{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0188,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001,
      "grad_norm": 5.9411797523498535,
      "learning_rate": 0.00019982,
      "loss": 10.402878570556641,
      "step": 10
    },
    {
      "epoch": 0.002,
      "grad_norm": 2.5087270736694336,
      "learning_rate": 0.00019962000000000002,
      "loss": 5.9839332580566404,
      "step": 20
    },
    {
      "epoch": 0.003,
      "grad_norm": 1.3757283687591553,
      "learning_rate": 0.00019942,
      "loss": 5.513664627075196,
      "step": 30
    },
    {
      "epoch": 0.004,
      "grad_norm": 1.2012906074523926,
      "learning_rate": 0.00019922,
      "loss": 5.281090545654297,
      "step": 40
    },
    {
      "epoch": 0.005,
      "grad_norm": 2.0055949687957764,
      "learning_rate": 0.00019902,
      "loss": 5.105018615722656,
      "step": 50
    },
    {
      "epoch": 0.006,
      "grad_norm": 2.674002170562744,
      "learning_rate": 0.00019882,
      "loss": 4.82950439453125,
      "step": 60
    },
    {
      "epoch": 0.007,
      "grad_norm": 1.7347458600997925,
      "learning_rate": 0.00019862000000000002,
      "loss": 4.698889923095703,
      "step": 70
    },
    {
      "epoch": 0.008,
      "grad_norm": 2.345951795578003,
      "learning_rate": 0.00019842000000000001,
      "loss": 4.519314956665039,
      "step": 80
    },
    {
      "epoch": 0.009,
      "grad_norm": 1.3420459032058716,
      "learning_rate": 0.00019822,
      "loss": 4.280438995361328,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4221423864364624,
      "learning_rate": 0.00019802,
      "loss": 4.258197784423828,
      "step": 100
    },
    {
      "epoch": 0.011,
      "grad_norm": 1.4795005321502686,
      "learning_rate": 0.00019782,
      "loss": 4.103477478027344,
      "step": 110
    },
    {
      "epoch": 0.012,
      "grad_norm": 2.135676860809326,
      "learning_rate": 0.00019762,
      "loss": 3.9324199676513674,
      "step": 120
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.9774000644683838,
      "learning_rate": 0.00019742000000000002,
      "loss": 3.872743606567383,
      "step": 130
    },
    {
      "epoch": 0.014,
      "grad_norm": 1.0677322149276733,
      "learning_rate": 0.00019722,
      "loss": 3.8740089416503904,
      "step": 140
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.956723153591156,
      "learning_rate": 0.00019702,
      "loss": 3.7239048004150392,
      "step": 150
    },
    {
      "epoch": 0.016,
      "grad_norm": 1.0390738248825073,
      "learning_rate": 0.00019682,
      "loss": 3.7527164459228515,
      "step": 160
    },
    {
      "epoch": 0.017,
      "grad_norm": 1.084784984588623,
      "learning_rate": 0.00019662,
      "loss": 3.5590255737304686,
      "step": 170
    },
    {
      "epoch": 0.018,
      "grad_norm": 1.1409659385681152,
      "learning_rate": 0.00019642,
      "loss": 3.6691215515136717,
      "step": 180
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.8642674088478088,
      "learning_rate": 0.00019622000000000002,
      "loss": 3.5040924072265627,
      "step": 190
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8515411019325256,
      "learning_rate": 0.00019602,
      "loss": 3.5071990966796873,
      "step": 200
    },
    {
      "epoch": 0.021,
      "grad_norm": 1.0194536447525024,
      "learning_rate": 0.00019582,
      "loss": 3.3893787384033205,
      "step": 210
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.9459556937217712,
      "learning_rate": 0.00019562,
      "loss": 3.3273845672607423,
      "step": 220
    },
    {
      "epoch": 0.023,
      "grad_norm": 0.8981431722640991,
      "learning_rate": 0.00019542,
      "loss": 3.3746875762939452,
      "step": 230
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.9751044511795044,
      "learning_rate": 0.00019522,
      "loss": 3.1989873886108398,
      "step": 240
    },
    {
      "epoch": 0.025,
      "grad_norm": 0.8557232022285461,
      "learning_rate": 0.00019502,
      "loss": 3.1978750228881836,
      "step": 250
    },
    {
      "epoch": 0.026,
      "grad_norm": 0.7834332585334778,
      "learning_rate": 0.00019482,
      "loss": 3.28112678527832,
      "step": 260
    },
    {
      "epoch": 0.027,
      "grad_norm": 0.8482623100280762,
      "learning_rate": 0.00019462,
      "loss": 3.2512683868408203,
      "step": 270
    },
    {
      "epoch": 0.028,
      "grad_norm": 1.1247769594192505,
      "learning_rate": 0.00019442,
      "loss": 3.167398452758789,
      "step": 280
    },
    {
      "epoch": 0.029,
      "grad_norm": 0.7613039612770081,
      "learning_rate": 0.00019422,
      "loss": 3.1816900253295897,
      "step": 290
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7643451690673828,
      "learning_rate": 0.00019402,
      "loss": 2.998531723022461,
      "step": 300
    },
    {
      "epoch": 0.031,
      "grad_norm": 0.8855999112129211,
      "learning_rate": 0.00019382,
      "loss": 3.131449317932129,
      "step": 310
    },
    {
      "epoch": 1.0008,
      "grad_norm": 0.8182146549224854,
      "learning_rate": 0.00019362,
      "loss": 3.362815093994141,
      "step": 320
    },
    {
      "epoch": 1.0018,
      "grad_norm": 0.8580504655838013,
      "learning_rate": 0.00019342,
      "loss": 3.060775947570801,
      "step": 330
    },
    {
      "epoch": 1.0028,
      "grad_norm": 0.691214382648468,
      "learning_rate": 0.00019322,
      "loss": 2.930637741088867,
      "step": 340
    },
    {
      "epoch": 1.0038,
      "grad_norm": 0.8206065893173218,
      "learning_rate": 0.00019302,
      "loss": 2.9964399337768555,
      "step": 350
    },
    {
      "epoch": 1.0048,
      "grad_norm": 0.7985368371009827,
      "learning_rate": 0.00019282000000000001,
      "loss": 2.869978141784668,
      "step": 360
    },
    {
      "epoch": 1.0058,
      "grad_norm": 0.8223126530647278,
      "learning_rate": 0.00019262,
      "loss": 2.929924964904785,
      "step": 370
    },
    {
      "epoch": 1.0068,
      "grad_norm": 0.7868787050247192,
      "learning_rate": 0.00019242,
      "loss": 2.8226070404052734,
      "step": 380
    },
    {
      "epoch": 1.0078,
      "grad_norm": 0.6478514671325684,
      "learning_rate": 0.00019222,
      "loss": 2.8945112228393555,
      "step": 390
    },
    {
      "epoch": 1.0088,
      "grad_norm": 0.8428190350532532,
      "learning_rate": 0.00019202,
      "loss": 2.852457618713379,
      "step": 400
    },
    {
      "epoch": 1.0098,
      "grad_norm": 0.7077657580375671,
      "learning_rate": 0.00019182,
      "loss": 2.887688636779785,
      "step": 410
    },
    {
      "epoch": 1.0108,
      "grad_norm": 0.6806844472885132,
      "learning_rate": 0.00019162,
      "loss": 2.8496984481811523,
      "step": 420
    },
    {
      "epoch": 1.0118,
      "grad_norm": 0.6535109877586365,
      "learning_rate": 0.00019142,
      "loss": 2.7463850021362304,
      "step": 430
    },
    {
      "epoch": 1.0128,
      "grad_norm": 0.611242949962616,
      "learning_rate": 0.00019122,
      "loss": 2.7267787933349608,
      "step": 440
    },
    {
      "epoch": 1.0138,
      "grad_norm": 0.7190207839012146,
      "learning_rate": 0.00019102,
      "loss": 2.7335268020629884,
      "step": 450
    },
    {
      "epoch": 1.0148,
      "grad_norm": 0.6520172357559204,
      "learning_rate": 0.00019082,
      "loss": 2.705989646911621,
      "step": 460
    },
    {
      "epoch": 1.0158,
      "grad_norm": 0.7822602987289429,
      "learning_rate": 0.00019062,
      "loss": 2.7904048919677735,
      "step": 470
    },
    {
      "epoch": 1.0168,
      "grad_norm": 0.6247469186782837,
      "learning_rate": 0.00019042,
      "loss": 2.627030372619629,
      "step": 480
    },
    {
      "epoch": 1.0178,
      "grad_norm": 0.7010595798492432,
      "learning_rate": 0.00019022,
      "loss": 2.7214162826538084,
      "step": 490
    },
    {
      "epoch": 1.0188,
      "grad_norm": 0.5612587928771973,
      "learning_rate": 0.00019002,
      "loss": 2.7336082458496094,
      "step": 500
    },
    {
      "epoch": 1.0198,
      "grad_norm": 0.6944838762283325,
      "learning_rate": 0.00018982000000000002,
      "loss": 2.67153377532959,
      "step": 510
    },
    {
      "epoch": 1.0208,
      "grad_norm": 0.7149194478988647,
      "learning_rate": 0.00018962000000000002,
      "loss": 2.589804267883301,
      "step": 520
    },
    {
      "epoch": 1.0218,
      "grad_norm": 0.7364429235458374,
      "learning_rate": 0.00018942,
      "loss": 2.5354875564575194,
      "step": 530
    },
    {
      "epoch": 1.0228,
      "grad_norm": 0.5923320651054382,
      "learning_rate": 0.00018922,
      "loss": 2.6230649948120117,
      "step": 540
    },
    {
      "epoch": 1.0238,
      "grad_norm": 0.7289549112319946,
      "learning_rate": 0.00018902000000000003,
      "loss": 2.5119558334350587,
      "step": 550
    },
    {
      "epoch": 1.0248,
      "grad_norm": 0.658437967300415,
      "learning_rate": 0.00018882000000000003,
      "loss": 2.46578369140625,
      "step": 560
    },
    {
      "epoch": 1.0258,
      "grad_norm": 0.6258341073989868,
      "learning_rate": 0.00018862000000000002,
      "loss": 2.585083770751953,
      "step": 570
    },
    {
      "epoch": 1.0268,
      "grad_norm": 0.6197068691253662,
      "learning_rate": 0.00018842000000000002,
      "loss": 2.590087127685547,
      "step": 580
    },
    {
      "epoch": 1.0278,
      "grad_norm": 0.6865118145942688,
      "learning_rate": 0.00018822,
      "loss": 2.4772613525390623,
      "step": 590
    },
    {
      "epoch": 1.0288,
      "grad_norm": 0.6310247778892517,
      "learning_rate": 0.00018802,
      "loss": 2.505804252624512,
      "step": 600
    },
    {
      "epoch": 1.0298,
      "grad_norm": 0.6243994832038879,
      "learning_rate": 0.00018782000000000003,
      "loss": 2.4622268676757812,
      "step": 610
    },
    {
      "epoch": 1.0308,
      "grad_norm": 0.6914836168289185,
      "learning_rate": 0.00018762000000000002,
      "loss": 2.4515810012817383,
      "step": 620
    },
    {
      "epoch": 2.0006,
      "grad_norm": 0.7113664746284485,
      "learning_rate": 0.00018742000000000002,
      "loss": 2.673104667663574,
      "step": 630
    },
    {
      "epoch": 2.0016,
      "grad_norm": 0.660602331161499,
      "learning_rate": 0.00018722,
      "loss": 2.4536123275756836,
      "step": 640
    },
    {
      "epoch": 2.0026,
      "grad_norm": 0.6207917928695679,
      "learning_rate": 0.00018702,
      "loss": 2.409420204162598,
      "step": 650
    },
    {
      "epoch": 2.0036,
      "grad_norm": 0.7043523192405701,
      "learning_rate": 0.00018682000000000003,
      "loss": 2.391322898864746,
      "step": 660
    },
    {
      "epoch": 2.0046,
      "grad_norm": 0.6161942481994629,
      "learning_rate": 0.00018662000000000003,
      "loss": 2.3142091751098635,
      "step": 670
    },
    {
      "epoch": 2.0056,
      "grad_norm": 0.7580428123474121,
      "learning_rate": 0.00018642000000000002,
      "loss": 2.345711326599121,
      "step": 680
    },
    {
      "epoch": 2.0066,
      "grad_norm": 0.7074770927429199,
      "learning_rate": 0.00018622000000000002,
      "loss": 2.318608283996582,
      "step": 690
    },
    {
      "epoch": 2.0076,
      "grad_norm": 0.6267780661582947,
      "learning_rate": 0.00018602,
      "loss": 2.3219219207763673,
      "step": 700
    },
    {
      "epoch": 2.0086,
      "grad_norm": 0.6792202591896057,
      "learning_rate": 0.00018582,
      "loss": 2.2827287673950196,
      "step": 710
    },
    {
      "epoch": 2.0096,
      "grad_norm": 0.671473503112793,
      "learning_rate": 0.00018562000000000003,
      "loss": 2.300338935852051,
      "step": 720
    },
    {
      "epoch": 2.0106,
      "grad_norm": 0.6469615697860718,
      "learning_rate": 0.00018542000000000002,
      "loss": 2.3048261642456054,
      "step": 730
    },
    {
      "epoch": 2.0116,
      "grad_norm": 0.6969014406204224,
      "learning_rate": 0.00018522000000000002,
      "loss": 2.217598533630371,
      "step": 740
    },
    {
      "epoch": 2.0126,
      "grad_norm": 0.641813337802887,
      "learning_rate": 0.00018502000000000001,
      "loss": 2.1945074081420897,
      "step": 750
    },
    {
      "epoch": 2.0136,
      "grad_norm": 0.6202750205993652,
      "learning_rate": 0.00018482,
      "loss": 2.2036405563354493,
      "step": 760
    },
    {
      "epoch": 2.0146,
      "grad_norm": 0.6696969866752625,
      "learning_rate": 0.00018462,
      "loss": 2.199502944946289,
      "step": 770
    },
    {
      "epoch": 2.0156,
      "grad_norm": 0.7106838226318359,
      "learning_rate": 0.00018442000000000003,
      "loss": 2.2403232574462892,
      "step": 780
    },
    {
      "epoch": 2.0166,
      "grad_norm": 0.6160182356834412,
      "learning_rate": 0.00018422000000000002,
      "loss": 2.1474706649780275,
      "step": 790
    },
    {
      "epoch": 2.0176,
      "grad_norm": 0.6780025362968445,
      "learning_rate": 0.00018402000000000002,
      "loss": 2.233156204223633,
      "step": 800
    },
    {
      "epoch": 2.0186,
      "grad_norm": 0.6111968755722046,
      "learning_rate": 0.00018382,
      "loss": 2.2461593627929686,
      "step": 810
    },
    {
      "epoch": 2.0196,
      "grad_norm": 0.7200072407722473,
      "learning_rate": 0.00018362,
      "loss": 2.1219438552856444,
      "step": 820
    },
    {
      "epoch": 2.0206,
      "grad_norm": 0.6476977467536926,
      "learning_rate": 0.00018342,
      "loss": 2.0821203231811523,
      "step": 830
    },
    {
      "epoch": 2.0216,
      "grad_norm": 0.6472647786140442,
      "learning_rate": 0.00018322000000000002,
      "loss": 2.0447275161743166,
      "step": 840
    },
    {
      "epoch": 2.0226,
      "grad_norm": 0.7900026440620422,
      "learning_rate": 0.00018302000000000002,
      "loss": 2.087274932861328,
      "step": 850
    },
    {
      "epoch": 2.0236,
      "grad_norm": 0.6523037552833557,
      "learning_rate": 0.00018282000000000001,
      "loss": 2.0110971450805666,
      "step": 860
    },
    {
      "epoch": 2.0246,
      "grad_norm": 0.5921297073364258,
      "learning_rate": 0.00018262,
      "loss": 1.981992530822754,
      "step": 870
    },
    {
      "epoch": 2.0256,
      "grad_norm": 0.7393470406532288,
      "learning_rate": 0.00018242,
      "loss": 2.044318962097168,
      "step": 880
    },
    {
      "epoch": 2.0266,
      "grad_norm": 0.697785496711731,
      "learning_rate": 0.00018222,
      "loss": 2.05950870513916,
      "step": 890
    },
    {
      "epoch": 2.0276,
      "grad_norm": 0.6266019940376282,
      "learning_rate": 0.00018202000000000002,
      "loss": 1.9614368438720704,
      "step": 900
    },
    {
      "epoch": 2.0286,
      "grad_norm": 0.6764276623725891,
      "learning_rate": 0.00018182000000000002,
      "loss": 1.9438285827636719,
      "step": 910
    },
    {
      "epoch": 2.0296,
      "grad_norm": 0.7023544907569885,
      "learning_rate": 0.00018162,
      "loss": 2.02318115234375,
      "step": 920
    },
    {
      "epoch": 2.0306,
      "grad_norm": 0.8344048857688904,
      "learning_rate": 0.00018142,
      "loss": 1.822168731689453,
      "step": 930
    },
    {
      "epoch": 3.0004,
      "grad_norm": 0.7352348566055298,
      "learning_rate": 0.00018122,
      "loss": 2.1745040893554686,
      "step": 940
    },
    {
      "epoch": 3.0014,
      "grad_norm": 0.7016635537147522,
      "learning_rate": 0.00018102000000000003,
      "loss": 1.9799747467041016,
      "step": 950
    },
    {
      "epoch": 3.0024,
      "grad_norm": 0.7483706474304199,
      "learning_rate": 0.00018082000000000002,
      "loss": 1.9175453186035156,
      "step": 960
    },
    {
      "epoch": 3.0034,
      "grad_norm": 0.776687502861023,
      "learning_rate": 0.00018062000000000002,
      "loss": 1.8681112289428712,
      "step": 970
    },
    {
      "epoch": 3.0044,
      "grad_norm": 0.7005246877670288,
      "learning_rate": 0.00018042,
      "loss": 1.8719995498657227,
      "step": 980
    },
    {
      "epoch": 3.0054,
      "grad_norm": 0.729756772518158,
      "learning_rate": 0.00018022,
      "loss": 1.8054256439208984,
      "step": 990
    },
    {
      "epoch": 3.0064,
      "grad_norm": 0.7293346524238586,
      "learning_rate": 0.00018002,
      "loss": 1.862677001953125,
      "step": 1000
    },
    {
      "epoch": 1.001,
      "grad_norm": 0.7271579504013062,
      "learning_rate": 0.00017982000000000002,
      "loss": 1.4198949813842774,
      "step": 1010
    },
    {
      "epoch": 1.002,
      "grad_norm": 0.7241926789283752,
      "learning_rate": 0.00017962000000000002,
      "loss": 1.290165328979492,
      "step": 1020
    },
    {
      "epoch": 1.003,
      "grad_norm": 0.6930139064788818,
      "learning_rate": 0.00017942,
      "loss": 1.2588088989257813,
      "step": 1030
    },
    {
      "epoch": 1.004,
      "grad_norm": 0.7046691179275513,
      "learning_rate": 0.00017922,
      "loss": 1.2508729934692382,
      "step": 1040
    },
    {
      "epoch": 1.005,
      "grad_norm": 0.7425150275230408,
      "learning_rate": 0.00017902,
      "loss": 1.2099505424499513,
      "step": 1050
    },
    {
      "epoch": 1.006,
      "grad_norm": 0.6843275427818298,
      "learning_rate": 0.00017882,
      "loss": 1.1226897239685059,
      "step": 1060
    },
    {
      "epoch": 1.007,
      "grad_norm": 0.6979613900184631,
      "learning_rate": 0.00017862000000000002,
      "loss": 1.6102104187011719,
      "step": 1070
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.6627645492553711,
      "learning_rate": 0.00017842000000000002,
      "loss": 1.9422037124633789,
      "step": 1080
    },
    {
      "epoch": 1.009,
      "grad_norm": 0.6664915084838867,
      "learning_rate": 0.00017822,
      "loss": 1.8065723419189452,
      "step": 1090
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.6870133280754089,
      "learning_rate": 0.00017802,
      "loss": 1.8182893753051759,
      "step": 1100
    },
    {
      "epoch": 1.011,
      "grad_norm": 0.6618905663490295,
      "learning_rate": 0.00017782,
      "loss": 1.814337158203125,
      "step": 1110
    },
    {
      "epoch": 1.012,
      "grad_norm": 1.1367806196212769,
      "learning_rate": 0.00017762,
      "loss": 1.73431453704834,
      "step": 1120
    },
    {
      "epoch": 1.013,
      "grad_norm": 0.685649573802948,
      "learning_rate": 0.00017742000000000002,
      "loss": 1.7108922958374024,
      "step": 1130
    },
    {
      "epoch": 1.014,
      "grad_norm": 0.6594149470329285,
      "learning_rate": 0.00017722000000000001,
      "loss": 1.779058074951172,
      "step": 1140
    },
    {
      "epoch": 1.015,
      "grad_norm": 0.6914166212081909,
      "learning_rate": 0.00017702,
      "loss": 1.7222532272338866,
      "step": 1150
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.6719418168067932,
      "learning_rate": 0.00017682,
      "loss": 1.730402946472168,
      "step": 1160
    },
    {
      "epoch": 1.017,
      "grad_norm": 0.7353265881538391,
      "learning_rate": 0.00017662,
      "loss": 1.711669921875,
      "step": 1170
    },
    {
      "epoch": 1.018,
      "grad_norm": 0.6861590147018433,
      "learning_rate": 0.00017642,
      "loss": 1.7661975860595702,
      "step": 1180
    },
    {
      "epoch": 1.019,
      "grad_norm": 0.6513635516166687,
      "learning_rate": 0.00017622000000000002,
      "loss": 1.6878995895385742,
      "step": 1190
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.6829492449760437,
      "learning_rate": 0.00017602,
      "loss": 1.6560598373413087,
      "step": 1200
    },
    {
      "epoch": 1.021,
      "grad_norm": 0.6806496381759644,
      "learning_rate": 0.00017582,
      "loss": 1.6022390365600585,
      "step": 1210
    },
    {
      "epoch": 1.022,
      "grad_norm": 0.6467958092689514,
      "learning_rate": 0.00017562,
      "loss": 1.527933406829834,
      "step": 1220
    },
    {
      "epoch": 1.023,
      "grad_norm": 0.7558693289756775,
      "learning_rate": 0.00017542,
      "loss": 1.6221937179565429,
      "step": 1230
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.7131749391555786,
      "learning_rate": 0.00017522000000000002,
      "loss": 1.5218177795410157,
      "step": 1240
    },
    {
      "epoch": 1.025,
      "grad_norm": 0.7794132828712463,
      "learning_rate": 0.00017502000000000001,
      "loss": 1.5550528526306153,
      "step": 1250
    },
    {
      "epoch": 1.026,
      "grad_norm": 0.7174275517463684,
      "learning_rate": 0.00017482,
      "loss": 1.5929729461669921,
      "step": 1260
    },
    {
      "epoch": 1.027,
      "grad_norm": 0.7710967659950256,
      "learning_rate": 0.00017462,
      "loss": 1.4883572578430175,
