epoch": 0.19406392694063926,
      "grad_norm": 1.4126307142610093,
      "learning_rate": 2e-05,
      "loss": 0.7352984428405762,
      "step": 1530
    },
    {
      "epoch": 0.19533231861998984,
      "grad_norm": 2.555568996382671,
      "learning_rate": 2e-05,
      "loss": 0.7223186492919922,
      "step": 1540
    },
    {
      "epoch": 0.19660071029934043,
      "grad_norm": 1.3153890277344187,
      "learning_rate": 2e-05,
      "loss": 0.7090831756591797,
      "step": 1550
    },
    {
      "epoch": 0.197869101978691,
      "grad_norm": 3.859084850993966,
      "learning_rate": 2e-05,
      "loss": 0.7268383979797364,
      "step": 1560
    },
    {
      "epoch": 0.1991374936580416,
      "grad_norm": 3.2098222352393506,
      "learning_rate": 2e-05,
      "loss": 0.7043148040771484,
      "step": 1570
    },
    {
      "epoch": 0.20040588533739218,
      "grad_norm": 7.724187482406483,
      "learning_rate": 2e-05,
      "loss": 0.7330468177795411,
      "step": 1580
    },
    {
      "epoch": 0.20167427701674276,
      "grad_norm": 1.2978782113134073,
      "learning_rate": 2e-05,
      "loss": 0.7313216686248779,
      "step": 1590
    },
    {
      "epoch": 0.20294266869609334,
      "grad_norm": 1.418496505826777,
      "learning_rate": 2e-05,
      "loss": 0.6993028163909912,
      "step": 1600
    },
    {
      "epoch": 0.20421106037544393,
      "grad_norm": 3.3144867771110054,
      "learning_rate": 2e-05,
      "loss": 0.7366094589233398,
      "step": 1610
    },
    {
      "epoch": 0.2054794520547945,
      "grad_norm": 3.9945844024628303,
      "learning_rate": 2e-05,
      "loss": 0.7112298965454101,
      "step": 1620
    },
    {
      "epoch": 0.2067478437341451,
      "grad_norm": 5.585977116213833,
      "learning_rate": 2e-05,
      "loss": 0.7064263820648193,
      "step": 1630
    },
    {
      "epoch": 0.20801623541349568,
      "grad_norm": 1.6312453458764222,
      "learning_rate": 2e-05,
      "loss": 0.6861506462097168,
      "step": 1640
    },
    {
      "epoch": 0.20928462709284626,
      "grad_norm": 1.7363788523388355,
      "learning_rate": 2e-05,
      "loss": 0.7475335597991943,
      "step": 1650
    },
    {
      "epoch": 0.21055301877219684,
      "grad_norm": 2.1479220254126483,
      "learning_rate": 2e-05,
      "loss": 0.7279925346374512,
      "step": 1660
    },
    {
      "epoch": 0.21182141045154743,
      "grad_norm": 1.88205862365636,
      "learning_rate": 2e-05,
      "loss": 0.7343072414398193,
      "step": 1670
    },
    {
      "epoch": 0.213089802130898,
      "grad_norm": 1.9552107389959414,
      "learning_rate": 2e-05,
      "loss": 0.7220296382904052,
      "step": 1680
    },
    {
      "epoch": 0.2143581938102486,
      "grad_norm": 1.6713484873020035,
      "learning_rate": 2e-05,
      "loss": 0.7099277019500733,
      "step": 1690
    },
    {
      "epoch": 0.21562658548959918,
      "grad_norm": 2.0564439747818066,
      "learning_rate": 2e-05,
      "loss": 0.7099870681762696,
      "step": 1700
    },
    {
      "epoch": 0.21689497716894976,
      "grad_norm": 1.1650425835118574,
      "learning_rate": 2e-05,
      "loss": 0.731694221496582,
      "step": 1710
    },
    {
      "epoch": 0.21816336884830034,
      "grad_norm": 1.7655637891911142,
      "learning_rate": 2e-05,
      "loss": 0.729124641418457,
      "step": 1720
    },
    {
      "epoch": 0.21943176052765093,
      "grad_norm": 1.0476006014812598,
      "learning_rate": 2e-05,
      "loss": 0.720085859298706,
      "step": 1730
    },
    {
      "epoch": 0.2207001522070015,
      "grad_norm": 1.0999549427992747,
      "learning_rate": 2e-05,
      "loss": 0.755762767791748,
      "step": 1740
    },
    {
      "epoch": 0.2219685438863521,
      "grad_norm": 1.5088812786776697,
      "learning_rate": 2e-05,
      "loss": 0.7249419212341308,
      "step": 1750
    },
    {
      "epoch": 0.22323693556570268,
      "grad_norm": 2.251935110841283,
      "learning_rate": 2e-05,
      "loss": 0.7105390071868897,
      "step": 1760
    },
    {
      "epoch": 0.22450532724505326,
      "grad_norm": 1.5950634433275381,
      "learning_rate": 2e-05,
      "loss": 0.721195125579834,
      "step": 1770
    },
    {
      "epoch": 0.22577371892440384,
      "grad_norm": 1.6110432413872982,
      "learning_rate": 2e-05,
      "loss": 0.7167281150817871,
      "step": 1780
    },
    {
      "epoch": 0.22704211060375443,
      "grad_norm": 2.9799697562913456,
      "learning_rate": 2e-05,
      "loss": 0.7384551048278809,
      "step": 1790
    },
    {
      "epoch": 0.228310502283105,
      "grad_norm": 3.528432886105676,
      "learning_rate": 2e-05,
      "loss": 0.7057220458984375,
      "step": 1800
    },
    {
      "epoch": 0.2295788939624556,
      "grad_norm": 2.24445322913813,
      "learning_rate": 2e-05,
      "loss": 0.7273429870605469,
      "step": 1810
    },
    {
      "epoch": 0.23084728564180618,
      "grad_norm": 1.341910260515698,
      "learning_rate": 2e-05,
      "loss": 0.7449875831604004,
      "step": 1820
    },
    {
      "epoch": 0.23211567732115676,
      "grad_norm": 2.947858969598419,
      "learning_rate": 2e-05,
      "loss": 0.7049365997314453,
      "step": 1830
    },
    {
      "epoch": 0.23338406900050734,
      "grad_norm": 1.6939139925030757,
      "learning_rate": 2e-05,
      "loss": 0.7313479423522949,
      "step": 1840
    },
    {
      "epoch": 0.23465246067985793,
      "grad_norm": 1.4223535978007373,
      "learning_rate": 2e-05,
      "loss": 0.7094035148620605,
      "step": 1850
    },
    {
      "epoch": 0.2359208523592085,
      "grad_norm": 2.041218983645498,
      "learning_rate": 2e-05,
      "loss": 0.716699743270874,
      "step": 1860
    },
    {
      "epoch": 0.2371892440385591,
      "grad_norm": 1.695792460316363,
      "learning_rate": 2e-05,
      "loss": 0.7320201396942139,
      "step": 1870
    },
    {
      "epoch": 0.23845763571790968,
      "grad_norm": 6.249869346100446,
      "learning_rate": 2e-05,
      "loss": 0.7185420036315918,
      "step": 1880
    },
    {
      "epoch": 0.23972602739726026,
      "grad_norm": 2.047223818317003,
      "learning_rate": 2e-05,
      "loss": 0.7600899696350097,
      "step": 1890
    },
    {
      "epoch": 0.24099441907661084,
      "grad_norm": 1.674008876941635,
      "learning_rate": 2e-05,
      "loss": 0.7453332424163819,
      "step": 1900
    },
    {
      "epoch": 0.24226281075596143,
      "grad_norm": 1.3861493421287727,
      "learning_rate": 2e-05,
      "loss": 0.7336289882659912,
      "step": 1910
    },
    {
      "epoch": 0.243531202435312,
      "grad_norm": 1.3239455071295139,
      "learning_rate": 2e-05,
      "loss": 0.7150239944458008,
      "step": 1920
    },
    {
      "epoch": 0.2447995941146626,
      "grad_norm": 1.2043124813437642,
      "learning_rate": 2e-05,
      "loss": 0.6964576721191407,
      "step": 1930
    },
    {
      "epoch": 0.24606798579401318,
      "grad_norm": 16.14821981447028,
      "learning_rate": 2e-05,
      "loss": 0.7451602935791015,
      "step": 1940
    },
    {
      "epoch": 0.24733637747336376,
      "grad_norm": 12.659813710174426,
      "learning_rate": 2e-05,
      "loss": 0.7130792140960693,
      "step": 1950
    },
    {
      "epoch": 0.24860476915271434,
      "grad_norm": 2.107096485448891,
      "learning_rate": 2e-05,
      "loss": 0.7027217388153076,
      "step": 1960
    },
    {
      "epoch": 0.24987316083206493,
      "grad_norm": 1.4289179116664272,
      "learning_rate": 2e-05,
      "loss": 0.6999739646911621,
      "step": 1970
    },
    {
      "epoch": 0.2511415525114155,
      "grad_norm": 2.4647165146521623,
      "learning_rate": 2e-05,
      "loss": 0.7236842155456543,
      "step": 1980
    },
    {
      "epoch": 0.2524099441907661,
      "grad_norm": 1.247644508895393,
      "learning_rate": 2e-05,
      "loss": 0.7129887580871582,
      "step": 1990
    },
    {
      "epoch": 0.2536783358701167,
      "grad_norm": 7.2870630530778335,
      "learning_rate": 2e-05,
      "loss": 0.7318644523620605,
      "step": 2000
    },
    {
      "epoch": 0.2549467275494673,
      "grad_norm": 497.897533935121,
      "learning_rate": 2e-05,
      "loss": 0.7083016872406006,
      "step": 2010
    },
    {
      "epoch": 0.25621511922881784,
      "grad_norm": 1.6424977467606583,
      "learning_rate": 2e-05,
      "loss": 0.752432918548584,
      "step": 2020
    },
    {
      "epoch": 0.25748351090816846,
      "grad_norm": 3.397364213682138,
      "learning_rate": 2e-05,
      "loss": 0.7201516628265381,
      "step": 2030
    },
    {
      "epoch": 0.258751902587519,
      "grad_norm": 1.5436173241827034,
      "learning_rate": 2e-05,
      "loss": 0.7620837211608886,
      "step": 2040
    },
    {
      "epoch": 0.2600202942668696,
      "grad_norm": 1.7654082191192209,
      "learning_rate": 2e-05,
      "loss": 0.7176222801208496,
      "step": 2050
    },
    {
      "epoch": 0.2612886859462202,
      "grad_norm": 2.3352854727589,
      "learning_rate": 2e-05,
      "loss": 0.7700258255004883,
      "step": 2060
    },
    {
      "epoch": 0.2625570776255708,
      "grad_norm": 1.5811874772774719,
      "learning_rate": 2e-05,
      "loss": 0.6814945220947266,
      "step": 2070
    },
    {
      "epoch": 0.26382546930492135,
      "grad_norm": 3.399074264943864,
      "learning_rate": 2e-05,
      "loss": 0.7022511959075928,
