{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9836540996572634,
  "eval_steps": 500,
  "global_step": 11193,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008788118463836892,
      "grad_norm": 0.7183610796928406,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 0.5856,
      "step": 10
    },
    {
      "epoch": 0.0017576236927673784,
      "grad_norm": 0.6834720373153687,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 0.5868,
      "step": 20
    },
    {
      "epoch": 0.002636435539151068,
      "grad_norm": 0.6469134092330933,
      "learning_rate": 3.342134831460675e-06,
      "loss": 0.5883,
      "step": 30
    },
    {
      "epoch": 0.003515247385534757,
      "grad_norm": 0.5836713910102844,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.6023,
      "step": 40
    },
    {
      "epoch": 0.004394059231918446,
      "grad_norm": 0.6964038610458374,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.5766,
      "step": 50
    },
    {
      "epoch": 0.005272871078302136,
      "grad_norm": 1.2740495204925537,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.5805,
      "step": 60
    },
    {
      "epoch": 0.006151682924685825,
      "grad_norm": 0.6438924670219421,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.5624,
      "step": 70
    },
    {
      "epoch": 0.007030494771069514,
      "grad_norm": 0.6378642320632935,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.561,
      "step": 80
    },
    {
      "epoch": 0.007909306617453203,
      "grad_norm": 0.6773518323898315,
      "learning_rate": 1.005e-05,
      "loss": 0.5738,
      "step": 90
    },
    {
      "epoch": 0.008788118463836892,
      "grad_norm": 0.5750642418861389,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.565,
      "step": 100
    },
    {
      "epoch": 0.009666930310220583,
      "grad_norm": 0.6147071719169617,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.5776,
      "step": 110
    },
    {
      "epoch": 0.010545742156604272,
      "grad_norm": 0.5244885087013245,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.5569,
      "step": 120
    },
    {
      "epoch": 0.01142455400298796,
      "grad_norm": 0.7714810371398926,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.5731,
      "step": 130
    },
    {
      "epoch": 0.01230336584937165,
      "grad_norm": 0.681827187538147,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.5487,
      "step": 140
    },
    {
      "epoch": 0.013182177695755339,
      "grad_norm": 0.6038880944252014,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.5655,
      "step": 150
    },
    {
      "epoch": 0.014060989542139028,
      "grad_norm": 0.5889508128166199,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.5804,
      "step": 160
    },
    {
      "epoch": 0.014939801388522717,
      "grad_norm": 0.6750523447990417,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.5946,
      "step": 170
    },
    {
      "epoch": 0.015818613234906406,
      "grad_norm": 0.5245691537857056,
      "learning_rate": 2e-05,
      "loss": 0.5724,
      "step": 180
    },
    {
      "epoch": 0.016697425081290095,
      "grad_norm": 0.6065133810043335,
      "learning_rate": 2e-05,
      "loss": 0.5716,
      "step": 190
    },
    {
      "epoch": 0.017576236927673784,
      "grad_norm": 0.5226805806159973,
      "learning_rate": 2e-05,
      "loss": 0.5864,
      "step": 200
    },
    {
      "epoch": 0.018455048774057473,
      "grad_norm": 0.5066975355148315,
      "learning_rate": 2e-05,
      "loss": 0.564,
      "step": 210
    },
    {
      "epoch": 0.019333860620441165,
      "grad_norm": 0.5714482069015503,
      "learning_rate": 2e-05,
      "loss": 0.5555,
      "step": 220
    },
    {
      "epoch": 0.020212672466824854,
      "grad_norm": 0.4541548788547516,
      "learning_rate": 2e-05,
      "loss": 0.5641,
      "step": 230
    },
    {
      "epoch": 0.021091484313208543,
      "grad_norm": 0.4934563934803009,
      "learning_rate": 2e-05,
      "loss": 0.5667,
      "step": 240
    },
    {
      "epoch": 0.021970296159592232,
      "grad_norm": 0.4963897168636322,
      "learning_rate": 2e-05,
      "loss": 0.5721,
      "step": 250
    },
    {
      "epoch": 0.02284910800597592,
      "grad_norm": 0.45243293046951294,
      "learning_rate": 2e-05,
      "loss": 0.5924,
      "step": 260
    },
    {
      "epoch": 0.02372791985235961,
      "grad_norm": 0.4372263550758362,
      "learning_rate": 2e-05,
      "loss": 0.6031,
      "step": 270
    },
    {
      "epoch": 0.0246067316987433,
      "grad_norm": 0.4153901934623718,
      "learning_rate": 2e-05,
      "loss": 0.5682,
      "step": 280
    },
    {
      "epoch": 0.025485543545126988,
      "grad_norm": 0.43816617131233215,
      "learning_rate": 2e-05,
      "loss": 0.5886,
      "step": 290
    },
    {
      "epoch": 0.026364355391510677,
      "grad_norm": 0.4889500141143799,
      "learning_rate": 2e-05,
      "loss": 0.577,
      "step": 300
    },
    {
      "epoch": 0.027243167237894366,
      "grad_norm": 0.38636863231658936,
      "learning_rate": 2e-05,
      "loss": 0.6108,
      "step": 310
    },
    {
      "epoch": 0.028121979084278055,
      "grad_norm": 0.46534135937690735,
      "learning_rate": 2e-05,
      "loss": 0.5867,
      "step": 320
    },
    {
      "epoch": 0.029000790930661744,
      "grad_norm": 0.9078437685966492,
      "learning_rate": 2e-05,
      "loss": 0.5642,
      "step": 330
    },
    {
      "epoch": 0.029879602777045433,
      "grad_norm": 0.5279733538627625,
      "learning_rate": 2e-05,
      "loss": 0.5767,
      "step": 340
    },
    {
      "epoch": 0.030758414623429122,
      "grad_norm": 0.4053654372692108,
      "learning_rate": 2e-05,
      "loss": 0.5714,
      "step": 350
    },
    {
      "epoch": 0.03163722646981281,
      "grad_norm": 0.3655475676059723,
      "learning_rate": 2e-05,
      "loss": 0.5658,
      "step": 360
    },
    {
      "epoch": 0.032516038316196504,
      "grad_norm": 0.42354220151901245,
      "learning_rate": 2e-05,
      "loss": 0.5735,
      "step": 370
    },
    {
      "epoch": 0.03339485016258019,
      "grad_norm": 0.4039507806301117,
      "learning_rate": 2e-05,
      "loss": 0.5713,
      "step": 380
    },
    {
      "epoch": 0.03427366200896388,
      "grad_norm": 0.3883609473705292,
      "learning_rate": 2e-05,
      "loss": 0.5736,
      "step": 390
    },
    {
      "epoch": 0.03515247385534757,
      "grad_norm": 0.43263545632362366,
      "learning_rate": 2e-05,
      "loss": 0.6093,
      "step": 400
    },
    {
      "epoch": 0.03603128570173126,
      "grad_norm": 0.4262617528438568,
      "learning_rate": 2e-05,
      "loss": 0.57,
      "step": 410
    },
    {
      "epoch": 0.036910097548114945,
      "grad_norm": 0.39129018783569336,
      "learning_rate": 2e-05,
      "loss": 0.5798,
      "step": 420
    },
    {
      "epoch": 0.03778890939449864,
      "grad_norm": 0.4544897973537445,
      "learning_rate": 2e-05,
      "loss": 0.5747,
      "step": 430
    },
    {
      "epoch": 0.03866772124088233,
      "grad_norm": 0.4992966651916504,
      "learning_rate": 2e-05,
      "loss": 0.6086,
      "step": 440
    },
    {
      "epoch": 0.039546533087266016,
      "grad_norm": 0.411885142326355,
      "learning_rate": 2e-05,
      "loss": 0.5822,
      "step": 450
    },
    {
      "epoch": 0.04042534493364971,
      "grad_norm": 0.3853900134563446,
      "learning_rate": 2e-05,
      "loss": 0.5449,
      "step": 460
    },
    {
      "epoch": 0.041304156780033394,
      "grad_norm": 0.810398519039154,
      "learning_rate": 2e-05,
      "loss": 0.5477,
      "step": 470
    },
    {
      "epoch": 0.042182968626417086,
      "grad_norm": 0.45868226885795593,
      "learning_rate": 2e-05,
      "loss": 0.5688,
      "step": 480
    },
    {
      "epoch": 0.04306178047280077,
      "grad_norm": 0.378400593996048,
      "learning_rate": 2e-05,
      "loss": 0.5556,
      "step": 490
    },
    {
      "epoch": 0.043940592319184464,
      "grad_norm": 0.37122344970703125,
      "learning_rate": 2e-05,
      "loss": 0.5601,
      "step": 500
    },
    {
      "epoch": 0.04481940416556815,
      "grad_norm": 0.390835165977478,
      "learning_rate": 2e-05,
      "loss": 0.5872,
      "step": 510
    },
    {
      "epoch": 0.04569821601195184,
      "grad_norm": 0.41930919885635376,
      "learning_rate": 2e-05,
      "loss": 0.5501,
      "step": 520
    },
    {
      "epoch": 0.04657702785833553,
      "grad_norm": 0.4113750755786896,
      "learning_rate": 2e-05,
      "loss": 0.5878,
      "step": 530
    },
    {
      "epoch": 0.04745583970471922,
      "grad_norm": 0.6010010838508606,
      "learning_rate": 2e-05,
      "loss": 0.5784,
      "step": 540
    },
    {
      "epoch": 0.048334651551102906,
      "grad_norm": 0.3737349510192871,
      "learning_rate": 2e-05,
      "loss": 0.583,
      "step": 550
    },
    {
      "epoch": 0.0492134633974866,
      "grad_norm": 0.4933265745639801,
      "learning_rate": 2e-05,
      "loss": 0.5771,
      "step": 560
    },
    {
      "epoch": 0.05009227524387029,
      "grad_norm": 0.4226865768432617,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 570
    },
    {
      "epoch": 0.050971087090253976,
      "grad_norm": 0.36813685297966003,
      "learning_rate": 2e-05,
      "loss": 0.5718,
      "step": 580
    },
    {
      "epoch": 0.05184989893663767,
      "grad_norm": 0.3392640948295593,
      "learning_rate": 2e-05,
      "loss": 0.5665,
      "step": 590
    },
    {
      "epoch": 0.052728710783021354,
      "grad_norm": 0.41001757979393005,
      "learning_rate": 2e-05,
      "loss": 0.5783,
      "step": 600
    },
    {
      "epoch": 0.05360752262940505,
      "grad_norm": 0.3661254346370697,
      "learning_rate": 2e-05,
      "loss": 0.5634,
      "step": 610
    },
    {
      "epoch": 0.05448633447578873,
      "grad_norm": 0.532596230506897,
      "learning_rate": 2e-05,
      "loss": 0.5944,
      "step": 620
    },
    {
      "epoch": 0.055365146322172425,
      "grad_norm": 0.3782620429992676,
      "learning_rate": 2e-05,
      "loss": 0.5559,
      "step": 630
    },
    {
      "epoch": 0.05624395816855611,
      "grad_norm": 0.36520808935165405,
      "learning_rate": 2e-05,
      "loss": 0.5781,
      "step": 640
    },
    {
      "epoch": 0.0571227700149398,
      "grad_norm": 0.422725111246109,
      "learning_rate": 2e-05,
      "loss": 0.5887,
      "step": 650
    },
    {
      "epoch": 0.05800158186132349,
      "grad_norm": 0.47393128275871277,
      "learning_rate": 2e-05,
      "loss": 0.5897,
      "step": 660
    },
    {
      "epoch": 0.05888039370770718,
      "grad_norm": 0.37145882844924927,
      "learning_rate": 2e-05,
      "loss": 0.6046,
      "step": 670
    },
    {
      "epoch": 0.059759205554090866,
      "grad_norm": 0.37511953711509705,
      "learning_rate": 2e-05,
      "loss": 0.5616,
      "step": 680
    },
    {
      "epoch": 0.06063801740047456,
      "grad_norm": 0.36933016777038574,
      "learning_rate": 2e-05,
      "loss": 0.575,
      "step": 690
    },
    {
      "epoch": 0.061516829246858244,
      "grad_norm": 0.36136773228645325,
      "learning_rate": 2e-05,
      "loss": 0.5674,
      "step": 700
    },
    {
      "epoch": 0.06239564109324194,
      "grad_norm": 0.387931227684021,
      "learning_rate": 2e-05,
      "loss": 0.552,
      "step": 710
    },
    {
      "epoch": 0.06327445293962562,
      "grad_norm": 0.357898473739624,
      "learning_rate": 2e-05,
      "loss": 0.5733,
      "step": 720
    },
    {
      "epoch": 0.06415326478600931,
      "grad_norm": 0.45435261726379395,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 730
    },
    {
      "epoch": 0.06503207663239301,
      "grad_norm": 0.3749592900276184,
      "learning_rate": 2e-05,
      "loss": 0.5828,
      "step": 740
    },
    {
      "epoch": 0.0659108884787767,
      "grad_norm": 0.552718997001648,
      "learning_rate": 2e-05,
      "loss": 0.5765,
      "step": 750
    },
    {
      "epoch": 0.06678970032516038,
      "grad_norm": 0.3859593868255615,
      "learning_rate": 2e-05,
      "loss": 0.5642,
      "step": 760
    },
    {
      "epoch": 0.06766851217154407,
      "grad_norm": 0.3709048628807068,
      "learning_rate": 2e-05,
      "loss": 0.5861,
      "step": 770
    },
    {
      "epoch": 0.06854732401792776,
      "grad_norm": 0.36113089323043823,
      "learning_rate": 2e-05,
      "loss": 0.5777,
      "step": 780
    },
    {
      "epoch": 0.06942613586431146,
      "grad_norm": 0.355157732963562,
      "learning_rate": 2e-05,
      "loss": 0.5779,
      "step": 790
    },
    {
      "epoch": 0.07030494771069513,
      "grad_norm": 0.34941405057907104,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 800
    },
    {
      "epoch": 0.07118375955707883,
      "grad_norm": 0.37520429491996765,
      "learning_rate": 2e-05,
      "loss": 0.5619,
      "step": 810
    },
    {
      "epoch": 0.07206257140346252,
      "grad_norm": 0.38500913977622986,
      "learning_rate": 2e-05,
      "loss": 0.5532,
      "step": 820
    },
    {
      "epoch": 0.07294138324984621,
      "grad_norm": 0.35824254155158997,
      "learning_rate": 2e-05,
      "loss": 0.5778,
      "step": 830
    },
    {
      "epoch": 0.07382019509622989,
      "grad_norm": 0.3841072618961334,
      "learning_rate": 2e-05,
      "loss": 0.5733,
      "step": 840
    },
    {
      "epoch": 0.07469900694261358,
      "grad_norm": 0.38087451457977295,
      "learning_rate": 2e-05,
      "loss": 0.5752,
      "step": 850
    },
    {
      "epoch": 0.07557781878899728,
      "grad_norm": 0.3607955873012543,
      "learning_rate": 2e-05,
      "loss": 0.5509,
      "step": 860
    },
    {
      "epoch": 0.07645663063538097,
      "grad_norm": 0.36730197072029114,
      "learning_rate": 2e-05,
      "loss": 0.5664,
      "step": 870
    },
    {
      "epoch": 0.07733544248176466,
      "grad_norm": 0.35436132550239563,
      "learning_rate": 2e-05,
      "loss": 0.5599,
      "step": 880
    },
    {
      "epoch": 0.07821425432814834,
      "grad_norm": 0.3620469570159912,
      "learning_rate": 2e-05,
      "loss": 0.5821,
      "step": 890
    },
    {
      "epoch": 0.07909306617453203,
      "grad_norm": 0.3576803505420685,
      "learning_rate": 2e-05,
      "loss": 0.5859,
      "step": 900
    },
    {
      "epoch": 0.07997187802091572,
      "grad_norm": 0.3401060104370117,
      "learning_rate": 2e-05,
      "loss": 0.5649,
      "step": 910
    },
    {
      "epoch": 0.08085068986729942,
      "grad_norm": 0.35253942012786865,
      "learning_rate": 2e-05,
      "loss": 0.5624,
      "step": 920
    },
    {
      "epoch": 0.0817295017136831,
      "grad_norm": 0.3855127692222595,
      "learning_rate": 2e-05,
      "loss": 0.5648,
      "step": 930
    },
    {
      "epoch": 0.08260831356006679,
      "grad_norm": 0.3717329502105713,
      "learning_rate": 2e-05,
      "loss": 0.5602,
      "step": 940
    },
    {
      "epoch": 0.08348712540645048,
      "grad_norm": 0.6942209601402283,
      "learning_rate": 2e-05,
      "loss": 0.5947,
      "step": 950
    },
    {
      "epoch": 0.08436593725283417,
      "grad_norm": 0.39715272188186646,
      "learning_rate": 2e-05,
      "loss": 0.5669,
      "step": 960
    },
    {
      "epoch": 0.08524474909921785,
      "grad_norm": 0.35998284816741943,
      "learning_rate": 2e-05,
      "loss": 0.5785,
      "step": 970
    },
    {
      "epoch": 0.08612356094560154,
      "grad_norm": 0.4870074391365051,
      "learning_rate": 2e-05,
      "loss": 0.6038,
      "step": 980
    },
    {
      "epoch": 0.08700237279198524,
      "grad_norm": 0.35796302556991577,
      "learning_rate": 2e-05,
      "loss": 0.5861,
      "step": 990
    },
    {
      "epoch": 0.08788118463836893,
      "grad_norm": 0.3613901436328888,
      "learning_rate": 2e-05,
      "loss": 0.5718,
      "step": 1000
    },
    {
      "epoch": 0.08875999648475262,
      "grad_norm": 0.3208099603652954,
      "learning_rate": 2e-05,
      "loss": 0.5484,
      "step": 1010
    },
    {
      "epoch": 0.0896388083311363,
      "grad_norm": 0.46393319964408875,
      "learning_rate": 2e-05,
      "loss": 0.5745,
      "step": 1020
    },
    {
      "epoch": 0.09051762017751999,
      "grad_norm": 0.39247334003448486,
      "learning_rate": 2e-05,
      "loss": 0.554,
      "step": 1030
    },
    {
      "epoch": 0.09139643202390368,
      "grad_norm": 0.3445221185684204,
      "learning_rate": 2e-05,
      "loss": 0.5677,
      "step": 1040
    },
    {
      "epoch": 0.09227524387028738,
      "grad_norm": 0.3665354549884796,
      "learning_rate": 2e-05,
      "loss": 0.5512,
      "step": 1050
    },
    {
      "epoch": 0.09315405571667106,
      "grad_norm": 0.4386295974254608,
      "learning_rate": 2e-05,
      "loss": 0.5789,
      "step": 1060
    },
    {
      "epoch": 0.09403286756305475,
      "grad_norm": 0.33483144640922546,
      "learning_rate": 2e-05,
      "loss": 0.5744,
      "step": 1070
    },
    {
      "epoch": 0.09491167940943844,
      "grad_norm": 0.36662590503692627,
      "learning_rate": 2e-05,
      "loss": 0.5612,
      "step": 1080
    },
    {
      "epoch": 0.09579049125582213,
      "grad_norm": 0.3931642770767212,
      "learning_rate": 2e-05,
      "loss": 0.5816,
      "step": 1090
    },
    {
      "epoch": 0.09666930310220581,
      "grad_norm": 0.37220466136932373,
      "learning_rate": 2e-05,
      "loss": 0.5612,
      "step": 1100
    },
    {
      "epoch": 0.0975481149485895,
      "grad_norm": 0.40914490818977356,
      "learning_rate": 2e-05,
      "loss": 0.574,
      "step": 1110
    },
    {
      "epoch": 0.0984269267949732,
      "grad_norm": 0.3663410246372223,
      "learning_rate": 2e-05,
      "loss": 0.5576,
      "step": 1120
    },
    {
      "epoch": 0.09930573864135689,
      "grad_norm": 0.3547806143760681,
      "learning_rate": 2e-05,
      "loss": 0.5779,
      "step": 1130
    },
    {
      "epoch": 0.10018455048774058,
      "grad_norm": 0.3315262794494629,
      "learning_rate": 2e-05,
      "loss": 0.5607,
      "step": 1140
    },
    {
      "epoch": 0.10106336233412426,
      "grad_norm": 0.34696242213249207,
      "learning_rate": 2e-05,
      "loss": 0.5796,
      "step": 1150
    },
    {
      "epoch": 0.10194217418050795,
      "grad_norm": 0.38509389758110046,
      "learning_rate": 2e-05,
      "loss": 0.5713,
      "step": 1160
    },
    {
      "epoch": 0.10282098602689164,
      "grad_norm": 0.37624210119247437,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 1170
    },
    {
      "epoch": 0.10369979787327534,
      "grad_norm": 0.34956589341163635,
      "learning_rate": 2e-05,
      "loss": 0.5578,
      "step": 1180
    },
    {
      "epoch": 0.10457860971965902,
      "grad_norm": 0.35008957982063293,
      "learning_rate": 2e-05,
      "loss": 0.5441,
      "step": 1190
    },
    {
      "epoch": 0.10545742156604271,
      "grad_norm": 0.34270432591438293,
      "learning_rate": 2e-05,
      "loss": 0.5645,
      "step": 1200
    },
    {
      "epoch": 0.1063362334124264,
      "grad_norm": 0.4176119863986969,
      "learning_rate": 2e-05,
      "loss": 0.5866,
      "step": 1210
    },
    {
      "epoch": 0.1072150452588101,
      "grad_norm": 0.4271017014980316,
      "learning_rate": 2e-05,
      "loss": 0.5854,
      "step": 1220
    },
    {
      "epoch": 0.10809385710519377,
      "grad_norm": 0.3339524269104004,
      "learning_rate": 2e-05,
      "loss": 0.5748,
      "step": 1230
    },
    {
      "epoch": 0.10897266895157746,
      "grad_norm": 0.4067085385322571,
      "learning_rate": 2e-05,
      "loss": 0.5779,
      "step": 1240
    },
    {
      "epoch": 0.10985148079796116,
      "grad_norm": 0.3834608197212219,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 1250
    },
    {
      "epoch": 0.11073029264434485,
      "grad_norm": 0.38505619764328003,
      "learning_rate": 2e-05,
      "loss": 0.5785,
      "step": 1260
    },
    {
      "epoch": 0.11160910449072854,
      "grad_norm": 0.3317438066005707,
      "learning_rate": 2e-05,
      "loss": 0.5673,
      "step": 1270
    },
    {
      "epoch": 0.11248791633711222,
      "grad_norm": 0.37222227454185486,
      "learning_rate": 2e-05,
      "loss": 0.5762,
      "step": 1280
    },
    {
      "epoch": 0.11336672818349591,
      "grad_norm": 0.40460169315338135,
      "learning_rate": 2e-05,
      "loss": 0.5747,
      "step": 1290
    },
    {
      "epoch": 0.1142455400298796,
      "grad_norm": 0.3579246997833252,
      "learning_rate": 2e-05,
      "loss": 0.5895,
      "step": 1300
    },
    {
      "epoch": 0.1151243518762633,
      "grad_norm": 0.34237009286880493,
      "learning_rate": 2e-05,
      "loss": 0.5574,
      "step": 1310
    },
    {
      "epoch": 0.11600316372264698,
      "grad_norm": 0.3422780930995941,
      "learning_rate": 2e-05,
      "loss": 0.5913,
      "step": 1320
    },
    {
      "epoch": 0.11688197556903067,
      "grad_norm": 0.3603161573410034,
      "learning_rate": 2e-05,
      "loss": 0.5465,
      "step": 1330
    },
    {
      "epoch": 0.11776078741541436,
      "grad_norm": 0.34054920077323914,
      "learning_rate": 2e-05,
      "loss": 0.5698,
      "step": 1340
    },
    {
      "epoch": 0.11863959926179805,
      "grad_norm": 0.35030534863471985,
      "learning_rate": 2e-05,
      "loss": 0.5953,
      "step": 1350
    },
    {
      "epoch": 0.11951841110818173,
      "grad_norm": 0.4028947353363037,
      "learning_rate": 2e-05,
      "loss": 0.5675,
      "step": 1360
    },
    {
      "epoch": 0.12039722295456542,
      "grad_norm": 0.3338577151298523,
      "learning_rate": 2e-05,
      "loss": 0.5659,
      "step": 1370
    },
    {
      "epoch": 0.12127603480094912,
      "grad_norm": 0.3435846269130707,
      "learning_rate": 2e-05,
      "loss": 0.5833,
      "step": 1380
    },
    {
      "epoch": 0.12215484664733281,
      "grad_norm": 0.6601398587226868,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 1390