      "step": 1270
    },
    {
      "epoch": 1.028,
      "grad_norm": 0.689930260181427,
      "learning_rate": 0.00017442,
      "loss": 1.4959463119506835,
      "step": 1280
    },
    {
      "epoch": 1.029,
      "grad_norm": 0.7309102416038513,
      "learning_rate": 0.00017422,
      "loss": 1.5185231208801269,
      "step": 1290
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.6255451440811157,
      "learning_rate": 0.00017402000000000002,
      "loss": 1.4664793968200684,
      "step": 1300
    },
    {
      "epoch": 1.031,
      "grad_norm": 0.7244739532470703,
      "learning_rate": 0.00017382,
      "loss": 1.4731544494628905,
      "step": 1310
    },
    {
      "epoch": 2.0008,
      "grad_norm": 0.7122154235839844,
      "learning_rate": 0.00017362,
      "loss": 1.3298330307006836,
      "step": 1320
    },
    {
      "epoch": 2.0018,
      "grad_norm": 0.7280747294425964,
      "learning_rate": 0.00017342,
      "loss": 1.094059658050537,
      "step": 1330
    },
    {
      "epoch": 2.0028,
      "grad_norm": 0.7428712844848633,
      "learning_rate": 0.00017322,
      "loss": 0.9706879615783691,
      "step": 1340
    },
    {
      "epoch": 2.0038,
      "grad_norm": 0.759975790977478,
      "learning_rate": 0.00017302,
      "loss": 1.02620849609375,
      "step": 1350
    },
    {
      "epoch": 2.0048,
      "grad_norm": 0.7684850692749023,
      "learning_rate": 0.00017282000000000002,
      "loss": 0.9361392974853515,
      "step": 1360
    },
    {
      "epoch": 2.0058,
      "grad_norm": 0.7847408652305603,
      "learning_rate": 0.00017262,
      "loss": 0.8885183334350586,
      "step": 1370
    },
    {
      "epoch": 2.0068,
      "grad_norm": 0.921360194683075,
      "learning_rate": 0.00017242,
      "loss": 1.0311158180236817,
      "step": 1380
    },
    {
      "epoch": 2.0078,
      "grad_norm": 0.6837793588638306,
      "learning_rate": 0.00017222,
      "loss": 1.2890826225280763,
      "step": 1390
    },
    {
      "epoch": 2.0088,
      "grad_norm": 0.7072200775146484,
      "learning_rate": 0.00017202,
      "loss": 1.209097957611084,
      "step": 1400
    },
    {
      "epoch": 2.0098,
      "grad_norm": 0.6607412695884705,
      "learning_rate": 0.00017182,
      "loss": 1.2470938682556152,
      "step": 1410
    },
    {
      "epoch": 2.0108,
      "grad_norm": 0.7476115226745605,
      "learning_rate": 0.00017162000000000001,
      "loss": 1.2279239654541017,
      "step": 1420
    },
    {
      "epoch": 2.0118,
      "grad_norm": 0.7690820693969727,
      "learning_rate": 0.00017142,
      "loss": 1.1757261276245117,
      "step": 1430
    },
    {
      "epoch": 2.0128,
      "grad_norm": 0.7281740307807922,
      "learning_rate": 0.00017122,
      "loss": 1.158098030090332,
      "step": 1440
    },
    {
      "epoch": 2.0138,
      "grad_norm": 0.7036393880844116,
      "learning_rate": 0.00017102,
      "loss": 1.1959182739257812,
      "step": 1450
    },
    {
      "epoch": 2.0148,
      "grad_norm": 0.7431781888008118,
      "learning_rate": 0.00017082,
      "loss": 1.189777946472168,
      "step": 1460
    },
    {
      "epoch": 2.0158,
      "grad_norm": 0.7411831021308899,
      "learning_rate": 0.00017062,
      "loss": 1.2544533729553222,
      "step": 1470
    },
    {
      "epoch": 2.0168,
      "grad_norm": 0.7163280844688416,
      "learning_rate": 0.00017042,
      "loss": 1.165062141418457,
      "step": 1480
    },
    {
      "epoch": 2.0178,
      "grad_norm": 0.7118193507194519,
      "learning_rate": 0.00017022,
      "loss": 1.2281935691833497,
      "step": 1490
    },
    {
      "epoch": 2.0188,
      "grad_norm": 0.6703944206237793,
      "learning_rate": 0.00017002,
      "loss": 1.1760727882385253,
      "step": 1500
    },
    {
      "epoch": 2.0198,
      "grad_norm": 0.7265837788581848,
      "learning_rate": 0.00016982,
      "loss": 1.1267470359802245,
      "step": 1510
    },
    {
      "epoch": 2.0208,
      "grad_norm": 0.8042080998420715,
      "learning_rate": 0.00016962,
      "loss": 1.0916626930236817,
      "step": 1520
    },
    {
      "epoch": 2.0218,
      "grad_norm": 0.7643582224845886,
      "learning_rate": 0.00016942000000000001,
      "loss": 1.0545287132263184,
      "step": 1530
    },
    {
      "epoch": 2.0228,
      "grad_norm": 0.7249123454093933,
      "learning_rate": 0.00016922,
      "loss": 1.1005349159240723,
      "step": 1540
    },
    {
      "epoch": 2.0238,
      "grad_norm": 0.7539435625076294,
      "learning_rate": 0.00016902,
      "loss": 1.0422410011291503,
      "step": 1550
    },
    {
      "epoch": 2.0248,
      "grad_norm": 0.7162719964981079,
      "learning_rate": 0.00016882,
      "loss": 1.039008617401123,
      "step": 1560
    },
    {
      "epoch": 2.0258,
      "grad_norm": 0.7562013268470764,
      "learning_rate": 0.00016862,
      "loss": 1.0739432334899903,
      "step": 1570
    },
    {
      "epoch": 2.0268,
      "grad_norm": 0.7008998394012451,
      "learning_rate": 0.00016842,
      "loss": 1.0126309394836426,
      "step": 1580
    },
    {
      "epoch": 2.0278,
      "grad_norm": 0.7990198731422424,
      "learning_rate": 0.00016822,
      "loss": 0.9999658584594726,
      "step": 1590
    },
    {
      "epoch": 2.0288,
      "grad_norm": 0.7978523969650269,
      "learning_rate": 0.00016802,
      "loss": 1.0024938583374023,
      "step": 1600
    },
    {
      "epoch": 2.0298,
      "grad_norm": 0.7431021928787231,
      "learning_rate": 0.00016782,
      "loss": 1.009383487701416,
      "step": 1610
    },
    {
      "epoch": 2.0308,
      "grad_norm": 0.8124121427536011,
      "learning_rate": 0.00016762,
      "loss": 0.9954951286315918,
      "step": 1620
    },
    {
      "epoch": 3.0006,
      "grad_norm": 0.7478945255279541,
      "learning_rate": 0.00016742,
      "loss": 0.9498451232910157,
      "step": 1630
    },
    {
      "epoch": 3.0016,
      "grad_norm": 0.7504536509513855,
      "learning_rate": 0.00016722,
      "loss": 0.7260428428649902,
      "step": 1640
    },
    {
      "epoch": 3.0026,
      "grad_norm": 0.6449979543685913,
      "learning_rate": 0.00016702,
      "loss": 0.6750858783721924,
      "step": 1650
    },
    {
      "epoch": 3.0036,
      "grad_norm": 0.8800742030143738,
      "learning_rate": 0.00016682,
      "loss": 0.6494207859039307,
      "step": 1660
    },
    {
      "epoch": 3.0046,
      "grad_norm": 0.674615204334259,
      "learning_rate": 0.00016662,
      "loss": 0.6307613372802734,
      "step": 1670
    },
    {
      "epoch": 3.0056,
      "grad_norm": 0.780610978603363,
      "learning_rate": 0.00016642,
      "loss": 0.5745514392852783,
      "step": 1680
    },
    {
      "epoch": 3.0066,
      "grad_norm": 0.8345136642456055,
      "learning_rate": 0.00016622,
      "loss": 0.6355406284332276,
      "step": 1690
    },
    {
      "epoch": 3.0076,
      "grad_norm": 0.7923954725265503,
      "learning_rate": 0.00016601999999999999,
      "loss": 0.7923160076141358,
      "step": 1700
    },
    {
      "epoch": 3.0086,
      "grad_norm": 0.725975513458252,
      "learning_rate": 0.00016582,
      "loss": 0.756317138671875,
      "step": 1710
    },
    {
      "epoch": 3.0096,
      "grad_norm": 0.8168895840644836,
      "learning_rate": 0.00016562,
      "loss": 0.758098840713501,
      "step": 1720
    },
    {
      "epoch": 3.0106,
      "grad_norm": 0.7553869485855103,
      "learning_rate": 0.00016542,
      "loss": 0.7652549743652344,
      "step": 1730
    },
    {
      "epoch": 3.0116,
      "grad_norm": 0.8063785433769226,
      "learning_rate": 0.00016522,
      "loss": 0.7233606815338135,
      "step": 1740
    },
    {
      "epoch": 3.0126,
      "grad_norm": 0.7430322170257568,
      "learning_rate": 0.00016502,
      "loss": 0.7092338085174561,
      "step": 1750
    },
    {
      "epoch": 3.0136,
      "grad_norm": 0.7664160132408142,
      "learning_rate": 0.00016482,
      "loss": 0.7311967372894287,
      "step": 1760
    },
    {
      "epoch": 3.0146,
      "grad_norm": 0.7792779803276062,
      "learning_rate": 0.00016462,
      "loss": 0.7741461277008057,
      "step": 1770
    },
    {
      "epoch": 3.0156,
      "grad_norm": 0.7967932224273682,
      "learning_rate": 0.00016442000000000003,
      "loss": 0.7999020576477051,
      "step": 1780
    },
    {
      "epoch": 3.0166,
      "grad_norm": 0.7274350523948669,
      "learning_rate": 0.00016422000000000002,
      "loss": 0.7347360610961914,
      "step": 1790
    },
    {
      "epoch": 3.0176,
      "grad_norm": 0.8932685256004333,
      "learning_rate": 0.00016402000000000002,
      "loss": 0.776170301437378,
      "step": 1800
    },
    {
      "epoch": 3.0186,
      "grad_norm": 0.7732952833175659,
      "learning_rate": 0.00016382000000000001,
      "loss": 0.729710578918457,
      "step": 1810
    },
    {
      "epoch": 3.0196,
      "grad_norm": 0.8386185765266418,
      "learning_rate": 0.00016362,
      "loss": 0.6981480121612549,
      "step": 1820
    },
    {
      "epoch": 3.0206,
      "grad_norm": 0.8548919558525085,
      "learning_rate": 0.00016342,
      "loss": 0.6844400405883789,
      "step": 1830
    },
    {
      "epoch": 3.0216,
      "grad_norm": 0.7728302478790283,
      "learning_rate": 0.00016322000000000003,
      "loss": 0.6626255989074707,
      "step": 1840
    },
    {
      "epoch": 3.0226,
      "grad_norm": 0.815079391002655,
      "learning_rate": 0.00016302000000000002,
      "loss": 0.6900956153869628,
      "step": 1850
    },
    {
      "epoch": 3.0236,
      "grad_norm": 0.7471504807472229,
      "learning_rate": 0.00016282000000000002,
      "loss": 0.6368459224700928,
      "step": 1860
    },
    {
      "epoch": 3.0246,
      "grad_norm": 0.7270839214324951,
      "learning_rate": 0.00016262,
      "loss": 0.6348861217498779,
      "step": 1870
    },
    {
      "epoch": 3.0256,
      "grad_norm": 0.7989698648452759,
      "learning_rate": 0.00016242,
      "loss": 0.6348659038543701,
      "step": 1880
    },
    {
      "epoch": 3.0266,
      "grad_norm": 0.7626523375511169,
      "learning_rate": 0.00016222000000000003,
      "loss": 0.5972939968109131,
      "step": 1890
    },
    {
      "epoch": 3.0276,
      "grad_norm": 0.7608401775360107,
      "learning_rate": 0.00016202000000000002,
      "loss": 0.608570671081543,
      "step": 1900
    },
    {
      "epoch": 3.0286,
      "grad_norm": 0.8012869358062744,
      "learning_rate": 0.00016182000000000002,
      "loss": 0.5898276329040527,
      "step": 1910
    },
    {
      "epoch": 3.0296,
      "grad_norm": 0.7275280356407166,
      "learning_rate": 0.00016162000000000001,
      "loss": 0.6227708339691163,
      "step": 1920
    },
    {
      "epoch": 3.0306,
      "grad_norm": 1.054669737815857,
      "learning_rate": 0.00016142,
      "loss": 0.551525354385376,
      "step": 1930
    },
    {
      "epoch": 4.0004,
      "grad_norm": 0.6590172052383423,
      "learning_rate": 0.00016122,
      "loss": 0.6316772937774658,
      "step": 1940
    },
    {
      "epoch": 4.0014,
      "grad_norm": 0.7991273999214172,
      "learning_rate": 0.00016102000000000003,
      "loss": 0.46561732292175295,
      "step": 1950
    },
    {
      "epoch": 4.0024,
      "grad_norm": 0.6764177680015564,
      "learning_rate": 0.00016082000000000002,
      "loss": 0.4100037574768066,
      "step": 1960
    },
    {
      "epoch": 4.0034,
      "grad_norm": 0.7882692813873291,
      "learning_rate": 0.00016062000000000002,
      "loss": 0.38791024684906006,
      "step": 1970
    },
    {
      "epoch": 4.0044,
      "grad_norm": 0.6683878898620605,
      "learning_rate": 0.00016042,
      "loss": 0.39162867069244384,
      "step": 1980
    },
    {
      "epoch": 4.0054,
      "grad_norm": 0.6484203934669495,
      "learning_rate": 0.00016022,
      "loss": 0.34521486759185793,
      "step": 1990
    },
    {
      "epoch": 4.0064,
      "grad_norm": 0.6811336874961853,
      "learning_rate": 0.00016002,
      "loss": 0.35950288772583006,
      "step": 2000
    },
    {
      "epoch": 1.001,
      "grad_norm": 0.4672287404537201,
      "learning_rate": 0.00015982000000000002,
      "loss": 0.3158791780471802,
      "step": 2010
    },
    {
      "epoch": 1.002,
      "grad_norm": 0.40676912665367126,
      "learning_rate": 0.00015962000000000002,
      "loss": 0.20427651405334474,
      "step": 2020
    },
    {
      "epoch": 1.003,
      "grad_norm": 0.4785628616809845,
      "learning_rate": 0.00015942000000000002,
      "loss": 0.20116724967956542,
      "step": 2030
    },
    {
      "epoch": 1.004,
      "grad_norm": 0.42470747232437134,
      "learning_rate": 0.00015922,
      "loss": 0.20162370204925537,
      "step": 2040
    },
    {
      "epoch": 1.005,
      "grad_norm": 0.44953426718711853,
      "learning_rate": 0.00015902,
      "loss": 0.19468295574188232,
      "step": 2050
    },
    {
      "epoch": 1.006,
      "grad_norm": 0.46356457471847534,
      "learning_rate": 0.00015882,
      "loss": 0.20782365798950195,
      "step": 2060
    },
    {
      "epoch": 1.007,
      "grad_norm": 0.6635887026786804,
      "learning_rate": 0.00015862000000000002,
      "loss": 0.33964028358459475,
      "step": 2070
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.7209072113037109,
      "learning_rate": 0.00015842000000000002,
      "loss": 0.45632095336914064,
      "step": 2080
    },
    {
      "epoch": 1.009,
      "grad_norm": 0.7033224105834961,
      "learning_rate": 0.00015822,
      "loss": 0.41698122024536133,
      "step": 2090
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.6875112056732178,
      "learning_rate": 0.00015802,
      "loss": 0.430418872833252,
      "step": 2100
    },
    {
      "epoch": 1.011,
      "grad_norm": 0.7253019213676453,
      "learning_rate": 0.00015782,
      "loss": 0.4426724910736084,
      "step": 2110
    },
    {
      "epoch": 1.012,
      "grad_norm": 0.7583150863647461,
      "learning_rate": 0.00015762,
      "loss": 0.41685285568237307,
      "step": 2120
    },
    {
      "epoch": 1.013,
      "grad_norm": 0.6832845211029053,
      "learning_rate": 0.00015742000000000002,
      "loss": 0.40614800453186034,
      "step": 2130
    },
    {
      "epoch": 1.014,
      "grad_norm": 0.7236650586128235,
      "learning_rate": 0.00015722000000000002,
      "loss": 0.4495222568511963,
      "step": 2140
    },
    {
      "epoch": 1.015,
      "grad_norm": 0.759689211845398,
      "learning_rate": 0.00015702,
      "loss": 0.44749298095703127,
      "step": 2150
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.7024709582328796,
      "learning_rate": 0.00015682,
      "loss": 0.44994359016418456,
      "step": 2160
    },
    {
      "epoch": 1.017,
      "grad_norm": 0.7905299663543701,
      "learning_rate": 0.00015662,
      "loss": 0.40801243782043456,
      "step": 2170
    },
    {
      "epoch": 1.018,
      "grad_norm": 0.7310312986373901,
      "learning_rate": 0.00015642000000000002,
      "loss": 0.44345922470092775,
      "step": 2180
    },
    {
      "epoch": 1.019,
      "grad_norm": 0.7063546776771545,
      "learning_rate": 0.00015622000000000002,
      "loss": 0.3925187587738037,
      "step": 2190
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.7559898495674133,
      "learning_rate": 0.00015602000000000001,
      "loss": 0.390599799156189,
      "step": 2200
    },
    {
      "epoch": 1.021,
      "grad_norm": 0.7375671863555908,
      "learning_rate": 0.00015582,
      "loss": 0.39828450679779054,
      "step": 2210
    },
    {
      "epoch": 1.022,
      "grad_norm": 0.7128247022628784,
      "learning_rate": 0.00015562,
      "loss": 0.38204762935638426,
      "step": 2220
    },
    {
      "epoch": 1.023,
      "grad_norm": 0.7096289396286011,
      "learning_rate": 0.00015542,
      "loss": 0.39499573707580565,
      "step": 2230
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.666587769985199,
      "learning_rate": 0.00015522000000000002,
      "loss": 0.3651753902435303,
      "step": 2240
    },
    {
      "epoch": 1.025,
      "grad_norm": 0.7085521817207336,
      "learning_rate": 0.00015502000000000002,
      "loss": 0.35499038696289065,
      "step": 2250
    },
    {
      "epoch": 1.026,
      "grad_norm": 0.6602509617805481,
      "learning_rate": 0.00015482,
      "loss": 0.34633963108062743,
      "step": 2260
    },
    {
      "epoch": 1.027,
      "grad_norm": 0.7021428942680359,
      "learning_rate": 0.00015462,
      "loss": 0.324503755569458,
      "step": 2270
    },
    {
      "epoch": 1.028,
      "grad_norm": 0.6079488396644592,
      "learning_rate": 0.00015442,
      "loss": 0.3342065095901489,
      "step": 2280
    },
    {
      "epoch": 1.029,
      "grad_norm": 0.6947924494743347,
      "learning_rate": 0.00015422,
      "loss": 0.3538747072219849,
      "step": 2290
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.6721782684326172,
      "learning_rate": 0.00015402000000000002,
      "loss": 0.32679312229156493,
      "step": 2300
    },
    {
      "epoch": 1.031,
      "grad_norm": 0.6266535520553589,
      "learning_rate": 0.00015382000000000001,
      "loss": 0.33827953338623046,
      "step": 2310
    },
    {
      "epoch": 2.0008,
      "grad_norm": 0.5401132702827454,
      "learning_rate": 0.00015362,
      "loss": 0.2531868934631348,
      "step": 2320
    },
    {
      "epoch": 2.0018,
      "grad_norm": 0.48838838934898376,
      "learning_rate": 0.00015342,
      "loss": 0.18140783309936523,
      "step": 2330
    },
    {
      "epoch": 2.0028,
      "grad_norm": 0.4247227609157562,
      "learning_rate": 0.00015322,
      "loss": 0.16909232139587402,
      "step": 2340
    },
    {
      "epoch": 2.0038,
      "grad_norm": 0.49411192536354065,
      "learning_rate": 0.00015302,
      "loss": 0.172965931892395,
      "step": 2350
    },
    {
      "epoch": 2.0048,
      "grad_norm": 0.4332464933395386,
      "learning_rate": 0.00015282000000000002,
      "loss": 0.1627725601196289,
      "step": 2360
    },
    {
      "epoch": 2.0058,
      "grad_norm": 0.44918233156204224,
      "learning_rate": 0.00015262,
      "loss": 0.150301730632782,
      "step": 2370
    },
    {
      "epoch": 2.0068,
      "grad_norm": 0.7051998376846313,
      "learning_rate": 0.00015242,
      "loss": 0.17362732887268068,
      "step": 2380
    },
    {
      "epoch": 2.0078,
      "grad_norm": 0.5488992929458618,
      "learning_rate": 0.00015222,
      "loss": 0.22459635734558106,
      "step": 2390
    },
    {
      "epoch": 2.0088,
      "grad_norm": 0.6187590956687927,
      "learning_rate": 0.00015202,
      "loss": 0.23692691326141357,
      "step": 2400
    },
    {
      "epoch": 2.0098,
      "grad_norm": 0.5993273258209229,
      "learning_rate": 0.00015182,
      "loss": 0.29602196216583254,
      "step": 2410
    },
    {
      "epoch": 2.0108,
      "grad_norm": 0.6054103970527649,
      "learning_rate": 0.00015162000000000002,
      "loss": 0.2574064016342163,
      "step": 2420
    },
    {
      "epoch": 2.0118,
      "grad_norm": 0.592335045337677,
      "learning_rate": 0.00015142,
      "loss": 0.24514660835266114,
      "step": 2430
    },
    {
      "epoch": 2.0128,
      "grad_norm": 0.5396589040756226,
      "learning_rate": 0.00015122,
      "loss": 0.23127524852752684,
      "step": 2440
    },
    {
      "epoch": 2.0138,
      "grad_norm": 0.5918888449668884,
      "learning_rate": 0.00015102,
      "loss": 0.24235455989837645,
      "step": 2450
    },
    {
      "epoch": 2.0148,
      "grad_norm": 0.5735516548156738,
      "learning_rate": 0.00015082,
      "loss": 0.25308713912963865,
      "step": 2460
    },
    {
      "epoch": 2.0158,
      "grad_norm": 0.5613916516304016,
      "learning_rate": 0.00015062000000000002,
      "loss": 0.292048716545105,
      "step": 2470
    },
    {
      "epoch": 2.0168,
      "grad_norm": 0.5100058913230896,
      "learning_rate": 0.00015042,
      "loss": 0.2264169454574585,
      "step": 2480
    },
    {
      "epoch": 2.0178,
      "grad_norm": 0.5964860320091248,
      "learning_rate": 0.00015022,
      "loss": 0.2513779878616333,
      "step": 2490
    },
    {
      "epoch": 2.0188,
      "grad_norm": 0.5261106491088867,
      "learning_rate": 0.00015002,
      "loss": 0.22608683109283448,
      "step": 2500
    },
    {
      "epoch": 2.0198,
      "grad_norm": 0.5102221369743347,
      "learning_rate": 0.00014982,
      "loss": 0.21579885482788086,
      "step": 2510
    },
    {
      "epoch": 2.0208,
      "grad_norm": 0.5863536596298218,
      "learning_rate": 0.00014962,
      "loss": 0.2347024440765381,
      "step": 2520
    },
    {
      "epoch": 2.0218,
      "grad_norm": 0.5884446501731873,
      "learning_rate": 0.00014942000000000002,
      "loss": 0.2166447401046753,
      "step": 2530
    },
    {
