{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0010548523206751054,
      "grad_norm": 1.2911320924758911,
      "learning_rate": 0.00015822784810126583,
      "loss": 7.5306,
      "step": 10
    },
    {
      "epoch": 0.002109704641350211,
      "grad_norm": 1.1738190650939941,
      "learning_rate": 0.00031645569620253165,
      "loss": 6.9264,
      "step": 20
    },
    {
      "epoch": 0.0031645569620253164,
      "grad_norm": 0.8675833940505981,
      "learning_rate": 0.00047468354430379745,
      "loss": 6.2658,
      "step": 30
    },
    {
      "epoch": 0.004219409282700422,
      "grad_norm": 1.2689286470413208,
      "learning_rate": 0.0006329113924050633,
      "loss": 5.7667,
      "step": 40
    },
    {
      "epoch": 0.005274261603375527,
      "grad_norm": 0.8328885436058044,
      "learning_rate": 0.0007911392405063291,
      "loss": 5.3006,
      "step": 50
    },
    {
      "epoch": 0.006329113924050633,
      "grad_norm": 0.6221523284912109,
      "learning_rate": 0.0009493670886075949,
      "loss": 4.7854,
      "step": 60
    },
    {
      "epoch": 0.007383966244725738,
      "grad_norm": 0.7859921455383301,
      "learning_rate": 0.0011075949367088608,
      "loss": 4.3737,
      "step": 70
    },
    {
      "epoch": 0.008438818565400843,
      "grad_norm": 1.4790822267532349,
      "learning_rate": 0.0012658227848101266,
      "loss": 4.1257,
      "step": 80
    },
    {
      "epoch": 0.00949367088607595,
      "grad_norm": 0.84088134765625,
      "learning_rate": 0.0014240506329113926,
      "loss": 3.927,
      "step": 90
    },
    {
      "epoch": 0.010548523206751054,
      "grad_norm": 0.9328136444091797,
      "learning_rate": 0.0015,
      "loss": 3.7791,
      "step": 100
    },
    {
      "epoch": 0.011603375527426161,
      "grad_norm": 0.8343794941902161,
      "learning_rate": 0.0015,
      "loss": 3.6241,
      "step": 110
    },
    {
      "epoch": 0.012658227848101266,
      "grad_norm": 1.082481026649475,
      "learning_rate": 0.0015,
      "loss": 3.513,
      "step": 120
    },
    {
      "epoch": 0.013713080168776372,
      "grad_norm": 0.7395880222320557,
      "learning_rate": 0.0015,
      "loss": 3.4026,
      "step": 130
    },
    {
      "epoch": 0.014767932489451477,
      "grad_norm": 0.8911088109016418,
      "learning_rate": 0.0015,
      "loss": 3.3082,
      "step": 140
    },
    {
      "epoch": 0.015822784810126583,
      "grad_norm": 1.0754770040512085,
      "learning_rate": 0.0015,
      "loss": 3.2323,
      "step": 150
    },
    {
      "epoch": 0.016877637130801686,
      "grad_norm": 0.7520929574966431,
      "learning_rate": 0.0015,
      "loss": 3.1652,
      "step": 160
    },
    {
      "epoch": 0.017932489451476793,
      "grad_norm": 1.0018035173416138,
      "learning_rate": 0.0015,
      "loss": 3.0882,
      "step": 170
    },
    {
      "epoch": 0.0189873417721519,
      "grad_norm": 0.8712095618247986,
      "learning_rate": 0.0015,
      "loss": 3.0627,
      "step": 180
    },
    {
      "epoch": 0.020042194092827006,
      "grad_norm": 0.6786903142929077,
      "learning_rate": 0.0015,
      "loss": 2.989,
      "step": 190
    },
    {
      "epoch": 0.02109704641350211,
      "grad_norm": 0.887618899345398,
      "learning_rate": 0.0015,
      "loss": 2.9474,
      "step": 200
    },
    {
      "epoch": 0.022151898734177215,
      "grad_norm": 0.6775170564651489,
      "learning_rate": 0.0015,
      "loss": 2.9038,
      "step": 210
    },
    {
      "epoch": 0.023206751054852322,
      "grad_norm": 0.7518839836120605,
      "learning_rate": 0.0015,
      "loss": 2.8697,
      "step": 220
    },
    {
      "epoch": 0.024261603375527425,
      "grad_norm": 0.7377467155456543,
      "learning_rate": 0.0015,
      "loss": 2.8184,
      "step": 230
    },
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.9767955541610718,
      "learning_rate": 0.0015,
      "loss": 2.7812,
      "step": 240
    },
    {
      "epoch": 0.026371308016877638,
      "grad_norm": 0.8166293501853943,
      "learning_rate": 0.0015,
      "loss": 2.7466,
      "step": 250
    },
    {
      "epoch": 0.027426160337552744,
      "grad_norm": 0.7962077260017395,
      "learning_rate": 0.0015,
      "loss": 2.7132,
      "step": 260
    },
    {
      "epoch": 0.028481012658227847,
      "grad_norm": 0.8260292410850525,
      "learning_rate": 0.0015,
      "loss": 2.6831,
      "step": 270
    },
    {
      "epoch": 0.029535864978902954,
      "grad_norm": 0.9678518176078796,
      "learning_rate": 0.0015,
      "loss": 2.6457,
      "step": 280
    },
    {
      "epoch": 0.03059071729957806,
      "grad_norm": 0.8574512004852295,
      "learning_rate": 0.0015,
      "loss": 2.6198,
      "step": 290
    },
    {
      "epoch": 0.03164556962025317,
      "grad_norm": 1.0233553647994995,
      "learning_rate": 0.0015,
      "loss": 2.5961,
      "step": 300
    },
    {
      "epoch": 0.03270042194092827,
      "grad_norm": 0.8460607528686523,
      "learning_rate": 0.0015,
      "loss": 2.574,
      "step": 310
    },
    {
      "epoch": 0.03375527426160337,
      "grad_norm": 0.9686404466629028,
      "learning_rate": 0.0015,
      "loss": 2.5576,
      "step": 320
    },
    {
      "epoch": 0.03481012658227848,
      "grad_norm": 0.731982946395874,
      "learning_rate": 0.0015,
      "loss": 2.5384,
      "step": 330
    },
    {
      "epoch": 0.035864978902953586,
      "grad_norm": 1.0275819301605225,
      "learning_rate": 0.0015,
      "loss": 2.5105,
      "step": 340
    },
    {
      "epoch": 0.03691983122362869,
      "grad_norm": 0.9130907654762268,
      "learning_rate": 0.0015,
      "loss": 2.4707,
      "step": 350
    },
    {
      "epoch": 0.0379746835443038,
      "grad_norm": 0.802019476890564,
      "learning_rate": 0.0015,
      "loss": 2.4571,
      "step": 360
    },
    {
      "epoch": 0.039029535864978905,
      "grad_norm": 1.0577914714813232,
      "learning_rate": 0.0015,
      "loss": 2.4457,
      "step": 370
    },
    {
      "epoch": 0.04008438818565401,
      "grad_norm": 0.8877514004707336,
      "learning_rate": 0.0015,
      "loss": 2.4274,
      "step": 380
    },
    {
      "epoch": 0.04113924050632911,
      "grad_norm": 1.0621153116226196,
      "learning_rate": 0.0015,
      "loss": 2.4058,
      "step": 390
    },
    {
      "epoch": 0.04219409282700422,
      "grad_norm": 0.8034183382987976,
      "learning_rate": 0.0015,
      "loss": 2.3812,
      "step": 400
    },
    {
      "epoch": 0.043248945147679324,
      "grad_norm": 1.1909996271133423,
      "learning_rate": 0.0015,
      "loss": 2.3706,
      "step": 410
    },
    {
      "epoch": 0.04430379746835443,
      "grad_norm": 0.8821457028388977,
      "learning_rate": 0.0015,
      "loss": 2.3519,
      "step": 420
    },
    {
      "epoch": 0.04535864978902954,
      "grad_norm": 0.877312183380127,
      "learning_rate": 0.0015,
      "loss": 2.3309,
      "step": 430
    },
    {
      "epoch": 0.046413502109704644,
      "grad_norm": 0.7695094347000122,
      "learning_rate": 0.0015,
      "loss": 2.3125,
      "step": 440
    },
    {
      "epoch": 0.04746835443037975,
      "grad_norm": 0.9509552717208862,
      "learning_rate": 0.0015,
      "loss": 2.3128,
      "step": 450
    },
    {
      "epoch": 0.04852320675105485,
      "grad_norm": 0.7557908296585083,
      "learning_rate": 0.0015,
      "loss": 2.293,
      "step": 460
    },
    {
      "epoch": 0.049578059071729956,
      "grad_norm": 0.9517701864242554,
      "learning_rate": 0.0015,
      "loss": 2.2694,
      "step": 470
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.8353269696235657,
      "learning_rate": 0.0015,
      "loss": 2.2722,
      "step": 480
    },
    {
      "epoch": 0.05168776371308017,
      "grad_norm": 0.8538005948066711,
      "learning_rate": 0.0015,
      "loss": 2.2391,
      "step": 490
    },
    {
      "epoch": 0.052742616033755275,
      "grad_norm": 0.7373113632202148,
      "learning_rate": 0.0015,
      "loss": 2.2339,
      "step": 500
    },
    {
      "epoch": 0.05379746835443038,
      "grad_norm": 1.0332047939300537,
      "learning_rate": 0.0015,
      "loss": 2.2247,
      "step": 510
    },
    {
      "epoch": 0.05485232067510549,
      "grad_norm": 0.8454136848449707,
      "learning_rate": 0.0015,
      "loss": 2.2092,
      "step": 520
    },
    {
      "epoch": 0.05590717299578059,
      "grad_norm": 0.7165494561195374,
      "learning_rate": 0.0015,
      "loss": 2.1931,
      "step": 530
    },
    {
      "epoch": 0.056962025316455694,
      "grad_norm": 0.7286007404327393,
      "learning_rate": 0.0015,
      "loss": 2.1751,
      "step": 540
    },
    {
      "epoch": 0.0580168776371308,
      "grad_norm": 0.7424426674842834,
      "learning_rate": 0.0015,
      "loss": 2.1783,
      "step": 550
    },
    {
      "epoch": 0.05907172995780591,
      "grad_norm": 0.7351111173629761,
      "learning_rate": 0.0015,
      "loss": 2.146,
      "step": 560
    },
    {
      "epoch": 0.060126582278481014,
      "grad_norm": 1.0205843448638916,
      "learning_rate": 0.0015,
      "loss": 2.1503,
      "step": 570
    },
    {
      "epoch": 0.06118143459915612,
      "grad_norm": 1.1331634521484375,
      "learning_rate": 0.0015,
      "loss": 2.1545,
      "step": 580
    },
    {
      "epoch": 0.06223628691983123,
      "grad_norm": 1.3081284761428833,
      "learning_rate": 0.0015,
      "loss": 2.1313,
      "step": 590
    },
    {
      "epoch": 0.06329113924050633,
      "grad_norm": 1.0923019647598267,
      "learning_rate": 0.0015,
      "loss": 2.1306,
      "step": 600
    },
    {
      "epoch": 0.06434599156118144,
      "grad_norm": 1.1899714469909668,
      "learning_rate": 0.0015,
      "loss": 2.1104,
      "step": 610
    },
    {
      "epoch": 0.06540084388185655,
      "grad_norm": 0.8574506044387817,
      "learning_rate": 0.0015,
      "loss": 2.0926,
      "step": 620
    },
    {
      "epoch": 0.06645569620253164,
      "grad_norm": 0.9749586582183838,
      "learning_rate": 0.0015,
      "loss": 2.0799,
      "step": 630
    },
    {
      "epoch": 0.06751054852320675,
      "grad_norm": 0.7181791663169861,
      "learning_rate": 0.0015,
      "loss": 2.0941,
      "step": 640
    },
    {
      "epoch": 0.06856540084388185,
      "grad_norm": 0.736224353313446,
      "learning_rate": 0.0015,
      "loss": 2.0857,
      "step": 650
    },
    {
      "epoch": 0.06962025316455696,
      "grad_norm": 0.9573279023170471,
      "learning_rate": 0.0015,
      "loss": 2.0748,
      "step": 660
    },
    {
      "epoch": 0.07067510548523206,
      "grad_norm": 0.7422716021537781,
      "learning_rate": 0.0015,
      "loss": 2.0621,
      "step": 670
    },
    {
      "epoch": 0.07172995780590717,
      "grad_norm": 0.7074818015098572,
      "learning_rate": 0.0015,
      "loss": 2.0523,
      "step": 680
    },
    {
      "epoch": 0.07278481012658228,
      "grad_norm": 0.7281100749969482,
      "learning_rate": 0.0015,
      "loss": 2.0635,
      "step": 690
    },
    {
      "epoch": 0.07383966244725738,
      "grad_norm": 0.7090765833854675,
      "learning_rate": 0.0015,
      "loss": 2.0448,
      "step": 700
    },
    {
      "epoch": 0.07489451476793249,
      "grad_norm": 0.727451741695404,
      "learning_rate": 0.0015,
      "loss": 2.0325,
      "step": 710
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.7485256195068359,
      "learning_rate": 0.0015,
      "loss": 2.0135,
      "step": 720
    },
    {
      "epoch": 0.0770042194092827,
      "grad_norm": 0.8138155341148376,
      "learning_rate": 0.0015,
      "loss": 2.0065,
      "step": 730
    },
    {
      "epoch": 0.07805907172995781,
      "grad_norm": 0.6835618615150452,
      "learning_rate": 0.0015,
      "loss": 2.014,
      "step": 740
    },
    {
      "epoch": 0.07911392405063292,
      "grad_norm": 1.0360156297683716,
      "learning_rate": 0.0015,
      "loss": 1.9957,
      "step": 750
    },
    {
      "epoch": 0.08016877637130802,
      "grad_norm": 1.390318751335144,
      "learning_rate": 0.0015,
      "loss": 2.0014,
      "step": 760
    },
    {
      "epoch": 0.08122362869198312,
      "grad_norm": 0.989683985710144,
      "learning_rate": 0.0015,
      "loss": 2.0001,
      "step": 770
    },
    {
      "epoch": 0.08227848101265822,
      "grad_norm": 0.9884315729141235,
      "learning_rate": 0.0015,
      "loss": 1.9788,
      "step": 780
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.6850045919418335,
      "learning_rate": 0.0015,
      "loss": 1.9818,
      "step": 790
    },
    {
      "epoch": 0.08438818565400844,
      "grad_norm": 0.7124843597412109,
      "learning_rate": 0.0015,
      "loss": 1.9591,
      "step": 800
    },
    {
      "epoch": 0.08544303797468354,
      "grad_norm": 0.7981064319610596,
      "learning_rate": 0.0015,
      "loss": 1.9474,
      "step": 810
    },
    {
      "epoch": 0.08649789029535865,
      "grad_norm": 0.8757826685905457,
      "learning_rate": 0.0015,
      "loss": 1.9577,
      "step": 820
    },
    {
      "epoch": 0.08755274261603375,
      "grad_norm": 0.9261611700057983,
      "learning_rate": 0.0015,
      "loss": 1.9614,
      "step": 830
    },
    {
      "epoch": 0.08860759493670886,
      "grad_norm": 0.7376549243927002,
      "learning_rate": 0.0015,
      "loss": 1.9399,
      "step": 840
    },
    {
      "epoch": 0.08966244725738397,
      "grad_norm": 0.6958174109458923,
      "learning_rate": 0.0015,
      "loss": 1.9405,
      "step": 850
    },
    {
      "epoch": 0.09071729957805907,
      "grad_norm": 0.7377433180809021,
      "learning_rate": 0.0015,
      "loss": 1.9374,
      "step": 860
    },
    {
      "epoch": 0.09177215189873418,
      "grad_norm": 0.659022867679596,
      "learning_rate": 0.0015,
      "loss": 1.9339,
      "step": 870
    },
    {
      "epoch": 0.09282700421940929,
      "grad_norm": 0.7705568075180054,
      "learning_rate": 0.0015,
      "loss": 1.9272,
      "step": 880
    },
    {
      "epoch": 0.0938818565400844,
      "grad_norm": 1.073669672012329,
      "learning_rate": 0.0015,
      "loss": 1.9159,
      "step": 890
    },
    {
      "epoch": 0.0949367088607595,
      "grad_norm": 1.3000624179840088,
      "learning_rate": 0.0015,
      "loss": 1.9168,
      "step": 900
    },
    {
      "epoch": 0.09599156118143459,
      "grad_norm": 0.817836344242096,
      "learning_rate": 0.0015,
      "loss": 1.9206,
      "step": 910
    },
    {
      "epoch": 0.0970464135021097,
      "grad_norm": 0.6906339526176453,
      "learning_rate": 0.0015,
      "loss": 1.9065,
      "step": 920
    },
    {
      "epoch": 0.0981012658227848,
      "grad_norm": 0.7800134420394897,
      "learning_rate": 0.0015,
      "loss": 1.8938,
      "step": 930
    },
    {
      "epoch": 0.09915611814345991,
      "grad_norm": 0.7053550481796265,
      "learning_rate": 0.0015,
      "loss": 1.8986,
      "step": 940
    },
    {
      "epoch": 0.10021097046413502,
      "grad_norm": 1.0190623998641968,
      "learning_rate": 0.0015,
      "loss": 1.8828,
      "step": 950
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.7127584218978882,
      "learning_rate": 0.0015,
      "loss": 1.8896,
      "step": 960
    },
    {
      "epoch": 0.10232067510548523,
      "grad_norm": 1.0691441297531128,
      "learning_rate": 0.0015,
      "loss": 1.8884,
      "step": 970
    },
    {
      "epoch": 0.10337552742616034,
      "grad_norm": 0.6816598773002625,
      "learning_rate": 0.0015,
      "loss": 1.8671,
      "step": 980
    },
    {
      "epoch": 0.10443037974683544,
      "grad_norm": 0.7380853295326233,
      "learning_rate": 0.0015,
      "loss": 1.867,
      "step": 990
    },
    {
      "epoch": 0.10548523206751055,
      "grad_norm": 0.8465728759765625,
      "learning_rate": 0.0015,
      "loss": 1.8785,
      "step": 1000
    },
    {
      "epoch": 0.10654008438818566,
      "grad_norm": 0.7769269943237305,
      "learning_rate": 0.0015,
      "loss": 1.8691,
      "step": 1010
    },
    {
      "epoch": 0.10759493670886076,
      "grad_norm": 1.0904594659805298,
      "learning_rate": 0.0015,
      "loss": 1.8588,
      "step": 1020
    },
    {
      "epoch": 0.10864978902953587,
      "grad_norm": 1.2031594514846802,
      "learning_rate": 0.0015,
      "loss": 1.8605,
      "step": 1030
    },
    {
      "epoch": 0.10970464135021098,
      "grad_norm": 0.7433950901031494,
      "learning_rate": 0.0015,
      "loss": 1.8515,
      "step": 1040
    },
    {
      "epoch": 0.11075949367088607,
      "grad_norm": 0.7069742679595947,
      "learning_rate": 0.0015,
      "loss": 1.854,
      "step": 1050
    },
    {
      "epoch": 0.11181434599156118,
      "grad_norm": 0.6981137990951538,
      "learning_rate": 0.0015,
      "loss": 1.8432,
      "step": 1060
    },
    {
      "epoch": 0.11286919831223628,
      "grad_norm": 0.6770772933959961,
      "learning_rate": 0.0015,
      "loss": 1.8311,
      "step": 1070
    },
    {
      "epoch": 0.11392405063291139,
      "grad_norm": 0.6839622855186462,
      "learning_rate": 0.0015,
      "loss": 1.8326,
      "step": 1080
    },
    {
      "epoch": 0.1149789029535865,
      "grad_norm": 0.6609824895858765,
      "learning_rate": 0.0015,
      "loss": 1.8409,
      "step": 1090
    },
    {
      "epoch": 0.1160337552742616,
      "grad_norm": 0.7277442216873169,
      "learning_rate": 0.0015,
      "loss": 1.8276,
      "step": 1100
    },
    {
      "epoch": 0.11708860759493671,
      "grad_norm": 0.7287368774414062,
      "learning_rate": 0.0015,
      "loss": 1.8329,
      "step": 1110
    },
    {
      "epoch": 0.11814345991561181,
      "grad_norm": 0.7724989652633667,
      "learning_rate": 0.0015,
      "loss": 1.8279,
      "step": 1120
    },
    {
      "epoch": 0.11919831223628692,
      "grad_norm": 0.7109525203704834,
      "learning_rate": 0.0015,
      "loss": 1.8134,
      "step": 1130
    },
    {
      "epoch": 0.12025316455696203,
      "grad_norm": 0.7559357285499573,
      "learning_rate": 0.0015,
      "loss": 1.8262,
      "step": 1140
    },
    {
      "epoch": 0.12130801687763713,
      "grad_norm": 1.1389662027359009,
      "learning_rate": 0.0015,
      "loss": 1.8134,
      "step": 1150
    },
    {
      "epoch": 0.12236286919831224,
      "grad_norm": 0.9472622871398926,
      "learning_rate": 0.0015,
      "loss": 1.8176,
      "step": 1160
    },
    {
      "epoch": 0.12341772151898735,
      "grad_norm": 0.6383090019226074,
      "learning_rate": 0.0015,
      "loss": 1.8033,
      "step": 1170
    },
    {
      "epoch": 0.12447257383966245,
      "grad_norm": 0.8101405501365662,
      "learning_rate": 0.0015,
      "loss": 1.7904,
      "step": 1180
    },
    {
      "epoch": 0.12552742616033755,
      "grad_norm": 0.655144453048706,
      "learning_rate": 0.0015,
      "loss": 1.7934,
      "step": 1190
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.9537879824638367,
      "learning_rate": 0.0015,
      "loss": 1.808,
      "step": 1200
    },
    {
      "epoch": 0.12763713080168776,
      "grad_norm": 0.8779239654541016,
      "learning_rate": 0.0015,
      "loss": 1.7989,
      "step": 1210
    },
    {
      "epoch": 0.12869198312236288,
      "grad_norm": 0.7360990643501282,