      "step": 2080
    },
    {
      "epoch": 0.26509386098427196,
      "grad_norm": 1.6919456742001187,
      "learning_rate": 2e-05,
      "loss": 0.7267668724060059,
      "step": 2090
    },
    {
      "epoch": 0.2663622526636225,
      "grad_norm": 1.640404098573094,
      "learning_rate": 2e-05,
      "loss": 0.7226474285125732,
      "step": 2100
    },
    {
      "epoch": 0.2676306443429731,
      "grad_norm": 22.655915856588393,
      "learning_rate": 2e-05,
      "loss": 0.6992047309875489,
      "step": 2110
    },
    {
      "epoch": 0.2688990360223237,
      "grad_norm": 1.395402888150166,
      "learning_rate": 2e-05,
      "loss": 0.710693645477295,
      "step": 2120
    },
    {
      "epoch": 0.2701674277016743,
      "grad_norm": 3.0264214357255854,
      "learning_rate": 2e-05,
      "loss": 0.7098849296569825,
      "step": 2130
    },
    {
      "epoch": 0.27143581938102485,
      "grad_norm": 1.5549201004782767,
      "learning_rate": 2e-05,
      "loss": 0.7324645042419433,
      "step": 2140
    },
    {
      "epoch": 0.27270421106037546,
      "grad_norm": 1.6402334284757991,
      "learning_rate": 2e-05,
      "loss": 0.7227683544158936,
      "step": 2150
    },
    {
      "epoch": 0.273972602739726,
      "grad_norm": 1.1166755897688916,
      "learning_rate": 2e-05,
      "loss": 0.7160517692565918,
      "step": 2160
    },
    {
      "epoch": 0.2752409944190766,
      "grad_norm": 1.807488373841244,
      "learning_rate": 2e-05,
      "loss": 0.7465599060058594,
      "step": 2170
    },
    {
      "epoch": 0.2765093860984272,
      "grad_norm": 1.1949058602206344,
      "learning_rate": 2e-05,
      "loss": 0.7288659095764161,
      "step": 2180
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 1.2604140373237862,
      "learning_rate": 2e-05,
      "loss": 0.7248151302337646,
      "step": 2190
    },
    {
      "epoch": 0.27904616945712835,
      "grad_norm": 1.4853739413549576,
      "learning_rate": 2e-05,
      "loss": 0.7410670280456543,
      "step": 2200
    },
    {
      "epoch": 0.28031456113647896,
      "grad_norm": 1.4048591876600616,
      "learning_rate": 2e-05,
      "loss": 0.733606481552124,
      "step": 2210
    },
    {
      "epoch": 0.2815829528158295,
      "grad_norm": 1.2175883604734996,
      "learning_rate": 2e-05,
      "loss": 0.6726350784301758,
      "step": 2220
    },
    {
      "epoch": 0.2828513444951801,
      "grad_norm": 3.2751860069739216,
      "learning_rate": 2e-05,
      "loss": 0.739863920211792,
      "step": 2230
    },
    {
      "epoch": 0.2841197361745307,
      "grad_norm": 1.262770690252382,
      "learning_rate": 2e-05,
      "loss": 0.7283521175384522,
      "step": 2240
    },
    {
      "epoch": 0.2853881278538813,
      "grad_norm": 1.4030781263402066,
      "learning_rate": 2e-05,
      "loss": 0.7302201747894287,
      "step": 2250
    },
    {
      "epoch": 0.28665651953323185,
      "grad_norm": 2.5692713162277645,
      "learning_rate": 2e-05,
      "loss": 0.7376399993896484,
      "step": 2260
    },
    {
      "epoch": 0.28792491121258246,
      "grad_norm": 1.5536228582466036,
      "learning_rate": 2e-05,
      "loss": 0.738432788848877,
      "step": 2270
    },
    {
      "epoch": 0.289193302891933,
      "grad_norm": 2.5167357217788884,
      "learning_rate": 2e-05,
      "loss": 0.7069236755371093,
      "step": 2280
    },
    {
      "epoch": 0.2904616945712836,
      "grad_norm": 3.803185283228567,
      "learning_rate": 2e-05,
      "loss": 0.7257742881774902,
      "step": 2290
    },
    {
      "epoch": 0.2917300862506342,
      "grad_norm": 3.6554616044174364,
      "learning_rate": 2e-05,
      "loss": 0.7072547912597656,
      "step": 2300
    },
    {
      "epoch": 0.2929984779299848,
      "grad_norm": 5.134785607460091,
      "learning_rate": 2e-05,
      "loss": 0.7284947872161865,
      "step": 2310
    },
    {
      "epoch": 0.29426686960933535,
      "grad_norm": 1.3510324368641156,
      "learning_rate": 2e-05,
      "loss": 0.7529828548431396,
      "step": 2320
    },
    {
      "epoch": 0.29553526128868596,
      "grad_norm": 1.183405585175162,
      "learning_rate": 2e-05,
      "loss": 0.7215205192565918,
      "step": 2330
    },
    {
      "epoch": 0.2968036529680365,
      "grad_norm": 2.3520501262314806,
      "learning_rate": 2e-05,
      "loss": 0.7097564220428467,
      "step": 2340
    },
    {
      "epoch": 0.2980720446473871,
      "grad_norm": 1.6341945691709432,
      "learning_rate": 2e-05,
      "loss": 0.73095703125,
      "step": 2350
    },
    {
      "epoch": 0.2993404363267377,
      "grad_norm": 1.3864179672047938,
      "learning_rate": 2e-05,
      "loss": 0.7215674877166748,
      "step": 2360
    },
    {
      "epoch": 0.3006088280060883,
      "grad_norm": 1.1824875413410951,
      "learning_rate": 2e-05,
      "loss": 0.6878623008728028,
      "step": 2370
    },
    {
      "epoch": 0.30187721968543885,
      "grad_norm": 2.6781623493923985,
      "learning_rate": 2e-05,
      "loss": 0.7135646343231201,
      "step": 2380
    },
    {
      "epoch": 0.30314561136478946,
      "grad_norm": 1.0033073751403543,
      "learning_rate": 2e-05,
      "loss": 0.6955747127532959,
      "step": 2390
    },
    {
      "epoch": 0.30441400304414,
      "grad_norm": 2.352671006703847,
      "learning_rate": 2e-05,
      "loss": 0.7020764350891113,
      "step": 2400
    },
    {
      "epoch": 0.3056823947234906,
      "grad_norm": 1.6278074402038472,
      "learning_rate": 2e-05,
      "loss": 0.7096355438232422,
      "step": 2410
    },
    {
      "epoch": 0.3069507864028412,
      "grad_norm": 20.28292981631277,
      "learning_rate": 2e-05,
      "loss": 0.728090476989746,
      "step": 2420
    },
    {
      "epoch": 0.3082191780821918,
      "grad_norm": 1.462297285534622,
      "learning_rate": 2e-05,
      "loss": 0.725699520111084,
      "step": 2430
    },
    {
      "epoch": 0.30948756976154235,
      "grad_norm": 3.1309035603601614,
      "learning_rate": 2e-05,
      "loss": 0.7337343215942382,
      "step": 2440
    },
    {
      "epoch": 0.31075596144089296,
      "grad_norm": 3.6288963836099186,
      "learning_rate": 2e-05,
      "loss": 0.7274514675140381,
      "step": 2450
    },
    {
      "epoch": 0.3120243531202435,
      "grad_norm": 1.3478303002346632,
      "learning_rate": 2e-05,
      "loss": 0.7406110286712646,
      "step": 2460
    },
    {
      "epoch": 0.3132927447995941,
      "grad_norm": 1.3770554160511144,
      "learning_rate": 2e-05,
      "loss": 0.6926933288574219,
      "step": 2470
    },
    {
      "epoch": 0.3145611364789447,
      "grad_norm": 4.428418325943364,
      "learning_rate": 2e-05,
      "loss": 0.6768437385559082,
      "step": 2480
    },
    {
      "epoch": 0.3158295281582953,
      "grad_norm": 1.5429669669495916,
      "learning_rate": 2e-05,
      "loss": 0.7363099575042724,
      "step": 2490
    },
    {
      "epoch": 0.31709791983764585,
      "grad_norm": 1.1766770848271058,
      "learning_rate": 2e-05,
      "loss": 0.696164608001709,
      "step": 2500
    },
    {
      "epoch": 0.31836631151699646,
      "grad_norm": 1.4117017579512698,
      "learning_rate": 2e-05,
      "loss": 0.7026076316833496,
      "step": 2510
    },
    {
      "epoch": 0.319634703196347,
      "grad_norm": 3.466803159925888,
      "learning_rate": 2e-05,
      "loss": 0.6960143089294434,
      "step": 2520
    },
    {
      "epoch": 0.3209030948756976,
      "grad_norm": 7.388915910161205,
      "learning_rate": 2e-05,
      "loss": 0.7391340255737304,
      "step": 2530
    },
    {
      "epoch": 0.3221714865550482,
      "grad_norm": 2.3726802661340662,
      "learning_rate": 2e-05,
      "loss": 0.6932723045349121,
      "step": 2540
    },
    {
      "epoch": 0.3234398782343988,
      "grad_norm": 4.908968345107884,
      "learning_rate": 2e-05,
      "loss": 0.7188965797424316,
      "step": 2550
    },
    {
      "epoch": 0.32470826991374935,
      "grad_norm": 1.7650720301188074,
      "learning_rate": 2e-05,
      "loss": 0.7650503158569336,
      "step": 2560
    },
    {
      "epoch": 0.32597666159309996,
      "grad_norm": 28.282675255846673,
      "learning_rate": 2e-05,
      "loss": 0.7194470882415771,
      "step": 2570
    },
    {
      "epoch": 0.3272450532724505,
      "grad_norm": 1.289620962298377,
      "learning_rate": 2e-05,
      "loss": 0.7174947261810303,
      "step": 2580
    },
    {
      "epoch": 0.3285134449518011,
      "grad_norm": 2.471745175505916,
      "learning_rate": 2e-05,
      "loss": 0.7201624870300293,
      "step": 2590
    },
    {
      "epoch": 0.3297818366311517,
      "grad_norm": 1.484982579212836,
      "learning_rate": 2e-05,
      "loss": 0.7049246788024902,
      "step": 2600
    },
    {
      "epoch": 0.3310502283105023,
      "grad_norm": 1.330985677959722,
      "learning_rate": 2e-05,
      "loss": 0.7232493400573731,
      "step": 2610
    },
    {
      "epoch": 0.33231861998985285,
      "grad_norm": 1.4423014794542133,
      "learning_rate": 2e-05,
      "loss": 0.7062403678894043,
      "step": 2620
    },
    {
      "epoch": 0.33358701166920346,
      "grad_norm": 1.5752871357066351,
      "learning_rate": 2e-05,
      "loss": 0.6895558834075928,
      "step": 2630
    },
    {
      "epoch": 0.334855403348554,
      "grad_norm": 1.2374500032039994,
      "learning_rate": 2e-05,