      "epoch": 2.0228,
      "grad_norm": 0.5152504444122314,
      "learning_rate": 0.00014922,
      "loss": 0.23897838592529297,
      "step": 2540
    },
    {
      "epoch": 2.0238,
      "grad_norm": 0.5819721221923828,
      "learning_rate": 0.00014902,
      "loss": 0.2348470449447632,
      "step": 2550
    },
    {
      "epoch": 2.0248,
      "grad_norm": 0.5081148147583008,
      "learning_rate": 0.00014882,
      "loss": 0.2127830982208252,
      "step": 2560
    },
    {
      "epoch": 2.0258,
      "grad_norm": 0.5137162804603577,
      "learning_rate": 0.00014862,
      "loss": 0.20368046760559083,
      "step": 2570
    },
    {
      "epoch": 2.0268,
      "grad_norm": 0.45256203413009644,
      "learning_rate": 0.00014842,
      "loss": 0.19960324764251708,
      "step": 2580
    },
    {
      "epoch": 2.0278,
      "grad_norm": 0.5101103782653809,
      "learning_rate": 0.00014822000000000001,
      "loss": 0.19512444734573364,
      "step": 2590
    },
    {
      "epoch": 2.0288,
      "grad_norm": 0.5388202667236328,
      "learning_rate": 0.00014802,
      "loss": 0.20622694492340088,
      "step": 2600
    },
    {
      "epoch": 2.0298,
      "grad_norm": 0.5038774013519287,
      "learning_rate": 0.00014782,
      "loss": 0.19988502264022828,
      "step": 2610
    },
    {
      "epoch": 2.0308,
      "grad_norm": 0.5558412671089172,
      "learning_rate": 0.00014762,
      "loss": 0.20110974311828614,
      "step": 2620
    },
    {
      "epoch": 3.0006,
      "grad_norm": 0.43655768036842346,
      "learning_rate": 0.00014742,
      "loss": 0.1968843460083008,
      "step": 2630
    },
    {
      "epoch": 3.0016,
      "grad_norm": 0.4135885536670685,
      "learning_rate": 0.00014722,
      "loss": 0.13574039936065674,
      "step": 2640
    },
    {
      "epoch": 3.0026,
      "grad_norm": 0.364497572183609,
      "learning_rate": 0.00014702,
      "loss": 0.12638025283813475,
      "step": 2650
    },
    {
      "epoch": 3.0036,
      "grad_norm": 0.4627242982387543,
      "learning_rate": 0.00014682,
      "loss": 0.1232415795326233,
      "step": 2660
    },
    {
      "epoch": 3.0046,
      "grad_norm": 0.4199890196323395,
      "learning_rate": 0.00014662,
      "loss": 0.1209063172340393,
      "step": 2670
    },
    {
      "epoch": 3.0056,
      "grad_norm": 0.3729947507381439,
      "learning_rate": 0.00014642,
      "loss": 0.11018155813217163,
      "step": 2680
    },
    {
      "epoch": 3.0066,
      "grad_norm": 0.49802160263061523,
      "learning_rate": 0.00014622,
      "loss": 0.12115752696990967,
      "step": 2690
    },
    {
      "epoch": 3.0076,
      "grad_norm": 0.5387603044509888,
      "learning_rate": 0.00014602,
      "loss": 0.1478494167327881,
      "step": 2700
    },
    {
      "epoch": 3.0086,
      "grad_norm": 0.46145540475845337,
      "learning_rate": 0.00014582,
      "loss": 0.14995505809783935,
      "step": 2710
    },
    {
      "epoch": 3.0096,
      "grad_norm": 0.4195813834667206,
      "learning_rate": 0.00014562,
      "loss": 0.16438782215118408,
      "step": 2720
    },
    {
      "epoch": 3.0106,
      "grad_norm": 0.4632911682128906,
      "learning_rate": 0.00014542,
      "loss": 0.16314655542373657,
      "step": 2730
    },
    {
      "epoch": 3.0116,
      "grad_norm": 0.49666211009025574,
      "learning_rate": 0.00014522,
      "loss": 0.1643773317337036,
      "step": 2740
    },
    {
      "epoch": 3.0126,
      "grad_norm": 0.4677850306034088,
      "learning_rate": 0.00014502,
      "loss": 0.14674897193908693,
      "step": 2750
    },
    {
      "epoch": 3.0136,
      "grad_norm": 0.48538899421691895,
      "learning_rate": 0.00014482,
      "loss": 0.151479172706604,
      "step": 2760
    },
    {
      "epoch": 3.0146,
      "grad_norm": 0.4430471360683441,
      "learning_rate": 0.00014462,
      "loss": 0.16841667890548706,
      "step": 2770
    },
    {
      "epoch": 3.0156,
      "grad_norm": 0.5144022107124329,
      "learning_rate": 0.00014442,
      "loss": 0.16647714376449585,
      "step": 2780
    },
    {
      "epoch": 3.0166,
      "grad_norm": 0.45523014664649963,
      "learning_rate": 0.00014422,
      "loss": 0.14415124654769898,
      "step": 2790
    },
    {
      "epoch": 3.0176,
      "grad_norm": 0.5218808054924011,
      "learning_rate": 0.00014402,
      "loss": 0.15597842931747435,
      "step": 2800
    },
    {
      "epoch": 3.0186,
      "grad_norm": 0.4452967941761017,
      "learning_rate": 0.00014382,
      "loss": 0.1488332986831665,
      "step": 2810
    },
    {
      "epoch": 3.0196,
      "grad_norm": 0.4646563231945038,
      "learning_rate": 0.00014362,
      "loss": 0.1394293189048767,
      "step": 2820
    },
    {
      "epoch": 3.0206,
      "grad_norm": 0.5117284059524536,
      "learning_rate": 0.00014342,
      "loss": 0.15330575704574584,
      "step": 2830
    },
    {
      "epoch": 3.0216,
      "grad_norm": 0.4975617825984955,
      "learning_rate": 0.00014322,
      "loss": 0.1486660957336426,
      "step": 2840
    },
    {
      "epoch": 3.0226,
      "grad_norm": 0.4680009186267853,
      "learning_rate": 0.00014302,
      "loss": 0.15574642419815063,
      "step": 2850
    },
    {
      "epoch": 3.0236,
      "grad_norm": 0.45793306827545166,
      "learning_rate": 0.00014282,
      "loss": 0.1538609027862549,
      "step": 2860
    },
    {
      "epoch": 3.0246,
      "grad_norm": 0.43385592103004456,
      "learning_rate": 0.00014261999999999999,
      "loss": 0.14147435426712035,
      "step": 2870
    },
    {
      "epoch": 3.0256,
      "grad_norm": 0.48055240511894226,
      "learning_rate": 0.00014242,
      "loss": 0.13285771608352662,
      "step": 2880
    },
    {
      "epoch": 3.0266,
      "grad_norm": 0.426279217004776,
      "learning_rate": 0.00014222,
      "loss": 0.13433635234832764,
      "step": 2890
    },
    {
      "epoch": 3.0276,
      "grad_norm": 0.4118101894855499,
      "learning_rate": 0.00014202,
      "loss": 0.13033357858657837,
      "step": 2900
    },
    {
      "epoch": 3.0286,
      "grad_norm": 0.4413125216960907,
      "learning_rate": 0.00014182,
      "loss": 0.1351193904876709,
      "step": 2910
    },
    {
      "epoch": 3.0296,
      "grad_norm": 0.41682061553001404,
      "learning_rate": 0.00014162,
      "loss": 0.13452460765838622,
      "step": 2920
    },
    {
      "epoch": 3.0306,
      "grad_norm": 0.5230453610420227,
      "learning_rate": 0.00014141999999999998,
      "loss": 0.1290982484817505,
      "step": 2930
    },
    {
      "epoch": 4.0004,
      "grad_norm": 0.3429013788700104,
      "learning_rate": 0.00014122,
      "loss": 0.14483762979507447,
      "step": 2940
    },
    {
      "epoch": 4.0014,
      "grad_norm": 0.32104936242103577,
      "learning_rate": 0.00014102,
      "loss": 0.10209068059921264,
      "step": 2950
    },
    {
      "epoch": 4.0024,
      "grad_norm": 0.29722949862480164,
      "learning_rate": 0.00014082,
      "loss": 0.0967288076877594,
      "step": 2960
    },
    {
      "epoch": 4.0034,
      "grad_norm": 0.4013476073741913,
      "learning_rate": 0.00014062,
      "loss": 0.08881974220275879,
      "step": 2970
    },
    {
      "epoch": 4.0044,
      "grad_norm": 0.37912485003471375,
      "learning_rate": 0.00014042,
      "loss": 0.09519890546798707,
      "step": 2980
    },
    {
      "epoch": 4.0054,
      "grad_norm": 0.31893491744995117,
      "learning_rate": 0.00014022,
      "loss": 0.08268740773200989,
      "step": 2990
    },
    {
      "epoch": 4.0064,
      "grad_norm": 0.3440733850002289,
      "learning_rate": 0.00014002,
      "loss": 0.08511788249015809,
      "step": 3000
    },
    {
      "epoch": 4.0074,
      "grad_norm": 0.37289556860923767,
      "learning_rate": 0.00013982000000000003,
      "loss": 0.10226565599441528,
      "step": 3010
    },
    {
      "epoch": 4.0084,
      "grad_norm": 0.36419790983200073,
      "learning_rate": 0.00013962000000000002,
      "loss": 0.0977319598197937,
      "step": 3020
    },
    {
      "epoch": 4.0094,
      "grad_norm": 0.347756952047348,
      "learning_rate": 0.00013942000000000002,
      "loss": 0.10832228660583496,
      "step": 3030
    },
    {
      "epoch": 4.0104,
      "grad_norm": 0.40426164865493774,
      "learning_rate": 0.00013922,
      "loss": 0.11001000404357911,
      "step": 3040
    },
    {
      "epoch": 4.0114,
      "grad_norm": 0.3765728771686554,
      "learning_rate": 0.00013902,
      "loss": 0.10214830636978149,
      "step": 3050
    },
    {
      "epoch": 4.0124,
      "grad_norm": 0.37442028522491455,
      "learning_rate": 0.00013882000000000003,
      "loss": 0.10177899599075317,
      "step": 3060
    },
    {
      "epoch": 4.0134,
      "grad_norm": 0.3760021924972534,
      "learning_rate": 0.00013862000000000002,
      "loss": 0.10235388278961181,
      "step": 3070
    },
    {
      "epoch": 4.0144,
      "grad_norm": 0.40187352895736694,
      "learning_rate": 0.00013842000000000002,
      "loss": 0.10720946788787841,
      "step": 3080
    },
    {
      "epoch": 4.0154,
      "grad_norm": 0.33189886808395386,
      "learning_rate": 0.00013822000000000001,
      "loss": 0.12659095525741576,
      "step": 3090
    },
    {
      "epoch": 4.0164,
      "grad_norm": 0.35277223587036133,
      "learning_rate": 0.00013802,
      "loss": 0.10271661281585694,
      "step": 3100
    },
    {
      "epoch": 4.0174,
      "grad_norm": 0.37806376814842224,
      "learning_rate": 0.00013782,
      "loss": 0.10171124935150147,
      "step": 3110
    },
    {
      "epoch": 4.0184,
      "grad_norm": 0.38206177949905396,
      "learning_rate": 0.00013762000000000003,
      "loss": 0.10048867464065551,
      "step": 3120
    },
    {
      "epoch": 4.0194,
      "grad_norm": 0.36296215653419495,
      "learning_rate": 0.00013742000000000002,
      "loss": 0.09387139678001404,
      "step": 3130
    },
    {
      "epoch": 4.0204,
      "grad_norm": 0.3241318166255951,
      "learning_rate": 0.00013722000000000002,
      "loss": 0.09321512579917908,
      "step": 3140
    },
    {
      "epoch": 4.0214,
      "grad_norm": 0.3707175552845001,
      "learning_rate": 0.00013702,
      "loss": 0.10409607887268066,
      "step": 3150
    },
    {
      "epoch": 4.0224,
      "grad_norm": 0.35573673248291016,
      "learning_rate": 0.00013682,
      "loss": 0.09896103143692017,
      "step": 3160
    },
    {
      "epoch": 4.0234,
      "grad_norm": 0.4057498276233673,
      "learning_rate": 0.00013662,
      "loss": 0.11238985061645508,
      "step": 3170
    },
    {
      "epoch": 4.0244,
      "grad_norm": 0.34688249230384827,
      "learning_rate": 0.00013642000000000003,
      "loss": 0.09726371765136718,
      "step": 3180
    },
    {
      "epoch": 4.0254,
      "grad_norm": 0.36548253893852234,
      "learning_rate": 0.00013622000000000002,
      "loss": 0.09118783473968506,
      "step": 3190
    },
    {
      "epoch": 4.0264,
      "grad_norm": 0.3322118818759918,
      "learning_rate": 0.00013602000000000002,
      "loss": 0.08653171062469482,
      "step": 3200
    },
    {
      "epoch": 4.0274,
      "grad_norm": 0.35557135939598083,
      "learning_rate": 0.00013582,
      "loss": 0.08778741359710693,
      "step": 3210
    },
    {
      "epoch": 4.0284,
      "grad_norm": 0.3009137511253357,
      "learning_rate": 0.00013562,
      "loss": 0.09119086265563965,
      "step": 3220
    },
    {
      "epoch": 4.0294,
      "grad_norm": 0.3856281340122223,
      "learning_rate": 0.00013542,
      "loss": 0.09039224982261658,
      "step": 3230
    },
    {
      "epoch": 4.0304,
      "grad_norm": 0.33204856514930725,
      "learning_rate": 0.00013522000000000002,
      "loss": 0.08515154123306275,
      "step": 3240
    },
    {
      "epoch": 5.0002,
      "grad_norm": 0.31765300035476685,
      "learning_rate": 0.00013502000000000002,
      "loss": 0.10655074119567871,
      "step": 3250
    },
    {
      "epoch": 5.0012,
      "grad_norm": 0.26351630687713623,
      "learning_rate": 0.00013482000000000001,
      "loss": 0.06924263834953308,
      "step": 3260
    },
    {
      "epoch": 5.0022,
      "grad_norm": 0.2613130211830139,
      "learning_rate": 0.00013462,
      "loss": 0.07073780298233032,
      "step": 3270
    },
    {
      "epoch": 5.0032,
      "grad_norm": 0.2842184603214264,
      "learning_rate": 0.00013442,
      "loss": 0.06662424206733704,
      "step": 3280
    },
    {
      "epoch": 5.0042,
      "grad_norm": 0.2726162075996399,
      "learning_rate": 0.00013422,
      "loss": 0.07099056243896484,
      "step": 3290
    },
    {
      "epoch": 5.0052,
      "grad_norm": 0.302075058221817,
      "learning_rate": 0.00013402000000000002,
      "loss": 0.06526976823806763,
      "step": 3300
    },
    {
      "epoch": 5.0062,
      "grad_norm": 0.310641884803772,
      "learning_rate": 0.00013382000000000002,
      "loss": 0.06162307262420654,
      "step": 3310
    },
    {
      "epoch": 5.0072,
      "grad_norm": 0.32952383160591125,
      "learning_rate": 0.00013362,
      "loss": 0.07002668380737305,
      "step": 3320
    },
    {
      "epoch": 5.0082,
      "grad_norm": 0.30953407287597656,
      "learning_rate": 0.00013342,
      "loss": 0.0669539988040924,
      "step": 3330
    },
    {
      "epoch": 5.0092,
      "grad_norm": 0.3306865692138672,
      "learning_rate": 0.00013322,
      "loss": 0.07151323556900024,
      "step": 3340
    },
    {
      "epoch": 5.0102,
      "grad_norm": 0.28720608353614807,
      "learning_rate": 0.00013302000000000002,
      "loss": 0.07206830978393555,
      "step": 3350
    },
    {
      "epoch": 5.0112,
      "grad_norm": 0.29193148016929626,
      "learning_rate": 0.00013282000000000002,
      "loss": 0.0766656219959259,
      "step": 3360
    },
    {
      "epoch": 5.0122,
      "grad_norm": 0.2924456298351288,
      "learning_rate": 0.00013262000000000001,
      "loss": 0.06823275685310363,
      "step": 3370
    },
    {
      "epoch": 5.0132,
      "grad_norm": 0.2984090745449066,
      "learning_rate": 0.00013242,
      "loss": 0.06823591589927673,
      "step": 3380
    },
    {
      "epoch": 5.0142,
      "grad_norm": 0.30468592047691345,
      "learning_rate": 0.00013222,
      "loss": 0.07523037791252137,
      "step": 3390
    },
    {
      "epoch": 5.0152,
      "grad_norm": 0.32693299651145935,
      "learning_rate": 0.00013202,
      "loss": 0.09835698008537293,
      "step": 3400
    },
    {
      "epoch": 5.0162,
      "grad_norm": 0.34058642387390137,
      "learning_rate": 0.00013182000000000002,
      "loss": 0.07129377722740174,
      "step": 3410
    },
    {
      "epoch": 5.0172,
      "grad_norm": 0.31692641973495483,
      "learning_rate": 0.00013162000000000002,
      "loss": 0.06967606544494628,
      "step": 3420
    },
    {
      "epoch": 5.0182,
      "grad_norm": 0.27472278475761414,
      "learning_rate": 0.00013142,
      "loss": 0.06811057925224304,
      "step": 3430
    },
    {
      "epoch": 5.0192,
      "grad_norm": 0.3070657253265381,
      "learning_rate": 0.00013122,
      "loss": 0.06674447655677795,
      "step": 3440
    },
    {
      "epoch": 5.0202,
      "grad_norm": 0.31556063890457153,
      "learning_rate": 0.00013102,
      "loss": 0.0634992241859436,
      "step": 3450
    },
    {
      "epoch": 5.0212,
      "grad_norm": 0.32814645767211914,
      "learning_rate": 0.00013082,
      "loss": 0.07011401653289795,
      "step": 3460
    },
    {
      "epoch": 5.0222,
      "grad_norm": 0.32829102873802185,
      "learning_rate": 0.00013062000000000002,
      "loss": 0.06834354400634765,
      "step": 3470
    },
    {
      "epoch": 5.0232,
      "grad_norm": 0.27184948325157166,
      "learning_rate": 0.00013042000000000002,
      "loss": 0.07986602783203126,
      "step": 3480
    },
    {
      "epoch": 5.0242,
      "grad_norm": 0.2695474624633789,
      "learning_rate": 0.00013022,
      "loss": 0.06707475185394288,
      "step": 3490
    },
    {
      "epoch": 5.0252,
      "grad_norm": 0.3358134627342224,
      "learning_rate": 0.00013002,
      "loss": 0.06365590691566467,
      "step": 3500
    },
    {
      "epoch": 5.0262,
      "grad_norm": 0.2963351309299469,
      "learning_rate": 0.00012982,
      "loss": 0.06292765736579894,
      "step": 3510
    },
    {
      "epoch": 5.0272,
      "grad_norm": 0.24942268431186676,
      "learning_rate": 0.00012962,
      "loss": 0.061947989463806155,
      "step": 3520
    },
    {
      "epoch": 5.0282,
      "grad_norm": 0.31199055910110474,
      "learning_rate": 0.00012942000000000002,
      "loss": 0.06328007578849792,
      "step": 3530
    },
    {
      "epoch": 5.0292,
      "grad_norm": 0.2745618522167206,
      "learning_rate": 0.00012922,
      "loss": 0.0613422691822052,
      "step": 3540
    },
    {
      "epoch": 5.0302,
      "grad_norm": 0.3146347999572754,
      "learning_rate": 0.00012902,
      "loss": 0.05851168036460876,
      "step": 3550
    },
    {
      "epoch": 5.0312,
      "grad_norm": 0.27351754903793335,
      "learning_rate": 0.00012882,
      "loss": 0.0684496819972992,
      "step": 3560
    },
    {
      "epoch": 6.001,
      "grad_norm": 0.2317938506603241,
      "learning_rate": 0.00012862,
      "loss": 0.06254401206970214,
      "step": 3570
    },
    {
      "epoch": 6.002,
      "grad_norm": 0.24452786147594452,
      "learning_rate": 0.00012842,
      "loss": 0.053379553556442264,
      "step": 3580
    },
    {
      "epoch": 6.003,
      "grad_norm": 0.2590080201625824,
      "learning_rate": 0.00012822000000000002,
      "loss": 0.052920466661453246,
      "step": 3590
    },
    {
      "epoch": 6.004,
      "grad_norm": 0.2468854933977127,
      "learning_rate": 0.00012802,
      "loss": 0.0584908664226532,
      "step": 3600
    },
    {
      "epoch": 6.005,
      "grad_norm": 0.24238893389701843,
      "learning_rate": 0.00012782,
      "loss": 0.05077446103096008,
      "step": 3610
    },
    {
      "epoch": 6.006,
      "grad_norm": 0.21615304052829742,
      "learning_rate": 0.00012762,
      "loss": 0.047195515036582945,
      "step": 3620
    },
    {
      "epoch": 6.007,
      "grad_norm": 0.2538570761680603,
      "learning_rate": 0.00012742,
      "loss": 0.05251326560974121,
      "step": 3630
    },
    {
      "epoch": 6.008,
      "grad_norm": 0.21952210366725922,
      "learning_rate": 0.00012722000000000002,
      "loss": 0.04804377257823944,
      "step": 3640
    },
    {
      "epoch": 6.009,
      "grad_norm": 0.2638396620750427,
      "learning_rate": 0.00012702000000000001,
      "loss": 0.05005112886428833,
      "step": 3650
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.2512870728969574,
      "learning_rate": 0.00012682,
      "loss": 0.05190543532371521,
      "step": 3660
    },
    {
      "epoch": 6.011,
      "grad_norm": 0.26794344186782837,
      "learning_rate": 0.00012662,
      "loss": 0.05652360320091247,
      "step": 3670
    },
    {
      "epoch": 6.012,
      "grad_norm": 0.24694757163524628,
      "learning_rate": 0.00012642,
      "loss": 0.05183016061782837,
      "step": 3680
    },
    {
      "epoch": 6.013,
      "grad_norm": 0.27098706364631653,
      "learning_rate": 0.00012622,
      "loss": 0.05123690366744995,
      "step": 3690
    },
    {
      "epoch": 6.014,
      "grad_norm": 0.23455461859703064,
      "learning_rate": 0.00012602000000000002,
      "loss": 0.05616133213043213,
      "step": 3700
    },
    {
      "epoch": 6.015,
      "grad_norm": 0.23430338501930237,
      "learning_rate": 0.00012582,
      "loss": 0.05355681180953979,
      "step": 3710
    },
    {
      "epoch": 6.016,
      "grad_norm": 0.23048514127731323,
      "learning_rate": 0.00012562,
      "loss": 0.11050782203674317,
      "step": 3720
    },
    {
      "epoch": 6.017,
      "grad_norm": 0.2567964494228363,
      "learning_rate": 0.00012542,
      "loss": 0.05631558299064636,
      "step": 3730
    },
    {
      "epoch": 6.018,
      "grad_norm": 0.3105704188346863,
      "learning_rate": 0.00012522,
      "loss": 0.05600916147232056,
      "step": 3740
    },
    {
      "epoch": 6.019,
      "grad_norm": 0.25742048025131226,
      "learning_rate": 0.00012502,
      "loss": 0.0482271134853363,
      "step": 3750
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.2826644480228424,
      "learning_rate": 0.00012482000000000001,
      "loss": 0.050260508060455324,
      "step": 3760
    },
    {
      "epoch": 6.021,
      "grad_norm": 0.2458411604166031,
      "learning_rate": 0.00012462,
      "loss": 0.05451898574829102,
      "step": 3770
    },
    {
      "epoch": 6.022,
      "grad_norm": 0.26720693707466125,
      "learning_rate": 0.00012442,
      "loss": 0.05155031681060791,
      "step": 3780
    },
    {
      "epoch": 6.023,
      "grad_norm": 0.24704712629318237,
