{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4918709904209509,
  "eval_steps": 500,
  "global_step": 5597,
  "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": 1.9999810479633703e-05,
      "loss": 0.5082,
      "step": 5090
    },
    {
      "epoch": 0.4481940416556815,
      "grad_norm": 0.3651531934738159,
      "learning_rate": 1.997707677020786e-05,
      "loss": 0.5472,
      "step": 5100
    },
    {
      "epoch": 0.4490728535020652,
      "grad_norm": 0.31714680790901184,
      "learning_rate": 1.9916538196124637e-05,
      "loss": 0.5246,
      "step": 5110
    },
    {
      "epoch": 0.4499516653484489,
      "grad_norm": 0.4181217551231384,
      "learning_rate": 1.9818425303863402e-05,
      "loss": 0.5312,
      "step": 5120
    },
    {
      "epoch": 0.4508304771948326,
      "grad_norm": 0.33540576696395874,
      "learning_rate": 1.9683111732582887e-05,
      "loss": 0.5482,
      "step": 5130
    },
    {
      "epoch": 0.45170928904121627,
      "grad_norm": 0.36196160316467285,
      "learning_rate": 1.951111279120706e-05,
      "loss": 0.5285,
      "step": 5140
    },
    {
      "epoch": 0.4525881008876,
      "grad_norm": 0.3670157790184021,
      "learning_rate": 1.9303083495996028e-05,
      "loss": 0.5497,
      "step": 5150
    },
    {
      "epoch": 0.45346691273398365,
      "grad_norm": 0.33631399273872375,
      "learning_rate": 1.9059816076075482e-05,
      "loss": 0.5532,
      "step": 5160
    },
    {
      "epoch": 0.4543457245803673,
      "grad_norm": 0.8250991106033325,
      "learning_rate": 1.8782236956424142e-05,
      "loss": 0.5559,
      "step": 5170
    },
    {
      "epoch": 0.45522453642675104,
      "grad_norm": 0.3488682806491852,
      "learning_rate": 1.8471403229808897e-05,
      "loss": 0.5287,
      "step": 5180
    },
    {
      "epoch": 0.4561033482731347,
      "grad_norm": 0.32487279176712036,
      "learning_rate": 1.8128498631103333e-05,
      "loss": 0.5641,
      "step": 5190
    },
    {
      "epoch": 0.4569821601195184,
      "grad_norm": 0.3495788276195526,
      "learning_rate": 1.7754829029320507e-05,
      "loss": 0.5421,
      "step": 5200
    },
    {
      "epoch": 0.4578609719659021,
      "grad_norm": 0.35946157574653625,
      "learning_rate": 1.7351817454527473e-05,
      "loss": 0.5065,
      "step": 5210
    },
    {
      "epoch": 0.4587397838122858,
      "grad_norm": 0.34108319878578186,
      "learning_rate": 1.6920998678580345e-05,
      "loss": 0.539,
      "step": 5220
    },
    {
      "epoch": 0.45961859565866947,
      "grad_norm": 0.34420546889305115,
      "learning_rate": 1.6464013370317854e-05,
      "loss": 0.5473,
      "step": 5230
    },
    {
      "epoch": 0.4604974075050532,
      "grad_norm": 0.3907243609428406,
      "learning_rate": 1.5982601847471915e-05,
      "loss": 0.5334,
      "step": 5240
    },
    {
      "epoch": 0.46137621935143686,
      "grad_norm": 0.4602360427379608,
      "learning_rate": 1.5478597449089553e-05,
      "loss": 0.5575,
      "step": 5250
    },
    {
      "epoch": 0.4622550311978205,
      "grad_norm": 0.5124063491821289,
      "learning_rate": 1.4953919553705725e-05,
      "loss": 0.5637,
      "step": 5260
    },
    {
      "epoch": 0.46313384304420424,
      "grad_norm": 0.3482801616191864,
      "learning_rate": 1.441056626985559e-05,
      "loss": 0.5529,
      "step": 5270
    },
    {
      "epoch": 0.4640126548905879,
      "grad_norm": 0.41699984669685364,
      "learning_rate": 1.3850606826762668e-05,
      "loss": 0.5645,
      "step": 5280
    },
    {
      "epoch": 0.4648914667369716,
      "grad_norm": 0.6701170206069946,
      "learning_rate": 1.3276173694181128e-05,
      "loss": 0.5369,
      "step": 5290
    },
    {
      "epoch": 0.4657702785833553,
      "grad_norm": 0.34473496675491333,
      "learning_rate": 1.2689454461401822e-05,
      "loss": 0.5277,
      "step": 5300
    },
    {
      "epoch": 0.466649090429739,
      "grad_norm": 0.32014894485473633,
      "learning_rate": 1.2092683506348939e-05,
      "loss": 0.5287,
      "step": 5310
    },
    {
      "epoch": 0.4675279022761227,
      "grad_norm": 0.349581778049469,
      "learning_rate": 1.14881334864935e-05,
      "loss": 0.5195,
      "step": 5320
    },
    {
      "epoch": 0.4684067141225064,
      "grad_norm": 0.3643857538700104,
      "learning_rate": 1.0878106683988481e-05,
      "loss": 0.5353,
      "step": 5330
    },
    {
      "epoch": 0.46928552596889006,
      "grad_norm": 0.35531625151634216,
      "learning_rate": 1.026492623798551e-05,
      "loss": 0.5464,
      "step": 5340
    },
    {
      "epoch": 0.4701643378152737,
      "grad_norm": 0.9461858868598938,