    },
    {
      "epoch": 0.12303365849371649,
      "grad_norm": 0.37921780347824097,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 1400
    },
    {
      "epoch": 0.12391247034010018,
      "grad_norm": 0.34458211064338684,
      "learning_rate": 2e-05,
      "loss": 0.5595,
      "step": 1410
    },
    {
      "epoch": 0.12479128218648387,
      "grad_norm": 0.4461982548236847,
      "learning_rate": 2e-05,
      "loss": 0.5616,
      "step": 1420
    },
    {
      "epoch": 0.12567009403286755,
      "grad_norm": 0.4057040512561798,
      "learning_rate": 2e-05,
      "loss": 0.5754,
      "step": 1430
    },
    {
      "epoch": 0.12654890587925124,
      "grad_norm": 0.35918596386909485,
      "learning_rate": 2e-05,
      "loss": 0.5781,
      "step": 1440
    },
    {
      "epoch": 0.12742771772563494,
      "grad_norm": 0.33282655477523804,
      "learning_rate": 2e-05,
      "loss": 0.5608,
      "step": 1450
    },
    {
      "epoch": 0.12830652957201863,
      "grad_norm": 0.4065190553665161,
      "learning_rate": 2e-05,
      "loss": 0.5608,
      "step": 1460
    },
    {
      "epoch": 0.12918534141840232,
      "grad_norm": 0.3624435365200043,
      "learning_rate": 2e-05,
      "loss": 0.5518,
      "step": 1470
    },
    {
      "epoch": 0.13006415326478601,
      "grad_norm": 0.37889572978019714,
      "learning_rate": 2e-05,
      "loss": 0.5649,
      "step": 1480
    },
    {
      "epoch": 0.1309429651111697,
      "grad_norm": 0.39061930775642395,
      "learning_rate": 2e-05,
      "loss": 0.5631,
      "step": 1490
    },
    {
      "epoch": 0.1318217769575534,
      "grad_norm": 0.3646930754184723,
      "learning_rate": 2e-05,
      "loss": 0.5544,
      "step": 1500
    },
    {
      "epoch": 0.13270058880393706,
      "grad_norm": 0.3735094964504242,
      "learning_rate": 2e-05,
      "loss": 0.5566,
      "step": 1510
    },
    {
      "epoch": 0.13357940065032076,
      "grad_norm": 0.32499170303344727,
      "learning_rate": 2e-05,
      "loss": 0.5561,
      "step": 1520
    },
    {
      "epoch": 0.13445821249670445,
      "grad_norm": 0.3363400399684906,
      "learning_rate": 2e-05,
      "loss": 0.5578,
      "step": 1530
    },
    {
      "epoch": 0.13533702434308814,
      "grad_norm": 0.3458295464515686,
      "learning_rate": 2e-05,
      "loss": 0.6095,
      "step": 1540
    },
    {
      "epoch": 0.13621583618947183,
      "grad_norm": 0.3249399662017822,
      "learning_rate": 2e-05,
      "loss": 0.5599,
      "step": 1550
    },
    {
      "epoch": 0.13709464803585553,
      "grad_norm": 0.33804813027381897,
      "learning_rate": 2e-05,
      "loss": 0.5561,
      "step": 1560
    },
    {
      "epoch": 0.13797345988223922,
      "grad_norm": 0.34222960472106934,
      "learning_rate": 2e-05,
      "loss": 0.6034,
      "step": 1570
    },
    {
      "epoch": 0.1388522717286229,
      "grad_norm": 0.33628180623054504,
      "learning_rate": 2e-05,
      "loss": 0.5703,
      "step": 1580
    },
    {
      "epoch": 0.1397310835750066,
      "grad_norm": 0.36276009678840637,
      "learning_rate": 2e-05,
      "loss": 0.5843,
      "step": 1590
    },
    {
      "epoch": 0.14060989542139027,
      "grad_norm": 0.355282723903656,
      "learning_rate": 2e-05,
      "loss": 0.5719,
      "step": 1600
    },
    {
      "epoch": 0.14148870726777396,
      "grad_norm": 0.3517269194126129,
      "learning_rate": 2e-05,
      "loss": 0.5629,
      "step": 1610
    },
    {
      "epoch": 0.14236751911415765,
      "grad_norm": 0.35511642694473267,
      "learning_rate": 2e-05,
      "loss": 0.5458,
      "step": 1620
    },
    {
      "epoch": 0.14324633096054135,
      "grad_norm": 0.3451294004917145,
      "learning_rate": 2e-05,
      "loss": 0.572,
      "step": 1630
    },
    {
      "epoch": 0.14412514280692504,
      "grad_norm": 0.3705539107322693,
      "learning_rate": 2e-05,
      "loss": 0.5335,
      "step": 1640
    },
    {
      "epoch": 0.14500395465330873,
      "grad_norm": 0.43997931480407715,
      "learning_rate": 2e-05,
      "loss": 0.5767,
      "step": 1650
    },
    {
      "epoch": 0.14588276649969242,
      "grad_norm": 0.35617420077323914,
      "learning_rate": 2e-05,
      "loss": 0.57,
      "step": 1660
    },
    {
      "epoch": 0.14676157834607612,
      "grad_norm": 0.37875717878341675,
      "learning_rate": 2e-05,
      "loss": 0.5734,
      "step": 1670
    },
    {
      "epoch": 0.14764039019245978,
      "grad_norm": 0.3402183949947357,
      "learning_rate": 2e-05,
      "loss": 0.5483,
      "step": 1680
    },
    {
      "epoch": 0.14851920203884347,
      "grad_norm": 0.3516283333301544,
      "learning_rate": 2e-05,
      "loss": 0.5653,
      "step": 1690
    },
    {
      "epoch": 0.14939801388522717,
      "grad_norm": 0.854839563369751,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 1700
    },
    {
      "epoch": 0.15027682573161086,
      "grad_norm": 0.3747766315937042,
      "learning_rate": 2e-05,
      "loss": 0.5819,
      "step": 1710
    },
    {
      "epoch": 0.15115563757799455,
      "grad_norm": 0.370029479265213,
      "learning_rate": 2e-05,
      "loss": 0.5952,
      "step": 1720
    },
    {
      "epoch": 0.15203444942437824,
      "grad_norm": 0.3330126404762268,
      "learning_rate": 2e-05,
      "loss": 0.5299,
      "step": 1730
    },
    {
      "epoch": 0.15291326127076194,
      "grad_norm": 0.36191222071647644,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 1740
    },
    {
      "epoch": 0.15379207311714563,
      "grad_norm": 0.3550053536891937,
      "learning_rate": 2e-05,
      "loss": 0.5432,
      "step": 1750
    },
    {
      "epoch": 0.15467088496352932,
      "grad_norm": 0.37736421823501587,
      "learning_rate": 2e-05,
      "loss": 0.5751,
      "step": 1760
    },
    {
      "epoch": 0.15554969680991299,
      "grad_norm": 0.42097368836402893,
      "learning_rate": 2e-05,
      "loss": 0.5758,
      "step": 1770
    },
    {
      "epoch": 0.15642850865629668,
      "grad_norm": 0.4581955671310425,
      "learning_rate": 2e-05,
      "loss": 0.5722,
      "step": 1780
    },
    {
      "epoch": 0.15730732050268037,
      "grad_norm": 0.329158216714859,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 1790
    },
    {
      "epoch": 0.15818613234906406,
      "grad_norm": 0.474506139755249,
      "learning_rate": 2e-05,
      "loss": 0.5526,
      "step": 1800
    },
    {
      "epoch": 0.15906494419544775,
      "grad_norm": 0.3612702488899231,
      "learning_rate": 2e-05,
      "loss": 0.5653,
      "step": 1810
    },
    {
      "epoch": 0.15994375604183145,
      "grad_norm": 0.38358527421951294,
      "learning_rate": 2e-05,
      "loss": 0.5698,
      "step": 1820
    },
    {
      "epoch": 0.16082256788821514,
      "grad_norm": 0.6245725750923157,
      "learning_rate": 2e-05,
      "loss": 0.5551,
      "step": 1830
    },
    {
      "epoch": 0.16170137973459883,
      "grad_norm": 0.47649702429771423,
      "learning_rate": 2e-05,
      "loss": 0.586,
      "step": 1840
    },
    {
      "epoch": 0.16258019158098252,
      "grad_norm": 0.3675776422023773,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 1850
    },
    {
      "epoch": 0.1634590034273662,
      "grad_norm": 0.498647004365921,
      "learning_rate": 2e-05,
      "loss": 0.5741,
      "step": 1860
    },
    {
      "epoch": 0.16433781527374988,
      "grad_norm": 0.3350846767425537,
      "learning_rate": 2e-05,
      "loss": 0.5476,
      "step": 1870
    },
    {
      "epoch": 0.16521662712013357,
      "grad_norm": 0.3603364825248718,
      "learning_rate": 2e-05,
      "loss": 0.5652,
      "step": 1880
    },
    {
      "epoch": 0.16609543896651727,
      "grad_norm": 0.35095348954200745,
      "learning_rate": 2e-05,
      "loss": 0.5746,
      "step": 1890
    },
    {
      "epoch": 0.16697425081290096,
      "grad_norm": 0.33752840757369995,
      "learning_rate": 2e-05,
      "loss": 0.5308,
      "step": 1900
    },
    {
      "epoch": 0.16785306265928465,
      "grad_norm": 0.3529576361179352,
      "learning_rate": 2e-05,
      "loss": 0.5671,
      "step": 1910
    },
    {
      "epoch": 0.16873187450566834,
      "grad_norm": 0.39925357699394226,
      "learning_rate": 2e-05,
      "loss": 0.5751,
      "step": 1920
    },
    {
      "epoch": 0.16961068635205204,
      "grad_norm": 0.34715285897254944,
      "learning_rate": 2e-05,
      "loss": 0.5669,
      "step": 1930
    },
    {
      "epoch": 0.1704894981984357,
      "grad_norm": 0.35600221157073975,
      "learning_rate": 2e-05,
      "loss": 0.5748,
      "step": 1940
    },
    {
      "epoch": 0.1713683100448194,
      "grad_norm": 0.33379316329956055,
      "learning_rate": 2e-05,
      "loss": 0.5785,
      "step": 1950
    },
    {
      "epoch": 0.1722471218912031,
      "grad_norm": 0.3700233995914459,
      "learning_rate": 2e-05,
      "loss": 0.5509,
      "step": 1960
    },
    {
      "epoch": 0.17312593373758678,
      "grad_norm": 0.3257662355899811,
      "learning_rate": 2e-05,
      "loss": 0.5586,
      "step": 1970
    },
    {
      "epoch": 0.17400474558397047,
      "grad_norm": 0.3797135055065155,
      "learning_rate": 2e-05,
      "loss": 0.5652,
      "step": 1980
    },
    {
      "epoch": 0.17488355743035416,
      "grad_norm": 0.42362040281295776,
      "learning_rate": 2e-05,
      "loss": 0.5493,
      "step": 1990
    },
    {
      "epoch": 0.17576236927673786,
      "grad_norm": 0.3780948519706726,
      "learning_rate": 2e-05,
      "loss": 0.5569,
      "step": 2000
    },
    {
      "epoch": 0.17664118112312155,
      "grad_norm": 0.3641783893108368,
      "learning_rate": 2e-05,
      "loss": 0.6153,
      "step": 2010
    },
    {
      "epoch": 0.17751999296950524,
      "grad_norm": 0.4090346097946167,
      "learning_rate": 2e-05,
      "loss": 0.5754,
      "step": 2020
    },
    {
      "epoch": 0.1783988048158889,
      "grad_norm": 0.38258033990859985,
      "learning_rate": 2e-05,
      "loss": 0.5628,
      "step": 2030
    },
    {
      "epoch": 0.1792776166622726,
      "grad_norm": 0.35445868968963623,
      "learning_rate": 2e-05,
      "loss": 0.5542,
      "step": 2040
    },
    {
      "epoch": 0.1801564285086563,
      "grad_norm": 0.31840959191322327,
      "learning_rate": 2e-05,
      "loss": 0.5689,
      "step": 2050
    },
    {
      "epoch": 0.18103524035503998,
      "grad_norm": 0.3384089767932892,
      "learning_rate": 2e-05,
      "loss": 0.5553,
      "step": 2060
    },
    {
      "epoch": 0.18191405220142368,
      "grad_norm": 0.4153691828250885,
      "learning_rate": 2e-05,
      "loss": 0.5831,
      "step": 2070
    },
    {
      "epoch": 0.18279286404780737,
      "grad_norm": 0.3473172187805176,
      "learning_rate": 2e-05,
      "loss": 0.5285,
      "step": 2080
    },
    {
      "epoch": 0.18367167589419106,
      "grad_norm": 1.0200769901275635,
      "learning_rate": 2e-05,
      "loss": 0.5793,
      "step": 2090
    },
    {
      "epoch": 0.18455048774057475,
      "grad_norm": 0.34419798851013184,
      "learning_rate": 2e-05,
      "loss": 0.5602,
      "step": 2100
    },
    {
      "epoch": 0.18542929958695845,
      "grad_norm": 0.3456267714500427,
      "learning_rate": 2e-05,
      "loss": 0.5545,
      "step": 2110
    },
    {
      "epoch": 0.1863081114333421,
      "grad_norm": 0.32470136880874634,
      "learning_rate": 2e-05,
      "loss": 0.5664,
      "step": 2120
    },
    {
      "epoch": 0.1871869232797258,
      "grad_norm": 0.34200549125671387,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 2130
    },
    {
      "epoch": 0.1880657351261095,
      "grad_norm": 0.3530149757862091,
      "learning_rate": 2e-05,
      "loss": 0.5596,
      "step": 2140
    },
    {
      "epoch": 0.1889445469724932,
      "grad_norm": 0.4087691605091095,
      "learning_rate": 2e-05,
      "loss": 0.563,
      "step": 2150
    },
    {
      "epoch": 0.18982335881887688,
      "grad_norm": 0.35635480284690857,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 2160
    },
    {
      "epoch": 0.19070217066526057,
      "grad_norm": 0.35621872544288635,
      "learning_rate": 2e-05,
      "loss": 0.5818,
      "step": 2170
    },
    {
      "epoch": 0.19158098251164427,
      "grad_norm": 0.3401750922203064,
      "learning_rate": 2e-05,
      "loss": 0.5909,
      "step": 2180
    },
    {
      "epoch": 0.19245979435802796,
      "grad_norm": 0.3132881820201874,
      "learning_rate": 2e-05,
      "loss": 0.5579,
      "step": 2190
    },
    {
      "epoch": 0.19333860620441162,
      "grad_norm": 0.3682171702384949,
      "learning_rate": 2e-05,
      "loss": 0.5912,
      "step": 2200
    },
    {
      "epoch": 0.19421741805079532,
      "grad_norm": 0.4018304646015167,
      "learning_rate": 2e-05,
      "loss": 0.5898,
      "step": 2210
    },
    {
      "epoch": 0.195096229897179,
      "grad_norm": 0.36482760310173035,
      "learning_rate": 2e-05,
      "loss": 0.5752,
      "step": 2220
    },
    {
      "epoch": 0.1959750417435627,
      "grad_norm": 0.29210537672042847,
      "learning_rate": 2e-05,
      "loss": 0.556,
      "step": 2230
    },
    {
      "epoch": 0.1968538535899464,
      "grad_norm": 0.35293927788734436,
      "learning_rate": 2e-05,
      "loss": 0.5914,
      "step": 2240
    },
    {
      "epoch": 0.19773266543633009,
      "grad_norm": 0.34668824076652527,
      "learning_rate": 2e-05,
      "loss": 0.577,
      "step": 2250
    },
    {
      "epoch": 0.19861147728271378,
      "grad_norm": 0.360667884349823,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 2260
    },
    {
      "epoch": 0.19949028912909747,
      "grad_norm": 0.3482297956943512,
      "learning_rate": 2e-05,
      "loss": 0.5465,
      "step": 2270
    },
    {
      "epoch": 0.20036910097548116,
      "grad_norm": 0.339021772146225,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 2280
    },
    {
      "epoch": 0.20124791282186483,
      "grad_norm": 0.36110997200012207,
      "learning_rate": 2e-05,
      "loss": 0.5538,
      "step": 2290
    },
    {
      "epoch": 0.20212672466824852,
      "grad_norm": 0.35237032175064087,
      "learning_rate": 2e-05,
      "loss": 0.5548,
      "step": 2300
    },
    {
      "epoch": 0.2030055365146322,
      "grad_norm": 0.4096246063709259,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 2310
    },
    {
      "epoch": 0.2038843483610159,
      "grad_norm": 0.41096365451812744,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 2320
    },
    {
      "epoch": 0.2047631602073996,
      "grad_norm": 0.3804115355014801,
      "learning_rate": 2e-05,
      "loss": 0.5508,
      "step": 2330
    },
    {
      "epoch": 0.2056419720537833,
      "grad_norm": 0.3373338580131531,
      "learning_rate": 2e-05,
      "loss": 0.5659,
      "step": 2340
    },
    {
      "epoch": 0.20652078390016698,
      "grad_norm": 0.37081179022789,
      "learning_rate": 2e-05,
      "loss": 0.5592,
      "step": 2350
    },
    {
      "epoch": 0.20739959574655067,
      "grad_norm": 0.34369438886642456,
      "learning_rate": 2e-05,
      "loss": 0.6028,
      "step": 2360
    },
    {
      "epoch": 0.20827840759293434,
      "grad_norm": 0.3445272147655487,
      "learning_rate": 2e-05,
      "loss": 0.536,
      "step": 2370
    },
    {
      "epoch": 0.20915721943931803,
      "grad_norm": 0.36156561970710754,
      "learning_rate": 2e-05,
      "loss": 0.5625,
      "step": 2380
    },
    {
      "epoch": 0.21003603128570172,
      "grad_norm": 0.3855374753475189,
      "learning_rate": 2e-05,
      "loss": 0.5516,
      "step": 2390
    },
    {
      "epoch": 0.21091484313208542,
      "grad_norm": 0.36775392293930054,
      "learning_rate": 2e-05,
      "loss": 0.5893,
      "step": 2400
    },
    {
      "epoch": 0.2117936549784691,
      "grad_norm": 0.3330851197242737,
      "learning_rate": 2e-05,
      "loss": 0.5767,
      "step": 2410
    },
    {
      "epoch": 0.2126724668248528,
      "grad_norm": 0.34514012932777405,
      "learning_rate": 2e-05,
      "loss": 0.5727,
      "step": 2420
    },
    {
      "epoch": 0.2135512786712365,
      "grad_norm": 0.47892290353775024,
      "learning_rate": 2e-05,
      "loss": 0.5875,
      "step": 2430
    },
    {
      "epoch": 0.2144300905176202,
      "grad_norm": 0.36140429973602295,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 2440
    },
    {
      "epoch": 0.21530890236400388,
      "grad_norm": 0.33918988704681396,
      "learning_rate": 2e-05,
      "loss": 0.5562,
      "step": 2450
    },
    {
      "epoch": 0.21618771421038754,
      "grad_norm": 0.3281526565551758,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 2460
    },
    {
      "epoch": 0.21706652605677124,
      "grad_norm": 0.3529514968395233,
      "learning_rate": 2e-05,
      "loss": 0.5498,
      "step": 2470
    },
    {
      "epoch": 0.21794533790315493,
      "grad_norm": 0.339227557182312,
      "learning_rate": 2e-05,
      "loss": 0.5498,
      "step": 2480
    },
    {
      "epoch": 0.21882414974953862,
      "grad_norm": 0.3628411293029785,
      "learning_rate": 2e-05,
      "loss": 0.557,
      "step": 2490
    },
    {
      "epoch": 0.2197029615959223,
      "grad_norm": 0.34096869826316833,
      "learning_rate": 2e-05,
      "loss": 0.5719,
      "step": 2500
    },
    {
      "epoch": 0.220581773442306,
      "grad_norm": 0.32310160994529724,
      "learning_rate": 2e-05,
      "loss": 0.5729,
      "step": 2510
    },
    {
      "epoch": 0.2214605852886897,
      "grad_norm": 0.3401024043560028,
      "learning_rate": 2e-05,
      "loss": 0.555,
      "step": 2520
    },
    {
      "epoch": 0.2223393971350734,
      "grad_norm": 0.46229088306427,
      "learning_rate": 2e-05,
      "loss": 0.5514,
      "step": 2530
    },
    {
      "epoch": 0.22321820898145708,
      "grad_norm": 0.3563999831676483,
      "learning_rate": 2e-05,
      "loss": 0.5748,
      "step": 2540
    },
    {
      "epoch": 0.22409702082784075,
      "grad_norm": 0.30827537178993225,
      "learning_rate": 2e-05,
      "loss": 0.5635,
      "step": 2550
    },
    {
      "epoch": 0.22497583267422444,
      "grad_norm": 0.4186650812625885,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 2560
    },
    {
      "epoch": 0.22585464452060813,
      "grad_norm": 0.3800259232521057,
      "learning_rate": 2e-05,
      "loss": 0.5712,
      "step": 2570
    },
    {
      "epoch": 0.22673345636699183,
      "grad_norm": 0.33005034923553467,
      "learning_rate": 2e-05,
      "loss": 0.5729,
      "step": 2580
    },
    {
      "epoch": 0.22761226821337552,
      "grad_norm": 0.32586944103240967,
      "learning_rate": 2e-05,
      "loss": 0.5703,
      "step": 2590
    },
    {
      "epoch": 0.2284910800597592,
      "grad_norm": 0.3634280562400818,
      "learning_rate": 2e-05,
      "loss": 0.5453,
      "step": 2600
    },
    {
      "epoch": 0.2293698919061429,
      "grad_norm": 0.34054821729660034,
      "learning_rate": 2e-05,
      "loss": 0.5432,
      "step": 2610
    },
    {
      "epoch": 0.2302487037525266,
      "grad_norm": 0.3557206094264984,
      "learning_rate": 2e-05,
      "loss": 0.551,
      "step": 2620
    },
    {
      "epoch": 0.23112751559891026,
      "grad_norm": 0.35411056876182556,
      "learning_rate": 2e-05,
      "loss": 0.5566,
      "step": 2630
    },
    {
      "epoch": 0.23200632744529395,
      "grad_norm": 0.38252395391464233,
      "learning_rate": 2e-05,
      "loss": 0.5462,
      "step": 2640
    },
    {
      "epoch": 0.23288513929167765,
      "grad_norm": 0.359747052192688,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 2650
    },
    {
      "epoch": 0.23376395113806134,
      "grad_norm": 0.3527311086654663,
      "learning_rate": 2e-05,
      "loss": 0.5575,
      "step": 2660
    },
    {
      "epoch": 0.23464276298444503,
      "grad_norm": 0.31489190459251404,
      "learning_rate": 2e-05,
      "loss": 0.5641,
      "step": 2670
    },
    {
      "epoch": 0.23552157483082872,
      "grad_norm": 0.3567199409008026,
      "learning_rate": 2e-05,
      "loss": 0.5614,
      "step": 2680
    },
    {
      "epoch": 0.23640038667721242,
      "grad_norm": 0.40155547857284546,
      "learning_rate": 2e-05,
      "loss": 0.5465,
      "step": 2690
    },
    {
      "epoch": 0.2372791985235961,
      "grad_norm": 0.3615034818649292,
      "learning_rate": 2e-05,
      "loss": 0.566,
      "step": 2700
    },
    {
      "epoch": 0.2381580103699798,
      "grad_norm": 0.38731303811073303,
      "learning_rate": 2e-05,
      "loss": 0.5571,
      "step": 2710
    },
    {
      "epoch": 0.23903682221636346,
      "grad_norm": 0.3839712142944336,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 2720
    },
    {
      "epoch": 0.23991563406274716,
      "grad_norm": 0.33872711658477783,
      "learning_rate": 2e-05,
      "loss": 0.5362,
      "step": 2730
    },
    {
      "epoch": 0.24079444590913085,
      "grad_norm": 0.3734551966190338,
      "learning_rate": 2e-05,
      "loss": 0.5652,
      "step": 2740
    },
    {
      "epoch": 0.24167325775551454,
      "grad_norm": 0.36234650015830994,
      "learning_rate": 2e-05,
      "loss": 0.5594,
      "step": 2750
    },
    {
      "epoch": 0.24255206960189823,
      "grad_norm": 0.36426880955696106,
      "learning_rate": 2e-05,
      "loss": 0.5413,
      "step": 2760
    },
    {
      "epoch": 0.24343088144828193,
      "grad_norm": 0.3386363983154297,
      "learning_rate": 2e-05,
      "loss": 0.5431,
      "step": 2770
    },
    {
      "epoch": 0.24430969329466562,
      "grad_norm": 0.33737513422966003,
      "learning_rate": 2e-05,
      "loss": 0.554,
      "step": 2780
    },
    {
      "epoch": 0.2451885051410493,
      "grad_norm": 0.4223691523075104,
      "learning_rate": 2e-05,
      "loss": 0.5708,
      "step": 2790
    },
    {
      "epoch": 0.24606731698743298,
      "grad_norm": 0.3341444134712219,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 2800
    },