      "loss": 0.6963345527648925,
      "step": 2640
    },
    {
      "epoch": 0.3361237950279046,
      "grad_norm": 1.3154361632180258,
      "learning_rate": 2e-05,
      "loss": 0.6861466407775879,
      "step": 2650
    },
    {
      "epoch": 0.3373921867072552,
      "grad_norm": 1.1130033889551345,
      "learning_rate": 2e-05,
      "loss": 0.6945218563079834,
      "step": 2660
    },
    {
      "epoch": 0.3386605783866058,
      "grad_norm": 1.0136853728786341,
      "learning_rate": 2e-05,
      "loss": 0.741546630859375,
      "step": 2670
    },
    {
      "epoch": 0.33992897006595635,
      "grad_norm": 1.328610164219176,
      "learning_rate": 2e-05,
      "loss": 0.6940020561218262,
      "step": 2680
    },
    {
      "epoch": 0.34119736174530696,
      "grad_norm": 0.9887684918869929,
      "learning_rate": 2e-05,
      "loss": 0.697452163696289,
      "step": 2690
    },
    {
      "epoch": 0.3424657534246575,
      "grad_norm": 1.2349031899002996,
      "learning_rate": 2e-05,
      "loss": 0.6839765548706055,
      "step": 2700
    },
    {
      "epoch": 0.3437341451040081,
      "grad_norm": 2.040544888336507,
      "learning_rate": 2e-05,
      "loss": 0.7359745025634765,
      "step": 2710
    },
    {
      "epoch": 0.3450025367833587,
      "grad_norm": 3.3977839991338086,
      "learning_rate": 2e-05,
      "loss": 0.6973296165466308,
      "step": 2720
    },
    {
      "epoch": 0.3462709284627093,
      "grad_norm": 1.1566932252315703,
      "learning_rate": 2e-05,
      "loss": 0.6969756126403809,
      "step": 2730
    },
    {
      "epoch": 0.34753932014205985,
      "grad_norm": 1.7500352429084352,
      "learning_rate": 2e-05,
      "loss": 0.7445779323577881,
      "step": 2740
    },
    {
      "epoch": 0.34880771182141046,
      "grad_norm": 1.3962533271825566,
      "learning_rate": 2e-05,
      "loss": 0.7302707672119141,
      "step": 2750
    },
    {
      "epoch": 0.350076103500761,
      "grad_norm": 1.4091056879529666,
      "learning_rate": 2e-05,
      "loss": 0.6887744903564453,
      "step": 2760
    },
    {
      "epoch": 0.3513444951801116,
      "grad_norm": 1.0985660040527039,
      "learning_rate": 2e-05,
      "loss": 0.7133283138275146,
      "step": 2770
    },
    {
      "epoch": 0.3526128868594622,
      "grad_norm": 1.0176711905357152,
      "learning_rate": 2e-05,
      "loss": 0.709499740600586,
      "step": 2780
    },
    {
      "epoch": 0.3538812785388128,
      "grad_norm": 2.3620044075313453,
      "learning_rate": 2e-05,
      "loss": 0.7214662075042725,
      "step": 2790
    },
    {
      "epoch": 0.35514967021816335,
      "grad_norm": 1.1010350150025432,
      "learning_rate": 2e-05,
      "loss": 0.7206092834472656,
      "step": 2800
    },
    {
      "epoch": 0.35641806189751396,
      "grad_norm": 4.8010833951493534,
      "learning_rate": 2e-05,
      "loss": 0.714292335510254,
      "step": 2810
    },
    {
      "epoch": 0.3576864535768645,
      "grad_norm": 2.2130545715247756,
      "learning_rate": 2e-05,
      "loss": 0.7084404945373535,
      "step": 2820
    },
    {
      "epoch": 0.3589548452562151,
      "grad_norm": 1.392266035838293,
      "learning_rate": 2e-05,
      "loss": 0.7139371395111084,
      "step": 2830
    },
    {
      "epoch": 0.3602232369355657,
      "grad_norm": 1.6544189266019673,
      "learning_rate": 2e-05,
      "loss": 0.7246771812438965,
      "step": 2840
    },
    {
      "epoch": 0.3614916286149163,
      "grad_norm": 1.319071839452812,
      "learning_rate": 2e-05,
      "loss": 0.695408821105957,
      "step": 2850
    },
    {
      "epoch": 0.36276002029426685,
      "grad_norm": 1.3874283258861055,
      "learning_rate": 2e-05,
      "loss": 0.7076705932617188,
      "step": 2860
    },
    {
      "epoch": 0.36402841197361746,
      "grad_norm": 1.2181583628923844,
      "learning_rate": 2e-05,
      "loss": 0.6892782211303711,
      "step": 2870
    },
    {
      "epoch": 0.365296803652968,
      "grad_norm": 1.7252472370456187,
      "learning_rate": 2e-05,
      "loss": 0.6980548858642578,
      "step": 2880
    },
    {
      "epoch": 0.3665651953323186,
      "grad_norm": 21.359238258057147,
      "learning_rate": 2e-05,
      "loss": 0.7311030387878418,
      "step": 2890
    },
    {
      "epoch": 0.3678335870116692,
      "grad_norm": 1.1827443065646854,
      "learning_rate": 2e-05,
      "loss": 0.6969143867492675,
      "step": 2900
    },
    {
      "epoch": 0.3691019786910198,
      "grad_norm": 1.104709618070998,
      "learning_rate": 2e-05,
      "loss": 0.7171304225921631,
      "step": 2910
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 1.0586174879515884,
      "learning_rate": 2e-05,
      "loss": 0.7084929943084717,
      "step": 2920
    },
    {
      "epoch": 0.37163876204972096,
      "grad_norm": 0.886323042785058,
      "learning_rate": 2e-05,
      "loss": 0.699922752380371,
      "step": 2930
    },
    {
      "epoch": 0.3729071537290715,
      "grad_norm": 0.960540365904056,
      "learning_rate": 2e-05,
      "loss": 0.698507833480835,
      "step": 2940
    },
    {
      "epoch": 0.3741755454084221,
      "grad_norm": 1.0762056987544082,
      "learning_rate": 2e-05,
      "loss": 0.6993095397949218,
      "step": 2950
    },
    {
      "epoch": 0.3754439370877727,
      "grad_norm": 1.429150249363788,
      "learning_rate": 2e-05,
      "loss": 0.6787753582000733,
      "step": 2960
    },
    {
      "epoch": 0.3767123287671233,
      "grad_norm": 1.204347391252279,
      "learning_rate": 2e-05,
      "loss": 0.6911252498626709,
      "step": 2970
    },
    {
      "epoch": 0.37798072044647385,
      "grad_norm": 1.3682149056871815,
      "learning_rate": 2e-05,
      "loss": 0.6893972396850586,
      "step": 2980
    },
    {
      "epoch": 0.37924911212582446,
      "grad_norm": 1.3971635605670971,
      "learning_rate": 2e-05,
      "loss": 0.727703857421875,
      "step": 2990
    },
    {
      "epoch": 0.380517503805175,
      "grad_norm": 0.9989880095150363,
      "learning_rate": 2e-05,
      "loss": 0.7104997158050537,
      "step": 3000
    },
    {
      "epoch": 0.38178589548452563,
      "grad_norm": 1.0576531938789513,
      "learning_rate": 2e-05,
      "loss": 0.6870818138122559,
      "step": 3010
    },
    {
      "epoch": 0.3830542871638762,
      "grad_norm": 1.704361684010784,
      "learning_rate": 2e-05,
      "loss": 0.6690594673156738,
      "step": 3020
    },
    {
      "epoch": 0.3843226788432268,
      "grad_norm": 1.0860218958894754,
      "learning_rate": 2e-05,
      "loss": 0.7140944480895997,
      "step": 3030
    },
    {
      "epoch": 0.38559107052257735,
      "grad_norm": 1.401148510321468,
      "learning_rate": 2e-05,
      "loss": 0.7057546138763428,
      "step": 3040
    },
    {
      "epoch": 0.38685946220192796,
      "grad_norm": 1.0830524148096021,
      "learning_rate": 2e-05,
      "loss": 0.7069300651550293,
      "step": 3050
    },
    {
      "epoch": 0.3881278538812785,
      "grad_norm": 3.085768618127829,
      "learning_rate": 2e-05,
      "loss": 0.7066722393035889,
      "step": 3060
    },
    {
      "epoch": 0.38939624556062913,
      "grad_norm": 3.7830498962816033,
      "learning_rate": 2e-05,
      "loss": 0.70030837059021,
      "step": 3070
    },
    {
      "epoch": 0.3906646372399797,
      "grad_norm": 3.047542736077345,
      "learning_rate": 2e-05,
      "loss": 0.701556921005249,
      "step": 3080
    },
    {
      "epoch": 0.3919330289193303,
      "grad_norm": 1.22387931594585,
      "learning_rate": 2e-05,
      "loss": 0.7121144771575928,
      "step": 3090
    },
    {
      "epoch": 0.39320142059868085,
      "grad_norm": 1.6534815407494718,
      "learning_rate": 2e-05,
      "loss": 0.6786773204803467,
      "step": 3100
    },
    {
      "epoch": 0.39446981227803146,
      "grad_norm": 1.7834868069971543,
      "learning_rate": 2e-05,
      "loss": 0.6908323287963867,
      "step": 3110
    },
    {
      "epoch": 0.395738203957382,
      "grad_norm": 1.7597910130147842,
      "learning_rate": 2e-05,
      "loss": 0.6861392498016358,
      "step": 3120
    },
    {
      "epoch": 0.39700659563673263,
      "grad_norm": 5.610653545290705,
      "learning_rate": 2e-05,
      "loss": 0.7183076858520507,
      "step": 3130
    },
    {
      "epoch": 0.3982749873160832,
      "grad_norm": 2.7535489186992637,
      "learning_rate": 2e-05,
      "loss": 0.6924714088439942,
      "step": 3140
    },
    {
      "epoch": 0.3995433789954338,
      "grad_norm": 1.2452752337551964,
      "learning_rate": 2e-05,
      "loss": 0.6809077739715577,
      "step": 3150
    },
    {
      "epoch": 0.40081177067478435,
      "grad_norm": 1.520805162428499,
      "learning_rate": 2e-05,
      "loss": 0.6977185249328614,
      "step": 3160
    },
    {
      "epoch": 0.40208016235413496,
      "grad_norm": 1.3363351805209807,
      "learning_rate": 2e-05,
      "loss": 0.7076436519622803,
      "step": 3170
    },
    {
      "epoch": 0.4033485540334855,
      "grad_norm": 1.497731777212216,
      "learning_rate": 2e-05,
      "loss": 0.6808995723724365,
      "step": 3180
    },
    {
      "epoch": 0.40461694571283613,
      "grad_norm": 0.916553937669652,
      "learning_rate": 2e-05,
      "loss": 0.7141075134277344,
      "step": 3190
    },
    {
      "epoch": 0.4058853373921867,
      "grad_norm": 1.193761872573663,
