{
  "best_metric": 0.034325700253248215,
  "best_model_checkpoint": "./cifar10_outputs/checkpoint-53130",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 53130,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 2.814899444580078,
      "learning_rate": 1.9996235648409565e-05,
      "loss": 2.327,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.6160950660705566,
      "learning_rate": 1.9992471296819125e-05,
      "loss": 2.2852,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.41660475730896,
      "learning_rate": 1.9988706945228688e-05,
      "loss": 2.2588,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.0623512268066406,
      "learning_rate": 1.9984942593638247e-05,
      "loss": 2.2605,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.4211559295654297,
      "learning_rate": 1.998117824204781e-05,
      "loss": 2.2211,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.8450746536254883,
      "learning_rate": 1.997741389045737e-05,
      "loss": 2.199,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.4961893558502197,
      "learning_rate": 1.9973649538866934e-05,
      "loss": 2.1407,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.477571964263916,
      "learning_rate": 1.9969885187276493e-05,
      "loss": 2.1167,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.67600679397583,
      "learning_rate": 1.9966120835686053e-05,
      "loss": 2.1076,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.6568191051483154,
      "learning_rate": 1.9962356484095616e-05,
      "loss": 2.0177,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.4204459190368652,
      "learning_rate": 1.9958592132505176e-05,
      "loss": 2.0285,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.031322717666626,
      "learning_rate": 1.995482778091474e-05,
      "loss": 1.9454,
      "step": 120
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.2625372409820557,
      "learning_rate": 1.99510634293243e-05,
      "loss": 1.8982,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.081059217453003,
      "learning_rate": 1.9947299077733862e-05,
      "loss": 1.8509,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.556724786758423,
      "learning_rate": 1.9943534726143422e-05,
      "loss": 1.8542,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.684180498123169,
      "learning_rate": 1.9939770374552985e-05,
      "loss": 1.7885,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.279416084289551,
      "learning_rate": 1.9936006022962545e-05,
      "loss": 1.7249,
      "step": 170
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.491975784301758,
      "learning_rate": 1.9932241671372108e-05,
      "loss": 1.6934,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.5165696144104,
      "learning_rate": 1.992847731978167e-05,
      "loss": 1.692,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.120129585266113,
      "learning_rate": 1.992471296819123e-05,
      "loss": 1.5837,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.18604040145874,
      "learning_rate": 1.9920948616600794e-05,
      "loss": 1.5069,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.8111982345581055,
      "learning_rate": 1.9917184265010354e-05,
      "loss": 1.6036,
      "step": 220
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.115821123123169,
      "learning_rate": 1.9913419913419917e-05,
      "loss": 1.4801,
      "step": 230
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.2015562057495117,
      "learning_rate": 1.9909655561829477e-05,
      "loss": 1.4055,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.093550682067871,
      "learning_rate": 1.990589121023904e-05,
      "loss": 1.4003,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.127725124359131,
      "learning_rate": 1.99021268586486e-05,
      "loss": 1.3889,
      "step": 260
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.6554887294769287,
      "learning_rate": 1.989836250705816e-05,
      "loss": 1.3906,
      "step": 270
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.2247841358184814,
      "learning_rate": 1.9894598155467723e-05,
      "loss": 1.2236,
      "step": 280
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.622257232666016,
      "learning_rate": 1.9890833803877282e-05,
      "loss": 1.2971,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.8183791637420654,
      "learning_rate": 1.9887069452286845e-05,
      "loss": 1.1908,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.8954896926879883,
      "learning_rate": 1.9883305100696405e-05,
      "loss": 1.2147,
      "step": 310
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.090787410736084,
      "learning_rate": 1.987954074910597e-05,
      "loss": 1.2619,
      "step": 320
    },
    {
      "epoch": 0.06,
      "grad_norm": 6.540658473968506,
      "learning_rate": 1.9875776397515528e-05,
      "loss": 1.1826,
      "step": 330
    },
    {
      "epoch": 0.06,
      "grad_norm": 5.227759838104248,
      "learning_rate": 1.987201204592509e-05,
      "loss": 1.1198,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 4.01509428024292,
      "learning_rate": 1.986824769433465e-05,
      "loss": 1.1363,
      "step": 350
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.1885578632354736,
      "learning_rate": 1.9864483342744214e-05,
      "loss": 1.1362,
      "step": 360
    },
    {
      "epoch": 0.07,
      "grad_norm": 4.961787700653076,
      "learning_rate": 1.9860718991153774e-05,
      "loss": 1.0776,
      "step": 370
    },
    {
      "epoch": 0.07,
      "grad_norm": 4.749518871307373,
      "learning_rate": 1.9856954639563337e-05,
      "loss": 1.1963,
      "step": 380
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.671156167984009,
      "learning_rate": 1.98531902879729e-05,
      "loss": 1.1087,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.822259902954102,
      "learning_rate": 1.984942593638246e-05,
      "loss": 1.0092,
      "step": 400
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.055830955505371,
      "learning_rate": 1.9845661584792023e-05,
      "loss": 1.1057,
      "step": 410
    },
    {
      "epoch": 0.08,
      "grad_norm": 7.269201278686523,
      "learning_rate": 1.9841897233201583e-05,
      "loss": 1.063,
      "step": 420
    },
    {
      "epoch": 0.08,
      "grad_norm": 6.632596969604492,
      "learning_rate": 1.9838132881611146e-05,
      "loss": 0.9014,
      "step": 430
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.082939624786377,
      "learning_rate": 1.9834368530020706e-05,
      "loss": 0.9152,
      "step": 440
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.534853935241699,
      "learning_rate": 1.9830604178430266e-05,
      "loss": 0.9754,
      "step": 450
    },
    {
      "epoch": 0.09,
      "grad_norm": 7.8275532722473145,
      "learning_rate": 1.982683982683983e-05,
      "loss": 0.9595,
      "step": 460
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.591826438903809,
      "learning_rate": 1.982307547524939e-05,
      "loss": 0.9536,
      "step": 470
    },
    {
      "epoch": 0.09,
      "grad_norm": 5.814584732055664,
      "learning_rate": 1.9819311123658952e-05,
      "loss": 0.9543,
      "step": 480
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.66559886932373,
      "learning_rate": 1.981554677206851e-05,
      "loss": 0.9508,
      "step": 490
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.177793502807617,
      "learning_rate": 1.9811782420478075e-05,
      "loss": 0.9477,
      "step": 500
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.47395133972168,
      "learning_rate": 1.9808018068887634e-05,
      "loss": 0.8307,
      "step": 510
    },
    {
      "epoch": 0.1,
      "grad_norm": 4.7389116287231445,
      "learning_rate": 1.9804253717297198e-05,
      "loss": 0.8478,
      "step": 520
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.453147888183594,
      "learning_rate": 1.9800489365706757e-05,
      "loss": 0.8604,
      "step": 530
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.6701579093933105,
      "learning_rate": 1.979672501411632e-05,
      "loss": 0.7919,
      "step": 540
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.018000602722168,
      "learning_rate": 1.979296066252588e-05,
      "loss": 0.8582,
      "step": 550
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.23560905456543,
      "learning_rate": 1.9789196310935443e-05,
      "loss": 0.7818,
      "step": 560
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.050342559814453,
      "learning_rate": 1.9785431959345007e-05,
      "loss": 0.835,
      "step": 570
    },
    {
      "epoch": 0.11,
      "grad_norm": 6.1569743156433105,
      "learning_rate": 1.9781667607754566e-05,
      "loss": 0.7801,
      "step": 580
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.508142471313477,
      "learning_rate": 1.977790325616413e-05,
      "loss": 0.808,
      "step": 590
    },
    {
      "epoch": 0.11,
      "grad_norm": 8.961487770080566,
      "learning_rate": 1.977413890457369e-05,
      "loss": 0.8116,
      "step": 600
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.894444465637207,
      "learning_rate": 1.977037455298325e-05,
      "loss": 0.8347,
      "step": 610
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.366480588912964,
      "learning_rate": 1.9766610201392812e-05,
      "loss": 0.7458,
      "step": 620
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.9548075199127197,
      "learning_rate": 1.9762845849802372e-05,
      "loss": 0.8069,
      "step": 630
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.245074272155762,
      "learning_rate": 1.9759081498211935e-05,
      "loss": 0.8055,
      "step": 640
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.90622615814209,
      "learning_rate": 1.9755317146621495e-05,
      "loss": 0.7178,
      "step": 650
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.3787405490875244,
      "learning_rate": 1.9751552795031058e-05,
      "loss": 0.6198,
      "step": 660
    },
    {
      "epoch": 0.13,
      "grad_norm": 8.888846397399902,
      "learning_rate": 1.9747788443440618e-05,
      "loss": 0.7308,
      "step": 670
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.188430309295654,
      "learning_rate": 1.974402409185018e-05,
      "loss": 0.7489,
      "step": 680
    },
    {
      "epoch": 0.13,
      "grad_norm": 6.952388286590576,
      "learning_rate": 1.974025974025974e-05,
      "loss": 0.695,
      "step": 690
    },
    {
      "epoch": 0.13,
      "grad_norm": 14.750497817993164,
      "learning_rate": 1.9736495388669304e-05,
      "loss": 0.7685,
      "step": 700
    },
    {
      "epoch": 0.13,
      "grad_norm": 9.111574172973633,
      "learning_rate": 1.9732731037078864e-05,
      "loss": 0.5338,
      "step": 710
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.903047561645508,
      "learning_rate": 1.9728966685488427e-05,
      "loss": 0.7381,
      "step": 720
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.304437637329102,
      "learning_rate": 1.9725202333897987e-05,
      "loss": 0.7738,
      "step": 730
    },
    {
      "epoch": 0.14,
      "grad_norm": 8.232093811035156,
      "learning_rate": 1.972143798230755e-05,
      "loss": 0.6408,
      "step": 740
    },
    {
      "epoch": 0.14,
      "grad_norm": 4.832240581512451,
      "learning_rate": 1.9717673630717113e-05,
      "loss": 0.7076,
      "step": 750
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.338514566421509,
      "learning_rate": 1.9713909279126673e-05,
      "loss": 0.5465,
      "step": 760
    },
    {
      "epoch": 0.14,
      "grad_norm": 7.6388468742370605,
      "learning_rate": 1.9710144927536236e-05,
      "loss": 0.629,
      "step": 770
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.158779144287109,
      "learning_rate": 1.9706380575945796e-05,
      "loss": 0.6122,
      "step": 780
    },
    {
      "epoch": 0.15,
      "grad_norm": 3.0482125282287598,
      "learning_rate": 1.9702616224355355e-05,
      "loss": 0.6316,
      "step": 790
    },
    {
      "epoch": 0.15,
      "grad_norm": 7.655091762542725,
      "learning_rate": 1.969885187276492e-05,
      "loss": 0.6038,
      "step": 800
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.0930566787719727,
      "learning_rate": 1.9695087521174478e-05,
      "loss": 0.5782,
      "step": 810
    },
    {
      "epoch": 0.15,
      "grad_norm": 6.42022180557251,
      "learning_rate": 1.969132316958404e-05,
      "loss": 0.5586,
      "step": 820
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.554563522338867,
      "learning_rate": 1.96875588179936e-05,
      "loss": 0.7011,
      "step": 830
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.002378940582275,
      "learning_rate": 1.9683794466403164e-05,
      "loss": 0.7199,
      "step": 840
    },
    {
      "epoch": 0.16,
      "grad_norm": 16.87588119506836,
      "learning_rate": 1.9680030114812724e-05,
      "loss": 0.7092,
      "step": 850
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.338812828063965,
      "learning_rate": 1.9676265763222287e-05,
      "loss": 0.7505,
      "step": 860
    },
    {
      "epoch": 0.16,
      "grad_norm": 8.91244888305664,
      "learning_rate": 1.9672501411631847e-05,
      "loss": 0.6793,
      "step": 870
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.374460697174072,
      "learning_rate": 1.966873706004141e-05,
      "loss": 0.5094,
      "step": 880
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.669682025909424,
      "learning_rate": 1.966497270845097e-05,
      "loss": 0.448,
      "step": 890
    },
    {
      "epoch": 0.17,
      "grad_norm": 9.208888053894043,
      "learning_rate": 1.9661208356860533e-05,
      "loss": 0.5325,
      "step": 900
    },
    {
      "epoch": 0.17,
      "grad_norm": 6.583241939544678,
      "learning_rate": 1.9657444005270093e-05,
      "loss": 0.5432,
      "step": 910
    },
    {
      "epoch": 0.17,
      "grad_norm": 15.520217895507812,
      "learning_rate": 1.9653679653679656e-05,
      "loss": 0.6806,
      "step": 920
    },
    {
      "epoch": 0.18,
      "grad_norm": 12.892702102661133,
      "learning_rate": 1.964991530208922e-05,
      "loss": 0.6994,
      "step": 930
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.393306255340576,
      "learning_rate": 1.964615095049878e-05,
      "loss": 0.7246,
      "step": 940
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.4449687004089355,
      "learning_rate": 1.9642386598908342e-05,
      "loss": 0.5296,
      "step": 950
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.519042491912842,
      "learning_rate": 1.96386222473179e-05,
      "loss": 0.6237,
      "step": 960
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.8469395637512207,
      "learning_rate": 1.963485789572746e-05,
      "loss": 0.4813,
      "step": 970
    },
    {
      "epoch": 0.18,
      "grad_norm": 4.901673316955566,
      "learning_rate": 1.9631093544137025e-05,
      "loss": 0.5266,
      "step": 980
    },
    {
      "epoch": 0.19,
      "grad_norm": 13.271441459655762,
      "learning_rate": 1.9627329192546585e-05,
      "loss": 0.5264,
      "step": 990
    },
    {
      "epoch": 0.19,
      "grad_norm": 18.88715362548828,
      "learning_rate": 1.9623564840956148e-05,
      "loss": 0.5944,
      "step": 1000
    },
    {
      "epoch": 0.19,
      "grad_norm": 9.696794509887695,
      "learning_rate": 1.9619800489365707e-05,
      "loss": 0.5291,
      "step": 1010
    },
    {
      "epoch": 0.19,
      "grad_norm": 3.8172876834869385,
      "learning_rate": 1.961603613777527e-05,
      "loss": 0.465,
      "step": 1020
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.899192810058594,
      "learning_rate": 1.961227178618483e-05,
      "loss": 0.5869,
      "step": 1030
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.844030380249023,
      "learning_rate": 1.9608507434594394e-05,
      "loss": 0.5845,
      "step": 1040
    },
    {
      "epoch": 0.2,
      "grad_norm": 5.3461833000183105,
      "learning_rate": 1.9604743083003953e-05,
      "loss": 0.4208,
      "step": 1050
    },
    {
      "epoch": 0.2,
      "grad_norm": 17.311580657958984,
      "learning_rate": 1.9600978731413516e-05,
      "loss": 0.5794,
      "step": 1060
    },
    {
      "epoch": 0.2,
      "grad_norm": 13.961825370788574,
      "learning_rate": 1.9597214379823076e-05,
      "loss": 0.5965,
      "step": 1070
    },
    {
      "epoch": 0.2,
      "grad_norm": 11.40803050994873,
      "learning_rate": 1.959345002823264e-05,
      "loss": 0.487,
      "step": 1080
    },
    {
      "epoch": 0.21,
      "grad_norm": 7.2968878746032715,
      "learning_rate": 1.95896856766422e-05,
      "loss": 0.4935,
      "step": 1090
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.641606330871582,
      "learning_rate": 1.9585921325051762e-05,
      "loss": 0.5702,
      "step": 1100
    },
    {
      "epoch": 0.21,
      "grad_norm": 20.537263870239258,
      "learning_rate": 1.9582156973461322e-05,
      "loss": 0.5886,
      "step": 1110
    },
    {
      "epoch": 0.21,
      "grad_norm": 13.335675239562988,
      "learning_rate": 1.9578392621870885e-05,
      "loss": 0.6493,
      "step": 1120
    },
    {
      "epoch": 0.21,
      "grad_norm": 3.4032161235809326,
      "learning_rate": 1.957462827028045e-05,
      "loss": 0.4274,
      "step": 1130
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.771735191345215,
      "learning_rate": 1.9570863918690005e-05,
      "loss": 0.4998,
      "step": 1140
    },
    {
      "epoch": 0.22,
      "grad_norm": 13.649166107177734,
      "learning_rate": 1.9567099567099568e-05,
      "loss": 0.5903,
      "step": 1150
    },
    {
      "epoch": 0.22,
      "grad_norm": 17.220354080200195,
      "learning_rate": 1.956333521550913e-05,
      "loss": 0.4822,
      "step": 1160
    },
    {
      "epoch": 0.22,
      "grad_norm": 24.886266708374023,
      "learning_rate": 1.955957086391869e-05,
      "loss": 0.5327,
      "step": 1170
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7781261801719666,
      "learning_rate": 1.9555806512328254e-05,
      "loss": 0.4998,
      "step": 1180
    },
    {
      "epoch": 0.22,
      "grad_norm": 15.118757247924805,
      "learning_rate": 1.9552042160737814e-05,
      "loss": 0.6799,
      "step": 1190
    },
    {
      "epoch": 0.23,
      "grad_norm": 31.076623916625977,
      "learning_rate": 1.9548277809147377e-05,
      "loss": 0.3816,
      "step": 1200
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.240067481994629,
      "learning_rate": 1.9544513457556937e-05,
      "loss": 0.623,
      "step": 1210
    },
    {
      "epoch": 0.23,
      "grad_norm": 15.31002140045166,
      "learning_rate": 1.95407491059665e-05,
      "loss": 0.7027,
      "step": 1220
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.10597562789917,
      "learning_rate": 1.953698475437606e-05,
      "loss": 0.4529,
      "step": 1230
    },
    {
      "epoch": 0.23,
      "grad_norm": 17.48518943786621,
      "learning_rate": 1.9533220402785623e-05,
      "loss": 0.6128,
      "step": 1240
    },
    {
      "epoch": 0.24,
      "grad_norm": 16.264951705932617,
      "learning_rate": 1.9529456051195183e-05,
      "loss": 0.5778,
      "step": 1250
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6286412477493286,
      "learning_rate": 1.9525691699604746e-05,
      "loss": 0.4268,
      "step": 1260
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.025686740875244,
      "learning_rate": 1.9521927348014305e-05,
      "loss": 0.5031,
      "step": 1270
    },
    {
      "epoch": 0.24,
      "grad_norm": 6.937544345855713,
      "learning_rate": 1.951816299642387e-05,
      "loss": 0.4877,
      "step": 1280
    },
    {
      "epoch": 0.24,
      "grad_norm": 12.527236938476562,
      "learning_rate": 1.951439864483343e-05,
      "loss": 0.5257,
      "step": 1290
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3396496772766113,
      "learning_rate": 1.951063429324299e-05,
      "loss": 0.5487,
      "step": 1300
    },
    {
      "epoch": 0.25,
      "grad_norm": 6.098325729370117,
      "learning_rate": 1.950686994165255e-05,
      "loss": 0.4214,
      "step": 1310
    },
    {
      "epoch": 0.25,
      "grad_norm": 15.815199851989746,
      "learning_rate": 1.950310559006211e-05,
      "loss": 0.4703,
      "step": 1320
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.978259086608887,
      "learning_rate": 1.9499341238471674e-05,
      "loss": 0.43,
      "step": 1330
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.7017760276794434,
      "learning_rate": 1.9495576886881234e-05,
      "loss": 0.5265,
      "step": 1340
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.4417724609375,
      "learning_rate": 1.9491812535290797e-05,
      "loss": 0.4193,
      "step": 1350
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.31155252456665,
      "learning_rate": 1.948804818370036e-05,
      "loss": 0.6,
      "step": 1360
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.697686672210693,
      "learning_rate": 1.948428383210992e-05,
      "loss": 0.3223,
      "step": 1370
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.571923732757568,
      "learning_rate": 1.9480519480519483e-05,
      "loss": 0.4124,
      "step": 1380
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.5974889397621155,
      "learning_rate": 1.9476755128929043e-05,
      "loss": 0.5792,
      "step": 1390
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.555315017700195,
      "learning_rate": 1.9472990777338606e-05,
      "loss": 0.6258,
      "step": 1400
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4411174058914185,
      "learning_rate": 1.9469226425748166e-05,
      "loss": 0.3901,
      "step": 1410
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.3852739334106445,
      "learning_rate": 1.946546207415773e-05,
      "loss": 0.4935,
      "step": 1420
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.584674835205078,
      "learning_rate": 1.946169772256729e-05,
      "loss": 0.4618,
      "step": 1430
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3752799034118652,
      "learning_rate": 1.9457933370976852e-05,
      "loss": 0.3981,
      "step": 1440
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.5314813852310181,
      "learning_rate": 1.9454169019386412e-05,
      "loss": 0.4034,
      "step": 1450
    },
    {
      "epoch": 0.27,
      "grad_norm": 7.783598899841309,
      "learning_rate": 1.9450404667795975e-05,
      "loss": 0.4625,
      "step": 1460
    },
    {
      "epoch": 0.28,
      "grad_norm": 15.772924423217773,
      "learning_rate": 1.9446640316205535e-05,
      "loss": 0.5227,
      "step": 1470
    },
    {
      "epoch": 0.28,
      "grad_norm": 13.923240661621094,
      "learning_rate": 1.9442875964615098e-05,
      "loss": 0.6646,
      "step": 1480
    },
    {
      "epoch": 0.28,
      "grad_norm": 15.943310737609863,
      "learning_rate": 1.9439111613024658e-05,
      "loss": 0.4111,
      "step": 1490
    },
    {
      "epoch": 0.28,
      "grad_norm": 4.564625263214111,
      "learning_rate": 1.9435347261434217e-05,
      "loss": 0.4224,
      "step": 1500
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.81473159790039,
      "learning_rate": 1.943158290984378e-05,
      "loss": 0.4566,
      "step": 1510
    },
    {
      "epoch": 0.29,
      "grad_norm": 4.042243957519531,
      "learning_rate": 1.942781855825334e-05,
      "loss": 0.341,
      "step": 1520
    },
    {
      "epoch": 0.29,
      "grad_norm": 3.45874285697937,
      "learning_rate": 1.9424054206662903e-05,
      "loss": 0.6423,
      "step": 1530
    },
    {
      "epoch": 0.29,
      "grad_norm": 7.619143009185791,
      "learning_rate": 1.9420289855072467e-05,
      "loss": 0.4225,
      "step": 1540
    },
    {
      "epoch": 0.29,
      "grad_norm": 18.439268112182617,
      "learning_rate": 1.9416525503482026e-05,
      "loss": 0.5887,
      "step": 1550
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.509594917297363,
      "learning_rate": 1.941276115189159e-05,
      "loss": 0.2852,
      "step": 1560
    },
    {
      "epoch": 0.3,
      "grad_norm": 20.117277145385742,
      "learning_rate": 1.940899680030115e-05,
      "loss": 0.3562,
      "step": 1570
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.0258663892745972,
      "learning_rate": 1.9405232448710712e-05,
      "loss": 0.3495,
      "step": 1580
    },
    {
      "epoch": 0.3,
      "grad_norm": 12.384246826171875,
      "learning_rate": 1.9401468097120272e-05,
      "loss": 0.3677,
      "step": 1590
    },
    {
      "epoch": 0.3,
      "grad_norm": 7.396050453186035,
      "learning_rate": 1.9397703745529835e-05,
      "loss": 0.5209,
      "step": 1600
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.182412147521973,
      "learning_rate": 1.9393939393939395e-05,
      "loss": 0.4951,
      "step": 1610
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.2450031042099,
      "learning_rate": 1.9390175042348958e-05,
      "loss": 0.3517,
      "step": 1620
    },
    {
      "epoch": 0.31,
      "grad_norm": 21.506317138671875,
      "learning_rate": 1.9386410690758518e-05,
      "loss": 0.3767,
      "step": 1630
    },
    {
      "epoch": 0.31,
      "grad_norm": 7.109480857849121,
      "learning_rate": 1.938264633916808e-05,
      "loss": 0.401,
      "step": 1640
    },
    {
      "epoch": 0.31,
      "grad_norm": 3.4579553604125977,
      "learning_rate": 1.937888198757764e-05,
      "loss": 0.4032,
      "step": 1650
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.582856178283691,
      "learning_rate": 1.93751176359872e-05,
      "loss": 0.2779,
      "step": 1660
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.766223907470703,
      "learning_rate": 1.9371353284396764e-05,
      "loss": 0.4978,
      "step": 1670
    },
    {
      "epoch": 0.32,
      "grad_norm": 29.351909637451172,
      "learning_rate": 1.9367588932806324e-05,
      "loss": 0.446,
      "step": 1680
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.514208793640137,
      "learning_rate": 1.9363824581215887e-05,
      "loss": 0.3558,
      "step": 1690
    },
    {
      "epoch": 0.32,
      "grad_norm": 4.894351959228516,
      "learning_rate": 1.9360060229625447e-05,
      "loss": 0.558,
      "step": 1700
    },
    {
      "epoch": 0.32,
      "grad_norm": 8.540889739990234,
      "learning_rate": 1.935629587803501e-05,
      "loss": 0.3761,
      "step": 1710
    },
    {
      "epoch": 0.32,
      "grad_norm": 14.161347389221191,
      "learning_rate": 1.9352531526444573e-05,
      "loss": 0.4687,
      "step": 1720
    },
    {
      "epoch": 0.33,
      "grad_norm": 26.30253028869629,
      "learning_rate": 1.9348767174854133e-05,
      "loss": 0.567,
      "step": 1730
    },
    {
      "epoch": 0.33,
      "grad_norm": 29.719797134399414,
      "learning_rate": 1.9345002823263696e-05,
      "loss": 0.533,
      "step": 1740
    },
    {
      "epoch": 0.33,
      "grad_norm": 5.152700901031494,
      "learning_rate": 1.9341238471673256e-05,
      "loss": 0.6312,
      "step": 1750
    },
    {
      "epoch": 0.33,
      "grad_norm": 19.56715202331543,
      "learning_rate": 1.933747412008282e-05,
      "loss": 0.532,
      "step": 1760
    },
    {
      "epoch": 0.33,
      "grad_norm": 14.732528686523438,
      "learning_rate": 1.933370976849238e-05,
      "loss": 0.3465,
      "step": 1770
    },
    {
      "epoch": 0.34,
      "grad_norm": 13.159880638122559,
      "learning_rate": 1.932994541690194e-05,
      "loss": 0.3621,
      "step": 1780
    },
    {
      "epoch": 0.34,
      "grad_norm": 14.20004653930664,
      "learning_rate": 1.93261810653115e-05,
      "loss": 0.3553,
      "step": 1790
    },
    {
      "epoch": 0.34,
      "grad_norm": 11.828947067260742,
      "learning_rate": 1.9322416713721065e-05,
      "loss": 0.523,
      "step": 1800
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.383852005004883,
      "learning_rate": 1.9318652362130624e-05,
      "loss": 0.5887,
      "step": 1810
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.9098563194274902,
      "learning_rate": 1.9314888010540187e-05,
      "loss": 0.4421,
      "step": 1820
    },
    {
      "epoch": 0.34,
      "grad_norm": 16.96078109741211,
      "learning_rate": 1.9311123658949747e-05,
      "loss": 0.4026,
      "step": 1830
    },
    {
      "epoch": 0.35,
      "grad_norm": 16.952852249145508,
      "learning_rate": 1.9307359307359307e-05,
      "loss": 0.4488,
      "step": 1840
    },
    {
      "epoch": 0.35,
      "grad_norm": 9.7103910446167,
      "learning_rate": 1.930359495576887e-05,
      "loss": 0.4461,
      "step": 1850
    },
    {
      "epoch": 0.35,
      "grad_norm": 14.61265754699707,
      "learning_rate": 1.929983060417843e-05,
      "loss": 0.4779,
      "step": 1860
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.043939590454102,
      "learning_rate": 1.9296066252587993e-05,
      "loss": 0.4594,
      "step": 1870
    },
    {
      "epoch": 0.35,
      "grad_norm": 18.581029891967773,
      "learning_rate": 1.9292301900997553e-05,
      "loss": 0.5699,
      "step": 1880
    },
    {
      "epoch": 0.36,
      "grad_norm": 22.796884536743164,
      "learning_rate": 1.9288537549407116e-05,
      "loss": 0.5695,
      "step": 1890
    },
    {
      "epoch": 0.36,
      "grad_norm": 21.6387939453125,
      "learning_rate": 1.9284773197816676e-05,
      "loss": 0.4768,
      "step": 1900
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.905150413513184,
      "learning_rate": 1.928100884622624e-05,
      "loss": 0.4545,
      "step": 1910
    },
    {
      "epoch": 0.36,
      "grad_norm": 3.67718768119812,
      "learning_rate": 1.9277244494635802e-05,
      "loss": 0.3285,
      "step": 1920
    },
    {
      "epoch": 0.36,
      "grad_norm": 5.294912815093994,
      "learning_rate": 1.9273480143045362e-05,
      "loss": 0.3583,
      "step": 1930
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.164522171020508,
      "learning_rate": 1.9269715791454925e-05,
      "loss": 0.3476,
      "step": 1940
    },
    {
      "epoch": 0.37,
      "grad_norm": 13.233643531799316,
      "learning_rate": 1.9265951439864485e-05,
      "loss": 0.4586,
      "step": 1950
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.404780864715576,
      "learning_rate": 1.9262187088274048e-05,
      "loss": 0.4162,
      "step": 1960
    },
    {
      "epoch": 0.37,
      "grad_norm": 18.41141128540039,
      "learning_rate": 1.9258422736683608e-05,
      "loss": 0.3693,
      "step": 1970
    },
    {
      "epoch": 0.37,
      "grad_norm": 4.590950012207031,
      "learning_rate": 1.925465838509317e-05,
      "loss": 0.4327,
      "step": 1980
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4611373543739319,
      "learning_rate": 1.925089403350273e-05,
      "loss": 0.3137,
      "step": 1990
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.073715209960938,
      "learning_rate": 1.9247129681912294e-05,
      "loss": 0.2616,
      "step": 2000
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.125640869140625,
      "learning_rate": 1.9243365330321854e-05,
      "loss": 0.4098,
      "step": 2010
    },
    {
      "epoch": 0.38,
      "grad_norm": 3.2404801845550537,
      "learning_rate": 1.9239600978731413e-05,
      "loss": 0.2879,
      "step": 2020
    },
    {
      "epoch": 0.38,
      "grad_norm": 8.871710777282715,
      "learning_rate": 1.9235836627140976e-05,
      "loss": 0.4856,
      "step": 2030
    },
    {
      "epoch": 0.38,
      "grad_norm": 15.049797058105469,
      "learning_rate": 1.9232072275550536e-05,
      "loss": 0.4209,
      "step": 2040
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.116982936859131,
      "learning_rate": 1.92283079239601e-05,
      "loss": 0.3569,
      "step": 2050
    },
    {
      "epoch": 0.39,
      "grad_norm": 16.39138412475586,
      "learning_rate": 1.922454357236966e-05,
      "loss": 0.4758,
      "step": 2060
    },
    {
      "epoch": 0.39,
      "grad_norm": 17.201120376586914,
      "learning_rate": 1.9220779220779222e-05,
      "loss": 0.5029,
      "step": 2070
    },
    {
      "epoch": 0.39,
      "grad_norm": 7.135878562927246,
      "learning_rate": 1.9217014869188782e-05,
      "loss": 0.404,
      "step": 2080
    },
    {
      "epoch": 0.39,
      "grad_norm": 16.335006713867188,
      "learning_rate": 1.9213250517598345e-05,
      "loss": 0.3514,
      "step": 2090
    },
    {
      "epoch": 0.4,
      "grad_norm": 19.623199462890625,
      "learning_rate": 1.920948616600791e-05,
      "loss": 0.359,
      "step": 2100
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.7051615715026855,
      "learning_rate": 1.9205721814417468e-05,
      "loss": 0.3522,
      "step": 2110
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.48463249206543,
      "learning_rate": 1.920195746282703e-05,
      "loss": 0.3984,
      "step": 2120
    },
    {
      "epoch": 0.4,
      "grad_norm": 15.596338272094727,
      "learning_rate": 1.919819311123659e-05,
      "loss": 0.4853,
      "step": 2130
    },
    {
      "epoch": 0.4,
      "grad_norm": 6.698812007904053,
      "learning_rate": 1.9194428759646154e-05,
      "loss": 0.5048,
      "step": 2140
    },
    {
      "epoch": 0.4,
      "grad_norm": 15.503188133239746,
      "learning_rate": 1.9190664408055714e-05,
      "loss": 0.4302,
      "step": 2150
    },
    {
      "epoch": 0.41,
      "grad_norm": 7.965035438537598,
      "learning_rate": 1.9186900056465277e-05,
      "loss": 0.4024,
      "step": 2160
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.394098281860352,
      "learning_rate": 1.9183135704874837e-05,
      "loss": 0.3822,
      "step": 2170
    },
    {
      "epoch": 0.41,
      "grad_norm": 6.874277591705322,
      "learning_rate": 1.9179371353284397e-05,
      "loss": 0.321,
      "step": 2180
    },
    {
      "epoch": 0.41,
      "grad_norm": 4.065316200256348,
      "learning_rate": 1.917560700169396e-05,
      "loss": 0.4858,
      "step": 2190
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.484227180480957,
      "learning_rate": 1.917184265010352e-05,
      "loss": 0.3184,
      "step": 2200
    },
    {
      "epoch": 0.42,
      "grad_norm": 15.838642120361328,
      "learning_rate": 1.9168078298513083e-05,
      "loss": 0.3317,
      "step": 2210
    },
    {
      "epoch": 0.42,
      "grad_norm": 14.35332202911377,
      "learning_rate": 1.9164313946922643e-05,
      "loss": 0.4181,
      "step": 2220
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.4296956062316895,
      "learning_rate": 1.9160549595332206e-05,
      "loss": 0.4403,
      "step": 2230
    },
    {
      "epoch": 0.42,
      "grad_norm": 20.951000213623047,
      "learning_rate": 1.9156785243741765e-05,
      "loss": 0.3265,
      "step": 2240
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.6506690979003906,
      "learning_rate": 1.915302089215133e-05,
      "loss": 0.2785,
      "step": 2250
    },
    {
      "epoch": 0.43,
      "grad_norm": 8.721875190734863,
      "learning_rate": 1.914925654056089e-05,
      "loss": 0.476,
      "step": 2260
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5788027048110962,
      "learning_rate": 1.914549218897045e-05,
      "loss": 0.4012,
      "step": 2270
    },
    {
      "epoch": 0.43,
      "grad_norm": 3.5817747116088867,
      "learning_rate": 1.9141727837380015e-05,
      "loss": 0.38,
      "step": 2280
    },
    {
      "epoch": 0.43,
      "grad_norm": 4.096621036529541,
      "learning_rate": 1.9137963485789574e-05,
      "loss": 0.2905,
      "step": 2290
    },
    {
      "epoch": 0.43,
      "grad_norm": 19.098316192626953,
      "learning_rate": 1.9134199134199138e-05,
      "loss": 0.414,
      "step": 2300
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.006946563720703,
      "learning_rate": 1.9130434782608697e-05,
      "loss": 0.4336,
      "step": 2310
    },
    {
      "epoch": 0.44,
      "grad_norm": 25.646982192993164,
      "learning_rate": 1.912667043101826e-05,
      "loss": 0.5114,
      "step": 2320
    },
    {
      "epoch": 0.44,
      "grad_norm": 12.100605010986328,
      "learning_rate": 1.912290607942782e-05,
      "loss": 0.2938,
      "step": 2330
    },
    {
      "epoch": 0.44,
      "grad_norm": 18.50227928161621,
      "learning_rate": 1.9119141727837383e-05,
      "loss": 0.3462,
      "step": 2340
    },
    {
      "epoch": 0.44,
      "grad_norm": 15.916492462158203,
      "learning_rate": 1.9115377376246943e-05,
      "loss": 0.4863,
      "step": 2350
    },
    {
      "epoch": 0.44,
      "grad_norm": 4.765915870666504,
      "learning_rate": 1.9111613024656503e-05,
      "loss": 0.2875,
      "step": 2360
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.605278015136719,
      "learning_rate": 1.9107848673066066e-05,
      "loss": 0.4483,
      "step": 2370
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.7430098056793213,
      "learning_rate": 1.9104084321475626e-05,
      "loss": 0.4582,
      "step": 2380
    },
    {
      "epoch": 0.45,
      "grad_norm": 53.96059799194336,
      "learning_rate": 1.910031996988519e-05,
      "loss": 0.3417,
      "step": 2390
    },
    {
      "epoch": 0.45,
      "grad_norm": 23.37925148010254,
      "learning_rate": 1.909655561829475e-05,
      "loss": 0.1993,
      "step": 2400
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.89754867553711,
      "learning_rate": 1.9092791266704312e-05,
      "loss": 0.4081,
      "step": 2410
    },
    {
      "epoch": 0.46,
      "grad_norm": 17.08811378479004,
      "learning_rate": 1.9089026915113872e-05,
      "loss": 0.5033,
      "step": 2420
    },
    {
      "epoch": 0.46,
      "grad_norm": 14.817082405090332,
      "learning_rate": 1.9085262563523435e-05,
      "loss": 0.5315,
      "step": 2430
    },
    {
      "epoch": 0.46,
      "grad_norm": 6.389044284820557,
      "learning_rate": 1.9081498211932995e-05,
      "loss": 0.4177,
      "step": 2440
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5137330889701843,
      "learning_rate": 1.9077733860342558e-05,
      "loss": 0.2788,
      "step": 2450
    },
    {
      "epoch": 0.46,
      "grad_norm": 4.673267841339111,
      "learning_rate": 1.907396950875212e-05,
      "loss": 0.3451,
      "step": 2460
    },
    {
      "epoch": 0.46,
      "grad_norm": 20.979839324951172,
      "learning_rate": 1.907020515716168e-05,
      "loss": 0.6803,
      "step": 2470
    },
    {
      "epoch": 0.47,
      "grad_norm": 4.662585258483887,
      "learning_rate": 1.9066440805571244e-05,
      "loss": 0.2544,
      "step": 2480
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.5298420190811157,
      "learning_rate": 1.9062676453980804e-05,
      "loss": 0.2808,
      "step": 2490
    },
    {
      "epoch": 0.47,
      "grad_norm": 13.276006698608398,
      "learning_rate": 1.9058912102390367e-05,
      "loss": 0.2642,
      "step": 2500
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.166401863098145,
      "learning_rate": 1.9055147750799927e-05,
      "loss": 0.4446,
      "step": 2510
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.779420375823975,
      "learning_rate": 1.905138339920949e-05,
      "loss": 0.2508,
      "step": 2520
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8076228499412537,
      "learning_rate": 1.904761904761905e-05,
      "loss": 0.352,
      "step": 2530
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.194454193115234,
      "learning_rate": 1.904385469602861e-05,
      "loss": 0.3334,
      "step": 2540
    },
    {
      "epoch": 0.48,
      "grad_norm": 20.605688095092773,
      "learning_rate": 1.9040090344438172e-05,
      "loss": 0.3524,
      "step": 2550
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3982861340045929,
      "learning_rate": 1.9036325992847732e-05,
      "loss": 0.2745,
      "step": 2560
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.395184516906738,
      "learning_rate": 1.9032561641257295e-05,
      "loss": 0.4688,
      "step": 2570
    },
    {
      "epoch": 0.49,
      "grad_norm": 8.443707466125488,
      "learning_rate": 1.9028797289666855e-05,
      "loss": 0.3585,
      "step": 2580
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.26993560791015625,
      "learning_rate": 1.9025032938076418e-05,
      "loss": 0.3365,
      "step": 2590
    },
    {
      "epoch": 0.49,
      "grad_norm": 19.354724884033203,
      "learning_rate": 1.9021268586485978e-05,
      "loss": 0.3284,
      "step": 2600
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.1320550441741943,
      "learning_rate": 1.901750423489554e-05,
      "loss": 0.3734,
      "step": 2610
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.4338151812553406,
      "learning_rate": 1.90137398833051e-05,
      "loss": 0.3889,
      "step": 2620
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.13483715057373,
      "learning_rate": 1.9009975531714664e-05,
      "loss": 0.4172,
      "step": 2630
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2432006597518921,
      "learning_rate": 1.9006211180124224e-05,
      "loss": 0.3854,
      "step": 2640
    },
    {
      "epoch": 0.5,
      "grad_norm": 6.900197505950928,
      "learning_rate": 1.9002446828533787e-05,
      "loss": 0.3931,
      "step": 2650
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.32918930053711,
      "learning_rate": 1.899868247694335e-05,
      "loss": 0.2212,
      "step": 2660
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.745707511901855,
      "learning_rate": 1.899491812535291e-05,
      "loss": 0.1989,
      "step": 2670
    },
    {
      "epoch": 0.5,
      "grad_norm": 20.327112197875977,
      "learning_rate": 1.8991153773762473e-05,
      "loss": 0.3687,
      "step": 2680
    },
    {
      "epoch": 0.51,
      "grad_norm": 2.3677399158477783,
      "learning_rate": 1.8987389422172033e-05,
      "loss": 0.4441,
      "step": 2690
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.20602819323539734,
      "learning_rate": 1.8983625070581596e-05,
      "loss": 0.4168,
      "step": 2700
    },
    {
      "epoch": 0.51,
      "grad_norm": 17.441272735595703,
      "learning_rate": 1.8979860718991156e-05,
      "loss": 0.3532,
      "step": 2710
    },
    {
      "epoch": 0.51,
      "grad_norm": 4.144612789154053,
      "learning_rate": 1.8976096367400716e-05,
      "loss": 0.3653,
      "step": 2720
    },
    {
      "epoch": 0.51,
      "grad_norm": 13.233855247497559,
      "learning_rate": 1.897233201581028e-05,
      "loss": 0.3939,
      "step": 2730
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.36828988790512085,
      "learning_rate": 1.896856766421984e-05,
      "loss": 0.2799,
      "step": 2740
    },
    {
      "epoch": 0.52,
      "grad_norm": 5.536131858825684,
      "learning_rate": 1.89648033126294e-05,
      "loss": 0.3448,
      "step": 2750
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.416645050048828,
      "learning_rate": 1.896103896103896e-05,
      "loss": 0.3487,
      "step": 2760
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.731974601745605,
      "learning_rate": 1.8957274609448525e-05,
      "loss": 0.4327,
      "step": 2770
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.806385040283203,
      "learning_rate": 1.8953510257858084e-05,
      "loss": 0.3772,
      "step": 2780
    },
    {
      "epoch": 0.53,
      "grad_norm": 8.177189826965332,
      "learning_rate": 1.8949745906267647e-05,
      "loss": 0.4002,
      "step": 2790
    },
    {
      "epoch": 0.53,
      "grad_norm": 5.6015424728393555,
      "learning_rate": 1.8945981554677207e-05,
      "loss": 0.3613,
      "step": 2800
    },
    {
      "epoch": 0.53,
      "grad_norm": 16.987865447998047,
      "learning_rate": 1.894221720308677e-05,
      "loss": 0.2126,
      "step": 2810
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4467201232910156,
      "learning_rate": 1.893845285149633e-05,
      "loss": 0.2995,
      "step": 2820
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.20903028547763824,
      "learning_rate": 1.8934688499905893e-05,
      "loss": 0.4346,
      "step": 2830
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.2488715648651123,
      "learning_rate": 1.8930924148315456e-05,
      "loss": 0.4355,
      "step": 2840
    },
    {
      "epoch": 0.54,
      "grad_norm": 7.609149932861328,
      "learning_rate": 1.8927159796725016e-05,
      "loss": 0.4929,
      "step": 2850
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.317333221435547,
      "learning_rate": 1.892339544513458e-05,
      "loss": 0.3358,
      "step": 2860
    },
    {
      "epoch": 0.54,
      "grad_norm": 6.560088157653809,
      "learning_rate": 1.891963109354414e-05,
      "loss": 0.4056,
      "step": 2870
    },
    {
      "epoch": 0.54,
      "grad_norm": 19.497058868408203,
      "learning_rate": 1.89158667419537e-05,
      "loss": 0.2039,
      "step": 2880
    },
    {
      "epoch": 0.54,
      "grad_norm": 19.92127799987793,
      "learning_rate": 1.8912102390363262e-05,
      "loss": 0.5016,
      "step": 2890
    },
    {
      "epoch": 0.55,
      "grad_norm": 20.401813507080078,
      "learning_rate": 1.8908338038772822e-05,
      "loss": 0.2783,
      "step": 2900
    },
    {
      "epoch": 0.55,
      "grad_norm": 2.5606870651245117,
      "learning_rate": 1.8904573687182385e-05,
      "loss": 0.1566,
      "step": 2910
    },
    {
      "epoch": 0.55,
      "grad_norm": 6.865498065948486,
      "learning_rate": 1.8900809335591945e-05,
      "loss": 0.2979,
      "step": 2920
    },
    {
      "epoch": 0.55,
      "grad_norm": 10.443206787109375,
      "learning_rate": 1.8897044984001508e-05,
      "loss": 0.441,
      "step": 2930
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.024031639099121,
      "learning_rate": 1.8893280632411068e-05,
      "loss": 0.3154,
      "step": 2940
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.16709014773368835,
      "learning_rate": 1.888951628082063e-05,
      "loss": 0.2852,
      "step": 2950
    },
    {
      "epoch": 0.56,
      "grad_norm": 11.639143943786621,
      "learning_rate": 1.888575192923019e-05,
      "loss": 0.4774,
      "step": 2960
    },
    {
      "epoch": 0.56,
      "grad_norm": 21.260581970214844,
      "learning_rate": 1.8881987577639754e-05,
      "loss": 0.2661,
      "step": 2970
    },
    {
      "epoch": 0.56,
      "grad_norm": 5.51666259765625,
      "learning_rate": 1.8878223226049314e-05,
      "loss": 0.3656,
      "step": 2980
    },
    {
      "epoch": 0.56,
      "grad_norm": 4.753450870513916,
      "learning_rate": 1.8874458874458877e-05,
      "loss": 0.3346,
      "step": 2990
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.0496236085891724,
      "learning_rate": 1.8870694522868436e-05,
      "loss": 0.5764,
      "step": 3000
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.624492645263672,
      "learning_rate": 1.8866930171278e-05,
      "loss": 0.218,
      "step": 3010
    },
    {
      "epoch": 0.57,
      "grad_norm": 4.713318347930908,
      "learning_rate": 1.8863165819687563e-05,
      "loss": 0.197,
      "step": 3020
    },
    {
      "epoch": 0.57,
      "grad_norm": 29.986032485961914,
      "learning_rate": 1.8859401468097123e-05,
      "loss": 0.4063,
      "step": 3030
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.1420813947916031,
      "learning_rate": 1.8855637116506686e-05,
      "loss": 0.3988,
      "step": 3040
    },
    {
      "epoch": 0.57,
      "grad_norm": 24.69119644165039,
      "learning_rate": 1.8851872764916242e-05,
      "loss": 0.5314,
      "step": 3050
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.815787315368652,
      "learning_rate": 1.8848108413325805e-05,
      "loss": 0.097,
      "step": 3060
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.896925926208496,
      "learning_rate": 1.884434406173537e-05,
      "loss": 0.2229,
      "step": 3070
    },
    {
      "epoch": 0.58,
      "grad_norm": 24.158226013183594,
      "learning_rate": 1.8840579710144928e-05,
      "loss": 0.4095,
      "step": 3080
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.1922907680273056,
      "learning_rate": 1.883681535855449e-05,
      "loss": 0.1617,
      "step": 3090
    },
    {
      "epoch": 0.58,
      "grad_norm": 5.08436393737793,
      "learning_rate": 1.883305100696405e-05,
      "loss": 0.4129,
      "step": 3100
    },
    {
      "epoch": 0.59,
      "grad_norm": 3.495609760284424,
      "learning_rate": 1.8829286655373614e-05,
      "loss": 0.4011,
      "step": 3110
    },
    {
      "epoch": 0.59,
      "grad_norm": 7.160621166229248,
      "learning_rate": 1.8825522303783174e-05,
      "loss": 0.2615,
      "step": 3120
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.2513090372085571,
      "learning_rate": 1.8821757952192737e-05,
      "loss": 0.3322,
      "step": 3130
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.97227954864502,
      "learning_rate": 1.8817993600602297e-05,
      "loss": 0.2852,
      "step": 3140
    },
    {
      "epoch": 0.59,
      "grad_norm": 21.775028228759766,
      "learning_rate": 1.881422924901186e-05,
      "loss": 0.3863,
      "step": 3150
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.059825420379639,
      "learning_rate": 1.881046489742142e-05,
      "loss": 0.5047,
      "step": 3160
    },
    {
      "epoch": 0.6,
      "grad_norm": 13.549070358276367,
      "learning_rate": 1.8806700545830983e-05,
      "loss": 0.2473,
      "step": 3170
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.092524528503418,
      "learning_rate": 1.8802936194240543e-05,
      "loss": 0.4863,
      "step": 3180
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.21894772350788116,
      "learning_rate": 1.8799171842650106e-05,
      "loss": 0.52,
      "step": 3190
    },
    {
      "epoch": 0.6,
      "grad_norm": 21.380390167236328,
      "learning_rate": 1.8795407491059666e-05,
      "loss": 0.3953,
      "step": 3200
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4426378309726715,
      "learning_rate": 1.879164313946923e-05,
      "loss": 0.1354,
      "step": 3210
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.654337882995605,
      "learning_rate": 1.8787878787878792e-05,
      "loss": 0.2518,
      "step": 3220
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.994779586791992,
      "learning_rate": 1.878411443628835e-05,
      "loss": 0.2122,
      "step": 3230
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.271393775939941,
      "learning_rate": 1.878035008469791e-05,
      "loss": 0.2757,
      "step": 3240
    },
    {
      "epoch": 0.61,
      "grad_norm": 5.836674690246582,
      "learning_rate": 1.8776585733107475e-05,
      "loss": 0.4413,
      "step": 3250
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.7096942663192749,
      "learning_rate": 1.8772821381517034e-05,
      "loss": 0.1442,
      "step": 3260
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.13442808389663696,
      "learning_rate": 1.8769057029926598e-05,
      "loss": 0.2355,
      "step": 3270
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.3720588684082031,
      "learning_rate": 1.8765292678336157e-05,
      "loss": 0.2432,
      "step": 3280
    },
    {
      "epoch": 0.62,
      "grad_norm": 17.1495418548584,
      "learning_rate": 1.876152832674572e-05,
      "loss": 0.3899,
      "step": 3290
    },
    {
      "epoch": 0.62,
      "grad_norm": 24.300983428955078,
      "learning_rate": 1.875776397515528e-05,
      "loss": 0.5492,
      "step": 3300
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.664377212524414,
      "learning_rate": 1.8753999623564843e-05,
      "loss": 0.2242,
      "step": 3310
    },
    {
      "epoch": 0.62,
      "grad_norm": 20.048418045043945,
      "learning_rate": 1.8750235271974403e-05,
      "loss": 0.4149,
      "step": 3320
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.6590203642845154,
      "learning_rate": 1.8746470920383966e-05,
      "loss": 0.3397,
      "step": 3330
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.602578043937683,
      "learning_rate": 1.8742706568793526e-05,
      "loss": 0.5012,
      "step": 3340
    },
    {
      "epoch": 0.63,
      "grad_norm": 8.189477920532227,
      "learning_rate": 1.873894221720309e-05,
      "loss": 0.3322,
      "step": 3350
    },
    {
      "epoch": 0.63,
      "grad_norm": 21.63746452331543,
      "learning_rate": 1.873517786561265e-05,
      "loss": 0.2193,
      "step": 3360
    },
    {
      "epoch": 0.63,
      "grad_norm": 24.855737686157227,
      "learning_rate": 1.8731413514022212e-05,
      "loss": 0.3602,
      "step": 3370
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6849569082260132,
      "learning_rate": 1.8727649162431772e-05,
      "loss": 0.2525,
      "step": 3380
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2710568606853485,
      "learning_rate": 1.8723884810841335e-05,
      "loss": 0.3558,
      "step": 3390
    },
    {
      "epoch": 0.64,
      "grad_norm": 26.25127601623535,
      "learning_rate": 1.8720120459250895e-05,
      "loss": 0.3094,
      "step": 3400
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.373831748962402,
      "learning_rate": 1.8716356107660455e-05,
      "loss": 0.4803,
      "step": 3410
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.13042207062244415,
      "learning_rate": 1.8712591756070018e-05,
      "loss": 0.2587,
      "step": 3420
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.513965368270874,
      "learning_rate": 1.8708827404479578e-05,
      "loss": 0.4412,
      "step": 3430
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.13519428670406342,
      "learning_rate": 1.870506305288914e-05,
      "loss": 0.3507,
      "step": 3440
    },
    {
      "epoch": 0.65,
      "grad_norm": 21.247726440429688,
      "learning_rate": 1.8701298701298704e-05,
      "loss": 0.3612,
      "step": 3450
    },
    {
      "epoch": 0.65,
      "grad_norm": 21.522607803344727,
      "learning_rate": 1.8697534349708264e-05,
      "loss": 0.407,
      "step": 3460
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1086030900478363,
      "learning_rate": 1.8693769998117827e-05,
      "loss": 0.0736,
      "step": 3470
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.68559741973877,
      "learning_rate": 1.8690005646527387e-05,
      "loss": 0.3317,
      "step": 3480
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.6514382362365723,
      "learning_rate": 1.868624129493695e-05,
      "loss": 0.4263,
      "step": 3490
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.15339209139347076,
      "learning_rate": 1.868247694334651e-05,
      "loss": 0.4605,
      "step": 3500
    },
    {
      "epoch": 0.66,
      "grad_norm": 27.09156036376953,
      "learning_rate": 1.8678712591756073e-05,
      "loss": 0.3307,
      "step": 3510
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.4358277320861816,
      "learning_rate": 1.8674948240165632e-05,
      "loss": 0.2532,
      "step": 3520
    },
    {
      "epoch": 0.66,
      "grad_norm": 18.586095809936523,
      "learning_rate": 1.8671183888575196e-05,
      "loss": 0.2632,
      "step": 3530
    },
    {
      "epoch": 0.67,
      "grad_norm": 14.750617980957031,
      "learning_rate": 1.8667419536984755e-05,
      "loss": 0.357,
      "step": 3540
    },
    {
      "epoch": 0.67,
      "grad_norm": 4.153351306915283,
      "learning_rate": 1.866365518539432e-05,
      "loss": 0.2523,
      "step": 3550
    },
    {
      "epoch": 0.67,
      "grad_norm": 5.1560211181640625,
      "learning_rate": 1.8659890833803878e-05,
      "loss": 0.3114,
      "step": 3560
    },
    {
      "epoch": 0.67,
      "grad_norm": 14.306260108947754,
      "learning_rate": 1.865612648221344e-05,
      "loss": 0.4807,
      "step": 3570
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.24774952232837677,
      "learning_rate": 1.8652362130623e-05,
      "loss": 0.5276,
      "step": 3580
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.14769458770751953,
      "learning_rate": 1.864859777903256e-05,
      "loss": 0.4019,
      "step": 3590
    },
    {
      "epoch": 0.68,
      "grad_norm": 18.842666625976562,
      "learning_rate": 1.8644833427442124e-05,
      "loss": 0.2624,
      "step": 3600
    },
    {
      "epoch": 0.68,
      "grad_norm": 7.942086219787598,
      "learning_rate": 1.8641069075851684e-05,
      "loss": 0.2449,
      "step": 3610
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.49323031306266785,
      "learning_rate": 1.8637304724261247e-05,
      "loss": 0.3452,
      "step": 3620
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.340002059936523,
      "learning_rate": 1.863354037267081e-05,
      "loss": 0.2158,
      "step": 3630
    },
    {
      "epoch": 0.69,
      "grad_norm": 41.97476577758789,
      "learning_rate": 1.862977602108037e-05,
      "loss": 0.2487,
      "step": 3640
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.5329411029815674,
      "learning_rate": 1.8626011669489933e-05,
      "loss": 0.2567,
      "step": 3650
    },
    {
      "epoch": 0.69,
      "grad_norm": 4.377205848693848,
      "learning_rate": 1.8622247317899493e-05,
      "loss": 0.6072,
      "step": 3660
    },
    {
      "epoch": 0.69,
      "grad_norm": 34.75514221191406,
      "learning_rate": 1.8618482966309056e-05,
      "loss": 0.578,
      "step": 3670
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9700514078140259,
      "learning_rate": 1.8614718614718616e-05,
      "loss": 0.235,
      "step": 3680
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.1095764636993408,
      "learning_rate": 1.861095426312818e-05,
      "loss": 0.1985,
      "step": 3690
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.2766005992889404,
      "learning_rate": 1.860718991153774e-05,
      "loss": 0.2345,
      "step": 3700
    },
    {
      "epoch": 0.7,
      "grad_norm": 19.965744018554688,
      "learning_rate": 1.8603425559947302e-05,
      "loss": 0.4046,
      "step": 3710
    },
    {
      "epoch": 0.7,
      "grad_norm": 39.0003662109375,
      "learning_rate": 1.859966120835686e-05,
      "loss": 0.2664,
      "step": 3720
    },
    {
      "epoch": 0.7,
      "grad_norm": 40.885902404785156,
      "learning_rate": 1.8595896856766425e-05,
      "loss": 0.3563,
      "step": 3730
    },
    {
      "epoch": 0.7,
      "grad_norm": 3.916372060775757,
      "learning_rate": 1.8592132505175985e-05,
      "loss": 0.3511,
      "step": 3740
    },
    {
      "epoch": 0.71,
      "grad_norm": 15.955424308776855,
      "learning_rate": 1.8588368153585544e-05,
      "loss": 0.2216,
      "step": 3750
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.960245132446289,
      "learning_rate": 1.8584603801995107e-05,
      "loss": 0.3281,
      "step": 3760
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.748043417930603,
      "learning_rate": 1.8580839450404667e-05,
      "loss": 0.2858,
      "step": 3770
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.13805866241455,
      "learning_rate": 1.857707509881423e-05,
      "loss": 0.4375,
      "step": 3780
    },
    {
      "epoch": 0.71,
      "grad_norm": 25.220966339111328,
      "learning_rate": 1.857331074722379e-05,
      "loss": 0.3347,
      "step": 3790
    },
    {
      "epoch": 0.72,
      "grad_norm": 8.935770988464355,
      "learning_rate": 1.8569546395633353e-05,
      "loss": 0.2879,
      "step": 3800
    },
    {
      "epoch": 0.72,
      "grad_norm": 23.336849212646484,
      "learning_rate": 1.8565782044042916e-05,
      "loss": 0.4093,
      "step": 3810
    },
    {
      "epoch": 0.72,
      "grad_norm": 36.227134704589844,
      "learning_rate": 1.8562017692452476e-05,
      "loss": 0.3455,
      "step": 3820
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.7966506481170654,
      "learning_rate": 1.855825334086204e-05,
      "loss": 0.3276,
      "step": 3830
    },
    {
      "epoch": 0.72,
      "grad_norm": 2.2726681232452393,
      "learning_rate": 1.85544889892716e-05,
      "loss": 0.3794,
      "step": 3840
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.8250223398208618,
      "learning_rate": 1.8550724637681162e-05,
      "loss": 0.1354,
      "step": 3850
    },
    {
      "epoch": 0.73,
      "grad_norm": 26.319286346435547,
      "learning_rate": 1.8546960286090722e-05,
      "loss": 0.3514,
      "step": 3860
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.805113792419434,
      "learning_rate": 1.8543195934500285e-05,
      "loss": 0.3346,
      "step": 3870
    },
    {
      "epoch": 0.73,
      "grad_norm": 25.349321365356445,
      "learning_rate": 1.8539431582909845e-05,
      "loss": 0.3901,
      "step": 3880
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.5271695256233215,
      "learning_rate": 1.8535667231319408e-05,
      "loss": 0.4922,
      "step": 3890
    },
    {
      "epoch": 0.73,
      "grad_norm": 16.128196716308594,
      "learning_rate": 1.8531902879728968e-05,
      "loss": 0.2922,
      "step": 3900
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.42348727583885193,
      "learning_rate": 1.852813852813853e-05,
      "loss": 0.2131,
      "step": 3910
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.019148826599121,
      "learning_rate": 1.852437417654809e-05,
      "loss": 0.391,
      "step": 3920
    },
    {
      "epoch": 0.74,
      "grad_norm": 4.70227575302124,
      "learning_rate": 1.852060982495765e-05,
      "loss": 0.3015,
      "step": 3930
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.684154987335205,
      "learning_rate": 1.8516845473367214e-05,
      "loss": 0.2834,
      "step": 3940
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.022817611694336,
      "learning_rate": 1.8513081121776774e-05,
      "loss": 0.1923,
      "step": 3950
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8067790865898132,
      "learning_rate": 1.8509316770186337e-05,
      "loss": 0.3325,
      "step": 3960
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.1488920450210571,
      "learning_rate": 1.8505552418595896e-05,
      "loss": 0.2968,
      "step": 3970
    },
    {
      "epoch": 0.75,
      "grad_norm": 12.178488731384277,
      "learning_rate": 1.850178806700546e-05,
      "loss": 0.3268,
      "step": 3980
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.5618786811828613,
      "learning_rate": 1.849802371541502e-05,
      "loss": 0.4774,
      "step": 3990
    },
    {
      "epoch": 0.75,
      "grad_norm": 5.661219120025635,
      "learning_rate": 1.8494259363824583e-05,
      "loss": 0.4162,
      "step": 4000
    },
    {
      "epoch": 0.75,
      "grad_norm": 17.216169357299805,
      "learning_rate": 1.8490495012234146e-05,
      "loss": 0.4605,
      "step": 4010
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.16571305692195892,
      "learning_rate": 1.8486730660643705e-05,
      "loss": 0.4482,
      "step": 4020
    },
    {
      "epoch": 0.76,
      "grad_norm": 21.004676818847656,
      "learning_rate": 1.848296630905327e-05,
      "loss": 0.3889,
      "step": 4030
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.08606118708848953,
      "learning_rate": 1.847920195746283e-05,
      "loss": 0.3311,
      "step": 4040
    },
    {
      "epoch": 0.76,
      "grad_norm": 26.108911514282227,
      "learning_rate": 1.847543760587239e-05,
      "loss": 0.3635,
      "step": 4050
    },
    {
      "epoch": 0.76,
      "grad_norm": 2.373267650604248,
      "learning_rate": 1.847167325428195e-05,
      "loss": 0.1791,
      "step": 4060
    },
    {
      "epoch": 0.77,
      "grad_norm": 18.549354553222656,
      "learning_rate": 1.8467908902691514e-05,
      "loss": 0.307,
      "step": 4070
    },
    {
      "epoch": 0.77,
      "grad_norm": 3.2640037536621094,
      "learning_rate": 1.8464144551101074e-05,
      "loss": 0.4454,
      "step": 4080
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8209856152534485,
      "learning_rate": 1.8460380199510637e-05,
      "loss": 0.4828,
      "step": 4090
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.123522758483887,
      "learning_rate": 1.8456615847920197e-05,
      "loss": 0.521,
      "step": 4100
    },
    {
      "epoch": 0.77,
      "grad_norm": 7.5985822677612305,
      "learning_rate": 1.8452851496329757e-05,
      "loss": 0.4325,
      "step": 4110
    },
    {
      "epoch": 0.78,
      "grad_norm": 21.7657527923584,
      "learning_rate": 1.844908714473932e-05,
      "loss": 0.4065,
      "step": 4120
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.558148384094238,
      "learning_rate": 1.844532279314888e-05,
      "loss": 0.2973,
      "step": 4130
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.80436897277832,
      "learning_rate": 1.8441558441558443e-05,
      "loss": 0.3028,
      "step": 4140
    },
    {
      "epoch": 0.78,
      "grad_norm": 3.0970754623413086,
      "learning_rate": 1.8437794089968003e-05,
      "loss": 0.3983,
      "step": 4150
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.241364479064941,
      "learning_rate": 1.8434029738377566e-05,
      "loss": 0.3504,
      "step": 4160
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.392250061035156,
      "learning_rate": 1.8430265386787126e-05,
      "loss": 0.3923,
      "step": 4170
    },
    {
      "epoch": 0.79,
      "grad_norm": 2.3413052558898926,
      "learning_rate": 1.842650103519669e-05,
      "loss": 0.5141,
      "step": 4180
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.053512573242188,
      "learning_rate": 1.8422736683606252e-05,
      "loss": 0.4397,
      "step": 4190
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3278385102748871,
      "learning_rate": 1.8418972332015812e-05,
      "loss": 0.2543,
      "step": 4200
    },
    {
      "epoch": 0.79,
      "grad_norm": 14.683547019958496,
      "learning_rate": 1.8415207980425375e-05,
      "loss": 0.5726,
      "step": 4210
    },
    {
      "epoch": 0.79,
      "grad_norm": 5.101531028747559,
      "learning_rate": 1.8411443628834935e-05,
      "loss": 0.4049,
      "step": 4220
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5422520041465759,
      "learning_rate": 1.8407679277244498e-05,
      "loss": 0.2672,
      "step": 4230
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.592452049255371,
      "learning_rate": 1.8403914925654058e-05,
      "loss": 0.3553,
      "step": 4240
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.72024154663086,
      "learning_rate": 1.840015057406362e-05,
      "loss": 0.5051,
      "step": 4250
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.038532257080078,
      "learning_rate": 1.839638622247318e-05,
      "loss": 0.3779,
      "step": 4260
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.403825759887695,
      "learning_rate": 1.839262187088274e-05,
      "loss": 0.4494,
      "step": 4270
    },
    {
      "epoch": 0.81,
      "grad_norm": 5.415940761566162,
      "learning_rate": 1.8388857519292303e-05,
      "loss": 0.4299,
      "step": 4280
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.3869456350803375,
      "learning_rate": 1.8385093167701863e-05,
      "loss": 0.3987,
      "step": 4290
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.26389607787132263,
      "learning_rate": 1.8381328816111426e-05,
      "loss": 0.3264,
      "step": 4300
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.805625915527344,
      "learning_rate": 1.8377564464520986e-05,
      "loss": 0.3879,
      "step": 4310
    },
    {
      "epoch": 0.81,
      "grad_norm": 35.108760833740234,
      "learning_rate": 1.837380011293055e-05,
      "loss": 0.2411,
      "step": 4320
    },
    {
      "epoch": 0.81,
      "grad_norm": 6.1817402839660645,
      "learning_rate": 1.837003576134011e-05,
      "loss": 0.3083,
      "step": 4330
    },
    {
      "epoch": 0.82,
      "grad_norm": 6.483433246612549,
      "learning_rate": 1.8366271409749672e-05,
      "loss": 0.3487,
      "step": 4340
    },
    {
      "epoch": 0.82,
      "grad_norm": 3.4147720336914062,
      "learning_rate": 1.8362507058159232e-05,
      "loss": 0.1697,
      "step": 4350
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.923584461212158,
      "learning_rate": 1.8358742706568795e-05,
      "loss": 0.2751,
      "step": 4360
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.098614901304245,
      "learning_rate": 1.8354978354978358e-05,
      "loss": 0.2673,
      "step": 4370
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.692312240600586,
      "learning_rate": 1.8351214003387918e-05,
      "loss": 0.4457,
      "step": 4380
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.104398727416992,
      "learning_rate": 1.834744965179748e-05,
      "loss": 0.5471,
      "step": 4390
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.12419943511486053,
      "learning_rate": 1.834368530020704e-05,
      "loss": 0.2705,
      "step": 4400
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8003888726234436,
      "learning_rate": 1.8339920948616604e-05,
      "loss": 0.4681,
      "step": 4410
    },
    {
      "epoch": 0.83,
      "grad_norm": 17.6849422454834,
      "learning_rate": 1.8336156597026164e-05,
      "loss": 0.2297,
      "step": 4420
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20031321048736572,
      "learning_rate": 1.8332392245435727e-05,
      "loss": 0.2845,
      "step": 4430
    },
    {
      "epoch": 0.84,
      "grad_norm": 6.4708662033081055,
      "learning_rate": 1.8328627893845287e-05,
      "loss": 0.2183,
      "step": 4440
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.16704165935516357,
      "learning_rate": 1.8324863542254847e-05,
      "loss": 0.3274,
      "step": 4450
    },
    {
      "epoch": 0.84,
      "grad_norm": 8.634613990783691,
      "learning_rate": 1.832109919066441e-05,
      "loss": 0.2829,
      "step": 4460
    },
    {
      "epoch": 0.84,
      "grad_norm": 14.103352546691895,
      "learning_rate": 1.831733483907397e-05,
      "loss": 0.2882,
      "step": 4470
    },
    {
      "epoch": 0.84,
      "grad_norm": 17.139183044433594,
      "learning_rate": 1.8313570487483533e-05,
      "loss": 0.3231,
      "step": 4480
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.439067840576172,
      "learning_rate": 1.8309806135893092e-05,
      "loss": 0.3116,
      "step": 4490
    },
    {
      "epoch": 0.85,
      "grad_norm": 16.523258209228516,
      "learning_rate": 1.8306041784302656e-05,
      "loss": 0.2644,
      "step": 4500
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.428323745727539,
      "learning_rate": 1.8302277432712215e-05,
      "loss": 0.4484,
      "step": 4510
    },
    {
      "epoch": 0.85,
      "grad_norm": 20.11658477783203,
      "learning_rate": 1.829851308112178e-05,
      "loss": 0.5045,
      "step": 4520
    },
    {
      "epoch": 0.85,
      "grad_norm": 7.064486503601074,
      "learning_rate": 1.8294748729531338e-05,
      "loss": 0.4726,
      "step": 4530
    },
    {
      "epoch": 0.85,
      "grad_norm": 6.159055233001709,
      "learning_rate": 1.82909843779409e-05,
      "loss": 0.3075,
      "step": 4540
    },
    {
      "epoch": 0.86,
      "grad_norm": 15.74612808227539,
      "learning_rate": 1.8287220026350465e-05,
      "loss": 0.6226,
      "step": 4550
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.9841431975364685,
      "learning_rate": 1.8283455674760024e-05,
      "loss": 0.5233,
      "step": 4560
    },
    {
      "epoch": 0.86,
      "grad_norm": 15.161498069763184,
      "learning_rate": 1.8279691323169587e-05,
      "loss": 0.2225,
      "step": 4570
    },
    {
      "epoch": 0.86,
      "grad_norm": 3.9527482986450195,
      "learning_rate": 1.8275926971579147e-05,
      "loss": 0.2669,
      "step": 4580
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1389472633600235,
      "learning_rate": 1.827216261998871e-05,
      "loss": 0.4668,
      "step": 4590
    },
    {
      "epoch": 0.87,
      "grad_norm": 20.399036407470703,
      "learning_rate": 1.826839826839827e-05,
      "loss": 0.3341,
      "step": 4600
    },
    {
      "epoch": 0.87,
      "grad_norm": 3.456237316131592,
      "learning_rate": 1.8264633916807833e-05,
      "loss": 0.3046,
      "step": 4610
    },
    {
      "epoch": 0.87,
      "grad_norm": 21.47395133972168,
      "learning_rate": 1.8260869565217393e-05,
      "loss": 0.2329,
      "step": 4620
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.514335632324219,
      "learning_rate": 1.8257105213626953e-05,
      "loss": 0.3929,
      "step": 4630
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6025896072387695,
      "learning_rate": 1.8253340862036516e-05,
      "loss": 0.2219,
      "step": 4640
    },
    {
      "epoch": 0.88,
      "grad_norm": 8.066771507263184,
      "learning_rate": 1.8249576510446076e-05,
      "loss": 0.4494,
      "step": 4650
    },
    {
      "epoch": 0.88,
      "grad_norm": 4.125711441040039,
      "learning_rate": 1.824581215885564e-05,
      "loss": 0.3851,
      "step": 4660
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.2077977657318115,
      "learning_rate": 1.82420478072652e-05,
      "loss": 0.421,
      "step": 4670
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2542494833469391,
      "learning_rate": 1.8238283455674762e-05,
      "loss": 0.2053,
      "step": 4680
    },
    {
      "epoch": 0.88,
      "grad_norm": 21.14381217956543,
      "learning_rate": 1.823451910408432e-05,
      "loss": 0.3124,
      "step": 4690
    },
    {
      "epoch": 0.88,
      "grad_norm": 14.048863410949707,
      "learning_rate": 1.8230754752493885e-05,
      "loss": 0.3861,
      "step": 4700
    },
    {
      "epoch": 0.89,
      "grad_norm": 3.9185659885406494,
      "learning_rate": 1.8226990400903445e-05,
      "loss": 0.333,
      "step": 4710
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.7607910633087158,
      "learning_rate": 1.8223226049313008e-05,
      "loss": 0.2568,
      "step": 4720
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6405158042907715,
      "learning_rate": 1.8219461697722567e-05,
      "loss": 0.1758,
      "step": 4730
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.658937454223633,
      "learning_rate": 1.821569734613213e-05,
      "loss": 0.298,
      "step": 4740
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.09561130404472351,
      "learning_rate": 1.8211932994541694e-05,
      "loss": 0.2881,
      "step": 4750
    },
    {
      "epoch": 0.9,
      "grad_norm": 23.781330108642578,
      "learning_rate": 1.8208168642951254e-05,
      "loss": 0.1772,
      "step": 4760
    },
    {
      "epoch": 0.9,
      "grad_norm": 14.084696769714355,
      "learning_rate": 1.8204404291360817e-05,
      "loss": 0.3676,
      "step": 4770
    },
    {
      "epoch": 0.9,
      "grad_norm": 6.580160617828369,
      "learning_rate": 1.8200639939770376e-05,
      "loss": 0.3677,
      "step": 4780
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1802600473165512,
      "learning_rate": 1.819687558817994e-05,
      "loss": 0.365,
      "step": 4790
    },
    {
      "epoch": 0.9,
      "grad_norm": 3.6714134216308594,
      "learning_rate": 1.81931112365895e-05,
      "loss": 0.1857,
      "step": 4800
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.0675513744354248,
      "learning_rate": 1.818934688499906e-05,
      "loss": 0.3331,
      "step": 4810
    },
    {
      "epoch": 0.91,
      "grad_norm": 22.301809310913086,
      "learning_rate": 1.8185582533408622e-05,
      "loss": 0.2882,
      "step": 4820
    },
    {
      "epoch": 0.91,
      "grad_norm": 5.7221574783325195,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.1879,
      "step": 4830
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.807055473327637,
      "learning_rate": 1.8178053830227745e-05,
      "loss": 0.3778,
      "step": 4840
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.226119041442871,
      "learning_rate": 1.8174289478637305e-05,
      "loss": 0.6181,
      "step": 4850
    },
    {
      "epoch": 0.91,
      "grad_norm": 1.7629563808441162,
      "learning_rate": 1.8170525127046868e-05,
      "loss": 0.2637,
      "step": 4860
    },
    {
      "epoch": 0.92,
      "grad_norm": 15.921305656433105,
      "learning_rate": 1.8166760775456428e-05,
      "loss": 0.3678,
      "step": 4870
    },
    {
      "epoch": 0.92,
      "grad_norm": 6.812640190124512,
      "learning_rate": 1.816299642386599e-05,
      "loss": 0.4655,
      "step": 4880
    },
    {
      "epoch": 0.92,
      "grad_norm": 3.452596426010132,
      "learning_rate": 1.815923207227555e-05,
      "loss": 0.5872,
      "step": 4890
    },
    {
      "epoch": 0.92,
      "grad_norm": 5.352724075317383,
      "learning_rate": 1.8155467720685114e-05,
      "loss": 0.3469,
      "step": 4900
    },
    {
      "epoch": 0.92,
      "grad_norm": 8.480611801147461,
      "learning_rate": 1.8151703369094674e-05,
      "loss": 0.2097,
      "step": 4910
    },
    {
      "epoch": 0.93,
      "grad_norm": 5.667534828186035,
      "learning_rate": 1.8147939017504237e-05,
      "loss": 0.3899,
      "step": 4920
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.45225715637207,
      "learning_rate": 1.81441746659138e-05,
      "loss": 0.5682,
      "step": 4930
    },
    {
      "epoch": 0.93,
      "grad_norm": 5.663761615753174,
      "learning_rate": 1.814041031432336e-05,
      "loss": 0.3908,
      "step": 4940
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.645633697509766,
      "learning_rate": 1.8136645962732923e-05,
      "loss": 0.3028,
      "step": 4950
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.09904290735721588,
      "learning_rate": 1.8132881611142483e-05,
      "loss": 0.2451,
      "step": 4960
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.155705690383911,
      "learning_rate": 1.8129117259552043e-05,
      "loss": 0.3438,
      "step": 4970
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.011092185974121,
      "learning_rate": 1.8125352907961606e-05,
      "loss": 0.4503,
      "step": 4980
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.32028958201408386,
      "learning_rate": 1.8121588556371165e-05,
      "loss": 0.2594,
      "step": 4990
    },
    {
      "epoch": 0.94,
      "grad_norm": 15.643269538879395,
      "learning_rate": 1.811782420478073e-05,
      "loss": 0.4123,
      "step": 5000
    },
    {
      "epoch": 0.94,
      "grad_norm": 20.600902557373047,
      "learning_rate": 1.811405985319029e-05,
      "loss": 0.2917,
      "step": 5010
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.0635320171713829,
      "learning_rate": 1.811029550159985e-05,
      "loss": 0.1802,
      "step": 5020
    },
    {
      "epoch": 0.95,
      "grad_norm": 8.02939510345459,
      "learning_rate": 1.810653115000941e-05,
      "loss": 0.3053,
      "step": 5030
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.3876296281814575,
      "learning_rate": 1.8102766798418974e-05,
      "loss": 0.3791,
      "step": 5040
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.618868827819824,
      "learning_rate": 1.8099002446828534e-05,
      "loss": 0.5907,
      "step": 5050
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.08411061018705368,
      "learning_rate": 1.8095238095238097e-05,
      "loss": 0.1999,
      "step": 5060
    },
    {
      "epoch": 0.95,
      "grad_norm": 16.635868072509766,
      "learning_rate": 1.8091473743647657e-05,
      "loss": 0.4062,
      "step": 5070
    },
    {
      "epoch": 0.96,
      "grad_norm": 34.237178802490234,
      "learning_rate": 1.808770939205722e-05,
      "loss": 0.3686,
      "step": 5080
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.4500527381896973,
      "learning_rate": 1.808394504046678e-05,
      "loss": 0.2305,
      "step": 5090
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.986656188964844,
      "learning_rate": 1.8080180688876343e-05,
      "loss": 0.2764,
      "step": 5100
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.787248611450195,
      "learning_rate": 1.8076416337285906e-05,
      "loss": 0.1672,
      "step": 5110
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.0679803267121315,
      "learning_rate": 1.8072651985695466e-05,
      "loss": 0.4352,
      "step": 5120
    },
    {
      "epoch": 0.97,
      "grad_norm": 9.353144645690918,
      "learning_rate": 1.806888763410503e-05,
      "loss": 0.3285,
      "step": 5130
    },
    {
      "epoch": 0.97,
      "grad_norm": 15.507201194763184,
      "learning_rate": 1.8065123282514586e-05,
      "loss": 0.4608,
      "step": 5140
    },
    {
      "epoch": 0.97,
      "grad_norm": 3.4180500507354736,
      "learning_rate": 1.806135893092415e-05,
      "loss": 0.2462,
      "step": 5150
    },
    {
      "epoch": 0.97,
      "grad_norm": 16.238187789916992,
      "learning_rate": 1.8057594579333712e-05,
      "loss": 0.2479,
      "step": 5160
    },
    {
      "epoch": 0.97,
      "grad_norm": 6.686985492706299,
      "learning_rate": 1.8053830227743272e-05,
      "loss": 0.3747,
      "step": 5170
    },
    {
      "epoch": 0.97,
      "grad_norm": 15.016100883483887,
      "learning_rate": 1.8050065876152835e-05,
      "loss": 0.4008,
      "step": 5180
    },
    {
      "epoch": 0.98,
      "grad_norm": 5.656367301940918,
      "learning_rate": 1.8046301524562395e-05,
      "loss": 0.2546,
      "step": 5190
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.6078419089317322,
      "learning_rate": 1.8042537172971958e-05,
      "loss": 0.2217,
      "step": 5200
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.915227890014648,
      "learning_rate": 1.8038772821381518e-05,
      "loss": 0.476,
      "step": 5210
    },
    {
      "epoch": 0.98,
      "grad_norm": 16.058279037475586,
      "learning_rate": 1.803500846979108e-05,
      "loss": 0.2404,
      "step": 5220
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.12465570867061615,
      "learning_rate": 1.803124411820064e-05,
      "loss": 0.3569,
      "step": 5230
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.23719528317451477,
      "learning_rate": 1.8027479766610204e-05,
      "loss": 0.1742,
      "step": 5240
    },
    {
      "epoch": 0.99,
      "grad_norm": 5.542939186096191,
      "learning_rate": 1.8023715415019763e-05,
      "loss": 0.2004,
      "step": 5250
    },
    {
      "epoch": 0.99,
      "grad_norm": 4.399273872375488,
      "learning_rate": 1.8019951063429327e-05,
      "loss": 0.339,
      "step": 5260
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.894686698913574,
      "learning_rate": 1.8016186711838886e-05,
      "loss": 0.3393,
      "step": 5270
    },
    {
      "epoch": 0.99,
      "grad_norm": 1.790652871131897,
      "learning_rate": 1.801242236024845e-05,
      "loss": 0.3117,
      "step": 5280
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.4967285692691803,
      "learning_rate": 1.800865800865801e-05,
      "loss": 0.3365,
      "step": 5290
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.29847145080566406,
      "learning_rate": 1.8004893657067572e-05,
      "loss": 0.3346,
      "step": 5300
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.142557621002197,
      "learning_rate": 1.8001129305477136e-05,
      "loss": 0.2906,
      "step": 5310
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.982,
      "eval_loss": 0.07272109389305115,
      "eval_runtime": 51.0848,
      "eval_samples_per_second": 146.815,
      "eval_steps_per_second": 18.362,
      "step": 5313
    },
    {
      "epoch": 1.0,
      "grad_norm": 32.06064987182617,
      "learning_rate": 1.7997364953886692e-05,
      "loss": 0.3597,
      "step": 5320
    },
    {
      "epoch": 1.0,
      "grad_norm": 10.492015838623047,
      "learning_rate": 1.7993600602296255e-05,
      "loss": 0.2226,
      "step": 5330
    },
    {
      "epoch": 1.01,
      "grad_norm": 7.4060211181640625,
      "learning_rate": 1.7989836250705818e-05,
      "loss": 0.2996,
      "step": 5340
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.06498083472251892,
      "learning_rate": 1.7986071899115378e-05,
      "loss": 0.2357,
      "step": 5350
    },
    {
      "epoch": 1.01,
      "grad_norm": 12.546557426452637,
      "learning_rate": 1.798230754752494e-05,
      "loss": 0.2816,
      "step": 5360
    },
    {
      "epoch": 1.01,
      "grad_norm": 1.1530576944351196,
      "learning_rate": 1.79785431959345e-05,
      "loss": 0.2292,
      "step": 5370
    },
    {
      "epoch": 1.01,
      "grad_norm": 10.668543815612793,
      "learning_rate": 1.7974778844344064e-05,
      "loss": 0.3786,
      "step": 5380
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.27988600730895996,
      "learning_rate": 1.7971014492753624e-05,
      "loss": 0.3764,
      "step": 5390
    },
    {
      "epoch": 1.02,
      "grad_norm": 12.905964851379395,
      "learning_rate": 1.7967250141163187e-05,
      "loss": 0.4022,
      "step": 5400
    },
    {
      "epoch": 1.02,
      "grad_norm": 8.057467460632324,
      "learning_rate": 1.7963485789572747e-05,
      "loss": 0.2954,
      "step": 5410
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.309386134147644,
      "learning_rate": 1.795972143798231e-05,
      "loss": 0.1192,
      "step": 5420
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.4962215423583984,
      "learning_rate": 1.795595708639187e-05,
      "loss": 0.467,
      "step": 5430
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.15860696136951447,
      "learning_rate": 1.7952192734801433e-05,
      "loss": 0.1493,
      "step": 5440
    },
    {
      "epoch": 1.03,
      "grad_norm": 14.491449356079102,
      "learning_rate": 1.7948428383210993e-05,
      "loss": 0.294,
      "step": 5450
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.07851121574640274,
      "learning_rate": 1.7944664031620556e-05,
      "loss": 0.2248,
      "step": 5460
    },
    {
      "epoch": 1.03,
      "grad_norm": 8.234402656555176,
      "learning_rate": 1.7940899680030116e-05,
      "loss": 0.4424,
      "step": 5470
    },
    {
      "epoch": 1.03,
      "grad_norm": 20.08556365966797,
      "learning_rate": 1.793713532843968e-05,
      "loss": 0.3341,
      "step": 5480
    },
    {
      "epoch": 1.03,
      "grad_norm": 15.090421676635742,
      "learning_rate": 1.793337097684924e-05,
      "loss": 0.3586,
      "step": 5490
    },
    {
      "epoch": 1.04,
      "grad_norm": 40.520565032958984,
      "learning_rate": 1.7929606625258798e-05,
      "loss": 0.5029,
      "step": 5500
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.7009685039520264,
      "learning_rate": 1.792584227366836e-05,
      "loss": 0.2354,
      "step": 5510
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.48183441162109375,
      "learning_rate": 1.792207792207792e-05,
      "loss": 0.1704,
      "step": 5520
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.48861971497535706,
      "learning_rate": 1.7918313570487484e-05,
      "loss": 0.2326,
      "step": 5530
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.09228648990392685,
      "learning_rate": 1.7914549218897047e-05,
      "loss": 0.2684,
      "step": 5540
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.060910988599061966,
      "learning_rate": 1.7910784867306607e-05,
      "loss": 0.2414,
      "step": 5550
    },
    {
      "epoch": 1.05,
      "grad_norm": 18.09152603149414,
      "learning_rate": 1.790702051571617e-05,
      "loss": 0.3235,
      "step": 5560
    },
    {
      "epoch": 1.05,
      "grad_norm": 1.8147934675216675,
      "learning_rate": 1.790325616412573e-05,
      "loss": 0.4586,
      "step": 5570
    },
    {
      "epoch": 1.05,
      "grad_norm": 10.438521385192871,
      "learning_rate": 1.7899491812535293e-05,
      "loss": 0.1875,
      "step": 5580
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.06755027174949646,
      "learning_rate": 1.7895727460944853e-05,
      "loss": 0.2427,
      "step": 5590
    },
    {
      "epoch": 1.05,
      "grad_norm": 7.72373628616333,
      "learning_rate": 1.7891963109354416e-05,
      "loss": 0.3591,
      "step": 5600
    },
    {
      "epoch": 1.06,
      "grad_norm": 23.316707611083984,
      "learning_rate": 1.7888198757763976e-05,
      "loss": 0.3075,
      "step": 5610
    },
    {
      "epoch": 1.06,
      "grad_norm": 6.044504165649414,
      "learning_rate": 1.788443440617354e-05,
      "loss": 0.3666,
      "step": 5620
    },
    {
      "epoch": 1.06,
      "grad_norm": 4.240804195404053,
      "learning_rate": 1.78806700545831e-05,
      "loss": 0.3303,
      "step": 5630
    },
    {
      "epoch": 1.06,
      "grad_norm": 13.91414737701416,
      "learning_rate": 1.7876905702992662e-05,
      "loss": 0.5335,
      "step": 5640
    },
    {
      "epoch": 1.06,
      "grad_norm": 22.953662872314453,
      "learning_rate": 1.7873141351402222e-05,
      "loss": 0.4226,
      "step": 5650
    },
    {
      "epoch": 1.07,
      "grad_norm": 17.735177993774414,
      "learning_rate": 1.7869376999811785e-05,
      "loss": 0.3761,
      "step": 5660
    },
    {
      "epoch": 1.07,
      "grad_norm": 12.626730918884277,
      "learning_rate": 1.7865612648221345e-05,
      "loss": 0.3366,
      "step": 5670
    },
    {
      "epoch": 1.07,
      "grad_norm": 7.499827861785889,
      "learning_rate": 1.7861848296630905e-05,
      "loss": 0.1857,
      "step": 5680
    },
    {
      "epoch": 1.07,
      "grad_norm": 18.860071182250977,
      "learning_rate": 1.7858083945040468e-05,
      "loss": 0.415,
      "step": 5690
    },
    {
      "epoch": 1.07,
      "grad_norm": 2.7299387454986572,
      "learning_rate": 1.7854319593450027e-05,
      "loss": 0.2967,
      "step": 5700
    },
    {
      "epoch": 1.07,
      "grad_norm": 1.9738844633102417,
      "learning_rate": 1.785055524185959e-05,
      "loss": 0.2457,
      "step": 5710
    },
    {
      "epoch": 1.08,
      "grad_norm": 18.41208839416504,
      "learning_rate": 1.7846790890269154e-05,
      "loss": 0.2698,
      "step": 5720
    },
    {
      "epoch": 1.08,
      "grad_norm": 6.7303900718688965,
      "learning_rate": 1.7843026538678714e-05,
      "loss": 0.2575,
      "step": 5730
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.12529726326465607,
      "learning_rate": 1.7839262187088277e-05,
      "loss": 0.4166,
      "step": 5740
    },
    {
      "epoch": 1.08,
      "grad_norm": 14.660829544067383,
      "learning_rate": 1.7835497835497836e-05,
      "loss": 0.3122,
      "step": 5750
    },
    {
      "epoch": 1.08,
      "grad_norm": 7.67722749710083,
      "learning_rate": 1.78317334839074e-05,
      "loss": 0.4591,
      "step": 5760
    },
    {
      "epoch": 1.09,
      "grad_norm": 6.773375511169434,
      "learning_rate": 1.782796913231696e-05,
      "loss": 0.3175,
      "step": 5770
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.21152108907699585,
      "learning_rate": 1.7824204780726523e-05,
      "loss": 0.175,
      "step": 5780
    },
    {
      "epoch": 1.09,
      "grad_norm": 1.0829007625579834,
      "learning_rate": 1.7820440429136082e-05,
      "loss": 0.179,
      "step": 5790
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.12647803127765656,
      "learning_rate": 1.7816676077545645e-05,
      "loss": 0.2343,
      "step": 5800
    },
    {
      "epoch": 1.09,
      "grad_norm": 4.451048851013184,
      "learning_rate": 1.7812911725955205e-05,
      "loss": 0.1965,
      "step": 5810
    },
    {
      "epoch": 1.1,
      "grad_norm": 15.420040130615234,
      "learning_rate": 1.780914737436477e-05,
      "loss": 0.4636,
      "step": 5820
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.901856005191803,
      "learning_rate": 1.7805383022774328e-05,
      "loss": 0.4255,
      "step": 5830
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.17663244903087616,
      "learning_rate": 1.7801618671183888e-05,
      "loss": 0.2197,
      "step": 5840
    },
    {
      "epoch": 1.1,
      "grad_norm": 28.10008430480957,
      "learning_rate": 1.779785431959345e-05,
      "loss": 0.2445,
      "step": 5850
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.7232682108879089,
      "learning_rate": 1.779408996800301e-05,
      "loss": 0.2407,
      "step": 5860
    },
    {
      "epoch": 1.1,
      "grad_norm": 17.44423484802246,
      "learning_rate": 1.7790325616412574e-05,
      "loss": 0.3047,
      "step": 5870
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.590700149536133,
      "learning_rate": 1.7786561264822134e-05,
      "loss": 0.3666,
      "step": 5880
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.543517112731934,
      "learning_rate": 1.7782796913231697e-05,
      "loss": 0.4344,
      "step": 5890
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.03803512454032898,
      "learning_rate": 1.777903256164126e-05,
      "loss": 0.3141,
      "step": 5900
    },
    {
      "epoch": 1.11,
      "grad_norm": 10.419922828674316,
      "learning_rate": 1.777526821005082e-05,
      "loss": 0.1945,
      "step": 5910
    },
    {
      "epoch": 1.11,
      "grad_norm": 8.903624534606934,
      "learning_rate": 1.7771503858460383e-05,
      "loss": 0.4987,
      "step": 5920
    },
    {
      "epoch": 1.12,
      "grad_norm": 12.289496421813965,
      "learning_rate": 1.7767739506869943e-05,
      "loss": 0.3577,
      "step": 5930
    },
    {
      "epoch": 1.12,
      "grad_norm": 14.816337585449219,
      "learning_rate": 1.7763975155279506e-05,
      "loss": 0.3124,
      "step": 5940
    },
    {
      "epoch": 1.12,
      "grad_norm": 8.756638526916504,
      "learning_rate": 1.7760210803689066e-05,
      "loss": 0.331,
      "step": 5950
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.9256272315979,
      "learning_rate": 1.775644645209863e-05,
      "loss": 0.314,
      "step": 5960
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.0664701461791992,
      "learning_rate": 1.775268210050819e-05,
      "loss": 0.2971,
      "step": 5970
    },
    {
      "epoch": 1.13,
      "grad_norm": 17.4043025970459,
      "learning_rate": 1.7748917748917752e-05,
      "loss": 0.2575,
      "step": 5980
    },
    {
      "epoch": 1.13,
      "grad_norm": 1.7042344808578491,
      "learning_rate": 1.774515339732731e-05,
      "loss": 0.1803,
      "step": 5990
    },
    {
      "epoch": 1.13,
      "grad_norm": 4.82717227935791,
      "learning_rate": 1.7741389045736875e-05,
      "loss": 0.3587,
      "step": 6000
    },
    {
      "epoch": 1.13,
      "grad_norm": 32.751041412353516,
      "learning_rate": 1.7737624694146434e-05,
      "loss": 0.1633,
      "step": 6010
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.17549344897270203,
      "learning_rate": 1.7733860342555994e-05,
      "loss": 0.2987,
      "step": 6020
    },
    {
      "epoch": 1.13,
      "grad_norm": 20.089223861694336,
      "learning_rate": 1.7730095990965557e-05,
      "loss": 0.4031,
      "step": 6030
    },
    {
      "epoch": 1.14,
      "grad_norm": 17.87177276611328,
      "learning_rate": 1.7726331639375117e-05,
      "loss": 0.3605,
      "step": 6040
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.05802224203944206,
      "learning_rate": 1.772256728778468e-05,
      "loss": 0.2113,
      "step": 6050
    },
    {
      "epoch": 1.14,
      "grad_norm": 27.633373260498047,
      "learning_rate": 1.771880293619424e-05,
      "loss": 0.458,
      "step": 6060
    },
    {
      "epoch": 1.14,
      "grad_norm": 13.443075180053711,
      "learning_rate": 1.7715038584603803e-05,
      "loss": 0.3516,
      "step": 6070
    },
    {
      "epoch": 1.14,
      "grad_norm": 20.25200843811035,
      "learning_rate": 1.7711274233013366e-05,
      "loss": 0.3123,
      "step": 6080
    },
    {
      "epoch": 1.15,
      "grad_norm": 5.744795322418213,
      "learning_rate": 1.7707509881422926e-05,
      "loss": 0.4988,
      "step": 6090
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.810657024383545,
      "learning_rate": 1.770374552983249e-05,
      "loss": 0.2438,
      "step": 6100
    },
    {
      "epoch": 1.15,
      "grad_norm": 23.06705093383789,
      "learning_rate": 1.769998117824205e-05,
      "loss": 0.3493,
      "step": 6110
    },
    {
      "epoch": 1.15,
      "grad_norm": 6.4262871742248535,
      "learning_rate": 1.7696216826651612e-05,
      "loss": 0.2268,
      "step": 6120
    },
    {
      "epoch": 1.15,
      "grad_norm": 5.230250835418701,
      "learning_rate": 1.7692452475061172e-05,
      "loss": 0.1774,
      "step": 6130
    },
    {
      "epoch": 1.16,
      "grad_norm": 15.252341270446777,
      "learning_rate": 1.7688688123470735e-05,
      "loss": 0.2452,
      "step": 6140
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.05415533483028412,
      "learning_rate": 1.7684923771880295e-05,
      "loss": 0.1875,
      "step": 6150
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.049749597907066345,
      "learning_rate": 1.7681159420289858e-05,
      "loss": 0.1644,
      "step": 6160
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.25254836678504944,
      "learning_rate": 1.7677395068699418e-05,
      "loss": 0.2607,
      "step": 6170
    },
    {
      "epoch": 1.16,
      "grad_norm": 9.684237480163574,
      "learning_rate": 1.767363071710898e-05,
      "loss": 0.2353,
      "step": 6180
    },
    {
      "epoch": 1.17,
      "grad_norm": 27.44820785522461,
      "learning_rate": 1.766986636551854e-05,
      "loss": 0.317,
      "step": 6190
    },
    {
      "epoch": 1.17,
      "grad_norm": 7.580427646636963,
      "learning_rate": 1.76661020139281e-05,
      "loss": 0.2825,
      "step": 6200
    },
    {
      "epoch": 1.17,
      "grad_norm": 6.638911724090576,
      "learning_rate": 1.7662337662337664e-05,
      "loss": 0.2671,
      "step": 6210
    },
    {
      "epoch": 1.17,
      "grad_norm": 7.344533920288086,
      "learning_rate": 1.7658573310747223e-05,
      "loss": 0.4158,
      "step": 6220
    },
    {
      "epoch": 1.17,
      "grad_norm": 21.25824737548828,
      "learning_rate": 1.7654808959156787e-05,
      "loss": 0.355,
      "step": 6230
    },
    {
      "epoch": 1.17,
      "grad_norm": 38.96686553955078,
      "learning_rate": 1.7651044607566346e-05,
      "loss": 0.2805,
      "step": 6240
    },
    {
      "epoch": 1.18,
      "grad_norm": 21.923730850219727,
      "learning_rate": 1.764728025597591e-05,
      "loss": 0.1912,
      "step": 6250
    },
    {
      "epoch": 1.18,
      "grad_norm": 2.206803560256958,
      "learning_rate": 1.764351590438547e-05,
      "loss": 0.1947,
      "step": 6260
    },
    {
      "epoch": 1.18,
      "grad_norm": 12.114392280578613,
      "learning_rate": 1.7639751552795032e-05,
      "loss": 0.3474,
      "step": 6270
    },
    {
      "epoch": 1.18,
      "grad_norm": 7.857627868652344,
      "learning_rate": 1.7635987201204596e-05,
      "loss": 0.2447,
      "step": 6280
    },
    {
      "epoch": 1.18,
      "grad_norm": 9.643177032470703,
      "learning_rate": 1.7632222849614155e-05,
      "loss": 0.3523,
      "step": 6290
    },
    {
      "epoch": 1.19,
      "grad_norm": 2.706428050994873,
      "learning_rate": 1.762845849802372e-05,
      "loss": 0.3901,
      "step": 6300
    },
    {
      "epoch": 1.19,
      "grad_norm": 12.305505752563477,
      "learning_rate": 1.7624694146433278e-05,
      "loss": 0.2936,
      "step": 6310
    },
    {
      "epoch": 1.19,
      "grad_norm": 9.531702995300293,
      "learning_rate": 1.762092979484284e-05,
      "loss": 0.4113,
      "step": 6320
    },
    {
      "epoch": 1.19,
      "grad_norm": 1.1557501554489136,
      "learning_rate": 1.76171654432524e-05,
      "loss": 0.2235,
      "step": 6330
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.36906567215919495,
      "learning_rate": 1.7613401091661964e-05,
      "loss": 0.335,
      "step": 6340
    },
    {
      "epoch": 1.2,
      "grad_norm": 29.542802810668945,
      "learning_rate": 1.7609636740071524e-05,
      "loss": 0.2363,
      "step": 6350
    },
    {
      "epoch": 1.2,
      "grad_norm": 14.587756156921387,
      "learning_rate": 1.7605872388481084e-05,
      "loss": 0.4073,
      "step": 6360
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.593445777893066,
      "learning_rate": 1.7602108036890647e-05,
      "loss": 0.3196,
      "step": 6370
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.05206476151943207,
      "learning_rate": 1.7598343685300207e-05,
      "loss": 0.3597,
      "step": 6380
    },
    {
      "epoch": 1.2,
      "grad_norm": 14.10525894165039,
      "learning_rate": 1.759457933370977e-05,
      "loss": 0.1895,
      "step": 6390
    },
    {
      "epoch": 1.2,
      "grad_norm": 5.895766735076904,
      "learning_rate": 1.759081498211933e-05,
      "loss": 0.2861,
      "step": 6400
    },
    {
      "epoch": 1.21,
      "grad_norm": 3.7865443229675293,
      "learning_rate": 1.7587050630528893e-05,
      "loss": 0.1253,
      "step": 6410
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.6739709377288818,
      "learning_rate": 1.7583286278938453e-05,
      "loss": 0.326,
      "step": 6420
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.2942664623260498,
      "learning_rate": 1.7579521927348016e-05,
      "loss": 0.2794,
      "step": 6430
    },
    {
      "epoch": 1.21,
      "grad_norm": 11.703619003295898,
      "learning_rate": 1.7575757575757576e-05,
      "loss": 0.3833,
      "step": 6440
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.1144811138510704,
      "learning_rate": 1.757199322416714e-05,
      "loss": 0.4558,
      "step": 6450
    },
    {
      "epoch": 1.22,
      "grad_norm": 44.68494415283203,
      "learning_rate": 1.7568228872576702e-05,
      "loss": 0.1525,
      "step": 6460
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.615532636642456,
      "learning_rate": 1.756446452098626e-05,
      "loss": 0.2398,
      "step": 6470
    },
    {
      "epoch": 1.22,
      "grad_norm": 4.4646406173706055,
      "learning_rate": 1.7560700169395825e-05,
      "loss": 0.545,
      "step": 6480
    },
    {
      "epoch": 1.22,
      "grad_norm": 5.854625225067139,
      "learning_rate": 1.7556935817805385e-05,
      "loss": 0.3327,
      "step": 6490
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.048285141587257385,
      "learning_rate": 1.7553171466214948e-05,
      "loss": 0.2132,
      "step": 6500
    },
    {
      "epoch": 1.23,
      "grad_norm": 14.727179527282715,
      "learning_rate": 1.7549407114624507e-05,
      "loss": 0.2916,
      "step": 6510
    },
    {
      "epoch": 1.23,
      "grad_norm": 6.532195568084717,
      "learning_rate": 1.754564276303407e-05,
      "loss": 0.3357,
      "step": 6520
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.2176375687122345,
      "learning_rate": 1.754187841144363e-05,
      "loss": 0.2793,
      "step": 6530
    },
    {
      "epoch": 1.23,
      "grad_norm": 3.899984121322632,
      "learning_rate": 1.753811405985319e-05,
      "loss": 0.3413,
      "step": 6540
    },
    {
      "epoch": 1.23,
      "grad_norm": 10.481951713562012,
      "learning_rate": 1.7534349708262753e-05,
      "loss": 0.2153,
      "step": 6550
    },
    {
      "epoch": 1.23,
      "grad_norm": 1.294416904449463,
      "learning_rate": 1.7530585356672313e-05,
      "loss": 0.2428,
      "step": 6560
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.10451066493988037,
      "learning_rate": 1.7526821005081876e-05,
      "loss": 0.1475,
      "step": 6570
    },
    {
      "epoch": 1.24,
      "grad_norm": 17.172317504882812,
      "learning_rate": 1.7523056653491436e-05,
      "loss": 0.1532,
      "step": 6580
    },
    {
      "epoch": 1.24,
      "grad_norm": 13.984004020690918,
      "learning_rate": 1.7519292301901e-05,
      "loss": 0.6267,
      "step": 6590
    },
    {
      "epoch": 1.24,
      "grad_norm": 3.260589122772217,
      "learning_rate": 1.751552795031056e-05,
      "loss": 0.1099,
      "step": 6600
    },
    {
      "epoch": 1.24,
      "grad_norm": 33.7536506652832,
      "learning_rate": 1.7511763598720122e-05,
      "loss": 0.4407,
      "step": 6610
    },
    {
      "epoch": 1.25,
      "grad_norm": 23.979394912719727,
      "learning_rate": 1.7507999247129682e-05,
      "loss": 0.3299,
      "step": 6620
    },
    {
      "epoch": 1.25,
      "grad_norm": 13.720709800720215,
      "learning_rate": 1.7504234895539245e-05,
      "loss": 0.2525,
      "step": 6630
    },
    {
      "epoch": 1.25,
      "grad_norm": 8.97732162475586,
      "learning_rate": 1.7500470543948808e-05,
      "loss": 0.3157,
      "step": 6640
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.060696665197610855,
      "learning_rate": 1.7496706192358368e-05,
      "loss": 0.4767,
      "step": 6650
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.10852402448654175,
      "learning_rate": 1.749294184076793e-05,
      "loss": 0.3016,
      "step": 6660
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.20572949945926666,
      "learning_rate": 1.748917748917749e-05,
      "loss": 0.3479,
      "step": 6670
    },
    {
      "epoch": 1.26,
      "grad_norm": 2.1876344680786133,
      "learning_rate": 1.7485413137587054e-05,
      "loss": 0.2702,
      "step": 6680
    },
    {
      "epoch": 1.26,
      "grad_norm": 22.546932220458984,
      "learning_rate": 1.7481648785996614e-05,
      "loss": 0.2548,
      "step": 6690
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.7548444867134094,
      "learning_rate": 1.7477884434406177e-05,
      "loss": 0.4816,
      "step": 6700
    },
    {
      "epoch": 1.26,
      "grad_norm": 35.48444366455078,
      "learning_rate": 1.7474120082815737e-05,
      "loss": 0.3318,
      "step": 6710
    },
    {
      "epoch": 1.26,
      "grad_norm": 19.06321144104004,
      "learning_rate": 1.7470355731225296e-05,
      "loss": 0.2922,
      "step": 6720
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.10732751339673996,