    {
      "epoch": 0.24694612883381667,
      "grad_norm": 0.3827393651008606,
      "learning_rate": 2e-05,
      "loss": 0.5748,
      "step": 2810
    },
    {
      "epoch": 0.24782494068020036,
      "grad_norm": 0.31838056445121765,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 2820
    },
    {
      "epoch": 0.24870375252658405,
      "grad_norm": 0.3411373496055603,
      "learning_rate": 2e-05,
      "loss": 0.5508,
      "step": 2830
    },
    {
      "epoch": 0.24958256437296775,
      "grad_norm": 0.3904740512371063,
      "learning_rate": 2e-05,
      "loss": 0.5469,
      "step": 2840
    },
    {
      "epoch": 0.25046137621935144,
      "grad_norm": 0.3458789885044098,
      "learning_rate": 2e-05,
      "loss": 0.5754,
      "step": 2850
    },
    {
      "epoch": 0.2513401880657351,
      "grad_norm": 0.3715694546699524,
      "learning_rate": 2e-05,
      "loss": 0.567,
      "step": 2860
    },
    {
      "epoch": 0.2522189999121188,
      "grad_norm": 0.40037286281585693,
      "learning_rate": 2e-05,
      "loss": 0.5428,
      "step": 2870
    },
    {
      "epoch": 0.2530978117585025,
      "grad_norm": 0.33550241589546204,
      "learning_rate": 2e-05,
      "loss": 0.5524,
      "step": 2880
    },
    {
      "epoch": 0.2539766236048862,
      "grad_norm": 0.35246455669403076,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 2890
    },
    {
      "epoch": 0.2548554354512699,
      "grad_norm": 0.37529873847961426,
      "learning_rate": 2e-05,
      "loss": 0.5855,
      "step": 2900
    },
    {
      "epoch": 0.2557342472976536,
      "grad_norm": 0.3769456744194031,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 2910
    },
    {
      "epoch": 0.25661305914403726,
      "grad_norm": 0.32500946521759033,
      "learning_rate": 2e-05,
      "loss": 0.5316,
      "step": 2920
    },
    {
      "epoch": 0.2574918709904209,
      "grad_norm": 0.7248839139938354,
      "learning_rate": 2e-05,
      "loss": 0.5609,
      "step": 2930
    },
    {
      "epoch": 0.25837068283680464,
      "grad_norm": 0.3718763291835785,
      "learning_rate": 2e-05,
      "loss": 0.5686,
      "step": 2940
    },
    {
      "epoch": 0.2592494946831883,
      "grad_norm": 0.34554237127304077,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 2950
    },
    {
      "epoch": 0.26012830652957203,
      "grad_norm": 0.48291119933128357,
      "learning_rate": 2e-05,
      "loss": 0.5575,
      "step": 2960
    },
    {
      "epoch": 0.2610071183759557,
      "grad_norm": 0.3251497149467468,
      "learning_rate": 2e-05,
      "loss": 0.5307,
      "step": 2970
    },
    {
      "epoch": 0.2618859302223394,
      "grad_norm": 0.3903724253177643,
      "learning_rate": 2e-05,
      "loss": 0.56,
      "step": 2980
    },
    {
      "epoch": 0.2627647420687231,
      "grad_norm": 0.6968425512313843,
      "learning_rate": 2e-05,
      "loss": 0.5503,
      "step": 2990
    },
    {
      "epoch": 0.2636435539151068,
      "grad_norm": 0.367001473903656,
      "learning_rate": 2e-05,
      "loss": 0.5347,
      "step": 3000
    },
    {
      "epoch": 0.26452236576149046,
      "grad_norm": 0.33539122343063354,
      "learning_rate": 2e-05,
      "loss": 0.5521,
      "step": 3010
    },
    {
      "epoch": 0.26540117760787413,
      "grad_norm": 0.326261430978775,
      "learning_rate": 2e-05,
      "loss": 0.5578,
      "step": 3020
    },
    {
      "epoch": 0.26627998945425785,
      "grad_norm": 0.34501734375953674,
      "learning_rate": 2e-05,
      "loss": 0.5787,
      "step": 3030
    },
    {
      "epoch": 0.2671588013006415,
      "grad_norm": 0.3778016269207001,
      "learning_rate": 2e-05,
      "loss": 0.5703,
      "step": 3040
    },
    {
      "epoch": 0.26803761314702523,
      "grad_norm": 0.356425017118454,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 3050
    },
    {
      "epoch": 0.2689164249934089,
      "grad_norm": 0.3525022268295288,
      "learning_rate": 2e-05,
      "loss": 0.5452,
      "step": 3060
    },
    {
      "epoch": 0.2697952368397926,
      "grad_norm": 0.33454424142837524,
      "learning_rate": 2e-05,
      "loss": 0.5678,
      "step": 3070
    },
    {
      "epoch": 0.2706740486861763,
      "grad_norm": 0.3657768964767456,
      "learning_rate": 2e-05,
      "loss": 0.5567,
      "step": 3080
    },
    {
      "epoch": 0.27155286053256,
      "grad_norm": 0.3132709264755249,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 3090
    },
    {
      "epoch": 0.27243167237894367,
      "grad_norm": 0.3250260651111603,
      "learning_rate": 2e-05,
      "loss": 0.5446,
      "step": 3100
    },
    {
      "epoch": 0.27331048422532733,
      "grad_norm": 0.35021060705184937,
      "learning_rate": 2e-05,
      "loss": 0.5642,
      "step": 3110
    },
    {
      "epoch": 0.27418929607171105,
      "grad_norm": 0.3359904885292053,
      "learning_rate": 2e-05,
      "loss": 0.532,
      "step": 3120
    },
    {
      "epoch": 0.2750681079180947,
      "grad_norm": 0.34872737526893616,
      "learning_rate": 2e-05,
      "loss": 0.5476,
      "step": 3130
    },
    {
      "epoch": 0.27594691976447844,
      "grad_norm": 0.3775680661201477,
      "learning_rate": 2e-05,
      "loss": 0.5493,
      "step": 3140
    },
    {
      "epoch": 0.2768257316108621,
      "grad_norm": 0.35976430773735046,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 3150
    },
    {
      "epoch": 0.2777045434572458,
      "grad_norm": 0.34369155764579773,
      "learning_rate": 2e-05,
      "loss": 0.5536,
      "step": 3160
    },
    {
      "epoch": 0.2785833553036295,
      "grad_norm": 0.3594971001148224,
      "learning_rate": 2e-05,
      "loss": 0.5643,
      "step": 3170
    },
    {
      "epoch": 0.2794621671500132,
      "grad_norm": 0.3360744118690491,
      "learning_rate": 2e-05,
      "loss": 0.5536,
      "step": 3180
    },
    {
      "epoch": 0.2803409789963969,
      "grad_norm": 0.36943694949150085,
      "learning_rate": 2e-05,
      "loss": 0.5542,
      "step": 3190
    },
    {
      "epoch": 0.28121979084278054,
      "grad_norm": 0.3765130937099457,
      "learning_rate": 2e-05,
      "loss": 0.574,
      "step": 3200
    },
    {
      "epoch": 0.28209860268916426,
      "grad_norm": 0.3518107831478119,
      "learning_rate": 2e-05,
      "loss": 0.5691,
      "step": 3210
    },
    {
      "epoch": 0.2829774145355479,
      "grad_norm": 0.3056955635547638,
      "learning_rate": 2e-05,
      "loss": 0.5569,
      "step": 3220
    },
    {
      "epoch": 0.28385622638193164,
      "grad_norm": 0.37627944350242615,
      "learning_rate": 2e-05,
      "loss": 0.5706,
      "step": 3230
    },
    {
      "epoch": 0.2847350382283153,
      "grad_norm": 0.3819272518157959,
      "learning_rate": 2e-05,
      "loss": 0.5674,
      "step": 3240
    },
    {
      "epoch": 0.285613850074699,
      "grad_norm": 0.39118707180023193,
      "learning_rate": 2e-05,
      "loss": 0.5573,
      "step": 3250
    },
    {
      "epoch": 0.2864926619210827,
      "grad_norm": 0.4317283034324646,
      "learning_rate": 2e-05,
      "loss": 0.564,
      "step": 3260
    },
    {
      "epoch": 0.2873714737674664,
      "grad_norm": 0.3278979957103729,
      "learning_rate": 2e-05,
      "loss": 0.5655,
      "step": 3270
    },
    {
      "epoch": 0.2882502856138501,
      "grad_norm": 0.35178691148757935,
      "learning_rate": 2e-05,
      "loss": 0.5539,
      "step": 3280
    },
    {
      "epoch": 0.28912909746023374,
      "grad_norm": 0.36217454075813293,
      "learning_rate": 2e-05,
      "loss": 0.558,
      "step": 3290
    },
    {
      "epoch": 0.29000790930661746,
      "grad_norm": 0.34998324513435364,
      "learning_rate": 2e-05,
      "loss": 0.5604,
      "step": 3300
    },
    {
      "epoch": 0.2908867211530011,
      "grad_norm": 0.44804877042770386,
      "learning_rate": 2e-05,
      "loss": 0.5341,
      "step": 3310
    },
    {
      "epoch": 0.29176553299938485,
      "grad_norm": 0.33786121010780334,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 3320
    },
    {
      "epoch": 0.2926443448457685,
      "grad_norm": 0.35405269265174866,
      "learning_rate": 2e-05,
      "loss": 0.567,
      "step": 3330
    },
    {
      "epoch": 0.29352315669215223,
      "grad_norm": 0.36850911378860474,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 3340
    },
    {
      "epoch": 0.2944019685385359,
      "grad_norm": 0.455915629863739,
      "learning_rate": 2e-05,
      "loss": 0.5569,
      "step": 3350
    },
    {
      "epoch": 0.29528078038491956,
      "grad_norm": 0.3683995008468628,
      "learning_rate": 2e-05,
      "loss": 0.5332,
      "step": 3360
    },
    {
      "epoch": 0.2961595922313033,
      "grad_norm": 0.36006689071655273,
      "learning_rate": 2e-05,
      "loss": 0.5607,
      "step": 3370
    },
    {
      "epoch": 0.29703840407768695,
      "grad_norm": 0.35571765899658203,
      "learning_rate": 2e-05,
      "loss": 0.5557,
      "step": 3380
    },
    {
      "epoch": 0.29791721592407067,
      "grad_norm": 0.4341544806957245,
      "learning_rate": 2e-05,
      "loss": 0.5611,
      "step": 3390
    },
    {
      "epoch": 0.29879602777045433,
      "grad_norm": 0.3267984092235565,
      "learning_rate": 2e-05,
      "loss": 0.5413,
      "step": 3400
    },
    {
      "epoch": 0.29967483961683805,
      "grad_norm": 0.35858282446861267,
      "learning_rate": 2e-05,
      "loss": 0.5479,
      "step": 3410
    },
    {
      "epoch": 0.3005536514632217,
      "grad_norm": 0.34086737036705017,
      "learning_rate": 2e-05,
      "loss": 0.5555,
      "step": 3420
    },
    {
      "epoch": 0.30143246330960544,
      "grad_norm": 0.3618117570877075,
      "learning_rate": 2e-05,
      "loss": 0.5544,
      "step": 3430
    },
    {
      "epoch": 0.3023112751559891,
      "grad_norm": 0.5729960799217224,
      "learning_rate": 2e-05,
      "loss": 0.5382,
      "step": 3440
    },
    {
      "epoch": 0.30319008700237277,
      "grad_norm": 0.3379969894886017,
      "learning_rate": 2e-05,
      "loss": 0.5678,
      "step": 3450
    },
    {
      "epoch": 0.3040688988487565,
      "grad_norm": 0.37390968203544617,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 3460
    },
    {
      "epoch": 0.30494771069514015,
      "grad_norm": 0.3556511402130127,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 3470
    },
    {
      "epoch": 0.30582652254152387,
      "grad_norm": 0.34213700890541077,
      "learning_rate": 2e-05,
      "loss": 0.5534,
      "step": 3480
    },
    {
      "epoch": 0.30670533438790754,
      "grad_norm": 0.3467865586280823,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 3490
    },
    {
      "epoch": 0.30758414623429126,
      "grad_norm": 0.3172374963760376,
      "learning_rate": 2e-05,
      "loss": 0.5648,
      "step": 3500
    },
    {
      "epoch": 0.3084629580806749,
      "grad_norm": 0.3406992554664612,
      "learning_rate": 2e-05,
      "loss": 0.5697,
      "step": 3510
    },
    {
      "epoch": 0.30934176992705864,
      "grad_norm": 0.3619944453239441,
      "learning_rate": 2e-05,
      "loss": 0.593,
      "step": 3520
    },
    {
      "epoch": 0.3102205817734423,
      "grad_norm": 0.3534380793571472,
      "learning_rate": 2e-05,
      "loss": 0.5632,
      "step": 3530
    },
    {
      "epoch": 0.31109939361982597,
      "grad_norm": 0.3550601303577423,
      "learning_rate": 2e-05,
      "loss": 0.5533,
      "step": 3540
    },
    {
      "epoch": 0.3119782054662097,
      "grad_norm": 0.4721556305885315,
      "learning_rate": 2e-05,
      "loss": 0.582,
      "step": 3550
    },
    {
      "epoch": 0.31285701731259336,
      "grad_norm": 0.3861790895462036,
      "learning_rate": 2e-05,
      "loss": 0.566,
      "step": 3560
    },
    {
      "epoch": 0.3137358291589771,
      "grad_norm": 0.3908025622367859,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 3570
    },
    {
      "epoch": 0.31461464100536074,
      "grad_norm": 0.3980657160282135,
      "learning_rate": 2e-05,
      "loss": 0.5802,
      "step": 3580
    },
    {
      "epoch": 0.31549345285174446,
      "grad_norm": 0.35337314009666443,
      "learning_rate": 2e-05,
      "loss": 0.5844,
      "step": 3590
    },
    {
      "epoch": 0.3163722646981281,
      "grad_norm": 0.3947729468345642,
      "learning_rate": 2e-05,
      "loss": 0.5317,
      "step": 3600
    },
    {
      "epoch": 0.31725107654451185,
      "grad_norm": 0.3219287097454071,
      "learning_rate": 2e-05,
      "loss": 0.5319,
      "step": 3610
    },
    {
      "epoch": 0.3181298883908955,
      "grad_norm": 0.3003973662853241,
      "learning_rate": 2e-05,
      "loss": 0.519,
      "step": 3620
    },
    {
      "epoch": 0.3190087002372792,
      "grad_norm": 0.33029764890670776,
      "learning_rate": 2e-05,
      "loss": 0.5516,
      "step": 3630
    },
    {
      "epoch": 0.3198875120836629,
      "grad_norm": 0.3382836878299713,
      "learning_rate": 2e-05,
      "loss": 0.5526,
      "step": 3640
    },
    {
      "epoch": 0.32076632393004656,
      "grad_norm": 0.3484552204608917,
      "learning_rate": 2e-05,
      "loss": 0.5653,
      "step": 3650
    },
    {
      "epoch": 0.3216451357764303,
      "grad_norm": 0.34894368052482605,
      "learning_rate": 2e-05,
      "loss": 0.5771,
      "step": 3660
    },
    {
      "epoch": 0.32252394762281394,
      "grad_norm": 0.34267646074295044,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 3670
    },
    {
      "epoch": 0.32340275946919766,
      "grad_norm": 0.4821334183216095,
      "learning_rate": 2e-05,
      "loss": 0.5357,
      "step": 3680
    },
    {
      "epoch": 0.32428157131558133,
      "grad_norm": 0.40147072076797485,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 3690
    },
    {
      "epoch": 0.32516038316196505,
      "grad_norm": 0.3397529125213623,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 3700
    },
    {
      "epoch": 0.3260391950083487,
      "grad_norm": 0.3210875988006592,
      "learning_rate": 2e-05,
      "loss": 0.5626,
      "step": 3710
    },
    {
      "epoch": 0.3269180068547324,
      "grad_norm": 0.3875472843647003,
      "learning_rate": 2e-05,
      "loss": 0.5732,
      "step": 3720
    },
    {
      "epoch": 0.3277968187011161,
      "grad_norm": 0.32561638951301575,
      "learning_rate": 2e-05,
      "loss": 0.5418,
      "step": 3730
    },
    {
      "epoch": 0.32867563054749976,
      "grad_norm": 0.3329981863498688,
      "learning_rate": 2e-05,
      "loss": 0.5547,
      "step": 3740
    },
    {
      "epoch": 0.3295544423938835,
      "grad_norm": 0.349528968334198,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 3750
    },
    {
      "epoch": 0.33043325424026715,
      "grad_norm": 0.4276971220970154,
      "learning_rate": 2e-05,
      "loss": 0.5711,
      "step": 3760
    },
    {
      "epoch": 0.33131206608665087,
      "grad_norm": 0.3583855628967285,
      "learning_rate": 2e-05,
      "loss": 0.544,
      "step": 3770
    },
    {
      "epoch": 0.33219087793303453,
      "grad_norm": 0.39071449637413025,
      "learning_rate": 2e-05,
      "loss": 0.5676,
      "step": 3780
    },
    {
      "epoch": 0.33306968977941825,
      "grad_norm": 0.3613530993461609,
      "learning_rate": 2e-05,
      "loss": 0.5444,
      "step": 3790
    },
    {
      "epoch": 0.3339485016258019,
      "grad_norm": 3.2899422645568848,
      "learning_rate": 2e-05,
      "loss": 0.5845,
      "step": 3800
    },
    {
      "epoch": 0.3348273134721856,
      "grad_norm": 0.36647123098373413,
      "learning_rate": 2e-05,
      "loss": 0.5525,
      "step": 3810
    },
    {
      "epoch": 0.3357061253185693,
      "grad_norm": 0.4115641117095947,
      "learning_rate": 2e-05,
      "loss": 0.5761,
      "step": 3820
    },
    {
      "epoch": 0.33658493716495297,
      "grad_norm": 0.35072311758995056,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 3830
    },
    {
      "epoch": 0.3374637490113367,
      "grad_norm": 0.40616434812545776,
      "learning_rate": 2e-05,
      "loss": 0.5703,
      "step": 3840
    },
    {
      "epoch": 0.33834256085772035,
      "grad_norm": 0.3501415550708771,
      "learning_rate": 2e-05,
      "loss": 0.5705,
      "step": 3850
    },
    {
      "epoch": 0.3392213727041041,
      "grad_norm": 0.3634043335914612,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 3860
    },
    {
      "epoch": 0.34010018455048774,
      "grad_norm": 0.37502527236938477,
      "learning_rate": 2e-05,
      "loss": 0.5602,
      "step": 3870
    },
    {
      "epoch": 0.3409789963968714,
      "grad_norm": 0.3518204391002655,
      "learning_rate": 2e-05,
      "loss": 0.5516,
      "step": 3880
    },
    {
      "epoch": 0.3418578082432551,
      "grad_norm": 0.3304819166660309,
      "learning_rate": 2e-05,
      "loss": 0.564,
      "step": 3890
    },
    {
      "epoch": 0.3427366200896388,
      "grad_norm": 0.3323899209499359,
      "learning_rate": 2e-05,
      "loss": 0.55,
      "step": 3900
    },
    {
      "epoch": 0.3436154319360225,
      "grad_norm": 0.3736434578895569,
      "learning_rate": 2e-05,
      "loss": 0.5657,
      "step": 3910
    },
    {
      "epoch": 0.3444942437824062,
      "grad_norm": 0.37221184372901917,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 3920
    },
    {
      "epoch": 0.3453730556287899,
      "grad_norm": 0.37677672505378723,
      "learning_rate": 2e-05,
      "loss": 0.5796,
      "step": 3930
    },
    {
      "epoch": 0.34625186747517356,
      "grad_norm": 0.3674095869064331,
      "learning_rate": 2e-05,
      "loss": 0.5534,
      "step": 3940
    },
    {
      "epoch": 0.3471306793215573,
      "grad_norm": 0.3323845863342285,
      "learning_rate": 2e-05,
      "loss": 0.5527,
      "step": 3950
    },
    {
      "epoch": 0.34800949116794094,
      "grad_norm": 0.4143407642841339,
      "learning_rate": 2e-05,
      "loss": 0.5314,
      "step": 3960
    },
    {
      "epoch": 0.3488883030143246,
      "grad_norm": 0.3421297073364258,
      "learning_rate": 2e-05,
      "loss": 0.586,
      "step": 3970
    },
    {
      "epoch": 0.34976711486070833,
      "grad_norm": 0.34102344512939453,
      "learning_rate": 2e-05,
      "loss": 0.5579,
      "step": 3980
    },
    {
      "epoch": 0.350645926707092,
      "grad_norm": 0.3486737906932831,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 3990
    },
    {
      "epoch": 0.3515247385534757,
      "grad_norm": 0.33497175574302673,
      "learning_rate": 2e-05,
      "loss": 0.5752,
      "step": 4000
    },
    {
      "epoch": 0.3524035503998594,
      "grad_norm": 0.386078417301178,
      "learning_rate": 2e-05,
      "loss": 0.5573,
      "step": 4010
    },
    {
      "epoch": 0.3532823622462431,
      "grad_norm": 0.3763893246650696,
      "learning_rate": 2e-05,
      "loss": 0.5693,
      "step": 4020
    },
    {
      "epoch": 0.35416117409262676,
      "grad_norm": 0.3679053485393524,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 4030
    },
    {
      "epoch": 0.3550399859390105,
      "grad_norm": 0.3901916742324829,
      "learning_rate": 2e-05,
      "loss": 0.5643,
      "step": 4040
    },
    {
      "epoch": 0.35591879778539415,
      "grad_norm": 0.4080599248409271,
      "learning_rate": 2e-05,
      "loss": 0.5743,
      "step": 4050
    },
    {
      "epoch": 0.3567976096317778,
      "grad_norm": 0.38750332593917847,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 4060
    },
    {
      "epoch": 0.35767642147816153,
      "grad_norm": 0.3243080973625183,
      "learning_rate": 2e-05,
      "loss": 0.5524,
      "step": 4070
    },
    {
      "epoch": 0.3585552333245452,
      "grad_norm": 0.3177420496940613,
      "learning_rate": 2e-05,
      "loss": 0.5508,
      "step": 4080
    },
    {
      "epoch": 0.3594340451709289,
      "grad_norm": 0.42822059988975525,
      "learning_rate": 2e-05,
      "loss": 0.5567,
      "step": 4090
    },
    {
      "epoch": 0.3603128570173126,
      "grad_norm": 0.33071213960647583,
      "learning_rate": 2e-05,
      "loss": 0.5463,
      "step": 4100
    },
    {
      "epoch": 0.3611916688636963,
      "grad_norm": 0.3083895146846771,
      "learning_rate": 2e-05,
      "loss": 0.5433,
      "step": 4110
    },
    {
      "epoch": 0.36207048071007997,
      "grad_norm": 0.3738991618156433,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 4120
    },
    {
      "epoch": 0.3629492925564637,
      "grad_norm": 0.3362555205821991,
      "learning_rate": 2e-05,
      "loss": 0.55,
      "step": 4130
    },
    {
      "epoch": 0.36382810440284735,
      "grad_norm": 0.39087429642677307,
      "learning_rate": 2e-05,
      "loss": 0.5662,
      "step": 4140
    },
    {
      "epoch": 0.364706916249231,
      "grad_norm": 0.5044722557067871,
      "learning_rate": 2e-05,
      "loss": 0.554,
      "step": 4150
    },
    {
      "epoch": 0.36558572809561474,
      "grad_norm": 0.38679587841033936,
      "learning_rate": 2e-05,
      "loss": 0.5716,
      "step": 4160
    },
    {
      "epoch": 0.3664645399419984,
      "grad_norm": 0.2997303903102875,
      "learning_rate": 2e-05,
      "loss": 0.5443,
      "step": 4170
    },
    {
      "epoch": 0.3673433517883821,
      "grad_norm": 0.5776960849761963,
      "learning_rate": 2e-05,
      "loss": 0.5433,
      "step": 4180
    },
    {
      "epoch": 0.3682221636347658,
      "grad_norm": 0.3290417194366455,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 4190
    },
    {
      "epoch": 0.3691009754811495,
      "grad_norm": 0.31029802560806274,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 4200
    },
    {
      "epoch": 0.36997978732753317,
      "grad_norm": 0.34994277358055115,
      "learning_rate": 2e-05,
      "loss": 0.5474,
      "step": 4210
    },
    {
      "epoch": 0.3708585991739169,
      "grad_norm": 0.34928035736083984,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 4220
    },
    {
      "epoch": 0.37173741102030056,