      "learning_rate": 2e-05,
      "loss": 0.6950091361999512,
      "step": 3200
    },
    {
      "epoch": 0.4071537290715373,
      "grad_norm": 3.078435775537744,
      "learning_rate": 2e-05,
      "loss": 0.7315812587738038,
      "step": 3210
    },
    {
      "epoch": 0.40842212075088785,
      "grad_norm": 1.1398733829418206,
      "learning_rate": 2e-05,
      "loss": 0.6878073692321778,
      "step": 3220
    },
    {
      "epoch": 0.40969051243023846,
      "grad_norm": 2.105776238463147,
      "learning_rate": 2e-05,
      "loss": 0.7109416961669922,
      "step": 3230
    },
    {
      "epoch": 0.410958904109589,
      "grad_norm": 0.9644679217277874,
      "learning_rate": 2e-05,
      "loss": 0.7372840881347656,
      "step": 3240
    },
    {
      "epoch": 0.41222729578893963,
      "grad_norm": 0.9833531396296312,
      "learning_rate": 2e-05,
      "loss": 0.718284797668457,
      "step": 3250
    },
    {
      "epoch": 0.4134956874682902,
      "grad_norm": 0.9908012192647654,
      "learning_rate": 2e-05,
      "loss": 0.7132099151611329,
      "step": 3260
    },
    {
      "epoch": 0.4147640791476408,
      "grad_norm": 1.0433903776964855,
      "learning_rate": 2e-05,
      "loss": 0.6863864421844482,
      "step": 3270
    },
    {
      "epoch": 0.41603247082699135,
      "grad_norm": 1.4720619862527746,
      "learning_rate": 2e-05,
      "loss": 0.7224568843841552,
      "step": 3280
    },
    {
      "epoch": 0.41730086250634196,
      "grad_norm": 0.9912124298589535,
      "learning_rate": 2e-05,
      "loss": 0.723963737487793,
      "step": 3290
    },
    {
      "epoch": 0.4185692541856925,
      "grad_norm": 3.1302052156932403,
      "learning_rate": 2e-05,
      "loss": 0.6804662704467773,
      "step": 3300
    },
    {
      "epoch": 0.41983764586504313,
      "grad_norm": 0.8762048338315703,
      "learning_rate": 2e-05,
      "loss": 0.6845558166503907,
      "step": 3310
    },
    {
      "epoch": 0.4211060375443937,
      "grad_norm": 2.0558689469248272,
      "learning_rate": 2e-05,
      "loss": 0.6893569946289062,
      "step": 3320
    },
    {
      "epoch": 0.4223744292237443,
      "grad_norm": 1.2827465626270889,
      "learning_rate": 2e-05,
      "loss": 0.6941981792449952,
      "step": 3330
    },
    {
      "epoch": 0.42364282090309485,
      "grad_norm": 1.9831179196012176,
      "learning_rate": 2e-05,
      "loss": 0.7198607444763183,
      "step": 3340
    },
    {
      "epoch": 0.42491121258244546,
      "grad_norm": 2.7578050838015957,
      "learning_rate": 2e-05,
      "loss": 0.6877538204193115,
      "step": 3350
    },
    {
      "epoch": 0.426179604261796,
      "grad_norm": 3.516644441107207,
      "learning_rate": 2e-05,
      "loss": 0.6907095909118652,
      "step": 3360
    },
    {
      "epoch": 0.42744799594114663,
      "grad_norm": 1.4756907284142395,
      "learning_rate": 2e-05,
      "loss": 0.6960198879241943,
      "step": 3370
    },
    {
      "epoch": 0.4287163876204972,
      "grad_norm": 0.8465441326246769,
      "learning_rate": 2e-05,
      "loss": 0.697170639038086,
      "step": 3380
    },
    {
      "epoch": 0.4299847792998478,
      "grad_norm": 1.2388530972413112,
      "learning_rate": 2e-05,
      "loss": 0.6809267997741699,
      "step": 3390
    },
    {
      "epoch": 0.43125317097919835,
      "grad_norm": 0.9512103437376884,
      "learning_rate": 2e-05,
      "loss": 0.687342643737793,
      "step": 3400
    },
    {
      "epoch": 0.43252156265854896,
      "grad_norm": 1.864844158042898,
      "learning_rate": 2e-05,
      "loss": 0.6892079830169677,
      "step": 3410
    },
    {
      "epoch": 0.4337899543378995,
      "grad_norm": 1.1895178830895066,
      "learning_rate": 2e-05,
      "loss": 0.662391185760498,
      "step": 3420
    },
    {
      "epoch": 0.43505834601725013,
      "grad_norm": 1.4626997726113433,
      "learning_rate": 2e-05,
      "loss": 0.6921442985534668,
      "step": 3430
    },
    {
      "epoch": 0.4363267376966007,
      "grad_norm": 15.4089446705098,
      "learning_rate": 2e-05,
      "loss": 0.6812989711761475,
      "step": 3440
    },
    {
      "epoch": 0.4375951293759513,
      "grad_norm": 7.769731283362733,
      "learning_rate": 2e-05,
      "loss": 0.6541043758392334,
      "step": 3450
    },
    {
      "epoch": 0.43886352105530185,
      "grad_norm": 2.79000349891637,
      "learning_rate": 2e-05,
      "loss": 0.6726436614990234,
      "step": 3460
    },
    {
      "epoch": 0.44013191273465246,
      "grad_norm": 2.7832146873160704,
      "learning_rate": 2e-05,
      "loss": 0.7052318572998046,
      "step": 3470
    },
    {
      "epoch": 0.441400304414003,
      "grad_norm": 2.0390719467146563,
      "learning_rate": 2e-05,
      "loss": 0.7008838176727294,
      "step": 3480
    },
    {
      "epoch": 0.44266869609335363,
      "grad_norm": 1.0775626726493017,
      "learning_rate": 2e-05,
      "loss": 0.692940092086792,
      "step": 3490
    },
    {
      "epoch": 0.4439370877727042,
      "grad_norm": 18.091179957270356,
      "learning_rate": 2e-05,
      "loss": 0.7115600109100342,
      "step": 3500
    },
    {
      "epoch": 0.4452054794520548,
      "grad_norm": 2.5589526986360167,
      "learning_rate": 2e-05,
      "loss": 0.7172080039978027,
      "step": 3510
    },
    {
      "epoch": 0.44647387113140535,
      "grad_norm": 1.7099885311647005,
      "learning_rate": 2e-05,
      "loss": 0.7292625427246093,
      "step": 3520
    },
    {
      "epoch": 0.44774226281075596,
      "grad_norm": 1.12896476674737,
      "learning_rate": 2e-05,
      "loss": 0.672574806213379,
      "step": 3530
    },
    {
      "epoch": 0.4490106544901065,
      "grad_norm": 1.6853541780022807,
      "learning_rate": 2e-05,
      "loss": 0.7191531181335449,
      "step": 3540
    },
    {
      "epoch": 0.45027904616945713,
      "grad_norm": 1.1972796783226842,
      "learning_rate": 2e-05,
      "loss": 0.6878252029418945,
      "step": 3550
    },
    {
      "epoch": 0.4515474378488077,
      "grad_norm": 2.187666096187072,
      "learning_rate": 2e-05,
      "loss": 0.6909091949462891,
      "step": 3560
    },
    {
      "epoch": 0.4528158295281583,
      "grad_norm": 2.431856344190387,
      "learning_rate": 2e-05,
      "loss": 0.6869796752929688,
      "step": 3570
    },
    {
      "epoch": 0.45408422120750885,
      "grad_norm": 1.1932911475975354,
      "learning_rate": 2e-05,
      "loss": 0.6749141693115235,
      "step": 3580
    },
    {
      "epoch": 0.45535261288685946,
      "grad_norm": 2.8248600194639204,
      "learning_rate": 2e-05,
      "loss": 0.6755324840545655,
      "step": 3590
    },
    {
      "epoch": 0.45662100456621,
      "grad_norm": 1.11098298635846,
      "learning_rate": 2e-05,
      "loss": 0.6984715938568116,
      "step": 3600
    },
    {
      "epoch": 0.45788939624556063,
      "grad_norm": 1.1307532084680774,
      "learning_rate": 2e-05,
      "loss": 0.6917080879211426,
      "step": 3610
    },
    {
      "epoch": 0.4591577879249112,
      "grad_norm": 1.8681572619789442,
      "learning_rate": 2e-05,
      "loss": 0.6871934413909913,
      "step": 3620
    },
    {
      "epoch": 0.4604261796042618,
      "grad_norm": 1.3776313601774413,
      "learning_rate": 2e-05,
      "loss": 0.7022416114807128,
      "step": 3630
    },
    {
      "epoch": 0.46169457128361235,
      "grad_norm": 1.2648969686349127,
      "learning_rate": 2e-05,
      "loss": 0.6842359066009521,
      "step": 3640
    },
    {
      "epoch": 0.46296296296296297,
      "grad_norm": 1.17951582986628,
      "learning_rate": 2e-05,
      "loss": 0.6830984115600586,
      "step": 3650
    },
    {
      "epoch": 0.4642313546423135,
      "grad_norm": 1.4238402401833237,
      "learning_rate": 2e-05,
      "loss": 0.7007670402526855,
      "step": 3660
    },
    {
      "epoch": 0.46549974632166413,
      "grad_norm": 1.778443414101243,
      "learning_rate": 2e-05,
      "loss": 0.708482313156128,
      "step": 3670
    },
    {
      "epoch": 0.4667681380010147,
      "grad_norm": 1.2533939905514058,
      "learning_rate": 2e-05,
      "loss": 0.6858346462249756,
      "step": 3680
    },
    {
      "epoch": 0.4680365296803653,
      "grad_norm": 1.434698822089564,
      "learning_rate": 2e-05,
      "loss": 0.6949050426483154,
      "step": 3690
    },
    {
      "epoch": 0.46930492135971585,
      "grad_norm": 1.1300378549911505,
      "learning_rate": 2e-05,
      "loss": 0.695467472076416,
      "step": 3700
    },
    {
      "epoch": 0.47057331303906647,
      "grad_norm": 1.2657008004698769,
      "learning_rate": 2e-05,
      "loss": 0.7693055152893067,
      "step": 3710
    },
    {
      "epoch": 0.471841704718417,
      "grad_norm": 1.513131398986265,
      "learning_rate": 2e-05,
      "loss": 0.6963194847106934,
      "step": 3720
    },
    {
      "epoch": 0.47311009639776763,
      "grad_norm": 1.5010628822555845,
      "learning_rate": 2e-05,
      "loss": 0.6754077911376953,
      "step": 3730
    },
    {
      "epoch": 0.4743784880771182,
      "grad_norm": 1.0597047725877526,
      "learning_rate": 2e-05,
      "loss": 0.6765080451965332,
      "step": 3740
    },
    {
      "epoch": 0.4756468797564688,
      "grad_norm": 1.742666783513689,
      "learning_rate": 2e-05,
      "loss": 0.6817559719085693,
      "step": 3750
    },
    {
      "epoch": 0.47691527143581935,