      "learning_rate": 1.746659137963486e-05,
      "loss": 0.159,
      "step": 6730
    },
    {
      "epoch": 1.27,
      "grad_norm": 6.251004695892334,
      "learning_rate": 1.746282702804442e-05,
      "loss": 0.2194,
      "step": 6740
    },
    {
      "epoch": 1.27,
      "grad_norm": 9.226393699645996,
      "learning_rate": 1.7459062676453983e-05,
      "loss": 0.2429,
      "step": 6750
    },
    {
      "epoch": 1.27,
      "grad_norm": 5.944048881530762,
      "learning_rate": 1.7455298324863542e-05,
      "loss": 0.2859,
      "step": 6760
    },
    {
      "epoch": 1.27,
      "grad_norm": 1.030680775642395,
      "learning_rate": 1.7451533973273105e-05,
      "loss": 0.2315,
      "step": 6770
    },
    {
      "epoch": 1.28,
      "grad_norm": 11.144603729248047,
      "learning_rate": 1.7447769621682665e-05,
      "loss": 0.5212,
      "step": 6780
    },
    {
      "epoch": 1.28,
      "grad_norm": 27.1265811920166,
      "learning_rate": 1.744400527009223e-05,
      "loss": 0.4771,
      "step": 6790
    },
    {
      "epoch": 1.28,
      "grad_norm": 12.093399047851562,
      "learning_rate": 1.7440240918501788e-05,
      "loss": 0.2112,
      "step": 6800
    },
    {
      "epoch": 1.28,
      "grad_norm": 15.382379531860352,
      "learning_rate": 1.743647656691135e-05,
      "loss": 0.3319,
      "step": 6810
    },
    {
      "epoch": 1.28,
      "grad_norm": 3.984701633453369,
      "learning_rate": 1.743271221532091e-05,
      "loss": 0.1868,
      "step": 6820
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.4144419729709625,
      "learning_rate": 1.7428947863730474e-05,
      "loss": 0.125,
      "step": 6830
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.045936811715364456,
      "learning_rate": 1.7425183512140037e-05,
      "loss": 0.199,
      "step": 6840
    },
    {
      "epoch": 1.29,
      "grad_norm": 8.621210098266602,
      "learning_rate": 1.7421419160549597e-05,
      "loss": 0.1761,
      "step": 6850
    },
    {
      "epoch": 1.29,
      "grad_norm": 1.2863411903381348,
      "learning_rate": 1.741765480895916e-05,
      "loss": 0.1838,
      "step": 6860
    },
    {
      "epoch": 1.29,
      "grad_norm": 18.248380661010742,
      "learning_rate": 1.741389045736872e-05,
      "loss": 0.1607,
      "step": 6870
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.08692477643489838,
      "learning_rate": 1.7410126105778283e-05,
      "loss": 0.3262,
      "step": 6880
    },
    {
      "epoch": 1.3,
      "grad_norm": 5.438815116882324,
      "learning_rate": 1.7406361754187843e-05,
      "loss": 0.3421,
      "step": 6890
    },
    {
      "epoch": 1.3,
      "grad_norm": 14.672286033630371,
      "learning_rate": 1.7402597402597403e-05,
      "loss": 0.1204,
      "step": 6900
    },
    {
      "epoch": 1.3,
      "grad_norm": 19.01220703125,
      "learning_rate": 1.7398833051006966e-05,
      "loss": 0.3222,
      "step": 6910
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.053402259945869446,
      "learning_rate": 1.7395068699416526e-05,
      "loss": 0.1932,
      "step": 6920
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.2856398820877075,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.228,
      "step": 6930
    },
    {
      "epoch": 1.31,
      "grad_norm": 35.01115036010742,
      "learning_rate": 1.738753999623565e-05,
      "loss": 0.2398,
      "step": 6940
    },
    {
      "epoch": 1.31,
      "grad_norm": 3.1591899394989014,
      "learning_rate": 1.7383775644645212e-05,
      "loss": 0.2069,
      "step": 6950
    },
    {
      "epoch": 1.31,
      "grad_norm": 2.1366374492645264,
      "learning_rate": 1.738001129305477e-05,
      "loss": 0.1715,
      "step": 6960
    },
    {
      "epoch": 1.31,
      "grad_norm": 1.7839075326919556,
      "learning_rate": 1.7376246941464335e-05,
      "loss": 0.2071,
      "step": 6970
    },
    {
      "epoch": 1.31,
      "grad_norm": 16.320722579956055,
      "learning_rate": 1.7372482589873894e-05,
      "loss": 0.1907,
      "step": 6980
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.4329834580421448,
      "learning_rate": 1.7368718238283458e-05,
      "loss": 0.2524,
      "step": 6990
    },
    {
      "epoch": 1.32,
      "grad_norm": 21.901504516601562,
      "learning_rate": 1.7364953886693017e-05,
      "loss": 0.1725,
      "step": 7000
    },
    {
      "epoch": 1.32,
      "grad_norm": 17.68385124206543,
      "learning_rate": 1.736118953510258e-05,
      "loss": 0.384,
      "step": 7010
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.611810564994812,
      "learning_rate": 1.7357425183512144e-05,
      "loss": 0.296,
      "step": 7020
    },
    {
      "epoch": 1.32,
      "grad_norm": 30.460451126098633,
      "learning_rate": 1.7353660831921703e-05,
      "loss": 0.4383,
      "step": 7030
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.03811188042163849,
      "learning_rate": 1.7349896480331267e-05,
      "loss": 0.1684,
      "step": 7040
    },
    {
      "epoch": 1.33,
      "grad_norm": 1.3913265466690063,
      "learning_rate": 1.7346132128740826e-05,
      "loss": 0.4592,
      "step": 7050
    },
    {
      "epoch": 1.33,
      "grad_norm": 24.42656135559082,
      "learning_rate": 1.7342367777150386e-05,
      "loss": 0.3685,
      "step": 7060
    },
    {
      "epoch": 1.33,
      "grad_norm": 12.958415985107422,
      "learning_rate": 1.733860342555995e-05,
      "loss": 0.3574,
      "step": 7070
    },
    {
      "epoch": 1.33,
      "grad_norm": 5.067093849182129,
      "learning_rate": 1.733483907396951e-05,
      "loss": 0.2793,
      "step": 7080
    },
    {
      "epoch": 1.33,
      "grad_norm": 15.73208999633789,
      "learning_rate": 1.7331074722379072e-05,
      "loss": 0.3948,
      "step": 7090
    },
    {
      "epoch": 1.34,
      "grad_norm": 15.483287811279297,
      "learning_rate": 1.7327310370788632e-05,
      "loss": 0.4851,
      "step": 7100
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.6781468391418457,
      "learning_rate": 1.7323546019198195e-05,
      "loss": 0.2212,
      "step": 7110
    },
    {
      "epoch": 1.34,
      "grad_norm": 7.565657138824463,
      "learning_rate": 1.7319781667607755e-05,
      "loss": 0.4148,
      "step": 7120
    },
    {
      "epoch": 1.34,
      "grad_norm": 25.396787643432617,
      "learning_rate": 1.7316017316017318e-05,
      "loss": 0.1434,
      "step": 7130
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.032858707010746,
      "learning_rate": 1.7312252964426878e-05,
      "loss": 0.3181,
      "step": 7140
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.03974176570773125,
      "learning_rate": 1.730848861283644e-05,
      "loss": 0.213,
      "step": 7150
    },
    {
      "epoch": 1.35,
      "grad_norm": 6.578794002532959,
      "learning_rate": 1.7304724261246e-05,
      "loss": 0.3964,
      "step": 7160
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.20288939774036407,
      "learning_rate": 1.7300959909655564e-05,
      "loss": 0.2931,
      "step": 7170
    },
    {
      "epoch": 1.35,
      "grad_norm": 30.366374969482422,
      "learning_rate": 1.7297195558065124e-05,
      "loss": 0.3999,
      "step": 7180
    },
    {
      "epoch": 1.35,
      "grad_norm": 3.8586971759796143,
      "learning_rate": 1.7293431206474687e-05,
      "loss": 0.261,
      "step": 7190
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.345582008361816,
      "learning_rate": 1.728966685488425e-05,
      "loss": 0.3604,
      "step": 7200
    },
    {
      "epoch": 1.36,
      "grad_norm": 27.920082092285156,
      "learning_rate": 1.728590250329381e-05,
      "loss": 0.3172,
      "step": 7210
    },
    {
      "epoch": 1.36,
      "grad_norm": 10.259176254272461,
      "learning_rate": 1.7282138151703373e-05,
      "loss": 0.1658,
      "step": 7220
    },
    {
      "epoch": 1.36,
      "grad_norm": 9.395998001098633,
      "learning_rate": 1.727837380011293e-05,
      "loss": 0.2039,
      "step": 7230
    },
    {
      "epoch": 1.36,
      "grad_norm": 24.532428741455078,
      "learning_rate": 1.7274609448522492e-05,
      "loss": 0.2799,
      "step": 7240
    },
    {
      "epoch": 1.36,
      "grad_norm": 11.422987937927246,
      "learning_rate": 1.7270845096932056e-05,
      "loss": 0.4088,
      "step": 7250
    },
    {
      "epoch": 1.37,
      "grad_norm": 9.457900047302246,
      "learning_rate": 1.7267080745341615e-05,
      "loss": 0.2298,
      "step": 7260
    },
    {
      "epoch": 1.37,
      "grad_norm": 5.66178560256958,
      "learning_rate": 1.726331639375118e-05,
      "loss": 0.3135,
      "step": 7270
    },
    {
      "epoch": 1.37,
      "grad_norm": 17.641002655029297,
      "learning_rate": 1.7259552042160738e-05,
      "loss": 0.3586,
      "step": 7280
    },
    {
      "epoch": 1.37,
      "grad_norm": 18.152624130249023,
      "learning_rate": 1.72557876905703e-05,
      "loss": 0.3042,
      "step": 7290
    },
    {
      "epoch": 1.37,
      "grad_norm": 14.392729759216309,
      "learning_rate": 1.725202333897986e-05,
      "loss": 0.2784,
      "step": 7300
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.5259966254234314,
      "learning_rate": 1.7248258987389424e-05,
      "loss": 0.5443,
      "step": 7310
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.3994278311729431,
      "learning_rate": 1.7244494635798984e-05,
      "loss": 0.2466,
      "step": 7320
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.05281887203454971,
      "learning_rate": 1.7240730284208547e-05,
      "loss": 0.1216,
      "step": 7330
    },
    {
      "epoch": 1.38,
      "grad_norm": 7.929999828338623,
      "learning_rate": 1.7236965932618107e-05,
      "loss": 0.34,
      "step": 7340
    },
    {
      "epoch": 1.38,
      "grad_norm": 6.2646484375,
      "learning_rate": 1.723320158102767e-05,
      "loss": 0.3634,
      "step": 7350
    },
    {
      "epoch": 1.39,
      "grad_norm": 15.034292221069336,
      "learning_rate": 1.722943722943723e-05,
      "loss": 0.4079,
      "step": 7360
    },
    {
      "epoch": 1.39,
      "grad_norm": 26.86743927001953,
      "learning_rate": 1.7225672877846793e-05,
      "loss": 0.3783,
      "step": 7370
    },
    {
      "epoch": 1.39,
      "grad_norm": 20.575267791748047,
      "learning_rate": 1.7221908526256356e-05,
      "loss": 0.171,
      "step": 7380
    },
    {
      "epoch": 1.39,
      "grad_norm": 23.95577049255371,
      "learning_rate": 1.7218144174665916e-05,
      "loss": 0.1819,
      "step": 7390
    },
    {
      "epoch": 1.39,
      "grad_norm": 4.162304878234863,
      "learning_rate": 1.721437982307548e-05,
      "loss": 0.2595,
      "step": 7400
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.04258643463253975,
      "learning_rate": 1.7210615471485036e-05,
      "loss": 0.2228,
      "step": 7410
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.6805218458175659,
      "learning_rate": 1.72068511198946e-05,
      "loss": 0.2522,
      "step": 7420
    },
    {
      "epoch": 1.4,
      "grad_norm": 11.200102806091309,
      "learning_rate": 1.7203086768304162e-05,
      "loss": 0.1603,
      "step": 7430
    },
    {
      "epoch": 1.4,
      "grad_norm": 13.226347923278809,
      "learning_rate": 1.719932241671372e-05,
      "loss": 0.3612,
      "step": 7440
    },
    {
      "epoch": 1.4,
      "grad_norm": 4.923086166381836,
      "learning_rate": 1.7195558065123285e-05,
      "loss": 0.2719,
      "step": 7450
    },
    {
      "epoch": 1.4,
      "grad_norm": 3.7026889324188232,
      "learning_rate": 1.7191793713532845e-05,
      "loss": 0.2951,
      "step": 7460
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.2390275001525879,
      "learning_rate": 1.7188029361942408e-05,
      "loss": 0.2423,
      "step": 7470
    },
    {
      "epoch": 1.41,
      "grad_norm": 4.925628662109375,
      "learning_rate": 1.7184265010351967e-05,
      "loss": 0.3349,
      "step": 7480
    },
    {
      "epoch": 1.41,
      "grad_norm": 11.171341896057129,
      "learning_rate": 1.718050065876153e-05,
      "loss": 0.1886,
      "step": 7490
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.08292572945356369,
      "learning_rate": 1.717673630717109e-05,
      "loss": 0.304,
      "step": 7500
    },
    {
      "epoch": 1.41,
      "grad_norm": 12.082964897155762,
      "learning_rate": 1.7172971955580654e-05,
      "loss": 0.3931,
      "step": 7510
    },
    {
      "epoch": 1.42,
      "grad_norm": 13.223511695861816,
      "learning_rate": 1.7169207603990213e-05,
      "loss": 0.3183,
      "step": 7520
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.2502141296863556,
      "learning_rate": 1.7165443252399776e-05,
      "loss": 0.2177,
      "step": 7530
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.24700814485549927,
      "learning_rate": 1.7161678900809336e-05,
      "loss": 0.3078,
      "step": 7540
    },
    {
      "epoch": 1.42,
      "grad_norm": 2.143354892730713,
      "learning_rate": 1.71579145492189e-05,
      "loss": 0.1793,
      "step": 7550
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.0370979830622673,
      "learning_rate": 1.715415019762846e-05,
      "loss": 0.2082,
      "step": 7560
    },
    {
      "epoch": 1.42,
      "grad_norm": 16.83991241455078,
      "learning_rate": 1.7150385846038022e-05,
      "loss": 0.2277,
      "step": 7570
    },
    {
      "epoch": 1.43,
      "grad_norm": 7.546878337860107,
      "learning_rate": 1.7146621494447582e-05,
      "loss": 0.2537,
      "step": 7580
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.17173852026462555,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 0.2453,
      "step": 7590
    },
    {
      "epoch": 1.43,
      "grad_norm": 11.809530258178711,
      "learning_rate": 1.7139092791266705e-05,
      "loss": 0.3286,
      "step": 7600
    },
    {
      "epoch": 1.43,
      "grad_norm": 6.450569152832031,
      "learning_rate": 1.7135328439676268e-05,
      "loss": 0.1443,
      "step": 7610
    },
    {
      "epoch": 1.43,
      "grad_norm": 24.40304183959961,
      "learning_rate": 1.7131564088085828e-05,
      "loss": 0.4091,
      "step": 7620
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.0310637466609478,
      "learning_rate": 1.712779973649539e-05,
      "loss": 0.3248,
      "step": 7630
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.098997116088867,
      "learning_rate": 1.712403538490495e-05,
      "loss": 0.3954,
      "step": 7640
    },
    {
      "epoch": 1.44,
      "grad_norm": 8.263358116149902,
      "learning_rate": 1.7120271033314514e-05,
      "loss": 0.2409,
      "step": 7650
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.03411370515823364,
      "learning_rate": 1.7116506681724074e-05,
      "loss": 0.2386,
      "step": 7660
    },
    {
      "epoch": 1.44,
      "grad_norm": 5.3482136726379395,
      "learning_rate": 1.7112742330133637e-05,
      "loss": 0.2472,
      "step": 7670
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.7274813055992126,
      "learning_rate": 1.7108977978543197e-05,
      "loss": 0.3562,
      "step": 7680
    },
    {
      "epoch": 1.45,
      "grad_norm": 8.806757926940918,
      "learning_rate": 1.710521362695276e-05,
      "loss": 0.3652,
      "step": 7690
    },
    {
      "epoch": 1.45,
      "grad_norm": 26.512208938598633,
      "learning_rate": 1.710144927536232e-05,
      "loss": 0.194,
      "step": 7700
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.30574938654899597,
      "learning_rate": 1.7097684923771883e-05,
      "loss": 0.4424,
      "step": 7710
    },
    {
      "epoch": 1.45,
      "grad_norm": 11.667891502380371,
      "learning_rate": 1.7093920572181443e-05,
      "loss": 0.3922,
      "step": 7720
    },
    {
      "epoch": 1.45,
      "grad_norm": 13.048910140991211,
      "learning_rate": 1.7090156220591006e-05,
      "loss": 0.3342,
      "step": 7730
    },
    {
      "epoch": 1.46,
      "grad_norm": 8.711930274963379,
      "learning_rate": 1.7086391869000565e-05,
      "loss": 0.3395,
      "step": 7740
    },
    {
      "epoch": 1.46,
      "grad_norm": 13.956701278686523,
      "learning_rate": 1.708262751741013e-05,
      "loss": 0.2758,
      "step": 7750
    },
    {
      "epoch": 1.46,
      "grad_norm": 2.6432230472564697,
      "learning_rate": 1.707886316581969e-05,
      "loss": 0.2649,
      "step": 7760
    },
    {
      "epoch": 1.46,
      "grad_norm": 11.148094177246094,
      "learning_rate": 1.7075098814229248e-05,
      "loss": 0.1806,
      "step": 7770
    },
    {
      "epoch": 1.46,
      "grad_norm": 23.22715950012207,
      "learning_rate": 1.707133446263881e-05,
      "loss": 0.2528,
      "step": 7780
    },
    {
      "epoch": 1.47,
      "grad_norm": 17.26124382019043,
      "learning_rate": 1.706757011104837e-05,
      "loss": 0.4776,
      "step": 7790
    },
    {
      "epoch": 1.47,
      "grad_norm": 2.2572834491729736,
      "learning_rate": 1.7063805759457934e-05,
      "loss": 0.2748,
      "step": 7800
    },
    {
      "epoch": 1.47,
      "grad_norm": 11.370808601379395,
      "learning_rate": 1.7060041407867497e-05,
      "loss": 0.3378,
      "step": 7810
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.7734615206718445,
      "learning_rate": 1.7056277056277057e-05,
      "loss": 0.2407,
      "step": 7820
    },
    {
      "epoch": 1.47,
      "grad_norm": 3.8331315517425537,
      "learning_rate": 1.705251270468662e-05,
      "loss": 0.3293,
      "step": 7830
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.3909129500389099,
      "learning_rate": 1.704874835309618e-05,
      "loss": 0.2185,
      "step": 7840
    },
    {
      "epoch": 1.48,
      "grad_norm": 13.099543571472168,
      "learning_rate": 1.7044984001505743e-05,
      "loss": 0.4775,
      "step": 7850
    },
    {
      "epoch": 1.48,
      "grad_norm": 13.489631652832031,
      "learning_rate": 1.7041219649915303e-05,
      "loss": 0.2033,
      "step": 7860
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.5569513440132141,
      "learning_rate": 1.7037455298324866e-05,
      "loss": 0.3073,
      "step": 7870
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.9824039340019226,
      "learning_rate": 1.7033690946734426e-05,
      "loss": 0.2277,
      "step": 7880
    },
    {
      "epoch": 1.49,
      "grad_norm": 2.805014133453369,
      "learning_rate": 1.702992659514399e-05,
      "loss": 0.2318,
      "step": 7890
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.04034963250160217,
      "learning_rate": 1.702616224355355e-05,
      "loss": 0.2073,
      "step": 7900
    },
    {
      "epoch": 1.49,
      "grad_norm": 17.369876861572266,
      "learning_rate": 1.7022397891963112e-05,
      "loss": 0.3733,
      "step": 7910
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.6833702325820923,
      "learning_rate": 1.7018633540372672e-05,
      "loss": 0.3551,
      "step": 7920
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.037343449890613556,
      "learning_rate": 1.701486918878223e-05,
      "loss": 0.1934,
      "step": 7930
    },
    {
      "epoch": 1.49,
      "grad_norm": 1.206510305404663,
      "learning_rate": 1.7011104837191795e-05,
      "loss": 0.5499,
      "step": 7940
    },
    {
      "epoch": 1.5,
      "grad_norm": 5.40426778793335,
      "learning_rate": 1.7007340485601354e-05,
      "loss": 0.4214,
      "step": 7950
    },
    {
      "epoch": 1.5,
      "grad_norm": 15.006440162658691,
      "learning_rate": 1.7003576134010918e-05,
      "loss": 0.5344,
      "step": 7960
    },
    {
      "epoch": 1.5,
      "grad_norm": 6.656860828399658,
      "learning_rate": 1.6999811782420477e-05,
      "loss": 0.1106,
      "step": 7970
    },
    {
      "epoch": 1.5,
      "grad_norm": 16.810333251953125,
      "learning_rate": 1.699604743083004e-05,
      "loss": 0.2688,
      "step": 7980
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.1971554458141327,
      "learning_rate": 1.6992283079239604e-05,
      "loss": 0.1379,
      "step": 7990
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.4042752683162689,
      "learning_rate": 1.6988518727649163e-05,
      "loss": 0.4471,
      "step": 8000
    },
    {
      "epoch": 1.51,
      "grad_norm": 6.206781387329102,
      "learning_rate": 1.6984754376058727e-05,
      "loss": 0.277,
      "step": 8010
    },
    {
      "epoch": 1.51,
      "grad_norm": 9.190951347351074,
      "learning_rate": 1.6980990024468286e-05,
      "loss": 0.479,
      "step": 8020
    },
    {
      "epoch": 1.51,
      "grad_norm": 15.295787811279297,
      "learning_rate": 1.697722567287785e-05,
      "loss": 0.5108,
      "step": 8030
    },
    {
      "epoch": 1.51,
      "grad_norm": 34.4173583984375,
      "learning_rate": 1.697346132128741e-05,
      "loss": 0.2704,
      "step": 8040
    },
    {
      "epoch": 1.52,
      "grad_norm": 10.382560729980469,
      "learning_rate": 1.6969696969696972e-05,
      "loss": 0.3755,
      "step": 8050
    },
    {
      "epoch": 1.52,
      "grad_norm": 5.533123016357422,
      "learning_rate": 1.6965932618106532e-05,
      "loss": 0.4195,
      "step": 8060
    },
    {
      "epoch": 1.52,
      "grad_norm": 8.101094245910645,
      "learning_rate": 1.6962168266516095e-05,
      "loss": 0.3951,
      "step": 8070
    },
    {
      "epoch": 1.52,
      "grad_norm": 19.448396682739258,
      "learning_rate": 1.6958403914925655e-05,
      "loss": 0.245,
      "step": 8080
    },
    {
      "epoch": 1.52,
      "grad_norm": 14.669295310974121,
      "learning_rate": 1.6954639563335218e-05,
      "loss": 0.2057,
      "step": 8090
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.21374733746051788,
      "learning_rate": 1.6950875211744778e-05,
      "loss": 0.4004,
      "step": 8100
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.482131004333496,
      "learning_rate": 1.6947110860154338e-05,
      "loss": 0.234,
      "step": 8110
    },
    {
      "epoch": 1.53,
      "grad_norm": 5.392820358276367,
      "learning_rate": 1.69433465085639e-05,
      "loss": 0.2771,
      "step": 8120
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.7049048542976379,
      "learning_rate": 1.693958215697346e-05,
      "loss": 0.1951,
      "step": 8130
    },
    {
      "epoch": 1.53,
      "grad_norm": 11.953749656677246,
      "learning_rate": 1.6935817805383024e-05,
      "loss": 0.3908,
      "step": 8140
    },
    {
      "epoch": 1.53,
      "grad_norm": 13.924515724182129,
      "learning_rate": 1.6932053453792584e-05,
      "loss": 0.138,
      "step": 8150
    },
    {
      "epoch": 1.54,
      "grad_norm": 4.300415992736816,
      "learning_rate": 1.6928289102202147e-05,
      "loss": 0.198,
      "step": 8160
    },
    {
      "epoch": 1.54,
      "grad_norm": 37.569576263427734,
      "learning_rate": 1.692452475061171e-05,
      "loss": 0.4148,
      "step": 8170
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.4307906925678253,
      "learning_rate": 1.692076039902127e-05,
      "loss": 0.4731,
      "step": 8180
    },
    {
      "epoch": 1.54,
      "grad_norm": 2.9502201080322266,
      "learning_rate": 1.6916996047430833e-05,
      "loss": 0.4447,
      "step": 8190
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.10897088050842285,
      "learning_rate": 1.6913231695840393e-05,
      "loss": 0.1276,
      "step": 8200
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.04244183376431465,
      "learning_rate": 1.6909467344249956e-05,
      "loss": 0.3534,
      "step": 8210
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.024120863527059555,
      "learning_rate": 1.6905702992659516e-05,
      "loss": 0.2333,
      "step": 8220
    },
    {
      "epoch": 1.55,
      "grad_norm": 14.947810173034668,
      "learning_rate": 1.690193864106908e-05,
      "loss": 0.3953,
      "step": 8230
    },
    {
      "epoch": 1.55,
      "grad_norm": 9.076094627380371,
      "learning_rate": 1.689817428947864e-05,
      "loss": 0.4225,
      "step": 8240
    },
    {
      "epoch": 1.55,
      "grad_norm": 2.168973684310913,
      "learning_rate": 1.68944099378882e-05,
      "loss": 0.2134,
      "step": 8250
    },
    {
      "epoch": 1.55,
      "grad_norm": 4.159278392791748,
      "learning_rate": 1.689064558629776e-05,
      "loss": 0.2796,
      "step": 8260
    },
    {
      "epoch": 1.56,
      "grad_norm": 22.78943634033203,
      "learning_rate": 1.6886881234707325e-05,
      "loss": 0.3678,
      "step": 8270
    },
    {
      "epoch": 1.56,
      "grad_norm": 6.811593532562256,
      "learning_rate": 1.6883116883116884e-05,
      "loss": 0.2859,
      "step": 8280
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.035676684230566025,
      "learning_rate": 1.6879352531526444e-05,
      "loss": 0.3349,
      "step": 8290
    },
    {
      "epoch": 1.56,
      "grad_norm": 21.20203971862793,
      "learning_rate": 1.6875588179936007e-05,
      "loss": 0.3243,
      "step": 8300
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.8749669194221497,
      "learning_rate": 1.6871823828345567e-05,
      "loss": 0.3414,
      "step": 8310
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.20766928791999817,
      "learning_rate": 1.686805947675513e-05,
      "loss": 0.1351,
      "step": 8320
    },
    {
      "epoch": 1.57,
      "grad_norm": 9.77857494354248,
      "learning_rate": 1.686429512516469e-05,
      "loss": 0.291,
      "step": 8330
    },
    {
      "epoch": 1.57,
      "grad_norm": 24.21440887451172,
      "learning_rate": 1.6860530773574253e-05,
      "loss": 0.3718,
      "step": 8340
    },
    {
      "epoch": 1.57,
      "grad_norm": 8.406005859375,
      "learning_rate": 1.6856766421983813e-05,
      "loss": 0.3942,
      "step": 8350
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.06927170604467392,
      "learning_rate": 1.6853002070393376e-05,
      "loss": 0.3111,
      "step": 8360
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.23874908685684204,
      "learning_rate": 1.684923771880294e-05,
      "loss": 0.2727,
      "step": 8370
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.11230555921792984,
      "learning_rate": 1.68454733672125e-05,
      "loss": 0.2099,
      "step": 8380
    },
    {
      "epoch": 1.58,
      "grad_norm": 9.465620040893555,
      "learning_rate": 1.6841709015622062e-05,
      "loss": 0.2183,
      "step": 8390
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.6930161118507385,
      "learning_rate": 1.6837944664031622e-05,
      "loss": 0.2956,
      "step": 8400
    },
    {
      "epoch": 1.58,
      "grad_norm": 6.657528400421143,
      "learning_rate": 1.6834180312441185e-05,
      "loss": 0.5441,
      "step": 8410
    },
    {
      "epoch": 1.58,
      "grad_norm": 3.47774338722229,
      "learning_rate": 1.6830415960850745e-05,
      "loss": 0.2989,
      "step": 8420
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.3893883228302,
      "learning_rate": 1.6826651609260308e-05,
      "loss": 0.1871,
      "step": 8430
    },
    {
      "epoch": 1.59,
      "grad_norm": 8.039647102355957,
      "learning_rate": 1.6822887257669868e-05,
      "loss": 0.405,
      "step": 8440
    },
    {
      "epoch": 1.59,
      "grad_norm": 33.54558563232422,
      "learning_rate": 1.6819122906079427e-05,
      "loss": 0.2776,
      "step": 8450
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.0786020755767822,
      "learning_rate": 1.681535855448899e-05,
      "loss": 0.1697,
      "step": 8460
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.049975261092185974,
      "learning_rate": 1.681159420289855e-05,
      "loss": 0.4664,
      "step": 8470
    },
    {
      "epoch": 1.6,
      "grad_norm": 4.228277206420898,
      "learning_rate": 1.6807829851308114e-05,
      "loss": 0.2151,
      "step": 8480
    },
    {
      "epoch": 1.6,
      "grad_norm": 7.7242937088012695,
      "learning_rate": 1.6804065499717673e-05,
      "loss": 0.3863,
      "step": 8490
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.358811855316162,
      "learning_rate": 1.6800301148127236e-05,
      "loss": 0.1874,
      "step": 8500
    },
    {
      "epoch": 1.6,
      "grad_norm": 10.4616060256958,
      "learning_rate": 1.6796536796536796e-05,
      "loss": 0.2912,
      "step": 8510
    },
    {
      "epoch": 1.6,
      "grad_norm": 16.447561264038086,
      "learning_rate": 1.679277244494636e-05,
      "loss": 0.2512,
      "step": 8520
    },
    {
      "epoch": 1.61,
      "grad_norm": 8.851737976074219,
      "learning_rate": 1.678900809335592e-05,
      "loss": 0.3897,
      "step": 8530
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.31811287999153137,
      "learning_rate": 1.6785243741765482e-05,
      "loss": 0.2693,
      "step": 8540
    },
    {
      "epoch": 1.61,
      "grad_norm": 9.860529899597168,
      "learning_rate": 1.6781479390175045e-05,
      "loss": 0.5254,
      "step": 8550
    },
    {
      "epoch": 1.61,
      "grad_norm": 12.245877265930176,
      "learning_rate": 1.6777715038584605e-05,
      "loss": 0.3246,
      "step": 8560
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.044164713472127914,
      "learning_rate": 1.677395068699417e-05,
      "loss": 0.222,
      "step": 8570
    },
    {
      "epoch": 1.61,
      "grad_norm": 13.014546394348145,
      "learning_rate": 1.6770186335403728e-05,
      "loss": 0.2424,
      "step": 8580
    },
    {
      "epoch": 1.62,
      "grad_norm": 1.0840725898742676,
      "learning_rate": 1.676642198381329e-05,
      "loss": 0.4446,
      "step": 8590
    },
    {
      "epoch": 1.62,
      "grad_norm": 7.20802116394043,
      "learning_rate": 1.676265763222285e-05,
      "loss": 0.2118,
      "step": 8600
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.39927688241004944,
      "learning_rate": 1.6758893280632414e-05,
      "loss": 0.1866,
      "step": 8610
    },
    {
      "epoch": 1.62,
      "grad_norm": 7.404331207275391,
      "learning_rate": 1.6755128929041974e-05,
      "loss": 0.3686,
      "step": 8620
    },
    {
      "epoch": 1.62,
      "grad_norm": 4.517202377319336,
      "learning_rate": 1.6751364577451534e-05,
      "loss": 0.2794,
      "step": 8630
    },
    {
      "epoch": 1.63,
      "grad_norm": 43.021644592285156,
      "learning_rate": 1.6747600225861097e-05,
      "loss": 0.1322,
      "step": 8640
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.12860260903835297,
      "learning_rate": 1.6743835874270657e-05,
      "loss": 0.2104,
      "step": 8650
    },
    {
      "epoch": 1.63,
      "grad_norm": 24.812345504760742,
      "learning_rate": 1.674007152268022e-05,
      "loss": 0.1685,
      "step": 8660
    },
    {
      "epoch": 1.63,
      "grad_norm": 12.117597579956055,
      "learning_rate": 1.673630717108978e-05,
      "loss": 0.2313,
      "step": 8670
    },
    {
      "epoch": 1.63,
      "grad_norm": 5.608607769012451,
      "learning_rate": 1.6732542819499343e-05,
      "loss": 0.2683,
      "step": 8680
    },
    {
      "epoch": 1.64,
      "grad_norm": 3.907818555831909,
      "learning_rate": 1.6728778467908903e-05,
      "loss": 0.1705,
      "step": 8690
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.08978606760501862,
      "learning_rate": 1.6725014116318466e-05,
      "loss": 0.1397,
      "step": 8700
    },
    {
      "epoch": 1.64,
      "grad_norm": 19.57398223876953,
      "learning_rate": 1.6721249764728025e-05,
      "loss": 0.3013,
      "step": 8710
    },
    {
      "epoch": 1.64,
      "grad_norm": 29.60633087158203,
      "learning_rate": 1.671748541313759e-05,
      "loss": 0.2959,
      "step": 8720
    },
    {
      "epoch": 1.64,
      "grad_norm": 3.5758614540100098,
      "learning_rate": 1.6713721061547152e-05,
      "loss": 0.2811,
      "step": 8730
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.04845660179853439,
      "learning_rate": 1.670995670995671e-05,
      "loss": 0.2088,
      "step": 8740
    },
    {
      "epoch": 1.65,
      "grad_norm": 20.819917678833008,
      "learning_rate": 1.6706192358366275e-05,
      "loss": 0.3688,
      "step": 8750
    },
    {
      "epoch": 1.65,
      "grad_norm": 11.952372550964355,
      "learning_rate": 1.6702428006775834e-05,
      "loss": 0.3736,
      "step": 8760
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.9629964828491211,
      "learning_rate": 1.6698663655185398e-05,
      "loss": 0.4909,
      "step": 8770
    },
    {
      "epoch": 1.65,
      "grad_norm": 18.00396156311035,
      "learning_rate": 1.6694899303594957e-05,
      "loss": 0.3258,
      "step": 8780
    },
    {
      "epoch": 1.65,
      "grad_norm": 3.549755096435547,
      "learning_rate": 1.669113495200452e-05,
      "loss": 0.3156,
      "step": 8790
    },
    {
      "epoch": 1.66,
      "grad_norm": 3.037888288497925,
      "learning_rate": 1.668737060041408e-05,
      "loss": 0.3395,
      "step": 8800
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.14833039045333862,
      "learning_rate": 1.668360624882364e-05,
      "loss": 0.2187,
      "step": 8810
    },
    {
      "epoch": 1.66,
      "grad_norm": 4.986731052398682,
      "learning_rate": 1.6679841897233203e-05,
      "loss": 0.2023,
      "step": 8820
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.32868432998657227,
      "learning_rate": 1.6676077545642763e-05,
      "loss": 0.1572,
      "step": 8830
    },
    {
      "epoch": 1.66,
      "grad_norm": 6.6207170486450195,
      "learning_rate": 1.6672313194052326e-05,
      "loss": 0.0347,
      "step": 8840
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.021260514855384827,
      "learning_rate": 1.6668548842461886e-05,
      "loss": 0.1426,
      "step": 8850
    },
    {
      "epoch": 1.67,
      "grad_norm": 30.467321395874023,
      "learning_rate": 1.666478449087145e-05,
      "loss": 0.3403,
      "step": 8860
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.06284291297197342,
      "learning_rate": 1.666102013928101e-05,
      "loss": 0.2808,
      "step": 8870
    },
    {
      "epoch": 1.67,
      "grad_norm": 7.406793594360352,
      "learning_rate": 1.6657255787690572e-05,
      "loss": 0.3507,
      "step": 8880
    },
    {
      "epoch": 1.67,
      "grad_norm": 8.369754791259766,
      "learning_rate": 1.6653491436100132e-05,
      "loss": 0.3574,
      "step": 8890
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.07993824034929276,
      "learning_rate": 1.6649727084509695e-05,
      "loss": 0.2279,
      "step": 8900
    },
    {
      "epoch": 1.68,
      "grad_norm": 10.284481048583984,
      "learning_rate": 1.6645962732919258e-05,
      "loss": 0.3397,
      "step": 8910
    },
    {
      "epoch": 1.68,
      "grad_norm": 9.481452941894531,
      "learning_rate": 1.6642198381328818e-05,
      "loss": 0.3879,
      "step": 8920
    },
    {
      "epoch": 1.68,
      "grad_norm": 13.257113456726074,
      "learning_rate": 1.663843402973838e-05,
      "loss": 0.4326,
      "step": 8930
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.95779800415039,
      "learning_rate": 1.663466967814794e-05,
      "loss": 0.2833,
      "step": 8940
    },
    {
      "epoch": 1.68,
      "grad_norm": 11.471521377563477,
      "learning_rate": 1.6630905326557504e-05,
      "loss": 0.1903,
      "step": 8950
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.06078150123357773,
      "learning_rate": 1.6627140974967064e-05,
      "loss": 0.3259,
      "step": 8960
    },
    {
      "epoch": 1.69,
      "grad_norm": 24.76637840270996,
      "learning_rate": 1.6623376623376627e-05,
      "loss": 0.326,
      "step": 8970
    },
    {
      "epoch": 1.69,
      "grad_norm": 2.2596938610076904,
      "learning_rate": 1.6619612271786187e-05,
      "loss": 0.243,
      "step": 8980
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.09513311833143234,
      "learning_rate": 1.6615847920195746e-05,
      "loss": 0.062,
      "step": 8990
    },
    {
      "epoch": 1.69,
      "grad_norm": 23.184696197509766,
      "learning_rate": 1.661208356860531e-05,
      "loss": 0.3259,
      "step": 9000
    },
    {
      "epoch": 1.7,
      "grad_norm": 8.123701095581055,
      "learning_rate": 1.660831921701487e-05,
      "loss": 0.4647,
      "step": 9010
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.2529292106628418,
      "learning_rate": 1.6604554865424432e-05,
      "loss": 0.2496,
      "step": 9020
    },
    {
      "epoch": 1.7,
      "grad_norm": 11.049469947814941,
      "learning_rate": 1.6600790513833992e-05,
      "loss": 0.22,
      "step": 9030
    },
    {
      "epoch": 1.7,
      "grad_norm": 29.820812225341797,
      "learning_rate": 1.6597026162243555e-05,
      "loss": 0.1622,
      "step": 9040
    },
    {
      "epoch": 1.7,
      "grad_norm": 4.785089015960693,
      "learning_rate": 1.6593261810653115e-05,
      "loss": 0.212,
      "step": 9050
    },
    {
      "epoch": 1.71,
      "grad_norm": 12.907205581665039,
      "learning_rate": 1.6589497459062678e-05,
      "loss": 0.3182,
      "step": 9060
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.026758868247270584,
      "learning_rate": 1.6585733107472238e-05,
      "loss": 0.2642,
      "step": 9070
    },
    {
      "epoch": 1.71,
      "grad_norm": 5.124535083770752,
      "learning_rate": 1.65819687558818e-05,
      "loss": 0.4348,
      "step": 9080
    },
    {
      "epoch": 1.71,
      "grad_norm": 9.858972549438477,
      "learning_rate": 1.657820440429136e-05,
      "loss": 0.35,
      "step": 9090
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.05167049169540405,
      "learning_rate": 1.6574440052700924e-05,
      "loss": 0.2781,
      "step": 9100
    },
    {
      "epoch": 1.71,
      "grad_norm": 7.978021621704102,
      "learning_rate": 1.6570675701110487e-05,
      "loss": 0.324,
      "step": 9110
    },
    {
      "epoch": 1.72,
      "grad_norm": 7.549741268157959,
      "learning_rate": 1.6566911349520047e-05,
      "loss": 0.2086,
      "step": 9120
    },
    {
      "epoch": 1.72,
      "grad_norm": 12.31412124633789,
      "learning_rate": 1.656314699792961e-05,
      "loss": 0.1692,
      "step": 9130
    },
    {
      "epoch": 1.72,
      "grad_norm": 28.019611358642578,
      "learning_rate": 1.655938264633917e-05,
      "loss": 0.2613,
      "step": 9140
    },
    {
      "epoch": 1.72,
      "grad_norm": 9.009861946105957,
      "learning_rate": 1.655561829474873e-05,
      "loss": 0.2828,
      "step": 9150
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.6212007999420166,
      "learning_rate": 1.6551853943158293e-05,
      "loss": 0.2516,
      "step": 9160
    },
    {
      "epoch": 1.73,
      "grad_norm": 18.18745994567871,
      "learning_rate": 1.6548089591567853e-05,
      "loss": 0.4031,
      "step": 9170
    },
    {
      "epoch": 1.73,
      "grad_norm": 5.736402988433838,
      "learning_rate": 1.6544325239977416e-05,
      "loss": 0.1736,
      "step": 9180
    },
    {
      "epoch": 1.73,
      "grad_norm": 16.646106719970703,
      "learning_rate": 1.6540560888386976e-05,
      "loss": 0.346,
      "step": 9190
    },
    {
      "epoch": 1.73,
      "grad_norm": 4.171016693115234,
      "learning_rate": 1.653679653679654e-05,
      "loss": 0.5091,
      "step": 9200
    },
    {
      "epoch": 1.73,
      "grad_norm": 2.88995361328125,
      "learning_rate": 1.65330321852061e-05,
      "loss": 0.3099,
      "step": 9210
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.406023979187012,
      "learning_rate": 1.652926783361566e-05,
      "loss": 0.1212,
      "step": 9220
    },
    {
      "epoch": 1.74,
      "grad_norm": 5.908947944641113,
      "learning_rate": 1.652550348202522e-05,
      "loss": 0.1641,
      "step": 9230
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.02778688445687294,
      "learning_rate": 1.6521739130434785e-05,
      "loss": 0.3078,
      "step": 9240
    },
    {
      "epoch": 1.74,
      "grad_norm": 12.006573677062988,
      "learning_rate": 1.6517974778844344e-05,
      "loss": 0.4784,
      "step": 9250
    },
    {
      "epoch": 1.74,
      "grad_norm": 8.120018005371094,
      "learning_rate": 1.6514210427253907e-05,
      "loss": 0.3427,
      "step": 9260
    },
    {
      "epoch": 1.74,
      "grad_norm": 11.392836570739746,
      "learning_rate": 1.6510446075663467e-05,
      "loss": 0.3247,
      "step": 9270
    },
    {
      "epoch": 1.75,
      "grad_norm": 3.501584529876709,
      "learning_rate": 1.650668172407303e-05,
      "loss": 0.1097,
      "step": 9280
    },
    {
      "epoch": 1.75,
      "grad_norm": 18.299657821655273,
      "learning_rate": 1.6502917372482594e-05,
      "loss": 0.407,
      "step": 9290
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.7153254151344299,
      "learning_rate": 1.6499153020892153e-05,
      "loss": 0.319,
      "step": 9300
    },
    {
      "epoch": 1.75,
      "grad_norm": 6.911435604095459,
      "learning_rate": 1.6495388669301716e-05,
      "loss": 0.2321,
      "step": 9310
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.033025797456502914,
      "learning_rate": 1.6491624317711273e-05,
      "loss": 0.2356,
      "step": 9320
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.16273614764213562,
      "learning_rate": 1.6487859966120836e-05,
      "loss": 0.5412,
      "step": 9330
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.027941910549998283,
      "learning_rate": 1.64840956145304e-05,
      "loss": 0.2849,
      "step": 9340
    },
    {
      "epoch": 1.76,
      "grad_norm": 18.502824783325195,
      "learning_rate": 1.648033126293996e-05,
      "loss": 0.2393,
      "step": 9350
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.194416046142578,
      "learning_rate": 1.6476566911349522e-05,
      "loss": 0.2999,
      "step": 9360
    },
    {
      "epoch": 1.76,
      "grad_norm": 9.125223159790039,
      "learning_rate": 1.6472802559759082e-05,
      "loss": 0.2455,
      "step": 9370
    },
    {
      "epoch": 1.77,
      "grad_norm": 10.938528060913086,
      "learning_rate": 1.6469038208168645e-05,
      "loss": 0.3403,
      "step": 9380
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.4583791494369507,
      "learning_rate": 1.6465273856578205e-05,
      "loss": 0.1412,
      "step": 9390
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.15345677733421326,
      "learning_rate": 1.6461509504987768e-05,
      "loss": 0.3129,
      "step": 9400
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.05747629329562187,
      "learning_rate": 1.6457745153397328e-05,
      "loss": 0.298,
      "step": 9410
    },
    {
      "epoch": 1.77,
      "grad_norm": 4.696110725402832,
      "learning_rate": 1.645398080180689e-05,
      "loss": 0.4108,
      "step": 9420
    },
    {
      "epoch": 1.77,
      "grad_norm": 2.812985420227051,
      "learning_rate": 1.645021645021645e-05,
      "loss": 0.2567,
      "step": 9430
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.03002488799393177,
      "learning_rate": 1.6446452098626014e-05,
      "loss": 0.2033,
      "step": 9440
    },
    {
      "epoch": 1.78,
      "grad_norm": 15.042752265930176,
      "learning_rate": 1.6442687747035574e-05,
      "loss": 0.2748,
      "step": 9450
    },
    {
      "epoch": 1.78,
      "grad_norm": 14.62966251373291,
      "learning_rate": 1.6438923395445137e-05,
      "loss": 0.3251,
      "step": 9460
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.028030524030327797,
      "learning_rate": 1.64351590438547e-05,
      "loss": 0.0999,
      "step": 9470
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.4641982316970825,
      "learning_rate": 1.643139469226426e-05,
      "loss": 0.3124,
      "step": 9480
    },
    {
      "epoch": 1.79,
      "grad_norm": 21.00873565673828,
      "learning_rate": 1.6427630340673823e-05,
      "loss": 0.3322,
      "step": 9490
    },
    {
      "epoch": 1.79,
      "grad_norm": 6.387186050415039,
      "learning_rate": 1.642386598908338e-05,
      "loss": 0.3565,
      "step": 9500
    },
    {
      "epoch": 1.79,
      "grad_norm": 24.23895835876465,
      "learning_rate": 1.6420101637492942e-05,
      "loss": 0.5073,
      "step": 9510
    },
    {
      "epoch": 1.79,
      "grad_norm": 24.894182205200195,
      "learning_rate": 1.6416337285902505e-05,
      "loss": 0.4015,
      "step": 9520
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.05306817218661308,
      "learning_rate": 1.6412572934312065e-05,
      "loss": 0.4002,
      "step": 9530
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.268216609954834,
      "learning_rate": 1.640880858272163e-05,
      "loss": 0.3182,
      "step": 9540
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5311167240142822,
      "learning_rate": 1.6405044231131188e-05,
      "loss": 0.1576,
      "step": 9550
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5603498816490173,
      "learning_rate": 1.640127987954075e-05,
      "loss": 0.2551,
      "step": 9560
    },
    {
      "epoch": 1.8,
      "grad_norm": 12.341197967529297,
      "learning_rate": 1.639751552795031e-05,
      "loss": 0.2108,
      "step": 9570
    },
    {
      "epoch": 1.8,
      "grad_norm": 18.991601943969727,
      "learning_rate": 1.6393751176359874e-05,
      "loss": 0.8104,
      "step": 9580
    },
    {
      "epoch": 1.81,
      "grad_norm": 4.916114330291748,
      "learning_rate": 1.6389986824769434e-05,
      "loss": 0.3694,
      "step": 9590
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.8138089179992676,
      "learning_rate": 1.6386222473178997e-05,
      "loss": 0.184,
      "step": 9600
    },
    {
      "epoch": 1.81,
      "grad_norm": 6.93763542175293,
      "learning_rate": 1.6382458121588557e-05,
      "loss": 0.207,
      "step": 9610
    },
    {
      "epoch": 1.81,
      "grad_norm": 3.088991403579712,
      "learning_rate": 1.637869376999812e-05,
      "loss": 0.1603,
      "step": 9620
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.494444876909256,
      "learning_rate": 1.637492941840768e-05,
      "loss": 0.3963,
      "step": 9630
    },
    {
      "epoch": 1.81,
      "grad_norm": 1.281357765197754,
      "learning_rate": 1.6371165066817243e-05,
      "loss": 0.3629,
      "step": 9640
    },
    {
      "epoch": 1.82,
      "grad_norm": 19.56279754638672,
      "learning_rate": 1.6367400715226803e-05,
      "loss": 0.2883,
      "step": 9650
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.1886323094367981,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 0.235,
      "step": 9660
    },
    {
      "epoch": 1.82,
      "grad_norm": 9.69140338897705,
      "learning_rate": 1.6359872012045926e-05,
      "loss": 0.3343,
      "step": 9670
    },
    {
      "epoch": 1.82,
      "grad_norm": 12.699943542480469,
      "learning_rate": 1.6356107660455485e-05,
      "loss": 0.2628,
      "step": 9680
    },
    {
      "epoch": 1.82,
      "grad_norm": 13.287461280822754,
      "learning_rate": 1.635234330886505e-05,
      "loss": 0.3181,
      "step": 9690
    },
    {
      "epoch": 1.83,
      "grad_norm": 14.643268585205078,
      "learning_rate": 1.6348578957274612e-05,
      "loss": 0.4688,
      "step": 9700
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.4771287739276886,
      "learning_rate": 1.634481460568417e-05,
      "loss": 0.1302,
      "step": 9710
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.6301877498626709,
      "learning_rate": 1.6341050254093735e-05,
      "loss": 0.2187,
      "step": 9720
    },
    {
      "epoch": 1.83,
      "grad_norm": 17.00600242614746,
      "learning_rate": 1.6337285902503294e-05,
      "loss": 0.19,
      "step": 9730
    },
    {
      "epoch": 1.83,
      "grad_norm": 10.562419891357422,
      "learning_rate": 1.6333521550912858e-05,
      "loss": 0.4001,
      "step": 9740
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.1650846153497696,
      "learning_rate": 1.6329757199322417e-05,
      "loss": 0.3258,
      "step": 9750
    },
    {
      "epoch": 1.84,
      "grad_norm": 16.823192596435547,
      "learning_rate": 1.632599284773198e-05,
      "loss": 0.2407,
      "step": 9760
    },
    {
      "epoch": 1.84,
      "grad_norm": 1.8425030708312988,
      "learning_rate": 1.632222849614154e-05,
      "loss": 0.2382,
      "step": 9770
    },
    {
      "epoch": 1.84,
      "grad_norm": 45.72221755981445,
      "learning_rate": 1.6318464144551103e-05,
      "loss": 0.3792,
      "step": 9780
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.10260268300771713,
      "learning_rate": 1.6314699792960663e-05,
      "loss": 0.3058,
      "step": 9790
    },
    {
      "epoch": 1.84,
      "grad_norm": 7.359597206115723,
      "learning_rate": 1.6310935441370226e-05,
      "loss": 0.3137,
      "step": 9800
    },
    {
      "epoch": 1.85,
      "grad_norm": 2.1723551750183105,
      "learning_rate": 1.6307171089779786e-05,
      "loss": 0.3994,
      "step": 9810
    },
    {
      "epoch": 1.85,
      "grad_norm": 1.1915785074234009,
      "learning_rate": 1.630340673818935e-05,
      "loss": 0.2574,
      "step": 9820
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.5839290618896484,
      "learning_rate": 1.629964238659891e-05,
      "loss": 0.2985,
      "step": 9830
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.7010390758514404,
      "learning_rate": 1.6295878035008472e-05,
      "loss": 0.3062,
      "step": 9840
    },
    {
      "epoch": 1.85,
      "grad_norm": 5.798739433288574,
      "learning_rate": 1.6292113683418032e-05,
      "loss": 0.2684,
      "step": 9850
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.149734228849411,
      "learning_rate": 1.6288349331827592e-05,
      "loss": 0.2038,
      "step": 9860
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.329597234725952,
      "learning_rate": 1.6284584980237155e-05,
      "loss": 0.1834,
      "step": 9870
    },
    {
      "epoch": 1.86,
      "grad_norm": 7.2279462814331055,
      "learning_rate": 1.6280820628646715e-05,
      "loss": 0.2785,
      "step": 9880
    },
    {
      "epoch": 1.86,
      "grad_norm": 2.4880359172821045,
      "learning_rate": 1.6277056277056278e-05,
      "loss": 0.3801,
      "step": 9890
    },
    {
      "epoch": 1.86,
      "grad_norm": 11.710917472839355,
      "learning_rate": 1.627329192546584e-05,
      "loss": 0.4175,
      "step": 9900
    },
    {
      "epoch": 1.87,
      "grad_norm": 8.251520156860352,
      "learning_rate": 1.62695275738754e-05,
      "loss": 0.2468,
      "step": 9910
    },
    {
      "epoch": 1.87,
      "grad_norm": 1.6254079341888428,
      "learning_rate": 1.6265763222284964e-05,
      "loss": 0.2869,
      "step": 9920
    },
    {
      "epoch": 1.87,
      "grad_norm": 2.328249216079712,
      "learning_rate": 1.6261998870694524e-05,
      "loss": 0.26,
      "step": 9930
    },
    {
      "epoch": 1.87,
      "grad_norm": 19.06911849975586,
      "learning_rate": 1.6258234519104087e-05,
      "loss": 0.1198,
      "step": 9940
    },
    {
      "epoch": 1.87,
      "grad_norm": 4.8381733894348145,
      "learning_rate": 1.6254470167513647e-05,
      "loss": 0.2999,
      "step": 9950
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.20530369877815247,
      "learning_rate": 1.625070581592321e-05,
      "loss": 0.325,
      "step": 9960
    },
    {
      "epoch": 1.88,
      "grad_norm": 7.393706321716309,
      "learning_rate": 1.624694146433277e-05,
      "loss": 0.2624,
      "step": 9970
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.08343791961669922,
      "learning_rate": 1.6243177112742333e-05,
      "loss": 0.3246,
      "step": 9980
    },
    {
      "epoch": 1.88,
      "grad_norm": 18.146913528442383,
      "learning_rate": 1.6239412761151892e-05,
      "loss": 0.1879,
      "step": 9990
    },
    {
      "epoch": 1.88,
      "grad_norm": 8.685861587524414,
      "learning_rate": 1.6235648409561456e-05,
      "loss": 0.2632,
      "step": 10000
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.5423463582992554,
      "learning_rate": 1.6231884057971015e-05,
      "loss": 0.1752,
      "step": 10010
    },
    {
      "epoch": 1.89,
      "grad_norm": 5.38092565536499,
      "learning_rate": 1.6228119706380575e-05,
      "loss": 0.1842,
      "step": 10020
    },
    {
      "epoch": 1.89,
      "grad_norm": 21.352991104125977,
      "learning_rate": 1.6224355354790138e-05,
      "loss": 0.5758,
      "step": 10030
    },
    {
      "epoch": 1.89,
      "grad_norm": 10.81428337097168,
      "learning_rate": 1.6220591003199698e-05,
      "loss": 0.2849,
      "step": 10040
    },
    {
      "epoch": 1.89,
      "grad_norm": 6.751648902893066,
      "learning_rate": 1.621682665160926e-05,
      "loss": 0.3465,
      "step": 10050
    },
    {
      "epoch": 1.89,
      "grad_norm": 11.685079574584961,
      "learning_rate": 1.621306230001882e-05,
      "loss": 0.411,
      "step": 10060
    },
    {
      "epoch": 1.9,
      "grad_norm": 16.50705909729004,
      "learning_rate": 1.6209297948428384e-05,
      "loss": 0.2783,
      "step": 10070
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.28239303827285767,
      "learning_rate": 1.6205533596837947e-05,
      "loss": 0.3107,
      "step": 10080
    },
    {
      "epoch": 1.9,
      "grad_norm": 9.995141983032227,
      "learning_rate": 1.6201769245247507e-05,
      "loss": 0.1654,
      "step": 10090
    },
    {
      "epoch": 1.9,
      "grad_norm": 6.747482776641846,
      "learning_rate": 1.619800489365707e-05,
      "loss": 0.3331,
      "step": 10100
    },
    {
      "epoch": 1.9,
      "grad_norm": 13.731993675231934,
      "learning_rate": 1.619424054206663e-05,
      "loss": 0.1953,
      "step": 10110
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.33005252480506897,
      "learning_rate": 1.6190476190476193e-05,
      "loss": 0.138,
      "step": 10120
    },
    {
      "epoch": 1.91,
      "grad_norm": 33.661216735839844,
      "learning_rate": 1.6186711838885753e-05,
      "loss": 0.2888,
      "step": 10130
    },
    {
      "epoch": 1.91,
      "grad_norm": 4.381628513336182,
      "learning_rate": 1.6182947487295316e-05,
      "loss": 0.3398,
      "step": 10140
    },
    {
      "epoch": 1.91,
      "grad_norm": 20.355709075927734,
      "learning_rate": 1.6179183135704876e-05,
      "loss": 0.2301,
      "step": 10150
    },
    {
      "epoch": 1.91,
      "grad_norm": 11.423288345336914,
      "learning_rate": 1.617541878411444e-05,
      "loss": 0.3247,
      "step": 10160
    },
    {
      "epoch": 1.91,
      "grad_norm": 2.360579490661621,
      "learning_rate": 1.6171654432524e-05,
      "loss": 0.4906,
      "step": 10170
    },
    {
      "epoch": 1.92,
      "grad_norm": 15.754151344299316,
      "learning_rate": 1.6167890080933562e-05,
      "loss": 0.2982,
      "step": 10180
    },
    {
      "epoch": 1.92,
      "grad_norm": 13.262439727783203,
      "learning_rate": 1.616412572934312e-05,
      "loss": 0.3827,
      "step": 10190
    },
    {
      "epoch": 1.92,
      "grad_norm": 4.707301616668701,
      "learning_rate": 1.616036137775268e-05,
      "loss": 0.3402,
      "step": 10200
    },
    {
      "epoch": 1.92,
      "grad_norm": 3.142564535140991,
      "learning_rate": 1.6156597026162245e-05,
      "loss": 0.2613,
      "step": 10210
    },
    {
      "epoch": 1.92,
      "grad_norm": 5.9899702072143555,
      "learning_rate": 1.6152832674571804e-05,
      "loss": 0.4478,
      "step": 10220
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.5889505743980408,
      "learning_rate": 1.6149068322981367e-05,
      "loss": 0.1249,
      "step": 10230
    },
    {
      "epoch": 1.93,
      "grad_norm": 10.380497932434082,
      "learning_rate": 1.6145303971390927e-05,
      "loss": 0.399,
      "step": 10240
    },
    {
      "epoch": 1.93,
      "grad_norm": 17.387714385986328,
      "learning_rate": 1.614153961980049e-05,
      "loss": 0.4684,
      "step": 10250
    },
    {
      "epoch": 1.93,
      "grad_norm": 3.8738033771514893,
      "learning_rate": 1.6137775268210054e-05,
      "loss": 0.3211,
      "step": 10260
    },
    {
      "epoch": 1.93,
      "grad_norm": 12.881617546081543,
      "learning_rate": 1.6134010916619613e-05,
      "loss": 0.2058,
      "step": 10270
    },
    {
      "epoch": 1.93,
      "grad_norm": 6.7992939949035645,
      "learning_rate": 1.6130246565029176e-05,
      "loss": 0.1671,
      "step": 10280
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.977795362472534,
      "learning_rate": 1.6126482213438736e-05,
      "loss": 0.236,
      "step": 10290
    },
    {
      "epoch": 1.94,
      "grad_norm": 4.797702312469482,
      "learning_rate": 1.61227178618483e-05,
      "loss": 0.4174,
      "step": 10300
    },
    {
      "epoch": 1.94,
      "grad_norm": 23.082847595214844,
      "learning_rate": 1.611895351025786e-05,
      "loss": 0.3731,
      "step": 10310
    },
    {
      "epoch": 1.94,
      "grad_norm": 2.9605114459991455,
      "learning_rate": 1.6115189158667422e-05,
      "loss": 0.2312,
      "step": 10320
    },
    {
      "epoch": 1.94,
      "grad_norm": 18.443361282348633,
      "learning_rate": 1.6111424807076982e-05,
      "loss": 0.1472,
      "step": 10330
    },
    {
      "epoch": 1.95,
      "grad_norm": 12.151503562927246,
      "learning_rate": 1.6107660455486545e-05,
      "loss": 0.3523,
      "step": 10340
    },
    {
      "epoch": 1.95,
      "grad_norm": 6.169518947601318,
      "learning_rate": 1.6103896103896105e-05,
      "loss": 0.2636,
      "step": 10350
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.883929252624512,
      "learning_rate": 1.6100131752305668e-05,
      "loss": 0.2747,
      "step": 10360
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.261861801147461,
      "learning_rate": 1.6096367400715228e-05,
      "loss": 0.3482,
      "step": 10370
    },
    {
      "epoch": 1.95,
      "grad_norm": 11.442970275878906,
      "learning_rate": 1.6092603049124788e-05,
      "loss": 0.3089,
      "step": 10380
    },
    {
      "epoch": 1.96,
      "grad_norm": 5.298076152801514,
      "learning_rate": 1.608883869753435e-05,
      "loss": 0.265,
      "step": 10390
    },
    {
      "epoch": 1.96,
      "grad_norm": 19.199138641357422,
      "learning_rate": 1.608507434594391e-05,
      "loss": 0.3449,
      "step": 10400
    },
    {
      "epoch": 1.96,
      "grad_norm": 16.81801986694336,
      "learning_rate": 1.6081309994353474e-05,
      "loss": 0.369,
      "step": 10410
    },
    {
      "epoch": 1.96,
      "grad_norm": 5.926851272583008,
      "learning_rate": 1.6077545642763034e-05,
      "loss": 0.158,
      "step": 10420
    },
    {
      "epoch": 1.96,
      "grad_norm": 12.396550178527832,
      "learning_rate": 1.6073781291172597e-05,
      "loss": 0.3471,
      "step": 10430
    },
    {
      "epoch": 1.96,
      "grad_norm": 15.123412132263184,
      "learning_rate": 1.6070016939582156e-05,
      "loss": 0.2831,
      "step": 10440
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.019230080768465996,
      "learning_rate": 1.606625258799172e-05,
      "loss": 0.4338,
      "step": 10450
    },
    {
      "epoch": 1.97,
      "grad_norm": 1.1439626216888428,
      "learning_rate": 1.6062488236401283e-05,
      "loss": 0.2244,
      "step": 10460
    },
    {
      "epoch": 1.97,
      "grad_norm": 8.582100868225098,
      "learning_rate": 1.6058723884810843e-05,
      "loss": 0.4041,
      "step": 10470
    },
    {
      "epoch": 1.97,
      "grad_norm": 2.7329792976379395,
      "learning_rate": 1.6054959533220406e-05,
      "loss": 0.3182,
      "step": 10480
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.6955661177635193,
      "learning_rate": 1.6051195181629965e-05,
      "loss": 0.1814,
      "step": 10490
    },
    {
      "epoch": 1.98,
      "grad_norm": 6.490228652954102,
      "learning_rate": 1.604743083003953e-05,
      "loss": 0.24,
      "step": 10500
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.3193872570991516,
      "learning_rate": 1.604366647844909e-05,
      "loss": 0.127,
      "step": 10510
    },
    {
      "epoch": 1.98,
      "grad_norm": 11.384047508239746,
      "learning_rate": 1.603990212685865e-05,
      "loss": 0.287,
      "step": 10520
    },
    {
      "epoch": 1.98,
      "grad_norm": 18.699859619140625,
      "learning_rate": 1.603613777526821e-05,
      "loss": 0.1549,
      "step": 10530
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.025467922911047935,
      "learning_rate": 1.603237342367777e-05,
      "loss": 0.3966,
      "step": 10540
    },
    {
      "epoch": 1.99,
      "grad_norm": 12.881343841552734,
      "learning_rate": 1.6028609072087334e-05,
      "loss": 0.1507,
      "step": 10550
    },
    {
      "epoch": 1.99,
      "grad_norm": 5.623108386993408,
      "learning_rate": 1.6024844720496894e-05,
      "loss": 0.4098,
      "step": 10560
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.02270541898906231,
      "learning_rate": 1.6021080368906457e-05,
      "loss": 0.3949,
      "step": 10570
    },
    {
      "epoch": 1.99,
      "grad_norm": 4.504025936126709,
      "learning_rate": 1.6017316017316017e-05,
      "loss": 0.3122,
      "step": 10580
    },
    {
      "epoch": 1.99,
      "grad_norm": 6.443422317504883,
      "learning_rate": 1.601355166572558e-05,
      "loss": 0.2997,
      "step": 10590
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.685037612915039,
      "learning_rate": 1.600978731413514e-05,
      "loss": 0.1032,
      "step": 10600
    },
    {
      "epoch": 2.0,
      "grad_norm": 14.032691955566406,
      "learning_rate": 1.6006022962544703e-05,
      "loss": 0.4188,
      "step": 10610
    },
    {
      "epoch": 2.0,
      "grad_norm": 39.16253662109375,
      "learning_rate": 1.6002258610954263e-05,
      "loss": 0.3249,
      "step": 10620
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9836,
      "eval_loss": 0.05493224412202835,
      "eval_runtime": 850.468,
      "eval_samples_per_second": 8.819,
      "eval_steps_per_second": 1.103,
      "step": 10626
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.720757484436035,
      "learning_rate": 1.5998494259363826e-05,
      "loss": 0.3579,
      "step": 10630
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.04844209924340248,
      "learning_rate": 1.599472990777339e-05,
      "loss": 0.1262,
      "step": 10640
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1008999347686768,
      "learning_rate": 1.599096555618295e-05,
      "loss": 0.1838,
      "step": 10650
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.19694870710372925,
      "learning_rate": 1.5987201204592512e-05,
      "loss": 0.1221,
      "step": 10660
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.7260994911193848,
      "learning_rate": 1.5983436853002072e-05,
      "loss": 0.2181,
      "step": 10670
    },
    {
      "epoch": 2.01,
      "grad_norm": 7.747636795043945,
      "learning_rate": 1.5979672501411635e-05,
      "loss": 0.3375,
      "step": 10680
    },
    {
      "epoch": 2.01,
      "grad_norm": 32.49991989135742,
      "learning_rate": 1.5975908149821195e-05,
      "loss": 0.4678,
      "step": 10690
    },
    {
      "epoch": 2.01,
      "grad_norm": 34.39246368408203,
      "learning_rate": 1.5972143798230758e-05,
      "loss": 0.4381,
      "step": 10700
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.0823490619659424,
      "learning_rate": 1.5968379446640318e-05,
      "loss": 0.1856,
      "step": 10710
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.1460677832365036,
      "learning_rate": 1.5964615095049877e-05,
      "loss": 0.1955,
      "step": 10720
    },
    {
      "epoch": 2.02,
      "grad_norm": 2.439439296722412,
      "learning_rate": 1.596085074345944e-05,
      "loss": 0.091,
      "step": 10730
    },
    {
      "epoch": 2.02,
      "grad_norm": 13.830291748046875,
      "learning_rate": 1.5957086391869e-05,
      "loss": 0.253,
      "step": 10740
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.7650459408760071,
      "learning_rate": 1.5953322040278563e-05,
      "loss": 0.2654,
      "step": 10750
    },
    {
      "epoch": 2.03,
      "grad_norm": 14.792553901672363,
      "learning_rate": 1.5949557688688123e-05,
      "loss": 0.2915,
      "step": 10760
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.025888020172715187,
      "learning_rate": 1.5945793337097686e-05,
      "loss": 0.2562,
      "step": 10770
    },
    {
      "epoch": 2.03,
      "grad_norm": 10.788376808166504,
      "learning_rate": 1.5942028985507246e-05,
      "loss": 0.3648,
      "step": 10780
    },
    {
      "epoch": 2.03,
      "grad_norm": 1.012776255607605,
      "learning_rate": 1.593826463391681e-05,
      "loss": 0.3397,
      "step": 10790
    },
    {
      "epoch": 2.03,
      "grad_norm": 27.082355499267578,
      "learning_rate": 1.593450028232637e-05,
      "loss": 0.4166,
      "step": 10800
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.1255490928888321,
      "learning_rate": 1.5930735930735932e-05,
      "loss": 0.2413,
      "step": 10810
    },
    {
      "epoch": 2.04,
      "grad_norm": 9.698269844055176,
      "learning_rate": 1.5926971579145495e-05,
      "loss": 0.5846,
      "step": 10820
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.056032951921224594,
      "learning_rate": 1.5923207227555055e-05,
      "loss": 0.2966,
      "step": 10830
    },
    {
      "epoch": 2.04,
      "grad_norm": 5.43375825881958,
      "learning_rate": 1.5919442875964618e-05,
      "loss": 0.2037,
      "step": 10840
    },
    {
      "epoch": 2.04,
      "grad_norm": 14.970441818237305,
      "learning_rate": 1.5915678524374178e-05,
      "loss": 0.2764,
      "step": 10850
    },
    {
      "epoch": 2.04,
      "grad_norm": 14.757416725158691,
      "learning_rate": 1.591191417278374e-05,
      "loss": 0.2707,
      "step": 10860
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.03912699967622757,
      "learning_rate": 1.59081498211933e-05,
      "loss": 0.0964,
      "step": 10870
    },
    {
      "epoch": 2.05,
      "grad_norm": 3.315324068069458,
      "learning_rate": 1.5904385469602864e-05,
      "loss": 0.3427,
      "step": 10880
    },
    {
      "epoch": 2.05,
      "grad_norm": 1.4862186908721924,
      "learning_rate": 1.5900621118012424e-05,
      "loss": 0.1062,
      "step": 10890
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.03727298974990845,
      "learning_rate": 1.5896856766421984e-05,
      "loss": 0.1726,
      "step": 10900
    },
    {
      "epoch": 2.05,
      "grad_norm": 27.964309692382812,
      "learning_rate": 1.5893092414831547e-05,
      "loss": 0.2408,
      "step": 10910
    },
    {
      "epoch": 2.06,
      "grad_norm": 8.196560859680176,
      "learning_rate": 1.5889328063241107e-05,
      "loss": 0.1679,
      "step": 10920
    },
    {
      "epoch": 2.06,
      "grad_norm": 8.691763877868652,
      "learning_rate": 1.588556371165067e-05,
      "loss": 0.4126,
      "step": 10930
    },
    {
      "epoch": 2.06,
      "grad_norm": 3.2335197925567627,
      "learning_rate": 1.588179936006023e-05,
      "loss": 0.2964,
      "step": 10940
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.09506280720233917,
      "learning_rate": 1.5878035008469793e-05,
      "loss": 0.2007,
      "step": 10950
    },
    {
      "epoch": 2.06,
      "grad_norm": 8.84326457977295,
      "learning_rate": 1.5874270656879352e-05,
      "loss": 0.2755,
      "step": 10960
    },
    {
      "epoch": 2.06,
      "grad_norm": 10.02408504486084,
      "learning_rate": 1.5870506305288916e-05,
      "loss": 0.1724,
      "step": 10970
    },
    {
      "epoch": 2.07,
      "grad_norm": 3.7499773502349854,
      "learning_rate": 1.5866741953698475e-05,
      "loss": 0.0843,
      "step": 10980
    },
    {
      "epoch": 2.07,
      "grad_norm": 4.502124786376953,
      "learning_rate": 1.586297760210804e-05,
      "loss": 0.2712,
      "step": 10990
    },
    {
      "epoch": 2.07,
      "grad_norm": 12.55028247833252,
      "learning_rate": 1.58592132505176e-05,
      "loss": 0.1953,
      "step": 11000
    },
    {
      "epoch": 2.07,
      "grad_norm": 11.702543258666992,
      "learning_rate": 1.585544889892716e-05,
      "loss": 0.3449,
      "step": 11010
    },
    {
      "epoch": 2.07,
      "grad_norm": 6.3455610275268555,
      "learning_rate": 1.5851684547336725e-05,
      "loss": 0.2738,
      "step": 11020
    },
    {
      "epoch": 2.08,
      "grad_norm": 20.903593063354492,
      "learning_rate": 1.5847920195746284e-05,
      "loss": 0.375,
      "step": 11030
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.24477726221084595,
      "learning_rate": 1.5844155844155847e-05,
      "loss": 0.4178,
      "step": 11040
    },
    {
      "epoch": 2.08,
      "grad_norm": 4.132774353027344,
      "learning_rate": 1.5840391492565407e-05,
      "loss": 0.3193,
      "step": 11050
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.4195270240306854,
      "learning_rate": 1.583662714097497e-05,
      "loss": 0.2587,
      "step": 11060
    },
    {
      "epoch": 2.08,
      "grad_norm": 7.835009574890137,
      "learning_rate": 1.583286278938453e-05,
      "loss": 0.103,
      "step": 11070
    },
    {
      "epoch": 2.09,
      "grad_norm": 7.435961723327637,
      "learning_rate": 1.582909843779409e-05,
      "loss": 0.3796,
      "step": 11080
    },
    {
      "epoch": 2.09,
      "grad_norm": 8.033628463745117,
      "learning_rate": 1.5825334086203653e-05,
      "loss": 0.2456,
      "step": 11090
    },
    {
      "epoch": 2.09,
      "grad_norm": 1.7867671251296997,
      "learning_rate": 1.5821569734613213e-05,
      "loss": 0.1728,
      "step": 11100
    },
    {
      "epoch": 2.09,
      "grad_norm": 8.073805809020996,
      "learning_rate": 1.5817805383022776e-05,
      "loss": 0.3218,
      "step": 11110
    },
    {
      "epoch": 2.09,
      "grad_norm": 16.999914169311523,
      "learning_rate": 1.5814041031432336e-05,
      "loss": 0.1918,
      "step": 11120
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.6607365012168884,
      "learning_rate": 1.58102766798419e-05,
      "loss": 0.3763,
      "step": 11130
    },
    {
      "epoch": 2.1,
      "grad_norm": 10.932785987854004,
      "learning_rate": 1.580651232825146e-05,
      "loss": 0.3011,
      "step": 11140
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.05750110000371933,
      "learning_rate": 1.5802747976661022e-05,
      "loss": 0.2824,
      "step": 11150
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.016762375831604,
      "learning_rate": 1.579898362507058e-05,
      "loss": 0.1903,
      "step": 11160
    },
    {
      "epoch": 2.1,
      "grad_norm": 19.681821823120117,
      "learning_rate": 1.5795219273480145e-05,
      "loss": 0.3866,
      "step": 11170
    },
    {
      "epoch": 2.1,
      "grad_norm": 4.102476596832275,
      "learning_rate": 1.5791454921889705e-05,
      "loss": 0.1491,
      "step": 11180
    },
    {
      "epoch": 2.11,
      "grad_norm": 4.127505302429199,
      "learning_rate": 1.5787690570299268e-05,
      "loss": 0.2247,
      "step": 11190
    },
    {
      "epoch": 2.11,
      "grad_norm": 11.143381118774414,
      "learning_rate": 1.578392621870883e-05,
      "loss": 0.1786,
      "step": 11200
    },
    {
      "epoch": 2.11,
      "grad_norm": 5.376995086669922,
      "learning_rate": 1.578016186711839e-05,
      "loss": 0.2119,
      "step": 11210
    },
    {
      "epoch": 2.11,
      "grad_norm": 11.91586685180664,
      "learning_rate": 1.5776397515527954e-05,
      "loss": 0.1476,
      "step": 11220
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.07449079304933548,
      "learning_rate": 1.5772633163937514e-05,
      "loss": 0.1573,
      "step": 11230
    },
    {
      "epoch": 2.12,
      "grad_norm": 34.288875579833984,
      "learning_rate": 1.5768868812347073e-05,
      "loss": 0.3537,
      "step": 11240
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.023568045347929,
      "learning_rate": 1.5765104460756636e-05,
      "loss": 0.1365,
      "step": 11250
    },
    {
      "epoch": 2.12,
      "grad_norm": 27.15109634399414,
      "learning_rate": 1.5761340109166196e-05,
      "loss": 0.3159,
      "step": 11260
    },
    {
      "epoch": 2.12,
      "grad_norm": 33.51382064819336,
      "learning_rate": 1.575757575757576e-05,
      "loss": 0.2789,
      "step": 11270
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.4020715057849884,
      "learning_rate": 1.575381140598532e-05,
      "loss": 0.2935,
      "step": 11280
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.6226245164871216,
      "learning_rate": 1.5750047054394882e-05,
      "loss": 0.0995,
      "step": 11290
    },
    {
      "epoch": 2.13,
      "grad_norm": 10.851766586303711,
      "learning_rate": 1.5746282702804442e-05,
      "loss": 0.2058,
      "step": 11300
    },
    {
      "epoch": 2.13,
      "grad_norm": 6.103633880615234,
      "learning_rate": 1.5742518351214005e-05,
      "loss": 0.3244,
      "step": 11310
    },
    {
      "epoch": 2.13,
      "grad_norm": 20.003332138061523,
      "learning_rate": 1.5738753999623565e-05,
      "loss": 0.2851,
      "step": 11320
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.19715793430805206,
      "learning_rate": 1.5734989648033128e-05,
      "loss": 0.295,
      "step": 11330
    },
    {
      "epoch": 2.13,
      "grad_norm": 45.27911376953125,
      "learning_rate": 1.5731225296442688e-05,
      "loss": 0.228,
      "step": 11340
    },
    {
      "epoch": 2.14,
      "grad_norm": 10.10094165802002,
      "learning_rate": 1.572746094485225e-05,
      "loss": 0.293,
      "step": 11350
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.4316902458667755,
      "learning_rate": 1.572369659326181e-05,
      "loss": 0.2179,
      "step": 11360
    },
    {
      "epoch": 2.14,
      "grad_norm": 16.158845901489258,
      "learning_rate": 1.5719932241671374e-05,
      "loss": 0.4242,
      "step": 11370
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.1779171228408813,
      "learning_rate": 1.5716167890080937e-05,
      "loss": 0.4027,
      "step": 11380
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.03330737724900246,
      "learning_rate": 1.5712403538490497e-05,
      "loss": 0.2985,
      "step": 11390
    },
    {
      "epoch": 2.15,
      "grad_norm": 12.913736343383789,
      "learning_rate": 1.570863918690006e-05,
      "loss": 0.2242,
      "step": 11400
    },
    {
      "epoch": 2.15,
      "grad_norm": 3.709968090057373,
      "learning_rate": 1.5704874835309616e-05,
      "loss": 0.2149,
      "step": 11410
    },
    {
      "epoch": 2.15,
      "grad_norm": 13.43122673034668,
      "learning_rate": 1.570111048371918e-05,
      "loss": 0.1671,
      "step": 11420
    },
    {
      "epoch": 2.15,
      "grad_norm": 17.276885986328125,
      "learning_rate": 1.5697346132128743e-05,
      "loss": 0.204,
      "step": 11430
    },
    {
      "epoch": 2.15,
      "grad_norm": 21.145790100097656,
      "learning_rate": 1.5693581780538303e-05,
      "loss": 0.3126,
      "step": 11440
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.13098600506782532,
      "learning_rate": 1.5689817428947866e-05,
      "loss": 0.2368,
      "step": 11450
    },
    {
      "epoch": 2.16,
      "grad_norm": 19.674537658691406,
      "learning_rate": 1.5686053077357425e-05,
      "loss": 0.2984,
      "step": 11460
    },
    {
      "epoch": 2.16,
      "grad_norm": 6.087719917297363,
      "learning_rate": 1.568228872576699e-05,
      "loss": 0.2699,
      "step": 11470
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.06172459200024605,
      "learning_rate": 1.567852437417655e-05,
      "loss": 0.3256,
      "step": 11480
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.03710721805691719,
      "learning_rate": 1.567476002258611e-05,
      "loss": 0.2581,
      "step": 11490
    },
    {
      "epoch": 2.16,
      "grad_norm": 7.31170129776001,
      "learning_rate": 1.567099567099567e-05,
      "loss": 0.3319,
      "step": 11500
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.08790650218725204,
      "learning_rate": 1.5667231319405234e-05,
      "loss": 0.3352,
      "step": 11510
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.06398595869541168,
      "learning_rate": 1.5663466967814794e-05,
      "loss": 0.4996,
      "step": 11520
    },
    {
      "epoch": 2.17,
      "grad_norm": 4.893272876739502,
      "learning_rate": 1.5659702616224357e-05,
      "loss": 0.2201,
      "step": 11530
    },
    {
      "epoch": 2.17,
      "grad_norm": 11.224406242370605,
      "learning_rate": 1.5655938264633917e-05,
      "loss": 0.3499,
      "step": 11540
    },
    {
      "epoch": 2.17,
      "grad_norm": 17.08206558227539,
      "learning_rate": 1.565217391304348e-05,
      "loss": 0.438,
      "step": 11550
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.05660124868154526,
      "learning_rate": 1.5648409561453043e-05,
      "loss": 0.1798,
      "step": 11560
    },
    {
      "epoch": 2.18,
      "grad_norm": 6.639573097229004,
      "learning_rate": 1.5644645209862603e-05,
      "loss": 0.1483,
      "step": 11570
    },
    {
      "epoch": 2.18,
      "grad_norm": 14.944281578063965,
      "learning_rate": 1.5640880858272166e-05,
      "loss": 0.2142,
      "step": 11580
    },
    {
      "epoch": 2.18,
      "grad_norm": 4.605640411376953,
      "learning_rate": 1.5637116506681723e-05,
      "loss": 0.1507,
      "step": 11590
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.45475712418556213,
      "learning_rate": 1.5633352155091286e-05,
      "loss": 0.2771,
      "step": 11600
    },
    {
      "epoch": 2.19,
      "grad_norm": 8.682119369506836,
      "learning_rate": 1.562958780350085e-05,
      "loss": 0.1995,
      "step": 11610
    },
    {
      "epoch": 2.19,
      "grad_norm": 30.454381942749023,
      "learning_rate": 1.562582345191041e-05,
      "loss": 0.1664,
      "step": 11620
    },
    {
      "epoch": 2.19,
      "grad_norm": 15.34536361694336,
      "learning_rate": 1.5622059100319972e-05,
      "loss": 0.2487,
      "step": 11630
    },
    {
      "epoch": 2.19,
      "grad_norm": 17.762868881225586,
      "learning_rate": 1.5618294748729532e-05,
      "loss": 0.2551,
      "step": 11640
    },
    {
      "epoch": 2.19,
      "grad_norm": 1.249190092086792,
      "learning_rate": 1.5614530397139095e-05,
      "loss": 0.194,
      "step": 11650
    },
    {
      "epoch": 2.19,
      "grad_norm": 17.51453971862793,
      "learning_rate": 1.5610766045548655e-05,
      "loss": 0.3595,
      "step": 11660
    },
    {
      "epoch": 2.2,
      "grad_norm": 5.651980400085449,
      "learning_rate": 1.5607001693958218e-05,
      "loss": 0.3132,
      "step": 11670
    },
    {
      "epoch": 2.2,
      "grad_norm": 1.3934309482574463,
      "learning_rate": 1.5603237342367778e-05,
      "loss": 0.2388,
      "step": 11680
    },
    {
      "epoch": 2.2,
      "grad_norm": 4.636242866516113,
      "learning_rate": 1.559947299077734e-05,
      "loss": 0.2819,
      "step": 11690
    },
    {
      "epoch": 2.2,
      "grad_norm": 18.109731674194336,
      "learning_rate": 1.55957086391869e-05,
      "loss": 0.1787,
      "step": 11700
    },
    {
      "epoch": 2.2,
      "grad_norm": 12.686445236206055,
      "learning_rate": 1.5591944287596464e-05,
      "loss": 0.462,
      "step": 11710
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.019882088527083397,
      "learning_rate": 1.5588179936006023e-05,
      "loss": 0.4454,
      "step": 11720
    },
    {
      "epoch": 2.21,
      "grad_norm": 23.785274505615234,
      "learning_rate": 1.5584415584415587e-05,
      "loss": 0.118,
      "step": 11730
    },
    {
      "epoch": 2.21,
      "grad_norm": 15.304242134094238,
      "learning_rate": 1.5580651232825146e-05,
      "loss": 0.219,
      "step": 11740
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.02687566541135311,
      "learning_rate": 1.557688688123471e-05,
      "loss": 0.302,
      "step": 11750
    },
    {
      "epoch": 2.21,
      "grad_norm": 7.878220081329346,
      "learning_rate": 1.557312252964427e-05,
      "loss": 0.2987,
      "step": 11760
    },
    {
      "epoch": 2.22,
      "grad_norm": 10.12842082977295,
      "learning_rate": 1.556935817805383e-05,
      "loss": 0.1431,
      "step": 11770
    },
    {
      "epoch": 2.22,
      "grad_norm": 5.604041576385498,
      "learning_rate": 1.5565593826463392e-05,
      "loss": 0.2718,
      "step": 11780
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.053012438118457794,
      "learning_rate": 1.5561829474872955e-05,
      "loss": 0.1872,
      "step": 11790
    },
    {
      "epoch": 2.22,
      "grad_norm": 4.071653366088867,
      "learning_rate": 1.5558065123282515e-05,
      "loss": 0.2144,
      "step": 11800
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.17044880986213684,
      "learning_rate": 1.5554300771692078e-05,
      "loss": 0.4359,
      "step": 11810
    },
    {
      "epoch": 2.22,
      "grad_norm": 13.346423149108887,
      "learning_rate": 1.5550536420101638e-05,
      "loss": 0.2969,
      "step": 11820
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.813899040222168,
      "learning_rate": 1.55467720685112e-05,
      "loss": 0.3415,
      "step": 11830
    },
    {
      "epoch": 2.23,
      "grad_norm": 14.354334831237793,
      "learning_rate": 1.554300771692076e-05,
      "loss": 0.399,
      "step": 11840
    },
    {
      "epoch": 2.23,
      "grad_norm": 6.637619495391846,
      "learning_rate": 1.5539243365330324e-05,
      "loss": 0.4078,
      "step": 11850
    },
    {
      "epoch": 2.23,
      "grad_norm": 2.074040174484253,
      "learning_rate": 1.5535479013739884e-05,
      "loss": 0.1784,
      "step": 11860
    },
    {
      "epoch": 2.23,
      "grad_norm": 1.1945127248764038,
      "learning_rate": 1.5531714662149447e-05,
      "loss": 0.218,
      "step": 11870
    },
    {
      "epoch": 2.24,
      "grad_norm": 9.09565544128418,
      "learning_rate": 1.5527950310559007e-05,
      "loss": 0.2696,
      "step": 11880
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.04204779490828514,
      "learning_rate": 1.552418595896857e-05,
      "loss": 0.4272,
      "step": 11890
    },
    {
      "epoch": 2.24,
      "grad_norm": 8.610626220703125,
      "learning_rate": 1.552042160737813e-05,
      "loss": 0.2247,
      "step": 11900
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.8028196096420288,
      "learning_rate": 1.5516657255787693e-05,
      "loss": 0.1584,
      "step": 11910
    },
    {
      "epoch": 2.24,
      "grad_norm": 10.798333168029785,
      "learning_rate": 1.5512892904197253e-05,
      "loss": 0.2658,
      "step": 11920
    },
    {
      "epoch": 2.25,
      "grad_norm": 2.247654676437378,
      "learning_rate": 1.5509128552606816e-05,
      "loss": 0.2047,
      "step": 11930
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.03462931141257286,
      "learning_rate": 1.5505364201016376e-05,
      "loss": 0.4586,
      "step": 11940
    },
    {
      "epoch": 2.25,
      "grad_norm": 12.61550235748291,
      "learning_rate": 1.5501599849425935e-05,
      "loss": 0.3065,
      "step": 11950
    },
    {
      "epoch": 2.25,
      "grad_norm": 37.06946563720703,
      "learning_rate": 1.54978354978355e-05,
      "loss": 0.2103,
      "step": 11960
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.018287409096956253,
      "learning_rate": 1.5494071146245058e-05,
      "loss": 0.2301,
      "step": 11970
    },
    {
      "epoch": 2.25,
      "grad_norm": 5.742473602294922,
      "learning_rate": 1.549030679465462e-05,
      "loss": 0.1518,
      "step": 11980
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.019625265151262283,
      "learning_rate": 1.5486542443064185e-05,
      "loss": 0.2602,
      "step": 11990
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.24242456257343292,
      "learning_rate": 1.5482778091473744e-05,
      "loss": 0.2306,
      "step": 12000
    },
    {
      "epoch": 2.26,
      "grad_norm": 4.092889308929443,
      "learning_rate": 1.5479013739883307e-05,
      "loss": 0.3122,
      "step": 12010
    },
    {
      "epoch": 2.26,
      "grad_norm": 6.170224666595459,
      "learning_rate": 1.5475249388292867e-05,
      "loss": 0.4591,
      "step": 12020
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.25296902656555176,
      "learning_rate": 1.547148503670243e-05,
      "loss": 0.1665,
      "step": 12030
    },
    {
      "epoch": 2.27,
      "grad_norm": 27.29945945739746,
      "learning_rate": 1.546772068511199e-05,
      "loss": 0.3636,
      "step": 12040
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.20417575538158417,
      "learning_rate": 1.5463956333521553e-05,
      "loss": 0.1329,
      "step": 12050
    },
    {
      "epoch": 2.27,
      "grad_norm": 11.166847229003906,
      "learning_rate": 1.5460191981931113e-05,
      "loss": 0.2383,
      "step": 12060
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.2848060727119446,
      "learning_rate": 1.5456427630340676e-05,
      "loss": 0.4669,
      "step": 12070
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.11858190596103668,
      "learning_rate": 1.5452663278750236e-05,
      "loss": 0.2716,
      "step": 12080
    },
    {
      "epoch": 2.28,
      "grad_norm": 1.5967313051223755,
      "learning_rate": 1.54488989271598e-05,
      "loss": 0.2944,
      "step": 12090
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.022934604436159134,
      "learning_rate": 1.544513457556936e-05,
      "loss": 0.3272,
      "step": 12100
    },
    {
      "epoch": 2.28,
      "grad_norm": 5.362580299377441,
      "learning_rate": 1.544137022397892e-05,
      "loss": 0.0747,
      "step": 12110
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.07393370568752289,
      "learning_rate": 1.5437605872388482e-05,
      "loss": 0.2064,
      "step": 12120
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.09617757797241211,
      "learning_rate": 1.543384152079804e-05,
      "loss": 0.0709,
      "step": 12130
    },
    {
      "epoch": 2.28,
      "grad_norm": 8.564157485961914,
      "learning_rate": 1.5430077169207605e-05,
      "loss": 0.1342,
      "step": 12140
    },
    {
      "epoch": 2.29,
      "grad_norm": 10.776432991027832,
      "learning_rate": 1.5426312817617165e-05,
      "loss": 0.1027,
      "step": 12150
    },
    {
      "epoch": 2.29,
      "grad_norm": 2.7552130222320557,
      "learning_rate": 1.5422548466026728e-05,
      "loss": 0.3862,
      "step": 12160
    },
    {
      "epoch": 2.29,
      "grad_norm": 5.673059463500977,
      "learning_rate": 1.541878411443629e-05,
      "loss": 0.2473,
      "step": 12170
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.06814634054899216,
      "learning_rate": 1.541501976284585e-05,
      "loss": 0.1984,
      "step": 12180
    },
    {
      "epoch": 2.29,
      "grad_norm": 9.430500030517578,
      "learning_rate": 1.5411255411255414e-05,
      "loss": 0.3962,
      "step": 12190
    },
    {
      "epoch": 2.3,
      "grad_norm": 3.9334232807159424,
      "learning_rate": 1.5407491059664974e-05,
      "loss": 0.0622,
      "step": 12200
    },
    {
      "epoch": 2.3,
      "grad_norm": 2.789630889892578,
      "learning_rate": 1.5403726708074537e-05,
      "loss": 0.2123,
      "step": 12210
    },
    {
      "epoch": 2.3,
      "grad_norm": 4.516412734985352,
      "learning_rate": 1.5399962356484096e-05,
      "loss": 0.2493,
      "step": 12220
    },
    {
      "epoch": 2.3,
      "grad_norm": 32.90705871582031,
      "learning_rate": 1.539619800489366e-05,
      "loss": 0.2589,
      "step": 12230
    },
    {
      "epoch": 2.3,
      "grad_norm": 25.8496036529541,
      "learning_rate": 1.539243365330322e-05,
      "loss": 0.1915,
      "step": 12240
    },
    {
      "epoch": 2.31,
      "grad_norm": 22.368656158447266,
      "learning_rate": 1.5388669301712783e-05,
      "loss": 0.2521,
      "step": 12250
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.5335608124732971,
      "learning_rate": 1.5384904950122342e-05,
      "loss": 0.385,
      "step": 12260
    },
    {
      "epoch": 2.31,
      "grad_norm": 12.594615936279297,
      "learning_rate": 1.5381140598531905e-05,
      "loss": 0.1973,
      "step": 12270
    },
    {
      "epoch": 2.31,
      "grad_norm": 11.365102767944336,
      "learning_rate": 1.5377376246941465e-05,
      "loss": 0.5312,
      "step": 12280
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.09339795261621475,
      "learning_rate": 1.5373611895351025e-05,
      "loss": 0.2688,
      "step": 12290
    },
    {
      "epoch": 2.32,
      "grad_norm": 10.312443733215332,
      "learning_rate": 1.5369847543760588e-05,
      "loss": 0.2137,
      "step": 12300
    },
    {
      "epoch": 2.32,
      "grad_norm": 20.362913131713867,
      "learning_rate": 1.5366083192170148e-05,
      "loss": 0.1916,
      "step": 12310
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.5917559862136841,
      "learning_rate": 1.536231884057971e-05,
      "loss": 0.3086,
      "step": 12320
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.031630489975214005,
      "learning_rate": 1.535855448898927e-05,
      "loss": 0.1217,
      "step": 12330
    },
    {
      "epoch": 2.32,
      "grad_norm": 8.389304161071777,
      "learning_rate": 1.5354790137398834e-05,
      "loss": 0.2131,
      "step": 12340
    },
    {
      "epoch": 2.32,
      "grad_norm": 12.692630767822266,
      "learning_rate": 1.5351025785808397e-05,
      "loss": 0.28,
      "step": 12350
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.1496800184249878,
      "learning_rate": 1.5347261434217957e-05,
      "loss": 0.4131,
      "step": 12360
    },
    {
      "epoch": 2.33,
      "grad_norm": 11.977798461914062,
      "learning_rate": 1.534349708262752e-05,
      "loss": 0.2663,
      "step": 12370
    },
    {
      "epoch": 2.33,
      "grad_norm": 5.761414051055908,
      "learning_rate": 1.533973273103708e-05,
      "loss": 0.2617,
      "step": 12380
    },
    {
      "epoch": 2.33,
      "grad_norm": 2.0424273014068604,
      "learning_rate": 1.5335968379446643e-05,
      "loss": 0.2062,
      "step": 12390
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.5645923614501953,
      "learning_rate": 1.5332204027856203e-05,
      "loss": 0.293,
      "step": 12400
    },
    {
      "epoch": 2.34,
      "grad_norm": 23.933063507080078,
      "learning_rate": 1.5328439676265766e-05,
      "loss": 0.3886,
      "step": 12410
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.15198612213134766,
      "learning_rate": 1.5324675324675326e-05,
      "loss": 0.3484,
      "step": 12420
    },
    {
      "epoch": 2.34,
      "grad_norm": 10.579571723937988,
      "learning_rate": 1.532091097308489e-05,
      "loss": 0.1587,
      "step": 12430
    },
    {
      "epoch": 2.34,
      "grad_norm": 19.620285034179688,
      "learning_rate": 1.531714662149445e-05,
      "loss": 0.1553,
      "step": 12440
    },
    {
      "epoch": 2.34,
      "grad_norm": 20.725751876831055,
      "learning_rate": 1.5313382269904012e-05,
      "loss": 0.3803,
      "step": 12450
    },
    {
      "epoch": 2.35,
      "grad_norm": 3.821186065673828,
      "learning_rate": 1.530961791831357e-05,
      "loss": 0.4889,
      "step": 12460
    },
    {
      "epoch": 2.35,
      "grad_norm": 8.757345199584961,
      "learning_rate": 1.530585356672313e-05,
      "loss": 0.3642,
      "step": 12470
    },
    {
      "epoch": 2.35,
      "grad_norm": 9.593331336975098,
      "learning_rate": 1.5302089215132694e-05,
      "loss": 0.2853,
      "step": 12480
    },
    {
      "epoch": 2.35,
      "grad_norm": 4.736720561981201,
      "learning_rate": 1.5298324863542254e-05,
      "loss": 0.1535,
      "step": 12490
    },
    {
      "epoch": 2.35,
      "grad_norm": 4.302800178527832,
      "learning_rate": 1.5294560511951817e-05,
      "loss": 0.1366,
      "step": 12500
    },
    {
      "epoch": 2.35,
      "grad_norm": 6.378731727600098,
      "learning_rate": 1.5290796160361377e-05,
      "loss": 0.3135,
      "step": 12510
    },
    {
      "epoch": 2.36,
      "grad_norm": 20.473796844482422,
      "learning_rate": 1.528703180877094e-05,
      "loss": 0.0776,
      "step": 12520
    },
    {
      "epoch": 2.36,
      "grad_norm": 21.092369079589844,
      "learning_rate": 1.5283267457180503e-05,
      "loss": 0.412,
      "step": 12530
    },
    {
      "epoch": 2.36,
      "grad_norm": 23.93069839477539,
      "learning_rate": 1.5279503105590063e-05,
      "loss": 0.3048,
      "step": 12540
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.014215127564966679,
      "learning_rate": 1.5275738753999626e-05,
      "loss": 0.2414,
      "step": 12550
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.1275087296962738,
      "learning_rate": 1.5271974402409186e-05,
      "loss": 0.3082,
      "step": 12560
    },
    {
      "epoch": 2.37,
      "grad_norm": 14.308680534362793,
      "learning_rate": 1.526821005081875e-05,
      "loss": 0.2825,
      "step": 12570
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.03820957988500595,
      "learning_rate": 1.526444569922831e-05,
      "loss": 0.285,
      "step": 12580
    },
    {
      "epoch": 2.37,
      "grad_norm": 35.3873291015625,
      "learning_rate": 1.5260681347637872e-05,
      "loss": 0.1261,
      "step": 12590
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.08324924111366272,
      "learning_rate": 1.5256916996047434e-05,
      "loss": 0.1017,
      "step": 12600
    },
    {
      "epoch": 2.37,
      "grad_norm": 12.333545684814453,
      "learning_rate": 1.5253152644456995e-05,
      "loss": 0.3062,
      "step": 12610
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.46221014857292175,
      "learning_rate": 1.5249388292866557e-05,
      "loss": 0.4337,
      "step": 12620
    },
    {
      "epoch": 2.38,
      "grad_norm": 6.795083045959473,
      "learning_rate": 1.5245623941276115e-05,
      "loss": 0.0959,
      "step": 12630
    },
    {
      "epoch": 2.38,
      "grad_norm": 13.661226272583008,
      "learning_rate": 1.5241859589685676e-05,
      "loss": 0.1676,
      "step": 12640
    },
    {
      "epoch": 2.38,
      "grad_norm": 9.006892204284668,
      "learning_rate": 1.523809523809524e-05,
      "loss": 0.2698,
      "step": 12650
    },
    {
      "epoch": 2.38,
      "grad_norm": 11.457891464233398,
      "learning_rate": 1.52343308865048e-05,
      "loss": 0.4495,
      "step": 12660
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.03691503405570984,
      "learning_rate": 1.5230566534914362e-05,
      "loss": 0.2196,
      "step": 12670
    },
    {
      "epoch": 2.39,
      "grad_norm": 1.6712708473205566,
      "learning_rate": 1.5226802183323924e-05,
      "loss": 0.1751,
      "step": 12680
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.25265273451805115,
      "learning_rate": 1.5223037831733485e-05,
      "loss": 0.2372,
      "step": 12690
    },
    {
      "epoch": 2.39,
      "grad_norm": 8.51880931854248,
      "learning_rate": 1.5219273480143047e-05,
      "loss": 0.2155,
      "step": 12700
    },
    {
      "epoch": 2.39,
      "grad_norm": 26.969141006469727,
      "learning_rate": 1.5215509128552608e-05,
      "loss": 0.3305,
      "step": 12710
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.24688658118247986,
      "learning_rate": 1.521174477696217e-05,
      "loss": 0.2505,
      "step": 12720
    },
    {
      "epoch": 2.4,
      "grad_norm": 7.455538272857666,
      "learning_rate": 1.5207980425371731e-05,
      "loss": 0.493,
      "step": 12730
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.042772602289915085,
      "learning_rate": 1.5204216073781292e-05,
      "loss": 0.1615,
      "step": 12740
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.060405321419239044,
      "learning_rate": 1.5200451722190854e-05,
      "loss": 0.3726,
      "step": 12750
    },
    {
      "epoch": 2.4,
      "grad_norm": 36.193660736083984,
      "learning_rate": 1.5196687370600415e-05,
      "loss": 0.1488,
      "step": 12760
    },
    {
      "epoch": 2.4,
      "grad_norm": 8.312196731567383,
      "learning_rate": 1.5192923019009977e-05,
      "loss": 0.3302,
      "step": 12770
    },
    {
      "epoch": 2.41,
      "grad_norm": 2.6934854984283447,
      "learning_rate": 1.5189158667419538e-05,
      "loss": 0.0748,
      "step": 12780
    },
    {
      "epoch": 2.41,
      "grad_norm": 25.02640151977539,
      "learning_rate": 1.5185394315829101e-05,
      "loss": 0.3117,
      "step": 12790
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.039934366941452026,
      "learning_rate": 1.5181629964238663e-05,
      "loss": 0.2948,
      "step": 12800
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.04921785369515419,
      "learning_rate": 1.5177865612648221e-05,
      "loss": 0.1635,
      "step": 12810
    },
    {
      "epoch": 2.41,
      "grad_norm": 13.582701683044434,
      "learning_rate": 1.5174101261057782e-05,
      "loss": 0.1183,
      "step": 12820
    },
    {
      "epoch": 2.41,
      "grad_norm": 8.666038513183594,
      "learning_rate": 1.5170336909467346e-05,
      "loss": 0.2173,
      "step": 12830
    },
    {
      "epoch": 2.42,
      "grad_norm": 23.320066452026367,
      "learning_rate": 1.5166572557876907e-05,
      "loss": 0.1321,
      "step": 12840
    },
    {
      "epoch": 2.42,
      "grad_norm": 3.018787145614624,
      "learning_rate": 1.5162808206286468e-05,
      "loss": 0.3733,
      "step": 12850
    },
    {
      "epoch": 2.42,
      "grad_norm": 17.136816024780273,
      "learning_rate": 1.515904385469603e-05,
      "loss": 0.2702,
      "step": 12860
    },
    {
      "epoch": 2.42,
      "grad_norm": 5.929393291473389,
      "learning_rate": 1.5155279503105591e-05,
      "loss": 0.2852,
      "step": 12870
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.1843037605285645,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.3144,
      "step": 12880
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.5411343574523926,
      "learning_rate": 1.5147750799924714e-05,
      "loss": 0.4353,
      "step": 12890
    },
    {
      "epoch": 2.43,
      "grad_norm": 20.22041893005371,
      "learning_rate": 1.5143986448334276e-05,
      "loss": 0.3757,
      "step": 12900
    },
    {
      "epoch": 2.43,
      "grad_norm": 4.425317764282227,
      "learning_rate": 1.5140222096743837e-05,
      "loss": 0.3327,
      "step": 12910
    },
    {
      "epoch": 2.43,
      "grad_norm": 18.777544021606445,
      "learning_rate": 1.5136457745153399e-05,
      "loss": 0.1719,
      "step": 12920
    },
    {
      "epoch": 2.43,
      "grad_norm": 2.6958069801330566,
      "learning_rate": 1.513269339356296e-05,
      "loss": 0.2756,
      "step": 12930
    },
    {
      "epoch": 2.44,
      "grad_norm": 3.262017250061035,
      "learning_rate": 1.5128929041972522e-05,
      "loss": 0.1243,
      "step": 12940
    },
    {
      "epoch": 2.44,
      "grad_norm": 5.671223163604736,
      "learning_rate": 1.5125164690382083e-05,
      "loss": 0.2296,
      "step": 12950
    },
    {
      "epoch": 2.44,
      "grad_norm": 7.30415153503418,
      "learning_rate": 1.5121400338791645e-05,
      "loss": 0.3656,
      "step": 12960
    },
    {
      "epoch": 2.44,
      "grad_norm": 14.557415962219238,
      "learning_rate": 1.5117635987201206e-05,
      "loss": 0.4846,
      "step": 12970
    },
    {
      "epoch": 2.44,
      "grad_norm": 18.728322982788086,
      "learning_rate": 1.5113871635610766e-05,
      "loss": 0.2723,
      "step": 12980
    },
    {
      "epoch": 2.44,
      "grad_norm": 1.7387354373931885,
      "learning_rate": 1.5110107284020327e-05,
      "loss": 0.4286,
      "step": 12990
    },
    {
      "epoch": 2.45,
      "grad_norm": 6.253885269165039,
      "learning_rate": 1.5106342932429889e-05,
      "loss": 0.2231,
      "step": 13000
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.04833545163273811,
      "learning_rate": 1.510257858083945e-05,
      "loss": 0.229,
      "step": 13010
    },
    {
      "epoch": 2.45,
      "grad_norm": 5.138844966888428,
      "learning_rate": 1.5098814229249013e-05,
      "loss": 0.1786,
      "step": 13020
    },
    {
      "epoch": 2.45,
      "grad_norm": 7.677248477935791,
      "learning_rate": 1.5095049877658575e-05,
      "loss": 0.3468,
      "step": 13030
    },
    {
      "epoch": 2.45,
      "grad_norm": 8.792970657348633,
      "learning_rate": 1.5091285526068136e-05,
      "loss": 0.408,
      "step": 13040
    },
    {
      "epoch": 2.46,
      "grad_norm": 31.7252197265625,
      "learning_rate": 1.5087521174477698e-05,
      "loss": 0.2573,
      "step": 13050
    },
    {
      "epoch": 2.46,
      "grad_norm": 18.50107765197754,
      "learning_rate": 1.5083756822887259e-05,
      "loss": 0.2775,
      "step": 13060
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.02156536467373371,
      "learning_rate": 1.507999247129682e-05,
      "loss": 0.1274,
      "step": 13070
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.030990909785032272,
      "learning_rate": 1.5076228119706382e-05,
      "loss": 0.1981,
      "step": 13080
    },
    {
      "epoch": 2.46,
      "grad_norm": 1.8555041551589966,
      "learning_rate": 1.5072463768115944e-05,
      "loss": 0.1907,
      "step": 13090
    },
    {
      "epoch": 2.47,
      "grad_norm": 19.9966983795166,
      "learning_rate": 1.5068699416525505e-05,
      "loss": 0.2913,
      "step": 13100
    },
    {
      "epoch": 2.47,
      "grad_norm": 4.910961151123047,
      "learning_rate": 1.5064935064935066e-05,
      "loss": 0.3144,
      "step": 13110
    },
    {
      "epoch": 2.47,
      "grad_norm": 2.8181371688842773,
      "learning_rate": 1.5061170713344628e-05,
      "loss": 0.2469,
      "step": 13120
    },
    {
      "epoch": 2.47,
      "grad_norm": 5.139735221862793,
      "learning_rate": 1.505740636175419e-05,
      "loss": 0.2083,
      "step": 13130
    },
    {
      "epoch": 2.47,
      "grad_norm": 14.704835891723633,
      "learning_rate": 1.505364201016375e-05,
      "loss": 0.2191,
      "step": 13140
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.04894518479704857,
      "learning_rate": 1.5049877658573312e-05,
      "loss": 0.2266,
      "step": 13150
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.4192140102386475,
      "learning_rate": 1.5046113306982872e-05,
      "loss": 0.2374,
      "step": 13160
    },
    {
      "epoch": 2.48,
      "grad_norm": 7.141027927398682,
      "learning_rate": 1.5042348955392434e-05,
      "loss": 0.1756,
      "step": 13170
    },
    {
      "epoch": 2.48,
      "grad_norm": 3.5215392112731934,
      "learning_rate": 1.5038584603801995e-05,
      "loss": 0.2789,
      "step": 13180
    },
    {
      "epoch": 2.48,
      "grad_norm": 6.916445255279541,
      "learning_rate": 1.5034820252211556e-05,
      "loss": 0.2489,
      "step": 13190
    },
    {
      "epoch": 2.48,
      "grad_norm": 2.5890514850616455,
      "learning_rate": 1.5031055900621118e-05,
      "loss": 0.2353,
      "step": 13200
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.021378763020038605,
      "learning_rate": 1.5027291549030681e-05,
      "loss": 0.4178,
      "step": 13210
    },
    {
      "epoch": 2.49,
      "grad_norm": 23.215621948242188,
      "learning_rate": 1.5023527197440243e-05,
      "loss": 0.2898,
      "step": 13220
    },
    {
      "epoch": 2.49,
      "grad_norm": 4.362763404846191,
      "learning_rate": 1.5019762845849804e-05,
      "loss": 0.3021,
      "step": 13230
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.017629658803343773,
      "learning_rate": 1.5015998494259365e-05,
      "loss": 0.2268,
      "step": 13240
    },
    {
      "epoch": 2.49,
      "grad_norm": 23.738834381103516,
      "learning_rate": 1.5012234142668927e-05,
      "loss": 0.1801,
      "step": 13250
    },
    {
      "epoch": 2.5,
      "grad_norm": 19.871145248413086,
      "learning_rate": 1.5008469791078488e-05,
      "loss": 0.287,
      "step": 13260
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.030364952981472015,
      "learning_rate": 1.500470543948805e-05,
      "loss": 0.1476,
      "step": 13270
    },
    {
      "epoch": 2.5,
      "grad_norm": 3.811730146408081,
      "learning_rate": 1.5000941087897611e-05,
      "loss": 0.4736,
      "step": 13280
    },
    {
      "epoch": 2.5,
      "grad_norm": 14.833952903747559,
      "learning_rate": 1.4997176736307173e-05,
      "loss": 0.2641,
      "step": 13290
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.6472156047821045,
      "learning_rate": 1.4993412384716734e-05,
      "loss": 0.1803,
      "step": 13300
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.09927736222743988,
      "learning_rate": 1.4989648033126296e-05,
      "loss": 0.2438,
      "step": 13310
    },
    {
      "epoch": 2.51,
      "grad_norm": 16.08211326599121,
      "learning_rate": 1.4985883681535857e-05,
      "loss": 0.2445,
      "step": 13320
    },
    {
      "epoch": 2.51,
      "grad_norm": 13.33660888671875,
      "learning_rate": 1.4982119329945417e-05,
      "loss": 0.2836,
      "step": 13330
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.4387320280075073,
      "learning_rate": 1.4978354978354978e-05,
      "loss": 0.1359,
      "step": 13340
    },
    {
      "epoch": 2.51,
      "grad_norm": 21.024782180786133,
      "learning_rate": 1.497459062676454e-05,
      "loss": 0.2203,
      "step": 13350
    },
    {
      "epoch": 2.51,
      "grad_norm": 13.724315643310547,
      "learning_rate": 1.4970826275174101e-05,
      "loss": 0.2202,
      "step": 13360
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.020193064585328102,
      "learning_rate": 1.4967061923583663e-05,
      "loss": 0.2333,
      "step": 13370
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.4651835262775421,
      "learning_rate": 1.4963297571993224e-05,
      "loss": 0.1371,
      "step": 13380
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.6778590083122253,
      "learning_rate": 1.4959533220402787e-05,
      "loss": 0.1746,