      "grad_norm": 0.39571091532707214,
      "learning_rate": 2e-05,
      "loss": 0.5651,
      "step": 4230
    },
    {
      "epoch": 0.3726162228666842,
      "grad_norm": 0.2939375042915344,
      "learning_rate": 2e-05,
      "loss": 0.5306,
      "step": 4240
    },
    {
      "epoch": 0.37349503471306794,
      "grad_norm": 0.362588495016098,
      "learning_rate": 2e-05,
      "loss": 0.5398,
      "step": 4250
    },
    {
      "epoch": 0.3743738465594516,
      "grad_norm": 0.3571527600288391,
      "learning_rate": 2e-05,
      "loss": 0.5562,
      "step": 4260
    },
    {
      "epoch": 0.3752526584058353,
      "grad_norm": 0.33316364884376526,
      "learning_rate": 2e-05,
      "loss": 0.5513,
      "step": 4270
    },
    {
      "epoch": 0.376131470252219,
      "grad_norm": 0.36481890082359314,
      "learning_rate": 2e-05,
      "loss": 0.526,
      "step": 4280
    },
    {
      "epoch": 0.3770102820986027,
      "grad_norm": 0.3915199041366577,
      "learning_rate": 2e-05,
      "loss": 0.5419,
      "step": 4290
    },
    {
      "epoch": 0.3778890939449864,
      "grad_norm": 0.35840851068496704,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 4300
    },
    {
      "epoch": 0.37876790579137004,
      "grad_norm": 0.3246406614780426,
      "learning_rate": 2e-05,
      "loss": 0.5471,
      "step": 4310
    },
    {
      "epoch": 0.37964671763775376,
      "grad_norm": 0.34937310218811035,
      "learning_rate": 2e-05,
      "loss": 0.5584,
      "step": 4320
    },
    {
      "epoch": 0.3805255294841374,
      "grad_norm": 0.343380331993103,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 4330
    },
    {
      "epoch": 0.38140434133052115,
      "grad_norm": 0.3456110954284668,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 4340
    },
    {
      "epoch": 0.3822831531769048,
      "grad_norm": 0.3094264268875122,
      "learning_rate": 2e-05,
      "loss": 0.5635,
      "step": 4350
    },
    {
      "epoch": 0.38316196502328853,
      "grad_norm": 0.3686951994895935,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 4360
    },
    {
      "epoch": 0.3840407768696722,
      "grad_norm": 0.38977304100990295,
      "learning_rate": 2e-05,
      "loss": 0.5718,
      "step": 4370
    },
    {
      "epoch": 0.3849195887160559,
      "grad_norm": 0.3388058543205261,
      "learning_rate": 2e-05,
      "loss": 0.5264,
      "step": 4380
    },
    {
      "epoch": 0.3857984005624396,
      "grad_norm": 0.34327635169029236,
      "learning_rate": 2e-05,
      "loss": 0.5526,
      "step": 4390
    },
    {
      "epoch": 0.38667721240882325,
      "grad_norm": 0.34426259994506836,
      "learning_rate": 2e-05,
      "loss": 0.5579,
      "step": 4400
    },
    {
      "epoch": 0.38755602425520697,
      "grad_norm": 0.3414023220539093,
      "learning_rate": 2e-05,
      "loss": 0.5582,
      "step": 4410
    },
    {
      "epoch": 0.38843483610159063,
      "grad_norm": 0.9653606414794922,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 4420
    },
    {
      "epoch": 0.38931364794797435,
      "grad_norm": 0.3341502249240875,
      "learning_rate": 2e-05,
      "loss": 0.5095,
      "step": 4430
    },
    {
      "epoch": 0.390192459794358,
      "grad_norm": 0.3617326617240906,
      "learning_rate": 2e-05,
      "loss": 0.5432,
      "step": 4440
    },
    {
      "epoch": 0.39107127164074174,
      "grad_norm": 0.426432728767395,
      "learning_rate": 2e-05,
      "loss": 0.5465,
      "step": 4450
    },
    {
      "epoch": 0.3919500834871254,
      "grad_norm": 0.3690936863422394,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 4460
    },
    {
      "epoch": 0.3928288953335091,
      "grad_norm": 0.35639163851737976,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 4470
    },
    {
      "epoch": 0.3937077071798928,
      "grad_norm": 0.3225020170211792,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 4480
    },
    {
      "epoch": 0.39458651902627645,
      "grad_norm": 0.31979331374168396,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 4490
    },
    {
      "epoch": 0.39546533087266017,
      "grad_norm": 0.3583161532878876,
      "learning_rate": 2e-05,
      "loss": 0.5784,
      "step": 4500
    },
    {
      "epoch": 0.39634414271904383,
      "grad_norm": 0.35459017753601074,
      "learning_rate": 2e-05,
      "loss": 0.5343,
      "step": 4510
    },
    {
      "epoch": 0.39722295456542756,
      "grad_norm": 0.31602829694747925,
      "learning_rate": 2e-05,
      "loss": 0.5378,
      "step": 4520
    },
    {
      "epoch": 0.3981017664118112,
      "grad_norm": 0.317524254322052,
      "learning_rate": 2e-05,
      "loss": 0.5525,
      "step": 4530
    },
    {
      "epoch": 0.39898057825819494,
      "grad_norm": 0.33391332626342773,
      "learning_rate": 2e-05,
      "loss": 0.5424,
      "step": 4540
    },
    {
      "epoch": 0.3998593901045786,
      "grad_norm": 0.3520185351371765,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 4550
    },
    {
      "epoch": 0.4007382019509623,
      "grad_norm": 0.5122516751289368,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 4560
    },
    {
      "epoch": 0.401617013797346,
      "grad_norm": 0.36817750334739685,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 4570
    },
    {
      "epoch": 0.40249582564372965,
      "grad_norm": 0.34956347942352295,
      "learning_rate": 2e-05,
      "loss": 0.5505,
      "step": 4580
    },
    {
      "epoch": 0.4033746374901134,
      "grad_norm": 0.3491576910018921,
      "learning_rate": 2e-05,
      "loss": 0.5245,
      "step": 4590
    },
    {
      "epoch": 0.40425344933649704,
      "grad_norm": 0.3160029947757721,
      "learning_rate": 2e-05,
      "loss": 0.5641,
      "step": 4600
    },
    {
      "epoch": 0.40513226118288076,
      "grad_norm": 0.3337414264678955,
      "learning_rate": 2e-05,
      "loss": 0.532,
      "step": 4610
    },
    {
      "epoch": 0.4060110730292644,
      "grad_norm": 0.3471808135509491,
      "learning_rate": 2e-05,
      "loss": 0.5491,
      "step": 4620
    },
    {
      "epoch": 0.40688988487564814,
      "grad_norm": 0.3394460678100586,
      "learning_rate": 2e-05,
      "loss": 0.5809,
      "step": 4630
    },
    {
      "epoch": 0.4077686967220318,
      "grad_norm": 0.3804316818714142,
      "learning_rate": 2e-05,
      "loss": 0.5703,
      "step": 4640
    },
    {
      "epoch": 0.40864750856841553,
      "grad_norm": 0.3334446847438812,
      "learning_rate": 2e-05,
      "loss": 0.5604,
      "step": 4650
    },
    {
      "epoch": 0.4095263204147992,
      "grad_norm": 0.3258700966835022,
      "learning_rate": 2e-05,
      "loss": 0.5511,
      "step": 4660
    },
    {
      "epoch": 0.41040513226118286,
      "grad_norm": 0.3032234311103821,
      "learning_rate": 2e-05,
      "loss": 0.5639,
      "step": 4670
    },
    {
      "epoch": 0.4112839441075666,
      "grad_norm": 0.3459280729293823,
      "learning_rate": 2e-05,
      "loss": 0.5547,
      "step": 4680
    },
    {
      "epoch": 0.41216275595395024,
      "grad_norm": 0.36971738934516907,
      "learning_rate": 2e-05,
      "loss": 0.5683,
      "step": 4690
    },
    {
      "epoch": 0.41304156780033396,
      "grad_norm": 0.420030802488327,
      "learning_rate": 2e-05,
      "loss": 0.569,
      "step": 4700
    },
    {
      "epoch": 0.41392037964671763,
      "grad_norm": 0.4020545184612274,
      "learning_rate": 2e-05,
      "loss": 0.5739,
      "step": 4710
    },
    {
      "epoch": 0.41479919149310135,
      "grad_norm": 0.3419502377510071,
      "learning_rate": 2e-05,
      "loss": 0.5152,
      "step": 4720
    },
    {
      "epoch": 0.415678003339485,
      "grad_norm": 0.38680633902549744,
      "learning_rate": 2e-05,
      "loss": 0.5336,
      "step": 4730
    },
    {
      "epoch": 0.4165568151858687,
      "grad_norm": 0.4457762539386749,
      "learning_rate": 2e-05,
      "loss": 0.5492,
      "step": 4740
    },
    {
      "epoch": 0.4174356270322524,
      "grad_norm": 0.3519948422908783,
      "learning_rate": 2e-05,
      "loss": 0.5639,
      "step": 4750
    },
    {
      "epoch": 0.41831443887863606,
      "grad_norm": 0.34173524379730225,
      "learning_rate": 2e-05,
      "loss": 0.5651,
      "step": 4760
    },
    {
      "epoch": 0.4191932507250198,
      "grad_norm": 0.36172765493392944,
      "learning_rate": 2e-05,
      "loss": 0.5579,
      "step": 4770
    },
    {
      "epoch": 0.42007206257140345,
      "grad_norm": 0.3582833409309387,
      "learning_rate": 2e-05,
      "loss": 0.5544,
      "step": 4780
    },
    {
      "epoch": 0.42095087441778717,
      "grad_norm": 0.3736724853515625,
      "learning_rate": 2e-05,
      "loss": 0.5195,
      "step": 4790
    },
    {
      "epoch": 0.42182968626417083,
      "grad_norm": 0.3538292348384857,
      "learning_rate": 2e-05,
      "loss": 0.5675,
      "step": 4800
    },
    {
      "epoch": 0.42270849811055455,
      "grad_norm": 0.39768341183662415,
      "learning_rate": 2e-05,
      "loss": 0.5742,
      "step": 4810
    },
    {
      "epoch": 0.4235873099569382,
      "grad_norm": 0.33936232328414917,
      "learning_rate": 2e-05,
      "loss": 0.5312,
      "step": 4820
    },
    {
      "epoch": 0.4244661218033219,
      "grad_norm": 0.37284401059150696,
      "learning_rate": 2e-05,
      "loss": 0.5649,
      "step": 4830
    },
    {
      "epoch": 0.4253449336497056,
      "grad_norm": 0.3544437885284424,
      "learning_rate": 2e-05,
      "loss": 0.5348,
      "step": 4840
    },
    {
      "epoch": 0.42622374549608927,
      "grad_norm": 0.674677312374115,
      "learning_rate": 2e-05,
      "loss": 0.5324,
      "step": 4850
    },
    {
      "epoch": 0.427102557342473,
      "grad_norm": 0.36129504442214966,
      "learning_rate": 2e-05,
      "loss": 0.5348,
      "step": 4860
    },
    {
      "epoch": 0.42798136918885665,
      "grad_norm": 0.32914307713508606,
      "learning_rate": 2e-05,
      "loss": 0.5573,
      "step": 4870
    },
    {
      "epoch": 0.4288601810352404,
      "grad_norm": 0.36696475744247437,
      "learning_rate": 2e-05,
      "loss": 0.5206,
      "step": 4880
    },
    {
      "epoch": 0.42973899288162404,
      "grad_norm": 0.3382825553417206,
      "learning_rate": 2e-05,
      "loss": 0.5502,
      "step": 4890
    },
    {
      "epoch": 0.43061780472800776,
      "grad_norm": 0.336336612701416,
      "learning_rate": 2e-05,
      "loss": 0.5729,
      "step": 4900
    },
    {
      "epoch": 0.4314966165743914,
      "grad_norm": 0.3556782603263855,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 4910
    },
    {
      "epoch": 0.4323754284207751,
      "grad_norm": 0.36311811208724976,
      "learning_rate": 2e-05,
      "loss": 0.5507,
      "step": 4920
    },
    {
      "epoch": 0.4332542402671588,
      "grad_norm": 0.38301485776901245,
      "learning_rate": 2e-05,
      "loss": 0.5546,
      "step": 4930
    },
    {
      "epoch": 0.4341330521135425,
      "grad_norm": 0.39091095328330994,
      "learning_rate": 2e-05,
      "loss": 0.5626,
      "step": 4940
    },
    {
      "epoch": 0.4350118639599262,
      "grad_norm": 0.36095932126045227,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 4950
    },
    {
      "epoch": 0.43589067580630986,
      "grad_norm": 0.35853439569473267,
      "learning_rate": 2e-05,
      "loss": 0.5512,
      "step": 4960
    },
    {
      "epoch": 0.4367694876526936,
      "grad_norm": 0.3357088565826416,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 4970
    },
    {
      "epoch": 0.43764829949907724,
      "grad_norm": 0.40412941575050354,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 4980
    },
    {
      "epoch": 0.43852711134546096,
      "grad_norm": 0.3572874069213867,
      "learning_rate": 2e-05,
      "loss": 0.5468,
      "step": 4990
    },
    {
      "epoch": 0.4394059231918446,
      "grad_norm": 0.3652222454547882,
      "learning_rate": 2e-05,
      "loss": 0.5473,
      "step": 5000
    },
    {
      "epoch": 0.4402847350382283,
      "grad_norm": 0.3919486105442047,
      "learning_rate": 2e-05,
      "loss": 0.5324,
      "step": 5010
    },
    {
      "epoch": 0.441163546884612,
      "grad_norm": 0.3372253477573395,
      "learning_rate": 2e-05,
      "loss": 0.5541,
      "step": 5020
    },
    {
      "epoch": 0.4420423587309957,
      "grad_norm": 0.43406522274017334,
      "learning_rate": 2e-05,
      "loss": 0.5449,
      "step": 5030
    },
    {
      "epoch": 0.4429211705773794,
      "grad_norm": 0.6748683452606201,
      "learning_rate": 2e-05,
      "loss": 0.5772,
      "step": 5040
    },
    {
      "epoch": 0.44379998242376306,
      "grad_norm": 0.4234529137611389,
      "learning_rate": 2e-05,
      "loss": 0.5642,
      "step": 5050
    },
    {
      "epoch": 0.4446787942701468,
      "grad_norm": 0.3501267731189728,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 5060
    },
    {
      "epoch": 0.44555760611653045,
      "grad_norm": 0.3795582056045532,
      "learning_rate": 2e-05,
      "loss": 0.5206,
      "step": 5070
    },
    {
      "epoch": 0.44643641796291417,
      "grad_norm": 0.33441901206970215,
      "learning_rate": 2e-05,
      "loss": 0.5596,
      "step": 5080
    },
    {
      "epoch": 0.44731522980929783,
      "grad_norm": 0.34686899185180664,
      "learning_rate": 2e-05,
      "loss": 0.5082,
      "step": 5090
    },
    {
      "epoch": 0.4481940416556815,
      "grad_norm": 0.3675048053264618,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 5100
    },
    {
      "epoch": 0.4490728535020652,
      "grad_norm": 0.3247295320034027,
      "learning_rate": 2e-05,
      "loss": 0.5246,
      "step": 5110
    },
    {
      "epoch": 0.4499516653484489,
      "grad_norm": 0.4007485508918762,
      "learning_rate": 2e-05,
      "loss": 0.5313,
      "step": 5120
    },
    {
      "epoch": 0.4508304771948326,
      "grad_norm": 0.34064239263534546,
      "learning_rate": 2e-05,
      "loss": 0.5482,
      "step": 5130
    },
    {
      "epoch": 0.45170928904121627,
      "grad_norm": 0.42627081274986267,
      "learning_rate": 2e-05,
      "loss": 0.5285,
      "step": 5140
    },
    {
      "epoch": 0.4525881008876,
      "grad_norm": 0.387762188911438,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 5150
    },
    {
      "epoch": 0.45346691273398365,
      "grad_norm": 0.33811521530151367,
      "learning_rate": 2e-05,
      "loss": 0.5534,
      "step": 5160
    },
    {
      "epoch": 0.4543457245803673,
      "grad_norm": 0.38338661193847656,
      "learning_rate": 2e-05,
      "loss": 0.5562,
      "step": 5170
    },
    {
      "epoch": 0.45522453642675104,
      "grad_norm": 0.3450353443622589,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 5180
    },
    {
      "epoch": 0.4561033482731347,
      "grad_norm": 0.3360109329223633,
      "learning_rate": 2e-05,
      "loss": 0.5645,
      "step": 5190
    },
    {
      "epoch": 0.4569821601195184,
      "grad_norm": 0.3548944294452667,
      "learning_rate": 2e-05,
      "loss": 0.5424,
      "step": 5200
    },
    {
      "epoch": 0.4578609719659021,
      "grad_norm": 0.3279622793197632,
      "learning_rate": 2e-05,
      "loss": 0.5071,
      "step": 5210
    },
    {
      "epoch": 0.4587397838122858,
      "grad_norm": 0.37016528844833374,
      "learning_rate": 2e-05,
      "loss": 0.5396,
      "step": 5220
    },
    {
      "epoch": 0.45961859565866947,
      "grad_norm": 0.338382363319397,
      "learning_rate": 2e-05,
      "loss": 0.5479,
      "step": 5230
    },
    {
      "epoch": 0.4604974075050532,
      "grad_norm": 0.37774530053138733,
      "learning_rate": 2e-05,
      "loss": 0.5343,
      "step": 5240
    },
    {
      "epoch": 0.46137621935143686,
      "grad_norm": 0.36135753989219666,
      "learning_rate": 2e-05,
      "loss": 0.5586,
      "step": 5250
    },
    {
      "epoch": 0.4622550311978205,
      "grad_norm": 0.35616615414619446,
      "learning_rate": 2e-05,
      "loss": 0.5648,
      "step": 5260
    },
    {
      "epoch": 0.46313384304420424,
      "grad_norm": 0.3263412117958069,
      "learning_rate": 2e-05,
      "loss": 0.5547,
      "step": 5270
    },
    {
      "epoch": 0.4640126548905879,
      "grad_norm": 0.35563138127326965,
      "learning_rate": 2e-05,
      "loss": 0.5665,
      "step": 5280
    },
    {
      "epoch": 0.4648914667369716,
      "grad_norm": 0.380193829536438,
      "learning_rate": 2e-05,
      "loss": 0.5387,
      "step": 5290
    },
    {
      "epoch": 0.4657702785833553,
      "grad_norm": 0.3608831465244293,
      "learning_rate": 2e-05,
      "loss": 0.53,
      "step": 5300
    },
    {
      "epoch": 0.466649090429739,
      "grad_norm": 0.3129928410053253,
      "learning_rate": 2e-05,
      "loss": 0.5307,
      "step": 5310
    },
    {
      "epoch": 0.4675279022761227,
      "grad_norm": 0.33062058687210083,
      "learning_rate": 2e-05,
      "loss": 0.522,
      "step": 5320
    },
    {
      "epoch": 0.4684067141225064,
      "grad_norm": 0.34377557039260864,
      "learning_rate": 2e-05,
      "loss": 0.5384,
      "step": 5330
    },
    {
      "epoch": 0.46928552596889006,
      "grad_norm": 0.3388850688934326,
      "learning_rate": 2e-05,
      "loss": 0.5495,
      "step": 5340
    },
    {
      "epoch": 0.4701643378152737,
      "grad_norm": 0.4327860176563263,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 5350
    },
    {
      "epoch": 0.47104314966165745,
      "grad_norm": 0.36854660511016846,
      "learning_rate": 2e-05,
      "loss": 0.542,
      "step": 5360
    },
    {
      "epoch": 0.4719219615080411,
      "grad_norm": 0.36846476793289185,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 5370
    },
    {
      "epoch": 0.47280077335442483,
      "grad_norm": 0.34063291549682617,
      "learning_rate": 2e-05,
      "loss": 0.5178,
      "step": 5380
    },
    {
      "epoch": 0.4736795852008085,
      "grad_norm": 0.39007407426834106,
      "learning_rate": 2e-05,
      "loss": 0.5503,
      "step": 5390
    },
    {
      "epoch": 0.4745583970471922,
      "grad_norm": 0.34226930141448975,
      "learning_rate": 2e-05,
      "loss": 0.5597,
      "step": 5400
    },
    {
      "epoch": 0.4754372088935759,
      "grad_norm": 0.35507798194885254,
      "learning_rate": 2e-05,
      "loss": 0.5424,
      "step": 5410
    },
    {
      "epoch": 0.4763160207399596,
      "grad_norm": 0.41991424560546875,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 5420
    },
    {
      "epoch": 0.47719483258634326,
      "grad_norm": 0.353799432516098,
      "learning_rate": 2e-05,
      "loss": 0.5466,
      "step": 5430
    },
    {
      "epoch": 0.47807364443272693,
      "grad_norm": 0.385633647441864,
      "learning_rate": 2e-05,
      "loss": 0.5601,
      "step": 5440
    },
    {
      "epoch": 0.47895245627911065,
      "grad_norm": 0.40804630517959595,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 5450
    },
    {
      "epoch": 0.4798312681254943,
      "grad_norm": 0.3325469493865967,
      "learning_rate": 2e-05,
      "loss": 0.5251,
      "step": 5460
    },
    {
      "epoch": 0.48071007997187803,
      "grad_norm": 0.35576412081718445,
      "learning_rate": 2e-05,
      "loss": 0.5284,
      "step": 5470
    },
    {
      "epoch": 0.4815888918182617,
      "grad_norm": 0.36081069707870483,
      "learning_rate": 2e-05,
      "loss": 0.5641,
      "step": 5480
    },
    {
      "epoch": 0.4824677036646454,
      "grad_norm": 0.3600965440273285,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 5490
    },
    {
      "epoch": 0.4833465155110291,
      "grad_norm": 0.3926527798175812,
      "learning_rate": 2e-05,
      "loss": 0.5309,
      "step": 5500
    },
    {
      "epoch": 0.4842253273574128,
      "grad_norm": 0.3193684220314026,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 5510
    },
    {
      "epoch": 0.48510413920379647,
      "grad_norm": 0.42473888397216797,
      "learning_rate": 2e-05,
      "loss": 0.5285,
      "step": 5520
    },
    {
      "epoch": 0.48598295105018013,
      "grad_norm": 0.32970696687698364,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 5530
    },
    {
      "epoch": 0.48686176289656385,
      "grad_norm": 0.3121568262577057,
      "learning_rate": 2e-05,
      "loss": 0.5285,
      "step": 5540
    },
    {
      "epoch": 0.4877405747429475,
      "grad_norm": 0.32342109084129333,
      "learning_rate": 2e-05,
      "loss": 0.5257,
      "step": 5550
    },
    {
      "epoch": 0.48861938658933124,
      "grad_norm": 0.37905725836753845,
      "learning_rate": 2e-05,
      "loss": 0.5136,
      "step": 5560
    },
    {
      "epoch": 0.4894981984357149,
      "grad_norm": 0.4027540683746338,
      "learning_rate": 2e-05,
      "loss": 0.5625,
      "step": 5570
    },
    {
      "epoch": 0.4903770102820986,
      "grad_norm": 0.328654021024704,
      "learning_rate": 2e-05,
      "loss": 0.5692,
      "step": 5580
    },
    {
      "epoch": 0.4912558221284823,
      "grad_norm": 0.35865482687950134,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 5590
    },
    {
      "epoch": 0.49213463397486595,
      "grad_norm": 0.34479567408561707,
      "learning_rate": 2e-05,
      "loss": 0.5636,
      "step": 5600
    },
    {
      "epoch": 0.4930134458212497,
      "grad_norm": 0.45242613554000854,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 5610
    },
    {
      "epoch": 0.49389225766763334,
      "grad_norm": 0.3208048641681671,
      "learning_rate": 2e-05,
      "loss": 0.4967,
      "step": 5620
    },
    {
      "epoch": 0.49477106951401706,
      "grad_norm": 0.35941752791404724,
      "learning_rate": 2e-05,
      "loss": 0.524,
      "step": 5630
    },
    {
      "epoch": 0.4956498813604007,
      "grad_norm": 0.3228263854980469,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 5640
    },
    {
      "epoch": 0.49652869320678444,
      "grad_norm": 0.3667476773262024,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 5650