      "learning_rate": 0.00012422,
      "loss": 0.06429644227027893,
      "step": 3790
    },
    {
      "epoch": 6.024,
      "grad_norm": 0.23193997144699097,
      "learning_rate": 0.00012402,
      "loss": 0.051171398162841795,
      "step": 3800
    },
    {
      "epoch": 6.025,
      "grad_norm": 0.2553248405456543,
      "learning_rate": 0.00012382,
      "loss": 0.04670791625976563,
      "step": 3810
    },
    {
      "epoch": 6.026,
      "grad_norm": 0.22213296592235565,
      "learning_rate": 0.00012362,
      "loss": 0.04988162517547608,
      "step": 3820
    },
    {
      "epoch": 6.027,
      "grad_norm": 0.288107305765152,
      "learning_rate": 0.00012342,
      "loss": 0.04665826857089996,
      "step": 3830
    },
    {
      "epoch": 6.028,
      "grad_norm": 0.21731558442115784,
      "learning_rate": 0.00012322,
      "loss": 0.04808914065361023,
      "step": 3840
    },
    {
      "epoch": 6.029,
      "grad_norm": 0.22236528992652893,
      "learning_rate": 0.00012302,
      "loss": 0.04851194024085999,
      "step": 3850
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.23472720384597778,
      "learning_rate": 0.00012282,
      "loss": 0.04830377697944641,
      "step": 3860
    },
    {
      "epoch": 6.031,
      "grad_norm": 0.2267661690711975,
      "learning_rate": 0.00012262,
      "loss": 0.050290358066558835,
      "step": 3870
    },
    {
      "epoch": 7.0008,
      "grad_norm": 0.24839870631694794,
      "learning_rate": 0.00012242,
      "loss": 0.051380503177642825,
      "step": 3880
    },
    {
      "epoch": 7.0018,
      "grad_norm": 0.21834354102611542,
      "learning_rate": 0.00012222,
      "loss": 0.04199742078781128,
      "step": 3890
    },
    {
      "epoch": 7.0028,
      "grad_norm": 0.22921745479106903,
      "learning_rate": 0.00012202,
      "loss": 0.04476028978824616,
      "step": 3900
    },
    {
      "epoch": 7.0038,
      "grad_norm": 0.2517320215702057,
      "learning_rate": 0.00012182,
      "loss": 0.04470941722393036,
      "step": 3910
    },
    {
      "epoch": 7.0048,
      "grad_norm": 0.23211771249771118,
      "learning_rate": 0.00012162,
      "loss": 0.04450752139091492,
      "step": 3920
    },
    {
      "epoch": 7.0058,
      "grad_norm": 0.21550092101097107,
      "learning_rate": 0.00012142,
      "loss": 0.03875753879547119,
      "step": 3930
    },
    {
      "epoch": 7.0068,
      "grad_norm": 0.25320014357566833,
      "learning_rate": 0.00012122,
      "loss": 0.04415590167045593,
      "step": 3940
    },
    {
      "epoch": 7.0078,
      "grad_norm": 0.19818426668643951,
      "learning_rate": 0.00012102,
      "loss": 0.04161977171897888,
      "step": 3950
    },
    {
      "epoch": 7.0088,
      "grad_norm": 0.20641763508319855,
      "learning_rate": 0.00012082,
      "loss": 0.04034121632575989,
      "step": 3960
    },
    {
      "epoch": 7.0098,
      "grad_norm": 0.19377565383911133,
      "learning_rate": 0.00012062,
      "loss": 0.04172587096691131,
      "step": 3970
    },
    {
      "epoch": 7.0108,
      "grad_norm": 0.18526528775691986,
      "learning_rate": 0.00012042,
      "loss": 0.04182737171649933,
      "step": 3980
    },
    {
      "epoch": 7.0118,
      "grad_norm": 0.2163708209991455,
      "learning_rate": 0.00012022,
      "loss": 0.042621687054634094,
      "step": 3990
    },
    {
      "epoch": 7.0128,
      "grad_norm": 0.19984810054302216,
      "learning_rate": 0.00012001999999999999,
      "loss": 0.04216248393058777,
      "step": 4000
    },
    {
      "epoch": 7.0138,
      "grad_norm": 0.20787842571735382,
      "learning_rate": 0.00011982,
      "loss": 0.04524537920951843,
      "step": 4010
    },
    {
      "epoch": 7.0148,
      "grad_norm": 0.22596226632595062,
      "learning_rate": 0.00011962,
      "loss": 0.04263944923877716,
      "step": 4020
    },
    {
      "epoch": 7.0158,
      "grad_norm": 0.2565840780735016,
      "learning_rate": 0.00011942,
      "loss": 0.07590364217758179,
      "step": 4030
    },
    {
      "epoch": 7.0168,
      "grad_norm": 0.21802914142608643,
      "learning_rate": 0.00011922,
      "loss": 0.04318772554397583,
      "step": 4040
    },
    {
      "epoch": 7.0178,
      "grad_norm": 0.24467627704143524,
      "learning_rate": 0.00011902,
      "loss": 0.04431358277797699,
      "step": 4050
    },
    {
      "epoch": 7.0188,
      "grad_norm": 0.18366847932338715,
      "learning_rate": 0.00011882,
      "loss": 0.0405062198638916,
      "step": 4060
    },
    {
      "epoch": 7.0198,
      "grad_norm": 0.1755930334329605,
      "learning_rate": 0.00011862,
      "loss": 0.03957487344741821,
      "step": 4070
    },
    {
      "epoch": 7.0208,
      "grad_norm": 0.2335081398487091,
      "learning_rate": 0.00011842,
      "loss": 0.04518919289112091,
      "step": 4080
    },
    {
      "epoch": 7.0218,
      "grad_norm": 0.21695205569267273,
      "learning_rate": 0.00011822,
      "loss": 0.03841679692268372,
      "step": 4090
    },
    {
      "epoch": 7.0228,
      "grad_norm": 0.21446512639522552,
      "learning_rate": 0.00011802,
      "loss": 0.04169888198375702,
      "step": 4100
    },
    {
      "epoch": 7.0238,
      "grad_norm": 0.2171398401260376,
      "learning_rate": 0.00011782,
      "loss": 0.04906872808933258,
      "step": 4110
    },
    {
      "epoch": 7.0248,
      "grad_norm": 0.20123597979545593,
      "learning_rate": 0.00011762,
      "loss": 0.03715299665927887,
      "step": 4120
    },
    {
      "epoch": 7.0258,
      "grad_norm": 0.18073002994060516,
      "learning_rate": 0.00011742,
      "loss": 0.03863258361816406,
      "step": 4130
    },
    {
      "epoch": 7.0268,
      "grad_norm": 0.19887232780456543,
      "learning_rate": 0.00011721999999999999,
      "loss": 0.040367615222930905,
      "step": 4140
    },
    {
      "epoch": 7.0278,
      "grad_norm": 0.19514504075050354,
      "learning_rate": 0.00011702,
      "loss": 0.039677688479423524,
      "step": 4150
    },
    {
      "epoch": 7.0288,
      "grad_norm": 0.22709649801254272,
      "learning_rate": 0.00011682,
      "loss": 0.03945083320140839,
      "step": 4160
    },
    {
      "epoch": 7.0298,
      "grad_norm": 0.18606211245059967,
      "learning_rate": 0.00011661999999999999,
      "loss": 0.04064054489135742,
      "step": 4170
    },
    {
      "epoch": 7.0308,
      "grad_norm": 0.21493422985076904,
      "learning_rate": 0.00011642,
      "loss": 0.04021798074245453,
      "step": 4180
    },
    {
      "epoch": 8.0006,
      "grad_norm": 0.19142334163188934,
      "learning_rate": 0.00011622,
      "loss": 0.03999471664428711,
      "step": 4190
    },
    {
      "epoch": 8.0016,
      "grad_norm": 0.1997303068637848,
      "learning_rate": 0.00011601999999999999,
      "loss": 0.03443270623683929,
      "step": 4200
    },
    {
      "epoch": 8.0026,
      "grad_norm": 0.18130983412265778,
      "learning_rate": 0.00011582,
      "loss": 0.039177173376083375,
      "step": 4210
    },
    {
      "epoch": 8.0036,
      "grad_norm": 0.49426183104515076,
      "learning_rate": 0.00011562,
      "loss": 0.038170018792152406,
      "step": 4220
    },
    {
      "epoch": 8.0046,
      "grad_norm": 0.20743317902088165,
      "learning_rate": 0.00011541999999999999,
      "loss": 0.0379817396402359,
      "step": 4230
    },
    {
      "epoch": 8.0056,
      "grad_norm": 0.18070366978645325,
      "learning_rate": 0.00011522,
      "loss": 0.036043813824653624,
      "step": 4240
    },
    {
      "epoch": 8.0066,
      "grad_norm": 0.22303594648838043,
      "learning_rate": 0.00011501999999999999,
      "loss": 0.0366193026304245,
      "step": 4250
    },
    {
      "epoch": 8.0076,
      "grad_norm": 0.23021455109119415,
      "learning_rate": 0.00011482000000000002,
      "loss": 0.03866423368453979,
      "step": 4260
    },
    {
      "epoch": 8.0086,
      "grad_norm": 0.18778809905052185,
      "learning_rate": 0.00011462000000000001,
      "loss": 0.035205498337745667,
      "step": 4270
    },
    {
      "epoch": 8.0096,
      "grad_norm": 0.20131027698516846,
      "learning_rate": 0.00011442000000000002,
      "loss": 0.03923934400081634,
      "step": 4280
    },
    {
      "epoch": 8.0106,
      "grad_norm": 0.21135014295578003,
      "learning_rate": 0.00011422000000000001,
      "loss": 0.0388818234205246,
      "step": 4290
    },
    {
      "epoch": 8.0116,
      "grad_norm": 0.21640531718730927,
      "learning_rate": 0.00011402000000000001,
      "loss": 0.038308563828468326,
      "step": 4300
    },
    {
      "epoch": 8.0126,
      "grad_norm": 0.19794313609600067,
      "learning_rate": 0.00011382000000000002,
      "loss": 0.03605532646179199,
      "step": 4310
    },
    {
      "epoch": 8.0136,
      "grad_norm": 0.20682232081890106,
      "learning_rate": 0.00011362000000000001,
      "loss": 0.03772588968276978,
      "step": 4320
    },
    {
      "epoch": 8.0146,
      "grad_norm": 0.1675058901309967,
      "learning_rate": 0.00011342000000000001,
      "loss": 0.03707289695739746,
      "step": 4330
    },
    {
      "epoch": 8.0156,
      "grad_norm": 0.1780337393283844,
      "learning_rate": 0.00011322000000000002,
      "loss": 0.04678710401058197,
      "step": 4340
    },
    {
      "epoch": 8.0166,
      "grad_norm": 0.22333677113056183,
      "learning_rate": 0.00011302000000000001,
      "loss": 0.03880314826965332,
      "step": 4350
    },
    {
      "epoch": 8.0176,
      "grad_norm": 0.2167593091726303,
      "learning_rate": 0.00011282000000000002,
      "loss": 0.038645395636558534,
      "step": 4360
    },
    {
      "epoch": 8.0186,
      "grad_norm": 0.21094417572021484,
      "learning_rate": 0.00011262000000000002,
      "loss": 0.03557896316051483,
      "step": 4370
    },
    {
      "epoch": 8.0196,
      "grad_norm": 0.2134038805961609,
      "learning_rate": 0.00011242000000000001,
      "loss": 0.035387244820594785,
      "step": 4380
    },
    {
      "epoch": 8.0206,
      "grad_norm": 0.18721550703048706,
      "learning_rate": 0.00011222000000000002,
      "loss": 0.0355907142162323,
      "step": 4390
    },
    {
      "epoch": 8.0216,
      "grad_norm": 0.20684556663036346,
      "learning_rate": 0.00011202000000000002,
      "loss": 0.0357794851064682,
      "step": 4400
    },
    {
      "epoch": 8.0226,
      "grad_norm": 0.2096317857503891,
      "learning_rate": 0.00011182000000000001,
      "loss": 0.036594313383102414,
      "step": 4410
    },
    {
      "epoch": 8.0236,
      "grad_norm": 0.1866813451051712,
      "learning_rate": 0.00011162000000000002,
      "loss": 0.038320142030715945,
      "step": 4420
    },
    {
      "epoch": 8.0246,
      "grad_norm": 0.17548422515392303,
      "learning_rate": 0.00011142000000000001,
      "loss": 0.03279055655002594,
      "step": 4430
    },
    {
      "epoch": 8.0256,
      "grad_norm": 0.1939040571451187,
      "learning_rate": 0.00011122000000000001,
      "loss": 0.033512675762176515,
      "step": 4440
    },
    {
      "epoch": 8.0266,
      "grad_norm": 0.1942121386528015,
      "learning_rate": 0.00011102000000000002,
      "loss": 0.03630416989326477,
      "step": 4450
    },
    {
      "epoch": 8.0276,
      "grad_norm": 0.18124482035636902,
      "learning_rate": 0.00011082000000000001,
      "loss": 0.0364689439535141,
      "step": 4460
    },
    {
      "epoch": 8.0286,
      "grad_norm": 0.2448221743106842,
      "learning_rate": 0.00011062000000000001,
      "loss": 0.036901557445526124,
      "step": 4470
    },
    {
      "epoch": 8.0296,
      "grad_norm": 0.20503127574920654,
      "learning_rate": 0.00011042000000000002,
      "loss": 0.037753531336784364,
      "step": 4480
    },
    {
      "epoch": 8.0306,
      "grad_norm": 0.20306122303009033,
      "learning_rate": 0.00011022000000000001,
      "loss": 0.036759817600250246,
      "step": 4490
    },
    {
      "epoch": 9.0004,
      "grad_norm": 0.16577443480491638,
      "learning_rate": 0.00011002000000000001,
      "loss": 0.03821190297603607,
      "step": 4500
    },
    {
      "epoch": 1.001,
      "grad_norm": 0.15719620883464813,
      "learning_rate": 0.00010982000000000002,
      "loss": 0.02730073928833008,
      "step": 4510
    },
    {
      "epoch": 1.002,
      "grad_norm": 0.19089031219482422,
      "learning_rate": 0.00010962000000000001,
      "loss": 0.036603128910064696,
      "step": 4520
    },
    {
      "epoch": 1.003,
      "grad_norm": 0.20838193595409393,
      "learning_rate": 0.00010942,
      "loss": 0.03440204858779907,
      "step": 4530
    },
    {
      "epoch": 1.004,
      "grad_norm": 0.2211591750383377,
      "learning_rate": 0.00010922000000000001,
      "loss": 0.03840194046497345,
      "step": 4540
    },
    {
      "epoch": 1.005,
      "grad_norm": 0.19744090735912323,
      "learning_rate": 0.00010902000000000001,
      "loss": 0.03325707018375397,
      "step": 4550
    },
    {
      "epoch": 1.006,
      "grad_norm": 0.2157205492258072,
      "learning_rate": 0.00010882,
      "loss": 0.03541966378688812,
      "step": 4560
    },
    {
      "epoch": 1.007,
      "grad_norm": 0.21323737502098083,
      "learning_rate": 0.00010862000000000001,
      "loss": 0.03531204462051392,
      "step": 4570
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.1734951287508011,
      "learning_rate": 0.00010842000000000001,
      "loss": 0.033580628037452695,
      "step": 4580
    },
    {
      "epoch": 1.009,
      "grad_norm": 0.19895830750465393,
      "learning_rate": 0.00010822,
      "loss": 0.03522237241268158,
      "step": 4590
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.21840578317642212,
      "learning_rate": 0.00010802000000000001,
      "loss": 0.03673352003097534,
      "step": 4600
    },
    {
      "epoch": 1.011,
      "grad_norm": 0.18891401588916779,
      "learning_rate": 0.00010782000000000001,
      "loss": 0.03417102098464966,
      "step": 4610
    },
    {
      "epoch": 1.012,
      "grad_norm": 0.21882204711437225,
      "learning_rate": 0.00010762,
      "loss": 0.03885083496570587,
      "step": 4620
    },
    {
      "epoch": 1.013,
      "grad_norm": 0.1923336237668991,
      "learning_rate": 0.00010742000000000001,
      "loss": 0.03595931828022003,
      "step": 4630
    },
    {
      "epoch": 1.014,
      "grad_norm": 0.21916937828063965,
      "learning_rate": 0.00010722000000000001,
      "loss": 0.03626834750175476,
      "step": 4640
    },
    {
      "epoch": 1.015,
      "grad_norm": 0.19531555473804474,
      "learning_rate": 0.00010702000000000002,
      "loss": 0.0359089732170105,
      "step": 4650
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.15870879590511322,
      "learning_rate": 0.00010682000000000001,
      "loss": 0.048665601015090945,
      "step": 4660
    },
    {
      "epoch": 1.017,
      "grad_norm": 0.23078610002994537,
      "learning_rate": 0.00010662,
      "loss": 0.034716612100601195,
      "step": 4670
    },
    {
      "epoch": 1.018,
      "grad_norm": 0.19514289498329163,
      "learning_rate": 0.00010642000000000001,
      "loss": 0.03471817970275879,
      "step": 4680
    },
    {
      "epoch": 1.019,
      "grad_norm": 0.20728954672813416,
      "learning_rate": 0.00010622000000000001,
      "loss": 0.032735931873321536,
      "step": 4690
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.18462544679641724,
      "learning_rate": 0.00010602,
      "loss": 0.032894554734230044,
      "step": 4700
    },
    {
      "epoch": 1.021,
      "grad_norm": 0.16550667583942413,
      "learning_rate": 0.00010582000000000001,
      "loss": 0.03383542597293854,
      "step": 4710
    },
    {
      "epoch": 1.022,
      "grad_norm": 0.1777375191450119,
      "learning_rate": 0.00010562000000000001,
      "loss": 0.0322774738073349,
      "step": 4720
    },
    {
      "epoch": 1.023,
      "grad_norm": 0.19512343406677246,
      "learning_rate": 0.00010542,
      "loss": 0.035142365097999576,
      "step": 4730
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.1756298989057541,
      "learning_rate": 0.00010522000000000001,
      "loss": 0.031283268332481386,
      "step": 4740
    },
    {
      "epoch": 1.025,
      "grad_norm": 0.18716566264629364,
      "learning_rate": 0.00010502000000000001,
      "loss": 0.03243010342121124,
      "step": 4750
    },
    {
      "epoch": 1.026,
      "grad_norm": 0.18821994960308075,
      "learning_rate": 0.00010482,
      "loss": 0.033910727500915526,
      "step": 4760
    },
    {
      "epoch": 1.027,
      "grad_norm": 0.22193297743797302,
      "learning_rate": 0.00010462000000000001,
      "loss": 0.033655932545661925,
      "step": 4770
    },
    {
      "epoch": 1.028,
      "grad_norm": 0.15854744613170624,
      "learning_rate": 0.00010442,
      "loss": 0.031897363066673276,
      "step": 4780
    },
    {
      "epoch": 1.029,
      "grad_norm": 0.20192238688468933,
      "learning_rate": 0.00010422,
      "loss": 0.034781771898269656,
      "step": 4790
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.17679820954799652,
      "learning_rate": 0.00010402000000000001,
      "loss": 0.03542966842651367,
      "step": 4800
    },
    {
      "epoch": 1.031,
      "grad_norm": 0.2050696164369583,
      "learning_rate": 0.00010382,
      "loss": 0.03525029718875885,
      "step": 4810
    },
    {
      "epoch": 2.0008,
      "grad_norm": 0.1847396343946457,
      "learning_rate": 0.00010362,
      "loss": 0.03324263691902161,
      "step": 4820
    },
    {
      "epoch": 2.0018,
      "grad_norm": 0.198337122797966,
      "learning_rate": 0.00010342000000000001,
      "loss": 0.031684267520904544,
      "step": 4830
    },
    {
      "epoch": 2.0028,
      "grad_norm": 0.18886806070804596,
      "learning_rate": 0.00010322,
      "loss": 0.03731386065483093,
      "step": 4840
    },
    {
      "epoch": 2.0038,
      "grad_norm": 0.1607206165790558,
      "learning_rate": 0.00010302,
      "loss": 0.03477532863616943,
      "step": 4850
    },
    {
      "epoch": 2.0048,
      "grad_norm": 0.18906278908252716,
      "learning_rate": 0.00010282000000000001,
      "loss": 0.033050671219825745,
      "step": 4860
    },
    {
      "epoch": 2.0058,
      "grad_norm": 0.20967625081539154,
      "learning_rate": 0.00010262,
      "loss": 0.03245969414710999,
      "step": 4870
    },
    {
      "epoch": 2.0068,
      "grad_norm": 0.2613036334514618,
      "learning_rate": 0.00010242,
      "loss": 0.0329026609659195,
      "step": 4880
    },
    {
      "epoch": 2.0078,
      "grad_norm": 0.22789430618286133,
      "learning_rate": 0.00010222000000000001,
      "loss": 0.033549562096595764,
      "step": 4890
    },
    {
      "epoch": 2.0088,
      "grad_norm": 0.24477814137935638,
      "learning_rate": 0.00010202,
      "loss": 0.03555153608322144,
      "step": 4900
    },
    {
      "epoch": 2.0098,
      "grad_norm": 0.185205340385437,
      "learning_rate": 0.00010182,
      "loss": 0.034665200114250186,
      "step": 4910
    },
    {
      "epoch": 2.0108,
      "grad_norm": 0.18530991673469543,
      "learning_rate": 0.00010162,
      "loss": 0.03328844904899597,
      "step": 4920
    },
    {
      "epoch": 2.0118,
      "grad_norm": 0.2243437021970749,
      "learning_rate": 0.00010142,
      "loss": 0.033396825194358826,
      "step": 4930
    },
    {
      "epoch": 2.0128,
      "grad_norm": 0.18864013254642487,
      "learning_rate": 0.00010122000000000001,
      "loss": 0.034027200937271115,
      "step": 4940
    },
    {
      "epoch": 2.0138,
      "grad_norm": 0.18260319530963898,
      "learning_rate": 0.00010102,
      "loss": 0.033051350712776185,
      "step": 4950
    },
    {
      "epoch": 2.0148,
      "grad_norm": 0.18739180266857147,
      "learning_rate": 0.00010082,
      "loss": 0.03482633233070374,
      "step": 4960
    },
    {
      "epoch": 2.0158,
      "grad_norm": 0.21303392946720123,
      "learning_rate": 0.00010064,
      "loss": 0.06499571204185486,
      "step": 4970
    },
    {
      "epoch": 2.0168,
      "grad_norm": 0.23845174908638,
      "learning_rate": 0.00010044000000000001,
      "loss": 0.03167785108089447,
      "step": 4980
    },
    {
      "epoch": 2.0178,
      "grad_norm": 0.22625099122524261,
      "learning_rate": 0.00010024,
      "loss": 0.03223021924495697,
      "step": 4990
    },
    {
      "epoch": 2.0188,
      "grad_norm": 0.17533113062381744,
      "learning_rate": 0.00010004,
      "loss": 0.03166678845882416,
      "step": 5000
    },
    {
      "epoch": 2.0198,
      "grad_norm": 0.16470959782600403,
      "learning_rate": 9.984e-05,
      "loss": 0.031538739800453186,
      "step": 5010
    },
    {
      "epoch": 2.0208,
      "grad_norm": 0.18001972138881683,
      "learning_rate": 9.964e-05,
      "loss": 0.03074006736278534,
      "step": 5020
    },