      "step": 13390
    },
    {
      "epoch": 2.52,
      "grad_norm": 19.456600189208984,
      "learning_rate": 1.4955768868812349e-05,
      "loss": 0.18,
      "step": 13400
    },
    {
      "epoch": 2.52,
      "grad_norm": 1.9717931747436523,
      "learning_rate": 1.495200451722191e-05,
      "loss": 0.4092,
      "step": 13410
    },
    {
      "epoch": 2.53,
      "grad_norm": 19.983518600463867,
      "learning_rate": 1.4948240165631472e-05,
      "loss": 0.2635,
      "step": 13420
    },
    {
      "epoch": 2.53,
      "grad_norm": 5.28502082824707,
      "learning_rate": 1.4944475814041033e-05,
      "loss": 0.145,
      "step": 13430
    },
    {
      "epoch": 2.53,
      "grad_norm": 5.6852521896362305,
      "learning_rate": 1.4940711462450595e-05,
      "loss": 0.4588,
      "step": 13440
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.05631304904818535,
      "learning_rate": 1.4936947110860156e-05,
      "loss": 0.1757,
      "step": 13450
    },
    {
      "epoch": 2.53,
      "grad_norm": 13.865554809570312,
      "learning_rate": 1.4933182759269718e-05,
      "loss": 0.2462,
      "step": 13460
    },
    {
      "epoch": 2.54,
      "grad_norm": 7.8119797706604,
      "learning_rate": 1.4929418407679279e-05,
      "loss": 0.2183,
      "step": 13470
    },
    {
      "epoch": 2.54,
      "grad_norm": 3.6569607257843018,
      "learning_rate": 1.492565405608884e-05,
      "loss": 0.278,
      "step": 13480
    },
    {
      "epoch": 2.54,
      "grad_norm": 9.667588233947754,
      "learning_rate": 1.4921889704498402e-05,
      "loss": 0.2342,
      "step": 13490
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.01957961730659008,
      "learning_rate": 1.4918125352907963e-05,
      "loss": 0.124,
      "step": 13500
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.24761758744716644,
      "learning_rate": 1.4914361001317523e-05,
      "loss": 0.0949,
      "step": 13510
    },
    {
      "epoch": 2.54,
      "grad_norm": 4.8042402267456055,
      "learning_rate": 1.4910596649727085e-05,
      "loss": 0.2248,
      "step": 13520
    },
    {
      "epoch": 2.55,
      "grad_norm": 25.022092819213867,
      "learning_rate": 1.4906832298136646e-05,
      "loss": 0.2172,
      "step": 13530
    },
    {
      "epoch": 2.55,
      "grad_norm": 1.6538975238800049,
      "learning_rate": 1.4903067946546208e-05,
      "loss": 0.1259,
      "step": 13540
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.026058997958898544,
      "learning_rate": 1.4899303594955769e-05,
      "loss": 0.225,
      "step": 13550
    },
    {
      "epoch": 2.55,
      "grad_norm": 70.18437957763672,
      "learning_rate": 1.489553924336533e-05,
      "loss": 0.3298,
      "step": 13560
    },
    {
      "epoch": 2.55,
      "grad_norm": 9.744485855102539,
      "learning_rate": 1.4891774891774892e-05,
      "loss": 0.3324,
      "step": 13570
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.047737713903188705,
      "learning_rate": 1.4888010540184455e-05,
      "loss": 0.237,
      "step": 13580
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.09070626646280289,
      "learning_rate": 1.4884246188594017e-05,
      "loss": 0.2349,
      "step": 13590
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.04411903768777847,
      "learning_rate": 1.4880481837003578e-05,
      "loss": 0.099,
      "step": 13600
    },
    {
      "epoch": 2.56,
      "grad_norm": 13.464763641357422,
      "learning_rate": 1.487671748541314e-05,
      "loss": 0.2013,
      "step": 13610
    },
    {
      "epoch": 2.56,
      "grad_norm": 5.462995529174805,
      "learning_rate": 1.4872953133822701e-05,
      "loss": 0.0885,
      "step": 13620
    },
    {
      "epoch": 2.57,
      "grad_norm": 7.152541637420654,
      "learning_rate": 1.4869188782232262e-05,
      "loss": 0.4785,
      "step": 13630
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.8324568271636963,
      "learning_rate": 1.4865424430641824e-05,
      "loss": 0.2718,
      "step": 13640
    },
    {
      "epoch": 2.57,
      "grad_norm": 15.500285148620605,
      "learning_rate": 1.4861660079051385e-05,
      "loss": 0.2873,
      "step": 13650
    },
    {
      "epoch": 2.57,
      "grad_norm": 3.914032220840454,
      "learning_rate": 1.4857895727460947e-05,
      "loss": 0.303,
      "step": 13660
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.012785468250513077,
      "learning_rate": 1.4854131375870508e-05,
      "loss": 0.1363,
      "step": 13670
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.37162280082702637,
      "learning_rate": 1.4850367024280068e-05,
      "loss": 0.3618,
      "step": 13680
    },
    {
      "epoch": 2.58,
      "grad_norm": 7.617442607879639,
      "learning_rate": 1.484660267268963e-05,
      "loss": 0.2989,
      "step": 13690
    },
    {
      "epoch": 2.58,
      "grad_norm": 4.410751819610596,
      "learning_rate": 1.4842838321099191e-05,
      "loss": 0.3414,
      "step": 13700
    },
    {
      "epoch": 2.58,
      "grad_norm": 30.439998626708984,
      "learning_rate": 1.4839073969508752e-05,
      "loss": 0.3571,
      "step": 13710
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.6361440420150757,
      "learning_rate": 1.4835309617918314e-05,
      "loss": 0.2506,
      "step": 13720
    },
    {
      "epoch": 2.58,
      "grad_norm": 1.465313196182251,
      "learning_rate": 1.4831545266327875e-05,
      "loss": 0.1576,
      "step": 13730
    },
    {
      "epoch": 2.59,
      "grad_norm": 23.712356567382812,
      "learning_rate": 1.4827780914737437e-05,
      "loss": 0.3379,
      "step": 13740
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.3175652027130127,
      "learning_rate": 1.4824016563146998e-05,
      "loss": 0.3623,
      "step": 13750
    },
    {
      "epoch": 2.59,
      "grad_norm": 2.747946262359619,
      "learning_rate": 1.4820252211556561e-05,
      "loss": 0.3166,
      "step": 13760
    },
    {
      "epoch": 2.59,
      "grad_norm": 4.122313499450684,
      "learning_rate": 1.4816487859966123e-05,
      "loss": 0.2164,
      "step": 13770
    },
    {
      "epoch": 2.59,
      "grad_norm": 9.40050220489502,
      "learning_rate": 1.4812723508375684e-05,
      "loss": 0.1246,
      "step": 13780
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.025036897510290146,
      "learning_rate": 1.4808959156785246e-05,
      "loss": 0.3535,
      "step": 13790
    },
    {
      "epoch": 2.6,
      "grad_norm": 14.766253471374512,
      "learning_rate": 1.4805194805194807e-05,
      "loss": 0.2566,
      "step": 13800
    },
    {
      "epoch": 2.6,
      "grad_norm": 4.560943603515625,
      "learning_rate": 1.4801430453604369e-05,
      "loss": 0.4398,
      "step": 13810
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.5036472678184509,
      "learning_rate": 1.479766610201393e-05,
      "loss": 0.2229,
      "step": 13820
    },
    {
      "epoch": 2.6,
      "grad_norm": 2.668865203857422,
      "learning_rate": 1.4793901750423492e-05,
      "loss": 0.3484,
      "step": 13830
    },
    {
      "epoch": 2.6,
      "grad_norm": 20.500673294067383,
      "learning_rate": 1.4790137398833053e-05,
      "loss": 0.3176,
      "step": 13840
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.18162061274051666,
      "learning_rate": 1.4786373047242613e-05,
      "loss": 0.2284,
      "step": 13850
    },
    {
      "epoch": 2.61,
      "grad_norm": 10.189698219299316,
      "learning_rate": 1.4782608695652174e-05,
      "loss": 0.3275,
      "step": 13860
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.785983681678772,
      "learning_rate": 1.4778844344061736e-05,
      "loss": 0.3027,
      "step": 13870
    },
    {
      "epoch": 2.61,
      "grad_norm": 5.628896713256836,
      "learning_rate": 1.4775079992471297e-05,
      "loss": 0.2041,
      "step": 13880
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.15929974615573883,
      "learning_rate": 1.4771315640880859e-05,
      "loss": 0.2343,
      "step": 13890
    },
    {
      "epoch": 2.62,
      "grad_norm": 21.28129005432129,
      "learning_rate": 1.476755128929042e-05,
      "loss": 0.1865,
      "step": 13900
    },
    {
      "epoch": 2.62,
      "grad_norm": 13.203634262084961,
      "learning_rate": 1.4763786937699982e-05,
      "loss": 0.3328,
      "step": 13910
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.042203355580568314,
      "learning_rate": 1.4760022586109543e-05,
      "loss": 0.1652,
      "step": 13920
    },
    {
      "epoch": 2.62,
      "grad_norm": 2.041217803955078,
      "learning_rate": 1.4756258234519105e-05,
      "loss": 0.066,
      "step": 13930
    },
    {
      "epoch": 2.62,
      "grad_norm": 25.57231903076172,
      "learning_rate": 1.4752493882928666e-05,
      "loss": 0.296,
      "step": 13940
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.45437097549438477,
      "learning_rate": 1.4748729531338229e-05,
      "loss": 0.3383,
      "step": 13950
    },
    {
      "epoch": 2.63,
      "grad_norm": 10.684845924377441,
      "learning_rate": 1.474496517974779e-05,
      "loss": 0.1926,
      "step": 13960
    },
    {
      "epoch": 2.63,
      "grad_norm": 14.951252937316895,
      "learning_rate": 1.4741200828157352e-05,
      "loss": 0.4879,
      "step": 13970
    },
    {
      "epoch": 2.63,
      "grad_norm": 16.90325355529785,
      "learning_rate": 1.4737436476566914e-05,
      "loss": 0.2833,
      "step": 13980
    },
    {
      "epoch": 2.63,
      "grad_norm": 4.032130718231201,
      "learning_rate": 1.4733672124976475e-05,
      "loss": 0.3423,
      "step": 13990
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.22152701020240784,
      "learning_rate": 1.4729907773386036e-05,
      "loss": 0.217,
      "step": 14000
    },
    {
      "epoch": 2.64,
      "grad_norm": 2.3801581859588623,
      "learning_rate": 1.4726143421795598e-05,
      "loss": 0.105,
      "step": 14010
    },
    {
      "epoch": 2.64,
      "grad_norm": 41.877166748046875,
      "learning_rate": 1.472237907020516e-05,
      "loss": 0.4313,
      "step": 14020
    },
    {
      "epoch": 2.64,
      "grad_norm": 25.373172760009766,
      "learning_rate": 1.4718614718614719e-05,
      "loss": 0.2936,
      "step": 14030
    },
    {
      "epoch": 2.64,
      "grad_norm": 7.812513828277588,
      "learning_rate": 1.471485036702428e-05,
      "loss": 0.2657,
      "step": 14040
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.031295355409383774,
      "learning_rate": 1.4711086015433842e-05,
      "loss": 0.1596,
      "step": 14050
    },
    {
      "epoch": 2.65,
      "grad_norm": 14.050597190856934,
      "learning_rate": 1.4707321663843404e-05,
      "loss": 0.2711,
      "step": 14060
    },
    {
      "epoch": 2.65,
      "grad_norm": 9.684741973876953,
      "learning_rate": 1.4703557312252965e-05,
      "loss": 0.3143,
      "step": 14070
    },
    {
      "epoch": 2.65,
      "grad_norm": 22.19061279296875,
      "learning_rate": 1.4699792960662526e-05,
      "loss": 0.2676,
      "step": 14080
    },
    {
      "epoch": 2.65,
      "grad_norm": 8.933333396911621,
      "learning_rate": 1.4696028609072088e-05,
      "loss": 0.2201,
      "step": 14090
    },
    {
      "epoch": 2.65,
      "grad_norm": 4.093243598937988,
      "learning_rate": 1.469226425748165e-05,
      "loss": 0.2919,
      "step": 14100
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.05602738633751869,
      "learning_rate": 1.468849990589121e-05,
      "loss": 0.5019,
      "step": 14110
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.09239928424358368,
      "learning_rate": 1.4684735554300772e-05,
      "loss": 0.237,
      "step": 14120
    },
    {
      "epoch": 2.66,
      "grad_norm": 11.792847633361816,
      "learning_rate": 1.4680971202710335e-05,
      "loss": 0.2587,
      "step": 14130
    },
    {
      "epoch": 2.66,
      "grad_norm": 19.159547805786133,
      "learning_rate": 1.4677206851119897e-05,
      "loss": 0.3432,
      "step": 14140
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.07479263097047806,
      "learning_rate": 1.4673442499529458e-05,
      "loss": 0.3457,
      "step": 14150
    },
    {
      "epoch": 2.67,
      "grad_norm": 13.874930381774902,
      "learning_rate": 1.466967814793902e-05,
      "loss": 0.3241,
      "step": 14160
    },
    {
      "epoch": 2.67,
      "grad_norm": 19.694456100463867,
      "learning_rate": 1.4665913796348581e-05,
      "loss": 0.2123,
      "step": 14170
    },
    {
      "epoch": 2.67,
      "grad_norm": 9.415377616882324,
      "learning_rate": 1.4662149444758143e-05,
      "loss": 0.3423,
      "step": 14180
    },
    {
      "epoch": 2.67,
      "grad_norm": 18.648759841918945,
      "learning_rate": 1.4658385093167704e-05,
      "loss": 0.1644,
      "step": 14190
    },
    {
      "epoch": 2.67,
      "grad_norm": 13.744484901428223,
      "learning_rate": 1.4654620741577264e-05,
      "loss": 0.2473,
      "step": 14200
    },
    {
      "epoch": 2.67,
      "grad_norm": 7.4315314292907715,
      "learning_rate": 1.4650856389986825e-05,
      "loss": 0.4236,
      "step": 14210
    },
    {
      "epoch": 2.68,
      "grad_norm": 5.316078186035156,
      "learning_rate": 1.4647092038396387e-05,
      "loss": 0.27,
      "step": 14220
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.37941107153892517,
      "learning_rate": 1.4643327686805948e-05,
      "loss": 0.2978,
      "step": 14230
    },
    {
      "epoch": 2.68,
      "grad_norm": 1.990936517715454,
      "learning_rate": 1.463956333521551e-05,
      "loss": 0.2606,
      "step": 14240
    },
    {
      "epoch": 2.68,
      "grad_norm": 4.790708541870117,
      "learning_rate": 1.4635798983625071e-05,
      "loss": 0.3215,
      "step": 14250
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.7783608436584473,
      "learning_rate": 1.4632034632034633e-05,
      "loss": 0.2051,
      "step": 14260
    },
    {
      "epoch": 2.69,
      "grad_norm": 3.5843749046325684,
      "learning_rate": 1.4628270280444194e-05,
      "loss": 0.1648,
      "step": 14270
    },
    {
      "epoch": 2.69,
      "grad_norm": 1.8680593967437744,
      "learning_rate": 1.4624505928853756e-05,
      "loss": 0.2276,
      "step": 14280
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.3123987913131714,
      "learning_rate": 1.4620741577263317e-05,
      "loss": 0.2943,
      "step": 14290
    },
    {
      "epoch": 2.69,
      "grad_norm": 4.276027202606201,
      "learning_rate": 1.4616977225672879e-05,
      "loss": 0.2315,
      "step": 14300
    },
    {
      "epoch": 2.69,
      "grad_norm": 3.1547861099243164,
      "learning_rate": 1.461321287408244e-05,
      "loss": 0.1279,
      "step": 14310
    },
    {
      "epoch": 2.7,
      "grad_norm": 8.543717384338379,
      "learning_rate": 1.4609448522492003e-05,
      "loss": 0.4393,
      "step": 14320
    },
    {
      "epoch": 2.7,
      "grad_norm": 4.619165420532227,
      "learning_rate": 1.4605684170901565e-05,
      "loss": 0.5069,
      "step": 14330
    },
    {
      "epoch": 2.7,
      "grad_norm": 15.962494850158691,
      "learning_rate": 1.4601919819311126e-05,
      "loss": 0.188,
      "step": 14340
    },
    {
      "epoch": 2.7,
      "grad_norm": 20.17991065979004,
      "learning_rate": 1.4598155467720688e-05,
      "loss": 0.3091,
      "step": 14350
    },
    {
      "epoch": 2.7,
      "grad_norm": 15.477872848510742,
      "learning_rate": 1.4594391116130249e-05,
      "loss": 0.3861,
      "step": 14360
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.1446465700864792,
      "learning_rate": 1.459062676453981e-05,
      "loss": 0.1801,
      "step": 14370
    },
    {
      "epoch": 2.71,
      "grad_norm": 14.846261024475098,
      "learning_rate": 1.458686241294937e-05,
      "loss": 0.267,
      "step": 14380
    },
    {
      "epoch": 2.71,
      "grad_norm": 17.88726043701172,
      "learning_rate": 1.4583098061358932e-05,
      "loss": 0.191,
      "step": 14390
    },
    {
      "epoch": 2.71,
      "grad_norm": 4.974720478057861,
      "learning_rate": 1.4579333709768493e-05,
      "loss": 0.2477,
      "step": 14400
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.33594509959220886,
      "learning_rate": 1.4575569358178055e-05,
      "loss": 0.2026,
      "step": 14410
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.27815142273902893,
      "learning_rate": 1.4571805006587616e-05,
      "loss": 0.1438,
      "step": 14420
    },
    {
      "epoch": 2.72,
      "grad_norm": 1.5972073078155518,
      "learning_rate": 1.4568040654997178e-05,
      "loss": 0.1744,
      "step": 14430
    },
    {
      "epoch": 2.72,
      "grad_norm": 16.22450065612793,
      "learning_rate": 1.4564276303406739e-05,
      "loss": 0.311,
      "step": 14440
    },
    {
      "epoch": 2.72,
      "grad_norm": 14.905293464660645,
      "learning_rate": 1.45605119518163e-05,
      "loss": 0.2125,
      "step": 14450
    },
    {
      "epoch": 2.72,
      "grad_norm": 7.292027950286865,
      "learning_rate": 1.4556747600225862e-05,
      "loss": 0.2593,
      "step": 14460
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.3765714764595032,
      "learning_rate": 1.4552983248635423e-05,
      "loss": 0.0721,
      "step": 14470
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.042076244950294495,
      "learning_rate": 1.4549218897044985e-05,
      "loss": 0.316,
      "step": 14480
    },
    {
      "epoch": 2.73,
      "grad_norm": 11.66748046875,
      "learning_rate": 1.4545454545454546e-05,
      "loss": 0.1896,
      "step": 14490
    },
    {
      "epoch": 2.73,
      "grad_norm": 23.502302169799805,
      "learning_rate": 1.4541690193864108e-05,
      "loss": 0.4005,
      "step": 14500
    },
    {
      "epoch": 2.73,
      "grad_norm": 5.241673469543457,
      "learning_rate": 1.4537925842273671e-05,
      "loss": 0.1638,
      "step": 14510
    },
    {
      "epoch": 2.73,
      "grad_norm": 2.6241190433502197,
      "learning_rate": 1.4534161490683232e-05,
      "loss": 0.2537,
      "step": 14520
    },
    {
      "epoch": 2.73,
      "grad_norm": 35.8802375793457,
      "learning_rate": 1.4530397139092794e-05,
      "loss": 0.4925,
      "step": 14530
    },
    {
      "epoch": 2.74,
      "grad_norm": 13.407919883728027,
      "learning_rate": 1.4526632787502355e-05,
      "loss": 0.319,
      "step": 14540
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.03427611663937569,
      "learning_rate": 1.4522868435911915e-05,
      "loss": 0.2774,
      "step": 14550
    },
    {
      "epoch": 2.74,
      "grad_norm": 7.008240699768066,
      "learning_rate": 1.4519104084321477e-05,
      "loss": 0.2269,
      "step": 14560
    },
    {
      "epoch": 2.74,
      "grad_norm": 29.55392074584961,
      "learning_rate": 1.4515339732731038e-05,
      "loss": 0.2999,
      "step": 14570
    },
    {
      "epoch": 2.74,
      "grad_norm": 9.6903076171875,
      "learning_rate": 1.45115753811406e-05,
      "loss": 0.2756,
      "step": 14580
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.038982655853033066,
      "learning_rate": 1.4507811029550161e-05,
      "loss": 0.2076,
      "step": 14590
    },
    {
      "epoch": 2.75,
      "grad_norm": 11.586057662963867,
      "learning_rate": 1.4504046677959722e-05,
      "loss": 0.2146,
      "step": 14600
    },
    {
      "epoch": 2.75,
      "grad_norm": 6.1121416091918945,
      "learning_rate": 1.4500282326369284e-05,
      "loss": 0.3332,
      "step": 14610
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.08903893828392029,
      "learning_rate": 1.4496517974778845e-05,
      "loss": 0.4252,
      "step": 14620
    },
    {
      "epoch": 2.75,
      "grad_norm": 8.873623847961426,
      "learning_rate": 1.4492753623188407e-05,
      "loss": 0.2117,
      "step": 14630
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.05344058573246002,
      "learning_rate": 1.4488989271597968e-05,
      "loss": 0.5194,
      "step": 14640
    },
    {
      "epoch": 2.76,
      "grad_norm": 5.713958263397217,
      "learning_rate": 1.448522492000753e-05,
      "loss": 0.067,
      "step": 14650
    },
    {
      "epoch": 2.76,
      "grad_norm": 26.4097843170166,
      "learning_rate": 1.4481460568417091e-05,
      "loss": 0.3602,
      "step": 14660
    },
    {
      "epoch": 2.76,
      "grad_norm": 20.79299545288086,
      "learning_rate": 1.4477696216826653e-05,
      "loss": 0.3055,
      "step": 14670
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.018442951142787933,
      "learning_rate": 1.4473931865236214e-05,
      "loss": 0.2437,
      "step": 14680
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.019529465585947037,
      "learning_rate": 1.4470167513645777e-05,
      "loss": 0.3347,
      "step": 14690
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.10015013813972473,
      "learning_rate": 1.4466403162055339e-05,
      "loss": 0.1687,
      "step": 14700
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.1515250205993652,
      "learning_rate": 1.44626388104649e-05,
      "loss": 0.1558,
      "step": 14710
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.013014870695769787,
      "learning_rate": 1.4458874458874458e-05,
      "loss": 0.2286,
      "step": 14720
    },
    {
      "epoch": 2.77,
      "grad_norm": 4.731417655944824,
      "learning_rate": 1.445511010728402e-05,
      "loss": 0.1119,
      "step": 14730
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.7149228453636169,
      "learning_rate": 1.4451345755693583e-05,
      "loss": 0.2606,
      "step": 14740
    },
    {
      "epoch": 2.78,
      "grad_norm": 4.5265045166015625,
      "learning_rate": 1.4447581404103144e-05,
      "loss": 0.4256,
      "step": 14750
    },
    {
      "epoch": 2.78,
      "grad_norm": 4.416591644287109,
      "learning_rate": 1.4443817052512706e-05,
      "loss": 0.3112,
      "step": 14760
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.0500420406460762,
      "learning_rate": 1.4440052700922267e-05,
      "loss": 0.2338,
      "step": 14770
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.03410278633236885,
      "learning_rate": 1.4436288349331829e-05,
      "loss": 0.2696,
      "step": 14780
    },
    {
      "epoch": 2.78,
      "grad_norm": 23.082719802856445,
      "learning_rate": 1.443252399774139e-05,
      "loss": 0.1625,
      "step": 14790
    },
    {
      "epoch": 2.79,
      "grad_norm": 14.70162296295166,
      "learning_rate": 1.4428759646150952e-05,
      "loss": 0.3814,
      "step": 14800
    },
    {
      "epoch": 2.79,
      "grad_norm": 11.135512351989746,
      "learning_rate": 1.4424995294560513e-05,
      "loss": 0.3143,
      "step": 14810
    },
    {
      "epoch": 2.79,
      "grad_norm": 15.387861251831055,
      "learning_rate": 1.4421230942970075e-05,
      "loss": 0.3579,
      "step": 14820
    },
    {
      "epoch": 2.79,
      "grad_norm": 3.7702314853668213,
      "learning_rate": 1.4417466591379636e-05,
      "loss": 0.3258,
      "step": 14830
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.06137000769376755,
      "learning_rate": 1.4413702239789197e-05,
      "loss": 0.2284,
      "step": 14840
    },
    {
      "epoch": 2.8,
      "grad_norm": 14.064306259155273,
      "learning_rate": 1.4409937888198759e-05,
      "loss": 0.1828,
      "step": 14850
    },
    {
      "epoch": 2.8,
      "grad_norm": 11.78044605255127,
      "learning_rate": 1.440617353660832e-05,
      "loss": 0.3042,
      "step": 14860
    },
    {
      "epoch": 2.8,
      "grad_norm": 5.5149245262146,
      "learning_rate": 1.4402409185017882e-05,
      "loss": 0.2738,
      "step": 14870
    },
    {
      "epoch": 2.8,
      "grad_norm": 7.215740203857422,
      "learning_rate": 1.4398644833427445e-05,
      "loss": 0.1571,
      "step": 14880
    },
    {
      "epoch": 2.8,
      "grad_norm": 11.280044555664062,
      "learning_rate": 1.4394880481837006e-05,
      "loss": 0.1821,
      "step": 14890
    },
    {
      "epoch": 2.8,
      "grad_norm": 7.479217052459717,
      "learning_rate": 1.4391116130246565e-05,
      "loss": 0.2666,
      "step": 14900
    },
    {
      "epoch": 2.81,
      "grad_norm": 2.4822704792022705,
      "learning_rate": 1.4387351778656126e-05,
      "loss": 0.2371,
      "step": 14910
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.1591014266014099,
      "learning_rate": 1.4383587427065689e-05,
      "loss": 0.1343,
      "step": 14920
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.017317062243819237,
      "learning_rate": 1.437982307547525e-05,
      "loss": 0.1844,
      "step": 14930
    },
    {
      "epoch": 2.81,
      "grad_norm": 7.22636604309082,
      "learning_rate": 1.4376058723884812e-05,
      "loss": 0.4701,
      "step": 14940
    },
    {
      "epoch": 2.81,
      "grad_norm": 8.931745529174805,
      "learning_rate": 1.4372294372294374e-05,
      "loss": 0.2363,
      "step": 14950
    },
    {
      "epoch": 2.82,
      "grad_norm": 5.314764499664307,
      "learning_rate": 1.4368530020703935e-05,
      "loss": 0.2057,
      "step": 14960
    },
    {
      "epoch": 2.82,
      "grad_norm": 17.26463508605957,
      "learning_rate": 1.4364765669113496e-05,
      "loss": 0.199,
      "step": 14970
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.013541669584810734,
      "learning_rate": 1.4361001317523058e-05,
      "loss": 0.2564,
      "step": 14980
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.026105977594852448,
      "learning_rate": 1.435723696593262e-05,
      "loss": 0.1071,
      "step": 14990
    },
    {
      "epoch": 2.82,
      "grad_norm": 7.403133392333984,
      "learning_rate": 1.435347261434218e-05,
      "loss": 0.3323,
      "step": 15000
    },
    {
      "epoch": 2.83,
      "grad_norm": 5.461624622344971,
      "learning_rate": 1.4349708262751742e-05,
      "loss": 0.2752,
      "step": 15010
    },
    {
      "epoch": 2.83,
      "grad_norm": 22.498125076293945,
      "learning_rate": 1.4345943911161304e-05,
      "loss": 0.4028,
      "step": 15020
    },
    {
      "epoch": 2.83,
      "grad_norm": 8.872709274291992,
      "learning_rate": 1.4342179559570865e-05,
      "loss": 0.3902,
      "step": 15030
    },
    {
      "epoch": 2.83,
      "grad_norm": 11.680445671081543,
      "learning_rate": 1.4338415207980427e-05,
      "loss": 0.3443,
      "step": 15040
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.0837346762418747,
      "learning_rate": 1.4334650856389988e-05,
      "loss": 0.0724,
      "step": 15050
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.03785625472664833,
      "learning_rate": 1.4330886504799551e-05,
      "loss": 0.0709,
      "step": 15060
    },
    {
      "epoch": 2.84,
      "grad_norm": 2.2610859870910645,
      "learning_rate": 1.432712215320911e-05,
      "loss": 0.2874,
      "step": 15070
    },
    {
      "epoch": 2.84,
      "grad_norm": 23.059499740600586,
      "learning_rate": 1.432335780161867e-05,
      "loss": 0.2303,
      "step": 15080
    },
    {
      "epoch": 2.84,
      "grad_norm": 2.434633493423462,
      "learning_rate": 1.4319593450028232e-05,
      "loss": 0.2706,
      "step": 15090
    },
    {
      "epoch": 2.84,
      "grad_norm": 5.018374919891357,
      "learning_rate": 1.4315829098437794e-05,
      "loss": 0.345,
      "step": 15100
    },
    {
      "epoch": 2.84,
      "grad_norm": 23.358627319335938,
      "learning_rate": 1.4312064746847357e-05,
      "loss": 0.3058,
      "step": 15110
    },
    {
      "epoch": 2.85,
      "grad_norm": 46.489986419677734,
      "learning_rate": 1.4308300395256918e-05,
      "loss": 0.26,
      "step": 15120
    },
    {
      "epoch": 2.85,
      "grad_norm": 19.33156394958496,
      "learning_rate": 1.430453604366648e-05,
      "loss": 0.3267,
      "step": 15130
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.784745454788208,
      "learning_rate": 1.4300771692076041e-05,
      "loss": 0.0729,
      "step": 15140
    },
    {
      "epoch": 2.85,
      "grad_norm": 1.5927478075027466,
      "learning_rate": 1.4297007340485603e-05,
      "loss": 0.1555,
      "step": 15150
    },
    {
      "epoch": 2.85,
      "grad_norm": 4.469014644622803,
      "learning_rate": 1.4293242988895164e-05,
      "loss": 0.1859,
      "step": 15160
    },
    {
      "epoch": 2.86,
      "grad_norm": 17.845458984375,
      "learning_rate": 1.4289478637304726e-05,
      "loss": 0.4426,
      "step": 15170
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.030550677329301834,
      "learning_rate": 1.4285714285714287e-05,
      "loss": 0.1846,
      "step": 15180
    },
    {
      "epoch": 2.86,
      "grad_norm": 8.859922409057617,
      "learning_rate": 1.4281949934123849e-05,
      "loss": 0.1728,
      "step": 15190
    },
    {
      "epoch": 2.86,
      "grad_norm": 3.0599591732025146,
      "learning_rate": 1.427818558253341e-05,
      "loss": 0.2201,
      "step": 15200
    },
    {
      "epoch": 2.86,
      "grad_norm": 13.119793891906738,
      "learning_rate": 1.4274421230942972e-05,
      "loss": 0.3582,
      "step": 15210
    },
    {
      "epoch": 2.86,
      "grad_norm": 14.222670555114746,
      "learning_rate": 1.4270656879352533e-05,
      "loss": 0.1647,
      "step": 15220
    },
    {
      "epoch": 2.87,
      "grad_norm": 17.192983627319336,
      "learning_rate": 1.4266892527762094e-05,
      "loss": 0.2499,
      "step": 15230
    },
    {
      "epoch": 2.87,
      "grad_norm": 31.67969512939453,
      "learning_rate": 1.4263128176171656e-05,
      "loss": 0.2021,
      "step": 15240
    },
    {
      "epoch": 2.87,
      "grad_norm": 16.927156448364258,
      "learning_rate": 1.4259363824581216e-05,
      "loss": 0.2052,
      "step": 15250
    },
    {
      "epoch": 2.87,
      "grad_norm": 21.820667266845703,
      "learning_rate": 1.4255599472990777e-05,
      "loss": 0.2315,
      "step": 15260
    },
    {
      "epoch": 2.87,
      "grad_norm": 8.232240676879883,
      "learning_rate": 1.4251835121400339e-05,
      "loss": 0.1971,
      "step": 15270
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.013723282143473625,
      "learning_rate": 1.42480707698099e-05,
      "loss": 0.2142,
      "step": 15280
    },
    {
      "epoch": 2.88,
      "grad_norm": 7.797157287597656,
      "learning_rate": 1.4244306418219463e-05,
      "loss": 0.3666,
      "step": 15290
    },
    {
      "epoch": 2.88,
      "grad_norm": 17.3726806640625,
      "learning_rate": 1.4240542066629025e-05,
      "loss": 0.1602,
      "step": 15300
    },
    {
      "epoch": 2.88,
      "grad_norm": 19.963117599487305,
      "learning_rate": 1.4236777715038586e-05,
      "loss": 0.3667,
      "step": 15310
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.2100875824689865,
      "learning_rate": 1.4233013363448148e-05,
      "loss": 0.2943,
      "step": 15320
    },
    {
      "epoch": 2.89,
      "grad_norm": 15.624946594238281,
      "learning_rate": 1.4229249011857709e-05,
      "loss": 0.2846,
      "step": 15330
    },
    {
      "epoch": 2.89,
      "grad_norm": 8.909420013427734,
      "learning_rate": 1.422548466026727e-05,
      "loss": 0.1652,
      "step": 15340
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.3456437885761261,
      "learning_rate": 1.4221720308676832e-05,
      "loss": 0.2664,
      "step": 15350
    },
    {
      "epoch": 2.89,
      "grad_norm": 6.245460510253906,
      "learning_rate": 1.4217955957086393e-05,
      "loss": 0.3548,
      "step": 15360
    },
    {
      "epoch": 2.89,
      "grad_norm": 15.549107551574707,
      "learning_rate": 1.4214191605495955e-05,
      "loss": 0.1583,
      "step": 15370
    },
    {
      "epoch": 2.89,
      "grad_norm": 6.741213798522949,
      "learning_rate": 1.4210427253905516e-05,
      "loss": 0.277,
      "step": 15380
    },
    {
      "epoch": 2.9,
      "grad_norm": 22.206668853759766,
      "learning_rate": 1.4206662902315078e-05,
      "loss": 0.2815,
      "step": 15390
    },
    {
      "epoch": 2.9,
      "grad_norm": 7.749658107757568,
      "learning_rate": 1.420289855072464e-05,
      "loss": 0.4437,
      "step": 15400
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.017829660326242447,
      "learning_rate": 1.41991341991342e-05,
      "loss": 0.1785,
      "step": 15410
    },
    {
      "epoch": 2.9,
      "grad_norm": 4.420356750488281,
      "learning_rate": 1.419536984754376e-05,
      "loss": 0.2917,
      "step": 15420
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.024433838203549385,
      "learning_rate": 1.4191605495953322e-05,
      "loss": 0.1737,
      "step": 15430
    },
    {
      "epoch": 2.91,
      "grad_norm": 8.703853607177734,
      "learning_rate": 1.4187841144362883e-05,
      "loss": 0.1617,
      "step": 15440
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.010950411669909954,
      "learning_rate": 1.4184076792772445e-05,
      "loss": 0.2292,
      "step": 15450
    },
    {
      "epoch": 2.91,
      "grad_norm": 1.8055084943771362,
      "learning_rate": 1.4180312441182006e-05,
      "loss": 0.3715,
      "step": 15460
    },
    {
      "epoch": 2.91,
      "grad_norm": 3.8196189403533936,
      "learning_rate": 1.4176548089591568e-05,
      "loss": 0.2765,
      "step": 15470
    },
    {
      "epoch": 2.91,
      "grad_norm": 37.889244079589844,
      "learning_rate": 1.4172783738001131e-05,
      "loss": 0.5024,
      "step": 15480
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.860675036907196,
      "learning_rate": 1.4169019386410692e-05,
      "loss": 0.3371,
      "step": 15490
    },
    {
      "epoch": 2.92,
      "grad_norm": 7.009801864624023,
      "learning_rate": 1.4165255034820254e-05,
      "loss": 0.2087,
      "step": 15500
    },
    {
      "epoch": 2.92,
      "grad_norm": 21.133445739746094,
      "learning_rate": 1.4161490683229815e-05,
      "loss": 0.4334,
      "step": 15510
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.5159934163093567,
      "learning_rate": 1.4157726331639377e-05,
      "loss": 0.314,
      "step": 15520
    },
    {
      "epoch": 2.92,
      "grad_norm": 19.11453628540039,
      "learning_rate": 1.4153961980048938e-05,
      "loss": 0.3607,
      "step": 15530
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.841351270675659,
      "learning_rate": 1.41501976284585e-05,
      "loss": 0.2422,
      "step": 15540
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.07019159197807312,
      "learning_rate": 1.4146433276868061e-05,
      "loss": 0.169,
      "step": 15550
    },
    {
      "epoch": 2.93,
      "grad_norm": 2.239062547683716,
      "learning_rate": 1.4142668925277623e-05,
      "loss": 0.0834,
      "step": 15560
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.10594528913497925,
      "learning_rate": 1.4138904573687184e-05,
      "loss": 0.378,
      "step": 15570
    },
    {
      "epoch": 2.93,
      "grad_norm": 9.946199417114258,
      "learning_rate": 1.4135140222096746e-05,
      "loss": 0.4877,
      "step": 15580
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.01927807554602623,
      "learning_rate": 1.4131375870506307e-05,
      "loss": 0.2749,
      "step": 15590
    },
    {
      "epoch": 2.94,
      "grad_norm": 9.410808563232422,
      "learning_rate": 1.4127611518915867e-05,
      "loss": 0.2195,
      "step": 15600
    },
    {
      "epoch": 2.94,
      "grad_norm": 2.9816925525665283,
      "learning_rate": 1.4123847167325428e-05,
      "loss": 0.4345,
      "step": 15610
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.8589879274368286,
      "learning_rate": 1.412008281573499e-05,
      "loss": 0.4044,
      "step": 15620
    },
    {
      "epoch": 2.94,
      "grad_norm": 10.39150333404541,
      "learning_rate": 1.4116318464144551e-05,
      "loss": 0.2055,
      "step": 15630
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.028857693076133728,
      "learning_rate": 1.4112554112554113e-05,
      "loss": 0.175,
      "step": 15640
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.0521932952105999,
      "learning_rate": 1.4108789760963674e-05,
      "loss": 0.2616,
      "step": 15650
    },
    {
      "epoch": 2.95,
      "grad_norm": 11.82694149017334,
      "learning_rate": 1.4105025409373236e-05,
      "loss": 0.3231,
      "step": 15660
    },
    {
      "epoch": 2.95,
      "grad_norm": 2.178129196166992,
      "learning_rate": 1.4101261057782799e-05,
      "loss": 0.2175,
      "step": 15670
    },
    {
      "epoch": 2.95,
      "grad_norm": 16.188840866088867,
      "learning_rate": 1.409749670619236e-05,
      "loss": 0.1606,
      "step": 15680
    },
    {
      "epoch": 2.95,
      "grad_norm": 7.363923072814941,
      "learning_rate": 1.4093732354601922e-05,
      "loss": 0.1934,
      "step": 15690
    },
    {
      "epoch": 2.96,
      "grad_norm": 14.87932300567627,
      "learning_rate": 1.4089968003011483e-05,
      "loss": 0.264,
      "step": 15700
    },
    {
      "epoch": 2.96,
      "grad_norm": 10.435835838317871,
      "learning_rate": 1.4086203651421045e-05,
      "loss": 0.2604,
      "step": 15710
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.04438408836722374,
      "learning_rate": 1.4082439299830606e-05,
      "loss": 0.133,
      "step": 15720
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.03822668269276619,
      "learning_rate": 1.4078674948240167e-05,
      "loss": 0.2259,
      "step": 15730
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.057425376027822495,
      "learning_rate": 1.4074910596649729e-05,
      "loss": 0.3544,
      "step": 15740
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.015939144417643547,
      "learning_rate": 1.407114624505929e-05,
      "loss": 0.3133,
      "step": 15750
    },
    {
      "epoch": 2.97,
      "grad_norm": 15.669068336486816,
      "learning_rate": 1.4067381893468852e-05,
      "loss": 0.2068,
      "step": 15760
    },
    {
      "epoch": 2.97,
      "grad_norm": 7.269059658050537,
      "learning_rate": 1.4063617541878412e-05,
      "loss": 0.3437,
      "step": 15770
    },
    {
      "epoch": 2.97,
      "grad_norm": 18.469810485839844,
      "learning_rate": 1.4059853190287973e-05,
      "loss": 0.1998,
      "step": 15780
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.4941658675670624,
      "learning_rate": 1.4056088838697535e-05,
      "loss": 0.2695,
      "step": 15790
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.02360748127102852,
      "learning_rate": 1.4052324487107096e-05,
      "loss": 0.1455,
      "step": 15800
    },
    {
      "epoch": 2.98,
      "grad_norm": 40.085227966308594,
      "learning_rate": 1.4048560135516657e-05,
      "loss": 0.1921,
      "step": 15810
    },
    {
      "epoch": 2.98,
      "grad_norm": 30.855789184570312,
      "learning_rate": 1.4044795783926219e-05,
      "loss": 0.2993,
      "step": 15820
    },
    {
      "epoch": 2.98,
      "grad_norm": 16.580869674682617,
      "learning_rate": 1.404103143233578e-05,
      "loss": 0.2446,
      "step": 15830
    },
    {
      "epoch": 2.98,
      "grad_norm": 7.5667219161987305,
      "learning_rate": 1.4037267080745342e-05,
      "loss": 0.2085,
      "step": 15840
    },
    {
      "epoch": 2.98,
      "grad_norm": 9.053364753723145,
      "learning_rate": 1.4033502729154905e-05,
      "loss": 0.3371,
      "step": 15850
    },
    {
      "epoch": 2.99,
      "grad_norm": 5.675618648529053,
      "learning_rate": 1.4029738377564466e-05,
      "loss": 0.1351,
      "step": 15860
    },
    {
      "epoch": 2.99,
      "grad_norm": 13.122978210449219,
      "learning_rate": 1.4025974025974028e-05,
      "loss": 0.2888,
      "step": 15870
    },
    {
      "epoch": 2.99,
      "grad_norm": 9.505881309509277,
      "learning_rate": 1.402220967438359e-05,
      "loss": 0.3771,
      "step": 15880
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.4093904495239258,
      "learning_rate": 1.401844532279315e-05,
      "loss": 0.2765,
      "step": 15890
    },
    {
      "epoch": 2.99,
      "grad_norm": 1.1487828493118286,
      "learning_rate": 1.4014680971202712e-05,
      "loss": 0.0946,
      "step": 15900
    },
    {
      "epoch": 2.99,
      "grad_norm": 23.867189407348633,
      "learning_rate": 1.4010916619612274e-05,
      "loss": 0.1645,
      "step": 15910
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.009584042243659496,
      "learning_rate": 1.4007152268021835e-05,
      "loss": 0.0572,
      "step": 15920
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.06809663772583,
      "learning_rate": 1.4003387916431397e-05,
      "loss": 0.1848,
      "step": 15930
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9873333333333333,
      "eval_loss": 0.0503651425242424,
      "eval_runtime": 50.738,
      "eval_samples_per_second": 147.818,
      "eval_steps_per_second": 18.487,
      "step": 15939
    },
    {
      "epoch": 3.0,
      "grad_norm": 14.04341983795166,
      "learning_rate": 1.3999623564840956e-05,
      "loss": 0.297,
      "step": 15940
    },
    {
      "epoch": 3.0,
      "grad_norm": 11.245918273925781,
      "learning_rate": 1.3995859213250518e-05,
      "loss": 0.122,
      "step": 15950
    },
    {
      "epoch": 3.0,
      "grad_norm": 8.632491111755371,
      "learning_rate": 1.399209486166008e-05,
      "loss": 0.0628,
      "step": 15960
    },
    {
      "epoch": 3.01,
      "grad_norm": 49.83925247192383,
      "learning_rate": 1.398833051006964e-05,
      "loss": 0.3203,
      "step": 15970
    },
    {
      "epoch": 3.01,
      "grad_norm": 8.499287605285645,
      "learning_rate": 1.3984566158479202e-05,
      "loss": 0.24,
      "step": 15980
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.10720015317201614,
      "learning_rate": 1.3980801806888764e-05,
      "loss": 0.4293,
      "step": 15990
    },
    {
      "epoch": 3.01,
      "grad_norm": 0.08821815252304077,
      "learning_rate": 1.3977037455298325e-05,
      "loss": 0.1175,
      "step": 16000
    },
    {
      "epoch": 3.01,
      "grad_norm": 19.80327796936035,
      "learning_rate": 1.3973273103707887e-05,
      "loss": 0.1843,
      "step": 16010
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.012754634954035282,
      "learning_rate": 1.3969508752117448e-05,
      "loss": 0.1514,
      "step": 16020
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.060094982385635376,
      "learning_rate": 1.396574440052701e-05,
      "loss": 0.2008,
      "step": 16030
    },
    {
      "epoch": 3.02,
      "grad_norm": 21.31043815612793,
      "learning_rate": 1.3961980048936573e-05,
      "loss": 0.3815,
      "step": 16040
    },
    {
      "epoch": 3.02,
      "grad_norm": 3.531586170196533,
      "learning_rate": 1.3958215697346134e-05,
      "loss": 0.0922,
      "step": 16050
    },
    {
      "epoch": 3.02,
      "grad_norm": 0.012437762692570686,
      "learning_rate": 1.3954451345755696e-05,
      "loss": 0.2058,
      "step": 16060
    },
    {
      "epoch": 3.02,
      "grad_norm": 18.43088722229004,
      "learning_rate": 1.3950686994165257e-05,
      "loss": 0.3206,
      "step": 16070
    },
    {
      "epoch": 3.03,
      "grad_norm": 7.141884803771973,
      "learning_rate": 1.3946922642574819e-05,
      "loss": 0.2616,
      "step": 16080
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.02548779919743538,
      "learning_rate": 1.394315829098438e-05,
      "loss": 0.2484,
      "step": 16090
    },
    {
      "epoch": 3.03,
      "grad_norm": 0.12823136150836945,
      "learning_rate": 1.3939393939393942e-05,
      "loss": 0.1434,
      "step": 16100
    },
    {
      "epoch": 3.03,
      "grad_norm": 17.013288497924805,
      "learning_rate": 1.3935629587803503e-05,
      "loss": 0.1813,
      "step": 16110
    },
    {
      "epoch": 3.03,
      "grad_norm": 1.77372407913208,
      "learning_rate": 1.3931865236213063e-05,
      "loss": 0.2091,
      "step": 16120
    },
    {
      "epoch": 3.04,
      "grad_norm": 27.1054744720459,
      "learning_rate": 1.3928100884622624e-05,
      "loss": 0.2163,
      "step": 16130
    },
    {
      "epoch": 3.04,
      "grad_norm": 9.572776794433594,
      "learning_rate": 1.3924336533032186e-05,
      "loss": 0.2059,
      "step": 16140
    },
    {
      "epoch": 3.04,
      "grad_norm": 20.8569278717041,
      "learning_rate": 1.3920572181441747e-05,
      "loss": 0.2118,
      "step": 16150
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.21719448268413544,
      "learning_rate": 1.3916807829851309e-05,
      "loss": 0.3636,
      "step": 16160
    },
    {
      "epoch": 3.04,
      "grad_norm": 0.471840500831604,
      "learning_rate": 1.391304347826087e-05,
      "loss": 0.3708,
      "step": 16170
    },
    {
      "epoch": 3.05,
      "grad_norm": 4.994062900543213,
      "learning_rate": 1.3909279126670432e-05,
      "loss": 0.2909,
      "step": 16180
    },
    {
      "epoch": 3.05,
      "grad_norm": 2.6290016174316406,
      "learning_rate": 1.3905514775079993e-05,
      "loss": 0.1041,
      "step": 16190
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.28102636337280273,
      "learning_rate": 1.3901750423489554e-05,
      "loss": 0.1655,
      "step": 16200
    },
    {
      "epoch": 3.05,
      "grad_norm": 12.94188117980957,
      "learning_rate": 1.3897986071899116e-05,
      "loss": 0.4226,
      "step": 16210
    },
    {
      "epoch": 3.05,
      "grad_norm": 7.222362995147705,
      "learning_rate": 1.3894221720308679e-05,
      "loss": 0.1502,
      "step": 16220
    },
    {
      "epoch": 3.05,
      "grad_norm": 11.334131240844727,
      "learning_rate": 1.389045736871824e-05,
      "loss": 0.3588,
      "step": 16230
    },
    {
      "epoch": 3.06,
      "grad_norm": 0.09283274412155151,
      "learning_rate": 1.3886693017127802e-05,
      "loss": 0.178,
      "step": 16240
    },
    {
      "epoch": 3.06,
      "grad_norm": 5.626291275024414,
      "learning_rate": 1.3882928665537363e-05,
      "loss": 0.3655,
      "step": 16250
    },
    {
      "epoch": 3.06,
      "grad_norm": 18.000261306762695,
      "learning_rate": 1.3879164313946925e-05,
      "loss": 0.1933,
      "step": 16260
    },
    {
      "epoch": 3.06,
      "grad_norm": 11.50313663482666,
      "learning_rate": 1.3875399962356486e-05,
      "loss": 0.3914,
      "step": 16270
    },
    {
      "epoch": 3.06,
      "grad_norm": 7.677784442901611,
      "learning_rate": 1.3871635610766048e-05,
      "loss": 0.0828,
      "step": 16280
    },
    {
      "epoch": 3.07,
      "grad_norm": 7.3539042472839355,
      "learning_rate": 1.3867871259175608e-05,
      "loss": 0.208,
      "step": 16290
    },
    {
      "epoch": 3.07,
      "grad_norm": 0.04681446775794029,
      "learning_rate": 1.3864106907585169e-05,
      "loss": 0.4903,
      "step": 16300
    },
    {
      "epoch": 3.07,
      "grad_norm": 20.07516098022461,
      "learning_rate": 1.386034255599473e-05,
      "loss": 0.2946,
      "step": 16310
    },
    {
      "epoch": 3.07,
      "grad_norm": 65.470458984375,
      "learning_rate": 1.3856578204404292e-05,
      "loss": 0.2727,
      "step": 16320
    },
    {
      "epoch": 3.07,
      "grad_norm": 5.789030075073242,
      "learning_rate": 1.3852813852813853e-05,
      "loss": 0.187,
      "step": 16330
    },
    {
      "epoch": 3.08,
      "grad_norm": 18.854379653930664,
      "learning_rate": 1.3849049501223415e-05,
      "loss": 0.1926,
      "step": 16340
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.7280802726745605,
      "learning_rate": 1.3845285149632976e-05,
      "loss": 0.0938,
      "step": 16350
    },
    {
      "epoch": 3.08,
      "grad_norm": 12.581260681152344,
      "learning_rate": 1.3841520798042538e-05,
      "loss": 0.3363,
      "step": 16360
    },
    {
      "epoch": 3.08,
      "grad_norm": 4.9129204750061035,
      "learning_rate": 1.38377564464521e-05,
      "loss": 0.1627,
      "step": 16370
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.02393791452050209,
      "learning_rate": 1.383399209486166e-05,
      "loss": 0.03,
      "step": 16380
    },
    {
      "epoch": 3.08,
      "grad_norm": 1.9745910167694092,
      "learning_rate": 1.3830227743271222e-05,
      "loss": 0.1582,
      "step": 16390
    },
    {
      "epoch": 3.09,
      "grad_norm": 11.230238914489746,
      "learning_rate": 1.3826463391680784e-05,
      "loss": 0.5885,
      "step": 16400
    },
    {
      "epoch": 3.09,
      "grad_norm": 16.097415924072266,
      "learning_rate": 1.3822699040090347e-05,
      "loss": 0.1778,
      "step": 16410
    },
    {
      "epoch": 3.09,
      "grad_norm": 4.409106731414795,
      "learning_rate": 1.3818934688499908e-05,
      "loss": 0.2779,
      "step": 16420
    },
    {
      "epoch": 3.09,
      "grad_norm": 5.562557220458984,
      "learning_rate": 1.381517033690947e-05,
      "loss": 0.2883,
      "step": 16430
    },
    {
      "epoch": 3.09,
      "grad_norm": 0.07913478463888168,
      "learning_rate": 1.3811405985319031e-05,
      "loss": 0.2199,
      "step": 16440
    },
    {
      "epoch": 3.1,
      "grad_norm": 16.408103942871094,
      "learning_rate": 1.3807641633728593e-05,
      "loss": 0.3274,
      "step": 16450
    },
    {
      "epoch": 3.1,
      "grad_norm": 7.725156307220459,
      "learning_rate": 1.3803877282138154e-05,
      "loss": 0.2693,
      "step": 16460
    },
    {
      "epoch": 3.1,
      "grad_norm": 3.448357105255127,
      "learning_rate": 1.3800112930547714e-05,
      "loss": 0.2836,
      "step": 16470
    },
    {
      "epoch": 3.1,
      "grad_norm": 7.673715114593506,
      "learning_rate": 1.3796348578957275e-05,
      "loss": 0.1253,
      "step": 16480
    },
    {
      "epoch": 3.1,
      "grad_norm": 2.890608072280884,
      "learning_rate": 1.3792584227366837e-05,
      "loss": 0.2119,
      "step": 16490
    },
    {
      "epoch": 3.11,
      "grad_norm": 10.583338737487793,
      "learning_rate": 1.3788819875776398e-05,
      "loss": 0.3395,
      "step": 16500
    },
    {
      "epoch": 3.11,
      "grad_norm": 13.629029273986816,
      "learning_rate": 1.378505552418596e-05,
      "loss": 0.4848,
      "step": 16510
    },
    {
      "epoch": 3.11,
      "grad_norm": 22.8969669342041,
      "learning_rate": 1.3781291172595521e-05,
      "loss": 0.302,
      "step": 16520
    },
    {
      "epoch": 3.11,
      "grad_norm": 4.932634353637695,
      "learning_rate": 1.3777526821005083e-05,
      "loss": 0.2823,
      "step": 16530
    },
    {
      "epoch": 3.11,
      "grad_norm": 4.923015594482422,
      "learning_rate": 1.3773762469414644e-05,
      "loss": 0.1328,
      "step": 16540
    },
    {
      "epoch": 3.12,
      "grad_norm": 8.096443176269531,
      "learning_rate": 1.3769998117824206e-05,
      "loss": 0.2901,
      "step": 16550
    },
    {
      "epoch": 3.12,
      "grad_norm": 22.236814498901367,
      "learning_rate": 1.3766233766233767e-05,
      "loss": 0.1568,
      "step": 16560
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.8296091556549072,
      "learning_rate": 1.3762469414643328e-05,
      "loss": 0.1746,
      "step": 16570
    },
    {
      "epoch": 3.12,
      "grad_norm": 0.6501277685165405,
      "learning_rate": 1.375870506305289e-05,
      "loss": 0.5321,
      "step": 16580
    },
    {
      "epoch": 3.12,
      "grad_norm": 33.04877853393555,
      "learning_rate": 1.3754940711462453e-05,
      "loss": 0.1835,
      "step": 16590
    },
    {
      "epoch": 3.12,
      "grad_norm": 21.99195098876953,
      "learning_rate": 1.3751176359872015e-05,
      "loss": 0.2625,
      "step": 16600
    },
    {
      "epoch": 3.13,
      "grad_norm": 16.687665939331055,
      "learning_rate": 1.3747412008281576e-05,
      "loss": 0.3849,
      "step": 16610
    },
    {
      "epoch": 3.13,
      "grad_norm": 0.14279034733772278,
      "learning_rate": 1.3743647656691137e-05,
      "loss": 0.3512,
      "step": 16620
    },
    {
      "epoch": 3.13,
      "grad_norm": 13.945484161376953,
      "learning_rate": 1.3739883305100699e-05,
      "loss": 0.477,
      "step": 16630
    },
    {
      "epoch": 3.13,
      "grad_norm": 11.20386791229248,
      "learning_rate": 1.3736118953510259e-05,
      "loss": 0.2205,
      "step": 16640
    },
    {
      "epoch": 3.13,
      "grad_norm": 10.552443504333496,
      "learning_rate": 1.373235460191982e-05,
      "loss": 0.2555,
      "step": 16650
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.16616837680339813,
      "learning_rate": 1.3728590250329382e-05,
      "loss": 0.3188,
      "step": 16660
    },
    {
      "epoch": 3.14,
      "grad_norm": 0.2319163829088211,
      "learning_rate": 1.3724825898738943e-05,
      "loss": 0.1594,
      "step": 16670
    },
    {
      "epoch": 3.14,
      "grad_norm": 6.049283027648926,
      "learning_rate": 1.3721061547148505e-05,
      "loss": 0.4116,
      "step": 16680
    },
    {
      "epoch": 3.14,
      "grad_norm": 21.742549896240234,
      "learning_rate": 1.3717297195558066e-05,
      "loss": 0.3298,
      "step": 16690
    },
    {
      "epoch": 3.14,
      "grad_norm": 10.678056716918945,
      "learning_rate": 1.3713532843967627e-05,
      "loss": 0.0963,
      "step": 16700
    },
    {
      "epoch": 3.15,
      "grad_norm": 3.0524823665618896,
      "learning_rate": 1.3709768492377189e-05,
      "loss": 0.2381,
      "step": 16710
    },
    {
      "epoch": 3.15,
      "grad_norm": 0.017938483506441116,
      "learning_rate": 1.370600414078675e-05,
      "loss": 0.0856,
      "step": 16720
    },
    {
      "epoch": 3.15,
      "grad_norm": 5.312686443328857,
      "learning_rate": 1.3702239789196312e-05,
      "loss": 0.2305,
      "step": 16730
    },
    {
      "epoch": 3.15,
      "grad_norm": 5.520197868347168,
      "learning_rate": 1.3698475437605873e-05,
      "loss": 0.3135,
      "step": 16740
    },
    {
      "epoch": 3.15,
      "grad_norm": 13.966022491455078,
      "learning_rate": 1.3694711086015435e-05,
      "loss": 0.2356,
      "step": 16750
    },
    {
      "epoch": 3.15,
      "grad_norm": 4.6176629066467285,
      "learning_rate": 1.3690946734424996e-05,
      "loss": 0.2254,
      "step": 16760
    },
    {
      "epoch": 3.16,
      "grad_norm": 20.774662017822266,
      "learning_rate": 1.3687182382834558e-05,
      "loss": 0.2353,
      "step": 16770
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.03736928105354309,
      "learning_rate": 1.368341803124412e-05,
      "loss": 0.0658,
      "step": 16780
    },
    {
      "epoch": 3.16,
      "grad_norm": 4.4706878662109375,
      "learning_rate": 1.3679653679653682e-05,
      "loss": 0.1358,
      "step": 16790
    },
    {
      "epoch": 3.16,
      "grad_norm": 0.9585201144218445,
      "learning_rate": 1.3675889328063244e-05,
      "loss": 0.1599,
      "step": 16800
    },
    {
      "epoch": 3.16,
      "grad_norm": 5.0444464683532715,
      "learning_rate": 1.3672124976472802e-05,
      "loss": 0.1536,
      "step": 16810
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.0728171095252037,
      "learning_rate": 1.3668360624882365e-05,
      "loss": 0.1311,
      "step": 16820
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.013270686380565166,
      "learning_rate": 1.3664596273291926e-05,
      "loss": 0.2346,
      "step": 16830
    },
    {
      "epoch": 3.17,
      "grad_norm": 0.05075976625084877,
      "learning_rate": 1.3660831921701488e-05,
      "loss": 0.0872,
      "step": 16840
    },
    {
      "epoch": 3.17,
      "grad_norm": 14.674505233764648,
      "learning_rate": 1.365706757011105e-05,
      "loss": 0.3827,
      "step": 16850
    },
    {
      "epoch": 3.17,
      "grad_norm": 15.35838508605957,
      "learning_rate": 1.365330321852061e-05,
      "loss": 0.3517,
      "step": 16860
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.08562055230140686,
      "learning_rate": 1.3649538866930172e-05,
      "loss": 0.1433,
      "step": 16870
    },
    {
      "epoch": 3.18,
      "grad_norm": 15.240372657775879,
      "learning_rate": 1.3645774515339734e-05,
      "loss": 0.2443,
      "step": 16880
    },
    {
      "epoch": 3.18,
      "grad_norm": 2.0486180782318115,
      "learning_rate": 1.3642010163749295e-05,
      "loss": 0.2364,
      "step": 16890
    },
    {
      "epoch": 3.18,
      "grad_norm": 8.199923515319824,
      "learning_rate": 1.3638245812158857e-05,
      "loss": 0.4277,
      "step": 16900
    },
    {
      "epoch": 3.18,
      "grad_norm": 0.2164665311574936,
      "learning_rate": 1.3634481460568418e-05,
      "loss": 0.1623,
      "step": 16910
    },
    {
      "epoch": 3.18,
      "grad_norm": 13.832843780517578,
      "learning_rate": 1.363071710897798e-05,
      "loss": 0.1919,
      "step": 16920
    },
    {
      "epoch": 3.19,
      "grad_norm": 13.660701751708984,
      "learning_rate": 1.3626952757387541e-05,
      "loss": 0.2038,
      "step": 16930
    },
    {
      "epoch": 3.19,
      "grad_norm": 7.076765537261963,
      "learning_rate": 1.3623188405797103e-05,
      "loss": 0.2752,
      "step": 16940
    },
    {
      "epoch": 3.19,
      "grad_norm": 9.872081756591797,
      "learning_rate": 1.3619424054206664e-05,
      "loss": 0.4344,
      "step": 16950
    },
    {
      "epoch": 3.19,
      "grad_norm": 3.2976200580596924,
      "learning_rate": 1.3615659702616225e-05,
      "loss": 0.2875,
      "step": 16960
    },
    {
      "epoch": 3.19,
      "grad_norm": 5.71697998046875,
      "learning_rate": 1.3611895351025789e-05,
      "loss": 0.2327,
      "step": 16970
    },
    {
      "epoch": 3.2,
      "grad_norm": 4.075228691101074,
      "learning_rate": 1.360813099943535e-05,
      "loss": 0.0918,
      "step": 16980
    },
    {
      "epoch": 3.2,
      "grad_norm": 13.780068397521973,
      "learning_rate": 1.3604366647844908e-05,
      "loss": 0.3671,
      "step": 16990
    },
    {
      "epoch": 3.2,
      "grad_norm": 3.851155996322632,
      "learning_rate": 1.360060229625447e-05,
      "loss": 0.132,
      "step": 17000
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.02201663702726364,
      "learning_rate": 1.3596837944664033e-05,
      "loss": 0.1272,
      "step": 17010
    },
    {
      "epoch": 3.2,
      "grad_norm": 5.157936096191406,
      "learning_rate": 1.3593073593073594e-05,
      "loss": 0.2701,
      "step": 17020
    },
    {
      "epoch": 3.21,
      "grad_norm": 19.526235580444336,
      "learning_rate": 1.3589309241483156e-05,
      "loss": 0.2988,
      "step": 17030
    },
    {
      "epoch": 3.21,
      "grad_norm": 7.205384254455566,
      "learning_rate": 1.3585544889892717e-05,
      "loss": 0.1331,
      "step": 17040
    },
    {
      "epoch": 3.21,
      "grad_norm": 12.1398286819458,
      "learning_rate": 1.3581780538302279e-05,
      "loss": 0.2703,
      "step": 17050
    },
    {
      "epoch": 3.21,
      "grad_norm": 5.234925270080566,
      "learning_rate": 1.357801618671184e-05,
      "loss": 0.2406,
      "step": 17060
    },
    {
      "epoch": 3.21,
      "grad_norm": 12.121842384338379,
      "learning_rate": 1.3574251835121402e-05,
      "loss": 0.3448,
      "step": 17070
    },
    {
      "epoch": 3.21,
      "grad_norm": 0.14474855363368988,
      "learning_rate": 1.3570487483530963e-05,
      "loss": 0.0401,
      "step": 17080
    },
    {
      "epoch": 3.22,
      "grad_norm": 0.6939184665679932,
      "learning_rate": 1.3566723131940524e-05,
      "loss": 0.2325,
      "step": 17090
    },
    {
      "epoch": 3.22,
      "grad_norm": 6.818851947784424,
      "learning_rate": 1.3562958780350086e-05,
      "loss": 0.3959,
      "step": 17100
    },
    {
      "epoch": 3.22,
      "grad_norm": 6.479343891143799,
      "learning_rate": 1.3559194428759647e-05,
      "loss": 0.3152,
      "step": 17110
    },
    {
      "epoch": 3.22,
      "grad_norm": 5.823625564575195,
      "learning_rate": 1.3555430077169209e-05,
      "loss": 0.1768,
      "step": 17120
    },
    {
      "epoch": 3.22,
      "grad_norm": 18.35344886779785,
      "learning_rate": 1.355166572557877e-05,
      "loss": 0.2804,
      "step": 17130
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.9095736145973206,
      "learning_rate": 1.3547901373988332e-05,
      "loss": 0.3543,
      "step": 17140
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.03011692687869072,
      "learning_rate": 1.3544137022397895e-05,
      "loss": 0.0432,
      "step": 17150
    },
    {
      "epoch": 3.23,
      "grad_norm": 12.35361099243164,
      "learning_rate": 1.3540372670807453e-05,
      "loss": 0.2581,
      "step": 17160
    },
    {
      "epoch": 3.23,
      "grad_norm": 5.524060249328613,
      "learning_rate": 1.3536608319217014e-05,
      "loss": 0.2649,
      "step": 17170
    },
    {
      "epoch": 3.23,
      "grad_norm": 0.011408457532525063,
      "learning_rate": 1.3532843967626576e-05,
      "loss": 0.0997,
      "step": 17180
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.05037633702158928,
      "learning_rate": 1.3529079616036137e-05,
      "loss": 0.0588,
      "step": 17190
    },
    {
      "epoch": 3.24,
      "grad_norm": 5.936678409576416,
      "learning_rate": 1.35253152644457e-05,
      "loss": 0.2105,
      "step": 17200
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.14888136088848114,
      "learning_rate": 1.3521550912855262e-05,
      "loss": 0.4399,
      "step": 17210
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.03191644325852394,
      "learning_rate": 1.3517786561264823e-05,
      "loss": 0.1542,
      "step": 17220
    },
    {
      "epoch": 3.24,
      "grad_norm": 8.847940444946289,
      "learning_rate": 1.3514022209674385e-05,
      "loss": 0.2058,
      "step": 17230
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.16067425906658173,
      "learning_rate": 1.3510257858083946e-05,
      "loss": 0.1524,
      "step": 17240
    },
    {
      "epoch": 3.25,
      "grad_norm": 1.3063693046569824,
      "learning_rate": 1.3506493506493508e-05,
      "loss": 0.1415,
      "step": 17250
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.17863279581069946,
      "learning_rate": 1.350272915490307e-05,
      "loss": 0.3714,
      "step": 17260
    },
    {
      "epoch": 3.25,
      "grad_norm": 7.577050685882568,
      "learning_rate": 1.349896480331263e-05,
      "loss": 0.2928,
      "step": 17270
    },
    {
      "epoch": 3.25,
      "grad_norm": 6.5331711769104,
      "learning_rate": 1.3495200451722192e-05,
      "loss": 0.2992,
      "step": 17280
    },
    {
      "epoch": 3.25,
      "grad_norm": 0.23796579241752625,
      "learning_rate": 1.3491436100131754e-05,
      "loss": 0.1404,
      "step": 17290
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.723618745803833,
      "learning_rate": 1.3487671748541315e-05,
      "loss": 0.2882,
      "step": 17300
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.9221251606941223,
      "learning_rate": 1.3483907396950877e-05,
      "loss": 0.1415,
      "step": 17310
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.02485170029103756,
      "learning_rate": 1.3480143045360438e-05,
      "loss": 0.1402,
      "step": 17320
    },
    {
      "epoch": 3.26,
      "grad_norm": 25.6627140045166,
      "learning_rate": 1.347637869377e-05,
      "loss": 0.2799,
      "step": 17330
    },
    {
      "epoch": 3.26,
      "grad_norm": 0.02598695643246174,
      "learning_rate": 1.347261434217956e-05,
      "loss": 0.1652,
      "step": 17340
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.1155676320195198,
      "learning_rate": 1.346884999058912e-05,
      "loss": 0.2108,
      "step": 17350
    },
    {
      "epoch": 3.27,
      "grad_norm": 16.159772872924805,
      "learning_rate": 1.3465085638998682e-05,
      "loss": 0.2744,
      "step": 17360
    },
    {
      "epoch": 3.27,
      "grad_norm": 20.831344604492188,
      "learning_rate": 1.3461321287408244e-05,
      "loss": 0.2171,
      "step": 17370
    },
    {
      "epoch": 3.27,
      "grad_norm": 8.271937370300293,
      "learning_rate": 1.3457556935817807e-05,
      "loss": 0.1868,
      "step": 17380
    },
    {
      "epoch": 3.27,
      "grad_norm": 5.880423069000244,
      "learning_rate": 1.3453792584227368e-05,
      "loss": 0.2685,
      "step": 17390
    },
    {
      "epoch": 3.27,
      "grad_norm": 1.3605464696884155,
      "learning_rate": 1.345002823263693e-05,
      "loss": 0.2187,
      "step": 17400
    },
    {
      "epoch": 3.28,
      "grad_norm": 25.923091888427734,
      "learning_rate": 1.3446263881046491e-05,
      "loss": 0.2892,
      "step": 17410
    },
    {
      "epoch": 3.28,
      "grad_norm": 6.927120208740234,
      "learning_rate": 1.3442499529456053e-05,
      "loss": 0.2079,
      "step": 17420
    },
    {
      "epoch": 3.28,
      "grad_norm": 0.01690254919230938,
      "learning_rate": 1.3438735177865614e-05,
      "loss": 0.3319,
      "step": 17430
    },
    {
      "epoch": 3.28,
      "grad_norm": 15.431009292602539,
      "learning_rate": 1.3434970826275176e-05,
      "loss": 0.1235,
      "step": 17440
    },
    {
      "epoch": 3.28,
      "grad_norm": 12.719295501708984,
      "learning_rate": 1.3431206474684737e-05,
      "loss": 0.3588,
      "step": 17450
    },
    {
      "epoch": 3.29,
      "grad_norm": 18.71914291381836,
      "learning_rate": 1.3427442123094298e-05,
      "loss": 0.2502,
      "step": 17460
    },
    {
      "epoch": 3.29,
      "grad_norm": 0.05563555657863617,
      "learning_rate": 1.342367777150386e-05,
      "loss": 0.3576,
      "step": 17470
    },
    {
      "epoch": 3.29,
      "grad_norm": 9.438433647155762,
      "learning_rate": 1.3419913419913421e-05,
      "loss": 0.2411,
      "step": 17480
    },
    {
      "epoch": 3.29,
      "grad_norm": 11.73392105102539,
      "learning_rate": 1.3416149068322983e-05,
      "loss": 0.3242,
      "step": 17490
    },
    {
      "epoch": 3.29,
      "grad_norm": 8.685367584228516,
      "learning_rate": 1.3412384716732544e-05,
      "loss": 0.219,
      "step": 17500
    },
    {
      "epoch": 3.3,
      "grad_norm": 11.01241397857666,
      "learning_rate": 1.3408620365142104e-05,
      "loss": 0.2004,
      "step": 17510
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.09492083638906479,
      "learning_rate": 1.3404856013551666e-05,
      "loss": 0.3801,
      "step": 17520
    },
    {
      "epoch": 3.3,
      "grad_norm": 0.8663870096206665,
      "learning_rate": 1.3401091661961227e-05,
      "loss": 0.1844,
      "step": 17530
    },
    {
      "epoch": 3.3,
      "grad_norm": 4.47327184677124,
      "learning_rate": 1.3397327310370788e-05,
      "loss": 0.1316,
      "step": 17540
    },
    {
      "epoch": 3.3,
      "grad_norm": 9.153975486755371,
      "learning_rate": 1.339356295878035e-05,
      "loss": 0.2003,
      "step": 17550
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.15328913927078247,
      "learning_rate": 1.3389798607189911e-05,
      "loss": 0.1238,
      "step": 17560
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.8576046228408813,
      "learning_rate": 1.3386034255599475e-05,
      "loss": 0.4624,
      "step": 17570
    },
    {
      "epoch": 3.31,
      "grad_norm": 10.703767776489258,
      "learning_rate": 1.3382269904009036e-05,
      "loss": 0.2,
      "step": 17580
    },
    {
      "epoch": 3.31,
      "grad_norm": 8.39234733581543,
      "learning_rate": 1.3378505552418597e-05,
      "loss": 0.2076,
      "step": 17590
    },
    {
      "epoch": 3.31,
      "grad_norm": 4.090664386749268,
      "learning_rate": 1.3374741200828159e-05,
      "loss": 0.2392,
      "step": 17600
    },
    {
      "epoch": 3.31,
      "grad_norm": 0.025757938623428345,
      "learning_rate": 1.337097684923772e-05,
      "loss": 0.3029,
      "step": 17610
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.19906479120254517,
      "learning_rate": 1.3367212497647282e-05,
      "loss": 0.2664,
      "step": 17620
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.07912864536046982,
      "learning_rate": 1.3363448146056843e-05,
      "loss": 0.2271,
      "step": 17630
    },
    {
      "epoch": 3.32,
      "grad_norm": 6.767550945281982,
      "learning_rate": 1.3359683794466405e-05,
      "loss": 0.0572,
      "step": 17640
    },
    {
      "epoch": 3.32,
      "grad_norm": 1.6497043371200562,
      "learning_rate": 1.3355919442875966e-05,
      "loss": 0.0348,
      "step": 17650
    },
    {
      "epoch": 3.32,
      "grad_norm": 4.075567722320557,
      "learning_rate": 1.3352155091285528e-05,
      "loss": 0.0201,
      "step": 17660
    },
    {
      "epoch": 3.33,
      "grad_norm": 5.05453634262085,
      "learning_rate": 1.3348390739695089e-05,
      "loss": 0.3208,
      "step": 17670
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.026265982538461685,
      "learning_rate": 1.334462638810465e-05,
      "loss": 0.157,
      "step": 17680
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.09509198367595673,
      "learning_rate": 1.334086203651421e-05,
      "loss": 0.1413,
      "step": 17690
    },
    {
      "epoch": 3.33,
      "grad_norm": 37.765716552734375,
      "learning_rate": 1.3337097684923772e-05,
      "loss": 0.2444,
      "step": 17700
    },
    {
      "epoch": 3.33,
      "grad_norm": 20.070215225219727,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0899,
      "step": 17710
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.03865547478199005,
      "learning_rate": 1.3329568981742895e-05,
      "loss": 0.0547,
      "step": 17720
    },
    {
      "epoch": 3.34,
      "grad_norm": 8.229589462280273,
      "learning_rate": 1.3325804630152456e-05,
      "loss": 0.4493,
      "step": 17730
    },
    {
      "epoch": 3.34,
      "grad_norm": 0.06470435857772827,
      "learning_rate": 1.3322040278562018e-05,
      "loss": 0.2374,
      "step": 17740
    },
    {
      "epoch": 3.34,
      "grad_norm": 10.119998931884766,
      "learning_rate": 1.331827592697158e-05,
      "loss": 0.0873,
      "step": 17750
    },
    {
      "epoch": 3.34,
      "grad_norm": 9.789290428161621,
      "learning_rate": 1.3314511575381142e-05,
      "loss": 0.2507,
      "step": 17760
    },
    {
      "epoch": 3.34,
      "grad_norm": 12.330989837646484,
      "learning_rate": 1.3310747223790704e-05,
      "loss": 0.1905,
      "step": 17770
    },
    {
      "epoch": 3.35,
      "grad_norm": 21.401165008544922,
      "learning_rate": 1.3306982872200265e-05,
      "loss": 0.361,
      "step": 17780
    },
    {
      "epoch": 3.35,
      "grad_norm": 8.611123085021973,
      "learning_rate": 1.3303218520609827e-05,
      "loss": 0.2034,
      "step": 17790
    },
    {
      "epoch": 3.35,
      "grad_norm": 25.793087005615234,
      "learning_rate": 1.3299454169019388e-05,
      "loss": 0.195,
      "step": 17800
    },
    {
      "epoch": 3.35,
      "grad_norm": 6.79320764541626,
      "learning_rate": 1.329568981742895e-05,
      "loss": 0.252,
      "step": 17810
    },
    {
      "epoch": 3.35,
      "grad_norm": 0.09510830789804459,
      "learning_rate": 1.3291925465838511e-05,
      "loss": 0.2624,
      "step": 17820
    },
    {
      "epoch": 3.36,
      "grad_norm": 6.306151390075684,
      "learning_rate": 1.3288161114248073e-05,
      "loss": 0.1709,
      "step": 17830
    },
    {
      "epoch": 3.36,
      "grad_norm": 2.490610122680664,
      "learning_rate": 1.3284396762657634e-05,
      "loss": 0.1579,
      "step": 17840
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.053604841232299805,
      "learning_rate": 1.3280632411067195e-05,
      "loss": 0.3011,
      "step": 17850
    },
    {
      "epoch": 3.36,
      "grad_norm": 26.880477905273438,
      "learning_rate": 1.3276868059476755e-05,
      "loss": 0.3319,
      "step": 17860
    },
    {
      "epoch": 3.36,
      "grad_norm": 8.298501968383789,
      "learning_rate": 1.3273103707886317e-05,
      "loss": 0.2387,
      "step": 17870
    },
    {
      "epoch": 3.37,
      "grad_norm": 0.022843526676297188,
      "learning_rate": 1.3269339356295878e-05,
      "loss": 0.296,
      "step": 17880
    },
    {
      "epoch": 3.37,
      "grad_norm": 6.171627044677734,
      "learning_rate": 1.326557500470544e-05,
      "loss": 0.2355,
      "step": 17890
    },
    {
      "epoch": 3.37,
      "grad_norm": 20.123029708862305,
      "learning_rate": 1.3261810653115001e-05,
      "loss": 0.1571,
      "step": 17900
    },
    {
      "epoch": 3.37,
      "grad_norm": 9.388335227966309,
      "learning_rate": 1.3258046301524563e-05,
      "loss": 0.2954,
      "step": 17910
    },
    {
      "epoch": 3.37,
      "grad_norm": 6.993913650512695,
      "learning_rate": 1.3254281949934124e-05,
      "loss": 0.328,
      "step": 17920
    },
    {
      "epoch": 3.37,
      "grad_norm": 8.310067176818848,
      "learning_rate": 1.3250517598343685e-05,
      "loss": 0.1283,
      "step": 17930
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.0118014020845294,
      "learning_rate": 1.3246753246753249e-05,
      "loss": 0.2404,
      "step": 17940
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.051254309713840485,
      "learning_rate": 1.324298889516281e-05,
      "loss": 0.0751,
      "step": 17950
    },
    {
      "epoch": 3.38,
      "grad_norm": 15.308034896850586,
      "learning_rate": 1.3239224543572372e-05,
      "loss": 0.2343,
      "step": 17960
    },
    {
      "epoch": 3.38,
      "grad_norm": 0.021542532369494438,
      "learning_rate": 1.3235460191981933e-05,
      "loss": 0.4817,
      "step": 17970
    },
    {
      "epoch": 3.38,
      "grad_norm": 16.53514862060547,
      "learning_rate": 1.3231695840391494e-05,
      "loss": 0.1758,
      "step": 17980
    },
    {
      "epoch": 3.39,
      "grad_norm": 19.60160255432129,
      "learning_rate": 1.3227931488801056e-05,
      "loss": 0.2552,
      "step": 17990
    },
    {
      "epoch": 3.39,
      "grad_norm": 11.5253324508667,
      "learning_rate": 1.3224167137210617e-05,
      "loss": 0.4783,
      "step": 18000
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.05425533279776573,
      "learning_rate": 1.3220402785620179e-05,
      "loss": 0.256,
      "step": 18010
    },
    {
      "epoch": 3.39,
      "grad_norm": 40.98523712158203,
      "learning_rate": 1.321663843402974e-05,
      "loss": 0.3216,
      "step": 18020
    },
    {
      "epoch": 3.39,
      "grad_norm": 0.02880512736737728,
      "learning_rate": 1.32128740824393e-05,
      "loss": 0.3365,
      "step": 18030
    },
    {
      "epoch": 3.4,
      "grad_norm": 15.215142250061035,
      "learning_rate": 1.3209109730848861e-05,
      "loss": 0.4386,
      "step": 18040
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.4875926971435547,
      "learning_rate": 1.3205345379258423e-05,
      "loss": 0.3964,
      "step": 18050
    },
    {
      "epoch": 3.4,
      "grad_norm": 11.352274894714355,
      "learning_rate": 1.3201581027667984e-05,
      "loss": 0.3201,
      "step": 18060
    },
    {
      "epoch": 3.4,
      "grad_norm": 17.443998336791992,
      "learning_rate": 1.3197816676077546e-05,
      "loss": 0.3349,
      "step": 18070
    },
    {
      "epoch": 3.4,
      "grad_norm": 0.17464424669742584,
      "learning_rate": 1.3194052324487107e-05,
      "loss": 0.3656,
      "step": 18080
    },
    {
      "epoch": 3.4,
      "grad_norm": 9.667675971984863,
      "learning_rate": 1.3190287972896669e-05,
      "loss": 0.2552,
      "step": 18090
    },
    {
      "epoch": 3.41,
      "grad_norm": 30.626279830932617,
      "learning_rate": 1.318652362130623e-05,
      "loss": 0.2463,
      "step": 18100
    },
    {
      "epoch": 3.41,
      "grad_norm": 15.08073616027832,
      "learning_rate": 1.3182759269715792e-05,
      "loss": 0.2439,
      "step": 18110
    },
    {
      "epoch": 3.41,
      "grad_norm": 11.383460998535156,
      "learning_rate": 1.3178994918125355e-05,
      "loss": 0.2845,
      "step": 18120
    },
    {
      "epoch": 3.41,
      "grad_norm": 20.405689239501953,
      "learning_rate": 1.3175230566534916e-05,
      "loss": 0.3568,
      "step": 18130
    },
    {
      "epoch": 3.41,
      "grad_norm": 8.989091873168945,
      "learning_rate": 1.3171466214944478e-05,
      "loss": 0.1156,
      "step": 18140
    },
    {
      "epoch": 3.42,
      "grad_norm": 1.81601881980896,
      "learning_rate": 1.316770186335404e-05,
      "loss": 0.1251,
      "step": 18150
    },
    {
      "epoch": 3.42,
      "grad_norm": 10.025184631347656,
      "learning_rate": 1.31639375117636e-05,
      "loss": 0.1863,
      "step": 18160
    },
    {
      "epoch": 3.42,
      "grad_norm": 18.12965202331543,
      "learning_rate": 1.3160173160173162e-05,
      "loss": 0.1719,
      "step": 18170
    },
    {
      "epoch": 3.42,
      "grad_norm": 10.812607765197754,
      "learning_rate": 1.3156408808582724e-05,
      "loss": 0.3166,
      "step": 18180
    },
    {
      "epoch": 3.42,
      "grad_norm": 3.363036632537842,
      "learning_rate": 1.3152644456992285e-05,
      "loss": 0.268,
      "step": 18190
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.7742383480072021,
      "learning_rate": 1.3148880105401847e-05,
      "loss": 0.5163,
      "step": 18200
    },
    {
      "epoch": 3.43,
      "grad_norm": 5.08019495010376,
      "learning_rate": 1.3145115753811406e-05,
      "loss": 0.0601,
      "step": 18210
    },
    {
      "epoch": 3.43,
      "grad_norm": 9.661688804626465,
      "learning_rate": 1.3141351402220968e-05,
      "loss": 0.0427,
      "step": 18220
    },
    {
      "epoch": 3.43,
      "grad_norm": 13.514941215515137,
      "learning_rate": 1.313758705063053e-05,
      "loss": 0.3507,
      "step": 18230
    },
    {
      "epoch": 3.43,
      "grad_norm": 1.3952182531356812,
      "learning_rate": 1.313382269904009e-05,
      "loss": 0.0955,
      "step": 18240
    },
    {
      "epoch": 3.43,
      "grad_norm": 13.047310829162598,
      "learning_rate": 1.3130058347449652e-05,
      "loss": 0.3308,
      "step": 18250
    },
    {
      "epoch": 3.44,
      "grad_norm": 14.103184700012207,
      "learning_rate": 1.3126293995859214e-05,
      "loss": 0.4961,
      "step": 18260
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.3597294092178345,
      "learning_rate": 1.3122529644268775e-05,
      "loss": 0.0991,
      "step": 18270
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.11809659004211426,
      "learning_rate": 1.3118765292678337e-05,
      "loss": 0.2138,
      "step": 18280
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.0671916007995605,
      "learning_rate": 1.3115000941087898e-05,
      "loss": 0.219,
      "step": 18290
    },
    {
      "epoch": 3.44,
      "grad_norm": 0.015438083559274673,
      "learning_rate": 1.311123658949746e-05,
      "loss": 0.2751,
      "step": 18300
    },
    {
      "epoch": 3.45,
      "grad_norm": 14.662311553955078,
      "learning_rate": 1.3107472237907023e-05,
      "loss": 0.18,
      "step": 18310
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.11608461290597916,
      "learning_rate": 1.3103707886316584e-05,
      "loss": 0.223,
      "step": 18320
    },
    {
      "epoch": 3.45,
      "grad_norm": 5.570398807525635,
      "learning_rate": 1.3099943534726146e-05,
      "loss": 0.2862,
      "step": 18330
    },
    {
      "epoch": 3.45,
      "grad_norm": 14.614316940307617,
      "learning_rate": 1.3096179183135707e-05,
      "loss": 0.2999,
      "step": 18340
    },
    {
      "epoch": 3.45,
      "grad_norm": 0.03411823883652687,
      "learning_rate": 1.3092414831545268e-05,
      "loss": 0.0488,
      "step": 18350
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.36282479763031,
      "learning_rate": 1.308865047995483e-05,
      "loss": 0.1488,
      "step": 18360
    },
    {
      "epoch": 3.46,
      "grad_norm": 40.979190826416016,
      "learning_rate": 1.3084886128364391e-05,
      "loss": 0.1307,
      "step": 18370
    },
    {
      "epoch": 3.46,
      "grad_norm": 3.8575668334960938,
      "learning_rate": 1.3081121776773951e-05,
      "loss": 0.2757,
      "step": 18380
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.03161320090293884,
      "learning_rate": 1.3077357425183513e-05,
      "loss": 0.0925,
      "step": 18390
    },
    {
      "epoch": 3.46,
      "grad_norm": 6.939659118652344,
      "learning_rate": 1.3073593073593074e-05,
      "loss": 0.1144,
      "step": 18400
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.3561144471168518,
      "learning_rate": 1.3069828722002636e-05,
      "loss": 0.3374,
      "step": 18410
    },
    {
      "epoch": 3.47,
      "grad_norm": 7.6319661140441895,
      "learning_rate": 1.3066064370412197e-05,
      "loss": 0.236,
      "step": 18420
    },
    {
      "epoch": 3.47,
      "grad_norm": 13.4546480178833,
      "learning_rate": 1.3062300018821758e-05,
      "loss": 0.2859,
      "step": 18430
    },
    {
      "epoch": 3.47,
      "grad_norm": 29.357236862182617,
      "learning_rate": 1.305853566723132e-05,
      "loss": 0.34,
      "step": 18440
    },
    {
      "epoch": 3.47,
      "grad_norm": 23.04450225830078,
      "learning_rate": 1.3054771315640881e-05,
      "loss": 0.2582,
      "step": 18450
    },
    {
      "epoch": 3.47,
      "grad_norm": 0.019754087552428246,
      "learning_rate": 1.3051006964050443e-05,
      "loss": 0.4005,
      "step": 18460
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.09689188003540039,
      "learning_rate": 1.3047242612460004e-05,
      "loss": 0.1699,
      "step": 18470
    },
    {
      "epoch": 3.48,
      "grad_norm": 0.018619855865836143,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.23,
      "step": 18480
    },
    {
      "epoch": 3.48,
      "grad_norm": 17.757638931274414,
      "learning_rate": 1.3039713909279127e-05,
      "loss": 0.2441,
      "step": 18490
    },
    {
      "epoch": 3.48,
      "grad_norm": 23.871299743652344,
      "learning_rate": 1.303594955768869e-05,
      "loss": 0.1921,
      "step": 18500
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.3272854089736938,
      "learning_rate": 1.3032185206098252e-05,
      "loss": 0.2013,
      "step": 18510
    },
    {
      "epoch": 3.49,
      "grad_norm": 10.177550315856934,
      "learning_rate": 1.3028420854507813e-05,
      "loss": 0.2397,
      "step": 18520
    },
    {
      "epoch": 3.49,
      "grad_norm": 22.167346954345703,
      "learning_rate": 1.3024656502917375e-05,
      "loss": 0.061,
      "step": 18530
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.4717220067977905,
      "learning_rate": 1.3020892151326936e-05,
      "loss": 0.1541,
      "step": 18540
    },
    {
      "epoch": 3.49,
      "grad_norm": 16.1068058013916,
      "learning_rate": 1.3017127799736498e-05,
      "loss": 0.2986,
      "step": 18550
    },
    {
      "epoch": 3.49,
      "grad_norm": 24.14872169494629,
      "learning_rate": 1.3013363448146057e-05,
      "loss": 0.1846,
      "step": 18560
    },
    {
      "epoch": 3.5,
      "grad_norm": 6.816590785980225,
      "learning_rate": 1.3009599096555619e-05,
      "loss": 0.2902,
      "step": 18570
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.017113568261265755,
      "learning_rate": 1.300583474496518e-05,
      "loss": 0.1645,
      "step": 18580
    },
    {
      "epoch": 3.5,
      "grad_norm": 21.744606018066406,
      "learning_rate": 1.3002070393374742e-05,
      "loss": 0.3113,
      "step": 18590
    },
    {
      "epoch": 3.5,
      "grad_norm": 10.586426734924316,
      "learning_rate": 1.2998306041784303e-05,
      "loss": 0.1857,
      "step": 18600
    },
    {
      "epoch": 3.5,
      "grad_norm": 0.038934022188186646,
      "learning_rate": 1.2994541690193865e-05,
      "loss": 0.1354,
      "step": 18610
    },
    {
      "epoch": 3.5,
      "grad_norm": 8.841143608093262,
      "learning_rate": 1.2990777338603426e-05,
      "loss": 0.4436,
      "step": 18620
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.7484642863273621,
      "learning_rate": 1.2987012987012988e-05,
      "loss": 0.2724,
      "step": 18630
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.030762378126382828,
      "learning_rate": 1.2983248635422549e-05,
      "loss": 0.2657,
      "step": 18640
    },
    {
      "epoch": 3.51,
      "grad_norm": 2.7259809970855713,
      "learning_rate": 1.297948428383211e-05,
      "loss": 0.2351,
      "step": 18650
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.07231257110834122,
      "learning_rate": 1.2975719932241672e-05,
      "loss": 0.3099,
      "step": 18660
    },
    {
      "epoch": 3.51,
      "grad_norm": 0.03723420202732086,
      "learning_rate": 1.2971955580651234e-05,
      "loss": 0.1125,
      "step": 18670
    },
    {
      "epoch": 3.52,
      "grad_norm": 13.549407005310059,
      "learning_rate": 1.2968191229060797e-05,
      "loss": 0.2583,
      "step": 18680
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.6006358861923218,
      "learning_rate": 1.2964426877470358e-05,
      "loss": 0.2372,
      "step": 18690
    },
    {
      "epoch": 3.52,
      "grad_norm": 4.2022786140441895,
      "learning_rate": 1.296066252587992e-05,
      "loss": 0.0551,
      "step": 18700
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.03636297583580017,
      "learning_rate": 1.2956898174289481e-05,
      "loss": 0.2499,
      "step": 18710
    },
    {
      "epoch": 3.52,
      "grad_norm": 0.07935820519924164,
      "learning_rate": 1.2953133822699043e-05,
      "loss": 0.2185,
      "step": 18720
    },
    {
      "epoch": 3.53,
      "grad_norm": 2.1129236221313477,
      "learning_rate": 1.2949369471108602e-05,
      "loss": 0.18,
      "step": 18730
    },
    {
      "epoch": 3.53,
      "grad_norm": 5.6079840660095215,
      "learning_rate": 1.2945605119518164e-05,
      "loss": 0.2065,
      "step": 18740
    },
    {
      "epoch": 3.53,
      "grad_norm": 8.630861282348633,
      "learning_rate": 1.2941840767927725e-05,
      "loss": 0.2866,
      "step": 18750
    },
    {
      "epoch": 3.53,
      "grad_norm": 9.242431640625,
      "learning_rate": 1.2938076416337287e-05,
      "loss": 0.3792,
      "step": 18760
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.06677576899528503,
      "learning_rate": 1.2934312064746848e-05,
      "loss": 0.2634,
      "step": 18770
    },
    {
      "epoch": 3.53,
      "grad_norm": 0.8697452545166016,
      "learning_rate": 1.293054771315641e-05,
      "loss": 0.1181,
      "step": 18780
    },
    {
      "epoch": 3.54,
      "grad_norm": 6.3697662353515625,
      "learning_rate": 1.2926783361565971e-05,
      "loss": 0.2249,
      "step": 18790
    },
    {
      "epoch": 3.54,
      "grad_norm": 10.377037048339844,
      "learning_rate": 1.2923019009975533e-05,
      "loss": 0.2175,
      "step": 18800
    },
    {
      "epoch": 3.54,
      "grad_norm": 0.03322554752230644,
      "learning_rate": 1.2919254658385094e-05,
      "loss": 0.1383,
      "step": 18810
    },
    {
      "epoch": 3.54,
      "grad_norm": 25.784677505493164,
      "learning_rate": 1.2915490306794655e-05,
      "loss": 0.254,
      "step": 18820
    },
    {
      "epoch": 3.54,
      "grad_norm": 6.488381862640381,
      "learning_rate": 1.2911725955204217e-05,
      "loss": 0.2625,
      "step": 18830
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.016591619700193405,
      "learning_rate": 1.2907961603613778e-05,
      "loss": 0.3053,
      "step": 18840
    },
    {
      "epoch": 3.55,
      "grad_norm": 2.5346968173980713,
      "learning_rate": 1.290419725202334e-05,
      "loss": 0.2021,
      "step": 18850
    },
    {
      "epoch": 3.55,
      "grad_norm": 8.318920135498047,
      "learning_rate": 1.2900432900432901e-05,
      "loss": 0.1265,
      "step": 18860
    },
    {
      "epoch": 3.55,
      "grad_norm": 10.090860366821289,
      "learning_rate": 1.2896668548842464e-05,
      "loss": 0.3277,
      "step": 18870
    },
    {
      "epoch": 3.55,
      "grad_norm": 0.010229360312223434,
      "learning_rate": 1.2892904197252026e-05,
      "loss": 0.1388,
      "step": 18880
    },
    {
      "epoch": 3.56,
      "grad_norm": 34.555633544921875,
      "learning_rate": 1.2889139845661587e-05,
      "loss": 0.3143,
      "step": 18890
    },
    {
      "epoch": 3.56,
      "grad_norm": 24.09465217590332,
      "learning_rate": 1.2885375494071149e-05,
      "loss": 0.2303,
      "step": 18900
    },
    {
      "epoch": 3.56,
      "grad_norm": 12.38953971862793,
      "learning_rate": 1.2881611142480709e-05,
      "loss": 0.1718,
      "step": 18910
    },
    {
      "epoch": 3.56,
      "grad_norm": 8.63287353515625,
      "learning_rate": 1.287784679089027e-05,
      "loss": 0.251,
      "step": 18920
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.019159475341439247,
      "learning_rate": 1.2874082439299832e-05,
      "loss": 0.3263,
      "step": 18930
    },
    {
      "epoch": 3.56,
      "grad_norm": 0.0978059396147728,
      "learning_rate": 1.2870318087709393e-05,
      "loss": 0.2126,
      "step": 18940
    },
    {
      "epoch": 3.57,
      "grad_norm": 11.365647315979004,
      "learning_rate": 1.2866553736118954e-05,
      "loss": 0.2144,
      "step": 18950
    },
    {
      "epoch": 3.57,
      "grad_norm": 9.961935043334961,
      "learning_rate": 1.2862789384528516e-05,
      "loss": 0.2335,
      "step": 18960
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.32675763964653015,
      "learning_rate": 1.2859025032938077e-05,
      "loss": 0.2283,
      "step": 18970
    },
    {
      "epoch": 3.57,
      "grad_norm": 0.020718161016702652,
      "learning_rate": 1.2855260681347639e-05,
      "loss": 0.2186,
      "step": 18980
    },
    {
      "epoch": 3.57,
      "grad_norm": 18.889326095581055,
      "learning_rate": 1.28514963297572e-05,
      "loss": 0.3825,
      "step": 18990
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.31656697392463684,
      "learning_rate": 1.2847731978166762e-05,
      "loss": 0.2011,
      "step": 19000
    },
    {
      "epoch": 3.58,
      "grad_norm": 24.44812774658203,
      "learning_rate": 1.2843967626576323e-05,
      "loss": 0.4098,
      "step": 19010
    },
    {
      "epoch": 3.58,
      "grad_norm": 7.814602851867676,
      "learning_rate": 1.2840203274985885e-05,
      "loss": 0.151,
      "step": 19020
    },
    {
      "epoch": 3.58,
      "grad_norm": 0.03191610053181648,
      "learning_rate": 1.2836438923395446e-05,
      "loss": 0.1617,
      "step": 19030
    },
    {
      "epoch": 3.58,
      "grad_norm": 13.205074310302734,
      "learning_rate": 1.2832674571805008e-05,
      "loss": 0.1369,
      "step": 19040
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.023082556203007698,
      "learning_rate": 1.282891022021457e-05,
      "loss": 0.2144,
      "step": 19050
    },
    {
      "epoch": 3.59,
      "grad_norm": 18.686664581298828,
      "learning_rate": 1.2825145868624132e-05,
      "loss": 0.264,
      "step": 19060
    },
    {
      "epoch": 3.59,
      "grad_norm": 3.568209409713745,
      "learning_rate": 1.2821381517033694e-05,
      "loss": 0.2137,
      "step": 19070
    },
    {
      "epoch": 3.59,
      "grad_norm": 3.681786298751831,
      "learning_rate": 1.2817617165443252e-05,
      "loss": 0.1617,
      "step": 19080
    },
    {
      "epoch": 3.59,
      "grad_norm": 1.6221308708190918,
      "learning_rate": 1.2813852813852813e-05,
      "loss": 0.0962,
      "step": 19090
    },
    {
      "epoch": 3.59,
      "grad_norm": 0.04819386079907417,
      "learning_rate": 1.2810088462262376e-05,
      "loss": 0.1628,
      "step": 19100
    },
    {
      "epoch": 3.6,
      "grad_norm": 24.68199348449707,
      "learning_rate": 1.2806324110671938e-05,
      "loss": 0.167,
      "step": 19110
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.04668772965669632,
      "learning_rate": 1.28025597590815e-05,
      "loss": 0.1642,
      "step": 19120
    },
    {
      "epoch": 3.6,
      "grad_norm": 24.059232711791992,
      "learning_rate": 1.279879540749106e-05,
      "loss": 0.2658,
      "step": 19130
    },
    {
      "epoch": 3.6,
      "grad_norm": 2.499809741973877,
      "learning_rate": 1.2795031055900622e-05,
      "loss": 0.1081,
      "step": 19140
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.023904163390398026,
      "learning_rate": 1.2791266704310184e-05,
      "loss": 0.3639,
      "step": 19150
    },
    {
      "epoch": 3.61,
      "grad_norm": 4.810695171356201,
      "learning_rate": 1.2787502352719745e-05,
      "loss": 0.3009,
      "step": 19160
    },
    {
      "epoch": 3.61,
      "grad_norm": 9.599764823913574,
      "learning_rate": 1.2783738001129307e-05,
      "loss": 0.1026,
      "step": 19170
    },
    {
      "epoch": 3.61,
      "grad_norm": 1.275760293006897,
      "learning_rate": 1.2779973649538868e-05,
      "loss": 0.2448,
      "step": 19180
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.007262041326612234,
      "learning_rate": 1.277620929794843e-05,
      "loss": 0.1134,
      "step": 19190
    },
    {
      "epoch": 3.61,
      "grad_norm": 9.75331974029541,
      "learning_rate": 1.2772444946357991e-05,
      "loss": 0.1418,
      "step": 19200
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.05042801797389984,
      "learning_rate": 1.2768680594767552e-05,
      "loss": 0.2094,
      "step": 19210
    },
    {
      "epoch": 3.62,
      "grad_norm": 3.5360124111175537,
      "learning_rate": 1.2764916243177114e-05,
      "loss": 0.3447,
      "step": 19220
    },
    {
      "epoch": 3.62,
      "grad_norm": 15.299063682556152,
      "learning_rate": 1.2761151891586675e-05,
      "loss": 0.2404,
      "step": 19230
    },
    {
      "epoch": 3.62,
      "grad_norm": 0.22137901186943054,
      "learning_rate": 1.2757387539996238e-05,
      "loss": 0.2584,
      "step": 19240
    },
    {
      "epoch": 3.62,
      "grad_norm": 27.29067039489746,
      "learning_rate": 1.2753623188405797e-05,
      "loss": 0.2296,
      "step": 19250
    },
    {
      "epoch": 3.63,
      "grad_norm": 3.783214807510376,
      "learning_rate": 1.2749858836815358e-05,
      "loss": 0.2145,
      "step": 19260
    },
    {
      "epoch": 3.63,
      "grad_norm": 0.05169679597020149,
      "learning_rate": 1.274609448522492e-05,
      "loss": 0.1959,
      "step": 19270
    },
    {
      "epoch": 3.63,
      "grad_norm": 16.714921951293945,
      "learning_rate": 1.2742330133634483e-05,
      "loss": 0.1651,
      "step": 19280
    },
    {
      "epoch": 3.63,
      "grad_norm": 3.3195347785949707,
      "learning_rate": 1.2738565782044044e-05,
      "loss": 0.187,
      "step": 19290
    },
    {
      "epoch": 3.63,
      "grad_norm": 23.594141006469727,
      "learning_rate": 1.2734801430453606e-05,
      "loss": 0.2738,
      "step": 19300
    },
    {
      "epoch": 3.63,
      "grad_norm": 9.604933738708496,
      "learning_rate": 1.2731037078863167e-05,
      "loss": 0.1777,
      "step": 19310
    },
    {
      "epoch": 3.64,
      "grad_norm": 8.786954879760742,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 0.3784,
      "step": 19320
    },
    {
      "epoch": 3.64,
      "grad_norm": 5.983619689941406,
      "learning_rate": 1.272350837568229e-05,
      "loss": 0.1124,
      "step": 19330
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.390774130821228,
      "learning_rate": 1.2719744024091851e-05,
      "loss": 0.1912,
      "step": 19340
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.6272754669189453,
      "learning_rate": 1.2715979672501413e-05,
      "loss": 0.2749,
      "step": 19350
    },
    {
      "epoch": 3.64,
      "grad_norm": 4.477691650390625,
      "learning_rate": 1.2712215320910974e-05,
      "loss": 0.3231,
      "step": 19360
    },
    {
      "epoch": 3.65,
      "grad_norm": 0.013151847757399082,
      "learning_rate": 1.2708450969320536e-05,
      "loss": 0.1796,
      "step": 19370
    },
    {
      "epoch": 3.65,
      "grad_norm": 18.439233779907227,
      "learning_rate": 1.2704686617730097e-05,
      "loss": 0.2412,
      "step": 19380
    },
    {
      "epoch": 3.65,
      "grad_norm": 2.475517511367798,
      "learning_rate": 1.2700922266139659e-05,
      "loss": 0.2153,
      "step": 19390
    },
    {
      "epoch": 3.65,
      "grad_norm": 14.672127723693848,
      "learning_rate": 1.269715791454922e-05,
      "loss": 0.2564,
      "step": 19400
    },
    {
      "epoch": 3.65,
      "grad_norm": 13.726595878601074,
      "learning_rate": 1.2693393562958782e-05,
      "loss": 0.1525,
      "step": 19410
    },
    {
      "epoch": 3.66,
      "grad_norm": 15.752335548400879,
      "learning_rate": 1.2689629211368343e-05,
      "loss": 0.1391,
      "step": 19420
    },
    {
      "epoch": 3.66,
      "grad_norm": 0.5268934965133667,
      "learning_rate": 1.2685864859777903e-05,
      "loss": 0.2104,
      "step": 19430
    },
    {
      "epoch": 3.66,
      "grad_norm": 5.071855545043945,
      "learning_rate": 1.2682100508187464e-05,
      "loss": 0.2811,
      "step": 19440
    },
    {
      "epoch": 3.66,
      "grad_norm": 19.39012336730957,
      "learning_rate": 1.2678336156597026e-05,
      "loss": 0.3444,
      "step": 19450
    },
    {
      "epoch": 3.66,
      "grad_norm": 56.40627670288086,
      "learning_rate": 1.2674571805006587e-05,
      "loss": 0.2374,
      "step": 19460
    },
    {
      "epoch": 3.66,
      "grad_norm": 16.291536331176758,
      "learning_rate": 1.267080745341615e-05,
      "loss": 0.2836,
      "step": 19470
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.2644953727722168,
      "learning_rate": 1.2667043101825712e-05,
      "loss": 0.3444,
      "step": 19480
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.04314287379384041,
      "learning_rate": 1.2663278750235273e-05,
      "loss": 0.2642,
      "step": 19490
    },
    {
      "epoch": 3.67,
      "grad_norm": 17.981103897094727,
      "learning_rate": 1.2659514398644835e-05,
      "loss": 0.4119,
      "step": 19500
    },
    {
      "epoch": 3.67,
      "grad_norm": 24.649415969848633,
      "learning_rate": 1.2655750047054396e-05,
      "loss": 0.1759,
      "step": 19510
    },
    {
      "epoch": 3.67,
      "grad_norm": 0.3834242522716522,
      "learning_rate": 1.2651985695463958e-05,
      "loss": 0.1658,
      "step": 19520
    },
    {
      "epoch": 3.68,
      "grad_norm": 32.82929992675781,
      "learning_rate": 1.2648221343873519e-05,
      "loss": 0.222,
      "step": 19530
    },
    {
      "epoch": 3.68,
      "grad_norm": 5.842701435089111,
      "learning_rate": 1.264445699228308e-05,
      "loss": 0.1484,
      "step": 19540
    },
    {
      "epoch": 3.68,
      "grad_norm": 5.63521671295166,
      "learning_rate": 1.2640692640692642e-05,
      "loss": 0.1124,
      "step": 19550
    },
    {
      "epoch": 3.68,
      "grad_norm": 6.555415630340576,
      "learning_rate": 1.2636928289102204e-05,
      "loss": 0.2431,
      "step": 19560
    },
    {
      "epoch": 3.68,
      "grad_norm": 13.561223983764648,
      "learning_rate": 1.2633163937511765e-05,
      "loss": 0.23,
      "step": 19570
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.8318793177604675,
      "learning_rate": 1.2629399585921326e-05,
      "loss": 0.1946,
      "step": 19580
    },
    {
      "epoch": 3.69,
      "grad_norm": 4.063101768493652,
      "learning_rate": 1.2625635234330888e-05,
      "loss": 0.1753,
      "step": 19590
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.02116585709154606,
      "learning_rate": 1.2621870882740448e-05,
      "loss": 0.2901,
      "step": 19600
    },
    {
      "epoch": 3.69,
      "grad_norm": 7.278772354125977,
      "learning_rate": 1.2618106531150009e-05,
      "loss": 0.1562,
      "step": 19610
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.08567207306623459,
      "learning_rate": 1.261434217955957e-05,
      "loss": 0.1559,
      "step": 19620
    },
    {
      "epoch": 3.69,
      "grad_norm": 0.09141389280557632,
      "learning_rate": 1.2610577827969132e-05,
      "loss": 0.071,
      "step": 19630
    },
    {
      "epoch": 3.7,
      "grad_norm": 9.606438636779785,
      "learning_rate": 1.2606813476378694e-05,
      "loss": 0.3498,
      "step": 19640
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.007803948130458593,
      "learning_rate": 1.2603049124788255e-05,
      "loss": 0.1711,
      "step": 19650
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.15334540605545044,
      "learning_rate": 1.2599284773197818e-05,
      "loss": 0.1254,
      "step": 19660
    },
    {
      "epoch": 3.7,
      "grad_norm": 0.22281348705291748,
      "learning_rate": 1.259552042160738e-05,
      "loss": 0.1671,
      "step": 19670
    },
    {
      "epoch": 3.7,
      "grad_norm": 12.31337833404541,
      "learning_rate": 1.2591756070016941e-05,
      "loss": 0.4899,
      "step": 19680
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.1865173727273941,
      "learning_rate": 1.2587991718426503e-05,
      "loss": 0.1203,
      "step": 19690
    },
    {
      "epoch": 3.71,
      "grad_norm": 6.36646032333374,
      "learning_rate": 1.2584227366836064e-05,
      "loss": 0.2935,
      "step": 19700
    },
    {
      "epoch": 3.71,
      "grad_norm": 0.03346915915608406,
      "learning_rate": 1.2580463015245625e-05,
      "loss": 0.1765,
      "step": 19710
    },
    {
      "epoch": 3.71,
      "grad_norm": 18.94300079345703,
      "learning_rate": 1.2576698663655187e-05,
      "loss": 0.3102,
      "step": 19720
    },
    {
      "epoch": 3.71,
      "grad_norm": 6.480704307556152,
      "learning_rate": 1.2572934312064748e-05,
      "loss": 0.2403,
      "step": 19730
    },
    {
      "epoch": 3.72,
      "grad_norm": 10.412813186645508,
      "learning_rate": 1.256916996047431e-05,
      "loss": 0.4047,
      "step": 19740
    },
    {
      "epoch": 3.72,
      "grad_norm": 13.385335922241211,
      "learning_rate": 1.2565405608883871e-05,
      "loss": 0.3237,
      "step": 19750
    },
    {
      "epoch": 3.72,
      "grad_norm": 6.5037360191345215,
      "learning_rate": 1.2561641257293433e-05,
      "loss": 0.2442,
      "step": 19760
    },
    {
      "epoch": 3.72,
      "grad_norm": 10.88359546661377,
      "learning_rate": 1.2557876905702994e-05,
      "loss": 0.321,
      "step": 19770
    },
    {
      "epoch": 3.72,
      "grad_norm": 10.45460319519043,
      "learning_rate": 1.2554112554112554e-05,
      "loss": 0.2046,
      "step": 19780
    },
    {
      "epoch": 3.72,
      "grad_norm": 13.531464576721191,
      "learning_rate": 1.2550348202522115e-05,
      "loss": 0.2528,
      "step": 19790
    },
    {
      "epoch": 3.73,
      "grad_norm": 0.05770757049322128,
      "learning_rate": 1.2546583850931677e-05,
      "loss": 0.2585,
      "step": 19800
    },
    {
      "epoch": 3.73,
      "grad_norm": 7.938133239746094,
      "learning_rate": 1.2542819499341238e-05,
      "loss": 0.3382,
      "step": 19810
    },
    {
      "epoch": 3.73,
      "grad_norm": 4.427739143371582,
      "learning_rate": 1.25390551477508e-05,
      "loss": 0.1261,
      "step": 19820
    },
    {
      "epoch": 3.73,
      "grad_norm": 12.087244033813477,
      "learning_rate": 1.2535290796160361e-05,
      "loss": 0.2642,
      "step": 19830
    },
    {
      "epoch": 3.73,
      "grad_norm": 26.63177490234375,
      "learning_rate": 1.2531526444569924e-05,
      "loss": 0.3847,
      "step": 19840
    },
    {
      "epoch": 3.74,
      "grad_norm": 20.069252014160156,
      "learning_rate": 1.2527762092979486e-05,
      "loss": 0.1637,
      "step": 19850
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.19314803183078766,
      "learning_rate": 1.2523997741389047e-05,
      "loss": 0.0505,
      "step": 19860
    },
    {
      "epoch": 3.74,
      "grad_norm": 3.8977651596069336,
      "learning_rate": 1.2520233389798609e-05,
      "loss": 0.163,
      "step": 19870
    },
    {
      "epoch": 3.74,
      "grad_norm": 9.880955696105957,
      "learning_rate": 1.251646903820817e-05,
      "loss": 0.4392,
      "step": 19880
    },
    {
      "epoch": 3.74,
      "grad_norm": 0.20131833851337433,
      "learning_rate": 1.2512704686617732e-05,
      "loss": 0.2005,
      "step": 19890
    },
    {
      "epoch": 3.75,
      "grad_norm": 14.037400245666504,
      "learning_rate": 1.2508940335027293e-05,
      "loss": 0.1982,
      "step": 19900
    },
    {
      "epoch": 3.75,
      "grad_norm": 15.88588809967041,
      "learning_rate": 1.2505175983436855e-05,
      "loss": 0.216,
      "step": 19910
    },
    {
      "epoch": 3.75,
      "grad_norm": 38.70572280883789,
      "learning_rate": 1.2501411631846416e-05,
      "loss": 0.1272,
      "step": 19920
    },
    {
      "epoch": 3.75,
      "grad_norm": 6.377044200897217,
      "learning_rate": 1.2497647280255978e-05,
      "loss": 0.1032,
      "step": 19930
    },
    {
      "epoch": 3.75,
      "grad_norm": 5.340588092803955,
      "learning_rate": 1.2493882928665539e-05,
      "loss": 0.1864,
      "step": 19940
    },
    {
      "epoch": 3.75,
      "grad_norm": 0.030735017731785774,
      "learning_rate": 1.2490118577075099e-05,
      "loss": 0.2025,
      "step": 19950
    },
    {
      "epoch": 3.76,
      "grad_norm": 8.229569435119629,
      "learning_rate": 1.248635422548466e-05,
      "loss": 0.3231,
      "step": 19960
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.25035369396209717,
      "learning_rate": 1.2482589873894222e-05,
      "loss": 0.4092,
      "step": 19970
    },
    {
      "epoch": 3.76,
      "grad_norm": 1.8606724739074707,
      "learning_rate": 1.2478825522303783e-05,
      "loss": 0.3048,
      "step": 19980
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.13971136510372162,
      "learning_rate": 1.2475061170713345e-05,
      "loss": 0.1336,
      "step": 19990
    },
    {
      "epoch": 3.76,
      "grad_norm": 0.05513539910316467,
      "learning_rate": 1.2471296819122906e-05,
      "loss": 0.233,
      "step": 20000
    },
    {
      "epoch": 3.77,
      "grad_norm": 2.627864360809326,
      "learning_rate": 1.2467532467532468e-05,
      "loss": 0.2722,
      "step": 20010
    },
    {
      "epoch": 3.77,
      "grad_norm": 28.602386474609375,
      "learning_rate": 1.2463768115942029e-05,