    },
    {
      "epoch": 0.4974075050531681,
      "grad_norm": 0.4038882255554199,
      "learning_rate": 2e-05,
      "loss": 0.562,
      "step": 5660
    },
    {
      "epoch": 0.49828631689955183,
      "grad_norm": 0.322954922914505,
      "learning_rate": 2e-05,
      "loss": 0.5331,
      "step": 5670
    },
    {
      "epoch": 0.4991651287459355,
      "grad_norm": 0.35359811782836914,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 5680
    },
    {
      "epoch": 0.5000439405923192,
      "grad_norm": 0.3503968119621277,
      "learning_rate": 2e-05,
      "loss": 0.5487,
      "step": 5690
    },
    {
      "epoch": 0.5009227524387029,
      "grad_norm": 0.34654828906059265,
      "learning_rate": 2e-05,
      "loss": 0.5238,
      "step": 5700
    },
    {
      "epoch": 0.5018015642850866,
      "grad_norm": 0.33823567628860474,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 5710
    },
    {
      "epoch": 0.5026803761314702,
      "grad_norm": 0.3126208186149597,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 5720
    },
    {
      "epoch": 0.5035591879778539,
      "grad_norm": 0.39850226044654846,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 5730
    },
    {
      "epoch": 0.5044379998242376,
      "grad_norm": 0.33085793256759644,
      "learning_rate": 2e-05,
      "loss": 0.5378,
      "step": 5740
    },
    {
      "epoch": 0.5053168116706214,
      "grad_norm": 0.34183770418167114,
      "learning_rate": 2e-05,
      "loss": 0.5801,
      "step": 5750
    },
    {
      "epoch": 0.506195623517005,
      "grad_norm": 0.3824526369571686,
      "learning_rate": 2e-05,
      "loss": 0.5577,
      "step": 5760
    },
    {
      "epoch": 0.5070744353633887,
      "grad_norm": 0.3536856770515442,
      "learning_rate": 2e-05,
      "loss": 0.555,
      "step": 5770
    },
    {
      "epoch": 0.5079532472097724,
      "grad_norm": 0.3598324954509735,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 5780
    },
    {
      "epoch": 0.508832059056156,
      "grad_norm": 0.3485099971294403,
      "learning_rate": 2e-05,
      "loss": 0.532,
      "step": 5790
    },
    {
      "epoch": 0.5097108709025397,
      "grad_norm": 0.9100950360298157,
      "learning_rate": 2e-05,
      "loss": 0.5273,
      "step": 5800
    },
    {
      "epoch": 0.5105896827489235,
      "grad_norm": 0.33965620398521423,
      "learning_rate": 2e-05,
      "loss": 0.5603,
      "step": 5810
    },
    {
      "epoch": 0.5114684945953072,
      "grad_norm": 0.39044642448425293,
      "learning_rate": 2e-05,
      "loss": 0.543,
      "step": 5820
    },
    {
      "epoch": 0.5123473064416908,
      "grad_norm": 0.366967111825943,
      "learning_rate": 2e-05,
      "loss": 0.526,
      "step": 5830
    },
    {
      "epoch": 0.5132261182880745,
      "grad_norm": 0.3692724108695984,
      "learning_rate": 2e-05,
      "loss": 0.5362,
      "step": 5840
    },
    {
      "epoch": 0.5141049301344582,
      "grad_norm": 0.3853226602077484,
      "learning_rate": 2e-05,
      "loss": 0.5306,
      "step": 5850
    },
    {
      "epoch": 0.5149837419808418,
      "grad_norm": 0.3449903428554535,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 5860
    },
    {
      "epoch": 0.5158625538272256,
      "grad_norm": 0.31989586353302,
      "learning_rate": 2e-05,
      "loss": 0.5382,
      "step": 5870
    },
    {
      "epoch": 0.5167413656736093,
      "grad_norm": 0.353564977645874,
      "learning_rate": 2e-05,
      "loss": 0.5182,
      "step": 5880
    },
    {
      "epoch": 0.517620177519993,
      "grad_norm": 0.33187544345855713,
      "learning_rate": 2e-05,
      "loss": 0.5089,
      "step": 5890
    },
    {
      "epoch": 0.5184989893663766,
      "grad_norm": 0.37009280920028687,
      "learning_rate": 2e-05,
      "loss": 0.546,
      "step": 5900
    },
    {
      "epoch": 0.5193778012127603,
      "grad_norm": 0.36561277508735657,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 5910
    },
    {
      "epoch": 0.5202566130591441,
      "grad_norm": 0.3187030553817749,
      "learning_rate": 2e-05,
      "loss": 0.5288,
      "step": 5920
    },
    {
      "epoch": 0.5211354249055278,
      "grad_norm": 0.33941999077796936,
      "learning_rate": 2e-05,
      "loss": 0.5108,
      "step": 5930
    },
    {
      "epoch": 0.5220142367519114,
      "grad_norm": 0.3412536680698395,
      "learning_rate": 2e-05,
      "loss": 0.5047,
      "step": 5940
    },
    {
      "epoch": 0.5228930485982951,
      "grad_norm": 0.3213081955909729,
      "learning_rate": 2e-05,
      "loss": 0.5651,
      "step": 5950
    },
    {
      "epoch": 0.5237718604446788,
      "grad_norm": 0.36378201842308044,
      "learning_rate": 2e-05,
      "loss": 0.5199,
      "step": 5960
    },
    {
      "epoch": 0.5246506722910624,
      "grad_norm": 0.35616523027420044,
      "learning_rate": 2e-05,
      "loss": 0.5488,
      "step": 5970
    },
    {
      "epoch": 0.5255294841374462,
      "grad_norm": 0.39030689001083374,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 5980
    },
    {
      "epoch": 0.5264082959838299,
      "grad_norm": 0.35906723141670227,
      "learning_rate": 2e-05,
      "loss": 0.546,
      "step": 5990
    },
    {
      "epoch": 0.5272871078302136,
      "grad_norm": 0.31921133399009705,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 6000
    },
    {
      "epoch": 0.5281659196765972,
      "grad_norm": 0.33798670768737793,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 6010
    },
    {
      "epoch": 0.5290447315229809,
      "grad_norm": 0.3745160400867462,
      "learning_rate": 2e-05,
      "loss": 0.5607,
      "step": 6020
    },
    {
      "epoch": 0.5299235433693646,
      "grad_norm": 0.3481859266757965,
      "learning_rate": 2e-05,
      "loss": 0.5701,
      "step": 6030
    },
    {
      "epoch": 0.5308023552157483,
      "grad_norm": 0.35364973545074463,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 6040
    },
    {
      "epoch": 0.531681167062132,
      "grad_norm": 0.3588162064552307,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 6050
    },
    {
      "epoch": 0.5325599789085157,
      "grad_norm": 0.301149845123291,
      "learning_rate": 2e-05,
      "loss": 0.5532,
      "step": 6060
    },
    {
      "epoch": 0.5334387907548994,
      "grad_norm": 0.3525964915752411,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 6070
    },
    {
      "epoch": 0.534317602601283,
      "grad_norm": 0.33862653374671936,
      "learning_rate": 2e-05,
      "loss": 0.5222,
      "step": 6080
    },
    {
      "epoch": 0.5351964144476667,
      "grad_norm": 0.40301474928855896,
      "learning_rate": 2e-05,
      "loss": 0.5479,
      "step": 6090
    },
    {
      "epoch": 0.5360752262940505,
      "grad_norm": 0.3276920020580292,
      "learning_rate": 2e-05,
      "loss": 0.5236,
      "step": 6100
    },
    {
      "epoch": 0.5369540381404342,
      "grad_norm": 0.38633087277412415,
      "learning_rate": 2e-05,
      "loss": 0.5267,
      "step": 6110
    },
    {
      "epoch": 0.5378328499868178,
      "grad_norm": 0.3037598729133606,
      "learning_rate": 2e-05,
      "loss": 0.5377,
      "step": 6120
    },
    {
      "epoch": 0.5387116618332015,
      "grad_norm": 0.35142871737480164,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 6130
    },
    {
      "epoch": 0.5395904736795852,
      "grad_norm": 0.3544119894504547,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 6140
    },
    {
      "epoch": 0.5404692855259688,
      "grad_norm": 0.3449588119983673,
      "learning_rate": 2e-05,
      "loss": 0.5303,
      "step": 6150
    },
    {
      "epoch": 0.5413480973723526,
      "grad_norm": 0.3818516433238983,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 6160
    },
    {
      "epoch": 0.5422269092187363,
      "grad_norm": 0.3502005338668823,
      "learning_rate": 2e-05,
      "loss": 0.5474,
      "step": 6170
    },
    {
      "epoch": 0.54310572106512,
      "grad_norm": 0.41313424706459045,
      "learning_rate": 2e-05,
      "loss": 0.5396,
      "step": 6180
    },
    {
      "epoch": 0.5439845329115036,
      "grad_norm": 0.34990477561950684,
      "learning_rate": 2e-05,
      "loss": 0.5448,
      "step": 6190
    },
    {
      "epoch": 0.5448633447578873,
      "grad_norm": 0.3134634792804718,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 6200
    },
    {
      "epoch": 0.5457421566042711,
      "grad_norm": 0.3239031136035919,
      "learning_rate": 2e-05,
      "loss": 0.535,
      "step": 6210
    },
    {
      "epoch": 0.5466209684506547,
      "grad_norm": 0.37899941205978394,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 6220
    },
    {
      "epoch": 0.5474997802970384,
      "grad_norm": 0.35871097445487976,
      "learning_rate": 2e-05,
      "loss": 0.5479,
      "step": 6230
    },
    {
      "epoch": 0.5483785921434221,
      "grad_norm": 0.3542439341545105,
      "learning_rate": 2e-05,
      "loss": 0.5351,
      "step": 6240
    },
    {
      "epoch": 0.5492574039898058,
      "grad_norm": 0.384600967168808,
      "learning_rate": 2e-05,
      "loss": 0.5674,
      "step": 6250
    },
    {
      "epoch": 0.5501362158361894,
      "grad_norm": 0.37704843282699585,
      "learning_rate": 2e-05,
      "loss": 0.5558,
      "step": 6260
    },
    {
      "epoch": 0.5510150276825732,
      "grad_norm": 0.3346363604068756,
      "learning_rate": 2e-05,
      "loss": 0.5463,
      "step": 6270
    },
    {
      "epoch": 0.5518938395289569,
      "grad_norm": 0.3544868230819702,
      "learning_rate": 2e-05,
      "loss": 0.534,
      "step": 6280
    },
    {
      "epoch": 0.5527726513753405,
      "grad_norm": 0.369394987821579,
      "learning_rate": 2e-05,
      "loss": 0.5315,
      "step": 6290
    },
    {
      "epoch": 0.5536514632217242,
      "grad_norm": 0.401042640209198,
      "learning_rate": 2e-05,
      "loss": 0.5389,
      "step": 6300
    },
    {
      "epoch": 0.5545302750681079,
      "grad_norm": 0.4422287046909332,
      "learning_rate": 2e-05,
      "loss": 0.5449,
      "step": 6310
    },
    {
      "epoch": 0.5554090869144916,
      "grad_norm": 0.3113273084163666,
      "learning_rate": 2e-05,
      "loss": 0.5309,
      "step": 6320
    },
    {
      "epoch": 0.5562878987608753,
      "grad_norm": 0.34816792607307434,
      "learning_rate": 2e-05,
      "loss": 0.524,
      "step": 6330
    },
    {
      "epoch": 0.557166710607259,
      "grad_norm": 0.45087605714797974,
      "learning_rate": 2e-05,
      "loss": 0.5281,
      "step": 6340
    },
    {
      "epoch": 0.5580455224536427,
      "grad_norm": 0.3481863737106323,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 6350
    },
    {
      "epoch": 0.5589243343000264,
      "grad_norm": 0.33030328154563904,
      "learning_rate": 2e-05,
      "loss": 0.5452,
      "step": 6360
    },
    {
      "epoch": 0.55980314614641,
      "grad_norm": 0.3922591805458069,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 6370
    },
    {
      "epoch": 0.5606819579927937,
      "grad_norm": 0.3547910451889038,
      "learning_rate": 2e-05,
      "loss": 0.5491,
      "step": 6380
    },
    {
      "epoch": 0.5615607698391775,
      "grad_norm": 0.4465925693511963,
      "learning_rate": 2e-05,
      "loss": 0.5413,
      "step": 6390
    },
    {
      "epoch": 0.5624395816855611,
      "grad_norm": 0.3475148379802704,
      "learning_rate": 2e-05,
      "loss": 0.5197,
      "step": 6400
    },
    {
      "epoch": 0.5633183935319448,
      "grad_norm": 0.3544149100780487,
      "learning_rate": 2e-05,
      "loss": 0.53,
      "step": 6410
    },
    {
      "epoch": 0.5641972053783285,
      "grad_norm": 0.3417002260684967,
      "learning_rate": 2e-05,
      "loss": 0.5526,
      "step": 6420
    },
    {
      "epoch": 0.5650760172247122,
      "grad_norm": 0.3288358747959137,
      "learning_rate": 2e-05,
      "loss": 0.5073,
      "step": 6430
    },
    {
      "epoch": 0.5659548290710958,
      "grad_norm": 0.3840493857860565,
      "learning_rate": 2e-05,
      "loss": 0.5387,
      "step": 6440
    },
    {
      "epoch": 0.5668336409174796,
      "grad_norm": 0.378919780254364,
      "learning_rate": 2e-05,
      "loss": 0.5377,
      "step": 6450
    },
    {
      "epoch": 0.5677124527638633,
      "grad_norm": 0.3764628767967224,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 6460
    },
    {
      "epoch": 0.5685912646102469,
      "grad_norm": 0.35947051644325256,
      "learning_rate": 2e-05,
      "loss": 0.5318,
      "step": 6470
    },
    {
      "epoch": 0.5694700764566306,
      "grad_norm": 0.3734472990036011,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 6480
    },
    {
      "epoch": 0.5703488883030143,
      "grad_norm": 0.33991938829421997,
      "learning_rate": 2e-05,
      "loss": 0.5253,
      "step": 6490
    },
    {
      "epoch": 0.571227700149398,
      "grad_norm": 0.3635921776294708,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 6500
    },
    {
      "epoch": 0.5721065119957817,
      "grad_norm": 0.372970849275589,
      "learning_rate": 2e-05,
      "loss": 0.5574,
      "step": 6510
    },
    {
      "epoch": 0.5729853238421654,
      "grad_norm": 0.36239272356033325,
      "learning_rate": 2e-05,
      "loss": 0.5213,
      "step": 6520
    },
    {
      "epoch": 0.5738641356885491,
      "grad_norm": 0.37176117300987244,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 6530
    },
    {
      "epoch": 0.5747429475349328,
      "grad_norm": 0.35930293798446655,
      "learning_rate": 2e-05,
      "loss": 0.5521,
      "step": 6540
    },
    {
      "epoch": 0.5756217593813164,
      "grad_norm": 0.3506200611591339,
      "learning_rate": 2e-05,
      "loss": 0.526,
      "step": 6550
    },
    {
      "epoch": 0.5765005712277002,
      "grad_norm": 0.3321985900402069,
      "learning_rate": 2e-05,
      "loss": 0.5424,
      "step": 6560
    },
    {
      "epoch": 0.5773793830740839,
      "grad_norm": 0.3998512923717499,
      "learning_rate": 2e-05,
      "loss": 0.5373,
      "step": 6570
    },
    {
      "epoch": 0.5782581949204675,
      "grad_norm": 0.3990097939968109,
      "learning_rate": 2e-05,
      "loss": 0.5327,
      "step": 6580
    },
    {
      "epoch": 0.5791370067668512,
      "grad_norm": 0.42080727219581604,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 6590
    },
    {
      "epoch": 0.5800158186132349,
      "grad_norm": 0.3026658296585083,
      "learning_rate": 2e-05,
      "loss": 0.5405,
      "step": 6600
    },
    {
      "epoch": 0.5808946304596186,
      "grad_norm": 0.4440387785434723,
      "learning_rate": 2e-05,
      "loss": 0.5334,
      "step": 6610
    },
    {
      "epoch": 0.5817734423060023,
      "grad_norm": 0.37209415435791016,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 6620
    },
    {
      "epoch": 0.582652254152386,
      "grad_norm": 0.3903203308582306,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 6630
    },
    {
      "epoch": 0.5835310659987697,
      "grad_norm": 0.3610727787017822,
      "learning_rate": 2e-05,
      "loss": 0.5505,
      "step": 6640
    },
    {
      "epoch": 0.5844098778451533,
      "grad_norm": 0.407707542181015,
      "learning_rate": 2e-05,
      "loss": 0.5381,
      "step": 6650
    },
    {
      "epoch": 0.585288689691537,
      "grad_norm": 0.3117853105068207,
      "learning_rate": 2e-05,
      "loss": 0.543,
      "step": 6660
    },
    {
      "epoch": 0.5861675015379207,
      "grad_norm": 0.4043383300304413,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 6670
    },
    {
      "epoch": 0.5870463133843045,
      "grad_norm": 0.3618474304676056,
      "learning_rate": 2e-05,
      "loss": 0.5334,
      "step": 6680
    },
    {
      "epoch": 0.5879251252306881,
      "grad_norm": 0.3382108509540558,
      "learning_rate": 2e-05,
      "loss": 0.5166,
      "step": 6690
    },
    {
      "epoch": 0.5888039370770718,
      "grad_norm": 0.44376009702682495,
      "learning_rate": 2e-05,
      "loss": 0.5625,
      "step": 6700
    },
    {
      "epoch": 0.5896827489234555,
      "grad_norm": 0.73813796043396,
      "learning_rate": 2e-05,
      "loss": 0.5159,
      "step": 6710
    },
    {
      "epoch": 0.5905615607698391,
      "grad_norm": 0.3754452168941498,
      "learning_rate": 2e-05,
      "loss": 0.5331,
      "step": 6720
    },
    {
      "epoch": 0.5914403726162228,
      "grad_norm": 0.35456764698028564,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 6730
    },
    {
      "epoch": 0.5923191844626066,
      "grad_norm": 0.38001108169555664,
      "learning_rate": 2e-05,
      "loss": 0.5444,
      "step": 6740
    },
    {
      "epoch": 0.5931979963089903,
      "grad_norm": 0.3459557890892029,
      "learning_rate": 2e-05,
      "loss": 0.5254,
      "step": 6750
    },
    {
      "epoch": 0.5940768081553739,
      "grad_norm": 0.3378278613090515,
      "learning_rate": 2e-05,
      "loss": 0.5221,
      "step": 6760
    },
    {
      "epoch": 0.5949556200017576,
      "grad_norm": 0.34229087829589844,
      "learning_rate": 2e-05,
      "loss": 0.5214,
      "step": 6770
    },
    {
      "epoch": 0.5958344318481413,
      "grad_norm": 0.37862640619277954,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 6780
    },
    {
      "epoch": 0.596713243694525,
      "grad_norm": 0.38429781794548035,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 6790
    },
    {
      "epoch": 0.5975920555409087,
      "grad_norm": 0.36466020345687866,
      "learning_rate": 2e-05,
      "loss": 0.5404,
      "step": 6800
    },
    {
      "epoch": 0.5984708673872924,
      "grad_norm": 0.3793875575065613,
      "learning_rate": 2e-05,
      "loss": 0.5231,
      "step": 6810
    },
    {
      "epoch": 0.5993496792336761,
      "grad_norm": 0.39953622221946716,
      "learning_rate": 2e-05,
      "loss": 0.5283,
      "step": 6820
    },
    {
      "epoch": 0.6002284910800597,
      "grad_norm": 0.364557683467865,
      "learning_rate": 2e-05,
      "loss": 0.5265,
      "step": 6830
    },
    {
      "epoch": 0.6011073029264434,
      "grad_norm": 0.35580718517303467,
      "learning_rate": 2e-05,
      "loss": 0.5247,
      "step": 6840
    },
    {
      "epoch": 0.6019861147728272,
      "grad_norm": 0.3517809510231018,
      "learning_rate": 2e-05,
      "loss": 0.5511,
      "step": 6850
    },
    {
      "epoch": 0.6028649266192109,
      "grad_norm": 0.38442671298980713,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 6860
    },
    {
      "epoch": 0.6037437384655945,
      "grad_norm": 0.34150364995002747,
      "learning_rate": 2e-05,
      "loss": 0.5244,
      "step": 6870
    },
    {
      "epoch": 0.6046225503119782,
      "grad_norm": 0.36824026703834534,
      "learning_rate": 2e-05,
      "loss": 0.541,
      "step": 6880
    },
    {
      "epoch": 0.6055013621583619,
      "grad_norm": 0.3460348844528198,
      "learning_rate": 2e-05,
      "loss": 0.5322,
      "step": 6890
    },
    {
      "epoch": 0.6063801740047455,
      "grad_norm": 0.32730546593666077,
      "learning_rate": 2e-05,
      "loss": 0.523,
      "step": 6900
    },
    {
      "epoch": 0.6072589858511293,
      "grad_norm": 0.3591601848602295,
      "learning_rate": 2e-05,
      "loss": 0.5124,
      "step": 6910
    },
    {
      "epoch": 0.608137797697513,
      "grad_norm": 0.3623739778995514,
      "learning_rate": 2e-05,
      "loss": 0.5281,
      "step": 6920
    },
    {
      "epoch": 0.6090166095438967,
      "grad_norm": 0.3208017349243164,
      "learning_rate": 2e-05,
      "loss": 0.5311,
      "step": 6930
    },
    {
      "epoch": 0.6098954213902803,
      "grad_norm": 0.3618086576461792,
      "learning_rate": 2e-05,
      "loss": 0.5298,
      "step": 6940
    },
    {
      "epoch": 0.610774233236664,
      "grad_norm": 0.363064706325531,
      "learning_rate": 2e-05,
      "loss": 0.5652,
      "step": 6950
    },
    {
      "epoch": 0.6116530450830477,
      "grad_norm": 0.32000935077667236,
      "learning_rate": 2e-05,
      "loss": 0.5255,
      "step": 6960
    },
    {
      "epoch": 0.6125318569294315,
      "grad_norm": 0.3850856125354767,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 6970
    },
    {
      "epoch": 0.6134106687758151,
      "grad_norm": 0.3467339277267456,
      "learning_rate": 2e-05,
      "loss": 0.551,
      "step": 6980
    },
    {
      "epoch": 0.6142894806221988,
      "grad_norm": 0.4356135427951813,
      "learning_rate": 2e-05,
      "loss": 0.5137,
      "step": 6990
    },
    {
      "epoch": 0.6151682924685825,
      "grad_norm": 0.3162763714790344,
      "learning_rate": 2e-05,
      "loss": 0.5405,
      "step": 7000
    },
    {
      "epoch": 0.6160471043149661,
      "grad_norm": 0.39379218220710754,
      "learning_rate": 2e-05,
      "loss": 0.5365,
      "step": 7010
    },
    {
      "epoch": 0.6169259161613498,
      "grad_norm": 0.383926123380661,
      "learning_rate": 2e-05,
      "loss": 0.5436,
      "step": 7020
    },
    {
      "epoch": 0.6178047280077336,
      "grad_norm": 0.37739527225494385,
      "learning_rate": 2e-05,
      "loss": 0.5533,
      "step": 7030
    },
    {
      "epoch": 0.6186835398541173,
      "grad_norm": 0.3387387990951538,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 7040
    },
    {
      "epoch": 0.6195623517005009,
      "grad_norm": 0.3752298355102539,
      "learning_rate": 2e-05,
      "loss": 0.5294,
      "step": 7050
    },
    {
      "epoch": 0.6204411635468846,
      "grad_norm": 0.4204939305782318,
      "learning_rate": 2e-05,
      "loss": 0.5452,
      "step": 7060
    },
    {
      "epoch": 0.6213199753932683,
      "grad_norm": 0.38599392771720886,
      "learning_rate": 2e-05,
      "loss": 0.5417,
      "step": 7070
    },
    {
      "epoch": 0.6221987872396519,
      "grad_norm": 0.3572133779525757,
      "learning_rate": 2e-05,
      "loss": 0.5147,
      "step": 7080
    },
    {
      "epoch": 0.6230775990860357,