      "grad_norm": 0.9898329281126279,
      "learning_rate": 2e-05,
      "loss": 0.6671422958374024,
      "step": 3760
    },
    {
      "epoch": 0.47818366311516997,
      "grad_norm": 1.031453002077121,
      "learning_rate": 2e-05,
      "loss": 0.6615818023681641,
      "step": 3770
    },
    {
      "epoch": 0.4794520547945205,
      "grad_norm": 1.2477349185599542,
      "learning_rate": 2e-05,
      "loss": 0.709809684753418,
      "step": 3780
    },
    {
      "epoch": 0.48072044647387113,
      "grad_norm": 1.039977776190847,
      "learning_rate": 2e-05,
      "loss": 0.6623955726623535,
      "step": 3790
    },
    {
      "epoch": 0.4819888381532217,
      "grad_norm": 1.3914719514401461,
      "learning_rate": 2e-05,
      "loss": 0.6831393718719483,
      "step": 3800
    },
    {
      "epoch": 0.4832572298325723,
      "grad_norm": 2.326804082523956,
      "learning_rate": 2e-05,
      "loss": 0.6469039916992188,
      "step": 3810
    },
    {
      "epoch": 0.48452562151192285,
      "grad_norm": 1.2477279483289905,
      "learning_rate": 2e-05,
      "loss": 0.7009859085083008,
      "step": 3820
    },
    {
      "epoch": 0.48579401319127347,
      "grad_norm": 1.6715755359957445,
      "learning_rate": 2e-05,
      "loss": 0.6580695152282715,
      "step": 3830
    },
    {
      "epoch": 0.487062404870624,
      "grad_norm": 1.5809954796169925,
      "learning_rate": 2e-05,
      "loss": 0.6405568599700928,
      "step": 3840
    },
    {
      "epoch": 0.48833079654997463,
      "grad_norm": 1.4290307302819198,
      "learning_rate": 2e-05,
      "loss": 0.6657804012298584,
      "step": 3850
    },
    {
      "epoch": 0.4895991882293252,
      "grad_norm": 1.1186105628343384,
      "learning_rate": 2e-05,
      "loss": 0.6682124137878418,
      "step": 3860
    },
    {
      "epoch": 0.4908675799086758,
      "grad_norm": 1.147714891977977,
      "learning_rate": 2e-05,
      "loss": 0.6971897125244141,
      "step": 3870
    },
    {
      "epoch": 0.49213597158802636,
      "grad_norm": 2.8795944143187957,
      "learning_rate": 2e-05,
      "loss": 0.7148414611816406,
      "step": 3880
    },
    {
      "epoch": 0.49340436326737697,
      "grad_norm": 1.9596334027841549,
      "learning_rate": 2e-05,
      "loss": 0.6601509094238281,
      "step": 3890
    },
    {
      "epoch": 0.4946727549467275,
      "grad_norm": 3.7300578751636295,
      "learning_rate": 2e-05,
      "loss": 0.7265310287475586,
      "step": 3900
    },
    {
      "epoch": 0.49594114662607813,
      "grad_norm": 1.1722052903351763,
      "learning_rate": 2e-05,
      "loss": 0.6697810173034668,
      "step": 3910
    },
    {
      "epoch": 0.4972095383054287,
      "grad_norm": 2.1097120503670768,
      "learning_rate": 2e-05,
      "loss": 0.6614665985107422,
      "step": 3920
    },
    {
      "epoch": 0.4984779299847793,
      "grad_norm": 1.2412727642087709,
      "learning_rate": 2e-05,
      "loss": 0.6702292442321778,
      "step": 3930
    },
    {
      "epoch": 0.49974632166412986,
      "grad_norm": 2.0000417105795516,
      "learning_rate": 2e-05,
      "loss": 0.7101007461547851,
      "step": 3940
    },
    {
      "epoch": 0.5010147133434805,
      "grad_norm": 2.825280343760477,
      "learning_rate": 2e-05,
      "loss": 0.6956410884857178,
      "step": 3950
    },
    {
      "epoch": 0.502283105022831,
      "grad_norm": 1.4125972178176738,
      "learning_rate": 2e-05,
      "loss": 0.6566802024841308,
      "step": 3960
    },
    {
      "epoch": 0.5035514967021817,
      "grad_norm": 4.617791094462311,
      "learning_rate": 2e-05,
      "loss": 0.7369752407073975,
      "step": 3970
    },
    {
      "epoch": 0.5048198883815322,
      "grad_norm": 3.0694228555094782,
      "learning_rate": 2e-05,
      "loss": 0.6708066463470459,
      "step": 3980
    },
    {
      "epoch": 0.5060882800608828,
      "grad_norm": 1.9369032200929064,
      "learning_rate": 2e-05,
      "loss": 0.6765737056732177,
      "step": 3990
    },
    {
      "epoch": 0.5073566717402334,
      "grad_norm": 1.1699082099872802,
      "learning_rate": 2e-05,
      "loss": 0.7113807678222657,
      "step": 4000
    },
    {
      "epoch": 0.508625063419584,
      "grad_norm": 1.0579681499136815,
      "learning_rate": 2e-05,
      "loss": 0.6752194881439209,
      "step": 4010
    },
    {
      "epoch": 0.5098934550989346,
      "grad_norm": 1.1176206054578501,
      "learning_rate": 2e-05,
      "loss": 0.6735630035400391,
      "step": 4020
    },
    {
      "epoch": 0.5111618467782851,
      "grad_norm": 1.1333954045614174,
      "learning_rate": 2e-05,
      "loss": 0.6953109264373779,
      "step": 4030
    },
    {
      "epoch": 0.5124302384576357,
      "grad_norm": 1.1006915128862755,
      "learning_rate": 2e-05,
      "loss": 0.6966275691986084,
      "step": 4040
    },
    {
      "epoch": 0.5136986301369864,
      "grad_norm": 0.9243412566963863,
      "learning_rate": 2e-05,
      "loss": 0.6951508522033691,
      "step": 4050
    },
    {
      "epoch": 0.5149670218163369,
      "grad_norm": 1.6486033984159483,
      "learning_rate": 2e-05,
      "loss": 0.6822213172912598,
      "step": 4060
    },
    {
      "epoch": 0.5162354134956875,
      "grad_norm": 1.0932345609181877,
      "learning_rate": 2e-05,
      "loss": 0.6890708446502686,
      "step": 4070
    },
    {
      "epoch": 0.517503805175038,
      "grad_norm": 1.9604458163631275,
      "learning_rate": 2e-05,
      "loss": 0.6652935981750489,
      "step": 4080
    },
    {
      "epoch": 0.5187721968543887,
      "grad_norm": 7.677802075046973,
      "learning_rate": 2e-05,
      "loss": 0.6910076141357422,
      "step": 4090
    },
    {
      "epoch": 0.5200405885337392,
      "grad_norm": 1.75918793060002,
      "learning_rate": 2e-05,
      "loss": 0.7019295692443848,
      "step": 4100
    },
    {
      "epoch": 0.5213089802130898,
      "grad_norm": 0.996252292913245,
      "learning_rate": 2e-05,
      "loss": 0.6862443447113037,
      "step": 4110
    },
    {
      "epoch": 0.5225773718924404,
      "grad_norm": 1.0496645094720474,
      "learning_rate": 2e-05,
      "loss": 0.7064913749694824,
      "step": 4120
    },
    {
      "epoch": 0.523845763571791,
      "grad_norm": 1.1578070388742652,
      "learning_rate": 2e-05,
      "loss": 0.6819875240325928,
      "step": 4130
    },
    {
      "epoch": 0.5251141552511416,
      "grad_norm": 1.1807679459864984,
      "learning_rate": 2e-05,
      "loss": 0.687638235092163,
      "step": 4140
    },
    {
      "epoch": 0.5263825469304921,
      "grad_norm": 3.517281329610864,
      "learning_rate": 2e-05,
      "loss": 0.6837367057800293,
      "step": 4150
    },
    {
      "epoch": 0.5276509386098427,
      "grad_norm": 1.4784198654391891,
      "learning_rate": 2e-05,
      "loss": 0.6798507690429687,
      "step": 4160
    },
    {
      "epoch": 0.5289193302891934,
      "grad_norm": 1.3972223830347672,
      "learning_rate": 2e-05,
      "loss": 0.7047917366027832,
      "step": 4170
    },
    {
      "epoch": 0.5301877219685439,
      "grad_norm": 17.23315557570745,
      "learning_rate": 2e-05,
      "loss": 0.7046325206756592,
      "step": 4180
    },
    {
      "epoch": 0.5314561136478945,
      "grad_norm": 1.8351926950782174,
      "learning_rate": 2e-05,
      "loss": 0.6856886863708496,
      "step": 4190
    },
    {
      "epoch": 0.532724505327245,
      "grad_norm": 2.3313387954898905,
      "learning_rate": 2e-05,
      "loss": 0.6624431133270263,
      "step": 4200
    },
    {
      "epoch": 0.5339928970065957,
      "grad_norm": 0.9752211193373607,
      "learning_rate": 2e-05,
      "loss": 0.6685439109802246,
      "step": 4210
    },
    {
      "epoch": 0.5352612886859462,
      "grad_norm": 7.754425935548507,
      "learning_rate": 2e-05,
      "loss": 0.6792545318603516,
      "step": 4220
    },
    {
      "epoch": 0.5365296803652968,
      "grad_norm": 1.1386275349589376,
      "learning_rate": 2e-05,
      "loss": 0.736000394821167,
      "step": 4230
    },
    {
      "epoch": 0.5377980720446474,
      "grad_norm": 1.5239025954398326,
      "learning_rate": 2e-05,
      "loss": 0.6889354705810546,
      "step": 4240
    },
    {
      "epoch": 0.539066463723998,
      "grad_norm": 1.1789351093238623,
      "learning_rate": 2e-05,
      "loss": 0.6724788665771484,
      "step": 4250
    },
    {
      "epoch": 0.5403348554033486,
      "grad_norm": 9.086921190150965,
      "learning_rate": 2e-05,
      "loss": 0.7048792839050293,
      "step": 4260
    },
    {
      "epoch": 0.5416032470826991,
      "grad_norm": 1.5566929497497912,
      "learning_rate": 2e-05,
      "loss": 0.6721823692321778,
      "step": 4270
    },
    {
      "epoch": 0.5428716387620497,
      "grad_norm": 8.26792611310342,
      "learning_rate": 2e-05,
      "loss": 0.6834479331970215,
      "step": 4280
    },
    {
      "epoch": 0.5441400304414004,
      "grad_norm": 8.125627249856743,
      "learning_rate": 2e-05,
      "loss": 0.6844344139099121,
      "step": 4290
    },
    {
      "epoch": 0.5454084221207509,
      "grad_norm": 2.8421738541864143,
      "learning_rate": 2e-05,
      "loss": 0.715084457397461,
      "step": 4300
    },
    {
      "epoch": 0.5466768138001015,
      "grad_norm": 3.025007111192603,
      "learning_rate": 2e-05,
      "loss": 0.6849889755249023,
      "step": 4310
    },