      "loss": 0.3108,
      "step": 20020
    },
    {
      "epoch": 3.77,
      "grad_norm": 5.80349063873291,
      "learning_rate": 1.2460003764351592e-05,
      "loss": 0.4041,
      "step": 20030
    },
    {
      "epoch": 3.77,
      "grad_norm": 13.325228691101074,
      "learning_rate": 1.2456239412761154e-05,
      "loss": 0.4849,
      "step": 20040
    },
    {
      "epoch": 3.77,
      "grad_norm": 6.453797817230225,
      "learning_rate": 1.2452475061170715e-05,
      "loss": 0.2502,
      "step": 20050
    },
    {
      "epoch": 3.78,
      "grad_norm": 19.60500144958496,
      "learning_rate": 1.2448710709580277e-05,
      "loss": 0.3656,
      "step": 20060
    },
    {
      "epoch": 3.78,
      "grad_norm": 2.4736807346343994,
      "learning_rate": 1.2444946357989838e-05,
      "loss": 0.2007,
      "step": 20070
    },
    {
      "epoch": 3.78,
      "grad_norm": 16.191587448120117,
      "learning_rate": 1.24411820063994e-05,
      "loss": 0.148,
      "step": 20080
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.07293638586997986,
      "learning_rate": 1.2437417654808961e-05,
      "loss": 0.184,
      "step": 20090
    },
    {
      "epoch": 3.78,
      "grad_norm": 0.013387536630034447,
      "learning_rate": 1.2433653303218522e-05,
      "loss": 0.3137,
      "step": 20100
    },
    {
      "epoch": 3.79,
      "grad_norm": 8.764649391174316,
      "learning_rate": 1.2429888951628084e-05,
      "loss": 0.1392,
      "step": 20110
    },
    {
      "epoch": 3.79,
      "grad_norm": 2.136277914047241,
      "learning_rate": 1.2426124600037644e-05,
      "loss": 0.1447,
      "step": 20120
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.6819177865982056,
      "learning_rate": 1.2422360248447205e-05,
      "loss": 0.2467,
      "step": 20130
    },
    {
      "epoch": 3.79,
      "grad_norm": 1.2957944869995117,
      "learning_rate": 1.2418595896856767e-05,
      "loss": 0.2476,
      "step": 20140
    },
    {
      "epoch": 3.79,
      "grad_norm": 20.906267166137695,
      "learning_rate": 1.2414831545266328e-05,
      "loss": 0.2737,
      "step": 20150
    },
    {
      "epoch": 3.79,
      "grad_norm": 3.3025755882263184,
      "learning_rate": 1.241106719367589e-05,
      "loss": 0.125,
      "step": 20160
    },
    {
      "epoch": 3.8,
      "grad_norm": 13.586214065551758,
      "learning_rate": 1.2407302842085451e-05,
      "loss": 0.2207,
      "step": 20170
    },
    {
      "epoch": 3.8,
      "grad_norm": 10.193732261657715,
      "learning_rate": 1.2403538490495012e-05,
      "loss": 0.1816,
      "step": 20180
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.14643257856369019,
      "learning_rate": 1.2399774138904574e-05,
      "loss": 0.3727,
      "step": 20190
    },
    {
      "epoch": 3.8,
      "grad_norm": 10.179524421691895,
      "learning_rate": 1.2396009787314135e-05,
      "loss": 0.4217,
      "step": 20200
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.08736412972211838,
      "learning_rate": 1.2392245435723698e-05,
      "loss": 0.1694,
      "step": 20210
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.021829910576343536,
      "learning_rate": 1.238848108413326e-05,
      "loss": 0.1736,
      "step": 20220
    },
    {
      "epoch": 3.81,
      "grad_norm": 14.177165031433105,
      "learning_rate": 1.2384716732542821e-05,
      "loss": 0.3512,
      "step": 20230
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.5568282604217529,
      "learning_rate": 1.2380952380952383e-05,
      "loss": 0.2266,
      "step": 20240
    },
    {
      "epoch": 3.81,
      "grad_norm": 3.7102112770080566,
      "learning_rate": 1.2377188029361944e-05,
      "loss": 0.1934,
      "step": 20250
    },
    {
      "epoch": 3.81,
      "grad_norm": 0.01905328407883644,
      "learning_rate": 1.2373423677771506e-05,
      "loss": 0.2398,
      "step": 20260
    },
    {
      "epoch": 3.82,
      "grad_norm": 7.7666707038879395,
      "learning_rate": 1.2369659326181067e-05,
      "loss": 0.1751,
      "step": 20270
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.038244348019361496,
      "learning_rate": 1.2365894974590629e-05,
      "loss": 0.1592,
      "step": 20280
    },
    {
      "epoch": 3.82,
      "grad_norm": 0.03324722871184349,
      "learning_rate": 1.236213062300019e-05,
      "loss": 0.0683,
      "step": 20290
    },
    {
      "epoch": 3.82,
      "grad_norm": 28.736143112182617,
      "learning_rate": 1.235836627140975e-05,
      "loss": 0.1295,
      "step": 20300
    },
    {
      "epoch": 3.82,
      "grad_norm": 10.538153648376465,
      "learning_rate": 1.2354601919819311e-05,
      "loss": 0.1338,
      "step": 20310
    },
    {
      "epoch": 3.82,
      "grad_norm": 11.442151069641113,
      "learning_rate": 1.2350837568228873e-05,
      "loss": 0.1683,
      "step": 20320
    },
    {
      "epoch": 3.83,
      "grad_norm": 22.82421112060547,
      "learning_rate": 1.2347073216638434e-05,
      "loss": 0.2892,
      "step": 20330
    },
    {
      "epoch": 3.83,
      "grad_norm": 4.035543918609619,
      "learning_rate": 1.2343308865047996e-05,
      "loss": 0.2665,
      "step": 20340
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.06633714586496353,
      "learning_rate": 1.2339544513457557e-05,
      "loss": 0.1393,
      "step": 20350
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.3337404131889343,
      "learning_rate": 1.2335780161867119e-05,
      "loss": 0.3548,
      "step": 20360
    },
    {
      "epoch": 3.83,
      "grad_norm": 0.9487078785896301,
      "learning_rate": 1.233201581027668e-05,
      "loss": 0.1779,
      "step": 20370
    },
    {
      "epoch": 3.84,
      "grad_norm": 13.234500885009766,
      "learning_rate": 1.2328251458686242e-05,
      "loss": 0.4221,
      "step": 20380
    },
    {
      "epoch": 3.84,
      "grad_norm": 0.07009129226207733,
      "learning_rate": 1.2324487107095803e-05,
      "loss": 0.2181,
      "step": 20390
    },
    {
      "epoch": 3.84,
      "grad_norm": 5.792457580566406,
      "learning_rate": 1.2320722755505366e-05,
      "loss": 0.192,
      "step": 20400
    },
    {
      "epoch": 3.84,
      "grad_norm": 20.98419189453125,
      "learning_rate": 1.2316958403914928e-05,
      "loss": 0.1477,
      "step": 20410
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.639689564704895,
      "learning_rate": 1.2313194052324489e-05,
      "loss": 0.1892,
      "step": 20420
    },
    {
      "epoch": 3.85,
      "grad_norm": 0.8850065469741821,
      "learning_rate": 1.230942970073405e-05,
      "loss": 0.1757,
      "step": 20430
    },
    {
      "epoch": 3.85,
      "grad_norm": 2.97404408454895,
      "learning_rate": 1.2305665349143612e-05,
      "loss": 0.2498,
      "step": 20440
    },
    {
      "epoch": 3.85,
      "grad_norm": 5.440379619598389,
      "learning_rate": 1.2301900997553174e-05,
      "loss": 0.114,
      "step": 20450
    },
    {
      "epoch": 3.85,
      "grad_norm": 2.4872937202453613,
      "learning_rate": 1.2298136645962735e-05,
      "loss": 0.2016,
      "step": 20460
    },
    {
      "epoch": 3.85,
      "grad_norm": 15.697461128234863,
      "learning_rate": 1.2294372294372295e-05,
      "loss": 0.2133,
      "step": 20470
    },
    {
      "epoch": 3.85,
      "grad_norm": 13.53938102722168,
      "learning_rate": 1.2290607942781856e-05,
      "loss": 0.2607,
      "step": 20480
    },
    {
      "epoch": 3.86,
      "grad_norm": 36.47553253173828,
      "learning_rate": 1.2286843591191418e-05,
      "loss": 0.4989,
      "step": 20490
    },
    {
      "epoch": 3.86,
      "grad_norm": 10.213708877563477,
      "learning_rate": 1.2283079239600979e-05,
      "loss": 0.2653,
      "step": 20500
    },
    {
      "epoch": 3.86,
      "grad_norm": 5.881499767303467,
      "learning_rate": 1.227931488801054e-05,
      "loss": 0.5525,
      "step": 20510
    },
    {
      "epoch": 3.86,
      "grad_norm": 17.465599060058594,
      "learning_rate": 1.2275550536420102e-05,
      "loss": 0.2786,
      "step": 20520
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.015777114778757095,
      "learning_rate": 1.2271786184829664e-05,
      "loss": 0.3084,
      "step": 20530
    },
    {
      "epoch": 3.87,
      "grad_norm": 14.611968040466309,
      "learning_rate": 1.2268021833239225e-05,
      "loss": 0.2673,
      "step": 20540
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.019662246108055115,
      "learning_rate": 1.2264257481648786e-05,
      "loss": 0.4029,
      "step": 20550
    },
    {
      "epoch": 3.87,
      "grad_norm": 4.351469039916992,
      "learning_rate": 1.2260493130058348e-05,
      "loss": 0.2717,
      "step": 20560
    },
    {
      "epoch": 3.87,
      "grad_norm": 5.781222820281982,
      "learning_rate": 1.225672877846791e-05,
      "loss": 0.3288,
      "step": 20570
    },
    {
      "epoch": 3.87,
      "grad_norm": 0.3423316478729248,
      "learning_rate": 1.2252964426877473e-05,
      "loss": 0.1642,
      "step": 20580
    },
    {
      "epoch": 3.88,
      "grad_norm": 18.565914154052734,
      "learning_rate": 1.2249200075287034e-05,
      "loss": 0.3323,
      "step": 20590
    },
    {
      "epoch": 3.88,
      "grad_norm": 6.6165266036987305,
      "learning_rate": 1.2245435723696595e-05,
      "loss": 0.2581,
      "step": 20600
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.23904529213905334,
      "learning_rate": 1.2241671372106157e-05,
      "loss": 0.2076,
      "step": 20610
    },
    {
      "epoch": 3.88,
      "grad_norm": 0.009082888253033161,
      "learning_rate": 1.2237907020515718e-05,
      "loss": 0.3663,
      "step": 20620
    },
    {
      "epoch": 3.88,
      "grad_norm": 16.895992279052734,
      "learning_rate": 1.223414266892528e-05,
      "loss": 0.1432,
      "step": 20630
    },
    {
      "epoch": 3.88,
      "grad_norm": 17.645965576171875,
      "learning_rate": 1.2230378317334841e-05,
      "loss": 0.479,
      "step": 20640
    },
    {
      "epoch": 3.89,
      "grad_norm": 16.875394821166992,
      "learning_rate": 1.2226613965744401e-05,
      "loss": 0.3799,
      "step": 20650
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.2799813747406006,
      "learning_rate": 1.2222849614153963e-05,
      "loss": 0.2124,
      "step": 20660
    },
    {
      "epoch": 3.89,
      "grad_norm": 23.067668914794922,
      "learning_rate": 1.2219085262563524e-05,
      "loss": 0.2195,
      "step": 20670
    },
    {
      "epoch": 3.89,
      "grad_norm": 0.025148926302790642,
      "learning_rate": 1.2215320910973085e-05,
      "loss": 0.2675,
      "step": 20680
    },
    {
      "epoch": 3.89,
      "grad_norm": 6.080964088439941,
      "learning_rate": 1.2211556559382647e-05,
      "loss": 0.2592,
      "step": 20690
    },
    {
      "epoch": 3.9,
      "grad_norm": 8.111252784729004,
      "learning_rate": 1.2207792207792208e-05,
      "loss": 0.178,
      "step": 20700
    },
    {
      "epoch": 3.9,
      "grad_norm": 72.01897430419922,
      "learning_rate": 1.220402785620177e-05,
      "loss": 0.2926,
      "step": 20710
    },
    {
      "epoch": 3.9,
      "grad_norm": 11.691640853881836,
      "learning_rate": 1.2200263504611331e-05,
      "loss": 0.1858,
      "step": 20720
    },
    {
      "epoch": 3.9,
      "grad_norm": 22.981218338012695,
      "learning_rate": 1.2196499153020893e-05,
      "loss": 0.4093,
      "step": 20730
    },
    {
      "epoch": 3.9,
      "grad_norm": 7.562507152557373,
      "learning_rate": 1.2192734801430454e-05,
      "loss": 0.1173,
      "step": 20740
    },
    {
      "epoch": 3.91,
      "grad_norm": 2.805633783340454,
      "learning_rate": 1.2188970449840016e-05,
      "loss": 0.1805,
      "step": 20750
    },
    {
      "epoch": 3.91,
      "grad_norm": 2.3204448223114014,
      "learning_rate": 1.2185206098249577e-05,
      "loss": 0.2171,
      "step": 20760
    },
    {
      "epoch": 3.91,
      "grad_norm": 11.684600830078125,
      "learning_rate": 1.218144174665914e-05,
      "loss": 0.3245,
      "step": 20770
    },
    {
      "epoch": 3.91,
      "grad_norm": 2.273707628250122,
      "learning_rate": 1.2177677395068702e-05,
      "loss": 0.1872,
      "step": 20780
    },
    {
      "epoch": 3.91,
      "grad_norm": 5.970222473144531,
      "learning_rate": 1.2173913043478263e-05,
      "loss": 0.3626,
      "step": 20790
    },
    {
      "epoch": 3.91,
      "grad_norm": 8.510573387145996,
      "learning_rate": 1.2170148691887825e-05,
      "loss": 0.2852,
      "step": 20800
    },
    {
      "epoch": 3.92,
      "grad_norm": 3.087899923324585,
      "learning_rate": 1.2166384340297386e-05,
      "loss": 0.3688,
      "step": 20810
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.1901853382587433,
      "learning_rate": 1.2162619988706946e-05,
      "loss": 0.2623,
      "step": 20820
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.24486812949180603,
      "learning_rate": 1.2158855637116507e-05,
      "loss": 0.2693,
      "step": 20830
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.3221144378185272,
      "learning_rate": 1.2155091285526069e-05,
      "loss": 0.1926,
      "step": 20840
    },
    {
      "epoch": 3.92,
      "grad_norm": 0.01102051418274641,
      "learning_rate": 1.215132693393563e-05,
      "loss": 0.2216,
      "step": 20850
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.023544609546661377,
      "learning_rate": 1.2147562582345192e-05,
      "loss": 0.419,
      "step": 20860
    },
    {
      "epoch": 3.93,
      "grad_norm": 33.68996810913086,
      "learning_rate": 1.2143798230754753e-05,
      "loss": 0.211,
      "step": 20870
    },
    {
      "epoch": 3.93,
      "grad_norm": 13.250447273254395,
      "learning_rate": 1.2140033879164315e-05,
      "loss": 0.3039,
      "step": 20880
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.08316286653280258,
      "learning_rate": 1.2136269527573876e-05,
      "loss": 0.3865,
      "step": 20890
    },
    {
      "epoch": 3.93,
      "grad_norm": 4.8279337882995605,
      "learning_rate": 1.2132505175983438e-05,
      "loss": 0.2297,
      "step": 20900
    },
    {
      "epoch": 3.94,
      "grad_norm": 28.902542114257812,
      "learning_rate": 1.2128740824392999e-05,
      "loss": 0.4013,
      "step": 20910
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.2603214979171753,
      "learning_rate": 1.212497647280256e-05,
      "loss": 0.1684,
      "step": 20920
    },
    {
      "epoch": 3.94,
      "grad_norm": 7.9300217628479,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 0.1987,
      "step": 20930
    },
    {
      "epoch": 3.94,
      "grad_norm": 0.4164702296257019,
      "learning_rate": 1.2117447769621683e-05,
      "loss": 0.1475,
      "step": 20940
    },
    {
      "epoch": 3.94,
      "grad_norm": 6.892146110534668,
      "learning_rate": 1.2113683418031245e-05,
      "loss": 0.2744,
      "step": 20950
    },
    {
      "epoch": 3.95,
      "grad_norm": 1.5755751132965088,
      "learning_rate": 1.2109919066440808e-05,
      "loss": 0.1774,
      "step": 20960
    },
    {
      "epoch": 3.95,
      "grad_norm": 22.156152725219727,
      "learning_rate": 1.210615471485037e-05,
      "loss": 0.2115,
      "step": 20970
    },
    {
      "epoch": 3.95,
      "grad_norm": 2.536865711212158,
      "learning_rate": 1.2102390363259931e-05,
      "loss": 0.1533,
      "step": 20980
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.05064632371068001,
      "learning_rate": 1.2098626011669492e-05,
      "loss": 0.0417,
      "step": 20990
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.009289571084082127,
      "learning_rate": 1.2094861660079052e-05,
      "loss": 0.2115,
      "step": 21000
    },
    {
      "epoch": 3.95,
      "grad_norm": 0.011655650101602077,
      "learning_rate": 1.2091097308488614e-05,
      "loss": 0.206,
      "step": 21010
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.13986341655254364,
      "learning_rate": 1.2087332956898175e-05,
      "loss": 0.2076,
      "step": 21020
    },
    {
      "epoch": 3.96,
      "grad_norm": 22.379758834838867,
      "learning_rate": 1.2083568605307737e-05,
      "loss": 0.1975,
      "step": 21030
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.04326295107603073,
      "learning_rate": 1.2079804253717298e-05,
      "loss": 0.1863,
      "step": 21040
    },
    {
      "epoch": 3.96,
      "grad_norm": 0.02726653404533863,
      "learning_rate": 1.207603990212686e-05,
      "loss": 0.1239,
      "step": 21050
    },
    {
      "epoch": 3.96,
      "grad_norm": 19.921545028686523,
      "learning_rate": 1.2072275550536421e-05,
      "loss": 0.4075,
      "step": 21060
    },
    {
      "epoch": 3.97,
      "grad_norm": 8.355828285217285,
      "learning_rate": 1.2068511198945982e-05,
      "loss": 0.1035,
      "step": 21070
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.6684707999229431,
      "learning_rate": 1.2064746847355544e-05,
      "loss": 0.2394,
      "step": 21080
    },
    {
      "epoch": 3.97,
      "grad_norm": 2.7742879390716553,
      "learning_rate": 1.2060982495765105e-05,
      "loss": 0.0795,
      "step": 21090
    },
    {
      "epoch": 3.97,
      "grad_norm": 12.321138381958008,
      "learning_rate": 1.2057218144174667e-05,
      "loss": 0.2183,
      "step": 21100
    },
    {
      "epoch": 3.97,
      "grad_norm": 0.07847824692726135,
      "learning_rate": 1.2053453792584228e-05,
      "loss": 0.1447,
      "step": 21110
    },
    {
      "epoch": 3.98,
      "grad_norm": 16.827484130859375,
      "learning_rate": 1.204968944099379e-05,
      "loss": 0.3889,
      "step": 21120
    },
    {
      "epoch": 3.98,
      "grad_norm": 5.333073616027832,
      "learning_rate": 1.2045925089403351e-05,
      "loss": 0.1176,
      "step": 21130
    },
    {
      "epoch": 3.98,
      "grad_norm": 2.125105857849121,
      "learning_rate": 1.2042160737812914e-05,
      "loss": 0.2434,
      "step": 21140
    },
    {
      "epoch": 3.98,
      "grad_norm": 3.0484583377838135,
      "learning_rate": 1.2038396386222476e-05,
      "loss": 0.2529,
      "step": 21150
    },
    {
      "epoch": 3.98,
      "grad_norm": 6.926855087280273,
      "learning_rate": 1.2034632034632037e-05,
      "loss": 0.1694,
      "step": 21160
    },
    {
      "epoch": 3.98,
      "grad_norm": 5.073517322540283,
      "learning_rate": 1.2030867683041595e-05,
      "loss": 0.0927,
      "step": 21170
    },
    {
      "epoch": 3.99,
      "grad_norm": 5.1270623207092285,
      "learning_rate": 1.2027103331451157e-05,
      "loss": 0.0619,
      "step": 21180
    },
    {
      "epoch": 3.99,
      "grad_norm": 22.070262908935547,
      "learning_rate": 1.202333897986072e-05,
      "loss": 0.4834,
      "step": 21190
    },
    {
      "epoch": 3.99,
      "grad_norm": 1.1617339849472046,
      "learning_rate": 1.2019574628270281e-05,
      "loss": 0.2686,
      "step": 21200
    },
    {
      "epoch": 3.99,
      "grad_norm": 11.449942588806152,
      "learning_rate": 1.2015810276679843e-05,
      "loss": 0.1163,
      "step": 21210
    },
    {
      "epoch": 3.99,
      "grad_norm": 10.040719032287598,
      "learning_rate": 1.2012045925089404e-05,
      "loss": 0.2621,
      "step": 21220
    },
    {
      "epoch": 4.0,
      "grad_norm": 24.085525512695312,
      "learning_rate": 1.2008281573498966e-05,
      "loss": 0.4181,
      "step": 21230
    },
    {
      "epoch": 4.0,
      "grad_norm": 20.60751724243164,
      "learning_rate": 1.2004517221908527e-05,
      "loss": 0.4568,
      "step": 21240
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.10301145166158676,
      "learning_rate": 1.2000752870318089e-05,
      "loss": 0.2047,
      "step": 21250
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9873333333333333,
      "eval_loss": 0.05053049325942993,
      "eval_runtime": 51.2878,
      "eval_samples_per_second": 146.234,
      "eval_steps_per_second": 18.289,
      "step": 21252
    },
    {
      "epoch": 4.0,
      "grad_norm": 11.256328582763672,
      "learning_rate": 1.199698851872765e-05,
      "loss": 0.1127,
      "step": 21260
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.888209819793701,
      "learning_rate": 1.1993224167137212e-05,
      "loss": 0.1102,
      "step": 21270
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.16677512228488922,
      "learning_rate": 1.1989459815546773e-05,
      "loss": 0.1287,
      "step": 21280
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.31754350662231445,
      "learning_rate": 1.1985695463956335e-05,
      "loss": 0.4239,
      "step": 21290
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.13572365045547485,
      "learning_rate": 1.1981931112365896e-05,
      "loss": 0.1791,
      "step": 21300
    },
    {
      "epoch": 4.01,
      "grad_norm": 0.010000101290643215,
      "learning_rate": 1.1978166760775457e-05,
      "loss": 0.1479,
      "step": 21310
    },
    {
      "epoch": 4.01,
      "grad_norm": 13.754838943481445,
      "learning_rate": 1.1974402409185019e-05,
      "loss": 0.2112,
      "step": 21320
    },
    {
      "epoch": 4.01,
      "grad_norm": 12.481016159057617,
      "learning_rate": 1.1970638057594582e-05,
      "loss": 0.4658,
      "step": 21330
    },
    {
      "epoch": 4.02,
      "grad_norm": 3.65488338470459,
      "learning_rate": 1.196687370600414e-05,
      "loss": 0.1942,
      "step": 21340
    },
    {
      "epoch": 4.02,
      "grad_norm": 2.0450339317321777,
      "learning_rate": 1.1963109354413702e-05,
      "loss": 0.2142,
      "step": 21350
    },
    {
      "epoch": 4.02,
      "grad_norm": 11.102145195007324,
      "learning_rate": 1.1959345002823263e-05,
      "loss": 0.397,
      "step": 21360
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.012182439677417278,
      "learning_rate": 1.1955580651232826e-05,
      "loss": 0.1885,
      "step": 21370
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.14489667117595673,
      "learning_rate": 1.1951816299642388e-05,
      "loss": 0.1325,
      "step": 21380
    },
    {
      "epoch": 4.03,
      "grad_norm": 1.2254389524459839,
      "learning_rate": 1.1948051948051949e-05,
      "loss": 0.2208,
      "step": 21390
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.04186956211924553,
      "learning_rate": 1.194428759646151e-05,
      "loss": 0.215,
      "step": 21400
    },
    {
      "epoch": 4.03,
      "grad_norm": 3.3606247901916504,
      "learning_rate": 1.1940523244871072e-05,
      "loss": 0.034,
      "step": 21410
    },
    {
      "epoch": 4.03,
      "grad_norm": 0.11042144894599915,
      "learning_rate": 1.1936758893280634e-05,
      "loss": 0.1612,
      "step": 21420
    },
    {
      "epoch": 4.03,
      "grad_norm": 10.517876625061035,
      "learning_rate": 1.1932994541690195e-05,
      "loss": 0.0921,
      "step": 21430
    },
    {
      "epoch": 4.04,
      "grad_norm": 6.6743011474609375,
      "learning_rate": 1.1929230190099756e-05,
      "loss": 0.3475,
      "step": 21440
    },
    {
      "epoch": 4.04,
      "grad_norm": 29.440778732299805,
      "learning_rate": 1.1925465838509318e-05,
      "loss": 0.2551,
      "step": 21450
    },
    {
      "epoch": 4.04,
      "grad_norm": 19.518089294433594,
      "learning_rate": 1.192170148691888e-05,
      "loss": 0.4252,
      "step": 21460
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.3689042329788208,
      "learning_rate": 1.191793713532844e-05,
      "loss": 0.1287,
      "step": 21470
    },
    {
      "epoch": 4.04,
      "grad_norm": 0.0140572814270854,
      "learning_rate": 1.1914172783738002e-05,
      "loss": 0.2083,
      "step": 21480
    },
    {
      "epoch": 4.04,
      "grad_norm": 2.641885280609131,
      "learning_rate": 1.1910408432147564e-05,
      "loss": 0.3142,
      "step": 21490
    },
    {
      "epoch": 4.05,
      "grad_norm": 7.5029096603393555,
      "learning_rate": 1.1906644080557125e-05,
      "loss": 0.1912,
      "step": 21500
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.026988331228494644,
      "learning_rate": 1.1902879728966688e-05,
      "loss": 0.2279,
      "step": 21510
    },
    {
      "epoch": 4.05,
      "grad_norm": 7.934142589569092,
      "learning_rate": 1.1899115377376246e-05,
      "loss": 0.3418,
      "step": 21520
    },
    {
      "epoch": 4.05,
      "grad_norm": 0.08143796026706696,
      "learning_rate": 1.1895351025785808e-05,
      "loss": 0.1926,
      "step": 21530
    },
    {
      "epoch": 4.05,
      "grad_norm": 15.19783878326416,
      "learning_rate": 1.189158667419537e-05,
      "loss": 0.3903,
      "step": 21540
    },
    {
      "epoch": 4.06,
      "grad_norm": 9.457151412963867,
      "learning_rate": 1.188782232260493e-05,
      "loss": 0.1553,
      "step": 21550
    },
    {
      "epoch": 4.06,
      "grad_norm": 10.047698974609375,
      "learning_rate": 1.1884057971014494e-05,
      "loss": 0.3549,
      "step": 21560
    },
    {
      "epoch": 4.06,
      "grad_norm": 13.964696884155273,
      "learning_rate": 1.1880293619424055e-05,
      "loss": 0.3348,
      "step": 21570
    },
    {
      "epoch": 4.06,
      "grad_norm": 10.343367576599121,
      "learning_rate": 1.1876529267833617e-05,
      "loss": 0.0766,
      "step": 21580
    },
    {
      "epoch": 4.06,
      "grad_norm": 0.11974823474884033,
      "learning_rate": 1.1872764916243178e-05,
      "loss": 0.3364,
      "step": 21590
    },
    {
      "epoch": 4.07,
      "grad_norm": 0.01039473433047533,
      "learning_rate": 1.186900056465274e-05,
      "loss": 0.1258,
      "step": 21600
    },
    {
      "epoch": 4.07,
      "grad_norm": 19.760635375976562,
      "learning_rate": 1.1865236213062301e-05,
      "loss": 0.1935,
      "step": 21610
    },
    {
      "epoch": 4.07,
      "grad_norm": 27.996318817138672,
      "learning_rate": 1.1861471861471863e-05,
      "loss": 0.3171,
      "step": 21620
    },
    {
      "epoch": 4.07,
      "grad_norm": 17.650455474853516,
      "learning_rate": 1.1857707509881424e-05,
      "loss": 0.4528,
      "step": 21630
    },
    {
      "epoch": 4.07,
      "grad_norm": 24.887271881103516,
      "learning_rate": 1.1853943158290986e-05,
      "loss": 0.1737,
      "step": 21640
    },
    {
      "epoch": 4.07,
      "grad_norm": 12.326953887939453,
      "learning_rate": 1.1850178806700547e-05,
      "loss": 0.2303,
      "step": 21650
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.014831902459263802,
      "learning_rate": 1.1846414455110109e-05,
      "loss": 0.0525,
      "step": 21660
    },
    {
      "epoch": 4.08,
      "grad_norm": 6.173110485076904,
      "learning_rate": 1.184265010351967e-05,
      "loss": 0.3749,
      "step": 21670
    },
    {
      "epoch": 4.08,
      "grad_norm": 11.917346954345703,
      "learning_rate": 1.1838885751929231e-05,
      "loss": 0.1496,
      "step": 21680
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.011464470997452736,
      "learning_rate": 1.1835121400338791e-05,
      "loss": 0.0891,
      "step": 21690
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.2203262746334076,
      "learning_rate": 1.1831357048748353e-05,
      "loss": 0.1775,
      "step": 21700
    },
    {
      "epoch": 4.09,
      "grad_norm": 2.9583704471588135,
      "learning_rate": 1.1827592697157914e-05,
      "loss": 0.1128,
      "step": 21710
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.19376824796199799,
      "learning_rate": 1.1823828345567476e-05,
      "loss": 0.1103,
      "step": 21720
    },
    {
      "epoch": 4.09,
      "grad_norm": 0.26481735706329346,
      "learning_rate": 1.1820063993977037e-05,
      "loss": 0.3913,
      "step": 21730
    },
    {
      "epoch": 4.09,
      "grad_norm": 20.049636840820312,
      "learning_rate": 1.18162996423866e-05,
      "loss": 0.1856,
      "step": 21740
    },
    {
      "epoch": 4.09,
      "grad_norm": 11.356122016906738,
      "learning_rate": 1.1812535290796162e-05,
      "loss": 0.2589,
      "step": 21750
    },
    {
      "epoch": 4.1,
      "grad_norm": 7.908446311950684,
      "learning_rate": 1.1808770939205723e-05,
      "loss": 0.2565,
      "step": 21760
    },
    {
      "epoch": 4.1,
      "grad_norm": 23.497894287109375,
      "learning_rate": 1.1805006587615285e-05,
      "loss": 0.3038,
      "step": 21770
    },
    {
      "epoch": 4.1,
      "grad_norm": 5.031545162200928,
      "learning_rate": 1.1801242236024846e-05,
      "loss": 0.2782,
      "step": 21780
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.08537227660417557,
      "learning_rate": 1.1797477884434408e-05,
      "loss": 0.0625,
      "step": 21790
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.015330007299780846,
      "learning_rate": 1.1793713532843969e-05,
      "loss": 0.1351,
      "step": 21800
    },
    {
      "epoch": 4.11,
      "grad_norm": 13.741814613342285,
      "learning_rate": 1.178994918125353e-05,
      "loss": 0.3825,
      "step": 21810
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.3080389499664307,
      "learning_rate": 1.1786184829663092e-05,
      "loss": 0.3208,
      "step": 21820
    },
    {
      "epoch": 4.11,
      "grad_norm": 7.8950018882751465,
      "learning_rate": 1.1782420478072653e-05,
      "loss": 0.534,
      "step": 21830
    },
    {
      "epoch": 4.11,
      "grad_norm": 11.260322570800781,
      "learning_rate": 1.1778656126482215e-05,
      "loss": 0.178,
      "step": 21840
    },
    {
      "epoch": 4.11,
      "grad_norm": 1.8164857625961304,
      "learning_rate": 1.1774891774891776e-05,
      "loss": 0.2529,
      "step": 21850
    },
    {
      "epoch": 4.11,
      "grad_norm": 8.548678398132324,
      "learning_rate": 1.1771127423301338e-05,
      "loss": 0.1477,
      "step": 21860
    },
    {
      "epoch": 4.12,
      "grad_norm": 12.745514869689941,
      "learning_rate": 1.1767363071710898e-05,
      "loss": 0.1484,
      "step": 21870
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.019689926877617836,
      "learning_rate": 1.1763598720120459e-05,
      "loss": 0.1781,
      "step": 21880
    },
    {
      "epoch": 4.12,
      "grad_norm": 7.882682800292969,
      "learning_rate": 1.175983436853002e-05,
      "loss": 0.2001,
      "step": 21890
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.22520756721496582,
      "learning_rate": 1.1756070016939582e-05,
      "loss": 0.241,
      "step": 21900
    },
    {
      "epoch": 4.12,
      "grad_norm": 0.009213916957378387,
      "learning_rate": 1.1752305665349143e-05,
      "loss": 0.0941,
      "step": 21910
    },
    {
      "epoch": 4.13,
      "grad_norm": 13.014591217041016,
      "learning_rate": 1.1748541313758705e-05,
      "loss": 0.2159,
      "step": 21920
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.3739505410194397,
      "learning_rate": 1.1744776962168268e-05,
      "loss": 0.0605,
      "step": 21930
    },
    {
      "epoch": 4.13,
      "grad_norm": 6.506744861602783,
      "learning_rate": 1.174101261057783e-05,
      "loss": 0.3816,
      "step": 21940
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.009016675874590874,
      "learning_rate": 1.1737248258987391e-05,
      "loss": 0.3077,
      "step": 21950
    },
    {
      "epoch": 4.13,
      "grad_norm": 0.027735993266105652,
      "learning_rate": 1.1733483907396952e-05,
      "loss": 0.1403,
      "step": 21960
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.5713468790054321,
      "learning_rate": 1.1729719555806514e-05,
      "loss": 0.1304,
      "step": 21970
    },
    {
      "epoch": 4.14,
      "grad_norm": 27.421262741088867,
      "learning_rate": 1.1725955204216075e-05,
      "loss": 0.2215,
      "step": 21980
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.058197326958179474,
      "learning_rate": 1.1722190852625637e-05,
      "loss": 0.1876,
      "step": 21990
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.14904719591140747,
      "learning_rate": 1.1718426501035198e-05,
      "loss": 0.1809,
      "step": 22000
    },
    {
      "epoch": 4.14,
      "grad_norm": 15.225375175476074,
      "learning_rate": 1.171466214944476e-05,
      "loss": 0.2475,
      "step": 22010
    },
    {
      "epoch": 4.14,
      "grad_norm": 0.37519049644470215,
      "learning_rate": 1.1710897797854321e-05,
      "loss": 0.2713,
      "step": 22020
    },
    {
      "epoch": 4.15,
      "grad_norm": 19.436750411987305,
      "learning_rate": 1.1707133446263883e-05,
      "loss": 0.3249,
      "step": 22030
    },
    {
      "epoch": 4.15,
      "grad_norm": 8.237166404724121,
      "learning_rate": 1.1703369094673442e-05,
      "loss": 0.3546,
      "step": 22040
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.040746577084064484,
      "learning_rate": 1.1699604743083004e-05,
      "loss": 0.1654,
      "step": 22050
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.1160728707909584,
      "learning_rate": 1.1695840391492565e-05,
      "loss": 0.1293,
      "step": 22060
    },
    {
      "epoch": 4.15,
      "grad_norm": 0.041916731745004654,
      "learning_rate": 1.1692076039902127e-05,
      "loss": 0.0764,
      "step": 22070
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.14755280315876007,
      "learning_rate": 1.1688311688311688e-05,
      "loss": 0.1887,
      "step": 22080
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.022658569738268852,
      "learning_rate": 1.168454733672125e-05,
      "loss": 0.1312,
      "step": 22090
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.39802196621894836,
      "learning_rate": 1.1680782985130811e-05,
      "loss": 0.204,
      "step": 22100
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.41548869013786316,
      "learning_rate": 1.1677018633540373e-05,
      "loss": 0.3189,
      "step": 22110
    },
    {
      "epoch": 4.16,
      "grad_norm": 0.07461607456207275,
      "learning_rate": 1.1673254281949936e-05,
      "loss": 0.1407,
      "step": 22120
    },
    {
      "epoch": 4.17,
      "grad_norm": 0.39251115918159485,
      "learning_rate": 1.1669489930359497e-05,
      "loss": 0.1018,
      "step": 22130
    },
    {
      "epoch": 4.17,
      "grad_norm": 2.411945104598999,
      "learning_rate": 1.1665725578769059e-05,
      "loss": 0.1428,
      "step": 22140
    },
    {
      "epoch": 4.17,
      "grad_norm": 4.685790061950684,
      "learning_rate": 1.166196122717862e-05,
      "loss": 0.0784,
      "step": 22150
    },
    {
      "epoch": 4.17,
      "grad_norm": 1.646402359008789,
      "learning_rate": 1.1658196875588182e-05,
      "loss": 0.2148,
      "step": 22160
    },
    {
      "epoch": 4.17,
      "grad_norm": 9.29836654663086,
      "learning_rate": 1.1654432523997743e-05,
      "loss": 0.4604,
      "step": 22170
    },
    {
      "epoch": 4.17,
      "grad_norm": 8.603370666503906,
      "learning_rate": 1.1650668172407305e-05,
      "loss": 0.2148,
      "step": 22180
    },
    {
      "epoch": 4.18,
      "grad_norm": 10.791486740112305,
      "learning_rate": 1.1646903820816866e-05,
      "loss": 0.2261,
      "step": 22190
    },
    {
      "epoch": 4.18,
      "grad_norm": 3.749446392059326,
      "learning_rate": 1.1643139469226427e-05,
      "loss": 0.1284,
      "step": 22200
    },
    {
      "epoch": 4.18,
      "grad_norm": 33.78459167480469,
      "learning_rate": 1.1639375117635989e-05,
      "loss": 0.2135,
      "step": 22210
    },
    {
      "epoch": 4.18,
      "grad_norm": 8.085159301757812,
      "learning_rate": 1.1635610766045549e-05,
      "loss": 0.3421,
      "step": 22220
    },
    {
      "epoch": 4.18,
      "grad_norm": 26.8583984375,
      "learning_rate": 1.163184641445511e-05,
      "loss": 0.1838,
      "step": 22230
    },
    {
      "epoch": 4.19,
      "grad_norm": 7.5462565422058105,
      "learning_rate": 1.1628082062864672e-05,
      "loss": 0.2633,
      "step": 22240
    },
    {
      "epoch": 4.19,
      "grad_norm": 2.241389036178589,
      "learning_rate": 1.1624317711274233e-05,
      "loss": 0.1967,
      "step": 22250
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.01358577236533165,
      "learning_rate": 1.1620553359683795e-05,
      "loss": 0.4087,
      "step": 22260
    },
    {
      "epoch": 4.19,
      "grad_norm": 0.7372399568557739,
      "learning_rate": 1.1616789008093356e-05,
      "loss": 0.2556,
      "step": 22270
    },
    {
      "epoch": 4.19,
      "grad_norm": 8.281350135803223,
      "learning_rate": 1.1613024656502917e-05,
      "loss": 0.1532,
      "step": 22280
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.2123878002166748,
      "learning_rate": 1.1609260304912479e-05,
      "loss": 0.0973,
      "step": 22290
    },
    {
      "epoch": 4.2,
      "grad_norm": 13.586172103881836,
      "learning_rate": 1.1605495953322042e-05,
      "loss": 0.4331,
      "step": 22300
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.2404799461364746,
      "learning_rate": 1.1601731601731604e-05,
      "loss": 0.1868,
      "step": 22310
    },
    {
      "epoch": 4.2,
      "grad_norm": 3.022454261779785,
      "learning_rate": 1.1597967250141165e-05,
      "loss": 0.094,
      "step": 22320
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.047070056200027466,
      "learning_rate": 1.1594202898550726e-05,
      "loss": 0.0895,
      "step": 22330
    },
    {
      "epoch": 4.2,
      "grad_norm": 7.1957173347473145,
      "learning_rate": 1.1590438546960288e-05,
      "loss": 0.1381,
      "step": 22340
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.2083633989095688,
      "learning_rate": 1.158667419536985e-05,
      "loss": 0.1078,
      "step": 22350
    },
    {
      "epoch": 4.21,
      "grad_norm": 2.4297268390655518,
      "learning_rate": 1.158290984377941e-05,
      "loss": 0.1663,
      "step": 22360
    },
    {
      "epoch": 4.21,
      "grad_norm": 4.850368499755859,
      "learning_rate": 1.1579145492188972e-05,
      "loss": 0.3041,
      "step": 22370
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.012373204343020916,
      "learning_rate": 1.1575381140598534e-05,
      "loss": 0.4188,
      "step": 22380
    },
    {
      "epoch": 4.21,
      "grad_norm": 6.824936866760254,
      "learning_rate": 1.1571616789008094e-05,
      "loss": 0.1667,
      "step": 22390
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.8328636884689331,
      "learning_rate": 1.1567852437417655e-05,
      "loss": 0.2576,
      "step": 22400
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.014268619939684868,
      "learning_rate": 1.1564088085827216e-05,
      "loss": 0.1038,
      "step": 22410
    },
    {
      "epoch": 4.22,
      "grad_norm": 23.970003128051758,
      "learning_rate": 1.1560323734236778e-05,
      "loss": 0.2485,
      "step": 22420
    },
    {
      "epoch": 4.22,
      "grad_norm": 23.070219039916992,
      "learning_rate": 1.155655938264634e-05,
      "loss": 0.424,
      "step": 22430
    },
    {
      "epoch": 4.22,
      "grad_norm": 25.870946884155273,
      "learning_rate": 1.15527950310559e-05,
      "loss": 0.1929,
      "step": 22440
    },
    {
      "epoch": 4.23,
      "grad_norm": 7.899535655975342,
      "learning_rate": 1.1549030679465462e-05,
      "loss": 0.2287,
      "step": 22450
    },
    {
      "epoch": 4.23,
      "grad_norm": 9.26675796508789,
      "learning_rate": 1.1545266327875024e-05,
      "loss": 0.0708,
      "step": 22460
    },
    {
      "epoch": 4.23,
      "grad_norm": 6.14299201965332,
      "learning_rate": 1.1541501976284585e-05,
      "loss": 0.238,
      "step": 22470
    },
    {
      "epoch": 4.23,
      "grad_norm": 7.844719409942627,
      "learning_rate": 1.1537737624694147e-05,
      "loss": 0.2155,
      "step": 22480
    },
    {
      "epoch": 4.23,
      "grad_norm": 2.6691994667053223,
      "learning_rate": 1.153397327310371e-05,
      "loss": 0.2154,
      "step": 22490
    },
    {
      "epoch": 4.23,
      "grad_norm": 0.007289974018931389,
      "learning_rate": 1.1530208921513271e-05,
      "loss": 0.2287,
      "step": 22500
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.08975914120674133,
      "learning_rate": 1.1526444569922833e-05,
      "loss": 0.1287,
      "step": 22510
    },
    {
      "epoch": 4.24,
      "grad_norm": 11.788939476013184,
      "learning_rate": 1.1522680218332394e-05,
      "loss": 0.2224,
      "step": 22520
    },
    {
      "epoch": 4.24,
      "grad_norm": 10.91174602508545,
      "learning_rate": 1.1518915866741956e-05,
      "loss": 0.3656,
      "step": 22530
    },
    {
      "epoch": 4.24,
      "grad_norm": 0.3714274764060974,
      "learning_rate": 1.1515151515151517e-05,
      "loss": 0.1501,
      "step": 22540
    },
    {
      "epoch": 4.24,
      "grad_norm": 12.692159652709961,
      "learning_rate": 1.1511387163561079e-05,
      "loss": 0.1593,
      "step": 22550
    },
    {
      "epoch": 4.25,
      "grad_norm": 13.324334144592285,
      "learning_rate": 1.1507622811970638e-05,
      "loss": 0.1825,
      "step": 22560
    },
    {
      "epoch": 4.25,
      "grad_norm": 11.951353073120117,
      "learning_rate": 1.15038584603802e-05,
      "loss": 0.252,
      "step": 22570
    },
    {
      "epoch": 4.25,
      "grad_norm": 10.039237976074219,
      "learning_rate": 1.1500094108789761e-05,
      "loss": 0.1689,
      "step": 22580
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.528128445148468,
      "learning_rate": 1.1496329757199323e-05,
      "loss": 0.2678,
      "step": 22590
    },
    {
      "epoch": 4.25,
      "grad_norm": 0.02286248654127121,
      "learning_rate": 1.1492565405608884e-05,
      "loss": 0.0455,
      "step": 22600
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.11917700618505478,
      "learning_rate": 1.1488801054018446e-05,
      "loss": 0.196,
      "step": 22610
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.4245049059391022,
      "learning_rate": 1.1485036702428007e-05,
      "loss": 0.2152,
      "step": 22620
    },
    {
      "epoch": 4.26,
      "grad_norm": 0.022747408598661423,
      "learning_rate": 1.1481272350837569e-05,
      "loss": 0.2778,
      "step": 22630
    },
    {
      "epoch": 4.26,
      "grad_norm": 9.09815502166748,
      "learning_rate": 1.147750799924713e-05,
      "loss": 0.2059,
      "step": 22640
    },
    {
      "epoch": 4.26,
      "grad_norm": 1.7970017194747925,
      "learning_rate": 1.1473743647656691e-05,
      "loss": 0.2439,
      "step": 22650
    },
    {
      "epoch": 4.27,
      "grad_norm": 10.266678810119629,
      "learning_rate": 1.1469979296066253e-05,
      "loss": 0.2262,
      "step": 22660
    },
    {
      "epoch": 4.27,
      "grad_norm": 20.968494415283203,
      "learning_rate": 1.1466214944475816e-05,
      "loss": 0.3918,
      "step": 22670
    },
    {
      "epoch": 4.27,
      "grad_norm": 6.589321613311768,
      "learning_rate": 1.1462450592885378e-05,
      "loss": 0.3094,
      "step": 22680
    },
    {
      "epoch": 4.27,
      "grad_norm": 29.038026809692383,
      "learning_rate": 1.1458686241294939e-05,
      "loss": 0.1701,
      "step": 22690
    },
    {
      "epoch": 4.27,
      "grad_norm": 7.420811176300049,
      "learning_rate": 1.14549218897045e-05,
      "loss": 0.2701,
      "step": 22700
    },
    {
      "epoch": 4.27,
      "grad_norm": 24.812026977539062,
      "learning_rate": 1.1451157538114062e-05,
      "loss": 0.2729,
      "step": 22710
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.042433302849531174,
      "learning_rate": 1.1447393186523623e-05,
      "loss": 0.1034,
      "step": 22720
    },
    {
      "epoch": 4.28,
      "grad_norm": 9.062884330749512,
      "learning_rate": 1.1443628834933185e-05,
      "loss": 0.1808,
      "step": 22730
    },
    {
      "epoch": 4.28,
      "grad_norm": 4.233896255493164,
      "learning_rate": 1.1439864483342745e-05,
      "loss": 0.1075,
      "step": 22740
    },
    {
      "epoch": 4.28,
      "grad_norm": 18.584291458129883,
      "learning_rate": 1.1436100131752306e-05,
      "loss": 0.3349,
      "step": 22750
    },
    {
      "epoch": 4.28,
      "grad_norm": 1.2275185585021973,
      "learning_rate": 1.1432335780161868e-05,
      "loss": 0.2436,
      "step": 22760
    },
    {
      "epoch": 4.29,
      "grad_norm": 0.6953242421150208,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 0.0775,
      "step": 22770
    },
    {
      "epoch": 4.29,
      "grad_norm": 9.839485168457031,
      "learning_rate": 1.142480707698099e-05,
      "loss": 0.2646,
      "step": 22780
    },
    {
      "epoch": 4.29,
      "grad_norm": 5.040310859680176,
      "learning_rate": 1.1421042725390552e-05,
      "loss": 0.215,
      "step": 22790
    },
    {
      "epoch": 4.29,
      "grad_norm": 10.868447303771973,
      "learning_rate": 1.1417278373800113e-05,
      "loss": 0.1511,
      "step": 22800
    },
    {
      "epoch": 4.29,
      "grad_norm": 27.953033447265625,
      "learning_rate": 1.1413514022209675e-05,
      "loss": 0.1996,
      "step": 22810
    },
    {
      "epoch": 4.3,
      "grad_norm": 27.858957290649414,
      "learning_rate": 1.1409749670619236e-05,
      "loss": 0.1903,
      "step": 22820
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.2206966131925583,
      "learning_rate": 1.1405985319028798e-05,
      "loss": 0.188,
      "step": 22830
    },
    {
      "epoch": 4.3,
      "grad_norm": 5.096489906311035,
      "learning_rate": 1.140222096743836e-05,
      "loss": 0.2266,
      "step": 22840
    },
    {
      "epoch": 4.3,
      "grad_norm": 26.24561882019043,
      "learning_rate": 1.139845661584792e-05,
      "loss": 0.2195,
      "step": 22850
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.016234176233410835,
      "learning_rate": 1.1394692264257484e-05,
      "loss": 0.2044,
      "step": 22860
    },
    {
      "epoch": 4.3,
      "grad_norm": 8.599238395690918,
      "learning_rate": 1.1390927912667045e-05,
      "loss": 0.3794,
      "step": 22870
    },
    {
      "epoch": 4.31,
      "grad_norm": 7.247493743896484,
      "learning_rate": 1.1387163561076607e-05,
      "loss": 0.1274,
      "step": 22880
    },
    {
      "epoch": 4.31,
      "grad_norm": 12.700387001037598,
      "learning_rate": 1.1383399209486168e-05,
      "loss": 0.3425,
      "step": 22890
    },
    {
      "epoch": 4.31,
      "grad_norm": 2.804699420928955,
      "learning_rate": 1.137963485789573e-05,
      "loss": 0.2704,
      "step": 22900
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.11394985765218735,
      "learning_rate": 1.137587050630529e-05,
      "loss": 0.166,
      "step": 22910
    },
    {
      "epoch": 4.31,
      "grad_norm": 0.04534439742565155,
      "learning_rate": 1.1372106154714851e-05,
      "loss": 0.2187,
      "step": 22920
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.922676682472229,
      "learning_rate": 1.1368341803124412e-05,
      "loss": 0.2488,
      "step": 22930
    },
    {
      "epoch": 4.32,
      "grad_norm": 0.7930464148521423,
      "learning_rate": 1.1364577451533974e-05,
      "loss": 0.209,
      "step": 22940
    },
    {
      "epoch": 4.32,
      "grad_norm": 33.50905990600586,
      "learning_rate": 1.1360813099943535e-05,
      "loss": 0.2782,
      "step": 22950
    },
    {
      "epoch": 4.32,
      "grad_norm": 5.41417932510376,
      "learning_rate": 1.1357048748353097e-05,
      "loss": 0.2144,
      "step": 22960
    },
    {
      "epoch": 4.32,
      "grad_norm": 10.081356048583984,
      "learning_rate": 1.1353284396762658e-05,
      "loss": 0.4466,
      "step": 22970
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.13117273151874542,
      "learning_rate": 1.134952004517222e-05,
      "loss": 0.2166,
      "step": 22980
    },
    {
      "epoch": 4.33,
      "grad_norm": 6.683539867401123,
      "learning_rate": 1.1345755693581781e-05,
      "loss": 0.2909,
      "step": 22990
    },
    {
      "epoch": 4.33,
      "grad_norm": 29.220008850097656,
      "learning_rate": 1.1341991341991343e-05,
      "loss": 0.1341,
      "step": 23000
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.3278818726539612,
      "learning_rate": 1.1338226990400904e-05,
      "loss": 0.1759,
      "step": 23010
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.08102814853191376,
      "learning_rate": 1.1334462638810466e-05,
      "loss": 0.1508,
      "step": 23020
    },
    {
      "epoch": 4.33,
      "grad_norm": 0.4836154580116272,
      "learning_rate": 1.1330698287220027e-05,
      "loss": 0.067,
      "step": 23030
    },
    {
      "epoch": 4.34,
      "grad_norm": 2.84322452545166,
      "learning_rate": 1.132693393562959e-05,
      "loss": 0.2907,
      "step": 23040
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.013905525207519531,
      "learning_rate": 1.1323169584039152e-05,
      "loss": 0.2241,
      "step": 23050
    },
    {
      "epoch": 4.34,
      "grad_norm": 1.152024269104004,
      "learning_rate": 1.1319405232448713e-05,
      "loss": 0.1682,
      "step": 23060
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.019293595105409622,
      "learning_rate": 1.1315640880858275e-05,
      "loss": 0.0729,
      "step": 23070
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.01004357635974884,
      "learning_rate": 1.1311876529267836e-05,
      "loss": 0.102,
      "step": 23080
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.1414073258638382,
      "learning_rate": 1.1308112177677396e-05,
      "loss": 0.2586,
      "step": 23090
    },
    {
      "epoch": 4.35,
      "grad_norm": 5.970057964324951,
      "learning_rate": 1.1304347826086957e-05,
      "loss": 0.3111,
      "step": 23100
    },
    {
      "epoch": 4.35,
      "grad_norm": 10.822305679321289,
      "learning_rate": 1.1300583474496519e-05,
      "loss": 0.1584,
      "step": 23110
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.008775382302701473,
      "learning_rate": 1.129681912290608e-05,
      "loss": 0.3109,
      "step": 23120
    },
    {
      "epoch": 4.35,
      "grad_norm": 0.07651323080062866,
      "learning_rate": 1.1293054771315642e-05,
      "loss": 0.2281,
      "step": 23130
    },
    {
      "epoch": 4.36,
      "grad_norm": 14.930325508117676,
      "learning_rate": 1.1289290419725203e-05,
      "loss": 0.2351,
      "step": 23140
    },
    {
      "epoch": 4.36,
      "grad_norm": 3.8356659412384033,
      "learning_rate": 1.1285526068134765e-05,
      "loss": 0.0801,
      "step": 23150
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.039029061794281006,
      "learning_rate": 1.1281761716544326e-05,
      "loss": 0.197,
      "step": 23160
    },
    {
      "epoch": 4.36,
      "grad_norm": 5.243347644805908,
      "learning_rate": 1.1277997364953887e-05,
      "loss": 0.3509,
      "step": 23170
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.05618220940232277,
      "learning_rate": 1.1274233013363449e-05,
      "loss": 0.1678,
      "step": 23180
    },
    {
      "epoch": 4.36,
      "grad_norm": 0.4089544713497162,
      "learning_rate": 1.127046866177301e-05,
      "loss": 0.0991,
      "step": 23190
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.009486131370067596,
      "learning_rate": 1.1266704310182572e-05,
      "loss": 0.1853,
      "step": 23200
    },
    {
      "epoch": 4.37,
      "grad_norm": 0.021975990384817123,
      "learning_rate": 1.1262939958592133e-05,
      "loss": 0.3832,
      "step": 23210
    },
    {
      "epoch": 4.37,
      "grad_norm": 12.18401050567627,
      "learning_rate": 1.1259175607001695e-05,
      "loss": 0.3816,
      "step": 23220
    },
    {
      "epoch": 4.37,
      "grad_norm": 1.4411247968673706,
      "learning_rate": 1.1255411255411258e-05,
      "loss": 0.1172,
      "step": 23230
    },
    {
      "epoch": 4.37,
      "grad_norm": 9.398138999938965,
      "learning_rate": 1.125164690382082e-05,
      "loss": 0.2342,
      "step": 23240
    },
    {
      "epoch": 4.38,
      "grad_norm": 13.337433815002441,
      "learning_rate": 1.124788255223038e-05,
      "loss": 0.3103,
      "step": 23250
    },
    {
      "epoch": 4.38,
      "grad_norm": 14.835237503051758,
      "learning_rate": 1.1244118200639939e-05,
      "loss": 0.3135,
      "step": 23260
    },
    {
      "epoch": 4.38,
      "grad_norm": 0.07300077378749847,
      "learning_rate": 1.1240353849049502e-05,
      "loss": 0.122,
      "step": 23270
    },
    {
      "epoch": 4.38,
      "grad_norm": 2.7908835411071777,
      "learning_rate": 1.1236589497459064e-05,
      "loss": 0.2173,
      "step": 23280
    },
    {
      "epoch": 4.38,
      "grad_norm": 6.906830787658691,
      "learning_rate": 1.1232825145868625e-05,
      "loss": 0.1688,
      "step": 23290
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.30790191888809204,
      "learning_rate": 1.1229060794278186e-05,
      "loss": 0.1867,
      "step": 23300
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.14113818109035492,
      "learning_rate": 1.1225296442687748e-05,
      "loss": 0.1772,
      "step": 23310
    },
    {
      "epoch": 4.39,
      "grad_norm": 1.8497709035873413,
      "learning_rate": 1.122153209109731e-05,
      "loss": 0.0955,
      "step": 23320
    },
    {
      "epoch": 4.39,
      "grad_norm": 16.056926727294922,
      "learning_rate": 1.121776773950687e-05,
      "loss": 0.3389,
      "step": 23330
    },
    {
      "epoch": 4.39,
      "grad_norm": 31.473207473754883,
      "learning_rate": 1.1214003387916432e-05,
      "loss": 0.2744,
      "step": 23340
    },
    {
      "epoch": 4.39,
      "grad_norm": 2.012946605682373,
      "learning_rate": 1.1210239036325994e-05,
      "loss": 0.0563,
      "step": 23350
    },
    {
      "epoch": 4.4,
      "grad_norm": 13.434127807617188,
      "learning_rate": 1.1206474684735555e-05,
      "loss": 0.1972,
      "step": 23360
    },
    {
      "epoch": 4.4,
      "grad_norm": 6.5082807540893555,
      "learning_rate": 1.1202710333145117e-05,
      "loss": 0.0863,
      "step": 23370
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.38314422965049744,
      "learning_rate": 1.1198945981554678e-05,
      "loss": 0.0989,
      "step": 23380
    },
    {
      "epoch": 4.4,
      "grad_norm": 6.983781337738037,
      "learning_rate": 1.119518162996424e-05,
      "loss": 0.4309,
      "step": 23390
    },
    {
      "epoch": 4.4,
      "grad_norm": 0.04084671661257744,
      "learning_rate": 1.1191417278373801e-05,
      "loss": 0.1873,
      "step": 23400
    },
    {
      "epoch": 4.41,
      "grad_norm": 11.806520462036133,
      "learning_rate": 1.1187652926783363e-05,
      "loss": 0.2918,
      "step": 23410
    },
    {
      "epoch": 4.41,
      "grad_norm": 3.0398361682891846,
      "learning_rate": 1.1183888575192926e-05,
      "loss": 0.2693,
      "step": 23420
    },
    {
      "epoch": 4.41,
      "grad_norm": 12.409751892089844,
      "learning_rate": 1.1180124223602484e-05,
      "loss": 0.1467,
      "step": 23430
    },
    {
      "epoch": 4.41,
      "grad_norm": 1.3978066444396973,
      "learning_rate": 1.1176359872012045e-05,
      "loss": 0.1872,
      "step": 23440
    },
    {
      "epoch": 4.41,
      "grad_norm": 6.554567337036133,
      "learning_rate": 1.1172595520421607e-05,
      "loss": 0.1525,
      "step": 23450
    },
    {
      "epoch": 4.42,
      "grad_norm": 9.716166496276855,
      "learning_rate": 1.116883116883117e-05,
      "loss": 0.1977,
      "step": 23460
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.3557051718235016,
      "learning_rate": 1.1165066817240731e-05,
      "loss": 0.2526,
      "step": 23470
    },
    {
      "epoch": 4.42,
      "grad_norm": 9.572507858276367,
      "learning_rate": 1.1161302465650293e-05,
      "loss": 0.2441,
      "step": 23480
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.08872227370738983,
      "learning_rate": 1.1157538114059854e-05,
      "loss": 0.0743,
      "step": 23490
    },
    {
      "epoch": 4.42,
      "grad_norm": 2.496833086013794,
      "learning_rate": 1.1153773762469416e-05,
      "loss": 0.227,
      "step": 23500
    },
    {
      "epoch": 4.42,
      "grad_norm": 1.8722954988479614,
      "learning_rate": 1.1150009410878977e-05,
      "loss": 0.1334,
      "step": 23510
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.04014293849468231,
      "learning_rate": 1.1146245059288539e-05,
      "loss": 0.051,
      "step": 23520
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.016257185488939285,
      "learning_rate": 1.11424807076981e-05,
      "loss": 0.273,
      "step": 23530
    },
    {
      "epoch": 4.43,
      "grad_norm": 0.29893383383750916,
      "learning_rate": 1.1138716356107661e-05,
      "loss": 0.2781,
      "step": 23540
    },
    {
      "epoch": 4.43,
      "grad_norm": 17.516292572021484,
      "learning_rate": 1.1134952004517223e-05,
      "loss": 0.1923,
      "step": 23550
    },
    {
      "epoch": 4.43,
      "grad_norm": 25.359737396240234,
      "learning_rate": 1.1131187652926784e-05,
      "loss": 0.124,
      "step": 23560
    },
    {
      "epoch": 4.44,
      "grad_norm": 0.060373593121767044,
      "learning_rate": 1.1127423301336346e-05,
      "loss": 0.1344,
      "step": 23570
    },
    {
      "epoch": 4.44,
      "grad_norm": 9.290319442749023,
      "learning_rate": 1.1123658949745907e-05,
      "loss": 0.1697,
      "step": 23580
    },
    {
      "epoch": 4.44,
      "grad_norm": 2.055382251739502,
      "learning_rate": 1.1119894598155469e-05,
      "loss": 0.0094,
      "step": 23590
    },
    {
      "epoch": 4.44,
      "grad_norm": 10.825538635253906,
      "learning_rate": 1.1116130246565032e-05,
      "loss": 0.3262,
      "step": 23600
    },
    {
      "epoch": 4.44,
      "grad_norm": 49.500179290771484,
      "learning_rate": 1.111236589497459e-05,
      "loss": 0.302,
      "step": 23610
    },
    {
      "epoch": 4.45,
      "grad_norm": 33.85306167602539,
      "learning_rate": 1.1108601543384151e-05,
      "loss": 0.303,
      "step": 23620
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.055347464978694916,
      "learning_rate": 1.1104837191793713e-05,
      "loss": 0.1266,
      "step": 23630
    },
    {
      "epoch": 4.45,
      "grad_norm": 36.57317352294922,
      "learning_rate": 1.1101072840203274e-05,
      "loss": 0.1434,
      "step": 23640
    },
    {
      "epoch": 4.45,
      "grad_norm": 14.755773544311523,
      "learning_rate": 1.1097308488612838e-05,
      "loss": 0.062,
      "step": 23650
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.24658195674419403,
      "learning_rate": 1.1093544137022399e-05,
      "loss": 0.2062,
      "step": 23660
    },
    {
      "epoch": 4.46,
      "grad_norm": 10.505647659301758,
      "learning_rate": 1.108977978543196e-05,
      "loss": 0.2494,
      "step": 23670
    },
    {
      "epoch": 4.46,
      "grad_norm": 18.995227813720703,
      "learning_rate": 1.1086015433841522e-05,
      "loss": 0.1783,
      "step": 23680
    },
    {
      "epoch": 4.46,
      "grad_norm": 8.232138633728027,
      "learning_rate": 1.1082251082251083e-05,
      "loss": 0.1827,
      "step": 23690
    },
    {
      "epoch": 4.46,
      "grad_norm": 6.295475959777832,
      "learning_rate": 1.1078486730660645e-05,
      "loss": 0.3916,
      "step": 23700
    },
    {
      "epoch": 4.46,
      "grad_norm": 12.453989028930664,
      "learning_rate": 1.1074722379070206e-05,
      "loss": 0.2101,
      "step": 23710
    },
    {
      "epoch": 4.46,
      "grad_norm": 17.880796432495117,
      "learning_rate": 1.1070958027479768e-05,
      "loss": 0.0997,
      "step": 23720
    },
    {
      "epoch": 4.47,
      "grad_norm": 19.19998550415039,
      "learning_rate": 1.106719367588933e-05,
      "loss": 0.2408,
      "step": 23730
    },
    {
      "epoch": 4.47,
      "grad_norm": 9.083600044250488,
      "learning_rate": 1.106342932429889e-05,
      "loss": 0.2782,
      "step": 23740
    },
    {
      "epoch": 4.47,
      "grad_norm": 6.125687122344971,
      "learning_rate": 1.1059664972708452e-05,
      "loss": 0.1476,
      "step": 23750
    },
    {
      "epoch": 4.47,
      "grad_norm": 0.11055906862020493,
      "learning_rate": 1.1055900621118014e-05,
      "loss": 0.1185,
      "step": 23760
    },
    {
      "epoch": 4.47,
      "grad_norm": 10.364335060119629,
      "learning_rate": 1.1052136269527575e-05,
      "loss": 0.1843,
      "step": 23770
    },
    {
      "epoch": 4.48,
      "grad_norm": 0.08239158987998962,
      "learning_rate": 1.1048371917937135e-05,
      "loss": 0.1355,
      "step": 23780
    },
    {
      "epoch": 4.48,
      "grad_norm": 3.4858827590942383,
      "learning_rate": 1.1044607566346696e-05,
      "loss": 0.2227,
      "step": 23790
    },
    {
      "epoch": 4.48,
      "grad_norm": 3.5528597831726074,
      "learning_rate": 1.1040843214756258e-05,
      "loss": 0.2895,
      "step": 23800
    },
    {
      "epoch": 4.48,
      "grad_norm": 9.332303047180176,
      "learning_rate": 1.103707886316582e-05,
      "loss": 0.2068,
      "step": 23810
    },
    {
      "epoch": 4.48,
      "grad_norm": 2.7906715869903564,
      "learning_rate": 1.103331451157538e-05,
      "loss": 0.2302,
      "step": 23820
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.00804638396948576,
      "learning_rate": 1.1029550159984944e-05,
      "loss": 0.1025,
      "step": 23830
    },
    {
      "epoch": 4.49,
      "grad_norm": 5.037013530731201,
      "learning_rate": 1.1025785808394505e-05,
      "loss": 0.2815,
      "step": 23840
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.15488336980342865,
      "learning_rate": 1.1022021456804067e-05,
      "loss": 0.139,
      "step": 23850
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.09613378345966339,
      "learning_rate": 1.1018257105213628e-05,
      "loss": 0.2258,
      "step": 23860
    },
    {
      "epoch": 4.49,
      "grad_norm": 24.697589874267578,
      "learning_rate": 1.101449275362319e-05,
      "loss": 0.186,
      "step": 23870
    },
    {
      "epoch": 4.49,
      "grad_norm": 0.02501186914741993,
      "learning_rate": 1.1010728402032751e-05,
      "loss": 0.2805,
      "step": 23880
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.1554475724697113,
      "learning_rate": 1.1006964050442313e-05,
      "loss": 0.0604,
      "step": 23890
    },
    {
      "epoch": 4.5,
      "grad_norm": 10.541495323181152,
      "learning_rate": 1.1003199698851874e-05,
      "loss": 0.2192,
      "step": 23900
    },
    {
      "epoch": 4.5,
      "grad_norm": 25.92186164855957,
      "learning_rate": 1.0999435347261436e-05,
      "loss": 0.2024,
      "step": 23910
    },
    {
      "epoch": 4.5,
      "grad_norm": 8.577042579650879,
      "learning_rate": 1.0995670995670997e-05,
      "loss": 0.3344,
      "step": 23920
    },
    {
      "epoch": 4.5,
      "grad_norm": 0.621545135974884,
      "learning_rate": 1.0991906644080558e-05,
      "loss": 0.2475,
      "step": 23930
    },
    {
      "epoch": 4.51,
      "grad_norm": 6.132960319519043,
      "learning_rate": 1.098814229249012e-05,
      "loss": 0.1284,
      "step": 23940
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.1436023712158203,
      "learning_rate": 1.0984377940899681e-05,
      "loss": 0.2287,
      "step": 23950
    },
    {
      "epoch": 4.51,
      "grad_norm": 0.011295100674033165,
      "learning_rate": 1.0980613589309241e-05,
      "loss": 0.1911,
      "step": 23960
    },
    {
      "epoch": 4.51,
      "grad_norm": 14.550305366516113,
      "learning_rate": 1.0976849237718803e-05,
      "loss": 0.2653,
      "step": 23970
    },
    {
      "epoch": 4.51,
      "grad_norm": 19.702686309814453,
      "learning_rate": 1.0973084886128364e-05,
      "loss": 0.2879,
      "step": 23980
    },
    {
      "epoch": 4.52,
      "grad_norm": 11.596700668334961,
      "learning_rate": 1.0969320534537926e-05,
      "loss": 0.5052,
      "step": 23990
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.1243089884519577,
      "learning_rate": 1.0965556182947487e-05,
      "loss": 0.1799,
      "step": 24000
    },
    {
      "epoch": 4.52,
      "grad_norm": 6.212554454803467,
      "learning_rate": 1.0961791831357048e-05,
      "loss": 0.1988,
      "step": 24010
    },
    {
      "epoch": 4.52,
      "grad_norm": 6.038482189178467,
      "learning_rate": 1.0958027479766612e-05,
      "loss": 0.3562,
      "step": 24020
    },
    {
      "epoch": 4.52,
      "grad_norm": 0.0440446138381958,
      "learning_rate": 1.0954263128176173e-05,
      "loss": 0.2838,
      "step": 24030
    },
    {
      "epoch": 4.52,
      "grad_norm": 3.4879238605499268,
      "learning_rate": 1.0950498776585735e-05,
      "loss": 0.1837,
      "step": 24040
    },
    {
      "epoch": 4.53,
      "grad_norm": 2.8905112743377686,
      "learning_rate": 1.0946734424995296e-05,
      "loss": 0.0584,
      "step": 24050
    },
    {
      "epoch": 4.53,
      "grad_norm": 20.633892059326172,
      "learning_rate": 1.0942970073404857e-05,
      "loss": 0.3264,
      "step": 24060
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.030348876491189003,
      "learning_rate": 1.0939205721814419e-05,
      "loss": 0.284,
      "step": 24070
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.2624492049217224,
      "learning_rate": 1.093544137022398e-05,
      "loss": 0.3051,
      "step": 24080
    },
    {
      "epoch": 4.53,
      "grad_norm": 0.009326426312327385,
      "learning_rate": 1.0931677018633542e-05,
      "loss": 0.0771,
      "step": 24090
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.014700816944241524,
      "learning_rate": 1.0927912667043103e-05,
      "loss": 0.1901,
      "step": 24100
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.057867903262376785,
      "learning_rate": 1.0924148315452665e-05,
      "loss": 0.1467,
      "step": 24110
    },
    {
      "epoch": 4.54,
      "grad_norm": 2.9014620780944824,
      "learning_rate": 1.0920383963862226e-05,
      "loss": 0.3516,
      "step": 24120
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.01991095021367073,
      "learning_rate": 1.0916619612271786e-05,
      "loss": 0.1087,
      "step": 24130
    },
    {
      "epoch": 4.54,
      "grad_norm": 0.27069026231765747,
      "learning_rate": 1.0912855260681347e-05,
      "loss": 0.281,
      "step": 24140
    },
    {
      "epoch": 4.55,
      "grad_norm": 5.222668170928955,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 0.4053,
      "step": 24150
    },
    {
      "epoch": 4.55,
      "grad_norm": 9.11081314086914,
      "learning_rate": 1.090532655750047e-05,
      "loss": 0.4558,
      "step": 24160
    },
    {
      "epoch": 4.55,
      "grad_norm": 10.640218734741211,
      "learning_rate": 1.0901562205910032e-05,
      "loss": 0.1701,
      "step": 24170
    },
    {
      "epoch": 4.55,
      "grad_norm": 33.777278900146484,
      "learning_rate": 1.0897797854319593e-05,
      "loss": 0.1742,
      "step": 24180
    },
    {
      "epoch": 4.55,
      "grad_norm": 3.164998769760132,
      "learning_rate": 1.0894033502729155e-05,
      "loss": 0.1999,
      "step": 24190
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.012437891215085983,
      "learning_rate": 1.0890269151138718e-05,
      "loss": 0.5403,
      "step": 24200
    },
    {
      "epoch": 4.56,
      "grad_norm": 29.227174758911133,
      "learning_rate": 1.088650479954828e-05,
      "loss": 0.1444,
      "step": 24210
    },
    {
      "epoch": 4.56,
      "grad_norm": 22.57036781311035,
      "learning_rate": 1.088274044795784e-05,
      "loss": 0.3008,
      "step": 24220
    },
    {
      "epoch": 4.56,
      "grad_norm": 0.2972855567932129,
      "learning_rate": 1.0878976096367402e-05,
      "loss": 0.2116,
      "step": 24230
    },
    {
      "epoch": 4.56,
      "grad_norm": 3.518953561782837,
      "learning_rate": 1.0875211744776964e-05,
      "loss": 0.168,
      "step": 24240
    },
    {
      "epoch": 4.56,
      "grad_norm": 6.8473100662231445,
      "learning_rate": 1.0871447393186525e-05,
      "loss": 0.1822,
      "step": 24250
    },
    {
      "epoch": 4.57,
      "grad_norm": 10.95061206817627,
      "learning_rate": 1.0867683041596087e-05,
      "loss": 0.2443,
      "step": 24260
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.10024960339069366,
      "learning_rate": 1.0863918690005648e-05,
      "loss": 0.1898,
      "step": 24270
    },
    {
      "epoch": 4.57,
      "grad_norm": 6.828205585479736,
      "learning_rate": 1.086015433841521e-05,
      "loss": 0.3355,
      "step": 24280
    },
    {
      "epoch": 4.57,
      "grad_norm": 1.8178046941757202,
      "learning_rate": 1.0856389986824771e-05,
      "loss": 0.1876,
      "step": 24290
    },
    {
      "epoch": 4.57,
      "grad_norm": 0.037091560661792755,
      "learning_rate": 1.0852625635234333e-05,
      "loss": 0.1215,
      "step": 24300
    },
    {
      "epoch": 4.58,
      "grad_norm": 9.223336219787598,
      "learning_rate": 1.0848861283643892e-05,
      "loss": 0.3339,
      "step": 24310
    },
    {
      "epoch": 4.58,
      "grad_norm": 19.945035934448242,
      "learning_rate": 1.0845096932053454e-05,
      "loss": 0.1453,
      "step": 24320
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.37255358695983887,
      "learning_rate": 1.0841332580463015e-05,
      "loss": 0.4249,
      "step": 24330
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.1979893445968628,
      "learning_rate": 1.0837568228872577e-05,
      "loss": 0.243,
      "step": 24340
    },
    {
      "epoch": 4.58,
      "grad_norm": 13.305809020996094,
      "learning_rate": 1.0833803877282138e-05,
      "loss": 0.2164,
      "step": 24350
    },
    {
      "epoch": 4.58,
      "grad_norm": 0.1872040033340454,
      "learning_rate": 1.08300395256917e-05,
      "loss": 0.2111,
      "step": 24360
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.9375900626182556,
      "learning_rate": 1.0826275174101261e-05,
      "loss": 0.1039,
      "step": 24370
    },
    {
      "epoch": 4.59,
      "grad_norm": 3.870849847793579,
      "learning_rate": 1.0822510822510823e-05,
      "loss": 0.2191,
      "step": 24380
    },
    {
      "epoch": 4.59,
      "grad_norm": 2.346743583679199,
      "learning_rate": 1.0818746470920386e-05,
      "loss": 0.226,
      "step": 24390
    },
    {
      "epoch": 4.59,
      "grad_norm": 0.029595842584967613,
      "learning_rate": 1.0814982119329947e-05,
      "loss": 0.1802,
      "step": 24400
    },
    {
      "epoch": 4.59,
      "grad_norm": 13.577653884887695,
      "learning_rate": 1.0811217767739509e-05,
      "loss": 0.4848,
      "step": 24410
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.6157576441764832,
      "learning_rate": 1.080745341614907e-05,
      "loss": 0.2715,
      "step": 24420
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.06574751436710358,
      "learning_rate": 1.0803689064558631e-05,
      "loss": 0.1078,
      "step": 24430
    },
    {
      "epoch": 4.6,
      "grad_norm": 0.5896967053413391,
      "learning_rate": 1.0799924712968193e-05,
      "loss": 0.3351,
      "step": 24440
    },
    {
      "epoch": 4.6,
      "grad_norm": 15.435544967651367,
      "learning_rate": 1.0796160361377754e-05,
      "loss": 0.1635,
      "step": 24450
    },
    {
      "epoch": 4.6,
      "grad_norm": 5.7727508544921875,
      "learning_rate": 1.0792396009787316e-05,
      "loss": 0.2418,
      "step": 24460
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.00785401277244091,
      "learning_rate": 1.0788631658196877e-05,
      "loss": 0.3465,
      "step": 24470
    },
    {
      "epoch": 4.61,
      "grad_norm": 14.369595527648926,
      "learning_rate": 1.0784867306606437e-05,
      "loss": 0.1781,
      "step": 24480
    },
    {
      "epoch": 4.61,
      "grad_norm": 229.5497589111328,
      "learning_rate": 1.0781102955015999e-05,
      "loss": 0.1623,
      "step": 24490
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.05957866087555885,
      "learning_rate": 1.077733860342556e-05,
      "loss": 0.155,
      "step": 24500
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.6214449405670166,
      "learning_rate": 1.0773574251835121e-05,
      "loss": 0.3319,
      "step": 24510
    },
    {
      "epoch": 4.62,
      "grad_norm": 10.48415756225586,
      "learning_rate": 1.0769809900244683e-05,
      "loss": 0.2911,
      "step": 24520
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.0801691859960556,
      "learning_rate": 1.0766045548654244e-05,
      "loss": 0.2096,
      "step": 24530
    },
    {
      "epoch": 4.62,
      "grad_norm": 0.7446441054344177,
      "learning_rate": 1.0762281197063806e-05,
      "loss": 0.0894,
      "step": 24540
    },
    {
      "epoch": 4.62,
      "grad_norm": 9.391862869262695,
      "learning_rate": 1.0758516845473367e-05,
      "loss": 0.1638,
      "step": 24550
    },
    {
      "epoch": 4.62,
      "grad_norm": 4.433223247528076,
      "learning_rate": 1.0754752493882929e-05,
      "loss": 0.1869,
      "step": 24560
    },
    {
      "epoch": 4.62,
      "grad_norm": 7.198757648468018,
      "learning_rate": 1.0750988142292492e-05,
      "loss": 0.1436,
      "step": 24570
    },
    {
      "epoch": 4.63,
      "grad_norm": 13.013691902160645,
      "learning_rate": 1.0747223790702053e-05,
      "loss": 0.2416,
      "step": 24580
    },
    {
      "epoch": 4.63,
      "grad_norm": 0.09882364422082901,
      "learning_rate": 1.0743459439111615e-05,
      "loss": 0.076,
      "step": 24590
    },
    {
      "epoch": 4.63,
      "grad_norm": 19.30021095275879,
      "learning_rate": 1.0739695087521176e-05,
      "loss": 0.2422,
      "step": 24600
    },
    {
      "epoch": 4.63,
      "grad_norm": 43.781986236572266,
      "learning_rate": 1.0735930735930738e-05,
      "loss": 0.3456,
      "step": 24610
    },
    {
      "epoch": 4.63,
      "grad_norm": 12.286011695861816,
      "learning_rate": 1.07321663843403e-05,
      "loss": 0.2774,
      "step": 24620
    },
    {
      "epoch": 4.64,
      "grad_norm": 1.8685818910598755,
      "learning_rate": 1.072840203274986e-05,
      "loss": 0.0517,
      "step": 24630
    },
    {
      "epoch": 4.64,
      "grad_norm": 8.74968433380127,
      "learning_rate": 1.0724637681159422e-05,
      "loss": 0.2518,
      "step": 24640
    },
    {
      "epoch": 4.64,
      "grad_norm": 29.713884353637695,
      "learning_rate": 1.0720873329568982e-05,
      "loss": 0.2321,
      "step": 24650
    },
    {
      "epoch": 4.64,
      "grad_norm": 17.7718448638916,
      "learning_rate": 1.0717108977978543e-05,
      "loss": 0.3452,
      "step": 24660
    },
    {
      "epoch": 4.64,
      "grad_norm": 3.187169313430786,
      "learning_rate": 1.0713344626388105e-05,
      "loss": 0.287,
      "step": 24670
    },
    {
      "epoch": 4.65,
      "grad_norm": 11.557219505310059,
      "learning_rate": 1.0709580274797666e-05,
      "loss": 0.1635,
      "step": 24680
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.15217438340187073,
      "learning_rate": 1.0705815923207228e-05,
      "loss": 0.1136,
      "step": 24690
    },
    {
      "epoch": 4.65,
      "grad_norm": 0.0595257543027401,
      "learning_rate": 1.070205157161679e-05,
      "loss": 0.2673,
      "step": 24700
    },
    {
      "epoch": 4.65,
      "grad_norm": 2.5820529460906982,
      "learning_rate": 1.069828722002635e-05,
      "loss": 0.2609,
      "step": 24710
    },
    {
      "epoch": 4.65,
      "grad_norm": 38.47654724121094,
      "learning_rate": 1.0694522868435912e-05,
      "loss": 0.3922,
      "step": 24720
    },
    {
      "epoch": 4.65,
      "grad_norm": 1.1166672706604004,
      "learning_rate": 1.0690758516845474e-05,
      "loss": 0.2918,
      "step": 24730
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.022154349833726883,
      "learning_rate": 1.0686994165255035e-05,
      "loss": 0.0872,
      "step": 24740
    },
    {
      "epoch": 4.66,
      "grad_norm": 15.219449996948242,
      "learning_rate": 1.0683229813664597e-05,
      "loss": 0.3271,
      "step": 24750
    },
    {
      "epoch": 4.66,
      "grad_norm": 11.99293327331543,
      "learning_rate": 1.067946546207416e-05,
      "loss": 0.1998,
      "step": 24760
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.01194020826369524,
      "learning_rate": 1.0675701110483721e-05,
      "loss": 0.1208,
      "step": 24770
    },
    {
      "epoch": 4.66,
      "grad_norm": 3.965404987335205,
      "learning_rate": 1.0671936758893283e-05,
      "loss": 0.1521,
      "step": 24780
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.5957226753234863,
      "learning_rate": 1.0668172407302844e-05,
      "loss": 0.1512,
      "step": 24790
    },
    {
      "epoch": 4.67,
      "grad_norm": 19.849149703979492,
      "learning_rate": 1.0664408055712406e-05,
      "loss": 0.1781,
      "step": 24800
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.01286950334906578,
      "learning_rate": 1.0660643704121967e-05,
      "loss": 0.2147,
      "step": 24810
    },
    {
      "epoch": 4.67,
      "grad_norm": 0.009870591573417187,
      "learning_rate": 1.0656879352531528e-05,
      "loss": 0.2186,
      "step": 24820
    },
    {
      "epoch": 4.67,
      "grad_norm": 19.90217399597168,
      "learning_rate": 1.0653115000941088e-05,
      "loss": 0.3188,
      "step": 24830
    },
    {
      "epoch": 4.68,
      "grad_norm": 26.972253799438477,
      "learning_rate": 1.064935064935065e-05,
      "loss": 0.163,
      "step": 24840
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.009081571362912655,
      "learning_rate": 1.0645586297760211e-05,
      "loss": 0.1319,
      "step": 24850
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.09101490676403046,
      "learning_rate": 1.0641821946169773e-05,
      "loss": 0.0143,
      "step": 24860
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.2634938359260559,
      "learning_rate": 1.0638057594579334e-05,
      "loss": 0.1272,
      "step": 24870
    },
    {
      "epoch": 4.68,
      "grad_norm": 0.7254877090454102,
      "learning_rate": 1.0634293242988896e-05,
      "loss": 0.2084,
      "step": 24880
    },
    {
      "epoch": 4.68,
      "grad_norm": 1.072479248046875,
      "learning_rate": 1.0630528891398457e-05,
      "loss": 0.1578,
      "step": 24890
    },
    {
      "epoch": 4.69,
      "grad_norm": 10.378267288208008,
      "learning_rate": 1.0626764539808018e-05,
      "loss": 0.234,
      "step": 24900
    },
    {
      "epoch": 4.69,
      "grad_norm": 1.0366445779800415,
      "learning_rate": 1.062300018821758e-05,
      "loss": 0.1024,
      "step": 24910
    },
    {
      "epoch": 4.69,
      "grad_norm": 15.672382354736328,
      "learning_rate": 1.0619235836627141e-05,
      "loss": 0.2245,
      "step": 24920
    },
    {
      "epoch": 4.69,
      "grad_norm": 11.247196197509766,
      "learning_rate": 1.0615471485036703e-05,
      "loss": 0.4058,
      "step": 24930
    },
    {
      "epoch": 4.69,
      "grad_norm": 0.07503703981637955,
      "learning_rate": 1.0611707133446264e-05,
      "loss": 0.2604,
      "step": 24940
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.09371542185544968,
      "learning_rate": 1.0607942781855827e-05,
      "loss": 0.2718,
      "step": 24950
    },
    {
      "epoch": 4.7,
      "grad_norm": 9.894743919372559,
      "learning_rate": 1.0604178430265389e-05,
      "loss": 0.176,
      "step": 24960
    },
    {
      "epoch": 4.7,
      "grad_norm": 33.84804916381836,
      "learning_rate": 1.060041407867495e-05,
      "loss": 0.2395,
      "step": 24970
    },
    {
      "epoch": 4.7,
      "grad_norm": 2.7390079498291016,
      "learning_rate": 1.0596649727084512e-05,
      "loss": 0.1421,
      "step": 24980
    },
    {
      "epoch": 4.7,
      "grad_norm": 3.079441547393799,
      "learning_rate": 1.0592885375494073e-05,
      "loss": 0.197,
      "step": 24990
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.009812173433601856,
      "learning_rate": 1.0589121023903633e-05,
      "loss": 0.2729,
      "step": 25000
    },
    {
      "epoch": 4.71,
      "grad_norm": 0.05104930326342583,
      "learning_rate": 1.0585356672313195e-05,
      "loss": 0.3048,
      "step": 25010
    },
    {
      "epoch": 4.71,
      "grad_norm": 10.214625358581543,
      "learning_rate": 1.0581592320722756e-05,
      "loss": 0.4729,
      "step": 25020
    },
    {
      "epoch": 4.71,
      "grad_norm": 1.1645557880401611,
      "learning_rate": 1.0577827969132317e-05,
      "loss": 0.1095,
      "step": 25030
    },
    {
      "epoch": 4.71,
      "grad_norm": 4.782618045806885,
      "learning_rate": 1.0574063617541879e-05,
      "loss": 0.1773,
      "step": 25040
    },
    {
      "epoch": 4.71,
      "grad_norm": 7.9820556640625,
      "learning_rate": 1.057029926595144e-05,
      "loss": 0.131,
      "step": 25050
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.014314346015453339,
      "learning_rate": 1.0566534914361002e-05,
      "loss": 0.3976,
      "step": 25060
    },
    {
      "epoch": 4.72,
      "grad_norm": 11.038930892944336,
      "learning_rate": 1.0562770562770563e-05,
      "loss": 0.379,
      "step": 25070
    },
    {
      "epoch": 4.72,
      "grad_norm": 0.05928812921047211,
      "learning_rate": 1.0559006211180125e-05,
      "loss": 0.3612,
      "step": 25080
    },
    {
      "epoch": 4.72,
      "grad_norm": 9.051021575927734,
      "learning_rate": 1.0555241859589686e-05,
      "loss": 0.1213,
      "step": 25090
    },
    {
      "epoch": 4.72,
      "grad_norm": 12.407479286193848,
      "learning_rate": 1.0551477507999248e-05,
      "loss": 0.065,
      "step": 25100
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.02989199012517929,
      "learning_rate": 1.0547713156408809e-05,
      "loss": 0.0298,
      "step": 25110
    },
    {
      "epoch": 4.73,
      "grad_norm": 4.557281970977783,
      "learning_rate": 1.054394880481837e-05,
      "loss": 0.1889,
      "step": 25120
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.2493649274110794,
      "learning_rate": 1.0540184453227934e-05,
      "loss": 0.1717,
      "step": 25130
    },
    {
      "epoch": 4.73,
      "grad_norm": 0.007906606420874596,
      "learning_rate": 1.0536420101637495e-05,
      "loss": 0.2326,
      "step": 25140
    },
    {
      "epoch": 4.73,
      "grad_norm": 1.782894253730774,
      "learning_rate": 1.0532655750047057e-05,
      "loss": 0.1024,
      "step": 25150
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.057912543416023254,
      "learning_rate": 1.0528891398456618e-05,
      "loss": 0.2582,
      "step": 25160
    },
    {
      "epoch": 4.74,
      "grad_norm": 11.297438621520996,
      "learning_rate": 1.052512704686618e-05,
      "loss": 0.2759,
      "step": 25170
    },
    {
      "epoch": 4.74,
      "grad_norm": 2.7628278732299805,
      "learning_rate": 1.052136269527574e-05,
      "loss": 0.3576,
      "step": 25180
    },
    {
      "epoch": 4.74,
      "grad_norm": 10.555049896240234,
      "learning_rate": 1.05175983436853e-05,
      "loss": 0.2708,
      "step": 25190
    },
    {
      "epoch": 4.74,
      "grad_norm": 5.503226280212402,
      "learning_rate": 1.0513833992094862e-05,
      "loss": 0.1974,
      "step": 25200
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.12337111681699753,
      "learning_rate": 1.0510069640504424e-05,
      "loss": 0.09,
      "step": 25210
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.07381723821163177,
      "learning_rate": 1.0506305288913985e-05,
      "loss": 0.2356,
      "step": 25220
    },
    {
      "epoch": 4.75,
      "grad_norm": 25.90159797668457,
      "learning_rate": 1.0502540937323547e-05,
      "loss": 0.2565,
      "step": 25230
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.17993327975273132,
      "learning_rate": 1.0498776585733108e-05,
      "loss": 0.1008,
      "step": 25240
    },
    {
      "epoch": 4.75,
      "grad_norm": 11.493851661682129,
      "learning_rate": 1.049501223414267e-05,
      "loss": 0.4373,
      "step": 25250
    },
    {
      "epoch": 4.75,
      "grad_norm": 0.05524396151304245,
      "learning_rate": 1.0491247882552231e-05,
      "loss": 0.1762,
      "step": 25260
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.8478713035583496,
      "learning_rate": 1.0487483530961793e-05,
      "loss": 0.1837,
      "step": 25270
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.7147059440612793,
      "learning_rate": 1.0483719179371354e-05,
      "loss": 0.2182,
      "step": 25280
    },
    {
      "epoch": 4.76,
      "grad_norm": 9.139369010925293,
      "learning_rate": 1.0479954827780915e-05,
      "loss": 0.179,
      "step": 25290
    },
    {
      "epoch": 4.76,
      "grad_norm": 0.23603707551956177,
      "learning_rate": 1.0476190476190477e-05,
      "loss": 0.2244,
      "step": 25300
    },
    {
      "epoch": 4.76,
      "grad_norm": 27.893936157226562,
      "learning_rate": 1.0472426124600038e-05,
      "loss": 0.2134,
      "step": 25310
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.016559859737753868,
      "learning_rate": 1.0468661773009602e-05,
      "loss": 0.1036,
      "step": 25320
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.027669677510857582,
      "learning_rate": 1.0464897421419163e-05,
      "loss": 0.0916,
      "step": 25330
    },
    {
      "epoch": 4.77,
      "grad_norm": 4.034351825714111,
      "learning_rate": 1.0461133069828724e-05,
      "loss": 0.087,
      "step": 25340
    },
    {
      "epoch": 4.77,
      "grad_norm": 1.3619165420532227,
      "learning_rate": 1.0457368718238282e-05,
      "loss": 0.2699,
      "step": 25350
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.28466182947158813,
      "learning_rate": 1.0453604366647846e-05,
      "loss": 0.0638,
      "step": 25360
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.13293221592903137,
      "learning_rate": 1.0449840015057407e-05,
      "loss": 0.1358,
      "step": 25370
    },
    {
      "epoch": 4.78,
      "grad_norm": 14.62092399597168,
      "learning_rate": 1.0446075663466969e-05,
      "loss": 0.456,
      "step": 25380
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.053989212960004807,
      "learning_rate": 1.044231131187653e-05,
      "loss": 0.1404,
      "step": 25390
    },
    {
      "epoch": 4.78,
      "grad_norm": 6.261455535888672,
      "learning_rate": 1.0438546960286091e-05,
      "loss": 0.4025,
      "step": 25400
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.009555619210004807,
      "learning_rate": 1.0434782608695653e-05,
      "loss": 0.2354,
      "step": 25410
    },
    {
      "epoch": 4.78,
      "grad_norm": 0.024539370089769363,
      "learning_rate": 1.0431018257105214e-05,
      "loss": 0.0747,
      "step": 25420
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.012147245928645134,
      "learning_rate": 1.0427253905514776e-05,
      "loss": 0.1068,
      "step": 25430
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.09999841451644897,
      "learning_rate": 1.0423489553924337e-05,
      "loss": 0.1129,
      "step": 25440
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.9210818409919739,
      "learning_rate": 1.0419725202333899e-05,
      "loss": 0.2383,
      "step": 25450
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.020714355632662773,
      "learning_rate": 1.041596085074346e-05,
      "loss": 0.1199,
      "step": 25460
    },
    {
      "epoch": 4.79,
      "grad_norm": 0.03243391588330269,
      "learning_rate": 1.0412196499153022e-05,
      "loss": 0.26,
      "step": 25470
    },
    {
      "epoch": 4.8,
      "grad_norm": 4.486505031585693,
      "learning_rate": 1.0408432147562583e-05,
      "loss": 0.1558,
      "step": 25480
    },
    {
      "epoch": 4.8,
      "grad_norm": 6.074058532714844,
      "learning_rate": 1.0404667795972145e-05,
      "loss": 0.3927,
      "step": 25490
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.010418667457997799,
      "learning_rate": 1.0400903444381708e-05,
      "loss": 0.153,
      "step": 25500
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.49134689569473267,
      "learning_rate": 1.039713909279127e-05,
      "loss": 0.108,
      "step": 25510
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.1607600450515747,
      "learning_rate": 1.0393374741200827e-05,
      "loss": 0.1805,
      "step": 25520
    },
    {
      "epoch": 4.81,
      "grad_norm": 17.637863159179688,
      "learning_rate": 1.0389610389610389e-05,
      "loss": 0.3653,
      "step": 25530
    },
    {
      "epoch": 4.81,
      "grad_norm": 9.467818260192871,
      "learning_rate": 1.038584603801995e-05,
      "loss": 0.3899,
      "step": 25540
    },
    {
      "epoch": 4.81,
      "grad_norm": 3.0589940547943115,
      "learning_rate": 1.0382081686429513e-05,
      "loss": 0.2005,
      "step": 25550
    },
    {
      "epoch": 4.81,
      "grad_norm": 38.20039749145508,
      "learning_rate": 1.0378317334839075e-05,
      "loss": 0.1632,
      "step": 25560
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.008848601020872593,
      "learning_rate": 1.0374552983248636e-05,
      "loss": 0.1743,
      "step": 25570
    },
    {
      "epoch": 4.81,
      "grad_norm": 0.007705410942435265,
      "learning_rate": 1.0370788631658198e-05,
      "loss": 0.0804,
      "step": 25580
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.23606906831264496,
      "learning_rate": 1.036702428006776e-05,
      "loss": 0.3069,
      "step": 25590
    },
    {
      "epoch": 4.82,
      "grad_norm": 12.146738052368164,
      "learning_rate": 1.036325992847732e-05,
      "loss": 0.2069,
      "step": 25600
    },
    {
      "epoch": 4.82,
      "grad_norm": 19.161128997802734,
      "learning_rate": 1.0359495576886882e-05,
      "loss": 0.2056,
      "step": 25610
    },
    {
      "epoch": 4.82,
      "grad_norm": 1.5885854959487915,
      "learning_rate": 1.0355731225296444e-05,
      "loss": 0.2473,
      "step": 25620
    },
    {
      "epoch": 4.82,
      "grad_norm": 0.09702637791633606,
      "learning_rate": 1.0351966873706005e-05,
      "loss": 0.1346,
      "step": 25630
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.005680921487510204,
      "learning_rate": 1.0348202522115567e-05,
      "loss": 0.244,
      "step": 25640
    },
    {
      "epoch": 4.83,
      "grad_norm": 13.652621269226074,
      "learning_rate": 1.0344438170525128e-05,
      "loss": 0.2597,
      "step": 25650
    },
    {
      "epoch": 4.83,
      "grad_norm": 21.32855224609375,
      "learning_rate": 1.034067381893469e-05,
      "loss": 0.2511,
      "step": 25660
    },
    {
      "epoch": 4.83,
      "grad_norm": 0.14206022024154663,
      "learning_rate": 1.0336909467344251e-05,
      "loss": 0.0523,
      "step": 25670
    },
    {
      "epoch": 4.83,
      "grad_norm": 10.144585609436035,
      "learning_rate": 1.0333145115753812e-05,
      "loss": 0.5012,
      "step": 25680
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.007245467975735664,
      "learning_rate": 1.0329380764163376e-05,
      "loss": 0.0064,
      "step": 25690
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.14499182999134064,
      "learning_rate": 1.0325616412572934e-05,
      "loss": 0.0898,
      "step": 25700
    },
    {
      "epoch": 4.84,
      "grad_norm": 5.461489200592041,
      "learning_rate": 1.0321852060982495e-05,
      "loss": 0.2811,
      "step": 25710
    },
    {
      "epoch": 4.84,
      "grad_norm": 17.594762802124023,
      "learning_rate": 1.0318087709392057e-05,
      "loss": 0.3865,
      "step": 25720
    },
    {
      "epoch": 4.84,
      "grad_norm": 0.059493035078048706,
      "learning_rate": 1.031432335780162e-05,
      "loss": 0.0256,
      "step": 25730
    },
    {
      "epoch": 4.84,
      "grad_norm": 1.8105108737945557,
      "learning_rate": 1.0310559006211181e-05,
      "loss": 0.1255,
      "step": 25740
    },
    {
      "epoch": 4.85,
      "grad_norm": 6.392402172088623,
      "learning_rate": 1.0306794654620743e-05,
      "loss": 0.3998,
      "step": 25750
    },
    {
      "epoch": 4.85,
      "grad_norm": 1.2073583602905273,
      "learning_rate": 1.0303030303030304e-05,
      "loss": 0.1566,
      "step": 25760
    },
    {
      "epoch": 4.85,
      "grad_norm": 11.218798637390137,
      "learning_rate": 1.0299265951439866e-05,
      "loss": 0.289,
      "step": 25770
    },
    {
      "epoch": 4.85,
      "grad_norm": 9.293582916259766,
      "learning_rate": 1.0295501599849427e-05,
      "loss": 0.2006,
      "step": 25780
    },
    {
      "epoch": 4.85,
      "grad_norm": 8.630620956420898,
      "learning_rate": 1.0291737248258988e-05,
      "loss": 0.2569,
      "step": 25790
    },
    {
      "epoch": 4.86,
      "grad_norm": 12.378422737121582,
      "learning_rate": 1.028797289666855e-05,
      "loss": 0.2268,
      "step": 25800
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.016018223017454147,
      "learning_rate": 1.0284208545078111e-05,
      "loss": 0.2004,
      "step": 25810
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.02995292842388153,
      "learning_rate": 1.0280444193487673e-05,
      "loss": 0.0296,
      "step": 25820
    },
    {
      "epoch": 4.86,
      "grad_norm": 0.03220521658658981,
      "learning_rate": 1.0276679841897234e-05,
      "loss": 0.2461,
      "step": 25830
    },
    {
      "epoch": 4.86,
      "grad_norm": 16.62986946105957,
      "learning_rate": 1.0272915490306796e-05,
      "loss": 0.3053,
      "step": 25840
    },
    {
      "epoch": 4.87,
      "grad_norm": 33.22449493408203,
      "learning_rate": 1.0269151138716357e-05,
      "loss": 0.2714,
      "step": 25850
    },
    {
      "epoch": 4.87,
      "grad_norm": 21.546985626220703,
      "learning_rate": 1.0265386787125919e-05,
      "loss": 0.1972,
      "step": 25860
    },
    {
      "epoch": 4.87,
      "grad_norm": 1.3915764093399048,
      "learning_rate": 1.0261622435535478e-05,
      "loss": 0.2042,
      "step": 25870
    },
    {
      "epoch": 4.87,
      "grad_norm": 6.126638412475586,
      "learning_rate": 1.025785808394504e-05,
      "loss": 0.5296,
      "step": 25880
    },
    {
      "epoch": 4.87,
      "grad_norm": 7.340539455413818,
      "learning_rate": 1.0254093732354601e-05,
      "loss": 0.2042,
      "step": 25890
    },
    {
      "epoch": 4.87,
      "grad_norm": 0.2391154170036316,
      "learning_rate": 1.0250329380764163e-05,
      "loss": 0.0963,
      "step": 25900
    },
    {
      "epoch": 4.88,
      "grad_norm": 17.44920539855957,
      "learning_rate": 1.0246565029173724e-05,
      "loss": 0.1022,
      "step": 25910
    },
    {
      "epoch": 4.88,
      "grad_norm": 6.229788303375244,
      "learning_rate": 1.0242800677583287e-05,
      "loss": 0.21,
      "step": 25920
    },
    {
      "epoch": 4.88,
      "grad_norm": 18.95992088317871,
      "learning_rate": 1.0239036325992849e-05,
      "loss": 0.1576,
      "step": 25930
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.009861952625215054,
      "learning_rate": 1.023527197440241e-05,
      "loss": 0.1158,
      "step": 25940
    },
    {
      "epoch": 4.88,
      "grad_norm": 0.545590877532959,
      "learning_rate": 1.0231507622811972e-05,
      "loss": 0.1932,
      "step": 25950
    },
    {
      "epoch": 4.89,
      "grad_norm": 8.52595329284668,
      "learning_rate": 1.0227743271221533e-05,
      "loss": 0.2013,
      "step": 25960
    },
    {
      "epoch": 4.89,
      "grad_norm": 0.057179082185029984,
      "learning_rate": 1.0223978919631095e-05,
      "loss": 0.2033,
      "step": 25970
    },
    {
      "epoch": 4.89,
      "grad_norm": 2.1691715717315674,
      "learning_rate": 1.0220214568040656e-05,
      "loss": 0.3131,
      "step": 25980
    },
    {
      "epoch": 4.89,
      "grad_norm": 7.432029724121094,
      "learning_rate": 1.0216450216450218e-05,
      "loss": 0.3596,
      "step": 25990
    },
    {
      "epoch": 4.89,
      "grad_norm": 17.12654685974121,
      "learning_rate": 1.0212685864859779e-05,
      "loss": 0.1601,
      "step": 26000
    },
    {
      "epoch": 4.9,
      "grad_norm": 6.129942893981934,
      "learning_rate": 1.020892151326934e-05,
      "loss": 0.0461,
      "step": 26010
    },
    {
      "epoch": 4.9,
      "grad_norm": 1.3694754838943481,
      "learning_rate": 1.0205157161678902e-05,
      "loss": 0.0612,
      "step": 26020
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.448512464761734,
      "learning_rate": 1.0201392810088464e-05,
      "loss": 0.1115,
      "step": 26030
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.008318577893078327,
      "learning_rate": 1.0197628458498025e-05,
      "loss": 0.2333,
      "step": 26040
    },
    {
      "epoch": 4.9,
      "grad_norm": 7.578540325164795,
      "learning_rate": 1.0193864106907585e-05,
      "loss": 0.2741,
      "step": 26050
    },
    {
      "epoch": 4.9,
      "grad_norm": 0.2919299602508545,
      "learning_rate": 1.0190099755317146e-05,
      "loss": 0.1572,
      "step": 26060
    },
    {
      "epoch": 4.91,
      "grad_norm": 7.527616024017334,
      "learning_rate": 1.0186335403726708e-05,
      "loss": 0.123,
      "step": 26070
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.013498072512447834,
      "learning_rate": 1.0182571052136269e-05,
      "loss": 0.0177,
      "step": 26080
    },
    {
      "epoch": 4.91,
      "grad_norm": 0.2069225311279297,
      "learning_rate": 1.017880670054583e-05,
      "loss": 0.1128,
      "step": 26090
    },
    {
      "epoch": 4.91,
      "grad_norm": 5.21842098236084,
      "learning_rate": 1.0175042348955392e-05,
      "loss": 0.3606,
      "step": 26100
    },
    {
      "epoch": 4.91,
      "grad_norm": 23.98952293395996,
      "learning_rate": 1.0171277997364955e-05,
      "loss": 0.246,
      "step": 26110
    },
    {
      "epoch": 4.92,
      "grad_norm": 27.18362808227539,
      "learning_rate": 1.0167513645774517e-05,
      "loss": 0.2323,
      "step": 26120
    },
    {
      "epoch": 4.92,
      "grad_norm": 9.9590425491333,
      "learning_rate": 1.0163749294184078e-05,
      "loss": 0.2232,
      "step": 26130
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.07880811393260956,
      "learning_rate": 1.015998494259364e-05,
      "loss": 0.1344,
      "step": 26140
    },
    {
      "epoch": 4.92,
      "grad_norm": 23.81707000732422,
      "learning_rate": 1.0156220591003201e-05,
      "loss": 0.3432,
      "step": 26150
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.02327106148004532,
      "learning_rate": 1.0152456239412763e-05,
      "loss": 0.1752,
      "step": 26160
    },
    {
      "epoch": 4.93,
      "grad_norm": 4.255736827850342,
      "learning_rate": 1.0148691887822324e-05,
      "loss": 0.2483,
      "step": 26170
    },
    {
      "epoch": 4.93,
      "grad_norm": 11.604669570922852,
      "learning_rate": 1.0144927536231885e-05,
      "loss": 0.3015,
      "step": 26180
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.10873466730117798,
      "learning_rate": 1.0141163184641447e-05,
      "loss": 0.1445,
      "step": 26190
    },
    {
      "epoch": 4.93,
      "grad_norm": 0.08252499997615814,
      "learning_rate": 1.0137398833051008e-05,
      "loss": 0.1755,
      "step": 26200
    },
    {
      "epoch": 4.93,
      "grad_norm": 11.015498161315918,
      "learning_rate": 1.013363448146057e-05,
      "loss": 0.2367,
      "step": 26210
    },
    {
      "epoch": 4.94,
      "grad_norm": 1.1592999696731567,
      "learning_rate": 1.012987012987013e-05,
      "loss": 0.1281,
      "step": 26220
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.27107176184654236,
      "learning_rate": 1.0126105778279691e-05,
      "loss": 0.3434,
      "step": 26230
    },
    {
      "epoch": 4.94,
      "grad_norm": 0.026509586721658707,
      "learning_rate": 1.0122341426689252e-05,
      "loss": 0.0837,
      "step": 26240
    },
    {
      "epoch": 4.94,
      "grad_norm": 12.319701194763184,
      "learning_rate": 1.0118577075098814e-05,
      "loss": 0.4696,
      "step": 26250
    },
    {
      "epoch": 4.94,
      "grad_norm": 2.395874261856079,
      "learning_rate": 1.0114812723508375e-05,
      "loss": 0.362,
      "step": 26260
    },
    {
      "epoch": 4.94,
      "grad_norm": 3.0828981399536133,
      "learning_rate": 1.0111048371917937e-05,
      "loss": 0.2572,
      "step": 26270
    },
    {
      "epoch": 4.95,
      "grad_norm": 9.118378639221191,
      "learning_rate": 1.0107284020327498e-05,
      "loss": 0.2115,
      "step": 26280
    },
    {
      "epoch": 4.95,
      "grad_norm": 7.700536727905273,
      "learning_rate": 1.0103519668737061e-05,
      "loss": 0.291,
      "step": 26290
    },
    {
      "epoch": 4.95,
      "grad_norm": 6.881040096282959,
      "learning_rate": 1.0099755317146623e-05,
      "loss": 0.2927,
      "step": 26300
    },
    {
      "epoch": 4.95,
      "grad_norm": 12.159699440002441,
      "learning_rate": 1.0095990965556184e-05,
      "loss": 0.2579,
      "step": 26310
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.08632916212081909,
      "learning_rate": 1.0092226613965746e-05,
      "loss": 0.2368,
      "step": 26320
    },
    {
      "epoch": 4.96,
      "grad_norm": 10.159748077392578,
      "learning_rate": 1.0088462262375307e-05,
      "loss": 0.1586,
      "step": 26330
    },
    {
      "epoch": 4.96,
      "grad_norm": 13.029118537902832,
      "learning_rate": 1.0084697910784869e-05,
      "loss": 0.0741,
      "step": 26340
    },
    {
      "epoch": 4.96,
      "grad_norm": 13.853853225708008,
      "learning_rate": 1.008093355919443e-05,
      "loss": 0.2624,
      "step": 26350
    },
    {
      "epoch": 4.96,
      "grad_norm": 4.758121013641357,
      "learning_rate": 1.0077169207603992e-05,
      "loss": 0.2503,
      "step": 26360
    },
    {
      "epoch": 4.96,
      "grad_norm": 4.312897682189941,
      "learning_rate": 1.0073404856013553e-05,
      "loss": 0.1365,
      "step": 26370
    },
    {
      "epoch": 4.97,
      "grad_norm": 19.921751022338867,
      "learning_rate": 1.0069640504423115e-05,
      "loss": 0.2726,
      "step": 26380
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.013266388326883316,
      "learning_rate": 1.0065876152832676e-05,
      "loss": 0.203,
      "step": 26390
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.08996579796075821,
      "learning_rate": 1.0062111801242236e-05,
      "loss": 0.1068,
      "step": 26400
    },
    {
      "epoch": 4.97,
      "grad_norm": 1.2653086185455322,
      "learning_rate": 1.0058347449651797e-05,
      "loss": 0.2392,
      "step": 26410
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.14362378418445587,
      "learning_rate": 1.0054583098061359e-05,
      "loss": 0.3355,
      "step": 26420
    },
    {
      "epoch": 4.97,
      "grad_norm": 0.02867070585489273,
      "learning_rate": 1.005081874647092e-05,
      "loss": 0.2582,
      "step": 26430
    },
    {
      "epoch": 4.98,
      "grad_norm": 9.663721084594727,
      "learning_rate": 1.0047054394880482e-05,
      "loss": 0.0884,
      "step": 26440
    },
    {
      "epoch": 4.98,
      "grad_norm": 8.673450469970703,
      "learning_rate": 1.0043290043290043e-05,
      "loss": 0.2128,
      "step": 26450
    },
    {
      "epoch": 4.98,
      "grad_norm": 6.491775035858154,
      "learning_rate": 1.0039525691699605e-05,
      "loss": 0.1542,
      "step": 26460
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.052366629242897034,
      "learning_rate": 1.0035761340109166e-05,
      "loss": 0.3163,
      "step": 26470
    },
    {
      "epoch": 4.98,
      "grad_norm": 4.831406593322754,
      "learning_rate": 1.003199698851873e-05,
      "loss": 0.2368,
      "step": 26480
    },
    {
      "epoch": 4.99,
      "grad_norm": 3.972709894180298,
      "learning_rate": 1.002823263692829e-05,
      "loss": 0.1382,
      "step": 26490
    },
    {
      "epoch": 4.99,
      "grad_norm": 2.6498327255249023,
      "learning_rate": 1.0024468285337852e-05,
      "loss": 0.2958,
      "step": 26500
    },
    {
      "epoch": 4.99,
      "grad_norm": 7.853219509124756,
      "learning_rate": 1.0020703933747414e-05,
      "loss": 0.1701,
      "step": 26510
    },
    {
      "epoch": 4.99,
      "grad_norm": 3.000911235809326,
      "learning_rate": 1.0016939582156975e-05,
      "loss": 0.191,
      "step": 26520
    },
    {
      "epoch": 4.99,
      "grad_norm": 0.012887305580079556,
      "learning_rate": 1.0013175230566537e-05,
      "loss": 0.182,
      "step": 26530
    },
    {
      "epoch": 5.0,
      "grad_norm": 14.602657318115234,
      "learning_rate": 1.0009410878976098e-05,
      "loss": 0.1928,
      "step": 26540
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.021365296095609665,
      "learning_rate": 1.000564652738566e-05,
      "loss": 0.2628,
      "step": 26550
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.007734401151537895,
      "learning_rate": 1.0001882175795221e-05,
      "loss": 0.2532,
      "step": 26560
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9893333333333333,
      "eval_loss": 0.04541068524122238,
      "eval_runtime": 51.1689,
      "eval_samples_per_second": 146.573,
      "eval_steps_per_second": 18.331,
      "step": 26565
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.009335489012300968,
      "learning_rate": 9.998117824204782e-06,
      "loss": 0.1508,
      "step": 26570
    },
    {
      "epoch": 5.0,
      "grad_norm": 38.42567443847656,
      "learning_rate": 9.994353472614344e-06,
      "loss": 0.2035,
      "step": 26580
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.015634404495358467,
      "learning_rate": 9.990589121023905e-06,
      "loss": 0.3505,
      "step": 26590
    },
    {
      "epoch": 5.01,
      "grad_norm": 1.9697314500808716,
      "learning_rate": 9.986824769433467e-06,
      "loss": 0.1516,
      "step": 26600
    },
    {
      "epoch": 5.01,
      "grad_norm": 15.117608070373535,
      "learning_rate": 9.983060417843027e-06,
      "loss": 0.3776,
      "step": 26610
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.01444531511515379,
      "learning_rate": 9.979296066252588e-06,
      "loss": 0.1925,
      "step": 26620
    },
    {