      "grad_norm": 0.38146448135375977,
      "learning_rate": 2e-05,
      "loss": 0.5096,
      "step": 7090
    },
    {
      "epoch": 0.6239564109324194,
      "grad_norm": 0.33024948835372925,
      "learning_rate": 2e-05,
      "loss": 0.519,
      "step": 7100
    },
    {
      "epoch": 0.6248352227788031,
      "grad_norm": 0.3807598650455475,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 7110
    },
    {
      "epoch": 0.6257140346251867,
      "grad_norm": 0.35228031873703003,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 7120
    },
    {
      "epoch": 0.6265928464715704,
      "grad_norm": 0.3528590500354767,
      "learning_rate": 2e-05,
      "loss": 0.5505,
      "step": 7130
    },
    {
      "epoch": 0.6274716583179542,
      "grad_norm": 0.31893283128738403,
      "learning_rate": 2e-05,
      "loss": 0.5241,
      "step": 7140
    },
    {
      "epoch": 0.6283504701643378,
      "grad_norm": 0.40453389286994934,
      "learning_rate": 2e-05,
      "loss": 0.5466,
      "step": 7150
    },
    {
      "epoch": 0.6292292820107215,
      "grad_norm": 0.3256462812423706,
      "learning_rate": 2e-05,
      "loss": 0.5496,
      "step": 7160
    },
    {
      "epoch": 0.6301080938571052,
      "grad_norm": 0.42262375354766846,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 7170
    },
    {
      "epoch": 0.6309869057034889,
      "grad_norm": 0.31683897972106934,
      "learning_rate": 2e-05,
      "loss": 0.559,
      "step": 7180
    },
    {
      "epoch": 0.6318657175498725,
      "grad_norm": 0.3651883006095886,
      "learning_rate": 2e-05,
      "loss": 0.5386,
      "step": 7190
    },
    {
      "epoch": 0.6327445293962563,
      "grad_norm": 0.36305928230285645,
      "learning_rate": 2e-05,
      "loss": 0.579,
      "step": 7200
    },
    {
      "epoch": 0.63362334124264,
      "grad_norm": 0.3868792951107025,
      "learning_rate": 2e-05,
      "loss": 0.5176,
      "step": 7210
    },
    {
      "epoch": 0.6345021530890237,
      "grad_norm": 0.3617139458656311,
      "learning_rate": 2e-05,
      "loss": 0.5162,
      "step": 7220
    },
    {
      "epoch": 0.6353809649354073,
      "grad_norm": 0.4091183543205261,
      "learning_rate": 2e-05,
      "loss": 0.5739,
      "step": 7230
    },
    {
      "epoch": 0.636259776781791,
      "grad_norm": 0.3827371597290039,
      "learning_rate": 2e-05,
      "loss": 0.5133,
      "step": 7240
    },
    {
      "epoch": 0.6371385886281747,
      "grad_norm": 0.32122868299484253,
      "learning_rate": 2e-05,
      "loss": 0.5102,
      "step": 7250
    },
    {
      "epoch": 0.6380174004745583,
      "grad_norm": 0.3225368857383728,
      "learning_rate": 2e-05,
      "loss": 0.5581,
      "step": 7260
    },
    {
      "epoch": 0.6388962123209421,
      "grad_norm": 0.39036616683006287,
      "learning_rate": 2e-05,
      "loss": 0.5392,
      "step": 7270
    },
    {
      "epoch": 0.6397750241673258,
      "grad_norm": 0.3332098722457886,
      "learning_rate": 2e-05,
      "loss": 0.5458,
      "step": 7280
    },
    {
      "epoch": 0.6406538360137095,
      "grad_norm": 0.3662400245666504,
      "learning_rate": 2e-05,
      "loss": 0.5478,
      "step": 7290
    },
    {
      "epoch": 0.6415326478600931,
      "grad_norm": 0.34433111548423767,
      "learning_rate": 2e-05,
      "loss": 0.5376,
      "step": 7300
    },
    {
      "epoch": 0.6424114597064768,
      "grad_norm": 0.4566636085510254,
      "learning_rate": 2e-05,
      "loss": 0.5431,
      "step": 7310
    },
    {
      "epoch": 0.6432902715528606,
      "grad_norm": 0.40051695704460144,
      "learning_rate": 2e-05,
      "loss": 0.5375,
      "step": 7320
    },
    {
      "epoch": 0.6441690833992442,
      "grad_norm": 0.3659587800502777,
      "learning_rate": 2e-05,
      "loss": 0.5195,
      "step": 7330
    },
    {
      "epoch": 0.6450478952456279,
      "grad_norm": 0.36354124546051025,
      "learning_rate": 2e-05,
      "loss": 0.5353,
      "step": 7340
    },
    {
      "epoch": 0.6459267070920116,
      "grad_norm": 0.3416467308998108,
      "learning_rate": 2e-05,
      "loss": 0.5629,
      "step": 7350
    },
    {
      "epoch": 0.6468055189383953,
      "grad_norm": 0.362836629152298,
      "learning_rate": 2e-05,
      "loss": 0.5435,
      "step": 7360
    },
    {
      "epoch": 0.6476843307847789,
      "grad_norm": 0.3795658051967621,
      "learning_rate": 2e-05,
      "loss": 0.5336,
      "step": 7370
    },
    {
      "epoch": 0.6485631426311627,
      "grad_norm": 0.33436691761016846,
      "learning_rate": 2e-05,
      "loss": 0.5171,
      "step": 7380
    },
    {
      "epoch": 0.6494419544775464,
      "grad_norm": 0.3471194803714752,
      "learning_rate": 2e-05,
      "loss": 0.5296,
      "step": 7390
    },
    {
      "epoch": 0.6503207663239301,
      "grad_norm": 0.3387486934661865,
      "learning_rate": 2e-05,
      "loss": 0.5325,
      "step": 7400
    },
    {
      "epoch": 0.6511995781703137,
      "grad_norm": 0.38308653235435486,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 7410
    },
    {
      "epoch": 0.6520783900166974,
      "grad_norm": 0.39202553033828735,
      "learning_rate": 2e-05,
      "loss": 0.5465,
      "step": 7420
    },
    {
      "epoch": 0.6529572018630811,
      "grad_norm": 0.38378140330314636,
      "learning_rate": 2e-05,
      "loss": 0.5576,
      "step": 7430
    },
    {
      "epoch": 0.6538360137094648,
      "grad_norm": 0.36536139249801636,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 7440
    },
    {
      "epoch": 0.6547148255558485,
      "grad_norm": 0.369819313287735,
      "learning_rate": 2e-05,
      "loss": 0.5144,
      "step": 7450
    },
    {
      "epoch": 0.6555936374022322,
      "grad_norm": 0.4547944962978363,
      "learning_rate": 2e-05,
      "loss": 0.5338,
      "step": 7460
    },
    {
      "epoch": 0.6564724492486159,
      "grad_norm": 0.34010234475135803,
      "learning_rate": 2e-05,
      "loss": 0.5599,
      "step": 7470
    },
    {
      "epoch": 0.6573512610949995,
      "grad_norm": 0.3957253098487854,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 7480
    },
    {
      "epoch": 0.6582300729413832,
      "grad_norm": 0.35059720277786255,
      "learning_rate": 2e-05,
      "loss": 0.5436,
      "step": 7490
    },
    {
      "epoch": 0.659108884787767,
      "grad_norm": 0.4258873164653778,
      "learning_rate": 2e-05,
      "loss": 0.5129,
      "step": 7500
    },
    {
      "epoch": 0.6599876966341506,
      "grad_norm": 0.3541068732738495,
      "learning_rate": 2e-05,
      "loss": 0.5328,
      "step": 7510
    },
    {
      "epoch": 0.6608665084805343,
      "grad_norm": 0.34808555245399475,
      "learning_rate": 2e-05,
      "loss": 0.5362,
      "step": 7520
    },
    {
      "epoch": 0.661745320326918,
      "grad_norm": 0.3891119956970215,
      "learning_rate": 2e-05,
      "loss": 0.5504,
      "step": 7530
    },
    {
      "epoch": 0.6626241321733017,
      "grad_norm": 0.46070876717567444,
      "learning_rate": 2e-05,
      "loss": 0.5567,
      "step": 7540
    },
    {
      "epoch": 0.6635029440196853,
      "grad_norm": 0.35119718313217163,
      "learning_rate": 2e-05,
      "loss": 0.5238,
      "step": 7550
    },
    {
      "epoch": 0.6643817558660691,
      "grad_norm": 0.35495510697364807,
      "learning_rate": 2e-05,
      "loss": 0.5495,
      "step": 7560
    },
    {
      "epoch": 0.6652605677124528,
      "grad_norm": 0.4240226447582245,
      "learning_rate": 2e-05,
      "loss": 0.5316,
      "step": 7570
    },
    {
      "epoch": 0.6661393795588365,
      "grad_norm": 0.3270878195762634,
      "learning_rate": 2e-05,
      "loss": 0.507,
      "step": 7580
    },
    {
      "epoch": 0.6670181914052201,
      "grad_norm": 0.3316537141799927,
      "learning_rate": 2e-05,
      "loss": 0.5563,
      "step": 7590
    },
    {
      "epoch": 0.6678970032516038,
      "grad_norm": 0.3727238178253174,
      "learning_rate": 2e-05,
      "loss": 0.5053,
      "step": 7600
    },
    {
      "epoch": 0.6687758150979876,
      "grad_norm": 0.3476908802986145,
      "learning_rate": 2e-05,
      "loss": 0.5371,
      "step": 7610
    },
    {
      "epoch": 0.6696546269443712,
      "grad_norm": 0.45026981830596924,
      "learning_rate": 2e-05,
      "loss": 0.5218,
      "step": 7620
    },
    {
      "epoch": 0.6705334387907549,
      "grad_norm": 0.3603118062019348,
      "learning_rate": 2e-05,
      "loss": 0.5477,
      "step": 7630
    },
    {
      "epoch": 0.6714122506371386,
      "grad_norm": 0.4562949240207672,
      "learning_rate": 2e-05,
      "loss": 0.5308,
      "step": 7640
    },
    {
      "epoch": 0.6722910624835223,
      "grad_norm": 0.33607155084609985,
      "learning_rate": 2e-05,
      "loss": 0.509,
      "step": 7650
    },
    {
      "epoch": 0.6731698743299059,
      "grad_norm": 0.3662949800491333,
      "learning_rate": 2e-05,
      "loss": 0.5528,
      "step": 7660
    },
    {
      "epoch": 0.6740486861762897,
      "grad_norm": 0.3606943190097809,
      "learning_rate": 2e-05,
      "loss": 0.5503,
      "step": 7670
    },
    {
      "epoch": 0.6749274980226734,
      "grad_norm": 0.3293268382549286,
      "learning_rate": 2e-05,
      "loss": 0.5064,
      "step": 7680
    },
    {
      "epoch": 0.675806309869057,
      "grad_norm": 0.37557536363601685,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 7690
    },
    {
      "epoch": 0.6766851217154407,
      "grad_norm": 0.3698790967464447,
      "learning_rate": 2e-05,
      "loss": 0.5188,
      "step": 7700
    },
    {
      "epoch": 0.6775639335618244,
      "grad_norm": 0.4006452262401581,
      "learning_rate": 2e-05,
      "loss": 0.5065,
      "step": 7710
    },
    {
      "epoch": 0.6784427454082081,
      "grad_norm": 0.38918977975845337,
      "learning_rate": 2e-05,
      "loss": 0.5347,
      "step": 7720
    },
    {
      "epoch": 0.6793215572545918,
      "grad_norm": 0.36206454038619995,
      "learning_rate": 2e-05,
      "loss": 0.5399,
      "step": 7730
    },
    {
      "epoch": 0.6802003691009755,
      "grad_norm": 0.38814258575439453,
      "learning_rate": 2e-05,
      "loss": 0.5468,
      "step": 7740
    },
    {
      "epoch": 0.6810791809473592,
      "grad_norm": 0.4193398058414459,
      "learning_rate": 2e-05,
      "loss": 0.5123,
      "step": 7750
    },
    {
      "epoch": 0.6819579927937428,
      "grad_norm": 0.3855031430721283,
      "learning_rate": 2e-05,
      "loss": 0.5448,
      "step": 7760
    },
    {
      "epoch": 0.6828368046401265,
      "grad_norm": 0.3836282789707184,
      "learning_rate": 2e-05,
      "loss": 0.5296,
      "step": 7770
    },
    {
      "epoch": 0.6837156164865102,
      "grad_norm": 0.36712026596069336,
      "learning_rate": 2e-05,
      "loss": 0.5368,
      "step": 7780
    },
    {
      "epoch": 0.684594428332894,
      "grad_norm": 0.38301706314086914,
      "learning_rate": 2e-05,
      "loss": 0.5322,
      "step": 7790
    },
    {
      "epoch": 0.6854732401792776,
      "grad_norm": 0.3586750030517578,
      "learning_rate": 2e-05,
      "loss": 0.5146,
      "step": 7800
    },
    {
      "epoch": 0.6863520520256613,
      "grad_norm": 0.3308457136154175,
      "learning_rate": 2e-05,
      "loss": 0.5316,
      "step": 7810
    },
    {
      "epoch": 0.687230863872045,
      "grad_norm": 0.4728618562221527,
      "learning_rate": 2e-05,
      "loss": 0.5058,
      "step": 7820
    },
    {
      "epoch": 0.6881096757184287,
      "grad_norm": 0.3827001750469208,
      "learning_rate": 2e-05,
      "loss": 0.5488,
      "step": 7830
    },
    {
      "epoch": 0.6889884875648123,
      "grad_norm": 0.36872726678848267,
      "learning_rate": 2e-05,
      "loss": 0.5095,
      "step": 7840
    },
    {
      "epoch": 0.6898672994111961,
      "grad_norm": 0.36616966128349304,
      "learning_rate": 2e-05,
      "loss": 0.5301,
      "step": 7850
    },
    {
      "epoch": 0.6907461112575798,
      "grad_norm": 0.4945073425769806,
      "learning_rate": 2e-05,
      "loss": 0.5123,
      "step": 7860
    },
    {
      "epoch": 0.6916249231039634,
      "grad_norm": 0.3636736273765564,
      "learning_rate": 2e-05,
      "loss": 0.5206,
      "step": 7870
    },
    {
      "epoch": 0.6925037349503471,
      "grad_norm": 0.41000962257385254,
      "learning_rate": 2e-05,
      "loss": 0.5174,
      "step": 7880
    },
    {
      "epoch": 0.6933825467967308,
      "grad_norm": 0.36956432461738586,
      "learning_rate": 2e-05,
      "loss": 0.522,
      "step": 7890
    },
    {
      "epoch": 0.6942613586431146,
      "grad_norm": 0.3563474118709564,
      "learning_rate": 2e-05,
      "loss": 0.5296,
      "step": 7900
    },
    {
      "epoch": 0.6951401704894982,
      "grad_norm": 0.4562670886516571,
      "learning_rate": 2e-05,
      "loss": 0.5197,
      "step": 7910
    },
    {
      "epoch": 0.6960189823358819,
      "grad_norm": 0.4118843078613281,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 7920
    },
    {
      "epoch": 0.6968977941822656,
      "grad_norm": 0.34662866592407227,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 7930
    },
    {
      "epoch": 0.6977766060286492,
      "grad_norm": 0.3601706027984619,
      "learning_rate": 2e-05,
      "loss": 0.5222,
      "step": 7940
    },
    {
      "epoch": 0.6986554178750329,
      "grad_norm": 0.33368995785713196,
      "learning_rate": 2e-05,
      "loss": 0.52,
      "step": 7950
    },
    {
      "epoch": 0.6995342297214167,
      "grad_norm": 0.33430880308151245,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 7960
    },
    {
      "epoch": 0.7004130415678004,
      "grad_norm": 0.3651207685470581,
      "learning_rate": 2e-05,
      "loss": 0.5099,
      "step": 7970
    },
    {
      "epoch": 0.701291853414184,
      "grad_norm": 0.3436855971813202,
      "learning_rate": 2e-05,
      "loss": 0.5302,
      "step": 7980
    },
    {
      "epoch": 0.7021706652605677,
      "grad_norm": 0.3660446107387543,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 7990
    },
    {
      "epoch": 0.7030494771069514,
      "grad_norm": 0.35243427753448486,
      "learning_rate": 2e-05,
      "loss": 0.5194,
      "step": 8000
    },
    {
      "epoch": 0.7039282889533351,
      "grad_norm": 0.33274373412132263,
      "learning_rate": 2e-05,
      "loss": 0.5259,
      "step": 8010
    },
    {
      "epoch": 0.7048071007997188,
      "grad_norm": 0.39158737659454346,
      "learning_rate": 2e-05,
      "loss": 0.5296,
      "step": 8020
    },
    {
      "epoch": 0.7056859126461025,
      "grad_norm": 0.3075087368488312,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 8030
    },
    {
      "epoch": 0.7065647244924862,
      "grad_norm": 0.3483194410800934,
      "learning_rate": 2e-05,
      "loss": 0.5396,
      "step": 8040
    },
    {
      "epoch": 0.7074435363388698,
      "grad_norm": 0.3626619577407837,
      "learning_rate": 2e-05,
      "loss": 0.5277,
      "step": 8050
    },
    {
      "epoch": 0.7083223481852535,
      "grad_norm": 0.40117600560188293,
      "learning_rate": 2e-05,
      "loss": 0.5255,
      "step": 8060
    },
    {
      "epoch": 0.7092011600316372,
      "grad_norm": 0.47729572653770447,
      "learning_rate": 2e-05,
      "loss": 0.5572,
      "step": 8070
    },
    {
      "epoch": 0.710079971878021,
      "grad_norm": 0.3313013017177582,
      "learning_rate": 2e-05,
      "loss": 0.5128,
      "step": 8080
    },
    {
      "epoch": 0.7109587837244046,
      "grad_norm": 0.37826141715049744,
      "learning_rate": 2e-05,
      "loss": 0.525,
      "step": 8090
    },
    {
      "epoch": 0.7118375955707883,
      "grad_norm": 0.4007148742675781,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 8100
    },
    {
      "epoch": 0.712716407417172,
      "grad_norm": 0.3478046953678131,
      "learning_rate": 2e-05,
      "loss": 0.5493,
      "step": 8110
    },
    {
      "epoch": 0.7135952192635556,
      "grad_norm": 0.37845683097839355,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 8120
    },
    {
      "epoch": 0.7144740311099393,
      "grad_norm": 0.34525442123413086,
      "learning_rate": 2e-05,
      "loss": 0.5337,
      "step": 8130
    },
    {
      "epoch": 0.7153528429563231,
      "grad_norm": 0.45192646980285645,
      "learning_rate": 2e-05,
      "loss": 0.5337,
      "step": 8140
    },
    {
      "epoch": 0.7162316548027068,
      "grad_norm": 0.3639838695526123,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 8150
    },
    {
      "epoch": 0.7171104666490904,
      "grad_norm": 0.3488052785396576,
      "learning_rate": 2e-05,
      "loss": 0.5156,
      "step": 8160
    },
    {
      "epoch": 0.7179892784954741,
      "grad_norm": 0.3622989356517792,
      "learning_rate": 2e-05,
      "loss": 0.5155,
      "step": 8170
    },
    {
      "epoch": 0.7188680903418578,
      "grad_norm": 0.3201085031032562,
      "learning_rate": 2e-05,
      "loss": 0.5109,
      "step": 8180
    },
    {
      "epoch": 0.7197469021882414,
      "grad_norm": 0.3609197735786438,
      "learning_rate": 2e-05,
      "loss": 0.5048,
      "step": 8190
    },
    {
      "epoch": 0.7206257140346252,
      "grad_norm": 0.3626137375831604,
      "learning_rate": 2e-05,
      "loss": 0.5104,
      "step": 8200
    },
    {
      "epoch": 0.7215045258810089,
      "grad_norm": 0.33866509795188904,
      "learning_rate": 2e-05,
      "loss": 0.4989,
      "step": 8210
    },
    {
      "epoch": 0.7223833377273926,
      "grad_norm": 0.37702131271362305,
      "learning_rate": 2e-05,
      "loss": 0.5418,
      "step": 8220
    },
    {
      "epoch": 0.7232621495737762,
      "grad_norm": 0.4207315444946289,
      "learning_rate": 2e-05,
      "loss": 0.5292,
      "step": 8230
    },
    {
      "epoch": 0.7241409614201599,
      "grad_norm": 0.3663068413734436,
      "learning_rate": 2e-05,
      "loss": 0.5286,
      "step": 8240
    },
    {
      "epoch": 0.7250197732665437,
      "grad_norm": 0.3579179346561432,
      "learning_rate": 2e-05,
      "loss": 0.5373,
      "step": 8250
    },
    {
      "epoch": 0.7258985851129274,
      "grad_norm": 0.35196587443351746,
      "learning_rate": 2e-05,
      "loss": 0.5341,
      "step": 8260
    },
    {
      "epoch": 0.726777396959311,
      "grad_norm": 0.3516296148300171,
      "learning_rate": 2e-05,
      "loss": 0.5051,
      "step": 8270
    },
    {
      "epoch": 0.7276562088056947,
      "grad_norm": 0.32521840929985046,
      "learning_rate": 2e-05,
      "loss": 0.5116,
      "step": 8280
    },
    {
      "epoch": 0.7285350206520784,
      "grad_norm": 0.3544333875179291,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 8290
    },
    {
      "epoch": 0.729413832498462,
      "grad_norm": 0.49609145522117615,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 8300
    },
    {
      "epoch": 0.7302926443448458,
      "grad_norm": 0.43958431482315063,
      "learning_rate": 2e-05,
      "loss": 0.5283,
      "step": 8310
    },
    {
      "epoch": 0.7311714561912295,
      "grad_norm": 0.37573879957199097,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 8320
    },
    {
      "epoch": 0.7320502680376132,
      "grad_norm": 0.33459755778312683,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 8330
    },
    {
      "epoch": 0.7329290798839968,
      "grad_norm": 0.3984433114528656,
      "learning_rate": 2e-05,
      "loss": 0.5763,
      "step": 8340
    },
    {
      "epoch": 0.7338078917303805,
      "grad_norm": 0.3672713041305542,
      "learning_rate": 2e-05,
      "loss": 0.5373,
      "step": 8350
    },
    {
      "epoch": 0.7346867035767642,
      "grad_norm": 0.3761705160140991,
      "learning_rate": 2e-05,
      "loss": 0.5202,
      "step": 8360
    },
    {
      "epoch": 0.7355655154231479,
      "grad_norm": 0.3487396538257599,
      "learning_rate": 2e-05,
      "loss": 0.5007,
      "step": 8370
    },
    {
      "epoch": 0.7364443272695316,
      "grad_norm": 0.33625277876853943,
      "learning_rate": 2e-05,
      "loss": 0.5087,
      "step": 8380
    },
    {
      "epoch": 0.7373231391159153,
      "grad_norm": 0.37763917446136475,
      "learning_rate": 2e-05,
      "loss": 0.52,
      "step": 8390
    },
    {
      "epoch": 0.738201950962299,
      "grad_norm": 0.3642916977405548,
      "learning_rate": 2e-05,
      "loss": 0.5243,
      "step": 8400
    },
    {
      "epoch": 0.7390807628086826,
      "grad_norm": 0.3551781177520752,
      "learning_rate": 2e-05,
      "loss": 0.5541,
      "step": 8410
    },
    {
      "epoch": 0.7399595746550663,
      "grad_norm": 0.35462257266044617,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 8420
    },
    {
      "epoch": 0.7408383865014501,
      "grad_norm": 0.6856582760810852,
      "learning_rate": 2e-05,
      "loss": 0.4976,
      "step": 8430
    },
    {
      "epoch": 0.7417171983478338,
      "grad_norm": 0.39249134063720703,
      "learning_rate": 2e-05,
      "loss": 0.5278,
      "step": 8440
    },
    {
      "epoch": 0.7425960101942174,
      "grad_norm": 0.3554900586605072,
      "learning_rate": 2e-05,
      "loss": 0.5097,
      "step": 8450
    },
    {
      "epoch": 0.7434748220406011,
      "grad_norm": 0.36453068256378174,
      "learning_rate": 2e-05,
      "loss": 0.5196,
      "step": 8460
    },
    {
      "epoch": 0.7443536338869848,
      "grad_norm": 0.3349120020866394,
      "learning_rate": 2e-05,
      "loss": 0.5154,
      "step": 8470
    },
    {
      "epoch": 0.7452324457333684,
      "grad_norm": 0.3538738787174225,
      "learning_rate": 2e-05,
      "loss": 0.5079,
      "step": 8480
    },
    {
      "epoch": 0.7461112575797522,
      "grad_norm": 0.3567652404308319,
      "learning_rate": 2e-05,
      "loss": 0.5254,
      "step": 8490
    },
    {
      "epoch": 0.7469900694261359,
      "grad_norm": 0.34555625915527344,
      "learning_rate": 2e-05,
      "loss": 0.5457,
      "step": 8500
    },
    {
      "epoch": 0.7478688812725196,