      "learning_rate": 9.650927297522743e-06,
      "loss": 0.553,
      "step": 5350
    },
    {
      "epoch": 0.47104314966165745,
      "grad_norm": 0.3360968232154846,
      "learning_rate": 9.038448128676012e-06,
      "loss": 0.5383,
      "step": 5360
    },
    {
      "epoch": 0.4719219615080411,
      "grad_norm": 0.3314628005027771,
      "learning_rate": 8.42982120983949e-06,
      "loss": 0.5321,
      "step": 5370
    },
    {
      "epoch": 0.47280077335442483,
      "grad_norm": 0.3270135521888733,
      "learning_rate": 7.827364349047303e-06,
      "loss": 0.5135,
      "step": 5380
    },
    {
      "epoch": 0.4736795852008085,
      "grad_norm": 0.3804806172847748,
      "learning_rate": 7.233371857163706e-06,
      "loss": 0.5462,
      "step": 5390
    },
    {
      "epoch": 0.4745583970471922,
      "grad_norm": 0.339309960603714,
      "learning_rate": 6.650105810556492e-06,
      "loss": 0.5554,
      "step": 5400
    },
    {
      "epoch": 0.4754372088935759,
      "grad_norm": 0.34906429052352905,
      "learning_rate": 6.079787436527681e-06,
      "loss": 0.5377,
      "step": 5410
    },
    {
      "epoch": 0.4763160207399596,
      "grad_norm": 0.3404780924320221,
      "learning_rate": 5.524588654307961e-06,
      "loss": 0.5456,
      "step": 5420
    },
    {
      "epoch": 0.47719483258634326,
      "grad_norm": 0.34517258405685425,
      "learning_rate": 4.986623803828985e-06,
      "loss": 0.541,
      "step": 5430
    },
    {
      "epoch": 0.47807364443272693,
      "grad_norm": 0.37149953842163086,
      "learning_rate": 4.467941593772423e-06,
      "loss": 0.5547,
      "step": 5440
    },
    {
      "epoch": 0.47895245627911065,
      "grad_norm": 0.44154414534568787,
      "learning_rate": 3.9705172995597606e-06,
      "loss": 0.5309,
      "step": 5450
    },
    {
      "epoch": 0.4798312681254943,
      "grad_norm": 0.3964569568634033,
      "learning_rate": 3.4962452409948346e-06,
      "loss": 0.5202,
      "step": 5460
    },
    {
      "epoch": 0.48071007997187803,
      "grad_norm": 0.3427949547767639,
      "learning_rate": 3.046931568206211e-06,
      "loss": 0.5236,
      "step": 5470
    },
    {
      "epoch": 0.4815888918182617,
      "grad_norm": 0.3578217327594757,
      "learning_rate": 2.624287383362404e-06,
      "loss": 0.5586,
      "step": 5480
    },
    {
      "epoch": 0.4824677036646454,
      "grad_norm": 0.3597162663936615,
      "learning_rate": 2.2299222243541845e-06,
      "loss": 0.5403,
      "step": 5490
    },
    {
      "epoch": 0.4833465155110291,
      "grad_norm": 0.3687967360019684,
      "learning_rate": 1.8653379352598246e-06,
      "loss": 0.526,
      "step": 5500
    },
    {
      "epoch": 0.4842253273574128,
      "grad_norm": 0.3392982482910156,
      "learning_rate": 1.5319229469361075e-06,
      "loss": 0.5205,
      "step": 5510
    },
    {
      "epoch": 0.48510413920379647,
      "grad_norm": 0.36678963899612427,
      "learning_rate": 1.2309469895160917e-06,
      "loss": 0.5236,
      "step": 5520
    },
    {
      "epoch": 0.48598295105018013,
      "grad_norm": 0.3331840932369232,
      "learning_rate": 9.635562569497974e-07,
      "loss": 0.5394,
      "step": 5530
    },
    {
      "epoch": 0.48686176289656385,
      "grad_norm": 0.35520151257514954,
      "learning_rate": 7.307690420024867e-07,
      "loss": 0.5227,
      "step": 5540
    },
    {
      "epoch": 0.4877405747429475,
      "grad_norm": 0.3363773822784424,
      "learning_rate": 5.334718583335942e-07,
      "loss": 0.5205,
      "step": 5550
    },
    {
      "epoch": 0.48861938658933124,
      "grad_norm": 0.38362252712249756,
      "learning_rate": 3.7241606442442763e-07,
      "loss": 0.5071,
      "step": 5560
    },
    {
      "epoch": 0.4894981984357149,
      "grad_norm": 0.3763801157474518,
      "learning_rate": 2.482150022115879e-07,
      "loss": 0.5554,
      "step": 5570
    },
    {
      "epoch": 0.4903770102820986,
      "grad_norm": 0.31886377930641174,
      "learning_rate": 1.6134166132293174e-07,
      "loss": 0.5646,
      "step": 5580
    },
    {
      "epoch": 0.4912558221284823,
      "grad_norm": 0.32389166951179504,
      "learning_rate": 1.1212687781127357e-07,
      "loss": 0.5438,
      "step": 5590
    },
    {
      "epoch": 0.4918709904209509,
      "step": 5597,
      "total_flos": 2.4985159616803373e+19,
      "train_loss": 0.04896302386780221,
      "train_runtime": 5739.7338,
      "train_samples_per_second": 11.702,
      "train_steps_per_second": 0.975
    }
  ],
  "logging_steps": 10,
  "max_steps": 5597,
  "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": 2.4985159616803373e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}