      "epoch": 5.01,
      "grad_norm": 0.4839775264263153,
      "learning_rate": 9.97553171466215e-06,
      "loss": 0.3626,
      "step": 26630
    },
    {
      "epoch": 5.01,
      "grad_norm": 11.297017097473145,
      "learning_rate": 9.971767363071711e-06,
      "loss": 0.1762,
      "step": 26640
    },
    {
      "epoch": 5.02,
      "grad_norm": 27.25918197631836,
      "learning_rate": 9.968003011481272e-06,
      "loss": 0.3497,
      "step": 26650
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.03226303681731224,
      "learning_rate": 9.964238659890836e-06,
      "loss": 0.0086,
      "step": 26660
    },
    {
      "epoch": 5.02,
      "grad_norm": 25.068052291870117,
      "learning_rate": 9.960474308300397e-06,
      "loss": 0.2652,
      "step": 26670
    },
    {
      "epoch": 5.02,
      "grad_norm": 0.6698800921440125,
      "learning_rate": 9.956709956709958e-06,
      "loss": 0.1642,
      "step": 26680
    },
    {
      "epoch": 5.02,
      "grad_norm": 10.702884674072266,
      "learning_rate": 9.95294560511952e-06,
      "loss": 0.318,
      "step": 26690
    },
    {
      "epoch": 5.03,
      "grad_norm": 11.635863304138184,
      "learning_rate": 9.94918125352908e-06,
      "loss": 0.2026,
      "step": 26700
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.018664950504899025,
      "learning_rate": 9.945416901938641e-06,
      "loss": 0.2274,
      "step": 26710
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.03811613470315933,
      "learning_rate": 9.941652550348203e-06,
      "loss": 0.0821,
      "step": 26720
    },
    {
      "epoch": 5.03,
      "grad_norm": 0.008922936394810677,
      "learning_rate": 9.937888198757764e-06,
      "loss": 0.0558,
      "step": 26730
    },
    {
      "epoch": 5.03,
      "grad_norm": 1.9243003129959106,
      "learning_rate": 9.934123847167326e-06,
      "loss": 0.1018,
      "step": 26740
    },
    {
      "epoch": 5.03,
      "grad_norm": 6.63258695602417,
      "learning_rate": 9.930359495576887e-06,
      "loss": 0.3456,
      "step": 26750
    },
    {
      "epoch": 5.04,
      "grad_norm": 9.270735740661621,
      "learning_rate": 9.92659514398645e-06,
      "loss": 0.1176,
      "step": 26760
    },
    {
      "epoch": 5.04,
      "grad_norm": 9.806709289550781,
      "learning_rate": 9.922830792396012e-06,
      "loss": 0.2691,
      "step": 26770
    },
    {
      "epoch": 5.04,
      "grad_norm": 30.470460891723633,
      "learning_rate": 9.919066440805573e-06,
      "loss": 0.3304,
      "step": 26780
    },
    {
      "epoch": 5.04,
      "grad_norm": 27.997514724731445,
      "learning_rate": 9.915302089215133e-06,
      "loss": 0.1514,
      "step": 26790
    },
    {
      "epoch": 5.04,
      "grad_norm": 10.272509574890137,
      "learning_rate": 9.911537737624694e-06,
      "loss": 0.2508,
      "step": 26800
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.30335912108421326,
      "learning_rate": 9.907773386034256e-06,
      "loss": 0.3002,
      "step": 26810
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.3048275411128998,
      "learning_rate": 9.904009034443817e-06,
      "loss": 0.138,
      "step": 26820
    },
    {
      "epoch": 5.05,
      "grad_norm": 4.221545219421387,
      "learning_rate": 9.900244682853379e-06,
      "loss": 0.233,
      "step": 26830
    },
    {
      "epoch": 5.05,
      "grad_norm": 8.116349220275879,
      "learning_rate": 9.89648033126294e-06,
      "loss": 0.1061,
      "step": 26840
    },
    {
      "epoch": 5.05,
      "grad_norm": 3.731807231903076,
      "learning_rate": 9.892715979672503e-06,
      "loss": 0.2508,
      "step": 26850
    },
    {
      "epoch": 5.06,
      "grad_norm": 12.052627563476562,
      "learning_rate": 9.888951628082065e-06,
      "loss": 0.1203,
      "step": 26860
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.7799234390258789,
      "learning_rate": 9.885187276491625e-06,
      "loss": 0.3819,
      "step": 26870
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.011619717814028263,
      "learning_rate": 9.881422924901186e-06,
      "loss": 0.2747,
      "step": 26880
    },
    {
      "epoch": 5.06,
      "grad_norm": 4.210769176483154,
      "learning_rate": 9.877658573310747e-06,
      "loss": 0.3195,
      "step": 26890
    },
    {
      "epoch": 5.06,
      "grad_norm": 12.908836364746094,
      "learning_rate": 9.873894221720309e-06,
      "loss": 0.2461,
      "step": 26900
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.027916645631194115,
      "learning_rate": 9.87012987012987e-06,
      "loss": 0.2426,
      "step": 26910
    },
    {
      "epoch": 5.07,
      "grad_norm": 16.792829513549805,
      "learning_rate": 9.866365518539432e-06,
      "loss": 0.197,
      "step": 26920
    },
    {
      "epoch": 5.07,
      "grad_norm": 26.003755569458008,
      "learning_rate": 9.862601166948993e-06,
      "loss": 0.2761,
      "step": 26930
    },
    {
      "epoch": 5.07,
      "grad_norm": 3.783144474029541,
      "learning_rate": 9.858836815358556e-06,
      "loss": 0.1702,
      "step": 26940
    },
    {
      "epoch": 5.07,
      "grad_norm": 3.190188407897949,
      "learning_rate": 9.855072463768118e-06,
      "loss": 0.2033,
      "step": 26950
    },
    {
      "epoch": 5.07,
      "grad_norm": 0.023463433608412743,
      "learning_rate": 9.851308112177678e-06,
      "loss": 0.258,
      "step": 26960
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.011378780007362366,
      "learning_rate": 9.847543760587239e-06,
      "loss": 0.0805,
      "step": 26970
    },
    {
      "epoch": 5.08,
      "grad_norm": 12.540125846862793,
      "learning_rate": 9.8437794089968e-06,
      "loss": 0.1487,
      "step": 26980
    },
    {
      "epoch": 5.08,
      "grad_norm": 1.9414715766906738,
      "learning_rate": 9.840015057406362e-06,
      "loss": 0.0364,
      "step": 26990
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.21213127672672272,
      "learning_rate": 9.836250705815924e-06,
      "loss": 0.1955,
      "step": 27000
    },
    {
      "epoch": 5.08,
      "grad_norm": 3.988600492477417,
      "learning_rate": 9.832486354225485e-06,
      "loss": 0.2294,
      "step": 27010
    },
    {
      "epoch": 5.09,
      "grad_norm": 8.902695655822754,
      "learning_rate": 9.828722002635046e-06,
      "loss": 0.0749,
      "step": 27020
    },
    {
      "epoch": 5.09,
      "grad_norm": 4.401754856109619,
      "learning_rate": 9.82495765104461e-06,
      "loss": 0.1511,
      "step": 27030
    },
    {
      "epoch": 5.09,
      "grad_norm": 19.244426727294922,
      "learning_rate": 9.821193299454171e-06,
      "loss": 0.2996,
      "step": 27040
    },
    {
      "epoch": 5.09,
      "grad_norm": 14.334905624389648,
      "learning_rate": 9.81742894786373e-06,
      "loss": 0.206,
      "step": 27050
    },
    {
      "epoch": 5.09,
      "grad_norm": 1.2225977182388306,
      "learning_rate": 9.813664596273292e-06,
      "loss": 0.119,
      "step": 27060
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.16205735504627228,
      "learning_rate": 9.809900244682854e-06,
      "loss": 0.1479,
      "step": 27070
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.008554784581065178,
      "learning_rate": 9.806135893092415e-06,
      "loss": 0.0395,
      "step": 27080
    },
    {
      "epoch": 5.1,
      "grad_norm": 24.27212905883789,
      "learning_rate": 9.802371541501977e-06,
      "loss": 0.109,
      "step": 27090
    },
    {
      "epoch": 5.1,
      "grad_norm": 8.759340286254883,
      "learning_rate": 9.798607189911538e-06,
      "loss": 0.1558,
      "step": 27100
    },
    {
      "epoch": 5.1,
      "grad_norm": 0.04053306579589844,
      "learning_rate": 9.7948428383211e-06,
      "loss": 0.2157,
      "step": 27110
    },
    {
      "epoch": 5.1,
      "grad_norm": 12.061295509338379,
      "learning_rate": 9.791078486730661e-06,
      "loss": 0.1332,
      "step": 27120
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.0496448278427124,
      "learning_rate": 9.787314135140224e-06,
      "loss": 0.2847,
      "step": 27130
    },
    {
      "epoch": 5.11,
      "grad_norm": 2.596406936645508,
      "learning_rate": 9.783549783549784e-06,
      "loss": 0.2521,
      "step": 27140
    },
    {
      "epoch": 5.11,
      "grad_norm": 5.5594162940979,
      "learning_rate": 9.779785431959345e-06,
      "loss": 0.2266,
      "step": 27150
    },
    {
      "epoch": 5.11,
      "grad_norm": 7.668999671936035,
      "learning_rate": 9.776021080368907e-06,
      "loss": 0.1229,
      "step": 27160
    },
    {
      "epoch": 5.11,
      "grad_norm": 11.121779441833496,
      "learning_rate": 9.772256728778468e-06,
      "loss": 0.25,
      "step": 27170
    },
    {
      "epoch": 5.12,
      "grad_norm": 6.381068706512451,
      "learning_rate": 9.76849237718803e-06,
      "loss": 0.2836,
      "step": 27180
    },
    {
      "epoch": 5.12,
      "grad_norm": 4.6819071769714355,
      "learning_rate": 9.764728025597591e-06,
      "loss": 0.1976,
      "step": 27190
    },
    {
      "epoch": 5.12,
      "grad_norm": 5.285612106323242,
      "learning_rate": 9.760963674007153e-06,
      "loss": 0.3107,
      "step": 27200
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.13070780038833618,
      "learning_rate": 9.757199322416714e-06,
      "loss": 0.1434,
      "step": 27210
    },
    {
      "epoch": 5.12,
      "grad_norm": 0.010976992547512054,
      "learning_rate": 9.753434970826276e-06,
      "loss": 0.0527,
      "step": 27220
    },
    {
      "epoch": 5.13,
      "grad_norm": 5.181121349334717,
      "learning_rate": 9.749670619235837e-06,
      "loss": 0.3073,
      "step": 27230
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.01172599196434021,
      "learning_rate": 9.745906267645399e-06,
      "loss": 0.1293,
      "step": 27240
    },
    {
      "epoch": 5.13,
      "grad_norm": 10.517288208007812,
      "learning_rate": 9.74214191605496e-06,
      "loss": 0.2111,
      "step": 27250
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.02627822756767273,
      "learning_rate": 9.738377564464521e-06,
      "loss": 0.157,
      "step": 27260
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.01754416711628437,
      "learning_rate": 9.734613212874083e-06,
      "loss": 0.1051,
      "step": 27270
    },
    {
      "epoch": 5.13,
      "grad_norm": 0.33240336179733276,
      "learning_rate": 9.730848861283644e-06,
      "loss": 0.2443,
      "step": 27280
    },
    {
      "epoch": 5.14,
      "grad_norm": 14.459914207458496,
      "learning_rate": 9.727084509693206e-06,
      "loss": 0.1693,
      "step": 27290
    },
    {
      "epoch": 5.14,
      "grad_norm": 4.136043548583984,
      "learning_rate": 9.723320158102767e-06,
      "loss": 0.1144,
      "step": 27300
    },
    {
      "epoch": 5.14,
      "grad_norm": 8.685572624206543,
      "learning_rate": 9.719555806512329e-06,
      "loss": 0.2932,
      "step": 27310
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.02122410014271736,
      "learning_rate": 9.71579145492189e-06,
      "loss": 0.0579,
      "step": 27320
    },
    {
      "epoch": 5.14,
      "grad_norm": 0.007427821401506662,
      "learning_rate": 9.712027103331452e-06,
      "loss": 0.2298,
      "step": 27330
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.04279862344264984,
      "learning_rate": 9.708262751741013e-06,
      "loss": 0.2272,
      "step": 27340
    },
    {
      "epoch": 5.15,
      "grad_norm": 4.248894691467285,
      "learning_rate": 9.704498400150575e-06,
      "loss": 0.1344,
      "step": 27350
    },
    {
      "epoch": 5.15,
      "grad_norm": 4.729659557342529,
      "learning_rate": 9.700734048560136e-06,
      "loss": 0.1951,
      "step": 27360
    },
    {
      "epoch": 5.15,
      "grad_norm": 10.804882049560547,
      "learning_rate": 9.696969696969698e-06,
      "loss": 0.3522,
      "step": 27370
    },
    {
      "epoch": 5.15,
      "grad_norm": 0.039111074060201645,
      "learning_rate": 9.693205345379259e-06,
      "loss": 0.1077,
      "step": 27380
    },
    {
      "epoch": 5.16,
      "grad_norm": 8.604037284851074,
      "learning_rate": 9.68944099378882e-06,
      "loss": 0.1344,
      "step": 27390
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.005857485346496105,
      "learning_rate": 9.685676642198382e-06,
      "loss": 0.1275,
      "step": 27400
    },
    {
      "epoch": 5.16,
      "grad_norm": 14.375760078430176,
      "learning_rate": 9.681912290607943e-06,
      "loss": 0.3415,
      "step": 27410
    },
    {
      "epoch": 5.16,
      "grad_norm": 13.976086616516113,
      "learning_rate": 9.678147939017505e-06,
      "loss": 0.2402,
      "step": 27420
    },
    {
      "epoch": 5.16,
      "grad_norm": 28.31745147705078,
      "learning_rate": 9.674383587427066e-06,
      "loss": 0.1684,
      "step": 27430
    },
    {
      "epoch": 5.16,
      "grad_norm": 0.7871201038360596,
      "learning_rate": 9.670619235836628e-06,
      "loss": 0.3231,
      "step": 27440
    },
    {
      "epoch": 5.17,
      "grad_norm": 3.794248342514038,
      "learning_rate": 9.66685488424619e-06,
      "loss": 0.1991,
      "step": 27450
    },
    {
      "epoch": 5.17,
      "grad_norm": 11.05125904083252,
      "learning_rate": 9.66309053265575e-06,
      "loss": 0.2647,
      "step": 27460
    },
    {
      "epoch": 5.17,
      "grad_norm": 29.934812545776367,
      "learning_rate": 9.659326181065312e-06,
      "loss": 0.2462,
      "step": 27470
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.011904767714440823,
      "learning_rate": 9.655561829474874e-06,
      "loss": 0.2889,
      "step": 27480
    },
    {
      "epoch": 5.17,
      "grad_norm": 10.814526557922363,
      "learning_rate": 9.651797477884435e-06,
      "loss": 0.3317,
      "step": 27490
    },
    {
      "epoch": 5.18,
      "grad_norm": 3.471715211868286,
      "learning_rate": 9.648033126293997e-06,
      "loss": 0.1247,
      "step": 27500
    },
    {
      "epoch": 5.18,
      "grad_norm": 8.675235748291016,
      "learning_rate": 9.644268774703558e-06,
      "loss": 0.2583,
      "step": 27510
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.14856240153312683,
      "learning_rate": 9.64050442311312e-06,
      "loss": 0.1792,
      "step": 27520
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.010126655921339989,
      "learning_rate": 9.636740071522681e-06,
      "loss": 0.1967,
      "step": 27530
    },
    {
      "epoch": 5.18,
      "grad_norm": 0.22748489677906036,
      "learning_rate": 9.632975719932242e-06,
      "loss": 0.2769,
      "step": 27540
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.014524069614708424,
      "learning_rate": 9.629211368341804e-06,
      "loss": 0.1775,
      "step": 27550
    },
    {
      "epoch": 5.19,
      "grad_norm": 6.692471504211426,
      "learning_rate": 9.625447016751365e-06,
      "loss": 0.2986,
      "step": 27560
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.09901455789804459,
      "learning_rate": 9.621682665160927e-06,
      "loss": 0.227,
      "step": 27570
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.00740487314760685,
      "learning_rate": 9.617918313570488e-06,
      "loss": 0.1713,
      "step": 27580
    },
    {
      "epoch": 5.19,
      "grad_norm": 0.1699577271938324,
      "learning_rate": 9.61415396198005e-06,
      "loss": 0.1381,
      "step": 27590
    },
    {
      "epoch": 5.19,
      "grad_norm": 3.192525863647461,
      "learning_rate": 9.610389610389611e-06,
      "loss": 0.1164,
      "step": 27600
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.1085520088672638,
      "learning_rate": 9.606625258799173e-06,
      "loss": 0.2018,
      "step": 27610
    },
    {
      "epoch": 5.2,
      "grad_norm": 0.45968666672706604,
      "learning_rate": 9.602860907208734e-06,
      "loss": 0.188,
      "step": 27620
    },
    {
      "epoch": 5.2,
      "grad_norm": 6.309462547302246,
      "learning_rate": 9.599096555618296e-06,
      "loss": 0.1927,
      "step": 27630
    },
    {
      "epoch": 5.2,
      "grad_norm": 10.3814058303833,
      "learning_rate": 9.595332204027857e-06,
      "loss": 0.1425,
      "step": 27640
    },
    {
      "epoch": 5.2,
      "grad_norm": 5.42204475402832,
      "learning_rate": 9.591567852437418e-06,
      "loss": 0.1822,
      "step": 27650
    },
    {
      "epoch": 5.21,
      "grad_norm": 20.324928283691406,
      "learning_rate": 9.58780350084698e-06,
      "loss": 0.3071,
      "step": 27660
    },
    {
      "epoch": 5.21,
      "grad_norm": 13.532119750976562,
      "learning_rate": 9.584039149256541e-06,
      "loss": 0.2024,
      "step": 27670
    },
    {
      "epoch": 5.21,
      "grad_norm": 15.923417091369629,
      "learning_rate": 9.580274797666103e-06,
      "loss": 0.1027,
      "step": 27680
    },
    {
      "epoch": 5.21,
      "grad_norm": 0.02121681347489357,
      "learning_rate": 9.576510446075664e-06,
      "loss": 0.148,
      "step": 27690
    },
    {
      "epoch": 5.21,
      "grad_norm": 7.156311511993408,
      "learning_rate": 9.572746094485226e-06,
      "loss": 0.1113,
      "step": 27700
    },
    {
      "epoch": 5.22,
      "grad_norm": 14.614628791809082,
      "learning_rate": 9.568981742894787e-06,
      "loss": 0.0681,
      "step": 27710
    },
    {
      "epoch": 5.22,
      "grad_norm": 10.054220199584961,
      "learning_rate": 9.565217391304349e-06,
      "loss": 0.0949,
      "step": 27720
    },
    {
      "epoch": 5.22,
      "grad_norm": 14.659689903259277,
      "learning_rate": 9.56145303971391e-06,
      "loss": 0.2635,
      "step": 27730
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.0067359162494540215,
      "learning_rate": 9.557688688123472e-06,
      "loss": 0.1383,
      "step": 27740
    },
    {
      "epoch": 5.22,
      "grad_norm": 0.17363157868385315,
      "learning_rate": 9.553924336533033e-06,
      "loss": 0.3372,
      "step": 27750
    },
    {
      "epoch": 5.22,
      "grad_norm": 3.175229549407959,
      "learning_rate": 9.550159984942595e-06,
      "loss": 0.2227,
      "step": 27760
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.4161076247692108,
      "learning_rate": 9.546395633352156e-06,
      "loss": 0.2293,
      "step": 27770
    },
    {
      "epoch": 5.23,
      "grad_norm": 1.8930001258850098,
      "learning_rate": 9.542631281761717e-06,
      "loss": 0.0762,
      "step": 27780
    },
    {
      "epoch": 5.23,
      "grad_norm": 21.78729820251465,
      "learning_rate": 9.538866930171279e-06,
      "loss": 0.2616,
      "step": 27790
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.007066166028380394,
      "learning_rate": 9.53510257858084e-06,
      "loss": 0.1513,
      "step": 27800
    },
    {
      "epoch": 5.23,
      "grad_norm": 8.604775428771973,
      "learning_rate": 9.531338226990402e-06,
      "loss": 0.1688,
      "step": 27810
    },
    {
      "epoch": 5.24,
      "grad_norm": 7.924352645874023,
      "learning_rate": 9.527573875399963e-06,
      "loss": 0.0635,
      "step": 27820
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.21233785152435303,
      "learning_rate": 9.523809523809525e-06,
      "loss": 0.0816,
      "step": 27830
    },
    {
      "epoch": 5.24,
      "grad_norm": 22.927818298339844,
      "learning_rate": 9.520045172219086e-06,
      "loss": 0.3386,
      "step": 27840
    },
    {
      "epoch": 5.24,
      "grad_norm": 9.30520248413086,
      "learning_rate": 9.516280820628648e-06,
      "loss": 0.0527,
      "step": 27850
    },
    {
      "epoch": 5.24,
      "grad_norm": 0.021669048815965652,
      "learning_rate": 9.512516469038209e-06,
      "loss": 0.1761,
      "step": 27860
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.008399346843361855,
      "learning_rate": 9.50875211744777e-06,
      "loss": 0.3285,
      "step": 27870
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.013145764358341694,
      "learning_rate": 9.504987765857332e-06,
      "loss": 0.0633,
      "step": 27880
    },
    {
      "epoch": 5.25,
      "grad_norm": 10.008935928344727,
      "learning_rate": 9.501223414266894e-06,
      "loss": 0.0688,
      "step": 27890
    },
    {
      "epoch": 5.25,
      "grad_norm": 7.382472038269043,
      "learning_rate": 9.497459062676455e-06,
      "loss": 0.4089,
      "step": 27900
    },
    {
      "epoch": 5.25,
      "grad_norm": 0.021265998482704163,
      "learning_rate": 9.493694711086016e-06,
      "loss": 0.1951,
      "step": 27910
    },
    {
      "epoch": 5.26,
      "grad_norm": 26.058290481567383,
      "learning_rate": 9.489930359495578e-06,
      "loss": 0.2485,
      "step": 27920
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.9384995102882385,
      "learning_rate": 9.48616600790514e-06,
      "loss": 0.3583,
      "step": 27930
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.253492534160614,
      "learning_rate": 9.4824016563147e-06,
      "loss": 0.178,
      "step": 27940
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.006237076595425606,
      "learning_rate": 9.478637304724262e-06,
      "loss": 0.2483,
      "step": 27950
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.4867120087146759,
      "learning_rate": 9.474872953133824e-06,
      "loss": 0.1165,
      "step": 27960
    },
    {
      "epoch": 5.26,
      "grad_norm": 34.590545654296875,
      "learning_rate": 9.471108601543385e-06,
      "loss": 0.2123,
      "step": 27970
    },
    {
      "epoch": 5.27,
      "grad_norm": 9.859752655029297,
      "learning_rate": 9.467344249952947e-06,
      "loss": 0.11,
      "step": 27980
    },
    {
      "epoch": 5.27,
      "grad_norm": 1.1678277254104614,
      "learning_rate": 9.463579898362508e-06,
      "loss": 0.081,
      "step": 27990
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.018770772963762283,
      "learning_rate": 9.45981554677207e-06,
      "loss": 0.2225,
      "step": 28000
    },
    {
      "epoch": 5.27,
      "grad_norm": 0.008414745330810547,
      "learning_rate": 9.456051195181631e-06,
      "loss": 0.1194,
      "step": 28010
    },
    {
      "epoch": 5.27,
      "grad_norm": 15.697770118713379,
      "learning_rate": 9.452286843591193e-06,
      "loss": 0.1138,
      "step": 28020
    },
    {
      "epoch": 5.28,
      "grad_norm": 1.1247235536575317,
      "learning_rate": 9.448522492000754e-06,
      "loss": 0.0621,
      "step": 28030
    },
    {
      "epoch": 5.28,
      "grad_norm": 2.586589813232422,
      "learning_rate": 9.444758140410315e-06,
      "loss": 0.0837,
      "step": 28040
    },
    {
      "epoch": 5.28,
      "grad_norm": 22.925575256347656,
      "learning_rate": 9.440993788819877e-06,
      "loss": 0.4135,
      "step": 28050
    },
    {
      "epoch": 5.28,
      "grad_norm": 6.155892848968506,
      "learning_rate": 9.437229437229438e-06,
      "loss": 0.0183,
      "step": 28060
    },
    {
      "epoch": 5.28,
      "grad_norm": 5.4548234939575195,
      "learning_rate": 9.433465085639e-06,
      "loss": 0.158,
      "step": 28070
    },
    {
      "epoch": 5.29,
      "grad_norm": 34.537620544433594,
      "learning_rate": 9.429700734048561e-06,
      "loss": 0.1634,
      "step": 28080
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.25412067770957947,
      "learning_rate": 9.425936382458121e-06,
      "loss": 0.1251,
      "step": 28090
    },
    {
      "epoch": 5.29,
      "grad_norm": 18.719083786010742,
      "learning_rate": 9.422172030867684e-06,
      "loss": 0.3072,
      "step": 28100
    },
    {
      "epoch": 5.29,
      "grad_norm": 4.264339923858643,
      "learning_rate": 9.418407679277246e-06,
      "loss": 0.3038,
      "step": 28110
    },
    {
      "epoch": 5.29,
      "grad_norm": 4.084651470184326,
      "learning_rate": 9.414643327686807e-06,
      "loss": 0.2721,
      "step": 28120
    },
    {
      "epoch": 5.29,
      "grad_norm": 0.40065842866897583,
      "learning_rate": 9.410878976096369e-06,
      "loss": 0.1111,
      "step": 28130
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.8203812837600708,
      "learning_rate": 9.40711462450593e-06,
      "loss": 0.1318,
      "step": 28140
    },
    {
      "epoch": 5.3,
      "grad_norm": 12.882645606994629,
      "learning_rate": 9.403350272915491e-06,
      "loss": 0.3131,
      "step": 28150
    },
    {
      "epoch": 5.3,
      "grad_norm": 7.430115222930908,
      "learning_rate": 9.399585921325053e-06,
      "loss": 0.2426,
      "step": 28160
    },
    {
      "epoch": 5.3,
      "grad_norm": 10.190829277038574,
      "learning_rate": 9.395821569734614e-06,
      "loss": 0.3214,
      "step": 28170
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.21115858852863312,
      "learning_rate": 9.392057218144174e-06,
      "loss": 0.0589,
      "step": 28180
    },
    {
      "epoch": 5.31,
      "grad_norm": 5.649243354797363,
      "learning_rate": 9.388292866553737e-06,
      "loss": 0.2862,
      "step": 28190
    },
    {
      "epoch": 5.31,
      "grad_norm": 17.122295379638672,
      "learning_rate": 9.384528514963299e-06,
      "loss": 0.2565,
      "step": 28200
    },
    {
      "epoch": 5.31,
      "grad_norm": 3.9809648990631104,
      "learning_rate": 9.38076416337286e-06,
      "loss": 0.3475,
      "step": 28210
    },
    {
      "epoch": 5.31,
      "grad_norm": 0.013284835033118725,
      "learning_rate": 9.376999811782422e-06,
      "loss": 0.2342,
      "step": 28220
    },
    {
      "epoch": 5.31,
      "grad_norm": 2.0136146545410156,
      "learning_rate": 9.373235460191983e-06,
      "loss": 0.2582,
      "step": 28230
    },
    {
      "epoch": 5.32,
      "grad_norm": 1.301734447479248,
      "learning_rate": 9.369471108601545e-06,
      "loss": 0.1077,
      "step": 28240
    },
    {
      "epoch": 5.32,
      "grad_norm": 14.69788646697998,
      "learning_rate": 9.365706757011106e-06,
      "loss": 0.2577,
      "step": 28250
    },
    {
      "epoch": 5.32,
      "grad_norm": 24.01556968688965,
      "learning_rate": 9.361942405420668e-06,
      "loss": 0.2131,
      "step": 28260
    },
    {
      "epoch": 5.32,
      "grad_norm": 0.03824397176504135,
      "learning_rate": 9.358178053830227e-06,
      "loss": 0.2897,
      "step": 28270
    },
    {
      "epoch": 5.32,
      "grad_norm": 10.120931625366211,
      "learning_rate": 9.354413702239789e-06,
      "loss": 0.1553,
      "step": 28280
    },
    {
      "epoch": 5.32,
      "grad_norm": 4.70806360244751,
      "learning_rate": 9.350649350649352e-06,
      "loss": 0.2308,
      "step": 28290
    },
    {
      "epoch": 5.33,
      "grad_norm": 11.94262981414795,
      "learning_rate": 9.346884999058913e-06,
      "loss": 0.0395,
      "step": 28300
    },
    {
      "epoch": 5.33,
      "grad_norm": 23.4514102935791,
      "learning_rate": 9.343120647468475e-06,
      "loss": 0.2968,
      "step": 28310
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.32038408517837524,
      "learning_rate": 9.339356295878036e-06,
      "loss": 0.2312,
      "step": 28320
    },
    {
      "epoch": 5.33,
      "grad_norm": 10.092565536499023,
      "learning_rate": 9.335591944287598e-06,
      "loss": 0.2923,
      "step": 28330
    },
    {
      "epoch": 5.33,
      "grad_norm": 24.80899429321289,
      "learning_rate": 9.33182759269716e-06,
      "loss": 0.2707,
      "step": 28340
    },
    {
      "epoch": 5.34,
      "grad_norm": 3.2424933910369873,
      "learning_rate": 9.32806324110672e-06,
      "loss": 0.2337,
      "step": 28350
    },
    {
      "epoch": 5.34,
      "grad_norm": 13.110199928283691,
      "learning_rate": 9.32429888951628e-06,
      "loss": 0.1659,
      "step": 28360
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.6325567960739136,
      "learning_rate": 9.320534537925842e-06,
      "loss": 0.1928,
      "step": 28370
    },
    {
      "epoch": 5.34,
      "grad_norm": 10.177062034606934,
      "learning_rate": 9.316770186335405e-06,
      "loss": 0.2291,
      "step": 28380
    },
    {
      "epoch": 5.34,
      "grad_norm": 10.371405601501465,
      "learning_rate": 9.313005834744967e-06,
      "loss": 0.247,
      "step": 28390
    },
    {
      "epoch": 5.35,
      "grad_norm": 21.302942276000977,
      "learning_rate": 9.309241483154528e-06,
      "loss": 0.1344,
      "step": 28400
    },
    {
      "epoch": 5.35,
      "grad_norm": 2.4723966121673584,
      "learning_rate": 9.30547713156409e-06,
      "loss": 0.0539,
      "step": 28410
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.08080439269542694,
      "learning_rate": 9.301712779973651e-06,
      "loss": 0.2974,
      "step": 28420
    },
    {
      "epoch": 5.35,
      "grad_norm": 12.073210716247559,
      "learning_rate": 9.297948428383212e-06,
      "loss": 0.4267,
      "step": 28430
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.02411097101867199,
      "learning_rate": 9.294184076792772e-06,
      "loss": 0.2717,
      "step": 28440
    },
    {
      "epoch": 5.35,
      "grad_norm": 0.1342860758304596,
      "learning_rate": 9.290419725202334e-06,
      "loss": 0.3478,
      "step": 28450
    },
    {
      "epoch": 5.36,
      "grad_norm": 12.694531440734863,
      "learning_rate": 9.286655373611895e-06,
      "loss": 0.191,
      "step": 28460
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.026467325165867805,
      "learning_rate": 9.282891022021458e-06,
      "loss": 0.0925,
      "step": 28470
    },
    {
      "epoch": 5.36,
      "grad_norm": 19.364316940307617,
      "learning_rate": 9.27912667043102e-06,
      "loss": 0.3616,
      "step": 28480
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.014195889234542847,
      "learning_rate": 9.275362318840581e-06,
      "loss": 0.2287,
      "step": 28490
    },
    {
      "epoch": 5.36,
      "grad_norm": 0.010434860363602638,
      "learning_rate": 9.271597967250143e-06,
      "loss": 0.3215,
      "step": 28500
    },
    {
      "epoch": 5.37,
      "grad_norm": 13.616914749145508,
      "learning_rate": 9.267833615659704e-06,
      "loss": 0.1723,
      "step": 28510
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.008647059090435505,
      "learning_rate": 9.264069264069266e-06,
      "loss": 0.3642,
      "step": 28520
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.22188574075698853,
      "learning_rate": 9.260304912478825e-06,
      "loss": 0.1869,
      "step": 28530
    },
    {
      "epoch": 5.37,
      "grad_norm": 37.893280029296875,
      "learning_rate": 9.256540560888387e-06,
      "loss": 0.1526,
      "step": 28540
    },
    {
      "epoch": 5.37,
      "grad_norm": 0.00978679209947586,
      "learning_rate": 9.252776209297948e-06,
      "loss": 0.1741,
      "step": 28550
    },
    {
      "epoch": 5.38,
      "grad_norm": 4.415532112121582,
      "learning_rate": 9.24901185770751e-06,
      "loss": 0.1888,
      "step": 28560
    },
    {
      "epoch": 5.38,
      "grad_norm": 3.4348366260528564,
      "learning_rate": 9.245247506117073e-06,
      "loss": 0.2308,
      "step": 28570
    },
    {
      "epoch": 5.38,
      "grad_norm": 3.7699170112609863,
      "learning_rate": 9.241483154526634e-06,
      "loss": 0.253,
      "step": 28580
    },
    {
      "epoch": 5.38,
      "grad_norm": 7.985724449157715,
      "learning_rate": 9.237718802936196e-06,
      "loss": 0.1935,
      "step": 28590
    },
    {
      "epoch": 5.38,
      "grad_norm": 21.77646255493164,
      "learning_rate": 9.233954451345757e-06,
      "loss": 0.2809,
      "step": 28600
    },
    {
      "epoch": 5.38,
      "grad_norm": 0.02481369860470295,
      "learning_rate": 9.230190099755319e-06,
      "loss": 0.1689,
      "step": 28610
    },
    {
      "epoch": 5.39,
      "grad_norm": 6.620485305786133,
      "learning_rate": 9.226425748164878e-06,
      "loss": 0.2342,
      "step": 28620
    },
    {
      "epoch": 5.39,
      "grad_norm": 7.075104236602783,
      "learning_rate": 9.22266139657444e-06,
      "loss": 0.1847,
      "step": 28630
    },
    {
      "epoch": 5.39,
      "grad_norm": 6.284593105316162,
      "learning_rate": 9.218897044984001e-06,
      "loss": 0.1344,
      "step": 28640
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.011582594364881516,
      "learning_rate": 9.215132693393563e-06,
      "loss": 0.1626,
      "step": 28650
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.00561127532273531,
      "learning_rate": 9.211368341803126e-06,
      "loss": 0.0939,
      "step": 28660
    },
    {
      "epoch": 5.4,
      "grad_norm": 19.15254020690918,
      "learning_rate": 9.207603990212687e-06,
      "loss": 0.2911,
      "step": 28670
    },
    {
      "epoch": 5.4,
      "grad_norm": 7.338189601898193,
      "learning_rate": 9.203839638622249e-06,
      "loss": 0.2077,
      "step": 28680
    },
    {
      "epoch": 5.4,
      "grad_norm": 5.405006408691406,
      "learning_rate": 9.20007528703181e-06,
      "loss": 0.1454,
      "step": 28690
    },
    {
      "epoch": 5.4,
      "grad_norm": 23.43243408203125,
      "learning_rate": 9.19631093544137e-06,
      "loss": 0.2196,
      "step": 28700
    },
    {
      "epoch": 5.4,
      "grad_norm": 0.03142094612121582,
      "learning_rate": 9.192546583850932e-06,
      "loss": 0.1788,
      "step": 28710
    },
    {
      "epoch": 5.41,
      "grad_norm": 6.420365810394287,
      "learning_rate": 9.188782232260493e-06,
      "loss": 0.1479,
      "step": 28720
    },
    {
      "epoch": 5.41,
      "grad_norm": 0.05389772728085518,
      "learning_rate": 9.185017880670055e-06,
      "loss": 0.0749,
      "step": 28730
    },
    {
      "epoch": 5.41,
      "grad_norm": 8.599747657775879,
      "learning_rate": 9.181253529079616e-06,
      "loss": 0.2755,
      "step": 28740
    },
    {
      "epoch": 5.41,
      "grad_norm": 30.944591522216797,
      "learning_rate": 9.177489177489179e-06,
      "loss": 0.2463,
      "step": 28750
    },
    {
      "epoch": 5.41,
      "grad_norm": 10.699268341064453,
      "learning_rate": 9.17372482589874e-06,
      "loss": 0.1579,
      "step": 28760
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.011417158879339695,
      "learning_rate": 9.169960474308302e-06,
      "loss": 0.2544,
      "step": 28770
    },
    {
      "epoch": 5.42,
      "grad_norm": 3.7936208248138428,
      "learning_rate": 9.166196122717864e-06,
      "loss": 0.0593,
      "step": 28780
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.006710621062666178,
      "learning_rate": 9.162431771127423e-06,
      "loss": 0.0834,
      "step": 28790
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.009677231311798096,
      "learning_rate": 9.158667419536985e-06,
      "loss": 0.1661,
      "step": 28800
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.3058014214038849,
      "learning_rate": 9.154903067946546e-06,
      "loss": 0.0539,
      "step": 28810
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.00502186082303524,
      "learning_rate": 9.151138716356108e-06,
      "loss": 0.1242,
      "step": 28820
    },
    {
      "epoch": 5.43,
      "grad_norm": 9.553414344787598,
      "learning_rate": 9.147374364765669e-06,
      "loss": 0.2826,
      "step": 28830
    },
    {
      "epoch": 5.43,
      "grad_norm": 9.712553024291992,
      "learning_rate": 9.143610013175232e-06,
      "loss": 0.1931,
      "step": 28840
    },
    {
      "epoch": 5.43,
      "grad_norm": 28.550477981567383,
      "learning_rate": 9.139845661584794e-06,
      "loss": 0.237,
      "step": 28850
    },
    {
      "epoch": 5.43,
      "grad_norm": 0.014167079702019691,
      "learning_rate": 9.136081309994355e-06,
      "loss": 0.252,
      "step": 28860
    },
    {
      "epoch": 5.43,
      "grad_norm": 12.660571098327637,
      "learning_rate": 9.132316958403917e-06,
      "loss": 0.1043,
      "step": 28870
    },
    {
      "epoch": 5.44,
      "grad_norm": 13.140114784240723,
      "learning_rate": 9.128552606813476e-06,
      "loss": 0.138,
      "step": 28880
    },
    {
      "epoch": 5.44,
      "grad_norm": 25.609630584716797,
      "learning_rate": 9.124788255223038e-06,
      "loss": 0.0986,
      "step": 28890
    },
    {
      "epoch": 5.44,
      "grad_norm": 9.639039993286133,
      "learning_rate": 9.1210239036326e-06,
      "loss": 0.4931,
      "step": 28900
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.3381165862083435,
      "learning_rate": 9.11725955204216e-06,
      "loss": 0.2057,
      "step": 28910
    },
    {
      "epoch": 5.44,
      "grad_norm": 0.1331396996974945,
      "learning_rate": 9.113495200451722e-06,
      "loss": 0.2599,
      "step": 28920
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.06031129136681557,
      "learning_rate": 9.109730848861284e-06,
      "loss": 0.1465,
      "step": 28930
    },
    {
      "epoch": 5.45,
      "grad_norm": 11.993691444396973,
      "learning_rate": 9.105966497270847e-06,
      "loss": 0.1043,
      "step": 28940
    },
    {
      "epoch": 5.45,
      "grad_norm": 27.820892333984375,
      "learning_rate": 9.102202145680408e-06,
      "loss": 0.1955,
      "step": 28950
    },
    {
      "epoch": 5.45,
      "grad_norm": 0.09441579133272171,
      "learning_rate": 9.09843779408997e-06,
      "loss": 0.0346,
      "step": 28960
    },
    {
      "epoch": 5.45,
      "grad_norm": 9.36015796661377,
      "learning_rate": 9.09467344249953e-06,
      "loss": 0.116,
      "step": 28970
    },
    {
      "epoch": 5.45,
      "grad_norm": 16.87674331665039,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.1039,
      "step": 28980
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.005927353631705046,
      "learning_rate": 9.087144739318652e-06,
      "loss": 0.2809,
      "step": 28990
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.02805718034505844,
      "learning_rate": 9.083380387728214e-06,
      "loss": 0.0769,
      "step": 29000
    },
    {
      "epoch": 5.46,
      "grad_norm": 0.007796235382556915,
      "learning_rate": 9.079616036137775e-06,
      "loss": 0.1315,
      "step": 29010
    },
    {
      "epoch": 5.46,
      "grad_norm": 2.393768072128296,
      "learning_rate": 9.075851684547337e-06,
      "loss": 0.2613,
      "step": 29020
    },
    {
      "epoch": 5.46,
      "grad_norm": 27.01193618774414,
      "learning_rate": 9.0720873329569e-06,
      "loss": 0.0657,
      "step": 29030
    },
    {
      "epoch": 5.47,
      "grad_norm": 4.923625946044922,
      "learning_rate": 9.068322981366461e-06,
      "loss": 0.1741,
      "step": 29040
    },
    {
      "epoch": 5.47,
      "grad_norm": 7.6751708984375,
      "learning_rate": 9.064558629776021e-06,
      "loss": 0.1935,
      "step": 29050
    },
    {
      "epoch": 5.47,
      "grad_norm": 14.95875358581543,
      "learning_rate": 9.060794278185583e-06,
      "loss": 0.218,
      "step": 29060
    },
    {
      "epoch": 5.47,
      "grad_norm": 6.555695533752441,
      "learning_rate": 9.057029926595144e-06,
      "loss": 0.2449,
      "step": 29070
    },
    {
      "epoch": 5.47,
      "grad_norm": 11.786258697509766,
      "learning_rate": 9.053265575004706e-06,
      "loss": 0.1999,
      "step": 29080
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.03370462730526924,
      "learning_rate": 9.049501223414267e-06,
      "loss": 0.0656,
      "step": 29090
    },
    {
      "epoch": 5.48,
      "grad_norm": 6.308347702026367,
      "learning_rate": 9.045736871823829e-06,
      "loss": 0.2209,
      "step": 29100
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.0574021078646183,
      "learning_rate": 9.04197252023339e-06,
      "loss": 0.2378,
      "step": 29110
    },
    {
      "epoch": 5.48,
      "grad_norm": 20.671274185180664,
      "learning_rate": 9.038208168642953e-06,
      "loss": 0.1645,
      "step": 29120
    },
    {
      "epoch": 5.48,
      "grad_norm": 37.490699768066406,
      "learning_rate": 9.034443817052515e-06,
      "loss": 0.0971,
      "step": 29130
    },
    {
      "epoch": 5.48,
      "grad_norm": 6.057639122009277,
      "learning_rate": 9.030679465462074e-06,
      "loss": 0.1228,
      "step": 29140
    },
    {
      "epoch": 5.49,
      "grad_norm": 1.933973789215088,
      "learning_rate": 9.026915113871636e-06,
      "loss": 0.1755,
      "step": 29150
    },
    {
      "epoch": 5.49,
      "grad_norm": 15.097574234008789,
      "learning_rate": 9.023150762281197e-06,
      "loss": 0.3344,
      "step": 29160
    },
    {
      "epoch": 5.49,
      "grad_norm": 2.808037281036377,
      "learning_rate": 9.019386410690759e-06,
      "loss": 0.1084,
      "step": 29170
    },
    {
      "epoch": 5.49,
      "grad_norm": 6.949810028076172,
      "learning_rate": 9.01562205910032e-06,
      "loss": 0.1654,
      "step": 29180
    },
    {
      "epoch": 5.49,
      "grad_norm": 0.009319888427853584,
      "learning_rate": 9.011857707509882e-06,
      "loss": 0.253,
      "step": 29190
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.033200185745954514,
      "learning_rate": 9.008093355919443e-06,
      "loss": 0.2674,
      "step": 29200
    },
    {
      "epoch": 5.5,
      "grad_norm": 43.68589401245117,
      "learning_rate": 9.004329004329005e-06,
      "loss": 0.4405,
      "step": 29210
    },
    {
      "epoch": 5.5,
      "grad_norm": 1.6882309913635254,
      "learning_rate": 9.000564652738568e-06,
      "loss": 0.1309,
      "step": 29220
    },
    {
      "epoch": 5.5,
      "grad_norm": 0.18247650563716888,
      "learning_rate": 8.996800301148128e-06,
      "loss": 0.0958,
      "step": 29230
    },
    {
      "epoch": 5.5,
      "grad_norm": 5.736489772796631,
      "learning_rate": 8.993035949557689e-06,
      "loss": 0.2529,
      "step": 29240
    },
    {
      "epoch": 5.51,
      "grad_norm": 11.175929069519043,
      "learning_rate": 8.98927159796725e-06,
      "loss": 0.2024,
      "step": 29250
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.5597444176673889,
      "learning_rate": 8.985507246376812e-06,
      "loss": 0.22,
      "step": 29260
    },
    {
      "epoch": 5.51,
      "grad_norm": 24.76885414123535,
      "learning_rate": 8.981742894786373e-06,
      "loss": 0.2176,
      "step": 29270
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.005698820576071739,
      "learning_rate": 8.977978543195935e-06,
      "loss": 0.0109,
      "step": 29280
    },
    {
      "epoch": 5.51,
      "grad_norm": 19.68682861328125,
      "learning_rate": 8.974214191605496e-06,
      "loss": 0.0787,
      "step": 29290
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.9808218479156494,
      "learning_rate": 8.970449840015058e-06,
      "loss": 0.0345,
      "step": 29300
    },
    {
      "epoch": 5.52,
      "grad_norm": 21.14544105529785,
      "learning_rate": 8.96668548842462e-06,
      "loss": 0.0937,
      "step": 29310
    },
    {
      "epoch": 5.52,
      "grad_norm": 25.65923309326172,
      "learning_rate": 8.96292113683418e-06,
      "loss": 0.4328,
      "step": 29320
    },
    {
      "epoch": 5.52,
      "grad_norm": 32.50546646118164,
      "learning_rate": 8.959156785243742e-06,
      "loss": 0.4352,
      "step": 29330
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.007387540768831968,
      "learning_rate": 8.955392433653304e-06,
      "loss": 0.1509,
      "step": 29340
    },
    {
      "epoch": 5.52,
      "grad_norm": 0.006670736242085695,
      "learning_rate": 8.951628082062865e-06,
      "loss": 0.0904,
      "step": 29350
    },
    {
      "epoch": 5.53,
      "grad_norm": 12.418031692504883,
      "learning_rate": 8.947863730472427e-06,
      "loss": 0.2056,
      "step": 29360
    },
    {
      "epoch": 5.53,
      "grad_norm": 2.558192014694214,
      "learning_rate": 8.944099378881988e-06,
      "loss": 0.059,
      "step": 29370
    },
    {
      "epoch": 5.53,
      "grad_norm": 11.413241386413574,
      "learning_rate": 8.94033502729155e-06,
      "loss": 0.2827,
      "step": 29380
    },
    {
      "epoch": 5.53,
      "grad_norm": 1.335037350654602,
      "learning_rate": 8.936570675701111e-06,
      "loss": 0.2188,
      "step": 29390
    },
    {
      "epoch": 5.53,
      "grad_norm": 19.74190330505371,
      "learning_rate": 8.932806324110672e-06,
      "loss": 0.173,
      "step": 29400
    },
    {
      "epoch": 5.54,
      "grad_norm": 3.8921146392822266,
      "learning_rate": 8.929041972520234e-06,
      "loss": 0.2343,
      "step": 29410
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.021860744804143906,
      "learning_rate": 8.925277620929795e-06,
      "loss": 0.1212,
      "step": 29420
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.06367145478725433,
      "learning_rate": 8.921513269339357e-06,
      "loss": 0.0481,
      "step": 29430
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.00437152711674571,
      "learning_rate": 8.917748917748918e-06,
      "loss": 0.1427,
      "step": 29440
    },
    {
      "epoch": 5.54,
      "grad_norm": 0.022850865498185158,
      "learning_rate": 8.91398456615848e-06,
      "loss": 0.1864,
      "step": 29450
    },
    {
      "epoch": 5.54,
      "grad_norm": 28.052215576171875,
      "learning_rate": 8.910220214568041e-06,
      "loss": 0.1998,
      "step": 29460
    },
    {
      "epoch": 5.55,
      "grad_norm": 3.947526216506958,
      "learning_rate": 8.906455862977603e-06,
      "loss": 0.2784,
      "step": 29470
    },
    {
      "epoch": 5.55,
      "grad_norm": 6.976309776306152,
      "learning_rate": 8.902691511387164e-06,
      "loss": 0.3018,
      "step": 29480
    },
    {
      "epoch": 5.55,
      "grad_norm": 10.273598670959473,
      "learning_rate": 8.898927159796726e-06,
      "loss": 0.3646,
      "step": 29490
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.08721452206373215,
      "learning_rate": 8.895162808206287e-06,
      "loss": 0.2888,
      "step": 29500
    },
    {
      "epoch": 5.55,
      "grad_norm": 38.66028594970703,
      "learning_rate": 8.891398456615848e-06,
      "loss": 0.3157,
      "step": 29510
    },
    {
      "epoch": 5.56,
      "grad_norm": 3.956404209136963,
      "learning_rate": 8.88763410502541e-06,
      "loss": 0.2244,
      "step": 29520
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.039385031908750534,
      "learning_rate": 8.883869753434971e-06,
      "loss": 0.201,
      "step": 29530
    },
    {
      "epoch": 5.56,
      "grad_norm": 16.372528076171875,
      "learning_rate": 8.880105401844533e-06,
      "loss": 0.1975,
      "step": 29540
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.08266725391149521,
      "learning_rate": 8.876341050254094e-06,
      "loss": 0.3196,
      "step": 29550
    },
    {
      "epoch": 5.56,
      "grad_norm": 0.055904895067214966,
      "learning_rate": 8.872576698663656e-06,
      "loss": 0.1899,
      "step": 29560
    },
    {
      "epoch": 5.57,
      "grad_norm": 26.705352783203125,
      "learning_rate": 8.868812347073217e-06,
      "loss": 0.2509,
      "step": 29570
    },
    {
      "epoch": 5.57,
      "grad_norm": 7.836174488067627,
      "learning_rate": 8.865047995482779e-06,
      "loss": 0.2491,
      "step": 29580
    },
    {
      "epoch": 5.57,
      "grad_norm": 1.2346054315567017,
      "learning_rate": 8.86128364389234e-06,
      "loss": 0.0957,
      "step": 29590
    },
    {
      "epoch": 5.57,
      "grad_norm": 16.422927856445312,
      "learning_rate": 8.857519292301902e-06,
      "loss": 0.1756,
      "step": 29600
    },
    {
      "epoch": 5.57,
      "grad_norm": 0.004819110967218876,
      "learning_rate": 8.853754940711463e-06,
      "loss": 0.0756,
      "step": 29610
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.4149014353752136,
      "learning_rate": 8.849990589121025e-06,
      "loss": 0.1843,
      "step": 29620
    },
    {
      "epoch": 5.58,
      "grad_norm": 0.01997578702867031,
      "learning_rate": 8.846226237530586e-06,
      "loss": 0.1705,
      "step": 29630
    },
    {
      "epoch": 5.58,
      "grad_norm": 11.239051818847656,
      "learning_rate": 8.842461885940147e-06,
      "loss": 0.2037,
      "step": 29640
    },
    {
      "epoch": 5.58,
      "grad_norm": 23.398822784423828,
      "learning_rate": 8.838697534349709e-06,
      "loss": 0.1891,
      "step": 29650
    },
    {
      "epoch": 5.58,
      "grad_norm": 7.845251560211182,
      "learning_rate": 8.83493318275927e-06,
      "loss": 0.3322,
      "step": 29660
    },
    {
      "epoch": 5.58,
      "grad_norm": 13.699748039245605,
      "learning_rate": 8.831168831168832e-06,
      "loss": 0.1884,
      "step": 29670
    },
    {
      "epoch": 5.59,
      "grad_norm": 2.5840084552764893,
      "learning_rate": 8.827404479578393e-06,
      "loss": 0.1415,
      "step": 29680
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.016256140545010567,
      "learning_rate": 8.823640127987955e-06,
      "loss": 0.3379,
      "step": 29690
    },
    {
      "epoch": 5.59,
      "grad_norm": 19.495197296142578,
      "learning_rate": 8.819875776397516e-06,
      "loss": 0.2615,
      "step": 29700
    },
    {
      "epoch": 5.59,
      "grad_norm": 0.027355168014764786,
      "learning_rate": 8.816111424807078e-06,
      "loss": 0.3446,
      "step": 29710
    },
    {
      "epoch": 5.59,
      "grad_norm": 9.003369331359863,
      "learning_rate": 8.812347073216639e-06,
      "loss": 0.2263,
      "step": 29720
    },
    {
      "epoch": 5.6,
      "grad_norm": 5.985177993774414,
      "learning_rate": 8.8085827216262e-06,
      "loss": 0.0691,
      "step": 29730
    },
    {
      "epoch": 5.6,
      "grad_norm": 7.364879131317139,
      "learning_rate": 8.804818370035762e-06,
      "loss": 0.2286,
      "step": 29740
    },
    {
      "epoch": 5.6,
      "grad_norm": 17.181093215942383,
      "learning_rate": 8.801054018445324e-06,
      "loss": 0.0607,
      "step": 29750
    },
    {
      "epoch": 5.6,
      "grad_norm": 11.913125038146973,
      "learning_rate": 8.797289666854885e-06,
      "loss": 0.2232,
      "step": 29760
    },
    {
      "epoch": 5.6,
      "grad_norm": 10.52779769897461,
      "learning_rate": 8.793525315264446e-06,
      "loss": 0.2599,
      "step": 29770
    },
    {
      "epoch": 5.61,
      "grad_norm": 27.054058074951172,
      "learning_rate": 8.789760963674008e-06,
      "loss": 0.3556,
      "step": 29780
    },
    {
      "epoch": 5.61,
      "grad_norm": 2.2026960849761963,
      "learning_rate": 8.78599661208357e-06,
      "loss": 0.2399,
      "step": 29790
    },
    {
      "epoch": 5.61,
      "grad_norm": 8.636094093322754,
      "learning_rate": 8.78223226049313e-06,
      "loss": 0.189,
      "step": 29800
    },
    {
      "epoch": 5.61,
      "grad_norm": 13.693768501281738,
      "learning_rate": 8.778467908902692e-06,
      "loss": 0.1984,
      "step": 29810
    },
    {
      "epoch": 5.61,
      "grad_norm": 18.401159286499023,
      "learning_rate": 8.774703557312254e-06,
      "loss": 0.2822,
      "step": 29820
    },
    {
      "epoch": 5.61,
      "grad_norm": 0.012890457175672054,
      "learning_rate": 8.770939205721815e-06,
      "loss": 0.2574,
      "step": 29830
    },
    {
      "epoch": 5.62,
      "grad_norm": 21.309545516967773,
      "learning_rate": 8.767174854131377e-06,
      "loss": 0.1443,
      "step": 29840
    },
    {
      "epoch": 5.62,
      "grad_norm": 11.88377857208252,
      "learning_rate": 8.763410502540938e-06,
      "loss": 0.2589,
      "step": 29850
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.05657990276813507,
      "learning_rate": 8.7596461509505e-06,
      "loss": 0.092,
      "step": 29860
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.37455877661705017,
      "learning_rate": 8.755881799360061e-06,
      "loss": 0.2226,
      "step": 29870
    },
    {
      "epoch": 5.62,
      "grad_norm": 0.005761316511780024,
      "learning_rate": 8.752117447769623e-06,
      "loss": 0.2344,
      "step": 29880
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.013984526507556438,
      "learning_rate": 8.748353096179184e-06,
      "loss": 0.3452,
      "step": 29890
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.022852474823594093,
      "learning_rate": 8.744588744588745e-06,
      "loss": 0.2365,
      "step": 29900
    },
    {
      "epoch": 5.63,
      "grad_norm": 0.7875692248344421,
      "learning_rate": 8.740824392998307e-06,
      "loss": 0.16,
      "step": 29910
    },
    {
      "epoch": 5.63,
      "grad_norm": 13.076730728149414,
      "learning_rate": 8.737060041407868e-06,
      "loss": 0.206,
      "step": 29920
    },
    {
      "epoch": 5.63,
      "grad_norm": 23.206405639648438,
      "learning_rate": 8.73329568981743e-06,
      "loss": 0.344,
      "step": 29930
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.08533094078302383,
      "learning_rate": 8.729531338226991e-06,
      "loss": 0.1654,
      "step": 29940
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.0452103354036808,
      "learning_rate": 8.725766986636553e-06,
      "loss": 0.2708,
      "step": 29950
    },
    {
      "epoch": 5.64,
      "grad_norm": 32.9577751159668,
      "learning_rate": 8.722002635046114e-06,
      "loss": 0.156,
      "step": 29960
    },
    {
      "epoch": 5.64,
      "grad_norm": 2.12272310256958,
      "learning_rate": 8.718238283455676e-06,
      "loss": 0.1226,
      "step": 29970
    },
    {
      "epoch": 5.64,
      "grad_norm": 0.009269097819924355,
      "learning_rate": 8.714473931865237e-06,
      "loss": 0.1321,
      "step": 29980
    },
    {
      "epoch": 5.64,
      "grad_norm": 2.1899476051330566,
      "learning_rate": 8.710709580274799e-06,
      "loss": 0.2647,
      "step": 29990
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.036416057497262955,
      "learning_rate": 8.70694522868436e-06,
      "loss": 0.2971,
      "step": 30000
    },
    {
      "epoch": 5.65,
      "grad_norm": 18.00596809387207,
      "learning_rate": 8.703180877093921e-06,
      "loss": 0.1702,
      "step": 30010
    },
    {
      "epoch": 5.65,
      "grad_norm": 0.850931704044342,
      "learning_rate": 8.699416525503483e-06,
      "loss": 0.241,
      "step": 30020
    },
    {
      "epoch": 5.65,
      "grad_norm": 2.545569658279419,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.1966,
      "step": 30030
    },
    {
      "epoch": 5.65,
      "grad_norm": 6.137225151062012,
      "learning_rate": 8.691887822322606e-06,
      "loss": 0.0799,
      "step": 30040
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.05489889159798622,
      "learning_rate": 8.688123470732167e-06,
      "loss": 0.1082,
      "step": 30050
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.05705985054373741,
      "learning_rate": 8.684359119141729e-06,
      "loss": 0.1996,
      "step": 30060
    },
    {
      "epoch": 5.66,
      "grad_norm": 2.177003860473633,
      "learning_rate": 8.68059476755129e-06,
      "loss": 0.1727,
      "step": 30070
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.4541029632091522,
      "learning_rate": 8.676830415960852e-06,
      "loss": 0.2375,
      "step": 30080
    },
    {
      "epoch": 5.66,
      "grad_norm": 0.38978058099746704,
      "learning_rate": 8.673066064370413e-06,
      "loss": 0.0175,
      "step": 30090
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.031389929354190826,
      "learning_rate": 8.669301712779975e-06,
      "loss": 0.1547,
      "step": 30100
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.16240645945072174,
      "learning_rate": 8.665537361189536e-06,
      "loss": 0.2859,
      "step": 30110
    },
    {
      "epoch": 5.67,
      "grad_norm": 15.24821662902832,
      "learning_rate": 8.661773009599098e-06,
      "loss": 0.1377,
      "step": 30120
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.06035977602005005,
      "learning_rate": 8.658008658008659e-06,
      "loss": 0.201,
      "step": 30130
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.0370708666741848,
      "learning_rate": 8.65424430641822e-06,
      "loss": 0.0424,
      "step": 30140
    },
    {
      "epoch": 5.67,
      "grad_norm": 0.47388291358947754,
      "learning_rate": 8.650479954827782e-06,
      "loss": 0.1436,
      "step": 30150
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.013754886575043201,
      "learning_rate": 8.646715603237343e-06,
      "loss": 0.1316,
      "step": 30160
    },
    {
      "epoch": 5.68,
      "grad_norm": 0.011090034618973732,
      "learning_rate": 8.642951251646905e-06,
      "loss": 0.2694,
      "step": 30170
    },
    {
      "epoch": 5.68,
      "grad_norm": 1.243071436882019,
      "learning_rate": 8.639186900056465e-06,
      "loss": 0.1673,
      "step": 30180
    },
    {
      "epoch": 5.68,
      "grad_norm": 2.265900135040283,
      "learning_rate": 8.635422548466028e-06,
      "loss": 0.0633,
      "step": 30190
    },
    {
      "epoch": 5.68,
      "grad_norm": 17.071086883544922,
      "learning_rate": 8.63165819687559e-06,
      "loss": 0.3256,
      "step": 30200
    },
    {
      "epoch": 5.69,
      "grad_norm": 12.47226333618164,
      "learning_rate": 8.62789384528515e-06,
      "loss": 0.2559,
      "step": 30210
    },
    {
      "epoch": 5.69,
      "grad_norm": 6.761903762817383,
      "learning_rate": 8.624129493694712e-06,
      "loss": 0.1803,
      "step": 30220
    },
    {
      "epoch": 5.69,
      "grad_norm": 15.36771297454834,
      "learning_rate": 8.620365142104274e-06,
      "loss": 0.2001,
      "step": 30230
    },
    {
      "epoch": 5.69,
      "grad_norm": 29.51740074157715,
      "learning_rate": 8.616600790513835e-06,
      "loss": 0.1732,
      "step": 30240
    },
    {
      "epoch": 5.69,
      "grad_norm": 11.280985832214355,
      "learning_rate": 8.612836438923397e-06,
      "loss": 0.2812,
      "step": 30250
    },
    {
      "epoch": 5.7,
      "grad_norm": 14.25707721710205,
      "learning_rate": 8.609072087332958e-06,
      "loss": 0.3334,
      "step": 30260
    },
    {
      "epoch": 5.7,
      "grad_norm": 8.348906517028809,
      "learning_rate": 8.605307735742518e-06,
      "loss": 0.2546,
      "step": 30270
    },
    {
      "epoch": 5.7,
      "grad_norm": 1.0092333555221558,
      "learning_rate": 8.601543384152081e-06,
      "loss": 0.0467,
      "step": 30280
    },
    {
      "epoch": 5.7,
      "grad_norm": 32.301658630371094,
      "learning_rate": 8.597779032561642e-06,
      "loss": 0.3246,
      "step": 30290
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.613673746585846,
      "learning_rate": 8.594014680971204e-06,
      "loss": 0.1761,
      "step": 30300
    },
    {
      "epoch": 5.7,
      "grad_norm": 0.019146479666233063,
      "learning_rate": 8.590250329380765e-06,
      "loss": 0.2205,
      "step": 30310
    },
    {
      "epoch": 5.71,
      "grad_norm": 14.915626525878906,
      "learning_rate": 8.586485977790327e-06,
      "loss": 0.026,
      "step": 30320
    },
    {
      "epoch": 5.71,
      "grad_norm": 14.830574989318848,
      "learning_rate": 8.582721626199888e-06,
      "loss": 0.1565,
      "step": 30330
    },
    {
      "epoch": 5.71,
      "grad_norm": 0.3777834177017212,
      "learning_rate": 8.57895727460945e-06,
      "loss": 0.3159,
      "step": 30340
    },
    {
      "epoch": 5.71,
      "grad_norm": 10.11447811126709,
      "learning_rate": 8.575192923019011e-06,
      "loss": 0.1371,
      "step": 30350
    },
    {
      "epoch": 5.71,
      "grad_norm": 6.1209635734558105,
      "learning_rate": 8.571428571428571e-06,
      "loss": 0.3541,
      "step": 30360
    },
    {
      "epoch": 5.72,
      "grad_norm": 1.4113259315490723,
      "learning_rate": 8.567664219838134e-06,
      "loss": 0.0549,
      "step": 30370
    },
    {
      "epoch": 5.72,
      "grad_norm": 6.999384880065918,
      "learning_rate": 8.563899868247696e-06,
      "loss": 0.1596,
      "step": 30380
    },
    {
      "epoch": 5.72,
      "grad_norm": 4.226133823394775,
      "learning_rate": 8.560135516657257e-06,
      "loss": 0.2414,
      "step": 30390
    },
    {
      "epoch": 5.72,
      "grad_norm": 3.7721009254455566,
      "learning_rate": 8.556371165066818e-06,
      "loss": 0.1437,
      "step": 30400
    },
    {
      "epoch": 5.72,
      "grad_norm": 1.0274720191955566,
      "learning_rate": 8.55260681347638e-06,
      "loss": 0.1179,
      "step": 30410
    },
    {
      "epoch": 5.73,
      "grad_norm": 8.594444274902344,
      "learning_rate": 8.548842461885941e-06,
      "loss": 0.4114,
      "step": 30420
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.42885348200798035,
      "learning_rate": 8.545078110295503e-06,
      "loss": 0.1744,
      "step": 30430
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.00945940613746643,
      "learning_rate": 8.541313758705064e-06,
      "loss": 0.2016,
      "step": 30440
    },
    {
      "epoch": 5.73,
      "grad_norm": 2.913994312286377,
      "learning_rate": 8.537549407114624e-06,
      "loss": 0.2011,
      "step": 30450
    },
    {
      "epoch": 5.73,
      "grad_norm": 7.888825416564941,
      "learning_rate": 8.533785055524186e-06,
      "loss": 0.2535,
      "step": 30460
    },
    {
      "epoch": 5.73,
      "grad_norm": 0.005579719785600901,
      "learning_rate": 8.530020703933749e-06,
      "loss": 0.4258,
      "step": 30470
    },
    {
      "epoch": 5.74,
      "grad_norm": 30.898923873901367,
      "learning_rate": 8.52625635234331e-06,
      "loss": 0.2701,
      "step": 30480
    },
    {
      "epoch": 5.74,
      "grad_norm": 0.005523036699742079,
      "learning_rate": 8.522492000752872e-06,
      "loss": 0.0823,
      "step": 30490
    },
    {
      "epoch": 5.74,
      "grad_norm": 23.156978607177734,
      "learning_rate": 8.518727649162433e-06,
      "loss": 0.3339,
      "step": 30500
    },
    {
      "epoch": 5.74,
      "grad_norm": 0.056860215961933136,
      "learning_rate": 8.514963297571995e-06,
      "loss": 0.1139,
      "step": 30510
    },
    {
      "epoch": 5.74,
      "grad_norm": 5.66157341003418,
      "learning_rate": 8.511198945981556e-06,
      "loss": 0.4304,
      "step": 30520
    },
    {
      "epoch": 5.75,
      "grad_norm": 8.250516891479492,
      "learning_rate": 8.507434594391116e-06,
      "loss": 0.3463,
      "step": 30530
    },
    {
      "epoch": 5.75,
      "grad_norm": 27.503793716430664,
      "learning_rate": 8.503670242800677e-06,
      "loss": 0.0847,
      "step": 30540
    },
    {
      "epoch": 5.75,
      "grad_norm": 19.425661087036133,
      "learning_rate": 8.499905891210239e-06,
      "loss": 0.3161,
      "step": 30550
    },
    {
      "epoch": 5.75,
      "grad_norm": 23.538179397583008,
      "learning_rate": 8.496141539619802e-06,
      "loss": 0.2981,
      "step": 30560
    },
    {
      "epoch": 5.75,
      "grad_norm": 0.014637521468102932,
      "learning_rate": 8.492377188029363e-06,
      "loss": 0.2376,
      "step": 30570
    },
    {
      "epoch": 5.76,
      "grad_norm": 1.1009384393692017,
      "learning_rate": 8.488612836438925e-06,
      "loss": 0.2123,
      "step": 30580
    },
    {
      "epoch": 5.76,
      "grad_norm": 13.660248756408691,
      "learning_rate": 8.484848484848486e-06,
      "loss": 0.3527,
      "step": 30590
    },
    {
      "epoch": 5.76,
      "grad_norm": 1.007034182548523,
      "learning_rate": 8.481084133258048e-06,
      "loss": 0.1026,
      "step": 30600
    },
    {
      "epoch": 5.76,
      "grad_norm": 17.017820358276367,
      "learning_rate": 8.477319781667609e-06,
      "loss": 0.3523,
      "step": 30610
    },
    {
      "epoch": 5.76,
      "grad_norm": 2.1739561557769775,
      "learning_rate": 8.473555430077169e-06,
      "loss": 0.2626,
      "step": 30620
    },
    {
      "epoch": 5.77,
      "grad_norm": 2.9290850162506104,
      "learning_rate": 8.46979107848673e-06,
      "loss": 0.1341,
      "step": 30630
    },
    {
      "epoch": 5.77,
      "grad_norm": 13.736824035644531,
      "learning_rate": 8.466026726896292e-06,
      "loss": 0.3587,
      "step": 30640
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.07666464149951935,
      "learning_rate": 8.462262375305855e-06,
      "loss": 0.1917,
      "step": 30650
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.006205015815794468,
      "learning_rate": 8.458498023715416e-06,
      "loss": 0.0996,
      "step": 30660
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.1765328049659729,
      "learning_rate": 8.454733672124978e-06,
      "loss": 0.0529,
      "step": 30670
    },
    {
      "epoch": 5.77,
      "grad_norm": 0.036436084657907486,
      "learning_rate": 8.45096932053454e-06,
      "loss": 0.0758,
      "step": 30680
    },
    {
      "epoch": 5.78,
      "grad_norm": 5.241800785064697,
      "learning_rate": 8.4472049689441e-06,
      "loss": 0.1611,
      "step": 30690
    },
    {
      "epoch": 5.78,
      "grad_norm": 15.640585899353027,
      "learning_rate": 8.443440617353662e-06,
      "loss": 0.1982,
      "step": 30700
    },
    {
      "epoch": 5.78,
      "grad_norm": 0.005324713885784149,
      "learning_rate": 8.439676265763222e-06,
      "loss": 0.2077,
      "step": 30710
    },
    {
      "epoch": 5.78,
      "grad_norm": 18.186288833618164,
      "learning_rate": 8.435911914172784e-06,
      "loss": 0.1897,
      "step": 30720
    },
    {
      "epoch": 5.78,
      "grad_norm": 0.556161642074585,
      "learning_rate": 8.432147562582345e-06,
      "loss": 0.1146,
      "step": 30730
    },
    {
      "epoch": 5.79,
      "grad_norm": 10.52157974243164,
      "learning_rate": 8.428383210991906e-06,
      "loss": 0.2051,
      "step": 30740
    },
    {
      "epoch": 5.79,
      "grad_norm": 14.694414138793945,
      "learning_rate": 8.42461885940147e-06,
      "loss": 0.1745,
      "step": 30750
    },
    {
      "epoch": 5.79,
      "grad_norm": 2.318531036376953,
      "learning_rate": 8.420854507811031e-06,
      "loss": 0.119,
      "step": 30760
    },
    {
      "epoch": 5.79,
      "grad_norm": 0.1005595475435257,
      "learning_rate": 8.417090156220593e-06,
      "loss": 0.1088,
      "step": 30770
    },
    {
      "epoch": 5.79,
      "grad_norm": 16.391887664794922,
      "learning_rate": 8.413325804630154e-06,
      "loss": 0.2854,
      "step": 30780
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.0275210402905941,
      "learning_rate": 8.409561453039714e-06,
      "loss": 0.1463,
      "step": 30790
    },
    {
      "epoch": 5.8,
      "grad_norm": 43.346412658691406,
      "learning_rate": 8.405797101449275e-06,
      "loss": 0.3322,
      "step": 30800
    },
    {
      "epoch": 5.8,
      "grad_norm": 9.857486724853516,
      "learning_rate": 8.402032749858837e-06,
      "loss": 0.1704,
      "step": 30810
    },
    {
      "epoch": 5.8,
      "grad_norm": 0.17444755136966705,
      "learning_rate": 8.398268398268398e-06,
      "loss": 0.3018,
      "step": 30820
    },
    {
      "epoch": 5.8,
      "grad_norm": 1.168224811553955,
      "learning_rate": 8.39450404667796e-06,
      "loss": 0.1043,
      "step": 30830
    },
    {
      "epoch": 5.8,
      "grad_norm": 24.487333297729492,
      "learning_rate": 8.390739695087523e-06,
      "loss": 0.0352,
      "step": 30840
    },
    {
      "epoch": 5.81,
      "grad_norm": 0.03991680592298508,
      "learning_rate": 8.386975343497084e-06,
      "loss": 0.182,
      "step": 30850
    },
    {
      "epoch": 5.81,
      "grad_norm": 7.108608722686768,
      "learning_rate": 8.383210991906646e-06,
      "loss": 0.1509,
      "step": 30860
    },
    {
      "epoch": 5.81,
      "grad_norm": 7.147695541381836,
      "learning_rate": 8.379446640316207e-06,
      "loss": 0.0697,
      "step": 30870
    },
    {
      "epoch": 5.81,
      "grad_norm": 0.009975707158446312,
      "learning_rate": 8.375682288725767e-06,
      "loss": 0.1118,
      "step": 30880
    },
    {
      "epoch": 5.81,
      "grad_norm": 11.378852844238281,
      "learning_rate": 8.371917937135328e-06,
      "loss": 0.3624,
      "step": 30890
    },
    {
      "epoch": 5.82,
      "grad_norm": 6.672840118408203,
      "learning_rate": 8.36815358554489e-06,
      "loss": 0.1089,
      "step": 30900
    },
    {
      "epoch": 5.82,
      "grad_norm": 7.33618688583374,
      "learning_rate": 8.364389233954451e-06,
      "loss": 0.2524,
      "step": 30910
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.010631621815264225,
      "learning_rate": 8.360624882364013e-06,
      "loss": 0.1703,
      "step": 30920
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.004941882565617561,
      "learning_rate": 8.356860530773576e-06,
      "loss": 0.2134,
      "step": 30930
    },
    {
      "epoch": 5.82,
      "grad_norm": 0.04475608468055725,
      "learning_rate": 8.353096179183137e-06,
      "loss": 0.1409,
      "step": 30940
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.030573207885026932,
      "learning_rate": 8.349331827592699e-06,
      "loss": 0.1934,
      "step": 30950
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.023497873917222023,
      "learning_rate": 8.34556747600226e-06,
      "loss": 0.201,
      "step": 30960
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.008009742014110088,
      "learning_rate": 8.34180312441182e-06,
      "loss": 0.3253,
      "step": 30970
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.0365145243704319,
      "learning_rate": 8.338038772821381e-06,
      "loss": 0.0868,
      "step": 30980
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.6706798076629639,
      "learning_rate": 8.334274421230943e-06,
      "loss": 0.2973,
      "step": 30990
    },
    {
      "epoch": 5.83,
      "grad_norm": 0.06727469712495804,
      "learning_rate": 8.330510069640504e-06,
      "loss": 0.2077,
      "step": 31000
    },
    {
      "epoch": 5.84,
      "grad_norm": 0.049850549548864365,
      "learning_rate": 8.326745718050066e-06,
      "loss": 0.1021,
      "step": 31010
    },
    {
      "epoch": 5.84,
      "grad_norm": 17.14585304260254,
      "learning_rate": 8.322981366459629e-06,
      "loss": 0.0976,
      "step": 31020
    },
    {
      "epoch": 5.84,
      "grad_norm": 17.85866928100586,
      "learning_rate": 8.31921701486919e-06,
      "loss": 0.2928,
      "step": 31030
    },
    {
      "epoch": 5.84,
      "grad_norm": 21.222166061401367,
      "learning_rate": 8.315452663278752e-06,
      "loss": 0.2203,
      "step": 31040
    },
    {
      "epoch": 5.84,
      "grad_norm": 20.12792205810547,
      "learning_rate": 8.311688311688313e-06,
      "loss": 0.2803,
      "step": 31050
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.6244221925735474,
      "learning_rate": 8.307923960097873e-06,
      "loss": 0.1354,
      "step": 31060
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.06850211322307587,
      "learning_rate": 8.304159608507435e-06,
      "loss": 0.1312,
      "step": 31070
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.011849544942378998,
      "learning_rate": 8.300395256916996e-06,
      "loss": 0.2976,
      "step": 31080
    },
    {
      "epoch": 5.85,
      "grad_norm": 0.09165441244840622,
      "learning_rate": 8.296630905326558e-06,
      "loss": 0.112,
      "step": 31090
    },
    {
      "epoch": 5.85,
      "grad_norm": 11.896322250366211,
      "learning_rate": 8.292866553736119e-06,
      "loss": 0.1499,
      "step": 31100
    },
    {
      "epoch": 5.86,
      "grad_norm": 0.6860090494155884,
      "learning_rate": 8.28910220214568e-06,
      "loss": 0.0999,
      "step": 31110
    },
    {
      "epoch": 5.86,
      "grad_norm": 22.199647903442383,
      "learning_rate": 8.285337850555244e-06,
      "loss": 0.4779,
      "step": 31120
    },
    {
      "epoch": 5.86,
      "grad_norm": 7.597449779510498,
      "learning_rate": 8.281573498964805e-06,
      "loss": 0.2471,
      "step": 31130
    },
    {
      "epoch": 5.86,
      "grad_norm": 0.362332284450531,
      "learning_rate": 8.277809147374365e-06,
      "loss": 0.109,
      "step": 31140
    },
    {
      "epoch": 5.86,
      "grad_norm": 18.790128707885742,
      "learning_rate": 8.274044795783926e-06,
      "loss": 0.1927,
      "step": 31150
    },
    {
      "epoch": 5.86,
      "grad_norm": 7.463292598724365,
      "learning_rate": 8.270280444193488e-06,
      "loss": 0.1028,
      "step": 31160
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.015491100959479809,
      "learning_rate": 8.26651609260305e-06,
      "loss": 0.1107,
      "step": 31170
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.008882504887878895,
      "learning_rate": 8.26275174101261e-06,
      "loss": 0.0729,
      "step": 31180
    },
    {
      "epoch": 5.87,
      "grad_norm": 24.92828369140625,
      "learning_rate": 8.258987389422172e-06,
      "loss": 0.1534,
      "step": 31190
    },
    {
      "epoch": 5.87,
      "grad_norm": 27.258203506469727,
      "learning_rate": 8.255223037831734e-06,
      "loss": 0.2341,
      "step": 31200
    },
    {
      "epoch": 5.87,
      "grad_norm": 0.013599599711596966,
      "learning_rate": 8.251458686241297e-06,
      "loss": 0.2832,
      "step": 31210
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.032255686819553375,
      "learning_rate": 8.247694334650858e-06,
      "loss": 0.3357,
      "step": 31220
    },
    {
      "epoch": 5.88,
      "grad_norm": 4.564566135406494,
      "learning_rate": 8.243929983060418e-06,
      "loss": 0.2728,
      "step": 31230
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.022260768339037895,
      "learning_rate": 8.24016563146998e-06,
      "loss": 0.0683,
      "step": 31240
    },
    {
      "epoch": 5.88,
      "grad_norm": 0.8031601905822754,
      "learning_rate": 8.236401279879541e-06,
      "loss": 0.205,
      "step": 31250
    },
    {
      "epoch": 5.88,
      "grad_norm": 33.895477294921875,
      "learning_rate": 8.232636928289102e-06,
      "loss": 0.2041,
      "step": 31260
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.016000116243958473,
      "learning_rate": 8.228872576698664e-06,
      "loss": 0.1395,
      "step": 31270
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.4374946653842926,
      "learning_rate": 8.225108225108225e-06,
      "loss": 0.3203,
      "step": 31280
    },
    {
      "epoch": 5.89,
      "grad_norm": 6.119121551513672,
      "learning_rate": 8.221343873517787e-06,
      "loss": 0.2823,
      "step": 31290
    },
    {
      "epoch": 5.89,
      "grad_norm": 7.773920059204102,
      "learning_rate": 8.21757952192735e-06,
      "loss": 0.3798,
      "step": 31300
    },
    {
      "epoch": 5.89,
      "grad_norm": 6.783458709716797,
      "learning_rate": 8.213815170336911e-06,
      "loss": 0.1039,
      "step": 31310
    },
    {
      "epoch": 5.89,
      "grad_norm": 0.011981431394815445,
      "learning_rate": 8.210050818746471e-06,
      "loss": 0.0832,
      "step": 31320
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.032474491745233536,
      "learning_rate": 8.206286467156033e-06,
      "loss": 0.287,
      "step": 31330
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.2800351679325104,
      "learning_rate": 8.202522115565594e-06,
      "loss": 0.1772,
      "step": 31340
    },
    {
      "epoch": 5.9,
      "grad_norm": 0.4671836197376251,
      "learning_rate": 8.198757763975156e-06,
      "loss": 0.0078,
      "step": 31350
    },
    {
      "epoch": 5.9,
      "grad_norm": 27.532869338989258,
      "learning_rate": 8.194993412384717e-06,
      "loss": 0.4374,
      "step": 31360
    },
    {
      "epoch": 5.9,
      "grad_norm": 6.066959381103516,
      "learning_rate": 8.191229060794278e-06,
      "loss": 0.39,
      "step": 31370
    },
    {
      "epoch": 5.91,
      "grad_norm": 0.024276822805404663,
      "learning_rate": 8.18746470920384e-06,
      "loss": 0.3776,
      "step": 31380
    },
    {
      "epoch": 5.91,
      "grad_norm": 0.008411074057221413,
      "learning_rate": 8.183700357613401e-06,
      "loss": 0.1086,
      "step": 31390
    },
    {
      "epoch": 5.91,
      "grad_norm": 3.630349636077881,
      "learning_rate": 8.179936006022963e-06,
      "loss": 0.1597,
      "step": 31400
    },
    {
      "epoch": 5.91,
      "grad_norm": 0.20165522396564484,
      "learning_rate": 8.176171654432524e-06,
      "loss": 0.3271,
      "step": 31410
    },
    {
      "epoch": 5.91,
      "grad_norm": 22.771024703979492,
      "learning_rate": 8.172407302842086e-06,
      "loss": 0.0859,
      "step": 31420
    },
    {
      "epoch": 5.92,
      "grad_norm": 23.93321990966797,
      "learning_rate": 8.168642951251647e-06,
      "loss": 0.2098,
      "step": 31430
    },
    {
      "epoch": 5.92,
      "grad_norm": 7.391737937927246,
      "learning_rate": 8.164878599661209e-06,
      "loss": 0.411,
      "step": 31440
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.2187797874212265,
      "learning_rate": 8.16111424807077e-06,
      "loss": 0.292,
      "step": 31450
    },
    {
      "epoch": 5.92,
      "grad_norm": 0.043489404022693634,
      "learning_rate": 8.157349896480332e-06,
      "loss": 0.2404,
      "step": 31460
    },
    {
      "epoch": 5.92,
      "grad_norm": 4.532589912414551,
      "learning_rate": 8.153585544889893e-06,
      "loss": 0.2849,
      "step": 31470
    },
    {
      "epoch": 5.93,
      "grad_norm": 11.369132041931152,
      "learning_rate": 8.149821193299455e-06,
      "loss": 0.13,
      "step": 31480
    },
    {
      "epoch": 5.93,
      "grad_norm": 11.958439826965332,
      "learning_rate": 8.146056841709016e-06,
      "loss": 0.2561,
      "step": 31490
    },
    {
      "epoch": 5.93,
      "grad_norm": 15.536293983459473,
      "learning_rate": 8.142292490118577e-06,
      "loss": 0.2889,
      "step": 31500
    },
    {
      "epoch": 5.93,
      "grad_norm": 8.54074764251709,
      "learning_rate": 8.138528138528139e-06,
      "loss": 0.1737,
      "step": 31510
    },
    {
      "epoch": 5.93,
      "grad_norm": 5.839772701263428,
      "learning_rate": 8.1347637869377e-06,
      "loss": 0.035,
      "step": 31520
    },
    {
      "epoch": 5.93,
      "grad_norm": 10.805840492248535,
      "learning_rate": 8.130999435347262e-06,
      "loss": 0.1951,
      "step": 31530
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.01644359715282917,
      "learning_rate": 8.127235083756823e-06,
      "loss": 0.1075,
      "step": 31540
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.8676337599754333,
      "learning_rate": 8.123470732166385e-06,
      "loss": 0.1434,
      "step": 31550
    },
    {
      "epoch": 5.94,
      "grad_norm": 0.006162993609905243,
      "learning_rate": 8.119706380575946e-06,
      "loss": 0.2423,
      "step": 31560
    },
    {
      "epoch": 5.94,
      "grad_norm": 6.256371974945068,
      "learning_rate": 8.115942028985508e-06,
      "loss": 0.0785,
      "step": 31570
    },
    {
      "epoch": 5.94,
      "grad_norm": 8.023968696594238,
      "learning_rate": 8.112177677395069e-06,
      "loss": 0.2243,
      "step": 31580
    },
    {
      "epoch": 5.95,
      "grad_norm": 0.41991788148880005,
      "learning_rate": 8.10841332580463e-06,
      "loss": 0.0969,
      "step": 31590
    },
    {
      "epoch": 5.95,
      "grad_norm": 5.869123458862305,
      "learning_rate": 8.104648974214192e-06,
      "loss": 0.27,
      "step": 31600
    },
    {
      "epoch": 5.95,
      "grad_norm": 11.712272644042969,
      "learning_rate": 8.100884622623754e-06,
      "loss": 0.3082,
      "step": 31610
    },
    {
      "epoch": 5.95,
      "grad_norm": 1.3723137378692627,
      "learning_rate": 8.097120271033315e-06,
      "loss": 0.2839,
      "step": 31620
    },
    {
      "epoch": 5.95,
      "grad_norm": 1.6338516473770142,
      "learning_rate": 8.093355919442876e-06,
      "loss": 0.1729,
      "step": 31630
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.6031217575073242,
      "learning_rate": 8.089591567852438e-06,
      "loss": 0.1169,
      "step": 31640
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.5117166042327881,
      "learning_rate": 8.085827216262e-06,
      "loss": 0.2052,
      "step": 31650
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.06870060414075851,
      "learning_rate": 8.08206286467156e-06,
      "loss": 0.2245,
      "step": 31660
    },
    {
      "epoch": 5.96,
      "grad_norm": 0.023770157247781754,
      "learning_rate": 8.078298513081122e-06,
      "loss": 0.2938,
      "step": 31670
    },
    {
      "epoch": 5.96,
      "grad_norm": 6.043656349182129,
      "learning_rate": 8.074534161490684e-06,
      "loss": 0.0853,
      "step": 31680
    },
    {
      "epoch": 5.96,
      "grad_norm": 3.3479888439178467,
      "learning_rate": 8.070769809900245e-06,
      "loss": 0.0854,
      "step": 31690
    },
    {
      "epoch": 5.97,
      "grad_norm": 4.874444961547852,
      "learning_rate": 8.067005458309807e-06,
      "loss": 0.173,
      "step": 31700
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.03719315677881241,
      "learning_rate": 8.063241106719368e-06,
      "loss": 0.1679,
      "step": 31710
    },
    {
      "epoch": 5.97,
      "grad_norm": 1.9241883754730225,
      "learning_rate": 8.05947675512893e-06,
      "loss": 0.2966,
      "step": 31720
    },
    {
      "epoch": 5.97,
      "grad_norm": 14.62259578704834,
      "learning_rate": 8.055712403538491e-06,
      "loss": 0.1798,
      "step": 31730
    },
    {
      "epoch": 5.97,
      "grad_norm": 0.015617160126566887,
      "learning_rate": 8.051948051948052e-06,
      "loss": 0.2855,
      "step": 31740
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.12502971291542053,
      "learning_rate": 8.048183700357614e-06,
      "loss": 0.1693,
      "step": 31750
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.014193184673786163,
      "learning_rate": 8.044419348767175e-06,
      "loss": 0.2331,
      "step": 31760
    },
    {
      "epoch": 5.98,
      "grad_norm": 9.47991943359375,
      "learning_rate": 8.040654997176737e-06,
      "loss": 0.3188,
      "step": 31770
    },
    {
      "epoch": 5.98,
      "grad_norm": 0.007529792375862598,
      "learning_rate": 8.036890645586298e-06,
      "loss": 0.1416,
      "step": 31780
    },
    {
      "epoch": 5.98,
      "grad_norm": 17.113964080810547,
      "learning_rate": 8.03312629399586e-06,
      "loss": 0.2743,
      "step": 31790
    },
    {
      "epoch": 5.99,
      "grad_norm": 8.697827339172363,
      "learning_rate": 8.029361942405421e-06,
      "loss": 0.1175,
      "step": 31800
    },
    {
      "epoch": 5.99,
      "grad_norm": 10.947951316833496,
      "learning_rate": 8.025597590814983e-06,
      "loss": 0.3277,
      "step": 31810
    },
    {
      "epoch": 5.99,
      "grad_norm": 9.364234924316406,
      "learning_rate": 8.021833239224544e-06,
      "loss": 0.1949,
      "step": 31820
    },
    {
      "epoch": 5.99,
      "grad_norm": 0.1207304522395134,
      "learning_rate": 8.018068887634106e-06,
      "loss": 0.198,
      "step": 31830
    },
    {
      "epoch": 5.99,
      "grad_norm": 1.75444757938385,
      "learning_rate": 8.014304536043667e-06,
      "loss": 0.2386,
      "step": 31840
    },
    {
      "epoch": 5.99,
      "grad_norm": 11.878552436828613,
      "learning_rate": 8.010540184453229e-06,
      "loss": 0.3483,
      "step": 31850
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9711681604385376,
      "learning_rate": 8.00677583286279e-06,
      "loss": 0.1746,
      "step": 31860
    },
    {
      "epoch": 6.0,
      "grad_norm": 22.212026596069336,
      "learning_rate": 8.003011481272351e-06,
      "loss": 0.2591,
      "step": 31870
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9888,
      "eval_loss": 0.04222671315073967,
      "eval_runtime": 51.0113,
      "eval_samples_per_second": 147.026,
      "eval_steps_per_second": 18.388,
      "step": 31878
    },
    {
      "epoch": 6.0,
      "grad_norm": 19.507991790771484,
      "learning_rate": 7.999247129681913e-06,
      "loss": 0.3921,
      "step": 31880
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.400970458984375,
      "learning_rate": 7.995482778091474e-06,
      "loss": 0.1635,
      "step": 31890
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.16339637339115143,
      "learning_rate": 7.991718426501036e-06,
      "loss": 0.1064,
      "step": 31900
    },
    {
      "epoch": 6.01,
      "grad_norm": 36.96540451049805,
      "learning_rate": 7.987954074910597e-06,
      "loss": 0.0972,
      "step": 31910
    },
    {
      "epoch": 6.01,
      "grad_norm": 14.071617126464844,
      "learning_rate": 7.984189723320159e-06,
      "loss": 0.1361,
      "step": 31920
    },
    {
      "epoch": 6.01,
      "grad_norm": 30.427289962768555,
      "learning_rate": 7.98042537172972e-06,
      "loss": 0.1523,
      "step": 31930
    },
    {
      "epoch": 6.01,
      "grad_norm": 27.549089431762695,
      "learning_rate": 7.976661020139282e-06,
      "loss": 0.1395,
      "step": 31940
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.05731729790568352,
      "learning_rate": 7.972896668548843e-06,
      "loss": 0.0515,
      "step": 31950
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.04269855096936226,
      "learning_rate": 7.969132316958405e-06,
      "loss": 0.1815,
      "step": 31960
    },
    {
      "epoch": 6.02,
      "grad_norm": 21.04296112060547,
      "learning_rate": 7.965367965367966e-06,
      "loss": 0.1875,
      "step": 31970
    },
    {
      "epoch": 6.02,
      "grad_norm": 8.330078125,
      "learning_rate": 7.961603613777528e-06,
      "loss": 0.1477,
      "step": 31980
    },
    {
      "epoch": 6.02,
      "grad_norm": 23.743839263916016,
      "learning_rate": 7.957839262187089e-06,
      "loss": 0.2419,
      "step": 31990
    },
    {
      "epoch": 6.02,
      "grad_norm": 4.648481845855713,
      "learning_rate": 7.95407491059665e-06,
      "loss": 0.1087,
      "step": 32000
    },
    {
      "epoch": 6.02,
      "grad_norm": 1.0855118036270142,
      "learning_rate": 7.950310559006212e-06,
      "loss": 0.1857,
      "step": 32010
    },
    {
      "epoch": 6.03,
      "grad_norm": 10.043625831604004,
      "learning_rate": 7.946546207415773e-06,
      "loss": 0.1279,
      "step": 32020
    },
    {
      "epoch": 6.03,
      "grad_norm": 16.099210739135742,
      "learning_rate": 7.942781855825335e-06,
      "loss": 0.2028,
      "step": 32030
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.003941613249480724,
      "learning_rate": 7.939017504234896e-06,
      "loss": 0.1887,
      "step": 32040
    },
    {
      "epoch": 6.03,
      "grad_norm": 1.094671607017517,
      "learning_rate": 7.935253152644458e-06,
      "loss": 0.1211,
      "step": 32050
    },
    {
      "epoch": 6.03,
      "grad_norm": 21.319555282592773,
      "learning_rate": 7.93148880105402e-06,
      "loss": 0.3216,
      "step": 32060
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.00910995714366436,
      "learning_rate": 7.92772444946358e-06,
      "loss": 0.1946,
      "step": 32070
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.037266381084918976,
      "learning_rate": 7.923960097873142e-06,
      "loss": 0.2916,
      "step": 32080
    },
    {
      "epoch": 6.04,
      "grad_norm": 8.356555938720703,
      "learning_rate": 7.920195746282704e-06,
      "loss": 0.0609,
      "step": 32090
    },
    {
      "epoch": 6.04,
      "grad_norm": 0.027218367904424667,
      "learning_rate": 7.916431394692265e-06,
      "loss": 0.1991,
      "step": 32100
    },
    {
      "epoch": 6.04,
      "grad_norm": 24.192346572875977,
      "learning_rate": 7.912667043101827e-06,
      "loss": 0.1389,
      "step": 32110
    },
    {
      "epoch": 6.05,
      "grad_norm": 1.3213474750518799,
      "learning_rate": 7.908902691511388e-06,
      "loss": 0.1137,
      "step": 32120
    },
    {
      "epoch": 6.05,
      "grad_norm": 23.115503311157227,
      "learning_rate": 7.90513833992095e-06,
      "loss": 0.1555,
      "step": 32130
    },
    {
      "epoch": 6.05,
      "grad_norm": 4.3735504150390625,
      "learning_rate": 7.901373988330511e-06,
      "loss": 0.1333,
      "step": 32140
    },
    {
      "epoch": 6.05,
      "grad_norm": 10.242804527282715,
      "learning_rate": 7.897609636740072e-06,
      "loss": 0.211,
      "step": 32150
    },
    {
      "epoch": 6.05,
      "grad_norm": 8.18448543548584,
      "learning_rate": 7.893845285149634e-06,
      "loss": 0.1541,
      "step": 32160
    },
    {
      "epoch": 6.05,
      "grad_norm": 1.3346278667449951,
      "learning_rate": 7.890080933559195e-06,
      "loss": 0.1466,
      "step": 32170
    },
    {
      "epoch": 6.06,
      "grad_norm": 3.4460432529449463,
      "learning_rate": 7.886316581968757e-06,
      "loss": 0.0746,
      "step": 32180
    },
    {
      "epoch": 6.06,
      "grad_norm": 3.7612664699554443,
      "learning_rate": 7.882552230378318e-06,
      "loss": 0.2867,
      "step": 32190
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.8583506941795349,
      "learning_rate": 7.87878787878788e-06,
      "loss": 0.0811,
      "step": 32200
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.21813808381557465,
      "learning_rate": 7.875023527197441e-06,
      "loss": 0.2676,
      "step": 32210
    },
    {
      "epoch": 6.06,
      "grad_norm": 0.07377442717552185,
      "learning_rate": 7.871259175607003e-06,
      "loss": 0.1689,
      "step": 32220
    },
    {
      "epoch": 6.07,
      "grad_norm": 8.187485694885254,
      "learning_rate": 7.867494824016564e-06,
      "loss": 0.1864,
      "step": 32230
    },
    {
      "epoch": 6.07,
      "grad_norm": 0.006011044140905142,
      "learning_rate": 7.863730472426126e-06,
      "loss": 0.2793,
      "step": 32240
    },
    {
      "epoch": 6.07,
      "grad_norm": 0.3004312515258789,
      "learning_rate": 7.859966120835687e-06,
      "loss": 0.0423,
      "step": 32250
    },
    {
      "epoch": 6.07,
      "grad_norm": 22.321874618530273,
      "learning_rate": 7.856201769245248e-06,
      "loss": 0.3117,
      "step": 32260
    },
    {
      "epoch": 6.07,
      "grad_norm": 2.7287864685058594,
      "learning_rate": 7.852437417654808e-06,
      "loss": 0.1218,
      "step": 32270
    },
    {
      "epoch": 6.08,
      "grad_norm": 43.60725402832031,
      "learning_rate": 7.848673066064371e-06,
      "loss": 0.2719,
      "step": 32280
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.6163210272789001,
      "learning_rate": 7.844908714473933e-06,
      "loss": 0.1151,
      "step": 32290
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.11249370872974396,
      "learning_rate": 7.841144362883494e-06,
      "loss": 0.1082,
      "step": 32300
    },
    {
      "epoch": 6.08,
      "grad_norm": 9.90340805053711,
      "learning_rate": 7.837380011293056e-06,
      "loss": 0.3271,
      "step": 32310
    },
    {
      "epoch": 6.08,
      "grad_norm": 0.32259848713874817,
      "learning_rate": 7.833615659702617e-06,
      "loss": 0.0718,
      "step": 32320
    },
    {
      "epoch": 6.09,
      "grad_norm": 6.870954513549805,
      "learning_rate": 7.829851308112179e-06,
      "loss": 0.1444,
      "step": 32330
    },
    {
      "epoch": 6.09,
      "grad_norm": 39.584938049316406,
      "learning_rate": 7.82608695652174e-06,
      "loss": 0.1652,
      "step": 32340
    },
    {
      "epoch": 6.09,
      "grad_norm": 9.253133773803711,
      "learning_rate": 7.822322604931302e-06,
      "loss": 0.2593,
      "step": 32350
    },
    {
      "epoch": 6.09,
      "grad_norm": 0.546807587146759,
      "learning_rate": 7.818558253340861e-06,
      "loss": 0.0632,
      "step": 32360
    },
    {
      "epoch": 6.09,
      "grad_norm": 12.067464828491211,
      "learning_rate": 7.814793901750425e-06,
      "loss": 0.151,
      "step": 32370
    },
    {
      "epoch": 6.09,
      "grad_norm": 7.653334617614746,
      "learning_rate": 7.811029550159986e-06,
      "loss": 0.2029,
      "step": 32380
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.005794860888272524,
      "learning_rate": 7.807265198569547e-06,
      "loss": 0.1498,
      "step": 32390
    },
    {
      "epoch": 6.1,
      "grad_norm": 36.327789306640625,
      "learning_rate": 7.803500846979109e-06,
      "loss": 0.2974,
      "step": 32400
    },
    {
      "epoch": 6.1,
      "grad_norm": 15.453861236572266,
      "learning_rate": 7.79973649538867e-06,
      "loss": 0.3446,
      "step": 32410
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.007539401762187481,
      "learning_rate": 7.795972143798232e-06,
      "loss": 0.1136,
      "step": 32420
    },
    {
      "epoch": 6.1,
      "grad_norm": 0.04490741342306137,
      "learning_rate": 7.792207792207793e-06,
      "loss": 0.34,
      "step": 32430
    },
    {
      "epoch": 6.11,
      "grad_norm": 25.18846893310547,
      "learning_rate": 7.788443440617355e-06,
      "loss": 0.1521,
      "step": 32440
    },
    {
      "epoch": 6.11,
      "grad_norm": 9.020671844482422,
      "learning_rate": 7.784679089026915e-06,
      "loss": 0.1885,
      "step": 32450
    },
    {
      "epoch": 6.11,
      "grad_norm": 142.74827575683594,
      "learning_rate": 7.780914737436478e-06,
      "loss": 0.3915,
      "step": 32460
    },
    {
      "epoch": 6.11,
      "grad_norm": 0.6103916764259338,
      "learning_rate": 7.777150385846039e-06,
      "loss": 0.2131,
      "step": 32470
    },
    {
      "epoch": 6.11,
      "grad_norm": 3.570294141769409,
      "learning_rate": 7.7733860342556e-06,
      "loss": 0.2806,
      "step": 32480
    },
    {
      "epoch": 6.12,
      "grad_norm": 9.565071105957031,
      "learning_rate": 7.769621682665162e-06,
      "loss": 0.1915,
      "step": 32490
    },
    {
      "epoch": 6.12,
      "grad_norm": 8.956217765808105,
      "learning_rate": 7.765857331074724e-06,
      "loss": 0.1787,
      "step": 32500
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.012830507010221481,
      "learning_rate": 7.762092979484285e-06,
      "loss": 0.1675,
      "step": 32510
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.13928885757923126,
      "learning_rate": 7.758328627893846e-06,
      "loss": 0.105,
      "step": 32520
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.6284078359603882,
      "learning_rate": 7.754564276303408e-06,
      "loss": 0.0569,
      "step": 32530
    },
    {
      "epoch": 6.12,
      "grad_norm": 0.013158991001546383,
      "learning_rate": 7.750799924712968e-06,
      "loss": 0.3265,
      "step": 32540
    },
    {
      "epoch": 6.13,
      "grad_norm": 1.298017978668213,
      "learning_rate": 7.747035573122529e-06,
      "loss": 0.0875,
      "step": 32550
    },
    {
      "epoch": 6.13,
      "grad_norm": 1.1146191358566284,
      "learning_rate": 7.743271221532092e-06,
      "loss": 0.329,
      "step": 32560
    },
    {
      "epoch": 6.13,
      "grad_norm": 0.07424399256706238,
      "learning_rate": 7.739506869941654e-06,
      "loss": 0.2612,
      "step": 32570
    },
    {
      "epoch": 6.13,
      "grad_norm": 2.0578954219818115,
      "learning_rate": 7.735742518351215e-06,
      "loss": 0.1382,
      "step": 32580
    },
    {
      "epoch": 6.13,
      "grad_norm": 2.8450489044189453,
      "learning_rate": 7.731978166760777e-06,
      "loss": 0.2337,
      "step": 32590
    },
    {
      "epoch": 6.14,
      "grad_norm": 15.191513061523438,
      "learning_rate": 7.728213815170338e-06,
      "loss": 0.1561,
      "step": 32600
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.0498182438313961,
      "learning_rate": 7.7244494635799e-06,
      "loss": 0.1388,
      "step": 32610
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.789818286895752,
      "learning_rate": 7.72068511198946e-06,
      "loss": 0.2002,
      "step": 32620
    },
    {
      "epoch": 6.14,
      "grad_norm": 0.011287186294794083,
      "learning_rate": 7.71692076039902e-06,
      "loss": 0.2178,
      "step": 32630
    },
    {
      "epoch": 6.14,
      "grad_norm": 4.630227565765381,
      "learning_rate": 7.713156408808582e-06,
      "loss": 0.2725,
      "step": 32640
    },
    {
      "epoch": 6.15,
      "grad_norm": 1.5468565225601196,
      "learning_rate": 7.709392057218145e-06,
      "loss": 0.2132,
      "step": 32650
    },
    {
      "epoch": 6.15,
      "grad_norm": 11.751218795776367,
      "learning_rate": 7.705627705627707e-06,
      "loss": 0.1028,
      "step": 32660
    },
    {
      "epoch": 6.15,
      "grad_norm": 2.1048460006713867,
      "learning_rate": 7.701863354037268e-06,
      "loss": 0.2284,
      "step": 32670
    },
    {
      "epoch": 6.15,
      "grad_norm": 18.849855422973633,
      "learning_rate": 7.69809900244683e-06,
      "loss": 0.0947,
      "step": 32680
    },
    {
      "epoch": 6.15,
      "grad_norm": 27.001802444458008,
      "learning_rate": 7.694334650856391e-06,
      "loss": 0.1903,
      "step": 32690
    },
    {
      "epoch": 6.15,
      "grad_norm": 10.998465538024902,
      "learning_rate": 7.690570299265953e-06,
      "loss": 0.3557,
      "step": 32700
    },
    {
      "epoch": 6.16,
      "grad_norm": 15.642669677734375,
      "learning_rate": 7.686805947675512e-06,
      "loss": 0.5379,
      "step": 32710
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.16382291913032532,
      "learning_rate": 7.683041596085074e-06,
      "loss": 0.0975,
      "step": 32720
    },
    {
      "epoch": 6.16,
      "grad_norm": 1.909355878829956,
      "learning_rate": 7.679277244494635e-06,
      "loss": 0.0831,
      "step": 32730
    },
    {
      "epoch": 6.16,
      "grad_norm": 18.203125,
      "learning_rate": 7.675512892904199e-06,
      "loss": 0.2104,
      "step": 32740
    },
    {
      "epoch": 6.16,
      "grad_norm": 0.060304295271635056,
      "learning_rate": 7.67174854131376e-06,
      "loss": 0.1364,
      "step": 32750
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.5184149146080017,
      "learning_rate": 7.667984189723321e-06,
      "loss": 0.0918,
      "step": 32760
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.45602720975875854,
      "learning_rate": 7.664219838132883e-06,
      "loss": 0.1221,
      "step": 32770
    },
    {
      "epoch": 6.17,
      "grad_norm": 5.950124263763428,
      "learning_rate": 7.660455486542444e-06,
      "loss": 0.1409,
      "step": 32780
    },
    {
      "epoch": 6.17,
      "grad_norm": 8.826350212097168,
      "learning_rate": 7.656691134952006e-06,
      "loss": 0.1224,
      "step": 32790
    },
    {
      "epoch": 6.17,
      "grad_norm": 0.004388615023344755,
      "learning_rate": 7.652926783361566e-06,
      "loss": 0.1772,
      "step": 32800
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.5683454275131226,
      "learning_rate": 7.649162431771127e-06,
      "loss": 0.1878,
      "step": 32810
    },
    {
      "epoch": 6.18,
      "grad_norm": 5.83488130569458,
      "learning_rate": 7.645398080180689e-06,
      "loss": 0.2699,
      "step": 32820
    },
    {
      "epoch": 6.18,
      "grad_norm": 14.224485397338867,
      "learning_rate": 7.641633728590252e-06,
      "loss": 0.3631,
      "step": 32830
    },
    {
      "epoch": 6.18,
      "grad_norm": 8.898087501525879,
      "learning_rate": 7.637869376999813e-06,
      "loss": 0.3195,
      "step": 32840
    },
    {
      "epoch": 6.18,
      "grad_norm": 1.7417521476745605,
      "learning_rate": 7.634105025409375e-06,
      "loss": 0.2201,
      "step": 32850
    },
    {
      "epoch": 6.18,
      "grad_norm": 0.009408744983375072,
      "learning_rate": 7.630340673818936e-06,
      "loss": 0.1209,
      "step": 32860
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.007678693160414696,
      "learning_rate": 7.6265763222284976e-06,
      "loss": 0.2692,
      "step": 32870
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.014336693100631237,
      "learning_rate": 7.622811970638057e-06,
      "loss": 0.0256,
      "step": 32880
    },
    {
      "epoch": 6.19,
      "grad_norm": 14.741891860961914,
      "learning_rate": 7.61904761904762e-06,
      "loss": 0.1638,
      "step": 32890
    },
    {
      "epoch": 6.19,
      "grad_norm": 0.27727121114730835,
      "learning_rate": 7.615283267457181e-06,
      "loss": 0.148,
      "step": 32900
    },
    {
      "epoch": 6.19,
      "grad_norm": 21.95084571838379,
      "learning_rate": 7.6115189158667426e-06,
      "loss": 0.2251,
      "step": 32910
    },
    {
      "epoch": 6.2,
      "grad_norm": 4.565463066101074,
      "learning_rate": 7.607754564276304e-06,
      "loss": 0.0549,
      "step": 32920
    },
    {
      "epoch": 6.2,
      "grad_norm": 27.1176815032959,
      "learning_rate": 7.6039902126858655e-06,
      "loss": 0.2184,
      "step": 32930
    },
    {
      "epoch": 6.2,
      "grad_norm": 12.660545349121094,
      "learning_rate": 7.600225861095427e-06,
      "loss": 0.0808,
      "step": 32940
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.009280041791498661,
      "learning_rate": 7.596461509504988e-06,
      "loss": 0.1276,
      "step": 32950
    },
    {
      "epoch": 6.2,
      "grad_norm": 12.521551132202148,
      "learning_rate": 7.592697157914551e-06,
      "loss": 0.2868,
      "step": 32960
    },
    {
      "epoch": 6.21,
      "grad_norm": 27.23976707458496,
      "learning_rate": 7.5889328063241105e-06,
      "loss": 0.1869,
      "step": 32970
    },
    {
      "epoch": 6.21,
      "grad_norm": 2.1055068969726562,
      "learning_rate": 7.585168454733673e-06,
      "loss": 0.0888,
      "step": 32980
    },
    {
      "epoch": 6.21,
      "grad_norm": 4.733766555786133,
      "learning_rate": 7.581404103143234e-06,
      "loss": 0.2759,
      "step": 32990
    },
    {
      "epoch": 6.21,
      "grad_norm": 10.490496635437012,
      "learning_rate": 7.577639751552796e-06,
      "loss": 0.2793,
      "step": 33000
    },
    {
      "epoch": 6.21,
      "grad_norm": 6.138078212738037,
      "learning_rate": 7.573875399962357e-06,
      "loss": 0.2049,
      "step": 33010
    },
    {
      "epoch": 6.21,
      "grad_norm": 3.176279067993164,
      "learning_rate": 7.570111048371919e-06,
      "loss": 0.2819,
      "step": 33020
    },
    {
      "epoch": 6.22,
      "grad_norm": 0.531818687915802,
      "learning_rate": 7.56634669678148e-06,
      "loss": 0.126,
      "step": 33030
    },
    {
      "epoch": 6.22,
      "grad_norm": 12.691883087158203,
      "learning_rate": 7.5625823451910415e-06,
      "loss": 0.1349,
      "step": 33040
    },
    {
      "epoch": 6.22,
      "grad_norm": 18.203638076782227,
      "learning_rate": 7.558817993600603e-06,
      "loss": 0.0834,
      "step": 33050
    },
    {
      "epoch": 6.22,
      "grad_norm": 0.011891626752912998,
      "learning_rate": 7.555053642010164e-06,
      "loss": 0.2416,
      "step": 33060
    },
    {
      "epoch": 6.22,
      "grad_norm": 1.7532778978347778,
      "learning_rate": 7.551289290419725e-06,
      "loss": 0.1791,
      "step": 33070
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.008245915174484253,
      "learning_rate": 7.547524938829287e-06,
      "loss": 0.2675,
      "step": 33080
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.03823555260896683,
      "learning_rate": 7.543760587238849e-06,
      "loss": 0.0171,
      "step": 33090
    },
    {
      "epoch": 6.23,
      "grad_norm": 3.9749739170074463,
      "learning_rate": 7.53999623564841e-06,
      "loss": 0.2163,
      "step": 33100
    },
    {
      "epoch": 6.23,
      "grad_norm": 4.280645847320557,
      "learning_rate": 7.536231884057972e-06,
      "loss": 0.1331,
      "step": 33110
    },
    {
      "epoch": 6.23,
      "grad_norm": 0.02994379773736,
      "learning_rate": 7.532467532467533e-06,
      "loss": 0.1546,
      "step": 33120
    },
    {
      "epoch": 6.24,
      "grad_norm": 2.4533801078796387,
      "learning_rate": 7.528703180877095e-06,
      "loss": 0.1665,
      "step": 33130
    },
    {
      "epoch": 6.24,
      "grad_norm": 8.12675666809082,
      "learning_rate": 7.524938829286656e-06,
      "loss": 0.3595,
      "step": 33140
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.010040095075964928,
      "learning_rate": 7.521174477696217e-06,
      "loss": 0.2651,
      "step": 33150
    },
    {
      "epoch": 6.24,
      "grad_norm": 9.189531326293945,
      "learning_rate": 7.517410126105778e-06,
      "loss": 0.2037,
      "step": 33160
    },
    {
      "epoch": 6.24,
      "grad_norm": 0.691264271736145,
      "learning_rate": 7.5136457745153405e-06,
      "loss": 0.2297,
      "step": 33170
    },
    {
      "epoch": 6.25,
      "grad_norm": 10.153438568115234,
      "learning_rate": 7.509881422924902e-06,
      "loss": 0.1548,
      "step": 33180
    },
    {
      "epoch": 6.25,
      "grad_norm": 17.05594253540039,
      "learning_rate": 7.5061170713344635e-06,
      "loss": 0.1857,
      "step": 33190
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.010721084661781788,
      "learning_rate": 7.502352719744025e-06,
      "loss": 0.0768,
      "step": 33200
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.012493197806179523,
      "learning_rate": 7.498588368153586e-06,
      "loss": 0.2203,
      "step": 33210
    },
    {
      "epoch": 6.25,
      "grad_norm": 8.492493629455566,
      "learning_rate": 7.494824016563148e-06,
      "loss": 0.2931,
      "step": 33220
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.5157361626625061,
      "learning_rate": 7.4910596649727084e-06,
      "loss": 0.3162,
      "step": 33230
    },
    {
      "epoch": 6.26,
      "grad_norm": 33.9202880859375,
      "learning_rate": 7.48729531338227e-06,
      "loss": 0.4978,
      "step": 33240
    },
    {
      "epoch": 6.26,
      "grad_norm": 4.415780067443848,
      "learning_rate": 7.483530961791831e-06,
      "loss": 0.1853,
      "step": 33250
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.41647252440452576,
      "learning_rate": 7.479766610201394e-06,
      "loss": 0.1219,
      "step": 33260
    },
    {
      "epoch": 6.26,
      "grad_norm": 0.003755794372409582,
      "learning_rate": 7.476002258610955e-06,
      "loss": 0.0544,
      "step": 33270
    },
    {
      "epoch": 6.26,
      "grad_norm": 18.30587387084961,
      "learning_rate": 7.472237907020517e-06,
      "loss": 0.3684,
      "step": 33280
    },