    {
      "epoch": 0.547945205479452,
      "grad_norm": 3.8628316345543796,
      "learning_rate": 2e-05,
      "loss": 0.7173240184783936,
      "step": 4320
    },
    {
      "epoch": 0.5492135971588027,
      "grad_norm": 1.689533549584356,
      "learning_rate": 2e-05,
      "loss": 0.6935465812683106,
      "step": 4330
    },
    {
      "epoch": 0.5504819888381532,
      "grad_norm": 2.90748230369957,
      "learning_rate": 2e-05,
      "loss": 0.6775613307952881,
      "step": 4340
    },
    {
      "epoch": 0.5517503805175038,
      "grad_norm": 4.0170434512741835,
      "learning_rate": 2e-05,
      "loss": 0.6890576839447021,
      "step": 4350
    },
    {
      "epoch": 0.5530187721968544,
      "grad_norm": 4.219311226101338,
      "learning_rate": 2e-05,
      "loss": 0.7097297191619873,
      "step": 4360
    },
    {
      "epoch": 0.554287163876205,
      "grad_norm": 1.2263261936938965,
      "learning_rate": 2e-05,
      "loss": 0.6754806518554688,
      "step": 4370
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 2.5362644904702085,
      "learning_rate": 2e-05,
      "loss": 0.645080280303955,
      "step": 4380
    },
    {
      "epoch": 0.5568239472349061,
      "grad_norm": 2.006976057340254,
      "learning_rate": 2e-05,
      "loss": 0.6952437400817871,
      "step": 4390
    },
    {
      "epoch": 0.5580923389142567,
      "grad_norm": 5.176979581829321,
      "learning_rate": 2e-05,
      "loss": 0.701282262802124,
      "step": 4400
    },
    {
      "epoch": 0.5593607305936074,
      "grad_norm": 4.3367971994356305,
      "learning_rate": 2e-05,
      "loss": 0.691312599182129,
      "step": 4410
    },
    {
      "epoch": 0.5606291222729579,
      "grad_norm": 3.6388186991505744,
      "learning_rate": 2e-05,
      "loss": 0.6821917533874512,
      "step": 4420
    },
    {
      "epoch": 0.5618975139523085,
      "grad_norm": 1.5780164171804072,
      "learning_rate": 2e-05,
      "loss": 0.7005750179290772,
      "step": 4430
    },
    {
      "epoch": 0.563165905631659,
      "grad_norm": 1.7369400904708359,
      "learning_rate": 2e-05,
      "loss": 0.6733996391296386,
      "step": 4440
    },
    {
      "epoch": 0.5644342973110097,
      "grad_norm": 5.708096264427102,
      "learning_rate": 2e-05,
      "loss": 0.7272424697875977,
      "step": 4450
    },
    {
      "epoch": 0.5657026889903602,
      "grad_norm": 3.2502364566959057,
      "learning_rate": 2e-05,
      "loss": 0.7025663375854492,
      "step": 4460
    },
    {
      "epoch": 0.5669710806697108,
      "grad_norm": 2.047117376423799,
      "learning_rate": 2e-05,
      "loss": 0.7046807765960693,
      "step": 4470
    },
    {
      "epoch": 0.5682394723490614,
      "grad_norm": 1.3855275017271353,
      "learning_rate": 2e-05,
      "loss": 0.7135118007659912,
      "step": 4480
    },
    {
      "epoch": 0.569507864028412,
      "grad_norm": 1.5726741076562636,
      "learning_rate": 2e-05,
      "loss": 0.6605092525482178,
      "step": 4490
    },
    {
      "epoch": 0.5707762557077626,
      "grad_norm": 1.1094613553602433,
      "learning_rate": 2e-05,
      "loss": 0.71207275390625,
      "step": 4500
    },
    {
      "epoch": 0.5720446473871131,
      "grad_norm": 1.351059113341295,
      "learning_rate": 2e-05,
      "loss": 0.6823588371276855,
      "step": 4510
    },
    {
      "epoch": 0.5733130390664637,
      "grad_norm": 1.9717574530541782,
      "learning_rate": 2e-05,
      "loss": 0.7126172065734864,
      "step": 4520
    },
    {
      "epoch": 0.5745814307458144,
      "grad_norm": 1.7701746741423257,
      "learning_rate": 2e-05,
      "loss": 0.6761096000671387,
      "step": 4530
    },
    {
      "epoch": 0.5758498224251649,
      "grad_norm": 3.4864687935860625,
      "learning_rate": 2e-05,
      "loss": 0.677631139755249,
      "step": 4540
    },
    {
      "epoch": 0.5771182141045155,
      "grad_norm": 1.3031740813126822,
      "learning_rate": 2e-05,
      "loss": 0.6485799789428711,
      "step": 4550
    },
    {
      "epoch": 0.578386605783866,
      "grad_norm": 1.203505453439311,
      "learning_rate": 2e-05,
      "loss": 0.6640719413757324,
      "step": 4560
    },
    {
      "epoch": 0.5796549974632167,
      "grad_norm": 1.4475259273352683,
      "learning_rate": 2e-05,
      "loss": 0.6618621349334717,
      "step": 4570
    },
    {
      "epoch": 0.5809233891425672,
      "grad_norm": 6.363407639764493,
      "learning_rate": 2e-05,
      "loss": 0.6911416530609131,
      "step": 4580
    },
    {
      "epoch": 0.5821917808219178,
      "grad_norm": 1.144449181032623,
      "learning_rate": 2e-05,
      "loss": 0.6547828674316406,
      "step": 4590
    },
    {
      "epoch": 0.5834601725012684,
      "grad_norm": 1.216319720807365,
      "learning_rate": 2e-05,
      "loss": 0.6870989799499512,
      "step": 4600
    },
    {
      "epoch": 0.584728564180619,
      "grad_norm": 1.129381667164184,
      "learning_rate": 2e-05,
      "loss": 0.6711385726928711,
      "step": 4610
    },
    {
      "epoch": 0.5859969558599696,
      "grad_norm": 3.1845439387268857,
      "learning_rate": 2e-05,
      "loss": 0.7018142223358155,
      "step": 4620
    },
    {
      "epoch": 0.5872653475393201,
      "grad_norm": 1.8144628126767592,
      "learning_rate": 2e-05,
      "loss": 0.6543365478515625,
      "step": 4630
    },
    {
      "epoch": 0.5885337392186707,
      "grad_norm": 1.3348537143873103,
      "learning_rate": 2e-05,
      "loss": 0.6841076850891114,
      "step": 4640
    },
    {
      "epoch": 0.5898021308980214,
      "grad_norm": 1.6291988164091082,
      "learning_rate": 2e-05,
      "loss": 0.6448442459106445,
      "step": 4650
    },
    {
      "epoch": 0.5910705225773719,
      "grad_norm": 1.9554526342878682,
      "learning_rate": 2e-05,
      "loss": 0.7052442550659179,
      "step": 4660
    },
    {
      "epoch": 0.5923389142567225,
      "grad_norm": 0.953998763331625,
      "learning_rate": 2e-05,
      "loss": 0.6536243438720704,
      "step": 4670
    },
    {
      "epoch": 0.593607305936073,
      "grad_norm": 1.226989803281022,
      "learning_rate": 2e-05,
      "loss": 0.7327777862548828,
      "step": 4680
    },
    {
      "epoch": 0.5948756976154237,
      "grad_norm": 24.15656132623872,
      "learning_rate": 2e-05,
      "loss": 0.7085799694061279,
      "step": 4690
    },
    {
      "epoch": 0.5961440892947742,
      "grad_norm": 1.633176277039922,
      "learning_rate": 2e-05,
      "loss": 0.6776537895202637,
      "step": 4700
    },
    {
      "epoch": 0.5974124809741248,
      "grad_norm": 1.5110153105799853,
      "learning_rate": 2e-05,
      "loss": 0.6916415691375732,
      "step": 4710
    },
    {
      "epoch": 0.5986808726534754,
      "grad_norm": 1.0634358859385251,
      "learning_rate": 2e-05,
      "loss": 0.6873381614685059,
      "step": 4720
    },
    {
      "epoch": 0.599949264332826,
      "grad_norm": 1.0870104320306402,
      "learning_rate": 2e-05,
      "loss": 0.6644130229949952,
      "step": 4730
    },
    {
      "epoch": 0.6012176560121766,
      "grad_norm": 0.8899051013550261,
      "learning_rate": 2e-05,
      "loss": 0.6844129085540771,
      "step": 4740
    },
    {
      "epoch": 0.6024860476915271,
      "grad_norm": 2.1471713854060446,
      "learning_rate": 2e-05,
      "loss": 0.6981614589691162,
      "step": 4750
    },
    {
      "epoch": 0.6037544393708777,
      "grad_norm": 1.1224910420807845,
      "learning_rate": 2e-05,
      "loss": 0.6662663459777832,
      "step": 4760
    },
    {
      "epoch": 0.6050228310502284,
      "grad_norm": 1.2468076373451553,
      "learning_rate": 2e-05,
      "loss": 0.6670088768005371,
      "step": 4770
    },
    {
      "epoch": 0.6062912227295789,
      "grad_norm": 1.350094821122234,
      "learning_rate": 2e-05,
      "loss": 0.6816661834716797,
      "step": 4780
    },
    {
      "epoch": 0.6075596144089295,
      "grad_norm": 2.253916078221231,
      "learning_rate": 2e-05,
      "loss": 0.7326220512390137,
      "step": 4790
    },
    {
      "epoch": 0.60882800608828,
      "grad_norm": 2.976444370496611,
      "learning_rate": 2e-05,
      "loss": 0.694272518157959,
      "step": 4800
    },
    {
      "epoch": 0.6100963977676307,
      "grad_norm": 1.612884727635378,
      "learning_rate": 2e-05,
      "loss": 0.6723672389984131,
      "step": 4810
    },
    {
      "epoch": 0.6113647894469812,
      "grad_norm": 14.875686984094994,
      "learning_rate": 2e-05,
      "loss": 0.6641634941101074,
      "step": 4820
    },
    {
      "epoch": 0.6126331811263318,
      "grad_norm": 1.5296622460918816,
      "learning_rate": 2e-05,
      "loss": 0.6817727088928223,
      "step": 4830
    },
    {
      "epoch": 0.6139015728056824,
      "grad_norm": 3.4684255276589884,
      "learning_rate": 2e-05,
      "loss": 0.6922375679016113,
      "step": 4840
    },
    {
      "epoch": 0.615169964485033,
      "grad_norm": 1.3298325605943906,
      "learning_rate": 2e-05,
      "loss": 0.6717410564422608,
      "step": 4850
    },
    {
      "epoch": 0.6164383561643836,
      "grad_norm": 1.312443616754889,
      "learning_rate": 2e-05,
      "loss": 0.7139492988586426,
      "step": 4860
    },
    {
      "epoch": 0.6177067478437341,
      "grad_norm": 6.070858047608295,
      "learning_rate": 2e-05,
      "loss": 0.667232894897461,
      "step": 4870
    },
    {