      "learning_rate": 0.0015,
      "loss": 1.7854,
      "step": 1220
    },
    {
      "epoch": 0.12974683544303797,
      "grad_norm": 0.6458498239517212,
      "learning_rate": 0.0015,
      "loss": 1.7767,
      "step": 1230
    },
    {
      "epoch": 0.1308016877637131,
      "grad_norm": 0.7693349123001099,
      "learning_rate": 0.0015,
      "loss": 1.7905,
      "step": 1240
    },
    {
      "epoch": 0.13185654008438819,
      "grad_norm": 0.6853756308555603,
      "learning_rate": 0.0015,
      "loss": 1.7936,
      "step": 1250
    },
    {
      "epoch": 0.13291139240506328,
      "grad_norm": 0.9074394106864929,
      "learning_rate": 0.0015,
      "loss": 1.7796,
      "step": 1260
    },
    {
      "epoch": 0.1339662447257384,
      "grad_norm": 0.6336771249771118,
      "learning_rate": 0.0015,
      "loss": 1.777,
      "step": 1270
    },
    {
      "epoch": 0.1350210970464135,
      "grad_norm": 0.785158097743988,
      "learning_rate": 0.0015,
      "loss": 1.7892,
      "step": 1280
    },
    {
      "epoch": 0.1360759493670886,
      "grad_norm": 0.9094076156616211,
      "learning_rate": 0.0015,
      "loss": 1.7746,
      "step": 1290
    },
    {
      "epoch": 0.1371308016877637,
      "grad_norm": 0.7252681851387024,
      "learning_rate": 0.0015,
      "loss": 1.761,
      "step": 1300
    },
    {
      "epoch": 0.13818565400843882,
      "grad_norm": 0.7606396079063416,
      "learning_rate": 0.0015,
      "loss": 1.7692,
      "step": 1310
    },
    {
      "epoch": 0.13924050632911392,
      "grad_norm": 0.7315070629119873,
      "learning_rate": 0.0015,
      "loss": 1.7717,
      "step": 1320
    },
    {
      "epoch": 0.14029535864978904,
      "grad_norm": 1.215480089187622,
      "learning_rate": 0.0015,
      "loss": 1.7627,
      "step": 1330
    },
    {
      "epoch": 0.14135021097046413,
      "grad_norm": 0.7350618839263916,
      "learning_rate": 0.0015,
      "loss": 1.7627,
      "step": 1340
    },
    {
      "epoch": 0.14240506329113925,
      "grad_norm": 0.7639700174331665,
      "learning_rate": 0.0015,
      "loss": 1.767,
      "step": 1350
    },
    {
      "epoch": 0.14345991561181434,
      "grad_norm": 0.6382935643196106,
      "learning_rate": 0.0015,
      "loss": 1.7592,
      "step": 1360
    },
    {
      "epoch": 0.14451476793248946,
      "grad_norm": 0.6957666873931885,
      "learning_rate": 0.0015,
      "loss": 1.7538,
      "step": 1370
    },
    {
      "epoch": 0.14556962025316456,
      "grad_norm": 0.7660850882530212,
      "learning_rate": 0.0015,
      "loss": 1.7445,
      "step": 1380
    },
    {
      "epoch": 0.14662447257383968,
      "grad_norm": 1.0359055995941162,
      "learning_rate": 0.0015,
      "loss": 1.7563,
      "step": 1390
    },
    {
      "epoch": 0.14767932489451477,
      "grad_norm": 1.3672610521316528,
      "learning_rate": 0.0015,
      "loss": 1.7391,
      "step": 1400
    },
    {
      "epoch": 0.14873417721518986,
      "grad_norm": 1.2046788930892944,
      "learning_rate": 0.0015,
      "loss": 1.7483,
      "step": 1410
    },
    {
      "epoch": 0.14978902953586498,
      "grad_norm": 0.7581397891044617,
      "learning_rate": 0.0015,
      "loss": 1.7504,
      "step": 1420
    },
    {
      "epoch": 0.15084388185654007,
      "grad_norm": 0.7376483678817749,
      "learning_rate": 0.0015,
      "loss": 1.7375,
      "step": 1430
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.661811888217926,
      "learning_rate": 0.0015,
      "loss": 1.7304,
      "step": 1440
    },
    {
      "epoch": 0.1529535864978903,
      "grad_norm": 0.6791737675666809,
      "learning_rate": 0.0015,
      "loss": 1.7312,
      "step": 1450
    },
    {
      "epoch": 0.1540084388185654,
      "grad_norm": 0.6737160682678223,
      "learning_rate": 0.0015,
      "loss": 1.7297,
      "step": 1460
    },
    {
      "epoch": 0.1550632911392405,
      "grad_norm": 0.8996027708053589,
      "learning_rate": 0.0015,
      "loss": 1.7299,
      "step": 1470
    },
    {
      "epoch": 0.15611814345991562,
      "grad_norm": 0.9724168181419373,
      "learning_rate": 0.0015,
      "loss": 1.7306,
      "step": 1480
    },
    {
      "epoch": 0.1571729957805907,
      "grad_norm": 1.112387776374817,
      "learning_rate": 0.0015,
      "loss": 1.7305,
      "step": 1490
    },
    {
      "epoch": 0.15822784810126583,
      "grad_norm": 0.6346387267112732,
      "learning_rate": 0.0015,
      "loss": 1.7287,
      "step": 1500
    },
    {
      "epoch": 0.15928270042194093,
      "grad_norm": 0.6662168502807617,
      "learning_rate": 0.0015,
      "loss": 1.723,
      "step": 1510
    },
    {
      "epoch": 0.16033755274261605,
      "grad_norm": 0.8017670512199402,
      "learning_rate": 0.0015,
      "loss": 1.7262,
      "step": 1520
    },
    {
      "epoch": 0.16139240506329114,
      "grad_norm": 0.805647611618042,
      "learning_rate": 0.0015,
      "loss": 1.7106,
      "step": 1530
    },
    {
      "epoch": 0.16244725738396623,
      "grad_norm": 0.6892802119255066,
      "learning_rate": 0.0015,
      "loss": 1.7026,
      "step": 1540
    },
    {
      "epoch": 0.16350210970464135,
      "grad_norm": 1.0336511135101318,
      "learning_rate": 0.0015,
      "loss": 1.7047,
      "step": 1550
    },
    {
      "epoch": 0.16455696202531644,
      "grad_norm": 0.6698900461196899,
      "learning_rate": 0.0015,
      "loss": 1.7082,
      "step": 1560
    },
    {
      "epoch": 0.16561181434599156,
      "grad_norm": 0.7654104828834534,
      "learning_rate": 0.0015,
      "loss": 1.6972,
      "step": 1570
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.6756095290184021,
      "learning_rate": 0.0015,
      "loss": 1.7089,
      "step": 1580
    },
    {
      "epoch": 0.16772151898734178,
      "grad_norm": 0.9509694576263428,
      "learning_rate": 0.0015,
      "loss": 1.7045,
      "step": 1590
    },
    {
      "epoch": 0.16877637130801687,
      "grad_norm": 0.712745726108551,
      "learning_rate": 0.0015,
      "loss": 1.7096,
      "step": 1600
    },
    {
      "epoch": 0.169831223628692,
      "grad_norm": 0.655924379825592,
      "learning_rate": 0.0015,
      "loss": 1.7065,
      "step": 1610
    },
    {
      "epoch": 0.17088607594936708,
      "grad_norm": 0.6995239853858948,
      "learning_rate": 0.0015,
      "loss": 1.6769,
      "step": 1620
    },
    {
      "epoch": 0.1719409282700422,
      "grad_norm": 0.8030900955200195,
      "learning_rate": 0.0015,
      "loss": 1.6908,
      "step": 1630
    },
    {
      "epoch": 0.1729957805907173,
      "grad_norm": 0.6829028129577637,
      "learning_rate": 0.0015,
      "loss": 1.6939,
      "step": 1640
    },
    {
      "epoch": 0.17405063291139242,
      "grad_norm": 1.173487901687622,
      "learning_rate": 0.0015,
      "loss": 1.6842,
      "step": 1650
    },
    {
      "epoch": 0.1751054852320675,
      "grad_norm": 0.6973627805709839,
      "learning_rate": 0.0015,
      "loss": 1.6833,
      "step": 1660
    },
    {
      "epoch": 0.17616033755274263,
      "grad_norm": 0.833595871925354,
      "learning_rate": 0.0015,
      "loss": 1.6944,
      "step": 1670
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 1.1782370805740356,
      "learning_rate": 0.0015,
      "loss": 1.6824,
      "step": 1680
    },
    {
      "epoch": 0.17827004219409281,
      "grad_norm": 0.7128448486328125,
      "learning_rate": 0.0015,
      "loss": 1.6761,
      "step": 1690
    },
    {
      "epoch": 0.17932489451476794,
      "grad_norm": 0.8215356469154358,
      "learning_rate": 0.0015,
      "loss": 1.6916,
      "step": 1700
    },
    {
      "epoch": 0.18037974683544303,
      "grad_norm": 0.6329225301742554,
      "learning_rate": 0.0015,
      "loss": 1.6724,
      "step": 1710
    },
    {
      "epoch": 0.18143459915611815,
      "grad_norm": 0.6323493123054504,
      "learning_rate": 0.0015,
      "loss": 1.6685,
      "step": 1720
    },
    {
      "epoch": 0.18248945147679324,
      "grad_norm": 0.6220436692237854,
      "learning_rate": 0.0015,
      "loss": 1.6703,
      "step": 1730
    },
    {
      "epoch": 0.18354430379746836,
      "grad_norm": 0.6482318043708801,
      "learning_rate": 0.0015,
      "loss": 1.6786,
      "step": 1740
    },
    {
      "epoch": 0.18459915611814345,
      "grad_norm": 0.6890340447425842,
      "learning_rate": 0.0015,
      "loss": 1.6931,
      "step": 1750
    },
    {
      "epoch": 0.18565400843881857,
      "grad_norm": 0.6083322167396545,
      "learning_rate": 0.0015,
      "loss": 1.6757,
      "step": 1760
    },
    {
      "epoch": 0.18670886075949367,
      "grad_norm": 0.6560311317443848,
      "learning_rate": 0.0015,
      "loss": 1.6763,
      "step": 1770
    },
    {
      "epoch": 0.1877637130801688,
      "grad_norm": 0.6504945158958435,
      "learning_rate": 0.0015,
      "loss": 1.677,
      "step": 1780
    },
    {
      "epoch": 0.18881856540084388,
      "grad_norm": 0.6659324169158936,
      "learning_rate": 0.0015,
      "loss": 1.6772,
      "step": 1790
    },
    {
      "epoch": 0.189873417721519,
      "grad_norm": 0.7314705848693848,
      "learning_rate": 0.0015,
      "loss": 1.6742,
      "step": 1800
    },
    {
      "epoch": 0.1909282700421941,
      "grad_norm": 0.7147253751754761,
      "learning_rate": 0.0015,
      "loss": 1.6619,
      "step": 1810
    },
    {
      "epoch": 0.19198312236286919,
      "grad_norm": 0.797636091709137,
      "learning_rate": 0.0015,
      "loss": 1.6628,
      "step": 1820
    },
    {
      "epoch": 0.1930379746835443,
      "grad_norm": 0.6524857878684998,
      "learning_rate": 0.0015,
      "loss": 1.658,
      "step": 1830
    },
    {
      "epoch": 0.1940928270042194,
      "grad_norm": 0.6256518363952637,
      "learning_rate": 0.0015,
      "loss": 1.6571,
      "step": 1840
    },
    {
      "epoch": 0.19514767932489452,
      "grad_norm": 0.8984813094139099,
      "learning_rate": 0.0015,
      "loss": 1.6586,
      "step": 1850
    },
    {
      "epoch": 0.1962025316455696,
      "grad_norm": 0.7350126504898071,
      "learning_rate": 0.0015,
      "loss": 1.6625,
      "step": 1860
    },
    {
      "epoch": 0.19725738396624473,
      "grad_norm": 0.7215333580970764,
      "learning_rate": 0.0015,
      "loss": 1.6628,
      "step": 1870
    },
    {
      "epoch": 0.19831223628691982,
      "grad_norm": 0.6555130481719971,
      "learning_rate": 0.0015,
      "loss": 1.6525,
      "step": 1880
    },
    {
      "epoch": 0.19936708860759494,
      "grad_norm": 0.7172640562057495,
      "learning_rate": 0.0015,
      "loss": 1.6563,
      "step": 1890
    },
    {
      "epoch": 0.20042194092827004,
      "grad_norm": 0.8645635843276978,
      "learning_rate": 0.0015,
      "loss": 1.6514,
      "step": 1900
    },
    {
      "epoch": 0.20147679324894516,
      "grad_norm": 0.8532048463821411,
      "learning_rate": 0.0015,
      "loss": 1.6634,
      "step": 1910
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.8084898591041565,
      "learning_rate": 0.0015,
      "loss": 1.6496,
      "step": 1920
    },
    {
      "epoch": 0.20358649789029537,
      "grad_norm": 0.6495212316513062,
      "learning_rate": 0.0015,
      "loss": 1.6468,
      "step": 1930
    },
    {
      "epoch": 0.20464135021097046,
      "grad_norm": 0.6420271396636963,
      "learning_rate": 0.0015,
      "loss": 1.641,
      "step": 1940
    },
    {
      "epoch": 0.20569620253164558,
      "grad_norm": 0.6756624579429626,
      "learning_rate": 0.0015,
      "loss": 1.6408,
      "step": 1950
    },
    {
      "epoch": 0.20675105485232068,
      "grad_norm": 0.632693350315094,
      "learning_rate": 0.0015,
      "loss": 1.6454,
      "step": 1960
    },
    {
      "epoch": 0.20780590717299577,
      "grad_norm": 0.7314813137054443,
      "learning_rate": 0.0015,
      "loss": 1.6523,
      "step": 1970
    },
    {
      "epoch": 0.2088607594936709,
      "grad_norm": 0.6397210955619812,
      "learning_rate": 0.0015,
      "loss": 1.6369,
      "step": 1980
    },
    {
      "epoch": 0.20991561181434598,
      "grad_norm": 0.6606935262680054,
      "learning_rate": 0.0015,
      "loss": 1.6346,
      "step": 1990
    },
    {
      "epoch": 0.2109704641350211,
      "grad_norm": 0.6740031242370605,
      "learning_rate": 0.0015,
      "loss": 1.6339,
      "step": 2000
    },
    {
      "epoch": 0.2120253164556962,
      "grad_norm": 0.720748245716095,
      "learning_rate": 0.0015,
      "loss": 1.6401,
      "step": 2010
    },
    {
      "epoch": 0.21308016877637131,
      "grad_norm": 1.1676092147827148,
      "learning_rate": 0.0015,
      "loss": 1.6498,
      "step": 2020
    },
    {
      "epoch": 0.2141350210970464,
      "grad_norm": 0.6337310075759888,
      "learning_rate": 0.0015,
      "loss": 1.6379,
      "step": 2030
    },
    {
      "epoch": 0.21518987341772153,
      "grad_norm": 0.8364405035972595,
      "learning_rate": 0.0015,
      "loss": 1.6366,
      "step": 2040
    },
    {
      "epoch": 0.21624472573839662,
      "grad_norm": 0.7344895005226135,
      "learning_rate": 0.0015,
      "loss": 1.6351,
      "step": 2050
    },
    {
      "epoch": 0.21729957805907174,
      "grad_norm": 0.9234893918037415,
      "learning_rate": 0.0015,
      "loss": 1.6269,
      "step": 2060
    },
    {
      "epoch": 0.21835443037974683,
      "grad_norm": 0.7972549796104431,
      "learning_rate": 0.0015,
      "loss": 1.6302,
      "step": 2070
    },
    {
      "epoch": 0.21940928270042195,
      "grad_norm": 0.6978487372398376,
      "learning_rate": 0.0015,
      "loss": 1.6337,
      "step": 2080
    },
    {
      "epoch": 0.22046413502109705,
      "grad_norm": 0.766801655292511,
      "learning_rate": 0.0015,
      "loss": 1.6191,
      "step": 2090
    },
    {
      "epoch": 0.22151898734177214,
      "grad_norm": 1.0701206922531128,
      "learning_rate": 0.0015,
      "loss": 1.6314,
      "step": 2100
    },
    {
      "epoch": 0.22257383966244726,
      "grad_norm": 0.6192570328712463,
      "learning_rate": 0.0015,
      "loss": 1.6205,
      "step": 2110
    },
    {
      "epoch": 0.22362869198312235,
      "grad_norm": 0.5801493525505066,
      "learning_rate": 0.0015,
      "loss": 1.6181,
      "step": 2120
    },
    {
      "epoch": 0.22468354430379747,
      "grad_norm": 0.6492977142333984,
      "learning_rate": 0.0015,
      "loss": 1.6231,
      "step": 2130
    },
    {
      "epoch": 0.22573839662447256,
      "grad_norm": 0.574461817741394,
      "learning_rate": 0.0015,
      "loss": 1.6173,
      "step": 2140
    },
    {
      "epoch": 0.22679324894514769,
      "grad_norm": 0.7691101431846619,
      "learning_rate": 0.0015,
      "loss": 1.6237,
      "step": 2150
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.9927956461906433,
      "learning_rate": 0.0015,
      "loss": 1.635,
      "step": 2160
    },
    {
      "epoch": 0.2289029535864979,
      "grad_norm": 0.6871766448020935,
      "learning_rate": 0.0015,
      "loss": 1.62,
      "step": 2170
    },
    {
      "epoch": 0.229957805907173,
      "grad_norm": 0.5886237025260925,
      "learning_rate": 0.0015,
      "loss": 1.6045,
      "step": 2180
    },
    {
      "epoch": 0.2310126582278481,
      "grad_norm": 0.6547583341598511,
      "learning_rate": 0.0015,
      "loss": 1.6228,
      "step": 2190
    },
    {
      "epoch": 0.2320675105485232,
      "grad_norm": 0.6600242853164673,
      "learning_rate": 0.0015,
      "loss": 1.6188,
      "step": 2200
    },
    {
      "epoch": 0.23312236286919832,
      "grad_norm": 0.797605037689209,
      "learning_rate": 0.0015,
      "loss": 1.6039,
      "step": 2210
    },
    {
      "epoch": 0.23417721518987342,
      "grad_norm": 0.6497875452041626,
      "learning_rate": 0.0015,
      "loss": 1.6185,
      "step": 2220
    },
    {
      "epoch": 0.23523206751054854,
      "grad_norm": 0.782981276512146,
      "learning_rate": 0.0015,
      "loss": 1.613,
      "step": 2230
    },
    {
      "epoch": 0.23628691983122363,
      "grad_norm": 0.7650336623191833,
      "learning_rate": 0.0015,
      "loss": 1.6109,
      "step": 2240
    },
    {
      "epoch": 0.23734177215189872,
      "grad_norm": 1.1431978940963745,
      "learning_rate": 0.0015,
      "loss": 1.6079,
      "step": 2250
    },
    {
      "epoch": 0.23839662447257384,
      "grad_norm": 0.9392552971839905,
      "learning_rate": 0.0015,
      "loss": 1.599,
      "step": 2260
    },
    {
      "epoch": 0.23945147679324894,
      "grad_norm": 0.7382718324661255,
      "learning_rate": 0.0015,
      "loss": 1.6302,
      "step": 2270
    },
    {
      "epoch": 0.24050632911392406,
      "grad_norm": 0.6001378893852234,
      "learning_rate": 0.0015,
      "loss": 1.617,
      "step": 2280
    },
    {
      "epoch": 0.24156118143459915,
      "grad_norm": 0.6126279830932617,
      "learning_rate": 0.0015,
      "loss": 1.5995,
      "step": 2290
    },
    {
      "epoch": 0.24261603375527427,
      "grad_norm": 0.6365258097648621,
      "learning_rate": 0.0015,
      "loss": 1.5984,
      "step": 2300
    },
    {
      "epoch": 0.24367088607594936,
      "grad_norm": 0.6386217474937439,
      "learning_rate": 0.0015,
      "loss": 1.5997,
      "step": 2310
    },
    {
      "epoch": 0.24472573839662448,
      "grad_norm": 0.6965967416763306,
      "learning_rate": 0.0015,
      "loss": 1.6123,
      "step": 2320
    },
    {
      "epoch": 0.24578059071729957,
      "grad_norm": 0.6952495574951172,
      "learning_rate": 0.0015,
      "loss": 1.5944,
      "step": 2330
    },
    {
      "epoch": 0.2468354430379747,
      "grad_norm": 0.8141445517539978,
      "learning_rate": 0.0015,
      "loss": 1.5891,
      "step": 2340
    },
    {
      "epoch": 0.2478902953586498,
      "grad_norm": 0.8598236441612244,
      "learning_rate": 0.0015,
      "loss": 1.5927,
      "step": 2350
    },
    {
      "epoch": 0.2489451476793249,
      "grad_norm": 0.9009906053543091,
      "learning_rate": 0.0015,
      "loss": 1.6005,
      "step": 2360
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.9204186201095581,
      "learning_rate": 0.0015,
      "loss": 1.599,
      "step": 2370
    },
    {
      "epoch": 0.2510548523206751,
      "grad_norm": 0.7122026085853577,
      "learning_rate": 0.0015,
      "loss": 1.597,
      "step": 2380
    },
    {
      "epoch": 0.2521097046413502,
      "grad_norm": 0.6895260214805603,
      "learning_rate": 0.0015,
      "loss": 1.5961,
      "step": 2390
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.5801357626914978,
      "learning_rate": 0.0015,
      "loss": 1.5932,
      "step": 2400
    },
    {
      "epoch": 0.2542194092827004,
      "grad_norm": 0.7699609398841858,
      "learning_rate": 0.0015,
      "loss": 1.5971,
      "step": 2410
    },
    {
      "epoch": 0.2552742616033755,
      "grad_norm": 0.7428116202354431,
      "learning_rate": 0.0015,
      "loss": 1.5941,
      "step": 2420
    },
    {
      "epoch": 0.2563291139240506,
      "grad_norm": 0.651231050491333,
      "learning_rate": 0.0015,
      "loss": 1.5936,
      "step": 2430
    },
    {
      "epoch": 0.25738396624472576,
      "grad_norm": 0.7375125885009766,
      "learning_rate": 0.0015,
      "loss": 1.5993,
      "step": 2440
    },
    {