    {
      "epoch": 6.27,
      "grad_norm": 18.167831420898438,
      "learning_rate": 7.468473555430078e-06,
      "loss": 0.2457,
      "step": 33290
    },
    {
      "epoch": 6.27,
      "grad_norm": 13.5750093460083,
      "learning_rate": 7.4647092038396395e-06,
      "loss": 0.1637,
      "step": 33300
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.6159980893135071,
      "learning_rate": 7.460944852249201e-06,
      "loss": 0.0868,
      "step": 33310
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.005957382265478373,
      "learning_rate": 7.457180500658762e-06,
      "loss": 0.117,
      "step": 33320
    },
    {
      "epoch": 6.27,
      "grad_norm": 0.005731220357120037,
      "learning_rate": 7.453416149068323e-06,
      "loss": 0.0331,
      "step": 33330
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.005144340451806784,
      "learning_rate": 7.4496517974778845e-06,
      "loss": 0.1449,
      "step": 33340
    },
    {
      "epoch": 6.28,
      "grad_norm": 7.810390949249268,
      "learning_rate": 7.445887445887446e-06,
      "loss": 0.1084,
      "step": 33350
    },
    {
      "epoch": 6.28,
      "grad_norm": 33.93387985229492,
      "learning_rate": 7.442123094297008e-06,
      "loss": 0.2389,
      "step": 33360
    },
    {
      "epoch": 6.28,
      "grad_norm": 1.1177371740341187,
      "learning_rate": 7.43835874270657e-06,
      "loss": 0.1105,
      "step": 33370
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.5171994566917419,
      "learning_rate": 7.434594391116131e-06,
      "loss": 0.333,
      "step": 33380
    },
    {
      "epoch": 6.28,
      "grad_norm": 0.385047972202301,
      "learning_rate": 7.430830039525693e-06,
      "loss": 0.1762,
      "step": 33390
    },
    {
      "epoch": 6.29,
      "grad_norm": 5.207344055175781,
      "learning_rate": 7.427065687935254e-06,
      "loss": 0.4671,
      "step": 33400
    },
    {
      "epoch": 6.29,
      "grad_norm": 3.675480604171753,
      "learning_rate": 7.423301336344815e-06,
      "loss": 0.1021,
      "step": 33410
    },
    {
      "epoch": 6.29,
      "grad_norm": 0.7879157662391663,
      "learning_rate": 7.419536984754376e-06,
      "loss": 0.2517,
      "step": 33420
    },
    {
      "epoch": 6.29,
      "grad_norm": 2.9064714908599854,
      "learning_rate": 7.415772633163938e-06,
      "loss": 0.1389,
      "step": 33430
    },
    {
      "epoch": 6.29,
      "grad_norm": 8.216279983520508,
      "learning_rate": 7.412008281573499e-06,
      "loss": 0.4367,
      "step": 33440
    },
    {
      "epoch": 6.3,
      "grad_norm": 11.451451301574707,
      "learning_rate": 7.4082439299830614e-06,
      "loss": 0.0675,
      "step": 33450
    },
    {
      "epoch": 6.3,
      "grad_norm": 5.495248794555664,
      "learning_rate": 7.404479578392623e-06,
      "loss": 0.2166,
      "step": 33460
    },
    {
      "epoch": 6.3,
      "grad_norm": 11.266733169555664,
      "learning_rate": 7.400715226802184e-06,
      "loss": 0.1589,
      "step": 33470
    },
    {
      "epoch": 6.3,
      "grad_norm": 5.522986888885498,
      "learning_rate": 7.396950875211746e-06,
      "loss": 0.1446,
      "step": 33480
    },
    {
      "epoch": 6.3,
      "grad_norm": 9.521207809448242,
      "learning_rate": 7.3931865236213064e-06,
      "loss": 0.2064,
      "step": 33490
    },
    {
      "epoch": 6.31,
      "grad_norm": 6.835692882537842,
      "learning_rate": 7.389422172030868e-06,
      "loss": 0.0831,
      "step": 33500
    },
    {
      "epoch": 6.31,
      "grad_norm": 1.175052285194397,
      "learning_rate": 7.385657820440429e-06,
      "loss": 0.3264,
      "step": 33510
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.006890713237226009,
      "learning_rate": 7.381893468849991e-06,
      "loss": 0.2206,
      "step": 33520
    },
    {
      "epoch": 6.31,
      "grad_norm": 0.0346723273396492,
      "learning_rate": 7.378129117259552e-06,
      "loss": 0.1901,
      "step": 33530
    },
    {
      "epoch": 6.31,
      "grad_norm": 2.298034191131592,
      "learning_rate": 7.3743647656691146e-06,
      "loss": 0.1313,
      "step": 33540
    },
    {
      "epoch": 6.31,
      "grad_norm": 4.210710525512695,
      "learning_rate": 7.370600414078676e-06,
      "loss": 0.409,
      "step": 33550
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.019281912595033646,
      "learning_rate": 7.3668360624882375e-06,
      "loss": 0.048,
      "step": 33560
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.03075498528778553,
      "learning_rate": 7.363071710897799e-06,
      "loss": 0.2481,
      "step": 33570
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.006136924494057894,
      "learning_rate": 7.3593073593073596e-06,
      "loss": 0.1589,
      "step": 33580
    },
    {
      "epoch": 6.32,
      "grad_norm": 11.766714096069336,
      "learning_rate": 7.355543007716921e-06,
      "loss": 0.1367,
      "step": 33590
    },
    {
      "epoch": 6.32,
      "grad_norm": 4.149861812591553,
      "learning_rate": 7.3517786561264825e-06,
      "loss": 0.1203,
      "step": 33600
    },
    {
      "epoch": 6.33,
      "grad_norm": 53.780887603759766,
      "learning_rate": 7.348014304536044e-06,
      "loss": 0.1481,
      "step": 33610
    },
    {
      "epoch": 6.33,
      "grad_norm": 13.354452133178711,
      "learning_rate": 7.344249952945605e-06,
      "loss": 0.1505,
      "step": 33620
    },
    {
      "epoch": 6.33,
      "grad_norm": 5.555115699768066,
      "learning_rate": 7.340485601355168e-06,
      "loss": 0.4317,
      "step": 33630
    },
    {
      "epoch": 6.33,
      "grad_norm": 0.08710916340351105,
      "learning_rate": 7.336721249764729e-06,
      "loss": 0.0743,
      "step": 33640
    },
    {
      "epoch": 6.33,
      "grad_norm": 3.1929399967193604,
      "learning_rate": 7.332956898174291e-06,
      "loss": 0.1753,
      "step": 33650
    },
    {
      "epoch": 6.34,
      "grad_norm": 0.1650117039680481,
      "learning_rate": 7.329192546583852e-06,
      "loss": 0.2272,
      "step": 33660
    },
    {
      "epoch": 6.34,
      "grad_norm": 11.5098237991333,
      "learning_rate": 7.325428194993413e-06,
      "loss": 0.2233,
      "step": 33670
    },
    {
      "epoch": 6.34,
      "grad_norm": 1.5177030563354492,
      "learning_rate": 7.321663843402974e-06,
      "loss": 0.2518,
      "step": 33680
    },
    {
      "epoch": 6.34,
      "grad_norm": 22.30908966064453,
      "learning_rate": 7.317899491812536e-06,
      "loss": 0.3899,
      "step": 33690
    },
    {
      "epoch": 6.34,
      "grad_norm": 12.805367469787598,
      "learning_rate": 7.314135140222097e-06,
      "loss": 0.33,
      "step": 33700
    },
    {
      "epoch": 6.34,
      "grad_norm": 10.133213996887207,
      "learning_rate": 7.3103707886316586e-06,
      "loss": 0.1466,
      "step": 33710
    },
    {
      "epoch": 6.35,
      "grad_norm": 5.361339569091797,
      "learning_rate": 7.30660643704122e-06,
      "loss": 0.1604,
      "step": 33720
    },
    {
      "epoch": 6.35,
      "grad_norm": 2.007650852203369,
      "learning_rate": 7.302842085450782e-06,
      "loss": 0.1714,
      "step": 33730
    },
    {
      "epoch": 6.35,
      "grad_norm": 27.39136505126953,
      "learning_rate": 7.299077733860344e-06,
      "loss": 0.3605,
      "step": 33740
    },
    {
      "epoch": 6.35,
      "grad_norm": 4.918207168579102,
      "learning_rate": 7.295313382269905e-06,
      "loss": 0.3357,
      "step": 33750
    },
    {
      "epoch": 6.35,
      "grad_norm": 1.3677923679351807,
      "learning_rate": 7.291549030679466e-06,
      "loss": 0.1061,
      "step": 33760
    },
    {
      "epoch": 6.36,
      "grad_norm": 7.9233808517456055,
      "learning_rate": 7.287784679089027e-06,
      "loss": 0.0599,
      "step": 33770
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.8261010050773621,
      "learning_rate": 7.284020327498589e-06,
      "loss": 0.0958,
      "step": 33780
    },
    {
      "epoch": 6.36,
      "grad_norm": 7.440674781799316,
      "learning_rate": 7.28025597590815e-06,
      "loss": 0.0786,
      "step": 33790
    },
    {
      "epoch": 6.36,
      "grad_norm": 0.004310086835175753,
      "learning_rate": 7.276491624317712e-06,
      "loss": 0.2031,
      "step": 33800
    },
    {
      "epoch": 6.36,
      "grad_norm": 55.272911071777344,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.2403,
      "step": 33810
    },
    {
      "epoch": 6.37,
      "grad_norm": 6.272849082946777,
      "learning_rate": 7.2689629211368355e-06,
      "loss": 0.2182,
      "step": 33820
    },
    {
      "epoch": 6.37,
      "grad_norm": 18.738779067993164,
      "learning_rate": 7.265198569546397e-06,
      "loss": 0.2882,
      "step": 33830
    },
    {
      "epoch": 6.37,
      "grad_norm": 22.96047019958496,
      "learning_rate": 7.2614342179559576e-06,
      "loss": 0.1166,
      "step": 33840
    },
    {
      "epoch": 6.37,
      "grad_norm": 7.087310314178467,
      "learning_rate": 7.257669866365519e-06,
      "loss": 0.1822,
      "step": 33850
    },
    {
      "epoch": 6.37,
      "grad_norm": 6.78462028503418,
      "learning_rate": 7.2539055147750805e-06,
      "loss": 0.2669,
      "step": 33860
    },
    {
      "epoch": 6.37,
      "grad_norm": 0.008043105714023113,
      "learning_rate": 7.250141163184642e-06,
      "loss": 0.1167,
      "step": 33870
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.8138658404350281,
      "learning_rate": 7.246376811594203e-06,
      "loss": 0.0521,
      "step": 33880
    },
    {
      "epoch": 6.38,
      "grad_norm": 1.1843488216400146,
      "learning_rate": 7.242612460003765e-06,
      "loss": 0.3437,
      "step": 33890
    },
    {
      "epoch": 6.38,
      "grad_norm": 1.5273874998092651,
      "learning_rate": 7.238848108413326e-06,
      "loss": 0.1901,
      "step": 33900
    },
    {
      "epoch": 6.38,
      "grad_norm": 0.005417229142040014,
      "learning_rate": 7.235083756822889e-06,
      "loss": 0.1041,
      "step": 33910
    },
    {
      "epoch": 6.38,
      "grad_norm": 4.723695278167725,
      "learning_rate": 7.23131940523245e-06,
      "loss": 0.109,
      "step": 33920
    },
    {
      "epoch": 6.39,
      "grad_norm": 10.436178207397461,
      "learning_rate": 7.22755505364201e-06,
      "loss": 0.3508,
      "step": 33930
    },
    {
      "epoch": 6.39,
      "grad_norm": 8.801382064819336,
      "learning_rate": 7.223790702051572e-06,
      "loss": 0.2371,
      "step": 33940
    },
    {
      "epoch": 6.39,
      "grad_norm": 9.424591064453125,
      "learning_rate": 7.220026350461134e-06,
      "loss": 0.1741,
      "step": 33950
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.013317758217453957,
      "learning_rate": 7.216261998870695e-06,
      "loss": 0.396,
      "step": 33960
    },
    {
      "epoch": 6.39,
      "grad_norm": 0.12028107792139053,
      "learning_rate": 7.2124976472802565e-06,
      "loss": 0.1125,
      "step": 33970
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.014682059176266193,
      "learning_rate": 7.208733295689818e-06,
      "loss": 0.1415,
      "step": 33980
    },
    {
      "epoch": 6.4,
      "grad_norm": 5.054403781890869,
      "learning_rate": 7.2049689440993795e-06,
      "loss": 0.1923,
      "step": 33990
    },
    {
      "epoch": 6.4,
      "grad_norm": 0.014526083134114742,
      "learning_rate": 7.201204592508941e-06,
      "loss": 0.1294,
      "step": 34000
    },
    {
      "epoch": 6.4,
      "grad_norm": 29.636356353759766,
      "learning_rate": 7.197440240918503e-06,
      "loss": 0.2142,
      "step": 34010
    },
    {
      "epoch": 6.4,
      "grad_norm": 5.067316055297852,
      "learning_rate": 7.193675889328063e-06,
      "loss": 0.1601,
      "step": 34020
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.005441443063318729,
      "learning_rate": 7.189911537737625e-06,
      "loss": 0.189,
      "step": 34030
    },
    {
      "epoch": 6.41,
      "grad_norm": 2.4027280807495117,
      "learning_rate": 7.186147186147187e-06,
      "loss": 0.1877,
      "step": 34040
    },
    {
      "epoch": 6.41,
      "grad_norm": 0.557399570941925,
      "learning_rate": 7.182382834556748e-06,
      "loss": 0.3136,
      "step": 34050
    },
    {
      "epoch": 6.41,
      "grad_norm": 21.743656158447266,
      "learning_rate": 7.17861848296631e-06,
      "loss": 0.1709,
      "step": 34060
    },
    {
      "epoch": 6.41,
      "grad_norm": 7.738515853881836,
      "learning_rate": 7.174854131375871e-06,
      "loss": 0.1167,
      "step": 34070
    },
    {
      "epoch": 6.41,
      "grad_norm": 6.397573947906494,
      "learning_rate": 7.171089779785433e-06,
      "loss": 0.1096,
      "step": 34080
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.7749409675598145,
      "learning_rate": 7.167325428194994e-06,
      "loss": 0.1107,
      "step": 34090
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.06625822931528091,
      "learning_rate": 7.163561076604555e-06,
      "loss": 0.2071,
      "step": 34100
    },
    {
      "epoch": 6.42,
      "grad_norm": 0.006089507136493921,
      "learning_rate": 7.159796725014116e-06,
      "loss": 0.1548,
      "step": 34110
    },
    {
      "epoch": 6.42,
      "grad_norm": 1.1297001838684082,
      "learning_rate": 7.1560323734236784e-06,
      "loss": 0.3373,
      "step": 34120
    },
    {
      "epoch": 6.42,
      "grad_norm": 8.861988067626953,
      "learning_rate": 7.15226802183324e-06,
      "loss": 0.3658,
      "step": 34130
    },
    {
      "epoch": 6.43,
      "grad_norm": 12.718072891235352,
      "learning_rate": 7.148503670242801e-06,
      "loss": 0.2785,
      "step": 34140
    },
    {
      "epoch": 6.43,
      "grad_norm": 10.766060829162598,
      "learning_rate": 7.144739318652363e-06,
      "loss": 0.1251,
      "step": 34150
    },
    {
      "epoch": 6.43,
      "grad_norm": 6.073504447937012,
      "learning_rate": 7.140974967061924e-06,
      "loss": 0.1653,
      "step": 34160
    },
    {
      "epoch": 6.43,
      "grad_norm": 21.271453857421875,
      "learning_rate": 7.137210615471486e-06,
      "loss": 0.0963,
      "step": 34170
    },
    {
      "epoch": 6.43,
      "grad_norm": 0.012385723181068897,
      "learning_rate": 7.133446263881047e-06,
      "loss": 0.1349,
      "step": 34180
    },
    {
      "epoch": 6.44,
      "grad_norm": 0.28191304206848145,
      "learning_rate": 7.129681912290608e-06,
      "loss": 0.1425,
      "step": 34190
    },
    {
      "epoch": 6.44,
      "grad_norm": 0.01502896472811699,
      "learning_rate": 7.125917560700169e-06,
      "loss": 0.318,
      "step": 34200
    },
    {
      "epoch": 6.44,
      "grad_norm": 3.8576014041900635,
      "learning_rate": 7.122153209109732e-06,
      "loss": 0.1965,
      "step": 34210
    },
    {
      "epoch": 6.44,
      "grad_norm": 21.344083786010742,
      "learning_rate": 7.118388857519293e-06,
      "loss": 0.3362,
      "step": 34220
    },
    {
      "epoch": 6.44,
      "grad_norm": 38.536373138427734,
      "learning_rate": 7.1146245059288545e-06,
      "loss": 0.0817,
      "step": 34230
    },
    {
      "epoch": 6.44,
      "grad_norm": 7.2858662605285645,
      "learning_rate": 7.110860154338416e-06,
      "loss": 0.1348,
      "step": 34240
    },
    {
      "epoch": 6.45,
      "grad_norm": 0.4228447675704956,
      "learning_rate": 7.1070958027479774e-06,
      "loss": 0.1352,
      "step": 34250
    },
    {
      "epoch": 6.45,
      "grad_norm": 3.5498600006103516,
      "learning_rate": 7.103331451157539e-06,
      "loss": 0.1883,
      "step": 34260
    },
    {
      "epoch": 6.45,
      "grad_norm": 1.1147761344909668,
      "learning_rate": 7.0995670995671e-06,
      "loss": 0.21,
      "step": 34270
    },
    {
      "epoch": 6.45,
      "grad_norm": 1.8100972175598145,
      "learning_rate": 7.095802747976661e-06,
      "loss": 0.1368,
      "step": 34280
    },
    {
      "epoch": 6.45,
      "grad_norm": 21.316986083984375,
      "learning_rate": 7.0920383963862224e-06,
      "loss": 0.1127,
      "step": 34290
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.14516809582710266,
      "learning_rate": 7.088274044795784e-06,
      "loss": 0.1124,
      "step": 34300
    },
    {
      "epoch": 6.46,
      "grad_norm": 3.6689908504486084,
      "learning_rate": 7.084509693205346e-06,
      "loss": 0.1858,
      "step": 34310
    },
    {
      "epoch": 6.46,
      "grad_norm": 0.015026669949293137,
      "learning_rate": 7.080745341614908e-06,
      "loss": 0.2424,
      "step": 34320
    },
    {
      "epoch": 6.46,
      "grad_norm": 50.35696029663086,
      "learning_rate": 7.076980990024469e-06,
      "loss": 0.3008,
      "step": 34330
    },
    {
      "epoch": 6.46,
      "grad_norm": 4.7025885581970215,
      "learning_rate": 7.073216638434031e-06,
      "loss": 0.3842,
      "step": 34340
    },
    {
      "epoch": 6.47,
      "grad_norm": 11.793999671936035,
      "learning_rate": 7.069452286843592e-06,
      "loss": 0.1987,
      "step": 34350
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.019232450053095818,
      "learning_rate": 7.0656879352531535e-06,
      "loss": 0.1266,
      "step": 34360
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.03133144974708557,
      "learning_rate": 7.061923583662714e-06,
      "loss": 0.0481,
      "step": 34370
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.012026842683553696,
      "learning_rate": 7.058159232072276e-06,
      "loss": 0.0997,
      "step": 34380
    },
    {
      "epoch": 6.47,
      "grad_norm": 0.012382052838802338,
      "learning_rate": 7.054394880481837e-06,
      "loss": 0.2791,
      "step": 34390
    },
    {
      "epoch": 6.47,
      "grad_norm": 10.344717979431152,
      "learning_rate": 7.050630528891399e-06,
      "loss": 0.1523,
      "step": 34400
    },
    {
      "epoch": 6.48,
      "grad_norm": 31.50136947631836,
      "learning_rate": 7.046866177300961e-06,
      "loss": 0.3091,
      "step": 34410
    },
    {
      "epoch": 6.48,
      "grad_norm": 1.0517066717147827,
      "learning_rate": 7.043101825710522e-06,
      "loss": 0.1413,
      "step": 34420
    },
    {
      "epoch": 6.48,
      "grad_norm": 0.20899175107479095,
      "learning_rate": 7.039337474120084e-06,
      "loss": 0.0787,
      "step": 34430
    },
    {
      "epoch": 6.48,
      "grad_norm": 3.0107178688049316,
      "learning_rate": 7.035573122529645e-06,
      "loss": 0.1353,
      "step": 34440
    },
    {
      "epoch": 6.48,
      "grad_norm": 5.831308841705322,
      "learning_rate": 7.031808770939206e-06,
      "loss": 0.1732,
      "step": 34450
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.05589495599269867,
      "learning_rate": 7.028044419348767e-06,
      "loss": 0.1139,
      "step": 34460
    },
    {
      "epoch": 6.49,
      "grad_norm": 0.0037380901630967855,
      "learning_rate": 7.024280067758329e-06,
      "loss": 0.1717,
      "step": 34470
    },
    {
      "epoch": 6.49,
      "grad_norm": 49.92671203613281,
      "learning_rate": 7.02051571616789e-06,
      "loss": 0.1382,
      "step": 34480
    },
    {
      "epoch": 6.49,
      "grad_norm": 43.958351135253906,
      "learning_rate": 7.0167513645774525e-06,
      "loss": 0.3517,
      "step": 34490
    },
    {
      "epoch": 6.49,
      "grad_norm": 16.357650756835938,
      "learning_rate": 7.012987012987014e-06,
      "loss": 0.288,
      "step": 34500
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.04773865267634392,
      "learning_rate": 7.009222661396575e-06,
      "loss": 0.1996,
      "step": 34510
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.010994565673172474,
      "learning_rate": 7.005458309806137e-06,
      "loss": 0.2472,
      "step": 34520
    },
    {
      "epoch": 6.5,
      "grad_norm": 17.143901824951172,
      "learning_rate": 7.001693958215698e-06,
      "loss": 0.1359,
      "step": 34530
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.04136435687541962,
      "learning_rate": 6.997929606625259e-06,
      "loss": 0.3072,
      "step": 34540
    },
    {
      "epoch": 6.5,
      "grad_norm": 3.3638510704040527,
      "learning_rate": 6.99416525503482e-06,
      "loss": 0.0657,
      "step": 34550
    },
    {
      "epoch": 6.5,
      "grad_norm": 0.30233657360076904,
      "learning_rate": 6.990400903444382e-06,
      "loss": 0.1688,
      "step": 34560
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.005954407621175051,
      "learning_rate": 6.986636551853943e-06,
      "loss": 0.1024,
      "step": 34570
    },
    {
      "epoch": 6.51,
      "grad_norm": 0.004195455461740494,
      "learning_rate": 6.982872200263505e-06,
      "loss": 0.2812,
      "step": 34580
    },
    {
      "epoch": 6.51,
      "grad_norm": 1.6767531633377075,
      "learning_rate": 6.979107848673067e-06,
      "loss": 0.2297,
      "step": 34590
    },
    {
      "epoch": 6.51,
      "grad_norm": 22.72420883178711,
      "learning_rate": 6.9753434970826286e-06,
      "loss": 0.1699,
      "step": 34600
    },
    {
      "epoch": 6.51,
      "grad_norm": 9.494135856628418,
      "learning_rate": 6.97157914549219e-06,
      "loss": 0.1964,
      "step": 34610
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.015217738226056099,
      "learning_rate": 6.9678147939017515e-06,
      "loss": 0.103,
      "step": 34620
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.010022526606917381,
      "learning_rate": 6.964050442311312e-06,
      "loss": 0.1105,
      "step": 34630
    },
    {
      "epoch": 6.52,
      "grad_norm": 9.281975746154785,
      "learning_rate": 6.9602860907208736e-06,
      "loss": 0.2062,
      "step": 34640
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.004425908904522657,
      "learning_rate": 6.956521739130435e-06,
      "loss": 0.2519,
      "step": 34650
    },
    {
      "epoch": 6.52,
      "grad_norm": 0.5666004419326782,
      "learning_rate": 6.9527573875399965e-06,
      "loss": 0.0367,
      "step": 34660
    },
    {
      "epoch": 6.53,
      "grad_norm": 14.099675178527832,
      "learning_rate": 6.948993035949558e-06,
      "loss": 0.2986,
      "step": 34670
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.021234363317489624,
      "learning_rate": 6.94522868435912e-06,
      "loss": 0.3401,
      "step": 34680
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.053624652326107025,
      "learning_rate": 6.941464332768682e-06,
      "loss": 0.0226,
      "step": 34690
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.08757674694061279,
      "learning_rate": 6.937699981178243e-06,
      "loss": 0.1375,
      "step": 34700
    },
    {
      "epoch": 6.53,
      "grad_norm": 0.017857970669865608,
      "learning_rate": 6.933935629587804e-06,
      "loss": 0.3195,
      "step": 34710
    },
    {
      "epoch": 6.53,
      "grad_norm": 4.985729694366455,
      "learning_rate": 6.930171277997365e-06,
      "loss": 0.0736,
      "step": 34720
    },
    {
      "epoch": 6.54,
      "grad_norm": 18.975849151611328,
      "learning_rate": 6.926406926406927e-06,
      "loss": 0.0708,
      "step": 34730
    },
    {
      "epoch": 6.54,
      "grad_norm": 0.05405601114034653,
      "learning_rate": 6.922642574816488e-06,
      "loss": 0.0227,
      "step": 34740
    },
    {
      "epoch": 6.54,
      "grad_norm": 16.134872436523438,
      "learning_rate": 6.91887822322605e-06,
      "loss": 0.1744,
      "step": 34750
    },
    {
      "epoch": 6.54,
      "grad_norm": 11.69938850402832,
      "learning_rate": 6.915113871635611e-06,
      "loss": 0.2777,
      "step": 34760
    },
    {
      "epoch": 6.54,
      "grad_norm": 0.008972265757620335,
      "learning_rate": 6.911349520045173e-06,
      "loss": 0.2021,
      "step": 34770
    },
    {
      "epoch": 6.55,
      "grad_norm": 21.85778045654297,
      "learning_rate": 6.907585168454735e-06,
      "loss": 0.2445,
      "step": 34780
    },
    {
      "epoch": 6.55,
      "grad_norm": 17.154335021972656,
      "learning_rate": 6.903820816864296e-06,
      "loss": 0.0852,
      "step": 34790
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.04921324923634529,
      "learning_rate": 6.900056465273857e-06,
      "loss": 0.2367,
      "step": 34800
    },
    {
      "epoch": 6.55,
      "grad_norm": 5.959379196166992,
      "learning_rate": 6.896292113683418e-06,
      "loss": 0.1985,
      "step": 34810
    },
    {
      "epoch": 6.55,
      "grad_norm": 0.3107336163520813,
      "learning_rate": 6.89252776209298e-06,
      "loss": 0.3237,
      "step": 34820
    },
    {
      "epoch": 6.56,
      "grad_norm": 0.017287978902459145,
      "learning_rate": 6.888763410502541e-06,
      "loss": 0.131,
      "step": 34830
    },
    {
      "epoch": 6.56,
      "grad_norm": 9.368523597717285,
      "learning_rate": 6.884999058912103e-06,
      "loss": 0.4686,
      "step": 34840
    },
    {
      "epoch": 6.56,
      "grad_norm": 0.04478839039802551,
      "learning_rate": 6.881234707321664e-06,
      "loss": 0.2685,
      "step": 34850
    },
    {
      "epoch": 6.56,
      "grad_norm": 4.961436748504639,
      "learning_rate": 6.8774703557312265e-06,
      "loss": 0.2084,
      "step": 34860
    },
    {
      "epoch": 6.56,
      "grad_norm": 11.356475830078125,
      "learning_rate": 6.873706004140788e-06,
      "loss": 0.0784,
      "step": 34870
    },
    {
      "epoch": 6.57,
      "grad_norm": 25.798919677734375,
      "learning_rate": 6.8699416525503495e-06,
      "loss": 0.2003,
      "step": 34880
    },
    {
      "epoch": 6.57,
      "grad_norm": 0.0043741860426962376,
      "learning_rate": 6.86617730095991e-06,
      "loss": 0.1415,
      "step": 34890
    },
    {
      "epoch": 6.57,
      "grad_norm": 13.247472763061523,
      "learning_rate": 6.8624129493694715e-06,
      "loss": 0.2257,
      "step": 34900
    },
    {
      "epoch": 6.57,
      "grad_norm": 21.602872848510742,
      "learning_rate": 6.858648597779033e-06,
      "loss": 0.1386,
      "step": 34910
    },
    {
      "epoch": 6.57,
      "grad_norm": 3.7519004344940186,
      "learning_rate": 6.8548842461885945e-06,
      "loss": 0.1936,
      "step": 34920
    },
    {
      "epoch": 6.57,
      "grad_norm": 15.357027053833008,
      "learning_rate": 6.851119894598156e-06,
      "loss": 0.216,
      "step": 34930
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.26526889204978943,
      "learning_rate": 6.847355543007717e-06,
      "loss": 0.134,
      "step": 34940
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.221304789185524,
      "learning_rate": 6.843591191417279e-06,
      "loss": 0.2773,
      "step": 34950
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.1353296935558319,
      "learning_rate": 6.839826839826841e-06,
      "loss": 0.2009,
      "step": 34960
    },
    {
      "epoch": 6.58,
      "grad_norm": 6.508285999298096,
      "learning_rate": 6.836062488236401e-06,
      "loss": 0.1253,
      "step": 34970
    },
    {
      "epoch": 6.58,
      "grad_norm": 0.019724830985069275,
      "learning_rate": 6.832298136645963e-06,
      "loss": 0.44,
      "step": 34980
    },
    {
      "epoch": 6.59,
      "grad_norm": 10.097959518432617,
      "learning_rate": 6.828533785055525e-06,
      "loss": 0.1864,
      "step": 34990
    },
    {
      "epoch": 6.59,
      "grad_norm": 4.429990291595459,
      "learning_rate": 6.824769433465086e-06,
      "loss": 0.1938,
      "step": 35000
    },
    {
      "epoch": 6.59,
      "grad_norm": 2.395200252532959,
      "learning_rate": 6.821005081874648e-06,
      "loss": 0.1799,
      "step": 35010
    },
    {
      "epoch": 6.59,
      "grad_norm": 0.00504745589569211,
      "learning_rate": 6.817240730284209e-06,
      "loss": 0.1087,
      "step": 35020
    },
    {
      "epoch": 6.59,
      "grad_norm": 6.203241348266602,
      "learning_rate": 6.8134763786937705e-06,
      "loss": 0.3631,
      "step": 35030
    },
    {
      "epoch": 6.6,
      "grad_norm": 0.18260100483894348,
      "learning_rate": 6.809712027103332e-06,
      "loss": 0.061,
      "step": 35040
    },
    {
      "epoch": 6.6,
      "grad_norm": 8.52432632446289,
      "learning_rate": 6.805947675512894e-06,
      "loss": 0.0645,
      "step": 35050
    },
    {
      "epoch": 6.6,
      "grad_norm": 14.328934669494629,
      "learning_rate": 6.802183323922454e-06,
      "loss": 0.2584,
      "step": 35060
    },
    {
      "epoch": 6.6,
      "grad_norm": 2.188633441925049,
      "learning_rate": 6.798418972332016e-06,
      "loss": 0.1731,
      "step": 35070
    },
    {
      "epoch": 6.6,
      "grad_norm": 3.8024954795837402,
      "learning_rate": 6.794654620741578e-06,
      "loss": 0.0109,
      "step": 35080
    },
    {
      "epoch": 6.6,
      "grad_norm": 3.058250665664673,
      "learning_rate": 6.790890269151139e-06,
      "loss": 0.2189,
      "step": 35090
    },
    {
      "epoch": 6.61,
      "grad_norm": 15.518731117248535,
      "learning_rate": 6.787125917560701e-06,
      "loss": 0.1135,
      "step": 35100
    },
    {
      "epoch": 6.61,
      "grad_norm": 19.791894912719727,
      "learning_rate": 6.783361565970262e-06,
      "loss": 0.2458,
      "step": 35110
    },
    {
      "epoch": 6.61,
      "grad_norm": 27.48121452331543,
      "learning_rate": 6.779597214379824e-06,
      "loss": 0.1116,
      "step": 35120
    },
    {
      "epoch": 6.61,
      "grad_norm": 8.906675338745117,
      "learning_rate": 6.775832862789385e-06,
      "loss": 0.3224,
      "step": 35130
    },
    {
      "epoch": 6.61,
      "grad_norm": 0.004278236534446478,
      "learning_rate": 6.7720685111989474e-06,
      "loss": 0.1691,
      "step": 35140
    },
    {
      "epoch": 6.62,
      "grad_norm": 29.998615264892578,
      "learning_rate": 6.768304159608507e-06,
      "loss": 0.1781,
      "step": 35150
    },
    {
      "epoch": 6.62,
      "grad_norm": 26.94253921508789,
      "learning_rate": 6.764539808018069e-06,
      "loss": 0.1376,
      "step": 35160
    },
    {
      "epoch": 6.62,
      "grad_norm": 43.9926643371582,
      "learning_rate": 6.760775456427631e-06,
      "loss": 0.2374,
      "step": 35170
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.05498330295085907,
      "learning_rate": 6.7570111048371924e-06,
      "loss": 0.0885,
      "step": 35180
    },
    {
      "epoch": 6.62,
      "grad_norm": 0.009778610430657864,
      "learning_rate": 6.753246753246754e-06,
      "loss": 0.3403,
      "step": 35190
    },
    {
      "epoch": 6.63,
      "grad_norm": 6.606904983520508,
      "learning_rate": 6.749482401656315e-06,
      "loss": 0.168,
      "step": 35200
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.004804588388651609,
      "learning_rate": 6.745718050065877e-06,
      "loss": 0.1156,
      "step": 35210
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.004437767900526524,
      "learning_rate": 6.741953698475438e-06,
      "loss": 0.259,
      "step": 35220
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.009007365442812443,
      "learning_rate": 6.738189346885e-06,
      "loss": 0.2081,
      "step": 35230
    },
    {
      "epoch": 6.63,
      "grad_norm": 11.829503059387207,
      "learning_rate": 6.73442499529456e-06,
      "loss": 0.073,
      "step": 35240
    },
    {
      "epoch": 6.63,
      "grad_norm": 0.09758754819631577,
      "learning_rate": 6.730660643704122e-06,
      "loss": 0.3146,
      "step": 35250
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.00381456152535975,
      "learning_rate": 6.726896292113684e-06,
      "loss": 0.0523,
      "step": 35260
    },
    {
      "epoch": 6.64,
      "grad_norm": 3.5328986644744873,
      "learning_rate": 6.723131940523246e-06,
      "loss": 0.2217,
      "step": 35270
    },
    {
      "epoch": 6.64,
      "grad_norm": 0.020386753603816032,
      "learning_rate": 6.719367588932807e-06,
      "loss": 0.0941,
      "step": 35280
    },
    {
      "epoch": 6.64,
      "grad_norm": 8.508111000061035,
      "learning_rate": 6.7156032373423685e-06,
      "loss": 0.2189,
      "step": 35290
    },
    {
      "epoch": 6.64,
      "grad_norm": 10.149103164672852,
      "learning_rate": 6.71183888575193e-06,
      "loss": 0.2313,
      "step": 35300
    },
    {
      "epoch": 6.65,
      "grad_norm": 14.21928596496582,
      "learning_rate": 6.7080745341614914e-06,
      "loss": 0.4364,
      "step": 35310
    },
    {
      "epoch": 6.65,
      "grad_norm": 0.0921817272901535,
      "learning_rate": 6.704310182571052e-06,
      "loss": 0.0866,
      "step": 35320
    },
    {
      "epoch": 6.65,
      "grad_norm": 4.239680290222168,
      "learning_rate": 6.7005458309806135e-06,
      "loss": 0.3377,
      "step": 35330
    },
    {
      "epoch": 6.65,
      "grad_norm": 0.30783897638320923,
      "learning_rate": 6.696781479390175e-06,
      "loss": 0.1448,
      "step": 35340
    },
    {
      "epoch": 6.65,
      "grad_norm": 22.301450729370117,
      "learning_rate": 6.693017127799737e-06,
      "loss": 0.2532,
      "step": 35350
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.5305352807044983,
      "learning_rate": 6.689252776209299e-06,
      "loss": 0.059,
      "step": 35360
    },
    {
      "epoch": 6.66,
      "grad_norm": 4.291310787200928,
      "learning_rate": 6.68548842461886e-06,
      "loss": 0.2113,
      "step": 35370
    },
    {
      "epoch": 6.66,
      "grad_norm": 21.63286018371582,
      "learning_rate": 6.681724073028422e-06,
      "loss": 0.205,
      "step": 35380
    },
    {
      "epoch": 6.66,
      "grad_norm": 30.439868927001953,
      "learning_rate": 6.677959721437983e-06,
      "loss": 0.5876,
      "step": 35390
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.0071540274657309055,
      "learning_rate": 6.6741953698475446e-06,
      "loss": 0.1576,
      "step": 35400
    },
    {
      "epoch": 6.66,
      "grad_norm": 0.005743447691202164,
      "learning_rate": 6.670431018257105e-06,
      "loss": 0.2129,
      "step": 35410
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.013085199519991875,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2839,
      "step": 35420
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.02649598754942417,
      "learning_rate": 6.662902315076228e-06,
      "loss": 0.0644,
      "step": 35430
    },
    {
      "epoch": 6.67,
      "grad_norm": 0.014947297051548958,
      "learning_rate": 6.65913796348579e-06,
      "loss": 0.1928,
      "step": 35440
    },
    {
      "epoch": 6.67,
      "grad_norm": 43.11162185668945,
      "learning_rate": 6.655373611895352e-06,
      "loss": 0.2744,
      "step": 35450
    },
    {
      "epoch": 6.67,
      "grad_norm": 2.5682084560394287,
      "learning_rate": 6.651609260304913e-06,
      "loss": 0.0948,
      "step": 35460
    },
    {
      "epoch": 6.68,
      "grad_norm": 1.8677304983139038,
      "learning_rate": 6.647844908714475e-06,
      "loss": 0.0569,
      "step": 35470
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.0039545283652842045,
      "learning_rate": 6.644080557124036e-06,
      "loss": 0.1661,
      "step": 35480
    },
    {
      "epoch": 6.68,
      "grad_norm": 2.2204582691192627,
      "learning_rate": 6.640316205533598e-06,
      "loss": 0.2194,
      "step": 35490
    },
    {
      "epoch": 6.68,
      "grad_norm": 3.4814329147338867,
      "learning_rate": 6.636551853943158e-06,
      "loss": 0.1445,
      "step": 35500
    },
    {
      "epoch": 6.68,
      "grad_norm": 0.6565182209014893,
      "learning_rate": 6.63278750235272e-06,
      "loss": 0.1496,
      "step": 35510
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.023735351860523224,
      "learning_rate": 6.629023150762281e-06,
      "loss": 0.0355,
      "step": 35520
    },
    {
      "epoch": 6.69,
      "grad_norm": 0.006357759237289429,
      "learning_rate": 6.625258799171843e-06,
      "loss": 0.1675,
      "step": 35530
    },
    {
      "epoch": 6.69,
      "grad_norm": 7.086661338806152,
      "learning_rate": 6.621494447581405e-06,
      "loss": 0.1904,
      "step": 35540
    },
    {
      "epoch": 6.69,
      "grad_norm": 20.267166137695312,
      "learning_rate": 6.6177300959909665e-06,
      "loss": 0.1659,
      "step": 35550
    },
    {
      "epoch": 6.69,
      "grad_norm": 7.262717247009277,
      "learning_rate": 6.613965744400528e-06,
      "loss": 0.2645,
      "step": 35560
    },
    {
      "epoch": 6.69,
      "grad_norm": 11.095940589904785,
      "learning_rate": 6.610201392810089e-06,
      "loss": 0.302,
      "step": 35570
    },
    {
      "epoch": 6.7,
      "grad_norm": 10.50123119354248,
      "learning_rate": 6.60643704121965e-06,
      "loss": 0.3752,
      "step": 35580
    },
    {
      "epoch": 6.7,
      "grad_norm": 0.10886739194393158,
      "learning_rate": 6.6026726896292115e-06,
      "loss": 0.169,
      "step": 35590
    },
    {
      "epoch": 6.7,
      "grad_norm": 10.391016006469727,
      "learning_rate": 6.598908338038773e-06,
      "loss": 0.1556,
      "step": 35600
    },
    {
      "epoch": 6.7,
      "grad_norm": 3.517181158065796,
      "learning_rate": 6.595143986448334e-06,
      "loss": 0.2495,
      "step": 35610
    },
    {
      "epoch": 6.7,
      "grad_norm": 15.055655479431152,
      "learning_rate": 6.591379634857896e-06,
      "loss": 0.2641,
      "step": 35620
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.5330414772033691,
      "learning_rate": 6.587615283267458e-06,
      "loss": 0.2444,
      "step": 35630
    },
    {
      "epoch": 6.71,
      "grad_norm": 9.935203552246094,
      "learning_rate": 6.58385093167702e-06,
      "loss": 0.3317,
      "step": 35640
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.014888529665768147,
      "learning_rate": 6.580086580086581e-06,
      "loss": 0.1926,
      "step": 35650
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.011109587736427784,
      "learning_rate": 6.5763222284961426e-06,
      "loss": 0.2389,
      "step": 35660
    },
    {
      "epoch": 6.71,
      "grad_norm": 0.33202558755874634,
      "learning_rate": 6.572557876905703e-06,
      "loss": 0.0555,
      "step": 35670
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.016494670882821083,
      "learning_rate": 6.568793525315265e-06,
      "loss": 0.0652,
      "step": 35680
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.0178840234875679,
      "learning_rate": 6.565029173724826e-06,
      "loss": 0.3228,
      "step": 35690
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.12291597574949265,
      "learning_rate": 6.5612648221343875e-06,
      "loss": 0.1106,
      "step": 35700
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.027833105996251106,
      "learning_rate": 6.557500470543949e-06,
      "loss": 0.0938,
      "step": 35710
    },
    {
      "epoch": 6.72,
      "grad_norm": 0.0056888265535235405,
      "learning_rate": 6.553736118953511e-06,
      "loss": 0.1588,
      "step": 35720
    },
    {
      "epoch": 6.73,
      "grad_norm": 4.431260585784912,
      "learning_rate": 6.549971767363073e-06,
      "loss": 0.0754,
      "step": 35730
    },
    {
      "epoch": 6.73,
      "grad_norm": 10.571792602539062,
      "learning_rate": 6.546207415772634e-06,
      "loss": 0.1547,
      "step": 35740
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.01899634674191475,
      "learning_rate": 6.542443064182196e-06,
      "loss": 0.1224,
      "step": 35750
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.07985693961381912,
      "learning_rate": 6.538678712591756e-06,
      "loss": 0.1667,
      "step": 35760
    },
    {
      "epoch": 6.73,
      "grad_norm": 7.1994123458862305,
      "learning_rate": 6.534914361001318e-06,
      "loss": 0.1662,
      "step": 35770
    },
    {
      "epoch": 6.73,
      "grad_norm": 0.008827066980302334,
      "learning_rate": 6.531150009410879e-06,
      "loss": 0.1123,
      "step": 35780
    },
    {
      "epoch": 6.74,
      "grad_norm": 22.66560173034668,
      "learning_rate": 6.527385657820441e-06,
      "loss": 0.1312,
      "step": 35790
    },
    {
      "epoch": 6.74,
      "grad_norm": 15.479928016662598,
      "learning_rate": 6.523621306230002e-06,
      "loss": 0.3253,
      "step": 35800
    },
    {
      "epoch": 6.74,
      "grad_norm": 6.807658672332764,
      "learning_rate": 6.519856954639564e-06,
      "loss": 0.3274,
      "step": 35810
    },
    {
      "epoch": 6.74,
      "grad_norm": 0.00910489447414875,
      "learning_rate": 6.516092603049126e-06,
      "loss": 0.0103,
      "step": 35820
    },
    {
      "epoch": 6.74,
      "grad_norm": 2.0244410037994385,
      "learning_rate": 6.512328251458687e-06,
      "loss": 0.1069,
      "step": 35830
    },
    {
      "epoch": 6.75,
      "grad_norm": 6.395997047424316,
      "learning_rate": 6.508563899868249e-06,
      "loss": 0.3788,
      "step": 35840
    },
    {
      "epoch": 6.75,
      "grad_norm": 15.104719161987305,
      "learning_rate": 6.5047995482778095e-06,
      "loss": 0.0808,
      "step": 35850
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.6855412125587463,
      "learning_rate": 6.501035196687371e-06,
      "loss": 0.0785,
      "step": 35860
    },
    {
      "epoch": 6.75,
      "grad_norm": 0.03595370426774025,
      "learning_rate": 6.497270845096932e-06,
      "loss": 0.2297,
      "step": 35870
    },
    {
      "epoch": 6.75,
      "grad_norm": 8.083109855651855,
      "learning_rate": 6.493506493506494e-06,
      "loss": 0.1771,
      "step": 35880
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.004782949108630419,
      "learning_rate": 6.489742141916055e-06,
      "loss": 0.4012,
      "step": 35890
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.8231896758079529,
      "learning_rate": 6.485977790325617e-06,
      "loss": 0.2431,
      "step": 35900
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.009242299012839794,
      "learning_rate": 6.482213438735179e-06,
      "loss": 0.1008,
      "step": 35910
    },
    {
      "epoch": 6.76,
      "grad_norm": 2.611624240875244,
      "learning_rate": 6.4784490871447405e-06,
      "loss": 0.0738,
      "step": 35920
    },
    {
      "epoch": 6.76,
      "grad_norm": 15.05661392211914,
      "learning_rate": 6.474684735554301e-06,
      "loss": 0.1998,
      "step": 35930
    },
    {
      "epoch": 6.76,
      "grad_norm": 8.65178108215332,
      "learning_rate": 6.470920383963863e-06,
      "loss": 0.1237,
      "step": 35940
    },
    {
      "epoch": 6.77,
      "grad_norm": 8.221214294433594,
      "learning_rate": 6.467156032373424e-06,
      "loss": 0.0281,
      "step": 35950
    },
    {
      "epoch": 6.77,
      "grad_norm": 33.08686447143555,
      "learning_rate": 6.4633916807829855e-06,
      "loss": 0.1069,
      "step": 35960
    },
    {
      "epoch": 6.77,
      "grad_norm": 28.531604766845703,
      "learning_rate": 6.459627329192547e-06,
      "loss": 0.4105,
      "step": 35970
    },
    {
      "epoch": 6.77,
      "grad_norm": 1.7071834802627563,
      "learning_rate": 6.4558629776021084e-06,
      "loss": 0.1428,
      "step": 35980
    },
    {
      "epoch": 6.77,
      "grad_norm": 1.8222893476486206,
      "learning_rate": 6.45209862601167e-06,
      "loss": 0.1487,
      "step": 35990
    },
    {
      "epoch": 6.78,
      "grad_norm": 2.041825771331787,
      "learning_rate": 6.448334274421232e-06,
      "loss": 0.034,
      "step": 36000
    },
    {
      "epoch": 6.78,
      "grad_norm": 17.23551368713379,
      "learning_rate": 6.444569922830794e-06,
      "loss": 0.2445,
      "step": 36010
    },
    {
      "epoch": 6.78,
      "grad_norm": 1.0482386350631714,
      "learning_rate": 6.440805571240354e-06,
      "loss": 0.0432,
      "step": 36020
    },
    {
      "epoch": 6.78,
      "grad_norm": 41.12699890136719,
      "learning_rate": 6.437041219649916e-06,
      "loss": 0.1796,
      "step": 36030
    },
    {
      "epoch": 6.78,
      "grad_norm": 0.029972318559885025,
      "learning_rate": 6.433276868059477e-06,
      "loss": 0.07,
      "step": 36040
    },
    {
      "epoch": 6.79,
      "grad_norm": 13.36446762084961,
      "learning_rate": 6.429512516469039e-06,
      "loss": 0.2117,
      "step": 36050
    },
    {
      "epoch": 6.79,
      "grad_norm": 38.04163360595703,
      "learning_rate": 6.4257481648786e-06,
      "loss": 0.0598,
      "step": 36060
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.019980570301413536,
      "learning_rate": 6.421983813288162e-06,
      "loss": 0.1573,
      "step": 36070
    },
    {
      "epoch": 6.79,
      "grad_norm": 8.11574649810791,
      "learning_rate": 6.418219461697723e-06,
      "loss": 0.1591,
      "step": 36080
    },
    {
      "epoch": 6.79,
      "grad_norm": 0.26194772124290466,
      "learning_rate": 6.414455110107285e-06,
      "loss": 0.1478,
      "step": 36090
    },
    {
      "epoch": 6.79,
      "grad_norm": 1.7264548540115356,
      "learning_rate": 6.410690758516847e-06,
      "loss": 0.2795,
      "step": 36100
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.0394383929669857,
      "learning_rate": 6.406926406926407e-06,
      "loss": 0.1848,
      "step": 36110
    },
    {
      "epoch": 6.8,
      "grad_norm": 12.573692321777344,
      "learning_rate": 6.403162055335969e-06,
      "loss": 0.178,
      "step": 36120
    },
    {
      "epoch": 6.8,
      "grad_norm": 0.005690202582627535,
      "learning_rate": 6.39939770374553e-06,
      "loss": 0.3066,
      "step": 36130
    },
    {
      "epoch": 6.8,
      "grad_norm": 8.21556282043457,
      "learning_rate": 6.395633352155092e-06,
      "loss": 0.3482,
      "step": 36140
    },
    {
      "epoch": 6.8,
      "grad_norm": 6.330270290374756,
      "learning_rate": 6.391869000564653e-06,
      "loss": 0.1825,
      "step": 36150
    },
    {
      "epoch": 6.81,
      "grad_norm": 18.168760299682617,
      "learning_rate": 6.388104648974215e-06,
      "loss": 0.2741,
      "step": 36160
    },
    {
      "epoch": 6.81,
      "grad_norm": 16.058412551879883,
      "learning_rate": 6.384340297383776e-06,
      "loss": 0.2353,
      "step": 36170
    },
    {
      "epoch": 6.81,
      "grad_norm": 1.3712855577468872,
      "learning_rate": 6.380575945793338e-06,
      "loss": 0.1798,
      "step": 36180
    },
    {
      "epoch": 6.81,
      "grad_norm": 1.4504889249801636,
      "learning_rate": 6.376811594202898e-06,
      "loss": 0.2119,
      "step": 36190
    },
    {
      "epoch": 6.81,
      "grad_norm": 0.6500701904296875,
      "learning_rate": 6.37304724261246e-06,
      "loss": 0.0641,
      "step": 36200
    },
    {
      "epoch": 6.82,
      "grad_norm": 1.3924716711044312,
      "learning_rate": 6.369282891022022e-06,
      "loss": 0.104,
      "step": 36210
    },
    {
      "epoch": 6.82,
      "grad_norm": 14.150341033935547,
      "learning_rate": 6.3655185394315835e-06,
      "loss": 0.0904,
      "step": 36220
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.3912964165210724,
      "learning_rate": 6.361754187841145e-06,
      "loss": 0.2247,
      "step": 36230
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.2767292261123657,
      "learning_rate": 6.3579898362507064e-06,
      "loss": 0.1233,
      "step": 36240
    },
    {
      "epoch": 6.82,
      "grad_norm": 0.07434406131505966,
      "learning_rate": 6.354225484660268e-06,
      "loss": 0.0578,
      "step": 36250
    },
    {
      "epoch": 6.82,
      "grad_norm": 23.58869171142578,
      "learning_rate": 6.350461133069829e-06,
      "loss": 0.1622,
      "step": 36260
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.019170990213751793,
      "learning_rate": 6.346696781479391e-06,
      "loss": 0.1481,
      "step": 36270
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.005171836819499731,
      "learning_rate": 6.342932429888951e-06,
      "loss": 0.2074,
      "step": 36280
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.0052894651889801025,
      "learning_rate": 6.339168078298513e-06,
      "loss": 0.0833,
      "step": 36290
    },
    {
      "epoch": 6.83,
      "grad_norm": 3.6812682151794434,
      "learning_rate": 6.335403726708075e-06,
      "loss": 0.5328,
      "step": 36300
    },
    {
      "epoch": 6.83,
      "grad_norm": 0.016765862703323364,
      "learning_rate": 6.331639375117637e-06,
      "loss": 0.1625,
      "step": 36310
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.00508911581709981,
      "learning_rate": 6.327875023527198e-06,
      "loss": 0.1575,
      "step": 36320
    },
    {
      "epoch": 6.84,
      "grad_norm": 6.976175308227539,
      "learning_rate": 6.3241106719367596e-06,
      "loss": 0.2262,
      "step": 36330
    },
    {
      "epoch": 6.84,
      "grad_norm": 13.620063781738281,
      "learning_rate": 6.320346320346321e-06,
      "loss": 0.2172,
      "step": 36340
    },
    {
      "epoch": 6.84,
      "grad_norm": 19.87889862060547,
      "learning_rate": 6.3165819687558825e-06,
      "loss": 0.2743,
      "step": 36350
    },
    {
      "epoch": 6.84,
      "grad_norm": 0.01080454420298338,
      "learning_rate": 6.312817617165444e-06,
      "loss": 0.0435,
      "step": 36360
    },
    {
      "epoch": 6.85,
      "grad_norm": 22.95772933959961,
      "learning_rate": 6.3090532655750046e-06,
      "loss": 0.2156,
      "step": 36370
    },
    {
      "epoch": 6.85,
      "grad_norm": 0.11217557638883591,
      "learning_rate": 6.305288913984566e-06,
      "loss": 0.0839,
      "step": 36380
    },
    {
      "epoch": 6.85,
      "grad_norm": 5.196349620819092,
      "learning_rate": 6.3015245623941275e-06,
      "loss": 0.2068,
      "step": 36390
    },
    {
      "epoch": 6.85,
      "grad_norm": 38.6831169128418,
      "learning_rate": 6.29776021080369e-06,
      "loss": 0.0542,
      "step": 36400
    },
    {
      "epoch": 6.85,
      "grad_norm": 38.81691360473633,
      "learning_rate": 6.293995859213251e-06,
      "loss": 0.3108,
      "step": 36410
    },
    {
      "epoch": 6.85,
      "grad_norm": 1.919084906578064,
      "learning_rate": 6.290231507622813e-06,
      "loss": 0.1508,
      "step": 36420
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.7956769466400146,
      "learning_rate": 6.286467156032374e-06,
      "loss": 0.1486,
      "step": 36430
    },
    {
      "epoch": 6.86,
      "grad_norm": 9.75143814086914,
      "learning_rate": 6.282702804441936e-06,
      "loss": 0.3031,
      "step": 36440
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.005922992713749409,
      "learning_rate": 6.278938452851497e-06,
      "loss": 0.188,
      "step": 36450
    },
    {
      "epoch": 6.86,
      "grad_norm": 0.017864853143692017,
      "learning_rate": 6.275174101261058e-06,
      "loss": 0.2062,
      "step": 36460
    },
    {
      "epoch": 6.86,
      "grad_norm": 23.99368667602539,
      "learning_rate": 6.271409749670619e-06,
      "loss": 0.2293,
      "step": 36470
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.023939380422234535,
      "learning_rate": 6.267645398080181e-06,
      "loss": 0.0579,
      "step": 36480
    },
    {
      "epoch": 6.87,
      "grad_norm": 0.004430987872183323,
      "learning_rate": 6.263881046489743e-06,
      "loss": 0.1909,
      "step": 36490
    },
    {
      "epoch": 6.87,
      "grad_norm": 19.863258361816406,
      "learning_rate": 6.260116694899304e-06,
      "loss": 0.4464,
      "step": 36500
    },
    {
      "epoch": 6.87,
      "grad_norm": 5.021882057189941,
      "learning_rate": 6.256352343308866e-06,
      "loss": 0.2766,
      "step": 36510
    },
    {
      "epoch": 6.87,
      "grad_norm": 1.7789100408554077,
      "learning_rate": 6.252587991718427e-06,
      "loss": 0.3443,
      "step": 36520
    },
    {
      "epoch": 6.88,
      "grad_norm": 15.799257278442383,
      "learning_rate": 6.248823640127989e-06,
      "loss": 0.5121,
      "step": 36530
    },
    {
      "epoch": 6.88,
      "grad_norm": 8.36252498626709,
      "learning_rate": 6.245059288537549e-06,
      "loss": 0.239,
      "step": 36540
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.8076105713844299,
      "learning_rate": 6.241294936947111e-06,
      "loss": 0.2852,
      "step": 36550
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.0374312698841095,
      "learning_rate": 6.237530585356672e-06,
      "loss": 0.2056,
      "step": 36560
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.00854515004903078,
      "learning_rate": 6.233766233766234e-06,
      "loss": 0.0939,
      "step": 36570
    },
    {
      "epoch": 6.88,
      "grad_norm": 0.08278057724237442,
      "learning_rate": 6.230001882175796e-06,
      "loss": 0.2267,
      "step": 36580
    },
    {
      "epoch": 6.89,
      "grad_norm": 11.034607887268066,
      "learning_rate": 6.2262375305853575e-06,
      "loss": 0.3064,
      "step": 36590
    },
    {
      "epoch": 6.89,
      "grad_norm": 81.32418823242188,
      "learning_rate": 6.222473178994919e-06,
      "loss": 0.2267,
      "step": 36600
    },
    {
      "epoch": 6.89,
      "grad_norm": 0.16295531392097473,
      "learning_rate": 6.2187088274044805e-06,
      "loss": 0.1216,
      "step": 36610
    },
    {
      "epoch": 6.89,
      "grad_norm": 1.7374696731567383,
      "learning_rate": 6.214944475814042e-06,
      "loss": 0.3677,
      "step": 36620
    },
    {
      "epoch": 6.89,
      "grad_norm": 3.87906551361084,
      "learning_rate": 6.2111801242236025e-06,
      "loss": 0.1676,
      "step": 36630
    },
    {
      "epoch": 6.9,
      "grad_norm": 0.015681099146604538,
      "learning_rate": 6.207415772633164e-06,
      "loss": 0.1863,
      "step": 36640
    },
    {
      "epoch": 6.9,
      "grad_norm": 21.523597717285156,
      "learning_rate": 6.2036514210427255e-06,
      "loss": 0.1306,
      "step": 36650
    },
    {
      "epoch": 6.9,
      "grad_norm": 16.166555404663086,
      "learning_rate": 6.199887069452287e-06,
      "loss": 0.2241,
      "step": 36660
    },
    {
      "epoch": 6.9,
      "grad_norm": 5.832311153411865,
      "learning_rate": 6.196122717861849e-06,
      "loss": 0.2362,
      "step": 36670
    },
    {
      "epoch": 6.9,
      "grad_norm": 0.6826976537704468,
      "learning_rate": 6.192358366271411e-06,
      "loss": 0.1775,
      "step": 36680
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.005314170382916927,
      "learning_rate": 6.188594014680972e-06,
      "loss": 0.1871,
      "step": 36690
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.1084243580698967,
      "learning_rate": 6.184829663090534e-06,
      "loss": 0.0476,
      "step": 36700
    },
    {
      "epoch": 6.91,
      "grad_norm": 9.639162063598633,
      "learning_rate": 6.181065311500095e-06,
      "loss": 0.1045,
      "step": 36710
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.057649411261081696,
      "learning_rate": 6.177300959909656e-06,
      "loss": 0.0178,
      "step": 36720
    },
    {
      "epoch": 6.91,
      "grad_norm": 0.1679777204990387,
      "learning_rate": 6.173536608319217e-06,
      "loss": 0.1744,
      "step": 36730
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.012088864110410213,
      "learning_rate": 6.169772256728779e-06,
      "loss": 0.2307,
      "step": 36740
    },
    {
      "epoch": 6.92,
      "grad_norm": 3.3160908222198486,
      "learning_rate": 6.16600790513834e-06,
      "loss": 0.1667,
      "step": 36750
    },
    {
      "epoch": 6.92,
      "grad_norm": 42.444156646728516,
      "learning_rate": 6.1622435535479015e-06,
      "loss": 0.2177,
      "step": 36760
    },
    {
      "epoch": 6.92,
      "grad_norm": 17.346935272216797,
      "learning_rate": 6.158479201957464e-06,
      "loss": 0.1661,
      "step": 36770
    },
    {
      "epoch": 6.92,
      "grad_norm": 0.3202081024646759,
      "learning_rate": 6.154714850367025e-06,
      "loss": 0.2907,
      "step": 36780
    },
    {
      "epoch": 6.92,
      "grad_norm": 10.56147289276123,
      "learning_rate": 6.150950498776587e-06,
      "loss": 0.2622,
      "step": 36790
    },
    {
      "epoch": 6.93,
      "grad_norm": 17.557096481323242,
      "learning_rate": 6.147186147186147e-06,
      "loss": 0.1995,
      "step": 36800
    },
    {
      "epoch": 6.93,
      "grad_norm": 18.097747802734375,
      "learning_rate": 6.143421795595709e-06,
      "loss": 0.1636,
      "step": 36810
    },
    {
      "epoch": 6.93,
      "grad_norm": 4.408967018127441,
      "learning_rate": 6.13965744400527e-06,
      "loss": 0.212,
      "step": 36820
    },
    {
      "epoch": 6.93,
      "grad_norm": 2.0536444187164307,
      "learning_rate": 6.135893092414832e-06,
      "loss": 0.0888,
      "step": 36830
    },
    {
      "epoch": 6.93,
      "grad_norm": 10.114295959472656,
      "learning_rate": 6.132128740824393e-06,
      "loss": 0.3205,
      "step": 36840
    },
    {
      "epoch": 6.94,
      "grad_norm": 24.806224822998047,
      "learning_rate": 6.128364389233955e-06,
      "loss": 0.1907,
      "step": 36850
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.10939033329486847,
      "learning_rate": 6.124600037643517e-06,
      "loss": 0.1276,
      "step": 36860
    },
    {
      "epoch": 6.94,
      "grad_norm": 27.757322311401367,
      "learning_rate": 6.1208356860530784e-06,
      "loss": 0.2743,
      "step": 36870
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.20790190994739532,
      "learning_rate": 6.11707133446264e-06,
      "loss": 0.0745,
      "step": 36880
    },
    {
      "epoch": 6.94,
      "grad_norm": 0.041102781891822815,
      "learning_rate": 6.1133069828722005e-06,
      "loss": 0.178,
      "step": 36890
    },
    {
      "epoch": 6.95,
      "grad_norm": 5.39219856262207,
      "learning_rate": 6.109542631281762e-06,
      "loss": 0.3331,
      "step": 36900
    },
    {
      "epoch": 6.95,
      "grad_norm": 6.225903034210205,
      "learning_rate": 6.1057782796913234e-06,
      "loss": 0.2216,
      "step": 36910
    },
    {
      "epoch": 6.95,
      "grad_norm": 4.442195415496826,
      "learning_rate": 6.102013928100885e-06,
      "loss": 0.232,
      "step": 36920
    },
    {
      "epoch": 6.95,
      "grad_norm": 41.77784729003906,
      "learning_rate": 6.098249576510446e-06,
      "loss": 0.2279,
      "step": 36930
    },
    {
      "epoch": 6.95,
      "grad_norm": 1.1963173151016235,
      "learning_rate": 6.094485224920008e-06,
      "loss": 0.1062,
      "step": 36940
    },
    {
      "epoch": 6.95,
      "grad_norm": 0.8327680230140686,
      "learning_rate": 6.09072087332957e-06,
      "loss": 0.1902,
      "step": 36950
    },
    {
      "epoch": 6.96,
      "grad_norm": 0.005166544578969479,
      "learning_rate": 6.086956521739132e-06,
      "loss": 0.1202,
      "step": 36960
    },
    {
      "epoch": 6.96,
      "grad_norm": 6.69047737121582,
      "learning_rate": 6.083192170148693e-06,
      "loss": 0.0722,
      "step": 36970
    },
    {
      "epoch": 6.96,
      "grad_norm": 4.629175186157227,
      "learning_rate": 6.079427818558254e-06,
      "loss": 0.1815,
      "step": 36980
    },
    {
      "epoch": 6.96,
      "grad_norm": 15.691399574279785,
      "learning_rate": 6.075663466967815e-06,
      "loss": 0.1927,
      "step": 36990
    },
    {
      "epoch": 6.96,
      "grad_norm": 3.9890987873077393,
      "learning_rate": 6.071899115377377e-06,
      "loss": 0.2116,
      "step": 37000
    },
    {
      "epoch": 6.97,
      "grad_norm": 13.684748649597168,
      "learning_rate": 6.068134763786938e-06,
      "loss": 0.1541,
      "step": 37010
    },
    {
      "epoch": 6.97,
      "grad_norm": 4.819939136505127,
      "learning_rate": 6.0643704121964995e-06,
      "loss": 0.301,
      "step": 37020
    },
    {
      "epoch": 6.97,
      "grad_norm": 3.2137558460235596,
      "learning_rate": 6.060606060606061e-06,
      "loss": 0.097,
      "step": 37030
    },
    {
      "epoch": 6.97,
      "grad_norm": 0.0069637601263821125,
      "learning_rate": 6.0568417090156224e-06,
      "loss": 0.296,
      "step": 37040
    },
    {
      "epoch": 6.97,
      "grad_norm": 0.006687845103442669,
      "learning_rate": 6.053077357425185e-06,
      "loss": 0.0992,
      "step": 37050
    },
    {
      "epoch": 6.98,
      "grad_norm": 0.07013238221406937,
      "learning_rate": 6.049313005834746e-06,
      "loss": 0.1701,
      "step": 37060
    },
    {
      "epoch": 6.98,
      "grad_norm": 8.007899284362793,
      "learning_rate": 6.045548654244307e-06,
      "loss": 0.1847,
      "step": 37070
    },
    {
      "epoch": 6.98,
      "grad_norm": 2.286176919937134,
      "learning_rate": 6.041784302653868e-06,
      "loss": 0.1452,
      "step": 37080
    },
    {
      "epoch": 6.98,
      "grad_norm": 7.135737895965576,
      "learning_rate": 6.03801995106343e-06,
      "loss": 0.2618,
      "step": 37090
    },
    {
      "epoch": 6.98,
      "grad_norm": 17.96920394897461,
      "learning_rate": 6.034255599472991e-06,
      "loss": 0.3619,
      "step": 37100
    },
    {
      "epoch": 6.98,
      "grad_norm": 9.395142555236816,
      "learning_rate": 6.030491247882553e-06,
      "loss": 0.3724,
      "step": 37110
    },
    {
      "epoch": 6.99,
      "grad_norm": 21.037466049194336,
      "learning_rate": 6.026726896292114e-06,
      "loss": 0.1191,
      "step": 37120
    },
    {
      "epoch": 6.99,
      "grad_norm": 11.869218826293945,
      "learning_rate": 6.022962544701676e-06,
      "loss": 0.257,
      "step": 37130
    },
    {
      "epoch": 6.99,
      "grad_norm": 5.827176094055176,
      "learning_rate": 6.019198193111238e-06,
      "loss": 0.3006,
      "step": 37140
    },
    {
      "epoch": 6.99,
      "grad_norm": 1.9480148553848267,
      "learning_rate": 6.015433841520798e-06,
      "loss": 0.1901,
      "step": 37150
    },
    {
      "epoch": 6.99,
      "grad_norm": 0.011721281334757805,
      "learning_rate": 6.01166948993036e-06,
      "loss": 0.2029,
      "step": 37160
    },
    {
      "epoch": 7.0,
      "grad_norm": 22.344335556030273,
      "learning_rate": 6.007905138339921e-06,
      "loss": 0.2494,
      "step": 37170
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.1331589221954346,
      "learning_rate": 6.004140786749483e-06,
      "loss": 0.0221,
      "step": 37180
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.8588600158691406,
      "learning_rate": 6.000376435159044e-06,
      "loss": 0.133,
      "step": 37190
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9910666666666667,
      "eval_loss": 0.03684931620955467,
      "eval_runtime": 51.2436,
      "eval_samples_per_second": 146.36,
      "eval_steps_per_second": 18.305,
      "step": 37191
    },
    {
      "epoch": 7.0,
      "grad_norm": 12.256184577941895,
      "learning_rate": 5.996612083568606e-06,
      "loss": 0.1332,
      "step": 37200
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.2593938112258911,
      "learning_rate": 5.992847731978167e-06,
      "loss": 0.1521,
      "step": 37210
    },
    {
      "epoch": 7.01,
      "grad_norm": 41.557220458984375,
      "learning_rate": 5.989083380387729e-06,
      "loss": 0.2337,
      "step": 37220
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.8613956570625305,
      "learning_rate": 5.985319028797291e-06,
      "loss": 0.0662,
      "step": 37230
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.005155636928975582,
      "learning_rate": 5.981554677206851e-06,
      "loss": 0.1394,
      "step": 37240
    },
    {
      "epoch": 7.01,
      "grad_norm": 6.470524311065674,
      "learning_rate": 5.977790325616413e-06,
      "loss": 0.268,
      "step": 37250
    },
    {
      "epoch": 7.01,
      "grad_norm": 6.44150447845459,
      "learning_rate": 5.9740259740259746e-06,
      "loss": 0.1806,
      "step": 37260
    },
    {
      "epoch": 7.01,
      "grad_norm": 0.006630315911024809,
      "learning_rate": 5.970261622435536e-06,
      "loss": 0.1476,
      "step": 37270
    },
    {
      "epoch": 7.02,
      "grad_norm": 14.538272857666016,
      "learning_rate": 5.9664972708450975e-06,
      "loss": 0.1498,
      "step": 37280
    },
    {
      "epoch": 7.02,
      "grad_norm": 5.568931579589844,
      "learning_rate": 5.962732919254659e-06,
      "loss": 0.171,
      "step": 37290
    },
    {
      "epoch": 7.02,
      "grad_norm": 0.06793810427188873,
      "learning_rate": 5.95896856766422e-06,
      "loss": 0.1745,
      "step": 37300
    },
    {
      "epoch": 7.02,
      "grad_norm": 19.724597930908203,
      "learning_rate": 5.955204216073782e-06,
      "loss": 0.1408,
      "step": 37310
    },
    {
      "epoch": 7.02,
      "grad_norm": 0.23028509318828583,
      "learning_rate": 5.951439864483344e-06,
      "loss": 0.0805,
      "step": 37320
    },
    {
      "epoch": 7.03,
      "grad_norm": 17.94396209716797,
      "learning_rate": 5.947675512892904e-06,
      "loss": 0.2682,
      "step": 37330
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.33757466077804565,
      "learning_rate": 5.943911161302465e-06,
      "loss": 0.0356,
      "step": 37340
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.007450920529663563,
      "learning_rate": 5.940146809712028e-06,
      "loss": 0.1651,
      "step": 37350
    },
    {
      "epoch": 7.03,
      "grad_norm": 0.006568220444023609,
      "learning_rate": 5.936382458121589e-06,
      "loss": 0.16,
      "step": 37360
    },
    {
      "epoch": 7.03,
      "grad_norm": 14.52227783203125,
      "learning_rate": 5.932618106531151e-06,
      "loss": 0.1751,
      "step": 37370
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.011770389974117279,
      "learning_rate": 5.928853754940712e-06,
      "loss": 0.2462,
      "step": 37380
    },
    {
      "epoch": 7.04,
      "grad_norm": 21.111907958984375,
      "learning_rate": 5.9250894033502736e-06,
      "loss": 0.1445,
      "step": 37390
    },
    {
      "epoch": 7.04,
      "grad_norm": 3.9244065284729004,
      "learning_rate": 5.921325051759835e-06,
      "loss": 0.2087,
      "step": 37400
    },
    {
      "epoch": 7.04,
      "grad_norm": 2.7848761081695557,
      "learning_rate": 5.917560700169396e-06,
      "loss": 0.1174,
      "step": 37410
    },
    {
      "epoch": 7.04,
      "grad_norm": 23.053918838500977,
      "learning_rate": 5.913796348578957e-06,
      "loss": 0.197,
      "step": 37420
    },
    {
      "epoch": 7.04,
      "grad_norm": 0.1026964783668518,
      "learning_rate": 5.9100319969885186e-06,
      "loss": 0.0763,
      "step": 37430
    },
    {
      "epoch": 7.05,
      "grad_norm": 0.631119966506958,
      "learning_rate": 5.906267645398081e-06,
      "loss": 0.2372,
      "step": 37440
    },
    {
      "epoch": 7.05,
      "grad_norm": 6.604733943939209,
      "learning_rate": 5.902503293807642e-06,
      "loss": 0.2685,
      "step": 37450
    },
    {
      "epoch": 7.05,
      "grad_norm": 1.1114252805709839,
      "learning_rate": 5.898738942217204e-06,
      "loss": 0.238,
      "step": 37460
    },
    {
      "epoch": 7.05,
      "grad_norm": 41.08172607421875,
      "learning_rate": 5.894974590626765e-06,
      "loss": 0.5514,
      "step": 37470
    },
    {
      "epoch": 7.05,
      "grad_norm": 0.004217694513499737,
      "learning_rate": 5.891210239036327e-06,
      "loss": 0.243,
      "step": 37480
    },
    {
      "epoch": 7.06,
      "grad_norm": 39.454524993896484,
      "learning_rate": 5.887445887445888e-06,
      "loss": 0.0654,
      "step": 37490
    },
    {
      "epoch": 7.06,
      "grad_norm": 5.350090980529785,
      "learning_rate": 5.883681535855449e-06,
      "loss": 0.3062,
      "step": 37500
    },
    {
      "epoch": 7.06,
      "grad_norm": 35.24092102050781,
      "learning_rate": 5.87991718426501e-06,
      "loss": 0.1597,
      "step": 37510
    },
    {
      "epoch": 7.06,
      "grad_norm": 0.008568137884140015,
      "learning_rate": 5.876152832674572e-06,
      "loss": 0.1961,
      "step": 37520
    },
    {
      "epoch": 7.06,
      "grad_norm": 3.959826946258545,
      "learning_rate": 5.872388481084134e-06,
      "loss": 0.1091,
      "step": 37530
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.20467731356620789,
      "learning_rate": 5.8686241294936955e-06,
      "loss": 0.043,
      "step": 37540
    },
    {
      "epoch": 7.07,
      "grad_norm": 19.6613712310791,
      "learning_rate": 5.864859777903257e-06,
      "loss": 0.1451,
      "step": 37550
    },
    {
      "epoch": 7.07,
      "grad_norm": 54.88874053955078,
      "learning_rate": 5.861095426312818e-06,
      "loss": 0.2519,
      "step": 37560
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.12221944332122803,
      "learning_rate": 5.85733107472238e-06,
      "loss": 0.2664,
      "step": 37570
    },
    {
      "epoch": 7.07,
      "grad_norm": 0.023488381877541542,
      "learning_rate": 5.853566723131941e-06,
      "loss": 0.2164,
      "step": 37580
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.008625051937997341,
      "learning_rate": 5.849802371541502e-06,
      "loss": 0.2866,
      "step": 37590
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.08634841442108154,
      "learning_rate": 5.846038019951063e-06,
      "loss": 0.2117,
      "step": 37600
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.016825608909130096,
      "learning_rate": 5.842273668360625e-06,
      "loss": 0.1398,
      "step": 37610
    },
    {
      "epoch": 7.08,
      "grad_norm": 26.56159782409668,
      "learning_rate": 5.838509316770186e-06,
      "loss": 0.3046,
      "step": 37620
    },
    {
      "epoch": 7.08,
      "grad_norm": 5.012177467346191,
      "learning_rate": 5.834744965179749e-06,
      "loss": 0.2082,
      "step": 37630
    },
    {
      "epoch": 7.08,
      "grad_norm": 0.01664702594280243,
      "learning_rate": 5.83098061358931e-06,
      "loss": 0.0195,
      "step": 37640
    },
    {
      "epoch": 7.09,
      "grad_norm": 14.833102226257324,
      "learning_rate": 5.8272162619988715e-06,
      "loss": 0.2715,
      "step": 37650
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.06048378720879555,
      "learning_rate": 5.823451910408433e-06,
      "loss": 0.1972,
      "step": 37660
    },
    {
      "epoch": 7.09,
      "grad_norm": 22.43869400024414,
      "learning_rate": 5.8196875588179945e-06,
      "loss": 0.1718,
      "step": 37670
    },
    {
      "epoch": 7.09,
      "grad_norm": 0.013256150297820568,
      "learning_rate": 5.815923207227555e-06,
      "loss": 0.0584,
      "step": 37680
    },
    {
      "epoch": 7.09,
      "grad_norm": 6.026370048522949,
      "learning_rate": 5.8121588556371165e-06,
      "loss": 0.2856,
      "step": 37690
    },
    {
      "epoch": 7.1,
      "grad_norm": 8.284187316894531,
      "learning_rate": 5.808394504046678e-06,
      "loss": 0.1835,
      "step": 37700
    },
    {
      "epoch": 7.1,
      "grad_norm": 3.917576313018799,
      "learning_rate": 5.8046301524562395e-06,
      "loss": 0.3282,
      "step": 37710
    },
    {
      "epoch": 7.1,
      "grad_norm": 30.30458641052246,
      "learning_rate": 5.800865800865802e-06,
      "loss": 0.3005,
      "step": 37720
    },
    {
      "epoch": 7.1,
      "grad_norm": 0.014058428816497326,
      "learning_rate": 5.797101449275363e-06,
      "loss": 0.0884,
      "step": 37730
    },
    {
      "epoch": 7.1,
      "grad_norm": 9.717737197875977,
      "learning_rate": 5.793337097684925e-06,
      "loss": 0.1097,
      "step": 37740
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.02633320912718773,
      "learning_rate": 5.789572746094486e-06,
      "loss": 0.2463,
      "step": 37750
    },
    {
      "epoch": 7.11,
      "grad_norm": 3.707240581512451,
      "learning_rate": 5.785808394504047e-06,
      "loss": 0.2937,
      "step": 37760
    },
    {
      "epoch": 7.11,
      "grad_norm": 11.806624412536621,
      "learning_rate": 5.782044042913608e-06,
      "loss": 0.2457,
      "step": 37770
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.0968085452914238,
      "learning_rate": 5.77827969132317e-06,
      "loss": 0.2157,
      "step": 37780
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.013852328062057495,
      "learning_rate": 5.774515339732731e-06,
      "loss": 0.1636,
      "step": 37790
    },
    {
      "epoch": 7.11,
      "grad_norm": 0.09347653388977051,
      "learning_rate": 5.770750988142293e-06,
      "loss": 0.3389,
      "step": 37800
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.006945687346160412,
      "learning_rate": 5.766986636551855e-06,
      "loss": 0.19,
      "step": 37810
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.007952738553285599,
      "learning_rate": 5.763222284961416e-06,
      "loss": 0.164,
      "step": 37820
    },
    {
      "epoch": 7.12,
      "grad_norm": 2.526045322418213,
      "learning_rate": 5.759457933370978e-06,
      "loss": 0.3376,
      "step": 37830
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.017537254840135574,
      "learning_rate": 5.755693581780539e-06,
      "loss": 0.1528,
      "step": 37840
    },
    {
      "epoch": 7.12,
      "grad_norm": 0.09951084852218628,
      "learning_rate": 5.7519292301901e-06,
      "loss": 0.2103,
      "step": 37850
    },
    {
      "epoch": 7.13,
      "grad_norm": 5.293554306030273,
      "learning_rate": 5.748164878599661e-06,
      "loss": 0.1028,
      "step": 37860
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.5685975551605225,
      "learning_rate": 5.744400527009223e-06,
      "loss": 0.0298,
      "step": 37870
    },
    {
      "epoch": 7.13,
      "grad_norm": 2.7828612327575684,
      "learning_rate": 5.740636175418784e-06,
      "loss": 0.1093,
      "step": 37880
    },
    {
      "epoch": 7.13,
      "grad_norm": 0.0037335713859647512,
      "learning_rate": 5.736871823828346e-06,
      "loss": 0.1121,
      "step": 37890
    },
    {
      "epoch": 7.13,
      "grad_norm": 16.625938415527344,
      "learning_rate": 5.733107472237908e-06,
      "loss": 0.2014,
      "step": 37900
    },
    {
      "epoch": 7.14,
      "grad_norm": 6.445840835571289,
      "learning_rate": 5.7293431206474695e-06,
      "loss": 0.203,
      "step": 37910
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.015050169080495834,
      "learning_rate": 5.725578769057031e-06,
      "loss": 0.1732,
      "step": 37920
    },
    {
      "epoch": 7.14,
      "grad_norm": 34.25404739379883,
      "learning_rate": 5.7218144174665924e-06,
      "loss": 0.264,
      "step": 37930
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.032222989946603775,
      "learning_rate": 5.718050065876153e-06,
      "loss": 0.1695,
      "step": 37940
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.0073930504731833935,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.1387,
      "step": 37950
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.007963989861309528,
      "learning_rate": 5.710521362695276e-06,
      "loss": 0.0913,
      "step": 37960
    },
    {
      "epoch": 7.15,
      "grad_norm": 12.056676864624023,
      "learning_rate": 5.7067570111048374e-06,
      "loss": 0.368,
      "step": 37970
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.9320905208587646,
      "learning_rate": 5.702992659514399e-06,
      "loss": 0.2999,
      "step": 37980
    },
    {
      "epoch": 7.15,
      "grad_norm": 0.01549438014626503,
      "learning_rate": 5.69922830792396e-06,
      "loss": 0.1467,
      "step": 37990
    },
    {
      "epoch": 7.15,
      "grad_norm": 6.155904769897461,
      "learning_rate": 5.695463956333523e-06,
      "loss": 0.2997,
      "step": 38000
    },
    {
      "epoch": 7.15,
      "grad_norm": 8.344841957092285,
      "learning_rate": 5.691699604743084e-06,
      "loss": 0.2235,
      "step": 38010
    },
    {
      "epoch": 7.16,
      "grad_norm": 33.447139739990234,
      "learning_rate": 5.687935253152645e-06,
      "loss": 0.1835,
      "step": 38020
    },
    {
      "epoch": 7.16,
      "grad_norm": 4.505063056945801,
      "learning_rate": 5.684170901562206e-06,
      "loss": 0.0958,
      "step": 38030
    },
    {
      "epoch": 7.16,
      "grad_norm": 34.33796691894531,
      "learning_rate": 5.680406549971768e-06,
      "loss": 0.4119,
      "step": 38040
    },
    {
      "epoch": 7.16,
      "grad_norm": 19.099681854248047,
      "learning_rate": 5.676642198381329e-06,
      "loss": 0.053,
      "step": 38050
    },
    {
      "epoch": 7.16,
      "grad_norm": 8.617927551269531,
      "learning_rate": 5.672877846790891e-06,
      "loss": 0.173,
      "step": 38060
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.004702289588749409,
      "learning_rate": 5.669113495200452e-06,
      "loss": 0.1499,
      "step": 38070
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.009395179338753223,
      "learning_rate": 5.6653491436100135e-06,
      "loss": 0.1153,
      "step": 38080
    },
    {
      "epoch": 7.17,
      "grad_norm": 19.420408248901367,
      "learning_rate": 5.661584792019576e-06,
      "loss": 0.2316,
      "step": 38090
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.05560816451907158,
      "learning_rate": 5.657820440429137e-06,
      "loss": 0.1575,
      "step": 38100
    },
    {
      "epoch": 7.17,
      "grad_norm": 1.3646773099899292,
      "learning_rate": 5.654056088838698e-06,
      "loss": 0.1049,
      "step": 38110
    },
    {
      "epoch": 7.17,
      "grad_norm": 0.0035281842574477196,
      "learning_rate": 5.650291737248259e-06,
      "loss": 0.1816,
      "step": 38120
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.008720306679606438,
      "learning_rate": 5.646527385657821e-06,
      "loss": 0.0643,
      "step": 38130
    },
    {
      "epoch": 7.18,
      "grad_norm": 5.0900702476501465,
      "learning_rate": 5.642763034067382e-06,
      "loss": 0.1657,
      "step": 38140
    },
    {
      "epoch": 7.18,
      "grad_norm": 24.471471786499023,
      "learning_rate": 5.638998682476944e-06,
      "loss": 0.4437,
      "step": 38150
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.01770561933517456,
      "learning_rate": 5.635234330886505e-06,
      "loss": 0.1038,
      "step": 38160
    },
    {
      "epoch": 7.18,
      "grad_norm": 0.13364748656749725,
      "learning_rate": 5.631469979296067e-06,
      "loss": 0.2925,
      "step": 38170
    },
    {
      "epoch": 7.19,
      "grad_norm": 5.666759490966797,
      "learning_rate": 5.627705627705629e-06,
      "loss": 0.183,
      "step": 38180
    },
    {
      "epoch": 7.19,
      "grad_norm": 3.8661653995513916,
      "learning_rate": 5.62394127611519e-06,
      "loss": 0.1065,
      "step": 38190
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.026401078328490257,
      "learning_rate": 5.620176924524751e-06,
      "loss": 0.1493,
      "step": 38200
    },
    {
      "epoch": 7.19,
      "grad_norm": 0.4152458906173706,
      "learning_rate": 5.6164125729343125e-06,
      "loss": 0.0069,
      "step": 38210
    },
    {
      "epoch": 7.19,
      "grad_norm": 5.415332317352295,
      "learning_rate": 5.612648221343874e-06,
      "loss": 0.217,
      "step": 38220
    },
    {
      "epoch": 7.2,
      "grad_norm": 8.30183219909668,
      "learning_rate": 5.608883869753435e-06,
      "loss": 0.2578,
      "step": 38230
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.0044572316110134125,
      "learning_rate": 5.605119518162997e-06,
      "loss": 0.1654,
      "step": 38240
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.724814236164093,
      "learning_rate": 5.601355166572558e-06,
      "loss": 0.1754,
      "step": 38250
    },
    {
      "epoch": 7.2,
      "grad_norm": 14.581521987915039,
      "learning_rate": 5.59759081498212e-06,
      "loss": 0.1867,
      "step": 38260
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.009991789236664772,
      "learning_rate": 5.593826463391681e-06,
      "loss": 0.3555,
      "step": 38270
    },
    {
      "epoch": 7.2,
      "grad_norm": 0.0072522773407399654,
      "learning_rate": 5.590062111801242e-06,
      "loss": 0.2164,
      "step": 38280
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.01714443601667881,
      "learning_rate": 5.586297760210803e-06,
      "loss": 0.2044,
      "step": 38290
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.019844001159071922,
      "learning_rate": 5.582533408620366e-06,
      "loss": 0.1575,
      "step": 38300
    },
    {
      "epoch": 7.21,
      "grad_norm": 3.470121145248413,
      "learning_rate": 5.578769057029927e-06,
      "loss": 0.1971,
      "step": 38310
    },
    {
      "epoch": 7.21,
      "grad_norm": 0.025619855150580406,
      "learning_rate": 5.5750047054394886e-06,
      "loss": 0.16,
      "step": 38320
    },
    {
      "epoch": 7.21,
      "grad_norm": 37.2904052734375,
      "learning_rate": 5.57124035384905e-06,
      "loss": 0.2031,
      "step": 38330
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.05162591487169266,
      "learning_rate": 5.5674760022586115e-06,
      "loss": 0.1201,
      "step": 38340
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.15618836879730225,
      "learning_rate": 5.563711650668173e-06,
      "loss": 0.2047,
      "step": 38350
    },
    {
      "epoch": 7.22,
      "grad_norm": 1.8339310884475708,
      "learning_rate": 5.559947299077734e-06,
      "loss": 0.066,
      "step": 38360
    },
    {
      "epoch": 7.22,
      "grad_norm": 29.168737411499023,
      "learning_rate": 5.556182947487295e-06,
      "loss": 0.1633,
      "step": 38370
    },
    {
      "epoch": 7.22,
      "grad_norm": 0.42933720350265503,
      "learning_rate": 5.5524185958968565e-06,
      "loss": 0.2286,
      "step": 38380
    },
    {
      "epoch": 7.23,
      "grad_norm": 17.730260848999023,
      "learning_rate": 5.548654244306419e-06,
      "loss": 0.0861,
      "step": 38390
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.013194596394896507,
      "learning_rate": 5.54488989271598e-06,
      "loss": 0.1177,
      "step": 38400
    },
    {
      "epoch": 7.23,
      "grad_norm": 17.099416732788086,
      "learning_rate": 5.541125541125542e-06,
      "loss": 0.1615,
      "step": 38410
    },
    {
      "epoch": 7.23,
      "grad_norm": 16.897329330444336,
      "learning_rate": 5.537361189535103e-06,
      "loss": 0.1462,
      "step": 38420
    },
    {
      "epoch": 7.23,
      "grad_norm": 0.030531231313943863,
      "learning_rate": 5.533596837944665e-06,
      "loss": 0.0584,
      "step": 38430
    },
    {
      "epoch": 7.24,
      "grad_norm": 32.75541305541992,
      "learning_rate": 5.529832486354226e-06,
      "loss": 0.1207,
      "step": 38440
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.03371430188417435,
      "learning_rate": 5.5260681347637875e-06,
      "loss": 0.1448,
      "step": 38450
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.13245323300361633,
      "learning_rate": 5.522303783173348e-06,
      "loss": 0.1089,
      "step": 38460
    },
    {
      "epoch": 7.24,
      "grad_norm": 14.70712661743164,
      "learning_rate": 5.51853943158291e-06,
      "loss": 0.146,
      "step": 38470
    },
    {
      "epoch": 7.24,
      "grad_norm": 2.548194408416748,
      "learning_rate": 5.514775079992472e-06,
      "loss": 0.1477,
      "step": 38480
    },
    {
      "epoch": 7.24,
      "grad_norm": 0.014603745192289352,
      "learning_rate": 5.511010728402033e-06,
      "loss": 0.1054,
      "step": 38490
    },
    {
      "epoch": 7.25,
      "grad_norm": 0.08214918524026871,
      "learning_rate": 5.507246376811595e-06,
      "loss": 0.1845,
      "step": 38500
    },
    {
      "epoch": 7.25,
      "grad_norm": 1.7221230268478394,
      "learning_rate": 5.503482025221156e-06,
      "loss": 0.2164,
      "step": 38510
    },
    {
      "epoch": 7.25,
      "grad_norm": 8.295744895935059,
      "learning_rate": 5.499717673630718e-06,
      "loss": 0.4063,
      "step": 38520
    },
    {
      "epoch": 7.25,
      "grad_norm": 1.0702894926071167,
      "learning_rate": 5.495953322040279e-06,
      "loss": 0.1585,
      "step": 38530
    },
    {
      "epoch": 7.25,
      "grad_norm": 16.897005081176758,
      "learning_rate": 5.492188970449841e-06,
      "loss": 0.1579,
      "step": 38540
    },
    {
      "epoch": 7.26,
      "grad_norm": 22.098968505859375,
      "learning_rate": 5.488424618859401e-06,
      "loss": 0.13,
      "step": 38550
    },
    {
      "epoch": 7.26,
      "grad_norm": 10.991353988647461,
      "learning_rate": 5.484660267268963e-06,
      "loss": 0.3069,
      "step": 38560
    },
    {
      "epoch": 7.26,
      "grad_norm": 0.004447883926331997,
      "learning_rate": 5.480895915678524e-06,
      "loss": 0.0452,
      "step": 38570
    },
    {
      "epoch": 7.26,
      "grad_norm": 8.194801330566406,
      "learning_rate": 5.4771315640880865e-06,
      "loss": 0.2908,
      "step": 38580
    },
    {
      "epoch": 7.26,
      "grad_norm": 11.853592872619629,
      "learning_rate": 5.473367212497648e-06,
      "loss": 0.277,
      "step": 38590
    },
    {
      "epoch": 7.27,
      "grad_norm": 5.993780136108398,
      "learning_rate": 5.4696028609072095e-06,
      "loss": 0.1399,
      "step": 38600
    },
    {
      "epoch": 7.27,
      "grad_norm": 18.60516357421875,
      "learning_rate": 5.465838509316771e-06,
      "loss": 0.23,
      "step": 38610
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.019330350682139397,
      "learning_rate": 5.462074157726332e-06,
      "loss": 0.0828,
      "step": 38620
    },
    {
      "epoch": 7.27,
      "grad_norm": 5.319974899291992,
      "learning_rate": 5.458309806135893e-06,
      "loss": 0.2111,
      "step": 38630
    },
    {
      "epoch": 7.27,
      "grad_norm": 25.8642635345459,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.3021,
      "step": 38640
    },
    {
      "epoch": 7.27,
      "grad_norm": 0.036091163754463196,
      "learning_rate": 5.450781102955016e-06,
      "loss": 0.154,
      "step": 38650
    },
    {
      "epoch": 7.28,
      "grad_norm": 18.453731536865234,
      "learning_rate": 5.447016751364577e-06,
      "loss": 0.0599,
      "step": 38660
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.015668820589780807,
      "learning_rate": 5.44325239977414e-06,
      "loss": 0.1508,
      "step": 38670
    },
    {
      "epoch": 7.28,
      "grad_norm": 0.02977585978806019,
      "learning_rate": 5.439488048183701e-06,
      "loss": 0.1399,
      "step": 38680
    },
    {
      "epoch": 7.28,
      "grad_norm": 10.193829536437988,
      "learning_rate": 5.435723696593263e-06,
      "loss": 0.3884,
      "step": 38690
    },
    {
      "epoch": 7.28,
      "grad_norm": 49.255496978759766,
      "learning_rate": 5.431959345002824e-06,
      "loss": 0.1249,
      "step": 38700
    },
    {
      "epoch": 7.29,
      "grad_norm": 0.08078165352344513,
      "learning_rate": 5.4281949934123855e-06,
      "loss": 0.1351,
      "step": 38710
    },
    {
      "epoch": 7.29,
      "grad_norm": 2.833468437194824,
      "learning_rate": 5.424430641821946e-06,
      "loss": 0.2557,
      "step": 38720
    },
    {
      "epoch": 7.29,
      "grad_norm": 22.911283493041992,
      "learning_rate": 5.420666290231508e-06,
      "loss": 0.1277,
      "step": 38730
    },
    {
      "epoch": 7.29,
      "grad_norm": 18.089767456054688,
      "learning_rate": 5.416901938641069e-06,
      "loss": 0.4499,
      "step": 38740
    },
    {
      "epoch": 7.29,
      "grad_norm": 1.4847962856292725,
      "learning_rate": 5.4131375870506305e-06,
      "loss": 0.1407,
      "step": 38750
    },
    {
      "epoch": 7.3,
      "grad_norm": 21.403898239135742,
      "learning_rate": 5.409373235460193e-06,
      "loss": 0.3395,
      "step": 38760
    },
    {
      "epoch": 7.3,
      "grad_norm": 9.80564022064209,
      "learning_rate": 5.405608883869754e-06,
      "loss": 0.293,
      "step": 38770
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.008273954503238201,
      "learning_rate": 5.401844532279316e-06,
      "loss": 0.0136,
      "step": 38780
    },
    {
      "epoch": 7.3,
      "grad_norm": 0.004594031721353531,
      "learning_rate": 5.398080180688877e-06,
      "loss": 0.0312,
      "step": 38790
    },
    {
      "epoch": 7.3,
      "grad_norm": 1.1656910181045532,
      "learning_rate": 5.394315829098439e-06,
      "loss": 0.0727,
      "step": 38800
    },
    {
      "epoch": 7.3,
      "grad_norm": 4.067411422729492,
      "learning_rate": 5.390551477507999e-06,
      "loss": 0.1429,
      "step": 38810
    },
    {
      "epoch": 7.31,
      "grad_norm": 30.5484676361084,
      "learning_rate": 5.386787125917561e-06,
      "loss": 0.3267,
      "step": 38820
    },
    {
      "epoch": 7.31,
      "grad_norm": 22.908485412597656,
      "learning_rate": 5.383022774327122e-06,
      "loss": 0.1667,
      "step": 38830
    },
    {
      "epoch": 7.31,
      "grad_norm": 13.712990760803223,
      "learning_rate": 5.379258422736684e-06,
      "loss": 0.2746,
      "step": 38840
    },
    {
      "epoch": 7.31,
      "grad_norm": 7.670235633850098,
      "learning_rate": 5.375494071146246e-06,
      "loss": 0.1715,
      "step": 38850
    },
    {
      "epoch": 7.31,
      "grad_norm": 0.007190734148025513,
      "learning_rate": 5.3717297195558074e-06,
      "loss": 0.0771,
      "step": 38860
    },
    {
      "epoch": 7.32,
      "grad_norm": 10.340008735656738,
      "learning_rate": 5.367965367965369e-06,
      "loss": 0.2075,
      "step": 38870
    },
    {
      "epoch": 7.32,
      "grad_norm": 0.0040046884678304195,
      "learning_rate": 5.36420101637493e-06,
      "loss": 0.2528,
      "step": 38880
    },
    {
      "epoch": 7.32,
      "grad_norm": 15.631004333496094,
      "learning_rate": 5.360436664784491e-06,
      "loss": 0.3905,
      "step": 38890
    },
    {
      "epoch": 7.32,
      "grad_norm": 12.118378639221191,
      "learning_rate": 5.3566723131940524e-06,
      "loss": 0.0807,
      "step": 38900
    },
    {
      "epoch": 7.32,
      "grad_norm": 22.30154037475586,
      "learning_rate": 5.352907961603614e-06,
      "loss": 0.1724,
      "step": 38910
    },
    {
      "epoch": 7.33,
      "grad_norm": 7.1235456466674805,
      "learning_rate": 5.349143610013175e-06,
      "loss": 0.0643,
      "step": 38920
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.2004988044500351,
      "learning_rate": 5.345379258422737e-06,
      "loss": 0.2372,
      "step": 38930
    },
    {
      "epoch": 7.33,
      "grad_norm": 6.889286041259766,
      "learning_rate": 5.341614906832298e-06,
      "loss": 0.2121,
      "step": 38940
    },
    {
      "epoch": 7.33,
      "grad_norm": 22.12773895263672,
      "learning_rate": 5.337850555241861e-06,
      "loss": 0.0549,
      "step": 38950
    },
    {
      "epoch": 7.33,
      "grad_norm": 0.061212118715047836,
      "learning_rate": 5.334086203651422e-06,
      "loss": 0.0728,
      "step": 38960
    },
    {
      "epoch": 7.33,
      "grad_norm": 17.19682502746582,
      "learning_rate": 5.3303218520609835e-06,
      "loss": 0.2462,
      "step": 38970
    },
    {
      "epoch": 7.34,
      "grad_norm": 19.97817039489746,
      "learning_rate": 5.326557500470544e-06,
      "loss": 0.1671,
      "step": 38980
    },
    {
      "epoch": 7.34,
      "grad_norm": 19.212127685546875,
      "learning_rate": 5.322793148880106e-06,
      "loss": 0.0732,
      "step": 38990
    },
    {
      "epoch": 7.34,
      "grad_norm": 0.18936102092266083,
      "learning_rate": 5.319028797289667e-06,
      "loss": 0.0838,
      "step": 39000
    },
    {
      "epoch": 7.34,
      "grad_norm": 0.012450424022972584,
      "learning_rate": 5.3152644456992285e-06,
      "loss": 0.1601,
      "step": 39010
    },
    {
      "epoch": 7.34,
      "grad_norm": 2.534027338027954,
      "learning_rate": 5.31150009410879e-06,
      "loss": 0.2486,
      "step": 39020
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.09095486998558044,
      "learning_rate": 5.307735742518351e-06,
      "loss": 0.1787,
      "step": 39030
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.004378501325845718,
      "learning_rate": 5.303971390927914e-06,
      "loss": 0.0379,
      "step": 39040
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.047294098883867264,
      "learning_rate": 5.300207039337475e-06,
      "loss": 0.3415,
      "step": 39050
    },
    {
      "epoch": 7.35,
      "grad_norm": 4.161035060882568,
      "learning_rate": 5.296442687747037e-06,
      "loss": 0.0718,
      "step": 39060
    },
    {
      "epoch": 7.35,
      "grad_norm": 0.05698128417134285,
      "learning_rate": 5.292678336156597e-06,
      "loss": 0.2909,
      "step": 39070
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.0046559725888073444,
      "learning_rate": 5.288913984566159e-06,
      "loss": 0.1281,
      "step": 39080
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.7239276170730591,
      "learning_rate": 5.28514963297572e-06,
      "loss": 0.1715,
      "step": 39090
    },
    {
      "epoch": 7.36,
      "grad_norm": 17.10782241821289,
      "learning_rate": 5.281385281385282e-06,
      "loss": 0.3368,
      "step": 39100
    },
    {
      "epoch": 7.36,
      "grad_norm": 9.772466659545898,
      "learning_rate": 5.277620929794843e-06,
      "loss": 0.0595,
      "step": 39110
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.04447704926133156,
      "learning_rate": 5.2738565782044046e-06,
      "loss": 0.0797,
      "step": 39120
    },
    {
      "epoch": 7.36,
      "grad_norm": 0.6292139291763306,
      "learning_rate": 5.270092226613967e-06,
      "loss": 0.1134,
      "step": 39130
    },
    {
      "epoch": 7.37,
      "grad_norm": 3.4142096042633057,
      "learning_rate": 5.266327875023528e-06,
      "loss": 0.2548,
      "step": 39140
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.0126120550557971,
      "learning_rate": 5.26256352343309e-06,
      "loss": 0.064,
      "step": 39150
    },
    {
      "epoch": 7.37,
      "grad_norm": 0.6419405341148376,
      "learning_rate": 5.25879917184265e-06,
      "loss": 0.228,
      "step": 39160
    },
    {
      "epoch": 7.37,
      "grad_norm": 26.479373931884766,
      "learning_rate": 5.255034820252212e-06,
      "loss": 0.1785,
      "step": 39170
    },
    {
      "epoch": 7.37,
      "grad_norm": 2.510162591934204,
      "learning_rate": 5.251270468661773e-06,
      "loss": 0.1844,
      "step": 39180
    },
    {
      "epoch": 7.38,
      "grad_norm": 0.00876823253929615,
      "learning_rate": 5.247506117071335e-06,
      "loss": 0.1882,
      "step": 39190
    },
    {
      "epoch": 7.38,
      "grad_norm": 32.219417572021484,
      "learning_rate": 5.243741765480896e-06,
      "loss": 0.2089,
      "step": 39200
    },
    {
      "epoch": 7.38,
      "grad_norm": 7.608548164367676,
      "learning_rate": 5.239977413890458e-06,
      "loss": 0.0466,
      "step": 39210
    },
    {
      "epoch": 7.38,
      "grad_norm": 37.02949905395508,
      "learning_rate": 5.236213062300019e-06,
      "loss": 0.46,
      "step": 39220
    },
    {
      "epoch": 7.38,
      "grad_norm": 0.134820356965065,
      "learning_rate": 5.2324487107095815e-06,
      "loss": 0.0893,
      "step": 39230
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.05084793269634247,
      "learning_rate": 5.228684359119141e-06,
      "loss": 0.0207,
      "step": 39240
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.052201513200998306,
      "learning_rate": 5.2249200075287036e-06,
      "loss": 0.0709,
      "step": 39250
    },
    {
      "epoch": 7.39,
      "grad_norm": 8.243937492370605,
      "learning_rate": 5.221155655938265e-06,
      "loss": 0.0923,
      "step": 39260
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.0638815239071846,
      "learning_rate": 5.2173913043478265e-06,
      "loss": 0.0378,
      "step": 39270
    },
    {
      "epoch": 7.39,
      "grad_norm": 0.16850541532039642,
      "learning_rate": 5.213626952757388e-06,
      "loss": 0.0786,
      "step": 39280
    },
    {
      "epoch": 7.4,
      "grad_norm": 1.1487064361572266,
      "learning_rate": 5.209862601166949e-06,
      "loss": 0.1133,
      "step": 39290
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.02061787061393261,
      "learning_rate": 5.206098249576511e-06,
      "loss": 0.0791,
      "step": 39300
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.05056507885456085,
      "learning_rate": 5.202333897986072e-06,
      "loss": 0.1283,
      "step": 39310
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.03451072797179222,
      "learning_rate": 5.198569546395635e-06,
      "loss": 0.1189,
      "step": 39320
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.02850453555583954,
      "learning_rate": 5.194805194805194e-06,
      "loss": 0.1355,
      "step": 39330
    },
    {
      "epoch": 7.4,
      "grad_norm": 19.00522232055664,
      "learning_rate": 5.191040843214757e-06,
      "loss": 0.2664,
      "step": 39340
    },
    {
      "epoch": 7.41,
      "grad_norm": 4.848191261291504,
      "learning_rate": 5.187276491624318e-06,
      "loss": 0.0276,
      "step": 39350
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.12228427827358246,
      "learning_rate": 5.18351214003388e-06,
      "loss": 0.0973,
      "step": 39360
    },
    {
      "epoch": 7.41,
      "grad_norm": 1.9151418209075928,
      "learning_rate": 5.179747788443441e-06,
      "loss": 0.2443,
      "step": 39370
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.8809740543365479,
      "learning_rate": 5.1759834368530025e-06,
      "loss": 0.2422,
      "step": 39380
    },
    {
      "epoch": 7.41,
      "grad_norm": 0.1613714098930359,
      "learning_rate": 5.172219085262564e-06,
      "loss": 0.0507,
      "step": 39390
    },
    {
      "epoch": 7.42,
      "grad_norm": 25.76273536682129,
      "learning_rate": 5.1684547336721255e-06,
      "loss": 0.2196,
      "step": 39400
    },
    {
      "epoch": 7.42,
      "grad_norm": 9.385035514831543,
      "learning_rate": 5.164690382081688e-06,
      "loss": 0.1323,
      "step": 39410
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.21381884813308716,
      "learning_rate": 5.1609260304912475e-06,
      "loss": 0.3214,
      "step": 39420
    },
    {
      "epoch": 7.42,
      "grad_norm": 0.01206116285175085,
      "learning_rate": 5.15716167890081e-06,
      "loss": 0.1036,
      "step": 39430
    },
    {
      "epoch": 7.42,
      "grad_norm": 26.895349502563477,
      "learning_rate": 5.153397327310371e-06,
      "loss": 0.2876,
      "step": 39440
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.6290885210037231,
      "learning_rate": 5.149632975719933e-06,
      "loss": 0.1774,
      "step": 39450
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.003787095658481121,
      "learning_rate": 5.145868624129494e-06,
      "loss": 0.1291,
      "step": 39460
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.5000643134117126,
      "learning_rate": 5.142104272539056e-06,
      "loss": 0.1641,
      "step": 39470
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.010159430094063282,
      "learning_rate": 5.138339920948617e-06,
      "loss": 0.2732,
      "step": 39480
    },
    {
      "epoch": 7.43,
      "grad_norm": 1.3747875690460205,
      "learning_rate": 5.134575569358179e-06,
      "loss": 0.1275,
      "step": 39490
    },
    {
      "epoch": 7.43,
      "grad_norm": 0.9450660943984985,
      "learning_rate": 5.130811217767739e-06,
      "loss": 0.1722,
      "step": 39500
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.012098622508347034,
      "learning_rate": 5.127046866177301e-06,
      "loss": 0.0835,
      "step": 39510
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.02264981158077717,
      "learning_rate": 5.123282514586862e-06,
      "loss": 0.1884,
      "step": 39520
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.027598433196544647,
      "learning_rate": 5.1195181629964245e-06,
      "loss": 0.1313,
      "step": 39530
    },
    {
      "epoch": 7.44,
      "grad_norm": 0.021403266116976738,
      "learning_rate": 5.115753811405986e-06,
      "loss": 0.058,
      "step": 39540
    },
    {
      "epoch": 7.44,
      "grad_norm": 2.5614991188049316,
      "learning_rate": 5.111989459815547e-06,
      "loss": 0.1454,
      "step": 39550
    },
    {
      "epoch": 7.45,
      "grad_norm": 1.690061092376709,
      "learning_rate": 5.108225108225109e-06,
      "loss": 0.1116,
      "step": 39560
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.06443335860967636,
      "learning_rate": 5.10446075663467e-06,
      "loss": 0.2111,
      "step": 39570
    },
    {
      "epoch": 7.45,
      "grad_norm": 29.796432495117188,
      "learning_rate": 5.100696405044232e-06,
      "loss": 0.4029,
      "step": 39580
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.18142405152320862,
      "learning_rate": 5.096932053453792e-06,
      "loss": 0.2187,
      "step": 39590
    },
    {
      "epoch": 7.45,
      "grad_norm": 0.11493239551782608,
      "learning_rate": 5.093167701863354e-06,
      "loss": 0.2016,
      "step": 39600
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.040252551436424255,
      "learning_rate": 5.089403350272915e-06,
      "loss": 0.146,
      "step": 39610
    },
    {
      "epoch": 7.46,
      "grad_norm": 4.120568752288818,
      "learning_rate": 5.085638998682478e-06,
      "loss": 0.2219,
      "step": 39620
    },
    {
      "epoch": 7.46,
      "grad_norm": 36.79468536376953,
      "learning_rate": 5.081874647092039e-06,
      "loss": 0.1821,
      "step": 39630
    },
    {
      "epoch": 7.46,
      "grad_norm": 20.557336807250977,
      "learning_rate": 5.0781102955016005e-06,
      "loss": 0.0479,
      "step": 39640
    },
    {
      "epoch": 7.46,
      "grad_norm": 10.032586097717285,
      "learning_rate": 5.074345943911162e-06,
      "loss": 0.3843,
      "step": 39650
    },
    {
      "epoch": 7.46,
      "grad_norm": 0.16239017248153687,
      "learning_rate": 5.0705815923207234e-06,
      "loss": 0.0597,
      "step": 39660
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.012492709793150425,
      "learning_rate": 5.066817240730285e-06,
      "loss": 0.0711,
      "step": 39670
    },
    {
      "epoch": 7.47,
      "grad_norm": 54.2467155456543,
      "learning_rate": 5.0630528891398455e-06,
      "loss": 0.1679,
      "step": 39680
    },
    {
      "epoch": 7.47,
      "grad_norm": 2.8411734104156494,
      "learning_rate": 5.059288537549407e-06,
      "loss": 0.1066,
      "step": 39690
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.10053247213363647,
      "learning_rate": 5.0555241859589684e-06,
      "loss": 0.1592,
      "step": 39700
    },
    {
      "epoch": 7.47,
      "grad_norm": 0.36260905861854553,
      "learning_rate": 5.051759834368531e-06,
      "loss": 0.2423,
      "step": 39710
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.08345061540603638,
      "learning_rate": 5.047995482778092e-06,
      "loss": 0.0928,
      "step": 39720
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.35413238406181335,
      "learning_rate": 5.044231131187654e-06,
      "loss": 0.1468,
      "step": 39730
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.04868623986840248,
      "learning_rate": 5.040466779597215e-06,
      "loss": 0.117,
      "step": 39740
    },
    {
      "epoch": 7.48,
      "grad_norm": 0.0113257160410285,
      "learning_rate": 5.036702428006777e-06,
      "loss": 0.232,
      "step": 39750
    },
    {
      "epoch": 7.48,
      "grad_norm": 7.5144243240356445,
      "learning_rate": 5.032938076416338e-06,
      "loss": 0.137,
      "step": 39760
    },
    {
      "epoch": 7.49,
      "grad_norm": 5.231700420379639,
      "learning_rate": 5.029173724825899e-06,
      "loss": 0.1531,
      "step": 39770
    },
    {
      "epoch": 7.49,
      "grad_norm": 5.67008113861084,
      "learning_rate": 5.02540937323546e-06,
      "loss": 0.6299,
      "step": 39780
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.008743156678974628,
      "learning_rate": 5.021645021645022e-06,
      "loss": 0.1135,
      "step": 39790
    },
    {
      "epoch": 7.49,
      "grad_norm": 3.0704047679901123,
      "learning_rate": 5.017880670054583e-06,
      "loss": 0.2179,
      "step": 39800
    },
    {
      "epoch": 7.49,
      "grad_norm": 10.4718599319458,
      "learning_rate": 5.014116318464145e-06,
      "loss": 0.1777,
      "step": 39810
    },
    {
      "epoch": 7.49,
      "grad_norm": 0.01104668341577053,
      "learning_rate": 5.010351966873707e-06,
      "loss": 0.3378,
      "step": 39820
    },
    {
      "epoch": 7.5,
      "grad_norm": 4.447186470031738,
      "learning_rate": 5.006587615283268e-06,
      "loss": 0.1662,
      "step": 39830
    },
    {
      "epoch": 7.5,
      "grad_norm": 19.11048126220703,
      "learning_rate": 5.00282326369283e-06,
      "loss": 0.1999,
      "step": 39840
    },
    {
      "epoch": 7.5,
      "grad_norm": 8.681451797485352,
      "learning_rate": 4.999058912102391e-06,
      "loss": 0.1903,
      "step": 39850
    },
    {
      "epoch": 7.5,
      "grad_norm": 6.181904315948486,
      "learning_rate": 4.995294560511953e-06,
      "loss": 0.1218,
      "step": 39860
    },
    {
      "epoch": 7.5,
      "grad_norm": 1.9590243101119995,
      "learning_rate": 4.991530208921513e-06,
      "loss": 0.1295,
      "step": 39870
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.7368337512016296,
      "learning_rate": 4.987765857331075e-06,
      "loss": 0.062,
      "step": 39880
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.05597082898020744,
      "learning_rate": 4.984001505740636e-06,
      "loss": 0.3442,
      "step": 39890
    },
    {
      "epoch": 7.51,
      "grad_norm": 33.57429504394531,
      "learning_rate": 4.9802371541501985e-06,
      "loss": 0.2824,
      "step": 39900
    },
    {
      "epoch": 7.51,
      "grad_norm": 12.05380916595459,
      "learning_rate": 4.97647280255976e-06,
      "loss": 0.0562,
      "step": 39910
    },
    {
      "epoch": 7.51,
      "grad_norm": 0.006315217819064856,
      "learning_rate": 4.9727084509693206e-06,
      "loss": 0.0901,
      "step": 39920
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.1153237521648407,