    {
      "epoch": 2.0218,
      "grad_norm": 0.19739823043346405,
      "learning_rate": 9.944e-05,
      "loss": 0.03161226511001587,
      "step": 5030
    },
    {
      "epoch": 2.0228,
      "grad_norm": 0.1822267472743988,
      "learning_rate": 9.924e-05,
      "loss": 0.03073483407497406,
      "step": 5040
    },
    {
      "epoch": 2.0238,
      "grad_norm": 0.14822076261043549,
      "learning_rate": 9.904e-05,
      "loss": 0.036596888303756715,
      "step": 5050
    },
    {
      "epoch": 2.0248,
      "grad_norm": 0.16567520797252655,
      "learning_rate": 9.884e-05,
      "loss": 0.0309079110622406,
      "step": 5060
    },
    {
      "epoch": 2.0258,
      "grad_norm": 0.20851384103298187,
      "learning_rate": 9.864e-05,
      "loss": 0.03173742294311523,
      "step": 5070
    },
    {
      "epoch": 2.0268,
      "grad_norm": 0.21927666664123535,
      "learning_rate": 9.844000000000001e-05,
      "loss": 0.033800792694091794,
      "step": 5080
    },
    {
      "epoch": 2.0278,
      "grad_norm": 0.17513330280780792,
      "learning_rate": 9.824000000000001e-05,
      "loss": 0.03414813280105591,
      "step": 5090
    },
    {
      "epoch": 2.0288,
      "grad_norm": 0.17805452644824982,
      "learning_rate": 9.804e-05,
      "loss": 0.0326084703207016,
      "step": 5100
    },
    {
      "epoch": 2.0298,
      "grad_norm": 0.17655085027217865,
      "learning_rate": 9.784000000000001e-05,
      "loss": 0.0349094420671463,
      "step": 5110
    },
    {
      "epoch": 2.0308,
      "grad_norm": 0.16686567664146423,
      "learning_rate": 9.764000000000001e-05,
      "loss": 0.032757118344306946,
      "step": 5120
    },
    {
      "epoch": 3.0006,
      "grad_norm": 0.2090245932340622,
      "learning_rate": 9.744000000000002e-05,
      "loss": 0.03307884931564331,
      "step": 5130
    },
    {
      "epoch": 3.0016,
      "grad_norm": 0.1725403368473053,
      "learning_rate": 9.724000000000001e-05,
      "loss": 0.028950795531272888,
      "step": 5140
    },
    {
      "epoch": 3.0026,
      "grad_norm": 0.20363931357860565,
      "learning_rate": 9.704e-05,
      "loss": 0.0335098922252655,
      "step": 5150
    },
    {
      "epoch": 3.0036,
      "grad_norm": 0.3671290874481201,
      "learning_rate": 9.684000000000001e-05,
      "loss": 0.03179095685482025,
      "step": 5160
    },
    {
      "epoch": 3.0046,
      "grad_norm": 0.19481417536735535,
      "learning_rate": 9.664000000000001e-05,
      "loss": 0.0307291179895401,
      "step": 5170
    },
    {
      "epoch": 3.0056,
      "grad_norm": 0.22754861414432526,
      "learning_rate": 9.644e-05,
      "loss": 0.03357323706150055,
      "step": 5180
    },
    {
      "epoch": 3.0066,
      "grad_norm": 0.23329274356365204,
      "learning_rate": 9.624000000000001e-05,
      "loss": 0.03400978147983551,
      "step": 5190
    },
    {
      "epoch": 3.0076,
      "grad_norm": 0.18839909136295319,
      "learning_rate": 9.604000000000001e-05,
      "loss": 0.03397571742534637,
      "step": 5200
    },
    {
      "epoch": 3.0086,
      "grad_norm": 0.2037007361650467,
      "learning_rate": 9.584e-05,
      "loss": 0.032344302535057066,
      "step": 5210
    },
    {
      "epoch": 3.0096,
      "grad_norm": 0.22633804380893707,
      "learning_rate": 9.564000000000001e-05,
      "loss": 0.03367453217506409,
      "step": 5220
    },
    {
      "epoch": 3.0106,
      "grad_norm": 0.21246115863323212,
      "learning_rate": 9.544000000000001e-05,
      "loss": 0.03286383748054504,
      "step": 5230
    },
    {
      "epoch": 3.0116,
      "grad_norm": 0.1776570826768875,
      "learning_rate": 9.524e-05,
      "loss": 0.031331130862236024,
      "step": 5240
    },
    {
      "epoch": 3.0126,
      "grad_norm": 0.18236294388771057,
      "learning_rate": 9.504000000000001e-05,
      "loss": 0.03371650278568268,
      "step": 5250
    },
    {
      "epoch": 3.0136,
      "grad_norm": 0.20707473158836365,
      "learning_rate": 9.484e-05,
      "loss": 0.0332234263420105,
      "step": 5260
    },
    {
      "epoch": 3.0146,
      "grad_norm": 0.22507430613040924,
      "learning_rate": 9.464e-05,
      "loss": 0.035472571849823,
      "step": 5270
    },
    {
      "epoch": 3.0156,
      "grad_norm": 0.1862962543964386,
      "learning_rate": 9.444000000000001e-05,
      "loss": 0.06388773322105408,
      "step": 5280
    },
    {
      "epoch": 3.0166,
      "grad_norm": 0.19476568698883057,
      "learning_rate": 9.424e-05,
      "loss": 0.030706775188446046,
      "step": 5290
    },
    {
      "epoch": 3.0176,
      "grad_norm": 0.1976071298122406,
      "learning_rate": 9.404e-05,
      "loss": 0.03262068927288055,
      "step": 5300
    },
    {
      "epoch": 3.0186,
      "grad_norm": 0.19985517859458923,
      "learning_rate": 9.384000000000001e-05,
      "loss": 0.0315820038318634,
      "step": 5310
    },
    {
      "epoch": 3.0196,
      "grad_norm": 0.22363457083702087,
      "learning_rate": 9.364e-05,
      "loss": 0.030417001247406004,
      "step": 5320
    },
    {
      "epoch": 3.0206,
      "grad_norm": 0.16263076663017273,
      "learning_rate": 9.344e-05,
      "loss": 0.03050481975078583,
      "step": 5330
    },
    {
      "epoch": 3.0216,
      "grad_norm": 0.22901087999343872,
      "learning_rate": 9.324000000000001e-05,
      "loss": 0.031095728278160095,
      "step": 5340
    },
    {
      "epoch": 3.0226,
      "grad_norm": 0.18322071433067322,
      "learning_rate": 9.304e-05,
      "loss": 0.029934582114219666,
      "step": 5350
    },
    {
      "epoch": 3.0236,
      "grad_norm": 0.20856066048145294,
      "learning_rate": 9.284e-05,
      "loss": 0.03481347560882568,
      "step": 5360
    },
    {
      "epoch": 3.0246,
      "grad_norm": 0.16634713113307953,
      "learning_rate": 9.264000000000001e-05,
      "loss": 0.03441186249256134,
      "step": 5370
    },
    {
      "epoch": 3.0256,
      "grad_norm": 0.16705217957496643,
      "learning_rate": 9.244e-05,
      "loss": 0.03096645176410675,
      "step": 5380
    },
    {
      "epoch": 3.0266,
      "grad_norm": 0.20979522168636322,
      "learning_rate": 9.224e-05,
      "loss": 0.03209489285945892,
      "step": 5390
    },
    {
      "epoch": 3.0276,
      "grad_norm": 0.1905623823404312,
      "learning_rate": 9.204e-05,
      "loss": 0.03437664210796356,
      "step": 5400
    },
    {
      "epoch": 3.0286,
      "grad_norm": 0.23262503743171692,
      "learning_rate": 9.184e-05,
      "loss": 0.033020347356796265,
      "step": 5410
    },
    {
      "epoch": 3.0296,
      "grad_norm": 0.1738559901714325,
      "learning_rate": 9.164000000000001e-05,
      "loss": 0.03307526707649231,
      "step": 5420
    },
    {
      "epoch": 3.0306,
      "grad_norm": 0.1965901404619217,
      "learning_rate": 9.144e-05,
      "loss": 0.032591798901557924,
      "step": 5430
    },
    {
      "epoch": 4.0004,
      "grad_norm": 0.18910537660121918,
      "learning_rate": 9.124e-05,
      "loss": 0.03326196372509003,
      "step": 5440
    },
    {
      "epoch": 4.0014,
      "grad_norm": 0.2184404879808426,
      "learning_rate": 9.104000000000001e-05,
      "loss": 0.029417768120765686,
      "step": 5450
    },
    {
      "epoch": 4.0024,
      "grad_norm": 0.22602999210357666,
      "learning_rate": 9.084e-05,
      "loss": 0.031762811541557315,
      "step": 5460
    },
    {
      "epoch": 4.0034,
      "grad_norm": 0.19832205772399902,
      "learning_rate": 9.064e-05,
      "loss": 0.03158574402332306,
      "step": 5470
    },
    {
      "epoch": 4.0044,
      "grad_norm": 0.19668439030647278,
      "learning_rate": 9.044000000000001e-05,
      "loss": 0.032255178689956664,
      "step": 5480
    },
    {
      "epoch": 4.0054,
      "grad_norm": 0.20277754962444305,
      "learning_rate": 9.024e-05,
      "loss": 0.031222817301750184,
      "step": 5490
    },
    {
      "epoch": 4.0064,
      "grad_norm": 0.21987652778625488,
      "learning_rate": 9.004e-05,
      "loss": 0.03465475142002106,
      "step": 5500
    },
    {
      "epoch": 4.0074,
      "grad_norm": 0.22462855279445648,
      "learning_rate": 8.984000000000001e-05,
      "loss": 0.03296079635620117,
      "step": 5510
    },
    {
      "epoch": 4.0084,
      "grad_norm": 0.22776798903942108,
      "learning_rate": 8.964e-05,
      "loss": 0.03229100704193115,
      "step": 5520
    },
    {
      "epoch": 4.0094,
      "grad_norm": 0.22231830656528473,
      "learning_rate": 8.944e-05,
      "loss": 0.03347973227500915,
      "step": 5530
    },
    {
      "epoch": 4.0104,
      "grad_norm": 0.2460055649280548,
      "learning_rate": 8.924e-05,
      "loss": 0.0332697331905365,
      "step": 5540
    },
    {
      "epoch": 4.0114,
      "grad_norm": 0.17865519225597382,
      "learning_rate": 8.904e-05,
      "loss": 0.03274652659893036,
      "step": 5550
    },
    {
      "epoch": 4.0124,
      "grad_norm": 0.19629009068012238,
      "learning_rate": 8.884e-05,
      "loss": 0.03690815269947052,
      "step": 5560
    },
    {
      "epoch": 4.0134,
      "grad_norm": 0.22446274757385254,
      "learning_rate": 8.864e-05,
      "loss": 0.03131195604801178,
      "step": 5570
    },
    {
      "epoch": 4.0144,
      "grad_norm": 0.2365735024213791,
      "learning_rate": 8.844e-05,
      "loss": 0.03362289369106293,
      "step": 5580
    },
    {
      "epoch": 4.0154,
      "grad_norm": 0.2169988453388214,
      "learning_rate": 8.824e-05,
      "loss": 0.04250128865242005,
      "step": 5590
    },
    {
      "epoch": 4.0164,
      "grad_norm": 0.2399117648601532,
      "learning_rate": 8.804e-05,
      "loss": 0.031144574284553528,
      "step": 5600
    },
    {
      "epoch": 4.0174,
      "grad_norm": 0.2245727777481079,
      "learning_rate": 8.784e-05,
      "loss": 0.03198579251766205,
      "step": 5610
    },
    {
      "epoch": 4.0184,
      "grad_norm": 0.19800491631031036,
      "learning_rate": 8.764e-05,
      "loss": 0.031690898537635806,
      "step": 5620
    },
    {
      "epoch": 4.0194,
      "grad_norm": 0.21474099159240723,
      "learning_rate": 8.744e-05,
      "loss": 0.034224125742912295,
      "step": 5630
    },
    {
      "epoch": 4.0204,
      "grad_norm": 0.23651978373527527,
      "learning_rate": 8.724e-05,
      "loss": 0.030482065677642823,
      "step": 5640
    },
    {
      "epoch": 4.0214,
      "grad_norm": 0.2008265107870102,
      "learning_rate": 8.704e-05,
      "loss": 0.029612430930137636,
      "step": 5650
    },
    {
      "epoch": 4.0224,
      "grad_norm": 0.18203487992286682,
      "learning_rate": 8.684e-05,
      "loss": 0.03142530620098114,
      "step": 5660
    },
    {
      "epoch": 4.0234,
      "grad_norm": 0.19077278673648834,
      "learning_rate": 8.664e-05,
      "loss": 0.03421863317489624,
      "step": 5670
    },
    {
      "epoch": 4.0244,
      "grad_norm": 0.18829531967639923,
      "learning_rate": 8.643999999999999e-05,
      "loss": 0.03218718469142914,
      "step": 5680
    },
    {
      "epoch": 4.0254,
      "grad_norm": 0.19440174102783203,
      "learning_rate": 8.624000000000001e-05,
      "loss": 0.030671533942222596,
      "step": 5690
    },
    {
      "epoch": 4.0264,
      "grad_norm": 0.20343391597270966,
      "learning_rate": 8.604000000000001e-05,
      "loss": 0.03245647549629212,
      "step": 5700
    },
    {
      "epoch": 4.0274,
      "grad_norm": 0.20841389894485474,
      "learning_rate": 8.584e-05,
      "loss": 0.032360246777534483,
      "step": 5710
    },
    {
      "epoch": 4.0284,
      "grad_norm": 0.20818255841732025,
      "learning_rate": 8.564000000000001e-05,
      "loss": 0.03382550477981568,
      "step": 5720
    },
    {
      "epoch": 4.0294,
      "grad_norm": 0.22749511897563934,
      "learning_rate": 8.544000000000001e-05,
      "loss": 0.033779263496398926,
      "step": 5730
    },
    {
      "epoch": 4.0304,
      "grad_norm": 0.15877234935760498,
      "learning_rate": 8.524e-05,
      "loss": 0.031108471751213073,
      "step": 5740
    },
    {
      "epoch": 5.0002,
      "grad_norm": 0.1577063500881195,
      "learning_rate": 8.504000000000001e-05,
      "loss": 0.033298739790916444,
      "step": 5750
    },
    {
      "epoch": 5.0012,
      "grad_norm": 0.19950023293495178,
      "learning_rate": 8.484000000000001e-05,
      "loss": 0.028189751505851745,
      "step": 5760
    },
    {
      "epoch": 5.0022,
      "grad_norm": 0.24177151918411255,
      "learning_rate": 8.464e-05,
      "loss": 0.031087157130241395,
      "step": 5770
    },
    {
      "epoch": 5.0032,
      "grad_norm": 0.20051972568035126,
      "learning_rate": 8.444000000000001e-05,
      "loss": 0.03331252634525299,
      "step": 5780
    },
    {
      "epoch": 5.0042,
      "grad_norm": 0.2181842178106308,
      "learning_rate": 8.424e-05,
      "loss": 0.032180684804916385,
      "step": 5790
    },
    {
      "epoch": 5.0052,
      "grad_norm": 0.1912951022386551,
      "learning_rate": 8.404e-05,
      "loss": 0.029671770334243775,
      "step": 5800
    },
    {
      "epoch": 5.0062,
      "grad_norm": 0.21211616694927216,
      "learning_rate": 8.384000000000001e-05,
      "loss": 0.032269975543022154,
      "step": 5810
    },
    {
      "epoch": 5.0072,
      "grad_norm": 0.20029117166996002,
      "learning_rate": 8.364e-05,
      "loss": 0.033931884169578555,
      "step": 5820
    },
    {
      "epoch": 5.0082,
      "grad_norm": 0.2266046702861786,
      "learning_rate": 8.344e-05,
      "loss": 0.03187280297279358,
      "step": 5830
    },
    {
      "epoch": 5.0092,
      "grad_norm": 0.22159722447395325,
      "learning_rate": 8.324000000000001e-05,
      "loss": 0.034100428223609924,
      "step": 5840
    },
    {
      "epoch": 5.0102,
      "grad_norm": 0.21007134020328522,
      "learning_rate": 8.304e-05,
      "loss": 0.033664295077323915,
      "step": 5850
    },
    {
      "epoch": 5.0112,
      "grad_norm": 0.2078692615032196,
      "learning_rate": 8.284000000000001e-05,
      "loss": 0.03211201131343842,
      "step": 5860
    },
    {
      "epoch": 5.0122,
      "grad_norm": 0.24859845638275146,
      "learning_rate": 8.264000000000001e-05,
      "loss": 0.03572002947330475,
      "step": 5870
    },
    {
      "epoch": 5.0132,
      "grad_norm": 0.23154641687870026,
      "learning_rate": 8.244e-05,
      "loss": 0.030712798237800598,
      "step": 5880
    },
    {
      "epoch": 5.0142,
      "grad_norm": 0.2343783676624298,
      "learning_rate": 8.224000000000001e-05,
      "loss": 0.03176690340042114,
      "step": 5890
    },
    {
      "epoch": 5.0152,
      "grad_norm": 0.196640282869339,
      "learning_rate": 8.204000000000001e-05,
      "loss": 0.04715462028980255,
      "step": 5900
    },
    {
      "epoch": 5.0162,
      "grad_norm": 0.1985359638929367,
      "learning_rate": 8.184e-05,
      "loss": 0.030827867984771728,
      "step": 5910
    },
    {
      "epoch": 5.0172,
      "grad_norm": 0.19483497738838196,
      "learning_rate": 8.164000000000001e-05,
      "loss": 0.032147163152694704,
      "step": 5920
    },
    {
      "epoch": 5.0182,
      "grad_norm": 0.19168393313884735,
      "learning_rate": 8.144e-05,
      "loss": 0.031657275557518,
      "step": 5930
    },
    {
      "epoch": 5.0192,
      "grad_norm": 0.18714462220668793,
      "learning_rate": 8.124e-05,
      "loss": 0.031624266505241395,
      "step": 5940
    },
    {
      "epoch": 5.0202,
      "grad_norm": 0.21911782026290894,
      "learning_rate": 8.104000000000001e-05,
      "loss": 0.03106756806373596,
      "step": 5950
    },
    {
      "epoch": 5.0212,
      "grad_norm": 0.19405068457126617,
      "learning_rate": 8.084e-05,
      "loss": 0.03130735754966736,
      "step": 5960
    },
    {
      "epoch": 5.0222,
      "grad_norm": 0.21102119982242584,
      "learning_rate": 8.064e-05,
      "loss": 0.03118746280670166,
      "step": 5970
    },
    {
      "epoch": 5.0232,
      "grad_norm": 0.19692668318748474,
      "learning_rate": 8.044000000000001e-05,
      "loss": 0.030439674854278564,
      "step": 5980
    },
    {
      "epoch": 5.0242,
      "grad_norm": 0.2191537320613861,
      "learning_rate": 8.024e-05,
      "loss": 0.030052882432937623,
      "step": 5990
    },
    {
      "epoch": 5.0252,
      "grad_norm": 0.19516703486442566,
      "learning_rate": 8.004e-05,
      "loss": 0.029340890049934388,
      "step": 6000
    },
    {
      "epoch": 5.0262,
      "grad_norm": 0.21514344215393066,
      "learning_rate": 7.984000000000001e-05,
      "loss": 0.029882851243019103,
      "step": 6010
    },
    {
      "epoch": 5.0272,
      "grad_norm": 0.18462449312210083,
      "learning_rate": 7.964e-05,
      "loss": 0.03178698718547821,
      "step": 6020
    },
    {
      "epoch": 5.0282,
      "grad_norm": 0.17087651789188385,
      "learning_rate": 7.944e-05,
      "loss": 0.031465652585029605,
      "step": 6030
    },
    {
      "epoch": 5.0292,
      "grad_norm": 0.17649878561496735,
      "learning_rate": 7.924000000000001e-05,
      "loss": 0.030974376201629638,
      "step": 6040
    },
    {
      "epoch": 5.0302,
      "grad_norm": 0.17699715495109558,
      "learning_rate": 7.904e-05,
      "loss": 0.03126508891582489,
      "step": 6050
    },
    {
      "epoch": 5.0312,
      "grad_norm": 0.1981772631406784,
      "learning_rate": 7.884e-05,
      "loss": 0.03145314157009125,
      "step": 6060
    },
    {
      "epoch": 6.001,
      "grad_norm": 0.20157794654369354,
      "learning_rate": 7.864e-05,
      "loss": 0.034029459953308104,
      "step": 6070
    },
    {
      "epoch": 6.002,
      "grad_norm": 0.19972266256809235,
      "learning_rate": 7.844e-05,
      "loss": 0.03382774591445923,
      "step": 6080
    },
    {
      "epoch": 6.003,
      "grad_norm": 0.2004992961883545,
      "learning_rate": 7.824e-05,
      "loss": 0.033232077956199646,
      "step": 6090
    },
    {
      "epoch": 6.004,
      "grad_norm": 0.19882363080978394,
      "learning_rate": 7.804e-05,
      "loss": 0.03139175176620483,
      "step": 6100
    },
    {
      "epoch": 6.005,
      "grad_norm": 0.17804212868213654,
      "learning_rate": 7.784e-05,
      "loss": 0.03176693022251129,
      "step": 6110
    },
    {
      "epoch": 6.006,
      "grad_norm": 0.19885767996311188,
      "learning_rate": 7.764e-05,
      "loss": 0.03188234269618988,
      "step": 6120
    },
    {
      "epoch": 6.007,
      "grad_norm": 0.231368288397789,
      "learning_rate": 7.744e-05,
      "loss": 0.03077830970287323,
      "step": 6130
    },
    {
      "epoch": 6.008,
      "grad_norm": 0.21649253368377686,
      "learning_rate": 7.724e-05,
      "loss": 0.030950218439102173,
      "step": 6140
    },
    {
      "epoch": 6.009,
      "grad_norm": 0.18937084078788757,
      "learning_rate": 7.704000000000001e-05,
      "loss": 0.03231707811355591,
      "step": 6150
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.1458628922700882,
      "learning_rate": 7.684e-05,
      "loss": 0.02982558310031891,
      "step": 6160
    },
    {
      "epoch": 6.011,
      "grad_norm": 0.18662185966968536,
      "learning_rate": 7.664e-05,
      "loss": 0.030502399802207945,
      "step": 6170
    },
    {
      "epoch": 6.012,
      "grad_norm": 0.9817757606506348,
      "learning_rate": 7.644e-05,
      "loss": 0.03566881716251373,
      "step": 6180
    },
    {
      "epoch": 6.013,
      "grad_norm": 0.21928751468658447,
      "learning_rate": 7.624e-05,
      "loss": 0.03193634152412415,
      "step": 6190
    },
    {
      "epoch": 6.014,
      "grad_norm": 0.2132958173751831,
      "learning_rate": 7.604e-05,
      "loss": 0.03002108633518219,
      "step": 6200
    },
    {
      "epoch": 6.015,
      "grad_norm": 0.19893419742584229,
      "learning_rate": 7.584e-05,
      "loss": 0.02982170581817627,
      "step": 6210
    },
    {
      "epoch": 6.016,
      "grad_norm": 0.17725782096385956,
      "learning_rate": 7.564e-05,
      "loss": 0.036422553658485415,
      "step": 6220
    },
    {
      "epoch": 6.017,
      "grad_norm": 0.1956397294998169,
      "learning_rate": 7.544e-05,
      "loss": 0.028982603549957277,
      "step": 6230
    },
    {
      "epoch": 6.018,
      "grad_norm": 0.1871957629919052,
      "learning_rate": 7.524e-05,
      "loss": 0.03049376904964447,
      "step": 6240
    },
    {
      "epoch": 6.019,
      "grad_norm": 0.15977296233177185,
      "learning_rate": 7.504e-05,
      "loss": 0.028836172819137574,
      "step": 6250
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.1604572832584381,
      "learning_rate": 7.484e-05,
      "loss": 0.031694543361663816,
      "step": 6260
    },
    {
      "epoch": 6.021,
      "grad_norm": 0.19708256423473358,
      "learning_rate": 7.464e-05,
      "loss": 0.0299109548330307,
      "step": 6270
    },
    {