      "grad_norm": 0.3603267967700958,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 8510
    },
    {
      "epoch": 0.7487476931189032,
      "grad_norm": 0.47140905261039734,
      "learning_rate": 2e-05,
      "loss": 0.5082,
      "step": 8520
    },
    {
      "epoch": 0.7496265049652869,
      "grad_norm": 0.3156260848045349,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 8530
    },
    {
      "epoch": 0.7505053168116707,
      "grad_norm": 0.35314199328422546,
      "learning_rate": 2e-05,
      "loss": 0.5131,
      "step": 8540
    },
    {
      "epoch": 0.7513841286580543,
      "grad_norm": 0.36993417143821716,
      "learning_rate": 2e-05,
      "loss": 0.5203,
      "step": 8550
    },
    {
      "epoch": 0.752262940504438,
      "grad_norm": 0.41483813524246216,
      "learning_rate": 2e-05,
      "loss": 0.5442,
      "step": 8560
    },
    {
      "epoch": 0.7531417523508217,
      "grad_norm": 0.403870165348053,
      "learning_rate": 2e-05,
      "loss": 0.5205,
      "step": 8570
    },
    {
      "epoch": 0.7540205641972054,
      "grad_norm": 0.35109028220176697,
      "learning_rate": 2e-05,
      "loss": 0.5049,
      "step": 8580
    },
    {
      "epoch": 0.754899376043589,
      "grad_norm": 0.518711268901825,
      "learning_rate": 2e-05,
      "loss": 0.5419,
      "step": 8590
    },
    {
      "epoch": 0.7557781878899728,
      "grad_norm": 0.3551972210407257,
      "learning_rate": 2e-05,
      "loss": 0.5301,
      "step": 8600
    },
    {
      "epoch": 0.7566569997363565,
      "grad_norm": 0.37170857191085815,
      "learning_rate": 2e-05,
      "loss": 0.5432,
      "step": 8610
    },
    {
      "epoch": 0.7575358115827401,
      "grad_norm": 0.41798412799835205,
      "learning_rate": 2e-05,
      "loss": 0.5133,
      "step": 8620
    },
    {
      "epoch": 0.7584146234291238,
      "grad_norm": 0.38754600286483765,
      "learning_rate": 2e-05,
      "loss": 0.5499,
      "step": 8630
    },
    {
      "epoch": 0.7592934352755075,
      "grad_norm": 0.3602105975151062,
      "learning_rate": 2e-05,
      "loss": 0.541,
      "step": 8640
    },
    {
      "epoch": 0.7601722471218912,
      "grad_norm": 0.35684725642204285,
      "learning_rate": 2e-05,
      "loss": 0.544,
      "step": 8650
    },
    {
      "epoch": 0.7610510589682749,
      "grad_norm": 0.34904223680496216,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 8660
    },
    {
      "epoch": 0.7619298708146586,
      "grad_norm": 0.37074458599090576,
      "learning_rate": 2e-05,
      "loss": 0.5481,
      "step": 8670
    },
    {
      "epoch": 0.7628086826610423,
      "grad_norm": 0.3503706753253937,
      "learning_rate": 2e-05,
      "loss": 0.5081,
      "step": 8680
    },
    {
      "epoch": 0.763687494507426,
      "grad_norm": 0.35190242528915405,
      "learning_rate": 2e-05,
      "loss": 0.5091,
      "step": 8690
    },
    {
      "epoch": 0.7645663063538096,
      "grad_norm": 0.37592995166778564,
      "learning_rate": 2e-05,
      "loss": 0.5507,
      "step": 8700
    },
    {
      "epoch": 0.7654451182001933,
      "grad_norm": 0.3844635486602783,
      "learning_rate": 2e-05,
      "loss": 0.5194,
      "step": 8710
    },
    {
      "epoch": 0.7663239300465771,
      "grad_norm": 0.3414153754711151,
      "learning_rate": 2e-05,
      "loss": 0.5198,
      "step": 8720
    },
    {
      "epoch": 0.7672027418929607,
      "grad_norm": 0.350923091173172,
      "learning_rate": 2e-05,
      "loss": 0.5373,
      "step": 8730
    },
    {
      "epoch": 0.7680815537393444,
      "grad_norm": 0.35310226678848267,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 8740
    },
    {
      "epoch": 0.7689603655857281,
      "grad_norm": 0.3671940565109253,
      "learning_rate": 2e-05,
      "loss": 0.5211,
      "step": 8750
    },
    {
      "epoch": 0.7698391774321118,
      "grad_norm": 0.3403673768043518,
      "learning_rate": 2e-05,
      "loss": 0.5078,
      "step": 8760
    },
    {
      "epoch": 0.7707179892784954,
      "grad_norm": 0.3576790392398834,
      "learning_rate": 2e-05,
      "loss": 0.5146,
      "step": 8770
    },
    {
      "epoch": 0.7715968011248792,
      "grad_norm": 0.3642694652080536,
      "learning_rate": 2e-05,
      "loss": 0.525,
      "step": 8780
    },
    {
      "epoch": 0.7724756129712629,
      "grad_norm": 0.38960033655166626,
      "learning_rate": 2e-05,
      "loss": 0.5247,
      "step": 8790
    },
    {
      "epoch": 0.7733544248176465,
      "grad_norm": 0.32385528087615967,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 8800
    },
    {
      "epoch": 0.7742332366640302,
      "grad_norm": 0.41729339957237244,
      "learning_rate": 2e-05,
      "loss": 0.5243,
      "step": 8810
    },
    {
      "epoch": 0.7751120485104139,
      "grad_norm": 0.41685882210731506,
      "learning_rate": 2e-05,
      "loss": 0.5278,
      "step": 8820
    },
    {
      "epoch": 0.7759908603567977,
      "grad_norm": 0.3382299542427063,
      "learning_rate": 2e-05,
      "loss": 0.5037,
      "step": 8830
    },
    {
      "epoch": 0.7768696722031813,
      "grad_norm": 0.37513384222984314,
      "learning_rate": 2e-05,
      "loss": 0.5206,
      "step": 8840
    },
    {
      "epoch": 0.777748484049565,
      "grad_norm": 0.39156413078308105,
      "learning_rate": 2e-05,
      "loss": 0.5018,
      "step": 8850
    },
    {
      "epoch": 0.7786272958959487,
      "grad_norm": 0.3638029098510742,
      "learning_rate": 2e-05,
      "loss": 0.5252,
      "step": 8860
    },
    {
      "epoch": 0.7795061077423324,
      "grad_norm": 0.390289306640625,
      "learning_rate": 2e-05,
      "loss": 0.5186,
      "step": 8870
    },
    {
      "epoch": 0.780384919588716,
      "grad_norm": 0.3644057810306549,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 8880
    },
    {
      "epoch": 0.7812637314350998,
      "grad_norm": 0.3338548541069031,
      "learning_rate": 2e-05,
      "loss": 0.5166,
      "step": 8890
    },
    {
      "epoch": 0.7821425432814835,
      "grad_norm": 0.36306431889533997,
      "learning_rate": 2e-05,
      "loss": 0.5417,
      "step": 8900
    },
    {
      "epoch": 0.7830213551278671,
      "grad_norm": 0.3919699192047119,
      "learning_rate": 2e-05,
      "loss": 0.5278,
      "step": 8910
    },
    {
      "epoch": 0.7839001669742508,
      "grad_norm": 0.4318430423736572,
      "learning_rate": 2e-05,
      "loss": 0.5309,
      "step": 8920
    },
    {
      "epoch": 0.7847789788206345,
      "grad_norm": 0.4534205198287964,
      "learning_rate": 2e-05,
      "loss": 0.5214,
      "step": 8930
    },
    {
      "epoch": 0.7856577906670182,
      "grad_norm": 0.3462096154689789,
      "learning_rate": 2e-05,
      "loss": 0.5209,
      "step": 8940
    },
    {
      "epoch": 0.7865366025134019,
      "grad_norm": 0.35275012254714966,
      "learning_rate": 2e-05,
      "loss": 0.5327,
      "step": 8950
    },
    {
      "epoch": 0.7874154143597856,
      "grad_norm": 0.3450658619403839,
      "learning_rate": 2e-05,
      "loss": 0.5147,
      "step": 8960
    },
    {
      "epoch": 0.7882942262061693,
      "grad_norm": 0.43237850069999695,
      "learning_rate": 2e-05,
      "loss": 0.5168,
      "step": 8970
    },
    {
      "epoch": 0.7891730380525529,
      "grad_norm": 0.34790584444999695,
      "learning_rate": 2e-05,
      "loss": 0.5153,
      "step": 8980
    },
    {
      "epoch": 0.7900518498989366,
      "grad_norm": 0.33977270126342773,
      "learning_rate": 2e-05,
      "loss": 0.5265,
      "step": 8990
    },
    {
      "epoch": 0.7909306617453203,
      "grad_norm": 0.3491854667663574,
      "learning_rate": 2e-05,
      "loss": 0.5377,
      "step": 9000
    },
    {
      "epoch": 0.7918094735917041,
      "grad_norm": 0.3819669485092163,
      "learning_rate": 2e-05,
      "loss": 0.5336,
      "step": 9010
    },
    {
      "epoch": 0.7926882854380877,
      "grad_norm": 0.38407352566719055,
      "learning_rate": 2e-05,
      "loss": 0.5163,
      "step": 9020
    },
    {
      "epoch": 0.7935670972844714,
      "grad_norm": 0.4173966646194458,
      "learning_rate": 2e-05,
      "loss": 0.5476,
      "step": 9030
    },
    {
      "epoch": 0.7944459091308551,
      "grad_norm": 0.33108294010162354,
      "learning_rate": 2e-05,
      "loss": 0.5118,
      "step": 9040
    },
    {
      "epoch": 0.7953247209772387,
      "grad_norm": 0.420549601316452,
      "learning_rate": 2e-05,
      "loss": 0.5137,
      "step": 9050
    },
    {
      "epoch": 0.7962035328236224,
      "grad_norm": 0.37894389033317566,
      "learning_rate": 2e-05,
      "loss": 0.5197,
      "step": 9060
    },
    {
      "epoch": 0.7970823446700062,
      "grad_norm": 0.3867136240005493,
      "learning_rate": 2e-05,
      "loss": 0.5148,
      "step": 9070
    },
    {
      "epoch": 0.7979611565163899,
      "grad_norm": 0.41404059529304504,
      "learning_rate": 2e-05,
      "loss": 0.5367,
      "step": 9080
    },
    {
      "epoch": 0.7988399683627735,
      "grad_norm": 0.3393182158470154,
      "learning_rate": 2e-05,
      "loss": 0.5013,
      "step": 9090
    },
    {
      "epoch": 0.7997187802091572,
      "grad_norm": 0.4352535009384155,
      "learning_rate": 2e-05,
      "loss": 0.523,
      "step": 9100
    },
    {
      "epoch": 0.8005975920555409,
      "grad_norm": 0.38436514139175415,
      "learning_rate": 2e-05,
      "loss": 0.5316,
      "step": 9110
    },
    {
      "epoch": 0.8014764039019246,
      "grad_norm": 0.37239721417427063,
      "learning_rate": 2e-05,
      "loss": 0.5384,
      "step": 9120
    },
    {
      "epoch": 0.8023552157483083,
      "grad_norm": 0.33295226097106934,
      "learning_rate": 2e-05,
      "loss": 0.5291,
      "step": 9130
    },
    {
      "epoch": 0.803234027594692,
      "grad_norm": 0.7052021026611328,
      "learning_rate": 2e-05,
      "loss": 0.5212,
      "step": 9140
    },
    {
      "epoch": 0.8041128394410757,
      "grad_norm": 0.3411831855773926,
      "learning_rate": 2e-05,
      "loss": 0.509,
      "step": 9150
    },
    {
      "epoch": 0.8049916512874593,
      "grad_norm": 0.40182483196258545,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 9160
    },
    {
      "epoch": 0.805870463133843,
      "grad_norm": 0.32982924580574036,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 9170
    },
    {
      "epoch": 0.8067492749802267,
      "grad_norm": 0.39872363209724426,
      "learning_rate": 2e-05,
      "loss": 0.5366,
      "step": 9180
    },
    {
      "epoch": 0.8076280868266105,
      "grad_norm": 0.37407493591308594,
      "learning_rate": 2e-05,
      "loss": 0.5343,
      "step": 9190
    },
    {
      "epoch": 0.8085068986729941,
      "grad_norm": 0.38882148265838623,
      "learning_rate": 2e-05,
      "loss": 0.5133,
      "step": 9200
    },
    {
      "epoch": 0.8093857105193778,
      "grad_norm": 0.3726615607738495,
      "learning_rate": 2e-05,
      "loss": 0.5095,
      "step": 9210
    },
    {
      "epoch": 0.8102645223657615,
      "grad_norm": 0.3771108090877533,
      "learning_rate": 2e-05,
      "loss": 0.5275,
      "step": 9220
    },
    {
      "epoch": 0.8111433342121451,
      "grad_norm": 0.4238576889038086,
      "learning_rate": 2e-05,
      "loss": 0.5411,
      "step": 9230
    },
    {
      "epoch": 0.8120221460585288,
      "grad_norm": 0.424276202917099,
      "learning_rate": 2e-05,
      "loss": 0.5069,
      "step": 9240
    },
    {
      "epoch": 0.8129009579049126,
      "grad_norm": 0.3614506423473358,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 9250
    },
    {
      "epoch": 0.8137797697512963,
      "grad_norm": 0.36140766739845276,
      "learning_rate": 2e-05,
      "loss": 0.5139,
      "step": 9260
    },
    {
      "epoch": 0.8146585815976799,
      "grad_norm": 0.40187183022499084,
      "learning_rate": 2e-05,
      "loss": 0.5593,
      "step": 9270
    },
    {
      "epoch": 0.8155373934440636,
      "grad_norm": 0.4235718548297882,
      "learning_rate": 2e-05,
      "loss": 0.5314,
      "step": 9280
    },
    {
      "epoch": 0.8164162052904473,
      "grad_norm": 0.3735692799091339,
      "learning_rate": 2e-05,
      "loss": 0.5208,
      "step": 9290
    },
    {
      "epoch": 0.8172950171368311,
      "grad_norm": 0.39952823519706726,
      "learning_rate": 2e-05,
      "loss": 0.5104,
      "step": 9300
    },
    {
      "epoch": 0.8181738289832147,
      "grad_norm": 0.35223764181137085,
      "learning_rate": 2e-05,
      "loss": 0.54,
      "step": 9310
    },
    {
      "epoch": 0.8190526408295984,
      "grad_norm": 0.370201051235199,
      "learning_rate": 2e-05,
      "loss": 0.5336,
      "step": 9320
    },
    {
      "epoch": 0.8199314526759821,
      "grad_norm": 0.34956488013267517,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 9330
    },
    {
      "epoch": 0.8208102645223657,
      "grad_norm": 0.38316792249679565,
      "learning_rate": 2e-05,
      "loss": 0.5203,
      "step": 9340
    },
    {
      "epoch": 0.8216890763687494,
      "grad_norm": 0.37181565165519714,
      "learning_rate": 2e-05,
      "loss": 0.5632,
      "step": 9350
    },
    {
      "epoch": 0.8225678882151332,
      "grad_norm": 0.35303905606269836,
      "learning_rate": 2e-05,
      "loss": 0.5421,
      "step": 9360
    },
    {
      "epoch": 0.8234467000615169,
      "grad_norm": 0.3235478401184082,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 9370
    },
    {
      "epoch": 0.8243255119079005,
      "grad_norm": 0.3267079293727875,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 9380
    },
    {
      "epoch": 0.8252043237542842,
      "grad_norm": 0.36280786991119385,
      "learning_rate": 2e-05,
      "loss": 0.5345,
      "step": 9390
    },
    {
      "epoch": 0.8260831356006679,
      "grad_norm": 0.37867870926856995,
      "learning_rate": 2e-05,
      "loss": 0.5314,
      "step": 9400
    },
    {
      "epoch": 0.8269619474470515,
      "grad_norm": 0.4049442410469055,
      "learning_rate": 2e-05,
      "loss": 0.5492,
      "step": 9410
    },
    {
      "epoch": 0.8278407592934353,
      "grad_norm": 0.3633390963077545,
      "learning_rate": 2e-05,
      "loss": 0.5687,
      "step": 9420
    },
    {
      "epoch": 0.828719571139819,
      "grad_norm": 0.37445178627967834,
      "learning_rate": 2e-05,
      "loss": 0.526,
      "step": 9430
    },
    {
      "epoch": 0.8295983829862027,
      "grad_norm": 0.34585848450660706,
      "learning_rate": 2e-05,
      "loss": 0.4881,
      "step": 9440
    },
    {
      "epoch": 0.8304771948325863,
      "grad_norm": 0.3532872200012207,
      "learning_rate": 2e-05,
      "loss": 0.4898,
      "step": 9450
    },
    {
      "epoch": 0.83135600667897,
      "grad_norm": 0.3550582826137543,
      "learning_rate": 2e-05,
      "loss": 0.5312,
      "step": 9460
    },
    {
      "epoch": 0.8322348185253537,
      "grad_norm": 0.34059351682662964,
      "learning_rate": 2e-05,
      "loss": 0.4878,
      "step": 9470
    },
    {
      "epoch": 0.8331136303717374,
      "grad_norm": 0.3726445436477661,
      "learning_rate": 2e-05,
      "loss": 0.4943,
      "step": 9480
    },
    {
      "epoch": 0.8339924422181211,
      "grad_norm": 0.3388146162033081,
      "learning_rate": 2e-05,
      "loss": 0.5121,
      "step": 9490
    },
    {
      "epoch": 0.8348712540645048,
      "grad_norm": 0.33791548013687134,
      "learning_rate": 2e-05,
      "loss": 0.5123,
      "step": 9500
    },
    {
      "epoch": 0.8357500659108885,
      "grad_norm": 0.3319653570652008,
      "learning_rate": 2e-05,
      "loss": 0.5079,
      "step": 9510
    },
    {
      "epoch": 0.8366288777572721,
      "grad_norm": 0.38049542903900146,
      "learning_rate": 2e-05,
      "loss": 0.5159,
      "step": 9520
    },
    {
      "epoch": 0.8375076896036558,
      "grad_norm": 0.37364909052848816,
      "learning_rate": 2e-05,
      "loss": 0.5107,
      "step": 9530
    },
    {
      "epoch": 0.8383865014500396,
      "grad_norm": 0.42299723625183105,
      "learning_rate": 2e-05,
      "loss": 0.508,
      "step": 9540
    },
    {
      "epoch": 0.8392653132964233,
      "grad_norm": 0.3644964098930359,
      "learning_rate": 2e-05,
      "loss": 0.484,
      "step": 9550
    },
    {
      "epoch": 0.8401441251428069,
      "grad_norm": 0.35114142298698425,
      "learning_rate": 2e-05,
      "loss": 0.5315,
      "step": 9560
    },
    {
      "epoch": 0.8410229369891906,
      "grad_norm": 0.40702614188194275,
      "learning_rate": 2e-05,
      "loss": 0.5074,
      "step": 9570
    },
    {
      "epoch": 0.8419017488355743,
      "grad_norm": 0.39313217997550964,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 9580
    },
    {
      "epoch": 0.842780560681958,
      "grad_norm": 0.37034666538238525,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 9590
    },
    {
      "epoch": 0.8436593725283417,
      "grad_norm": 0.3564906120300293,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 9600
    },
    {
      "epoch": 0.8445381843747254,
      "grad_norm": 0.35559988021850586,
      "learning_rate": 2e-05,
      "loss": 0.5258,
      "step": 9610
    },
    {
      "epoch": 0.8454169962211091,
      "grad_norm": 0.3448827266693115,
      "learning_rate": 2e-05,
      "loss": 0.5078,
      "step": 9620
    },
    {
      "epoch": 0.8462958080674927,
      "grad_norm": 0.3611103296279907,
      "learning_rate": 2e-05,
      "loss": 0.5034,
      "step": 9630
    },
    {
      "epoch": 0.8471746199138764,
      "grad_norm": 0.33421850204467773,
      "learning_rate": 2e-05,
      "loss": 0.5045,
      "step": 9640
    },
    {
      "epoch": 0.8480534317602602,
      "grad_norm": 0.4717365503311157,
      "learning_rate": 2e-05,
      "loss": 0.514,
      "step": 9650
    },
    {
      "epoch": 0.8489322436066438,
      "grad_norm": 0.34485507011413574,
      "learning_rate": 2e-05,
      "loss": 0.5088,
      "step": 9660
    },
    {
      "epoch": 0.8498110554530275,
      "grad_norm": 0.3937644064426422,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 9670
    },
    {
      "epoch": 0.8506898672994112,
      "grad_norm": 0.346683144569397,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 9680
    },
    {
      "epoch": 0.8515686791457949,
      "grad_norm": 0.36385565996170044,
      "learning_rate": 2e-05,
      "loss": 0.523,
      "step": 9690
    },
    {
      "epoch": 0.8524474909921785,
      "grad_norm": 0.3494904339313507,
      "learning_rate": 2e-05,
      "loss": 0.5061,
      "step": 9700
    },
    {
      "epoch": 0.8533263028385623,
      "grad_norm": 0.43281424045562744,
      "learning_rate": 2e-05,
      "loss": 0.5347,
      "step": 9710
    },
    {
      "epoch": 0.854205114684946,
      "grad_norm": 0.3808765411376953,
      "learning_rate": 2e-05,
      "loss": 0.5215,
      "step": 9720
    },
    {
      "epoch": 0.8550839265313297,
      "grad_norm": 0.4449610412120819,
      "learning_rate": 2e-05,
      "loss": 0.5147,
      "step": 9730
    },
    {
      "epoch": 0.8559627383777133,
      "grad_norm": 0.3478802442550659,
      "learning_rate": 2e-05,
      "loss": 0.4877,
      "step": 9740
    },
    {
      "epoch": 0.856841550224097,
      "grad_norm": 0.4621484577655792,
      "learning_rate": 2e-05,
      "loss": 0.5541,
      "step": 9750
    },
    {
      "epoch": 0.8577203620704807,
      "grad_norm": 0.39242249727249146,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 9760
    },
    {
      "epoch": 0.8585991739168644,
      "grad_norm": 0.39698585867881775,
      "learning_rate": 2e-05,
      "loss": 0.5186,
      "step": 9770
    },
    {
      "epoch": 0.8594779857632481,
      "grad_norm": 0.358394980430603,
      "learning_rate": 2e-05,
      "loss": 0.5542,
      "step": 9780
    },
    {
      "epoch": 0.8603567976096318,
      "grad_norm": 0.45696741342544556,
      "learning_rate": 2e-05,
      "loss": 0.4819,
      "step": 9790
    },
    {
      "epoch": 0.8612356094560155,
      "grad_norm": 0.38220515847206116,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 9800
    },
    {
      "epoch": 0.8621144213023991,
      "grad_norm": 0.37910938262939453,
      "learning_rate": 2e-05,
      "loss": 0.5212,
      "step": 9810
    },
    {
      "epoch": 0.8629932331487828,
      "grad_norm": 0.34992673993110657,
      "learning_rate": 2e-05,
      "loss": 0.5102,
      "step": 9820
    },
    {
      "epoch": 0.8638720449951666,
      "grad_norm": 0.34306803345680237,
      "learning_rate": 2e-05,
      "loss": 0.5089,
      "step": 9830
    },
    {
      "epoch": 0.8647508568415502,
      "grad_norm": 0.33596840500831604,
      "learning_rate": 2e-05,
      "loss": 0.4921,
      "step": 9840
    },
    {
      "epoch": 0.8656296686879339,
      "grad_norm": 0.41012999415397644,
      "learning_rate": 2e-05,
      "loss": 0.525,
      "step": 9850
    },
    {
      "epoch": 0.8665084805343176,
      "grad_norm": 0.4343957304954529,
      "learning_rate": 2e-05,
      "loss": 0.5106,
      "step": 9860
    },
    {
      "epoch": 0.8673872923807013,
      "grad_norm": 0.3645520508289337,
      "learning_rate": 2e-05,
      "loss": 0.5045,
      "step": 9870
    },
    {
      "epoch": 0.868266104227085,
      "grad_norm": 0.3806425929069519,
      "learning_rate": 2e-05,
      "loss": 0.5127,
      "step": 9880
    },
    {
      "epoch": 0.8691449160734687,
      "grad_norm": 0.3844059109687805,
      "learning_rate": 2e-05,
      "loss": 0.5307,
      "step": 9890
    },
    {
      "epoch": 0.8700237279198524,
      "grad_norm": 0.3582404553890228,
      "learning_rate": 2e-05,
      "loss": 0.5303,
      "step": 9900
    },
    {
      "epoch": 0.870902539766236,
      "grad_norm": 0.39040327072143555,
      "learning_rate": 2e-05,
      "loss": 0.5166,
      "step": 9910
    },
    {
      "epoch": 0.8717813516126197,
      "grad_norm": 0.33157432079315186,
      "learning_rate": 2e-05,
      "loss": 0.492,
      "step": 9920
    },
    {
      "epoch": 0.8726601634590034,
      "grad_norm": 0.3557230532169342,