      "learning_rate": 4.968944099378882e-06,
      "loss": 0.1684,
      "step": 39930
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.00791832897812128,
      "learning_rate": 4.9651797477884435e-06,
      "loss": 0.3444,
      "step": 39940
    },
    {
      "epoch": 7.52,
      "grad_norm": 10.468094825744629,
      "learning_rate": 4.961415396198006e-06,
      "loss": 0.0613,
      "step": 39950
    },
    {
      "epoch": 7.52,
      "grad_norm": 19.32353401184082,
      "learning_rate": 4.957651044607566e-06,
      "loss": 0.2235,
      "step": 39960
    },
    {
      "epoch": 7.52,
      "grad_norm": 0.023652248084545135,
      "learning_rate": 4.953886693017128e-06,
      "loss": 0.157,
      "step": 39970
    },
    {
      "epoch": 7.52,
      "grad_norm": 6.914787769317627,
      "learning_rate": 4.950122341426689e-06,
      "loss": 0.2788,
      "step": 39980
    },
    {
      "epoch": 7.53,
      "grad_norm": 13.747199058532715,
      "learning_rate": 4.946357989836252e-06,
      "loss": 0.3791,
      "step": 39990
    },
    {
      "epoch": 7.53,
      "grad_norm": 2.324418783187866,
      "learning_rate": 4.942593638245812e-06,
      "loss": 0.1125,
      "step": 40000
    },
    {
      "epoch": 7.53,
      "grad_norm": 11.055949211120605,
      "learning_rate": 4.938829286655374e-06,
      "loss": 0.1841,
      "step": 40010
    },
    {
      "epoch": 7.53,
      "grad_norm": 1.663029670715332,
      "learning_rate": 4.935064935064935e-06,
      "loss": 0.1484,
      "step": 40020
    },
    {
      "epoch": 7.53,
      "grad_norm": 11.327677726745605,
      "learning_rate": 4.931300583474497e-06,
      "loss": 0.2824,
      "step": 40030
    },
    {
      "epoch": 7.54,
      "grad_norm": 4.7016777992248535,
      "learning_rate": 4.927536231884059e-06,
      "loss": 0.2717,
      "step": 40040
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.9359035491943359,
      "learning_rate": 4.9237718802936196e-06,
      "loss": 0.0378,
      "step": 40050
    },
    {
      "epoch": 7.54,
      "grad_norm": 31.073671340942383,
      "learning_rate": 4.920007528703181e-06,
      "loss": 0.2035,
      "step": 40060
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.003570201341062784,
      "learning_rate": 4.9162431771127425e-06,
      "loss": 0.2046,
      "step": 40070
    },
    {
      "epoch": 7.54,
      "grad_norm": 0.45042768120765686,
      "learning_rate": 4.912478825522305e-06,
      "loss": 0.1853,
      "step": 40080
    },
    {
      "epoch": 7.55,
      "grad_norm": 9.497610092163086,
      "learning_rate": 4.908714473931865e-06,
      "loss": 0.1349,
      "step": 40090
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.05903337523341179,
      "learning_rate": 4.904950122341427e-06,
      "loss": 0.0855,
      "step": 40100
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.007909747771918774,
      "learning_rate": 4.901185770750988e-06,
      "loss": 0.1239,
      "step": 40110
    },
    {
      "epoch": 7.55,
      "grad_norm": 0.05143298953771591,
      "learning_rate": 4.89742141916055e-06,
      "loss": 0.0509,
      "step": 40120
    },
    {
      "epoch": 7.55,
      "grad_norm": 3.066380500793457,
      "learning_rate": 4.893657067570112e-06,
      "loss": 0.2185,
      "step": 40130
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.12315992265939713,
      "learning_rate": 4.889892715979673e-06,
      "loss": 0.0555,
      "step": 40140
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.47490784525871277,
      "learning_rate": 4.886128364389234e-06,
      "loss": 0.0877,
      "step": 40150
    },
    {
      "epoch": 7.56,
      "grad_norm": 26.085853576660156,
      "learning_rate": 4.882364012798796e-06,
      "loss": 0.1132,
      "step": 40160
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.013324011117219925,
      "learning_rate": 4.878599661208357e-06,
      "loss": 0.1274,
      "step": 40170
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.06495288014411926,
      "learning_rate": 4.8748353096179186e-06,
      "loss": 0.0386,
      "step": 40180
    },
    {
      "epoch": 7.56,
      "grad_norm": 0.11250825226306915,
      "learning_rate": 4.87107095802748e-06,
      "loss": 0.1004,
      "step": 40190
    },
    {
      "epoch": 7.57,
      "grad_norm": 18.87899398803711,
      "learning_rate": 4.8673066064370415e-06,
      "loss": 0.2099,
      "step": 40200
    },
    {
      "epoch": 7.57,
      "grad_norm": 18.417194366455078,
      "learning_rate": 4.863542254846603e-06,
      "loss": 0.147,
      "step": 40210
    },
    {
      "epoch": 7.57,
      "grad_norm": 8.690559387207031,
      "learning_rate": 4.859777903256164e-06,
      "loss": 0.2025,
      "step": 40220
    },
    {
      "epoch": 7.57,
      "grad_norm": 31.1717472076416,
      "learning_rate": 4.856013551665726e-06,
      "loss": 0.2798,
      "step": 40230
    },
    {
      "epoch": 7.57,
      "grad_norm": 3.0252695083618164,
      "learning_rate": 4.852249200075287e-06,
      "loss": 0.1158,
      "step": 40240
    },
    {
      "epoch": 7.58,
      "grad_norm": 6.495982646942139,
      "learning_rate": 4.848484848484849e-06,
      "loss": 0.2548,
      "step": 40250
    },
    {
      "epoch": 7.58,
      "grad_norm": 14.491064071655273,
      "learning_rate": 4.84472049689441e-06,
      "loss": 0.1639,
      "step": 40260
    },
    {
      "epoch": 7.58,
      "grad_norm": 11.582160949707031,
      "learning_rate": 4.840956145303972e-06,
      "loss": 0.0549,
      "step": 40270
    },
    {
      "epoch": 7.58,
      "grad_norm": 0.12002509087324142,
      "learning_rate": 4.837191793713533e-06,
      "loss": 0.1021,
      "step": 40280
    },
    {
      "epoch": 7.58,
      "grad_norm": 0.6293095350265503,
      "learning_rate": 4.833427442123095e-06,
      "loss": 0.1939,
      "step": 40290
    },
    {
      "epoch": 7.59,
      "grad_norm": 14.691259384155273,
      "learning_rate": 4.829663090532656e-06,
      "loss": 0.1787,
      "step": 40300
    },
    {
      "epoch": 7.59,
      "grad_norm": 29.179521560668945,
      "learning_rate": 4.8258987389422175e-06,
      "loss": 0.5372,
      "step": 40310
    },
    {
      "epoch": 7.59,
      "grad_norm": 1.0774072408676147,
      "learning_rate": 4.822134387351779e-06,
      "loss": 0.1153,
      "step": 40320
    },
    {
      "epoch": 7.59,
      "grad_norm": 3.074634313583374,
      "learning_rate": 4.8183700357613405e-06,
      "loss": 0.1982,
      "step": 40330
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.542233943939209,
      "learning_rate": 4.814605684170902e-06,
      "loss": 0.0578,
      "step": 40340
    },
    {
      "epoch": 7.59,
      "grad_norm": 37.34349822998047,
      "learning_rate": 4.810841332580463e-06,
      "loss": 0.1273,
      "step": 40350
    },
    {
      "epoch": 7.6,
      "grad_norm": 1.0747956037521362,
      "learning_rate": 4.807076980990025e-06,
      "loss": 0.0898,
      "step": 40360
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.1217997670173645,
      "learning_rate": 4.803312629399586e-06,
      "loss": 0.3409,
      "step": 40370
    },
    {
      "epoch": 7.6,
      "grad_norm": 0.09144926816225052,
      "learning_rate": 4.799548277809148e-06,
      "loss": 0.2615,
      "step": 40380
    },
    {
      "epoch": 7.6,
      "grad_norm": 16.74960708618164,
      "learning_rate": 4.795783926218709e-06,
      "loss": 0.323,
      "step": 40390
    },
    {
      "epoch": 7.6,
      "grad_norm": 2.2639901638031006,
      "learning_rate": 4.792019574628271e-06,
      "loss": 0.1759,
      "step": 40400
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.4358842968940735,
      "learning_rate": 4.788255223037832e-06,
      "loss": 0.0515,
      "step": 40410
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.003917002584785223,
      "learning_rate": 4.784490871447394e-06,
      "loss": 0.0877,
      "step": 40420
    },
    {
      "epoch": 7.61,
      "grad_norm": 2.73468017578125,
      "learning_rate": 4.780726519856955e-06,
      "loss": 0.0804,
      "step": 40430
    },
    {
      "epoch": 7.61,
      "grad_norm": 7.225642681121826,
      "learning_rate": 4.7769621682665165e-06,
      "loss": 0.1217,
      "step": 40440
    },
    {
      "epoch": 7.61,
      "grad_norm": 0.6331893801689148,
      "learning_rate": 4.773197816676078e-06,
      "loss": 0.114,
      "step": 40450
    },
    {
      "epoch": 7.62,
      "grad_norm": 5.758376598358154,
      "learning_rate": 4.7694334650856395e-06,
      "loss": 0.2163,
      "step": 40460
    },
    {
      "epoch": 7.62,
      "grad_norm": 3.2266881465911865,
      "learning_rate": 4.765669113495201e-06,
      "loss": 0.1127,
      "step": 40470
    },
    {
      "epoch": 7.62,
      "grad_norm": 5.861881732940674,
      "learning_rate": 4.761904761904762e-06,
      "loss": 0.1016,
      "step": 40480
    },
    {
      "epoch": 7.62,
      "grad_norm": 0.022121459245681763,
      "learning_rate": 4.758140410314324e-06,
      "loss": 0.14,
      "step": 40490
    },
    {
      "epoch": 7.62,
      "grad_norm": 4.89031982421875,
      "learning_rate": 4.754376058723885e-06,
      "loss": 0.303,
      "step": 40500
    },
    {
      "epoch": 7.62,
      "grad_norm": 21.865713119506836,
      "learning_rate": 4.750611707133447e-06,
      "loss": 0.2289,
      "step": 40510
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.4887808859348297,
      "learning_rate": 4.746847355543008e-06,
      "loss": 0.0503,
      "step": 40520
    },
    {
      "epoch": 7.63,
      "grad_norm": 7.459065914154053,
      "learning_rate": 4.74308300395257e-06,
      "loss": 0.1477,
      "step": 40530
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.043394479900598526,
      "learning_rate": 4.739318652362131e-06,
      "loss": 0.1742,
      "step": 40540
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.3963397741317749,
      "learning_rate": 4.735554300771693e-06,
      "loss": 0.1035,
      "step": 40550
    },
    {
      "epoch": 7.63,
      "grad_norm": 0.008845110423862934,
      "learning_rate": 4.731789949181254e-06,
      "loss": 0.0659,
      "step": 40560
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.046063486486673355,
      "learning_rate": 4.7280255975908155e-06,
      "loss": 0.1613,
      "step": 40570
    },
    {
      "epoch": 7.64,
      "grad_norm": 3.114840030670166,
      "learning_rate": 4.724261246000377e-06,
      "loss": 0.2443,
      "step": 40580
    },
    {
      "epoch": 7.64,
      "grad_norm": 1.3649768829345703,
      "learning_rate": 4.7204968944099384e-06,
      "loss": 0.3564,
      "step": 40590
    },
    {
      "epoch": 7.64,
      "grad_norm": 12.85567855834961,
      "learning_rate": 4.7167325428195e-06,
      "loss": 0.1596,
      "step": 40600
    },
    {
      "epoch": 7.64,
      "grad_norm": 0.0079322075471282,
      "learning_rate": 4.7129681912290605e-06,
      "loss": 0.1313,
      "step": 40610
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.0295166727155447,
      "learning_rate": 4.709203839638623e-06,
      "loss": 0.2005,
      "step": 40620
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.2343994826078415,
      "learning_rate": 4.705439488048184e-06,
      "loss": 0.1638,
      "step": 40630
    },
    {
      "epoch": 7.65,
      "grad_norm": 31.965316772460938,
      "learning_rate": 4.701675136457746e-06,
      "loss": 0.2956,
      "step": 40640
    },
    {
      "epoch": 7.65,
      "grad_norm": 13.869183540344238,
      "learning_rate": 4.697910784867307e-06,
      "loss": 0.2812,
      "step": 40650
    },
    {
      "epoch": 7.65,
      "grad_norm": 0.08952552825212479,
      "learning_rate": 4.694146433276869e-06,
      "loss": 0.1198,
      "step": 40660
    },
    {
      "epoch": 7.65,
      "grad_norm": 36.9962158203125,
      "learning_rate": 4.69038208168643e-06,
      "loss": 0.1694,
      "step": 40670
    },
    {
      "epoch": 7.66,
      "grad_norm": 10.865767478942871,
      "learning_rate": 4.686617730095992e-06,
      "loss": 0.2095,
      "step": 40680
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.009248302318155766,
      "learning_rate": 4.682853378505553e-06,
      "loss": 0.1163,
      "step": 40690
    },
    {
      "epoch": 7.66,
      "grad_norm": 21.18915557861328,
      "learning_rate": 4.679089026915114e-06,
      "loss": 0.2841,
      "step": 40700
    },
    {
      "epoch": 7.66,
      "grad_norm": 0.5230027437210083,
      "learning_rate": 4.675324675324676e-06,
      "loss": 0.1791,
      "step": 40710
    },
    {
      "epoch": 7.66,
      "grad_norm": 32.820159912109375,
      "learning_rate": 4.6715603237342374e-06,
      "loss": 0.3476,
      "step": 40720
    },
    {
      "epoch": 7.67,
      "grad_norm": 5.358686447143555,
      "learning_rate": 4.667795972143799e-06,
      "loss": 0.202,
      "step": 40730
    },
    {
      "epoch": 7.67,
      "grad_norm": 17.790075302124023,
      "learning_rate": 4.66403162055336e-06,
      "loss": 0.2912,
      "step": 40740
    },
    {
      "epoch": 7.67,
      "grad_norm": 0.2017471045255661,
      "learning_rate": 4.660267268962921e-06,
      "loss": 0.1351,
      "step": 40750
    },
    {
      "epoch": 7.67,
      "grad_norm": 3.2144434452056885,
      "learning_rate": 4.656502917372483e-06,
      "loss": 0.0955,
      "step": 40760
    },
    {
      "epoch": 7.67,
      "grad_norm": 10.137898445129395,
      "learning_rate": 4.652738565782045e-06,
      "loss": 0.2191,
      "step": 40770
    },
    {
      "epoch": 7.68,
      "grad_norm": 4.99985408782959,
      "learning_rate": 4.648974214191606e-06,
      "loss": 0.019,
      "step": 40780
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.010862770490348339,
      "learning_rate": 4.645209862601167e-06,
      "loss": 0.2113,
      "step": 40790
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.08114121854305267,
      "learning_rate": 4.641445511010729e-06,
      "loss": 0.2564,
      "step": 40800
    },
    {
      "epoch": 7.68,
      "grad_norm": 4.430674076080322,
      "learning_rate": 4.637681159420291e-06,
      "loss": 0.0608,
      "step": 40810
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.011408709920942783,
      "learning_rate": 4.633916807829852e-06,
      "loss": 0.1259,
      "step": 40820
    },
    {
      "epoch": 7.68,
      "grad_norm": 0.2039790004491806,
      "learning_rate": 4.630152456239413e-06,
      "loss": 0.198,
      "step": 40830
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.02268284000456333,
      "learning_rate": 4.626388104648974e-06,
      "loss": 0.0444,
      "step": 40840
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.004437791649252176,
      "learning_rate": 4.622623753058536e-06,
      "loss": 0.0262,
      "step": 40850
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.0457296222448349,
      "learning_rate": 4.618859401468098e-06,
      "loss": 0.1314,
      "step": 40860
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.02917664684355259,
      "learning_rate": 4.615095049877659e-06,
      "loss": 0.1192,
      "step": 40870
    },
    {
      "epoch": 7.69,
      "grad_norm": 0.010614867322146893,
      "learning_rate": 4.61133069828722e-06,
      "loss": 0.0605,
      "step": 40880
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.06369317322969437,
      "learning_rate": 4.607566346696781e-06,
      "loss": 0.0245,
      "step": 40890
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.007895002141594887,
      "learning_rate": 4.603801995106344e-06,
      "loss": 0.1949,
      "step": 40900
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.21658062934875488,
      "learning_rate": 4.600037643515905e-06,
      "loss": 0.2505,
      "step": 40910
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.34792912006378174,
      "learning_rate": 4.596273291925466e-06,
      "loss": 0.2174,
      "step": 40920
    },
    {
      "epoch": 7.7,
      "grad_norm": 0.8741416931152344,
      "learning_rate": 4.592508940335027e-06,
      "loss": 0.2502,
      "step": 40930
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.044663406908512115,
      "learning_rate": 4.5887445887445896e-06,
      "loss": 0.2541,
      "step": 40940
    },
    {
      "epoch": 7.71,
      "grad_norm": 48.81891632080078,
      "learning_rate": 4.584980237154151e-06,
      "loss": 0.1395,
      "step": 40950
    },
    {
      "epoch": 7.71,
      "grad_norm": 14.240748405456543,
      "learning_rate": 4.581215885563712e-06,
      "loss": 0.1971,
      "step": 40960
    },
    {
      "epoch": 7.71,
      "grad_norm": 0.017767028883099556,
      "learning_rate": 4.577451533973273e-06,
      "loss": 0.1657,
      "step": 40970
    },
    {
      "epoch": 7.71,
      "grad_norm": 102.5630874633789,
      "learning_rate": 4.5736871823828346e-06,
      "loss": 0.1712,
      "step": 40980
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.1316167116165161,
      "learning_rate": 4.569922830792397e-06,
      "loss": 0.1874,
      "step": 40990
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.035472970455884933,
      "learning_rate": 4.566158479201958e-06,
      "loss": 0.1896,
      "step": 41000
    },
    {
      "epoch": 7.72,
      "grad_norm": 3.409102439880371,
      "learning_rate": 4.562394127611519e-06,
      "loss": 0.2869,
      "step": 41010
    },
    {
      "epoch": 7.72,
      "grad_norm": 0.003950661513954401,
      "learning_rate": 4.55862977602108e-06,
      "loss": 0.1255,
      "step": 41020
    },
    {
      "epoch": 7.72,
      "grad_norm": 12.85799789428711,
      "learning_rate": 4.554865424430642e-06,
      "loss": 0.2498,
      "step": 41030
    },
    {
      "epoch": 7.72,
      "grad_norm": 1.692420482635498,
      "learning_rate": 4.551101072840204e-06,
      "loss": 0.2518,
      "step": 41040
    },
    {
      "epoch": 7.73,
      "grad_norm": 15.429850578308105,
      "learning_rate": 4.547336721249765e-06,
      "loss": 0.2418,
      "step": 41050
    },
    {
      "epoch": 7.73,
      "grad_norm": 9.768464088439941,
      "learning_rate": 4.543572369659326e-06,
      "loss": 0.1701,
      "step": 41060
    },
    {
      "epoch": 7.73,
      "grad_norm": 9.696982383728027,
      "learning_rate": 4.539808018068888e-06,
      "loss": 0.2075,
      "step": 41070
    },
    {
      "epoch": 7.73,
      "grad_norm": 26.402116775512695,
      "learning_rate": 4.53604366647845e-06,
      "loss": 0.2712,
      "step": 41080
    },
    {
      "epoch": 7.73,
      "grad_norm": 49.9360466003418,
      "learning_rate": 4.532279314888011e-06,
      "loss": 0.1898,
      "step": 41090
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.46538323163986206,
      "learning_rate": 4.528514963297572e-06,
      "loss": 0.2007,
      "step": 41100
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.08006254583597183,
      "learning_rate": 4.5247506117071336e-06,
      "loss": 0.0286,
      "step": 41110
    },
    {
      "epoch": 7.74,
      "grad_norm": 10.88635540008545,
      "learning_rate": 4.520986260116695e-06,
      "loss": 0.1312,
      "step": 41120
    },
    {
      "epoch": 7.74,
      "grad_norm": 0.005502830259501934,
      "learning_rate": 4.517221908526257e-06,
      "loss": 0.1481,
      "step": 41130
    },
    {
      "epoch": 7.74,
      "grad_norm": 12.660408973693848,
      "learning_rate": 4.513457556935818e-06,
      "loss": 0.2221,
      "step": 41140
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.9698376655578613,
      "learning_rate": 4.509693205345379e-06,
      "loss": 0.1364,
      "step": 41150
    },
    {
      "epoch": 7.75,
      "grad_norm": 10.659947395324707,
      "learning_rate": 4.505928853754941e-06,
      "loss": 0.0999,
      "step": 41160
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.07503397762775421,
      "learning_rate": 4.502164502164502e-06,
      "loss": 0.0565,
      "step": 41170
    },
    {
      "epoch": 7.75,
      "grad_norm": 3.507465362548828,
      "learning_rate": 4.498400150574064e-06,
      "loss": 0.0429,
      "step": 41180
    },
    {
      "epoch": 7.75,
      "grad_norm": 0.00909991841763258,
      "learning_rate": 4.494635798983625e-06,
      "loss": 0.2449,
      "step": 41190
    },
    {
      "epoch": 7.75,
      "grad_norm": 10.228785514831543,
      "learning_rate": 4.490871447393187e-06,
      "loss": 0.0768,
      "step": 41200
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.1803985834121704,
      "learning_rate": 4.487107095802748e-06,
      "loss": 0.076,
      "step": 41210
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.006753656081855297,
      "learning_rate": 4.48334274421231e-06,
      "loss": 0.0534,
      "step": 41220
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.0033660547342151403,
      "learning_rate": 4.479578392621871e-06,
      "loss": 0.1227,
      "step": 41230
    },
    {
      "epoch": 7.76,
      "grad_norm": 20.917356491088867,
      "learning_rate": 4.4758140410314325e-06,
      "loss": 0.3178,
      "step": 41240
    },
    {
      "epoch": 7.76,
      "grad_norm": 0.2599021792411804,
      "learning_rate": 4.472049689440994e-06,
      "loss": 0.1153,
      "step": 41250
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.004145992454141378,
      "learning_rate": 4.4682853378505555e-06,
      "loss": 0.1879,
      "step": 41260
    },
    {
      "epoch": 7.77,
      "grad_norm": 20.99785804748535,
      "learning_rate": 4.464520986260117e-06,
      "loss": 0.1902,
      "step": 41270
    },
    {
      "epoch": 7.77,
      "grad_norm": 42.95526885986328,
      "learning_rate": 4.460756634669678e-06,
      "loss": 0.2394,
      "step": 41280
    },
    {
      "epoch": 7.77,
      "grad_norm": 0.006982968654483557,
      "learning_rate": 4.45699228307924e-06,
      "loss": 0.1859,
      "step": 41290
    },
    {
      "epoch": 7.77,
      "grad_norm": 16.12358856201172,
      "learning_rate": 4.453227931488801e-06,
      "loss": 0.1467,
      "step": 41300
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.06538297981023788,
      "learning_rate": 4.449463579898363e-06,
      "loss": 0.2926,
      "step": 41310
    },
    {
      "epoch": 7.78,
      "grad_norm": 2.2056219577789307,
      "learning_rate": 4.445699228307924e-06,
      "loss": 0.0708,
      "step": 41320
    },
    {
      "epoch": 7.78,
      "grad_norm": 2.2926669120788574,
      "learning_rate": 4.441934876717486e-06,
      "loss": 0.1185,
      "step": 41330
    },
    {
      "epoch": 7.78,
      "grad_norm": 20.53550148010254,
      "learning_rate": 4.438170525127047e-06,
      "loss": 0.2463,
      "step": 41340
    },
    {
      "epoch": 7.78,
      "grad_norm": 8.850163459777832,
      "learning_rate": 4.434406173536609e-06,
      "loss": 0.1102,
      "step": 41350
    },
    {
      "epoch": 7.78,
      "grad_norm": 0.6292569041252136,
      "learning_rate": 4.43064182194617e-06,
      "loss": 0.0927,
      "step": 41360
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.013768848963081837,
      "learning_rate": 4.4268774703557315e-06,
      "loss": 0.0357,
      "step": 41370
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.1272488832473755,
      "learning_rate": 4.423113118765293e-06,
      "loss": 0.1108,
      "step": 41380
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.011392001062631607,
      "learning_rate": 4.4193487671748545e-06,
      "loss": 0.0231,
      "step": 41390
    },
    {
      "epoch": 7.79,
      "grad_norm": 4.394837856292725,
      "learning_rate": 4.415584415584416e-06,
      "loss": 0.2386,
      "step": 41400
    },
    {
      "epoch": 7.79,
      "grad_norm": 0.01600014604628086,
      "learning_rate": 4.411820063993977e-06,
      "loss": 0.1785,
      "step": 41410
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.03302409499883652,
      "learning_rate": 4.408055712403539e-06,
      "loss": 0.1347,
      "step": 41420
    },
    {
      "epoch": 7.8,
      "grad_norm": 0.0050651999190449715,
      "learning_rate": 4.4042913608131e-06,
      "loss": 0.1611,
      "step": 41430
    },
    {
      "epoch": 7.8,
      "grad_norm": 7.309347152709961,
      "learning_rate": 4.400527009222662e-06,
      "loss": 0.3294,
      "step": 41440
    },
    {
      "epoch": 7.8,
      "grad_norm": 1.0347834825515747,
      "learning_rate": 4.396762657632223e-06,
      "loss": 0.4042,
      "step": 41450
    },
    {
      "epoch": 7.8,
      "grad_norm": 13.506361961364746,
      "learning_rate": 4.392998306041785e-06,
      "loss": 0.3109,
      "step": 41460
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.283466100692749,
      "learning_rate": 4.389233954451346e-06,
      "loss": 0.0933,
      "step": 41470
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.0035902801901102066,
      "learning_rate": 4.385469602860908e-06,
      "loss": 0.0838,
      "step": 41480
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.010214058682322502,
      "learning_rate": 4.381705251270469e-06,
      "loss": 0.0846,
      "step": 41490
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.004873054567724466,
      "learning_rate": 4.3779408996800305e-06,
      "loss": 0.2445,
      "step": 41500
    },
    {
      "epoch": 7.81,
      "grad_norm": 18.807485580444336,
      "learning_rate": 4.374176548089592e-06,
      "loss": 0.2998,
      "step": 41510
    },
    {
      "epoch": 7.81,
      "grad_norm": 0.11603605002164841,
      "learning_rate": 4.3704121964991534e-06,
      "loss": 0.074,
      "step": 41520
    },
    {
      "epoch": 7.82,
      "grad_norm": 9.478109359741211,
      "learning_rate": 4.366647844908715e-06,
      "loss": 0.0601,
      "step": 41530
    },
    {
      "epoch": 7.82,
      "grad_norm": 3.4649152755737305,
      "learning_rate": 4.362883493318276e-06,
      "loss": 0.2347,
      "step": 41540
    },
    {
      "epoch": 7.82,
      "grad_norm": 2.037090301513672,
      "learning_rate": 4.359119141727838e-06,
      "loss": 0.1916,
      "step": 41550
    },
    {
      "epoch": 7.82,
      "grad_norm": 19.36328125,
      "learning_rate": 4.355354790137399e-06,
      "loss": 0.2158,
      "step": 41560
    },
    {
      "epoch": 7.82,
      "grad_norm": 12.321585655212402,
      "learning_rate": 4.351590438546961e-06,
      "loss": 0.1471,
      "step": 41570
    },
    {
      "epoch": 7.83,
      "grad_norm": 2.3927690982818604,
      "learning_rate": 4.347826086956522e-06,
      "loss": 0.0431,
      "step": 41580
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.0066087860614061356,
      "learning_rate": 4.344061735366084e-06,
      "loss": 0.0995,
      "step": 41590
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.0037947907112538815,
      "learning_rate": 4.340297383775645e-06,
      "loss": 0.2117,
      "step": 41600
    },
    {
      "epoch": 7.83,
      "grad_norm": 0.09212931990623474,
      "learning_rate": 4.336533032185207e-06,
      "loss": 0.0421,
      "step": 41610
    },
    {
      "epoch": 7.83,
      "grad_norm": 10.32288932800293,
      "learning_rate": 4.332768680594768e-06,
      "loss": 0.4011,
      "step": 41620
    },
    {
      "epoch": 7.84,
      "grad_norm": 13.36526870727539,
      "learning_rate": 4.3290043290043295e-06,
      "loss": 0.1238,
      "step": 41630
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.01559041440486908,
      "learning_rate": 4.325239977413891e-06,
      "loss": 0.0839,
      "step": 41640
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.03942855820059776,
      "learning_rate": 4.3214756258234524e-06,
      "loss": 0.0198,
      "step": 41650
    },
    {
      "epoch": 7.84,
      "grad_norm": 1.0315455198287964,
      "learning_rate": 4.317711274233014e-06,
      "loss": 0.181,
      "step": 41660
    },
    {
      "epoch": 7.84,
      "grad_norm": 9.042343139648438,
      "learning_rate": 4.313946922642575e-06,
      "loss": 0.0907,
      "step": 41670
    },
    {
      "epoch": 7.84,
      "grad_norm": 0.005992177873849869,
      "learning_rate": 4.310182571052137e-06,
      "loss": 0.1481,
      "step": 41680
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.006543765310198069,
      "learning_rate": 4.306418219461698e-06,
      "loss": 0.1224,
      "step": 41690
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.2681175172328949,
      "learning_rate": 4.302653867871259e-06,
      "loss": 0.0775,
      "step": 41700
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.011450330726802349,
      "learning_rate": 4.298889516280821e-06,
      "loss": 0.1285,
      "step": 41710
    },
    {
      "epoch": 7.85,
      "grad_norm": 7.963943004608154,
      "learning_rate": 4.295125164690383e-06,
      "loss": 0.2908,
      "step": 41720
    },
    {
      "epoch": 7.85,
      "grad_norm": 0.0033576050773262978,
      "learning_rate": 4.291360813099944e-06,
      "loss": 0.0152,
      "step": 41730
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.007216492667794228,
      "learning_rate": 4.287596461509506e-06,
      "loss": 0.2566,
      "step": 41740
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.01159477885812521,
      "learning_rate": 4.283832109919067e-06,
      "loss": 0.0119,
      "step": 41750
    },
    {
      "epoch": 7.86,
      "grad_norm": 6.214239597320557,
      "learning_rate": 4.2800677583286285e-06,
      "loss": 0.154,
      "step": 41760
    },
    {
      "epoch": 7.86,
      "grad_norm": 7.1128458976745605,
      "learning_rate": 4.27630340673819e-06,
      "loss": 0.1613,
      "step": 41770
    },
    {
      "epoch": 7.86,
      "grad_norm": 0.004408534616231918,
      "learning_rate": 4.272539055147751e-06,
      "loss": 0.1454,
      "step": 41780
    },
    {
      "epoch": 7.87,
      "grad_norm": 4.641051769256592,
      "learning_rate": 4.268774703557312e-06,
      "loss": 0.2247,
      "step": 41790
    },
    {
      "epoch": 7.87,
      "grad_norm": 6.919368267059326,
      "learning_rate": 4.265010351966874e-06,
      "loss": 0.0979,
      "step": 41800
    },
    {
      "epoch": 7.87,
      "grad_norm": 0.01333671621978283,
      "learning_rate": 4.261246000376436e-06,
      "loss": 0.1764,
      "step": 41810
    },
    {
      "epoch": 7.87,
      "grad_norm": 10.831182479858398,
      "learning_rate": 4.257481648785997e-06,
      "loss": 0.0533,
      "step": 41820
    },
    {
      "epoch": 7.87,
      "grad_norm": 0.05763298645615578,
      "learning_rate": 4.253717297195558e-06,
      "loss": 0.2911,
      "step": 41830
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.002890286734327674,
      "learning_rate": 4.249952945605119e-06,
      "loss": 0.3562,
      "step": 41840
    },
    {
      "epoch": 7.88,
      "grad_norm": 10.597702026367188,
      "learning_rate": 4.246188594014682e-06,
      "loss": 0.4088,
      "step": 41850
    },
    {
      "epoch": 7.88,
      "grad_norm": 1.7763631343841553,
      "learning_rate": 4.242424242424243e-06,
      "loss": 0.1458,
      "step": 41860
    },
    {
      "epoch": 7.88,
      "grad_norm": 3.9730441570281982,
      "learning_rate": 4.2386598908338046e-06,
      "loss": 0.1998,
      "step": 41870
    },
    {
      "epoch": 7.88,
      "grad_norm": 0.014345121569931507,
      "learning_rate": 4.234895539243365e-06,
      "loss": 0.1456,
      "step": 41880
    },
    {
      "epoch": 7.88,
      "grad_norm": 4.2904558181762695,
      "learning_rate": 4.2311311876529275e-06,
      "loss": 0.1787,
      "step": 41890
    },
    {
      "epoch": 7.89,
      "grad_norm": 20.931791305541992,
      "learning_rate": 4.227366836062489e-06,
      "loss": 0.2628,
      "step": 41900
    },
    {
      "epoch": 7.89,
      "grad_norm": 5.29111385345459,
      "learning_rate": 4.22360248447205e-06,
      "loss": 0.1079,
      "step": 41910
    },
    {
      "epoch": 7.89,
      "grad_norm": 2.9756925106048584,
      "learning_rate": 4.219838132881611e-06,
      "loss": 0.137,
      "step": 41920
    },
    {
      "epoch": 7.89,
      "grad_norm": 0.9376057386398315,
      "learning_rate": 4.2160737812911725e-06,
      "loss": 0.0968,
      "step": 41930
    },
    {
      "epoch": 7.89,
      "grad_norm": 25.012758255004883,
      "learning_rate": 4.212309429700735e-06,
      "loss": 0.1335,
      "step": 41940
    },
    {
      "epoch": 7.9,
      "grad_norm": 0.1295500099658966,
      "learning_rate": 4.208545078110296e-06,
      "loss": 0.1825,
      "step": 41950
    },
    {
      "epoch": 7.9,
      "grad_norm": 11.511603355407715,
      "learning_rate": 4.204780726519857e-06,
      "loss": 0.2339,
      "step": 41960
    },
    {
      "epoch": 7.9,
      "grad_norm": 5.113443374633789,
      "learning_rate": 4.201016374929418e-06,
      "loss": 0.122,
      "step": 41970
    },
    {
      "epoch": 7.9,
      "grad_norm": 0.009820988401770592,
      "learning_rate": 4.19725202333898e-06,
      "loss": 0.2625,
      "step": 41980
    },
    {
      "epoch": 7.9,
      "grad_norm": 33.33989334106445,
      "learning_rate": 4.193487671748542e-06,
      "loss": 0.2158,
      "step": 41990
    },
    {
      "epoch": 7.91,
      "grad_norm": 1.3190946578979492,
      "learning_rate": 4.1897233201581036e-06,
      "loss": 0.2384,
      "step": 42000
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.6824606657028198,
      "learning_rate": 4.185958968567664e-06,
      "loss": 0.1505,
      "step": 42010
    },
    {
      "epoch": 7.91,
      "grad_norm": 10.540101051330566,
      "learning_rate": 4.182194616977226e-06,
      "loss": 0.0501,
      "step": 42020
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.2917954623699188,
      "learning_rate": 4.178430265386788e-06,
      "loss": 0.0668,
      "step": 42030
    },
    {
      "epoch": 7.91,
      "grad_norm": 28.170421600341797,
      "learning_rate": 4.174665913796349e-06,
      "loss": 0.1846,
      "step": 42040
    },
    {
      "epoch": 7.91,
      "grad_norm": 0.3870393931865692,
      "learning_rate": 4.17090156220591e-06,
      "loss": 0.0365,
      "step": 42050
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.003879326395690441,
      "learning_rate": 4.1671372106154715e-06,
      "loss": 0.1848,
      "step": 42060
    },
    {
      "epoch": 7.92,
      "grad_norm": 1.7813441753387451,
      "learning_rate": 4.163372859025033e-06,
      "loss": 0.2072,
      "step": 42070
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.09556812047958374,
      "learning_rate": 4.159608507434595e-06,
      "loss": 0.1643,
      "step": 42080
    },
    {
      "epoch": 7.92,
      "grad_norm": 70.2798080444336,
      "learning_rate": 4.155844155844157e-06,
      "loss": 0.1369,
      "step": 42090
    },
    {
      "epoch": 7.92,
      "grad_norm": 0.026978734880685806,
      "learning_rate": 4.152079804253717e-06,
      "loss": 0.1361,
      "step": 42100
    },
    {
      "epoch": 7.93,
      "grad_norm": 12.48388957977295,
      "learning_rate": 4.148315452663279e-06,
      "loss": 0.1213,
      "step": 42110
    },
    {
      "epoch": 7.93,
      "grad_norm": 24.48952293395996,
      "learning_rate": 4.14455110107284e-06,
      "loss": 0.1674,
      "step": 42120
    },
    {
      "epoch": 7.93,
      "grad_norm": 32.2702751159668,
      "learning_rate": 4.1407867494824025e-06,
      "loss": 0.1858,
      "step": 42130
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.24107985198497772,
      "learning_rate": 4.137022397891963e-06,
      "loss": 0.1853,
      "step": 42140
    },
    {
      "epoch": 7.93,
      "grad_norm": 0.0030275702010840178,
      "learning_rate": 4.133258046301525e-06,
      "loss": 0.2443,
      "step": 42150
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.012879923917353153,
      "learning_rate": 4.129493694711086e-06,
      "loss": 0.0964,
      "step": 42160
    },
    {
      "epoch": 7.94,
      "grad_norm": 18.813873291015625,
      "learning_rate": 4.125729343120648e-06,
      "loss": 0.2615,
      "step": 42170
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.08030563592910767,
      "learning_rate": 4.121964991530209e-06,
      "loss": 0.06,
      "step": 42180
    },
    {
      "epoch": 7.94,
      "grad_norm": 0.03864583000540733,
      "learning_rate": 4.1182006399397705e-06,
      "loss": 0.1565,
      "step": 42190
    },
    {
      "epoch": 7.94,
      "grad_norm": 10.535635948181152,
      "learning_rate": 4.114436288349332e-06,
      "loss": 0.0784,
      "step": 42200
    },
    {
      "epoch": 7.94,
      "grad_norm": 26.722627639770508,
      "learning_rate": 4.110671936758893e-06,
      "loss": 0.2365,
      "step": 42210
    },
    {
      "epoch": 7.95,
      "grad_norm": 0.13719183206558228,
      "learning_rate": 4.106907585168456e-06,
      "loss": 0.413,
      "step": 42220
    },
    {
      "epoch": 7.95,
      "grad_norm": 16.92498207092285,
      "learning_rate": 4.103143233578016e-06,
      "loss": 0.1212,
      "step": 42230
    },
    {
      "epoch": 7.95,
      "grad_norm": 9.468924522399902,
      "learning_rate": 4.099378881987578e-06,
      "loss": 0.3614,
      "step": 42240
    },
    {
      "epoch": 7.95,
      "grad_norm": 13.808631896972656,
      "learning_rate": 4.095614530397139e-06,
      "loss": 0.2332,
      "step": 42250
    },
    {
      "epoch": 7.95,
      "grad_norm": 7.491482257843018,
      "learning_rate": 4.091850178806701e-06,
      "loss": 0.1907,
      "step": 42260
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.019621191546320915,
      "learning_rate": 4.088085827216262e-06,
      "loss": 0.1038,
      "step": 42270
    },
    {
      "epoch": 7.96,
      "grad_norm": 37.89130783081055,
      "learning_rate": 4.084321475625824e-06,
      "loss": 0.151,
      "step": 42280
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.12550175189971924,
      "learning_rate": 4.080557124035385e-06,
      "loss": 0.2375,
      "step": 42290
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.10164400935173035,
      "learning_rate": 4.0767927724449465e-06,
      "loss": 0.1122,
      "step": 42300
    },
    {
      "epoch": 7.96,
      "grad_norm": 0.027707286179065704,
      "learning_rate": 4.073028420854508e-06,
      "loss": 0.1297,
      "step": 42310
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.4114322066307068,
      "learning_rate": 4.0692640692640695e-06,
      "loss": 0.2077,
      "step": 42320
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.01688412018120289,
      "learning_rate": 4.065499717673631e-06,
      "loss": 0.3069,
      "step": 42330
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.5761536955833435,
      "learning_rate": 4.061735366083192e-06,
      "loss": 0.1042,
      "step": 42340
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.17285460233688354,
      "learning_rate": 4.057971014492754e-06,
      "loss": 0.1219,
      "step": 42350
    },
    {
      "epoch": 7.97,
      "grad_norm": 22.63449478149414,
      "learning_rate": 4.054206662902315e-06,
      "loss": 0.2138,
      "step": 42360
    },
    {
      "epoch": 7.97,
      "grad_norm": 0.04067244008183479,
      "learning_rate": 4.050442311311877e-06,
      "loss": 0.3191,
      "step": 42370
    },
    {
      "epoch": 7.98,
      "grad_norm": 1.4968725442886353,
      "learning_rate": 4.046677959721438e-06,
      "loss": 0.1681,
      "step": 42380
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.010270068421959877,
      "learning_rate": 4.042913608131e-06,
      "loss": 0.2307,
      "step": 42390
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.0842098817229271,
      "learning_rate": 4.039149256540561e-06,
      "loss": 0.2105,
      "step": 42400
    },
    {
      "epoch": 7.98,
      "grad_norm": 0.25274622440338135,
      "learning_rate": 4.035384904950123e-06,
      "loss": 0.2348,
      "step": 42410
    },
    {
      "epoch": 7.98,
      "grad_norm": 18.5978946685791,
      "learning_rate": 4.031620553359684e-06,
      "loss": 0.0106,
      "step": 42420
    },
    {
      "epoch": 7.99,
      "grad_norm": 17.57866668701172,
      "learning_rate": 4.0278562017692455e-06,
      "loss": 0.1669,
      "step": 42430
    },
    {
      "epoch": 7.99,
      "grad_norm": 14.896883010864258,
      "learning_rate": 4.024091850178807e-06,
      "loss": 0.2011,
      "step": 42440
    },
    {
      "epoch": 7.99,
      "grad_norm": 10.013394355773926,
      "learning_rate": 4.0203274985883684e-06,
      "loss": 0.1008,
      "step": 42450
    },
    {
      "epoch": 7.99,
      "grad_norm": 13.902844429016113,
      "learning_rate": 4.01656314699793e-06,
      "loss": 0.2072,
      "step": 42460
    },
    {
      "epoch": 7.99,
      "grad_norm": 7.964358806610107,
      "learning_rate": 4.012798795407491e-06,
      "loss": 0.1994,
      "step": 42470
    },
    {
      "epoch": 8.0,
      "grad_norm": 47.48723220825195,
      "learning_rate": 4.009034443817053e-06,
      "loss": 0.3791,
      "step": 42480
    },
    {
      "epoch": 8.0,
      "grad_norm": 11.336153984069824,
      "learning_rate": 4.005270092226614e-06,
      "loss": 0.3419,
      "step": 42490
    },
    {
      "epoch": 8.0,
      "grad_norm": 48.610130310058594,
      "learning_rate": 4.001505740636176e-06,
      "loss": 0.1669,
      "step": 42500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.99,
      "eval_loss": 0.04177061468362808,
      "eval_runtime": 27.2966,
      "eval_samples_per_second": 274.76,
      "eval_steps_per_second": 34.363,
      "step": 42504
    },
    {
      "epoch": 8.0,
      "grad_norm": 10.000838279724121,
      "learning_rate": 3.997741389045737e-06,
      "loss": 0.2257,
      "step": 42510
    },
    {
      "epoch": 8.0,
      "grad_norm": 34.99166488647461,
      "learning_rate": 3.993977037455299e-06,
      "loss": 0.1457,
      "step": 42520
    },
    {
      "epoch": 8.0,
      "grad_norm": 4.323693752288818,
      "learning_rate": 3.99021268586486e-06,
      "loss": 0.3136,
      "step": 42530
    },
    {
      "epoch": 8.01,
      "grad_norm": 3.930790424346924,
      "learning_rate": 3.986448334274422e-06,
      "loss": 0.2414,
      "step": 42540
    },
    {
      "epoch": 8.01,
      "grad_norm": 10.141860961914062,
      "learning_rate": 3.982683982683983e-06,
      "loss": 0.1779,
      "step": 42550
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.00868989434093237,
      "learning_rate": 3.9789196310935445e-06,
      "loss": 0.0817,
      "step": 42560
    },
    {
      "epoch": 8.01,
      "grad_norm": 0.012134138494729996,
      "learning_rate": 3.975155279503106e-06,
      "loss": 0.1732,
      "step": 42570
    },
    {
      "epoch": 8.01,
      "grad_norm": 12.137693405151367,
      "learning_rate": 3.9713909279126674e-06,
      "loss": 0.1525,
      "step": 42580
    },
    {
      "epoch": 8.02,
      "grad_norm": 0.2462858259677887,
      "learning_rate": 3.967626576322229e-06,
      "loss": 0.2085,
      "step": 42590
    },
    {
      "epoch": 8.02,
      "grad_norm": 7.281343936920166,
      "learning_rate": 3.96386222473179e-06,
      "loss": 0.2421,
      "step": 42600
    },
    {
      "epoch": 8.02,
      "grad_norm": 6.51716947555542,
      "learning_rate": 3.960097873141352e-06,
      "loss": 0.3202,
      "step": 42610
    },
    {
      "epoch": 8.02,
      "grad_norm": 7.488397121429443,
      "learning_rate": 3.956333521550913e-06,
      "loss": 0.1214,
      "step": 42620
    },
    {
      "epoch": 8.02,
      "grad_norm": 0.010680852457880974,
      "learning_rate": 3.952569169960475e-06,
      "loss": 0.0021,
      "step": 42630
    },
    {
      "epoch": 8.03,
      "grad_norm": 2.9098191261291504,
      "learning_rate": 3.948804818370036e-06,
      "loss": 0.3818,
      "step": 42640
    },
    {
      "epoch": 8.03,
      "grad_norm": 11.461164474487305,
      "learning_rate": 3.945040466779598e-06,
      "loss": 0.306,
      "step": 42650
    },
    {
      "epoch": 8.03,
      "grad_norm": 9.904094696044922,
      "learning_rate": 3.941276115189159e-06,
      "loss": 0.384,
      "step": 42660
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.020291132852435112,
      "learning_rate": 3.9375117635987206e-06,
      "loss": 0.1546,
      "step": 42670
    },
    {
      "epoch": 8.03,
      "grad_norm": 0.029828594997525215,
      "learning_rate": 3.933747412008282e-06,
      "loss": 0.3926,
      "step": 42680
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.18995775282382965,
      "learning_rate": 3.9299830604178435e-06,
      "loss": 0.1347,
      "step": 42690
    },
    {
      "epoch": 8.04,
      "grad_norm": 0.021530447527766228,
      "learning_rate": 3.926218708827404e-06,
      "loss": 0.1274,
      "step": 42700
    },
    {
      "epoch": 8.04,
      "grad_norm": 5.946957588195801,
      "learning_rate": 3.922454357236966e-06,
      "loss": 0.111,
      "step": 42710
    },
    {
      "epoch": 8.04,
      "grad_norm": 13.218122482299805,
      "learning_rate": 3.918690005646528e-06,
      "loss": 0.1915,
      "step": 42720
    },
    {
      "epoch": 8.04,
      "grad_norm": 15.745772361755371,
      "learning_rate": 3.914925654056089e-06,
      "loss": 0.2255,
      "step": 42730
    },
    {
      "epoch": 8.04,
      "grad_norm": 24.266033172607422,
      "learning_rate": 3.911161302465651e-06,
      "loss": 0.1078,
      "step": 42740
    },
    {
      "epoch": 8.05,
      "grad_norm": 27.176345825195312,
      "learning_rate": 3.907396950875212e-06,
      "loss": 0.2987,
      "step": 42750
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.07400763779878616,
      "learning_rate": 3.903632599284774e-06,
      "loss": 0.1243,
      "step": 42760
    },
    {
      "epoch": 8.05,
      "grad_norm": 25.495607376098633,
      "learning_rate": 3.899868247694335e-06,
      "loss": 0.0908,
      "step": 42770
    },
    {
      "epoch": 8.05,
      "grad_norm": 0.36077016592025757,
      "learning_rate": 3.896103896103897e-06,
      "loss": 0.0975,
      "step": 42780
    },
    {
      "epoch": 8.05,
      "grad_norm": 1.1561613082885742,
      "learning_rate": 3.892339544513457e-06,
      "loss": 0.1583,
      "step": 42790
    },
    {
      "epoch": 8.06,
      "grad_norm": 9.118714332580566,
      "learning_rate": 3.8885751929230196e-06,
      "loss": 0.0893,
      "step": 42800
    },
    {
      "epoch": 8.06,
      "grad_norm": 0.004794972948729992,
      "learning_rate": 3.884810841332581e-06,
      "loss": 0.0906,
      "step": 42810
    },
    {
      "epoch": 8.06,
      "grad_norm": 0.018586119636893272,
      "learning_rate": 3.8810464897421425e-06,
      "loss": 0.3521,
      "step": 42820
    },
    {
      "epoch": 8.06,
      "grad_norm": 2.728970766067505,
      "learning_rate": 3.877282138151704e-06,
      "loss": 0.1733,
      "step": 42830
    },
    {
      "epoch": 8.06,
      "grad_norm": 1.3522286415100098,
      "learning_rate": 3.8735177865612646e-06,
      "loss": 0.0996,
      "step": 42840
    },
    {
      "epoch": 8.07,
      "grad_norm": 11.444347381591797,
      "learning_rate": 3.869753434970827e-06,
      "loss": 0.2641,
      "step": 42850
    },
    {
      "epoch": 8.07,
      "grad_norm": 4.517781734466553,
      "learning_rate": 3.865989083380388e-06,
      "loss": 0.1393,
      "step": 42860
    },
    {
      "epoch": 8.07,
      "grad_norm": 18.18436050415039,
      "learning_rate": 3.86222473178995e-06,
      "loss": 0.3756,
      "step": 42870
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.6042503714561462,
      "learning_rate": 3.85846038019951e-06,
      "loss": 0.199,
      "step": 42880
    },
    {
      "epoch": 8.07,
      "grad_norm": 12.210646629333496,
      "learning_rate": 3.854696028609073e-06,
      "loss": 0.1736,
      "step": 42890
    },
    {
      "epoch": 8.07,
      "grad_norm": 0.010609155520796776,
      "learning_rate": 3.850931677018634e-06,
      "loss": 0.1265,
      "step": 42900
    },
    {
      "epoch": 8.08,
      "grad_norm": 19.26944351196289,
      "learning_rate": 3.847167325428196e-06,
      "loss": 0.1777,
      "step": 42910
    },
    {
      "epoch": 8.08,
      "grad_norm": 7.758819580078125,
      "learning_rate": 3.843402973837756e-06,
      "loss": 0.146,
      "step": 42920
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.0070473612286150455,
      "learning_rate": 3.839638622247318e-06,
      "loss": 0.1384,
      "step": 42930
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.00868208147585392,
      "learning_rate": 3.83587427065688e-06,
      "loss": 0.1712,
      "step": 42940
    },
    {
      "epoch": 8.08,
      "grad_norm": 0.010056274011731148,
      "learning_rate": 3.8321099190664415e-06,
      "loss": 0.3088,
      "step": 42950
    },
    {
      "epoch": 8.09,
      "grad_norm": 0.10152862221002579,
      "learning_rate": 3.828345567476003e-06,
      "loss": 0.0231,
      "step": 42960
    },
    {
      "epoch": 8.09,
      "grad_norm": 3.6894495487213135,
      "learning_rate": 3.8245812158855635e-06,
      "loss": 0.0879,
      "step": 42970
    },
    {
      "epoch": 8.09,
      "grad_norm": 9.06627368927002,
      "learning_rate": 3.820816864295126e-06,
      "loss": 0.1828,
      "step": 42980
    },
    {
      "epoch": 8.09,
      "grad_norm": 12.548944473266602,
      "learning_rate": 3.817052512704687e-06,
      "loss": 0.1051,
      "step": 42990
    },
    {
      "epoch": 8.09,
      "grad_norm": 3.3518998622894287,
      "learning_rate": 3.8132881611142488e-06,
      "loss": 0.2074,
      "step": 43000
    },
    {
      "epoch": 8.1,
      "grad_norm": 1.467064619064331,
      "learning_rate": 3.80952380952381e-06,
      "loss": 0.1982,
      "step": 43010
    },
    {
      "epoch": 8.1,
      "grad_norm": 2.348451614379883,
      "learning_rate": 3.8057594579333713e-06,
      "loss": 0.1433,
      "step": 43020
    },
    {
      "epoch": 8.1,
      "grad_norm": 0.4999294579029083,
      "learning_rate": 3.8019951063429327e-06,
      "loss": 0.0676,
      "step": 43030
    },
    {
      "epoch": 8.1,
      "grad_norm": 20.775815963745117,
      "learning_rate": 3.798230754752494e-06,
      "loss": 0.2197,
      "step": 43040
    },
    {
      "epoch": 8.1,
      "grad_norm": 23.86593246459961,
      "learning_rate": 3.7944664031620552e-06,
      "loss": 0.2657,
      "step": 43050
    },
    {
      "epoch": 8.1,
      "grad_norm": 27.783098220825195,
      "learning_rate": 3.790702051571617e-06,
      "loss": 0.2326,
      "step": 43060
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.4397278130054474,
      "learning_rate": 3.7869376999811786e-06,
      "loss": 0.0435,
      "step": 43070
    },
    {
      "epoch": 8.11,
      "grad_norm": 16.699403762817383,
      "learning_rate": 3.78317334839074e-06,
      "loss": 0.4028,
      "step": 43080
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.026661699637770653,
      "learning_rate": 3.7794089968003015e-06,
      "loss": 0.1988,
      "step": 43090
    },
    {
      "epoch": 8.11,
      "grad_norm": 2.2638251781463623,
      "learning_rate": 3.7756446452098625e-06,
      "loss": 0.1748,
      "step": 43100
    },
    {
      "epoch": 8.11,
      "grad_norm": 0.34065449237823486,
      "learning_rate": 3.7718802936194244e-06,
      "loss": 0.1507,
      "step": 43110
    },
    {
      "epoch": 8.12,
      "grad_norm": 7.436203956604004,
      "learning_rate": 3.768115942028986e-06,
      "loss": 0.1716,
      "step": 43120
    },
    {
      "epoch": 8.12,
      "grad_norm": 26.779584884643555,
      "learning_rate": 3.7643515904385473e-06,
      "loss": 0.1111,
      "step": 43130
    },
    {
      "epoch": 8.12,
      "grad_norm": 0.09941428154706955,
      "learning_rate": 3.7605872388481084e-06,
      "loss": 0.2019,
      "step": 43140
    },
    {
      "epoch": 8.12,
      "grad_norm": 18.337732315063477,
      "learning_rate": 3.7568228872576703e-06,
      "loss": 0.2386,
      "step": 43150
    },
    {
      "epoch": 8.12,
      "grad_norm": 4.331237316131592,
      "learning_rate": 3.7530585356672317e-06,
      "loss": 0.1599,
      "step": 43160
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.1650846302509308,
      "learning_rate": 3.749294184076793e-06,
      "loss": 0.0413,
      "step": 43170
    },
    {
      "epoch": 8.13,
      "grad_norm": 3.763246774673462,
      "learning_rate": 3.7455298324863542e-06,
      "loss": 0.2378,
      "step": 43180
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.011877230368554592,
      "learning_rate": 3.7417654808959157e-06,
      "loss": 0.2499,
      "step": 43190
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.008318454027175903,
      "learning_rate": 3.7380011293054776e-06,
      "loss": 0.0882,
      "step": 43200
    },
    {
      "epoch": 8.13,
      "grad_norm": 0.0260122362524271,
      "learning_rate": 3.734236777715039e-06,
      "loss": 0.3061,
      "step": 43210
    },
    {
      "epoch": 8.13,
      "grad_norm": 23.79955291748047,
      "learning_rate": 3.7304724261246005e-06,
      "loss": 0.1601,
      "step": 43220
    },
    {
      "epoch": 8.14,
      "grad_norm": 14.120388984680176,
      "learning_rate": 3.7267080745341615e-06,
      "loss": 0.1463,
      "step": 43230
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.27815568447113037,
      "learning_rate": 3.722943722943723e-06,
      "loss": 0.0604,
      "step": 43240
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.004799243528395891,
      "learning_rate": 3.719179371353285e-06,
      "loss": 0.1291,
      "step": 43250
    },
    {
      "epoch": 8.14,
      "grad_norm": 12.247443199157715,
      "learning_rate": 3.7154150197628463e-06,
      "loss": 0.424,
      "step": 43260
    },
    {
      "epoch": 8.14,
      "grad_norm": 0.14927129447460175,
      "learning_rate": 3.7116506681724074e-06,
      "loss": 0.2076,
      "step": 43270
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.02195632830262184,
      "learning_rate": 3.707886316581969e-06,
      "loss": 0.1282,
      "step": 43280
    },
    {
      "epoch": 8.15,
      "grad_norm": 0.05117698386311531,
      "learning_rate": 3.7041219649915307e-06,
      "loss": 0.2206,
      "step": 43290
    },
    {
      "epoch": 8.15,
      "grad_norm": 7.994806289672852,
      "learning_rate": 3.700357613401092e-06,
      "loss": 0.3987,
      "step": 43300
    },
    {
      "epoch": 8.15,
      "grad_norm": 45.457313537597656,
      "learning_rate": 3.6965932618106532e-06,
      "loss": 0.0826,
      "step": 43310
    },
    {
      "epoch": 8.15,
      "grad_norm": 15.824580192565918,
      "learning_rate": 3.6928289102202147e-06,
      "loss": 0.1186,
      "step": 43320
    },
    {
      "epoch": 8.16,
      "grad_norm": 10.439388275146484,
      "learning_rate": 3.689064558629776e-06,
      "loss": 0.2521,
      "step": 43330
    },
    {
      "epoch": 8.16,
      "grad_norm": 7.78478479385376,
      "learning_rate": 3.685300207039338e-06,
      "loss": 0.2203,
      "step": 43340
    },
    {
      "epoch": 8.16,
      "grad_norm": 2.6825344562530518,
      "learning_rate": 3.6815358554488995e-06,
      "loss": 0.0055,
      "step": 43350
    },
    {
      "epoch": 8.16,
      "grad_norm": 25.168540954589844,
      "learning_rate": 3.6777715038584605e-06,
      "loss": 0.155,
      "step": 43360
    },
    {
      "epoch": 8.16,
      "grad_norm": 1.3383409976959229,
      "learning_rate": 3.674007152268022e-06,
      "loss": 0.0898,
      "step": 43370
    },
    {
      "epoch": 8.16,
      "grad_norm": 1.126628041267395,
      "learning_rate": 3.670242800677584e-06,
      "loss": 0.0585,
      "step": 43380
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.36692163348197937,
      "learning_rate": 3.6664784490871453e-06,
      "loss": 0.1733,
      "step": 43390
    },
    {
      "epoch": 8.17,
      "grad_norm": 6.034840106964111,
      "learning_rate": 3.6627140974967064e-06,
      "loss": 0.2417,
      "step": 43400
    },
    {
      "epoch": 8.17,
      "grad_norm": 8.345094680786133,
      "learning_rate": 3.658949745906268e-06,
      "loss": 0.3707,
      "step": 43410
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.010809577070176601,
      "learning_rate": 3.6551853943158293e-06,
      "loss": 0.1603,
      "step": 43420
    },
    {
      "epoch": 8.17,
      "grad_norm": 0.023334119468927383,
      "learning_rate": 3.651421042725391e-06,
      "loss": 0.184,
      "step": 43430
    },
    {
      "epoch": 8.18,
      "grad_norm": 28.007986068725586,
      "learning_rate": 3.6476566911349526e-06,
      "loss": 0.3185,
      "step": 43440
    },
    {
      "epoch": 8.18,
      "grad_norm": 16.53616714477539,
      "learning_rate": 3.6438923395445137e-06,
      "loss": 0.1116,
      "step": 43450
    },
    {
      "epoch": 8.18,
      "grad_norm": 2.9128172397613525,
      "learning_rate": 3.640127987954075e-06,
      "loss": 0.1081,
      "step": 43460
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.013568619266152382,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.2147,
      "step": 43470
    },
    {
      "epoch": 8.18,
      "grad_norm": 0.01935197412967682,
      "learning_rate": 3.6325992847731985e-06,
      "loss": 0.0377,
      "step": 43480
    },
    {
      "epoch": 8.19,
      "grad_norm": 18.504405975341797,
      "learning_rate": 3.6288349331827595e-06,
      "loss": 0.3072,
      "step": 43490
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.715424656867981,
      "learning_rate": 3.625070581592321e-06,
      "loss": 0.0823,
      "step": 43500
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.004430097993463278,
      "learning_rate": 3.6213062300018824e-06,
      "loss": 0.0962,
      "step": 43510
    },
    {
      "epoch": 8.19,
      "grad_norm": 6.181677341461182,
      "learning_rate": 3.6175418784114443e-06,
      "loss": 0.0979,
      "step": 43520
    },
    {
      "epoch": 8.19,
      "grad_norm": 9.798259735107422,
      "learning_rate": 3.613777526821005e-06,
      "loss": 0.3633,
      "step": 43530
    },
    {
      "epoch": 8.19,
      "grad_norm": 0.00450526550412178,
      "learning_rate": 3.610013175230567e-06,
      "loss": 0.146,
      "step": 43540
    },
    {
      "epoch": 8.2,
      "grad_norm": 9.980013847351074,
      "learning_rate": 3.6062488236401283e-06,
      "loss": 0.0956,
      "step": 43550
    },
    {
      "epoch": 8.2,
      "grad_norm": 23.384153366088867,
      "learning_rate": 3.6024844720496897e-06,
      "loss": 0.307,
      "step": 43560
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.008191576227545738,
      "learning_rate": 3.5987201204592516e-06,
      "loss": 0.1429,
      "step": 43570
    },
    {
      "epoch": 8.2,
      "grad_norm": 19.599523544311523,
      "learning_rate": 3.5949557688688127e-06,
      "loss": 0.0657,
      "step": 43580
    },
    {
      "epoch": 8.2,
      "grad_norm": 0.03506388142704964,
      "learning_rate": 3.591191417278374e-06,
      "loss": 0.1111,
      "step": 43590
    },
    {
      "epoch": 8.21,
      "grad_norm": 13.614469528198242,
      "learning_rate": 3.5874270656879356e-06,
      "loss": 0.2665,
      "step": 43600
    },
    {
      "epoch": 8.21,
      "grad_norm": 0.14039622247219086,
      "learning_rate": 3.583662714097497e-06,
      "loss": 0.1293,
      "step": 43610
    },
    {
      "epoch": 8.21,
      "grad_norm": 14.7056245803833,
      "learning_rate": 3.579898362507058e-06,
      "loss": 0.2022,
      "step": 43620
    },
    {
      "epoch": 8.21,
      "grad_norm": 0.032121799886226654,
      "learning_rate": 3.57613401091662e-06,
      "loss": 0.4404,
      "step": 43630
    },
    {
      "epoch": 8.21,
      "grad_norm": 20.435836791992188,
      "learning_rate": 3.5723696593261814e-06,
      "loss": 0.2724,
      "step": 43640
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.007248232141137123,
      "learning_rate": 3.568605307735743e-06,
      "loss": 0.2155,
      "step": 43650
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.01702449657022953,
      "learning_rate": 3.564840956145304e-06,
      "loss": 0.1873,
      "step": 43660
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.5429418683052063,
      "learning_rate": 3.561076604554866e-06,
      "loss": 0.0335,
      "step": 43670
    },
    {
      "epoch": 8.22,
      "grad_norm": 0.784830629825592,
      "learning_rate": 3.5573122529644273e-06,
      "loss": 0.1424,
      "step": 43680
    },
    {
      "epoch": 8.22,
      "grad_norm": 11.71828842163086,
      "learning_rate": 3.5535479013739887e-06,
      "loss": 0.2129,
      "step": 43690
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.035924457013607025,
      "learning_rate": 3.54978354978355e-06,
      "loss": 0.3795,
      "step": 43700
    },
    {
      "epoch": 8.23,
      "grad_norm": 4.6747002601623535,
      "learning_rate": 3.5460191981931112e-06,
      "loss": 0.0856,
      "step": 43710
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.004435141570866108,
      "learning_rate": 3.542254846602673e-06,
      "loss": 0.0961,
      "step": 43720
    },
    {
      "epoch": 8.23,
      "grad_norm": 19.135223388671875,
      "learning_rate": 3.5384904950122346e-06,
      "loss": 0.2805,
      "step": 43730
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.33678820729255676,
      "learning_rate": 3.534726143421796e-06,
      "loss": 0.0731,
      "step": 43740
    },
    {
      "epoch": 8.23,
      "grad_norm": 0.48058488965034485,
      "learning_rate": 3.530961791831357e-06,
      "loss": 0.1596,
      "step": 43750
    },
    {
      "epoch": 8.24,
      "grad_norm": 30.067209243774414,
      "learning_rate": 3.5271974402409185e-06,
      "loss": 0.1827,
      "step": 43760
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.0827186331152916,
      "learning_rate": 3.5234330886504804e-06,
      "loss": 0.2381,
      "step": 43770
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.0070457360707223415,
      "learning_rate": 3.519668737060042e-06,
      "loss": 0.1289,
      "step": 43780
    },
    {
      "epoch": 8.24,
      "grad_norm": 32.270381927490234,
      "learning_rate": 3.515904385469603e-06,
      "loss": 0.1803,
      "step": 43790
    },
    {
      "epoch": 8.24,
      "grad_norm": 0.022102050483226776,
      "learning_rate": 3.5121400338791644e-06,
      "loss": 0.1773,
      "step": 43800
    },
    {
      "epoch": 8.25,
      "grad_norm": 2.47894549369812,
      "learning_rate": 3.5083756822887262e-06,
      "loss": 0.0773,
      "step": 43810
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.020889459177851677,
      "learning_rate": 3.5046113306982877e-06,
      "loss": 0.1227,
      "step": 43820
    },
    {
      "epoch": 8.25,
      "grad_norm": 3.402500867843628,
      "learning_rate": 3.500846979107849e-06,
      "loss": 0.2006,
      "step": 43830
    },
    {
      "epoch": 8.25,
      "grad_norm": 0.0040222215466201305,
      "learning_rate": 3.49708262751741e-06,
      "loss": 0.0229,
      "step": 43840
    },
    {
      "epoch": 8.25,
      "grad_norm": 3.594181537628174,
      "learning_rate": 3.4933182759269717e-06,
      "loss": 0.1281,
      "step": 43850
    },
    {
      "epoch": 8.26,
      "grad_norm": 1.8714030981063843,
      "learning_rate": 3.4895539243365336e-06,
      "loss": 0.0969,
      "step": 43860
    },
    {
      "epoch": 8.26,
      "grad_norm": 32.60432052612305,
      "learning_rate": 3.485789572746095e-06,
      "loss": 0.3657,
      "step": 43870
    },
    {
      "epoch": 8.26,
      "grad_norm": 0.08047598600387573,
      "learning_rate": 3.482025221155656e-06,
      "loss": 0.2175,
      "step": 43880
    },
    {
      "epoch": 8.26,
      "grad_norm": 2.551795721054077,
      "learning_rate": 3.4782608695652175e-06,
      "loss": 0.3632,
      "step": 43890
    },
    {
      "epoch": 8.26,
      "grad_norm": 18.31554412841797,
      "learning_rate": 3.474496517974779e-06,
      "loss": 0.2306,
      "step": 43900
    },
    {
      "epoch": 8.26,
      "grad_norm": 0.042112212628126144,
      "learning_rate": 3.470732166384341e-06,
      "loss": 0.1219,
      "step": 43910
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.3941866457462311,
      "learning_rate": 3.466967814793902e-06,
      "loss": 0.0199,
      "step": 43920
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.023738371208310127,
      "learning_rate": 3.4632034632034634e-06,
      "loss": 0.2052,
      "step": 43930
    },
    {
      "epoch": 8.27,
      "grad_norm": 14.010457038879395,
      "learning_rate": 3.459439111613025e-06,
      "loss": 0.185,
      "step": 43940
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.030247213318943977,
      "learning_rate": 3.4556747600225867e-06,
      "loss": 0.305,
      "step": 43950
    },
    {
      "epoch": 8.27,
      "grad_norm": 0.008810865692794323,
      "learning_rate": 3.451910408432148e-06,
      "loss": 0.075,
      "step": 43960
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.004307066556066275,
      "learning_rate": 3.448146056841709e-06,
      "loss": 0.2084,
      "step": 43970
    },
    {
      "epoch": 8.28,
      "grad_norm": 13.912513732910156,
      "learning_rate": 3.4443817052512707e-06,
      "loss": 0.1705,
      "step": 43980
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.049022890627384186,
      "learning_rate": 3.440617353660832e-06,
      "loss": 0.1843,
      "step": 43990
    },
    {
      "epoch": 8.28,
      "grad_norm": 0.03220395743846893,
      "learning_rate": 3.436853002070394e-06,
      "loss": 0.1263,
      "step": 44000
    },
    {
      "epoch": 8.28,
      "grad_norm": 16.93511962890625,
      "learning_rate": 3.433088650479955e-06,
      "loss": 0.1332,
      "step": 44010
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.023529820144176483,
      "learning_rate": 3.4293242988895165e-06,
      "loss": 0.1281,
      "step": 44020
    },
    {
      "epoch": 8.29,
      "grad_norm": 5.805478572845459,
      "learning_rate": 3.425559947299078e-06,
      "loss": 0.2241,
      "step": 44030
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.004796687513589859,
      "learning_rate": 3.4217955957086394e-06,
      "loss": 0.0706,
      "step": 44040
    },
    {
      "epoch": 8.29,
      "grad_norm": 24.967557907104492,
      "learning_rate": 3.4180312441182005e-06,
      "loss": 0.1528,
      "step": 44050
    },
    {
      "epoch": 8.29,
      "grad_norm": 1.9152361154556274,
      "learning_rate": 3.4142668925277623e-06,
      "loss": 0.213,
      "step": 44060
    },
    {
      "epoch": 8.29,
      "grad_norm": 0.015526873990893364,
      "learning_rate": 3.410502540937324e-06,
      "loss": 0.1662,
      "step": 44070
    },
    {
      "epoch": 8.3,
      "grad_norm": 11.994935035705566,
      "learning_rate": 3.4067381893468853e-06,
      "loss": 0.0835,
      "step": 44080
    },
    {
      "epoch": 8.3,
      "grad_norm": 47.373329162597656,
      "learning_rate": 3.402973837756447e-06,
      "loss": 0.1609,
      "step": 44090
    },
    {
      "epoch": 8.3,
      "grad_norm": 6.6731157302856445,
      "learning_rate": 3.399209486166008e-06,
      "loss": 0.135,
      "step": 44100
    },
    {
      "epoch": 8.3,
      "grad_norm": 0.010556275025010109,
      "learning_rate": 3.3954451345755696e-06,
      "loss": 0.3108,
      "step": 44110
    },
    {
      "epoch": 8.3,
      "grad_norm": 7.970505237579346,
      "learning_rate": 3.391680782985131e-06,
      "loss": 0.3037,
      "step": 44120
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.1269792914390564,
      "learning_rate": 3.3879164313946926e-06,
      "loss": 0.1661,
      "step": 44130
    },
    {
      "epoch": 8.31,
      "grad_norm": 3.5626070499420166,
      "learning_rate": 3.3841520798042536e-06,
      "loss": 0.1705,
      "step": 44140
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.03423117846250534,
      "learning_rate": 3.3803877282138155e-06,
      "loss": 0.2632,
      "step": 44150
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.00549786863848567,
      "learning_rate": 3.376623376623377e-06,
      "loss": 0.3962,
      "step": 44160
    },
    {
      "epoch": 8.31,
      "grad_norm": 0.060872212052345276,
      "learning_rate": 3.3728590250329384e-06,
      "loss": 0.0354,
      "step": 44170
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.3541152775287628,
      "learning_rate": 3.3690946734425e-06,
      "loss": 0.2526,
      "step": 44180
    },
    {
      "epoch": 8.32,
      "grad_norm": 3.1486310958862305,
      "learning_rate": 3.365330321852061e-06,
      "loss": 0.3104,
      "step": 44190
    },
    {
      "epoch": 8.32,
      "grad_norm": 17.281293869018555,
      "learning_rate": 3.361565970261623e-06,
      "loss": 0.1232,
      "step": 44200
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.003846403444185853,
      "learning_rate": 3.3578016186711843e-06,
      "loss": 0.1167,
      "step": 44210
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.024919727817177773,
      "learning_rate": 3.3540372670807457e-06,
      "loss": 0.1287,
      "step": 44220
    },
    {
      "epoch": 8.32,
      "grad_norm": 0.15599241852760315,
      "learning_rate": 3.3502729154903068e-06,
      "loss": 0.1736,
      "step": 44230
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.00442164670675993,
      "learning_rate": 3.3465085638998686e-06,
      "loss": 0.0902,
      "step": 44240
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.004450839012861252,
      "learning_rate": 3.34274421230943e-06,
      "loss": 0.0821,
      "step": 44250
    },
    {
      "epoch": 8.33,
      "grad_norm": 1.4127049446105957,
      "learning_rate": 3.3389798607189916e-06,
      "loss": 0.0919,
      "step": 44260
    },
    {
      "epoch": 8.33,
      "grad_norm": 2.6160573959350586,
      "learning_rate": 3.3352155091285526e-06,
      "loss": 0.1823,
      "step": 44270
    },
    {
      "epoch": 8.33,
      "grad_norm": 0.1458774209022522,
      "learning_rate": 3.331451157538114e-06,
      "loss": 0.1894,
      "step": 44280
    },
    {
      "epoch": 8.34,
      "grad_norm": 19.849157333374023,
      "learning_rate": 3.327686805947676e-06,
      "loss": 0.2297,
      "step": 44290
    },
    {
      "epoch": 8.34,
      "grad_norm": 9.788932800292969,
      "learning_rate": 3.3239224543572374e-06,
      "loss": 0.273,
      "step": 44300
    },
    {
      "epoch": 8.34,
      "grad_norm": 0.6595255136489868,
      "learning_rate": 3.320158102766799e-06,
      "loss": 0.2011,
      "step": 44310
    },
    {
      "epoch": 8.34,
      "grad_norm": 1.6199009418487549,
      "learning_rate": 3.31639375117636e-06,
      "loss": 0.2349,
      "step": 44320
    },
    {
      "epoch": 8.34,
      "grad_norm": 26.475521087646484,
      "learning_rate": 3.3126293995859214e-06,
      "loss": 0.0925,
      "step": 44330
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.03217468038201332,
      "learning_rate": 3.3088650479954832e-06,
      "loss": 0.3266,
      "step": 44340
    },
    {
      "epoch": 8.35,
      "grad_norm": 0.10134441405534744,
      "learning_rate": 3.3051006964050447e-06,
      "loss": 0.199,
      "step": 44350
    },
    {
      "epoch": 8.35,
      "grad_norm": 3.4470391273498535,
      "learning_rate": 3.3013363448146057e-06,
      "loss": 0.205,
      "step": 44360
    },
    {
      "epoch": 8.35,
      "grad_norm": 7.111494541168213,
      "learning_rate": 3.297571993224167e-06,
      "loss": 0.1529,
      "step": 44370
    },
    {
      "epoch": 8.35,
      "grad_norm": 6.095053195953369,
      "learning_rate": 3.293807641633729e-06,
      "loss": 0.0604,
      "step": 44380
    },
    {
      "epoch": 8.35,
      "grad_norm": 19.563438415527344,
      "learning_rate": 3.2900432900432905e-06,
      "loss": 0.2798,
      "step": 44390
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.0039287409745156765,
      "learning_rate": 3.2862789384528516e-06,
      "loss": 0.1762,
      "step": 44400
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.3113458752632141,
      "learning_rate": 3.282514586862413e-06,
      "loss": 0.1477,
      "step": 44410
    },
    {
      "epoch": 8.36,
      "grad_norm": 3.928264856338501,
      "learning_rate": 3.2787502352719745e-06,
      "loss": 0.08,
      "step": 44420
    },
    {
      "epoch": 8.36,
      "grad_norm": 2.486176013946533,
      "learning_rate": 3.2749858836815364e-06,
      "loss": 0.0735,
      "step": 44430
    },
    {
      "epoch": 8.36,
      "grad_norm": 0.9972230792045593,
      "learning_rate": 3.271221532091098e-06,
      "loss": 0.1744,
      "step": 44440
    },
    {
      "epoch": 8.37,
      "grad_norm": 3.1484148502349854,
      "learning_rate": 3.267457180500659e-06,
      "loss": 0.3232,
      "step": 44450
    },
    {
      "epoch": 8.37,
      "grad_norm": 2.4035751819610596,
      "learning_rate": 3.2636928289102203e-06,
      "loss": 0.1661,
      "step": 44460
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.18072912096977234,
      "learning_rate": 3.259928477319782e-06,
      "loss": 0.0712,
      "step": 44470
    },
    {
      "epoch": 8.37,
      "grad_norm": 1.292542576789856,
      "learning_rate": 3.2561641257293437e-06,
      "loss": 0.1699,
      "step": 44480
    },
    {
      "epoch": 8.37,
      "grad_norm": 0.009601183235645294,
      "learning_rate": 3.2523997741389047e-06,
      "loss": 0.1745,
      "step": 44490
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.005314834415912628,
      "learning_rate": 3.248635422548466e-06,
      "loss": 0.1383,
      "step": 44500
    },
    {
      "epoch": 8.38,
      "grad_norm": 1.388582706451416,
      "learning_rate": 3.2448710709580277e-06,
      "loss": 0.3848,
      "step": 44510
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.6272100210189819,
      "learning_rate": 3.2411067193675895e-06,
      "loss": 0.0734,
      "step": 44520
    },
    {
      "epoch": 8.38,
      "grad_norm": 0.6307181119918823,
      "learning_rate": 3.2373423677771506e-06,
      "loss": 0.1429,
      "step": 44530
    },
    {
      "epoch": 8.38,
      "grad_norm": 57.35911178588867,
      "learning_rate": 3.233578016186712e-06,
      "loss": 0.082,
      "step": 44540
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.41133126616477966,
      "learning_rate": 3.2298136645962735e-06,
      "loss": 0.0921,
      "step": 44550
    },
    {
      "epoch": 8.39,
      "grad_norm": 1.0685217380523682,
      "learning_rate": 3.226049313005835e-06,
      "loss": 0.0813,
      "step": 44560
    },
    {
      "epoch": 8.39,
      "grad_norm": 6.20578145980835,
      "learning_rate": 3.222284961415397e-06,
      "loss": 0.237,
      "step": 44570
    },
    {
      "epoch": 8.39,
      "grad_norm": 2.740791082382202,
      "learning_rate": 3.218520609824958e-06,
      "loss": 0.3265,
      "step": 44580
    },
    {
      "epoch": 8.39,
      "grad_norm": 0.0075504048727452755,
      "learning_rate": 3.2147562582345193e-06,
      "loss": 0.0197,
      "step": 44590
    },
    {
      "epoch": 8.39,
      "grad_norm": 18.999364852905273,
      "learning_rate": 3.210991906644081e-06,
      "loss": 0.0784,
      "step": 44600
    },
    {
      "epoch": 8.4,
      "grad_norm": 5.307258605957031,
      "learning_rate": 3.2072275550536427e-06,
      "loss": 0.135,
      "step": 44610
    },
    {
      "epoch": 8.4,
      "grad_norm": 7.78779411315918,
      "learning_rate": 3.2034632034632033e-06,
      "loss": 0.2344,
      "step": 44620
    },
    {
      "epoch": 8.4,
      "grad_norm": 6.372420310974121,
      "learning_rate": 3.199698851872765e-06,
      "loss": 0.3044,
      "step": 44630
    },
    {
      "epoch": 8.4,
      "grad_norm": 6.207488536834717,
      "learning_rate": 3.1959345002823266e-06,
      "loss": 0.2296,
      "step": 44640
    },
    {
      "epoch": 8.4,
      "grad_norm": 8.6488037109375,
      "learning_rate": 3.192170148691888e-06,
      "loss": 0.2034,
      "step": 44650
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.005694823805242777,
      "learning_rate": 3.188405797101449e-06,
      "loss": 0.0135,
      "step": 44660
    },
    {
      "epoch": 8.41,
      "grad_norm": 5.9774088859558105,
      "learning_rate": 3.184641445511011e-06,
      "loss": 0.14,
      "step": 44670
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.31471145153045654,
      "learning_rate": 3.1808770939205725e-06,
      "loss": 0.1417,
      "step": 44680
    },
    {
      "epoch": 8.41,
      "grad_norm": 0.3139224946498871,
      "learning_rate": 3.177112742330134e-06,
      "loss": 0.2549,
      "step": 44690
    },
    {
      "epoch": 8.41,
      "grad_norm": 5.157718181610107,
      "learning_rate": 3.1733483907396954e-06,
      "loss": 0.1614,
      "step": 44700
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.015323806554079056,
      "learning_rate": 3.1695840391492564e-06,
      "loss": 0.0931,
      "step": 44710
    },
    {
      "epoch": 8.42,
      "grad_norm": 10.106663703918457,
      "learning_rate": 3.1658196875588183e-06,
      "loss": 0.1721,
      "step": 44720
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.0030023783911019564,
      "learning_rate": 3.1620553359683798e-06,
      "loss": 0.0969,
      "step": 44730
    },
    {
      "epoch": 8.42,
      "grad_norm": 8.230118751525879,
      "learning_rate": 3.1582909843779412e-06,
      "loss": 0.1501,
      "step": 44740
    },
    {
      "epoch": 8.42,
      "grad_norm": 1.0531959533691406,
      "learning_rate": 3.1545266327875023e-06,
      "loss": 0.1122,
      "step": 44750
    },
    {
      "epoch": 8.42,
      "grad_norm": 0.8612009882926941,
      "learning_rate": 3.1507622811970637e-06,
      "loss": 0.3296,
      "step": 44760
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.09927431493997574,
      "learning_rate": 3.1469979296066256e-06,
      "loss": 0.1973,
      "step": 44770
    },
    {
      "epoch": 8.43,
      "grad_norm": 4.921050071716309,
      "learning_rate": 3.143233578016187e-06,
      "loss": 0.1439,
      "step": 44780
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.23526562750339508,
      "learning_rate": 3.1394692264257485e-06,
      "loss": 0.1628,
      "step": 44790
    },
    {
      "epoch": 8.43,
      "grad_norm": 0.003873916110023856,
      "learning_rate": 3.1357048748353096e-06,
      "loss": 0.2202,
      "step": 44800
    },
    {
      "epoch": 8.43,
      "grad_norm": 3.2889604568481445,
      "learning_rate": 3.1319405232448715e-06,
      "loss": 0.059,
      "step": 44810
    },
    {
      "epoch": 8.44,
      "grad_norm": 7.945717811584473,
      "learning_rate": 3.128176171654433e-06,
      "loss": 0.2193,
      "step": 44820
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.004206668119877577,
      "learning_rate": 3.1244118200639944e-06,
      "loss": 0.1642,
      "step": 44830
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.11788476258516312,
      "learning_rate": 3.1206474684735554e-06,
      "loss": 0.0731,
      "step": 44840
    },
    {
      "epoch": 8.44,
      "grad_norm": 0.13546054065227509,
      "learning_rate": 3.116883116883117e-06,
      "loss": 0.095,
      "step": 44850
    },
    {
      "epoch": 8.44,
      "grad_norm": 24.633913040161133,
      "learning_rate": 3.1131187652926788e-06,
      "loss": 0.306,
      "step": 44860
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.6971827149391174,
      "learning_rate": 3.1093544137022402e-06,
      "loss": 0.1036,
      "step": 44870
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.011020142585039139,
      "learning_rate": 3.1055900621118013e-06,
      "loss": 0.0155,
      "step": 44880
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.010694989934563637,
      "learning_rate": 3.1018257105213627e-06,
      "loss": 0.1848,
      "step": 44890
    },
    {
      "epoch": 8.45,
      "grad_norm": 2.7448291778564453,
      "learning_rate": 3.0980613589309246e-06,
      "loss": 0.0424,
      "step": 44900
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.005120688583701849,
      "learning_rate": 3.094297007340486e-06,
      "loss": 0.0762,
      "step": 44910
    },
    {
      "epoch": 8.45,
      "grad_norm": 0.9466463327407837,
      "learning_rate": 3.0905326557500475e-06,
      "loss": 0.0285,
      "step": 44920
    },
    {
      "epoch": 8.46,
      "grad_norm": 15.811626434326172,
      "learning_rate": 3.0867683041596086e-06,
      "loss": 0.1777,
      "step": 44930
    },
    {
      "epoch": 8.46,
      "grad_norm": 16.42203712463379,
      "learning_rate": 3.08300395256917e-06,
      "loss": 0.2163,
      "step": 44940
    },
    {
      "epoch": 8.46,
      "grad_norm": 7.775193214416504,
      "learning_rate": 3.079239600978732e-06,
      "loss": 0.2141,
      "step": 44950
    },
    {
      "epoch": 8.46,
      "grad_norm": 17.66362190246582,
      "learning_rate": 3.0754752493882934e-06,
      "loss": 0.3812,
      "step": 44960
    },
    {
      "epoch": 8.46,
      "grad_norm": 0.023539405316114426,
      "learning_rate": 3.0717108977978544e-06,
      "loss": 0.2688,
      "step": 44970
    },
    {
      "epoch": 8.47,
      "grad_norm": 1.2292659282684326,
      "learning_rate": 3.067946546207416e-06,
      "loss": 0.1505,
      "step": 44980
    },
    {
      "epoch": 8.47,
      "grad_norm": 13.065359115600586,
      "learning_rate": 3.0641821946169773e-06,
      "loss": 0.1703,
      "step": 44990
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.004679796751588583,
      "learning_rate": 3.0604178430265392e-06,
      "loss": 0.0925,
      "step": 45000
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.036459118127822876,
      "learning_rate": 3.0566534914361003e-06,
      "loss": 0.2301,
      "step": 45010
    },
    {
      "epoch": 8.47,
      "grad_norm": 0.010220129042863846,
      "learning_rate": 3.0528891398456617e-06,
      "loss": 0.1274,
      "step": 45020
    },
    {
      "epoch": 8.48,
      "grad_norm": 10.476090431213379,
      "learning_rate": 3.049124788255223e-06,
      "loss": 0.3372,
      "step": 45030
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.45662033557891846,
      "learning_rate": 3.045360436664785e-06,
      "loss": 0.0596,
      "step": 45040
    },
    {
      "epoch": 8.48,
      "grad_norm": 16.36342430114746,
      "learning_rate": 3.0415960850743465e-06,
      "loss": 0.14,
      "step": 45050
    },
    {
      "epoch": 8.48,
      "grad_norm": 26.2547664642334,
      "learning_rate": 3.0378317334839076e-06,
      "loss": 0.2302,
      "step": 45060
    },
    {
      "epoch": 8.48,
      "grad_norm": 12.077350616455078,
      "learning_rate": 3.034067381893469e-06,
      "loss": 0.2678,
      "step": 45070
    },
    {
      "epoch": 8.48,
      "grad_norm": 0.011667963117361069,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 0.2061,
      "step": 45080
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.010367770679295063,
      "learning_rate": 3.0265386787125924e-06,
      "loss": 0.1761,
      "step": 45090
    },
    {
      "epoch": 8.49,
      "grad_norm": 32.7271842956543,
      "learning_rate": 3.0227743271221534e-06,
      "loss": 0.1248,
      "step": 45100
    },
    {
      "epoch": 8.49,
      "grad_norm": 2.3465240001678467,
      "learning_rate": 3.019009975531715e-06,
      "loss": 0.0338,
      "step": 45110
    },
    {
      "epoch": 8.49,
      "grad_norm": 0.2795522212982178,
      "learning_rate": 3.0152456239412763e-06,
      "loss": 0.3559,
      "step": 45120
    },
    {
      "epoch": 8.49,
      "grad_norm": 27.78301429748535,
      "learning_rate": 3.011481272350838e-06,
      "loss": 0.312,
      "step": 45130
    },
    {
      "epoch": 8.5,
      "grad_norm": 9.866744041442871,
      "learning_rate": 3.007716920760399e-06,
      "loss": 0.1065,
      "step": 45140
    },
    {
      "epoch": 8.5,
      "grad_norm": 16.968303680419922,
      "learning_rate": 3.0039525691699607e-06,
      "loss": 0.1926,
      "step": 45150
    },
    {
      "epoch": 8.5,
      "grad_norm": 10.868888854980469,
      "learning_rate": 3.000188217579522e-06,
      "loss": 0.2971,
      "step": 45160
    },
    {
      "epoch": 8.5,
      "grad_norm": 0.5675098299980164,
      "learning_rate": 2.9964238659890836e-06,
      "loss": 0.2119,
      "step": 45170
    },
    {
      "epoch": 8.5,
      "grad_norm": 4.395773887634277,
      "learning_rate": 2.9926595143986455e-06,
      "loss": 0.1184,
      "step": 45180
    },
    {
      "epoch": 8.51,
      "grad_norm": 17.195354461669922,
      "learning_rate": 2.9888951628082066e-06,
      "loss": 0.0754,
      "step": 45190
    },
    {
      "epoch": 8.51,
      "grad_norm": 14.842548370361328,
      "learning_rate": 2.985130811217768e-06,
      "loss": 0.305,
      "step": 45200
    },
    {
      "epoch": 8.51,
      "grad_norm": 6.892485618591309,
      "learning_rate": 2.9813664596273295e-06,
      "loss": 0.184,
      "step": 45210
    },
    {
      "epoch": 8.51,
      "grad_norm": 1.4683300256729126,
      "learning_rate": 2.977602108036891e-06,
      "loss": 0.0571,
      "step": 45220
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.05412657931447029,
      "learning_rate": 2.973837756446452e-06,
      "loss": 0.1239,
      "step": 45230
    },
    {
      "epoch": 8.51,
      "grad_norm": 0.006969267502427101,
      "learning_rate": 2.970073404856014e-06,
      "loss": 0.21,
      "step": 45240
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.017576254904270172,
      "learning_rate": 2.9663090532655753e-06,
      "loss": 0.1513,
      "step": 45250
    },
    {
      "epoch": 8.52,
      "grad_norm": 1.1532460451126099,
      "learning_rate": 2.9625447016751368e-06,
      "loss": 0.2187,
      "step": 45260
    },
    {
      "epoch": 8.52,
      "grad_norm": 0.023211367428302765,
      "learning_rate": 2.958780350084698e-06,
      "loss": 0.2789,
      "step": 45270
    },
    {
      "epoch": 8.52,
      "grad_norm": 4.530707359313965,
      "learning_rate": 2.9550159984942593e-06,
      "loss": 0.2529,
      "step": 45280
    },
    {
      "epoch": 8.52,
      "grad_norm": 9.255178451538086,
      "learning_rate": 2.951251646903821e-06,
      "loss": 0.0488,
      "step": 45290
    },
    {
      "epoch": 8.53,
      "grad_norm": 17.951570510864258,
      "learning_rate": 2.9474872953133826e-06,
      "loss": 0.1728,
      "step": 45300
    },
    {
      "epoch": 8.53,
      "grad_norm": 1.0469212532043457,
      "learning_rate": 2.943722943722944e-06,
      "loss": 0.0923,
      "step": 45310
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.013357494957745075,
      "learning_rate": 2.939958592132505e-06,
      "loss": 0.0506,
      "step": 45320
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.004213337320834398,
      "learning_rate": 2.936194240542067e-06,
      "loss": 0.2935,
      "step": 45330
    },
    {
      "epoch": 8.53,
      "grad_norm": 0.005573914386332035,
      "learning_rate": 2.9324298889516285e-06,
      "loss": 0.193,
      "step": 45340
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.00582647230476141,
      "learning_rate": 2.92866553736119e-06,
      "loss": 0.037,
      "step": 45350
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.017139622941613197,
      "learning_rate": 2.924901185770751e-06,
      "loss": 0.2735,
      "step": 45360
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.003541667712852359,
      "learning_rate": 2.9211368341803124e-06,
      "loss": 0.1129,
      "step": 45370
    },
    {
      "epoch": 8.54,
      "grad_norm": 0.006987821776419878,
      "learning_rate": 2.9173724825898743e-06,
      "loss": 0.1957,
      "step": 45380
    },
    {
      "epoch": 8.54,
      "grad_norm": 7.132236480712891,
      "learning_rate": 2.9136081309994358e-06,
      "loss": 0.1669,
      "step": 45390
    },
    {
      "epoch": 8.55,
      "grad_norm": 6.705604076385498,
      "learning_rate": 2.9098437794089972e-06,
      "loss": 0.2187,
      "step": 45400
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.006447926163673401,
      "learning_rate": 2.9060794278185583e-06,
      "loss": 0.1194,
      "step": 45410
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.022328652441501617,
      "learning_rate": 2.9023150762281197e-06,
      "loss": 0.26,
      "step": 45420
    },
    {
      "epoch": 8.55,
      "grad_norm": 7.96865177154541,
      "learning_rate": 2.8985507246376816e-06,
      "loss": 0.1744,
      "step": 45430
    },
    {
      "epoch": 8.55,
      "grad_norm": 15.768625259399414,
      "learning_rate": 2.894786373047243e-06,
      "loss": 0.2592,
      "step": 45440
    },
    {
      "epoch": 8.55,
      "grad_norm": 0.013202544301748276,
      "learning_rate": 2.891022021456804e-06,
      "loss": 0.132,
      "step": 45450
    },
    {
      "epoch": 8.56,
      "grad_norm": 23.192529678344727,
      "learning_rate": 2.8872576698663656e-06,
      "loss": 0.3129,
      "step": 45460
    },
    {
      "epoch": 8.56,
      "grad_norm": 0.6409586071968079,
      "learning_rate": 2.8834933182759275e-06,
      "loss": 0.1619,
      "step": 45470
    },
    {
      "epoch": 8.56,
      "grad_norm": 15.344633102416992,
      "learning_rate": 2.879728966685489e-06,
      "loss": 0.1935,
      "step": 45480
    },
    {
      "epoch": 8.56,
      "grad_norm": 13.032301902770996,
      "learning_rate": 2.87596461509505e-06,
      "loss": 0.3162,
      "step": 45490
    },
    {
      "epoch": 8.56,
      "grad_norm": 25.099225997924805,
      "learning_rate": 2.8722002635046114e-06,
      "loss": 0.1703,
      "step": 45500
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.04938386380672455,
      "learning_rate": 2.868435911914173e-06,
      "loss": 0.1025,
      "step": 45510
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.004378327634185553,
      "learning_rate": 2.8646715603237348e-06,
      "loss": 0.1551,
      "step": 45520
    },
    {
      "epoch": 8.57,
      "grad_norm": 11.770549774169922,
      "learning_rate": 2.8609072087332962e-06,
      "loss": 0.2056,
      "step": 45530
    },
    {
      "epoch": 8.57,
      "grad_norm": 1.2891154289245605,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 0.2307,
      "step": 45540
    },
    {
      "epoch": 8.57,
      "grad_norm": 0.08139143884181976,
      "learning_rate": 2.8533785055524187e-06,
      "loss": 0.3123,
      "step": 45550
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.0120173254981637,
      "learning_rate": 2.84961415396198e-06,
      "loss": 0.1385,
      "step": 45560
    },
    {
      "epoch": 8.58,
      "grad_norm": 4.636909008026123,
      "learning_rate": 2.845849802371542e-06,
      "loss": 0.1068,
      "step": 45570
    },
    {
      "epoch": 8.58,
      "grad_norm": 1.510550618171692,
      "learning_rate": 2.842085450781103e-06,
      "loss": 0.1774,
      "step": 45580
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.013908603228628635,
      "learning_rate": 2.8383210991906646e-06,
      "loss": 0.2061,
      "step": 45590
    },
    {
      "epoch": 8.58,
      "grad_norm": 4.3199872970581055,
      "learning_rate": 2.834556747600226e-06,
      "loss": 0.0438,
      "step": 45600
    },
    {
      "epoch": 8.58,
      "grad_norm": 23.759498596191406,
      "learning_rate": 2.830792396009788e-06,
      "loss": 0.2392,
      "step": 45610
    },
    {
      "epoch": 8.59,
      "grad_norm": 0.058604270219802856,
      "learning_rate": 2.827028044419349e-06,
      "loss": 0.1409,
      "step": 45620
    },
    {
      "epoch": 8.59,
      "grad_norm": 25.992321014404297,
      "learning_rate": 2.8232636928289104e-06,
      "loss": 0.2509,
      "step": 45630
    },
    {
      "epoch": 8.59,
      "grad_norm": 9.984472274780273,
      "learning_rate": 2.819499341238472e-06,
      "loss": 0.1589,
      "step": 45640
    },
    {
      "epoch": 8.59,
      "grad_norm": 21.208911895751953,
      "learning_rate": 2.8157349896480333e-06,
      "loss": 0.1405,
      "step": 45650
    },
    {
      "epoch": 8.59,
      "grad_norm": 0.31339210271835327,
      "learning_rate": 2.811970638057595e-06,
      "loss": 0.2325,
      "step": 45660
    },
    {
      "epoch": 8.6,
      "grad_norm": 19.97664451599121,
      "learning_rate": 2.8082062864671562e-06,
      "loss": 0.3261,
      "step": 45670
    },
    {
      "epoch": 8.6,
      "grad_norm": 14.917105674743652,
      "learning_rate": 2.8044419348767177e-06,
      "loss": 0.0516,
      "step": 45680
    },
    {
      "epoch": 8.6,
      "grad_norm": 2.4222874641418457,
      "learning_rate": 2.800677583286279e-06,
      "loss": 0.1418,
      "step": 45690
    },
    {
      "epoch": 8.6,
      "grad_norm": 9.627737045288086,
      "learning_rate": 2.7969132316958406e-06,
      "loss": 0.191,
      "step": 45700
    },
    {
      "epoch": 8.6,
      "grad_norm": 0.46485409140586853,
      "learning_rate": 2.7931488801054017e-06,
      "loss": 0.1743,
      "step": 45710
    },
    {
      "epoch": 8.61,
      "grad_norm": 6.60011100769043,
      "learning_rate": 2.7893845285149635e-06,
      "loss": 0.3793,
      "step": 45720
    },
    {
      "epoch": 8.61,
      "grad_norm": 2.756089448928833,
      "learning_rate": 2.785620176924525e-06,
      "loss": 0.0489,
      "step": 45730
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.05293891206383705,
      "learning_rate": 2.7818558253340865e-06,
      "loss": 0.1094,
      "step": 45740
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.056367985904216766,
      "learning_rate": 2.7780914737436475e-06,
      "loss": 0.1845,
      "step": 45750
    },
    {
      "epoch": 8.61,
      "grad_norm": 32.57680130004883,
      "learning_rate": 2.7743271221532094e-06,
      "loss": 0.3286,
      "step": 45760
    },
    {
      "epoch": 8.61,
      "grad_norm": 0.35892462730407715,
      "learning_rate": 2.770562770562771e-06,
      "loss": 0.0907,
      "step": 45770
    },
    {
      "epoch": 8.62,
      "grad_norm": 11.20388412475586,
      "learning_rate": 2.7667984189723323e-06,
      "loss": 0.1567,
      "step": 45780
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.008292987011373043,
      "learning_rate": 2.7630340673818938e-06,
      "loss": 0.1083,
      "step": 45790
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.6199795603752136,
      "learning_rate": 2.759269715791455e-06,
      "loss": 0.1768,
      "step": 45800
    },
    {
      "epoch": 8.62,
      "grad_norm": 27.761178970336914,
      "learning_rate": 2.7555053642010167e-06,
      "loss": 0.0761,
      "step": 45810
    },
    {
      "epoch": 8.62,
      "grad_norm": 0.31952348351478577,
      "learning_rate": 2.751741012610578e-06,
      "loss": 0.1591,
      "step": 45820
    },
    {
      "epoch": 8.63,
      "grad_norm": 0.004031436052173376,
      "learning_rate": 2.7479766610201396e-06,
      "loss": 0.1197,
      "step": 45830
    },
    {
      "epoch": 8.63,
      "grad_norm": 14.182780265808105,
      "learning_rate": 2.7442123094297007e-06,
      "loss": 0.115,
      "step": 45840
    },
    {
      "epoch": 8.63,
      "grad_norm": 5.576279163360596,
      "learning_rate": 2.740447957839262e-06,
      "loss": 0.1402,
      "step": 45850
    },
    {
      "epoch": 8.63,
      "grad_norm": 4.379868984222412,
      "learning_rate": 2.736683606248824e-06,
      "loss": 0.1156,
      "step": 45860
    },
    {
      "epoch": 8.63,
      "grad_norm": 32.59175109863281,
      "learning_rate": 2.7329192546583855e-06,
      "loss": 0.2746,
      "step": 45870
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.01854350045323372,
      "learning_rate": 2.7291549030679465e-06,
      "loss": 0.0898,
      "step": 45880
    },
    {
      "epoch": 8.64,
      "grad_norm": 6.422102928161621,
      "learning_rate": 2.725390551477508e-06,
      "loss": 0.1894,
      "step": 45890
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.46982038021087646,
      "learning_rate": 2.72162619988707e-06,
      "loss": 0.2446,
      "step": 45900
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.05521256849169731,
      "learning_rate": 2.7178618482966313e-06,
      "loss": 0.0716,
      "step": 45910
    },
    {
      "epoch": 8.64,
      "grad_norm": 2.0827395915985107,
      "learning_rate": 2.7140974967061928e-06,
      "loss": 0.3358,
      "step": 45920
    },
    {
      "epoch": 8.64,
      "grad_norm": 0.1304616779088974,
      "learning_rate": 2.710333145115754e-06,
      "loss": 0.1046,
      "step": 45930
    },
    {
      "epoch": 8.65,
      "grad_norm": 8.427474975585938,
      "learning_rate": 2.7065687935253153e-06,
      "loss": 0.1209,
      "step": 45940
    },
    {
      "epoch": 8.65,
      "grad_norm": 2.8755972385406494,
      "learning_rate": 2.702804441934877e-06,
      "loss": 0.1367,
      "step": 45950
    },
    {
      "epoch": 8.65,
      "grad_norm": 7.7070817947387695,
      "learning_rate": 2.6990400903444386e-06,
      "loss": 0.2677,
      "step": 45960
    },
    {
      "epoch": 8.65,
      "grad_norm": 0.15384049713611603,
      "learning_rate": 2.6952757387539996e-06,
      "loss": 0.1566,
      "step": 45970
    },
    {
      "epoch": 8.65,
      "grad_norm": 0.16432297229766846,
      "learning_rate": 2.691511387163561e-06,
      "loss": 0.1092,
      "step": 45980
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.004089550580829382,
      "learning_rate": 2.687747035573123e-06,
      "loss": 0.1567,
      "step": 45990
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.06510978937149048,
      "learning_rate": 2.6839826839826844e-06,
      "loss": 0.1217,
      "step": 46000
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.0147783812135458,
      "learning_rate": 2.6802183323922455e-06,
      "loss": 0.0989,
      "step": 46010
    },
    {
      "epoch": 8.66,
      "grad_norm": 3.4192616939544678,
      "learning_rate": 2.676453980801807e-06,
      "loss": 0.1176,
      "step": 46020
    },
    {
      "epoch": 8.66,
      "grad_norm": 0.016071872785687447,
      "learning_rate": 2.6726896292113684e-06,
      "loss": 0.1426,
      "step": 46030
    },
    {
      "epoch": 8.67,
      "grad_norm": 3.0639872550964355,
      "learning_rate": 2.6689252776209303e-06,
      "loss": 0.1394,
      "step": 46040
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.028606731444597244,
      "learning_rate": 2.6651609260304918e-06,
      "loss": 0.2118,
      "step": 46050
    },
    {
      "epoch": 8.67,
      "grad_norm": 4.988860607147217,
      "learning_rate": 2.661396574440053e-06,
      "loss": 0.132,
      "step": 46060
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.005040525458753109,
      "learning_rate": 2.6576322228496142e-06,
      "loss": 0.1482,
      "step": 46070
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.00744453864172101,
      "learning_rate": 2.6538678712591757e-06,
      "loss": 0.2448,
      "step": 46080
    },
    {
      "epoch": 8.67,
      "grad_norm": 0.007808149326592684,
      "learning_rate": 2.6501035196687376e-06,
      "loss": 0.1718,
      "step": 46090
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.21629448235034943,
      "learning_rate": 2.6463391680782986e-06,
      "loss": 0.2682,
      "step": 46100
    },
    {
      "epoch": 8.68,
      "grad_norm": 34.93745803833008,
      "learning_rate": 2.64257481648786e-06,
      "loss": 0.2212,
      "step": 46110
    },
    {
      "epoch": 8.68,
      "grad_norm": 34.2768669128418,
      "learning_rate": 2.6388104648974216e-06,
      "loss": 0.159,
      "step": 46120
    },
    {
      "epoch": 8.68,
      "grad_norm": 0.011613433249294758,
      "learning_rate": 2.6350461133069834e-06,
      "loss": 0.1231,
      "step": 46130
    },
    {
      "epoch": 8.68,
      "grad_norm": 1.2315365076065063,
      "learning_rate": 2.631281761716545e-06,
      "loss": 0.2162,
      "step": 46140
    },
    {
      "epoch": 8.69,
      "grad_norm": 11.116827011108398,
      "learning_rate": 2.627517410126106e-06,
      "loss": 0.2153,
      "step": 46150
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.030044695362448692,
      "learning_rate": 2.6237530585356674e-06,
      "loss": 0.0453,
      "step": 46160
    },
    {
      "epoch": 8.69,
      "grad_norm": 8.410063743591309,
      "learning_rate": 2.619988706945229e-06,
      "loss": 0.0547,
      "step": 46170
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.006890262942761183,
      "learning_rate": 2.6162243553547907e-06,
      "loss": 0.0899,
      "step": 46180
    },
    {
      "epoch": 8.69,
      "grad_norm": 0.010163234546780586,
      "learning_rate": 2.6124600037643518e-06,
      "loss": 0.2548,
      "step": 46190
    },
    {
      "epoch": 8.7,
      "grad_norm": 19.27224349975586,
      "learning_rate": 2.6086956521739132e-06,
      "loss": 0.3776,
      "step": 46200
    },
    {
      "epoch": 8.7,
      "grad_norm": 2.746565580368042,
      "learning_rate": 2.6049313005834747e-06,
      "loss": 0.1028,
      "step": 46210
    },
    {
      "epoch": 8.7,
      "grad_norm": 0.004154515452682972,
      "learning_rate": 2.601166948993036e-06,
      "loss": 0.2355,
      "step": 46220
    },
    {
      "epoch": 8.7,
      "grad_norm": 7.875269889831543,
      "learning_rate": 2.597402597402597e-06,
      "loss": 0.1833,
      "step": 46230
    },
    {
      "epoch": 8.7,
      "grad_norm": 9.690305709838867,
      "learning_rate": 2.593638245812159e-06,
      "loss": 0.0936,
      "step": 46240
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.0041099716909229755,
      "learning_rate": 2.5898738942217205e-06,
      "loss": 0.0935,
      "step": 46250
    },
    {
      "epoch": 8.71,
      "grad_norm": 6.63506555557251,
      "learning_rate": 2.586109542631282e-06,
      "loss": 0.0388,
      "step": 46260
    },
    {
      "epoch": 8.71,
      "grad_norm": 12.502037048339844,
      "learning_rate": 2.582345191040844e-06,
      "loss": 0.1522,
      "step": 46270
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.004462874494493008,
      "learning_rate": 2.578580839450405e-06,
      "loss": 0.0509,
      "step": 46280
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.07180631160736084,
      "learning_rate": 2.5748164878599664e-06,
      "loss": 0.1441,
      "step": 46290
    },
    {
      "epoch": 8.71,
      "grad_norm": 0.1847943216562271,
      "learning_rate": 2.571052136269528e-06,
      "loss": 0.208,
      "step": 46300
    },
    {
      "epoch": 8.72,
      "grad_norm": 27.239225387573242,
      "learning_rate": 2.5672877846790893e-06,
      "loss": 0.2557,
      "step": 46310
    },
    {
      "epoch": 8.72,
      "grad_norm": 6.99392557144165,
      "learning_rate": 2.5635234330886503e-06,
      "loss": 0.0806,
      "step": 46320
    },
    {
      "epoch": 8.72,
      "grad_norm": 6.173346042633057,
      "learning_rate": 2.5597590814982122e-06,
      "loss": 0.1426,
      "step": 46330
    },
    {
      "epoch": 8.72,
      "grad_norm": 0.1534779965877533,
      "learning_rate": 2.5559947299077737e-06,
      "loss": 0.1511,
      "step": 46340
    },
    {
      "epoch": 8.72,
      "grad_norm": 2.7229299545288086,
      "learning_rate": 2.552230378317335e-06,
      "loss": 0.1939,
      "step": 46350
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.008107272908091545,
      "learning_rate": 2.548466026726896e-06,
      "loss": 0.2814,
      "step": 46360
    },
    {
      "epoch": 8.73,
      "grad_norm": 15.321224212646484,
      "learning_rate": 2.5447016751364576e-06,
      "loss": 0.2167,
      "step": 46370
    },
    {
      "epoch": 8.73,
      "grad_norm": 6.417855739593506,
      "learning_rate": 2.5409373235460195e-06,
      "loss": 0.3211,
      "step": 46380
    },
    {
      "epoch": 8.73,
      "grad_norm": 0.5385270714759827,
      "learning_rate": 2.537172971955581e-06,
      "loss": 0.1316,
      "step": 46390
    },
    {
      "epoch": 8.73,
      "grad_norm": 16.703105926513672,
      "learning_rate": 2.5334086203651425e-06,
      "loss": 0.2515,
      "step": 46400
    },
    {
      "epoch": 8.74,
      "grad_norm": 6.621845245361328,
      "learning_rate": 2.5296442687747035e-06,
      "loss": 0.1828,
      "step": 46410
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.00556070264428854,
      "learning_rate": 2.5258799171842654e-06,
      "loss": 0.2235,
      "step": 46420
    },
    {
      "epoch": 8.74,
      "grad_norm": 20.055273056030273,
      "learning_rate": 2.522115565593827e-06,
      "loss": 0.1556,
      "step": 46430
    },
    {
      "epoch": 8.74,
      "grad_norm": 8.877673149108887,
      "learning_rate": 2.5183512140033883e-06,
      "loss": 0.1678,
      "step": 46440
    },
    {
      "epoch": 8.74,
      "grad_norm": 0.08181773126125336,
      "learning_rate": 2.5145868624129493e-06,
      "loss": 0.2501,
      "step": 46450
    },
    {
      "epoch": 8.74,
      "grad_norm": 23.303861618041992,
      "learning_rate": 2.510822510822511e-06,
      "loss": 0.1171,
      "step": 46460
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.0711456760764122,
      "learning_rate": 2.5070581592320727e-06,
      "loss": 0.2447,
      "step": 46470
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.0075196451507508755,
      "learning_rate": 2.503293807641634e-06,
      "loss": 0.0671,
      "step": 46480
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.23987708985805511,
      "learning_rate": 2.4995294560511956e-06,
      "loss": 0.353,
      "step": 46490
    },
    {
      "epoch": 8.75,
      "grad_norm": 18.07382583618164,
      "learning_rate": 2.4957651044607566e-06,
      "loss": 0.3534,
      "step": 46500
    },
    {
      "epoch": 8.75,
      "grad_norm": 3.4529011249542236,
      "learning_rate": 2.492000752870318e-06,
      "loss": 0.225,
      "step": 46510
    },
    {
      "epoch": 8.76,
      "grad_norm": 17.33688735961914,
      "learning_rate": 2.48823640127988e-06,
      "loss": 0.4387,
      "step": 46520
    },
    {
      "epoch": 8.76,
      "grad_norm": 0.09231207519769669,
      "learning_rate": 2.484472049689441e-06,
      "loss": 0.2064,
      "step": 46530
    },
    {
      "epoch": 8.76,
      "grad_norm": 12.045492172241211,
      "learning_rate": 2.480707698099003e-06,
      "loss": 0.1439,
      "step": 46540
    },
    {
      "epoch": 8.76,
      "grad_norm": 4.1995930671691895,
      "learning_rate": 2.476943346508564e-06,
      "loss": 0.1281,
      "step": 46550
    },
    {
      "epoch": 8.76,
      "grad_norm": 13.70324420928955,
      "learning_rate": 2.473178994918126e-06,
      "loss": 0.1381,