      "epoch": 0.25843881856540085,
      "grad_norm": 0.635988712310791,
      "learning_rate": 0.0015,
      "loss": 1.5849,
      "step": 2450
    },
    {
      "epoch": 0.25949367088607594,
      "grad_norm": 0.601569414138794,
      "learning_rate": 0.0015,
      "loss": 1.5997,
      "step": 2460
    },
    {
      "epoch": 0.26054852320675104,
      "grad_norm": 0.8157471418380737,
      "learning_rate": 0.0015,
      "loss": 1.5897,
      "step": 2470
    },
    {
      "epoch": 0.2616033755274262,
      "grad_norm": 0.7209903597831726,
      "learning_rate": 0.0015,
      "loss": 1.5945,
      "step": 2480
    },
    {
      "epoch": 0.2626582278481013,
      "grad_norm": 0.6519895792007446,
      "learning_rate": 0.0015,
      "loss": 1.5806,
      "step": 2490
    },
    {
      "epoch": 0.26371308016877637,
      "grad_norm": 0.7864737510681152,
      "learning_rate": 0.0015,
      "loss": 1.5839,
      "step": 2500
    },
    {
      "epoch": 0.26476793248945146,
      "grad_norm": 0.6227979063987732,
      "learning_rate": 0.0015,
      "loss": 1.5865,
      "step": 2510
    },
    {
      "epoch": 0.26582278481012656,
      "grad_norm": 0.6954833269119263,
      "learning_rate": 0.0015,
      "loss": 1.5903,
      "step": 2520
    },
    {
      "epoch": 0.2668776371308017,
      "grad_norm": 0.640419065952301,
      "learning_rate": 0.0015,
      "loss": 1.5808,
      "step": 2530
    },
    {
      "epoch": 0.2679324894514768,
      "grad_norm": 1.0571954250335693,
      "learning_rate": 0.0015,
      "loss": 1.5849,
      "step": 2540
    },
    {
      "epoch": 0.2689873417721519,
      "grad_norm": 0.7716902494430542,
      "learning_rate": 0.0015,
      "loss": 1.5787,
      "step": 2550
    },
    {
      "epoch": 0.270042194092827,
      "grad_norm": 0.627837598323822,
      "learning_rate": 0.0015,
      "loss": 1.5874,
      "step": 2560
    },
    {
      "epoch": 0.27109704641350213,
      "grad_norm": 0.6312415599822998,
      "learning_rate": 0.0015,
      "loss": 1.5804,
      "step": 2570
    },
    {
      "epoch": 0.2721518987341772,
      "grad_norm": 0.5654999017715454,
      "learning_rate": 0.0015,
      "loss": 1.5793,
      "step": 2580
    },
    {
      "epoch": 0.2732067510548523,
      "grad_norm": 0.6043466329574585,
      "learning_rate": 0.0015,
      "loss": 1.5861,
      "step": 2590
    },
    {
      "epoch": 0.2742616033755274,
      "grad_norm": 0.6311189532279968,
      "learning_rate": 0.0015,
      "loss": 1.5872,
      "step": 2600
    },
    {
      "epoch": 0.27531645569620256,
      "grad_norm": 0.6845487952232361,
      "learning_rate": 0.0015,
      "loss": 1.5839,
      "step": 2610
    },
    {
      "epoch": 0.27637130801687765,
      "grad_norm": 0.6696321368217468,
      "learning_rate": 0.0015,
      "loss": 1.5808,
      "step": 2620
    },
    {
      "epoch": 0.27742616033755274,
      "grad_norm": 0.7458539009094238,
      "learning_rate": 0.0015,
      "loss": 1.5731,
      "step": 2630
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.6243004202842712,
      "learning_rate": 0.0015,
      "loss": 1.582,
      "step": 2640
    },
    {
      "epoch": 0.2795358649789029,
      "grad_norm": 0.7745714783668518,
      "learning_rate": 0.0015,
      "loss": 1.5747,
      "step": 2650
    },
    {
      "epoch": 0.2805907172995781,
      "grad_norm": 0.6416304707527161,
      "learning_rate": 0.0015,
      "loss": 1.5663,
      "step": 2660
    },
    {
      "epoch": 0.28164556962025317,
      "grad_norm": 0.5748112201690674,
      "learning_rate": 0.0015,
      "loss": 1.5675,
      "step": 2670
    },
    {
      "epoch": 0.28270042194092826,
      "grad_norm": 0.662480890750885,
      "learning_rate": 0.0015,
      "loss": 1.5698,
      "step": 2680
    },
    {
      "epoch": 0.28375527426160335,
      "grad_norm": 0.7659668326377869,
      "learning_rate": 0.0015,
      "loss": 1.5688,
      "step": 2690
    },
    {
      "epoch": 0.2848101265822785,
      "grad_norm": 0.5924683809280396,
      "learning_rate": 0.0015,
      "loss": 1.5646,
      "step": 2700
    },
    {
      "epoch": 0.2858649789029536,
      "grad_norm": 0.6555392742156982,
      "learning_rate": 0.0015,
      "loss": 1.5703,
      "step": 2710
    },
    {
      "epoch": 0.2869198312236287,
      "grad_norm": 0.56369549036026,
      "learning_rate": 0.0015,
      "loss": 1.5708,
      "step": 2720
    },
    {
      "epoch": 0.2879746835443038,
      "grad_norm": 0.674506664276123,
      "learning_rate": 0.0015,
      "loss": 1.5524,
      "step": 2730
    },
    {
      "epoch": 0.2890295358649789,
      "grad_norm": 0.6215850710868835,
      "learning_rate": 0.0015,
      "loss": 1.5631,
      "step": 2740
    },
    {
      "epoch": 0.290084388185654,
      "grad_norm": 0.5844721794128418,
      "learning_rate": 0.0015,
      "loss": 1.5713,
      "step": 2750
    },
    {
      "epoch": 0.2911392405063291,
      "grad_norm": 0.8595011234283447,
      "learning_rate": 0.0015,
      "loss": 1.577,
      "step": 2760
    },
    {
      "epoch": 0.2921940928270042,
      "grad_norm": 0.6657257080078125,
      "learning_rate": 0.0015,
      "loss": 1.5717,
      "step": 2770
    },
    {
      "epoch": 0.29324894514767935,
      "grad_norm": 0.6913807988166809,
      "learning_rate": 0.0015,
      "loss": 1.5685,
      "step": 2780
    },
    {
      "epoch": 0.29430379746835444,
      "grad_norm": 0.5601217150688171,
      "learning_rate": 0.0015,
      "loss": 1.5645,
      "step": 2790
    },
    {
      "epoch": 0.29535864978902954,
      "grad_norm": 0.6473618745803833,
      "learning_rate": 0.0015,
      "loss": 1.562,
      "step": 2800
    },
    {
      "epoch": 0.29641350210970463,
      "grad_norm": 0.5763059854507446,
      "learning_rate": 0.0015,
      "loss": 1.5653,
      "step": 2810
    },
    {
      "epoch": 0.2974683544303797,
      "grad_norm": 0.6512886881828308,
      "learning_rate": 0.0015,
      "loss": 1.5714,
      "step": 2820
    },
    {
      "epoch": 0.29852320675105487,
      "grad_norm": 0.9278571605682373,
      "learning_rate": 0.0015,
      "loss": 1.5471,
      "step": 2830
    },
    {
      "epoch": 0.29957805907172996,
      "grad_norm": 1.1715987920761108,
      "learning_rate": 0.0015,
      "loss": 1.5512,
      "step": 2840
    },
    {
      "epoch": 0.30063291139240506,
      "grad_norm": 0.7548184990882874,
      "learning_rate": 0.0015,
      "loss": 1.5553,
      "step": 2850
    },
    {
      "epoch": 0.30168776371308015,
      "grad_norm": 0.6434464454650879,
      "learning_rate": 0.0015,
      "loss": 1.5572,
      "step": 2860
    },
    {
      "epoch": 0.3027426160337553,
      "grad_norm": 0.9384394288063049,
      "learning_rate": 0.0015,
      "loss": 1.5584,
      "step": 2870
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.7197995781898499,
      "learning_rate": 0.0015,
      "loss": 1.5463,
      "step": 2880
    },
    {
      "epoch": 0.3048523206751055,
      "grad_norm": 0.6241549849510193,
      "learning_rate": 0.0015,
      "loss": 1.5513,
      "step": 2890
    },
    {
      "epoch": 0.3059071729957806,
      "grad_norm": 0.6179733872413635,
      "learning_rate": 0.0015,
      "loss": 1.5582,
      "step": 2900
    },
    {
      "epoch": 0.3069620253164557,
      "grad_norm": 0.7129524946212769,
      "learning_rate": 0.0015,
      "loss": 1.5592,
      "step": 2910
    },
    {
      "epoch": 0.3080168776371308,
      "grad_norm": 0.7332343459129333,
      "learning_rate": 0.0015,
      "loss": 1.5523,
      "step": 2920
    },
    {
      "epoch": 0.3090717299578059,
      "grad_norm": 0.6103625893592834,
      "learning_rate": 0.0015,
      "loss": 1.5479,
      "step": 2930
    },
    {
      "epoch": 0.310126582278481,
      "grad_norm": 0.6277995705604553,
      "learning_rate": 0.0015,
      "loss": 1.5548,
      "step": 2940
    },
    {
      "epoch": 0.3111814345991561,
      "grad_norm": 0.6909558773040771,
      "learning_rate": 0.0015,
      "loss": 1.5628,
      "step": 2950
    },
    {
      "epoch": 0.31223628691983124,
      "grad_norm": 1.1366920471191406,
      "learning_rate": 0.0015,
      "loss": 1.548,
      "step": 2960
    },
    {
      "epoch": 0.31329113924050633,
      "grad_norm": 1.1065442562103271,
      "learning_rate": 0.0015,
      "loss": 1.5391,
      "step": 2970
    },
    {
      "epoch": 0.3143459915611814,
      "grad_norm": 0.6319011449813843,
      "learning_rate": 0.0015,
      "loss": 1.5451,
      "step": 2980
    },
    {
      "epoch": 0.3154008438818565,
      "grad_norm": 0.6630522608757019,
      "learning_rate": 0.0015,
      "loss": 1.5607,
      "step": 2990
    },
    {
      "epoch": 0.31645569620253167,
      "grad_norm": 0.6209017634391785,
      "learning_rate": 0.0015,
      "loss": 1.5387,
      "step": 3000
    },
    {
      "epoch": 0.31751054852320676,
      "grad_norm": 0.5443456768989563,
      "learning_rate": 0.0015,
      "loss": 1.546,
      "step": 3010
    },
    {
      "epoch": 0.31856540084388185,
      "grad_norm": 0.6264317035675049,
      "learning_rate": 0.0015,
      "loss": 1.5462,
      "step": 3020
    },
    {
      "epoch": 0.31962025316455694,
      "grad_norm": 0.6058927774429321,
      "learning_rate": 0.0015,
      "loss": 1.5356,
      "step": 3030
    },
    {
      "epoch": 0.3206751054852321,
      "grad_norm": 0.6220033168792725,
      "learning_rate": 0.0015,
      "loss": 1.5442,
      "step": 3040
    },
    {
      "epoch": 0.3217299578059072,
      "grad_norm": 0.6217190027236938,
      "learning_rate": 0.0015,
      "loss": 1.5394,
      "step": 3050
    },
    {
      "epoch": 0.3227848101265823,
      "grad_norm": 0.5777936577796936,
      "learning_rate": 0.0015,
      "loss": 1.5479,
      "step": 3060
    },
    {
      "epoch": 0.32383966244725737,
      "grad_norm": 0.6267879605293274,
      "learning_rate": 0.0015,
      "loss": 1.541,
      "step": 3070
    },
    {
      "epoch": 0.32489451476793246,
      "grad_norm": 0.6611675024032593,
      "learning_rate": 0.0015,
      "loss": 1.5295,
      "step": 3080
    },
    {
      "epoch": 0.3259493670886076,
      "grad_norm": 0.6568845510482788,
      "learning_rate": 0.0015,
      "loss": 1.5486,
      "step": 3090
    },
    {
      "epoch": 0.3270042194092827,
      "grad_norm": 0.6635953783988953,
      "learning_rate": 0.0015,
      "loss": 1.5361,
      "step": 3100
    },
    {
      "epoch": 0.3280590717299578,
      "grad_norm": 0.6701586842536926,
      "learning_rate": 0.0015,
      "loss": 1.539,
      "step": 3110
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.5741431713104248,
      "learning_rate": 0.0015,
      "loss": 1.5402,
      "step": 3120
    },
    {
      "epoch": 0.33016877637130804,
      "grad_norm": 0.5856887698173523,
      "learning_rate": 0.0015,
      "loss": 1.5504,
      "step": 3130
    },
    {
      "epoch": 0.33122362869198313,
      "grad_norm": 0.6227253675460815,
      "learning_rate": 0.0015,
      "loss": 1.5286,
      "step": 3140
    },
    {
      "epoch": 0.3322784810126582,
      "grad_norm": 0.5713142156600952,
      "learning_rate": 0.0015,
      "loss": 1.5379,
      "step": 3150
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.5559601783752441,
      "learning_rate": 0.0015,
      "loss": 1.5388,
      "step": 3160
    },
    {
      "epoch": 0.33438818565400846,
      "grad_norm": 0.693455159664154,
      "learning_rate": 0.0015,
      "loss": 1.5439,
      "step": 3170
    },
    {
      "epoch": 0.33544303797468356,
      "grad_norm": 0.6805558800697327,
      "learning_rate": 0.0015,
      "loss": 1.5373,
      "step": 3180
    },
    {
      "epoch": 0.33649789029535865,
      "grad_norm": 0.8250734210014343,
      "learning_rate": 0.0015,
      "loss": 1.5352,
      "step": 3190
    },
    {
      "epoch": 0.33755274261603374,
      "grad_norm": 0.6972458958625793,
      "learning_rate": 0.0015,
      "loss": 1.5329,
      "step": 3200
    },
    {
      "epoch": 0.33860759493670883,
      "grad_norm": 0.6945505142211914,
      "learning_rate": 0.0015,
      "loss": 1.5413,
      "step": 3210
    },
    {
      "epoch": 0.339662447257384,
      "grad_norm": 0.7304695248603821,
      "learning_rate": 0.0015,
      "loss": 1.536,
      "step": 3220
    },
    {
      "epoch": 0.3407172995780591,
      "grad_norm": 0.7209477424621582,
      "learning_rate": 0.0015,
      "loss": 1.529,
      "step": 3230
    },
    {
      "epoch": 0.34177215189873417,
      "grad_norm": 0.5435463190078735,
      "learning_rate": 0.0015,
      "loss": 1.5086,
      "step": 3240
    },
    {
      "epoch": 0.34282700421940926,
      "grad_norm": 0.6081704497337341,
      "learning_rate": 0.0015,
      "loss": 1.5356,
      "step": 3250
    },
    {
      "epoch": 0.3438818565400844,
      "grad_norm": 0.6042312979698181,
      "learning_rate": 0.0015,
      "loss": 1.523,
      "step": 3260
    },
    {
      "epoch": 0.3449367088607595,
      "grad_norm": 0.6200316548347473,
      "learning_rate": 0.0015,
      "loss": 1.5428,
      "step": 3270
    },
    {
      "epoch": 0.3459915611814346,
      "grad_norm": 0.9771542549133301,
      "learning_rate": 0.0015,
      "loss": 1.5421,
      "step": 3280
    },
    {
      "epoch": 0.3470464135021097,
      "grad_norm": 1.0858914852142334,
      "learning_rate": 0.0015,
      "loss": 1.5248,
      "step": 3290
    },
    {
      "epoch": 0.34810126582278483,
      "grad_norm": 0.5964863300323486,
      "learning_rate": 0.0015,
      "loss": 1.5354,
      "step": 3300
    },
    {
      "epoch": 0.3491561181434599,
      "grad_norm": 0.6484059691429138,
      "learning_rate": 0.0015,
      "loss": 1.5339,
      "step": 3310
    },
    {
      "epoch": 0.350210970464135,
      "grad_norm": 0.6149209141731262,
      "learning_rate": 0.0015,
      "loss": 1.5415,
      "step": 3320
    },
    {
      "epoch": 0.3512658227848101,
      "grad_norm": 0.616062581539154,
      "learning_rate": 0.0015,
      "loss": 1.526,
      "step": 3330
    },
    {
      "epoch": 0.35232067510548526,
      "grad_norm": 0.6185603141784668,
      "learning_rate": 0.0015,
      "loss": 1.5272,
      "step": 3340
    },
    {
      "epoch": 0.35337552742616035,
      "grad_norm": 0.6322174072265625,
      "learning_rate": 0.0015,
      "loss": 1.5195,
      "step": 3350
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.5539165735244751,
      "learning_rate": 0.0015,
      "loss": 1.5197,
      "step": 3360
    },
    {
      "epoch": 0.35548523206751054,
      "grad_norm": 0.6246862411499023,
      "learning_rate": 0.0015,
      "loss": 1.5295,
      "step": 3370
    },
    {
      "epoch": 0.35654008438818563,
      "grad_norm": 0.563215434551239,
      "learning_rate": 0.0015,
      "loss": 1.5246,
      "step": 3380
    },
    {
      "epoch": 0.3575949367088608,
      "grad_norm": 0.6254818439483643,
      "learning_rate": 0.0015,
      "loss": 1.5194,
      "step": 3390
    },
    {
      "epoch": 0.35864978902953587,
      "grad_norm": 0.6142661571502686,
      "learning_rate": 0.0015,
      "loss": 1.529,
      "step": 3400
    },
    {
      "epoch": 0.35970464135021096,
      "grad_norm": 0.6170629262924194,
      "learning_rate": 0.0015,
      "loss": 1.5256,
      "step": 3410
    },
    {
      "epoch": 0.36075949367088606,
      "grad_norm": 0.6925488710403442,
      "learning_rate": 0.0015,
      "loss": 1.5072,
      "step": 3420
    },
    {
      "epoch": 0.3618143459915612,
      "grad_norm": 0.6214223504066467,
      "learning_rate": 0.0015,
      "loss": 1.525,
      "step": 3430
    },
    {
      "epoch": 0.3628691983122363,
      "grad_norm": 0.6367167830467224,
      "learning_rate": 0.0015,
      "loss": 1.5188,
      "step": 3440
    },
    {
      "epoch": 0.3639240506329114,
      "grad_norm": 0.5955284237861633,
      "learning_rate": 0.0015,
      "loss": 1.5162,
      "step": 3450
    },
    {
      "epoch": 0.3649789029535865,
      "grad_norm": 0.8075214624404907,
      "learning_rate": 0.0015,
      "loss": 1.5109,
      "step": 3460
    },
    {
      "epoch": 0.36603375527426163,
      "grad_norm": 0.6733569502830505,
      "learning_rate": 0.0015,
      "loss": 1.5271,
      "step": 3470
    },
    {
      "epoch": 0.3670886075949367,
      "grad_norm": 0.6443074941635132,
      "learning_rate": 0.0015,
      "loss": 1.5324,
      "step": 3480
    },
    {
      "epoch": 0.3681434599156118,
      "grad_norm": 0.5668088793754578,
      "learning_rate": 0.0015,
      "loss": 1.5227,
      "step": 3490
    },
    {
      "epoch": 0.3691983122362869,
      "grad_norm": 0.5876608490943909,
      "learning_rate": 0.0015,
      "loss": 1.5037,
      "step": 3500
    },
    {
      "epoch": 0.370253164556962,
      "grad_norm": 0.6632211804389954,
      "learning_rate": 0.0015,
      "loss": 1.5236,
      "step": 3510
    },
    {
      "epoch": 0.37130801687763715,
      "grad_norm": 0.5315051078796387,
      "learning_rate": 0.0015,
      "loss": 1.5217,
      "step": 3520
    },
    {
      "epoch": 0.37236286919831224,
      "grad_norm": 0.652911365032196,
      "learning_rate": 0.0015,
      "loss": 1.5295,
      "step": 3530
    },
    {
      "epoch": 0.37341772151898733,
      "grad_norm": 0.9318296909332275,
      "learning_rate": 0.0015,
      "loss": 1.523,
      "step": 3540
    },
    {
      "epoch": 0.3744725738396624,
      "grad_norm": 0.6587705612182617,
      "learning_rate": 0.0015,
      "loss": 1.5133,
      "step": 3550
    },
    {
      "epoch": 0.3755274261603376,
      "grad_norm": 0.586134672164917,
      "learning_rate": 0.0015,
      "loss": 1.5155,
      "step": 3560
    },
    {
      "epoch": 0.37658227848101267,
      "grad_norm": 0.6934296488761902,
      "learning_rate": 0.0015,
      "loss": 1.5097,
      "step": 3570
    },
    {
      "epoch": 0.37763713080168776,
      "grad_norm": 0.591506838798523,
      "learning_rate": 0.0015,
      "loss": 1.5117,
      "step": 3580
    },
    {
      "epoch": 0.37869198312236285,
      "grad_norm": 0.6162989735603333,
      "learning_rate": 0.0015,
      "loss": 1.5228,
      "step": 3590
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.7782261371612549,
      "learning_rate": 0.0015,
      "loss": 1.5094,
      "step": 3600
    },
    {
      "epoch": 0.3808016877637131,
      "grad_norm": 0.6871771216392517,
      "learning_rate": 0.0015,
      "loss": 1.5094,
      "step": 3610
    },
    {
      "epoch": 0.3818565400843882,
      "grad_norm": 0.6823848485946655,
      "learning_rate": 0.0015,
      "loss": 1.5118,
      "step": 3620
    },
    {
      "epoch": 0.3829113924050633,
      "grad_norm": 0.6270849108695984,
      "learning_rate": 0.0015,
      "loss": 1.5111,
      "step": 3630
    },
    {
      "epoch": 0.38396624472573837,
      "grad_norm": 0.8293321132659912,
      "learning_rate": 0.0015,
      "loss": 1.5135,
      "step": 3640
    },
    {
      "epoch": 0.3850210970464135,
      "grad_norm": 0.697428822517395,
      "learning_rate": 0.0015,
      "loss": 1.5199,
      "step": 3650
    },
    {
      "epoch": 0.3860759493670886,
      "grad_norm": 0.5967044234275818,
      "learning_rate": 0.0015,
      "loss": 1.5125,
      "step": 3660
    },
    {
      "epoch": 0.3871308016877637,
      "grad_norm": 0.5544715523719788,
      "learning_rate": 0.0015,