      "epoch": 6.022,
      "grad_norm": 0.15963052213191986,
      "learning_rate": 7.444e-05,
      "loss": 0.02851861119270325,
      "step": 6280
    },
    {
      "epoch": 6.023,
      "grad_norm": 0.19732700288295746,
      "learning_rate": 7.424e-05,
      "loss": 0.02974964678287506,
      "step": 6290
    },
    {
      "epoch": 6.024,
      "grad_norm": 0.22040079534053802,
      "learning_rate": 7.404e-05,
      "loss": 0.02835639417171478,
      "step": 6300
    },
    {
      "epoch": 6.025,
      "grad_norm": 0.19101078808307648,
      "learning_rate": 7.384e-05,
      "loss": 0.028555047512054444,
      "step": 6310
    },
    {
      "epoch": 6.026,
      "grad_norm": 0.17046277225017548,
      "learning_rate": 7.364e-05,
      "loss": 0.029239752888679506,
      "step": 6320
    },
    {
      "epoch": 6.027,
      "grad_norm": 0.17496943473815918,
      "learning_rate": 7.344000000000002e-05,
      "loss": 0.028895750641822815,
      "step": 6330
    },
    {
      "epoch": 6.028,
      "grad_norm": 0.1851808875799179,
      "learning_rate": 7.324000000000001e-05,
      "loss": 0.029496818780899048,
      "step": 6340
    },
    {
      "epoch": 6.029,
      "grad_norm": 0.1762508600950241,
      "learning_rate": 7.304e-05,
      "loss": 0.028394830226898194,
      "step": 6350
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.2357683628797531,
      "learning_rate": 7.284000000000001e-05,
      "loss": 0.02871710658073425,
      "step": 6360
    },
    {
      "epoch": 6.031,
      "grad_norm": 0.16987094283103943,
      "learning_rate": 7.264000000000001e-05,
      "loss": 0.030784130096435547,
      "step": 6370
    },
    {
      "epoch": 7.0008,
      "grad_norm": 0.18693174421787262,
      "learning_rate": 7.244e-05,
      "loss": 0.03147324323654175,
      "step": 6380
    },
    {
      "epoch": 7.0018,
      "grad_norm": 0.21501696109771729,
      "learning_rate": 7.224000000000001e-05,
      "loss": 0.029030478000640868,
      "step": 6390
    },
    {
      "epoch": 7.0028,
      "grad_norm": 0.18433253467082977,
      "learning_rate": 7.204000000000001e-05,
      "loss": 0.031191644072532655,
      "step": 6400
    },
    {
      "epoch": 7.0038,
      "grad_norm": 0.22278091311454773,
      "learning_rate": 7.184e-05,
      "loss": 0.028717485070228577,
      "step": 6410
    },
    {
      "epoch": 7.0048,
      "grad_norm": 0.1910051554441452,
      "learning_rate": 7.164000000000001e-05,
      "loss": 0.027972865104675292,
      "step": 6420
    },
    {
      "epoch": 7.0058,
      "grad_norm": 0.227296844124794,
      "learning_rate": 7.144000000000001e-05,
      "loss": 0.027867498993873595,
      "step": 6430
    },
    {
      "epoch": 7.0068,
      "grad_norm": 0.21170516312122345,
      "learning_rate": 7.124e-05,
      "loss": 0.028536897897720338,
      "step": 6440
    },
    {
      "epoch": 7.0078,
      "grad_norm": 0.22283098101615906,
      "learning_rate": 7.104000000000001e-05,
      "loss": 0.027557718753814697,
      "step": 6450
    },
    {
      "epoch": 7.0088,
      "grad_norm": 0.19392724335193634,
      "learning_rate": 7.084e-05,
      "loss": 0.027928680181503296,
      "step": 6460
    },
    {
      "epoch": 7.0098,
      "grad_norm": 0.15986867249011993,
      "learning_rate": 7.064e-05,
      "loss": 0.028359243273735048,
      "step": 6470
    },
    {
      "epoch": 7.0108,
      "grad_norm": 0.16614486277103424,
      "learning_rate": 7.044000000000001e-05,
      "loss": 0.027036306262016297,
      "step": 6480
    },
    {
      "epoch": 7.0118,
      "grad_norm": 0.192184180021286,
      "learning_rate": 7.024e-05,
      "loss": 0.026826027035713195,
      "step": 6490
    },
    {
      "epoch": 7.0128,
      "grad_norm": 0.1559998244047165,
      "learning_rate": 7.004e-05,
      "loss": 0.030304345488548278,
      "step": 6500
    },
    {
      "epoch": 1.001,
      "grad_norm": 0.17337968945503235,
      "learning_rate": 6.984000000000001e-05,
      "loss": 0.04341032803058624,
      "step": 6510
    },
    {
      "epoch": 1.002,
      "grad_norm": 0.14447154104709625,
      "learning_rate": 6.964e-05,
      "loss": 0.02765703499317169,
      "step": 6520
    },
    {
      "epoch": 1.003,
      "grad_norm": 0.1359182447195053,
      "learning_rate": 6.944e-05,
      "loss": 0.027671322226524353,
      "step": 6530
    },
    {
      "epoch": 1.004,
      "grad_norm": 0.1486900895833969,
      "learning_rate": 6.924000000000001e-05,
      "loss": 0.024222801625728606,
      "step": 6540
    },
    {
      "epoch": 1.005,
      "grad_norm": 0.17833292484283447,
      "learning_rate": 6.904e-05,
      "loss": 0.024685877561569213,
      "step": 6550
    },
    {
      "epoch": 1.006,
      "grad_norm": 0.1444295197725296,
      "learning_rate": 6.884e-05,
      "loss": 0.027162906527519227,
      "step": 6560
    },
    {
      "epoch": 1.007,
      "grad_norm": 0.18214137852191925,
      "learning_rate": 6.864000000000001e-05,
      "loss": 0.02484011501073837,
      "step": 6570
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.16924667358398438,
      "learning_rate": 6.844e-05,
      "loss": 0.025604432821273802,
      "step": 6580
    },
    {
      "epoch": 1.009,
      "grad_norm": 0.16066834330558777,
      "learning_rate": 6.824e-05,
      "loss": 0.025216665863990784,
      "step": 6590
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.14215150475502014,
      "learning_rate": 6.804e-05,
      "loss": 0.024471497535705565,
      "step": 6600
    },
    {
      "epoch": 1.011,
      "grad_norm": 0.128558948636055,
      "learning_rate": 6.784e-05,
      "loss": 0.02575332820415497,
      "step": 6610
    },
    {
      "epoch": 1.012,
      "grad_norm": 0.33806535601615906,
      "learning_rate": 6.764000000000001e-05,
      "loss": 0.025676625967025756,
      "step": 6620
    },
    {
      "epoch": 1.013,
      "grad_norm": 0.16000686585903168,
      "learning_rate": 6.744e-05,
      "loss": 0.025471991300582884,
      "step": 6630
    },
    {
      "epoch": 1.014,
      "grad_norm": 0.19417209923267365,
      "learning_rate": 6.724e-05,
      "loss": 0.0291011780500412,
      "step": 6640
    },
    {
      "epoch": 1.015,
      "grad_norm": 0.2124422937631607,
      "learning_rate": 6.704000000000001e-05,
      "loss": 0.02770865261554718,
      "step": 6650
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.16012416779994965,
      "learning_rate": 6.684e-05,
      "loss": 0.03058260679244995,
      "step": 6660
    },
    {
      "epoch": 1.017,
      "grad_norm": 0.1718786507844925,
      "learning_rate": 6.664e-05,
      "loss": 0.02695387601852417,
      "step": 6670
    },
    {
      "epoch": 1.018,
      "grad_norm": 0.18105700612068176,
      "learning_rate": 6.644000000000001e-05,
      "loss": 0.026461568474769593,
      "step": 6680
    },
    {
      "epoch": 1.019,
      "grad_norm": 0.15785875916481018,
      "learning_rate": 6.624e-05,
      "loss": 0.02598581314086914,
      "step": 6690
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.15605275332927704,
      "learning_rate": 6.604e-05,
      "loss": 0.026878923177719116,
      "step": 6700
    },
    {
      "epoch": 1.021,
      "grad_norm": 0.1681290715932846,
      "learning_rate": 6.584e-05,
      "loss": 0.026533681154251098,
      "step": 6710
    },
    {
      "epoch": 1.022,
      "grad_norm": 0.17714430391788483,
      "learning_rate": 6.564e-05,
      "loss": 0.026042383909225465,
      "step": 6720
    },
    {
      "epoch": 1.023,
      "grad_norm": 0.17928007245063782,
      "learning_rate": 6.544e-05,
      "loss": 0.026147887110710144,
      "step": 6730
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.21016575396060944,
      "learning_rate": 6.524e-05,
      "loss": 0.026558607816696167,
      "step": 6740
    },
    {
      "epoch": 1.025,
      "grad_norm": 0.19502075016498566,
      "learning_rate": 6.504e-05,
      "loss": 0.02726261019706726,
      "step": 6750
    },
    {
      "epoch": 1.026,
      "grad_norm": 0.1563853919506073,
      "learning_rate": 6.484e-05,
      "loss": 0.025972676277160645,
      "step": 6760
    },
    {
      "epoch": 1.027,
      "grad_norm": 0.15548229217529297,
      "learning_rate": 6.464e-05,
      "loss": 0.027564069628715514,
      "step": 6770
    },
    {
      "epoch": 1.028,
      "grad_norm": 0.18457072973251343,
      "learning_rate": 6.444e-05,
      "loss": 0.027495378255844118,
      "step": 6780
    },
    {
      "epoch": 1.029,
      "grad_norm": 0.15842361748218536,
      "learning_rate": 6.424e-05,
      "loss": 0.026171448826789855,
      "step": 6790
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.19476866722106934,
      "learning_rate": 6.404e-05,
      "loss": 0.031318637728691104,
      "step": 6800
    },
    {
      "epoch": 1.031,
      "grad_norm": 0.1272721141576767,
      "learning_rate": 6.384e-05,
      "loss": 0.026132452487945556,
      "step": 6810
    },
    {
      "epoch": 2.0008,
      "grad_norm": 0.14176732301712036,
      "learning_rate": 6.364e-05,
      "loss": 0.03041217029094696,
      "step": 6820
    },
    {
      "epoch": 2.0018,
      "grad_norm": 0.14128795266151428,
      "learning_rate": 6.344e-05,
      "loss": 0.02396068423986435,
      "step": 6830
    },
    {
      "epoch": 2.0028,
      "grad_norm": 0.14129574596881866,
      "learning_rate": 6.324e-05,
      "loss": 0.02366064041852951,
      "step": 6840
    },
    {
      "epoch": 2.0038,
      "grad_norm": 0.1310533732175827,
      "learning_rate": 6.303999999999999e-05,
      "loss": 0.02287290096282959,
      "step": 6850
    },
    {
      "epoch": 2.0048,
      "grad_norm": 0.11340674012899399,
      "learning_rate": 6.284e-05,
      "loss": 0.02178637236356735,
      "step": 6860
    },
    {
      "epoch": 2.0058,
      "grad_norm": 0.10936598479747772,
      "learning_rate": 6.264e-05,
      "loss": 0.02486586421728134,
      "step": 6870
    },
    {
      "epoch": 2.0068,
      "grad_norm": 0.14033988118171692,
      "learning_rate": 6.244e-05,
      "loss": 0.022671811282634735,
      "step": 6880
    },
    {
      "epoch": 2.0078,
      "grad_norm": 0.1289263367652893,
      "learning_rate": 6.224e-05,
      "loss": 0.021725392341613768,
      "step": 6890
    },
    {
      "epoch": 2.0088,
      "grad_norm": 0.13452701270580292,
      "learning_rate": 6.204e-05,
      "loss": 0.02472696304321289,
      "step": 6900
    },
    {
      "epoch": 2.0098,
      "grad_norm": 0.14499792456626892,
      "learning_rate": 6.184e-05,
      "loss": 0.022322843968868255,
      "step": 6910
    },
    {
      "epoch": 2.0108,
      "grad_norm": 0.1398196518421173,
      "learning_rate": 6.164e-05,
      "loss": 0.022544071078300476,
      "step": 6920
    },
    {
      "epoch": 2.0118,
      "grad_norm": 0.1454416811466217,
      "learning_rate": 6.144e-05,
      "loss": 0.022087126970291138,
      "step": 6930
    },
    {
      "epoch": 2.0128,
      "grad_norm": 0.12601584196090698,
      "learning_rate": 6.124e-05,
      "loss": 0.023739957809448244,
      "step": 6940
    },
    {
      "epoch": 2.0138,
      "grad_norm": 0.13553965091705322,
      "learning_rate": 6.104000000000001e-05,
      "loss": 0.024189202487468718,
      "step": 6950
    },
    {
      "epoch": 2.0148,
      "grad_norm": 0.17853738367557526,
      "learning_rate": 6.084000000000001e-05,
      "loss": 0.024137826263904573,
      "step": 6960
    },
    {
      "epoch": 2.0158,
      "grad_norm": 0.1292908936738968,
      "learning_rate": 6.064000000000001e-05,
      "loss": 0.029603716731071473,
      "step": 6970
    },
    {
      "epoch": 2.0168,
      "grad_norm": 0.13389606773853302,
      "learning_rate": 6.044000000000001e-05,
      "loss": 0.02270825654268265,
      "step": 6980
    },
    {
      "epoch": 2.0178,
      "grad_norm": 0.10673966258764267,
      "learning_rate": 6.0240000000000006e-05,
      "loss": 0.0229933500289917,
      "step": 6990
    },
    {
      "epoch": 2.0188,
      "grad_norm": 0.13245923817157745,
      "learning_rate": 6.004000000000001e-05,
      "loss": 0.02357790470123291,
      "step": 7000
    },
    {
      "epoch": 2.0198,
      "grad_norm": 0.17045632004737854,
      "learning_rate": 5.984000000000001e-05,
      "loss": 0.02287629395723343,
      "step": 7010
    },
    {
      "epoch": 2.0208,
      "grad_norm": 0.12083585560321808,
      "learning_rate": 5.9640000000000005e-05,
      "loss": 0.02358904778957367,
      "step": 7020
    },
    {
      "epoch": 2.0218,
      "grad_norm": 0.13300317525863647,
      "learning_rate": 5.944000000000001e-05,
      "loss": 0.023620645701885223,
      "step": 7030
    },
    {
      "epoch": 2.0228,
      "grad_norm": 0.15547838807106018,
      "learning_rate": 5.924000000000001e-05,
      "loss": 0.023183451592922212,
      "step": 7040
    },
    {
      "epoch": 2.0238,
      "grad_norm": 0.16687844693660736,
      "learning_rate": 5.9040000000000004e-05,
      "loss": 0.023912055790424346,
      "step": 7050
    },
    {
      "epoch": 2.0248,
      "grad_norm": 0.12975580990314484,
      "learning_rate": 5.8840000000000006e-05,
      "loss": 0.023545092344284056,
      "step": 7060
    },
    {
      "epoch": 2.0258,
      "grad_norm": 0.11908718943595886,
      "learning_rate": 5.864000000000001e-05,
      "loss": 0.024130003154277803,
      "step": 7070
    },
    {
      "epoch": 2.0268,
      "grad_norm": 0.13249467313289642,
      "learning_rate": 5.844e-05,
      "loss": 0.02512686550617218,
      "step": 7080
    },
    {
      "epoch": 2.0278,
      "grad_norm": 0.12178528308868408,
      "learning_rate": 5.8240000000000005e-05,
      "loss": 0.02444648593664169,
      "step": 7090
    },
    {
      "epoch": 2.0288,
      "grad_norm": 0.12476595491170883,
      "learning_rate": 5.804000000000001e-05,
      "loss": 0.022948546707630156,
      "step": 7100
    },
    {
      "epoch": 2.0298,
      "grad_norm": 0.15463438630104065,
      "learning_rate": 5.784000000000001e-05,
      "loss": 0.025882315635681153,
      "step": 7110
    },
    {
      "epoch": 2.0308,
      "grad_norm": 0.16765928268432617,
      "learning_rate": 5.7640000000000004e-05,
      "loss": 0.024091361463069914,
      "step": 7120
    },
    {
      "epoch": 3.0006,
      "grad_norm": 0.08396776765584946,
      "learning_rate": 5.7440000000000006e-05,
      "loss": 0.02553088665008545,
      "step": 7130
    },
    {
      "epoch": 3.0016,
      "grad_norm": 0.10819737613201141,
      "learning_rate": 5.724000000000001e-05,
      "loss": 0.019945000112056733,
      "step": 7140
    },
    {
      "epoch": 3.0026,
      "grad_norm": 0.1035761684179306,
      "learning_rate": 5.704e-05,
      "loss": 0.021518710255622863,
      "step": 7150
    },
    {
      "epoch": 3.0036,
      "grad_norm": 0.11577849835157394,
      "learning_rate": 5.6840000000000005e-05,
      "loss": 0.020588450133800507,
      "step": 7160
    },
    {
      "epoch": 3.0046,
      "grad_norm": 0.0943162813782692,
      "learning_rate": 5.6640000000000007e-05,
      "loss": 0.019012409448623657,
      "step": 7170
    },
    {
      "epoch": 3.0056,
      "grad_norm": 0.10145782679319382,
      "learning_rate": 5.644e-05,
      "loss": 0.020459675788879396,
      "step": 7180
    },
    {
      "epoch": 3.0066,
      "grad_norm": 0.11473723500967026,
      "learning_rate": 5.6240000000000004e-05,
      "loss": 0.02082264721393585,
      "step": 7190
    },
    {
      "epoch": 3.0076,
      "grad_norm": 0.14834633469581604,
      "learning_rate": 5.6040000000000006e-05,
      "loss": 0.02009241431951523,
      "step": 7200
    },
    {
      "epoch": 3.0086,
      "grad_norm": 0.18625617027282715,
      "learning_rate": 5.584e-05,
      "loss": 0.022254915535449983,
      "step": 7210
    },
    {
      "epoch": 3.0096,
      "grad_norm": 0.09577304124832153,
      "learning_rate": 5.564e-05,
      "loss": 0.02103894203901291,
      "step": 7220
    },
    {
      "epoch": 3.0106,
      "grad_norm": 0.11397113651037216,
      "learning_rate": 5.5440000000000005e-05,
      "loss": 0.020081984996795654,
      "step": 7230
    },
    {
      "epoch": 3.0116,
      "grad_norm": 0.11981774121522903,
      "learning_rate": 5.524e-05,
      "loss": 0.020449602603912355,
      "step": 7240
    },
    {
      "epoch": 3.0126,
      "grad_norm": 0.12391442805528641,
      "learning_rate": 5.504e-05,
      "loss": 0.021178384125232697,
      "step": 7250
    },
    {
      "epoch": 3.0136,
      "grad_norm": 0.10961490869522095,
      "learning_rate": 5.4840000000000003e-05,
      "loss": 0.020945684611797334,
      "step": 7260
    },
    {
      "epoch": 3.0146,
      "grad_norm": 0.13801056146621704,
      "learning_rate": 5.4640000000000005e-05,
      "loss": 0.01994608938694,
      "step": 7270
    },
    {
      "epoch": 3.0156,
      "grad_norm": 0.09954143315553665,
      "learning_rate": 5.444e-05,
      "loss": 0.0260534793138504,
      "step": 7280
    },
    {
      "epoch": 3.0166,
      "grad_norm": 0.09468942880630493,
      "learning_rate": 5.424e-05,
      "loss": 0.019537252187728883,
      "step": 7290
    },
    {
      "epoch": 3.0176,
      "grad_norm": 0.09855340421199799,
      "learning_rate": 5.4040000000000004e-05,
      "loss": 0.021300722658634186,
      "step": 7300
    },
    {
      "epoch": 3.0186,
      "grad_norm": 0.14724481105804443,
      "learning_rate": 5.384e-05,
      "loss": 0.021099399030208587,
      "step": 7310
    },
    {
      "epoch": 3.0196,
      "grad_norm": 0.1300877034664154,
      "learning_rate": 5.364e-05,
      "loss": 0.021501822769641875,
      "step": 7320
    },
    {
      "epoch": 3.0206,
      "grad_norm": 0.09704280644655228,
      "learning_rate": 5.344e-05,
      "loss": 0.020881634950637818,
      "step": 7330
    },
    {
      "epoch": 3.0216,
      "grad_norm": 0.11273547261953354,
      "learning_rate": 5.324e-05,
      "loss": 0.02265263944864273,
      "step": 7340
    },
    {
      "epoch": 3.0226,
      "grad_norm": 0.116827592253685,
      "learning_rate": 5.304e-05,
      "loss": 0.021155886352062225,
      "step": 7350
    },
    {
      "epoch": 3.0236,
      "grad_norm": 0.13560475409030914,
      "learning_rate": 5.284e-05,
      "loss": 0.02109742909669876,
      "step": 7360
    },
    {
      "epoch": 3.0246,
      "grad_norm": 0.11204122006893158,
      "learning_rate": 5.264e-05,
      "loss": 0.021917839348316193,
      "step": 7370
    },
    {
      "epoch": 3.0256,
      "grad_norm": 0.126216858625412,
      "learning_rate": 5.244e-05,
      "loss": 0.021687588095664977,
      "step": 7380
    },
    {
      "epoch": 3.0266,
      "grad_norm": 0.10559123009443283,
      "learning_rate": 5.224e-05,
      "loss": 0.023697364330291747,
      "step": 7390
    },
    {
      "epoch": 3.0276,
      "grad_norm": 0.11427237838506699,
      "learning_rate": 5.204e-05,
      "loss": 0.022155900299549103,
      "step": 7400
    },
    {
      "epoch": 3.0286,
      "grad_norm": 0.14890673756599426,
      "learning_rate": 5.184e-05,
      "loss": 0.021498112380504607,
      "step": 7410
    },
    {
      "epoch": 3.0296,
      "grad_norm": 0.09329762309789658,
      "learning_rate": 5.164e-05,
      "loss": 0.021474449336528777,
      "step": 7420
    },
    {
      "epoch": 3.0306,
      "grad_norm": 0.09695407003164291,
      "learning_rate": 5.144e-05,
      "loss": 0.021487128734588624,
      "step": 7430
    },
    {
      "epoch": 4.0004,
      "grad_norm": 0.1085561066865921,
      "learning_rate": 5.124e-05,
      "loss": 0.024187105894088744,
      "step": 7440
    },
    {
      "epoch": 4.0014,
      "grad_norm": 0.09849539399147034,
      "learning_rate": 5.104e-05,
      "loss": 0.01914684921503067,
      "step": 7450
    },
    {
      "epoch": 4.0024,
      "grad_norm": 0.12541185319423676,
      "learning_rate": 5.084e-05,
      "loss": 0.020627421140670777,
      "step": 7460
    },
    {
      "epoch": 4.0034,
      "grad_norm": 0.08327507227659225,
      "learning_rate": 5.0639999999999996e-05,
      "loss": 0.019001516699790954,
      "step": 7470
    },
    {
      "epoch": 4.0044,
      "grad_norm": 0.10637800395488739,
      "learning_rate": 5.044e-05,
      "loss": 0.017808449268341065,
      "step": 7480
    },
    {
      "epoch": 4.0054,
      "grad_norm": 0.09127842634916306,
      "learning_rate": 5.024e-05,
      "loss": 0.019047291576862337,
      "step": 7490
    },
    {
      "epoch": 4.0064,
      "grad_norm": 0.0757402703166008,
      "learning_rate": 5.0039999999999995e-05,
      "loss": 0.018700633943080903,
      "step": 7500
    },
    {
      "epoch": 4.0074,
      "grad_norm": 0.08042102307081223,
      "learning_rate": 4.9840000000000004e-05,
      "loss": 0.018940824270248412,
      "step": 7510
    },
    {
      "epoch": 4.0084,
      "grad_norm": 0.21641090512275696,
      "learning_rate": 4.9640000000000006e-05,