      "epoch": 0.6189751395230847,
      "grad_norm": 1.6367824483707443,
      "learning_rate": 2e-05,
      "loss": 0.6844718933105469,
      "step": 4880
    },
    {
      "epoch": 0.6202435312024354,
      "grad_norm": 1.3920003075389475,
      "learning_rate": 2e-05,
      "loss": 0.7157072067260742,
      "step": 4890
    },
    {
      "epoch": 0.6215119228817859,
      "grad_norm": 1.7570734473626832,
      "learning_rate": 2e-05,
      "loss": 0.679513692855835,
      "step": 4900
    },
    {
      "epoch": 0.6227803145611365,
      "grad_norm": 1.1895461124330857,
      "learning_rate": 2e-05,
      "loss": 0.658949613571167,
      "step": 4910
    },
    {
      "epoch": 0.624048706240487,
      "grad_norm": 0.8258801455447918,
      "learning_rate": 2e-05,
      "loss": 0.6867888450622559,
      "step": 4920
    },
    {
      "epoch": 0.6253170979198377,
      "grad_norm": 0.8939437479317008,
      "learning_rate": 2e-05,
      "loss": 0.671229076385498,
      "step": 4930
    },
    {
      "epoch": 0.6265854895991883,
      "grad_norm": 1.0206016302390069,
      "learning_rate": 2e-05,
      "loss": 0.6558897972106934,
      "step": 4940
    },
    {
      "epoch": 0.6278538812785388,
      "grad_norm": 2.6527501622634353,
      "learning_rate": 2e-05,
      "loss": 0.7233300685882569,
      "step": 4950
    },
    {
      "epoch": 0.6291222729578894,
      "grad_norm": 1.1705886205838638,
      "learning_rate": 2e-05,
      "loss": 0.6886804103851318,
      "step": 4960
    },
    {
      "epoch": 0.63039066463724,
      "grad_norm": 5.581396061720732,
      "learning_rate": 2e-05,
      "loss": 0.6646123409271241,
      "step": 4970
    },
    {
      "epoch": 0.6316590563165906,
      "grad_norm": 0.9781867843425378,
      "learning_rate": 2e-05,
      "loss": 0.7057945251464843,
      "step": 4980
    },
    {
      "epoch": 0.6329274479959411,
      "grad_norm": 1.367117833665198,
      "learning_rate": 2e-05,
      "loss": 0.6685174465179443,
      "step": 4990
    },
    {
      "epoch": 0.6341958396752917,
      "grad_norm": 1.568832858316303,
      "learning_rate": 2e-05,
      "loss": 0.6930606842041016,
      "step": 5000
    },
    {
      "epoch": 0.6354642313546424,
      "grad_norm": 5.0736860015214775,
      "learning_rate": 2e-05,
      "loss": 0.6832860469818115,
      "step": 5010
    },
    {
      "epoch": 0.6367326230339929,
      "grad_norm": 1.150482675497788,
      "learning_rate": 2e-05,
      "loss": 0.6540486812591553,
      "step": 5020
    },
    {
      "epoch": 0.6380010147133435,
      "grad_norm": 1.116635529204696,
      "learning_rate": 2e-05,
      "loss": 0.6667252540588379,
      "step": 5030
    },
    {
      "epoch": 0.639269406392694,
      "grad_norm": 1.4249408859120571,
      "learning_rate": 2e-05,
      "loss": 0.6510902404785156,
      "step": 5040
    },
    {
      "epoch": 0.6405377980720447,
      "grad_norm": 1.3865013546325862,
      "learning_rate": 2e-05,
      "loss": 0.6763072490692139,
      "step": 5050
    },
    {
      "epoch": 0.6418061897513953,
      "grad_norm": 5.499428752017872,
      "learning_rate": 2e-05,
      "loss": 0.6979012966156006,
      "step": 5060
    },
    {
      "epoch": 0.6430745814307458,
      "grad_norm": 1.1307475408050165,
      "learning_rate": 2e-05,
      "loss": 0.690005111694336,
      "step": 5070
    },
    {
      "epoch": 0.6443429731100964,
      "grad_norm": 1.0964079448997324,
      "learning_rate": 2e-05,
      "loss": 0.6874360084533692,
      "step": 5080
    },
    {
      "epoch": 0.645611364789447,
      "grad_norm": 1.2963478903565528,
      "learning_rate": 2e-05,
      "loss": 0.687739086151123,
      "step": 5090
    },
    {
      "epoch": 0.6468797564687976,
      "grad_norm": 1.1972986331441962,
      "learning_rate": 2e-05,
      "loss": 0.709494686126709,
      "step": 5100
    },
    {
      "epoch": 0.6481481481481481,
      "grad_norm": 5.2200938804409684,
      "learning_rate": 2e-05,
      "loss": 0.6661120414733886,
      "step": 5110
    },
    {
      "epoch": 0.6494165398274987,
      "grad_norm": 1.9150807483090664,
      "learning_rate": 2e-05,
      "loss": 0.6963963508605957,
      "step": 5120
    },
    {
      "epoch": 0.6506849315068494,
      "grad_norm": 1.009028360307343,
      "learning_rate": 2e-05,
      "loss": 0.692938756942749,
      "step": 5130
    },
    {
      "epoch": 0.6519533231861999,
      "grad_norm": 2.415145405396824,
      "learning_rate": 2e-05,
      "loss": 0.7020602703094483,
      "step": 5140
    },
    {
      "epoch": 0.6532217148655505,
      "grad_norm": 1.8271724773310307,
      "learning_rate": 2e-05,
      "loss": 0.688006591796875,
      "step": 5150
    },
    {
      "epoch": 0.654490106544901,
      "grad_norm": 1.5127454073424815,
      "learning_rate": 2e-05,
      "loss": 0.7013001918792725,
      "step": 5160
    },
    {
      "epoch": 0.6557584982242517,
      "grad_norm": 0.9348550839174586,
      "learning_rate": 2e-05,
      "loss": 0.6823058128356934,
      "step": 5170
    },
    {
      "epoch": 0.6570268899036023,
      "grad_norm": 3.376017382348284,
      "learning_rate": 2e-05,
      "loss": 0.6459805488586425,
      "step": 5180
    },
    {
      "epoch": 0.6582952815829528,
      "grad_norm": 0.990310747840206,
      "learning_rate": 2e-05,
      "loss": 0.6828865051269531,
      "step": 5190
    },
    {
      "epoch": 0.6595636732623034,
      "grad_norm": 0.9826772310251697,
      "learning_rate": 2e-05,
      "loss": 0.6518405914306641,
      "step": 5200
    },
    {
      "epoch": 0.660832064941654,
      "grad_norm": 1.704675141244084,
      "learning_rate": 2e-05,
      "loss": 0.6593197822570801,
      "step": 5210
    },
    {
      "epoch": 0.6621004566210046,
      "grad_norm": 1.1661397575269763,
      "learning_rate": 2e-05,
      "loss": 0.6678568840026855,
      "step": 5220
    },
    {
      "epoch": 0.6633688483003551,
      "grad_norm": 6.271302371052588,
      "learning_rate": 2e-05,
      "loss": 0.6900056362152099,
      "step": 5230
    },
    {
      "epoch": 0.6646372399797057,
      "grad_norm": 1.1052176432903738,
      "learning_rate": 2e-05,
      "loss": 0.7045561790466308,
      "step": 5240
    },
    {
      "epoch": 0.6659056316590564,
      "grad_norm": 0.9674064303813003,
      "learning_rate": 2e-05,
      "loss": 0.6891006469726563,
      "step": 5250
    },
    {
      "epoch": 0.6671740233384069,
      "grad_norm": 1.2266507838357632,
      "learning_rate": 2e-05,
      "loss": 0.6916210174560546,
      "step": 5260
    },
    {
      "epoch": 0.6684424150177575,
      "grad_norm": 1.0221703491240415,
      "learning_rate": 2e-05,
      "loss": 0.6675290584564209,
      "step": 5270
    },
    {
      "epoch": 0.669710806697108,
      "grad_norm": 3.2539296956204824,
      "learning_rate": 2e-05,
      "loss": 0.6716821670532227,
      "step": 5280
    },
    {
      "epoch": 0.6709791983764587,
      "grad_norm": 1.0124582951069105,
      "learning_rate": 2e-05,
      "loss": 0.6884188652038574,
      "step": 5290
    },
    {
      "epoch": 0.6722475900558093,
      "grad_norm": 0.8654753848059636,
      "learning_rate": 2e-05,
      "loss": 0.6870044708251953,
      "step": 5300
    },
    {
      "epoch": 0.6735159817351598,
      "grad_norm": 1.6701035955772598,
      "learning_rate": 2e-05,
      "loss": 0.6423444271087646,
      "step": 5310
    },
    {
      "epoch": 0.6747843734145104,
      "grad_norm": 0.9640894047657678,
      "learning_rate": 2e-05,
      "loss": 0.6536365509033203,
      "step": 5320
    },
    {
      "epoch": 0.676052765093861,
      "grad_norm": 1.187753217003705,
      "learning_rate": 2e-05,
      "loss": 0.6517406463623047,
      "step": 5330
    },
    {
      "epoch": 0.6773211567732116,
      "grad_norm": 1.3258589248438108,
      "learning_rate": 2e-05,
      "loss": 0.6949641227722168,
      "step": 5340
    },
    {
      "epoch": 0.6785895484525621,
      "grad_norm": 1.22025073967006,
      "learning_rate": 1.999194520388485e-05,
      "loss": 0.709574031829834,
      "step": 5350
    },
    {
      "epoch": 0.6798579401319127,
      "grad_norm": 1.2848215133141674,
      "learning_rate": 1.9959247303157763e-05,
      "loss": 0.6613364219665527,
      "step": 5360
    },
    {
      "epoch": 0.6811263318112634,
      "grad_norm": 1.429374604283802,
      "learning_rate": 1.990149424459817e-05,
      "loss": 0.6761489868164062,
      "step": 5370
    },
    {
      "epoch": 0.6823947234906139,
      "grad_norm": 1.5988897061277771,
      "learning_rate": 1.9818847538648688e-05,
      "loss": 0.6596675872802734,
      "step": 5380
    },
    {
      "epoch": 0.6836631151699645,
      "grad_norm": 1.3947399786149248,
      "learning_rate": 1.971153831256721e-05,
      "loss": 0.6783432006835938,
      "step": 5390
    },
    {
      "epoch": 0.684931506849315,
      "grad_norm": 2.0453755438074137,
      "learning_rate": 1.9579866664063435e-05,
      "loss": 0.6344878196716308,
      "step": 5400
    },
    {
      "epoch": 0.6861998985286657,
      "grad_norm": 2.3712249095706746,
      "learning_rate": 1.9424200822054782e-05,
      "loss": 0.6518111228942871,
      "step": 5410
    },
    {
      "epoch": 0.6874682902080163,
      "grad_norm": 2.5905617956082523,
      "learning_rate": 1.9244976116888626e-05,
      "loss": 0.6491889953613281,
      "step": 5420
    },
    {