      "loss": 1.5075,
      "step": 3670
    },
    {
      "epoch": 0.3881856540084388,
      "grad_norm": 0.9727868437767029,
      "learning_rate": 0.0015,
      "loss": 1.5113,
      "step": 3680
    },
    {
      "epoch": 0.38924050632911394,
      "grad_norm": 0.7310069799423218,
      "learning_rate": 0.0015,
      "loss": 1.519,
      "step": 3690
    },
    {
      "epoch": 0.39029535864978904,
      "grad_norm": 0.7747533917427063,
      "learning_rate": 0.0015,
      "loss": 1.5223,
      "step": 3700
    },
    {
      "epoch": 0.39135021097046413,
      "grad_norm": 0.5666539072990417,
      "learning_rate": 0.0015,
      "loss": 1.5085,
      "step": 3710
    },
    {
      "epoch": 0.3924050632911392,
      "grad_norm": 0.59869384765625,
      "learning_rate": 0.0015,
      "loss": 1.4887,
      "step": 3720
    },
    {
      "epoch": 0.39345991561181437,
      "grad_norm": 0.520443856716156,
      "learning_rate": 0.0015,
      "loss": 1.5077,
      "step": 3730
    },
    {
      "epoch": 0.39451476793248946,
      "grad_norm": 0.6021689176559448,
      "learning_rate": 0.0015,
      "loss": 1.5085,
      "step": 3740
    },
    {
      "epoch": 0.39556962025316456,
      "grad_norm": 0.7624187469482422,
      "learning_rate": 0.0015,
      "loss": 1.5086,
      "step": 3750
    },
    {
      "epoch": 0.39662447257383965,
      "grad_norm": 0.8453581929206848,
      "learning_rate": 0.0015,
      "loss": 1.5013,
      "step": 3760
    },
    {
      "epoch": 0.39767932489451474,
      "grad_norm": 0.7012314796447754,
      "learning_rate": 0.0015,
      "loss": 1.518,
      "step": 3770
    },
    {
      "epoch": 0.3987341772151899,
      "grad_norm": 0.6165980696678162,
      "learning_rate": 0.0015,
      "loss": 1.5028,
      "step": 3780
    },
    {
      "epoch": 0.399789029535865,
      "grad_norm": 0.7216675281524658,
      "learning_rate": 0.0015,
      "loss": 1.5087,
      "step": 3790
    },
    {
      "epoch": 0.4008438818565401,
      "grad_norm": 0.6874868273735046,
      "learning_rate": 0.0015,
      "loss": 1.5116,
      "step": 3800
    },
    {
      "epoch": 0.40189873417721517,
      "grad_norm": 0.5580333471298218,
      "learning_rate": 0.0015,
      "loss": 1.5012,
      "step": 3810
    },
    {
      "epoch": 0.4029535864978903,
      "grad_norm": 0.6233204007148743,
      "learning_rate": 0.0015,
      "loss": 1.4946,
      "step": 3820
    },
    {
      "epoch": 0.4040084388185654,
      "grad_norm": 0.6824832558631897,
      "learning_rate": 0.0015,
      "loss": 1.4943,
      "step": 3830
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.5857206583023071,
      "learning_rate": 0.0015,
      "loss": 1.5137,
      "step": 3840
    },
    {
      "epoch": 0.4061181434599156,
      "grad_norm": 0.6413280367851257,
      "learning_rate": 0.0015,
      "loss": 1.497,
      "step": 3850
    },
    {
      "epoch": 0.40717299578059074,
      "grad_norm": 0.6493369340896606,
      "learning_rate": 0.0015,
      "loss": 1.5088,
      "step": 3860
    },
    {
      "epoch": 0.40822784810126583,
      "grad_norm": 0.6379006505012512,
      "learning_rate": 0.0015,
      "loss": 1.4983,
      "step": 3870
    },
    {
      "epoch": 0.4092827004219409,
      "grad_norm": 0.6437706351280212,
      "learning_rate": 0.0015,
      "loss": 1.4893,
      "step": 3880
    },
    {
      "epoch": 0.410337552742616,
      "grad_norm": 0.5746003985404968,
      "learning_rate": 0.0015,
      "loss": 1.4896,
      "step": 3890
    },
    {
      "epoch": 0.41139240506329117,
      "grad_norm": 1.0675814151763916,
      "learning_rate": 0.0015,
      "loss": 1.4927,
      "step": 3900
    },
    {
      "epoch": 0.41244725738396626,
      "grad_norm": 0.7103888392448425,
      "learning_rate": 0.0015,
      "loss": 1.4975,
      "step": 3910
    },
    {
      "epoch": 0.41350210970464135,
      "grad_norm": 0.53952556848526,
      "learning_rate": 0.0015,
      "loss": 1.4904,
      "step": 3920
    },
    {
      "epoch": 0.41455696202531644,
      "grad_norm": 0.5914176106452942,
      "learning_rate": 0.0015,
      "loss": 1.4952,
      "step": 3930
    },
    {
      "epoch": 0.41561181434599154,
      "grad_norm": 0.6269722580909729,
      "learning_rate": 0.0015,
      "loss": 1.4931,
      "step": 3940
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.6420876383781433,
      "learning_rate": 0.0015,
      "loss": 1.5002,
      "step": 3950
    },
    {
      "epoch": 0.4177215189873418,
      "grad_norm": 0.5700493454933167,
      "learning_rate": 0.0015,
      "loss": 1.5001,
      "step": 3960
    },
    {
      "epoch": 0.41877637130801687,
      "grad_norm": 0.6639719009399414,
      "learning_rate": 0.0015,
      "loss": 1.4879,
      "step": 3970
    },
    {
      "epoch": 0.41983122362869196,
      "grad_norm": 0.6735414266586304,
      "learning_rate": 0.0015,
      "loss": 1.4985,
      "step": 3980
    },
    {
      "epoch": 0.4208860759493671,
      "grad_norm": 0.6394850015640259,
      "learning_rate": 0.0015,
      "loss": 1.488,
      "step": 3990
    },
    {
      "epoch": 0.4219409282700422,
      "grad_norm": 0.6736827492713928,
      "learning_rate": 0.0015,
      "loss": 1.5041,
      "step": 4000
    },
    {
      "epoch": 0.4229957805907173,
      "grad_norm": 0.6361777186393738,
      "learning_rate": 0.0015,
      "loss": 1.489,
      "step": 4010
    },
    {
      "epoch": 0.4240506329113924,
      "grad_norm": 0.5799570679664612,
      "learning_rate": 0.0015,
      "loss": 1.4927,
      "step": 4020
    },
    {
      "epoch": 0.42510548523206754,
      "grad_norm": 0.5939344763755798,
      "learning_rate": 0.0015,
      "loss": 1.4866,
      "step": 4030
    },
    {
      "epoch": 0.42616033755274263,
      "grad_norm": 0.5622565746307373,
      "learning_rate": 0.0015,
      "loss": 1.4965,
      "step": 4040
    },
    {
      "epoch": 0.4272151898734177,
      "grad_norm": 0.5914807915687561,
      "learning_rate": 0.0015,
      "loss": 1.4961,
      "step": 4050
    },
    {
      "epoch": 0.4282700421940928,
      "grad_norm": 0.6562783122062683,
      "learning_rate": 0.0015,
      "loss": 1.4998,
      "step": 4060
    },
    {
      "epoch": 0.4293248945147679,
      "grad_norm": 0.6392592191696167,
      "learning_rate": 0.0015,
      "loss": 1.4958,
      "step": 4070
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.8016908168792725,
      "learning_rate": 0.0015,
      "loss": 1.5008,
      "step": 4080
    },
    {
      "epoch": 0.43143459915611815,
      "grad_norm": 0.6368083953857422,
      "learning_rate": 0.0015,
      "loss": 1.501,
      "step": 4090
    },
    {
      "epoch": 0.43248945147679324,
      "grad_norm": 0.6188535690307617,
      "learning_rate": 0.0015,
      "loss": 1.4872,
      "step": 4100
    },
    {
      "epoch": 0.43354430379746833,
      "grad_norm": 0.7229503393173218,
      "learning_rate": 0.0015,
      "loss": 1.4801,
      "step": 4110
    },
    {
      "epoch": 0.4345991561181435,
      "grad_norm": 0.6336178183555603,
      "learning_rate": 0.0015,
      "loss": 1.4864,
      "step": 4120
    },
    {
      "epoch": 0.4356540084388186,
      "grad_norm": 0.5833616256713867,
      "learning_rate": 0.0015,
      "loss": 1.4815,
      "step": 4130
    },
    {
      "epoch": 0.43670886075949367,
      "grad_norm": 0.5828094482421875,
      "learning_rate": 0.0015,
      "loss": 1.4856,
      "step": 4140
    },
    {
      "epoch": 0.43776371308016876,
      "grad_norm": 0.6903582215309143,
      "learning_rate": 0.0015,
      "loss": 1.4899,
      "step": 4150
    },
    {
      "epoch": 0.4388185654008439,
      "grad_norm": 0.6231018900871277,
      "learning_rate": 0.0015,
      "loss": 1.4857,
      "step": 4160
    },
    {
      "epoch": 0.439873417721519,
      "grad_norm": 0.5356478095054626,
      "learning_rate": 0.0015,
      "loss": 1.4836,
      "step": 4170
    },
    {
      "epoch": 0.4409282700421941,
      "grad_norm": 0.5384558439254761,
      "learning_rate": 0.0015,
      "loss": 1.4791,
      "step": 4180
    },
    {
      "epoch": 0.4419831223628692,
      "grad_norm": 0.7372967600822449,
      "learning_rate": 0.0015,
      "loss": 1.4871,
      "step": 4190
    },
    {
      "epoch": 0.4430379746835443,
      "grad_norm": 0.700654923915863,
      "learning_rate": 0.0015,
      "loss": 1.4888,
      "step": 4200
    },
    {
      "epoch": 0.4440928270042194,
      "grad_norm": 0.5394490957260132,
      "learning_rate": 0.0015,
      "loss": 1.4772,
      "step": 4210
    },
    {
      "epoch": 0.4451476793248945,
      "grad_norm": 0.614011824131012,
      "learning_rate": 0.0015,
      "loss": 1.487,
      "step": 4220
    },
    {
      "epoch": 0.4462025316455696,
      "grad_norm": 0.7026930451393127,
      "learning_rate": 0.0015,
      "loss": 1.4813,
      "step": 4230
    },
    {
      "epoch": 0.4472573839662447,
      "grad_norm": 0.5955589413642883,
      "learning_rate": 0.0015,
      "loss": 1.4815,
      "step": 4240
    },
    {
      "epoch": 0.44831223628691985,
      "grad_norm": 0.8915395140647888,
      "learning_rate": 0.0015,
      "loss": 1.4865,
      "step": 4250
    },
    {
      "epoch": 0.44936708860759494,
      "grad_norm": 0.663708508014679,
      "learning_rate": 0.0015,
      "loss": 1.4911,
      "step": 4260
    },
    {
      "epoch": 0.45042194092827004,
      "grad_norm": 0.5448995232582092,
      "learning_rate": 0.0015,
      "loss": 1.4808,
      "step": 4270
    },
    {
      "epoch": 0.45147679324894513,
      "grad_norm": 0.6886418461799622,
      "learning_rate": 0.0015,
      "loss": 1.475,
      "step": 4280
    },
    {
      "epoch": 0.4525316455696203,
      "grad_norm": 0.7126356363296509,
      "learning_rate": 0.0015,
      "loss": 1.486,
      "step": 4290
    },
    {
      "epoch": 0.45358649789029537,
      "grad_norm": 0.5966187715530396,
      "learning_rate": 0.0015,
      "loss": 1.4828,
      "step": 4300
    },
    {
      "epoch": 0.45464135021097046,
      "grad_norm": 0.5577530860900879,
      "learning_rate": 0.0015,
      "loss": 1.4824,
      "step": 4310
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.7881457209587097,
      "learning_rate": 0.0015,
      "loss": 1.4776,
      "step": 4320
    },
    {
      "epoch": 0.45675105485232065,
      "grad_norm": 0.5317174196243286,
      "learning_rate": 0.0015,
      "loss": 1.4862,
      "step": 4330
    },
    {
      "epoch": 0.4578059071729958,
      "grad_norm": 0.5694594383239746,
      "learning_rate": 0.0015,
      "loss": 1.4766,
      "step": 4340
    },
    {
      "epoch": 0.4588607594936709,
      "grad_norm": 0.6109514832496643,
      "learning_rate": 0.0015,
      "loss": 1.4795,
      "step": 4350
    },
    {
      "epoch": 0.459915611814346,
      "grad_norm": 0.5422523021697998,
      "learning_rate": 0.0015,
      "loss": 1.4936,
      "step": 4360
    },
    {
      "epoch": 0.4609704641350211,
      "grad_norm": 0.5639380216598511,
      "learning_rate": 0.0015,
      "loss": 1.4682,
      "step": 4370
    },
    {
      "epoch": 0.4620253164556962,
      "grad_norm": 0.5603376626968384,
      "learning_rate": 0.0015,
      "loss": 1.4781,
      "step": 4380
    },
    {
      "epoch": 0.4630801687763713,
      "grad_norm": 0.5457561016082764,
      "learning_rate": 0.0015,
      "loss": 1.4733,
      "step": 4390
    },
    {
      "epoch": 0.4641350210970464,
      "grad_norm": 0.5526235103607178,
      "learning_rate": 0.0015,
      "loss": 1.4797,
      "step": 4400
    },
    {
      "epoch": 0.4651898734177215,
      "grad_norm": 0.6319583058357239,
      "learning_rate": 0.0015,
      "loss": 1.4939,
      "step": 4410
    },
    {
      "epoch": 0.46624472573839665,
      "grad_norm": 0.6389647722244263,
      "learning_rate": 0.0015,
      "loss": 1.4749,
      "step": 4420
    },
    {
      "epoch": 0.46729957805907174,
      "grad_norm": 0.6226533651351929,
      "learning_rate": 0.0015,
      "loss": 1.4792,
      "step": 4430
    },
    {
      "epoch": 0.46835443037974683,
      "grad_norm": 0.6680729985237122,
      "learning_rate": 0.0015,
      "loss": 1.4782,
      "step": 4440
    },
    {
      "epoch": 0.4694092827004219,
      "grad_norm": 0.699388861656189,
      "learning_rate": 0.0015,
      "loss": 1.4839,
      "step": 4450
    },
    {
      "epoch": 0.4704641350210971,
      "grad_norm": 0.5822103023529053,
      "learning_rate": 0.0015,
      "loss": 1.4698,
      "step": 4460
    },
    {
      "epoch": 0.47151898734177217,
      "grad_norm": 0.5479047894477844,
      "learning_rate": 0.0015,
      "loss": 1.4802,
      "step": 4470
    },
    {
      "epoch": 0.47257383966244726,
      "grad_norm": 0.5745465159416199,
      "learning_rate": 0.0015,
      "loss": 1.4838,
      "step": 4480
    },
    {
      "epoch": 0.47362869198312235,
      "grad_norm": 0.6743048429489136,
      "learning_rate": 0.0015,
      "loss": 1.4712,
      "step": 4490
    },
    {
      "epoch": 0.47468354430379744,
      "grad_norm": 0.6289355158805847,
      "learning_rate": 0.0015,
      "loss": 1.4706,
      "step": 4500
    },
    {
      "epoch": 0.4757383966244726,
      "grad_norm": 0.5514520406723022,
      "learning_rate": 0.0015,
      "loss": 1.4788,
      "step": 4510
    },
    {
      "epoch": 0.4767932489451477,
      "grad_norm": 0.6398304104804993,
      "learning_rate": 0.0015,
      "loss": 1.4609,
      "step": 4520
    },
    {
      "epoch": 0.4778481012658228,
      "grad_norm": 0.6768178939819336,
      "learning_rate": 0.0015,
      "loss": 1.479,
      "step": 4530
    },
    {
      "epoch": 0.47890295358649787,
      "grad_norm": 0.5781738758087158,
      "learning_rate": 0.0015,
      "loss": 1.4804,
      "step": 4540
    },
    {
      "epoch": 0.479957805907173,
      "grad_norm": 0.6289318203926086,
      "learning_rate": 0.0015,
      "loss": 1.4785,
      "step": 4550
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.6006326675415039,
      "learning_rate": 0.0015,
      "loss": 1.4664,
      "step": 4560
    },
    {
      "epoch": 0.4820675105485232,
      "grad_norm": 0.5571173429489136,
      "learning_rate": 0.0015,
      "loss": 1.4487,
      "step": 4570
    },
    {
      "epoch": 0.4831223628691983,
      "grad_norm": 0.7171228528022766,
      "learning_rate": 0.0015,
      "loss": 1.4739,
      "step": 4580
    },
    {
      "epoch": 0.48417721518987344,
      "grad_norm": 0.7601715922355652,
      "learning_rate": 0.0015,
      "loss": 1.4601,
      "step": 4590
    },
    {
      "epoch": 0.48523206751054854,
      "grad_norm": 0.6442757248878479,
      "learning_rate": 0.0015,
      "loss": 1.4807,
      "step": 4600
    },
    {
      "epoch": 0.48628691983122363,
      "grad_norm": 0.8816727995872498,
      "learning_rate": 0.0015,
      "loss": 1.4832,
      "step": 4610
    },
    {
      "epoch": 0.4873417721518987,
      "grad_norm": 0.6388178467750549,
      "learning_rate": 0.0015,
      "loss": 1.4599,
      "step": 4620
    },
    {
      "epoch": 0.4883966244725738,
      "grad_norm": 0.6010212898254395,
      "learning_rate": 0.0015,
      "loss": 1.476,
      "step": 4630
    },
    {
      "epoch": 0.48945147679324896,
      "grad_norm": 0.5553440451622009,
      "learning_rate": 0.0015,
      "loss": 1.4746,
      "step": 4640
    },
    {
      "epoch": 0.49050632911392406,
      "grad_norm": 0.5826687216758728,
      "learning_rate": 0.0015,
      "loss": 1.4771,
      "step": 4650
    },
    {
      "epoch": 0.49156118143459915,
      "grad_norm": 0.565367579460144,
      "learning_rate": 0.0015,
      "loss": 1.4675,
      "step": 4660
    },
    {
      "epoch": 0.49261603375527424,
      "grad_norm": 0.8090323805809021,
      "learning_rate": 0.0015,
      "loss": 1.4669,
      "step": 4670
    },
    {
      "epoch": 0.4936708860759494,
      "grad_norm": 0.5813798904418945,
      "learning_rate": 0.0015,
      "loss": 1.4625,
      "step": 4680
    },
    {
      "epoch": 0.4947257383966245,
      "grad_norm": 0.8920308947563171,
      "learning_rate": 0.0015,
      "loss": 1.4601,
      "step": 4690
    },
    {
      "epoch": 0.4957805907172996,
      "grad_norm": 0.5381173491477966,
      "learning_rate": 0.0015,
      "loss": 1.4686,
      "step": 4700
    },
    {
      "epoch": 0.49683544303797467,
      "grad_norm": 0.5849639773368835,
      "learning_rate": 0.0015,
      "loss": 1.4717,
      "step": 4710
    },
    {
      "epoch": 0.4978902953586498,
      "grad_norm": 0.6569827795028687,
      "learning_rate": 0.0015,
      "loss": 1.4651,
      "step": 4720
    },
    {
      "epoch": 0.4989451476793249,
      "grad_norm": 0.5914872884750366,
      "learning_rate": 0.0015,
      "loss": 1.462,
      "step": 4730
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.7037190794944763,
      "learning_rate": 0.0015,
      "loss": 1.4619,
      "step": 4740
    },
    {
      "epoch": 0.5010548523206751,
      "grad_norm": 0.7166423797607422,
      "learning_rate": 0.0015,
      "loss": 1.4828,
      "step": 4750
    },
    {
      "epoch": 0.5021097046413502,
      "grad_norm": 0.5268374681472778,
      "learning_rate": 0.0015,
      "loss": 1.4669,
      "step": 4760
    },
    {
      "epoch": 0.5031645569620253,
      "grad_norm": 0.5577636361122131,
      "learning_rate": 0.0015,
      "loss": 1.4635,
      "step": 4770
    },
    {
      "epoch": 0.5042194092827004,
      "grad_norm": 0.5795514583587646,
      "learning_rate": 0.0015,
      "loss": 1.475,
      "step": 4780
    },
    {
      "epoch": 0.5052742616033755,
      "grad_norm": 0.670405924320221,
      "learning_rate": 0.0015,
      "loss": 1.4563,
      "step": 4790
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.7040602564811707,
      "learning_rate": 0.0015,
      "loss": 1.4665,
      "step": 4800
    },
    {
      "epoch": 0.5073839662447257,
      "grad_norm": 0.6669322848320007,
      "learning_rate": 0.0015,
      "loss": 1.4647,
      "step": 4810
    },
    {
      "epoch": 0.5084388185654009,
      "grad_norm": 0.7414965629577637,
      "learning_rate": 0.0015,
      "loss": 1.4835,
      "step": 4820
    },
    {
      "epoch": 0.509493670886076,
      "grad_norm": 0.5885231494903564,
      "learning_rate": 0.0015,
      "loss": 1.4645,
      "step": 4830
    },
    {
      "epoch": 0.510548523206751,
      "grad_norm": 0.6001221537590027,
      "learning_rate": 0.0015,
      "loss": 1.4597,
      "step": 4840
    },
    {
      "epoch": 0.5116033755274262,
      "grad_norm": 0.5307698249816895,
      "learning_rate": 0.0015,
      "loss": 1.4736,
      "step": 4850
    },
    {
      "epoch": 0.5126582278481012,
      "grad_norm": 0.5808476209640503,
      "learning_rate": 0.0015,
      "loss": 1.4604,
      "step": 4860
    },
    {
      "epoch": 0.5137130801687764,
      "grad_norm": 0.8642584085464478,
      "learning_rate": 0.0015,
      "loss": 1.454,
      "step": 4870
    },
    {
      "epoch": 0.5147679324894515,
      "grad_norm": 0.5445716381072998,
      "learning_rate": 0.0015,
      "loss": 1.4775,
      "step": 4880
    },
    {
      "epoch": 0.5158227848101266,
      "grad_norm": 0.5095697045326233,
      "learning_rate": 0.0015,
      "loss": 1.4624,
      "step": 4890
    },
    {
      "epoch": 0.5168776371308017,
      "grad_norm": 0.633233368396759,