      "step": 46560
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.007012187037616968,
      "learning_rate": 2.469414643327687e-06,
      "loss": 0.272,
      "step": 46570
    },
    {
      "epoch": 8.77,
      "grad_norm": 44.52003479003906,
      "learning_rate": 2.4656502917372483e-06,
      "loss": 0.1375,
      "step": 46580
    },
    {
      "epoch": 8.77,
      "grad_norm": 21.349685668945312,
      "learning_rate": 2.4618859401468098e-06,
      "loss": 0.3543,
      "step": 46590
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.020648472011089325,
      "learning_rate": 2.4581215885563712e-06,
      "loss": 0.0916,
      "step": 46600
    },
    {
      "epoch": 8.77,
      "grad_norm": 5.295027256011963,
      "learning_rate": 2.4543572369659327e-06,
      "loss": 0.1277,
      "step": 46610
    },
    {
      "epoch": 8.77,
      "grad_norm": 0.03927454352378845,
      "learning_rate": 2.450592885375494e-06,
      "loss": 0.3085,
      "step": 46620
    },
    {
      "epoch": 8.78,
      "grad_norm": 2.292860269546509,
      "learning_rate": 2.446828533785056e-06,
      "loss": 0.1304,
      "step": 46630
    },
    {
      "epoch": 8.78,
      "grad_norm": 9.905180931091309,
      "learning_rate": 2.443064182194617e-06,
      "loss": 0.1017,
      "step": 46640
    },
    {
      "epoch": 8.78,
      "grad_norm": 0.009300472214818,
      "learning_rate": 2.4392998306041785e-06,
      "loss": 0.1845,
      "step": 46650
    },
    {
      "epoch": 8.78,
      "grad_norm": 9.025555610656738,
      "learning_rate": 2.43553547901374e-06,
      "loss": 0.2676,
      "step": 46660
    },
    {
      "epoch": 8.78,
      "grad_norm": 8.237606048583984,
      "learning_rate": 2.4317711274233015e-06,
      "loss": 0.1665,
      "step": 46670
    },
    {
      "epoch": 8.79,
      "grad_norm": 1.6117913722991943,
      "learning_rate": 2.428006775832863e-06,
      "loss": 0.225,
      "step": 46680
    },
    {
      "epoch": 8.79,
      "grad_norm": 7.155617713928223,
      "learning_rate": 2.4242424242424244e-06,
      "loss": 0.2417,
      "step": 46690
    },
    {
      "epoch": 8.79,
      "grad_norm": 9.043313980102539,
      "learning_rate": 2.420478072651986e-06,
      "loss": 0.016,
      "step": 46700
    },
    {
      "epoch": 8.79,
      "grad_norm": 5.589144706726074,
      "learning_rate": 2.4167137210615473e-06,
      "loss": 0.2292,
      "step": 46710
    },
    {
      "epoch": 8.79,
      "grad_norm": 9.714485168457031,
      "learning_rate": 2.4129493694711088e-06,
      "loss": 0.1287,
      "step": 46720
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.8655574917793274,
      "learning_rate": 2.4091850178806702e-06,
      "loss": 0.1116,
      "step": 46730
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.02492569573223591,
      "learning_rate": 2.4054206662902317e-06,
      "loss": 0.1961,
      "step": 46740
    },
    {
      "epoch": 8.8,
      "grad_norm": 27.022062301635742,
      "learning_rate": 2.401656314699793e-06,
      "loss": 0.2987,
      "step": 46750
    },
    {
      "epoch": 8.8,
      "grad_norm": 0.24443423748016357,
      "learning_rate": 2.3978919631093546e-06,
      "loss": 0.1629,
      "step": 46760
    },
    {
      "epoch": 8.8,
      "grad_norm": 6.803404331207275,
      "learning_rate": 2.394127611518916e-06,
      "loss": 0.3537,
      "step": 46770
    },
    {
      "epoch": 8.8,
      "grad_norm": 8.455885887145996,
      "learning_rate": 2.3903632599284775e-06,
      "loss": 0.1323,
      "step": 46780
    },
    {
      "epoch": 8.81,
      "grad_norm": 18.13623046875,
      "learning_rate": 2.386598908338039e-06,
      "loss": 0.0821,
      "step": 46790
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.008826698176562786,
      "learning_rate": 2.3828345567476005e-06,
      "loss": 0.1238,
      "step": 46800
    },
    {
      "epoch": 8.81,
      "grad_norm": 3.118528366088867,
      "learning_rate": 2.379070205157162e-06,
      "loss": 0.1266,
      "step": 46810
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.003516155993565917,
      "learning_rate": 2.3753058535667234e-06,
      "loss": 0.1796,
      "step": 46820
    },
    {
      "epoch": 8.81,
      "grad_norm": 0.03456592559814453,
      "learning_rate": 2.371541501976285e-06,
      "loss": 0.1428,
      "step": 46830
    },
    {
      "epoch": 8.82,
      "grad_norm": 0.04014725610613823,
      "learning_rate": 2.3677771503858463e-06,
      "loss": 0.1776,
      "step": 46840
    },
    {
      "epoch": 8.82,
      "grad_norm": 0.031509879976511,
      "learning_rate": 2.3640127987954078e-06,
      "loss": 0.0647,
      "step": 46850
    },
    {
      "epoch": 8.82,
      "grad_norm": 9.007281303405762,
      "learning_rate": 2.3602484472049692e-06,
      "loss": 0.1278,
      "step": 46860
    },
    {
      "epoch": 8.82,
      "grad_norm": 4.06545877456665,
      "learning_rate": 2.3564840956145303e-06,
      "loss": 0.1576,
      "step": 46870
    },
    {
      "epoch": 8.82,
      "grad_norm": 15.420404434204102,
      "learning_rate": 2.352719744024092e-06,
      "loss": 0.4458,
      "step": 46880
    },
    {
      "epoch": 8.83,
      "grad_norm": 15.1631498336792,
      "learning_rate": 2.3489553924336536e-06,
      "loss": 0.1589,
      "step": 46890
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.006104353815317154,
      "learning_rate": 2.345191040843215e-06,
      "loss": 0.175,
      "step": 46900
    },
    {
      "epoch": 8.83,
      "grad_norm": 9.437883377075195,
      "learning_rate": 2.3414266892527765e-06,
      "loss": 0.1312,
      "step": 46910
    },
    {
      "epoch": 8.83,
      "grad_norm": 9.362249374389648,
      "learning_rate": 2.337662337662338e-06,
      "loss": 0.1159,
      "step": 46920
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.00566528458148241,
      "learning_rate": 2.3338979860718994e-06,
      "loss": 0.1631,
      "step": 46930
    },
    {
      "epoch": 8.83,
      "grad_norm": 0.0037751495838165283,
      "learning_rate": 2.3301336344814605e-06,
      "loss": 0.0245,
      "step": 46940
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.05096269026398659,
      "learning_rate": 2.3263692828910224e-06,
      "loss": 0.0463,
      "step": 46950
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.30596205592155457,
      "learning_rate": 2.3226049313005834e-06,
      "loss": 0.2923,
      "step": 46960
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.10496103018522263,
      "learning_rate": 2.3188405797101453e-06,
      "loss": 0.1214,
      "step": 46970
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.006855998653918505,
      "learning_rate": 2.3150762281197063e-06,
      "loss": 0.212,
      "step": 46980
    },
    {
      "epoch": 8.84,
      "grad_norm": 0.010659908875823021,
      "learning_rate": 2.311311876529268e-06,
      "loss": 0.0308,
      "step": 46990
    },
    {
      "epoch": 8.85,
      "grad_norm": 28.5120849609375,
      "learning_rate": 2.3075475249388297e-06,
      "loss": 0.0913,
      "step": 47000
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.02101234346628189,
      "learning_rate": 2.3037831733483907e-06,
      "loss": 0.1668,
      "step": 47010
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.016702543944120407,
      "learning_rate": 2.3000188217579526e-06,
      "loss": 0.0734,
      "step": 47020
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.013027558103203773,
      "learning_rate": 2.2962544701675136e-06,
      "loss": 0.2104,
      "step": 47030
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.004345022141933441,
      "learning_rate": 2.2924901185770755e-06,
      "loss": 0.4069,
      "step": 47040
    },
    {
      "epoch": 8.86,
      "grad_norm": 5.609530448913574,
      "learning_rate": 2.2887257669866366e-06,
      "loss": 0.1431,
      "step": 47050
    },
    {
      "epoch": 8.86,
      "grad_norm": 11.015730857849121,
      "learning_rate": 2.2849614153961984e-06,
      "loss": 0.187,
      "step": 47060
    },
    {
      "epoch": 8.86,
      "grad_norm": 15.835850715637207,
      "learning_rate": 2.2811970638057595e-06,
      "loss": 0.2656,
      "step": 47070
    },
    {
      "epoch": 8.86,
      "grad_norm": 0.05781165510416031,
      "learning_rate": 2.277432712215321e-06,
      "loss": 0.2114,
      "step": 47080
    },
    {
      "epoch": 8.86,
      "grad_norm": 3.5304646492004395,
      "learning_rate": 2.2736683606248824e-06,
      "loss": 0.1518,
      "step": 47090
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.035017747431993484,
      "learning_rate": 2.269904009034444e-06,
      "loss": 0.0391,
      "step": 47100
    },
    {
      "epoch": 8.87,
      "grad_norm": 14.017739295959473,
      "learning_rate": 2.2661396574440053e-06,
      "loss": 0.2579,
      "step": 47110
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.07805225998163223,
      "learning_rate": 2.2623753058535668e-06,
      "loss": 0.0875,
      "step": 47120
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.5247780680656433,
      "learning_rate": 2.2586109542631287e-06,
      "loss": 0.1694,
      "step": 47130
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.22645582258701324,
      "learning_rate": 2.2548466026726897e-06,
      "loss": 0.0927,
      "step": 47140
    },
    {
      "epoch": 8.87,
      "grad_norm": 0.004124061204493046,
      "learning_rate": 2.251082251082251e-06,
      "loss": 0.1331,
      "step": 47150
    },
    {
      "epoch": 8.88,
      "grad_norm": 6.155834674835205,
      "learning_rate": 2.2473178994918126e-06,
      "loss": 0.082,
      "step": 47160
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.5222640037536621,
      "learning_rate": 2.243553547901374e-06,
      "loss": 0.0174,
      "step": 47170
    },
    {
      "epoch": 8.88,
      "grad_norm": 6.241584300994873,
      "learning_rate": 2.2397891963109355e-06,
      "loss": 0.106,
      "step": 47180
    },
    {
      "epoch": 8.88,
      "grad_norm": 27.596872329711914,
      "learning_rate": 2.236024844720497e-06,
      "loss": 0.2802,
      "step": 47190
    },
    {
      "epoch": 8.88,
      "grad_norm": 0.026385100558400154,
      "learning_rate": 2.2322604931300585e-06,
      "loss": 0.2643,
      "step": 47200
    },
    {
      "epoch": 8.89,
      "grad_norm": 3.5821003913879395,
      "learning_rate": 2.22849614153962e-06,
      "loss": 0.1007,
      "step": 47210
    },
    {
      "epoch": 8.89,
      "grad_norm": 5.696538925170898,
      "learning_rate": 2.2247317899491814e-06,
      "loss": 0.1882,
      "step": 47220
    },
    {
      "epoch": 8.89,
      "grad_norm": 17.851669311523438,
      "learning_rate": 2.220967438358743e-06,
      "loss": 0.1454,
      "step": 47230
    },
    {
      "epoch": 8.89,
      "grad_norm": 0.10345958918333054,
      "learning_rate": 2.2172030867683043e-06,
      "loss": 0.3214,
      "step": 47240
    },
    {
      "epoch": 8.89,
      "grad_norm": 0.2727500796318054,
      "learning_rate": 2.2134387351778658e-06,
      "loss": 0.2572,
      "step": 47250
    },
    {
      "epoch": 8.9,
      "grad_norm": 16.08303451538086,
      "learning_rate": 2.2096743835874272e-06,
      "loss": 0.3028,
      "step": 47260
    },
    {
      "epoch": 8.9,
      "grad_norm": 13.197850227355957,
      "learning_rate": 2.2059100319969887e-06,
      "loss": 0.1395,
      "step": 47270
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.030838772654533386,
      "learning_rate": 2.20214568040655e-06,
      "loss": 0.1485,
      "step": 47280
    },
    {
      "epoch": 8.9,
      "grad_norm": 2.794558525085449,
      "learning_rate": 2.1983813288161116e-06,
      "loss": 0.0787,
      "step": 47290
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.003674957202747464,
      "learning_rate": 2.194616977225673e-06,
      "loss": 0.1407,
      "step": 47300
    },
    {
      "epoch": 8.9,
      "grad_norm": 0.058988042175769806,
      "learning_rate": 2.1908526256352345e-06,
      "loss": 0.0725,
      "step": 47310
    },
    {
      "epoch": 8.91,
      "grad_norm": 24.179073333740234,
      "learning_rate": 2.187088274044796e-06,
      "loss": 0.3709,
      "step": 47320
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.003784383414313197,
      "learning_rate": 2.1833239224543575e-06,
      "loss": 0.0652,
      "step": 47330
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.3423250615596771,
      "learning_rate": 2.179559570863919e-06,
      "loss": 0.123,
      "step": 47340
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.02199295535683632,
      "learning_rate": 2.1757952192734804e-06,
      "loss": 0.1729,
      "step": 47350
    },
    {
      "epoch": 8.91,
      "grad_norm": 0.25307801365852356,
      "learning_rate": 2.172030867683042e-06,
      "loss": 0.181,
      "step": 47360
    },
    {
      "epoch": 8.92,
      "grad_norm": 3.5200555324554443,
      "learning_rate": 2.1682665160926033e-06,
      "loss": 0.1249,
      "step": 47370
    },
    {
      "epoch": 8.92,
      "grad_norm": 5.519104957580566,
      "learning_rate": 2.1645021645021648e-06,
      "loss": 0.0962,
      "step": 47380
    },
    {
      "epoch": 8.92,
      "grad_norm": 11.502580642700195,
      "learning_rate": 2.1607378129117262e-06,
      "loss": 0.1738,
      "step": 47390
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.003758196020498872,
      "learning_rate": 2.1569734613212877e-06,
      "loss": 0.1468,
      "step": 47400
    },
    {
      "epoch": 8.92,
      "grad_norm": 0.005495910998433828,
      "learning_rate": 2.153209109730849e-06,
      "loss": 0.0442,
      "step": 47410
    },
    {
      "epoch": 8.93,
      "grad_norm": 19.5859432220459,
      "learning_rate": 2.1494447581404106e-06,
      "loss": 0.1398,
      "step": 47420
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.015495622530579567,
      "learning_rate": 2.145680406549972e-06,
      "loss": 0.0523,
      "step": 47430
    },
    {
      "epoch": 8.93,
      "grad_norm": 38.67420959472656,
      "learning_rate": 2.1419160549595335e-06,
      "loss": 0.0921,
      "step": 47440
    },
    {
      "epoch": 8.93,
      "grad_norm": 11.818008422851562,
      "learning_rate": 2.138151703369095e-06,
      "loss": 0.1415,
      "step": 47450
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.0201152004301548,
      "learning_rate": 2.134387351778656e-06,
      "loss": 0.0445,
      "step": 47460
    },
    {
      "epoch": 8.93,
      "grad_norm": 11.320472717285156,
      "learning_rate": 2.130623000188218e-06,
      "loss": 0.2491,
      "step": 47470
    },
    {
      "epoch": 8.94,
      "grad_norm": 4.665805339813232,
      "learning_rate": 2.126858648597779e-06,
      "loss": 0.2942,
      "step": 47480
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.1507522165775299,
      "learning_rate": 2.123094297007341e-06,
      "loss": 0.0431,
      "step": 47490
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.008316931314766407,
      "learning_rate": 2.1193299454169023e-06,
      "loss": 0.1722,
      "step": 47500
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.6588090658187866,
      "learning_rate": 2.1155655938264637e-06,
      "loss": 0.0886,
      "step": 47510
    },
    {
      "epoch": 8.94,
      "grad_norm": 0.031189607456326485,
      "learning_rate": 2.111801242236025e-06,
      "loss": 0.2764,
      "step": 47520
    },
    {
      "epoch": 8.95,
      "grad_norm": 8.895737648010254,
      "learning_rate": 2.1080368906455862e-06,
      "loss": 0.1812,
      "step": 47530
    },
    {
      "epoch": 8.95,
      "grad_norm": 25.882984161376953,
      "learning_rate": 2.104272539055148e-06,
      "loss": 0.3084,
      "step": 47540
    },
    {
      "epoch": 8.95,
      "grad_norm": 4.4941911697387695,
      "learning_rate": 2.100508187464709e-06,
      "loss": 0.1027,
      "step": 47550
    },
    {
      "epoch": 8.95,
      "grad_norm": 0.008610288612544537,
      "learning_rate": 2.096743835874271e-06,
      "loss": 0.1269,
      "step": 47560
    },
    {
      "epoch": 8.95,
      "grad_norm": 27.184947967529297,
      "learning_rate": 2.092979484283832e-06,
      "loss": 0.2601,
      "step": 47570
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.00580189935863018,
      "learning_rate": 2.089215132693394e-06,
      "loss": 0.2305,
      "step": 47580
    },
    {
      "epoch": 8.96,
      "grad_norm": 1.897900938987732,
      "learning_rate": 2.085450781102955e-06,
      "loss": 0.2375,
      "step": 47590
    },
    {
      "epoch": 8.96,
      "grad_norm": 17.680147171020508,
      "learning_rate": 2.0816864295125165e-06,
      "loss": 0.1856,
      "step": 47600
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.009452820755541325,
      "learning_rate": 2.0779220779220784e-06,
      "loss": 0.1802,
      "step": 47610
    },
    {
      "epoch": 8.96,
      "grad_norm": 0.19355718791484833,
      "learning_rate": 2.0741577263316394e-06,
      "loss": 0.1233,
      "step": 47620
    },
    {
      "epoch": 8.96,
      "grad_norm": 11.466160774230957,
      "learning_rate": 2.0703933747412013e-06,
      "loss": 0.3482,
      "step": 47630
    },
    {
      "epoch": 8.97,
      "grad_norm": 11.396308898925781,
      "learning_rate": 2.0666290231507623e-06,
      "loss": 0.3488,
      "step": 47640
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.006663864012807608,
      "learning_rate": 2.062864671560324e-06,
      "loss": 0.0394,
      "step": 47650
    },
    {
      "epoch": 8.97,
      "grad_norm": 13.607287406921387,
      "learning_rate": 2.0591003199698852e-06,
      "loss": 0.2121,
      "step": 47660
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.08018065243959427,
      "learning_rate": 2.0553359683794467e-06,
      "loss": 0.1893,
      "step": 47670
    },
    {
      "epoch": 8.97,
      "grad_norm": 0.12857113778591156,
      "learning_rate": 2.051571616789008e-06,
      "loss": 0.0724,
      "step": 47680
    },
    {
      "epoch": 8.98,
      "grad_norm": 8.457551956176758,
      "learning_rate": 2.0478072651985696e-06,
      "loss": 0.0716,
      "step": 47690
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.007121187634766102,
      "learning_rate": 2.044042913608131e-06,
      "loss": 0.1761,
      "step": 47700
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.003906609490513802,
      "learning_rate": 2.0402785620176925e-06,
      "loss": 0.2633,
      "step": 47710
    },
    {
      "epoch": 8.98,
      "grad_norm": 0.00582027668133378,
      "learning_rate": 2.036514210427254e-06,
      "loss": 0.1506,
      "step": 47720
    },
    {
      "epoch": 8.98,
      "grad_norm": 17.560644149780273,
      "learning_rate": 2.0327498588368155e-06,
      "loss": 0.1054,
      "step": 47730
    },
    {
      "epoch": 8.99,
      "grad_norm": 11.83836841583252,
      "learning_rate": 2.028985507246377e-06,
      "loss": 0.1503,
      "step": 47740
    },
    {
      "epoch": 8.99,
      "grad_norm": 13.822301864624023,
      "learning_rate": 2.0252211556559384e-06,
      "loss": 0.0987,
      "step": 47750
    },
    {
      "epoch": 8.99,
      "grad_norm": 0.003371230326592922,
      "learning_rate": 2.0214568040655e-06,
      "loss": 0.1753,
      "step": 47760
    },
    {
      "epoch": 8.99,
      "grad_norm": 51.78420639038086,
      "learning_rate": 2.0176924524750613e-06,
      "loss": 0.1508,
      "step": 47770
    },
    {
      "epoch": 8.99,
      "grad_norm": 0.015421545132994652,
      "learning_rate": 2.0139281008846228e-06,
      "loss": 0.1412,
      "step": 47780
    },
    {
      "epoch": 8.99,
      "grad_norm": 46.56858825683594,
      "learning_rate": 2.0101637492941842e-06,
      "loss": 0.2422,
      "step": 47790
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.02757568471133709,
      "learning_rate": 2.0063993977037457e-06,
      "loss": 0.1982,
      "step": 47800
    },
    {
      "epoch": 9.0,
      "grad_norm": 7.849656581878662,
      "learning_rate": 2.002635046113307e-06,
      "loss": 0.2412,
      "step": 47810
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9910666666666667,
      "eval_loss": 0.03504617512226105,
      "eval_runtime": 26.9936,
      "eval_samples_per_second": 277.844,
      "eval_steps_per_second": 34.749,
      "step": 47817
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.004640958271920681,
      "learning_rate": 1.9988706945228686e-06,
      "loss": 0.2645,
      "step": 47820
    },
    {
      "epoch": 9.0,
      "grad_norm": 10.68244743347168,
      "learning_rate": 1.99510634293243e-06,
      "loss": 0.1922,
      "step": 47830
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.08029619604349136,
      "learning_rate": 1.9913419913419915e-06,
      "loss": 0.1353,
      "step": 47840
    },
    {
      "epoch": 9.01,
      "grad_norm": 15.328018188476562,
      "learning_rate": 1.987577639751553e-06,
      "loss": 0.2377,
      "step": 47850
    },
    {
      "epoch": 9.01,
      "grad_norm": 0.01895495131611824,
      "learning_rate": 1.9838132881611144e-06,
      "loss": 0.0869,
      "step": 47860
    },
    {
      "epoch": 9.01,
      "grad_norm": 17.427936553955078,
      "learning_rate": 1.980048936570676e-06,
      "loss": 0.2751,
      "step": 47870
    },
    {
      "epoch": 9.01,
      "grad_norm": 0.0033607317600399256,
      "learning_rate": 1.9762845849802374e-06,
      "loss": 0.0622,
      "step": 47880
    },
    {
      "epoch": 9.01,
      "grad_norm": 0.03174758329987526,
      "learning_rate": 1.972520233389799e-06,
      "loss": 0.1007,
      "step": 47890
    },
    {
      "epoch": 9.02,
      "grad_norm": 14.103028297424316,
      "learning_rate": 1.9687558817993603e-06,
      "loss": 0.2044,
      "step": 47900
    },
    {
      "epoch": 9.02,
      "grad_norm": 0.17187176644802094,
      "learning_rate": 1.9649915302089217e-06,
      "loss": 0.1578,
      "step": 47910
    },
    {
      "epoch": 9.02,
      "grad_norm": 9.40506362915039,
      "learning_rate": 1.961227178618483e-06,
      "loss": 0.1422,
      "step": 47920
    },
    {
      "epoch": 9.02,
      "grad_norm": 9.147233009338379,
      "learning_rate": 1.9574628270280447e-06,
      "loss": 0.1162,
      "step": 47930
    },
    {
      "epoch": 9.02,
      "grad_norm": 5.818715572357178,
      "learning_rate": 1.953698475437606e-06,
      "loss": 0.1892,
      "step": 47940
    },
    {
      "epoch": 9.03,
      "grad_norm": 6.80541467666626,
      "learning_rate": 1.9499341238471676e-06,
      "loss": 0.1598,
      "step": 47950
    },
    {
      "epoch": 9.03,
      "grad_norm": 9.4486083984375,
      "learning_rate": 1.9461697722567286e-06,
      "loss": 0.2799,
      "step": 47960
    },
    {
      "epoch": 9.03,
      "grad_norm": 0.003626194316893816,
      "learning_rate": 1.9424054206662905e-06,
      "loss": 0.2563,
      "step": 47970
    },
    {
      "epoch": 9.03,
      "grad_norm": 11.851634979248047,
      "learning_rate": 1.938641069075852e-06,
      "loss": 0.1941,
      "step": 47980
    },
    {
      "epoch": 9.03,
      "grad_norm": 8.557991027832031,
      "learning_rate": 1.9348767174854134e-06,
      "loss": 0.1685,
      "step": 47990
    },
    {
      "epoch": 9.03,
      "grad_norm": 11.349964141845703,
      "learning_rate": 1.931112365894975e-06,
      "loss": 0.2633,
      "step": 48000
    },
    {
      "epoch": 9.04,
      "grad_norm": 0.0050938986241817474,
      "learning_rate": 1.9273480143045364e-06,
      "loss": 0.0403,
      "step": 48010
    },
    {
      "epoch": 9.04,
      "grad_norm": 6.052948474884033,
      "learning_rate": 1.923583662714098e-06,
      "loss": 0.2877,
      "step": 48020
    },
    {
      "epoch": 9.04,
      "grad_norm": 0.04581421613693237,
      "learning_rate": 1.919819311123659e-06,
      "loss": 0.1231,
      "step": 48030
    },
    {
      "epoch": 9.04,
      "grad_norm": 26.62656593322754,
      "learning_rate": 1.9160549595332207e-06,
      "loss": 0.0541,
      "step": 48040
    },
    {
      "epoch": 9.04,
      "grad_norm": 0.16635794937610626,
      "learning_rate": 1.9122906079427818e-06,
      "loss": 0.1832,
      "step": 48050
    },
    {
      "epoch": 9.05,
      "grad_norm": 0.006478332448750734,
      "learning_rate": 1.9085262563523437e-06,
      "loss": 0.0118,
      "step": 48060
    },
    {
      "epoch": 9.05,
      "grad_norm": 0.017105581238865852,
      "learning_rate": 1.904761904761905e-06,
      "loss": 0.1048,
      "step": 48070
    },
    {
      "epoch": 9.05,
      "grad_norm": 0.5691583156585693,
      "learning_rate": 1.9009975531714664e-06,
      "loss": 0.0347,
      "step": 48080
    },
    {
      "epoch": 9.05,
      "grad_norm": 0.007283061742782593,
      "learning_rate": 1.8972332015810276e-06,
      "loss": 0.2108,
      "step": 48090
    },
    {
      "epoch": 9.05,
      "grad_norm": 31.505123138427734,
      "learning_rate": 1.8934688499905893e-06,
      "loss": 0.3265,
      "step": 48100
    },
    {
      "epoch": 9.06,
      "grad_norm": 6.634415626525879,
      "learning_rate": 1.8897044984001508e-06,
      "loss": 0.1965,
      "step": 48110
    },
    {
      "epoch": 9.06,
      "grad_norm": 0.011021504178643227,
      "learning_rate": 1.8859401468097122e-06,
      "loss": 0.0424,
      "step": 48120
    },
    {
      "epoch": 9.06,
      "grad_norm": 0.8117303252220154,
      "learning_rate": 1.8821757952192737e-06,
      "loss": 0.0976,
      "step": 48130
    },
    {
      "epoch": 9.06,
      "grad_norm": 7.521407604217529,
      "learning_rate": 1.8784114436288351e-06,
      "loss": 0.2084,
      "step": 48140
    },
    {
      "epoch": 9.06,
      "grad_norm": 2.6066713333129883,
      "learning_rate": 1.8746470920383966e-06,
      "loss": 0.0889,
      "step": 48150
    },
    {
      "epoch": 9.06,
      "grad_norm": 8.570279121398926,
      "learning_rate": 1.8708827404479578e-06,
      "loss": 0.2024,
      "step": 48160
    },
    {
      "epoch": 9.07,
      "grad_norm": 0.010919656604528427,
      "learning_rate": 1.8671183888575195e-06,
      "loss": 0.0592,
      "step": 48170
    },
    {
      "epoch": 9.07,
      "grad_norm": 1.7765733003616333,
      "learning_rate": 1.8633540372670808e-06,
      "loss": 0.1693,
      "step": 48180
    },
    {
      "epoch": 9.07,
      "grad_norm": 19.637720108032227,
      "learning_rate": 1.8595896856766424e-06,
      "loss": 0.2533,
      "step": 48190
    },
    {
      "epoch": 9.07,
      "grad_norm": 21.648691177368164,
      "learning_rate": 1.8558253340862037e-06,
      "loss": 0.1586,
      "step": 48200
    },
    {
      "epoch": 9.07,
      "grad_norm": 7.346878528594971,
      "learning_rate": 1.8520609824957654e-06,
      "loss": 0.0794,
      "step": 48210
    },
    {
      "epoch": 9.08,
      "grad_norm": 0.053309366106987,
      "learning_rate": 1.8482966309053266e-06,
      "loss": 0.1157,
      "step": 48220
    },
    {
      "epoch": 9.08,
      "grad_norm": 0.34328165650367737,
      "learning_rate": 1.844532279314888e-06,
      "loss": 0.2867,
      "step": 48230
    },
    {
      "epoch": 9.08,
      "grad_norm": 0.004103637300431728,
      "learning_rate": 1.8407679277244497e-06,
      "loss": 0.0573,
      "step": 48240
    },
    {
      "epoch": 9.08,
      "grad_norm": 0.0029933671467006207,
      "learning_rate": 1.837003576134011e-06,
      "loss": 0.2313,
      "step": 48250
    },
    {
      "epoch": 9.08,
      "grad_norm": 14.814998626708984,
      "learning_rate": 1.8332392245435727e-06,
      "loss": 0.2871,
      "step": 48260
    },
    {
      "epoch": 9.09,
      "grad_norm": 0.01592802256345749,
      "learning_rate": 1.829474872953134e-06,
      "loss": 0.3287,
      "step": 48270
    },
    {
      "epoch": 9.09,
      "grad_norm": 29.115724563598633,
      "learning_rate": 1.8257105213626956e-06,
      "loss": 0.1494,
      "step": 48280
    },
    {
      "epoch": 9.09,
      "grad_norm": 0.6358739733695984,
      "learning_rate": 1.8219461697722568e-06,
      "loss": 0.2015,
      "step": 48290
    },
    {
      "epoch": 9.09,
      "grad_norm": 0.005010218825191259,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.0738,
      "step": 48300
    },
    {
      "epoch": 9.09,
      "grad_norm": 0.08787321299314499,
      "learning_rate": 1.8144174665913798e-06,
      "loss": 0.2415,
      "step": 48310
    },
    {
      "epoch": 9.09,
      "grad_norm": 0.056446775794029236,
      "learning_rate": 1.8106531150009412e-06,
      "loss": 0.1585,
      "step": 48320
    },
    {
      "epoch": 9.1,
      "grad_norm": 4.861025333404541,
      "learning_rate": 1.8068887634105025e-06,
      "loss": 0.209,
      "step": 48330
    },
    {
      "epoch": 9.1,
      "grad_norm": 7.4785051345825195,
      "learning_rate": 1.8031244118200641e-06,
      "loss": 0.1957,
      "step": 48340
    },
    {
      "epoch": 9.1,
      "grad_norm": 1.6227164268493652,
      "learning_rate": 1.7993600602296258e-06,
      "loss": 0.0919,
      "step": 48350
    },
    {
      "epoch": 9.1,
      "grad_norm": 9.888586044311523,
      "learning_rate": 1.795595708639187e-06,
      "loss": 0.2067,
      "step": 48360
    },
    {
      "epoch": 9.1,
      "grad_norm": 0.005892313085496426,
      "learning_rate": 1.7918313570487485e-06,
      "loss": 0.1219,
      "step": 48370
    },
    {
      "epoch": 9.11,
      "grad_norm": 8.569474220275879,
      "learning_rate": 1.78806700545831e-06,
      "loss": 0.1779,
      "step": 48380
    },
    {
      "epoch": 9.11,
      "grad_norm": 0.021995043382048607,
      "learning_rate": 1.7843026538678714e-06,
      "loss": 0.2269,
      "step": 48390
    },
    {
      "epoch": 9.11,
      "grad_norm": 9.449252128601074,
      "learning_rate": 1.780538302277433e-06,
      "loss": 0.1374,
      "step": 48400
    },
    {
      "epoch": 9.11,
      "grad_norm": 0.6005410552024841,
      "learning_rate": 1.7767739506869944e-06,
      "loss": 0.1719,
      "step": 48410
    },
    {
      "epoch": 9.11,
      "grad_norm": 12.711328506469727,
      "learning_rate": 1.7730095990965556e-06,
      "loss": 0.1686,
      "step": 48420
    },
    {
      "epoch": 9.12,
      "grad_norm": 4.178927421569824,
      "learning_rate": 1.7692452475061173e-06,
      "loss": 0.137,
      "step": 48430
    },
    {
      "epoch": 9.12,
      "grad_norm": 9.198250770568848,
      "learning_rate": 1.7654808959156785e-06,
      "loss": 0.0278,
      "step": 48440
    },
    {
      "epoch": 9.12,
      "grad_norm": 0.004214898683130741,
      "learning_rate": 1.7617165443252402e-06,
      "loss": 0.1591,
      "step": 48450
    },
    {
      "epoch": 9.12,
      "grad_norm": 4.409656047821045,
      "learning_rate": 1.7579521927348015e-06,
      "loss": 0.106,
      "step": 48460
    },
    {
      "epoch": 9.12,
      "grad_norm": 0.30126702785491943,
      "learning_rate": 1.7541878411443631e-06,
      "loss": 0.0934,
      "step": 48470
    },
    {
      "epoch": 9.12,
      "grad_norm": 0.006339331157505512,
      "learning_rate": 1.7504234895539246e-06,
      "loss": 0.0668,
      "step": 48480
    },
    {
      "epoch": 9.13,
      "grad_norm": 4.0888566970825195,
      "learning_rate": 1.7466591379634858e-06,
      "loss": 0.0333,
      "step": 48490
    },
    {
      "epoch": 9.13,
      "grad_norm": 13.63100528717041,
      "learning_rate": 1.7428947863730475e-06,
      "loss": 0.1864,
      "step": 48500
    },
    {
      "epoch": 9.13,
      "grad_norm": 28.85260772705078,
      "learning_rate": 1.7391304347826088e-06,
      "loss": 0.1535,
      "step": 48510
    },
    {
      "epoch": 9.13,
      "grad_norm": 6.96646785736084,
      "learning_rate": 1.7353660831921704e-06,
      "loss": 0.1209,
      "step": 48520
    },
    {
      "epoch": 9.13,
      "grad_norm": 0.005375972483307123,
      "learning_rate": 1.7316017316017317e-06,
      "loss": 0.2367,
      "step": 48530
    },
    {
      "epoch": 9.14,
      "grad_norm": 6.989639759063721,
      "learning_rate": 1.7278373800112933e-06,
      "loss": 0.1139,
      "step": 48540
    },
    {
      "epoch": 9.14,
      "grad_norm": 0.0035316795110702515,
      "learning_rate": 1.7240730284208546e-06,
      "loss": 0.1868,
      "step": 48550
    },
    {
      "epoch": 9.14,
      "grad_norm": 9.89670181274414,
      "learning_rate": 1.720308676830416e-06,
      "loss": 0.1947,
      "step": 48560
    },
    {
      "epoch": 9.14,
      "grad_norm": 0.007525062188506126,
      "learning_rate": 1.7165443252399775e-06,
      "loss": 0.1528,
      "step": 48570
    },
    {
      "epoch": 9.14,
      "grad_norm": 0.004582991823554039,
      "learning_rate": 1.712779973649539e-06,
      "loss": 0.2047,
      "step": 48580
    },
    {
      "epoch": 9.15,
      "grad_norm": 14.427157402038574,
      "learning_rate": 1.7090156220591002e-06,
      "loss": 0.0542,
      "step": 48590
    },
    {
      "epoch": 9.15,
      "grad_norm": 0.003179016290232539,
      "learning_rate": 1.705251270468662e-06,
      "loss": 0.2009,
      "step": 48600
    },
    {
      "epoch": 9.15,
      "grad_norm": 0.6060037016868591,
      "learning_rate": 1.7014869188782236e-06,
      "loss": 0.1303,
      "step": 48610
    },
    {
      "epoch": 9.15,
      "grad_norm": 0.0028638881631195545,
      "learning_rate": 1.6977225672877848e-06,
      "loss": 0.1199,
      "step": 48620
    },
    {
      "epoch": 9.15,
      "grad_norm": 8.555558204650879,
      "learning_rate": 1.6939582156973463e-06,
      "loss": 0.1832,
      "step": 48630
    },
    {
      "epoch": 9.15,
      "grad_norm": 4.52678108215332,
      "learning_rate": 1.6901938641069077e-06,
      "loss": 0.3724,
      "step": 48640
    },
    {
      "epoch": 9.16,
      "grad_norm": 23.873172760009766,
      "learning_rate": 1.6864295125164692e-06,
      "loss": 0.1527,
      "step": 48650
    },
    {
      "epoch": 9.16,
      "grad_norm": 0.02074374631047249,
      "learning_rate": 1.6826651609260305e-06,
      "loss": 0.0581,
      "step": 48660
    },
    {
      "epoch": 9.16,
      "grad_norm": 7.155594348907471,
      "learning_rate": 1.6789008093355921e-06,
      "loss": 0.2501,
      "step": 48670
    },
    {
      "epoch": 9.16,
      "grad_norm": 36.21196746826172,
      "learning_rate": 1.6751364577451534e-06,
      "loss": 0.0746,
      "step": 48680
    },
    {
      "epoch": 9.16,
      "grad_norm": 1.0303316116333008,
      "learning_rate": 1.671372106154715e-06,
      "loss": 0.1672,
      "step": 48690
    },
    {
      "epoch": 9.17,
      "grad_norm": 0.6452359557151794,
      "learning_rate": 1.6676077545642763e-06,
      "loss": 0.3059,
      "step": 48700
    },
    {
      "epoch": 9.17,
      "grad_norm": 10.828946113586426,
      "learning_rate": 1.663843402973838e-06,
      "loss": 0.1629,
      "step": 48710
    },
    {
      "epoch": 9.17,
      "grad_norm": 1.0610060691833496,
      "learning_rate": 1.6600790513833994e-06,
      "loss": 0.1631,
      "step": 48720
    },
    {
      "epoch": 9.17,
      "grad_norm": 0.009052175097167492,
      "learning_rate": 1.6563146997929607e-06,
      "loss": 0.051,
      "step": 48730
    },
    {
      "epoch": 9.17,
      "grad_norm": 5.882680416107178,
      "learning_rate": 1.6525503482025224e-06,
      "loss": 0.1118,
      "step": 48740
    },
    {
      "epoch": 9.18,
      "grad_norm": 0.0040578986518085,
      "learning_rate": 1.6487859966120836e-06,
      "loss": 0.1412,
      "step": 48750
    },
    {
      "epoch": 9.18,
      "grad_norm": 0.017706580460071564,
      "learning_rate": 1.6450216450216453e-06,
      "loss": 0.0995,
      "step": 48760
    },
    {
      "epoch": 9.18,
      "grad_norm": 2.1415858268737793,
      "learning_rate": 1.6412572934312065e-06,
      "loss": 0.0621,
      "step": 48770
    },
    {
      "epoch": 9.18,
      "grad_norm": 6.102647304534912,
      "learning_rate": 1.6374929418407682e-06,
      "loss": 0.1649,
      "step": 48780
    },
    {
      "epoch": 9.18,
      "grad_norm": 0.0037551650311797857,
      "learning_rate": 1.6337285902503294e-06,
      "loss": 0.1647,
      "step": 48790
    },
    {
      "epoch": 9.19,
      "grad_norm": 0.15284423530101776,
      "learning_rate": 1.629964238659891e-06,
      "loss": 0.1899,
      "step": 48800
    },
    {
      "epoch": 9.19,
      "grad_norm": 2.7790141105651855,
      "learning_rate": 1.6261998870694524e-06,
      "loss": 0.2667,
      "step": 48810
    },
    {
      "epoch": 9.19,
      "grad_norm": 19.01610565185547,
      "learning_rate": 1.6224355354790138e-06,
      "loss": 0.1514,
      "step": 48820
    },
    {
      "epoch": 9.19,
      "grad_norm": 2.2549121379852295,
      "learning_rate": 1.6186711838885753e-06,
      "loss": 0.1137,
      "step": 48830
    },
    {
      "epoch": 9.19,
      "grad_norm": 0.0233939066529274,
      "learning_rate": 1.6149068322981367e-06,
      "loss": 0.1688,
      "step": 48840
    },
    {
      "epoch": 9.19,
      "grad_norm": 10.12087631225586,
      "learning_rate": 1.6111424807076984e-06,
      "loss": 0.109,
      "step": 48850
    },
    {
      "epoch": 9.2,
      "grad_norm": 0.7435272932052612,
      "learning_rate": 1.6073781291172597e-06,
      "loss": 0.0951,
      "step": 48860
    },
    {
      "epoch": 9.2,
      "grad_norm": 3.996717929840088,
      "learning_rate": 1.6036137775268213e-06,
      "loss": 0.1121,
      "step": 48870
    },
    {
      "epoch": 9.2,
      "grad_norm": 10.919347763061523,
      "learning_rate": 1.5998494259363826e-06,
      "loss": 0.3466,
      "step": 48880
    },
    {
      "epoch": 9.2,
      "grad_norm": 38.23072814941406,
      "learning_rate": 1.596085074345944e-06,
      "loss": 0.1006,
      "step": 48890
    },
    {
      "epoch": 9.2,
      "grad_norm": 11.066908836364746,
      "learning_rate": 1.5923207227555055e-06,
      "loss": 0.0717,
      "step": 48900
    },
    {
      "epoch": 9.21,
      "grad_norm": 29.55499839782715,
      "learning_rate": 1.588556371165067e-06,
      "loss": 0.1729,
      "step": 48910
    },
    {
      "epoch": 9.21,
      "grad_norm": 6.713967323303223,
      "learning_rate": 1.5847920195746282e-06,
      "loss": 0.2194,
      "step": 48920
    },
    {
      "epoch": 9.21,
      "grad_norm": 1.155410647392273,
      "learning_rate": 1.5810276679841899e-06,
      "loss": 0.2505,
      "step": 48930
    },
    {
      "epoch": 9.21,
      "grad_norm": 0.03431905806064606,
      "learning_rate": 1.5772633163937511e-06,
      "loss": 0.1576,
      "step": 48940
    },
    {
      "epoch": 9.21,
      "grad_norm": 0.0038003171794116497,
      "learning_rate": 1.5734989648033128e-06,
      "loss": 0.1722,
      "step": 48950
    },
    {
      "epoch": 9.22,
      "grad_norm": 17.588560104370117,
      "learning_rate": 1.5697346132128743e-06,
      "loss": 0.2669,
      "step": 48960
    },
    {
      "epoch": 9.22,
      "grad_norm": 0.014295457862317562,
      "learning_rate": 1.5659702616224357e-06,
      "loss": 0.0449,
      "step": 48970
    },
    {
      "epoch": 9.22,
      "grad_norm": 6.059362888336182,
      "learning_rate": 1.5622059100319972e-06,
      "loss": 0.2459,
      "step": 48980
    },
    {
      "epoch": 9.22,
      "grad_norm": 0.23920145630836487,
      "learning_rate": 1.5584415584415584e-06,
      "loss": 0.1983,
      "step": 48990
    },
    {
      "epoch": 9.22,
      "grad_norm": 0.007495987694710493,
      "learning_rate": 1.5546772068511201e-06,
      "loss": 0.2355,
      "step": 49000
    },
    {
      "epoch": 9.22,
      "grad_norm": 5.33493185043335,
      "learning_rate": 1.5509128552606814e-06,
      "loss": 0.1973,
      "step": 49010
    },
    {
      "epoch": 9.23,
      "grad_norm": 33.67094802856445,
      "learning_rate": 1.547148503670243e-06,
      "loss": 0.1861,
      "step": 49020
    },
    {
      "epoch": 9.23,
      "grad_norm": 21.95941925048828,
      "learning_rate": 1.5433841520798043e-06,
      "loss": 0.3058,
      "step": 49030
    },
    {
      "epoch": 9.23,
      "grad_norm": 12.926125526428223,
      "learning_rate": 1.539619800489366e-06,
      "loss": 0.3159,
      "step": 49040
    },
    {
      "epoch": 9.23,
      "grad_norm": 0.018391737714409828,
      "learning_rate": 1.5358554488989272e-06,
      "loss": 0.2065,
      "step": 49050
    },
    {
      "epoch": 9.23,
      "grad_norm": 4.158997058868408,
      "learning_rate": 1.5320910973084887e-06,
      "loss": 0.1346,
      "step": 49060
    },
    {
      "epoch": 9.24,
      "grad_norm": 0.5835167765617371,
      "learning_rate": 1.5283267457180501e-06,
      "loss": 0.0145,
      "step": 49070
    },
    {
      "epoch": 9.24,
      "grad_norm": 0.004218466579914093,
      "learning_rate": 1.5245623941276116e-06,
      "loss": 0.15,
      "step": 49080
    },
    {
      "epoch": 9.24,
      "grad_norm": 0.912668764591217,
      "learning_rate": 1.5207980425371733e-06,
      "loss": 0.3786,
      "step": 49090
    },
    {
      "epoch": 9.24,
      "grad_norm": 9.824188232421875,
      "learning_rate": 1.5170336909467345e-06,
      "loss": 0.3177,
      "step": 49100
    },
    {
      "epoch": 9.24,
      "grad_norm": 2.483562469482422,
      "learning_rate": 1.5132693393562962e-06,
      "loss": 0.4222,
      "step": 49110
    },
    {
      "epoch": 9.25,
      "grad_norm": 10.508928298950195,
      "learning_rate": 1.5095049877658574e-06,
      "loss": 0.0972,
      "step": 49120
    },
    {
      "epoch": 9.25,
      "grad_norm": 25.997400283813477,
      "learning_rate": 1.505740636175419e-06,
      "loss": 0.0592,
      "step": 49130
    },
    {
      "epoch": 9.25,
      "grad_norm": 19.286235809326172,
      "learning_rate": 1.5019762845849804e-06,
      "loss": 0.2108,
      "step": 49140
    },
    {
      "epoch": 9.25,
      "grad_norm": 0.006492584012448788,
      "learning_rate": 1.4982119329945418e-06,
      "loss": 0.131,
      "step": 49150
    },
    {
      "epoch": 9.25,
      "grad_norm": 0.02029990591108799,
      "learning_rate": 1.4944475814041033e-06,
      "loss": 0.0742,
      "step": 49160
    },
    {
      "epoch": 9.25,
      "grad_norm": 7.444857597351074,
      "learning_rate": 1.4906832298136647e-06,
      "loss": 0.1238,
      "step": 49170
    },
    {
      "epoch": 9.26,
      "grad_norm": 14.532998085021973,
      "learning_rate": 1.486918878223226e-06,
      "loss": 0.1678,
      "step": 49180
    },
    {
      "epoch": 9.26,
      "grad_norm": 0.015678424388170242,
      "learning_rate": 1.4831545266327877e-06,
      "loss": 0.1815,
      "step": 49190
    },
    {
      "epoch": 9.26,
      "grad_norm": 19.47397804260254,
      "learning_rate": 1.479390175042349e-06,
      "loss": 0.2612,
      "step": 49200
    },
    {
      "epoch": 9.26,
      "grad_norm": 0.05409480631351471,
      "learning_rate": 1.4756258234519106e-06,
      "loss": 0.1281,
      "step": 49210
    },
    {
      "epoch": 9.26,
      "grad_norm": 1.8059682846069336,
      "learning_rate": 1.471861471861472e-06,
      "loss": 0.0624,
      "step": 49220
    },
    {
      "epoch": 9.27,
      "grad_norm": 0.025014258921146393,
      "learning_rate": 1.4680971202710335e-06,
      "loss": 0.1748,
      "step": 49230
    },
    {
      "epoch": 9.27,
      "grad_norm": 0.002997885225340724,
      "learning_rate": 1.464332768680595e-06,
      "loss": 0.2587,
      "step": 49240
    },
    {
      "epoch": 9.27,
      "grad_norm": 0.03804245591163635,
      "learning_rate": 1.4605684170901562e-06,
      "loss": 0.0594,
      "step": 49250
    },
    {
      "epoch": 9.27,
      "grad_norm": 1.2966951131820679,
      "learning_rate": 1.4568040654997179e-06,
      "loss": 0.2569,
      "step": 49260
    },
    {
      "epoch": 9.27,
      "grad_norm": 22.892065048217773,
      "learning_rate": 1.4530397139092791e-06,
      "loss": 0.1701,
      "step": 49270
    },
    {
      "epoch": 9.28,
      "grad_norm": 5.166015625,
      "learning_rate": 1.4492753623188408e-06,
      "loss": 0.0167,
      "step": 49280
    },
    {
      "epoch": 9.28,
      "grad_norm": 22.783777236938477,
      "learning_rate": 1.445511010728402e-06,
      "loss": 0.1226,
      "step": 49290
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.0027043980080634356,
      "learning_rate": 1.4417466591379637e-06,
      "loss": 0.0532,
      "step": 49300
    },
    {
      "epoch": 9.28,
      "grad_norm": 2.246042490005493,
      "learning_rate": 1.437982307547525e-06,
      "loss": 0.3735,
      "step": 49310
    },
    {
      "epoch": 9.28,
      "grad_norm": 10.176612854003906,
      "learning_rate": 1.4342179559570864e-06,
      "loss": 0.2525,
      "step": 49320
    },
    {
      "epoch": 9.28,
      "grad_norm": 0.31273147463798523,
      "learning_rate": 1.4304536043666481e-06,
      "loss": 0.1622,
      "step": 49330
    },
    {
      "epoch": 9.29,
      "grad_norm": 0.046601302921772,
      "learning_rate": 1.4266892527762094e-06,
      "loss": 0.1761,
      "step": 49340
    },
    {
      "epoch": 9.29,
      "grad_norm": 6.128500938415527,
      "learning_rate": 1.422924901185771e-06,
      "loss": 0.1705,
      "step": 49350
    },
    {
      "epoch": 9.29,
      "grad_norm": 0.013611139729619026,
      "learning_rate": 1.4191605495953323e-06,
      "loss": 0.2087,
      "step": 49360
    },
    {
      "epoch": 9.29,
      "grad_norm": 0.10704396665096283,
      "learning_rate": 1.415396198004894e-06,
      "loss": 0.0525,
      "step": 49370
    },
    {
      "epoch": 9.29,
      "grad_norm": 0.38533052802085876,
      "learning_rate": 1.4116318464144552e-06,
      "loss": 0.0749,
      "step": 49380
    },
    {
      "epoch": 9.3,
      "grad_norm": 5.009214401245117,
      "learning_rate": 1.4078674948240167e-06,
      "loss": 0.2508,
      "step": 49390
    },
    {
      "epoch": 9.3,
      "grad_norm": 0.04618481174111366,
      "learning_rate": 1.4041031432335781e-06,
      "loss": 0.2825,
      "step": 49400
    },
    {
      "epoch": 9.3,
      "grad_norm": 0.004270387347787619,
      "learning_rate": 1.4003387916431396e-06,
      "loss": 0.0415,
      "step": 49410
    },
    {
      "epoch": 9.3,
      "grad_norm": 3.027621269226074,
      "learning_rate": 1.3965744400527008e-06,
      "loss": 0.0915,
      "step": 49420
    },
    {
      "epoch": 9.3,
      "grad_norm": 5.719045639038086,
      "learning_rate": 1.3928100884622625e-06,
      "loss": 0.1118,
      "step": 49430
    },
    {
      "epoch": 9.31,
      "grad_norm": 12.184414863586426,
      "learning_rate": 1.3890457368718238e-06,
      "loss": 0.1467,
      "step": 49440
    },
    {
      "epoch": 9.31,
      "grad_norm": 8.88697338104248,
      "learning_rate": 1.3852813852813854e-06,
      "loss": 0.1369,
      "step": 49450
    },
    {
      "epoch": 9.31,
      "grad_norm": 21.216154098510742,
      "learning_rate": 1.3815170336909469e-06,
      "loss": 0.2886,
      "step": 49460
    },
    {
      "epoch": 9.31,
      "grad_norm": 28.933076858520508,
      "learning_rate": 1.3777526821005083e-06,
      "loss": 0.1965,
      "step": 49470
    },
    {
      "epoch": 9.31,
      "grad_norm": 0.003988806623965502,
      "learning_rate": 1.3739883305100698e-06,
      "loss": 0.1452,
      "step": 49480
    },
    {
      "epoch": 9.31,
      "grad_norm": 5.434657573699951,
      "learning_rate": 1.370223978919631e-06,
      "loss": 0.1927,
      "step": 49490
    },
    {
      "epoch": 9.32,
      "grad_norm": 22.743406295776367,
      "learning_rate": 1.3664596273291927e-06,
      "loss": 0.1979,
      "step": 49500
    },
    {
      "epoch": 9.32,
      "grad_norm": 1.0955983400344849,
      "learning_rate": 1.362695275738754e-06,
      "loss": 0.1662,
      "step": 49510
    },
    {
      "epoch": 9.32,
      "grad_norm": 0.0766681358218193,
      "learning_rate": 1.3589309241483157e-06,
      "loss": 0.0688,
      "step": 49520
    },
    {
      "epoch": 9.32,
      "grad_norm": 0.008558296598494053,
      "learning_rate": 1.355166572557877e-06,
      "loss": 0.2808,
      "step": 49530
    },
    {
      "epoch": 9.32,
      "grad_norm": 0.04636925458908081,
      "learning_rate": 1.3514022209674386e-06,
      "loss": 0.0217,
      "step": 49540
    },
    {
      "epoch": 9.33,
      "grad_norm": 7.185703277587891,
      "learning_rate": 1.3476378693769998e-06,
      "loss": 0.1822,
      "step": 49550
    },
    {
      "epoch": 9.33,
      "grad_norm": 0.018046174198389053,
      "learning_rate": 1.3438735177865615e-06,
      "loss": 0.1938,
      "step": 49560
    },
    {
      "epoch": 9.33,
      "grad_norm": 0.009080019779503345,
      "learning_rate": 1.3401091661961227e-06,
      "loss": 0.1291,
      "step": 49570
    },
    {
      "epoch": 9.33,
      "grad_norm": 0.07508441060781479,
      "learning_rate": 1.3363448146056842e-06,
      "loss": 0.2514,
      "step": 49580
    },
    {
      "epoch": 9.33,
      "grad_norm": 23.7309513092041,
      "learning_rate": 1.3325804630152459e-06,
      "loss": 0.1583,
      "step": 49590
    },
    {
      "epoch": 9.34,
      "grad_norm": 0.01442288514226675,
      "learning_rate": 1.3288161114248071e-06,
      "loss": 0.0395,
      "step": 49600
    },
    {
      "epoch": 9.34,
      "grad_norm": 3.1011149883270264,
      "learning_rate": 1.3250517598343688e-06,
      "loss": 0.2429,
      "step": 49610
    },
    {
      "epoch": 9.34,
      "grad_norm": 0.0179927796125412,
      "learning_rate": 1.32128740824393e-06,
      "loss": 0.1883,
      "step": 49620
    },
    {
      "epoch": 9.34,
      "grad_norm": 0.014444788917899132,
      "learning_rate": 1.3175230566534917e-06,
      "loss": 0.1264,
      "step": 49630
    },
    {
      "epoch": 9.34,
      "grad_norm": 5.354752540588379,
      "learning_rate": 1.313758705063053e-06,
      "loss": 0.0372,
      "step": 49640
    },
    {
      "epoch": 9.35,
      "grad_norm": 0.05424873158335686,
      "learning_rate": 1.3099943534726144e-06,
      "loss": 0.1853,
      "step": 49650
    },
    {
      "epoch": 9.35,
      "grad_norm": 24.28473663330078,
      "learning_rate": 1.3062300018821759e-06,
      "loss": 0.0994,
      "step": 49660
    },
    {
      "epoch": 9.35,
      "grad_norm": 1.3614308834075928,
      "learning_rate": 1.3024656502917373e-06,
      "loss": 0.1133,
      "step": 49670
    },
    {
      "epoch": 9.35,
      "grad_norm": 0.3897532522678375,
      "learning_rate": 1.2987012987012986e-06,
      "loss": 0.1837,
      "step": 49680
    },
    {
      "epoch": 9.35,
      "grad_norm": 0.15229752659797668,
      "learning_rate": 1.2949369471108603e-06,
      "loss": 0.1784,
      "step": 49690
    },
    {
      "epoch": 9.35,
      "grad_norm": 2.612488031387329,
      "learning_rate": 1.291172595520422e-06,
      "loss": 0.041,
      "step": 49700
    },
    {
      "epoch": 9.36,
      "grad_norm": 8.151073455810547,
      "learning_rate": 1.2874082439299832e-06,
      "loss": 0.074,
      "step": 49710
    },
    {
      "epoch": 9.36,
      "grad_norm": 0.014114074409008026,
      "learning_rate": 1.2836438923395447e-06,
      "loss": 0.1503,
      "step": 49720
    },
    {
      "epoch": 9.36,
      "grad_norm": 0.9798872470855713,
      "learning_rate": 1.2798795407491061e-06,
      "loss": 0.2701,
      "step": 49730
    },
    {
      "epoch": 9.36,
      "grad_norm": 30.948965072631836,
      "learning_rate": 1.2761151891586676e-06,
      "loss": 0.2419,
      "step": 49740
    },
    {
      "epoch": 9.36,
      "grad_norm": 0.006186028476804495,
      "learning_rate": 1.2723508375682288e-06,
      "loss": 0.113,
      "step": 49750
    },
    {
      "epoch": 9.37,
      "grad_norm": 7.58191442489624,
      "learning_rate": 1.2685864859777905e-06,
      "loss": 0.1978,
      "step": 49760
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.008680231869220734,
      "learning_rate": 1.2648221343873517e-06,
      "loss": 0.1253,
      "step": 49770
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.0067359572276473045,
      "learning_rate": 1.2610577827969134e-06,
      "loss": 0.3312,
      "step": 49780
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.003633724292740226,
      "learning_rate": 1.2572934312064747e-06,
      "loss": 0.0611,
      "step": 49790
    },
    {
      "epoch": 9.37,
      "grad_norm": 0.13362182676792145,
      "learning_rate": 1.2535290796160363e-06,
      "loss": 0.2222,
      "step": 49800
    },
    {
      "epoch": 9.38,
      "grad_norm": 0.003943243995308876,
      "learning_rate": 1.2497647280255978e-06,
      "loss": 0.4244,
      "step": 49810
    },
    {
      "epoch": 9.38,
      "grad_norm": 29.752687454223633,
      "learning_rate": 1.246000376435159e-06,
      "loss": 0.2152,
      "step": 49820
    },
    {
      "epoch": 9.38,
      "grad_norm": 0.01345011405646801,
      "learning_rate": 1.2422360248447205e-06,
      "loss": 0.1275,
      "step": 49830
    },
    {
      "epoch": 9.38,
      "grad_norm": 0.05559426173567772,
      "learning_rate": 1.238471673254282e-06,
      "loss": 0.1979,
      "step": 49840
    },
    {
      "epoch": 9.38,
      "grad_norm": 0.011537078768014908,
      "learning_rate": 1.2347073216638434e-06,
      "loss": 0.0426,
      "step": 49850
    },
    {
      "epoch": 9.38,
      "grad_norm": 5.634771823883057,
      "learning_rate": 1.2309429700734049e-06,
      "loss": 0.3468,
      "step": 49860
    },
    {
      "epoch": 9.39,
      "grad_norm": 0.797957181930542,
      "learning_rate": 1.2271786184829664e-06,
      "loss": 0.1726,
      "step": 49870
    },
    {
      "epoch": 9.39,
      "grad_norm": 9.324555397033691,
      "learning_rate": 1.223414266892528e-06,
      "loss": 0.1118,
      "step": 49880
    },
    {
      "epoch": 9.39,
      "grad_norm": 9.931642532348633,
      "learning_rate": 1.2196499153020893e-06,
      "loss": 0.1549,
      "step": 49890
    },
    {
      "epoch": 9.39,
      "grad_norm": 0.7265287637710571,
      "learning_rate": 1.2158855637116507e-06,
      "loss": 0.1494,
      "step": 49900
    },
    {
      "epoch": 9.39,
      "grad_norm": 0.006699044723063707,
      "learning_rate": 1.2121212121212122e-06,
      "loss": 0.0992,
      "step": 49910
    },
    {
      "epoch": 9.4,
      "grad_norm": 0.366902619600296,
      "learning_rate": 1.2083568605307737e-06,
      "loss": 0.0977,
      "step": 49920
    },
    {
      "epoch": 9.4,
      "grad_norm": 31.29306411743164,
      "learning_rate": 1.2045925089403351e-06,
      "loss": 0.1368,
      "step": 49930
    },
    {
      "epoch": 9.4,
      "grad_norm": 0.013638084754347801,
      "learning_rate": 1.2008281573498966e-06,
      "loss": 0.1185,
      "step": 49940
    },
    {
      "epoch": 9.4,
      "grad_norm": 0.37019070982933044,
      "learning_rate": 1.197063805759458e-06,
      "loss": 0.1058,
      "step": 49950
    },
    {
      "epoch": 9.4,
      "grad_norm": 0.0033348985016345978,
      "learning_rate": 1.1932994541690195e-06,
      "loss": 0.2326,
      "step": 49960
    },
    {
      "epoch": 9.41,
      "grad_norm": 4.079520225524902,
      "learning_rate": 1.189535102578581e-06,
      "loss": 0.2197,
      "step": 49970
    },
    {
      "epoch": 9.41,
      "grad_norm": 0.0073106358759105206,
      "learning_rate": 1.1857707509881424e-06,
      "loss": 0.1406,
      "step": 49980
    },
    {
      "epoch": 9.41,
      "grad_norm": 0.005739548243582249,
      "learning_rate": 1.1820063993977039e-06,
      "loss": 0.1482,
      "step": 49990
    },
    {
      "epoch": 9.41,
      "grad_norm": 16.797861099243164,
      "learning_rate": 1.1782420478072651e-06,
      "loss": 0.0474,
      "step": 50000
    },
    {
      "epoch": 9.41,
      "grad_norm": 0.03905607387423515,
      "learning_rate": 1.1744776962168268e-06,
      "loss": 0.2154,
      "step": 50010
    },
    {
      "epoch": 9.41,
      "grad_norm": 9.904181480407715,
      "learning_rate": 1.1707133446263883e-06,
      "loss": 0.1878,
      "step": 50020
    },
    {
      "epoch": 9.42,
      "grad_norm": 0.5066874027252197,
      "learning_rate": 1.1669489930359497e-06,
      "loss": 0.1394,
      "step": 50030
    },
    {
      "epoch": 9.42,
      "grad_norm": 0.4699447453022003,
      "learning_rate": 1.1631846414455112e-06,
      "loss": 0.126,
      "step": 50040
    },
    {
      "epoch": 9.42,
      "grad_norm": 0.042065080255270004,
      "learning_rate": 1.1594202898550726e-06,
      "loss": 0.1301,
      "step": 50050
    },
    {
      "epoch": 9.42,
      "grad_norm": 10.649992942810059,
      "learning_rate": 1.155655938264634e-06,
      "loss": 0.1723,
      "step": 50060
    },
    {
      "epoch": 9.42,
      "grad_norm": 0.08726591616868973,
      "learning_rate": 1.1518915866741954e-06,
      "loss": 0.0221,
      "step": 50070
    },
    {
      "epoch": 9.43,
      "grad_norm": 0.17898158729076385,
      "learning_rate": 1.1481272350837568e-06,
      "loss": 0.0697,
      "step": 50080
    },
    {
      "epoch": 9.43,
      "grad_norm": 10.242304801940918,
      "learning_rate": 1.1443628834933183e-06,
      "loss": 0.2728,
      "step": 50090
    },
    {
      "epoch": 9.43,
      "grad_norm": 38.10251235961914,
      "learning_rate": 1.1405985319028797e-06,
      "loss": 0.0946,
      "step": 50100
    },
    {
      "epoch": 9.43,
      "grad_norm": 0.007783820852637291,
      "learning_rate": 1.1368341803124412e-06,
      "loss": 0.0923,
      "step": 50110
    },
    {
      "epoch": 9.43,
      "grad_norm": 0.0033500604331493378,
      "learning_rate": 1.1330698287220027e-06,
      "loss": 0.1469,
      "step": 50120
    },
    {
      "epoch": 9.44,
      "grad_norm": 1.0940799713134766,
      "learning_rate": 1.1293054771315643e-06,
      "loss": 0.1244,
      "step": 50130
    },
    {
      "epoch": 9.44,
      "grad_norm": 13.593926429748535,
      "learning_rate": 1.1255411255411256e-06,
      "loss": 0.1965,
      "step": 50140
    },
    {
      "epoch": 9.44,
      "grad_norm": 0.004496280569583178,
      "learning_rate": 1.121776773950687e-06,
      "loss": 0.1533,
      "step": 50150
    },
    {
      "epoch": 9.44,
      "grad_norm": 14.750734329223633,
      "learning_rate": 1.1180124223602485e-06,
      "loss": 0.0817,
      "step": 50160
    },
    {
      "epoch": 9.44,
      "grad_norm": 0.10755278170108795,
      "learning_rate": 1.11424807076981e-06,
      "loss": 0.0592,
      "step": 50170
    },
    {
      "epoch": 9.44,
      "grad_norm": 4.955772399902344,
      "learning_rate": 1.1104837191793714e-06,
      "loss": 0.1231,
      "step": 50180
    },
    {
      "epoch": 9.45,
      "grad_norm": 0.29413166642189026,
      "learning_rate": 1.1067193675889329e-06,
      "loss": 0.1379,
      "step": 50190
    },
    {
      "epoch": 9.45,
      "grad_norm": 1.0984842777252197,
      "learning_rate": 1.1029550159984943e-06,
      "loss": 0.1228,
      "step": 50200
    },
    {
      "epoch": 9.45,
      "grad_norm": 10.957953453063965,
      "learning_rate": 1.0991906644080558e-06,
      "loss": 0.0743,
      "step": 50210
    },
    {
      "epoch": 9.45,
      "grad_norm": 0.0077149225398898125,
      "learning_rate": 1.0954263128176173e-06,
      "loss": 0.0237,
      "step": 50220
    },
    {
      "epoch": 9.45,
      "grad_norm": 0.6509040594100952,
      "learning_rate": 1.0916619612271787e-06,
      "loss": 0.3099,
      "step": 50230
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.19901898503303528,
      "learning_rate": 1.0878976096367402e-06,
      "loss": 0.0645,
      "step": 50240
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.023920951411128044,
      "learning_rate": 1.0841332580463016e-06,
      "loss": 0.1552,
      "step": 50250
    },
    {
      "epoch": 9.46,
      "grad_norm": 6.93363618850708,
      "learning_rate": 1.0803689064558631e-06,
      "loss": 0.196,
      "step": 50260
    },
    {
      "epoch": 9.46,
      "grad_norm": 0.03984972462058067,
      "learning_rate": 1.0766045548654246e-06,
      "loss": 0.1239,
      "step": 50270
    },
    {
      "epoch": 9.46,
      "grad_norm": 6.994830131530762,
      "learning_rate": 1.072840203274986e-06,
      "loss": 0.1865,
      "step": 50280
    },
    {
      "epoch": 9.47,
      "grad_norm": 0.9028927087783813,
      "learning_rate": 1.0690758516845475e-06,
      "loss": 0.0045,
      "step": 50290
    },
    {
      "epoch": 9.47,
      "grad_norm": 19.097110748291016,
      "learning_rate": 1.065311500094109e-06,
      "loss": 0.2255,
      "step": 50300
    },
    {
      "epoch": 9.47,
      "grad_norm": 0.0035785511136054993,
      "learning_rate": 1.0615471485036704e-06,
      "loss": 0.1536,
      "step": 50310
    },
    {
      "epoch": 9.47,
      "grad_norm": 0.004918965045362711,
      "learning_rate": 1.0577827969132319e-06,
      "loss": 0.3612,
      "step": 50320
    },
    {
      "epoch": 9.47,
      "grad_norm": 0.011572405695915222,
      "learning_rate": 1.0540184453227931e-06,
      "loss": 0.174,
      "step": 50330
    },
    {
      "epoch": 9.47,
      "grad_norm": 0.47949331998825073,
      "learning_rate": 1.0502540937323546e-06,
      "loss": 0.1387,
      "step": 50340
    },
    {
      "epoch": 9.48,
      "grad_norm": 0.7120999693870544,
      "learning_rate": 1.046489742141916e-06,
      "loss": 0.2182,
      "step": 50350
    },
    {
      "epoch": 9.48,
      "grad_norm": 0.015505412593483925,
      "learning_rate": 1.0427253905514775e-06,
      "loss": 0.0099,
      "step": 50360
    },
    {
      "epoch": 9.48,
      "grad_norm": 1.0895874500274658,
      "learning_rate": 1.0389610389610392e-06,
      "loss": 0.1563,
      "step": 50370
    },
    {
      "epoch": 9.48,
      "grad_norm": 14.657905578613281,
      "learning_rate": 1.0351966873706006e-06,
      "loss": 0.1918,
      "step": 50380
    },
    {
      "epoch": 9.48,
      "grad_norm": 0.3927984833717346,
      "learning_rate": 1.031432335780162e-06,
      "loss": 0.1652,
      "step": 50390
    },
    {
      "epoch": 9.49,
      "grad_norm": 0.43692031502723694,
      "learning_rate": 1.0276679841897233e-06,
      "loss": 0.1009,
      "step": 50400
    },
    {
      "epoch": 9.49,
      "grad_norm": 1.0691211223602295,
      "learning_rate": 1.0239036325992848e-06,
      "loss": 0.2564,
      "step": 50410
    },
    {
      "epoch": 9.49,
      "grad_norm": 17.350324630737305,
      "learning_rate": 1.0201392810088463e-06,
      "loss": 0.1796,
      "step": 50420
    },
    {
      "epoch": 9.49,
      "grad_norm": 0.039267878979444504,
      "learning_rate": 1.0163749294184077e-06,
      "loss": 0.221,
      "step": 50430
    },
    {
      "epoch": 9.49,
      "grad_norm": 14.294244766235352,
      "learning_rate": 1.0126105778279692e-06,
      "loss": 0.2201,
      "step": 50440
    },
    {
      "epoch": 9.5,
      "grad_norm": 6.4382829666137695,
      "learning_rate": 1.0088462262375306e-06,
      "loss": 0.1404,
      "step": 50450
    },
    {
      "epoch": 9.5,
      "grad_norm": 18.890953063964844,
      "learning_rate": 1.0050818746470921e-06,
      "loss": 0.1059,
      "step": 50460
    },
    {
      "epoch": 9.5,
      "grad_norm": 11.253399848937988,
      "learning_rate": 1.0013175230566536e-06,
      "loss": 0.1933,
      "step": 50470
    },
    {
      "epoch": 9.5,
      "grad_norm": 0.7060115337371826,
      "learning_rate": 9.97553171466215e-07,
      "loss": 0.0639,
      "step": 50480
    },
    {
      "epoch": 9.5,
      "grad_norm": 10.807540893554688,
      "learning_rate": 9.937888198757765e-07,
      "loss": 0.2468,
      "step": 50490
    },
    {
      "epoch": 9.5,
      "grad_norm": 0.011891942471265793,
      "learning_rate": 9.90024468285338e-07,
      "loss": 0.1363,
      "step": 50500
    },
    {
      "epoch": 9.51,
      "grad_norm": 10.976309776306152,
      "learning_rate": 9.862601166948994e-07,
      "loss": 0.2212,
      "step": 50510
    },
    {
      "epoch": 9.51,
      "grad_norm": 12.86689567565918,
      "learning_rate": 9.824957651044609e-07,
      "loss": 0.0503,
      "step": 50520
    },
    {
      "epoch": 9.51,
      "grad_norm": 1.714057207107544,
      "learning_rate": 9.787314135140223e-07,
      "loss": 0.2065,
      "step": 50530
    },
    {
      "epoch": 9.51,
      "grad_norm": 16.20371437072754,
      "learning_rate": 9.749670619235838e-07,
      "loss": 0.1964,
      "step": 50540
    },
    {
      "epoch": 9.51,
      "grad_norm": 20.7742919921875,
      "learning_rate": 9.712027103331453e-07,
      "loss": 0.1582,
      "step": 50550
    },
    {
      "epoch": 9.52,
      "grad_norm": 2.2414963245391846,
      "learning_rate": 9.674383587427067e-07,
      "loss": 0.1347,
      "step": 50560
    },
    {
      "epoch": 9.52,
      "grad_norm": 0.007517498917877674,
      "learning_rate": 9.636740071522682e-07,
      "loss": 0.2661,
      "step": 50570
    },
    {
      "epoch": 9.52,
      "grad_norm": 0.008685542270541191,
      "learning_rate": 9.599096555618294e-07,
      "loss": 0.1882,
      "step": 50580
    },
    {
      "epoch": 9.52,
      "grad_norm": 16.764934539794922,
      "learning_rate": 9.561453039713909e-07,
      "loss": 0.2041,
      "step": 50590
    },
    {
      "epoch": 9.52,
      "grad_norm": 3.8847382068634033,
      "learning_rate": 9.523809523809525e-07,
      "loss": 0.2251,
      "step": 50600
    },
    {
      "epoch": 9.53,
      "grad_norm": 16.056352615356445,
      "learning_rate": 9.486166007905138e-07,
      "loss": 0.2243,
      "step": 50610
    },
    {
      "epoch": 9.53,
      "grad_norm": 7.7409820556640625,
      "learning_rate": 9.448522492000754e-07,
      "loss": 0.1496,
      "step": 50620
    },
    {
      "epoch": 9.53,
      "grad_norm": 0.051682520657777786,
      "learning_rate": 9.410878976096368e-07,
      "loss": 0.1441,
      "step": 50630
    },
    {
      "epoch": 9.53,
      "grad_norm": 42.59306335449219,
      "learning_rate": 9.373235460191983e-07,
      "loss": 0.2264,
      "step": 50640
    },
    {
      "epoch": 9.53,
      "grad_norm": 12.349274635314941,
      "learning_rate": 9.335591944287598e-07,
      "loss": 0.1528,
      "step": 50650
    },
    {
      "epoch": 9.54,
      "grad_norm": 0.0068712178617715836,
      "learning_rate": 9.297948428383212e-07,
      "loss": 0.0964,
      "step": 50660
    },
    {
      "epoch": 9.54,
      "grad_norm": 5.418402671813965,
      "learning_rate": 9.260304912478827e-07,
      "loss": 0.0719,
      "step": 50670
    },
    {
      "epoch": 9.54,
      "grad_norm": 1.4298168420791626,
      "learning_rate": 9.22266139657444e-07,
      "loss": 0.0135,
      "step": 50680
    },
    {
      "epoch": 9.54,
      "grad_norm": 4.483101844787598,
      "learning_rate": 9.185017880670055e-07,
      "loss": 0.2935,
      "step": 50690
    },
    {
      "epoch": 9.54,
      "grad_norm": 13.930105209350586,
      "learning_rate": 9.14737436476567e-07,
      "loss": 0.0511,
      "step": 50700
    },
    {
      "epoch": 9.54,
      "grad_norm": 12.606245994567871,
      "learning_rate": 9.109730848861284e-07,
      "loss": 0.0998,
      "step": 50710
    },
    {
      "epoch": 9.55,
      "grad_norm": 0.010567332617938519,
      "learning_rate": 9.072087332956899e-07,
      "loss": 0.3006,
      "step": 50720
    },
    {
      "epoch": 9.55,
      "grad_norm": 0.005626684986054897,
      "learning_rate": 9.034443817052512e-07,
      "loss": 0.2377,
      "step": 50730
    },
    {
      "epoch": 9.55,
      "grad_norm": 0.023403815925121307,
      "learning_rate": 8.996800301148129e-07,
      "loss": 0.2369,
      "step": 50740
    },
    {
      "epoch": 9.55,
      "grad_norm": 11.312474250793457,
      "learning_rate": 8.959156785243743e-07,
      "loss": 0.136,
      "step": 50750
    },
    {
      "epoch": 9.55,
      "grad_norm": 10.266181945800781,
      "learning_rate": 8.921513269339357e-07,
      "loss": 0.1517,
      "step": 50760
    },
    {
      "epoch": 9.56,
      "grad_norm": 0.006531096994876862,
      "learning_rate": 8.883869753434972e-07,
      "loss": 0.0376,
      "step": 50770
    },
    {
      "epoch": 9.56,
      "grad_norm": 0.27332255244255066,
      "learning_rate": 8.846226237530586e-07,
      "loss": 0.0908,
      "step": 50780
    },
    {
      "epoch": 9.56,
      "grad_norm": 1.7999335527420044,
      "learning_rate": 8.808582721626201e-07,
      "loss": 0.0707,
      "step": 50790
    },
    {
      "epoch": 9.56,
      "grad_norm": 0.26361072063446045,
      "learning_rate": 8.770939205721816e-07,
      "loss": 0.1168,
      "step": 50800
    },
    {
      "epoch": 9.56,
      "grad_norm": 2.829091787338257,
      "learning_rate": 8.733295689817429e-07,
      "loss": 0.1751,
      "step": 50810
    },
    {
      "epoch": 9.57,
      "grad_norm": 4.492772102355957,
      "learning_rate": 8.695652173913044e-07,
      "loss": 0.1672,
      "step": 50820
    },
    {
      "epoch": 9.57,
      "grad_norm": 4.916472434997559,
      "learning_rate": 8.658008658008658e-07,
      "loss": 0.4034,
      "step": 50830
    },
    {
      "epoch": 9.57,
      "grad_norm": 0.007297304924577475,
      "learning_rate": 8.620365142104273e-07,
      "loss": 0.0416,
      "step": 50840
    },
    {
      "epoch": 9.57,
      "grad_norm": 5.492473125457764,
      "learning_rate": 8.582721626199888e-07,
      "loss": 0.3457,
      "step": 50850
    },
    {
      "epoch": 9.57,
      "grad_norm": 4.431015491485596,
      "learning_rate": 8.545078110295501e-07,
      "loss": 0.1651,
      "step": 50860
    },
    {
      "epoch": 9.57,
      "grad_norm": 0.013440125621855259,
      "learning_rate": 8.507434594391118e-07,
      "loss": 0.0921,
      "step": 50870
    },
    {
      "epoch": 9.58,
      "grad_norm": 19.33757972717285,
      "learning_rate": 8.469791078486731e-07,
      "loss": 0.2312,
      "step": 50880
    },
    {
      "epoch": 9.58,
      "grad_norm": 13.427885055541992,
      "learning_rate": 8.432147562582346e-07,
      "loss": 0.1889,
      "step": 50890
    },
    {
      "epoch": 9.58,
      "grad_norm": 26.11310386657715,
      "learning_rate": 8.394504046677961e-07,
      "loss": 0.1789,
      "step": 50900
    },
    {
      "epoch": 9.58,
      "grad_norm": 11.193914413452148,
      "learning_rate": 8.356860530773575e-07,
      "loss": 0.1954,
      "step": 50910
    },
    {
      "epoch": 9.58,
      "grad_norm": 0.6571615934371948,
      "learning_rate": 8.31921701486919e-07,
      "loss": 0.1427,
      "step": 50920
    },
    {
      "epoch": 9.59,
      "grad_norm": 8.842864036560059,
      "learning_rate": 8.281573498964803e-07,
      "loss": 0.0947,
      "step": 50930
    },
    {
      "epoch": 9.59,
      "grad_norm": 0.2615185081958771,
      "learning_rate": 8.243929983060418e-07,
      "loss": 0.0733,
      "step": 50940
    },
    {
      "epoch": 9.59,
      "grad_norm": 47.92326736450195,
      "learning_rate": 8.206286467156033e-07,
      "loss": 0.1024,
      "step": 50950
    },
    {
      "epoch": 9.59,
      "grad_norm": 6.793840408325195,
      "learning_rate": 8.168642951251647e-07,
      "loss": 0.1865,
      "step": 50960
    },
    {
      "epoch": 9.59,
      "grad_norm": 0.26016518473625183,
      "learning_rate": 8.130999435347262e-07,
      "loss": 0.0954,
      "step": 50970
    },
    {
      "epoch": 9.6,
      "grad_norm": 0.2108842432498932,
      "learning_rate": 8.093355919442876e-07,
      "loss": 0.2342,
      "step": 50980
    },
    {
      "epoch": 9.6,
      "grad_norm": 10.547240257263184,
      "learning_rate": 8.055712403538492e-07,
      "loss": 0.2274,
      "step": 50990
    },
    {
      "epoch": 9.6,
      "grad_norm": 0.00384813966229558,
      "learning_rate": 8.018068887634107e-07,
      "loss": 0.2912,
      "step": 51000
    },
    {
      "epoch": 9.6,
      "grad_norm": 0.020009305328130722,
      "learning_rate": 7.98042537172972e-07,
      "loss": 0.229,
      "step": 51010
    },
    {
      "epoch": 9.6,
      "grad_norm": 20.299415588378906,
      "learning_rate": 7.942781855825335e-07,
      "loss": 0.0841,
      "step": 51020
    },
    {
      "epoch": 9.6,
      "grad_norm": 23.06370735168457,
      "learning_rate": 7.905138339920949e-07,
      "loss": 0.293,
      "step": 51030
    },
    {
      "epoch": 9.61,
      "grad_norm": 13.652506828308105,
      "learning_rate": 7.867494824016564e-07,
      "loss": 0.3722,
      "step": 51040
    },
    {
      "epoch": 9.61,
      "grad_norm": 0.0038510989397764206,
      "learning_rate": 7.829851308112179e-07,
      "loss": 0.0872,
      "step": 51050
    },
    {
      "epoch": 9.61,
      "grad_norm": 10.801384925842285,
      "learning_rate": 7.792207792207792e-07,
      "loss": 0.2585,
      "step": 51060
    },
    {
      "epoch": 9.61,
      "grad_norm": 0.035926658660173416,
      "learning_rate": 7.754564276303407e-07,
      "loss": 0.1479,
      "step": 51070
    },
    {
      "epoch": 9.61,
      "grad_norm": 0.00522974319756031,
      "learning_rate": 7.716920760399021e-07,
      "loss": 0.1585,
      "step": 51080
    },
    {
      "epoch": 9.62,
      "grad_norm": 18.04073715209961,
      "learning_rate": 7.679277244494636e-07,
      "loss": 0.1684,
      "step": 51090
    },
    {
      "epoch": 9.62,
      "grad_norm": 0.8207247257232666,
      "learning_rate": 7.641633728590251e-07,
      "loss": 0.1494,
      "step": 51100
    },
    {
      "epoch": 9.62,
      "grad_norm": 12.734384536743164,
      "learning_rate": 7.603990212685866e-07,
      "loss": 0.1631,
      "step": 51110
    },
    {
      "epoch": 9.62,
      "grad_norm": 0.011319233104586601,
      "learning_rate": 7.566346696781481e-07,
      "loss": 0.2232,
      "step": 51120
    },
    {
      "epoch": 9.62,
      "grad_norm": 0.7226599454879761,
      "learning_rate": 7.528703180877094e-07,
      "loss": 0.0158,
      "step": 51130
    },
    {
      "epoch": 9.63,
      "grad_norm": 0.07509887218475342,
      "learning_rate": 7.491059664972709e-07,
      "loss": 0.1553,
      "step": 51140
    },
    {
      "epoch": 9.63,
      "grad_norm": 0.0459410585463047,
      "learning_rate": 7.453416149068324e-07,
      "loss": 0.172,
      "step": 51150
    },
    {
      "epoch": 9.63,
      "grad_norm": 0.0034712806809693575,
      "learning_rate": 7.415772633163938e-07,
      "loss": 0.2042,
      "step": 51160
    },
    {
      "epoch": 9.63,
      "grad_norm": 0.14592990279197693,
      "learning_rate": 7.378129117259553e-07,
      "loss": 0.1377,
      "step": 51170
    },
    {
      "epoch": 9.63,
      "grad_norm": 0.019890498369932175,
      "learning_rate": 7.340485601355168e-07,
      "loss": 0.0763,
      "step": 51180
    },
    {
      "epoch": 9.63,
      "grad_norm": 1.3121528625488281,
      "learning_rate": 7.302842085450781e-07,
      "loss": 0.0364,
      "step": 51190
    },
    {
      "epoch": 9.64,
      "grad_norm": 0.14124996960163116,
      "learning_rate": 7.265198569546396e-07,
      "loss": 0.247,
      "step": 51200
    },
    {
      "epoch": 9.64,
      "grad_norm": 0.010441433638334274,
      "learning_rate": 7.22755505364201e-07,
      "loss": 0.1165,
      "step": 51210
    },
    {
      "epoch": 9.64,
      "grad_norm": 0.003246891777962446,
      "learning_rate": 7.189911537737625e-07,
      "loss": 0.1306,
      "step": 51220
    },
    {
      "epoch": 9.64,
      "grad_norm": 11.543542861938477,
      "learning_rate": 7.152268021833241e-07,
      "loss": 0.1273,
      "step": 51230
    },
    {
      "epoch": 9.64,
      "grad_norm": 0.1644592434167862,
      "learning_rate": 7.114624505928855e-07,
      "loss": 0.0724,
      "step": 51240
    },
    {
      "epoch": 9.65,
      "grad_norm": 6.949909210205078,
      "learning_rate": 7.07698099002447e-07,
      "loss": 0.2412,
      "step": 51250
    },
    {
      "epoch": 9.65,
      "grad_norm": 0.0055879284627735615,
      "learning_rate": 7.039337474120083e-07,
      "loss": 0.1925,
      "step": 51260
    },
    {
      "epoch": 9.65,
      "grad_norm": 0.05285622552037239,
      "learning_rate": 7.001693958215698e-07,
      "loss": 0.1611,
      "step": 51270
    },
    {
      "epoch": 9.65,
      "grad_norm": 0.005291651003062725,
      "learning_rate": 6.964050442311313e-07,
      "loss": 0.2302,
      "step": 51280
    },
    {
      "epoch": 9.65,
      "grad_norm": 5.34735107421875,
      "learning_rate": 6.926406926406927e-07,
      "loss": 0.0567,
      "step": 51290
    },
    {
      "epoch": 9.66,
      "grad_norm": 32.558326721191406,
      "learning_rate": 6.888763410502542e-07,
      "loss": 0.2322,
      "step": 51300
    },
    {
      "epoch": 9.66,
      "grad_norm": 0.014381973072886467,
      "learning_rate": 6.851119894598155e-07,
      "loss": 0.0981,
      "step": 51310
    },
    {
      "epoch": 9.66,
      "grad_norm": 0.013572474010288715,
      "learning_rate": 6.81347637869377e-07,
      "loss": 0.2635,
      "step": 51320
    },
    {
      "epoch": 9.66,
      "grad_norm": 1.752585530281067,
      "learning_rate": 6.775832862789384e-07,
      "loss": 0.1356,
      "step": 51330
    },
    {
      "epoch": 9.66,
      "grad_norm": 0.003971709869801998,
      "learning_rate": 6.738189346884999e-07,
      "loss": 0.0642,
      "step": 51340
    },
    {
      "epoch": 9.66,
      "grad_norm": 1.5294979810714722,
      "learning_rate": 6.700545830980614e-07,
      "loss": 0.3311,
      "step": 51350
    },
    {
      "epoch": 9.67,
      "grad_norm": 13.638725280761719,
      "learning_rate": 6.662902315076229e-07,
      "loss": 0.1808,
      "step": 51360
    },
    {
      "epoch": 9.67,
      "grad_norm": 4.505469799041748,
      "learning_rate": 6.625258799171844e-07,
      "loss": 0.1915,
      "step": 51370
    },
    {
      "epoch": 9.67,
      "grad_norm": 32.34842300415039,
      "learning_rate": 6.587615283267459e-07,
      "loss": 0.04,
      "step": 51380
    },
    {
      "epoch": 9.67,
      "grad_norm": 0.005380976479500532,
      "learning_rate": 6.549971767363072e-07,
      "loss": 0.1854,
      "step": 51390
    },
    {
      "epoch": 9.67,
      "grad_norm": 0.03604097291827202,
      "learning_rate": 6.512328251458687e-07,
      "loss": 0.2484,
      "step": 51400
    },
    {
      "epoch": 9.68,
      "grad_norm": 6.1091227531433105,
      "learning_rate": 6.474684735554301e-07,
      "loss": 0.0607,
      "step": 51410
    },
    {
      "epoch": 9.68,
      "grad_norm": 10.915803909301758,
      "learning_rate": 6.437041219649916e-07,
      "loss": 0.0364,
      "step": 51420
    },
    {
      "epoch": 9.68,
      "grad_norm": 8.31409740447998,
      "learning_rate": 6.399397703745531e-07,
      "loss": 0.2094,
      "step": 51430
    },
    {
      "epoch": 9.68,
      "grad_norm": 4.822686195373535,
      "learning_rate": 6.361754187841144e-07,
      "loss": 0.1171,
      "step": 51440
    },
    {
      "epoch": 9.68,
      "grad_norm": 93.29821014404297,
      "learning_rate": 6.324110671936759e-07,
      "loss": 0.222,
      "step": 51450
    },
    {
      "epoch": 9.69,
      "grad_norm": 0.006585160735994577,
      "learning_rate": 6.286467156032373e-07,
      "loss": 0.2475,
      "step": 51460
    },
    {
      "epoch": 9.69,
      "grad_norm": 20.083065032958984,
      "learning_rate": 6.248823640127989e-07,
      "loss": 0.0696,
      "step": 51470
    },
    {
      "epoch": 9.69,
      "grad_norm": 0.006351159885525703,
      "learning_rate": 6.211180124223603e-07,
      "loss": 0.1731,
      "step": 51480
    },
    {
      "epoch": 9.69,
      "grad_norm": 4.1066718101501465,
      "learning_rate": 6.173536608319217e-07,
      "loss": 0.0991,
      "step": 51490
    },
    {
      "epoch": 9.69,
      "grad_norm": 9.786056518554688,
      "learning_rate": 6.135893092414832e-07,
      "loss": 0.0326,
      "step": 51500
    },
    {
      "epoch": 9.7,
      "grad_norm": 6.347681045532227,
      "learning_rate": 6.098249576510446e-07,
      "loss": 0.1304,
      "step": 51510
    },
    {
      "epoch": 9.7,
      "grad_norm": 15.366695404052734,
      "learning_rate": 6.060606060606061e-07,
      "loss": 0.0845,
      "step": 51520
    },
    {
      "epoch": 9.7,
      "grad_norm": 5.126430034637451,
      "learning_rate": 6.022962544701676e-07,
      "loss": 0.2675,
      "step": 51530
    },
    {
      "epoch": 9.7,
      "grad_norm": 9.330107688903809,
      "learning_rate": 5.98531902879729e-07,
      "loss": 0.1826,
      "step": 51540
    },
    {
      "epoch": 9.7,
      "grad_norm": 0.013792709447443485,
      "learning_rate": 5.947675512892905e-07,
      "loss": 0.073,
      "step": 51550
    },
    {
      "epoch": 9.7,
      "grad_norm": 3.519915819168091,
      "learning_rate": 5.910031996988519e-07,
      "loss": 0.0835,
      "step": 51560
    },
    {
      "epoch": 9.71,
      "grad_norm": 0.020572861656546593,
      "learning_rate": 5.872388481084134e-07,
      "loss": 0.1187,
      "step": 51570
    },
    {
      "epoch": 9.71,
      "grad_norm": 46.59688186645508,
      "learning_rate": 5.834744965179749e-07,
      "loss": 0.2574,
      "step": 51580
    },
    {
      "epoch": 9.71,
      "grad_norm": 0.7556085586547852,
      "learning_rate": 5.797101449275363e-07,
      "loss": 0.1012,
      "step": 51590
    },
    {
      "epoch": 9.71,
      "grad_norm": 0.5482608675956726,
      "learning_rate": 5.759457933370977e-07,
      "loss": 0.1829,
      "step": 51600
    },
    {
      "epoch": 9.71,
      "grad_norm": 0.008136688731610775,
      "learning_rate": 5.721814417466591e-07,
      "loss": 0.101,
      "step": 51610
    },
    {
      "epoch": 9.72,
      "grad_norm": 9.163705825805664,
      "learning_rate": 5.684170901562206e-07,
      "loss": 0.2156,
      "step": 51620
    },
    {
      "epoch": 9.72,
      "grad_norm": 3.596132278442383,
      "learning_rate": 5.646527385657822e-07,
      "loss": 0.2913,
      "step": 51630
    },
    {
      "epoch": 9.72,
      "grad_norm": 2.0683116912841797,
      "learning_rate": 5.608883869753435e-07,
      "loss": 0.2158,
      "step": 51640
    },
    {
      "epoch": 9.72,
      "grad_norm": 0.0037856632843613625,
      "learning_rate": 5.57124035384905e-07,
      "loss": 0.1773,
      "step": 51650
    },
    {
      "epoch": 9.72,
      "grad_norm": 0.11805713921785355,
      "learning_rate": 5.533596837944664e-07,
      "loss": 0.1783,
      "step": 51660
    },
    {
      "epoch": 9.73,
      "grad_norm": 10.261449813842773,
      "learning_rate": 5.495953322040279e-07,
      "loss": 0.0492,
      "step": 51670
    },
    {
      "epoch": 9.73,
      "grad_norm": 0.002996019320562482,
      "learning_rate": 5.458309806135894e-07,
      "loss": 0.0651,
      "step": 51680
    },
    {
      "epoch": 9.73,
      "grad_norm": 0.20321513712406158,
      "learning_rate": 5.420666290231508e-07,
      "loss": 0.0689,
      "step": 51690
    },
    {
      "epoch": 9.73,
      "grad_norm": 5.545836925506592,
      "learning_rate": 5.383022774327123e-07,
      "loss": 0.1641,
      "step": 51700
    },
    {
      "epoch": 9.73,
      "grad_norm": 35.26544952392578,
      "learning_rate": 5.345379258422737e-07,
      "loss": 0.1865,
      "step": 51710
    },
    {
      "epoch": 9.73,
      "grad_norm": 0.012471206486225128,
      "learning_rate": 5.307735742518352e-07,
      "loss": 0.1785,
      "step": 51720
    },
    {
      "epoch": 9.74,
      "grad_norm": 0.0055159395560622215,
      "learning_rate": 5.270092226613966e-07,
      "loss": 0.0591,
      "step": 51730
    },
    {
      "epoch": 9.74,
      "grad_norm": 0.010178087279200554,
      "learning_rate": 5.23244871070958e-07,
      "loss": 0.0698,
      "step": 51740
    },
    {
      "epoch": 9.74,
      "grad_norm": 0.04755941778421402,
      "learning_rate": 5.194805194805196e-07,
      "loss": 0.3396,
      "step": 51750
    },
    {
      "epoch": 9.74,
      "grad_norm": 0.10165128856897354,
      "learning_rate": 5.15716167890081e-07,
      "loss": 0.1838,
      "step": 51760
    },
    {
      "epoch": 9.74,
      "grad_norm": 37.38896942138672,
      "learning_rate": 5.119518162996424e-07,
      "loss": 0.1275,
      "step": 51770
    },
    {
      "epoch": 9.75,
      "grad_norm": 0.004757017362862825,
      "learning_rate": 5.081874647092039e-07,
      "loss": 0.2773,
      "step": 51780
    },
    {
      "epoch": 9.75,
      "grad_norm": 2.926429271697998,
      "learning_rate": 5.044231131187653e-07,
      "loss": 0.065,
      "step": 51790
    },
    {
      "epoch": 9.75,
      "grad_norm": 0.0037423772737383842,
      "learning_rate": 5.006587615283268e-07,
      "loss": 0.1023,
      "step": 51800
    },
    {
      "epoch": 9.75,
      "grad_norm": 0.006536235101521015,
      "learning_rate": 4.968944099378882e-07,
      "loss": 0.1078,
      "step": 51810
    },
    {
      "epoch": 9.75,
      "grad_norm": 20.170753479003906,
      "learning_rate": 4.931300583474497e-07,
      "loss": 0.1146,
      "step": 51820
    },
    {
      "epoch": 9.76,
      "grad_norm": 25.14588737487793,
      "learning_rate": 4.893657067570112e-07,
      "loss": 0.1489,
      "step": 51830
    },
    {
      "epoch": 9.76,
      "grad_norm": 0.08180654048919678,
      "learning_rate": 4.856013551665726e-07,
      "loss": 0.2522,
      "step": 51840
    },
    {
      "epoch": 9.76,
      "grad_norm": 0.2996702492237091,
      "learning_rate": 4.818370035761341e-07,
      "loss": 0.0222,
      "step": 51850
    },
    {
      "epoch": 9.76,
      "grad_norm": 0.00418334174901247,
      "learning_rate": 4.780726519856954e-07,
      "loss": 0.1374,
      "step": 51860
    },
    {
      "epoch": 9.76,
      "grad_norm": 8.42982292175293,
      "learning_rate": 4.743083003952569e-07,
      "loss": 0.2895,
      "step": 51870
    },
    {
      "epoch": 9.76,
      "grad_norm": 17.296483993530273,
      "learning_rate": 4.705439488048184e-07,
      "loss": 0.2635,
      "step": 51880
    },
    {
      "epoch": 9.77,
      "grad_norm": 0.0035486321430653334,
      "learning_rate": 4.667795972143799e-07,
      "loss": 0.0345,
      "step": 51890
    },
    {
      "epoch": 9.77,
      "grad_norm": 32.99420928955078,
      "learning_rate": 4.6301524562394134e-07,
      "loss": 0.2259,
      "step": 51900
    },
    {
      "epoch": 9.77,
      "grad_norm": 7.924685001373291,
      "learning_rate": 4.5925089403350275e-07,
      "loss": 0.1172,
      "step": 51910
    },
    {
      "epoch": 9.77,
      "grad_norm": 0.38119980692863464,
      "learning_rate": 4.554865424430642e-07,
      "loss": 0.1071,
      "step": 51920
    },
    {
      "epoch": 9.77,
      "grad_norm": 0.007368507329374552,
      "learning_rate": 4.517221908526256e-07,
      "loss": 0.3026,
      "step": 51930
    },
    {
      "epoch": 9.78,
      "grad_norm": 28.610759735107422,
      "learning_rate": 4.4795783926218713e-07,
      "loss": 0.172,
      "step": 51940
    },
    {
      "epoch": 9.78,
      "grad_norm": 0.004183626733720303,
      "learning_rate": 4.441934876717486e-07,
      "loss": 0.114,
      "step": 51950
    },
    {
      "epoch": 9.78,
      "grad_norm": 0.005035290028899908,
      "learning_rate": 4.4042913608131005e-07,
      "loss": 0.0716,
      "step": 51960
    },
    {
      "epoch": 9.78,
      "grad_norm": 12.789113998413086,
      "learning_rate": 4.3666478449087146e-07,
      "loss": 0.0636,
      "step": 51970
    },
    {
      "epoch": 9.78,
      "grad_norm": 0.09953998774290085,
      "learning_rate": 4.329004329004329e-07,
      "loss": 0.2301,
      "step": 51980
    },
    {
      "epoch": 9.79,
      "grad_norm": 24.431516647338867,
      "learning_rate": 4.291360813099944e-07,
      "loss": 0.1681,
      "step": 51990
    },
    {
      "epoch": 9.79,
      "grad_norm": 22.394184112548828,
      "learning_rate": 4.253717297195559e-07,
      "loss": 0.1704,
      "step": 52000
    },
    {
      "epoch": 9.79,
      "grad_norm": 7.764616966247559,
      "learning_rate": 4.216073781291173e-07,
      "loss": 0.1019,
      "step": 52010
    },
    {
      "epoch": 9.79,
      "grad_norm": 23.924365997314453,
      "learning_rate": 4.1784302653867876e-07,
      "loss": 0.2098,
      "step": 52020
    },
    {
      "epoch": 9.79,
      "grad_norm": 0.019631855189800262,
      "learning_rate": 4.1407867494824017e-07,
      "loss": 0.1062,
      "step": 52030
    },
    {
      "epoch": 9.79,
      "grad_norm": 0.40625402331352234,
      "learning_rate": 4.1031432335780163e-07,
      "loss": 0.2127,
      "step": 52040
    },
    {
      "epoch": 9.8,
      "grad_norm": 22.197633743286133,
      "learning_rate": 4.065499717673631e-07,
      "loss": 0.0686,
      "step": 52050
    },
    {
      "epoch": 9.8,
      "grad_norm": 0.11276934295892715,
      "learning_rate": 4.027856201769246e-07,
      "loss": 0.0746,
      "step": 52060
    },
    {
      "epoch": 9.8,
      "grad_norm": 0.005711324047297239,
      "learning_rate": 3.99021268586486e-07,
      "loss": 0.294,
      "step": 52070
    },
    {
      "epoch": 9.8,
      "grad_norm": 42.46678161621094,
      "learning_rate": 3.9525691699604747e-07,
      "loss": 0.3137,
      "step": 52080
    },
    {
      "epoch": 9.8,
      "grad_norm": 0.19246163964271545,
      "learning_rate": 3.9149256540560893e-07,
      "loss": 0.1359,
      "step": 52090
    },
    {
      "epoch": 9.81,
      "grad_norm": 9.77893352508545,
      "learning_rate": 3.8772821381517034e-07,
      "loss": 0.0887,
      "step": 52100
    },
    {
      "epoch": 9.81,
      "grad_norm": 0.6603289246559143,
      "learning_rate": 3.839638622247318e-07,
      "loss": 0.1524,
      "step": 52110
    },
    {
      "epoch": 9.81,
      "grad_norm": 0.12937724590301514,
      "learning_rate": 3.801995106342933e-07,
      "loss": 0.0394,
      "step": 52120
    },
    {
      "epoch": 9.81,
      "grad_norm": 5.976528167724609,
      "learning_rate": 3.764351590438547e-07,
      "loss": 0.1841,
      "step": 52130
    },
    {
      "epoch": 9.81,
      "grad_norm": 0.1112789437174797,
      "learning_rate": 3.726708074534162e-07,
      "loss": 0.0867,
      "step": 52140
    },
    {
      "epoch": 9.82,
      "grad_norm": 0.026705702766776085,
      "learning_rate": 3.6890645586297765e-07,
      "loss": 0.1203,
      "step": 52150
    },
    {
      "epoch": 9.82,
      "grad_norm": 38.82207489013672,
      "learning_rate": 3.6514210427253905e-07,
      "loss": 0.3227,
      "step": 52160
    },
    {
      "epoch": 9.82,
      "grad_norm": 0.12452086806297302,
      "learning_rate": 3.613777526821005e-07,
      "loss": 0.1435,
      "step": 52170
    },
    {
      "epoch": 9.82,
      "grad_norm": 0.02174833044409752,
      "learning_rate": 3.5761340109166203e-07,
      "loss": 0.1322,
      "step": 52180
    },
    {
      "epoch": 9.82,
      "grad_norm": 0.3516334891319275,
      "learning_rate": 3.538490495012235e-07,
      "loss": 0.1111,
      "step": 52190
    },
    {
      "epoch": 9.82,
      "grad_norm": 0.06238023191690445,
      "learning_rate": 3.500846979107849e-07,
      "loss": 0.0516,
      "step": 52200
    },
    {
      "epoch": 9.83,
      "grad_norm": 0.00469348905608058,
      "learning_rate": 3.4632034632034636e-07,
      "loss": 0.1729,
      "step": 52210
    },
    {
      "epoch": 9.83,
      "grad_norm": 0.03575940430164337,
      "learning_rate": 3.4255599472990776e-07,
      "loss": 0.0981,
      "step": 52220
    },
    {
      "epoch": 9.83,
      "grad_norm": 0.06887777894735336,
      "learning_rate": 3.387916431394692e-07,
      "loss": 0.149,
      "step": 52230
    },
    {
      "epoch": 9.83,
      "grad_norm": 0.10040636360645294,
      "learning_rate": 3.350272915490307e-07,
      "loss": 0.1145,
      "step": 52240
    },
    {
      "epoch": 9.83,
      "grad_norm": 33.31528854370117,
      "learning_rate": 3.312629399585922e-07,
      "loss": 0.2321,
      "step": 52250
    },
    {
      "epoch": 9.84,
      "grad_norm": 0.14812682569026947,
      "learning_rate": 3.274985883681536e-07,
      "loss": 0.1468,
      "step": 52260
    },
    {
      "epoch": 9.84,
      "grad_norm": 0.004853300750255585,
      "learning_rate": 3.2373423677771507e-07,
      "loss": 0.1178,
      "step": 52270
    },
    {
      "epoch": 9.84,
      "grad_norm": 3.8380541801452637,
      "learning_rate": 3.1996988518727653e-07,
      "loss": 0.1603,
      "step": 52280
    },
    {
      "epoch": 9.84,
      "grad_norm": 0.005686253309249878,
      "learning_rate": 3.1620553359683794e-07,
      "loss": 0.1868,
      "step": 52290
    },
    {
      "epoch": 9.84,
      "grad_norm": 0.03527418524026871,
      "learning_rate": 3.1244118200639945e-07,
      "loss": 0.3835,
      "step": 52300
    },
    {
      "epoch": 9.85,
      "grad_norm": 28.722721099853516,
      "learning_rate": 3.0867683041596086e-07,
      "loss": 0.3077,
      "step": 52310
    },
    {
      "epoch": 9.85,
      "grad_norm": 0.0313204862177372,
      "learning_rate": 3.049124788255223e-07,
      "loss": 0.1192,
      "step": 52320
    },
    {
      "epoch": 9.85,
      "grad_norm": 8.833168983459473,
      "learning_rate": 3.011481272350838e-07,
      "loss": 0.1018,
      "step": 52330
    },
    {
      "epoch": 9.85,
      "grad_norm": 0.06888779997825623,
      "learning_rate": 2.9738377564464524e-07,
      "loss": 0.1309,
      "step": 52340
    },
    {
      "epoch": 9.85,
      "grad_norm": 0.0031095892190933228,
      "learning_rate": 2.936194240542067e-07,
      "loss": 0.1128,
      "step": 52350
    },
    {
      "epoch": 9.86,
      "grad_norm": 8.762362480163574,
      "learning_rate": 2.8985507246376816e-07,
      "loss": 0.1908,
      "step": 52360
    },
    {
      "epoch": 9.86,
      "grad_norm": 0.016843749210238457,
      "learning_rate": 2.8609072087332957e-07,
      "loss": 0.12,
      "step": 52370
    },
    {
      "epoch": 9.86,
      "grad_norm": 0.004038303624838591,
      "learning_rate": 2.823263692828911e-07,
      "loss": 0.1202,
      "step": 52380
    },
    {
      "epoch": 9.86,
      "grad_norm": 1.118934988975525,
      "learning_rate": 2.785620176924525e-07,
      "loss": 0.1261,
      "step": 52390
    },
    {
      "epoch": 9.86,
      "grad_norm": 0.008919217623770237,
      "learning_rate": 2.7479766610201395e-07,
      "loss": 0.1095,
      "step": 52400
    },
    {
      "epoch": 9.86,
      "grad_norm": 0.009541160427033901,
      "learning_rate": 2.710333145115754e-07,
      "loss": 0.0291,
      "step": 52410
    },
    {
      "epoch": 9.87,
      "grad_norm": 43.4287109375,
      "learning_rate": 2.6726896292113687e-07,
      "loss": 0.1021,
      "step": 52420
    },
    {
      "epoch": 9.87,
      "grad_norm": 0.004459756892174482,
      "learning_rate": 2.635046113306983e-07,
      "loss": 0.2306,
      "step": 52430
    },
    {
      "epoch": 9.87,
      "grad_norm": 1.1663109064102173,
      "learning_rate": 2.597402597402598e-07,
      "loss": 0.0871,
      "step": 52440
    },
    {
      "epoch": 9.87,
      "grad_norm": 11.554471969604492,
      "learning_rate": 2.559759081498212e-07,
      "loss": 0.0719,
      "step": 52450
    },
    {
      "epoch": 9.87,
      "grad_norm": 27.942886352539062,
      "learning_rate": 2.5221155655938266e-07,
      "loss": 0.2675,
      "step": 52460
    },
    {
      "epoch": 9.88,
      "grad_norm": 9.43742561340332,
      "learning_rate": 2.484472049689441e-07,
      "loss": 0.088,
      "step": 52470
    },
    {
      "epoch": 9.88,
      "grad_norm": 0.003390817902982235,
      "learning_rate": 2.446828533785056e-07,
      "loss": 0.3097,
      "step": 52480
    },
    {
      "epoch": 9.88,
      "grad_norm": 0.01334055233746767,
      "learning_rate": 2.4091850178806704e-07,
      "loss": 0.0207,
      "step": 52490
    },
    {
      "epoch": 9.88,
      "grad_norm": 0.004762040916830301,
      "learning_rate": 2.3715415019762845e-07,
      "loss": 0.116,
      "step": 52500
    },
    {
      "epoch": 9.88,
      "grad_norm": 25.891761779785156,
      "learning_rate": 2.3338979860718994e-07,
      "loss": 0.1182,
      "step": 52510
    },
    {
      "epoch": 9.89,
      "grad_norm": 0.005869030952453613,
      "learning_rate": 2.2962544701675137e-07,
      "loss": 0.0157,
      "step": 52520
    },
    {
      "epoch": 9.89,
      "grad_norm": 1.5118013620376587,
      "learning_rate": 2.258610954263128e-07,
      "loss": 0.032,
      "step": 52530
    },
    {
      "epoch": 9.89,
      "grad_norm": 28.971708297729492,
      "learning_rate": 2.220967438358743e-07,
      "loss": 0.3216,
      "step": 52540
    },
    {
      "epoch": 9.89,
      "grad_norm": 0.03376523405313492,
      "learning_rate": 2.1833239224543573e-07,
      "loss": 0.0156,
      "step": 52550
    },
    {
      "epoch": 9.89,
      "grad_norm": 0.051541559398174286,
      "learning_rate": 2.145680406549972e-07,
      "loss": 0.1726,
      "step": 52560
    },
    {
      "epoch": 9.89,
      "grad_norm": 22.35733413696289,
      "learning_rate": 2.1080368906455865e-07,
      "loss": 0.0944,
      "step": 52570
    },
    {
      "epoch": 9.9,
      "grad_norm": 0.005431021563708782,
      "learning_rate": 2.0703933747412008e-07,
      "loss": 0.1852,
      "step": 52580
    },
    {
      "epoch": 9.9,
      "grad_norm": 0.003926573786884546,
      "learning_rate": 2.0327498588368155e-07,
      "loss": 0.1449,
      "step": 52590
    },
    {
      "epoch": 9.9,
      "grad_norm": 10.595260620117188,
      "learning_rate": 1.99510634293243e-07,
      "loss": 0.0857,
      "step": 52600
    },
    {
      "epoch": 9.9,
      "grad_norm": 30.172321319580078,
      "learning_rate": 1.9574628270280447e-07,
      "loss": 0.0988,
      "step": 52610
    },
    {
      "epoch": 9.9,
      "grad_norm": 0.8399246335029602,
      "learning_rate": 1.919819311123659e-07,
      "loss": 0.147,
      "step": 52620
    },
    {
      "epoch": 9.91,
      "grad_norm": 0.004909265320748091,
      "learning_rate": 1.8821757952192736e-07,
      "loss": 0.2406,
      "step": 52630
    },
    {
      "epoch": 9.91,
      "grad_norm": 0.021434729918837547,
      "learning_rate": 1.8445322793148882e-07,
      "loss": 0.0874,
      "step": 52640
    },
    {
      "epoch": 9.91,
      "grad_norm": 8.405227661132812,
      "learning_rate": 1.8068887634105026e-07,
      "loss": 0.1004,
      "step": 52650
    },
    {
      "epoch": 9.91,
      "grad_norm": 11.655117988586426,
      "learning_rate": 1.7692452475061174e-07,
      "loss": 0.0985,
      "step": 52660
    },
    {
      "epoch": 9.91,
      "grad_norm": 0.07113207131624222,
      "learning_rate": 1.7316017316017318e-07,
      "loss": 0.1807,
      "step": 52670
    },
    {
      "epoch": 9.92,
      "grad_norm": 6.747674465179443,
      "learning_rate": 1.693958215697346e-07,
      "loss": 0.1336,
      "step": 52680
    },
    {
      "epoch": 9.92,
      "grad_norm": 0.20838651061058044,
      "learning_rate": 1.656314699792961e-07,
      "loss": 0.0943,
      "step": 52690
    },
    {
      "epoch": 9.92,
      "grad_norm": 35.7362174987793,
      "learning_rate": 1.6186711838885753e-07,
      "loss": 0.1659,
      "step": 52700
    },
    {
      "epoch": 9.92,
      "grad_norm": 0.009511086158454418,
      "learning_rate": 1.5810276679841897e-07,
      "loss": 0.0139,
      "step": 52710
    },
    {
      "epoch": 9.92,
      "grad_norm": 10.669073104858398,
      "learning_rate": 1.5433841520798043e-07,
      "loss": 0.1537,
      "step": 52720
    },
    {
      "epoch": 9.92,
      "grad_norm": 1.4942444562911987,
      "learning_rate": 1.505740636175419e-07,
      "loss": 0.0931,
      "step": 52730
    },
    {
      "epoch": 9.93,
      "grad_norm": 8.831548690795898,
      "learning_rate": 1.4680971202710335e-07,
      "loss": 0.0596,
      "step": 52740
    },
    {
      "epoch": 9.93,
      "grad_norm": 19.278507232666016,
      "learning_rate": 1.4304536043666478e-07,
      "loss": 0.3693,
      "step": 52750
    },
    {
      "epoch": 9.93,
      "grad_norm": 5.302980899810791,
      "learning_rate": 1.3928100884622625e-07,
      "loss": 0.036,
      "step": 52760
    },
    {
      "epoch": 9.93,
      "grad_norm": 0.01890048198401928,
      "learning_rate": 1.355166572557877e-07,
      "loss": 0.151,
      "step": 52770
    },
    {
      "epoch": 9.93,
      "grad_norm": 0.00646629836410284,
      "learning_rate": 1.3175230566534914e-07,
      "loss": 0.1642,
      "step": 52780
    },
    {
      "epoch": 9.94,
      "grad_norm": 5.075627326965332,
      "learning_rate": 1.279879540749106e-07,
      "loss": 0.2049,
      "step": 52790
    },
    {
      "epoch": 9.94,
      "grad_norm": 29.73031997680664,
      "learning_rate": 1.2422360248447206e-07,
      "loss": 0.1374,
      "step": 52800
    },
    {
      "epoch": 9.94,
      "grad_norm": 0.31691157817840576,
      "learning_rate": 1.2045925089403352e-07,
      "loss": 0.1492,
      "step": 52810
    },
    {
      "epoch": 9.94,
      "grad_norm": 4.610744953155518,
      "learning_rate": 1.1669489930359497e-07,
      "loss": 0.2632,
      "step": 52820
    },
    {
      "epoch": 9.94,
      "grad_norm": 0.044259361922740936,
      "learning_rate": 1.129305477131564e-07,
      "loss": 0.1004,
      "step": 52830
    },
    {
      "epoch": 9.95,
      "grad_norm": 0.0451117679476738,
      "learning_rate": 1.0916619612271786e-07,
      "loss": 0.0607,
      "step": 52840
    },
    {
      "epoch": 9.95,
      "grad_norm": 5.806632995605469,
      "learning_rate": 1.0540184453227933e-07,
      "loss": 0.1482,
      "step": 52850
    },
    {
      "epoch": 9.95,
      "grad_norm": 13.765084266662598,
      "learning_rate": 1.0163749294184077e-07,
      "loss": 0.2953,
      "step": 52860
    },
    {
      "epoch": 9.95,
      "grad_norm": 0.20953358709812164,
      "learning_rate": 9.787314135140223e-08,
      "loss": 0.3749,
      "step": 52870
    },
    {
      "epoch": 9.95,
      "grad_norm": 12.787135124206543,
      "learning_rate": 9.410878976096368e-08,
      "loss": 0.1901,
      "step": 52880
    },
    {
      "epoch": 9.95,
      "grad_norm": 7.956745147705078,
      "learning_rate": 9.034443817052513e-08,
      "loss": 0.2469,
      "step": 52890
    },
    {
      "epoch": 9.96,
      "grad_norm": 0.0052265203557908535,
      "learning_rate": 8.658008658008659e-08,
      "loss": 0.0559,
      "step": 52900
    },
    {
      "epoch": 9.96,
      "grad_norm": 18.069608688354492,
      "learning_rate": 8.281573498964805e-08,
      "loss": 0.3897,
      "step": 52910
    },
    {
      "epoch": 9.96,
      "grad_norm": 5.714245319366455,
      "learning_rate": 7.905138339920948e-08,
      "loss": 0.0411,
      "step": 52920
    },
    {
      "epoch": 9.96,
      "grad_norm": 2.3237922191619873,
      "learning_rate": 7.528703180877094e-08,
      "loss": 0.1545,
      "step": 52930
    },
    {
      "epoch": 9.96,
      "grad_norm": 0.017527252435684204,
      "learning_rate": 7.152268021833239e-08,
      "loss": 0.2306,
      "step": 52940
    },
    {
      "epoch": 9.97,
      "grad_norm": 16.88936996459961,
      "learning_rate": 6.775832862789385e-08,
      "loss": 0.3103,
      "step": 52950
    },
    {
      "epoch": 9.97,
      "grad_norm": 0.025157351046800613,
      "learning_rate": 6.39939770374553e-08,
      "loss": 0.1518,
      "step": 52960
    },
    {
      "epoch": 9.97,
      "grad_norm": 0.009600149467587471,
      "learning_rate": 6.022962544701676e-08,
      "loss": 0.11,
      "step": 52970
    },
    {
      "epoch": 9.97,
      "grad_norm": 3.076256275177002,
      "learning_rate": 5.64652738565782e-08,
      "loss": 0.0252,
      "step": 52980
    },
    {
      "epoch": 9.97,
      "grad_norm": 0.1476135104894638,
      "learning_rate": 5.270092226613966e-08,
      "loss": 0.1602,
      "step": 52990
    },
    {
      "epoch": 9.98,
      "grad_norm": 0.003721782937645912,
      "learning_rate": 4.893657067570112e-08,
      "loss": 0.1003,
      "step": 53000
    },
    {
      "epoch": 9.98,
      "grad_norm": 0.02556477300822735,
      "learning_rate": 4.5172219085262564e-08,
      "loss": 0.1424,
      "step": 53010
    },
    {
      "epoch": 9.98,
      "grad_norm": 0.003916356712579727,
      "learning_rate": 4.1407867494824025e-08,
      "loss": 0.2292,
      "step": 53020
    },
    {
      "epoch": 9.98,
      "grad_norm": 3.7394144535064697,
      "learning_rate": 3.764351590438547e-08,
      "loss": 0.1305,
      "step": 53030
    },
    {
      "epoch": 9.98,
      "grad_norm": 0.01022093091160059,
      "learning_rate": 3.3879164313946926e-08,
      "loss": 0.0459,
      "step": 53040
    },
    {
      "epoch": 9.98,
      "grad_norm": 9.358291625976562,
      "learning_rate": 3.011481272350838e-08,
      "loss": 0.0851,
      "step": 53050
    },
    {
      "epoch": 9.99,
      "grad_norm": 0.006715327501296997,
      "learning_rate": 2.635046113306983e-08,
      "loss": 0.2074,
      "step": 53060
    },
    {
      "epoch": 9.99,
      "grad_norm": 2.4220783710479736,
      "learning_rate": 2.2586109542631282e-08,
      "loss": 0.1077,
      "step": 53070
    },
    {
      "epoch": 9.99,
      "grad_norm": 0.0027555287815630436,
      "learning_rate": 1.8821757952192736e-08,
      "loss": 0.1622,
      "step": 53080
    },
    {
      "epoch": 9.99,
      "grad_norm": 0.00825362466275692,
      "learning_rate": 1.505740636175419e-08,
      "loss": 0.0077,
      "step": 53090
    },
    {
      "epoch": 9.99,
      "grad_norm": 1.5795022249221802,
      "learning_rate": 1.1293054771315641e-08,
      "loss": 0.1304,
      "step": 53100
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.026925668120384216,
      "learning_rate": 7.528703180877095e-09,
      "loss": 0.0562,
      "step": 53110
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.03492143005132675,
      "learning_rate": 3.7643515904385476e-09,
      "loss": 0.1939,
      "step": 53120
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.003115452826023102,
      "learning_rate": 0.0,
      "loss": 0.2009,
      "step": 53130
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9916,
      "eval_loss": 0.034325700253248215,
      "eval_runtime": 26.735,
      "eval_samples_per_second": 280.531,
      "eval_steps_per_second": 35.085,
      "step": 53130
    },
    {
      "epoch": 10.0,
      "step": 53130,
      "total_flos": 3.29364570092544e+19,
      "train_loss": 0.24297491548954178,
      "train_runtime": 9405.5545,
      "train_samples_per_second": 45.186,
      "train_steps_per_second": 5.649
    }
  ],
  "logging_steps": 10,
  "max_steps": 53130,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 3.29364570092544e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}