      "epoch": 0.6887366818873668,
      "grad_norm": 1.8499902870344829,
      "learning_rate": 1.904269376291071e-05,
      "loss": 0.6815225124359131,
      "step": 5430
    },
    {
      "epoch": 0.6900050735667174,
      "grad_norm": 1.9833413632934562,
      "learning_rate": 1.88179194567844e-05,
      "loss": 0.6821569442749024,
      "step": 5440
    },
    {
      "epoch": 0.691273465246068,
      "grad_norm": 1.8536405474230584,
      "learning_rate": 1.8571281795480632e-05,
      "loss": 0.6728893280029297,
      "step": 5450
    },
    {
      "epoch": 0.6925418569254186,
      "grad_norm": 1.2929086769576823,
      "learning_rate": 1.8303470518362808e-05,
      "loss": 0.6345015525817871,
      "step": 5460
    },
    {
      "epoch": 0.6938102486047691,
      "grad_norm": 0.9052943676071513,
      "learning_rate": 1.801523457828271e-05,
      "loss": 0.6461944580078125,
      "step": 5470
    },
    {
      "epoch": 0.6950786402841197,
      "grad_norm": 1.709022042633873,
      "learning_rate": 1.7707380047081796e-05,
      "loss": 0.6566840648651123,
      "step": 5480
    },
    {
      "epoch": 0.6963470319634704,
      "grad_norm": 1.6896173650660717,
      "learning_rate": 1.7380767861355252e-05,
      "loss": 0.7078164100646973,
      "step": 5490
    },
    {
      "epoch": 0.6976154236428209,
      "grad_norm": 2.4017575721056454,
      "learning_rate": 1.703631141478299e-05,
      "loss": 0.6496434688568116,
      "step": 5500
    },
    {
      "epoch": 0.6988838153221715,
      "grad_norm": 1.3078714120253687,
      "learning_rate": 1.6674974003760705e-05,
      "loss": 0.6476611614227294,
      "step": 5510
    },
    {
      "epoch": 0.700152207001522,
      "grad_norm": 1.0265645089433744,
      "learning_rate": 1.6297766133474556e-05,
      "loss": 0.6869290828704834,
      "step": 5520
    },
    {
      "epoch": 0.7014205986808727,
      "grad_norm": 1.0533999833724916,
      "learning_rate": 1.5905742691953186e-05,
      "loss": 0.6464869976043701,
      "step": 5530
    },
    {
      "epoch": 0.7026889903602233,
      "grad_norm": 1.0641928720205718,
      "learning_rate": 1.55e-05,
      "loss": 0.6754951477050781,
      "step": 5540
    },
    {
      "epoch": 0.7039573820395738,
      "grad_norm": 0.9658246422845503,
      "learning_rate": 1.5081672745255862e-05,
      "loss": 0.6540566444396972,
      "step": 5550
    },
    {
      "epoch": 0.7052257737189244,
      "grad_norm": 0.9303411714613762,
      "learning_rate": 1.4651930808966206e-05,
      "loss": 0.6702482223510742,
      "step": 5560
    },
    {
      "epoch": 0.706494165398275,
      "grad_norm": 1.9119266152294763,
      "learning_rate": 1.4211975994326848e-05,
      "loss": 0.6635209560394287,
      "step": 5570
    },
    {
      "epoch": 0.7077625570776256,
      "grad_norm": 0.993900034058341,
      "learning_rate": 1.376303866555779e-05,
      "loss": 0.6339982032775879,
      "step": 5580
    },
    {
      "epoch": 0.7090309487569761,
      "grad_norm": 1.0136725201352181,
      "learning_rate": 1.3306374307104164e-05,
      "loss": 0.6624653816223145,
      "step": 5590
    },
    {
      "epoch": 0.7102993404363267,
      "grad_norm": 0.7828238983133698,
      "learning_rate": 1.2843260012586718e-05,
      "loss": 0.675888204574585,
      "step": 5600
    },
    {
      "epoch": 0.7115677321156774,
      "grad_norm": 0.8511866645286336,
      "learning_rate": 1.2374990913320726e-05,
      "loss": 0.6653191566467285,
      "step": 5610
    },
    {
      "epoch": 0.7128361237950279,
      "grad_norm": 1.135902245440002,
      "learning_rate": 1.1902876556391247e-05,
      "loss": 0.6719844818115235,
      "step": 5620
    },
    {
      "epoch": 0.7141045154743785,
      "grad_norm": 1.1103387021188629,
      "learning_rate": 1.1428237242413685e-05,
      "loss": 0.6222578048706054,
      "step": 5630
    },
    {
      "epoch": 0.715372907153729,
      "grad_norm": 1.003512178862491,
      "learning_rate": 1.095240033322132e-05,
      "loss": 0.6572049617767334,
      "step": 5640
    },
    {
      "epoch": 0.7166412988330797,
      "grad_norm": 3.4720054588268647,
      "learning_rate": 1.047669653980572e-05,
      "loss": 0.6619221687316894,
      "step": 5650
    },
    {
      "epoch": 0.7179096905124303,
      "grad_norm": 1.0655700875623892,
      "learning_rate": 1.0002456200890902e-05,
      "loss": 0.6615418910980224,
      "step": 5660
    },
    {
      "epoch": 0.7191780821917808,
      "grad_norm": 1.0423235290834962,
      "learning_rate": 9.531005562548668e-06,
      "loss": 0.6552172660827636,
      "step": 5670
    },
    {
      "epoch": 0.7204464738711314,
      "grad_norm": 1.261909703904696,
      "learning_rate": 9.06366306925925e-06,
      "loss": 0.6493484497070312,
      "step": 5680
    },
    {
      "epoch": 0.721714865550482,
      "grad_norm": 0.994101900474543,
      "learning_rate": 8.601735676789686e-06,
      "loss": 0.6309298515319824,
      "step": 5690
    },
    {
      "epoch": 0.7229832572298326,
      "grad_norm": 1.3079491980180147,
      "learning_rate": 8.14651519720118e-06,
      "loss": 0.63365797996521,
      "step": 5700
    },
    {
      "epoch": 0.7242516489091831,
      "grad_norm": 4.076710598420271,
      "learning_rate": 7.699274686206866e-06,
      "loss": 0.6313989162445068,
      "step": 5710
    },
    {
      "epoch": 0.7255200405885337,
      "grad_norm": 1.0560963223608322,
      "learning_rate": 7.261264882983024e-06,
      "loss": 0.6326074123382568,
      "step": 5720
    },
    {
      "epoch": 0.7267884322678844,
      "grad_norm": 5.367859446648986,
      "learning_rate": 6.8337107123900386e-06,
      "loss": 0.6793523788452148,
      "step": 5730
    },
    {
      "epoch": 0.7280568239472349,
      "grad_norm": 1.3179805962100755,
      "learning_rate": 6.417807859384881e-06,
      "loss": 0.644274091720581,
      "step": 5740
    },
    {
      "epoch": 0.7293252156265855,
      "grad_norm": 1.4722671848572835,
      "learning_rate": 6.014719425205009e-06,
      "loss": 0.6411468029022217,
      "step": 5750
    },
    {
      "epoch": 0.730593607305936,
      "grad_norm": 1.4388894679589062,
      "learning_rate": 5.6255726746749275e-06,
      "loss": 0.6132498741149902,
      "step": 5760
    },
    {
      "epoch": 0.7318619989852867,
      "grad_norm": 1.039120091854845,
      "learning_rate": 5.251455883731728e-06,
      "loss": 0.6519311428070068,
      "step": 5770
    },
    {
      "epoch": 0.7331303906646373,
      "grad_norm": 0.8827993538543202,
      "learning_rate": 4.893415295985812e-06,
      "loss": 0.6502262115478515,
      "step": 5780
    },
    {
      "epoch": 0.7343987823439878,
      "grad_norm": 2.3658211155191338,
      "learning_rate": 4.552452196827885e-06,
      "loss": 0.6125197410583496,
      "step": 5790
    },
    {
      "epoch": 0.7356671740233384,
      "grad_norm": 1.2092225742684777,
      "learning_rate": 4.229520113264785e-06,
      "loss": 0.6368597030639649,
      "step": 5800
    },
    {
      "epoch": 0.736935565702689,
      "grad_norm": 0.8474949458209987,
      "learning_rate": 3.925522147314915e-06,
      "loss": 0.6502898216247559,
      "step": 5810
    },
    {
      "epoch": 0.7382039573820396,
      "grad_norm": 18.081845705968906,
      "learning_rate": 3.641308450420675e-06,
      "loss": 0.6471766471862793,
      "step": 5820
    },
    {
      "epoch": 0.7394723490613901,
      "grad_norm": 1.4876005168207684,
      "learning_rate": 3.37767384594087e-06,
      "loss": 0.6519950866699219,
      "step": 5830
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.7669750133736558,
      "learning_rate": 3.1353556063719357e-06,
      "loss": 0.680915355682373,
      "step": 5840
    },
    {
      "epoch": 0.7420091324200914,
      "grad_norm": 1.5200232782410792,
      "learning_rate": 2.9150313915141474e-06,
      "loss": 0.6112856864929199,
      "step": 5850
    },
    {
      "epoch": 0.7432775240994419,
      "grad_norm": 1.064276645635058,
      "learning_rate": 2.717317353348891e-06,
      "loss": 0.629461669921875,
      "step": 5860
    },
    {
      "epoch": 0.7445459157787925,
      "grad_norm": 0.8033814702188683,
      "learning_rate": 2.5427664129268253e-06,
      "loss": 0.6689444541931152,
      "step": 5870
    },
    {
      "epoch": 0.745814307458143,
      "grad_norm": 1.3523943764460118,
      "learning_rate": 2.391866714085717e-06,
      "loss": 0.6370697021484375,
      "step": 5880
    },
    {
      "epoch": 0.7470826991374937,
      "grad_norm": 1.486680810206009,
      "learning_rate": 2.2650402583222753e-06,
      "loss": 0.6321832180023194,
      "step": 5890
    },
    {
      "epoch": 0.7483510908168443,
      "grad_norm": 0.977879650591291,
      "learning_rate": 2.1626417246356398e-06,
      "loss": 0.6477483749389649,
      "step": 5900
    },
    {
      "epoch": 0.7496194824961948,
      "grad_norm": 1.2083852912305637,
      "learning_rate": 2.084957477642894e-06,
      "loss": 0.6267468929290771,
      "step": 5910
    },
    {
      "epoch": 0.7508878741755454,
      "grad_norm": 1.1652000362254096,
      "learning_rate": 2.0322047667405147e-06,
      "loss": 0.6213836669921875,
      "step": 5920
    },
    {
      "epoch": 0.752156265854896,
      "grad_norm": 1.157234756002711,
      "learning_rate": 2.0045311185513342e-06,
      "loss": 0.6436711311340332,
      "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
}