      "learning_rate": 0.0015,
      "loss": 1.474,
      "step": 4900
    },
    {
      "epoch": 0.5179324894514767,
      "grad_norm": 0.5968495607376099,
      "learning_rate": 0.0015,
      "loss": 1.4661,
      "step": 4910
    },
    {
      "epoch": 0.5189873417721519,
      "grad_norm": 0.7888312339782715,
      "learning_rate": 0.0015,
      "loss": 1.4584,
      "step": 4920
    },
    {
      "epoch": 0.520042194092827,
      "grad_norm": 0.7476255893707275,
      "learning_rate": 0.0015,
      "loss": 1.4646,
      "step": 4930
    },
    {
      "epoch": 0.5210970464135021,
      "grad_norm": 0.6436273455619812,
      "learning_rate": 0.0015,
      "loss": 1.4643,
      "step": 4940
    },
    {
      "epoch": 0.5221518987341772,
      "grad_norm": 0.5799338221549988,
      "learning_rate": 0.0015,
      "loss": 1.47,
      "step": 4950
    },
    {
      "epoch": 0.5232067510548524,
      "grad_norm": 0.6167886853218079,
      "learning_rate": 0.0015,
      "loss": 1.4514,
      "step": 4960
    },
    {
      "epoch": 0.5242616033755274,
      "grad_norm": 0.57327800989151,
      "learning_rate": 0.0015,
      "loss": 1.4614,
      "step": 4970
    },
    {
      "epoch": 0.5253164556962026,
      "grad_norm": 0.595667839050293,
      "learning_rate": 0.0015,
      "loss": 1.4537,
      "step": 4980
    },
    {
      "epoch": 0.5263713080168776,
      "grad_norm": 0.5202828049659729,
      "learning_rate": 0.0015,
      "loss": 1.4572,
      "step": 4990
    },
    {
      "epoch": 0.5274261603375527,
      "grad_norm": 0.561742901802063,
      "learning_rate": 0.0015,
      "loss": 1.4678,
      "step": 5000
    },
    {
      "epoch": 0.5284810126582279,
      "grad_norm": 0.6367873549461365,
      "learning_rate": 0.0015,
      "loss": 1.4594,
      "step": 5010
    },
    {
      "epoch": 0.5295358649789029,
      "grad_norm": 0.5855687260627747,
      "learning_rate": 0.0015,
      "loss": 1.4534,
      "step": 5020
    },
    {
      "epoch": 0.5305907172995781,
      "grad_norm": 0.5318405032157898,
      "learning_rate": 0.0015,
      "loss": 1.4396,
      "step": 5030
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.5266106128692627,
      "learning_rate": 0.0015,
      "loss": 1.4657,
      "step": 5040
    },
    {
      "epoch": 0.5327004219409283,
      "grad_norm": 0.6130898594856262,
      "learning_rate": 0.0015,
      "loss": 1.4534,
      "step": 5050
    },
    {
      "epoch": 0.5337552742616034,
      "grad_norm": 0.5058093667030334,
      "learning_rate": 0.0015,
      "loss": 1.465,
      "step": 5060
    },
    {
      "epoch": 0.5348101265822784,
      "grad_norm": 0.7037327289581299,
      "learning_rate": 0.0015,
      "loss": 1.452,
      "step": 5070
    },
    {
      "epoch": 0.5358649789029536,
      "grad_norm": 0.5680815577507019,
      "learning_rate": 0.0015,
      "loss": 1.4676,
      "step": 5080
    },
    {
      "epoch": 0.5369198312236287,
      "grad_norm": 0.7030700445175171,
      "learning_rate": 0.0015,
      "loss": 1.4598,
      "step": 5090
    },
    {
      "epoch": 0.5379746835443038,
      "grad_norm": 0.6677927374839783,
      "learning_rate": 0.0015,
      "loss": 1.442,
      "step": 5100
    },
    {
      "epoch": 0.5390295358649789,
      "grad_norm": 0.5371360182762146,
      "learning_rate": 0.0015,
      "loss": 1.456,
      "step": 5110
    },
    {
      "epoch": 0.540084388185654,
      "grad_norm": 0.6903518438339233,
      "learning_rate": 0.0015,
      "loss": 1.4524,
      "step": 5120
    },
    {
      "epoch": 0.5411392405063291,
      "grad_norm": 0.5381430387496948,
      "learning_rate": 0.0015,
      "loss": 1.4489,
      "step": 5130
    },
    {
      "epoch": 0.5421940928270043,
      "grad_norm": 0.7357153296470642,
      "learning_rate": 0.0015,
      "loss": 1.4644,
      "step": 5140
    },
    {
      "epoch": 0.5432489451476793,
      "grad_norm": 0.5630563497543335,
      "learning_rate": 0.0015,
      "loss": 1.4599,
      "step": 5150
    },
    {
      "epoch": 0.5443037974683544,
      "grad_norm": 0.5637800693511963,
      "learning_rate": 0.0015,
      "loss": 1.4546,
      "step": 5160
    },
    {
      "epoch": 0.5453586497890295,
      "grad_norm": 0.5217317342758179,
      "learning_rate": 0.0015,
      "loss": 1.4565,
      "step": 5170
    },
    {
      "epoch": 0.5464135021097046,
      "grad_norm": 0.6386004090309143,
      "learning_rate": 0.0015,
      "loss": 1.4663,
      "step": 5180
    },
    {
      "epoch": 0.5474683544303798,
      "grad_norm": 0.6506286859512329,
      "learning_rate": 0.0015,
      "loss": 1.4465,
      "step": 5190
    },
    {
      "epoch": 0.5485232067510548,
      "grad_norm": 0.7016057372093201,
      "learning_rate": 0.0015,
      "loss": 1.4591,
      "step": 5200
    },
    {
      "epoch": 0.54957805907173,
      "grad_norm": 0.6944337487220764,
      "learning_rate": 0.0015,
      "loss": 1.4463,
      "step": 5210
    },
    {
      "epoch": 0.5506329113924051,
      "grad_norm": 0.6351850628852844,
      "learning_rate": 0.0015,
      "loss": 1.4553,
      "step": 5220
    },
    {
      "epoch": 0.5516877637130801,
      "grad_norm": 0.6268001794815063,
      "learning_rate": 0.0015,
      "loss": 1.4496,
      "step": 5230
    },
    {
      "epoch": 0.5527426160337553,
      "grad_norm": 0.5437992215156555,
      "learning_rate": 0.0015,
      "loss": 1.4564,
      "step": 5240
    },
    {
      "epoch": 0.5537974683544303,
      "grad_norm": 0.6136021614074707,
      "learning_rate": 0.0015,
      "loss": 1.4477,
      "step": 5250
    },
    {
      "epoch": 0.5548523206751055,
      "grad_norm": 0.5644496083259583,
      "learning_rate": 0.0015,
      "loss": 1.4539,
      "step": 5260
    },
    {
      "epoch": 0.5559071729957806,
      "grad_norm": 0.5912958979606628,
      "learning_rate": 0.0015,
      "loss": 1.461,
      "step": 5270
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.725299596786499,
      "learning_rate": 0.0015,
      "loss": 1.4503,
      "step": 5280
    },
    {
      "epoch": 0.5580168776371308,
      "grad_norm": 0.527134358882904,
      "learning_rate": 0.0015,
      "loss": 1.4504,
      "step": 5290
    },
    {
      "epoch": 0.5590717299578059,
      "grad_norm": 0.706976056098938,
      "learning_rate": 0.0015,
      "loss": 1.4573,
      "step": 5300
    },
    {
      "epoch": 0.560126582278481,
      "grad_norm": 0.8941502571105957,
      "learning_rate": 0.0015,
      "loss": 1.4439,
      "step": 5310
    },
    {
      "epoch": 0.5611814345991561,
      "grad_norm": 0.6011282205581665,
      "learning_rate": 0.0015,
      "loss": 1.4446,
      "step": 5320
    },
    {
      "epoch": 0.5622362869198312,
      "grad_norm": 0.5284645557403564,
      "learning_rate": 0.0015,
      "loss": 1.4557,
      "step": 5330
    },
    {
      "epoch": 0.5632911392405063,
      "grad_norm": 0.5965940952301025,
      "learning_rate": 0.0015,
      "loss": 1.4487,
      "step": 5340
    },
    {
      "epoch": 0.5643459915611815,
      "grad_norm": 0.5459398627281189,
      "learning_rate": 0.0015,
      "loss": 1.4382,
      "step": 5350
    },
    {
      "epoch": 0.5654008438818565,
      "grad_norm": 0.5910132527351379,
      "learning_rate": 0.0015,
      "loss": 1.4485,
      "step": 5360
    },
    {
      "epoch": 0.5664556962025317,
      "grad_norm": 0.5403808951377869,
      "learning_rate": 0.0015,
      "loss": 1.4485,
      "step": 5370
    },
    {
      "epoch": 0.5675105485232067,
      "grad_norm": 0.743107259273529,
      "learning_rate": 0.0015,
      "loss": 1.4454,
      "step": 5380
    },
    {
      "epoch": 0.5685654008438819,
      "grad_norm": 0.7197443842887878,
      "learning_rate": 0.0015,
      "loss": 1.4446,
      "step": 5390
    },
    {
      "epoch": 0.569620253164557,
      "grad_norm": 0.5556684732437134,
      "learning_rate": 0.0015,
      "loss": 1.4439,
      "step": 5400
    },
    {
      "epoch": 0.570675105485232,
      "grad_norm": 0.5397831201553345,
      "learning_rate": 0.0015,
      "loss": 1.4466,
      "step": 5410
    },
    {
      "epoch": 0.5717299578059072,
      "grad_norm": 0.5782108902931213,
      "learning_rate": 0.0015,
      "loss": 1.4414,
      "step": 5420
    },
    {
      "epoch": 0.5727848101265823,
      "grad_norm": 0.7732884287834167,
      "learning_rate": 0.0015,
      "loss": 1.4402,
      "step": 5430
    },
    {
      "epoch": 0.5738396624472574,
      "grad_norm": 0.5555076003074646,
      "learning_rate": 0.0015,
      "loss": 1.4392,
      "step": 5440
    },
    {
      "epoch": 0.5748945147679325,
      "grad_norm": 0.5733400583267212,
      "learning_rate": 0.0015,
      "loss": 1.4459,
      "step": 5450
    },
    {
      "epoch": 0.5759493670886076,
      "grad_norm": 0.6663935780525208,
      "learning_rate": 0.0015,
      "loss": 1.448,
      "step": 5460
    },
    {
      "epoch": 0.5770042194092827,
      "grad_norm": 0.5980082750320435,
      "learning_rate": 0.0015,
      "loss": 1.4468,
      "step": 5470
    },
    {
      "epoch": 0.5780590717299579,
      "grad_norm": 0.7896722555160522,
      "learning_rate": 0.0015,
      "loss": 1.4477,
      "step": 5480
    },
    {
      "epoch": 0.5791139240506329,
      "grad_norm": 0.5920150876045227,
      "learning_rate": 0.0015,
      "loss": 1.4514,
      "step": 5490
    },
    {
      "epoch": 0.580168776371308,
      "grad_norm": 0.6280402541160583,
      "learning_rate": 0.0015,
      "loss": 1.4431,
      "step": 5500
    },
    {
      "epoch": 0.5812236286919831,
      "grad_norm": 0.6328909993171692,
      "learning_rate": 0.0015,
      "loss": 1.4502,
      "step": 5510
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.599063515663147,
      "learning_rate": 0.0015,
      "loss": 1.44,
      "step": 5520
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.6359413266181946,
      "learning_rate": 0.0015,
      "loss": 1.436,
      "step": 5530
    },
    {
      "epoch": 0.5843881856540084,
      "grad_norm": 0.5926364064216614,
      "learning_rate": 0.0015,
      "loss": 1.4371,
      "step": 5540
    },
    {
      "epoch": 0.5854430379746836,
      "grad_norm": 0.608309805393219,
      "learning_rate": 0.0015,
      "loss": 1.4529,
      "step": 5550
    },
    {
      "epoch": 0.5864978902953587,
      "grad_norm": 0.5794243812561035,
      "learning_rate": 0.0015,
      "loss": 1.4546,
      "step": 5560
    },
    {
      "epoch": 0.5875527426160337,
      "grad_norm": 0.5262650847434998,
      "learning_rate": 0.0015,
      "loss": 1.4407,
      "step": 5570
    },
    {
      "epoch": 0.5886075949367089,
      "grad_norm": 0.5861445665359497,
      "learning_rate": 0.0015,
      "loss": 1.4393,
      "step": 5580
    },
    {
      "epoch": 0.5896624472573839,
      "grad_norm": 0.5474296808242798,
      "learning_rate": 0.0015,
      "loss": 1.4427,
      "step": 5590
    },
    {
      "epoch": 0.5907172995780591,
      "grad_norm": 0.8075207471847534,
      "learning_rate": 0.0015,
      "loss": 1.4387,
      "step": 5600
    },
    {
      "epoch": 0.5917721518987342,
      "grad_norm": 0.5450811386108398,
      "learning_rate": 0.0015,
      "loss": 1.4435,
      "step": 5610
    },
    {
      "epoch": 0.5928270042194093,
      "grad_norm": 0.6229982376098633,
      "learning_rate": 0.0015,
      "loss": 1.4514,
      "step": 5620
    },
    {
      "epoch": 0.5938818565400844,
      "grad_norm": 0.7292780876159668,
      "learning_rate": 0.0015,
      "loss": 1.4442,
      "step": 5630
    },
    {
      "epoch": 0.5949367088607594,
      "grad_norm": 0.5265942215919495,
      "learning_rate": 0.0015,
      "loss": 1.4549,
      "step": 5640
    },
    {
      "epoch": 0.5959915611814346,
      "grad_norm": 0.5641297101974487,
      "learning_rate": 0.0015,
      "loss": 1.4506,
      "step": 5650
    },
    {
      "epoch": 0.5970464135021097,
      "grad_norm": 0.8600500822067261,
      "learning_rate": 0.0015,
      "loss": 1.446,
      "step": 5660
    },
    {
      "epoch": 0.5981012658227848,
      "grad_norm": 0.5192574858665466,
      "learning_rate": 0.0015,
      "loss": 1.4502,
      "step": 5670
    },
    {
      "epoch": 0.5991561181434599,
      "grad_norm": 0.5317137241363525,
      "learning_rate": 0.0015,
      "loss": 1.4482,
      "step": 5680
    },
    {
      "epoch": 0.6002109704641351,
      "grad_norm": 0.8360275030136108,
      "learning_rate": 0.0015,
      "loss": 1.4275,
      "step": 5690
    },
    {
      "epoch": 0.6012658227848101,
      "grad_norm": 0.6226162314414978,
      "learning_rate": 0.0015,
      "loss": 1.438,
      "step": 5700
    },
    {
      "epoch": 0.6023206751054853,
      "grad_norm": 0.6044192910194397,
      "learning_rate": 0.0015,
      "loss": 1.4403,
      "step": 5710
    },
    {
      "epoch": 0.6033755274261603,
      "grad_norm": 0.5660203099250793,
      "learning_rate": 0.0015,
      "loss": 1.4417,
      "step": 5720
    },
    {
      "epoch": 0.6044303797468354,
      "grad_norm": 0.5431947708129883,
      "learning_rate": 0.0015,
      "loss": 1.4419,
      "step": 5730
    },
    {
      "epoch": 0.6054852320675106,
      "grad_norm": 0.6212903261184692,
      "learning_rate": 0.0015,
      "loss": 1.44,
      "step": 5740
    },
    {
      "epoch": 0.6065400843881856,
      "grad_norm": 0.6808838248252869,
      "learning_rate": 0.0015,
      "loss": 1.414,
      "step": 5750
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.561962366104126,
      "learning_rate": 0.0015,
      "loss": 1.4223,
      "step": 5760
    },
    {
      "epoch": 0.6086497890295358,
      "grad_norm": 0.6142343282699585,
      "learning_rate": 0.0015,
      "loss": 1.4247,
      "step": 5770
    },
    {
      "epoch": 0.609704641350211,
      "grad_norm": 0.5520644783973694,
      "learning_rate": 0.0015,
      "loss": 1.4344,
      "step": 5780
    },
    {
      "epoch": 0.6107594936708861,
      "grad_norm": 0.5744478702545166,
      "learning_rate": 0.0015,
      "loss": 1.4308,
      "step": 5790
    },
    {
      "epoch": 0.6118143459915611,
      "grad_norm": 0.6295973658561707,
      "learning_rate": 0.0015,
      "loss": 1.4377,
      "step": 5800
    },
    {
      "epoch": 0.6128691983122363,
      "grad_norm": 0.5243735313415527,
      "learning_rate": 0.0015,
      "loss": 1.4401,
      "step": 5810
    },
    {
      "epoch": 0.6139240506329114,
      "grad_norm": 0.5113600492477417,
      "learning_rate": 0.0015,
      "loss": 1.4511,
      "step": 5820
    },
    {
      "epoch": 0.6149789029535865,
      "grad_norm": 0.585241973400116,
      "learning_rate": 0.0015,
      "loss": 1.4345,
      "step": 5830
    },
    {
      "epoch": 0.6160337552742616,
      "grad_norm": 0.6048781871795654,
      "learning_rate": 0.0015,
      "loss": 1.4481,
      "step": 5840
    },
    {
      "epoch": 0.6170886075949367,
      "grad_norm": 0.5778324604034424,
      "learning_rate": 0.0015,
      "loss": 1.4512,
      "step": 5850
    },
    {
      "epoch": 0.6181434599156118,
      "grad_norm": 0.5663139820098877,
      "learning_rate": 0.0015,
      "loss": 1.4372,
      "step": 5860
    },
    {
      "epoch": 0.619198312236287,
      "grad_norm": 0.516270637512207,
      "learning_rate": 0.0015,
      "loss": 1.4333,
      "step": 5870
    },
    {
      "epoch": 0.620253164556962,
      "grad_norm": 0.6882340312004089,
      "learning_rate": 0.0015,
      "loss": 1.4428,
      "step": 5880
    },
    {
      "epoch": 0.6213080168776371,
      "grad_norm": 0.6412220001220703,
      "learning_rate": 0.0015,
      "loss": 1.4334,
      "step": 5890
    },
    {
      "epoch": 0.6223628691983122,
      "grad_norm": 0.5238667130470276,
      "learning_rate": 0.0015,
      "loss": 1.435,
      "step": 5900
    },
    {
      "epoch": 0.6234177215189873,
      "grad_norm": 0.5808557868003845,
      "learning_rate": 0.0015,
      "loss": 1.4282,
      "step": 5910
    },
    {
      "epoch": 0.6244725738396625,
      "grad_norm": 0.7341051697731018,
      "learning_rate": 0.0015,
      "loss": 1.4274,
      "step": 5920
    },
    {
      "epoch": 0.6255274261603375,
      "grad_norm": 0.8130826950073242,
      "learning_rate": 0.0015,
      "loss": 1.4306,
      "step": 5930
    },
    {
      "epoch": 0.6265822784810127,
      "grad_norm": 0.6295958161354065,
      "learning_rate": 0.0015,
      "loss": 1.4392,
      "step": 5940
    },
    {
      "epoch": 0.6276371308016878,
      "grad_norm": 0.6690378189086914,
      "learning_rate": 0.0015,
      "loss": 1.427,
      "step": 5950
    },
    {
      "epoch": 0.6286919831223629,
      "grad_norm": 0.5623489618301392,
      "learning_rate": 0.0015,
      "loss": 1.4341,
      "step": 5960
    },
    {
      "epoch": 0.629746835443038,
      "grad_norm": 0.5505332946777344,
      "learning_rate": 0.0015,
      "loss": 1.437,
      "step": 5970
    },
    {
      "epoch": 0.630801687763713,
      "grad_norm": 0.5139110684394836,
      "learning_rate": 0.0015,
      "loss": 1.4171,
      "step": 5980
    },
    {
      "epoch": 0.6318565400843882,
      "grad_norm": 0.6092433333396912,
      "learning_rate": 0.0015,
      "loss": 1.438,
      "step": 5990
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.5405699610710144,
      "learning_rate": 0.0015,
      "loss": 1.4483,
      "step": 6000
    },
    {
      "epoch": 0.6339662447257384,
      "grad_norm": 0.5922624468803406,
      "learning_rate": 0.0015,
      "loss": 1.4344,
      "step": 6010
    },
    {
      "epoch": 0.6350210970464135,
      "grad_norm": 0.6210659146308899,
      "learning_rate": 0.0015,
      "loss": 1.4326,
      "step": 6020
    },
    {
      "epoch": 0.6360759493670886,
      "grad_norm": 0.5855004191398621,
      "learning_rate": 0.0015,
      "loss": 1.4157,
      "step": 6030
    },
    {
      "epoch": 0.6371308016877637,
      "grad_norm": 0.5853936076164246,
      "learning_rate": 0.0015,
      "loss": 1.435,
      "step": 6040
    },
    {
      "epoch": 0.6381856540084389,
      "grad_norm": 0.5281463861465454,
      "learning_rate": 0.0015,
      "loss": 1.419,
      "step": 6050
    },
    {
      "epoch": 0.6392405063291139,
      "grad_norm": 0.7227752804756165,
      "learning_rate": 0.0015,
      "loss": 1.4388,
      "step": 6060
    },
    {
      "epoch": 0.640295358649789,
      "grad_norm": 0.7425885200500488,
      "learning_rate": 0.0015,
      "loss": 1.4294,
      "step": 6070
    },
    {
      "epoch": 0.6413502109704642,
      "grad_norm": 0.5315224528312683,
      "learning_rate": 0.0015,
      "loss": 1.4294,
      "step": 6080
    },
    {
      "epoch": 0.6424050632911392,
      "grad_norm": 0.5397734642028809,
      "learning_rate": 0.0015,
      "loss": 1.4278,
      "step": 6090
    },
    {
      "epoch": 0.6434599156118144,
      "grad_norm": 0.5076785087585449,
      "learning_rate": 0.0015,
      "loss": 1.4216,
      "step": 6100
    },
    {
      "epoch": 0.6445147679324894,
      "grad_norm": 0.5564759373664856,
      "learning_rate": 0.0015,
      "loss": 1.4333,
      "step": 6110
    },
    {
      "epoch": 0.6455696202531646,
      "grad_norm": 0.5860690474510193,
      "learning_rate": 0.0015,
      "loss": 1.4296,
      "step": 6120
    },
    {
      "epoch": 0.6466244725738397,
      "grad_norm": 0.5914776921272278,