      "loss": 0.020995579659938812,
      "step": 7520
    },
    {
      "epoch": 4.0094,
      "grad_norm": 0.12526147067546844,
      "learning_rate": 4.944e-05,
      "loss": 0.019232508540153504,
      "step": 7530
    },
    {
      "epoch": 4.0104,
      "grad_norm": 0.10949081927537918,
      "learning_rate": 4.924e-05,
      "loss": 0.01903458535671234,
      "step": 7540
    },
    {
      "epoch": 4.0114,
      "grad_norm": 0.09758574515581131,
      "learning_rate": 4.9040000000000005e-05,
      "loss": 0.01993790864944458,
      "step": 7550
    },
    {
      "epoch": 4.0124,
      "grad_norm": 0.09602127224206924,
      "learning_rate": 4.884e-05,
      "loss": 0.021270765364170073,
      "step": 7560
    },
    {
      "epoch": 4.0134,
      "grad_norm": 0.08903051167726517,
      "learning_rate": 4.864e-05,
      "loss": 0.018507078289985657,
      "step": 7570
    },
    {
      "epoch": 4.0144,
      "grad_norm": 0.09527137875556946,
      "learning_rate": 4.8440000000000004e-05,
      "loss": 0.01924394816160202,
      "step": 7580
    },
    {
      "epoch": 4.0154,
      "grad_norm": 0.1218629702925682,
      "learning_rate": 4.824e-05,
      "loss": 0.02654995024204254,
      "step": 7590
    },
    {
      "epoch": 4.0164,
      "grad_norm": 0.0813978984951973,
      "learning_rate": 4.804e-05,
      "loss": 0.018305253982543946,
      "step": 7600
    },
    {
      "epoch": 4.0174,
      "grad_norm": 0.08926694095134735,
      "learning_rate": 4.784e-05,
      "loss": 0.01904298961162567,
      "step": 7610
    },
    {
      "epoch": 4.0184,
      "grad_norm": 0.084304079413414,
      "learning_rate": 4.7640000000000005e-05,
      "loss": 0.019423790276050568,
      "step": 7620
    },
    {
      "epoch": 4.0194,
      "grad_norm": 0.08805521577596664,
      "learning_rate": 4.744e-05,
      "loss": 0.019075404107570647,
      "step": 7630
    },
    {
      "epoch": 4.0204,
      "grad_norm": 0.07542338967323303,
      "learning_rate": 4.724e-05,
      "loss": 0.01921938955783844,
      "step": 7640
    },
    {
      "epoch": 4.0214,
      "grad_norm": 0.11425633728504181,
      "learning_rate": 4.7040000000000004e-05,
      "loss": 0.01996878832578659,
      "step": 7650
    },
    {
      "epoch": 4.0224,
      "grad_norm": 0.11465906351804733,
      "learning_rate": 4.684e-05,
      "loss": 0.019843164086341857,
      "step": 7660
    },
    {
      "epoch": 4.0234,
      "grad_norm": 0.12781135737895966,
      "learning_rate": 4.664e-05,
      "loss": 0.01992042511701584,
      "step": 7670
    },
    {
      "epoch": 4.0244,
      "grad_norm": 0.09651028364896774,
      "learning_rate": 4.644e-05,
      "loss": 0.019879813492298126,
      "step": 7680
    },
    {
      "epoch": 4.0254,
      "grad_norm": 0.10908477753400803,
      "learning_rate": 4.624e-05,
      "loss": 0.020540766417980194,
      "step": 7690
    },
    {
      "epoch": 4.0264,
      "grad_norm": 0.11591225117444992,
      "learning_rate": 4.604e-05,
      "loss": 0.020340165495872496,
      "step": 7700
    },
    {
      "epoch": 4.0274,
      "grad_norm": 0.0748429000377655,
      "learning_rate": 4.584e-05,
      "loss": 0.02119535654783249,
      "step": 7710
    },
    {
      "epoch": 4.0284,
      "grad_norm": 0.06569884717464447,
      "learning_rate": 4.564e-05,
      "loss": 0.019903770089149474,
      "step": 7720
    },
    {
      "epoch": 4.0294,
      "grad_norm": 0.08006953448057175,
      "learning_rate": 4.5440000000000005e-05,
      "loss": 0.020622527599334715,
      "step": 7730
    },
    {
      "epoch": 4.0304,
      "grad_norm": 0.12711670994758606,
      "learning_rate": 4.524000000000001e-05,
      "loss": 0.019541466236114503,
      "step": 7740
    },
    {
      "epoch": 5.0002,
      "grad_norm": 0.13896505534648895,
      "learning_rate": 4.504e-05,
      "loss": 0.020583076775074004,
      "step": 7750
    },
    {
      "epoch": 5.0012,
      "grad_norm": 0.08660919964313507,
      "learning_rate": 4.4840000000000004e-05,
      "loss": 0.016970406472682952,
      "step": 7760
    },
    {
      "epoch": 5.0022,
      "grad_norm": 0.11170201748609543,
      "learning_rate": 4.4640000000000006e-05,
      "loss": 0.018784117698669434,
      "step": 7770
    },
    {
      "epoch": 5.0032,
      "grad_norm": 0.08062026649713516,
      "learning_rate": 4.444e-05,
      "loss": 0.01944620758295059,
      "step": 7780
    },
    {
      "epoch": 5.0042,
      "grad_norm": 0.05483711510896683,
      "learning_rate": 4.424e-05,
      "loss": 0.01771644353866577,
      "step": 7790
    },
    {
      "epoch": 5.0052,
      "grad_norm": 0.08323200792074203,
      "learning_rate": 4.4040000000000005e-05,
      "loss": 0.01714482456445694,
      "step": 7800
    },
    {
      "epoch": 5.0062,
      "grad_norm": 0.11725310981273651,
      "learning_rate": 4.384e-05,
      "loss": 0.01833456605672836,
      "step": 7810
    },
    {
      "epoch": 5.0072,
      "grad_norm": 0.08281055092811584,
      "learning_rate": 4.364e-05,
      "loss": 0.017857730388641357,
      "step": 7820
    },
    {
      "epoch": 5.0082,
      "grad_norm": 0.08471981436014175,
      "learning_rate": 4.3440000000000004e-05,
      "loss": 0.0180238738656044,
      "step": 7830
    },
    {
      "epoch": 5.0092,
      "grad_norm": 0.07369330525398254,
      "learning_rate": 4.324e-05,
      "loss": 0.020367558300495147,
      "step": 7840
    },
    {
      "epoch": 5.0102,
      "grad_norm": 0.14078500866889954,
      "learning_rate": 4.304e-05,
      "loss": 0.018425200879573823,
      "step": 7850
    },
    {
      "epoch": 5.0112,
      "grad_norm": 0.09964534640312195,
      "learning_rate": 4.284e-05,
      "loss": 0.01857619732618332,
      "step": 7860
    },
    {
      "epoch": 5.0122,
      "grad_norm": 0.11316290497779846,
      "learning_rate": 4.2640000000000005e-05,
      "loss": 0.01818966567516327,
      "step": 7870
    },
    {
      "epoch": 5.0132,
      "grad_norm": 0.08515363186597824,
      "learning_rate": 4.244e-05,
      "loss": 0.01757376194000244,
      "step": 7880
    },
    {
      "epoch": 5.0142,
      "grad_norm": 0.07923150807619095,
      "learning_rate": 4.224e-05,
      "loss": 0.017273730039596556,
      "step": 7890
    },
    {
      "epoch": 5.0152,
      "grad_norm": 0.13433587551116943,
      "learning_rate": 4.2040000000000004e-05,
      "loss": 0.02332073152065277,
      "step": 7900
    },
    {
      "epoch": 5.0162,
      "grad_norm": 0.08601813018321991,
      "learning_rate": 4.184e-05,
      "loss": 0.017112986743450166,
      "step": 7910
    },
    {
      "epoch": 5.0172,
      "grad_norm": 0.09130693227052689,
      "learning_rate": 4.164e-05,
      "loss": 0.017460432648658753,
      "step": 7920
    },
    {
      "epoch": 5.0182,
      "grad_norm": 0.09017442166805267,
      "learning_rate": 4.144e-05,
      "loss": 0.01737082749605179,
      "step": 7930
    },
    {
      "epoch": 5.0192,
      "grad_norm": 0.08579915761947632,
      "learning_rate": 4.124e-05,
      "loss": 0.018616248667240144,
      "step": 7940
    },
    {
      "epoch": 5.0202,
      "grad_norm": 0.08709147572517395,
      "learning_rate": 4.104e-05,
      "loss": 0.018517845869064332,
      "step": 7950
    },
    {
      "epoch": 5.0212,
      "grad_norm": 0.1221432164311409,
      "learning_rate": 4.084e-05,
      "loss": 0.017741505801677705,
      "step": 7960
    },
    {
      "epoch": 5.0222,
      "grad_norm": 0.08435460925102234,
      "learning_rate": 4.064e-05,
      "loss": 0.018242837488651277,
      "step": 7970
    },
    {
      "epoch": 5.0232,
      "grad_norm": 0.07412870973348618,
      "learning_rate": 4.044e-05,
      "loss": 0.017411427199840547,
      "step": 7980
    },
    {
      "epoch": 5.0242,
      "grad_norm": 0.08610522001981735,
      "learning_rate": 4.024e-05,
      "loss": 0.01860019862651825,
      "step": 7990
    },
    {
      "epoch": 5.0252,
      "grad_norm": 0.06387773156166077,
      "learning_rate": 4.004e-05,
      "loss": 0.01812358796596527,
      "step": 8000
    },
    {
      "epoch": 5.0262,
      "grad_norm": 0.09226816147565842,
      "learning_rate": 3.984e-05,
      "loss": 0.018133127689361574,
      "step": 8010
    },
    {
      "epoch": 5.0272,
      "grad_norm": 0.08140015602111816,
      "learning_rate": 3.964e-05,
      "loss": 0.018813768029212953,
      "step": 8020
    },
    {
      "epoch": 5.0282,
      "grad_norm": 0.06501121819019318,
      "learning_rate": 3.944e-05,
      "loss": 0.019245068728923797,
      "step": 8030
    },
    {
      "epoch": 5.0292,
      "grad_norm": 0.09711839258670807,
      "learning_rate": 3.9240000000000004e-05,
      "loss": 0.01863838583230972,
      "step": 8040
    },
    {
      "epoch": 5.0302,
      "grad_norm": 0.0660022497177124,
      "learning_rate": 3.9040000000000006e-05,
      "loss": 0.017671407759189607,
      "step": 8050
    },
    {
      "epoch": 5.0312,
      "grad_norm": 0.08400487899780273,
      "learning_rate": 3.884e-05,
      "loss": 0.017706127464771272,
      "step": 8060
    },
    {
      "epoch": 6.001,
      "grad_norm": 0.07883397489786148,
      "learning_rate": 3.864e-05,
      "loss": 0.01782989501953125,
      "step": 8070
    },
    {
      "epoch": 6.002,
      "grad_norm": 0.08091418445110321,
      "learning_rate": 3.8440000000000005e-05,
      "loss": 0.016622547805309296,
      "step": 8080
    },
    {
      "epoch": 6.003,
      "grad_norm": 0.07292184978723526,
      "learning_rate": 3.8240000000000007e-05,
      "loss": 0.017438323795795442,
      "step": 8090
    },
    {
      "epoch": 6.004,
      "grad_norm": 0.07673656195402145,
      "learning_rate": 3.804e-05,
      "loss": 0.01731536090373993,
      "step": 8100
    },
    {
      "epoch": 6.005,
      "grad_norm": 0.07159565389156342,
      "learning_rate": 3.7840000000000004e-05,
      "loss": 0.01596030592918396,
      "step": 8110
    },
    {
      "epoch": 6.006,
      "grad_norm": 0.06656063348054886,
      "learning_rate": 3.7640000000000006e-05,
      "loss": 0.017606176435947418,
      "step": 8120
    },
    {
      "epoch": 6.007,
      "grad_norm": 0.07145673036575317,
      "learning_rate": 3.744e-05,
      "loss": 0.016346777975559234,
      "step": 8130
    },
    {
      "epoch": 6.008,
      "grad_norm": 0.09691258519887924,
      "learning_rate": 3.724e-05,
      "loss": 0.016676793992519378,
      "step": 8140
    },
    {
      "epoch": 6.009,
      "grad_norm": 0.11464672535657883,
      "learning_rate": 3.7040000000000005e-05,
      "loss": 0.017788544297218323,
      "step": 8150
    },
    {
      "epoch": 6.01,
      "grad_norm": 0.07520481944084167,
      "learning_rate": 3.684e-05,
      "loss": 0.01642948091030121,
      "step": 8160
    },
    {
      "epoch": 6.011,
      "grad_norm": 0.0775238573551178,
      "learning_rate": 3.664e-05,
      "loss": 0.017403733730316163,
      "step": 8170
    },
    {
      "epoch": 6.012,
      "grad_norm": 0.06533999741077423,
      "learning_rate": 3.6440000000000003e-05,
      "loss": 0.0171025887131691,
      "step": 8180
    },
    {
      "epoch": 6.013,
      "grad_norm": 0.09260275959968567,
      "learning_rate": 3.624e-05,
      "loss": 0.017036963999271394,
      "step": 8190
    },
    {
      "epoch": 6.014,
      "grad_norm": 0.10016588866710663,
      "learning_rate": 3.604e-05,
      "loss": 0.017642612755298614,
      "step": 8200
    },
    {
      "epoch": 6.015,
      "grad_norm": 0.07862694561481476,
      "learning_rate": 3.584e-05,
      "loss": 0.017080143094062805,
      "step": 8210
    },
    {
      "epoch": 6.016,
      "grad_norm": 0.07004833966493607,
      "learning_rate": 3.5640000000000004e-05,
      "loss": 0.019901210069656373,
      "step": 8220
    },
    {
      "epoch": 6.017,
      "grad_norm": 0.07592538744211197,
      "learning_rate": 3.544e-05,
      "loss": 0.016654495894908906,
      "step": 8230
    },
    {
      "epoch": 6.018,
      "grad_norm": 0.08571147918701172,
      "learning_rate": 3.524e-05,
      "loss": 0.01710202395915985,
      "step": 8240
    },
    {
      "epoch": 6.019,
      "grad_norm": 0.077202208340168,
      "learning_rate": 3.504e-05,
      "loss": 0.017121993005275726,
      "step": 8250
    },
    {
      "epoch": 6.02,
      "grad_norm": 0.08266453444957733,
      "learning_rate": 3.484e-05,
      "loss": 0.017408430576324463,
      "step": 8260
    },
    {
      "epoch": 6.021,
      "grad_norm": 0.06414631009101868,
      "learning_rate": 3.464e-05,
      "loss": 0.01690840870141983,
      "step": 8270
    },
    {
      "epoch": 6.022,
      "grad_norm": 0.06548868864774704,
      "learning_rate": 3.444e-05,
      "loss": 0.017626507580280303,
      "step": 8280
    },
    {
      "epoch": 6.023,
      "grad_norm": 0.10201598703861237,
      "learning_rate": 3.424e-05,
      "loss": 0.01741800308227539,
      "step": 8290
    },
    {
      "epoch": 6.024,
      "grad_norm": 0.06847354769706726,
      "learning_rate": 3.404e-05,
      "loss": 0.018401354551315308,
      "step": 8300
    },
    {
      "epoch": 6.025,
      "grad_norm": 0.07972142100334167,
      "learning_rate": 3.384e-05,
      "loss": 0.017628875374794007,
      "step": 8310
    },
    {
      "epoch": 6.026,
      "grad_norm": 0.08320039510726929,
      "learning_rate": 3.3639999999999996e-05,
      "loss": 0.0165034145116806,
      "step": 8320
    },
    {
      "epoch": 6.027,
      "grad_norm": 0.07521474361419678,
      "learning_rate": 3.344e-05,
      "loss": 0.017265401780605316,
      "step": 8330
    },
    {
      "epoch": 6.028,
      "grad_norm": 0.05939231067895889,
      "learning_rate": 3.324e-05,
      "loss": 0.01859947293996811,
      "step": 8340
    },
    {
      "epoch": 6.029,
      "grad_norm": 0.06058882176876068,
      "learning_rate": 3.304e-05,
      "loss": 0.018016797304153443,
      "step": 8350
    },
    {
      "epoch": 6.03,
      "grad_norm": 0.1014842763543129,
      "learning_rate": 3.2840000000000004e-05,
      "loss": 0.01612357497215271,
      "step": 8360
    },
    {
      "epoch": 6.031,
      "grad_norm": 0.07388660311698914,
      "learning_rate": 3.2640000000000006e-05,
      "loss": 0.01661892682313919,
      "step": 8370
    },
    {
      "epoch": 7.0008,
      "grad_norm": 0.06829363852739334,
      "learning_rate": 3.244e-05,
      "loss": 0.01718652993440628,
      "step": 8380
    },
    {
      "epoch": 7.0018,
      "grad_norm": 0.08097092807292938,
      "learning_rate": 3.224e-05,
      "loss": 0.015461650490760804,
      "step": 8390
    },
    {
      "epoch": 7.0028,
      "grad_norm": 0.07897481322288513,
      "learning_rate": 3.2040000000000005e-05,
      "loss": 0.01638838052749634,
      "step": 8400
    },
    {
      "epoch": 7.0038,
      "grad_norm": 0.07321804761886597,
      "learning_rate": 3.184e-05,
      "loss": 0.016494518518447875,
      "step": 8410
    },
    {
      "epoch": 7.0048,
      "grad_norm": 0.06354601681232452,
      "learning_rate": 3.164e-05,
      "loss": 0.016121280193328858,
      "step": 8420
    },
    {
      "epoch": 7.0058,
      "grad_norm": 0.05840650200843811,
      "learning_rate": 3.1440000000000004e-05,
      "loss": 0.016373760998249054,
      "step": 8430
    },
    {
      "epoch": 7.0068,
      "grad_norm": 0.06818988174200058,
      "learning_rate": 3.1240000000000006e-05,
      "loss": 0.015844850242137908,
      "step": 8440
    },
    {
      "epoch": 7.0078,
      "grad_norm": 0.07914569228887558,
      "learning_rate": 3.104e-05,
      "loss": 0.015179328620433807,
      "step": 8450
    },
    {
      "epoch": 7.0088,
      "grad_norm": 0.06279801577329636,
      "learning_rate": 3.084e-05,
      "loss": 0.01618899405002594,
      "step": 8460
    },
    {
      "epoch": 7.0098,
      "grad_norm": 0.06824712455272675,
      "learning_rate": 3.0640000000000005e-05,
      "loss": 0.015947595238685608,
      "step": 8470
    },
    {
      "epoch": 7.0108,
      "grad_norm": 0.07371484488248825,
      "learning_rate": 3.0440000000000003e-05,
      "loss": 0.017725233733654023,
      "step": 8480
    },
    {
      "epoch": 7.0118,
      "grad_norm": 0.06131076440215111,
      "learning_rate": 3.0240000000000002e-05,
      "loss": 0.016513384878635406,
      "step": 8490
    },
    {
      "epoch": 7.0128,
      "grad_norm": 0.07285471260547638,
      "learning_rate": 3.004e-05,
      "loss": 0.016372813284397124,
      "step": 8500
    },
    {
      "epoch": 7.0138,
      "grad_norm": 0.07750345766544342,
      "learning_rate": 2.9840000000000002e-05,
      "loss": 0.01583592891693115,
      "step": 8510
    },
    {
      "epoch": 7.0148,
      "grad_norm": 0.08645971119403839,
      "learning_rate": 2.964e-05,
      "loss": 0.015735095739364623,
      "step": 8520
    },
    {
      "epoch": 7.0158,
      "grad_norm": 0.07818005234003067,
      "learning_rate": 2.944e-05,
      "loss": 0.018969209492206575,
      "step": 8530
    },
    {
      "epoch": 7.0168,
      "grad_norm": 0.09315695613622665,
      "learning_rate": 2.924e-05,
      "loss": 0.015404698252677918,
      "step": 8540
    },
    {
      "epoch": 7.0178,
      "grad_norm": 0.07144665718078613,
      "learning_rate": 2.904e-05,
      "loss": 0.01541641354560852,
      "step": 8550
    },
    {
      "epoch": 7.0188,
      "grad_norm": 0.06964854151010513,
      "learning_rate": 2.8840000000000002e-05,
      "loss": 0.016346700489521027,
      "step": 8560
    },
    {
      "epoch": 7.0198,
      "grad_norm": 0.06505166739225388,
      "learning_rate": 2.864e-05,
      "loss": 0.01625395566225052,
      "step": 8570
    },
    {
      "epoch": 7.0208,
      "grad_norm": 0.06147618219256401,
      "learning_rate": 2.844e-05,
      "loss": 0.016307108104228973,
      "step": 8580
    },
    {
      "epoch": 7.0218,
      "grad_norm": 0.08299954235553741,
      "learning_rate": 2.824e-05,
      "loss": 0.016882283985614775,
      "step": 8590
    },
    {
      "epoch": 7.0228,
      "grad_norm": 0.07017214596271515,
      "learning_rate": 2.804e-05,
      "loss": 0.015500307083129883,
      "step": 8600
    },
    {
      "epoch": 7.0238,
      "grad_norm": 0.06280285865068436,
      "learning_rate": 2.7839999999999998e-05,
      "loss": 0.01642789989709854,
      "step": 8610
    },
    {
      "epoch": 7.0248,
      "grad_norm": 0.07035724818706512,
      "learning_rate": 2.764e-05,
      "loss": 0.0161510095000267,
      "step": 8620
    },
    {
      "epoch": 7.0258,
      "grad_norm": 0.05377668887376785,
      "learning_rate": 2.7439999999999998e-05,
      "loss": 0.01608278453350067,
      "step": 8630
    },
    {
      "epoch": 7.0268,
      "grad_norm": 0.05881371721625328,
      "learning_rate": 2.724e-05,
      "loss": 0.01537017673254013,
      "step": 8640
    },
    {
      "epoch": 7.0278,
      "grad_norm": 0.05430138483643532,
      "learning_rate": 2.704e-05,
      "loss": 0.016539952158927916,
      "step": 8650
    },
    {
      "epoch": 7.0288,
      "grad_norm": 0.056010618805885315,
      "learning_rate": 2.6840000000000004e-05,
      "loss": 0.016414332389831542,
      "step": 8660
    },
    {
      "epoch": 7.0298,
      "grad_norm": 0.08144864439964294,
      "learning_rate": 2.6640000000000002e-05,
      "loss": 0.015341292321681976,
      "step": 8670
    },
    {
      "epoch": 7.0308,
      "grad_norm": 0.06344097852706909,
      "learning_rate": 2.6440000000000004e-05,
      "loss": 0.016318969428539276,
      "step": 8680
    },
    {
      "epoch": 8.0006,
      "grad_norm": 0.05204887315630913,
      "learning_rate": 2.6240000000000003e-05,
      "loss": 0.01676938831806183,
      "step": 8690
    },
    {
      "epoch": 8.0016,
      "grad_norm": 0.07362329214811325,
      "learning_rate": 2.6040000000000005e-05,
      "loss": 0.01465071588754654,
      "step": 8700
    },
    {
      "epoch": 8.0026,
      "grad_norm": 0.05992142856121063,
      "learning_rate": 2.5840000000000003e-05,
      "loss": 0.01574086844921112,
      "step": 8710
    },
    {
      "epoch": 8.0036,
      "grad_norm": 0.056651197373867035,
      "learning_rate": 2.5640000000000002e-05,
      "loss": 0.015629269182682037,
      "step": 8720
    },
    {
      "epoch": 8.0046,
      "grad_norm": 0.057444050908088684,
      "learning_rate": 2.5440000000000004e-05,
      "loss": 0.015164247155189515,
      "step": 8730
    },
    {
      "epoch": 8.0056,
      "grad_norm": 0.09881485253572464,
      "learning_rate": 2.5240000000000002e-05,
      "loss": 0.016363966464996337,
      "step": 8740
    },
    {
      "epoch": 8.0066,
      "grad_norm": 0.07986856251955032,
      "learning_rate": 2.504e-05,
      "loss": 0.01639338582754135,
      "step": 8750
    },
    {
      "epoch": 8.0076,
      "grad_norm": 0.0682622492313385,
      "learning_rate": 2.4840000000000003e-05,
      "loss": 0.015499250590801239,