      "learning_rate": 2e-05,
      "loss": 0.5284,
      "step": 9930
    },
    {
      "epoch": 0.8735389753053872,
      "grad_norm": 0.39875340461730957,
      "learning_rate": 2e-05,
      "loss": 0.5288,
      "step": 9940
    },
    {
      "epoch": 0.8744177871517708,
      "grad_norm": 0.3925156891345978,
      "learning_rate": 2e-05,
      "loss": 0.5319,
      "step": 9950
    },
    {
      "epoch": 0.8752965989981545,
      "grad_norm": 0.33891162276268005,
      "learning_rate": 2e-05,
      "loss": 0.5033,
      "step": 9960
    },
    {
      "epoch": 0.8761754108445382,
      "grad_norm": 0.402402400970459,
      "learning_rate": 2e-05,
      "loss": 0.5224,
      "step": 9970
    },
    {
      "epoch": 0.8770542226909219,
      "grad_norm": 0.3467569947242737,
      "learning_rate": 2e-05,
      "loss": 0.5143,
      "step": 9980
    },
    {
      "epoch": 0.8779330345373055,
      "grad_norm": 0.41573092341423035,
      "learning_rate": 2e-05,
      "loss": 0.5343,
      "step": 9990
    },
    {
      "epoch": 0.8788118463836893,
      "grad_norm": 0.3786810636520386,
      "learning_rate": 2e-05,
      "loss": 0.5138,
      "step": 10000
    },
    {
      "epoch": 0.879690658230073,
      "grad_norm": 0.35712116956710815,
      "learning_rate": 2e-05,
      "loss": 0.4941,
      "step": 10010
    },
    {
      "epoch": 0.8805694700764566,
      "grad_norm": 0.34305137395858765,
      "learning_rate": 2e-05,
      "loss": 0.4921,
      "step": 10020
    },
    {
      "epoch": 0.8814482819228403,
      "grad_norm": 0.5074466466903687,
      "learning_rate": 2e-05,
      "loss": 0.5421,
      "step": 10030
    },
    {
      "epoch": 0.882327093769224,
      "grad_norm": 0.3444838225841522,
      "learning_rate": 2e-05,
      "loss": 0.5184,
      "step": 10040
    },
    {
      "epoch": 0.8832059056156077,
      "grad_norm": 0.34227412939071655,
      "learning_rate": 2e-05,
      "loss": 0.5189,
      "step": 10050
    },
    {
      "epoch": 0.8840847174619914,
      "grad_norm": 0.34151414036750793,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 10060
    },
    {
      "epoch": 0.8849635293083751,
      "grad_norm": 0.3533826768398285,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 10070
    },
    {
      "epoch": 0.8858423411547588,
      "grad_norm": 0.35011520981788635,
      "learning_rate": 2e-05,
      "loss": 0.5134,
      "step": 10080
    },
    {
      "epoch": 0.8867211530011424,
      "grad_norm": 0.376728892326355,
      "learning_rate": 2e-05,
      "loss": 0.5322,
      "step": 10090
    },
    {
      "epoch": 0.8875999648475261,
      "grad_norm": 0.3572444021701813,
      "learning_rate": 2e-05,
      "loss": 0.529,
      "step": 10100
    },
    {
      "epoch": 0.8884787766939098,
      "grad_norm": 0.4010150134563446,
      "learning_rate": 2e-05,
      "loss": 0.5112,
      "step": 10110
    },
    {
      "epoch": 0.8893575885402936,
      "grad_norm": 0.39358046650886536,
      "learning_rate": 2e-05,
      "loss": 0.5199,
      "step": 10120
    },
    {
      "epoch": 0.8902364003866772,
      "grad_norm": 0.383069783449173,
      "learning_rate": 2e-05,
      "loss": 0.5124,
      "step": 10130
    },
    {
      "epoch": 0.8911152122330609,
      "grad_norm": 0.35765737295150757,
      "learning_rate": 2e-05,
      "loss": 0.5185,
      "step": 10140
    },
    {
      "epoch": 0.8919940240794446,
      "grad_norm": 0.3592289090156555,
      "learning_rate": 2e-05,
      "loss": 0.5062,
      "step": 10150
    },
    {
      "epoch": 0.8928728359258283,
      "grad_norm": 0.3950502276420593,
      "learning_rate": 2e-05,
      "loss": 0.5057,
      "step": 10160
    },
    {
      "epoch": 0.8937516477722119,
      "grad_norm": 0.3476558029651642,
      "learning_rate": 2e-05,
      "loss": 0.5085,
      "step": 10170
    },
    {
      "epoch": 0.8946304596185957,
      "grad_norm": 0.3434862494468689,
      "learning_rate": 1.999957274187931e-05,
      "loss": 0.532,
      "step": 10180
    },
    {
      "epoch": 0.8955092714649794,
      "grad_norm": 0.3749314546585083,
      "learning_rate": 1.9991978062682906e-05,
      "loss": 0.5335,
      "step": 10190
    },
    {
      "epoch": 0.896388083311363,
      "grad_norm": 0.3856378197669983,
      "learning_rate": 1.997489709987378e-05,
      "loss": 0.5436,
      "step": 10200
    },
    {
      "epoch": 0.8972668951577467,
      "grad_norm": 0.3595348000526428,
      "learning_rate": 1.9948346151502135e-05,
      "loss": 0.5198,
      "step": 10210
    },
    {
      "epoch": 0.8981457070041304,
      "grad_norm": 0.3432980179786682,
      "learning_rate": 1.991235055154325e-05,
      "loss": 0.5277,
      "step": 10220
    },
    {
      "epoch": 0.8990245188505142,
      "grad_norm": 0.370986670255661,
      "learning_rate": 1.9866944645724693e-05,
      "loss": 0.5211,
      "step": 10230
    },
    {
      "epoch": 0.8999033306968978,
      "grad_norm": 0.3482705354690552,
      "learning_rate": 1.9812171758754842e-05,
      "loss": 0.5254,
      "step": 10240
    },
    {
      "epoch": 0.9007821425432815,
      "grad_norm": 0.34930917620658875,
      "learning_rate": 1.9748084152983987e-05,
      "loss": 0.5275,
      "step": 10250
    },
    {
      "epoch": 0.9016609543896652,
      "grad_norm": 0.497697651386261,
      "learning_rate": 1.967474297853744e-05,
      "loss": 0.5254,
      "step": 10260
    },
    {
      "epoch": 0.9025397662360488,
      "grad_norm": 0.3974209725856781,
      "learning_rate": 1.9592218214968256e-05,
      "loss": 0.5023,
      "step": 10270
    },
    {
      "epoch": 0.9034185780824325,
      "grad_norm": 0.3622397482395172,
      "learning_rate": 1.9500588604485213e-05,
      "loss": 0.5134,
      "step": 10280
    },
    {
      "epoch": 0.9042973899288163,
      "grad_norm": 0.3650231957435608,
      "learning_rate": 1.939994157681979e-05,
      "loss": 0.5302,
      "step": 10290
    },
    {
      "epoch": 0.9051762017752,
      "grad_norm": 0.3823395371437073,
      "learning_rate": 1.9290373165803803e-05,
      "loss": 0.5276,
      "step": 10300
    },
    {
      "epoch": 0.9060550136215836,
      "grad_norm": 0.3724977672100067,
      "learning_rate": 1.9171987917737328e-05,
      "loss": 0.5399,
      "step": 10310
    },
    {
      "epoch": 0.9069338254679673,
      "grad_norm": 0.3680291473865509,
      "learning_rate": 1.9044898791634322e-05,
      "loss": 0.5162,
      "step": 10320
    },
    {
      "epoch": 0.907812637314351,
      "grad_norm": 0.35422542691230774,
      "learning_rate": 1.8909227051441126e-05,
      "loss": 0.5455,
      "step": 10330
    },
    {
      "epoch": 0.9086914491607346,
      "grad_norm": 0.3823006749153137,
      "learning_rate": 1.8765102150330692e-05,
      "loss": 0.5297,
      "step": 10340
    },
    {
      "epoch": 0.9095702610071184,
      "grad_norm": 0.3467950224876404,
      "learning_rate": 1.861266160718297e-05,
      "loss": 0.5089,
      "step": 10350
    },
    {
      "epoch": 0.9104490728535021,
      "grad_norm": 0.37045881152153015,
      "learning_rate": 1.845205087536923e-05,
      "loss": 0.5401,
      "step": 10360
    },
    {
      "epoch": 0.9113278846998858,
      "grad_norm": 0.35239335894584656,
      "learning_rate": 1.8283423203965633e-05,
      "loss": 0.4951,
      "step": 10370
    },
    {
      "epoch": 0.9122066965462694,
      "grad_norm": 0.40544581413269043,
      "learning_rate": 1.810693949152835e-05,
      "loss": 0.525,
      "step": 10380
    },
    {
      "epoch": 0.9130855083926531,
      "grad_norm": 0.34518319368362427,
      "learning_rate": 1.7922768132569902e-05,
      "loss": 0.5096,
      "step": 10390
    },
    {
      "epoch": 0.9139643202390368,
      "grad_norm": 0.33353886008262634,
      "learning_rate": 1.773108485688302e-05,
      "loss": 0.5156,
      "step": 10400
    },
    {
      "epoch": 0.9148431320854206,
      "grad_norm": 0.3997364044189453,
      "learning_rate": 1.7532072561865537e-05,
      "loss": 0.5056,
      "step": 10410
    },
    {
      "epoch": 0.9157219439318042,
      "grad_norm": 0.3702492117881775,
      "learning_rate": 1.732592113800616e-05,
      "loss": 0.4916,
      "step": 10420
    },
    {
      "epoch": 0.9166007557781879,
      "grad_norm": 0.3793681859970093,
      "learning_rate": 1.7112827287697668e-05,
      "loss": 0.5515,
      "step": 10430
    },
    {
      "epoch": 0.9174795676245716,
      "grad_norm": 0.3883363902568817,
      "learning_rate": 1.6892994337550452e-05,
      "loss": 0.5261,
      "step": 10440
    },
    {
      "epoch": 0.9183583794709552,
      "grad_norm": 0.3734261989593506,
      "learning_rate": 1.6666632044385487e-05,
      "loss": 0.5129,
      "step": 10450
    },
    {
      "epoch": 0.9192371913173389,
      "grad_norm": 0.3334291875362396,
      "learning_rate": 1.6433956395091757e-05,
      "loss": 0.5198,
      "step": 10460
    },
    {
      "epoch": 0.9201160031637227,
      "grad_norm": 0.34962722659111023,
      "learning_rate": 1.6195189400539238e-05,
      "loss": 0.5366,
      "step": 10470
    },
    {
      "epoch": 0.9209948150101064,
      "grad_norm": 0.362252801656723,
      "learning_rate": 1.595055888374398e-05,
      "loss": 0.5125,
      "step": 10480
    },
    {
      "epoch": 0.92187362685649,
      "grad_norm": 0.35282111167907715,
      "learning_rate": 1.5700298262487434e-05,
      "loss": 0.5579,
      "step": 10490
    },
    {
      "epoch": 0.9227524387028737,
      "grad_norm": 0.35117098689079285,
      "learning_rate": 1.5444646326597467e-05,
      "loss": 0.5082,
      "step": 10500
    },
    {
      "epoch": 0.9236312505492574,
      "grad_norm": 0.42528095841407776,
      "learning_rate": 1.5183847010103586e-05,
      "loss": 0.5131,
      "step": 10510
    },
    {
      "epoch": 0.924510062395641,
      "grad_norm": 0.3605803847312927,
      "learning_rate": 1.4918149158483686e-05,
      "loss": 0.5227,
      "step": 10520
    },
    {
      "epoch": 0.9253888742420248,
      "grad_norm": 0.3642340898513794,
      "learning_rate": 1.464780629122452e-05,
      "loss": 0.5016,
      "step": 10530
    },
    {
      "epoch": 0.9262676860884085,
      "grad_norm": 0.3786945044994354,
      "learning_rate": 1.4373076359922374e-05,
      "loss": 0.4987,
      "step": 10540
    },
    {
      "epoch": 0.9271464979347922,
      "grad_norm": 0.39350253343582153,
      "learning_rate": 1.4094221502154774e-05,
      "loss": 0.5186,
      "step": 10550
    },
    {
      "epoch": 0.9280253097811758,
      "grad_norm": 0.368342787027359,
      "learning_rate": 1.3811507791358113e-05,
      "loss": 0.5039,
      "step": 10560
    },
    {
      "epoch": 0.9289041216275595,
      "grad_norm": 0.35634204745292664,
      "learning_rate": 1.3525204982949753e-05,
      "loss": 0.4901,
      "step": 10570
    },
    {
      "epoch": 0.9297829334739433,
      "grad_norm": 0.3847208619117737,
      "learning_rate": 1.3235586256936996e-05,
      "loss": 0.5125,
      "step": 10580
    },
    {
      "epoch": 0.930661745320327,
      "grad_norm": 0.33358925580978394,
      "learning_rate": 1.294292795725836e-05,
      "loss": 0.5179,
      "step": 10590
    },
    {
      "epoch": 0.9315405571667106,
      "grad_norm": 0.3911134898662567,
      "learning_rate": 1.2647509328105978e-05,
      "loss": 0.5009,
      "step": 10600
    },
    {
      "epoch": 0.9324193690130943,
      "grad_norm": 0.3557060658931732,
      "learning_rate": 1.2349612247480686e-05,
      "loss": 0.5406,
      "step": 10610
    },
    {
      "epoch": 0.933298180859478,
      "grad_norm": 0.38635414838790894,
      "learning_rate": 1.2049520958234e-05,
      "loss": 0.55,
      "step": 10620
    },
    {
      "epoch": 0.9341769927058616,
      "grad_norm": 0.3531319797039032,
      "learning_rate": 1.1747521796853638e-05,
      "loss": 0.4919,
      "step": 10630
    },
    {
      "epoch": 0.9350558045522454,
      "grad_norm": 0.4289895296096802,
      "learning_rate": 1.1443902920251395e-05,
      "loss": 0.5358,
      "step": 10640
    },
    {
      "epoch": 0.9359346163986291,
      "grad_norm": 0.4065803587436676,
      "learning_rate": 1.1138954030814009e-05,
      "loss": 0.5249,
      "step": 10650
    },
    {
      "epoch": 0.9368134282450128,
      "grad_norm": 0.3788784146308899,
      "learning_rate": 1.0832966099979388e-05,
      "loss": 0.5078,
      "step": 10660
    },
    {
      "epoch": 0.9376922400913964,
      "grad_norm": 0.3955523669719696,
      "learning_rate": 1.0526231090601969e-05,
      "loss": 0.5243,
      "step": 10670
    },
    {
      "epoch": 0.9385710519377801,
      "grad_norm": 0.37966659665107727,
      "learning_rate": 1.0219041678372106e-05,
      "loss": 0.5159,
      "step": 10680
    },
    {
      "epoch": 0.9394498637841638,
      "grad_norm": 0.36736294627189636,
      "learning_rate": 9.911690972555268e-06,
      "loss": 0.5389,
      "step": 10690
    },
    {
      "epoch": 0.9403286756305475,
      "grad_norm": 0.3827562928199768,
      "learning_rate": 9.604472236317541e-06,
      "loss": 0.4962,
      "step": 10700
    },
    {
      "epoch": 0.9412074874769312,
      "grad_norm": 0.32662737369537354,
      "learning_rate": 9.297678606904335e-06,
      "loss": 0.5055,
      "step": 10710
    },
    {
      "epoch": 0.9420862993233149,
      "grad_norm": 0.3985399901866913,
      "learning_rate": 8.991602815939155e-06,
      "loss": 0.534,
      "step": 10720
    },
    {
      "epoch": 0.9429651111696986,
      "grad_norm": 0.3279280662536621,
      "learning_rate": 8.686536910109464e-06,
      "loss": 0.5167,
      "step": 10730
    },
    {
      "epoch": 0.9438439230160822,
      "grad_norm": 0.4124777317047119,
      "learning_rate": 8.382771972506084e-06,
      "loss": 0.5223,
      "step": 10740
    },
    {
      "epoch": 0.9447227348624659,
      "grad_norm": 0.37235507369041443,
      "learning_rate": 8.08059784488198e-06,
      "loss": 0.5051,
      "step": 10750
    },
    {
      "epoch": 0.9456015467088497,
      "grad_norm": 0.36804771423339844,
      "learning_rate": 7.780302851095544e-06,
      "loss": 0.5196,
      "step": 10760
    },
    {
      "epoch": 0.9464803585552333,
      "grad_norm": 0.3435978591442108,
      "learning_rate": 7.482173522002166e-06,
      "loss": 0.5202,
      "step": 10770
    },
    {
      "epoch": 0.947359170401617,
      "grad_norm": 0.33227425813674927,
      "learning_rate": 7.186494322056624e-06,
      "loss": 0.4983,
      "step": 10780
    },
    {
      "epoch": 0.9482379822480007,
      "grad_norm": 0.34496209025382996,
      "learning_rate": 6.893547377887183e-06,
      "loss": 0.4801,
      "step": 10790
    },
    {
      "epoch": 0.9491167940943844,
      "grad_norm": 0.453265517950058,
      "learning_rate": 6.6036122091003365e-06,
      "loss": 0.5255,
      "step": 10800
    },
    {
      "epoch": 0.949995605940768,
      "grad_norm": 0.37690088152885437,
      "learning_rate": 6.316965461573097e-06,
      "loss": 0.5129,
      "step": 10810
    },
    {
      "epoch": 0.9508744177871518,
      "grad_norm": 0.3494386374950409,
      "learning_rate": 6.0338806434872965e-06,
      "loss": 0.5201,
      "step": 10820
    },
    {
      "epoch": 0.9517532296335355,
      "grad_norm": 0.3726600110530853,
      "learning_rate": 5.754627864357749e-06,
      "loss": 0.4998,
      "step": 10830
    },
    {
      "epoch": 0.9526320414799192,
      "grad_norm": 0.38944002985954285,
      "learning_rate": 5.479473577303319e-06,
      "loss": 0.5004,
      "step": 10840
    },
    {
      "epoch": 0.9535108533263028,
      "grad_norm": 0.40071582794189453,
      "learning_rate": 5.208680324806778e-06,
      "loss": 0.5102,
      "step": 10850
    },
    {
      "epoch": 0.9543896651726865,
      "grad_norm": 0.34419965744018555,
      "learning_rate": 4.94250648820606e-06,
      "loss": 0.4972,
      "step": 10860
    },
    {
      "epoch": 0.9552684770190703,
      "grad_norm": 0.3507113456726074,
      "learning_rate": 4.681206041155951e-06,
      "loss": 0.5133,
      "step": 10870
    },
    {
      "epoch": 0.9561472888654539,
      "grad_norm": 0.3372703194618225,
      "learning_rate": 4.42502830729538e-06,
      "loss": 0.5236,
      "step": 10880
    },
    {
      "epoch": 0.9570261007118376,
      "grad_norm": 0.3428637683391571,
      "learning_rate": 4.174217722351669e-06,
      "loss": 0.479,
      "step": 10890
    },
    {
      "epoch": 0.9579049125582213,
      "grad_norm": 0.4205319881439209,
      "learning_rate": 3.929013600908575e-06,
      "loss": 0.499,
      "step": 10900
    },
    {
      "epoch": 0.958783724404605,
      "grad_norm": 0.3800400495529175,
      "learning_rate": 3.689649908060807e-06,
      "loss": 0.5212,
      "step": 10910
    },
    {
      "epoch": 0.9596625362509886,
      "grad_norm": 0.34339600801467896,
      "learning_rate": 3.4563550361728028e-06,
      "loss": 0.517,
      "step": 10920
    },
    {
      "epoch": 0.9605413480973723,
      "grad_norm": 0.3964458107948303,
      "learning_rate": 3.2293515869548216e-06,
      "loss": 0.535,
      "step": 10930
    },
    {
      "epoch": 0.9614201599437561,
      "grad_norm": 0.412043958902359,
      "learning_rate": 3.0088561590642746e-06,
      "loss": 0.5297,
      "step": 10940
    },
    {
      "epoch": 0.9622989717901397,
      "grad_norm": 0.3887500464916229,
      "learning_rate": 2.795079141434941e-06,
      "loss": 0.5318,
      "step": 10950
    },
    {
      "epoch": 0.9631777836365234,
      "grad_norm": 0.3815765082836151,
      "learning_rate": 2.588224512531299e-06,
      "loss": 0.4617,
      "step": 10960
    },
    {
      "epoch": 0.9640565954829071,
      "grad_norm": 0.41199082136154175,
      "learning_rate": 2.388489645719495e-06,
      "loss": 0.5041,
      "step": 10970
    },
    {
      "epoch": 0.9649354073292908,
      "grad_norm": 0.34162992238998413,
      "learning_rate": 2.196065120940663e-06,
      "loss": 0.5148,
      "step": 10980
    },
    {
      "epoch": 0.9658142191756744,
      "grad_norm": 0.35256466269493103,
      "learning_rate": 2.0111345428662808e-06,
      "loss": 0.5035,
      "step": 10990
    },
    {
      "epoch": 0.9666930310220582,
      "grad_norm": 0.3714001774787903,
      "learning_rate": 1.8338743657091102e-06,
      "loss": 0.5083,
      "step": 11000
    },
    {
      "epoch": 0.9675718428684419,
      "grad_norm": 0.3553823232650757,
      "learning_rate": 1.6644537248568113e-06,
      "loss": 0.5241,
      "step": 11010
    },
    {
      "epoch": 0.9684506547148256,
      "grad_norm": 0.4539598822593689,
      "learning_rate": 1.503034275488951e-06,
      "loss": 0.5256,
      "step": 11020
    },
    {
      "epoch": 0.9693294665612092,
      "grad_norm": 0.3575528562068939,
      "learning_rate": 1.3497700383313495e-06,
      "loss": 0.5029,
      "step": 11030
    },
    {
      "epoch": 0.9702082784075929,
      "grad_norm": 0.37538033723831177,
      "learning_rate": 1.2048072526949562e-06,
      "loss": 0.523,
      "step": 11040
    },
    {
      "epoch": 0.9710870902539767,
      "grad_norm": 0.39639803767204285,
      "learning_rate": 1.0682842369394756e-06,
      "loss": 0.5129,
      "step": 11050
    },
    {
      "epoch": 0.9719659021003603,
      "grad_norm": 0.37807974219322205,
      "learning_rate": 9.403312564948792e-07,
      "loss": 0.5162,
      "step": 11060
    },
    {
      "epoch": 0.972844713946744,
      "grad_norm": 0.3329496383666992,
      "learning_rate": 8.210703995667471e-07,
      "loss": 0.4884,
      "step": 11070
    },
    {
      "epoch": 0.9737235257931277,
      "grad_norm": 0.3586650490760803,
      "learning_rate": 7.106154606440245e-07,
      "loss": 0.5087,
      "step": 11080
    },
    {
      "epoch": 0.9746023376395114,
      "grad_norm": 0.40165334939956665,
      "learning_rate": 6.090718319203393e-07,
      "loss": 0.5343,
      "step": 11090
    },
    {
      "epoch": 0.975481149485895,
      "grad_norm": 0.3793430030345917,
      "learning_rate": 5.165364027325089e-07,
      "loss": 0.5177,
      "step": 11100
    },
    {
      "epoch": 0.9763599613322788,
      "grad_norm": 0.3327946662902832,
      "learning_rate": 4.330974671121585e-07,
      "loss": 0.5074,
      "step": 11110
    },
    {
      "epoch": 0.9772387731786625,
      "grad_norm": 0.30681028962135315,
      "learning_rate": 3.588346395386846e-07,
      "loss": 0.4959,
      "step": 11120
    },
    {
      "epoch": 0.9781175850250461,
      "grad_norm": 0.34503671526908875,
      "learning_rate": 2.938187789739395e-07,
      "loss": 0.5335,
      "step": 11130
    },
    {
      "epoch": 0.9789963968714298,
      "grad_norm": 0.3517617881298065,
      "learning_rate": 2.3811192125110724e-07,
      "loss": 0.4869,
      "step": 11140
    },
    {
      "epoch": 0.9798752087178135,
      "grad_norm": 0.3626396954059601,
      "learning_rate": 1.917672198823177e-07,
      "loss": 0.5032,
      "step": 11150
    },
    {
      "epoch": 0.9807540205641972,
      "grad_norm": 0.3735167682170868,
      "learning_rate": 1.5482889534143795e-07,
      "loss": 0.5058,
      "step": 11160
    },
    {
      "epoch": 0.9816328324105809,
      "grad_norm": 0.36062005162239075,
      "learning_rate": 1.273321928704694e-07,
      "loss": 0.4883,
      "step": 11170
    },
    {
      "epoch": 0.9825116442569646,
      "grad_norm": 0.4451676905155182,
      "learning_rate": 1.0930334884977866e-07,
      "loss": 0.529,
      "step": 11180
    },
    {
      "epoch": 0.9833904561033483,
      "grad_norm": 0.3877953886985779,
      "learning_rate": 1.0075956576427487e-07,
      "loss": 0.5133,
      "step": 11190
    },
    {
      "epoch": 0.9836540996572634,
      "step": 11193,
      "total_flos": 4.9965855215861105e+19,
      "train_loss": 0.04686111353958813,
      "train_runtime": 11462.5506,
      "train_samples_per_second": 11.718,
      "train_steps_per_second": 0.976
    }
  ],
  "logging_steps": 10,
  "max_steps": 11193,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.9965855215861105e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}