      "learning_rate": 0.0015,
      "loss": 1.4267,
      "step": 6130
    },
    {
      "epoch": 0.6476793248945147,
      "grad_norm": 0.9994964599609375,
      "learning_rate": 0.0015,
      "loss": 1.4279,
      "step": 6140
    },
    {
      "epoch": 0.6487341772151899,
      "grad_norm": 0.6222503781318665,
      "learning_rate": 0.0015,
      "loss": 1.441,
      "step": 6150
    },
    {
      "epoch": 0.6497890295358649,
      "grad_norm": 0.5461714863777161,
      "learning_rate": 0.0015,
      "loss": 1.4296,
      "step": 6160
    },
    {
      "epoch": 0.6508438818565401,
      "grad_norm": 0.5366678237915039,
      "learning_rate": 0.0014854972418331944,
      "loss": 1.4111,
      "step": 6170
    },
    {
      "epoch": 0.6518987341772152,
      "grad_norm": 0.5969879627227783,
      "learning_rate": 0.0014650219182191931,
      "loss": 1.4339,
      "step": 6180
    },
    {
      "epoch": 0.6529535864978903,
      "grad_norm": 0.5103928446769714,
      "learning_rate": 0.001444828815847542,
      "loss": 1.4352,
      "step": 6190
    },
    {
      "epoch": 0.6540084388185654,
      "grad_norm": 0.6483280658721924,
      "learning_rate": 0.0014249140447269945,
      "loss": 1.4254,
      "step": 6200
    },
    {
      "epoch": 0.6550632911392406,
      "grad_norm": 0.5745824575424194,
      "learning_rate": 0.0014052737684839257,
      "loss": 1.422,
      "step": 6210
    },
    {
      "epoch": 0.6561181434599156,
      "grad_norm": 0.6464072465896606,
      "learning_rate": 0.0013859042036232954,
      "loss": 1.4291,
      "step": 6220
    },
    {
      "epoch": 0.6571729957805907,
      "grad_norm": 0.9690040946006775,
      "learning_rate": 0.001366801618799797,
      "loss": 1.416,
      "step": 6230
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.6810333132743835,
      "learning_rate": 0.001347962334099052,
      "loss": 1.4176,
      "step": 6240
    },
    {
      "epoch": 0.6592827004219409,
      "grad_norm": 0.4878316819667816,
      "learning_rate": 0.0013293827203287143,
      "loss": 1.4245,
      "step": 6250
    },
    {
      "epoch": 0.6603375527426161,
      "grad_norm": 0.5093593001365662,
      "learning_rate": 0.0013110591983193423,
      "loss": 1.4125,
      "step": 6260
    },
    {
      "epoch": 0.6613924050632911,
      "grad_norm": 0.6860863566398621,
      "learning_rate": 0.0012929882382349102,
      "loss": 1.4218,
      "step": 6270
    },
    {
      "epoch": 0.6624472573839663,
      "grad_norm": 0.5376235842704773,
      "learning_rate": 0.0012751663588928214,
      "loss": 1.417,
      "step": 6280
    },
    {
      "epoch": 0.6635021097046413,
      "grad_norm": 0.5228108763694763,
      "learning_rate": 0.0012575901270932943,
      "loss": 1.4114,
      "step": 6290
    },
    {
      "epoch": 0.6645569620253164,
      "grad_norm": 0.6319876313209534,
      "learning_rate": 0.0012402561569579936,
      "loss": 1.4056,
      "step": 6300
    },
    {
      "epoch": 0.6656118143459916,
      "grad_norm": 0.7265379428863525,
      "learning_rate": 0.0012231611092777745,
      "loss": 1.4061,
      "step": 6310
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.5666329264640808,
      "learning_rate": 0.0012063016908694193,
      "loss": 1.3991,
      "step": 6320
    },
    {
      "epoch": 0.6677215189873418,
      "grad_norm": 0.546264111995697,
      "learning_rate": 0.0011896746539412405,
      "loss": 1.4098,
      "step": 6330
    },
    {
      "epoch": 0.6687763713080169,
      "grad_norm": 0.5310012102127075,
      "learning_rate": 0.0011732767954674265,
      "loss": 1.4009,
      "step": 6340
    },
    {
      "epoch": 0.669831223628692,
      "grad_norm": 0.6444507241249084,
      "learning_rate": 0.0011571049565710122,
      "loss": 1.4051,
      "step": 6350
    },
    {
      "epoch": 0.6708860759493671,
      "grad_norm": 0.7691652774810791,
      "learning_rate": 0.001141156021915355,
      "loss": 1.4021,
      "step": 6360
    },
    {
      "epoch": 0.6719409282700421,
      "grad_norm": 0.5308224558830261,
      "learning_rate": 0.001125426919103997,
      "loss": 1.3832,
      "step": 6370
    },
    {
      "epoch": 0.6729957805907173,
      "grad_norm": 0.49382564425468445,
      "learning_rate": 0.001109914618088799,
      "loss": 1.3857,
      "step": 6380
    },
    {
      "epoch": 0.6740506329113924,
      "grad_norm": 0.7249054312705994,
      "learning_rate": 0.0010946161305862348,
      "loss": 1.3995,
      "step": 6390
    },
    {
      "epoch": 0.6751054852320675,
      "grad_norm": 0.7635771036148071,
      "learning_rate": 0.001079528509501728,
      "loss": 1.3984,
      "step": 6400
    },
    {
      "epoch": 0.6761603375527426,
      "grad_norm": 0.5237568020820618,
      "learning_rate": 0.0010646488483619261,
      "loss": 1.3913,
      "step": 6410
    },
    {
      "epoch": 0.6772151898734177,
      "grad_norm": 0.5200632810592651,
      "learning_rate": 0.0010499742807547976,
      "loss": 1.3801,
      "step": 6420
    },
    {
      "epoch": 0.6782700421940928,
      "grad_norm": 0.4805111587047577,
      "learning_rate": 0.0010355019797774478,
      "loss": 1.3887,
      "step": 6430
    },
    {
      "epoch": 0.679324894514768,
      "grad_norm": 0.6184947490692139,
      "learning_rate": 0.001021229157491546,
      "loss": 1.3747,
      "step": 6440
    },
    {
      "epoch": 0.680379746835443,
      "grad_norm": 0.5290433764457703,
      "learning_rate": 0.0010071530643862578,
      "loss": 1.3779,
      "step": 6450
    },
    {
      "epoch": 0.6814345991561181,
      "grad_norm": 0.4848653972148895,
      "learning_rate": 0.000993270988848579,
      "loss": 1.3894,
      "step": 6460
    },
    {
      "epoch": 0.6824894514767933,
      "grad_norm": 0.5046072602272034,
      "learning_rate": 0.0009795802566409742,
      "loss": 1.3842,
      "step": 6470
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.6274845004081726,
      "learning_rate": 0.0009660782303862109,
      "loss": 1.3926,
      "step": 6480
    },
    {
      "epoch": 0.6845991561181435,
      "grad_norm": 0.49987682700157166,
      "learning_rate": 0.0009527623090592963,
      "loss": 1.3827,
      "step": 6490
    },
    {
      "epoch": 0.6856540084388185,
      "grad_norm": 0.5083965063095093,
      "learning_rate": 0.0009396299274864177,
      "loss": 1.398,
      "step": 6500
    },
    {
      "epoch": 0.6867088607594937,
      "grad_norm": 0.49454379081726074,
      "learning_rate": 0.0009266785558507877,
      "loss": 1.3962,
      "step": 6510
    },
    {
      "epoch": 0.6877637130801688,
      "grad_norm": 0.5040996074676514,
      "learning_rate": 0.0009139056992053016,
      "loss": 1.3786,
      "step": 6520
    },
    {
      "epoch": 0.6888185654008439,
      "grad_norm": 0.667354941368103,
      "learning_rate": 0.000901308896991912,
      "loss": 1.3717,
      "step": 6530
    },
    {
      "epoch": 0.689873417721519,
      "grad_norm": 0.5207562446594238,
      "learning_rate": 0.000888885722567627,
      "loss": 1.3789,
      "step": 6540
    },
    {
      "epoch": 0.6909282700421941,
      "grad_norm": 0.519875705242157,
      "learning_rate": 0.0008766337827370438,
      "loss": 1.3727,
      "step": 6550
    },
    {
      "epoch": 0.6919831223628692,
      "grad_norm": 0.7908602952957153,
      "learning_rate": 0.000864550717291324,
      "loss": 1.3749,
      "step": 6560
    },
    {
      "epoch": 0.6930379746835443,
      "grad_norm": 0.762065589427948,
      "learning_rate": 0.0008526341985535229,
      "loss": 1.3702,
      "step": 6570
    },
    {
      "epoch": 0.6940928270042194,
      "grad_norm": 0.5344306230545044,
      "learning_rate": 0.0008408819309301891,
      "loss": 1.3655,
      "step": 6580
    },
    {
      "epoch": 0.6951476793248945,
      "grad_norm": 0.6269177198410034,
      "learning_rate": 0.0008292916504691397,
      "loss": 1.3701,
      "step": 6590
    },
    {
      "epoch": 0.6962025316455697,
      "grad_norm": 0.4671165347099304,
      "learning_rate": 0.0008178611244233354,
      "loss": 1.3761,
      "step": 6600
    },
    {
      "epoch": 0.6972573839662447,
      "grad_norm": 0.5067280530929565,
      "learning_rate": 0.0008065881508207637,
      "loss": 1.3606,
      "step": 6610
    },
    {
      "epoch": 0.6983122362869199,
      "grad_norm": 0.4908502995967865,
      "learning_rate": 0.0007954705580402523,
      "loss": 1.3697,
      "step": 6620
    },
    {
      "epoch": 0.6993670886075949,
      "grad_norm": 0.49855223298072815,
      "learning_rate": 0.0007845062043931298,
      "loss": 1.36,
      "step": 6630
    },
    {
      "epoch": 0.70042194092827,
      "grad_norm": 0.5829418897628784,
      "learning_rate": 0.0007736929777106497,
      "loss": 1.3663,
      "step": 6640
    },
    {
      "epoch": 0.7014767932489452,
      "grad_norm": 0.5269045233726501,
      "learning_rate": 0.000763028794937105,
      "loss": 1.3586,
      "step": 6650
    },
    {
      "epoch": 0.7025316455696202,
      "grad_norm": 0.46957603096961975,
      "learning_rate": 0.0007525116017285476,
      "loss": 1.3576,
      "step": 6660
    },
    {
      "epoch": 0.7035864978902954,
      "grad_norm": 0.4709470272064209,
      "learning_rate": 0.0007421393720570417,
      "loss": 1.3605,
      "step": 6670
    },
    {
      "epoch": 0.7046413502109705,
      "grad_norm": 0.4660401940345764,
      "learning_rate": 0.0007319101078203694,
      "loss": 1.3594,
      "step": 6680
    },
    {
      "epoch": 0.7056962025316456,
      "grad_norm": 0.6550858616828918,
      "learning_rate": 0.0007218218384571178,
      "loss": 1.3543,
      "step": 6690
    },
    {
      "epoch": 0.7067510548523207,
      "grad_norm": 0.55482017993927,
      "learning_rate": 0.0007118726205670703,
      "loss": 1.355,
      "step": 6700
    },
    {
      "epoch": 0.7078059071729957,
      "grad_norm": 0.5752652883529663,
      "learning_rate": 0.0007020605375368316,
      "loss": 1.3582,
      "step": 6710
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.6009365320205688,
      "learning_rate": 0.000692383699170611,
      "loss": 1.3487,
      "step": 6720
    },
    {
      "epoch": 0.709915611814346,
      "grad_norm": 0.508026659488678,
      "learning_rate": 0.0006828402413260966,
      "loss": 1.3564,
      "step": 6730
    },
    {
      "epoch": 0.7109704641350211,
      "grad_norm": 0.5703988075256348,
      "learning_rate": 0.0006734283255553471,
      "loss": 1.3563,
      "step": 6740
    },
    {
      "epoch": 0.7120253164556962,
      "grad_norm": 0.4827798008918762,
      "learning_rate": 0.0006641461387506347,
      "loss": 1.3582,
      "step": 6750
    },
    {
      "epoch": 0.7130801687763713,
      "grad_norm": 0.5830769538879395,
      "learning_rate": 0.0006549918927951678,
      "loss": 1.3562,
      "step": 6760
    },
    {
      "epoch": 0.7141350210970464,
      "grad_norm": 0.5382652878761292,
      "learning_rate": 0.0006459638242186297,
      "loss": 1.3544,
      "step": 6770
    },
    {
      "epoch": 0.7151898734177216,
      "grad_norm": 0.4844338893890381,
      "learning_rate": 0.0006370601938574639,
      "loss": 1.3535,
      "step": 6780
    },
    {
      "epoch": 0.7162447257383966,
      "grad_norm": 0.5328819751739502,
      "learning_rate": 0.0006282792865198421,
      "loss": 1.3573,
      "step": 6790
    },
    {
      "epoch": 0.7172995780590717,
      "grad_norm": 0.48490241169929504,
      "learning_rate": 0.0006196194106552512,
      "loss": 1.3444,
      "step": 6800
    },
    {
      "epoch": 0.7183544303797469,
      "grad_norm": 0.5487757921218872,
      "learning_rate": 0.0006110788980286328,
      "loss": 1.3464,
      "step": 6810
    },
    {
      "epoch": 0.7194092827004219,
      "grad_norm": 0.5482735633850098,
      "learning_rate": 0.0006026561033990158,
      "loss": 1.3486,
      "step": 6820
    },
    {
      "epoch": 0.7204641350210971,
      "grad_norm": 0.5385942459106445,
      "learning_rate": 0.000594349404202577,
      "loss": 1.3443,
      "step": 6830
    },
    {
      "epoch": 0.7215189873417721,
      "grad_norm": 0.6625657081604004,
      "learning_rate": 0.0005861572002400716,
      "loss": 1.3345,
      "step": 6840
    },
    {
      "epoch": 0.7225738396624473,
      "grad_norm": 0.5083481073379517,
      "learning_rate": 0.0005780779133685717,
      "loss": 1.343,
      "step": 6850
    },
    {
      "epoch": 0.7236286919831224,
      "grad_norm": 0.5370842218399048,
      "learning_rate": 0.0005701099871974524,
      "loss": 1.3387,
      "step": 6860
    },
    {
      "epoch": 0.7246835443037974,
      "grad_norm": 0.4709780812263489,
      "learning_rate": 0.0005622518867885708,
      "loss": 1.3348,
      "step": 6870
    },
    {
      "epoch": 0.7257383966244726,
      "grad_norm": 0.5011778473854065,
      "learning_rate": 0.0005545020983605748,
      "loss": 1.3441,
      "step": 6880
    },
    {
      "epoch": 0.7267932489451476,
      "grad_norm": 0.5303826928138733,
      "learning_rate": 0.0005468591289972898,
      "loss": 1.3316,
      "step": 6890
    },
    {
      "epoch": 0.7278481012658228,
      "grad_norm": 0.48876091837882996,
      "learning_rate": 0.0005393215063601232,
      "loss": 1.3325,
      "step": 6900
    },
    {
      "epoch": 0.7289029535864979,
      "grad_norm": 0.5018401145935059,
      "learning_rate": 0.0005318877784044343,
      "loss": 1.356,
      "step": 6910
    },
    {
      "epoch": 0.729957805907173,
      "grad_norm": 0.5491061806678772,
      "learning_rate": 0.0005245565130998126,
      "loss": 1.3361,
      "step": 6920
    },
    {
      "epoch": 0.7310126582278481,
      "grad_norm": 0.4945247769355774,
      "learning_rate": 0.000517326298154212,
      "loss": 1.3428,
      "step": 6930
    },
    {
      "epoch": 0.7320675105485233,
      "grad_norm": 0.5948100686073303,
      "learning_rate": 0.0005101957407418877,
      "loss": 1.3448,
      "step": 6940
    },
    {
      "epoch": 0.7331223628691983,
      "grad_norm": 0.4961825907230377,
      "learning_rate": 0.0005031634672350829,
      "loss": 1.3417,
      "step": 6950
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.5309960246086121,
      "learning_rate": 0.0004962281229394129,
      "loss": 1.3315,
      "step": 6960
    },
    {
      "epoch": 0.7352320675105485,
      "grad_norm": 0.4714023470878601,
      "learning_rate": 0.0004893883718328983,
      "loss": 1.3264,
      "step": 6970
    },
    {
      "epoch": 0.7362869198312236,
      "grad_norm": 0.5240657329559326,
      "learning_rate": 0.0004826428963085938,
      "loss": 1.3293,
      "step": 6980
    },
    {
      "epoch": 0.7373417721518988,
      "grad_norm": 0.5004906058311462,
      "learning_rate": 0.00047599039692076457,
      "loss": 1.3311,
      "step": 6990
    },
    {
      "epoch": 0.7383966244725738,
      "grad_norm": 0.4975900948047638,
      "learning_rate": 0.0004694295921345622,
      "loss": 1.3408,
      "step": 7000
    },
    {
      "epoch": 0.739451476793249,
      "grad_norm": 0.505084216594696,
      "learning_rate": 0.00046295921807915015,
      "loss": 1.3135,
      "step": 7010
    },
    {
      "epoch": 0.740506329113924,
      "grad_norm": 0.6532091498374939,
      "learning_rate": 0.00045657802830423164,
      "loss": 1.3279,
      "step": 7020
    },
    {
      "epoch": 0.7415611814345991,
      "grad_norm": 0.5637933611869812,
      "learning_rate": 0.00045028479353993473,
      "loss": 1.3303,
      "step": 7030
    },
    {
      "epoch": 0.7426160337552743,
      "grad_norm": 0.691018283367157,
      "learning_rate": 0.00044407830146000587,
      "loss": 1.3362,
      "step": 7040
    },
    {
      "epoch": 0.7436708860759493,
      "grad_norm": 0.569522500038147,
      "learning_rate": 0.0004379573564482676,
      "loss": 1.3213,
      "step": 7050
    },
    {
      "epoch": 0.7447257383966245,
      "grad_norm": 0.5943568348884583,
      "learning_rate": 0.0004319207793682963,
      "loss": 1.329,
      "step": 7060
    },
    {
      "epoch": 0.7457805907172996,
      "grad_norm": 0.6255069375038147,
      "learning_rate": 0.0004259674073362731,
      "loss": 1.335,
      "step": 7070
    },
    {
      "epoch": 0.7468354430379747,
      "grad_norm": 0.5183704495429993,
      "learning_rate": 0.00042009609349696626,
      "loss": 1.3295,
      "step": 7080
    },
    {
      "epoch": 0.7478902953586498,
      "grad_norm": 0.4704875946044922,
      "learning_rate": 0.00041430570680280233,
      "loss": 1.3277,
      "step": 7090
    },
    {
      "epoch": 0.7489451476793249,
      "grad_norm": 0.5074374079704285,
      "learning_rate": 0.0004085951317959809,
      "loss": 1.3227,
      "step": 7100
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5200400948524475,
      "learning_rate": 0.00040296326839359315,
      "loss": 1.332,
      "step": 7110
    },
    {
      "epoch": 0.7510548523206751,
      "grad_norm": 0.6528633832931519,
      "learning_rate": 0.000397409031675703,
      "loss": 1.3265,
      "step": 7120
    },
    {
      "epoch": 0.7521097046413502,
      "grad_norm": 0.49102628231048584,
      "learning_rate": 0.00039193135167634786,
      "loss": 1.335,
      "step": 7130
    },
    {
      "epoch": 0.7531645569620253,
      "grad_norm": 0.5010769367218018,
      "learning_rate": 0.00038652917317742123,
      "loss": 1.318,
      "step": 7140
    },
    {
      "epoch": 0.7542194092827004,
      "grad_norm": 0.5290884375572205,
      "learning_rate": 0.0003812014555053956,
      "loss": 1.3173,
      "step": 7150
    },
    {
      "epoch": 0.7552742616033755,
      "grad_norm": 0.46309688687324524,
      "learning_rate": 0.00037594717233084774,
      "loss": 1.3241,
      "step": 7160
    },
    {
      "epoch": 0.7563291139240507,
      "grad_norm": 0.5092751979827881,
      "learning_rate": 0.0003707653114707471,
      "loss": 1.3304,
      "step": 7170
    },
    {
      "epoch": 0.7573839662447257,
      "grad_norm": 0.5297157764434814,
      "learning_rate": 0.00036565487469346906,
      "loss": 1.3099,
      "step": 7180
    },
    {
      "epoch": 0.7584388185654009,
      "grad_norm": 0.47773805260658264,
      "learning_rate": 0.0003606148775264958,
      "loss": 1.3127,
      "step": 7190
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.5312460660934448,
      "learning_rate": 0.0003556443490667684,
      "loss": 1.3177,
      "step": 7200
    },
    {
      "epoch": 0.760548523206751,
      "grad_norm": 0.6486009359359741,
      "learning_rate": 0.0003507423317936521,
      "loss": 1.321,
      "step": 7210
    },
    {
      "epoch": 0.7616033755274262,
      "grad_norm": 0.47344207763671875,
      "learning_rate": 0.00034590788138448006,
      "loss": 1.335,
      "step": 7220
    },
    {
      "epoch": 0.7626582278481012,
      "grad_norm": 0.47290611267089844,
      "learning_rate": 0.0003411400665326393,
      "loss": 1.3292,
      "step": 7230
    },
    {
      "epoch": 0.7637130801687764,
      "grad_norm": 0.4823418855667114,
      "learning_rate": 0.00033643796876816424,
      "loss": 1.3247,
      "step": 7240
    },
    {
      "epoch": 0.7647679324894515,
      "grad_norm": 0.5188544392585754,
      "learning_rate": 0.000331800682280803,
      "loss": 1.3189,
      "step": 7250
    },
    {
      "epoch": 0.7658227848101266,