      "step": 8760
    },
    {
      "epoch": 8.0086,
      "grad_norm": 0.09517238289117813,
      "learning_rate": 2.464e-05,
      "loss": 0.016006806492805482,
      "step": 8770
    },
    {
      "epoch": 8.0096,
      "grad_norm": 0.07619570940732956,
      "learning_rate": 2.4440000000000003e-05,
      "loss": 0.015566851198673248,
      "step": 8780
    },
    {
      "epoch": 8.0106,
      "grad_norm": 0.0652938038110733,
      "learning_rate": 2.4240000000000002e-05,
      "loss": 0.016661147773265838,
      "step": 8790
    },
    {
      "epoch": 8.0116,
      "grad_norm": 0.06141452118754387,
      "learning_rate": 2.404e-05,
      "loss": 0.01581350266933441,
      "step": 8800
    },
    {
      "epoch": 8.0126,
      "grad_norm": 0.06693390756845474,
      "learning_rate": 2.3840000000000002e-05,
      "loss": 0.016309624910354613,
      "step": 8810
    },
    {
      "epoch": 8.0136,
      "grad_norm": 0.0599714033305645,
      "learning_rate": 2.364e-05,
      "loss": 0.014834089577198029,
      "step": 8820
    },
    {
      "epoch": 8.0146,
      "grad_norm": 0.06985117495059967,
      "learning_rate": 2.344e-05,
      "loss": 0.01527884155511856,
      "step": 8830
    },
    {
      "epoch": 8.0156,
      "grad_norm": 0.07937555760145187,
      "learning_rate": 2.324e-05,
      "loss": 0.017245960235595704,
      "step": 8840
    },
    {
      "epoch": 8.0166,
      "grad_norm": 0.07799092680215836,
      "learning_rate": 2.304e-05,
      "loss": 0.014394041895866395,
      "step": 8850
    },
    {
      "epoch": 8.0176,
      "grad_norm": 0.06856916099786758,
      "learning_rate": 2.284e-05,
      "loss": 0.01503336727619171,
      "step": 8860
    },
    {
      "epoch": 8.0186,
      "grad_norm": 0.09324207156896591,
      "learning_rate": 2.264e-05,
      "loss": 0.016311252117156984,
      "step": 8870
    },
    {
      "epoch": 8.0196,
      "grad_norm": 0.06528984755277634,
      "learning_rate": 2.244e-05,
      "loss": 0.01599731594324112,
      "step": 8880
    },
    {
      "epoch": 8.0206,
      "grad_norm": 0.07623006403446198,
      "learning_rate": 2.224e-05,
      "loss": 0.016098654270172118,
      "step": 8890
    },
    {
      "epoch": 8.0216,
      "grad_norm": 0.0691269040107727,
      "learning_rate": 2.2040000000000002e-05,
      "loss": 0.016098056733608247,
      "step": 8900
    },
    {
      "epoch": 8.0226,
      "grad_norm": 0.06822088360786438,
      "learning_rate": 2.184e-05,
      "loss": 0.015027979016304016,
      "step": 8910
    },
    {
      "epoch": 8.0236,
      "grad_norm": 0.06876049190759659,
      "learning_rate": 2.1640000000000003e-05,
      "loss": 0.015229566395282746,
      "step": 8920
    },
    {
      "epoch": 8.0246,
      "grad_norm": 0.0731753334403038,
      "learning_rate": 2.144e-05,
      "loss": 0.014861150085926056,
      "step": 8930
    },
    {
      "epoch": 8.0256,
      "grad_norm": 0.05666269734501839,
      "learning_rate": 2.124e-05,
      "loss": 0.015309785306453706,
      "step": 8940
    },
    {
      "epoch": 8.0266,
      "grad_norm": 0.0933213010430336,
      "learning_rate": 2.1040000000000002e-05,
      "loss": 0.01527441293001175,
      "step": 8950
    },
    {
      "epoch": 8.0276,
      "grad_norm": 0.06596820801496506,
      "learning_rate": 2.084e-05,
      "loss": 0.015828940272331237,
      "step": 8960
    },
    {
      "epoch": 8.0286,
      "grad_norm": 0.08432097733020782,
      "learning_rate": 2.0640000000000002e-05,
      "loss": 0.016046036779880524,
      "step": 8970
    },
    {
      "epoch": 8.0296,
      "grad_norm": 0.06969591975212097,
      "learning_rate": 2.044e-05,
      "loss": 0.014962480962276458,
      "step": 8980
    },
    {
      "epoch": 8.0306,
      "grad_norm": 0.0731443241238594,
      "learning_rate": 2.024e-05,
      "loss": 0.01611236482858658,
      "step": 8990
    },
    {
      "epoch": 9.0004,
      "grad_norm": 0.07312334328889847,
      "learning_rate": 2.004e-05,
      "loss": 0.01628260165452957,
      "step": 9000
    },
    {
      "epoch": 9.0014,
      "grad_norm": 0.08494532853364944,
      "learning_rate": 1.984e-05,
      "loss": 0.014085091650485992,
      "step": 9010
    },
    {
      "epoch": 9.0024,
      "grad_norm": 0.06271655112504959,
      "learning_rate": 1.9640000000000002e-05,
      "loss": 0.014741827547550202,
      "step": 9020
    },
    {
      "epoch": 9.0034,
      "grad_norm": 0.05553199723362923,
      "learning_rate": 1.944e-05,
      "loss": 0.015085065364837646,
      "step": 9030
    },
    {
      "epoch": 9.0044,
      "grad_norm": 0.07189421355724335,
      "learning_rate": 1.924e-05,
      "loss": 0.015170367062091827,
      "step": 9040
    },
    {
      "epoch": 9.0054,
      "grad_norm": 0.07952412962913513,
      "learning_rate": 1.904e-05,
      "loss": 0.015609446167945861,
      "step": 9050
    },
    {
      "epoch": 9.0064,
      "grad_norm": 0.07108260691165924,
      "learning_rate": 1.8840000000000003e-05,
      "loss": 0.016404539346694946,
      "step": 9060
    },
    {
      "epoch": 9.0074,
      "grad_norm": 0.06802374124526978,
      "learning_rate": 1.864e-05,
      "loss": 0.015850432217121124,
      "step": 9070
    },
    {
      "epoch": 9.0084,
      "grad_norm": 0.050249889492988586,
      "learning_rate": 1.8440000000000003e-05,
      "loss": 0.015042982995510101,
      "step": 9080
    },
    {
      "epoch": 9.0094,
      "grad_norm": 0.06494880467653275,
      "learning_rate": 1.824e-05,
      "loss": 0.015259303152561188,
      "step": 9090
    },
    {
      "epoch": 9.0104,
      "grad_norm": 0.06114324554800987,
      "learning_rate": 1.804e-05,
      "loss": 0.015477609634399415,
      "step": 9100
    },
    {
      "epoch": 9.0114,
      "grad_norm": 0.06710667163133621,
      "learning_rate": 1.7840000000000002e-05,
      "loss": 0.015735583007335664,
      "step": 9110
    },
    {
      "epoch": 9.0124,
      "grad_norm": 0.058985739946365356,
      "learning_rate": 1.764e-05,
      "loss": 0.01523323804140091,
      "step": 9120
    },
    {
      "epoch": 9.0134,
      "grad_norm": 0.06335913389921188,
      "learning_rate": 1.7440000000000002e-05,
      "loss": 0.014204534888267516,
      "step": 9130
    },
    {
      "epoch": 9.0144,
      "grad_norm": 0.0716351643204689,
      "learning_rate": 1.724e-05,
      "loss": 0.014392800629138947,
      "step": 9140
    },
    {
      "epoch": 9.0154,
      "grad_norm": 0.08384208381175995,
      "learning_rate": 1.704e-05,
      "loss": 0.016598230600357054,
      "step": 9150
    },
    {
      "epoch": 9.0164,
      "grad_norm": 0.085759237408638,
      "learning_rate": 1.684e-05,
      "loss": 0.014103879034519196,
      "step": 9160
    },
    {
      "epoch": 9.0174,
      "grad_norm": 0.12926150858402252,
      "learning_rate": 1.664e-05,
      "loss": 0.015359596908092498,
      "step": 9170
    },
    {
      "epoch": 9.0184,
      "grad_norm": 0.07225863635540009,
      "learning_rate": 1.644e-05,
      "loss": 0.014975431561470031,
      "step": 9180
    },
    {
      "epoch": 9.0194,
      "grad_norm": 0.07080310583114624,
      "learning_rate": 1.624e-05,
      "loss": 0.015512678027153014,
      "step": 9190
    },
    {
      "epoch": 9.0204,
      "grad_norm": 0.06747028976678848,
      "learning_rate": 1.604e-05,
      "loss": 0.01568002849817276,
      "step": 9200
    },
    {
      "epoch": 9.0214,
      "grad_norm": 0.05957520753145218,
      "learning_rate": 1.584e-05,
      "loss": 0.015441593527793885,
      "step": 9210
    },
    {
      "epoch": 9.0224,
      "grad_norm": 0.06704767793416977,
      "learning_rate": 1.5640000000000003e-05,
      "loss": 0.014649063348770142,
      "step": 9220
    },
    {
      "epoch": 9.0234,
      "grad_norm": 0.06555522978305817,
      "learning_rate": 1.544e-05,
      "loss": 0.01482708305120468,
      "step": 9230
    },
    {
      "epoch": 9.0244,
      "grad_norm": 0.07707587629556656,
      "learning_rate": 1.5240000000000001e-05,
      "loss": 0.014254559576511384,
      "step": 9240
    },
    {
      "epoch": 9.0254,
      "grad_norm": 0.07337791472673416,
      "learning_rate": 1.5040000000000002e-05,
      "loss": 0.015043064951896667,
      "step": 9250
    },
    {
      "epoch": 9.0264,
      "grad_norm": 0.08570864051580429,
      "learning_rate": 1.4840000000000002e-05,
      "loss": 0.01574891060590744,
      "step": 9260
    },
    {
      "epoch": 9.0274,
      "grad_norm": 0.07305701076984406,
      "learning_rate": 1.464e-05,
      "loss": 0.015572662651538848,
      "step": 9270
    },
    {
      "epoch": 9.0284,
      "grad_norm": 0.059258975088596344,
      "learning_rate": 1.444e-05,
      "loss": 0.01588566303253174,
      "step": 9280
    },
    {
      "epoch": 9.0294,
      "grad_norm": 0.08041700720787048,
      "learning_rate": 1.4240000000000001e-05,
      "loss": 0.016104544699192046,
      "step": 9290
    },
    {
      "epoch": 9.0304,
      "grad_norm": 0.12365186959505081,
      "learning_rate": 1.4040000000000001e-05,
      "loss": 0.01534939706325531,
      "step": 9300
    },
    {
      "epoch": 10.0002,
      "grad_norm": 0.057674337178468704,
      "learning_rate": 1.384e-05,
      "loss": 0.015914039313793184,
      "step": 9310
    },
    {
      "epoch": 10.0012,
      "grad_norm": 0.06685669720172882,
      "learning_rate": 1.364e-05,
      "loss": 0.014044062793254852,
      "step": 9320
    },
    {
      "epoch": 10.0022,
      "grad_norm": 0.07209975272417068,
      "learning_rate": 1.344e-05,
      "loss": 0.014324517548084259,
      "step": 9330
    },
    {
      "epoch": 10.0032,
      "grad_norm": 0.07383687049150467,
      "learning_rate": 1.324e-05,
      "loss": 0.014187808334827422,
      "step": 9340
    },
    {
      "epoch": 10.0042,
      "grad_norm": 0.05301659554243088,
      "learning_rate": 1.3039999999999999e-05,
      "loss": 0.014343148469924927,
      "step": 9350
    },
    {
      "epoch": 10.0052,
      "grad_norm": 0.07555723190307617,
      "learning_rate": 1.2839999999999999e-05,
      "loss": 0.014555636048316955,
      "step": 9360
    },
    {
      "epoch": 10.0062,
      "grad_norm": 0.07009563595056534,
      "learning_rate": 1.2640000000000003e-05,
      "loss": 0.015680891275405884,
      "step": 9370
    },
    {
      "epoch": 10.0072,
      "grad_norm": 0.07948929071426392,
      "learning_rate": 1.244e-05,
      "loss": 0.015406323969364167,
      "step": 9380
    },
    {
      "epoch": 10.0082,
      "grad_norm": 0.07286491245031357,
      "learning_rate": 1.224e-05,
      "loss": 0.014841750264167786,
      "step": 9390
    },
    {
      "epoch": 10.0092,
      "grad_norm": 0.0699690729379654,
      "learning_rate": 1.204e-05,
      "loss": 0.01526234596967697,
      "step": 9400
    },
    {
      "epoch": 10.0102,
      "grad_norm": 0.06607243418693542,
      "learning_rate": 1.1840000000000002e-05,
      "loss": 0.014856207370758056,
      "step": 9410
    },
    {
      "epoch": 10.0112,
      "grad_norm": 0.06737522780895233,
      "learning_rate": 1.164e-05,
      "loss": 0.014999723434448243,
      "step": 9420
    },
    {
      "epoch": 10.0122,
      "grad_norm": 0.06653918325901031,
      "learning_rate": 1.144e-05,
      "loss": 0.015131491422653198,
      "step": 9430
    },
    {
      "epoch": 10.0132,
      "grad_norm": 0.06990286707878113,
      "learning_rate": 1.124e-05,
      "loss": 0.014309829473495484,
      "step": 9440
    },
    {
      "epoch": 10.0142,
      "grad_norm": 0.07211795449256897,
      "learning_rate": 1.1040000000000001e-05,
      "loss": 0.01395113468170166,
      "step": 9450
    },
    {
      "epoch": 10.0152,
      "grad_norm": 0.09780146181583405,
      "learning_rate": 1.084e-05,
      "loss": 0.016415250301361085,
      "step": 9460
    },
    {
      "epoch": 10.0162,
      "grad_norm": 0.05811549723148346,
      "learning_rate": 1.064e-05,
      "loss": 0.01397673338651657,
      "step": 9470
    },
    {
      "epoch": 10.0172,
      "grad_norm": 0.06327290832996368,
      "learning_rate": 1.0440000000000002e-05,
      "loss": 0.014413160085678101,
      "step": 9480
    },
    {
      "epoch": 10.0182,
      "grad_norm": 0.05637902393937111,
      "learning_rate": 1.024e-05,
      "loss": 0.014348730444908142,
      "step": 9490
    },
    {
      "epoch": 10.0192,
      "grad_norm": 0.06641823798418045,
      "learning_rate": 1.004e-05,
      "loss": 0.014750830829143524,
      "step": 9500
    },
    {
      "epoch": 1.001,
      "grad_norm": 0.09777779877185822,
      "learning_rate": 9.84e-06,
      "loss": 0.025202512741088867,
      "step": 9510
    },
    {
      "epoch": 1.002,
      "grad_norm": 0.08488325029611588,
      "learning_rate": 9.640000000000001e-06,
      "loss": 0.017650887370109558,
      "step": 9520
    },
    {
      "epoch": 1.003,
      "grad_norm": 0.05732690170407295,
      "learning_rate": 9.44e-06,
      "loss": 0.01551215648651123,
      "step": 9530
    },
    {
      "epoch": 1.004,
      "grad_norm": 0.0768381655216217,
      "learning_rate": 9.24e-06,
      "loss": 0.015778852999210356,
      "step": 9540
    },
    {
      "epoch": 1.005,
      "grad_norm": 0.06292874366044998,
      "learning_rate": 9.04e-06,
      "loss": 0.015921856462955474,
      "step": 9550
    },
    {
      "epoch": 1.006,
      "grad_norm": 0.05529346317052841,
      "learning_rate": 8.840000000000002e-06,
      "loss": 0.016185162961483,
      "step": 9560
    },
    {
      "epoch": 1.007,
      "grad_norm": 0.07324839383363724,
      "learning_rate": 8.64e-06,
      "loss": 0.014256039261817932,
      "step": 9570
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.07430668920278549,
      "learning_rate": 8.44e-06,
      "loss": 0.01393248736858368,
      "step": 9580
    },
    {
      "epoch": 1.009,
      "grad_norm": 0.05947092920541763,
      "learning_rate": 8.24e-06,
      "loss": 0.014715026319026946,
      "step": 9590
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.05222109332680702,
      "learning_rate": 8.040000000000001e-06,
      "loss": 0.014007921516895293,
      "step": 9600
    },
    {
      "epoch": 1.011,
      "grad_norm": 0.06184246763586998,
      "learning_rate": 7.84e-06,
      "loss": 0.014500398933887482,
      "step": 9610
    },
    {
      "epoch": 1.012,
      "grad_norm": 0.06747237592935562,
      "learning_rate": 7.64e-06,
      "loss": 0.014911913871765136,
      "step": 9620
    },
    {
      "epoch": 1.013,
      "grad_norm": 0.07077904045581818,
      "learning_rate": 7.44e-06,
      "loss": 0.01411098539829254,
      "step": 9630
    },
    {
      "epoch": 1.014,
      "grad_norm": 0.07017918676137924,
      "learning_rate": 7.240000000000001e-06,
      "loss": 0.013570456206798554,
      "step": 9640
    },
    {
      "epoch": 1.015,
      "grad_norm": 0.08114173263311386,
      "learning_rate": 7.04e-06,
      "loss": 0.013726058602333068,
      "step": 9650
    },
    {
      "epoch": 1.016,
      "grad_norm": 0.07676446437835693,
      "learning_rate": 6.840000000000001e-06,
      "loss": 0.016030623018741606,
      "step": 9660
    },
    {
      "epoch": 1.017,
      "grad_norm": 0.05721115320920944,
      "learning_rate": 6.640000000000001e-06,
      "loss": 0.014686207473278045,
      "step": 9670
    },
    {
      "epoch": 1.018,
      "grad_norm": 0.08408799022436142,
      "learning_rate": 6.44e-06,
      "loss": 0.015205322206020356,
      "step": 9680
    },
    {
      "epoch": 1.019,
      "grad_norm": 0.08085671067237854,
      "learning_rate": 6.24e-06,
      "loss": 0.01706176847219467,
      "step": 9690
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.09435422718524933,
      "learning_rate": 6.040000000000001e-06,
      "loss": 0.017468377947807312,
      "step": 9700
    },
    {
      "epoch": 1.021,
      "grad_norm": 0.0752953365445137,
      "learning_rate": 5.84e-06,
      "loss": 0.0176930233836174,
      "step": 9710
    },
    {
      "epoch": 1.022,
      "grad_norm": 0.07494309544563293,
      "learning_rate": 5.64e-06,
      "loss": 0.016850067675113677,
      "step": 9720
    },
    {
      "epoch": 1.023,
      "grad_norm": 0.08355386555194855,
      "learning_rate": 5.44e-06,
      "loss": 0.015157189965248109,
      "step": 9730
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.08324045687913895,
      "learning_rate": 5.240000000000001e-06,
      "loss": 0.015080301463603974,
      "step": 9740
    },
    {
      "epoch": 1.025,
      "grad_norm": 0.07760655134916306,
      "learning_rate": 5.04e-06,
      "loss": 0.01745104193687439,
      "step": 9750
    },
    {
      "epoch": 1.026,
      "grad_norm": 0.08728311955928802,
      "learning_rate": 4.84e-06,
      "loss": 0.01614951491355896,
      "step": 9760
    },
    {
      "epoch": 1.027,
      "grad_norm": 0.08304473012685776,
      "learning_rate": 4.64e-06,
      "loss": 0.01723228096961975,
      "step": 9770
    },
    {
      "epoch": 1.028,
      "grad_norm": 0.0646970123052597,
      "learning_rate": 4.440000000000001e-06,
      "loss": 0.01564721018075943,
      "step": 9780
    },
    {
      "epoch": 1.029,
      "grad_norm": 0.11390984058380127,
      "learning_rate": 4.24e-06,
      "loss": 0.01521741896867752,
      "step": 9790
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.08274708688259125,
      "learning_rate": 4.04e-06,
      "loss": 0.015826791524887085,
      "step": 9800
    },
    {
      "epoch": 1.031,
      "grad_norm": 0.07968990504741669,
      "learning_rate": 3.84e-06,
      "loss": 0.016426579654216768,
      "step": 9810
    },
    {
      "epoch": 2.0008,
      "grad_norm": 0.06300608068704605,
      "learning_rate": 3.6400000000000003e-06,
      "loss": 0.015782804787158967,
      "step": 9820
    },
    {
      "epoch": 2.0018,
      "grad_norm": 0.08467205613851547,
      "learning_rate": 3.44e-06,
      "loss": 0.014097627997398377,
      "step": 9830
    },
    {
      "epoch": 2.0028,
      "grad_norm": 0.09556452929973602,
      "learning_rate": 3.24e-06,
      "loss": 0.01549130529165268,
      "step": 9840
    },
    {
      "epoch": 2.0038,
      "grad_norm": 0.07436936348676682,
      "learning_rate": 3.04e-06,
      "loss": 0.015328900516033172,
      "step": 9850
    },
    {
      "epoch": 2.0048,
      "grad_norm": 0.07495304942131042,
      "learning_rate": 2.8400000000000003e-06,
      "loss": 0.015208512544631958,
      "step": 9860
    },
    {
      "epoch": 2.0058,
      "grad_norm": 0.10458475351333618,
      "learning_rate": 2.64e-06,
      "loss": 0.015952044725418092,
      "step": 9870
    },
    {
      "epoch": 2.0068,
      "grad_norm": 0.13968482613563538,
      "learning_rate": 2.4400000000000004e-06,
      "loss": 0.015596236288547515,
      "step": 9880
    },
    {
      "epoch": 2.0078,
      "grad_norm": 0.09523583203554153,
      "learning_rate": 2.24e-06,
      "loss": 0.014299987256526947,
      "step": 9890
    },
    {
      "epoch": 2.0088,
      "grad_norm": 0.08944617211818695,
      "learning_rate": 2.0400000000000004e-06,
      "loss": 0.014749652147293091,
      "step": 9900
    },
    {
      "epoch": 2.0098,
      "grad_norm": 0.27204543352127075,
      "learning_rate": 1.84e-06,
      "loss": 0.013910163938999177,
      "step": 9910
    },
    {
      "epoch": 2.0108,
      "grad_norm": 0.3379315137863159,
      "learning_rate": 1.6400000000000002e-06,
      "loss": 0.015016606450080872,
      "step": 9920
    },
    {
      "epoch": 2.0118,
      "grad_norm": 0.10731206089258194,
      "learning_rate": 1.44e-06,
      "loss": 0.016171526908874512,
      "step": 9930
    },
    {
      "epoch": 2.0128,
      "grad_norm": 0.15397508442401886,
      "learning_rate": 1.24e-06,
      "loss": 0.01619943231344223,
      "step": 9940
    },
    {
      "epoch": 2.0138,
      "grad_norm": 0.07552789896726608,
      "learning_rate": 1.04e-06,
      "loss": 0.015332949161529542,
      "step": 9950
    },
    {
      "epoch": 2.0148,
      "grad_norm": 0.13802768290042877,
      "learning_rate": 8.4e-07,
      "loss": 0.0159247025847435,
      "step": 9960
    },
    {
      "epoch": 2.0158,
      "grad_norm": 0.18125946819782257,
      "learning_rate": 6.4e-07,
      "loss": 0.01946225166320801,
      "step": 9970
    },
    {
      "epoch": 2.0168,
      "grad_norm": 0.1346278339624405,
      "learning_rate": 4.4e-07,
      "loss": 0.019359566271305084,
      "step": 9980
    },
    {
      "epoch": 2.0178,
      "grad_norm": 0.11318124085664749,
      "learning_rate": 2.4e-07,
      "loss": 0.022229455411434174,
      "step": 9990
    },
    {
      "epoch": 2.0188,
      "grad_norm": 0.13011205196380615,
      "learning_rate": 4e-08,
      "loss": 0.025185179710388184,
      "step": 10000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.465117552687677e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}