      "grad_norm": 0.46605202555656433,
      "learning_rate": 0.0003272273137455226,
      "loss": 1.3199,
      "step": 7260
    },
    {
      "epoch": 0.7668776371308017,
      "grad_norm": 0.5110488533973694,
      "learning_rate": 0.00032271698215041863,
      "loss": 1.3141,
      "step": 7270
    },
    {
      "epoch": 0.7679324894514767,
      "grad_norm": 0.4910179078578949,
      "learning_rate": 0.0003182688186269984,
      "loss": 1.3187,
      "step": 7280
    },
    {
      "epoch": 0.7689873417721519,
      "grad_norm": 0.5152925252914429,
      "learning_rate": 0.0003138819662828017,
      "loss": 1.326,
      "step": 7290
    },
    {
      "epoch": 0.770042194092827,
      "grad_norm": 0.4966948926448822,
      "learning_rate": 0.00030955558003632966,
      "loss": 1.3289,
      "step": 7300
    },
    {
      "epoch": 0.7710970464135021,
      "grad_norm": 0.4542575776576996,
      "learning_rate": 0.0003052888264542483,
      "loss": 1.3136,
      "step": 7310
    },
    {
      "epoch": 0.7721518987341772,
      "grad_norm": 0.5378862023353577,
      "learning_rate": 0.0003010808835908368,
      "loss": 1.3061,
      "step": 7320
    },
    {
      "epoch": 0.7732067510548524,
      "grad_norm": 0.4870184361934662,
      "learning_rate": 0.00029693094082964785,
      "loss": 1.3139,
      "step": 7330
    },
    {
      "epoch": 0.7742616033755274,
      "grad_norm": 0.5277137160301208,
      "learning_rate": 0.0002928381987273508,
      "loss": 1.3095,
      "step": 7340
    },
    {
      "epoch": 0.7753164556962026,
      "grad_norm": 0.4717441499233246,
      "learning_rate": 0.0002888018688597272,
      "loss": 1.3065,
      "step": 7350
    },
    {
      "epoch": 0.7763713080168776,
      "grad_norm": 0.49216166138648987,
      "learning_rate": 0.0002848211736697894,
      "loss": 1.3147,
      "step": 7360
    },
    {
      "epoch": 0.7774261603375527,
      "grad_norm": 0.5425299406051636,
      "learning_rate": 0.00028089534631799183,
      "loss": 1.3152,
      "step": 7370
    },
    {
      "epoch": 0.7784810126582279,
      "grad_norm": 0.45568835735321045,
      "learning_rate": 0.0002770236305345076,
      "loss": 1.3028,
      "step": 7380
    },
    {
      "epoch": 0.7795358649789029,
      "grad_norm": 0.47306162118911743,
      "learning_rate": 0.00027320528047354093,
      "loss": 1.3118,
      "step": 7390
    },
    {
      "epoch": 0.7805907172995781,
      "grad_norm": 0.480543315410614,
      "learning_rate": 0.00026943956056964773,
      "loss": 1.3191,
      "step": 7400
    },
    {
      "epoch": 0.7816455696202531,
      "grad_norm": 0.48114633560180664,
      "learning_rate": 0.0002657257453960364,
      "loss": 1.2981,
      "step": 7410
    },
    {
      "epoch": 0.7827004219409283,
      "grad_norm": 0.49501311779022217,
      "learning_rate": 0.0002620631195248222,
      "loss": 1.3052,
      "step": 7420
    },
    {
      "epoch": 0.7837552742616034,
      "grad_norm": 0.48266351222991943,
      "learning_rate": 0.00025845097738920735,
      "loss": 1.2905,
      "step": 7430
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.5152450799942017,
      "learning_rate": 0.0002548886231475606,
      "loss": 1.2963,
      "step": 7440
    },
    {
      "epoch": 0.7858649789029536,
      "grad_norm": 0.4615631401538849,
      "learning_rate": 0.0002513753705493713,
      "loss": 1.2994,
      "step": 7450
    },
    {
      "epoch": 0.7869198312236287,
      "grad_norm": 0.47058191895484924,
      "learning_rate": 0.0002479105428030497,
      "loss": 1.3083,
      "step": 7460
    },
    {
      "epoch": 0.7879746835443038,
      "grad_norm": 0.4936390221118927,
      "learning_rate": 0.00024449347244555043,
      "loss": 1.3,
      "step": 7470
    },
    {
      "epoch": 0.7890295358649789,
      "grad_norm": 0.4663834869861603,
      "learning_rate": 0.00024112350121379254,
      "loss": 1.3077,
      "step": 7480
    },
    {
      "epoch": 0.790084388185654,
      "grad_norm": 0.464754581451416,
      "learning_rate": 0.000237799979917852,
      "loss": 1.3138,
      "step": 7490
    },
    {
      "epoch": 0.7911392405063291,
      "grad_norm": 0.49306800961494446,
      "learning_rate": 0.00023452226831590227,
      "loss": 1.3078,
      "step": 7500
    },
    {
      "epoch": 0.7921940928270043,
      "grad_norm": 0.5315954685211182,
      "learning_rate": 0.00023128973499087779,
      "loss": 1.3156,
      "step": 7510
    },
    {
      "epoch": 0.7932489451476793,
      "grad_norm": 0.45483651757240295,
      "learning_rate": 0.00022810175722883858,
      "loss": 1.2992,
      "step": 7520
    },
    {
      "epoch": 0.7943037974683544,
      "grad_norm": 0.5130757093429565,
      "learning_rate": 0.0002249577208990106,
      "loss": 1.3006,
      "step": 7530
    },
    {
      "epoch": 0.7953586497890295,
      "grad_norm": 0.5637947916984558,
      "learning_rate": 0.00022185702033547996,
      "loss": 1.2995,
      "step": 7540
    },
    {
      "epoch": 0.7964135021097046,
      "grad_norm": 0.5045897960662842,
      "learning_rate": 0.00021879905822051756,
      "loss": 1.2965,
      "step": 7550
    },
    {
      "epoch": 0.7974683544303798,
      "grad_norm": 0.5311060547828674,
      "learning_rate": 0.00021578324546951222,
      "loss": 1.2967,
      "step": 7560
    },
    {
      "epoch": 0.7985232067510548,
      "grad_norm": 0.5004249811172485,
      "learning_rate": 0.00021280900111748948,
      "loss": 1.3082,
      "step": 7570
    },
    {
      "epoch": 0.79957805907173,
      "grad_norm": 0.456435889005661,
      "learning_rate": 0.00020987575220719483,
      "loss": 1.303,
      "step": 7580
    },
    {
      "epoch": 0.8006329113924051,
      "grad_norm": 0.46554267406463623,
      "learning_rate": 0.00020698293367871933,
      "loss": 1.2948,
      "step": 7590
    },
    {
      "epoch": 0.8016877637130801,
      "grad_norm": 0.47276362776756287,
      "learning_rate": 0.00020412998826064692,
      "loss": 1.2991,
      "step": 7600
    },
    {
      "epoch": 0.8027426160337553,
      "grad_norm": 0.4799243211746216,
      "learning_rate": 0.00020131636636270178,
      "loss": 1.3118,
      "step": 7610
    },
    {
      "epoch": 0.8037974683544303,
      "grad_norm": 0.44948357343673706,
      "learning_rate": 0.00019854152596987523,
      "loss": 1.2852,
      "step": 7620
    },
    {
      "epoch": 0.8048523206751055,
      "grad_norm": 0.5160753130912781,
      "learning_rate": 0.00019580493253801255,
      "loss": 1.3087,
      "step": 7630
    },
    {
      "epoch": 0.8059071729957806,
      "grad_norm": 0.5136882066726685,
      "learning_rate": 0.00019310605889083838,
      "loss": 1.2994,
      "step": 7640
    },
    {
      "epoch": 0.8069620253164557,
      "grad_norm": 0.4668951630592346,
      "learning_rate": 0.0001904443851184018,
      "loss": 1.315,
      "step": 7650
    },
    {
      "epoch": 0.8080168776371308,
      "grad_norm": 0.4726894795894623,
      "learning_rate": 0.00018781939847692096,
      "loss": 1.2922,
      "step": 7660
    },
    {
      "epoch": 0.8090717299578059,
      "grad_norm": 0.4939499795436859,
      "learning_rate": 0.00018523059329000844,
      "loss": 1.2987,
      "step": 7670
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.5483400225639343,
      "learning_rate": 0.0001826774708512579,
      "loss": 1.3123,
      "step": 7680
    },
    {
      "epoch": 0.8111814345991561,
      "grad_norm": 0.5137965679168701,
      "learning_rate": 0.00018015953932817348,
      "loss": 1.2875,
      "step": 7690
    },
    {
      "epoch": 0.8122362869198312,
      "grad_norm": 0.4608552157878876,
      "learning_rate": 0.00017767631366742332,
      "loss": 1.2954,
      "step": 7700
    },
    {
      "epoch": 0.8132911392405063,
      "grad_norm": 0.5277746915817261,
      "learning_rate": 0.00017522731550139922,
      "loss": 1.2946,
      "step": 7710
    },
    {
      "epoch": 0.8143459915611815,
      "grad_norm": 0.5126060843467712,
      "learning_rate": 0.00017281207305606407,
      "loss": 1.2962,
      "step": 7720
    },
    {
      "epoch": 0.8154008438818565,
      "grad_norm": 0.5055632591247559,
      "learning_rate": 0.00017043012106006926,
      "loss": 1.3032,
      "step": 7730
    },
    {
      "epoch": 0.8164556962025317,
      "grad_norm": 0.4589047431945801,
      "learning_rate": 0.00016808100065512528,
      "loss": 1.3038,
      "step": 7740
    },
    {
      "epoch": 0.8175105485232067,
      "grad_norm": 0.5196783542633057,
      "learning_rate": 0.00016576425930760734,
      "loss": 1.2913,
      "step": 7750
    },
    {
      "epoch": 0.8185654008438819,
      "grad_norm": 0.5244337320327759,
      "learning_rate": 0.00016347945072137934,
      "loss": 1.2892,
      "step": 7760
    },
    {
      "epoch": 0.819620253164557,
      "grad_norm": 0.5240455865859985,
      "learning_rate": 0.00016122613475181977,
      "loss": 1.2988,
      "step": 7770
    },
    {
      "epoch": 0.820675105485232,
      "grad_norm": 0.5089477896690369,
      "learning_rate": 0.00015900387732103232,
      "loss": 1.291,
      "step": 7780
    },
    {
      "epoch": 0.8217299578059072,
      "grad_norm": 0.5250300765037537,
      "learning_rate": 0.00015681225033422526,
      "loss": 1.3059,
      "step": 7790
    },
    {
      "epoch": 0.8227848101265823,
      "grad_norm": 0.44551143050193787,
      "learning_rate": 0.00015465083159724345,
      "loss": 1.2953,
      "step": 7800
    },
    {
      "epoch": 0.8238396624472574,
      "grad_norm": 0.5375047922134399,
      "learning_rate": 0.0001525192047352371,
      "loss": 1.2942,
      "step": 7810
    },
    {
      "epoch": 0.8248945147679325,
      "grad_norm": 0.4749961793422699,
      "learning_rate": 0.00015041695911245136,
      "loss": 1.3014,
      "step": 7820
    },
    {
      "epoch": 0.8259493670886076,
      "grad_norm": 0.4921550154685974,
      "learning_rate": 0.00014834368975312172,
      "loss": 1.2897,
      "step": 7830
    },
    {
      "epoch": 0.8270042194092827,
      "grad_norm": 0.4852724075317383,
      "learning_rate": 0.00014629899726345958,
      "loss": 1.2765,
      "step": 7840
    },
    {
      "epoch": 0.8280590717299579,
      "grad_norm": 0.4684356153011322,
      "learning_rate": 0.00014428248775471316,
      "loss": 1.2851,
      "step": 7850
    },
    {
      "epoch": 0.8291139240506329,
      "grad_norm": 0.4646863341331482,
      "learning_rate": 0.000142293772767289,
      "loss": 1.274,
      "step": 7860
    },
    {
      "epoch": 0.830168776371308,
      "grad_norm": 0.5266587138175964,
      "learning_rate": 0.00014033246919591922,
      "loss": 1.2888,
      "step": 7870
    },
    {
      "epoch": 0.8312236286919831,
      "grad_norm": 0.4876491129398346,
      "learning_rate": 0.00013839819921586025,
      "loss": 1.3102,
      "step": 7880
    },
    {
      "epoch": 0.8322784810126582,
      "grad_norm": 0.49130967259407043,
      "learning_rate": 0.00013649059021010894,
      "loss": 1.305,
      "step": 7890
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.49542373418807983,
      "learning_rate": 0.00013460927469762155,
      "loss": 1.2852,
      "step": 7900
    },
    {
      "epoch": 0.8343881856540084,
      "grad_norm": 0.4820687174797058,
      "learning_rate": 0.00013275389026252255,
      "loss": 1.2938,
      "step": 7910
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.4949972629547119,
      "learning_rate": 0.0001309240794842889,
      "loss": 1.2911,
      "step": 7920
    },
    {
      "epoch": 0.8364978902953587,
      "grad_norm": 0.5055283308029175,
      "learning_rate": 0.00012911948986889664,
      "loss": 1.3017,
      "step": 7930
    },
    {
      "epoch": 0.8375527426160337,
      "grad_norm": 0.4902411997318268,
      "learning_rate": 0.00012733977378091664,
      "loss": 1.2943,
      "step": 7940
    },
    {
      "epoch": 0.8386075949367089,
      "grad_norm": 0.5116524696350098,
      "learning_rate": 0.00012558458837654633,
      "loss": 1.3001,
      "step": 7950
    },
    {
      "epoch": 0.8396624472573839,
      "grad_norm": 0.46216875314712524,
      "learning_rate": 0.00012385359553756422,
      "loss": 1.283,
      "step": 7960
    },
    {
      "epoch": 0.8407172995780591,
      "grad_norm": 0.4756139814853668,
      "learning_rate": 0.0001221464618061951,
      "loss": 1.2841,
      "step": 7970
    },
    {
      "epoch": 0.8417721518987342,
      "grad_norm": 0.4843553602695465,
      "learning_rate": 0.0001204628583208727,
      "loss": 1.2809,
      "step": 7980
    },
    {
      "epoch": 0.8428270042194093,
      "grad_norm": 0.5512832999229431,
      "learning_rate": 0.00011880246075288824,
      "loss": 1.2936,
      "step": 7990
    },
    {
      "epoch": 0.8438818565400844,
      "grad_norm": 0.5163074135780334,
      "learning_rate": 0.00011716494924391148,
      "loss": 1.2894,
      "step": 8000
    },
    {
      "epoch": 0.8449367088607594,
      "grad_norm": 0.47564107179641724,
      "learning_rate": 0.00011555000834437363,
      "loss": 1.2964,
      "step": 8010
    },
    {
      "epoch": 0.8459915611814346,
      "grad_norm": 0.5144618153572083,
      "learning_rate": 0.00011395732695269907,
      "loss": 1.2881,
      "step": 8020
    },
    {
      "epoch": 0.8470464135021097,
      "grad_norm": 0.5242222547531128,
      "learning_rate": 0.00011238659825537507,
      "loss": 1.2879,
      "step": 8030
    },
    {
      "epoch": 0.8481012658227848,
      "grad_norm": 0.46482113003730774,
      "learning_rate": 0.00011083751966784716,
      "loss": 1.3019,
      "step": 8040
    },
    {
      "epoch": 0.8491561181434599,
      "grad_norm": 0.46565577387809753,
      "learning_rate": 0.00010930979277622952,
      "loss": 1.2812,
      "step": 8050
    },
    {
      "epoch": 0.8502109704641351,
      "grad_norm": 0.4554700553417206,
      "learning_rate": 0.00010780312327981853,
      "loss": 1.2898,
      "step": 8060
    },
    {
      "epoch": 0.8512658227848101,
      "grad_norm": 0.46447324752807617,
      "learning_rate": 0.0001063172209343989,
      "loss": 1.2946,
      "step": 8070
    },
    {
      "epoch": 0.8523206751054853,
      "grad_norm": 0.4921494126319885,
      "learning_rate": 0.000104851799496331,
      "loss": 1.2951,
      "step": 8080
    },
    {
      "epoch": 0.8533755274261603,
      "grad_norm": 0.48273053765296936,
      "learning_rate": 0.00010340657666740917,
      "loss": 1.283,
      "step": 8090
    },
    {
      "epoch": 0.8544303797468354,
      "grad_norm": 0.4731471538543701,
      "learning_rate": 0.00010198127404047976,
      "loss": 1.2775,
      "step": 8100
    },
    {
      "epoch": 0.8554852320675106,
      "grad_norm": 0.5514680743217468,
      "learning_rate": 0.00010057561704580898,
      "loss": 1.2962,
      "step": 8110
    },
    {
      "epoch": 0.8565400843881856,
      "grad_norm": 0.47624343633651733,
      "learning_rate": 9.918933489818986e-05,
      "loss": 1.2869,
      "step": 8120
    },
    {
      "epoch": 0.8575949367088608,
      "grad_norm": 0.5039873123168945,
      "learning_rate": 9.782216054477828e-05,
      "loss": 1.2844,
      "step": 8130
    },
    {
      "epoch": 0.8586497890295358,
      "grad_norm": 0.4922039806842804,
      "learning_rate": 9.647383061364803e-05,
      "loss": 1.2843,
      "step": 8140
    },
    {
      "epoch": 0.859704641350211,
      "grad_norm": 0.4668273627758026,
      "learning_rate": 9.514408536305497e-05,
      "loss": 1.2804,
      "step": 8150
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.4708327054977417,
      "learning_rate": 9.383266863140043e-05,
      "loss": 1.2913,
      "step": 8160
    },
    {
      "epoch": 0.8618143459915611,
      "grad_norm": 0.47959965467453003,
      "learning_rate": 9.25393277878844e-05,
      "loss": 1.282,
      "step": 8170
    },
    {
      "epoch": 0.8628691983122363,
      "grad_norm": 0.46921077370643616,
      "learning_rate": 9.126381368383881e-05,
      "loss": 1.288,
      "step": 8180
    },
    {
      "epoch": 0.8639240506329114,
      "grad_norm": 0.5404256582260132,
      "learning_rate": 9.000588060473158e-05,
      "loss": 1.2832,
      "step": 8190
    },
    {
      "epoch": 0.8649789029535865,
      "grad_norm": 0.47403275966644287,
      "learning_rate": 8.876528622283232e-05,
      "loss": 1.2952,
      "step": 8200
    },
    {
      "epoch": 0.8660337552742616,
      "grad_norm": 0.4553932249546051,
      "learning_rate": 8.754179155053052e-05,
      "loss": 1.2874,
      "step": 8210
    },
    {
      "epoch": 0.8670886075949367,
      "grad_norm": 0.47285547852516174,
      "learning_rate": 8.63351608942968e-05,
      "loss": 1.2934,
      "step": 8220
    },
    {
      "epoch": 0.8681434599156118,
      "grad_norm": 0.460549920797348,
      "learning_rate": 8.514516180927926e-05,
      "loss": 1.2891,
      "step": 8230
    },
    {
      "epoch": 0.869198312236287,
      "grad_norm": 0.46874484419822693,
      "learning_rate": 8.397156505452524e-05,
      "loss": 1.2865,
      "step": 8240
    },
    {
      "epoch": 0.870253164556962,
      "grad_norm": 0.4736628830432892,
      "learning_rate": 8.28141445488205e-05,
      "loss": 1.3032,
      "step": 8250
    },
    {
      "epoch": 0.8713080168776371,
      "grad_norm": 0.46797215938568115,
      "learning_rate": 8.167267732713705e-05,
      "loss": 1.2918,
      "step": 8260
    },
    {
      "epoch": 0.8723628691983122,
      "grad_norm": 0.46798843145370483,
      "learning_rate": 8.054694349768114e-05,
      "loss": 1.2877,
      "step": 8270
    },
    {
      "epoch": 0.8734177215189873,
      "grad_norm": 0.4596850872039795,
      "learning_rate": 7.943672619953359e-05,
      "loss": 1.2736,
      "step": 8280
    },
    {
      "epoch": 0.8744725738396625,
      "grad_norm": 0.4654959440231323,
      "learning_rate": 7.834181156087357e-05,
      "loss": 1.2728,
      "step": 8290
    },
    {
      "epoch": 0.8755274261603375,
      "grad_norm": 0.4837932884693146,
      "learning_rate": 7.726198865777852e-05,
      "loss": 1.298,
      "step": 8300
    },
    {
      "epoch": 0.8765822784810127,
      "grad_norm": 0.48890531063079834,
      "learning_rate": 7.61970494735919e-05,
      "loss": 1.2768,
      "step": 8310
    },
    {
      "epoch": 0.8776371308016878,
      "grad_norm": 0.4484415650367737,
      "learning_rate": 7.514678885885086e-05,
      "loss": 1.2956,
      "step": 8320
    },
    {
      "epoch": 0.8786919831223629,
      "grad_norm": 0.45066314935684204,
      "learning_rate": 7.411100449176634e-05,
      "loss": 1.288,
      "step": 8330
    },
    {
      "epoch": 0.879746835443038,
      "grad_norm": 0.470375120639801,
      "learning_rate": 7.308949683924792e-05,
      "loss": 1.2746,
      "step": 8340
    },
    {
      "epoch": 0.880801687763713,
      "grad_norm": 0.48986098170280457,
      "learning_rate": 7.208206911846581e-05,
      "loss": 1.2796,
      "step": 8350
    },
    {
      "epoch": 0.8818565400843882,
      "grad_norm": 0.4666622579097748,
      "learning_rate": 7.10885272589427e-05,
      "loss": 1.2766,
      "step": 8360
    },
    {
      "epoch": 0.8829113924050633,
      "grad_norm": 0.4520988464355469,
      "learning_rate": 7.010867986516811e-05,
      "loss": 1.2841,
      "step": 8370
    },
    {
      "epoch": 0.8839662447257384,
      "grad_norm": 0.48233267664909363,
      "learning_rate": 6.914233817972799e-05,
      "loss": 1.2908,
      "step": 8380
    },
    {
      "epoch": 0.8850210970464135,
      "grad_norm": 0.4731663763523102,
      "learning_rate": 6.818931604694264e-05,
      "loss": 1.2797,
      "step": 8390
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.5251386165618896,
      "learning_rate": 6.724942987700563e-05,
      "loss": 1.2814,
      "step": 8400
    },
    {
      "epoch": 0.8871308016877637,
      "grad_norm": 0.5002477765083313,
      "learning_rate": 6.632249861061733e-05,
      "loss": 1.3002,
      "step": 8410
    },
    {
      "epoch": 0.8881856540084389,
      "grad_norm": 0.4532804787158966,
      "learning_rate": 6.540834368410549e-05,
      "loss": 1.2869,
      "step": 8420
    },
    {
      "epoch": 0.8892405063291139,
      "grad_norm": 0.46717965602874756,
      "learning_rate": 6.4506788995027e-05,
      "loss": 1.2841,
      "step": 8430
    },
    {
      "epoch": 0.890295358649789,
      "grad_norm": 0.4740357995033264,
      "learning_rate": 6.361766086824344e-05,
      "loss": 1.2787,
      "step": 8440
    },
    {
      "epoch": 0.8913502109704642,
      "grad_norm": 0.48714375495910645,
      "learning_rate": 6.274078802246449e-05,
      "loss": 1.288,
      "step": 8450
    },
    {
      "epoch": 0.8924050632911392,
      "grad_norm": 0.48944953083992004,
      "learning_rate": 6.187600153725223e-05,
      "loss": 1.2818,
      "step": 8460
    },
    {
      "epoch": 0.8934599156118144,
      "grad_norm": 0.4496278464794159,
      "learning_rate": 6.1023134820480546e-05,
      "loss": 1.2844,
      "step": 8470
    },
    {
      "epoch": 0.8945147679324894,
      "grad_norm": 0.45108094811439514,
      "learning_rate": 6.0182023576242725e-05,
      "loss": 1.2788,
      "step": 8480
    },
    {
      "epoch": 0.8955696202531646,
      "grad_norm": 0.4812398850917816,
      "learning_rate": 5.9352505773201664e-05,
      "loss": 1.2836,
      "step": 8490
    },
    {
      "epoch": 0.8966244725738397,
      "grad_norm": 0.4708564579486847,
      "learning_rate": 5.8534421613376175e-05,
      "loss": 1.2729,
      "step": 8500
    },
    {
      "epoch": 0.8976793248945147,
      "grad_norm": 0.47171875834465027,
      "learning_rate": 5.772761350135759e-05,
      "loss": 1.2842,
      "step": 8510
    },
    {
      "epoch": 0.8987341772151899,
      "grad_norm": 0.47209882736206055,
      "learning_rate": 5.6931926013950586e-05,
      "loss": 1.2815,
      "step": 8520
    },
    {
      "epoch": 0.8997890295358649,
      "grad_norm": 0.4671887159347534,
      "learning_rate": 5.61472058702326e-05,
      "loss": 1.2951,
      "step": 8530
    },
    {
      "epoch": 0.9008438818565401,
      "grad_norm": 0.46785426139831543,
      "learning_rate": 5.53733019020258e-05,
      "loss": 1.2906,
      "step": 8540
    },
    {
      "epoch": 0.9018987341772152,
      "grad_norm": 0.49801570177078247,
      "learning_rate": 5.4610065024776125e-05,
      "loss": 1.2817,
      "step": 8550
    },
    {
      "epoch": 0.9029535864978903,
      "grad_norm": 0.5166123509407043,
      "learning_rate": 5.38573482088337e-05,
      "loss": 1.2739,
      "step": 8560
    },
    {
      "epoch": 0.9040084388185654,
      "grad_norm": 0.5468111038208008,
      "learning_rate": 5.3115006451129075e-05,
      "loss": 1.2913,
      "step": 8570
    },
    {
      "epoch": 0.9050632911392406,
      "grad_norm": 0.4588988423347473,
      "learning_rate": 5.2382896747239935e-05,
      "loss": 1.2836,
      "step": 8580
    },
    {
      "epoch": 0.9061181434599156,
      "grad_norm": 0.5207972526550293,
      "learning_rate": 5.166087806384275e-05,
      "loss": 1.2896,
      "step": 8590
    },
    {
      "epoch": 0.9071729957805907,
      "grad_norm": 0.4933626353740692,
      "learning_rate": 5.0948811311544186e-05,
      "loss": 1.278,
      "step": 8600
    },
    {
      "epoch": 0.9082278481012658,
      "grad_norm": 0.4750603437423706,
      "learning_rate": 5.024655931808697e-05,
      "loss": 1.2799,
      "step": 8610
    },
    {
      "epoch": 0.9092827004219409,
      "grad_norm": 0.480411171913147,
      "learning_rate": 4.955398680192509e-05,
      "loss": 1.2792,
      "step": 8620
    },
    {
      "epoch": 0.9103375527426161,
      "grad_norm": 0.4989491403102875,
      "learning_rate": 4.887096034616319e-05,
      "loss": 1.2883,
      "step": 8630
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.5354823470115662,
      "learning_rate": 4.819734837285529e-05,
      "loss": 1.2703,
      "step": 8640
    },
    {
      "epoch": 0.9124472573839663,
      "grad_norm": 0.4586617350578308,
      "learning_rate": 4.7533021117657475e-05,
      "loss": 1.2742,
      "step": 8650
    },
    {
      "epoch": 0.9135021097046413,
      "grad_norm": 0.49164167046546936,
      "learning_rate": 4.687785060483031e-05,
      "loss": 1.277,
      "step": 8660
    },
    {
      "epoch": 0.9145569620253164,
      "grad_norm": 0.4514996409416199,
      "learning_rate": 4.623171062258557e-05,
      "loss": 1.2834,
      "step": 8670
    },
    {
      "epoch": 0.9156118143459916,
      "grad_norm": 0.4464404284954071,
      "learning_rate": 4.559447669877288e-05,
      "loss": 1.2895,
      "step": 8680
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.4650578200817108,
      "learning_rate": 4.496602607690141e-05,
      "loss": 1.2868,
      "step": 8690
    },
    {
      "epoch": 0.9177215189873418,
      "grad_norm": 0.48441407084465027,
      "learning_rate": 4.434623769249217e-05,
      "loss": 1.282,
      "step": 8700
    },
    {
      "epoch": 0.9187763713080169,
      "grad_norm": 0.46642568707466125,
      "learning_rate": 4.373499214975615e-05,
      "loss": 1.2803,
      "step": 8710
    },
    {
      "epoch": 0.919831223628692,
      "grad_norm": 0.4851605296134949,
      "learning_rate": 4.313217169859397e-05,
      "loss": 1.2756,
      "step": 8720
    },
    {
      "epoch": 0.9208860759493671,
      "grad_norm": 0.469745934009552,
      "learning_rate": 4.253766021191256e-05,
      "loss": 1.2882,
      "step": 8730
    },
    {
      "epoch": 0.9219409282700421,
      "grad_norm": 0.47525298595428467,
      "learning_rate": 4.19513431632545e-05,
      "loss": 1.2776,
      "step": 8740
    },
    {
      "epoch": 0.9229957805907173,
      "grad_norm": 0.5118164420127869,
      "learning_rate": 4.1373107604735626e-05,
      "loss": 1.2749,
      "step": 8750
    },
    {
      "epoch": 0.9240506329113924,
      "grad_norm": 0.5095266103744507,
      "learning_rate": 4.0802842145286876e-05,
      "loss": 1.2656,
      "step": 8760
    },
    {
      "epoch": 0.9251054852320675,
      "grad_norm": 0.4578050673007965,
      "learning_rate": 4.024043692919589e-05,
      "loss": 1.276,
      "step": 8770
    },
    {
      "epoch": 0.9261603375527426,
      "grad_norm": 0.46929845213890076,
      "learning_rate": 3.968578361494449e-05,
      "loss": 1.2813,
      "step": 8780
    },
    {
      "epoch": 0.9272151898734177,
      "grad_norm": 0.4484109580516815,
      "learning_rate": 3.91387753543378e-05,
      "loss": 1.2757,
      "step": 8790
    },
    {
      "epoch": 0.9282700421940928,
      "grad_norm": 0.46053996682167053,
      "learning_rate": 3.859930677192103e-05,
      "loss": 1.2739,
      "step": 8800
    },
    {
      "epoch": 0.929324894514768,
      "grad_norm": 0.4593728184700012,
      "learning_rate": 3.806727394468005e-05,
      "loss": 1.2912,
      "step": 8810
    },
    {
      "epoch": 0.930379746835443,
      "grad_norm": 0.4655931890010834,
      "learning_rate": 3.7542574382021635e-05,
      "loss": 1.2773,
      "step": 8820
    },
    {
      "epoch": 0.9314345991561181,
      "grad_norm": 0.4637735188007355,
      "learning_rate": 3.702510700602975e-05,
      "loss": 1.2937,
      "step": 8830
    },
    {
      "epoch": 0.9324894514767933,
      "grad_norm": 0.47105222940444946,
      "learning_rate": 3.651477213199394e-05,
      "loss": 1.2797,
      "step": 8840
    },
    {
      "epoch": 0.9335443037974683,
      "grad_norm": 0.5045534372329712,
      "learning_rate": 3.601147144920609e-05,
      "loss": 1.2798,
      "step": 8850
    },
    {
      "epoch": 0.9345991561181435,
      "grad_norm": 0.517986536026001,
      "learning_rate": 3.5515108002021946e-05,
      "loss": 1.279,
      "step": 8860
    },
    {
      "epoch": 0.9356540084388185,
      "grad_norm": 0.46878963708877563,
      "learning_rate": 3.502558617118352e-05,
      "loss": 1.2809,
      "step": 8870
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.4535967707633972,
      "learning_rate": 3.454281165539913e-05,
      "loss": 1.2826,
      "step": 8880
    },
    {
      "epoch": 0.9377637130801688,
      "grad_norm": 0.4552432596683502,
      "learning_rate": 3.406669145317717e-05,
      "loss": 1.2841,
      "step": 8890
    },
    {
      "epoch": 0.9388185654008439,
      "grad_norm": 0.46975669264793396,
      "learning_rate": 3.359713384491036e-05,
      "loss": 1.2745,
      "step": 8900
    },
    {
      "epoch": 0.939873417721519,
      "grad_norm": 0.4926530122756958,
      "learning_rate": 3.313404837520694e-05,
      "loss": 1.2733,
      "step": 8910
    },
    {
      "epoch": 0.9409282700421941,
      "grad_norm": 0.43800973892211914,
      "learning_rate": 3.267734583546536e-05,
      "loss": 1.2807,
      "step": 8920
    },
    {
      "epoch": 0.9419831223628692,
      "grad_norm": 0.4557863473892212,
      "learning_rate": 3.222693824668916e-05,
      "loss": 1.2766,
      "step": 8930
    },
    {
      "epoch": 0.9430379746835443,
      "grad_norm": 0.468304306268692,
      "learning_rate": 3.178273884253874e-05,
      "loss": 1.2771,
      "step": 8940
    },
    {
      "epoch": 0.9440928270042194,
      "grad_norm": 0.4833139479160309,
      "learning_rate": 3.134466205261674e-05,
      "loss": 1.2791,
      "step": 8950
    },
    {
      "epoch": 0.9451476793248945,
      "grad_norm": 0.45856285095214844,
      "learning_rate": 3.0912623485983774e-05,
      "loss": 1.2837,
      "step": 8960
    },
    {
      "epoch": 0.9462025316455697,
      "grad_norm": 0.5141430497169495,
      "learning_rate": 3.048653991490141e-05,
      "loss": 1.2798,
      "step": 8970
    },
    {
      "epoch": 0.9472573839662447,
      "grad_norm": 0.5297994613647461,
      "learning_rate": 3.0066329258799184e-05,
      "loss": 1.2774,
      "step": 8980
    },
    {
      "epoch": 0.9483122362869199,
      "grad_norm": 0.4530332684516907,
      "learning_rate": 2.965191056846266e-05,
      "loss": 1.2842,
      "step": 8990
    },
    {
      "epoch": 0.9493670886075949,
      "grad_norm": 0.4460042715072632,
      "learning_rate": 2.9243204010439396e-05,
      "loss": 1.2763,
      "step": 9000
    },
    {
      "epoch": 0.95042194092827,
      "grad_norm": 0.4769385755062103,
      "learning_rate": 2.8840130851659852e-05,
      "loss": 1.2765,
      "step": 9010
    },
    {
      "epoch": 0.9514767932489452,
      "grad_norm": 0.47912001609802246,
      "learning_rate": 2.844261344427029e-05,
      "loss": 1.2743,
      "step": 9020
    },
    {
      "epoch": 0.9525316455696202,
      "grad_norm": 0.5202059745788574,
      "learning_rate": 2.805057521067472e-05,
      "loss": 1.2942,
      "step": 9030
    },
    {
      "epoch": 0.9535864978902954,
      "grad_norm": 0.46697160601615906,
      "learning_rate": 2.766394062878302e-05,
      "loss": 1.2704,
      "step": 9040
    },
    {
      "epoch": 0.9546413502109705,
      "grad_norm": 0.4940732419490814,
      "learning_rate": 2.7282635217462405e-05,
      "loss": 1.2817,
      "step": 9050
    },
    {
      "epoch": 0.9556962025316456,
      "grad_norm": 0.4639584422111511,
      "learning_rate": 2.6906585522189378e-05,
      "loss": 1.2827,
      "step": 9060
    },
    {
      "epoch": 0.9567510548523207,
      "grad_norm": 0.4650475084781647,
      "learning_rate": 2.653571910089951e-05,
      "loss": 1.2828,
      "step": 9070
    },
    {
      "epoch": 0.9578059071729957,
      "grad_norm": 0.4645904302597046,
      "learning_rate": 2.6169964510032243e-05,
      "loss": 1.2831,
      "step": 9080
    },
    {
      "epoch": 0.9588607594936709,
      "grad_norm": 0.47728443145751953,
      "learning_rate": 2.580925129076798e-05,
      "loss": 1.2845,
      "step": 9090
    },
    {
      "epoch": 0.959915611814346,
      "grad_norm": 0.4670153856277466,
      "learning_rate": 2.5453509955454954e-05,
      "loss": 1.2631,
      "step": 9100
    },
    {
      "epoch": 0.9609704641350211,
      "grad_norm": 0.4669053554534912,
      "learning_rate": 2.510267197422317e-05,
      "loss": 1.2765,
      "step": 9110
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.4864523410797119,
      "learning_rate": 2.4756669761782806e-05,
      "loss": 1.2864,
      "step": 9120
    },
    {
      "epoch": 0.9630801687763713,
      "grad_norm": 0.46119776368141174,
      "learning_rate": 2.4415436664404643e-05,
      "loss": 1.2686,
      "step": 9130
    },
    {
      "epoch": 0.9641350210970464,
      "grad_norm": 0.5017858743667603,
      "learning_rate": 2.4078906947079882e-05,
      "loss": 1.2902,
      "step": 9140
    },
    {
      "epoch": 0.9651898734177216,
      "grad_norm": 0.49536940455436707,
      "learning_rate": 2.3747015780857007e-05,
      "loss": 1.2777,
      "step": 9150
    },
    {
      "epoch": 0.9662447257383966,
      "grad_norm": 0.46517708897590637,
      "learning_rate": 2.3419699230353144e-05,
      "loss": 1.2829,
      "step": 9160
    },
    {
      "epoch": 0.9672995780590717,
      "grad_norm": 0.47231581807136536,
      "learning_rate": 2.3096894241437583e-05,
      "loss": 1.2837,
      "step": 9170
    },
    {
      "epoch": 0.9683544303797469,
      "grad_norm": 0.45664721727371216,
      "learning_rate": 2.2778538629085057e-05,
      "loss": 1.2796,
      "step": 9180
    },
    {
      "epoch": 0.9694092827004219,
      "grad_norm": 0.48494887351989746,
      "learning_rate": 2.2464571065396428e-05,
      "loss": 1.2743,
      "step": 9190
    },
    {
      "epoch": 0.9704641350210971,
      "grad_norm": 0.4595504403114319,
      "learning_rate": 2.2154931067784525e-05,
      "loss": 1.2776,
      "step": 9200
    },
    {
      "epoch": 0.9715189873417721,
      "grad_norm": 0.4692482650279999,
      "learning_rate": 2.1849558987322783e-05,
      "loss": 1.2775,
      "step": 9210
    },
    {
      "epoch": 0.9725738396624473,
      "grad_norm": 0.4805602431297302,
      "learning_rate": 2.1548395997254516e-05,
      "loss": 1.2744,
      "step": 9220
    },
    {
      "epoch": 0.9736286919831224,
      "grad_norm": 0.44999340176582336,
      "learning_rate": 2.1251384081660546e-05,
      "loss": 1.2674,
      "step": 9230
    },
    {
      "epoch": 0.9746835443037974,
      "grad_norm": 0.45670419931411743,
      "learning_rate": 2.0958466024283035e-05,
      "loss": 1.261,
      "step": 9240
    },
    {
      "epoch": 0.9757383966244726,
      "grad_norm": 0.4626820981502533,
      "learning_rate": 2.0669585397503362e-05,
      "loss": 1.2735,
      "step": 9250
    },
    {
      "epoch": 0.9767932489451476,
      "grad_norm": 0.4641728699207306,
      "learning_rate": 2.0384686551471954e-05,
      "loss": 1.2683,
      "step": 9260
    },
    {
      "epoch": 0.9778481012658228,
      "grad_norm": 0.45692700147628784,
      "learning_rate": 2.0103714603387898e-05,
      "loss": 1.2813,
      "step": 9270
    },
    {
      "epoch": 0.9789029535864979,
      "grad_norm": 0.44312015175819397,
      "learning_rate": 1.9826615426926342e-05,
      "loss": 1.2699,
      "step": 9280
    },
    {
      "epoch": 0.979957805907173,
      "grad_norm": 0.46310457587242126,
      "learning_rate": 1.9553335641811623e-05,
      "loss": 1.276,
      "step": 9290
    },
    {
      "epoch": 0.9810126582278481,
      "grad_norm": 0.47141996026039124,
      "learning_rate": 1.9283822603534143e-05,
      "loss": 1.2746,
      "step": 9300
    },
    {
      "epoch": 0.9820675105485233,
      "grad_norm": 0.48251011967658997,
      "learning_rate": 1.90180243932089e-05,
      "loss": 1.2757,
      "step": 9310
    },
    {
      "epoch": 0.9831223628691983,
      "grad_norm": 0.4870384633541107,
      "learning_rate": 1.8755889807573868e-05,
      "loss": 1.2703,
      "step": 9320
    },
    {
      "epoch": 0.9841772151898734,
      "grad_norm": 0.4840216338634491,
      "learning_rate": 1.8497368349126255e-05,
      "loss": 1.2829,
      "step": 9330
    },
    {
      "epoch": 0.9852320675105485,
      "grad_norm": 0.46854716539382935,
      "learning_rate": 1.824241021639465e-05,
      "loss": 1.282,
      "step": 9340
    },
    {
      "epoch": 0.9862869198312236,
      "grad_norm": 0.46142417192459106,
      "learning_rate": 1.799096629434529e-05,
      "loss": 1.2914,
      "step": 9350
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.4705776274204254,
      "learning_rate": 1.7742988144920578e-05,
      "loss": 1.284,
      "step": 9360
    },
    {
      "epoch": 0.9883966244725738,
      "grad_norm": 0.4654499292373657,
      "learning_rate": 1.7498427997707978e-05,
      "loss": 1.282,
      "step": 9370
    },
    {
      "epoch": 0.989451476793249,
      "grad_norm": 0.46457529067993164,
      "learning_rate": 1.7257238740737548e-05,
      "loss": 1.2679,
      "step": 9380
    },
    {
      "epoch": 0.990506329113924,
      "grad_norm": 0.49462538957595825,
      "learning_rate": 1.7019373911406307e-05,
      "loss": 1.2722,
      "step": 9390
    },
    {
      "epoch": 0.9915611814345991,
      "grad_norm": 0.4786498546600342,
      "learning_rate": 1.67847876875277e-05,
      "loss": 1.2831,
      "step": 9400
    },
    {
      "epoch": 0.9926160337552743,
      "grad_norm": 0.4924539625644684,
      "learning_rate": 1.655343487850443e-05,
      "loss": 1.2585,
      "step": 9410
    },
    {
      "epoch": 0.9936708860759493,
      "grad_norm": 0.4809725284576416,
      "learning_rate": 1.6325270916622947e-05,
      "loss": 1.2792,
      "step": 9420
    },
    {
      "epoch": 0.9947257383966245,
      "grad_norm": 0.4645226299762726,
      "learning_rate": 1.610025184846797e-05,
      "loss": 1.2787,
      "step": 9430
    },
    {
      "epoch": 0.9957805907172996,
      "grad_norm": 0.4642578661441803,
      "learning_rate": 1.587833432645528e-05,
      "loss": 1.2749,
      "step": 9440
    },
    {
      "epoch": 0.9968354430379747,
      "grad_norm": 0.4650630056858063,
      "learning_rate": 1.5659475600481297e-05,
      "loss": 1.2838,
      "step": 9450
    },
    {
      "epoch": 0.9978902953586498,
      "grad_norm": 0.4606262743473053,
      "learning_rate": 1.544363350968769e-05,
      "loss": 1.2703,
      "step": 9460
    },
    {
      "epoch": 0.9989451476793249,
      "grad_norm": 0.4388413727283478,
      "learning_rate": 1.523076647433954e-05,
      "loss": 1.2801,
      "step": 9470
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.3256233930587769,
      "learning_rate": 1.5020833487815421e-05,
      "loss": 1.2745,
      "step": 9480
    }
  ],
  "logging_steps": 10,
  "max_steps": 9480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.832308198648013e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}